{ "best_global_step": null, "best_metric": null, "best_model_checkpoint": null, "epoch": 0.5, "eval_steps": 500, "global_step": 15000, "is_hyper_param_search": false, "is_local_process_zero": true, "is_world_process_zero": true, "log_history": [ { "grad_norm": 1.4286223649978638, "learning_rate": 6.000000000000001e-07, "loss": 1.1338, "step": 10 }, { "grad_norm": 1.3188368082046509, "learning_rate": 1.2666666666666667e-06, "loss": 1.1311, "step": 20 }, { "grad_norm": 1.0014344453811646, "learning_rate": 1.9333333333333336e-06, "loss": 1.1227, "step": 30 }, { "grad_norm": 0.5054961442947388, "learning_rate": 2.6e-06, "loss": 1.0995, "step": 40 }, { "grad_norm": 0.3656136989593506, "learning_rate": 3.2666666666666666e-06, "loss": 1.0826, "step": 50 }, { "grad_norm": 0.2424938678741455, "learning_rate": 3.9333333333333335e-06, "loss": 1.0659, "step": 60 }, { "grad_norm": 0.17607522010803223, "learning_rate": 4.6e-06, "loss": 1.0579, "step": 70 }, { "grad_norm": 0.22626137733459473, "learning_rate": 5.266666666666667e-06, "loss": 1.0616, "step": 80 }, { "grad_norm": 0.18578225374221802, "learning_rate": 5.933333333333334e-06, "loss": 1.0542, "step": 90 }, { "grad_norm": 0.22231262922286987, "learning_rate": 6.6e-06, "loss": 1.0443, "step": 100 }, { "grad_norm": 0.3583531975746155, "learning_rate": 7.266666666666668e-06, "loss": 1.0464, "step": 110 }, { "grad_norm": 0.28736764192581177, "learning_rate": 7.933333333333334e-06, "loss": 1.0383, "step": 120 }, { "grad_norm": 0.2196534126996994, "learning_rate": 8.599999999999999e-06, "loss": 1.0376, "step": 130 }, { "grad_norm": 0.26178935170173645, "learning_rate": 9.266666666666667e-06, "loss": 1.0213, "step": 140 }, { "grad_norm": 0.24020500481128693, "learning_rate": 9.933333333333334e-06, "loss": 1.0201, "step": 150 }, { "grad_norm": 0.2486283779144287, "learning_rate": 1.06e-05, "loss": 1.0153, "step": 160 }, { "grad_norm": 0.24497820436954498, "learning_rate": 1.1266666666666667e-05, "loss": 1.0176, "step": 170 }, { "grad_norm": 0.23539046943187714, "learning_rate": 1.1933333333333333e-05, "loss": 1.0224, "step": 180 }, { "grad_norm": 0.2657565176486969, "learning_rate": 1.2600000000000001e-05, "loss": 1.0164, "step": 190 }, { "grad_norm": 0.25632473826408386, "learning_rate": 1.3266666666666666e-05, "loss": 1.017, "step": 200 }, { "grad_norm": 0.26856011152267456, "learning_rate": 1.3933333333333334e-05, "loss": 1.0115, "step": 210 }, { "grad_norm": 0.26619046926498413, "learning_rate": 1.4599999999999999e-05, "loss": 1.0093, "step": 220 }, { "grad_norm": 0.8444914221763611, "learning_rate": 1.5266666666666667e-05, "loss": 0.9958, "step": 230 }, { "grad_norm": 0.3969593048095703, "learning_rate": 1.5933333333333332e-05, "loss": 0.9828, "step": 240 }, { "grad_norm": 0.6749173998832703, "learning_rate": 1.66e-05, "loss": 0.9603, "step": 250 }, { "grad_norm": 0.7760488986968994, "learning_rate": 1.726666666666667e-05, "loss": 0.909, "step": 260 }, { "grad_norm": 0.8203310966491699, "learning_rate": 1.7933333333333337e-05, "loss": 0.8739, "step": 270 }, { "grad_norm": 0.8459580540657043, "learning_rate": 1.86e-05, "loss": 0.8284, "step": 280 }, { "grad_norm": 0.773054301738739, "learning_rate": 1.926666666666667e-05, "loss": 0.7857, "step": 290 }, { "grad_norm": 0.9689459204673767, "learning_rate": 1.9933333333333334e-05, "loss": 0.7308, "step": 300 }, { "grad_norm": 1.0795235633850098, "learning_rate": 2.06e-05, "loss": 0.6831, "step": 310 }, { "grad_norm": 1.120764136314392, "learning_rate": 2.1266666666666667e-05, "loss": 0.6348, "step": 320 }, { "grad_norm": 1.217294692993164, "learning_rate": 2.1933333333333332e-05, "loss": 0.582, "step": 330 }, { "grad_norm": 1.1336292028427124, "learning_rate": 2.26e-05, "loss": 0.5147, "step": 340 }, { "grad_norm": 1.336414098739624, "learning_rate": 2.326666666666667e-05, "loss": 0.4546, "step": 350 }, { "grad_norm": 1.2505450248718262, "learning_rate": 2.3933333333333337e-05, "loss": 0.4079, "step": 360 }, { "grad_norm": 1.1749523878097534, "learning_rate": 2.46e-05, "loss": 0.3668, "step": 370 }, { "grad_norm": 1.255007028579712, "learning_rate": 2.5266666666666666e-05, "loss": 0.3165, "step": 380 }, { "grad_norm": 1.3134664297103882, "learning_rate": 2.5933333333333338e-05, "loss": 0.277, "step": 390 }, { "grad_norm": 1.2453639507293701, "learning_rate": 2.6600000000000003e-05, "loss": 0.2371, "step": 400 }, { "grad_norm": 1.6057149171829224, "learning_rate": 2.7266666666666668e-05, "loss": 0.2103, "step": 410 }, { "grad_norm": 1.4706779718399048, "learning_rate": 2.7933333333333332e-05, "loss": 0.1866, "step": 420 }, { "grad_norm": 1.2687115669250488, "learning_rate": 2.86e-05, "loss": 0.1825, "step": 430 }, { "grad_norm": 1.4025218486785889, "learning_rate": 2.926666666666667e-05, "loss": 0.1626, "step": 440 }, { "grad_norm": 1.6982293128967285, "learning_rate": 2.9933333333333337e-05, "loss": 0.1467, "step": 450 }, { "grad_norm": 1.3343334197998047, "learning_rate": 3.06e-05, "loss": 0.1289, "step": 460 }, { "grad_norm": 1.2774659395217896, "learning_rate": 3.126666666666666e-05, "loss": 0.1269, "step": 470 }, { "grad_norm": 1.0952668190002441, "learning_rate": 3.1933333333333335e-05, "loss": 0.1141, "step": 480 }, { "grad_norm": 1.2600839138031006, "learning_rate": 3.26e-05, "loss": 0.1065, "step": 490 }, { "grad_norm": 0.895696759223938, "learning_rate": 3.326666666666667e-05, "loss": 0.1014, "step": 500 }, { "grad_norm": 0.7864219546318054, "learning_rate": 3.3933333333333336e-05, "loss": 0.1048, "step": 510 }, { "grad_norm": 1.1830556392669678, "learning_rate": 3.46e-05, "loss": 0.098, "step": 520 }, { "grad_norm": 1.2950283288955688, "learning_rate": 3.526666666666667e-05, "loss": 0.1046, "step": 530 }, { "grad_norm": 0.8309084177017212, "learning_rate": 3.593333333333334e-05, "loss": 0.0901, "step": 540 }, { "grad_norm": 0.8505648970603943, "learning_rate": 3.66e-05, "loss": 0.0905, "step": 550 }, { "grad_norm": 0.8784123659133911, "learning_rate": 3.726666666666667e-05, "loss": 0.0859, "step": 560 }, { "grad_norm": 1.155542254447937, "learning_rate": 3.793333333333334e-05, "loss": 0.0878, "step": 570 }, { "grad_norm": 0.8971678614616394, "learning_rate": 3.86e-05, "loss": 0.0871, "step": 580 }, { "grad_norm": 1.2177577018737793, "learning_rate": 3.926666666666667e-05, "loss": 0.0886, "step": 590 }, { "grad_norm": 1.2923951148986816, "learning_rate": 3.993333333333333e-05, "loss": 0.0873, "step": 600 }, { "grad_norm": 1.00050687789917, "learning_rate": 4.0600000000000004e-05, "loss": 0.0982, "step": 610 }, { "grad_norm": 0.9716435670852661, "learning_rate": 4.126666666666667e-05, "loss": 0.0837, "step": 620 }, { "grad_norm": 0.8162665367126465, "learning_rate": 4.1933333333333334e-05, "loss": 0.0763, "step": 630 }, { "grad_norm": 0.9773716926574707, "learning_rate": 4.26e-05, "loss": 0.0753, "step": 640 }, { "grad_norm": 1.102065920829773, "learning_rate": 4.3266666666666664e-05, "loss": 0.0804, "step": 650 }, { "grad_norm": 0.9951865077018738, "learning_rate": 4.3933333333333335e-05, "loss": 0.0815, "step": 660 }, { "grad_norm": 0.9913735389709473, "learning_rate": 4.46e-05, "loss": 0.0922, "step": 670 }, { "grad_norm": 0.9612625241279602, "learning_rate": 4.526666666666667e-05, "loss": 0.0776, "step": 680 }, { "grad_norm": 1.025493860244751, "learning_rate": 4.5933333333333336e-05, "loss": 0.0744, "step": 690 }, { "grad_norm": 0.9520700573921204, "learning_rate": 4.660000000000001e-05, "loss": 0.074, "step": 700 }, { "grad_norm": 0.9577100872993469, "learning_rate": 4.726666666666667e-05, "loss": 0.0874, "step": 710 }, { "grad_norm": 0.7608726620674133, "learning_rate": 4.793333333333334e-05, "loss": 0.0748, "step": 720 }, { "grad_norm": 0.7298006415367126, "learning_rate": 4.86e-05, "loss": 0.0759, "step": 730 }, { "grad_norm": 0.8136162757873535, "learning_rate": 4.926666666666667e-05, "loss": 0.0647, "step": 740 }, { "grad_norm": 0.7462270259857178, "learning_rate": 4.993333333333334e-05, "loss": 0.0716, "step": 750 }, { "grad_norm": 0.7353017926216125, "learning_rate": 5.0600000000000003e-05, "loss": 0.0701, "step": 760 }, { "grad_norm": 0.7542075514793396, "learning_rate": 5.1266666666666675e-05, "loss": 0.0688, "step": 770 }, { "grad_norm": 0.976772665977478, "learning_rate": 5.193333333333333e-05, "loss": 0.0696, "step": 780 }, { "grad_norm": 0.7531357407569885, "learning_rate": 5.2600000000000005e-05, "loss": 0.0656, "step": 790 }, { "grad_norm": 0.9906631112098694, "learning_rate": 5.326666666666666e-05, "loss": 0.0739, "step": 800 }, { "grad_norm": 0.7879554033279419, "learning_rate": 5.3933333333333334e-05, "loss": 0.077, "step": 810 }, { "grad_norm": 0.7727330327033997, "learning_rate": 5.4600000000000006e-05, "loss": 0.0715, "step": 820 }, { "grad_norm": 0.6929222941398621, "learning_rate": 5.5266666666666664e-05, "loss": 0.0709, "step": 830 }, { "grad_norm": 0.9249580502510071, "learning_rate": 5.5933333333333335e-05, "loss": 0.065, "step": 840 }, { "grad_norm": 0.7341945171356201, "learning_rate": 5.66e-05, "loss": 0.0772, "step": 850 }, { "grad_norm": 0.9434566497802734, "learning_rate": 5.726666666666667e-05, "loss": 0.0681, "step": 860 }, { "grad_norm": 0.7272666692733765, "learning_rate": 5.7933333333333337e-05, "loss": 0.0692, "step": 870 }, { "grad_norm": 0.7985471487045288, "learning_rate": 5.86e-05, "loss": 0.0655, "step": 880 }, { "grad_norm": 0.6332736611366272, "learning_rate": 5.926666666666667e-05, "loss": 0.0611, "step": 890 }, { "grad_norm": 0.6844629645347595, "learning_rate": 5.9933333333333345e-05, "loss": 0.0703, "step": 900 }, { "grad_norm": 0.5473362803459167, "learning_rate": 6.06e-05, "loss": 0.0593, "step": 910 }, { "grad_norm": 0.6990862488746643, "learning_rate": 6.126666666666667e-05, "loss": 0.0622, "step": 920 }, { "grad_norm": 0.7508540749549866, "learning_rate": 6.193333333333333e-05, "loss": 0.0624, "step": 930 }, { "grad_norm": 0.46849414706230164, "learning_rate": 6.26e-05, "loss": 0.0607, "step": 940 }, { "grad_norm": 0.744961678981781, "learning_rate": 6.326666666666667e-05, "loss": 0.0673, "step": 950 }, { "grad_norm": 0.5163121819496155, "learning_rate": 6.393333333333333e-05, "loss": 0.0549, "step": 960 }, { "grad_norm": 0.6850712895393372, "learning_rate": 6.460000000000001e-05, "loss": 0.0594, "step": 970 }, { "grad_norm": 0.7848864197731018, "learning_rate": 6.526666666666666e-05, "loss": 0.0645, "step": 980 }, { "grad_norm": 0.5131463408470154, "learning_rate": 6.593333333333334e-05, "loss": 0.0636, "step": 990 }, { "grad_norm": 0.672545850276947, "learning_rate": 6.66e-05, "loss": 0.06, "step": 1000 }, { "grad_norm": 0.5691325068473816, "learning_rate": 6.726666666666667e-05, "loss": 0.0609, "step": 1010 }, { "grad_norm": 0.720917284488678, "learning_rate": 6.793333333333334e-05, "loss": 0.0579, "step": 1020 }, { "grad_norm": 0.6977890729904175, "learning_rate": 6.860000000000001e-05, "loss": 0.0611, "step": 1030 }, { "grad_norm": 0.4413103759288788, "learning_rate": 6.926666666666667e-05, "loss": 0.0595, "step": 1040 }, { "grad_norm": 0.5616114139556885, "learning_rate": 6.993333333333334e-05, "loss": 0.054, "step": 1050 }, { "grad_norm": 0.6126264333724976, "learning_rate": 7.06e-05, "loss": 0.0656, "step": 1060 }, { "grad_norm": 0.5445641875267029, "learning_rate": 7.126666666666667e-05, "loss": 0.0637, "step": 1070 }, { "grad_norm": 0.7580947279930115, "learning_rate": 7.193333333333334e-05, "loss": 0.0654, "step": 1080 }, { "grad_norm": 0.6648939847946167, "learning_rate": 7.26e-05, "loss": 0.0532, "step": 1090 }, { "grad_norm": 0.7424170970916748, "learning_rate": 7.326666666666667e-05, "loss": 0.0568, "step": 1100 }, { "grad_norm": 0.4409730136394501, "learning_rate": 7.393333333333333e-05, "loss": 0.0584, "step": 1110 }, { "grad_norm": 0.5734094381332397, "learning_rate": 7.46e-05, "loss": 0.0568, "step": 1120 }, { "grad_norm": 0.7434169054031372, "learning_rate": 7.526666666666668e-05, "loss": 0.0614, "step": 1130 }, { "grad_norm": 0.6002821922302246, "learning_rate": 7.593333333333334e-05, "loss": 0.0605, "step": 1140 }, { "grad_norm": 0.43884825706481934, "learning_rate": 7.66e-05, "loss": 0.0579, "step": 1150 }, { "grad_norm": 0.6982753276824951, "learning_rate": 7.726666666666667e-05, "loss": 0.0545, "step": 1160 }, { "grad_norm": 0.571456253528595, "learning_rate": 7.793333333333333e-05, "loss": 0.0551, "step": 1170 }, { "grad_norm": 0.4162715673446655, "learning_rate": 7.860000000000001e-05, "loss": 0.0542, "step": 1180 }, { "grad_norm": 0.4186014235019684, "learning_rate": 7.926666666666666e-05, "loss": 0.0534, "step": 1190 }, { "grad_norm": 0.6683947443962097, "learning_rate": 7.993333333333334e-05, "loss": 0.06, "step": 1200 }, { "grad_norm": 0.6045103669166565, "learning_rate": 8.060000000000001e-05, "loss": 0.0602, "step": 1210 }, { "grad_norm": 0.4470864236354828, "learning_rate": 8.126666666666667e-05, "loss": 0.0524, "step": 1220 }, { "grad_norm": 0.5444450974464417, "learning_rate": 8.193333333333334e-05, "loss": 0.0566, "step": 1230 }, { "grad_norm": 0.4147312641143799, "learning_rate": 8.26e-05, "loss": 0.053, "step": 1240 }, { "grad_norm": 0.669206440448761, "learning_rate": 8.326666666666667e-05, "loss": 0.0557, "step": 1250 }, { "grad_norm": 0.5752815008163452, "learning_rate": 8.393333333333335e-05, "loss": 0.0523, "step": 1260 }, { "grad_norm": 0.5641756057739258, "learning_rate": 8.46e-05, "loss": 0.0619, "step": 1270 }, { "grad_norm": 0.5776059627532959, "learning_rate": 8.526666666666667e-05, "loss": 0.0586, "step": 1280 }, { "grad_norm": 0.4552633464336395, "learning_rate": 8.593333333333333e-05, "loss": 0.0587, "step": 1290 }, { "grad_norm": 0.5360649228096008, "learning_rate": 8.66e-05, "loss": 0.0546, "step": 1300 }, { "grad_norm": 0.49262380599975586, "learning_rate": 8.726666666666667e-05, "loss": 0.0515, "step": 1310 }, { "grad_norm": 0.5333666801452637, "learning_rate": 8.793333333333333e-05, "loss": 0.0594, "step": 1320 }, { "grad_norm": 0.597510814666748, "learning_rate": 8.86e-05, "loss": 0.0506, "step": 1330 }, { "grad_norm": 0.5688159465789795, "learning_rate": 8.926666666666668e-05, "loss": 0.055, "step": 1340 }, { "grad_norm": 0.5717595219612122, "learning_rate": 8.993333333333334e-05, "loss": 0.0466, "step": 1350 }, { "grad_norm": 0.5456644892692566, "learning_rate": 9.06e-05, "loss": 0.0508, "step": 1360 }, { "grad_norm": 0.49075713753700256, "learning_rate": 9.126666666666667e-05, "loss": 0.0461, "step": 1370 }, { "grad_norm": 0.5612489581108093, "learning_rate": 9.193333333333334e-05, "loss": 0.051, "step": 1380 }, { "grad_norm": 0.5526825189590454, "learning_rate": 9.260000000000001e-05, "loss": 0.0474, "step": 1390 }, { "grad_norm": 0.530838131904602, "learning_rate": 9.326666666666667e-05, "loss": 0.0541, "step": 1400 }, { "grad_norm": 0.468610018491745, "learning_rate": 9.393333333333334e-05, "loss": 0.0474, "step": 1410 }, { "grad_norm": 0.5206702947616577, "learning_rate": 9.46e-05, "loss": 0.0568, "step": 1420 }, { "grad_norm": 0.491825133562088, "learning_rate": 9.526666666666667e-05, "loss": 0.0546, "step": 1430 }, { "grad_norm": 0.47865086793899536, "learning_rate": 9.593333333333334e-05, "loss": 0.0517, "step": 1440 }, { "grad_norm": 0.5672942399978638, "learning_rate": 9.66e-05, "loss": 0.0588, "step": 1450 }, { "grad_norm": 0.3895394206047058, "learning_rate": 9.726666666666667e-05, "loss": 0.0527, "step": 1460 }, { "grad_norm": 0.5083832144737244, "learning_rate": 9.793333333333333e-05, "loss": 0.054, "step": 1470 }, { "grad_norm": 0.4451282322406769, "learning_rate": 9.86e-05, "loss": 0.0584, "step": 1480 }, { "grad_norm": 0.5525972843170166, "learning_rate": 9.926666666666668e-05, "loss": 0.0509, "step": 1490 }, { "grad_norm": 0.5539897084236145, "learning_rate": 9.993333333333334e-05, "loss": 0.0642, "step": 1500 }, { "grad_norm": 0.36530470848083496, "learning_rate": 9.999997539434007e-05, "loss": 0.0554, "step": 1510 }, { "grad_norm": 0.47777435183525085, "learning_rate": 9.999989033776898e-05, "loss": 0.0518, "step": 1520 }, { "grad_norm": 0.34993427991867065, "learning_rate": 9.999974452661641e-05, "loss": 0.0513, "step": 1530 }, { "grad_norm": 0.6030260324478149, "learning_rate": 9.999953796105959e-05, "loss": 0.0567, "step": 1540 }, { "grad_norm": 0.42229360342025757, "learning_rate": 9.999927064134949e-05, "loss": 0.0534, "step": 1550 }, { "grad_norm": 0.5018796324729919, "learning_rate": 9.999894256781095e-05, "loss": 0.0493, "step": 1560 }, { "grad_norm": 0.4494084417819977, "learning_rate": 9.99985537408426e-05, "loss": 0.0454, "step": 1570 }, { "grad_norm": 0.38180992007255554, "learning_rate": 9.999810416091688e-05, "loss": 0.0459, "step": 1580 }, { "grad_norm": 0.31819790601730347, "learning_rate": 9.99975938285801e-05, "loss": 0.0514, "step": 1590 }, { "grad_norm": 0.4726041555404663, "learning_rate": 9.999702274445236e-05, "loss": 0.0486, "step": 1600 }, { "grad_norm": 0.5705097913742065, "learning_rate": 9.999639090922756e-05, "loss": 0.0458, "step": 1610 }, { "grad_norm": 0.49578216671943665, "learning_rate": 9.999569832367346e-05, "loss": 0.0421, "step": 1620 }, { "grad_norm": 0.47529393434524536, "learning_rate": 9.999494498863162e-05, "loss": 0.0457, "step": 1630 }, { "grad_norm": 0.4710843861103058, "learning_rate": 9.99941309050174e-05, "loss": 0.0474, "step": 1640 }, { "grad_norm": 0.4547995626926422, "learning_rate": 9.999325607381999e-05, "loss": 0.0518, "step": 1650 }, { "grad_norm": 0.42555010318756104, "learning_rate": 9.999232049610238e-05, "loss": 0.0406, "step": 1660 }, { "grad_norm": 0.42643633484840393, "learning_rate": 9.999132417300142e-05, "loss": 0.0487, "step": 1670 }, { "grad_norm": 0.36292701959609985, "learning_rate": 9.99902671057277e-05, "loss": 0.0491, "step": 1680 }, { "grad_norm": 0.4643704891204834, "learning_rate": 9.998914929556569e-05, "loss": 0.0539, "step": 1690 }, { "grad_norm": 0.5451205372810364, "learning_rate": 9.998797074387361e-05, "loss": 0.0457, "step": 1700 }, { "grad_norm": 0.4272986948490143, "learning_rate": 9.99867314520835e-05, "loss": 0.0468, "step": 1710 }, { "grad_norm": 0.510099470615387, "learning_rate": 9.998543142170126e-05, "loss": 0.0521, "step": 1720 }, { "grad_norm": 0.42392948269844055, "learning_rate": 9.99840706543065e-05, "loss": 0.0484, "step": 1730 }, { "grad_norm": 0.44271978735923767, "learning_rate": 9.998264915155274e-05, "loss": 0.0437, "step": 1740 }, { "grad_norm": 0.391396701335907, "learning_rate": 9.998116691516718e-05, "loss": 0.0471, "step": 1750 }, { "grad_norm": 0.45806899666786194, "learning_rate": 9.997962394695091e-05, "loss": 0.0496, "step": 1760 }, { "grad_norm": 0.43387362360954285, "learning_rate": 9.997802024877875e-05, "loss": 0.0472, "step": 1770 }, { "grad_norm": 0.39978858828544617, "learning_rate": 9.99763558225994e-05, "loss": 0.0485, "step": 1780 }, { "grad_norm": 0.48166730999946594, "learning_rate": 9.997463067043526e-05, "loss": 0.0542, "step": 1790 }, { "grad_norm": 0.46206942200660706, "learning_rate": 9.997284479438253e-05, "loss": 0.0483, "step": 1800 }, { "grad_norm": 0.34230849146842957, "learning_rate": 9.997099819661127e-05, "loss": 0.0464, "step": 1810 }, { "grad_norm": 0.3898741900920868, "learning_rate": 9.996909087936524e-05, "loss": 0.0436, "step": 1820 }, { "grad_norm": 0.45722997188568115, "learning_rate": 9.996712284496201e-05, "loss": 0.0415, "step": 1830 }, { "grad_norm": 0.35175952315330505, "learning_rate": 9.996509409579293e-05, "loss": 0.0457, "step": 1840 }, { "grad_norm": 0.36876893043518066, "learning_rate": 9.996300463432312e-05, "loss": 0.0416, "step": 1850 }, { "grad_norm": 0.41966450214385986, "learning_rate": 9.996085446309148e-05, "loss": 0.0474, "step": 1860 }, { "grad_norm": 0.29449400305747986, "learning_rate": 9.995864358471066e-05, "loss": 0.0491, "step": 1870 }, { "grad_norm": 0.3837015628814697, "learning_rate": 9.99563720018671e-05, "loss": 0.041, "step": 1880 }, { "grad_norm": 0.3597424626350403, "learning_rate": 9.995403971732098e-05, "loss": 0.0435, "step": 1890 }, { "grad_norm": 0.4648176431655884, "learning_rate": 9.995164673390625e-05, "loss": 0.0541, "step": 1900 }, { "grad_norm": 0.3706319034099579, "learning_rate": 9.994919305453059e-05, "loss": 0.0493, "step": 1910 }, { "grad_norm": 0.3526769280433655, "learning_rate": 9.994667868217548e-05, "loss": 0.0421, "step": 1920 }, { "grad_norm": 0.3949790894985199, "learning_rate": 9.99441036198961e-05, "loss": 0.0495, "step": 1930 }, { "grad_norm": 0.48242396116256714, "learning_rate": 9.99414678708214e-05, "loss": 0.0439, "step": 1940 }, { "grad_norm": 0.2822151482105255, "learning_rate": 9.993877143815407e-05, "loss": 0.0445, "step": 1950 }, { "grad_norm": 0.34638455510139465, "learning_rate": 9.993601432517053e-05, "loss": 0.0387, "step": 1960 }, { "grad_norm": 0.3569153845310211, "learning_rate": 9.993319653522091e-05, "loss": 0.0448, "step": 1970 }, { "grad_norm": 0.5560488700866699, "learning_rate": 9.993031807172911e-05, "loss": 0.0541, "step": 1980 }, { "grad_norm": 0.3920804262161255, "learning_rate": 9.992737893819273e-05, "loss": 0.0413, "step": 1990 }, { "grad_norm": 0.36530837416648865, "learning_rate": 9.992437913818312e-05, "loss": 0.0379, "step": 2000 }, { "grad_norm": 0.43708130717277527, "learning_rate": 9.992131867534526e-05, "loss": 0.0373, "step": 2010 }, { "grad_norm": 0.3348710536956787, "learning_rate": 9.991819755339796e-05, "loss": 0.0426, "step": 2020 }, { "grad_norm": 0.4954042136669159, "learning_rate": 9.991501577613365e-05, "loss": 0.0473, "step": 2030 }, { "grad_norm": 0.3213751018047333, "learning_rate": 9.99117733474185e-05, "loss": 0.0466, "step": 2040 }, { "grad_norm": 0.3377479314804077, "learning_rate": 9.990847027119234e-05, "loss": 0.0411, "step": 2050 }, { "grad_norm": 0.24278947710990906, "learning_rate": 9.990510655146877e-05, "loss": 0.0393, "step": 2060 }, { "grad_norm": 0.3590817153453827, "learning_rate": 9.990168219233496e-05, "loss": 0.0387, "step": 2070 }, { "grad_norm": 0.4050121605396271, "learning_rate": 9.989819719795188e-05, "loss": 0.0365, "step": 2080 }, { "grad_norm": 0.4315040111541748, "learning_rate": 9.989465157255412e-05, "loss": 0.0399, "step": 2090 }, { "grad_norm": 0.3367268741130829, "learning_rate": 9.989104532044994e-05, "loss": 0.038, "step": 2100 }, { "grad_norm": 0.26074960827827454, "learning_rate": 9.988737844602128e-05, "loss": 0.0407, "step": 2110 }, { "grad_norm": 0.31182852387428284, "learning_rate": 9.988365095372372e-05, "loss": 0.0391, "step": 2120 }, { "grad_norm": 0.35750165581703186, "learning_rate": 9.987986284808654e-05, "loss": 0.0424, "step": 2130 }, { "grad_norm": 0.36125466227531433, "learning_rate": 9.987601413371264e-05, "loss": 0.0446, "step": 2140 }, { "grad_norm": 0.4450468420982361, "learning_rate": 9.987210481527855e-05, "loss": 0.0511, "step": 2150 }, { "grad_norm": 0.2671463191509247, "learning_rate": 9.98681348975345e-05, "loss": 0.0437, "step": 2160 }, { "grad_norm": 0.3591150939464569, "learning_rate": 9.986410438530427e-05, "loss": 0.0395, "step": 2170 }, { "grad_norm": 0.42698219418525696, "learning_rate": 9.986001328348534e-05, "loss": 0.0467, "step": 2180 }, { "grad_norm": 0.3475445806980133, "learning_rate": 9.985586159704878e-05, "loss": 0.0393, "step": 2190 }, { "grad_norm": 0.3523111343383789, "learning_rate": 9.985164933103929e-05, "loss": 0.0396, "step": 2200 }, { "grad_norm": 0.3586738407611847, "learning_rate": 9.984737649057513e-05, "loss": 0.0373, "step": 2210 }, { "grad_norm": 0.3778880834579468, "learning_rate": 9.984304308084827e-05, "loss": 0.0371, "step": 2220 }, { "grad_norm": 0.39608579874038696, "learning_rate": 9.983864910712416e-05, "loss": 0.0419, "step": 2230 }, { "grad_norm": 0.45594409108161926, "learning_rate": 9.98341945747419e-05, "loss": 0.0368, "step": 2240 }, { "grad_norm": 0.37032201886177063, "learning_rate": 9.98296794891142e-05, "loss": 0.0436, "step": 2250 }, { "grad_norm": 0.413915753364563, "learning_rate": 9.982510385572725e-05, "loss": 0.04, "step": 2260 }, { "grad_norm": 0.32911771535873413, "learning_rate": 9.982046768014094e-05, "loss": 0.038, "step": 2270 }, { "grad_norm": 0.32144561409950256, "learning_rate": 9.981577096798863e-05, "loss": 0.0439, "step": 2280 }, { "grad_norm": 0.28206518292427063, "learning_rate": 9.981101372497727e-05, "loss": 0.0401, "step": 2290 }, { "grad_norm": 0.3610568344593048, "learning_rate": 9.980619595688737e-05, "loss": 0.037, "step": 2300 }, { "grad_norm": 0.34978851675987244, "learning_rate": 9.980131766957295e-05, "loss": 0.0444, "step": 2310 }, { "grad_norm": 0.28932616114616394, "learning_rate": 9.979637886896163e-05, "loss": 0.0382, "step": 2320 }, { "grad_norm": 0.36608782410621643, "learning_rate": 9.979137956105447e-05, "loss": 0.0499, "step": 2330 }, { "grad_norm": 0.32772204279899597, "learning_rate": 9.978631975192613e-05, "loss": 0.0365, "step": 2340 }, { "grad_norm": 0.35400906205177307, "learning_rate": 9.978119944772475e-05, "loss": 0.0443, "step": 2350 }, { "grad_norm": 0.32196030020713806, "learning_rate": 9.977601865467197e-05, "loss": 0.0393, "step": 2360 }, { "grad_norm": 0.42715147137641907, "learning_rate": 9.977077737906297e-05, "loss": 0.0437, "step": 2370 }, { "grad_norm": 0.3944530487060547, "learning_rate": 9.976547562726636e-05, "loss": 0.0395, "step": 2380 }, { "grad_norm": 0.34627705812454224, "learning_rate": 9.976011340572429e-05, "loss": 0.0424, "step": 2390 }, { "grad_norm": 0.35061323642730713, "learning_rate": 9.975469072095237e-05, "loss": 0.0435, "step": 2400 }, { "grad_norm": 0.3443011939525604, "learning_rate": 9.974920757953965e-05, "loss": 0.0379, "step": 2410 }, { "grad_norm": 0.3471289277076721, "learning_rate": 9.97436639881487e-05, "loss": 0.0399, "step": 2420 }, { "grad_norm": 0.43773412704467773, "learning_rate": 9.973805995351545e-05, "loss": 0.0364, "step": 2430 }, { "grad_norm": 0.24965138733386993, "learning_rate": 9.973239548244939e-05, "loss": 0.0368, "step": 2440 }, { "grad_norm": 0.2570035457611084, "learning_rate": 9.972667058183333e-05, "loss": 0.0342, "step": 2450 }, { "grad_norm": 0.3479859530925751, "learning_rate": 9.972088525862362e-05, "loss": 0.0403, "step": 2460 }, { "grad_norm": 0.262636661529541, "learning_rate": 9.971503951984995e-05, "loss": 0.0352, "step": 2470 }, { "grad_norm": 0.3546050190925598, "learning_rate": 9.970913337261543e-05, "loss": 0.033, "step": 2480 }, { "grad_norm": 0.36946389079093933, "learning_rate": 9.97031668240966e-05, "loss": 0.0314, "step": 2490 }, { "grad_norm": 0.35801342129707336, "learning_rate": 9.969713988154339e-05, "loss": 0.0406, "step": 2500 }, { "grad_norm": 0.3424617350101471, "learning_rate": 9.969105255227906e-05, "loss": 0.0358, "step": 2510 }, { "grad_norm": 0.3694426119327545, "learning_rate": 9.968490484370035e-05, "loss": 0.0356, "step": 2520 }, { "grad_norm": 0.29051831364631653, "learning_rate": 9.967869676327726e-05, "loss": 0.0334, "step": 2530 }, { "grad_norm": 0.33100759983062744, "learning_rate": 9.96724283185532e-05, "loss": 0.0364, "step": 2540 }, { "grad_norm": 0.305808961391449, "learning_rate": 9.966609951714494e-05, "loss": 0.0408, "step": 2550 }, { "grad_norm": 0.37323492765426636, "learning_rate": 9.965971036674255e-05, "loss": 0.0354, "step": 2560 }, { "grad_norm": 0.26690348982810974, "learning_rate": 9.965326087510947e-05, "loss": 0.0334, "step": 2570 }, { "grad_norm": 0.2663111090660095, "learning_rate": 9.964675105008244e-05, "loss": 0.0382, "step": 2580 }, { "grad_norm": 0.3225189745426178, "learning_rate": 9.964018089957147e-05, "loss": 0.0362, "step": 2590 }, { "grad_norm": 0.21992123126983643, "learning_rate": 9.963355043155996e-05, "loss": 0.0357, "step": 2600 }, { "grad_norm": 0.3507784605026245, "learning_rate": 9.962685965410455e-05, "loss": 0.0349, "step": 2610 }, { "grad_norm": 0.32163360714912415, "learning_rate": 9.962010857533514e-05, "loss": 0.0367, "step": 2620 }, { "grad_norm": 0.32095786929130554, "learning_rate": 9.961329720345493e-05, "loss": 0.0334, "step": 2630 }, { "grad_norm": 0.31472834944725037, "learning_rate": 9.96064255467404e-05, "loss": 0.0339, "step": 2640 }, { "grad_norm": 0.392196387052536, "learning_rate": 9.959949361354126e-05, "loss": 0.0392, "step": 2650 }, { "grad_norm": 0.3518916964530945, "learning_rate": 9.959250141228045e-05, "loss": 0.045, "step": 2660 }, { "grad_norm": 0.29921600222587585, "learning_rate": 9.958544895145414e-05, "loss": 0.038, "step": 2670 }, { "grad_norm": 0.2692553699016571, "learning_rate": 9.957833623963177e-05, "loss": 0.0391, "step": 2680 }, { "grad_norm": 0.2813042104244232, "learning_rate": 9.957116328545593e-05, "loss": 0.0353, "step": 2690 }, { "grad_norm": 0.27075234055519104, "learning_rate": 9.956393009764244e-05, "loss": 0.039, "step": 2700 }, { "grad_norm": 0.2775849401950836, "learning_rate": 9.955663668498032e-05, "loss": 0.0395, "step": 2710 }, { "grad_norm": 0.3429386615753174, "learning_rate": 9.954928305633173e-05, "loss": 0.0382, "step": 2720 }, { "grad_norm": 0.36746281385421753, "learning_rate": 9.954186922063204e-05, "loss": 0.0379, "step": 2730 }, { "grad_norm": 0.2106514871120453, "learning_rate": 9.953439518688974e-05, "loss": 0.0375, "step": 2740 }, { "grad_norm": 0.2520779073238373, "learning_rate": 9.952686096418652e-05, "loss": 0.0332, "step": 2750 }, { "grad_norm": 0.28839102387428284, "learning_rate": 9.951926656167714e-05, "loss": 0.0346, "step": 2760 }, { "grad_norm": 0.28207844495773315, "learning_rate": 9.951161198858953e-05, "loss": 0.0343, "step": 2770 }, { "grad_norm": 0.25479984283447266, "learning_rate": 9.950389725422471e-05, "loss": 0.0359, "step": 2780 }, { "grad_norm": 0.2918780446052551, "learning_rate": 9.949612236795682e-05, "loss": 0.0328, "step": 2790 }, { "grad_norm": 0.3814399540424347, "learning_rate": 9.948828733923305e-05, "loss": 0.0352, "step": 2800 }, { "grad_norm": 0.3272506296634674, "learning_rate": 9.948039217757374e-05, "loss": 0.0368, "step": 2810 }, { "grad_norm": 0.2723280191421509, "learning_rate": 9.947243689257225e-05, "loss": 0.0343, "step": 2820 }, { "grad_norm": 0.3470291197299957, "learning_rate": 9.946442149389497e-05, "loss": 0.0355, "step": 2830 }, { "grad_norm": 0.3108155131340027, "learning_rate": 9.945634599128139e-05, "loss": 0.039, "step": 2840 }, { "grad_norm": 0.2814142405986786, "learning_rate": 9.944821039454402e-05, "loss": 0.0369, "step": 2850 }, { "grad_norm": 0.27483293414115906, "learning_rate": 9.944001471356835e-05, "loss": 0.0362, "step": 2860 }, { "grad_norm": 0.2819496691226959, "learning_rate": 9.94317589583129e-05, "loss": 0.0341, "step": 2870 }, { "grad_norm": 0.35509470105171204, "learning_rate": 9.942344313880922e-05, "loss": 0.0332, "step": 2880 }, { "grad_norm": 0.3057287037372589, "learning_rate": 9.941506726516179e-05, "loss": 0.033, "step": 2890 }, { "grad_norm": 0.30556419491767883, "learning_rate": 9.94066313475481e-05, "loss": 0.036, "step": 2900 }, { "grad_norm": 0.2953285574913025, "learning_rate": 9.939813539621857e-05, "loss": 0.045, "step": 2910 }, { "grad_norm": 0.3086806535720825, "learning_rate": 9.93895794214966e-05, "loss": 0.032, "step": 2920 }, { "grad_norm": 0.2891356945037842, "learning_rate": 9.938096343377852e-05, "loss": 0.0348, "step": 2930 }, { "grad_norm": 0.23821333050727844, "learning_rate": 9.937228744353353e-05, "loss": 0.035, "step": 2940 }, { "grad_norm": 0.35334500670433044, "learning_rate": 9.936355146130379e-05, "loss": 0.0347, "step": 2950 }, { "grad_norm": 0.36387619376182556, "learning_rate": 9.935475549770435e-05, "loss": 0.0427, "step": 2960 }, { "grad_norm": 0.3732975125312805, "learning_rate": 9.934589956342315e-05, "loss": 0.041, "step": 2970 }, { "grad_norm": 0.27346840500831604, "learning_rate": 9.933698366922093e-05, "loss": 0.0346, "step": 2980 }, { "grad_norm": 0.272949755191803, "learning_rate": 9.93280078259314e-05, "loss": 0.0337, "step": 2990 }, { "grad_norm": 0.2521970570087433, "learning_rate": 9.931897204446104e-05, "loss": 0.038, "step": 3000 }, { "grad_norm": 0.2809789478778839, "learning_rate": 9.930987633578915e-05, "loss": 0.0364, "step": 3010 }, { "grad_norm": 0.2556251287460327, "learning_rate": 9.93007207109679e-05, "loss": 0.0395, "step": 3020 }, { "grad_norm": 0.30563926696777344, "learning_rate": 9.929150518112224e-05, "loss": 0.0333, "step": 3030 }, { "grad_norm": 0.33915895223617554, "learning_rate": 9.928222975744991e-05, "loss": 0.0326, "step": 3040 }, { "grad_norm": 0.33186429738998413, "learning_rate": 9.92728944512214e-05, "loss": 0.0369, "step": 3050 }, { "grad_norm": 0.2459770143032074, "learning_rate": 9.926349927378001e-05, "loss": 0.0374, "step": 3060 }, { "grad_norm": 0.29682546854019165, "learning_rate": 9.925404423654174e-05, "loss": 0.0388, "step": 3070 }, { "grad_norm": 0.32139870524406433, "learning_rate": 9.924452935099537e-05, "loss": 0.0367, "step": 3080 }, { "grad_norm": 0.21356557309627533, "learning_rate": 9.92349546287024e-05, "loss": 0.0357, "step": 3090 }, { "grad_norm": 0.2548346519470215, "learning_rate": 9.9225320081297e-05, "loss": 0.0359, "step": 3100 }, { "grad_norm": 0.2944643199443817, "learning_rate": 9.921562572048606e-05, "loss": 0.0394, "step": 3110 }, { "grad_norm": 0.36908605694770813, "learning_rate": 9.920587155804913e-05, "loss": 0.0336, "step": 3120 }, { "grad_norm": 0.38233789801597595, "learning_rate": 9.919605760583845e-05, "loss": 0.0409, "step": 3130 }, { "grad_norm": 0.2826779782772064, "learning_rate": 9.91861838757789e-05, "loss": 0.0381, "step": 3140 }, { "grad_norm": 0.2811231315135956, "learning_rate": 9.917625037986798e-05, "loss": 0.0369, "step": 3150 }, { "grad_norm": 0.32433658838272095, "learning_rate": 9.916625713017583e-05, "loss": 0.0333, "step": 3160 }, { "grad_norm": 0.31853899359703064, "learning_rate": 9.915620413884519e-05, "loss": 0.0363, "step": 3170 }, { "grad_norm": 0.2896600365638733, "learning_rate": 9.914609141809139e-05, "loss": 0.0332, "step": 3180 }, { "grad_norm": 0.23982714116573334, "learning_rate": 9.913591898020235e-05, "loss": 0.0368, "step": 3190 }, { "grad_norm": 0.2892548739910126, "learning_rate": 9.912568683753853e-05, "loss": 0.038, "step": 3200 }, { "grad_norm": 0.2549099326133728, "learning_rate": 9.911539500253295e-05, "loss": 0.0382, "step": 3210 }, { "grad_norm": 0.26167428493499756, "learning_rate": 9.910504348769118e-05, "loss": 0.0339, "step": 3220 }, { "grad_norm": 0.2887308895587921, "learning_rate": 9.909463230559127e-05, "loss": 0.0356, "step": 3230 }, { "grad_norm": 0.2307373583316803, "learning_rate": 9.908416146888376e-05, "loss": 0.0313, "step": 3240 }, { "grad_norm": 0.4429417550563812, "learning_rate": 9.907363099029175e-05, "loss": 0.0354, "step": 3250 }, { "grad_norm": 0.27956104278564453, "learning_rate": 9.906304088261073e-05, "loss": 0.0409, "step": 3260 }, { "grad_norm": 0.2723853588104248, "learning_rate": 9.905239115870872e-05, "loss": 0.0379, "step": 3270 }, { "grad_norm": 0.22352485358715057, "learning_rate": 9.90416818315261e-05, "loss": 0.0326, "step": 3280 }, { "grad_norm": 0.25841546058654785, "learning_rate": 9.903091291407573e-05, "loss": 0.0349, "step": 3290 }, { "grad_norm": 0.2908726930618286, "learning_rate": 9.902008441944286e-05, "loss": 0.0337, "step": 3300 }, { "grad_norm": 0.24134330451488495, "learning_rate": 9.900919636078512e-05, "loss": 0.0336, "step": 3310 }, { "grad_norm": 0.22915515303611755, "learning_rate": 9.899824875133255e-05, "loss": 0.0335, "step": 3320 }, { "grad_norm": 0.2795277535915375, "learning_rate": 9.898724160438749e-05, "loss": 0.035, "step": 3330 }, { "grad_norm": 0.315969854593277, "learning_rate": 9.89761749333247e-05, "loss": 0.033, "step": 3340 }, { "grad_norm": 0.32318344712257385, "learning_rate": 9.896504875159122e-05, "loss": 0.0394, "step": 3350 }, { "grad_norm": 0.3505741059780121, "learning_rate": 9.89538630727064e-05, "loss": 0.0344, "step": 3360 }, { "grad_norm": 0.24356842041015625, "learning_rate": 9.894261791026189e-05, "loss": 0.0448, "step": 3370 }, { "grad_norm": 0.2837693691253662, "learning_rate": 9.893131327792165e-05, "loss": 0.0355, "step": 3380 }, { "grad_norm": 0.2266145944595337, "learning_rate": 9.891994918942182e-05, "loss": 0.0378, "step": 3390 }, { "grad_norm": 0.2038327008485794, "learning_rate": 9.890852565857092e-05, "loss": 0.04, "step": 3400 }, { "grad_norm": 0.2771011292934418, "learning_rate": 9.889704269924954e-05, "loss": 0.0352, "step": 3410 }, { "grad_norm": 0.26260632276535034, "learning_rate": 9.888550032541059e-05, "loss": 0.0337, "step": 3420 }, { "grad_norm": 0.245091512799263, "learning_rate": 9.887389855107916e-05, "loss": 0.0309, "step": 3430 }, { "grad_norm": 0.28305497765541077, "learning_rate": 9.886223739035248e-05, "loss": 0.0338, "step": 3440 }, { "grad_norm": 0.28816455602645874, "learning_rate": 9.885051685739997e-05, "loss": 0.0313, "step": 3450 }, { "grad_norm": 0.2417427897453308, "learning_rate": 9.883873696646316e-05, "loss": 0.0364, "step": 3460 }, { "grad_norm": 0.30100715160369873, "learning_rate": 9.882689773185575e-05, "loss": 0.0318, "step": 3470 }, { "grad_norm": 0.25989899039268494, "learning_rate": 9.881499916796353e-05, "loss": 0.0366, "step": 3480 }, { "grad_norm": 0.2707766592502594, "learning_rate": 9.880304128924434e-05, "loss": 0.0345, "step": 3490 }, { "grad_norm": 0.2442186027765274, "learning_rate": 9.879102411022817e-05, "loss": 0.0392, "step": 3500 }, { "grad_norm": 0.26353660225868225, "learning_rate": 9.877894764551703e-05, "loss": 0.0343, "step": 3510 }, { "grad_norm": 0.3113507926464081, "learning_rate": 9.876681190978494e-05, "loss": 0.0323, "step": 3520 }, { "grad_norm": 0.21240299940109253, "learning_rate": 9.875461691777797e-05, "loss": 0.0339, "step": 3530 }, { "grad_norm": 0.3515579104423523, "learning_rate": 9.874236268431417e-05, "loss": 0.0386, "step": 3540 }, { "grad_norm": 0.23459798097610474, "learning_rate": 9.873004922428361e-05, "loss": 0.0323, "step": 3550 }, { "grad_norm": 0.23396295309066772, "learning_rate": 9.871767655264829e-05, "loss": 0.0329, "step": 3560 }, { "grad_norm": 0.339568167924881, "learning_rate": 9.87052446844422e-05, "loss": 0.0357, "step": 3570 }, { "grad_norm": 0.30799129605293274, "learning_rate": 9.869275363477122e-05, "loss": 0.0343, "step": 3580 }, { "grad_norm": 0.25435152649879456, "learning_rate": 9.868020341881312e-05, "loss": 0.0308, "step": 3590 }, { "grad_norm": 0.2983786165714264, "learning_rate": 9.866759405181765e-05, "loss": 0.033, "step": 3600 }, { "grad_norm": 0.2190297394990921, "learning_rate": 9.865492554910633e-05, "loss": 0.0308, "step": 3610 }, { "grad_norm": 0.20390067994594574, "learning_rate": 9.864219792607262e-05, "loss": 0.0327, "step": 3620 }, { "grad_norm": 0.22389930486679077, "learning_rate": 9.862941119818177e-05, "loss": 0.0331, "step": 3630 }, { "grad_norm": 0.229415625333786, "learning_rate": 9.861656538097086e-05, "loss": 0.0334, "step": 3640 }, { "grad_norm": 0.2973964214324951, "learning_rate": 9.860366049004877e-05, "loss": 0.0352, "step": 3650 }, { "grad_norm": 0.31495925784111023, "learning_rate": 9.859069654109615e-05, "loss": 0.0308, "step": 3660 }, { "grad_norm": 0.27302220463752747, "learning_rate": 9.857767354986545e-05, "loss": 0.0342, "step": 3670 }, { "grad_norm": 0.3227456510066986, "learning_rate": 9.856459153218078e-05, "loss": 0.0358, "step": 3680 }, { "grad_norm": 0.26731786131858826, "learning_rate": 9.855145050393808e-05, "loss": 0.0377, "step": 3690 }, { "grad_norm": 0.29686787724494934, "learning_rate": 9.85382504811049e-05, "loss": 0.0308, "step": 3700 }, { "grad_norm": 0.21109698712825775, "learning_rate": 9.852499147972054e-05, "loss": 0.0385, "step": 3710 }, { "grad_norm": 0.2166232317686081, "learning_rate": 9.851167351589592e-05, "loss": 0.036, "step": 3720 }, { "grad_norm": 0.19541573524475098, "learning_rate": 9.849829660581363e-05, "loss": 0.0297, "step": 3730 }, { "grad_norm": 0.3190360963344574, "learning_rate": 9.848486076572787e-05, "loss": 0.0373, "step": 3740 }, { "grad_norm": 0.17807696759700775, "learning_rate": 9.847136601196446e-05, "loss": 0.0355, "step": 3750 }, { "grad_norm": 0.21160131692886353, "learning_rate": 9.845781236092078e-05, "loss": 0.0305, "step": 3760 }, { "grad_norm": 0.24084891378879547, "learning_rate": 9.844419982906583e-05, "loss": 0.036, "step": 3770 }, { "grad_norm": 0.31141653656959534, "learning_rate": 9.843052843294008e-05, "loss": 0.0317, "step": 3780 }, { "grad_norm": 0.24087437987327576, "learning_rate": 9.841679818915559e-05, "loss": 0.0308, "step": 3790 }, { "grad_norm": 0.2501969039440155, "learning_rate": 9.840300911439591e-05, "loss": 0.0344, "step": 3800 }, { "grad_norm": 0.23600023984909058, "learning_rate": 9.838916122541603e-05, "loss": 0.0299, "step": 3810 }, { "grad_norm": 0.27542638778686523, "learning_rate": 9.837525453904246e-05, "loss": 0.0342, "step": 3820 }, { "grad_norm": 0.26895633339881897, "learning_rate": 9.836128907217314e-05, "loss": 0.0356, "step": 3830 }, { "grad_norm": 0.272287517786026, "learning_rate": 9.834726484177743e-05, "loss": 0.0296, "step": 3840 }, { "grad_norm": 0.2281694859266281, "learning_rate": 9.833318186489609e-05, "loss": 0.0306, "step": 3850 }, { "grad_norm": 0.2514426112174988, "learning_rate": 9.831904015864126e-05, "loss": 0.0346, "step": 3860 }, { "grad_norm": 0.28251543641090393, "learning_rate": 9.830483974019645e-05, "loss": 0.032, "step": 3870 }, { "grad_norm": 0.16133616864681244, "learning_rate": 9.82905806268165e-05, "loss": 0.0309, "step": 3880 }, { "grad_norm": 0.29323726892471313, "learning_rate": 9.82762628358276e-05, "loss": 0.0282, "step": 3890 }, { "grad_norm": 0.27166035771369934, "learning_rate": 9.826188638462718e-05, "loss": 0.0307, "step": 3900 }, { "grad_norm": 0.22072918713092804, "learning_rate": 9.824745129068402e-05, "loss": 0.0338, "step": 3910 }, { "grad_norm": 0.2643439769744873, "learning_rate": 9.82329575715381e-05, "loss": 0.0313, "step": 3920 }, { "grad_norm": 0.3166283369064331, "learning_rate": 9.821840524480066e-05, "loss": 0.0313, "step": 3930 }, { "grad_norm": 0.2126985788345337, "learning_rate": 9.820379432815414e-05, "loss": 0.0324, "step": 3940 }, { "grad_norm": 0.26648762822151184, "learning_rate": 9.81891248393522e-05, "loss": 0.0334, "step": 3950 }, { "grad_norm": 0.23768727481365204, "learning_rate": 9.817439679621963e-05, "loss": 0.0293, "step": 3960 }, { "grad_norm": 0.28346073627471924, "learning_rate": 9.815961021665243e-05, "loss": 0.0321, "step": 3970 }, { "grad_norm": 0.3498678505420685, "learning_rate": 9.814476511861763e-05, "loss": 0.0344, "step": 3980 }, { "grad_norm": 0.2795715034008026, "learning_rate": 9.812986152015348e-05, "loss": 0.0339, "step": 3990 }, { "grad_norm": 0.2788242995738983, "learning_rate": 9.811489943936922e-05, "loss": 0.0342, "step": 4000 }, { "grad_norm": 0.21484830975532532, "learning_rate": 9.809987889444522e-05, "loss": 0.032, "step": 4010 }, { "grad_norm": 0.2867797613143921, "learning_rate": 9.808479990363282e-05, "loss": 0.0297, "step": 4020 }, { "grad_norm": 0.26604515314102173, "learning_rate": 9.806966248525445e-05, "loss": 0.0341, "step": 4030 }, { "grad_norm": 0.15827511250972748, "learning_rate": 9.805446665770348e-05, "loss": 0.0316, "step": 4040 }, { "grad_norm": 0.26560184359550476, "learning_rate": 9.803921243944429e-05, "loss": 0.0303, "step": 4050 }, { "grad_norm": 0.28975242376327515, "learning_rate": 9.802389984901218e-05, "loss": 0.0362, "step": 4060 }, { "grad_norm": 0.2958698570728302, "learning_rate": 9.80085289050134e-05, "loss": 0.0337, "step": 4070 }, { "grad_norm": 0.2920568585395813, "learning_rate": 9.799309962612508e-05, "loss": 0.0299, "step": 4080 }, { "grad_norm": 0.22438588738441467, "learning_rate": 9.797761203109527e-05, "loss": 0.0336, "step": 4090 }, { "grad_norm": 0.3101698160171509, "learning_rate": 9.796206613874283e-05, "loss": 0.0375, "step": 4100 }, { "grad_norm": 0.2402053326368332, "learning_rate": 9.794646196795754e-05, "loss": 0.0333, "step": 4110 }, { "grad_norm": 0.25514283776283264, "learning_rate": 9.793079953769987e-05, "loss": 0.0345, "step": 4120 }, { "grad_norm": 0.2619798481464386, "learning_rate": 9.79150788670012e-05, "loss": 0.0316, "step": 4130 }, { "grad_norm": 0.2864750027656555, "learning_rate": 9.78992999749636e-05, "loss": 0.035, "step": 4140 }, { "grad_norm": 0.29329100251197815, "learning_rate": 9.788346288075994e-05, "loss": 0.0328, "step": 4150 }, { "grad_norm": 0.32842564582824707, "learning_rate": 9.786756760363373e-05, "loss": 0.0302, "step": 4160 }, { "grad_norm": 0.2717845141887665, "learning_rate": 9.78516141628993e-05, "loss": 0.0368, "step": 4170 }, { "grad_norm": 0.29956188797950745, "learning_rate": 9.783560257794154e-05, "loss": 0.0357, "step": 4180 }, { "grad_norm": 0.239421546459198, "learning_rate": 9.781953286821603e-05, "loss": 0.034, "step": 4190 }, { "grad_norm": 0.18612802028656006, "learning_rate": 9.780340505324901e-05, "loss": 0.029, "step": 4200 }, { "grad_norm": 0.14744599163532257, "learning_rate": 9.778721915263727e-05, "loss": 0.0267, "step": 4210 }, { "grad_norm": 0.22997815907001495, "learning_rate": 9.777097518604824e-05, "loss": 0.0297, "step": 4220 }, { "grad_norm": 0.35401955246925354, "learning_rate": 9.775467317321984e-05, "loss": 0.0281, "step": 4230 }, { "grad_norm": 0.277066707611084, "learning_rate": 9.773831313396055e-05, "loss": 0.0315, "step": 4240 }, { "grad_norm": 0.215666264295578, "learning_rate": 9.77218950881494e-05, "loss": 0.0332, "step": 4250 }, { "grad_norm": 0.27352502942085266, "learning_rate": 9.770541905573583e-05, "loss": 0.0325, "step": 4260 }, { "grad_norm": 0.2619900405406952, "learning_rate": 9.768888505673976e-05, "loss": 0.0294, "step": 4270 }, { "grad_norm": 0.19816714525222778, "learning_rate": 9.767229311125162e-05, "loss": 0.0306, "step": 4280 }, { "grad_norm": 0.20372048020362854, "learning_rate": 9.765564323943211e-05, "loss": 0.0277, "step": 4290 }, { "grad_norm": 0.19429892301559448, "learning_rate": 9.763893546151244e-05, "loss": 0.0291, "step": 4300 }, { "grad_norm": 0.18286654353141785, "learning_rate": 9.762216979779412e-05, "loss": 0.0262, "step": 4310 }, { "grad_norm": 0.27293694019317627, "learning_rate": 9.760534626864902e-05, "loss": 0.0285, "step": 4320 }, { "grad_norm": 0.20325680077075958, "learning_rate": 9.758846489451931e-05, "loss": 0.0286, "step": 4330 }, { "grad_norm": 0.2597692608833313, "learning_rate": 9.757152569591748e-05, "loss": 0.0293, "step": 4340 }, { "grad_norm": 0.24201206862926483, "learning_rate": 9.75545286934262e-05, "loss": 0.0292, "step": 4350 }, { "grad_norm": 0.2706036865711212, "learning_rate": 9.753747390769847e-05, "loss": 0.0298, "step": 4360 }, { "grad_norm": 0.26244068145751953, "learning_rate": 9.752036135945744e-05, "loss": 0.0276, "step": 4370 }, { "grad_norm": 0.25983375310897827, "learning_rate": 9.750319106949649e-05, "loss": 0.0273, "step": 4380 }, { "grad_norm": 0.22491343319416046, "learning_rate": 9.748596305867913e-05, "loss": 0.0322, "step": 4390 }, { "grad_norm": 0.29062047600746155, "learning_rate": 9.746867734793903e-05, "loss": 0.0343, "step": 4400 }, { "grad_norm": 0.19285595417022705, "learning_rate": 9.745133395827993e-05, "loss": 0.0304, "step": 4410 }, { "grad_norm": 0.22780440747737885, "learning_rate": 9.743393291077572e-05, "loss": 0.0355, "step": 4420 }, { "grad_norm": 0.24806730449199677, "learning_rate": 9.741647422657028e-05, "loss": 0.035, "step": 4430 }, { "grad_norm": 0.1997525990009308, "learning_rate": 9.739895792687758e-05, "loss": 0.0384, "step": 4440 }, { "grad_norm": 0.23704089224338531, "learning_rate": 9.738138403298157e-05, "loss": 0.0307, "step": 4450 }, { "grad_norm": 0.2078835368156433, "learning_rate": 9.736375256623619e-05, "loss": 0.0303, "step": 4460 }, { "grad_norm": 0.2367943674325943, "learning_rate": 9.734606354806533e-05, "loss": 0.0282, "step": 4470 }, { "grad_norm": 0.2054702192544937, "learning_rate": 9.73283169999628e-05, "loss": 0.0304, "step": 4480 }, { "grad_norm": 0.25434809923171997, "learning_rate": 9.731051294349238e-05, "loss": 0.0299, "step": 4490 }, { "grad_norm": 0.25925928354263306, "learning_rate": 9.729265140028762e-05, "loss": 0.0319, "step": 4500 }, { "grad_norm": 0.27313098311424255, "learning_rate": 9.727473239205201e-05, "loss": 0.0359, "step": 4510 }, { "grad_norm": 0.2613179385662079, "learning_rate": 9.725675594055883e-05, "loss": 0.0322, "step": 4520 }, { "grad_norm": 0.23891250789165497, "learning_rate": 9.723872206765116e-05, "loss": 0.0302, "step": 4530 }, { "grad_norm": 0.1775926649570465, "learning_rate": 9.722063079524185e-05, "loss": 0.0288, "step": 4540 }, { "grad_norm": 0.23229677975177765, "learning_rate": 9.720248214531351e-05, "loss": 0.0276, "step": 4550 }, { "grad_norm": 0.27787381410598755, "learning_rate": 9.718427613991848e-05, "loss": 0.0292, "step": 4560 }, { "grad_norm": 0.28231915831565857, "learning_rate": 9.716601280117873e-05, "loss": 0.0315, "step": 4570 }, { "grad_norm": 0.28531673550605774, "learning_rate": 9.714769215128596e-05, "loss": 0.03, "step": 4580 }, { "grad_norm": 0.19408665597438812, "learning_rate": 9.712931421250152e-05, "loss": 0.0274, "step": 4590 }, { "grad_norm": 0.29593369364738464, "learning_rate": 9.711087900715627e-05, "loss": 0.0275, "step": 4600 }, { "grad_norm": 0.2201319932937622, "learning_rate": 9.709238655765078e-05, "loss": 0.0278, "step": 4610 }, { "grad_norm": 0.19389986991882324, "learning_rate": 9.707383688645511e-05, "loss": 0.0312, "step": 4620 }, { "grad_norm": 0.20424561202526093, "learning_rate": 9.705523001610883e-05, "loss": 0.0279, "step": 4630 }, { "grad_norm": 0.22056585550308228, "learning_rate": 9.703656596922107e-05, "loss": 0.0298, "step": 4640 }, { "grad_norm": 0.26575881242752075, "learning_rate": 9.70178447684704e-05, "loss": 0.0307, "step": 4650 }, { "grad_norm": 0.3156203627586365, "learning_rate": 9.699906643660483e-05, "loss": 0.0384, "step": 4660 }, { "grad_norm": 0.1968419849872589, "learning_rate": 9.698023099644185e-05, "loss": 0.0331, "step": 4670 }, { "grad_norm": 0.2618681490421295, "learning_rate": 9.696133847086823e-05, "loss": 0.0268, "step": 4680 }, { "grad_norm": 0.21128517389297485, "learning_rate": 9.694238888284022e-05, "loss": 0.0337, "step": 4690 }, { "grad_norm": 0.24556384980678558, "learning_rate": 9.692338225538333e-05, "loss": 0.0369, "step": 4700 }, { "grad_norm": 0.2761615812778473, "learning_rate": 9.690431861159241e-05, "loss": 0.0292, "step": 4710 }, { "grad_norm": 0.1896582394838333, "learning_rate": 9.688519797463161e-05, "loss": 0.0305, "step": 4720 }, { "grad_norm": 0.2554270923137665, "learning_rate": 9.686602036773426e-05, "loss": 0.0315, "step": 4730 }, { "grad_norm": 0.20166057348251343, "learning_rate": 9.684678581420302e-05, "loss": 0.0266, "step": 4740 }, { "grad_norm": 0.23929958045482635, "learning_rate": 9.682749433740962e-05, "loss": 0.0284, "step": 4750 }, { "grad_norm": 0.23918992280960083, "learning_rate": 9.680814596079507e-05, "loss": 0.0287, "step": 4760 }, { "grad_norm": 0.2302311211824417, "learning_rate": 9.678874070786945e-05, "loss": 0.0293, "step": 4770 }, { "grad_norm": 0.25526782870292664, "learning_rate": 9.676927860221199e-05, "loss": 0.032, "step": 4780 }, { "grad_norm": 0.2136276364326477, "learning_rate": 9.674975966747097e-05, "loss": 0.0302, "step": 4790 }, { "grad_norm": 0.2523040771484375, "learning_rate": 9.673018392736374e-05, "loss": 0.0331, "step": 4800 }, { "grad_norm": 0.20681831240653992, "learning_rate": 9.671055140567667e-05, "loss": 0.0272, "step": 4810 }, { "grad_norm": 0.2259078323841095, "learning_rate": 9.669086212626511e-05, "loss": 0.0261, "step": 4820 }, { "grad_norm": 0.2749622166156769, "learning_rate": 9.667111611305341e-05, "loss": 0.0324, "step": 4830 }, { "grad_norm": 0.24899350106716156, "learning_rate": 9.665131339003486e-05, "loss": 0.0402, "step": 4840 }, { "grad_norm": 0.27027639746665955, "learning_rate": 9.663145398127158e-05, "loss": 0.0316, "step": 4850 }, { "grad_norm": 0.2038101851940155, "learning_rate": 9.661153791089467e-05, "loss": 0.0313, "step": 4860 }, { "grad_norm": 0.20739710330963135, "learning_rate": 9.659156520310402e-05, "loss": 0.0289, "step": 4870 }, { "grad_norm": 0.19637799263000488, "learning_rate": 9.657153588216835e-05, "loss": 0.0336, "step": 4880 }, { "grad_norm": 0.22023335099220276, "learning_rate": 9.655144997242516e-05, "loss": 0.0317, "step": 4890 }, { "grad_norm": 0.29066839814186096, "learning_rate": 9.653130749828075e-05, "loss": 0.0323, "step": 4900 }, { "grad_norm": 0.22450421750545502, "learning_rate": 9.65111084842101e-05, "loss": 0.0284, "step": 4910 }, { "grad_norm": 0.23808757960796356, "learning_rate": 9.649085295475695e-05, "loss": 0.0275, "step": 4920 }, { "grad_norm": 0.20806172490119934, "learning_rate": 9.647054093453365e-05, "loss": 0.0316, "step": 4930 }, { "grad_norm": 0.2135685831308365, "learning_rate": 9.645017244822123e-05, "loss": 0.028, "step": 4940 }, { "grad_norm": 0.24151793122291565, "learning_rate": 9.642974752056931e-05, "loss": 0.0305, "step": 4950 }, { "grad_norm": 0.21959035098552704, "learning_rate": 9.640926617639613e-05, "loss": 0.0273, "step": 4960 }, { "grad_norm": 0.2654542624950409, "learning_rate": 9.638872844058843e-05, "loss": 0.0338, "step": 4970 }, { "grad_norm": 0.25512394309043884, "learning_rate": 9.63681343381015e-05, "loss": 0.0323, "step": 4980 }, { "grad_norm": 0.26256680488586426, "learning_rate": 9.634748389395914e-05, "loss": 0.0268, "step": 4990 }, { "grad_norm": 0.24870331585407257, "learning_rate": 9.632677713325353e-05, "loss": 0.0298, "step": 5000 }, { "grad_norm": 0.19756197929382324, "learning_rate": 9.63060140811454e-05, "loss": 0.0308, "step": 5010 }, { "grad_norm": 0.2211374193429947, "learning_rate": 9.628519476286379e-05, "loss": 0.0286, "step": 5020 }, { "grad_norm": 0.20613257586956024, "learning_rate": 9.626431920370612e-05, "loss": 0.0277, "step": 5030 }, { "grad_norm": 0.2055036872625351, "learning_rate": 9.624338742903819e-05, "loss": 0.0274, "step": 5040 }, { "grad_norm": 0.2792186439037323, "learning_rate": 9.622239946429406e-05, "loss": 0.0296, "step": 5050 }, { "grad_norm": 0.24794518947601318, "learning_rate": 9.620135533497609e-05, "loss": 0.0276, "step": 5060 }, { "grad_norm": 0.2708359658718109, "learning_rate": 9.61802550666549e-05, "loss": 0.0257, "step": 5070 }, { "grad_norm": 0.23689919710159302, "learning_rate": 9.615909868496928e-05, "loss": 0.0323, "step": 5080 }, { "grad_norm": 0.26406893134117126, "learning_rate": 9.613788621562622e-05, "loss": 0.0301, "step": 5090 }, { "grad_norm": 0.2077074646949768, "learning_rate": 9.611661768440093e-05, "loss": 0.0307, "step": 5100 }, { "grad_norm": 0.2537584900856018, "learning_rate": 9.609529311713661e-05, "loss": 0.0266, "step": 5110 }, { "grad_norm": 0.21507342159748077, "learning_rate": 9.607391253974466e-05, "loss": 0.0339, "step": 5120 }, { "grad_norm": 0.1875399351119995, "learning_rate": 9.605247597820448e-05, "loss": 0.0248, "step": 5130 }, { "grad_norm": 0.21934334933757782, "learning_rate": 9.603098345856354e-05, "loss": 0.0275, "step": 5140 }, { "grad_norm": 0.25349143147468567, "learning_rate": 9.600943500693725e-05, "loss": 0.025, "step": 5150 }, { "grad_norm": 0.19082336127758026, "learning_rate": 9.598783064950902e-05, "loss": 0.0281, "step": 5160 }, { "grad_norm": 0.2549755573272705, "learning_rate": 9.596617041253018e-05, "loss": 0.0243, "step": 5170 }, { "grad_norm": 0.2391955852508545, "learning_rate": 9.594445432231996e-05, "loss": 0.0287, "step": 5180 }, { "grad_norm": 0.19563959538936615, "learning_rate": 9.592268240526547e-05, "loss": 0.0306, "step": 5190 }, { "grad_norm": 0.18902570009231567, "learning_rate": 9.590085468782162e-05, "loss": 0.0333, "step": 5200 }, { "grad_norm": 0.21910232305526733, "learning_rate": 9.587897119651116e-05, "loss": 0.0314, "step": 5210 }, { "grad_norm": 0.23597417771816254, "learning_rate": 9.585703195792459e-05, "loss": 0.0323, "step": 5220 }, { "grad_norm": 0.22345948219299316, "learning_rate": 9.583503699872016e-05, "loss": 0.0304, "step": 5230 }, { "grad_norm": 0.2097024768590927, "learning_rate": 9.581298634562381e-05, "loss": 0.0279, "step": 5240 }, { "grad_norm": 0.195930615067482, "learning_rate": 9.579088002542917e-05, "loss": 0.028, "step": 5250 }, { "grad_norm": 0.24371151626110077, "learning_rate": 9.57687180649975e-05, "loss": 0.0316, "step": 5260 }, { "grad_norm": 0.24893689155578613, "learning_rate": 9.574650049125768e-05, "loss": 0.0318, "step": 5270 }, { "grad_norm": 0.24635446071624756, "learning_rate": 9.572422733120614e-05, "loss": 0.0315, "step": 5280 }, { "grad_norm": 0.23562034964561462, "learning_rate": 9.570189861190689e-05, "loss": 0.0345, "step": 5290 }, { "grad_norm": 0.21763862669467926, "learning_rate": 9.56795143604914e-05, "loss": 0.0311, "step": 5300 }, { "grad_norm": 0.2498406618833542, "learning_rate": 9.565707460415869e-05, "loss": 0.0313, "step": 5310 }, { "grad_norm": 0.2415490299463272, "learning_rate": 9.563457937017515e-05, "loss": 0.0333, "step": 5320 }, { "grad_norm": 0.23412320017814636, "learning_rate": 9.56120286858746e-05, "loss": 0.0276, "step": 5330 }, { "grad_norm": 0.20872823894023895, "learning_rate": 9.558942257865829e-05, "loss": 0.0323, "step": 5340 }, { "grad_norm": 0.2229330688714981, "learning_rate": 9.556676107599472e-05, "loss": 0.0261, "step": 5350 }, { "grad_norm": 0.32876741886138916, "learning_rate": 9.554404420541978e-05, "loss": 0.0312, "step": 5360 }, { "grad_norm": 0.2125527709722519, "learning_rate": 9.55212719945366e-05, "loss": 0.0332, "step": 5370 }, { "grad_norm": 0.17500363290309906, "learning_rate": 9.549844447101559e-05, "loss": 0.0282, "step": 5380 }, { "grad_norm": 0.19760490953922272, "learning_rate": 9.547556166259433e-05, "loss": 0.0271, "step": 5390 }, { "grad_norm": 0.23126821219921112, "learning_rate": 9.545262359707756e-05, "loss": 0.029, "step": 5400 }, { "grad_norm": 0.18249264359474182, "learning_rate": 9.542963030233724e-05, "loss": 0.0295, "step": 5410 }, { "grad_norm": 0.279461532831192, "learning_rate": 9.540658180631237e-05, "loss": 0.0263, "step": 5420 }, { "grad_norm": 0.2667381167411804, "learning_rate": 9.538347813700904e-05, "loss": 0.0279, "step": 5430 }, { "grad_norm": 0.2437836080789566, "learning_rate": 9.536031932250036e-05, "loss": 0.0276, "step": 5440 }, { "grad_norm": 0.2382870763540268, "learning_rate": 9.533710539092653e-05, "loss": 0.0298, "step": 5450 }, { "grad_norm": 0.2902277708053589, "learning_rate": 9.531383637049464e-05, "loss": 0.0275, "step": 5460 }, { "grad_norm": 0.24929934740066528, "learning_rate": 9.529051228947875e-05, "loss": 0.028, "step": 5470 }, { "grad_norm": 0.20127563178539276, "learning_rate": 9.52671331762198e-05, "loss": 0.026, "step": 5480 }, { "grad_norm": 0.22337953746318817, "learning_rate": 9.524369905912565e-05, "loss": 0.0248, "step": 5490 }, { "grad_norm": 0.1920941323041916, "learning_rate": 9.522020996667092e-05, "loss": 0.0243, "step": 5500 }, { "grad_norm": 0.19813159108161926, "learning_rate": 9.519666592739709e-05, "loss": 0.0275, "step": 5510 }, { "grad_norm": 0.21987876296043396, "learning_rate": 9.517306696991241e-05, "loss": 0.0238, "step": 5520 }, { "grad_norm": 0.19997544586658478, "learning_rate": 9.51494131228918e-05, "loss": 0.0324, "step": 5530 }, { "grad_norm": 0.18335038423538208, "learning_rate": 9.512570441507695e-05, "loss": 0.0272, "step": 5540 }, { "grad_norm": 0.21134313941001892, "learning_rate": 9.510194087527615e-05, "loss": 0.0246, "step": 5550 }, { "grad_norm": 0.2130938321352005, "learning_rate": 9.507812253236435e-05, "loss": 0.0279, "step": 5560 }, { "grad_norm": 0.2252819687128067, "learning_rate": 9.505424941528309e-05, "loss": 0.0254, "step": 5570 }, { "grad_norm": 0.24322855472564697, "learning_rate": 9.503032155304046e-05, "loss": 0.0288, "step": 5580 }, { "grad_norm": 0.2520354688167572, "learning_rate": 9.500633897471106e-05, "loss": 0.0283, "step": 5590 }, { "grad_norm": 0.24814407527446747, "learning_rate": 9.498230170943596e-05, "loss": 0.0316, "step": 5600 }, { "grad_norm": 0.18254344165325165, "learning_rate": 9.495820978642275e-05, "loss": 0.0299, "step": 5610 }, { "grad_norm": 0.2603053152561188, "learning_rate": 9.493406323494535e-05, "loss": 0.0298, "step": 5620 }, { "grad_norm": 0.23372521996498108, "learning_rate": 9.490986208434413e-05, "loss": 0.0274, "step": 5630 }, { "grad_norm": 0.271401047706604, "learning_rate": 9.488560636402577e-05, "loss": 0.0327, "step": 5640 }, { "grad_norm": 0.24841144680976868, "learning_rate": 9.486129610346321e-05, "loss": 0.0253, "step": 5650 }, { "grad_norm": 0.22093693912029266, "learning_rate": 9.483693133219576e-05, "loss": 0.0286, "step": 5660 }, { "grad_norm": 0.20731639862060547, "learning_rate": 9.481251207982888e-05, "loss": 0.0281, "step": 5670 }, { "grad_norm": 0.20849162340164185, "learning_rate": 9.47880383760343e-05, "loss": 0.0253, "step": 5680 }, { "grad_norm": 0.2080790251493454, "learning_rate": 9.476351025054983e-05, "loss": 0.0296, "step": 5690 }, { "grad_norm": 0.2280687689781189, "learning_rate": 9.473892773317952e-05, "loss": 0.0228, "step": 5700 }, { "grad_norm": 0.23908893764019012, "learning_rate": 9.471429085379338e-05, "loss": 0.0292, "step": 5710 }, { "grad_norm": 0.19067363440990448, "learning_rate": 9.468959964232757e-05, "loss": 0.029, "step": 5720 }, { "grad_norm": 0.19661135971546173, "learning_rate": 9.466485412878425e-05, "loss": 0.0325, "step": 5730 }, { "grad_norm": 0.23355215787887573, "learning_rate": 9.464005434323154e-05, "loss": 0.0271, "step": 5740 }, { "grad_norm": 0.21555110812187195, "learning_rate": 9.461520031580352e-05, "loss": 0.0259, "step": 5750 }, { "grad_norm": 0.19003425538539886, "learning_rate": 9.459029207670019e-05, "loss": 0.0264, "step": 5760 }, { "grad_norm": 0.19415143132209778, "learning_rate": 9.456532965618737e-05, "loss": 0.0255, "step": 5770 }, { "grad_norm": 0.2283756136894226, "learning_rate": 9.454031308459681e-05, "loss": 0.0299, "step": 5780 }, { "grad_norm": 0.16991578042507172, "learning_rate": 9.451524239232595e-05, "loss": 0.0281, "step": 5790 }, { "grad_norm": 0.1799631416797638, "learning_rate": 9.449011760983809e-05, "loss": 0.0271, "step": 5800 }, { "grad_norm": 0.216788187623024, "learning_rate": 9.446493876766218e-05, "loss": 0.0241, "step": 5810 }, { "grad_norm": 0.2394324392080307, "learning_rate": 9.44397058963929e-05, "loss": 0.0264, "step": 5820 }, { "grad_norm": 0.1983470320701599, "learning_rate": 9.441441902669056e-05, "loss": 0.0288, "step": 5830 }, { "grad_norm": 0.22269758582115173, "learning_rate": 9.43890781892811e-05, "loss": 0.0279, "step": 5840 }, { "grad_norm": 0.22559042274951935, "learning_rate": 9.436368341495603e-05, "loss": 0.0363, "step": 5850 }, { "grad_norm": 0.23857998847961426, "learning_rate": 9.43382347345724e-05, "loss": 0.0282, "step": 5860 }, { "grad_norm": 0.1977320909500122, "learning_rate": 9.431273217905273e-05, "loss": 0.0273, "step": 5870 }, { "grad_norm": 0.2392677366733551, "learning_rate": 9.428717577938504e-05, "loss": 0.0351, "step": 5880 }, { "grad_norm": 0.21501755714416504, "learning_rate": 9.426156556662276e-05, "loss": 0.0255, "step": 5890 }, { "grad_norm": 0.2095785290002823, "learning_rate": 9.423590157188474e-05, "loss": 0.0268, "step": 5900 }, { "grad_norm": 0.2505526542663574, "learning_rate": 9.421018382635513e-05, "loss": 0.0259, "step": 5910 }, { "grad_norm": 0.17474409937858582, "learning_rate": 9.418441236128343e-05, "loss": 0.028, "step": 5920 }, { "grad_norm": 0.23237846791744232, "learning_rate": 9.41585872079844e-05, "loss": 0.0247, "step": 5930 }, { "grad_norm": 0.23669125139713287, "learning_rate": 9.413270839783802e-05, "loss": 0.0279, "step": 5940 }, { "grad_norm": 0.21580657362937927, "learning_rate": 9.41067759622895e-05, "loss": 0.0262, "step": 5950 }, { "grad_norm": 0.20861980319023132, "learning_rate": 9.408078993284917e-05, "loss": 0.0239, "step": 5960 }, { "grad_norm": 0.244109109044075, "learning_rate": 9.405475034109254e-05, "loss": 0.0229, "step": 5970 }, { "grad_norm": 0.1993977427482605, "learning_rate": 9.402865721866015e-05, "loss": 0.0252, "step": 5980 }, { "grad_norm": 0.18189267814159393, "learning_rate": 9.400251059725762e-05, "loss": 0.0246, "step": 5990 }, { "grad_norm": 0.17691929638385773, "learning_rate": 9.397631050865554e-05, "loss": 0.0251, "step": 6000 }, { "grad_norm": 0.16933465003967285, "learning_rate": 9.395005698468949e-05, "loss": 0.0236, "step": 6010 }, { "grad_norm": 0.2661931812763214, "learning_rate": 9.392375005726e-05, "loss": 0.0255, "step": 6020 }, { "grad_norm": 0.24974742531776428, "learning_rate": 9.389738975833242e-05, "loss": 0.0237, "step": 6030 }, { "grad_norm": 0.19385753571987152, "learning_rate": 9.387097611993707e-05, "loss": 0.0282, "step": 6040 }, { "grad_norm": 0.19951003789901733, "learning_rate": 9.384450917416894e-05, "loss": 0.0224, "step": 6050 }, { "grad_norm": 0.1801142543554306, "learning_rate": 9.381798895318792e-05, "loss": 0.0259, "step": 6060 }, { "grad_norm": 0.1975366324186325, "learning_rate": 9.379141548921853e-05, "loss": 0.026, "step": 6070 }, { "grad_norm": 0.24601046741008759, "learning_rate": 9.376478881455009e-05, "loss": 0.0273, "step": 6080 }, { "grad_norm": 0.2139693647623062, "learning_rate": 9.373810896153647e-05, "loss": 0.0252, "step": 6090 }, { "grad_norm": 0.19255594909191132, "learning_rate": 9.371137596259623e-05, "loss": 0.0209, "step": 6100 }, { "grad_norm": 0.1453535556793213, "learning_rate": 9.368458985021248e-05, "loss": 0.0219, "step": 6110 }, { "grad_norm": 0.2992496192455292, "learning_rate": 9.365775065693287e-05, "loss": 0.0253, "step": 6120 }, { "grad_norm": 0.14981943368911743, "learning_rate": 9.363085841536957e-05, "loss": 0.0237, "step": 6130 }, { "grad_norm": 0.21052013337612152, "learning_rate": 9.360391315819917e-05, "loss": 0.0221, "step": 6140 }, { "grad_norm": 0.18235358595848083, "learning_rate": 9.357691491816269e-05, "loss": 0.0212, "step": 6150 }, { "grad_norm": 0.18436577916145325, "learning_rate": 9.354986372806557e-05, "loss": 0.0271, "step": 6160 }, { "grad_norm": 0.20776143670082092, "learning_rate": 9.352275962077752e-05, "loss": 0.0242, "step": 6170 }, { "grad_norm": 0.1481848806142807, "learning_rate": 9.349560262923262e-05, "loss": 0.0245, "step": 6180 }, { "grad_norm": 0.24450261890888214, "learning_rate": 9.346839278642913e-05, "loss": 0.0263, "step": 6190 }, { "grad_norm": 0.23981431126594543, "learning_rate": 9.344113012542963e-05, "loss": 0.022, "step": 6200 }, { "grad_norm": 0.32588204741477966, "learning_rate": 9.341381467936079e-05, "loss": 0.0226, "step": 6210 }, { "grad_norm": 0.1978333592414856, "learning_rate": 9.338644648141346e-05, "loss": 0.0224, "step": 6220 }, { "grad_norm": 0.2076939046382904, "learning_rate": 9.335902556484257e-05, "loss": 0.0232, "step": 6230 }, { "grad_norm": 0.2210085093975067, "learning_rate": 9.333155196296712e-05, "loss": 0.0244, "step": 6240 }, { "grad_norm": 0.16166488826274872, "learning_rate": 9.330402570917016e-05, "loss": 0.0257, "step": 6250 }, { "grad_norm": 0.20394763350486755, "learning_rate": 9.327644683689865e-05, "loss": 0.0273, "step": 6260 }, { "grad_norm": 0.31421348452568054, "learning_rate": 9.324881537966354e-05, "loss": 0.0247, "step": 6270 }, { "grad_norm": 0.3034772574901581, "learning_rate": 9.322113137103964e-05, "loss": 0.0275, "step": 6280 }, { "grad_norm": 0.1900094896554947, "learning_rate": 9.319339484466564e-05, "loss": 0.0279, "step": 6290 }, { "grad_norm": 0.17801028490066528, "learning_rate": 9.316560583424403e-05, "loss": 0.0224, "step": 6300 }, { "grad_norm": 0.22439001500606537, "learning_rate": 9.313776437354108e-05, "loss": 0.024, "step": 6310 }, { "grad_norm": 0.22454480826854706, "learning_rate": 9.31098704963868e-05, "loss": 0.0232, "step": 6320 }, { "grad_norm": 0.22801874577999115, "learning_rate": 9.308192423667485e-05, "loss": 0.0224, "step": 6330 }, { "grad_norm": 0.2176494002342224, "learning_rate": 9.305392562836262e-05, "loss": 0.0248, "step": 6340 }, { "grad_norm": 0.17874518036842346, "learning_rate": 9.3025874705471e-05, "loss": 0.0224, "step": 6350 }, { "grad_norm": 0.23807986080646515, "learning_rate": 9.299777150208457e-05, "loss": 0.022, "step": 6360 }, { "grad_norm": 0.20688492059707642, "learning_rate": 9.296961605235133e-05, "loss": 0.0231, "step": 6370 }, { "grad_norm": 0.222052201628685, "learning_rate": 9.29414083904828e-05, "loss": 0.0184, "step": 6380 }, { "grad_norm": 0.23517931997776031, "learning_rate": 9.291314855075397e-05, "loss": 0.0223, "step": 6390 }, { "grad_norm": 0.20439916849136353, "learning_rate": 9.288483656750322e-05, "loss": 0.0203, "step": 6400 }, { "grad_norm": 0.2226622849702835, "learning_rate": 9.285647247513225e-05, "loss": 0.0225, "step": 6410 }, { "grad_norm": 0.20026129484176636, "learning_rate": 9.282805630810614e-05, "loss": 0.0235, "step": 6420 }, { "grad_norm": 0.21602140367031097, "learning_rate": 9.279958810095317e-05, "loss": 0.019, "step": 6430 }, { "grad_norm": 0.26315730810165405, "learning_rate": 9.277106788826494e-05, "loss": 0.021, "step": 6440 }, { "grad_norm": 0.2178439348936081, "learning_rate": 9.274249570469618e-05, "loss": 0.0207, "step": 6450 }, { "grad_norm": 0.21189862489700317, "learning_rate": 9.271387158496476e-05, "loss": 0.0205, "step": 6460 }, { "grad_norm": 0.18360303342342377, "learning_rate": 9.268519556385173e-05, "loss": 0.0252, "step": 6470 }, { "grad_norm": 0.21340900659561157, "learning_rate": 9.265646767620113e-05, "loss": 0.0209, "step": 6480 }, { "grad_norm": 0.227768212556839, "learning_rate": 9.262768795692006e-05, "loss": 0.0225, "step": 6490 }, { "grad_norm": 0.22897422313690186, "learning_rate": 9.259885644097861e-05, "loss": 0.0219, "step": 6500 }, { "grad_norm": 0.20174172520637512, "learning_rate": 9.256997316340976e-05, "loss": 0.026, "step": 6510 }, { "grad_norm": 0.2745364308357239, "learning_rate": 9.254103815930943e-05, "loss": 0.0289, "step": 6520 }, { "grad_norm": 0.20963944494724274, "learning_rate": 9.251205146383636e-05, "loss": 0.0249, "step": 6530 }, { "grad_norm": 0.1974862962961197, "learning_rate": 9.248301311221216e-05, "loss": 0.0245, "step": 6540 }, { "grad_norm": 0.16240490972995758, "learning_rate": 9.245392313972115e-05, "loss": 0.0227, "step": 6550 }, { "grad_norm": 0.23451045155525208, "learning_rate": 9.242478158171037e-05, "loss": 0.025, "step": 6560 }, { "grad_norm": 0.20655550062656403, "learning_rate": 9.239558847358957e-05, "loss": 0.0275, "step": 6570 }, { "grad_norm": 0.19521363079547882, "learning_rate": 9.236634385083114e-05, "loss": 0.028, "step": 6580 }, { "grad_norm": 0.1746223121881485, "learning_rate": 9.233704774897005e-05, "loss": 0.0236, "step": 6590 }, { "grad_norm": 0.24007980525493622, "learning_rate": 9.230770020360382e-05, "loss": 0.0225, "step": 6600 }, { "grad_norm": 0.21855007112026215, "learning_rate": 9.227830125039247e-05, "loss": 0.0221, "step": 6610 }, { "grad_norm": 0.17761805653572083, "learning_rate": 9.224885092505853e-05, "loss": 0.0227, "step": 6620 }, { "grad_norm": 0.18426430225372314, "learning_rate": 9.221934926338691e-05, "loss": 0.0254, "step": 6630 }, { "grad_norm": 0.2656153440475464, "learning_rate": 9.21897963012249e-05, "loss": 0.0252, "step": 6640 }, { "grad_norm": 0.24706673622131348, "learning_rate": 9.216019207448217e-05, "loss": 0.0213, "step": 6650 }, { "grad_norm": 0.21878854930400848, "learning_rate": 9.21305366191306e-05, "loss": 0.0242, "step": 6660 }, { "grad_norm": 0.18987292051315308, "learning_rate": 9.21008299712044e-05, "loss": 0.0222, "step": 6670 }, { "grad_norm": 0.14003948867321014, "learning_rate": 9.207107216679993e-05, "loss": 0.0247, "step": 6680 }, { "grad_norm": 0.1884041428565979, "learning_rate": 9.204126324207574e-05, "loss": 0.0228, "step": 6690 }, { "grad_norm": 0.21157129108905792, "learning_rate": 9.201140323325248e-05, "loss": 0.0229, "step": 6700 }, { "grad_norm": 0.18778017163276672, "learning_rate": 9.198149217661287e-05, "loss": 0.0231, "step": 6710 }, { "grad_norm": 0.18699568510055542, "learning_rate": 9.195153010850167e-05, "loss": 0.0229, "step": 6720 }, { "grad_norm": 0.2003507763147354, "learning_rate": 9.192151706532562e-05, "loss": 0.0234, "step": 6730 }, { "grad_norm": 0.18195615708827972, "learning_rate": 9.189145308355337e-05, "loss": 0.0252, "step": 6740 }, { "grad_norm": 0.18145667016506195, "learning_rate": 9.186133819971555e-05, "loss": 0.022, "step": 6750 }, { "grad_norm": 0.20983484387397766, "learning_rate": 9.183117245040455e-05, "loss": 0.0236, "step": 6760 }, { "grad_norm": 0.21380503475666046, "learning_rate": 9.180095587227458e-05, "loss": 0.0207, "step": 6770 }, { "grad_norm": 0.20138464868068695, "learning_rate": 9.177068850204167e-05, "loss": 0.0213, "step": 6780 }, { "grad_norm": 0.22054819762706757, "learning_rate": 9.174037037648351e-05, "loss": 0.0222, "step": 6790 }, { "grad_norm": 0.21752966940402985, "learning_rate": 9.171000153243947e-05, "loss": 0.0263, "step": 6800 }, { "grad_norm": 0.2018434703350067, "learning_rate": 9.167958200681057e-05, "loss": 0.0198, "step": 6810 }, { "grad_norm": 0.24211491644382477, "learning_rate": 9.164911183655943e-05, "loss": 0.0232, "step": 6820 }, { "grad_norm": 0.22846050560474396, "learning_rate": 9.161859105871013e-05, "loss": 0.0201, "step": 6830 }, { "grad_norm": 0.14948436617851257, "learning_rate": 9.158801971034832e-05, "loss": 0.0205, "step": 6840 }, { "grad_norm": 0.20917269587516785, "learning_rate": 9.155739782862107e-05, "loss": 0.0212, "step": 6850 }, { "grad_norm": 0.1984214186668396, "learning_rate": 9.152672545073687e-05, "loss": 0.0266, "step": 6860 }, { "grad_norm": 0.21800899505615234, "learning_rate": 9.149600261396551e-05, "loss": 0.0261, "step": 6870 }, { "grad_norm": 0.2503814399242401, "learning_rate": 9.146522935563817e-05, "loss": 0.0271, "step": 6880 }, { "grad_norm": 0.22620424628257751, "learning_rate": 9.143440571314723e-05, "loss": 0.0234, "step": 6890 }, { "grad_norm": 0.17918923497200012, "learning_rate": 9.140353172394636e-05, "loss": 0.0226, "step": 6900 }, { "grad_norm": 0.17830803990364075, "learning_rate": 9.137260742555032e-05, "loss": 0.0233, "step": 6910 }, { "grad_norm": 0.24200162291526794, "learning_rate": 9.13416328555351e-05, "loss": 0.0295, "step": 6920 }, { "grad_norm": 0.24629199504852295, "learning_rate": 9.131060805153769e-05, "loss": 0.0211, "step": 6930 }, { "grad_norm": 0.21278508007526398, "learning_rate": 9.127953305125618e-05, "loss": 0.0214, "step": 6940 }, { "grad_norm": 0.16768500208854675, "learning_rate": 9.124840789244958e-05, "loss": 0.0254, "step": 6950 }, { "grad_norm": 0.20481671392917633, "learning_rate": 9.121723261293794e-05, "loss": 0.0242, "step": 6960 }, { "grad_norm": 0.17720167338848114, "learning_rate": 9.118600725060214e-05, "loss": 0.0222, "step": 6970 }, { "grad_norm": 0.17534491419792175, "learning_rate": 9.115473184338392e-05, "loss": 0.0216, "step": 6980 }, { "grad_norm": 0.24215716123580933, "learning_rate": 9.11234064292859e-05, "loss": 0.0231, "step": 6990 }, { "grad_norm": 0.16277913749217987, "learning_rate": 9.109203104637137e-05, "loss": 0.0194, "step": 7000 }, { "grad_norm": 0.2186107486486435, "learning_rate": 9.10606057327644e-05, "loss": 0.0213, "step": 7010 }, { "grad_norm": 0.19274434447288513, "learning_rate": 9.10291305266497e-05, "loss": 0.0171, "step": 7020 }, { "grad_norm": 0.16718555986881256, "learning_rate": 9.099760546627261e-05, "loss": 0.0208, "step": 7030 }, { "grad_norm": 0.18804149329662323, "learning_rate": 9.096603058993907e-05, "loss": 0.0216, "step": 7040 }, { "grad_norm": 0.2088651806116104, "learning_rate": 9.093440593601552e-05, "loss": 0.0268, "step": 7050 }, { "grad_norm": 0.20545841753482819, "learning_rate": 9.090273154292889e-05, "loss": 0.0197, "step": 7060 }, { "grad_norm": 0.21352049708366394, "learning_rate": 9.087100744916656e-05, "loss": 0.0219, "step": 7070 }, { "grad_norm": 0.19377663731575012, "learning_rate": 9.08392336932763e-05, "loss": 0.018, "step": 7080 }, { "grad_norm": 0.19949892163276672, "learning_rate": 9.080741031386619e-05, "loss": 0.019, "step": 7090 }, { "grad_norm": 0.20815543830394745, "learning_rate": 9.077553734960468e-05, "loss": 0.0211, "step": 7100 }, { "grad_norm": 0.1725156158208847, "learning_rate": 9.074361483922041e-05, "loss": 0.0207, "step": 7110 }, { "grad_norm": 0.20371699333190918, "learning_rate": 9.071164282150224e-05, "loss": 0.0195, "step": 7120 }, { "grad_norm": 0.16127830743789673, "learning_rate": 9.067962133529919e-05, "loss": 0.0185, "step": 7130 }, { "grad_norm": 0.15961019694805145, "learning_rate": 9.064755041952036e-05, "loss": 0.0213, "step": 7140 }, { "grad_norm": 0.2014247477054596, "learning_rate": 9.061543011313497e-05, "loss": 0.0221, "step": 7150 }, { "grad_norm": 0.18235579133033752, "learning_rate": 9.05832604551722e-05, "loss": 0.0248, "step": 7160 }, { "grad_norm": 0.22609268128871918, "learning_rate": 9.055104148472124e-05, "loss": 0.0226, "step": 7170 }, { "grad_norm": 0.18519537150859833, "learning_rate": 9.051877324093114e-05, "loss": 0.0216, "step": 7180 }, { "grad_norm": 0.2602714002132416, "learning_rate": 9.04864557630109e-05, "loss": 0.0275, "step": 7190 }, { "grad_norm": 0.1583767831325531, "learning_rate": 9.045408909022928e-05, "loss": 0.021, "step": 7200 }, { "grad_norm": 0.1936240941286087, "learning_rate": 9.042167326191484e-05, "loss": 0.019, "step": 7210 }, { "grad_norm": 0.16838723421096802, "learning_rate": 9.038920831745587e-05, "loss": 0.0229, "step": 7220 }, { "grad_norm": 0.17987263202667236, "learning_rate": 9.035669429630036e-05, "loss": 0.0191, "step": 7230 }, { "grad_norm": 0.12038452923297882, "learning_rate": 9.032413123795588e-05, "loss": 0.0181, "step": 7240 }, { "grad_norm": 0.23048876225948334, "learning_rate": 9.029151918198962e-05, "loss": 0.0181, "step": 7250 }, { "grad_norm": 0.2060045748949051, "learning_rate": 9.025885816802833e-05, "loss": 0.0199, "step": 7260 }, { "grad_norm": 0.2478746622800827, "learning_rate": 9.02261482357582e-05, "loss": 0.0205, "step": 7270 }, { "grad_norm": 0.18881921470165253, "learning_rate": 9.019338942492485e-05, "loss": 0.0197, "step": 7280 }, { "grad_norm": 0.14296559989452362, "learning_rate": 9.01605817753334e-05, "loss": 0.0213, "step": 7290 }, { "grad_norm": 0.20327723026275635, "learning_rate": 9.012772532684818e-05, "loss": 0.022, "step": 7300 }, { "grad_norm": 0.221468985080719, "learning_rate": 9.00948201193929e-05, "loss": 0.0262, "step": 7310 }, { "grad_norm": 0.1641695201396942, "learning_rate": 9.006186619295047e-05, "loss": 0.0208, "step": 7320 }, { "grad_norm": 0.19850783050060272, "learning_rate": 9.002886358756305e-05, "loss": 0.0231, "step": 7330 }, { "grad_norm": 0.20414942502975464, "learning_rate": 8.999581234333188e-05, "loss": 0.0209, "step": 7340 }, { "grad_norm": 0.21204650402069092, "learning_rate": 8.996271250041734e-05, "loss": 0.0205, "step": 7350 }, { "grad_norm": 0.17479778826236725, "learning_rate": 8.99295640990389e-05, "loss": 0.0195, "step": 7360 }, { "grad_norm": 0.17730750143527985, "learning_rate": 8.989636717947494e-05, "loss": 0.0227, "step": 7370 }, { "grad_norm": 0.2271052747964859, "learning_rate": 8.986312178206291e-05, "loss": 0.0223, "step": 7380 }, { "grad_norm": 0.2005060315132141, "learning_rate": 8.982982794719904e-05, "loss": 0.0222, "step": 7390 }, { "grad_norm": 0.23985624313354492, "learning_rate": 8.979648571533852e-05, "loss": 0.0204, "step": 7400 }, { "grad_norm": 0.1692429780960083, "learning_rate": 8.976309512699528e-05, "loss": 0.0194, "step": 7410 }, { "grad_norm": 0.15606026351451874, "learning_rate": 8.972965622274205e-05, "loss": 0.0221, "step": 7420 }, { "grad_norm": 0.19871702790260315, "learning_rate": 8.969616904321026e-05, "loss": 0.0189, "step": 7430 }, { "grad_norm": 0.1780184656381607, "learning_rate": 8.966263362908998e-05, "loss": 0.0223, "step": 7440 }, { "grad_norm": 0.21088911592960358, "learning_rate": 8.962905002112989e-05, "loss": 0.0174, "step": 7450 }, { "grad_norm": 0.1882866621017456, "learning_rate": 8.959541826013724e-05, "loss": 0.0209, "step": 7460 }, { "grad_norm": 0.15823718905448914, "learning_rate": 8.956173838697781e-05, "loss": 0.0192, "step": 7470 }, { "grad_norm": 0.1296790987253189, "learning_rate": 8.952801044257581e-05, "loss": 0.021, "step": 7480 }, { "grad_norm": 0.18659135699272156, "learning_rate": 8.949423446791387e-05, "loss": 0.0209, "step": 7490 }, { "grad_norm": 0.19962072372436523, "learning_rate": 8.9460410504033e-05, "loss": 0.0183, "step": 7500 }, { "grad_norm": 0.1446976512670517, "learning_rate": 8.942653859203248e-05, "loss": 0.016, "step": 7510 }, { "grad_norm": 0.14320243895053864, "learning_rate": 8.93926187730699e-05, "loss": 0.0203, "step": 7520 }, { "grad_norm": 0.19179947674274445, "learning_rate": 8.935865108836102e-05, "loss": 0.0224, "step": 7530 }, { "grad_norm": 0.1815337985754013, "learning_rate": 8.932463557917981e-05, "loss": 0.0169, "step": 7540 }, { "grad_norm": 0.15696613490581512, "learning_rate": 8.929057228685829e-05, "loss": 0.0176, "step": 7550 }, { "grad_norm": 0.15198162198066711, "learning_rate": 8.925646125278658e-05, "loss": 0.0163, "step": 7560 }, { "grad_norm": 0.21881291270256042, "learning_rate": 8.92223025184128e-05, "loss": 0.0202, "step": 7570 }, { "grad_norm": 0.21143382787704468, "learning_rate": 8.918809612524303e-05, "loss": 0.0195, "step": 7580 }, { "grad_norm": 0.17254997789859772, "learning_rate": 8.91538421148413e-05, "loss": 0.0181, "step": 7590 }, { "grad_norm": 0.17870792746543884, "learning_rate": 8.91195405288294e-05, "loss": 0.0177, "step": 7600 }, { "grad_norm": 0.17141395807266235, "learning_rate": 8.908519140888705e-05, "loss": 0.0151, "step": 7610 }, { "grad_norm": 0.12811297178268433, "learning_rate": 8.905079479675164e-05, "loss": 0.0155, "step": 7620 }, { "grad_norm": 0.18393060564994812, "learning_rate": 8.901635073421831e-05, "loss": 0.0171, "step": 7630 }, { "grad_norm": 0.16364826261997223, "learning_rate": 8.898185926313983e-05, "loss": 0.017, "step": 7640 }, { "grad_norm": 0.16526290774345398, "learning_rate": 8.894732042542659e-05, "loss": 0.0208, "step": 7650 }, { "grad_norm": 0.21665020287036896, "learning_rate": 8.891273426304654e-05, "loss": 0.0182, "step": 7660 }, { "grad_norm": 0.19239342212677002, "learning_rate": 8.887810081802515e-05, "loss": 0.0203, "step": 7670 }, { "grad_norm": 0.2164546698331833, "learning_rate": 8.884342013244529e-05, "loss": 0.0196, "step": 7680 }, { "grad_norm": 0.19294947385787964, "learning_rate": 8.880869224844725e-05, "loss": 0.0197, "step": 7690 }, { "grad_norm": 0.1807573139667511, "learning_rate": 8.877391720822875e-05, "loss": 0.0197, "step": 7700 }, { "grad_norm": 0.19228847324848175, "learning_rate": 8.873909505404468e-05, "loss": 0.0214, "step": 7710 }, { "grad_norm": 0.19902953505516052, "learning_rate": 8.870422582820726e-05, "loss": 0.0177, "step": 7720 }, { "grad_norm": 0.18675169348716736, "learning_rate": 8.866930957308588e-05, "loss": 0.021, "step": 7730 }, { "grad_norm": 0.12282244861125946, "learning_rate": 8.86343463311071e-05, "loss": 0.0178, "step": 7740 }, { "grad_norm": 0.2047216296195984, "learning_rate": 8.859933614475452e-05, "loss": 0.019, "step": 7750 }, { "grad_norm": 0.19102618098258972, "learning_rate": 8.85642790565689e-05, "loss": 0.0189, "step": 7760 }, { "grad_norm": 0.18001291155815125, "learning_rate": 8.852917510914782e-05, "loss": 0.0184, "step": 7770 }, { "grad_norm": 0.21290306746959686, "learning_rate": 8.84940243451459e-05, "loss": 0.0206, "step": 7780 }, { "grad_norm": 0.14980022609233856, "learning_rate": 8.845882680727469e-05, "loss": 0.0183, "step": 7790 }, { "grad_norm": 0.1888519674539566, "learning_rate": 8.842358253830246e-05, "loss": 0.0203, "step": 7800 }, { "grad_norm": 0.18766991794109344, "learning_rate": 8.838829158105434e-05, "loss": 0.0208, "step": 7810 }, { "grad_norm": 0.20158043503761292, "learning_rate": 8.835295397841217e-05, "loss": 0.0243, "step": 7820 }, { "grad_norm": 0.17740388214588165, "learning_rate": 8.831756977331447e-05, "loss": 0.0268, "step": 7830 }, { "grad_norm": 0.19577540457248688, "learning_rate": 8.828213900875638e-05, "loss": 0.022, "step": 7840 }, { "grad_norm": 0.1603468656539917, "learning_rate": 8.824666172778964e-05, "loss": 0.0204, "step": 7850 }, { "grad_norm": 0.2140481323003769, "learning_rate": 8.821113797352245e-05, "loss": 0.0184, "step": 7860 }, { "grad_norm": 0.18681864440441132, "learning_rate": 8.817556778911956e-05, "loss": 0.0203, "step": 7870 }, { "grad_norm": 0.18044528365135193, "learning_rate": 8.81399512178021e-05, "loss": 0.0214, "step": 7880 }, { "grad_norm": 0.2368387132883072, "learning_rate": 8.810428830284752e-05, "loss": 0.02, "step": 7890 }, { "grad_norm": 0.17725932598114014, "learning_rate": 8.806857908758967e-05, "loss": 0.0175, "step": 7900 }, { "grad_norm": 0.21024276316165924, "learning_rate": 8.80328236154186e-05, "loss": 0.0239, "step": 7910 }, { "grad_norm": 0.2215568870306015, "learning_rate": 8.799702192978056e-05, "loss": 0.019, "step": 7920 }, { "grad_norm": 0.19028519093990326, "learning_rate": 8.7961174074178e-05, "loss": 0.0211, "step": 7930 }, { "grad_norm": 0.14912118017673492, "learning_rate": 8.792528009216942e-05, "loss": 0.0204, "step": 7940 }, { "grad_norm": 0.18664078414440155, "learning_rate": 8.788934002736943e-05, "loss": 0.0246, "step": 7950 }, { "grad_norm": 0.23980988562107086, "learning_rate": 8.785335392344857e-05, "loss": 0.0254, "step": 7960 }, { "grad_norm": 0.17429415881633759, "learning_rate": 8.781732182413336e-05, "loss": 0.0213, "step": 7970 }, { "grad_norm": 0.21017597615718842, "learning_rate": 8.778124377320618e-05, "loss": 0.0284, "step": 7980 }, { "grad_norm": 0.18990711867809296, "learning_rate": 8.774511981450529e-05, "loss": 0.0206, "step": 7990 }, { "grad_norm": 0.19316713511943817, "learning_rate": 8.770894999192467e-05, "loss": 0.0199, "step": 8000 }, { "grad_norm": 0.17406108975410461, "learning_rate": 8.767273434941412e-05, "loss": 0.0199, "step": 8010 }, { "grad_norm": 0.13996517658233643, "learning_rate": 8.7636472930979e-05, "loss": 0.0195, "step": 8020 }, { "grad_norm": 0.20978151261806488, "learning_rate": 8.76001657806804e-05, "loss": 0.0196, "step": 8030 }, { "grad_norm": 0.1951679140329361, "learning_rate": 8.756381294263491e-05, "loss": 0.0173, "step": 8040 }, { "grad_norm": 0.1744164079427719, "learning_rate": 8.752741446101464e-05, "loss": 0.0185, "step": 8050 }, { "grad_norm": 0.15912789106369019, "learning_rate": 8.749097038004722e-05, "loss": 0.0203, "step": 8060 }, { "grad_norm": 0.11596682667732239, "learning_rate": 8.745448074401561e-05, "loss": 0.019, "step": 8070 }, { "grad_norm": 0.15478064119815826, "learning_rate": 8.741794559725818e-05, "loss": 0.0192, "step": 8080 }, { "grad_norm": 0.20767928659915924, "learning_rate": 8.738136498416856e-05, "loss": 0.0197, "step": 8090 }, { "grad_norm": 0.14943239092826843, "learning_rate": 8.734473894919564e-05, "loss": 0.0175, "step": 8100 }, { "grad_norm": 0.1986924707889557, "learning_rate": 8.730806753684353e-05, "loss": 0.0205, "step": 8110 }, { "grad_norm": 0.2328224927186966, "learning_rate": 8.727135079167144e-05, "loss": 0.0207, "step": 8120 }, { "grad_norm": 0.18529537320137024, "learning_rate": 8.723458875829368e-05, "loss": 0.0176, "step": 8130 }, { "grad_norm": 0.19131989777088165, "learning_rate": 8.719778148137958e-05, "loss": 0.0207, "step": 8140 }, { "grad_norm": 0.20103728771209717, "learning_rate": 8.716092900565346e-05, "loss": 0.019, "step": 8150 }, { "grad_norm": 0.16092371940612793, "learning_rate": 8.712403137589456e-05, "loss": 0.0186, "step": 8160 }, { "grad_norm": 0.20263512432575226, "learning_rate": 8.708708863693697e-05, "loss": 0.0186, "step": 8170 }, { "grad_norm": 0.17780674993991852, "learning_rate": 8.70501008336696e-05, "loss": 0.021, "step": 8180 }, { "grad_norm": 0.24662983417510986, "learning_rate": 8.701306801103611e-05, "loss": 0.019, "step": 8190 }, { "grad_norm": 0.14739036560058594, "learning_rate": 8.697599021403489e-05, "loss": 0.0194, "step": 8200 }, { "grad_norm": 0.15499988198280334, "learning_rate": 8.693886748771896e-05, "loss": 0.0206, "step": 8210 }, { "grad_norm": 0.18738755583763123, "learning_rate": 8.690169987719593e-05, "loss": 0.0199, "step": 8220 }, { "grad_norm": 0.20156879723072052, "learning_rate": 8.686448742762791e-05, "loss": 0.0195, "step": 8230 }, { "grad_norm": 0.18990211188793182, "learning_rate": 8.682723018423161e-05, "loss": 0.017, "step": 8240 }, { "grad_norm": 0.1515296846628189, "learning_rate": 8.678992819227804e-05, "loss": 0.0164, "step": 8250 }, { "grad_norm": 0.1356753259897232, "learning_rate": 8.675258149709265e-05, "loss": 0.0178, "step": 8260 }, { "grad_norm": 0.24106022715568542, "learning_rate": 8.67151901440552e-05, "loss": 0.0244, "step": 8270 }, { "grad_norm": 0.1422320008277893, "learning_rate": 8.667775417859971e-05, "loss": 0.0187, "step": 8280 }, { "grad_norm": 0.2098095566034317, "learning_rate": 8.664027364621441e-05, "loss": 0.0213, "step": 8290 }, { "grad_norm": 0.20014576613903046, "learning_rate": 8.660274859244168e-05, "loss": 0.019, "step": 8300 }, { "grad_norm": 0.18342290818691254, "learning_rate": 8.656517906287799e-05, "loss": 0.0235, "step": 8310 }, { "grad_norm": 0.15552179515361786, "learning_rate": 8.652756510317387e-05, "loss": 0.0214, "step": 8320 }, { "grad_norm": 0.23454642295837402, "learning_rate": 8.648990675903382e-05, "loss": 0.0211, "step": 8330 }, { "grad_norm": 0.17904022336006165, "learning_rate": 8.64522040762163e-05, "loss": 0.0195, "step": 8340 }, { "grad_norm": 0.17843596637248993, "learning_rate": 8.64144571005336e-05, "loss": 0.0206, "step": 8350 }, { "grad_norm": 0.13968932628631592, "learning_rate": 8.637666587785184e-05, "loss": 0.0174, "step": 8360 }, { "grad_norm": 0.15208643674850464, "learning_rate": 8.633883045409096e-05, "loss": 0.016, "step": 8370 }, { "grad_norm": 0.185285747051239, "learning_rate": 8.630095087522457e-05, "loss": 0.0184, "step": 8380 }, { "grad_norm": 0.12871551513671875, "learning_rate": 8.62630271872799e-05, "loss": 0.0183, "step": 8390 }, { "grad_norm": 0.15651968121528625, "learning_rate": 8.62250594363378e-05, "loss": 0.0166, "step": 8400 }, { "grad_norm": 0.23154032230377197, "learning_rate": 8.61870476685327e-05, "loss": 0.0189, "step": 8410 }, { "grad_norm": 0.15319015085697174, "learning_rate": 8.614899193005248e-05, "loss": 0.018, "step": 8420 }, { "grad_norm": 0.19146592915058136, "learning_rate": 8.611089226713842e-05, "loss": 0.0173, "step": 8430 }, { "grad_norm": 0.24015076458454132, "learning_rate": 8.607274872608521e-05, "loss": 0.0152, "step": 8440 }, { "grad_norm": 0.1892993301153183, "learning_rate": 8.603456135324088e-05, "loss": 0.017, "step": 8450 }, { "grad_norm": 0.14253613352775574, "learning_rate": 8.599633019500665e-05, "loss": 0.0149, "step": 8460 }, { "grad_norm": 0.1766420602798462, "learning_rate": 8.595805529783702e-05, "loss": 0.0164, "step": 8470 }, { "grad_norm": 0.20875249803066254, "learning_rate": 8.591973670823958e-05, "loss": 0.0222, "step": 8480 }, { "grad_norm": 0.20444685220718384, "learning_rate": 8.588137447277503e-05, "loss": 0.0199, "step": 8490 }, { "grad_norm": 0.2152489572763443, "learning_rate": 8.584296863805709e-05, "loss": 0.0163, "step": 8500 }, { "grad_norm": 0.18338677287101746, "learning_rate": 8.580451925075248e-05, "loss": 0.0171, "step": 8510 }, { "grad_norm": 0.175130695104599, "learning_rate": 8.576602635758085e-05, "loss": 0.0167, "step": 8520 }, { "grad_norm": 0.12490704655647278, "learning_rate": 8.572749000531469e-05, "loss": 0.0138, "step": 8530 }, { "grad_norm": 0.15031969547271729, "learning_rate": 8.568891024077926e-05, "loss": 0.0168, "step": 8540 }, { "grad_norm": 0.1513349711894989, "learning_rate": 8.565028711085265e-05, "loss": 0.0184, "step": 8550 }, { "grad_norm": 0.11322110891342163, "learning_rate": 8.561162066246561e-05, "loss": 0.0163, "step": 8560 }, { "grad_norm": 0.15831130743026733, "learning_rate": 8.557291094260152e-05, "loss": 0.0179, "step": 8570 }, { "grad_norm": 0.1503530889749527, "learning_rate": 8.553415799829629e-05, "loss": 0.0184, "step": 8580 }, { "grad_norm": 0.18862949311733246, "learning_rate": 8.549536187663847e-05, "loss": 0.016, "step": 8590 }, { "grad_norm": 0.19520673155784607, "learning_rate": 8.545652262476897e-05, "loss": 0.0219, "step": 8600 }, { "grad_norm": 0.2226811796426773, "learning_rate": 8.541764028988115e-05, "loss": 0.0201, "step": 8610 }, { "grad_norm": 0.19885985553264618, "learning_rate": 8.537871491922071e-05, "loss": 0.0214, "step": 8620 }, { "grad_norm": 0.14974670112133026, "learning_rate": 8.533974656008566e-05, "loss": 0.0188, "step": 8630 }, { "grad_norm": 0.17418095469474792, "learning_rate": 8.530073525982621e-05, "loss": 0.0187, "step": 8640 }, { "grad_norm": 0.148734912276268, "learning_rate": 8.526168106584476e-05, "loss": 0.0228, "step": 8650 }, { "grad_norm": 0.15583516657352448, "learning_rate": 8.522258402559586e-05, "loss": 0.0191, "step": 8660 }, { "grad_norm": 0.17789092659950256, "learning_rate": 8.518344418658613e-05, "loss": 0.0189, "step": 8670 }, { "grad_norm": 0.15345166623592377, "learning_rate": 8.51442615963741e-05, "loss": 0.0233, "step": 8680 }, { "grad_norm": 0.15245819091796875, "learning_rate": 8.510503630257034e-05, "loss": 0.0185, "step": 8690 }, { "grad_norm": 0.14769645035266876, "learning_rate": 8.50657683528373e-05, "loss": 0.019, "step": 8700 }, { "grad_norm": 0.17710916697978973, "learning_rate": 8.502645779488922e-05, "loss": 0.0162, "step": 8710 }, { "grad_norm": 0.2001744508743286, "learning_rate": 8.498710467649213e-05, "loss": 0.0177, "step": 8720 }, { "grad_norm": 0.21207177639007568, "learning_rate": 8.494770904546381e-05, "loss": 0.0204, "step": 8730 }, { "grad_norm": 0.2165505290031433, "learning_rate": 8.490827094967363e-05, "loss": 0.0185, "step": 8740 }, { "grad_norm": 0.18312351405620575, "learning_rate": 8.486879043704263e-05, "loss": 0.0201, "step": 8750 }, { "grad_norm": 0.17763440310955048, "learning_rate": 8.482926755554332e-05, "loss": 0.0183, "step": 8760 }, { "grad_norm": 0.16518349945545197, "learning_rate": 8.478970235319975e-05, "loss": 0.0165, "step": 8770 }, { "grad_norm": 0.12790963053703308, "learning_rate": 8.475009487808737e-05, "loss": 0.0191, "step": 8780 }, { "grad_norm": 0.21840934455394745, "learning_rate": 8.471044517833298e-05, "loss": 0.02, "step": 8790 }, { "grad_norm": 0.16252566874027252, "learning_rate": 8.467075330211474e-05, "loss": 0.0188, "step": 8800 }, { "grad_norm": 0.16122661530971527, "learning_rate": 8.463101929766198e-05, "loss": 0.0218, "step": 8810 }, { "grad_norm": 0.1732473224401474, "learning_rate": 8.459124321325528e-05, "loss": 0.0188, "step": 8820 }, { "grad_norm": 0.2225600779056549, "learning_rate": 8.455142509722634e-05, "loss": 0.0173, "step": 8830 }, { "grad_norm": 0.16469556093215942, "learning_rate": 8.451156499795792e-05, "loss": 0.0152, "step": 8840 }, { "grad_norm": 0.23507104814052582, "learning_rate": 8.447166296388381e-05, "loss": 0.0193, "step": 8850 }, { "grad_norm": 0.16938462853431702, "learning_rate": 8.443171904348872e-05, "loss": 0.0187, "step": 8860 }, { "grad_norm": 0.1656828224658966, "learning_rate": 8.43917332853083e-05, "loss": 0.0163, "step": 8870 }, { "grad_norm": 0.15269413590431213, "learning_rate": 8.435170573792901e-05, "loss": 0.0171, "step": 8880 }, { "grad_norm": 0.17971271276474, "learning_rate": 8.431163644998808e-05, "loss": 0.0165, "step": 8890 }, { "grad_norm": 0.13559399545192719, "learning_rate": 8.42715254701735e-05, "loss": 0.0222, "step": 8900 }, { "grad_norm": 0.14407405257225037, "learning_rate": 8.42313728472239e-05, "loss": 0.0166, "step": 8910 }, { "grad_norm": 0.13460685312747955, "learning_rate": 8.419117862992844e-05, "loss": 0.0168, "step": 8920 }, { "grad_norm": 0.17215795814990997, "learning_rate": 8.415094286712694e-05, "loss": 0.0161, "step": 8930 }, { "grad_norm": 0.1673564612865448, "learning_rate": 8.411066560770965e-05, "loss": 0.0174, "step": 8940 }, { "grad_norm": 0.1827922761440277, "learning_rate": 8.407034690061721e-05, "loss": 0.0204, "step": 8950 }, { "grad_norm": 0.2238367199897766, "learning_rate": 8.402998679484067e-05, "loss": 0.0197, "step": 8960 }, { "grad_norm": 0.17654259502887726, "learning_rate": 8.398958533942135e-05, "loss": 0.017, "step": 8970 }, { "grad_norm": 0.14827457070350647, "learning_rate": 8.394914258345083e-05, "loss": 0.0203, "step": 8980 }, { "grad_norm": 0.1583746075630188, "learning_rate": 8.390865857607089e-05, "loss": 0.018, "step": 8990 }, { "grad_norm": 0.17074327170848846, "learning_rate": 8.386813336647339e-05, "loss": 0.0201, "step": 9000 }, { "grad_norm": 0.13477647304534912, "learning_rate": 8.38275670039003e-05, "loss": 0.0168, "step": 9010 }, { "grad_norm": 0.14958685636520386, "learning_rate": 8.378695953764356e-05, "loss": 0.0165, "step": 9020 }, { "grad_norm": 0.1930977702140808, "learning_rate": 8.374631101704509e-05, "loss": 0.0195, "step": 9030 }, { "grad_norm": 0.2033226191997528, "learning_rate": 8.370562149149665e-05, "loss": 0.0163, "step": 9040 }, { "grad_norm": 0.12579543888568878, "learning_rate": 8.366489101043989e-05, "loss": 0.0158, "step": 9050 }, { "grad_norm": 0.1525481939315796, "learning_rate": 8.362411962336614e-05, "loss": 0.0163, "step": 9060 }, { "grad_norm": 0.14942318201065063, "learning_rate": 8.358330737981651e-05, "loss": 0.0152, "step": 9070 }, { "grad_norm": 0.15851037204265594, "learning_rate": 8.35424543293817e-05, "loss": 0.016, "step": 9080 }, { "grad_norm": 0.12418448179960251, "learning_rate": 8.350156052170206e-05, "loss": 0.0165, "step": 9090 }, { "grad_norm": 0.1900465488433838, "learning_rate": 8.346062600646738e-05, "loss": 0.0184, "step": 9100 }, { "grad_norm": 0.15701457858085632, "learning_rate": 8.341965083341697e-05, "loss": 0.0171, "step": 9110 }, { "grad_norm": 0.22644421458244324, "learning_rate": 8.337863505233953e-05, "loss": 0.017, "step": 9120 }, { "grad_norm": 0.1112697571516037, "learning_rate": 8.33375787130731e-05, "loss": 0.0133, "step": 9130 }, { "grad_norm": 0.15834389626979828, "learning_rate": 8.3296481865505e-05, "loss": 0.0164, "step": 9140 }, { "grad_norm": 0.16640350222587585, "learning_rate": 8.325534455957181e-05, "loss": 0.018, "step": 9150 }, { "grad_norm": 0.1528049260377884, "learning_rate": 8.321416684525917e-05, "loss": 0.0166, "step": 9160 }, { "grad_norm": 0.16163884103298187, "learning_rate": 8.317294877260193e-05, "loss": 0.0208, "step": 9170 }, { "grad_norm": 0.2087302803993225, "learning_rate": 8.313169039168394e-05, "loss": 0.0182, "step": 9180 }, { "grad_norm": 0.180209681391716, "learning_rate": 8.3090391752638e-05, "loss": 0.0144, "step": 9190 }, { "grad_norm": 0.16800659894943237, "learning_rate": 8.304905290564586e-05, "loss": 0.0193, "step": 9200 }, { "grad_norm": 0.18370914459228516, "learning_rate": 8.300767390093813e-05, "loss": 0.0172, "step": 9210 }, { "grad_norm": 0.1277659833431244, "learning_rate": 8.296625478879417e-05, "loss": 0.0226, "step": 9220 }, { "grad_norm": 0.18925487995147705, "learning_rate": 8.292479561954214e-05, "loss": 0.0201, "step": 9230 }, { "grad_norm": 0.160203218460083, "learning_rate": 8.288329644355883e-05, "loss": 0.0172, "step": 9240 }, { "grad_norm": 0.20062503218650818, "learning_rate": 8.284175731126965e-05, "loss": 0.0197, "step": 9250 }, { "grad_norm": 0.14890024065971375, "learning_rate": 8.280017827314854e-05, "loss": 0.0178, "step": 9260 }, { "grad_norm": 0.15183892846107483, "learning_rate": 8.2758559379718e-05, "loss": 0.0199, "step": 9270 }, { "grad_norm": 0.13133327662944794, "learning_rate": 8.271690068154887e-05, "loss": 0.0158, "step": 9280 }, { "grad_norm": 0.15951848030090332, "learning_rate": 8.26752022292604e-05, "loss": 0.017, "step": 9290 }, { "grad_norm": 0.14169055223464966, "learning_rate": 8.263346407352018e-05, "loss": 0.0189, "step": 9300 }, { "grad_norm": 0.1700456440448761, "learning_rate": 8.259168626504395e-05, "loss": 0.019, "step": 9310 }, { "grad_norm": 0.14417074620723724, "learning_rate": 8.254986885459572e-05, "loss": 0.0189, "step": 9320 }, { "grad_norm": 0.18141387403011322, "learning_rate": 8.250801189298757e-05, "loss": 0.0206, "step": 9330 }, { "grad_norm": 0.1875029057264328, "learning_rate": 8.246611543107967e-05, "loss": 0.0181, "step": 9340 }, { "grad_norm": 0.14666245877742767, "learning_rate": 8.242417951978018e-05, "loss": 0.0162, "step": 9350 }, { "grad_norm": 0.1684499979019165, "learning_rate": 8.238220421004517e-05, "loss": 0.0212, "step": 9360 }, { "grad_norm": 0.13578934967517853, "learning_rate": 8.234018955287859e-05, "loss": 0.02, "step": 9370 }, { "grad_norm": 0.18024009466171265, "learning_rate": 8.229813559933224e-05, "loss": 0.02, "step": 9380 }, { "grad_norm": 0.1557052582502365, "learning_rate": 8.22560424005056e-05, "loss": 0.0187, "step": 9390 }, { "grad_norm": 0.19705241918563843, "learning_rate": 8.22139100075459e-05, "loss": 0.0147, "step": 9400 }, { "grad_norm": 0.13115115463733673, "learning_rate": 8.217173847164799e-05, "loss": 0.0149, "step": 9410 }, { "grad_norm": 0.20436514914035797, "learning_rate": 8.212952784405422e-05, "loss": 0.016, "step": 9420 }, { "grad_norm": 0.18175333738327026, "learning_rate": 8.208727817605451e-05, "loss": 0.0162, "step": 9430 }, { "grad_norm": 0.15672509372234344, "learning_rate": 8.204498951898618e-05, "loss": 0.02, "step": 9440 }, { "grad_norm": 0.11858717352151871, "learning_rate": 8.200266192423395e-05, "loss": 0.0154, "step": 9450 }, { "grad_norm": 0.1615431308746338, "learning_rate": 8.196029544322981e-05, "loss": 0.0164, "step": 9460 }, { "grad_norm": 0.15926645696163177, "learning_rate": 8.191789012745305e-05, "loss": 0.0168, "step": 9470 }, { "grad_norm": 0.1226617619395256, "learning_rate": 8.187544602843014e-05, "loss": 0.0152, "step": 9480 }, { "grad_norm": 0.1228213682770729, "learning_rate": 8.183296319773466e-05, "loss": 0.0179, "step": 9490 }, { "grad_norm": 0.17046307027339935, "learning_rate": 8.179044168698721e-05, "loss": 0.0195, "step": 9500 }, { "grad_norm": 0.17297028005123138, "learning_rate": 8.174788154785548e-05, "loss": 0.0175, "step": 9510 }, { "grad_norm": 0.21975557506084442, "learning_rate": 8.170528283205403e-05, "loss": 0.0201, "step": 9520 }, { "grad_norm": 0.19070559740066528, "learning_rate": 8.166264559134435e-05, "loss": 0.0173, "step": 9530 }, { "grad_norm": 0.21318069100379944, "learning_rate": 8.161996987753467e-05, "loss": 0.0212, "step": 9540 }, { "grad_norm": 0.18302153050899506, "learning_rate": 8.157725574248e-05, "loss": 0.0187, "step": 9550 }, { "grad_norm": 0.19001249969005585, "learning_rate": 8.153450323808209e-05, "loss": 0.0189, "step": 9560 }, { "grad_norm": 0.19049763679504395, "learning_rate": 8.149171241628923e-05, "loss": 0.0174, "step": 9570 }, { "grad_norm": 0.18711541593074799, "learning_rate": 8.144888332909631e-05, "loss": 0.0187, "step": 9580 }, { "grad_norm": 0.14924682676792145, "learning_rate": 8.14060160285447e-05, "loss": 0.0146, "step": 9590 }, { "grad_norm": 0.17385224997997284, "learning_rate": 8.136311056672223e-05, "loss": 0.0175, "step": 9600 }, { "grad_norm": 0.19561457633972168, "learning_rate": 8.132016699576308e-05, "loss": 0.021, "step": 9610 }, { "grad_norm": 0.14182467758655548, "learning_rate": 8.127718536784771e-05, "loss": 0.017, "step": 9620 }, { "grad_norm": 0.1822485476732254, "learning_rate": 8.12341657352029e-05, "loss": 0.0176, "step": 9630 }, { "grad_norm": 0.17293494939804077, "learning_rate": 8.119110815010151e-05, "loss": 0.0151, "step": 9640 }, { "grad_norm": 0.17283374071121216, "learning_rate": 8.114801266486261e-05, "loss": 0.0171, "step": 9650 }, { "grad_norm": 0.18012084066867828, "learning_rate": 8.110487933185122e-05, "loss": 0.0166, "step": 9660 }, { "grad_norm": 0.145748108625412, "learning_rate": 8.106170820347847e-05, "loss": 0.0174, "step": 9670 }, { "grad_norm": 0.15392549335956573, "learning_rate": 8.101849933220132e-05, "loss": 0.0173, "step": 9680 }, { "grad_norm": 0.15722332894802094, "learning_rate": 8.097525277052264e-05, "loss": 0.0149, "step": 9690 }, { "grad_norm": 0.18393827974796295, "learning_rate": 8.093196857099106e-05, "loss": 0.0161, "step": 9700 }, { "grad_norm": 0.1566927433013916, "learning_rate": 8.088864678620096e-05, "loss": 0.0178, "step": 9710 }, { "grad_norm": 0.20222510397434235, "learning_rate": 8.084528746879243e-05, "loss": 0.0176, "step": 9720 }, { "grad_norm": 0.21331623196601868, "learning_rate": 8.080189067145108e-05, "loss": 0.0183, "step": 9730 }, { "grad_norm": 0.16727013885974884, "learning_rate": 8.075845644690813e-05, "loss": 0.0163, "step": 9740 }, { "grad_norm": 0.12542326748371124, "learning_rate": 8.07149848479403e-05, "loss": 0.0156, "step": 9750 }, { "grad_norm": 0.15650072693824768, "learning_rate": 8.067147592736962e-05, "loss": 0.0162, "step": 9760 }, { "grad_norm": 0.15538135170936584, "learning_rate": 8.062792973806357e-05, "loss": 0.0164, "step": 9770 }, { "grad_norm": 0.17493726313114166, "learning_rate": 8.058434633293485e-05, "loss": 0.0171, "step": 9780 }, { "grad_norm": 0.1729980856180191, "learning_rate": 8.054072576494143e-05, "loss": 0.0195, "step": 9790 }, { "grad_norm": 0.1555272936820984, "learning_rate": 8.04970680870864e-05, "loss": 0.0171, "step": 9800 }, { "grad_norm": 0.1758786290884018, "learning_rate": 8.045337335241793e-05, "loss": 0.0186, "step": 9810 }, { "grad_norm": 0.1696571707725525, "learning_rate": 8.040964161402932e-05, "loss": 0.0152, "step": 9820 }, { "grad_norm": 0.144557386636734, "learning_rate": 8.036587292505869e-05, "loss": 0.0207, "step": 9830 }, { "grad_norm": 0.17098776996135712, "learning_rate": 8.032206733868912e-05, "loss": 0.0191, "step": 9840 }, { "grad_norm": 0.17922714352607727, "learning_rate": 8.027822490814858e-05, "loss": 0.0196, "step": 9850 }, { "grad_norm": 0.15973518788814545, "learning_rate": 8.023434568670972e-05, "loss": 0.0196, "step": 9860 }, { "grad_norm": 0.15593810379505157, "learning_rate": 8.019042972768992e-05, "loss": 0.016, "step": 9870 }, { "grad_norm": 0.12878966331481934, "learning_rate": 8.014647708445124e-05, "loss": 0.0195, "step": 9880 }, { "grad_norm": 0.14637725055217743, "learning_rate": 8.010248781040027e-05, "loss": 0.0174, "step": 9890 }, { "grad_norm": 0.17673252522945404, "learning_rate": 8.005846195898814e-05, "loss": 0.0166, "step": 9900 }, { "grad_norm": 0.13231772184371948, "learning_rate": 8.00143995837104e-05, "loss": 0.0163, "step": 9910 }, { "grad_norm": 0.1475759893655777, "learning_rate": 7.997030073810698e-05, "loss": 0.0189, "step": 9920 }, { "grad_norm": 0.13605175912380219, "learning_rate": 7.992616547576217e-05, "loss": 0.0137, "step": 9930 }, { "grad_norm": 0.1351393312215805, "learning_rate": 7.988199385030445e-05, "loss": 0.0194, "step": 9940 }, { "grad_norm": 0.2152746021747589, "learning_rate": 7.98377859154065e-05, "loss": 0.0189, "step": 9950 }, { "grad_norm": 0.1466880887746811, "learning_rate": 7.979354172478516e-05, "loss": 0.0169, "step": 9960 }, { "grad_norm": 0.18999730050563812, "learning_rate": 7.974926133220127e-05, "loss": 0.0169, "step": 9970 }, { "grad_norm": 0.16213664412498474, "learning_rate": 7.970494479145967e-05, "loss": 0.0153, "step": 9980 }, { "grad_norm": 0.1596134901046753, "learning_rate": 7.966059215640919e-05, "loss": 0.0173, "step": 9990 }, { "grad_norm": 0.12695173919200897, "learning_rate": 7.961620348094241e-05, "loss": 0.0191, "step": 10000 }, { "grad_norm": 0.1576731652021408, "learning_rate": 7.957177881899577e-05, "loss": 0.0183, "step": 10010 }, { "grad_norm": 0.15233685076236725, "learning_rate": 7.952731822454944e-05, "loss": 0.0172, "step": 10020 }, { "grad_norm": 0.1819033920764923, "learning_rate": 7.948282175162722e-05, "loss": 0.0164, "step": 10030 }, { "grad_norm": 0.21924735605716705, "learning_rate": 7.943828945429652e-05, "loss": 0.0193, "step": 10040 }, { "grad_norm": 0.1372692734003067, "learning_rate": 7.939372138666827e-05, "loss": 0.0151, "step": 10050 }, { "grad_norm": 0.1539071798324585, "learning_rate": 7.934911760289692e-05, "loss": 0.016, "step": 10060 }, { "grad_norm": 0.13403171300888062, "learning_rate": 7.930447815718022e-05, "loss": 0.0168, "step": 10070 }, { "grad_norm": 0.13129118084907532, "learning_rate": 7.925980310375934e-05, "loss": 0.0155, "step": 10080 }, { "grad_norm": 0.10144665092229843, "learning_rate": 7.921509249691865e-05, "loss": 0.0142, "step": 10090 }, { "grad_norm": 0.13893793523311615, "learning_rate": 7.917034639098579e-05, "loss": 0.0146, "step": 10100 }, { "grad_norm": 0.1405605971813202, "learning_rate": 7.912556484033146e-05, "loss": 0.0155, "step": 10110 }, { "grad_norm": 0.15397761762142181, "learning_rate": 7.908074789936952e-05, "loss": 0.0147, "step": 10120 }, { "grad_norm": 0.2006307691335678, "learning_rate": 7.903589562255673e-05, "loss": 0.0145, "step": 10130 }, { "grad_norm": 0.13343793153762817, "learning_rate": 7.899100806439285e-05, "loss": 0.0146, "step": 10140 }, { "grad_norm": 0.12620557844638824, "learning_rate": 7.894608527942049e-05, "loss": 0.0158, "step": 10150 }, { "grad_norm": 0.17013972997665405, "learning_rate": 7.89011273222251e-05, "loss": 0.015, "step": 10160 }, { "grad_norm": 0.1736205369234085, "learning_rate": 7.885613424743482e-05, "loss": 0.0166, "step": 10170 }, { "grad_norm": 0.19461636245250702, "learning_rate": 7.881110610972044e-05, "loss": 0.0165, "step": 10180 }, { "grad_norm": 0.20558954775333405, "learning_rate": 7.876604296379544e-05, "loss": 0.0164, "step": 10190 }, { "grad_norm": 0.17039096355438232, "learning_rate": 7.872094486441579e-05, "loss": 0.0165, "step": 10200 }, { "grad_norm": 0.17398861050605774, "learning_rate": 7.867581186637991e-05, "loss": 0.0174, "step": 10210 }, { "grad_norm": 0.16876599192619324, "learning_rate": 7.863064402452866e-05, "loss": 0.0164, "step": 10220 }, { "grad_norm": 0.1418675035238266, "learning_rate": 7.858544139374524e-05, "loss": 0.0168, "step": 10230 }, { "grad_norm": 0.14635999500751495, "learning_rate": 7.854020402895508e-05, "loss": 0.0143, "step": 10240 }, { "grad_norm": 0.16767016053199768, "learning_rate": 7.849493198512587e-05, "loss": 0.0156, "step": 10250 }, { "grad_norm": 0.13987155258655548, "learning_rate": 7.84496253172674e-05, "loss": 0.017, "step": 10260 }, { "grad_norm": 0.12955763936042786, "learning_rate": 7.840428408043155e-05, "loss": 0.0162, "step": 10270 }, { "grad_norm": 0.2022266387939453, "learning_rate": 7.835890832971218e-05, "loss": 0.023, "step": 10280 }, { "grad_norm": 0.16967825591564178, "learning_rate": 7.831349812024512e-05, "loss": 0.0189, "step": 10290 }, { "grad_norm": 0.18302561342716217, "learning_rate": 7.826805350720807e-05, "loss": 0.0162, "step": 10300 }, { "grad_norm": 0.15891887247562408, "learning_rate": 7.822257454582049e-05, "loss": 0.0155, "step": 10310 }, { "grad_norm": 0.15891647338867188, "learning_rate": 7.817706129134363e-05, "loss": 0.0193, "step": 10320 }, { "grad_norm": 0.15116211771965027, "learning_rate": 7.813151379908036e-05, "loss": 0.0149, "step": 10330 }, { "grad_norm": 0.1312730312347412, "learning_rate": 7.808593212437523e-05, "loss": 0.0148, "step": 10340 }, { "grad_norm": 0.15143375098705292, "learning_rate": 7.804031632261421e-05, "loss": 0.0158, "step": 10350 }, { "grad_norm": 0.1251329481601715, "learning_rate": 7.799466644922484e-05, "loss": 0.0134, "step": 10360 }, { "grad_norm": 0.14516355097293854, "learning_rate": 7.794898255967602e-05, "loss": 0.0179, "step": 10370 }, { "grad_norm": 0.17855021357536316, "learning_rate": 7.790326470947795e-05, "loss": 0.0157, "step": 10380 }, { "grad_norm": 0.19353042542934418, "learning_rate": 7.785751295418217e-05, "loss": 0.0154, "step": 10390 }, { "grad_norm": 0.16677510738372803, "learning_rate": 7.781172734938136e-05, "loss": 0.0179, "step": 10400 }, { "grad_norm": 0.13778038322925568, "learning_rate": 7.776590795070933e-05, "loss": 0.0159, "step": 10410 }, { "grad_norm": 0.15052427351474762, "learning_rate": 7.772005481384099e-05, "loss": 0.0176, "step": 10420 }, { "grad_norm": 0.15833227336406708, "learning_rate": 7.767416799449223e-05, "loss": 0.0141, "step": 10430 }, { "grad_norm": 0.14772963523864746, "learning_rate": 7.762824754841984e-05, "loss": 0.017, "step": 10440 }, { "grad_norm": 0.10708561539649963, "learning_rate": 7.758229353142152e-05, "loss": 0.0146, "step": 10450 }, { "grad_norm": 0.17031414806842804, "learning_rate": 7.753630599933572e-05, "loss": 0.015, "step": 10460 }, { "grad_norm": 0.12879209220409393, "learning_rate": 7.749028500804161e-05, "loss": 0.0161, "step": 10470 }, { "grad_norm": 0.1105239987373352, "learning_rate": 7.744423061345906e-05, "loss": 0.0155, "step": 10480 }, { "grad_norm": 0.14285174012184143, "learning_rate": 7.73981428715485e-05, "loss": 0.0175, "step": 10490 }, { "grad_norm": 0.20577920973300934, "learning_rate": 7.735202183831085e-05, "loss": 0.0147, "step": 10500 }, { "grad_norm": 0.12677986919879913, "learning_rate": 7.730586756978757e-05, "loss": 0.0193, "step": 10510 }, { "grad_norm": 0.13733656704425812, "learning_rate": 7.72596801220604e-05, "loss": 0.0179, "step": 10520 }, { "grad_norm": 0.15463125705718994, "learning_rate": 7.721345955125147e-05, "loss": 0.0183, "step": 10530 }, { "grad_norm": 0.12580810487270355, "learning_rate": 7.71672059135231e-05, "loss": 0.0143, "step": 10540 }, { "grad_norm": 0.14246700704097748, "learning_rate": 7.712091926507787e-05, "loss": 0.0144, "step": 10550 }, { "grad_norm": 0.1542811542749405, "learning_rate": 7.70745996621584e-05, "loss": 0.0161, "step": 10560 }, { "grad_norm": 0.1362553983926773, "learning_rate": 7.702824716104735e-05, "loss": 0.0172, "step": 10570 }, { "grad_norm": 0.17228937149047852, "learning_rate": 7.698186181806743e-05, "loss": 0.0147, "step": 10580 }, { "grad_norm": 0.15507374703884125, "learning_rate": 7.693544368958116e-05, "loss": 0.0172, "step": 10590 }, { "grad_norm": 0.178826242685318, "learning_rate": 7.688899283199096e-05, "loss": 0.0156, "step": 10600 }, { "grad_norm": 0.18902835249900818, "learning_rate": 7.684250930173901e-05, "loss": 0.0149, "step": 10610 }, { "grad_norm": 0.18242670595645905, "learning_rate": 7.679599315530718e-05, "loss": 0.0151, "step": 10620 }, { "grad_norm": 0.17494769394397736, "learning_rate": 7.674944444921695e-05, "loss": 0.0155, "step": 10630 }, { "grad_norm": 0.1760735660791397, "learning_rate": 7.670286324002944e-05, "loss": 0.0144, "step": 10640 }, { "grad_norm": 0.17471027374267578, "learning_rate": 7.665624958434514e-05, "loss": 0.013, "step": 10650 }, { "grad_norm": 0.20825478434562683, "learning_rate": 7.66096035388041e-05, "loss": 0.0168, "step": 10660 }, { "grad_norm": 0.1576198786497116, "learning_rate": 7.656292516008563e-05, "loss": 0.0157, "step": 10670 }, { "grad_norm": 0.13856124877929688, "learning_rate": 7.651621450490837e-05, "loss": 0.018, "step": 10680 }, { "grad_norm": 0.20667997002601624, "learning_rate": 7.646947163003017e-05, "loss": 0.0162, "step": 10690 }, { "grad_norm": 0.1312311887741089, "learning_rate": 7.642269659224803e-05, "loss": 0.0154, "step": 10700 }, { "grad_norm": 0.1647927463054657, "learning_rate": 7.637588944839803e-05, "loss": 0.0182, "step": 10710 }, { "grad_norm": 0.1906055361032486, "learning_rate": 7.632905025535528e-05, "loss": 0.0147, "step": 10720 }, { "grad_norm": 0.14057046175003052, "learning_rate": 7.628217907003378e-05, "loss": 0.0143, "step": 10730 }, { "grad_norm": 0.19318006932735443, "learning_rate": 7.623527594938649e-05, "loss": 0.0166, "step": 10740 }, { "grad_norm": 0.15807001292705536, "learning_rate": 7.618834095040509e-05, "loss": 0.0158, "step": 10750 }, { "grad_norm": 0.1434830278158188, "learning_rate": 7.614137413012e-05, "loss": 0.017, "step": 10760 }, { "grad_norm": 0.16429640352725983, "learning_rate": 7.609437554560042e-05, "loss": 0.0154, "step": 10770 }, { "grad_norm": 0.14951099455356598, "learning_rate": 7.604734525395398e-05, "loss": 0.0141, "step": 10780 }, { "grad_norm": 0.12771296501159668, "learning_rate": 7.600028331232699e-05, "loss": 0.0177, "step": 10790 }, { "grad_norm": 0.17279879748821259, "learning_rate": 7.595318977790408e-05, "loss": 0.0155, "step": 10800 }, { "grad_norm": 0.16328291594982147, "learning_rate": 7.590606470790836e-05, "loss": 0.015, "step": 10810 }, { "grad_norm": 0.14049586653709412, "learning_rate": 7.585890815960125e-05, "loss": 0.0181, "step": 10820 }, { "grad_norm": 0.1585731953382492, "learning_rate": 7.581172019028238e-05, "loss": 0.0174, "step": 10830 }, { "grad_norm": 0.19293445348739624, "learning_rate": 7.576450085728959e-05, "loss": 0.0172, "step": 10840 }, { "grad_norm": 0.1369207203388214, "learning_rate": 7.571725021799885e-05, "loss": 0.0149, "step": 10850 }, { "grad_norm": 0.14955170452594757, "learning_rate": 7.566996832982409e-05, "loss": 0.0153, "step": 10860 }, { "grad_norm": 0.1861666887998581, "learning_rate": 7.56226552502173e-05, "loss": 0.015, "step": 10870 }, { "grad_norm": 0.14656293392181396, "learning_rate": 7.557531103666833e-05, "loss": 0.0151, "step": 10880 }, { "grad_norm": 0.1546252816915512, "learning_rate": 7.552793574670485e-05, "loss": 0.0144, "step": 10890 }, { "grad_norm": 0.1577860713005066, "learning_rate": 7.54805294378923e-05, "loss": 0.0149, "step": 10900 }, { "grad_norm": 0.16198919713497162, "learning_rate": 7.543309216783384e-05, "loss": 0.0145, "step": 10910 }, { "grad_norm": 0.13103517889976501, "learning_rate": 7.53856239941702e-05, "loss": 0.0138, "step": 10920 }, { "grad_norm": 0.14876404404640198, "learning_rate": 7.533812497457972e-05, "loss": 0.0147, "step": 10930 }, { "grad_norm": 0.16990211606025696, "learning_rate": 7.529059516677814e-05, "loss": 0.0139, "step": 10940 }, { "grad_norm": 0.1636996865272522, "learning_rate": 7.524303462851872e-05, "loss": 0.015, "step": 10950 }, { "grad_norm": 0.1450711190700531, "learning_rate": 7.519544341759192e-05, "loss": 0.0142, "step": 10960 }, { "grad_norm": 0.14677101373672485, "learning_rate": 7.514782159182562e-05, "loss": 0.0154, "step": 10970 }, { "grad_norm": 0.1463291049003601, "learning_rate": 7.510016920908481e-05, "loss": 0.0157, "step": 10980 }, { "grad_norm": 0.18587912619113922, "learning_rate": 7.505248632727158e-05, "loss": 0.016, "step": 10990 }, { "grad_norm": 0.20752379298210144, "learning_rate": 7.50047730043252e-05, "loss": 0.0175, "step": 11000 }, { "grad_norm": 0.149189755320549, "learning_rate": 7.495702929822183e-05, "loss": 0.0135, "step": 11010 }, { "grad_norm": 0.12091613560914993, "learning_rate": 7.490925526697455e-05, "loss": 0.0154, "step": 11020 }, { "grad_norm": 0.1546352654695511, "learning_rate": 7.486145096863334e-05, "loss": 0.0171, "step": 11030 }, { "grad_norm": 0.12371781468391418, "learning_rate": 7.481361646128491e-05, "loss": 0.0127, "step": 11040 }, { "grad_norm": 0.18197447061538696, "learning_rate": 7.476575180305271e-05, "loss": 0.0145, "step": 11050 }, { "grad_norm": 0.14250914752483368, "learning_rate": 7.471785705209682e-05, "loss": 0.0132, "step": 11060 }, { "grad_norm": 0.17896141111850739, "learning_rate": 7.466993226661387e-05, "loss": 0.017, "step": 11070 }, { "grad_norm": 0.1633039265871048, "learning_rate": 7.462197750483699e-05, "loss": 0.0162, "step": 11080 }, { "grad_norm": 0.2075720727443695, "learning_rate": 7.457399282503574e-05, "loss": 0.0162, "step": 11090 }, { "grad_norm": 0.15906935930252075, "learning_rate": 7.452597828551604e-05, "loss": 0.0162, "step": 11100 }, { "grad_norm": 0.11626747995615005, "learning_rate": 7.447793394462006e-05, "loss": 0.0192, "step": 11110 }, { "grad_norm": 0.1463293731212616, "learning_rate": 7.442985986072624e-05, "loss": 0.0156, "step": 11120 }, { "grad_norm": 0.16984815895557404, "learning_rate": 7.438175609224908e-05, "loss": 0.0148, "step": 11130 }, { "grad_norm": 0.15094295144081116, "learning_rate": 7.433362269763924e-05, "loss": 0.0173, "step": 11140 }, { "grad_norm": 0.1686314046382904, "learning_rate": 7.428545973538329e-05, "loss": 0.0182, "step": 11150 }, { "grad_norm": 0.15050043165683746, "learning_rate": 7.42372672640038e-05, "loss": 0.0185, "step": 11160 }, { "grad_norm": 0.1391567885875702, "learning_rate": 7.418904534205917e-05, "loss": 0.0159, "step": 11170 }, { "grad_norm": 0.11792457103729248, "learning_rate": 7.414079402814356e-05, "loss": 0.0143, "step": 11180 }, { "grad_norm": 0.11837995797395706, "learning_rate": 7.409251338088687e-05, "loss": 0.0184, "step": 11190 }, { "grad_norm": 0.13303539156913757, "learning_rate": 7.404420345895467e-05, "loss": 0.0195, "step": 11200 }, { "grad_norm": 0.18187622725963593, "learning_rate": 7.399586432104804e-05, "loss": 0.0195, "step": 11210 }, { "grad_norm": 0.19466258585453033, "learning_rate": 7.394749602590358e-05, "loss": 0.0161, "step": 11220 }, { "grad_norm": 0.16977420449256897, "learning_rate": 7.389909863229336e-05, "loss": 0.0182, "step": 11230 }, { "grad_norm": 0.15293395519256592, "learning_rate": 7.385067219902477e-05, "loss": 0.0145, "step": 11240 }, { "grad_norm": 0.16522392630577087, "learning_rate": 7.380221678494049e-05, "loss": 0.0198, "step": 11250 }, { "grad_norm": 0.162870392203331, "learning_rate": 7.37537324489184e-05, "loss": 0.0142, "step": 11260 }, { "grad_norm": 0.19071362912654877, "learning_rate": 7.370521924987155e-05, "loss": 0.0149, "step": 11270 }, { "grad_norm": 0.20187543332576752, "learning_rate": 7.365667724674806e-05, "loss": 0.0142, "step": 11280 }, { "grad_norm": 0.21472644805908203, "learning_rate": 7.360810649853105e-05, "loss": 0.0169, "step": 11290 }, { "grad_norm": 0.15106754004955292, "learning_rate": 7.355950706423853e-05, "loss": 0.0159, "step": 11300 }, { "grad_norm": 0.15346235036849976, "learning_rate": 7.351087900292342e-05, "loss": 0.0143, "step": 11310 }, { "grad_norm": 0.12961652874946594, "learning_rate": 7.346222237367339e-05, "loss": 0.0139, "step": 11320 }, { "grad_norm": 0.13104784488677979, "learning_rate": 7.341353723561081e-05, "loss": 0.0173, "step": 11330 }, { "grad_norm": 0.1532108634710312, "learning_rate": 7.336482364789277e-05, "loss": 0.0178, "step": 11340 }, { "grad_norm": 0.16394267976284027, "learning_rate": 7.331608166971082e-05, "loss": 0.0144, "step": 11350 }, { "grad_norm": 0.15703049302101135, "learning_rate": 7.326731136029108e-05, "loss": 0.0142, "step": 11360 }, { "grad_norm": 0.20859843492507935, "learning_rate": 7.321851277889408e-05, "loss": 0.017, "step": 11370 }, { "grad_norm": 0.10460738837718964, "learning_rate": 7.316968598481469e-05, "loss": 0.0147, "step": 11380 }, { "grad_norm": 0.18429197371006012, "learning_rate": 7.312083103738206e-05, "loss": 0.0137, "step": 11390 }, { "grad_norm": 0.1391548365354538, "learning_rate": 7.307194799595958e-05, "loss": 0.0138, "step": 11400 }, { "grad_norm": 0.13803179562091827, "learning_rate": 7.302303691994473e-05, "loss": 0.0146, "step": 11410 }, { "grad_norm": 0.1712299883365631, "learning_rate": 7.29740978687691e-05, "loss": 0.0155, "step": 11420 }, { "grad_norm": 0.14177455008029938, "learning_rate": 7.292513090189825e-05, "loss": 0.0142, "step": 11430 }, { "grad_norm": 0.13595370948314667, "learning_rate": 7.287613607883163e-05, "loss": 0.0138, "step": 11440 }, { "grad_norm": 0.15623168647289276, "learning_rate": 7.282711345910263e-05, "loss": 0.0149, "step": 11450 }, { "grad_norm": 0.17877443134784698, "learning_rate": 7.27780631022783e-05, "loss": 0.018, "step": 11460 }, { "grad_norm": 0.15840503573417664, "learning_rate": 7.272898506795948e-05, "loss": 0.0166, "step": 11470 }, { "grad_norm": 0.13009849190711975, "learning_rate": 7.267987941578058e-05, "loss": 0.0146, "step": 11480 }, { "grad_norm": 0.10580310225486755, "learning_rate": 7.263074620540962e-05, "loss": 0.0126, "step": 11490 }, { "grad_norm": 0.12560226023197174, "learning_rate": 7.25815854965481e-05, "loss": 0.017, "step": 11500 }, { "grad_norm": 0.15345901250839233, "learning_rate": 7.253239734893089e-05, "loss": 0.0149, "step": 11510 }, { "grad_norm": 0.15255114436149597, "learning_rate": 7.248318182232623e-05, "loss": 0.0155, "step": 11520 }, { "grad_norm": 0.13418224453926086, "learning_rate": 7.243393897653565e-05, "loss": 0.0179, "step": 11530 }, { "grad_norm": 0.19078662991523743, "learning_rate": 7.23846688713938e-05, "loss": 0.014, "step": 11540 }, { "grad_norm": 0.15991859138011932, "learning_rate": 7.233537156676854e-05, "loss": 0.0151, "step": 11550 }, { "grad_norm": 0.13785888254642487, "learning_rate": 7.228604712256076e-05, "loss": 0.0149, "step": 11560 }, { "grad_norm": 0.15634296834468842, "learning_rate": 7.223669559870426e-05, "loss": 0.0142, "step": 11570 }, { "grad_norm": 0.1302725225687027, "learning_rate": 7.218731705516585e-05, "loss": 0.0123, "step": 11580 }, { "grad_norm": 0.13888607919216156, "learning_rate": 7.21379115519451e-05, "loss": 0.0164, "step": 11590 }, { "grad_norm": 0.1503426730632782, "learning_rate": 7.20884791490743e-05, "loss": 0.0132, "step": 11600 }, { "grad_norm": 0.11740218847990036, "learning_rate": 7.203901990661857e-05, "loss": 0.0151, "step": 11610 }, { "grad_norm": 0.12841694056987762, "learning_rate": 7.198953388467549e-05, "loss": 0.0159, "step": 11620 }, { "grad_norm": 0.18207523226737976, "learning_rate": 7.194002114337526e-05, "loss": 0.0149, "step": 11630 }, { "grad_norm": 0.1117168515920639, "learning_rate": 7.189048174288054e-05, "loss": 0.0156, "step": 11640 }, { "grad_norm": 0.14173917472362518, "learning_rate": 7.184091574338636e-05, "loss": 0.0144, "step": 11650 }, { "grad_norm": 0.14573988318443298, "learning_rate": 7.179132320512009e-05, "loss": 0.0146, "step": 11660 }, { "grad_norm": 0.14315329492092133, "learning_rate": 7.174170418834134e-05, "loss": 0.0169, "step": 11670 }, { "grad_norm": 0.1393398493528366, "learning_rate": 7.169205875334189e-05, "loss": 0.0131, "step": 11680 }, { "grad_norm": 0.16658659279346466, "learning_rate": 7.164238696044562e-05, "loss": 0.017, "step": 11690 }, { "grad_norm": 0.1520349234342575, "learning_rate": 7.159268887000846e-05, "loss": 0.0162, "step": 11700 }, { "grad_norm": 0.1171252503991127, "learning_rate": 7.154296454241827e-05, "loss": 0.018, "step": 11710 }, { "grad_norm": 0.1164456307888031, "learning_rate": 7.149321403809479e-05, "loss": 0.0128, "step": 11720 }, { "grad_norm": 0.13201266527175903, "learning_rate": 7.144343741748956e-05, "loss": 0.0171, "step": 11730 }, { "grad_norm": 0.18779927492141724, "learning_rate": 7.139363474108589e-05, "loss": 0.0147, "step": 11740 }, { "grad_norm": 0.15118777751922607, "learning_rate": 7.134380606939874e-05, "loss": 0.0156, "step": 11750 }, { "grad_norm": 0.1983243077993393, "learning_rate": 7.12939514629746e-05, "loss": 0.0173, "step": 11760 }, { "grad_norm": 0.1482674926519394, "learning_rate": 7.124407098239155e-05, "loss": 0.0132, "step": 11770 }, { "grad_norm": 0.1395879089832306, "learning_rate": 7.119416468825908e-05, "loss": 0.0132, "step": 11780 }, { "grad_norm": 0.18356308341026306, "learning_rate": 7.114423264121804e-05, "loss": 0.0158, "step": 11790 }, { "grad_norm": 0.1251779943704605, "learning_rate": 7.109427490194056e-05, "loss": 0.0149, "step": 11800 }, { "grad_norm": 0.15252582728862762, "learning_rate": 7.104429153113001e-05, "loss": 0.0136, "step": 11810 }, { "grad_norm": 0.17591118812561035, "learning_rate": 7.099428258952092e-05, "loss": 0.0167, "step": 11820 }, { "grad_norm": 0.14766688644886017, "learning_rate": 7.094424813787883e-05, "loss": 0.0152, "step": 11830 }, { "grad_norm": 0.162902370095253, "learning_rate": 7.089418823700034e-05, "loss": 0.0136, "step": 11840 }, { "grad_norm": 0.14132168889045715, "learning_rate": 7.084410294771298e-05, "loss": 0.0178, "step": 11850 }, { "grad_norm": 0.16053487360477448, "learning_rate": 7.079399233087504e-05, "loss": 0.0183, "step": 11860 }, { "grad_norm": 0.1863558441400528, "learning_rate": 7.074385644737568e-05, "loss": 0.0182, "step": 11870 }, { "grad_norm": 0.164927676320076, "learning_rate": 7.069369535813473e-05, "loss": 0.0139, "step": 11880 }, { "grad_norm": 0.12510685622692108, "learning_rate": 7.06435091241026e-05, "loss": 0.0163, "step": 11890 }, { "grad_norm": 0.11255963146686554, "learning_rate": 7.059329780626034e-05, "loss": 0.0169, "step": 11900 }, { "grad_norm": 0.12491125613451004, "learning_rate": 7.054306146561944e-05, "loss": 0.0153, "step": 11910 }, { "grad_norm": 0.15426325798034668, "learning_rate": 7.049280016322178e-05, "loss": 0.0137, "step": 11920 }, { "grad_norm": 0.1622738391160965, "learning_rate": 7.044251396013957e-05, "loss": 0.013, "step": 11930 }, { "grad_norm": 0.18275178968906403, "learning_rate": 7.039220291747528e-05, "loss": 0.0153, "step": 11940 }, { "grad_norm": 0.1087409034371376, "learning_rate": 7.034186709636159e-05, "loss": 0.0142, "step": 11950 }, { "grad_norm": 0.17648498713970184, "learning_rate": 7.029150655796129e-05, "loss": 0.016, "step": 11960 }, { "grad_norm": 0.13088186085224152, "learning_rate": 7.024112136346712e-05, "loss": 0.0181, "step": 11970 }, { "grad_norm": 0.16950279474258423, "learning_rate": 7.01907115741019e-05, "loss": 0.0148, "step": 11980 }, { "grad_norm": 0.14888674020767212, "learning_rate": 7.014027725111825e-05, "loss": 0.015, "step": 11990 }, { "grad_norm": 0.17238861322402954, "learning_rate": 7.008981845579864e-05, "loss": 0.0137, "step": 12000 }, { "grad_norm": 0.14826509356498718, "learning_rate": 7.003933524945528e-05, "loss": 0.0175, "step": 12010 }, { "grad_norm": 0.1542300432920456, "learning_rate": 6.998882769342998e-05, "loss": 0.0184, "step": 12020 }, { "grad_norm": 0.19278976321220398, "learning_rate": 6.993829584909423e-05, "loss": 0.0164, "step": 12030 }, { "grad_norm": 0.13790328800678253, "learning_rate": 6.988773977784895e-05, "loss": 0.014, "step": 12040 }, { "grad_norm": 0.18205852806568146, "learning_rate": 6.983715954112454e-05, "loss": 0.0156, "step": 12050 }, { "grad_norm": 0.2010582834482193, "learning_rate": 6.978655520038079e-05, "loss": 0.0172, "step": 12060 }, { "grad_norm": 0.14734847843647003, "learning_rate": 6.97359268171067e-05, "loss": 0.015, "step": 12070 }, { "grad_norm": 0.19450758397579193, "learning_rate": 6.968527445282056e-05, "loss": 0.0127, "step": 12080 }, { "grad_norm": 0.10665734857320786, "learning_rate": 6.963459816906977e-05, "loss": 0.0117, "step": 12090 }, { "grad_norm": 0.16044332087039948, "learning_rate": 6.958389802743077e-05, "loss": 0.0147, "step": 12100 }, { "grad_norm": 0.21721936762332916, "learning_rate": 6.953317408950903e-05, "loss": 0.0147, "step": 12110 }, { "grad_norm": 0.16376623511314392, "learning_rate": 6.94824264169389e-05, "loss": 0.0149, "step": 12120 }, { "grad_norm": 0.1396978199481964, "learning_rate": 6.94316550713836e-05, "loss": 0.0127, "step": 12130 }, { "grad_norm": 0.19907212257385254, "learning_rate": 6.938086011453513e-05, "loss": 0.0138, "step": 12140 }, { "grad_norm": 0.17342379689216614, "learning_rate": 6.93300416081141e-05, "loss": 0.015, "step": 12150 }, { "grad_norm": 0.16404886543750763, "learning_rate": 6.927919961386984e-05, "loss": 0.0193, "step": 12160 }, { "grad_norm": 0.1265455186367035, "learning_rate": 6.922833419358013e-05, "loss": 0.0129, "step": 12170 }, { "grad_norm": 0.15899139642715454, "learning_rate": 6.917744540905124e-05, "loss": 0.0161, "step": 12180 }, { "grad_norm": 0.10044904053211212, "learning_rate": 6.912653332211787e-05, "loss": 0.0165, "step": 12190 }, { "grad_norm": 0.13889937102794647, "learning_rate": 6.9075597994643e-05, "loss": 0.0122, "step": 12200 }, { "grad_norm": 0.11229631304740906, "learning_rate": 6.902463948851787e-05, "loss": 0.0153, "step": 12210 }, { "grad_norm": 0.12157677859067917, "learning_rate": 6.897365786566184e-05, "loss": 0.0142, "step": 12220 }, { "grad_norm": 0.12959924340248108, "learning_rate": 6.892265318802242e-05, "loss": 0.0166, "step": 12230 }, { "grad_norm": 0.11857452988624573, "learning_rate": 6.887162551757507e-05, "loss": 0.0124, "step": 12240 }, { "grad_norm": 0.17677631974220276, "learning_rate": 6.882057491632326e-05, "loss": 0.0129, "step": 12250 }, { "grad_norm": 0.14577698707580566, "learning_rate": 6.876950144629824e-05, "loss": 0.0147, "step": 12260 }, { "grad_norm": 0.1642228662967682, "learning_rate": 6.87184051695591e-05, "loss": 0.0139, "step": 12270 }, { "grad_norm": 0.11652180552482605, "learning_rate": 6.866728614819268e-05, "loss": 0.0142, "step": 12280 }, { "grad_norm": 0.17365996539592743, "learning_rate": 6.861614444431337e-05, "loss": 0.0132, "step": 12290 }, { "grad_norm": 0.14925585687160492, "learning_rate": 6.856498012006318e-05, "loss": 0.0116, "step": 12300 }, { "grad_norm": 0.14255370199680328, "learning_rate": 6.851379323761157e-05, "loss": 0.0124, "step": 12310 }, { "grad_norm": 0.1941661238670349, "learning_rate": 6.846258385915545e-05, "loss": 0.012, "step": 12320 }, { "grad_norm": 0.1662083864212036, "learning_rate": 6.841135204691902e-05, "loss": 0.0122, "step": 12330 }, { "grad_norm": 0.11528697609901428, "learning_rate": 6.836009786315377e-05, "loss": 0.0133, "step": 12340 }, { "grad_norm": 0.1502029150724411, "learning_rate": 6.830882137013839e-05, "loss": 0.0153, "step": 12350 }, { "grad_norm": 0.11722052097320557, "learning_rate": 6.825752263017863e-05, "loss": 0.0122, "step": 12360 }, { "grad_norm": 0.1105099767446518, "learning_rate": 6.820620170560731e-05, "loss": 0.0118, "step": 12370 }, { "grad_norm": 0.11951989680528641, "learning_rate": 6.815485865878418e-05, "loss": 0.0157, "step": 12380 }, { "grad_norm": 0.1188378781080246, "learning_rate": 6.810349355209588e-05, "loss": 0.0135, "step": 12390 }, { "grad_norm": 0.14664003252983093, "learning_rate": 6.805210644795588e-05, "loss": 0.0169, "step": 12400 }, { "grad_norm": 0.2002658098936081, "learning_rate": 6.800069740880433e-05, "loss": 0.0145, "step": 12410 }, { "grad_norm": 0.15516047179698944, "learning_rate": 6.794926649710807e-05, "loss": 0.0158, "step": 12420 }, { "grad_norm": 0.16627612709999084, "learning_rate": 6.789781377536052e-05, "loss": 0.0154, "step": 12430 }, { "grad_norm": 0.1236342266201973, "learning_rate": 6.784633930608158e-05, "loss": 0.0136, "step": 12440 }, { "grad_norm": 0.13713333010673523, "learning_rate": 6.779484315181759e-05, "loss": 0.0122, "step": 12450 }, { "grad_norm": 0.15563635528087616, "learning_rate": 6.774332537514122e-05, "loss": 0.015, "step": 12460 }, { "grad_norm": 0.14479276537895203, "learning_rate": 6.769178603865143e-05, "loss": 0.0144, "step": 12470 }, { "grad_norm": 0.14545829594135284, "learning_rate": 6.764022520497337e-05, "loss": 0.016, "step": 12480 }, { "grad_norm": 0.14648547768592834, "learning_rate": 6.758864293675833e-05, "loss": 0.0144, "step": 12490 }, { "grad_norm": 0.12413424998521805, "learning_rate": 6.753703929668362e-05, "loss": 0.0136, "step": 12500 }, { "grad_norm": 0.11210532486438751, "learning_rate": 6.748541434745253e-05, "loss": 0.0129, "step": 12510 }, { "grad_norm": 0.1259201169013977, "learning_rate": 6.743376815179424e-05, "loss": 0.0129, "step": 12520 }, { "grad_norm": 0.12379930168390274, "learning_rate": 6.738210077246376e-05, "loss": 0.0136, "step": 12530 }, { "grad_norm": 0.12437110394239426, "learning_rate": 6.733041227224181e-05, "loss": 0.0131, "step": 12540 }, { "grad_norm": 0.1146075427532196, "learning_rate": 6.72787027139348e-05, "loss": 0.015, "step": 12550 }, { "grad_norm": 0.1567237377166748, "learning_rate": 6.72269721603747e-05, "loss": 0.0127, "step": 12560 }, { "grad_norm": 0.14170853793621063, "learning_rate": 6.717522067441904e-05, "loss": 0.0155, "step": 12570 }, { "grad_norm": 0.08846498280763626, "learning_rate": 6.712344831895074e-05, "loss": 0.0135, "step": 12580 }, { "grad_norm": 0.14020200073719025, "learning_rate": 6.707165515687811e-05, "loss": 0.0133, "step": 12590 }, { "grad_norm": 0.1718560755252838, "learning_rate": 6.70198412511347e-05, "loss": 0.015, "step": 12600 }, { "grad_norm": 0.1788918673992157, "learning_rate": 6.69680066646793e-05, "loss": 0.0135, "step": 12610 }, { "grad_norm": 0.10800201445817947, "learning_rate": 6.691615146049583e-05, "loss": 0.0134, "step": 12620 }, { "grad_norm": 0.13151435554027557, "learning_rate": 6.686427570159324e-05, "loss": 0.0142, "step": 12630 }, { "grad_norm": 0.13880765438079834, "learning_rate": 6.681237945100548e-05, "loss": 0.0135, "step": 12640 }, { "grad_norm": 0.1480298638343811, "learning_rate": 6.676046277179139e-05, "loss": 0.0141, "step": 12650 }, { "grad_norm": 0.14690539240837097, "learning_rate": 6.670852572703462e-05, "loss": 0.0151, "step": 12660 }, { "grad_norm": 0.13856522738933563, "learning_rate": 6.665656837984359e-05, "loss": 0.0157, "step": 12670 }, { "grad_norm": 0.1528443992137909, "learning_rate": 6.660459079335135e-05, "loss": 0.0114, "step": 12680 }, { "grad_norm": 0.21458014845848083, "learning_rate": 6.655259303071558e-05, "loss": 0.016, "step": 12690 }, { "grad_norm": 0.21686221659183502, "learning_rate": 6.650057515511848e-05, "loss": 0.0158, "step": 12700 }, { "grad_norm": 0.11105874180793762, "learning_rate": 6.644853722976667e-05, "loss": 0.0121, "step": 12710 }, { "grad_norm": 0.16096022725105286, "learning_rate": 6.639647931789113e-05, "loss": 0.0149, "step": 12720 }, { "grad_norm": 0.12798424065113068, "learning_rate": 6.634440148274713e-05, "loss": 0.0138, "step": 12730 }, { "grad_norm": 0.1334688365459442, "learning_rate": 6.629230378761415e-05, "loss": 0.0132, "step": 12740 }, { "grad_norm": 0.13614094257354736, "learning_rate": 6.624018629579581e-05, "loss": 0.0126, "step": 12750 }, { "grad_norm": 0.12507551908493042, "learning_rate": 6.618804907061976e-05, "loss": 0.0144, "step": 12760 }, { "grad_norm": 0.16404476761817932, "learning_rate": 6.613589217543766e-05, "loss": 0.0123, "step": 12770 }, { "grad_norm": 0.14546753466129303, "learning_rate": 6.608371567362504e-05, "loss": 0.0166, "step": 12780 }, { "grad_norm": 0.16433314979076385, "learning_rate": 6.60315196285813e-05, "loss": 0.016, "step": 12790 }, { "grad_norm": 0.1400090456008911, "learning_rate": 6.597930410372954e-05, "loss": 0.0144, "step": 12800 }, { "grad_norm": 0.1639266014099121, "learning_rate": 6.592706916251653e-05, "loss": 0.0163, "step": 12810 }, { "grad_norm": 0.1893301159143448, "learning_rate": 6.587481486841267e-05, "loss": 0.0132, "step": 12820 }, { "grad_norm": 0.1139996275305748, "learning_rate": 6.582254128491184e-05, "loss": 0.0124, "step": 12830 }, { "grad_norm": 0.12463530898094177, "learning_rate": 6.577024847553139e-05, "loss": 0.0141, "step": 12840 }, { "grad_norm": 0.09059487283229828, "learning_rate": 6.571793650381201e-05, "loss": 0.0117, "step": 12850 }, { "grad_norm": 0.16477645933628082, "learning_rate": 6.566560543331766e-05, "loss": 0.0128, "step": 12860 }, { "grad_norm": 0.18300490081310272, "learning_rate": 6.561325532763554e-05, "loss": 0.0138, "step": 12870 }, { "grad_norm": 0.12580551207065582, "learning_rate": 6.556088625037597e-05, "loss": 0.0155, "step": 12880 }, { "grad_norm": 0.13589827716350555, "learning_rate": 6.55084982651723e-05, "loss": 0.0138, "step": 12890 }, { "grad_norm": 0.10748524218797684, "learning_rate": 6.54560914356809e-05, "loss": 0.0123, "step": 12900 }, { "grad_norm": 0.14537277817726135, "learning_rate": 6.540366582558097e-05, "loss": 0.0138, "step": 12910 }, { "grad_norm": 0.14521481096744537, "learning_rate": 6.53512214985746e-05, "loss": 0.0153, "step": 12920 }, { "grad_norm": 0.16869473457336426, "learning_rate": 6.529875851838658e-05, "loss": 0.0138, "step": 12930 }, { "grad_norm": 0.15394854545593262, "learning_rate": 6.52462769487644e-05, "loss": 0.0116, "step": 12940 }, { "grad_norm": 0.13840527832508087, "learning_rate": 6.519377685347807e-05, "loss": 0.0139, "step": 12950 }, { "grad_norm": 0.11503861844539642, "learning_rate": 6.514125829632021e-05, "loss": 0.0137, "step": 12960 }, { "grad_norm": 0.13427236676216125, "learning_rate": 6.508872134110577e-05, "loss": 0.0147, "step": 12970 }, { "grad_norm": 0.14702676236629486, "learning_rate": 6.503616605167213e-05, "loss": 0.0142, "step": 12980 }, { "grad_norm": 0.12994484603405, "learning_rate": 6.498359249187893e-05, "loss": 0.0134, "step": 12990 }, { "grad_norm": 0.16966113448143005, "learning_rate": 6.493100072560799e-05, "loss": 0.0131, "step": 13000 }, { "grad_norm": 0.14742599427700043, "learning_rate": 6.487839081676327e-05, "loss": 0.0178, "step": 13010 }, { "grad_norm": 0.12516973912715912, "learning_rate": 6.482576282927076e-05, "loss": 0.0145, "step": 13020 }, { "grad_norm": 0.11760807782411575, "learning_rate": 6.477311682707845e-05, "loss": 0.0129, "step": 13030 }, { "grad_norm": 0.1574585884809494, "learning_rate": 6.472045287415616e-05, "loss": 0.0134, "step": 13040 }, { "grad_norm": 0.19859068095684052, "learning_rate": 6.466777103449558e-05, "loss": 0.0144, "step": 13050 }, { "grad_norm": 0.18259218335151672, "learning_rate": 6.461507137211012e-05, "loss": 0.0155, "step": 13060 }, { "grad_norm": 0.13117575645446777, "learning_rate": 6.456235395103483e-05, "loss": 0.0106, "step": 13070 }, { "grad_norm": 0.12406468391418457, "learning_rate": 6.450961883532634e-05, "loss": 0.0123, "step": 13080 }, { "grad_norm": 0.14203764498233795, "learning_rate": 6.445686608906283e-05, "loss": 0.0138, "step": 13090 }, { "grad_norm": 0.14848960936069489, "learning_rate": 6.44040957763438e-05, "loss": 0.0146, "step": 13100 }, { "grad_norm": 0.1603003591299057, "learning_rate": 6.435130796129018e-05, "loss": 0.0124, "step": 13110 }, { "grad_norm": 0.1501852124929428, "learning_rate": 6.429850270804416e-05, "loss": 0.0149, "step": 13120 }, { "grad_norm": 0.12447606027126312, "learning_rate": 6.424568008076909e-05, "loss": 0.015, "step": 13130 }, { "grad_norm": 0.15847934782505035, "learning_rate": 6.419284014364943e-05, "loss": 0.013, "step": 13140 }, { "grad_norm": 0.1448795050382614, "learning_rate": 6.41399829608907e-05, "loss": 0.0128, "step": 13150 }, { "grad_norm": 0.15055163204669952, "learning_rate": 6.408710859671938e-05, "loss": 0.0107, "step": 13160 }, { "grad_norm": 0.16741704940795898, "learning_rate": 6.403421711538278e-05, "loss": 0.0141, "step": 13170 }, { "grad_norm": 0.1724359393119812, "learning_rate": 6.398130858114903e-05, "loss": 0.0123, "step": 13180 }, { "grad_norm": 0.15216979384422302, "learning_rate": 6.392838305830701e-05, "loss": 0.0105, "step": 13190 }, { "grad_norm": 0.14182543754577637, "learning_rate": 6.387544061116621e-05, "loss": 0.0159, "step": 13200 }, { "grad_norm": 0.14061155915260315, "learning_rate": 6.382248130405671e-05, "loss": 0.0128, "step": 13210 }, { "grad_norm": 0.13544021546840668, "learning_rate": 6.376950520132905e-05, "loss": 0.0146, "step": 13220 }, { "grad_norm": 0.12956292927265167, "learning_rate": 6.371651236735418e-05, "loss": 0.0145, "step": 13230 }, { "grad_norm": 0.15257036685943604, "learning_rate": 6.366350286652341e-05, "loss": 0.0178, "step": 13240 }, { "grad_norm": 0.11732316017150879, "learning_rate": 6.361047676324827e-05, "loss": 0.0164, "step": 13250 }, { "grad_norm": 0.12449666857719421, "learning_rate": 6.355743412196047e-05, "loss": 0.0169, "step": 13260 }, { "grad_norm": 0.16973118484020233, "learning_rate": 6.350437500711184e-05, "loss": 0.0164, "step": 13270 }, { "grad_norm": 0.13323314487934113, "learning_rate": 6.345129948317418e-05, "loss": 0.0138, "step": 13280 }, { "grad_norm": 0.11428669095039368, "learning_rate": 6.33982076146393e-05, "loss": 0.016, "step": 13290 }, { "grad_norm": 0.15164785087108612, "learning_rate": 6.334509946601879e-05, "loss": 0.0158, "step": 13300 }, { "grad_norm": 0.10353098809719086, "learning_rate": 6.329197510184406e-05, "loss": 0.011, "step": 13310 }, { "grad_norm": 0.12243316322565079, "learning_rate": 6.323883458666624e-05, "loss": 0.0129, "step": 13320 }, { "grad_norm": 0.14008428156375885, "learning_rate": 6.318567798505605e-05, "loss": 0.0144, "step": 13330 }, { "grad_norm": 0.17914903163909912, "learning_rate": 6.313250536160378e-05, "loss": 0.0128, "step": 13340 }, { "grad_norm": 0.12127115577459335, "learning_rate": 6.307931678091917e-05, "loss": 0.011, "step": 13350 }, { "grad_norm": 0.1700526624917984, "learning_rate": 6.302611230763138e-05, "loss": 0.0138, "step": 13360 }, { "grad_norm": 0.12140721827745438, "learning_rate": 6.297289200638887e-05, "loss": 0.0118, "step": 13370 }, { "grad_norm": 0.13347800076007843, "learning_rate": 6.29196559418593e-05, "loss": 0.0129, "step": 13380 }, { "grad_norm": 0.08938933908939362, "learning_rate": 6.286640417872951e-05, "loss": 0.014, "step": 13390 }, { "grad_norm": 0.11133460700511932, "learning_rate": 6.281313678170542e-05, "loss": 0.0128, "step": 13400 }, { "grad_norm": 0.15579645335674286, "learning_rate": 6.275985381551194e-05, "loss": 0.0144, "step": 13410 }, { "grad_norm": 0.17702628672122955, "learning_rate": 6.270655534489292e-05, "loss": 0.0146, "step": 13420 }, { "grad_norm": 0.13929539918899536, "learning_rate": 6.265324143461098e-05, "loss": 0.0161, "step": 13430 }, { "grad_norm": 0.13590005040168762, "learning_rate": 6.259991214944758e-05, "loss": 0.0183, "step": 13440 }, { "grad_norm": 0.13402917981147766, "learning_rate": 6.254656755420283e-05, "loss": 0.0144, "step": 13450 }, { "grad_norm": 0.14229172468185425, "learning_rate": 6.249320771369545e-05, "loss": 0.0159, "step": 13460 }, { "grad_norm": 0.16742588579654694, "learning_rate": 6.243983269276263e-05, "loss": 0.0161, "step": 13470 }, { "grad_norm": 0.1398928463459015, "learning_rate": 6.238644255626012e-05, "loss": 0.0132, "step": 13480 }, { "grad_norm": 0.16091951727867126, "learning_rate": 6.233303736906193e-05, "loss": 0.0176, "step": 13490 }, { "grad_norm": 0.12479384988546371, "learning_rate": 6.227961719606039e-05, "loss": 0.0162, "step": 13500 }, { "grad_norm": 0.14217862486839294, "learning_rate": 6.222618210216608e-05, "loss": 0.0174, "step": 13510 }, { "grad_norm": 0.10018392652273178, "learning_rate": 6.217273215230767e-05, "loss": 0.011, "step": 13520 }, { "grad_norm": 0.13556398451328278, "learning_rate": 6.211926741143188e-05, "loss": 0.0128, "step": 13530 }, { "grad_norm": 0.13120292127132416, "learning_rate": 6.20657879445034e-05, "loss": 0.016, "step": 13540 }, { "grad_norm": 0.14653383195400238, "learning_rate": 6.201229381650485e-05, "loss": 0.0139, "step": 13550 }, { "grad_norm": 0.17242716252803802, "learning_rate": 6.195878509243661e-05, "loss": 0.0148, "step": 13560 }, { "grad_norm": 0.12292077392339706, "learning_rate": 6.190526183731685e-05, "loss": 0.0112, "step": 13570 }, { "grad_norm": 0.1788325011730194, "learning_rate": 6.185172411618138e-05, "loss": 0.0164, "step": 13580 }, { "grad_norm": 0.1450984925031662, "learning_rate": 6.179817199408355e-05, "loss": 0.0122, "step": 13590 }, { "grad_norm": 0.1401987373828888, "learning_rate": 6.174460553609426e-05, "loss": 0.0132, "step": 13600 }, { "grad_norm": 0.16496452689170837, "learning_rate": 6.16910248073018e-05, "loss": 0.0133, "step": 13610 }, { "grad_norm": 0.12823578715324402, "learning_rate": 6.16374298728118e-05, "loss": 0.0125, "step": 13620 }, { "grad_norm": 0.1620214581489563, "learning_rate": 6.158382079774716e-05, "loss": 0.0135, "step": 13630 }, { "grad_norm": 0.14681223034858704, "learning_rate": 6.153019764724799e-05, "loss": 0.0143, "step": 13640 }, { "grad_norm": 0.16912101209163666, "learning_rate": 6.147656048647144e-05, "loss": 0.012, "step": 13650 }, { "grad_norm": 0.14713923633098602, "learning_rate": 6.142290938059173e-05, "loss": 0.0144, "step": 13660 }, { "grad_norm": 0.16031445562839508, "learning_rate": 6.136924439480001e-05, "loss": 0.0114, "step": 13670 }, { "grad_norm": 0.14857104420661926, "learning_rate": 6.13155655943043e-05, "loss": 0.0149, "step": 13680 }, { "grad_norm": 0.10818526148796082, "learning_rate": 6.126187304432941e-05, "loss": 0.013, "step": 13690 }, { "grad_norm": 0.1844847947359085, "learning_rate": 6.120816681011684e-05, "loss": 0.0127, "step": 13700 }, { "grad_norm": 0.11603015661239624, "learning_rate": 6.115444695692474e-05, "loss": 0.0142, "step": 13710 }, { "grad_norm": 0.14891298115253448, "learning_rate": 6.110071355002779e-05, "loss": 0.0117, "step": 13720 }, { "grad_norm": 0.13027285039424896, "learning_rate": 6.104696665471714e-05, "loss": 0.0135, "step": 13730 }, { "grad_norm": 0.1440996527671814, "learning_rate": 6.099320633630036e-05, "loss": 0.0126, "step": 13740 }, { "grad_norm": 0.14623671770095825, "learning_rate": 6.093943266010128e-05, "loss": 0.0151, "step": 13750 }, { "grad_norm": 0.11499848961830139, "learning_rate": 6.088564569146e-05, "loss": 0.0118, "step": 13760 }, { "grad_norm": 0.13576751947402954, "learning_rate": 6.083184549573274e-05, "loss": 0.0124, "step": 13770 }, { "grad_norm": 0.16443251073360443, "learning_rate": 6.077803213829184e-05, "loss": 0.013, "step": 13780 }, { "grad_norm": 0.14421872794628143, "learning_rate": 6.0724205684525594e-05, "loss": 0.0139, "step": 13790 }, { "grad_norm": 0.12139511108398438, "learning_rate": 6.0670366199838215e-05, "loss": 0.0148, "step": 13800 }, { "grad_norm": 0.1670086681842804, "learning_rate": 6.061651374964974e-05, "loss": 0.0145, "step": 13810 }, { "grad_norm": 0.14992676675319672, "learning_rate": 6.0562648399396004e-05, "loss": 0.0131, "step": 13820 }, { "grad_norm": 0.1757311373949051, "learning_rate": 6.050877021452845e-05, "loss": 0.0144, "step": 13830 }, { "grad_norm": 0.10971327126026154, "learning_rate": 6.04548792605142e-05, "loss": 0.0132, "step": 13840 }, { "grad_norm": 0.11077089607715607, "learning_rate": 6.04009756028358e-05, "loss": 0.0118, "step": 13850 }, { "grad_norm": 0.1602349579334259, "learning_rate": 6.0347059306991294e-05, "loss": 0.0157, "step": 13860 }, { "grad_norm": 0.089258573949337, "learning_rate": 6.029313043849407e-05, "loss": 0.0107, "step": 13870 }, { "grad_norm": 0.15238426625728607, "learning_rate": 6.023918906287279e-05, "loss": 0.0113, "step": 13880 }, { "grad_norm": 0.17031070590019226, "learning_rate": 6.01852352456713e-05, "loss": 0.014, "step": 13890 }, { "grad_norm": 0.13752290606498718, "learning_rate": 6.013126905244858e-05, "loss": 0.014, "step": 13900 }, { "grad_norm": 0.13070815801620483, "learning_rate": 6.007729054877864e-05, "loss": 0.0108, "step": 13910 }, { "grad_norm": 0.10863208770751953, "learning_rate": 6.0023299800250466e-05, "loss": 0.0133, "step": 13920 }, { "grad_norm": 0.0826619416475296, "learning_rate": 5.9969296872467894e-05, "loss": 0.0118, "step": 13930 }, { "grad_norm": 0.18416371941566467, "learning_rate": 5.9915281831049585e-05, "loss": 0.0134, "step": 13940 }, { "grad_norm": 0.18143093585968018, "learning_rate": 5.98612547416289e-05, "loss": 0.0144, "step": 13950 }, { "grad_norm": 0.14473626017570496, "learning_rate": 5.9807215669853855e-05, "loss": 0.0117, "step": 13960 }, { "grad_norm": 0.1288524568080902, "learning_rate": 5.9753164681387e-05, "loss": 0.0118, "step": 13970 }, { "grad_norm": 0.1043073758482933, "learning_rate": 5.969910184190539e-05, "loss": 0.0118, "step": 13980 }, { "grad_norm": 0.1317167431116104, "learning_rate": 5.9645027217100477e-05, "loss": 0.0126, "step": 13990 }, { "grad_norm": 0.11090383678674698, "learning_rate": 5.959094087267804e-05, "loss": 0.0099, "step": 14000 }, { "grad_norm": 0.16584862768650055, "learning_rate": 5.9536842874358065e-05, "loss": 0.0127, "step": 14010 }, { "grad_norm": 0.15128959715366364, "learning_rate": 5.9482733287874734e-05, "loss": 0.0134, "step": 14020 }, { "grad_norm": 0.11339466273784637, "learning_rate": 5.942861217897631e-05, "loss": 0.0119, "step": 14030 }, { "grad_norm": 0.1567685902118683, "learning_rate": 5.9374479613425e-05, "loss": 0.0119, "step": 14040 }, { "grad_norm": 0.12866921722888947, "learning_rate": 5.932033565699704e-05, "loss": 0.0117, "step": 14050 }, { "grad_norm": 0.12666508555412292, "learning_rate": 5.926618037548237e-05, "loss": 0.0129, "step": 14060 }, { "grad_norm": 0.11791911721229553, "learning_rate": 5.9212013834684824e-05, "loss": 0.0142, "step": 14070 }, { "grad_norm": 0.11760513484477997, "learning_rate": 5.9157836100421824e-05, "loss": 0.0117, "step": 14080 }, { "grad_norm": 0.10780870169401169, "learning_rate": 5.910364723852444e-05, "loss": 0.0129, "step": 14090 }, { "grad_norm": 0.1526738405227661, "learning_rate": 5.904944731483723e-05, "loss": 0.0126, "step": 14100 }, { "grad_norm": 0.13516376912593842, "learning_rate": 5.899523639521825e-05, "loss": 0.0107, "step": 14110 }, { "grad_norm": 0.10689202696084976, "learning_rate": 5.894101454553883e-05, "loss": 0.0106, "step": 14120 }, { "grad_norm": 0.13959378004074097, "learning_rate": 5.888678183168368e-05, "loss": 0.013, "step": 14130 }, { "grad_norm": 0.13614320755004883, "learning_rate": 5.883253831955061e-05, "loss": 0.0145, "step": 14140 }, { "grad_norm": 0.1757003515958786, "learning_rate": 5.8778284075050625e-05, "loss": 0.0116, "step": 14150 }, { "grad_norm": 0.14913739264011383, "learning_rate": 5.872401916410777e-05, "loss": 0.0133, "step": 14160 }, { "grad_norm": 0.11475637555122375, "learning_rate": 5.8669743652659014e-05, "loss": 0.0142, "step": 14170 }, { "grad_norm": 0.13099338114261627, "learning_rate": 5.86154576066542e-05, "loss": 0.0119, "step": 14180 }, { "grad_norm": 0.16083528101444244, "learning_rate": 5.856116109205602e-05, "loss": 0.0117, "step": 14190 }, { "grad_norm": 0.14924566447734833, "learning_rate": 5.850685417483983e-05, "loss": 0.013, "step": 14200 }, { "grad_norm": 0.14049433171749115, "learning_rate": 5.8452536920993684e-05, "loss": 0.0127, "step": 14210 }, { "grad_norm": 0.10586835443973541, "learning_rate": 5.839820939651817e-05, "loss": 0.0138, "step": 14220 }, { "grad_norm": 0.1351652294397354, "learning_rate": 5.834387166742632e-05, "loss": 0.0141, "step": 14230 }, { "grad_norm": 0.14196577668190002, "learning_rate": 5.828952379974364e-05, "loss": 0.0119, "step": 14240 }, { "grad_norm": 0.14962828159332275, "learning_rate": 5.8235165859507864e-05, "loss": 0.0122, "step": 14250 }, { "grad_norm": 0.11779846996068954, "learning_rate": 5.818079791276907e-05, "loss": 0.0143, "step": 14260 }, { "grad_norm": 0.145252525806427, "learning_rate": 5.8126420025589415e-05, "loss": 0.0115, "step": 14270 }, { "grad_norm": 0.1043713316321373, "learning_rate": 5.807203226404313e-05, "loss": 0.0116, "step": 14280 }, { "grad_norm": 0.12059379369020462, "learning_rate": 5.801763469421652e-05, "loss": 0.0097, "step": 14290 }, { "grad_norm": 0.12893664836883545, "learning_rate": 5.796322738220774e-05, "loss": 0.0129, "step": 14300 }, { "grad_norm": 0.13573160767555237, "learning_rate": 5.79088103941268e-05, "loss": 0.0117, "step": 14310 }, { "grad_norm": 0.13926677405834198, "learning_rate": 5.785438379609549e-05, "loss": 0.0161, "step": 14320 }, { "grad_norm": 0.1391792893409729, "learning_rate": 5.779994765424724e-05, "loss": 0.0137, "step": 14330 }, { "grad_norm": 0.13279008865356445, "learning_rate": 5.7745502034727114e-05, "loss": 0.0139, "step": 14340 }, { "grad_norm": 0.1325315237045288, "learning_rate": 5.769104700369164e-05, "loss": 0.0118, "step": 14350 }, { "grad_norm": 0.15189999341964722, "learning_rate": 5.7636582627308854e-05, "loss": 0.0156, "step": 14360 }, { "grad_norm": 0.151927188038826, "learning_rate": 5.7582108971758095e-05, "loss": 0.0131, "step": 14370 }, { "grad_norm": 0.13276609778404236, "learning_rate": 5.752762610322999e-05, "loss": 0.0132, "step": 14380 }, { "grad_norm": 0.11099401861429214, "learning_rate": 5.747313408792636e-05, "loss": 0.0128, "step": 14390 }, { "grad_norm": 0.13411420583724976, "learning_rate": 5.741863299206014e-05, "loss": 0.0123, "step": 14400 }, { "grad_norm": 0.13604989647865295, "learning_rate": 5.73641228818553e-05, "loss": 0.0142, "step": 14410 }, { "grad_norm": 0.10835506021976471, "learning_rate": 5.730960382354676e-05, "loss": 0.0114, "step": 14420 }, { "grad_norm": 0.10736110061407089, "learning_rate": 5.725507588338035e-05, "loss": 0.0105, "step": 14430 }, { "grad_norm": 0.1525639295578003, "learning_rate": 5.7200539127612604e-05, "loss": 0.0105, "step": 14440 }, { "grad_norm": 0.16826744377613068, "learning_rate": 5.714599362251087e-05, "loss": 0.0153, "step": 14450 }, { "grad_norm": 0.1438940167427063, "learning_rate": 5.709143943435307e-05, "loss": 0.0118, "step": 14460 }, { "grad_norm": 0.15250886976718903, "learning_rate": 5.7036876629427646e-05, "loss": 0.0139, "step": 14470 }, { "grad_norm": 0.16099700331687927, "learning_rate": 5.698230527403361e-05, "loss": 0.0159, "step": 14480 }, { "grad_norm": 0.1154017224907875, "learning_rate": 5.6927725434480274e-05, "loss": 0.0116, "step": 14490 }, { "grad_norm": 0.14721696078777313, "learning_rate": 5.687313717708728e-05, "loss": 0.0103, "step": 14500 }, { "grad_norm": 0.13831916451454163, "learning_rate": 5.681854056818453e-05, "loss": 0.0114, "step": 14510 }, { "grad_norm": 0.15794281661510468, "learning_rate": 5.676393567411205e-05, "loss": 0.0102, "step": 14520 }, { "grad_norm": 0.1276630163192749, "learning_rate": 5.670932256121991e-05, "loss": 0.0124, "step": 14530 }, { "grad_norm": 0.15986688435077667, "learning_rate": 5.6654701295868216e-05, "loss": 0.0124, "step": 14540 }, { "grad_norm": 0.15216626226902008, "learning_rate": 5.660007194442697e-05, "loss": 0.0127, "step": 14550 }, { "grad_norm": 0.16958504915237427, "learning_rate": 5.6545434573275946e-05, "loss": 0.0115, "step": 14560 }, { "grad_norm": 0.13685940206050873, "learning_rate": 5.649078924880472e-05, "loss": 0.0121, "step": 14570 }, { "grad_norm": 0.11162809282541275, "learning_rate": 5.643613603741252e-05, "loss": 0.0123, "step": 14580 }, { "grad_norm": 0.08890724182128906, "learning_rate": 5.6381475005508154e-05, "loss": 0.0103, "step": 14590 }, { "grad_norm": 0.12545965611934662, "learning_rate": 5.632680621950992e-05, "loss": 0.0115, "step": 14600 }, { "grad_norm": 0.11660564690828323, "learning_rate": 5.627212974584555e-05, "loss": 0.0123, "step": 14610 }, { "grad_norm": 0.11655749380588531, "learning_rate": 5.62174456509521e-05, "loss": 0.0114, "step": 14620 }, { "grad_norm": 0.11715232580900192, "learning_rate": 5.616275400127594e-05, "loss": 0.0146, "step": 14630 }, { "grad_norm": 0.1264192759990692, "learning_rate": 5.610805486327254e-05, "loss": 0.0129, "step": 14640 }, { "grad_norm": 0.14592042565345764, "learning_rate": 5.6053348303406536e-05, "loss": 0.0119, "step": 14650 }, { "grad_norm": 0.11013375222682953, "learning_rate": 5.599863438815156e-05, "loss": 0.0103, "step": 14660 }, { "grad_norm": 0.11289285123348236, "learning_rate": 5.594391318399017e-05, "loss": 0.0104, "step": 14670 }, { "grad_norm": 0.131998673081398, "learning_rate": 5.588918475741377e-05, "loss": 0.0119, "step": 14680 }, { "grad_norm": 0.12553125619888306, "learning_rate": 5.5834449174922586e-05, "loss": 0.0104, "step": 14690 }, { "grad_norm": 0.1162482500076294, "learning_rate": 5.57797065030255e-05, "loss": 0.0125, "step": 14700 }, { "grad_norm": 0.07996292412281036, "learning_rate": 5.5724956808240016e-05, "loss": 0.0198, "step": 14710 }, { "grad_norm": 0.16195502877235413, "learning_rate": 5.567020015709219e-05, "loss": 0.0166, "step": 14720 }, { "grad_norm": 0.14080706238746643, "learning_rate": 5.561543661611649e-05, "loss": 0.0132, "step": 14730 }, { "grad_norm": 0.11416575312614441, "learning_rate": 5.556066625185583e-05, "loss": 0.0179, "step": 14740 }, { "grad_norm": 0.13234585523605347, "learning_rate": 5.550588913086131e-05, "loss": 0.0131, "step": 14750 }, { "grad_norm": 0.11887134611606598, "learning_rate": 5.545110531969234e-05, "loss": 0.014, "step": 14760 }, { "grad_norm": 0.11097018420696259, "learning_rate": 5.539631488491641e-05, "loss": 0.0144, "step": 14770 }, { "grad_norm": 0.13666154444217682, "learning_rate": 5.534151789310904e-05, "loss": 0.0127, "step": 14780 }, { "grad_norm": 0.11645801365375519, "learning_rate": 5.528671441085376e-05, "loss": 0.0129, "step": 14790 }, { "grad_norm": 0.14585749804973602, "learning_rate": 5.523190450474197e-05, "loss": 0.0104, "step": 14800 }, { "grad_norm": 0.10783863812685013, "learning_rate": 5.5177088241372865e-05, "loss": 0.0113, "step": 14810 }, { "grad_norm": 0.1614600419998169, "learning_rate": 5.512226568735338e-05, "loss": 0.012, "step": 14820 }, { "grad_norm": 0.1063212901353836, "learning_rate": 5.506743690929809e-05, "loss": 0.0127, "step": 14830 }, { "grad_norm": 0.13056474924087524, "learning_rate": 5.501260197382912e-05, "loss": 0.011, "step": 14840 }, { "grad_norm": 0.13377559185028076, "learning_rate": 5.49577609475761e-05, "loss": 0.0106, "step": 14850 }, { "grad_norm": 0.16174156963825226, "learning_rate": 5.490291389717603e-05, "loss": 0.011, "step": 14860 }, { "grad_norm": 0.1528310924768448, "learning_rate": 5.484806088927329e-05, "loss": 0.0123, "step": 14870 }, { "grad_norm": 0.13397316634655, "learning_rate": 5.479320199051942e-05, "loss": 0.012, "step": 14880 }, { "grad_norm": 0.11971265077590942, "learning_rate": 5.473833726757314e-05, "loss": 0.0138, "step": 14890 }, { "grad_norm": 0.10590039938688278, "learning_rate": 5.4683466787100304e-05, "loss": 0.0112, "step": 14900 }, { "grad_norm": 0.1543235331773758, "learning_rate": 5.462859061577369e-05, "loss": 0.0116, "step": 14910 }, { "grad_norm": 0.11589136719703674, "learning_rate": 5.4573708820273026e-05, "loss": 0.0177, "step": 14920 }, { "grad_norm": 0.1168501079082489, "learning_rate": 5.451882146728488e-05, "loss": 0.0121, "step": 14930 }, { "grad_norm": 0.17091704905033112, "learning_rate": 5.446392862350255e-05, "loss": 0.0138, "step": 14940 }, { "grad_norm": 0.12951578199863434, "learning_rate": 5.440903035562603e-05, "loss": 0.0131, "step": 14950 }, { "grad_norm": 0.13843896985054016, "learning_rate": 5.435412673036188e-05, "loss": 0.0111, "step": 14960 }, { "grad_norm": 0.14586523175239563, "learning_rate": 5.429921781442318e-05, "loss": 0.0111, "step": 14970 }, { "grad_norm": 0.09353245049715042, "learning_rate": 5.424430367452945e-05, "loss": 0.0119, "step": 14980 }, { "grad_norm": 0.11588440090417862, "learning_rate": 5.418938437740655e-05, "loss": 0.0159, "step": 14990 }, { "grad_norm": 0.11895736306905746, "learning_rate": 5.413445998978658e-05, "loss": 0.0124, "step": 15000 } ], "logging_steps": 10, "max_steps": 30000, "num_input_tokens_seen": 0, "num_train_epochs": 9223372036854775807, "save_steps": 15000, "stateful_callbacks": { "TrainerControl": { "args": { "should_epoch_stop": false, "should_evaluate": false, "should_log": false, "should_save": true, "should_training_stop": false }, "attributes": {} } }, "total_flos": 0.0, "train_batch_size": 32, "trial_name": null, "trial_params": null }