{ "best_global_step": null, "best_metric": null, "best_model_checkpoint": null, "epoch": 1.0, "eval_steps": 500, "global_step": 60000, "is_hyper_param_search": false, "is_local_process_zero": true, "is_world_process_zero": true, "log_history": [ { "grad_norm": 0.4963434934616089, "learning_rate": 3.0000000000000004e-07, "loss": 1.1852, "step": 10 }, { "grad_norm": 0.5362529158592224, "learning_rate": 6.333333333333333e-07, "loss": 1.191, "step": 20 }, { "grad_norm": 0.5185095071792603, "learning_rate": 9.666666666666668e-07, "loss": 1.1875, "step": 30 }, { "grad_norm": 0.5237129330635071, "learning_rate": 1.3e-06, "loss": 1.1809, "step": 40 }, { "grad_norm": 0.3315047025680542, "learning_rate": 1.6333333333333333e-06, "loss": 1.1496, "step": 50 }, { "grad_norm": 0.4051373600959778, "learning_rate": 1.9666666666666668e-06, "loss": 1.1254, "step": 60 }, { "grad_norm": 0.28214961290359497, "learning_rate": 2.3e-06, "loss": 1.1258, "step": 70 }, { "grad_norm": 0.28722986578941345, "learning_rate": 2.6333333333333337e-06, "loss": 1.1285, "step": 80 }, { "grad_norm": 0.40627187490463257, "learning_rate": 2.966666666666667e-06, "loss": 1.102, "step": 90 }, { "grad_norm": 0.36213696002960205, "learning_rate": 3.3e-06, "loss": 1.0813, "step": 100 }, { "grad_norm": 0.693615198135376, "learning_rate": 3.633333333333334e-06, "loss": 1.0766, "step": 110 }, { "grad_norm": 0.3118543028831482, "learning_rate": 3.966666666666667e-06, "loss": 1.0766, "step": 120 }, { "grad_norm": 0.26597896218299866, "learning_rate": 4.2999999999999995e-06, "loss": 1.0664, "step": 130 }, { "grad_norm": 0.8201600313186646, "learning_rate": 4.633333333333334e-06, "loss": 1.0637, "step": 140 }, { "grad_norm": 0.421932190656662, "learning_rate": 4.966666666666667e-06, "loss": 1.059, "step": 150 }, { "grad_norm": 0.5572344064712524, "learning_rate": 5.3e-06, "loss": 1.0543, "step": 160 }, { "grad_norm": 0.6793390512466431, "learning_rate": 5.633333333333333e-06, "loss": 1.0566, "step": 170 }, { "grad_norm": 0.7040989398956299, "learning_rate": 5.9666666666666666e-06, "loss": 1.0391, "step": 180 }, { "grad_norm": 0.5722596645355225, "learning_rate": 6.300000000000001e-06, "loss": 1.023, "step": 190 }, { "grad_norm": 0.8514906167984009, "learning_rate": 6.633333333333333e-06, "loss": 0.9844, "step": 200 }, { "grad_norm": 1.5842680931091309, "learning_rate": 6.966666666666667e-06, "loss": 0.9338, "step": 210 }, { "grad_norm": 1.1300703287124634, "learning_rate": 7.2999999999999996e-06, "loss": 0.8832, "step": 220 }, { "grad_norm": 0.9567537903785706, "learning_rate": 7.633333333333334e-06, "loss": 0.8342, "step": 230 }, { "grad_norm": 1.3872158527374268, "learning_rate": 7.966666666666666e-06, "loss": 0.7941, "step": 240 }, { "grad_norm": 1.3385834693908691, "learning_rate": 8.3e-06, "loss": 0.732, "step": 250 }, { "grad_norm": 1.5443499088287354, "learning_rate": 8.633333333333334e-06, "loss": 0.6844, "step": 260 }, { "grad_norm": 1.637166142463684, "learning_rate": 8.966666666666668e-06, "loss": 0.6297, "step": 270 }, { "grad_norm": 1.6781947612762451, "learning_rate": 9.3e-06, "loss": 0.5842, "step": 280 }, { "grad_norm": 1.577732801437378, "learning_rate": 9.633333333333335e-06, "loss": 0.5311, "step": 290 }, { "grad_norm": 1.4956679344177246, "learning_rate": 9.966666666666667e-06, "loss": 0.4924, "step": 300 }, { "grad_norm": 1.534165620803833, "learning_rate": 1.03e-05, "loss": 0.4595, "step": 310 }, { "grad_norm": 2.0054097175598145, "learning_rate": 1.0633333333333334e-05, "loss": 0.4091, "step": 320 }, { "grad_norm": 2.390655755996704, "learning_rate": 1.0966666666666666e-05, "loss": 0.3768, "step": 330 }, { "grad_norm": 1.7814544439315796, "learning_rate": 1.13e-05, "loss": 0.3529, "step": 340 }, { "grad_norm": 2.2316360473632812, "learning_rate": 1.1633333333333334e-05, "loss": 0.318, "step": 350 }, { "grad_norm": 1.839164137840271, "learning_rate": 1.1966666666666668e-05, "loss": 0.294, "step": 360 }, { "grad_norm": 1.4398057460784912, "learning_rate": 1.23e-05, "loss": 0.2568, "step": 370 }, { "grad_norm": 1.8941164016723633, "learning_rate": 1.2633333333333333e-05, "loss": 0.2419, "step": 380 }, { "grad_norm": 1.580346941947937, "learning_rate": 1.2966666666666669e-05, "loss": 0.2313, "step": 390 }, { "grad_norm": 1.739145040512085, "learning_rate": 1.3300000000000001e-05, "loss": 0.2246, "step": 400 }, { "grad_norm": 1.8966467380523682, "learning_rate": 1.3633333333333334e-05, "loss": 0.2226, "step": 410 }, { "grad_norm": 1.600353717803955, "learning_rate": 1.3966666666666666e-05, "loss": 0.2061, "step": 420 }, { "grad_norm": 2.2946321964263916, "learning_rate": 1.43e-05, "loss": 0.1956, "step": 430 }, { "grad_norm": 1.7680151462554932, "learning_rate": 1.4633333333333334e-05, "loss": 0.1661, "step": 440 }, { "grad_norm": 1.8046908378601074, "learning_rate": 1.4966666666666668e-05, "loss": 0.1616, "step": 450 }, { "grad_norm": 1.9970968961715698, "learning_rate": 1.53e-05, "loss": 0.1517, "step": 460 }, { "grad_norm": 1.9537441730499268, "learning_rate": 1.563333333333333e-05, "loss": 0.1426, "step": 470 }, { "grad_norm": 1.796105980873108, "learning_rate": 1.5966666666666667e-05, "loss": 0.1438, "step": 480 }, { "grad_norm": 2.432992935180664, "learning_rate": 1.63e-05, "loss": 0.1406, "step": 490 }, { "grad_norm": 1.744086742401123, "learning_rate": 1.6633333333333336e-05, "loss": 0.1399, "step": 500 }, { "grad_norm": 1.9467130899429321, "learning_rate": 1.6966666666666668e-05, "loss": 0.1281, "step": 510 }, { "grad_norm": 1.7084479331970215, "learning_rate": 1.73e-05, "loss": 0.1315, "step": 520 }, { "grad_norm": 2.1384193897247314, "learning_rate": 1.7633333333333336e-05, "loss": 0.1305, "step": 530 }, { "grad_norm": 1.63255774974823, "learning_rate": 1.796666666666667e-05, "loss": 0.128, "step": 540 }, { "grad_norm": 2.028160333633423, "learning_rate": 1.83e-05, "loss": 0.1122, "step": 550 }, { "grad_norm": 1.3919050693511963, "learning_rate": 1.8633333333333333e-05, "loss": 0.1254, "step": 560 }, { "grad_norm": 1.6799205541610718, "learning_rate": 1.896666666666667e-05, "loss": 0.1283, "step": 570 }, { "grad_norm": 2.0520362854003906, "learning_rate": 1.93e-05, "loss": 0.1227, "step": 580 }, { "grad_norm": 1.6150975227355957, "learning_rate": 1.9633333333333334e-05, "loss": 0.117, "step": 590 }, { "grad_norm": 1.59844970703125, "learning_rate": 1.9966666666666666e-05, "loss": 0.1177, "step": 600 }, { "grad_norm": 1.66582453250885, "learning_rate": 2.0300000000000002e-05, "loss": 0.1102, "step": 610 }, { "grad_norm": 1.7572534084320068, "learning_rate": 2.0633333333333335e-05, "loss": 0.1116, "step": 620 }, { "grad_norm": 1.3142567873001099, "learning_rate": 2.0966666666666667e-05, "loss": 0.1128, "step": 630 }, { "grad_norm": 2.120042324066162, "learning_rate": 2.13e-05, "loss": 0.1149, "step": 640 }, { "grad_norm": 1.4696060419082642, "learning_rate": 2.1633333333333332e-05, "loss": 0.1113, "step": 650 }, { "grad_norm": 2.0135626792907715, "learning_rate": 2.1966666666666668e-05, "loss": 0.1095, "step": 660 }, { "grad_norm": 1.8148534297943115, "learning_rate": 2.23e-05, "loss": 0.1024, "step": 670 }, { "grad_norm": 1.617483139038086, "learning_rate": 2.2633333333333336e-05, "loss": 0.1028, "step": 680 }, { "grad_norm": 1.5676223039627075, "learning_rate": 2.2966666666666668e-05, "loss": 0.0944, "step": 690 }, { "grad_norm": 1.9585235118865967, "learning_rate": 2.3300000000000004e-05, "loss": 0.0998, "step": 700 }, { "grad_norm": 1.6630173921585083, "learning_rate": 2.3633333333333336e-05, "loss": 0.0978, "step": 710 }, { "grad_norm": 1.5976413488388062, "learning_rate": 2.396666666666667e-05, "loss": 0.1005, "step": 720 }, { "grad_norm": 1.2446086406707764, "learning_rate": 2.43e-05, "loss": 0.1009, "step": 730 }, { "grad_norm": 1.3185960054397583, "learning_rate": 2.4633333333333334e-05, "loss": 0.1025, "step": 740 }, { "grad_norm": 1.559364676475525, "learning_rate": 2.496666666666667e-05, "loss": 0.1023, "step": 750 }, { "grad_norm": 1.7643721103668213, "learning_rate": 2.5300000000000002e-05, "loss": 0.0961, "step": 760 }, { "grad_norm": 1.574176549911499, "learning_rate": 2.5633333333333338e-05, "loss": 0.1046, "step": 770 }, { "grad_norm": 1.7484185695648193, "learning_rate": 2.5966666666666667e-05, "loss": 0.0984, "step": 780 }, { "grad_norm": 1.8233050107955933, "learning_rate": 2.6300000000000002e-05, "loss": 0.0904, "step": 790 }, { "grad_norm": 1.6157609224319458, "learning_rate": 2.663333333333333e-05, "loss": 0.0951, "step": 800 }, { "grad_norm": 1.644906997680664, "learning_rate": 2.6966666666666667e-05, "loss": 0.0881, "step": 810 }, { "grad_norm": 1.5373362302780151, "learning_rate": 2.7300000000000003e-05, "loss": 0.0901, "step": 820 }, { "grad_norm": 0.9688888192176819, "learning_rate": 2.7633333333333332e-05, "loss": 0.0907, "step": 830 }, { "grad_norm": 1.6994704008102417, "learning_rate": 2.7966666666666668e-05, "loss": 0.092, "step": 840 }, { "grad_norm": 1.7064887285232544, "learning_rate": 2.83e-05, "loss": 0.0936, "step": 850 }, { "grad_norm": 1.608164668083191, "learning_rate": 2.8633333333333336e-05, "loss": 0.0928, "step": 860 }, { "grad_norm": 1.383835792541504, "learning_rate": 2.8966666666666668e-05, "loss": 0.0911, "step": 870 }, { "grad_norm": 1.505488634109497, "learning_rate": 2.93e-05, "loss": 0.0926, "step": 880 }, { "grad_norm": 1.7351610660552979, "learning_rate": 2.9633333333333336e-05, "loss": 0.0869, "step": 890 }, { "grad_norm": 1.476628303527832, "learning_rate": 2.9966666666666672e-05, "loss": 0.0928, "step": 900 }, { "grad_norm": 1.2343003749847412, "learning_rate": 3.03e-05, "loss": 0.0835, "step": 910 }, { "grad_norm": 1.4476467370986938, "learning_rate": 3.063333333333334e-05, "loss": 0.0843, "step": 920 }, { "grad_norm": 1.4148728847503662, "learning_rate": 3.096666666666666e-05, "loss": 0.0884, "step": 930 }, { "grad_norm": 1.0850536823272705, "learning_rate": 3.13e-05, "loss": 0.0813, "step": 940 }, { "grad_norm": 1.400771975517273, "learning_rate": 3.1633333333333334e-05, "loss": 0.0864, "step": 950 }, { "grad_norm": 1.9069130420684814, "learning_rate": 3.196666666666667e-05, "loss": 0.0827, "step": 960 }, { "grad_norm": 1.4011884927749634, "learning_rate": 3.2300000000000006e-05, "loss": 0.0844, "step": 970 }, { "grad_norm": 1.4135830402374268, "learning_rate": 3.263333333333333e-05, "loss": 0.0802, "step": 980 }, { "grad_norm": 1.1292967796325684, "learning_rate": 3.296666666666667e-05, "loss": 0.0902, "step": 990 }, { "grad_norm": 1.3411247730255127, "learning_rate": 3.33e-05, "loss": 0.077, "step": 1000 }, { "grad_norm": 1.1278730630874634, "learning_rate": 3.3633333333333335e-05, "loss": 0.0767, "step": 1010 }, { "grad_norm": 1.4784291982650757, "learning_rate": 3.396666666666667e-05, "loss": 0.0887, "step": 1020 }, { "grad_norm": 1.6353739500045776, "learning_rate": 3.430000000000001e-05, "loss": 0.0777, "step": 1030 }, { "grad_norm": 1.3922699689865112, "learning_rate": 3.463333333333333e-05, "loss": 0.0737, "step": 1040 }, { "grad_norm": 1.1614177227020264, "learning_rate": 3.496666666666667e-05, "loss": 0.0749, "step": 1050 }, { "grad_norm": 1.1485652923583984, "learning_rate": 3.53e-05, "loss": 0.0728, "step": 1060 }, { "grad_norm": 1.1936993598937988, "learning_rate": 3.563333333333334e-05, "loss": 0.0783, "step": 1070 }, { "grad_norm": 1.160801887512207, "learning_rate": 3.596666666666667e-05, "loss": 0.0811, "step": 1080 }, { "grad_norm": 1.5741806030273438, "learning_rate": 3.63e-05, "loss": 0.0775, "step": 1090 }, { "grad_norm": 1.000142216682434, "learning_rate": 3.6633333333333334e-05, "loss": 0.0739, "step": 1100 }, { "grad_norm": 1.4253404140472412, "learning_rate": 3.6966666666666666e-05, "loss": 0.0732, "step": 1110 }, { "grad_norm": 1.1558648347854614, "learning_rate": 3.73e-05, "loss": 0.0755, "step": 1120 }, { "grad_norm": 1.3830760717391968, "learning_rate": 3.763333333333334e-05, "loss": 0.0789, "step": 1130 }, { "grad_norm": 1.2591768503189087, "learning_rate": 3.796666666666667e-05, "loss": 0.0746, "step": 1140 }, { "grad_norm": 1.252992868423462, "learning_rate": 3.83e-05, "loss": 0.0698, "step": 1150 }, { "grad_norm": 1.1146215200424194, "learning_rate": 3.8633333333333335e-05, "loss": 0.0753, "step": 1160 }, { "grad_norm": 1.210828423500061, "learning_rate": 3.896666666666667e-05, "loss": 0.0762, "step": 1170 }, { "grad_norm": 1.0407379865646362, "learning_rate": 3.9300000000000007e-05, "loss": 0.0723, "step": 1180 }, { "grad_norm": 1.085291862487793, "learning_rate": 3.963333333333333e-05, "loss": 0.0691, "step": 1190 }, { "grad_norm": 1.0675197839736938, "learning_rate": 3.996666666666667e-05, "loss": 0.0733, "step": 1200 }, { "grad_norm": 1.234538197517395, "learning_rate": 4.0300000000000004e-05, "loss": 0.0706, "step": 1210 }, { "grad_norm": 1.1045076847076416, "learning_rate": 4.0633333333333336e-05, "loss": 0.0703, "step": 1220 }, { "grad_norm": 0.8566645383834839, "learning_rate": 4.096666666666667e-05, "loss": 0.0727, "step": 1230 }, { "grad_norm": 1.1666266918182373, "learning_rate": 4.13e-05, "loss": 0.0691, "step": 1240 }, { "grad_norm": 1.1437296867370605, "learning_rate": 4.1633333333333333e-05, "loss": 0.0705, "step": 1250 }, { "grad_norm": 1.1349562406539917, "learning_rate": 4.196666666666667e-05, "loss": 0.0706, "step": 1260 }, { "grad_norm": 1.197243571281433, "learning_rate": 4.23e-05, "loss": 0.0666, "step": 1270 }, { "grad_norm": 1.2493517398834229, "learning_rate": 4.263333333333334e-05, "loss": 0.0742, "step": 1280 }, { "grad_norm": 1.0570238828659058, "learning_rate": 4.296666666666666e-05, "loss": 0.0704, "step": 1290 }, { "grad_norm": 1.050927758216858, "learning_rate": 4.33e-05, "loss": 0.0709, "step": 1300 }, { "grad_norm": 1.0574309825897217, "learning_rate": 4.3633333333333335e-05, "loss": 0.0685, "step": 1310 }, { "grad_norm": 0.8983197808265686, "learning_rate": 4.396666666666667e-05, "loss": 0.0643, "step": 1320 }, { "grad_norm": 1.207806944847107, "learning_rate": 4.43e-05, "loss": 0.0771, "step": 1330 }, { "grad_norm": 0.9312703609466553, "learning_rate": 4.463333333333334e-05, "loss": 0.0731, "step": 1340 }, { "grad_norm": 1.2369251251220703, "learning_rate": 4.496666666666667e-05, "loss": 0.0698, "step": 1350 }, { "grad_norm": 0.8722283840179443, "learning_rate": 4.53e-05, "loss": 0.0709, "step": 1360 }, { "grad_norm": 0.9897818565368652, "learning_rate": 4.5633333333333336e-05, "loss": 0.0711, "step": 1370 }, { "grad_norm": 1.1966822147369385, "learning_rate": 4.596666666666667e-05, "loss": 0.0631, "step": 1380 }, { "grad_norm": 1.0266767740249634, "learning_rate": 4.630000000000001e-05, "loss": 0.0641, "step": 1390 }, { "grad_norm": 0.7966419458389282, "learning_rate": 4.663333333333333e-05, "loss": 0.0637, "step": 1400 }, { "grad_norm": 0.9530993700027466, "learning_rate": 4.696666666666667e-05, "loss": 0.0677, "step": 1410 }, { "grad_norm": 0.9805204272270203, "learning_rate": 4.73e-05, "loss": 0.0657, "step": 1420 }, { "grad_norm": 0.8988294005393982, "learning_rate": 4.763333333333334e-05, "loss": 0.0679, "step": 1430 }, { "grad_norm": 0.9517422318458557, "learning_rate": 4.796666666666667e-05, "loss": 0.0629, "step": 1440 }, { "grad_norm": 1.0493371486663818, "learning_rate": 4.83e-05, "loss": 0.064, "step": 1450 }, { "grad_norm": 0.9309093952178955, "learning_rate": 4.8633333333333334e-05, "loss": 0.0704, "step": 1460 }, { "grad_norm": 1.0646703243255615, "learning_rate": 4.8966666666666667e-05, "loss": 0.0628, "step": 1470 }, { "grad_norm": 0.9575940370559692, "learning_rate": 4.93e-05, "loss": 0.0664, "step": 1480 }, { "grad_norm": 0.8289860486984253, "learning_rate": 4.963333333333334e-05, "loss": 0.0629, "step": 1490 }, { "grad_norm": 0.9999306201934814, "learning_rate": 4.996666666666667e-05, "loss": 0.0636, "step": 1500 }, { "grad_norm": 0.6606645584106445, "learning_rate": 5.03e-05, "loss": 0.0689, "step": 1510 }, { "grad_norm": 0.9491574168205261, "learning_rate": 5.0633333333333335e-05, "loss": 0.0622, "step": 1520 }, { "grad_norm": 1.1991971731185913, "learning_rate": 5.0966666666666674e-05, "loss": 0.0663, "step": 1530 }, { "grad_norm": 0.9637433290481567, "learning_rate": 5.130000000000001e-05, "loss": 0.0656, "step": 1540 }, { "grad_norm": 0.8015892505645752, "learning_rate": 5.163333333333333e-05, "loss": 0.0641, "step": 1550 }, { "grad_norm": 0.9023792147636414, "learning_rate": 5.196666666666667e-05, "loss": 0.0626, "step": 1560 }, { "grad_norm": 0.7096395492553711, "learning_rate": 5.2300000000000004e-05, "loss": 0.063, "step": 1570 }, { "grad_norm": 0.8274873495101929, "learning_rate": 5.2633333333333336e-05, "loss": 0.0617, "step": 1580 }, { "grad_norm": 0.8809630274772644, "learning_rate": 5.296666666666666e-05, "loss": 0.0643, "step": 1590 }, { "grad_norm": 0.8721491694450378, "learning_rate": 5.330000000000001e-05, "loss": 0.0679, "step": 1600 }, { "grad_norm": 0.7451794147491455, "learning_rate": 5.3633333333333334e-05, "loss": 0.0608, "step": 1610 }, { "grad_norm": 1.0013816356658936, "learning_rate": 5.3966666666666666e-05, "loss": 0.0565, "step": 1620 }, { "grad_norm": 0.9440953135490417, "learning_rate": 5.4300000000000005e-05, "loss": 0.0638, "step": 1630 }, { "grad_norm": 0.7555800080299377, "learning_rate": 5.463333333333334e-05, "loss": 0.0596, "step": 1640 }, { "grad_norm": 0.8619276285171509, "learning_rate": 5.496666666666666e-05, "loss": 0.0604, "step": 1650 }, { "grad_norm": 0.7645530104637146, "learning_rate": 5.530000000000001e-05, "loss": 0.0582, "step": 1660 }, { "grad_norm": 0.8340891599655151, "learning_rate": 5.5633333333333335e-05, "loss": 0.059, "step": 1670 }, { "grad_norm": 0.8476892709732056, "learning_rate": 5.596666666666667e-05, "loss": 0.0623, "step": 1680 }, { "grad_norm": 0.9234489798545837, "learning_rate": 5.63e-05, "loss": 0.0688, "step": 1690 }, { "grad_norm": 0.8480938673019409, "learning_rate": 5.663333333333334e-05, "loss": 0.0675, "step": 1700 }, { "grad_norm": 0.8579698801040649, "learning_rate": 5.696666666666667e-05, "loss": 0.0592, "step": 1710 }, { "grad_norm": 0.8718623518943787, "learning_rate": 5.73e-05, "loss": 0.06, "step": 1720 }, { "grad_norm": 1.0157506465911865, "learning_rate": 5.7633333333333336e-05, "loss": 0.06, "step": 1730 }, { "grad_norm": 0.8724209070205688, "learning_rate": 5.796666666666667e-05, "loss": 0.0582, "step": 1740 }, { "grad_norm": 0.9149057865142822, "learning_rate": 5.83e-05, "loss": 0.0641, "step": 1750 }, { "grad_norm": 0.8602994084358215, "learning_rate": 5.863333333333334e-05, "loss": 0.062, "step": 1760 }, { "grad_norm": 0.8449149131774902, "learning_rate": 5.896666666666667e-05, "loss": 0.0618, "step": 1770 }, { "grad_norm": 0.7762284278869629, "learning_rate": 5.93e-05, "loss": 0.0602, "step": 1780 }, { "grad_norm": 0.6767029762268066, "learning_rate": 5.9633333333333344e-05, "loss": 0.0609, "step": 1790 }, { "grad_norm": 0.6424612402915955, "learning_rate": 5.996666666666667e-05, "loss": 0.0596, "step": 1800 }, { "grad_norm": 0.942735493183136, "learning_rate": 6.03e-05, "loss": 0.0558, "step": 1810 }, { "grad_norm": 0.9018301367759705, "learning_rate": 6.063333333333333e-05, "loss": 0.0598, "step": 1820 }, { "grad_norm": 0.8495177030563354, "learning_rate": 6.0966666666666674e-05, "loss": 0.0576, "step": 1830 }, { "grad_norm": 0.7344351410865784, "learning_rate": 6.13e-05, "loss": 0.0591, "step": 1840 }, { "grad_norm": 1.0043690204620361, "learning_rate": 6.163333333333333e-05, "loss": 0.0579, "step": 1850 }, { "grad_norm": 0.7147799730300903, "learning_rate": 6.196666666666668e-05, "loss": 0.059, "step": 1860 }, { "grad_norm": 0.8345941305160522, "learning_rate": 6.23e-05, "loss": 0.0577, "step": 1870 }, { "grad_norm": 0.654281497001648, "learning_rate": 6.263333333333333e-05, "loss": 0.0582, "step": 1880 }, { "grad_norm": 0.7362072467803955, "learning_rate": 6.296666666666667e-05, "loss": 0.0578, "step": 1890 }, { "grad_norm": 0.7964036464691162, "learning_rate": 6.330000000000001e-05, "loss": 0.057, "step": 1900 }, { "grad_norm": 0.7486265897750854, "learning_rate": 6.363333333333334e-05, "loss": 0.0574, "step": 1910 }, { "grad_norm": 0.6489279270172119, "learning_rate": 6.396666666666667e-05, "loss": 0.0583, "step": 1920 }, { "grad_norm": 0.7810956239700317, "learning_rate": 6.43e-05, "loss": 0.0589, "step": 1930 }, { "grad_norm": 0.7831915616989136, "learning_rate": 6.463333333333334e-05, "loss": 0.057, "step": 1940 }, { "grad_norm": 0.6389955878257751, "learning_rate": 6.496666666666667e-05, "loss": 0.0564, "step": 1950 }, { "grad_norm": 0.7881708145141602, "learning_rate": 6.53e-05, "loss": 0.053, "step": 1960 }, { "grad_norm": 0.6984956860542297, "learning_rate": 6.563333333333333e-05, "loss": 0.0555, "step": 1970 }, { "grad_norm": 0.5442981123924255, "learning_rate": 6.596666666666667e-05, "loss": 0.0539, "step": 1980 }, { "grad_norm": 0.974873423576355, "learning_rate": 6.630000000000001e-05, "loss": 0.0554, "step": 1990 }, { "grad_norm": 0.6696648597717285, "learning_rate": 6.663333333333333e-05, "loss": 0.0545, "step": 2000 }, { "grad_norm": 0.7061467170715332, "learning_rate": 6.696666666666666e-05, "loss": 0.0518, "step": 2010 }, { "grad_norm": 0.7191210985183716, "learning_rate": 6.730000000000001e-05, "loss": 0.0562, "step": 2020 }, { "grad_norm": 0.831903338432312, "learning_rate": 6.763333333333334e-05, "loss": 0.0566, "step": 2030 }, { "grad_norm": 0.6326937675476074, "learning_rate": 6.796666666666666e-05, "loss": 0.0548, "step": 2040 }, { "grad_norm": 0.8980246782302856, "learning_rate": 6.83e-05, "loss": 0.0581, "step": 2050 }, { "grad_norm": 0.8011890053749084, "learning_rate": 6.863333333333334e-05, "loss": 0.0581, "step": 2060 }, { "grad_norm": 0.7703070640563965, "learning_rate": 6.896666666666667e-05, "loss": 0.0574, "step": 2070 }, { "grad_norm": 0.5740545392036438, "learning_rate": 6.93e-05, "loss": 0.0527, "step": 2080 }, { "grad_norm": 0.6876325607299805, "learning_rate": 6.963333333333334e-05, "loss": 0.0578, "step": 2090 }, { "grad_norm": 0.6046603322029114, "learning_rate": 6.996666666666667e-05, "loss": 0.0551, "step": 2100 }, { "grad_norm": 0.591177761554718, "learning_rate": 7.03e-05, "loss": 0.0499, "step": 2110 }, { "grad_norm": 0.7132079601287842, "learning_rate": 7.063333333333333e-05, "loss": 0.0538, "step": 2120 }, { "grad_norm": 0.49045032262802124, "learning_rate": 7.096666666666667e-05, "loss": 0.0559, "step": 2130 }, { "grad_norm": 0.6506680846214294, "learning_rate": 7.13e-05, "loss": 0.0553, "step": 2140 }, { "grad_norm": 0.715941309928894, "learning_rate": 7.163333333333334e-05, "loss": 0.0533, "step": 2150 }, { "grad_norm": 0.7913379073143005, "learning_rate": 7.196666666666668e-05, "loss": 0.0501, "step": 2160 }, { "grad_norm": 0.6303556561470032, "learning_rate": 7.23e-05, "loss": 0.0502, "step": 2170 }, { "grad_norm": 0.9289034605026245, "learning_rate": 7.263333333333334e-05, "loss": 0.0519, "step": 2180 }, { "grad_norm": 0.6582747101783752, "learning_rate": 7.296666666666667e-05, "loss": 0.0498, "step": 2190 }, { "grad_norm": 0.592159628868103, "learning_rate": 7.33e-05, "loss": 0.0593, "step": 2200 }, { "grad_norm": 0.7080442905426025, "learning_rate": 7.363333333333334e-05, "loss": 0.0539, "step": 2210 }, { "grad_norm": 0.6412515044212341, "learning_rate": 7.396666666666667e-05, "loss": 0.0578, "step": 2220 }, { "grad_norm": 0.7321165800094604, "learning_rate": 7.43e-05, "loss": 0.0526, "step": 2230 }, { "grad_norm": 0.6292983889579773, "learning_rate": 7.463333333333334e-05, "loss": 0.0508, "step": 2240 }, { "grad_norm": 0.5690622329711914, "learning_rate": 7.496666666666667e-05, "loss": 0.054, "step": 2250 }, { "grad_norm": 0.535159170627594, "learning_rate": 7.53e-05, "loss": 0.0559, "step": 2260 }, { "grad_norm": 0.5076817274093628, "learning_rate": 7.563333333333333e-05, "loss": 0.0532, "step": 2270 }, { "grad_norm": 0.627003014087677, "learning_rate": 7.596666666666668e-05, "loss": 0.0497, "step": 2280 }, { "grad_norm": 0.6225648522377014, "learning_rate": 7.630000000000001e-05, "loss": 0.0506, "step": 2290 }, { "grad_norm": 0.716613233089447, "learning_rate": 7.663333333333333e-05, "loss": 0.0505, "step": 2300 }, { "grad_norm": 0.633381187915802, "learning_rate": 7.696666666666668e-05, "loss": 0.0541, "step": 2310 }, { "grad_norm": 0.6307882070541382, "learning_rate": 7.730000000000001e-05, "loss": 0.0531, "step": 2320 }, { "grad_norm": 0.6148698925971985, "learning_rate": 7.763333333333334e-05, "loss": 0.0569, "step": 2330 }, { "grad_norm": 0.5976276397705078, "learning_rate": 7.796666666666666e-05, "loss": 0.0482, "step": 2340 }, { "grad_norm": 0.4870288074016571, "learning_rate": 7.83e-05, "loss": 0.0547, "step": 2350 }, { "grad_norm": 0.6811262369155884, "learning_rate": 7.863333333333334e-05, "loss": 0.05, "step": 2360 }, { "grad_norm": 0.617865264415741, "learning_rate": 7.896666666666667e-05, "loss": 0.0503, "step": 2370 }, { "grad_norm": 0.5618493556976318, "learning_rate": 7.93e-05, "loss": 0.0504, "step": 2380 }, { "grad_norm": 0.8332135677337646, "learning_rate": 7.963333333333334e-05, "loss": 0.0525, "step": 2390 }, { "grad_norm": 0.7149859070777893, "learning_rate": 7.996666666666667e-05, "loss": 0.051, "step": 2400 }, { "grad_norm": 0.6374902129173279, "learning_rate": 8.030000000000001e-05, "loss": 0.0481, "step": 2410 }, { "grad_norm": 0.603898286819458, "learning_rate": 8.063333333333333e-05, "loss": 0.056, "step": 2420 }, { "grad_norm": 0.6916307210922241, "learning_rate": 8.096666666666667e-05, "loss": 0.0552, "step": 2430 }, { "grad_norm": 0.5881582498550415, "learning_rate": 8.13e-05, "loss": 0.0529, "step": 2440 }, { "grad_norm": 0.621459424495697, "learning_rate": 8.163333333333334e-05, "loss": 0.0523, "step": 2450 }, { "grad_norm": 0.609537661075592, "learning_rate": 8.196666666666668e-05, "loss": 0.0462, "step": 2460 }, { "grad_norm": 0.5870870351791382, "learning_rate": 8.23e-05, "loss": 0.0517, "step": 2470 }, { "grad_norm": 0.46851658821105957, "learning_rate": 8.263333333333334e-05, "loss": 0.0495, "step": 2480 }, { "grad_norm": 0.6007828116416931, "learning_rate": 8.296666666666667e-05, "loss": 0.0568, "step": 2490 }, { "grad_norm": 0.5860142707824707, "learning_rate": 8.33e-05, "loss": 0.0492, "step": 2500 }, { "grad_norm": 0.5511006712913513, "learning_rate": 8.363333333333334e-05, "loss": 0.0532, "step": 2510 }, { "grad_norm": 0.7166287302970886, "learning_rate": 8.396666666666667e-05, "loss": 0.051, "step": 2520 }, { "grad_norm": 0.5528836250305176, "learning_rate": 8.43e-05, "loss": 0.0503, "step": 2530 }, { "grad_norm": 0.5070132613182068, "learning_rate": 8.463333333333335e-05, "loss": 0.0458, "step": 2540 }, { "grad_norm": 0.5816045999526978, "learning_rate": 8.496666666666667e-05, "loss": 0.0513, "step": 2550 }, { "grad_norm": 0.5596718192100525, "learning_rate": 8.53e-05, "loss": 0.0561, "step": 2560 }, { "grad_norm": 0.496566504240036, "learning_rate": 8.563333333333333e-05, "loss": 0.051, "step": 2570 }, { "grad_norm": 0.7508627772331238, "learning_rate": 8.596666666666668e-05, "loss": 0.0489, "step": 2580 }, { "grad_norm": 0.6426692008972168, "learning_rate": 8.63e-05, "loss": 0.0454, "step": 2590 }, { "grad_norm": 0.6039537787437439, "learning_rate": 8.663333333333333e-05, "loss": 0.0524, "step": 2600 }, { "grad_norm": 0.5783808827400208, "learning_rate": 8.696666666666668e-05, "loss": 0.0507, "step": 2610 }, { "grad_norm": 0.5825155973434448, "learning_rate": 8.730000000000001e-05, "loss": 0.0475, "step": 2620 }, { "grad_norm": 0.5732088685035706, "learning_rate": 8.763333333333334e-05, "loss": 0.0494, "step": 2630 }, { "grad_norm": 0.568000316619873, "learning_rate": 8.796666666666667e-05, "loss": 0.0509, "step": 2640 }, { "grad_norm": 0.536655068397522, "learning_rate": 8.83e-05, "loss": 0.049, "step": 2650 }, { "grad_norm": 0.5456947684288025, "learning_rate": 8.863333333333334e-05, "loss": 0.048, "step": 2660 }, { "grad_norm": 0.6681229472160339, "learning_rate": 8.896666666666667e-05, "loss": 0.0474, "step": 2670 }, { "grad_norm": 0.46725228428840637, "learning_rate": 8.93e-05, "loss": 0.049, "step": 2680 }, { "grad_norm": 0.5373956561088562, "learning_rate": 8.963333333333333e-05, "loss": 0.051, "step": 2690 }, { "grad_norm": 0.6221674680709839, "learning_rate": 8.996666666666667e-05, "loss": 0.0476, "step": 2700 }, { "grad_norm": 0.44114041328430176, "learning_rate": 9.030000000000001e-05, "loss": 0.048, "step": 2710 }, { "grad_norm": 0.6916643977165222, "learning_rate": 9.063333333333333e-05, "loss": 0.0448, "step": 2720 }, { "grad_norm": 0.5402345061302185, "learning_rate": 9.096666666666666e-05, "loss": 0.0482, "step": 2730 }, { "grad_norm": 0.5246310234069824, "learning_rate": 9.130000000000001e-05, "loss": 0.0456, "step": 2740 }, { "grad_norm": 0.4353185296058655, "learning_rate": 9.163333333333334e-05, "loss": 0.0435, "step": 2750 }, { "grad_norm": 0.46436789631843567, "learning_rate": 9.196666666666666e-05, "loss": 0.0453, "step": 2760 }, { "grad_norm": 0.58781898021698, "learning_rate": 9.230000000000001e-05, "loss": 0.0511, "step": 2770 }, { "grad_norm": 0.47681376338005066, "learning_rate": 9.263333333333334e-05, "loss": 0.0471, "step": 2780 }, { "grad_norm": 0.49050113558769226, "learning_rate": 9.296666666666667e-05, "loss": 0.0447, "step": 2790 }, { "grad_norm": 0.4632444679737091, "learning_rate": 9.33e-05, "loss": 0.0482, "step": 2800 }, { "grad_norm": 0.33257052302360535, "learning_rate": 9.363333333333334e-05, "loss": 0.0495, "step": 2810 }, { "grad_norm": 0.5827874541282654, "learning_rate": 9.396666666666667e-05, "loss": 0.0503, "step": 2820 }, { "grad_norm": 0.6805890798568726, "learning_rate": 9.43e-05, "loss": 0.0476, "step": 2830 }, { "grad_norm": 0.5234531164169312, "learning_rate": 9.463333333333333e-05, "loss": 0.0478, "step": 2840 }, { "grad_norm": 0.5681066513061523, "learning_rate": 9.496666666666667e-05, "loss": 0.0483, "step": 2850 }, { "grad_norm": 0.5555163025856018, "learning_rate": 9.53e-05, "loss": 0.0517, "step": 2860 }, { "grad_norm": 0.5606744289398193, "learning_rate": 9.563333333333334e-05, "loss": 0.05, "step": 2870 }, { "grad_norm": 0.5716975331306458, "learning_rate": 9.596666666666668e-05, "loss": 0.047, "step": 2880 }, { "grad_norm": 0.6978026628494263, "learning_rate": 9.63e-05, "loss": 0.0506, "step": 2890 }, { "grad_norm": 0.5656298995018005, "learning_rate": 9.663333333333334e-05, "loss": 0.0476, "step": 2900 }, { "grad_norm": 0.5590419769287109, "learning_rate": 9.696666666666667e-05, "loss": 0.0479, "step": 2910 }, { "grad_norm": 0.4916408658027649, "learning_rate": 9.730000000000001e-05, "loss": 0.0496, "step": 2920 }, { "grad_norm": 0.41652727127075195, "learning_rate": 9.763333333333334e-05, "loss": 0.0512, "step": 2930 }, { "grad_norm": 0.4865264594554901, "learning_rate": 9.796666666666667e-05, "loss": 0.0525, "step": 2940 }, { "grad_norm": 0.4580366313457489, "learning_rate": 9.83e-05, "loss": 0.049, "step": 2950 }, { "grad_norm": 0.4808882474899292, "learning_rate": 9.863333333333334e-05, "loss": 0.049, "step": 2960 }, { "grad_norm": 0.553287148475647, "learning_rate": 9.896666666666667e-05, "loss": 0.0484, "step": 2970 }, { "grad_norm": 0.5048560500144958, "learning_rate": 9.93e-05, "loss": 0.0454, "step": 2980 }, { "grad_norm": 0.4227713346481323, "learning_rate": 9.963333333333333e-05, "loss": 0.046, "step": 2990 }, { "grad_norm": 0.5952905416488647, "learning_rate": 9.996666666666668e-05, "loss": 0.0448, "step": 3000 }, { "grad_norm": 0.5548657774925232, "learning_rate": 9.999999384858465e-05, "loss": 0.0462, "step": 3010 }, { "grad_norm": 0.6064696311950684, "learning_rate": 9.999997258443473e-05, "loss": 0.0465, "step": 3020 }, { "grad_norm": 0.5892049670219421, "learning_rate": 9.999993613161331e-05, "loss": 0.0459, "step": 3030 }, { "grad_norm": 0.5532440543174744, "learning_rate": 9.999988449013146e-05, "loss": 0.0481, "step": 3040 }, { "grad_norm": 0.565274178981781, "learning_rate": 9.99998176600049e-05, "loss": 0.0461, "step": 3050 }, { "grad_norm": 0.583250880241394, "learning_rate": 9.999973564125389e-05, "loss": 0.0473, "step": 3060 }, { "grad_norm": 0.470424085855484, "learning_rate": 9.999963843390335e-05, "loss": 0.0447, "step": 3070 }, { "grad_norm": 0.4779054522514343, "learning_rate": 9.999952603798282e-05, "loss": 0.0457, "step": 3080 }, { "grad_norm": 0.48773813247680664, "learning_rate": 9.999939845352646e-05, "loss": 0.0471, "step": 3090 }, { "grad_norm": 0.40086629986763, "learning_rate": 9.999925568057298e-05, "loss": 0.0461, "step": 3100 }, { "grad_norm": 0.5165130496025085, "learning_rate": 9.999909771916578e-05, "loss": 0.0444, "step": 3110 }, { "grad_norm": 0.453973650932312, "learning_rate": 9.999892456935285e-05, "loss": 0.0445, "step": 3120 }, { "grad_norm": 0.5748994946479797, "learning_rate": 9.999873623118679e-05, "loss": 0.0463, "step": 3130 }, { "grad_norm": 0.5203547477722168, "learning_rate": 9.999853270472479e-05, "loss": 0.046, "step": 3140 }, { "grad_norm": 0.38782161474227905, "learning_rate": 9.999831399002871e-05, "loss": 0.0454, "step": 3150 }, { "grad_norm": 0.5043109059333801, "learning_rate": 9.999808008716494e-05, "loss": 0.0431, "step": 3160 }, { "grad_norm": 0.45340457558631897, "learning_rate": 9.999783099620459e-05, "loss": 0.0419, "step": 3170 }, { "grad_norm": 0.39430150389671326, "learning_rate": 9.999756671722328e-05, "loss": 0.0417, "step": 3180 }, { "grad_norm": 0.5276562571525574, "learning_rate": 9.99972872503013e-05, "loss": 0.0441, "step": 3190 }, { "grad_norm": 0.49642807245254517, "learning_rate": 9.999699259552359e-05, "loss": 0.0456, "step": 3200 }, { "grad_norm": 0.38196271657943726, "learning_rate": 9.99966827529796e-05, "loss": 0.0449, "step": 3210 }, { "grad_norm": 0.605189859867096, "learning_rate": 9.999635772276348e-05, "loss": 0.0424, "step": 3220 }, { "grad_norm": 0.48224836587905884, "learning_rate": 9.999601750497396e-05, "loss": 0.0417, "step": 3230 }, { "grad_norm": 0.47880685329437256, "learning_rate": 9.99956620997144e-05, "loss": 0.0464, "step": 3240 }, { "grad_norm": 0.47011637687683105, "learning_rate": 9.999529150709275e-05, "loss": 0.0426, "step": 3250 }, { "grad_norm": 0.5222616791725159, "learning_rate": 9.999490572722158e-05, "loss": 0.0471, "step": 3260 }, { "grad_norm": 0.36061999201774597, "learning_rate": 9.99945047602181e-05, "loss": 0.0437, "step": 3270 }, { "grad_norm": 0.42162564396858215, "learning_rate": 9.99940886062041e-05, "loss": 0.0428, "step": 3280 }, { "grad_norm": 0.526419997215271, "learning_rate": 9.999365726530599e-05, "loss": 0.0513, "step": 3290 }, { "grad_norm": 0.46384626626968384, "learning_rate": 9.999321073765481e-05, "loss": 0.0445, "step": 3300 }, { "grad_norm": 0.41664591431617737, "learning_rate": 9.99927490233862e-05, "loss": 0.0435, "step": 3310 }, { "grad_norm": 0.49868083000183105, "learning_rate": 9.999227212264043e-05, "loss": 0.0461, "step": 3320 }, { "grad_norm": 0.494498074054718, "learning_rate": 9.999178003556236e-05, "loss": 0.0436, "step": 3330 }, { "grad_norm": 0.39971697330474854, "learning_rate": 9.999127276230146e-05, "loss": 0.0446, "step": 3340 }, { "grad_norm": 0.4619086682796478, "learning_rate": 9.999075030301184e-05, "loss": 0.0435, "step": 3350 }, { "grad_norm": 0.3608328700065613, "learning_rate": 9.999021265785221e-05, "loss": 0.0426, "step": 3360 }, { "grad_norm": 0.4919787347316742, "learning_rate": 9.998965982698589e-05, "loss": 0.0459, "step": 3370 }, { "grad_norm": 0.456455260515213, "learning_rate": 9.998909181058082e-05, "loss": 0.0429, "step": 3380 }, { "grad_norm": 0.4113308787345886, "learning_rate": 9.998850860880953e-05, "loss": 0.0418, "step": 3390 }, { "grad_norm": 0.4666517376899719, "learning_rate": 9.998791022184922e-05, "loss": 0.0467, "step": 3400 }, { "grad_norm": 0.42678144574165344, "learning_rate": 9.99872966498816e-05, "loss": 0.0401, "step": 3410 }, { "grad_norm": 0.444285124540329, "learning_rate": 9.998666789309313e-05, "loss": 0.0404, "step": 3420 }, { "grad_norm": 0.4688093066215515, "learning_rate": 9.998602395167475e-05, "loss": 0.0415, "step": 3430 }, { "grad_norm": 0.45351341366767883, "learning_rate": 9.998536482582213e-05, "loss": 0.0467, "step": 3440 }, { "grad_norm": 0.5043035745620728, "learning_rate": 9.998469051573544e-05, "loss": 0.0409, "step": 3450 }, { "grad_norm": 0.3743304908275604, "learning_rate": 9.998400102161954e-05, "loss": 0.0459, "step": 3460 }, { "grad_norm": 0.3941773474216461, "learning_rate": 9.998329634368388e-05, "loss": 0.039, "step": 3470 }, { "grad_norm": 0.482013076543808, "learning_rate": 9.998257648214253e-05, "loss": 0.0446, "step": 3480 }, { "grad_norm": 0.3901718854904175, "learning_rate": 9.998184143721417e-05, "loss": 0.0415, "step": 3490 }, { "grad_norm": 0.38475993275642395, "learning_rate": 9.998109120912206e-05, "loss": 0.0424, "step": 3500 }, { "grad_norm": 0.2945215404033661, "learning_rate": 9.998032579809411e-05, "loss": 0.0378, "step": 3510 }, { "grad_norm": 0.4998665452003479, "learning_rate": 9.997954520436286e-05, "loss": 0.0425, "step": 3520 }, { "grad_norm": 0.503100574016571, "learning_rate": 9.997874942816538e-05, "loss": 0.0444, "step": 3530 }, { "grad_norm": 0.39165961742401123, "learning_rate": 9.997793846974345e-05, "loss": 0.0404, "step": 3540 }, { "grad_norm": 0.24502801895141602, "learning_rate": 9.997711232934341e-05, "loss": 0.0376, "step": 3550 }, { "grad_norm": 0.4482492506504059, "learning_rate": 9.99762710072162e-05, "loss": 0.0401, "step": 3560 }, { "grad_norm": 0.46007633209228516, "learning_rate": 9.997541450361743e-05, "loss": 0.0401, "step": 3570 }, { "grad_norm": 0.5245420932769775, "learning_rate": 9.997454281880723e-05, "loss": 0.0461, "step": 3580 }, { "grad_norm": 0.4401470124721527, "learning_rate": 9.997365595305044e-05, "loss": 0.0418, "step": 3590 }, { "grad_norm": 0.4964672923088074, "learning_rate": 9.997275390661644e-05, "loss": 0.0444, "step": 3600 }, { "grad_norm": 0.4256664216518402, "learning_rate": 9.997183667977926e-05, "loss": 0.0388, "step": 3610 }, { "grad_norm": 0.299241304397583, "learning_rate": 9.997090427281752e-05, "loss": 0.0379, "step": 3620 }, { "grad_norm": 0.4567224085330963, "learning_rate": 9.996995668601448e-05, "loss": 0.0391, "step": 3630 }, { "grad_norm": 0.41271957755088806, "learning_rate": 9.996899391965798e-05, "loss": 0.0391, "step": 3640 }, { "grad_norm": 0.4606080949306488, "learning_rate": 9.996801597404048e-05, "loss": 0.0404, "step": 3650 }, { "grad_norm": 0.41207239031791687, "learning_rate": 9.996702284945905e-05, "loss": 0.0414, "step": 3660 }, { "grad_norm": 0.43558233976364136, "learning_rate": 9.996601454621539e-05, "loss": 0.0427, "step": 3670 }, { "grad_norm": 0.35699373483657837, "learning_rate": 9.996499106461577e-05, "loss": 0.0413, "step": 3680 }, { "grad_norm": 0.30618250370025635, "learning_rate": 9.996395240497112e-05, "loss": 0.0389, "step": 3690 }, { "grad_norm": 0.38777586817741394, "learning_rate": 9.996289856759696e-05, "loss": 0.0383, "step": 3700 }, { "grad_norm": 0.4662986099720001, "learning_rate": 9.996182955281342e-05, "loss": 0.038, "step": 3710 }, { "grad_norm": 0.4156440794467926, "learning_rate": 9.996074536094519e-05, "loss": 0.0405, "step": 3720 }, { "grad_norm": 0.4254329204559326, "learning_rate": 9.995964599232168e-05, "loss": 0.0384, "step": 3730 }, { "grad_norm": 0.3496548533439636, "learning_rate": 9.995853144727683e-05, "loss": 0.0384, "step": 3740 }, { "grad_norm": 0.30544644594192505, "learning_rate": 9.99574017261492e-05, "loss": 0.0386, "step": 3750 }, { "grad_norm": 0.399122029542923, "learning_rate": 9.995625682928198e-05, "loss": 0.0366, "step": 3760 }, { "grad_norm": 0.44310951232910156, "learning_rate": 9.995509675702295e-05, "loss": 0.037, "step": 3770 }, { "grad_norm": 0.4059154689311981, "learning_rate": 9.995392150972451e-05, "loss": 0.037, "step": 3780 }, { "grad_norm": 0.42718711495399475, "learning_rate": 9.995273108774366e-05, "loss": 0.0397, "step": 3790 }, { "grad_norm": 0.4127197265625, "learning_rate": 9.995152549144205e-05, "loss": 0.0406, "step": 3800 }, { "grad_norm": 0.4541732966899872, "learning_rate": 9.995030472118587e-05, "loss": 0.0365, "step": 3810 }, { "grad_norm": 0.45974621176719666, "learning_rate": 9.9949068777346e-05, "loss": 0.0379, "step": 3820 }, { "grad_norm": 0.3700809180736542, "learning_rate": 9.994781766029786e-05, "loss": 0.0396, "step": 3830 }, { "grad_norm": 0.41689249873161316, "learning_rate": 9.994655137042151e-05, "loss": 0.037, "step": 3840 }, { "grad_norm": 0.4805205166339874, "learning_rate": 9.99452699081016e-05, "loss": 0.0382, "step": 3850 }, { "grad_norm": 0.39892929792404175, "learning_rate": 9.994397327372743e-05, "loss": 0.0383, "step": 3860 }, { "grad_norm": 0.30671384930610657, "learning_rate": 9.994266146769286e-05, "loss": 0.0394, "step": 3870 }, { "grad_norm": 0.31128573417663574, "learning_rate": 9.994133449039642e-05, "loss": 0.036, "step": 3880 }, { "grad_norm": 0.4541941285133362, "learning_rate": 9.993999234224118e-05, "loss": 0.037, "step": 3890 }, { "grad_norm": 0.39358237385749817, "learning_rate": 9.993863502363485e-05, "loss": 0.0387, "step": 3900 }, { "grad_norm": 0.4683881998062134, "learning_rate": 9.993726253498976e-05, "loss": 0.0425, "step": 3910 }, { "grad_norm": 0.4125257134437561, "learning_rate": 9.993587487672282e-05, "loss": 0.0451, "step": 3920 }, { "grad_norm": 0.39217472076416016, "learning_rate": 9.993447204925558e-05, "loss": 0.036, "step": 3930 }, { "grad_norm": 0.455281525850296, "learning_rate": 9.993305405301416e-05, "loss": 0.0401, "step": 3940 }, { "grad_norm": 0.4450373947620392, "learning_rate": 9.993162088842935e-05, "loss": 0.0392, "step": 3950 }, { "grad_norm": 0.4121532440185547, "learning_rate": 9.993017255593646e-05, "loss": 0.0402, "step": 3960 }, { "grad_norm": 0.3618178367614746, "learning_rate": 9.992870905597548e-05, "loss": 0.0405, "step": 3970 }, { "grad_norm": 0.3545474112033844, "learning_rate": 9.9927230388991e-05, "loss": 0.0446, "step": 3980 }, { "grad_norm": 0.32528674602508545, "learning_rate": 9.992573655543215e-05, "loss": 0.0414, "step": 3990 }, { "grad_norm": 0.30988138914108276, "learning_rate": 9.992422755575277e-05, "loss": 0.0378, "step": 4000 }, { "grad_norm": 0.34045785665512085, "learning_rate": 9.992270339041123e-05, "loss": 0.0389, "step": 4010 }, { "grad_norm": 0.29022184014320374, "learning_rate": 9.992116405987053e-05, "loss": 0.0356, "step": 4020 }, { "grad_norm": 0.3762984275817871, "learning_rate": 9.991960956459828e-05, "loss": 0.0346, "step": 4030 }, { "grad_norm": 0.3813821077346802, "learning_rate": 9.991803990506669e-05, "loss": 0.0344, "step": 4040 }, { "grad_norm": 0.2847200334072113, "learning_rate": 9.991645508175258e-05, "loss": 0.0319, "step": 4050 }, { "grad_norm": 0.38433638215065, "learning_rate": 9.99148550951374e-05, "loss": 0.0347, "step": 4060 }, { "grad_norm": 0.22351059317588806, "learning_rate": 9.991323994570716e-05, "loss": 0.0352, "step": 4070 }, { "grad_norm": 0.2622181177139282, "learning_rate": 9.99116096339525e-05, "loss": 0.0353, "step": 4080 }, { "grad_norm": 0.4043204188346863, "learning_rate": 9.990996416036869e-05, "loss": 0.037, "step": 4090 }, { "grad_norm": 0.36474737524986267, "learning_rate": 9.990830352545555e-05, "loss": 0.0359, "step": 4100 }, { "grad_norm": 0.3563602566719055, "learning_rate": 9.990662772971756e-05, "loss": 0.035, "step": 4110 }, { "grad_norm": 0.38627561926841736, "learning_rate": 9.990493677366376e-05, "loss": 0.0361, "step": 4120 }, { "grad_norm": 0.37022942304611206, "learning_rate": 9.990323065780786e-05, "loss": 0.0345, "step": 4130 }, { "grad_norm": 0.3938691020011902, "learning_rate": 9.990150938266808e-05, "loss": 0.0398, "step": 4140 }, { "grad_norm": 0.3102794587612152, "learning_rate": 9.989977294876733e-05, "loss": 0.0346, "step": 4150 }, { "grad_norm": 0.26242783665657043, "learning_rate": 9.989802135663308e-05, "loss": 0.0385, "step": 4160 }, { "grad_norm": 0.31987178325653076, "learning_rate": 9.989625460679743e-05, "loss": 0.0362, "step": 4170 }, { "grad_norm": 0.35966524481773376, "learning_rate": 9.989447269979706e-05, "loss": 0.0368, "step": 4180 }, { "grad_norm": 0.38331690430641174, "learning_rate": 9.989267563617328e-05, "loss": 0.0344, "step": 4190 }, { "grad_norm": 0.28971797227859497, "learning_rate": 9.989086341647198e-05, "loss": 0.0352, "step": 4200 }, { "grad_norm": 0.4560658633708954, "learning_rate": 9.988903604124366e-05, "loss": 0.0365, "step": 4210 }, { "grad_norm": 0.42155134677886963, "learning_rate": 9.988719351104343e-05, "loss": 0.0354, "step": 4220 }, { "grad_norm": 0.3382489085197449, "learning_rate": 9.9885335826431e-05, "loss": 0.033, "step": 4230 }, { "grad_norm": 0.34199121594429016, "learning_rate": 9.988346298797071e-05, "loss": 0.036, "step": 4240 }, { "grad_norm": 0.2737998068332672, "learning_rate": 9.988157499623146e-05, "loss": 0.0347, "step": 4250 }, { "grad_norm": 0.30714258551597595, "learning_rate": 9.987967185178677e-05, "loss": 0.0348, "step": 4260 }, { "grad_norm": 0.4649922847747803, "learning_rate": 9.987775355521476e-05, "loss": 0.0321, "step": 4270 }, { "grad_norm": 0.4174307584762573, "learning_rate": 9.987582010709817e-05, "loss": 0.0328, "step": 4280 }, { "grad_norm": 0.31184932589530945, "learning_rate": 9.987387150802431e-05, "loss": 0.0365, "step": 4290 }, { "grad_norm": 0.30284440517425537, "learning_rate": 9.987190775858517e-05, "loss": 0.0357, "step": 4300 }, { "grad_norm": 0.46506810188293457, "learning_rate": 9.98699288593772e-05, "loss": 0.037, "step": 4310 }, { "grad_norm": 0.36110562086105347, "learning_rate": 9.986793481100161e-05, "loss": 0.0342, "step": 4320 }, { "grad_norm": 0.43162137269973755, "learning_rate": 9.986592561406412e-05, "loss": 0.0376, "step": 4330 }, { "grad_norm": 0.3947873115539551, "learning_rate": 9.986390126917503e-05, "loss": 0.0367, "step": 4340 }, { "grad_norm": 0.4391649663448334, "learning_rate": 9.986186177694933e-05, "loss": 0.0341, "step": 4350 }, { "grad_norm": 0.2995440661907196, "learning_rate": 9.985980713800656e-05, "loss": 0.0337, "step": 4360 }, { "grad_norm": 0.3485530912876129, "learning_rate": 9.985773735297084e-05, "loss": 0.0338, "step": 4370 }, { "grad_norm": 0.3196663558483124, "learning_rate": 9.985565242247092e-05, "loss": 0.0365, "step": 4380 }, { "grad_norm": 0.38279029726982117, "learning_rate": 9.985355234714016e-05, "loss": 0.0358, "step": 4390 }, { "grad_norm": 0.3354455232620239, "learning_rate": 9.985143712761652e-05, "loss": 0.0345, "step": 4400 }, { "grad_norm": 0.47447797656059265, "learning_rate": 9.984930676454252e-05, "loss": 0.0376, "step": 4410 }, { "grad_norm": 0.3036918342113495, "learning_rate": 9.984716125856532e-05, "loss": 0.0337, "step": 4420 }, { "grad_norm": 0.36662983894348145, "learning_rate": 9.984500061033667e-05, "loss": 0.0362, "step": 4430 }, { "grad_norm": 0.37459877133369446, "learning_rate": 9.984282482051293e-05, "loss": 0.0334, "step": 4440 }, { "grad_norm": 0.3251151442527771, "learning_rate": 9.9840633889755e-05, "loss": 0.0356, "step": 4450 }, { "grad_norm": 0.3246462345123291, "learning_rate": 9.983842781872848e-05, "loss": 0.032, "step": 4460 }, { "grad_norm": 0.30838528275489807, "learning_rate": 9.98362066081035e-05, "loss": 0.0324, "step": 4470 }, { "grad_norm": 0.369288831949234, "learning_rate": 9.983397025855479e-05, "loss": 0.0355, "step": 4480 }, { "grad_norm": 0.35837599635124207, "learning_rate": 9.983171877076171e-05, "loss": 0.0334, "step": 4490 }, { "grad_norm": 0.2374315857887268, "learning_rate": 9.98294521454082e-05, "loss": 0.0355, "step": 4500 }, { "grad_norm": 0.29828155040740967, "learning_rate": 9.98271703831828e-05, "loss": 0.0308, "step": 4510 }, { "grad_norm": 0.36845704913139343, "learning_rate": 9.982487348477865e-05, "loss": 0.033, "step": 4520 }, { "grad_norm": 0.3268306255340576, "learning_rate": 9.982256145089347e-05, "loss": 0.0333, "step": 4530 }, { "grad_norm": 0.31096935272216797, "learning_rate": 9.982023428222962e-05, "loss": 0.0329, "step": 4540 }, { "grad_norm": 0.286575049161911, "learning_rate": 9.981789197949403e-05, "loss": 0.0334, "step": 4550 }, { "grad_norm": 0.37902337312698364, "learning_rate": 9.98155345433982e-05, "loss": 0.0344, "step": 4560 }, { "grad_norm": 0.3421473503112793, "learning_rate": 9.981316197465831e-05, "loss": 0.0313, "step": 4570 }, { "grad_norm": 0.33145880699157715, "learning_rate": 9.981077427399504e-05, "loss": 0.0333, "step": 4580 }, { "grad_norm": 0.39878609776496887, "learning_rate": 9.980837144213371e-05, "loss": 0.036, "step": 4590 }, { "grad_norm": 0.2780810296535492, "learning_rate": 9.980595347980426e-05, "loss": 0.0311, "step": 4600 }, { "grad_norm": 0.3285324275493622, "learning_rate": 9.980352038774119e-05, "loss": 0.034, "step": 4610 }, { "grad_norm": 0.27813512086868286, "learning_rate": 9.98010721666836e-05, "loss": 0.0325, "step": 4620 }, { "grad_norm": 0.3706181049346924, "learning_rate": 9.979860881737523e-05, "loss": 0.0335, "step": 4630 }, { "grad_norm": 0.3429867625236511, "learning_rate": 9.979613034056434e-05, "loss": 0.0331, "step": 4640 }, { "grad_norm": 0.3327844738960266, "learning_rate": 9.979363673700386e-05, "loss": 0.0331, "step": 4650 }, { "grad_norm": 0.3036394715309143, "learning_rate": 9.979112800745124e-05, "loss": 0.0333, "step": 4660 }, { "grad_norm": 0.36698612570762634, "learning_rate": 9.978860415266861e-05, "loss": 0.0347, "step": 4670 }, { "grad_norm": 0.3933042287826538, "learning_rate": 9.978606517342262e-05, "loss": 0.0321, "step": 4680 }, { "grad_norm": 0.29829883575439453, "learning_rate": 9.978351107048456e-05, "loss": 0.0357, "step": 4690 }, { "grad_norm": 0.299515962600708, "learning_rate": 9.978094184463029e-05, "loss": 0.0336, "step": 4700 }, { "grad_norm": 0.3219625949859619, "learning_rate": 9.977835749664029e-05, "loss": 0.033, "step": 4710 }, { "grad_norm": 0.33897021412849426, "learning_rate": 9.97757580272996e-05, "loss": 0.0368, "step": 4720 }, { "grad_norm": 0.26534420251846313, "learning_rate": 9.977314343739786e-05, "loss": 0.0343, "step": 4730 }, { "grad_norm": 0.2728618085384369, "learning_rate": 9.977051372772934e-05, "loss": 0.032, "step": 4740 }, { "grad_norm": 0.31712114810943604, "learning_rate": 9.976786889909286e-05, "loss": 0.0354, "step": 4750 }, { "grad_norm": 0.3562261164188385, "learning_rate": 9.976520895229185e-05, "loss": 0.0359, "step": 4760 }, { "grad_norm": 0.43291378021240234, "learning_rate": 9.976253388813433e-05, "loss": 0.0353, "step": 4770 }, { "grad_norm": 0.35367798805236816, "learning_rate": 9.975984370743293e-05, "loss": 0.0361, "step": 4780 }, { "grad_norm": 0.27532193064689636, "learning_rate": 9.975713841100485e-05, "loss": 0.036, "step": 4790 }, { "grad_norm": 0.34626683592796326, "learning_rate": 9.975441799967187e-05, "loss": 0.0296, "step": 4800 }, { "grad_norm": 0.3537181317806244, "learning_rate": 9.975168247426039e-05, "loss": 0.0338, "step": 4810 }, { "grad_norm": 0.29291361570358276, "learning_rate": 9.974893183560139e-05, "loss": 0.0322, "step": 4820 }, { "grad_norm": 0.3106731176376343, "learning_rate": 9.974616608453045e-05, "loss": 0.0341, "step": 4830 }, { "grad_norm": 0.2886616289615631, "learning_rate": 9.974338522188772e-05, "loss": 0.0329, "step": 4840 }, { "grad_norm": 0.30567654967308044, "learning_rate": 9.974058924851797e-05, "loss": 0.0319, "step": 4850 }, { "grad_norm": 0.31702059507369995, "learning_rate": 9.973777816527051e-05, "loss": 0.0332, "step": 4860 }, { "grad_norm": 0.4321846663951874, "learning_rate": 9.973495197299931e-05, "loss": 0.0336, "step": 4870 }, { "grad_norm": 0.3722715973854065, "learning_rate": 9.973211067256287e-05, "loss": 0.0332, "step": 4880 }, { "grad_norm": 0.35731828212738037, "learning_rate": 9.97292542648243e-05, "loss": 0.0308, "step": 4890 }, { "grad_norm": 0.293606698513031, "learning_rate": 9.972638275065131e-05, "loss": 0.0326, "step": 4900 }, { "grad_norm": 0.28698161244392395, "learning_rate": 9.972349613091621e-05, "loss": 0.0318, "step": 4910 }, { "grad_norm": 0.37889936566352844, "learning_rate": 9.972059440649584e-05, "loss": 0.0345, "step": 4920 }, { "grad_norm": 0.4578428864479065, "learning_rate": 9.971767757827168e-05, "loss": 0.0356, "step": 4930 }, { "grad_norm": 0.3379064202308655, "learning_rate": 9.971474564712982e-05, "loss": 0.033, "step": 4940 }, { "grad_norm": 0.29651397466659546, "learning_rate": 9.971179861396084e-05, "loss": 0.0332, "step": 4950 }, { "grad_norm": 0.2975715100765228, "learning_rate": 9.970883647966003e-05, "loss": 0.035, "step": 4960 }, { "grad_norm": 0.2567862272262573, "learning_rate": 9.970585924512717e-05, "loss": 0.0317, "step": 4970 }, { "grad_norm": 0.3089312016963959, "learning_rate": 9.970286691126669e-05, "loss": 0.0346, "step": 4980 }, { "grad_norm": 0.29556214809417725, "learning_rate": 9.969985947898756e-05, "loss": 0.0333, "step": 4990 }, { "grad_norm": 0.23461288213729858, "learning_rate": 9.969683694920337e-05, "loss": 0.0327, "step": 5000 }, { "grad_norm": 0.28776630759239197, "learning_rate": 9.969379932283228e-05, "loss": 0.0327, "step": 5010 }, { "grad_norm": 0.2413591593503952, "learning_rate": 9.969074660079704e-05, "loss": 0.0329, "step": 5020 }, { "grad_norm": 0.2925525903701782, "learning_rate": 9.968767878402501e-05, "loss": 0.03, "step": 5030 }, { "grad_norm": 0.2720029950141907, "learning_rate": 9.968459587344808e-05, "loss": 0.0302, "step": 5040 }, { "grad_norm": 0.31064996123313904, "learning_rate": 9.968149787000278e-05, "loss": 0.0308, "step": 5050 }, { "grad_norm": 0.298651784658432, "learning_rate": 9.967838477463018e-05, "loss": 0.031, "step": 5060 }, { "grad_norm": 0.39292728900909424, "learning_rate": 9.967525658827597e-05, "loss": 0.0297, "step": 5070 }, { "grad_norm": 0.29544663429260254, "learning_rate": 9.967211331189042e-05, "loss": 0.0293, "step": 5080 }, { "grad_norm": 0.293852299451828, "learning_rate": 9.966895494642834e-05, "loss": 0.0325, "step": 5090 }, { "grad_norm": 0.3156912326812744, "learning_rate": 9.96657814928492e-05, "loss": 0.0318, "step": 5100 }, { "grad_norm": 0.39724916219711304, "learning_rate": 9.966259295211697e-05, "loss": 0.0312, "step": 5110 }, { "grad_norm": 0.31109297275543213, "learning_rate": 9.965938932520028e-05, "loss": 0.0313, "step": 5120 }, { "grad_norm": 0.30387452244758606, "learning_rate": 9.965617061307229e-05, "loss": 0.0318, "step": 5130 }, { "grad_norm": 0.4248841106891632, "learning_rate": 9.965293681671077e-05, "loss": 0.0343, "step": 5140 }, { "grad_norm": 0.28117674589157104, "learning_rate": 9.964968793709804e-05, "loss": 0.0329, "step": 5150 }, { "grad_norm": 0.3433448076248169, "learning_rate": 9.964642397522106e-05, "loss": 0.0339, "step": 5160 }, { "grad_norm": 0.29606741666793823, "learning_rate": 9.96431449320713e-05, "loss": 0.0308, "step": 5170 }, { "grad_norm": 0.29844459891319275, "learning_rate": 9.963985080864486e-05, "loss": 0.0299, "step": 5180 }, { "grad_norm": 0.4215967357158661, "learning_rate": 9.96365416059424e-05, "loss": 0.033, "step": 5190 }, { "grad_norm": 0.2810239791870117, "learning_rate": 9.963321732496919e-05, "loss": 0.0311, "step": 5200 }, { "grad_norm": 0.3246335983276367, "learning_rate": 9.962987796673506e-05, "loss": 0.0328, "step": 5210 }, { "grad_norm": 0.32506823539733887, "learning_rate": 9.962652353225438e-05, "loss": 0.0334, "step": 5220 }, { "grad_norm": 0.24024590849876404, "learning_rate": 9.962315402254619e-05, "loss": 0.033, "step": 5230 }, { "grad_norm": 0.20617049932479858, "learning_rate": 9.9619769438634e-05, "loss": 0.0337, "step": 5240 }, { "grad_norm": 0.25945571064949036, "learning_rate": 9.9616369781546e-05, "loss": 0.0307, "step": 5250 }, { "grad_norm": 0.3110930621623993, "learning_rate": 9.961295505231491e-05, "loss": 0.0341, "step": 5260 }, { "grad_norm": 0.35372045636177063, "learning_rate": 9.960952525197804e-05, "loss": 0.0329, "step": 5270 }, { "grad_norm": 0.28020918369293213, "learning_rate": 9.960608038157724e-05, "loss": 0.0298, "step": 5280 }, { "grad_norm": 0.23485516011714935, "learning_rate": 9.960262044215901e-05, "loss": 0.0293, "step": 5290 }, { "grad_norm": 0.24987754225730896, "learning_rate": 9.959914543477435e-05, "loss": 0.0307, "step": 5300 }, { "grad_norm": 0.28317955136299133, "learning_rate": 9.959565536047892e-05, "loss": 0.0297, "step": 5310 }, { "grad_norm": 0.376844197511673, "learning_rate": 9.959215022033288e-05, "loss": 0.0301, "step": 5320 }, { "grad_norm": 0.2781023383140564, "learning_rate": 9.9588630015401e-05, "loss": 0.0296, "step": 5330 }, { "grad_norm": 0.2356785535812378, "learning_rate": 9.958509474675264e-05, "loss": 0.0302, "step": 5340 }, { "grad_norm": 0.3924737274646759, "learning_rate": 9.958154441546171e-05, "loss": 0.0298, "step": 5350 }, { "grad_norm": 0.28657618165016174, "learning_rate": 9.957797902260673e-05, "loss": 0.0316, "step": 5360 }, { "grad_norm": 0.26018309593200684, "learning_rate": 9.957439856927073e-05, "loss": 0.0307, "step": 5370 }, { "grad_norm": 0.3272460103034973, "learning_rate": 9.957080305654139e-05, "loss": 0.0289, "step": 5380 }, { "grad_norm": 0.26612427830696106, "learning_rate": 9.956719248551092e-05, "loss": 0.0295, "step": 5390 }, { "grad_norm": 0.277058482170105, "learning_rate": 9.956356685727612e-05, "loss": 0.0307, "step": 5400 }, { "grad_norm": 0.5037674307823181, "learning_rate": 9.955992617293836e-05, "loss": 0.0329, "step": 5410 }, { "grad_norm": 0.42472344636917114, "learning_rate": 9.955627043360358e-05, "loss": 0.0428, "step": 5420 }, { "grad_norm": 0.3288707435131073, "learning_rate": 9.955259964038231e-05, "loss": 0.0393, "step": 5430 }, { "grad_norm": 0.25461727380752563, "learning_rate": 9.954891379438962e-05, "loss": 0.0333, "step": 5440 }, { "grad_norm": 0.25542396306991577, "learning_rate": 9.954521289674519e-05, "loss": 0.0335, "step": 5450 }, { "grad_norm": 0.33876514434814453, "learning_rate": 9.954149694857325e-05, "loss": 0.0309, "step": 5460 }, { "grad_norm": 0.29498764872550964, "learning_rate": 9.953776595100258e-05, "loss": 0.0345, "step": 5470 }, { "grad_norm": 0.3145694434642792, "learning_rate": 9.95340199051666e-05, "loss": 0.0322, "step": 5480 }, { "grad_norm": 0.3394213020801544, "learning_rate": 9.953025881220325e-05, "loss": 0.031, "step": 5490 }, { "grad_norm": 0.3469878137111664, "learning_rate": 9.952648267325504e-05, "loss": 0.0339, "step": 5500 }, { "grad_norm": 0.3754296898841858, "learning_rate": 9.952269148946905e-05, "loss": 0.0305, "step": 5510 }, { "grad_norm": 0.2941936254501343, "learning_rate": 9.951888526199697e-05, "loss": 0.0313, "step": 5520 }, { "grad_norm": 0.2704320251941681, "learning_rate": 9.951506399199501e-05, "loss": 0.0294, "step": 5530 }, { "grad_norm": 0.321056604385376, "learning_rate": 9.951122768062399e-05, "loss": 0.0321, "step": 5540 }, { "grad_norm": 0.289594441652298, "learning_rate": 9.950737632904927e-05, "loss": 0.0317, "step": 5550 }, { "grad_norm": 0.26476195454597473, "learning_rate": 9.950350993844077e-05, "loss": 0.0318, "step": 5560 }, { "grad_norm": 0.2952621281147003, "learning_rate": 9.949962850997303e-05, "loss": 0.0311, "step": 5570 }, { "grad_norm": 0.29692786931991577, "learning_rate": 9.949573204482512e-05, "loss": 0.0362, "step": 5580 }, { "grad_norm": 0.26673075556755066, "learning_rate": 9.949182054418064e-05, "loss": 0.0316, "step": 5590 }, { "grad_norm": 0.33851301670074463, "learning_rate": 9.948789400922787e-05, "loss": 0.0343, "step": 5600 }, { "grad_norm": 0.2985857129096985, "learning_rate": 9.948395244115953e-05, "loss": 0.0335, "step": 5610 }, { "grad_norm": 0.2432062327861786, "learning_rate": 9.9479995841173e-05, "loss": 0.0322, "step": 5620 }, { "grad_norm": 0.2555656135082245, "learning_rate": 9.947602421047017e-05, "loss": 0.0335, "step": 5630 }, { "grad_norm": 0.33901143074035645, "learning_rate": 9.947203755025753e-05, "loss": 0.029, "step": 5640 }, { "grad_norm": 0.292293906211853, "learning_rate": 9.946803586174611e-05, "loss": 0.0309, "step": 5650 }, { "grad_norm": 0.3219669759273529, "learning_rate": 9.946401914615151e-05, "loss": 0.0289, "step": 5660 }, { "grad_norm": 0.3153766691684723, "learning_rate": 9.945998740469394e-05, "loss": 0.0311, "step": 5670 }, { "grad_norm": 0.3149682581424713, "learning_rate": 9.945594063859809e-05, "loss": 0.0283, "step": 5680 }, { "grad_norm": 0.3125959038734436, "learning_rate": 9.94518788490933e-05, "loss": 0.0305, "step": 5690 }, { "grad_norm": 0.24282237887382507, "learning_rate": 9.944780203741341e-05, "loss": 0.0282, "step": 5700 }, { "grad_norm": 0.27554070949554443, "learning_rate": 9.944371020479686e-05, "loss": 0.0315, "step": 5710 }, { "grad_norm": 0.28677812218666077, "learning_rate": 9.943960335248662e-05, "loss": 0.0296, "step": 5720 }, { "grad_norm": 0.3136039674282074, "learning_rate": 9.943548148173027e-05, "loss": 0.0323, "step": 5730 }, { "grad_norm": 0.29758548736572266, "learning_rate": 9.943134459377992e-05, "loss": 0.0314, "step": 5740 }, { "grad_norm": 0.22512240707874298, "learning_rate": 9.942719268989222e-05, "loss": 0.0316, "step": 5750 }, { "grad_norm": 0.3116750419139862, "learning_rate": 9.942302577132844e-05, "loss": 0.031, "step": 5760 }, { "grad_norm": 0.25314706563949585, "learning_rate": 9.941884383935438e-05, "loss": 0.0319, "step": 5770 }, { "grad_norm": 0.28800520300865173, "learning_rate": 9.941464689524039e-05, "loss": 0.0295, "step": 5780 }, { "grad_norm": 0.1957715004682541, "learning_rate": 9.941043494026139e-05, "loss": 0.0286, "step": 5790 }, { "grad_norm": 0.30325552821159363, "learning_rate": 9.940620797569685e-05, "loss": 0.0292, "step": 5800 }, { "grad_norm": 0.28528892993927, "learning_rate": 9.940196600283082e-05, "loss": 0.0297, "step": 5810 }, { "grad_norm": 0.2773872911930084, "learning_rate": 9.939770902295192e-05, "loss": 0.0303, "step": 5820 }, { "grad_norm": 0.2709914445877075, "learning_rate": 9.939343703735329e-05, "loss": 0.0275, "step": 5830 }, { "grad_norm": 0.2251684069633484, "learning_rate": 9.938915004733264e-05, "loss": 0.0308, "step": 5840 }, { "grad_norm": 0.3052627146244049, "learning_rate": 9.938484805419224e-05, "loss": 0.0288, "step": 5850 }, { "grad_norm": 0.30160462856292725, "learning_rate": 9.938053105923894e-05, "loss": 0.0301, "step": 5860 }, { "grad_norm": 0.23271365463733673, "learning_rate": 9.937619906378413e-05, "loss": 0.0288, "step": 5870 }, { "grad_norm": 0.3219897747039795, "learning_rate": 9.937185206914374e-05, "loss": 0.028, "step": 5880 }, { "grad_norm": 0.3126720190048218, "learning_rate": 9.936749007663829e-05, "loss": 0.028, "step": 5890 }, { "grad_norm": 0.3252781331539154, "learning_rate": 9.93631130875928e-05, "loss": 0.0297, "step": 5900 }, { "grad_norm": 0.236862450838089, "learning_rate": 9.935872110333692e-05, "loss": 0.0308, "step": 5910 }, { "grad_norm": 0.2367679625749588, "learning_rate": 9.935431412520484e-05, "loss": 0.0272, "step": 5920 }, { "grad_norm": 0.3256148099899292, "learning_rate": 9.934989215453523e-05, "loss": 0.0322, "step": 5930 }, { "grad_norm": 0.3288554847240448, "learning_rate": 9.934545519267139e-05, "loss": 0.0308, "step": 5940 }, { "grad_norm": 0.32576510310173035, "learning_rate": 9.934100324096117e-05, "loss": 0.0318, "step": 5950 }, { "grad_norm": 0.3180586099624634, "learning_rate": 9.933653630075692e-05, "loss": 0.0283, "step": 5960 }, { "grad_norm": 0.33035334944725037, "learning_rate": 9.93320543734156e-05, "loss": 0.0278, "step": 5970 }, { "grad_norm": 0.28793492913246155, "learning_rate": 9.932755746029871e-05, "loss": 0.0298, "step": 5980 }, { "grad_norm": 0.28880974650382996, "learning_rate": 9.932304556277228e-05, "loss": 0.0306, "step": 5990 }, { "grad_norm": 0.3549385070800781, "learning_rate": 9.93185186822069e-05, "loss": 0.0281, "step": 6000 }, { "grad_norm": 0.24114742875099182, "learning_rate": 9.931397681997773e-05, "loss": 0.0333, "step": 6010 }, { "grad_norm": 0.23648403584957123, "learning_rate": 9.930941997746446e-05, "loss": 0.0285, "step": 6020 }, { "grad_norm": 0.3019183576107025, "learning_rate": 9.930484815605134e-05, "loss": 0.031, "step": 6030 }, { "grad_norm": 0.24309030175209045, "learning_rate": 9.930026135712717e-05, "loss": 0.0283, "step": 6040 }, { "grad_norm": 0.2921231687068939, "learning_rate": 9.92956595820853e-05, "loss": 0.029, "step": 6050 }, { "grad_norm": 0.27002057433128357, "learning_rate": 9.929104283232362e-05, "loss": 0.0271, "step": 6060 }, { "grad_norm": 0.26852428913116455, "learning_rate": 9.92864111092446e-05, "loss": 0.0276, "step": 6070 }, { "grad_norm": 0.24168333411216736, "learning_rate": 9.92817644142552e-05, "loss": 0.0319, "step": 6080 }, { "grad_norm": 0.30342474579811096, "learning_rate": 9.927710274876698e-05, "loss": 0.031, "step": 6090 }, { "grad_norm": 0.24785834550857544, "learning_rate": 9.927242611419603e-05, "loss": 0.0286, "step": 6100 }, { "grad_norm": 0.2929899990558624, "learning_rate": 9.926773451196301e-05, "loss": 0.0296, "step": 6110 }, { "grad_norm": 0.26850274205207825, "learning_rate": 9.926302794349306e-05, "loss": 0.03, "step": 6120 }, { "grad_norm": 0.23336976766586304, "learning_rate": 9.925830641021594e-05, "loss": 0.0296, "step": 6130 }, { "grad_norm": 0.2559879422187805, "learning_rate": 9.925356991356593e-05, "loss": 0.0291, "step": 6140 }, { "grad_norm": 0.2502081096172333, "learning_rate": 9.924881845498184e-05, "loss": 0.0275, "step": 6150 }, { "grad_norm": 0.2867584228515625, "learning_rate": 9.924405203590705e-05, "loss": 0.0276, "step": 6160 }, { "grad_norm": 0.18329083919525146, "learning_rate": 9.923927065778946e-05, "loss": 0.0266, "step": 6170 }, { "grad_norm": 0.2786738872528076, "learning_rate": 9.923447432208154e-05, "loss": 0.027, "step": 6180 }, { "grad_norm": 0.26015129685401917, "learning_rate": 9.922966303024027e-05, "loss": 0.0282, "step": 6190 }, { "grad_norm": 0.21109814941883087, "learning_rate": 9.922483678372721e-05, "loss": 0.0278, "step": 6200 }, { "grad_norm": 0.28378602862358093, "learning_rate": 9.921999558400845e-05, "loss": 0.0284, "step": 6210 }, { "grad_norm": 0.25496265292167664, "learning_rate": 9.92151394325546e-05, "loss": 0.0288, "step": 6220 }, { "grad_norm": 0.23152996599674225, "learning_rate": 9.921026833084084e-05, "loss": 0.0261, "step": 6230 }, { "grad_norm": 0.30074837803840637, "learning_rate": 9.920538228034689e-05, "loss": 0.0278, "step": 6240 }, { "grad_norm": 0.2410011738538742, "learning_rate": 9.920048128255699e-05, "loss": 0.0269, "step": 6250 }, { "grad_norm": 0.2800215184688568, "learning_rate": 9.919556533895995e-05, "loss": 0.0265, "step": 6260 }, { "grad_norm": 0.20820626616477966, "learning_rate": 9.919063445104907e-05, "loss": 0.0273, "step": 6270 }, { "grad_norm": 0.2688838839530945, "learning_rate": 9.918568862032227e-05, "loss": 0.0268, "step": 6280 }, { "grad_norm": 0.2645457983016968, "learning_rate": 9.918072784828194e-05, "loss": 0.0295, "step": 6290 }, { "grad_norm": 0.2566224932670593, "learning_rate": 9.917575213643501e-05, "loss": 0.0274, "step": 6300 }, { "grad_norm": 0.2945362329483032, "learning_rate": 9.917076148629302e-05, "loss": 0.0288, "step": 6310 }, { "grad_norm": 0.2866954207420349, "learning_rate": 9.916575589937196e-05, "loss": 0.0287, "step": 6320 }, { "grad_norm": 0.25271373987197876, "learning_rate": 9.916073537719239e-05, "loss": 0.0295, "step": 6330 }, { "grad_norm": 0.31302255392074585, "learning_rate": 9.915569992127944e-05, "loss": 0.0283, "step": 6340 }, { "grad_norm": 0.32733944058418274, "learning_rate": 9.915064953316273e-05, "loss": 0.0325, "step": 6350 }, { "grad_norm": 0.22395510971546173, "learning_rate": 9.914558421437645e-05, "loss": 0.03, "step": 6360 }, { "grad_norm": 0.28774917125701904, "learning_rate": 9.914050396645929e-05, "loss": 0.029, "step": 6370 }, { "grad_norm": 0.3277343511581421, "learning_rate": 9.913540879095452e-05, "loss": 0.0273, "step": 6380 }, { "grad_norm": 0.2067689448595047, "learning_rate": 9.913029868940987e-05, "loss": 0.0306, "step": 6390 }, { "grad_norm": 0.2617696225643158, "learning_rate": 9.912517366337772e-05, "loss": 0.0302, "step": 6400 }, { "grad_norm": 0.28904247283935547, "learning_rate": 9.912003371441487e-05, "loss": 0.0294, "step": 6410 }, { "grad_norm": 0.291431188583374, "learning_rate": 9.911487884408271e-05, "loss": 0.0285, "step": 6420 }, { "grad_norm": 0.29333722591400146, "learning_rate": 9.910970905394719e-05, "loss": 0.0281, "step": 6430 }, { "grad_norm": 0.2977822721004486, "learning_rate": 9.91045243455787e-05, "loss": 0.0302, "step": 6440 }, { "grad_norm": 0.2928118109703064, "learning_rate": 9.909932472055225e-05, "loss": 0.0288, "step": 6450 }, { "grad_norm": 0.3452310264110565, "learning_rate": 9.909411018044734e-05, "loss": 0.0298, "step": 6460 }, { "grad_norm": 0.33308646082878113, "learning_rate": 9.908888072684802e-05, "loss": 0.0272, "step": 6470 }, { "grad_norm": 0.3039981424808502, "learning_rate": 9.908363636134285e-05, "loss": 0.0296, "step": 6480 }, { "grad_norm": 0.3938208222389221, "learning_rate": 9.907837708552493e-05, "loss": 0.0317, "step": 6490 }, { "grad_norm": 0.32274535298347473, "learning_rate": 9.90731029009919e-05, "loss": 0.0307, "step": 6500 }, { "grad_norm": 0.2410590648651123, "learning_rate": 9.906781380934589e-05, "loss": 0.0304, "step": 6510 }, { "grad_norm": 0.2110500931739807, "learning_rate": 9.906250981219362e-05, "loss": 0.029, "step": 6520 }, { "grad_norm": 0.34025058150291443, "learning_rate": 9.905719091114628e-05, "loss": 0.0292, "step": 6530 }, { "grad_norm": 0.23945483565330505, "learning_rate": 9.905185710781964e-05, "loss": 0.026, "step": 6540 }, { "grad_norm": 0.32492420077323914, "learning_rate": 9.904650840383392e-05, "loss": 0.0282, "step": 6550 }, { "grad_norm": 0.22718916833400726, "learning_rate": 9.904114480081397e-05, "loss": 0.0292, "step": 6560 }, { "grad_norm": 0.2434980720281601, "learning_rate": 9.903576630038906e-05, "loss": 0.0297, "step": 6570 }, { "grad_norm": 0.2574143707752228, "learning_rate": 9.903037290419309e-05, "loss": 0.0257, "step": 6580 }, { "grad_norm": 0.2736647427082062, "learning_rate": 9.902496461386439e-05, "loss": 0.0284, "step": 6590 }, { "grad_norm": 0.3033333420753479, "learning_rate": 9.901954143104588e-05, "loss": 0.0291, "step": 6600 }, { "grad_norm": 0.26594457030296326, "learning_rate": 9.901410335738496e-05, "loss": 0.0291, "step": 6610 }, { "grad_norm": 0.2306976318359375, "learning_rate": 9.900865039453358e-05, "loss": 0.0295, "step": 6620 }, { "grad_norm": 0.24834689497947693, "learning_rate": 9.900318254414821e-05, "loss": 0.0305, "step": 6630 }, { "grad_norm": 0.25717562437057495, "learning_rate": 9.899769980788985e-05, "loss": 0.031, "step": 6640 }, { "grad_norm": 0.22926220297813416, "learning_rate": 9.899220218742398e-05, "loss": 0.0318, "step": 6650 }, { "grad_norm": 0.3093836009502411, "learning_rate": 9.898668968442066e-05, "loss": 0.0271, "step": 6660 }, { "grad_norm": 0.2784537076950073, "learning_rate": 9.898116230055443e-05, "loss": 0.0285, "step": 6670 }, { "grad_norm": 0.2754233777523041, "learning_rate": 9.897562003750437e-05, "loss": 0.0282, "step": 6680 }, { "grad_norm": 0.21481990814208984, "learning_rate": 9.897006289695407e-05, "loss": 0.0274, "step": 6690 }, { "grad_norm": 0.20389050245285034, "learning_rate": 9.896449088059164e-05, "loss": 0.026, "step": 6700 }, { "grad_norm": 0.2640131711959839, "learning_rate": 9.89589039901097e-05, "loss": 0.0276, "step": 6710 }, { "grad_norm": 0.2913075089454651, "learning_rate": 9.895330222720542e-05, "loss": 0.0307, "step": 6720 }, { "grad_norm": 0.27000707387924194, "learning_rate": 9.894768559358047e-05, "loss": 0.0296, "step": 6730 }, { "grad_norm": 0.24074703454971313, "learning_rate": 9.894205409094101e-05, "loss": 0.0281, "step": 6740 }, { "grad_norm": 0.27539604902267456, "learning_rate": 9.893640772099777e-05, "loss": 0.0333, "step": 6750 }, { "grad_norm": 0.22955359518527985, "learning_rate": 9.893074648546595e-05, "loss": 0.0294, "step": 6760 }, { "grad_norm": 0.24148526787757874, "learning_rate": 9.892507038606528e-05, "loss": 0.0287, "step": 6770 }, { "grad_norm": 0.23778176307678223, "learning_rate": 9.891937942452003e-05, "loss": 0.0301, "step": 6780 }, { "grad_norm": 0.26316407322883606, "learning_rate": 9.891367360255895e-05, "loss": 0.0288, "step": 6790 }, { "grad_norm": 0.2313854694366455, "learning_rate": 9.890795292191532e-05, "loss": 0.0284, "step": 6800 }, { "grad_norm": 0.27560916543006897, "learning_rate": 9.890221738432694e-05, "loss": 0.0291, "step": 6810 }, { "grad_norm": 0.25388261675834656, "learning_rate": 9.88964669915361e-05, "loss": 0.031, "step": 6820 }, { "grad_norm": 0.21308894455432892, "learning_rate": 9.889070174528963e-05, "loss": 0.0275, "step": 6830 }, { "grad_norm": 0.26833102107048035, "learning_rate": 9.888492164733883e-05, "loss": 0.0282, "step": 6840 }, { "grad_norm": 0.24352921545505524, "learning_rate": 9.88791266994396e-05, "loss": 0.0277, "step": 6850 }, { "grad_norm": 0.26232218742370605, "learning_rate": 9.887331690335223e-05, "loss": 0.0274, "step": 6860 }, { "grad_norm": 0.3173314034938812, "learning_rate": 9.886749226084163e-05, "loss": 0.0274, "step": 6870 }, { "grad_norm": 0.28463318943977356, "learning_rate": 9.886165277367714e-05, "loss": 0.0287, "step": 6880 }, { "grad_norm": 0.2516433298587799, "learning_rate": 9.885579844363265e-05, "loss": 0.0259, "step": 6890 }, { "grad_norm": 0.21570511162281036, "learning_rate": 9.884992927248656e-05, "loss": 0.0299, "step": 6900 }, { "grad_norm": 0.2926516532897949, "learning_rate": 9.884404526202178e-05, "loss": 0.0289, "step": 6910 }, { "grad_norm": 0.25230905413627625, "learning_rate": 9.883814641402568e-05, "loss": 0.0274, "step": 6920 }, { "grad_norm": 0.30464574694633484, "learning_rate": 9.88322327302902e-05, "loss": 0.0288, "step": 6930 }, { "grad_norm": 0.2477879673242569, "learning_rate": 9.882630421261176e-05, "loss": 0.0261, "step": 6940 }, { "grad_norm": 0.35531020164489746, "learning_rate": 9.88203608627913e-05, "loss": 0.0284, "step": 6950 }, { "grad_norm": 0.2945120334625244, "learning_rate": 9.881440268263422e-05, "loss": 0.028, "step": 6960 }, { "grad_norm": 0.2614474594593048, "learning_rate": 9.880842967395048e-05, "loss": 0.0278, "step": 6970 }, { "grad_norm": 0.23492522537708282, "learning_rate": 9.880244183855452e-05, "loss": 0.0287, "step": 6980 }, { "grad_norm": 0.23125509917736053, "learning_rate": 9.879643917826527e-05, "loss": 0.026, "step": 6990 }, { "grad_norm": 0.2307005226612091, "learning_rate": 9.87904216949062e-05, "loss": 0.0243, "step": 7000 }, { "grad_norm": 0.3021189570426941, "learning_rate": 9.878438939030526e-05, "loss": 0.0293, "step": 7010 }, { "grad_norm": 0.28832754492759705, "learning_rate": 9.877834226629489e-05, "loss": 0.0278, "step": 7020 }, { "grad_norm": 0.2800067961215973, "learning_rate": 9.877228032471206e-05, "loss": 0.0261, "step": 7030 }, { "grad_norm": 0.31534942984580994, "learning_rate": 9.876620356739823e-05, "loss": 0.0256, "step": 7040 }, { "grad_norm": 0.26259201765060425, "learning_rate": 9.876011199619935e-05, "loss": 0.0279, "step": 7050 }, { "grad_norm": 0.30858129262924194, "learning_rate": 9.875400561296589e-05, "loss": 0.0253, "step": 7060 }, { "grad_norm": 0.1915603131055832, "learning_rate": 9.874788441955278e-05, "loss": 0.0283, "step": 7070 }, { "grad_norm": 0.20796778798103333, "learning_rate": 9.874174841781951e-05, "loss": 0.0258, "step": 7080 }, { "grad_norm": 0.27555519342422485, "learning_rate": 9.873559760963003e-05, "loss": 0.028, "step": 7090 }, { "grad_norm": 0.22024813294410706, "learning_rate": 9.872943199685278e-05, "loss": 0.025, "step": 7100 }, { "grad_norm": 0.3280911445617676, "learning_rate": 9.872325158136071e-05, "loss": 0.0291, "step": 7110 }, { "grad_norm": 0.26900649070739746, "learning_rate": 9.871705636503128e-05, "loss": 0.0274, "step": 7120 }, { "grad_norm": 0.18032699823379517, "learning_rate": 9.871084634974641e-05, "loss": 0.0277, "step": 7130 }, { "grad_norm": 0.23416954278945923, "learning_rate": 9.870462153739257e-05, "loss": 0.0251, "step": 7140 }, { "grad_norm": 0.25608623027801514, "learning_rate": 9.869838192986067e-05, "loss": 0.0248, "step": 7150 }, { "grad_norm": 0.1867072582244873, "learning_rate": 9.869212752904616e-05, "loss": 0.0269, "step": 7160 }, { "grad_norm": 0.22921614348888397, "learning_rate": 9.868585833684894e-05, "loss": 0.0269, "step": 7170 }, { "grad_norm": 0.31323370337486267, "learning_rate": 9.867957435517342e-05, "loss": 0.0279, "step": 7180 }, { "grad_norm": 0.2607262134552002, "learning_rate": 9.867327558592854e-05, "loss": 0.0301, "step": 7190 }, { "grad_norm": 0.24962440133094788, "learning_rate": 9.866696203102766e-05, "loss": 0.029, "step": 7200 }, { "grad_norm": 0.2805942893028259, "learning_rate": 9.86606336923887e-05, "loss": 0.0295, "step": 7210 }, { "grad_norm": 0.22196993231773376, "learning_rate": 9.865429057193403e-05, "loss": 0.0256, "step": 7220 }, { "grad_norm": 0.2509216070175171, "learning_rate": 9.864793267159053e-05, "loss": 0.0306, "step": 7230 }, { "grad_norm": 0.2777119576931, "learning_rate": 9.864155999328957e-05, "loss": 0.03, "step": 7240 }, { "grad_norm": 0.2071262151002884, "learning_rate": 9.8635172538967e-05, "loss": 0.0262, "step": 7250 }, { "grad_norm": 0.2440653294324875, "learning_rate": 9.862877031056312e-05, "loss": 0.0249, "step": 7260 }, { "grad_norm": 0.22216929495334625, "learning_rate": 9.862235331002279e-05, "loss": 0.027, "step": 7270 }, { "grad_norm": 0.23377594351768494, "learning_rate": 9.861592153929533e-05, "loss": 0.0252, "step": 7280 }, { "grad_norm": 0.22563529014587402, "learning_rate": 9.860947500033455e-05, "loss": 0.028, "step": 7290 }, { "grad_norm": 0.21081647276878357, "learning_rate": 9.86030136950987e-05, "loss": 0.0265, "step": 7300 }, { "grad_norm": 0.24316123127937317, "learning_rate": 9.85965376255506e-05, "loss": 0.026, "step": 7310 }, { "grad_norm": 0.25097841024398804, "learning_rate": 9.859004679365747e-05, "loss": 0.028, "step": 7320 }, { "grad_norm": 0.23964770138263702, "learning_rate": 9.858354120139108e-05, "loss": 0.0258, "step": 7330 }, { "grad_norm": 0.2085687667131424, "learning_rate": 9.857702085072764e-05, "loss": 0.0243, "step": 7340 }, { "grad_norm": 0.23856808245182037, "learning_rate": 9.857048574364787e-05, "loss": 0.0257, "step": 7350 }, { "grad_norm": 0.24985067546367645, "learning_rate": 9.856393588213698e-05, "loss": 0.0316, "step": 7360 }, { "grad_norm": 0.2426789551973343, "learning_rate": 9.855737126818458e-05, "loss": 0.03, "step": 7370 }, { "grad_norm": 0.1956608146429062, "learning_rate": 9.855079190378491e-05, "loss": 0.0278, "step": 7380 }, { "grad_norm": 0.2360762357711792, "learning_rate": 9.854419779093655e-05, "loss": 0.0285, "step": 7390 }, { "grad_norm": 0.21629855036735535, "learning_rate": 9.853758893164264e-05, "loss": 0.0269, "step": 7400 }, { "grad_norm": 0.2763349413871765, "learning_rate": 9.853096532791078e-05, "loss": 0.0248, "step": 7410 }, { "grad_norm": 0.2899736166000366, "learning_rate": 9.852432698175304e-05, "loss": 0.0275, "step": 7420 }, { "grad_norm": 0.18535150587558746, "learning_rate": 9.851767389518597e-05, "loss": 0.0271, "step": 7430 }, { "grad_norm": 0.25067365169525146, "learning_rate": 9.85110060702306e-05, "loss": 0.0268, "step": 7440 }, { "grad_norm": 0.2756042182445526, "learning_rate": 9.850432350891245e-05, "loss": 0.0272, "step": 7450 }, { "grad_norm": 0.31211787462234497, "learning_rate": 9.84976262132615e-05, "loss": 0.0292, "step": 7460 }, { "grad_norm": 0.34506872296333313, "learning_rate": 9.849091418531222e-05, "loss": 0.0271, "step": 7470 }, { "grad_norm": 0.2571541965007782, "learning_rate": 9.848418742710353e-05, "loss": 0.0257, "step": 7480 }, { "grad_norm": 0.1918298751115799, "learning_rate": 9.847744594067885e-05, "loss": 0.0239, "step": 7490 }, { "grad_norm": 0.2672421634197235, "learning_rate": 9.847068972808607e-05, "loss": 0.0287, "step": 7500 }, { "grad_norm": 0.2653951346874237, "learning_rate": 9.846391879137756e-05, "loss": 0.0252, "step": 7510 }, { "grad_norm": 0.19639992713928223, "learning_rate": 9.845713313261012e-05, "loss": 0.0292, "step": 7520 }, { "grad_norm": 0.3040771186351776, "learning_rate": 9.845033275384505e-05, "loss": 0.0274, "step": 7530 }, { "grad_norm": 0.22853480279445648, "learning_rate": 9.844351765714818e-05, "loss": 0.0257, "step": 7540 }, { "grad_norm": 0.21130892634391785, "learning_rate": 9.843668784458971e-05, "loss": 0.0262, "step": 7550 }, { "grad_norm": 0.2420453280210495, "learning_rate": 9.842984331824437e-05, "loss": 0.0305, "step": 7560 }, { "grad_norm": 0.23674866557121277, "learning_rate": 9.842298408019133e-05, "loss": 0.0254, "step": 7570 }, { "grad_norm": 0.2004309743642807, "learning_rate": 9.841611013251429e-05, "loss": 0.0252, "step": 7580 }, { "grad_norm": 0.2963436543941498, "learning_rate": 9.840922147730133e-05, "loss": 0.0267, "step": 7590 }, { "grad_norm": 0.19782225787639618, "learning_rate": 9.840231811664506e-05, "loss": 0.0258, "step": 7600 }, { "grad_norm": 0.17377980053424835, "learning_rate": 9.839540005264252e-05, "loss": 0.0246, "step": 7610 }, { "grad_norm": 0.2999553084373474, "learning_rate": 9.838846728739527e-05, "loss": 0.0261, "step": 7620 }, { "grad_norm": 0.26307234168052673, "learning_rate": 9.838151982300927e-05, "loss": 0.0275, "step": 7630 }, { "grad_norm": 0.2792263925075531, "learning_rate": 9.8374557661595e-05, "loss": 0.0274, "step": 7640 }, { "grad_norm": 0.24399735033512115, "learning_rate": 9.836758080526735e-05, "loss": 0.0266, "step": 7650 }, { "grad_norm": 0.26408475637435913, "learning_rate": 9.836058925614575e-05, "loss": 0.0273, "step": 7660 }, { "grad_norm": 0.19073951244354248, "learning_rate": 9.8353583016354e-05, "loss": 0.0275, "step": 7670 }, { "grad_norm": 0.2250853180885315, "learning_rate": 9.834656208802044e-05, "loss": 0.0263, "step": 7680 }, { "grad_norm": 0.22960993647575378, "learning_rate": 9.833952647327784e-05, "loss": 0.0256, "step": 7690 }, { "grad_norm": 0.272298663854599, "learning_rate": 9.833247617426342e-05, "loss": 0.0288, "step": 7700 }, { "grad_norm": 0.25993505120277405, "learning_rate": 9.832541119311889e-05, "loss": 0.0274, "step": 7710 }, { "grad_norm": 0.2274685502052307, "learning_rate": 9.83183315319904e-05, "loss": 0.0257, "step": 7720 }, { "grad_norm": 0.3074868321418762, "learning_rate": 9.831123719302855e-05, "loss": 0.0278, "step": 7730 }, { "grad_norm": 0.2599892020225525, "learning_rate": 9.830412817838842e-05, "loss": 0.0259, "step": 7740 }, { "grad_norm": 0.3126978576183319, "learning_rate": 9.829700449022956e-05, "loss": 0.0283, "step": 7750 }, { "grad_norm": 0.1913173943758011, "learning_rate": 9.828986613071593e-05, "loss": 0.0267, "step": 7760 }, { "grad_norm": 0.26414754986763, "learning_rate": 9.828271310201601e-05, "loss": 0.0335, "step": 7770 }, { "grad_norm": 0.2751868665218353, "learning_rate": 9.827554540630268e-05, "loss": 0.0285, "step": 7780 }, { "grad_norm": 0.26543620228767395, "learning_rate": 9.826836304575329e-05, "loss": 0.0261, "step": 7790 }, { "grad_norm": 0.24553710222244263, "learning_rate": 9.826116602254966e-05, "loss": 0.0263, "step": 7800 }, { "grad_norm": 0.3040827214717865, "learning_rate": 9.825395433887805e-05, "loss": 0.0269, "step": 7810 }, { "grad_norm": 0.2862716019153595, "learning_rate": 9.824672799692917e-05, "loss": 0.0255, "step": 7820 }, { "grad_norm": 0.22945339977741241, "learning_rate": 9.823948699889823e-05, "loss": 0.0286, "step": 7830 }, { "grad_norm": 0.2710127830505371, "learning_rate": 9.823223134698483e-05, "loss": 0.0269, "step": 7840 }, { "grad_norm": 0.2895180881023407, "learning_rate": 9.822496104339303e-05, "loss": 0.0268, "step": 7850 }, { "grad_norm": 0.23108135163784027, "learning_rate": 9.821767609033138e-05, "loss": 0.0274, "step": 7860 }, { "grad_norm": 0.21866479516029358, "learning_rate": 9.821037649001284e-05, "loss": 0.0263, "step": 7870 }, { "grad_norm": 0.22328267991542816, "learning_rate": 9.820306224465486e-05, "loss": 0.0254, "step": 7880 }, { "grad_norm": 0.22696486115455627, "learning_rate": 9.819573335647928e-05, "loss": 0.0281, "step": 7890 }, { "grad_norm": 0.23738110065460205, "learning_rate": 9.818838982771246e-05, "loss": 0.0271, "step": 7900 }, { "grad_norm": 0.2795117199420929, "learning_rate": 9.818103166058514e-05, "loss": 0.0275, "step": 7910 }, { "grad_norm": 0.26686641573905945, "learning_rate": 9.817365885733254e-05, "loss": 0.0255, "step": 7920 }, { "grad_norm": 0.26420319080352783, "learning_rate": 9.816627142019434e-05, "loss": 0.0269, "step": 7930 }, { "grad_norm": 0.24678432941436768, "learning_rate": 9.815886935141463e-05, "loss": 0.0284, "step": 7940 }, { "grad_norm": 0.24574162065982819, "learning_rate": 9.8151452653242e-05, "loss": 0.0261, "step": 7950 }, { "grad_norm": 0.2821238040924072, "learning_rate": 9.814402132792939e-05, "loss": 0.0286, "step": 7960 }, { "grad_norm": 0.2535751163959503, "learning_rate": 9.813657537773428e-05, "loss": 0.0271, "step": 7970 }, { "grad_norm": 0.238338440656662, "learning_rate": 9.812911480491854e-05, "loss": 0.028, "step": 7980 }, { "grad_norm": 0.255340576171875, "learning_rate": 9.81216396117485e-05, "loss": 0.0277, "step": 7990 }, { "grad_norm": 0.20367497205734253, "learning_rate": 9.811414980049491e-05, "loss": 0.0254, "step": 8000 }, { "grad_norm": 0.21424312889575958, "learning_rate": 9.810664537343301e-05, "loss": 0.0256, "step": 8010 }, { "grad_norm": 0.28432998061180115, "learning_rate": 9.809912633284243e-05, "loss": 0.0263, "step": 8020 }, { "grad_norm": 0.24180522561073303, "learning_rate": 9.809159268100725e-05, "loss": 0.0261, "step": 8030 }, { "grad_norm": 0.1890978217124939, "learning_rate": 9.808404442021599e-05, "loss": 0.0263, "step": 8040 }, { "grad_norm": 0.26292476058006287, "learning_rate": 9.807648155276163e-05, "loss": 0.0261, "step": 8050 }, { "grad_norm": 0.2646186053752899, "learning_rate": 9.806890408094156e-05, "loss": 0.0264, "step": 8060 }, { "grad_norm": 0.23747938871383667, "learning_rate": 9.806131200705761e-05, "loss": 0.025, "step": 8070 }, { "grad_norm": 0.21323783695697784, "learning_rate": 9.805370533341605e-05, "loss": 0.0271, "step": 8080 }, { "grad_norm": 0.2354414165019989, "learning_rate": 9.804608406232762e-05, "loss": 0.0267, "step": 8090 }, { "grad_norm": 0.21193870902061462, "learning_rate": 9.803844819610741e-05, "loss": 0.0268, "step": 8100 }, { "grad_norm": 0.22799676656723022, "learning_rate": 9.803079773707504e-05, "loss": 0.0265, "step": 8110 }, { "grad_norm": 0.28898417949676514, "learning_rate": 9.802313268755447e-05, "loss": 0.0265, "step": 8120 }, { "grad_norm": 0.2433178424835205, "learning_rate": 9.801545304987419e-05, "loss": 0.0263, "step": 8130 }, { "grad_norm": 0.22598324716091156, "learning_rate": 9.800775882636704e-05, "loss": 0.0251, "step": 8140 }, { "grad_norm": 0.3256177604198456, "learning_rate": 9.800005001937034e-05, "loss": 0.0278, "step": 8150 }, { "grad_norm": 0.22237522900104523, "learning_rate": 9.79923266312258e-05, "loss": 0.0269, "step": 8160 }, { "grad_norm": 0.2742556929588318, "learning_rate": 9.79845886642796e-05, "loss": 0.0273, "step": 8170 }, { "grad_norm": 0.20090492069721222, "learning_rate": 9.797683612088233e-05, "loss": 0.027, "step": 8180 }, { "grad_norm": 0.19968654215335846, "learning_rate": 9.796906900338898e-05, "loss": 0.026, "step": 8190 }, { "grad_norm": 0.22295857965946198, "learning_rate": 9.796128731415903e-05, "loss": 0.0262, "step": 8200 }, { "grad_norm": 0.2231748253107071, "learning_rate": 9.795349105555634e-05, "loss": 0.026, "step": 8210 }, { "grad_norm": 0.24936574697494507, "learning_rate": 9.794568022994922e-05, "loss": 0.0243, "step": 8220 }, { "grad_norm": 0.20294170081615448, "learning_rate": 9.793785483971034e-05, "loss": 0.026, "step": 8230 }, { "grad_norm": 0.19491326808929443, "learning_rate": 9.793001488721691e-05, "loss": 0.0259, "step": 8240 }, { "grad_norm": 0.24245037138462067, "learning_rate": 9.792216037485047e-05, "loss": 0.0274, "step": 8250 }, { "grad_norm": 0.17787957191467285, "learning_rate": 9.791429130499704e-05, "loss": 0.0239, "step": 8260 }, { "grad_norm": 0.19711439311504364, "learning_rate": 9.790640768004698e-05, "loss": 0.0255, "step": 8270 }, { "grad_norm": 0.2796955704689026, "learning_rate": 9.789850950239518e-05, "loss": 0.026, "step": 8280 }, { "grad_norm": 0.3108673095703125, "learning_rate": 9.789059677444089e-05, "loss": 0.028, "step": 8290 }, { "grad_norm": 0.3072154223918915, "learning_rate": 9.788266949858776e-05, "loss": 0.0269, "step": 8300 }, { "grad_norm": 0.2593250274658203, "learning_rate": 9.787472767724392e-05, "loss": 0.0269, "step": 8310 }, { "grad_norm": 0.1853855699300766, "learning_rate": 9.786677131282185e-05, "loss": 0.0249, "step": 8320 }, { "grad_norm": 0.23580563068389893, "learning_rate": 9.785880040773853e-05, "loss": 0.0261, "step": 8330 }, { "grad_norm": 0.21695177257061005, "learning_rate": 9.785081496441527e-05, "loss": 0.0248, "step": 8340 }, { "grad_norm": 0.2349850982427597, "learning_rate": 9.784281498527785e-05, "loss": 0.028, "step": 8350 }, { "grad_norm": 0.20918238162994385, "learning_rate": 9.783480047275646e-05, "loss": 0.0262, "step": 8360 }, { "grad_norm": 0.24386315047740936, "learning_rate": 9.78267714292857e-05, "loss": 0.0241, "step": 8370 }, { "grad_norm": 0.23874667286872864, "learning_rate": 9.781872785730454e-05, "loss": 0.0265, "step": 8380 }, { "grad_norm": 0.2388232797384262, "learning_rate": 9.781066975925646e-05, "loss": 0.0263, "step": 8390 }, { "grad_norm": 0.1774873435497284, "learning_rate": 9.780259713758928e-05, "loss": 0.0253, "step": 8400 }, { "grad_norm": 0.1880027800798416, "learning_rate": 9.779450999475524e-05, "loss": 0.0274, "step": 8410 }, { "grad_norm": 0.21869270503520966, "learning_rate": 9.7786408333211e-05, "loss": 0.0241, "step": 8420 }, { "grad_norm": 0.17756371200084686, "learning_rate": 9.777829215541764e-05, "loss": 0.0243, "step": 8430 }, { "grad_norm": 0.21194060146808624, "learning_rate": 9.777016146384064e-05, "loss": 0.0252, "step": 8440 }, { "grad_norm": 0.3014346957206726, "learning_rate": 9.776201626094988e-05, "loss": 0.0297, "step": 8450 }, { "grad_norm": 0.2449604868888855, "learning_rate": 9.775385654921965e-05, "loss": 0.0245, "step": 8460 }, { "grad_norm": 0.21766744554042816, "learning_rate": 9.774568233112868e-05, "loss": 0.0245, "step": 8470 }, { "grad_norm": 0.22468304634094238, "learning_rate": 9.773749360916007e-05, "loss": 0.0265, "step": 8480 }, { "grad_norm": 0.23219642043113708, "learning_rate": 9.772929038580134e-05, "loss": 0.0246, "step": 8490 }, { "grad_norm": 0.22052878141403198, "learning_rate": 9.772107266354439e-05, "loss": 0.0295, "step": 8500 }, { "grad_norm": 0.24451607465744019, "learning_rate": 9.77128404448856e-05, "loss": 0.0259, "step": 8510 }, { "grad_norm": 0.24188332259655, "learning_rate": 9.770459373232565e-05, "loss": 0.0243, "step": 8520 }, { "grad_norm": 0.24459144473075867, "learning_rate": 9.769633252836969e-05, "loss": 0.0237, "step": 8530 }, { "grad_norm": 0.2951214015483856, "learning_rate": 9.768805683552724e-05, "loss": 0.0278, "step": 8540 }, { "grad_norm": 0.24373754858970642, "learning_rate": 9.767976665631228e-05, "loss": 0.0272, "step": 8550 }, { "grad_norm": 0.3392750918865204, "learning_rate": 9.767146199324311e-05, "loss": 0.029, "step": 8560 }, { "grad_norm": 0.20371083915233612, "learning_rate": 9.766314284884249e-05, "loss": 0.0293, "step": 8570 }, { "grad_norm": 0.26837071776390076, "learning_rate": 9.765480922563752e-05, "loss": 0.0265, "step": 8580 }, { "grad_norm": 0.2512645125389099, "learning_rate": 9.764646112615978e-05, "loss": 0.0244, "step": 8590 }, { "grad_norm": 0.2224784940481186, "learning_rate": 9.763809855294517e-05, "loss": 0.0247, "step": 8600 }, { "grad_norm": 0.23505403101444244, "learning_rate": 9.762972150853404e-05, "loss": 0.0263, "step": 8610 }, { "grad_norm": 0.21979005634784698, "learning_rate": 9.762132999547111e-05, "loss": 0.0249, "step": 8620 }, { "grad_norm": 0.16457732021808624, "learning_rate": 9.761292401630549e-05, "loss": 0.0262, "step": 8630 }, { "grad_norm": 0.24565093219280243, "learning_rate": 9.76045035735907e-05, "loss": 0.0258, "step": 8640 }, { "grad_norm": 0.26288679242134094, "learning_rate": 9.759606866988464e-05, "loss": 0.0258, "step": 8650 }, { "grad_norm": 0.269319087266922, "learning_rate": 9.758761930774963e-05, "loss": 0.0243, "step": 8660 }, { "grad_norm": 0.19415752589702606, "learning_rate": 9.757915548975235e-05, "loss": 0.0258, "step": 8670 }, { "grad_norm": 0.27752652764320374, "learning_rate": 9.757067721846389e-05, "loss": 0.0248, "step": 8680 }, { "grad_norm": 0.15380871295928955, "learning_rate": 9.756218449645971e-05, "loss": 0.024, "step": 8690 }, { "grad_norm": 0.23814809322357178, "learning_rate": 9.75536773263197e-05, "loss": 0.0224, "step": 8700 }, { "grad_norm": 0.24076582491397858, "learning_rate": 9.75451557106281e-05, "loss": 0.0254, "step": 8710 }, { "grad_norm": 0.2243119776248932, "learning_rate": 9.753661965197354e-05, "loss": 0.0264, "step": 8720 }, { "grad_norm": 0.23268432915210724, "learning_rate": 9.752806915294908e-05, "loss": 0.0251, "step": 8730 }, { "grad_norm": 0.2044651061296463, "learning_rate": 9.75195042161521e-05, "loss": 0.0246, "step": 8740 }, { "grad_norm": 0.22495262324810028, "learning_rate": 9.751092484418442e-05, "loss": 0.0231, "step": 8750 }, { "grad_norm": 0.19704985618591309, "learning_rate": 9.750233103965224e-05, "loss": 0.0236, "step": 8760 }, { "grad_norm": 0.21915596723556519, "learning_rate": 9.749372280516611e-05, "loss": 0.0263, "step": 8770 }, { "grad_norm": 0.2315548062324524, "learning_rate": 9.748510014334097e-05, "loss": 0.0255, "step": 8780 }, { "grad_norm": 0.22854577004909515, "learning_rate": 9.747646305679621e-05, "loss": 0.0258, "step": 8790 }, { "grad_norm": 0.21719229221343994, "learning_rate": 9.74678115481555e-05, "loss": 0.0227, "step": 8800 }, { "grad_norm": 0.2399194985628128, "learning_rate": 9.745914562004696e-05, "loss": 0.0233, "step": 8810 }, { "grad_norm": 0.22333671152591705, "learning_rate": 9.745046527510307e-05, "loss": 0.0239, "step": 8820 }, { "grad_norm": 0.19889762997627258, "learning_rate": 9.744177051596068e-05, "loss": 0.0237, "step": 8830 }, { "grad_norm": 0.21021801233291626, "learning_rate": 9.743306134526105e-05, "loss": 0.0265, "step": 8840 }, { "grad_norm": 0.16545449197292328, "learning_rate": 9.742433776564977e-05, "loss": 0.0255, "step": 8850 }, { "grad_norm": 0.26331043243408203, "learning_rate": 9.741559977977683e-05, "loss": 0.024, "step": 8860 }, { "grad_norm": 0.21327252686023712, "learning_rate": 9.740684739029661e-05, "loss": 0.0257, "step": 8870 }, { "grad_norm": 0.1923227161169052, "learning_rate": 9.739808059986789e-05, "loss": 0.024, "step": 8880 }, { "grad_norm": 0.24084725975990295, "learning_rate": 9.738929941115373e-05, "loss": 0.0276, "step": 8890 }, { "grad_norm": 0.2175656259059906, "learning_rate": 9.738050382682167e-05, "loss": 0.0262, "step": 8900 }, { "grad_norm": 0.26175469160079956, "learning_rate": 9.737169384954355e-05, "loss": 0.0247, "step": 8910 }, { "grad_norm": 0.23860718309879303, "learning_rate": 9.736286948199562e-05, "loss": 0.0259, "step": 8920 }, { "grad_norm": 0.27065762877464294, "learning_rate": 9.735403072685848e-05, "loss": 0.0288, "step": 8930 }, { "grad_norm": 0.29785311222076416, "learning_rate": 9.734517758681712e-05, "loss": 0.0247, "step": 8940 }, { "grad_norm": 0.17506177723407745, "learning_rate": 9.733631006456088e-05, "loss": 0.0246, "step": 8950 }, { "grad_norm": 0.2281722128391266, "learning_rate": 9.732742816278348e-05, "loss": 0.0263, "step": 8960 }, { "grad_norm": 0.169467031955719, "learning_rate": 9.731853188418302e-05, "loss": 0.0272, "step": 8970 }, { "grad_norm": 0.22951051592826843, "learning_rate": 9.730962123146194e-05, "loss": 0.0252, "step": 8980 }, { "grad_norm": 0.21849635243415833, "learning_rate": 9.730069620732709e-05, "loss": 0.0263, "step": 8990 }, { "grad_norm": 0.258720725774765, "learning_rate": 9.72917568144896e-05, "loss": 0.0268, "step": 9000 }, { "grad_norm": 0.2440410703420639, "learning_rate": 9.728280305566509e-05, "loss": 0.0269, "step": 9010 }, { "grad_norm": 0.26378151774406433, "learning_rate": 9.727383493357343e-05, "loss": 0.0277, "step": 9020 }, { "grad_norm": 0.2159527689218521, "learning_rate": 9.726485245093891e-05, "loss": 0.0277, "step": 9030 }, { "grad_norm": 0.234807550907135, "learning_rate": 9.725585561049018e-05, "loss": 0.0244, "step": 9040 }, { "grad_norm": 0.19746145606040955, "learning_rate": 9.724684441496022e-05, "loss": 0.0235, "step": 9050 }, { "grad_norm": 0.1995624154806137, "learning_rate": 9.72378188670864e-05, "loss": 0.0257, "step": 9060 }, { "grad_norm": 0.2618711292743683, "learning_rate": 9.722877896961047e-05, "loss": 0.0246, "step": 9070 }, { "grad_norm": 0.2237137407064438, "learning_rate": 9.721972472527848e-05, "loss": 0.024, "step": 9080 }, { "grad_norm": 0.22288453578948975, "learning_rate": 9.721065613684089e-05, "loss": 0.0255, "step": 9090 }, { "grad_norm": 0.21933037042617798, "learning_rate": 9.72015732070525e-05, "loss": 0.0236, "step": 9100 }, { "grad_norm": 0.21873369812965393, "learning_rate": 9.719247593867244e-05, "loss": 0.0225, "step": 9110 }, { "grad_norm": 0.18032406270503998, "learning_rate": 9.718336433446423e-05, "loss": 0.024, "step": 9120 }, { "grad_norm": 0.19633795320987701, "learning_rate": 9.717423839719574e-05, "loss": 0.024, "step": 9130 }, { "grad_norm": 0.17154338955879211, "learning_rate": 9.71650981296392e-05, "loss": 0.0245, "step": 9140 }, { "grad_norm": 0.18777093291282654, "learning_rate": 9.715594353457118e-05, "loss": 0.0268, "step": 9150 }, { "grad_norm": 0.21173278987407684, "learning_rate": 9.714677461477257e-05, "loss": 0.0228, "step": 9160 }, { "grad_norm": 0.20722296833992004, "learning_rate": 9.713759137302869e-05, "loss": 0.0251, "step": 9170 }, { "grad_norm": 0.2329133003950119, "learning_rate": 9.712839381212914e-05, "loss": 0.0239, "step": 9180 }, { "grad_norm": 0.20653030276298523, "learning_rate": 9.71191819348679e-05, "loss": 0.0272, "step": 9190 }, { "grad_norm": 0.23651926219463348, "learning_rate": 9.710995574404331e-05, "loss": 0.0256, "step": 9200 }, { "grad_norm": 0.21964240074157715, "learning_rate": 9.710071524245802e-05, "loss": 0.0271, "step": 9210 }, { "grad_norm": 0.2320842146873474, "learning_rate": 9.709146043291906e-05, "loss": 0.0255, "step": 9220 }, { "grad_norm": 0.22971674799919128, "learning_rate": 9.70821913182378e-05, "loss": 0.0259, "step": 9230 }, { "grad_norm": 0.18382683396339417, "learning_rate": 9.707290790122995e-05, "loss": 0.0227, "step": 9240 }, { "grad_norm": 0.2045517861843109, "learning_rate": 9.706361018471557e-05, "loss": 0.0262, "step": 9250 }, { "grad_norm": 0.19605664908885956, "learning_rate": 9.705429817151906e-05, "loss": 0.0263, "step": 9260 }, { "grad_norm": 0.20786935091018677, "learning_rate": 9.704497186446917e-05, "loss": 0.0267, "step": 9270 }, { "grad_norm": 0.2572239339351654, "learning_rate": 9.703563126639896e-05, "loss": 0.0255, "step": 9280 }, { "grad_norm": 0.2394401580095291, "learning_rate": 9.70262763801459e-05, "loss": 0.0245, "step": 9290 }, { "grad_norm": 0.2872943878173828, "learning_rate": 9.701690720855171e-05, "loss": 0.0269, "step": 9300 }, { "grad_norm": 0.21681372821331024, "learning_rate": 9.700752375446253e-05, "loss": 0.0229, "step": 9310 }, { "grad_norm": 0.18413199484348297, "learning_rate": 9.69981260207288e-05, "loss": 0.0225, "step": 9320 }, { "grad_norm": 0.20549488067626953, "learning_rate": 9.698871401020529e-05, "loss": 0.023, "step": 9330 }, { "grad_norm": 0.19364100694656372, "learning_rate": 9.697928772575112e-05, "loss": 0.0286, "step": 9340 }, { "grad_norm": 0.2138899266719818, "learning_rate": 9.696984717022976e-05, "loss": 0.0276, "step": 9350 }, { "grad_norm": 0.17383013665676117, "learning_rate": 9.6960392346509e-05, "loss": 0.0259, "step": 9360 }, { "grad_norm": 0.1857437938451767, "learning_rate": 9.695092325746097e-05, "loss": 0.0257, "step": 9370 }, { "grad_norm": 0.19300729036331177, "learning_rate": 9.694143990596211e-05, "loss": 0.0247, "step": 9380 }, { "grad_norm": 0.27623745799064636, "learning_rate": 9.693194229489325e-05, "loss": 0.0254, "step": 9390 }, { "grad_norm": 0.21323850750923157, "learning_rate": 9.692243042713944e-05, "loss": 0.0255, "step": 9400 }, { "grad_norm": 0.17742779850959778, "learning_rate": 9.691290430559022e-05, "loss": 0.0258, "step": 9410 }, { "grad_norm": 0.19076861441135406, "learning_rate": 9.690336393313932e-05, "loss": 0.0247, "step": 9420 }, { "grad_norm": 0.16451215744018555, "learning_rate": 9.689380931268487e-05, "loss": 0.023, "step": 9430 }, { "grad_norm": 0.20674556493759155, "learning_rate": 9.688424044712932e-05, "loss": 0.0248, "step": 9440 }, { "grad_norm": 0.2186637669801712, "learning_rate": 9.687465733937942e-05, "loss": 0.0248, "step": 9450 }, { "grad_norm": 0.20830538868904114, "learning_rate": 9.686505999234627e-05, "loss": 0.0267, "step": 9460 }, { "grad_norm": 0.22493033111095428, "learning_rate": 9.685544840894529e-05, "loss": 0.0249, "step": 9470 }, { "grad_norm": 0.26134350895881653, "learning_rate": 9.684582259209624e-05, "loss": 0.0243, "step": 9480 }, { "grad_norm": 0.19404476881027222, "learning_rate": 9.683618254472317e-05, "loss": 0.0243, "step": 9490 }, { "grad_norm": 0.20572520792484283, "learning_rate": 9.682652826975449e-05, "loss": 0.0209, "step": 9500 }, { "grad_norm": 0.24530282616615295, "learning_rate": 9.681685977012291e-05, "loss": 0.0241, "step": 9510 }, { "grad_norm": 0.20027892291545868, "learning_rate": 9.680717704876546e-05, "loss": 0.0246, "step": 9520 }, { "grad_norm": 0.26983577013015747, "learning_rate": 9.679748010862349e-05, "loss": 0.0224, "step": 9530 }, { "grad_norm": 0.2666393518447876, "learning_rate": 9.678776895264267e-05, "loss": 0.0246, "step": 9540 }, { "grad_norm": 0.21481679379940033, "learning_rate": 9.6778043583773e-05, "loss": 0.023, "step": 9550 }, { "grad_norm": 0.18214459717273712, "learning_rate": 9.67683040049688e-05, "loss": 0.022, "step": 9560 }, { "grad_norm": 0.19719089567661285, "learning_rate": 9.675855021918869e-05, "loss": 0.0243, "step": 9570 }, { "grad_norm": 0.2184448093175888, "learning_rate": 9.674878222939561e-05, "loss": 0.0245, "step": 9580 }, { "grad_norm": 0.21128423511981964, "learning_rate": 9.673900003855681e-05, "loss": 0.0252, "step": 9590 }, { "grad_norm": 0.2768596410751343, "learning_rate": 9.672920364964389e-05, "loss": 0.0255, "step": 9600 }, { "grad_norm": 0.2514042854309082, "learning_rate": 9.671939306563269e-05, "loss": 0.0256, "step": 9610 }, { "grad_norm": 0.27891525626182556, "learning_rate": 9.670956828950345e-05, "loss": 0.0269, "step": 9620 }, { "grad_norm": 0.24384701251983643, "learning_rate": 9.669972932424065e-05, "loss": 0.026, "step": 9630 }, { "grad_norm": 0.2860582768917084, "learning_rate": 9.668987617283312e-05, "loss": 0.0252, "step": 9640 }, { "grad_norm": 0.2250538170337677, "learning_rate": 9.668000883827397e-05, "loss": 0.0238, "step": 9650 }, { "grad_norm": 0.17937421798706055, "learning_rate": 9.667012732356067e-05, "loss": 0.0234, "step": 9660 }, { "grad_norm": 0.21082086861133575, "learning_rate": 9.666023163169493e-05, "loss": 0.0248, "step": 9670 }, { "grad_norm": 0.17801125347614288, "learning_rate": 9.665032176568281e-05, "loss": 0.0248, "step": 9680 }, { "grad_norm": 0.18104681372642517, "learning_rate": 9.664039772853469e-05, "loss": 0.0249, "step": 9690 }, { "grad_norm": 0.2184102088212967, "learning_rate": 9.663045952326518e-05, "loss": 0.0249, "step": 9700 }, { "grad_norm": 0.2059715837240219, "learning_rate": 9.662050715289328e-05, "loss": 0.0244, "step": 9710 }, { "grad_norm": 0.2436496764421463, "learning_rate": 9.661054062044226e-05, "loss": 0.0245, "step": 9720 }, { "grad_norm": 0.21226784586906433, "learning_rate": 9.660055992893968e-05, "loss": 0.0239, "step": 9730 }, { "grad_norm": 0.16355401277542114, "learning_rate": 9.659056508141739e-05, "loss": 0.028, "step": 9740 }, { "grad_norm": 0.22865939140319824, "learning_rate": 9.658055608091161e-05, "loss": 0.0258, "step": 9750 }, { "grad_norm": 0.191495880484581, "learning_rate": 9.657053293046276e-05, "loss": 0.0247, "step": 9760 }, { "grad_norm": 0.20319706201553345, "learning_rate": 9.656049563311564e-05, "loss": 0.0241, "step": 9770 }, { "grad_norm": 0.18627376854419708, "learning_rate": 9.655044419191929e-05, "loss": 0.026, "step": 9780 }, { "grad_norm": 0.19765964150428772, "learning_rate": 9.654037860992711e-05, "loss": 0.0258, "step": 9790 }, { "grad_norm": 0.26249754428863525, "learning_rate": 9.653029889019672e-05, "loss": 0.0215, "step": 9800 }, { "grad_norm": 0.2472931295633316, "learning_rate": 9.65202050357901e-05, "loss": 0.0221, "step": 9810 }, { "grad_norm": 0.15659360587596893, "learning_rate": 9.651009704977347e-05, "loss": 0.0221, "step": 9820 }, { "grad_norm": 0.2030465304851532, "learning_rate": 9.649997493521738e-05, "loss": 0.0226, "step": 9830 }, { "grad_norm": 0.21079076826572418, "learning_rate": 9.64898386951967e-05, "loss": 0.0229, "step": 9840 }, { "grad_norm": 0.19143669307231903, "learning_rate": 9.647968833279049e-05, "loss": 0.0264, "step": 9850 }, { "grad_norm": 0.22624020278453827, "learning_rate": 9.646952385108218e-05, "loss": 0.024, "step": 9860 }, { "grad_norm": 0.21752111613750458, "learning_rate": 9.645934525315951e-05, "loss": 0.0254, "step": 9870 }, { "grad_norm": 0.18393002450466156, "learning_rate": 9.644915254211442e-05, "loss": 0.0259, "step": 9880 }, { "grad_norm": 0.26040682196617126, "learning_rate": 9.643894572104321e-05, "loss": 0.0262, "step": 9890 }, { "grad_norm": 0.21727989614009857, "learning_rate": 9.642872479304644e-05, "loss": 0.0243, "step": 9900 }, { "grad_norm": 0.2519247531890869, "learning_rate": 9.641848976122895e-05, "loss": 0.0242, "step": 9910 }, { "grad_norm": 0.19745922088623047, "learning_rate": 9.64082406286999e-05, "loss": 0.0256, "step": 9920 }, { "grad_norm": 0.17450639605522156, "learning_rate": 9.639797739857269e-05, "loss": 0.0262, "step": 9930 }, { "grad_norm": 0.23589584231376648, "learning_rate": 9.638770007396498e-05, "loss": 0.0244, "step": 9940 }, { "grad_norm": 0.15132682025432587, "learning_rate": 9.63774086579988e-05, "loss": 0.025, "step": 9950 }, { "grad_norm": 0.23773826658725739, "learning_rate": 9.63671031538004e-05, "loss": 0.0234, "step": 9960 }, { "grad_norm": 0.2097502201795578, "learning_rate": 9.635678356450031e-05, "loss": 0.0218, "step": 9970 }, { "grad_norm": 0.19507254660129547, "learning_rate": 9.634644989323336e-05, "loss": 0.0236, "step": 9980 }, { "grad_norm": 0.1822517067193985, "learning_rate": 9.633610214313861e-05, "loss": 0.0236, "step": 9990 }, { "grad_norm": 0.2621082067489624, "learning_rate": 9.632574031735951e-05, "loss": 0.0236, "step": 10000 }, { "grad_norm": 0.22120538353919983, "learning_rate": 9.631536441904364e-05, "loss": 0.023, "step": 10010 }, { "grad_norm": 0.29765114188194275, "learning_rate": 9.630497445134293e-05, "loss": 0.0243, "step": 10020 }, { "grad_norm": 0.22464990615844727, "learning_rate": 9.62945704174136e-05, "loss": 0.0236, "step": 10030 }, { "grad_norm": 0.1848776936531067, "learning_rate": 9.628415232041612e-05, "loss": 0.0227, "step": 10040 }, { "grad_norm": 0.22639726102352142, "learning_rate": 9.627372016351524e-05, "loss": 0.0235, "step": 10050 }, { "grad_norm": 0.21659281849861145, "learning_rate": 9.626327394987995e-05, "loss": 0.0231, "step": 10060 }, { "grad_norm": 0.24341022968292236, "learning_rate": 9.625281368268355e-05, "loss": 0.0241, "step": 10070 }, { "grad_norm": 0.24536924064159393, "learning_rate": 9.624233936510357e-05, "loss": 0.0261, "step": 10080 }, { "grad_norm": 0.16954895853996277, "learning_rate": 9.623185100032187e-05, "loss": 0.0257, "step": 10090 }, { "grad_norm": 0.2304326742887497, "learning_rate": 9.62213485915245e-05, "loss": 0.0267, "step": 10100 }, { "grad_norm": 0.21295581758022308, "learning_rate": 9.621083214190186e-05, "loss": 0.0247, "step": 10110 }, { "grad_norm": 0.23144371807575226, "learning_rate": 9.62003016546485e-05, "loss": 0.0241, "step": 10120 }, { "grad_norm": 0.15858319401741028, "learning_rate": 9.618975713296339e-05, "loss": 0.0219, "step": 10130 }, { "grad_norm": 0.16969725489616394, "learning_rate": 9.61791985800496e-05, "loss": 0.0234, "step": 10140 }, { "grad_norm": 0.21040590107440948, "learning_rate": 9.616862599911458e-05, "loss": 0.0207, "step": 10150 }, { "grad_norm": 0.22385390102863312, "learning_rate": 9.615803939337e-05, "loss": 0.0226, "step": 10160 }, { "grad_norm": 0.276739239692688, "learning_rate": 9.614743876603178e-05, "loss": 0.0246, "step": 10170 }, { "grad_norm": 0.21306945383548737, "learning_rate": 9.613682412032013e-05, "loss": 0.0243, "step": 10180 }, { "grad_norm": 0.2325744479894638, "learning_rate": 9.612619545945947e-05, "loss": 0.0218, "step": 10190 }, { "grad_norm": 0.1973913311958313, "learning_rate": 9.611555278667852e-05, "loss": 0.0234, "step": 10200 }, { "grad_norm": 0.15424209833145142, "learning_rate": 9.610489610521024e-05, "loss": 0.0231, "step": 10210 }, { "grad_norm": 0.235724076628685, "learning_rate": 9.609422541829187e-05, "loss": 0.0249, "step": 10220 }, { "grad_norm": 0.22412417829036713, "learning_rate": 9.608354072916486e-05, "loss": 0.0233, "step": 10230 }, { "grad_norm": 0.2348853200674057, "learning_rate": 9.607284204107493e-05, "loss": 0.025, "step": 10240 }, { "grad_norm": 0.18065209686756134, "learning_rate": 9.606212935727208e-05, "loss": 0.0235, "step": 10250 }, { "grad_norm": 0.2132440209388733, "learning_rate": 9.605140268101052e-05, "loss": 0.0242, "step": 10260 }, { "grad_norm": 0.18179506063461304, "learning_rate": 9.604066201554875e-05, "loss": 0.0234, "step": 10270 }, { "grad_norm": 0.2003496140241623, "learning_rate": 9.60299073641495e-05, "loss": 0.0223, "step": 10280 }, { "grad_norm": 0.18924939632415771, "learning_rate": 9.601913873007974e-05, "loss": 0.0236, "step": 10290 }, { "grad_norm": 0.24431423842906952, "learning_rate": 9.60083561166107e-05, "loss": 0.0227, "step": 10300 }, { "grad_norm": 0.2394518107175827, "learning_rate": 9.599755952701783e-05, "loss": 0.0241, "step": 10310 }, { "grad_norm": 0.19605711102485657, "learning_rate": 9.598674896458089e-05, "loss": 0.021, "step": 10320 }, { "grad_norm": 0.22973045706748962, "learning_rate": 9.597592443258383e-05, "loss": 0.0244, "step": 10330 }, { "grad_norm": 0.1940940022468567, "learning_rate": 9.596508593431483e-05, "loss": 0.0231, "step": 10340 }, { "grad_norm": 0.19747428596019745, "learning_rate": 9.59542334730664e-05, "loss": 0.0235, "step": 10350 }, { "grad_norm": 0.20239882171154022, "learning_rate": 9.594336705213516e-05, "loss": 0.0232, "step": 10360 }, { "grad_norm": 0.22723989188671112, "learning_rate": 9.593248667482208e-05, "loss": 0.0225, "step": 10370 }, { "grad_norm": 0.22701680660247803, "learning_rate": 9.592159234443233e-05, "loss": 0.0211, "step": 10380 }, { "grad_norm": 0.20151005685329437, "learning_rate": 9.59106840642753e-05, "loss": 0.0243, "step": 10390 }, { "grad_norm": 0.1859389990568161, "learning_rate": 9.589976183766467e-05, "loss": 0.025, "step": 10400 }, { "grad_norm": 0.26498833298683167, "learning_rate": 9.58888256679183e-05, "loss": 0.0231, "step": 10410 }, { "grad_norm": 0.22523465752601624, "learning_rate": 9.587787555835832e-05, "loss": 0.0217, "step": 10420 }, { "grad_norm": 0.21564655005931854, "learning_rate": 9.586691151231107e-05, "loss": 0.0209, "step": 10430 }, { "grad_norm": 0.24744805693626404, "learning_rate": 9.585593353310715e-05, "loss": 0.0221, "step": 10440 }, { "grad_norm": 0.14762556552886963, "learning_rate": 9.58449416240814e-05, "loss": 0.0219, "step": 10450 }, { "grad_norm": 0.2562410831451416, "learning_rate": 9.583393578857283e-05, "loss": 0.0233, "step": 10460 }, { "grad_norm": 0.20244388282299042, "learning_rate": 9.582291602992474e-05, "loss": 0.0226, "step": 10470 }, { "grad_norm": 0.30276015400886536, "learning_rate": 9.581188235148466e-05, "loss": 0.0229, "step": 10480 }, { "grad_norm": 0.2569456994533539, "learning_rate": 9.58008347566043e-05, "loss": 0.0246, "step": 10490 }, { "grad_norm": 0.22443097829818726, "learning_rate": 9.578977324863965e-05, "loss": 0.0279, "step": 10500 }, { "grad_norm": 0.18456150591373444, "learning_rate": 9.577869783095089e-05, "loss": 0.0207, "step": 10510 }, { "grad_norm": 0.18784189224243164, "learning_rate": 9.576760850690245e-05, "loss": 0.0225, "step": 10520 }, { "grad_norm": 0.22759097814559937, "learning_rate": 9.575650527986298e-05, "loss": 0.0248, "step": 10530 }, { "grad_norm": 0.2248920500278473, "learning_rate": 9.574538815320531e-05, "loss": 0.0255, "step": 10540 }, { "grad_norm": 0.20510439574718475, "learning_rate": 9.573425713030656e-05, "loss": 0.0236, "step": 10550 }, { "grad_norm": 0.198405921459198, "learning_rate": 9.572311221454806e-05, "loss": 0.0231, "step": 10560 }, { "grad_norm": 0.2072466015815735, "learning_rate": 9.57119534093153e-05, "loss": 0.0229, "step": 10570 }, { "grad_norm": 0.2070150524377823, "learning_rate": 9.570078071799806e-05, "loss": 0.0217, "step": 10580 }, { "grad_norm": 0.20615889132022858, "learning_rate": 9.568959414399028e-05, "loss": 0.0234, "step": 10590 }, { "grad_norm": 0.20734259486198425, "learning_rate": 9.567839369069018e-05, "loss": 0.0224, "step": 10600 }, { "grad_norm": 0.22603845596313477, "learning_rate": 9.566717936150013e-05, "loss": 0.022, "step": 10610 }, { "grad_norm": 0.2668154239654541, "learning_rate": 9.565595115982678e-05, "loss": 0.0253, "step": 10620 }, { "grad_norm": 0.20744670927524567, "learning_rate": 9.564470908908094e-05, "loss": 0.0246, "step": 10630 }, { "grad_norm": 0.24253258109092712, "learning_rate": 9.563345315267764e-05, "loss": 0.0255, "step": 10640 }, { "grad_norm": 0.21072982251644135, "learning_rate": 9.562218335403616e-05, "loss": 0.0257, "step": 10650 }, { "grad_norm": 0.22287288308143616, "learning_rate": 9.561089969657999e-05, "loss": 0.0217, "step": 10660 }, { "grad_norm": 0.2206498682498932, "learning_rate": 9.559960218373673e-05, "loss": 0.0224, "step": 10670 }, { "grad_norm": 0.22690285742282867, "learning_rate": 9.558829081893836e-05, "loss": 0.0247, "step": 10680 }, { "grad_norm": 0.19857670366764069, "learning_rate": 9.55769656056209e-05, "loss": 0.0239, "step": 10690 }, { "grad_norm": 0.23083597421646118, "learning_rate": 9.556562654722469e-05, "loss": 0.024, "step": 10700 }, { "grad_norm": 0.1819840520620346, "learning_rate": 9.555427364719422e-05, "loss": 0.0228, "step": 10710 }, { "grad_norm": 0.24647486209869385, "learning_rate": 9.55429069089782e-05, "loss": 0.0228, "step": 10720 }, { "grad_norm": 0.20032599568367004, "learning_rate": 9.553152633602956e-05, "loss": 0.0235, "step": 10730 }, { "grad_norm": 0.24271705746650696, "learning_rate": 9.552013193180543e-05, "loss": 0.0235, "step": 10740 }, { "grad_norm": 0.21165233850479126, "learning_rate": 9.550872369976707e-05, "loss": 0.0235, "step": 10750 }, { "grad_norm": 0.19811086356639862, "learning_rate": 9.549730164338007e-05, "loss": 0.024, "step": 10760 }, { "grad_norm": 0.15980523824691772, "learning_rate": 9.548586576611408e-05, "loss": 0.0225, "step": 10770 }, { "grad_norm": 0.26575037837028503, "learning_rate": 9.54744160714431e-05, "loss": 0.0207, "step": 10780 }, { "grad_norm": 0.17397673428058624, "learning_rate": 9.546295256284516e-05, "loss": 0.0223, "step": 10790 }, { "grad_norm": 0.25228866934776306, "learning_rate": 9.545147524380265e-05, "loss": 0.0246, "step": 10800 }, { "grad_norm": 0.22187045216560364, "learning_rate": 9.543998411780201e-05, "loss": 0.0241, "step": 10810 }, { "grad_norm": 0.1733245998620987, "learning_rate": 9.542847918833397e-05, "loss": 0.0245, "step": 10820 }, { "grad_norm": 0.1939026117324829, "learning_rate": 9.541696045889343e-05, "loss": 0.0225, "step": 10830 }, { "grad_norm": 0.22879402339458466, "learning_rate": 9.540542793297947e-05, "loss": 0.0239, "step": 10840 }, { "grad_norm": 0.23731470108032227, "learning_rate": 9.539388161409537e-05, "loss": 0.0235, "step": 10850 }, { "grad_norm": 0.22554703056812286, "learning_rate": 9.538232150574857e-05, "loss": 0.0227, "step": 10860 }, { "grad_norm": 0.2639988958835602, "learning_rate": 9.537074761145076e-05, "loss": 0.0222, "step": 10870 }, { "grad_norm": 0.25801172852516174, "learning_rate": 9.535915993471778e-05, "loss": 0.0233, "step": 10880 }, { "grad_norm": 0.22832754254341125, "learning_rate": 9.534755847906964e-05, "loss": 0.0243, "step": 10890 }, { "grad_norm": 0.16066108644008636, "learning_rate": 9.533594324803057e-05, "loss": 0.0241, "step": 10900 }, { "grad_norm": 0.23492644727230072, "learning_rate": 9.532431424512895e-05, "loss": 0.0237, "step": 10910 }, { "grad_norm": 0.15438887476921082, "learning_rate": 9.531267147389741e-05, "loss": 0.0237, "step": 10920 }, { "grad_norm": 0.15217554569244385, "learning_rate": 9.530101493787266e-05, "loss": 0.023, "step": 10930 }, { "grad_norm": 0.23945382237434387, "learning_rate": 9.528934464059571e-05, "loss": 0.0224, "step": 10940 }, { "grad_norm": 0.20542562007904053, "learning_rate": 9.527766058561163e-05, "loss": 0.0218, "step": 10950 }, { "grad_norm": 0.30280596017837524, "learning_rate": 9.526596277646976e-05, "loss": 0.0219, "step": 10960 }, { "grad_norm": 0.23920884728431702, "learning_rate": 9.525425121672358e-05, "loss": 0.0236, "step": 10970 }, { "grad_norm": 0.19906403124332428, "learning_rate": 9.524252590993074e-05, "loss": 0.0221, "step": 10980 }, { "grad_norm": 0.20981965959072113, "learning_rate": 9.523078685965309e-05, "loss": 0.0243, "step": 10990 }, { "grad_norm": 0.18567872047424316, "learning_rate": 9.521903406945664e-05, "loss": 0.0224, "step": 11000 }, { "grad_norm": 0.16702602803707123, "learning_rate": 9.520726754291158e-05, "loss": 0.0233, "step": 11010 }, { "grad_norm": 0.1966935247182846, "learning_rate": 9.519548728359227e-05, "loss": 0.0228, "step": 11020 }, { "grad_norm": 0.20051413774490356, "learning_rate": 9.518369329507726e-05, "loss": 0.0226, "step": 11030 }, { "grad_norm": 0.2119501382112503, "learning_rate": 9.51718855809492e-05, "loss": 0.0224, "step": 11040 }, { "grad_norm": 0.2289399653673172, "learning_rate": 9.516006414479502e-05, "loss": 0.0213, "step": 11050 }, { "grad_norm": 0.1762104034423828, "learning_rate": 9.514822899020572e-05, "loss": 0.0232, "step": 11060 }, { "grad_norm": 0.19655989110469818, "learning_rate": 9.513638012077654e-05, "loss": 0.0225, "step": 11070 }, { "grad_norm": 0.21538858115673065, "learning_rate": 9.512451754010683e-05, "loss": 0.0234, "step": 11080 }, { "grad_norm": 0.2250618189573288, "learning_rate": 9.511264125180013e-05, "loss": 0.0219, "step": 11090 }, { "grad_norm": 0.20006605982780457, "learning_rate": 9.510075125946414e-05, "loss": 0.0227, "step": 11100 }, { "grad_norm": 0.22623956203460693, "learning_rate": 9.508884756671075e-05, "loss": 0.0219, "step": 11110 }, { "grad_norm": 0.2014467865228653, "learning_rate": 9.507693017715596e-05, "loss": 0.023, "step": 11120 }, { "grad_norm": 0.17659781873226166, "learning_rate": 9.506499909441997e-05, "loss": 0.0228, "step": 11130 }, { "grad_norm": 0.23288530111312866, "learning_rate": 9.505305432212713e-05, "loss": 0.022, "step": 11140 }, { "grad_norm": 0.20657971501350403, "learning_rate": 9.504109586390595e-05, "loss": 0.0213, "step": 11150 }, { "grad_norm": 0.23453475534915924, "learning_rate": 9.502912372338908e-05, "loss": 0.0222, "step": 11160 }, { "grad_norm": 0.23364225029945374, "learning_rate": 9.501713790421335e-05, "loss": 0.0241, "step": 11170 }, { "grad_norm": 0.18457402288913727, "learning_rate": 9.500513841001974e-05, "loss": 0.0224, "step": 11180 }, { "grad_norm": 0.15238507091999054, "learning_rate": 9.499312524445336e-05, "loss": 0.0241, "step": 11190 }, { "grad_norm": 0.16876555979251862, "learning_rate": 9.498109841116351e-05, "loss": 0.0234, "step": 11200 }, { "grad_norm": 0.20508114993572235, "learning_rate": 9.496905791380363e-05, "loss": 0.0238, "step": 11210 }, { "grad_norm": 0.18755832314491272, "learning_rate": 9.495700375603129e-05, "loss": 0.0204, "step": 11220 }, { "grad_norm": 0.2162061482667923, "learning_rate": 9.494493594150822e-05, "loss": 0.0271, "step": 11230 }, { "grad_norm": 0.23550331592559814, "learning_rate": 9.493285447390032e-05, "loss": 0.0237, "step": 11240 }, { "grad_norm": 0.17017552256584167, "learning_rate": 9.492075935687761e-05, "loss": 0.0227, "step": 11250 }, { "grad_norm": 0.19242319464683533, "learning_rate": 9.490865059411427e-05, "loss": 0.021, "step": 11260 }, { "grad_norm": 0.19795194268226624, "learning_rate": 9.489652818928863e-05, "loss": 0.0229, "step": 11270 }, { "grad_norm": 0.1895112842321396, "learning_rate": 9.488439214608315e-05, "loss": 0.0201, "step": 11280 }, { "grad_norm": 0.16596272587776184, "learning_rate": 9.487224246818444e-05, "loss": 0.0203, "step": 11290 }, { "grad_norm": 0.17833520472049713, "learning_rate": 9.486007915928325e-05, "loss": 0.0217, "step": 11300 }, { "grad_norm": 0.2037050575017929, "learning_rate": 9.484790222307448e-05, "loss": 0.0214, "step": 11310 }, { "grad_norm": 0.20408669114112854, "learning_rate": 9.483571166325716e-05, "loss": 0.0204, "step": 11320 }, { "grad_norm": 0.17844931781291962, "learning_rate": 9.482350748353444e-05, "loss": 0.0212, "step": 11330 }, { "grad_norm": 0.14081020653247833, "learning_rate": 9.481128968761363e-05, "loss": 0.0224, "step": 11340 }, { "grad_norm": 0.2772957682609558, "learning_rate": 9.479905827920621e-05, "loss": 0.0244, "step": 11350 }, { "grad_norm": 0.17641322314739227, "learning_rate": 9.478681326202773e-05, "loss": 0.0216, "step": 11360 }, { "grad_norm": 0.21517391502857208, "learning_rate": 9.477455463979791e-05, "loss": 0.0231, "step": 11370 }, { "grad_norm": 0.188696950674057, "learning_rate": 9.476228241624059e-05, "loss": 0.0225, "step": 11380 }, { "grad_norm": 0.1795019656419754, "learning_rate": 9.474999659508374e-05, "loss": 0.0215, "step": 11390 }, { "grad_norm": 0.2513219714164734, "learning_rate": 9.47376971800595e-05, "loss": 0.0231, "step": 11400 }, { "grad_norm": 0.16435174643993378, "learning_rate": 9.472538417490409e-05, "loss": 0.0228, "step": 11410 }, { "grad_norm": 0.1879328489303589, "learning_rate": 9.471305758335784e-05, "loss": 0.0224, "step": 11420 }, { "grad_norm": 0.2365105003118515, "learning_rate": 9.47007174091653e-05, "loss": 0.0229, "step": 11430 }, { "grad_norm": 0.23260411620140076, "learning_rate": 9.468836365607507e-05, "loss": 0.0226, "step": 11440 }, { "grad_norm": 0.14718595147132874, "learning_rate": 9.467599632783988e-05, "loss": 0.0224, "step": 11450 }, { "grad_norm": 0.14891362190246582, "learning_rate": 9.466361542821662e-05, "loss": 0.0241, "step": 11460 }, { "grad_norm": 0.18281899392604828, "learning_rate": 9.465122096096625e-05, "loss": 0.0228, "step": 11470 }, { "grad_norm": 0.23442664742469788, "learning_rate": 9.463881292985391e-05, "loss": 0.0219, "step": 11480 }, { "grad_norm": 0.20522290468215942, "learning_rate": 9.462639133864881e-05, "loss": 0.0226, "step": 11490 }, { "grad_norm": 0.2007414996623993, "learning_rate": 9.461395619112432e-05, "loss": 0.0213, "step": 11500 }, { "grad_norm": 0.20403718948364258, "learning_rate": 9.460150749105791e-05, "loss": 0.0225, "step": 11510 }, { "grad_norm": 0.23522843420505524, "learning_rate": 9.458904524223116e-05, "loss": 0.0227, "step": 11520 }, { "grad_norm": 0.20109832286834717, "learning_rate": 9.457656944842976e-05, "loss": 0.0268, "step": 11530 }, { "grad_norm": 0.22612975537776947, "learning_rate": 9.456408011344353e-05, "loss": 0.0288, "step": 11540 }, { "grad_norm": 0.25399428606033325, "learning_rate": 9.455157724106643e-05, "loss": 0.0253, "step": 11550 }, { "grad_norm": 0.21062739193439484, "learning_rate": 9.453906083509647e-05, "loss": 0.025, "step": 11560 }, { "grad_norm": 0.19980891048908234, "learning_rate": 9.45265308993358e-05, "loss": 0.0233, "step": 11570 }, { "grad_norm": 0.1985275149345398, "learning_rate": 9.451398743759071e-05, "loss": 0.0243, "step": 11580 }, { "grad_norm": 0.2316037267446518, "learning_rate": 9.450143045367156e-05, "loss": 0.0232, "step": 11590 }, { "grad_norm": 0.22844737768173218, "learning_rate": 9.448885995139283e-05, "loss": 0.0217, "step": 11600 }, { "grad_norm": 0.26093271374702454, "learning_rate": 9.44762759345731e-05, "loss": 0.0236, "step": 11610 }, { "grad_norm": 0.1994665563106537, "learning_rate": 9.446367840703509e-05, "loss": 0.0259, "step": 11620 }, { "grad_norm": 0.2980620563030243, "learning_rate": 9.445106737260556e-05, "loss": 0.0268, "step": 11630 }, { "grad_norm": 0.3013952076435089, "learning_rate": 9.443844283511543e-05, "loss": 0.0249, "step": 11640 }, { "grad_norm": 0.2278793603181839, "learning_rate": 9.442580479839968e-05, "loss": 0.0243, "step": 11650 }, { "grad_norm": 0.20143979787826538, "learning_rate": 9.441315326629745e-05, "loss": 0.0227, "step": 11660 }, { "grad_norm": 0.22047773003578186, "learning_rate": 9.44004882426519e-05, "loss": 0.0227, "step": 11670 }, { "grad_norm": 0.19129633903503418, "learning_rate": 9.438780973131037e-05, "loss": 0.0216, "step": 11680 }, { "grad_norm": 0.19554078578948975, "learning_rate": 9.437511773612423e-05, "loss": 0.0226, "step": 11690 }, { "grad_norm": 0.18456418812274933, "learning_rate": 9.436241226094896e-05, "loss": 0.0242, "step": 11700 }, { "grad_norm": 0.22191280126571655, "learning_rate": 9.434969330964418e-05, "loss": 0.0226, "step": 11710 }, { "grad_norm": 0.2635849714279175, "learning_rate": 9.433696088607356e-05, "loss": 0.022, "step": 11720 }, { "grad_norm": 0.2306859791278839, "learning_rate": 9.432421499410486e-05, "loss": 0.0229, "step": 11730 }, { "grad_norm": 0.20622603595256805, "learning_rate": 9.431145563760998e-05, "loss": 0.0226, "step": 11740 }, { "grad_norm": 0.1941065490245819, "learning_rate": 9.429868282046484e-05, "loss": 0.0237, "step": 11750 }, { "grad_norm": 0.1654898226261139, "learning_rate": 9.428589654654951e-05, "loss": 0.022, "step": 11760 }, { "grad_norm": 0.16515076160430908, "learning_rate": 9.42730968197481e-05, "loss": 0.0208, "step": 11770 }, { "grad_norm": 0.1725461333990097, "learning_rate": 9.426028364394883e-05, "loss": 0.0193, "step": 11780 }, { "grad_norm": 0.19497522711753845, "learning_rate": 9.424745702304402e-05, "loss": 0.0213, "step": 11790 }, { "grad_norm": 0.1991053968667984, "learning_rate": 9.423461696093006e-05, "loss": 0.0219, "step": 11800 }, { "grad_norm": 0.15381337702274323, "learning_rate": 9.422176346150741e-05, "loss": 0.022, "step": 11810 }, { "grad_norm": 0.1840803623199463, "learning_rate": 9.420889652868063e-05, "loss": 0.0224, "step": 11820 }, { "grad_norm": 0.18339043855667114, "learning_rate": 9.419601616635836e-05, "loss": 0.0221, "step": 11830 }, { "grad_norm": 0.22838932275772095, "learning_rate": 9.418312237845331e-05, "loss": 0.0218, "step": 11840 }, { "grad_norm": 0.20200853049755096, "learning_rate": 9.417021516888225e-05, "loss": 0.0223, "step": 11850 }, { "grad_norm": 0.19280128180980682, "learning_rate": 9.415729454156608e-05, "loss": 0.0209, "step": 11860 }, { "grad_norm": 0.17377230525016785, "learning_rate": 9.414436050042973e-05, "loss": 0.0242, "step": 11870 }, { "grad_norm": 0.18575547635555267, "learning_rate": 9.413141304940223e-05, "loss": 0.0227, "step": 11880 }, { "grad_norm": 0.21215969324111938, "learning_rate": 9.411845219241666e-05, "loss": 0.0216, "step": 11890 }, { "grad_norm": 0.20252998173236847, "learning_rate": 9.410547793341021e-05, "loss": 0.0216, "step": 11900 }, { "grad_norm": 0.20027774572372437, "learning_rate": 9.409249027632408e-05, "loss": 0.0235, "step": 11910 }, { "grad_norm": 0.18770843744277954, "learning_rate": 9.407948922510362e-05, "loss": 0.0226, "step": 11920 }, { "grad_norm": 0.17700974643230438, "learning_rate": 9.406647478369817e-05, "loss": 0.0265, "step": 11930 }, { "grad_norm": 0.20503321290016174, "learning_rate": 9.405344695606118e-05, "loss": 0.0251, "step": 11940 }, { "grad_norm": 0.17017248272895813, "learning_rate": 9.404040574615018e-05, "loss": 0.0219, "step": 11950 }, { "grad_norm": 0.22365881502628326, "learning_rate": 9.402735115792674e-05, "loss": 0.0235, "step": 11960 }, { "grad_norm": 0.18757282197475433, "learning_rate": 9.401428319535649e-05, "loss": 0.0235, "step": 11970 }, { "grad_norm": 0.22226639091968536, "learning_rate": 9.400120186240912e-05, "loss": 0.0237, "step": 11980 }, { "grad_norm": 0.22200770676136017, "learning_rate": 9.398810716305844e-05, "loss": 0.0224, "step": 11990 }, { "grad_norm": 0.1476600617170334, "learning_rate": 9.397499910128222e-05, "loss": 0.0233, "step": 12000 }, { "grad_norm": 0.16704696416854858, "learning_rate": 9.396187768106237e-05, "loss": 0.024, "step": 12010 }, { "grad_norm": 0.23709622025489807, "learning_rate": 9.394874290638482e-05, "loss": 0.0243, "step": 12020 }, { "grad_norm": 0.2517925500869751, "learning_rate": 9.393559478123959e-05, "loss": 0.0238, "step": 12030 }, { "grad_norm": 0.22059540450572968, "learning_rate": 9.39224333096207e-05, "loss": 0.0218, "step": 12040 }, { "grad_norm": 0.21604569256305695, "learning_rate": 9.390925849552629e-05, "loss": 0.0236, "step": 12050 }, { "grad_norm": 0.20057010650634766, "learning_rate": 9.389607034295849e-05, "loss": 0.0237, "step": 12060 }, { "grad_norm": 0.16262216866016388, "learning_rate": 9.388286885592355e-05, "loss": 0.0193, "step": 12070 }, { "grad_norm": 0.15294203162193298, "learning_rate": 9.386965403843168e-05, "loss": 0.0213, "step": 12080 }, { "grad_norm": 0.1813019961118698, "learning_rate": 9.385642589449726e-05, "loss": 0.0208, "step": 12090 }, { "grad_norm": 0.18069206178188324, "learning_rate": 9.38431844281386e-05, "loss": 0.0234, "step": 12100 }, { "grad_norm": 0.19239652156829834, "learning_rate": 9.38299296433781e-05, "loss": 0.0215, "step": 12110 }, { "grad_norm": 0.19542378187179565, "learning_rate": 9.381666154424226e-05, "loss": 0.0223, "step": 12120 }, { "grad_norm": 0.19123722612857819, "learning_rate": 9.380338013476157e-05, "loss": 0.0223, "step": 12130 }, { "grad_norm": 0.20124734938144684, "learning_rate": 9.379008541897054e-05, "loss": 0.0215, "step": 12140 }, { "grad_norm": 0.1485995054244995, "learning_rate": 9.377677740090777e-05, "loss": 0.0211, "step": 12150 }, { "grad_norm": 0.17511525750160217, "learning_rate": 9.376345608461588e-05, "loss": 0.0218, "step": 12160 }, { "grad_norm": 0.2075629085302353, "learning_rate": 9.375012147414155e-05, "loss": 0.0208, "step": 12170 }, { "grad_norm": 0.20952248573303223, "learning_rate": 9.373677357353545e-05, "loss": 0.0223, "step": 12180 }, { "grad_norm": 0.16846400499343872, "learning_rate": 9.372341238685237e-05, "loss": 0.0218, "step": 12190 }, { "grad_norm": 0.16817380487918854, "learning_rate": 9.371003791815102e-05, "loss": 0.024, "step": 12200 }, { "grad_norm": 0.2600792646408081, "learning_rate": 9.369665017149429e-05, "loss": 0.024, "step": 12210 }, { "grad_norm": 0.19537785649299622, "learning_rate": 9.368324915094895e-05, "loss": 0.0204, "step": 12220 }, { "grad_norm": 0.1688484400510788, "learning_rate": 9.366983486058591e-05, "loss": 0.0224, "step": 12230 }, { "grad_norm": 0.14917485415935516, "learning_rate": 9.365640730448009e-05, "loss": 0.0226, "step": 12240 }, { "grad_norm": 0.18543988466262817, "learning_rate": 9.36429664867104e-05, "loss": 0.0206, "step": 12250 }, { "grad_norm": 0.1989470273256302, "learning_rate": 9.362951241135982e-05, "loss": 0.0211, "step": 12260 }, { "grad_norm": 0.1653849184513092, "learning_rate": 9.361604508251534e-05, "loss": 0.024, "step": 12270 }, { "grad_norm": 0.22731025516986847, "learning_rate": 9.360256450426799e-05, "loss": 0.0209, "step": 12280 }, { "grad_norm": 0.2192097306251526, "learning_rate": 9.358907068071279e-05, "loss": 0.0199, "step": 12290 }, { "grad_norm": 0.16190972924232483, "learning_rate": 9.357556361594882e-05, "loss": 0.0229, "step": 12300 }, { "grad_norm": 0.1829512119293213, "learning_rate": 9.356204331407917e-05, "loss": 0.0216, "step": 12310 }, { "grad_norm": 0.2171289324760437, "learning_rate": 9.354850977921094e-05, "loss": 0.02, "step": 12320 }, { "grad_norm": 0.17619696259498596, "learning_rate": 9.353496301545529e-05, "loss": 0.0211, "step": 12330 }, { "grad_norm": 0.19023440778255463, "learning_rate": 9.352140302692733e-05, "loss": 0.0228, "step": 12340 }, { "grad_norm": 0.1756919026374817, "learning_rate": 9.350782981774627e-05, "loss": 0.0235, "step": 12350 }, { "grad_norm": 0.21991142630577087, "learning_rate": 9.349424339203526e-05, "loss": 0.0219, "step": 12360 }, { "grad_norm": 0.17241425812244415, "learning_rate": 9.34806437539215e-05, "loss": 0.0207, "step": 12370 }, { "grad_norm": 0.18971291184425354, "learning_rate": 9.346703090753622e-05, "loss": 0.0205, "step": 12380 }, { "grad_norm": 0.16315272450447083, "learning_rate": 9.345340485701461e-05, "loss": 0.0223, "step": 12390 }, { "grad_norm": 0.22968946397304535, "learning_rate": 9.343976560649595e-05, "loss": 0.0222, "step": 12400 }, { "grad_norm": 0.18043498694896698, "learning_rate": 9.342611316012344e-05, "loss": 0.0194, "step": 12410 }, { "grad_norm": 0.22545841336250305, "learning_rate": 9.341244752204437e-05, "loss": 0.0212, "step": 12420 }, { "grad_norm": 0.168023943901062, "learning_rate": 9.339876869640995e-05, "loss": 0.0237, "step": 12430 }, { "grad_norm": 0.1769919991493225, "learning_rate": 9.33850766873755e-05, "loss": 0.0217, "step": 12440 }, { "grad_norm": 0.21271079778671265, "learning_rate": 9.337137149910028e-05, "loss": 0.021, "step": 12450 }, { "grad_norm": 0.1382354199886322, "learning_rate": 9.335765313574753e-05, "loss": 0.0198, "step": 12460 }, { "grad_norm": 0.1980958729982376, "learning_rate": 9.334392160148457e-05, "loss": 0.0238, "step": 12470 }, { "grad_norm": 0.25242364406585693, "learning_rate": 9.333017690048264e-05, "loss": 0.0219, "step": 12480 }, { "grad_norm": 0.17042601108551025, "learning_rate": 9.331641903691706e-05, "loss": 0.0249, "step": 12490 }, { "grad_norm": 0.20257918536663055, "learning_rate": 9.330264801496707e-05, "loss": 0.0208, "step": 12500 }, { "grad_norm": 0.18695613741874695, "learning_rate": 9.328886383881594e-05, "loss": 0.023, "step": 12510 }, { "grad_norm": 0.21379873156547546, "learning_rate": 9.327506651265095e-05, "loss": 0.0199, "step": 12520 }, { "grad_norm": 0.16410128772258759, "learning_rate": 9.326125604066338e-05, "loss": 0.0221, "step": 12530 }, { "grad_norm": 0.2095990777015686, "learning_rate": 9.324743242704847e-05, "loss": 0.0193, "step": 12540 }, { "grad_norm": 0.23675695061683655, "learning_rate": 9.323359567600546e-05, "loss": 0.0209, "step": 12550 }, { "grad_norm": 0.17854854464530945, "learning_rate": 9.321974579173761e-05, "loss": 0.0244, "step": 12560 }, { "grad_norm": 0.17351028323173523, "learning_rate": 9.320588277845213e-05, "loss": 0.0222, "step": 12570 }, { "grad_norm": 0.19612039625644684, "learning_rate": 9.319200664036026e-05, "loss": 0.0237, "step": 12580 }, { "grad_norm": 0.1560988873243332, "learning_rate": 9.31781173816772e-05, "loss": 0.0211, "step": 12590 }, { "grad_norm": 0.1899024397134781, "learning_rate": 9.316421500662212e-05, "loss": 0.0211, "step": 12600 }, { "grad_norm": 0.18084338307380676, "learning_rate": 9.31502995194182e-05, "loss": 0.0192, "step": 12610 }, { "grad_norm": 0.17258961498737335, "learning_rate": 9.31363709242926e-05, "loss": 0.0201, "step": 12620 }, { "grad_norm": 0.18802444636821747, "learning_rate": 9.312242922547647e-05, "loss": 0.0219, "step": 12630 }, { "grad_norm": 0.15474705398082733, "learning_rate": 9.310847442720492e-05, "loss": 0.0238, "step": 12640 }, { "grad_norm": 0.17870064079761505, "learning_rate": 9.309450653371706e-05, "loss": 0.0195, "step": 12650 }, { "grad_norm": 0.1603943258523941, "learning_rate": 9.308052554925595e-05, "loss": 0.0213, "step": 12660 }, { "grad_norm": 0.16342006623744965, "learning_rate": 9.306653147806867e-05, "loss": 0.0207, "step": 12670 }, { "grad_norm": 0.19775249063968658, "learning_rate": 9.305252432440622e-05, "loss": 0.02, "step": 12680 }, { "grad_norm": 0.18262609839439392, "learning_rate": 9.303850409252361e-05, "loss": 0.0231, "step": 12690 }, { "grad_norm": 0.16530878841876984, "learning_rate": 9.302447078667985e-05, "loss": 0.0201, "step": 12700 }, { "grad_norm": 0.1626727133989334, "learning_rate": 9.301042441113783e-05, "loss": 0.0203, "step": 12710 }, { "grad_norm": 0.17158496379852295, "learning_rate": 9.299636497016451e-05, "loss": 0.0245, "step": 12720 }, { "grad_norm": 0.13333721458911896, "learning_rate": 9.298229246803076e-05, "loss": 0.0229, "step": 12730 }, { "grad_norm": 0.18029141426086426, "learning_rate": 9.296820690901144e-05, "loss": 0.0205, "step": 12740 }, { "grad_norm": 0.21378786861896515, "learning_rate": 9.295410829738539e-05, "loss": 0.0205, "step": 12750 }, { "grad_norm": 0.16401782631874084, "learning_rate": 9.293999663743535e-05, "loss": 0.0229, "step": 12760 }, { "grad_norm": 0.19252713024616241, "learning_rate": 9.292587193344813e-05, "loss": 0.0232, "step": 12770 }, { "grad_norm": 0.17664219439029694, "learning_rate": 9.291173418971437e-05, "loss": 0.0224, "step": 12780 }, { "grad_norm": 0.23638670146465302, "learning_rate": 9.28975834105288e-05, "loss": 0.0216, "step": 12790 }, { "grad_norm": 0.1937420517206192, "learning_rate": 9.288341960019004e-05, "loss": 0.0233, "step": 12800 }, { "grad_norm": 0.16753357648849487, "learning_rate": 9.286924276300067e-05, "loss": 0.0223, "step": 12810 }, { "grad_norm": 0.16836613416671753, "learning_rate": 9.285505290326726e-05, "loss": 0.0222, "step": 12820 }, { "grad_norm": 0.23786278069019318, "learning_rate": 9.284085002530027e-05, "loss": 0.0222, "step": 12830 }, { "grad_norm": 0.18798497319221497, "learning_rate": 9.282663413341422e-05, "loss": 0.0218, "step": 12840 }, { "grad_norm": 0.2114873230457306, "learning_rate": 9.281240523192747e-05, "loss": 0.0235, "step": 12850 }, { "grad_norm": 0.18666036427021027, "learning_rate": 9.279816332516242e-05, "loss": 0.0216, "step": 12860 }, { "grad_norm": 0.21087005734443665, "learning_rate": 9.278390841744536e-05, "loss": 0.0228, "step": 12870 }, { "grad_norm": 0.18436792492866516, "learning_rate": 9.276964051310658e-05, "loss": 0.0209, "step": 12880 }, { "grad_norm": 0.24088819324970245, "learning_rate": 9.275535961648027e-05, "loss": 0.022, "step": 12890 }, { "grad_norm": 0.23252959549427032, "learning_rate": 9.274106573190459e-05, "loss": 0.0205, "step": 12900 }, { "grad_norm": 0.21083371341228485, "learning_rate": 9.272675886372168e-05, "loss": 0.0197, "step": 12910 }, { "grad_norm": 0.19365862011909485, "learning_rate": 9.271243901627754e-05, "loss": 0.0226, "step": 12920 }, { "grad_norm": 0.2163909375667572, "learning_rate": 9.269810619392219e-05, "loss": 0.0235, "step": 12930 }, { "grad_norm": 0.20691226422786713, "learning_rate": 9.268376040100955e-05, "loss": 0.0212, "step": 12940 }, { "grad_norm": 0.1537151038646698, "learning_rate": 9.266940164189752e-05, "loss": 0.0209, "step": 12950 }, { "grad_norm": 0.21854056417942047, "learning_rate": 9.265502992094787e-05, "loss": 0.0215, "step": 12960 }, { "grad_norm": 0.2511041760444641, "learning_rate": 9.264064524252638e-05, "loss": 0.0204, "step": 12970 }, { "grad_norm": 0.18629005551338196, "learning_rate": 9.262624761100271e-05, "loss": 0.0206, "step": 12980 }, { "grad_norm": 0.17011821269989014, "learning_rate": 9.261183703075051e-05, "loss": 0.0206, "step": 12990 }, { "grad_norm": 0.21689336001873016, "learning_rate": 9.259741350614733e-05, "loss": 0.0228, "step": 13000 }, { "grad_norm": 0.19618691504001617, "learning_rate": 9.258297704157464e-05, "loss": 0.0203, "step": 13010 }, { "grad_norm": 0.24279211461544037, "learning_rate": 9.256852764141786e-05, "loss": 0.0211, "step": 13020 }, { "grad_norm": 0.14758485555648804, "learning_rate": 9.255406531006634e-05, "loss": 0.0206, "step": 13030 }, { "grad_norm": 0.17723466455936432, "learning_rate": 9.253959005191335e-05, "loss": 0.0206, "step": 13040 }, { "grad_norm": 0.20843887329101562, "learning_rate": 9.25251018713561e-05, "loss": 0.022, "step": 13050 }, { "grad_norm": 0.20984332263469696, "learning_rate": 9.251060077279571e-05, "loss": 0.0207, "step": 13060 }, { "grad_norm": 0.18230381608009338, "learning_rate": 9.249608676063724e-05, "loss": 0.0202, "step": 13070 }, { "grad_norm": 0.21951518952846527, "learning_rate": 9.248155983928964e-05, "loss": 0.0228, "step": 13080 }, { "grad_norm": 0.1923452615737915, "learning_rate": 9.246702001316583e-05, "loss": 0.0222, "step": 13090 }, { "grad_norm": 0.2526964545249939, "learning_rate": 9.245246728668262e-05, "loss": 0.0206, "step": 13100 }, { "grad_norm": 0.15881836414337158, "learning_rate": 9.243790166426073e-05, "loss": 0.0221, "step": 13110 }, { "grad_norm": 0.16374972462654114, "learning_rate": 9.242332315032484e-05, "loss": 0.0209, "step": 13120 }, { "grad_norm": 0.16730506718158722, "learning_rate": 9.240873174930349e-05, "loss": 0.0211, "step": 13130 }, { "grad_norm": 0.17086216807365417, "learning_rate": 9.239412746562917e-05, "loss": 0.0203, "step": 13140 }, { "grad_norm": 0.17622530460357666, "learning_rate": 9.237951030373828e-05, "loss": 0.0208, "step": 13150 }, { "grad_norm": 0.13874153792858124, "learning_rate": 9.236488026807113e-05, "loss": 0.0201, "step": 13160 }, { "grad_norm": 0.21005238592624664, "learning_rate": 9.235023736307193e-05, "loss": 0.0209, "step": 13170 }, { "grad_norm": 0.2092410922050476, "learning_rate": 9.233558159318881e-05, "loss": 0.0206, "step": 13180 }, { "grad_norm": 0.20777270197868347, "learning_rate": 9.232091296287382e-05, "loss": 0.0228, "step": 13190 }, { "grad_norm": 0.19073449075222015, "learning_rate": 9.230623147658288e-05, "loss": 0.0186, "step": 13200 }, { "grad_norm": 0.18911026418209076, "learning_rate": 9.229153713877586e-05, "loss": 0.0208, "step": 13210 }, { "grad_norm": 0.16413156688213348, "learning_rate": 9.227682995391649e-05, "loss": 0.0212, "step": 13220 }, { "grad_norm": 0.15851205587387085, "learning_rate": 9.226210992647243e-05, "loss": 0.0214, "step": 13230 }, { "grad_norm": 0.21001698076725006, "learning_rate": 9.224737706091525e-05, "loss": 0.0196, "step": 13240 }, { "grad_norm": 0.20838552713394165, "learning_rate": 9.223263136172039e-05, "loss": 0.0211, "step": 13250 }, { "grad_norm": 0.2243349701166153, "learning_rate": 9.22178728333672e-05, "loss": 0.0226, "step": 13260 }, { "grad_norm": 0.18550893664360046, "learning_rate": 9.220310148033897e-05, "loss": 0.0205, "step": 13270 }, { "grad_norm": 0.16472965478897095, "learning_rate": 9.21883173071228e-05, "loss": 0.0205, "step": 13280 }, { "grad_norm": 0.19648420810699463, "learning_rate": 9.217352031820976e-05, "loss": 0.0224, "step": 13290 }, { "grad_norm": 0.1998908966779709, "learning_rate": 9.215871051809477e-05, "loss": 0.0224, "step": 13300 }, { "grad_norm": 0.18998336791992188, "learning_rate": 9.214388791127666e-05, "loss": 0.0201, "step": 13310 }, { "grad_norm": 0.20774134993553162, "learning_rate": 9.212905250225814e-05, "loss": 0.0222, "step": 13320 }, { "grad_norm": 0.14587517082691193, "learning_rate": 9.211420429554583e-05, "loss": 0.0212, "step": 13330 }, { "grad_norm": 0.18984627723693848, "learning_rate": 9.209934329565022e-05, "loss": 0.0208, "step": 13340 }, { "grad_norm": 0.22099103033542633, "learning_rate": 9.208446950708568e-05, "loss": 0.0202, "step": 13350 }, { "grad_norm": 0.1919621229171753, "learning_rate": 9.20695829343705e-05, "loss": 0.0235, "step": 13360 }, { "grad_norm": 0.1949273943901062, "learning_rate": 9.205468358202678e-05, "loss": 0.0204, "step": 13370 }, { "grad_norm": 0.18176433444023132, "learning_rate": 9.203977145458059e-05, "loss": 0.0213, "step": 13380 }, { "grad_norm": 0.15301965177059174, "learning_rate": 9.202484655656182e-05, "loss": 0.0205, "step": 13390 }, { "grad_norm": 0.21924468874931335, "learning_rate": 9.200990889250427e-05, "loss": 0.0225, "step": 13400 }, { "grad_norm": 0.17362439632415771, "learning_rate": 9.19949584669456e-05, "loss": 0.02, "step": 13410 }, { "grad_norm": 0.1951644867658615, "learning_rate": 9.197999528442738e-05, "loss": 0.0225, "step": 13420 }, { "grad_norm": 0.21179279685020447, "learning_rate": 9.196501934949499e-05, "loss": 0.0232, "step": 13430 }, { "grad_norm": 0.2083195298910141, "learning_rate": 9.195003066669776e-05, "loss": 0.0234, "step": 13440 }, { "grad_norm": 0.2054193913936615, "learning_rate": 9.193502924058884e-05, "loss": 0.0215, "step": 13450 }, { "grad_norm": 0.17548969388008118, "learning_rate": 9.192001507572526e-05, "loss": 0.022, "step": 13460 }, { "grad_norm": 0.17012512683868408, "learning_rate": 9.190498817666793e-05, "loss": 0.0226, "step": 13470 }, { "grad_norm": 0.2605862617492676, "learning_rate": 9.188994854798163e-05, "loss": 0.0216, "step": 13480 }, { "grad_norm": 0.25528469681739807, "learning_rate": 9.187489619423499e-05, "loss": 0.0213, "step": 13490 }, { "grad_norm": 0.16232194006443024, "learning_rate": 9.185983112000056e-05, "loss": 0.0228, "step": 13500 }, { "grad_norm": 0.16952311992645264, "learning_rate": 9.184475332985464e-05, "loss": 0.0216, "step": 13510 }, { "grad_norm": 0.21221880614757538, "learning_rate": 9.182966282837754e-05, "loss": 0.0214, "step": 13520 }, { "grad_norm": 0.19102156162261963, "learning_rate": 9.18145596201533e-05, "loss": 0.0215, "step": 13530 }, { "grad_norm": 0.15706069767475128, "learning_rate": 9.179944370976991e-05, "loss": 0.0214, "step": 13540 }, { "grad_norm": 0.17228557169437408, "learning_rate": 9.178431510181918e-05, "loss": 0.0234, "step": 13550 }, { "grad_norm": 0.195745050907135, "learning_rate": 9.176917380089675e-05, "loss": 0.0205, "step": 13560 }, { "grad_norm": 0.18961197137832642, "learning_rate": 9.175401981160219e-05, "loss": 0.0211, "step": 13570 }, { "grad_norm": 0.22162708640098572, "learning_rate": 9.173885313853885e-05, "loss": 0.0205, "step": 13580 }, { "grad_norm": 0.2289978563785553, "learning_rate": 9.172367378631398e-05, "loss": 0.0234, "step": 13590 }, { "grad_norm": 0.17536041140556335, "learning_rate": 9.170848175953866e-05, "loss": 0.021, "step": 13600 }, { "grad_norm": 0.2121567279100418, "learning_rate": 9.169327706282784e-05, "loss": 0.0201, "step": 13610 }, { "grad_norm": 0.13538363575935364, "learning_rate": 9.167805970080029e-05, "loss": 0.0192, "step": 13620 }, { "grad_norm": 0.17157135903835297, "learning_rate": 9.166282967807864e-05, "loss": 0.0214, "step": 13630 }, { "grad_norm": 0.16220702230930328, "learning_rate": 9.16475869992894e-05, "loss": 0.0199, "step": 13640 }, { "grad_norm": 0.2218475192785263, "learning_rate": 9.163233166906284e-05, "loss": 0.0206, "step": 13650 }, { "grad_norm": 0.1566193699836731, "learning_rate": 9.161706369203317e-05, "loss": 0.0194, "step": 13660 }, { "grad_norm": 0.22172869741916656, "learning_rate": 9.16017830728384e-05, "loss": 0.021, "step": 13670 }, { "grad_norm": 0.13503649830818176, "learning_rate": 9.158648981612035e-05, "loss": 0.022, "step": 13680 }, { "grad_norm": 0.17284207046031952, "learning_rate": 9.157118392652472e-05, "loss": 0.0198, "step": 13690 }, { "grad_norm": 0.19371064007282257, "learning_rate": 9.155586540870104e-05, "loss": 0.0212, "step": 13700 }, { "grad_norm": 0.2375413030385971, "learning_rate": 9.154053426730267e-05, "loss": 0.021, "step": 13710 }, { "grad_norm": 0.1735040843486786, "learning_rate": 9.15251905069868e-05, "loss": 0.0201, "step": 13720 }, { "grad_norm": 0.17579878866672516, "learning_rate": 9.150983413241446e-05, "loss": 0.0223, "step": 13730 }, { "grad_norm": 0.18616656959056854, "learning_rate": 9.149446514825051e-05, "loss": 0.0204, "step": 13740 }, { "grad_norm": 0.21046799421310425, "learning_rate": 9.147908355916365e-05, "loss": 0.0227, "step": 13750 }, { "grad_norm": 0.2113187611103058, "learning_rate": 9.146368936982642e-05, "loss": 0.0211, "step": 13760 }, { "grad_norm": 0.15286517143249512, "learning_rate": 9.144828258491511e-05, "loss": 0.0188, "step": 13770 }, { "grad_norm": 0.2228587120771408, "learning_rate": 9.143286320910996e-05, "loss": 0.0192, "step": 13780 }, { "grad_norm": 0.18112996220588684, "learning_rate": 9.141743124709491e-05, "loss": 0.0206, "step": 13790 }, { "grad_norm": 0.17142711579799652, "learning_rate": 9.140198670355784e-05, "loss": 0.0198, "step": 13800 }, { "grad_norm": 0.22127939760684967, "learning_rate": 9.138652958319034e-05, "loss": 0.0189, "step": 13810 }, { "grad_norm": 0.20492693781852722, "learning_rate": 9.137105989068791e-05, "loss": 0.0205, "step": 13820 }, { "grad_norm": 0.15384820103645325, "learning_rate": 9.135557763074983e-05, "loss": 0.0212, "step": 13830 }, { "grad_norm": 0.16295140981674194, "learning_rate": 9.13400828080792e-05, "loss": 0.0207, "step": 13840 }, { "grad_norm": 0.19692298769950867, "learning_rate": 9.132457542738292e-05, "loss": 0.0197, "step": 13850 }, { "grad_norm": 0.12177105247974396, "learning_rate": 9.130905549337174e-05, "loss": 0.0174, "step": 13860 }, { "grad_norm": 0.16590318083763123, "learning_rate": 9.129352301076021e-05, "loss": 0.0224, "step": 13870 }, { "grad_norm": 0.1472773253917694, "learning_rate": 9.127797798426668e-05, "loss": 0.0203, "step": 13880 }, { "grad_norm": 0.16238150000572205, "learning_rate": 9.126242041861333e-05, "loss": 0.0212, "step": 13890 }, { "grad_norm": 0.21918395161628723, "learning_rate": 9.124685031852611e-05, "loss": 0.0202, "step": 13900 }, { "grad_norm": 0.173313170671463, "learning_rate": 9.123126768873482e-05, "loss": 0.0227, "step": 13910 }, { "grad_norm": 0.2066342830657959, "learning_rate": 9.121567253397308e-05, "loss": 0.0195, "step": 13920 }, { "grad_norm": 0.20048469305038452, "learning_rate": 9.120006485897824e-05, "loss": 0.0191, "step": 13930 }, { "grad_norm": 0.17632928490638733, "learning_rate": 9.118444466849152e-05, "loss": 0.0206, "step": 13940 }, { "grad_norm": 0.16820386052131653, "learning_rate": 9.116881196725793e-05, "loss": 0.02, "step": 13950 }, { "grad_norm": 0.15587937831878662, "learning_rate": 9.115316676002627e-05, "loss": 0.0209, "step": 13960 }, { "grad_norm": 0.21872137486934662, "learning_rate": 9.113750905154911e-05, "loss": 0.0228, "step": 13970 }, { "grad_norm": 0.18950915336608887, "learning_rate": 9.112183884658289e-05, "loss": 0.02, "step": 13980 }, { "grad_norm": 0.211467444896698, "learning_rate": 9.11061561498878e-05, "loss": 0.0199, "step": 13990 }, { "grad_norm": 0.18643061816692352, "learning_rate": 9.109046096622779e-05, "loss": 0.0198, "step": 14000 }, { "grad_norm": 0.2909282445907593, "learning_rate": 9.107475330037069e-05, "loss": 0.0214, "step": 14010 }, { "grad_norm": 0.21855011582374573, "learning_rate": 9.105903315708806e-05, "loss": 0.0214, "step": 14020 }, { "grad_norm": 0.23888733983039856, "learning_rate": 9.104330054115524e-05, "loss": 0.0193, "step": 14030 }, { "grad_norm": 0.1977103352546692, "learning_rate": 9.102755545735141e-05, "loss": 0.019, "step": 14040 }, { "grad_norm": 0.17959746718406677, "learning_rate": 9.10117979104595e-05, "loss": 0.0224, "step": 14050 }, { "grad_norm": 0.1588594615459442, "learning_rate": 9.099602790526624e-05, "loss": 0.0199, "step": 14060 }, { "grad_norm": 0.19006766378879547, "learning_rate": 9.098024544656212e-05, "loss": 0.0186, "step": 14070 }, { "grad_norm": 0.18835924565792084, "learning_rate": 9.096445053914148e-05, "loss": 0.0194, "step": 14080 }, { "grad_norm": 0.16474823653697968, "learning_rate": 9.094864318780236e-05, "loss": 0.0197, "step": 14090 }, { "grad_norm": 0.18655794858932495, "learning_rate": 9.093282339734663e-05, "loss": 0.0201, "step": 14100 }, { "grad_norm": 0.1797541081905365, "learning_rate": 9.091699117257992e-05, "loss": 0.0199, "step": 14110 }, { "grad_norm": 0.14991870522499084, "learning_rate": 9.090114651831163e-05, "loss": 0.022, "step": 14120 }, { "grad_norm": 0.19553634524345398, "learning_rate": 9.088528943935497e-05, "loss": 0.0206, "step": 14130 }, { "grad_norm": 0.2211083173751831, "learning_rate": 9.086941994052689e-05, "loss": 0.0216, "step": 14140 }, { "grad_norm": 0.1577509194612503, "learning_rate": 9.085353802664813e-05, "loss": 0.0198, "step": 14150 }, { "grad_norm": 0.18055951595306396, "learning_rate": 9.08376437025432e-05, "loss": 0.0204, "step": 14160 }, { "grad_norm": 0.16323840618133545, "learning_rate": 9.082173697304035e-05, "loss": 0.0213, "step": 14170 }, { "grad_norm": 0.16851693391799927, "learning_rate": 9.080581784297166e-05, "loss": 0.0201, "step": 14180 }, { "grad_norm": 0.18645814061164856, "learning_rate": 9.078988631717291e-05, "loss": 0.021, "step": 14190 }, { "grad_norm": 0.16068001091480255, "learning_rate": 9.077394240048369e-05, "loss": 0.0209, "step": 14200 }, { "grad_norm": 0.20946133136749268, "learning_rate": 9.075798609774736e-05, "loss": 0.0213, "step": 14210 }, { "grad_norm": 0.20037980377674103, "learning_rate": 9.0742017413811e-05, "loss": 0.0199, "step": 14220 }, { "grad_norm": 0.23566202819347382, "learning_rate": 9.072603635352548e-05, "loss": 0.0217, "step": 14230 }, { "grad_norm": 0.1777312010526657, "learning_rate": 9.071004292174541e-05, "loss": 0.0206, "step": 14240 }, { "grad_norm": 0.17888927459716797, "learning_rate": 9.06940371233292e-05, "loss": 0.0222, "step": 14250 }, { "grad_norm": 0.20505917072296143, "learning_rate": 9.067801896313898e-05, "loss": 0.0213, "step": 14260 }, { "grad_norm": 0.19522526860237122, "learning_rate": 9.066198844604064e-05, "loss": 0.0208, "step": 14270 }, { "grad_norm": 0.15199574828147888, "learning_rate": 9.06459455769038e-05, "loss": 0.019, "step": 14280 }, { "grad_norm": 0.13799329102039337, "learning_rate": 9.062989036060193e-05, "loss": 0.019, "step": 14290 }, { "grad_norm": 0.13052935898303986, "learning_rate": 9.061382280201212e-05, "loss": 0.0197, "step": 14300 }, { "grad_norm": 0.1900930553674698, "learning_rate": 9.059774290601528e-05, "loss": 0.0213, "step": 14310 }, { "grad_norm": 0.2358272224664688, "learning_rate": 9.058165067749606e-05, "loss": 0.0198, "step": 14320 }, { "grad_norm": 0.18012677133083344, "learning_rate": 9.056554612134288e-05, "loss": 0.0183, "step": 14330 }, { "grad_norm": 0.17915154993534088, "learning_rate": 9.054942924244785e-05, "loss": 0.0207, "step": 14340 }, { "grad_norm": 0.24472691118717194, "learning_rate": 9.053330004570686e-05, "loss": 0.0185, "step": 14350 }, { "grad_norm": 0.15270081162452698, "learning_rate": 9.051715853601955e-05, "loss": 0.0204, "step": 14360 }, { "grad_norm": 0.14979304373264313, "learning_rate": 9.050100471828926e-05, "loss": 0.0188, "step": 14370 }, { "grad_norm": 0.19442619383335114, "learning_rate": 9.048483859742311e-05, "loss": 0.0192, "step": 14380 }, { "grad_norm": 0.24304355680942535, "learning_rate": 9.046866017833193e-05, "loss": 0.0219, "step": 14390 }, { "grad_norm": 0.16488982737064362, "learning_rate": 9.045246946593029e-05, "loss": 0.0208, "step": 14400 }, { "grad_norm": 0.17566612362861633, "learning_rate": 9.043626646513652e-05, "loss": 0.0204, "step": 14410 }, { "grad_norm": 0.22212184965610504, "learning_rate": 9.042005118087267e-05, "loss": 0.0191, "step": 14420 }, { "grad_norm": 0.13297192752361298, "learning_rate": 9.040382361806448e-05, "loss": 0.0214, "step": 14430 }, { "grad_norm": 0.17459163069725037, "learning_rate": 9.038758378164148e-05, "loss": 0.02, "step": 14440 }, { "grad_norm": 0.18427127599716187, "learning_rate": 9.037133167653691e-05, "loss": 0.0206, "step": 14450 }, { "grad_norm": 0.19167561829090118, "learning_rate": 9.035506730768771e-05, "loss": 0.0199, "step": 14460 }, { "grad_norm": 0.18420273065567017, "learning_rate": 9.033879068003458e-05, "loss": 0.0185, "step": 14470 }, { "grad_norm": 0.1973770707845688, "learning_rate": 9.032250179852193e-05, "loss": 0.0189, "step": 14480 }, { "grad_norm": 0.1661621481180191, "learning_rate": 9.030620066809787e-05, "loss": 0.0201, "step": 14490 }, { "grad_norm": 0.18298542499542236, "learning_rate": 9.028988729371428e-05, "loss": 0.0211, "step": 14500 }, { "grad_norm": 0.24471165239810944, "learning_rate": 9.027356168032673e-05, "loss": 0.0208, "step": 14510 }, { "grad_norm": 0.1802057921886444, "learning_rate": 9.02572238328945e-05, "loss": 0.0203, "step": 14520 }, { "grad_norm": 0.1643144190311432, "learning_rate": 9.02408737563806e-05, "loss": 0.0205, "step": 14530 }, { "grad_norm": 0.23928986489772797, "learning_rate": 9.022451145575174e-05, "loss": 0.0208, "step": 14540 }, { "grad_norm": 0.172225683927536, "learning_rate": 9.02081369359784e-05, "loss": 0.0214, "step": 14550 }, { "grad_norm": 0.13840679824352264, "learning_rate": 9.019175020203465e-05, "loss": 0.0214, "step": 14560 }, { "grad_norm": 0.193657785654068, "learning_rate": 9.017535125889842e-05, "loss": 0.019, "step": 14570 }, { "grad_norm": 0.16321435570716858, "learning_rate": 9.015894011155124e-05, "loss": 0.0193, "step": 14580 }, { "grad_norm": 0.2297980636358261, "learning_rate": 9.014251676497838e-05, "loss": 0.0199, "step": 14590 }, { "grad_norm": 0.137435182929039, "learning_rate": 9.012608122416884e-05, "loss": 0.0207, "step": 14600 }, { "grad_norm": 0.18379445374011993, "learning_rate": 9.010963349411529e-05, "loss": 0.0203, "step": 14610 }, { "grad_norm": 0.20301984250545502, "learning_rate": 9.00931735798141e-05, "loss": 0.0215, "step": 14620 }, { "grad_norm": 0.14655670523643494, "learning_rate": 9.00767014862654e-05, "loss": 0.022, "step": 14630 }, { "grad_norm": 0.1987801343202591, "learning_rate": 9.006021721847295e-05, "loss": 0.0216, "step": 14640 }, { "grad_norm": 0.2081887125968933, "learning_rate": 9.004372078144423e-05, "loss": 0.0219, "step": 14650 }, { "grad_norm": 0.17046445608139038, "learning_rate": 9.002721218019043e-05, "loss": 0.0201, "step": 14660 }, { "grad_norm": 0.2499013990163803, "learning_rate": 9.001069141972642e-05, "loss": 0.0233, "step": 14670 }, { "grad_norm": 0.1597931981086731, "learning_rate": 8.99941585050708e-05, "loss": 0.0219, "step": 14680 }, { "grad_norm": 0.1711793690919876, "learning_rate": 8.997761344124578e-05, "loss": 0.0203, "step": 14690 }, { "grad_norm": 0.21443621814250946, "learning_rate": 8.996105623327737e-05, "loss": 0.0215, "step": 14700 }, { "grad_norm": 0.15340003371238708, "learning_rate": 8.994448688619517e-05, "loss": 0.0186, "step": 14710 }, { "grad_norm": 0.1607234627008438, "learning_rate": 8.992790540503253e-05, "loss": 0.0194, "step": 14720 }, { "grad_norm": 0.2616494596004486, "learning_rate": 8.991131179482648e-05, "loss": 0.0206, "step": 14730 }, { "grad_norm": 0.20689433813095093, "learning_rate": 8.989470606061768e-05, "loss": 0.0198, "step": 14740 }, { "grad_norm": 0.19681957364082336, "learning_rate": 8.987808820745056e-05, "loss": 0.0179, "step": 14750 }, { "grad_norm": 0.142257958650589, "learning_rate": 8.986145824037315e-05, "loss": 0.0222, "step": 14760 }, { "grad_norm": 0.17655016481876373, "learning_rate": 8.984481616443721e-05, "loss": 0.0196, "step": 14770 }, { "grad_norm": 0.14680133759975433, "learning_rate": 8.982816198469815e-05, "loss": 0.0193, "step": 14780 }, { "grad_norm": 0.20092009007930756, "learning_rate": 8.98114957062151e-05, "loss": 0.0194, "step": 14790 }, { "grad_norm": 0.14575102925300598, "learning_rate": 8.97948173340508e-05, "loss": 0.0204, "step": 14800 }, { "grad_norm": 0.1310509294271469, "learning_rate": 8.977812687327172e-05, "loss": 0.0189, "step": 14810 }, { "grad_norm": 0.15700282156467438, "learning_rate": 8.976142432894798e-05, "loss": 0.0173, "step": 14820 }, { "grad_norm": 0.1377997100353241, "learning_rate": 8.974470970615336e-05, "loss": 0.02, "step": 14830 }, { "grad_norm": 0.19083549082279205, "learning_rate": 8.972798300996534e-05, "loss": 0.0205, "step": 14840 }, { "grad_norm": 0.1446482539176941, "learning_rate": 8.971124424546504e-05, "loss": 0.0182, "step": 14850 }, { "grad_norm": 0.14845944941043854, "learning_rate": 8.969449341773724e-05, "loss": 0.0194, "step": 14860 }, { "grad_norm": 0.18451641499996185, "learning_rate": 8.967773053187042e-05, "loss": 0.0203, "step": 14870 }, { "grad_norm": 0.17487062513828278, "learning_rate": 8.966095559295668e-05, "loss": 0.0205, "step": 14880 }, { "grad_norm": 0.1717108190059662, "learning_rate": 8.964416860609184e-05, "loss": 0.0193, "step": 14890 }, { "grad_norm": 0.18187230825424194, "learning_rate": 8.962736957637532e-05, "loss": 0.0205, "step": 14900 }, { "grad_norm": 0.1857156753540039, "learning_rate": 8.96105585089102e-05, "loss": 0.0208, "step": 14910 }, { "grad_norm": 0.16381433606147766, "learning_rate": 8.959373540880329e-05, "loss": 0.0208, "step": 14920 }, { "grad_norm": 0.20627617835998535, "learning_rate": 8.957690028116495e-05, "loss": 0.0217, "step": 14930 }, { "grad_norm": 0.1948920339345932, "learning_rate": 8.956005313110928e-05, "loss": 0.0203, "step": 14940 }, { "grad_norm": 0.19320446252822876, "learning_rate": 8.9543193963754e-05, "loss": 0.0191, "step": 14950 }, { "grad_norm": 0.1503916233778, "learning_rate": 8.952632278422048e-05, "loss": 0.0176, "step": 14960 }, { "grad_norm": 0.18438377976417542, "learning_rate": 8.95094395976337e-05, "loss": 0.0203, "step": 14970 }, { "grad_norm": 0.12800560891628265, "learning_rate": 8.949254440912239e-05, "loss": 0.0201, "step": 14980 }, { "grad_norm": 0.16441760957241058, "learning_rate": 8.94756372238188e-05, "loss": 0.0201, "step": 14990 }, { "grad_norm": 0.13720183074474335, "learning_rate": 8.945871804685892e-05, "loss": 0.0185, "step": 15000 }, { "grad_norm": 0.18670746684074402, "learning_rate": 8.944178688338236e-05, "loss": 0.0223, "step": 15010 }, { "grad_norm": 0.18198449909687042, "learning_rate": 8.942484373853233e-05, "loss": 0.0198, "step": 15020 }, { "grad_norm": 0.1564316600561142, "learning_rate": 8.940788861745572e-05, "loss": 0.0179, "step": 15030 }, { "grad_norm": 0.14333869516849518, "learning_rate": 8.939092152530308e-05, "loss": 0.0182, "step": 15040 }, { "grad_norm": 0.17132720351219177, "learning_rate": 8.937394246722853e-05, "loss": 0.0198, "step": 15050 }, { "grad_norm": 0.20579476654529572, "learning_rate": 8.935695144838984e-05, "loss": 0.0199, "step": 15060 }, { "grad_norm": 0.20989054441452026, "learning_rate": 8.933994847394849e-05, "loss": 0.0228, "step": 15070 }, { "grad_norm": 0.1672840714454651, "learning_rate": 8.932293354906949e-05, "loss": 0.023, "step": 15080 }, { "grad_norm": 0.18315064907073975, "learning_rate": 8.930590667892153e-05, "loss": 0.0211, "step": 15090 }, { "grad_norm": 0.20042181015014648, "learning_rate": 8.928886786867696e-05, "loss": 0.0196, "step": 15100 }, { "grad_norm": 0.24735920131206512, "learning_rate": 8.927181712351168e-05, "loss": 0.0217, "step": 15110 }, { "grad_norm": 0.24245309829711914, "learning_rate": 8.925475444860527e-05, "loss": 0.0211, "step": 15120 }, { "grad_norm": 0.15206386148929596, "learning_rate": 8.923767984914092e-05, "loss": 0.0188, "step": 15130 }, { "grad_norm": 0.13136599957942963, "learning_rate": 8.922059333030545e-05, "loss": 0.02, "step": 15140 }, { "grad_norm": 0.14674656093120575, "learning_rate": 8.920349489728928e-05, "loss": 0.0184, "step": 15150 }, { "grad_norm": 0.13886937499046326, "learning_rate": 8.918638455528646e-05, "loss": 0.0182, "step": 15160 }, { "grad_norm": 0.15690912306308746, "learning_rate": 8.916926230949468e-05, "loss": 0.0192, "step": 15170 }, { "grad_norm": 0.17780441045761108, "learning_rate": 8.915212816511522e-05, "loss": 0.0183, "step": 15180 }, { "grad_norm": 0.13196273148059845, "learning_rate": 8.913498212735296e-05, "loss": 0.0208, "step": 15190 }, { "grad_norm": 0.18611915409564972, "learning_rate": 8.911782420141643e-05, "loss": 0.02, "step": 15200 }, { "grad_norm": 0.2115909457206726, "learning_rate": 8.910065439251775e-05, "loss": 0.0194, "step": 15210 }, { "grad_norm": 0.20121575891971588, "learning_rate": 8.908347270587268e-05, "loss": 0.0203, "step": 15220 }, { "grad_norm": 0.14296680688858032, "learning_rate": 8.906627914670054e-05, "loss": 0.02, "step": 15230 }, { "grad_norm": 0.1609913557767868, "learning_rate": 8.904907372022427e-05, "loss": 0.0194, "step": 15240 }, { "grad_norm": 0.19867253303527832, "learning_rate": 8.903185643167042e-05, "loss": 0.0199, "step": 15250 }, { "grad_norm": 0.16673727333545685, "learning_rate": 8.901462728626919e-05, "loss": 0.0202, "step": 15260 }, { "grad_norm": 0.12701119482517242, "learning_rate": 8.899738628925429e-05, "loss": 0.0201, "step": 15270 }, { "grad_norm": 0.20236699283123016, "learning_rate": 8.898013344586312e-05, "loss": 0.02, "step": 15280 }, { "grad_norm": 0.20409028232097626, "learning_rate": 8.896286876133661e-05, "loss": 0.0202, "step": 15290 }, { "grad_norm": 0.18544454872608185, "learning_rate": 8.894559224091933e-05, "loss": 0.0222, "step": 15300 }, { "grad_norm": 0.1669916808605194, "learning_rate": 8.892830388985942e-05, "loss": 0.02, "step": 15310 }, { "grad_norm": 0.12971745431423187, "learning_rate": 8.891100371340864e-05, "loss": 0.0172, "step": 15320 }, { "grad_norm": 0.14125752449035645, "learning_rate": 8.889369171682231e-05, "loss": 0.02, "step": 15330 }, { "grad_norm": 0.20893257856369019, "learning_rate": 8.887636790535936e-05, "loss": 0.0207, "step": 15340 }, { "grad_norm": 0.21518157422542572, "learning_rate": 8.885903228428231e-05, "loss": 0.02, "step": 15350 }, { "grad_norm": 0.2259586900472641, "learning_rate": 8.884168485885727e-05, "loss": 0.0207, "step": 15360 }, { "grad_norm": 0.15682543814182281, "learning_rate": 8.882432563435393e-05, "loss": 0.0206, "step": 15370 }, { "grad_norm": 0.1543176919221878, "learning_rate": 8.880695461604556e-05, "loss": 0.0195, "step": 15380 }, { "grad_norm": 0.16915027797222137, "learning_rate": 8.878957180920901e-05, "loss": 0.0223, "step": 15390 }, { "grad_norm": 0.19746403396129608, "learning_rate": 8.877217721912473e-05, "loss": 0.0217, "step": 15400 }, { "grad_norm": 0.21321958303451538, "learning_rate": 8.875477085107673e-05, "loss": 0.0179, "step": 15410 }, { "grad_norm": 0.15522576868534088, "learning_rate": 8.87373527103526e-05, "loss": 0.0228, "step": 15420 }, { "grad_norm": 0.19282957911491394, "learning_rate": 8.871992280224353e-05, "loss": 0.019, "step": 15430 }, { "grad_norm": 0.1828882247209549, "learning_rate": 8.870248113204422e-05, "loss": 0.0205, "step": 15440 }, { "grad_norm": 0.18143954873085022, "learning_rate": 8.868502770505306e-05, "loss": 0.0194, "step": 15450 }, { "grad_norm": 0.12737339735031128, "learning_rate": 8.86675625265719e-05, "loss": 0.019, "step": 15460 }, { "grad_norm": 0.15166927874088287, "learning_rate": 8.865008560190618e-05, "loss": 0.0172, "step": 15470 }, { "grad_norm": 0.18578849732875824, "learning_rate": 8.863259693636496e-05, "loss": 0.0204, "step": 15480 }, { "grad_norm": 0.13772077858448029, "learning_rate": 8.861509653526083e-05, "loss": 0.0207, "step": 15490 }, { "grad_norm": 0.14849108457565308, "learning_rate": 8.859758440390993e-05, "loss": 0.0224, "step": 15500 }, { "grad_norm": 0.2336774319410324, "learning_rate": 8.858006054763202e-05, "loss": 0.0191, "step": 15510 }, { "grad_norm": 0.16751374304294586, "learning_rate": 8.856252497175035e-05, "loss": 0.0215, "step": 15520 }, { "grad_norm": 0.20301948487758636, "learning_rate": 8.854497768159178e-05, "loss": 0.0196, "step": 15530 }, { "grad_norm": 0.16680701076984406, "learning_rate": 8.852741868248671e-05, "loss": 0.0197, "step": 15540 }, { "grad_norm": 0.16218437254428864, "learning_rate": 8.85098479797691e-05, "loss": 0.0196, "step": 15550 }, { "grad_norm": 0.1985892653465271, "learning_rate": 8.849226557877646e-05, "loss": 0.0218, "step": 15560 }, { "grad_norm": 0.16140227019786835, "learning_rate": 8.84746714848499e-05, "loss": 0.0211, "step": 15570 }, { "grad_norm": 0.17910778522491455, "learning_rate": 8.845706570333397e-05, "loss": 0.0207, "step": 15580 }, { "grad_norm": 0.19862820208072662, "learning_rate": 8.84394482395769e-05, "loss": 0.0184, "step": 15590 }, { "grad_norm": 0.16029182076454163, "learning_rate": 8.842181909893038e-05, "loss": 0.0184, "step": 15600 }, { "grad_norm": 0.199184849858284, "learning_rate": 8.840417828674969e-05, "loss": 0.0202, "step": 15610 }, { "grad_norm": 0.1743202805519104, "learning_rate": 8.838652580839364e-05, "loss": 0.019, "step": 15620 }, { "grad_norm": 0.20809918642044067, "learning_rate": 8.836886166922458e-05, "loss": 0.0182, "step": 15630 }, { "grad_norm": 0.22118262946605682, "learning_rate": 8.835118587460844e-05, "loss": 0.0196, "step": 15640 }, { "grad_norm": 0.18216988444328308, "learning_rate": 8.83334984299146e-05, "loss": 0.0195, "step": 15650 }, { "grad_norm": 0.1699240654706955, "learning_rate": 8.83157993405161e-05, "loss": 0.02, "step": 15660 }, { "grad_norm": 0.17496618628501892, "learning_rate": 8.829808861178943e-05, "loss": 0.0173, "step": 15670 }, { "grad_norm": 0.14186778664588928, "learning_rate": 8.828036624911464e-05, "loss": 0.0181, "step": 15680 }, { "grad_norm": 0.1963379830121994, "learning_rate": 8.826263225787532e-05, "loss": 0.0186, "step": 15690 }, { "grad_norm": 0.17164470255374908, "learning_rate": 8.824488664345858e-05, "loss": 0.0195, "step": 15700 }, { "grad_norm": 0.16703549027442932, "learning_rate": 8.822712941125508e-05, "loss": 0.0199, "step": 15710 }, { "grad_norm": 0.20178264379501343, "learning_rate": 8.820936056665898e-05, "loss": 0.0175, "step": 15720 }, { "grad_norm": 0.16481028497219086, "learning_rate": 8.819158011506801e-05, "loss": 0.019, "step": 15730 }, { "grad_norm": 0.17657610774040222, "learning_rate": 8.81737880618834e-05, "loss": 0.0187, "step": 15740 }, { "grad_norm": 0.19900932908058167, "learning_rate": 8.815598441250987e-05, "loss": 0.0208, "step": 15750 }, { "grad_norm": 0.15964458882808685, "learning_rate": 8.813816917235576e-05, "loss": 0.0178, "step": 15760 }, { "grad_norm": 0.13332031667232513, "learning_rate": 8.812034234683282e-05, "loss": 0.0182, "step": 15770 }, { "grad_norm": 0.16929411888122559, "learning_rate": 8.810250394135637e-05, "loss": 0.019, "step": 15780 }, { "grad_norm": 0.1824825257062912, "learning_rate": 8.808465396134529e-05, "loss": 0.0203, "step": 15790 }, { "grad_norm": 0.16732333600521088, "learning_rate": 8.806679241222189e-05, "loss": 0.0208, "step": 15800 }, { "grad_norm": 0.18084771931171417, "learning_rate": 8.804891929941203e-05, "loss": 0.0187, "step": 15810 }, { "grad_norm": 0.1398552507162094, "learning_rate": 8.803103462834514e-05, "loss": 0.0201, "step": 15820 }, { "grad_norm": 0.18907906115055084, "learning_rate": 8.801313840445408e-05, "loss": 0.0201, "step": 15830 }, { "grad_norm": 0.14863446354866028, "learning_rate": 8.799523063317524e-05, "loss": 0.019, "step": 15840 }, { "grad_norm": 0.1610252410173416, "learning_rate": 8.797731131994854e-05, "loss": 0.0197, "step": 15850 }, { "grad_norm": 0.1414487510919571, "learning_rate": 8.795938047021739e-05, "loss": 0.0212, "step": 15860 }, { "grad_norm": 0.1981615573167801, "learning_rate": 8.794143808942872e-05, "loss": 0.0196, "step": 15870 }, { "grad_norm": 0.18116694688796997, "learning_rate": 8.792348418303296e-05, "loss": 0.0201, "step": 15880 }, { "grad_norm": 0.17669722437858582, "learning_rate": 8.790551875648398e-05, "loss": 0.0216, "step": 15890 }, { "grad_norm": 0.17416279017925262, "learning_rate": 8.788754181523926e-05, "loss": 0.0193, "step": 15900 }, { "grad_norm": 0.14827567338943481, "learning_rate": 8.78695533647597e-05, "loss": 0.0186, "step": 15910 }, { "grad_norm": 0.1885354220867157, "learning_rate": 8.785155341050972e-05, "loss": 0.0178, "step": 15920 }, { "grad_norm": 0.1856997013092041, "learning_rate": 8.783354195795721e-05, "loss": 0.018, "step": 15930 }, { "grad_norm": 0.22733627259731293, "learning_rate": 8.78155190125736e-05, "loss": 0.0185, "step": 15940 }, { "grad_norm": 0.16336102783679962, "learning_rate": 8.779748457983378e-05, "loss": 0.0205, "step": 15950 }, { "grad_norm": 0.1782795488834381, "learning_rate": 8.777943866521612e-05, "loss": 0.0182, "step": 15960 }, { "grad_norm": 0.19504116475582123, "learning_rate": 8.77613812742025e-05, "loss": 0.0199, "step": 15970 }, { "grad_norm": 0.20365996658802032, "learning_rate": 8.774331241227829e-05, "loss": 0.0211, "step": 15980 }, { "grad_norm": 0.14330361783504486, "learning_rate": 8.772523208493232e-05, "loss": 0.0193, "step": 15990 }, { "grad_norm": 0.18158234655857086, "learning_rate": 8.770714029765692e-05, "loss": 0.0185, "step": 16000 }, { "grad_norm": 0.2145855873823166, "learning_rate": 8.768903705594789e-05, "loss": 0.0186, "step": 16010 }, { "grad_norm": 0.16859211027622223, "learning_rate": 8.767092236530453e-05, "loss": 0.0194, "step": 16020 }, { "grad_norm": 0.1334458887577057, "learning_rate": 8.76527962312296e-05, "loss": 0.0183, "step": 16030 }, { "grad_norm": 0.187560573220253, "learning_rate": 8.763465865922934e-05, "loss": 0.018, "step": 16040 }, { "grad_norm": 0.1774243861436844, "learning_rate": 8.761650965481347e-05, "loss": 0.0191, "step": 16050 }, { "grad_norm": 0.19841890037059784, "learning_rate": 8.759834922349516e-05, "loss": 0.0205, "step": 16060 }, { "grad_norm": 0.17925971746444702, "learning_rate": 8.758017737079108e-05, "loss": 0.0178, "step": 16070 }, { "grad_norm": 0.1837424784898758, "learning_rate": 8.756199410222137e-05, "loss": 0.0197, "step": 16080 }, { "grad_norm": 0.14789670705795288, "learning_rate": 8.754379942330963e-05, "loss": 0.0214, "step": 16090 }, { "grad_norm": 0.1629207730293274, "learning_rate": 8.75255933395829e-05, "loss": 0.0167, "step": 16100 }, { "grad_norm": 0.1936439573764801, "learning_rate": 8.750737585657171e-05, "loss": 0.0186, "step": 16110 }, { "grad_norm": 0.1616566777229309, "learning_rate": 8.748914697981008e-05, "loss": 0.0191, "step": 16120 }, { "grad_norm": 0.23598231375217438, "learning_rate": 8.747090671483542e-05, "loss": 0.0196, "step": 16130 }, { "grad_norm": 0.166209414601326, "learning_rate": 8.745265506718869e-05, "loss": 0.0196, "step": 16140 }, { "grad_norm": 0.16866441071033478, "learning_rate": 8.74343920424142e-05, "loss": 0.02, "step": 16150 }, { "grad_norm": 0.17817828059196472, "learning_rate": 8.741611764605982e-05, "loss": 0.0189, "step": 16160 }, { "grad_norm": 0.14370323717594147, "learning_rate": 8.739783188367682e-05, "loss": 0.019, "step": 16170 }, { "grad_norm": 0.19262759387493134, "learning_rate": 8.737953476081991e-05, "loss": 0.0192, "step": 16180 }, { "grad_norm": 0.20574204623699188, "learning_rate": 8.73612262830473e-05, "loss": 0.0203, "step": 16190 }, { "grad_norm": 0.1604091078042984, "learning_rate": 8.734290645592061e-05, "loss": 0.0213, "step": 16200 }, { "grad_norm": 0.20952162146568298, "learning_rate": 8.732457528500493e-05, "loss": 0.0203, "step": 16210 }, { "grad_norm": 0.1800132542848587, "learning_rate": 8.730623277586875e-05, "loss": 0.02, "step": 16220 }, { "grad_norm": 0.19656197726726532, "learning_rate": 8.72878789340841e-05, "loss": 0.0197, "step": 16230 }, { "grad_norm": 0.14395250380039215, "learning_rate": 8.726951376522635e-05, "loss": 0.0171, "step": 16240 }, { "grad_norm": 0.16094273328781128, "learning_rate": 8.725113727487435e-05, "loss": 0.0203, "step": 16250 }, { "grad_norm": 0.15495269000530243, "learning_rate": 8.723274946861042e-05, "loss": 0.0189, "step": 16260 }, { "grad_norm": 0.16637873649597168, "learning_rate": 8.721435035202026e-05, "loss": 0.0188, "step": 16270 }, { "grad_norm": 0.1523182988166809, "learning_rate": 8.719593993069306e-05, "loss": 0.0196, "step": 16280 }, { "grad_norm": 0.13317546248435974, "learning_rate": 8.717751821022139e-05, "loss": 0.0227, "step": 16290 }, { "grad_norm": 0.15847554802894592, "learning_rate": 8.715908519620134e-05, "loss": 0.0197, "step": 16300 }, { "grad_norm": 0.13504846394062042, "learning_rate": 8.71406408942323e-05, "loss": 0.0182, "step": 16310 }, { "grad_norm": 0.18850216269493103, "learning_rate": 8.712218530991723e-05, "loss": 0.0186, "step": 16320 }, { "grad_norm": 0.19622641801834106, "learning_rate": 8.710371844886241e-05, "loss": 0.0191, "step": 16330 }, { "grad_norm": 0.2031058371067047, "learning_rate": 8.708524031667758e-05, "loss": 0.0193, "step": 16340 }, { "grad_norm": 0.16580776870250702, "learning_rate": 8.706675091897592e-05, "loss": 0.0188, "step": 16350 }, { "grad_norm": 0.19419561326503754, "learning_rate": 8.704825026137404e-05, "loss": 0.0186, "step": 16360 }, { "grad_norm": 0.22281606495380402, "learning_rate": 8.702973834949192e-05, "loss": 0.0176, "step": 16370 }, { "grad_norm": 0.19277890026569366, "learning_rate": 8.701121518895301e-05, "loss": 0.0181, "step": 16380 }, { "grad_norm": 0.15415024757385254, "learning_rate": 8.699268078538414e-05, "loss": 0.0205, "step": 16390 }, { "grad_norm": 0.18921923637390137, "learning_rate": 8.69741351444156e-05, "loss": 0.018, "step": 16400 }, { "grad_norm": 0.13499388098716736, "learning_rate": 8.695557827168101e-05, "loss": 0.0178, "step": 16410 }, { "grad_norm": 0.15215185284614563, "learning_rate": 8.693701017281753e-05, "loss": 0.0189, "step": 16420 }, { "grad_norm": 0.18757693469524384, "learning_rate": 8.691843085346563e-05, "loss": 0.0186, "step": 16430 }, { "grad_norm": 0.16884706914424896, "learning_rate": 8.689984031926919e-05, "loss": 0.0171, "step": 16440 }, { "grad_norm": 0.14635847508907318, "learning_rate": 8.688123857587555e-05, "loss": 0.0188, "step": 16450 }, { "grad_norm": 0.1413511484861374, "learning_rate": 8.686262562893544e-05, "loss": 0.0173, "step": 16460 }, { "grad_norm": 0.18319034576416016, "learning_rate": 8.684400148410294e-05, "loss": 0.0177, "step": 16470 }, { "grad_norm": 0.1257772147655487, "learning_rate": 8.682536614703562e-05, "loss": 0.0203, "step": 16480 }, { "grad_norm": 0.15668097138404846, "learning_rate": 8.680671962339437e-05, "loss": 0.0198, "step": 16490 }, { "grad_norm": 0.18519707024097443, "learning_rate": 8.678806191884352e-05, "loss": 0.0168, "step": 16500 }, { "grad_norm": 0.18309788405895233, "learning_rate": 8.67693930390508e-05, "loss": 0.0191, "step": 16510 }, { "grad_norm": 0.17532067000865936, "learning_rate": 8.67507129896873e-05, "loss": 0.0186, "step": 16520 }, { "grad_norm": 0.12245668470859528, "learning_rate": 8.673202177642757e-05, "loss": 0.018, "step": 16530 }, { "grad_norm": 0.1571715623140335, "learning_rate": 8.671331940494945e-05, "loss": 0.0175, "step": 16540 }, { "grad_norm": 0.11659911274909973, "learning_rate": 8.669460588093427e-05, "loss": 0.0192, "step": 16550 }, { "grad_norm": 0.1438630223274231, "learning_rate": 8.667588121006667e-05, "loss": 0.0188, "step": 16560 }, { "grad_norm": 0.18790404498577118, "learning_rate": 8.665714539803475e-05, "loss": 0.0186, "step": 16570 }, { "grad_norm": 0.17243094742298126, "learning_rate": 8.663839845052993e-05, "loss": 0.0206, "step": 16580 }, { "grad_norm": 0.14998884499073029, "learning_rate": 8.661964037324703e-05, "loss": 0.0183, "step": 16590 }, { "grad_norm": 0.20449574291706085, "learning_rate": 8.660087117188427e-05, "loss": 0.0203, "step": 16600 }, { "grad_norm": 0.18462421000003815, "learning_rate": 8.658209085214325e-05, "loss": 0.0182, "step": 16610 }, { "grad_norm": 0.17586970329284668, "learning_rate": 8.656329941972891e-05, "loss": 0.0183, "step": 16620 }, { "grad_norm": 0.15966619551181793, "learning_rate": 8.654449688034963e-05, "loss": 0.0179, "step": 16630 }, { "grad_norm": 0.14018505811691284, "learning_rate": 8.652568323971706e-05, "loss": 0.0225, "step": 16640 }, { "grad_norm": 0.18777945637702942, "learning_rate": 8.650685850354636e-05, "loss": 0.0196, "step": 16650 }, { "grad_norm": 0.17876574397087097, "learning_rate": 8.648802267755593e-05, "loss": 0.0188, "step": 16660 }, { "grad_norm": 0.1162569597363472, "learning_rate": 8.646917576746764e-05, "loss": 0.0199, "step": 16670 }, { "grad_norm": 0.16034741699695587, "learning_rate": 8.645031777900666e-05, "loss": 0.0186, "step": 16680 }, { "grad_norm": 0.14928686618804932, "learning_rate": 8.643144871790154e-05, "loss": 0.0218, "step": 16690 }, { "grad_norm": 0.14352992177009583, "learning_rate": 8.641256858988424e-05, "loss": 0.0192, "step": 16700 }, { "grad_norm": 0.14934340119361877, "learning_rate": 8.639367740069e-05, "loss": 0.0187, "step": 16710 }, { "grad_norm": 0.173346608877182, "learning_rate": 8.63747751560575e-05, "loss": 0.0192, "step": 16720 }, { "grad_norm": 0.16789305210113525, "learning_rate": 8.635586186172871e-05, "loss": 0.017, "step": 16730 }, { "grad_norm": 0.21070556342601776, "learning_rate": 8.633693752344902e-05, "loss": 0.019, "step": 16740 }, { "grad_norm": 0.16993585228919983, "learning_rate": 8.631800214696713e-05, "loss": 0.0209, "step": 16750 }, { "grad_norm": 0.15362548828125, "learning_rate": 8.629905573803511e-05, "loss": 0.019, "step": 16760 }, { "grad_norm": 0.1566566675901413, "learning_rate": 8.628009830240839e-05, "loss": 0.0191, "step": 16770 }, { "grad_norm": 0.17701980471611023, "learning_rate": 8.626112984584571e-05, "loss": 0.0174, "step": 16780 }, { "grad_norm": 0.20455768704414368, "learning_rate": 8.62421503741092e-05, "loss": 0.0185, "step": 16790 }, { "grad_norm": 0.15522322058677673, "learning_rate": 8.622315989296432e-05, "loss": 0.0201, "step": 16800 }, { "grad_norm": 0.12917190790176392, "learning_rate": 8.62041584081799e-05, "loss": 0.0204, "step": 16810 }, { "grad_norm": 0.13908936083316803, "learning_rate": 8.618514592552807e-05, "loss": 0.0191, "step": 16820 }, { "grad_norm": 0.14038647711277008, "learning_rate": 8.616612245078431e-05, "loss": 0.0202, "step": 16830 }, { "grad_norm": 0.17827188968658447, "learning_rate": 8.614708798972746e-05, "loss": 0.0185, "step": 16840 }, { "grad_norm": 0.17689454555511475, "learning_rate": 8.61280425481397e-05, "loss": 0.0193, "step": 16850 }, { "grad_norm": 0.2018277943134308, "learning_rate": 8.61089861318065e-05, "loss": 0.0176, "step": 16860 }, { "grad_norm": 0.14417847990989685, "learning_rate": 8.608991874651673e-05, "loss": 0.0184, "step": 16870 }, { "grad_norm": 0.16444915533065796, "learning_rate": 8.607084039806255e-05, "loss": 0.0191, "step": 16880 }, { "grad_norm": 0.18273749947547913, "learning_rate": 8.605175109223944e-05, "loss": 0.018, "step": 16890 }, { "grad_norm": 0.19097900390625, "learning_rate": 8.603265083484624e-05, "loss": 0.0203, "step": 16900 }, { "grad_norm": 0.1614203304052353, "learning_rate": 8.60135396316851e-05, "loss": 0.0187, "step": 16910 }, { "grad_norm": 0.20279064774513245, "learning_rate": 8.599441748856152e-05, "loss": 0.0184, "step": 16920 }, { "grad_norm": 0.1676226258277893, "learning_rate": 8.597528441128427e-05, "loss": 0.0168, "step": 16930 }, { "grad_norm": 0.16738449037075043, "learning_rate": 8.595614040566549e-05, "loss": 0.0185, "step": 16940 }, { "grad_norm": 0.17902116477489471, "learning_rate": 8.593698547752063e-05, "loss": 0.0186, "step": 16950 }, { "grad_norm": 0.19360844790935516, "learning_rate": 8.591781963266843e-05, "loss": 0.0185, "step": 16960 }, { "grad_norm": 0.1757708340883255, "learning_rate": 8.5898642876931e-05, "loss": 0.0164, "step": 16970 }, { "grad_norm": 0.14694832265377045, "learning_rate": 8.587945521613369e-05, "loss": 0.0185, "step": 16980 }, { "grad_norm": 0.16929800808429718, "learning_rate": 8.586025665610524e-05, "loss": 0.0207, "step": 16990 }, { "grad_norm": 0.18267367780208588, "learning_rate": 8.584104720267765e-05, "loss": 0.0197, "step": 17000 }, { "grad_norm": 0.12433760613203049, "learning_rate": 8.582182686168625e-05, "loss": 0.0187, "step": 17010 }, { "grad_norm": 0.16481339931488037, "learning_rate": 8.580259563896967e-05, "loss": 0.02, "step": 17020 }, { "grad_norm": 0.1658940464258194, "learning_rate": 8.578335354036983e-05, "loss": 0.0178, "step": 17030 }, { "grad_norm": 0.17817649245262146, "learning_rate": 8.576410057173201e-05, "loss": 0.0189, "step": 17040 }, { "grad_norm": 0.17035417258739471, "learning_rate": 8.574483673890474e-05, "loss": 0.0185, "step": 17050 }, { "grad_norm": 0.16201211512088776, "learning_rate": 8.572556204773983e-05, "loss": 0.0197, "step": 17060 }, { "grad_norm": 0.15194272994995117, "learning_rate": 8.570627650409246e-05, "loss": 0.0171, "step": 17070 }, { "grad_norm": 0.155680313706398, "learning_rate": 8.568698011382107e-05, "loss": 0.0192, "step": 17080 }, { "grad_norm": 0.17689122259616852, "learning_rate": 8.566767288278738e-05, "loss": 0.0187, "step": 17090 }, { "grad_norm": 0.16162672638893127, "learning_rate": 8.56483548168564e-05, "loss": 0.0186, "step": 17100 }, { "grad_norm": 0.20084232091903687, "learning_rate": 8.562902592189648e-05, "loss": 0.0209, "step": 17110 }, { "grad_norm": 0.19845648109912872, "learning_rate": 8.560968620377921e-05, "loss": 0.0181, "step": 17120 }, { "grad_norm": 0.13613177835941315, "learning_rate": 8.559033566837951e-05, "loss": 0.0189, "step": 17130 }, { "grad_norm": 0.13881714642047882, "learning_rate": 8.557097432157551e-05, "loss": 0.0191, "step": 17140 }, { "grad_norm": 0.1435401290655136, "learning_rate": 8.555160216924872e-05, "loss": 0.0184, "step": 17150 }, { "grad_norm": 0.12074500322341919, "learning_rate": 8.55322192172839e-05, "loss": 0.0188, "step": 17160 }, { "grad_norm": 0.15266303718090057, "learning_rate": 8.551282547156902e-05, "loss": 0.0173, "step": 17170 }, { "grad_norm": 0.21362906694412231, "learning_rate": 8.549342093799544e-05, "loss": 0.0193, "step": 17180 }, { "grad_norm": 0.11179491877555847, "learning_rate": 8.547400562245773e-05, "loss": 0.0174, "step": 17190 }, { "grad_norm": 0.16753625869750977, "learning_rate": 8.545457953085374e-05, "loss": 0.019, "step": 17200 }, { "grad_norm": 0.15215229988098145, "learning_rate": 8.543514266908463e-05, "loss": 0.0183, "step": 17210 }, { "grad_norm": 0.14082776010036469, "learning_rate": 8.541569504305478e-05, "loss": 0.0204, "step": 17220 }, { "grad_norm": 0.17079460620880127, "learning_rate": 8.539623665867187e-05, "loss": 0.0191, "step": 17230 }, { "grad_norm": 0.18420930206775665, "learning_rate": 8.537676752184685e-05, "loss": 0.0186, "step": 17240 }, { "grad_norm": 0.1808854341506958, "learning_rate": 8.53572876384939e-05, "loss": 0.0195, "step": 17250 }, { "grad_norm": 0.13440734148025513, "learning_rate": 8.533779701453056e-05, "loss": 0.0194, "step": 17260 }, { "grad_norm": 0.18178264796733856, "learning_rate": 8.53182956558775e-05, "loss": 0.0171, "step": 17270 }, { "grad_norm": 0.16962319612503052, "learning_rate": 8.529878356845877e-05, "loss": 0.0186, "step": 17280 }, { "grad_norm": 0.12969762086868286, "learning_rate": 8.527926075820158e-05, "loss": 0.0176, "step": 17290 }, { "grad_norm": 0.16265268623828888, "learning_rate": 8.525972723103648e-05, "loss": 0.0199, "step": 17300 }, { "grad_norm": 0.1351538896560669, "learning_rate": 8.524018299289722e-05, "loss": 0.0193, "step": 17310 }, { "grad_norm": 0.13774438202381134, "learning_rate": 8.522062804972083e-05, "loss": 0.0184, "step": 17320 }, { "grad_norm": 0.1678866446018219, "learning_rate": 8.520106240744759e-05, "loss": 0.0191, "step": 17330 }, { "grad_norm": 0.15069261193275452, "learning_rate": 8.518148607202102e-05, "loss": 0.0176, "step": 17340 }, { "grad_norm": 0.1315113753080368, "learning_rate": 8.51618990493879e-05, "loss": 0.0175, "step": 17350 }, { "grad_norm": 0.15925846993923187, "learning_rate": 8.514230134549823e-05, "loss": 0.0177, "step": 17360 }, { "grad_norm": 0.17008955776691437, "learning_rate": 8.51226929663053e-05, "loss": 0.0192, "step": 17370 }, { "grad_norm": 0.1570618897676468, "learning_rate": 8.51030739177656e-05, "loss": 0.0184, "step": 17380 }, { "grad_norm": 0.14862844347953796, "learning_rate": 8.508344420583889e-05, "loss": 0.0175, "step": 17390 }, { "grad_norm": 0.15381406247615814, "learning_rate": 8.506380383648816e-05, "loss": 0.0215, "step": 17400 }, { "grad_norm": 0.14987334609031677, "learning_rate": 8.504415281567963e-05, "loss": 0.0167, "step": 17410 }, { "grad_norm": 0.1551356017589569, "learning_rate": 8.502449114938275e-05, "loss": 0.0166, "step": 17420 }, { "grad_norm": 0.12499264627695084, "learning_rate": 8.500481884357025e-05, "loss": 0.0178, "step": 17430 }, { "grad_norm": 0.1636708378791809, "learning_rate": 8.498513590421801e-05, "loss": 0.0173, "step": 17440 }, { "grad_norm": 0.12537921965122223, "learning_rate": 8.496544233730522e-05, "loss": 0.017, "step": 17450 }, { "grad_norm": 0.17338445782661438, "learning_rate": 8.494573814881426e-05, "loss": 0.0193, "step": 17460 }, { "grad_norm": 0.26010167598724365, "learning_rate": 8.492602334473074e-05, "loss": 0.0192, "step": 17470 }, { "grad_norm": 0.12318045645952225, "learning_rate": 8.49062979310435e-05, "loss": 0.0179, "step": 17480 }, { "grad_norm": 0.1570289433002472, "learning_rate": 8.488656191374458e-05, "loss": 0.0195, "step": 17490 }, { "grad_norm": 0.1562681496143341, "learning_rate": 8.48668152988293e-05, "loss": 0.017, "step": 17500 }, { "grad_norm": 0.166804701089859, "learning_rate": 8.484705809229612e-05, "loss": 0.0179, "step": 17510 }, { "grad_norm": 0.15143132209777832, "learning_rate": 8.482729030014677e-05, "loss": 0.0193, "step": 17520 }, { "grad_norm": 0.1832364946603775, "learning_rate": 8.48075119283862e-05, "loss": 0.0206, "step": 17530 }, { "grad_norm": 0.1598597913980484, "learning_rate": 8.478772298302254e-05, "loss": 0.019, "step": 17540 }, { "grad_norm": 0.13815036416053772, "learning_rate": 8.476792347006716e-05, "loss": 0.018, "step": 17550 }, { "grad_norm": 0.1641858071088791, "learning_rate": 8.474811339553462e-05, "loss": 0.0188, "step": 17560 }, { "grad_norm": 0.13203714787960052, "learning_rate": 8.47282927654427e-05, "loss": 0.02, "step": 17570 }, { "grad_norm": 0.17201906442642212, "learning_rate": 8.470846158581238e-05, "loss": 0.0187, "step": 17580 }, { "grad_norm": 0.14768919348716736, "learning_rate": 8.468861986266787e-05, "loss": 0.019, "step": 17590 }, { "grad_norm": 0.1614309698343277, "learning_rate": 8.466876760203654e-05, "loss": 0.0189, "step": 17600 }, { "grad_norm": 0.1569884568452835, "learning_rate": 8.464890480994898e-05, "loss": 0.0182, "step": 17610 }, { "grad_norm": 0.20373238623142242, "learning_rate": 8.462903149243899e-05, "loss": 0.0169, "step": 17620 }, { "grad_norm": 0.190693661570549, "learning_rate": 8.460914765554357e-05, "loss": 0.0206, "step": 17630 }, { "grad_norm": 0.18291744589805603, "learning_rate": 8.458925330530288e-05, "loss": 0.0196, "step": 17640 }, { "grad_norm": 0.19819577038288116, "learning_rate": 8.456934844776032e-05, "loss": 0.019, "step": 17650 }, { "grad_norm": 0.1608300507068634, "learning_rate": 8.454943308896246e-05, "loss": 0.019, "step": 17660 }, { "grad_norm": 0.17973732948303223, "learning_rate": 8.452950723495905e-05, "loss": 0.017, "step": 17670 }, { "grad_norm": 0.16509553790092468, "learning_rate": 8.450957089180303e-05, "loss": 0.0178, "step": 17680 }, { "grad_norm": 0.1325170248746872, "learning_rate": 8.448962406555055e-05, "loss": 0.0205, "step": 17690 }, { "grad_norm": 0.17134705185890198, "learning_rate": 8.446966676226093e-05, "loss": 0.0169, "step": 17700 }, { "grad_norm": 0.1423754096031189, "learning_rate": 8.444969898799667e-05, "loss": 0.0169, "step": 17710 }, { "grad_norm": 0.131002277135849, "learning_rate": 8.442972074882343e-05, "loss": 0.0173, "step": 17720 }, { "grad_norm": 0.10662694275379181, "learning_rate": 8.44097320508101e-05, "loss": 0.0186, "step": 17730 }, { "grad_norm": 0.1768508106470108, "learning_rate": 8.43897329000287e-05, "loss": 0.0173, "step": 17740 }, { "grad_norm": 0.13949155807495117, "learning_rate": 8.436972330255448e-05, "loss": 0.0181, "step": 17750 }, { "grad_norm": 0.163404643535614, "learning_rate": 8.434970326446579e-05, "loss": 0.0174, "step": 17760 }, { "grad_norm": 0.1768132746219635, "learning_rate": 8.432967279184418e-05, "loss": 0.0169, "step": 17770 }, { "grad_norm": 0.13147827982902527, "learning_rate": 8.430963189077441e-05, "loss": 0.0179, "step": 17780 }, { "grad_norm": 0.15014804899692535, "learning_rate": 8.428958056734437e-05, "loss": 0.0181, "step": 17790 }, { "grad_norm": 0.15863245725631714, "learning_rate": 8.426951882764513e-05, "loss": 0.0175, "step": 17800 }, { "grad_norm": 0.13578324019908905, "learning_rate": 8.424944667777089e-05, "loss": 0.0183, "step": 17810 }, { "grad_norm": 0.20078982412815094, "learning_rate": 8.422936412381905e-05, "loss": 0.0178, "step": 17820 }, { "grad_norm": 0.13534514605998993, "learning_rate": 8.420927117189017e-05, "loss": 0.0162, "step": 17830 }, { "grad_norm": 0.15421311557292938, "learning_rate": 8.418916782808795e-05, "loss": 0.0172, "step": 17840 }, { "grad_norm": 0.2620203197002411, "learning_rate": 8.416905409851926e-05, "loss": 0.0188, "step": 17850 }, { "grad_norm": 0.14506006240844727, "learning_rate": 8.41489299892941e-05, "loss": 0.0178, "step": 17860 }, { "grad_norm": 0.17832057178020477, "learning_rate": 8.412879550652566e-05, "loss": 0.0194, "step": 17870 }, { "grad_norm": 0.13113294541835785, "learning_rate": 8.410865065633029e-05, "loss": 0.0169, "step": 17880 }, { "grad_norm": 0.13034039735794067, "learning_rate": 8.408849544482742e-05, "loss": 0.0179, "step": 17890 }, { "grad_norm": 0.14607436954975128, "learning_rate": 8.406832987813968e-05, "loss": 0.0161, "step": 17900 }, { "grad_norm": 0.16336943209171295, "learning_rate": 8.404815396239286e-05, "loss": 0.017, "step": 17910 }, { "grad_norm": 0.15855196118354797, "learning_rate": 8.402796770371587e-05, "loss": 0.0197, "step": 17920 }, { "grad_norm": 0.12089693546295166, "learning_rate": 8.400777110824071e-05, "loss": 0.0168, "step": 17930 }, { "grad_norm": 0.1496332734823227, "learning_rate": 8.398756418210263e-05, "loss": 0.0176, "step": 17940 }, { "grad_norm": 0.1768798530101776, "learning_rate": 8.396734693143993e-05, "loss": 0.0173, "step": 17950 }, { "grad_norm": 0.14949655532836914, "learning_rate": 8.39471193623941e-05, "loss": 0.0162, "step": 17960 }, { "grad_norm": 0.16064870357513428, "learning_rate": 8.392688148110974e-05, "loss": 0.0182, "step": 17970 }, { "grad_norm": 0.191933736205101, "learning_rate": 8.390663329373456e-05, "loss": 0.0164, "step": 17980 }, { "grad_norm": 0.19220155477523804, "learning_rate": 8.388637480641944e-05, "loss": 0.0162, "step": 17990 }, { "grad_norm": 0.20162422955036163, "learning_rate": 8.386610602531837e-05, "loss": 0.0185, "step": 18000 }, { "grad_norm": 0.12759189307689667, "learning_rate": 8.384582695658847e-05, "loss": 0.0185, "step": 18010 }, { "grad_norm": 0.1386285126209259, "learning_rate": 8.382553760638999e-05, "loss": 0.017, "step": 18020 }, { "grad_norm": 0.12710784375667572, "learning_rate": 8.380523798088631e-05, "loss": 0.0183, "step": 18030 }, { "grad_norm": 0.18567436933517456, "learning_rate": 8.378492808624389e-05, "loss": 0.0188, "step": 18040 }, { "grad_norm": 0.133214071393013, "learning_rate": 8.376460792863237e-05, "loss": 0.0183, "step": 18050 }, { "grad_norm": 0.1737290322780609, "learning_rate": 8.374427751422444e-05, "loss": 0.016, "step": 18060 }, { "grad_norm": 0.13522285223007202, "learning_rate": 8.3723936849196e-05, "loss": 0.0175, "step": 18070 }, { "grad_norm": 0.17123019695281982, "learning_rate": 8.370358593972595e-05, "loss": 0.0176, "step": 18080 }, { "grad_norm": 0.13955757021903992, "learning_rate": 8.36832247919964e-05, "loss": 0.0179, "step": 18090 }, { "grad_norm": 0.11356417834758759, "learning_rate": 8.36628534121925e-05, "loss": 0.0153, "step": 18100 }, { "grad_norm": 0.16936413943767548, "learning_rate": 8.364247180650254e-05, "loss": 0.017, "step": 18110 }, { "grad_norm": 0.11767575144767761, "learning_rate": 8.362207998111794e-05, "loss": 0.0171, "step": 18120 }, { "grad_norm": 0.1649600714445114, "learning_rate": 8.360167794223318e-05, "loss": 0.0175, "step": 18130 }, { "grad_norm": 0.14098607003688812, "learning_rate": 8.358126569604586e-05, "loss": 0.0211, "step": 18140 }, { "grad_norm": 0.13402214646339417, "learning_rate": 8.356084324875668e-05, "loss": 0.0174, "step": 18150 }, { "grad_norm": 0.15741050243377686, "learning_rate": 8.354041060656945e-05, "loss": 0.0167, "step": 18160 }, { "grad_norm": 0.12084929645061493, "learning_rate": 8.351996777569106e-05, "loss": 0.0192, "step": 18170 }, { "grad_norm": 0.20018422603607178, "learning_rate": 8.349951476233148e-05, "loss": 0.0182, "step": 18180 }, { "grad_norm": 0.1803124099969864, "learning_rate": 8.347905157270386e-05, "loss": 0.0173, "step": 18190 }, { "grad_norm": 0.18925639986991882, "learning_rate": 8.345857821302432e-05, "loss": 0.0173, "step": 18200 }, { "grad_norm": 0.16147059202194214, "learning_rate": 8.343809468951213e-05, "loss": 0.0165, "step": 18210 }, { "grad_norm": 0.16076327860355377, "learning_rate": 8.341760100838965e-05, "loss": 0.0192, "step": 18220 }, { "grad_norm": 0.11758031696081161, "learning_rate": 8.339709717588233e-05, "loss": 0.0181, "step": 18230 }, { "grad_norm": 0.12737654149532318, "learning_rate": 8.33765831982187e-05, "loss": 0.018, "step": 18240 }, { "grad_norm": 0.1565479189157486, "learning_rate": 8.335605908163035e-05, "loss": 0.0173, "step": 18250 }, { "grad_norm": 0.16566571593284607, "learning_rate": 8.333552483235196e-05, "loss": 0.017, "step": 18260 }, { "grad_norm": 0.17229695618152618, "learning_rate": 8.33149804566213e-05, "loss": 0.0186, "step": 18270 }, { "grad_norm": 0.18268543481826782, "learning_rate": 8.329442596067921e-05, "loss": 0.0179, "step": 18280 }, { "grad_norm": 0.19127899408340454, "learning_rate": 8.32738613507696e-05, "loss": 0.0179, "step": 18290 }, { "grad_norm": 0.15941046178340912, "learning_rate": 8.325328663313946e-05, "loss": 0.0185, "step": 18300 }, { "grad_norm": 0.1667836755514145, "learning_rate": 8.323270181403884e-05, "loss": 0.0191, "step": 18310 }, { "grad_norm": 0.19056449830532074, "learning_rate": 8.321210689972086e-05, "loss": 0.0175, "step": 18320 }, { "grad_norm": 0.1959446668624878, "learning_rate": 8.319150189644174e-05, "loss": 0.02, "step": 18330 }, { "grad_norm": 0.19335758686065674, "learning_rate": 8.31708868104607e-05, "loss": 0.0189, "step": 18340 }, { "grad_norm": 0.12450661510229111, "learning_rate": 8.315026164804007e-05, "loss": 0.0172, "step": 18350 }, { "grad_norm": 0.12622705101966858, "learning_rate": 8.312962641544524e-05, "loss": 0.017, "step": 18360 }, { "grad_norm": 0.15717725455760956, "learning_rate": 8.310898111894465e-05, "loss": 0.0179, "step": 18370 }, { "grad_norm": 0.22763867676258087, "learning_rate": 8.308832576480977e-05, "loss": 0.0206, "step": 18380 }, { "grad_norm": 0.18744565546512604, "learning_rate": 8.306766035931519e-05, "loss": 0.0185, "step": 18390 }, { "grad_norm": 0.1947813779115677, "learning_rate": 8.304698490873847e-05, "loss": 0.0186, "step": 18400 }, { "grad_norm": 0.19152528047561646, "learning_rate": 8.30262994193603e-05, "loss": 0.0203, "step": 18410 }, { "grad_norm": 0.19945698976516724, "learning_rate": 8.300560389746438e-05, "loss": 0.0187, "step": 18420 }, { "grad_norm": 0.191675066947937, "learning_rate": 8.298489834933745e-05, "loss": 0.0197, "step": 18430 }, { "grad_norm": 0.14425909519195557, "learning_rate": 8.296418278126934e-05, "loss": 0.0157, "step": 18440 }, { "grad_norm": 0.14474010467529297, "learning_rate": 8.294345719955284e-05, "loss": 0.0189, "step": 18450 }, { "grad_norm": 0.15248392522335052, "learning_rate": 8.29227216104839e-05, "loss": 0.0182, "step": 18460 }, { "grad_norm": 0.13034822046756744, "learning_rate": 8.290197602036137e-05, "loss": 0.0169, "step": 18470 }, { "grad_norm": 0.15819640457630157, "learning_rate": 8.288122043548725e-05, "loss": 0.0171, "step": 18480 }, { "grad_norm": 0.13519206643104553, "learning_rate": 8.286045486216657e-05, "loss": 0.0171, "step": 18490 }, { "grad_norm": 0.14735107123851776, "learning_rate": 8.283967930670733e-05, "loss": 0.0176, "step": 18500 }, { "grad_norm": 0.14250603318214417, "learning_rate": 8.281889377542058e-05, "loss": 0.0178, "step": 18510 }, { "grad_norm": 0.16108062863349915, "learning_rate": 8.279809827462045e-05, "loss": 0.0201, "step": 18520 }, { "grad_norm": 0.1849130541086197, "learning_rate": 8.277729281062402e-05, "loss": 0.0168, "step": 18530 }, { "grad_norm": 0.20297059416770935, "learning_rate": 8.27564773897515e-05, "loss": 0.0197, "step": 18540 }, { "grad_norm": 0.12446214258670807, "learning_rate": 8.273565201832602e-05, "loss": 0.0174, "step": 18550 }, { "grad_norm": 0.12623544037342072, "learning_rate": 8.27148167026738e-05, "loss": 0.0174, "step": 18560 }, { "grad_norm": 0.1503695696592331, "learning_rate": 8.269397144912405e-05, "loss": 0.0196, "step": 18570 }, { "grad_norm": 0.13779893517494202, "learning_rate": 8.267311626400899e-05, "loss": 0.0151, "step": 18580 }, { "grad_norm": 0.1581096351146698, "learning_rate": 8.26522511536639e-05, "loss": 0.018, "step": 18590 }, { "grad_norm": 0.17423023283481598, "learning_rate": 8.263137612442706e-05, "loss": 0.0182, "step": 18600 }, { "grad_norm": 0.1772536337375641, "learning_rate": 8.261049118263971e-05, "loss": 0.02, "step": 18610 }, { "grad_norm": 0.1506762057542801, "learning_rate": 8.258959633464619e-05, "loss": 0.0162, "step": 18620 }, { "grad_norm": 0.11234478652477264, "learning_rate": 8.256869158679377e-05, "loss": 0.0162, "step": 18630 }, { "grad_norm": 0.18053117394447327, "learning_rate": 8.254777694543278e-05, "loss": 0.0175, "step": 18640 }, { "grad_norm": 0.17005904018878937, "learning_rate": 8.252685241691651e-05, "loss": 0.0175, "step": 18650 }, { "grad_norm": 0.13809168338775635, "learning_rate": 8.250591800760133e-05, "loss": 0.0174, "step": 18660 }, { "grad_norm": 0.1586449295282364, "learning_rate": 8.248497372384649e-05, "loss": 0.0157, "step": 18670 }, { "grad_norm": 0.15523825585842133, "learning_rate": 8.246401957201437e-05, "loss": 0.0169, "step": 18680 }, { "grad_norm": 0.15902727842330933, "learning_rate": 8.244305555847027e-05, "loss": 0.0174, "step": 18690 }, { "grad_norm": 0.19730624556541443, "learning_rate": 8.24220816895825e-05, "loss": 0.0168, "step": 18700 }, { "grad_norm": 0.18366295099258423, "learning_rate": 8.240109797172237e-05, "loss": 0.0173, "step": 18710 }, { "grad_norm": 0.18363070487976074, "learning_rate": 8.238010441126416e-05, "loss": 0.0186, "step": 18720 }, { "grad_norm": 0.14882680773735046, "learning_rate": 8.23591010145852e-05, "loss": 0.0158, "step": 18730 }, { "grad_norm": 0.1461612582206726, "learning_rate": 8.233808778806571e-05, "loss": 0.0192, "step": 18740 }, { "grad_norm": 0.13420948386192322, "learning_rate": 8.231706473808903e-05, "loss": 0.018, "step": 18750 }, { "grad_norm": 0.16780459880828857, "learning_rate": 8.229603187104133e-05, "loss": 0.0172, "step": 18760 }, { "grad_norm": 0.1275257021188736, "learning_rate": 8.22749891933119e-05, "loss": 0.0163, "step": 18770 }, { "grad_norm": 0.16159769892692566, "learning_rate": 8.225393671129291e-05, "loss": 0.0174, "step": 18780 }, { "grad_norm": 0.1500135064125061, "learning_rate": 8.223287443137957e-05, "loss": 0.0185, "step": 18790 }, { "grad_norm": 0.14705809950828552, "learning_rate": 8.221180235997004e-05, "loss": 0.0187, "step": 18800 }, { "grad_norm": 0.15717177093029022, "learning_rate": 8.219072050346544e-05, "loss": 0.0177, "step": 18810 }, { "grad_norm": 0.15889877080917358, "learning_rate": 8.216962886826992e-05, "loss": 0.0172, "step": 18820 }, { "grad_norm": 0.1909739077091217, "learning_rate": 8.214852746079054e-05, "loss": 0.0177, "step": 18830 }, { "grad_norm": 0.13824215531349182, "learning_rate": 8.212741628743732e-05, "loss": 0.0173, "step": 18840 }, { "grad_norm": 0.2219124138355255, "learning_rate": 8.210629535462333e-05, "loss": 0.0182, "step": 18850 }, { "grad_norm": 0.18464413285255432, "learning_rate": 8.208516466876453e-05, "loss": 0.0175, "step": 18860 }, { "grad_norm": 0.14760734140872955, "learning_rate": 8.206402423627986e-05, "loss": 0.0167, "step": 18870 }, { "grad_norm": 0.13902021944522858, "learning_rate": 8.204287406359124e-05, "loss": 0.0175, "step": 18880 }, { "grad_norm": 0.13508738577365875, "learning_rate": 8.20217141571235e-05, "loss": 0.0161, "step": 18890 }, { "grad_norm": 0.1483321189880371, "learning_rate": 8.200054452330449e-05, "loss": 0.0173, "step": 18900 }, { "grad_norm": 0.1744116097688675, "learning_rate": 8.197936516856499e-05, "loss": 0.0167, "step": 18910 }, { "grad_norm": 0.14464326202869415, "learning_rate": 8.195817609933871e-05, "loss": 0.0185, "step": 18920 }, { "grad_norm": 0.1865975558757782, "learning_rate": 8.193697732206233e-05, "loss": 0.018, "step": 18930 }, { "grad_norm": 0.13743190467357635, "learning_rate": 8.19157688431755e-05, "loss": 0.016, "step": 18940 }, { "grad_norm": 0.20856626331806183, "learning_rate": 8.189455066912077e-05, "loss": 0.0181, "step": 18950 }, { "grad_norm": 0.14324577152729034, "learning_rate": 8.187332280634369e-05, "loss": 0.0172, "step": 18960 }, { "grad_norm": 0.12022917717695236, "learning_rate": 8.18520852612927e-05, "loss": 0.0181, "step": 18970 }, { "grad_norm": 0.16590313613414764, "learning_rate": 8.183083804041921e-05, "loss": 0.0173, "step": 18980 }, { "grad_norm": 0.1957574039697647, "learning_rate": 8.180958115017757e-05, "loss": 0.0176, "step": 18990 }, { "grad_norm": 0.1752672642469406, "learning_rate": 8.178831459702505e-05, "loss": 0.0166, "step": 19000 }, { "grad_norm": 0.1256168782711029, "learning_rate": 8.17670383874219e-05, "loss": 0.016, "step": 19010 }, { "grad_norm": 0.16103404760360718, "learning_rate": 8.174575252783124e-05, "loss": 0.0167, "step": 19020 }, { "grad_norm": 0.17317014932632446, "learning_rate": 8.172445702471914e-05, "loss": 0.0188, "step": 19030 }, { "grad_norm": 0.16849832236766815, "learning_rate": 8.170315188455466e-05, "loss": 0.0188, "step": 19040 }, { "grad_norm": 0.15258818864822388, "learning_rate": 8.168183711380969e-05, "loss": 0.0185, "step": 19050 }, { "grad_norm": 0.1533595472574234, "learning_rate": 8.166051271895913e-05, "loss": 0.0203, "step": 19060 }, { "grad_norm": 0.11963828653097153, "learning_rate": 8.163917870648075e-05, "loss": 0.0174, "step": 19070 }, { "grad_norm": 0.15465417504310608, "learning_rate": 8.161783508285526e-05, "loss": 0.0188, "step": 19080 }, { "grad_norm": 0.20868778228759766, "learning_rate": 8.159648185456628e-05, "loss": 0.0178, "step": 19090 }, { "grad_norm": 0.20067472755908966, "learning_rate": 8.157511902810038e-05, "loss": 0.0194, "step": 19100 }, { "grad_norm": 0.15341167151927948, "learning_rate": 8.155374660994701e-05, "loss": 0.0203, "step": 19110 }, { "grad_norm": 0.12722370028495789, "learning_rate": 8.153236460659857e-05, "loss": 0.018, "step": 19120 }, { "grad_norm": 0.1719716489315033, "learning_rate": 8.151097302455031e-05, "loss": 0.0182, "step": 19130 }, { "grad_norm": 0.18281249701976776, "learning_rate": 8.148957187030044e-05, "loss": 0.0187, "step": 19140 }, { "grad_norm": 0.15474078059196472, "learning_rate": 8.146816115035006e-05, "loss": 0.0178, "step": 19150 }, { "grad_norm": 0.14829373359680176, "learning_rate": 8.14467408712032e-05, "loss": 0.0169, "step": 19160 }, { "grad_norm": 0.15897439420223236, "learning_rate": 8.142531103936678e-05, "loss": 0.0165, "step": 19170 }, { "grad_norm": 0.18321825563907623, "learning_rate": 8.14038716613506e-05, "loss": 0.0171, "step": 19180 }, { "grad_norm": 0.17595145106315613, "learning_rate": 8.138242274366736e-05, "loss": 0.0188, "step": 19190 }, { "grad_norm": 0.1818808615207672, "learning_rate": 8.136096429283271e-05, "loss": 0.0197, "step": 19200 }, { "grad_norm": 0.1415441483259201, "learning_rate": 8.133949631536515e-05, "loss": 0.0173, "step": 19210 }, { "grad_norm": 0.16264726221561432, "learning_rate": 8.131801881778607e-05, "loss": 0.0177, "step": 19220 }, { "grad_norm": 0.1584453135728836, "learning_rate": 8.129653180661978e-05, "loss": 0.0178, "step": 19230 }, { "grad_norm": 0.1713601052761078, "learning_rate": 8.127503528839346e-05, "loss": 0.0177, "step": 19240 }, { "grad_norm": 0.17089775204658508, "learning_rate": 8.125352926963721e-05, "loss": 0.0207, "step": 19250 }, { "grad_norm": 0.16454310715198517, "learning_rate": 8.123201375688395e-05, "loss": 0.0185, "step": 19260 }, { "grad_norm": 0.12414990365505219, "learning_rate": 8.121048875666954e-05, "loss": 0.0158, "step": 19270 }, { "grad_norm": 0.23520468175411224, "learning_rate": 8.118895427553274e-05, "loss": 0.0176, "step": 19280 }, { "grad_norm": 0.1889357715845108, "learning_rate": 8.116741032001511e-05, "loss": 0.017, "step": 19290 }, { "grad_norm": 0.15471017360687256, "learning_rate": 8.114585689666114e-05, "loss": 0.0191, "step": 19300 }, { "grad_norm": 0.15161989629268646, "learning_rate": 8.112429401201821e-05, "loss": 0.0179, "step": 19310 }, { "grad_norm": 0.15892978012561798, "learning_rate": 8.110272167263656e-05, "loss": 0.0193, "step": 19320 }, { "grad_norm": 0.19524875283241272, "learning_rate": 8.108113988506929e-05, "loss": 0.0175, "step": 19330 }, { "grad_norm": 0.1780458241701126, "learning_rate": 8.105954865587235e-05, "loss": 0.0178, "step": 19340 }, { "grad_norm": 0.1255573183298111, "learning_rate": 8.103794799160463e-05, "loss": 0.0167, "step": 19350 }, { "grad_norm": 0.18433284759521484, "learning_rate": 8.101633789882781e-05, "loss": 0.0172, "step": 19360 }, { "grad_norm": 0.20372837781906128, "learning_rate": 8.099471838410648e-05, "loss": 0.0215, "step": 19370 }, { "grad_norm": 0.14901095628738403, "learning_rate": 8.097308945400806e-05, "loss": 0.0181, "step": 19380 }, { "grad_norm": 0.1505717933177948, "learning_rate": 8.095145111510288e-05, "loss": 0.0198, "step": 19390 }, { "grad_norm": 0.2069382518529892, "learning_rate": 8.092980337396406e-05, "loss": 0.0193, "step": 19400 }, { "grad_norm": 0.1692270040512085, "learning_rate": 8.090814623716763e-05, "loss": 0.0196, "step": 19410 }, { "grad_norm": 0.14527542889118195, "learning_rate": 8.088647971129246e-05, "loss": 0.0186, "step": 19420 }, { "grad_norm": 0.14234298467636108, "learning_rate": 8.086480380292026e-05, "loss": 0.0163, "step": 19430 }, { "grad_norm": 0.1757863312959671, "learning_rate": 8.084311851863562e-05, "loss": 0.0198, "step": 19440 }, { "grad_norm": 0.12920241057872772, "learning_rate": 8.082142386502591e-05, "loss": 0.0145, "step": 19450 }, { "grad_norm": 0.13859409093856812, "learning_rate": 8.079971984868145e-05, "loss": 0.0174, "step": 19460 }, { "grad_norm": 0.15345250070095062, "learning_rate": 8.077800647619532e-05, "loss": 0.0173, "step": 19470 }, { "grad_norm": 0.13000968098640442, "learning_rate": 8.075628375416345e-05, "loss": 0.0173, "step": 19480 }, { "grad_norm": 0.12269225716590881, "learning_rate": 8.073455168918464e-05, "loss": 0.0168, "step": 19490 }, { "grad_norm": 0.15551269054412842, "learning_rate": 8.071281028786055e-05, "loss": 0.016, "step": 19500 }, { "grad_norm": 0.15278345346450806, "learning_rate": 8.069105955679562e-05, "loss": 0.0195, "step": 19510 }, { "grad_norm": 0.16734318435192108, "learning_rate": 8.066929950259713e-05, "loss": 0.0178, "step": 19520 }, { "grad_norm": 0.14915885031223297, "learning_rate": 8.064753013187522e-05, "loss": 0.018, "step": 19530 }, { "grad_norm": 0.11067048460245132, "learning_rate": 8.062575145124289e-05, "loss": 0.0153, "step": 19540 }, { "grad_norm": 0.15374866127967834, "learning_rate": 8.060396346731587e-05, "loss": 0.0176, "step": 19550 }, { "grad_norm": 0.1604742854833603, "learning_rate": 8.058216618671281e-05, "loss": 0.0183, "step": 19560 }, { "grad_norm": 0.17490923404693604, "learning_rate": 8.056035961605514e-05, "loss": 0.0185, "step": 19570 }, { "grad_norm": 0.12837116420269012, "learning_rate": 8.05385437619671e-05, "loss": 0.0163, "step": 19580 }, { "grad_norm": 0.16874344646930695, "learning_rate": 8.05167186310758e-05, "loss": 0.0164, "step": 19590 }, { "grad_norm": 0.1686919629573822, "learning_rate": 8.049488423001113e-05, "loss": 0.0167, "step": 19600 }, { "grad_norm": 0.1727740317583084, "learning_rate": 8.047304056540581e-05, "loss": 0.0181, "step": 19610 }, { "grad_norm": 0.17198307812213898, "learning_rate": 8.045118764389534e-05, "loss": 0.0174, "step": 19620 }, { "grad_norm": 0.11281390488147736, "learning_rate": 8.042932547211809e-05, "loss": 0.0177, "step": 19630 }, { "grad_norm": 0.12333542853593826, "learning_rate": 8.04074540567152e-05, "loss": 0.0158, "step": 19640 }, { "grad_norm": 0.2056383639574051, "learning_rate": 8.038557340433063e-05, "loss": 0.018, "step": 19650 }, { "grad_norm": 0.20535627007484436, "learning_rate": 8.036368352161115e-05, "loss": 0.0189, "step": 19660 }, { "grad_norm": 0.14558768272399902, "learning_rate": 8.034178441520633e-05, "loss": 0.0178, "step": 19670 }, { "grad_norm": 0.11788477748632431, "learning_rate": 8.031987609176852e-05, "loss": 0.0156, "step": 19680 }, { "grad_norm": 0.1336362063884735, "learning_rate": 8.02979585579529e-05, "loss": 0.0152, "step": 19690 }, { "grad_norm": 0.1231703907251358, "learning_rate": 8.027603182041745e-05, "loss": 0.0153, "step": 19700 }, { "grad_norm": 0.12274188548326492, "learning_rate": 8.025409588582292e-05, "loss": 0.0165, "step": 19710 }, { "grad_norm": 0.1576191782951355, "learning_rate": 8.023215076083288e-05, "loss": 0.016, "step": 19720 }, { "grad_norm": 0.15183144807815552, "learning_rate": 8.021019645211367e-05, "loss": 0.0154, "step": 19730 }, { "grad_norm": 0.13746856153011322, "learning_rate": 8.018823296633441e-05, "loss": 0.0164, "step": 19740 }, { "grad_norm": 0.15856437385082245, "learning_rate": 8.016626031016708e-05, "loss": 0.0172, "step": 19750 }, { "grad_norm": 0.13529403507709503, "learning_rate": 8.014427849028636e-05, "loss": 0.019, "step": 19760 }, { "grad_norm": 0.20985563099384308, "learning_rate": 8.012228751336974e-05, "loss": 0.0159, "step": 19770 }, { "grad_norm": 0.13614489138126373, "learning_rate": 8.01002873860975e-05, "loss": 0.0166, "step": 19780 }, { "grad_norm": 0.1875324249267578, "learning_rate": 8.00782781151527e-05, "loss": 0.0171, "step": 19790 }, { "grad_norm": 0.16590102016925812, "learning_rate": 8.005625970722119e-05, "loss": 0.0176, "step": 19800 }, { "grad_norm": 0.15985627472400665, "learning_rate": 8.003423216899158e-05, "loss": 0.0183, "step": 19810 }, { "grad_norm": 0.14091581106185913, "learning_rate": 8.001219550715522e-05, "loss": 0.0158, "step": 19820 }, { "grad_norm": 0.15997998416423798, "learning_rate": 7.999014972840632e-05, "loss": 0.0172, "step": 19830 }, { "grad_norm": 0.14372213184833527, "learning_rate": 7.996809483944174e-05, "loss": 0.0161, "step": 19840 }, { "grad_norm": 0.1467854082584381, "learning_rate": 7.994603084696124e-05, "loss": 0.0168, "step": 19850 }, { "grad_norm": 0.1270149052143097, "learning_rate": 7.992395775766724e-05, "loss": 0.0168, "step": 19860 }, { "grad_norm": 0.16234122216701508, "learning_rate": 7.990187557826497e-05, "loss": 0.0171, "step": 19870 }, { "grad_norm": 0.15593156218528748, "learning_rate": 7.987978431546242e-05, "loss": 0.0171, "step": 19880 }, { "grad_norm": 0.23258616030216217, "learning_rate": 7.985768397597031e-05, "loss": 0.0168, "step": 19890 }, { "grad_norm": 0.17183367908000946, "learning_rate": 7.983557456650216e-05, "loss": 0.0169, "step": 19900 }, { "grad_norm": 0.15630951523780823, "learning_rate": 7.981345609377422e-05, "loss": 0.0191, "step": 19910 }, { "grad_norm": 0.13887670636177063, "learning_rate": 7.97913285645055e-05, "loss": 0.0179, "step": 19920 }, { "grad_norm": 0.16426490247249603, "learning_rate": 7.976919198541776e-05, "loss": 0.0207, "step": 19930 }, { "grad_norm": 0.11428659409284592, "learning_rate": 7.974704636323548e-05, "loss": 0.0168, "step": 19940 }, { "grad_norm": 0.16037525236606598, "learning_rate": 7.972489170468597e-05, "loss": 0.0173, "step": 19950 }, { "grad_norm": 0.19330358505249023, "learning_rate": 7.970272801649918e-05, "loss": 0.0201, "step": 19960 }, { "grad_norm": 0.15521469712257385, "learning_rate": 7.96805553054079e-05, "loss": 0.0167, "step": 19970 }, { "grad_norm": 0.11561542749404907, "learning_rate": 7.965837357814756e-05, "loss": 0.0176, "step": 19980 }, { "grad_norm": 0.1404503732919693, "learning_rate": 7.963618284145643e-05, "loss": 0.0153, "step": 19990 }, { "grad_norm": 0.13663038611412048, "learning_rate": 7.961398310207544e-05, "loss": 0.0167, "step": 20000 }, { "grad_norm": 0.10442283749580383, "learning_rate": 7.95917743667483e-05, "loss": 0.0165, "step": 20010 }, { "grad_norm": 0.16094942390918732, "learning_rate": 7.956955664222144e-05, "loss": 0.0169, "step": 20020 }, { "grad_norm": 0.1596398502588272, "learning_rate": 7.954732993524399e-05, "loss": 0.0169, "step": 20030 }, { "grad_norm": 0.16139499843120575, "learning_rate": 7.952509425256786e-05, "loss": 0.0164, "step": 20040 }, { "grad_norm": 0.13502337038516998, "learning_rate": 7.950284960094767e-05, "loss": 0.015, "step": 20050 }, { "grad_norm": 0.13608887791633606, "learning_rate": 7.948059598714076e-05, "loss": 0.016, "step": 20060 }, { "grad_norm": 0.1702307015657425, "learning_rate": 7.945833341790717e-05, "loss": 0.0169, "step": 20070 }, { "grad_norm": 0.14769992232322693, "learning_rate": 7.94360619000097e-05, "loss": 0.0167, "step": 20080 }, { "grad_norm": 0.13465870916843414, "learning_rate": 7.941378144021381e-05, "loss": 0.0163, "step": 20090 }, { "grad_norm": 0.14820978045463562, "learning_rate": 7.939149204528777e-05, "loss": 0.017, "step": 20100 }, { "grad_norm": 0.153712198138237, "learning_rate": 7.936919372200246e-05, "loss": 0.0174, "step": 20110 }, { "grad_norm": 0.13536189496517181, "learning_rate": 7.934688647713158e-05, "loss": 0.0169, "step": 20120 }, { "grad_norm": 0.14004050195217133, "learning_rate": 7.932457031745143e-05, "loss": 0.0174, "step": 20130 }, { "grad_norm": 0.13416919112205505, "learning_rate": 7.930224524974108e-05, "loss": 0.0154, "step": 20140 }, { "grad_norm": 0.14657267928123474, "learning_rate": 7.927991128078232e-05, "loss": 0.0155, "step": 20150 }, { "grad_norm": 0.18203569948673248, "learning_rate": 7.925756841735958e-05, "loss": 0.0181, "step": 20160 }, { "grad_norm": 0.14496992528438568, "learning_rate": 7.923521666626008e-05, "loss": 0.017, "step": 20170 }, { "grad_norm": 0.1578647494316101, "learning_rate": 7.921285603427366e-05, "loss": 0.0168, "step": 20180 }, { "grad_norm": 0.1253279745578766, "learning_rate": 7.91904865281929e-05, "loss": 0.0163, "step": 20190 }, { "grad_norm": 0.14332450926303864, "learning_rate": 7.916810815481307e-05, "loss": 0.0162, "step": 20200 }, { "grad_norm": 0.15613189339637756, "learning_rate": 7.914572092093211e-05, "loss": 0.018, "step": 20210 }, { "grad_norm": 0.1540517956018448, "learning_rate": 7.912332483335068e-05, "loss": 0.0181, "step": 20220 }, { "grad_norm": 0.16431786119937897, "learning_rate": 7.910091989887213e-05, "loss": 0.017, "step": 20230 }, { "grad_norm": 0.1502264142036438, "learning_rate": 7.907850612430248e-05, "loss": 0.0165, "step": 20240 }, { "grad_norm": 0.10734792798757553, "learning_rate": 7.905608351645044e-05, "loss": 0.0167, "step": 20250 }, { "grad_norm": 0.14203225076198578, "learning_rate": 7.90336520821274e-05, "loss": 0.0167, "step": 20260 }, { "grad_norm": 0.11666298657655716, "learning_rate": 7.901121182814746e-05, "loss": 0.0158, "step": 20270 }, { "grad_norm": 0.12608981132507324, "learning_rate": 7.898876276132736e-05, "loss": 0.0155, "step": 20280 }, { "grad_norm": 0.11489413678646088, "learning_rate": 7.896630488848654e-05, "loss": 0.0161, "step": 20290 }, { "grad_norm": 0.14162549376487732, "learning_rate": 7.89438382164471e-05, "loss": 0.0175, "step": 20300 }, { "grad_norm": 0.1526932716369629, "learning_rate": 7.892136275203383e-05, "loss": 0.018, "step": 20310 }, { "grad_norm": 0.12993039190769196, "learning_rate": 7.889887850207418e-05, "loss": 0.0154, "step": 20320 }, { "grad_norm": 0.14560335874557495, "learning_rate": 7.887638547339827e-05, "loss": 0.0166, "step": 20330 }, { "grad_norm": 0.12962156534194946, "learning_rate": 7.885388367283891e-05, "loss": 0.0155, "step": 20340 }, { "grad_norm": 0.12268178910017014, "learning_rate": 7.88313731072315e-05, "loss": 0.0156, "step": 20350 }, { "grad_norm": 0.13534368574619293, "learning_rate": 7.88088537834142e-05, "loss": 0.0164, "step": 20360 }, { "grad_norm": 0.15658266842365265, "learning_rate": 7.878632570822778e-05, "loss": 0.0171, "step": 20370 }, { "grad_norm": 0.16040511429309845, "learning_rate": 7.876378888851567e-05, "loss": 0.018, "step": 20380 }, { "grad_norm": 0.16697491705417633, "learning_rate": 7.874124333112396e-05, "loss": 0.0164, "step": 20390 }, { "grad_norm": 0.19050318002700806, "learning_rate": 7.871868904290138e-05, "loss": 0.017, "step": 20400 }, { "grad_norm": 0.1167401522397995, "learning_rate": 7.869612603069935e-05, "loss": 0.0157, "step": 20410 }, { "grad_norm": 0.19851596653461456, "learning_rate": 7.867355430137192e-05, "loss": 0.0153, "step": 20420 }, { "grad_norm": 0.1447136253118515, "learning_rate": 7.865097386177577e-05, "loss": 0.0162, "step": 20430 }, { "grad_norm": 0.11941421031951904, "learning_rate": 7.862838471877023e-05, "loss": 0.0167, "step": 20440 }, { "grad_norm": 0.12841686606407166, "learning_rate": 7.860578687921731e-05, "loss": 0.0167, "step": 20450 }, { "grad_norm": 0.16882175207138062, "learning_rate": 7.858318034998164e-05, "loss": 0.0158, "step": 20460 }, { "grad_norm": 0.177010178565979, "learning_rate": 7.856056513793046e-05, "loss": 0.0171, "step": 20470 }, { "grad_norm": 0.19331367313861847, "learning_rate": 7.85379412499337e-05, "loss": 0.0194, "step": 20480 }, { "grad_norm": 0.15515536069869995, "learning_rate": 7.851530869286389e-05, "loss": 0.0162, "step": 20490 }, { "grad_norm": 0.12299145013093948, "learning_rate": 7.849266747359619e-05, "loss": 0.0143, "step": 20500 }, { "grad_norm": 0.13760411739349365, "learning_rate": 7.847001759900843e-05, "loss": 0.0149, "step": 20510 }, { "grad_norm": 0.15656259655952454, "learning_rate": 7.844735907598102e-05, "loss": 0.0149, "step": 20520 }, { "grad_norm": 0.14480698108673096, "learning_rate": 7.842469191139703e-05, "loss": 0.0155, "step": 20530 }, { "grad_norm": 0.22269374132156372, "learning_rate": 7.840201611214215e-05, "loss": 0.0166, "step": 20540 }, { "grad_norm": 0.1773843616247177, "learning_rate": 7.837933168510469e-05, "loss": 0.0156, "step": 20550 }, { "grad_norm": 0.15220049023628235, "learning_rate": 7.835663863717559e-05, "loss": 0.0163, "step": 20560 }, { "grad_norm": 0.13410979509353638, "learning_rate": 7.833393697524838e-05, "loss": 0.0157, "step": 20570 }, { "grad_norm": 0.16713345050811768, "learning_rate": 7.831122670621922e-05, "loss": 0.0177, "step": 20580 }, { "grad_norm": 0.13928134739398956, "learning_rate": 7.82885078369869e-05, "loss": 0.0172, "step": 20590 }, { "grad_norm": 0.1266593039035797, "learning_rate": 7.826578037445283e-05, "loss": 0.0164, "step": 20600 }, { "grad_norm": 0.10893411189317703, "learning_rate": 7.824304432552097e-05, "loss": 0.0169, "step": 20610 }, { "grad_norm": 0.21607547998428345, "learning_rate": 7.822029969709798e-05, "loss": 0.0149, "step": 20620 }, { "grad_norm": 0.16667936742305756, "learning_rate": 7.819754649609306e-05, "loss": 0.0166, "step": 20630 }, { "grad_norm": 0.09149135649204254, "learning_rate": 7.817478472941802e-05, "loss": 0.0157, "step": 20640 }, { "grad_norm": 0.13164924085140228, "learning_rate": 7.815201440398727e-05, "loss": 0.0159, "step": 20650 }, { "grad_norm": 0.1510406881570816, "learning_rate": 7.812923552671789e-05, "loss": 0.0168, "step": 20660 }, { "grad_norm": 0.1452333778142929, "learning_rate": 7.810644810452945e-05, "loss": 0.0155, "step": 20670 }, { "grad_norm": 0.15970097482204437, "learning_rate": 7.808365214434417e-05, "loss": 0.0168, "step": 20680 }, { "grad_norm": 0.1743118017911911, "learning_rate": 7.80608476530869e-05, "loss": 0.0183, "step": 20690 }, { "grad_norm": 0.1299510896205902, "learning_rate": 7.8038034637685e-05, "loss": 0.0165, "step": 20700 }, { "grad_norm": 0.15243247151374817, "learning_rate": 7.801521310506848e-05, "loss": 0.018, "step": 20710 }, { "grad_norm": 0.17118072509765625, "learning_rate": 7.799238306216994e-05, "loss": 0.0187, "step": 20720 }, { "grad_norm": 0.1848677545785904, "learning_rate": 7.796954451592448e-05, "loss": 0.0166, "step": 20730 }, { "grad_norm": 0.16708092391490936, "learning_rate": 7.794669747326992e-05, "loss": 0.017, "step": 20740 }, { "grad_norm": 0.142473503947258, "learning_rate": 7.792384194114654e-05, "loss": 0.0168, "step": 20750 }, { "grad_norm": 0.13191570341587067, "learning_rate": 7.790097792649729e-05, "loss": 0.0157, "step": 20760 }, { "grad_norm": 0.15330636501312256, "learning_rate": 7.787810543626762e-05, "loss": 0.0171, "step": 20770 }, { "grad_norm": 0.12461375445127487, "learning_rate": 7.785522447740558e-05, "loss": 0.0157, "step": 20780 }, { "grad_norm": 0.12616316974163055, "learning_rate": 7.783233505686182e-05, "loss": 0.015, "step": 20790 }, { "grad_norm": 0.1091356948018074, "learning_rate": 7.780943718158955e-05, "loss": 0.0155, "step": 20800 }, { "grad_norm": 0.13054943084716797, "learning_rate": 7.778653085854453e-05, "loss": 0.0163, "step": 20810 }, { "grad_norm": 0.17507362365722656, "learning_rate": 7.77636160946851e-05, "loss": 0.0159, "step": 20820 }, { "grad_norm": 0.14576679468154907, "learning_rate": 7.774069289697215e-05, "loss": 0.0174, "step": 20830 }, { "grad_norm": 0.14211009442806244, "learning_rate": 7.771776127236913e-05, "loss": 0.015, "step": 20840 }, { "grad_norm": 0.2057868391275406, "learning_rate": 7.769482122784212e-05, "loss": 0.0172, "step": 20850 }, { "grad_norm": 0.10729412734508514, "learning_rate": 7.767187277035963e-05, "loss": 0.0173, "step": 20860 }, { "grad_norm": 0.14604346454143524, "learning_rate": 7.764891590689285e-05, "loss": 0.017, "step": 20870 }, { "grad_norm": 0.13591526448726654, "learning_rate": 7.762595064441542e-05, "loss": 0.0148, "step": 20880 }, { "grad_norm": 0.09015917032957077, "learning_rate": 7.760297698990362e-05, "loss": 0.0178, "step": 20890 }, { "grad_norm": 0.15870964527130127, "learning_rate": 7.757999495033623e-05, "loss": 0.0163, "step": 20900 }, { "grad_norm": 0.12033005803823471, "learning_rate": 7.755700453269456e-05, "loss": 0.0159, "step": 20910 }, { "grad_norm": 0.12765003740787506, "learning_rate": 7.753400574396254e-05, "loss": 0.0169, "step": 20920 }, { "grad_norm": 0.14100174605846405, "learning_rate": 7.751099859112655e-05, "loss": 0.0158, "step": 20930 }, { "grad_norm": 0.16290126740932465, "learning_rate": 7.748798308117557e-05, "loss": 0.0172, "step": 20940 }, { "grad_norm": 0.1531531661748886, "learning_rate": 7.746495922110112e-05, "loss": 0.0163, "step": 20950 }, { "grad_norm": 0.17133207619190216, "learning_rate": 7.744192701789723e-05, "loss": 0.0159, "step": 20960 }, { "grad_norm": 0.1627005934715271, "learning_rate": 7.741888647856046e-05, "loss": 0.0164, "step": 20970 }, { "grad_norm": 0.14071305096149445, "learning_rate": 7.739583761008994e-05, "loss": 0.0178, "step": 20980 }, { "grad_norm": 0.1253565400838852, "learning_rate": 7.73727804194873e-05, "loss": 0.0157, "step": 20990 }, { "grad_norm": 0.16924230754375458, "learning_rate": 7.734971491375671e-05, "loss": 0.0162, "step": 21000 }, { "grad_norm": 0.14282935857772827, "learning_rate": 7.732664109990485e-05, "loss": 0.0182, "step": 21010 }, { "grad_norm": 0.1435413360595703, "learning_rate": 7.730355898494095e-05, "loss": 0.017, "step": 21020 }, { "grad_norm": 0.11313315480947495, "learning_rate": 7.728046857587673e-05, "loss": 0.0164, "step": 21030 }, { "grad_norm": 0.16372685134410858, "learning_rate": 7.725736987972647e-05, "loss": 0.0172, "step": 21040 }, { "grad_norm": 0.15106415748596191, "learning_rate": 7.723426290350691e-05, "loss": 0.0163, "step": 21050 }, { "grad_norm": 0.17983123660087585, "learning_rate": 7.721114765423736e-05, "loss": 0.0164, "step": 21060 }, { "grad_norm": 0.14677876234054565, "learning_rate": 7.718802413893963e-05, "loss": 0.0155, "step": 21070 }, { "grad_norm": 0.17981624603271484, "learning_rate": 7.716489236463802e-05, "loss": 0.017, "step": 21080 }, { "grad_norm": 0.1561695635318756, "learning_rate": 7.714175233835936e-05, "loss": 0.0167, "step": 21090 }, { "grad_norm": 0.15154369175434113, "learning_rate": 7.711860406713299e-05, "loss": 0.0153, "step": 21100 }, { "grad_norm": 0.13859103620052338, "learning_rate": 7.70954475579907e-05, "loss": 0.0158, "step": 21110 }, { "grad_norm": 0.15795496106147766, "learning_rate": 7.707228281796688e-05, "loss": 0.0154, "step": 21120 }, { "grad_norm": 0.1836368888616562, "learning_rate": 7.704910985409833e-05, "loss": 0.0174, "step": 21130 }, { "grad_norm": 0.15643572807312012, "learning_rate": 7.702592867342439e-05, "loss": 0.0164, "step": 21140 }, { "grad_norm": 0.16182266175746918, "learning_rate": 7.700273928298691e-05, "loss": 0.0167, "step": 21150 }, { "grad_norm": 0.15196631848812103, "learning_rate": 7.697954168983021e-05, "loss": 0.016, "step": 21160 }, { "grad_norm": 0.17934900522232056, "learning_rate": 7.695633590100109e-05, "loss": 0.0165, "step": 21170 }, { "grad_norm": 0.15319840610027313, "learning_rate": 7.693312192354886e-05, "loss": 0.0156, "step": 21180 }, { "grad_norm": 0.18525315821170807, "learning_rate": 7.690989976452532e-05, "loss": 0.0171, "step": 21190 }, { "grad_norm": 0.14564163982868195, "learning_rate": 7.688666943098475e-05, "loss": 0.0156, "step": 21200 }, { "grad_norm": 0.11834743618965149, "learning_rate": 7.686343092998389e-05, "loss": 0.0162, "step": 21210 }, { "grad_norm": 0.15082572400569916, "learning_rate": 7.684018426858202e-05, "loss": 0.0163, "step": 21220 }, { "grad_norm": 0.11912848800420761, "learning_rate": 7.681692945384084e-05, "loss": 0.0165, "step": 21230 }, { "grad_norm": 0.1619499772787094, "learning_rate": 7.679366649282456e-05, "loss": 0.0161, "step": 21240 }, { "grad_norm": 0.14897842705249786, "learning_rate": 7.677039539259983e-05, "loss": 0.0151, "step": 21250 }, { "grad_norm": 0.14898884296417236, "learning_rate": 7.674711616023581e-05, "loss": 0.0149, "step": 21260 }, { "grad_norm": 0.16964808106422424, "learning_rate": 7.672382880280413e-05, "loss": 0.0153, "step": 21270 }, { "grad_norm": 0.19095735251903534, "learning_rate": 7.670053332737885e-05, "loss": 0.016, "step": 21280 }, { "grad_norm": 0.11442751437425613, "learning_rate": 7.667722974103654e-05, "loss": 0.0159, "step": 21290 }, { "grad_norm": 0.12459485977888107, "learning_rate": 7.66539180508562e-05, "loss": 0.0158, "step": 21300 }, { "grad_norm": 0.12710531055927277, "learning_rate": 7.663059826391932e-05, "loss": 0.0179, "step": 21310 }, { "grad_norm": 0.17561843991279602, "learning_rate": 7.660727038730981e-05, "loss": 0.0153, "step": 21320 }, { "grad_norm": 0.15498638153076172, "learning_rate": 7.65839344281141e-05, "loss": 0.0148, "step": 21330 }, { "grad_norm": 0.14678983390331268, "learning_rate": 7.656059039342101e-05, "loss": 0.0168, "step": 21340 }, { "grad_norm": 0.1639057844877243, "learning_rate": 7.653723829032187e-05, "loss": 0.0166, "step": 21350 }, { "grad_norm": 0.19102337956428528, "learning_rate": 7.65138781259104e-05, "loss": 0.0173, "step": 21360 }, { "grad_norm": 0.19636496901512146, "learning_rate": 7.649050990728279e-05, "loss": 0.0155, "step": 21370 }, { "grad_norm": 0.12973672151565552, "learning_rate": 7.646713364153774e-05, "loss": 0.0162, "step": 21380 }, { "grad_norm": 0.13515767455101013, "learning_rate": 7.64437493357763e-05, "loss": 0.0167, "step": 21390 }, { "grad_norm": 0.1337616741657257, "learning_rate": 7.642035699710202e-05, "loss": 0.0179, "step": 21400 }, { "grad_norm": 0.1314256340265274, "learning_rate": 7.639695663262089e-05, "loss": 0.0163, "step": 21410 }, { "grad_norm": 0.17424185574054718, "learning_rate": 7.637354824944128e-05, "loss": 0.0152, "step": 21420 }, { "grad_norm": 0.14942826330661774, "learning_rate": 7.635013185467408e-05, "loss": 0.0151, "step": 21430 }, { "grad_norm": 0.1336805522441864, "learning_rate": 7.632670745543256e-05, "loss": 0.0159, "step": 21440 }, { "grad_norm": 0.15946343541145325, "learning_rate": 7.630327505883242e-05, "loss": 0.0159, "step": 21450 }, { "grad_norm": 0.143310546875, "learning_rate": 7.627983467199182e-05, "loss": 0.0164, "step": 21460 }, { "grad_norm": 0.13095837831497192, "learning_rate": 7.625638630203132e-05, "loss": 0.0177, "step": 21470 }, { "grad_norm": 0.1594979166984558, "learning_rate": 7.623292995607394e-05, "loss": 0.0154, "step": 21480 }, { "grad_norm": 0.14474527537822723, "learning_rate": 7.620946564124507e-05, "loss": 0.016, "step": 21490 }, { "grad_norm": 0.16005025804042816, "learning_rate": 7.618599336467256e-05, "loss": 0.0156, "step": 21500 }, { "grad_norm": 0.15429633855819702, "learning_rate": 7.616251313348666e-05, "loss": 0.0155, "step": 21510 }, { "grad_norm": 0.16238924860954285, "learning_rate": 7.613902495482005e-05, "loss": 0.0174, "step": 21520 }, { "grad_norm": 0.13891378045082092, "learning_rate": 7.611552883580784e-05, "loss": 0.016, "step": 21530 }, { "grad_norm": 0.14432859420776367, "learning_rate": 7.609202478358748e-05, "loss": 0.0151, "step": 21540 }, { "grad_norm": 0.13155576586723328, "learning_rate": 7.606851280529895e-05, "loss": 0.0157, "step": 21550 }, { "grad_norm": 0.1338668167591095, "learning_rate": 7.604499290808449e-05, "loss": 0.0154, "step": 21560 }, { "grad_norm": 0.13660474121570587, "learning_rate": 7.602146509908888e-05, "loss": 0.0162, "step": 21570 }, { "grad_norm": 0.14064179360866547, "learning_rate": 7.599792938545921e-05, "loss": 0.0162, "step": 21580 }, { "grad_norm": 0.1611737161874771, "learning_rate": 7.597438577434506e-05, "loss": 0.0158, "step": 21590 }, { "grad_norm": 0.15534789860248566, "learning_rate": 7.595083427289831e-05, "loss": 0.0155, "step": 21600 }, { "grad_norm": 0.12324900180101395, "learning_rate": 7.59272748882733e-05, "loss": 0.0165, "step": 21610 }, { "grad_norm": 0.14111334085464478, "learning_rate": 7.590370762762675e-05, "loss": 0.0155, "step": 21620 }, { "grad_norm": 0.1471528261899948, "learning_rate": 7.588013249811777e-05, "loss": 0.0156, "step": 21630 }, { "grad_norm": 0.17576511204242706, "learning_rate": 7.585654950690786e-05, "loss": 0.0153, "step": 21640 }, { "grad_norm": 0.12911246716976166, "learning_rate": 7.583295866116091e-05, "loss": 0.0159, "step": 21650 }, { "grad_norm": 0.10703636705875397, "learning_rate": 7.580935996804321e-05, "loss": 0.0147, "step": 21660 }, { "grad_norm": 0.18186138570308685, "learning_rate": 7.57857534347234e-05, "loss": 0.015, "step": 21670 }, { "grad_norm": 0.1449519544839859, "learning_rate": 7.576213906837254e-05, "loss": 0.0156, "step": 21680 }, { "grad_norm": 0.1187816709280014, "learning_rate": 7.573851687616403e-05, "loss": 0.0135, "step": 21690 }, { "grad_norm": 0.13698357343673706, "learning_rate": 7.571488686527368e-05, "loss": 0.0168, "step": 21700 }, { "grad_norm": 0.13713209331035614, "learning_rate": 7.569124904287968e-05, "loss": 0.0153, "step": 21710 }, { "grad_norm": 0.17456135153770447, "learning_rate": 7.566760341616254e-05, "loss": 0.017, "step": 21720 }, { "grad_norm": 0.23437562584877014, "learning_rate": 7.564394999230519e-05, "loss": 0.0168, "step": 21730 }, { "grad_norm": 0.1322067528963089, "learning_rate": 7.562028877849294e-05, "loss": 0.0159, "step": 21740 }, { "grad_norm": 0.14238502085208893, "learning_rate": 7.559661978191341e-05, "loss": 0.0185, "step": 21750 }, { "grad_norm": 0.14026497304439545, "learning_rate": 7.557294300975664e-05, "loss": 0.0173, "step": 21760 }, { "grad_norm": 0.12039074301719666, "learning_rate": 7.554925846921499e-05, "loss": 0.0156, "step": 21770 }, { "grad_norm": 0.11091282963752747, "learning_rate": 7.552556616748321e-05, "loss": 0.0179, "step": 21780 }, { "grad_norm": 0.1313733160495758, "learning_rate": 7.550186611175838e-05, "loss": 0.0156, "step": 21790 }, { "grad_norm": 0.14460667967796326, "learning_rate": 7.547815830923998e-05, "loss": 0.0156, "step": 21800 }, { "grad_norm": 0.191311314702034, "learning_rate": 7.54544427671298e-05, "loss": 0.0165, "step": 21810 }, { "grad_norm": 0.16732260584831238, "learning_rate": 7.543071949263198e-05, "loss": 0.0153, "step": 21820 }, { "grad_norm": 0.19138768315315247, "learning_rate": 7.540698849295305e-05, "loss": 0.0156, "step": 21830 }, { "grad_norm": 0.19327177107334137, "learning_rate": 7.538324977530183e-05, "loss": 0.0158, "step": 21840 }, { "grad_norm": 0.11378506571054459, "learning_rate": 7.535950334688955e-05, "loss": 0.0169, "step": 21850 }, { "grad_norm": 0.16245345771312714, "learning_rate": 7.533574921492972e-05, "loss": 0.0159, "step": 21860 }, { "grad_norm": 0.12186422944068909, "learning_rate": 7.531198738663824e-05, "loss": 0.0161, "step": 21870 }, { "grad_norm": 0.13833458721637726, "learning_rate": 7.528821786923333e-05, "loss": 0.0148, "step": 21880 }, { "grad_norm": 0.17040807008743286, "learning_rate": 7.52644406699355e-05, "loss": 0.0144, "step": 21890 }, { "grad_norm": 0.10093621909618378, "learning_rate": 7.524065579596766e-05, "loss": 0.0145, "step": 21900 }, { "grad_norm": 0.1489272564649582, "learning_rate": 7.521686325455506e-05, "loss": 0.0161, "step": 21910 }, { "grad_norm": 0.13570107519626617, "learning_rate": 7.51930630529252e-05, "loss": 0.0153, "step": 21920 }, { "grad_norm": 0.18888375163078308, "learning_rate": 7.516925519830797e-05, "loss": 0.0178, "step": 21930 }, { "grad_norm": 0.16515080630779266, "learning_rate": 7.514543969793557e-05, "loss": 0.0161, "step": 21940 }, { "grad_norm": 0.12233969569206238, "learning_rate": 7.512161655904251e-05, "loss": 0.0149, "step": 21950 }, { "grad_norm": 0.11930069327354431, "learning_rate": 7.509778578886563e-05, "loss": 0.0142, "step": 21960 }, { "grad_norm": 0.15218539535999298, "learning_rate": 7.507394739464412e-05, "loss": 0.0159, "step": 21970 }, { "grad_norm": 0.1362581104040146, "learning_rate": 7.50501013836194e-05, "loss": 0.0158, "step": 21980 }, { "grad_norm": 0.12008487433195114, "learning_rate": 7.50262477630353e-05, "loss": 0.0156, "step": 21990 }, { "grad_norm": 0.1444818526506424, "learning_rate": 7.500238654013794e-05, "loss": 0.0164, "step": 22000 }, { "grad_norm": 0.13979201018810272, "learning_rate": 7.497851772217566e-05, "loss": 0.0151, "step": 22010 }, { "grad_norm": 0.14000213146209717, "learning_rate": 7.495464131639924e-05, "loss": 0.0166, "step": 22020 }, { "grad_norm": 0.11631902307271957, "learning_rate": 7.493075733006166e-05, "loss": 0.0159, "step": 22030 }, { "grad_norm": 0.16735002398490906, "learning_rate": 7.490686577041828e-05, "loss": 0.0165, "step": 22040 }, { "grad_norm": 0.15002210438251495, "learning_rate": 7.488296664472668e-05, "loss": 0.0161, "step": 22050 }, { "grad_norm": 0.10318573564291, "learning_rate": 7.485905996024682e-05, "loss": 0.0148, "step": 22060 }, { "grad_norm": 0.1302022486925125, "learning_rate": 7.483514572424093e-05, "loss": 0.0158, "step": 22070 }, { "grad_norm": 0.13579973578453064, "learning_rate": 7.481122394397349e-05, "loss": 0.0164, "step": 22080 }, { "grad_norm": 0.12642227113246918, "learning_rate": 7.478729462671131e-05, "loss": 0.0167, "step": 22090 }, { "grad_norm": 0.16332171857357025, "learning_rate": 7.47633577797235e-05, "loss": 0.016, "step": 22100 }, { "grad_norm": 0.16210277378559113, "learning_rate": 7.473941341028144e-05, "loss": 0.0151, "step": 22110 }, { "grad_norm": 0.10208547860383987, "learning_rate": 7.471546152565879e-05, "loss": 0.0149, "step": 22120 }, { "grad_norm": 0.11225123703479767, "learning_rate": 7.46915021331315e-05, "loss": 0.0154, "step": 22130 }, { "grad_norm": 0.1589273363351822, "learning_rate": 7.466753523997778e-05, "loss": 0.0144, "step": 22140 }, { "grad_norm": 0.15485580265522003, "learning_rate": 7.464356085347819e-05, "loss": 0.0155, "step": 22150 }, { "grad_norm": 0.1427338421344757, "learning_rate": 7.461957898091548e-05, "loss": 0.0166, "step": 22160 }, { "grad_norm": 0.19351637363433838, "learning_rate": 7.459558962957473e-05, "loss": 0.0169, "step": 22170 }, { "grad_norm": 0.1399211287498474, "learning_rate": 7.457159280674326e-05, "loss": 0.0159, "step": 22180 }, { "grad_norm": 0.1464521735906601, "learning_rate": 7.454758851971066e-05, "loss": 0.0148, "step": 22190 }, { "grad_norm": 0.11952314525842667, "learning_rate": 7.45235767757688e-05, "loss": 0.0158, "step": 22200 }, { "grad_norm": 0.16247905790805817, "learning_rate": 7.449955758221183e-05, "loss": 0.0171, "step": 22210 }, { "grad_norm": 0.13141930103302002, "learning_rate": 7.447553094633615e-05, "loss": 0.0148, "step": 22220 }, { "grad_norm": 0.14372862875461578, "learning_rate": 7.445149687544039e-05, "loss": 0.0157, "step": 22230 }, { "grad_norm": 0.1376771330833435, "learning_rate": 7.44274553768255e-05, "loss": 0.0147, "step": 22240 }, { "grad_norm": 0.17815595865249634, "learning_rate": 7.440340645779464e-05, "loss": 0.0147, "step": 22250 }, { "grad_norm": 0.1688447743654251, "learning_rate": 7.437935012565322e-05, "loss": 0.0151, "step": 22260 }, { "grad_norm": 0.14238017797470093, "learning_rate": 7.435528638770893e-05, "loss": 0.0155, "step": 22270 }, { "grad_norm": 0.1271003782749176, "learning_rate": 7.433121525127171e-05, "loss": 0.0164, "step": 22280 }, { "grad_norm": 0.12493068724870682, "learning_rate": 7.430713672365371e-05, "loss": 0.0154, "step": 22290 }, { "grad_norm": 0.12806792557239532, "learning_rate": 7.428305081216938e-05, "loss": 0.0165, "step": 22300 }, { "grad_norm": 0.10768955200910568, "learning_rate": 7.425895752413536e-05, "loss": 0.0151, "step": 22310 }, { "grad_norm": 0.1401643455028534, "learning_rate": 7.423485686687057e-05, "loss": 0.0152, "step": 22320 }, { "grad_norm": 0.12901321053504944, "learning_rate": 7.421074884769616e-05, "loss": 0.0168, "step": 22330 }, { "grad_norm": 0.13048218190670013, "learning_rate": 7.418663347393548e-05, "loss": 0.0149, "step": 22340 }, { "grad_norm": 0.12500008940696716, "learning_rate": 7.416251075291418e-05, "loss": 0.0167, "step": 22350 }, { "grad_norm": 0.11420731246471405, "learning_rate": 7.413838069196007e-05, "loss": 0.0141, "step": 22360 }, { "grad_norm": 0.15969106554985046, "learning_rate": 7.411424329840324e-05, "loss": 0.0164, "step": 22370 }, { "grad_norm": 0.18551531434059143, "learning_rate": 7.409009857957601e-05, "loss": 0.0158, "step": 22380 }, { "grad_norm": 0.14582133293151855, "learning_rate": 7.40659465428129e-05, "loss": 0.017, "step": 22390 }, { "grad_norm": 0.13852602243423462, "learning_rate": 7.404178719545063e-05, "loss": 0.0165, "step": 22400 }, { "grad_norm": 0.13065893948078156, "learning_rate": 7.401762054482822e-05, "loss": 0.0172, "step": 22410 }, { "grad_norm": 0.14783795177936554, "learning_rate": 7.39934465982868e-05, "loss": 0.0173, "step": 22420 }, { "grad_norm": 0.14460934698581696, "learning_rate": 7.396926536316984e-05, "loss": 0.015, "step": 22430 }, { "grad_norm": 0.17236049473285675, "learning_rate": 7.394507684682293e-05, "loss": 0.0173, "step": 22440 }, { "grad_norm": 0.13066741824150085, "learning_rate": 7.392088105659393e-05, "loss": 0.0167, "step": 22450 }, { "grad_norm": 0.15916860103607178, "learning_rate": 7.389667799983284e-05, "loss": 0.0165, "step": 22460 }, { "grad_norm": 0.12872053682804108, "learning_rate": 7.387246768389193e-05, "loss": 0.0151, "step": 22470 }, { "grad_norm": 0.11178658157587051, "learning_rate": 7.384825011612563e-05, "loss": 0.0144, "step": 22480 }, { "grad_norm": 0.17505145072937012, "learning_rate": 7.382402530389066e-05, "loss": 0.0143, "step": 22490 }, { "grad_norm": 0.12421488761901855, "learning_rate": 7.379979325454582e-05, "loss": 0.0152, "step": 22500 }, { "grad_norm": 0.19548317790031433, "learning_rate": 7.37755539754522e-05, "loss": 0.0162, "step": 22510 }, { "grad_norm": 0.1806824505329132, "learning_rate": 7.375130747397302e-05, "loss": 0.0173, "step": 22520 }, { "grad_norm": 0.16203494369983673, "learning_rate": 7.372705375747377e-05, "loss": 0.014, "step": 22530 }, { "grad_norm": 0.2023986130952835, "learning_rate": 7.370279283332205e-05, "loss": 0.0164, "step": 22540 }, { "grad_norm": 0.1363319456577301, "learning_rate": 7.36785247088877e-05, "loss": 0.0158, "step": 22550 }, { "grad_norm": 0.14770320057868958, "learning_rate": 7.365424939154275e-05, "loss": 0.0149, "step": 22560 }, { "grad_norm": 0.16442622244358063, "learning_rate": 7.362996688866138e-05, "loss": 0.0153, "step": 22570 }, { "grad_norm": 0.1950012892484665, "learning_rate": 7.360567720761999e-05, "loss": 0.016, "step": 22580 }, { "grad_norm": 0.14236487448215485, "learning_rate": 7.358138035579711e-05, "loss": 0.0148, "step": 22590 }, { "grad_norm": 0.15577837824821472, "learning_rate": 7.355707634057354e-05, "loss": 0.0152, "step": 22600 }, { "grad_norm": 0.13771292567253113, "learning_rate": 7.353276516933215e-05, "loss": 0.0183, "step": 22610 }, { "grad_norm": 0.13554902374744415, "learning_rate": 7.350844684945806e-05, "loss": 0.0169, "step": 22620 }, { "grad_norm": 0.18045207858085632, "learning_rate": 7.348412138833851e-05, "loss": 0.0161, "step": 22630 }, { "grad_norm": 0.14033959805965424, "learning_rate": 7.345978879336295e-05, "loss": 0.0148, "step": 22640 }, { "grad_norm": 0.17879468202590942, "learning_rate": 7.343544907192296e-05, "loss": 0.0167, "step": 22650 }, { "grad_norm": 0.1713511347770691, "learning_rate": 7.341110223141235e-05, "loss": 0.0168, "step": 22660 }, { "grad_norm": 0.1345522403717041, "learning_rate": 7.3386748279227e-05, "loss": 0.0169, "step": 22670 }, { "grad_norm": 0.14886410534381866, "learning_rate": 7.336238722276501e-05, "loss": 0.015, "step": 22680 }, { "grad_norm": 0.1521235555410385, "learning_rate": 7.333801906942663e-05, "loss": 0.017, "step": 22690 }, { "grad_norm": 0.13093332946300507, "learning_rate": 7.331364382661428e-05, "loss": 0.0169, "step": 22700 }, { "grad_norm": 0.14056967198848724, "learning_rate": 7.328926150173248e-05, "loss": 0.0184, "step": 22710 }, { "grad_norm": 0.1518784463405609, "learning_rate": 7.326487210218795e-05, "loss": 0.0165, "step": 22720 }, { "grad_norm": 0.15561331808567047, "learning_rate": 7.324047563538955e-05, "loss": 0.0155, "step": 22730 }, { "grad_norm": 0.14480867981910706, "learning_rate": 7.321607210874828e-05, "loss": 0.0173, "step": 22740 }, { "grad_norm": 0.16603484749794006, "learning_rate": 7.31916615296773e-05, "loss": 0.016, "step": 22750 }, { "grad_norm": 0.19833211600780487, "learning_rate": 7.316724390559188e-05, "loss": 0.0175, "step": 22760 }, { "grad_norm": 0.14573749899864197, "learning_rate": 7.314281924390946e-05, "loss": 0.0169, "step": 22770 }, { "grad_norm": 0.17846426367759705, "learning_rate": 7.311838755204959e-05, "loss": 0.0162, "step": 22780 }, { "grad_norm": 0.1342628300189972, "learning_rate": 7.3093948837434e-05, "loss": 0.0151, "step": 22790 }, { "grad_norm": 0.1693459153175354, "learning_rate": 7.306950310748651e-05, "loss": 0.0147, "step": 22800 }, { "grad_norm": 0.13928698003292084, "learning_rate": 7.304505036963311e-05, "loss": 0.0157, "step": 22810 }, { "grad_norm": 0.11785650998353958, "learning_rate": 7.302059063130186e-05, "loss": 0.0139, "step": 22820 }, { "grad_norm": 0.12002760916948318, "learning_rate": 7.2996123899923e-05, "loss": 0.0147, "step": 22830 }, { "grad_norm": 0.1491849273443222, "learning_rate": 7.297165018292886e-05, "loss": 0.0142, "step": 22840 }, { "grad_norm": 0.12174184620380402, "learning_rate": 7.294716948775396e-05, "loss": 0.0148, "step": 22850 }, { "grad_norm": 0.15187907218933105, "learning_rate": 7.292268182183484e-05, "loss": 0.0151, "step": 22860 }, { "grad_norm": 0.13952019810676575, "learning_rate": 7.28981871926102e-05, "loss": 0.0152, "step": 22870 }, { "grad_norm": 0.12833110988140106, "learning_rate": 7.28736856075209e-05, "loss": 0.0143, "step": 22880 }, { "grad_norm": 0.11547885090112686, "learning_rate": 7.284917707400985e-05, "loss": 0.0154, "step": 22890 }, { "grad_norm": 0.13767002522945404, "learning_rate": 7.282466159952212e-05, "loss": 0.0166, "step": 22900 }, { "grad_norm": 0.1447310447692871, "learning_rate": 7.280013919150483e-05, "loss": 0.0179, "step": 22910 }, { "grad_norm": 0.13281172513961792, "learning_rate": 7.277560985740728e-05, "loss": 0.0169, "step": 22920 }, { "grad_norm": 0.16000992059707642, "learning_rate": 7.275107360468079e-05, "loss": 0.0188, "step": 22930 }, { "grad_norm": 0.15275979042053223, "learning_rate": 7.272653044077885e-05, "loss": 0.0162, "step": 22940 }, { "grad_norm": 0.15955378115177155, "learning_rate": 7.270198037315703e-05, "loss": 0.0157, "step": 22950 }, { "grad_norm": 0.13666319847106934, "learning_rate": 7.267742340927297e-05, "loss": 0.0165, "step": 22960 }, { "grad_norm": 0.11717550456523895, "learning_rate": 7.265285955658645e-05, "loss": 0.0156, "step": 22970 }, { "grad_norm": 0.11569499969482422, "learning_rate": 7.26282888225593e-05, "loss": 0.0163, "step": 22980 }, { "grad_norm": 0.10758174955844879, "learning_rate": 7.260371121465548e-05, "loss": 0.0159, "step": 22990 }, { "grad_norm": 0.15655909478664398, "learning_rate": 7.2579126740341e-05, "loss": 0.0156, "step": 23000 }, { "grad_norm": 0.14527031779289246, "learning_rate": 7.2554535407084e-05, "loss": 0.0148, "step": 23010 }, { "grad_norm": 0.15783196687698364, "learning_rate": 7.252993722235464e-05, "loss": 0.0154, "step": 23020 }, { "grad_norm": 0.12609219551086426, "learning_rate": 7.250533219362523e-05, "loss": 0.0161, "step": 23030 }, { "grad_norm": 0.14424355328083038, "learning_rate": 7.248072032837012e-05, "loss": 0.016, "step": 23040 }, { "grad_norm": 0.15460063517093658, "learning_rate": 7.245610163406575e-05, "loss": 0.0169, "step": 23050 }, { "grad_norm": 0.15033333003520966, "learning_rate": 7.243147611819061e-05, "loss": 0.0149, "step": 23060 }, { "grad_norm": 0.14878469705581665, "learning_rate": 7.240684378822531e-05, "loss": 0.0152, "step": 23070 }, { "grad_norm": 0.1296839714050293, "learning_rate": 7.238220465165248e-05, "loss": 0.0151, "step": 23080 }, { "grad_norm": 0.12935301661491394, "learning_rate": 7.235755871595684e-05, "loss": 0.0157, "step": 23090 }, { "grad_norm": 0.10890597105026245, "learning_rate": 7.233290598862517e-05, "loss": 0.0161, "step": 23100 }, { "grad_norm": 0.1458607167005539, "learning_rate": 7.230824647714635e-05, "loss": 0.016, "step": 23110 }, { "grad_norm": 0.15057748556137085, "learning_rate": 7.228358018901124e-05, "loss": 0.0164, "step": 23120 }, { "grad_norm": 0.1576462835073471, "learning_rate": 7.225890713171286e-05, "loss": 0.0162, "step": 23130 }, { "grad_norm": 0.18061427772045135, "learning_rate": 7.223422731274618e-05, "loss": 0.0146, "step": 23140 }, { "grad_norm": 0.1260305792093277, "learning_rate": 7.220954073960832e-05, "loss": 0.017, "step": 23150 }, { "grad_norm": 0.15426026284694672, "learning_rate": 7.218484741979838e-05, "loss": 0.0165, "step": 23160 }, { "grad_norm": 0.14842012524604797, "learning_rate": 7.216014736081756e-05, "loss": 0.015, "step": 23170 }, { "grad_norm": 0.148368239402771, "learning_rate": 7.213544057016906e-05, "loss": 0.0154, "step": 23180 }, { "grad_norm": 0.15344922244548798, "learning_rate": 7.211072705535819e-05, "loss": 0.0157, "step": 23190 }, { "grad_norm": 0.1533629447221756, "learning_rate": 7.208600682389224e-05, "loss": 0.0147, "step": 23200 }, { "grad_norm": 0.11811333149671555, "learning_rate": 7.206127988328055e-05, "loss": 0.014, "step": 23210 }, { "grad_norm": 0.11474567651748657, "learning_rate": 7.203654624103453e-05, "loss": 0.0162, "step": 23220 }, { "grad_norm": 0.15110738575458527, "learning_rate": 7.201180590466761e-05, "loss": 0.0145, "step": 23230 }, { "grad_norm": 0.1920979917049408, "learning_rate": 7.198705888169523e-05, "loss": 0.0155, "step": 23240 }, { "grad_norm": 0.16238120198249817, "learning_rate": 7.196230517963491e-05, "loss": 0.015, "step": 23250 }, { "grad_norm": 0.1356951743364334, "learning_rate": 7.193754480600615e-05, "loss": 0.0152, "step": 23260 }, { "grad_norm": 0.111125148832798, "learning_rate": 7.19127777683305e-05, "loss": 0.0143, "step": 23270 }, { "grad_norm": 0.15892057120800018, "learning_rate": 7.188800407413156e-05, "loss": 0.0163, "step": 23280 }, { "grad_norm": 0.13304315507411957, "learning_rate": 7.186322373093489e-05, "loss": 0.0148, "step": 23290 }, { "grad_norm": 0.14843633770942688, "learning_rate": 7.18384367462681e-05, "loss": 0.0149, "step": 23300 }, { "grad_norm": 0.09539689123630524, "learning_rate": 7.181364312766085e-05, "loss": 0.0157, "step": 23310 }, { "grad_norm": 0.12376078963279724, "learning_rate": 7.178884288264477e-05, "loss": 0.0146, "step": 23320 }, { "grad_norm": 0.15702535212039948, "learning_rate": 7.176403601875353e-05, "loss": 0.0148, "step": 23330 }, { "grad_norm": 0.14357459545135498, "learning_rate": 7.173922254352279e-05, "loss": 0.0166, "step": 23340 }, { "grad_norm": 0.11890248954296112, "learning_rate": 7.171440246449024e-05, "loss": 0.0146, "step": 23350 }, { "grad_norm": 0.15929991006851196, "learning_rate": 7.168957578919555e-05, "loss": 0.0149, "step": 23360 }, { "grad_norm": 0.10694242268800735, "learning_rate": 7.16647425251804e-05, "loss": 0.0151, "step": 23370 }, { "grad_norm": 0.16668719053268433, "learning_rate": 7.163990267998852e-05, "loss": 0.0146, "step": 23380 }, { "grad_norm": 0.1996365189552307, "learning_rate": 7.161505626116556e-05, "loss": 0.0155, "step": 23390 }, { "grad_norm": 0.13602328300476074, "learning_rate": 7.159020327625923e-05, "loss": 0.015, "step": 23400 }, { "grad_norm": 0.13140715658664703, "learning_rate": 7.15653437328192e-05, "loss": 0.0156, "step": 23410 }, { "grad_norm": 0.1212695986032486, "learning_rate": 7.154047763839713e-05, "loss": 0.017, "step": 23420 }, { "grad_norm": 0.118385910987854, "learning_rate": 7.15156050005467e-05, "loss": 0.0164, "step": 23430 }, { "grad_norm": 0.12978337705135345, "learning_rate": 7.149072582682357e-05, "loss": 0.015, "step": 23440 }, { "grad_norm": 0.14538291096687317, "learning_rate": 7.146584012478535e-05, "loss": 0.0158, "step": 23450 }, { "grad_norm": 0.1611538976430893, "learning_rate": 7.144094790199169e-05, "loss": 0.0163, "step": 23460 }, { "grad_norm": 0.15272387862205505, "learning_rate": 7.141604916600415e-05, "loss": 0.017, "step": 23470 }, { "grad_norm": 0.18747879564762115, "learning_rate": 7.139114392438635e-05, "loss": 0.016, "step": 23480 }, { "grad_norm": 0.1214374452829361, "learning_rate": 7.136623218470382e-05, "loss": 0.014, "step": 23490 }, { "grad_norm": 0.13278457522392273, "learning_rate": 7.13413139545241e-05, "loss": 0.0157, "step": 23500 }, { "grad_norm": 0.11479038745164871, "learning_rate": 7.131638924141668e-05, "loss": 0.016, "step": 23510 }, { "grad_norm": 0.1751219928264618, "learning_rate": 7.129145805295304e-05, "loss": 0.0153, "step": 23520 }, { "grad_norm": 0.1386396884918213, "learning_rate": 7.126652039670661e-05, "loss": 0.0157, "step": 23530 }, { "grad_norm": 0.13280238211154938, "learning_rate": 7.124157628025278e-05, "loss": 0.0169, "step": 23540 }, { "grad_norm": 0.1333092451095581, "learning_rate": 7.121662571116894e-05, "loss": 0.0148, "step": 23550 }, { "grad_norm": 0.11542718857526779, "learning_rate": 7.119166869703441e-05, "loss": 0.0157, "step": 23560 }, { "grad_norm": 0.13332056999206543, "learning_rate": 7.116670524543044e-05, "loss": 0.0164, "step": 23570 }, { "grad_norm": 0.1613406240940094, "learning_rate": 7.114173536394032e-05, "loss": 0.0152, "step": 23580 }, { "grad_norm": 0.1496785283088684, "learning_rate": 7.111675906014917e-05, "loss": 0.0154, "step": 23590 }, { "grad_norm": 0.15425457060337067, "learning_rate": 7.109177634164421e-05, "loss": 0.0162, "step": 23600 }, { "grad_norm": 0.1740262508392334, "learning_rate": 7.106678721601449e-05, "loss": 0.0145, "step": 23610 }, { "grad_norm": 0.13920289278030396, "learning_rate": 7.104179169085103e-05, "loss": 0.0167, "step": 23620 }, { "grad_norm": 0.12565596401691437, "learning_rate": 7.101678977374683e-05, "loss": 0.0143, "step": 23630 }, { "grad_norm": 0.16589482128620148, "learning_rate": 7.099178147229685e-05, "loss": 0.0145, "step": 23640 }, { "grad_norm": 0.22040744125843048, "learning_rate": 7.096676679409789e-05, "loss": 0.0156, "step": 23650 }, { "grad_norm": 0.15430815517902374, "learning_rate": 7.094174574674877e-05, "loss": 0.0177, "step": 23660 }, { "grad_norm": 0.11731414496898651, "learning_rate": 7.091671833785025e-05, "loss": 0.0147, "step": 23670 }, { "grad_norm": 0.12272735685110092, "learning_rate": 7.089168457500493e-05, "loss": 0.0138, "step": 23680 }, { "grad_norm": 0.12449896335601807, "learning_rate": 7.086664446581747e-05, "loss": 0.0139, "step": 23690 }, { "grad_norm": 0.1449020951986313, "learning_rate": 7.084159801789438e-05, "loss": 0.0152, "step": 23700 }, { "grad_norm": 0.14557400345802307, "learning_rate": 7.081654523884411e-05, "loss": 0.0149, "step": 23710 }, { "grad_norm": 0.11788138747215271, "learning_rate": 7.0791486136277e-05, "loss": 0.0149, "step": 23720 }, { "grad_norm": 0.10622803121805191, "learning_rate": 7.07664207178054e-05, "loss": 0.0154, "step": 23730 }, { "grad_norm": 0.11996057629585266, "learning_rate": 7.074134899104345e-05, "loss": 0.0145, "step": 23740 }, { "grad_norm": 0.1296061873435974, "learning_rate": 7.071627096360735e-05, "loss": 0.0161, "step": 23750 }, { "grad_norm": 0.17812217772006989, "learning_rate": 7.069118664311511e-05, "loss": 0.0171, "step": 23760 }, { "grad_norm": 0.1289563626050949, "learning_rate": 7.06660960371867e-05, "loss": 0.0143, "step": 23770 }, { "grad_norm": 0.12483984977006912, "learning_rate": 7.064099915344396e-05, "loss": 0.0142, "step": 23780 }, { "grad_norm": 0.14612038433551788, "learning_rate": 7.061589599951066e-05, "loss": 0.0148, "step": 23790 }, { "grad_norm": 0.10147907584905624, "learning_rate": 7.05907865830125e-05, "loss": 0.0149, "step": 23800 }, { "grad_norm": 0.18411338329315186, "learning_rate": 7.056567091157703e-05, "loss": 0.0161, "step": 23810 }, { "grad_norm": 0.13409295678138733, "learning_rate": 7.054054899283375e-05, "loss": 0.0159, "step": 23820 }, { "grad_norm": 0.15733815729618073, "learning_rate": 7.051542083441403e-05, "loss": 0.0161, "step": 23830 }, { "grad_norm": 0.17750674486160278, "learning_rate": 7.049028644395113e-05, "loss": 0.0157, "step": 23840 }, { "grad_norm": 0.14815104007720947, "learning_rate": 7.046514582908024e-05, "loss": 0.0171, "step": 23850 }, { "grad_norm": 0.17626193165779114, "learning_rate": 7.043999899743838e-05, "loss": 0.016, "step": 23860 }, { "grad_norm": 0.13830138742923737, "learning_rate": 7.041484595666451e-05, "loss": 0.0147, "step": 23870 }, { "grad_norm": 0.15032213926315308, "learning_rate": 7.038968671439948e-05, "loss": 0.0138, "step": 23880 }, { "grad_norm": 0.13153432309627533, "learning_rate": 7.036452127828596e-05, "loss": 0.0145, "step": 23890 }, { "grad_norm": 0.14863787591457367, "learning_rate": 7.033934965596859e-05, "loss": 0.0148, "step": 23900 }, { "grad_norm": 0.103289894759655, "learning_rate": 7.031417185509381e-05, "loss": 0.0149, "step": 23910 }, { "grad_norm": 0.12403608858585358, "learning_rate": 7.028898788331e-05, "loss": 0.0136, "step": 23920 }, { "grad_norm": 0.1603798121213913, "learning_rate": 7.026379774826736e-05, "loss": 0.0148, "step": 23930 }, { "grad_norm": 0.11158356070518494, "learning_rate": 7.0238601457618e-05, "loss": 0.0137, "step": 23940 }, { "grad_norm": 0.11662998050451279, "learning_rate": 7.02133990190159e-05, "loss": 0.0153, "step": 23950 }, { "grad_norm": 0.1335730105638504, "learning_rate": 7.018819044011687e-05, "loss": 0.0155, "step": 23960 }, { "grad_norm": 0.1460486501455307, "learning_rate": 7.016297572857863e-05, "loss": 0.0165, "step": 23970 }, { "grad_norm": 0.13645152747631073, "learning_rate": 7.013775489206072e-05, "loss": 0.0151, "step": 23980 }, { "grad_norm": 0.1340470314025879, "learning_rate": 7.01125279382246e-05, "loss": 0.0141, "step": 23990 }, { "grad_norm": 0.10482069104909897, "learning_rate": 7.008729487473351e-05, "loss": 0.0132, "step": 24000 }, { "grad_norm": 0.1355832815170288, "learning_rate": 7.006205570925263e-05, "loss": 0.0145, "step": 24010 }, { "grad_norm": 0.1464596539735794, "learning_rate": 7.003681044944892e-05, "loss": 0.0145, "step": 24020 }, { "grad_norm": 0.12969475984573364, "learning_rate": 7.001155910299126e-05, "loss": 0.0144, "step": 24030 }, { "grad_norm": 0.14273637533187866, "learning_rate": 6.99863016775503e-05, "loss": 0.014, "step": 24040 }, { "grad_norm": 0.1509455442428589, "learning_rate": 6.996103818079859e-05, "loss": 0.0143, "step": 24050 }, { "grad_norm": 0.13145430386066437, "learning_rate": 6.993576862041054e-05, "loss": 0.0148, "step": 24060 }, { "grad_norm": 0.1503261923789978, "learning_rate": 6.991049300406235e-05, "loss": 0.0137, "step": 24070 }, { "grad_norm": 0.15079106390476227, "learning_rate": 6.988521133943209e-05, "loss": 0.0154, "step": 24080 }, { "grad_norm": 0.12630610167980194, "learning_rate": 6.985992363419966e-05, "loss": 0.0137, "step": 24090 }, { "grad_norm": 0.17128385603427887, "learning_rate": 6.983462989604682e-05, "loss": 0.0164, "step": 24100 }, { "grad_norm": 0.12249492108821869, "learning_rate": 6.980933013265709e-05, "loss": 0.0163, "step": 24110 }, { "grad_norm": 0.12856662273406982, "learning_rate": 6.978402435171592e-05, "loss": 0.0143, "step": 24120 }, { "grad_norm": 0.14155857264995575, "learning_rate": 6.975871256091052e-05, "loss": 0.0146, "step": 24130 }, { "grad_norm": 0.11392487585544586, "learning_rate": 6.973339476792995e-05, "loss": 0.0145, "step": 24140 }, { "grad_norm": 0.13846147060394287, "learning_rate": 6.970807098046505e-05, "loss": 0.0132, "step": 24150 }, { "grad_norm": 0.12173926085233688, "learning_rate": 6.968274120620858e-05, "loss": 0.017, "step": 24160 }, { "grad_norm": 0.12840484082698822, "learning_rate": 6.965740545285499e-05, "loss": 0.0158, "step": 24170 }, { "grad_norm": 0.1254098117351532, "learning_rate": 6.963206372810068e-05, "loss": 0.0152, "step": 24180 }, { "grad_norm": 0.14779256284236908, "learning_rate": 6.960671603964375e-05, "loss": 0.0127, "step": 24190 }, { "grad_norm": 0.1209101527929306, "learning_rate": 6.958136239518418e-05, "loss": 0.0163, "step": 24200 }, { "grad_norm": 0.11053764075040817, "learning_rate": 6.955600280242371e-05, "loss": 0.0148, "step": 24210 }, { "grad_norm": 0.13978496193885803, "learning_rate": 6.953063726906596e-05, "loss": 0.0161, "step": 24220 }, { "grad_norm": 0.1100136861205101, "learning_rate": 6.950526580281626e-05, "loss": 0.0146, "step": 24230 }, { "grad_norm": 0.12123428285121918, "learning_rate": 6.947988841138184e-05, "loss": 0.0142, "step": 24240 }, { "grad_norm": 0.12584172189235687, "learning_rate": 6.945450510247165e-05, "loss": 0.0145, "step": 24250 }, { "grad_norm": 0.1176532506942749, "learning_rate": 6.942911588379647e-05, "loss": 0.0142, "step": 24260 }, { "grad_norm": 0.11486071348190308, "learning_rate": 6.940372076306888e-05, "loss": 0.0146, "step": 24270 }, { "grad_norm": 0.12862242758274078, "learning_rate": 6.937831974800326e-05, "loss": 0.0138, "step": 24280 }, { "grad_norm": 0.13675642013549805, "learning_rate": 6.935291284631574e-05, "loss": 0.0151, "step": 24290 }, { "grad_norm": 0.14967191219329834, "learning_rate": 6.932750006572428e-05, "loss": 0.016, "step": 24300 }, { "grad_norm": 0.15749849379062653, "learning_rate": 6.930208141394863e-05, "loss": 0.0165, "step": 24310 }, { "grad_norm": 0.19227208197116852, "learning_rate": 6.927665689871026e-05, "loss": 0.0144, "step": 24320 }, { "grad_norm": 0.15075188875198364, "learning_rate": 6.925122652773253e-05, "loss": 0.0137, "step": 24330 }, { "grad_norm": 0.1493096798658371, "learning_rate": 6.922579030874046e-05, "loss": 0.0161, "step": 24340 }, { "grad_norm": 0.1316913664340973, "learning_rate": 6.920034824946093e-05, "loss": 0.0158, "step": 24350 }, { "grad_norm": 0.11547382920980453, "learning_rate": 6.917490035762255e-05, "loss": 0.0152, "step": 24360 }, { "grad_norm": 0.1389610469341278, "learning_rate": 6.914944664095573e-05, "loss": 0.0151, "step": 24370 }, { "grad_norm": 0.1383991688489914, "learning_rate": 6.912398710719264e-05, "loss": 0.0144, "step": 24380 }, { "grad_norm": 0.13138329982757568, "learning_rate": 6.90985217640672e-05, "loss": 0.0142, "step": 24390 }, { "grad_norm": 0.10295641422271729, "learning_rate": 6.90730506193151e-05, "loss": 0.0158, "step": 24400 }, { "grad_norm": 0.15542179346084595, "learning_rate": 6.904757368067384e-05, "loss": 0.0159, "step": 24410 }, { "grad_norm": 0.11307438462972641, "learning_rate": 6.90220909558826e-05, "loss": 0.0141, "step": 24420 }, { "grad_norm": 0.15734776854515076, "learning_rate": 6.899660245268237e-05, "loss": 0.0153, "step": 24430 }, { "grad_norm": 0.15337461233139038, "learning_rate": 6.897110817881592e-05, "loss": 0.0165, "step": 24440 }, { "grad_norm": 0.12195641547441483, "learning_rate": 6.894560814202769e-05, "loss": 0.0177, "step": 24450 }, { "grad_norm": 0.1482124626636505, "learning_rate": 6.892010235006394e-05, "loss": 0.013, "step": 24460 }, { "grad_norm": 0.11869191378355026, "learning_rate": 6.889459081067264e-05, "loss": 0.0125, "step": 24470 }, { "grad_norm": 0.10186059027910233, "learning_rate": 6.886907353160356e-05, "loss": 0.0152, "step": 24480 }, { "grad_norm": 0.13749027252197266, "learning_rate": 6.884355052060814e-05, "loss": 0.0136, "step": 24490 }, { "grad_norm": 0.1493672877550125, "learning_rate": 6.88180217854396e-05, "loss": 0.0146, "step": 24500 }, { "grad_norm": 0.14634902775287628, "learning_rate": 6.87924873338529e-05, "loss": 0.0155, "step": 24510 }, { "grad_norm": 0.16400721669197083, "learning_rate": 6.876694717360475e-05, "loss": 0.014, "step": 24520 }, { "grad_norm": 0.1281619369983673, "learning_rate": 6.874140131245355e-05, "loss": 0.0137, "step": 24530 }, { "grad_norm": 0.11502408236265182, "learning_rate": 6.871584975815948e-05, "loss": 0.0135, "step": 24540 }, { "grad_norm": 0.10140327364206314, "learning_rate": 6.86902925184844e-05, "loss": 0.0152, "step": 24550 }, { "grad_norm": 0.1509280651807785, "learning_rate": 6.866472960119195e-05, "loss": 0.0141, "step": 24560 }, { "grad_norm": 0.14211788773536682, "learning_rate": 6.863916101404748e-05, "loss": 0.0142, "step": 24570 }, { "grad_norm": 0.10703028738498688, "learning_rate": 6.8613586764818e-05, "loss": 0.015, "step": 24580 }, { "grad_norm": 0.10851027816534042, "learning_rate": 6.858800686127233e-05, "loss": 0.0129, "step": 24590 }, { "grad_norm": 0.19138889014720917, "learning_rate": 6.856242131118097e-05, "loss": 0.014, "step": 24600 }, { "grad_norm": 0.15478762984275818, "learning_rate": 6.853683012231614e-05, "loss": 0.0149, "step": 24610 }, { "grad_norm": 0.15265996754169464, "learning_rate": 6.851123330245173e-05, "loss": 0.016, "step": 24620 }, { "grad_norm": 0.12883824110031128, "learning_rate": 6.848563085936343e-05, "loss": 0.0146, "step": 24630 }, { "grad_norm": 0.13937440514564514, "learning_rate": 6.846002280082853e-05, "loss": 0.0148, "step": 24640 }, { "grad_norm": 0.07851199060678482, "learning_rate": 6.843440913462614e-05, "loss": 0.0131, "step": 24650 }, { "grad_norm": 0.14278462529182434, "learning_rate": 6.840878986853698e-05, "loss": 0.0158, "step": 24660 }, { "grad_norm": 0.12444581091403961, "learning_rate": 6.838316501034352e-05, "loss": 0.0141, "step": 24670 }, { "grad_norm": 0.15552595257759094, "learning_rate": 6.83575345678299e-05, "loss": 0.013, "step": 24680 }, { "grad_norm": 0.14100898802280426, "learning_rate": 6.833189854878196e-05, "loss": 0.0149, "step": 24690 }, { "grad_norm": 0.14719392359256744, "learning_rate": 6.83062569609873e-05, "loss": 0.014, "step": 24700 }, { "grad_norm": 0.14998115599155426, "learning_rate": 6.828060981223512e-05, "loss": 0.016, "step": 24710 }, { "grad_norm": 0.14057794213294983, "learning_rate": 6.825495711031634e-05, "loss": 0.0136, "step": 24720 }, { "grad_norm": 0.13479599356651306, "learning_rate": 6.822929886302359e-05, "loss": 0.0136, "step": 24730 }, { "grad_norm": 0.1556205451488495, "learning_rate": 6.820363507815116e-05, "loss": 0.0157, "step": 24740 }, { "grad_norm": 0.0983293429017067, "learning_rate": 6.817796576349501e-05, "loss": 0.0145, "step": 24750 }, { "grad_norm": 0.1278478056192398, "learning_rate": 6.815229092685285e-05, "loss": 0.0153, "step": 24760 }, { "grad_norm": 0.12017686665058136, "learning_rate": 6.812661057602399e-05, "loss": 0.0151, "step": 24770 }, { "grad_norm": 0.12805457413196564, "learning_rate": 6.810092471880943e-05, "loss": 0.015, "step": 24780 }, { "grad_norm": 0.1621965616941452, "learning_rate": 6.807523336301187e-05, "loss": 0.015, "step": 24790 }, { "grad_norm": 0.16740508377552032, "learning_rate": 6.804953651643566e-05, "loss": 0.0148, "step": 24800 }, { "grad_norm": 0.14873848855495453, "learning_rate": 6.802383418688685e-05, "loss": 0.0148, "step": 24810 }, { "grad_norm": 0.17369695007801056, "learning_rate": 6.799812638217309e-05, "loss": 0.0169, "step": 24820 }, { "grad_norm": 0.14560076594352722, "learning_rate": 6.797241311010373e-05, "loss": 0.0155, "step": 24830 }, { "grad_norm": 0.1569187045097351, "learning_rate": 6.794669437848982e-05, "loss": 0.0149, "step": 24840 }, { "grad_norm": 0.1098608747124672, "learning_rate": 6.792097019514402e-05, "loss": 0.0143, "step": 24850 }, { "grad_norm": 0.1799505352973938, "learning_rate": 6.789524056788064e-05, "loss": 0.0165, "step": 24860 }, { "grad_norm": 0.12012193351984024, "learning_rate": 6.786950550451567e-05, "loss": 0.0151, "step": 24870 }, { "grad_norm": 0.13340996205806732, "learning_rate": 6.784376501286676e-05, "loss": 0.0155, "step": 24880 }, { "grad_norm": 0.1337195783853531, "learning_rate": 6.781801910075316e-05, "loss": 0.0143, "step": 24890 }, { "grad_norm": 0.12534157931804657, "learning_rate": 6.779226777599581e-05, "loss": 0.0135, "step": 24900 }, { "grad_norm": 0.11624719202518463, "learning_rate": 6.776651104641729e-05, "loss": 0.0158, "step": 24910 }, { "grad_norm": 0.13757279515266418, "learning_rate": 6.774074891984183e-05, "loss": 0.0158, "step": 24920 }, { "grad_norm": 0.14497989416122437, "learning_rate": 6.771498140409526e-05, "loss": 0.0147, "step": 24930 }, { "grad_norm": 0.17389312386512756, "learning_rate": 6.768920850700506e-05, "loss": 0.0165, "step": 24940 }, { "grad_norm": 0.1402270793914795, "learning_rate": 6.766343023640039e-05, "loss": 0.0124, "step": 24950 }, { "grad_norm": 0.11056479811668396, "learning_rate": 6.763764660011198e-05, "loss": 0.0159, "step": 24960 }, { "grad_norm": 0.158872589468956, "learning_rate": 6.761185760597223e-05, "loss": 0.0161, "step": 24970 }, { "grad_norm": 0.10051701217889786, "learning_rate": 6.758606326181515e-05, "loss": 0.0138, "step": 24980 }, { "grad_norm": 0.18824106454849243, "learning_rate": 6.75602635754764e-05, "loss": 0.0142, "step": 24990 }, { "grad_norm": 0.11183835566043854, "learning_rate": 6.75344585547932e-05, "loss": 0.0175, "step": 25000 }, { "grad_norm": 0.12356320023536682, "learning_rate": 6.750864820760449e-05, "loss": 0.0134, "step": 25010 }, { "grad_norm": 0.14180156588554382, "learning_rate": 6.748283254175072e-05, "loss": 0.015, "step": 25020 }, { "grad_norm": 0.11907284706830978, "learning_rate": 6.745701156507404e-05, "loss": 0.0146, "step": 25030 }, { "grad_norm": 0.12666861712932587, "learning_rate": 6.743118528541818e-05, "loss": 0.0142, "step": 25040 }, { "grad_norm": 0.14550036191940308, "learning_rate": 6.740535371062846e-05, "loss": 0.0132, "step": 25050 }, { "grad_norm": 0.12904518842697144, "learning_rate": 6.737951684855185e-05, "loss": 0.0151, "step": 25060 }, { "grad_norm": 0.15417882800102234, "learning_rate": 6.735367470703691e-05, "loss": 0.0134, "step": 25070 }, { "grad_norm": 0.16486957669258118, "learning_rate": 6.732782729393379e-05, "loss": 0.0162, "step": 25080 }, { "grad_norm": 0.13741804659366608, "learning_rate": 6.730197461709425e-05, "loss": 0.0143, "step": 25090 }, { "grad_norm": 0.13793663680553436, "learning_rate": 6.727611668437164e-05, "loss": 0.0149, "step": 25100 }, { "grad_norm": 0.15173321962356567, "learning_rate": 6.725025350362094e-05, "loss": 0.0149, "step": 25110 }, { "grad_norm": 0.14626544713974, "learning_rate": 6.72243850826987e-05, "loss": 0.0147, "step": 25120 }, { "grad_norm": 0.12927299737930298, "learning_rate": 6.719851142946305e-05, "loss": 0.0154, "step": 25130 }, { "grad_norm": 0.18863347172737122, "learning_rate": 6.717263255177372e-05, "loss": 0.0155, "step": 25140 }, { "grad_norm": 0.1269787847995758, "learning_rate": 6.714674845749205e-05, "loss": 0.0141, "step": 25150 }, { "grad_norm": 0.15944191813468933, "learning_rate": 6.712085915448092e-05, "loss": 0.0154, "step": 25160 }, { "grad_norm": 0.13353508710861206, "learning_rate": 6.709496465060486e-05, "loss": 0.0138, "step": 25170 }, { "grad_norm": 0.15381145477294922, "learning_rate": 6.706906495372987e-05, "loss": 0.0143, "step": 25180 }, { "grad_norm": 0.12346772849559784, "learning_rate": 6.704316007172365e-05, "loss": 0.0141, "step": 25190 }, { "grad_norm": 0.12663406133651733, "learning_rate": 6.701725001245539e-05, "loss": 0.0131, "step": 25200 }, { "grad_norm": 0.12053855508565903, "learning_rate": 6.699133478379588e-05, "loss": 0.0149, "step": 25210 }, { "grad_norm": 0.16417041420936584, "learning_rate": 6.69654143936175e-05, "loss": 0.0154, "step": 25220 }, { "grad_norm": 0.09002434462308884, "learning_rate": 6.693948884979419e-05, "loss": 0.0137, "step": 25230 }, { "grad_norm": 0.1356407105922699, "learning_rate": 6.691355816020142e-05, "loss": 0.0148, "step": 25240 }, { "grad_norm": 0.1400856077671051, "learning_rate": 6.688762233271624e-05, "loss": 0.0152, "step": 25250 }, { "grad_norm": 0.12258244305849075, "learning_rate": 6.68616813752173e-05, "loss": 0.0135, "step": 25260 }, { "grad_norm": 0.1288895457983017, "learning_rate": 6.683573529558477e-05, "loss": 0.0164, "step": 25270 }, { "grad_norm": 0.1280437856912613, "learning_rate": 6.680978410170037e-05, "loss": 0.0131, "step": 25280 }, { "grad_norm": 0.12703993916511536, "learning_rate": 6.678382780144741e-05, "loss": 0.0145, "step": 25290 }, { "grad_norm": 0.11514900624752045, "learning_rate": 6.675786640271071e-05, "loss": 0.0135, "step": 25300 }, { "grad_norm": 0.13055408000946045, "learning_rate": 6.673189991337665e-05, "loss": 0.0134, "step": 25310 }, { "grad_norm": 0.1083674356341362, "learning_rate": 6.670592834133317e-05, "loss": 0.0141, "step": 25320 }, { "grad_norm": 0.1344500333070755, "learning_rate": 6.667995169446979e-05, "loss": 0.0146, "step": 25330 }, { "grad_norm": 0.17210665345191956, "learning_rate": 6.665396998067747e-05, "loss": 0.0138, "step": 25340 }, { "grad_norm": 0.16522987186908722, "learning_rate": 6.66279832078488e-05, "loss": 0.0148, "step": 25350 }, { "grad_norm": 0.14839310944080353, "learning_rate": 6.660199138387786e-05, "loss": 0.013, "step": 25360 }, { "grad_norm": 0.15190757811069489, "learning_rate": 6.65759945166603e-05, "loss": 0.0145, "step": 25370 }, { "grad_norm": 0.13364103436470032, "learning_rate": 6.654999261409326e-05, "loss": 0.0138, "step": 25380 }, { "grad_norm": 0.1275002360343933, "learning_rate": 6.652398568407544e-05, "loss": 0.0142, "step": 25390 }, { "grad_norm": 0.10817262530326843, "learning_rate": 6.649797373450707e-05, "loss": 0.0137, "step": 25400 }, { "grad_norm": 0.09964597970247269, "learning_rate": 6.647195677328988e-05, "loss": 0.0139, "step": 25410 }, { "grad_norm": 0.12753471732139587, "learning_rate": 6.644593480832712e-05, "loss": 0.016, "step": 25420 }, { "grad_norm": 0.10733731091022491, "learning_rate": 6.641990784752363e-05, "loss": 0.0151, "step": 25430 }, { "grad_norm": 0.10740269720554352, "learning_rate": 6.639387589878566e-05, "loss": 0.0137, "step": 25440 }, { "grad_norm": 0.13507792353630066, "learning_rate": 6.636783897002103e-05, "loss": 0.0154, "step": 25450 }, { "grad_norm": 0.11790130287408829, "learning_rate": 6.63417970691391e-05, "loss": 0.014, "step": 25460 }, { "grad_norm": 0.1523648351430893, "learning_rate": 6.63157502040507e-05, "loss": 0.0148, "step": 25470 }, { "grad_norm": 0.13547024130821228, "learning_rate": 6.628969838266819e-05, "loss": 0.0154, "step": 25480 }, { "grad_norm": 0.13724058866500854, "learning_rate": 6.626364161290541e-05, "loss": 0.0148, "step": 25490 }, { "grad_norm": 0.14916926622390747, "learning_rate": 6.623757990267774e-05, "loss": 0.0158, "step": 25500 }, { "grad_norm": 0.1683669537305832, "learning_rate": 6.621151325990201e-05, "loss": 0.0139, "step": 25510 }, { "grad_norm": 0.1181870773434639, "learning_rate": 6.618544169249657e-05, "loss": 0.0154, "step": 25520 }, { "grad_norm": 0.09949585795402527, "learning_rate": 6.615936520838133e-05, "loss": 0.0165, "step": 25530 }, { "grad_norm": 0.12397053837776184, "learning_rate": 6.613328381547759e-05, "loss": 0.0146, "step": 25540 }, { "grad_norm": 0.120151586830616, "learning_rate": 6.610719752170821e-05, "loss": 0.0143, "step": 25550 }, { "grad_norm": 0.14967674016952515, "learning_rate": 6.60811063349975e-05, "loss": 0.014, "step": 25560 }, { "grad_norm": 0.1232839897274971, "learning_rate": 6.605501026327127e-05, "loss": 0.0155, "step": 25570 }, { "grad_norm": 0.17688031494617462, "learning_rate": 6.602890931445685e-05, "loss": 0.0131, "step": 25580 }, { "grad_norm": 0.15731385350227356, "learning_rate": 6.6002803496483e-05, "loss": 0.0162, "step": 25590 }, { "grad_norm": 0.10480009764432907, "learning_rate": 6.597669281727997e-05, "loss": 0.0152, "step": 25600 }, { "grad_norm": 0.1501259207725525, "learning_rate": 6.595057728477949e-05, "loss": 0.0142, "step": 25610 }, { "grad_norm": 0.11314816772937775, "learning_rate": 6.59244569069148e-05, "loss": 0.0143, "step": 25620 }, { "grad_norm": 0.14642198383808136, "learning_rate": 6.589833169162054e-05, "loss": 0.0147, "step": 25630 }, { "grad_norm": 0.1256912648677826, "learning_rate": 6.587220164683291e-05, "loss": 0.0136, "step": 25640 }, { "grad_norm": 0.14237262308597565, "learning_rate": 6.58460667804895e-05, "loss": 0.013, "step": 25650 }, { "grad_norm": 0.11905910819768906, "learning_rate": 6.581992710052938e-05, "loss": 0.0131, "step": 25660 }, { "grad_norm": 0.12049750983715057, "learning_rate": 6.579378261489311e-05, "loss": 0.0144, "step": 25670 }, { "grad_norm": 0.15063337981700897, "learning_rate": 6.576763333152268e-05, "loss": 0.0146, "step": 25680 }, { "grad_norm": 0.12624162435531616, "learning_rate": 6.574147925836159e-05, "loss": 0.0148, "step": 25690 }, { "grad_norm": 0.13423289358615875, "learning_rate": 6.571532040335472e-05, "loss": 0.0144, "step": 25700 }, { "grad_norm": 0.1536361277103424, "learning_rate": 6.568915677444845e-05, "loss": 0.0149, "step": 25710 }, { "grad_norm": 0.11716196686029434, "learning_rate": 6.56629883795906e-05, "loss": 0.0148, "step": 25720 }, { "grad_norm": 0.128257617354393, "learning_rate": 6.563681522673043e-05, "loss": 0.0134, "step": 25730 }, { "grad_norm": 0.126731738448143, "learning_rate": 6.561063732381867e-05, "loss": 0.0156, "step": 25740 }, { "grad_norm": 0.12094669044017792, "learning_rate": 6.558445467880745e-05, "loss": 0.0138, "step": 25750 }, { "grad_norm": 0.12856487929821014, "learning_rate": 6.55582672996504e-05, "loss": 0.0146, "step": 25760 }, { "grad_norm": 0.14068599045276642, "learning_rate": 6.553207519430253e-05, "loss": 0.0141, "step": 25770 }, { "grad_norm": 0.12087903916835785, "learning_rate": 6.550587837072032e-05, "loss": 0.015, "step": 25780 }, { "grad_norm": 0.1596233993768692, "learning_rate": 6.547967683686166e-05, "loss": 0.017, "step": 25790 }, { "grad_norm": 0.17903533577919006, "learning_rate": 6.545347060068591e-05, "loss": 0.0147, "step": 25800 }, { "grad_norm": 0.15781138837337494, "learning_rate": 6.542725967015382e-05, "loss": 0.0147, "step": 25810 }, { "grad_norm": 0.1425895392894745, "learning_rate": 6.540104405322757e-05, "loss": 0.0145, "step": 25820 }, { "grad_norm": 0.09551291912794113, "learning_rate": 6.537482375787077e-05, "loss": 0.0132, "step": 25830 }, { "grad_norm": 0.11613930016756058, "learning_rate": 6.534859879204845e-05, "loss": 0.015, "step": 25840 }, { "grad_norm": 0.12906460464000702, "learning_rate": 6.532236916372709e-05, "loss": 0.0145, "step": 25850 }, { "grad_norm": 0.11622576415538788, "learning_rate": 6.529613488087454e-05, "loss": 0.0136, "step": 25860 }, { "grad_norm": 0.11194437742233276, "learning_rate": 6.526989595146009e-05, "loss": 0.0133, "step": 25870 }, { "grad_norm": 0.13365888595581055, "learning_rate": 6.524365238345441e-05, "loss": 0.0145, "step": 25880 }, { "grad_norm": 0.13666586577892303, "learning_rate": 6.521740418482964e-05, "loss": 0.014, "step": 25890 }, { "grad_norm": 0.1660020649433136, "learning_rate": 6.519115136355925e-05, "loss": 0.0151, "step": 25900 }, { "grad_norm": 0.12786507606506348, "learning_rate": 6.51648939276182e-05, "loss": 0.0137, "step": 25910 }, { "grad_norm": 0.1915593445301056, "learning_rate": 6.513863188498277e-05, "loss": 0.0137, "step": 25920 }, { "grad_norm": 0.12985239923000336, "learning_rate": 6.511236524363068e-05, "loss": 0.015, "step": 25930 }, { "grad_norm": 0.10974463820457458, "learning_rate": 6.508609401154104e-05, "loss": 0.015, "step": 25940 }, { "grad_norm": 0.10836710035800934, "learning_rate": 6.505981819669439e-05, "loss": 0.0135, "step": 25950 }, { "grad_norm": 0.1286924034357071, "learning_rate": 6.503353780707258e-05, "loss": 0.0141, "step": 25960 }, { "grad_norm": 0.10530363768339157, "learning_rate": 6.500725285065895e-05, "loss": 0.0134, "step": 25970 }, { "grad_norm": 0.13220733404159546, "learning_rate": 6.498096333543813e-05, "loss": 0.016, "step": 25980 }, { "grad_norm": 0.11522872745990753, "learning_rate": 6.49546692693962e-05, "loss": 0.0154, "step": 25990 }, { "grad_norm": 0.14137409627437592, "learning_rate": 6.492837066052059e-05, "loss": 0.0142, "step": 26000 }, { "grad_norm": 0.15287239849567413, "learning_rate": 6.490206751680014e-05, "loss": 0.0149, "step": 26010 }, { "grad_norm": 0.11994336545467377, "learning_rate": 6.487575984622505e-05, "loss": 0.0134, "step": 26020 }, { "grad_norm": 0.12334554642438889, "learning_rate": 6.484944765678689e-05, "loss": 0.0131, "step": 26030 }, { "grad_norm": 0.12619556486606598, "learning_rate": 6.482313095647861e-05, "loss": 0.0158, "step": 26040 }, { "grad_norm": 0.12835629284381866, "learning_rate": 6.479680975329451e-05, "loss": 0.0135, "step": 26050 }, { "grad_norm": 0.14944173395633698, "learning_rate": 6.477048405523031e-05, "loss": 0.0155, "step": 26060 }, { "grad_norm": 0.131228506565094, "learning_rate": 6.474415387028304e-05, "loss": 0.0148, "step": 26070 }, { "grad_norm": 0.13824519515037537, "learning_rate": 6.471781920645114e-05, "loss": 0.0148, "step": 26080 }, { "grad_norm": 0.12714692950248718, "learning_rate": 6.469148007173434e-05, "loss": 0.0134, "step": 26090 }, { "grad_norm": 0.11998619139194489, "learning_rate": 6.466513647413381e-05, "loss": 0.0129, "step": 26100 }, { "grad_norm": 0.0902145728468895, "learning_rate": 6.463878842165203e-05, "loss": 0.0143, "step": 26110 }, { "grad_norm": 0.10985346138477325, "learning_rate": 6.461243592229286e-05, "loss": 0.0145, "step": 26120 }, { "grad_norm": 0.1089840903878212, "learning_rate": 6.458607898406146e-05, "loss": 0.013, "step": 26130 }, { "grad_norm": 0.147661492228508, "learning_rate": 6.455971761496439e-05, "loss": 0.0134, "step": 26140 }, { "grad_norm": 0.14625956118106842, "learning_rate": 6.453335182300953e-05, "loss": 0.014, "step": 26150 }, { "grad_norm": 0.11408927291631699, "learning_rate": 6.450698161620612e-05, "loss": 0.0141, "step": 26160 }, { "grad_norm": 0.17757412791252136, "learning_rate": 6.448060700256473e-05, "loss": 0.0154, "step": 26170 }, { "grad_norm": 0.1036655381321907, "learning_rate": 6.445422799009726e-05, "loss": 0.0132, "step": 26180 }, { "grad_norm": 0.12874548137187958, "learning_rate": 6.442784458681699e-05, "loss": 0.0133, "step": 26190 }, { "grad_norm": 0.17817460000514984, "learning_rate": 6.440145680073847e-05, "loss": 0.0149, "step": 26200 }, { "grad_norm": 0.13412071764469147, "learning_rate": 6.437506463987762e-05, "loss": 0.0137, "step": 26210 }, { "grad_norm": 0.13790759444236755, "learning_rate": 6.434866811225168e-05, "loss": 0.014, "step": 26220 }, { "grad_norm": 0.11255796998739243, "learning_rate": 6.432226722587923e-05, "loss": 0.0148, "step": 26230 }, { "grad_norm": 0.12701386213302612, "learning_rate": 6.429586198878015e-05, "loss": 0.0143, "step": 26240 }, { "grad_norm": 0.13082651793956757, "learning_rate": 6.426945240897566e-05, "loss": 0.0129, "step": 26250 }, { "grad_norm": 0.12915834784507751, "learning_rate": 6.424303849448829e-05, "loss": 0.0173, "step": 26260 }, { "grad_norm": 0.1578271985054016, "learning_rate": 6.42166202533419e-05, "loss": 0.0138, "step": 26270 }, { "grad_norm": 0.1450699418783188, "learning_rate": 6.419019769356164e-05, "loss": 0.0143, "step": 26280 }, { "grad_norm": 0.1396867334842682, "learning_rate": 6.416377082317398e-05, "loss": 0.013, "step": 26290 }, { "grad_norm": 0.13847072422504425, "learning_rate": 6.413733965020674e-05, "loss": 0.0141, "step": 26300 }, { "grad_norm": 0.1325882524251938, "learning_rate": 6.411090418268896e-05, "loss": 0.0165, "step": 26310 }, { "grad_norm": 0.157831072807312, "learning_rate": 6.408446442865109e-05, "loss": 0.0148, "step": 26320 }, { "grad_norm": 0.10266545414924622, "learning_rate": 6.405802039612479e-05, "loss": 0.015, "step": 26330 }, { "grad_norm": 0.1828683465719223, "learning_rate": 6.403157209314308e-05, "loss": 0.0173, "step": 26340 }, { "grad_norm": 0.1354888528585434, "learning_rate": 6.400511952774024e-05, "loss": 0.0138, "step": 26350 }, { "grad_norm": 0.12455739080905914, "learning_rate": 6.397866270795187e-05, "loss": 0.0126, "step": 26360 }, { "grad_norm": 0.15703319013118744, "learning_rate": 6.395220164181489e-05, "loss": 0.0143, "step": 26370 }, { "grad_norm": 0.11198889464139938, "learning_rate": 6.39257363373674e-05, "loss": 0.0136, "step": 26380 }, { "grad_norm": 0.12600895762443542, "learning_rate": 6.389926680264892e-05, "loss": 0.0156, "step": 26390 }, { "grad_norm": 0.1399548053741455, "learning_rate": 6.387279304570017e-05, "loss": 0.0148, "step": 26400 }, { "grad_norm": 0.1533374935388565, "learning_rate": 6.384631507456319e-05, "loss": 0.0141, "step": 26410 }, { "grad_norm": 0.15909691154956818, "learning_rate": 6.381983289728126e-05, "loss": 0.0137, "step": 26420 }, { "grad_norm": 0.1051863506436348, "learning_rate": 6.3793346521899e-05, "loss": 0.0132, "step": 26430 }, { "grad_norm": 0.11947225779294968, "learning_rate": 6.376685595646226e-05, "loss": 0.0145, "step": 26440 }, { "grad_norm": 0.09353002160787582, "learning_rate": 6.374036120901816e-05, "loss": 0.0125, "step": 26450 }, { "grad_norm": 0.09746420383453369, "learning_rate": 6.371386228761514e-05, "loss": 0.0125, "step": 26460 }, { "grad_norm": 0.11990372836589813, "learning_rate": 6.368735920030283e-05, "loss": 0.0142, "step": 26470 }, { "grad_norm": 0.13107693195343018, "learning_rate": 6.366085195513218e-05, "loss": 0.0144, "step": 26480 }, { "grad_norm": 0.12200316786766052, "learning_rate": 6.363434056015543e-05, "loss": 0.0137, "step": 26490 }, { "grad_norm": 0.1222308874130249, "learning_rate": 6.360782502342599e-05, "loss": 0.0145, "step": 26500 }, { "grad_norm": 0.15410880744457245, "learning_rate": 6.358130535299862e-05, "loss": 0.0135, "step": 26510 }, { "grad_norm": 0.13357797265052795, "learning_rate": 6.355478155692926e-05, "loss": 0.0131, "step": 26520 }, { "grad_norm": 0.14116141200065613, "learning_rate": 6.352825364327517e-05, "loss": 0.0159, "step": 26530 }, { "grad_norm": 0.14104260504245758, "learning_rate": 6.350172162009482e-05, "loss": 0.0137, "step": 26540 }, { "grad_norm": 0.1602293699979782, "learning_rate": 6.347518549544793e-05, "loss": 0.0143, "step": 26550 }, { "grad_norm": 0.20674452185630798, "learning_rate": 6.344864527739547e-05, "loss": 0.0141, "step": 26560 }, { "grad_norm": 0.15860231220722198, "learning_rate": 6.342210097399966e-05, "loss": 0.0161, "step": 26570 }, { "grad_norm": 0.12952157855033875, "learning_rate": 6.339555259332398e-05, "loss": 0.0159, "step": 26580 }, { "grad_norm": 0.10684487223625183, "learning_rate": 6.33690001434331e-05, "loss": 0.0128, "step": 26590 }, { "grad_norm": 0.15154141187667847, "learning_rate": 6.334244363239296e-05, "loss": 0.0161, "step": 26600 }, { "grad_norm": 0.14748041331768036, "learning_rate": 6.331588306827073e-05, "loss": 0.0141, "step": 26610 }, { "grad_norm": 0.15194977819919586, "learning_rate": 6.328931845913483e-05, "loss": 0.0136, "step": 26620 }, { "grad_norm": 0.1413513422012329, "learning_rate": 6.326274981305484e-05, "loss": 0.014, "step": 26630 }, { "grad_norm": 0.12148913741111755, "learning_rate": 6.323617713810166e-05, "loss": 0.014, "step": 26640 }, { "grad_norm": 0.12809360027313232, "learning_rate": 6.320960044234734e-05, "loss": 0.0146, "step": 26650 }, { "grad_norm": 0.15847551822662354, "learning_rate": 6.318301973386518e-05, "loss": 0.0142, "step": 26660 }, { "grad_norm": 0.12364203482866287, "learning_rate": 6.315643502072971e-05, "loss": 0.0133, "step": 26670 }, { "grad_norm": 0.1030455231666565, "learning_rate": 6.312984631101667e-05, "loss": 0.0144, "step": 26680 }, { "grad_norm": 0.13635441660881042, "learning_rate": 6.310325361280297e-05, "loss": 0.0138, "step": 26690 }, { "grad_norm": 0.1433105319738388, "learning_rate": 6.30766569341668e-05, "loss": 0.0141, "step": 26700 }, { "grad_norm": 0.14611361920833588, "learning_rate": 6.305005628318753e-05, "loss": 0.0154, "step": 26710 }, { "grad_norm": 0.15780439972877502, "learning_rate": 6.302345166794572e-05, "loss": 0.013, "step": 26720 }, { "grad_norm": 0.14463217556476593, "learning_rate": 6.299684309652316e-05, "loss": 0.0168, "step": 26730 }, { "grad_norm": 0.12483277171850204, "learning_rate": 6.297023057700283e-05, "loss": 0.0142, "step": 26740 }, { "grad_norm": 0.13504746556282043, "learning_rate": 6.294361411746891e-05, "loss": 0.0137, "step": 26750 }, { "grad_norm": 0.12059379369020462, "learning_rate": 6.291699372600677e-05, "loss": 0.0144, "step": 26760 }, { "grad_norm": 0.12112843245267868, "learning_rate": 6.2890369410703e-05, "loss": 0.0158, "step": 26770 }, { "grad_norm": 0.1305561363697052, "learning_rate": 6.286374117964534e-05, "loss": 0.0138, "step": 26780 }, { "grad_norm": 0.17924416065216064, "learning_rate": 6.283710904092277e-05, "loss": 0.0169, "step": 26790 }, { "grad_norm": 0.1467931717634201, "learning_rate": 6.281047300262542e-05, "loss": 0.0146, "step": 26800 }, { "grad_norm": 0.16566410660743713, "learning_rate": 6.278383307284461e-05, "loss": 0.0134, "step": 26810 }, { "grad_norm": 0.1823885440826416, "learning_rate": 6.275718925967284e-05, "loss": 0.0135, "step": 26820 }, { "grad_norm": 0.13110755383968353, "learning_rate": 6.273054157120382e-05, "loss": 0.0138, "step": 26830 }, { "grad_norm": 0.11227846145629883, "learning_rate": 6.270389001553238e-05, "loss": 0.0122, "step": 26840 }, { "grad_norm": 0.13735710084438324, "learning_rate": 6.26772346007546e-05, "loss": 0.0133, "step": 26850 }, { "grad_norm": 0.15341021120548248, "learning_rate": 6.265057533496767e-05, "loss": 0.0138, "step": 26860 }, { "grad_norm": 0.12804710865020752, "learning_rate": 6.262391222626997e-05, "loss": 0.0143, "step": 26870 }, { "grad_norm": 0.12793833017349243, "learning_rate": 6.259724528276106e-05, "loss": 0.0129, "step": 26880 }, { "grad_norm": 0.15472182631492615, "learning_rate": 6.257057451254162e-05, "loss": 0.0139, "step": 26890 }, { "grad_norm": 0.12173131853342056, "learning_rate": 6.254389992371357e-05, "loss": 0.0121, "step": 26900 }, { "grad_norm": 0.1466502845287323, "learning_rate": 6.25172215243799e-05, "loss": 0.0154, "step": 26910 }, { "grad_norm": 0.12101048976182938, "learning_rate": 6.249053932264486e-05, "loss": 0.013, "step": 26920 }, { "grad_norm": 0.15198369324207306, "learning_rate": 6.246385332661376e-05, "loss": 0.0134, "step": 26930 }, { "grad_norm": 0.14029085636138916, "learning_rate": 6.24371635443931e-05, "loss": 0.0129, "step": 26940 }, { "grad_norm": 0.12447595596313477, "learning_rate": 6.241046998409054e-05, "loss": 0.0138, "step": 26950 }, { "grad_norm": 0.09268131852149963, "learning_rate": 6.238377265381489e-05, "loss": 0.0135, "step": 26960 }, { "grad_norm": 0.1278015822172165, "learning_rate": 6.235707156167607e-05, "loss": 0.0132, "step": 26970 }, { "grad_norm": 0.11253609508275986, "learning_rate": 6.233036671578519e-05, "loss": 0.0145, "step": 26980 }, { "grad_norm": 0.15327882766723633, "learning_rate": 6.230365812425445e-05, "loss": 0.0138, "step": 26990 }, { "grad_norm": 0.10582620650529861, "learning_rate": 6.227694579519724e-05, "loss": 0.0157, "step": 27000 }, { "grad_norm": 0.11349297314882278, "learning_rate": 6.225022973672805e-05, "loss": 0.0124, "step": 27010 }, { "grad_norm": 0.1226930096745491, "learning_rate": 6.222350995696253e-05, "loss": 0.014, "step": 27020 }, { "grad_norm": 0.11484721302986145, "learning_rate": 6.21967864640174e-05, "loss": 0.0165, "step": 27030 }, { "grad_norm": 0.12924185395240784, "learning_rate": 6.217005926601059e-05, "loss": 0.0146, "step": 27040 }, { "grad_norm": 0.1204303503036499, "learning_rate": 6.214332837106111e-05, "loss": 0.0139, "step": 27050 }, { "grad_norm": 0.1378452479839325, "learning_rate": 6.21165937872891e-05, "loss": 0.0148, "step": 27060 }, { "grad_norm": 0.111416757106781, "learning_rate": 6.208985552281582e-05, "loss": 0.0147, "step": 27070 }, { "grad_norm": 0.15055502951145172, "learning_rate": 6.206311358576364e-05, "loss": 0.0143, "step": 27080 }, { "grad_norm": 0.12344326078891754, "learning_rate": 6.203636798425608e-05, "loss": 0.0143, "step": 27090 }, { "grad_norm": 0.15594226121902466, "learning_rate": 6.20096187264177e-05, "loss": 0.014, "step": 27100 }, { "grad_norm": 0.15713737905025482, "learning_rate": 6.198286582037425e-05, "loss": 0.0132, "step": 27110 }, { "grad_norm": 0.13416826725006104, "learning_rate": 6.195610927425256e-05, "loss": 0.0127, "step": 27120 }, { "grad_norm": 0.12832528352737427, "learning_rate": 6.192934909618056e-05, "loss": 0.0147, "step": 27130 }, { "grad_norm": 0.1044548898935318, "learning_rate": 6.190258529428728e-05, "loss": 0.0128, "step": 27140 }, { "grad_norm": 0.10437142848968506, "learning_rate": 6.187581787670285e-05, "loss": 0.012, "step": 27150 }, { "grad_norm": 0.11186501383781433, "learning_rate": 6.184904685155852e-05, "loss": 0.0115, "step": 27160 }, { "grad_norm": 0.13109669089317322, "learning_rate": 6.18222722269866e-05, "loss": 0.0133, "step": 27170 }, { "grad_norm": 0.13405057787895203, "learning_rate": 6.179549401112053e-05, "loss": 0.015, "step": 27180 }, { "grad_norm": 0.141972616314888, "learning_rate": 6.176871221209482e-05, "loss": 0.0136, "step": 27190 }, { "grad_norm": 0.11617706716060638, "learning_rate": 6.174192683804508e-05, "loss": 0.014, "step": 27200 }, { "grad_norm": 0.13845041394233704, "learning_rate": 6.1715137897108e-05, "loss": 0.0136, "step": 27210 }, { "grad_norm": 0.11973416060209274, "learning_rate": 6.168834539742134e-05, "loss": 0.0127, "step": 27220 }, { "grad_norm": 0.11850995570421219, "learning_rate": 6.166154934712397e-05, "loss": 0.0116, "step": 27230 }, { "grad_norm": 0.1307802051305771, "learning_rate": 6.163474975435581e-05, "loss": 0.0126, "step": 27240 }, { "grad_norm": 0.1611192673444748, "learning_rate": 6.160794662725787e-05, "loss": 0.0145, "step": 27250 }, { "grad_norm": 0.13585561513900757, "learning_rate": 6.158113997397222e-05, "loss": 0.0135, "step": 27260 }, { "grad_norm": 0.11925270408391953, "learning_rate": 6.155432980264205e-05, "loss": 0.0138, "step": 27270 }, { "grad_norm": 0.12130072712898254, "learning_rate": 6.152751612141156e-05, "loss": 0.014, "step": 27280 }, { "grad_norm": 0.09719984978437424, "learning_rate": 6.150069893842602e-05, "loss": 0.0132, "step": 27290 }, { "grad_norm": 0.1376597136259079, "learning_rate": 6.147387826183182e-05, "loss": 0.0139, "step": 27300 }, { "grad_norm": 0.10015659779310226, "learning_rate": 6.144705409977635e-05, "loss": 0.0128, "step": 27310 }, { "grad_norm": 0.1418314427137375, "learning_rate": 6.142022646040808e-05, "loss": 0.0133, "step": 27320 }, { "grad_norm": 0.11100880801677704, "learning_rate": 6.139339535187653e-05, "loss": 0.0138, "step": 27330 }, { "grad_norm": 0.11365501582622528, "learning_rate": 6.136656078233232e-05, "loss": 0.0137, "step": 27340 }, { "grad_norm": 0.10195240378379822, "learning_rate": 6.133972275992707e-05, "loss": 0.012, "step": 27350 }, { "grad_norm": 0.14036957919597626, "learning_rate": 6.131288129281342e-05, "loss": 0.0124, "step": 27360 }, { "grad_norm": 0.09392168372869492, "learning_rate": 6.128603638914516e-05, "loss": 0.0125, "step": 27370 }, { "grad_norm": 0.12420224398374557, "learning_rate": 6.125918805707704e-05, "loss": 0.0134, "step": 27380 }, { "grad_norm": 0.11364063620567322, "learning_rate": 6.123233630476485e-05, "loss": 0.0118, "step": 27390 }, { "grad_norm": 0.12576347589492798, "learning_rate": 6.120548114036547e-05, "loss": 0.0136, "step": 27400 }, { "grad_norm": 0.1342453807592392, "learning_rate": 6.117862257203679e-05, "loss": 0.0133, "step": 27410 }, { "grad_norm": 0.11642465740442276, "learning_rate": 6.115176060793771e-05, "loss": 0.0132, "step": 27420 }, { "grad_norm": 0.1894666850566864, "learning_rate": 6.112489525622822e-05, "loss": 0.0137, "step": 27430 }, { "grad_norm": 0.1429445892572403, "learning_rate": 6.109802652506928e-05, "loss": 0.0149, "step": 27440 }, { "grad_norm": 0.12769481539726257, "learning_rate": 6.107115442262291e-05, "loss": 0.015, "step": 27450 }, { "grad_norm": 0.1700403392314911, "learning_rate": 6.104427895705214e-05, "loss": 0.0142, "step": 27460 }, { "grad_norm": 0.13927753269672394, "learning_rate": 6.101740013652103e-05, "loss": 0.0132, "step": 27470 }, { "grad_norm": 0.12165240198373795, "learning_rate": 6.099051796919465e-05, "loss": 0.0122, "step": 27480 }, { "grad_norm": 0.1589633673429489, "learning_rate": 6.096363246323911e-05, "loss": 0.0138, "step": 27490 }, { "grad_norm": 0.14048641920089722, "learning_rate": 6.0936743626821504e-05, "loss": 0.0131, "step": 27500 }, { "grad_norm": 0.1370524913072586, "learning_rate": 6.090985146810996e-05, "loss": 0.0139, "step": 27510 }, { "grad_norm": 0.12870363891124725, "learning_rate": 6.088295599527357e-05, "loss": 0.015, "step": 27520 }, { "grad_norm": 0.15376198291778564, "learning_rate": 6.085605721648252e-05, "loss": 0.0125, "step": 27530 }, { "grad_norm": 0.1262531280517578, "learning_rate": 6.082915513990792e-05, "loss": 0.0136, "step": 27540 }, { "grad_norm": 0.10585503280162811, "learning_rate": 6.080224977372192e-05, "loss": 0.0126, "step": 27550 }, { "grad_norm": 0.14086584746837616, "learning_rate": 6.0775341126097666e-05, "loss": 0.0143, "step": 27560 }, { "grad_norm": 0.1463734209537506, "learning_rate": 6.074842920520926e-05, "loss": 0.0137, "step": 27570 }, { "grad_norm": 0.11114492267370224, "learning_rate": 6.072151401923186e-05, "loss": 0.0119, "step": 27580 }, { "grad_norm": 0.09039999544620514, "learning_rate": 6.069459557634159e-05, "loss": 0.0132, "step": 27590 }, { "grad_norm": 0.12636172771453857, "learning_rate": 6.066767388471557e-05, "loss": 0.0132, "step": 27600 }, { "grad_norm": 0.09579668194055557, "learning_rate": 6.064074895253188e-05, "loss": 0.0123, "step": 27610 }, { "grad_norm": 0.1385866105556488, "learning_rate": 6.061382078796961e-05, "loss": 0.0133, "step": 27620 }, { "grad_norm": 0.10194917768239975, "learning_rate": 6.0586889399208814e-05, "loss": 0.0127, "step": 27630 }, { "grad_norm": 0.1369892805814743, "learning_rate": 6.0559954794430565e-05, "loss": 0.0137, "step": 27640 }, { "grad_norm": 0.12790325284004211, "learning_rate": 6.053301698181687e-05, "loss": 0.0125, "step": 27650 }, { "grad_norm": 0.13511186838150024, "learning_rate": 6.0506075969550725e-05, "loss": 0.0144, "step": 27660 }, { "grad_norm": 0.16938012838363647, "learning_rate": 6.047913176581609e-05, "loss": 0.013, "step": 27670 }, { "grad_norm": 0.13385778665542603, "learning_rate": 6.0452184378797904e-05, "loss": 0.014, "step": 27680 }, { "grad_norm": 0.10821831971406937, "learning_rate": 6.042523381668209e-05, "loss": 0.0144, "step": 27690 }, { "grad_norm": 0.08668509125709534, "learning_rate": 6.03982800876555e-05, "loss": 0.0119, "step": 27700 }, { "grad_norm": 0.10926337540149689, "learning_rate": 6.0371323199905975e-05, "loss": 0.0147, "step": 27710 }, { "grad_norm": 0.15266600251197815, "learning_rate": 6.03443631616223e-05, "loss": 0.0151, "step": 27720 }, { "grad_norm": 0.13100217282772064, "learning_rate": 6.031739998099421e-05, "loss": 0.013, "step": 27730 }, { "grad_norm": 0.11526843905448914, "learning_rate": 6.029043366621243e-05, "loss": 0.0136, "step": 27740 }, { "grad_norm": 0.20999619364738464, "learning_rate": 6.0263464225468615e-05, "loss": 0.0139, "step": 27750 }, { "grad_norm": 0.1545216143131256, "learning_rate": 6.023649166695534e-05, "loss": 0.0147, "step": 27760 }, { "grad_norm": 0.09839431196451187, "learning_rate": 6.0209515998866186e-05, "loss": 0.0128, "step": 27770 }, { "grad_norm": 0.10197681188583374, "learning_rate": 6.018253722939563e-05, "loss": 0.0144, "step": 27780 }, { "grad_norm": 0.10998216271400452, "learning_rate": 6.015555536673914e-05, "loss": 0.0136, "step": 27790 }, { "grad_norm": 0.1395629197359085, "learning_rate": 6.0128570419093054e-05, "loss": 0.0142, "step": 27800 }, { "grad_norm": 0.09883852303028107, "learning_rate": 6.010158239465471e-05, "loss": 0.0123, "step": 27810 }, { "grad_norm": 0.13194090127944946, "learning_rate": 6.007459130162235e-05, "loss": 0.0135, "step": 27820 }, { "grad_norm": 0.13749879598617554, "learning_rate": 6.004759714819516e-05, "loss": 0.0144, "step": 27830 }, { "grad_norm": 0.11537046730518341, "learning_rate": 6.002059994257323e-05, "loss": 0.0115, "step": 27840 }, { "grad_norm": 0.14245359599590302, "learning_rate": 5.999359969295764e-05, "loss": 0.013, "step": 27850 }, { "grad_norm": 0.13024647533893585, "learning_rate": 5.9966596407550314e-05, "loss": 0.013, "step": 27860 }, { "grad_norm": 0.12605896592140198, "learning_rate": 5.993959009455416e-05, "loss": 0.0135, "step": 27870 }, { "grad_norm": 0.13947083055973053, "learning_rate": 5.991258076217298e-05, "loss": 0.0152, "step": 27880 }, { "grad_norm": 0.13134361803531647, "learning_rate": 5.988556841861147e-05, "loss": 0.0144, "step": 27890 }, { "grad_norm": 0.14372359216213226, "learning_rate": 5.985855307207531e-05, "loss": 0.0143, "step": 27900 }, { "grad_norm": 0.14442217350006104, "learning_rate": 5.9831534730771e-05, "loss": 0.0162, "step": 27910 }, { "grad_norm": 0.11976713687181473, "learning_rate": 5.980451340290605e-05, "loss": 0.0156, "step": 27920 }, { "grad_norm": 0.11688560992479324, "learning_rate": 5.97774890966888e-05, "loss": 0.0123, "step": 27930 }, { "grad_norm": 0.1426674723625183, "learning_rate": 5.975046182032851e-05, "loss": 0.0134, "step": 27940 }, { "grad_norm": 0.11215920001268387, "learning_rate": 5.972343158203537e-05, "loss": 0.0121, "step": 27950 }, { "grad_norm": 0.13427235186100006, "learning_rate": 5.969639839002045e-05, "loss": 0.0147, "step": 27960 }, { "grad_norm": 0.1401950716972351, "learning_rate": 5.966936225249572e-05, "loss": 0.0131, "step": 27970 }, { "grad_norm": 0.10126380622386932, "learning_rate": 5.9642323177674044e-05, "loss": 0.0125, "step": 27980 }, { "grad_norm": 0.13676881790161133, "learning_rate": 5.9615281173769154e-05, "loss": 0.0148, "step": 27990 }, { "grad_norm": 0.11760519444942474, "learning_rate": 5.958823624899574e-05, "loss": 0.0143, "step": 28000 }, { "grad_norm": 0.14263060688972473, "learning_rate": 5.956118841156933e-05, "loss": 0.014, "step": 28010 }, { "grad_norm": 0.11990267038345337, "learning_rate": 5.953413766970631e-05, "loss": 0.014, "step": 28020 }, { "grad_norm": 0.10758461803197861, "learning_rate": 5.9507084031624e-05, "loss": 0.0129, "step": 28030 }, { "grad_norm": 0.14549042284488678, "learning_rate": 5.948002750554058e-05, "loss": 0.0142, "step": 28040 }, { "grad_norm": 0.10712353885173798, "learning_rate": 5.9452968099675124e-05, "loss": 0.0128, "step": 28050 }, { "grad_norm": 0.11610545217990875, "learning_rate": 5.9425905822247527e-05, "loss": 0.0134, "step": 28060 }, { "grad_norm": 0.108518086373806, "learning_rate": 5.939884068147864e-05, "loss": 0.012, "step": 28070 }, { "grad_norm": 0.13236810266971588, "learning_rate": 5.937177268559011e-05, "loss": 0.0148, "step": 28080 }, { "grad_norm": 0.16037513315677643, "learning_rate": 5.934470184280448e-05, "loss": 0.0145, "step": 28090 }, { "grad_norm": 0.10309937596321106, "learning_rate": 5.931762816134516e-05, "loss": 0.0125, "step": 28100 }, { "grad_norm": 0.14088772237300873, "learning_rate": 5.9290551649436434e-05, "loss": 0.0161, "step": 28110 }, { "grad_norm": 0.14175443351268768, "learning_rate": 5.9263472315303416e-05, "loss": 0.0139, "step": 28120 }, { "grad_norm": 0.15937262773513794, "learning_rate": 5.9236390167172096e-05, "loss": 0.0147, "step": 28130 }, { "grad_norm": 0.12827889621257782, "learning_rate": 5.920930521326932e-05, "loss": 0.014, "step": 28140 }, { "grad_norm": 0.15988044440746307, "learning_rate": 5.918221746182276e-05, "loss": 0.0142, "step": 28150 }, { "grad_norm": 0.1184987872838974, "learning_rate": 5.9155126921061e-05, "loss": 0.0138, "step": 28160 }, { "grad_norm": 0.10295838117599487, "learning_rate": 5.91280335992134e-05, "loss": 0.0131, "step": 28170 }, { "grad_norm": 0.10622610151767731, "learning_rate": 5.91009375045102e-05, "loss": 0.0136, "step": 28180 }, { "grad_norm": 0.15098433196544647, "learning_rate": 5.9073838645182476e-05, "loss": 0.0126, "step": 28190 }, { "grad_norm": 0.1333729475736618, "learning_rate": 5.904673702946217e-05, "loss": 0.0118, "step": 28200 }, { "grad_norm": 0.13606004416942596, "learning_rate": 5.9019632665582004e-05, "loss": 0.0126, "step": 28210 }, { "grad_norm": 0.1733241081237793, "learning_rate": 5.899252556177559e-05, "loss": 0.0125, "step": 28220 }, { "grad_norm": 0.1442093849182129, "learning_rate": 5.896541572627735e-05, "loss": 0.0117, "step": 28230 }, { "grad_norm": 0.11618904024362564, "learning_rate": 5.893830316732253e-05, "loss": 0.0135, "step": 28240 }, { "grad_norm": 0.20123997330665588, "learning_rate": 5.8911187893147214e-05, "loss": 0.0158, "step": 28250 }, { "grad_norm": 0.16240975260734558, "learning_rate": 5.888406991198828e-05, "loss": 0.016, "step": 28260 }, { "grad_norm": 0.1517639309167862, "learning_rate": 5.885694923208349e-05, "loss": 0.0148, "step": 28270 }, { "grad_norm": 0.0962788462638855, "learning_rate": 5.882982586167138e-05, "loss": 0.0126, "step": 28280 }, { "grad_norm": 0.11824341863393784, "learning_rate": 5.880269980899131e-05, "loss": 0.013, "step": 28290 }, { "grad_norm": 0.11046972125768661, "learning_rate": 5.8775571082283465e-05, "loss": 0.0123, "step": 28300 }, { "grad_norm": 0.15933232009410858, "learning_rate": 5.8748439689788824e-05, "loss": 0.0142, "step": 28310 }, { "grad_norm": 0.11928635090589523, "learning_rate": 5.87213056397492e-05, "loss": 0.0124, "step": 28320 }, { "grad_norm": 0.11275859922170639, "learning_rate": 5.869416894040719e-05, "loss": 0.0127, "step": 28330 }, { "grad_norm": 0.13171160221099854, "learning_rate": 5.866702960000621e-05, "loss": 0.0139, "step": 28340 }, { "grad_norm": 0.09624268114566803, "learning_rate": 5.863988762679048e-05, "loss": 0.012, "step": 28350 }, { "grad_norm": 0.1613750457763672, "learning_rate": 5.8612743029005e-05, "loss": 0.0138, "step": 28360 }, { "grad_norm": 0.14632144570350647, "learning_rate": 5.858559581489561e-05, "loss": 0.0131, "step": 28370 }, { "grad_norm": 0.10738196223974228, "learning_rate": 5.85584459927089e-05, "loss": 0.0135, "step": 28380 }, { "grad_norm": 0.10804779827594757, "learning_rate": 5.853129357069227e-05, "loss": 0.0141, "step": 28390 }, { "grad_norm": 0.11524812877178192, "learning_rate": 5.8504138557093913e-05, "loss": 0.0119, "step": 28400 }, { "grad_norm": 0.11252560466527939, "learning_rate": 5.8476980960162784e-05, "loss": 0.0125, "step": 28410 }, { "grad_norm": 0.14973148703575134, "learning_rate": 5.844982078814868e-05, "loss": 0.0124, "step": 28420 }, { "grad_norm": 0.09626806527376175, "learning_rate": 5.842265804930211e-05, "loss": 0.0108, "step": 28430 }, { "grad_norm": 0.10744381695985794, "learning_rate": 5.839549275187444e-05, "loss": 0.0127, "step": 28440 }, { "grad_norm": 0.14850005507469177, "learning_rate": 5.836832490411771e-05, "loss": 0.012, "step": 28450 }, { "grad_norm": 0.11999146640300751, "learning_rate": 5.834115451428485e-05, "loss": 0.0113, "step": 28460 }, { "grad_norm": 0.12118370085954666, "learning_rate": 5.831398159062946e-05, "loss": 0.0124, "step": 28470 }, { "grad_norm": 0.14566949009895325, "learning_rate": 5.828680614140599e-05, "loss": 0.0134, "step": 28480 }, { "grad_norm": 0.11880069971084595, "learning_rate": 5.825962817486962e-05, "loss": 0.0128, "step": 28490 }, { "grad_norm": 0.19775637984275818, "learning_rate": 5.823244769927629e-05, "loss": 0.0147, "step": 28500 }, { "grad_norm": 0.13626743853092194, "learning_rate": 5.8205264722882716e-05, "loss": 0.0128, "step": 28510 }, { "grad_norm": 0.08862190693616867, "learning_rate": 5.817807925394636e-05, "loss": 0.0121, "step": 28520 }, { "grad_norm": 0.13211706280708313, "learning_rate": 5.815089130072546e-05, "loss": 0.0117, "step": 28530 }, { "grad_norm": 0.1838804930448532, "learning_rate": 5.8123700871479e-05, "loss": 0.013, "step": 28540 }, { "grad_norm": 0.13195794820785522, "learning_rate": 5.809650797446671e-05, "loss": 0.0124, "step": 28550 }, { "grad_norm": 0.15224407613277435, "learning_rate": 5.806931261794907e-05, "loss": 0.0135, "step": 28560 }, { "grad_norm": 0.1506509780883789, "learning_rate": 5.804211481018731e-05, "loss": 0.0148, "step": 28570 }, { "grad_norm": 0.16694751381874084, "learning_rate": 5.801491455944341e-05, "loss": 0.0161, "step": 28580 }, { "grad_norm": 0.12198687344789505, "learning_rate": 5.79877118739801e-05, "loss": 0.0123, "step": 28590 }, { "grad_norm": 0.1126212403178215, "learning_rate": 5.7960506762060816e-05, "loss": 0.0136, "step": 28600 }, { "grad_norm": 0.12036170065402985, "learning_rate": 5.793329923194977e-05, "loss": 0.0121, "step": 28610 }, { "grad_norm": 0.13302257657051086, "learning_rate": 5.790608929191187e-05, "loss": 0.0131, "step": 28620 }, { "grad_norm": 0.1306198537349701, "learning_rate": 5.78788769502128e-05, "loss": 0.0136, "step": 28630 }, { "grad_norm": 0.14256523549556732, "learning_rate": 5.785166221511894e-05, "loss": 0.013, "step": 28640 }, { "grad_norm": 0.12976056337356567, "learning_rate": 5.7824445094897415e-05, "loss": 0.0125, "step": 28650 }, { "grad_norm": 0.1407020539045334, "learning_rate": 5.7797225597816065e-05, "loss": 0.0126, "step": 28660 }, { "grad_norm": 0.11716905981302261, "learning_rate": 5.777000373214345e-05, "loss": 0.0143, "step": 28670 }, { "grad_norm": 0.12056748569011688, "learning_rate": 5.774277950614885e-05, "loss": 0.0125, "step": 28680 }, { "grad_norm": 0.12786298990249634, "learning_rate": 5.771555292810227e-05, "loss": 0.0132, "step": 28690 }, { "grad_norm": 0.13770811259746552, "learning_rate": 5.768832400627444e-05, "loss": 0.0135, "step": 28700 }, { "grad_norm": 0.09897526353597641, "learning_rate": 5.7661092748936775e-05, "loss": 0.0136, "step": 28710 }, { "grad_norm": 0.15078744292259216, "learning_rate": 5.76338591643614e-05, "loss": 0.0132, "step": 28720 }, { "grad_norm": 0.1244683712720871, "learning_rate": 5.760662326082118e-05, "loss": 0.0124, "step": 28730 }, { "grad_norm": 0.14995937049388885, "learning_rate": 5.757938504658965e-05, "loss": 0.0141, "step": 28740 }, { "grad_norm": 0.13233621418476105, "learning_rate": 5.755214452994107e-05, "loss": 0.0146, "step": 28750 }, { "grad_norm": 0.18509243428707123, "learning_rate": 5.752490171915039e-05, "loss": 0.0152, "step": 28760 }, { "grad_norm": 0.12229160964488983, "learning_rate": 5.749765662249324e-05, "loss": 0.012, "step": 28770 }, { "grad_norm": 0.11450058966875076, "learning_rate": 5.747040924824596e-05, "loss": 0.0133, "step": 28780 }, { "grad_norm": 0.14134079217910767, "learning_rate": 5.7443159604685613e-05, "loss": 0.0128, "step": 28790 }, { "grad_norm": 0.12041401863098145, "learning_rate": 5.74159077000899e-05, "loss": 0.0125, "step": 28800 }, { "grad_norm": 0.17427939176559448, "learning_rate": 5.7388653542737235e-05, "loss": 0.0148, "step": 28810 }, { "grad_norm": 0.11827235668897629, "learning_rate": 5.736139714090672e-05, "loss": 0.0143, "step": 28820 }, { "grad_norm": 0.13481317460536957, "learning_rate": 5.73341385028781e-05, "loss": 0.012, "step": 28830 }, { "grad_norm": 0.1300133615732193, "learning_rate": 5.7306877636931855e-05, "loss": 0.0128, "step": 28840 }, { "grad_norm": 0.19497841596603394, "learning_rate": 5.7279614551349125e-05, "loss": 0.0143, "step": 28850 }, { "grad_norm": 0.12655901908874512, "learning_rate": 5.725234925441169e-05, "loss": 0.0146, "step": 28860 }, { "grad_norm": 0.10500729829072952, "learning_rate": 5.7225081754402044e-05, "loss": 0.0153, "step": 28870 }, { "grad_norm": 0.1846269965171814, "learning_rate": 5.7197812059603326e-05, "loss": 0.0144, "step": 28880 }, { "grad_norm": 0.1102713793516159, "learning_rate": 5.717054017829934e-05, "loss": 0.0119, "step": 28890 }, { "grad_norm": 0.12103429436683655, "learning_rate": 5.7143266118774584e-05, "loss": 0.0138, "step": 28900 }, { "grad_norm": 0.11076052486896515, "learning_rate": 5.711598988931418e-05, "loss": 0.0148, "step": 28910 }, { "grad_norm": 0.10336827486753464, "learning_rate": 5.7088711498203954e-05, "loss": 0.0133, "step": 28920 }, { "grad_norm": 0.11881240457296371, "learning_rate": 5.706143095373033e-05, "loss": 0.0129, "step": 28930 }, { "grad_norm": 0.14723661541938782, "learning_rate": 5.703414826418042e-05, "loss": 0.0132, "step": 28940 }, { "grad_norm": 0.1266966164112091, "learning_rate": 5.7006863437842007e-05, "loss": 0.0142, "step": 28950 }, { "grad_norm": 0.1140362098813057, "learning_rate": 5.697957648300348e-05, "loss": 0.0143, "step": 28960 }, { "grad_norm": 0.12039662152528763, "learning_rate": 5.695228740795391e-05, "loss": 0.0132, "step": 28970 }, { "grad_norm": 0.1323864907026291, "learning_rate": 5.6924996220982985e-05, "loss": 0.0131, "step": 28980 }, { "grad_norm": 0.1448695808649063, "learning_rate": 5.6897702930381045e-05, "loss": 0.0129, "step": 28990 }, { "grad_norm": 0.09877323359251022, "learning_rate": 5.687040754443908e-05, "loss": 0.0133, "step": 29000 }, { "grad_norm": 0.11814340949058533, "learning_rate": 5.6843110071448725e-05, "loss": 0.0131, "step": 29010 }, { "grad_norm": 0.14848698675632477, "learning_rate": 5.6815810519702194e-05, "loss": 0.0131, "step": 29020 }, { "grad_norm": 0.09856764227151871, "learning_rate": 5.6788508897492396e-05, "loss": 0.0128, "step": 29030 }, { "grad_norm": 0.13559354841709137, "learning_rate": 5.676120521311282e-05, "loss": 0.0128, "step": 29040 }, { "grad_norm": 0.1282961666584015, "learning_rate": 5.6733899474857634e-05, "loss": 0.0136, "step": 29050 }, { "grad_norm": 0.10631696879863739, "learning_rate": 5.670659169102157e-05, "loss": 0.014, "step": 29060 }, { "grad_norm": 0.11334876716136932, "learning_rate": 5.6679281869900044e-05, "loss": 0.0128, "step": 29070 }, { "grad_norm": 0.13783904910087585, "learning_rate": 5.6651970019789045e-05, "loss": 0.0137, "step": 29080 }, { "grad_norm": 0.12057912349700928, "learning_rate": 5.662465614898519e-05, "loss": 0.0116, "step": 29090 }, { "grad_norm": 0.11570675671100616, "learning_rate": 5.6597340265785695e-05, "loss": 0.0127, "step": 29100 }, { "grad_norm": 0.133440300822258, "learning_rate": 5.657002237848843e-05, "loss": 0.0123, "step": 29110 }, { "grad_norm": 0.11938497424125671, "learning_rate": 5.654270249539183e-05, "loss": 0.0137, "step": 29120 }, { "grad_norm": 0.09008229523897171, "learning_rate": 5.651538062479498e-05, "loss": 0.013, "step": 29130 }, { "grad_norm": 0.10146951675415039, "learning_rate": 5.648805677499751e-05, "loss": 0.0128, "step": 29140 }, { "grad_norm": 0.15678051114082336, "learning_rate": 5.646073095429969e-05, "loss": 0.0139, "step": 29150 }, { "grad_norm": 0.16352586448192596, "learning_rate": 5.643340317100241e-05, "loss": 0.0133, "step": 29160 }, { "grad_norm": 0.11023875325918198, "learning_rate": 5.64060734334071e-05, "loss": 0.0127, "step": 29170 }, { "grad_norm": 0.10674617439508438, "learning_rate": 5.637874174981583e-05, "loss": 0.0134, "step": 29180 }, { "grad_norm": 0.12369517982006073, "learning_rate": 5.635140812853124e-05, "loss": 0.0129, "step": 29190 }, { "grad_norm": 0.08020330965518951, "learning_rate": 5.6324072577856544e-05, "loss": 0.0117, "step": 29200 }, { "grad_norm": 0.141194686293602, "learning_rate": 5.629673510609559e-05, "loss": 0.014, "step": 29210 }, { "grad_norm": 0.14369738101959229, "learning_rate": 5.626939572155276e-05, "loss": 0.0139, "step": 29220 }, { "grad_norm": 0.10301477462053299, "learning_rate": 5.6242054432533054e-05, "loss": 0.0132, "step": 29230 }, { "grad_norm": 0.12129510939121246, "learning_rate": 5.621471124734201e-05, "loss": 0.0144, "step": 29240 }, { "grad_norm": 0.11981639266014099, "learning_rate": 5.6187366174285794e-05, "loss": 0.0119, "step": 29250 }, { "grad_norm": 0.08494983613491058, "learning_rate": 5.616001922167109e-05, "loss": 0.0137, "step": 29260 }, { "grad_norm": 0.14038506150245667, "learning_rate": 5.61326703978052e-05, "loss": 0.0138, "step": 29270 }, { "grad_norm": 0.14121751487255096, "learning_rate": 5.6105319710995964e-05, "loss": 0.013, "step": 29280 }, { "grad_norm": 0.11085744947195053, "learning_rate": 5.60779671695518e-05, "loss": 0.0135, "step": 29290 }, { "grad_norm": 0.10101135820150375, "learning_rate": 5.6050612781781684e-05, "loss": 0.0124, "step": 29300 }, { "grad_norm": 0.16417823731899261, "learning_rate": 5.602325655599516e-05, "loss": 0.0127, "step": 29310 }, { "grad_norm": 0.1448914259672165, "learning_rate": 5.599589850050234e-05, "loss": 0.0135, "step": 29320 }, { "grad_norm": 0.1162995919585228, "learning_rate": 5.5968538623613874e-05, "loss": 0.0121, "step": 29330 }, { "grad_norm": 0.13375632464885712, "learning_rate": 5.594117693364095e-05, "loss": 0.0136, "step": 29340 }, { "grad_norm": 0.16385184228420258, "learning_rate": 5.591381343889535e-05, "loss": 0.0139, "step": 29350 }, { "grad_norm": 0.15355230867862701, "learning_rate": 5.5886448147689355e-05, "loss": 0.0124, "step": 29360 }, { "grad_norm": 0.11595812439918518, "learning_rate": 5.585908106833585e-05, "loss": 0.012, "step": 29370 }, { "grad_norm": 0.14732283353805542, "learning_rate": 5.5831712209148226e-05, "loss": 0.0125, "step": 29380 }, { "grad_norm": 0.13654351234436035, "learning_rate": 5.58043415784404e-05, "loss": 0.0127, "step": 29390 }, { "grad_norm": 0.15145660936832428, "learning_rate": 5.577696918452686e-05, "loss": 0.0124, "step": 29400 }, { "grad_norm": 0.11690567433834076, "learning_rate": 5.5749595035722604e-05, "loss": 0.012, "step": 29410 }, { "grad_norm": 0.15763896703720093, "learning_rate": 5.5722219140343193e-05, "loss": 0.0133, "step": 29420 }, { "grad_norm": 0.14827577769756317, "learning_rate": 5.56948415067047e-05, "loss": 0.0157, "step": 29430 }, { "grad_norm": 0.10706058144569397, "learning_rate": 5.5667462143123704e-05, "loss": 0.0114, "step": 29440 }, { "grad_norm": 0.13983319699764252, "learning_rate": 5.564008105791737e-05, "loss": 0.0123, "step": 29450 }, { "grad_norm": 0.11875741183757782, "learning_rate": 5.5612698259403316e-05, "loss": 0.0141, "step": 29460 }, { "grad_norm": 0.10348864644765854, "learning_rate": 5.5585313755899724e-05, "loss": 0.0143, "step": 29470 }, { "grad_norm": 0.10690343379974365, "learning_rate": 5.5557927555725285e-05, "loss": 0.0132, "step": 29480 }, { "grad_norm": 0.14350125193595886, "learning_rate": 5.55305396671992e-05, "loss": 0.0132, "step": 29490 }, { "grad_norm": 0.0971696749329567, "learning_rate": 5.55031500986412e-05, "loss": 0.0142, "step": 29500 }, { "grad_norm": 0.12159064412117004, "learning_rate": 5.547575885837149e-05, "loss": 0.0132, "step": 29510 }, { "grad_norm": 0.09155336022377014, "learning_rate": 5.5448365954710825e-05, "loss": 0.013, "step": 29520 }, { "grad_norm": 0.10538607090711594, "learning_rate": 5.5420971395980446e-05, "loss": 0.0117, "step": 29530 }, { "grad_norm": 0.09843547642230988, "learning_rate": 5.539357519050209e-05, "loss": 0.0128, "step": 29540 }, { "grad_norm": 0.14216919243335724, "learning_rate": 5.536617734659799e-05, "loss": 0.0111, "step": 29550 }, { "grad_norm": 0.1281747967004776, "learning_rate": 5.533877787259091e-05, "loss": 0.0123, "step": 29560 }, { "grad_norm": 0.1136767566204071, "learning_rate": 5.5311376776804044e-05, "loss": 0.013, "step": 29570 }, { "grad_norm": 0.09506802260875702, "learning_rate": 5.528397406756118e-05, "loss": 0.0154, "step": 29580 }, { "grad_norm": 0.11037635058164597, "learning_rate": 5.525656975318652e-05, "loss": 0.0127, "step": 29590 }, { "grad_norm": 0.15846095979213715, "learning_rate": 5.522916384200474e-05, "loss": 0.0128, "step": 29600 }, { "grad_norm": 0.09485381096601486, "learning_rate": 5.520175634234106e-05, "loss": 0.0121, "step": 29610 }, { "grad_norm": 0.12081679701805115, "learning_rate": 5.517434726252113e-05, "loss": 0.0128, "step": 29620 }, { "grad_norm": 0.16066402196884155, "learning_rate": 5.514693661087113e-05, "loss": 0.0143, "step": 29630 }, { "grad_norm": 0.1384875476360321, "learning_rate": 5.511952439571769e-05, "loss": 0.0146, "step": 29640 }, { "grad_norm": 0.12071863561868668, "learning_rate": 5.509211062538791e-05, "loss": 0.0127, "step": 29650 }, { "grad_norm": 0.13418205082416534, "learning_rate": 5.506469530820939e-05, "loss": 0.0123, "step": 29660 }, { "grad_norm": 0.11232169717550278, "learning_rate": 5.503727845251014e-05, "loss": 0.0116, "step": 29670 }, { "grad_norm": 0.1171649917960167, "learning_rate": 5.50098600666187e-05, "loss": 0.0128, "step": 29680 }, { "grad_norm": 0.11509578675031662, "learning_rate": 5.498244015886406e-05, "loss": 0.0122, "step": 29690 }, { "grad_norm": 0.12904879450798035, "learning_rate": 5.495501873757565e-05, "loss": 0.0135, "step": 29700 }, { "grad_norm": 0.1627238541841507, "learning_rate": 5.492759581108336e-05, "loss": 0.0127, "step": 29710 }, { "grad_norm": 0.11782065778970718, "learning_rate": 5.490017138771759e-05, "loss": 0.0118, "step": 29720 }, { "grad_norm": 0.13731995224952698, "learning_rate": 5.487274547580912e-05, "loss": 0.0124, "step": 29730 }, { "grad_norm": 0.10016407817602158, "learning_rate": 5.484531808368923e-05, "loss": 0.0113, "step": 29740 }, { "grad_norm": 0.10147760808467865, "learning_rate": 5.4817889219689656e-05, "loss": 0.0124, "step": 29750 }, { "grad_norm": 0.146714448928833, "learning_rate": 5.4790458892142536e-05, "loss": 0.0122, "step": 29760 }, { "grad_norm": 0.12389257550239563, "learning_rate": 5.476302710938048e-05, "loss": 0.0128, "step": 29770 }, { "grad_norm": 0.11727964878082275, "learning_rate": 5.473559387973657e-05, "loss": 0.0132, "step": 29780 }, { "grad_norm": 0.14293889701366425, "learning_rate": 5.470815921154425e-05, "loss": 0.0126, "step": 29790 }, { "grad_norm": 0.1293559968471527, "learning_rate": 5.468072311313749e-05, "loss": 0.0141, "step": 29800 }, { "grad_norm": 0.12025792896747589, "learning_rate": 5.465328559285063e-05, "loss": 0.0134, "step": 29810 }, { "grad_norm": 0.09346222132444382, "learning_rate": 5.462584665901849e-05, "loss": 0.0122, "step": 29820 }, { "grad_norm": 0.13229501247406006, "learning_rate": 5.4598406319976235e-05, "loss": 0.0131, "step": 29830 }, { "grad_norm": 0.11045314371585846, "learning_rate": 5.457096458405958e-05, "loss": 0.013, "step": 29840 }, { "grad_norm": 0.11867164820432663, "learning_rate": 5.454352145960457e-05, "loss": 0.0132, "step": 29850 }, { "grad_norm": 0.09939072281122208, "learning_rate": 5.4516076954947715e-05, "loss": 0.0131, "step": 29860 }, { "grad_norm": 0.14665110409259796, "learning_rate": 5.448863107842591e-05, "loss": 0.0127, "step": 29870 }, { "grad_norm": 0.1234838217496872, "learning_rate": 5.446118383837651e-05, "loss": 0.0132, "step": 29880 }, { "grad_norm": 0.11418157815933228, "learning_rate": 5.443373524313722e-05, "loss": 0.0133, "step": 29890 }, { "grad_norm": 0.12914375960826874, "learning_rate": 5.440628530104626e-05, "loss": 0.0128, "step": 29900 }, { "grad_norm": 0.10947774350643158, "learning_rate": 5.4378834020442146e-05, "loss": 0.0124, "step": 29910 }, { "grad_norm": 0.12258150428533554, "learning_rate": 5.4351381409663884e-05, "loss": 0.0138, "step": 29920 }, { "grad_norm": 0.11474057286977768, "learning_rate": 5.432392747705084e-05, "loss": 0.0139, "step": 29930 }, { "grad_norm": 0.14778386056423187, "learning_rate": 5.429647223094278e-05, "loss": 0.0131, "step": 29940 }, { "grad_norm": 0.10678399354219437, "learning_rate": 5.4269015679679924e-05, "loss": 0.0133, "step": 29950 }, { "grad_norm": 0.10904999077320099, "learning_rate": 5.424155783160281e-05, "loss": 0.0111, "step": 29960 }, { "grad_norm": 0.11282510310411453, "learning_rate": 5.4214098695052415e-05, "loss": 0.0123, "step": 29970 }, { "grad_norm": 0.120896115899086, "learning_rate": 5.418663827837012e-05, "loss": 0.0118, "step": 29980 }, { "grad_norm": 0.11351677775382996, "learning_rate": 5.415917658989763e-05, "loss": 0.0112, "step": 29990 }, { "grad_norm": 0.1382218599319458, "learning_rate": 5.413171363797713e-05, "loss": 0.0117, "step": 30000 }, { "grad_norm": 0.11813255399465561, "learning_rate": 5.4104249430951116e-05, "loss": 0.0134, "step": 30010 }, { "grad_norm": 0.12482161819934845, "learning_rate": 5.4076783977162494e-05, "loss": 0.0131, "step": 30020 }, { "grad_norm": 0.10644228011369705, "learning_rate": 5.4049317284954525e-05, "loss": 0.0122, "step": 30030 }, { "grad_norm": 0.10211630165576935, "learning_rate": 5.4021849362670884e-05, "loss": 0.0131, "step": 30040 }, { "grad_norm": 0.11399929970502853, "learning_rate": 5.3994380218655604e-05, "loss": 0.0136, "step": 30050 }, { "grad_norm": 0.11692990362644196, "learning_rate": 5.396690986125309e-05, "loss": 0.0123, "step": 30060 }, { "grad_norm": 0.12755197286605835, "learning_rate": 5.3939438298808075e-05, "loss": 0.0141, "step": 30070 }, { "grad_norm": 0.12847363948822021, "learning_rate": 5.3911965539665744e-05, "loss": 0.0131, "step": 30080 }, { "grad_norm": 0.14217343926429749, "learning_rate": 5.388449159217156e-05, "loss": 0.0136, "step": 30090 }, { "grad_norm": 0.15776130557060242, "learning_rate": 5.3857016464671385e-05, "loss": 0.0164, "step": 30100 }, { "grad_norm": 0.1292692869901657, "learning_rate": 5.382954016551146e-05, "loss": 0.0128, "step": 30110 }, { "grad_norm": 0.15386663377285004, "learning_rate": 5.380206270303835e-05, "loss": 0.0119, "step": 30120 }, { "grad_norm": 0.13796702027320862, "learning_rate": 5.377458408559897e-05, "loss": 0.0141, "step": 30130 }, { "grad_norm": 0.13664153218269348, "learning_rate": 5.374710432154061e-05, "loss": 0.0135, "step": 30140 }, { "grad_norm": 0.15443019568920135, "learning_rate": 5.3719623419210886e-05, "loss": 0.0133, "step": 30150 }, { "grad_norm": 0.14151450991630554, "learning_rate": 5.3692141386957786e-05, "loss": 0.0125, "step": 30160 }, { "grad_norm": 0.12328960001468658, "learning_rate": 5.3664658233129616e-05, "loss": 0.0122, "step": 30170 }, { "grad_norm": 0.16392770409584045, "learning_rate": 5.363717396607504e-05, "loss": 0.0136, "step": 30180 }, { "grad_norm": 0.08714262396097183, "learning_rate": 5.360968859414305e-05, "loss": 0.0109, "step": 30190 }, { "grad_norm": 0.1405133455991745, "learning_rate": 5.358220212568295e-05, "loss": 0.0134, "step": 30200 }, { "grad_norm": 0.12865212559700012, "learning_rate": 5.355471456904444e-05, "loss": 0.0129, "step": 30210 }, { "grad_norm": 0.14412392675876617, "learning_rate": 5.3527225932577495e-05, "loss": 0.0122, "step": 30220 }, { "grad_norm": 0.12046249955892563, "learning_rate": 5.349973622463246e-05, "loss": 0.0126, "step": 30230 }, { "grad_norm": 0.14408273994922638, "learning_rate": 5.3472245453559956e-05, "loss": 0.0133, "step": 30240 }, { "grad_norm": 0.1214710995554924, "learning_rate": 5.3444753627710955e-05, "loss": 0.0126, "step": 30250 }, { "grad_norm": 0.13290360569953918, "learning_rate": 5.341726075543676e-05, "loss": 0.0129, "step": 30260 }, { "grad_norm": 0.11194293946027756, "learning_rate": 5.338976684508898e-05, "loss": 0.0146, "step": 30270 }, { "grad_norm": 0.08996652066707611, "learning_rate": 5.336227190501953e-05, "loss": 0.0115, "step": 30280 }, { "grad_norm": 0.1344270408153534, "learning_rate": 5.3334775943580664e-05, "loss": 0.0114, "step": 30290 }, { "grad_norm": 0.12503117322921753, "learning_rate": 5.330727896912491e-05, "loss": 0.014, "step": 30300 }, { "grad_norm": 0.12292784452438354, "learning_rate": 5.327978099000511e-05, "loss": 0.0124, "step": 30310 }, { "grad_norm": 0.09785839170217514, "learning_rate": 5.3252282014574465e-05, "loss": 0.0116, "step": 30320 }, { "grad_norm": 0.13018609583377838, "learning_rate": 5.322478205118641e-05, "loss": 0.0152, "step": 30330 }, { "grad_norm": 0.09454476833343506, "learning_rate": 5.3197281108194704e-05, "loss": 0.0131, "step": 30340 }, { "grad_norm": 0.11244232952594757, "learning_rate": 5.316977919395342e-05, "loss": 0.0135, "step": 30350 }, { "grad_norm": 0.12500597536563873, "learning_rate": 5.314227631681691e-05, "loss": 0.0121, "step": 30360 }, { "grad_norm": 0.12153227627277374, "learning_rate": 5.311477248513982e-05, "loss": 0.0119, "step": 30370 }, { "grad_norm": 0.10739487409591675, "learning_rate": 5.30872677072771e-05, "loss": 0.0117, "step": 30380 }, { "grad_norm": 0.10127223283052444, "learning_rate": 5.3059761991583954e-05, "loss": 0.0126, "step": 30390 }, { "grad_norm": 0.11690594255924225, "learning_rate": 5.303225534641592e-05, "loss": 0.0147, "step": 30400 }, { "grad_norm": 0.11100757867097855, "learning_rate": 5.300474778012875e-05, "loss": 0.0133, "step": 30410 }, { "grad_norm": 0.1432199776172638, "learning_rate": 5.297723930107855e-05, "loss": 0.0134, "step": 30420 }, { "grad_norm": 0.11487796902656555, "learning_rate": 5.294972991762167e-05, "loss": 0.0117, "step": 30430 }, { "grad_norm": 0.14220869541168213, "learning_rate": 5.292221963811472e-05, "loss": 0.0113, "step": 30440 }, { "grad_norm": 0.09959404170513153, "learning_rate": 5.28947084709146e-05, "loss": 0.0118, "step": 30450 }, { "grad_norm": 0.12596140801906586, "learning_rate": 5.2867196424378465e-05, "loss": 0.0124, "step": 30460 }, { "grad_norm": 0.10860495269298553, "learning_rate": 5.2839683506863765e-05, "loss": 0.0103, "step": 30470 }, { "grad_norm": 0.1281159222126007, "learning_rate": 5.281216972672821e-05, "loss": 0.0122, "step": 30480 }, { "grad_norm": 0.10227783024311066, "learning_rate": 5.278465509232973e-05, "loss": 0.012, "step": 30490 }, { "grad_norm": 0.12227421998977661, "learning_rate": 5.275713961202655e-05, "loss": 0.0108, "step": 30500 }, { "grad_norm": 0.08172754943370819, "learning_rate": 5.2729623294177165e-05, "loss": 0.0113, "step": 30510 }, { "grad_norm": 0.10351616144180298, "learning_rate": 5.270210614714028e-05, "loss": 0.0138, "step": 30520 }, { "grad_norm": 0.13295747339725494, "learning_rate": 5.267458817927491e-05, "loss": 0.0121, "step": 30530 }, { "grad_norm": 0.10439648479223251, "learning_rate": 5.264706939894026e-05, "loss": 0.0131, "step": 30540 }, { "grad_norm": 0.12415914237499237, "learning_rate": 5.261954981449584e-05, "loss": 0.0127, "step": 30550 }, { "grad_norm": 0.09792747348546982, "learning_rate": 5.2592029434301324e-05, "loss": 0.0119, "step": 30560 }, { "grad_norm": 0.10448843985795975, "learning_rate": 5.256450826671672e-05, "loss": 0.0129, "step": 30570 }, { "grad_norm": 0.14374326169490814, "learning_rate": 5.253698632010221e-05, "loss": 0.0117, "step": 30580 }, { "grad_norm": 0.11477640271186829, "learning_rate": 5.2509463602818246e-05, "loss": 0.0129, "step": 30590 }, { "grad_norm": 0.12171835452318192, "learning_rate": 5.248194012322549e-05, "loss": 0.0121, "step": 30600 }, { "grad_norm": 0.133353590965271, "learning_rate": 5.245441588968486e-05, "loss": 0.0138, "step": 30610 }, { "grad_norm": 0.15438875555992126, "learning_rate": 5.242689091055748e-05, "loss": 0.013, "step": 30620 }, { "grad_norm": 0.15410687029361725, "learning_rate": 5.239936519420473e-05, "loss": 0.0122, "step": 30630 }, { "grad_norm": 0.14394892752170563, "learning_rate": 5.2371838748988175e-05, "loss": 0.0118, "step": 30640 }, { "grad_norm": 0.13653340935707092, "learning_rate": 5.234431158326965e-05, "loss": 0.0113, "step": 30650 }, { "grad_norm": 0.11846642941236496, "learning_rate": 5.231678370541115e-05, "loss": 0.0126, "step": 30660 }, { "grad_norm": 0.09519076347351074, "learning_rate": 5.228925512377495e-05, "loss": 0.0123, "step": 30670 }, { "grad_norm": 0.10857737064361572, "learning_rate": 5.2261725846723465e-05, "loss": 0.0136, "step": 30680 }, { "grad_norm": 0.12200060486793518, "learning_rate": 5.22341958826194e-05, "loss": 0.0122, "step": 30690 }, { "grad_norm": 0.1185055747628212, "learning_rate": 5.22066652398256e-05, "loss": 0.0124, "step": 30700 }, { "grad_norm": 0.10046360641717911, "learning_rate": 5.2179133926705185e-05, "loss": 0.0126, "step": 30710 }, { "grad_norm": 0.11991281062364578, "learning_rate": 5.215160195162141e-05, "loss": 0.0118, "step": 30720 }, { "grad_norm": 0.13203392922878265, "learning_rate": 5.212406932293776e-05, "loss": 0.0142, "step": 30730 }, { "grad_norm": 0.12405151128768921, "learning_rate": 5.209653604901795e-05, "loss": 0.0122, "step": 30740 }, { "grad_norm": 0.12018381059169769, "learning_rate": 5.206900213822584e-05, "loss": 0.0128, "step": 30750 }, { "grad_norm": 0.14499947428703308, "learning_rate": 5.204146759892551e-05, "loss": 0.0126, "step": 30760 }, { "grad_norm": 0.09111843258142471, "learning_rate": 5.2013932439481216e-05, "loss": 0.0107, "step": 30770 }, { "grad_norm": 0.12213164567947388, "learning_rate": 5.198639666825743e-05, "loss": 0.0126, "step": 30780 }, { "grad_norm": 0.09845653176307678, "learning_rate": 5.195886029361877e-05, "loss": 0.0116, "step": 30790 }, { "grad_norm": 0.12384506314992905, "learning_rate": 5.193132332393009e-05, "loss": 0.012, "step": 30800 }, { "grad_norm": 0.12010736018419266, "learning_rate": 5.1903785767556376e-05, "loss": 0.0115, "step": 30810 }, { "grad_norm": 0.11920826882123947, "learning_rate": 5.187624763286282e-05, "loss": 0.0123, "step": 30820 }, { "grad_norm": 0.13283337652683258, "learning_rate": 5.184870892821475e-05, "loss": 0.0121, "step": 30830 }, { "grad_norm": 0.13848824799060822, "learning_rate": 5.182116966197773e-05, "loss": 0.0123, "step": 30840 }, { "grad_norm": 0.11084029078483582, "learning_rate": 5.1793629842517466e-05, "loss": 0.0112, "step": 30850 }, { "grad_norm": 0.12020634859800339, "learning_rate": 5.17660894781998e-05, "loss": 0.0115, "step": 30860 }, { "grad_norm": 0.12504246830940247, "learning_rate": 5.173854857739079e-05, "loss": 0.0111, "step": 30870 }, { "grad_norm": 0.09432642161846161, "learning_rate": 5.171100714845661e-05, "loss": 0.0119, "step": 30880 }, { "grad_norm": 0.09581892937421799, "learning_rate": 5.1683465199763646e-05, "loss": 0.0128, "step": 30890 }, { "grad_norm": 0.11631260812282562, "learning_rate": 5.16559227396784e-05, "loss": 0.0145, "step": 30900 }, { "grad_norm": 0.10708407312631607, "learning_rate": 5.1628379776567556e-05, "loss": 0.0109, "step": 30910 }, { "grad_norm": 0.09053546190261841, "learning_rate": 5.160083631879792e-05, "loss": 0.0107, "step": 30920 }, { "grad_norm": 0.11517383903265, "learning_rate": 5.1573292374736484e-05, "loss": 0.0114, "step": 30930 }, { "grad_norm": 0.1230253130197525, "learning_rate": 5.1545747952750356e-05, "loss": 0.013, "step": 30940 }, { "grad_norm": 0.10152684152126312, "learning_rate": 5.151820306120682e-05, "loss": 0.0132, "step": 30950 }, { "grad_norm": 0.13534167408943176, "learning_rate": 5.149065770847328e-05, "loss": 0.0125, "step": 30960 }, { "grad_norm": 0.12804102897644043, "learning_rate": 5.1463111902917297e-05, "loss": 0.0105, "step": 30970 }, { "grad_norm": 0.10177768021821976, "learning_rate": 5.143556565290654e-05, "loss": 0.0117, "step": 30980 }, { "grad_norm": 0.14341196417808533, "learning_rate": 5.140801896680882e-05, "loss": 0.0126, "step": 30990 }, { "grad_norm": 0.09558701515197754, "learning_rate": 5.1380471852992144e-05, "loss": 0.0112, "step": 31000 }, { "grad_norm": 0.1238846406340599, "learning_rate": 5.135292431982457e-05, "loss": 0.0132, "step": 31010 }, { "grad_norm": 0.12095954269170761, "learning_rate": 5.1325376375674294e-05, "loss": 0.0127, "step": 31020 }, { "grad_norm": 0.13635127246379852, "learning_rate": 5.129782802890968e-05, "loss": 0.0119, "step": 31030 }, { "grad_norm": 0.1322680115699768, "learning_rate": 5.127027928789916e-05, "loss": 0.0119, "step": 31040 }, { "grad_norm": 0.11070023477077484, "learning_rate": 5.124273016101135e-05, "loss": 0.0123, "step": 31050 }, { "grad_norm": 0.1374744176864624, "learning_rate": 5.121518065661492e-05, "loss": 0.0133, "step": 31060 }, { "grad_norm": 0.09450557827949524, "learning_rate": 5.11876307830787e-05, "loss": 0.0104, "step": 31070 }, { "grad_norm": 0.08550598472356796, "learning_rate": 5.1160080548771596e-05, "loss": 0.0137, "step": 31080 }, { "grad_norm": 0.1129220500588417, "learning_rate": 5.1132529962062656e-05, "loss": 0.0115, "step": 31090 }, { "grad_norm": 0.10707992315292358, "learning_rate": 5.110497903132101e-05, "loss": 0.0127, "step": 31100 }, { "grad_norm": 0.13823017477989197, "learning_rate": 5.107742776491592e-05, "loss": 0.012, "step": 31110 }, { "grad_norm": 0.12280257791280746, "learning_rate": 5.104987617121673e-05, "loss": 0.0133, "step": 31120 }, { "grad_norm": 0.14402829110622406, "learning_rate": 5.102232425859287e-05, "loss": 0.0119, "step": 31130 }, { "grad_norm": 0.11509326100349426, "learning_rate": 5.09947720354139e-05, "loss": 0.0119, "step": 31140 }, { "grad_norm": 0.09934159368276596, "learning_rate": 5.096721951004942e-05, "loss": 0.0116, "step": 31150 }, { "grad_norm": 0.11101450026035309, "learning_rate": 5.0939666690869227e-05, "loss": 0.0117, "step": 31160 }, { "grad_norm": 0.16573111712932587, "learning_rate": 5.0912113586243096e-05, "loss": 0.0121, "step": 31170 }, { "grad_norm": 0.12488952279090881, "learning_rate": 5.0884560204540935e-05, "loss": 0.013, "step": 31180 }, { "grad_norm": 0.12959297001361847, "learning_rate": 5.0857006554132736e-05, "loss": 0.0138, "step": 31190 }, { "grad_norm": 0.11642313003540039, "learning_rate": 5.0829452643388575e-05, "loss": 0.0119, "step": 31200 }, { "grad_norm": 0.12681515514850616, "learning_rate": 5.08018984806786e-05, "loss": 0.0107, "step": 31210 }, { "grad_norm": 0.13956435024738312, "learning_rate": 5.0774344074373036e-05, "loss": 0.0133, "step": 31220 }, { "grad_norm": 0.09502816200256348, "learning_rate": 5.07467894328422e-05, "loss": 0.0121, "step": 31230 }, { "grad_norm": 0.1303938329219818, "learning_rate": 5.0719234564456454e-05, "loss": 0.0118, "step": 31240 }, { "grad_norm": 0.14131437242031097, "learning_rate": 5.0691679477586216e-05, "loss": 0.0133, "step": 31250 }, { "grad_norm": 0.10502777993679047, "learning_rate": 5.0664124180602035e-05, "loss": 0.0113, "step": 31260 }, { "grad_norm": 0.13593809306621552, "learning_rate": 5.063656868187447e-05, "loss": 0.0122, "step": 31270 }, { "grad_norm": 0.1394820362329483, "learning_rate": 5.060901298977413e-05, "loss": 0.012, "step": 31280 }, { "grad_norm": 0.11158270388841629, "learning_rate": 5.0581457112671725e-05, "loss": 0.0131, "step": 31290 }, { "grad_norm": 0.1141449585556984, "learning_rate": 5.0553901058938016e-05, "loss": 0.0109, "step": 31300 }, { "grad_norm": 0.11511746793985367, "learning_rate": 5.052634483694377e-05, "loss": 0.0129, "step": 31310 }, { "grad_norm": 0.15860669314861298, "learning_rate": 5.049878845505988e-05, "loss": 0.0117, "step": 31320 }, { "grad_norm": 0.1108119785785675, "learning_rate": 5.047123192165721e-05, "loss": 0.0134, "step": 31330 }, { "grad_norm": 0.09586359560489655, "learning_rate": 5.0443675245106735e-05, "loss": 0.0135, "step": 31340 }, { "grad_norm": 0.10518722236156464, "learning_rate": 5.0416118433779426e-05, "loss": 0.0117, "step": 31350 }, { "grad_norm": 0.11846031248569489, "learning_rate": 5.038856149604633e-05, "loss": 0.0119, "step": 31360 }, { "grad_norm": 0.11810018122196198, "learning_rate": 5.03610044402785e-05, "loss": 0.0139, "step": 31370 }, { "grad_norm": 0.10267230123281479, "learning_rate": 5.033344727484707e-05, "loss": 0.0107, "step": 31380 }, { "grad_norm": 0.12032891064882278, "learning_rate": 5.030589000812315e-05, "loss": 0.0118, "step": 31390 }, { "grad_norm": 0.10718847811222076, "learning_rate": 5.027833264847793e-05, "loss": 0.0125, "step": 31400 }, { "grad_norm": 0.11073343455791473, "learning_rate": 5.025077520428258e-05, "loss": 0.0128, "step": 31410 }, { "grad_norm": 0.10412100702524185, "learning_rate": 5.022321768390837e-05, "loss": 0.0104, "step": 31420 }, { "grad_norm": 0.1255733072757721, "learning_rate": 5.0195660095726516e-05, "loss": 0.0124, "step": 31430 }, { "grad_norm": 0.1262773722410202, "learning_rate": 5.016810244810829e-05, "loss": 0.012, "step": 31440 }, { "grad_norm": 0.10708712786436081, "learning_rate": 5.0140544749424976e-05, "loss": 0.012, "step": 31450 }, { "grad_norm": 0.09246911853551865, "learning_rate": 5.0112987008047874e-05, "loss": 0.0115, "step": 31460 }, { "grad_norm": 0.13003093004226685, "learning_rate": 5.008542923234831e-05, "loss": 0.012, "step": 31470 }, { "grad_norm": 0.12146245688199997, "learning_rate": 5.00578714306976e-05, "loss": 0.0128, "step": 31480 }, { "grad_norm": 0.11913906782865524, "learning_rate": 5.0030313611467084e-05, "loss": 0.0115, "step": 31490 }, { "grad_norm": 0.08999604731798172, "learning_rate": 5.0002755783028074e-05, "loss": 0.0119, "step": 31500 }, { "grad_norm": 0.09834427386522293, "learning_rate": 4.997519795375194e-05, "loss": 0.0118, "step": 31510 }, { "grad_norm": 0.16506195068359375, "learning_rate": 4.9947640132010016e-05, "loss": 0.0119, "step": 31520 }, { "grad_norm": 0.12295001745223999, "learning_rate": 4.9920082326173625e-05, "loss": 0.0126, "step": 31530 }, { "grad_norm": 0.07874514162540436, "learning_rate": 4.9892524544614114e-05, "loss": 0.0109, "step": 31540 }, { "grad_norm": 0.15584039688110352, "learning_rate": 4.986496679570283e-05, "loss": 0.0126, "step": 31550 }, { "grad_norm": 0.08996797353029251, "learning_rate": 4.983740908781105e-05, "loss": 0.0115, "step": 31560 }, { "grad_norm": 0.1394093781709671, "learning_rate": 4.9809851429310116e-05, "loss": 0.0118, "step": 31570 }, { "grad_norm": 0.09455794841051102, "learning_rate": 4.9782293828571275e-05, "loss": 0.0117, "step": 31580 }, { "grad_norm": 0.10787267237901688, "learning_rate": 4.9754736293965846e-05, "loss": 0.0114, "step": 31590 }, { "grad_norm": 0.10659544169902802, "learning_rate": 4.972717883386502e-05, "loss": 0.0114, "step": 31600 }, { "grad_norm": 0.10915317386388779, "learning_rate": 4.9699621456640075e-05, "loss": 0.0122, "step": 31610 }, { "grad_norm": 0.16976141929626465, "learning_rate": 4.9672064170662214e-05, "loss": 0.0124, "step": 31620 }, { "grad_norm": 0.1092367172241211, "learning_rate": 4.9644506984302583e-05, "loss": 0.0116, "step": 31630 }, { "grad_norm": 0.14908914268016815, "learning_rate": 4.9616949905932356e-05, "loss": 0.0121, "step": 31640 }, { "grad_norm": 0.12731052935123444, "learning_rate": 4.9589392943922615e-05, "loss": 0.0122, "step": 31650 }, { "grad_norm": 0.1455388069152832, "learning_rate": 4.956183610664447e-05, "loss": 0.0116, "step": 31660 }, { "grad_norm": 0.11949457973241806, "learning_rate": 4.9534279402468945e-05, "loss": 0.0117, "step": 31670 }, { "grad_norm": 0.13987672328948975, "learning_rate": 4.9506722839767036e-05, "loss": 0.0124, "step": 31680 }, { "grad_norm": 0.11865437030792236, "learning_rate": 4.947916642690972e-05, "loss": 0.012, "step": 31690 }, { "grad_norm": 0.1359686553478241, "learning_rate": 4.9451610172267874e-05, "loss": 0.0118, "step": 31700 }, { "grad_norm": 0.1848181039094925, "learning_rate": 4.9424054084212376e-05, "loss": 0.0109, "step": 31710 }, { "grad_norm": 0.1002398356795311, "learning_rate": 4.939649817111407e-05, "loss": 0.0116, "step": 31720 }, { "grad_norm": 0.11657331883907318, "learning_rate": 4.936894244134365e-05, "loss": 0.0112, "step": 31730 }, { "grad_norm": 0.09356453269720078, "learning_rate": 4.9341386903271886e-05, "loss": 0.0135, "step": 31740 }, { "grad_norm": 0.12656940519809723, "learning_rate": 4.931383156526936e-05, "loss": 0.0124, "step": 31750 }, { "grad_norm": 0.11122498661279678, "learning_rate": 4.92862764357067e-05, "loss": 0.0108, "step": 31760 }, { "grad_norm": 0.09200171381235123, "learning_rate": 4.925872152295443e-05, "loss": 0.0116, "step": 31770 }, { "grad_norm": 0.10176485031843185, "learning_rate": 4.923116683538296e-05, "loss": 0.0133, "step": 31780 }, { "grad_norm": 0.10598864406347275, "learning_rate": 4.920361238136273e-05, "loss": 0.0132, "step": 31790 }, { "grad_norm": 0.12137220054864883, "learning_rate": 4.9176058169264014e-05, "loss": 0.0115, "step": 31800 }, { "grad_norm": 0.12195398658514023, "learning_rate": 4.9148504207457074e-05, "loss": 0.0115, "step": 31810 }, { "grad_norm": 0.12112552672624588, "learning_rate": 4.912095050431208e-05, "loss": 0.0123, "step": 31820 }, { "grad_norm": 0.09571657329797745, "learning_rate": 4.909339706819911e-05, "loss": 0.0123, "step": 31830 }, { "grad_norm": 0.10145220160484314, "learning_rate": 4.906584390748819e-05, "loss": 0.012, "step": 31840 }, { "grad_norm": 0.11323689669370651, "learning_rate": 4.9038291030549195e-05, "loss": 0.0129, "step": 31850 }, { "grad_norm": 0.17751698195934296, "learning_rate": 4.9010738445751995e-05, "loss": 0.0118, "step": 31860 }, { "grad_norm": 0.11836729943752289, "learning_rate": 4.8983186161466364e-05, "loss": 0.011, "step": 31870 }, { "grad_norm": 0.13710467517375946, "learning_rate": 4.89556341860619e-05, "loss": 0.0125, "step": 31880 }, { "grad_norm": 0.10589227825403214, "learning_rate": 4.892808252790822e-05, "loss": 0.0141, "step": 31890 }, { "grad_norm": 0.0941091775894165, "learning_rate": 4.890053119537475e-05, "loss": 0.0127, "step": 31900 }, { "grad_norm": 0.1201840415596962, "learning_rate": 4.887298019683087e-05, "loss": 0.0121, "step": 31910 }, { "grad_norm": 0.09576025605201721, "learning_rate": 4.884542954064587e-05, "loss": 0.0103, "step": 31920 }, { "grad_norm": 0.11272925138473511, "learning_rate": 4.881787923518887e-05, "loss": 0.0111, "step": 31930 }, { "grad_norm": 0.09577170759439468, "learning_rate": 4.879032928882896e-05, "loss": 0.0126, "step": 31940 }, { "grad_norm": 0.10079604387283325, "learning_rate": 4.876277970993505e-05, "loss": 0.0144, "step": 31950 }, { "grad_norm": 0.12186563014984131, "learning_rate": 4.873523050687602e-05, "loss": 0.0117, "step": 31960 }, { "grad_norm": 0.1578465849161148, "learning_rate": 4.870768168802056e-05, "loss": 0.0133, "step": 31970 }, { "grad_norm": 0.15036335587501526, "learning_rate": 4.868013326173728e-05, "loss": 0.0121, "step": 31980 }, { "grad_norm": 0.09163108468055725, "learning_rate": 4.865258523639468e-05, "loss": 0.0129, "step": 31990 }, { "grad_norm": 0.08312440663576126, "learning_rate": 4.862503762036109e-05, "loss": 0.011, "step": 32000 }, { "grad_norm": 0.09898731857538223, "learning_rate": 4.859749042200478e-05, "loss": 0.0119, "step": 32010 }, { "grad_norm": 0.1058415099978447, "learning_rate": 4.856994364969384e-05, "loss": 0.013, "step": 32020 }, { "grad_norm": 0.10688041150569916, "learning_rate": 4.854239731179625e-05, "loss": 0.0137, "step": 32030 }, { "grad_norm": 0.11521990597248077, "learning_rate": 4.85148514166799e-05, "loss": 0.0128, "step": 32040 }, { "grad_norm": 0.10729952901601791, "learning_rate": 4.8487305972712456e-05, "loss": 0.0128, "step": 32050 }, { "grad_norm": 0.10539475828409195, "learning_rate": 4.8459760988261526e-05, "loss": 0.0114, "step": 32060 }, { "grad_norm": 0.09328841418027878, "learning_rate": 4.843221647169453e-05, "loss": 0.0116, "step": 32070 }, { "grad_norm": 0.09977594763040543, "learning_rate": 4.840467243137878e-05, "loss": 0.012, "step": 32080 }, { "grad_norm": 0.12624965608119965, "learning_rate": 4.837712887568143e-05, "loss": 0.0123, "step": 32090 }, { "grad_norm": 0.15514419972896576, "learning_rate": 4.8349585812969464e-05, "loss": 0.0131, "step": 32100 }, { "grad_norm": 0.11996051669120789, "learning_rate": 4.8322043251609775e-05, "loss": 0.0131, "step": 32110 }, { "grad_norm": 0.10812821239233017, "learning_rate": 4.8294501199969015e-05, "loss": 0.012, "step": 32120 }, { "grad_norm": 0.12253760546445847, "learning_rate": 4.826695966641376e-05, "loss": 0.013, "step": 32130 }, { "grad_norm": 0.11848776042461395, "learning_rate": 4.823941865931043e-05, "loss": 0.0119, "step": 32140 }, { "grad_norm": 0.1372193843126297, "learning_rate": 4.82118781870252e-05, "loss": 0.0114, "step": 32150 }, { "grad_norm": 0.10556928813457489, "learning_rate": 4.8184338257924185e-05, "loss": 0.0131, "step": 32160 }, { "grad_norm": 0.10864689946174622, "learning_rate": 4.815679888037324e-05, "loss": 0.0113, "step": 32170 }, { "grad_norm": 0.1306663155555725, "learning_rate": 4.8129260062738135e-05, "loss": 0.0116, "step": 32180 }, { "grad_norm": 0.08893033117055893, "learning_rate": 4.810172181338445e-05, "loss": 0.0119, "step": 32190 }, { "grad_norm": 0.12388873100280762, "learning_rate": 4.807418414067753e-05, "loss": 0.0108, "step": 32200 }, { "grad_norm": 0.09643559902906418, "learning_rate": 4.804664705298264e-05, "loss": 0.0109, "step": 32210 }, { "grad_norm": 0.13503438234329224, "learning_rate": 4.80191105586648e-05, "loss": 0.0114, "step": 32220 }, { "grad_norm": 0.1426885575056076, "learning_rate": 4.799157466608886e-05, "loss": 0.0116, "step": 32230 }, { "grad_norm": 0.12082669883966446, "learning_rate": 4.796403938361951e-05, "loss": 0.0112, "step": 32240 }, { "grad_norm": 0.11513566225767136, "learning_rate": 4.793650471962123e-05, "loss": 0.011, "step": 32250 }, { "grad_norm": 0.11477836966514587, "learning_rate": 4.790897068245835e-05, "loss": 0.0112, "step": 32260 }, { "grad_norm": 0.12317775189876556, "learning_rate": 4.7881437280494954e-05, "loss": 0.0114, "step": 32270 }, { "grad_norm": 0.10206786543130875, "learning_rate": 4.7853904522094965e-05, "loss": 0.0116, "step": 32280 }, { "grad_norm": 0.1169847697019577, "learning_rate": 4.782637241562215e-05, "loss": 0.0119, "step": 32290 }, { "grad_norm": 0.10677891969680786, "learning_rate": 4.779884096943997e-05, "loss": 0.0122, "step": 32300 }, { "grad_norm": 0.12312310189008713, "learning_rate": 4.777131019191182e-05, "loss": 0.0131, "step": 32310 }, { "grad_norm": 0.1286705583333969, "learning_rate": 4.774378009140076e-05, "loss": 0.0122, "step": 32320 }, { "grad_norm": 0.11905037611722946, "learning_rate": 4.7716250676269735e-05, "loss": 0.0109, "step": 32330 }, { "grad_norm": 0.15195226669311523, "learning_rate": 4.7688721954881485e-05, "loss": 0.0128, "step": 32340 }, { "grad_norm": 0.11058328300714493, "learning_rate": 4.7661193935598446e-05, "loss": 0.0128, "step": 32350 }, { "grad_norm": 0.11913904547691345, "learning_rate": 4.763366662678296e-05, "loss": 0.0111, "step": 32360 }, { "grad_norm": 0.11439547687768936, "learning_rate": 4.7606140036797064e-05, "loss": 0.0104, "step": 32370 }, { "grad_norm": 0.0939631536602974, "learning_rate": 4.7578614174002614e-05, "loss": 0.0121, "step": 32380 }, { "grad_norm": 0.09576010704040527, "learning_rate": 4.755108904676125e-05, "loss": 0.0115, "step": 32390 }, { "grad_norm": 0.11212846636772156, "learning_rate": 4.752356466343436e-05, "loss": 0.0125, "step": 32400 }, { "grad_norm": 0.10289540141820908, "learning_rate": 4.7496041032383174e-05, "loss": 0.0116, "step": 32410 }, { "grad_norm": 0.14021141827106476, "learning_rate": 4.746851816196858e-05, "loss": 0.0156, "step": 32420 }, { "grad_norm": 0.12228015810251236, "learning_rate": 4.744099606055135e-05, "loss": 0.0128, "step": 32430 }, { "grad_norm": 0.12483000010251999, "learning_rate": 4.741347473649193e-05, "loss": 0.0119, "step": 32440 }, { "grad_norm": 0.13067686557769775, "learning_rate": 4.738595419815058e-05, "loss": 0.0117, "step": 32450 }, { "grad_norm": 0.11500085890293121, "learning_rate": 4.7358434453887365e-05, "loss": 0.0117, "step": 32460 }, { "grad_norm": 0.13211826980113983, "learning_rate": 4.7330915512061976e-05, "loss": 0.0117, "step": 32470 }, { "grad_norm": 0.11433693021535873, "learning_rate": 4.730339738103402e-05, "loss": 0.0114, "step": 32480 }, { "grad_norm": 0.12024365365505219, "learning_rate": 4.727588006916271e-05, "loss": 0.0108, "step": 32490 }, { "grad_norm": 0.12115886062383652, "learning_rate": 4.724836358480711e-05, "loss": 0.0125, "step": 32500 }, { "grad_norm": 0.13539890944957733, "learning_rate": 4.722084793632601e-05, "loss": 0.0116, "step": 32510 }, { "grad_norm": 0.11544118821620941, "learning_rate": 4.719333313207792e-05, "loss": 0.0125, "step": 32520 }, { "grad_norm": 0.09232006967067719, "learning_rate": 4.716581918042114e-05, "loss": 0.0117, "step": 32530 }, { "grad_norm": 0.11369549483060837, "learning_rate": 4.7138306089713636e-05, "loss": 0.0112, "step": 32540 }, { "grad_norm": 0.10811856389045715, "learning_rate": 4.7110793868313183e-05, "loss": 0.0114, "step": 32550 }, { "grad_norm": 0.11372952908277512, "learning_rate": 4.708328252457729e-05, "loss": 0.0136, "step": 32560 }, { "grad_norm": 0.1275101900100708, "learning_rate": 4.7055772066863135e-05, "loss": 0.0136, "step": 32570 }, { "grad_norm": 0.11376892030239105, "learning_rate": 4.702826250352771e-05, "loss": 0.0127, "step": 32580 }, { "grad_norm": 0.10979195684194565, "learning_rate": 4.7000753842927653e-05, "loss": 0.0108, "step": 32590 }, { "grad_norm": 0.10664445161819458, "learning_rate": 4.6973246093419384e-05, "loss": 0.0113, "step": 32600 }, { "grad_norm": 0.11735732108354568, "learning_rate": 4.694573926335906e-05, "loss": 0.0124, "step": 32610 }, { "grad_norm": 0.1031445637345314, "learning_rate": 4.6918233361102476e-05, "loss": 0.0111, "step": 32620 }, { "grad_norm": 0.12567274272441864, "learning_rate": 4.689072839500525e-05, "loss": 0.0116, "step": 32630 }, { "grad_norm": 0.09992280602455139, "learning_rate": 4.6863224373422635e-05, "loss": 0.0109, "step": 32640 }, { "grad_norm": 0.1256716549396515, "learning_rate": 4.683572130470962e-05, "loss": 0.0121, "step": 32650 }, { "grad_norm": 0.12216776609420776, "learning_rate": 4.680821919722094e-05, "loss": 0.0111, "step": 32660 }, { "grad_norm": 0.10822340846061707, "learning_rate": 4.6780718059310975e-05, "loss": 0.0116, "step": 32670 }, { "grad_norm": 0.14677147567272186, "learning_rate": 4.675321789933389e-05, "loss": 0.0124, "step": 32680 }, { "grad_norm": 0.10515834391117096, "learning_rate": 4.6725718725643464e-05, "loss": 0.0117, "step": 32690 }, { "grad_norm": 0.13803502917289734, "learning_rate": 4.669822054659323e-05, "loss": 0.0128, "step": 32700 }, { "grad_norm": 0.12334632128477097, "learning_rate": 4.667072337053644e-05, "loss": 0.0113, "step": 32710 }, { "grad_norm": 0.1255631148815155, "learning_rate": 4.6643227205825965e-05, "loss": 0.0103, "step": 32720 }, { "grad_norm": 0.09857525676488876, "learning_rate": 4.6615732060814454e-05, "loss": 0.0116, "step": 32730 }, { "grad_norm": 0.08333688974380493, "learning_rate": 4.658823794385417e-05, "loss": 0.0111, "step": 32740 }, { "grad_norm": 0.11410971730947495, "learning_rate": 4.6560744863297115e-05, "loss": 0.0114, "step": 32750 }, { "grad_norm": 0.10423851758241653, "learning_rate": 4.653325282749498e-05, "loss": 0.0115, "step": 32760 }, { "grad_norm": 0.10102195292711258, "learning_rate": 4.6505761844799075e-05, "loss": 0.0124, "step": 32770 }, { "grad_norm": 0.14929386973381042, "learning_rate": 4.647827192356048e-05, "loss": 0.0118, "step": 32780 }, { "grad_norm": 0.11191932111978531, "learning_rate": 4.645078307212989e-05, "loss": 0.0127, "step": 32790 }, { "grad_norm": 0.11909688264131546, "learning_rate": 4.642329529885768e-05, "loss": 0.0107, "step": 32800 }, { "grad_norm": 0.1071220263838768, "learning_rate": 4.639580861209393e-05, "loss": 0.0117, "step": 32810 }, { "grad_norm": 0.11752036958932877, "learning_rate": 4.636832302018835e-05, "loss": 0.0109, "step": 32820 }, { "grad_norm": 0.13046634197235107, "learning_rate": 4.6340838531490365e-05, "loss": 0.0097, "step": 32830 }, { "grad_norm": 0.11799152195453644, "learning_rate": 4.6313355154349e-05, "loss": 0.0119, "step": 32840 }, { "grad_norm": 0.11579578369855881, "learning_rate": 4.6285872897113025e-05, "loss": 0.0135, "step": 32850 }, { "grad_norm": 0.11241395771503448, "learning_rate": 4.625839176813077e-05, "loss": 0.0115, "step": 32860 }, { "grad_norm": 0.12119366228580475, "learning_rate": 4.623091177575031e-05, "loss": 0.0127, "step": 32870 }, { "grad_norm": 0.10671067982912064, "learning_rate": 4.620343292831936e-05, "loss": 0.0116, "step": 32880 }, { "grad_norm": 0.1208292618393898, "learning_rate": 4.6175955234185206e-05, "loss": 0.0122, "step": 32890 }, { "grad_norm": 0.13538141548633575, "learning_rate": 4.614847870169492e-05, "loss": 0.0115, "step": 32900 }, { "grad_norm": 0.13829636573791504, "learning_rate": 4.612100333919509e-05, "loss": 0.012, "step": 32910 }, { "grad_norm": 0.13333743810653687, "learning_rate": 4.609352915503202e-05, "loss": 0.0105, "step": 32920 }, { "grad_norm": 0.16404420137405396, "learning_rate": 4.606605615755166e-05, "loss": 0.0123, "step": 32930 }, { "grad_norm": 0.10462877154350281, "learning_rate": 4.6038584355099576e-05, "loss": 0.0109, "step": 32940 }, { "grad_norm": 0.12345992028713226, "learning_rate": 4.6011113756020964e-05, "loss": 0.0119, "step": 32950 }, { "grad_norm": 0.12230605632066727, "learning_rate": 4.598364436866066e-05, "loss": 0.0125, "step": 32960 }, { "grad_norm": 0.1253153681755066, "learning_rate": 4.595617620136316e-05, "loss": 0.0116, "step": 32970 }, { "grad_norm": 0.11153454333543777, "learning_rate": 4.592870926247257e-05, "loss": 0.011, "step": 32980 }, { "grad_norm": 0.09016207605600357, "learning_rate": 4.5901243560332594e-05, "loss": 0.0119, "step": 32990 }, { "grad_norm": 0.10933098942041397, "learning_rate": 4.587377910328662e-05, "loss": 0.0113, "step": 33000 }, { "grad_norm": 0.12482864409685135, "learning_rate": 4.5846315899677586e-05, "loss": 0.0113, "step": 33010 }, { "grad_norm": 0.10001932084560394, "learning_rate": 4.5818853957848114e-05, "loss": 0.0104, "step": 33020 }, { "grad_norm": 0.09943970292806625, "learning_rate": 4.579139328614043e-05, "loss": 0.0109, "step": 33030 }, { "grad_norm": 0.11617375910282135, "learning_rate": 4.576393389289633e-05, "loss": 0.0109, "step": 33040 }, { "grad_norm": 0.11423061788082123, "learning_rate": 4.573647578645728e-05, "loss": 0.0114, "step": 33050 }, { "grad_norm": 0.10537318140268326, "learning_rate": 4.57090189751643e-05, "loss": 0.0108, "step": 33060 }, { "grad_norm": 0.09740602225065231, "learning_rate": 4.568156346735806e-05, "loss": 0.0103, "step": 33070 }, { "grad_norm": 0.1683834195137024, "learning_rate": 4.565410927137882e-05, "loss": 0.0119, "step": 33080 }, { "grad_norm": 0.09070882946252823, "learning_rate": 4.562665639556644e-05, "loss": 0.0112, "step": 33090 }, { "grad_norm": 0.14982272684574127, "learning_rate": 4.559920484826037e-05, "loss": 0.0115, "step": 33100 }, { "grad_norm": 0.1001325398683548, "learning_rate": 4.5571754637799665e-05, "loss": 0.011, "step": 33110 }, { "grad_norm": 0.10254066437482834, "learning_rate": 4.554430577252298e-05, "loss": 0.0097, "step": 33120 }, { "grad_norm": 0.14519137144088745, "learning_rate": 4.551685826076858e-05, "loss": 0.0111, "step": 33130 }, { "grad_norm": 0.09510723501443863, "learning_rate": 4.5489412110874246e-05, "loss": 0.0098, "step": 33140 }, { "grad_norm": 0.11343372613191605, "learning_rate": 4.5461967331177444e-05, "loss": 0.0126, "step": 33150 }, { "grad_norm": 0.10185418277978897, "learning_rate": 4.5434523930015115e-05, "loss": 0.0104, "step": 33160 }, { "grad_norm": 0.13460800051689148, "learning_rate": 4.540708191572388e-05, "loss": 0.0123, "step": 33170 }, { "grad_norm": 0.10012591630220413, "learning_rate": 4.537964129663991e-05, "loss": 0.0112, "step": 33180 }, { "grad_norm": 0.10670509934425354, "learning_rate": 4.535220208109889e-05, "loss": 0.0117, "step": 33190 }, { "grad_norm": 0.0969107523560524, "learning_rate": 4.5324764277436194e-05, "loss": 0.0115, "step": 33200 }, { "grad_norm": 0.10241895169019699, "learning_rate": 4.529732789398664e-05, "loss": 0.0123, "step": 33210 }, { "grad_norm": 0.15347357094287872, "learning_rate": 4.526989293908472e-05, "loss": 0.0119, "step": 33220 }, { "grad_norm": 0.1411547213792801, "learning_rate": 4.524245942106442e-05, "loss": 0.0119, "step": 33230 }, { "grad_norm": 0.10168654471635818, "learning_rate": 4.5215027348259345e-05, "loss": 0.0127, "step": 33240 }, { "grad_norm": 0.11630523949861526, "learning_rate": 4.5187596729002616e-05, "loss": 0.0121, "step": 33250 }, { "grad_norm": 0.1266487091779709, "learning_rate": 4.516016757162693e-05, "loss": 0.0117, "step": 33260 }, { "grad_norm": 0.08914290368556976, "learning_rate": 4.513273988446457e-05, "loss": 0.0111, "step": 33270 }, { "grad_norm": 0.10967351496219635, "learning_rate": 4.5105313675847296e-05, "loss": 0.0104, "step": 33280 }, { "grad_norm": 0.10494846850633621, "learning_rate": 4.5077888954106495e-05, "loss": 0.0139, "step": 33290 }, { "grad_norm": 0.21603912115097046, "learning_rate": 4.505046572757309e-05, "loss": 0.0116, "step": 33300 }, { "grad_norm": 0.1185331642627716, "learning_rate": 4.502304400457749e-05, "loss": 0.0123, "step": 33310 }, { "grad_norm": 0.12002453207969666, "learning_rate": 4.499562379344973e-05, "loss": 0.0115, "step": 33320 }, { "grad_norm": 0.10384371876716614, "learning_rate": 4.4968205102519306e-05, "loss": 0.0114, "step": 33330 }, { "grad_norm": 0.10492204129695892, "learning_rate": 4.494078794011532e-05, "loss": 0.0113, "step": 33340 }, { "grad_norm": 0.11265677213668823, "learning_rate": 4.491337231456639e-05, "loss": 0.0122, "step": 33350 }, { "grad_norm": 0.1286594569683075, "learning_rate": 4.4885958234200634e-05, "loss": 0.012, "step": 33360 }, { "grad_norm": 0.13986752927303314, "learning_rate": 4.485854570734575e-05, "loss": 0.0108, "step": 33370 }, { "grad_norm": 0.11216746270656586, "learning_rate": 4.483113474232891e-05, "loss": 0.0106, "step": 33380 }, { "grad_norm": 0.12012241780757904, "learning_rate": 4.480372534747688e-05, "loss": 0.0102, "step": 33390 }, { "grad_norm": 0.13043545186519623, "learning_rate": 4.477631753111588e-05, "loss": 0.0105, "step": 33400 }, { "grad_norm": 0.0913296490907669, "learning_rate": 4.4748911301571686e-05, "loss": 0.0121, "step": 33410 }, { "grad_norm": 0.11800259351730347, "learning_rate": 4.472150666716961e-05, "loss": 0.0115, "step": 33420 }, { "grad_norm": 0.11870235949754715, "learning_rate": 4.469410363623442e-05, "loss": 0.0118, "step": 33430 }, { "grad_norm": 0.12500214576721191, "learning_rate": 4.466670221709044e-05, "loss": 0.013, "step": 33440 }, { "grad_norm": 0.08117754012346268, "learning_rate": 4.463930241806154e-05, "loss": 0.0114, "step": 33450 }, { "grad_norm": 0.0964612290263176, "learning_rate": 4.4611904247471006e-05, "loss": 0.0122, "step": 33460 }, { "grad_norm": 0.09400574117898941, "learning_rate": 4.458450771364171e-05, "loss": 0.0105, "step": 33470 }, { "grad_norm": 0.08617919683456421, "learning_rate": 4.4557112824895965e-05, "loss": 0.012, "step": 33480 }, { "grad_norm": 0.09143482148647308, "learning_rate": 4.452971958955563e-05, "loss": 0.012, "step": 33490 }, { "grad_norm": 0.09284518659114838, "learning_rate": 4.450232801594208e-05, "loss": 0.0113, "step": 33500 }, { "grad_norm": 0.10502566397190094, "learning_rate": 4.447493811237609e-05, "loss": 0.0115, "step": 33510 }, { "grad_norm": 0.0910014659166336, "learning_rate": 4.444754988717804e-05, "loss": 0.0109, "step": 33520 }, { "grad_norm": 0.1088210940361023, "learning_rate": 4.442016334866771e-05, "loss": 0.0117, "step": 33530 }, { "grad_norm": 0.11908374726772308, "learning_rate": 4.4392778505164445e-05, "loss": 0.0109, "step": 33540 }, { "grad_norm": 0.10808266699314117, "learning_rate": 4.436539536498702e-05, "loss": 0.0119, "step": 33550 }, { "grad_norm": 0.11792311817407608, "learning_rate": 4.433801393645369e-05, "loss": 0.0107, "step": 33560 }, { "grad_norm": 0.13141676783561707, "learning_rate": 4.431063422788226e-05, "loss": 0.0107, "step": 33570 }, { "grad_norm": 0.1082032322883606, "learning_rate": 4.428325624758991e-05, "loss": 0.0118, "step": 33580 }, { "grad_norm": 0.10906095802783966, "learning_rate": 4.4255880003893366e-05, "loss": 0.0104, "step": 33590 }, { "grad_norm": 0.12103837728500366, "learning_rate": 4.422850550510884e-05, "loss": 0.0115, "step": 33600 }, { "grad_norm": 0.09104857593774796, "learning_rate": 4.4201132759551934e-05, "loss": 0.0103, "step": 33610 }, { "grad_norm": 0.10324760526418686, "learning_rate": 4.4173761775537804e-05, "loss": 0.0124, "step": 33620 }, { "grad_norm": 0.14390477538108826, "learning_rate": 4.414639256138099e-05, "loss": 0.0131, "step": 33630 }, { "grad_norm": 0.13126368820667267, "learning_rate": 4.411902512539557e-05, "loss": 0.0119, "step": 33640 }, { "grad_norm": 0.10588022321462631, "learning_rate": 4.4091659475895044e-05, "loss": 0.0129, "step": 33650 }, { "grad_norm": 0.0886785238981247, "learning_rate": 4.406429562119235e-05, "loss": 0.0118, "step": 33660 }, { "grad_norm": 0.12034635990858078, "learning_rate": 4.4036933569599945e-05, "loss": 0.011, "step": 33670 }, { "grad_norm": 0.09720934927463531, "learning_rate": 4.400957332942965e-05, "loss": 0.0105, "step": 33680 }, { "grad_norm": 0.11989627033472061, "learning_rate": 4.3982214908992844e-05, "loss": 0.0115, "step": 33690 }, { "grad_norm": 0.10999913513660431, "learning_rate": 4.3954858316600235e-05, "loss": 0.0135, "step": 33700 }, { "grad_norm": 0.09063597023487091, "learning_rate": 4.392750356056205e-05, "loss": 0.0103, "step": 33710 }, { "grad_norm": 0.11201436817646027, "learning_rate": 4.390015064918798e-05, "loss": 0.0114, "step": 33720 }, { "grad_norm": 0.11897926032543182, "learning_rate": 4.387279959078705e-05, "loss": 0.011, "step": 33730 }, { "grad_norm": 0.12464333325624466, "learning_rate": 4.384545039366786e-05, "loss": 0.0116, "step": 33740 }, { "grad_norm": 0.11496912688016891, "learning_rate": 4.381810306613831e-05, "loss": 0.0115, "step": 33750 }, { "grad_norm": 0.11356748640537262, "learning_rate": 4.3790757616505826e-05, "loss": 0.0113, "step": 33760 }, { "grad_norm": 0.11700179427862167, "learning_rate": 4.376341405307725e-05, "loss": 0.0104, "step": 33770 }, { "grad_norm": 0.1020607203245163, "learning_rate": 4.37360723841588e-05, "loss": 0.012, "step": 33780 }, { "grad_norm": 0.1260097324848175, "learning_rate": 4.370873261805619e-05, "loss": 0.0113, "step": 33790 }, { "grad_norm": 0.11335933208465576, "learning_rate": 4.368139476307449e-05, "loss": 0.0111, "step": 33800 }, { "grad_norm": 0.09305644780397415, "learning_rate": 4.365405882751822e-05, "loss": 0.0109, "step": 33810 }, { "grad_norm": 0.13559772074222565, "learning_rate": 4.3626724819691326e-05, "loss": 0.0108, "step": 33820 }, { "grad_norm": 0.12181299924850464, "learning_rate": 4.359939274789715e-05, "loss": 0.0119, "step": 33830 }, { "grad_norm": 0.12000253796577454, "learning_rate": 4.357206262043848e-05, "loss": 0.0115, "step": 33840 }, { "grad_norm": 0.1147080734372139, "learning_rate": 4.354473444561745e-05, "loss": 0.0102, "step": 33850 }, { "grad_norm": 0.1279354840517044, "learning_rate": 4.3517408231735644e-05, "loss": 0.0114, "step": 33860 }, { "grad_norm": 0.10259544849395752, "learning_rate": 4.3490083987094086e-05, "loss": 0.013, "step": 33870 }, { "grad_norm": 0.09961477667093277, "learning_rate": 4.34627617199931e-05, "loss": 0.0106, "step": 33880 }, { "grad_norm": 0.10700345784425735, "learning_rate": 4.3435441438732526e-05, "loss": 0.0122, "step": 33890 }, { "grad_norm": 0.13737943768501282, "learning_rate": 4.340812315161149e-05, "loss": 0.0114, "step": 33900 }, { "grad_norm": 0.0906800776720047, "learning_rate": 4.338080686692859e-05, "loss": 0.0105, "step": 33910 }, { "grad_norm": 0.12245111167430878, "learning_rate": 4.3353492592981816e-05, "loss": 0.0099, "step": 33920 }, { "grad_norm": 0.09660937637090683, "learning_rate": 4.3326180338068485e-05, "loss": 0.0102, "step": 33930 }, { "grad_norm": 0.13750088214874268, "learning_rate": 4.3298870110485356e-05, "loss": 0.0128, "step": 33940 }, { "grad_norm": 0.21700477600097656, "learning_rate": 4.3271561918528567e-05, "loss": 0.0118, "step": 33950 }, { "grad_norm": 0.1425265073776245, "learning_rate": 4.324425577049359e-05, "loss": 0.0116, "step": 33960 }, { "grad_norm": 0.13416333496570587, "learning_rate": 4.321695167467535e-05, "loss": 0.0098, "step": 33970 }, { "grad_norm": 0.09105820953845978, "learning_rate": 4.3189649639368093e-05, "loss": 0.012, "step": 33980 }, { "grad_norm": 0.0972747951745987, "learning_rate": 4.316234967286547e-05, "loss": 0.0104, "step": 33990 }, { "grad_norm": 0.0926378145813942, "learning_rate": 4.313505178346046e-05, "loss": 0.0099, "step": 34000 }, { "grad_norm": 0.139136403799057, "learning_rate": 4.3107755979445465e-05, "loss": 0.0105, "step": 34010 }, { "grad_norm": 0.09003983438014984, "learning_rate": 4.308046226911224e-05, "loss": 0.0103, "step": 34020 }, { "grad_norm": 0.1816934198141098, "learning_rate": 4.305317066075185e-05, "loss": 0.0117, "step": 34030 }, { "grad_norm": 0.1212204322218895, "learning_rate": 4.302588116265482e-05, "loss": 0.011, "step": 34040 }, { "grad_norm": 0.12296416610479355, "learning_rate": 4.299859378311094e-05, "loss": 0.0122, "step": 34050 }, { "grad_norm": 0.14175789058208466, "learning_rate": 4.2971308530409424e-05, "loss": 0.0113, "step": 34060 }, { "grad_norm": 0.09126793593168259, "learning_rate": 4.2944025412838765e-05, "loss": 0.0111, "step": 34070 }, { "grad_norm": 0.1222928911447525, "learning_rate": 4.291674443868689e-05, "loss": 0.0098, "step": 34080 }, { "grad_norm": 0.12249375134706497, "learning_rate": 4.288946561624104e-05, "loss": 0.0124, "step": 34090 }, { "grad_norm": 0.11950691044330597, "learning_rate": 4.2862188953787794e-05, "loss": 0.0107, "step": 34100 }, { "grad_norm": 0.08818893879652023, "learning_rate": 4.283491445961308e-05, "loss": 0.0112, "step": 34110 }, { "grad_norm": 0.0950581282377243, "learning_rate": 4.2807642142002155e-05, "loss": 0.0106, "step": 34120 }, { "grad_norm": 0.12265228480100632, "learning_rate": 4.278037200923966e-05, "loss": 0.0108, "step": 34130 }, { "grad_norm": 0.15368865430355072, "learning_rate": 4.275310406960953e-05, "loss": 0.0131, "step": 34140 }, { "grad_norm": 0.12743782997131348, "learning_rate": 4.272583833139502e-05, "loss": 0.0117, "step": 34150 }, { "grad_norm": 0.1184421107172966, "learning_rate": 4.2698574802878794e-05, "loss": 0.0113, "step": 34160 }, { "grad_norm": 0.11769257485866547, "learning_rate": 4.2671313492342734e-05, "loss": 0.0103, "step": 34170 }, { "grad_norm": 0.1629980206489563, "learning_rate": 4.264405440806813e-05, "loss": 0.0107, "step": 34180 }, { "grad_norm": 0.1134582981467247, "learning_rate": 4.26167975583356e-05, "loss": 0.011, "step": 34190 }, { "grad_norm": 0.1312858760356903, "learning_rate": 4.2589542951425e-05, "loss": 0.0133, "step": 34200 }, { "grad_norm": 0.08281660079956055, "learning_rate": 4.2562290595615615e-05, "loss": 0.0112, "step": 34210 }, { "grad_norm": 0.12150486558675766, "learning_rate": 4.2535040499185946e-05, "loss": 0.0111, "step": 34220 }, { "grad_norm": 0.11422853916883469, "learning_rate": 4.250779267041387e-05, "loss": 0.0116, "step": 34230 }, { "grad_norm": 0.13424287736415863, "learning_rate": 4.248054711757657e-05, "loss": 0.0112, "step": 34240 }, { "grad_norm": 0.11963766068220139, "learning_rate": 4.245330384895052e-05, "loss": 0.011, "step": 34250 }, { "grad_norm": 0.13532735407352448, "learning_rate": 4.242606287281151e-05, "loss": 0.0116, "step": 34260 }, { "grad_norm": 0.09888887405395508, "learning_rate": 4.2398824197434595e-05, "loss": 0.0107, "step": 34270 }, { "grad_norm": 0.1254492700099945, "learning_rate": 4.23715878310942e-05, "loss": 0.0117, "step": 34280 }, { "grad_norm": 0.11804278194904327, "learning_rate": 4.234435378206402e-05, "loss": 0.0108, "step": 34290 }, { "grad_norm": 0.12836091220378876, "learning_rate": 4.2317122058617006e-05, "loss": 0.0117, "step": 34300 }, { "grad_norm": 0.12835296988487244, "learning_rate": 4.2289892669025485e-05, "loss": 0.0133, "step": 34310 }, { "grad_norm": 0.10684549063444138, "learning_rate": 4.226266562156097e-05, "loss": 0.0121, "step": 34320 }, { "grad_norm": 0.13176073133945465, "learning_rate": 4.223544092449435e-05, "loss": 0.0128, "step": 34330 }, { "grad_norm": 0.11782565712928772, "learning_rate": 4.2208218586095784e-05, "loss": 0.0127, "step": 34340 }, { "grad_norm": 0.12096721678972244, "learning_rate": 4.218099861463466e-05, "loss": 0.013, "step": 34350 }, { "grad_norm": 0.11754336953163147, "learning_rate": 4.215378101837972e-05, "loss": 0.0113, "step": 34360 }, { "grad_norm": 0.13538914918899536, "learning_rate": 4.2126565805598937e-05, "loss": 0.011, "step": 34370 }, { "grad_norm": 0.1523521989583969, "learning_rate": 4.209935298455957e-05, "loss": 0.0112, "step": 34380 }, { "grad_norm": 0.12177921831607819, "learning_rate": 4.207214256352817e-05, "loss": 0.0105, "step": 34390 }, { "grad_norm": 0.10860244929790497, "learning_rate": 4.2044934550770524e-05, "loss": 0.0116, "step": 34400 }, { "grad_norm": 0.10994383692741394, "learning_rate": 4.201772895455174e-05, "loss": 0.0116, "step": 34410 }, { "grad_norm": 0.1149163693189621, "learning_rate": 4.199052578313613e-05, "loss": 0.0105, "step": 34420 }, { "grad_norm": 0.13153187930583954, "learning_rate": 4.1963325044787294e-05, "loss": 0.01, "step": 34430 }, { "grad_norm": 0.1103607565164566, "learning_rate": 4.193612674776814e-05, "loss": 0.0111, "step": 34440 }, { "grad_norm": 0.10553408414125443, "learning_rate": 4.1908930900340745e-05, "loss": 0.011, "step": 34450 }, { "grad_norm": 0.1027817577123642, "learning_rate": 4.1881737510766536e-05, "loss": 0.0113, "step": 34460 }, { "grad_norm": 0.19077734649181366, "learning_rate": 4.185454658730609e-05, "loss": 0.0105, "step": 34470 }, { "grad_norm": 0.11502246558666229, "learning_rate": 4.1827358138219355e-05, "loss": 0.0116, "step": 34480 }, { "grad_norm": 0.10506708920001984, "learning_rate": 4.1800172171765404e-05, "loss": 0.0105, "step": 34490 }, { "grad_norm": 0.10233186930418015, "learning_rate": 4.177298869620264e-05, "loss": 0.0119, "step": 34500 }, { "grad_norm": 0.1224413812160492, "learning_rate": 4.1745807719788705e-05, "loss": 0.0108, "step": 34510 }, { "grad_norm": 0.10127171874046326, "learning_rate": 4.1718629250780445e-05, "loss": 0.0101, "step": 34520 }, { "grad_norm": 0.12019231170415878, "learning_rate": 4.1691453297433956e-05, "loss": 0.0117, "step": 34530 }, { "grad_norm": 0.10949630290269852, "learning_rate": 4.166427986800457e-05, "loss": 0.0126, "step": 34540 }, { "grad_norm": 0.10138779133558273, "learning_rate": 4.163710897074688e-05, "loss": 0.0125, "step": 34550 }, { "grad_norm": 0.1575637310743332, "learning_rate": 4.1609940613914686e-05, "loss": 0.0118, "step": 34560 }, { "grad_norm": 0.11811704188585281, "learning_rate": 4.1582774805760996e-05, "loss": 0.0106, "step": 34570 }, { "grad_norm": 0.10277058929204941, "learning_rate": 4.155561155453809e-05, "loss": 0.0108, "step": 34580 }, { "grad_norm": 0.1580645889043808, "learning_rate": 4.15284508684974e-05, "loss": 0.013, "step": 34590 }, { "grad_norm": 0.10618416965007782, "learning_rate": 4.1501292755889675e-05, "loss": 0.012, "step": 34600 }, { "grad_norm": 0.1274794489145279, "learning_rate": 4.1474137224964833e-05, "loss": 0.0111, "step": 34610 }, { "grad_norm": 0.10976561903953552, "learning_rate": 4.144698428397197e-05, "loss": 0.0108, "step": 34620 }, { "grad_norm": 0.13406285643577576, "learning_rate": 4.1419833941159466e-05, "loss": 0.0104, "step": 34630 }, { "grad_norm": 0.12519139051437378, "learning_rate": 4.1392686204774846e-05, "loss": 0.011, "step": 34640 }, { "grad_norm": 0.10369113832712173, "learning_rate": 4.13655410830649e-05, "loss": 0.0117, "step": 34650 }, { "grad_norm": 0.12038713693618774, "learning_rate": 4.1338398584275594e-05, "loss": 0.0102, "step": 34660 }, { "grad_norm": 0.11796201765537262, "learning_rate": 4.1311258716652104e-05, "loss": 0.0104, "step": 34670 }, { "grad_norm": 0.11831456422805786, "learning_rate": 4.128412148843881e-05, "loss": 0.0107, "step": 34680 }, { "grad_norm": 0.12215907126665115, "learning_rate": 4.125698690787926e-05, "loss": 0.0103, "step": 34690 }, { "grad_norm": 0.11347462236881256, "learning_rate": 4.1229854983216245e-05, "loss": 0.0107, "step": 34700 }, { "grad_norm": 0.09278062731027603, "learning_rate": 4.120272572269175e-05, "loss": 0.0105, "step": 34710 }, { "grad_norm": 0.08483899384737015, "learning_rate": 4.117559913454687e-05, "loss": 0.0104, "step": 34720 }, { "grad_norm": 0.10284934937953949, "learning_rate": 4.114847522702201e-05, "loss": 0.0112, "step": 34730 }, { "grad_norm": 0.13920283317565918, "learning_rate": 4.112135400835664e-05, "loss": 0.0119, "step": 34740 }, { "grad_norm": 0.10815886408090591, "learning_rate": 4.109423548678949e-05, "loss": 0.0123, "step": 34750 }, { "grad_norm": 0.11545385420322418, "learning_rate": 4.106711967055848e-05, "loss": 0.0126, "step": 34760 }, { "grad_norm": 0.09429657459259033, "learning_rate": 4.1040006567900636e-05, "loss": 0.0113, "step": 34770 }, { "grad_norm": 0.11421538144350052, "learning_rate": 4.101289618705224e-05, "loss": 0.0112, "step": 34780 }, { "grad_norm": 0.09696895629167557, "learning_rate": 4.0985788536248675e-05, "loss": 0.0109, "step": 34790 }, { "grad_norm": 0.1507321000099182, "learning_rate": 4.095868362372454e-05, "loss": 0.0111, "step": 34800 }, { "grad_norm": 0.09934264421463013, "learning_rate": 4.0931581457713614e-05, "loss": 0.0123, "step": 34810 }, { "grad_norm": 0.12254150211811066, "learning_rate": 4.09044820464488e-05, "loss": 0.0107, "step": 34820 }, { "grad_norm": 0.14121417701244354, "learning_rate": 4.087738539816219e-05, "loss": 0.0118, "step": 34830 }, { "grad_norm": 0.10405011475086212, "learning_rate": 4.085029152108501e-05, "loss": 0.0104, "step": 34840 }, { "grad_norm": 0.10669048130512238, "learning_rate": 4.0823200423447714e-05, "loss": 0.0113, "step": 34850 }, { "grad_norm": 0.08707300573587418, "learning_rate": 4.079611211347981e-05, "loss": 0.011, "step": 34860 }, { "grad_norm": 0.11757978796958923, "learning_rate": 4.076902659941002e-05, "loss": 0.0093, "step": 34870 }, { "grad_norm": 0.07832923531532288, "learning_rate": 4.074194388946624e-05, "loss": 0.0104, "step": 34880 }, { "grad_norm": 0.11118625849485397, "learning_rate": 4.071486399187545e-05, "loss": 0.0098, "step": 34890 }, { "grad_norm": 0.12358112633228302, "learning_rate": 4.0687786914863836e-05, "loss": 0.0101, "step": 34900 }, { "grad_norm": 0.1351807713508606, "learning_rate": 4.0660712666656666e-05, "loss": 0.0104, "step": 34910 }, { "grad_norm": 0.10393640398979187, "learning_rate": 4.0633641255478394e-05, "loss": 0.0114, "step": 34920 }, { "grad_norm": 0.10244626551866531, "learning_rate": 4.0606572689552624e-05, "loss": 0.0113, "step": 34930 }, { "grad_norm": 0.1120007112622261, "learning_rate": 4.0579506977102036e-05, "loss": 0.0116, "step": 34940 }, { "grad_norm": 0.08981763571500778, "learning_rate": 4.055244412634849e-05, "loss": 0.0111, "step": 34950 }, { "grad_norm": 0.09719004482030869, "learning_rate": 4.052538414551298e-05, "loss": 0.0115, "step": 34960 }, { "grad_norm": 0.09248580783605576, "learning_rate": 4.0498327042815596e-05, "loss": 0.0118, "step": 34970 }, { "grad_norm": 0.10686587542295456, "learning_rate": 4.047127282647559e-05, "loss": 0.0099, "step": 34980 }, { "grad_norm": 0.10007361322641373, "learning_rate": 4.04442215047113e-05, "loss": 0.0114, "step": 34990 }, { "grad_norm": 0.12641745805740356, "learning_rate": 4.041717308574023e-05, "loss": 0.0121, "step": 35000 }, { "grad_norm": 0.1404709815979004, "learning_rate": 4.039012757777893e-05, "loss": 0.0106, "step": 35010 }, { "grad_norm": 0.12470653653144836, "learning_rate": 4.036308498904314e-05, "loss": 0.0106, "step": 35020 }, { "grad_norm": 0.1022089421749115, "learning_rate": 4.033604532774771e-05, "loss": 0.0115, "step": 35030 }, { "grad_norm": 0.09750610589981079, "learning_rate": 4.030900860210652e-05, "loss": 0.0102, "step": 35040 }, { "grad_norm": 0.18184396624565125, "learning_rate": 4.028197482033266e-05, "loss": 0.0116, "step": 35050 }, { "grad_norm": 0.09313593059778214, "learning_rate": 4.0254943990638246e-05, "loss": 0.0102, "step": 35060 }, { "grad_norm": 0.09923060238361359, "learning_rate": 4.022791612123454e-05, "loss": 0.0109, "step": 35070 }, { "grad_norm": 0.10669462382793427, "learning_rate": 4.020089122033192e-05, "loss": 0.0109, "step": 35080 }, { "grad_norm": 0.1067902222275734, "learning_rate": 4.01738692961398e-05, "loss": 0.0108, "step": 35090 }, { "grad_norm": 0.10739213228225708, "learning_rate": 4.014685035686675e-05, "loss": 0.0113, "step": 35100 }, { "grad_norm": 0.10472454875707626, "learning_rate": 4.011983441072039e-05, "loss": 0.0107, "step": 35110 }, { "grad_norm": 0.13603855669498444, "learning_rate": 4.0092821465907485e-05, "loss": 0.0103, "step": 35120 }, { "grad_norm": 0.09206308424472809, "learning_rate": 4.006581153063383e-05, "loss": 0.0097, "step": 35130 }, { "grad_norm": 0.14239998161792755, "learning_rate": 4.003880461310432e-05, "loss": 0.0093, "step": 35140 }, { "grad_norm": 0.12960045039653778, "learning_rate": 4.001180072152298e-05, "loss": 0.0102, "step": 35150 }, { "grad_norm": 0.09797478467226028, "learning_rate": 3.998479986409285e-05, "loss": 0.0115, "step": 35160 }, { "grad_norm": 0.11362149566411972, "learning_rate": 3.995780204901607e-05, "loss": 0.009, "step": 35170 }, { "grad_norm": 0.15104220807552338, "learning_rate": 3.993080728449391e-05, "loss": 0.0097, "step": 35180 }, { "grad_norm": 0.11350920796394348, "learning_rate": 3.990381557872661e-05, "loss": 0.0097, "step": 35190 }, { "grad_norm": 0.11736364662647247, "learning_rate": 3.987682693991359e-05, "loss": 0.0106, "step": 35200 }, { "grad_norm": 0.14063936471939087, "learning_rate": 3.9849841376253226e-05, "loss": 0.0107, "step": 35210 }, { "grad_norm": 0.11033650487661362, "learning_rate": 3.982285889594306e-05, "loss": 0.011, "step": 35220 }, { "grad_norm": 0.10613057762384415, "learning_rate": 3.9795879507179665e-05, "loss": 0.0113, "step": 35230 }, { "grad_norm": 0.10741548240184784, "learning_rate": 3.9768903218158634e-05, "loss": 0.0111, "step": 35240 }, { "grad_norm": 0.10991934686899185, "learning_rate": 3.974193003707468e-05, "loss": 0.0097, "step": 35250 }, { "grad_norm": 0.11416126787662506, "learning_rate": 3.971495997212152e-05, "loss": 0.01, "step": 35260 }, { "grad_norm": 0.11133500188589096, "learning_rate": 3.9687993031491985e-05, "loss": 0.0101, "step": 35270 }, { "grad_norm": 0.13363096117973328, "learning_rate": 3.966102922337787e-05, "loss": 0.0121, "step": 35280 }, { "grad_norm": 0.12417926639318466, "learning_rate": 3.963406855597009e-05, "loss": 0.0112, "step": 35290 }, { "grad_norm": 0.1121039092540741, "learning_rate": 3.960711103745861e-05, "loss": 0.0113, "step": 35300 }, { "grad_norm": 0.09430599957704544, "learning_rate": 3.958015667603237e-05, "loss": 0.0098, "step": 35310 }, { "grad_norm": 0.14074674248695374, "learning_rate": 3.955320547987943e-05, "loss": 0.0113, "step": 35320 }, { "grad_norm": 0.10180644690990448, "learning_rate": 3.952625745718681e-05, "loss": 0.0112, "step": 35330 }, { "grad_norm": 0.09728924930095673, "learning_rate": 3.949931261614064e-05, "loss": 0.011, "step": 35340 }, { "grad_norm": 0.1186748594045639, "learning_rate": 3.947237096492605e-05, "loss": 0.0109, "step": 35350 }, { "grad_norm": 0.13336898386478424, "learning_rate": 3.944543251172719e-05, "loss": 0.0111, "step": 35360 }, { "grad_norm": 0.13283567130565643, "learning_rate": 3.941849726472725e-05, "loss": 0.0119, "step": 35370 }, { "grad_norm": 0.12204480171203613, "learning_rate": 3.939156523210846e-05, "loss": 0.0109, "step": 35380 }, { "grad_norm": 0.08672457188367844, "learning_rate": 3.9364636422052046e-05, "loss": 0.0099, "step": 35390 }, { "grad_norm": 0.10920189321041107, "learning_rate": 3.933771084273828e-05, "loss": 0.012, "step": 35400 }, { "grad_norm": 0.09168273210525513, "learning_rate": 3.931078850234643e-05, "loss": 0.0117, "step": 35410 }, { "grad_norm": 0.13044968247413635, "learning_rate": 3.928386940905483e-05, "loss": 0.0112, "step": 35420 }, { "grad_norm": 0.10957841575145721, "learning_rate": 3.925695357104073e-05, "loss": 0.0104, "step": 35430 }, { "grad_norm": 0.11780714243650436, "learning_rate": 3.923004099648049e-05, "loss": 0.0114, "step": 35440 }, { "grad_norm": 0.11091303825378418, "learning_rate": 3.920313169354944e-05, "loss": 0.0104, "step": 35450 }, { "grad_norm": 0.1258143037557602, "learning_rate": 3.9176225670421897e-05, "loss": 0.0103, "step": 35460 }, { "grad_norm": 0.09433948248624802, "learning_rate": 3.9149322935271224e-05, "loss": 0.0117, "step": 35470 }, { "grad_norm": 0.10179925709962845, "learning_rate": 3.9122423496269725e-05, "loss": 0.0096, "step": 35480 }, { "grad_norm": 0.11236243695020676, "learning_rate": 3.909552736158877e-05, "loss": 0.0108, "step": 35490 }, { "grad_norm": 0.10308012366294861, "learning_rate": 3.90686345393987e-05, "loss": 0.0103, "step": 35500 }, { "grad_norm": 0.2030266672372818, "learning_rate": 3.9041745037868816e-05, "loss": 0.0107, "step": 35510 }, { "grad_norm": 0.10856010019779205, "learning_rate": 3.9014858865167465e-05, "loss": 0.012, "step": 35520 }, { "grad_norm": 0.10318373888731003, "learning_rate": 3.8987976029461935e-05, "loss": 0.0112, "step": 35530 }, { "grad_norm": 0.13218319416046143, "learning_rate": 3.896109653891853e-05, "loss": 0.0109, "step": 35540 }, { "grad_norm": 0.14351743459701538, "learning_rate": 3.893422040170254e-05, "loss": 0.0106, "step": 35550 }, { "grad_norm": 0.1244596391916275, "learning_rate": 3.8907347625978207e-05, "loss": 0.0118, "step": 35560 }, { "grad_norm": 0.08720546960830688, "learning_rate": 3.88804782199088e-05, "loss": 0.0097, "step": 35570 }, { "grad_norm": 0.10067716985940933, "learning_rate": 3.8853612191656495e-05, "loss": 0.0114, "step": 35580 }, { "grad_norm": 0.11071848124265671, "learning_rate": 3.88267495493825e-05, "loss": 0.0103, "step": 35590 }, { "grad_norm": 0.10725226253271103, "learning_rate": 3.8799890301247004e-05, "loss": 0.0102, "step": 35600 }, { "grad_norm": 0.1311970353126526, "learning_rate": 3.8773034455409096e-05, "loss": 0.0109, "step": 35610 }, { "grad_norm": 0.10704036802053452, "learning_rate": 3.8746182020026904e-05, "loss": 0.01, "step": 35620 }, { "grad_norm": 0.11182044446468353, "learning_rate": 3.871933300325745e-05, "loss": 0.01, "step": 35630 }, { "grad_norm": 0.11048778146505356, "learning_rate": 3.869248741325679e-05, "loss": 0.0108, "step": 35640 }, { "grad_norm": 0.1553860455751419, "learning_rate": 3.866564525817992e-05, "loss": 0.0102, "step": 35650 }, { "grad_norm": 0.10476844012737274, "learning_rate": 3.8638806546180725e-05, "loss": 0.01, "step": 35660 }, { "grad_norm": 0.08737237751483917, "learning_rate": 3.861197128541213e-05, "loss": 0.0105, "step": 35670 }, { "grad_norm": 0.11952287703752518, "learning_rate": 3.858513948402599e-05, "loss": 0.0102, "step": 35680 }, { "grad_norm": 0.09575600177049637, "learning_rate": 3.8558311150173077e-05, "loss": 0.0109, "step": 35690 }, { "grad_norm": 0.09194452315568924, "learning_rate": 3.853148629200312e-05, "loss": 0.0097, "step": 35700 }, { "grad_norm": 0.21060103178024292, "learning_rate": 3.850466491766482e-05, "loss": 0.0118, "step": 35710 }, { "grad_norm": 0.11243345588445663, "learning_rate": 3.847784703530583e-05, "loss": 0.0112, "step": 35720 }, { "grad_norm": 0.10014107078313828, "learning_rate": 3.845103265307266e-05, "loss": 0.0102, "step": 35730 }, { "grad_norm": 0.09797878563404083, "learning_rate": 3.842422177911086e-05, "loss": 0.0101, "step": 35740 }, { "grad_norm": 0.08139631897211075, "learning_rate": 3.8397414421564826e-05, "loss": 0.01, "step": 35750 }, { "grad_norm": 0.1096576452255249, "learning_rate": 3.8370610588577935e-05, "loss": 0.01, "step": 35760 }, { "grad_norm": 0.08976994454860687, "learning_rate": 3.834381028829251e-05, "loss": 0.0107, "step": 35770 }, { "grad_norm": 0.12989942729473114, "learning_rate": 3.8317013528849745e-05, "loss": 0.0104, "step": 35780 }, { "grad_norm": 0.09747249633073807, "learning_rate": 3.8290220318389815e-05, "loss": 0.0096, "step": 35790 }, { "grad_norm": 0.10984247922897339, "learning_rate": 3.8263430665051746e-05, "loss": 0.01, "step": 35800 }, { "grad_norm": 0.12515997886657715, "learning_rate": 3.8236644576973554e-05, "loss": 0.011, "step": 35810 }, { "grad_norm": 0.11276265233755112, "learning_rate": 3.820986206229217e-05, "loss": 0.0104, "step": 35820 }, { "grad_norm": 0.08480217307806015, "learning_rate": 3.8183083129143384e-05, "loss": 0.0096, "step": 35830 }, { "grad_norm": 0.08225753903388977, "learning_rate": 3.815630778566193e-05, "loss": 0.0105, "step": 35840 }, { "grad_norm": 0.12601469457149506, "learning_rate": 3.812953603998145e-05, "loss": 0.0117, "step": 35850 }, { "grad_norm": 0.09771319478750229, "learning_rate": 3.8102767900234504e-05, "loss": 0.0103, "step": 35860 }, { "grad_norm": 0.10719819366931915, "learning_rate": 3.807600337455256e-05, "loss": 0.012, "step": 35870 }, { "grad_norm": 0.08477601408958435, "learning_rate": 3.804924247106593e-05, "loss": 0.0093, "step": 35880 }, { "grad_norm": 0.09423492103815079, "learning_rate": 3.8022485197903925e-05, "loss": 0.0102, "step": 35890 }, { "grad_norm": 0.09016136825084686, "learning_rate": 3.799573156319464e-05, "loss": 0.011, "step": 35900 }, { "grad_norm": 0.10810288786888123, "learning_rate": 3.796898157506515e-05, "loss": 0.0107, "step": 35910 }, { "grad_norm": 0.08636783063411713, "learning_rate": 3.794223524164143e-05, "loss": 0.0108, "step": 35920 }, { "grad_norm": 0.11618486791849136, "learning_rate": 3.7915492571048245e-05, "loss": 0.0101, "step": 35930 }, { "grad_norm": 0.10701614618301392, "learning_rate": 3.788875357140937e-05, "loss": 0.0099, "step": 35940 }, { "grad_norm": 0.1037122830748558, "learning_rate": 3.786201825084736e-05, "loss": 0.0108, "step": 35950 }, { "grad_norm": 0.10846612602472305, "learning_rate": 3.783528661748372e-05, "loss": 0.009, "step": 35960 }, { "grad_norm": 0.07894027233123779, "learning_rate": 3.780855867943882e-05, "loss": 0.0107, "step": 35970 }, { "grad_norm": 0.1056571751832962, "learning_rate": 3.778183444483189e-05, "loss": 0.0114, "step": 35980 }, { "grad_norm": 0.15882694721221924, "learning_rate": 3.775511392178108e-05, "loss": 0.0096, "step": 35990 }, { "grad_norm": 0.14964187145233154, "learning_rate": 3.772839711840332e-05, "loss": 0.0127, "step": 36000 }, { "grad_norm": 0.10607660561800003, "learning_rate": 3.7701684042814515e-05, "loss": 0.0096, "step": 36010 }, { "grad_norm": 0.11530221998691559, "learning_rate": 3.76749747031294e-05, "loss": 0.0109, "step": 36020 }, { "grad_norm": 0.13556542992591858, "learning_rate": 3.764826910746152e-05, "loss": 0.0102, "step": 36030 }, { "grad_norm": 0.1022723987698555, "learning_rate": 3.762156726392338e-05, "loss": 0.0101, "step": 36040 }, { "grad_norm": 0.10879649966955185, "learning_rate": 3.759486918062625e-05, "loss": 0.0092, "step": 36050 }, { "grad_norm": 0.10358066111803055, "learning_rate": 3.756817486568033e-05, "loss": 0.0103, "step": 36060 }, { "grad_norm": 0.1112111434340477, "learning_rate": 3.7541484327194654e-05, "loss": 0.0105, "step": 36070 }, { "grad_norm": 0.10270626097917557, "learning_rate": 3.751479757327707e-05, "loss": 0.0116, "step": 36080 }, { "grad_norm": 0.10817980766296387, "learning_rate": 3.7488114612034345e-05, "loss": 0.0104, "step": 36090 }, { "grad_norm": 0.1119508296251297, "learning_rate": 3.7461435451572044e-05, "loss": 0.0109, "step": 36100 }, { "grad_norm": 0.12856552004814148, "learning_rate": 3.743476009999459e-05, "loss": 0.0109, "step": 36110 }, { "grad_norm": 0.12636546790599823, "learning_rate": 3.7408088565405245e-05, "loss": 0.0101, "step": 36120 }, { "grad_norm": 0.10177326202392578, "learning_rate": 3.738142085590612e-05, "loss": 0.0095, "step": 36130 }, { "grad_norm": 0.09761515259742737, "learning_rate": 3.7354756979598194e-05, "loss": 0.0091, "step": 36140 }, { "grad_norm": 0.1383141279220581, "learning_rate": 3.7328096944581187e-05, "loss": 0.0106, "step": 36150 }, { "grad_norm": 0.08088533580303192, "learning_rate": 3.730144075895377e-05, "loss": 0.0097, "step": 36160 }, { "grad_norm": 0.11391834169626236, "learning_rate": 3.727478843081335e-05, "loss": 0.0104, "step": 36170 }, { "grad_norm": 0.10469455271959305, "learning_rate": 3.72481399682562e-05, "loss": 0.011, "step": 36180 }, { "grad_norm": 0.10512982308864594, "learning_rate": 3.722149537937747e-05, "loss": 0.011, "step": 36190 }, { "grad_norm": 0.07727988064289093, "learning_rate": 3.7194854672271015e-05, "loss": 0.0102, "step": 36200 }, { "grad_norm": 0.08017571270465851, "learning_rate": 3.7168217855029644e-05, "loss": 0.0106, "step": 36210 }, { "grad_norm": 0.10905525833368301, "learning_rate": 3.7141584935744856e-05, "loss": 0.0109, "step": 36220 }, { "grad_norm": 0.10291799157857895, "learning_rate": 3.7114955922507055e-05, "loss": 0.011, "step": 36230 }, { "grad_norm": 0.08978650718927383, "learning_rate": 3.708833082340545e-05, "loss": 0.0115, "step": 36240 }, { "grad_norm": 0.10352697223424911, "learning_rate": 3.7061709646528034e-05, "loss": 0.0097, "step": 36250 }, { "grad_norm": 0.10448192059993744, "learning_rate": 3.7035092399961604e-05, "loss": 0.0101, "step": 36260 }, { "grad_norm": 0.15601372718811035, "learning_rate": 3.700847909179177e-05, "loss": 0.0113, "step": 36270 }, { "grad_norm": 0.11054005473852158, "learning_rate": 3.698186973010297e-05, "loss": 0.0106, "step": 36280 }, { "grad_norm": 0.11900482326745987, "learning_rate": 3.695526432297844e-05, "loss": 0.0121, "step": 36290 }, { "grad_norm": 0.10646956413984299, "learning_rate": 3.692866287850017e-05, "loss": 0.0101, "step": 36300 }, { "grad_norm": 0.10669980198144913, "learning_rate": 3.6902065404749006e-05, "loss": 0.0105, "step": 36310 }, { "grad_norm": 0.08970361948013306, "learning_rate": 3.6875471909804516e-05, "loss": 0.0091, "step": 36320 }, { "grad_norm": 0.09791141748428345, "learning_rate": 3.6848882401745135e-05, "loss": 0.0109, "step": 36330 }, { "grad_norm": 0.11675461381673813, "learning_rate": 3.682229688864806e-05, "loss": 0.01, "step": 36340 }, { "grad_norm": 0.08364350348711014, "learning_rate": 3.6795715378589235e-05, "loss": 0.0103, "step": 36350 }, { "grad_norm": 0.11415330320596695, "learning_rate": 3.676913787964345e-05, "loss": 0.0092, "step": 36360 }, { "grad_norm": 0.09999009966850281, "learning_rate": 3.674256439988423e-05, "loss": 0.0109, "step": 36370 }, { "grad_norm": 0.10588013380765915, "learning_rate": 3.6715994947383904e-05, "loss": 0.0107, "step": 36380 }, { "grad_norm": 0.11165808141231537, "learning_rate": 3.668942953021357e-05, "loss": 0.0118, "step": 36390 }, { "grad_norm": 0.19141091406345367, "learning_rate": 3.66628681564431e-05, "loss": 0.0113, "step": 36400 }, { "grad_norm": 0.09978413581848145, "learning_rate": 3.663631083414114e-05, "loss": 0.0099, "step": 36410 }, { "grad_norm": 0.09929484874010086, "learning_rate": 3.660975757137509e-05, "loss": 0.0109, "step": 36420 }, { "grad_norm": 0.1504095047712326, "learning_rate": 3.658320837621114e-05, "loss": 0.0095, "step": 36430 }, { "grad_norm": 0.11445922404527664, "learning_rate": 3.655666325671426e-05, "loss": 0.0104, "step": 36440 }, { "grad_norm": 0.1339639276266098, "learning_rate": 3.65301222209481e-05, "loss": 0.0097, "step": 36450 }, { "grad_norm": 0.14801914989948273, "learning_rate": 3.650358527697519e-05, "loss": 0.012, "step": 36460 }, { "grad_norm": 0.0844438448548317, "learning_rate": 3.64770524328567e-05, "loss": 0.0095, "step": 36470 }, { "grad_norm": 0.12124589085578918, "learning_rate": 3.645052369665265e-05, "loss": 0.0102, "step": 36480 }, { "grad_norm": 0.12116198241710663, "learning_rate": 3.6423999076421724e-05, "loss": 0.0112, "step": 36490 }, { "grad_norm": 0.11808203905820847, "learning_rate": 3.639747858022142e-05, "loss": 0.0104, "step": 36500 }, { "grad_norm": 0.09329617768526077, "learning_rate": 3.637096221610799e-05, "loss": 0.0099, "step": 36510 }, { "grad_norm": 0.09505262225866318, "learning_rate": 3.634444999213638e-05, "loss": 0.0091, "step": 36520 }, { "grad_norm": 0.13192321360111237, "learning_rate": 3.6317941916360296e-05, "loss": 0.012, "step": 36530 }, { "grad_norm": 0.09447940438985825, "learning_rate": 3.629143799683221e-05, "loss": 0.0092, "step": 36540 }, { "grad_norm": 0.10106628388166428, "learning_rate": 3.626493824160331e-05, "loss": 0.0115, "step": 36550 }, { "grad_norm": 0.09049133211374283, "learning_rate": 3.623844265872352e-05, "loss": 0.009, "step": 36560 }, { "grad_norm": 0.1253533512353897, "learning_rate": 3.621195125624149e-05, "loss": 0.0101, "step": 36570 }, { "grad_norm": 0.12062305957078934, "learning_rate": 3.618546404220463e-05, "loss": 0.0106, "step": 36580 }, { "grad_norm": 0.13573428988456726, "learning_rate": 3.615898102465903e-05, "loss": 0.0114, "step": 36590 }, { "grad_norm": 0.13696764409542084, "learning_rate": 3.6132502211649544e-05, "loss": 0.009, "step": 36600 }, { "grad_norm": 0.08077044785022736, "learning_rate": 3.610602761121975e-05, "loss": 0.009, "step": 36610 }, { "grad_norm": 0.09625018388032913, "learning_rate": 3.6079557231411897e-05, "loss": 0.0106, "step": 36620 }, { "grad_norm": 0.13667479157447815, "learning_rate": 3.6053091080267035e-05, "loss": 0.0122, "step": 36630 }, { "grad_norm": 0.11931179463863373, "learning_rate": 3.602662916582483e-05, "loss": 0.0124, "step": 36640 }, { "grad_norm": 0.0975610688328743, "learning_rate": 3.600017149612375e-05, "loss": 0.0106, "step": 36650 }, { "grad_norm": 0.10196629166603088, "learning_rate": 3.5973718079200935e-05, "loss": 0.0108, "step": 36660 }, { "grad_norm": 0.10949734598398209, "learning_rate": 3.5947268923092216e-05, "loss": 0.0101, "step": 36670 }, { "grad_norm": 0.10661862045526505, "learning_rate": 3.592082403583216e-05, "loss": 0.0099, "step": 36680 }, { "grad_norm": 0.10183314234018326, "learning_rate": 3.5894383425454004e-05, "loss": 0.0087, "step": 36690 }, { "grad_norm": 0.11366202682256699, "learning_rate": 3.586794709998975e-05, "loss": 0.0097, "step": 36700 }, { "grad_norm": 0.1042177677154541, "learning_rate": 3.584151506747002e-05, "loss": 0.0116, "step": 36710 }, { "grad_norm": 0.09542471170425415, "learning_rate": 3.581508733592418e-05, "loss": 0.0091, "step": 36720 }, { "grad_norm": 0.11574020236730576, "learning_rate": 3.5788663913380297e-05, "loss": 0.0113, "step": 36730 }, { "grad_norm": 0.11457521468400955, "learning_rate": 3.576224480786506e-05, "loss": 0.0103, "step": 36740 }, { "grad_norm": 0.11881379783153534, "learning_rate": 3.573583002740393e-05, "loss": 0.0108, "step": 36750 }, { "grad_norm": 0.09511375427246094, "learning_rate": 3.570941958002103e-05, "loss": 0.0086, "step": 36760 }, { "grad_norm": 0.10276708751916885, "learning_rate": 3.568301347373912e-05, "loss": 0.0112, "step": 36770 }, { "grad_norm": 0.09791537374258041, "learning_rate": 3.5656611716579726e-05, "loss": 0.009, "step": 36780 }, { "grad_norm": 0.1305982768535614, "learning_rate": 3.5630214316562946e-05, "loss": 0.0109, "step": 36790 }, { "grad_norm": 0.12352203577756882, "learning_rate": 3.560382128170766e-05, "loss": 0.0087, "step": 36800 }, { "grad_norm": 0.11370077729225159, "learning_rate": 3.5577432620031374e-05, "loss": 0.0125, "step": 36810 }, { "grad_norm": 0.10902313143014908, "learning_rate": 3.5551048339550216e-05, "loss": 0.0093, "step": 36820 }, { "grad_norm": 0.12435229122638702, "learning_rate": 3.55246684482791e-05, "loss": 0.0095, "step": 36830 }, { "grad_norm": 0.12295302003622055, "learning_rate": 3.5498292954231496e-05, "loss": 0.0106, "step": 36840 }, { "grad_norm": 0.12953004240989685, "learning_rate": 3.54719218654196e-05, "loss": 0.01, "step": 36850 }, { "grad_norm": 0.10515987128019333, "learning_rate": 3.544555518985425e-05, "loss": 0.0101, "step": 36860 }, { "grad_norm": 0.10978373140096664, "learning_rate": 3.541919293554494e-05, "loss": 0.0108, "step": 36870 }, { "grad_norm": 0.14987950026988983, "learning_rate": 3.539283511049985e-05, "loss": 0.0104, "step": 36880 }, { "grad_norm": 0.11195291578769684, "learning_rate": 3.5366481722725755e-05, "loss": 0.0107, "step": 36890 }, { "grad_norm": 0.09611068665981293, "learning_rate": 3.534013278022816e-05, "loss": 0.0114, "step": 36900 }, { "grad_norm": 0.09412197768688202, "learning_rate": 3.531378829101113e-05, "loss": 0.0102, "step": 36910 }, { "grad_norm": 0.08248183876276016, "learning_rate": 3.528744826307746e-05, "loss": 0.01, "step": 36920 }, { "grad_norm": 0.08914297074079514, "learning_rate": 3.5261112704428554e-05, "loss": 0.0099, "step": 36930 }, { "grad_norm": 0.09144608676433563, "learning_rate": 3.523478162306443e-05, "loss": 0.01, "step": 36940 }, { "grad_norm": 0.09649288654327393, "learning_rate": 3.520845502698381e-05, "loss": 0.0103, "step": 36950 }, { "grad_norm": 0.10550790280103683, "learning_rate": 3.5182132924184005e-05, "loss": 0.0101, "step": 36960 }, { "grad_norm": 0.08305709809064865, "learning_rate": 3.5155815322660966e-05, "loss": 0.0088, "step": 36970 }, { "grad_norm": 0.10765886306762695, "learning_rate": 3.512950223040931e-05, "loss": 0.01, "step": 36980 }, { "grad_norm": 0.08971162140369415, "learning_rate": 3.5103193655422216e-05, "loss": 0.009, "step": 36990 }, { "grad_norm": 0.11067219823598862, "learning_rate": 3.5076889605691596e-05, "loss": 0.0108, "step": 37000 }, { "grad_norm": 0.13332979381084442, "learning_rate": 3.505059008920787e-05, "loss": 0.0102, "step": 37010 }, { "grad_norm": 0.08699944615364075, "learning_rate": 3.502429511396016e-05, "loss": 0.0092, "step": 37020 }, { "grad_norm": 0.10465023666620255, "learning_rate": 3.4998004687936196e-05, "loss": 0.0095, "step": 37030 }, { "grad_norm": 0.12957516312599182, "learning_rate": 3.497171881912229e-05, "loss": 0.0117, "step": 37040 }, { "grad_norm": 0.13005532324314117, "learning_rate": 3.494543751550342e-05, "loss": 0.0098, "step": 37050 }, { "grad_norm": 0.15024293959140778, "learning_rate": 3.491916078506313e-05, "loss": 0.0122, "step": 37060 }, { "grad_norm": 0.09015185385942459, "learning_rate": 3.489288863578361e-05, "loss": 0.0104, "step": 37070 }, { "grad_norm": 0.1684410721063614, "learning_rate": 3.4866621075645646e-05, "loss": 0.0097, "step": 37080 }, { "grad_norm": 0.11698035150766373, "learning_rate": 3.4840358112628614e-05, "loss": 0.0099, "step": 37090 }, { "grad_norm": 0.11065094918012619, "learning_rate": 3.481409975471053e-05, "loss": 0.0105, "step": 37100 }, { "grad_norm": 0.1096998006105423, "learning_rate": 3.4787846009867986e-05, "loss": 0.0098, "step": 37110 }, { "grad_norm": 0.1271166205406189, "learning_rate": 3.476159688607615e-05, "loss": 0.0112, "step": 37120 }, { "grad_norm": 0.14475767314434052, "learning_rate": 3.4735352391308854e-05, "loss": 0.0108, "step": 37130 }, { "grad_norm": 0.1661035567522049, "learning_rate": 3.4709112533538446e-05, "loss": 0.0093, "step": 37140 }, { "grad_norm": 0.1518985629081726, "learning_rate": 3.4682877320735934e-05, "loss": 0.0085, "step": 37150 }, { "grad_norm": 0.09706536680459976, "learning_rate": 3.465664676087085e-05, "loss": 0.0108, "step": 37160 }, { "grad_norm": 0.06477616727352142, "learning_rate": 3.463042086191136e-05, "loss": 0.0094, "step": 37170 }, { "grad_norm": 0.11479288339614868, "learning_rate": 3.460419963182423e-05, "loss": 0.011, "step": 37180 }, { "grad_norm": 0.09813254326581955, "learning_rate": 3.457798307857473e-05, "loss": 0.0116, "step": 37190 }, { "grad_norm": 0.08786960691213608, "learning_rate": 3.455177121012678e-05, "loss": 0.0106, "step": 37200 }, { "grad_norm": 0.119801364839077, "learning_rate": 3.452556403444285e-05, "loss": 0.0101, "step": 37210 }, { "grad_norm": 0.12415111809968948, "learning_rate": 3.4499361559483975e-05, "loss": 0.0095, "step": 37220 }, { "grad_norm": 0.12462235987186432, "learning_rate": 3.44731637932098e-05, "loss": 0.0108, "step": 37230 }, { "grad_norm": 0.12089183926582336, "learning_rate": 3.44469707435785e-05, "loss": 0.0098, "step": 37240 }, { "grad_norm": 0.14064083993434906, "learning_rate": 3.4420782418546835e-05, "loss": 0.0114, "step": 37250 }, { "grad_norm": 0.11819592863321304, "learning_rate": 3.439459882607012e-05, "loss": 0.0106, "step": 37260 }, { "grad_norm": 0.08362634479999542, "learning_rate": 3.436841997410225e-05, "loss": 0.0117, "step": 37270 }, { "grad_norm": 0.09011656045913696, "learning_rate": 3.434224587059567e-05, "loss": 0.0107, "step": 37280 }, { "grad_norm": 0.10526768863201141, "learning_rate": 3.431607652350136e-05, "loss": 0.0105, "step": 37290 }, { "grad_norm": 0.12617656588554382, "learning_rate": 3.428991194076891e-05, "loss": 0.0105, "step": 37300 }, { "grad_norm": 0.11803601682186127, "learning_rate": 3.4263752130346394e-05, "loss": 0.0105, "step": 37310 }, { "grad_norm": 0.08280165493488312, "learning_rate": 3.4237597100180515e-05, "loss": 0.0098, "step": 37320 }, { "grad_norm": 0.11769554018974304, "learning_rate": 3.4211446858216427e-05, "loss": 0.0106, "step": 37330 }, { "grad_norm": 0.11990980803966522, "learning_rate": 3.4185301412397915e-05, "loss": 0.0099, "step": 37340 }, { "grad_norm": 0.09012394398450851, "learning_rate": 3.415916077066729e-05, "loss": 0.0102, "step": 37350 }, { "grad_norm": 0.08005359023809433, "learning_rate": 3.413302494096535e-05, "loss": 0.0104, "step": 37360 }, { "grad_norm": 0.1422845870256424, "learning_rate": 3.410689393123151e-05, "loss": 0.0102, "step": 37370 }, { "grad_norm": 0.09608494490385056, "learning_rate": 3.408076774940364e-05, "loss": 0.01, "step": 37380 }, { "grad_norm": 0.11452942341566086, "learning_rate": 3.40546464034182e-05, "loss": 0.01, "step": 37390 }, { "grad_norm": 0.1492355912923813, "learning_rate": 3.4028529901210185e-05, "loss": 0.0102, "step": 37400 }, { "grad_norm": 0.1284862458705902, "learning_rate": 3.4002418250713086e-05, "loss": 0.0101, "step": 37410 }, { "grad_norm": 0.12115644663572311, "learning_rate": 3.3976311459858936e-05, "loss": 0.0101, "step": 37420 }, { "grad_norm": 0.10069438070058823, "learning_rate": 3.395020953657826e-05, "loss": 0.009, "step": 37430 }, { "grad_norm": 0.12021742761135101, "learning_rate": 3.3924112488800165e-05, "loss": 0.0098, "step": 37440 }, { "grad_norm": 0.16998648643493652, "learning_rate": 3.389802032445225e-05, "loss": 0.0112, "step": 37450 }, { "grad_norm": 0.1430702954530716, "learning_rate": 3.38719330514606e-05, "loss": 0.0106, "step": 37460 }, { "grad_norm": 0.11845054477453232, "learning_rate": 3.3845850677749866e-05, "loss": 0.0096, "step": 37470 }, { "grad_norm": 0.09895824640989304, "learning_rate": 3.3819773211243157e-05, "loss": 0.0092, "step": 37480 }, { "grad_norm": 0.10123758763074875, "learning_rate": 3.379370065986213e-05, "loss": 0.01, "step": 37490 }, { "grad_norm": 0.09642719477415085, "learning_rate": 3.3767633031526955e-05, "loss": 0.0109, "step": 37500 }, { "grad_norm": 0.08760936558246613, "learning_rate": 3.374157033415626e-05, "loss": 0.0123, "step": 37510 }, { "grad_norm": 0.09867319464683533, "learning_rate": 3.371551257566723e-05, "loss": 0.0108, "step": 37520 }, { "grad_norm": 0.1138637512922287, "learning_rate": 3.36894597639755e-05, "loss": 0.01, "step": 37530 }, { "grad_norm": 0.13156558573246002, "learning_rate": 3.366341190699523e-05, "loss": 0.0104, "step": 37540 }, { "grad_norm": 0.13353896141052246, "learning_rate": 3.36373690126391e-05, "loss": 0.0094, "step": 37550 }, { "grad_norm": 0.0905652642250061, "learning_rate": 3.3611331088818234e-05, "loss": 0.0098, "step": 37560 }, { "grad_norm": 0.10061235725879669, "learning_rate": 3.3585298143442265e-05, "loss": 0.0106, "step": 37570 }, { "grad_norm": 0.09870488941669464, "learning_rate": 3.35592701844193e-05, "loss": 0.0096, "step": 37580 }, { "grad_norm": 0.10667940974235535, "learning_rate": 3.353324721965596e-05, "loss": 0.0104, "step": 37590 }, { "grad_norm": 0.1120419129729271, "learning_rate": 3.350722925705736e-05, "loss": 0.0124, "step": 37600 }, { "grad_norm": 0.12946437299251556, "learning_rate": 3.348121630452703e-05, "loss": 0.0104, "step": 37610 }, { "grad_norm": 0.10541676729917526, "learning_rate": 3.3455208369967044e-05, "loss": 0.0106, "step": 37620 }, { "grad_norm": 0.11439742892980576, "learning_rate": 3.34292054612779e-05, "loss": 0.0104, "step": 37630 }, { "grad_norm": 0.10018013417720795, "learning_rate": 3.340320758635861e-05, "loss": 0.0095, "step": 37640 }, { "grad_norm": 0.09116560220718384, "learning_rate": 3.337721475310666e-05, "loss": 0.0098, "step": 37650 }, { "grad_norm": 0.10054893046617508, "learning_rate": 3.335122696941795e-05, "loss": 0.0095, "step": 37660 }, { "grad_norm": 0.09293455630540848, "learning_rate": 3.332524424318692e-05, "loss": 0.0093, "step": 37670 }, { "grad_norm": 0.10901442915201187, "learning_rate": 3.32992665823064e-05, "loss": 0.0095, "step": 37680 }, { "grad_norm": 0.11695107817649841, "learning_rate": 3.327329399466774e-05, "loss": 0.0096, "step": 37690 }, { "grad_norm": 0.10078023374080658, "learning_rate": 3.324732648816072e-05, "loss": 0.0096, "step": 37700 }, { "grad_norm": 0.10546889901161194, "learning_rate": 3.322136407067358e-05, "loss": 0.0099, "step": 37710 }, { "grad_norm": 0.11596306413412094, "learning_rate": 3.3195406750093036e-05, "loss": 0.0098, "step": 37720 }, { "grad_norm": 0.13749371469020844, "learning_rate": 3.3169454534304205e-05, "loss": 0.01, "step": 37730 }, { "grad_norm": 0.10601814091205597, "learning_rate": 3.3143507431190725e-05, "loss": 0.0113, "step": 37740 }, { "grad_norm": 0.11428157240152359, "learning_rate": 3.311756544863459e-05, "loss": 0.0097, "step": 37750 }, { "grad_norm": 0.12392271310091019, "learning_rate": 3.309162859451633e-05, "loss": 0.0102, "step": 37760 }, { "grad_norm": 0.17883527278900146, "learning_rate": 3.306569687671487e-05, "loss": 0.0104, "step": 37770 }, { "grad_norm": 0.08966252207756042, "learning_rate": 3.303977030310756e-05, "loss": 0.0097, "step": 37780 }, { "grad_norm": 0.08577735722064972, "learning_rate": 3.3013848881570245e-05, "loss": 0.0095, "step": 37790 }, { "grad_norm": 0.09261380136013031, "learning_rate": 3.298793261997712e-05, "loss": 0.0097, "step": 37800 }, { "grad_norm": 0.09453614056110382, "learning_rate": 3.2962021526200893e-05, "loss": 0.0096, "step": 37810 }, { "grad_norm": 0.11144179105758667, "learning_rate": 3.293611560811268e-05, "loss": 0.0091, "step": 37820 }, { "grad_norm": 0.11276240646839142, "learning_rate": 3.291021487358199e-05, "loss": 0.0116, "step": 37830 }, { "grad_norm": 0.10538658499717712, "learning_rate": 3.28843193304768e-05, "loss": 0.0095, "step": 37840 }, { "grad_norm": 0.08979925513267517, "learning_rate": 3.2858428986663456e-05, "loss": 0.0112, "step": 37850 }, { "grad_norm": 0.0844816341996193, "learning_rate": 3.283254385000681e-05, "loss": 0.0094, "step": 37860 }, { "grad_norm": 0.09911663830280304, "learning_rate": 3.2806663928370076e-05, "loss": 0.0105, "step": 37870 }, { "grad_norm": 0.13272453844547272, "learning_rate": 3.278078922961485e-05, "loss": 0.0113, "step": 37880 }, { "grad_norm": 0.10609358549118042, "learning_rate": 3.275491976160123e-05, "loss": 0.0088, "step": 37890 }, { "grad_norm": 0.08057059347629547, "learning_rate": 3.2729055532187645e-05, "loss": 0.0098, "step": 37900 }, { "grad_norm": 0.09328616410493851, "learning_rate": 3.270319654923097e-05, "loss": 0.0089, "step": 37910 }, { "grad_norm": 0.12745627760887146, "learning_rate": 3.2677342820586506e-05, "loss": 0.0113, "step": 37920 }, { "grad_norm": 0.12336496263742447, "learning_rate": 3.2651494354107905e-05, "loss": 0.0106, "step": 37930 }, { "grad_norm": 0.1029326319694519, "learning_rate": 3.2625651157647266e-05, "loss": 0.0097, "step": 37940 }, { "grad_norm": 0.08339185267686844, "learning_rate": 3.259981323905505e-05, "loss": 0.0094, "step": 37950 }, { "grad_norm": 0.10858765989542007, "learning_rate": 3.257398060618014e-05, "loss": 0.0099, "step": 37960 }, { "grad_norm": 0.11084870994091034, "learning_rate": 3.254815326686983e-05, "loss": 0.0111, "step": 37970 }, { "grad_norm": 0.10073333978652954, "learning_rate": 3.2522331228969774e-05, "loss": 0.0097, "step": 37980 }, { "grad_norm": 0.09602466970682144, "learning_rate": 3.2496514500324006e-05, "loss": 0.0097, "step": 37990 }, { "grad_norm": 0.09011806547641754, "learning_rate": 3.247070308877498e-05, "loss": 0.01, "step": 38000 }, { "grad_norm": 0.1070069819688797, "learning_rate": 3.2444897002163515e-05, "loss": 0.0089, "step": 38010 }, { "grad_norm": 0.12185488641262054, "learning_rate": 3.241909624832885e-05, "loss": 0.01, "step": 38020 }, { "grad_norm": 0.11698693782091141, "learning_rate": 3.239330083510852e-05, "loss": 0.0099, "step": 38030 }, { "grad_norm": 0.1171850711107254, "learning_rate": 3.236751077033855e-05, "loss": 0.0102, "step": 38040 }, { "grad_norm": 0.1189458817243576, "learning_rate": 3.234172606185322e-05, "loss": 0.0103, "step": 38050 }, { "grad_norm": 0.0991196483373642, "learning_rate": 3.231594671748528e-05, "loss": 0.0101, "step": 38060 }, { "grad_norm": 0.11596853286027908, "learning_rate": 3.2290172745065815e-05, "loss": 0.0096, "step": 38070 }, { "grad_norm": 0.10670359432697296, "learning_rate": 3.226440415242426e-05, "loss": 0.0114, "step": 38080 }, { "grad_norm": 0.08217617124319077, "learning_rate": 3.223864094738846e-05, "loss": 0.01, "step": 38090 }, { "grad_norm": 0.11026831716299057, "learning_rate": 3.221288313778456e-05, "loss": 0.0091, "step": 38100 }, { "grad_norm": 0.10046049952507019, "learning_rate": 3.2187130731437125e-05, "loss": 0.0096, "step": 38110 }, { "grad_norm": 0.11803407222032547, "learning_rate": 3.216138373616905e-05, "loss": 0.0096, "step": 38120 }, { "grad_norm": 0.11443085968494415, "learning_rate": 3.21356421598016e-05, "loss": 0.0105, "step": 38130 }, { "grad_norm": 0.11482931673526764, "learning_rate": 3.210990601015438e-05, "loss": 0.0089, "step": 38140 }, { "grad_norm": 0.10097600519657135, "learning_rate": 3.208417529504535e-05, "loss": 0.0115, "step": 38150 }, { "grad_norm": 0.09068775922060013, "learning_rate": 3.205845002229084e-05, "loss": 0.0114, "step": 38160 }, { "grad_norm": 0.1116911917924881, "learning_rate": 3.203273019970547e-05, "loss": 0.0107, "step": 38170 }, { "grad_norm": 0.1186850368976593, "learning_rate": 3.200701583510227e-05, "loss": 0.0103, "step": 38180 }, { "grad_norm": 0.11702167242765427, "learning_rate": 3.198130693629261e-05, "loss": 0.0113, "step": 38190 }, { "grad_norm": 0.14377012848854065, "learning_rate": 3.195560351108612e-05, "loss": 0.0103, "step": 38200 }, { "grad_norm": 0.09657768905162811, "learning_rate": 3.1929905567290865e-05, "loss": 0.0102, "step": 38210 }, { "grad_norm": 0.1128900870680809, "learning_rate": 3.1904213112713164e-05, "loss": 0.0094, "step": 38220 }, { "grad_norm": 0.17640157043933868, "learning_rate": 3.187852615515774e-05, "loss": 0.0094, "step": 38230 }, { "grad_norm": 0.09488674253225327, "learning_rate": 3.1852844702427606e-05, "loss": 0.0082, "step": 38240 }, { "grad_norm": 0.08122143149375916, "learning_rate": 3.18271687623241e-05, "loss": 0.0085, "step": 38250 }, { "grad_norm": 0.14181676506996155, "learning_rate": 3.1801498342646896e-05, "loss": 0.0098, "step": 38260 }, { "grad_norm": 0.11309841275215149, "learning_rate": 3.177583345119398e-05, "loss": 0.0097, "step": 38270 }, { "grad_norm": 0.1158507764339447, "learning_rate": 3.17501740957617e-05, "loss": 0.0091, "step": 38280 }, { "grad_norm": 0.09934771806001663, "learning_rate": 3.172452028414467e-05, "loss": 0.0094, "step": 38290 }, { "grad_norm": 0.10484591126441956, "learning_rate": 3.169887202413583e-05, "loss": 0.0107, "step": 38300 }, { "grad_norm": 0.12969696521759033, "learning_rate": 3.167322932352646e-05, "loss": 0.0099, "step": 38310 }, { "grad_norm": 0.11161065846681595, "learning_rate": 3.164759219010613e-05, "loss": 0.01, "step": 38320 }, { "grad_norm": 0.151212677359581, "learning_rate": 3.1621960631662725e-05, "loss": 0.0095, "step": 38330 }, { "grad_norm": 0.15113751590251923, "learning_rate": 3.159633465598245e-05, "loss": 0.0108, "step": 38340 }, { "grad_norm": 0.12659431993961334, "learning_rate": 3.1570714270849767e-05, "loss": 0.0097, "step": 38350 }, { "grad_norm": 0.08770876377820969, "learning_rate": 3.1545099484047516e-05, "loss": 0.0087, "step": 38360 }, { "grad_norm": 0.10582201182842255, "learning_rate": 3.151949030335674e-05, "loss": 0.009, "step": 38370 }, { "grad_norm": 0.10734758526086807, "learning_rate": 3.149388673655687e-05, "loss": 0.0082, "step": 38380 }, { "grad_norm": 0.10677269101142883, "learning_rate": 3.146828879142559e-05, "loss": 0.0099, "step": 38390 }, { "grad_norm": 0.12978726625442505, "learning_rate": 3.1442696475738866e-05, "loss": 0.0099, "step": 38400 }, { "grad_norm": 0.1128922551870346, "learning_rate": 3.141710979727098e-05, "loss": 0.0093, "step": 38410 }, { "grad_norm": 0.09922391921281815, "learning_rate": 3.139152876379447e-05, "loss": 0.0091, "step": 38420 }, { "grad_norm": 0.09936517477035522, "learning_rate": 3.1365953383080214e-05, "loss": 0.0109, "step": 38430 }, { "grad_norm": 0.09774452447891235, "learning_rate": 3.134038366289731e-05, "loss": 0.0097, "step": 38440 }, { "grad_norm": 0.10678347945213318, "learning_rate": 3.131481961101317e-05, "loss": 0.0111, "step": 38450 }, { "grad_norm": 0.102236807346344, "learning_rate": 3.128926123519349e-05, "loss": 0.0101, "step": 38460 }, { "grad_norm": 0.2018788605928421, "learning_rate": 3.1263708543202194e-05, "loss": 0.0098, "step": 38470 }, { "grad_norm": 0.14604151248931885, "learning_rate": 3.123816154280155e-05, "loss": 0.01, "step": 38480 }, { "grad_norm": 0.1088051050901413, "learning_rate": 3.121262024175207e-05, "loss": 0.0103, "step": 38490 }, { "grad_norm": 0.10513612627983093, "learning_rate": 3.118708464781248e-05, "loss": 0.0094, "step": 38500 }, { "grad_norm": 0.09196211397647858, "learning_rate": 3.116155476873987e-05, "loss": 0.0096, "step": 38510 }, { "grad_norm": 0.07485824078321457, "learning_rate": 3.11360306122895e-05, "loss": 0.0094, "step": 38520 }, { "grad_norm": 0.12433882802724838, "learning_rate": 3.1110512186214975e-05, "loss": 0.01, "step": 38530 }, { "grad_norm": 0.09333372861146927, "learning_rate": 3.1084999498268095e-05, "loss": 0.0096, "step": 38540 }, { "grad_norm": 0.10067777335643768, "learning_rate": 3.1059492556198934e-05, "loss": 0.0096, "step": 38550 }, { "grad_norm": 0.14318367838859558, "learning_rate": 3.103399136775586e-05, "loss": 0.0101, "step": 38560 }, { "grad_norm": 0.11159172654151917, "learning_rate": 3.100849594068541e-05, "loss": 0.0096, "step": 38570 }, { "grad_norm": 0.12268111109733582, "learning_rate": 3.0983006282732484e-05, "loss": 0.0095, "step": 38580 }, { "grad_norm": 0.13375642895698547, "learning_rate": 3.0957522401640116e-05, "loss": 0.0105, "step": 38590 }, { "grad_norm": 0.0841287299990654, "learning_rate": 3.0932044305149645e-05, "loss": 0.0103, "step": 38600 }, { "grad_norm": 0.13181647658348083, "learning_rate": 3.090657200100068e-05, "loss": 0.011, "step": 38610 }, { "grad_norm": 0.11322209984064102, "learning_rate": 3.088110549693099e-05, "loss": 0.0098, "step": 38620 }, { "grad_norm": 0.08949697762727737, "learning_rate": 3.085564480067667e-05, "loss": 0.0089, "step": 38630 }, { "grad_norm": 0.10812582820653915, "learning_rate": 3.0830189919971955e-05, "loss": 0.0093, "step": 38640 }, { "grad_norm": 0.10045907646417618, "learning_rate": 3.080474086254939e-05, "loss": 0.0103, "step": 38650 }, { "grad_norm": 0.08906853944063187, "learning_rate": 3.077929763613975e-05, "loss": 0.0094, "step": 38660 }, { "grad_norm": 0.08079928904771805, "learning_rate": 3.075386024847198e-05, "loss": 0.0092, "step": 38670 }, { "grad_norm": 0.09535568952560425, "learning_rate": 3.072842870727331e-05, "loss": 0.0105, "step": 38680 }, { "grad_norm": 0.11709211021661758, "learning_rate": 3.070300302026916e-05, "loss": 0.0094, "step": 38690 }, { "grad_norm": 0.14285118877887726, "learning_rate": 3.067758319518318e-05, "loss": 0.0099, "step": 38700 }, { "grad_norm": 0.0859043076634407, "learning_rate": 3.065216923973725e-05, "loss": 0.0088, "step": 38710 }, { "grad_norm": 0.176706925034523, "learning_rate": 3.062676116165145e-05, "loss": 0.0096, "step": 38720 }, { "grad_norm": 0.13021668791770935, "learning_rate": 3.06013589686441e-05, "loss": 0.011, "step": 38730 }, { "grad_norm": 0.07497522234916687, "learning_rate": 3.05759626684317e-05, "loss": 0.0101, "step": 38740 }, { "grad_norm": 0.10880602896213531, "learning_rate": 3.055057226872896e-05, "loss": 0.0101, "step": 38750 }, { "grad_norm": 0.10156495124101639, "learning_rate": 3.052518777724887e-05, "loss": 0.011, "step": 38760 }, { "grad_norm": 0.08869756758213043, "learning_rate": 3.04998092017025e-05, "loss": 0.0091, "step": 38770 }, { "grad_norm": 0.090314581990242, "learning_rate": 3.0474436549799246e-05, "loss": 0.0092, "step": 38780 }, { "grad_norm": 0.08099276572465897, "learning_rate": 3.044906982924661e-05, "loss": 0.0096, "step": 38790 }, { "grad_norm": 0.07965502142906189, "learning_rate": 3.0423709047750337e-05, "loss": 0.0091, "step": 38800 }, { "grad_norm": 0.08661014586687088, "learning_rate": 3.03983542130144e-05, "loss": 0.0088, "step": 38810 }, { "grad_norm": 0.07766523957252502, "learning_rate": 3.0373005332740877e-05, "loss": 0.0102, "step": 38820 }, { "grad_norm": 0.1880192905664444, "learning_rate": 3.034766241463013e-05, "loss": 0.0116, "step": 38830 }, { "grad_norm": 0.1170065850019455, "learning_rate": 3.032232546638064e-05, "loss": 0.0091, "step": 38840 }, { "grad_norm": 0.10240728408098221, "learning_rate": 3.0296994495689114e-05, "loss": 0.0106, "step": 38850 }, { "grad_norm": 0.07370693981647491, "learning_rate": 3.0271669510250444e-05, "loss": 0.009, "step": 38860 }, { "grad_norm": 0.09272375702857971, "learning_rate": 3.024635051775766e-05, "loss": 0.0119, "step": 38870 }, { "grad_norm": 0.08158188313245773, "learning_rate": 3.022103752590205e-05, "loss": 0.0095, "step": 38880 }, { "grad_norm": 0.11465699225664139, "learning_rate": 3.0195730542372992e-05, "loss": 0.0096, "step": 38890 }, { "grad_norm": 0.08561345934867859, "learning_rate": 3.0170429574858084e-05, "loss": 0.0097, "step": 38900 }, { "grad_norm": 0.08349812030792236, "learning_rate": 3.0145134631043127e-05, "loss": 0.0091, "step": 38910 }, { "grad_norm": 0.11100789159536362, "learning_rate": 3.0119845718612018e-05, "loss": 0.0091, "step": 38920 }, { "grad_norm": 0.08739574253559113, "learning_rate": 3.009456284524688e-05, "loss": 0.0103, "step": 38930 }, { "grad_norm": 0.08078927546739578, "learning_rate": 3.0069286018627967e-05, "loss": 0.0095, "step": 38940 }, { "grad_norm": 0.09932278841733932, "learning_rate": 3.0044015246433743e-05, "loss": 0.0089, "step": 38950 }, { "grad_norm": 0.13017672300338745, "learning_rate": 3.0018750536340755e-05, "loss": 0.0098, "step": 38960 }, { "grad_norm": 0.09361559897661209, "learning_rate": 2.999349189602378e-05, "loss": 0.0087, "step": 38970 }, { "grad_norm": 0.10115499794483185, "learning_rate": 2.9968239333155733e-05, "loss": 0.0089, "step": 38980 }, { "grad_norm": 0.09623198956251144, "learning_rate": 2.994299285540767e-05, "loss": 0.0107, "step": 38990 }, { "grad_norm": 0.09939143806695938, "learning_rate": 2.9917752470448813e-05, "loss": 0.0104, "step": 39000 }, { "grad_norm": 0.13682256639003754, "learning_rate": 2.9892518185946495e-05, "loss": 0.0087, "step": 39010 }, { "grad_norm": 0.11826064437627792, "learning_rate": 2.986729000956624e-05, "loss": 0.0103, "step": 39020 }, { "grad_norm": 0.09219695627689362, "learning_rate": 2.9842067948971736e-05, "loss": 0.0097, "step": 39030 }, { "grad_norm": 0.10773426294326782, "learning_rate": 2.9816852011824727e-05, "loss": 0.0094, "step": 39040 }, { "grad_norm": 0.11936911195516586, "learning_rate": 2.979164220578519e-05, "loss": 0.0095, "step": 39050 }, { "grad_norm": 0.09312200546264648, "learning_rate": 2.9766438538511165e-05, "loss": 0.0087, "step": 39060 }, { "grad_norm": 0.11061162501573563, "learning_rate": 2.9741241017658873e-05, "loss": 0.0095, "step": 39070 }, { "grad_norm": 0.08999381959438324, "learning_rate": 2.971604965088267e-05, "loss": 0.0084, "step": 39080 }, { "grad_norm": 0.15067310631275177, "learning_rate": 2.9690864445835008e-05, "loss": 0.0089, "step": 39090 }, { "grad_norm": 0.0777694582939148, "learning_rate": 2.966568541016651e-05, "loss": 0.0086, "step": 39100 }, { "grad_norm": 0.0928432047367096, "learning_rate": 2.9640512551525867e-05, "loss": 0.0091, "step": 39110 }, { "grad_norm": 0.10852544009685516, "learning_rate": 2.961534587755995e-05, "loss": 0.0097, "step": 39120 }, { "grad_norm": 0.10802886635065079, "learning_rate": 2.959018539591375e-05, "loss": 0.0086, "step": 39130 }, { "grad_norm": 0.10936875641345978, "learning_rate": 2.9565031114230325e-05, "loss": 0.01, "step": 39140 }, { "grad_norm": 0.11347989737987518, "learning_rate": 2.9539883040150895e-05, "loss": 0.01, "step": 39150 }, { "grad_norm": 0.1073213517665863, "learning_rate": 2.9514741181314774e-05, "loss": 0.0095, "step": 39160 }, { "grad_norm": 0.0882730707526207, "learning_rate": 2.94896055453594e-05, "loss": 0.0095, "step": 39170 }, { "grad_norm": 0.1290077418088913, "learning_rate": 2.9464476139920332e-05, "loss": 0.0103, "step": 39180 }, { "grad_norm": 0.10116493701934814, "learning_rate": 2.9439352972631186e-05, "loss": 0.0095, "step": 39190 }, { "grad_norm": 0.12761111557483673, "learning_rate": 2.9414236051123757e-05, "loss": 0.0093, "step": 39200 }, { "grad_norm": 0.08273224532604218, "learning_rate": 2.938912538302785e-05, "loss": 0.0102, "step": 39210 }, { "grad_norm": 0.08621333539485931, "learning_rate": 2.9364020975971464e-05, "loss": 0.0108, "step": 39220 }, { "grad_norm": 0.09572761505842209, "learning_rate": 2.9338922837580657e-05, "loss": 0.0095, "step": 39230 }, { "grad_norm": 0.1147368848323822, "learning_rate": 2.931383097547955e-05, "loss": 0.0086, "step": 39240 }, { "grad_norm": 0.11396279186010361, "learning_rate": 2.928874539729043e-05, "loss": 0.0097, "step": 39250 }, { "grad_norm": 0.10755399614572525, "learning_rate": 2.926366611063358e-05, "loss": 0.0101, "step": 39260 }, { "grad_norm": 0.12613223493099213, "learning_rate": 2.9238593123127463e-05, "loss": 0.0096, "step": 39270 }, { "grad_norm": 0.11785943061113358, "learning_rate": 2.9213526442388583e-05, "loss": 0.0087, "step": 39280 }, { "grad_norm": 0.10648606717586517, "learning_rate": 2.9188466076031545e-05, "loss": 0.0092, "step": 39290 }, { "grad_norm": 0.10324763506650925, "learning_rate": 2.9163412031669012e-05, "loss": 0.0087, "step": 39300 }, { "grad_norm": 0.09405351430177689, "learning_rate": 2.913836431691175e-05, "loss": 0.0126, "step": 39310 }, { "grad_norm": 0.13705521821975708, "learning_rate": 2.9113322939368583e-05, "loss": 0.0108, "step": 39320 }, { "grad_norm": 0.10655614733695984, "learning_rate": 2.9088287906646427e-05, "loss": 0.0099, "step": 39330 }, { "grad_norm": 0.09666675329208374, "learning_rate": 2.906325922635024e-05, "loss": 0.0105, "step": 39340 }, { "grad_norm": 0.1365690976381302, "learning_rate": 2.903823690608313e-05, "loss": 0.0102, "step": 39350 }, { "grad_norm": 0.07947803288698196, "learning_rate": 2.9013220953446174e-05, "loss": 0.0102, "step": 39360 }, { "grad_norm": 0.12085644900798798, "learning_rate": 2.8988211376038564e-05, "loss": 0.0112, "step": 39370 }, { "grad_norm": 0.1169067919254303, "learning_rate": 2.8963208181457564e-05, "loss": 0.01, "step": 39380 }, { "grad_norm": 0.11159297823905945, "learning_rate": 2.8938211377298453e-05, "loss": 0.0099, "step": 39390 }, { "grad_norm": 0.11641982942819595, "learning_rate": 2.8913220971154652e-05, "loss": 0.0094, "step": 39400 }, { "grad_norm": 0.13109464943408966, "learning_rate": 2.888823697061753e-05, "loss": 0.0103, "step": 39410 }, { "grad_norm": 0.13754840195178986, "learning_rate": 2.8863259383276618e-05, "loss": 0.0102, "step": 39420 }, { "grad_norm": 0.09212520718574524, "learning_rate": 2.8838288216719395e-05, "loss": 0.0088, "step": 39430 }, { "grad_norm": 0.08870722353458405, "learning_rate": 2.8813323478531484e-05, "loss": 0.0087, "step": 39440 }, { "grad_norm": 0.09964487701654434, "learning_rate": 2.8788365176296496e-05, "loss": 0.0104, "step": 39450 }, { "grad_norm": 0.10837205499410629, "learning_rate": 2.876341331759611e-05, "loss": 0.0098, "step": 39460 }, { "grad_norm": 0.10643362998962402, "learning_rate": 2.8738467910010036e-05, "loss": 0.0092, "step": 39470 }, { "grad_norm": 0.1611931025981903, "learning_rate": 2.8713528961116032e-05, "loss": 0.0092, "step": 39480 }, { "grad_norm": 0.08848592638969421, "learning_rate": 2.8688596478489875e-05, "loss": 0.0088, "step": 39490 }, { "grad_norm": 0.08059196919202805, "learning_rate": 2.8663670469705434e-05, "loss": 0.0089, "step": 39500 }, { "grad_norm": 0.09900093078613281, "learning_rate": 2.8638750942334546e-05, "loss": 0.0098, "step": 39510 }, { "grad_norm": 0.0869378000497818, "learning_rate": 2.8613837903947115e-05, "loss": 0.0085, "step": 39520 }, { "grad_norm": 0.11051120609045029, "learning_rate": 2.858893136211106e-05, "loss": 0.0099, "step": 39530 }, { "grad_norm": 0.12631767988204956, "learning_rate": 2.8564031324392315e-05, "loss": 0.0104, "step": 39540 }, { "grad_norm": 0.12248482555150986, "learning_rate": 2.85391377983549e-05, "loss": 0.0092, "step": 39550 }, { "grad_norm": 0.10966453701257706, "learning_rate": 2.851425079156075e-05, "loss": 0.0112, "step": 39560 }, { "grad_norm": 0.12280403822660446, "learning_rate": 2.848937031156994e-05, "loss": 0.0116, "step": 39570 }, { "grad_norm": 0.10288117825984955, "learning_rate": 2.846449636594044e-05, "loss": 0.009, "step": 39580 }, { "grad_norm": 0.07886363565921783, "learning_rate": 2.843962896222836e-05, "loss": 0.0086, "step": 39590 }, { "grad_norm": 0.11285851895809174, "learning_rate": 2.8414768107987722e-05, "loss": 0.0097, "step": 39600 }, { "grad_norm": 0.09158672392368317, "learning_rate": 2.838991381077061e-05, "loss": 0.0098, "step": 39610 }, { "grad_norm": 0.15867890417575836, "learning_rate": 2.83650660781271e-05, "loss": 0.0107, "step": 39620 }, { "grad_norm": 0.0855463296175003, "learning_rate": 2.8340224917605285e-05, "loss": 0.0085, "step": 39630 }, { "grad_norm": 0.11550524830818176, "learning_rate": 2.831539033675122e-05, "loss": 0.0096, "step": 39640 }, { "grad_norm": 0.1322818547487259, "learning_rate": 2.8290562343109038e-05, "loss": 0.0095, "step": 39650 }, { "grad_norm": 0.12006908655166626, "learning_rate": 2.826574094422082e-05, "loss": 0.0093, "step": 39660 }, { "grad_norm": 0.14159999787807465, "learning_rate": 2.8240926147626645e-05, "loss": 0.0089, "step": 39670 }, { "grad_norm": 0.09097032248973846, "learning_rate": 2.8216117960864586e-05, "loss": 0.0104, "step": 39680 }, { "grad_norm": 0.1277998387813568, "learning_rate": 2.8191316391470703e-05, "loss": 0.0122, "step": 39690 }, { "grad_norm": 0.13749685883522034, "learning_rate": 2.816652144697911e-05, "loss": 0.0099, "step": 39700 }, { "grad_norm": 0.10789791494607925, "learning_rate": 2.8141733134921783e-05, "loss": 0.0088, "step": 39710 }, { "grad_norm": 0.11971724033355713, "learning_rate": 2.811695146282884e-05, "loss": 0.0086, "step": 39720 }, { "grad_norm": 0.09221330285072327, "learning_rate": 2.8092176438228212e-05, "loss": 0.0096, "step": 39730 }, { "grad_norm": 0.10816189646720886, "learning_rate": 2.806740806864598e-05, "loss": 0.0096, "step": 39740 }, { "grad_norm": 0.07955048233270645, "learning_rate": 2.804264636160604e-05, "loss": 0.009, "step": 39750 }, { "grad_norm": 0.13376136124134064, "learning_rate": 2.8017891324630402e-05, "loss": 0.0091, "step": 39760 }, { "grad_norm": 0.1140747219324112, "learning_rate": 2.7993142965238976e-05, "loss": 0.0105, "step": 39770 }, { "grad_norm": 0.11699637025594711, "learning_rate": 2.7968401290949665e-05, "loss": 0.0095, "step": 39780 }, { "grad_norm": 0.12046948075294495, "learning_rate": 2.7943666309278328e-05, "loss": 0.0097, "step": 39790 }, { "grad_norm": 0.10431944578886032, "learning_rate": 2.7918938027738783e-05, "loss": 0.0101, "step": 39800 }, { "grad_norm": 0.08655036240816116, "learning_rate": 2.789421645384287e-05, "loss": 0.0082, "step": 39810 }, { "grad_norm": 0.11763322353363037, "learning_rate": 2.786950159510032e-05, "loss": 0.0132, "step": 39820 }, { "grad_norm": 0.12178865075111389, "learning_rate": 2.7844793459018876e-05, "loss": 0.0091, "step": 39830 }, { "grad_norm": 0.12443860620260239, "learning_rate": 2.7820092053104195e-05, "loss": 0.0092, "step": 39840 }, { "grad_norm": 0.10210840404033661, "learning_rate": 2.7795397384859933e-05, "loss": 0.0103, "step": 39850 }, { "grad_norm": 0.11504758894443512, "learning_rate": 2.7770709461787638e-05, "loss": 0.0089, "step": 39860 }, { "grad_norm": 0.08758552372455597, "learning_rate": 2.7746028291386915e-05, "loss": 0.0092, "step": 39870 }, { "grad_norm": 0.11168774962425232, "learning_rate": 2.772135388115519e-05, "loss": 0.0102, "step": 39880 }, { "grad_norm": 0.07416561990976334, "learning_rate": 2.7696686238587945e-05, "loss": 0.0093, "step": 39890 }, { "grad_norm": 0.09529265016317368, "learning_rate": 2.7672025371178505e-05, "loss": 0.0097, "step": 39900 }, { "grad_norm": 0.08872034400701523, "learning_rate": 2.7647371286418238e-05, "loss": 0.0093, "step": 39910 }, { "grad_norm": 0.11593221127986908, "learning_rate": 2.762272399179639e-05, "loss": 0.0089, "step": 39920 }, { "grad_norm": 0.08026319742202759, "learning_rate": 2.7598083494800154e-05, "loss": 0.0095, "step": 39930 }, { "grad_norm": 0.09726352989673615, "learning_rate": 2.7573449802914664e-05, "loss": 0.0084, "step": 39940 }, { "grad_norm": 0.1148471087217331, "learning_rate": 2.7548822923622964e-05, "loss": 0.0094, "step": 39950 }, { "grad_norm": 0.10463254898786545, "learning_rate": 2.752420286440609e-05, "loss": 0.0081, "step": 39960 }, { "grad_norm": 0.10407759249210358, "learning_rate": 2.749958963274295e-05, "loss": 0.0094, "step": 39970 }, { "grad_norm": 0.2061525285243988, "learning_rate": 2.747498323611039e-05, "loss": 0.0102, "step": 39980 }, { "grad_norm": 0.11649911105632782, "learning_rate": 2.7450383681983184e-05, "loss": 0.0094, "step": 39990 }, { "grad_norm": 0.12335991114377975, "learning_rate": 2.742579097783403e-05, "loss": 0.0106, "step": 40000 }, { "grad_norm": 0.12366156280040741, "learning_rate": 2.7401205131133512e-05, "loss": 0.0101, "step": 40010 }, { "grad_norm": 0.11140535771846771, "learning_rate": 2.7376626149350238e-05, "loss": 0.0091, "step": 40020 }, { "grad_norm": 0.13204988837242126, "learning_rate": 2.735205403995056e-05, "loss": 0.0108, "step": 40030 }, { "grad_norm": 0.09891951829195023, "learning_rate": 2.7327488810398917e-05, "loss": 0.0096, "step": 40040 }, { "grad_norm": 0.10252448916435242, "learning_rate": 2.7302930468157507e-05, "loss": 0.0098, "step": 40050 }, { "grad_norm": 0.11066973209381104, "learning_rate": 2.727837902068655e-05, "loss": 0.0094, "step": 40060 }, { "grad_norm": 0.10116082429885864, "learning_rate": 2.7253834475444123e-05, "loss": 0.0092, "step": 40070 }, { "grad_norm": 0.1166367307305336, "learning_rate": 2.7229296839886204e-05, "loss": 0.0121, "step": 40080 }, { "grad_norm": 0.10995452851057053, "learning_rate": 2.720476612146668e-05, "loss": 0.0095, "step": 40090 }, { "grad_norm": 0.13732461631298065, "learning_rate": 2.7180242327637317e-05, "loss": 0.0104, "step": 40100 }, { "grad_norm": 0.09715225547552109, "learning_rate": 2.7155725465847826e-05, "loss": 0.009, "step": 40110 }, { "grad_norm": 0.08621211349964142, "learning_rate": 2.713121554354578e-05, "loss": 0.0088, "step": 40120 }, { "grad_norm": 0.07965235412120819, "learning_rate": 2.7106712568176628e-05, "loss": 0.0095, "step": 40130 }, { "grad_norm": 0.11461851745843887, "learning_rate": 2.708221654718374e-05, "loss": 0.0089, "step": 40140 }, { "grad_norm": 0.10814819484949112, "learning_rate": 2.7057727488008357e-05, "loss": 0.0096, "step": 40150 }, { "grad_norm": 0.11252165585756302, "learning_rate": 2.703324539808961e-05, "loss": 0.01, "step": 40160 }, { "grad_norm": 0.09988726675510406, "learning_rate": 2.7008770284864505e-05, "loss": 0.0108, "step": 40170 }, { "grad_norm": 0.09268272668123245, "learning_rate": 2.6984302155767916e-05, "loss": 0.007, "step": 40180 }, { "grad_norm": 0.06928343325853348, "learning_rate": 2.6959841018232683e-05, "loss": 0.0093, "step": 40190 }, { "grad_norm": 0.10550318658351898, "learning_rate": 2.693538687968937e-05, "loss": 0.0088, "step": 40200 }, { "grad_norm": 0.10325470566749573, "learning_rate": 2.6910939747566556e-05, "loss": 0.0096, "step": 40210 }, { "grad_norm": 0.06944762170314789, "learning_rate": 2.6886499629290607e-05, "loss": 0.0091, "step": 40220 }, { "grad_norm": 0.10070493072271347, "learning_rate": 2.6862066532285802e-05, "loss": 0.0086, "step": 40230 }, { "grad_norm": 0.11219349503517151, "learning_rate": 2.6837640463974262e-05, "loss": 0.0109, "step": 40240 }, { "grad_norm": 0.07843859493732452, "learning_rate": 2.681322143177596e-05, "loss": 0.0106, "step": 40250 }, { "grad_norm": 0.0952722504734993, "learning_rate": 2.678880944310882e-05, "loss": 0.0088, "step": 40260 }, { "grad_norm": 0.09705138206481934, "learning_rate": 2.6764404505388474e-05, "loss": 0.0082, "step": 40270 }, { "grad_norm": 0.1133960410952568, "learning_rate": 2.6740006626028558e-05, "loss": 0.0087, "step": 40280 }, { "grad_norm": 0.10674937814474106, "learning_rate": 2.671561581244048e-05, "loss": 0.0104, "step": 40290 }, { "grad_norm": 0.11312322318553925, "learning_rate": 2.6691232072033536e-05, "loss": 0.0106, "step": 40300 }, { "grad_norm": 0.1010110154747963, "learning_rate": 2.6666855412214852e-05, "loss": 0.0101, "step": 40310 }, { "grad_norm": 0.10276558250188828, "learning_rate": 2.664248584038942e-05, "loss": 0.0087, "step": 40320 }, { "grad_norm": 0.09104864299297333, "learning_rate": 2.6618123363960047e-05, "loss": 0.0087, "step": 40330 }, { "grad_norm": 0.0875755175948143, "learning_rate": 2.659376799032748e-05, "loss": 0.0094, "step": 40340 }, { "grad_norm": 0.08979669958353043, "learning_rate": 2.6569419726890145e-05, "loss": 0.0087, "step": 40350 }, { "grad_norm": 0.12276755273342133, "learning_rate": 2.654507858104447e-05, "loss": 0.0094, "step": 40360 }, { "grad_norm": 0.13116762042045593, "learning_rate": 2.652074456018463e-05, "loss": 0.0103, "step": 40370 }, { "grad_norm": 0.10247345268726349, "learning_rate": 2.6496417671702646e-05, "loss": 0.0088, "step": 40380 }, { "grad_norm": 0.07501775771379471, "learning_rate": 2.6472097922988427e-05, "loss": 0.0093, "step": 40390 }, { "grad_norm": 0.08576138317584991, "learning_rate": 2.6447785321429607e-05, "loss": 0.0089, "step": 40400 }, { "grad_norm": 0.09066689014434814, "learning_rate": 2.6423479874411784e-05, "loss": 0.0099, "step": 40410 }, { "grad_norm": 0.1011170744895935, "learning_rate": 2.6399181589318234e-05, "loss": 0.0092, "step": 40420 }, { "grad_norm": 0.08231613039970398, "learning_rate": 2.6374890473530188e-05, "loss": 0.0091, "step": 40430 }, { "grad_norm": 0.12963344156742096, "learning_rate": 2.635060653442664e-05, "loss": 0.0094, "step": 40440 }, { "grad_norm": 0.12517762184143066, "learning_rate": 2.6326329779384395e-05, "loss": 0.0081, "step": 40450 }, { "grad_norm": 0.0931471660733223, "learning_rate": 2.63020602157781e-05, "loss": 0.0085, "step": 40460 }, { "grad_norm": 0.126634418964386, "learning_rate": 2.62777978509802e-05, "loss": 0.0093, "step": 40470 }, { "grad_norm": 0.10217034071683884, "learning_rate": 2.6253542692360954e-05, "loss": 0.0092, "step": 40480 }, { "grad_norm": 0.08596835285425186, "learning_rate": 2.6229294747288458e-05, "loss": 0.0095, "step": 40490 }, { "grad_norm": 0.08828848600387573, "learning_rate": 2.6205054023128596e-05, "loss": 0.0091, "step": 40500 }, { "grad_norm": 0.10089730471372604, "learning_rate": 2.6180820527245043e-05, "loss": 0.0091, "step": 40510 }, { "grad_norm": 0.09610413759946823, "learning_rate": 2.6156594266999313e-05, "loss": 0.0087, "step": 40520 }, { "grad_norm": 0.13520556688308716, "learning_rate": 2.6132375249750672e-05, "loss": 0.0087, "step": 40530 }, { "grad_norm": 0.10603528469800949, "learning_rate": 2.6108163482856286e-05, "loss": 0.0097, "step": 40540 }, { "grad_norm": 0.10328272730112076, "learning_rate": 2.6083958973670964e-05, "loss": 0.0082, "step": 40550 }, { "grad_norm": 0.0993652418255806, "learning_rate": 2.6059761729547483e-05, "loss": 0.0096, "step": 40560 }, { "grad_norm": 0.09908875077962875, "learning_rate": 2.603557175783624e-05, "loss": 0.009, "step": 40570 }, { "grad_norm": 0.09844764322042465, "learning_rate": 2.601138906588559e-05, "loss": 0.0086, "step": 40580 }, { "grad_norm": 0.0796075165271759, "learning_rate": 2.598721366104152e-05, "loss": 0.009, "step": 40590 }, { "grad_norm": 0.12595996260643005, "learning_rate": 2.5963045550647945e-05, "loss": 0.0096, "step": 40600 }, { "grad_norm": 0.09758905321359634, "learning_rate": 2.5938884742046466e-05, "loss": 0.0101, "step": 40610 }, { "grad_norm": 0.09375780075788498, "learning_rate": 2.5914731242576507e-05, "loss": 0.0096, "step": 40620 }, { "grad_norm": 0.09567918628454208, "learning_rate": 2.5890585059575268e-05, "loss": 0.0084, "step": 40630 }, { "grad_norm": 0.09650884568691254, "learning_rate": 2.5866446200377688e-05, "loss": 0.0088, "step": 40640 }, { "grad_norm": 0.1033874973654747, "learning_rate": 2.5842314672316566e-05, "loss": 0.0083, "step": 40650 }, { "grad_norm": 0.09166180342435837, "learning_rate": 2.581819048272239e-05, "loss": 0.01, "step": 40660 }, { "grad_norm": 0.1279258131980896, "learning_rate": 2.5794073638923478e-05, "loss": 0.0078, "step": 40670 }, { "grad_norm": 0.10286806523799896, "learning_rate": 2.576996414824586e-05, "loss": 0.009, "step": 40680 }, { "grad_norm": 0.09418420493602753, "learning_rate": 2.574586201801339e-05, "loss": 0.0088, "step": 40690 }, { "grad_norm": 0.10908480733633041, "learning_rate": 2.572176725554762e-05, "loss": 0.009, "step": 40700 }, { "grad_norm": 0.07426350563764572, "learning_rate": 2.5697679868167966e-05, "loss": 0.0095, "step": 40710 }, { "grad_norm": 0.11275395005941391, "learning_rate": 2.5673599863191468e-05, "loss": 0.0092, "step": 40720 }, { "grad_norm": 0.08083369582891464, "learning_rate": 2.564952724793306e-05, "loss": 0.01, "step": 40730 }, { "grad_norm": 0.09079216420650482, "learning_rate": 2.5625462029705306e-05, "loss": 0.0094, "step": 40740 }, { "grad_norm": 0.0913400799036026, "learning_rate": 2.5601404215818624e-05, "loss": 0.0082, "step": 40750 }, { "grad_norm": 0.1127767264842987, "learning_rate": 2.5577353813581144e-05, "loss": 0.0101, "step": 40760 }, { "grad_norm": 0.1247892901301384, "learning_rate": 2.5553310830298733e-05, "loss": 0.0092, "step": 40770 }, { "grad_norm": 0.08737004548311234, "learning_rate": 2.5529275273275012e-05, "loss": 0.0082, "step": 40780 }, { "grad_norm": 0.093137726187706, "learning_rate": 2.550524714981133e-05, "loss": 0.0089, "step": 40790 }, { "grad_norm": 0.23635566234588623, "learning_rate": 2.5481226467206837e-05, "loss": 0.0087, "step": 40800 }, { "grad_norm": 0.10633683949708939, "learning_rate": 2.5457213232758365e-05, "loss": 0.009, "step": 40810 }, { "grad_norm": 0.1128067746758461, "learning_rate": 2.5433207453760498e-05, "loss": 0.01, "step": 40820 }, { "grad_norm": 0.07951171696186066, "learning_rate": 2.5409209137505552e-05, "loss": 0.0091, "step": 40830 }, { "grad_norm": 0.0871494710445404, "learning_rate": 2.5385218291283597e-05, "loss": 0.0089, "step": 40840 }, { "grad_norm": 0.12579388916492462, "learning_rate": 2.5361234922382383e-05, "loss": 0.0091, "step": 40850 }, { "grad_norm": 0.1259220540523529, "learning_rate": 2.533725903808749e-05, "loss": 0.0093, "step": 40860 }, { "grad_norm": 0.09113803505897522, "learning_rate": 2.5313290645682085e-05, "loss": 0.0094, "step": 40870 }, { "grad_norm": 0.10588453710079193, "learning_rate": 2.52893297524472e-05, "loss": 0.0106, "step": 40880 }, { "grad_norm": 0.09576789289712906, "learning_rate": 2.526537636566145e-05, "loss": 0.0091, "step": 40890 }, { "grad_norm": 0.08128255605697632, "learning_rate": 2.5241430492601305e-05, "loss": 0.0091, "step": 40900 }, { "grad_norm": 0.14229537546634674, "learning_rate": 2.5217492140540867e-05, "loss": 0.0093, "step": 40910 }, { "grad_norm": 0.10810411721467972, "learning_rate": 2.5193561316751967e-05, "loss": 0.0078, "step": 40920 }, { "grad_norm": 0.09866168349981308, "learning_rate": 2.516963802850416e-05, "loss": 0.0089, "step": 40930 }, { "grad_norm": 0.10833265632390976, "learning_rate": 2.5145722283064698e-05, "loss": 0.0091, "step": 40940 }, { "grad_norm": 0.1466643214225769, "learning_rate": 2.5121814087698602e-05, "loss": 0.0103, "step": 40950 }, { "grad_norm": 0.0906122624874115, "learning_rate": 2.509791344966848e-05, "loss": 0.009, "step": 40960 }, { "grad_norm": 0.08739256858825684, "learning_rate": 2.5074020376234768e-05, "loss": 0.0095, "step": 40970 }, { "grad_norm": 0.12458745390176773, "learning_rate": 2.5050134874655534e-05, "loss": 0.0094, "step": 40980 }, { "grad_norm": 0.12266471236944199, "learning_rate": 2.5026256952186566e-05, "loss": 0.009, "step": 40990 }, { "grad_norm": 0.0821719691157341, "learning_rate": 2.5002386616081335e-05, "loss": 0.0087, "step": 41000 }, { "grad_norm": 0.13080893456935883, "learning_rate": 2.497852387359103e-05, "loss": 0.0085, "step": 41010 }, { "grad_norm": 0.08872547000646591, "learning_rate": 2.4954668731964496e-05, "loss": 0.0091, "step": 41020 }, { "grad_norm": 0.21346019208431244, "learning_rate": 2.4930821198448364e-05, "loss": 0.0086, "step": 41030 }, { "grad_norm": 0.08049291372299194, "learning_rate": 2.4906981280286796e-05, "loss": 0.01, "step": 41040 }, { "grad_norm": 0.11499039828777313, "learning_rate": 2.488314898472179e-05, "loss": 0.0099, "step": 41050 }, { "grad_norm": 0.1308450698852539, "learning_rate": 2.485932431899295e-05, "loss": 0.0083, "step": 41060 }, { "grad_norm": 0.14047996699810028, "learning_rate": 2.4835507290337584e-05, "loss": 0.0085, "step": 41070 }, { "grad_norm": 0.10752282291650772, "learning_rate": 2.4811697905990672e-05, "loss": 0.0088, "step": 41080 }, { "grad_norm": 0.0933104008436203, "learning_rate": 2.4787896173184854e-05, "loss": 0.0086, "step": 41090 }, { "grad_norm": 0.11202853173017502, "learning_rate": 2.4764102099150534e-05, "loss": 0.0091, "step": 41100 }, { "grad_norm": 0.15104462206363678, "learning_rate": 2.4740315691115644e-05, "loss": 0.0107, "step": 41110 }, { "grad_norm": 0.14951327443122864, "learning_rate": 2.4716536956305918e-05, "loss": 0.0107, "step": 41120 }, { "grad_norm": 0.09664437919855118, "learning_rate": 2.4692765901944697e-05, "loss": 0.0087, "step": 41130 }, { "grad_norm": 0.07977304607629776, "learning_rate": 2.4669002535253e-05, "loss": 0.0092, "step": 41140 }, { "grad_norm": 0.07377739250659943, "learning_rate": 2.46452468634495e-05, "loss": 0.0082, "step": 41150 }, { "grad_norm": 0.1141965389251709, "learning_rate": 2.462149889375055e-05, "loss": 0.009, "step": 41160 }, { "grad_norm": 0.09995610266923904, "learning_rate": 2.459775863337014e-05, "loss": 0.0086, "step": 41170 }, { "grad_norm": 0.12827236950397491, "learning_rate": 2.4574026089519985e-05, "loss": 0.0106, "step": 41180 }, { "grad_norm": 0.07345187664031982, "learning_rate": 2.4550301269409333e-05, "loss": 0.0096, "step": 41190 }, { "grad_norm": 0.10183284431695938, "learning_rate": 2.4526584180245216e-05, "loss": 0.0087, "step": 41200 }, { "grad_norm": 0.0984855368733406, "learning_rate": 2.4502874829232236e-05, "loss": 0.0088, "step": 41210 }, { "grad_norm": 0.12944814562797546, "learning_rate": 2.447917322357267e-05, "loss": 0.009, "step": 41220 }, { "grad_norm": 0.10977257788181305, "learning_rate": 2.4455479370466443e-05, "loss": 0.0081, "step": 41230 }, { "grad_norm": 0.0937311127781868, "learning_rate": 2.4431793277111097e-05, "loss": 0.0088, "step": 41240 }, { "grad_norm": 0.08251911401748657, "learning_rate": 2.4408114950701905e-05, "loss": 0.0082, "step": 41250 }, { "grad_norm": 0.0931416004896164, "learning_rate": 2.4384444398431634e-05, "loss": 0.0094, "step": 41260 }, { "grad_norm": 0.08001729100942612, "learning_rate": 2.4360781627490837e-05, "loss": 0.009, "step": 41270 }, { "grad_norm": 0.08801668137311935, "learning_rate": 2.433712664506762e-05, "loss": 0.0086, "step": 41280 }, { "grad_norm": 0.09570646286010742, "learning_rate": 2.431347945834774e-05, "loss": 0.0091, "step": 41290 }, { "grad_norm": 0.12819042801856995, "learning_rate": 2.428984007451458e-05, "loss": 0.0099, "step": 41300 }, { "grad_norm": 0.10662689059972763, "learning_rate": 2.426620850074917e-05, "loss": 0.0084, "step": 41310 }, { "grad_norm": 0.12344709783792496, "learning_rate": 2.424258474423014e-05, "loss": 0.0101, "step": 41320 }, { "grad_norm": 0.07664330303668976, "learning_rate": 2.421896881213382e-05, "loss": 0.0098, "step": 41330 }, { "grad_norm": 0.09105949848890305, "learning_rate": 2.419536071163402e-05, "loss": 0.0086, "step": 41340 }, { "grad_norm": 0.09320907294750214, "learning_rate": 2.417176044990233e-05, "loss": 0.0095, "step": 41350 }, { "grad_norm": 0.11308491230010986, "learning_rate": 2.4148168034107855e-05, "loss": 0.0088, "step": 41360 }, { "grad_norm": 0.1139211654663086, "learning_rate": 2.4124583471417355e-05, "loss": 0.0096, "step": 41370 }, { "grad_norm": 0.11721143126487732, "learning_rate": 2.41010067689952e-05, "loss": 0.0088, "step": 41380 }, { "grad_norm": 0.11819519102573395, "learning_rate": 2.4077437934003338e-05, "loss": 0.0091, "step": 41390 }, { "grad_norm": 0.08837316930294037, "learning_rate": 2.405387697360143e-05, "loss": 0.0103, "step": 41400 }, { "grad_norm": 0.11435148119926453, "learning_rate": 2.4030323894946595e-05, "loss": 0.0093, "step": 41410 }, { "grad_norm": 0.07483425736427307, "learning_rate": 2.40067787051937e-05, "loss": 0.0085, "step": 41420 }, { "grad_norm": 0.1145155057311058, "learning_rate": 2.3983241411495087e-05, "loss": 0.0093, "step": 41430 }, { "grad_norm": 0.14636245369911194, "learning_rate": 2.3959712021000823e-05, "loss": 0.0093, "step": 41440 }, { "grad_norm": 0.098744235932827, "learning_rate": 2.3936190540858495e-05, "loss": 0.0089, "step": 41450 }, { "grad_norm": 0.10887198895215988, "learning_rate": 2.39126769782133e-05, "loss": 0.0078, "step": 41460 }, { "grad_norm": 0.11592046171426773, "learning_rate": 2.388917134020805e-05, "loss": 0.0093, "step": 41470 }, { "grad_norm": 0.10318901389837265, "learning_rate": 2.3865673633983128e-05, "loss": 0.0088, "step": 41480 }, { "grad_norm": 0.07596325129270554, "learning_rate": 2.3842183866676492e-05, "loss": 0.0094, "step": 41490 }, { "grad_norm": 0.14976325631141663, "learning_rate": 2.381870204542377e-05, "loss": 0.0083, "step": 41500 }, { "grad_norm": 0.12330807000398636, "learning_rate": 2.379522817735808e-05, "loss": 0.009, "step": 41510 }, { "grad_norm": 0.1160857081413269, "learning_rate": 2.377176226961018e-05, "loss": 0.0089, "step": 41520 }, { "grad_norm": 0.09847122430801392, "learning_rate": 2.3748304329308384e-05, "loss": 0.0085, "step": 41530 }, { "grad_norm": 0.12052367627620697, "learning_rate": 2.372485436357858e-05, "loss": 0.0085, "step": 41540 }, { "grad_norm": 0.10461955517530441, "learning_rate": 2.3701412379544296e-05, "loss": 0.009, "step": 41550 }, { "grad_norm": 0.07156571000814438, "learning_rate": 2.367797838432653e-05, "loss": 0.0097, "step": 41560 }, { "grad_norm": 0.0758877694606781, "learning_rate": 2.3654552385043967e-05, "loss": 0.0085, "step": 41570 }, { "grad_norm": 0.11131278425455093, "learning_rate": 2.3631134388812742e-05, "loss": 0.0088, "step": 41580 }, { "grad_norm": 0.11446335166692734, "learning_rate": 2.3607724402746684e-05, "loss": 0.0088, "step": 41590 }, { "grad_norm": 0.10288359224796295, "learning_rate": 2.35843224339571e-05, "loss": 0.0084, "step": 41600 }, { "grad_norm": 0.12310170382261276, "learning_rate": 2.3560928489552897e-05, "loss": 0.0082, "step": 41610 }, { "grad_norm": 0.08165334165096283, "learning_rate": 2.353754257664053e-05, "loss": 0.0099, "step": 41620 }, { "grad_norm": 0.09426391124725342, "learning_rate": 2.3514164702324037e-05, "loss": 0.0092, "step": 41630 }, { "grad_norm": 0.1185532808303833, "learning_rate": 2.3490794873704963e-05, "loss": 0.0093, "step": 41640 }, { "grad_norm": 0.09561561793088913, "learning_rate": 2.3467433097882496e-05, "loss": 0.0082, "step": 41650 }, { "grad_norm": 0.08419639617204666, "learning_rate": 2.34440793819533e-05, "loss": 0.0093, "step": 41660 }, { "grad_norm": 0.10638894140720367, "learning_rate": 2.3420733733011617e-05, "loss": 0.0089, "step": 41670 }, { "grad_norm": 0.077288918197155, "learning_rate": 2.3397396158149243e-05, "loss": 0.0088, "step": 41680 }, { "grad_norm": 0.08084065467119217, "learning_rate": 2.3374066664455498e-05, "loss": 0.0089, "step": 41690 }, { "grad_norm": 0.10650511831045151, "learning_rate": 2.3350745259017315e-05, "loss": 0.0084, "step": 41700 }, { "grad_norm": 0.08738865703344345, "learning_rate": 2.332743194891906e-05, "loss": 0.0086, "step": 41710 }, { "grad_norm": 0.10172358900308609, "learning_rate": 2.330412674124276e-05, "loss": 0.0087, "step": 41720 }, { "grad_norm": 0.11454646289348602, "learning_rate": 2.328082964306786e-05, "loss": 0.0093, "step": 41730 }, { "grad_norm": 0.07396628707647324, "learning_rate": 2.325754066147145e-05, "loss": 0.0085, "step": 41740 }, { "grad_norm": 0.13184583187103271, "learning_rate": 2.32342598035281e-05, "loss": 0.0081, "step": 41750 }, { "grad_norm": 0.12617342174053192, "learning_rate": 2.321098707630991e-05, "loss": 0.009, "step": 41760 }, { "grad_norm": 0.10888636112213135, "learning_rate": 2.318772248688652e-05, "loss": 0.0092, "step": 41770 }, { "grad_norm": 0.14922916889190674, "learning_rate": 2.3164466042325107e-05, "loss": 0.01, "step": 41780 }, { "grad_norm": 0.1041097342967987, "learning_rate": 2.3141217749690353e-05, "loss": 0.0094, "step": 41790 }, { "grad_norm": 0.11583396047353745, "learning_rate": 2.3117977616044466e-05, "loss": 0.009, "step": 41800 }, { "grad_norm": 0.10113994032144547, "learning_rate": 2.309474564844722e-05, "loss": 0.0084, "step": 41810 }, { "grad_norm": 0.10394023358821869, "learning_rate": 2.307152185395585e-05, "loss": 0.0084, "step": 41820 }, { "grad_norm": 0.12701134383678436, "learning_rate": 2.3048306239625144e-05, "loss": 0.0095, "step": 41830 }, { "grad_norm": 0.09401512891054153, "learning_rate": 2.3025098812507378e-05, "loss": 0.0103, "step": 41840 }, { "grad_norm": 0.09068065136671066, "learning_rate": 2.3001899579652366e-05, "loss": 0.0101, "step": 41850 }, { "grad_norm": 0.1278364211320877, "learning_rate": 2.2978708548107393e-05, "loss": 0.0088, "step": 41860 }, { "grad_norm": 0.12303652614355087, "learning_rate": 2.2955525724917348e-05, "loss": 0.0083, "step": 41870 }, { "grad_norm": 0.13119786977767944, "learning_rate": 2.2932351117124477e-05, "loss": 0.0091, "step": 41880 }, { "grad_norm": 0.11706846952438354, "learning_rate": 2.29091847317687e-05, "loss": 0.009, "step": 41890 }, { "grad_norm": 0.1150641217827797, "learning_rate": 2.2886026575887277e-05, "loss": 0.0094, "step": 41900 }, { "grad_norm": 0.07534681260585785, "learning_rate": 2.2862876656515094e-05, "loss": 0.0088, "step": 41910 }, { "grad_norm": 0.12495265156030655, "learning_rate": 2.2839734980684464e-05, "loss": 0.0089, "step": 41920 }, { "grad_norm": 0.09464380145072937, "learning_rate": 2.281660155542522e-05, "loss": 0.0092, "step": 41930 }, { "grad_norm": 0.10924342274665833, "learning_rate": 2.279347638776469e-05, "loss": 0.0093, "step": 41940 }, { "grad_norm": 0.10180512815713882, "learning_rate": 2.2770359484727665e-05, "loss": 0.0098, "step": 41950 }, { "grad_norm": 0.08978216350078583, "learning_rate": 2.27472508533365e-05, "loss": 0.0089, "step": 41960 }, { "grad_norm": 0.09733512252569199, "learning_rate": 2.2724150500610948e-05, "loss": 0.0104, "step": 41970 }, { "grad_norm": 0.09229733049869537, "learning_rate": 2.2701058433568302e-05, "loss": 0.0085, "step": 41980 }, { "grad_norm": 0.07925434410572052, "learning_rate": 2.2677974659223318e-05, "loss": 0.0084, "step": 41990 }, { "grad_norm": 0.09955676645040512, "learning_rate": 2.2654899184588235e-05, "loss": 0.008, "step": 42000 }, { "grad_norm": 0.09409314393997192, "learning_rate": 2.2631832016672756e-05, "loss": 0.0092, "step": 42010 }, { "grad_norm": 0.09501435607671738, "learning_rate": 2.2608773162484127e-05, "loss": 0.0094, "step": 42020 }, { "grad_norm": 0.08888161927461624, "learning_rate": 2.2585722629026958e-05, "loss": 0.0075, "step": 42030 }, { "grad_norm": 0.09542372077703476, "learning_rate": 2.2562680423303457e-05, "loss": 0.0089, "step": 42040 }, { "grad_norm": 0.077475905418396, "learning_rate": 2.2539646552313165e-05, "loss": 0.0089, "step": 42050 }, { "grad_norm": 0.10310214012861252, "learning_rate": 2.251662102305322e-05, "loss": 0.0101, "step": 42060 }, { "grad_norm": 0.11605220288038254, "learning_rate": 2.2493603842518152e-05, "loss": 0.0115, "step": 42070 }, { "grad_norm": 0.11864302307367325, "learning_rate": 2.2470595017699974e-05, "loss": 0.0087, "step": 42080 }, { "grad_norm": 0.09197449684143066, "learning_rate": 2.244759455558816e-05, "loss": 0.0081, "step": 42090 }, { "grad_norm": 0.08028306066989899, "learning_rate": 2.2424602463169614e-05, "loss": 0.0071, "step": 42100 }, { "grad_norm": 0.10098932683467865, "learning_rate": 2.2401618747428776e-05, "loss": 0.0101, "step": 42110 }, { "grad_norm": 0.1156972125172615, "learning_rate": 2.237864341534747e-05, "loss": 0.0079, "step": 42120 }, { "grad_norm": 0.09495553374290466, "learning_rate": 2.2355676473904998e-05, "loss": 0.0081, "step": 42130 }, { "grad_norm": 0.08725311607122421, "learning_rate": 2.2332717930078108e-05, "loss": 0.0086, "step": 42140 }, { "grad_norm": 0.17635317146778107, "learning_rate": 2.2309767790840992e-05, "loss": 0.0089, "step": 42150 }, { "grad_norm": 0.12068769335746765, "learning_rate": 2.228682606316529e-05, "loss": 0.0089, "step": 42160 }, { "grad_norm": 0.09883856028318405, "learning_rate": 2.2263892754020138e-05, "loss": 0.0099, "step": 42170 }, { "grad_norm": 0.08691506087779999, "learning_rate": 2.2240967870372004e-05, "loss": 0.0086, "step": 42180 }, { "grad_norm": 0.08862181007862091, "learning_rate": 2.2218051419184933e-05, "loss": 0.0082, "step": 42190 }, { "grad_norm": 0.09101497381925583, "learning_rate": 2.219514340742026e-05, "loss": 0.0084, "step": 42200 }, { "grad_norm": 0.15790918469429016, "learning_rate": 2.2172243842036898e-05, "loss": 0.0086, "step": 42210 }, { "grad_norm": 0.11794659495353699, "learning_rate": 2.2149352729991107e-05, "loss": 0.0086, "step": 42220 }, { "grad_norm": 0.09789764881134033, "learning_rate": 2.2126470078236605e-05, "loss": 0.0083, "step": 42230 }, { "grad_norm": 0.08221873641014099, "learning_rate": 2.2103595893724533e-05, "loss": 0.0087, "step": 42240 }, { "grad_norm": 0.1081944927573204, "learning_rate": 2.208073018340345e-05, "loss": 0.0082, "step": 42250 }, { "grad_norm": 0.11023375391960144, "learning_rate": 2.2057872954219405e-05, "loss": 0.0103, "step": 42260 }, { "grad_norm": 0.06730427592992783, "learning_rate": 2.203502421311575e-05, "loss": 0.0084, "step": 42270 }, { "grad_norm": 0.13992400467395782, "learning_rate": 2.2012183967033388e-05, "loss": 0.0085, "step": 42280 }, { "grad_norm": 0.08761804550886154, "learning_rate": 2.198935222291056e-05, "loss": 0.0091, "step": 42290 }, { "grad_norm": 0.14950822293758392, "learning_rate": 2.1966528987682948e-05, "loss": 0.0081, "step": 42300 }, { "grad_norm": 0.11099277436733246, "learning_rate": 2.194371426828365e-05, "loss": 0.0091, "step": 42310 }, { "grad_norm": 0.1112222820520401, "learning_rate": 2.192090807164317e-05, "loss": 0.0097, "step": 42320 }, { "grad_norm": 0.1008671298623085, "learning_rate": 2.1898110404689422e-05, "loss": 0.0097, "step": 42330 }, { "grad_norm": 0.08303642272949219, "learning_rate": 2.1875321274347776e-05, "loss": 0.0081, "step": 42340 }, { "grad_norm": 0.1291600614786148, "learning_rate": 2.18525406875409e-05, "loss": 0.0085, "step": 42350 }, { "grad_norm": 0.09161799401044846, "learning_rate": 2.1829768651188997e-05, "loss": 0.0081, "step": 42360 }, { "grad_norm": 0.16654761135578156, "learning_rate": 2.180700517220958e-05, "loss": 0.0115, "step": 42370 }, { "grad_norm": 0.09701117128133774, "learning_rate": 2.1784250257517603e-05, "loss": 0.009, "step": 42380 }, { "grad_norm": 0.09634874761104584, "learning_rate": 2.1761503914025406e-05, "loss": 0.0089, "step": 42390 }, { "grad_norm": 0.10380770266056061, "learning_rate": 2.1738766148642705e-05, "loss": 0.0085, "step": 42400 }, { "grad_norm": 0.11648375540971756, "learning_rate": 2.1716036968276683e-05, "loss": 0.0092, "step": 42410 }, { "grad_norm": 0.10670964419841766, "learning_rate": 2.1693316379831808e-05, "loss": 0.0097, "step": 42420 }, { "grad_norm": 0.1141318529844284, "learning_rate": 2.1670604390210037e-05, "loss": 0.0104, "step": 42430 }, { "grad_norm": 0.1013164296746254, "learning_rate": 2.1647901006310656e-05, "loss": 0.01, "step": 42440 }, { "grad_norm": 0.09619897603988647, "learning_rate": 2.1625206235030353e-05, "loss": 0.0095, "step": 42450 }, { "grad_norm": 0.08157110214233398, "learning_rate": 2.160252008326321e-05, "loss": 0.0082, "step": 42460 }, { "grad_norm": 0.08699237555265427, "learning_rate": 2.157984255790067e-05, "loss": 0.009, "step": 42470 }, { "grad_norm": 0.14115916192531586, "learning_rate": 2.1557173665831553e-05, "loss": 0.0084, "step": 42480 }, { "grad_norm": 0.11124745011329651, "learning_rate": 2.153451341394212e-05, "loss": 0.0092, "step": 42490 }, { "grad_norm": 0.09206725656986237, "learning_rate": 2.151186180911589e-05, "loss": 0.0092, "step": 42500 }, { "grad_norm": 0.11385335773229599, "learning_rate": 2.1489218858233877e-05, "loss": 0.0079, "step": 42510 }, { "grad_norm": 0.09017297625541687, "learning_rate": 2.1466584568174392e-05, "loss": 0.0097, "step": 42520 }, { "grad_norm": 0.1247175857424736, "learning_rate": 2.1443958945813132e-05, "loss": 0.0091, "step": 42530 }, { "grad_norm": 0.09715007990598679, "learning_rate": 2.1421341998023163e-05, "loss": 0.0089, "step": 42540 }, { "grad_norm": 0.10581659525632858, "learning_rate": 2.139873373167491e-05, "loss": 0.0092, "step": 42550 }, { "grad_norm": 0.12695546448230743, "learning_rate": 2.13761341536362e-05, "loss": 0.0089, "step": 42560 }, { "grad_norm": 0.09334596246480942, "learning_rate": 2.1353543270772136e-05, "loss": 0.0083, "step": 42570 }, { "grad_norm": 0.10757017880678177, "learning_rate": 2.1330961089945297e-05, "loss": 0.0083, "step": 42580 }, { "grad_norm": 0.09450910985469818, "learning_rate": 2.130838761801548e-05, "loss": 0.0088, "step": 42590 }, { "grad_norm": 0.10367206484079361, "learning_rate": 2.1285822861839966e-05, "loss": 0.0086, "step": 42600 }, { "grad_norm": 0.08027610927820206, "learning_rate": 2.126326682827331e-05, "loss": 0.0086, "step": 42610 }, { "grad_norm": 0.07407664507627487, "learning_rate": 2.124071952416744e-05, "loss": 0.008, "step": 42620 }, { "grad_norm": 0.0992777943611145, "learning_rate": 2.1218180956371634e-05, "loss": 0.0094, "step": 42630 }, { "grad_norm": 0.09324034303426743, "learning_rate": 2.119565113173252e-05, "loss": 0.0098, "step": 42640 }, { "grad_norm": 0.08556697517633438, "learning_rate": 2.1173130057094033e-05, "loss": 0.0092, "step": 42650 }, { "grad_norm": 0.08288302272558212, "learning_rate": 2.115061773929753e-05, "loss": 0.0095, "step": 42660 }, { "grad_norm": 0.09461479634046555, "learning_rate": 2.1128114185181623e-05, "loss": 0.0086, "step": 42670 }, { "grad_norm": 0.0708969309926033, "learning_rate": 2.1105619401582317e-05, "loss": 0.009, "step": 42680 }, { "grad_norm": 0.08233848214149475, "learning_rate": 2.1083133395332928e-05, "loss": 0.0091, "step": 42690 }, { "grad_norm": 0.1284565031528473, "learning_rate": 2.1060656173264082e-05, "loss": 0.0089, "step": 42700 }, { "grad_norm": 0.07737396657466888, "learning_rate": 2.103818774220383e-05, "loss": 0.0084, "step": 42710 }, { "grad_norm": 0.10312902182340622, "learning_rate": 2.1015728108977412e-05, "loss": 0.0087, "step": 42720 }, { "grad_norm": 0.10039869695901871, "learning_rate": 2.0993277280407548e-05, "loss": 0.0089, "step": 42730 }, { "grad_norm": 0.08260869979858398, "learning_rate": 2.0970835263314132e-05, "loss": 0.0089, "step": 42740 }, { "grad_norm": 0.12096147239208221, "learning_rate": 2.094840206451451e-05, "loss": 0.0104, "step": 42750 }, { "grad_norm": 0.08691838383674622, "learning_rate": 2.0925977690823273e-05, "loss": 0.0076, "step": 42760 }, { "grad_norm": 0.08389068394899368, "learning_rate": 2.0903562149052364e-05, "loss": 0.0078, "step": 42770 }, { "grad_norm": 0.09932637959718704, "learning_rate": 2.0881155446011025e-05, "loss": 0.0098, "step": 42780 }, { "grad_norm": 0.15785658359527588, "learning_rate": 2.0858757588505823e-05, "loss": 0.0086, "step": 42790 }, { "grad_norm": 0.08701347559690475, "learning_rate": 2.0836368583340622e-05, "loss": 0.0076, "step": 42800 }, { "grad_norm": 0.11759622395038605, "learning_rate": 2.081398843731664e-05, "loss": 0.0088, "step": 42810 }, { "grad_norm": 0.10574840009212494, "learning_rate": 2.0791617157232357e-05, "loss": 0.0089, "step": 42820 }, { "grad_norm": 0.11051811277866364, "learning_rate": 2.0769254749883576e-05, "loss": 0.0078, "step": 42830 }, { "grad_norm": 0.08074811100959778, "learning_rate": 2.0746901222063415e-05, "loss": 0.0087, "step": 42840 }, { "grad_norm": 0.11509141325950623, "learning_rate": 2.072455658056226e-05, "loss": 0.0098, "step": 42850 }, { "grad_norm": 0.10634036362171173, "learning_rate": 2.0702220832167873e-05, "loss": 0.0084, "step": 42860 }, { "grad_norm": 0.07554837316274643, "learning_rate": 2.0679893983665205e-05, "loss": 0.0083, "step": 42870 }, { "grad_norm": 0.0933537483215332, "learning_rate": 2.0657576041836622e-05, "loss": 0.0089, "step": 42880 }, { "grad_norm": 0.10818547755479813, "learning_rate": 2.0635267013461666e-05, "loss": 0.0089, "step": 42890 }, { "grad_norm": 0.09461770951747894, "learning_rate": 2.061296690531728e-05, "loss": 0.0086, "step": 42900 }, { "grad_norm": 0.08212464302778244, "learning_rate": 2.0590675724177622e-05, "loss": 0.0078, "step": 42910 }, { "grad_norm": 0.07647073268890381, "learning_rate": 2.0568393476814167e-05, "loss": 0.0108, "step": 42920 }, { "grad_norm": 0.09231719374656677, "learning_rate": 2.0546120169995685e-05, "loss": 0.0082, "step": 42930 }, { "grad_norm": 0.08865072578191757, "learning_rate": 2.0523855810488214e-05, "loss": 0.008, "step": 42940 }, { "grad_norm": 0.10347356647253036, "learning_rate": 2.050160040505505e-05, "loss": 0.008, "step": 42950 }, { "grad_norm": 0.09182704985141754, "learning_rate": 2.0479353960456843e-05, "loss": 0.0081, "step": 42960 }, { "grad_norm": 0.11978094279766083, "learning_rate": 2.0457116483451456e-05, "loss": 0.009, "step": 42970 }, { "grad_norm": 0.08887254446744919, "learning_rate": 2.0434887980794043e-05, "loss": 0.0082, "step": 42980 }, { "grad_norm": 0.11394865065813065, "learning_rate": 2.0412668459237043e-05, "loss": 0.0083, "step": 42990 }, { "grad_norm": 0.1033642366528511, "learning_rate": 2.039045792553016e-05, "loss": 0.0084, "step": 43000 }, { "grad_norm": 0.10543355345726013, "learning_rate": 2.036825638642036e-05, "loss": 0.0082, "step": 43010 }, { "grad_norm": 0.07086704671382904, "learning_rate": 2.0346063848651868e-05, "loss": 0.0091, "step": 43020 }, { "grad_norm": 0.09125356376171112, "learning_rate": 2.0323880318966254e-05, "loss": 0.0081, "step": 43030 }, { "grad_norm": 0.1333448737859726, "learning_rate": 2.030170580410221e-05, "loss": 0.0097, "step": 43040 }, { "grad_norm": 0.11381541937589645, "learning_rate": 2.0279540310795837e-05, "loss": 0.009, "step": 43050 }, { "grad_norm": 0.09366831183433533, "learning_rate": 2.0257383845780365e-05, "loss": 0.0087, "step": 43060 }, { "grad_norm": 0.13478589057922363, "learning_rate": 2.0235236415786384e-05, "loss": 0.0088, "step": 43070 }, { "grad_norm": 0.13689330220222473, "learning_rate": 2.021309802754169e-05, "loss": 0.0084, "step": 43080 }, { "grad_norm": 0.1370251178741455, "learning_rate": 2.0190968687771332e-05, "loss": 0.0091, "step": 43090 }, { "grad_norm": 0.07403159886598587, "learning_rate": 2.016884840319763e-05, "loss": 0.0084, "step": 43100 }, { "grad_norm": 0.0832274928689003, "learning_rate": 2.0146737180540122e-05, "loss": 0.0079, "step": 43110 }, { "grad_norm": 0.0870756283402443, "learning_rate": 2.012463502651564e-05, "loss": 0.0083, "step": 43120 }, { "grad_norm": 0.10859192162752151, "learning_rate": 2.0102541947838228e-05, "loss": 0.009, "step": 43130 }, { "grad_norm": 0.22402243316173553, "learning_rate": 2.0080457951219173e-05, "loss": 0.0087, "step": 43140 }, { "grad_norm": 0.08597385883331299, "learning_rate": 2.0058383043367017e-05, "loss": 0.0086, "step": 43150 }, { "grad_norm": 0.09617085754871368, "learning_rate": 2.0036317230987528e-05, "loss": 0.0087, "step": 43160 }, { "grad_norm": 0.10713227093219757, "learning_rate": 2.0014260520783696e-05, "loss": 0.0101, "step": 43170 }, { "grad_norm": 0.1178223118185997, "learning_rate": 1.9992212919455834e-05, "loss": 0.0088, "step": 43180 }, { "grad_norm": 0.10046166181564331, "learning_rate": 1.9970174433701333e-05, "loss": 0.0084, "step": 43190 }, { "grad_norm": 0.10994692146778107, "learning_rate": 1.9948145070214992e-05, "loss": 0.009, "step": 43200 }, { "grad_norm": 0.10183536261320114, "learning_rate": 1.9926124835688663e-05, "loss": 0.0085, "step": 43210 }, { "grad_norm": 0.08791600912809372, "learning_rate": 1.9904113736811576e-05, "loss": 0.0094, "step": 43220 }, { "grad_norm": 0.09956230223178864, "learning_rate": 1.9882111780270096e-05, "loss": 0.0094, "step": 43230 }, { "grad_norm": 0.09914709627628326, "learning_rate": 1.986011897274784e-05, "loss": 0.0095, "step": 43240 }, { "grad_norm": 0.08330279588699341, "learning_rate": 1.983813532092565e-05, "loss": 0.0081, "step": 43250 }, { "grad_norm": 0.08895440399646759, "learning_rate": 1.981616083148155e-05, "loss": 0.0081, "step": 43260 }, { "grad_norm": 0.11717627942562103, "learning_rate": 1.9794195511090845e-05, "loss": 0.0096, "step": 43270 }, { "grad_norm": 0.11852401494979858, "learning_rate": 1.977223936642601e-05, "loss": 0.0097, "step": 43280 }, { "grad_norm": 0.12290389090776443, "learning_rate": 1.975029240415674e-05, "loss": 0.0078, "step": 43290 }, { "grad_norm": 0.10664394497871399, "learning_rate": 1.9728354630949936e-05, "loss": 0.0084, "step": 43300 }, { "grad_norm": 0.09403664618730545, "learning_rate": 1.9706426053469716e-05, "loss": 0.0083, "step": 43310 }, { "grad_norm": 0.07792652398347855, "learning_rate": 1.9684506678377396e-05, "loss": 0.0092, "step": 43320 }, { "grad_norm": 0.07692208141088486, "learning_rate": 1.9662596512331544e-05, "loss": 0.0087, "step": 43330 }, { "grad_norm": 0.10563261806964874, "learning_rate": 1.964069556198782e-05, "loss": 0.0083, "step": 43340 }, { "grad_norm": 0.07840511202812195, "learning_rate": 1.9618803833999232e-05, "loss": 0.0088, "step": 43350 }, { "grad_norm": 0.0991751030087471, "learning_rate": 1.9596921335015838e-05, "loss": 0.0078, "step": 43360 }, { "grad_norm": 0.1098765954375267, "learning_rate": 1.957504807168501e-05, "loss": 0.0091, "step": 43370 }, { "grad_norm": 0.16266150772571564, "learning_rate": 1.9553184050651253e-05, "loss": 0.0095, "step": 43380 }, { "grad_norm": 0.11826381087303162, "learning_rate": 1.953132927855628e-05, "loss": 0.0079, "step": 43390 }, { "grad_norm": 0.09046684950590134, "learning_rate": 1.9509483762038995e-05, "loss": 0.0081, "step": 43400 }, { "grad_norm": 0.13377420604228973, "learning_rate": 1.9487647507735467e-05, "loss": 0.0082, "step": 43410 }, { "grad_norm": 0.1016717180609703, "learning_rate": 1.9465820522279032e-05, "loss": 0.008, "step": 43420 }, { "grad_norm": 0.1038299947977066, "learning_rate": 1.9444002812300078e-05, "loss": 0.0078, "step": 43430 }, { "grad_norm": 0.10812173783779144, "learning_rate": 1.94221943844263e-05, "loss": 0.0078, "step": 43440 }, { "grad_norm": 0.05880017951130867, "learning_rate": 1.9400395245282515e-05, "loss": 0.0079, "step": 43450 }, { "grad_norm": 0.10510660707950592, "learning_rate": 1.937860540149071e-05, "loss": 0.0085, "step": 43460 }, { "grad_norm": 0.0856894701719284, "learning_rate": 1.9356824859670082e-05, "loss": 0.0078, "step": 43470 }, { "grad_norm": 0.10419727116823196, "learning_rate": 1.9335053626436967e-05, "loss": 0.0086, "step": 43480 }, { "grad_norm": 0.07330849021673203, "learning_rate": 1.9313291708404885e-05, "loss": 0.0084, "step": 43490 }, { "grad_norm": 0.09077832102775574, "learning_rate": 1.9291539112184587e-05, "loss": 0.0086, "step": 43500 }, { "grad_norm": 0.09506651014089584, "learning_rate": 1.9269795844383854e-05, "loss": 0.0088, "step": 43510 }, { "grad_norm": 0.08157604932785034, "learning_rate": 1.9248061911607777e-05, "loss": 0.0078, "step": 43520 }, { "grad_norm": 0.08943606168031693, "learning_rate": 1.9226337320458538e-05, "loss": 0.0082, "step": 43530 }, { "grad_norm": 0.15302832424640656, "learning_rate": 1.9204622077535488e-05, "loss": 0.0086, "step": 43540 }, { "grad_norm": 0.12547142803668976, "learning_rate": 1.9182916189435147e-05, "loss": 0.009, "step": 43550 }, { "grad_norm": 0.13356749713420868, "learning_rate": 1.916121966275117e-05, "loss": 0.0103, "step": 43560 }, { "grad_norm": 0.13387569785118103, "learning_rate": 1.9139532504074443e-05, "loss": 0.0091, "step": 43570 }, { "grad_norm": 0.09757493436336517, "learning_rate": 1.9117854719992885e-05, "loss": 0.0081, "step": 43580 }, { "grad_norm": 0.11675860732793808, "learning_rate": 1.9096186317091687e-05, "loss": 0.0095, "step": 43590 }, { "grad_norm": 0.09309670329093933, "learning_rate": 1.9074527301953116e-05, "loss": 0.0102, "step": 43600 }, { "grad_norm": 0.10051661729812622, "learning_rate": 1.9052877681156607e-05, "loss": 0.0089, "step": 43610 }, { "grad_norm": 0.09386927634477615, "learning_rate": 1.903123746127875e-05, "loss": 0.0086, "step": 43620 }, { "grad_norm": 0.07792777568101883, "learning_rate": 1.900960664889327e-05, "loss": 0.0087, "step": 43630 }, { "grad_norm": 0.11377809941768646, "learning_rate": 1.8987985250571015e-05, "loss": 0.0085, "step": 43640 }, { "grad_norm": 0.1704939752817154, "learning_rate": 1.8966373272880054e-05, "loss": 0.0092, "step": 43650 }, { "grad_norm": 0.09624747186899185, "learning_rate": 1.8944770722385462e-05, "loss": 0.0083, "step": 43660 }, { "grad_norm": 0.10708208382129669, "learning_rate": 1.8923177605649576e-05, "loss": 0.0076, "step": 43670 }, { "grad_norm": 0.07155726850032806, "learning_rate": 1.8901593929231802e-05, "loss": 0.0081, "step": 43680 }, { "grad_norm": 0.10921071469783783, "learning_rate": 1.8880019699688684e-05, "loss": 0.0095, "step": 43690 }, { "grad_norm": 0.07839854806661606, "learning_rate": 1.8858454923573904e-05, "loss": 0.008, "step": 43700 }, { "grad_norm": 0.09275859594345093, "learning_rate": 1.8836899607438253e-05, "loss": 0.0085, "step": 43710 }, { "grad_norm": 0.10850876569747925, "learning_rate": 1.8815353757829723e-05, "loss": 0.0081, "step": 43720 }, { "grad_norm": 0.13591726124286652, "learning_rate": 1.879381738129331e-05, "loss": 0.0091, "step": 43730 }, { "grad_norm": 0.09617806226015091, "learning_rate": 1.8772290484371236e-05, "loss": 0.0076, "step": 43740 }, { "grad_norm": 0.09053424745798111, "learning_rate": 1.8750773073602795e-05, "loss": 0.0087, "step": 43750 }, { "grad_norm": 0.10610753297805786, "learning_rate": 1.8729265155524405e-05, "loss": 0.0094, "step": 43760 }, { "grad_norm": 0.1011284589767456, "learning_rate": 1.8707766736669607e-05, "loss": 0.0084, "step": 43770 }, { "grad_norm": 0.0755339041352272, "learning_rate": 1.8686277823569055e-05, "loss": 0.0083, "step": 43780 }, { "grad_norm": 0.08372417837381363, "learning_rate": 1.8664798422750484e-05, "loss": 0.0078, "step": 43790 }, { "grad_norm": 0.1106048971414566, "learning_rate": 1.8643328540738832e-05, "loss": 0.0097, "step": 43800 }, { "grad_norm": 0.12298431247472763, "learning_rate": 1.862186818405601e-05, "loss": 0.0088, "step": 43810 }, { "grad_norm": 0.08465375006198883, "learning_rate": 1.8600417359221156e-05, "loss": 0.0075, "step": 43820 }, { "grad_norm": 0.111348457634449, "learning_rate": 1.8578976072750454e-05, "loss": 0.0079, "step": 43830 }, { "grad_norm": 0.10940355807542801, "learning_rate": 1.8557544331157194e-05, "loss": 0.0081, "step": 43840 }, { "grad_norm": 0.10756520926952362, "learning_rate": 1.8536122140951785e-05, "loss": 0.0085, "step": 43850 }, { "grad_norm": 0.12416192889213562, "learning_rate": 1.8514709508641688e-05, "loss": 0.0087, "step": 43860 }, { "grad_norm": 0.09640637040138245, "learning_rate": 1.8493306440731555e-05, "loss": 0.0093, "step": 43870 }, { "grad_norm": 0.07639364153146744, "learning_rate": 1.8471912943723013e-05, "loss": 0.0086, "step": 43880 }, { "grad_norm": 0.09639996290206909, "learning_rate": 1.8450529024114894e-05, "loss": 0.0093, "step": 43890 }, { "grad_norm": 0.10705514997243881, "learning_rate": 1.842915468840301e-05, "loss": 0.0083, "step": 43900 }, { "grad_norm": 0.08804488927125931, "learning_rate": 1.840778994308037e-05, "loss": 0.0094, "step": 43910 }, { "grad_norm": 0.0700061097741127, "learning_rate": 1.8386434794637004e-05, "loss": 0.0085, "step": 43920 }, { "grad_norm": 0.13025395572185516, "learning_rate": 1.8365089249560034e-05, "loss": 0.0084, "step": 43930 }, { "grad_norm": 0.10023177415132523, "learning_rate": 1.8343753314333683e-05, "loss": 0.0083, "step": 43940 }, { "grad_norm": 0.09606650471687317, "learning_rate": 1.8322426995439236e-05, "loss": 0.0081, "step": 43950 }, { "grad_norm": 0.11621370166540146, "learning_rate": 1.8301110299355058e-05, "loss": 0.008, "step": 43960 }, { "grad_norm": 0.09666003286838531, "learning_rate": 1.8279803232556625e-05, "loss": 0.0083, "step": 43970 }, { "grad_norm": 0.09617764502763748, "learning_rate": 1.8258505801516444e-05, "loss": 0.0087, "step": 43980 }, { "grad_norm": 0.12298761308193207, "learning_rate": 1.8237218012704117e-05, "loss": 0.0078, "step": 43990 }, { "grad_norm": 0.09720395505428314, "learning_rate": 1.821593987258631e-05, "loss": 0.0076, "step": 44000 }, { "grad_norm": 0.0864095613360405, "learning_rate": 1.8194671387626744e-05, "loss": 0.0079, "step": 44010 }, { "grad_norm": 0.08301006257534027, "learning_rate": 1.8173412564286276e-05, "loss": 0.0076, "step": 44020 }, { "grad_norm": 0.09412654489278793, "learning_rate": 1.8152163409022697e-05, "loss": 0.0076, "step": 44030 }, { "grad_norm": 0.0881810337305069, "learning_rate": 1.8130923928291023e-05, "loss": 0.0077, "step": 44040 }, { "grad_norm": 0.10274794697761536, "learning_rate": 1.8109694128543163e-05, "loss": 0.0085, "step": 44050 }, { "grad_norm": 0.08727657049894333, "learning_rate": 1.8088474016228237e-05, "loss": 0.008, "step": 44060 }, { "grad_norm": 0.0996268093585968, "learning_rate": 1.8067263597792328e-05, "loss": 0.0084, "step": 44070 }, { "grad_norm": 0.09648140519857407, "learning_rate": 1.80460628796786e-05, "loss": 0.0079, "step": 44080 }, { "grad_norm": 0.07191640138626099, "learning_rate": 1.8024871868327276e-05, "loss": 0.0088, "step": 44090 }, { "grad_norm": 0.07791652530431747, "learning_rate": 1.8003690570175608e-05, "loss": 0.0085, "step": 44100 }, { "grad_norm": 0.12277932465076447, "learning_rate": 1.7982518991657943e-05, "loss": 0.0091, "step": 44110 }, { "grad_norm": 0.0702192559838295, "learning_rate": 1.7961357139205643e-05, "loss": 0.0082, "step": 44120 }, { "grad_norm": 0.13950008153915405, "learning_rate": 1.7940205019247108e-05, "loss": 0.0087, "step": 44130 }, { "grad_norm": 0.08039215952157974, "learning_rate": 1.79190626382078e-05, "loss": 0.0082, "step": 44140 }, { "grad_norm": 0.07693962007761002, "learning_rate": 1.7897930002510215e-05, "loss": 0.0092, "step": 44150 }, { "grad_norm": 0.14186570048332214, "learning_rate": 1.787680711857387e-05, "loss": 0.01, "step": 44160 }, { "grad_norm": 0.09046454727649689, "learning_rate": 1.7855693992815398e-05, "loss": 0.009, "step": 44170 }, { "grad_norm": 0.08609931170940399, "learning_rate": 1.7834590631648328e-05, "loss": 0.0078, "step": 44180 }, { "grad_norm": 0.09191552549600601, "learning_rate": 1.7813497041483384e-05, "loss": 0.0079, "step": 44190 }, { "grad_norm": 0.10272148996591568, "learning_rate": 1.779241322872817e-05, "loss": 0.0078, "step": 44200 }, { "grad_norm": 0.10048980265855789, "learning_rate": 1.777133919978744e-05, "loss": 0.01, "step": 44210 }, { "grad_norm": 0.11732835322618484, "learning_rate": 1.7750274961062912e-05, "loss": 0.0089, "step": 44220 }, { "grad_norm": 0.13603082299232483, "learning_rate": 1.772922051895335e-05, "loss": 0.0082, "step": 44230 }, { "grad_norm": 0.11886387318372726, "learning_rate": 1.770817587985453e-05, "loss": 0.0098, "step": 44240 }, { "grad_norm": 0.10106755048036575, "learning_rate": 1.7687141050159246e-05, "loss": 0.009, "step": 44250 }, { "grad_norm": 0.08273979276418686, "learning_rate": 1.7666116036257375e-05, "loss": 0.0089, "step": 44260 }, { "grad_norm": 0.1135011613368988, "learning_rate": 1.764510084453569e-05, "loss": 0.0095, "step": 44270 }, { "grad_norm": 0.10923902690410614, "learning_rate": 1.76240954813781e-05, "loss": 0.0087, "step": 44280 }, { "grad_norm": 0.08873992413282394, "learning_rate": 1.7603099953165476e-05, "loss": 0.0088, "step": 44290 }, { "grad_norm": 0.09025993943214417, "learning_rate": 1.7582114266275683e-05, "loss": 0.0076, "step": 44300 }, { "grad_norm": 0.1005314439535141, "learning_rate": 1.756113842708364e-05, "loss": 0.007, "step": 44310 }, { "grad_norm": 0.1046430915594101, "learning_rate": 1.7540172441961245e-05, "loss": 0.0086, "step": 44320 }, { "grad_norm": 0.06538405269384384, "learning_rate": 1.7519216317277387e-05, "loss": 0.0077, "step": 44330 }, { "grad_norm": 0.12610019743442535, "learning_rate": 1.7498270059398046e-05, "loss": 0.0083, "step": 44340 }, { "grad_norm": 0.09807900339365005, "learning_rate": 1.7477333674686062e-05, "loss": 0.0083, "step": 44350 }, { "grad_norm": 0.13791286945343018, "learning_rate": 1.745640716950142e-05, "loss": 0.0087, "step": 44360 }, { "grad_norm": 0.11614176630973816, "learning_rate": 1.7435490550201017e-05, "loss": 0.008, "step": 44370 }, { "grad_norm": 0.09261858463287354, "learning_rate": 1.7414583823138762e-05, "loss": 0.008, "step": 44380 }, { "grad_norm": 0.08876516669988632, "learning_rate": 1.739368699466558e-05, "loss": 0.0079, "step": 44390 }, { "grad_norm": 0.1130189597606659, "learning_rate": 1.737280007112935e-05, "loss": 0.0088, "step": 44400 }, { "grad_norm": 0.11018086969852448, "learning_rate": 1.735192305887502e-05, "loss": 0.0077, "step": 44410 }, { "grad_norm": 0.09776149690151215, "learning_rate": 1.733105596424441e-05, "loss": 0.0091, "step": 44420 }, { "grad_norm": 0.09230037778615952, "learning_rate": 1.7310198793576437e-05, "loss": 0.0084, "step": 44430 }, { "grad_norm": 0.09516985714435577, "learning_rate": 1.7289351553206952e-05, "loss": 0.0077, "step": 44440 }, { "grad_norm": 0.10631605237722397, "learning_rate": 1.7268514249468788e-05, "loss": 0.0087, "step": 44450 }, { "grad_norm": 0.08029060065746307, "learning_rate": 1.7247686888691765e-05, "loss": 0.008, "step": 44460 }, { "grad_norm": 0.07683534175157547, "learning_rate": 1.7226869477202694e-05, "loss": 0.0081, "step": 44470 }, { "grad_norm": 0.06690363585948944, "learning_rate": 1.7206062021325336e-05, "loss": 0.0085, "step": 44480 }, { "grad_norm": 0.09765618294477463, "learning_rate": 1.7185264527380502e-05, "loss": 0.0072, "step": 44490 }, { "grad_norm": 0.0839313343167305, "learning_rate": 1.716447700168584e-05, "loss": 0.01, "step": 44500 }, { "grad_norm": 0.09941757470369339, "learning_rate": 1.714369945055611e-05, "loss": 0.0082, "step": 44510 }, { "grad_norm": 0.07708317786455154, "learning_rate": 1.7122931880302968e-05, "loss": 0.0081, "step": 44520 }, { "grad_norm": 0.11785705387592316, "learning_rate": 1.710217429723505e-05, "loss": 0.01, "step": 44530 }, { "grad_norm": 0.09805954992771149, "learning_rate": 1.7081426707657972e-05, "loss": 0.0086, "step": 44540 }, { "grad_norm": 0.13515876233577728, "learning_rate": 1.7060689117874275e-05, "loss": 0.0092, "step": 44550 }, { "grad_norm": 0.10910271108150482, "learning_rate": 1.703996153418354e-05, "loss": 0.0082, "step": 44560 }, { "grad_norm": 0.09647136181592941, "learning_rate": 1.7019243962882205e-05, "loss": 0.0098, "step": 44570 }, { "grad_norm": 0.09224417805671692, "learning_rate": 1.6998536410263754e-05, "loss": 0.0085, "step": 44580 }, { "grad_norm": 0.1039653867483139, "learning_rate": 1.6977838882618596e-05, "loss": 0.0072, "step": 44590 }, { "grad_norm": 0.07909560203552246, "learning_rate": 1.6957151386234088e-05, "loss": 0.0074, "step": 44600 }, { "grad_norm": 0.07343347370624542, "learning_rate": 1.6936473927394536e-05, "loss": 0.0083, "step": 44610 }, { "grad_norm": 0.09898126870393753, "learning_rate": 1.6915806512381222e-05, "loss": 0.0076, "step": 44620 }, { "grad_norm": 0.09942566603422165, "learning_rate": 1.6895149147472344e-05, "loss": 0.0081, "step": 44630 }, { "grad_norm": 0.0803518146276474, "learning_rate": 1.6874501838943073e-05, "loss": 0.0071, "step": 44640 }, { "grad_norm": 0.13182444870471954, "learning_rate": 1.6853864593065506e-05, "loss": 0.0088, "step": 44650 }, { "grad_norm": 0.07790658622980118, "learning_rate": 1.683323741610871e-05, "loss": 0.0079, "step": 44660 }, { "grad_norm": 0.13553547859191895, "learning_rate": 1.6812620314338674e-05, "loss": 0.0081, "step": 44670 }, { "grad_norm": 0.10210052132606506, "learning_rate": 1.6792013294018326e-05, "loss": 0.011, "step": 44680 }, { "grad_norm": 0.12371189147233963, "learning_rate": 1.6771416361407526e-05, "loss": 0.0103, "step": 44690 }, { "grad_norm": 0.07594550400972366, "learning_rate": 1.675082952276308e-05, "loss": 0.0088, "step": 44700 }, { "grad_norm": 0.09979367256164551, "learning_rate": 1.6730252784338757e-05, "loss": 0.0074, "step": 44710 }, { "grad_norm": 0.08695927262306213, "learning_rate": 1.6709686152385166e-05, "loss": 0.0098, "step": 44720 }, { "grad_norm": 0.11876411736011505, "learning_rate": 1.668912963314998e-05, "loss": 0.0092, "step": 44730 }, { "grad_norm": 0.09120726585388184, "learning_rate": 1.6668583232877653e-05, "loss": 0.0077, "step": 44740 }, { "grad_norm": 0.09672640264034271, "learning_rate": 1.6648046957809698e-05, "loss": 0.0082, "step": 44750 }, { "grad_norm": 0.11130114644765854, "learning_rate": 1.6627520814184462e-05, "loss": 0.0077, "step": 44760 }, { "grad_norm": 0.09594104439020157, "learning_rate": 1.660700480823726e-05, "loss": 0.0078, "step": 44770 }, { "grad_norm": 0.12121020257472992, "learning_rate": 1.65864989462003e-05, "loss": 0.0073, "step": 44780 }, { "grad_norm": 0.09905296564102173, "learning_rate": 1.656600323430273e-05, "loss": 0.0081, "step": 44790 }, { "grad_norm": 0.07720570266246796, "learning_rate": 1.654551767877059e-05, "loss": 0.0075, "step": 44800 }, { "grad_norm": 0.08667176216840744, "learning_rate": 1.6525042285826874e-05, "loss": 0.0088, "step": 44810 }, { "grad_norm": 0.10095273703336716, "learning_rate": 1.6504577061691468e-05, "loss": 0.0081, "step": 44820 }, { "grad_norm": 0.10613366216421127, "learning_rate": 1.6484122012581143e-05, "loss": 0.0081, "step": 44830 }, { "grad_norm": 0.1170521005988121, "learning_rate": 1.6463677144709623e-05, "loss": 0.0082, "step": 44840 }, { "grad_norm": 0.09116159379482269, "learning_rate": 1.6443242464287493e-05, "loss": 0.0081, "step": 44850 }, { "grad_norm": 0.1160278245806694, "learning_rate": 1.642281797752232e-05, "loss": 0.008, "step": 44860 }, { "grad_norm": 0.13097450137138367, "learning_rate": 1.6402403690618456e-05, "loss": 0.0079, "step": 44870 }, { "grad_norm": 0.08471664786338806, "learning_rate": 1.6381999609777295e-05, "loss": 0.0083, "step": 44880 }, { "grad_norm": 0.10321301966905594, "learning_rate": 1.6361605741196983e-05, "loss": 0.0082, "step": 44890 }, { "grad_norm": 0.0840587466955185, "learning_rate": 1.63412220910727e-05, "loss": 0.0093, "step": 44900 }, { "grad_norm": 0.0800672098994255, "learning_rate": 1.6320848665596433e-05, "loss": 0.008, "step": 44910 }, { "grad_norm": 0.08458888530731201, "learning_rate": 1.6300485470957095e-05, "loss": 0.0089, "step": 44920 }, { "grad_norm": 0.09286421537399292, "learning_rate": 1.6280132513340483e-05, "loss": 0.0087, "step": 44930 }, { "grad_norm": 0.09347322583198547, "learning_rate": 1.62597897989293e-05, "loss": 0.0087, "step": 44940 }, { "grad_norm": 0.05862990766763687, "learning_rate": 1.623945733390309e-05, "loss": 0.0078, "step": 44950 }, { "grad_norm": 0.08372863382101059, "learning_rate": 1.6219135124438374e-05, "loss": 0.0075, "step": 44960 }, { "grad_norm": 0.09851189702749252, "learning_rate": 1.6198823176708465e-05, "loss": 0.0072, "step": 44970 }, { "grad_norm": 0.09969606250524521, "learning_rate": 1.6178521496883613e-05, "loss": 0.0074, "step": 44980 }, { "grad_norm": 0.11189743131399155, "learning_rate": 1.6158230091130926e-05, "loss": 0.0091, "step": 44990 }, { "grad_norm": 0.10790669173002243, "learning_rate": 1.613794896561438e-05, "loss": 0.0081, "step": 45000 }, { "grad_norm": 0.10143999010324478, "learning_rate": 1.6117678126494894e-05, "loss": 0.0093, "step": 45010 }, { "grad_norm": 0.09917719662189484, "learning_rate": 1.6097417579930153e-05, "loss": 0.0089, "step": 45020 }, { "grad_norm": 0.07164749503135681, "learning_rate": 1.6077167332074834e-05, "loss": 0.0083, "step": 45030 }, { "grad_norm": 0.24777211248874664, "learning_rate": 1.605692738908037e-05, "loss": 0.0102, "step": 45040 }, { "grad_norm": 0.10920451581478119, "learning_rate": 1.6036697757095176e-05, "loss": 0.0085, "step": 45050 }, { "grad_norm": 0.08526674658060074, "learning_rate": 1.6016478442264428e-05, "loss": 0.0079, "step": 45060 }, { "grad_norm": 0.07420652359724045, "learning_rate": 1.599626945073026e-05, "loss": 0.0091, "step": 45070 }, { "grad_norm": 0.08035583049058914, "learning_rate": 1.597607078863162e-05, "loss": 0.0072, "step": 45080 }, { "grad_norm": 0.10823197662830353, "learning_rate": 1.595588246210432e-05, "loss": 0.0081, "step": 45090 }, { "grad_norm": 0.08439475297927856, "learning_rate": 1.5935704477281048e-05, "loss": 0.0084, "step": 45100 }, { "grad_norm": 0.09406109154224396, "learning_rate": 1.5915536840291323e-05, "loss": 0.0085, "step": 45110 }, { "grad_norm": 0.08800594508647919, "learning_rate": 1.5895379557261576e-05, "loss": 0.0085, "step": 45120 }, { "grad_norm": 0.09712237864732742, "learning_rate": 1.5875232634315033e-05, "loss": 0.0077, "step": 45130 }, { "grad_norm": 0.09003540128469467, "learning_rate": 1.5855096077571812e-05, "loss": 0.0078, "step": 45140 }, { "grad_norm": 0.0906175971031189, "learning_rate": 1.5834969893148855e-05, "loss": 0.0078, "step": 45150 }, { "grad_norm": 0.13094189763069153, "learning_rate": 1.581485408715997e-05, "loss": 0.008, "step": 45160 }, { "grad_norm": 0.08977936953306198, "learning_rate": 1.5794748665715785e-05, "loss": 0.0071, "step": 45170 }, { "grad_norm": 0.09576491266489029, "learning_rate": 1.5774653634923857e-05, "loss": 0.0072, "step": 45180 }, { "grad_norm": 0.06994877010583878, "learning_rate": 1.575456900088845e-05, "loss": 0.0082, "step": 45190 }, { "grad_norm": 0.11063791811466217, "learning_rate": 1.5734494769710816e-05, "loss": 0.0084, "step": 45200 }, { "grad_norm": 0.11118052899837494, "learning_rate": 1.5714430947488912e-05, "loss": 0.0079, "step": 45210 }, { "grad_norm": 0.09426844865083694, "learning_rate": 1.5694377540317645e-05, "loss": 0.0077, "step": 45220 }, { "grad_norm": 0.1328115314245224, "learning_rate": 1.5674334554288694e-05, "loss": 0.0077, "step": 45230 }, { "grad_norm": 0.11861923336982727, "learning_rate": 1.5654301995490582e-05, "loss": 0.0091, "step": 45240 }, { "grad_norm": 0.09680282324552536, "learning_rate": 1.5634279870008685e-05, "loss": 0.0095, "step": 45250 }, { "grad_norm": 0.10562320798635483, "learning_rate": 1.5614268183925174e-05, "loss": 0.0086, "step": 45260 }, { "grad_norm": 0.10031858831644058, "learning_rate": 1.5594266943319097e-05, "loss": 0.0088, "step": 45270 }, { "grad_norm": 0.0861479863524437, "learning_rate": 1.5574276154266294e-05, "loss": 0.008, "step": 45280 }, { "grad_norm": 0.1025102362036705, "learning_rate": 1.5554295822839437e-05, "loss": 0.0089, "step": 45290 }, { "grad_norm": 0.08775115013122559, "learning_rate": 1.5534325955108025e-05, "loss": 0.0094, "step": 45300 }, { "grad_norm": 0.07271944731473923, "learning_rate": 1.5514366557138373e-05, "loss": 0.0103, "step": 45310 }, { "grad_norm": 0.07112082093954086, "learning_rate": 1.5494417634993602e-05, "loss": 0.0079, "step": 45320 }, { "grad_norm": 0.05325872451066971, "learning_rate": 1.547447919473372e-05, "loss": 0.0081, "step": 45330 }, { "grad_norm": 0.1068110466003418, "learning_rate": 1.5454551242415434e-05, "loss": 0.0082, "step": 45340 }, { "grad_norm": 0.1163855791091919, "learning_rate": 1.543463378409239e-05, "loss": 0.0095, "step": 45350 }, { "grad_norm": 0.09215781837701797, "learning_rate": 1.541472682581493e-05, "loss": 0.007, "step": 45360 }, { "grad_norm": 0.15766826272010803, "learning_rate": 1.5394830373630298e-05, "loss": 0.0082, "step": 45370 }, { "grad_norm": 0.08452796936035156, "learning_rate": 1.5374944433582506e-05, "loss": 0.0079, "step": 45380 }, { "grad_norm": 0.07716108858585358, "learning_rate": 1.5355069011712375e-05, "loss": 0.0077, "step": 45390 }, { "grad_norm": 0.08408577740192413, "learning_rate": 1.5335204114057526e-05, "loss": 0.0072, "step": 45400 }, { "grad_norm": 0.08079823106527328, "learning_rate": 1.5315349746652387e-05, "loss": 0.0076, "step": 45410 }, { "grad_norm": 0.08140718936920166, "learning_rate": 1.5295505915528212e-05, "loss": 0.0083, "step": 45420 }, { "grad_norm": 0.07381081581115723, "learning_rate": 1.5275672626713024e-05, "loss": 0.008, "step": 45430 }, { "grad_norm": 0.06615664064884186, "learning_rate": 1.5255849886231643e-05, "loss": 0.0072, "step": 45440 }, { "grad_norm": 0.08769706636667252, "learning_rate": 1.523603770010571e-05, "loss": 0.0082, "step": 45450 }, { "grad_norm": 0.1364046335220337, "learning_rate": 1.521623607435363e-05, "loss": 0.0092, "step": 45460 }, { "grad_norm": 0.07670975476503372, "learning_rate": 1.5196445014990612e-05, "loss": 0.0082, "step": 45470 }, { "grad_norm": 0.1079300194978714, "learning_rate": 1.5176664528028672e-05, "loss": 0.0085, "step": 45480 }, { "grad_norm": 0.10170458257198334, "learning_rate": 1.5156894619476574e-05, "loss": 0.0076, "step": 45490 }, { "grad_norm": 0.07667063921689987, "learning_rate": 1.5137135295339938e-05, "loss": 0.0073, "step": 45500 }, { "grad_norm": 0.10767762362957001, "learning_rate": 1.5117386561621073e-05, "loss": 0.0075, "step": 45510 }, { "grad_norm": 0.1152133122086525, "learning_rate": 1.5097648424319167e-05, "loss": 0.0105, "step": 45520 }, { "grad_norm": 0.10506229847669601, "learning_rate": 1.5077920889430119e-05, "loss": 0.0078, "step": 45530 }, { "grad_norm": 0.11498407274484634, "learning_rate": 1.5058203962946644e-05, "loss": 0.0081, "step": 45540 }, { "grad_norm": 0.09129980951547623, "learning_rate": 1.503849765085822e-05, "loss": 0.0096, "step": 45550 }, { "grad_norm": 0.09398215264081955, "learning_rate": 1.501880195915109e-05, "loss": 0.0082, "step": 45560 }, { "grad_norm": 0.1064908504486084, "learning_rate": 1.499911689380833e-05, "loss": 0.0083, "step": 45570 }, { "grad_norm": 0.09293391555547714, "learning_rate": 1.4979442460809683e-05, "loss": 0.0074, "step": 45580 }, { "grad_norm": 0.13132184743881226, "learning_rate": 1.4959778666131763e-05, "loss": 0.0082, "step": 45590 }, { "grad_norm": 0.13729245960712433, "learning_rate": 1.4940125515747905e-05, "loss": 0.0075, "step": 45600 }, { "grad_norm": 0.10337276756763458, "learning_rate": 1.4920483015628211e-05, "loss": 0.0074, "step": 45610 }, { "grad_norm": 0.09997235238552094, "learning_rate": 1.490085117173956e-05, "loss": 0.007, "step": 45620 }, { "grad_norm": 0.0871305763721466, "learning_rate": 1.488122999004558e-05, "loss": 0.0078, "step": 45630 }, { "grad_norm": 0.18214866518974304, "learning_rate": 1.486161947650666e-05, "loss": 0.0089, "step": 45640 }, { "grad_norm": 0.09293709695339203, "learning_rate": 1.4842019637079995e-05, "loss": 0.0074, "step": 45650 }, { "grad_norm": 0.09837053716182709, "learning_rate": 1.482243047771944e-05, "loss": 0.0085, "step": 45660 }, { "grad_norm": 0.08246543258428574, "learning_rate": 1.4802852004375712e-05, "loss": 0.0066, "step": 45670 }, { "grad_norm": 0.08171962201595306, "learning_rate": 1.4783284222996218e-05, "loss": 0.0084, "step": 45680 }, { "grad_norm": 0.07975315302610397, "learning_rate": 1.4763727139525135e-05, "loss": 0.0068, "step": 45690 }, { "grad_norm": 0.08724761009216309, "learning_rate": 1.4744180759903392e-05, "loss": 0.0073, "step": 45700 }, { "grad_norm": 0.08444114029407501, "learning_rate": 1.4724645090068635e-05, "loss": 0.0079, "step": 45710 }, { "grad_norm": 0.08337394148111343, "learning_rate": 1.4705120135955341e-05, "loss": 0.0076, "step": 45720 }, { "grad_norm": 0.15297167003154755, "learning_rate": 1.4685605903494614e-05, "loss": 0.0074, "step": 45730 }, { "grad_norm": 0.08575006574392319, "learning_rate": 1.46661023986144e-05, "loss": 0.0075, "step": 45740 }, { "grad_norm": 0.12399069964885712, "learning_rate": 1.4646609627239344e-05, "loss": 0.0071, "step": 45750 }, { "grad_norm": 0.11666364222764969, "learning_rate": 1.4627127595290835e-05, "loss": 0.01, "step": 45760 }, { "grad_norm": 0.12356600910425186, "learning_rate": 1.460765630868699e-05, "loss": 0.0069, "step": 45770 }, { "grad_norm": 0.08672311156988144, "learning_rate": 1.4588195773342678e-05, "loss": 0.0081, "step": 45780 }, { "grad_norm": 0.05521054193377495, "learning_rate": 1.4568745995169485e-05, "loss": 0.0075, "step": 45790 }, { "grad_norm": 0.07775723189115524, "learning_rate": 1.4549306980075778e-05, "loss": 0.0081, "step": 45800 }, { "grad_norm": 0.11069182306528091, "learning_rate": 1.4529878733966557e-05, "loss": 0.0087, "step": 45810 }, { "grad_norm": 0.13270391523838043, "learning_rate": 1.4510461262743658e-05, "loss": 0.0078, "step": 45820 }, { "grad_norm": 0.07920683175325394, "learning_rate": 1.4491054572305585e-05, "loss": 0.0071, "step": 45830 }, { "grad_norm": 0.0816236063838005, "learning_rate": 1.4471658668547566e-05, "loss": 0.0079, "step": 45840 }, { "grad_norm": 0.08603721112012863, "learning_rate": 1.4452273557361579e-05, "loss": 0.0077, "step": 45850 }, { "grad_norm": 0.10471361875534058, "learning_rate": 1.4432899244636282e-05, "loss": 0.0068, "step": 45860 }, { "grad_norm": 0.10997368395328522, "learning_rate": 1.4413535736257134e-05, "loss": 0.0078, "step": 45870 }, { "grad_norm": 0.09762686491012573, "learning_rate": 1.439418303810619e-05, "loss": 0.0079, "step": 45880 }, { "grad_norm": 0.09496574848890305, "learning_rate": 1.4374841156062352e-05, "loss": 0.008, "step": 45890 }, { "grad_norm": 0.11529764533042908, "learning_rate": 1.4355510096001112e-05, "loss": 0.0084, "step": 45900 }, { "grad_norm": 0.13211382925510406, "learning_rate": 1.4336189863794786e-05, "loss": 0.0078, "step": 45910 }, { "grad_norm": 0.11546671390533447, "learning_rate": 1.4316880465312327e-05, "loss": 0.0087, "step": 45920 }, { "grad_norm": 0.11262481659650803, "learning_rate": 1.4297581906419426e-05, "loss": 0.0081, "step": 45930 }, { "grad_norm": 0.07001084834337234, "learning_rate": 1.4278294192978475e-05, "loss": 0.0079, "step": 45940 }, { "grad_norm": 0.11927627772092819, "learning_rate": 1.4259017330848574e-05, "loss": 0.0075, "step": 45950 }, { "grad_norm": 0.09157055616378784, "learning_rate": 1.4239751325885498e-05, "loss": 0.008, "step": 45960 }, { "grad_norm": 0.08896784484386444, "learning_rate": 1.4220496183941795e-05, "loss": 0.0083, "step": 45970 }, { "grad_norm": 0.1288035660982132, "learning_rate": 1.4201251910866648e-05, "loss": 0.0097, "step": 45980 }, { "grad_norm": 0.08928923308849335, "learning_rate": 1.4182018512505957e-05, "loss": 0.0074, "step": 45990 }, { "grad_norm": 0.09890435636043549, "learning_rate": 1.4162795994702327e-05, "loss": 0.0082, "step": 46000 }, { "grad_norm": 0.08885309100151062, "learning_rate": 1.4143584363295032e-05, "loss": 0.0083, "step": 46010 }, { "grad_norm": 0.10157953947782516, "learning_rate": 1.4124383624120101e-05, "loss": 0.0083, "step": 46020 }, { "grad_norm": 0.08723445236682892, "learning_rate": 1.4105193783010151e-05, "loss": 0.0074, "step": 46030 }, { "grad_norm": 0.0917944684624672, "learning_rate": 1.4086014845794621e-05, "loss": 0.0092, "step": 46040 }, { "grad_norm": 0.11699741333723068, "learning_rate": 1.4066846818299489e-05, "loss": 0.0089, "step": 46050 }, { "grad_norm": 0.08529973775148392, "learning_rate": 1.4047689706347555e-05, "loss": 0.0071, "step": 46060 }, { "grad_norm": 0.13493378460407257, "learning_rate": 1.402854351575822e-05, "loss": 0.0076, "step": 46070 }, { "grad_norm": 0.08059375733137131, "learning_rate": 1.4009408252347588e-05, "loss": 0.0082, "step": 46080 }, { "grad_norm": 0.07868693768978119, "learning_rate": 1.399028392192846e-05, "loss": 0.0072, "step": 46090 }, { "grad_norm": 0.06582308560609818, "learning_rate": 1.397117053031029e-05, "loss": 0.0076, "step": 46100 }, { "grad_norm": 0.1083463579416275, "learning_rate": 1.3952068083299213e-05, "loss": 0.0083, "step": 46110 }, { "grad_norm": 0.13232183456420898, "learning_rate": 1.3932976586698082e-05, "loss": 0.0078, "step": 46120 }, { "grad_norm": 0.07693936675786972, "learning_rate": 1.3913896046306363e-05, "loss": 0.007, "step": 46130 }, { "grad_norm": 0.11224581301212311, "learning_rate": 1.389482646792023e-05, "loss": 0.007, "step": 46140 }, { "grad_norm": 0.09952748566865921, "learning_rate": 1.387576785733251e-05, "loss": 0.0079, "step": 46150 }, { "grad_norm": 0.08605305850505829, "learning_rate": 1.3856720220332703e-05, "loss": 0.0083, "step": 46160 }, { "grad_norm": 0.12098980695009232, "learning_rate": 1.383768356270701e-05, "loss": 0.0085, "step": 46170 }, { "grad_norm": 0.07432253658771515, "learning_rate": 1.3818657890238207e-05, "loss": 0.0083, "step": 46180 }, { "grad_norm": 0.07782593369483948, "learning_rate": 1.3799643208705859e-05, "loss": 0.0065, "step": 46190 }, { "grad_norm": 0.13439594209194183, "learning_rate": 1.3780639523886058e-05, "loss": 0.0079, "step": 46200 }, { "grad_norm": 0.12203123420476913, "learning_rate": 1.3761646841551668e-05, "loss": 0.0072, "step": 46210 }, { "grad_norm": 0.08866378664970398, "learning_rate": 1.3742665167472146e-05, "loss": 0.0074, "step": 46220 }, { "grad_norm": 0.10331785678863525, "learning_rate": 1.372369450741363e-05, "loss": 0.0078, "step": 46230 }, { "grad_norm": 0.10449614375829697, "learning_rate": 1.3704734867138901e-05, "loss": 0.0084, "step": 46240 }, { "grad_norm": 0.08824499696493149, "learning_rate": 1.36857862524074e-05, "loss": 0.0074, "step": 46250 }, { "grad_norm": 0.09340275824069977, "learning_rate": 1.3666848668975213e-05, "loss": 0.0075, "step": 46260 }, { "grad_norm": 0.07438614964485168, "learning_rate": 1.3647922122595063e-05, "loss": 0.008, "step": 46270 }, { "grad_norm": 0.07680188864469528, "learning_rate": 1.3629006619016366e-05, "loss": 0.0078, "step": 46280 }, { "grad_norm": 0.08749763667583466, "learning_rate": 1.3610102163985139e-05, "loss": 0.0074, "step": 46290 }, { "grad_norm": 0.09326242655515671, "learning_rate": 1.3591208763244057e-05, "loss": 0.0093, "step": 46300 }, { "grad_norm": 0.0868803858757019, "learning_rate": 1.3572326422532428e-05, "loss": 0.0074, "step": 46310 }, { "grad_norm": 0.1066623255610466, "learning_rate": 1.355345514758622e-05, "loss": 0.0084, "step": 46320 }, { "grad_norm": 0.09983659535646439, "learning_rate": 1.3534594944138007e-05, "loss": 0.0077, "step": 46330 }, { "grad_norm": 0.09927991777658463, "learning_rate": 1.3515745817917069e-05, "loss": 0.008, "step": 46340 }, { "grad_norm": 0.0740000382065773, "learning_rate": 1.3496907774649208e-05, "loss": 0.0072, "step": 46350 }, { "grad_norm": 0.09413564950227737, "learning_rate": 1.3478080820056987e-05, "loss": 0.0078, "step": 46360 }, { "grad_norm": 0.1041647419333458, "learning_rate": 1.3459264959859474e-05, "loss": 0.0078, "step": 46370 }, { "grad_norm": 0.0831630751490593, "learning_rate": 1.3440460199772487e-05, "loss": 0.0073, "step": 46380 }, { "grad_norm": 0.10727182030677795, "learning_rate": 1.3421666545508382e-05, "loss": 0.0079, "step": 46390 }, { "grad_norm": 0.112169548869133, "learning_rate": 1.3402884002776194e-05, "loss": 0.0084, "step": 46400 }, { "grad_norm": 0.08472611755132675, "learning_rate": 1.3384112577281555e-05, "loss": 0.008, "step": 46410 }, { "grad_norm": 0.09335540235042572, "learning_rate": 1.3365352274726711e-05, "loss": 0.0087, "step": 46420 }, { "grad_norm": 0.07834829390048981, "learning_rate": 1.3346603100810578e-05, "loss": 0.0082, "step": 46430 }, { "grad_norm": 0.07359655946493149, "learning_rate": 1.3327865061228645e-05, "loss": 0.0078, "step": 46440 }, { "grad_norm": 0.06841517984867096, "learning_rate": 1.330913816167304e-05, "loss": 0.0081, "step": 46450 }, { "grad_norm": 0.075351782143116, "learning_rate": 1.3290422407832492e-05, "loss": 0.0074, "step": 46460 }, { "grad_norm": 0.1310766190290451, "learning_rate": 1.3271717805392354e-05, "loss": 0.0095, "step": 46470 }, { "grad_norm": 0.15687507390975952, "learning_rate": 1.3253024360034582e-05, "loss": 0.0097, "step": 46480 }, { "grad_norm": 0.11917435377836227, "learning_rate": 1.323434207743779e-05, "loss": 0.0072, "step": 46490 }, { "grad_norm": 0.09866341203451157, "learning_rate": 1.3215670963277105e-05, "loss": 0.0073, "step": 46500 }, { "grad_norm": 0.08656994253396988, "learning_rate": 1.3197011023224376e-05, "loss": 0.0071, "step": 46510 }, { "grad_norm": 0.0834977999329567, "learning_rate": 1.3178362262947941e-05, "loss": 0.0076, "step": 46520 }, { "grad_norm": 0.11210134625434875, "learning_rate": 1.3159724688112845e-05, "loss": 0.0069, "step": 46530 }, { "grad_norm": 0.10891559720039368, "learning_rate": 1.3141098304380683e-05, "loss": 0.0089, "step": 46540 }, { "grad_norm": 0.09334724396467209, "learning_rate": 1.3122483117409651e-05, "loss": 0.0081, "step": 46550 }, { "grad_norm": 0.11228296160697937, "learning_rate": 1.3103879132854552e-05, "loss": 0.0078, "step": 46560 }, { "grad_norm": 0.07609348744153976, "learning_rate": 1.3085286356366771e-05, "loss": 0.0071, "step": 46570 }, { "grad_norm": 0.08567269891500473, "learning_rate": 1.3066704793594337e-05, "loss": 0.0077, "step": 46580 }, { "grad_norm": 0.11520097404718399, "learning_rate": 1.3048134450181816e-05, "loss": 0.0076, "step": 46590 }, { "grad_norm": 0.08703876286745071, "learning_rate": 1.3029575331770394e-05, "loss": 0.0083, "step": 46600 }, { "grad_norm": 0.07979939877986908, "learning_rate": 1.3011027443997837e-05, "loss": 0.0071, "step": 46610 }, { "grad_norm": 0.08878061920404434, "learning_rate": 1.2992490792498507e-05, "loss": 0.0084, "step": 46620 }, { "grad_norm": 0.07728613168001175, "learning_rate": 1.297396538290333e-05, "loss": 0.0073, "step": 46630 }, { "grad_norm": 0.09788134694099426, "learning_rate": 1.2955451220839888e-05, "loss": 0.0084, "step": 46640 }, { "grad_norm": 0.14270737767219543, "learning_rate": 1.2936948311932223e-05, "loss": 0.008, "step": 46650 }, { "grad_norm": 0.09308801591396332, "learning_rate": 1.2918456661801104e-05, "loss": 0.0079, "step": 46660 }, { "grad_norm": 0.09085658937692642, "learning_rate": 1.2899976276063736e-05, "loss": 0.0078, "step": 46670 }, { "grad_norm": 0.13597427308559418, "learning_rate": 1.2881507160334022e-05, "loss": 0.0087, "step": 46680 }, { "grad_norm": 0.12547248601913452, "learning_rate": 1.286304932022238e-05, "loss": 0.0075, "step": 46690 }, { "grad_norm": 0.10926640033721924, "learning_rate": 1.2844602761335806e-05, "loss": 0.0068, "step": 46700 }, { "grad_norm": 0.08277672529220581, "learning_rate": 1.2826167489277885e-05, "loss": 0.0067, "step": 46710 }, { "grad_norm": 0.08938310295343399, "learning_rate": 1.2807743509648745e-05, "loss": 0.0085, "step": 46720 }, { "grad_norm": 0.09817346185445786, "learning_rate": 1.2789330828045149e-05, "loss": 0.0083, "step": 46730 }, { "grad_norm": 0.07830076664686203, "learning_rate": 1.2770929450060332e-05, "loss": 0.0068, "step": 46740 }, { "grad_norm": 0.08477799594402313, "learning_rate": 1.2752539381284184e-05, "loss": 0.009, "step": 46750 }, { "grad_norm": 0.08365502208471298, "learning_rate": 1.273416062730311e-05, "loss": 0.0074, "step": 46760 }, { "grad_norm": 0.06740681827068329, "learning_rate": 1.2715793193700088e-05, "loss": 0.0075, "step": 46770 }, { "grad_norm": 0.10823223739862442, "learning_rate": 1.2697437086054664e-05, "loss": 0.0109, "step": 46780 }, { "grad_norm": 0.06625979393720627, "learning_rate": 1.2679092309942937e-05, "loss": 0.007, "step": 46790 }, { "grad_norm": 0.09109287708997726, "learning_rate": 1.266075887093755e-05, "loss": 0.0073, "step": 46800 }, { "grad_norm": 0.13149085640907288, "learning_rate": 1.2642436774607757e-05, "loss": 0.0093, "step": 46810 }, { "grad_norm": 0.10332132130861282, "learning_rate": 1.2624126026519278e-05, "loss": 0.0077, "step": 46820 }, { "grad_norm": 0.091984823346138, "learning_rate": 1.2605826632234474e-05, "loss": 0.0074, "step": 46830 }, { "grad_norm": 0.10921743512153625, "learning_rate": 1.2587538597312198e-05, "loss": 0.0074, "step": 46840 }, { "grad_norm": 0.0852888748049736, "learning_rate": 1.2569261927307884e-05, "loss": 0.0075, "step": 46850 }, { "grad_norm": 0.08495848625898361, "learning_rate": 1.2550996627773493e-05, "loss": 0.0076, "step": 46860 }, { "grad_norm": 0.09160926192998886, "learning_rate": 1.2532742704257527e-05, "loss": 0.0074, "step": 46870 }, { "grad_norm": 0.09846768528223038, "learning_rate": 1.2514500162305087e-05, "loss": 0.008, "step": 46880 }, { "grad_norm": 0.08578508347272873, "learning_rate": 1.2496269007457728e-05, "loss": 0.007, "step": 46890 }, { "grad_norm": 0.07054868340492249, "learning_rate": 1.2478049245253625e-05, "loss": 0.0088, "step": 46900 }, { "grad_norm": 0.10397154837846756, "learning_rate": 1.2459840881227459e-05, "loss": 0.0085, "step": 46910 }, { "grad_norm": 0.07565024495124817, "learning_rate": 1.2441643920910435e-05, "loss": 0.0091, "step": 46920 }, { "grad_norm": 0.10260505229234695, "learning_rate": 1.2423458369830322e-05, "loss": 0.0079, "step": 46930 }, { "grad_norm": 0.09546954184770584, "learning_rate": 1.2405284233511406e-05, "loss": 0.0083, "step": 46940 }, { "grad_norm": 0.10657002031803131, "learning_rate": 1.2387121517474487e-05, "loss": 0.008, "step": 46950 }, { "grad_norm": 0.08079894632101059, "learning_rate": 1.2368970227236975e-05, "loss": 0.0073, "step": 46960 }, { "grad_norm": 0.09385395050048828, "learning_rate": 1.2350830368312688e-05, "loss": 0.0075, "step": 46970 }, { "grad_norm": 0.05768147110939026, "learning_rate": 1.2332701946212083e-05, "loss": 0.0076, "step": 46980 }, { "grad_norm": 0.10739149153232574, "learning_rate": 1.2314584966442077e-05, "loss": 0.0089, "step": 46990 }, { "grad_norm": 0.11761824786663055, "learning_rate": 1.2296479434506136e-05, "loss": 0.0075, "step": 47000 }, { "grad_norm": 0.11135967820882797, "learning_rate": 1.2278385355904232e-05, "loss": 0.0072, "step": 47010 }, { "grad_norm": 0.08161471039056778, "learning_rate": 1.2260302736132867e-05, "loss": 0.0075, "step": 47020 }, { "grad_norm": 0.0911211147904396, "learning_rate": 1.2242231580685098e-05, "loss": 0.0073, "step": 47030 }, { "grad_norm": 0.10379238426685333, "learning_rate": 1.2224171895050413e-05, "loss": 0.0071, "step": 47040 }, { "grad_norm": 0.0716460719704628, "learning_rate": 1.2206123684714903e-05, "loss": 0.0069, "step": 47050 }, { "grad_norm": 0.08689910173416138, "learning_rate": 1.2188086955161132e-05, "loss": 0.0078, "step": 47060 }, { "grad_norm": 0.08777876198291779, "learning_rate": 1.2170061711868175e-05, "loss": 0.0084, "step": 47070 }, { "grad_norm": 0.08769103139638901, "learning_rate": 1.215204796031163e-05, "loss": 0.0081, "step": 47080 }, { "grad_norm": 0.08341474086046219, "learning_rate": 1.2134045705963599e-05, "loss": 0.0077, "step": 47090 }, { "grad_norm": 0.0983022004365921, "learning_rate": 1.2116054954292689e-05, "loss": 0.0074, "step": 47100 }, { "grad_norm": 0.0932549461722374, "learning_rate": 1.2098075710764011e-05, "loss": 0.0069, "step": 47110 }, { "grad_norm": 0.09935521334409714, "learning_rate": 1.2080107980839183e-05, "loss": 0.0083, "step": 47120 }, { "grad_norm": 0.10492289811372757, "learning_rate": 1.2062151769976343e-05, "loss": 0.0077, "step": 47130 }, { "grad_norm": 0.10098592191934586, "learning_rate": 1.204420708363011e-05, "loss": 0.0078, "step": 47140 }, { "grad_norm": 0.09375602751970291, "learning_rate": 1.2026273927251597e-05, "loss": 0.0076, "step": 47150 }, { "grad_norm": 0.09894810616970062, "learning_rate": 1.2008352306288424e-05, "loss": 0.007, "step": 47160 }, { "grad_norm": 0.12101606279611588, "learning_rate": 1.1990442226184695e-05, "loss": 0.0088, "step": 47170 }, { "grad_norm": 0.11579109728336334, "learning_rate": 1.1972543692381066e-05, "loss": 0.0085, "step": 47180 }, { "grad_norm": 0.10097382962703705, "learning_rate": 1.1954656710314576e-05, "loss": 0.0075, "step": 47190 }, { "grad_norm": 0.10343515872955322, "learning_rate": 1.1936781285418875e-05, "loss": 0.0072, "step": 47200 }, { "grad_norm": 0.09662214666604996, "learning_rate": 1.1918917423123993e-05, "loss": 0.0082, "step": 47210 }, { "grad_norm": 0.0717419981956482, "learning_rate": 1.1901065128856537e-05, "loss": 0.0075, "step": 47220 }, { "grad_norm": 0.07614130526781082, "learning_rate": 1.1883224408039551e-05, "loss": 0.0081, "step": 47230 }, { "grad_norm": 0.10732440650463104, "learning_rate": 1.1865395266092578e-05, "loss": 0.007, "step": 47240 }, { "grad_norm": 0.08892860263586044, "learning_rate": 1.1847577708431633e-05, "loss": 0.0073, "step": 47250 }, { "grad_norm": 0.11328285932540894, "learning_rate": 1.1829771740469225e-05, "loss": 0.0069, "step": 47260 }, { "grad_norm": 0.09025495499372482, "learning_rate": 1.1811977367614324e-05, "loss": 0.0081, "step": 47270 }, { "grad_norm": 0.12224732339382172, "learning_rate": 1.1794194595272412e-05, "loss": 0.0069, "step": 47280 }, { "grad_norm": 0.10017135739326477, "learning_rate": 1.1776423428845423e-05, "loss": 0.008, "step": 47290 }, { "grad_norm": 0.07890235632658005, "learning_rate": 1.1758663873731756e-05, "loss": 0.0072, "step": 47300 }, { "grad_norm": 0.08199049532413483, "learning_rate": 1.1740915935326302e-05, "loss": 0.0066, "step": 47310 }, { "grad_norm": 0.08959024399518967, "learning_rate": 1.1723179619020396e-05, "loss": 0.0076, "step": 47320 }, { "grad_norm": 0.09780322760343552, "learning_rate": 1.1705454930201914e-05, "loss": 0.0077, "step": 47330 }, { "grad_norm": 0.0864727571606636, "learning_rate": 1.1687741874255087e-05, "loss": 0.0073, "step": 47340 }, { "grad_norm": 0.09894942492246628, "learning_rate": 1.1670040456560728e-05, "loss": 0.0071, "step": 47350 }, { "grad_norm": 0.08115042001008987, "learning_rate": 1.1652350682496005e-05, "loss": 0.0075, "step": 47360 }, { "grad_norm": 0.09564850479364395, "learning_rate": 1.163467255743465e-05, "loss": 0.008, "step": 47370 }, { "grad_norm": 0.08341984450817108, "learning_rate": 1.1617006086746796e-05, "loss": 0.008, "step": 47380 }, { "grad_norm": 0.10967917740345001, "learning_rate": 1.1599351275799047e-05, "loss": 0.0085, "step": 47390 }, { "grad_norm": 0.09869559854269028, "learning_rate": 1.1581708129954466e-05, "loss": 0.0071, "step": 47400 }, { "grad_norm": 0.09728734195232391, "learning_rate": 1.1564076654572587e-05, "loss": 0.0078, "step": 47410 }, { "grad_norm": 0.09066098928451538, "learning_rate": 1.1546456855009358e-05, "loss": 0.0064, "step": 47420 }, { "grad_norm": 0.12405041605234146, "learning_rate": 1.1528848736617248e-05, "loss": 0.008, "step": 47430 }, { "grad_norm": 0.08042220026254654, "learning_rate": 1.1511252304745112e-05, "loss": 0.0091, "step": 47440 }, { "grad_norm": 0.10464373975992203, "learning_rate": 1.1493667564738297e-05, "loss": 0.0071, "step": 47450 }, { "grad_norm": 0.1065056249499321, "learning_rate": 1.1476094521938574e-05, "loss": 0.0075, "step": 47460 }, { "grad_norm": 0.1160193458199501, "learning_rate": 1.1458533181684167e-05, "loss": 0.0077, "step": 47470 }, { "grad_norm": 0.08490706980228424, "learning_rate": 1.1440983549309753e-05, "loss": 0.0073, "step": 47480 }, { "grad_norm": 0.09067990630865097, "learning_rate": 1.1423445630146434e-05, "loss": 0.009, "step": 47490 }, { "grad_norm": 0.08043748885393143, "learning_rate": 1.1405919429521799e-05, "loss": 0.0083, "step": 47500 }, { "grad_norm": 0.12170736491680145, "learning_rate": 1.1388404952759802e-05, "loss": 0.0084, "step": 47510 }, { "grad_norm": 0.13587550818920135, "learning_rate": 1.1370902205180923e-05, "loss": 0.0084, "step": 47520 }, { "grad_norm": 0.09665148705244064, "learning_rate": 1.1353411192101987e-05, "loss": 0.0077, "step": 47530 }, { "grad_norm": 0.12647727131843567, "learning_rate": 1.133593191883634e-05, "loss": 0.0072, "step": 47540 }, { "grad_norm": 0.10903589427471161, "learning_rate": 1.1318464390693711e-05, "loss": 0.0087, "step": 47550 }, { "grad_norm": 0.09228506684303284, "learning_rate": 1.1301008612980257e-05, "loss": 0.0073, "step": 47560 }, { "grad_norm": 0.07134042680263519, "learning_rate": 1.128356459099863e-05, "loss": 0.0071, "step": 47570 }, { "grad_norm": 0.10089093446731567, "learning_rate": 1.1266132330047802e-05, "loss": 0.0095, "step": 47580 }, { "grad_norm": 0.1629858762025833, "learning_rate": 1.1248711835423281e-05, "loss": 0.0089, "step": 47590 }, { "grad_norm": 0.07872585207223892, "learning_rate": 1.123130311241693e-05, "loss": 0.0082, "step": 47600 }, { "grad_norm": 0.1002306193113327, "learning_rate": 1.1213906166317068e-05, "loss": 0.0075, "step": 47610 }, { "grad_norm": 0.09616225212812424, "learning_rate": 1.1196521002408427e-05, "loss": 0.0081, "step": 47620 }, { "grad_norm": 0.08740242570638657, "learning_rate": 1.1179147625972159e-05, "loss": 0.009, "step": 47630 }, { "grad_norm": 0.14309640228748322, "learning_rate": 1.1161786042285822e-05, "loss": 0.0072, "step": 47640 }, { "grad_norm": 0.12337392568588257, "learning_rate": 1.1144436256623447e-05, "loss": 0.0075, "step": 47650 }, { "grad_norm": 0.09112820029258728, "learning_rate": 1.1127098274255392e-05, "loss": 0.0081, "step": 47660 }, { "grad_norm": 0.08995769172906876, "learning_rate": 1.1109772100448512e-05, "loss": 0.0082, "step": 47670 }, { "grad_norm": 0.08884314447641373, "learning_rate": 1.1092457740466033e-05, "loss": 0.0075, "step": 47680 }, { "grad_norm": 0.0960092544555664, "learning_rate": 1.10751551995676e-05, "loss": 0.0076, "step": 47690 }, { "grad_norm": 0.08888611942529678, "learning_rate": 1.1057864483009262e-05, "loss": 0.0077, "step": 47700 }, { "grad_norm": 0.10844328999519348, "learning_rate": 1.1040585596043473e-05, "loss": 0.0069, "step": 47710 }, { "grad_norm": 0.1366826295852661, "learning_rate": 1.1023318543919148e-05, "loss": 0.0079, "step": 47720 }, { "grad_norm": 0.09837433695793152, "learning_rate": 1.10060633318815e-05, "loss": 0.0081, "step": 47730 }, { "grad_norm": 0.18845577538013458, "learning_rate": 1.0988819965172248e-05, "loss": 0.0073, "step": 47740 }, { "grad_norm": 0.2090606689453125, "learning_rate": 1.0971588449029462e-05, "loss": 0.0074, "step": 47750 }, { "grad_norm": 0.09195965528488159, "learning_rate": 1.095436878868762e-05, "loss": 0.0069, "step": 47760 }, { "grad_norm": 0.09950554370880127, "learning_rate": 1.0937160989377598e-05, "loss": 0.0079, "step": 47770 }, { "grad_norm": 0.07529947906732559, "learning_rate": 1.0919965056326676e-05, "loss": 0.0069, "step": 47780 }, { "grad_norm": 0.08743876218795776, "learning_rate": 1.0902780994758504e-05, "loss": 0.0076, "step": 47790 }, { "grad_norm": 0.10091172903776169, "learning_rate": 1.0885608809893193e-05, "loss": 0.0073, "step": 47800 }, { "grad_norm": 0.10079560428857803, "learning_rate": 1.0868448506947142e-05, "loss": 0.0077, "step": 47810 }, { "grad_norm": 0.08829555660486221, "learning_rate": 1.0851300091133243e-05, "loss": 0.0086, "step": 47820 }, { "grad_norm": 0.09072194248437881, "learning_rate": 1.083416356766071e-05, "loss": 0.0073, "step": 47830 }, { "grad_norm": 0.13030940294265747, "learning_rate": 1.0817038941735175e-05, "loss": 0.0075, "step": 47840 }, { "grad_norm": 0.10398148745298386, "learning_rate": 1.0799926218558642e-05, "loss": 0.0067, "step": 47850 }, { "grad_norm": 0.10847322642803192, "learning_rate": 1.0782825403329488e-05, "loss": 0.0078, "step": 47860 }, { "grad_norm": 0.09136518836021423, "learning_rate": 1.076573650124254e-05, "loss": 0.0085, "step": 47870 }, { "grad_norm": 0.06728344410657883, "learning_rate": 1.0748659517488891e-05, "loss": 0.0066, "step": 47880 }, { "grad_norm": 0.08563566207885742, "learning_rate": 1.0731594457256138e-05, "loss": 0.0085, "step": 47890 }, { "grad_norm": 0.07354899495840073, "learning_rate": 1.0714541325728139e-05, "loss": 0.0092, "step": 47900 }, { "grad_norm": 0.07887601107358932, "learning_rate": 1.0697500128085231e-05, "loss": 0.0082, "step": 47910 }, { "grad_norm": 0.09529290348291397, "learning_rate": 1.0680470869504055e-05, "loss": 0.0075, "step": 47920 }, { "grad_norm": 0.09145822376012802, "learning_rate": 1.066345355515766e-05, "loss": 0.0083, "step": 47930 }, { "grad_norm": 0.07119141519069672, "learning_rate": 1.0646448190215453e-05, "loss": 0.0078, "step": 47940 }, { "grad_norm": 0.09827467799186707, "learning_rate": 1.0629454779843217e-05, "loss": 0.0077, "step": 47950 }, { "grad_norm": 0.09813418984413147, "learning_rate": 1.0612473329203082e-05, "loss": 0.0073, "step": 47960 }, { "grad_norm": 0.13286833465099335, "learning_rate": 1.0595503843453596e-05, "loss": 0.0079, "step": 47970 }, { "grad_norm": 0.09725504368543625, "learning_rate": 1.0578546327749634e-05, "loss": 0.0072, "step": 47980 }, { "grad_norm": 0.08370431512594223, "learning_rate": 1.0561600787242425e-05, "loss": 0.007, "step": 47990 }, { "grad_norm": 0.11226341128349304, "learning_rate": 1.0544667227079591e-05, "loss": 0.0074, "step": 48000 }, { "grad_norm": 0.14705833792686462, "learning_rate": 1.0527745652405085e-05, "loss": 0.0076, "step": 48010 }, { "grad_norm": 0.08551838994026184, "learning_rate": 1.051083606835927e-05, "loss": 0.0078, "step": 48020 }, { "grad_norm": 0.07082061469554901, "learning_rate": 1.049393848007878e-05, "loss": 0.0069, "step": 48030 }, { "grad_norm": 0.08791515231132507, "learning_rate": 1.0477052892696709e-05, "loss": 0.0076, "step": 48040 }, { "grad_norm": 0.06583862751722336, "learning_rate": 1.0460179311342394e-05, "loss": 0.0063, "step": 48050 }, { "grad_norm": 0.07247745245695114, "learning_rate": 1.0443317741141634e-05, "loss": 0.0088, "step": 48060 }, { "grad_norm": 0.0981542319059372, "learning_rate": 1.0426468187216514e-05, "loss": 0.0075, "step": 48070 }, { "grad_norm": 0.0837881937623024, "learning_rate": 1.0409630654685477e-05, "loss": 0.0071, "step": 48080 }, { "grad_norm": 0.0745236873626709, "learning_rate": 1.039280514866332e-05, "loss": 0.0067, "step": 48090 }, { "grad_norm": 0.07392985373735428, "learning_rate": 1.0375991674261198e-05, "loss": 0.0076, "step": 48100 }, { "grad_norm": 0.07306677848100662, "learning_rate": 1.0359190236586575e-05, "loss": 0.0082, "step": 48110 }, { "grad_norm": 0.10318374633789062, "learning_rate": 1.0342400840743322e-05, "loss": 0.0071, "step": 48120 }, { "grad_norm": 0.08596009016036987, "learning_rate": 1.0325623491831593e-05, "loss": 0.0065, "step": 48130 }, { "grad_norm": 0.10693556815385818, "learning_rate": 1.0308858194947906e-05, "loss": 0.0077, "step": 48140 }, { "grad_norm": 0.10292617976665497, "learning_rate": 1.0292104955185111e-05, "loss": 0.0081, "step": 48150 }, { "grad_norm": 0.08130055665969849, "learning_rate": 1.0275363777632396e-05, "loss": 0.0076, "step": 48160 }, { "grad_norm": 0.11242352426052094, "learning_rate": 1.0258634667375321e-05, "loss": 0.008, "step": 48170 }, { "grad_norm": 0.10144525021314621, "learning_rate": 1.02419176294957e-05, "loss": 0.008, "step": 48180 }, { "grad_norm": 0.15333238244056702, "learning_rate": 1.0225212669071782e-05, "loss": 0.0071, "step": 48190 }, { "grad_norm": 0.14725419878959656, "learning_rate": 1.0208519791178029e-05, "loss": 0.0069, "step": 48200 }, { "grad_norm": 0.08082422614097595, "learning_rate": 1.019183900088535e-05, "loss": 0.0077, "step": 48210 }, { "grad_norm": 0.09888952970504761, "learning_rate": 1.0175170303260906e-05, "loss": 0.0079, "step": 48220 }, { "grad_norm": 0.10619301348924637, "learning_rate": 1.0158513703368206e-05, "loss": 0.0082, "step": 48230 }, { "grad_norm": 0.0891444981098175, "learning_rate": 1.0141869206267095e-05, "loss": 0.0074, "step": 48240 }, { "grad_norm": 0.0972660481929779, "learning_rate": 1.0125236817013723e-05, "loss": 0.0081, "step": 48250 }, { "grad_norm": 0.08850546926259995, "learning_rate": 1.010861654066056e-05, "loss": 0.0071, "step": 48260 }, { "grad_norm": 0.10474809259176254, "learning_rate": 1.0092008382256434e-05, "loss": 0.008, "step": 48270 }, { "grad_norm": 0.10276864469051361, "learning_rate": 1.0075412346846458e-05, "loss": 0.0074, "step": 48280 }, { "grad_norm": 0.11557828634977341, "learning_rate": 1.0058828439472056e-05, "loss": 0.0091, "step": 48290 }, { "grad_norm": 0.06753983348608017, "learning_rate": 1.0042256665170996e-05, "loss": 0.0078, "step": 48300 }, { "grad_norm": 0.07971667498350143, "learning_rate": 1.0025697028977332e-05, "loss": 0.0078, "step": 48310 }, { "grad_norm": 0.1029374748468399, "learning_rate": 1.0009149535921454e-05, "loss": 0.0094, "step": 48320 }, { "grad_norm": 0.10586707293987274, "learning_rate": 9.992614191030031e-06, "loss": 0.0076, "step": 48330 }, { "grad_norm": 0.08666208386421204, "learning_rate": 9.976090999326115e-06, "loss": 0.0068, "step": 48340 }, { "grad_norm": 0.09924063831567764, "learning_rate": 9.959579965828952e-06, "loss": 0.0073, "step": 48350 }, { "grad_norm": 0.09375698119401932, "learning_rate": 9.943081095554218e-06, "loss": 0.0069, "step": 48360 }, { "grad_norm": 0.06902839243412018, "learning_rate": 9.926594393513783e-06, "loss": 0.0076, "step": 48370 }, { "grad_norm": 0.08504469692707062, "learning_rate": 9.910119864715906e-06, "loss": 0.0073, "step": 48380 }, { "grad_norm": 0.11163599044084549, "learning_rate": 9.8936575141651e-06, "loss": 0.0072, "step": 48390 }, { "grad_norm": 0.09219501912593842, "learning_rate": 9.877207346862194e-06, "loss": 0.0069, "step": 48400 }, { "grad_norm": 0.12276306003332138, "learning_rate": 9.860769367804312e-06, "loss": 0.0082, "step": 48410 }, { "grad_norm": 0.0928073301911354, "learning_rate": 9.844343581984877e-06, "loss": 0.0072, "step": 48420 }, { "grad_norm": 0.10104979574680328, "learning_rate": 9.82792999439362e-06, "loss": 0.0079, "step": 48430 }, { "grad_norm": 0.10265081375837326, "learning_rate": 9.811528610016546e-06, "loss": 0.0073, "step": 48440 }, { "grad_norm": 0.09275677800178528, "learning_rate": 9.79513943383597e-06, "loss": 0.0087, "step": 48450 }, { "grad_norm": 0.10595712065696716, "learning_rate": 9.778762470830489e-06, "loss": 0.0067, "step": 48460 }, { "grad_norm": 0.10018648207187653, "learning_rate": 9.762397725974982e-06, "loss": 0.0069, "step": 48470 }, { "grad_norm": 0.10709920525550842, "learning_rate": 9.746045204240622e-06, "loss": 0.008, "step": 48480 }, { "grad_norm": 0.08321667462587357, "learning_rate": 9.729704910594917e-06, "loss": 0.0072, "step": 48490 }, { "grad_norm": 0.12651894986629486, "learning_rate": 9.713376850001554e-06, "loss": 0.008, "step": 48500 }, { "grad_norm": 0.1377585530281067, "learning_rate": 9.697061027420622e-06, "loss": 0.0085, "step": 48510 }, { "grad_norm": 0.06786521524190903, "learning_rate": 9.680757447808385e-06, "loss": 0.0079, "step": 48520 }, { "grad_norm": 0.08453483134508133, "learning_rate": 9.664466116117488e-06, "loss": 0.0079, "step": 48530 }, { "grad_norm": 0.148398756980896, "learning_rate": 9.64818703729678e-06, "loss": 0.0068, "step": 48540 }, { "grad_norm": 0.09801020473241806, "learning_rate": 9.631920216291423e-06, "loss": 0.0085, "step": 48550 }, { "grad_norm": 0.06952415406703949, "learning_rate": 9.615665658042849e-06, "loss": 0.0079, "step": 48560 }, { "grad_norm": 0.11901947855949402, "learning_rate": 9.599423367488747e-06, "loss": 0.008, "step": 48570 }, { "grad_norm": 0.1305549442768097, "learning_rate": 9.583193349563124e-06, "loss": 0.0077, "step": 48580 }, { "grad_norm": 0.07759435474872589, "learning_rate": 9.566975609196216e-06, "loss": 0.0072, "step": 48590 }, { "grad_norm": 0.1073249876499176, "learning_rate": 9.550770151314548e-06, "loss": 0.0081, "step": 48600 }, { "grad_norm": 0.0887698382139206, "learning_rate": 9.53457698084091e-06, "loss": 0.0077, "step": 48610 }, { "grad_norm": 0.11318765580654144, "learning_rate": 9.518396102694355e-06, "loss": 0.0082, "step": 48620 }, { "grad_norm": 0.13316088914871216, "learning_rate": 9.502227521790198e-06, "loss": 0.0081, "step": 48630 }, { "grad_norm": 0.09828708320856094, "learning_rate": 9.486071243040063e-06, "loss": 0.0073, "step": 48640 }, { "grad_norm": 0.11205976456403732, "learning_rate": 9.469927271351747e-06, "loss": 0.009, "step": 48650 }, { "grad_norm": 0.09473271667957306, "learning_rate": 9.453795611629419e-06, "loss": 0.007, "step": 48660 }, { "grad_norm": 0.09183043241500854, "learning_rate": 9.437676268773399e-06, "loss": 0.0067, "step": 48670 }, { "grad_norm": 0.10311179608106613, "learning_rate": 9.421569247680357e-06, "loss": 0.007, "step": 48680 }, { "grad_norm": 0.08754825592041016, "learning_rate": 9.40547455324316e-06, "loss": 0.0071, "step": 48690 }, { "grad_norm": 0.08613447099924088, "learning_rate": 9.389392190350965e-06, "loss": 0.0079, "step": 48700 }, { "grad_norm": 0.08976814150810242, "learning_rate": 9.373322163889153e-06, "loss": 0.0072, "step": 48710 }, { "grad_norm": 0.0768485963344574, "learning_rate": 9.357264478739375e-06, "loss": 0.0078, "step": 48720 }, { "grad_norm": 0.09123486280441284, "learning_rate": 9.341219139779567e-06, "loss": 0.0077, "step": 48730 }, { "grad_norm": 0.06887520849704742, "learning_rate": 9.325186151883824e-06, "loss": 0.0066, "step": 48740 }, { "grad_norm": 0.10201363265514374, "learning_rate": 9.30916551992258e-06, "loss": 0.0081, "step": 48750 }, { "grad_norm": 0.08039131760597229, "learning_rate": 9.293157248762479e-06, "loss": 0.0082, "step": 48760 }, { "grad_norm": 0.11626965552568436, "learning_rate": 9.2771613432664e-06, "loss": 0.0075, "step": 48770 }, { "grad_norm": 0.1129831075668335, "learning_rate": 9.261177808293481e-06, "loss": 0.0093, "step": 48780 }, { "grad_norm": 0.18270833790302277, "learning_rate": 9.245206648699096e-06, "loss": 0.0086, "step": 48790 }, { "grad_norm": 0.09663613885641098, "learning_rate": 9.22924786933485e-06, "loss": 0.0087, "step": 48800 }, { "grad_norm": 0.09907644242048264, "learning_rate": 9.213301475048642e-06, "loss": 0.008, "step": 48810 }, { "grad_norm": 0.08068189024925232, "learning_rate": 9.197367470684504e-06, "loss": 0.0071, "step": 48820 }, { "grad_norm": 0.09733671694993973, "learning_rate": 9.181445861082816e-06, "loss": 0.0082, "step": 48830 }, { "grad_norm": 0.07405564934015274, "learning_rate": 9.16553665108012e-06, "loss": 0.0074, "step": 48840 }, { "grad_norm": 0.09006550908088684, "learning_rate": 9.149639845509223e-06, "loss": 0.0077, "step": 48850 }, { "grad_norm": 0.11555993556976318, "learning_rate": 9.133755449199144e-06, "loss": 0.0073, "step": 48860 }, { "grad_norm": 0.07177426666021347, "learning_rate": 9.117883466975135e-06, "loss": 0.0064, "step": 48870 }, { "grad_norm": 0.08033198118209839, "learning_rate": 9.10202390365873e-06, "loss": 0.0081, "step": 48880 }, { "grad_norm": 0.08269289880990982, "learning_rate": 9.086176764067583e-06, "loss": 0.0072, "step": 48890 }, { "grad_norm": 0.10300498455762863, "learning_rate": 9.070342053015684e-06, "loss": 0.007, "step": 48900 }, { "grad_norm": 0.08850324898958206, "learning_rate": 9.054519775313187e-06, "loss": 0.0074, "step": 48910 }, { "grad_norm": 0.08488582819700241, "learning_rate": 9.038709935766476e-06, "loss": 0.0072, "step": 48920 }, { "grad_norm": 0.11519724875688553, "learning_rate": 9.02291253917817e-06, "loss": 0.0083, "step": 48930 }, { "grad_norm": 0.08711555600166321, "learning_rate": 9.007127590347091e-06, "loss": 0.0068, "step": 48940 }, { "grad_norm": 0.07978121936321259, "learning_rate": 8.991355094068288e-06, "loss": 0.0067, "step": 48950 }, { "grad_norm": 0.07234961539506912, "learning_rate": 8.975595055133062e-06, "loss": 0.0074, "step": 48960 }, { "grad_norm": 0.07542755454778671, "learning_rate": 8.959847478328848e-06, "loss": 0.0059, "step": 48970 }, { "grad_norm": 0.08797026425600052, "learning_rate": 8.944112368439378e-06, "loss": 0.0076, "step": 48980 }, { "grad_norm": 0.08005770295858383, "learning_rate": 8.928389730244552e-06, "loss": 0.0071, "step": 48990 }, { "grad_norm": 0.09686636179685593, "learning_rate": 8.912679568520494e-06, "loss": 0.0068, "step": 49000 }, { "grad_norm": 0.08501020818948746, "learning_rate": 8.896981888039534e-06, "loss": 0.0075, "step": 49010 }, { "grad_norm": 0.11157085001468658, "learning_rate": 8.881296693570201e-06, "loss": 0.0107, "step": 49020 }, { "grad_norm": 0.12437169998884201, "learning_rate": 8.865623989877281e-06, "loss": 0.0073, "step": 49030 }, { "grad_norm": 0.16260984539985657, "learning_rate": 8.849963781721681e-06, "loss": 0.0073, "step": 49040 }, { "grad_norm": 0.20308364927768707, "learning_rate": 8.834316073860588e-06, "loss": 0.0086, "step": 49050 }, { "grad_norm": 0.10268037021160126, "learning_rate": 8.818680871047357e-06, "loss": 0.0083, "step": 49060 }, { "grad_norm": 0.07524459064006805, "learning_rate": 8.803058178031549e-06, "loss": 0.0071, "step": 49070 }, { "grad_norm": 0.07795002311468124, "learning_rate": 8.787447999558922e-06, "loss": 0.0073, "step": 49080 }, { "grad_norm": 0.09746777266263962, "learning_rate": 8.77185034037144e-06, "loss": 0.0073, "step": 49090 }, { "grad_norm": 0.07950593531131744, "learning_rate": 8.756265205207259e-06, "loss": 0.0075, "step": 49100 }, { "grad_norm": 0.09284735471010208, "learning_rate": 8.740692598800732e-06, "loss": 0.0078, "step": 49110 }, { "grad_norm": 0.09122414886951447, "learning_rate": 8.72513252588239e-06, "loss": 0.0075, "step": 49120 }, { "grad_norm": 0.09416573494672775, "learning_rate": 8.709584991178998e-06, "loss": 0.0067, "step": 49130 }, { "grad_norm": 0.07217219471931458, "learning_rate": 8.694049999413479e-06, "loss": 0.0067, "step": 49140 }, { "grad_norm": 0.08505766838788986, "learning_rate": 8.678527555304945e-06, "loss": 0.0073, "step": 49150 }, { "grad_norm": 0.09120585769414902, "learning_rate": 8.663017663568712e-06, "loss": 0.0072, "step": 49160 }, { "grad_norm": 0.09942683577537537, "learning_rate": 8.647520328916259e-06, "loss": 0.0075, "step": 49170 }, { "grad_norm": 0.097938172519207, "learning_rate": 8.632035556055307e-06, "loss": 0.0072, "step": 49180 }, { "grad_norm": 0.07197552174329758, "learning_rate": 8.616563349689672e-06, "loss": 0.007, "step": 49190 }, { "grad_norm": 0.08749597519636154, "learning_rate": 8.601103714519448e-06, "loss": 0.0085, "step": 49200 }, { "grad_norm": 0.06893855333328247, "learning_rate": 8.58565665524082e-06, "loss": 0.009, "step": 49210 }, { "grad_norm": 0.058028582483530045, "learning_rate": 8.570222176546222e-06, "loss": 0.0062, "step": 49220 }, { "grad_norm": 0.06618591398000717, "learning_rate": 8.554800283124242e-06, "loss": 0.0076, "step": 49230 }, { "grad_norm": 0.11963631957769394, "learning_rate": 8.539390979659639e-06, "loss": 0.0075, "step": 49240 }, { "grad_norm": 0.12096055597066879, "learning_rate": 8.523994270833352e-06, "loss": 0.0071, "step": 49250 }, { "grad_norm": 0.09713368117809296, "learning_rate": 8.5086101613225e-06, "loss": 0.0073, "step": 49260 }, { "grad_norm": 0.08314722031354904, "learning_rate": 8.493238655800346e-06, "loss": 0.0088, "step": 49270 }, { "grad_norm": 0.0912584513425827, "learning_rate": 8.47787975893638e-06, "loss": 0.0073, "step": 49280 }, { "grad_norm": 0.08144976943731308, "learning_rate": 8.462533475396211e-06, "loss": 0.0084, "step": 49290 }, { "grad_norm": 0.17707374691963196, "learning_rate": 8.447199809841643e-06, "loss": 0.0078, "step": 49300 }, { "grad_norm": 0.12160832434892654, "learning_rate": 8.431878766930635e-06, "loss": 0.0063, "step": 49310 }, { "grad_norm": 0.08210331201553345, "learning_rate": 8.416570351317304e-06, "loss": 0.0068, "step": 49320 }, { "grad_norm": 0.09489522129297256, "learning_rate": 8.401274567651973e-06, "loss": 0.0079, "step": 49330 }, { "grad_norm": 0.08995247632265091, "learning_rate": 8.385991420581058e-06, "loss": 0.0072, "step": 49340 }, { "grad_norm": 0.09784683585166931, "learning_rate": 8.370720914747215e-06, "loss": 0.007, "step": 49350 }, { "grad_norm": 0.0817648321390152, "learning_rate": 8.355463054789181e-06, "loss": 0.0078, "step": 49360 }, { "grad_norm": 0.08435776084661484, "learning_rate": 8.340217845341919e-06, "loss": 0.0071, "step": 49370 }, { "grad_norm": 0.08279773592948914, "learning_rate": 8.324985291036514e-06, "loss": 0.0071, "step": 49380 }, { "grad_norm": 0.0978565663099289, "learning_rate": 8.309765396500213e-06, "loss": 0.0068, "step": 49390 }, { "grad_norm": 0.06961297243833542, "learning_rate": 8.294558166356419e-06, "loss": 0.007, "step": 49400 }, { "grad_norm": 0.11139857769012451, "learning_rate": 8.279363605224683e-06, "loss": 0.008, "step": 49410 }, { "grad_norm": 0.09599832445383072, "learning_rate": 8.264181717720704e-06, "loss": 0.007, "step": 49420 }, { "grad_norm": 0.07898857444524765, "learning_rate": 8.249012508456361e-06, "loss": 0.0067, "step": 49430 }, { "grad_norm": 0.11400584131479263, "learning_rate": 8.233855982039646e-06, "loss": 0.0077, "step": 49440 }, { "grad_norm": 0.09155984967947006, "learning_rate": 8.218712143074708e-06, "loss": 0.0077, "step": 49450 }, { "grad_norm": 0.1308976709842682, "learning_rate": 8.203580996161858e-06, "loss": 0.0069, "step": 49460 }, { "grad_norm": 0.08733382076025009, "learning_rate": 8.188462545897512e-06, "loss": 0.0069, "step": 49470 }, { "grad_norm": 0.07364660501480103, "learning_rate": 8.173356796874304e-06, "loss": 0.0073, "step": 49480 }, { "grad_norm": 0.10970933735370636, "learning_rate": 8.158263753680906e-06, "loss": 0.0084, "step": 49490 }, { "grad_norm": 0.07661843299865723, "learning_rate": 8.143183420902239e-06, "loss": 0.0067, "step": 49500 }, { "grad_norm": 0.09575986862182617, "learning_rate": 8.128115803119258e-06, "loss": 0.007, "step": 49510 }, { "grad_norm": 0.12948782742023468, "learning_rate": 8.11306090490916e-06, "loss": 0.0067, "step": 49520 }, { "grad_norm": 0.0792299211025238, "learning_rate": 8.098018730845169e-06, "loss": 0.0063, "step": 49530 }, { "grad_norm": 0.06654148548841476, "learning_rate": 8.082989285496745e-06, "loss": 0.0072, "step": 49540 }, { "grad_norm": 0.11726550757884979, "learning_rate": 8.067972573429416e-06, "loss": 0.0087, "step": 49550 }, { "grad_norm": 0.14874686300754547, "learning_rate": 8.052968599204874e-06, "loss": 0.0073, "step": 49560 }, { "grad_norm": 0.11717138439416885, "learning_rate": 8.037977367380922e-06, "loss": 0.0077, "step": 49570 }, { "grad_norm": 0.09713300317525864, "learning_rate": 8.022998882511495e-06, "loss": 0.0076, "step": 49580 }, { "grad_norm": 0.07808985561132431, "learning_rate": 8.008033149146677e-06, "loss": 0.0086, "step": 49590 }, { "grad_norm": 0.11114270240068436, "learning_rate": 7.993080171832656e-06, "loss": 0.0075, "step": 49600 }, { "grad_norm": 0.06848088651895523, "learning_rate": 7.978139955111752e-06, "loss": 0.007, "step": 49610 }, { "grad_norm": 0.08171261847019196, "learning_rate": 7.9632125035224e-06, "loss": 0.0067, "step": 49620 }, { "grad_norm": 0.07789295166730881, "learning_rate": 7.948297821599177e-06, "loss": 0.0069, "step": 49630 }, { "grad_norm": 0.07947584241628647, "learning_rate": 7.933395913872755e-06, "loss": 0.008, "step": 49640 }, { "grad_norm": 0.07973094284534454, "learning_rate": 7.918506784869972e-06, "loss": 0.0071, "step": 49650 }, { "grad_norm": 0.0743638277053833, "learning_rate": 7.903630439113707e-06, "loss": 0.0073, "step": 49660 }, { "grad_norm": 0.09465686231851578, "learning_rate": 7.888766881123044e-06, "loss": 0.007, "step": 49670 }, { "grad_norm": 0.171615868806839, "learning_rate": 7.873916115413099e-06, "loss": 0.007, "step": 49680 }, { "grad_norm": 0.11173985153436661, "learning_rate": 7.85907814649518e-06, "loss": 0.0072, "step": 49690 }, { "grad_norm": 0.10208505392074585, "learning_rate": 7.844252978876649e-06, "loss": 0.0072, "step": 49700 }, { "grad_norm": 0.14011292159557343, "learning_rate": 7.829440617061001e-06, "loss": 0.0067, "step": 49710 }, { "grad_norm": 0.07746374607086182, "learning_rate": 7.814641065547851e-06, "loss": 0.0077, "step": 49720 }, { "grad_norm": 0.07682264596223831, "learning_rate": 7.79985432883289e-06, "loss": 0.0077, "step": 49730 }, { "grad_norm": 0.23086713254451752, "learning_rate": 7.78508041140797e-06, "loss": 0.0064, "step": 49740 }, { "grad_norm": 0.11293099820613861, "learning_rate": 7.770319317760993e-06, "loss": 0.0078, "step": 49750 }, { "grad_norm": 0.07092736661434174, "learning_rate": 7.755571052376004e-06, "loss": 0.0067, "step": 49760 }, { "grad_norm": 0.09284719824790955, "learning_rate": 7.740835619733128e-06, "loss": 0.0062, "step": 49770 }, { "grad_norm": 0.06584557890892029, "learning_rate": 7.726113024308601e-06, "loss": 0.0078, "step": 49780 }, { "grad_norm": 0.11729367822408676, "learning_rate": 7.711403270574746e-06, "loss": 0.0072, "step": 49790 }, { "grad_norm": 0.09176429361104965, "learning_rate": 7.696706363000039e-06, "loss": 0.0078, "step": 49800 }, { "grad_norm": 0.09507030248641968, "learning_rate": 7.682022306048959e-06, "loss": 0.007, "step": 49810 }, { "grad_norm": 0.09906809031963348, "learning_rate": 7.667351104182186e-06, "loss": 0.0086, "step": 49820 }, { "grad_norm": 0.06715523451566696, "learning_rate": 7.652692761856395e-06, "loss": 0.0087, "step": 49830 }, { "grad_norm": 0.11296674609184265, "learning_rate": 7.63804728352444e-06, "loss": 0.0082, "step": 49840 }, { "grad_norm": 0.0776083767414093, "learning_rate": 7.623414673635215e-06, "loss": 0.0077, "step": 49850 }, { "grad_norm": 0.06864812225103378, "learning_rate": 7.608794936633723e-06, "loss": 0.0079, "step": 49860 }, { "grad_norm": 0.0772215947508812, "learning_rate": 7.594188076961056e-06, "loss": 0.0072, "step": 49870 }, { "grad_norm": 0.12972265481948853, "learning_rate": 7.579594099054382e-06, "loss": 0.0074, "step": 49880 }, { "grad_norm": 0.07601156830787659, "learning_rate": 7.565013007346983e-06, "loss": 0.0071, "step": 49890 }, { "grad_norm": 0.08532393723726273, "learning_rate": 7.5504448062682035e-06, "loss": 0.0072, "step": 49900 }, { "grad_norm": 0.1080586388707161, "learning_rate": 7.53588950024347e-06, "loss": 0.0072, "step": 49910 }, { "grad_norm": 0.09874293953180313, "learning_rate": 7.5213470936943145e-06, "loss": 0.0084, "step": 49920 }, { "grad_norm": 0.1027517095208168, "learning_rate": 7.506817591038323e-06, "loss": 0.0065, "step": 49930 }, { "grad_norm": 0.11837929487228394, "learning_rate": 7.492300996689183e-06, "loss": 0.0084, "step": 49940 }, { "grad_norm": 0.1327010840177536, "learning_rate": 7.477797315056645e-06, "loss": 0.0085, "step": 49950 }, { "grad_norm": 0.10083869844675064, "learning_rate": 7.463306550546539e-06, "loss": 0.0061, "step": 49960 }, { "grad_norm": 0.10326127707958221, "learning_rate": 7.448828707560812e-06, "loss": 0.0071, "step": 49970 }, { "grad_norm": 0.07987620681524277, "learning_rate": 7.4343637904974e-06, "loss": 0.0067, "step": 49980 }, { "grad_norm": 0.08025246858596802, "learning_rate": 7.419911803750401e-06, "loss": 0.0075, "step": 49990 }, { "grad_norm": 0.13282868266105652, "learning_rate": 7.405472751709935e-06, "loss": 0.0069, "step": 50000 }, { "grad_norm": 0.1511332392692566, "learning_rate": 7.3910466387622e-06, "loss": 0.0078, "step": 50010 }, { "grad_norm": 0.08759816735982895, "learning_rate": 7.3766334692894735e-06, "loss": 0.008, "step": 50020 }, { "grad_norm": 0.07849559187889099, "learning_rate": 7.3622332476700865e-06, "loss": 0.0063, "step": 50030 }, { "grad_norm": 0.09641242772340775, "learning_rate": 7.347845978278472e-06, "loss": 0.0074, "step": 50040 }, { "grad_norm": 0.1729421764612198, "learning_rate": 7.333471665485065e-06, "loss": 0.0075, "step": 50050 }, { "grad_norm": 0.08117533475160599, "learning_rate": 7.31911031365643e-06, "loss": 0.008, "step": 50060 }, { "grad_norm": 0.08694131672382355, "learning_rate": 7.304761927155157e-06, "loss": 0.0082, "step": 50070 }, { "grad_norm": 0.08120147138834, "learning_rate": 7.29042651033991e-06, "loss": 0.0073, "step": 50080 }, { "grad_norm": 0.09353798627853394, "learning_rate": 7.276104067565409e-06, "loss": 0.0074, "step": 50090 }, { "grad_norm": 0.14872314035892487, "learning_rate": 7.261794603182431e-06, "loss": 0.0074, "step": 50100 }, { "grad_norm": 0.06662523746490479, "learning_rate": 7.24749812153781e-06, "loss": 0.0067, "step": 50110 }, { "grad_norm": 0.09572334587574005, "learning_rate": 7.2332146269744605e-06, "loss": 0.0078, "step": 50120 }, { "grad_norm": 0.08327328413724899, "learning_rate": 7.218944123831295e-06, "loss": 0.008, "step": 50130 }, { "grad_norm": 0.11309750378131866, "learning_rate": 7.204686616443351e-06, "loss": 0.0078, "step": 50140 }, { "grad_norm": 0.14688318967819214, "learning_rate": 7.190442109141665e-06, "loss": 0.0075, "step": 50150 }, { "grad_norm": 0.08945779502391815, "learning_rate": 7.176210606253347e-06, "loss": 0.0067, "step": 50160 }, { "grad_norm": 0.11497724056243896, "learning_rate": 7.161992112101551e-06, "loss": 0.0076, "step": 50170 }, { "grad_norm": 0.09693590551614761, "learning_rate": 7.147786631005465e-06, "loss": 0.0077, "step": 50180 }, { "grad_norm": 0.09532888978719711, "learning_rate": 7.1335941672803775e-06, "loss": 0.0082, "step": 50190 }, { "grad_norm": 0.10551656782627106, "learning_rate": 7.1194147252375384e-06, "loss": 0.009, "step": 50200 }, { "grad_norm": 0.11508921533823013, "learning_rate": 7.10524830918432e-06, "loss": 0.0065, "step": 50210 }, { "grad_norm": 0.0752260610461235, "learning_rate": 7.091094923424097e-06, "loss": 0.0075, "step": 50220 }, { "grad_norm": 0.08752299845218658, "learning_rate": 7.0769545722562894e-06, "loss": 0.0078, "step": 50230 }, { "grad_norm": 0.07636725902557373, "learning_rate": 7.0628272599763675e-06, "loss": 0.0066, "step": 50240 }, { "grad_norm": 0.11398456245660782, "learning_rate": 7.048712990875828e-06, "loss": 0.0062, "step": 50250 }, { "grad_norm": 0.08762766420841217, "learning_rate": 7.034611769242216e-06, "loss": 0.0069, "step": 50260 }, { "grad_norm": 0.13235361874103546, "learning_rate": 7.02052359935913e-06, "loss": 0.0081, "step": 50270 }, { "grad_norm": 0.09236205369234085, "learning_rate": 7.006448485506145e-06, "loss": 0.0068, "step": 50280 }, { "grad_norm": 0.0809023380279541, "learning_rate": 6.992386431958942e-06, "loss": 0.007, "step": 50290 }, { "grad_norm": 0.09602903574705124, "learning_rate": 6.978337442989197e-06, "loss": 0.0076, "step": 50300 }, { "grad_norm": 0.11067143827676773, "learning_rate": 6.964301522864608e-06, "loss": 0.0064, "step": 50310 }, { "grad_norm": 0.07426702976226807, "learning_rate": 6.950278675848926e-06, "loss": 0.0069, "step": 50320 }, { "grad_norm": 0.10365297645330429, "learning_rate": 6.9362689062019145e-06, "loss": 0.007, "step": 50330 }, { "grad_norm": 0.07545022666454315, "learning_rate": 6.922272218179393e-06, "loss": 0.0077, "step": 50340 }, { "grad_norm": 0.09602078050374985, "learning_rate": 6.908288616033148e-06, "loss": 0.0073, "step": 50350 }, { "grad_norm": 0.06549808382987976, "learning_rate": 6.894318104011077e-06, "loss": 0.0071, "step": 50360 }, { "grad_norm": 0.09616003930568695, "learning_rate": 6.880360686357007e-06, "loss": 0.0083, "step": 50370 }, { "grad_norm": 0.07569947093725204, "learning_rate": 6.8664163673108575e-06, "loss": 0.0069, "step": 50380 }, { "grad_norm": 0.07906121760606766, "learning_rate": 6.85248515110854e-06, "loss": 0.0075, "step": 50390 }, { "grad_norm": 0.10245776176452637, "learning_rate": 6.838567041981992e-06, "loss": 0.0069, "step": 50400 }, { "grad_norm": 0.09207752346992493, "learning_rate": 6.8246620441591634e-06, "loss": 0.0079, "step": 50410 }, { "grad_norm": 0.11170899122953415, "learning_rate": 6.8107701618640275e-06, "loss": 0.0073, "step": 50420 }, { "grad_norm": 0.1214391216635704, "learning_rate": 6.796891399316557e-06, "loss": 0.0067, "step": 50430 }, { "grad_norm": 0.1773514300584793, "learning_rate": 6.7830257607327804e-06, "loss": 0.0066, "step": 50440 }, { "grad_norm": 0.15533800423145294, "learning_rate": 6.7691732503247e-06, "loss": 0.0075, "step": 50450 }, { "grad_norm": 0.10176550596952438, "learning_rate": 6.755333872300346e-06, "loss": 0.0081, "step": 50460 }, { "grad_norm": 0.1143990308046341, "learning_rate": 6.741507630863747e-06, "loss": 0.0088, "step": 50470 }, { "grad_norm": 0.09177005290985107, "learning_rate": 6.727694530214945e-06, "loss": 0.0076, "step": 50480 }, { "grad_norm": 0.10194992274045944, "learning_rate": 6.713894574550028e-06, "loss": 0.0067, "step": 50490 }, { "grad_norm": 0.0960036888718605, "learning_rate": 6.700107768061015e-06, "loss": 0.0062, "step": 50500 }, { "grad_norm": 0.07170554250478745, "learning_rate": 6.686334114936016e-06, "loss": 0.0078, "step": 50510 }, { "grad_norm": 0.11033116281032562, "learning_rate": 6.672573619359063e-06, "loss": 0.0073, "step": 50520 }, { "grad_norm": 0.09158472716808319, "learning_rate": 6.658826285510256e-06, "loss": 0.0067, "step": 50530 }, { "grad_norm": 0.07814428955316544, "learning_rate": 6.645092117565666e-06, "loss": 0.0067, "step": 50540 }, { "grad_norm": 0.11598428338766098, "learning_rate": 6.631371119697371e-06, "loss": 0.0073, "step": 50550 }, { "grad_norm": 0.08964760601520538, "learning_rate": 6.6176632960734505e-06, "loss": 0.0073, "step": 50560 }, { "grad_norm": 0.09407909959554672, "learning_rate": 6.603968650857978e-06, "loss": 0.0085, "step": 50570 }, { "grad_norm": 0.0846467912197113, "learning_rate": 6.5902871882110085e-06, "loss": 0.0066, "step": 50580 }, { "grad_norm": 0.07180361449718475, "learning_rate": 6.576618912288635e-06, "loss": 0.0069, "step": 50590 }, { "grad_norm": 0.08272477239370346, "learning_rate": 6.562963827242913e-06, "loss": 0.0072, "step": 50600 }, { "grad_norm": 0.12045469880104065, "learning_rate": 6.549321937221886e-06, "loss": 0.0074, "step": 50610 }, { "grad_norm": 0.14023371040821075, "learning_rate": 6.5356932463696064e-06, "loss": 0.008, "step": 50620 }, { "grad_norm": 0.09749675542116165, "learning_rate": 6.522077758826101e-06, "loss": 0.0072, "step": 50630 }, { "grad_norm": 0.07075260579586029, "learning_rate": 6.5084754787274275e-06, "loss": 0.0073, "step": 50640 }, { "grad_norm": 0.1107567772269249, "learning_rate": 6.494886410205553e-06, "loss": 0.0079, "step": 50650 }, { "grad_norm": 0.0690002515912056, "learning_rate": 6.481310557388521e-06, "loss": 0.0064, "step": 50660 }, { "grad_norm": 0.08176928013563156, "learning_rate": 6.46774792440028e-06, "loss": 0.007, "step": 50670 }, { "grad_norm": 0.08105358481407166, "learning_rate": 6.4541985153608206e-06, "loss": 0.007, "step": 50680 }, { "grad_norm": 0.08161893486976624, "learning_rate": 6.4406623343861e-06, "loss": 0.0067, "step": 50690 }, { "grad_norm": 0.08060893416404724, "learning_rate": 6.427139385588038e-06, "loss": 0.0071, "step": 50700 }, { "grad_norm": 0.10963960736989975, "learning_rate": 6.413629673074561e-06, "loss": 0.0076, "step": 50710 }, { "grad_norm": 0.09419215470552444, "learning_rate": 6.400133200949554e-06, "loss": 0.0089, "step": 50720 }, { "grad_norm": 0.0559268519282341, "learning_rate": 6.386649973312897e-06, "loss": 0.007, "step": 50730 }, { "grad_norm": 0.18568755686283112, "learning_rate": 6.37317999426042e-06, "loss": 0.0067, "step": 50740 }, { "grad_norm": 0.09198205918073654, "learning_rate": 6.359723267883977e-06, "loss": 0.0075, "step": 50750 }, { "grad_norm": 0.09363270550966263, "learning_rate": 6.346279798271343e-06, "loss": 0.0081, "step": 50760 }, { "grad_norm": 0.07675231248140335, "learning_rate": 6.332849589506301e-06, "loss": 0.0074, "step": 50770 }, { "grad_norm": 0.0755520612001419, "learning_rate": 6.319432645668588e-06, "loss": 0.0083, "step": 50780 }, { "grad_norm": 0.09340973943471909, "learning_rate": 6.306028970833922e-06, "loss": 0.0074, "step": 50790 }, { "grad_norm": 0.0683051198720932, "learning_rate": 6.2926385690739665e-06, "loss": 0.0077, "step": 50800 }, { "grad_norm": 0.08662232011556625, "learning_rate": 6.279261444456413e-06, "loss": 0.0065, "step": 50810 }, { "grad_norm": 0.12897954881191254, "learning_rate": 6.265897601044829e-06, "loss": 0.0082, "step": 50820 }, { "grad_norm": 0.08884245902299881, "learning_rate": 6.2525470428988434e-06, "loss": 0.007, "step": 50830 }, { "grad_norm": 0.09758014231920242, "learning_rate": 6.239209774073962e-06, "loss": 0.0068, "step": 50840 }, { "grad_norm": 0.07775433361530304, "learning_rate": 6.225885798621728e-06, "loss": 0.0068, "step": 50850 }, { "grad_norm": 0.06901920586824417, "learning_rate": 6.2125751205895925e-06, "loss": 0.0079, "step": 50860 }, { "grad_norm": 0.09309474378824234, "learning_rate": 6.199277744020998e-06, "loss": 0.0074, "step": 50870 }, { "grad_norm": 0.07457337528467178, "learning_rate": 6.185993672955337e-06, "loss": 0.0074, "step": 50880 }, { "grad_norm": 0.09116276353597641, "learning_rate": 6.172722911427947e-06, "loss": 0.007, "step": 50890 }, { "grad_norm": 0.10521034151315689, "learning_rate": 6.159465463470149e-06, "loss": 0.0069, "step": 50900 }, { "grad_norm": 0.06128000095486641, "learning_rate": 6.146221333109204e-06, "loss": 0.0068, "step": 50910 }, { "grad_norm": 0.08510337024927139, "learning_rate": 6.132990524368326e-06, "loss": 0.0079, "step": 50920 }, { "grad_norm": 0.11301115900278091, "learning_rate": 6.119773041266685e-06, "loss": 0.0087, "step": 50930 }, { "grad_norm": 0.1488543450832367, "learning_rate": 6.106568887819402e-06, "loss": 0.0071, "step": 50940 }, { "grad_norm": 0.13358086347579956, "learning_rate": 6.093378068037548e-06, "loss": 0.0069, "step": 50950 }, { "grad_norm": 0.06691758334636688, "learning_rate": 6.080200585928164e-06, "loss": 0.0077, "step": 50960 }, { "grad_norm": 0.09570054709911346, "learning_rate": 6.06703644549419e-06, "loss": 0.0068, "step": 50970 }, { "grad_norm": 0.09456378221511841, "learning_rate": 6.053885650734576e-06, "loss": 0.0076, "step": 50980 }, { "grad_norm": 0.09842625260353088, "learning_rate": 6.040748205644153e-06, "loss": 0.007, "step": 50990 }, { "grad_norm": 0.12996666133403778, "learning_rate": 6.0276241142137646e-06, "loss": 0.0078, "step": 51000 }, { "grad_norm": 0.06688697636127472, "learning_rate": 6.014513380430142e-06, "loss": 0.0066, "step": 51010 }, { "grad_norm": 0.0909523218870163, "learning_rate": 6.001416008275984e-06, "loss": 0.0062, "step": 51020 }, { "grad_norm": 0.08399869501590729, "learning_rate": 5.988332001729929e-06, "loss": 0.0069, "step": 51030 }, { "grad_norm": 0.13594874739646912, "learning_rate": 5.975261364766543e-06, "loss": 0.007, "step": 51040 }, { "grad_norm": 0.1008608341217041, "learning_rate": 5.962204101356356e-06, "loss": 0.0072, "step": 51050 }, { "grad_norm": 0.10046559572219849, "learning_rate": 5.94916021546581e-06, "loss": 0.0081, "step": 51060 }, { "grad_norm": 0.09525258839130402, "learning_rate": 5.936129711057298e-06, "loss": 0.0065, "step": 51070 }, { "grad_norm": 0.10156415402889252, "learning_rate": 5.923112592089142e-06, "loss": 0.0073, "step": 51080 }, { "grad_norm": 0.08360809832811356, "learning_rate": 5.9101088625155954e-06, "loss": 0.0065, "step": 51090 }, { "grad_norm": 0.18873606622219086, "learning_rate": 5.897118526286843e-06, "loss": 0.0073, "step": 51100 }, { "grad_norm": 0.10846707224845886, "learning_rate": 5.884141587349035e-06, "loss": 0.0073, "step": 51110 }, { "grad_norm": 0.09156131744384766, "learning_rate": 5.871178049644177e-06, "loss": 0.0068, "step": 51120 }, { "grad_norm": 0.09618978202342987, "learning_rate": 5.858227917110293e-06, "loss": 0.0077, "step": 51130 }, { "grad_norm": 0.06494889408349991, "learning_rate": 5.845291193681252e-06, "loss": 0.0067, "step": 51140 }, { "grad_norm": 0.10279817879199982, "learning_rate": 5.832367883286921e-06, "loss": 0.0077, "step": 51150 }, { "grad_norm": 0.07721703499555588, "learning_rate": 5.81945798985305e-06, "loss": 0.0074, "step": 51160 }, { "grad_norm": 0.10857842117547989, "learning_rate": 5.806561517301306e-06, "loss": 0.0077, "step": 51170 }, { "grad_norm": 0.1026272401213646, "learning_rate": 5.793678469549335e-06, "loss": 0.007, "step": 51180 }, { "grad_norm": 0.06813972443342209, "learning_rate": 5.780808850510627e-06, "loss": 0.0069, "step": 51190 }, { "grad_norm": 0.1326109915971756, "learning_rate": 5.767952664094673e-06, "loss": 0.0068, "step": 51200 }, { "grad_norm": 0.09378644078969955, "learning_rate": 5.755109914206791e-06, "loss": 0.0071, "step": 51210 }, { "grad_norm": 0.12421488016843796, "learning_rate": 5.7422806047483125e-06, "loss": 0.0078, "step": 51220 }, { "grad_norm": 0.09531793743371964, "learning_rate": 5.72946473961643e-06, "loss": 0.0072, "step": 51230 }, { "grad_norm": 0.07702741771936417, "learning_rate": 5.716662322704264e-06, "loss": 0.0072, "step": 51240 }, { "grad_norm": 0.07792489230632782, "learning_rate": 5.703873357900852e-06, "loss": 0.0067, "step": 51250 }, { "grad_norm": 0.07594843953847885, "learning_rate": 5.691097849091143e-06, "loss": 0.0075, "step": 51260 }, { "grad_norm": 0.08051889389753342, "learning_rate": 5.678335800155982e-06, "loss": 0.0066, "step": 51270 }, { "grad_norm": 0.10361427068710327, "learning_rate": 5.665587214972174e-06, "loss": 0.0069, "step": 51280 }, { "grad_norm": 0.07237331569194794, "learning_rate": 5.652852097412386e-06, "loss": 0.007, "step": 51290 }, { "grad_norm": 0.10569741576910019, "learning_rate": 5.640130451345216e-06, "loss": 0.0071, "step": 51300 }, { "grad_norm": 0.09014610201120377, "learning_rate": 5.627422280635159e-06, "loss": 0.0068, "step": 51310 }, { "grad_norm": 0.06239914521574974, "learning_rate": 5.61472758914261e-06, "loss": 0.0063, "step": 51320 }, { "grad_norm": 0.07137764990329742, "learning_rate": 5.602046380723918e-06, "loss": 0.0066, "step": 51330 }, { "grad_norm": 0.10434781014919281, "learning_rate": 5.5893786592312535e-06, "loss": 0.0069, "step": 51340 }, { "grad_norm": 0.07951932400465012, "learning_rate": 5.576724428512775e-06, "loss": 0.0069, "step": 51350 }, { "grad_norm": 0.10324549674987793, "learning_rate": 5.56408369241247e-06, "loss": 0.0069, "step": 51360 }, { "grad_norm": 0.12864328920841217, "learning_rate": 5.5514564547702875e-06, "loss": 0.0068, "step": 51370 }, { "grad_norm": 0.09469643980264664, "learning_rate": 5.538842719422038e-06, "loss": 0.009, "step": 51380 }, { "grad_norm": 0.09133519232273102, "learning_rate": 5.52624249019944e-06, "loss": 0.0071, "step": 51390 }, { "grad_norm": 0.07444380223751068, "learning_rate": 5.51365577093011e-06, "loss": 0.0062, "step": 51400 }, { "grad_norm": 0.08257024735212326, "learning_rate": 5.501082565437565e-06, "loss": 0.0062, "step": 51410 }, { "grad_norm": 0.08074094355106354, "learning_rate": 5.488522877541202e-06, "loss": 0.0073, "step": 51420 }, { "grad_norm": 0.09301789104938507, "learning_rate": 5.475976711056341e-06, "loss": 0.0072, "step": 51430 }, { "grad_norm": 0.07320655882358551, "learning_rate": 5.463444069794166e-06, "loss": 0.0067, "step": 51440 }, { "grad_norm": 0.10537134855985641, "learning_rate": 5.4509249575617594e-06, "loss": 0.0073, "step": 51450 }, { "grad_norm": 0.09004320949316025, "learning_rate": 5.438419378162107e-06, "loss": 0.0069, "step": 51460 }, { "grad_norm": 0.08225300908088684, "learning_rate": 5.425927335394054e-06, "loss": 0.0072, "step": 51470 }, { "grad_norm": 0.08109220117330551, "learning_rate": 5.413448833052387e-06, "loss": 0.0075, "step": 51480 }, { "grad_norm": 0.06379739195108414, "learning_rate": 5.400983874927701e-06, "loss": 0.0066, "step": 51490 }, { "grad_norm": 0.08663645386695862, "learning_rate": 5.388532464806567e-06, "loss": 0.007, "step": 51500 }, { "grad_norm": 0.10736437141895294, "learning_rate": 5.3760946064713546e-06, "loss": 0.0068, "step": 51510 }, { "grad_norm": 0.0873037725687027, "learning_rate": 5.363670303700386e-06, "loss": 0.0065, "step": 51520 }, { "grad_norm": 0.1089303195476532, "learning_rate": 5.351259560267824e-06, "loss": 0.0069, "step": 51530 }, { "grad_norm": 0.1140020340681076, "learning_rate": 5.338862379943721e-06, "loss": 0.0095, "step": 51540 }, { "grad_norm": 0.10178012400865555, "learning_rate": 5.326478766494025e-06, "loss": 0.0064, "step": 51550 }, { "grad_norm": 0.08952657133340836, "learning_rate": 5.3141087236805385e-06, "loss": 0.0073, "step": 51560 }, { "grad_norm": 0.10291656851768494, "learning_rate": 5.3017522552609615e-06, "loss": 0.0072, "step": 51570 }, { "grad_norm": 0.0741816908121109, "learning_rate": 5.289409364988851e-06, "loss": 0.0066, "step": 51580 }, { "grad_norm": 0.07908357679843903, "learning_rate": 5.277080056613671e-06, "loss": 0.0072, "step": 51590 }, { "grad_norm": 0.06740886718034744, "learning_rate": 5.264764333880729e-06, "loss": 0.007, "step": 51600 }, { "grad_norm": 0.1465711146593094, "learning_rate": 5.252462200531216e-06, "loss": 0.0091, "step": 51610 }, { "grad_norm": 0.12497805058956146, "learning_rate": 5.240173660302194e-06, "loss": 0.0076, "step": 51620 }, { "grad_norm": 0.0885649174451828, "learning_rate": 5.2278987169266044e-06, "loss": 0.0082, "step": 51630 }, { "grad_norm": 0.108170285820961, "learning_rate": 5.215637374133231e-06, "loss": 0.0076, "step": 51640 }, { "grad_norm": 0.1686466485261917, "learning_rate": 5.203389635646782e-06, "loss": 0.0073, "step": 51650 }, { "grad_norm": 0.0657997652888298, "learning_rate": 5.191155505187756e-06, "loss": 0.0065, "step": 51660 }, { "grad_norm": 0.15335363149642944, "learning_rate": 5.1789349864726e-06, "loss": 0.0072, "step": 51670 }, { "grad_norm": 0.12219926714897156, "learning_rate": 5.16672808321354e-06, "loss": 0.0076, "step": 51680 }, { "grad_norm": 0.059420399367809296, "learning_rate": 5.154534799118749e-06, "loss": 0.0071, "step": 51690 }, { "grad_norm": 0.11239825189113617, "learning_rate": 5.142355137892207e-06, "loss": 0.0069, "step": 51700 }, { "grad_norm": 0.10076773166656494, "learning_rate": 5.130189103233779e-06, "loss": 0.0072, "step": 51710 }, { "grad_norm": 0.08456834405660629, "learning_rate": 5.118036698839179e-06, "loss": 0.0071, "step": 51720 }, { "grad_norm": 0.07382411509752274, "learning_rate": 5.105897928399983e-06, "loss": 0.0071, "step": 51730 }, { "grad_norm": 0.10143926739692688, "learning_rate": 5.0937727956036464e-06, "loss": 0.0075, "step": 51740 }, { "grad_norm": 0.11834175884723663, "learning_rate": 5.081661304133456e-06, "loss": 0.0062, "step": 51750 }, { "grad_norm": 0.1198297068476677, "learning_rate": 5.069563457668558e-06, "loss": 0.0065, "step": 51760 }, { "grad_norm": 0.09424241632223129, "learning_rate": 5.0574792598839624e-06, "loss": 0.007, "step": 51770 }, { "grad_norm": 0.09405184537172318, "learning_rate": 5.0454087144505276e-06, "loss": 0.0061, "step": 51780 }, { "grad_norm": 0.11474620550870895, "learning_rate": 5.0333518250349655e-06, "loss": 0.008, "step": 51790 }, { "grad_norm": 0.06390753388404846, "learning_rate": 5.021308595299856e-06, "loss": 0.0061, "step": 51800 }, { "grad_norm": 0.11144302040338516, "learning_rate": 5.009279028903585e-06, "loss": 0.0065, "step": 51810 }, { "grad_norm": 0.10389246046543121, "learning_rate": 4.997263129500452e-06, "loss": 0.007, "step": 51820 }, { "grad_norm": 0.09946025162935257, "learning_rate": 4.985260900740535e-06, "loss": 0.0069, "step": 51830 }, { "grad_norm": 0.09757271409034729, "learning_rate": 4.973272346269814e-06, "loss": 0.0074, "step": 51840 }, { "grad_norm": 0.07720234990119934, "learning_rate": 4.961297469730097e-06, "loss": 0.007, "step": 51850 }, { "grad_norm": 0.11286091804504395, "learning_rate": 4.949336274759031e-06, "loss": 0.0072, "step": 51860 }, { "grad_norm": 0.08271538466215134, "learning_rate": 4.9373887649901144e-06, "loss": 0.0077, "step": 51870 }, { "grad_norm": 0.08826260268688202, "learning_rate": 4.925454944052666e-06, "loss": 0.0069, "step": 51880 }, { "grad_norm": 0.07284500449895859, "learning_rate": 4.913534815571891e-06, "loss": 0.0062, "step": 51890 }, { "grad_norm": 0.07719916105270386, "learning_rate": 4.901628383168805e-06, "loss": 0.0068, "step": 51900 }, { "grad_norm": 0.09993357211351395, "learning_rate": 4.8897356504602585e-06, "loss": 0.0065, "step": 51910 }, { "grad_norm": 0.0717623233795166, "learning_rate": 4.877856621058957e-06, "loss": 0.0067, "step": 51920 }, { "grad_norm": 0.06432857364416122, "learning_rate": 4.86599129857343e-06, "loss": 0.0074, "step": 51930 }, { "grad_norm": 0.08778847008943558, "learning_rate": 4.85413968660805e-06, "loss": 0.0068, "step": 51940 }, { "grad_norm": 0.07581619173288345, "learning_rate": 4.842301788763031e-06, "loss": 0.0068, "step": 51950 }, { "grad_norm": 0.10676049441099167, "learning_rate": 4.830477608634393e-06, "loss": 0.0077, "step": 51960 }, { "grad_norm": 0.10429869592189789, "learning_rate": 4.818667149814049e-06, "loss": 0.0064, "step": 51970 }, { "grad_norm": 0.06660967320203781, "learning_rate": 4.806870415889664e-06, "loss": 0.008, "step": 51980 }, { "grad_norm": 0.08793795853853226, "learning_rate": 4.795087410444798e-06, "loss": 0.0067, "step": 51990 }, { "grad_norm": 0.10638824105262756, "learning_rate": 4.783318137058807e-06, "loss": 0.0077, "step": 52000 }, { "grad_norm": 0.2183021903038025, "learning_rate": 4.771562599306895e-06, "loss": 0.0065, "step": 52010 }, { "grad_norm": 0.11327159404754639, "learning_rate": 4.759820800760073e-06, "loss": 0.0066, "step": 52020 }, { "grad_norm": 0.09099167585372925, "learning_rate": 4.7480927449851834e-06, "loss": 0.0066, "step": 52030 }, { "grad_norm": 0.10847894102334976, "learning_rate": 4.73637843554493e-06, "loss": 0.0081, "step": 52040 }, { "grad_norm": 0.08788828551769257, "learning_rate": 4.724677875997774e-06, "loss": 0.0066, "step": 52050 }, { "grad_norm": 0.07360027730464935, "learning_rate": 4.712991069898065e-06, "loss": 0.0076, "step": 52060 }, { "grad_norm": 0.10865523666143417, "learning_rate": 4.7013180207959305e-06, "loss": 0.0072, "step": 52070 }, { "grad_norm": 0.10127915441989899, "learning_rate": 4.6896587322373395e-06, "loss": 0.0061, "step": 52080 }, { "grad_norm": 0.06303189694881439, "learning_rate": 4.678013207764081e-06, "loss": 0.0078, "step": 52090 }, { "grad_norm": 0.09413555264472961, "learning_rate": 4.666381450913748e-06, "loss": 0.0066, "step": 52100 }, { "grad_norm": 0.07439156621694565, "learning_rate": 4.654763465219752e-06, "loss": 0.0067, "step": 52110 }, { "grad_norm": 0.07192070782184601, "learning_rate": 4.643159254211371e-06, "loss": 0.0066, "step": 52120 }, { "grad_norm": 0.10017028450965881, "learning_rate": 4.631568821413606e-06, "loss": 0.0075, "step": 52130 }, { "grad_norm": 0.07355307042598724, "learning_rate": 4.619992170347359e-06, "loss": 0.0069, "step": 52140 }, { "grad_norm": 0.1136491522192955, "learning_rate": 4.608429304529305e-06, "loss": 0.0063, "step": 52150 }, { "grad_norm": 0.09735345095396042, "learning_rate": 4.596880227471928e-06, "loss": 0.0065, "step": 52160 }, { "grad_norm": 0.06935755163431168, "learning_rate": 4.585344942683539e-06, "loss": 0.0068, "step": 52170 }, { "grad_norm": 0.05606888234615326, "learning_rate": 4.573823453668241e-06, "loss": 0.0065, "step": 52180 }, { "grad_norm": 0.06838293373584747, "learning_rate": 4.562315763925995e-06, "loss": 0.0063, "step": 52190 }, { "grad_norm": 0.09546571224927902, "learning_rate": 4.5508218769524825e-06, "loss": 0.0071, "step": 52200 }, { "grad_norm": 0.10124648362398148, "learning_rate": 4.539341796239277e-06, "loss": 0.0073, "step": 52210 }, { "grad_norm": 0.0648503378033638, "learning_rate": 4.527875525273717e-06, "loss": 0.0071, "step": 52220 }, { "grad_norm": 0.10023131221532822, "learning_rate": 4.51642306753895e-06, "loss": 0.0088, "step": 52230 }, { "grad_norm": 0.10984856635332108, "learning_rate": 4.5049844265139306e-06, "loss": 0.0066, "step": 52240 }, { "grad_norm": 0.07985052466392517, "learning_rate": 4.4935596056734144e-06, "loss": 0.0069, "step": 52250 }, { "grad_norm": 0.09279972314834595, "learning_rate": 4.482148608487957e-06, "loss": 0.007, "step": 52260 }, { "grad_norm": 0.13023674488067627, "learning_rate": 4.4707514384239365e-06, "loss": 0.0068, "step": 52270 }, { "grad_norm": 0.11166000366210938, "learning_rate": 4.459368098943484e-06, "loss": 0.0069, "step": 52280 }, { "grad_norm": 0.0848127231001854, "learning_rate": 4.447998593504582e-06, "loss": 0.0072, "step": 52290 }, { "grad_norm": 0.08194334805011749, "learning_rate": 4.4366429255609744e-06, "loss": 0.0068, "step": 52300 }, { "grad_norm": 0.06641946732997894, "learning_rate": 4.425301098562212e-06, "loss": 0.0058, "step": 52310 }, { "grad_norm": 0.10294413566589355, "learning_rate": 4.413973115953651e-06, "loss": 0.0072, "step": 52320 }, { "grad_norm": 0.0698760449886322, "learning_rate": 4.402658981176416e-06, "loss": 0.0069, "step": 52330 }, { "grad_norm": 0.08914092928171158, "learning_rate": 4.391358697667475e-06, "loss": 0.0064, "step": 52340 }, { "grad_norm": 0.15783469378948212, "learning_rate": 4.3800722688595195e-06, "loss": 0.0065, "step": 52350 }, { "grad_norm": 0.13390862941741943, "learning_rate": 4.368799698181097e-06, "loss": 0.0065, "step": 52360 }, { "grad_norm": 0.1340954750776291, "learning_rate": 4.357540989056486e-06, "loss": 0.0076, "step": 52370 }, { "grad_norm": 0.06642642617225647, "learning_rate": 4.346296144905815e-06, "loss": 0.007, "step": 52380 }, { "grad_norm": 0.11989904195070267, "learning_rate": 4.335065169144958e-06, "loss": 0.007, "step": 52390 }, { "grad_norm": 0.08739382773637772, "learning_rate": 4.323848065185593e-06, "loss": 0.0072, "step": 52400 }, { "grad_norm": 0.1024213582277298, "learning_rate": 4.31264483643517e-06, "loss": 0.007, "step": 52410 }, { "grad_norm": 0.090553879737854, "learning_rate": 4.301455486296946e-06, "loss": 0.0073, "step": 52420 }, { "grad_norm": 0.09094113111495972, "learning_rate": 4.290280018169935e-06, "loss": 0.0075, "step": 52430 }, { "grad_norm": 0.06934148818254471, "learning_rate": 4.27911843544897e-06, "loss": 0.0066, "step": 52440 }, { "grad_norm": 0.09963347017765045, "learning_rate": 4.2679707415246294e-06, "loss": 0.0078, "step": 52450 }, { "grad_norm": 0.06594947725534439, "learning_rate": 4.256836939783299e-06, "loss": 0.006, "step": 52460 }, { "grad_norm": 0.08733976632356644, "learning_rate": 4.245717033607127e-06, "loss": 0.0064, "step": 52470 }, { "grad_norm": 0.06868626177310944, "learning_rate": 4.234611026374035e-06, "loss": 0.006, "step": 52480 }, { "grad_norm": 0.11910957843065262, "learning_rate": 4.2235189214577694e-06, "loss": 0.0071, "step": 52490 }, { "grad_norm": 0.10014530271291733, "learning_rate": 4.212440722227779e-06, "loss": 0.0071, "step": 52500 }, { "grad_norm": 0.0806703045964241, "learning_rate": 4.201376432049364e-06, "loss": 0.0062, "step": 52510 }, { "grad_norm": 0.16421522200107574, "learning_rate": 4.1903260542835275e-06, "loss": 0.0083, "step": 52520 }, { "grad_norm": 0.08527974784374237, "learning_rate": 4.1792895922871114e-06, "loss": 0.0068, "step": 52530 }, { "grad_norm": 0.11264849454164505, "learning_rate": 4.168267049412694e-06, "loss": 0.0068, "step": 52540 }, { "grad_norm": 0.09684063494205475, "learning_rate": 4.157258429008626e-06, "loss": 0.0062, "step": 52550 }, { "grad_norm": 0.11856500804424286, "learning_rate": 4.146263734419043e-06, "loss": 0.0071, "step": 52560 }, { "grad_norm": 0.07094676792621613, "learning_rate": 4.135282968983839e-06, "loss": 0.0073, "step": 52570 }, { "grad_norm": 0.0787983387708664, "learning_rate": 4.124316136038675e-06, "loss": 0.007, "step": 52580 }, { "grad_norm": 0.12545627355575562, "learning_rate": 4.1133632389149965e-06, "loss": 0.0072, "step": 52590 }, { "grad_norm": 0.09943102300167084, "learning_rate": 4.102424280939998e-06, "loss": 0.0074, "step": 52600 }, { "grad_norm": 0.06831976771354675, "learning_rate": 4.091499265436649e-06, "loss": 0.006, "step": 52610 }, { "grad_norm": 0.09069642424583435, "learning_rate": 4.080588195723684e-06, "loss": 0.0077, "step": 52620 }, { "grad_norm": 0.08667636662721634, "learning_rate": 4.069691075115578e-06, "loss": 0.0068, "step": 52630 }, { "grad_norm": 0.12187305092811584, "learning_rate": 4.0588079069226235e-06, "loss": 0.0076, "step": 52640 }, { "grad_norm": 0.09417682886123657, "learning_rate": 4.0479386944508034e-06, "loss": 0.0065, "step": 52650 }, { "grad_norm": 0.08806106448173523, "learning_rate": 4.037083441001932e-06, "loss": 0.0077, "step": 52660 }, { "grad_norm": 0.11876464635133743, "learning_rate": 4.026242149873516e-06, "loss": 0.008, "step": 52670 }, { "grad_norm": 0.084327831864357, "learning_rate": 4.015414824358871e-06, "loss": 0.0093, "step": 52680 }, { "grad_norm": 0.07642091810703278, "learning_rate": 4.004601467747054e-06, "loss": 0.0069, "step": 52690 }, { "grad_norm": 0.09505632519721985, "learning_rate": 3.993802083322873e-06, "loss": 0.0074, "step": 52700 }, { "grad_norm": 0.0894627645611763, "learning_rate": 3.9830166743668906e-06, "loss": 0.0068, "step": 52710 }, { "grad_norm": 0.07374744862318039, "learning_rate": 3.9722452441554425e-06, "loss": 0.0074, "step": 52720 }, { "grad_norm": 0.0746525377035141, "learning_rate": 3.961487795960584e-06, "loss": 0.0068, "step": 52730 }, { "grad_norm": 0.0825076624751091, "learning_rate": 3.95074433305016e-06, "loss": 0.0068, "step": 52740 }, { "grad_norm": 0.07359005510807037, "learning_rate": 3.940014858687752e-06, "loss": 0.0067, "step": 52750 }, { "grad_norm": 0.14504112303256989, "learning_rate": 3.929299376132689e-06, "loss": 0.0069, "step": 52760 }, { "grad_norm": 0.05617407336831093, "learning_rate": 3.918597888640047e-06, "loss": 0.0065, "step": 52770 }, { "grad_norm": 0.09034714847803116, "learning_rate": 3.907910399460657e-06, "loss": 0.0066, "step": 52780 }, { "grad_norm": 0.08156097680330276, "learning_rate": 3.8972369118410956e-06, "loss": 0.0069, "step": 52790 }, { "grad_norm": 0.08509653806686401, "learning_rate": 3.88657742902368e-06, "loss": 0.0069, "step": 52800 }, { "grad_norm": 0.08423787355422974, "learning_rate": 3.875931954246504e-06, "loss": 0.0062, "step": 52810 }, { "grad_norm": 0.08930908888578415, "learning_rate": 3.865300490743351e-06, "loss": 0.008, "step": 52820 }, { "grad_norm": 0.06919664889574051, "learning_rate": 3.854683041743806e-06, "loss": 0.0064, "step": 52830 }, { "grad_norm": 0.10685590654611588, "learning_rate": 3.844079610473139e-06, "loss": 0.0069, "step": 52840 }, { "grad_norm": 0.07869380712509155, "learning_rate": 3.833490200152423e-06, "loss": 0.0076, "step": 52850 }, { "grad_norm": 0.08474131673574448, "learning_rate": 3.822914813998424e-06, "loss": 0.0065, "step": 52860 }, { "grad_norm": 0.10241293162107468, "learning_rate": 3.812353455223666e-06, "loss": 0.0068, "step": 52870 }, { "grad_norm": 0.08490537852048874, "learning_rate": 3.8018061270364225e-06, "loss": 0.0069, "step": 52880 }, { "grad_norm": 0.0858398899435997, "learning_rate": 3.7912728326406688e-06, "loss": 0.0067, "step": 52890 }, { "grad_norm": 0.09453046321868896, "learning_rate": 3.7807535752361732e-06, "loss": 0.0074, "step": 52900 }, { "grad_norm": 0.08280699700117111, "learning_rate": 3.7702483580183855e-06, "loss": 0.0066, "step": 52910 }, { "grad_norm": 0.10532735288143158, "learning_rate": 3.759757184178525e-06, "loss": 0.006, "step": 52920 }, { "grad_norm": 0.10637693852186203, "learning_rate": 3.7492800569035312e-06, "loss": 0.0064, "step": 52930 }, { "grad_norm": 0.1183207556605339, "learning_rate": 3.7388169793760754e-06, "loss": 0.009, "step": 52940 }, { "grad_norm": 0.08161366730928421, "learning_rate": 3.728367954774553e-06, "loss": 0.0063, "step": 52950 }, { "grad_norm": 0.08451514691114426, "learning_rate": 3.7179329862731317e-06, "loss": 0.0066, "step": 52960 }, { "grad_norm": 0.09938981384038925, "learning_rate": 3.707512077041647e-06, "loss": 0.0066, "step": 52970 }, { "grad_norm": 0.07373546063899994, "learning_rate": 3.6971052302457288e-06, "loss": 0.0061, "step": 52980 }, { "grad_norm": 0.07848802208900452, "learning_rate": 3.6867124490466698e-06, "loss": 0.0063, "step": 52990 }, { "grad_norm": 0.08298575133085251, "learning_rate": 3.6763337366015393e-06, "loss": 0.0065, "step": 53000 }, { "grad_norm": 0.09981504827737808, "learning_rate": 3.665969096063121e-06, "loss": 0.0064, "step": 53010 }, { "grad_norm": 0.06918756663799286, "learning_rate": 3.6556185305799074e-06, "loss": 0.0068, "step": 53020 }, { "grad_norm": 0.06865866482257843, "learning_rate": 3.645282043296133e-06, "loss": 0.007, "step": 53030 }, { "grad_norm": 0.08207204192876816, "learning_rate": 3.6349596373517427e-06, "loss": 0.0089, "step": 53040 }, { "grad_norm": 0.0765247642993927, "learning_rate": 3.6246513158824215e-06, "loss": 0.0071, "step": 53050 }, { "grad_norm": 0.1344248503446579, "learning_rate": 3.6143570820195593e-06, "loss": 0.0066, "step": 53060 }, { "grad_norm": 0.09648381918668747, "learning_rate": 3.6040769388902773e-06, "loss": 0.0072, "step": 53070 }, { "grad_norm": 0.13116218149662018, "learning_rate": 3.593810889617405e-06, "loss": 0.007, "step": 53080 }, { "grad_norm": 0.09560489654541016, "learning_rate": 3.5835589373194978e-06, "loss": 0.0063, "step": 53090 }, { "grad_norm": 0.10450621694326401, "learning_rate": 3.5733210851108257e-06, "loss": 0.0071, "step": 53100 }, { "grad_norm": 0.14036564528942108, "learning_rate": 3.5630973361014008e-06, "loss": 0.0067, "step": 53110 }, { "grad_norm": 0.10961825400590897, "learning_rate": 3.552887693396889e-06, "loss": 0.0078, "step": 53120 }, { "grad_norm": 0.08965662866830826, "learning_rate": 3.542692160098754e-06, "loss": 0.0077, "step": 53130 }, { "grad_norm": 0.06275569647550583, "learning_rate": 3.5325107393040846e-06, "loss": 0.0068, "step": 53140 }, { "grad_norm": 0.06292834877967834, "learning_rate": 3.522343434105757e-06, "loss": 0.0063, "step": 53150 }, { "grad_norm": 0.07003703713417053, "learning_rate": 3.512190247592323e-06, "loss": 0.0064, "step": 53160 }, { "grad_norm": 0.0868585854768753, "learning_rate": 3.502051182848054e-06, "loss": 0.0066, "step": 53170 }, { "grad_norm": 0.07445335388183594, "learning_rate": 3.4919262429529308e-06, "loss": 0.0073, "step": 53180 }, { "grad_norm": 0.1222672238945961, "learning_rate": 3.4818154309826325e-06, "loss": 0.0067, "step": 53190 }, { "grad_norm": 0.08905103802680969, "learning_rate": 3.4717187500085734e-06, "loss": 0.008, "step": 53200 }, { "grad_norm": 0.09154606610536575, "learning_rate": 3.46163620309784e-06, "loss": 0.0066, "step": 53210 }, { "grad_norm": 0.08282520622015, "learning_rate": 3.4515677933132595e-06, "loss": 0.0059, "step": 53220 }, { "grad_norm": 0.0770614817738533, "learning_rate": 3.4415135237133466e-06, "loss": 0.0061, "step": 53230 }, { "grad_norm": 0.10050298273563385, "learning_rate": 3.4314733973523196e-06, "loss": 0.0078, "step": 53240 }, { "grad_norm": 0.07717259228229523, "learning_rate": 3.4214474172800993e-06, "loss": 0.0069, "step": 53250 }, { "grad_norm": 0.06855445355176926, "learning_rate": 3.411435586542322e-06, "loss": 0.0068, "step": 53260 }, { "grad_norm": 0.09865173697471619, "learning_rate": 3.4014379081802995e-06, "loss": 0.0077, "step": 53270 }, { "grad_norm": 0.07638122141361237, "learning_rate": 3.391454385231102e-06, "loss": 0.0059, "step": 53280 }, { "grad_norm": 0.080743208527565, "learning_rate": 3.3814850207274095e-06, "loss": 0.0073, "step": 53290 }, { "grad_norm": 0.09275735169649124, "learning_rate": 3.371529817697694e-06, "loss": 0.0064, "step": 53300 }, { "grad_norm": 0.09678293019533157, "learning_rate": 3.3615887791660585e-06, "loss": 0.0075, "step": 53310 }, { "grad_norm": 0.08662738651037216, "learning_rate": 3.3516619081523426e-06, "loss": 0.0102, "step": 53320 }, { "grad_norm": 0.07440298795700073, "learning_rate": 3.3417492076720567e-06, "loss": 0.0067, "step": 53330 }, { "grad_norm": 0.07485619187355042, "learning_rate": 3.3318506807364147e-06, "loss": 0.0071, "step": 53340 }, { "grad_norm": 0.1955333948135376, "learning_rate": 3.3219663303523553e-06, "loss": 0.0067, "step": 53350 }, { "grad_norm": 0.11159359663724899, "learning_rate": 3.3120961595224374e-06, "loss": 0.0074, "step": 53360 }, { "grad_norm": 0.08612841367721558, "learning_rate": 3.302240171245002e-06, "loss": 0.0069, "step": 53370 }, { "grad_norm": 0.06404294818639755, "learning_rate": 3.29239836851401e-06, "loss": 0.006, "step": 53380 }, { "grad_norm": 0.08215433359146118, "learning_rate": 3.2825707543191588e-06, "loss": 0.0064, "step": 53390 }, { "grad_norm": 0.08283287286758423, "learning_rate": 3.272757331645804e-06, "loss": 0.0077, "step": 53400 }, { "grad_norm": 0.07642621546983719, "learning_rate": 3.2629581034750166e-06, "loss": 0.0061, "step": 53410 }, { "grad_norm": 0.09055845439434052, "learning_rate": 3.2531730727835218e-06, "loss": 0.007, "step": 53420 }, { "grad_norm": 0.08369120210409164, "learning_rate": 3.2434022425437914e-06, "loss": 0.0064, "step": 53430 }, { "grad_norm": 0.08768624067306519, "learning_rate": 3.233645615723907e-06, "loss": 0.0066, "step": 53440 }, { "grad_norm": 0.10973634570837021, "learning_rate": 3.2239031952876918e-06, "loss": 0.0068, "step": 53450 }, { "grad_norm": 0.07467276602983475, "learning_rate": 3.21417498419464e-06, "loss": 0.0059, "step": 53460 }, { "grad_norm": 0.09549110382795334, "learning_rate": 3.204460985399921e-06, "loss": 0.0069, "step": 53470 }, { "grad_norm": 0.07391805201768875, "learning_rate": 3.1947612018543903e-06, "loss": 0.0068, "step": 53480 }, { "grad_norm": 0.05780543386936188, "learning_rate": 3.1850756365045753e-06, "loss": 0.0064, "step": 53490 }, { "grad_norm": 0.058412134647369385, "learning_rate": 3.175404292292722e-06, "loss": 0.0064, "step": 53500 }, { "grad_norm": 0.10332342237234116, "learning_rate": 3.1657471721566965e-06, "loss": 0.0066, "step": 53510 }, { "grad_norm": 0.09765785932540894, "learning_rate": 3.1561042790300977e-06, "loss": 0.0074, "step": 53520 }, { "grad_norm": 0.08818579465150833, "learning_rate": 3.1464756158421816e-06, "loss": 0.0062, "step": 53530 }, { "grad_norm": 0.099019855260849, "learning_rate": 3.136861185517875e-06, "loss": 0.0072, "step": 53540 }, { "grad_norm": 0.08589791506528854, "learning_rate": 3.127260990977798e-06, "loss": 0.0063, "step": 53550 }, { "grad_norm": 0.09240444749593735, "learning_rate": 3.1176750351382235e-06, "loss": 0.0069, "step": 53560 }, { "grad_norm": 0.059128742665052414, "learning_rate": 3.1081033209111153e-06, "loss": 0.0067, "step": 53570 }, { "grad_norm": 0.08843738585710526, "learning_rate": 3.0985458512041155e-06, "loss": 0.0071, "step": 53580 }, { "grad_norm": 0.08211127668619156, "learning_rate": 3.089002628920512e-06, "loss": 0.0062, "step": 53590 }, { "grad_norm": 0.08774330466985703, "learning_rate": 3.079473656959303e-06, "loss": 0.0067, "step": 53600 }, { "grad_norm": 0.08242606371641159, "learning_rate": 3.0699589382151393e-06, "loss": 0.0076, "step": 53610 }, { "grad_norm": 0.10238399356603622, "learning_rate": 3.060458475578326e-06, "loss": 0.0066, "step": 53620 }, { "grad_norm": 0.10800259560346603, "learning_rate": 3.0509722719348656e-06, "loss": 0.0074, "step": 53630 }, { "grad_norm": 0.10483478009700775, "learning_rate": 3.041500330166408e-06, "loss": 0.0059, "step": 53640 }, { "grad_norm": 0.08152922987937927, "learning_rate": 3.032042653150291e-06, "loss": 0.006, "step": 53650 }, { "grad_norm": 0.07475046813488007, "learning_rate": 3.0225992437594887e-06, "loss": 0.0077, "step": 53660 }, { "grad_norm": 0.08376822620630264, "learning_rate": 3.0131701048626783e-06, "loss": 0.0064, "step": 53670 }, { "grad_norm": 0.07817422598600388, "learning_rate": 3.003755239324163e-06, "loss": 0.0097, "step": 53680 }, { "grad_norm": 0.08522661030292511, "learning_rate": 2.9943546500039553e-06, "loss": 0.0066, "step": 53690 }, { "grad_norm": 0.07556244730949402, "learning_rate": 2.9849683397576877e-06, "loss": 0.0067, "step": 53700 }, { "grad_norm": 0.11324623972177505, "learning_rate": 2.975596311436679e-06, "loss": 0.0076, "step": 53710 }, { "grad_norm": 0.10593577474355698, "learning_rate": 2.966238567887902e-06, "loss": 0.0065, "step": 53720 }, { "grad_norm": 0.09447363018989563, "learning_rate": 2.9568951119539943e-06, "loss": 0.0067, "step": 53730 }, { "grad_norm": 0.10171335190534592, "learning_rate": 2.947565946473241e-06, "loss": 0.0065, "step": 53740 }, { "grad_norm": 0.06687603890895844, "learning_rate": 2.9382510742796188e-06, "loss": 0.0066, "step": 53750 }, { "grad_norm": 0.08219102025032043, "learning_rate": 2.9289504982027204e-06, "loss": 0.0073, "step": 53760 }, { "grad_norm": 0.09443873912096024, "learning_rate": 2.9196642210678248e-06, "loss": 0.0075, "step": 53770 }, { "grad_norm": 0.06981617212295532, "learning_rate": 2.910392245695853e-06, "loss": 0.0064, "step": 53780 }, { "grad_norm": 0.08990593254566193, "learning_rate": 2.9011345749033804e-06, "loss": 0.0066, "step": 53790 }, { "grad_norm": 0.09589669853448868, "learning_rate": 2.8918912115026677e-06, "loss": 0.0066, "step": 53800 }, { "grad_norm": 0.09814558178186417, "learning_rate": 2.8826621583015743e-06, "loss": 0.0071, "step": 53810 }, { "grad_norm": 0.06456232070922852, "learning_rate": 2.8734474181036643e-06, "loss": 0.0067, "step": 53820 }, { "grad_norm": 0.07859565317630768, "learning_rate": 2.864246993708114e-06, "loss": 0.0066, "step": 53830 }, { "grad_norm": 0.0739869475364685, "learning_rate": 2.8550608879097884e-06, "loss": 0.0063, "step": 53840 }, { "grad_norm": 0.05580425634980202, "learning_rate": 2.845889103499172e-06, "loss": 0.0074, "step": 53850 }, { "grad_norm": 0.07240869104862213, "learning_rate": 2.8367316432624138e-06, "loss": 0.006, "step": 53860 }, { "grad_norm": 0.11095333844423294, "learning_rate": 2.8275885099813105e-06, "loss": 0.0073, "step": 53870 }, { "grad_norm": 0.08988582342863083, "learning_rate": 2.8184597064332963e-06, "loss": 0.0065, "step": 53880 }, { "grad_norm": 0.10304915904998779, "learning_rate": 2.809345235391464e-06, "loss": 0.0062, "step": 53890 }, { "grad_norm": 0.08842562139034271, "learning_rate": 2.80024509962456e-06, "loss": 0.0073, "step": 53900 }, { "grad_norm": 0.08592023700475693, "learning_rate": 2.7911593018969563e-06, "loss": 0.0071, "step": 53910 }, { "grad_norm": 0.07104314863681793, "learning_rate": 2.7820878449686838e-06, "loss": 0.0073, "step": 53920 }, { "grad_norm": 0.11302211135625839, "learning_rate": 2.7730307315953995e-06, "loss": 0.0074, "step": 53930 }, { "grad_norm": 0.12291014194488525, "learning_rate": 2.763987964528425e-06, "loss": 0.0068, "step": 53940 }, { "grad_norm": 0.06602567434310913, "learning_rate": 2.754959546514718e-06, "loss": 0.0069, "step": 53950 }, { "grad_norm": 0.07411353290081024, "learning_rate": 2.7459454802968576e-06, "loss": 0.0063, "step": 53960 }, { "grad_norm": 0.10386361926794052, "learning_rate": 2.7369457686131028e-06, "loss": 0.0084, "step": 53970 }, { "grad_norm": 0.07687769830226898, "learning_rate": 2.7279604141973004e-06, "loss": 0.0067, "step": 53980 }, { "grad_norm": 0.07706665992736816, "learning_rate": 2.7189894197789946e-06, "loss": 0.0068, "step": 53990 }, { "grad_norm": 0.07307067513465881, "learning_rate": 2.7100327880833055e-06, "loss": 0.0067, "step": 54000 }, { "grad_norm": 0.09946658462285995, "learning_rate": 2.70109052183104e-06, "loss": 0.0073, "step": 54010 }, { "grad_norm": 0.07827576249837875, "learning_rate": 2.69216262373862e-06, "loss": 0.0067, "step": 54020 }, { "grad_norm": 0.08696825057268143, "learning_rate": 2.6832490965181036e-06, "loss": 0.0076, "step": 54030 }, { "grad_norm": 0.11613178998231888, "learning_rate": 2.6743499428771857e-06, "loss": 0.0068, "step": 54040 }, { "grad_norm": 0.14550314843654633, "learning_rate": 2.6654651655191875e-06, "loss": 0.0063, "step": 54050 }, { "grad_norm": 0.08475998044013977, "learning_rate": 2.6565947671430836e-06, "loss": 0.0069, "step": 54060 }, { "grad_norm": 0.0734202191233635, "learning_rate": 2.647738750443457e-06, "loss": 0.0067, "step": 54070 }, { "grad_norm": 0.11473103612661362, "learning_rate": 2.6388971181105393e-06, "loss": 0.0065, "step": 54080 }, { "grad_norm": 0.15310992300510406, "learning_rate": 2.630069872830171e-06, "loss": 0.0084, "step": 54090 }, { "grad_norm": 0.07955727726221085, "learning_rate": 2.6212570172838514e-06, "loss": 0.0072, "step": 54100 }, { "grad_norm": 0.08350920677185059, "learning_rate": 2.6124585541486778e-06, "loss": 0.0069, "step": 54110 }, { "grad_norm": 0.09024476259946823, "learning_rate": 2.6036744860974127e-06, "loss": 0.0074, "step": 54120 }, { "grad_norm": 0.09972351789474487, "learning_rate": 2.594904815798399e-06, "loss": 0.0077, "step": 54130 }, { "grad_norm": 0.07630178332328796, "learning_rate": 2.586149545915656e-06, "loss": 0.007, "step": 54140 }, { "grad_norm": 0.06989982724189758, "learning_rate": 2.577408679108778e-06, "loss": 0.0067, "step": 54150 }, { "grad_norm": 0.14371342957019806, "learning_rate": 2.5686822180330306e-06, "loss": 0.0065, "step": 54160 }, { "grad_norm": 0.07450506836175919, "learning_rate": 2.5599701653392703e-06, "loss": 0.007, "step": 54170 }, { "grad_norm": 0.08717894554138184, "learning_rate": 2.551272523673992e-06, "loss": 0.0062, "step": 54180 }, { "grad_norm": 0.11756730079650879, "learning_rate": 2.542589295679315e-06, "loss": 0.006, "step": 54190 }, { "grad_norm": 0.10050801932811737, "learning_rate": 2.5339204839929575e-06, "loss": 0.0079, "step": 54200 }, { "grad_norm": 0.07355595380067825, "learning_rate": 2.525266091248296e-06, "loss": 0.0065, "step": 54210 }, { "grad_norm": 0.09575111418962479, "learning_rate": 2.5166261200743e-06, "loss": 0.0071, "step": 54220 }, { "grad_norm": 0.08814669400453568, "learning_rate": 2.5080005730955646e-06, "loss": 0.0064, "step": 54230 }, { "grad_norm": 0.09133997559547424, "learning_rate": 2.499389452932299e-06, "loss": 0.0075, "step": 54240 }, { "grad_norm": 0.07209242880344391, "learning_rate": 2.4907927622003336e-06, "loss": 0.0077, "step": 54250 }, { "grad_norm": 0.10588867217302322, "learning_rate": 2.4822105035111177e-06, "loss": 0.0069, "step": 54260 }, { "grad_norm": 0.1602409929037094, "learning_rate": 2.4736426794717273e-06, "loss": 0.0063, "step": 54270 }, { "grad_norm": 0.09679620712995529, "learning_rate": 2.4650892926848135e-06, "loss": 0.0065, "step": 54280 }, { "grad_norm": 0.08800387382507324, "learning_rate": 2.456550345748704e-06, "loss": 0.0064, "step": 54290 }, { "grad_norm": 0.07781227678060532, "learning_rate": 2.4480258412572733e-06, "loss": 0.0073, "step": 54300 }, { "grad_norm": 0.09143031388521194, "learning_rate": 2.4395157818000612e-06, "loss": 0.0063, "step": 54310 }, { "grad_norm": 0.09497950226068497, "learning_rate": 2.431020169962189e-06, "loss": 0.0079, "step": 54320 }, { "grad_norm": 0.09557909518480301, "learning_rate": 2.422539008324409e-06, "loss": 0.0076, "step": 54330 }, { "grad_norm": 0.08605042845010757, "learning_rate": 2.414072299463066e-06, "loss": 0.0063, "step": 54340 }, { "grad_norm": 0.16522593796253204, "learning_rate": 2.4056200459501186e-06, "loss": 0.0076, "step": 54350 }, { "grad_norm": 0.08577273041009903, "learning_rate": 2.397182250353147e-06, "loss": 0.0063, "step": 54360 }, { "grad_norm": 0.09778233617544174, "learning_rate": 2.388758915235334e-06, "loss": 0.0065, "step": 54370 }, { "grad_norm": 0.0649503543972969, "learning_rate": 2.380350043155455e-06, "loss": 0.0059, "step": 54380 }, { "grad_norm": 0.08384349942207336, "learning_rate": 2.371955636667911e-06, "loss": 0.0065, "step": 54390 }, { "grad_norm": 0.08687585592269897, "learning_rate": 2.3635756983227008e-06, "loss": 0.0068, "step": 54400 }, { "grad_norm": 0.0950048640370369, "learning_rate": 2.3552102306654278e-06, "loss": 0.0064, "step": 54410 }, { "grad_norm": 0.09557800740003586, "learning_rate": 2.346859236237292e-06, "loss": 0.0075, "step": 54420 }, { "grad_norm": 0.0816459059715271, "learning_rate": 2.3385227175751145e-06, "loss": 0.0064, "step": 54430 }, { "grad_norm": 0.0936082974076271, "learning_rate": 2.330200677211314e-06, "loss": 0.0071, "step": 54440 }, { "grad_norm": 0.08134140074253082, "learning_rate": 2.3218931176738847e-06, "loss": 0.0069, "step": 54450 }, { "grad_norm": 0.0922972559928894, "learning_rate": 2.313600041486469e-06, "loss": 0.0068, "step": 54460 }, { "grad_norm": 0.10928235948085785, "learning_rate": 2.3053214511682743e-06, "loss": 0.0066, "step": 54470 }, { "grad_norm": 0.15097621083259583, "learning_rate": 2.2970573492341163e-06, "loss": 0.007, "step": 54480 }, { "grad_norm": 0.07472358644008636, "learning_rate": 2.288807738194415e-06, "loss": 0.0065, "step": 54490 }, { "grad_norm": 0.07589073479175568, "learning_rate": 2.2805726205551768e-06, "loss": 0.0065, "step": 54500 }, { "grad_norm": 0.05966024100780487, "learning_rate": 2.272351998818023e-06, "loss": 0.0075, "step": 54510 }, { "grad_norm": 0.11968756467103958, "learning_rate": 2.2641458754801505e-06, "loss": 0.0057, "step": 54520 }, { "grad_norm": 0.08263376355171204, "learning_rate": 2.2559542530343756e-06, "loss": 0.0066, "step": 54530 }, { "grad_norm": 0.10027847439050674, "learning_rate": 2.247777133969087e-06, "loss": 0.0063, "step": 54540 }, { "grad_norm": 0.08452827483415604, "learning_rate": 2.2396145207682795e-06, "loss": 0.007, "step": 54550 }, { "grad_norm": 0.0882616862654686, "learning_rate": 2.231466415911543e-06, "loss": 0.0069, "step": 54560 }, { "grad_norm": 0.07916360348463058, "learning_rate": 2.2233328218740524e-06, "loss": 0.0065, "step": 54570 }, { "grad_norm": 0.19461828470230103, "learning_rate": 2.215213741126576e-06, "loss": 0.006, "step": 54580 }, { "grad_norm": 0.06803559511899948, "learning_rate": 2.2071091761354912e-06, "loss": 0.0069, "step": 54590 }, { "grad_norm": 0.06701695919036865, "learning_rate": 2.1990191293627337e-06, "loss": 0.0068, "step": 54600 }, { "grad_norm": 0.069117471575737, "learning_rate": 2.1909436032658548e-06, "loss": 0.0062, "step": 54610 }, { "grad_norm": 0.06607058644294739, "learning_rate": 2.1828826002979806e-06, "loss": 0.0086, "step": 54620 }, { "grad_norm": 0.07815950363874435, "learning_rate": 2.1748361229078362e-06, "loss": 0.0061, "step": 54630 }, { "grad_norm": 0.12356841564178467, "learning_rate": 2.1668041735397327e-06, "loss": 0.0072, "step": 54640 }, { "grad_norm": 0.09498274326324463, "learning_rate": 2.1587867546335514e-06, "loss": 0.0069, "step": 54650 }, { "grad_norm": 0.11743852496147156, "learning_rate": 2.1507838686247894e-06, "loss": 0.0063, "step": 54660 }, { "grad_norm": 0.07577979564666748, "learning_rate": 2.1427955179444957e-06, "loss": 0.0065, "step": 54670 }, { "grad_norm": 0.08013017475605011, "learning_rate": 2.13482170501933e-06, "loss": 0.0072, "step": 54680 }, { "grad_norm": 0.07705949246883392, "learning_rate": 2.1268624322715202e-06, "loss": 0.0061, "step": 54690 }, { "grad_norm": 0.0784299373626709, "learning_rate": 2.118917702118889e-06, "loss": 0.0063, "step": 54700 }, { "grad_norm": 0.1559659093618393, "learning_rate": 2.1109875169748327e-06, "loss": 0.0071, "step": 54710 }, { "grad_norm": 0.0972360149025917, "learning_rate": 2.103071879248336e-06, "loss": 0.0062, "step": 54720 }, { "grad_norm": 0.0962008461356163, "learning_rate": 2.0951707913439478e-06, "loss": 0.0072, "step": 54730 }, { "grad_norm": 0.06742817163467407, "learning_rate": 2.0872842556618255e-06, "loss": 0.0061, "step": 54740 }, { "grad_norm": 0.07192512601613998, "learning_rate": 2.079412274597686e-06, "loss": 0.0081, "step": 54750 }, { "grad_norm": 0.0718342587351799, "learning_rate": 2.0715548505428284e-06, "loss": 0.0067, "step": 54760 }, { "grad_norm": 0.06932628899812698, "learning_rate": 2.0637119858841258e-06, "loss": 0.007, "step": 54770 }, { "grad_norm": 0.07715391367673874, "learning_rate": 2.055883683004034e-06, "loss": 0.0065, "step": 54780 }, { "grad_norm": 0.06251344084739685, "learning_rate": 2.0480699442806006e-06, "loss": 0.0063, "step": 54790 }, { "grad_norm": 0.08791982382535934, "learning_rate": 2.0402707720874105e-06, "loss": 0.0068, "step": 54800 }, { "grad_norm": 0.14207379519939423, "learning_rate": 2.032486168793668e-06, "loss": 0.0073, "step": 54810 }, { "grad_norm": 0.06524446606636047, "learning_rate": 2.024716136764104e-06, "loss": 0.0063, "step": 54820 }, { "grad_norm": 0.08367756009101868, "learning_rate": 2.0169606783590757e-06, "loss": 0.0068, "step": 54830 }, { "grad_norm": 0.08890235424041748, "learning_rate": 2.0092197959344638e-06, "loss": 0.0068, "step": 54840 }, { "grad_norm": 0.10040078312158585, "learning_rate": 2.0014934918417606e-06, "loss": 0.0063, "step": 54850 }, { "grad_norm": 0.09832065552473068, "learning_rate": 1.99378176842801e-06, "loss": 0.0066, "step": 54860 }, { "grad_norm": 0.07954158633947372, "learning_rate": 1.9860846280358224e-06, "loss": 0.0069, "step": 54870 }, { "grad_norm": 0.06939823180437088, "learning_rate": 1.978402073003394e-06, "loss": 0.0065, "step": 54880 }, { "grad_norm": 0.12632448971271515, "learning_rate": 1.9707341056644736e-06, "loss": 0.0071, "step": 54890 }, { "grad_norm": 0.10118179023265839, "learning_rate": 1.963080728348399e-06, "loss": 0.008, "step": 54900 }, { "grad_norm": 0.08620642870664597, "learning_rate": 1.955441943380054e-06, "loss": 0.0061, "step": 54910 }, { "grad_norm": 0.09989457577466965, "learning_rate": 1.947817753079906e-06, "loss": 0.0083, "step": 54920 }, { "grad_norm": 0.09953618794679642, "learning_rate": 1.9402081597639785e-06, "loss": 0.0065, "step": 54930 }, { "grad_norm": 0.07769595086574554, "learning_rate": 1.9326131657438683e-06, "loss": 0.0065, "step": 54940 }, { "grad_norm": 0.11342782527208328, "learning_rate": 1.925032773326724e-06, "loss": 0.0082, "step": 54950 }, { "grad_norm": 0.07340776920318604, "learning_rate": 1.9174669848152916e-06, "loss": 0.0067, "step": 54960 }, { "grad_norm": 0.07751715928316116, "learning_rate": 1.9099158025078336e-06, "loss": 0.0056, "step": 54970 }, { "grad_norm": 0.09477534145116806, "learning_rate": 1.90237922869822e-06, "loss": 0.0067, "step": 54980 }, { "grad_norm": 0.08267254382371902, "learning_rate": 1.8948572656758367e-06, "loss": 0.0062, "step": 54990 }, { "grad_norm": 0.08143631368875504, "learning_rate": 1.8873499157256834e-06, "loss": 0.0071, "step": 55000 }, { "grad_norm": 0.07553713023662567, "learning_rate": 1.879857181128286e-06, "loss": 0.0069, "step": 55010 }, { "grad_norm": 0.0714937150478363, "learning_rate": 1.8723790641597349e-06, "loss": 0.0071, "step": 55020 }, { "grad_norm": 0.08215180039405823, "learning_rate": 1.8649155670916906e-06, "loss": 0.0066, "step": 55030 }, { "grad_norm": 0.10332271456718445, "learning_rate": 1.8574666921913565e-06, "loss": 0.007, "step": 55040 }, { "grad_norm": 0.07196410745382309, "learning_rate": 1.8500324417215166e-06, "loss": 0.0066, "step": 55050 }, { "grad_norm": 0.0593918077647686, "learning_rate": 1.8426128179404977e-06, "loss": 0.006, "step": 55060 }, { "grad_norm": 0.11955700814723969, "learning_rate": 1.8352078231021807e-06, "loss": 0.0066, "step": 55070 }, { "grad_norm": 0.06415495276451111, "learning_rate": 1.8278174594560049e-06, "loss": 0.0068, "step": 55080 }, { "grad_norm": 0.08330652862787247, "learning_rate": 1.82044172924698e-06, "loss": 0.0077, "step": 55090 }, { "grad_norm": 0.05689512565732002, "learning_rate": 1.813080634715636e-06, "loss": 0.0062, "step": 55100 }, { "grad_norm": 0.06648816913366318, "learning_rate": 1.8057341780981119e-06, "loss": 0.0062, "step": 55110 }, { "grad_norm": 0.06892815977334976, "learning_rate": 1.7984023616260338e-06, "loss": 0.0072, "step": 55120 }, { "grad_norm": 0.13851222395896912, "learning_rate": 1.7910851875266421e-06, "loss": 0.0063, "step": 55130 }, { "grad_norm": 0.08204196393489838, "learning_rate": 1.7837826580226757e-06, "loss": 0.0068, "step": 55140 }, { "grad_norm": 0.07271276414394379, "learning_rate": 1.7764947753324656e-06, "loss": 0.007, "step": 55150 }, { "grad_norm": 0.09011995792388916, "learning_rate": 1.7692215416698799e-06, "loss": 0.0064, "step": 55160 }, { "grad_norm": 0.09278269112110138, "learning_rate": 1.7619629592443233e-06, "loss": 0.0067, "step": 55170 }, { "grad_norm": 0.10280214250087738, "learning_rate": 1.7547190302607709e-06, "loss": 0.007, "step": 55180 }, { "grad_norm": 0.06069023162126541, "learning_rate": 1.747489756919729e-06, "loss": 0.0077, "step": 55190 }, { "grad_norm": 0.08900225162506104, "learning_rate": 1.7402751414172802e-06, "loss": 0.0068, "step": 55200 }, { "grad_norm": 0.1180942952632904, "learning_rate": 1.7330751859450044e-06, "loss": 0.0065, "step": 55210 }, { "grad_norm": 0.07986314594745636, "learning_rate": 1.7258898926900801e-06, "loss": 0.0068, "step": 55220 }, { "grad_norm": 0.09759928286075592, "learning_rate": 1.7187192638352002e-06, "loss": 0.0062, "step": 55230 }, { "grad_norm": 0.08612411469221115, "learning_rate": 1.7115633015586163e-06, "loss": 0.0065, "step": 55240 }, { "grad_norm": 0.056574348360300064, "learning_rate": 1.7044220080341178e-06, "loss": 0.0075, "step": 55250 }, { "grad_norm": 0.23324468731880188, "learning_rate": 1.6972953854310414e-06, "loss": 0.0065, "step": 55260 }, { "grad_norm": 0.06701164692640305, "learning_rate": 1.690183435914261e-06, "loss": 0.0062, "step": 55270 }, { "grad_norm": 0.07171176373958588, "learning_rate": 1.6830861616442206e-06, "loss": 0.0075, "step": 55280 }, { "grad_norm": 0.07370687276124954, "learning_rate": 1.6760035647768568e-06, "loss": 0.007, "step": 55290 }, { "grad_norm": 0.06475590169429779, "learning_rate": 1.6689356474636985e-06, "loss": 0.007, "step": 55300 }, { "grad_norm": 0.12068800628185272, "learning_rate": 1.6618824118517784e-06, "loss": 0.0078, "step": 55310 }, { "grad_norm": 0.07037116587162018, "learning_rate": 1.654843860083688e-06, "loss": 0.0066, "step": 55320 }, { "grad_norm": 0.09563259780406952, "learning_rate": 1.647819994297556e-06, "loss": 0.0071, "step": 55330 }, { "grad_norm": 0.0652458667755127, "learning_rate": 1.6408108166270363e-06, "loss": 0.0063, "step": 55340 }, { "grad_norm": 0.08314502239227295, "learning_rate": 1.6338163292013486e-06, "loss": 0.0078, "step": 55350 }, { "grad_norm": 0.09694121778011322, "learning_rate": 1.6268365341452208e-06, "loss": 0.0062, "step": 55360 }, { "grad_norm": 0.07696012407541275, "learning_rate": 1.6198714335789345e-06, "loss": 0.007, "step": 55370 }, { "grad_norm": 0.09043518453836441, "learning_rate": 1.612921029618303e-06, "loss": 0.0074, "step": 55380 }, { "grad_norm": 0.09846509993076324, "learning_rate": 1.6059853243746815e-06, "loss": 0.0074, "step": 55390 }, { "grad_norm": 0.06378920376300812, "learning_rate": 1.5990643199549404e-06, "loss": 0.0077, "step": 55400 }, { "grad_norm": 0.0975816547870636, "learning_rate": 1.5921580184615147e-06, "loss": 0.0073, "step": 55410 }, { "grad_norm": 0.09774018079042435, "learning_rate": 1.5852664219923374e-06, "loss": 0.0071, "step": 55420 }, { "grad_norm": 0.11615429073572159, "learning_rate": 1.5783895326409172e-06, "loss": 0.0066, "step": 55430 }, { "grad_norm": 0.0817328691482544, "learning_rate": 1.5715273524962438e-06, "loss": 0.0055, "step": 55440 }, { "grad_norm": 0.10505195707082748, "learning_rate": 1.564679883642889e-06, "loss": 0.006, "step": 55450 }, { "grad_norm": 0.08232593536376953, "learning_rate": 1.5578471281609276e-06, "loss": 0.0071, "step": 55460 }, { "grad_norm": 0.11371239274740219, "learning_rate": 1.551029088125966e-06, "loss": 0.0065, "step": 55470 }, { "grad_norm": 0.05959309637546539, "learning_rate": 1.544225765609142e-06, "loss": 0.0067, "step": 55480 }, { "grad_norm": 0.07949582487344742, "learning_rate": 1.53743716267713e-06, "loss": 0.0068, "step": 55490 }, { "grad_norm": 0.07580636441707611, "learning_rate": 1.5306632813921361e-06, "loss": 0.0084, "step": 55500 }, { "grad_norm": 0.09714339673519135, "learning_rate": 1.52390412381187e-06, "loss": 0.0065, "step": 55510 }, { "grad_norm": 0.08901567757129669, "learning_rate": 1.5171596919895948e-06, "loss": 0.0065, "step": 55520 }, { "grad_norm": 0.06963931024074554, "learning_rate": 1.5104299879740936e-06, "loss": 0.0065, "step": 55530 }, { "grad_norm": 0.10649492591619492, "learning_rate": 1.5037150138096701e-06, "loss": 0.0063, "step": 55540 }, { "grad_norm": 0.10018295794725418, "learning_rate": 1.4970147715361538e-06, "loss": 0.0062, "step": 55550 }, { "grad_norm": 0.0969586968421936, "learning_rate": 1.4903292631889054e-06, "loss": 0.0065, "step": 55560 }, { "grad_norm": 0.10291577130556107, "learning_rate": 1.483658490798795e-06, "loss": 0.0073, "step": 55570 }, { "grad_norm": 0.08762995898723602, "learning_rate": 1.4770024563922457e-06, "loss": 0.0069, "step": 55580 }, { "grad_norm": 0.08211983740329742, "learning_rate": 1.4703611619911628e-06, "loss": 0.0068, "step": 55590 }, { "grad_norm": 0.08384262025356293, "learning_rate": 1.4637346096130155e-06, "loss": 0.0067, "step": 55600 }, { "grad_norm": 0.06532596051692963, "learning_rate": 1.4571228012707662e-06, "loss": 0.0079, "step": 55610 }, { "grad_norm": 0.1394178420305252, "learning_rate": 1.4505257389729132e-06, "loss": 0.0071, "step": 55620 }, { "grad_norm": 0.07461176812648773, "learning_rate": 1.4439434247234596e-06, "loss": 0.0072, "step": 55630 }, { "grad_norm": 0.07999102026224136, "learning_rate": 1.4373758605219445e-06, "loss": 0.007, "step": 55640 }, { "grad_norm": 0.07125391066074371, "learning_rate": 1.4308230483634333e-06, "loss": 0.0067, "step": 55650 }, { "grad_norm": 0.0699881985783577, "learning_rate": 1.4242849902384724e-06, "loss": 0.0066, "step": 55660 }, { "grad_norm": 0.08980733901262283, "learning_rate": 1.4177616881331734e-06, "loss": 0.0084, "step": 55670 }, { "grad_norm": 0.07207569479942322, "learning_rate": 1.4112531440291233e-06, "loss": 0.0073, "step": 55680 }, { "grad_norm": 0.07893015444278717, "learning_rate": 1.4047593599034624e-06, "loss": 0.007, "step": 55690 }, { "grad_norm": 0.0789785385131836, "learning_rate": 1.3982803377288246e-06, "loss": 0.0077, "step": 55700 }, { "grad_norm": 0.07448220252990723, "learning_rate": 1.3918160794733681e-06, "loss": 0.0067, "step": 55710 }, { "grad_norm": 0.08137443661689758, "learning_rate": 1.3853665871007615e-06, "loss": 0.0064, "step": 55720 }, { "grad_norm": 0.08556751161813736, "learning_rate": 1.378931862570193e-06, "loss": 0.0076, "step": 55730 }, { "grad_norm": 0.08780090510845184, "learning_rate": 1.372511907836349e-06, "loss": 0.0075, "step": 55740 }, { "grad_norm": 0.07185449451208115, "learning_rate": 1.3661067248494586e-06, "loss": 0.0062, "step": 55750 }, { "grad_norm": 0.08049163222312927, "learning_rate": 1.3597163155552429e-06, "loss": 0.0066, "step": 55760 }, { "grad_norm": 0.07339568436145782, "learning_rate": 1.3533406818949434e-06, "loss": 0.0069, "step": 55770 }, { "grad_norm": 0.08724047243595123, "learning_rate": 1.3469798258053002e-06, "loss": 0.0062, "step": 55780 }, { "grad_norm": 0.10605154186487198, "learning_rate": 1.3406337492185672e-06, "loss": 0.0063, "step": 55790 }, { "grad_norm": 0.0715060755610466, "learning_rate": 1.3343024540625414e-06, "loss": 0.0067, "step": 55800 }, { "grad_norm": 0.06358008831739426, "learning_rate": 1.3279859422604735e-06, "loss": 0.0066, "step": 55810 }, { "grad_norm": 0.14788328111171722, "learning_rate": 1.3216842157311781e-06, "loss": 0.007, "step": 55820 }, { "grad_norm": 0.06562565267086029, "learning_rate": 1.3153972763889355e-06, "loss": 0.0068, "step": 55830 }, { "grad_norm": 0.09801765531301498, "learning_rate": 1.3091251261435566e-06, "loss": 0.0065, "step": 55840 }, { "grad_norm": 0.058186501264572144, "learning_rate": 1.3028677669003564e-06, "loss": 0.0064, "step": 55850 }, { "grad_norm": 0.07322006672620773, "learning_rate": 1.2966252005601587e-06, "loss": 0.0072, "step": 55860 }, { "grad_norm": 0.09712930023670197, "learning_rate": 1.2903974290192855e-06, "loss": 0.0071, "step": 55870 }, { "grad_norm": 0.0959840640425682, "learning_rate": 1.284184454169568e-06, "loss": 0.0067, "step": 55880 }, { "grad_norm": 0.08823607861995697, "learning_rate": 1.2779862778983464e-06, "loss": 0.0064, "step": 55890 }, { "grad_norm": 0.0726696252822876, "learning_rate": 1.2718029020884647e-06, "loss": 0.006, "step": 55900 }, { "grad_norm": 0.07998093217611313, "learning_rate": 1.2656343286182703e-06, "loss": 0.007, "step": 55910 }, { "grad_norm": 0.09564083814620972, "learning_rate": 1.2594805593616088e-06, "loss": 0.0065, "step": 55920 }, { "grad_norm": 0.07469171285629272, "learning_rate": 1.2533415961878404e-06, "loss": 0.0068, "step": 55930 }, { "grad_norm": 0.06893724203109741, "learning_rate": 1.2472174409618009e-06, "loss": 0.0064, "step": 55940 }, { "grad_norm": 0.096547432243824, "learning_rate": 1.2411080955438747e-06, "loss": 0.0074, "step": 55950 }, { "grad_norm": 0.08493895083665848, "learning_rate": 1.235013561789894e-06, "loss": 0.0071, "step": 55960 }, { "grad_norm": 0.18256776034832, "learning_rate": 1.2289338415512385e-06, "loss": 0.0063, "step": 55970 }, { "grad_norm": 0.07855137437582016, "learning_rate": 1.222868936674748e-06, "loss": 0.0078, "step": 55980 }, { "grad_norm": 0.08691594749689102, "learning_rate": 1.2168188490027876e-06, "loss": 0.0068, "step": 55990 }, { "grad_norm": 0.06970813125371933, "learning_rate": 1.2107835803732204e-06, "loss": 0.007, "step": 56000 }, { "grad_norm": 0.10036446154117584, "learning_rate": 1.2047631326193964e-06, "loss": 0.0067, "step": 56010 }, { "grad_norm": 0.07959169894456863, "learning_rate": 1.1987575075701696e-06, "loss": 0.0063, "step": 56020 }, { "grad_norm": 0.0829826220870018, "learning_rate": 1.1927667070498916e-06, "loss": 0.0064, "step": 56030 }, { "grad_norm": 0.09225902706384659, "learning_rate": 1.1867907328784067e-06, "loss": 0.0088, "step": 56040 }, { "grad_norm": 0.11042263358831406, "learning_rate": 1.1808295868710518e-06, "loss": 0.007, "step": 56050 }, { "grad_norm": 0.051620516926050186, "learning_rate": 1.174883270838678e-06, "loss": 0.0067, "step": 56060 }, { "grad_norm": 0.08826003223657608, "learning_rate": 1.1689517865876187e-06, "loss": 0.0072, "step": 56070 }, { "grad_norm": 0.0924379825592041, "learning_rate": 1.1630351359196933e-06, "loss": 0.0061, "step": 56080 }, { "grad_norm": 0.1052364781498909, "learning_rate": 1.1571333206322255e-06, "loss": 0.0077, "step": 56090 }, { "grad_norm": 0.13848213851451874, "learning_rate": 1.1512463425180365e-06, "loss": 0.0076, "step": 56100 }, { "grad_norm": 0.07570188492536545, "learning_rate": 1.145374203365429e-06, "loss": 0.0066, "step": 56110 }, { "grad_norm": 0.11927185952663422, "learning_rate": 1.1395169049582155e-06, "loss": 0.0077, "step": 56120 }, { "grad_norm": 0.12985555827617645, "learning_rate": 1.1336744490756722e-06, "loss": 0.0073, "step": 56130 }, { "grad_norm": 0.06996335089206696, "learning_rate": 1.1278468374925967e-06, "loss": 0.0076, "step": 56140 }, { "grad_norm": 0.1127023696899414, "learning_rate": 1.12203407197925e-06, "loss": 0.0072, "step": 56150 }, { "grad_norm": 0.09284178912639618, "learning_rate": 1.116236154301409e-06, "loss": 0.0059, "step": 56160 }, { "grad_norm": 0.0716613158583641, "learning_rate": 1.11045308622032e-06, "loss": 0.0061, "step": 56170 }, { "grad_norm": 0.10125179588794708, "learning_rate": 1.1046848694927337e-06, "loss": 0.0073, "step": 56180 }, { "grad_norm": 0.06912051141262054, "learning_rate": 1.098931505870876e-06, "loss": 0.0062, "step": 56190 }, { "grad_norm": 0.09391280263662338, "learning_rate": 1.0931929971024657e-06, "loss": 0.0063, "step": 56200 }, { "grad_norm": 0.10351191461086273, "learning_rate": 1.0874693449307193e-06, "loss": 0.0072, "step": 56210 }, { "grad_norm": 0.0839790478348732, "learning_rate": 1.081760551094324e-06, "loss": 0.0076, "step": 56220 }, { "grad_norm": 0.0893120989203453, "learning_rate": 1.0760666173274592e-06, "loss": 0.0065, "step": 56230 }, { "grad_norm": 0.07393425703048706, "learning_rate": 1.0703875453597967e-06, "loss": 0.006, "step": 56240 }, { "grad_norm": 0.08614543825387955, "learning_rate": 1.0647233369164845e-06, "loss": 0.0061, "step": 56250 }, { "grad_norm": 0.07021966576576233, "learning_rate": 1.0590739937181625e-06, "loss": 0.0083, "step": 56260 }, { "grad_norm": 0.10566668957471848, "learning_rate": 1.053439517480953e-06, "loss": 0.0083, "step": 56270 }, { "grad_norm": 0.12141069024801254, "learning_rate": 1.047819909916453e-06, "loss": 0.0062, "step": 56280 }, { "grad_norm": 0.09011437743902206, "learning_rate": 1.0422151727317697e-06, "loss": 0.0061, "step": 56290 }, { "grad_norm": 0.07036305963993073, "learning_rate": 1.0366253076294462e-06, "loss": 0.0087, "step": 56300 }, { "grad_norm": 0.09520483762025833, "learning_rate": 1.031050316307558e-06, "loss": 0.0082, "step": 56310 }, { "grad_norm": 0.11909790337085724, "learning_rate": 1.0254902004596333e-06, "loss": 0.0084, "step": 56320 }, { "grad_norm": 0.06439816206693649, "learning_rate": 1.0199449617746882e-06, "loss": 0.0067, "step": 56330 }, { "grad_norm": 0.056243788450956345, "learning_rate": 1.0144146019372247e-06, "loss": 0.0066, "step": 56340 }, { "grad_norm": 0.1481783241033554, "learning_rate": 1.0088991226272048e-06, "loss": 0.007, "step": 56350 }, { "grad_norm": 0.09512222558259964, "learning_rate": 1.003398525520105e-06, "loss": 0.007, "step": 56360 }, { "grad_norm": 0.08098314702510834, "learning_rate": 9.979128122868552e-07, "loss": 0.0076, "step": 56370 }, { "grad_norm": 0.09390042722225189, "learning_rate": 9.924419845938614e-07, "loss": 0.0067, "step": 56380 }, { "grad_norm": 0.0631728246808052, "learning_rate": 9.869860441030276e-07, "loss": 0.0069, "step": 56390 }, { "grad_norm": 0.06777040660381317, "learning_rate": 9.815449924717169e-07, "loss": 0.0075, "step": 56400 }, { "grad_norm": 0.11688526719808578, "learning_rate": 9.761188313527791e-07, "loss": 0.0065, "step": 56410 }, { "grad_norm": 0.08602908998727798, "learning_rate": 9.70707562394546e-07, "loss": 0.0068, "step": 56420 }, { "grad_norm": 0.0791676938533783, "learning_rate": 9.653111872408027e-07, "loss": 0.0069, "step": 56430 }, { "grad_norm": 0.0979098305106163, "learning_rate": 9.599297075308434e-07, "loss": 0.0068, "step": 56440 }, { "grad_norm": 0.07731470465660095, "learning_rate": 9.545631248994048e-07, "loss": 0.0071, "step": 56450 }, { "grad_norm": 0.0984492227435112, "learning_rate": 9.492114409767217e-07, "loss": 0.0063, "step": 56460 }, { "grad_norm": 0.07731679826974869, "learning_rate": 9.438746573884938e-07, "loss": 0.0066, "step": 56470 }, { "grad_norm": 0.06842419505119324, "learning_rate": 9.385527757558909e-07, "loss": 0.0065, "step": 56480 }, { "grad_norm": 0.08894044160842896, "learning_rate": 9.332457976955644e-07, "loss": 0.007, "step": 56490 }, { "grad_norm": 0.08273139595985413, "learning_rate": 9.279537248196247e-07, "loss": 0.0067, "step": 56500 }, { "grad_norm": 0.12962137162685394, "learning_rate": 9.2267655873568e-07, "loss": 0.0065, "step": 56510 }, { "grad_norm": 0.10345017164945602, "learning_rate": 9.174143010467762e-07, "loss": 0.0076, "step": 56520 }, { "grad_norm": 0.09837885200977325, "learning_rate": 9.12166953351462e-07, "loss": 0.0069, "step": 56530 }, { "grad_norm": 0.0914769247174263, "learning_rate": 9.069345172437404e-07, "loss": 0.0086, "step": 56540 }, { "grad_norm": 0.09863147884607315, "learning_rate": 9.017169943130843e-07, "loss": 0.0073, "step": 56550 }, { "grad_norm": 0.060432303696870804, "learning_rate": 8.965143861444425e-07, "loss": 0.0063, "step": 56560 }, { "grad_norm": 0.09878372400999069, "learning_rate": 8.91326694318223e-07, "loss": 0.007, "step": 56570 }, { "grad_norm": 0.1188717931509018, "learning_rate": 8.861539204103098e-07, "loss": 0.0062, "step": 56580 }, { "grad_norm": 0.04473715275526047, "learning_rate": 8.809960659920735e-07, "loss": 0.0071, "step": 56590 }, { "grad_norm": 0.07093077898025513, "learning_rate": 8.758531326303055e-07, "loss": 0.0062, "step": 56600 }, { "grad_norm": 0.06662583351135254, "learning_rate": 8.707251218873169e-07, "loss": 0.0057, "step": 56610 }, { "grad_norm": 0.11496781557798386, "learning_rate": 8.656120353208507e-07, "loss": 0.0073, "step": 56620 }, { "grad_norm": 0.07249018549919128, "learning_rate": 8.605138744841312e-07, "loss": 0.0061, "step": 56630 }, { "grad_norm": 0.09010723978281021, "learning_rate": 8.554306409258417e-07, "loss": 0.0068, "step": 56640 }, { "grad_norm": 0.0665837749838829, "learning_rate": 8.503623361901358e-07, "loss": 0.0073, "step": 56650 }, { "grad_norm": 0.12349724769592285, "learning_rate": 8.453089618166377e-07, "loss": 0.0064, "step": 56660 }, { "grad_norm": 0.06128938868641853, "learning_rate": 8.402705193404137e-07, "loss": 0.0071, "step": 56670 }, { "grad_norm": 0.10685227811336517, "learning_rate": 8.352470102920174e-07, "loss": 0.0064, "step": 56680 }, { "grad_norm": 0.050753213465213776, "learning_rate": 8.302384361974669e-07, "loss": 0.0059, "step": 56690 }, { "grad_norm": 0.09634862840175629, "learning_rate": 8.252447985782231e-07, "loss": 0.0063, "step": 56700 }, { "grad_norm": 0.08832686394453049, "learning_rate": 8.202660989512279e-07, "loss": 0.0067, "step": 56710 }, { "grad_norm": 0.08318780362606049, "learning_rate": 8.153023388288772e-07, "loss": 0.0066, "step": 56720 }, { "grad_norm": 0.08207773417234421, "learning_rate": 8.103535197190204e-07, "loss": 0.0068, "step": 56730 }, { "grad_norm": 0.12475507706403732, "learning_rate": 8.054196431249993e-07, "loss": 0.0072, "step": 56740 }, { "grad_norm": 0.09767181426286697, "learning_rate": 8.005007105455709e-07, "loss": 0.0063, "step": 56750 }, { "grad_norm": 0.22097747027873993, "learning_rate": 7.95596723474995e-07, "loss": 0.0066, "step": 56760 }, { "grad_norm": 0.06441277265548706, "learning_rate": 7.907076834029692e-07, "loss": 0.0069, "step": 56770 }, { "grad_norm": 0.13839489221572876, "learning_rate": 7.858335918146498e-07, "loss": 0.0067, "step": 56780 }, { "grad_norm": 0.05615455284714699, "learning_rate": 7.809744501906635e-07, "loss": 0.0069, "step": 56790 }, { "grad_norm": 0.07372274994850159, "learning_rate": 7.761302600070797e-07, "loss": 0.0069, "step": 56800 }, { "grad_norm": 0.10382505506277084, "learning_rate": 7.713010227354545e-07, "loss": 0.0069, "step": 56810 }, { "grad_norm": 0.07702865451574326, "learning_rate": 7.664867398427589e-07, "loss": 0.0065, "step": 56820 }, { "grad_norm": 0.08267597109079361, "learning_rate": 7.616874127914619e-07, "loss": 0.0062, "step": 56830 }, { "grad_norm": 0.08257434517145157, "learning_rate": 7.569030430394641e-07, "loss": 0.0073, "step": 56840 }, { "grad_norm": 0.0914105474948883, "learning_rate": 7.521336320401306e-07, "loss": 0.0066, "step": 56850 }, { "grad_norm": 0.0640789195895195, "learning_rate": 7.473791812422915e-07, "loss": 0.0059, "step": 56860 }, { "grad_norm": 0.07536149024963379, "learning_rate": 7.426396920902134e-07, "loss": 0.0067, "step": 56870 }, { "grad_norm": 0.07993943244218826, "learning_rate": 7.379151660236283e-07, "loss": 0.0079, "step": 56880 }, { "grad_norm": 0.11387239396572113, "learning_rate": 7.332056044777324e-07, "loss": 0.0071, "step": 56890 }, { "grad_norm": 0.11038673669099808, "learning_rate": 7.285110088831537e-07, "loss": 0.0067, "step": 56900 }, { "grad_norm": 0.11529604345560074, "learning_rate": 7.238313806659902e-07, "loss": 0.0059, "step": 56910 }, { "grad_norm": 0.07461941242218018, "learning_rate": 7.191667212477993e-07, "loss": 0.0064, "step": 56920 }, { "grad_norm": 0.07695499807596207, "learning_rate": 7.145170320455697e-07, "loss": 0.0063, "step": 56930 }, { "grad_norm": 0.10374815762042999, "learning_rate": 7.098823144717604e-07, "loss": 0.0066, "step": 56940 }, { "grad_norm": 0.09594341367483139, "learning_rate": 7.052625699342674e-07, "loss": 0.0071, "step": 56950 }, { "grad_norm": 0.11771302670240402, "learning_rate": 7.006577998364628e-07, "loss": 0.0063, "step": 56960 }, { "grad_norm": 0.06776461750268936, "learning_rate": 6.96068005577133e-07, "loss": 0.0066, "step": 56970 }, { "grad_norm": 0.08294625580310822, "learning_rate": 6.914931885505627e-07, "loss": 0.0077, "step": 56980 }, { "grad_norm": 0.06416236609220505, "learning_rate": 6.869333501464347e-07, "loss": 0.0062, "step": 56990 }, { "grad_norm": 0.05639949440956116, "learning_rate": 6.823884917499246e-07, "loss": 0.0073, "step": 57000 }, { "grad_norm": 0.07018250226974487, "learning_rate": 6.778586147416278e-07, "loss": 0.0062, "step": 57010 }, { "grad_norm": 0.06064383313059807, "learning_rate": 6.733437204976156e-07, "loss": 0.0064, "step": 57020 }, { "grad_norm": 0.06304698437452316, "learning_rate": 6.688438103893857e-07, "loss": 0.0062, "step": 57030 }, { "grad_norm": 0.07796572148799896, "learning_rate": 6.64358885783889e-07, "loss": 0.0068, "step": 57040 }, { "grad_norm": 0.06624588370323181, "learning_rate": 6.598889480435299e-07, "loss": 0.006, "step": 57050 }, { "grad_norm": 0.08518269658088684, "learning_rate": 6.554339985261615e-07, "loss": 0.0066, "step": 57060 }, { "grad_norm": 0.07016580551862717, "learning_rate": 6.509940385850733e-07, "loss": 0.0071, "step": 57070 }, { "grad_norm": 0.12230491638183594, "learning_rate": 6.465690695690141e-07, "loss": 0.0065, "step": 57080 }, { "grad_norm": 0.11913203448057175, "learning_rate": 6.421590928221699e-07, "loss": 0.0072, "step": 57090 }, { "grad_norm": 0.07403715699911118, "learning_rate": 6.377641096841691e-07, "loss": 0.007, "step": 57100 }, { "grad_norm": 0.0772608295083046, "learning_rate": 6.333841214901048e-07, "loss": 0.0066, "step": 57110 }, { "grad_norm": 0.08149489015340805, "learning_rate": 6.290191295704906e-07, "loss": 0.0069, "step": 57120 }, { "grad_norm": 0.06387846916913986, "learning_rate": 6.246691352513046e-07, "loss": 0.0066, "step": 57130 }, { "grad_norm": 0.08202209323644638, "learning_rate": 6.203341398539452e-07, "loss": 0.0061, "step": 57140 }, { "grad_norm": 0.07786236703395844, "learning_rate": 6.160141446952872e-07, "loss": 0.0062, "step": 57150 }, { "grad_norm": 0.09851007908582687, "learning_rate": 6.11709151087625e-07, "loss": 0.0072, "step": 57160 }, { "grad_norm": 0.08990272134542465, "learning_rate": 6.074191603386958e-07, "loss": 0.0063, "step": 57170 }, { "grad_norm": 0.08209814876317978, "learning_rate": 6.031441737516907e-07, "loss": 0.0065, "step": 57180 }, { "grad_norm": 0.07998160272836685, "learning_rate": 5.988841926252431e-07, "loss": 0.0063, "step": 57190 }, { "grad_norm": 0.0931289866566658, "learning_rate": 5.946392182534066e-07, "loss": 0.0068, "step": 57200 }, { "grad_norm": 0.14176729321479797, "learning_rate": 5.90409251925711e-07, "loss": 0.0065, "step": 57210 }, { "grad_norm": 0.09748245030641556, "learning_rate": 5.861942949270949e-07, "loss": 0.0066, "step": 57220 }, { "grad_norm": 0.06421276181936264, "learning_rate": 5.819943485379564e-07, "loss": 0.0091, "step": 57230 }, { "grad_norm": 0.10831364244222641, "learning_rate": 5.778094140341306e-07, "loss": 0.0062, "step": 57240 }, { "grad_norm": 0.08965663611888885, "learning_rate": 5.736394926868893e-07, "loss": 0.0065, "step": 57250 }, { "grad_norm": 0.0971086174249649, "learning_rate": 5.694845857629416e-07, "loss": 0.0072, "step": 57260 }, { "grad_norm": 0.11526734381914139, "learning_rate": 5.653446945244334e-07, "loss": 0.0065, "step": 57270 }, { "grad_norm": 0.07492523640394211, "learning_rate": 5.612198202289698e-07, "loss": 0.0076, "step": 57280 }, { "grad_norm": 0.08224108070135117, "learning_rate": 5.571099641295596e-07, "loss": 0.0059, "step": 57290 }, { "grad_norm": 0.05440908670425415, "learning_rate": 5.530151274746875e-07, "loss": 0.006, "step": 57300 }, { "grad_norm": 0.06181987002491951, "learning_rate": 5.489353115082418e-07, "loss": 0.0072, "step": 57310 }, { "grad_norm": 0.08729622513055801, "learning_rate": 5.4487051746957e-07, "loss": 0.0075, "step": 57320 }, { "grad_norm": 0.08969301730394363, "learning_rate": 5.408207465934511e-07, "loss": 0.0058, "step": 57330 }, { "grad_norm": 0.11120494455099106, "learning_rate": 5.3678600011009e-07, "loss": 0.007, "step": 57340 }, { "grad_norm": 0.0754445493221283, "learning_rate": 5.327662792451449e-07, "loss": 0.0067, "step": 57350 }, { "grad_norm": 0.08885379135608673, "learning_rate": 5.287615852196947e-07, "loss": 0.0077, "step": 57360 }, { "grad_norm": 0.19289530813694, "learning_rate": 5.247719192502665e-07, "loss": 0.0075, "step": 57370 }, { "grad_norm": 0.08636406064033508, "learning_rate": 5.207972825488128e-07, "loss": 0.0065, "step": 57380 }, { "grad_norm": 0.0884607583284378, "learning_rate": 5.168376763227178e-07, "loss": 0.0064, "step": 57390 }, { "grad_norm": 0.08407892286777496, "learning_rate": 5.12893101774814e-07, "loss": 0.0071, "step": 57400 }, { "grad_norm": 0.13885800540447235, "learning_rate": 5.089635601033483e-07, "loss": 0.0075, "step": 57410 }, { "grad_norm": 0.0555858351290226, "learning_rate": 5.050490525020213e-07, "loss": 0.0067, "step": 57420 }, { "grad_norm": 0.052732668817043304, "learning_rate": 5.011495801599541e-07, "loss": 0.0065, "step": 57430 }, { "grad_norm": 0.09781700372695923, "learning_rate": 4.972651442616994e-07, "loss": 0.0068, "step": 57440 }, { "grad_norm": 0.08312767744064331, "learning_rate": 4.933957459872574e-07, "loss": 0.0073, "step": 57450 }, { "grad_norm": 0.10724243521690369, "learning_rate": 4.895413865120324e-07, "loss": 0.0058, "step": 57460 }, { "grad_norm": 0.08003838360309601, "learning_rate": 4.857020670068935e-07, "loss": 0.0061, "step": 57470 }, { "grad_norm": 0.09726476669311523, "learning_rate": 4.818777886381132e-07, "loss": 0.0065, "step": 57480 }, { "grad_norm": 0.08255153149366379, "learning_rate": 4.780685525674122e-07, "loss": 0.0064, "step": 57490 }, { "grad_norm": 0.07267855107784271, "learning_rate": 4.7427435995193724e-07, "loss": 0.006, "step": 57500 }, { "grad_norm": 0.12834331393241882, "learning_rate": 4.7049521194425515e-07, "loss": 0.0061, "step": 57510 }, { "grad_norm": 0.06528450548648834, "learning_rate": 4.667311096923754e-07, "loss": 0.0061, "step": 57520 }, { "grad_norm": 0.12124704569578171, "learning_rate": 4.6298205433973895e-07, "loss": 0.0064, "step": 57530 }, { "grad_norm": 0.07690822333097458, "learning_rate": 4.5924804702520696e-07, "loss": 0.0068, "step": 57540 }, { "grad_norm": 0.09083429723978043, "learning_rate": 4.5552908888306655e-07, "loss": 0.0063, "step": 57550 }, { "grad_norm": 0.07356053590774536, "learning_rate": 4.5182518104304185e-07, "loss": 0.0066, "step": 57560 }, { "grad_norm": 0.16749705374240875, "learning_rate": 4.4813632463028277e-07, "loss": 0.0062, "step": 57570 }, { "grad_norm": 0.09135382622480392, "learning_rate": 4.444625207653763e-07, "loss": 0.0061, "step": 57580 }, { "grad_norm": 0.06889518350362778, "learning_rate": 4.4080377056430753e-07, "loss": 0.0065, "step": 57590 }, { "grad_norm": 0.07720253616571426, "learning_rate": 4.371600751385263e-07, "loss": 0.0061, "step": 57600 }, { "grad_norm": 0.09701453149318695, "learning_rate": 4.3353143559487495e-07, "loss": 0.007, "step": 57610 }, { "grad_norm": 0.12307962775230408, "learning_rate": 4.2991785303565513e-07, "loss": 0.0066, "step": 57620 }, { "grad_norm": 0.06467270106077194, "learning_rate": 4.2631932855856647e-07, "loss": 0.006, "step": 57630 }, { "grad_norm": 0.05582967773079872, "learning_rate": 4.2273586325674576e-07, "loss": 0.0064, "step": 57640 }, { "grad_norm": 0.06540671736001968, "learning_rate": 4.1916745821876103e-07, "loss": 0.0059, "step": 57650 }, { "grad_norm": 0.05841785669326782, "learning_rate": 4.156141145285897e-07, "loss": 0.0062, "step": 57660 }, { "grad_norm": 0.07720349729061127, "learning_rate": 4.1207583326566267e-07, "loss": 0.0067, "step": 57670 }, { "grad_norm": 0.07085420936346054, "learning_rate": 4.085526155047925e-07, "loss": 0.0066, "step": 57680 }, { "grad_norm": 0.09714122116565704, "learning_rate": 4.050444623162564e-07, "loss": 0.0058, "step": 57690 }, { "grad_norm": 0.10612824559211731, "learning_rate": 4.015513747657351e-07, "loss": 0.0072, "step": 57700 }, { "grad_norm": 0.07002730667591095, "learning_rate": 3.9807335391433554e-07, "loss": 0.0056, "step": 57710 }, { "grad_norm": 0.09327173233032227, "learning_rate": 3.946104008185847e-07, "loss": 0.0062, "step": 57720 }, { "grad_norm": 0.10749252140522003, "learning_rate": 3.9116251653044113e-07, "loss": 0.0067, "step": 57730 }, { "grad_norm": 0.09144523739814758, "learning_rate": 3.877297020972781e-07, "loss": 0.0062, "step": 57740 }, { "grad_norm": 0.08527007699012756, "learning_rate": 3.8431195856190036e-07, "loss": 0.0061, "step": 57750 }, { "grad_norm": 0.10273458063602448, "learning_rate": 3.8090928696251635e-07, "loss": 0.0062, "step": 57760 }, { "grad_norm": 0.10987849533557892, "learning_rate": 3.77521688332777e-07, "loss": 0.0062, "step": 57770 }, { "grad_norm": 0.08150089532136917, "learning_rate": 3.741491637017425e-07, "loss": 0.0061, "step": 57780 }, { "grad_norm": 0.09712319076061249, "learning_rate": 3.707917140939043e-07, "loss": 0.0071, "step": 57790 }, { "grad_norm": 0.09814527630805969, "learning_rate": 3.6744934052915235e-07, "loss": 0.007, "step": 57800 }, { "grad_norm": 0.07020124793052673, "learning_rate": 3.641220440228188e-07, "loss": 0.0064, "step": 57810 }, { "grad_norm": 0.08599499613046646, "learning_rate": 3.608098255856562e-07, "loss": 0.0055, "step": 57820 }, { "grad_norm": 0.07563385367393494, "learning_rate": 3.575126862238154e-07, "loss": 0.0068, "step": 57830 }, { "grad_norm": 0.10861341655254364, "learning_rate": 3.5423062693888955e-07, "loss": 0.0064, "step": 57840 }, { "grad_norm": 0.199043869972229, "learning_rate": 3.509636487278756e-07, "loss": 0.0078, "step": 57850 }, { "grad_norm": 0.11987337470054626, "learning_rate": 3.4771175258320186e-07, "loss": 0.0061, "step": 57860 }, { "grad_norm": 0.08255770057439804, "learning_rate": 3.4447493949270047e-07, "loss": 0.006, "step": 57870 }, { "grad_norm": 0.09173721820116043, "learning_rate": 3.4125321043964045e-07, "loss": 0.0059, "step": 57880 }, { "grad_norm": 0.08233096450567245, "learning_rate": 3.380465664026833e-07, "loss": 0.0071, "step": 57890 }, { "grad_norm": 0.07975858449935913, "learning_rate": 3.348550083559388e-07, "loss": 0.007, "step": 57900 }, { "grad_norm": 0.10749393701553345, "learning_rate": 3.316785372689091e-07, "loss": 0.0066, "step": 57910 }, { "grad_norm": 0.0757325068116188, "learning_rate": 3.285171541065224e-07, "loss": 0.0071, "step": 57920 }, { "grad_norm": 0.07532931864261627, "learning_rate": 3.253708598291272e-07, "loss": 0.0059, "step": 57930 }, { "grad_norm": 0.057099174708127975, "learning_rate": 3.2223965539248116e-07, "loss": 0.0063, "step": 57940 }, { "grad_norm": 0.06004190444946289, "learning_rate": 3.1912354174776227e-07, "loss": 0.0059, "step": 57950 }, { "grad_norm": 0.0815768912434578, "learning_rate": 3.1602251984155786e-07, "loss": 0.0068, "step": 57960 }, { "grad_norm": 0.104821115732193, "learning_rate": 3.1293659061589207e-07, "loss": 0.0066, "step": 57970 }, { "grad_norm": 0.11018777638673782, "learning_rate": 3.098657550081707e-07, "loss": 0.0064, "step": 57980 }, { "grad_norm": 0.09055496752262115, "learning_rate": 3.068100139512475e-07, "loss": 0.0062, "step": 57990 }, { "grad_norm": 0.07189319282770157, "learning_rate": 3.037693683733689e-07, "loss": 0.0062, "step": 58000 }, { "grad_norm": 0.08086642622947693, "learning_rate": 3.007438191982015e-07, "loss": 0.0061, "step": 58010 }, { "grad_norm": 0.05561289191246033, "learning_rate": 2.9773336734482684e-07, "loss": 0.0063, "step": 58020 }, { "grad_norm": 0.06345638632774353, "learning_rate": 2.9473801372774667e-07, "loss": 0.0061, "step": 58030 }, { "grad_norm": 0.08700120449066162, "learning_rate": 2.91757759256861e-07, "loss": 0.0061, "step": 58040 }, { "grad_norm": 0.08842670172452927, "learning_rate": 2.887926048375067e-07, "loss": 0.0069, "step": 58050 }, { "grad_norm": 0.06747233867645264, "learning_rate": 2.858425513704022e-07, "loss": 0.007, "step": 58060 }, { "grad_norm": 0.09668099135160446, "learning_rate": 2.8290759975170834e-07, "loss": 0.0071, "step": 58070 }, { "grad_norm": 0.07792861014604568, "learning_rate": 2.799877508729787e-07, "loss": 0.006, "step": 58080 }, { "grad_norm": 0.057999078184366226, "learning_rate": 2.770830056211926e-07, "loss": 0.0063, "step": 58090 }, { "grad_norm": 0.08370506018400192, "learning_rate": 2.741933648787331e-07, "loss": 0.0091, "step": 58100 }, { "grad_norm": 0.07984357327222824, "learning_rate": 2.7131882952339263e-07, "loss": 0.0066, "step": 58110 }, { "grad_norm": 0.07073558121919632, "learning_rate": 2.684594004283836e-07, "loss": 0.0068, "step": 58120 }, { "grad_norm": 0.08486814051866531, "learning_rate": 2.6561507846232234e-07, "loss": 0.0072, "step": 58130 }, { "grad_norm": 0.08944904059171677, "learning_rate": 2.6278586448924005e-07, "loss": 0.0063, "step": 58140 }, { "grad_norm": 0.08689256757497787, "learning_rate": 2.5997175936857685e-07, "loss": 0.0056, "step": 58150 }, { "grad_norm": 0.11266276985406876, "learning_rate": 2.57172763955188e-07, "loss": 0.0077, "step": 58160 }, { "grad_norm": 0.10343094915151596, "learning_rate": 2.543888790993265e-07, "loss": 0.0066, "step": 58170 }, { "grad_norm": 0.06526963412761688, "learning_rate": 2.5162010564666607e-07, "loss": 0.0067, "step": 58180 }, { "grad_norm": 0.1367470622062683, "learning_rate": 2.488664444382893e-07, "loss": 0.0074, "step": 58190 }, { "grad_norm": 0.11682809889316559, "learning_rate": 2.461278963106828e-07, "loss": 0.0083, "step": 58200 }, { "grad_norm": 0.059595342725515366, "learning_rate": 2.434044620957421e-07, "loss": 0.0081, "step": 58210 }, { "grad_norm": 0.08532867580652237, "learning_rate": 2.406961426207832e-07, "loss": 0.0066, "step": 58220 }, { "grad_norm": 0.0781676322221756, "learning_rate": 2.380029387085203e-07, "loss": 0.0064, "step": 58230 }, { "grad_norm": 0.09775584936141968, "learning_rate": 2.353248511770767e-07, "loss": 0.0065, "step": 58240 }, { "grad_norm": 0.10317344218492508, "learning_rate": 2.3266188083997942e-07, "loss": 0.0071, "step": 58250 }, { "grad_norm": 0.08131147176027298, "learning_rate": 2.3001402850617027e-07, "loss": 0.0061, "step": 58260 }, { "grad_norm": 0.06844092160463333, "learning_rate": 2.2738129498000581e-07, "loss": 0.0057, "step": 58270 }, { "grad_norm": 0.07482045888900757, "learning_rate": 2.2476368106122414e-07, "loss": 0.0055, "step": 58280 }, { "grad_norm": 0.08089126646518707, "learning_rate": 2.2216118754500582e-07, "loss": 0.0068, "step": 58290 }, { "grad_norm": 0.07371719181537628, "learning_rate": 2.1957381522190735e-07, "loss": 0.008, "step": 58300 }, { "grad_norm": 0.0999072790145874, "learning_rate": 2.1700156487790558e-07, "loss": 0.0066, "step": 58310 }, { "grad_norm": 0.07756821811199188, "learning_rate": 2.1444443729439213e-07, "loss": 0.0062, "step": 58320 }, { "grad_norm": 0.09447060525417328, "learning_rate": 2.1190243324814007e-07, "loss": 0.0067, "step": 58330 }, { "grad_norm": 0.07132629305124283, "learning_rate": 2.0937555351135395e-07, "loss": 0.0063, "step": 58340 }, { "grad_norm": 0.09806942939758301, "learning_rate": 2.0686379885162532e-07, "loss": 0.0066, "step": 58350 }, { "grad_norm": 0.07524121552705765, "learning_rate": 2.0436717003196604e-07, "loss": 0.0068, "step": 58360 }, { "grad_norm": 0.08632530272006989, "learning_rate": 2.0188566781078057e-07, "loss": 0.0064, "step": 58370 }, { "grad_norm": 0.07583147287368774, "learning_rate": 1.994192929418881e-07, "loss": 0.0064, "step": 58380 }, { "grad_norm": 0.0966869667172432, "learning_rate": 1.9696804617451158e-07, "loss": 0.0062, "step": 58390 }, { "grad_norm": 0.17969784140586853, "learning_rate": 1.9453192825326093e-07, "loss": 0.0076, "step": 58400 }, { "grad_norm": 0.10078578442335129, "learning_rate": 1.9211093991817753e-07, "loss": 0.0075, "step": 58410 }, { "grad_norm": 0.0755714699625969, "learning_rate": 1.8970508190468973e-07, "loss": 0.0069, "step": 58420 }, { "grad_norm": 0.1278602033853531, "learning_rate": 1.8731435494362958e-07, "loss": 0.0065, "step": 58430 }, { "grad_norm": 0.07923802733421326, "learning_rate": 1.849387597612495e-07, "loss": 0.0068, "step": 58440 }, { "grad_norm": 0.05802714079618454, "learning_rate": 1.8257829707917228e-07, "loss": 0.006, "step": 58450 }, { "grad_norm": 0.08512682467699051, "learning_rate": 1.8023296761446317e-07, "loss": 0.007, "step": 58460 }, { "grad_norm": 0.08137208968400955, "learning_rate": 1.7790277207956341e-07, "loss": 0.0061, "step": 58470 }, { "grad_norm": 0.055818308144807816, "learning_rate": 1.7558771118232343e-07, "loss": 0.0066, "step": 58480 }, { "grad_norm": 0.08104794472455978, "learning_rate": 1.7328778562599734e-07, "loss": 0.0062, "step": 58490 }, { "grad_norm": 0.10077280551195145, "learning_rate": 1.7100299610924298e-07, "loss": 0.0065, "step": 58500 }, { "grad_norm": 0.06608523428440094, "learning_rate": 1.6873334332612733e-07, "loss": 0.007, "step": 58510 }, { "grad_norm": 0.07475642114877701, "learning_rate": 1.6647882796609894e-07, "loss": 0.0061, "step": 58520 }, { "grad_norm": 0.09704010933637619, "learning_rate": 1.6423945071402102e-07, "loss": 0.0068, "step": 58530 }, { "grad_norm": 0.05010969936847687, "learning_rate": 1.6201521225016614e-07, "loss": 0.0083, "step": 58540 }, { "grad_norm": 0.1029738262295723, "learning_rate": 1.598061132501938e-07, "loss": 0.007, "step": 58550 }, { "grad_norm": 0.069417804479599, "learning_rate": 1.576121543851672e-07, "loss": 0.0065, "step": 58560 }, { "grad_norm": 0.058729398995637894, "learning_rate": 1.5543333632155876e-07, "loss": 0.0085, "step": 58570 }, { "grad_norm": 0.050871629267930984, "learning_rate": 1.5326965972123352e-07, "loss": 0.0071, "step": 58580 }, { "grad_norm": 0.08076394349336624, "learning_rate": 1.5112112524146016e-07, "loss": 0.0067, "step": 58590 }, { "grad_norm": 0.10271620005369186, "learning_rate": 1.4898773353489992e-07, "loss": 0.0062, "step": 58600 }, { "grad_norm": 0.07129647582769394, "learning_rate": 1.4686948524962884e-07, "loss": 0.0065, "step": 58610 }, { "grad_norm": 0.10104458034038544, "learning_rate": 1.4476638102911556e-07, "loss": 0.007, "step": 58620 }, { "grad_norm": 0.09434467554092407, "learning_rate": 1.4267842151222123e-07, "loss": 0.0066, "step": 58630 }, { "grad_norm": 0.08337137848138809, "learning_rate": 1.4060560733321626e-07, "loss": 0.0058, "step": 58640 }, { "grad_norm": 0.07345499098300934, "learning_rate": 1.385479391217692e-07, "loss": 0.0054, "step": 58650 }, { "grad_norm": 0.10439223796129227, "learning_rate": 1.365054175029412e-07, "loss": 0.0067, "step": 58660 }, { "grad_norm": 0.09391757100820541, "learning_rate": 1.3447804309719702e-07, "loss": 0.0071, "step": 58670 }, { "grad_norm": 0.05111059173941612, "learning_rate": 1.3246581652040512e-07, "loss": 0.0074, "step": 58680 }, { "grad_norm": 0.07161308825016022, "learning_rate": 1.3046873838381546e-07, "loss": 0.0067, "step": 58690 }, { "grad_norm": 0.057411499321460724, "learning_rate": 1.2848680929409828e-07, "loss": 0.0058, "step": 58700 }, { "grad_norm": 0.06516104191541672, "learning_rate": 1.2652002985331091e-07, "loss": 0.0069, "step": 58710 }, { "grad_norm": 0.07948994636535645, "learning_rate": 1.2456840065889764e-07, "loss": 0.0073, "step": 58720 }, { "grad_norm": 0.057799383997917175, "learning_rate": 1.226319223037231e-07, "loss": 0.0058, "step": 58730 }, { "grad_norm": 0.07574566453695297, "learning_rate": 1.2071059537603902e-07, "loss": 0.0071, "step": 58740 }, { "grad_norm": 0.10327783972024918, "learning_rate": 1.1880442045948403e-07, "loss": 0.0068, "step": 58750 }, { "grad_norm": 0.09514527022838593, "learning_rate": 1.1691339813311164e-07, "loss": 0.0066, "step": 58760 }, { "grad_norm": 0.10251381993293762, "learning_rate": 1.1503752897136233e-07, "loss": 0.007, "step": 58770 }, { "grad_norm": 0.08634960651397705, "learning_rate": 1.1317681354407472e-07, "loss": 0.0067, "step": 58780 }, { "grad_norm": 0.10292153805494308, "learning_rate": 1.1133125241649111e-07, "loss": 0.0068, "step": 58790 }, { "grad_norm": 0.06777399033308029, "learning_rate": 1.0950084614922973e-07, "loss": 0.0076, "step": 58800 }, { "grad_norm": 0.06316947937011719, "learning_rate": 1.0768559529834021e-07, "loss": 0.0068, "step": 58810 }, { "grad_norm": 0.05832381546497345, "learning_rate": 1.0588550041522594e-07, "loss": 0.0066, "step": 58820 }, { "grad_norm": 0.09500697255134583, "learning_rate": 1.0410056204672169e-07, "loss": 0.0063, "step": 58830 }, { "grad_norm": 0.06461962312459946, "learning_rate": 1.0233078073504376e-07, "loss": 0.0069, "step": 58840 }, { "grad_norm": 0.10950976610183716, "learning_rate": 1.0057615701780654e-07, "loss": 0.0066, "step": 58850 }, { "grad_norm": 0.1397624909877777, "learning_rate": 9.883669142801144e-08, "loss": 0.0057, "step": 58860 }, { "grad_norm": 0.06939654797315598, "learning_rate": 9.711238449406356e-08, "loss": 0.0058, "step": 58870 }, { "grad_norm": 0.11528625339269638, "learning_rate": 9.540323673976614e-08, "loss": 0.0068, "step": 58880 }, { "grad_norm": 0.0657789409160614, "learning_rate": 9.370924868430942e-08, "loss": 0.0073, "step": 58890 }, { "grad_norm": 0.08440042287111282, "learning_rate": 9.203042084228175e-08, "loss": 0.0063, "step": 58900 }, { "grad_norm": 0.09838616102933884, "learning_rate": 9.036675372366965e-08, "loss": 0.0065, "step": 58910 }, { "grad_norm": 0.11643793433904648, "learning_rate": 8.871824783385218e-08, "loss": 0.0061, "step": 58920 }, { "grad_norm": 0.06182451546192169, "learning_rate": 8.7084903673601e-08, "loss": 0.0085, "step": 58930 }, { "grad_norm": 0.11108464747667313, "learning_rate": 8.546672173908032e-08, "loss": 0.0072, "step": 58940 }, { "grad_norm": 0.11002903431653976, "learning_rate": 8.386370252185249e-08, "loss": 0.0068, "step": 58950 }, { "grad_norm": 0.10755521804094315, "learning_rate": 8.227584650887243e-08, "loss": 0.006, "step": 58960 }, { "grad_norm": 0.08747213333845139, "learning_rate": 8.070315418249319e-08, "loss": 0.0063, "step": 58970 }, { "grad_norm": 0.06906667351722717, "learning_rate": 7.914562602044929e-08, "loss": 0.0061, "step": 58980 }, { "grad_norm": 0.06407483667135239, "learning_rate": 7.76032624958789e-08, "loss": 0.0066, "step": 58990 }, { "grad_norm": 0.06635560095310211, "learning_rate": 7.607606407731282e-08, "loss": 0.006, "step": 59000 }, { "grad_norm": 0.1569792777299881, "learning_rate": 7.45640312286744e-08, "loss": 0.0067, "step": 59010 }, { "grad_norm": 0.06721513718366623, "learning_rate": 7.306716440927952e-08, "loss": 0.0062, "step": 59020 }, { "grad_norm": 0.062348950654268265, "learning_rate": 7.15854640738367e-08, "loss": 0.0061, "step": 59030 }, { "grad_norm": 0.1348564624786377, "learning_rate": 7.011893067244701e-08, "loss": 0.0077, "step": 59040 }, { "grad_norm": 0.11814457178115845, "learning_rate": 6.866756465060408e-08, "loss": 0.0065, "step": 59050 }, { "grad_norm": 0.08580672740936279, "learning_rate": 6.723136644918859e-08, "loss": 0.0069, "step": 59060 }, { "grad_norm": 0.1048821210861206, "learning_rate": 6.581033650449042e-08, "loss": 0.0067, "step": 59070 }, { "grad_norm": 0.15308940410614014, "learning_rate": 6.440447524817539e-08, "loss": 0.0075, "step": 59080 }, { "grad_norm": 0.06501541286706924, "learning_rate": 6.301378310730743e-08, "loss": 0.0062, "step": 59090 }, { "grad_norm": 0.09396933019161224, "learning_rate": 6.163826050434307e-08, "loss": 0.006, "step": 59100 }, { "grad_norm": 0.07993936538696289, "learning_rate": 6.027790785713139e-08, "loss": 0.0067, "step": 59110 }, { "grad_norm": 0.0688878744840622, "learning_rate": 5.89327255789085e-08, "loss": 0.0061, "step": 59120 }, { "grad_norm": 0.12497468292713165, "learning_rate": 5.7602714078303085e-08, "loss": 0.0067, "step": 59130 }, { "grad_norm": 0.08893644064664841, "learning_rate": 5.628787375934197e-08, "loss": 0.0071, "step": 59140 }, { "grad_norm": 0.07909281551837921, "learning_rate": 5.498820502143898e-08, "loss": 0.0063, "step": 59150 }, { "grad_norm": 0.08777375519275665, "learning_rate": 5.370370825939497e-08, "loss": 0.0063, "step": 59160 }, { "grad_norm": 0.13492035865783691, "learning_rate": 5.243438386340893e-08, "loss": 0.0069, "step": 59170 }, { "grad_norm": 0.12447843700647354, "learning_rate": 5.118023221907242e-08, "loss": 0.0065, "step": 59180 }, { "grad_norm": 0.08382692188024521, "learning_rate": 4.994125370735292e-08, "loss": 0.006, "step": 59190 }, { "grad_norm": 0.15887857973575592, "learning_rate": 4.871744870462713e-08, "loss": 0.0098, "step": 59200 }, { "grad_norm": 0.0890391543507576, "learning_rate": 4.7508817582658794e-08, "loss": 0.0072, "step": 59210 }, { "grad_norm": 0.1113073006272316, "learning_rate": 4.631536070858755e-08, "loss": 0.0059, "step": 59220 }, { "grad_norm": 0.06564503163099289, "learning_rate": 4.513707844495674e-08, "loss": 0.0061, "step": 59230 }, { "grad_norm": 0.10341241210699081, "learning_rate": 4.3973971149702255e-08, "loss": 0.0056, "step": 59240 }, { "grad_norm": 0.09534814208745956, "learning_rate": 4.2826039176147025e-08, "loss": 0.0067, "step": 59250 }, { "grad_norm": 0.07210744172334671, "learning_rate": 4.169328287299545e-08, "loss": 0.0062, "step": 59260 }, { "grad_norm": 0.10298675298690796, "learning_rate": 4.057570258435006e-08, "loss": 0.0078, "step": 59270 }, { "grad_norm": 0.09522845596075058, "learning_rate": 3.947329864970595e-08, "loss": 0.0065, "step": 59280 }, { "grad_norm": 0.09617350995540619, "learning_rate": 3.8386071403939686e-08, "loss": 0.0077, "step": 59290 }, { "grad_norm": 0.07568628340959549, "learning_rate": 3.731402117733152e-08, "loss": 0.0054, "step": 59300 }, { "grad_norm": 0.11274855583906174, "learning_rate": 3.625714829552651e-08, "loss": 0.0061, "step": 59310 }, { "grad_norm": 0.0647929459810257, "learning_rate": 3.5215453079590065e-08, "loss": 0.0065, "step": 59320 }, { "grad_norm": 0.09109309315681458, "learning_rate": 3.4188935845952396e-08, "loss": 0.0066, "step": 59330 }, { "grad_norm": 0.06753165274858475, "learning_rate": 3.3177596906447396e-08, "loss": 0.0072, "step": 59340 }, { "grad_norm": 0.08678609877824783, "learning_rate": 3.218143656829043e-08, "loss": 0.0065, "step": 59350 }, { "grad_norm": 0.10893617570400238, "learning_rate": 3.120045513408387e-08, "loss": 0.0068, "step": 59360 }, { "grad_norm": 0.1089242696762085, "learning_rate": 3.02346529018338e-08, "loss": 0.0063, "step": 59370 }, { "grad_norm": 0.07509353011846542, "learning_rate": 2.9284030164922204e-08, "loss": 0.0063, "step": 59380 }, { "grad_norm": 0.08842888474464417, "learning_rate": 2.8348587212123634e-08, "loss": 0.0063, "step": 59390 }, { "grad_norm": 0.07812711596488953, "learning_rate": 2.7428324327594125e-08, "loss": 0.0066, "step": 59400 }, { "grad_norm": 0.07112139463424683, "learning_rate": 2.6523241790893383e-08, "loss": 0.0072, "step": 59410 }, { "grad_norm": 0.08781395852565765, "learning_rate": 2.563333987695704e-08, "loss": 0.007, "step": 59420 }, { "grad_norm": 0.06430795043706894, "learning_rate": 2.4758618856118852e-08, "loss": 0.0062, "step": 59430 }, { "grad_norm": 0.07358712702989578, "learning_rate": 2.3899078994088497e-08, "loss": 0.0069, "step": 59440 }, { "grad_norm": 0.07902322709560394, "learning_rate": 2.3054720551973775e-08, "loss": 0.0067, "step": 59450 }, { "grad_norm": 0.06555849313735962, "learning_rate": 2.222554378627506e-08, "loss": 0.0061, "step": 59460 }, { "grad_norm": 0.11285444349050522, "learning_rate": 2.1411548948868653e-08, "loss": 0.0069, "step": 59470 }, { "grad_norm": 0.0856691300868988, "learning_rate": 2.0612736287023426e-08, "loss": 0.0081, "step": 59480 }, { "grad_norm": 0.10188251733779907, "learning_rate": 1.9829106043400826e-08, "loss": 0.0069, "step": 59490 }, { "grad_norm": 0.11103513091802597, "learning_rate": 1.9060658456043768e-08, "loss": 0.0069, "step": 59500 }, { "grad_norm": 0.07826752960681915, "learning_rate": 1.830739375838775e-08, "loss": 0.0061, "step": 59510 }, { "grad_norm": 0.07328028231859207, "learning_rate": 1.7569312179260832e-08, "loss": 0.0081, "step": 59520 }, { "grad_norm": 0.08589682728052139, "learning_rate": 1.684641394286146e-08, "loss": 0.007, "step": 59530 }, { "grad_norm": 0.08497127890586853, "learning_rate": 1.6138699268797296e-08, "loss": 0.0066, "step": 59540 }, { "grad_norm": 0.09096864610910416, "learning_rate": 1.5446168372046376e-08, "loss": 0.0071, "step": 59550 }, { "grad_norm": 0.07711384445428848, "learning_rate": 1.4768821462984861e-08, "loss": 0.0067, "step": 59560 }, { "grad_norm": 0.09986124187707901, "learning_rate": 1.4106658747370383e-08, "loss": 0.0075, "step": 59570 }, { "grad_norm": 0.09459879994392395, "learning_rate": 1.3459680426353149e-08, "loss": 0.0062, "step": 59580 }, { "grad_norm": 0.07766538858413696, "learning_rate": 1.2827886696464841e-08, "loss": 0.0071, "step": 59590 }, { "grad_norm": 0.08930337429046631, "learning_rate": 1.2211277749635264e-08, "loss": 0.007, "step": 59600 }, { "grad_norm": 0.1989632397890091, "learning_rate": 1.1609853773164592e-08, "loss": 0.0064, "step": 59610 }, { "grad_norm": 0.07478109747171402, "learning_rate": 1.1023614949751127e-08, "loss": 0.0069, "step": 59620 }, { "grad_norm": 0.09114154428243637, "learning_rate": 1.0452561457485744e-08, "loss": 0.0067, "step": 59630 }, { "grad_norm": 0.08945520222187042, "learning_rate": 9.896693469829688e-09, "loss": 0.0069, "step": 59640 }, { "grad_norm": 0.06247179955244064, "learning_rate": 9.35601115564788e-09, "loss": 0.0066, "step": 59650 }, { "grad_norm": 0.08339472115039825, "learning_rate": 8.830514679186719e-09, "loss": 0.0071, "step": 59660 }, { "grad_norm": 0.08225713670253754, "learning_rate": 8.320204200074066e-09, "loss": 0.0063, "step": 59670 }, { "grad_norm": 0.09439466893672943, "learning_rate": 7.825079873324814e-09, "loss": 0.0059, "step": 59680 }, { "grad_norm": 0.08109460771083832, "learning_rate": 7.345141849351977e-09, "loss": 0.0061, "step": 59690 }, { "grad_norm": 0.09043644368648529, "learning_rate": 6.880390273944493e-09, "loss": 0.006, "step": 59700 }, { "grad_norm": 0.07349655777215958, "learning_rate": 6.4308252882838704e-09, "loss": 0.0064, "step": 59710 }, { "grad_norm": 0.09995745867490768, "learning_rate": 5.9964470289386455e-09, "loss": 0.0067, "step": 59720 }, { "grad_norm": 0.07995537668466568, "learning_rate": 5.577255627853273e-09, "loss": 0.0058, "step": 59730 }, { "grad_norm": 0.13117307424545288, "learning_rate": 5.173251212370334e-09, "loss": 0.0069, "step": 59740 }, { "grad_norm": 0.07958384603261948, "learning_rate": 4.784433905219432e-09, "loss": 0.0063, "step": 59750 }, { "grad_norm": 0.06621520221233368, "learning_rate": 4.4108038245060934e-09, "loss": 0.006, "step": 59760 }, { "grad_norm": 0.10258115082979202, "learning_rate": 4.05236108373952e-09, "loss": 0.0072, "step": 59770 }, { "grad_norm": 0.1267351657152176, "learning_rate": 3.7091057917937324e-09, "loss": 0.0065, "step": 59780 }, { "grad_norm": 0.09282677620649338, "learning_rate": 3.381038052946428e-09, "loss": 0.007, "step": 59790 }, { "grad_norm": 0.05297244340181351, "learning_rate": 3.0681579668623283e-09, "loss": 0.006, "step": 59800 }, { "grad_norm": 0.09727036952972412, "learning_rate": 2.7704656285709727e-09, "loss": 0.0073, "step": 59810 }, { "grad_norm": 0.08810915052890778, "learning_rate": 2.4879611285166803e-09, "loss": 0.0064, "step": 59820 }, { "grad_norm": 0.07829294353723526, "learning_rate": 2.2206445525085883e-09, "loss": 0.007, "step": 59830 }, { "grad_norm": 0.0612417496740818, "learning_rate": 1.9685159817595112e-09, "loss": 0.0066, "step": 59840 }, { "grad_norm": 0.07441193610429764, "learning_rate": 1.7315754928470817e-09, "loss": 0.0058, "step": 59850 }, { "grad_norm": 0.0944516509771347, "learning_rate": 1.5098231577581611e-09, "loss": 0.0071, "step": 59860 }, { "grad_norm": 0.08045581728219986, "learning_rate": 1.3032590438499804e-09, "loss": 0.0072, "step": 59870 }, { "grad_norm": 0.1103404313325882, "learning_rate": 1.1118832138723444e-09, "loss": 0.0066, "step": 59880 }, { "grad_norm": 0.08065799623727798, "learning_rate": 9.356957259620825e-10, "loss": 0.0064, "step": 59890 }, { "grad_norm": 0.10253100842237473, "learning_rate": 7.74696633637495e-10, "loss": 0.0066, "step": 59900 }, { "grad_norm": 0.0952172577381134, "learning_rate": 6.288859858039064e-10, "loss": 0.0061, "step": 59910 }, { "grad_norm": 0.08060165494680405, "learning_rate": 4.982638267647665e-10, "loss": 0.0068, "step": 59920 }, { "grad_norm": 0.06330575793981552, "learning_rate": 3.8283019618834406e-10, "loss": 0.0063, "step": 59930 }, { "grad_norm": 0.10888925194740295, "learning_rate": 2.825851291410331e-10, "loss": 0.0064, "step": 59940 }, { "grad_norm": 0.0835295170545578, "learning_rate": 1.975286560873535e-10, "loss": 0.0069, "step": 59950 }, { "grad_norm": 0.06054665148258209, "learning_rate": 1.2766080285109284e-10, "loss": 0.0062, "step": 59960 }, { "grad_norm": 0.14083178341388702, "learning_rate": 7.298159065971533e-11, "loss": 0.0074, "step": 59970 }, { "grad_norm": 0.07209176570177078, "learning_rate": 3.349103612770854e-11, "loss": 0.0064, "step": 59980 }, { "grad_norm": 0.09501320123672485, "learning_rate": 9.189151245481142e-12, "loss": 0.007, "step": 59990 }, { "grad_norm": 0.09062499552965164, "learning_rate": 7.594340156735769e-14, "loss": 0.0068, "step": 60000 } ], "logging_steps": 10, "max_steps": 60000, "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": 64, "trial_name": null, "trial_params": null }