{ "best_global_step": null, "best_metric": null, "best_model_checkpoint": null, "epoch": 0.8, "eval_steps": 500, "global_step": 80000, "is_hyper_param_search": false, "is_local_process_zero": true, "is_world_process_zero": true, "log_history": [ { "grad_norm": 1.0110338926315308, "learning_rate": 1.8e-07, "loss": 1.5771, "step": 10 }, { "grad_norm": 1.259131669998169, "learning_rate": 3.8e-07, "loss": 1.597, "step": 20 }, { "grad_norm": 1.0786699056625366, "learning_rate": 5.8e-07, "loss": 1.5407, "step": 30 }, { "grad_norm": 1.4142526388168335, "learning_rate": 7.8e-07, "loss": 1.567, "step": 40 }, { "grad_norm": 1.115562081336975, "learning_rate": 9.8e-07, "loss": 1.5521, "step": 50 }, { "grad_norm": 1.4360578060150146, "learning_rate": 1.18e-06, "loss": 1.5265, "step": 60 }, { "grad_norm": 1.9106122255325317, "learning_rate": 1.3800000000000001e-06, "loss": 1.4242, "step": 70 }, { "grad_norm": 1.8124233484268188, "learning_rate": 1.5800000000000003e-06, "loss": 1.3718, "step": 80 }, { "grad_norm": 1.4542392492294312, "learning_rate": 1.7800000000000001e-06, "loss": 1.3039, "step": 90 }, { "grad_norm": 1.5422126054763794, "learning_rate": 1.98e-06, "loss": 1.3157, "step": 100 }, { "grad_norm": 1.612699031829834, "learning_rate": 2.1800000000000003e-06, "loss": 1.29, "step": 110 }, { "grad_norm": 3.707186698913574, "learning_rate": 2.38e-06, "loss": 1.2677, "step": 120 }, { "grad_norm": 1.4454550743103027, "learning_rate": 2.5800000000000003e-06, "loss": 1.2418, "step": 130 }, { "grad_norm": 2.0081264972686768, "learning_rate": 2.78e-06, "loss": 1.2051, "step": 140 }, { "grad_norm": 1.4187254905700684, "learning_rate": 2.9800000000000003e-06, "loss": 1.1414, "step": 150 }, { "grad_norm": 2.3556902408599854, "learning_rate": 3.1800000000000005e-06, "loss": 1.1719, "step": 160 }, { "grad_norm": 2.3005118370056152, "learning_rate": 3.38e-06, "loss": 1.1206, "step": 170 }, { "grad_norm": 1.7349478006362915, "learning_rate": 3.58e-06, "loss": 1.1016, "step": 180 }, { "grad_norm": 2.4811720848083496, "learning_rate": 3.7800000000000002e-06, "loss": 1.0943, "step": 190 }, { "grad_norm": 2.271969795227051, "learning_rate": 3.98e-06, "loss": 1.1256, "step": 200 }, { "grad_norm": 1.8641239404678345, "learning_rate": 4.18e-06, "loss": 1.114, "step": 210 }, { "grad_norm": 1.7965772151947021, "learning_rate": 4.38e-06, "loss": 1.1191, "step": 220 }, { "grad_norm": 1.6025229692459106, "learning_rate": 4.58e-06, "loss": 1.1106, "step": 230 }, { "grad_norm": 1.7413052320480347, "learning_rate": 4.780000000000001e-06, "loss": 1.0901, "step": 240 }, { "grad_norm": 1.3394341468811035, "learning_rate": 4.98e-06, "loss": 1.0856, "step": 250 }, { "grad_norm": 1.8484092950820923, "learning_rate": 5.18e-06, "loss": 1.0652, "step": 260 }, { "grad_norm": 1.4424114227294922, "learning_rate": 5.38e-06, "loss": 1.0816, "step": 270 }, { "grad_norm": 1.5816348791122437, "learning_rate": 5.580000000000001e-06, "loss": 1.0716, "step": 280 }, { "grad_norm": 1.6192833185195923, "learning_rate": 5.78e-06, "loss": 1.1043, "step": 290 }, { "grad_norm": 1.5688928365707397, "learning_rate": 5.98e-06, "loss": 1.073, "step": 300 }, { "grad_norm": 1.693131923675537, "learning_rate": 6.18e-06, "loss": 1.1094, "step": 310 }, { "grad_norm": 1.7114369869232178, "learning_rate": 6.38e-06, "loss": 1.092, "step": 320 }, { "grad_norm": 1.5655052661895752, "learning_rate": 6.58e-06, "loss": 1.073, "step": 330 }, { "grad_norm": 1.9529173374176025, "learning_rate": 6.78e-06, "loss": 1.0907, "step": 340 }, { "grad_norm": 1.731724500656128, "learning_rate": 6.98e-06, "loss": 1.0819, "step": 350 }, { "grad_norm": 1.3083642721176147, "learning_rate": 7.180000000000001e-06, "loss": 1.0858, "step": 360 }, { "grad_norm": 1.7053009271621704, "learning_rate": 7.3800000000000005e-06, "loss": 1.0848, "step": 370 }, { "grad_norm": 1.6039392948150635, "learning_rate": 7.580000000000001e-06, "loss": 1.0984, "step": 380 }, { "grad_norm": 1.6705094575881958, "learning_rate": 7.78e-06, "loss": 1.0864, "step": 390 }, { "grad_norm": 1.6706316471099854, "learning_rate": 7.98e-06, "loss": 1.0769, "step": 400 }, { "grad_norm": 1.6198359727859497, "learning_rate": 8.18e-06, "loss": 1.0917, "step": 410 }, { "grad_norm": 1.2895989418029785, "learning_rate": 8.380000000000001e-06, "loss": 1.0782, "step": 420 }, { "grad_norm": 1.5910062789916992, "learning_rate": 8.580000000000001e-06, "loss": 1.0618, "step": 430 }, { "grad_norm": 1.979642391204834, "learning_rate": 8.78e-06, "loss": 1.0607, "step": 440 }, { "grad_norm": 1.4569953680038452, "learning_rate": 8.98e-06, "loss": 1.0275, "step": 450 }, { "grad_norm": 1.5524741411209106, "learning_rate": 9.180000000000002e-06, "loss": 1.0343, "step": 460 }, { "grad_norm": 1.7159589529037476, "learning_rate": 9.38e-06, "loss": 1.0455, "step": 470 }, { "grad_norm": 2.153380870819092, "learning_rate": 9.58e-06, "loss": 1.0281, "step": 480 }, { "grad_norm": 1.9474912881851196, "learning_rate": 9.78e-06, "loss": 1.0289, "step": 490 }, { "grad_norm": 1.9878839254379272, "learning_rate": 9.980000000000001e-06, "loss": 0.9977, "step": 500 }, { "grad_norm": 2.071010112762451, "learning_rate": 1.018e-05, "loss": 0.9696, "step": 510 }, { "grad_norm": 2.145448684692383, "learning_rate": 1.038e-05, "loss": 0.9226, "step": 520 }, { "grad_norm": 2.667541027069092, "learning_rate": 1.058e-05, "loss": 0.9094, "step": 530 }, { "grad_norm": 3.0407752990722656, "learning_rate": 1.0780000000000002e-05, "loss": 0.8671, "step": 540 }, { "grad_norm": 2.7672817707061768, "learning_rate": 1.098e-05, "loss": 0.8519, "step": 550 }, { "grad_norm": 2.3314497470855713, "learning_rate": 1.118e-05, "loss": 0.8334, "step": 560 }, { "grad_norm": 2.1435508728027344, "learning_rate": 1.1380000000000001e-05, "loss": 0.8056, "step": 570 }, { "grad_norm": 2.3873069286346436, "learning_rate": 1.1580000000000001e-05, "loss": 0.7545, "step": 580 }, { "grad_norm": 2.6689839363098145, "learning_rate": 1.178e-05, "loss": 0.7625, "step": 590 }, { "grad_norm": 2.417330026626587, "learning_rate": 1.198e-05, "loss": 0.7229, "step": 600 }, { "grad_norm": 2.834223508834839, "learning_rate": 1.2180000000000002e-05, "loss": 0.7312, "step": 610 }, { "grad_norm": 2.5682950019836426, "learning_rate": 1.238e-05, "loss": 0.6881, "step": 620 }, { "grad_norm": 2.3975839614868164, "learning_rate": 1.258e-05, "loss": 0.6814, "step": 630 }, { "grad_norm": 2.562290906906128, "learning_rate": 1.278e-05, "loss": 0.6203, "step": 640 }, { "grad_norm": 2.9845962524414062, "learning_rate": 1.2980000000000001e-05, "loss": 0.5855, "step": 650 }, { "grad_norm": 2.941396713256836, "learning_rate": 1.3180000000000001e-05, "loss": 0.5918, "step": 660 }, { "grad_norm": 2.814932107925415, "learning_rate": 1.338e-05, "loss": 0.5383, "step": 670 }, { "grad_norm": 2.8579673767089844, "learning_rate": 1.358e-05, "loss": 0.5144, "step": 680 }, { "grad_norm": 2.9469640254974365, "learning_rate": 1.3780000000000002e-05, "loss": 0.5467, "step": 690 }, { "grad_norm": 3.7387378215789795, "learning_rate": 1.3980000000000002e-05, "loss": 0.5358, "step": 700 }, { "grad_norm": 4.025566577911377, "learning_rate": 1.4180000000000001e-05, "loss": 0.4598, "step": 710 }, { "grad_norm": 4.127574920654297, "learning_rate": 1.4380000000000001e-05, "loss": 0.4858, "step": 720 }, { "grad_norm": 3.254312753677368, "learning_rate": 1.4580000000000003e-05, "loss": 0.4378, "step": 730 }, { "grad_norm": 2.837331771850586, "learning_rate": 1.4779999999999999e-05, "loss": 0.4033, "step": 740 }, { "grad_norm": 3.437325954437256, "learning_rate": 1.4979999999999999e-05, "loss": 0.4119, "step": 750 }, { "grad_norm": 3.035741090774536, "learning_rate": 1.518e-05, "loss": 0.3813, "step": 760 }, { "grad_norm": 2.8543033599853516, "learning_rate": 1.538e-05, "loss": 0.3709, "step": 770 }, { "grad_norm": 2.824770212173462, "learning_rate": 1.558e-05, "loss": 0.385, "step": 780 }, { "grad_norm": 2.268631935119629, "learning_rate": 1.578e-05, "loss": 0.321, "step": 790 }, { "grad_norm": 2.8071677684783936, "learning_rate": 1.598e-05, "loss": 0.2964, "step": 800 }, { "grad_norm": 3.2190346717834473, "learning_rate": 1.618e-05, "loss": 0.2889, "step": 810 }, { "grad_norm": 2.936357259750366, "learning_rate": 1.6380000000000002e-05, "loss": 0.2838, "step": 820 }, { "grad_norm": 3.112809181213379, "learning_rate": 1.658e-05, "loss": 0.2977, "step": 830 }, { "grad_norm": 2.827674388885498, "learning_rate": 1.6780000000000002e-05, "loss": 0.283, "step": 840 }, { "grad_norm": 2.5919880867004395, "learning_rate": 1.698e-05, "loss": 0.2923, "step": 850 }, { "grad_norm": 3.120465040206909, "learning_rate": 1.718e-05, "loss": 0.2436, "step": 860 }, { "grad_norm": 2.4982752799987793, "learning_rate": 1.7380000000000003e-05, "loss": 0.283, "step": 870 }, { "grad_norm": 3.573584794998169, "learning_rate": 1.758e-05, "loss": 0.2568, "step": 880 }, { "grad_norm": 2.477536201477051, "learning_rate": 1.7780000000000003e-05, "loss": 0.268, "step": 890 }, { "grad_norm": 2.679007053375244, "learning_rate": 1.798e-05, "loss": 0.2547, "step": 900 }, { "grad_norm": 2.9270520210266113, "learning_rate": 1.818e-05, "loss": 0.2261, "step": 910 }, { "grad_norm": 2.275438070297241, "learning_rate": 1.838e-05, "loss": 0.2047, "step": 920 }, { "grad_norm": 3.007584810256958, "learning_rate": 1.858e-05, "loss": 0.2222, "step": 930 }, { "grad_norm": 3.7348198890686035, "learning_rate": 1.878e-05, "loss": 0.245, "step": 940 }, { "grad_norm": 2.18304181098938, "learning_rate": 1.898e-05, "loss": 0.2189, "step": 950 }, { "grad_norm": 2.9451920986175537, "learning_rate": 1.918e-05, "loss": 0.235, "step": 960 }, { "grad_norm": 3.7052807807922363, "learning_rate": 1.938e-05, "loss": 0.2489, "step": 970 }, { "grad_norm": 3.216968297958374, "learning_rate": 1.9580000000000002e-05, "loss": 0.2226, "step": 980 }, { "grad_norm": 2.8407645225524902, "learning_rate": 1.978e-05, "loss": 0.1896, "step": 990 }, { "grad_norm": 3.230043888092041, "learning_rate": 1.9980000000000002e-05, "loss": 0.2044, "step": 1000 }, { "grad_norm": 2.4371490478515625, "learning_rate": 2.0180000000000003e-05, "loss": 0.2249, "step": 1010 }, { "grad_norm": 2.4402239322662354, "learning_rate": 2.038e-05, "loss": 0.2263, "step": 1020 }, { "grad_norm": 2.8809869289398193, "learning_rate": 2.0580000000000003e-05, "loss": 0.1876, "step": 1030 }, { "grad_norm": 2.1531450748443604, "learning_rate": 2.078e-05, "loss": 0.2182, "step": 1040 }, { "grad_norm": 2.848485231399536, "learning_rate": 2.098e-05, "loss": 0.2044, "step": 1050 }, { "grad_norm": 1.8288562297821045, "learning_rate": 2.118e-05, "loss": 0.2175, "step": 1060 }, { "grad_norm": 2.008026599884033, "learning_rate": 2.138e-05, "loss": 0.2149, "step": 1070 }, { "grad_norm": 3.099989891052246, "learning_rate": 2.158e-05, "loss": 0.2262, "step": 1080 }, { "grad_norm": 1.4661778211593628, "learning_rate": 2.178e-05, "loss": 0.1902, "step": 1090 }, { "grad_norm": 2.7481231689453125, "learning_rate": 2.198e-05, "loss": 0.1586, "step": 1100 }, { "grad_norm": 2.848923444747925, "learning_rate": 2.218e-05, "loss": 0.2363, "step": 1110 }, { "grad_norm": 1.875160813331604, "learning_rate": 2.2380000000000003e-05, "loss": 0.1629, "step": 1120 }, { "grad_norm": 2.5567705631256104, "learning_rate": 2.258e-05, "loss": 0.2079, "step": 1130 }, { "grad_norm": 2.768261432647705, "learning_rate": 2.2780000000000002e-05, "loss": 0.253, "step": 1140 }, { "grad_norm": 2.6382968425750732, "learning_rate": 2.298e-05, "loss": 0.227, "step": 1150 }, { "grad_norm": 1.71817147731781, "learning_rate": 2.318e-05, "loss": 0.1974, "step": 1160 }, { "grad_norm": 1.9722336530685425, "learning_rate": 2.3380000000000003e-05, "loss": 0.1878, "step": 1170 }, { "grad_norm": 2.3601298332214355, "learning_rate": 2.358e-05, "loss": 0.1728, "step": 1180 }, { "grad_norm": 2.8796558380126953, "learning_rate": 2.3780000000000003e-05, "loss": 0.1958, "step": 1190 }, { "grad_norm": 3.66633677482605, "learning_rate": 2.398e-05, "loss": 0.1945, "step": 1200 }, { "grad_norm": 4.199319362640381, "learning_rate": 2.418e-05, "loss": 0.212, "step": 1210 }, { "grad_norm": 2.45011043548584, "learning_rate": 2.438e-05, "loss": 0.1749, "step": 1220 }, { "grad_norm": 2.844465494155884, "learning_rate": 2.4580000000000002e-05, "loss": 0.1938, "step": 1230 }, { "grad_norm": 2.5903515815734863, "learning_rate": 2.478e-05, "loss": 0.1899, "step": 1240 }, { "grad_norm": 2.650827407836914, "learning_rate": 2.498e-05, "loss": 0.1891, "step": 1250 }, { "grad_norm": 3.0194389820098877, "learning_rate": 2.5180000000000003e-05, "loss": 0.1856, "step": 1260 }, { "grad_norm": 2.1181724071502686, "learning_rate": 2.5380000000000004e-05, "loss": 0.1517, "step": 1270 }, { "grad_norm": 1.7006850242614746, "learning_rate": 2.5580000000000002e-05, "loss": 0.1859, "step": 1280 }, { "grad_norm": 2.6850388050079346, "learning_rate": 2.5779999999999997e-05, "loss": 0.1899, "step": 1290 }, { "grad_norm": 2.0383410453796387, "learning_rate": 2.598e-05, "loss": 0.1649, "step": 1300 }, { "grad_norm": 2.400217056274414, "learning_rate": 2.618e-05, "loss": 0.1982, "step": 1310 }, { "grad_norm": 2.8230690956115723, "learning_rate": 2.6379999999999998e-05, "loss": 0.1797, "step": 1320 }, { "grad_norm": 1.939424753189087, "learning_rate": 2.658e-05, "loss": 0.151, "step": 1330 }, { "grad_norm": 1.7687132358551025, "learning_rate": 2.678e-05, "loss": 0.1514, "step": 1340 }, { "grad_norm": 1.8889333009719849, "learning_rate": 2.698e-05, "loss": 0.1854, "step": 1350 }, { "grad_norm": 2.479250907897949, "learning_rate": 2.718e-05, "loss": 0.1699, "step": 1360 }, { "grad_norm": 2.454704999923706, "learning_rate": 2.738e-05, "loss": 0.1627, "step": 1370 }, { "grad_norm": 2.0168662071228027, "learning_rate": 2.758e-05, "loss": 0.1597, "step": 1380 }, { "grad_norm": 2.818643808364868, "learning_rate": 2.778e-05, "loss": 0.1726, "step": 1390 }, { "grad_norm": 1.8962608575820923, "learning_rate": 2.798e-05, "loss": 0.1588, "step": 1400 }, { "grad_norm": 1.7337509393692017, "learning_rate": 2.818e-05, "loss": 0.1735, "step": 1410 }, { "grad_norm": 2.1658122539520264, "learning_rate": 2.8380000000000003e-05, "loss": 0.1835, "step": 1420 }, { "grad_norm": 1.5655173063278198, "learning_rate": 2.858e-05, "loss": 0.1489, "step": 1430 }, { "grad_norm": 1.6020054817199707, "learning_rate": 2.8780000000000002e-05, "loss": 0.1666, "step": 1440 }, { "grad_norm": 1.8239387273788452, "learning_rate": 2.898e-05, "loss": 0.1855, "step": 1450 }, { "grad_norm": 1.6526334285736084, "learning_rate": 2.9180000000000002e-05, "loss": 0.1885, "step": 1460 }, { "grad_norm": 2.0222830772399902, "learning_rate": 2.9380000000000003e-05, "loss": 0.1637, "step": 1470 }, { "grad_norm": 1.835827350616455, "learning_rate": 2.958e-05, "loss": 0.1477, "step": 1480 }, { "grad_norm": 1.8704229593276978, "learning_rate": 2.9780000000000003e-05, "loss": 0.1614, "step": 1490 }, { "grad_norm": 1.9323128461837769, "learning_rate": 2.998e-05, "loss": 0.1793, "step": 1500 }, { "grad_norm": 1.4997694492340088, "learning_rate": 3.0180000000000002e-05, "loss": 0.1589, "step": 1510 }, { "grad_norm": 2.2113840579986572, "learning_rate": 3.0380000000000004e-05, "loss": 0.179, "step": 1520 }, { "grad_norm": 1.5503649711608887, "learning_rate": 3.058e-05, "loss": 0.165, "step": 1530 }, { "grad_norm": 1.8989194631576538, "learning_rate": 3.078e-05, "loss": 0.1591, "step": 1540 }, { "grad_norm": 1.3033192157745361, "learning_rate": 3.0980000000000005e-05, "loss": 0.1394, "step": 1550 }, { "grad_norm": 2.518665075302124, "learning_rate": 3.118e-05, "loss": 0.188, "step": 1560 }, { "grad_norm": 1.6907649040222168, "learning_rate": 3.138e-05, "loss": 0.1531, "step": 1570 }, { "grad_norm": 1.6662031412124634, "learning_rate": 3.1580000000000006e-05, "loss": 0.1553, "step": 1580 }, { "grad_norm": 1.9404993057250977, "learning_rate": 3.1780000000000004e-05, "loss": 0.1822, "step": 1590 }, { "grad_norm": 2.2974061965942383, "learning_rate": 3.198e-05, "loss": 0.1609, "step": 1600 }, { "grad_norm": 1.799367070198059, "learning_rate": 3.218e-05, "loss": 0.1415, "step": 1610 }, { "grad_norm": 1.7656210660934448, "learning_rate": 3.238e-05, "loss": 0.1588, "step": 1620 }, { "grad_norm": 1.892472267150879, "learning_rate": 3.2579999999999996e-05, "loss": 0.1938, "step": 1630 }, { "grad_norm": 1.8163261413574219, "learning_rate": 3.278e-05, "loss": 0.163, "step": 1640 }, { "grad_norm": 2.160885810852051, "learning_rate": 3.298e-05, "loss": 0.1686, "step": 1650 }, { "grad_norm": 1.8545888662338257, "learning_rate": 3.318e-05, "loss": 0.1641, "step": 1660 }, { "grad_norm": 1.6920684576034546, "learning_rate": 3.338e-05, "loss": 0.1322, "step": 1670 }, { "grad_norm": 1.5214707851409912, "learning_rate": 3.358e-05, "loss": 0.1447, "step": 1680 }, { "grad_norm": 2.2308502197265625, "learning_rate": 3.378e-05, "loss": 0.1665, "step": 1690 }, { "grad_norm": 1.5913013219833374, "learning_rate": 3.398e-05, "loss": 0.1628, "step": 1700 }, { "grad_norm": 2.3169703483581543, "learning_rate": 3.418e-05, "loss": 0.1871, "step": 1710 }, { "grad_norm": 1.6720435619354248, "learning_rate": 3.438e-05, "loss": 0.1681, "step": 1720 }, { "grad_norm": 1.5319087505340576, "learning_rate": 3.4580000000000004e-05, "loss": 0.1624, "step": 1730 }, { "grad_norm": 2.0419209003448486, "learning_rate": 3.478e-05, "loss": 0.1463, "step": 1740 }, { "grad_norm": 1.8104362487792969, "learning_rate": 3.498e-05, "loss": 0.1614, "step": 1750 }, { "grad_norm": 2.210909843444824, "learning_rate": 3.518e-05, "loss": 0.14, "step": 1760 }, { "grad_norm": 1.5166616439819336, "learning_rate": 3.5380000000000003e-05, "loss": 0.1482, "step": 1770 }, { "grad_norm": 1.294833779335022, "learning_rate": 3.558e-05, "loss": 0.1554, "step": 1780 }, { "grad_norm": 1.5207653045654297, "learning_rate": 3.578e-05, "loss": 0.1343, "step": 1790 }, { "grad_norm": 1.7003600597381592, "learning_rate": 3.5980000000000004e-05, "loss": 0.1588, "step": 1800 }, { "grad_norm": 1.515056848526001, "learning_rate": 3.618e-05, "loss": 0.1396, "step": 1810 }, { "grad_norm": 2.173182725906372, "learning_rate": 3.638e-05, "loss": 0.1579, "step": 1820 }, { "grad_norm": 2.411917209625244, "learning_rate": 3.6580000000000006e-05, "loss": 0.1741, "step": 1830 }, { "grad_norm": 1.6638277769088745, "learning_rate": 3.6780000000000004e-05, "loss": 0.1678, "step": 1840 }, { "grad_norm": 1.498815894126892, "learning_rate": 3.698e-05, "loss": 0.1808, "step": 1850 }, { "grad_norm": 2.4209070205688477, "learning_rate": 3.7180000000000007e-05, "loss": 0.1778, "step": 1860 }, { "grad_norm": 1.016111969947815, "learning_rate": 3.7380000000000005e-05, "loss": 0.13, "step": 1870 }, { "grad_norm": 1.529037594795227, "learning_rate": 3.758e-05, "loss": 0.1252, "step": 1880 }, { "grad_norm": 1.5931735038757324, "learning_rate": 3.778000000000001e-05, "loss": 0.155, "step": 1890 }, { "grad_norm": 1.2025028467178345, "learning_rate": 3.7980000000000006e-05, "loss": 0.1424, "step": 1900 }, { "grad_norm": 2.2061703205108643, "learning_rate": 3.818e-05, "loss": 0.17, "step": 1910 }, { "grad_norm": 1.1986604928970337, "learning_rate": 3.838e-05, "loss": 0.1459, "step": 1920 }, { "grad_norm": 1.473808765411377, "learning_rate": 3.858e-05, "loss": 0.1618, "step": 1930 }, { "grad_norm": 1.9677156209945679, "learning_rate": 3.878e-05, "loss": 0.1538, "step": 1940 }, { "grad_norm": 1.3102089166641235, "learning_rate": 3.898e-05, "loss": 0.1332, "step": 1950 }, { "grad_norm": 1.4948713779449463, "learning_rate": 3.918e-05, "loss": 0.1689, "step": 1960 }, { "grad_norm": 1.3206911087036133, "learning_rate": 3.938e-05, "loss": 0.1376, "step": 1970 }, { "grad_norm": 1.8772028684616089, "learning_rate": 3.958e-05, "loss": 0.1807, "step": 1980 }, { "grad_norm": 2.0223987102508545, "learning_rate": 3.978e-05, "loss": 0.1359, "step": 1990 }, { "grad_norm": 1.118356466293335, "learning_rate": 3.998e-05, "loss": 0.1528, "step": 2000 }, { "grad_norm": 1.484339952468872, "learning_rate": 4.018e-05, "loss": 0.1905, "step": 2010 }, { "grad_norm": 1.7186830043792725, "learning_rate": 4.038e-05, "loss": 0.1499, "step": 2020 }, { "grad_norm": 1.4902114868164062, "learning_rate": 4.058e-05, "loss": 0.1416, "step": 2030 }, { "grad_norm": 1.2969028949737549, "learning_rate": 4.078e-05, "loss": 0.1635, "step": 2040 }, { "grad_norm": 1.5498123168945312, "learning_rate": 4.0980000000000004e-05, "loss": 0.1549, "step": 2050 }, { "grad_norm": 1.9674773216247559, "learning_rate": 4.118e-05, "loss": 0.1432, "step": 2060 }, { "grad_norm": 1.1921190023422241, "learning_rate": 4.138e-05, "loss": 0.1293, "step": 2070 }, { "grad_norm": 1.039497971534729, "learning_rate": 4.1580000000000005e-05, "loss": 0.1493, "step": 2080 }, { "grad_norm": 1.3206233978271484, "learning_rate": 4.178e-05, "loss": 0.1474, "step": 2090 }, { "grad_norm": 1.3809046745300293, "learning_rate": 4.198e-05, "loss": 0.1486, "step": 2100 }, { "grad_norm": 1.4369730949401855, "learning_rate": 4.2180000000000006e-05, "loss": 0.1745, "step": 2110 }, { "grad_norm": 1.5424410104751587, "learning_rate": 4.2380000000000004e-05, "loss": 0.1492, "step": 2120 }, { "grad_norm": 1.3721860647201538, "learning_rate": 4.258e-05, "loss": 0.1549, "step": 2130 }, { "grad_norm": 1.4496523141860962, "learning_rate": 4.278e-05, "loss": 0.1084, "step": 2140 }, { "grad_norm": 1.3654372692108154, "learning_rate": 4.2980000000000005e-05, "loss": 0.1682, "step": 2150 }, { "grad_norm": 1.1786983013153076, "learning_rate": 4.318e-05, "loss": 0.1449, "step": 2160 }, { "grad_norm": 0.911827802658081, "learning_rate": 4.338e-05, "loss": 0.1586, "step": 2170 }, { "grad_norm": 1.2714409828186035, "learning_rate": 4.3580000000000006e-05, "loss": 0.1668, "step": 2180 }, { "grad_norm": 1.204265832901001, "learning_rate": 4.3780000000000004e-05, "loss": 0.14, "step": 2190 }, { "grad_norm": 1.4662424325942993, "learning_rate": 4.398e-05, "loss": 0.1508, "step": 2200 }, { "grad_norm": 1.298799753189087, "learning_rate": 4.418000000000001e-05, "loss": 0.1453, "step": 2210 }, { "grad_norm": 1.5639859437942505, "learning_rate": 4.438e-05, "loss": 0.1332, "step": 2220 }, { "grad_norm": 1.4191402196884155, "learning_rate": 4.458e-05, "loss": 0.1505, "step": 2230 }, { "grad_norm": 1.0988991260528564, "learning_rate": 4.478e-05, "loss": 0.1336, "step": 2240 }, { "grad_norm": 1.2630939483642578, "learning_rate": 4.498e-05, "loss": 0.1615, "step": 2250 }, { "grad_norm": 1.4778082370758057, "learning_rate": 4.518e-05, "loss": 0.1306, "step": 2260 }, { "grad_norm": 1.2400002479553223, "learning_rate": 4.538e-05, "loss": 0.1404, "step": 2270 }, { "grad_norm": 1.360198736190796, "learning_rate": 4.558e-05, "loss": 0.1458, "step": 2280 }, { "grad_norm": 1.4954631328582764, "learning_rate": 4.578e-05, "loss": 0.1931, "step": 2290 }, { "grad_norm": 1.3447328805923462, "learning_rate": 4.5980000000000004e-05, "loss": 0.1344, "step": 2300 }, { "grad_norm": 1.2067679166793823, "learning_rate": 4.618e-05, "loss": 0.1377, "step": 2310 }, { "grad_norm": 1.526206612586975, "learning_rate": 4.638e-05, "loss": 0.1259, "step": 2320 }, { "grad_norm": 1.8354443311691284, "learning_rate": 4.6580000000000005e-05, "loss": 0.1575, "step": 2330 }, { "grad_norm": 1.2835040092468262, "learning_rate": 4.678e-05, "loss": 0.1364, "step": 2340 }, { "grad_norm": 0.9974143505096436, "learning_rate": 4.698e-05, "loss": 0.1423, "step": 2350 }, { "grad_norm": 1.0167367458343506, "learning_rate": 4.718e-05, "loss": 0.1502, "step": 2360 }, { "grad_norm": 1.2501623630523682, "learning_rate": 4.7380000000000004e-05, "loss": 0.1409, "step": 2370 }, { "grad_norm": 1.2125962972640991, "learning_rate": 4.758e-05, "loss": 0.1679, "step": 2380 }, { "grad_norm": 1.011176347732544, "learning_rate": 4.778e-05, "loss": 0.161, "step": 2390 }, { "grad_norm": 1.2078500986099243, "learning_rate": 4.7980000000000005e-05, "loss": 0.1386, "step": 2400 }, { "grad_norm": 0.9965975880622864, "learning_rate": 4.818e-05, "loss": 0.1492, "step": 2410 }, { "grad_norm": 1.1778655052185059, "learning_rate": 4.838e-05, "loss": 0.1868, "step": 2420 }, { "grad_norm": 1.3350353240966797, "learning_rate": 4.8580000000000006e-05, "loss": 0.1442, "step": 2430 }, { "grad_norm": 1.7902277708053589, "learning_rate": 4.8780000000000004e-05, "loss": 0.1547, "step": 2440 }, { "grad_norm": 1.3632577657699585, "learning_rate": 4.898e-05, "loss": 0.1182, "step": 2450 }, { "grad_norm": 1.0892266035079956, "learning_rate": 4.918000000000001e-05, "loss": 0.1338, "step": 2460 }, { "grad_norm": 1.9413392543792725, "learning_rate": 4.9380000000000005e-05, "loss": 0.1263, "step": 2470 }, { "grad_norm": 1.3337883949279785, "learning_rate": 4.958e-05, "loss": 0.1593, "step": 2480 }, { "grad_norm": 1.6492254734039307, "learning_rate": 4.978e-05, "loss": 0.1479, "step": 2490 }, { "grad_norm": 1.2550112009048462, "learning_rate": 4.9980000000000006e-05, "loss": 0.1394, "step": 2500 }, { "grad_norm": 0.934120774269104, "learning_rate": 5.0180000000000004e-05, "loss": 0.1166, "step": 2510 }, { "grad_norm": 1.1338491439819336, "learning_rate": 5.038e-05, "loss": 0.136, "step": 2520 }, { "grad_norm": 1.0920917987823486, "learning_rate": 5.058000000000001e-05, "loss": 0.1204, "step": 2530 }, { "grad_norm": 1.0532567501068115, "learning_rate": 5.0780000000000005e-05, "loss": 0.1378, "step": 2540 }, { "grad_norm": 1.3725870847702026, "learning_rate": 5.098e-05, "loss": 0.1499, "step": 2550 }, { "grad_norm": 0.8068469166755676, "learning_rate": 5.118000000000001e-05, "loss": 0.1433, "step": 2560 }, { "grad_norm": 1.0369091033935547, "learning_rate": 5.1380000000000006e-05, "loss": 0.1367, "step": 2570 }, { "grad_norm": 1.0220218896865845, "learning_rate": 5.1580000000000004e-05, "loss": 0.1499, "step": 2580 }, { "grad_norm": 0.8764722347259521, "learning_rate": 5.178000000000001e-05, "loss": 0.1248, "step": 2590 }, { "grad_norm": 0.9021809697151184, "learning_rate": 5.198000000000001e-05, "loss": 0.1354, "step": 2600 }, { "grad_norm": 1.0313009023666382, "learning_rate": 5.2180000000000005e-05, "loss": 0.1469, "step": 2610 }, { "grad_norm": 1.0240452289581299, "learning_rate": 5.238000000000001e-05, "loss": 0.1137, "step": 2620 }, { "grad_norm": 1.3650296926498413, "learning_rate": 5.258000000000001e-05, "loss": 0.1358, "step": 2630 }, { "grad_norm": 1.0514270067214966, "learning_rate": 5.2780000000000006e-05, "loss": 0.1354, "step": 2640 }, { "grad_norm": 1.0270189046859741, "learning_rate": 5.2980000000000004e-05, "loss": 0.1444, "step": 2650 }, { "grad_norm": 0.9946691989898682, "learning_rate": 5.318000000000001e-05, "loss": 0.1393, "step": 2660 }, { "grad_norm": 0.9106272459030151, "learning_rate": 5.338000000000001e-05, "loss": 0.15, "step": 2670 }, { "grad_norm": 1.145870327949524, "learning_rate": 5.3580000000000005e-05, "loss": 0.155, "step": 2680 }, { "grad_norm": 0.901906430721283, "learning_rate": 5.378e-05, "loss": 0.1285, "step": 2690 }, { "grad_norm": 1.2542273998260498, "learning_rate": 5.3979999999999995e-05, "loss": 0.1332, "step": 2700 }, { "grad_norm": 1.1694765090942383, "learning_rate": 5.418e-05, "loss": 0.1389, "step": 2710 }, { "grad_norm": 1.2834876775741577, "learning_rate": 5.438e-05, "loss": 0.1568, "step": 2720 }, { "grad_norm": 0.955867350101471, "learning_rate": 5.4579999999999996e-05, "loss": 0.1225, "step": 2730 }, { "grad_norm": 0.9468156695365906, "learning_rate": 5.478e-05, "loss": 0.1253, "step": 2740 }, { "grad_norm": 1.103444218635559, "learning_rate": 5.498e-05, "loss": 0.1499, "step": 2750 }, { "grad_norm": 1.2074106931686401, "learning_rate": 5.518e-05, "loss": 0.1498, "step": 2760 }, { "grad_norm": 1.4492613077163696, "learning_rate": 5.538e-05, "loss": 0.1447, "step": 2770 }, { "grad_norm": 1.2879247665405273, "learning_rate": 5.558e-05, "loss": 0.1467, "step": 2780 }, { "grad_norm": 1.1696854829788208, "learning_rate": 5.578e-05, "loss": 0.1428, "step": 2790 }, { "grad_norm": 1.072102427482605, "learning_rate": 5.5979999999999996e-05, "loss": 0.1259, "step": 2800 }, { "grad_norm": 0.9171052575111389, "learning_rate": 5.618e-05, "loss": 0.1646, "step": 2810 }, { "grad_norm": 1.2331427335739136, "learning_rate": 5.638e-05, "loss": 0.1379, "step": 2820 }, { "grad_norm": 0.749969482421875, "learning_rate": 5.658e-05, "loss": 0.1284, "step": 2830 }, { "grad_norm": 1.4788274765014648, "learning_rate": 5.678e-05, "loss": 0.1564, "step": 2840 }, { "grad_norm": 0.9650305509567261, "learning_rate": 5.698e-05, "loss": 0.151, "step": 2850 }, { "grad_norm": 1.052051305770874, "learning_rate": 5.718e-05, "loss": 0.1392, "step": 2860 }, { "grad_norm": 0.9744260907173157, "learning_rate": 5.738e-05, "loss": 0.1391, "step": 2870 }, { "grad_norm": 0.9245352745056152, "learning_rate": 5.758e-05, "loss": 0.1347, "step": 2880 }, { "grad_norm": 1.0018616914749146, "learning_rate": 5.778e-05, "loss": 0.1285, "step": 2890 }, { "grad_norm": 1.0095096826553345, "learning_rate": 5.7980000000000004e-05, "loss": 0.151, "step": 2900 }, { "grad_norm": 1.4693821668624878, "learning_rate": 5.818e-05, "loss": 0.1218, "step": 2910 }, { "grad_norm": 1.0671226978302002, "learning_rate": 5.838e-05, "loss": 0.1372, "step": 2920 }, { "grad_norm": 1.1393638849258423, "learning_rate": 5.858e-05, "loss": 0.1095, "step": 2930 }, { "grad_norm": 1.0051038265228271, "learning_rate": 5.878e-05, "loss": 0.1126, "step": 2940 }, { "grad_norm": 1.1293619871139526, "learning_rate": 5.898e-05, "loss": 0.148, "step": 2950 }, { "grad_norm": 0.822479248046875, "learning_rate": 5.918e-05, "loss": 0.1517, "step": 2960 }, { "grad_norm": 1.0195116996765137, "learning_rate": 5.9380000000000004e-05, "loss": 0.1151, "step": 2970 }, { "grad_norm": 1.048513412475586, "learning_rate": 5.958e-05, "loss": 0.1414, "step": 2980 }, { "grad_norm": 1.0660037994384766, "learning_rate": 5.978e-05, "loss": 0.1451, "step": 2990 }, { "grad_norm": 1.013846755027771, "learning_rate": 5.9980000000000005e-05, "loss": 0.1496, "step": 3000 }, { "grad_norm": 1.1453099250793457, "learning_rate": 6.018e-05, "loss": 0.1123, "step": 3010 }, { "grad_norm": 1.3683503866195679, "learning_rate": 6.038e-05, "loss": 0.1237, "step": 3020 }, { "grad_norm": 1.9718518257141113, "learning_rate": 6.0580000000000006e-05, "loss": 0.1154, "step": 3030 }, { "grad_norm": 0.7798859477043152, "learning_rate": 6.0780000000000004e-05, "loss": 0.1253, "step": 3040 }, { "grad_norm": 0.8521268963813782, "learning_rate": 6.098e-05, "loss": 0.1402, "step": 3050 }, { "grad_norm": 0.9051283597946167, "learning_rate": 6.118000000000001e-05, "loss": 0.1399, "step": 3060 }, { "grad_norm": 0.96747887134552, "learning_rate": 6.138e-05, "loss": 0.1014, "step": 3070 }, { "grad_norm": 1.2045093774795532, "learning_rate": 6.158e-05, "loss": 0.1355, "step": 3080 }, { "grad_norm": 0.9787114858627319, "learning_rate": 6.178000000000001e-05, "loss": 0.1477, "step": 3090 }, { "grad_norm": 1.4234079122543335, "learning_rate": 6.198e-05, "loss": 0.1343, "step": 3100 }, { "grad_norm": 1.0001394748687744, "learning_rate": 6.218e-05, "loss": 0.1157, "step": 3110 }, { "grad_norm": 1.0707814693450928, "learning_rate": 6.238000000000001e-05, "loss": 0.12, "step": 3120 }, { "grad_norm": 1.0366233587265015, "learning_rate": 6.258e-05, "loss": 0.1213, "step": 3130 }, { "grad_norm": 1.3225457668304443, "learning_rate": 6.278e-05, "loss": 0.1445, "step": 3140 }, { "grad_norm": 1.2268407344818115, "learning_rate": 6.298000000000001e-05, "loss": 0.1365, "step": 3150 }, { "grad_norm": 1.5974562168121338, "learning_rate": 6.318e-05, "loss": 0.136, "step": 3160 }, { "grad_norm": 1.207676649093628, "learning_rate": 6.338e-05, "loss": 0.1252, "step": 3170 }, { "grad_norm": 0.6624413728713989, "learning_rate": 6.358000000000001e-05, "loss": 0.1396, "step": 3180 }, { "grad_norm": 0.9676418304443359, "learning_rate": 6.378e-05, "loss": 0.1365, "step": 3190 }, { "grad_norm": 0.6820864677429199, "learning_rate": 6.398000000000001e-05, "loss": 0.1163, "step": 3200 }, { "grad_norm": 0.7851729989051819, "learning_rate": 6.418000000000001e-05, "loss": 0.1131, "step": 3210 }, { "grad_norm": 1.1592669486999512, "learning_rate": 6.438e-05, "loss": 0.1434, "step": 3220 }, { "grad_norm": 0.9831502437591553, "learning_rate": 6.458000000000001e-05, "loss": 0.1444, "step": 3230 }, { "grad_norm": 0.8144361972808838, "learning_rate": 6.478000000000001e-05, "loss": 0.1061, "step": 3240 }, { "grad_norm": 0.7268911004066467, "learning_rate": 6.498e-05, "loss": 0.1443, "step": 3250 }, { "grad_norm": 0.9120867252349854, "learning_rate": 6.518000000000001e-05, "loss": 0.1185, "step": 3260 }, { "grad_norm": 1.0019721984863281, "learning_rate": 6.538000000000001e-05, "loss": 0.1186, "step": 3270 }, { "grad_norm": 0.9531659483909607, "learning_rate": 6.558e-05, "loss": 0.1134, "step": 3280 }, { "grad_norm": 0.885077178478241, "learning_rate": 6.578000000000001e-05, "loss": 0.129, "step": 3290 }, { "grad_norm": 0.8348997831344604, "learning_rate": 6.598e-05, "loss": 0.112, "step": 3300 }, { "grad_norm": 0.9166581034660339, "learning_rate": 6.618e-05, "loss": 0.1294, "step": 3310 }, { "grad_norm": 0.9142540097236633, "learning_rate": 6.638e-05, "loss": 0.1414, "step": 3320 }, { "grad_norm": 0.9303827881813049, "learning_rate": 6.658e-05, "loss": 0.1151, "step": 3330 }, { "grad_norm": 0.8739721775054932, "learning_rate": 6.678e-05, "loss": 0.1087, "step": 3340 }, { "grad_norm": 0.7763402462005615, "learning_rate": 6.698e-05, "loss": 0.1099, "step": 3350 }, { "grad_norm": 0.7974467873573303, "learning_rate": 6.718e-05, "loss": 0.1131, "step": 3360 }, { "grad_norm": 1.1086950302124023, "learning_rate": 6.738e-05, "loss": 0.1174, "step": 3370 }, { "grad_norm": 1.0140881538391113, "learning_rate": 6.758e-05, "loss": 0.1669, "step": 3380 }, { "grad_norm": 1.0410892963409424, "learning_rate": 6.778e-05, "loss": 0.1244, "step": 3390 }, { "grad_norm": 2.193319797515869, "learning_rate": 6.798e-05, "loss": 0.1335, "step": 3400 }, { "grad_norm": 1.0268808603286743, "learning_rate": 6.818e-05, "loss": 0.1242, "step": 3410 }, { "grad_norm": 1.4365521669387817, "learning_rate": 6.838e-05, "loss": 0.1271, "step": 3420 }, { "grad_norm": 0.971278727054596, "learning_rate": 6.858e-05, "loss": 0.12, "step": 3430 }, { "grad_norm": 0.8352065682411194, "learning_rate": 6.878e-05, "loss": 0.1412, "step": 3440 }, { "grad_norm": 0.9304540753364563, "learning_rate": 6.898e-05, "loss": 0.1195, "step": 3450 }, { "grad_norm": 0.6810506582260132, "learning_rate": 6.918e-05, "loss": 0.1081, "step": 3460 }, { "grad_norm": 1.1806529760360718, "learning_rate": 6.938e-05, "loss": 0.1543, "step": 3470 }, { "grad_norm": 0.7576145529747009, "learning_rate": 6.958e-05, "loss": 0.1389, "step": 3480 }, { "grad_norm": 1.0792121887207031, "learning_rate": 6.978e-05, "loss": 0.1293, "step": 3490 }, { "grad_norm": 0.9848029613494873, "learning_rate": 6.998e-05, "loss": 0.121, "step": 3500 }, { "grad_norm": 0.7988865971565247, "learning_rate": 7.018e-05, "loss": 0.127, "step": 3510 }, { "grad_norm": 0.8640171885490417, "learning_rate": 7.038e-05, "loss": 0.1294, "step": 3520 }, { "grad_norm": 1.4798474311828613, "learning_rate": 7.058e-05, "loss": 0.1289, "step": 3530 }, { "grad_norm": 0.8693019151687622, "learning_rate": 7.078e-05, "loss": 0.1184, "step": 3540 }, { "grad_norm": 0.818209171295166, "learning_rate": 7.098e-05, "loss": 0.1301, "step": 3550 }, { "grad_norm": 0.8386824727058411, "learning_rate": 7.118e-05, "loss": 0.1054, "step": 3560 }, { "grad_norm": 0.7377365231513977, "learning_rate": 7.138e-05, "loss": 0.1361, "step": 3570 }, { "grad_norm": 0.8842955827713013, "learning_rate": 7.158e-05, "loss": 0.1392, "step": 3580 }, { "grad_norm": 1.0598783493041992, "learning_rate": 7.178000000000001e-05, "loss": 0.1334, "step": 3590 }, { "grad_norm": 0.9092361330986023, "learning_rate": 7.198e-05, "loss": 0.1212, "step": 3600 }, { "grad_norm": 1.0006873607635498, "learning_rate": 7.218e-05, "loss": 0.1319, "step": 3610 }, { "grad_norm": 0.9288651943206787, "learning_rate": 7.238000000000001e-05, "loss": 0.1269, "step": 3620 }, { "grad_norm": 0.9101852774620056, "learning_rate": 7.258e-05, "loss": 0.1203, "step": 3630 }, { "grad_norm": 1.2454873323440552, "learning_rate": 7.278e-05, "loss": 0.1281, "step": 3640 }, { "grad_norm": 1.0717108249664307, "learning_rate": 7.298000000000001e-05, "loss": 0.1533, "step": 3650 }, { "grad_norm": 0.8773550391197205, "learning_rate": 7.318e-05, "loss": 0.1256, "step": 3660 }, { "grad_norm": 0.9743457436561584, "learning_rate": 7.338e-05, "loss": 0.1555, "step": 3670 }, { "grad_norm": 0.887226939201355, "learning_rate": 7.358000000000001e-05, "loss": 0.1377, "step": 3680 }, { "grad_norm": 0.8402469158172607, "learning_rate": 7.378e-05, "loss": 0.1291, "step": 3690 }, { "grad_norm": 1.0275211334228516, "learning_rate": 7.398e-05, "loss": 0.1095, "step": 3700 }, { "grad_norm": 0.9039138555526733, "learning_rate": 7.418000000000001e-05, "loss": 0.1276, "step": 3710 }, { "grad_norm": 1.0156646966934204, "learning_rate": 7.438e-05, "loss": 0.1244, "step": 3720 }, { "grad_norm": 0.8830580711364746, "learning_rate": 7.458000000000001e-05, "loss": 0.1261, "step": 3730 }, { "grad_norm": 0.8695610165596008, "learning_rate": 7.478e-05, "loss": 0.1147, "step": 3740 }, { "grad_norm": 1.0456292629241943, "learning_rate": 7.498e-05, "loss": 0.133, "step": 3750 }, { "grad_norm": 0.7072977423667908, "learning_rate": 7.518000000000001e-05, "loss": 0.1103, "step": 3760 }, { "grad_norm": 1.1195286512374878, "learning_rate": 7.538e-05, "loss": 0.1255, "step": 3770 }, { "grad_norm": 0.8446875214576721, "learning_rate": 7.558e-05, "loss": 0.118, "step": 3780 }, { "grad_norm": 0.8646713495254517, "learning_rate": 7.578000000000001e-05, "loss": 0.1205, "step": 3790 }, { "grad_norm": 0.858875572681427, "learning_rate": 7.598e-05, "loss": 0.1601, "step": 3800 }, { "grad_norm": 1.0679351091384888, "learning_rate": 7.618e-05, "loss": 0.1286, "step": 3810 }, { "grad_norm": 0.8327000141143799, "learning_rate": 7.638000000000001e-05, "loss": 0.1324, "step": 3820 }, { "grad_norm": 1.1590311527252197, "learning_rate": 7.658e-05, "loss": 0.1339, "step": 3830 }, { "grad_norm": 1.2492202520370483, "learning_rate": 7.678000000000001e-05, "loss": 0.1167, "step": 3840 }, { "grad_norm": 1.2065457105636597, "learning_rate": 7.698000000000001e-05, "loss": 0.1496, "step": 3850 }, { "grad_norm": 1.1240663528442383, "learning_rate": 7.718e-05, "loss": 0.1182, "step": 3860 }, { "grad_norm": 0.8275880813598633, "learning_rate": 7.738000000000001e-05, "loss": 0.1192, "step": 3870 }, { "grad_norm": 0.7966970205307007, "learning_rate": 7.758000000000001e-05, "loss": 0.1291, "step": 3880 }, { "grad_norm": 0.9573147892951965, "learning_rate": 7.778e-05, "loss": 0.1226, "step": 3890 }, { "grad_norm": 0.9154716730117798, "learning_rate": 7.798000000000001e-05, "loss": 0.1256, "step": 3900 }, { "grad_norm": 1.0389207601547241, "learning_rate": 7.818000000000001e-05, "loss": 0.1325, "step": 3910 }, { "grad_norm": 1.3391849994659424, "learning_rate": 7.838e-05, "loss": 0.1434, "step": 3920 }, { "grad_norm": 0.6097239255905151, "learning_rate": 7.858000000000001e-05, "loss": 0.1106, "step": 3930 }, { "grad_norm": 0.8953226208686829, "learning_rate": 7.878e-05, "loss": 0.1346, "step": 3940 }, { "grad_norm": 0.9161803126335144, "learning_rate": 7.897999999999999e-05, "loss": 0.1254, "step": 3950 }, { "grad_norm": 0.868558406829834, "learning_rate": 7.918e-05, "loss": 0.1414, "step": 3960 }, { "grad_norm": 0.9133217334747314, "learning_rate": 7.938e-05, "loss": 0.1119, "step": 3970 }, { "grad_norm": 0.8137634992599487, "learning_rate": 7.958e-05, "loss": 0.1274, "step": 3980 }, { "grad_norm": 1.0631729364395142, "learning_rate": 7.978e-05, "loss": 0.1465, "step": 3990 }, { "grad_norm": 1.065684199333191, "learning_rate": 7.998e-05, "loss": 0.1156, "step": 4000 }, { "grad_norm": 0.740567147731781, "learning_rate": 8.018e-05, "loss": 0.123, "step": 4010 }, { "grad_norm": 0.6977625489234924, "learning_rate": 8.038e-05, "loss": 0.1144, "step": 4020 }, { "grad_norm": 1.371604323387146, "learning_rate": 8.058e-05, "loss": 0.1296, "step": 4030 }, { "grad_norm": 1.0815871953964233, "learning_rate": 8.078e-05, "loss": 0.1056, "step": 4040 }, { "grad_norm": 0.681778609752655, "learning_rate": 8.098e-05, "loss": 0.1246, "step": 4050 }, { "grad_norm": 1.0329557657241821, "learning_rate": 8.118e-05, "loss": 0.1389, "step": 4060 }, { "grad_norm": 1.1097177267074585, "learning_rate": 8.138e-05, "loss": 0.1243, "step": 4070 }, { "grad_norm": 0.7777077555656433, "learning_rate": 8.158e-05, "loss": 0.1227, "step": 4080 }, { "grad_norm": 0.6955239176750183, "learning_rate": 8.178e-05, "loss": 0.1169, "step": 4090 }, { "grad_norm": 0.7929132580757141, "learning_rate": 8.198e-05, "loss": 0.1221, "step": 4100 }, { "grad_norm": 1.0469409227371216, "learning_rate": 8.218e-05, "loss": 0.1261, "step": 4110 }, { "grad_norm": 0.9672197103500366, "learning_rate": 8.238000000000001e-05, "loss": 0.1343, "step": 4120 }, { "grad_norm": 0.6930410861968994, "learning_rate": 8.258e-05, "loss": 0.1267, "step": 4130 }, { "grad_norm": 1.1181491613388062, "learning_rate": 8.278e-05, "loss": 0.1298, "step": 4140 }, { "grad_norm": 0.8251775503158569, "learning_rate": 8.298000000000001e-05, "loss": 0.1328, "step": 4150 }, { "grad_norm": 0.8624600768089294, "learning_rate": 8.318e-05, "loss": 0.1135, "step": 4160 }, { "grad_norm": 0.8188206553459167, "learning_rate": 8.338e-05, "loss": 0.1156, "step": 4170 }, { "grad_norm": 1.5977686643600464, "learning_rate": 8.358e-05, "loss": 0.1195, "step": 4180 }, { "grad_norm": 1.3531616926193237, "learning_rate": 8.378e-05, "loss": 0.1273, "step": 4190 }, { "grad_norm": 0.6988369226455688, "learning_rate": 8.398e-05, "loss": 0.1277, "step": 4200 }, { "grad_norm": 0.9405227899551392, "learning_rate": 8.418e-05, "loss": 0.1222, "step": 4210 }, { "grad_norm": 0.8639581799507141, "learning_rate": 8.438e-05, "loss": 0.1282, "step": 4220 }, { "grad_norm": 0.9261648058891296, "learning_rate": 8.458e-05, "loss": 0.139, "step": 4230 }, { "grad_norm": 0.9019935131072998, "learning_rate": 8.478e-05, "loss": 0.1336, "step": 4240 }, { "grad_norm": 1.1266398429870605, "learning_rate": 8.498e-05, "loss": 0.1456, "step": 4250 }, { "grad_norm": 0.7474331259727478, "learning_rate": 8.518000000000001e-05, "loss": 0.1357, "step": 4260 }, { "grad_norm": 0.7591966390609741, "learning_rate": 8.538e-05, "loss": 0.1273, "step": 4270 }, { "grad_norm": 0.8251168131828308, "learning_rate": 8.558e-05, "loss": 0.1108, "step": 4280 }, { "grad_norm": 1.0645939111709595, "learning_rate": 8.578000000000001e-05, "loss": 0.1106, "step": 4290 }, { "grad_norm": 0.7694222927093506, "learning_rate": 8.598e-05, "loss": 0.1046, "step": 4300 }, { "grad_norm": 0.7720732688903809, "learning_rate": 8.618e-05, "loss": 0.126, "step": 4310 }, { "grad_norm": 1.082078218460083, "learning_rate": 8.638000000000001e-05, "loss": 0.132, "step": 4320 }, { "grad_norm": 0.9850585460662842, "learning_rate": 8.658e-05, "loss": 0.1147, "step": 4330 }, { "grad_norm": 0.9314911961555481, "learning_rate": 8.678e-05, "loss": 0.1031, "step": 4340 }, { "grad_norm": 1.101064920425415, "learning_rate": 8.698000000000001e-05, "loss": 0.1277, "step": 4350 }, { "grad_norm": 0.8495332598686218, "learning_rate": 8.718e-05, "loss": 0.1202, "step": 4360 }, { "grad_norm": 0.8741047382354736, "learning_rate": 8.738000000000001e-05, "loss": 0.1532, "step": 4370 }, { "grad_norm": 0.7280544638633728, "learning_rate": 8.758000000000001e-05, "loss": 0.1125, "step": 4380 }, { "grad_norm": 0.6018278002738953, "learning_rate": 8.778e-05, "loss": 0.1067, "step": 4390 }, { "grad_norm": 0.7994418740272522, "learning_rate": 8.798000000000001e-05, "loss": 0.1362, "step": 4400 }, { "grad_norm": 0.5727084279060364, "learning_rate": 8.818000000000001e-05, "loss": 0.1179, "step": 4410 }, { "grad_norm": 0.6630362868309021, "learning_rate": 8.838e-05, "loss": 0.1276, "step": 4420 }, { "grad_norm": 1.2539128065109253, "learning_rate": 8.858000000000001e-05, "loss": 0.1218, "step": 4430 }, { "grad_norm": 0.7600860595703125, "learning_rate": 8.878000000000001e-05, "loss": 0.1217, "step": 4440 }, { "grad_norm": 0.7381993532180786, "learning_rate": 8.898e-05, "loss": 0.1402, "step": 4450 }, { "grad_norm": 0.7390010356903076, "learning_rate": 8.918000000000001e-05, "loss": 0.1297, "step": 4460 }, { "grad_norm": 0.8174942135810852, "learning_rate": 8.938e-05, "loss": 0.1164, "step": 4470 }, { "grad_norm": 0.686290442943573, "learning_rate": 8.958e-05, "loss": 0.1457, "step": 4480 }, { "grad_norm": 0.8980652689933777, "learning_rate": 8.978000000000001e-05, "loss": 0.1167, "step": 4490 }, { "grad_norm": 0.8899937868118286, "learning_rate": 8.998e-05, "loss": 0.1191, "step": 4500 }, { "grad_norm": 0.8233340978622437, "learning_rate": 9.018000000000001e-05, "loss": 0.1403, "step": 4510 }, { "grad_norm": 0.8600900769233704, "learning_rate": 9.038000000000001e-05, "loss": 0.1314, "step": 4520 }, { "grad_norm": 0.7801728844642639, "learning_rate": 9.058e-05, "loss": 0.1126, "step": 4530 }, { "grad_norm": 0.9974315166473389, "learning_rate": 9.078000000000001e-05, "loss": 0.1445, "step": 4540 }, { "grad_norm": 0.7019931077957153, "learning_rate": 9.098000000000001e-05, "loss": 0.1415, "step": 4550 }, { "grad_norm": 0.8548590540885925, "learning_rate": 9.118e-05, "loss": 0.1182, "step": 4560 }, { "grad_norm": 0.9260722994804382, "learning_rate": 9.138e-05, "loss": 0.1417, "step": 4570 }, { "grad_norm": 1.07524836063385, "learning_rate": 9.158e-05, "loss": 0.1357, "step": 4580 }, { "grad_norm": 1.0134586095809937, "learning_rate": 9.178e-05, "loss": 0.1236, "step": 4590 }, { "grad_norm": 0.6599453687667847, "learning_rate": 9.198e-05, "loss": 0.1262, "step": 4600 }, { "grad_norm": 0.8587327599525452, "learning_rate": 9.218e-05, "loss": 0.1394, "step": 4610 }, { "grad_norm": 0.8813931941986084, "learning_rate": 9.238e-05, "loss": 0.1309, "step": 4620 }, { "grad_norm": 1.0108038187026978, "learning_rate": 9.258e-05, "loss": 0.1419, "step": 4630 }, { "grad_norm": 0.7282737493515015, "learning_rate": 9.278e-05, "loss": 0.1279, "step": 4640 }, { "grad_norm": 0.8769378066062927, "learning_rate": 9.298e-05, "loss": 0.1198, "step": 4650 }, { "grad_norm": 0.9639085531234741, "learning_rate": 9.318e-05, "loss": 0.1252, "step": 4660 }, { "grad_norm": 1.0832815170288086, "learning_rate": 9.338e-05, "loss": 0.1288, "step": 4670 }, { "grad_norm": 1.0542241334915161, "learning_rate": 9.358e-05, "loss": 0.1302, "step": 4680 }, { "grad_norm": 0.7635073065757751, "learning_rate": 9.378e-05, "loss": 0.146, "step": 4690 }, { "grad_norm": 0.8174298405647278, "learning_rate": 9.398e-05, "loss": 0.1202, "step": 4700 }, { "grad_norm": 0.7352933883666992, "learning_rate": 9.418e-05, "loss": 0.1235, "step": 4710 }, { "grad_norm": 0.7789106965065002, "learning_rate": 9.438e-05, "loss": 0.1138, "step": 4720 }, { "grad_norm": 0.8112280368804932, "learning_rate": 9.458e-05, "loss": 0.1501, "step": 4730 }, { "grad_norm": 0.9860435724258423, "learning_rate": 9.478e-05, "loss": 0.1394, "step": 4740 }, { "grad_norm": 0.7318524122238159, "learning_rate": 9.498e-05, "loss": 0.1113, "step": 4750 }, { "grad_norm": 0.7842848896980286, "learning_rate": 9.518000000000001e-05, "loss": 0.1209, "step": 4760 }, { "grad_norm": 0.7748559713363647, "learning_rate": 9.538e-05, "loss": 0.1225, "step": 4770 }, { "grad_norm": 0.8918251991271973, "learning_rate": 9.558e-05, "loss": 0.1404, "step": 4780 }, { "grad_norm": 0.8908780217170715, "learning_rate": 9.578000000000001e-05, "loss": 0.1014, "step": 4790 }, { "grad_norm": 0.748260498046875, "learning_rate": 9.598e-05, "loss": 0.1158, "step": 4800 }, { "grad_norm": 1.0323406457901, "learning_rate": 9.618e-05, "loss": 0.1219, "step": 4810 }, { "grad_norm": 0.6863218545913696, "learning_rate": 9.638000000000001e-05, "loss": 0.1147, "step": 4820 }, { "grad_norm": 0.8324155807495117, "learning_rate": 9.658e-05, "loss": 0.1195, "step": 4830 }, { "grad_norm": 0.9350183010101318, "learning_rate": 9.678e-05, "loss": 0.1291, "step": 4840 }, { "grad_norm": 0.7406456470489502, "learning_rate": 9.698000000000001e-05, "loss": 0.1202, "step": 4850 }, { "grad_norm": 0.8074131011962891, "learning_rate": 9.718e-05, "loss": 0.1085, "step": 4860 }, { "grad_norm": 0.7790124416351318, "learning_rate": 9.738e-05, "loss": 0.113, "step": 4870 }, { "grad_norm": 0.8743554949760437, "learning_rate": 9.758000000000001e-05, "loss": 0.121, "step": 4880 }, { "grad_norm": 1.009361743927002, "learning_rate": 9.778e-05, "loss": 0.1479, "step": 4890 }, { "grad_norm": 1.01250159740448, "learning_rate": 9.798000000000001e-05, "loss": 0.1237, "step": 4900 }, { "grad_norm": 0.7195025682449341, "learning_rate": 9.818000000000001e-05, "loss": 0.11, "step": 4910 }, { "grad_norm": 0.8566808104515076, "learning_rate": 9.838e-05, "loss": 0.1293, "step": 4920 }, { "grad_norm": 0.747453510761261, "learning_rate": 9.858000000000001e-05, "loss": 0.1347, "step": 4930 }, { "grad_norm": 0.590721845626831, "learning_rate": 9.878e-05, "loss": 0.1328, "step": 4940 }, { "grad_norm": 1.488671064376831, "learning_rate": 9.898e-05, "loss": 0.1341, "step": 4950 }, { "grad_norm": 0.694701075553894, "learning_rate": 9.918000000000001e-05, "loss": 0.1411, "step": 4960 }, { "grad_norm": 0.7694219350814819, "learning_rate": 9.938e-05, "loss": 0.145, "step": 4970 }, { "grad_norm": 0.6370720267295837, "learning_rate": 9.958e-05, "loss": 0.1393, "step": 4980 }, { "grad_norm": 0.8336985111236572, "learning_rate": 9.978000000000001e-05, "loss": 0.1217, "step": 4990 }, { "grad_norm": 0.8597829937934875, "learning_rate": 9.998e-05, "loss": 0.1415, "step": 5000 }, { "grad_norm": 0.7965955138206482, "learning_rate": 9.999999778549045e-05, "loss": 0.1307, "step": 5010 }, { "grad_norm": 0.7805999517440796, "learning_rate": 9.999999013039593e-05, "loss": 0.1264, "step": 5020 }, { "grad_norm": 0.8059232234954834, "learning_rate": 9.999997700737766e-05, "loss": 0.1333, "step": 5030 }, { "grad_norm": 0.7773878574371338, "learning_rate": 9.999995841643709e-05, "loss": 0.1152, "step": 5040 }, { "grad_norm": 0.8747403025627136, "learning_rate": 9.999993435757623e-05, "loss": 0.1348, "step": 5050 }, { "grad_norm": 0.8386016488075256, "learning_rate": 9.999990483079773e-05, "loss": 0.1045, "step": 5060 }, { "grad_norm": 0.8992975354194641, "learning_rate": 9.999986983610481e-05, "loss": 0.1369, "step": 5070 }, { "grad_norm": 0.988239586353302, "learning_rate": 9.99998293735013e-05, "loss": 0.1422, "step": 5080 }, { "grad_norm": 0.9136562943458557, "learning_rate": 9.999978344299161e-05, "loss": 0.1301, "step": 5090 }, { "grad_norm": 1.0883816480636597, "learning_rate": 9.99997320445808e-05, "loss": 0.1463, "step": 5100 }, { "grad_norm": 0.7157883644104004, "learning_rate": 9.999967517827444e-05, "loss": 0.1286, "step": 5110 }, { "grad_norm": 0.9233700633049011, "learning_rate": 9.999961284407879e-05, "loss": 0.1393, "step": 5120 }, { "grad_norm": 0.9364409446716309, "learning_rate": 9.999954504200067e-05, "loss": 0.1241, "step": 5130 }, { "grad_norm": 0.8195462226867676, "learning_rate": 9.999947177204744e-05, "loss": 0.1217, "step": 5140 }, { "grad_norm": 0.7723621129989624, "learning_rate": 9.999939303422718e-05, "loss": 0.1078, "step": 5150 }, { "grad_norm": 0.7673085927963257, "learning_rate": 9.999930882854847e-05, "loss": 0.1367, "step": 5160 }, { "grad_norm": 0.9088662266731262, "learning_rate": 9.999921915502051e-05, "loss": 0.1191, "step": 5170 }, { "grad_norm": 0.8567297458648682, "learning_rate": 9.99991240136531e-05, "loss": 0.1304, "step": 5180 }, { "grad_norm": 0.8236304521560669, "learning_rate": 9.999902340445668e-05, "loss": 0.1505, "step": 5190 }, { "grad_norm": 0.6094697713851929, "learning_rate": 9.999891732744224e-05, "loss": 0.1121, "step": 5200 }, { "grad_norm": 1.145401954650879, "learning_rate": 9.999880578262135e-05, "loss": 0.13, "step": 5210 }, { "grad_norm": 0.7238115072250366, "learning_rate": 9.999868877000624e-05, "loss": 0.1219, "step": 5220 }, { "grad_norm": 0.7501238584518433, "learning_rate": 9.99985662896097e-05, "loss": 0.1266, "step": 5230 }, { "grad_norm": 0.7312914729118347, "learning_rate": 9.999843834144513e-05, "loss": 0.1219, "step": 5240 }, { "grad_norm": 0.8951148390769958, "learning_rate": 9.99983049255265e-05, "loss": 0.1389, "step": 5250 }, { "grad_norm": 0.7238635420799255, "learning_rate": 9.999816604186843e-05, "loss": 0.1228, "step": 5260 }, { "grad_norm": 0.8946648240089417, "learning_rate": 9.999802169048609e-05, "loss": 0.1179, "step": 5270 }, { "grad_norm": 0.8110727667808533, "learning_rate": 9.999787187139527e-05, "loss": 0.126, "step": 5280 }, { "grad_norm": 1.3602198362350464, "learning_rate": 9.999771658461234e-05, "loss": 0.1225, "step": 5290 }, { "grad_norm": 0.9229188561439514, "learning_rate": 9.999755583015431e-05, "loss": 0.1165, "step": 5300 }, { "grad_norm": 1.0744552612304688, "learning_rate": 9.999738960803874e-05, "loss": 0.1204, "step": 5310 }, { "grad_norm": 0.825234055519104, "learning_rate": 9.99972179182838e-05, "loss": 0.1243, "step": 5320 }, { "grad_norm": 1.0203521251678467, "learning_rate": 9.99970407609083e-05, "loss": 0.1489, "step": 5330 }, { "grad_norm": 1.1254745721817017, "learning_rate": 9.999685813593159e-05, "loss": 0.1296, "step": 5340 }, { "grad_norm": 0.6163673400878906, "learning_rate": 9.999667004337362e-05, "loss": 0.1037, "step": 5350 }, { "grad_norm": 0.8265554308891296, "learning_rate": 9.9996476483255e-05, "loss": 0.14, "step": 5360 }, { "grad_norm": 0.7337207794189453, "learning_rate": 9.999627745559688e-05, "loss": 0.1017, "step": 5370 }, { "grad_norm": 1.0118019580841064, "learning_rate": 9.999607296042101e-05, "loss": 0.1068, "step": 5380 }, { "grad_norm": 1.0362396240234375, "learning_rate": 9.99958629977498e-05, "loss": 0.1135, "step": 5390 }, { "grad_norm": 0.8317501544952393, "learning_rate": 9.999564756760615e-05, "loss": 0.1028, "step": 5400 }, { "grad_norm": 0.49388036131858826, "learning_rate": 9.999542667001366e-05, "loss": 0.0952, "step": 5410 }, { "grad_norm": 0.6897886395454407, "learning_rate": 9.999520030499647e-05, "loss": 0.114, "step": 5420 }, { "grad_norm": 0.9575714468955994, "learning_rate": 9.999496847257936e-05, "loss": 0.0976, "step": 5430 }, { "grad_norm": 0.9008278846740723, "learning_rate": 9.999473117278764e-05, "loss": 0.1249, "step": 5440 }, { "grad_norm": 0.7779574394226074, "learning_rate": 9.999448840564731e-05, "loss": 0.1005, "step": 5450 }, { "grad_norm": 0.5869126915931702, "learning_rate": 9.999424017118488e-05, "loss": 0.1003, "step": 5460 }, { "grad_norm": 0.5574921369552612, "learning_rate": 9.999398646942751e-05, "loss": 0.0964, "step": 5470 }, { "grad_norm": 0.6402806639671326, "learning_rate": 9.999372730040296e-05, "loss": 0.1056, "step": 5480 }, { "grad_norm": 0.9397607445716858, "learning_rate": 9.999346266413953e-05, "loss": 0.1276, "step": 5490 }, { "grad_norm": 0.7753353714942932, "learning_rate": 9.99931925606662e-05, "loss": 0.1276, "step": 5500 }, { "grad_norm": 0.8112704157829285, "learning_rate": 9.99929169900125e-05, "loss": 0.1181, "step": 5510 }, { "grad_norm": 0.9652935266494751, "learning_rate": 9.999263595220855e-05, "loss": 0.1003, "step": 5520 }, { "grad_norm": 0.9066533446311951, "learning_rate": 9.99923494472851e-05, "loss": 0.1179, "step": 5530 }, { "grad_norm": 0.628116250038147, "learning_rate": 9.999205747527348e-05, "loss": 0.0989, "step": 5540 }, { "grad_norm": 0.8200430870056152, "learning_rate": 9.999176003620561e-05, "loss": 0.1121, "step": 5550 }, { "grad_norm": 0.9216690063476562, "learning_rate": 9.999145713011405e-05, "loss": 0.1217, "step": 5560 }, { "grad_norm": 0.7976466417312622, "learning_rate": 9.999114875703186e-05, "loss": 0.1137, "step": 5570 }, { "grad_norm": 1.3702178001403809, "learning_rate": 9.999083491699281e-05, "loss": 0.1352, "step": 5580 }, { "grad_norm": 0.8018144965171814, "learning_rate": 9.999051561003123e-05, "loss": 0.1418, "step": 5590 }, { "grad_norm": 0.6641047596931458, "learning_rate": 9.999019083618202e-05, "loss": 0.1328, "step": 5600 }, { "grad_norm": 0.6992581486701965, "learning_rate": 9.99898605954807e-05, "loss": 0.1294, "step": 5610 }, { "grad_norm": 1.0200265645980835, "learning_rate": 9.998952488796338e-05, "loss": 0.1097, "step": 5620 }, { "grad_norm": 0.9592655897140503, "learning_rate": 9.998918371366676e-05, "loss": 0.1181, "step": 5630 }, { "grad_norm": 0.7333269715309143, "learning_rate": 9.99888370726282e-05, "loss": 0.1124, "step": 5640 }, { "grad_norm": 0.9539761543273926, "learning_rate": 9.998848496488556e-05, "loss": 0.1423, "step": 5650 }, { "grad_norm": 0.702399492263794, "learning_rate": 9.998812739047736e-05, "loss": 0.1024, "step": 5660 }, { "grad_norm": 0.7301600575447083, "learning_rate": 9.99877643494427e-05, "loss": 0.1135, "step": 5670 }, { "grad_norm": 0.7247178554534912, "learning_rate": 9.998739584182128e-05, "loss": 0.1216, "step": 5680 }, { "grad_norm": 1.1607660055160522, "learning_rate": 9.998702186765342e-05, "loss": 0.1101, "step": 5690 }, { "grad_norm": 0.6966094970703125, "learning_rate": 9.998664242698e-05, "loss": 0.1113, "step": 5700 }, { "grad_norm": 0.6662808656692505, "learning_rate": 9.998625751984251e-05, "loss": 0.1172, "step": 5710 }, { "grad_norm": 0.62990403175354, "learning_rate": 9.998586714628307e-05, "loss": 0.1023, "step": 5720 }, { "grad_norm": 0.5920556783676147, "learning_rate": 9.998547130634432e-05, "loss": 0.1163, "step": 5730 }, { "grad_norm": 0.6731451749801636, "learning_rate": 9.99850700000696e-05, "loss": 0.1296, "step": 5740 }, { "grad_norm": 0.9072434306144714, "learning_rate": 9.998466322750278e-05, "loss": 0.115, "step": 5750 }, { "grad_norm": 0.555267870426178, "learning_rate": 9.998425098868834e-05, "loss": 0.1092, "step": 5760 }, { "grad_norm": 0.8171665668487549, "learning_rate": 9.998383328367136e-05, "loss": 0.128, "step": 5770 }, { "grad_norm": 0.7420629262924194, "learning_rate": 9.99834101124975e-05, "loss": 0.1337, "step": 5780 }, { "grad_norm": 0.6687968373298645, "learning_rate": 9.998298147521309e-05, "loss": 0.1082, "step": 5790 }, { "grad_norm": 0.7406355738639832, "learning_rate": 9.998254737186496e-05, "loss": 0.1377, "step": 5800 }, { "grad_norm": 0.6449888348579407, "learning_rate": 9.99821078025006e-05, "loss": 0.1266, "step": 5810 }, { "grad_norm": 0.8682430386543274, "learning_rate": 9.998166276716807e-05, "loss": 0.1195, "step": 5820 }, { "grad_norm": 0.6073471903800964, "learning_rate": 9.998121226591606e-05, "loss": 0.1229, "step": 5830 }, { "grad_norm": 0.8002408742904663, "learning_rate": 9.998075629879382e-05, "loss": 0.121, "step": 5840 }, { "grad_norm": 1.041951298713684, "learning_rate": 9.99802948658512e-05, "loss": 0.1153, "step": 5850 }, { "grad_norm": 0.8126014471054077, "learning_rate": 9.99798279671387e-05, "loss": 0.1292, "step": 5860 }, { "grad_norm": 0.5861836075782776, "learning_rate": 9.997935560270734e-05, "loss": 0.1073, "step": 5870 }, { "grad_norm": 0.8085689544677734, "learning_rate": 9.997887777260879e-05, "loss": 0.1248, "step": 5880 }, { "grad_norm": 0.5967084765434265, "learning_rate": 9.997839447689532e-05, "loss": 0.1324, "step": 5890 }, { "grad_norm": 0.7226241230964661, "learning_rate": 9.997790571561978e-05, "loss": 0.1288, "step": 5900 }, { "grad_norm": 0.7241542339324951, "learning_rate": 9.99774114888356e-05, "loss": 0.1448, "step": 5910 }, { "grad_norm": 0.7323874831199646, "learning_rate": 9.997691179659684e-05, "loss": 0.1196, "step": 5920 }, { "grad_norm": 0.5521307587623596, "learning_rate": 9.997640663895815e-05, "loss": 0.0997, "step": 5930 }, { "grad_norm": 1.282619595527649, "learning_rate": 9.997589601597477e-05, "loss": 0.13, "step": 5940 }, { "grad_norm": 0.712039053440094, "learning_rate": 9.997537992770252e-05, "loss": 0.1122, "step": 5950 }, { "grad_norm": 1.2250233888626099, "learning_rate": 9.997485837419788e-05, "loss": 0.1163, "step": 5960 }, { "grad_norm": 0.8074131011962891, "learning_rate": 9.997433135551786e-05, "loss": 0.1371, "step": 5970 }, { "grad_norm": 0.7480318546295166, "learning_rate": 9.997379887172009e-05, "loss": 0.1101, "step": 5980 }, { "grad_norm": 0.707831621170044, "learning_rate": 9.997326092286281e-05, "loss": 0.1372, "step": 5990 }, { "grad_norm": 0.664130687713623, "learning_rate": 9.997271750900486e-05, "loss": 0.1411, "step": 6000 }, { "grad_norm": 0.694898784160614, "learning_rate": 9.997216863020565e-05, "loss": 0.1235, "step": 6010 }, { "grad_norm": 0.7486855983734131, "learning_rate": 9.99716142865252e-05, "loss": 0.1278, "step": 6020 }, { "grad_norm": 0.9695789217948914, "learning_rate": 9.997105447802415e-05, "loss": 0.1286, "step": 6030 }, { "grad_norm": 0.6348419785499573, "learning_rate": 9.997048920476373e-05, "loss": 0.1024, "step": 6040 }, { "grad_norm": 0.8883818984031677, "learning_rate": 9.996991846680572e-05, "loss": 0.116, "step": 6050 }, { "grad_norm": 0.5976727604866028, "learning_rate": 9.996934226421257e-05, "loss": 0.1109, "step": 6060 }, { "grad_norm": 0.6815117001533508, "learning_rate": 9.996876059704726e-05, "loss": 0.115, "step": 6070 }, { "grad_norm": 0.7129004597663879, "learning_rate": 9.996817346537343e-05, "loss": 0.1244, "step": 6080 }, { "grad_norm": 0.5811524987220764, "learning_rate": 9.996758086925526e-05, "loss": 0.1207, "step": 6090 }, { "grad_norm": 0.6503962278366089, "learning_rate": 9.996698280875759e-05, "loss": 0.1361, "step": 6100 }, { "grad_norm": 0.5890845060348511, "learning_rate": 9.99663792839458e-05, "loss": 0.1016, "step": 6110 }, { "grad_norm": 1.0201342105865479, "learning_rate": 9.99657702948859e-05, "loss": 0.1298, "step": 6120 }, { "grad_norm": 0.9941924810409546, "learning_rate": 9.996515584164448e-05, "loss": 0.11, "step": 6130 }, { "grad_norm": 0.9371047019958496, "learning_rate": 9.996453592428873e-05, "loss": 0.125, "step": 6140 }, { "grad_norm": 0.648521900177002, "learning_rate": 9.996391054288646e-05, "loss": 0.1435, "step": 6150 }, { "grad_norm": 0.9010971784591675, "learning_rate": 9.996327969750605e-05, "loss": 0.141, "step": 6160 }, { "grad_norm": 0.6948021650314331, "learning_rate": 9.996264338821649e-05, "loss": 0.1187, "step": 6170 }, { "grad_norm": 1.0414139032363892, "learning_rate": 9.996200161508735e-05, "loss": 0.1058, "step": 6180 }, { "grad_norm": 1.0328900814056396, "learning_rate": 9.996135437818885e-05, "loss": 0.1308, "step": 6190 }, { "grad_norm": 0.6147107481956482, "learning_rate": 9.996070167759175e-05, "loss": 0.1037, "step": 6200 }, { "grad_norm": 0.7473794221878052, "learning_rate": 9.996004351336743e-05, "loss": 0.1195, "step": 6210 }, { "grad_norm": 1.2042683362960815, "learning_rate": 9.995937988558785e-05, "loss": 0.114, "step": 6220 }, { "grad_norm": 0.6565845608711243, "learning_rate": 9.995871079432561e-05, "loss": 0.1344, "step": 6230 }, { "grad_norm": 0.964577853679657, "learning_rate": 9.995803623965389e-05, "loss": 0.1232, "step": 6240 }, { "grad_norm": 1.288548231124878, "learning_rate": 9.995735622164641e-05, "loss": 0.1296, "step": 6250 }, { "grad_norm": 0.7540532350540161, "learning_rate": 9.995667074037758e-05, "loss": 0.0854, "step": 6260 }, { "grad_norm": 1.0522816181182861, "learning_rate": 9.995597979592232e-05, "loss": 0.1152, "step": 6270 }, { "grad_norm": 0.7395334839820862, "learning_rate": 9.995528338835625e-05, "loss": 0.1048, "step": 6280 }, { "grad_norm": 0.7035649418830872, "learning_rate": 9.995458151775547e-05, "loss": 0.0967, "step": 6290 }, { "grad_norm": 0.8658909201622009, "learning_rate": 9.995387418419677e-05, "loss": 0.1165, "step": 6300 }, { "grad_norm": 0.662102222442627, "learning_rate": 9.99531613877575e-05, "loss": 0.1011, "step": 6310 }, { "grad_norm": 0.7158274054527283, "learning_rate": 9.995244312851559e-05, "loss": 0.1009, "step": 6320 }, { "grad_norm": 0.814746081829071, "learning_rate": 9.995171940654961e-05, "loss": 0.1157, "step": 6330 }, { "grad_norm": 0.8238362073898315, "learning_rate": 9.995099022193871e-05, "loss": 0.1159, "step": 6340 }, { "grad_norm": 0.7947238087654114, "learning_rate": 9.995025557476261e-05, "loss": 0.1013, "step": 6350 }, { "grad_norm": 0.5835409164428711, "learning_rate": 9.994951546510165e-05, "loss": 0.1055, "step": 6360 }, { "grad_norm": 0.5217671394348145, "learning_rate": 9.994876989303679e-05, "loss": 0.1058, "step": 6370 }, { "grad_norm": 0.6872901916503906, "learning_rate": 9.994801885864955e-05, "loss": 0.1031, "step": 6380 }, { "grad_norm": 1.1314911842346191, "learning_rate": 9.994726236202205e-05, "loss": 0.1292, "step": 6390 }, { "grad_norm": 0.7140918970108032, "learning_rate": 9.994650040323704e-05, "loss": 0.1381, "step": 6400 }, { "grad_norm": 0.561817467212677, "learning_rate": 9.994573298237784e-05, "loss": 0.0917, "step": 6410 }, { "grad_norm": 0.5761979818344116, "learning_rate": 9.994496009952837e-05, "loss": 0.1068, "step": 6420 }, { "grad_norm": 0.7313412427902222, "learning_rate": 9.994418175477316e-05, "loss": 0.1158, "step": 6430 }, { "grad_norm": 0.9691706299781799, "learning_rate": 9.994339794819733e-05, "loss": 0.1092, "step": 6440 }, { "grad_norm": 0.6203092932701111, "learning_rate": 9.994260867988658e-05, "loss": 0.1117, "step": 6450 }, { "grad_norm": 0.5419911742210388, "learning_rate": 9.994181394992723e-05, "loss": 0.1043, "step": 6460 }, { "grad_norm": 0.6320573091506958, "learning_rate": 9.994101375840618e-05, "loss": 0.0998, "step": 6470 }, { "grad_norm": 0.7574360370635986, "learning_rate": 9.994020810541098e-05, "loss": 0.0884, "step": 6480 }, { "grad_norm": 0.8061933517456055, "learning_rate": 9.99393969910297e-05, "loss": 0.1286, "step": 6490 }, { "grad_norm": 0.8992792367935181, "learning_rate": 9.993858041535104e-05, "loss": 0.1007, "step": 6500 }, { "grad_norm": 0.556893527507782, "learning_rate": 9.99377583784643e-05, "loss": 0.0983, "step": 6510 }, { "grad_norm": 0.8459044694900513, "learning_rate": 9.993693088045939e-05, "loss": 0.1101, "step": 6520 }, { "grad_norm": 0.7852718234062195, "learning_rate": 9.99360979214268e-05, "loss": 0.0998, "step": 6530 }, { "grad_norm": 0.5549641251564026, "learning_rate": 9.99352595014576e-05, "loss": 0.0946, "step": 6540 }, { "grad_norm": 1.069543480873108, "learning_rate": 9.993441562064354e-05, "loss": 0.1021, "step": 6550 }, { "grad_norm": 0.752232015132904, "learning_rate": 9.993356627907685e-05, "loss": 0.1001, "step": 6560 }, { "grad_norm": 0.5835233926773071, "learning_rate": 9.99327114768504e-05, "loss": 0.1092, "step": 6570 }, { "grad_norm": 0.7325602769851685, "learning_rate": 9.99318512140577e-05, "loss": 0.1015, "step": 6580 }, { "grad_norm": 0.8654665946960449, "learning_rate": 9.993098549079284e-05, "loss": 0.1273, "step": 6590 }, { "grad_norm": 0.7755988240242004, "learning_rate": 9.993011430715047e-05, "loss": 0.11, "step": 6600 }, { "grad_norm": 0.8433821201324463, "learning_rate": 9.992923766322586e-05, "loss": 0.0981, "step": 6610 }, { "grad_norm": 0.65975421667099, "learning_rate": 9.99283555591149e-05, "loss": 0.0878, "step": 6620 }, { "grad_norm": 0.621774435043335, "learning_rate": 9.992746799491404e-05, "loss": 0.1137, "step": 6630 }, { "grad_norm": 0.7892833948135376, "learning_rate": 9.992657497072033e-05, "loss": 0.1036, "step": 6640 }, { "grad_norm": 0.978022575378418, "learning_rate": 9.992567648663147e-05, "loss": 0.1419, "step": 6650 }, { "grad_norm": 0.587442934513092, "learning_rate": 9.992477254274568e-05, "loss": 0.1081, "step": 6660 }, { "grad_norm": 0.7455980181694031, "learning_rate": 9.992386313916183e-05, "loss": 0.1003, "step": 6670 }, { "grad_norm": 0.6935366988182068, "learning_rate": 9.992294827597934e-05, "loss": 0.1266, "step": 6680 }, { "grad_norm": 0.7139703631401062, "learning_rate": 9.992202795329831e-05, "loss": 0.1081, "step": 6690 }, { "grad_norm": 0.8006525039672852, "learning_rate": 9.992110217121936e-05, "loss": 0.0901, "step": 6700 }, { "grad_norm": 0.9518051743507385, "learning_rate": 9.992017092984372e-05, "loss": 0.1275, "step": 6710 }, { "grad_norm": 0.5613544583320618, "learning_rate": 9.991923422927326e-05, "loss": 0.1187, "step": 6720 }, { "grad_norm": 0.596446692943573, "learning_rate": 9.991829206961037e-05, "loss": 0.116, "step": 6730 }, { "grad_norm": 0.6030424237251282, "learning_rate": 9.991734445095813e-05, "loss": 0.1299, "step": 6740 }, { "grad_norm": 0.8293495774269104, "learning_rate": 9.991639137342015e-05, "loss": 0.1026, "step": 6750 }, { "grad_norm": 0.726409912109375, "learning_rate": 9.991543283710064e-05, "loss": 0.1023, "step": 6760 }, { "grad_norm": 0.6132926940917969, "learning_rate": 9.991446884210445e-05, "loss": 0.1074, "step": 6770 }, { "grad_norm": 0.7211570143699646, "learning_rate": 9.9913499388537e-05, "loss": 0.1215, "step": 6780 }, { "grad_norm": 0.7051658630371094, "learning_rate": 9.99125244765043e-05, "loss": 0.0982, "step": 6790 }, { "grad_norm": 0.6629385352134705, "learning_rate": 9.991154410611296e-05, "loss": 0.1295, "step": 6800 }, { "grad_norm": 0.5842995643615723, "learning_rate": 9.99105582774702e-05, "loss": 0.1034, "step": 6810 }, { "grad_norm": 0.8101098537445068, "learning_rate": 9.990956699068384e-05, "loss": 0.1266, "step": 6820 }, { "grad_norm": 0.5886058807373047, "learning_rate": 9.990857024586224e-05, "loss": 0.0889, "step": 6830 }, { "grad_norm": 0.6968764662742615, "learning_rate": 9.990756804311446e-05, "loss": 0.1086, "step": 6840 }, { "grad_norm": 0.5483924150466919, "learning_rate": 9.990656038255006e-05, "loss": 0.1068, "step": 6850 }, { "grad_norm": 0.77763831615448, "learning_rate": 9.990554726427926e-05, "loss": 0.1022, "step": 6860 }, { "grad_norm": 0.6800340414047241, "learning_rate": 9.990452868841284e-05, "loss": 0.1468, "step": 6870 }, { "grad_norm": 0.6538575291633606, "learning_rate": 9.99035046550622e-05, "loss": 0.1353, "step": 6880 }, { "grad_norm": 0.6112777590751648, "learning_rate": 9.99024751643393e-05, "loss": 0.1128, "step": 6890 }, { "grad_norm": 0.6728190183639526, "learning_rate": 9.990144021635677e-05, "loss": 0.1086, "step": 6900 }, { "grad_norm": 0.9313942193984985, "learning_rate": 9.990039981122775e-05, "loss": 0.0902, "step": 6910 }, { "grad_norm": 1.012069582939148, "learning_rate": 9.989935394906602e-05, "loss": 0.1057, "step": 6920 }, { "grad_norm": 1.1900042295455933, "learning_rate": 9.989830262998598e-05, "loss": 0.1224, "step": 6930 }, { "grad_norm": 0.7808098196983337, "learning_rate": 9.989724585410259e-05, "loss": 0.0971, "step": 6940 }, { "grad_norm": 0.9429182410240173, "learning_rate": 9.989618362153139e-05, "loss": 0.1004, "step": 6950 }, { "grad_norm": 0.607153594493866, "learning_rate": 9.989511593238859e-05, "loss": 0.1096, "step": 6960 }, { "grad_norm": 0.5641108155250549, "learning_rate": 9.98940427867909e-05, "loss": 0.1296, "step": 6970 }, { "grad_norm": 0.5232972502708435, "learning_rate": 9.989296418485573e-05, "loss": 0.1126, "step": 6980 }, { "grad_norm": 0.5714141726493835, "learning_rate": 9.989188012670101e-05, "loss": 0.0935, "step": 6990 }, { "grad_norm": 0.7658705711364746, "learning_rate": 9.989079061244528e-05, "loss": 0.0961, "step": 7000 }, { "grad_norm": 0.4802422523498535, "learning_rate": 9.988969564220769e-05, "loss": 0.1191, "step": 7010 }, { "grad_norm": 0.7541338205337524, "learning_rate": 9.988859521610801e-05, "loss": 0.0956, "step": 7020 }, { "grad_norm": 0.7258757948875427, "learning_rate": 9.988748933426656e-05, "loss": 0.1351, "step": 7030 }, { "grad_norm": 0.6400884389877319, "learning_rate": 9.988637799680428e-05, "loss": 0.1003, "step": 7040 }, { "grad_norm": 0.6287823915481567, "learning_rate": 9.98852612038427e-05, "loss": 0.0845, "step": 7050 }, { "grad_norm": 0.6686583161354065, "learning_rate": 9.988413895550397e-05, "loss": 0.094, "step": 7060 }, { "grad_norm": 0.7766510248184204, "learning_rate": 9.98830112519108e-05, "loss": 0.1107, "step": 7070 }, { "grad_norm": 0.5099185109138489, "learning_rate": 9.98818780931865e-05, "loss": 0.1224, "step": 7080 }, { "grad_norm": 0.6829915046691895, "learning_rate": 9.988073947945502e-05, "loss": 0.0954, "step": 7090 }, { "grad_norm": 0.7623870372772217, "learning_rate": 9.987959541084087e-05, "loss": 0.0976, "step": 7100 }, { "grad_norm": 0.6556048393249512, "learning_rate": 9.987844588746915e-05, "loss": 0.1121, "step": 7110 }, { "grad_norm": 0.6888645887374878, "learning_rate": 9.987729090946558e-05, "loss": 0.0978, "step": 7120 }, { "grad_norm": 0.506439208984375, "learning_rate": 9.987613047695647e-05, "loss": 0.0798, "step": 7130 }, { "grad_norm": 0.9923000335693359, "learning_rate": 9.987496459006871e-05, "loss": 0.1051, "step": 7140 }, { "grad_norm": 0.8087775111198425, "learning_rate": 9.987379324892982e-05, "loss": 0.1363, "step": 7150 }, { "grad_norm": 0.8249192237854004, "learning_rate": 9.987261645366788e-05, "loss": 0.1214, "step": 7160 }, { "grad_norm": 0.5981132388114929, "learning_rate": 9.987143420441158e-05, "loss": 0.1257, "step": 7170 }, { "grad_norm": 0.8988814949989319, "learning_rate": 9.987024650129022e-05, "loss": 0.1119, "step": 7180 }, { "grad_norm": 0.4252508878707886, "learning_rate": 9.986905334443368e-05, "loss": 0.1104, "step": 7190 }, { "grad_norm": 0.8178889751434326, "learning_rate": 9.986785473397245e-05, "loss": 0.1151, "step": 7200 }, { "grad_norm": 0.5311968922615051, "learning_rate": 9.98666506700376e-05, "loss": 0.1085, "step": 7210 }, { "grad_norm": 0.7115851044654846, "learning_rate": 9.986544115276081e-05, "loss": 0.1085, "step": 7220 }, { "grad_norm": 0.9514298439025879, "learning_rate": 9.986422618227433e-05, "loss": 0.1194, "step": 7230 }, { "grad_norm": 1.0879358053207397, "learning_rate": 9.986300575871106e-05, "loss": 0.1186, "step": 7240 }, { "grad_norm": 1.1102224588394165, "learning_rate": 9.986177988220444e-05, "loss": 0.1131, "step": 7250 }, { "grad_norm": 0.5540478229522705, "learning_rate": 9.986054855288856e-05, "loss": 0.0999, "step": 7260 }, { "grad_norm": 0.5746134519577026, "learning_rate": 9.985931177089802e-05, "loss": 0.1165, "step": 7270 }, { "grad_norm": 1.0142121315002441, "learning_rate": 9.985806953636814e-05, "loss": 0.1123, "step": 7280 }, { "grad_norm": 0.7715495228767395, "learning_rate": 9.985682184943471e-05, "loss": 0.1128, "step": 7290 }, { "grad_norm": 0.5204121470451355, "learning_rate": 9.98555687102342e-05, "loss": 0.1059, "step": 7300 }, { "grad_norm": 0.5932508111000061, "learning_rate": 9.985431011890367e-05, "loss": 0.1226, "step": 7310 }, { "grad_norm": 0.9034420847892761, "learning_rate": 9.985304607558075e-05, "loss": 0.1014, "step": 7320 }, { "grad_norm": 0.6295595169067383, "learning_rate": 9.985177658040364e-05, "loss": 0.1067, "step": 7330 }, { "grad_norm": 0.6840813755989075, "learning_rate": 9.985050163351119e-05, "loss": 0.1067, "step": 7340 }, { "grad_norm": 1.035874366760254, "learning_rate": 9.984922123504286e-05, "loss": 0.0928, "step": 7350 }, { "grad_norm": 0.8838710188865662, "learning_rate": 9.984793538513862e-05, "loss": 0.1043, "step": 7360 }, { "grad_norm": 0.6107223629951477, "learning_rate": 9.984664408393912e-05, "loss": 0.0987, "step": 7370 }, { "grad_norm": 0.6101014018058777, "learning_rate": 9.984534733158556e-05, "loss": 0.1142, "step": 7380 }, { "grad_norm": 0.9922347664833069, "learning_rate": 9.984404512821977e-05, "loss": 0.1141, "step": 7390 }, { "grad_norm": 0.6618730425834656, "learning_rate": 9.984273747398411e-05, "loss": 0.1157, "step": 7400 }, { "grad_norm": 0.7814968228340149, "learning_rate": 9.984142436902165e-05, "loss": 0.1164, "step": 7410 }, { "grad_norm": 0.5846209526062012, "learning_rate": 9.984010581347596e-05, "loss": 0.118, "step": 7420 }, { "grad_norm": 0.4901368319988251, "learning_rate": 9.983878180749121e-05, "loss": 0.1131, "step": 7430 }, { "grad_norm": 0.6900031566619873, "learning_rate": 9.983745235121222e-05, "loss": 0.0953, "step": 7440 }, { "grad_norm": 0.8769095540046692, "learning_rate": 9.983611744478438e-05, "loss": 0.1012, "step": 7450 }, { "grad_norm": 0.8622338175773621, "learning_rate": 9.983477708835365e-05, "loss": 0.0984, "step": 7460 }, { "grad_norm": 0.7723805904388428, "learning_rate": 9.983343128206664e-05, "loss": 0.1287, "step": 7470 }, { "grad_norm": 0.8298131823539734, "learning_rate": 9.983208002607049e-05, "loss": 0.1234, "step": 7480 }, { "grad_norm": 0.814434289932251, "learning_rate": 9.9830723320513e-05, "loss": 0.1104, "step": 7490 }, { "grad_norm": 0.8116990923881531, "learning_rate": 9.982936116554254e-05, "loss": 0.1137, "step": 7500 }, { "grad_norm": 0.581466019153595, "learning_rate": 9.982799356130803e-05, "loss": 0.0969, "step": 7510 }, { "grad_norm": 1.2287850379943848, "learning_rate": 9.982662050795908e-05, "loss": 0.0992, "step": 7520 }, { "grad_norm": 0.7631779909133911, "learning_rate": 9.982524200564583e-05, "loss": 0.0977, "step": 7530 }, { "grad_norm": 0.5566184520721436, "learning_rate": 9.982385805451901e-05, "loss": 0.1014, "step": 7540 }, { "grad_norm": 0.6487776637077332, "learning_rate": 9.982246865472998e-05, "loss": 0.0811, "step": 7550 }, { "grad_norm": 0.5688511729240417, "learning_rate": 9.982107380643069e-05, "loss": 0.0963, "step": 7560 }, { "grad_norm": 0.7942225337028503, "learning_rate": 9.981967350977368e-05, "loss": 0.1226, "step": 7570 }, { "grad_norm": 0.9356681108474731, "learning_rate": 9.981826776491208e-05, "loss": 0.099, "step": 7580 }, { "grad_norm": 0.7838374376296997, "learning_rate": 9.98168565719996e-05, "loss": 0.1241, "step": 7590 }, { "grad_norm": 0.5456328392028809, "learning_rate": 9.98154399311906e-05, "loss": 0.1056, "step": 7600 }, { "grad_norm": 0.6352971196174622, "learning_rate": 9.981401784263997e-05, "loss": 0.117, "step": 7610 }, { "grad_norm": 0.7170358896255493, "learning_rate": 9.981259030650326e-05, "loss": 0.1109, "step": 7620 }, { "grad_norm": 0.6229663491249084, "learning_rate": 9.981115732293655e-05, "loss": 0.1024, "step": 7630 }, { "grad_norm": 0.5718104243278503, "learning_rate": 9.980971889209659e-05, "loss": 0.1182, "step": 7640 }, { "grad_norm": 0.4560031592845917, "learning_rate": 9.980827501414064e-05, "loss": 0.0939, "step": 7650 }, { "grad_norm": 0.8125146627426147, "learning_rate": 9.980682568922663e-05, "loss": 0.1005, "step": 7660 }, { "grad_norm": 0.7822394967079163, "learning_rate": 9.980537091751304e-05, "loss": 0.1005, "step": 7670 }, { "grad_norm": 0.6683025360107422, "learning_rate": 9.980391069915897e-05, "loss": 0.0985, "step": 7680 }, { "grad_norm": 0.9345847368240356, "learning_rate": 9.98024450343241e-05, "loss": 0.113, "step": 7690 }, { "grad_norm": 0.6876957416534424, "learning_rate": 9.980097392316872e-05, "loss": 0.0928, "step": 7700 }, { "grad_norm": 0.7057859301567078, "learning_rate": 9.97994973658537e-05, "loss": 0.1151, "step": 7710 }, { "grad_norm": 0.4898459315299988, "learning_rate": 9.979801536254054e-05, "loss": 0.0923, "step": 7720 }, { "grad_norm": 0.48689892888069153, "learning_rate": 9.979652791339127e-05, "loss": 0.1019, "step": 7730 }, { "grad_norm": 0.5078398585319519, "learning_rate": 9.97950350185686e-05, "loss": 0.0878, "step": 7740 }, { "grad_norm": 0.4612702429294586, "learning_rate": 9.979353667823574e-05, "loss": 0.0995, "step": 7750 }, { "grad_norm": 0.6433703899383545, "learning_rate": 9.979203289255658e-05, "loss": 0.1167, "step": 7760 }, { "grad_norm": 0.7522123456001282, "learning_rate": 9.979052366169557e-05, "loss": 0.1053, "step": 7770 }, { "grad_norm": 0.5727981925010681, "learning_rate": 9.978900898581775e-05, "loss": 0.1079, "step": 7780 }, { "grad_norm": 0.7353187203407288, "learning_rate": 9.978748886508875e-05, "loss": 0.09, "step": 7790 }, { "grad_norm": 0.6950666308403015, "learning_rate": 9.978596329967484e-05, "loss": 0.1008, "step": 7800 }, { "grad_norm": 0.8629905581474304, "learning_rate": 9.978443228974284e-05, "loss": 0.1083, "step": 7810 }, { "grad_norm": 0.5860158205032349, "learning_rate": 9.978289583546015e-05, "loss": 0.0954, "step": 7820 }, { "grad_norm": 0.9976047873497009, "learning_rate": 9.978135393699484e-05, "loss": 0.0954, "step": 7830 }, { "grad_norm": 0.5356164574623108, "learning_rate": 9.977980659451548e-05, "loss": 0.1021, "step": 7840 }, { "grad_norm": 0.6996997594833374, "learning_rate": 9.977825380819135e-05, "loss": 0.0909, "step": 7850 }, { "grad_norm": 0.6459090709686279, "learning_rate": 9.97766955781922e-05, "loss": 0.1029, "step": 7860 }, { "grad_norm": 0.7628395557403564, "learning_rate": 9.977513190468848e-05, "loss": 0.1122, "step": 7870 }, { "grad_norm": 0.7793400883674622, "learning_rate": 9.977356278785116e-05, "loss": 0.1118, "step": 7880 }, { "grad_norm": 1.119162678718567, "learning_rate": 9.977198822785184e-05, "loss": 0.1037, "step": 7890 }, { "grad_norm": 0.6923109889030457, "learning_rate": 9.977040822486273e-05, "loss": 0.1116, "step": 7900 }, { "grad_norm": 0.69620680809021, "learning_rate": 9.97688227790566e-05, "loss": 0.1175, "step": 7910 }, { "grad_norm": 0.5662199854850769, "learning_rate": 9.976723189060684e-05, "loss": 0.1241, "step": 7920 }, { "grad_norm": 0.5330952405929565, "learning_rate": 9.976563555968742e-05, "loss": 0.0812, "step": 7930 }, { "grad_norm": 0.8391351699829102, "learning_rate": 9.976403378647292e-05, "loss": 0.1106, "step": 7940 }, { "grad_norm": 0.7841686606407166, "learning_rate": 9.97624265711385e-05, "loss": 0.1081, "step": 7950 }, { "grad_norm": 0.7189682126045227, "learning_rate": 9.976081391385993e-05, "loss": 0.1166, "step": 7960 }, { "grad_norm": 0.7061703205108643, "learning_rate": 9.975919581481356e-05, "loss": 0.099, "step": 7970 }, { "grad_norm": 0.8697271943092346, "learning_rate": 9.975757227417634e-05, "loss": 0.0913, "step": 7980 }, { "grad_norm": 0.849801242351532, "learning_rate": 9.975594329212586e-05, "loss": 0.1042, "step": 7990 }, { "grad_norm": 0.8496559858322144, "learning_rate": 9.97543088688402e-05, "loss": 0.1195, "step": 8000 }, { "grad_norm": 0.8961564302444458, "learning_rate": 9.975266900449814e-05, "loss": 0.1057, "step": 8010 }, { "grad_norm": 0.6627760529518127, "learning_rate": 9.975102369927898e-05, "loss": 0.1098, "step": 8020 }, { "grad_norm": 0.6266964077949524, "learning_rate": 9.974937295336269e-05, "loss": 0.1043, "step": 8030 }, { "grad_norm": 0.8486300706863403, "learning_rate": 9.974771676692975e-05, "loss": 0.1031, "step": 8040 }, { "grad_norm": 0.5857316255569458, "learning_rate": 9.974605514016131e-05, "loss": 0.1126, "step": 8050 }, { "grad_norm": 0.9259185791015625, "learning_rate": 9.974438807323907e-05, "loss": 0.0928, "step": 8060 }, { "grad_norm": 0.7047679424285889, "learning_rate": 9.974271556634535e-05, "loss": 0.0871, "step": 8070 }, { "grad_norm": 0.9173114895820618, "learning_rate": 9.974103761966302e-05, "loss": 0.1144, "step": 8080 }, { "grad_norm": 0.5835506319999695, "learning_rate": 9.973935423337563e-05, "loss": 0.1182, "step": 8090 }, { "grad_norm": 0.7149184346199036, "learning_rate": 9.973766540766722e-05, "loss": 0.1171, "step": 8100 }, { "grad_norm": 0.7974682450294495, "learning_rate": 9.97359711427225e-05, "loss": 0.1065, "step": 8110 }, { "grad_norm": 0.7056710720062256, "learning_rate": 9.973427143872677e-05, "loss": 0.1269, "step": 8120 }, { "grad_norm": 1.0548120737075806, "learning_rate": 9.973256629586589e-05, "loss": 0.1194, "step": 8130 }, { "grad_norm": 0.6672472357749939, "learning_rate": 9.973085571432632e-05, "loss": 0.1264, "step": 8140 }, { "grad_norm": 0.7087439298629761, "learning_rate": 9.972913969429513e-05, "loss": 0.1157, "step": 8150 }, { "grad_norm": 0.5842851400375366, "learning_rate": 9.972741823596e-05, "loss": 0.1064, "step": 8160 }, { "grad_norm": 0.7374566197395325, "learning_rate": 9.972569133950917e-05, "loss": 0.1114, "step": 8170 }, { "grad_norm": 0.6519105434417725, "learning_rate": 9.972395900513151e-05, "loss": 0.0894, "step": 8180 }, { "grad_norm": 0.6684066653251648, "learning_rate": 9.972222123301645e-05, "loss": 0.0969, "step": 8190 }, { "grad_norm": 0.7879840731620789, "learning_rate": 9.972047802335403e-05, "loss": 0.0985, "step": 8200 }, { "grad_norm": 0.5757349133491516, "learning_rate": 9.971872937633488e-05, "loss": 0.0854, "step": 8210 }, { "grad_norm": 0.5529012084007263, "learning_rate": 9.971697529215024e-05, "loss": 0.0989, "step": 8220 }, { "grad_norm": 0.9517374038696289, "learning_rate": 9.971521577099192e-05, "loss": 0.0923, "step": 8230 }, { "grad_norm": 0.5697001218795776, "learning_rate": 9.971345081305236e-05, "loss": 0.1018, "step": 8240 }, { "grad_norm": 0.7999972701072693, "learning_rate": 9.971168041852456e-05, "loss": 0.1034, "step": 8250 }, { "grad_norm": 0.4628225862979889, "learning_rate": 9.970990458760215e-05, "loss": 0.0741, "step": 8260 }, { "grad_norm": 0.4225960075855255, "learning_rate": 9.970812332047929e-05, "loss": 0.0948, "step": 8270 }, { "grad_norm": 0.6361268758773804, "learning_rate": 9.97063366173508e-05, "loss": 0.0979, "step": 8280 }, { "grad_norm": 0.5875067710876465, "learning_rate": 9.970454447841207e-05, "loss": 0.0845, "step": 8290 }, { "grad_norm": 0.5274432897567749, "learning_rate": 9.970274690385909e-05, "loss": 0.0954, "step": 8300 }, { "grad_norm": 0.5099095106124878, "learning_rate": 9.970094389388844e-05, "loss": 0.0882, "step": 8310 }, { "grad_norm": 0.631130576133728, "learning_rate": 9.969913544869728e-05, "loss": 0.0859, "step": 8320 }, { "grad_norm": 1.2376586198806763, "learning_rate": 9.96973215684834e-05, "loss": 0.0947, "step": 8330 }, { "grad_norm": 0.5785284042358398, "learning_rate": 9.969550225344513e-05, "loss": 0.09, "step": 8340 }, { "grad_norm": 0.7308444380760193, "learning_rate": 9.969367750378147e-05, "loss": 0.1202, "step": 8350 }, { "grad_norm": 0.815016508102417, "learning_rate": 9.969184731969194e-05, "loss": 0.0903, "step": 8360 }, { "grad_norm": 0.6024046540260315, "learning_rate": 9.96900117013767e-05, "loss": 0.104, "step": 8370 }, { "grad_norm": 0.6680266857147217, "learning_rate": 9.96881706490365e-05, "loss": 0.1057, "step": 8380 }, { "grad_norm": 0.620744526386261, "learning_rate": 9.968632416287265e-05, "loss": 0.0889, "step": 8390 }, { "grad_norm": 0.5326526165008545, "learning_rate": 9.96844722430871e-05, "loss": 0.0863, "step": 8400 }, { "grad_norm": 0.7138882875442505, "learning_rate": 9.968261488988235e-05, "loss": 0.1124, "step": 8410 }, { "grad_norm": 0.7610456943511963, "learning_rate": 9.968075210346155e-05, "loss": 0.0955, "step": 8420 }, { "grad_norm": 0.6758374571800232, "learning_rate": 9.967888388402839e-05, "loss": 0.0994, "step": 8430 }, { "grad_norm": 0.7018752694129944, "learning_rate": 9.967701023178717e-05, "loss": 0.0944, "step": 8440 }, { "grad_norm": 0.770243227481842, "learning_rate": 9.967513114694282e-05, "loss": 0.093, "step": 8450 }, { "grad_norm": 0.5307789444923401, "learning_rate": 9.967324662970079e-05, "loss": 0.0914, "step": 8460 }, { "grad_norm": 1.6835143566131592, "learning_rate": 9.96713566802672e-05, "loss": 0.1029, "step": 8470 }, { "grad_norm": 0.610233724117279, "learning_rate": 9.966946129884873e-05, "loss": 0.0993, "step": 8480 }, { "grad_norm": 0.9027944803237915, "learning_rate": 9.966756048565265e-05, "loss": 0.1007, "step": 8490 }, { "grad_norm": 0.7992764115333557, "learning_rate": 9.966565424088681e-05, "loss": 0.1053, "step": 8500 }, { "grad_norm": 0.9420221447944641, "learning_rate": 9.96637425647597e-05, "loss": 0.0947, "step": 8510 }, { "grad_norm": 0.7534971237182617, "learning_rate": 9.966182545748038e-05, "loss": 0.0912, "step": 8520 }, { "grad_norm": 0.9829698801040649, "learning_rate": 9.96599029192585e-05, "loss": 0.1, "step": 8530 }, { "grad_norm": 0.6079791784286499, "learning_rate": 9.965797495030428e-05, "loss": 0.1059, "step": 8540 }, { "grad_norm": 0.6917238235473633, "learning_rate": 9.96560415508286e-05, "loss": 0.0975, "step": 8550 }, { "grad_norm": 0.5558812618255615, "learning_rate": 9.965410272104286e-05, "loss": 0.0992, "step": 8560 }, { "grad_norm": 0.7545018792152405, "learning_rate": 9.96521584611591e-05, "loss": 0.1307, "step": 8570 }, { "grad_norm": 0.6179091334342957, "learning_rate": 9.965020877138994e-05, "loss": 0.1111, "step": 8580 }, { "grad_norm": 0.5655855536460876, "learning_rate": 9.964825365194861e-05, "loss": 0.0977, "step": 8590 }, { "grad_norm": 0.6434353590011597, "learning_rate": 9.96462931030489e-05, "loss": 0.0851, "step": 8600 }, { "grad_norm": 0.7213357090950012, "learning_rate": 9.96443271249052e-05, "loss": 0.0871, "step": 8610 }, { "grad_norm": 0.47484347224235535, "learning_rate": 9.964235571773255e-05, "loss": 0.1122, "step": 8620 }, { "grad_norm": 0.7276164293289185, "learning_rate": 9.96403788817465e-05, "loss": 0.1012, "step": 8630 }, { "grad_norm": 0.5810372233390808, "learning_rate": 9.963839661716325e-05, "loss": 0.1011, "step": 8640 }, { "grad_norm": 0.6699657440185547, "learning_rate": 9.963640892419958e-05, "loss": 0.0996, "step": 8650 }, { "grad_norm": 0.863141655921936, "learning_rate": 9.963441580307286e-05, "loss": 0.0842, "step": 8660 }, { "grad_norm": 0.5700920224189758, "learning_rate": 9.963241725400104e-05, "loss": 0.0999, "step": 8670 }, { "grad_norm": 0.642315149307251, "learning_rate": 9.963041327720271e-05, "loss": 0.0946, "step": 8680 }, { "grad_norm": 0.724109947681427, "learning_rate": 9.962840387289697e-05, "loss": 0.1036, "step": 8690 }, { "grad_norm": 1.12128484249115, "learning_rate": 9.962638904130363e-05, "loss": 0.1017, "step": 8700 }, { "grad_norm": 0.7590301632881165, "learning_rate": 9.962436878264298e-05, "loss": 0.1139, "step": 8710 }, { "grad_norm": 0.6856169700622559, "learning_rate": 9.962234309713598e-05, "loss": 0.115, "step": 8720 }, { "grad_norm": 0.6301792860031128, "learning_rate": 9.962031198500414e-05, "loss": 0.1067, "step": 8730 }, { "grad_norm": 0.5966845750808716, "learning_rate": 9.961827544646958e-05, "loss": 0.0837, "step": 8740 }, { "grad_norm": 0.6393066644668579, "learning_rate": 9.961623348175501e-05, "loss": 0.1044, "step": 8750 }, { "grad_norm": 0.6562419533729553, "learning_rate": 9.961418609108377e-05, "loss": 0.0823, "step": 8760 }, { "grad_norm": 0.6470655798912048, "learning_rate": 9.961213327467971e-05, "loss": 0.1006, "step": 8770 }, { "grad_norm": 0.6223496794700623, "learning_rate": 9.961007503276736e-05, "loss": 0.0888, "step": 8780 }, { "grad_norm": 0.6744070053100586, "learning_rate": 9.960801136557179e-05, "loss": 0.0763, "step": 8790 }, { "grad_norm": 0.5943198204040527, "learning_rate": 9.960594227331866e-05, "loss": 0.0841, "step": 8800 }, { "grad_norm": 0.6733149886131287, "learning_rate": 9.960386775623429e-05, "loss": 0.0762, "step": 8810 }, { "grad_norm": 0.7981215119361877, "learning_rate": 9.96017878145455e-05, "loss": 0.0943, "step": 8820 }, { "grad_norm": 0.5511945486068726, "learning_rate": 9.959970244847977e-05, "loss": 0.0953, "step": 8830 }, { "grad_norm": 0.6404156684875488, "learning_rate": 9.959761165826518e-05, "loss": 0.0964, "step": 8840 }, { "grad_norm": 1.2016335725784302, "learning_rate": 9.959551544413033e-05, "loss": 0.0979, "step": 8850 }, { "grad_norm": 0.6200894117355347, "learning_rate": 9.959341380630448e-05, "loss": 0.0865, "step": 8860 }, { "grad_norm": 0.728049099445343, "learning_rate": 9.959130674501746e-05, "loss": 0.101, "step": 8870 }, { "grad_norm": 0.7066789269447327, "learning_rate": 9.958919426049968e-05, "loss": 0.1122, "step": 8880 }, { "grad_norm": 0.6681591272354126, "learning_rate": 9.958707635298219e-05, "loss": 0.085, "step": 8890 }, { "grad_norm": 0.4522216320037842, "learning_rate": 9.958495302269657e-05, "loss": 0.0892, "step": 8900 }, { "grad_norm": 0.777679443359375, "learning_rate": 9.958282426987503e-05, "loss": 0.0978, "step": 8910 }, { "grad_norm": 0.6849967241287231, "learning_rate": 9.95806900947504e-05, "loss": 0.1041, "step": 8920 }, { "grad_norm": 0.694968044757843, "learning_rate": 9.957855049755604e-05, "loss": 0.0946, "step": 8930 }, { "grad_norm": 0.6121436357498169, "learning_rate": 9.957640547852593e-05, "loss": 0.1075, "step": 8940 }, { "grad_norm": 0.5592698454856873, "learning_rate": 9.957425503789466e-05, "loss": 0.0833, "step": 8950 }, { "grad_norm": 0.7299503087997437, "learning_rate": 9.957209917589738e-05, "loss": 0.0829, "step": 8960 }, { "grad_norm": 0.6931419372558594, "learning_rate": 9.956993789276987e-05, "loss": 0.0856, "step": 8970 }, { "grad_norm": 0.7075669169425964, "learning_rate": 9.956777118874847e-05, "loss": 0.0977, "step": 8980 }, { "grad_norm": 0.9466031789779663, "learning_rate": 9.956559906407016e-05, "loss": 0.1158, "step": 8990 }, { "grad_norm": 0.6903146505355835, "learning_rate": 9.956342151897245e-05, "loss": 0.1015, "step": 9000 }, { "grad_norm": 0.6409417986869812, "learning_rate": 9.956123855369346e-05, "loss": 0.0917, "step": 9010 }, { "grad_norm": 0.591174840927124, "learning_rate": 9.955905016847196e-05, "loss": 0.0817, "step": 9020 }, { "grad_norm": 0.5664951801300049, "learning_rate": 9.955685636354723e-05, "loss": 0.1159, "step": 9030 }, { "grad_norm": 1.3977245092391968, "learning_rate": 9.95546571391592e-05, "loss": 0.1101, "step": 9040 }, { "grad_norm": 0.5829405188560486, "learning_rate": 9.955245249554837e-05, "loss": 0.0885, "step": 9050 }, { "grad_norm": 0.5317845940589905, "learning_rate": 9.955024243295582e-05, "loss": 0.1172, "step": 9060 }, { "grad_norm": 0.8096945881843567, "learning_rate": 9.954802695162328e-05, "loss": 0.0912, "step": 9070 }, { "grad_norm": 1.2249860763549805, "learning_rate": 9.954580605179302e-05, "loss": 0.0973, "step": 9080 }, { "grad_norm": 0.925112783908844, "learning_rate": 9.954357973370788e-05, "loss": 0.1146, "step": 9090 }, { "grad_norm": 0.7666730284690857, "learning_rate": 9.954134799761135e-05, "loss": 0.1305, "step": 9100 }, { "grad_norm": 0.6946660876274109, "learning_rate": 9.953911084374748e-05, "loss": 0.1147, "step": 9110 }, { "grad_norm": 0.7279186844825745, "learning_rate": 9.953686827236093e-05, "loss": 0.1226, "step": 9120 }, { "grad_norm": 0.8083150386810303, "learning_rate": 9.953462028369695e-05, "loss": 0.1274, "step": 9130 }, { "grad_norm": 0.562310516834259, "learning_rate": 9.953236687800136e-05, "loss": 0.1085, "step": 9140 }, { "grad_norm": 0.6887985467910767, "learning_rate": 9.95301080555206e-05, "loss": 0.1149, "step": 9150 }, { "grad_norm": 0.6727374792098999, "learning_rate": 9.952784381650171e-05, "loss": 0.0848, "step": 9160 }, { "grad_norm": 0.7545431852340698, "learning_rate": 9.952557416119226e-05, "loss": 0.1064, "step": 9170 }, { "grad_norm": 0.8033350706100464, "learning_rate": 9.95232990898405e-05, "loss": 0.0955, "step": 9180 }, { "grad_norm": 0.6551624536514282, "learning_rate": 9.95210186026952e-05, "loss": 0.0804, "step": 9190 }, { "grad_norm": 0.5391742587089539, "learning_rate": 9.951873270000576e-05, "loss": 0.0942, "step": 9200 }, { "grad_norm": 0.673103392124176, "learning_rate": 9.951644138202216e-05, "loss": 0.1142, "step": 9210 }, { "grad_norm": 0.8074935078620911, "learning_rate": 9.951414464899498e-05, "loss": 0.0884, "step": 9220 }, { "grad_norm": 0.6120774745941162, "learning_rate": 9.951184250117538e-05, "loss": 0.0908, "step": 9230 }, { "grad_norm": 0.7149393558502197, "learning_rate": 9.950953493881513e-05, "loss": 0.0956, "step": 9240 }, { "grad_norm": 0.65717613697052, "learning_rate": 9.950722196216658e-05, "loss": 0.0828, "step": 9250 }, { "grad_norm": 0.4489409327507019, "learning_rate": 9.950490357148265e-05, "loss": 0.0717, "step": 9260 }, { "grad_norm": 0.5834887623786926, "learning_rate": 9.950257976701692e-05, "loss": 0.1034, "step": 9270 }, { "grad_norm": 0.6864001154899597, "learning_rate": 9.950025054902348e-05, "loss": 0.0935, "step": 9280 }, { "grad_norm": 0.8112936019897461, "learning_rate": 9.949791591775706e-05, "loss": 0.0817, "step": 9290 }, { "grad_norm": 0.5659216642379761, "learning_rate": 9.949557587347298e-05, "loss": 0.1056, "step": 9300 }, { "grad_norm": 0.7121293544769287, "learning_rate": 9.949323041642713e-05, "loss": 0.0917, "step": 9310 }, { "grad_norm": 0.6299290060997009, "learning_rate": 9.949087954687602e-05, "loss": 0.1123, "step": 9320 }, { "grad_norm": 0.7765110731124878, "learning_rate": 9.948852326507672e-05, "loss": 0.0847, "step": 9330 }, { "grad_norm": 0.41125205159187317, "learning_rate": 9.948616157128694e-05, "loss": 0.0856, "step": 9340 }, { "grad_norm": 0.7849204540252686, "learning_rate": 9.948379446576493e-05, "loss": 0.0994, "step": 9350 }, { "grad_norm": 0.6797839999198914, "learning_rate": 9.948142194876952e-05, "loss": 0.0887, "step": 9360 }, { "grad_norm": 0.5288428068161011, "learning_rate": 9.947904402056024e-05, "loss": 0.0941, "step": 9370 }, { "grad_norm": 0.7014575600624084, "learning_rate": 9.947666068139708e-05, "loss": 0.0906, "step": 9380 }, { "grad_norm": 0.6349895596504211, "learning_rate": 9.947427193154071e-05, "loss": 0.0946, "step": 9390 }, { "grad_norm": 0.5305426716804504, "learning_rate": 9.947187777125233e-05, "loss": 0.1194, "step": 9400 }, { "grad_norm": 0.5786422491073608, "learning_rate": 9.946947820079377e-05, "loss": 0.0946, "step": 9410 }, { "grad_norm": 0.6425585746765137, "learning_rate": 9.946707322042747e-05, "loss": 0.0959, "step": 9420 }, { "grad_norm": 0.7194210290908813, "learning_rate": 9.94646628304164e-05, "loss": 0.0979, "step": 9430 }, { "grad_norm": 0.6283953785896301, "learning_rate": 9.946224703102418e-05, "loss": 0.0871, "step": 9440 }, { "grad_norm": 0.6466428637504578, "learning_rate": 9.945982582251498e-05, "loss": 0.086, "step": 9450 }, { "grad_norm": 0.8085620403289795, "learning_rate": 9.94573992051536e-05, "loss": 0.0863, "step": 9460 }, { "grad_norm": 0.3764719069004059, "learning_rate": 9.94549671792054e-05, "loss": 0.1025, "step": 9470 }, { "grad_norm": 0.9364333748817444, "learning_rate": 9.945252974493635e-05, "loss": 0.0937, "step": 9480 }, { "grad_norm": 0.46322792768478394, "learning_rate": 9.9450086902613e-05, "loss": 0.1056, "step": 9490 }, { "grad_norm": 0.5176541209220886, "learning_rate": 9.944763865250248e-05, "loss": 0.0918, "step": 9500 }, { "grad_norm": 0.5653219819068909, "learning_rate": 9.944518499487254e-05, "loss": 0.0771, "step": 9510 }, { "grad_norm": 0.6783185601234436, "learning_rate": 9.944272592999151e-05, "loss": 0.0969, "step": 9520 }, { "grad_norm": 0.4181970953941345, "learning_rate": 9.94402614581283e-05, "loss": 0.0819, "step": 9530 }, { "grad_norm": 0.7092964053153992, "learning_rate": 9.943779157955244e-05, "loss": 0.09, "step": 9540 }, { "grad_norm": 0.5458682179450989, "learning_rate": 9.943531629453403e-05, "loss": 0.0833, "step": 9550 }, { "grad_norm": 0.6088822484016418, "learning_rate": 9.943283560334375e-05, "loss": 0.0984, "step": 9560 }, { "grad_norm": 0.7618441581726074, "learning_rate": 9.943034950625288e-05, "loss": 0.1026, "step": 9570 }, { "grad_norm": 0.5219953060150146, "learning_rate": 9.942785800353332e-05, "loss": 0.0831, "step": 9580 }, { "grad_norm": 0.564534604549408, "learning_rate": 9.942536109545751e-05, "loss": 0.085, "step": 9590 }, { "grad_norm": 0.6728559136390686, "learning_rate": 9.942285878229853e-05, "loss": 0.0895, "step": 9600 }, { "grad_norm": 0.8944361805915833, "learning_rate": 9.942035106433001e-05, "loss": 0.0923, "step": 9610 }, { "grad_norm": 0.9594665169715881, "learning_rate": 9.94178379418262e-05, "loss": 0.1162, "step": 9620 }, { "grad_norm": 0.7198472023010254, "learning_rate": 9.941531941506194e-05, "loss": 0.0868, "step": 9630 }, { "grad_norm": 0.8888493776321411, "learning_rate": 9.941279548431263e-05, "loss": 0.0884, "step": 9640 }, { "grad_norm": 0.5520808100700378, "learning_rate": 9.941026614985431e-05, "loss": 0.1088, "step": 9650 }, { "grad_norm": 0.40390366315841675, "learning_rate": 9.940773141196357e-05, "loss": 0.0805, "step": 9660 }, { "grad_norm": 0.8417918682098389, "learning_rate": 9.94051912709176e-05, "loss": 0.0932, "step": 9670 }, { "grad_norm": 0.6527343988418579, "learning_rate": 9.940264572699421e-05, "loss": 0.0859, "step": 9680 }, { "grad_norm": 0.6594552993774414, "learning_rate": 9.940009478047174e-05, "loss": 0.0911, "step": 9690 }, { "grad_norm": 0.6845465898513794, "learning_rate": 9.939753843162918e-05, "loss": 0.0817, "step": 9700 }, { "grad_norm": 0.6693693399429321, "learning_rate": 9.939497668074609e-05, "loss": 0.0801, "step": 9710 }, { "grad_norm": 0.7172420620918274, "learning_rate": 9.93924095281026e-05, "loss": 0.0893, "step": 9720 }, { "grad_norm": 0.4730941653251648, "learning_rate": 9.938983697397948e-05, "loss": 0.0889, "step": 9730 }, { "grad_norm": 0.6871257424354553, "learning_rate": 9.938725901865805e-05, "loss": 0.0898, "step": 9740 }, { "grad_norm": 0.6079258918762207, "learning_rate": 9.93846756624202e-05, "loss": 0.0713, "step": 9750 }, { "grad_norm": 0.5539264678955078, "learning_rate": 9.938208690554849e-05, "loss": 0.1053, "step": 9760 }, { "grad_norm": 0.6089736223220825, "learning_rate": 9.9379492748326e-05, "loss": 0.1001, "step": 9770 }, { "grad_norm": 0.6496771574020386, "learning_rate": 9.937689319103641e-05, "loss": 0.0997, "step": 9780 }, { "grad_norm": 0.7330746054649353, "learning_rate": 9.937428823396404e-05, "loss": 0.0905, "step": 9790 }, { "grad_norm": 0.7146828770637512, "learning_rate": 9.937167787739372e-05, "loss": 0.0772, "step": 9800 }, { "grad_norm": 0.6447967290878296, "learning_rate": 9.936906212161095e-05, "loss": 0.0902, "step": 9810 }, { "grad_norm": 0.604912281036377, "learning_rate": 9.936644096690176e-05, "loss": 0.1124, "step": 9820 }, { "grad_norm": 0.6339369416236877, "learning_rate": 9.936381441355282e-05, "loss": 0.0885, "step": 9830 }, { "grad_norm": 0.7272171974182129, "learning_rate": 9.936118246185136e-05, "loss": 0.1163, "step": 9840 }, { "grad_norm": 1.0237667560577393, "learning_rate": 9.935854511208518e-05, "loss": 0.1029, "step": 9850 }, { "grad_norm": 0.5988005995750427, "learning_rate": 9.935590236454272e-05, "loss": 0.0925, "step": 9860 }, { "grad_norm": 0.5084041357040405, "learning_rate": 9.935325421951298e-05, "loss": 0.0702, "step": 9870 }, { "grad_norm": 0.6705012321472168, "learning_rate": 9.935060067728557e-05, "loss": 0.0774, "step": 9880 }, { "grad_norm": 0.5576768517494202, "learning_rate": 9.934794173815067e-05, "loss": 0.0719, "step": 9890 }, { "grad_norm": 0.6446492075920105, "learning_rate": 9.934527740239906e-05, "loss": 0.0938, "step": 9900 }, { "grad_norm": 0.5578262805938721, "learning_rate": 9.934260767032209e-05, "loss": 0.1003, "step": 9910 }, { "grad_norm": 0.6541255712509155, "learning_rate": 9.933993254221172e-05, "loss": 0.0907, "step": 9920 }, { "grad_norm": 0.9449304938316345, "learning_rate": 9.933725201836053e-05, "loss": 0.0828, "step": 9930 }, { "grad_norm": 0.8372334241867065, "learning_rate": 9.933456609906162e-05, "loss": 0.0822, "step": 9940 }, { "grad_norm": 0.5283045172691345, "learning_rate": 9.933187478460875e-05, "loss": 0.0852, "step": 9950 }, { "grad_norm": 0.46858125925064087, "learning_rate": 9.93291780752962e-05, "loss": 0.0871, "step": 9960 }, { "grad_norm": 0.5356470346450806, "learning_rate": 9.932647597141893e-05, "loss": 0.0867, "step": 9970 }, { "grad_norm": 0.7328178882598877, "learning_rate": 9.932376847327239e-05, "loss": 0.0957, "step": 9980 }, { "grad_norm": 0.5243496894836426, "learning_rate": 9.932105558115268e-05, "loss": 0.093, "step": 9990 }, { "grad_norm": 0.589771568775177, "learning_rate": 9.931833729535651e-05, "loss": 0.0968, "step": 10000 }, { "grad_norm": 0.9032073020935059, "learning_rate": 9.931561361618111e-05, "loss": 0.1112, "step": 10010 }, { "grad_norm": 0.7196782827377319, "learning_rate": 9.931288454392435e-05, "loss": 0.1078, "step": 10020 }, { "grad_norm": 0.564315140247345, "learning_rate": 9.931015007888467e-05, "loss": 0.0917, "step": 10030 }, { "grad_norm": 0.7003421783447266, "learning_rate": 9.930741022136112e-05, "loss": 0.0985, "step": 10040 }, { "grad_norm": 0.5274190902709961, "learning_rate": 9.930466497165333e-05, "loss": 0.0933, "step": 10050 }, { "grad_norm": 0.5254102945327759, "learning_rate": 9.93019143300615e-05, "loss": 0.085, "step": 10060 }, { "grad_norm": 0.6338934898376465, "learning_rate": 9.929915829688644e-05, "loss": 0.079, "step": 10070 }, { "grad_norm": 0.4571680426597595, "learning_rate": 9.929639687242955e-05, "loss": 0.0796, "step": 10080 }, { "grad_norm": 0.5711126327514648, "learning_rate": 9.929363005699281e-05, "loss": 0.0881, "step": 10090 }, { "grad_norm": 0.46332505345344543, "learning_rate": 9.92908578508788e-05, "loss": 0.0872, "step": 10100 }, { "grad_norm": 0.7888131737709045, "learning_rate": 9.928808025439069e-05, "loss": 0.0944, "step": 10110 }, { "grad_norm": 0.5799462199211121, "learning_rate": 9.928529726783223e-05, "loss": 0.0798, "step": 10120 }, { "grad_norm": 1.2339792251586914, "learning_rate": 9.928250889150774e-05, "loss": 0.0972, "step": 10130 }, { "grad_norm": 0.8268458247184753, "learning_rate": 9.92797151257222e-05, "loss": 0.0947, "step": 10140 }, { "grad_norm": 0.6478838324546814, "learning_rate": 9.927691597078108e-05, "loss": 0.0948, "step": 10150 }, { "grad_norm": 0.9554468989372253, "learning_rate": 9.927411142699053e-05, "loss": 0.0869, "step": 10160 }, { "grad_norm": 0.6501837968826294, "learning_rate": 9.927130149465725e-05, "loss": 0.0955, "step": 10170 }, { "grad_norm": 0.744405210018158, "learning_rate": 9.92684861740885e-05, "loss": 0.1176, "step": 10180 }, { "grad_norm": 0.8967577219009399, "learning_rate": 9.926566546559217e-05, "loss": 0.1131, "step": 10190 }, { "grad_norm": 0.6334688067436218, "learning_rate": 9.926283936947673e-05, "loss": 0.0903, "step": 10200 }, { "grad_norm": 0.5052849054336548, "learning_rate": 9.926000788605126e-05, "loss": 0.0915, "step": 10210 }, { "grad_norm": 0.6340295672416687, "learning_rate": 9.92571710156254e-05, "loss": 0.0864, "step": 10220 }, { "grad_norm": 0.7967654466629028, "learning_rate": 9.925432875850936e-05, "loss": 0.0955, "step": 10230 }, { "grad_norm": 0.5531471371650696, "learning_rate": 9.925148111501396e-05, "loss": 0.0913, "step": 10240 }, { "grad_norm": 0.6474778056144714, "learning_rate": 9.924862808545066e-05, "loss": 0.0937, "step": 10250 }, { "grad_norm": 0.7338302731513977, "learning_rate": 9.924576967013141e-05, "loss": 0.0819, "step": 10260 }, { "grad_norm": 0.7101407051086426, "learning_rate": 9.924290586936887e-05, "loss": 0.0861, "step": 10270 }, { "grad_norm": 0.5965178608894348, "learning_rate": 9.924003668347614e-05, "loss": 0.0978, "step": 10280 }, { "grad_norm": 0.5726165175437927, "learning_rate": 9.923716211276704e-05, "loss": 0.1007, "step": 10290 }, { "grad_norm": 0.5329661965370178, "learning_rate": 9.923428215755594e-05, "loss": 0.0826, "step": 10300 }, { "grad_norm": 0.7797105312347412, "learning_rate": 9.923139681815775e-05, "loss": 0.0862, "step": 10310 }, { "grad_norm": 1.1847047805786133, "learning_rate": 9.922850609488801e-05, "loss": 0.0857, "step": 10320 }, { "grad_norm": 1.457945704460144, "learning_rate": 9.922560998806287e-05, "loss": 0.0883, "step": 10330 }, { "grad_norm": 0.5286403894424438, "learning_rate": 9.922270849799905e-05, "loss": 0.0996, "step": 10340 }, { "grad_norm": 0.6854425072669983, "learning_rate": 9.92198016250138e-05, "loss": 0.0977, "step": 10350 }, { "grad_norm": 0.7909789085388184, "learning_rate": 9.921688936942506e-05, "loss": 0.0884, "step": 10360 }, { "grad_norm": 0.6100190281867981, "learning_rate": 9.921397173155129e-05, "loss": 0.0837, "step": 10370 }, { "grad_norm": 0.9997747540473938, "learning_rate": 9.921104871171157e-05, "loss": 0.0854, "step": 10380 }, { "grad_norm": 0.6728078126907349, "learning_rate": 9.920812031022554e-05, "loss": 0.0917, "step": 10390 }, { "grad_norm": 0.7395618557929993, "learning_rate": 9.920518652741348e-05, "loss": 0.0911, "step": 10400 }, { "grad_norm": 0.742508590221405, "learning_rate": 9.920224736359618e-05, "loss": 0.1064, "step": 10410 }, { "grad_norm": 0.7427561283111572, "learning_rate": 9.91993028190951e-05, "loss": 0.1139, "step": 10420 }, { "grad_norm": 0.5491934418678284, "learning_rate": 9.919635289423222e-05, "loss": 0.1046, "step": 10430 }, { "grad_norm": 0.6035587191581726, "learning_rate": 9.919339758933015e-05, "loss": 0.0874, "step": 10440 }, { "grad_norm": 0.6937395334243774, "learning_rate": 9.919043690471209e-05, "loss": 0.102, "step": 10450 }, { "grad_norm": 0.7348670363426208, "learning_rate": 9.91874708407018e-05, "loss": 0.0954, "step": 10460 }, { "grad_norm": 0.5677726864814758, "learning_rate": 9.918449939762367e-05, "loss": 0.0747, "step": 10470 }, { "grad_norm": 0.41703683137893677, "learning_rate": 9.91815225758026e-05, "loss": 0.0884, "step": 10480 }, { "grad_norm": 0.4205474853515625, "learning_rate": 9.917854037556419e-05, "loss": 0.0777, "step": 10490 }, { "grad_norm": 0.8487289547920227, "learning_rate": 9.917555279723454e-05, "loss": 0.0784, "step": 10500 }, { "grad_norm": 0.672531008720398, "learning_rate": 9.917255984114036e-05, "loss": 0.1006, "step": 10510 }, { "grad_norm": 0.6012411117553711, "learning_rate": 9.916956150760896e-05, "loss": 0.078, "step": 10520 }, { "grad_norm": 0.5511800646781921, "learning_rate": 9.916655779696826e-05, "loss": 0.0911, "step": 10530 }, { "grad_norm": 0.5767893195152283, "learning_rate": 9.916354870954671e-05, "loss": 0.0787, "step": 10540 }, { "grad_norm": 0.6310960054397583, "learning_rate": 9.91605342456734e-05, "loss": 0.092, "step": 10550 }, { "grad_norm": 0.6062616109848022, "learning_rate": 9.915751440567795e-05, "loss": 0.086, "step": 10560 }, { "grad_norm": 0.5932069420814514, "learning_rate": 9.915448918989066e-05, "loss": 0.0903, "step": 10570 }, { "grad_norm": 0.5992730259895325, "learning_rate": 9.915145859864232e-05, "loss": 0.0913, "step": 10580 }, { "grad_norm": 0.9128912091255188, "learning_rate": 9.914842263226437e-05, "loss": 0.1078, "step": 10590 }, { "grad_norm": 0.4662901759147644, "learning_rate": 9.914538129108882e-05, "loss": 0.0986, "step": 10600 }, { "grad_norm": 0.46132755279541016, "learning_rate": 9.914233457544825e-05, "loss": 0.1226, "step": 10610 }, { "grad_norm": 0.6273329854011536, "learning_rate": 9.913928248567586e-05, "loss": 0.1107, "step": 10620 }, { "grad_norm": 0.7369409799575806, "learning_rate": 9.913622502210542e-05, "loss": 0.0746, "step": 10630 }, { "grad_norm": 0.7161555290222168, "learning_rate": 9.913316218507128e-05, "loss": 0.1073, "step": 10640 }, { "grad_norm": 0.5825491547584534, "learning_rate": 9.91300939749084e-05, "loss": 0.0937, "step": 10650 }, { "grad_norm": 0.5724256038665771, "learning_rate": 9.91270203919523e-05, "loss": 0.1097, "step": 10660 }, { "grad_norm": 1.6355373859405518, "learning_rate": 9.912394143653912e-05, "loss": 0.0997, "step": 10670 }, { "grad_norm": 0.6595706343650818, "learning_rate": 9.912085710900555e-05, "loss": 0.0888, "step": 10680 }, { "grad_norm": 0.43008577823638916, "learning_rate": 9.911776740968892e-05, "loss": 0.0962, "step": 10690 }, { "grad_norm": 0.7034228444099426, "learning_rate": 9.911467233892709e-05, "loss": 0.1188, "step": 10700 }, { "grad_norm": 0.6416076421737671, "learning_rate": 9.911157189705853e-05, "loss": 0.11, "step": 10710 }, { "grad_norm": 0.5858029127120972, "learning_rate": 9.910846608442229e-05, "loss": 0.0817, "step": 10720 }, { "grad_norm": 0.5364898443222046, "learning_rate": 9.910535490135805e-05, "loss": 0.082, "step": 10730 }, { "grad_norm": 0.7833541631698608, "learning_rate": 9.910223834820603e-05, "loss": 0.0846, "step": 10740 }, { "grad_norm": 0.6812909245491028, "learning_rate": 9.909911642530703e-05, "loss": 0.1016, "step": 10750 }, { "grad_norm": 0.6339329481124878, "learning_rate": 9.909598913300249e-05, "loss": 0.1133, "step": 10760 }, { "grad_norm": 0.7353823184967041, "learning_rate": 9.909285647163438e-05, "loss": 0.1043, "step": 10770 }, { "grad_norm": 0.8524032235145569, "learning_rate": 9.908971844154531e-05, "loss": 0.1026, "step": 10780 }, { "grad_norm": 0.6327078938484192, "learning_rate": 9.908657504307843e-05, "loss": 0.0867, "step": 10790 }, { "grad_norm": 0.5982388854026794, "learning_rate": 9.908342627657751e-05, "loss": 0.0952, "step": 10800 }, { "grad_norm": 0.4949437379837036, "learning_rate": 9.908027214238689e-05, "loss": 0.0857, "step": 10810 }, { "grad_norm": 0.8168694972991943, "learning_rate": 9.90771126408515e-05, "loss": 0.1005, "step": 10820 }, { "grad_norm": 0.5668489933013916, "learning_rate": 9.907394777231685e-05, "loss": 0.0935, "step": 10830 }, { "grad_norm": 0.5885536670684814, "learning_rate": 9.907077753712905e-05, "loss": 0.1165, "step": 10840 }, { "grad_norm": 0.6564896106719971, "learning_rate": 9.906760193563482e-05, "loss": 0.095, "step": 10850 }, { "grad_norm": 0.44471174478530884, "learning_rate": 9.906442096818139e-05, "loss": 0.0768, "step": 10860 }, { "grad_norm": 0.8035774827003479, "learning_rate": 9.906123463511665e-05, "loss": 0.1083, "step": 10870 }, { "grad_norm": 0.46552789211273193, "learning_rate": 9.905804293678907e-05, "loss": 0.0852, "step": 10880 }, { "grad_norm": 0.7778896689414978, "learning_rate": 9.905484587354766e-05, "loss": 0.1023, "step": 10890 }, { "grad_norm": 0.7081640958786011, "learning_rate": 9.905164344574205e-05, "loss": 0.0904, "step": 10900 }, { "grad_norm": 0.6852129101753235, "learning_rate": 9.904843565372248e-05, "loss": 0.0778, "step": 10910 }, { "grad_norm": 0.5052136182785034, "learning_rate": 9.904522249783972e-05, "loss": 0.0951, "step": 10920 }, { "grad_norm": 0.6952522993087769, "learning_rate": 9.904200397844517e-05, "loss": 0.1164, "step": 10930 }, { "grad_norm": 0.6441379189491272, "learning_rate": 9.903878009589078e-05, "loss": 0.0943, "step": 10940 }, { "grad_norm": 0.7453576922416687, "learning_rate": 9.903555085052915e-05, "loss": 0.0898, "step": 10950 }, { "grad_norm": 0.7150877118110657, "learning_rate": 9.903231624271338e-05, "loss": 0.0902, "step": 10960 }, { "grad_norm": 0.6644208431243896, "learning_rate": 9.902907627279724e-05, "loss": 0.0918, "step": 10970 }, { "grad_norm": 1.0247702598571777, "learning_rate": 9.902583094113504e-05, "loss": 0.0896, "step": 10980 }, { "grad_norm": 0.677595317363739, "learning_rate": 9.902258024808168e-05, "loss": 0.0927, "step": 10990 }, { "grad_norm": 0.49574771523475647, "learning_rate": 9.901932419399264e-05, "loss": 0.0829, "step": 11000 }, { "grad_norm": 0.6988652944564819, "learning_rate": 9.9016062779224e-05, "loss": 0.0985, "step": 11010 }, { "grad_norm": 0.562371015548706, "learning_rate": 9.901279600413242e-05, "loss": 0.0856, "step": 11020 }, { "grad_norm": 0.7550119161605835, "learning_rate": 9.900952386907518e-05, "loss": 0.0891, "step": 11030 }, { "grad_norm": 0.6152151226997375, "learning_rate": 9.90062463744101e-05, "loss": 0.0833, "step": 11040 }, { "grad_norm": 0.832251250743866, "learning_rate": 9.900296352049558e-05, "loss": 0.0737, "step": 11050 }, { "grad_norm": 0.5416218042373657, "learning_rate": 9.899967530769065e-05, "loss": 0.109, "step": 11060 }, { "grad_norm": 0.7143786549568176, "learning_rate": 9.899638173635489e-05, "loss": 0.0808, "step": 11070 }, { "grad_norm": 0.5256372690200806, "learning_rate": 9.899308280684849e-05, "loss": 0.0883, "step": 11080 }, { "grad_norm": 0.4698849320411682, "learning_rate": 9.898977851953222e-05, "loss": 0.0905, "step": 11090 }, { "grad_norm": 0.6024422645568848, "learning_rate": 9.898646887476741e-05, "loss": 0.0971, "step": 11100 }, { "grad_norm": 0.6381545662879944, "learning_rate": 9.898315387291603e-05, "loss": 0.0928, "step": 11110 }, { "grad_norm": 0.4339900016784668, "learning_rate": 9.89798335143406e-05, "loss": 0.082, "step": 11120 }, { "grad_norm": 0.689571738243103, "learning_rate": 9.897650779940419e-05, "loss": 0.0948, "step": 11130 }, { "grad_norm": 0.865577757358551, "learning_rate": 9.897317672847054e-05, "loss": 0.0834, "step": 11140 }, { "grad_norm": 0.6916906237602234, "learning_rate": 9.89698403019039e-05, "loss": 0.0898, "step": 11150 }, { "grad_norm": 0.5971739292144775, "learning_rate": 9.896649852006917e-05, "loss": 0.0841, "step": 11160 }, { "grad_norm": 0.675421953201294, "learning_rate": 9.896315138333177e-05, "loss": 0.0876, "step": 11170 }, { "grad_norm": 0.6561713814735413, "learning_rate": 9.895979889205774e-05, "loss": 0.0771, "step": 11180 }, { "grad_norm": 0.5636487007141113, "learning_rate": 9.895644104661372e-05, "loss": 0.081, "step": 11190 }, { "grad_norm": 0.6385423541069031, "learning_rate": 9.895307784736691e-05, "loss": 0.0945, "step": 11200 }, { "grad_norm": 0.7720340490341187, "learning_rate": 9.894970929468512e-05, "loss": 0.0864, "step": 11210 }, { "grad_norm": 0.8801654577255249, "learning_rate": 9.89463353889367e-05, "loss": 0.1012, "step": 11220 }, { "grad_norm": 0.5283602476119995, "learning_rate": 9.894295613049065e-05, "loss": 0.1097, "step": 11230 }, { "grad_norm": 0.7596815824508667, "learning_rate": 9.893957151971649e-05, "loss": 0.0817, "step": 11240 }, { "grad_norm": 0.6260648965835571, "learning_rate": 9.893618155698436e-05, "loss": 0.0791, "step": 11250 }, { "grad_norm": 0.46614870429039, "learning_rate": 9.8932786242665e-05, "loss": 0.0732, "step": 11260 }, { "grad_norm": 0.4166390001773834, "learning_rate": 9.89293855771297e-05, "loss": 0.0983, "step": 11270 }, { "grad_norm": 0.6330810189247131, "learning_rate": 9.892597956075036e-05, "loss": 0.0904, "step": 11280 }, { "grad_norm": 0.5587881803512573, "learning_rate": 9.892256819389947e-05, "loss": 0.0848, "step": 11290 }, { "grad_norm": 0.5527504086494446, "learning_rate": 9.891915147695006e-05, "loss": 0.0841, "step": 11300 }, { "grad_norm": 0.6631771922111511, "learning_rate": 9.891572941027577e-05, "loss": 0.1005, "step": 11310 }, { "grad_norm": 0.9096765518188477, "learning_rate": 9.89123019942509e-05, "loss": 0.0863, "step": 11320 }, { "grad_norm": 0.5705186128616333, "learning_rate": 9.89088692292502e-05, "loss": 0.0786, "step": 11330 }, { "grad_norm": 0.5528226494789124, "learning_rate": 9.89054311156491e-05, "loss": 0.0711, "step": 11340 }, { "grad_norm": 1.031301498413086, "learning_rate": 9.890198765382357e-05, "loss": 0.113, "step": 11350 }, { "grad_norm": 0.6659577488899231, "learning_rate": 9.889853884415021e-05, "loss": 0.0791, "step": 11360 }, { "grad_norm": 0.8011700510978699, "learning_rate": 9.889508468700614e-05, "loss": 0.072, "step": 11370 }, { "grad_norm": 0.5197322368621826, "learning_rate": 9.889162518276915e-05, "loss": 0.081, "step": 11380 }, { "grad_norm": 0.7026739716529846, "learning_rate": 9.888816033181752e-05, "loss": 0.0839, "step": 11390 }, { "grad_norm": 0.48718753457069397, "learning_rate": 9.888469013453018e-05, "loss": 0.1022, "step": 11400 }, { "grad_norm": 0.9166862964630127, "learning_rate": 9.888121459128663e-05, "loss": 0.0901, "step": 11410 }, { "grad_norm": 0.61580890417099, "learning_rate": 9.887773370246693e-05, "loss": 0.1103, "step": 11420 }, { "grad_norm": 0.7463096976280212, "learning_rate": 9.887424746845177e-05, "loss": 0.1024, "step": 11430 }, { "grad_norm": 1.1077747344970703, "learning_rate": 9.887075588962239e-05, "loss": 0.1013, "step": 11440 }, { "grad_norm": 0.6854669451713562, "learning_rate": 9.88672589663606e-05, "loss": 0.1049, "step": 11450 }, { "grad_norm": 0.5610270500183105, "learning_rate": 9.886375669904886e-05, "loss": 0.0959, "step": 11460 }, { "grad_norm": 0.9059871435165405, "learning_rate": 9.886024908807014e-05, "loss": 0.0894, "step": 11470 }, { "grad_norm": 0.8849126100540161, "learning_rate": 9.885673613380806e-05, "loss": 0.0926, "step": 11480 }, { "grad_norm": 0.639836311340332, "learning_rate": 9.885321783664676e-05, "loss": 0.0848, "step": 11490 }, { "grad_norm": 0.6698444485664368, "learning_rate": 9.884969419697101e-05, "loss": 0.0831, "step": 11500 }, { "grad_norm": 0.5852279663085938, "learning_rate": 9.884616521516614e-05, "loss": 0.0921, "step": 11510 }, { "grad_norm": 0.7710043787956238, "learning_rate": 9.88426308916181e-05, "loss": 0.0902, "step": 11520 }, { "grad_norm": 0.6154401898384094, "learning_rate": 9.883909122671335e-05, "loss": 0.084, "step": 11530 }, { "grad_norm": 0.57790607213974, "learning_rate": 9.883554622083904e-05, "loss": 0.0886, "step": 11540 }, { "grad_norm": 0.7243199944496155, "learning_rate": 9.88319958743828e-05, "loss": 0.1, "step": 11550 }, { "grad_norm": 0.6134152412414551, "learning_rate": 9.882844018773291e-05, "loss": 0.0872, "step": 11560 }, { "grad_norm": 0.5147788524627686, "learning_rate": 9.882487916127823e-05, "loss": 0.104, "step": 11570 }, { "grad_norm": 0.47779351472854614, "learning_rate": 9.882131279540815e-05, "loss": 0.1192, "step": 11580 }, { "grad_norm": 0.6924574971199036, "learning_rate": 9.881774109051271e-05, "loss": 0.1013, "step": 11590 }, { "grad_norm": 0.7237248420715332, "learning_rate": 9.881416404698252e-05, "loss": 0.1007, "step": 11600 }, { "grad_norm": 0.639965295791626, "learning_rate": 9.881058166520873e-05, "loss": 0.0998, "step": 11610 }, { "grad_norm": 0.7262994647026062, "learning_rate": 9.880699394558311e-05, "loss": 0.1134, "step": 11620 }, { "grad_norm": 0.5015500783920288, "learning_rate": 9.880340088849801e-05, "loss": 0.0985, "step": 11630 }, { "grad_norm": 0.9133360981941223, "learning_rate": 9.879980249434637e-05, "loss": 0.0817, "step": 11640 }, { "grad_norm": 0.6461959481239319, "learning_rate": 9.879619876352168e-05, "loss": 0.0868, "step": 11650 }, { "grad_norm": 0.558456301689148, "learning_rate": 9.879258969641809e-05, "loss": 0.0908, "step": 11660 }, { "grad_norm": 0.5077624917030334, "learning_rate": 9.878897529343023e-05, "loss": 0.0838, "step": 11670 }, { "grad_norm": 0.7831261157989502, "learning_rate": 9.878535555495338e-05, "loss": 0.09, "step": 11680 }, { "grad_norm": 0.7270698547363281, "learning_rate": 9.87817304813834e-05, "loss": 0.1017, "step": 11690 }, { "grad_norm": 0.7171099781990051, "learning_rate": 9.877810007311671e-05, "loss": 0.0905, "step": 11700 }, { "grad_norm": 0.5814310908317566, "learning_rate": 9.877446433055035e-05, "loss": 0.0929, "step": 11710 }, { "grad_norm": 0.7103462815284729, "learning_rate": 9.877082325408191e-05, "loss": 0.0963, "step": 11720 }, { "grad_norm": 0.6575279831886292, "learning_rate": 9.876717684410954e-05, "loss": 0.0798, "step": 11730 }, { "grad_norm": 0.6807323694229126, "learning_rate": 9.876352510103204e-05, "loss": 0.0895, "step": 11740 }, { "grad_norm": 0.6407753825187683, "learning_rate": 9.875986802524875e-05, "loss": 0.0758, "step": 11750 }, { "grad_norm": 0.47827789187431335, "learning_rate": 9.87562056171596e-05, "loss": 0.0764, "step": 11760 }, { "grad_norm": 0.508963406085968, "learning_rate": 9.875253787716511e-05, "loss": 0.1079, "step": 11770 }, { "grad_norm": 0.6678486466407776, "learning_rate": 9.874886480566637e-05, "loss": 0.0937, "step": 11780 }, { "grad_norm": 0.5714801549911499, "learning_rate": 9.874518640306507e-05, "loss": 0.0825, "step": 11790 }, { "grad_norm": 1.3847829103469849, "learning_rate": 9.874150266976347e-05, "loss": 0.1109, "step": 11800 }, { "grad_norm": 0.6041643619537354, "learning_rate": 9.873781360616443e-05, "loss": 0.1102, "step": 11810 }, { "grad_norm": 1.371284008026123, "learning_rate": 9.873411921267137e-05, "loss": 0.0913, "step": 11820 }, { "grad_norm": 0.6781241297721863, "learning_rate": 9.873041948968829e-05, "loss": 0.0994, "step": 11830 }, { "grad_norm": 0.601996898651123, "learning_rate": 9.872671443761981e-05, "loss": 0.0903, "step": 11840 }, { "grad_norm": 0.6447106003761292, "learning_rate": 9.872300405687109e-05, "loss": 0.0747, "step": 11850 }, { "grad_norm": 0.5442571640014648, "learning_rate": 9.871928834784792e-05, "loss": 0.0863, "step": 11860 }, { "grad_norm": 0.5410860180854797, "learning_rate": 9.871556731095661e-05, "loss": 0.0968, "step": 11870 }, { "grad_norm": 0.8378733992576599, "learning_rate": 9.871184094660411e-05, "loss": 0.1013, "step": 11880 }, { "grad_norm": 0.5923352837562561, "learning_rate": 9.870810925519791e-05, "loss": 0.0946, "step": 11890 }, { "grad_norm": 0.5305511951446533, "learning_rate": 9.870437223714612e-05, "loss": 0.0886, "step": 11900 }, { "grad_norm": 0.6191926598548889, "learning_rate": 9.87006298928574e-05, "loss": 0.0841, "step": 11910 }, { "grad_norm": 0.4471346139907837, "learning_rate": 9.869688222274103e-05, "loss": 0.0806, "step": 11920 }, { "grad_norm": 0.6359806656837463, "learning_rate": 9.869312922720681e-05, "loss": 0.091, "step": 11930 }, { "grad_norm": 0.4082539677619934, "learning_rate": 9.868937090666521e-05, "loss": 0.0846, "step": 11940 }, { "grad_norm": 0.4674593210220337, "learning_rate": 9.86856072615272e-05, "loss": 0.0743, "step": 11950 }, { "grad_norm": 0.7569797039031982, "learning_rate": 9.868183829220438e-05, "loss": 0.0796, "step": 11960 }, { "grad_norm": 0.8129593133926392, "learning_rate": 9.867806399910893e-05, "loss": 0.0823, "step": 11970 }, { "grad_norm": 0.7206203937530518, "learning_rate": 9.867428438265356e-05, "loss": 0.0878, "step": 11980 }, { "grad_norm": 0.5477662682533264, "learning_rate": 9.867049944325165e-05, "loss": 0.1017, "step": 11990 }, { "grad_norm": 0.5399656891822815, "learning_rate": 9.86667091813171e-05, "loss": 0.1065, "step": 12000 }, { "grad_norm": 0.9958348274230957, "learning_rate": 9.866291359726438e-05, "loss": 0.1059, "step": 12010 }, { "grad_norm": 0.6398419737815857, "learning_rate": 9.865911269150861e-05, "loss": 0.0946, "step": 12020 }, { "grad_norm": 0.7046775817871094, "learning_rate": 9.865530646446544e-05, "loss": 0.0774, "step": 12030 }, { "grad_norm": 0.6426643133163452, "learning_rate": 9.86514949165511e-05, "loss": 0.0825, "step": 12040 }, { "grad_norm": 0.5303687453269958, "learning_rate": 9.864767804818243e-05, "loss": 0.0938, "step": 12050 }, { "grad_norm": 0.5873979330062866, "learning_rate": 9.86438558597768e-05, "loss": 0.1002, "step": 12060 }, { "grad_norm": 0.47462475299835205, "learning_rate": 9.864002835175225e-05, "loss": 0.0989, "step": 12070 }, { "grad_norm": 0.528041660785675, "learning_rate": 9.863619552452734e-05, "loss": 0.075, "step": 12080 }, { "grad_norm": 0.7164462804794312, "learning_rate": 9.863235737852119e-05, "loss": 0.0631, "step": 12090 }, { "grad_norm": 0.8704277276992798, "learning_rate": 9.862851391415356e-05, "loss": 0.0909, "step": 12100 }, { "grad_norm": 0.6154713034629822, "learning_rate": 9.862466513184477e-05, "loss": 0.0867, "step": 12110 }, { "grad_norm": 0.6298216581344604, "learning_rate": 9.86208110320157e-05, "loss": 0.0774, "step": 12120 }, { "grad_norm": 0.687930166721344, "learning_rate": 9.861695161508784e-05, "loss": 0.0729, "step": 12130 }, { "grad_norm": 0.6237350702285767, "learning_rate": 9.861308688148324e-05, "loss": 0.0808, "step": 12140 }, { "grad_norm": 0.6533467769622803, "learning_rate": 9.860921683162455e-05, "loss": 0.082, "step": 12150 }, { "grad_norm": 0.7509920597076416, "learning_rate": 9.860534146593499e-05, "loss": 0.0914, "step": 12160 }, { "grad_norm": 0.6291418075561523, "learning_rate": 9.860146078483836e-05, "loss": 0.0892, "step": 12170 }, { "grad_norm": 0.7818060517311096, "learning_rate": 9.859757478875905e-05, "loss": 0.0852, "step": 12180 }, { "grad_norm": 0.6196910738945007, "learning_rate": 9.859368347812204e-05, "loss": 0.0804, "step": 12190 }, { "grad_norm": 0.7738250494003296, "learning_rate": 9.858978685335285e-05, "loss": 0.0761, "step": 12200 }, { "grad_norm": 0.6418511867523193, "learning_rate": 9.858588491487763e-05, "loss": 0.0911, "step": 12210 }, { "grad_norm": 0.7004272937774658, "learning_rate": 9.858197766312308e-05, "loss": 0.0792, "step": 12220 }, { "grad_norm": 0.6220501661300659, "learning_rate": 9.857806509851649e-05, "loss": 0.0913, "step": 12230 }, { "grad_norm": 0.46282294392585754, "learning_rate": 9.857414722148574e-05, "loss": 0.0889, "step": 12240 }, { "grad_norm": 0.6458562612533569, "learning_rate": 9.857022403245928e-05, "loss": 0.101, "step": 12250 }, { "grad_norm": 0.7262240052223206, "learning_rate": 9.856629553186615e-05, "loss": 0.1051, "step": 12260 }, { "grad_norm": 0.7930904626846313, "learning_rate": 9.856236172013595e-05, "loss": 0.0852, "step": 12270 }, { "grad_norm": 0.8438039422035217, "learning_rate": 9.85584225976989e-05, "loss": 0.0913, "step": 12280 }, { "grad_norm": 1.9157934188842773, "learning_rate": 9.855447816498575e-05, "loss": 0.11, "step": 12290 }, { "grad_norm": 0.6903286576271057, "learning_rate": 9.855052842242787e-05, "loss": 0.0833, "step": 12300 }, { "grad_norm": 1.1261025667190552, "learning_rate": 9.85465733704572e-05, "loss": 0.0947, "step": 12310 }, { "grad_norm": 0.7845144271850586, "learning_rate": 9.854261300950624e-05, "loss": 0.107, "step": 12320 }, { "grad_norm": 0.6049818396568298, "learning_rate": 9.853864734000813e-05, "loss": 0.0871, "step": 12330 }, { "grad_norm": 0.6876636743545532, "learning_rate": 9.85346763623965e-05, "loss": 0.1024, "step": 12340 }, { "grad_norm": 0.8187164068222046, "learning_rate": 9.853070007710564e-05, "loss": 0.0791, "step": 12350 }, { "grad_norm": 0.6657841205596924, "learning_rate": 9.85267184845704e-05, "loss": 0.0754, "step": 12360 }, { "grad_norm": 0.8798332810401917, "learning_rate": 9.852273158522616e-05, "loss": 0.0943, "step": 12370 }, { "grad_norm": 0.714044451713562, "learning_rate": 9.851873937950896e-05, "loss": 0.0789, "step": 12380 }, { "grad_norm": 0.7471494674682617, "learning_rate": 9.851474186785537e-05, "loss": 0.0934, "step": 12390 }, { "grad_norm": 0.8479434847831726, "learning_rate": 9.851073905070254e-05, "loss": 0.1011, "step": 12400 }, { "grad_norm": 0.9192229509353638, "learning_rate": 9.850673092848824e-05, "loss": 0.1195, "step": 12410 }, { "grad_norm": 0.48660174012184143, "learning_rate": 9.850271750165077e-05, "loss": 0.0714, "step": 12420 }, { "grad_norm": 0.7612777948379517, "learning_rate": 9.849869877062902e-05, "loss": 0.0938, "step": 12430 }, { "grad_norm": 0.6323807239532471, "learning_rate": 9.849467473586252e-05, "loss": 0.0951, "step": 12440 }, { "grad_norm": 0.7707821130752563, "learning_rate": 9.849064539779127e-05, "loss": 0.0863, "step": 12450 }, { "grad_norm": 0.6459088921546936, "learning_rate": 9.848661075685594e-05, "loss": 0.0874, "step": 12460 }, { "grad_norm": 0.5579776167869568, "learning_rate": 9.848257081349778e-05, "loss": 0.1006, "step": 12470 }, { "grad_norm": 0.6172691583633423, "learning_rate": 9.847852556815856e-05, "loss": 0.0785, "step": 12480 }, { "grad_norm": 0.5087842345237732, "learning_rate": 9.847447502128067e-05, "loss": 0.0931, "step": 12490 }, { "grad_norm": 0.8653514981269836, "learning_rate": 9.847041917330708e-05, "loss": 0.0829, "step": 12500 }, { "grad_norm": 0.5954230427742004, "learning_rate": 9.846635802468132e-05, "loss": 0.095, "step": 12510 }, { "grad_norm": 0.6622465252876282, "learning_rate": 9.84622915758475e-05, "loss": 0.0837, "step": 12520 }, { "grad_norm": 0.5057116746902466, "learning_rate": 9.845821982725034e-05, "loss": 0.0765, "step": 12530 }, { "grad_norm": 0.6372841000556946, "learning_rate": 9.845414277933514e-05, "loss": 0.0854, "step": 12540 }, { "grad_norm": 0.5693028569221497, "learning_rate": 9.845006043254771e-05, "loss": 0.0966, "step": 12550 }, { "grad_norm": 0.6676989197731018, "learning_rate": 9.844597278733451e-05, "loss": 0.0905, "step": 12560 }, { "grad_norm": 0.5364214181900024, "learning_rate": 9.844187984414259e-05, "loss": 0.0842, "step": 12570 }, { "grad_norm": 0.767753005027771, "learning_rate": 9.84377816034195e-05, "loss": 0.0864, "step": 12580 }, { "grad_norm": 0.6840788125991821, "learning_rate": 9.843367806561345e-05, "loss": 0.0857, "step": 12590 }, { "grad_norm": 0.7047858834266663, "learning_rate": 9.842956923117317e-05, "loss": 0.0774, "step": 12600 }, { "grad_norm": 0.5570169687271118, "learning_rate": 9.842545510054802e-05, "loss": 0.0693, "step": 12610 }, { "grad_norm": 0.6688730120658875, "learning_rate": 9.842133567418792e-05, "loss": 0.1002, "step": 12620 }, { "grad_norm": 0.5159227848052979, "learning_rate": 9.841721095254333e-05, "loss": 0.0752, "step": 12630 }, { "grad_norm": 0.6659371256828308, "learning_rate": 9.841308093606537e-05, "loss": 0.0816, "step": 12640 }, { "grad_norm": 0.5862628817558289, "learning_rate": 9.840894562520565e-05, "loss": 0.0842, "step": 12650 }, { "grad_norm": 0.7087007761001587, "learning_rate": 9.840480502041642e-05, "loss": 0.0999, "step": 12660 }, { "grad_norm": 0.6793761849403381, "learning_rate": 9.840065912215049e-05, "loss": 0.0855, "step": 12670 }, { "grad_norm": 0.5588470101356506, "learning_rate": 9.839650793086124e-05, "loss": 0.1101, "step": 12680 }, { "grad_norm": 0.4648589491844177, "learning_rate": 9.839235144700265e-05, "loss": 0.0772, "step": 12690 }, { "grad_norm": 0.5373817682266235, "learning_rate": 9.838818967102926e-05, "loss": 0.0775, "step": 12700 }, { "grad_norm": 0.5558391809463501, "learning_rate": 9.83840226033962e-05, "loss": 0.0822, "step": 12710 }, { "grad_norm": 0.7367396354675293, "learning_rate": 9.837985024455918e-05, "loss": 0.0769, "step": 12720 }, { "grad_norm": 0.5164972543716431, "learning_rate": 9.837567259497447e-05, "loss": 0.0764, "step": 12730 }, { "grad_norm": 0.5480327606201172, "learning_rate": 9.837148965509894e-05, "loss": 0.0789, "step": 12740 }, { "grad_norm": 0.8281675577163696, "learning_rate": 9.836730142539001e-05, "loss": 0.0742, "step": 12750 }, { "grad_norm": 0.6801111698150635, "learning_rate": 9.836310790630574e-05, "loss": 0.1039, "step": 12760 }, { "grad_norm": 0.48807647824287415, "learning_rate": 9.83589090983047e-05, "loss": 0.0926, "step": 12770 }, { "grad_norm": 0.7344358563423157, "learning_rate": 9.835470500184605e-05, "loss": 0.0878, "step": 12780 }, { "grad_norm": 0.6067625880241394, "learning_rate": 9.835049561738957e-05, "loss": 0.0875, "step": 12790 }, { "grad_norm": 0.8148003816604614, "learning_rate": 9.834628094539558e-05, "loss": 0.0768, "step": 12800 }, { "grad_norm": 0.7411630749702454, "learning_rate": 9.834206098632499e-05, "loss": 0.0935, "step": 12810 }, { "grad_norm": 0.6135327816009521, "learning_rate": 9.833783574063931e-05, "loss": 0.1011, "step": 12820 }, { "grad_norm": 0.7145399451255798, "learning_rate": 9.833360520880058e-05, "loss": 0.096, "step": 12830 }, { "grad_norm": 0.48376473784446716, "learning_rate": 9.832936939127144e-05, "loss": 0.0848, "step": 12840 }, { "grad_norm": 0.7872751951217651, "learning_rate": 9.832512828851515e-05, "loss": 0.1023, "step": 12850 }, { "grad_norm": 0.5693261027336121, "learning_rate": 9.832088190099546e-05, "loss": 0.0849, "step": 12860 }, { "grad_norm": 0.7178019285202026, "learning_rate": 9.831663022917679e-05, "loss": 0.0911, "step": 12870 }, { "grad_norm": 0.7939070463180542, "learning_rate": 9.831237327352407e-05, "loss": 0.0872, "step": 12880 }, { "grad_norm": 0.5806344151496887, "learning_rate": 9.830811103450286e-05, "loss": 0.0819, "step": 12890 }, { "grad_norm": 0.7448586821556091, "learning_rate": 9.830384351257924e-05, "loss": 0.087, "step": 12900 }, { "grad_norm": 0.8233267068862915, "learning_rate": 9.829957070821993e-05, "loss": 0.0908, "step": 12910 }, { "grad_norm": 0.7933453917503357, "learning_rate": 9.829529262189218e-05, "loss": 0.0745, "step": 12920 }, { "grad_norm": 0.6219724416732788, "learning_rate": 9.829100925406385e-05, "loss": 0.0867, "step": 12930 }, { "grad_norm": 0.8425785303115845, "learning_rate": 9.828672060520333e-05, "loss": 0.091, "step": 12940 }, { "grad_norm": 0.6705944538116455, "learning_rate": 9.828242667577966e-05, "loss": 0.0713, "step": 12950 }, { "grad_norm": 0.6028018593788147, "learning_rate": 9.82781274662624e-05, "loss": 0.0801, "step": 12960 }, { "grad_norm": 0.6921923160552979, "learning_rate": 9.82738229771217e-05, "loss": 0.0854, "step": 12970 }, { "grad_norm": 0.578669548034668, "learning_rate": 9.826951320882829e-05, "loss": 0.0778, "step": 12980 }, { "grad_norm": 0.8973339796066284, "learning_rate": 9.826519816185351e-05, "loss": 0.0816, "step": 12990 }, { "grad_norm": 0.684145987033844, "learning_rate": 9.826087783666921e-05, "loss": 0.0898, "step": 13000 }, { "grad_norm": 0.5883568525314331, "learning_rate": 9.825655223374787e-05, "loss": 0.1015, "step": 13010 }, { "grad_norm": 0.7912386059761047, "learning_rate": 9.825222135356253e-05, "loss": 0.1059, "step": 13020 }, { "grad_norm": 0.5652675628662109, "learning_rate": 9.82478851965868e-05, "loss": 0.1042, "step": 13030 }, { "grad_norm": 0.5420261025428772, "learning_rate": 9.82435437632949e-05, "loss": 0.0733, "step": 13040 }, { "grad_norm": 0.8878403306007385, "learning_rate": 9.823919705416158e-05, "loss": 0.0813, "step": 13050 }, { "grad_norm": 1.2653801441192627, "learning_rate": 9.82348450696622e-05, "loss": 0.0945, "step": 13060 }, { "grad_norm": 0.7577270269393921, "learning_rate": 9.823048781027268e-05, "loss": 0.1111, "step": 13070 }, { "grad_norm": 0.5977872610092163, "learning_rate": 9.822612527646953e-05, "loss": 0.0926, "step": 13080 }, { "grad_norm": 0.7647903561592102, "learning_rate": 9.822175746872984e-05, "loss": 0.1022, "step": 13090 }, { "grad_norm": 0.7010514140129089, "learning_rate": 9.821738438753123e-05, "loss": 0.0954, "step": 13100 }, { "grad_norm": 0.6210164427757263, "learning_rate": 9.821300603335196e-05, "loss": 0.0918, "step": 13110 }, { "grad_norm": 0.6894271969795227, "learning_rate": 9.820862240667085e-05, "loss": 0.1009, "step": 13120 }, { "grad_norm": 0.5698705315589905, "learning_rate": 9.820423350796726e-05, "loss": 0.0836, "step": 13130 }, { "grad_norm": 0.9347293972969055, "learning_rate": 9.819983933772118e-05, "loss": 0.095, "step": 13140 }, { "grad_norm": 0.6870325803756714, "learning_rate": 9.819543989641314e-05, "loss": 0.0872, "step": 13150 }, { "grad_norm": 0.622481107711792, "learning_rate": 9.819103518452423e-05, "loss": 0.0765, "step": 13160 }, { "grad_norm": 0.6743023991584778, "learning_rate": 9.818662520253618e-05, "loss": 0.085, "step": 13170 }, { "grad_norm": 0.47075915336608887, "learning_rate": 9.818220995093126e-05, "loss": 0.0744, "step": 13180 }, { "grad_norm": 1.420383334159851, "learning_rate": 9.817778943019228e-05, "loss": 0.084, "step": 13190 }, { "grad_norm": 0.6788074374198914, "learning_rate": 9.81733636408027e-05, "loss": 0.087, "step": 13200 }, { "grad_norm": 0.690026044845581, "learning_rate": 9.816893258324649e-05, "loss": 0.0746, "step": 13210 }, { "grad_norm": 0.664071798324585, "learning_rate": 9.816449625800823e-05, "loss": 0.0711, "step": 13220 }, { "grad_norm": 0.5430083870887756, "learning_rate": 9.816005466557308e-05, "loss": 0.0834, "step": 13230 }, { "grad_norm": 0.8451903462409973, "learning_rate": 9.815560780642674e-05, "loss": 0.1046, "step": 13240 }, { "grad_norm": 0.5001671314239502, "learning_rate": 9.815115568105555e-05, "loss": 0.075, "step": 13250 }, { "grad_norm": 0.6210177540779114, "learning_rate": 9.814669828994638e-05, "loss": 0.0948, "step": 13260 }, { "grad_norm": 0.5229415893554688, "learning_rate": 9.814223563358665e-05, "loss": 0.0789, "step": 13270 }, { "grad_norm": 0.6623302102088928, "learning_rate": 9.813776771246443e-05, "loss": 0.0735, "step": 13280 }, { "grad_norm": 0.49298763275146484, "learning_rate": 9.813329452706829e-05, "loss": 0.08, "step": 13290 }, { "grad_norm": 0.5429501533508301, "learning_rate": 9.812881607788744e-05, "loss": 0.0821, "step": 13300 }, { "grad_norm": 0.5490180850028992, "learning_rate": 9.812433236541163e-05, "loss": 0.0919, "step": 13310 }, { "grad_norm": 0.7244142889976501, "learning_rate": 9.811984339013116e-05, "loss": 0.103, "step": 13320 }, { "grad_norm": 0.5283586382865906, "learning_rate": 9.811534915253698e-05, "loss": 0.0805, "step": 13330 }, { "grad_norm": 0.8400384783744812, "learning_rate": 9.811084965312056e-05, "loss": 0.0898, "step": 13340 }, { "grad_norm": 0.7503861784934998, "learning_rate": 9.810634489237396e-05, "loss": 0.0767, "step": 13350 }, { "grad_norm": 1.068619728088379, "learning_rate": 9.81018348707898e-05, "loss": 0.1157, "step": 13360 }, { "grad_norm": 0.8376982808113098, "learning_rate": 9.809731958886131e-05, "loss": 0.0882, "step": 13370 }, { "grad_norm": 0.8785576224327087, "learning_rate": 9.809279904708224e-05, "loss": 0.1045, "step": 13380 }, { "grad_norm": 0.5690907835960388, "learning_rate": 9.808827324594699e-05, "loss": 0.095, "step": 13390 }, { "grad_norm": 0.5493364930152893, "learning_rate": 9.808374218595046e-05, "loss": 0.1023, "step": 13400 }, { "grad_norm": 0.6902622580528259, "learning_rate": 9.80792058675882e-05, "loss": 0.0787, "step": 13410 }, { "grad_norm": 0.5730109214782715, "learning_rate": 9.807466429135627e-05, "loss": 0.0881, "step": 13420 }, { "grad_norm": 0.5097213983535767, "learning_rate": 9.807011745775132e-05, "loss": 0.0868, "step": 13430 }, { "grad_norm": 0.47436195611953735, "learning_rate": 9.806556536727061e-05, "loss": 0.0767, "step": 13440 }, { "grad_norm": 0.5391947627067566, "learning_rate": 9.806100802041193e-05, "loss": 0.1025, "step": 13450 }, { "grad_norm": 0.48519548773765564, "learning_rate": 9.805644541767368e-05, "loss": 0.0916, "step": 13460 }, { "grad_norm": 0.5522710680961609, "learning_rate": 9.805187755955478e-05, "loss": 0.0969, "step": 13470 }, { "grad_norm": 0.36373400688171387, "learning_rate": 9.804730444655483e-05, "loss": 0.0811, "step": 13480 }, { "grad_norm": 0.6765844225883484, "learning_rate": 9.804272607917388e-05, "loss": 0.0912, "step": 13490 }, { "grad_norm": 0.45654118061065674, "learning_rate": 9.803814245791265e-05, "loss": 0.0762, "step": 13500 }, { "grad_norm": 0.7552376389503479, "learning_rate": 9.803355358327239e-05, "loss": 0.094, "step": 13510 }, { "grad_norm": 0.6177218556404114, "learning_rate": 9.802895945575492e-05, "loss": 0.0869, "step": 13520 }, { "grad_norm": 0.6022366285324097, "learning_rate": 9.802436007586266e-05, "loss": 0.1033, "step": 13530 }, { "grad_norm": 0.812938928604126, "learning_rate": 9.801975544409858e-05, "loss": 0.0762, "step": 13540 }, { "grad_norm": 0.9656152725219727, "learning_rate": 9.801514556096625e-05, "loss": 0.0856, "step": 13550 }, { "grad_norm": 0.5348899960517883, "learning_rate": 9.801053042696977e-05, "loss": 0.0925, "step": 13560 }, { "grad_norm": 0.583312451839447, "learning_rate": 9.800591004261388e-05, "loss": 0.0799, "step": 13570 }, { "grad_norm": 0.9333008527755737, "learning_rate": 9.800128440840385e-05, "loss": 0.098, "step": 13580 }, { "grad_norm": 0.8183982372283936, "learning_rate": 9.799665352484552e-05, "loss": 0.1141, "step": 13590 }, { "grad_norm": 0.8303657174110413, "learning_rate": 9.799201739244532e-05, "loss": 0.118, "step": 13600 }, { "grad_norm": 0.5109256505966187, "learning_rate": 9.798737601171025e-05, "loss": 0.0833, "step": 13610 }, { "grad_norm": 0.7961120009422302, "learning_rate": 9.79827293831479e-05, "loss": 0.1053, "step": 13620 }, { "grad_norm": 0.6309503316879272, "learning_rate": 9.797807750726638e-05, "loss": 0.0753, "step": 13630 }, { "grad_norm": 0.5027253031730652, "learning_rate": 9.797342038457446e-05, "loss": 0.0848, "step": 13640 }, { "grad_norm": 0.6558240652084351, "learning_rate": 9.796875801558141e-05, "loss": 0.0947, "step": 13650 }, { "grad_norm": 0.5056107640266418, "learning_rate": 9.79640904007971e-05, "loss": 0.0859, "step": 13660 }, { "grad_norm": 0.5818803906440735, "learning_rate": 9.795941754073199e-05, "loss": 0.076, "step": 13670 }, { "grad_norm": 0.5293084979057312, "learning_rate": 9.795473943589705e-05, "loss": 0.071, "step": 13680 }, { "grad_norm": 0.5760930180549622, "learning_rate": 9.795005608680394e-05, "loss": 0.081, "step": 13690 }, { "grad_norm": 0.645124077796936, "learning_rate": 9.794536749396477e-05, "loss": 0.1028, "step": 13700 }, { "grad_norm": 0.6086113452911377, "learning_rate": 9.79406736578923e-05, "loss": 0.0876, "step": 13710 }, { "grad_norm": 0.6724046468734741, "learning_rate": 9.793597457909984e-05, "loss": 0.0898, "step": 13720 }, { "grad_norm": 0.48771706223487854, "learning_rate": 9.793127025810127e-05, "loss": 0.0798, "step": 13730 }, { "grad_norm": 0.44543346762657166, "learning_rate": 9.792656069541104e-05, "loss": 0.0739, "step": 13740 }, { "grad_norm": 0.5584957599639893, "learning_rate": 9.79218458915442e-05, "loss": 0.0813, "step": 13750 }, { "grad_norm": 0.4843728840351105, "learning_rate": 9.791712584701634e-05, "loss": 0.0915, "step": 13760 }, { "grad_norm": 0.39872196316719055, "learning_rate": 9.791240056234364e-05, "loss": 0.0813, "step": 13770 }, { "grad_norm": 0.4397197365760803, "learning_rate": 9.790767003804283e-05, "loss": 0.078, "step": 13780 }, { "grad_norm": 0.6173840761184692, "learning_rate": 9.790293427463126e-05, "loss": 0.0694, "step": 13790 }, { "grad_norm": 0.6799091696739197, "learning_rate": 9.789819327262684e-05, "loss": 0.0719, "step": 13800 }, { "grad_norm": 0.6879981160163879, "learning_rate": 9.7893447032548e-05, "loss": 0.0848, "step": 13810 }, { "grad_norm": 0.7187581062316895, "learning_rate": 9.78886955549138e-05, "loss": 0.0873, "step": 13820 }, { "grad_norm": 0.4548802375793457, "learning_rate": 9.788393884024387e-05, "loss": 0.0864, "step": 13830 }, { "grad_norm": 0.5232139825820923, "learning_rate": 9.787917688905836e-05, "loss": 0.0921, "step": 13840 }, { "grad_norm": 0.6425222754478455, "learning_rate": 9.787440970187807e-05, "loss": 0.0731, "step": 13850 }, { "grad_norm": 0.7438345551490784, "learning_rate": 9.786963727922429e-05, "loss": 0.0834, "step": 13860 }, { "grad_norm": 2.322280168533325, "learning_rate": 9.786485962161897e-05, "loss": 0.0797, "step": 13870 }, { "grad_norm": 0.6560822129249573, "learning_rate": 9.786007672958455e-05, "loss": 0.0841, "step": 13880 }, { "grad_norm": 0.6065496802330017, "learning_rate": 9.78552886036441e-05, "loss": 0.0855, "step": 13890 }, { "grad_norm": 0.5556075572967529, "learning_rate": 9.785049524432124e-05, "loss": 0.0723, "step": 13900 }, { "grad_norm": 0.6091442108154297, "learning_rate": 9.784569665214016e-05, "loss": 0.0931, "step": 13910 }, { "grad_norm": 0.9051782488822937, "learning_rate": 9.784089282762563e-05, "loss": 0.0864, "step": 13920 }, { "grad_norm": 0.7290116548538208, "learning_rate": 9.7836083771303e-05, "loss": 0.0936, "step": 13930 }, { "grad_norm": 0.7290383577346802, "learning_rate": 9.783126948369817e-05, "loss": 0.093, "step": 13940 }, { "grad_norm": 0.5443789958953857, "learning_rate": 9.78264499653376e-05, "loss": 0.0856, "step": 13950 }, { "grad_norm": 0.7003943920135498, "learning_rate": 9.782162521674838e-05, "loss": 0.0821, "step": 13960 }, { "grad_norm": 0.5798090696334839, "learning_rate": 9.781679523845812e-05, "loss": 0.1005, "step": 13970 }, { "grad_norm": 0.8239958882331848, "learning_rate": 9.781196003099502e-05, "loss": 0.0978, "step": 13980 }, { "grad_norm": 0.47152402997016907, "learning_rate": 9.780711959488786e-05, "loss": 0.0689, "step": 13990 }, { "grad_norm": 0.39769887924194336, "learning_rate": 9.780227393066599e-05, "loss": 0.0735, "step": 14000 }, { "grad_norm": 0.561138391494751, "learning_rate": 9.77974230388593e-05, "loss": 0.092, "step": 14010 }, { "grad_norm": 0.584466278553009, "learning_rate": 9.779256691999829e-05, "loss": 0.0973, "step": 14020 }, { "grad_norm": 0.8250386714935303, "learning_rate": 9.778770557461403e-05, "loss": 0.1065, "step": 14030 }, { "grad_norm": 0.6236258149147034, "learning_rate": 9.778283900323812e-05, "loss": 0.0776, "step": 14040 }, { "grad_norm": 0.5815744996070862, "learning_rate": 9.777796720640277e-05, "loss": 0.0888, "step": 14050 }, { "grad_norm": 0.4902327358722687, "learning_rate": 9.777309018464078e-05, "loss": 0.0703, "step": 14060 }, { "grad_norm": 0.4979132115840912, "learning_rate": 9.776820793848547e-05, "loss": 0.0781, "step": 14070 }, { "grad_norm": 0.6972362995147705, "learning_rate": 9.776332046847075e-05, "loss": 0.0712, "step": 14080 }, { "grad_norm": 0.8066966533660889, "learning_rate": 9.775842777513111e-05, "loss": 0.0817, "step": 14090 }, { "grad_norm": 0.6084311604499817, "learning_rate": 9.775352985900163e-05, "loss": 0.0925, "step": 14100 }, { "grad_norm": 0.6869087815284729, "learning_rate": 9.774862672061791e-05, "loss": 0.0834, "step": 14110 }, { "grad_norm": 0.5165315270423889, "learning_rate": 9.774371836051616e-05, "loss": 0.0887, "step": 14120 }, { "grad_norm": 0.8290349245071411, "learning_rate": 9.773880477923315e-05, "loss": 0.0809, "step": 14130 }, { "grad_norm": 0.7077469825744629, "learning_rate": 9.773388597730623e-05, "loss": 0.0953, "step": 14140 }, { "grad_norm": 0.8022360801696777, "learning_rate": 9.77289619552733e-05, "loss": 0.0929, "step": 14150 }, { "grad_norm": 0.7381979823112488, "learning_rate": 9.772403271367285e-05, "loss": 0.0924, "step": 14160 }, { "grad_norm": 0.4629620611667633, "learning_rate": 9.771909825304396e-05, "loss": 0.0881, "step": 14170 }, { "grad_norm": 0.5582678318023682, "learning_rate": 9.771415857392619e-05, "loss": 0.0849, "step": 14180 }, { "grad_norm": 0.8234919905662537, "learning_rate": 9.770921367685978e-05, "loss": 0.0991, "step": 14190 }, { "grad_norm": 0.6782072186470032, "learning_rate": 9.770426356238551e-05, "loss": 0.0816, "step": 14200 }, { "grad_norm": 0.5733367204666138, "learning_rate": 9.769930823104469e-05, "loss": 0.095, "step": 14210 }, { "grad_norm": 0.6977033615112305, "learning_rate": 9.769434768337926e-05, "loss": 0.0873, "step": 14220 }, { "grad_norm": 0.6239496469497681, "learning_rate": 9.768938191993164e-05, "loss": 0.0869, "step": 14230 }, { "grad_norm": 0.6437998414039612, "learning_rate": 9.768441094124494e-05, "loss": 0.0803, "step": 14240 }, { "grad_norm": 0.6548511385917664, "learning_rate": 9.767943474786275e-05, "loss": 0.0919, "step": 14250 }, { "grad_norm": 0.5916740894317627, "learning_rate": 9.767445334032923e-05, "loss": 0.0641, "step": 14260 }, { "grad_norm": 0.8452873826026917, "learning_rate": 9.766946671918919e-05, "loss": 0.0915, "step": 14270 }, { "grad_norm": 0.7266057133674622, "learning_rate": 9.766447488498796e-05, "loss": 0.075, "step": 14280 }, { "grad_norm": 0.5714737772941589, "learning_rate": 9.765947783827139e-05, "loss": 0.0787, "step": 14290 }, { "grad_norm": 0.7536842823028564, "learning_rate": 9.765447557958599e-05, "loss": 0.0825, "step": 14300 }, { "grad_norm": 0.5640653371810913, "learning_rate": 9.764946810947879e-05, "loss": 0.072, "step": 14310 }, { "grad_norm": 0.43459635972976685, "learning_rate": 9.764445542849738e-05, "loss": 0.0941, "step": 14320 }, { "grad_norm": 0.6248339414596558, "learning_rate": 9.763943753718998e-05, "loss": 0.0773, "step": 14330 }, { "grad_norm": 0.4828435182571411, "learning_rate": 9.76344144361053e-05, "loss": 0.083, "step": 14340 }, { "grad_norm": 0.699201226234436, "learning_rate": 9.762938612579269e-05, "loss": 0.0915, "step": 14350 }, { "grad_norm": 0.6715221405029297, "learning_rate": 9.762435260680202e-05, "loss": 0.0811, "step": 14360 }, { "grad_norm": 0.6505060195922852, "learning_rate": 9.761931387968373e-05, "loss": 0.0919, "step": 14370 }, { "grad_norm": 0.33009329438209534, "learning_rate": 9.76142699449889e-05, "loss": 0.0735, "step": 14380 }, { "grad_norm": 0.5193867087364197, "learning_rate": 9.760922080326908e-05, "loss": 0.0823, "step": 14390 }, { "grad_norm": 0.7354573607444763, "learning_rate": 9.760416645507644e-05, "loss": 0.0859, "step": 14400 }, { "grad_norm": 0.5504603981971741, "learning_rate": 9.759910690096375e-05, "loss": 0.0774, "step": 14410 }, { "grad_norm": 0.647148847579956, "learning_rate": 9.759404214148429e-05, "loss": 0.0766, "step": 14420 }, { "grad_norm": 0.5669891238212585, "learning_rate": 9.758897217719191e-05, "loss": 0.0794, "step": 14430 }, { "grad_norm": 0.5186252593994141, "learning_rate": 9.758389700864113e-05, "loss": 0.0668, "step": 14440 }, { "grad_norm": 0.878113329410553, "learning_rate": 9.757881663638688e-05, "loss": 0.0845, "step": 14450 }, { "grad_norm": 0.5453450083732605, "learning_rate": 9.757373106098478e-05, "loss": 0.079, "step": 14460 }, { "grad_norm": 0.6025773882865906, "learning_rate": 9.756864028299097e-05, "loss": 0.0839, "step": 14470 }, { "grad_norm": 0.47140592336654663, "learning_rate": 9.75635443029622e-05, "loss": 0.075, "step": 14480 }, { "grad_norm": 0.6567625999450684, "learning_rate": 9.755844312145572e-05, "loss": 0.083, "step": 14490 }, { "grad_norm": 0.7144903540611267, "learning_rate": 9.755333673902941e-05, "loss": 0.1096, "step": 14500 }, { "grad_norm": 0.6257352232933044, "learning_rate": 9.75482251562417e-05, "loss": 0.0847, "step": 14510 }, { "grad_norm": 0.4366791844367981, "learning_rate": 9.754310837365155e-05, "loss": 0.0776, "step": 14520 }, { "grad_norm": 0.49567192792892456, "learning_rate": 9.753798639181856e-05, "loss": 0.1057, "step": 14530 }, { "grad_norm": 0.7401109337806702, "learning_rate": 9.753285921130286e-05, "loss": 0.068, "step": 14540 }, { "grad_norm": 0.7701501250267029, "learning_rate": 9.752772683266512e-05, "loss": 0.0766, "step": 14550 }, { "grad_norm": 0.8468741178512573, "learning_rate": 9.752258925646665e-05, "loss": 0.0805, "step": 14560 }, { "grad_norm": 0.628916323184967, "learning_rate": 9.751744648326926e-05, "loss": 0.0665, "step": 14570 }, { "grad_norm": 0.6352697014808655, "learning_rate": 9.751229851363536e-05, "loss": 0.0799, "step": 14580 }, { "grad_norm": 0.7047574520111084, "learning_rate": 9.750714534812793e-05, "loss": 0.0976, "step": 14590 }, { "grad_norm": 0.5749344229698181, "learning_rate": 9.750198698731053e-05, "loss": 0.0866, "step": 14600 }, { "grad_norm": 0.5863544940948486, "learning_rate": 9.749682343174722e-05, "loss": 0.0755, "step": 14610 }, { "grad_norm": 0.390047162771225, "learning_rate": 9.749165468200272e-05, "loss": 0.0774, "step": 14620 }, { "grad_norm": 0.5303566455841064, "learning_rate": 9.748648073864229e-05, "loss": 0.0815, "step": 14630 }, { "grad_norm": 0.5209925174713135, "learning_rate": 9.748130160223168e-05, "loss": 0.0722, "step": 14640 }, { "grad_norm": 0.8584150075912476, "learning_rate": 9.747611727333734e-05, "loss": 0.0887, "step": 14650 }, { "grad_norm": 0.5791692733764648, "learning_rate": 9.74709277525262e-05, "loss": 0.0846, "step": 14660 }, { "grad_norm": 0.4431767463684082, "learning_rate": 9.746573304036576e-05, "loss": 0.0846, "step": 14670 }, { "grad_norm": 0.8075158596038818, "learning_rate": 9.746053313742412e-05, "loss": 0.0696, "step": 14680 }, { "grad_norm": 0.8913453221321106, "learning_rate": 9.745532804426994e-05, "loss": 0.0941, "step": 14690 }, { "grad_norm": 0.6127343773841858, "learning_rate": 9.745011776147242e-05, "loss": 0.0775, "step": 14700 }, { "grad_norm": 0.7772984504699707, "learning_rate": 9.744490228960138e-05, "loss": 0.0909, "step": 14710 }, { "grad_norm": 0.6666492223739624, "learning_rate": 9.743968162922713e-05, "loss": 0.0774, "step": 14720 }, { "grad_norm": 0.5625427961349487, "learning_rate": 9.743445578092064e-05, "loss": 0.0962, "step": 14730 }, { "grad_norm": 0.596257746219635, "learning_rate": 9.742922474525338e-05, "loss": 0.1027, "step": 14740 }, { "grad_norm": 0.6389324069023132, "learning_rate": 9.742398852279741e-05, "loss": 0.0713, "step": 14750 }, { "grad_norm": 0.5979225635528564, "learning_rate": 9.741874711412535e-05, "loss": 0.0933, "step": 14760 }, { "grad_norm": 0.6446190476417542, "learning_rate": 9.741350051981042e-05, "loss": 0.0737, "step": 14770 }, { "grad_norm": 0.6321583390235901, "learning_rate": 9.740824874042633e-05, "loss": 0.0654, "step": 14780 }, { "grad_norm": 1.0629420280456543, "learning_rate": 9.740299177654746e-05, "loss": 0.0839, "step": 14790 }, { "grad_norm": 0.6516163945198059, "learning_rate": 9.739772962874867e-05, "loss": 0.0835, "step": 14800 }, { "grad_norm": 0.6588464379310608, "learning_rate": 9.739246229760541e-05, "loss": 0.0765, "step": 14810 }, { "grad_norm": 1.0547996759414673, "learning_rate": 9.738718978369376e-05, "loss": 0.0887, "step": 14820 }, { "grad_norm": 0.579478919506073, "learning_rate": 9.738191208759025e-05, "loss": 0.0931, "step": 14830 }, { "grad_norm": 0.7284175157546997, "learning_rate": 9.73766292098721e-05, "loss": 0.0989, "step": 14840 }, { "grad_norm": 0.5170187950134277, "learning_rate": 9.737134115111699e-05, "loss": 0.0961, "step": 14850 }, { "grad_norm": 0.6415605545043945, "learning_rate": 9.736604791190323e-05, "loss": 0.0813, "step": 14860 }, { "grad_norm": 0.582787275314331, "learning_rate": 9.73607494928097e-05, "loss": 0.0703, "step": 14870 }, { "grad_norm": 0.5411022901535034, "learning_rate": 9.735544589441581e-05, "loss": 0.0831, "step": 14880 }, { "grad_norm": 0.696735143661499, "learning_rate": 9.735013711730154e-05, "loss": 0.071, "step": 14890 }, { "grad_norm": 1.382435917854309, "learning_rate": 9.734482316204747e-05, "loss": 0.0773, "step": 14900 }, { "grad_norm": 0.7975710034370422, "learning_rate": 9.733950402923473e-05, "loss": 0.0803, "step": 14910 }, { "grad_norm": 0.5648559331893921, "learning_rate": 9.7334179719445e-05, "loss": 0.0779, "step": 14920 }, { "grad_norm": 0.6834574937820435, "learning_rate": 9.732885023326053e-05, "loss": 0.095, "step": 14930 }, { "grad_norm": 0.5370402932167053, "learning_rate": 9.732351557126418e-05, "loss": 0.105, "step": 14940 }, { "grad_norm": 0.6276283264160156, "learning_rate": 9.731817573403929e-05, "loss": 0.084, "step": 14950 }, { "grad_norm": 0.8168245553970337, "learning_rate": 9.731283072216985e-05, "loss": 0.0879, "step": 14960 }, { "grad_norm": 0.688232958316803, "learning_rate": 9.730748053624039e-05, "loss": 0.0697, "step": 14970 }, { "grad_norm": 0.5113072991371155, "learning_rate": 9.730212517683598e-05, "loss": 0.0923, "step": 14980 }, { "grad_norm": 0.498046338558197, "learning_rate": 9.729676464454228e-05, "loss": 0.0808, "step": 14990 }, { "grad_norm": 0.4776412844657898, "learning_rate": 9.72913989399455e-05, "loss": 0.0745, "step": 15000 }, { "grad_norm": 0.5702039003372192, "learning_rate": 9.728602806363242e-05, "loss": 0.0716, "step": 15010 }, { "grad_norm": 0.563701868057251, "learning_rate": 9.728065201619043e-05, "loss": 0.0806, "step": 15020 }, { "grad_norm": 0.5489841103553772, "learning_rate": 9.727527079820742e-05, "loss": 0.078, "step": 15030 }, { "grad_norm": 0.6193112134933472, "learning_rate": 9.726988441027186e-05, "loss": 0.0931, "step": 15040 }, { "grad_norm": 2.0933797359466553, "learning_rate": 9.726449285297281e-05, "loss": 0.098, "step": 15050 }, { "grad_norm": 0.6876230835914612, "learning_rate": 9.72590961268999e-05, "loss": 0.0988, "step": 15060 }, { "grad_norm": 0.7375438809394836, "learning_rate": 9.725369423264328e-05, "loss": 0.0905, "step": 15070 }, { "grad_norm": 0.6674543023109436, "learning_rate": 9.72482871707937e-05, "loss": 0.0796, "step": 15080 }, { "grad_norm": 0.658123254776001, "learning_rate": 9.724287494194247e-05, "loss": 0.0961, "step": 15090 }, { "grad_norm": 0.47452130913734436, "learning_rate": 9.723745754668147e-05, "loss": 0.0872, "step": 15100 }, { "grad_norm": 0.5692800879478455, "learning_rate": 9.723203498560313e-05, "loss": 0.0906, "step": 15110 }, { "grad_norm": 0.35905471444129944, "learning_rate": 9.722660725930046e-05, "loss": 0.07, "step": 15120 }, { "grad_norm": 0.5711699724197388, "learning_rate": 9.722117436836702e-05, "loss": 0.0797, "step": 15130 }, { "grad_norm": 0.4502703547477722, "learning_rate": 9.721573631339696e-05, "loss": 0.0989, "step": 15140 }, { "grad_norm": 1.1277436017990112, "learning_rate": 9.721029309498494e-05, "loss": 0.1052, "step": 15150 }, { "grad_norm": 0.6442707777023315, "learning_rate": 9.720484471372627e-05, "loss": 0.0913, "step": 15160 }, { "grad_norm": 0.5886473059654236, "learning_rate": 9.719939117021673e-05, "loss": 0.0904, "step": 15170 }, { "grad_norm": 0.643998920917511, "learning_rate": 9.719393246505275e-05, "loss": 0.096, "step": 15180 }, { "grad_norm": 0.8694250583648682, "learning_rate": 9.718846859883128e-05, "loss": 0.0829, "step": 15190 }, { "grad_norm": 0.530680775642395, "learning_rate": 9.718299957214982e-05, "loss": 0.0988, "step": 15200 }, { "grad_norm": 0.5598941445350647, "learning_rate": 9.717752538560646e-05, "loss": 0.0896, "step": 15210 }, { "grad_norm": 0.6579728722572327, "learning_rate": 9.717204603979986e-05, "loss": 0.0929, "step": 15220 }, { "grad_norm": 0.8543526530265808, "learning_rate": 9.716656153532922e-05, "loss": 0.0801, "step": 15230 }, { "grad_norm": 0.602909505367279, "learning_rate": 9.716107187279434e-05, "loss": 0.0939, "step": 15240 }, { "grad_norm": 0.6986809968948364, "learning_rate": 9.715557705279555e-05, "loss": 0.1015, "step": 15250 }, { "grad_norm": 0.6116282343864441, "learning_rate": 9.715007707593372e-05, "loss": 0.102, "step": 15260 }, { "grad_norm": 0.6827844977378845, "learning_rate": 9.714457194281036e-05, "loss": 0.0765, "step": 15270 }, { "grad_norm": 0.6216919422149658, "learning_rate": 9.713906165402751e-05, "loss": 0.084, "step": 15280 }, { "grad_norm": 0.5648477673530579, "learning_rate": 9.713354621018774e-05, "loss": 0.0987, "step": 15290 }, { "grad_norm": 0.66669762134552, "learning_rate": 9.712802561189422e-05, "loss": 0.0708, "step": 15300 }, { "grad_norm": 0.5520405769348145, "learning_rate": 9.712249985975069e-05, "loss": 0.0896, "step": 15310 }, { "grad_norm": 0.5307103991508484, "learning_rate": 9.71169689543614e-05, "loss": 0.0758, "step": 15320 }, { "grad_norm": 0.6976937651634216, "learning_rate": 9.711143289633123e-05, "loss": 0.0901, "step": 15330 }, { "grad_norm": 0.7356052398681641, "learning_rate": 9.710589168626561e-05, "loss": 0.0892, "step": 15340 }, { "grad_norm": 0.7153535485267639, "learning_rate": 9.710034532477048e-05, "loss": 0.0935, "step": 15350 }, { "grad_norm": 0.5132650136947632, "learning_rate": 9.709479381245239e-05, "loss": 0.0737, "step": 15360 }, { "grad_norm": 0.5689718723297119, "learning_rate": 9.708923714991847e-05, "loss": 0.0879, "step": 15370 }, { "grad_norm": 0.6526865363121033, "learning_rate": 9.708367533777638e-05, "loss": 0.0788, "step": 15380 }, { "grad_norm": 0.49547407031059265, "learning_rate": 9.707810837663431e-05, "loss": 0.0773, "step": 15390 }, { "grad_norm": 0.7782779932022095, "learning_rate": 9.707253626710113e-05, "loss": 0.0893, "step": 15400 }, { "grad_norm": 0.7622187733650208, "learning_rate": 9.706695900978613e-05, "loss": 0.0758, "step": 15410 }, { "grad_norm": 1.0457110404968262, "learning_rate": 9.706137660529926e-05, "loss": 0.0901, "step": 15420 }, { "grad_norm": 0.6016637682914734, "learning_rate": 9.705578905425101e-05, "loss": 0.0963, "step": 15430 }, { "grad_norm": 0.5284295082092285, "learning_rate": 9.705019635725241e-05, "loss": 0.0972, "step": 15440 }, { "grad_norm": 0.5151387453079224, "learning_rate": 9.704459851491508e-05, "loss": 0.0967, "step": 15450 }, { "grad_norm": 0.8921122550964355, "learning_rate": 9.703899552785118e-05, "loss": 0.0799, "step": 15460 }, { "grad_norm": 0.7843775749206543, "learning_rate": 9.703338739667346e-05, "loss": 0.0819, "step": 15470 }, { "grad_norm": 0.633996844291687, "learning_rate": 9.70277741219952e-05, "loss": 0.0896, "step": 15480 }, { "grad_norm": 0.7445108890533447, "learning_rate": 9.702215570443027e-05, "loss": 0.0917, "step": 15490 }, { "grad_norm": 0.48475146293640137, "learning_rate": 9.701653214459309e-05, "loss": 0.0864, "step": 15500 }, { "grad_norm": 0.6225465536117554, "learning_rate": 9.701090344309865e-05, "loss": 0.0918, "step": 15510 }, { "grad_norm": 1.0989521741867065, "learning_rate": 9.700526960056247e-05, "loss": 0.0779, "step": 15520 }, { "grad_norm": 0.5786524415016174, "learning_rate": 9.699963061760068e-05, "loss": 0.07, "step": 15530 }, { "grad_norm": 0.5878774523735046, "learning_rate": 9.699398649482997e-05, "loss": 0.0812, "step": 15540 }, { "grad_norm": 0.8789093494415283, "learning_rate": 9.698833723286753e-05, "loss": 0.0718, "step": 15550 }, { "grad_norm": 0.5616891384124756, "learning_rate": 9.698268283233118e-05, "loss": 0.089, "step": 15560 }, { "grad_norm": 0.9504624605178833, "learning_rate": 9.697702329383929e-05, "loss": 0.1028, "step": 15570 }, { "grad_norm": 0.4226412773132324, "learning_rate": 9.697135861801074e-05, "loss": 0.0922, "step": 15580 }, { "grad_norm": 0.5309758186340332, "learning_rate": 9.696568880546505e-05, "loss": 0.0771, "step": 15590 }, { "grad_norm": 0.8663474917411804, "learning_rate": 9.696001385682223e-05, "loss": 0.0813, "step": 15600 }, { "grad_norm": 0.6590249538421631, "learning_rate": 9.695433377270291e-05, "loss": 0.0779, "step": 15610 }, { "grad_norm": 0.5204740166664124, "learning_rate": 9.694864855372824e-05, "loss": 0.0858, "step": 15620 }, { "grad_norm": 0.7677971720695496, "learning_rate": 9.694295820051995e-05, "loss": 0.0861, "step": 15630 }, { "grad_norm": 0.7705398201942444, "learning_rate": 9.693726271370032e-05, "loss": 0.0839, "step": 15640 }, { "grad_norm": 0.7519357800483704, "learning_rate": 9.693156209389221e-05, "loss": 0.0797, "step": 15650 }, { "grad_norm": 0.5605280995368958, "learning_rate": 9.692585634171905e-05, "loss": 0.0801, "step": 15660 }, { "grad_norm": 0.4859650135040283, "learning_rate": 9.692014545780476e-05, "loss": 0.0752, "step": 15670 }, { "grad_norm": 0.6952943205833435, "learning_rate": 9.691442944277393e-05, "loss": 0.0781, "step": 15680 }, { "grad_norm": 0.558408796787262, "learning_rate": 9.690870829725162e-05, "loss": 0.0695, "step": 15690 }, { "grad_norm": 0.697536826133728, "learning_rate": 9.69029820218635e-05, "loss": 0.0881, "step": 15700 }, { "grad_norm": 0.6255780458450317, "learning_rate": 9.689725061723579e-05, "loss": 0.0902, "step": 15710 }, { "grad_norm": 0.5730352997779846, "learning_rate": 9.689151408399527e-05, "loss": 0.0813, "step": 15720 }, { "grad_norm": 0.739517092704773, "learning_rate": 9.688577242276924e-05, "loss": 0.0984, "step": 15730 }, { "grad_norm": 0.9657573699951172, "learning_rate": 9.688002563418566e-05, "loss": 0.0944, "step": 15740 }, { "grad_norm": 0.5917686820030212, "learning_rate": 9.687427371887293e-05, "loss": 0.0901, "step": 15750 }, { "grad_norm": 0.5938595533370972, "learning_rate": 9.686851667746012e-05, "loss": 0.081, "step": 15760 }, { "grad_norm": 0.5236178636550903, "learning_rate": 9.686275451057677e-05, "loss": 0.0699, "step": 15770 }, { "grad_norm": 0.8366506695747375, "learning_rate": 9.685698721885308e-05, "loss": 0.0811, "step": 15780 }, { "grad_norm": 0.9042981266975403, "learning_rate": 9.68512148029197e-05, "loss": 0.0822, "step": 15790 }, { "grad_norm": 0.5983405113220215, "learning_rate": 9.684543726340791e-05, "loss": 0.0881, "step": 15800 }, { "grad_norm": 0.6780979037284851, "learning_rate": 9.683965460094952e-05, "loss": 0.0785, "step": 15810 }, { "grad_norm": 0.5872125625610352, "learning_rate": 9.683386681617694e-05, "loss": 0.0794, "step": 15820 }, { "grad_norm": 0.5438132286071777, "learning_rate": 9.68280739097231e-05, "loss": 0.0808, "step": 15830 }, { "grad_norm": 0.6467993855476379, "learning_rate": 9.682227588222148e-05, "loss": 0.0729, "step": 15840 }, { "grad_norm": 0.9512289762496948, "learning_rate": 9.681647273430618e-05, "loss": 0.0642, "step": 15850 }, { "grad_norm": 0.7438040971755981, "learning_rate": 9.681066446661182e-05, "loss": 0.0785, "step": 15860 }, { "grad_norm": 0.8150305151939392, "learning_rate": 9.680485107977357e-05, "loss": 0.0816, "step": 15870 }, { "grad_norm": 0.668017566204071, "learning_rate": 9.679903257442716e-05, "loss": 0.092, "step": 15880 }, { "grad_norm": 0.7997297048568726, "learning_rate": 9.679320895120891e-05, "loss": 0.0843, "step": 15890 }, { "grad_norm": 0.4689437747001648, "learning_rate": 9.67873802107557e-05, "loss": 0.0838, "step": 15900 }, { "grad_norm": 0.5046187043190002, "learning_rate": 9.67815463537049e-05, "loss": 0.0791, "step": 15910 }, { "grad_norm": 0.6389027237892151, "learning_rate": 9.677570738069457e-05, "loss": 0.0714, "step": 15920 }, { "grad_norm": 0.5926241278648376, "learning_rate": 9.676986329236318e-05, "loss": 0.0976, "step": 15930 }, { "grad_norm": 0.7680691480636597, "learning_rate": 9.676401408934987e-05, "loss": 0.0824, "step": 15940 }, { "grad_norm": 0.5442052483558655, "learning_rate": 9.675815977229428e-05, "loss": 0.0906, "step": 15950 }, { "grad_norm": 0.6126217246055603, "learning_rate": 9.675230034183664e-05, "loss": 0.0729, "step": 15960 }, { "grad_norm": 0.6875811815261841, "learning_rate": 9.674643579861773e-05, "loss": 0.1032, "step": 15970 }, { "grad_norm": 0.9597428441047668, "learning_rate": 9.674056614327886e-05, "loss": 0.0873, "step": 15980 }, { "grad_norm": 0.5300121307373047, "learning_rate": 9.673469137646198e-05, "loss": 0.0956, "step": 15990 }, { "grad_norm": 0.7614507079124451, "learning_rate": 9.67288114988095e-05, "loss": 0.0787, "step": 16000 }, { "grad_norm": 0.6395676136016846, "learning_rate": 9.672292651096447e-05, "loss": 0.0839, "step": 16010 }, { "grad_norm": 0.4866299629211426, "learning_rate": 9.671703641357042e-05, "loss": 0.0714, "step": 16020 }, { "grad_norm": 0.5323402285575867, "learning_rate": 9.67111412072715e-05, "loss": 0.0901, "step": 16030 }, { "grad_norm": 1.4280627965927124, "learning_rate": 9.670524089271242e-05, "loss": 0.0894, "step": 16040 }, { "grad_norm": 0.56093829870224, "learning_rate": 9.669933547053842e-05, "loss": 0.0758, "step": 16050 }, { "grad_norm": 0.5921887159347534, "learning_rate": 9.669342494139531e-05, "loss": 0.084, "step": 16060 }, { "grad_norm": 0.5239952206611633, "learning_rate": 9.668750930592943e-05, "loss": 0.077, "step": 16070 }, { "grad_norm": 0.5640283823013306, "learning_rate": 9.668158856478775e-05, "loss": 0.0777, "step": 16080 }, { "grad_norm": 0.71526038646698, "learning_rate": 9.66756627186177e-05, "loss": 0.0934, "step": 16090 }, { "grad_norm": 0.6623452305793762, "learning_rate": 9.666973176806737e-05, "loss": 0.077, "step": 16100 }, { "grad_norm": 0.5103299617767334, "learning_rate": 9.666379571378534e-05, "loss": 0.0818, "step": 16110 }, { "grad_norm": 0.8203185200691223, "learning_rate": 9.665785455642076e-05, "loss": 0.1031, "step": 16120 }, { "grad_norm": 0.6974210143089294, "learning_rate": 9.665190829662337e-05, "loss": 0.0993, "step": 16130 }, { "grad_norm": 0.5391071438789368, "learning_rate": 9.664595693504342e-05, "loss": 0.073, "step": 16140 }, { "grad_norm": 0.5298978090286255, "learning_rate": 9.664000047233175e-05, "loss": 0.1037, "step": 16150 }, { "grad_norm": 0.48936086893081665, "learning_rate": 9.663403890913976e-05, "loss": 0.0731, "step": 16160 }, { "grad_norm": 0.8031927347183228, "learning_rate": 9.662807224611938e-05, "loss": 0.0844, "step": 16170 }, { "grad_norm": 0.5076702833175659, "learning_rate": 9.662210048392311e-05, "loss": 0.0763, "step": 16180 }, { "grad_norm": 0.6915735602378845, "learning_rate": 9.661612362320405e-05, "loss": 0.0787, "step": 16190 }, { "grad_norm": 0.7320131659507751, "learning_rate": 9.661014166461579e-05, "loss": 0.0827, "step": 16200 }, { "grad_norm": 0.5613679885864258, "learning_rate": 9.66041546088125e-05, "loss": 0.089, "step": 16210 }, { "grad_norm": 0.385267972946167, "learning_rate": 9.659816245644895e-05, "loss": 0.0684, "step": 16220 }, { "grad_norm": 0.957634449005127, "learning_rate": 9.65921652081804e-05, "loss": 0.0796, "step": 16230 }, { "grad_norm": 1.0101913213729858, "learning_rate": 9.658616286466271e-05, "loss": 0.0896, "step": 16240 }, { "grad_norm": 0.49439898133277893, "learning_rate": 9.65801554265523e-05, "loss": 0.0907, "step": 16250 }, { "grad_norm": 0.8610667586326599, "learning_rate": 9.657414289450612e-05, "loss": 0.0856, "step": 16260 }, { "grad_norm": 0.9769374132156372, "learning_rate": 9.656812526918171e-05, "loss": 0.0937, "step": 16270 }, { "grad_norm": 0.6994975209236145, "learning_rate": 9.656210255123712e-05, "loss": 0.0901, "step": 16280 }, { "grad_norm": 0.6854737997055054, "learning_rate": 9.6556074741331e-05, "loss": 0.0767, "step": 16290 }, { "grad_norm": 0.45677322149276733, "learning_rate": 9.655004184012256e-05, "loss": 0.0752, "step": 16300 }, { "grad_norm": 0.7112623453140259, "learning_rate": 9.654400384827152e-05, "loss": 0.0896, "step": 16310 }, { "grad_norm": 0.8396608829498291, "learning_rate": 9.653796076643818e-05, "loss": 0.0693, "step": 16320 }, { "grad_norm": 0.5503895878791809, "learning_rate": 9.653191259528344e-05, "loss": 0.0717, "step": 16330 }, { "grad_norm": 0.47286441922187805, "learning_rate": 9.65258593354687e-05, "loss": 0.0773, "step": 16340 }, { "grad_norm": 0.5246687531471252, "learning_rate": 9.651980098765591e-05, "loss": 0.0703, "step": 16350 }, { "grad_norm": 0.7861930131912231, "learning_rate": 9.651373755250765e-05, "loss": 0.0954, "step": 16360 }, { "grad_norm": 0.6841748952865601, "learning_rate": 9.650766903068697e-05, "loss": 0.0857, "step": 16370 }, { "grad_norm": 0.6214224696159363, "learning_rate": 9.650159542285753e-05, "loss": 0.0675, "step": 16380 }, { "grad_norm": 0.7060206532478333, "learning_rate": 9.649551672968353e-05, "loss": 0.1006, "step": 16390 }, { "grad_norm": 0.6354573965072632, "learning_rate": 9.648943295182973e-05, "loss": 0.0836, "step": 16400 }, { "grad_norm": 0.4723997712135315, "learning_rate": 9.648334408996144e-05, "loss": 0.0797, "step": 16410 }, { "grad_norm": 0.4688185155391693, "learning_rate": 9.647725014474452e-05, "loss": 0.0801, "step": 16420 }, { "grad_norm": 0.4305747449398041, "learning_rate": 9.64711511168454e-05, "loss": 0.0848, "step": 16430 }, { "grad_norm": 0.5460051894187927, "learning_rate": 9.646504700693108e-05, "loss": 0.0857, "step": 16440 }, { "grad_norm": 0.7157851457595825, "learning_rate": 9.645893781566907e-05, "loss": 0.0807, "step": 16450 }, { "grad_norm": 0.7469001412391663, "learning_rate": 9.645282354372744e-05, "loss": 0.1072, "step": 16460 }, { "grad_norm": 0.8174909353256226, "learning_rate": 9.644670419177491e-05, "loss": 0.0864, "step": 16470 }, { "grad_norm": 0.7567616105079651, "learning_rate": 9.644057976048062e-05, "loss": 0.0818, "step": 16480 }, { "grad_norm": 0.8452866077423096, "learning_rate": 9.643445025051435e-05, "loss": 0.0846, "step": 16490 }, { "grad_norm": 0.975002110004425, "learning_rate": 9.642831566254641e-05, "loss": 0.0765, "step": 16500 }, { "grad_norm": 0.7489939332008362, "learning_rate": 9.642217599724769e-05, "loss": 0.0739, "step": 16510 }, { "grad_norm": 0.49997368454933167, "learning_rate": 9.64160312552896e-05, "loss": 0.0801, "step": 16520 }, { "grad_norm": 0.4247101843357086, "learning_rate": 9.64098814373441e-05, "loss": 0.0726, "step": 16530 }, { "grad_norm": 0.6261866092681885, "learning_rate": 9.640372654408374e-05, "loss": 0.0808, "step": 16540 }, { "grad_norm": 0.6364086866378784, "learning_rate": 9.639756657618162e-05, "loss": 0.0915, "step": 16550 }, { "grad_norm": 0.6566647291183472, "learning_rate": 9.639140153431138e-05, "loss": 0.0797, "step": 16560 }, { "grad_norm": 0.7240087389945984, "learning_rate": 9.638523141914721e-05, "loss": 0.0885, "step": 16570 }, { "grad_norm": 0.7135177850723267, "learning_rate": 9.637905623136388e-05, "loss": 0.092, "step": 16580 }, { "grad_norm": 0.9516217708587646, "learning_rate": 9.637287597163669e-05, "loss": 0.079, "step": 16590 }, { "grad_norm": 0.8034681677818298, "learning_rate": 9.63666906406415e-05, "loss": 0.0786, "step": 16600 }, { "grad_norm": 0.4640374779701233, "learning_rate": 9.636050023905473e-05, "loss": 0.0735, "step": 16610 }, { "grad_norm": 0.7402284741401672, "learning_rate": 9.635430476755336e-05, "loss": 0.0711, "step": 16620 }, { "grad_norm": 0.45507776737213135, "learning_rate": 9.63481042268149e-05, "loss": 0.0702, "step": 16630 }, { "grad_norm": 0.5533487796783447, "learning_rate": 9.634189861751745e-05, "loss": 0.0793, "step": 16640 }, { "grad_norm": 0.4644881784915924, "learning_rate": 9.633568794033967e-05, "loss": 0.0855, "step": 16650 }, { "grad_norm": 0.48521167039871216, "learning_rate": 9.63294721959607e-05, "loss": 0.0737, "step": 16660 }, { "grad_norm": 0.47457435727119446, "learning_rate": 9.63232513850603e-05, "loss": 0.0959, "step": 16670 }, { "grad_norm": 0.6862094402313232, "learning_rate": 9.631702550831878e-05, "loss": 0.0899, "step": 16680 }, { "grad_norm": 0.633672833442688, "learning_rate": 9.631079456641698e-05, "loss": 0.0766, "step": 16690 }, { "grad_norm": 0.5358520746231079, "learning_rate": 9.630455856003632e-05, "loss": 0.085, "step": 16700 }, { "grad_norm": 0.5803004503250122, "learning_rate": 9.629831748985876e-05, "loss": 0.0879, "step": 16710 }, { "grad_norm": 0.741970956325531, "learning_rate": 9.629207135656679e-05, "loss": 0.0884, "step": 16720 }, { "grad_norm": 0.5222972631454468, "learning_rate": 9.628582016084353e-05, "loss": 0.0762, "step": 16730 }, { "grad_norm": 0.6948007941246033, "learning_rate": 9.627956390337254e-05, "loss": 0.0698, "step": 16740 }, { "grad_norm": 0.6036452651023865, "learning_rate": 9.627330258483802e-05, "loss": 0.0773, "step": 16750 }, { "grad_norm": 0.6053318977355957, "learning_rate": 9.62670362059247e-05, "loss": 0.064, "step": 16760 }, { "grad_norm": 0.6019712090492249, "learning_rate": 9.626076476731786e-05, "loss": 0.0747, "step": 16770 }, { "grad_norm": 0.6910335421562195, "learning_rate": 9.625448826970336e-05, "loss": 0.087, "step": 16780 }, { "grad_norm": 1.5603195428848267, "learning_rate": 9.624820671376755e-05, "loss": 0.0813, "step": 16790 }, { "grad_norm": 0.7658224701881409, "learning_rate": 9.62419201001974e-05, "loss": 0.0753, "step": 16800 }, { "grad_norm": 0.5137438178062439, "learning_rate": 9.623562842968037e-05, "loss": 0.0727, "step": 16810 }, { "grad_norm": 0.7007177472114563, "learning_rate": 9.622933170290454e-05, "loss": 0.0754, "step": 16820 }, { "grad_norm": 0.5349214673042297, "learning_rate": 9.622302992055849e-05, "loss": 0.0778, "step": 16830 }, { "grad_norm": 0.6326510906219482, "learning_rate": 9.62167230833314e-05, "loss": 0.0736, "step": 16840 }, { "grad_norm": 0.6422044038772583, "learning_rate": 9.621041119191295e-05, "loss": 0.08, "step": 16850 }, { "grad_norm": 0.7414842844009399, "learning_rate": 9.620409424699342e-05, "loss": 0.0822, "step": 16860 }, { "grad_norm": 0.7114547491073608, "learning_rate": 9.619777224926359e-05, "loss": 0.0613, "step": 16870 }, { "grad_norm": 0.5258206129074097, "learning_rate": 9.619144519941485e-05, "loss": 0.0776, "step": 16880 }, { "grad_norm": 0.5179718732833862, "learning_rate": 9.618511309813912e-05, "loss": 0.0903, "step": 16890 }, { "grad_norm": 0.7423545718193054, "learning_rate": 9.617877594612886e-05, "loss": 0.0721, "step": 16900 }, { "grad_norm": 0.5123944282531738, "learning_rate": 9.617243374407707e-05, "loss": 0.0693, "step": 16910 }, { "grad_norm": 0.5722377896308899, "learning_rate": 9.616608649267736e-05, "loss": 0.0764, "step": 16920 }, { "grad_norm": 0.7407156825065613, "learning_rate": 9.615973419262385e-05, "loss": 0.069, "step": 16930 }, { "grad_norm": 1.4975885152816772, "learning_rate": 9.615337684461119e-05, "loss": 0.0752, "step": 16940 }, { "grad_norm": 0.6063380241394043, "learning_rate": 9.614701444933465e-05, "loss": 0.0823, "step": 16950 }, { "grad_norm": 0.6922670006752014, "learning_rate": 9.614064700748997e-05, "loss": 0.0823, "step": 16960 }, { "grad_norm": 0.593688428401947, "learning_rate": 9.613427451977352e-05, "loss": 0.0668, "step": 16970 }, { "grad_norm": 1.1764311790466309, "learning_rate": 9.612789698688216e-05, "loss": 0.0832, "step": 16980 }, { "grad_norm": 0.4586765766143799, "learning_rate": 9.612151440951334e-05, "loss": 0.0804, "step": 16990 }, { "grad_norm": 0.6854137778282166, "learning_rate": 9.611512678836506e-05, "loss": 0.0851, "step": 17000 }, { "grad_norm": 0.6050657629966736, "learning_rate": 9.610873412413584e-05, "loss": 0.0814, "step": 17010 }, { "grad_norm": 1.2833877801895142, "learning_rate": 9.610233641752476e-05, "loss": 0.0889, "step": 17020 }, { "grad_norm": 0.8903701901435852, "learning_rate": 9.609593366923151e-05, "loss": 0.0758, "step": 17030 }, { "grad_norm": 1.347121238708496, "learning_rate": 9.608952587995625e-05, "loss": 0.074, "step": 17040 }, { "grad_norm": 0.6425021290779114, "learning_rate": 9.608311305039972e-05, "loss": 0.0886, "step": 17050 }, { "grad_norm": 0.9512551426887512, "learning_rate": 9.607669518126326e-05, "loss": 0.0922, "step": 17060 }, { "grad_norm": 0.5573267340660095, "learning_rate": 9.607027227324866e-05, "loss": 0.0771, "step": 17070 }, { "grad_norm": 0.5517371296882629, "learning_rate": 9.606384432705837e-05, "loss": 0.0792, "step": 17080 }, { "grad_norm": 0.8499151468276978, "learning_rate": 9.60574113433953e-05, "loss": 0.0908, "step": 17090 }, { "grad_norm": 0.5929647088050842, "learning_rate": 9.6050973322963e-05, "loss": 0.0831, "step": 17100 }, { "grad_norm": 0.7105236053466797, "learning_rate": 9.604453026646547e-05, "loss": 0.0814, "step": 17110 }, { "grad_norm": 0.5556855797767639, "learning_rate": 9.603808217460735e-05, "loss": 0.0783, "step": 17120 }, { "grad_norm": 0.7662156820297241, "learning_rate": 9.603162904809377e-05, "loss": 0.0761, "step": 17130 }, { "grad_norm": 0.6917667984962463, "learning_rate": 9.602517088763045e-05, "loss": 0.0716, "step": 17140 }, { "grad_norm": 0.48653119802474976, "learning_rate": 9.601870769392365e-05, "loss": 0.0962, "step": 17150 }, { "grad_norm": 0.616385817527771, "learning_rate": 9.601223946768017e-05, "loss": 0.0718, "step": 17160 }, { "grad_norm": 0.539509654045105, "learning_rate": 9.600576620960734e-05, "loss": 0.078, "step": 17170 }, { "grad_norm": 0.712546706199646, "learning_rate": 9.599928792041308e-05, "loss": 0.0842, "step": 17180 }, { "grad_norm": 0.9613418579101562, "learning_rate": 9.599280460080587e-05, "loss": 0.0802, "step": 17190 }, { "grad_norm": 0.5210888385772705, "learning_rate": 9.59863162514947e-05, "loss": 0.0813, "step": 17200 }, { "grad_norm": 0.7027076482772827, "learning_rate": 9.597982287318911e-05, "loss": 0.0778, "step": 17210 }, { "grad_norm": 0.48687177896499634, "learning_rate": 9.597332446659923e-05, "loss": 0.0827, "step": 17220 }, { "grad_norm": 0.5163702964782715, "learning_rate": 9.59668210324357e-05, "loss": 0.0741, "step": 17230 }, { "grad_norm": 0.49903133511543274, "learning_rate": 9.596031257140974e-05, "loss": 0.0671, "step": 17240 }, { "grad_norm": 0.6437112092971802, "learning_rate": 9.59537990842331e-05, "loss": 0.0896, "step": 17250 }, { "grad_norm": 0.35939860343933105, "learning_rate": 9.594728057161806e-05, "loss": 0.0644, "step": 17260 }, { "grad_norm": 0.6319924592971802, "learning_rate": 9.594075703427752e-05, "loss": 0.0771, "step": 17270 }, { "grad_norm": 0.4568161964416504, "learning_rate": 9.593422847292486e-05, "loss": 0.062, "step": 17280 }, { "grad_norm": 0.5610899329185486, "learning_rate": 9.592769488827402e-05, "loss": 0.0809, "step": 17290 }, { "grad_norm": 0.6896939277648926, "learning_rate": 9.592115628103952e-05, "loss": 0.076, "step": 17300 }, { "grad_norm": 0.5014094114303589, "learning_rate": 9.591461265193643e-05, "loss": 0.0779, "step": 17310 }, { "grad_norm": 1.2296240329742432, "learning_rate": 9.590806400168032e-05, "loss": 0.0712, "step": 17320 }, { "grad_norm": 0.43985268473625183, "learning_rate": 9.590151033098735e-05, "loss": 0.0688, "step": 17330 }, { "grad_norm": 0.5324223041534424, "learning_rate": 9.589495164057423e-05, "loss": 0.0764, "step": 17340 }, { "grad_norm": 0.39354345202445984, "learning_rate": 9.58883879311582e-05, "loss": 0.0798, "step": 17350 }, { "grad_norm": 0.5804018378257751, "learning_rate": 9.588181920345705e-05, "loss": 0.0761, "step": 17360 }, { "grad_norm": 0.4319194257259369, "learning_rate": 9.587524545818913e-05, "loss": 0.0791, "step": 17370 }, { "grad_norm": 0.6432393193244934, "learning_rate": 9.586866669607335e-05, "loss": 0.0729, "step": 17380 }, { "grad_norm": 0.6919527649879456, "learning_rate": 9.586208291782915e-05, "loss": 0.0763, "step": 17390 }, { "grad_norm": 0.9745420813560486, "learning_rate": 9.58554941241765e-05, "loss": 0.0953, "step": 17400 }, { "grad_norm": 0.7022535800933838, "learning_rate": 9.584890031583596e-05, "loss": 0.0715, "step": 17410 }, { "grad_norm": 0.766466498374939, "learning_rate": 9.584230149352861e-05, "loss": 0.0717, "step": 17420 }, { "grad_norm": 0.9267482161521912, "learning_rate": 9.58356976579761e-05, "loss": 0.081, "step": 17430 }, { "grad_norm": 0.537125289440155, "learning_rate": 9.58290888099006e-05, "loss": 0.0801, "step": 17440 }, { "grad_norm": 0.5466538071632385, "learning_rate": 9.582247495002486e-05, "loss": 0.0779, "step": 17450 }, { "grad_norm": 0.7949823141098022, "learning_rate": 9.581585607907214e-05, "loss": 0.0763, "step": 17460 }, { "grad_norm": 0.6005920171737671, "learning_rate": 9.580923219776628e-05, "loss": 0.0724, "step": 17470 }, { "grad_norm": 0.5089340806007385, "learning_rate": 9.580260330683167e-05, "loss": 0.0757, "step": 17480 }, { "grad_norm": 0.6333505511283875, "learning_rate": 9.579596940699322e-05, "loss": 0.0817, "step": 17490 }, { "grad_norm": 0.8350095152854919, "learning_rate": 9.578933049897643e-05, "loss": 0.0747, "step": 17500 }, { "grad_norm": 0.7558724284172058, "learning_rate": 9.578268658350728e-05, "loss": 0.0755, "step": 17510 }, { "grad_norm": 0.5625174641609192, "learning_rate": 9.577603766131235e-05, "loss": 0.0828, "step": 17520 }, { "grad_norm": 0.7140642404556274, "learning_rate": 9.576938373311878e-05, "loss": 0.0757, "step": 17530 }, { "grad_norm": 0.7920885682106018, "learning_rate": 9.576272479965421e-05, "loss": 0.0837, "step": 17540 }, { "grad_norm": 0.5660022497177124, "learning_rate": 9.575606086164687e-05, "loss": 0.0774, "step": 17550 }, { "grad_norm": 0.505185067653656, "learning_rate": 9.57493919198255e-05, "loss": 0.091, "step": 17560 }, { "grad_norm": 0.5719190239906311, "learning_rate": 9.57427179749194e-05, "loss": 0.0881, "step": 17570 }, { "grad_norm": 0.6765936613082886, "learning_rate": 9.573603902765846e-05, "loss": 0.0654, "step": 17580 }, { "grad_norm": 0.5854878425598145, "learning_rate": 9.572935507877304e-05, "loss": 0.0787, "step": 17590 }, { "grad_norm": 0.7716030478477478, "learning_rate": 9.57226661289941e-05, "loss": 0.0968, "step": 17600 }, { "grad_norm": 0.697904646396637, "learning_rate": 9.571597217905315e-05, "loss": 0.0906, "step": 17610 }, { "grad_norm": 0.6068164110183716, "learning_rate": 9.57092732296822e-05, "loss": 0.0837, "step": 17620 }, { "grad_norm": 0.5099666714668274, "learning_rate": 9.570256928161385e-05, "loss": 0.0668, "step": 17630 }, { "grad_norm": 0.48178523778915405, "learning_rate": 9.569586033558126e-05, "loss": 0.0822, "step": 17640 }, { "grad_norm": 0.6979930996894836, "learning_rate": 9.568914639231807e-05, "loss": 0.0696, "step": 17650 }, { "grad_norm": 0.8073990941047668, "learning_rate": 9.568242745255852e-05, "loss": 0.0765, "step": 17660 }, { "grad_norm": 0.5181501507759094, "learning_rate": 9.567570351703739e-05, "loss": 0.0828, "step": 17670 }, { "grad_norm": 0.6642955541610718, "learning_rate": 9.566897458649001e-05, "loss": 0.08, "step": 17680 }, { "grad_norm": 0.577215850353241, "learning_rate": 9.566224066165221e-05, "loss": 0.082, "step": 17690 }, { "grad_norm": 0.7169397473335266, "learning_rate": 9.565550174326043e-05, "loss": 0.0835, "step": 17700 }, { "grad_norm": 0.7316238284111023, "learning_rate": 9.564875783205162e-05, "loss": 0.0823, "step": 17710 }, { "grad_norm": 0.620303213596344, "learning_rate": 9.564200892876328e-05, "loss": 0.0852, "step": 17720 }, { "grad_norm": 0.8688331842422485, "learning_rate": 9.563525503413348e-05, "loss": 0.0814, "step": 17730 }, { "grad_norm": 0.5453299283981323, "learning_rate": 9.562849614890079e-05, "loss": 0.0869, "step": 17740 }, { "grad_norm": 0.4877909719944, "learning_rate": 9.562173227380436e-05, "loss": 0.091, "step": 17750 }, { "grad_norm": 0.6150606870651245, "learning_rate": 9.561496340958389e-05, "loss": 0.0766, "step": 17760 }, { "grad_norm": 0.60789954662323, "learning_rate": 9.560818955697959e-05, "loss": 0.0859, "step": 17770 }, { "grad_norm": 1.037854552268982, "learning_rate": 9.560141071673228e-05, "loss": 0.0816, "step": 17780 }, { "grad_norm": 0.4974856376647949, "learning_rate": 9.559462688958323e-05, "loss": 0.0519, "step": 17790 }, { "grad_norm": 0.6730712056159973, "learning_rate": 9.558783807627434e-05, "loss": 0.0642, "step": 17800 }, { "grad_norm": 0.7191355228424072, "learning_rate": 9.558104427754801e-05, "loss": 0.1035, "step": 17810 }, { "grad_norm": 0.7579252123832703, "learning_rate": 9.557424549414722e-05, "loss": 0.0977, "step": 17820 }, { "grad_norm": 0.7531139850616455, "learning_rate": 9.556744172681546e-05, "loss": 0.0835, "step": 17830 }, { "grad_norm": 1.3199944496154785, "learning_rate": 9.556063297629677e-05, "loss": 0.0849, "step": 17840 }, { "grad_norm": 0.6219087839126587, "learning_rate": 9.555381924333578e-05, "loss": 0.0667, "step": 17850 }, { "grad_norm": 0.6003133654594421, "learning_rate": 9.554700052867758e-05, "loss": 0.0753, "step": 17860 }, { "grad_norm": 0.6185053586959839, "learning_rate": 9.554017683306789e-05, "loss": 0.0753, "step": 17870 }, { "grad_norm": 0.6464651823043823, "learning_rate": 9.553334815725294e-05, "loss": 0.0752, "step": 17880 }, { "grad_norm": 0.5755116939544678, "learning_rate": 9.552651450197949e-05, "loss": 0.0773, "step": 17890 }, { "grad_norm": 0.749776303768158, "learning_rate": 9.551967586799486e-05, "loss": 0.0765, "step": 17900 }, { "grad_norm": 0.6839806437492371, "learning_rate": 9.551283225604692e-05, "loss": 0.0817, "step": 17910 }, { "grad_norm": 0.6761952638626099, "learning_rate": 9.550598366688406e-05, "loss": 0.0827, "step": 17920 }, { "grad_norm": 0.7028699517250061, "learning_rate": 9.549913010125526e-05, "loss": 0.0803, "step": 17930 }, { "grad_norm": 0.5326547026634216, "learning_rate": 9.549227155990999e-05, "loss": 0.0798, "step": 17940 }, { "grad_norm": 0.6168500185012817, "learning_rate": 9.548540804359828e-05, "loss": 0.0814, "step": 17950 }, { "grad_norm": 0.6133250594139099, "learning_rate": 9.547853955307077e-05, "loss": 0.0808, "step": 17960 }, { "grad_norm": 0.5494985580444336, "learning_rate": 9.547166608907853e-05, "loss": 0.0743, "step": 17970 }, { "grad_norm": 0.5061123371124268, "learning_rate": 9.546478765237326e-05, "loss": 0.0905, "step": 17980 }, { "grad_norm": 0.6344610452651978, "learning_rate": 9.545790424370715e-05, "loss": 0.0797, "step": 17990 }, { "grad_norm": 0.5010362267494202, "learning_rate": 9.5451015863833e-05, "loss": 0.0847, "step": 18000 }, { "grad_norm": 1.044573426246643, "learning_rate": 9.544412251350408e-05, "loss": 0.0814, "step": 18010 }, { "grad_norm": 0.7170330286026001, "learning_rate": 9.543722419347422e-05, "loss": 0.0698, "step": 18020 }, { "grad_norm": 0.8107560276985168, "learning_rate": 9.543032090449788e-05, "loss": 0.0837, "step": 18030 }, { "grad_norm": 0.5840674638748169, "learning_rate": 9.542341264732992e-05, "loss": 0.0817, "step": 18040 }, { "grad_norm": 0.6397941708564758, "learning_rate": 9.541649942272585e-05, "loss": 0.091, "step": 18050 }, { "grad_norm": 0.5581071376800537, "learning_rate": 9.54095812314417e-05, "loss": 0.0652, "step": 18060 }, { "grad_norm": 0.5659245848655701, "learning_rate": 9.540265807423401e-05, "loss": 0.0716, "step": 18070 }, { "grad_norm": 0.567302942276001, "learning_rate": 9.53957299518599e-05, "loss": 0.0706, "step": 18080 }, { "grad_norm": 0.43655678629875183, "learning_rate": 9.5388796865077e-05, "loss": 0.0754, "step": 18090 }, { "grad_norm": 0.7666684985160828, "learning_rate": 9.538185881464353e-05, "loss": 0.0818, "step": 18100 }, { "grad_norm": 0.5883358120918274, "learning_rate": 9.537491580131821e-05, "loss": 0.0711, "step": 18110 }, { "grad_norm": 0.5522922277450562, "learning_rate": 9.53679678258603e-05, "loss": 0.0659, "step": 18120 }, { "grad_norm": 0.7438483834266663, "learning_rate": 9.536101488902966e-05, "loss": 0.0773, "step": 18130 }, { "grad_norm": 0.8441523313522339, "learning_rate": 9.535405699158663e-05, "loss": 0.0708, "step": 18140 }, { "grad_norm": 0.5300365686416626, "learning_rate": 9.53470941342921e-05, "loss": 0.0596, "step": 18150 }, { "grad_norm": 0.47717419266700745, "learning_rate": 9.534012631790756e-05, "loss": 0.0661, "step": 18160 }, { "grad_norm": 1.1807502508163452, "learning_rate": 9.533315354319494e-05, "loss": 0.0846, "step": 18170 }, { "grad_norm": 0.4733075201511383, "learning_rate": 9.532617581091682e-05, "loss": 0.0813, "step": 18180 }, { "grad_norm": 0.6541879177093506, "learning_rate": 9.531919312183629e-05, "loss": 0.0759, "step": 18190 }, { "grad_norm": 1.1740798950195312, "learning_rate": 9.531220547671688e-05, "loss": 0.0723, "step": 18200 }, { "grad_norm": 0.6075194478034973, "learning_rate": 9.530521287632285e-05, "loss": 0.0889, "step": 18210 }, { "grad_norm": 0.7506512999534607, "learning_rate": 9.529821532141884e-05, "loss": 0.0824, "step": 18220 }, { "grad_norm": 0.4564567804336548, "learning_rate": 9.52912128127701e-05, "loss": 0.0723, "step": 18230 }, { "grad_norm": 0.6262651681900024, "learning_rate": 9.528420535114244e-05, "loss": 0.0718, "step": 18240 }, { "grad_norm": 0.835283637046814, "learning_rate": 9.527719293730215e-05, "loss": 0.0846, "step": 18250 }, { "grad_norm": 0.43515247106552124, "learning_rate": 9.527017557201611e-05, "loss": 0.0798, "step": 18260 }, { "grad_norm": 0.6187634468078613, "learning_rate": 9.526315325605176e-05, "loss": 0.0915, "step": 18270 }, { "grad_norm": 0.6420512199401855, "learning_rate": 9.525612599017699e-05, "loss": 0.0661, "step": 18280 }, { "grad_norm": 0.6314780712127686, "learning_rate": 9.524909377516033e-05, "loss": 0.0839, "step": 18290 }, { "grad_norm": 0.747739315032959, "learning_rate": 9.524205661177081e-05, "loss": 0.076, "step": 18300 }, { "grad_norm": 2.194931983947754, "learning_rate": 9.523501450077801e-05, "loss": 0.0828, "step": 18310 }, { "grad_norm": 0.6634487509727478, "learning_rate": 9.522796744295202e-05, "loss": 0.0886, "step": 18320 }, { "grad_norm": 0.8746833205223083, "learning_rate": 9.522091543906352e-05, "loss": 0.0806, "step": 18330 }, { "grad_norm": 0.9650812745094299, "learning_rate": 9.521385848988369e-05, "loss": 0.0788, "step": 18340 }, { "grad_norm": 0.5440700054168701, "learning_rate": 9.520679659618428e-05, "loss": 0.0658, "step": 18350 }, { "grad_norm": 1.369491457939148, "learning_rate": 9.519972975873754e-05, "loss": 0.0905, "step": 18360 }, { "grad_norm": 0.6395269632339478, "learning_rate": 9.519265797831633e-05, "loss": 0.0776, "step": 18370 }, { "grad_norm": 0.5921618938446045, "learning_rate": 9.518558125569399e-05, "loss": 0.0701, "step": 18380 }, { "grad_norm": 0.6684671640396118, "learning_rate": 9.517849959164442e-05, "loss": 0.0824, "step": 18390 }, { "grad_norm": 0.6309144496917725, "learning_rate": 9.517141298694205e-05, "loss": 0.0876, "step": 18400 }, { "grad_norm": 0.7020927667617798, "learning_rate": 9.516432144236188e-05, "loss": 0.0937, "step": 18410 }, { "grad_norm": 0.5651669502258301, "learning_rate": 9.515722495867941e-05, "loss": 0.0869, "step": 18420 }, { "grad_norm": 0.8768492937088013, "learning_rate": 9.515012353667072e-05, "loss": 0.0753, "step": 18430 }, { "grad_norm": 0.6955447196960449, "learning_rate": 9.51430171771124e-05, "loss": 0.0745, "step": 18440 }, { "grad_norm": 0.7817549705505371, "learning_rate": 9.513590588078159e-05, "loss": 0.074, "step": 18450 }, { "grad_norm": 0.6807875633239746, "learning_rate": 9.512878964845597e-05, "loss": 0.077, "step": 18460 }, { "grad_norm": 0.6252421140670776, "learning_rate": 9.512166848091377e-05, "loss": 0.0904, "step": 18470 }, { "grad_norm": 0.5429680347442627, "learning_rate": 9.511454237893376e-05, "loss": 0.0873, "step": 18480 }, { "grad_norm": 0.6821470260620117, "learning_rate": 9.51074113432952e-05, "loss": 0.0682, "step": 18490 }, { "grad_norm": 0.6004303693771362, "learning_rate": 9.510027537477797e-05, "loss": 0.092, "step": 18500 }, { "grad_norm": 0.5330809354782104, "learning_rate": 9.509313447416242e-05, "loss": 0.0827, "step": 18510 }, { "grad_norm": 0.5853548645973206, "learning_rate": 9.508598864222949e-05, "loss": 0.0815, "step": 18520 }, { "grad_norm": 0.6269851326942444, "learning_rate": 9.507883787976062e-05, "loss": 0.0722, "step": 18530 }, { "grad_norm": 0.7212691903114319, "learning_rate": 9.507168218753781e-05, "loss": 0.0668, "step": 18540 }, { "grad_norm": 0.8989888429641724, "learning_rate": 9.506452156634362e-05, "loss": 0.0869, "step": 18550 }, { "grad_norm": 0.7227135300636292, "learning_rate": 9.505735601696109e-05, "loss": 0.0889, "step": 18560 }, { "grad_norm": 0.5979358553886414, "learning_rate": 9.505018554017385e-05, "loss": 0.0635, "step": 18570 }, { "grad_norm": 0.7372714281082153, "learning_rate": 9.504301013676604e-05, "loss": 0.0673, "step": 18580 }, { "grad_norm": 0.6942089200019836, "learning_rate": 9.503582980752238e-05, "loss": 0.0736, "step": 18590 }, { "grad_norm": 0.7014951109886169, "learning_rate": 9.502864455322809e-05, "loss": 0.092, "step": 18600 }, { "grad_norm": 0.5249555706977844, "learning_rate": 9.502145437466891e-05, "loss": 0.0812, "step": 18610 }, { "grad_norm": 0.39330583810806274, "learning_rate": 9.501425927263116e-05, "loss": 0.0894, "step": 18620 }, { "grad_norm": 0.8087279200553894, "learning_rate": 9.500705924790172e-05, "loss": 0.1016, "step": 18630 }, { "grad_norm": 0.7905981540679932, "learning_rate": 9.499985430126794e-05, "loss": 0.0783, "step": 18640 }, { "grad_norm": 0.6402820944786072, "learning_rate": 9.499264443351775e-05, "loss": 0.07, "step": 18650 }, { "grad_norm": 0.7209077477455139, "learning_rate": 9.498542964543961e-05, "loss": 0.0801, "step": 18660 }, { "grad_norm": 0.6563429236412048, "learning_rate": 9.497820993782252e-05, "loss": 0.089, "step": 18670 }, { "grad_norm": 0.5585435032844543, "learning_rate": 9.497098531145601e-05, "loss": 0.0747, "step": 18680 }, { "grad_norm": 0.8628050684928894, "learning_rate": 9.496375576713017e-05, "loss": 0.0823, "step": 18690 }, { "grad_norm": 0.5201553106307983, "learning_rate": 9.49565213056356e-05, "loss": 0.0702, "step": 18700 }, { "grad_norm": 0.44587787985801697, "learning_rate": 9.494928192776342e-05, "loss": 0.0728, "step": 18710 }, { "grad_norm": 0.732751727104187, "learning_rate": 9.494203763430538e-05, "loss": 0.0734, "step": 18720 }, { "grad_norm": 0.506830096244812, "learning_rate": 9.493478842605366e-05, "loss": 0.1016, "step": 18730 }, { "grad_norm": 0.7801128625869751, "learning_rate": 9.492753430380105e-05, "loss": 0.0716, "step": 18740 }, { "grad_norm": 1.2896815538406372, "learning_rate": 9.492027526834083e-05, "loss": 0.0644, "step": 18750 }, { "grad_norm": 0.7434746026992798, "learning_rate": 9.491301132046684e-05, "loss": 0.077, "step": 18760 }, { "grad_norm": 0.6537661552429199, "learning_rate": 9.490574246097345e-05, "loss": 0.094, "step": 18770 }, { "grad_norm": 0.776452362537384, "learning_rate": 9.48984686906556e-05, "loss": 0.0777, "step": 18780 }, { "grad_norm": 0.6402115821838379, "learning_rate": 9.489119001030871e-05, "loss": 0.0665, "step": 18790 }, { "grad_norm": 0.5339945554733276, "learning_rate": 9.488390642072878e-05, "loss": 0.0878, "step": 18800 }, { "grad_norm": 0.6355206966400146, "learning_rate": 9.48766179227123e-05, "loss": 0.0849, "step": 18810 }, { "grad_norm": 1.3796287775039673, "learning_rate": 9.486932451705636e-05, "loss": 0.1036, "step": 18820 }, { "grad_norm": 0.4877069294452667, "learning_rate": 9.486202620455857e-05, "loss": 0.0804, "step": 18830 }, { "grad_norm": 0.6883272528648376, "learning_rate": 9.485472298601704e-05, "loss": 0.0773, "step": 18840 }, { "grad_norm": 0.757693350315094, "learning_rate": 9.484741486223043e-05, "loss": 0.0849, "step": 18850 }, { "grad_norm": 0.5753677487373352, "learning_rate": 9.484010183399797e-05, "loss": 0.0711, "step": 18860 }, { "grad_norm": 0.5443217754364014, "learning_rate": 9.483278390211938e-05, "loss": 0.0737, "step": 18870 }, { "grad_norm": 0.4883957505226135, "learning_rate": 9.482546106739496e-05, "loss": 0.058, "step": 18880 }, { "grad_norm": 0.5927470922470093, "learning_rate": 9.48181333306255e-05, "loss": 0.0669, "step": 18890 }, { "grad_norm": 0.7575588822364807, "learning_rate": 9.481080069261237e-05, "loss": 0.0862, "step": 18900 }, { "grad_norm": 0.6970444917678833, "learning_rate": 9.480346315415745e-05, "loss": 0.0716, "step": 18910 }, { "grad_norm": 0.9205489754676819, "learning_rate": 9.479612071606314e-05, "loss": 0.0761, "step": 18920 }, { "grad_norm": 0.451866090297699, "learning_rate": 9.478877337913244e-05, "loss": 0.0772, "step": 18930 }, { "grad_norm": 0.6014230251312256, "learning_rate": 9.478142114416881e-05, "loss": 0.0679, "step": 18940 }, { "grad_norm": 0.7106685042381287, "learning_rate": 9.47740640119763e-05, "loss": 0.0679, "step": 18950 }, { "grad_norm": 0.5910253524780273, "learning_rate": 9.476670198335947e-05, "loss": 0.0771, "step": 18960 }, { "grad_norm": 0.661345362663269, "learning_rate": 9.47593350591234e-05, "loss": 0.0739, "step": 18970 }, { "grad_norm": 0.4759232699871063, "learning_rate": 9.475196324007376e-05, "loss": 0.0728, "step": 18980 }, { "grad_norm": 0.7268799543380737, "learning_rate": 9.474458652701669e-05, "loss": 0.0769, "step": 18990 }, { "grad_norm": 2.7869324684143066, "learning_rate": 9.473720492075892e-05, "loss": 0.096, "step": 19000 }, { "grad_norm": 0.9572181105613708, "learning_rate": 9.472981842210768e-05, "loss": 0.084, "step": 19010 }, { "grad_norm": 1.346091389656067, "learning_rate": 9.472242703187074e-05, "loss": 0.0731, "step": 19020 }, { "grad_norm": 0.5666488409042358, "learning_rate": 9.471503075085643e-05, "loss": 0.0724, "step": 19030 }, { "grad_norm": 1.0911468267440796, "learning_rate": 9.470762957987359e-05, "loss": 0.0843, "step": 19040 }, { "grad_norm": 0.42398419976234436, "learning_rate": 9.470022351973158e-05, "loss": 0.0909, "step": 19050 }, { "grad_norm": 0.5329900979995728, "learning_rate": 9.469281257124034e-05, "loss": 0.0722, "step": 19060 }, { "grad_norm": 0.8803191184997559, "learning_rate": 9.46853967352103e-05, "loss": 0.0743, "step": 19070 }, { "grad_norm": 0.7061172127723694, "learning_rate": 9.467797601245246e-05, "loss": 0.0732, "step": 19080 }, { "grad_norm": 1.0629857778549194, "learning_rate": 9.467055040377834e-05, "loss": 0.0819, "step": 19090 }, { "grad_norm": 0.456546813249588, "learning_rate": 9.466311990999999e-05, "loss": 0.0787, "step": 19100 }, { "grad_norm": 0.9606502652168274, "learning_rate": 9.465568453193e-05, "loss": 0.0739, "step": 19110 }, { "grad_norm": 0.5565006136894226, "learning_rate": 9.464824427038148e-05, "loss": 0.0647, "step": 19120 }, { "grad_norm": 0.6263604164123535, "learning_rate": 9.46407991261681e-05, "loss": 0.0918, "step": 19130 }, { "grad_norm": 0.6010046005249023, "learning_rate": 9.463334910010404e-05, "loss": 0.0609, "step": 19140 }, { "grad_norm": 0.4884326457977295, "learning_rate": 9.462589419300403e-05, "loss": 0.0575, "step": 19150 }, { "grad_norm": 1.0287669897079468, "learning_rate": 9.461843440568333e-05, "loss": 0.0838, "step": 19160 }, { "grad_norm": 0.45480605959892273, "learning_rate": 9.461096973895773e-05, "loss": 0.0706, "step": 19170 }, { "grad_norm": 0.40564635396003723, "learning_rate": 9.460350019364355e-05, "loss": 0.067, "step": 19180 }, { "grad_norm": 0.6491615772247314, "learning_rate": 9.459602577055764e-05, "loss": 0.0935, "step": 19190 }, { "grad_norm": 0.4966188371181488, "learning_rate": 9.45885464705174e-05, "loss": 0.0681, "step": 19200 }, { "grad_norm": 0.5327675342559814, "learning_rate": 9.458106229434076e-05, "loss": 0.0834, "step": 19210 }, { "grad_norm": 0.6317367553710938, "learning_rate": 9.457357324284617e-05, "loss": 0.0819, "step": 19220 }, { "grad_norm": 0.6310253739356995, "learning_rate": 9.456607931685262e-05, "loss": 0.0746, "step": 19230 }, { "grad_norm": 0.5961248278617859, "learning_rate": 9.455858051717965e-05, "loss": 0.0716, "step": 19240 }, { "grad_norm": 0.5195487141609192, "learning_rate": 9.45510768446473e-05, "loss": 0.061, "step": 19250 }, { "grad_norm": 0.69919353723526, "learning_rate": 9.454356830007618e-05, "loss": 0.0646, "step": 19260 }, { "grad_norm": 0.6645986437797546, "learning_rate": 9.45360548842874e-05, "loss": 0.0637, "step": 19270 }, { "grad_norm": 0.571408748626709, "learning_rate": 9.452853659810261e-05, "loss": 0.0656, "step": 19280 }, { "grad_norm": 1.283707618713379, "learning_rate": 9.452101344234401e-05, "loss": 0.084, "step": 19290 }, { "grad_norm": 0.44467243552207947, "learning_rate": 9.451348541783431e-05, "loss": 0.0699, "step": 19300 }, { "grad_norm": 0.7892866730690002, "learning_rate": 9.450595252539678e-05, "loss": 0.0673, "step": 19310 }, { "grad_norm": 0.7981111407279968, "learning_rate": 9.449841476585518e-05, "loss": 0.0813, "step": 19320 }, { "grad_norm": 0.8367054462432861, "learning_rate": 9.449087214003384e-05, "loss": 0.0935, "step": 19330 }, { "grad_norm": 0.8195981383323669, "learning_rate": 9.448332464875765e-05, "loss": 0.0995, "step": 19340 }, { "grad_norm": 0.4666959345340729, "learning_rate": 9.447577229285192e-05, "loss": 0.0955, "step": 19350 }, { "grad_norm": 0.5781537294387817, "learning_rate": 9.446821507314261e-05, "loss": 0.0684, "step": 19360 }, { "grad_norm": 0.8951953053474426, "learning_rate": 9.446065299045617e-05, "loss": 0.0813, "step": 19370 }, { "grad_norm": 0.4861597418785095, "learning_rate": 9.445308604561955e-05, "loss": 0.0842, "step": 19380 }, { "grad_norm": 0.7744137644767761, "learning_rate": 9.444551423946028e-05, "loss": 0.0695, "step": 19390 }, { "grad_norm": 0.7491458058357239, "learning_rate": 9.443793757280638e-05, "loss": 0.0818, "step": 19400 }, { "grad_norm": 0.6744327545166016, "learning_rate": 9.443035604648646e-05, "loss": 0.0799, "step": 19410 }, { "grad_norm": 0.5715939402580261, "learning_rate": 9.44227696613296e-05, "loss": 0.0689, "step": 19420 }, { "grad_norm": 0.6133467555046082, "learning_rate": 9.441517841816542e-05, "loss": 0.0942, "step": 19430 }, { "grad_norm": 0.5786106586456299, "learning_rate": 9.440758231782413e-05, "loss": 0.069, "step": 19440 }, { "grad_norm": 0.7720746994018555, "learning_rate": 9.439998136113639e-05, "loss": 0.0859, "step": 19450 }, { "grad_norm": 0.6088905334472656, "learning_rate": 9.439237554893344e-05, "loss": 0.0815, "step": 19460 }, { "grad_norm": 0.8006205558776855, "learning_rate": 9.438476488204705e-05, "loss": 0.0804, "step": 19470 }, { "grad_norm": 0.656761109828949, "learning_rate": 9.43771493613095e-05, "loss": 0.0791, "step": 19480 }, { "grad_norm": 0.5129543542861938, "learning_rate": 9.436952898755362e-05, "loss": 0.0782, "step": 19490 }, { "grad_norm": 0.6467081904411316, "learning_rate": 9.436190376161276e-05, "loss": 0.0737, "step": 19500 }, { "grad_norm": 0.7527676820755005, "learning_rate": 9.43542736843208e-05, "loss": 0.1067, "step": 19510 }, { "grad_norm": 0.7923919558525085, "learning_rate": 9.434663875651216e-05, "loss": 0.0671, "step": 19520 }, { "grad_norm": 0.6907520890235901, "learning_rate": 9.433899897902177e-05, "loss": 0.0752, "step": 19530 }, { "grad_norm": 0.6254045367240906, "learning_rate": 9.433135435268511e-05, "loss": 0.0617, "step": 19540 }, { "grad_norm": 0.44734668731689453, "learning_rate": 9.432370487833819e-05, "loss": 0.0727, "step": 19550 }, { "grad_norm": 0.36812785267829895, "learning_rate": 9.431605055681756e-05, "loss": 0.0668, "step": 19560 }, { "grad_norm": 0.5322237610816956, "learning_rate": 9.430839138896026e-05, "loss": 0.0752, "step": 19570 }, { "grad_norm": 0.8367741107940674, "learning_rate": 9.43007273756039e-05, "loss": 0.0792, "step": 19580 }, { "grad_norm": 0.5068053007125854, "learning_rate": 9.429305851758658e-05, "loss": 0.0758, "step": 19590 }, { "grad_norm": 0.575678288936615, "learning_rate": 9.428538481574699e-05, "loss": 0.0705, "step": 19600 }, { "grad_norm": 0.6874651908874512, "learning_rate": 9.42777062709243e-05, "loss": 0.076, "step": 19610 }, { "grad_norm": 0.47348281741142273, "learning_rate": 9.427002288395821e-05, "loss": 0.0657, "step": 19620 }, { "grad_norm": 0.6239636540412903, "learning_rate": 9.426233465568898e-05, "loss": 0.0688, "step": 19630 }, { "grad_norm": 0.4214997887611389, "learning_rate": 9.42546415869574e-05, "loss": 0.0684, "step": 19640 }, { "grad_norm": 0.5701404213905334, "learning_rate": 9.424694367860473e-05, "loss": 0.0568, "step": 19650 }, { "grad_norm": 0.9219241142272949, "learning_rate": 9.423924093147284e-05, "loss": 0.0596, "step": 19660 }, { "grad_norm": 0.5609411597251892, "learning_rate": 9.423153334640407e-05, "loss": 0.0767, "step": 19670 }, { "grad_norm": 0.4715821146965027, "learning_rate": 9.42238209242413e-05, "loss": 0.0557, "step": 19680 }, { "grad_norm": 0.4635026752948761, "learning_rate": 9.421610366582798e-05, "loss": 0.0803, "step": 19690 }, { "grad_norm": 0.6611576676368713, "learning_rate": 9.420838157200803e-05, "loss": 0.0704, "step": 19700 }, { "grad_norm": 0.5918153524398804, "learning_rate": 9.420065464362594e-05, "loss": 0.0878, "step": 19710 }, { "grad_norm": 0.904037594795227, "learning_rate": 9.419292288152673e-05, "loss": 0.0747, "step": 19720 }, { "grad_norm": 0.4759719669818878, "learning_rate": 9.418518628655588e-05, "loss": 0.0706, "step": 19730 }, { "grad_norm": 0.40850257873535156, "learning_rate": 9.417744485955951e-05, "loss": 0.065, "step": 19740 }, { "grad_norm": 0.6249006986618042, "learning_rate": 9.41696986013842e-05, "loss": 0.0696, "step": 19750 }, { "grad_norm": 0.5116675496101379, "learning_rate": 9.416194751287705e-05, "loss": 0.0764, "step": 19760 }, { "grad_norm": 0.5606229305267334, "learning_rate": 9.415419159488572e-05, "loss": 0.0582, "step": 19770 }, { "grad_norm": 0.962415337562561, "learning_rate": 9.414643084825837e-05, "loss": 0.084, "step": 19780 }, { "grad_norm": 0.631696879863739, "learning_rate": 9.413866527384372e-05, "loss": 0.0736, "step": 19790 }, { "grad_norm": 0.684045135974884, "learning_rate": 9.4130894872491e-05, "loss": 0.0728, "step": 19800 }, { "grad_norm": 0.8575758934020996, "learning_rate": 9.412311964504998e-05, "loss": 0.0822, "step": 19810 }, { "grad_norm": 0.5079210996627808, "learning_rate": 9.411533959237091e-05, "loss": 0.0872, "step": 19820 }, { "grad_norm": 0.42947250604629517, "learning_rate": 9.410755471530464e-05, "loss": 0.0558, "step": 19830 }, { "grad_norm": 0.7782686352729797, "learning_rate": 9.40997650147025e-05, "loss": 0.0807, "step": 19840 }, { "grad_norm": 0.6048728823661804, "learning_rate": 9.409197049141637e-05, "loss": 0.0842, "step": 19850 }, { "grad_norm": 0.5531384944915771, "learning_rate": 9.408417114629863e-05, "loss": 0.0771, "step": 19860 }, { "grad_norm": 0.44086959958076477, "learning_rate": 9.40763669802022e-05, "loss": 0.0647, "step": 19870 }, { "grad_norm": 0.6868392825126648, "learning_rate": 9.406855799398056e-05, "loss": 0.0866, "step": 19880 }, { "grad_norm": 0.5410473942756653, "learning_rate": 9.406074418848767e-05, "loss": 0.0737, "step": 19890 }, { "grad_norm": 0.5795853137969971, "learning_rate": 9.405292556457805e-05, "loss": 0.086, "step": 19900 }, { "grad_norm": 0.5196983814239502, "learning_rate": 9.404510212310671e-05, "loss": 0.0744, "step": 19910 }, { "grad_norm": 0.4335140883922577, "learning_rate": 9.403727386492924e-05, "loss": 0.0644, "step": 19920 }, { "grad_norm": 0.3633750379085541, "learning_rate": 9.40294407909017e-05, "loss": 0.0646, "step": 19930 }, { "grad_norm": 0.7141996026039124, "learning_rate": 9.40216029018807e-05, "loss": 0.0765, "step": 19940 }, { "grad_norm": 0.6215440034866333, "learning_rate": 9.401376019872338e-05, "loss": 0.0685, "step": 19950 }, { "grad_norm": 0.535444974899292, "learning_rate": 9.400591268228746e-05, "loss": 0.0737, "step": 19960 }, { "grad_norm": 0.6179869771003723, "learning_rate": 9.399806035343106e-05, "loss": 0.071, "step": 19970 }, { "grad_norm": 0.5015203952789307, "learning_rate": 9.399020321301294e-05, "loss": 0.0644, "step": 19980 }, { "grad_norm": 0.5322580337524414, "learning_rate": 9.398234126189234e-05, "loss": 0.0838, "step": 19990 }, { "grad_norm": 1.18955659866333, "learning_rate": 9.397447450092902e-05, "loss": 0.0794, "step": 20000 }, { "grad_norm": 0.4671486020088196, "learning_rate": 9.39666029309833e-05, "loss": 0.0686, "step": 20010 }, { "grad_norm": 1.3183058500289917, "learning_rate": 9.395872655291596e-05, "loss": 0.0713, "step": 20020 }, { "grad_norm": 0.6121578812599182, "learning_rate": 9.395084536758838e-05, "loss": 0.0823, "step": 20030 }, { "grad_norm": 0.7549139857292175, "learning_rate": 9.394295937586243e-05, "loss": 0.1067, "step": 20040 }, { "grad_norm": 0.5350010395050049, "learning_rate": 9.393506857860052e-05, "loss": 0.0761, "step": 20050 }, { "grad_norm": 0.6291171312332153, "learning_rate": 9.392717297666555e-05, "loss": 0.0597, "step": 20060 }, { "grad_norm": 0.7362361550331116, "learning_rate": 9.391927257092101e-05, "loss": 0.0797, "step": 20070 }, { "grad_norm": 0.5058185458183289, "learning_rate": 9.391136736223085e-05, "loss": 0.0699, "step": 20080 }, { "grad_norm": 0.6190244555473328, "learning_rate": 9.390345735145956e-05, "loss": 0.0724, "step": 20090 }, { "grad_norm": 0.8898542523384094, "learning_rate": 9.389554253947219e-05, "loss": 0.0761, "step": 20100 }, { "grad_norm": 0.59061598777771, "learning_rate": 9.388762292713428e-05, "loss": 0.0747, "step": 20110 }, { "grad_norm": 0.7893803715705872, "learning_rate": 9.38796985153119e-05, "loss": 0.0737, "step": 20120 }, { "grad_norm": 0.6462571620941162, "learning_rate": 9.387176930487169e-05, "loss": 0.0775, "step": 20130 }, { "grad_norm": 1.131975769996643, "learning_rate": 9.386383529668072e-05, "loss": 0.0761, "step": 20140 }, { "grad_norm": 0.7881686687469482, "learning_rate": 9.385589649160669e-05, "loss": 0.0751, "step": 20150 }, { "grad_norm": 0.6547479629516602, "learning_rate": 9.384795289051775e-05, "loss": 0.0834, "step": 20160 }, { "grad_norm": 0.6665324568748474, "learning_rate": 9.384000449428261e-05, "loss": 0.0807, "step": 20170 }, { "grad_norm": 0.776921808719635, "learning_rate": 9.383205130377048e-05, "loss": 0.0766, "step": 20180 }, { "grad_norm": 0.39616653323173523, "learning_rate": 9.382409331985114e-05, "loss": 0.0683, "step": 20190 }, { "grad_norm": 0.8157426118850708, "learning_rate": 9.381613054339482e-05, "loss": 0.0734, "step": 20200 }, { "grad_norm": 0.7011055946350098, "learning_rate": 9.380816297527235e-05, "loss": 0.0833, "step": 20210 }, { "grad_norm": 1.0632249116897583, "learning_rate": 9.380019061635506e-05, "loss": 0.0873, "step": 20220 }, { "grad_norm": 0.48175162076950073, "learning_rate": 9.379221346751474e-05, "loss": 0.0948, "step": 20230 }, { "grad_norm": 0.9022938013076782, "learning_rate": 9.378423152962382e-05, "loss": 0.0734, "step": 20240 }, { "grad_norm": 0.6373438835144043, "learning_rate": 9.377624480355517e-05, "loss": 0.0755, "step": 20250 }, { "grad_norm": 0.5668415427207947, "learning_rate": 9.376825329018219e-05, "loss": 0.0714, "step": 20260 }, { "grad_norm": 0.8299984335899353, "learning_rate": 9.376025699037884e-05, "loss": 0.0811, "step": 20270 }, { "grad_norm": 0.6291079521179199, "learning_rate": 9.37522559050196e-05, "loss": 0.0709, "step": 20280 }, { "grad_norm": 0.6116538643836975, "learning_rate": 9.37442500349794e-05, "loss": 0.0711, "step": 20290 }, { "grad_norm": 0.5274567604064941, "learning_rate": 9.373623938113381e-05, "loss": 0.0725, "step": 20300 }, { "grad_norm": 0.5823306441307068, "learning_rate": 9.372822394435883e-05, "loss": 0.084, "step": 20310 }, { "grad_norm": 0.5671553015708923, "learning_rate": 9.372020372553102e-05, "loss": 0.0755, "step": 20320 }, { "grad_norm": 0.5772253274917603, "learning_rate": 9.371217872552746e-05, "loss": 0.0758, "step": 20330 }, { "grad_norm": 0.4981764853000641, "learning_rate": 9.370414894522576e-05, "loss": 0.0715, "step": 20340 }, { "grad_norm": 0.649815022945404, "learning_rate": 9.369611438550406e-05, "loss": 0.0763, "step": 20350 }, { "grad_norm": 0.494174599647522, "learning_rate": 9.368807504724095e-05, "loss": 0.0747, "step": 20360 }, { "grad_norm": 1.0196694135665894, "learning_rate": 9.368003093131565e-05, "loss": 0.0855, "step": 20370 }, { "grad_norm": 0.6751478910446167, "learning_rate": 9.367198203860785e-05, "loss": 0.0934, "step": 20380 }, { "grad_norm": 0.47204768657684326, "learning_rate": 9.366392836999774e-05, "loss": 0.065, "step": 20390 }, { "grad_norm": 0.6154266595840454, "learning_rate": 9.365586992636607e-05, "loss": 0.0766, "step": 20400 }, { "grad_norm": 0.5457310080528259, "learning_rate": 9.364780670859412e-05, "loss": 0.0657, "step": 20410 }, { "grad_norm": 0.4791911542415619, "learning_rate": 9.363973871756364e-05, "loss": 0.0728, "step": 20420 }, { "grad_norm": 0.8376659750938416, "learning_rate": 9.363166595415696e-05, "loss": 0.0714, "step": 20430 }, { "grad_norm": 0.9137346744537354, "learning_rate": 9.362358841925686e-05, "loss": 0.0722, "step": 20440 }, { "grad_norm": 0.7111172676086426, "learning_rate": 9.361550611374674e-05, "loss": 0.0827, "step": 20450 }, { "grad_norm": 0.5347897410392761, "learning_rate": 9.360741903851043e-05, "loss": 0.1073, "step": 20460 }, { "grad_norm": 0.6960870027542114, "learning_rate": 9.359932719443236e-05, "loss": 0.0981, "step": 20470 }, { "grad_norm": 0.40164914727211, "learning_rate": 9.35912305823974e-05, "loss": 0.0785, "step": 20480 }, { "grad_norm": 0.5624121427536011, "learning_rate": 9.358312920329101e-05, "loss": 0.0765, "step": 20490 }, { "grad_norm": 0.5357300639152527, "learning_rate": 9.357502305799914e-05, "loss": 0.0606, "step": 20500 }, { "grad_norm": 0.4984969198703766, "learning_rate": 9.356691214740824e-05, "loss": 0.0589, "step": 20510 }, { "grad_norm": 0.5170319080352783, "learning_rate": 9.355879647240535e-05, "loss": 0.0664, "step": 20520 }, { "grad_norm": 0.6428793668746948, "learning_rate": 9.355067603387798e-05, "loss": 0.0727, "step": 20530 }, { "grad_norm": 0.6144243478775024, "learning_rate": 9.354255083271412e-05, "loss": 0.0826, "step": 20540 }, { "grad_norm": 0.6209420561790466, "learning_rate": 9.353442086980239e-05, "loss": 0.0735, "step": 20550 }, { "grad_norm": 0.48147982358932495, "learning_rate": 9.352628614603185e-05, "loss": 0.0776, "step": 20560 }, { "grad_norm": 0.6584048271179199, "learning_rate": 9.351814666229209e-05, "loss": 0.0755, "step": 20570 }, { "grad_norm": 0.9115719199180603, "learning_rate": 9.351000241947324e-05, "loss": 0.069, "step": 20580 }, { "grad_norm": 0.6566423177719116, "learning_rate": 9.350185341846594e-05, "loss": 0.0702, "step": 20590 }, { "grad_norm": 0.38663777709007263, "learning_rate": 9.349369966016134e-05, "loss": 0.0686, "step": 20600 }, { "grad_norm": 0.6900171041488647, "learning_rate": 9.348554114545117e-05, "loss": 0.1048, "step": 20610 }, { "grad_norm": 0.5591286420822144, "learning_rate": 9.347737787522758e-05, "loss": 0.081, "step": 20620 }, { "grad_norm": 0.4795606732368469, "learning_rate": 9.346920985038332e-05, "loss": 0.0783, "step": 20630 }, { "grad_norm": 0.7608350515365601, "learning_rate": 9.346103707181162e-05, "loss": 0.0802, "step": 20640 }, { "grad_norm": 0.48254290223121643, "learning_rate": 9.345285954040626e-05, "loss": 0.0659, "step": 20650 }, { "grad_norm": 0.6861363649368286, "learning_rate": 9.34446772570615e-05, "loss": 0.0718, "step": 20660 }, { "grad_norm": 0.6414633393287659, "learning_rate": 9.343649022267214e-05, "loss": 0.072, "step": 20670 }, { "grad_norm": 2.5364573001861572, "learning_rate": 9.342829843813353e-05, "loss": 0.0751, "step": 20680 }, { "grad_norm": 0.4841941297054291, "learning_rate": 9.342010190434149e-05, "loss": 0.0835, "step": 20690 }, { "grad_norm": 0.765929639339447, "learning_rate": 9.34119006221924e-05, "loss": 0.0889, "step": 20700 }, { "grad_norm": 0.7140114307403564, "learning_rate": 9.340369459258313e-05, "loss": 0.0667, "step": 20710 }, { "grad_norm": 0.6852638721466064, "learning_rate": 9.339548381641106e-05, "loss": 0.0741, "step": 20720 }, { "grad_norm": 0.6302875876426697, "learning_rate": 9.338726829457413e-05, "loss": 0.0886, "step": 20730 }, { "grad_norm": 0.9010934829711914, "learning_rate": 9.337904802797078e-05, "loss": 0.0638, "step": 20740 }, { "grad_norm": 0.7430992126464844, "learning_rate": 9.337082301749993e-05, "loss": 0.0706, "step": 20750 }, { "grad_norm": 0.4072350859642029, "learning_rate": 9.336259326406109e-05, "loss": 0.0767, "step": 20760 }, { "grad_norm": 0.7173612117767334, "learning_rate": 9.335435876855427e-05, "loss": 0.0844, "step": 20770 }, { "grad_norm": 0.5499603152275085, "learning_rate": 9.334611953187994e-05, "loss": 0.0769, "step": 20780 }, { "grad_norm": 0.6591431498527527, "learning_rate": 9.333787555493914e-05, "loss": 0.0661, "step": 20790 }, { "grad_norm": 0.6080230474472046, "learning_rate": 9.332962683863345e-05, "loss": 0.0837, "step": 20800 }, { "grad_norm": 0.5198175311088562, "learning_rate": 9.332137338386489e-05, "loss": 0.0632, "step": 20810 }, { "grad_norm": 0.5624736547470093, "learning_rate": 9.33131151915361e-05, "loss": 0.0797, "step": 20820 }, { "grad_norm": 0.6196860671043396, "learning_rate": 9.330485226255012e-05, "loss": 0.0678, "step": 20830 }, { "grad_norm": 0.45435237884521484, "learning_rate": 9.329658459781061e-05, "loss": 0.0715, "step": 20840 }, { "grad_norm": 0.33298033475875854, "learning_rate": 9.328831219822172e-05, "loss": 0.0713, "step": 20850 }, { "grad_norm": 0.843715250492096, "learning_rate": 9.328003506468808e-05, "loss": 0.0876, "step": 20860 }, { "grad_norm": 0.6181851029396057, "learning_rate": 9.327175319811488e-05, "loss": 0.0687, "step": 20870 }, { "grad_norm": 0.44718730449676514, "learning_rate": 9.326346659940781e-05, "loss": 0.0722, "step": 20880 }, { "grad_norm": 0.5313760042190552, "learning_rate": 9.325517526947308e-05, "loss": 0.0781, "step": 20890 }, { "grad_norm": 0.6500895023345947, "learning_rate": 9.32468792092174e-05, "loss": 0.0796, "step": 20900 }, { "grad_norm": 0.5128629803657532, "learning_rate": 9.323857841954803e-05, "loss": 0.0747, "step": 20910 }, { "grad_norm": 0.8372079133987427, "learning_rate": 9.323027290137276e-05, "loss": 0.0763, "step": 20920 }, { "grad_norm": 0.43043503165245056, "learning_rate": 9.322196265559981e-05, "loss": 0.0772, "step": 20930 }, { "grad_norm": 0.5596475601196289, "learning_rate": 9.321364768313803e-05, "loss": 0.0715, "step": 20940 }, { "grad_norm": 0.5733376145362854, "learning_rate": 9.32053279848967e-05, "loss": 0.0827, "step": 20950 }, { "grad_norm": 0.5773961544036865, "learning_rate": 9.319700356178567e-05, "loss": 0.0755, "step": 20960 }, { "grad_norm": 0.583153486251831, "learning_rate": 9.318867441471527e-05, "loss": 0.0681, "step": 20970 }, { "grad_norm": 0.7347154021263123, "learning_rate": 9.318034054459637e-05, "loss": 0.0605, "step": 20980 }, { "grad_norm": 0.7350125312805176, "learning_rate": 9.317200195234034e-05, "loss": 0.0974, "step": 20990 }, { "grad_norm": 0.5036658644676208, "learning_rate": 9.316365863885909e-05, "loss": 0.094, "step": 21000 }, { "grad_norm": 0.7836974263191223, "learning_rate": 9.315531060506502e-05, "loss": 0.0915, "step": 21010 }, { "grad_norm": 0.5435460209846497, "learning_rate": 9.314695785187108e-05, "loss": 0.0814, "step": 21020 }, { "grad_norm": 0.50474613904953, "learning_rate": 9.313860038019069e-05, "loss": 0.074, "step": 21030 }, { "grad_norm": 1.327816367149353, "learning_rate": 9.313023819093782e-05, "loss": 0.1041, "step": 21040 }, { "grad_norm": 0.7143599390983582, "learning_rate": 9.312187128502695e-05, "loss": 0.0819, "step": 21050 }, { "grad_norm": 0.9605470299720764, "learning_rate": 9.311349966337307e-05, "loss": 0.0863, "step": 21060 }, { "grad_norm": 0.6344403028488159, "learning_rate": 9.310512332689169e-05, "loss": 0.0819, "step": 21070 }, { "grad_norm": 0.828437328338623, "learning_rate": 9.309674227649883e-05, "loss": 0.0719, "step": 21080 }, { "grad_norm": 0.613994836807251, "learning_rate": 9.308835651311103e-05, "loss": 0.0837, "step": 21090 }, { "grad_norm": 0.6954238414764404, "learning_rate": 9.307996603764533e-05, "loss": 0.0852, "step": 21100 }, { "grad_norm": 0.6285877227783203, "learning_rate": 9.307157085101932e-05, "loss": 0.0769, "step": 21110 }, { "grad_norm": 1.2409348487854004, "learning_rate": 9.306317095415109e-05, "loss": 0.0711, "step": 21120 }, { "grad_norm": 0.5163137316703796, "learning_rate": 9.305476634795922e-05, "loss": 0.0765, "step": 21130 }, { "grad_norm": 0.6738860607147217, "learning_rate": 9.304635703336284e-05, "loss": 0.0789, "step": 21140 }, { "grad_norm": 0.5468680262565613, "learning_rate": 9.303794301128157e-05, "loss": 0.0718, "step": 21150 }, { "grad_norm": 0.4567112326622009, "learning_rate": 9.302952428263555e-05, "loss": 0.0739, "step": 21160 }, { "grad_norm": 0.8333083391189575, "learning_rate": 9.302110084834545e-05, "loss": 0.0741, "step": 21170 }, { "grad_norm": 0.5968900918960571, "learning_rate": 9.301267270933245e-05, "loss": 0.0597, "step": 21180 }, { "grad_norm": 0.8278944492340088, "learning_rate": 9.300423986651823e-05, "loss": 0.0829, "step": 21190 }, { "grad_norm": 0.7444412112236023, "learning_rate": 9.299580232082501e-05, "loss": 0.0835, "step": 21200 }, { "grad_norm": 0.7642545104026794, "learning_rate": 9.298736007317547e-05, "loss": 0.0788, "step": 21210 }, { "grad_norm": 0.6753015518188477, "learning_rate": 9.297891312449288e-05, "loss": 0.0836, "step": 21220 }, { "grad_norm": 0.6480400562286377, "learning_rate": 9.297046147570094e-05, "loss": 0.1227, "step": 21230 }, { "grad_norm": 0.6421347260475159, "learning_rate": 9.296200512772396e-05, "loss": 0.0725, "step": 21240 }, { "grad_norm": 0.5971709489822388, "learning_rate": 9.295354408148668e-05, "loss": 0.0692, "step": 21250 }, { "grad_norm": 0.5424904823303223, "learning_rate": 9.294507833791441e-05, "loss": 0.0714, "step": 21260 }, { "grad_norm": 1.4248454570770264, "learning_rate": 9.293660789793295e-05, "loss": 0.0825, "step": 21270 }, { "grad_norm": 0.49347326159477234, "learning_rate": 9.292813276246858e-05, "loss": 0.0794, "step": 21280 }, { "grad_norm": 0.5868753790855408, "learning_rate": 9.291965293244816e-05, "loss": 0.0661, "step": 21290 }, { "grad_norm": 0.6279816031455994, "learning_rate": 9.291116840879904e-05, "loss": 0.0632, "step": 21300 }, { "grad_norm": 0.5918261408805847, "learning_rate": 9.290267919244904e-05, "loss": 0.0754, "step": 21310 }, { "grad_norm": 0.9085792303085327, "learning_rate": 9.289418528432655e-05, "loss": 0.0741, "step": 21320 }, { "grad_norm": 0.8476642370223999, "learning_rate": 9.288568668536045e-05, "loss": 0.0917, "step": 21330 }, { "grad_norm": 0.5891579985618591, "learning_rate": 9.287718339648013e-05, "loss": 0.071, "step": 21340 }, { "grad_norm": 0.823903501033783, "learning_rate": 9.28686754186155e-05, "loss": 0.0788, "step": 21350 }, { "grad_norm": 0.5141103267669678, "learning_rate": 9.286016275269698e-05, "loss": 0.1034, "step": 21360 }, { "grad_norm": 0.8613061308860779, "learning_rate": 9.285164539965551e-05, "loss": 0.0781, "step": 21370 }, { "grad_norm": 0.419220507144928, "learning_rate": 9.284312336042251e-05, "loss": 0.0669, "step": 21380 }, { "grad_norm": 0.7294270396232605, "learning_rate": 9.283459663592996e-05, "loss": 0.0719, "step": 21390 }, { "grad_norm": 0.5389617085456848, "learning_rate": 9.282606522711033e-05, "loss": 0.0757, "step": 21400 }, { "grad_norm": 0.7794533967971802, "learning_rate": 9.281752913489657e-05, "loss": 0.0903, "step": 21410 }, { "grad_norm": 0.551843523979187, "learning_rate": 9.280898836022222e-05, "loss": 0.0712, "step": 21420 }, { "grad_norm": 0.6295802593231201, "learning_rate": 9.280044290402126e-05, "loss": 0.0657, "step": 21430 }, { "grad_norm": 0.5039911270141602, "learning_rate": 9.279189276722821e-05, "loss": 0.07, "step": 21440 }, { "grad_norm": 0.4892365336418152, "learning_rate": 9.278333795077812e-05, "loss": 0.0719, "step": 21450 }, { "grad_norm": 0.6449318528175354, "learning_rate": 9.27747784556065e-05, "loss": 0.0596, "step": 21460 }, { "grad_norm": 0.5512619018554688, "learning_rate": 9.276621428264942e-05, "loss": 0.0858, "step": 21470 }, { "grad_norm": 0.6605653166770935, "learning_rate": 9.275764543284345e-05, "loss": 0.0698, "step": 21480 }, { "grad_norm": 0.48013994097709656, "learning_rate": 9.274907190712566e-05, "loss": 0.0652, "step": 21490 }, { "grad_norm": 1.7867152690887451, "learning_rate": 9.274049370643363e-05, "loss": 0.0846, "step": 21500 }, { "grad_norm": 1.0330541133880615, "learning_rate": 9.273191083170547e-05, "loss": 0.0851, "step": 21510 }, { "grad_norm": 0.669091522693634, "learning_rate": 9.27233232838798e-05, "loss": 0.0798, "step": 21520 }, { "grad_norm": 0.5925121307373047, "learning_rate": 9.27147310638957e-05, "loss": 0.0886, "step": 21530 }, { "grad_norm": 0.5001270771026611, "learning_rate": 9.270613417269286e-05, "loss": 0.0845, "step": 21540 }, { "grad_norm": 0.6983121037483215, "learning_rate": 9.269753261121138e-05, "loss": 0.0816, "step": 21550 }, { "grad_norm": 0.47562795877456665, "learning_rate": 9.268892638039194e-05, "loss": 0.0711, "step": 21560 }, { "grad_norm": 0.5844866037368774, "learning_rate": 9.268031548117569e-05, "loss": 0.0873, "step": 21570 }, { "grad_norm": 0.5478889346122742, "learning_rate": 9.26716999145043e-05, "loss": 0.068, "step": 21580 }, { "grad_norm": 0.675346851348877, "learning_rate": 9.266307968131998e-05, "loss": 0.0727, "step": 21590 }, { "grad_norm": 0.8605477809906006, "learning_rate": 9.26544547825654e-05, "loss": 0.0651, "step": 21600 }, { "grad_norm": 0.6524551510810852, "learning_rate": 9.264582521918376e-05, "loss": 0.0699, "step": 21610 }, { "grad_norm": 0.7174085974693298, "learning_rate": 9.263719099211881e-05, "loss": 0.0629, "step": 21620 }, { "grad_norm": 0.4799724817276001, "learning_rate": 9.262855210231476e-05, "loss": 0.0758, "step": 21630 }, { "grad_norm": 0.5004292130470276, "learning_rate": 9.261990855071633e-05, "loss": 0.0703, "step": 21640 }, { "grad_norm": 0.6859302520751953, "learning_rate": 9.261126033826878e-05, "loss": 0.063, "step": 21650 }, { "grad_norm": 0.45742931962013245, "learning_rate": 9.260260746591786e-05, "loss": 0.0641, "step": 21660 }, { "grad_norm": 0.5294561982154846, "learning_rate": 9.259394993460985e-05, "loss": 0.0688, "step": 21670 }, { "grad_norm": 0.43496090173721313, "learning_rate": 9.258528774529151e-05, "loss": 0.0656, "step": 21680 }, { "grad_norm": 0.6110910177230835, "learning_rate": 9.257662089891013e-05, "loss": 0.0608, "step": 21690 }, { "grad_norm": 0.6755549907684326, "learning_rate": 9.25679493964135e-05, "loss": 0.0749, "step": 21700 }, { "grad_norm": 0.412927508354187, "learning_rate": 9.255927323874994e-05, "loss": 0.0589, "step": 21710 }, { "grad_norm": 0.67795330286026, "learning_rate": 9.255059242686822e-05, "loss": 0.086, "step": 21720 }, { "grad_norm": 0.45270615816116333, "learning_rate": 9.254190696171769e-05, "loss": 0.0693, "step": 21730 }, { "grad_norm": 0.7006128430366516, "learning_rate": 9.25332168442482e-05, "loss": 0.0833, "step": 21740 }, { "grad_norm": 0.6041440367698669, "learning_rate": 9.252452207541004e-05, "loss": 0.0564, "step": 21750 }, { "grad_norm": 1.0629066228866577, "learning_rate": 9.251582265615409e-05, "loss": 0.0899, "step": 21760 }, { "grad_norm": 0.7483131289482117, "learning_rate": 9.250711858743169e-05, "loss": 0.0897, "step": 21770 }, { "grad_norm": 0.5585393905639648, "learning_rate": 9.24984098701947e-05, "loss": 0.0952, "step": 21780 }, { "grad_norm": 0.4645824432373047, "learning_rate": 9.248969650539552e-05, "loss": 0.0707, "step": 21790 }, { "grad_norm": 0.551457405090332, "learning_rate": 9.2480978493987e-05, "loss": 0.0702, "step": 21800 }, { "grad_norm": 0.5604602098464966, "learning_rate": 9.247225583692256e-05, "loss": 0.0664, "step": 21810 }, { "grad_norm": 0.6677916049957275, "learning_rate": 9.246352853515607e-05, "loss": 0.0659, "step": 21820 }, { "grad_norm": 0.5936517715454102, "learning_rate": 9.245479658964194e-05, "loss": 0.0644, "step": 21830 }, { "grad_norm": 0.7193658351898193, "learning_rate": 9.244606000133507e-05, "loss": 0.0667, "step": 21840 }, { "grad_norm": 0.47185876965522766, "learning_rate": 9.24373187711909e-05, "loss": 0.0638, "step": 21850 }, { "grad_norm": 0.6717560887336731, "learning_rate": 9.242857290016537e-05, "loss": 0.0761, "step": 21860 }, { "grad_norm": 0.5930526256561279, "learning_rate": 9.241982238921488e-05, "loss": 0.0819, "step": 21870 }, { "grad_norm": 0.5179972648620605, "learning_rate": 9.24110672392964e-05, "loss": 0.0563, "step": 21880 }, { "grad_norm": 0.35789981484413147, "learning_rate": 9.240230745136737e-05, "loss": 0.074, "step": 21890 }, { "grad_norm": 0.5145551562309265, "learning_rate": 9.239354302638575e-05, "loss": 0.063, "step": 21900 }, { "grad_norm": 0.7070452570915222, "learning_rate": 9.238477396531e-05, "loss": 0.0822, "step": 21910 }, { "grad_norm": 0.4778318405151367, "learning_rate": 9.23760002690991e-05, "loss": 0.0658, "step": 21920 }, { "grad_norm": 0.6271228790283203, "learning_rate": 9.236722193871252e-05, "loss": 0.0676, "step": 21930 }, { "grad_norm": 0.5469778776168823, "learning_rate": 9.235843897511023e-05, "loss": 0.0763, "step": 21940 }, { "grad_norm": 0.7241756916046143, "learning_rate": 9.234965137925276e-05, "loss": 0.0578, "step": 21950 }, { "grad_norm": 0.5537359118461609, "learning_rate": 9.234085915210108e-05, "loss": 0.0796, "step": 21960 }, { "grad_norm": 0.5766344666481018, "learning_rate": 9.23320622946167e-05, "loss": 0.0739, "step": 21970 }, { "grad_norm": 0.45937615633010864, "learning_rate": 9.232326080776163e-05, "loss": 0.058, "step": 21980 }, { "grad_norm": 0.4530909061431885, "learning_rate": 9.23144546924984e-05, "loss": 0.0588, "step": 21990 }, { "grad_norm": 0.5588677525520325, "learning_rate": 9.230564394979e-05, "loss": 0.0598, "step": 22000 }, { "grad_norm": 0.5315656065940857, "learning_rate": 9.22968285806e-05, "loss": 0.0508, "step": 22010 }, { "grad_norm": 0.5548833608627319, "learning_rate": 9.228800858589242e-05, "loss": 0.0679, "step": 22020 }, { "grad_norm": 0.5314732789993286, "learning_rate": 9.227918396663179e-05, "loss": 0.0721, "step": 22030 }, { "grad_norm": 1.0940173864364624, "learning_rate": 9.227035472378319e-05, "loss": 0.0735, "step": 22040 }, { "grad_norm": 0.6494933366775513, "learning_rate": 9.226152085831213e-05, "loss": 0.0687, "step": 22050 }, { "grad_norm": 0.575812578201294, "learning_rate": 9.22526823711847e-05, "loss": 0.0628, "step": 22060 }, { "grad_norm": 0.7759490013122559, "learning_rate": 9.224383926336745e-05, "loss": 0.0701, "step": 22070 }, { "grad_norm": 0.47884437441825867, "learning_rate": 9.223499153582744e-05, "loss": 0.0503, "step": 22080 }, { "grad_norm": 0.4982638359069824, "learning_rate": 9.222613918953226e-05, "loss": 0.068, "step": 22090 }, { "grad_norm": 0.6098856329917908, "learning_rate": 9.221728222544999e-05, "loss": 0.0699, "step": 22100 }, { "grad_norm": 0.5329023599624634, "learning_rate": 9.22084206445492e-05, "loss": 0.0775, "step": 22110 }, { "grad_norm": 0.5499184727668762, "learning_rate": 9.2199554447799e-05, "loss": 0.0848, "step": 22120 }, { "grad_norm": 0.4468596279621124, "learning_rate": 9.219068363616897e-05, "loss": 0.0698, "step": 22130 }, { "grad_norm": 0.48252415657043457, "learning_rate": 9.218180821062919e-05, "loss": 0.0709, "step": 22140 }, { "grad_norm": 0.4900990426540375, "learning_rate": 9.21729281721503e-05, "loss": 0.0774, "step": 22150 }, { "grad_norm": 0.6203591227531433, "learning_rate": 9.216404352170339e-05, "loss": 0.0864, "step": 22160 }, { "grad_norm": 0.606970489025116, "learning_rate": 9.215515426026007e-05, "loss": 0.0687, "step": 22170 }, { "grad_norm": 0.4599743187427521, "learning_rate": 9.214626038879246e-05, "loss": 0.0609, "step": 22180 }, { "grad_norm": 0.7704225778579712, "learning_rate": 9.21373619082732e-05, "loss": 0.0748, "step": 22190 }, { "grad_norm": 1.5927897691726685, "learning_rate": 9.212845881967535e-05, "loss": 0.0793, "step": 22200 }, { "grad_norm": 0.7843074202537537, "learning_rate": 9.211955112397262e-05, "loss": 0.0749, "step": 22210 }, { "grad_norm": 0.4227408468723297, "learning_rate": 9.211063882213909e-05, "loss": 0.0778, "step": 22220 }, { "grad_norm": 0.7614023089408875, "learning_rate": 9.210172191514942e-05, "loss": 0.0724, "step": 22230 }, { "grad_norm": 0.4736369848251343, "learning_rate": 9.209280040397874e-05, "loss": 0.0781, "step": 22240 }, { "grad_norm": 0.5439819097518921, "learning_rate": 9.208387428960268e-05, "loss": 0.0587, "step": 22250 }, { "grad_norm": 0.5778505802154541, "learning_rate": 9.20749435729974e-05, "loss": 0.0694, "step": 22260 }, { "grad_norm": 0.42079466581344604, "learning_rate": 9.206600825513957e-05, "loss": 0.0706, "step": 22270 }, { "grad_norm": 0.5054510235786438, "learning_rate": 9.20570683370063e-05, "loss": 0.0811, "step": 22280 }, { "grad_norm": 0.7712993025779724, "learning_rate": 9.204812381957528e-05, "loss": 0.0648, "step": 22290 }, { "grad_norm": 0.6860432624816895, "learning_rate": 9.203917470382465e-05, "loss": 0.0638, "step": 22300 }, { "grad_norm": 0.634657084941864, "learning_rate": 9.203022099073309e-05, "loss": 0.067, "step": 22310 }, { "grad_norm": 1.6261208057403564, "learning_rate": 9.202126268127976e-05, "loss": 0.0723, "step": 22320 }, { "grad_norm": 0.6968977451324463, "learning_rate": 9.20122997764443e-05, "loss": 0.0762, "step": 22330 }, { "grad_norm": 0.8041656017303467, "learning_rate": 9.200333227720692e-05, "loss": 0.0856, "step": 22340 }, { "grad_norm": 0.7481971979141235, "learning_rate": 9.199436018454826e-05, "loss": 0.0843, "step": 22350 }, { "grad_norm": 0.6296408176422119, "learning_rate": 9.198538349944952e-05, "loss": 0.0688, "step": 22360 }, { "grad_norm": 0.5773420333862305, "learning_rate": 9.197640222289234e-05, "loss": 0.0791, "step": 22370 }, { "grad_norm": 0.6660444140434265, "learning_rate": 9.196741635585895e-05, "loss": 0.0728, "step": 22380 }, { "grad_norm": 0.6050229668617249, "learning_rate": 9.195842589933199e-05, "loss": 0.0615, "step": 22390 }, { "grad_norm": 0.5271933674812317, "learning_rate": 9.194943085429466e-05, "loss": 0.0785, "step": 22400 }, { "grad_norm": 0.45671388506889343, "learning_rate": 9.194043122173065e-05, "loss": 0.0894, "step": 22410 }, { "grad_norm": 0.8878074288368225, "learning_rate": 9.193142700262413e-05, "loss": 0.0702, "step": 22420 }, { "grad_norm": 0.8566411137580872, "learning_rate": 9.192241819795979e-05, "loss": 0.0771, "step": 22430 }, { "grad_norm": 0.5179269909858704, "learning_rate": 9.191340480872284e-05, "loss": 0.0887, "step": 22440 }, { "grad_norm": 0.6500645875930786, "learning_rate": 9.190438683589895e-05, "loss": 0.0895, "step": 22450 }, { "grad_norm": 0.5372434258460999, "learning_rate": 9.189536428047432e-05, "loss": 0.0771, "step": 22460 }, { "grad_norm": 0.5353057384490967, "learning_rate": 9.188633714343564e-05, "loss": 0.0664, "step": 22470 }, { "grad_norm": 1.0635985136032104, "learning_rate": 9.18773054257701e-05, "loss": 0.0842, "step": 22480 }, { "grad_norm": 0.6459642648696899, "learning_rate": 9.18682691284654e-05, "loss": 0.0657, "step": 22490 }, { "grad_norm": 0.5176106691360474, "learning_rate": 9.185922825250974e-05, "loss": 0.0753, "step": 22500 }, { "grad_norm": 0.715826690196991, "learning_rate": 9.185018279889181e-05, "loss": 0.0576, "step": 22510 }, { "grad_norm": 1.3188633918762207, "learning_rate": 9.184113276860082e-05, "loss": 0.0726, "step": 22520 }, { "grad_norm": 0.6852982044219971, "learning_rate": 9.183207816262645e-05, "loss": 0.0706, "step": 22530 }, { "grad_norm": 0.5655123591423035, "learning_rate": 9.182301898195891e-05, "loss": 0.0678, "step": 22540 }, { "grad_norm": 0.3872191309928894, "learning_rate": 9.181395522758889e-05, "loss": 0.0765, "step": 22550 }, { "grad_norm": 0.6788960695266724, "learning_rate": 9.180488690050759e-05, "loss": 0.0854, "step": 22560 }, { "grad_norm": 0.49551090598106384, "learning_rate": 9.179581400170671e-05, "loss": 0.0685, "step": 22570 }, { "grad_norm": 0.5145838260650635, "learning_rate": 9.178673653217845e-05, "loss": 0.0654, "step": 22580 }, { "grad_norm": 0.5672488212585449, "learning_rate": 9.177765449291551e-05, "loss": 0.0683, "step": 22590 }, { "grad_norm": 1.1381841897964478, "learning_rate": 9.176856788491109e-05, "loss": 0.0863, "step": 22600 }, { "grad_norm": 0.5419642925262451, "learning_rate": 9.175947670915887e-05, "loss": 0.0751, "step": 22610 }, { "grad_norm": 0.6868036389350891, "learning_rate": 9.175038096665309e-05, "loss": 0.0743, "step": 22620 }, { "grad_norm": 0.5282338857650757, "learning_rate": 9.17412806583884e-05, "loss": 0.0601, "step": 22630 }, { "grad_norm": 0.6955615878105164, "learning_rate": 9.173217578536002e-05, "loss": 0.0844, "step": 22640 }, { "grad_norm": 0.6009126305580139, "learning_rate": 9.172306634856362e-05, "loss": 0.0736, "step": 22650 }, { "grad_norm": 0.526646077632904, "learning_rate": 9.171395234899545e-05, "loss": 0.0845, "step": 22660 }, { "grad_norm": 0.8313404321670532, "learning_rate": 9.170483378765214e-05, "loss": 0.0738, "step": 22670 }, { "grad_norm": 0.5959355235099792, "learning_rate": 9.169571066553091e-05, "loss": 0.073, "step": 22680 }, { "grad_norm": 0.5820912718772888, "learning_rate": 9.168658298362946e-05, "loss": 0.0741, "step": 22690 }, { "grad_norm": 0.6142944097518921, "learning_rate": 9.167745074294598e-05, "loss": 0.0722, "step": 22700 }, { "grad_norm": 0.7064096331596375, "learning_rate": 9.166831394447913e-05, "loss": 0.0882, "step": 22710 }, { "grad_norm": 0.577438473701477, "learning_rate": 9.165917258922812e-05, "loss": 0.0679, "step": 22720 }, { "grad_norm": 0.7500472664833069, "learning_rate": 9.165002667819262e-05, "loss": 0.0799, "step": 22730 }, { "grad_norm": 1.1997051239013672, "learning_rate": 9.164087621237282e-05, "loss": 0.0631, "step": 22740 }, { "grad_norm": 0.5851187109947205, "learning_rate": 9.163172119276942e-05, "loss": 0.0893, "step": 22750 }, { "grad_norm": 0.5997381210327148, "learning_rate": 9.162256162038358e-05, "loss": 0.0827, "step": 22760 }, { "grad_norm": 0.6324490904808044, "learning_rate": 9.161339749621698e-05, "loss": 0.0739, "step": 22770 }, { "grad_norm": 0.4740958511829376, "learning_rate": 9.160422882127177e-05, "loss": 0.0724, "step": 22780 }, { "grad_norm": 0.5932860970497131, "learning_rate": 9.159505559655069e-05, "loss": 0.0812, "step": 22790 }, { "grad_norm": 0.7220009565353394, "learning_rate": 9.158587782305684e-05, "loss": 0.0731, "step": 22800 }, { "grad_norm": 0.6193075180053711, "learning_rate": 9.157669550179391e-05, "loss": 0.0855, "step": 22810 }, { "grad_norm": 0.5628538131713867, "learning_rate": 9.156750863376609e-05, "loss": 0.0825, "step": 22820 }, { "grad_norm": 0.5332059860229492, "learning_rate": 9.155831721997801e-05, "loss": 0.0873, "step": 22830 }, { "grad_norm": 0.8026899695396423, "learning_rate": 9.154912126143484e-05, "loss": 0.0823, "step": 22840 }, { "grad_norm": 1.0164507627487183, "learning_rate": 9.153992075914224e-05, "loss": 0.0916, "step": 22850 }, { "grad_norm": 0.6661425828933716, "learning_rate": 9.153071571410635e-05, "loss": 0.0716, "step": 22860 }, { "grad_norm": 0.6746891140937805, "learning_rate": 9.152150612733384e-05, "loss": 0.0657, "step": 22870 }, { "grad_norm": 0.4369620978832245, "learning_rate": 9.151229199983184e-05, "loss": 0.0665, "step": 22880 }, { "grad_norm": 0.5678942203521729, "learning_rate": 9.150307333260802e-05, "loss": 0.0713, "step": 22890 }, { "grad_norm": 0.5693389177322388, "learning_rate": 9.149385012667048e-05, "loss": 0.086, "step": 22900 }, { "grad_norm": 0.8287730813026428, "learning_rate": 9.148462238302788e-05, "loss": 0.0893, "step": 22910 }, { "grad_norm": 0.7145648002624512, "learning_rate": 9.147539010268936e-05, "loss": 0.0877, "step": 22920 }, { "grad_norm": 0.8164507746696472, "learning_rate": 9.14661532866645e-05, "loss": 0.0832, "step": 22930 }, { "grad_norm": 0.7681787014007568, "learning_rate": 9.145691193596348e-05, "loss": 0.0613, "step": 22940 }, { "grad_norm": 0.7754125595092773, "learning_rate": 9.144766605159691e-05, "loss": 0.0974, "step": 22950 }, { "grad_norm": 0.8410392999649048, "learning_rate": 9.14384156345759e-05, "loss": 0.076, "step": 22960 }, { "grad_norm": 0.8691170811653137, "learning_rate": 9.142916068591204e-05, "loss": 0.0692, "step": 22970 }, { "grad_norm": 0.4546382427215576, "learning_rate": 9.141990120661746e-05, "loss": 0.0692, "step": 22980 }, { "grad_norm": 0.43914639949798584, "learning_rate": 9.141063719770475e-05, "loss": 0.0731, "step": 22990 }, { "grad_norm": 0.5515855550765991, "learning_rate": 9.140136866018704e-05, "loss": 0.0737, "step": 23000 }, { "grad_norm": 0.45674845576286316, "learning_rate": 9.139209559507788e-05, "loss": 0.0673, "step": 23010 }, { "grad_norm": 0.4801057279109955, "learning_rate": 9.13828180033914e-05, "loss": 0.0679, "step": 23020 }, { "grad_norm": 1.2573009729385376, "learning_rate": 9.137353588614212e-05, "loss": 0.0757, "step": 23030 }, { "grad_norm": 0.6160606145858765, "learning_rate": 9.136424924434519e-05, "loss": 0.0817, "step": 23040 }, { "grad_norm": 0.6953532695770264, "learning_rate": 9.135495807901615e-05, "loss": 0.0905, "step": 23050 }, { "grad_norm": 0.4345764219760895, "learning_rate": 9.134566239117108e-05, "loss": 0.0814, "step": 23060 }, { "grad_norm": 0.738008975982666, "learning_rate": 9.13363621818265e-05, "loss": 0.0715, "step": 23070 }, { "grad_norm": 1.564078450202942, "learning_rate": 9.132705745199953e-05, "loss": 0.0866, "step": 23080 }, { "grad_norm": 0.5483231544494629, "learning_rate": 9.131774820270768e-05, "loss": 0.0739, "step": 23090 }, { "grad_norm": 0.7390759587287903, "learning_rate": 9.130843443496901e-05, "loss": 0.0702, "step": 23100 }, { "grad_norm": 1.1025139093399048, "learning_rate": 9.129911614980206e-05, "loss": 0.0714, "step": 23110 }, { "grad_norm": 0.6329725980758667, "learning_rate": 9.128979334822584e-05, "loss": 0.0619, "step": 23120 }, { "grad_norm": 0.7594168186187744, "learning_rate": 9.128046603125992e-05, "loss": 0.0627, "step": 23130 }, { "grad_norm": 0.4410029947757721, "learning_rate": 9.12711341999243e-05, "loss": 0.0647, "step": 23140 }, { "grad_norm": 0.43977710604667664, "learning_rate": 9.12617978552395e-05, "loss": 0.0846, "step": 23150 }, { "grad_norm": 0.4387228488922119, "learning_rate": 9.12524569982265e-05, "loss": 0.0606, "step": 23160 }, { "grad_norm": 0.6078830361366272, "learning_rate": 9.124311162990684e-05, "loss": 0.0689, "step": 23170 }, { "grad_norm": 0.5704004764556885, "learning_rate": 9.12337617513025e-05, "loss": 0.0669, "step": 23180 }, { "grad_norm": 0.636344313621521, "learning_rate": 9.122440736343596e-05, "loss": 0.0636, "step": 23190 }, { "grad_norm": 0.43964460492134094, "learning_rate": 9.12150484673302e-05, "loss": 0.0664, "step": 23200 }, { "grad_norm": 0.9016568064689636, "learning_rate": 9.120568506400873e-05, "loss": 0.0695, "step": 23210 }, { "grad_norm": 0.5422278046607971, "learning_rate": 9.119631715449548e-05, "loss": 0.0662, "step": 23220 }, { "grad_norm": 0.9787745475769043, "learning_rate": 9.118694473981493e-05, "loss": 0.0783, "step": 23230 }, { "grad_norm": 0.7218133211135864, "learning_rate": 9.117756782099203e-05, "loss": 0.0919, "step": 23240 }, { "grad_norm": 1.1451122760772705, "learning_rate": 9.11681863990522e-05, "loss": 0.0761, "step": 23250 }, { "grad_norm": 1.2113710641860962, "learning_rate": 9.115880047502142e-05, "loss": 0.0704, "step": 23260 }, { "grad_norm": 0.6766378283500671, "learning_rate": 9.114941004992609e-05, "loss": 0.0899, "step": 23270 }, { "grad_norm": 0.6688621640205383, "learning_rate": 9.114001512479317e-05, "loss": 0.0745, "step": 23280 }, { "grad_norm": 0.9939049482345581, "learning_rate": 9.113061570065003e-05, "loss": 0.092, "step": 23290 }, { "grad_norm": 0.6072201132774353, "learning_rate": 9.112121177852459e-05, "loss": 0.0732, "step": 23300 }, { "grad_norm": 0.5840886831283569, "learning_rate": 9.111180335944527e-05, "loss": 0.0766, "step": 23310 }, { "grad_norm": 0.6445093154907227, "learning_rate": 9.110239044444093e-05, "loss": 0.0732, "step": 23320 }, { "grad_norm": 0.6979678273200989, "learning_rate": 9.109297303454099e-05, "loss": 0.0729, "step": 23330 }, { "grad_norm": 0.5881196856498718, "learning_rate": 9.108355113077526e-05, "loss": 0.0679, "step": 23340 }, { "grad_norm": 0.6627374291419983, "learning_rate": 9.107412473417419e-05, "loss": 0.0851, "step": 23350 }, { "grad_norm": 0.6113040447235107, "learning_rate": 9.106469384576858e-05, "loss": 0.0734, "step": 23360 }, { "grad_norm": 1.9890719652175903, "learning_rate": 9.105525846658978e-05, "loss": 0.0876, "step": 23370 }, { "grad_norm": 0.47123873233795166, "learning_rate": 9.104581859766965e-05, "loss": 0.0842, "step": 23380 }, { "grad_norm": 0.5555893182754517, "learning_rate": 9.10363742400405e-05, "loss": 0.0765, "step": 23390 }, { "grad_norm": 0.7413575053215027, "learning_rate": 9.102692539473518e-05, "loss": 0.0775, "step": 23400 }, { "grad_norm": 0.8111522793769836, "learning_rate": 9.101747206278697e-05, "loss": 0.0838, "step": 23410 }, { "grad_norm": 0.5060809850692749, "learning_rate": 9.100801424522968e-05, "loss": 0.0772, "step": 23420 }, { "grad_norm": 0.4792167544364929, "learning_rate": 9.099855194309762e-05, "loss": 0.0763, "step": 23430 }, { "grad_norm": 0.5924479961395264, "learning_rate": 9.098908515742554e-05, "loss": 0.0705, "step": 23440 }, { "grad_norm": 0.5454757809638977, "learning_rate": 9.097961388924873e-05, "loss": 0.078, "step": 23450 }, { "grad_norm": 0.6844363212585449, "learning_rate": 9.097013813960298e-05, "loss": 0.0746, "step": 23460 }, { "grad_norm": 0.9061334133148193, "learning_rate": 9.09606579095245e-05, "loss": 0.0858, "step": 23470 }, { "grad_norm": 0.6271457076072693, "learning_rate": 9.095117320005008e-05, "loss": 0.0729, "step": 23480 }, { "grad_norm": 0.8584481477737427, "learning_rate": 9.094168401221691e-05, "loss": 0.0781, "step": 23490 }, { "grad_norm": 0.7123224139213562, "learning_rate": 9.093219034706273e-05, "loss": 0.0936, "step": 23500 }, { "grad_norm": 0.8250432014465332, "learning_rate": 9.092269220562577e-05, "loss": 0.0832, "step": 23510 }, { "grad_norm": 0.5382727980613708, "learning_rate": 9.09131895889447e-05, "loss": 0.094, "step": 23520 }, { "grad_norm": 0.48873722553253174, "learning_rate": 9.090368249805873e-05, "loss": 0.0574, "step": 23530 }, { "grad_norm": 0.5057679414749146, "learning_rate": 9.089417093400754e-05, "loss": 0.0705, "step": 23540 }, { "grad_norm": 0.5364145040512085, "learning_rate": 9.088465489783131e-05, "loss": 0.0536, "step": 23550 }, { "grad_norm": 0.48550790548324585, "learning_rate": 9.087513439057068e-05, "loss": 0.0671, "step": 23560 }, { "grad_norm": 0.4962548315525055, "learning_rate": 9.08656094132668e-05, "loss": 0.0751, "step": 23570 }, { "grad_norm": 0.7993443608283997, "learning_rate": 9.085607996696134e-05, "loss": 0.0641, "step": 23580 }, { "grad_norm": 0.6363505721092224, "learning_rate": 9.084654605269639e-05, "loss": 0.0859, "step": 23590 }, { "grad_norm": 0.5400476455688477, "learning_rate": 9.083700767151457e-05, "loss": 0.0697, "step": 23600 }, { "grad_norm": 0.397003710269928, "learning_rate": 9.082746482445898e-05, "loss": 0.0666, "step": 23610 }, { "grad_norm": 0.6601568460464478, "learning_rate": 9.081791751257325e-05, "loss": 0.0662, "step": 23620 }, { "grad_norm": 0.5797028541564941, "learning_rate": 9.080836573690142e-05, "loss": 0.07, "step": 23630 }, { "grad_norm": 0.5888232588768005, "learning_rate": 9.079880949848805e-05, "loss": 0.0754, "step": 23640 }, { "grad_norm": 0.477382630109787, "learning_rate": 9.078924879837822e-05, "loss": 0.0715, "step": 23650 }, { "grad_norm": 0.720299482345581, "learning_rate": 9.077968363761747e-05, "loss": 0.0761, "step": 23660 }, { "grad_norm": 0.897735595703125, "learning_rate": 9.077011401725182e-05, "loss": 0.0581, "step": 23670 }, { "grad_norm": 0.42502620816230774, "learning_rate": 9.07605399383278e-05, "loss": 0.0725, "step": 23680 }, { "grad_norm": 0.5451716780662537, "learning_rate": 9.075096140189243e-05, "loss": 0.0716, "step": 23690 }, { "grad_norm": 0.9873819351196289, "learning_rate": 9.074137840899318e-05, "loss": 0.0648, "step": 23700 }, { "grad_norm": 0.9549939632415771, "learning_rate": 9.073179096067804e-05, "loss": 0.0832, "step": 23710 }, { "grad_norm": 0.5597211122512817, "learning_rate": 9.072219905799549e-05, "loss": 0.0737, "step": 23720 }, { "grad_norm": 0.5776735544204712, "learning_rate": 9.071260270199447e-05, "loss": 0.0821, "step": 23730 }, { "grad_norm": 0.6205254197120667, "learning_rate": 9.070300189372441e-05, "loss": 0.075, "step": 23740 }, { "grad_norm": 0.8351466059684753, "learning_rate": 9.069339663423528e-05, "loss": 0.0661, "step": 23750 }, { "grad_norm": 0.5697287321090698, "learning_rate": 9.068378692457747e-05, "loss": 0.0784, "step": 23760 }, { "grad_norm": 0.4857098162174225, "learning_rate": 9.067417276580189e-05, "loss": 0.0682, "step": 23770 }, { "grad_norm": 0.4696704149246216, "learning_rate": 9.066455415895993e-05, "loss": 0.0619, "step": 23780 }, { "grad_norm": 0.5089378952980042, "learning_rate": 9.065493110510346e-05, "loss": 0.0772, "step": 23790 }, { "grad_norm": 0.5441782474517822, "learning_rate": 9.064530360528484e-05, "loss": 0.0685, "step": 23800 }, { "grad_norm": 0.5228333473205566, "learning_rate": 9.063567166055695e-05, "loss": 0.061, "step": 23810 }, { "grad_norm": 0.9231544733047485, "learning_rate": 9.062603527197308e-05, "loss": 0.0633, "step": 23820 }, { "grad_norm": 0.9049921631813049, "learning_rate": 9.06163944405871e-05, "loss": 0.0684, "step": 23830 }, { "grad_norm": 0.6938779354095459, "learning_rate": 9.060674916745327e-05, "loss": 0.0747, "step": 23840 }, { "grad_norm": 0.6415536999702454, "learning_rate": 9.05970994536264e-05, "loss": 0.0679, "step": 23850 }, { "grad_norm": 0.6960239410400391, "learning_rate": 9.05874453001618e-05, "loss": 0.072, "step": 23860 }, { "grad_norm": 0.7511563897132874, "learning_rate": 9.057778670811517e-05, "loss": 0.0729, "step": 23870 }, { "grad_norm": 0.7568001747131348, "learning_rate": 9.056812367854281e-05, "loss": 0.0774, "step": 23880 }, { "grad_norm": 0.619308590888977, "learning_rate": 9.055845621250143e-05, "loss": 0.0679, "step": 23890 }, { "grad_norm": 0.4153578579425812, "learning_rate": 9.054878431104825e-05, "loss": 0.0725, "step": 23900 }, { "grad_norm": 0.8051710724830627, "learning_rate": 9.0539107975241e-05, "loss": 0.0693, "step": 23910 }, { "grad_norm": 0.7750213742256165, "learning_rate": 9.052942720613784e-05, "loss": 0.0681, "step": 23920 }, { "grad_norm": 0.6410121321678162, "learning_rate": 9.051974200479745e-05, "loss": 0.0611, "step": 23930 }, { "grad_norm": 0.5618603825569153, "learning_rate": 9.051005237227901e-05, "loss": 0.0518, "step": 23940 }, { "grad_norm": 0.6413738131523132, "learning_rate": 9.050035830964215e-05, "loss": 0.0548, "step": 23950 }, { "grad_norm": 0.743028998374939, "learning_rate": 9.049065981794698e-05, "loss": 0.0841, "step": 23960 }, { "grad_norm": 0.96608966588974, "learning_rate": 9.048095689825414e-05, "loss": 0.0703, "step": 23970 }, { "grad_norm": 0.971559464931488, "learning_rate": 9.047124955162472e-05, "loss": 0.0667, "step": 23980 }, { "grad_norm": 0.6935936212539673, "learning_rate": 9.046153777912028e-05, "loss": 0.0809, "step": 23990 }, { "grad_norm": 0.7656402587890625, "learning_rate": 9.045182158180292e-05, "loss": 0.0648, "step": 24000 }, { "grad_norm": 0.553074836730957, "learning_rate": 9.044210096073516e-05, "loss": 0.0579, "step": 24010 }, { "grad_norm": 0.46362611651420593, "learning_rate": 9.043237591698004e-05, "loss": 0.0701, "step": 24020 }, { "grad_norm": 0.5826522707939148, "learning_rate": 9.04226464516011e-05, "loss": 0.062, "step": 24030 }, { "grad_norm": 0.6514103412628174, "learning_rate": 9.041291256566229e-05, "loss": 0.0776, "step": 24040 }, { "grad_norm": 0.6746326684951782, "learning_rate": 9.040317426022814e-05, "loss": 0.0631, "step": 24050 }, { "grad_norm": 0.7547581195831299, "learning_rate": 9.03934315363636e-05, "loss": 0.0768, "step": 24060 }, { "grad_norm": 0.5346916317939758, "learning_rate": 9.038368439513409e-05, "loss": 0.0593, "step": 24070 }, { "grad_norm": 0.555508553981781, "learning_rate": 9.03739328376056e-05, "loss": 0.0776, "step": 24080 }, { "grad_norm": 0.7448802590370178, "learning_rate": 9.036417686484451e-05, "loss": 0.065, "step": 24090 }, { "grad_norm": 0.4679732322692871, "learning_rate": 9.035441647791773e-05, "loss": 0.0573, "step": 24100 }, { "grad_norm": 0.45687738060951233, "learning_rate": 9.034465167789263e-05, "loss": 0.0691, "step": 24110 }, { "grad_norm": 0.4689781069755554, "learning_rate": 9.033488246583706e-05, "loss": 0.0623, "step": 24120 }, { "grad_norm": 0.3693999946117401, "learning_rate": 9.032510884281941e-05, "loss": 0.0751, "step": 24130 }, { "grad_norm": 0.46886783838272095, "learning_rate": 9.031533080990848e-05, "loss": 0.0676, "step": 24140 }, { "grad_norm": 1.2756651639938354, "learning_rate": 9.030554836817358e-05, "loss": 0.0709, "step": 24150 }, { "grad_norm": 0.4688771367073059, "learning_rate": 9.029576151868451e-05, "loss": 0.0588, "step": 24160 }, { "grad_norm": 0.6463944315910339, "learning_rate": 9.028597026251155e-05, "loss": 0.0596, "step": 24170 }, { "grad_norm": 0.5918083786964417, "learning_rate": 9.027617460072547e-05, "loss": 0.0747, "step": 24180 }, { "grad_norm": 0.5483765006065369, "learning_rate": 9.026637453439745e-05, "loss": 0.0706, "step": 24190 }, { "grad_norm": 0.5491152405738831, "learning_rate": 9.025657006459927e-05, "loss": 0.0627, "step": 24200 }, { "grad_norm": 0.7887707352638245, "learning_rate": 9.024676119240311e-05, "loss": 0.0612, "step": 24210 }, { "grad_norm": 0.46420595049858093, "learning_rate": 9.023694791888166e-05, "loss": 0.0632, "step": 24220 }, { "grad_norm": 0.4344674348831177, "learning_rate": 9.022713024510808e-05, "loss": 0.069, "step": 24230 }, { "grad_norm": 0.7547239661216736, "learning_rate": 9.021730817215601e-05, "loss": 0.0654, "step": 24240 }, { "grad_norm": 0.7640465497970581, "learning_rate": 9.02074817010996e-05, "loss": 0.0846, "step": 24250 }, { "grad_norm": 0.5737211108207703, "learning_rate": 9.019765083301342e-05, "loss": 0.0726, "step": 24260 }, { "grad_norm": 0.5528131127357483, "learning_rate": 9.01878155689726e-05, "loss": 0.0712, "step": 24270 }, { "grad_norm": 0.6369527578353882, "learning_rate": 9.017797591005268e-05, "loss": 0.0592, "step": 24280 }, { "grad_norm": 0.4378754198551178, "learning_rate": 9.016813185732972e-05, "loss": 0.0786, "step": 24290 }, { "grad_norm": 0.6043114066123962, "learning_rate": 9.015828341188027e-05, "loss": 0.0787, "step": 24300 }, { "grad_norm": 1.1113321781158447, "learning_rate": 9.01484305747813e-05, "loss": 0.0674, "step": 24310 }, { "grad_norm": 0.6171440482139587, "learning_rate": 9.013857334711033e-05, "loss": 0.0695, "step": 24320 }, { "grad_norm": 0.8844628930091858, "learning_rate": 9.012871172994534e-05, "loss": 0.0609, "step": 24330 }, { "grad_norm": 0.5889569520950317, "learning_rate": 9.011884572436476e-05, "loss": 0.0717, "step": 24340 }, { "grad_norm": 0.808537483215332, "learning_rate": 9.010897533144754e-05, "loss": 0.0709, "step": 24350 }, { "grad_norm": 0.8214558362960815, "learning_rate": 9.009910055227306e-05, "loss": 0.0654, "step": 24360 }, { "grad_norm": 0.5474997758865356, "learning_rate": 9.008922138792124e-05, "loss": 0.0567, "step": 24370 }, { "grad_norm": 1.424925684928894, "learning_rate": 9.007933783947244e-05, "loss": 0.0624, "step": 24380 }, { "grad_norm": 0.6656487584114075, "learning_rate": 9.006944990800752e-05, "loss": 0.06, "step": 24390 }, { "grad_norm": 0.9707643389701843, "learning_rate": 9.005955759460779e-05, "loss": 0.0707, "step": 24400 }, { "grad_norm": 0.8849921822547913, "learning_rate": 9.004966090035508e-05, "loss": 0.0621, "step": 24410 }, { "grad_norm": 0.3558885157108307, "learning_rate": 9.003975982633166e-05, "loss": 0.0605, "step": 24420 }, { "grad_norm": 0.5648341774940491, "learning_rate": 9.00298543736203e-05, "loss": 0.066, "step": 24430 }, { "grad_norm": 0.6956040263175964, "learning_rate": 9.001994454330427e-05, "loss": 0.0606, "step": 24440 }, { "grad_norm": 0.6014798283576965, "learning_rate": 9.001003033646727e-05, "loss": 0.0709, "step": 24450 }, { "grad_norm": 0.5286750793457031, "learning_rate": 9.00001117541935e-05, "loss": 0.0666, "step": 24460 }, { "grad_norm": 0.6592404246330261, "learning_rate": 8.999018879756764e-05, "loss": 0.0694, "step": 24470 }, { "grad_norm": 0.44022148847579956, "learning_rate": 8.998026146767487e-05, "loss": 0.054, "step": 24480 }, { "grad_norm": 0.8470253944396973, "learning_rate": 8.99703297656008e-05, "loss": 0.0741, "step": 24490 }, { "grad_norm": 0.5269303321838379, "learning_rate": 8.996039369243156e-05, "loss": 0.071, "step": 24500 }, { "grad_norm": 0.4949968457221985, "learning_rate": 8.995045324925378e-05, "loss": 0.0556, "step": 24510 }, { "grad_norm": 1.013107180595398, "learning_rate": 8.994050843715448e-05, "loss": 0.0741, "step": 24520 }, { "grad_norm": 0.6601544618606567, "learning_rate": 8.993055925722121e-05, "loss": 0.0594, "step": 24530 }, { "grad_norm": 0.49490442872047424, "learning_rate": 8.992060571054202e-05, "loss": 0.0713, "step": 24540 }, { "grad_norm": 0.558987557888031, "learning_rate": 8.991064779820542e-05, "loss": 0.0673, "step": 24550 }, { "grad_norm": 0.5020607113838196, "learning_rate": 8.990068552130036e-05, "loss": 0.0711, "step": 24560 }, { "grad_norm": 0.6138238906860352, "learning_rate": 8.989071888091634e-05, "loss": 0.0697, "step": 24570 }, { "grad_norm": 0.6060677766799927, "learning_rate": 8.988074787814329e-05, "loss": 0.0594, "step": 24580 }, { "grad_norm": 0.9281871318817139, "learning_rate": 8.987077251407158e-05, "loss": 0.0647, "step": 24590 }, { "grad_norm": 0.5124716758728027, "learning_rate": 8.986079278979216e-05, "loss": 0.0649, "step": 24600 }, { "grad_norm": 0.8450791835784912, "learning_rate": 8.985080870639635e-05, "loss": 0.0789, "step": 24610 }, { "grad_norm": 0.5546930432319641, "learning_rate": 8.984082026497603e-05, "loss": 0.0583, "step": 24620 }, { "grad_norm": 0.586409330368042, "learning_rate": 8.98308274666235e-05, "loss": 0.0765, "step": 24630 }, { "grad_norm": 0.478360116481781, "learning_rate": 8.982083031243155e-05, "loss": 0.0641, "step": 24640 }, { "grad_norm": 0.5618711113929749, "learning_rate": 8.98108288034935e-05, "loss": 0.0623, "step": 24650 }, { "grad_norm": 0.6673212647438049, "learning_rate": 8.980082294090305e-05, "loss": 0.073, "step": 24660 }, { "grad_norm": 0.6260270476341248, "learning_rate": 8.979081272575443e-05, "loss": 0.0801, "step": 24670 }, { "grad_norm": 0.5744110345840454, "learning_rate": 8.978079815914236e-05, "loss": 0.0785, "step": 24680 }, { "grad_norm": 0.6523445248603821, "learning_rate": 8.977077924216202e-05, "loss": 0.08, "step": 24690 }, { "grad_norm": 0.6605185270309448, "learning_rate": 8.976075597590905e-05, "loss": 0.078, "step": 24700 }, { "grad_norm": 0.5087756514549255, "learning_rate": 8.975072836147958e-05, "loss": 0.061, "step": 24710 }, { "grad_norm": 0.6633077263832092, "learning_rate": 8.974069639997025e-05, "loss": 0.0648, "step": 24720 }, { "grad_norm": 0.44740593433380127, "learning_rate": 8.973066009247808e-05, "loss": 0.0686, "step": 24730 }, { "grad_norm": 0.5814014077186584, "learning_rate": 8.972061944010066e-05, "loss": 0.0703, "step": 24740 }, { "grad_norm": 0.4551445245742798, "learning_rate": 8.971057444393603e-05, "loss": 0.0672, "step": 24750 }, { "grad_norm": 0.46391183137893677, "learning_rate": 8.970052510508268e-05, "loss": 0.0685, "step": 24760 }, { "grad_norm": 0.47318828105926514, "learning_rate": 8.969047142463959e-05, "loss": 0.0765, "step": 24770 }, { "grad_norm": 0.6260510087013245, "learning_rate": 8.968041340370621e-05, "loss": 0.0749, "step": 24780 }, { "grad_norm": 1.3486322164535522, "learning_rate": 8.96703510433825e-05, "loss": 0.0686, "step": 24790 }, { "grad_norm": 0.940890371799469, "learning_rate": 8.966028434476883e-05, "loss": 0.0725, "step": 24800 }, { "grad_norm": 0.6510806083679199, "learning_rate": 8.96502133089661e-05, "loss": 0.0666, "step": 24810 }, { "grad_norm": 0.9841303825378418, "learning_rate": 8.964013793707564e-05, "loss": 0.0708, "step": 24820 }, { "grad_norm": 0.5641397833824158, "learning_rate": 8.963005823019932e-05, "loss": 0.0749, "step": 24830 }, { "grad_norm": 0.5515604019165039, "learning_rate": 8.961997418943939e-05, "loss": 0.0675, "step": 24840 }, { "grad_norm": 0.5005578398704529, "learning_rate": 8.960988581589865e-05, "loss": 0.0707, "step": 24850 }, { "grad_norm": 0.6923040747642517, "learning_rate": 8.959979311068037e-05, "loss": 0.0577, "step": 24860 }, { "grad_norm": 0.5227382779121399, "learning_rate": 8.958969607488823e-05, "loss": 0.0837, "step": 24870 }, { "grad_norm": 0.7320078611373901, "learning_rate": 8.957959470962647e-05, "loss": 0.0672, "step": 24880 }, { "grad_norm": 0.6237249970436096, "learning_rate": 8.956948901599971e-05, "loss": 0.0724, "step": 24890 }, { "grad_norm": 0.5318652987480164, "learning_rate": 8.955937899511315e-05, "loss": 0.0678, "step": 24900 }, { "grad_norm": 0.7321961522102356, "learning_rate": 8.954926464807238e-05, "loss": 0.0608, "step": 24910 }, { "grad_norm": 0.6523841023445129, "learning_rate": 8.953914597598347e-05, "loss": 0.0672, "step": 24920 }, { "grad_norm": 0.5915120244026184, "learning_rate": 8.952902297995303e-05, "loss": 0.0671, "step": 24930 }, { "grad_norm": 0.4446980357170105, "learning_rate": 8.951889566108804e-05, "loss": 0.0791, "step": 24940 }, { "grad_norm": 0.7745651006698608, "learning_rate": 8.950876402049606e-05, "loss": 0.0741, "step": 24950 }, { "grad_norm": 0.8403397798538208, "learning_rate": 8.949862805928504e-05, "loss": 0.0635, "step": 24960 }, { "grad_norm": 0.5406023263931274, "learning_rate": 8.948848777856343e-05, "loss": 0.0691, "step": 24970 }, { "grad_norm": 1.1009830236434937, "learning_rate": 8.947834317944017e-05, "loss": 0.0793, "step": 24980 }, { "grad_norm": 0.49969956278800964, "learning_rate": 8.946819426302466e-05, "loss": 0.0795, "step": 24990 }, { "grad_norm": 0.5630857944488525, "learning_rate": 8.945804103042676e-05, "loss": 0.0586, "step": 25000 }, { "grad_norm": 0.6787123084068298, "learning_rate": 8.944788348275681e-05, "loss": 0.0694, "step": 25010 }, { "grad_norm": 0.6768282651901245, "learning_rate": 8.943772162112565e-05, "loss": 0.0768, "step": 25020 }, { "grad_norm": 0.7509788870811462, "learning_rate": 8.942755544664454e-05, "loss": 0.0653, "step": 25030 }, { "grad_norm": 0.7596451640129089, "learning_rate": 8.941738496042525e-05, "loss": 0.0672, "step": 25040 }, { "grad_norm": 0.512336790561676, "learning_rate": 8.940721016357999e-05, "loss": 0.0736, "step": 25050 }, { "grad_norm": 0.5261013507843018, "learning_rate": 8.939703105722148e-05, "loss": 0.0741, "step": 25060 }, { "grad_norm": 0.6555437445640564, "learning_rate": 8.93868476424629e-05, "loss": 0.0668, "step": 25070 }, { "grad_norm": 0.7358865141868591, "learning_rate": 8.937665992041786e-05, "loss": 0.0696, "step": 25080 }, { "grad_norm": 2.3693082332611084, "learning_rate": 8.93664678922005e-05, "loss": 0.0704, "step": 25090 }, { "grad_norm": 0.7477633357048035, "learning_rate": 8.93562715589254e-05, "loss": 0.0635, "step": 25100 }, { "grad_norm": 0.4290209412574768, "learning_rate": 8.934607092170762e-05, "loss": 0.0697, "step": 25110 }, { "grad_norm": 0.5877124071121216, "learning_rate": 8.933586598166266e-05, "loss": 0.0755, "step": 25120 }, { "grad_norm": 1.9414174556732178, "learning_rate": 8.932565673990655e-05, "loss": 0.0823, "step": 25130 }, { "grad_norm": 0.48882949352264404, "learning_rate": 8.931544319755574e-05, "loss": 0.0656, "step": 25140 }, { "grad_norm": 0.5139110684394836, "learning_rate": 8.930522535572718e-05, "loss": 0.068, "step": 25150 }, { "grad_norm": 0.5766383409500122, "learning_rate": 8.929500321553826e-05, "loss": 0.0671, "step": 25160 }, { "grad_norm": 0.5368491411209106, "learning_rate": 8.928477677810686e-05, "loss": 0.0715, "step": 25170 }, { "grad_norm": 0.5115478038787842, "learning_rate": 8.927454604455137e-05, "loss": 0.0717, "step": 25180 }, { "grad_norm": 0.5670711994171143, "learning_rate": 8.926431101599053e-05, "loss": 0.0754, "step": 25190 }, { "grad_norm": 0.4880731999874115, "learning_rate": 8.925407169354369e-05, "loss": 0.0652, "step": 25200 }, { "grad_norm": 0.618969738483429, "learning_rate": 8.92438280783306e-05, "loss": 0.0924, "step": 25210 }, { "grad_norm": 0.6917199492454529, "learning_rate": 8.923358017147146e-05, "loss": 0.0846, "step": 25220 }, { "grad_norm": 0.490884006023407, "learning_rate": 8.922332797408697e-05, "loss": 0.0774, "step": 25230 }, { "grad_norm": 0.5402829051017761, "learning_rate": 8.921307148729831e-05, "loss": 0.0674, "step": 25240 }, { "grad_norm": 0.7879059910774231, "learning_rate": 8.920281071222712e-05, "loss": 0.0653, "step": 25250 }, { "grad_norm": 0.5028330087661743, "learning_rate": 8.919254564999548e-05, "loss": 0.0583, "step": 25260 }, { "grad_norm": 0.5869463086128235, "learning_rate": 8.918227630172598e-05, "loss": 0.0559, "step": 25270 }, { "grad_norm": 0.5636641979217529, "learning_rate": 8.917200266854165e-05, "loss": 0.0617, "step": 25280 }, { "grad_norm": 0.48663005232810974, "learning_rate": 8.9161724751566e-05, "loss": 0.063, "step": 25290 }, { "grad_norm": 0.42904192209243774, "learning_rate": 8.915144255192302e-05, "loss": 0.0845, "step": 25300 }, { "grad_norm": 0.789818286895752, "learning_rate": 8.914115607073714e-05, "loss": 0.0816, "step": 25310 }, { "grad_norm": 0.9868838787078857, "learning_rate": 8.913086530913327e-05, "loss": 0.069, "step": 25320 }, { "grad_norm": 0.391253262758255, "learning_rate": 8.912057026823681e-05, "loss": 0.0747, "step": 25330 }, { "grad_norm": 0.6564717292785645, "learning_rate": 8.91102709491736e-05, "loss": 0.079, "step": 25340 }, { "grad_norm": 0.3920595943927765, "learning_rate": 8.909996735306996e-05, "loss": 0.0697, "step": 25350 }, { "grad_norm": 0.6344684958457947, "learning_rate": 8.908965948105268e-05, "loss": 0.0688, "step": 25360 }, { "grad_norm": 0.607720136642456, "learning_rate": 8.907934733424901e-05, "loss": 0.0664, "step": 25370 }, { "grad_norm": 0.5137752294540405, "learning_rate": 8.906903091378666e-05, "loss": 0.0707, "step": 25380 }, { "grad_norm": 0.500272274017334, "learning_rate": 8.905871022079384e-05, "loss": 0.0721, "step": 25390 }, { "grad_norm": 0.46093374490737915, "learning_rate": 8.90483852563992e-05, "loss": 0.0697, "step": 25400 }, { "grad_norm": 0.5576540231704712, "learning_rate": 8.903805602173185e-05, "loss": 0.0643, "step": 25410 }, { "grad_norm": 0.4748271703720093, "learning_rate": 8.902772251792137e-05, "loss": 0.0653, "step": 25420 }, { "grad_norm": 0.6372000575065613, "learning_rate": 8.901738474609786e-05, "loss": 0.072, "step": 25430 }, { "grad_norm": 1.0231540203094482, "learning_rate": 8.900704270739179e-05, "loss": 0.0744, "step": 25440 }, { "grad_norm": 0.6008325815200806, "learning_rate": 8.89966964029342e-05, "loss": 0.0809, "step": 25450 }, { "grad_norm": 0.5445907115936279, "learning_rate": 8.898634583385652e-05, "loss": 0.0509, "step": 25460 }, { "grad_norm": 0.5752477645874023, "learning_rate": 8.897599100129065e-05, "loss": 0.0765, "step": 25470 }, { "grad_norm": 0.44000670313835144, "learning_rate": 8.896563190636903e-05, "loss": 0.06, "step": 25480 }, { "grad_norm": 1.012205958366394, "learning_rate": 8.895526855022448e-05, "loss": 0.0687, "step": 25490 }, { "grad_norm": 0.5207434296607971, "learning_rate": 8.894490093399033e-05, "loss": 0.0786, "step": 25500 }, { "grad_norm": 0.4551694393157959, "learning_rate": 8.893452905880035e-05, "loss": 0.0693, "step": 25510 }, { "grad_norm": 0.6975159645080566, "learning_rate": 8.892415292578883e-05, "loss": 0.0764, "step": 25520 }, { "grad_norm": 0.5432602167129517, "learning_rate": 8.891377253609046e-05, "loss": 0.0645, "step": 25530 }, { "grad_norm": 0.49642452597618103, "learning_rate": 8.890338789084043e-05, "loss": 0.0634, "step": 25540 }, { "grad_norm": 0.8859421014785767, "learning_rate": 8.88929989911744e-05, "loss": 0.0591, "step": 25550 }, { "grad_norm": 0.7280784845352173, "learning_rate": 8.888260583822847e-05, "loss": 0.0737, "step": 25560 }, { "grad_norm": 0.45391029119491577, "learning_rate": 8.887220843313921e-05, "loss": 0.0674, "step": 25570 }, { "grad_norm": 0.5965172052383423, "learning_rate": 8.88618067770437e-05, "loss": 0.0715, "step": 25580 }, { "grad_norm": 0.5366000533103943, "learning_rate": 8.885140087107942e-05, "loss": 0.0711, "step": 25590 }, { "grad_norm": 0.3925701081752777, "learning_rate": 8.884099071638436e-05, "loss": 0.0694, "step": 25600 }, { "grad_norm": 0.5584710836410522, "learning_rate": 8.883057631409695e-05, "loss": 0.0809, "step": 25610 }, { "grad_norm": 0.7892029285430908, "learning_rate": 8.882015766535608e-05, "loss": 0.0695, "step": 25620 }, { "grad_norm": 0.770428478717804, "learning_rate": 8.880973477130115e-05, "loss": 0.0713, "step": 25630 }, { "grad_norm": 0.5827586054801941, "learning_rate": 8.879930763307197e-05, "loss": 0.0567, "step": 25640 }, { "grad_norm": 0.5463842153549194, "learning_rate": 8.878887625180884e-05, "loss": 0.0767, "step": 25650 }, { "grad_norm": 0.9607470035552979, "learning_rate": 8.877844062865253e-05, "loss": 0.0601, "step": 25660 }, { "grad_norm": 0.624697744846344, "learning_rate": 8.876800076474424e-05, "loss": 0.0654, "step": 25670 }, { "grad_norm": 0.6233333349227905, "learning_rate": 8.875755666122568e-05, "loss": 0.0588, "step": 25680 }, { "grad_norm": 0.3627215027809143, "learning_rate": 8.8747108319239e-05, "loss": 0.058, "step": 25690 }, { "grad_norm": 0.49674174189567566, "learning_rate": 8.87366557399268e-05, "loss": 0.0579, "step": 25700 }, { "grad_norm": 0.6185820698738098, "learning_rate": 8.872619892443217e-05, "loss": 0.0744, "step": 25710 }, { "grad_norm": 0.4489392042160034, "learning_rate": 8.871573787389865e-05, "loss": 0.0546, "step": 25720 }, { "grad_norm": 0.5194432139396667, "learning_rate": 8.870527258947024e-05, "loss": 0.069, "step": 25730 }, { "grad_norm": 0.4506649971008301, "learning_rate": 8.869480307229143e-05, "loss": 0.0478, "step": 25740 }, { "grad_norm": 0.6084566712379456, "learning_rate": 8.868432932350712e-05, "loss": 0.0558, "step": 25750 }, { "grad_norm": 0.6853962540626526, "learning_rate": 8.867385134426272e-05, "loss": 0.074, "step": 25760 }, { "grad_norm": 0.45347675681114197, "learning_rate": 8.866336913570407e-05, "loss": 0.0585, "step": 25770 }, { "grad_norm": 0.4691256582736969, "learning_rate": 8.865288269897751e-05, "loss": 0.0523, "step": 25780 }, { "grad_norm": 1.6430861949920654, "learning_rate": 8.864239203522981e-05, "loss": 0.0542, "step": 25790 }, { "grad_norm": 0.4218238592147827, "learning_rate": 8.863189714560822e-05, "loss": 0.0624, "step": 25800 }, { "grad_norm": 0.46750184893608093, "learning_rate": 8.862139803126043e-05, "loss": 0.0532, "step": 25810 }, { "grad_norm": 0.35286951065063477, "learning_rate": 8.861089469333463e-05, "loss": 0.0654, "step": 25820 }, { "grad_norm": 0.3817032277584076, "learning_rate": 8.860038713297944e-05, "loss": 0.0479, "step": 25830 }, { "grad_norm": 0.43371158838272095, "learning_rate": 8.858987535134394e-05, "loss": 0.0648, "step": 25840 }, { "grad_norm": 0.4235866963863373, "learning_rate": 8.857935934957769e-05, "loss": 0.0579, "step": 25850 }, { "grad_norm": 0.44315797090530396, "learning_rate": 8.856883912883071e-05, "loss": 0.0586, "step": 25860 }, { "grad_norm": 0.4763559401035309, "learning_rate": 8.855831469025346e-05, "loss": 0.0568, "step": 25870 }, { "grad_norm": 0.5011365413665771, "learning_rate": 8.854778603499689e-05, "loss": 0.0847, "step": 25880 }, { "grad_norm": 0.9246358871459961, "learning_rate": 8.85372531642124e-05, "loss": 0.0587, "step": 25890 }, { "grad_norm": 0.4478399157524109, "learning_rate": 8.852671607905185e-05, "loss": 0.0569, "step": 25900 }, { "grad_norm": 0.707643985748291, "learning_rate": 8.851617478066754e-05, "loss": 0.0742, "step": 25910 }, { "grad_norm": 0.8090428113937378, "learning_rate": 8.850562927021227e-05, "loss": 0.0653, "step": 25920 }, { "grad_norm": 0.3791167140007019, "learning_rate": 8.849507954883928e-05, "loss": 0.0617, "step": 25930 }, { "grad_norm": 0.6015363335609436, "learning_rate": 8.848452561770226e-05, "loss": 0.0711, "step": 25940 }, { "grad_norm": 0.5325243473052979, "learning_rate": 8.847396747795538e-05, "loss": 0.0584, "step": 25950 }, { "grad_norm": 0.43408775329589844, "learning_rate": 8.846340513075327e-05, "loss": 0.0654, "step": 25960 }, { "grad_norm": 1.041923999786377, "learning_rate": 8.845283857725099e-05, "loss": 0.0661, "step": 25970 }, { "grad_norm": 0.7331128716468811, "learning_rate": 8.844226781860409e-05, "loss": 0.0676, "step": 25980 }, { "grad_norm": 0.6678151488304138, "learning_rate": 8.84316928559686e-05, "loss": 0.0795, "step": 25990 }, { "grad_norm": 1.163169264793396, "learning_rate": 8.842111369050094e-05, "loss": 0.0787, "step": 26000 }, { "grad_norm": 0.4725159704685211, "learning_rate": 8.841053032335808e-05, "loss": 0.0591, "step": 26010 }, { "grad_norm": 0.505184531211853, "learning_rate": 8.839994275569735e-05, "loss": 0.0678, "step": 26020 }, { "grad_norm": 0.6028842926025391, "learning_rate": 8.838935098867662e-05, "loss": 0.0624, "step": 26030 }, { "grad_norm": 0.6118471026420593, "learning_rate": 8.837875502345418e-05, "loss": 0.0825, "step": 26040 }, { "grad_norm": 0.742149829864502, "learning_rate": 8.83681548611888e-05, "loss": 0.0831, "step": 26050 }, { "grad_norm": 0.6987989544868469, "learning_rate": 8.835755050303969e-05, "loss": 0.0648, "step": 26060 }, { "grad_norm": 0.7084911465644836, "learning_rate": 8.834694195016653e-05, "loss": 0.0734, "step": 26070 }, { "grad_norm": 0.6850001215934753, "learning_rate": 8.833632920372942e-05, "loss": 0.0933, "step": 26080 }, { "grad_norm": 0.73182213306427, "learning_rate": 8.832571226488903e-05, "loss": 0.0712, "step": 26090 }, { "grad_norm": 0.5694886445999146, "learning_rate": 8.831509113480634e-05, "loss": 0.0635, "step": 26100 }, { "grad_norm": 0.6480735540390015, "learning_rate": 8.83044658146429e-05, "loss": 0.0789, "step": 26110 }, { "grad_norm": 0.5057871341705322, "learning_rate": 8.829383630556067e-05, "loss": 0.0568, "step": 26120 }, { "grad_norm": 0.5673575401306152, "learning_rate": 8.828320260872207e-05, "loss": 0.0752, "step": 26130 }, { "grad_norm": 0.5603436231613159, "learning_rate": 8.827256472529e-05, "loss": 0.0585, "step": 26140 }, { "grad_norm": 0.5709559917449951, "learning_rate": 8.826192265642778e-05, "loss": 0.0671, "step": 26150 }, { "grad_norm": 0.4593040943145752, "learning_rate": 8.825127640329923e-05, "loss": 0.0605, "step": 26160 }, { "grad_norm": 0.8776322603225708, "learning_rate": 8.824062596706861e-05, "loss": 0.0678, "step": 26170 }, { "grad_norm": 0.5943191051483154, "learning_rate": 8.822997134890062e-05, "loss": 0.0742, "step": 26180 }, { "grad_norm": 0.8475016355514526, "learning_rate": 8.821931254996044e-05, "loss": 0.068, "step": 26190 }, { "grad_norm": 0.5641717314720154, "learning_rate": 8.82086495714137e-05, "loss": 0.0807, "step": 26200 }, { "grad_norm": 0.6628923416137695, "learning_rate": 8.81979824144265e-05, "loss": 0.0794, "step": 26210 }, { "grad_norm": 0.5394722819328308, "learning_rate": 8.818731108016536e-05, "loss": 0.0752, "step": 26220 }, { "grad_norm": 0.8586941361427307, "learning_rate": 8.81766355697973e-05, "loss": 0.082, "step": 26230 }, { "grad_norm": 0.9208213686943054, "learning_rate": 8.816595588448977e-05, "loss": 0.0776, "step": 26240 }, { "grad_norm": 0.7482990026473999, "learning_rate": 8.81552720254107e-05, "loss": 0.0851, "step": 26250 }, { "grad_norm": 0.7203786969184875, "learning_rate": 8.814458399372842e-05, "loss": 0.0905, "step": 26260 }, { "grad_norm": 1.1740182638168335, "learning_rate": 8.813389179061181e-05, "loss": 0.0706, "step": 26270 }, { "grad_norm": 0.5924995541572571, "learning_rate": 8.812319541723012e-05, "loss": 0.0866, "step": 26280 }, { "grad_norm": 0.5324624180793762, "learning_rate": 8.811249487475309e-05, "loss": 0.0801, "step": 26290 }, { "grad_norm": 0.8027909994125366, "learning_rate": 8.810179016435092e-05, "loss": 0.0774, "step": 26300 }, { "grad_norm": 0.5562375783920288, "learning_rate": 8.809108128719428e-05, "loss": 0.0889, "step": 26310 }, { "grad_norm": 0.6337056159973145, "learning_rate": 8.808036824445424e-05, "loss": 0.0725, "step": 26320 }, { "grad_norm": 0.510868489742279, "learning_rate": 8.806965103730238e-05, "loss": 0.0737, "step": 26330 }, { "grad_norm": 0.6325877904891968, "learning_rate": 8.805892966691074e-05, "loss": 0.0637, "step": 26340 }, { "grad_norm": 0.5785706043243408, "learning_rate": 8.804820413445175e-05, "loss": 0.0641, "step": 26350 }, { "grad_norm": 0.4969007968902588, "learning_rate": 8.803747444109837e-05, "loss": 0.078, "step": 26360 }, { "grad_norm": 0.648205578327179, "learning_rate": 8.802674058802399e-05, "loss": 0.0535, "step": 26370 }, { "grad_norm": 0.5883035659790039, "learning_rate": 8.801600257640241e-05, "loss": 0.0698, "step": 26380 }, { "grad_norm": 0.6222942471504211, "learning_rate": 8.800526040740795e-05, "loss": 0.075, "step": 26390 }, { "grad_norm": 0.6722877025604248, "learning_rate": 8.799451408221535e-05, "loss": 0.0708, "step": 26400 }, { "grad_norm": 0.8004640340805054, "learning_rate": 8.798376360199982e-05, "loss": 0.0717, "step": 26410 }, { "grad_norm": 0.43244174122810364, "learning_rate": 8.797300896793701e-05, "loss": 0.0619, "step": 26420 }, { "grad_norm": 0.962490975856781, "learning_rate": 8.796225018120302e-05, "loss": 0.07, "step": 26430 }, { "grad_norm": 0.5857201218605042, "learning_rate": 8.795148724297444e-05, "loss": 0.0665, "step": 26440 }, { "grad_norm": 0.5217369794845581, "learning_rate": 8.794072015442825e-05, "loss": 0.0751, "step": 26450 }, { "grad_norm": 0.6086981296539307, "learning_rate": 8.792994891674198e-05, "loss": 0.065, "step": 26460 }, { "grad_norm": 1.0166343450546265, "learning_rate": 8.79191735310935e-05, "loss": 0.0643, "step": 26470 }, { "grad_norm": 0.6260571479797363, "learning_rate": 8.790839399866122e-05, "loss": 0.0573, "step": 26480 }, { "grad_norm": 0.678663432598114, "learning_rate": 8.789761032062397e-05, "loss": 0.06, "step": 26490 }, { "grad_norm": 0.5133363008499146, "learning_rate": 8.788682249816103e-05, "loss": 0.0585, "step": 26500 }, { "grad_norm": 0.5658761262893677, "learning_rate": 8.787603053245215e-05, "loss": 0.0755, "step": 26510 }, { "grad_norm": 0.6781460642814636, "learning_rate": 8.78652344246775e-05, "loss": 0.0677, "step": 26520 }, { "grad_norm": 0.5662279725074768, "learning_rate": 8.785443417601776e-05, "loss": 0.0657, "step": 26530 }, { "grad_norm": 0.638142466545105, "learning_rate": 8.784362978765401e-05, "loss": 0.0544, "step": 26540 }, { "grad_norm": 0.7392207980155945, "learning_rate": 8.783282126076779e-05, "loss": 0.0595, "step": 26550 }, { "grad_norm": 0.6087008118629456, "learning_rate": 8.782200859654112e-05, "loss": 0.0629, "step": 26560 }, { "grad_norm": 0.6471968293190002, "learning_rate": 8.781119179615646e-05, "loss": 0.0622, "step": 26570 }, { "grad_norm": 0.6490806937217712, "learning_rate": 8.780037086079674e-05, "loss": 0.069, "step": 26580 }, { "grad_norm": 0.5824667811393738, "learning_rate": 8.778954579164527e-05, "loss": 0.0721, "step": 26590 }, { "grad_norm": 0.6805433034896851, "learning_rate": 8.777871658988588e-05, "loss": 0.0663, "step": 26600 }, { "grad_norm": 0.6053360104560852, "learning_rate": 8.776788325670285e-05, "loss": 0.0633, "step": 26610 }, { "grad_norm": 1.850480318069458, "learning_rate": 8.775704579328089e-05, "loss": 0.0658, "step": 26620 }, { "grad_norm": 0.7773674130439758, "learning_rate": 8.774620420080517e-05, "loss": 0.0689, "step": 26630 }, { "grad_norm": 0.516532301902771, "learning_rate": 8.773535848046131e-05, "loss": 0.0563, "step": 26640 }, { "grad_norm": 0.6262345314025879, "learning_rate": 8.772450863343538e-05, "loss": 0.0801, "step": 26650 }, { "grad_norm": 0.8200851678848267, "learning_rate": 8.77136546609139e-05, "loss": 0.0878, "step": 26660 }, { "grad_norm": 1.5589559078216553, "learning_rate": 8.770279656408385e-05, "loss": 0.0632, "step": 26670 }, { "grad_norm": 0.5321199297904968, "learning_rate": 8.769193434413265e-05, "loss": 0.0652, "step": 26680 }, { "grad_norm": 0.5711780786514282, "learning_rate": 8.76810680022482e-05, "loss": 0.0644, "step": 26690 }, { "grad_norm": 0.5386244654655457, "learning_rate": 8.767019753961878e-05, "loss": 0.0761, "step": 26700 }, { "grad_norm": 0.8170008063316345, "learning_rate": 8.765932295743321e-05, "loss": 0.0654, "step": 26710 }, { "grad_norm": 0.5388513803482056, "learning_rate": 8.764844425688068e-05, "loss": 0.0709, "step": 26720 }, { "grad_norm": 0.7621394395828247, "learning_rate": 8.763756143915092e-05, "loss": 0.0602, "step": 26730 }, { "grad_norm": 0.6403313875198364, "learning_rate": 8.7626674505434e-05, "loss": 0.0753, "step": 26740 }, { "grad_norm": 0.5036414265632629, "learning_rate": 8.761578345692053e-05, "loss": 0.074, "step": 26750 }, { "grad_norm": 0.5687412023544312, "learning_rate": 8.760488829480156e-05, "loss": 0.0674, "step": 26760 }, { "grad_norm": 0.5710969567298889, "learning_rate": 8.759398902026854e-05, "loss": 0.0559, "step": 26770 }, { "grad_norm": 0.5662939548492432, "learning_rate": 8.758308563451339e-05, "loss": 0.074, "step": 26780 }, { "grad_norm": 0.6317121386528015, "learning_rate": 8.75721781387285e-05, "loss": 0.0749, "step": 26790 }, { "grad_norm": 0.655112624168396, "learning_rate": 8.75612665341067e-05, "loss": 0.0664, "step": 26800 }, { "grad_norm": 0.8076235055923462, "learning_rate": 8.755035082184126e-05, "loss": 0.0633, "step": 26810 }, { "grad_norm": 0.457131564617157, "learning_rate": 8.753943100312592e-05, "loss": 0.0708, "step": 26820 }, { "grad_norm": 0.43306419253349304, "learning_rate": 8.752850707915484e-05, "loss": 0.0548, "step": 26830 }, { "grad_norm": 0.8432632088661194, "learning_rate": 8.751757905112264e-05, "loss": 0.0863, "step": 26840 }, { "grad_norm": 0.659329891204834, "learning_rate": 8.75066469202244e-05, "loss": 0.0723, "step": 26850 }, { "grad_norm": 0.6231181025505066, "learning_rate": 8.749571068765567e-05, "loss": 0.0596, "step": 26860 }, { "grad_norm": 0.41967475414276123, "learning_rate": 8.748477035461238e-05, "loss": 0.0696, "step": 26870 }, { "grad_norm": 0.5496764779090881, "learning_rate": 8.747382592229095e-05, "loss": 0.0676, "step": 26880 }, { "grad_norm": 0.5658892393112183, "learning_rate": 8.746287739188828e-05, "loss": 0.0807, "step": 26890 }, { "grad_norm": 0.46106913685798645, "learning_rate": 8.745192476460165e-05, "loss": 0.0617, "step": 26900 }, { "grad_norm": 0.7922160029411316, "learning_rate": 8.744096804162882e-05, "loss": 0.0769, "step": 26910 }, { "grad_norm": 0.7176406979560852, "learning_rate": 8.743000722416804e-05, "loss": 0.0697, "step": 26920 }, { "grad_norm": 0.4715535342693329, "learning_rate": 8.741904231341793e-05, "loss": 0.0655, "step": 26930 }, { "grad_norm": 0.7089669108390808, "learning_rate": 8.740807331057762e-05, "loss": 0.0832, "step": 26940 }, { "grad_norm": 0.5087477564811707, "learning_rate": 8.739710021684667e-05, "loss": 0.0681, "step": 26950 }, { "grad_norm": 0.9545490741729736, "learning_rate": 8.738612303342503e-05, "loss": 0.0751, "step": 26960 }, { "grad_norm": 0.9390375018119812, "learning_rate": 8.73751417615132e-05, "loss": 0.0703, "step": 26970 }, { "grad_norm": 0.3234761357307434, "learning_rate": 8.736415640231208e-05, "loss": 0.0587, "step": 26980 }, { "grad_norm": 0.5299890637397766, "learning_rate": 8.735316695702297e-05, "loss": 0.0578, "step": 26990 }, { "grad_norm": 0.4237622022628784, "learning_rate": 8.734217342684769e-05, "loss": 0.0594, "step": 27000 }, { "grad_norm": 0.5525263547897339, "learning_rate": 8.733117581298847e-05, "loss": 0.0573, "step": 27010 }, { "grad_norm": 0.541837751865387, "learning_rate": 8.732017411664796e-05, "loss": 0.063, "step": 27020 }, { "grad_norm": 0.9543559551239014, "learning_rate": 8.730916833902936e-05, "loss": 0.0605, "step": 27030 }, { "grad_norm": 0.3862018883228302, "learning_rate": 8.729815848133618e-05, "loss": 0.0761, "step": 27040 }, { "grad_norm": 0.6103121638298035, "learning_rate": 8.728714454477247e-05, "loss": 0.0507, "step": 27050 }, { "grad_norm": 0.5325585007667542, "learning_rate": 8.727612653054269e-05, "loss": 0.0663, "step": 27060 }, { "grad_norm": 0.6545608639717102, "learning_rate": 8.726510443985176e-05, "loss": 0.0564, "step": 27070 }, { "grad_norm": 0.7580714225769043, "learning_rate": 8.725407827390503e-05, "loss": 0.0695, "step": 27080 }, { "grad_norm": 0.6217553019523621, "learning_rate": 8.724304803390833e-05, "loss": 0.0772, "step": 27090 }, { "grad_norm": 0.3944689929485321, "learning_rate": 8.723201372106788e-05, "loss": 0.0565, "step": 27100 }, { "grad_norm": 0.8000084757804871, "learning_rate": 8.722097533659038e-05, "loss": 0.0641, "step": 27110 }, { "grad_norm": 0.5955978631973267, "learning_rate": 8.720993288168299e-05, "loss": 0.0662, "step": 27120 }, { "grad_norm": 0.6297118067741394, "learning_rate": 8.719888635755327e-05, "loss": 0.0573, "step": 27130 }, { "grad_norm": 0.5251537561416626, "learning_rate": 8.718783576540928e-05, "loss": 0.0644, "step": 27140 }, { "grad_norm": 0.8819257020950317, "learning_rate": 8.717678110645948e-05, "loss": 0.0816, "step": 27150 }, { "grad_norm": 0.6928134560585022, "learning_rate": 8.716572238191279e-05, "loss": 0.0634, "step": 27160 }, { "grad_norm": 0.4998243451118469, "learning_rate": 8.715465959297857e-05, "loss": 0.0711, "step": 27170 }, { "grad_norm": 0.6364076733589172, "learning_rate": 8.714359274086665e-05, "loss": 0.0797, "step": 27180 }, { "grad_norm": 0.4652683138847351, "learning_rate": 8.713252182678726e-05, "loss": 0.0597, "step": 27190 }, { "grad_norm": 0.4916773736476898, "learning_rate": 8.712144685195112e-05, "loss": 0.0906, "step": 27200 }, { "grad_norm": 0.5786927342414856, "learning_rate": 8.711036781756936e-05, "loss": 0.0679, "step": 27210 }, { "grad_norm": 0.6148709654808044, "learning_rate": 8.709928472485357e-05, "loss": 0.0657, "step": 27220 }, { "grad_norm": 0.4373016059398651, "learning_rate": 8.708819757501579e-05, "loss": 0.0533, "step": 27230 }, { "grad_norm": 0.4881892502307892, "learning_rate": 8.707710636926846e-05, "loss": 0.061, "step": 27240 }, { "grad_norm": 0.6268231272697449, "learning_rate": 8.706601110882455e-05, "loss": 0.0766, "step": 27250 }, { "grad_norm": 0.5852571725845337, "learning_rate": 8.705491179489738e-05, "loss": 0.0615, "step": 27260 }, { "grad_norm": 0.5413491725921631, "learning_rate": 8.704380842870077e-05, "loss": 0.0664, "step": 27270 }, { "grad_norm": 0.4976763427257538, "learning_rate": 8.703270101144895e-05, "loss": 0.0618, "step": 27280 }, { "grad_norm": 0.3963862657546997, "learning_rate": 8.702158954435664e-05, "loss": 0.0665, "step": 27290 }, { "grad_norm": 2.495596408843994, "learning_rate": 8.701047402863896e-05, "loss": 0.0638, "step": 27300 }, { "grad_norm": 1.0609592199325562, "learning_rate": 8.699935446551148e-05, "loss": 0.0768, "step": 27310 }, { "grad_norm": 0.5673977732658386, "learning_rate": 8.698823085619022e-05, "loss": 0.0617, "step": 27320 }, { "grad_norm": 0.5892161130905151, "learning_rate": 8.697710320189166e-05, "loss": 0.0507, "step": 27330 }, { "grad_norm": 0.757548987865448, "learning_rate": 8.696597150383268e-05, "loss": 0.0633, "step": 27340 }, { "grad_norm": 0.5665348768234253, "learning_rate": 8.695483576323063e-05, "loss": 0.0686, "step": 27350 }, { "grad_norm": 0.8522955775260925, "learning_rate": 8.69436959813033e-05, "loss": 0.0729, "step": 27360 }, { "grad_norm": 0.4643760621547699, "learning_rate": 8.693255215926892e-05, "loss": 0.0711, "step": 27370 }, { "grad_norm": 0.6870155334472656, "learning_rate": 8.692140429834617e-05, "loss": 0.082, "step": 27380 }, { "grad_norm": 0.6063968539237976, "learning_rate": 8.691025239975415e-05, "loss": 0.0767, "step": 27390 }, { "grad_norm": 0.7789193391799927, "learning_rate": 8.689909646471243e-05, "loss": 0.0761, "step": 27400 }, { "grad_norm": 0.6684615612030029, "learning_rate": 8.688793649444099e-05, "loss": 0.0751, "step": 27410 }, { "grad_norm": 0.5062749981880188, "learning_rate": 8.687677249016029e-05, "loss": 0.0672, "step": 27420 }, { "grad_norm": 0.7071911692619324, "learning_rate": 8.686560445309118e-05, "loss": 0.072, "step": 27430 }, { "grad_norm": 0.8356448411941528, "learning_rate": 8.685443238445499e-05, "loss": 0.0842, "step": 27440 }, { "grad_norm": 0.4730253517627716, "learning_rate": 8.68432562854735e-05, "loss": 0.0839, "step": 27450 }, { "grad_norm": 0.5329523682594299, "learning_rate": 8.683207615736887e-05, "loss": 0.0689, "step": 27460 }, { "grad_norm": 1.0315521955490112, "learning_rate": 8.682089200136379e-05, "loss": 0.0788, "step": 27470 }, { "grad_norm": 0.5497618913650513, "learning_rate": 8.680970381868132e-05, "loss": 0.0714, "step": 27480 }, { "grad_norm": 0.4937511682510376, "learning_rate": 8.679851161054498e-05, "loss": 0.0512, "step": 27490 }, { "grad_norm": 0.5946037769317627, "learning_rate": 8.678731537817873e-05, "loss": 0.076, "step": 27500 }, { "grad_norm": 0.507576048374176, "learning_rate": 8.677611512280697e-05, "loss": 0.0762, "step": 27510 }, { "grad_norm": 0.8027248978614807, "learning_rate": 8.676491084565457e-05, "loss": 0.0809, "step": 27520 }, { "grad_norm": 0.4222778379917145, "learning_rate": 8.675370254794678e-05, "loss": 0.0835, "step": 27530 }, { "grad_norm": 0.437376469373703, "learning_rate": 8.674249023090935e-05, "loss": 0.0666, "step": 27540 }, { "grad_norm": 1.3058453798294067, "learning_rate": 8.673127389576843e-05, "loss": 0.0734, "step": 27550 }, { "grad_norm": 0.6317036151885986, "learning_rate": 8.67200535437506e-05, "loss": 0.0672, "step": 27560 }, { "grad_norm": 0.7142379879951477, "learning_rate": 8.670882917608296e-05, "loss": 0.0744, "step": 27570 }, { "grad_norm": 0.9153020977973938, "learning_rate": 8.669760079399292e-05, "loss": 0.0804, "step": 27580 }, { "grad_norm": 0.40644246339797974, "learning_rate": 8.668636839870845e-05, "loss": 0.0741, "step": 27590 }, { "grad_norm": 1.043641209602356, "learning_rate": 8.667513199145789e-05, "loss": 0.0833, "step": 27600 }, { "grad_norm": 0.6362689733505249, "learning_rate": 8.666389157347002e-05, "loss": 0.0794, "step": 27610 }, { "grad_norm": 0.5768572092056274, "learning_rate": 8.66526471459741e-05, "loss": 0.0771, "step": 27620 }, { "grad_norm": 0.5244368314743042, "learning_rate": 8.66413987101998e-05, "loss": 0.0802, "step": 27630 }, { "grad_norm": 0.5382435917854309, "learning_rate": 8.663014626737723e-05, "loss": 0.0687, "step": 27640 }, { "grad_norm": 0.5102038979530334, "learning_rate": 8.661888981873691e-05, "loss": 0.0777, "step": 27650 }, { "grad_norm": 0.6637563705444336, "learning_rate": 8.660762936550988e-05, "loss": 0.0786, "step": 27660 }, { "grad_norm": 0.6722251772880554, "learning_rate": 8.659636490892753e-05, "loss": 0.078, "step": 27670 }, { "grad_norm": 0.6223692893981934, "learning_rate": 8.658509645022174e-05, "loss": 0.0677, "step": 27680 }, { "grad_norm": 0.6796510219573975, "learning_rate": 8.657382399062481e-05, "loss": 0.0542, "step": 27690 }, { "grad_norm": 1.516087293624878, "learning_rate": 8.656254753136946e-05, "loss": 0.0667, "step": 27700 }, { "grad_norm": 0.6374524235725403, "learning_rate": 8.655126707368891e-05, "loss": 0.0554, "step": 27710 }, { "grad_norm": 0.5544005632400513, "learning_rate": 8.653998261881672e-05, "loss": 0.0685, "step": 27720 }, { "grad_norm": 0.5420922636985779, "learning_rate": 8.652869416798699e-05, "loss": 0.0644, "step": 27730 }, { "grad_norm": 0.6666470170021057, "learning_rate": 8.651740172243417e-05, "loss": 0.0758, "step": 27740 }, { "grad_norm": 0.4070053696632385, "learning_rate": 8.65061052833932e-05, "loss": 0.071, "step": 27750 }, { "grad_norm": 0.5062949061393738, "learning_rate": 8.649480485209945e-05, "loss": 0.0652, "step": 27760 }, { "grad_norm": 0.809837281703949, "learning_rate": 8.64835004297887e-05, "loss": 0.083, "step": 27770 }, { "grad_norm": 0.43970343470573425, "learning_rate": 8.64721920176972e-05, "loss": 0.0664, "step": 27780 }, { "grad_norm": 0.6611307263374329, "learning_rate": 8.646087961706164e-05, "loss": 0.0781, "step": 27790 }, { "grad_norm": 0.5641986727714539, "learning_rate": 8.644956322911908e-05, "loss": 0.0743, "step": 27800 }, { "grad_norm": 0.48887744545936584, "learning_rate": 8.643824285510709e-05, "loss": 0.0744, "step": 27810 }, { "grad_norm": 0.5227836966514587, "learning_rate": 8.642691849626364e-05, "loss": 0.0561, "step": 27820 }, { "grad_norm": 0.5393952131271362, "learning_rate": 8.641559015382717e-05, "loss": 0.0678, "step": 27830 }, { "grad_norm": 0.7378115057945251, "learning_rate": 8.640425782903649e-05, "loss": 0.0615, "step": 27840 }, { "grad_norm": 0.5269822478294373, "learning_rate": 8.639292152313091e-05, "loss": 0.0672, "step": 27850 }, { "grad_norm": 0.6007342338562012, "learning_rate": 8.638158123735015e-05, "loss": 0.0705, "step": 27860 }, { "grad_norm": 0.41696587204933167, "learning_rate": 8.637023697293436e-05, "loss": 0.0574, "step": 27870 }, { "grad_norm": 0.6142715811729431, "learning_rate": 8.635888873112414e-05, "loss": 0.0884, "step": 27880 }, { "grad_norm": 0.48746633529663086, "learning_rate": 8.634753651316052e-05, "loss": 0.0758, "step": 27890 }, { "grad_norm": 0.485826313495636, "learning_rate": 8.633618032028496e-05, "loss": 0.0636, "step": 27900 }, { "grad_norm": 0.6542759537696838, "learning_rate": 8.632482015373934e-05, "loss": 0.0689, "step": 27910 }, { "grad_norm": 0.6926559805870056, "learning_rate": 8.6313456014766e-05, "loss": 0.0656, "step": 27920 }, { "grad_norm": 0.7650174498558044, "learning_rate": 8.630208790460771e-05, "loss": 0.0644, "step": 27930 }, { "grad_norm": 0.9154366254806519, "learning_rate": 8.629071582450768e-05, "loss": 0.0751, "step": 27940 }, { "grad_norm": 0.6418443918228149, "learning_rate": 8.62793397757095e-05, "loss": 0.0725, "step": 27950 }, { "grad_norm": 0.5921344757080078, "learning_rate": 8.626795975945729e-05, "loss": 0.0655, "step": 27960 }, { "grad_norm": 0.753243088722229, "learning_rate": 8.625657577699551e-05, "loss": 0.0499, "step": 27970 }, { "grad_norm": 0.38074421882629395, "learning_rate": 8.624518782956914e-05, "loss": 0.059, "step": 27980 }, { "grad_norm": 0.6314891576766968, "learning_rate": 8.62337959184235e-05, "loss": 0.0689, "step": 27990 }, { "grad_norm": 0.6042144298553467, "learning_rate": 8.622240004480441e-05, "loss": 0.0738, "step": 28000 }, { "grad_norm": 0.5418338775634766, "learning_rate": 8.621100020995814e-05, "loss": 0.0613, "step": 28010 }, { "grad_norm": 0.7359590530395508, "learning_rate": 8.619959641513132e-05, "loss": 0.0772, "step": 28020 }, { "grad_norm": 0.6513711810112, "learning_rate": 8.618818866157105e-05, "loss": 0.0688, "step": 28030 }, { "grad_norm": 0.7985348105430603, "learning_rate": 8.617677695052487e-05, "loss": 0.0702, "step": 28040 }, { "grad_norm": 0.3760780096054077, "learning_rate": 8.616536128324078e-05, "loss": 0.0659, "step": 28050 }, { "grad_norm": 0.6570676565170288, "learning_rate": 8.615394166096712e-05, "loss": 0.0634, "step": 28060 }, { "grad_norm": 0.47540098428726196, "learning_rate": 8.614251808495279e-05, "loss": 0.0631, "step": 28070 }, { "grad_norm": 0.6205892562866211, "learning_rate": 8.6131090556447e-05, "loss": 0.0603, "step": 28080 }, { "grad_norm": 0.7915050387382507, "learning_rate": 8.611965907669947e-05, "loss": 0.0671, "step": 28090 }, { "grad_norm": 0.5496023893356323, "learning_rate": 8.610822364696034e-05, "loss": 0.0649, "step": 28100 }, { "grad_norm": 0.5436834692955017, "learning_rate": 8.609678426848015e-05, "loss": 0.0689, "step": 28110 }, { "grad_norm": 0.6179702281951904, "learning_rate": 8.60853409425099e-05, "loss": 0.0657, "step": 28120 }, { "grad_norm": 0.5722633600234985, "learning_rate": 8.607389367030104e-05, "loss": 0.063, "step": 28130 }, { "grad_norm": 0.7609144449234009, "learning_rate": 8.606244245310538e-05, "loss": 0.0623, "step": 28140 }, { "grad_norm": 0.8369426131248474, "learning_rate": 8.605098729217525e-05, "loss": 0.0696, "step": 28150 }, { "grad_norm": 0.6476860046386719, "learning_rate": 8.603952818876335e-05, "loss": 0.0716, "step": 28160 }, { "grad_norm": 0.7896018624305725, "learning_rate": 8.602806514412281e-05, "loss": 0.0607, "step": 28170 }, { "grad_norm": 0.6295753717422485, "learning_rate": 8.601659815950726e-05, "loss": 0.0768, "step": 28180 }, { "grad_norm": 0.7183090448379517, "learning_rate": 8.600512723617067e-05, "loss": 0.092, "step": 28190 }, { "grad_norm": 0.7595628499984741, "learning_rate": 8.59936523753675e-05, "loss": 0.0792, "step": 28200 }, { "grad_norm": 0.4802616536617279, "learning_rate": 8.598217357835264e-05, "loss": 0.0766, "step": 28210 }, { "grad_norm": 0.5430293679237366, "learning_rate": 8.597069084638135e-05, "loss": 0.0708, "step": 28220 }, { "grad_norm": 0.5390089154243469, "learning_rate": 8.595920418070939e-05, "loss": 0.0646, "step": 28230 }, { "grad_norm": 0.7228562831878662, "learning_rate": 8.594771358259295e-05, "loss": 0.0638, "step": 28240 }, { "grad_norm": 0.8467001914978027, "learning_rate": 8.593621905328858e-05, "loss": 0.0838, "step": 28250 }, { "grad_norm": 0.48162487149238586, "learning_rate": 8.592472059405333e-05, "loss": 0.0726, "step": 28260 }, { "grad_norm": 0.788550615310669, "learning_rate": 8.591321820614464e-05, "loss": 0.0731, "step": 28270 }, { "grad_norm": 0.5492426156997681, "learning_rate": 8.590171189082041e-05, "loss": 0.0876, "step": 28280 }, { "grad_norm": 0.6581282019615173, "learning_rate": 8.589020164933894e-05, "loss": 0.0732, "step": 28290 }, { "grad_norm": 0.9886089563369751, "learning_rate": 8.587868748295898e-05, "loss": 0.0596, "step": 28300 }, { "grad_norm": 0.5830642580986023, "learning_rate": 8.586716939293971e-05, "loss": 0.06, "step": 28310 }, { "grad_norm": 0.9170995950698853, "learning_rate": 8.58556473805407e-05, "loss": 0.0612, "step": 28320 }, { "grad_norm": 0.5219195485115051, "learning_rate": 8.584412144702202e-05, "loss": 0.0687, "step": 28330 }, { "grad_norm": 0.5240620970726013, "learning_rate": 8.58325915936441e-05, "loss": 0.0716, "step": 28340 }, { "grad_norm": 0.4965552091598511, "learning_rate": 8.582105782166783e-05, "loss": 0.0626, "step": 28350 }, { "grad_norm": 0.319658100605011, "learning_rate": 8.580952013235455e-05, "loss": 0.0591, "step": 28360 }, { "grad_norm": 0.9856499433517456, "learning_rate": 8.579797852696596e-05, "loss": 0.0649, "step": 28370 }, { "grad_norm": 0.5247169733047485, "learning_rate": 8.578643300676428e-05, "loss": 0.0655, "step": 28380 }, { "grad_norm": 0.3696332275867462, "learning_rate": 8.577488357301209e-05, "loss": 0.074, "step": 28390 }, { "grad_norm": 0.8279262781143188, "learning_rate": 8.576333022697242e-05, "loss": 0.064, "step": 28400 }, { "grad_norm": 1.0140248537063599, "learning_rate": 8.575177296990873e-05, "loss": 0.0687, "step": 28410 }, { "grad_norm": 0.7704429626464844, "learning_rate": 8.574021180308489e-05, "loss": 0.0771, "step": 28420 }, { "grad_norm": 0.7003499269485474, "learning_rate": 8.572864672776523e-05, "loss": 0.0747, "step": 28430 }, { "grad_norm": 0.4181438088417053, "learning_rate": 8.571707774521447e-05, "loss": 0.0583, "step": 28440 }, { "grad_norm": 0.49323317408561707, "learning_rate": 8.57055048566978e-05, "loss": 0.0575, "step": 28450 }, { "grad_norm": 0.8538905382156372, "learning_rate": 8.569392806348078e-05, "loss": 0.0735, "step": 28460 }, { "grad_norm": 0.5662397146224976, "learning_rate": 8.568234736682947e-05, "loss": 0.0606, "step": 28470 }, { "grad_norm": 0.5984818935394287, "learning_rate": 8.567076276801029e-05, "loss": 0.0925, "step": 28480 }, { "grad_norm": 0.5225776433944702, "learning_rate": 8.565917426829013e-05, "loss": 0.0475, "step": 28490 }, { "grad_norm": 1.1648647785186768, "learning_rate": 8.564758186893628e-05, "loss": 0.0663, "step": 28500 }, { "grad_norm": 0.5547174215316772, "learning_rate": 8.563598557121649e-05, "loss": 0.068, "step": 28510 }, { "grad_norm": 0.576730489730835, "learning_rate": 8.562438537639888e-05, "loss": 0.0637, "step": 28520 }, { "grad_norm": 1.1267502307891846, "learning_rate": 8.561278128575206e-05, "loss": 0.075, "step": 28530 }, { "grad_norm": 0.8739811182022095, "learning_rate": 8.5601173300545e-05, "loss": 0.0717, "step": 28540 }, { "grad_norm": 1.044455885887146, "learning_rate": 8.558956142204717e-05, "loss": 0.0793, "step": 28550 }, { "grad_norm": 0.6528717279434204, "learning_rate": 8.55779456515284e-05, "loss": 0.0789, "step": 28560 }, { "grad_norm": 0.538497269153595, "learning_rate": 8.556632599025898e-05, "loss": 0.0733, "step": 28570 }, { "grad_norm": 0.4428345561027527, "learning_rate": 8.555470243950964e-05, "loss": 0.062, "step": 28580 }, { "grad_norm": 0.38542068004608154, "learning_rate": 8.554307500055148e-05, "loss": 0.0642, "step": 28590 }, { "grad_norm": 0.6418411135673523, "learning_rate": 8.553144367465609e-05, "loss": 0.0547, "step": 28600 }, { "grad_norm": 0.4177907705307007, "learning_rate": 8.551980846309544e-05, "loss": 0.0645, "step": 28610 }, { "grad_norm": 0.6307201385498047, "learning_rate": 8.550816936714193e-05, "loss": 0.0786, "step": 28620 }, { "grad_norm": 0.4070395231246948, "learning_rate": 8.549652638806841e-05, "loss": 0.0624, "step": 28630 }, { "grad_norm": 1.2973576784133911, "learning_rate": 8.548487952714812e-05, "loss": 0.0865, "step": 28640 }, { "grad_norm": 0.7623012661933899, "learning_rate": 8.547322878565478e-05, "loss": 0.0573, "step": 28650 }, { "grad_norm": 0.5875428915023804, "learning_rate": 8.546157416486245e-05, "loss": 0.0711, "step": 28660 }, { "grad_norm": 0.6702502965927124, "learning_rate": 8.54499156660457e-05, "loss": 0.0712, "step": 28670 }, { "grad_norm": 0.7465789318084717, "learning_rate": 8.543825329047947e-05, "loss": 0.0495, "step": 28680 }, { "grad_norm": 0.57878178358078, "learning_rate": 8.542658703943913e-05, "loss": 0.0652, "step": 28690 }, { "grad_norm": 0.6976574659347534, "learning_rate": 8.541491691420051e-05, "loss": 0.0731, "step": 28700 }, { "grad_norm": 0.7211926579475403, "learning_rate": 8.54032429160398e-05, "loss": 0.0673, "step": 28710 }, { "grad_norm": 0.5913512110710144, "learning_rate": 8.539156504623369e-05, "loss": 0.0804, "step": 28720 }, { "grad_norm": 0.4746174216270447, "learning_rate": 8.537988330605923e-05, "loss": 0.084, "step": 28730 }, { "grad_norm": 0.57920241355896, "learning_rate": 8.536819769679393e-05, "loss": 0.0724, "step": 28740 }, { "grad_norm": 0.4880947470664978, "learning_rate": 8.53565082197157e-05, "loss": 0.061, "step": 28750 }, { "grad_norm": 0.507161557674408, "learning_rate": 8.534481487610289e-05, "loss": 0.081, "step": 28760 }, { "grad_norm": 0.4921027719974518, "learning_rate": 8.533311766723428e-05, "loss": 0.0516, "step": 28770 }, { "grad_norm": 0.47411978244781494, "learning_rate": 8.532141659438901e-05, "loss": 0.0548, "step": 28780 }, { "grad_norm": 0.7879480123519897, "learning_rate": 8.530971165884675e-05, "loss": 0.0619, "step": 28790 }, { "grad_norm": 0.5220322012901306, "learning_rate": 8.529800286188752e-05, "loss": 0.068, "step": 28800 }, { "grad_norm": 0.46317383646965027, "learning_rate": 8.528629020479175e-05, "loss": 0.063, "step": 28810 }, { "grad_norm": 0.4930605888366699, "learning_rate": 8.527457368884033e-05, "loss": 0.0561, "step": 28820 }, { "grad_norm": 0.5797176361083984, "learning_rate": 8.526285331531458e-05, "loss": 0.0443, "step": 28830 }, { "grad_norm": 0.47961023449897766, "learning_rate": 8.525112908549621e-05, "loss": 0.0514, "step": 28840 }, { "grad_norm": 0.6735710501670837, "learning_rate": 8.523940100066735e-05, "loss": 0.072, "step": 28850 }, { "grad_norm": 0.3661622405052185, "learning_rate": 8.52276690621106e-05, "loss": 0.0542, "step": 28860 }, { "grad_norm": 0.48217490315437317, "learning_rate": 8.521593327110889e-05, "loss": 0.0614, "step": 28870 }, { "grad_norm": 0.45128026604652405, "learning_rate": 8.520419362894569e-05, "loss": 0.0796, "step": 28880 }, { "grad_norm": 0.49552062153816223, "learning_rate": 8.51924501369048e-05, "loss": 0.0526, "step": 28890 }, { "grad_norm": 0.5213350057601929, "learning_rate": 8.518070279627047e-05, "loss": 0.0635, "step": 28900 }, { "grad_norm": 0.8032299876213074, "learning_rate": 8.516895160832737e-05, "loss": 0.0675, "step": 28910 }, { "grad_norm": 0.641455888748169, "learning_rate": 8.515719657436061e-05, "loss": 0.0564, "step": 28920 }, { "grad_norm": 0.8379350304603577, "learning_rate": 8.514543769565568e-05, "loss": 0.0609, "step": 28930 }, { "grad_norm": 1.0671738386154175, "learning_rate": 8.513367497349853e-05, "loss": 0.0685, "step": 28940 }, { "grad_norm": 0.46227020025253296, "learning_rate": 8.51219084091755e-05, "loss": 0.0568, "step": 28950 }, { "grad_norm": 0.4879419803619385, "learning_rate": 8.511013800397338e-05, "loss": 0.0624, "step": 28960 }, { "grad_norm": 0.5376954078674316, "learning_rate": 8.509836375917937e-05, "loss": 0.0671, "step": 28970 }, { "grad_norm": 1.28252375125885, "learning_rate": 8.508658567608104e-05, "loss": 0.073, "step": 28980 }, { "grad_norm": 0.40752360224723816, "learning_rate": 8.507480375596647e-05, "loss": 0.0482, "step": 28990 }, { "grad_norm": 0.5315948128700256, "learning_rate": 8.506301800012408e-05, "loss": 0.0547, "step": 29000 }, { "grad_norm": 0.733142077922821, "learning_rate": 8.505122840984278e-05, "loss": 0.0679, "step": 29010 }, { "grad_norm": 0.5112869143486023, "learning_rate": 8.503943498641182e-05, "loss": 0.0754, "step": 29020 }, { "grad_norm": 0.8917662501335144, "learning_rate": 8.502763773112095e-05, "loss": 0.0712, "step": 29030 }, { "grad_norm": 0.7163795828819275, "learning_rate": 8.501583664526026e-05, "loss": 0.0605, "step": 29040 }, { "grad_norm": 0.511347234249115, "learning_rate": 8.500403173012032e-05, "loss": 0.0704, "step": 29050 }, { "grad_norm": 0.5265129208564758, "learning_rate": 8.499222298699211e-05, "loss": 0.0644, "step": 29060 }, { "grad_norm": 0.6848763823509216, "learning_rate": 8.498041041716701e-05, "loss": 0.0539, "step": 29070 }, { "grad_norm": 0.934781014919281, "learning_rate": 8.496859402193681e-05, "loss": 0.0642, "step": 29080 }, { "grad_norm": 1.27518892288208, "learning_rate": 8.495677380259374e-05, "loss": 0.0678, "step": 29090 }, { "grad_norm": 0.43382129073143005, "learning_rate": 8.494494976043045e-05, "loss": 0.0704, "step": 29100 }, { "grad_norm": 0.5453978776931763, "learning_rate": 8.493312189673998e-05, "loss": 0.062, "step": 29110 }, { "grad_norm": 0.9223178625106812, "learning_rate": 8.492129021281584e-05, "loss": 0.0652, "step": 29120 }, { "grad_norm": 0.5449254512786865, "learning_rate": 8.490945470995188e-05, "loss": 0.0603, "step": 29130 }, { "grad_norm": 0.552552342414856, "learning_rate": 8.489761538944247e-05, "loss": 0.0643, "step": 29140 }, { "grad_norm": 0.5011366009712219, "learning_rate": 8.48857722525823e-05, "loss": 0.0583, "step": 29150 }, { "grad_norm": 0.36235395073890686, "learning_rate": 8.487392530066652e-05, "loss": 0.0599, "step": 29160 }, { "grad_norm": 0.7316604852676392, "learning_rate": 8.486207453499069e-05, "loss": 0.0684, "step": 29170 }, { "grad_norm": 0.5457270741462708, "learning_rate": 8.485021995685082e-05, "loss": 0.0764, "step": 29180 }, { "grad_norm": 0.6403481364250183, "learning_rate": 8.483836156754328e-05, "loss": 0.0594, "step": 29190 }, { "grad_norm": 1.635912537574768, "learning_rate": 8.482649936836491e-05, "loss": 0.0759, "step": 29200 }, { "grad_norm": 0.5719951391220093, "learning_rate": 8.481463336061293e-05, "loss": 0.0623, "step": 29210 }, { "grad_norm": 0.3596203923225403, "learning_rate": 8.480276354558496e-05, "loss": 0.0803, "step": 29220 }, { "grad_norm": 0.8067580461502075, "learning_rate": 8.479088992457913e-05, "loss": 0.0634, "step": 29230 }, { "grad_norm": 0.5357276797294617, "learning_rate": 8.477901249889387e-05, "loss": 0.0669, "step": 29240 }, { "grad_norm": 0.5240119695663452, "learning_rate": 8.47671312698281e-05, "loss": 0.0566, "step": 29250 }, { "grad_norm": 0.44254422187805176, "learning_rate": 8.475524623868112e-05, "loss": 0.0591, "step": 29260 }, { "grad_norm": 0.5193842649459839, "learning_rate": 8.474335740675266e-05, "loss": 0.0639, "step": 29270 }, { "grad_norm": 0.8331404328346252, "learning_rate": 8.473146477534289e-05, "loss": 0.07, "step": 29280 }, { "grad_norm": 0.6537258625030518, "learning_rate": 8.471956834575232e-05, "loss": 0.0637, "step": 29290 }, { "grad_norm": 0.7742031216621399, "learning_rate": 8.470766811928197e-05, "loss": 0.0585, "step": 29300 }, { "grad_norm": 0.552273690700531, "learning_rate": 8.469576409723323e-05, "loss": 0.0586, "step": 29310 }, { "grad_norm": 0.8197983503341675, "learning_rate": 8.468385628090788e-05, "loss": 0.0629, "step": 29320 }, { "grad_norm": 0.6964063048362732, "learning_rate": 8.467194467160815e-05, "loss": 0.068, "step": 29330 }, { "grad_norm": 0.9334434866905212, "learning_rate": 8.466002927063667e-05, "loss": 0.0626, "step": 29340 }, { "grad_norm": 0.518998384475708, "learning_rate": 8.464811007929651e-05, "loss": 0.0661, "step": 29350 }, { "grad_norm": 0.8369948863983154, "learning_rate": 8.463618709889114e-05, "loss": 0.0703, "step": 29360 }, { "grad_norm": 0.6575129628181458, "learning_rate": 8.462426033072442e-05, "loss": 0.067, "step": 29370 }, { "grad_norm": 0.6333649754524231, "learning_rate": 8.461232977610061e-05, "loss": 0.0633, "step": 29380 }, { "grad_norm": 1.0606105327606201, "learning_rate": 8.46003954363245e-05, "loss": 0.055, "step": 29390 }, { "grad_norm": 0.6164672374725342, "learning_rate": 8.458845731270115e-05, "loss": 0.0718, "step": 29400 }, { "grad_norm": 0.5484440326690674, "learning_rate": 8.45765154065361e-05, "loss": 0.0621, "step": 29410 }, { "grad_norm": 0.39173394441604614, "learning_rate": 8.456456971913532e-05, "loss": 0.0556, "step": 29420 }, { "grad_norm": 0.6159932613372803, "learning_rate": 8.455262025180517e-05, "loss": 0.0591, "step": 29430 }, { "grad_norm": 0.9735493659973145, "learning_rate": 8.454066700585242e-05, "loss": 0.0647, "step": 29440 }, { "grad_norm": 0.5815010070800781, "learning_rate": 8.452870998258423e-05, "loss": 0.0762, "step": 29450 }, { "grad_norm": 0.4439769983291626, "learning_rate": 8.451674918330825e-05, "loss": 0.0684, "step": 29460 }, { "grad_norm": 0.47763603925704956, "learning_rate": 8.450478460933246e-05, "loss": 0.0632, "step": 29470 }, { "grad_norm": 0.44555678963661194, "learning_rate": 8.449281626196532e-05, "loss": 0.0735, "step": 29480 }, { "grad_norm": 0.5396831631660461, "learning_rate": 8.448084414251564e-05, "loss": 0.0756, "step": 29490 }, { "grad_norm": 0.5142137408256531, "learning_rate": 8.446886825229271e-05, "loss": 0.0525, "step": 29500 }, { "grad_norm": 0.7320088744163513, "learning_rate": 8.445688859260615e-05, "loss": 0.0718, "step": 29510 }, { "grad_norm": 0.9142640233039856, "learning_rate": 8.444490516476606e-05, "loss": 0.0589, "step": 29520 }, { "grad_norm": 0.6582891345024109, "learning_rate": 8.443291797008293e-05, "loss": 0.0739, "step": 29530 }, { "grad_norm": 0.37354978919029236, "learning_rate": 8.442092700986765e-05, "loss": 0.072, "step": 29540 }, { "grad_norm": 1.4315778017044067, "learning_rate": 8.440893228543156e-05, "loss": 0.0719, "step": 29550 }, { "grad_norm": 1.8432642221450806, "learning_rate": 8.439693379808638e-05, "loss": 0.0675, "step": 29560 }, { "grad_norm": 0.7221134305000305, "learning_rate": 8.43849315491442e-05, "loss": 0.0713, "step": 29570 }, { "grad_norm": 0.728077232837677, "learning_rate": 8.437292553991763e-05, "loss": 0.0832, "step": 29580 }, { "grad_norm": 0.9619923830032349, "learning_rate": 8.436091577171959e-05, "loss": 0.0822, "step": 29590 }, { "grad_norm": 0.6396620869636536, "learning_rate": 8.434890224586347e-05, "loss": 0.0773, "step": 29600 }, { "grad_norm": 0.7776797413825989, "learning_rate": 8.433688496366303e-05, "loss": 0.0705, "step": 29610 }, { "grad_norm": 0.5140421390533447, "learning_rate": 8.432486392643248e-05, "loss": 0.0575, "step": 29620 }, { "grad_norm": 0.6854820251464844, "learning_rate": 8.431283913548643e-05, "loss": 0.0658, "step": 29630 }, { "grad_norm": 0.7104924321174622, "learning_rate": 8.430081059213985e-05, "loss": 0.0677, "step": 29640 }, { "grad_norm": 0.6623669266700745, "learning_rate": 8.428877829770823e-05, "loss": 0.0737, "step": 29650 }, { "grad_norm": 1.4008464813232422, "learning_rate": 8.427674225350735e-05, "loss": 0.0852, "step": 29660 }, { "grad_norm": 1.561189889907837, "learning_rate": 8.426470246085347e-05, "loss": 0.0813, "step": 29670 }, { "grad_norm": 0.7013693451881409, "learning_rate": 8.425265892106324e-05, "loss": 0.0765, "step": 29680 }, { "grad_norm": 0.49982813000679016, "learning_rate": 8.424061163545374e-05, "loss": 0.0766, "step": 29690 }, { "grad_norm": 0.49794748425483704, "learning_rate": 8.422856060534243e-05, "loss": 0.0676, "step": 29700 }, { "grad_norm": 0.6545180678367615, "learning_rate": 8.421650583204718e-05, "loss": 0.0684, "step": 29710 }, { "grad_norm": 0.480111300945282, "learning_rate": 8.420444731688633e-05, "loss": 0.0727, "step": 29720 }, { "grad_norm": 0.3959406912326813, "learning_rate": 8.419238506117852e-05, "loss": 0.0789, "step": 29730 }, { "grad_norm": 0.6014459729194641, "learning_rate": 8.418031906624289e-05, "loss": 0.0583, "step": 29740 }, { "grad_norm": 0.7419065237045288, "learning_rate": 8.416824933339898e-05, "loss": 0.0765, "step": 29750 }, { "grad_norm": 0.5831483006477356, "learning_rate": 8.415617586396667e-05, "loss": 0.0562, "step": 29760 }, { "grad_norm": 0.6170328259468079, "learning_rate": 8.414409865926632e-05, "loss": 0.0751, "step": 29770 }, { "grad_norm": 0.5288792252540588, "learning_rate": 8.413201772061867e-05, "loss": 0.0613, "step": 29780 }, { "grad_norm": 0.4023608863353729, "learning_rate": 8.411993304934488e-05, "loss": 0.0547, "step": 29790 }, { "grad_norm": 0.420283704996109, "learning_rate": 8.410784464676654e-05, "loss": 0.0613, "step": 29800 }, { "grad_norm": 0.8241367936134338, "learning_rate": 8.409575251420556e-05, "loss": 0.0661, "step": 29810 }, { "grad_norm": 0.3530080020427704, "learning_rate": 8.408365665298435e-05, "loss": 0.0574, "step": 29820 }, { "grad_norm": 0.5384415984153748, "learning_rate": 8.40715570644257e-05, "loss": 0.0782, "step": 29830 }, { "grad_norm": 0.5980070233345032, "learning_rate": 8.40594537498528e-05, "loss": 0.0792, "step": 29840 }, { "grad_norm": 0.5290810465812683, "learning_rate": 8.404734671058924e-05, "loss": 0.0653, "step": 29850 }, { "grad_norm": 0.5834527611732483, "learning_rate": 8.403523594795902e-05, "loss": 0.0687, "step": 29860 }, { "grad_norm": 0.9834302067756653, "learning_rate": 8.402312146328659e-05, "loss": 0.0589, "step": 29870 }, { "grad_norm": 0.6289923787117004, "learning_rate": 8.401100325789675e-05, "loss": 0.0687, "step": 29880 }, { "grad_norm": 0.5266468524932861, "learning_rate": 8.399888133311472e-05, "loss": 0.0604, "step": 29890 }, { "grad_norm": 0.7000634670257568, "learning_rate": 8.398675569026613e-05, "loss": 0.0701, "step": 29900 }, { "grad_norm": 0.5357553362846375, "learning_rate": 8.397462633067705e-05, "loss": 0.0617, "step": 29910 }, { "grad_norm": 0.5340375304222107, "learning_rate": 8.396249325567392e-05, "loss": 0.074, "step": 29920 }, { "grad_norm": 0.9202122092247009, "learning_rate": 8.395035646658357e-05, "loss": 0.0898, "step": 29930 }, { "grad_norm": 0.8691175580024719, "learning_rate": 8.39382159647333e-05, "loss": 0.0828, "step": 29940 }, { "grad_norm": 0.6958047151565552, "learning_rate": 8.392607175145075e-05, "loss": 0.072, "step": 29950 }, { "grad_norm": 0.7746875286102295, "learning_rate": 8.3913923828064e-05, "loss": 0.0749, "step": 29960 }, { "grad_norm": 0.3795734643936157, "learning_rate": 8.390177219590152e-05, "loss": 0.0554, "step": 29970 }, { "grad_norm": 0.674047589302063, "learning_rate": 8.388961685629222e-05, "loss": 0.0766, "step": 29980 }, { "grad_norm": 0.8626571297645569, "learning_rate": 8.387745781056536e-05, "loss": 0.0578, "step": 29990 }, { "grad_norm": 0.4072800278663635, "learning_rate": 8.386529506005065e-05, "loss": 0.0523, "step": 30000 }, { "grad_norm": 0.42974036931991577, "learning_rate": 8.38531286060782e-05, "loss": 0.0593, "step": 30010 }, { "grad_norm": 0.6742606163024902, "learning_rate": 8.384095844997849e-05, "loss": 0.0529, "step": 30020 }, { "grad_norm": 0.4667675197124481, "learning_rate": 8.382878459308245e-05, "loss": 0.0575, "step": 30030 }, { "grad_norm": 0.572085440158844, "learning_rate": 8.381660703672138e-05, "loss": 0.0596, "step": 30040 }, { "grad_norm": 0.687226414680481, "learning_rate": 8.380442578222702e-05, "loss": 0.0795, "step": 30050 }, { "grad_norm": 0.5829780101776123, "learning_rate": 8.379224083093146e-05, "loss": 0.0633, "step": 30060 }, { "grad_norm": 0.47972506284713745, "learning_rate": 8.378005218416727e-05, "loss": 0.0694, "step": 30070 }, { "grad_norm": 0.5909480452537537, "learning_rate": 8.376785984326735e-05, "loss": 0.0666, "step": 30080 }, { "grad_norm": 0.546733021736145, "learning_rate": 8.375566380956506e-05, "loss": 0.0591, "step": 30090 }, { "grad_norm": 0.9117475152015686, "learning_rate": 8.374346408439411e-05, "loss": 0.0781, "step": 30100 }, { "grad_norm": 0.6261569261550903, "learning_rate": 8.373126066908868e-05, "loss": 0.0662, "step": 30110 }, { "grad_norm": 0.5941520929336548, "learning_rate": 8.371905356498326e-05, "loss": 0.0708, "step": 30120 }, { "grad_norm": 0.7495465278625488, "learning_rate": 8.370684277341288e-05, "loss": 0.058, "step": 30130 }, { "grad_norm": 0.5517815947532654, "learning_rate": 8.369462829571282e-05, "loss": 0.06, "step": 30140 }, { "grad_norm": 0.695868194103241, "learning_rate": 8.36824101332189e-05, "loss": 0.0558, "step": 30150 }, { "grad_norm": 0.46416175365448, "learning_rate": 8.367018828726721e-05, "loss": 0.0656, "step": 30160 }, { "grad_norm": 0.683056116104126, "learning_rate": 8.365796275919438e-05, "loss": 0.0754, "step": 30170 }, { "grad_norm": 0.8852797746658325, "learning_rate": 8.364573355033734e-05, "loss": 0.0533, "step": 30180 }, { "grad_norm": 0.4983210265636444, "learning_rate": 8.363350066203346e-05, "loss": 0.0512, "step": 30190 }, { "grad_norm": 0.75600266456604, "learning_rate": 8.362126409562053e-05, "loss": 0.0667, "step": 30200 }, { "grad_norm": 0.560215950012207, "learning_rate": 8.360902385243667e-05, "loss": 0.0572, "step": 30210 }, { "grad_norm": 0.4819476306438446, "learning_rate": 8.359677993382052e-05, "loss": 0.0495, "step": 30220 }, { "grad_norm": 0.47125717997550964, "learning_rate": 8.358453234111103e-05, "loss": 0.0569, "step": 30230 }, { "grad_norm": 0.6872603297233582, "learning_rate": 8.357228107564756e-05, "loss": 0.0608, "step": 30240 }, { "grad_norm": 0.6511592268943787, "learning_rate": 8.356002613876993e-05, "loss": 0.0573, "step": 30250 }, { "grad_norm": 0.3567284941673279, "learning_rate": 8.35477675318183e-05, "loss": 0.0475, "step": 30260 }, { "grad_norm": 0.8979402780532837, "learning_rate": 8.353550525613323e-05, "loss": 0.0689, "step": 30270 }, { "grad_norm": 0.5589401721954346, "learning_rate": 8.352323931305572e-05, "loss": 0.0731, "step": 30280 }, { "grad_norm": 0.5282641649246216, "learning_rate": 8.351096970392717e-05, "loss": 0.0615, "step": 30290 }, { "grad_norm": 0.5733829140663147, "learning_rate": 8.349869643008937e-05, "loss": 0.0685, "step": 30300 }, { "grad_norm": 0.8556593060493469, "learning_rate": 8.348641949288449e-05, "loss": 0.0835, "step": 30310 }, { "grad_norm": 0.5515826344490051, "learning_rate": 8.34741388936551e-05, "loss": 0.0694, "step": 30320 }, { "grad_norm": 0.5863445997238159, "learning_rate": 8.346185463374423e-05, "loss": 0.0838, "step": 30330 }, { "grad_norm": 0.398624986410141, "learning_rate": 8.344956671449524e-05, "loss": 0.0667, "step": 30340 }, { "grad_norm": 0.5662630200386047, "learning_rate": 8.343727513725192e-05, "loss": 0.0635, "step": 30350 }, { "grad_norm": 0.6248458027839661, "learning_rate": 8.342497990335847e-05, "loss": 0.0617, "step": 30360 }, { "grad_norm": 0.9602419137954712, "learning_rate": 8.341268101415946e-05, "loss": 0.0691, "step": 30370 }, { "grad_norm": 0.4520256519317627, "learning_rate": 8.34003784709999e-05, "loss": 0.0607, "step": 30380 }, { "grad_norm": 0.4142971932888031, "learning_rate": 8.338807227522516e-05, "loss": 0.0669, "step": 30390 }, { "grad_norm": 0.579619288444519, "learning_rate": 8.337576242818103e-05, "loss": 0.0641, "step": 30400 }, { "grad_norm": 0.4290531277656555, "learning_rate": 8.336344893121372e-05, "loss": 0.0634, "step": 30410 }, { "grad_norm": 0.41254356503486633, "learning_rate": 8.335113178566977e-05, "loss": 0.053, "step": 30420 }, { "grad_norm": 0.6120868921279907, "learning_rate": 8.33388109928962e-05, "loss": 0.0661, "step": 30430 }, { "grad_norm": 0.8690680861473083, "learning_rate": 8.33264865542404e-05, "loss": 0.0712, "step": 30440 }, { "grad_norm": 0.22986970841884613, "learning_rate": 8.331415847105013e-05, "loss": 0.0634, "step": 30450 }, { "grad_norm": 0.4803661108016968, "learning_rate": 8.330182674467357e-05, "loss": 0.0599, "step": 30460 }, { "grad_norm": 0.5044216513633728, "learning_rate": 8.32894913764593e-05, "loss": 0.0588, "step": 30470 }, { "grad_norm": 2.6239402294158936, "learning_rate": 8.327715236775633e-05, "loss": 0.0595, "step": 30480 }, { "grad_norm": 0.4774385392665863, "learning_rate": 8.326480971991398e-05, "loss": 0.0529, "step": 30490 }, { "grad_norm": 0.6061402559280396, "learning_rate": 8.325246343428206e-05, "loss": 0.0644, "step": 30500 }, { "grad_norm": 0.47286897897720337, "learning_rate": 8.324011351221072e-05, "loss": 0.0618, "step": 30510 }, { "grad_norm": 0.4470769464969635, "learning_rate": 8.322775995505057e-05, "loss": 0.0639, "step": 30520 }, { "grad_norm": 0.8520841598510742, "learning_rate": 8.321540276415254e-05, "loss": 0.0644, "step": 30530 }, { "grad_norm": 0.4632759690284729, "learning_rate": 8.320304194086798e-05, "loss": 0.0593, "step": 30540 }, { "grad_norm": 0.6443447470664978, "learning_rate": 8.31906774865487e-05, "loss": 0.0497, "step": 30550 }, { "grad_norm": 0.6109731793403625, "learning_rate": 8.317830940254682e-05, "loss": 0.066, "step": 30560 }, { "grad_norm": 0.6172118186950684, "learning_rate": 8.316593769021491e-05, "loss": 0.0661, "step": 30570 }, { "grad_norm": 0.5871868133544922, "learning_rate": 8.315356235090592e-05, "loss": 0.0668, "step": 30580 }, { "grad_norm": 1.4748598337173462, "learning_rate": 8.314118338597319e-05, "loss": 0.0746, "step": 30590 }, { "grad_norm": 0.409973680973053, "learning_rate": 8.312880079677048e-05, "loss": 0.0643, "step": 30600 }, { "grad_norm": 0.7021113038063049, "learning_rate": 8.311641458465191e-05, "loss": 0.0732, "step": 30610 }, { "grad_norm": 0.831434428691864, "learning_rate": 8.310402475097205e-05, "loss": 0.0752, "step": 30620 }, { "grad_norm": 0.644754946231842, "learning_rate": 8.309163129708581e-05, "loss": 0.0729, "step": 30630 }, { "grad_norm": 0.5594604015350342, "learning_rate": 8.307923422434852e-05, "loss": 0.0597, "step": 30640 }, { "grad_norm": 0.6957604885101318, "learning_rate": 8.30668335341159e-05, "loss": 0.0705, "step": 30650 }, { "grad_norm": 0.5475778579711914, "learning_rate": 8.305442922774408e-05, "loss": 0.0614, "step": 30660 }, { "grad_norm": 0.43954840302467346, "learning_rate": 8.304202130658959e-05, "loss": 0.0545, "step": 30670 }, { "grad_norm": 0.4806738495826721, "learning_rate": 8.302960977200931e-05, "loss": 0.0451, "step": 30680 }, { "grad_norm": 0.761177659034729, "learning_rate": 8.301719462536058e-05, "loss": 0.0623, "step": 30690 }, { "grad_norm": 0.6453550457954407, "learning_rate": 8.300477586800108e-05, "loss": 0.0622, "step": 30700 }, { "grad_norm": 0.5456706881523132, "learning_rate": 8.299235350128892e-05, "loss": 0.0567, "step": 30710 }, { "grad_norm": 0.5170758962631226, "learning_rate": 8.297992752658258e-05, "loss": 0.045, "step": 30720 }, { "grad_norm": 0.42820197343826294, "learning_rate": 8.296749794524094e-05, "loss": 0.0595, "step": 30730 }, { "grad_norm": 0.6274471282958984, "learning_rate": 8.295506475862332e-05, "loss": 0.0498, "step": 30740 }, { "grad_norm": 0.6953765749931335, "learning_rate": 8.294262796808933e-05, "loss": 0.0568, "step": 30750 }, { "grad_norm": 0.4891088306903839, "learning_rate": 8.293018757499909e-05, "loss": 0.0647, "step": 30760 }, { "grad_norm": 0.9502344131469727, "learning_rate": 8.291774358071305e-05, "loss": 0.0597, "step": 30770 }, { "grad_norm": 0.6571516990661621, "learning_rate": 8.290529598659205e-05, "loss": 0.0716, "step": 30780 }, { "grad_norm": 0.5805913805961609, "learning_rate": 8.289284479399738e-05, "loss": 0.0559, "step": 30790 }, { "grad_norm": 0.5635450482368469, "learning_rate": 8.288039000429064e-05, "loss": 0.0556, "step": 30800 }, { "grad_norm": 0.807903528213501, "learning_rate": 8.286793161883388e-05, "loss": 0.07, "step": 30810 }, { "grad_norm": 0.7088199257850647, "learning_rate": 8.285546963898954e-05, "loss": 0.0661, "step": 30820 }, { "grad_norm": 0.7220513224601746, "learning_rate": 8.284300406612044e-05, "loss": 0.0618, "step": 30830 }, { "grad_norm": 0.7176201939582825, "learning_rate": 8.283053490158978e-05, "loss": 0.059, "step": 30840 }, { "grad_norm": 0.46162649989128113, "learning_rate": 8.28180621467612e-05, "loss": 0.0491, "step": 30850 }, { "grad_norm": 0.39098769426345825, "learning_rate": 8.280558580299867e-05, "loss": 0.0605, "step": 30860 }, { "grad_norm": 0.6503124237060547, "learning_rate": 8.279310587166661e-05, "loss": 0.0608, "step": 30870 }, { "grad_norm": 0.4264712333679199, "learning_rate": 8.278062235412978e-05, "loss": 0.0556, "step": 30880 }, { "grad_norm": 0.5197107195854187, "learning_rate": 8.276813525175339e-05, "loss": 0.0673, "step": 30890 }, { "grad_norm": 0.5674344301223755, "learning_rate": 8.2755644565903e-05, "loss": 0.0655, "step": 30900 }, { "grad_norm": 0.48921000957489014, "learning_rate": 8.274315029794454e-05, "loss": 0.0537, "step": 30910 }, { "grad_norm": 0.963762104511261, "learning_rate": 8.273065244924443e-05, "loss": 0.0644, "step": 30920 }, { "grad_norm": 0.8427486419677734, "learning_rate": 8.271815102116936e-05, "loss": 0.0587, "step": 30930 }, { "grad_norm": 0.6259502172470093, "learning_rate": 8.270564601508648e-05, "loss": 0.0605, "step": 30940 }, { "grad_norm": 0.5256237387657166, "learning_rate": 8.269313743236333e-05, "loss": 0.0585, "step": 30950 }, { "grad_norm": 0.3973122239112854, "learning_rate": 8.268062527436783e-05, "loss": 0.0564, "step": 30960 }, { "grad_norm": 0.5921386480331421, "learning_rate": 8.266810954246828e-05, "loss": 0.0627, "step": 30970 }, { "grad_norm": 0.6510362029075623, "learning_rate": 8.265559023803338e-05, "loss": 0.0426, "step": 30980 }, { "grad_norm": 0.5246044397354126, "learning_rate": 8.264306736243223e-05, "loss": 0.0582, "step": 30990 }, { "grad_norm": 1.1215221881866455, "learning_rate": 8.263054091703432e-05, "loss": 0.0701, "step": 31000 }, { "grad_norm": 0.6603337526321411, "learning_rate": 8.26180109032095e-05, "loss": 0.0589, "step": 31010 }, { "grad_norm": 0.7412529587745667, "learning_rate": 8.260547732232807e-05, "loss": 0.0588, "step": 31020 }, { "grad_norm": 0.4913075566291809, "learning_rate": 8.259294017576063e-05, "loss": 0.0572, "step": 31030 }, { "grad_norm": 0.6259194612503052, "learning_rate": 8.258039946487828e-05, "loss": 0.0564, "step": 31040 }, { "grad_norm": 0.7174782156944275, "learning_rate": 8.256785519105241e-05, "loss": 0.058, "step": 31050 }, { "grad_norm": 1.2596399784088135, "learning_rate": 8.255530735565488e-05, "loss": 0.0661, "step": 31060 }, { "grad_norm": 0.9048771262168884, "learning_rate": 8.254275596005787e-05, "loss": 0.0643, "step": 31070 }, { "grad_norm": 0.7417299747467041, "learning_rate": 8.2530201005634e-05, "loss": 0.0797, "step": 31080 }, { "grad_norm": 0.8996754884719849, "learning_rate": 8.251764249375626e-05, "loss": 0.0752, "step": 31090 }, { "grad_norm": 0.6596020460128784, "learning_rate": 8.250508042579803e-05, "loss": 0.0638, "step": 31100 }, { "grad_norm": 1.1588207483291626, "learning_rate": 8.249251480313307e-05, "loss": 0.0725, "step": 31110 }, { "grad_norm": 0.6397458910942078, "learning_rate": 8.247994562713555e-05, "loss": 0.0663, "step": 31120 }, { "grad_norm": 0.4614182114601135, "learning_rate": 8.246737289917998e-05, "loss": 0.0781, "step": 31130 }, { "grad_norm": 0.5875397324562073, "learning_rate": 8.245479662064135e-05, "loss": 0.0598, "step": 31140 }, { "grad_norm": 0.7296130657196045, "learning_rate": 8.244221679289496e-05, "loss": 0.0527, "step": 31150 }, { "grad_norm": 0.9107297658920288, "learning_rate": 8.242963341731652e-05, "loss": 0.0531, "step": 31160 }, { "grad_norm": 0.4980422258377075, "learning_rate": 8.24170464952821e-05, "loss": 0.0576, "step": 31170 }, { "grad_norm": 0.47804152965545654, "learning_rate": 8.240445602816824e-05, "loss": 0.0572, "step": 31180 }, { "grad_norm": 0.8463224768638611, "learning_rate": 8.239186201735179e-05, "loss": 0.0633, "step": 31190 }, { "grad_norm": 0.3799625337123871, "learning_rate": 8.237926446420998e-05, "loss": 0.0633, "step": 31200 }, { "grad_norm": 0.6695966124534607, "learning_rate": 8.236666337012052e-05, "loss": 0.0546, "step": 31210 }, { "grad_norm": 0.5857899188995361, "learning_rate": 8.23540587364614e-05, "loss": 0.0677, "step": 31220 }, { "grad_norm": 0.43054115772247314, "learning_rate": 8.234145056461107e-05, "loss": 0.0623, "step": 31230 }, { "grad_norm": 0.48688727617263794, "learning_rate": 8.232883885594832e-05, "loss": 0.0749, "step": 31240 }, { "grad_norm": 0.8363404273986816, "learning_rate": 8.231622361185236e-05, "loss": 0.0642, "step": 31250 }, { "grad_norm": 0.7385798096656799, "learning_rate": 8.230360483370278e-05, "loss": 0.0898, "step": 31260 }, { "grad_norm": 1.248925805091858, "learning_rate": 8.229098252287953e-05, "loss": 0.0891, "step": 31270 }, { "grad_norm": 0.7399126887321472, "learning_rate": 8.2278356680763e-05, "loss": 0.0785, "step": 31280 }, { "grad_norm": 0.986995279788971, "learning_rate": 8.22657273087339e-05, "loss": 0.0684, "step": 31290 }, { "grad_norm": 0.8218235969543457, "learning_rate": 8.225309440817336e-05, "loss": 0.0789, "step": 31300 }, { "grad_norm": 0.6975133419036865, "learning_rate": 8.224045798046293e-05, "loss": 0.0829, "step": 31310 }, { "grad_norm": 0.46218305826187134, "learning_rate": 8.22278180269845e-05, "loss": 0.0741, "step": 31320 }, { "grad_norm": 0.5645705461502075, "learning_rate": 8.221517454912031e-05, "loss": 0.0621, "step": 31330 }, { "grad_norm": 0.49541449546813965, "learning_rate": 8.220252754825308e-05, "loss": 0.053, "step": 31340 }, { "grad_norm": 0.4886656105518341, "learning_rate": 8.218987702576586e-05, "loss": 0.0571, "step": 31350 }, { "grad_norm": 0.6288127303123474, "learning_rate": 8.217722298304207e-05, "loss": 0.0631, "step": 31360 }, { "grad_norm": 0.42478856444358826, "learning_rate": 8.216456542146557e-05, "loss": 0.0617, "step": 31370 }, { "grad_norm": 0.5844901204109192, "learning_rate": 8.215190434242055e-05, "loss": 0.0492, "step": 31380 }, { "grad_norm": 0.6201057434082031, "learning_rate": 8.213923974729161e-05, "loss": 0.0644, "step": 31390 }, { "grad_norm": 0.5705466270446777, "learning_rate": 8.212657163746373e-05, "loss": 0.0555, "step": 31400 }, { "grad_norm": 0.43598687648773193, "learning_rate": 8.211390001432227e-05, "loss": 0.0527, "step": 31410 }, { "grad_norm": 0.4700496196746826, "learning_rate": 8.210122487925297e-05, "loss": 0.0652, "step": 31420 }, { "grad_norm": 0.7546883225440979, "learning_rate": 8.208854623364202e-05, "loss": 0.0613, "step": 31430 }, { "grad_norm": 0.6857277154922485, "learning_rate": 8.207586407887589e-05, "loss": 0.0612, "step": 31440 }, { "grad_norm": 0.4637499749660492, "learning_rate": 8.206317841634148e-05, "loss": 0.0548, "step": 31450 }, { "grad_norm": 0.3781781792640686, "learning_rate": 8.205048924742609e-05, "loss": 0.0613, "step": 31460 }, { "grad_norm": 0.7109658718109131, "learning_rate": 8.203779657351738e-05, "loss": 0.0578, "step": 31470 }, { "grad_norm": 0.5567046403884888, "learning_rate": 8.20251003960034e-05, "loss": 0.0647, "step": 31480 }, { "grad_norm": 0.45603275299072266, "learning_rate": 8.201240071627258e-05, "loss": 0.0632, "step": 31490 }, { "grad_norm": 0.6087816953659058, "learning_rate": 8.199969753571377e-05, "loss": 0.0662, "step": 31500 }, { "grad_norm": 0.7270925641059875, "learning_rate": 8.198699085571615e-05, "loss": 0.0619, "step": 31510 }, { "grad_norm": 0.5052213072776794, "learning_rate": 8.197428067766928e-05, "loss": 0.0657, "step": 31520 }, { "grad_norm": 1.3032892942428589, "learning_rate": 8.196156700296316e-05, "loss": 0.0791, "step": 31530 }, { "grad_norm": 0.5802841186523438, "learning_rate": 8.194884983298814e-05, "loss": 0.0544, "step": 31540 }, { "grad_norm": 0.46217095851898193, "learning_rate": 8.193612916913491e-05, "loss": 0.0604, "step": 31550 }, { "grad_norm": 0.5040650367736816, "learning_rate": 8.192340501279463e-05, "loss": 0.0543, "step": 31560 }, { "grad_norm": 0.4824482202529907, "learning_rate": 8.191067736535876e-05, "loss": 0.065, "step": 31570 }, { "grad_norm": 0.7232305407524109, "learning_rate": 8.18979462282192e-05, "loss": 0.072, "step": 31580 }, { "grad_norm": 0.5331870913505554, "learning_rate": 8.188521160276819e-05, "loss": 0.064, "step": 31590 }, { "grad_norm": 0.6014114618301392, "learning_rate": 8.187247349039837e-05, "loss": 0.0683, "step": 31600 }, { "grad_norm": 0.5712324380874634, "learning_rate": 8.185973189250278e-05, "loss": 0.0674, "step": 31610 }, { "grad_norm": 0.7188146114349365, "learning_rate": 8.184698681047482e-05, "loss": 0.0755, "step": 31620 }, { "grad_norm": 0.5027790069580078, "learning_rate": 8.183423824570827e-05, "loss": 0.073, "step": 31630 }, { "grad_norm": 0.47802743315696716, "learning_rate": 8.182148619959725e-05, "loss": 0.0643, "step": 31640 }, { "grad_norm": 0.5288105607032776, "learning_rate": 8.180873067353636e-05, "loss": 0.061, "step": 31650 }, { "grad_norm": 0.5775167346000671, "learning_rate": 8.179597166892052e-05, "loss": 0.0546, "step": 31660 }, { "grad_norm": 0.5416283011436462, "learning_rate": 8.178320918714501e-05, "loss": 0.0572, "step": 31670 }, { "grad_norm": 1.237322211265564, "learning_rate": 8.177044322960554e-05, "loss": 0.0635, "step": 31680 }, { "grad_norm": 0.7337195873260498, "learning_rate": 8.175767379769816e-05, "loss": 0.0659, "step": 31690 }, { "grad_norm": 0.7094279527664185, "learning_rate": 8.174490089281932e-05, "loss": 0.0681, "step": 31700 }, { "grad_norm": 0.5245396494865417, "learning_rate": 8.173212451636584e-05, "loss": 0.0798, "step": 31710 }, { "grad_norm": 0.7550330758094788, "learning_rate": 8.171934466973493e-05, "loss": 0.084, "step": 31720 }, { "grad_norm": 0.32033416628837585, "learning_rate": 8.170656135432418e-05, "loss": 0.066, "step": 31730 }, { "grad_norm": 0.7557323575019836, "learning_rate": 8.169377457153155e-05, "loss": 0.0707, "step": 31740 }, { "grad_norm": 0.4966825246810913, "learning_rate": 8.168098432275539e-05, "loss": 0.0692, "step": 31750 }, { "grad_norm": 1.0723055601119995, "learning_rate": 8.166819060939442e-05, "loss": 0.0737, "step": 31760 }, { "grad_norm": 0.5362398028373718, "learning_rate": 8.165539343284772e-05, "loss": 0.0683, "step": 31770 }, { "grad_norm": 2.154308795928955, "learning_rate": 8.16425927945148e-05, "loss": 0.0802, "step": 31780 }, { "grad_norm": 0.6499096155166626, "learning_rate": 8.162978869579551e-05, "loss": 0.0672, "step": 31790 }, { "grad_norm": 0.7184817790985107, "learning_rate": 8.161698113809007e-05, "loss": 0.0536, "step": 31800 }, { "grad_norm": 0.6060891151428223, "learning_rate": 8.160417012279911e-05, "loss": 0.0797, "step": 31810 }, { "grad_norm": 0.538266122341156, "learning_rate": 8.159135565132363e-05, "loss": 0.0675, "step": 31820 }, { "grad_norm": 1.0959933996200562, "learning_rate": 8.157853772506498e-05, "loss": 0.0856, "step": 31830 }, { "grad_norm": 0.6370545625686646, "learning_rate": 8.156571634542494e-05, "loss": 0.0683, "step": 31840 }, { "grad_norm": 0.7328980565071106, "learning_rate": 8.15528915138056e-05, "loss": 0.0573, "step": 31850 }, { "grad_norm": 0.5141611099243164, "learning_rate": 8.154006323160949e-05, "loss": 0.0586, "step": 31860 }, { "grad_norm": 0.6159022450447083, "learning_rate": 8.152723150023949e-05, "loss": 0.068, "step": 31870 }, { "grad_norm": 0.5514340996742249, "learning_rate": 8.151439632109886e-05, "loss": 0.0791, "step": 31880 }, { "grad_norm": 0.6228532791137695, "learning_rate": 8.150155769559122e-05, "loss": 0.0661, "step": 31890 }, { "grad_norm": 0.7446309924125671, "learning_rate": 8.148871562512058e-05, "loss": 0.0629, "step": 31900 }, { "grad_norm": 0.48306989669799805, "learning_rate": 8.147587011109136e-05, "loss": 0.0759, "step": 31910 }, { "grad_norm": 0.5960713028907776, "learning_rate": 8.14630211549083e-05, "loss": 0.0734, "step": 31920 }, { "grad_norm": 0.6195316910743713, "learning_rate": 8.145016875797655e-05, "loss": 0.0472, "step": 31930 }, { "grad_norm": 0.5467785000801086, "learning_rate": 8.143731292170164e-05, "loss": 0.0681, "step": 31940 }, { "grad_norm": 0.7786531448364258, "learning_rate": 8.142445364748944e-05, "loss": 0.0659, "step": 31950 }, { "grad_norm": 0.46391335129737854, "learning_rate": 8.141159093674624e-05, "loss": 0.068, "step": 31960 }, { "grad_norm": 0.6061497926712036, "learning_rate": 8.139872479087869e-05, "loss": 0.0627, "step": 31970 }, { "grad_norm": 0.5510544180870056, "learning_rate": 8.13858552112938e-05, "loss": 0.05, "step": 31980 }, { "grad_norm": 0.4526425898075104, "learning_rate": 8.137298219939895e-05, "loss": 0.0703, "step": 31990 }, { "grad_norm": 0.45976680517196655, "learning_rate": 8.136010575660196e-05, "loss": 0.0504, "step": 32000 }, { "grad_norm": 0.559728741645813, "learning_rate": 8.134722588431095e-05, "loss": 0.0598, "step": 32010 }, { "grad_norm": 0.5533867478370667, "learning_rate": 8.133434258393444e-05, "loss": 0.0747, "step": 32020 }, { "grad_norm": 0.806084394454956, "learning_rate": 8.132145585688134e-05, "loss": 0.0711, "step": 32030 }, { "grad_norm": 0.6117070317268372, "learning_rate": 8.130856570456093e-05, "loss": 0.0728, "step": 32040 }, { "grad_norm": 0.597743809223175, "learning_rate": 8.129567212838283e-05, "loss": 0.069, "step": 32050 }, { "grad_norm": 0.47435471415519714, "learning_rate": 8.128277512975708e-05, "loss": 0.0743, "step": 32060 }, { "grad_norm": 0.6246134042739868, "learning_rate": 8.126987471009408e-05, "loss": 0.0693, "step": 32070 }, { "grad_norm": 0.4549911618232727, "learning_rate": 8.125697087080459e-05, "loss": 0.0557, "step": 32080 }, { "grad_norm": 0.6597950458526611, "learning_rate": 8.124406361329976e-05, "loss": 0.0708, "step": 32090 }, { "grad_norm": 0.6228259801864624, "learning_rate": 8.123115293899111e-05, "loss": 0.0597, "step": 32100 }, { "grad_norm": 0.46445396542549133, "learning_rate": 8.121823884929055e-05, "loss": 0.0512, "step": 32110 }, { "grad_norm": 0.4957263767719269, "learning_rate": 8.12053213456103e-05, "loss": 0.0479, "step": 32120 }, { "grad_norm": 0.3983455002307892, "learning_rate": 8.119240042936303e-05, "loss": 0.0449, "step": 32130 }, { "grad_norm": 0.6514474749565125, "learning_rate": 8.117947610196175e-05, "loss": 0.0557, "step": 32140 }, { "grad_norm": 0.8327457308769226, "learning_rate": 8.116654836481982e-05, "loss": 0.0591, "step": 32150 }, { "grad_norm": 0.802550196647644, "learning_rate": 8.115361721935106e-05, "loss": 0.0718, "step": 32160 }, { "grad_norm": 0.45887959003448486, "learning_rate": 8.114068266696953e-05, "loss": 0.065, "step": 32170 }, { "grad_norm": 0.4848642945289612, "learning_rate": 8.112774470908978e-05, "loss": 0.0476, "step": 32180 }, { "grad_norm": 0.5335450172424316, "learning_rate": 8.111480334712665e-05, "loss": 0.0568, "step": 32190 }, { "grad_norm": 0.6430681943893433, "learning_rate": 8.110185858249542e-05, "loss": 0.0526, "step": 32200 }, { "grad_norm": 0.7181524038314819, "learning_rate": 8.108891041661168e-05, "loss": 0.0551, "step": 32210 }, { "grad_norm": 0.5055475234985352, "learning_rate": 8.107595885089146e-05, "loss": 0.0677, "step": 32220 }, { "grad_norm": 0.4380852282047272, "learning_rate": 8.106300388675108e-05, "loss": 0.066, "step": 32230 }, { "grad_norm": 0.44915422797203064, "learning_rate": 8.105004552560731e-05, "loss": 0.0659, "step": 32240 }, { "grad_norm": 0.6195862293243408, "learning_rate": 8.103708376887724e-05, "loss": 0.0613, "step": 32250 }, { "grad_norm": 0.4620864987373352, "learning_rate": 8.102411861797836e-05, "loss": 0.0535, "step": 32260 }, { "grad_norm": 0.4617632031440735, "learning_rate": 8.101115007432851e-05, "loss": 0.0523, "step": 32270 }, { "grad_norm": 0.472964346408844, "learning_rate": 8.09981781393459e-05, "loss": 0.0633, "step": 32280 }, { "grad_norm": 0.7062981724739075, "learning_rate": 8.098520281444915e-05, "loss": 0.0537, "step": 32290 }, { "grad_norm": 0.5595211386680603, "learning_rate": 8.09722241010572e-05, "loss": 0.0505, "step": 32300 }, { "grad_norm": 0.7220812439918518, "learning_rate": 8.095924200058939e-05, "loss": 0.0576, "step": 32310 }, { "grad_norm": 0.4799515902996063, "learning_rate": 8.094625651446541e-05, "loss": 0.0517, "step": 32320 }, { "grad_norm": 0.38046854734420776, "learning_rate": 8.093326764410536e-05, "loss": 0.0672, "step": 32330 }, { "grad_norm": 0.5883759260177612, "learning_rate": 8.092027539092966e-05, "loss": 0.0623, "step": 32340 }, { "grad_norm": 0.5766520500183105, "learning_rate": 8.090727975635913e-05, "loss": 0.0643, "step": 32350 }, { "grad_norm": 0.6695365309715271, "learning_rate": 8.089428074181497e-05, "loss": 0.0631, "step": 32360 }, { "grad_norm": 0.4530394971370697, "learning_rate": 8.088127834871871e-05, "loss": 0.0523, "step": 32370 }, { "grad_norm": 1.0022375583648682, "learning_rate": 8.086827257849226e-05, "loss": 0.0579, "step": 32380 }, { "grad_norm": 0.7327443361282349, "learning_rate": 8.085526343255795e-05, "loss": 0.0471, "step": 32390 }, { "grad_norm": 2.983544111251831, "learning_rate": 8.084225091233842e-05, "loss": 0.0686, "step": 32400 }, { "grad_norm": 0.7980770468711853, "learning_rate": 8.082923501925668e-05, "loss": 0.0528, "step": 32410 }, { "grad_norm": 1.3209221363067627, "learning_rate": 8.081621575473617e-05, "loss": 0.0758, "step": 32420 }, { "grad_norm": 1.0472174882888794, "learning_rate": 8.080319312020064e-05, "loss": 0.075, "step": 32430 }, { "grad_norm": 0.6154917478561401, "learning_rate": 8.079016711707421e-05, "loss": 0.0712, "step": 32440 }, { "grad_norm": 0.582047164440155, "learning_rate": 8.077713774678139e-05, "loss": 0.076, "step": 32450 }, { "grad_norm": 0.7085886597633362, "learning_rate": 8.076410501074707e-05, "loss": 0.0609, "step": 32460 }, { "grad_norm": 0.5470997095108032, "learning_rate": 8.075106891039647e-05, "loss": 0.0702, "step": 32470 }, { "grad_norm": 0.6717728972434998, "learning_rate": 8.073802944715523e-05, "loss": 0.0695, "step": 32480 }, { "grad_norm": 0.5317618250846863, "learning_rate": 8.072498662244929e-05, "loss": 0.0589, "step": 32490 }, { "grad_norm": 0.4982012212276459, "learning_rate": 8.0711940437705e-05, "loss": 0.0594, "step": 32500 }, { "grad_norm": 0.45083028078079224, "learning_rate": 8.06988908943491e-05, "loss": 0.0496, "step": 32510 }, { "grad_norm": 0.7214428186416626, "learning_rate": 8.068583799380863e-05, "loss": 0.0506, "step": 32520 }, { "grad_norm": 0.7369725704193115, "learning_rate": 8.067278173751104e-05, "loss": 0.0678, "step": 32530 }, { "grad_norm": 0.5750062465667725, "learning_rate": 8.065972212688417e-05, "loss": 0.0626, "step": 32540 }, { "grad_norm": 1.973327398300171, "learning_rate": 8.064665916335618e-05, "loss": 0.0746, "step": 32550 }, { "grad_norm": 0.6474663019180298, "learning_rate": 8.063359284835564e-05, "loss": 0.0678, "step": 32560 }, { "grad_norm": 0.49793824553489685, "learning_rate": 8.062052318331142e-05, "loss": 0.0734, "step": 32570 }, { "grad_norm": 0.592557966709137, "learning_rate": 8.060745016965283e-05, "loss": 0.0678, "step": 32580 }, { "grad_norm": 0.6167080402374268, "learning_rate": 8.059437380880952e-05, "loss": 0.0545, "step": 32590 }, { "grad_norm": 0.5147932767868042, "learning_rate": 8.058129410221146e-05, "loss": 0.0602, "step": 32600 }, { "grad_norm": 0.5826452970504761, "learning_rate": 8.056821105128908e-05, "loss": 0.0616, "step": 32610 }, { "grad_norm": 0.7077281475067139, "learning_rate": 8.05551246574731e-05, "loss": 0.0633, "step": 32620 }, { "grad_norm": 0.9841307997703552, "learning_rate": 8.05420349221946e-05, "loss": 0.0648, "step": 32630 }, { "grad_norm": 0.5032424330711365, "learning_rate": 8.05289418468851e-05, "loss": 0.0483, "step": 32640 }, { "grad_norm": 0.4320400357246399, "learning_rate": 8.051584543297642e-05, "loss": 0.0686, "step": 32650 }, { "grad_norm": 0.5254572033882141, "learning_rate": 8.050274568190074e-05, "loss": 0.061, "step": 32660 }, { "grad_norm": 0.9165874719619751, "learning_rate": 8.048964259509067e-05, "loss": 0.0588, "step": 32670 }, { "grad_norm": 0.8515697717666626, "learning_rate": 8.047653617397914e-05, "loss": 0.0604, "step": 32680 }, { "grad_norm": 0.6157923340797424, "learning_rate": 8.046342641999941e-05, "loss": 0.0557, "step": 32690 }, { "grad_norm": 0.512851357460022, "learning_rate": 8.045031333458517e-05, "loss": 0.0606, "step": 32700 }, { "grad_norm": 0.7117455005645752, "learning_rate": 8.043719691917047e-05, "loss": 0.0594, "step": 32710 }, { "grad_norm": 0.4815244674682617, "learning_rate": 8.042407717518966e-05, "loss": 0.057, "step": 32720 }, { "grad_norm": 0.8371496796607971, "learning_rate": 8.041095410407751e-05, "loss": 0.0554, "step": 32730 }, { "grad_norm": 0.36789369583129883, "learning_rate": 8.039782770726913e-05, "loss": 0.0503, "step": 32740 }, { "grad_norm": 0.37544170022010803, "learning_rate": 8.038469798620004e-05, "loss": 0.0606, "step": 32750 }, { "grad_norm": 0.5017421841621399, "learning_rate": 8.037156494230604e-05, "loss": 0.061, "step": 32760 }, { "grad_norm": 0.5243842005729675, "learning_rate": 8.035842857702338e-05, "loss": 0.0477, "step": 32770 }, { "grad_norm": 0.6204031705856323, "learning_rate": 8.03452888917886e-05, "loss": 0.0625, "step": 32780 }, { "grad_norm": 0.4380861222743988, "learning_rate": 8.033214588803866e-05, "loss": 0.0583, "step": 32790 }, { "grad_norm": 0.5538873076438904, "learning_rate": 8.031899956721083e-05, "loss": 0.0498, "step": 32800 }, { "grad_norm": 0.48366567492485046, "learning_rate": 8.030584993074282e-05, "loss": 0.0599, "step": 32810 }, { "grad_norm": 0.43511515855789185, "learning_rate": 8.02926969800726e-05, "loss": 0.0544, "step": 32820 }, { "grad_norm": 0.5634382963180542, "learning_rate": 8.027954071663859e-05, "loss": 0.0745, "step": 32830 }, { "grad_norm": 0.6337878704071045, "learning_rate": 8.026638114187954e-05, "loss": 0.0713, "step": 32840 }, { "grad_norm": 0.6279234290122986, "learning_rate": 8.025321825723456e-05, "loss": 0.0637, "step": 32850 }, { "grad_norm": 0.5519091486930847, "learning_rate": 8.02400520641431e-05, "loss": 0.0821, "step": 32860 }, { "grad_norm": 0.4331096410751343, "learning_rate": 8.022688256404501e-05, "loss": 0.0586, "step": 32870 }, { "grad_norm": 0.8189354538917542, "learning_rate": 8.02137097583805e-05, "loss": 0.0678, "step": 32880 }, { "grad_norm": 0.7407180070877075, "learning_rate": 8.02005336485901e-05, "loss": 0.0634, "step": 32890 }, { "grad_norm": 0.830105185508728, "learning_rate": 8.018735423611476e-05, "loss": 0.0733, "step": 32900 }, { "grad_norm": 0.8280872702598572, "learning_rate": 8.017417152239574e-05, "loss": 0.0645, "step": 32910 }, { "grad_norm": 0.8645308613777161, "learning_rate": 8.01609855088747e-05, "loss": 0.0701, "step": 32920 }, { "grad_norm": 0.5992599129676819, "learning_rate": 8.014779619699362e-05, "loss": 0.0566, "step": 32930 }, { "grad_norm": 0.5058514475822449, "learning_rate": 8.013460358819489e-05, "loss": 0.0536, "step": 32940 }, { "grad_norm": 0.7037409543991089, "learning_rate": 8.01214076839212e-05, "loss": 0.0625, "step": 32950 }, { "grad_norm": 0.57801353931427, "learning_rate": 8.010820848561565e-05, "loss": 0.0578, "step": 32960 }, { "grad_norm": 0.4878549873828888, "learning_rate": 8.009500599472171e-05, "loss": 0.0533, "step": 32970 }, { "grad_norm": 0.6551867723464966, "learning_rate": 8.008180021268314e-05, "loss": 0.0542, "step": 32980 }, { "grad_norm": 0.4705713093280792, "learning_rate": 8.006859114094414e-05, "loss": 0.0563, "step": 32990 }, { "grad_norm": 0.6312029361724854, "learning_rate": 8.005537878094921e-05, "loss": 0.0591, "step": 33000 }, { "grad_norm": 0.6050772666931152, "learning_rate": 8.004216313414323e-05, "loss": 0.064, "step": 33010 }, { "grad_norm": 0.36997732520103455, "learning_rate": 8.002894420197149e-05, "loss": 0.0508, "step": 33020 }, { "grad_norm": 0.9097073078155518, "learning_rate": 8.001572198587954e-05, "loss": 0.053, "step": 33030 }, { "grad_norm": 1.1083859205245972, "learning_rate": 8.000249648731338e-05, "loss": 0.0664, "step": 33040 }, { "grad_norm": 0.5568991303443909, "learning_rate": 7.998926770771928e-05, "loss": 0.0592, "step": 33050 }, { "grad_norm": 0.6283136010169983, "learning_rate": 7.997603564854397e-05, "loss": 0.0747, "step": 33060 }, { "grad_norm": 0.5179421305656433, "learning_rate": 7.996280031123448e-05, "loss": 0.059, "step": 33070 }, { "grad_norm": 0.5548040270805359, "learning_rate": 7.994956169723818e-05, "loss": 0.0757, "step": 33080 }, { "grad_norm": 0.5292304754257202, "learning_rate": 7.993631980800285e-05, "loss": 0.0548, "step": 33090 }, { "grad_norm": 0.5478576421737671, "learning_rate": 7.992307464497659e-05, "loss": 0.0735, "step": 33100 }, { "grad_norm": 0.385050505399704, "learning_rate": 7.990982620960787e-05, "loss": 0.0473, "step": 33110 }, { "grad_norm": 1.508987545967102, "learning_rate": 7.989657450334554e-05, "loss": 0.0673, "step": 33120 }, { "grad_norm": 0.5395138263702393, "learning_rate": 7.988331952763877e-05, "loss": 0.0564, "step": 33130 }, { "grad_norm": 0.5912910103797913, "learning_rate": 7.987006128393709e-05, "loss": 0.0586, "step": 33140 }, { "grad_norm": 0.6408900022506714, "learning_rate": 7.985679977369043e-05, "loss": 0.0526, "step": 33150 }, { "grad_norm": 0.6693406105041504, "learning_rate": 7.984353499834902e-05, "loss": 0.0483, "step": 33160 }, { "grad_norm": 0.45313870906829834, "learning_rate": 7.983026695936351e-05, "loss": 0.0564, "step": 33170 }, { "grad_norm": 0.5467361211776733, "learning_rate": 7.981699565818486e-05, "loss": 0.0551, "step": 33180 }, { "grad_norm": 0.7137870788574219, "learning_rate": 7.980372109626437e-05, "loss": 0.0592, "step": 33190 }, { "grad_norm": 0.5941634178161621, "learning_rate": 7.979044327505375e-05, "loss": 0.0503, "step": 33200 }, { "grad_norm": 0.5411359667778015, "learning_rate": 7.977716219600506e-05, "loss": 0.0777, "step": 33210 }, { "grad_norm": 0.5983999967575073, "learning_rate": 7.97638778605707e-05, "loss": 0.0617, "step": 33220 }, { "grad_norm": 0.7011352777481079, "learning_rate": 7.975059027020338e-05, "loss": 0.0772, "step": 33230 }, { "grad_norm": 0.7033697366714478, "learning_rate": 7.973729942635623e-05, "loss": 0.0679, "step": 33240 }, { "grad_norm": 0.6053556203842163, "learning_rate": 7.972400533048273e-05, "loss": 0.0674, "step": 33250 }, { "grad_norm": 0.5282509326934814, "learning_rate": 7.97107079840367e-05, "loss": 0.064, "step": 33260 }, { "grad_norm": 0.7225149869918823, "learning_rate": 7.969740738847231e-05, "loss": 0.0586, "step": 33270 }, { "grad_norm": 0.5453532338142395, "learning_rate": 7.968410354524411e-05, "loss": 0.0542, "step": 33280 }, { "grad_norm": 0.31824320554733276, "learning_rate": 7.967079645580697e-05, "loss": 0.0546, "step": 33290 }, { "grad_norm": 0.5460740327835083, "learning_rate": 7.965748612161612e-05, "loss": 0.0503, "step": 33300 }, { "grad_norm": 0.6889947056770325, "learning_rate": 7.96441725441272e-05, "loss": 0.0703, "step": 33310 }, { "grad_norm": 0.9192836284637451, "learning_rate": 7.963085572479614e-05, "loss": 0.0624, "step": 33320 }, { "grad_norm": 0.47843503952026367, "learning_rate": 7.961753566507924e-05, "loss": 0.0521, "step": 33330 }, { "grad_norm": 0.5801183581352234, "learning_rate": 7.960421236643316e-05, "loss": 0.0604, "step": 33340 }, { "grad_norm": 0.49161696434020996, "learning_rate": 7.959088583031496e-05, "loss": 0.0508, "step": 33350 }, { "grad_norm": 0.6040979027748108, "learning_rate": 7.957755605818194e-05, "loss": 0.0659, "step": 33360 }, { "grad_norm": 0.42923182249069214, "learning_rate": 7.956422305149185e-05, "loss": 0.0626, "step": 33370 }, { "grad_norm": 0.593882143497467, "learning_rate": 7.95508868117028e-05, "loss": 0.072, "step": 33380 }, { "grad_norm": 0.6170177459716797, "learning_rate": 7.953754734027318e-05, "loss": 0.0698, "step": 33390 }, { "grad_norm": 0.3516123592853546, "learning_rate": 7.952420463866182e-05, "loss": 0.0584, "step": 33400 }, { "grad_norm": 0.8761959075927734, "learning_rate": 7.951085870832782e-05, "loss": 0.0806, "step": 33410 }, { "grad_norm": 0.4401461184024811, "learning_rate": 7.949750955073067e-05, "loss": 0.0584, "step": 33420 }, { "grad_norm": 0.8025989532470703, "learning_rate": 7.948415716733022e-05, "loss": 0.0708, "step": 33430 }, { "grad_norm": 0.6890272498130798, "learning_rate": 7.947080155958669e-05, "loss": 0.0686, "step": 33440 }, { "grad_norm": 0.6562023758888245, "learning_rate": 7.94574427289606e-05, "loss": 0.0559, "step": 33450 }, { "grad_norm": 0.6618228554725647, "learning_rate": 7.944408067691284e-05, "loss": 0.064, "step": 33460 }, { "grad_norm": 0.7461862564086914, "learning_rate": 7.943071540490473e-05, "loss": 0.0591, "step": 33470 }, { "grad_norm": 0.5274271368980408, "learning_rate": 7.94173469143978e-05, "loss": 0.0599, "step": 33480 }, { "grad_norm": 0.29373452067375183, "learning_rate": 7.940397520685406e-05, "loss": 0.0513, "step": 33490 }, { "grad_norm": 1.1975047588348389, "learning_rate": 7.939060028373577e-05, "loss": 0.0678, "step": 33500 }, { "grad_norm": 0.7162962555885315, "learning_rate": 7.937722214650565e-05, "loss": 0.0727, "step": 33510 }, { "grad_norm": 0.49214839935302734, "learning_rate": 7.936384079662666e-05, "loss": 0.0646, "step": 33520 }, { "grad_norm": 0.5917534232139587, "learning_rate": 7.93504562355622e-05, "loss": 0.0613, "step": 33530 }, { "grad_norm": 0.4982898235321045, "learning_rate": 7.933706846477599e-05, "loss": 0.0593, "step": 33540 }, { "grad_norm": 0.4194295108318329, "learning_rate": 7.932367748573206e-05, "loss": 0.0602, "step": 33550 }, { "grad_norm": 0.7222688794136047, "learning_rate": 7.931028329989485e-05, "loss": 0.0525, "step": 33560 }, { "grad_norm": 0.6069767475128174, "learning_rate": 7.929688590872913e-05, "loss": 0.0782, "step": 33570 }, { "grad_norm": 0.4714096188545227, "learning_rate": 7.928348531370003e-05, "loss": 0.0482, "step": 33580 }, { "grad_norm": 0.6936464309692383, "learning_rate": 7.927008151627297e-05, "loss": 0.0623, "step": 33590 }, { "grad_norm": 0.5374143719673157, "learning_rate": 7.925667451791383e-05, "loss": 0.0562, "step": 33600 }, { "grad_norm": 0.8038157820701599, "learning_rate": 7.924326432008874e-05, "loss": 0.0623, "step": 33610 }, { "grad_norm": 0.46002209186553955, "learning_rate": 7.922985092426422e-05, "loss": 0.0623, "step": 33620 }, { "grad_norm": 0.6885645985603333, "learning_rate": 7.921643433190717e-05, "loss": 0.0737, "step": 33630 }, { "grad_norm": 1.5971355438232422, "learning_rate": 7.920301454448478e-05, "loss": 0.0693, "step": 33640 }, { "grad_norm": 0.4369424283504486, "learning_rate": 7.918959156346461e-05, "loss": 0.0508, "step": 33650 }, { "grad_norm": 0.6685049533843994, "learning_rate": 7.91761653903146e-05, "loss": 0.0522, "step": 33660 }, { "grad_norm": 0.5072528719902039, "learning_rate": 7.916273602650302e-05, "loss": 0.0695, "step": 33670 }, { "grad_norm": 0.761286735534668, "learning_rate": 7.914930347349847e-05, "loss": 0.0599, "step": 33680 }, { "grad_norm": 0.6374523639678955, "learning_rate": 7.913586773276992e-05, "loss": 0.0686, "step": 33690 }, { "grad_norm": 1.3705322742462158, "learning_rate": 7.912242880578667e-05, "loss": 0.0577, "step": 33700 }, { "grad_norm": 0.6541709899902344, "learning_rate": 7.910898669401839e-05, "loss": 0.0615, "step": 33710 }, { "grad_norm": 0.3483155071735382, "learning_rate": 7.909554139893511e-05, "loss": 0.0543, "step": 33720 }, { "grad_norm": 0.6966603398323059, "learning_rate": 7.908209292200715e-05, "loss": 0.0527, "step": 33730 }, { "grad_norm": 0.6311540007591248, "learning_rate": 7.906864126470523e-05, "loss": 0.0523, "step": 33740 }, { "grad_norm": 0.9448833465576172, "learning_rate": 7.905518642850041e-05, "loss": 0.0658, "step": 33750 }, { "grad_norm": 0.413742333650589, "learning_rate": 7.904172841486409e-05, "loss": 0.0646, "step": 33760 }, { "grad_norm": 0.9168839454650879, "learning_rate": 7.902826722526801e-05, "loss": 0.064, "step": 33770 }, { "grad_norm": 0.4952515959739685, "learning_rate": 7.901480286118427e-05, "loss": 0.0644, "step": 33780 }, { "grad_norm": 0.49730005860328674, "learning_rate": 7.900133532408531e-05, "loss": 0.058, "step": 33790 }, { "grad_norm": 0.546584963798523, "learning_rate": 7.898786461544395e-05, "loss": 0.0599, "step": 33800 }, { "grad_norm": 0.7584212422370911, "learning_rate": 7.897439073673325e-05, "loss": 0.0686, "step": 33810 }, { "grad_norm": 0.6144772171974182, "learning_rate": 7.896091368942677e-05, "loss": 0.0708, "step": 33820 }, { "grad_norm": 0.7348586320877075, "learning_rate": 7.894743347499832e-05, "loss": 0.0525, "step": 33830 }, { "grad_norm": 0.3909585177898407, "learning_rate": 7.893395009492203e-05, "loss": 0.0619, "step": 33840 }, { "grad_norm": 0.4282349944114685, "learning_rate": 7.892046355067248e-05, "loss": 0.0561, "step": 33850 }, { "grad_norm": 0.6319012641906738, "learning_rate": 7.890697384372451e-05, "loss": 0.0552, "step": 33860 }, { "grad_norm": 0.4868464171886444, "learning_rate": 7.889348097555336e-05, "loss": 0.0524, "step": 33870 }, { "grad_norm": 0.6101924777030945, "learning_rate": 7.887998494763455e-05, "loss": 0.0616, "step": 33880 }, { "grad_norm": 0.4807738959789276, "learning_rate": 7.886648576144404e-05, "loss": 0.069, "step": 33890 }, { "grad_norm": 0.42077016830444336, "learning_rate": 7.885298341845802e-05, "loss": 0.056, "step": 33900 }, { "grad_norm": 0.5611105561256409, "learning_rate": 7.883947792015311e-05, "loss": 0.0586, "step": 33910 }, { "grad_norm": 0.6422563195228577, "learning_rate": 7.882596926800628e-05, "loss": 0.0857, "step": 33920 }, { "grad_norm": 0.45785534381866455, "learning_rate": 7.881245746349477e-05, "loss": 0.0682, "step": 33930 }, { "grad_norm": 0.4387468099594116, "learning_rate": 7.879894250809623e-05, "loss": 0.0595, "step": 33940 }, { "grad_norm": 0.7658165693283081, "learning_rate": 7.878542440328865e-05, "loss": 0.0753, "step": 33950 }, { "grad_norm": 0.4884980022907257, "learning_rate": 7.877190315055031e-05, "loss": 0.0439, "step": 33960 }, { "grad_norm": 0.48236560821533203, "learning_rate": 7.875837875135991e-05, "loss": 0.0511, "step": 33970 }, { "grad_norm": 0.5447328686714172, "learning_rate": 7.874485120719646e-05, "loss": 0.0677, "step": 33980 }, { "grad_norm": 0.41498610377311707, "learning_rate": 7.873132051953928e-05, "loss": 0.0569, "step": 33990 }, { "grad_norm": 0.8391163945198059, "learning_rate": 7.87177866898681e-05, "loss": 0.065, "step": 34000 }, { "grad_norm": 0.3977004587650299, "learning_rate": 7.870424971966294e-05, "loss": 0.0534, "step": 34010 }, { "grad_norm": 0.4462828040122986, "learning_rate": 7.869070961040419e-05, "loss": 0.0492, "step": 34020 }, { "grad_norm": 0.5268566608428955, "learning_rate": 7.867716636357257e-05, "loss": 0.0637, "step": 34030 }, { "grad_norm": 0.4812169075012207, "learning_rate": 7.866361998064915e-05, "loss": 0.0623, "step": 34040 }, { "grad_norm": 0.5851898789405823, "learning_rate": 7.865007046311534e-05, "loss": 0.0571, "step": 34050 }, { "grad_norm": 0.7649716138839722, "learning_rate": 7.86365178124529e-05, "loss": 0.0589, "step": 34060 }, { "grad_norm": 0.5058276653289795, "learning_rate": 7.862296203014394e-05, "loss": 0.0596, "step": 34070 }, { "grad_norm": 0.6655906438827515, "learning_rate": 7.860940311767088e-05, "loss": 0.0693, "step": 34080 }, { "grad_norm": 0.5940271615982056, "learning_rate": 7.85958410765165e-05, "loss": 0.0589, "step": 34090 }, { "grad_norm": 0.9324275851249695, "learning_rate": 7.858227590816394e-05, "loss": 0.0778, "step": 34100 }, { "grad_norm": 0.49981066584587097, "learning_rate": 7.856870761409664e-05, "loss": 0.061, "step": 34110 }, { "grad_norm": 0.6935362219810486, "learning_rate": 7.855513619579846e-05, "loss": 0.0661, "step": 34120 }, { "grad_norm": 0.6775828003883362, "learning_rate": 7.85415616547535e-05, "loss": 0.0591, "step": 34130 }, { "grad_norm": 0.42510920763015747, "learning_rate": 7.852798399244627e-05, "loss": 0.0474, "step": 34140 }, { "grad_norm": 0.47470661997795105, "learning_rate": 7.851440321036161e-05, "loss": 0.0532, "step": 34150 }, { "grad_norm": 0.4053608477115631, "learning_rate": 7.850081930998468e-05, "loss": 0.0579, "step": 34160 }, { "grad_norm": 0.9006706476211548, "learning_rate": 7.8487232292801e-05, "loss": 0.0694, "step": 34170 }, { "grad_norm": 0.8468952775001526, "learning_rate": 7.847364216029642e-05, "loss": 0.0697, "step": 34180 }, { "grad_norm": 0.5269449353218079, "learning_rate": 7.846004891395716e-05, "loss": 0.0569, "step": 34190 }, { "grad_norm": 0.7114869952201843, "learning_rate": 7.844645255526972e-05, "loss": 0.0754, "step": 34200 }, { "grad_norm": 0.6202364563941956, "learning_rate": 7.843285308572101e-05, "loss": 0.0655, "step": 34210 }, { "grad_norm": 0.5408256649971008, "learning_rate": 7.841925050679823e-05, "loss": 0.047, "step": 34220 }, { "grad_norm": 1.1178292036056519, "learning_rate": 7.840564481998895e-05, "loss": 0.065, "step": 34230 }, { "grad_norm": 1.4284214973449707, "learning_rate": 7.839203602678105e-05, "loss": 0.0604, "step": 34240 }, { "grad_norm": 0.6114625930786133, "learning_rate": 7.837842412866279e-05, "loss": 0.0702, "step": 34250 }, { "grad_norm": 0.3780896067619324, "learning_rate": 7.836480912712272e-05, "loss": 0.0617, "step": 34260 }, { "grad_norm": 0.6659774780273438, "learning_rate": 7.835119102364976e-05, "loss": 0.0526, "step": 34270 }, { "grad_norm": 0.5631633400917053, "learning_rate": 7.83375698197332e-05, "loss": 0.0509, "step": 34280 }, { "grad_norm": 0.5023578405380249, "learning_rate": 7.83239455168626e-05, "loss": 0.0579, "step": 34290 }, { "grad_norm": 0.47594740986824036, "learning_rate": 7.83103181165279e-05, "loss": 0.0599, "step": 34300 }, { "grad_norm": 0.7967323660850525, "learning_rate": 7.829668762021937e-05, "loss": 0.0711, "step": 34310 }, { "grad_norm": 0.8257389068603516, "learning_rate": 7.828305402942764e-05, "loss": 0.058, "step": 34320 }, { "grad_norm": 1.0901217460632324, "learning_rate": 7.826941734564363e-05, "loss": 0.0523, "step": 34330 }, { "grad_norm": 0.43226972222328186, "learning_rate": 7.825577757035865e-05, "loss": 0.0618, "step": 34340 }, { "grad_norm": 0.7060384154319763, "learning_rate": 7.824213470506431e-05, "loss": 0.0591, "step": 34350 }, { "grad_norm": 0.8250828981399536, "learning_rate": 7.822848875125257e-05, "loss": 0.0561, "step": 34360 }, { "grad_norm": 0.46589893102645874, "learning_rate": 7.821483971041576e-05, "loss": 0.0623, "step": 34370 }, { "grad_norm": 0.3651362955570221, "learning_rate": 7.820118758404649e-05, "loss": 0.053, "step": 34380 }, { "grad_norm": 0.38777676224708557, "learning_rate": 7.818753237363776e-05, "loss": 0.0583, "step": 34390 }, { "grad_norm": 0.5587208867073059, "learning_rate": 7.817387408068286e-05, "loss": 0.065, "step": 34400 }, { "grad_norm": 0.4003011882305145, "learning_rate": 7.816021270667544e-05, "loss": 0.0592, "step": 34410 }, { "grad_norm": 0.6860589385032654, "learning_rate": 7.81465482531095e-05, "loss": 0.0525, "step": 34420 }, { "grad_norm": 0.6813696026802063, "learning_rate": 7.813288072147938e-05, "loss": 0.0631, "step": 34430 }, { "grad_norm": 0.6989443302154541, "learning_rate": 7.811921011327972e-05, "loss": 0.0618, "step": 34440 }, { "grad_norm": 0.31851187348365784, "learning_rate": 7.810553643000549e-05, "loss": 0.0524, "step": 34450 }, { "grad_norm": 0.6338070631027222, "learning_rate": 7.809185967315206e-05, "loss": 0.0551, "step": 34460 }, { "grad_norm": 0.40000173449516296, "learning_rate": 7.80781798442151e-05, "loss": 0.0619, "step": 34470 }, { "grad_norm": 0.7706928253173828, "learning_rate": 7.806449694469061e-05, "loss": 0.047, "step": 34480 }, { "grad_norm": 0.4803677201271057, "learning_rate": 7.805081097607492e-05, "loss": 0.0626, "step": 34490 }, { "grad_norm": 0.5402969121932983, "learning_rate": 7.803712193986474e-05, "loss": 0.0564, "step": 34500 }, { "grad_norm": 0.5626657009124756, "learning_rate": 7.802342983755702e-05, "loss": 0.0532, "step": 34510 }, { "grad_norm": 0.7007023692131042, "learning_rate": 7.800973467064918e-05, "loss": 0.0667, "step": 34520 }, { "grad_norm": 0.7506843209266663, "learning_rate": 7.799603644063884e-05, "loss": 0.0567, "step": 34530 }, { "grad_norm": 0.5163552761077881, "learning_rate": 7.798233514902405e-05, "loss": 0.0482, "step": 34540 }, { "grad_norm": 0.5251017808914185, "learning_rate": 7.796863079730318e-05, "loss": 0.0638, "step": 34550 }, { "grad_norm": 0.48171624541282654, "learning_rate": 7.795492338697488e-05, "loss": 0.0531, "step": 34560 }, { "grad_norm": 0.4692471921443939, "learning_rate": 7.794121291953819e-05, "loss": 0.0615, "step": 34570 }, { "grad_norm": 0.754643976688385, "learning_rate": 7.792749939649246e-05, "loss": 0.073, "step": 34580 }, { "grad_norm": 0.4809252917766571, "learning_rate": 7.79137828193374e-05, "loss": 0.0495, "step": 34590 }, { "grad_norm": 0.5285003781318665, "learning_rate": 7.790006318957301e-05, "loss": 0.0708, "step": 34600 }, { "grad_norm": 0.6760608553886414, "learning_rate": 7.788634050869965e-05, "loss": 0.0574, "step": 34610 }, { "grad_norm": 0.6827071309089661, "learning_rate": 7.787261477821803e-05, "loss": 0.0588, "step": 34620 }, { "grad_norm": 0.611402690410614, "learning_rate": 7.785888599962916e-05, "loss": 0.0481, "step": 34630 }, { "grad_norm": 0.8440324664115906, "learning_rate": 7.784515417443439e-05, "loss": 0.0546, "step": 34640 }, { "grad_norm": 0.5182258486747742, "learning_rate": 7.783141930413545e-05, "loss": 0.0564, "step": 34650 }, { "grad_norm": 0.5040369629859924, "learning_rate": 7.78176813902343e-05, "loss": 0.0639, "step": 34660 }, { "grad_norm": 0.5329881310462952, "learning_rate": 7.780394043423336e-05, "loss": 0.0517, "step": 34670 }, { "grad_norm": 0.6762109398841858, "learning_rate": 7.77901964376353e-05, "loss": 0.0653, "step": 34680 }, { "grad_norm": 0.6270642280578613, "learning_rate": 7.777644940194316e-05, "loss": 0.0564, "step": 34690 }, { "grad_norm": 0.49662381410598755, "learning_rate": 7.776269932866023e-05, "loss": 0.0621, "step": 34700 }, { "grad_norm": 2.129028797149658, "learning_rate": 7.774894621929026e-05, "loss": 0.0613, "step": 34710 }, { "grad_norm": 0.4848383367061615, "learning_rate": 7.773519007533725e-05, "loss": 0.0727, "step": 34720 }, { "grad_norm": 0.33896440267562866, "learning_rate": 7.772143089830556e-05, "loss": 0.0505, "step": 34730 }, { "grad_norm": 0.38742735981941223, "learning_rate": 7.770766868969985e-05, "loss": 0.055, "step": 34740 }, { "grad_norm": 0.38518884778022766, "learning_rate": 7.769390345102518e-05, "loss": 0.0683, "step": 34750 }, { "grad_norm": 0.6428626775741577, "learning_rate": 7.768013518378683e-05, "loss": 0.066, "step": 34760 }, { "grad_norm": 0.45874103903770447, "learning_rate": 7.766636388949053e-05, "loss": 0.0435, "step": 34770 }, { "grad_norm": 0.5414170026779175, "learning_rate": 7.765258956964229e-05, "loss": 0.0669, "step": 34780 }, { "grad_norm": 0.46733614802360535, "learning_rate": 7.76388122257484e-05, "loss": 0.0587, "step": 34790 }, { "grad_norm": 0.5623935461044312, "learning_rate": 7.762503185931558e-05, "loss": 0.0546, "step": 34800 }, { "grad_norm": 0.48423752188682556, "learning_rate": 7.76112484718508e-05, "loss": 0.0624, "step": 34810 }, { "grad_norm": 0.493928998708725, "learning_rate": 7.75974620648614e-05, "loss": 0.0615, "step": 34820 }, { "grad_norm": 0.7452479004859924, "learning_rate": 7.758367263985503e-05, "loss": 0.0546, "step": 34830 }, { "grad_norm": 0.5239875316619873, "learning_rate": 7.75698801983397e-05, "loss": 0.0541, "step": 34840 }, { "grad_norm": 0.42448312044143677, "learning_rate": 7.755608474182372e-05, "loss": 0.0618, "step": 34850 }, { "grad_norm": 0.9499935507774353, "learning_rate": 7.754228627181574e-05, "loss": 0.0621, "step": 34860 }, { "grad_norm": 0.5955123901367188, "learning_rate": 7.752848478982476e-05, "loss": 0.0593, "step": 34870 }, { "grad_norm": 0.5386947989463806, "learning_rate": 7.751468029736006e-05, "loss": 0.0549, "step": 34880 }, { "grad_norm": 0.5528023838996887, "learning_rate": 7.750087279593129e-05, "loss": 0.0609, "step": 34890 }, { "grad_norm": 0.40995854139328003, "learning_rate": 7.748706228704843e-05, "loss": 0.0451, "step": 34900 }, { "grad_norm": 0.5131848454475403, "learning_rate": 7.747324877222176e-05, "loss": 0.0656, "step": 34910 }, { "grad_norm": 0.5752193927764893, "learning_rate": 7.745943225296188e-05, "loss": 0.0743, "step": 34920 }, { "grad_norm": 0.6011455059051514, "learning_rate": 7.744561273077981e-05, "loss": 0.049, "step": 34930 }, { "grad_norm": 0.4226621985435486, "learning_rate": 7.743179020718678e-05, "loss": 0.0568, "step": 34940 }, { "grad_norm": 0.5892205834388733, "learning_rate": 7.741796468369443e-05, "loss": 0.0569, "step": 34950 }, { "grad_norm": 0.529667317867279, "learning_rate": 7.740413616181466e-05, "loss": 0.0678, "step": 34960 }, { "grad_norm": 0.760266900062561, "learning_rate": 7.739030464305978e-05, "loss": 0.0743, "step": 34970 }, { "grad_norm": 1.032281756401062, "learning_rate": 7.737647012894235e-05, "loss": 0.074, "step": 34980 }, { "grad_norm": 0.7901782393455505, "learning_rate": 7.736263262097532e-05, "loss": 0.0671, "step": 34990 }, { "grad_norm": 0.34999290108680725, "learning_rate": 7.734879212067192e-05, "loss": 0.0644, "step": 35000 }, { "grad_norm": 0.573792040348053, "learning_rate": 7.733494862954573e-05, "loss": 0.0555, "step": 35010 }, { "grad_norm": 0.525324285030365, "learning_rate": 7.732110214911066e-05, "loss": 0.0572, "step": 35020 }, { "grad_norm": 0.7890287041664124, "learning_rate": 7.730725268088094e-05, "loss": 0.0703, "step": 35030 }, { "grad_norm": 0.5277650356292725, "learning_rate": 7.729340022637112e-05, "loss": 0.0586, "step": 35040 }, { "grad_norm": 0.6065695285797119, "learning_rate": 7.727954478709607e-05, "loss": 0.0661, "step": 35050 }, { "grad_norm": 0.6796467900276184, "learning_rate": 7.726568636457103e-05, "loss": 0.0666, "step": 35060 }, { "grad_norm": 0.7338840961456299, "learning_rate": 7.725182496031153e-05, "loss": 0.0662, "step": 35070 }, { "grad_norm": 0.6641966700553894, "learning_rate": 7.723796057583341e-05, "loss": 0.0677, "step": 35080 }, { "grad_norm": 0.6530977487564087, "learning_rate": 7.722409321265287e-05, "loss": 0.0693, "step": 35090 }, { "grad_norm": 0.3265199661254883, "learning_rate": 7.721022287228645e-05, "loss": 0.0737, "step": 35100 }, { "grad_norm": 0.7707527875900269, "learning_rate": 7.719634955625094e-05, "loss": 0.0481, "step": 35110 }, { "grad_norm": 1.0815304517745972, "learning_rate": 7.718247326606355e-05, "loss": 0.0548, "step": 35120 }, { "grad_norm": 0.5795065760612488, "learning_rate": 7.716859400324175e-05, "loss": 0.0596, "step": 35130 }, { "grad_norm": 1.3759492635726929, "learning_rate": 7.715471176930335e-05, "loss": 0.0836, "step": 35140 }, { "grad_norm": 0.6413960456848145, "learning_rate": 7.714082656576651e-05, "loss": 0.0665, "step": 35150 }, { "grad_norm": 0.509839653968811, "learning_rate": 7.712693839414968e-05, "loss": 0.057, "step": 35160 }, { "grad_norm": 0.4073674976825714, "learning_rate": 7.711304725597164e-05, "loss": 0.0459, "step": 35170 }, { "grad_norm": 0.5609623789787292, "learning_rate": 7.709915315275151e-05, "loss": 0.0678, "step": 35180 }, { "grad_norm": 0.5601547956466675, "learning_rate": 7.708525608600876e-05, "loss": 0.0534, "step": 35190 }, { "grad_norm": 0.4387545883655548, "learning_rate": 7.707135605726311e-05, "loss": 0.0637, "step": 35200 }, { "grad_norm": 0.6054499745368958, "learning_rate": 7.705745306803466e-05, "loss": 0.0824, "step": 35210 }, { "grad_norm": 0.5158323049545288, "learning_rate": 7.704354711984383e-05, "loss": 0.0545, "step": 35220 }, { "grad_norm": 0.516686737537384, "learning_rate": 7.702963821421133e-05, "loss": 0.055, "step": 35230 }, { "grad_norm": 0.7722005844116211, "learning_rate": 7.701572635265826e-05, "loss": 0.0577, "step": 35240 }, { "grad_norm": 0.5367758870124817, "learning_rate": 7.700181153670596e-05, "loss": 0.0448, "step": 35250 }, { "grad_norm": 0.5210641622543335, "learning_rate": 7.698789376787616e-05, "loss": 0.0615, "step": 35260 }, { "grad_norm": 2.371680498123169, "learning_rate": 7.697397304769085e-05, "loss": 0.0586, "step": 35270 }, { "grad_norm": 0.5160402059555054, "learning_rate": 7.696004937767241e-05, "loss": 0.0656, "step": 35280 }, { "grad_norm": 0.6171307563781738, "learning_rate": 7.694612275934352e-05, "loss": 0.0634, "step": 35290 }, { "grad_norm": 0.5731187462806702, "learning_rate": 7.693219319422714e-05, "loss": 0.0622, "step": 35300 }, { "grad_norm": 0.586962878704071, "learning_rate": 7.69182606838466e-05, "loss": 0.057, "step": 35310 }, { "grad_norm": 0.554574728012085, "learning_rate": 7.690432522972558e-05, "loss": 0.0549, "step": 35320 }, { "grad_norm": 0.38907578587532043, "learning_rate": 7.689038683338796e-05, "loss": 0.0553, "step": 35330 }, { "grad_norm": 0.4903872609138489, "learning_rate": 7.687644549635808e-05, "loss": 0.0573, "step": 35340 }, { "grad_norm": 1.290209174156189, "learning_rate": 7.686250122016053e-05, "loss": 0.0529, "step": 35350 }, { "grad_norm": 1.1501718759536743, "learning_rate": 7.684855400632022e-05, "loss": 0.063, "step": 35360 }, { "grad_norm": 0.842215895652771, "learning_rate": 7.683460385636243e-05, "loss": 0.0445, "step": 35370 }, { "grad_norm": 0.6925368309020996, "learning_rate": 7.68206507718127e-05, "loss": 0.0589, "step": 35380 }, { "grad_norm": 0.6148405075073242, "learning_rate": 7.680669475419693e-05, "loss": 0.0701, "step": 35390 }, { "grad_norm": 0.613761305809021, "learning_rate": 7.679273580504132e-05, "loss": 0.0472, "step": 35400 }, { "grad_norm": 0.7603937387466431, "learning_rate": 7.67787739258724e-05, "loss": 0.0677, "step": 35410 }, { "grad_norm": 0.8744277358055115, "learning_rate": 7.676480911821705e-05, "loss": 0.0706, "step": 35420 }, { "grad_norm": 0.6087354421615601, "learning_rate": 7.675084138360238e-05, "loss": 0.0622, "step": 35430 }, { "grad_norm": 0.7712079882621765, "learning_rate": 7.673687072355592e-05, "loss": 0.0606, "step": 35440 }, { "grad_norm": 0.5261895060539246, "learning_rate": 7.67228971396055e-05, "loss": 0.0513, "step": 35450 }, { "grad_norm": 0.6617346405982971, "learning_rate": 7.670892063327922e-05, "loss": 0.0588, "step": 35460 }, { "grad_norm": 0.6694092154502869, "learning_rate": 7.669494120610552e-05, "loss": 0.0473, "step": 35470 }, { "grad_norm": 0.4242016077041626, "learning_rate": 7.66809588596132e-05, "loss": 0.0504, "step": 35480 }, { "grad_norm": 0.41302087903022766, "learning_rate": 7.666697359533132e-05, "loss": 0.0627, "step": 35490 }, { "grad_norm": 1.0639562606811523, "learning_rate": 7.665298541478932e-05, "loss": 0.068, "step": 35500 }, { "grad_norm": 0.8393462300300598, "learning_rate": 7.66389943195169e-05, "loss": 0.062, "step": 35510 }, { "grad_norm": 0.6468695998191833, "learning_rate": 7.66250003110441e-05, "loss": 0.0657, "step": 35520 }, { "grad_norm": 0.56600022315979, "learning_rate": 7.66110033909013e-05, "loss": 0.0599, "step": 35530 }, { "grad_norm": 0.6763574481010437, "learning_rate": 7.659700356061918e-05, "loss": 0.0613, "step": 35540 }, { "grad_norm": 0.577262818813324, "learning_rate": 7.658300082172875e-05, "loss": 0.0503, "step": 35550 }, { "grad_norm": 0.6103348731994629, "learning_rate": 7.65689951757613e-05, "loss": 0.053, "step": 35560 }, { "grad_norm": 0.4974537789821625, "learning_rate": 7.655498662424849e-05, "loss": 0.0582, "step": 35570 }, { "grad_norm": 0.5468505024909973, "learning_rate": 7.654097516872227e-05, "loss": 0.0506, "step": 35580 }, { "grad_norm": 0.6305457949638367, "learning_rate": 7.65269608107149e-05, "loss": 0.066, "step": 35590 }, { "grad_norm": 1.1115919351577759, "learning_rate": 7.6512943551759e-05, "loss": 0.0722, "step": 35600 }, { "grad_norm": 1.8334866762161255, "learning_rate": 7.649892339338743e-05, "loss": 0.0555, "step": 35610 }, { "grad_norm": 0.6927986741065979, "learning_rate": 7.648490033713344e-05, "loss": 0.0447, "step": 35620 }, { "grad_norm": 0.7203492522239685, "learning_rate": 7.647087438453058e-05, "loss": 0.0612, "step": 35630 }, { "grad_norm": 0.40524524450302124, "learning_rate": 7.64568455371127e-05, "loss": 0.0561, "step": 35640 }, { "grad_norm": 0.5473031401634216, "learning_rate": 7.644281379641396e-05, "loss": 0.0498, "step": 35650 }, { "grad_norm": 0.4497602880001068, "learning_rate": 7.642877916396887e-05, "loss": 0.0608, "step": 35660 }, { "grad_norm": 0.5068642497062683, "learning_rate": 7.641474164131221e-05, "loss": 0.0589, "step": 35670 }, { "grad_norm": 0.5251486897468567, "learning_rate": 7.640070122997913e-05, "loss": 0.0557, "step": 35680 }, { "grad_norm": 0.7964940667152405, "learning_rate": 7.638665793150505e-05, "loss": 0.0461, "step": 35690 }, { "grad_norm": 0.5508981347084045, "learning_rate": 7.637261174742574e-05, "loss": 0.0686, "step": 35700 }, { "grad_norm": 0.5162283182144165, "learning_rate": 7.635856267927725e-05, "loss": 0.0434, "step": 35710 }, { "grad_norm": 0.5082821846008301, "learning_rate": 7.634451072859597e-05, "loss": 0.0538, "step": 35720 }, { "grad_norm": 0.7265266180038452, "learning_rate": 7.633045589691863e-05, "loss": 0.0438, "step": 35730 }, { "grad_norm": 0.7105903029441833, "learning_rate": 7.63163981857822e-05, "loss": 0.0583, "step": 35740 }, { "grad_norm": 0.3022024631500244, "learning_rate": 7.630233759672403e-05, "loss": 0.0553, "step": 35750 }, { "grad_norm": 0.628068208694458, "learning_rate": 7.628827413128177e-05, "loss": 0.0646, "step": 35760 }, { "grad_norm": 0.5834246277809143, "learning_rate": 7.627420779099338e-05, "loss": 0.0614, "step": 35770 }, { "grad_norm": 0.8156839609146118, "learning_rate": 7.626013857739711e-05, "loss": 0.0585, "step": 35780 }, { "grad_norm": 0.830119252204895, "learning_rate": 7.624606649203158e-05, "loss": 0.0536, "step": 35790 }, { "grad_norm": 0.3847174048423767, "learning_rate": 7.623199153643569e-05, "loss": 0.0448, "step": 35800 }, { "grad_norm": 0.7880823016166687, "learning_rate": 7.621791371214863e-05, "loss": 0.0588, "step": 35810 }, { "grad_norm": 0.38556841015815735, "learning_rate": 7.620383302070995e-05, "loss": 0.0503, "step": 35820 }, { "grad_norm": 0.4909726679325104, "learning_rate": 7.61897494636595e-05, "loss": 0.0486, "step": 35830 }, { "grad_norm": 0.8964859843254089, "learning_rate": 7.617566304253739e-05, "loss": 0.0549, "step": 35840 }, { "grad_norm": 0.5513015389442444, "learning_rate": 7.616157375888416e-05, "loss": 0.0519, "step": 35850 }, { "grad_norm": 0.514920175075531, "learning_rate": 7.614748161424053e-05, "loss": 0.0441, "step": 35860 }, { "grad_norm": 0.704578697681427, "learning_rate": 7.613338661014763e-05, "loss": 0.0512, "step": 35870 }, { "grad_norm": 0.7799713611602783, "learning_rate": 7.611928874814686e-05, "loss": 0.0568, "step": 35880 }, { "grad_norm": 0.5222952961921692, "learning_rate": 7.610518802977996e-05, "loss": 0.047, "step": 35890 }, { "grad_norm": 0.5122066736221313, "learning_rate": 7.609108445658893e-05, "loss": 0.0519, "step": 35900 }, { "grad_norm": 0.7955411076545715, "learning_rate": 7.607697803011612e-05, "loss": 0.0578, "step": 35910 }, { "grad_norm": 0.6749039888381958, "learning_rate": 7.606286875190421e-05, "loss": 0.0594, "step": 35920 }, { "grad_norm": 0.4908110201358795, "learning_rate": 7.604875662349617e-05, "loss": 0.0451, "step": 35930 }, { "grad_norm": 0.37080973386764526, "learning_rate": 7.603464164643525e-05, "loss": 0.0632, "step": 35940 }, { "grad_norm": 0.8619739413261414, "learning_rate": 7.602052382226507e-05, "loss": 0.0602, "step": 35950 }, { "grad_norm": 0.530018150806427, "learning_rate": 7.600640315252954e-05, "loss": 0.0612, "step": 35960 }, { "grad_norm": 0.6536946892738342, "learning_rate": 7.599227963877284e-05, "loss": 0.0663, "step": 35970 }, { "grad_norm": 0.6493270397186279, "learning_rate": 7.597815328253954e-05, "loss": 0.0833, "step": 35980 }, { "grad_norm": 0.4887933135032654, "learning_rate": 7.596402408537443e-05, "loss": 0.0607, "step": 35990 }, { "grad_norm": 0.5395995378494263, "learning_rate": 7.59498920488227e-05, "loss": 0.067, "step": 36000 }, { "grad_norm": 0.9955996870994568, "learning_rate": 7.593575717442979e-05, "loss": 0.0565, "step": 36010 }, { "grad_norm": 0.5258169770240784, "learning_rate": 7.592161946374147e-05, "loss": 0.0643, "step": 36020 }, { "grad_norm": 0.4428810775279999, "learning_rate": 7.590747891830381e-05, "loss": 0.0528, "step": 36030 }, { "grad_norm": 0.7226118445396423, "learning_rate": 7.58933355396632e-05, "loss": 0.0561, "step": 36040 }, { "grad_norm": 0.5601562261581421, "learning_rate": 7.587918932936636e-05, "loss": 0.0551, "step": 36050 }, { "grad_norm": 0.5012474656105042, "learning_rate": 7.586504028896028e-05, "loss": 0.0596, "step": 36060 }, { "grad_norm": 0.5510597825050354, "learning_rate": 7.585088841999228e-05, "loss": 0.0522, "step": 36070 }, { "grad_norm": 0.8063371777534485, "learning_rate": 7.583673372400999e-05, "loss": 0.0658, "step": 36080 }, { "grad_norm": 0.6301459074020386, "learning_rate": 7.582257620256134e-05, "loss": 0.0506, "step": 36090 }, { "grad_norm": 0.6500322222709656, "learning_rate": 7.580841585719458e-05, "loss": 0.0638, "step": 36100 }, { "grad_norm": 0.7039864659309387, "learning_rate": 7.579425268945825e-05, "loss": 0.0633, "step": 36110 }, { "grad_norm": 0.6854655742645264, "learning_rate": 7.578008670090127e-05, "loss": 0.0504, "step": 36120 }, { "grad_norm": 0.5582349300384521, "learning_rate": 7.576591789307272e-05, "loss": 0.0515, "step": 36130 }, { "grad_norm": 0.6370102167129517, "learning_rate": 7.575174626752216e-05, "loss": 0.0458, "step": 36140 }, { "grad_norm": 0.7926601767539978, "learning_rate": 7.573757182579934e-05, "loss": 0.052, "step": 36150 }, { "grad_norm": 0.8336763978004456, "learning_rate": 7.572339456945435e-05, "loss": 0.0594, "step": 36160 }, { "grad_norm": 0.6872701048851013, "learning_rate": 7.570921450003762e-05, "loss": 0.0694, "step": 36170 }, { "grad_norm": 1.3301112651824951, "learning_rate": 7.569503161909983e-05, "loss": 0.0641, "step": 36180 }, { "grad_norm": 0.9784895181655884, "learning_rate": 7.568084592819202e-05, "loss": 0.0628, "step": 36190 }, { "grad_norm": 1.9726595878601074, "learning_rate": 7.566665742886551e-05, "loss": 0.0601, "step": 36200 }, { "grad_norm": 9.925285339355469, "learning_rate": 7.565246612267194e-05, "loss": 0.0724, "step": 36210 }, { "grad_norm": 1.0514612197875977, "learning_rate": 7.563827201116325e-05, "loss": 0.0948, "step": 36220 }, { "grad_norm": 0.636044442653656, "learning_rate": 7.562407509589164e-05, "loss": 0.058, "step": 36230 }, { "grad_norm": 0.4483020007610321, "learning_rate": 7.560987537840973e-05, "loss": 0.0606, "step": 36240 }, { "grad_norm": 0.4980296492576599, "learning_rate": 7.559567286027036e-05, "loss": 0.0506, "step": 36250 }, { "grad_norm": 0.8295484185218811, "learning_rate": 7.558146754302668e-05, "loss": 0.0647, "step": 36260 }, { "grad_norm": 0.6634542942047119, "learning_rate": 7.556725942823217e-05, "loss": 0.0467, "step": 36270 }, { "grad_norm": 0.5636085867881775, "learning_rate": 7.555304851744061e-05, "loss": 0.0652, "step": 36280 }, { "grad_norm": 0.8122960329055786, "learning_rate": 7.553883481220608e-05, "loss": 0.067, "step": 36290 }, { "grad_norm": 0.47398754954338074, "learning_rate": 7.552461831408298e-05, "loss": 0.0513, "step": 36300 }, { "grad_norm": 0.5699141621589661, "learning_rate": 7.551039902462599e-05, "loss": 0.0517, "step": 36310 }, { "grad_norm": 0.49678629636764526, "learning_rate": 7.549617694539011e-05, "loss": 0.039, "step": 36320 }, { "grad_norm": 0.693930447101593, "learning_rate": 7.548195207793066e-05, "loss": 0.055, "step": 36330 }, { "grad_norm": 0.4870619773864746, "learning_rate": 7.546772442380323e-05, "loss": 0.0539, "step": 36340 }, { "grad_norm": 0.47047704458236694, "learning_rate": 7.545349398456376e-05, "loss": 0.0724, "step": 36350 }, { "grad_norm": 0.46629393100738525, "learning_rate": 7.543926076176845e-05, "loss": 0.0626, "step": 36360 }, { "grad_norm": 0.42437949776649475, "learning_rate": 7.542502475697385e-05, "loss": 0.0534, "step": 36370 }, { "grad_norm": 0.5878036618232727, "learning_rate": 7.541078597173675e-05, "loss": 0.0659, "step": 36380 }, { "grad_norm": 0.5803481936454773, "learning_rate": 7.539654440761431e-05, "loss": 0.0655, "step": 36390 }, { "grad_norm": 0.45873382687568665, "learning_rate": 7.538230006616395e-05, "loss": 0.0514, "step": 36400 }, { "grad_norm": 0.4717300236225128, "learning_rate": 7.536805294894342e-05, "loss": 0.0708, "step": 36410 }, { "grad_norm": 0.5367945432662964, "learning_rate": 7.535380305751076e-05, "loss": 0.048, "step": 36420 }, { "grad_norm": 0.39414724707603455, "learning_rate": 7.533955039342431e-05, "loss": 0.0615, "step": 36430 }, { "grad_norm": 1.1053262948989868, "learning_rate": 7.532529495824274e-05, "loss": 0.0623, "step": 36440 }, { "grad_norm": 0.5325299501419067, "learning_rate": 7.531103675352497e-05, "loss": 0.047, "step": 36450 }, { "grad_norm": 0.6037616729736328, "learning_rate": 7.529677578083029e-05, "loss": 0.0604, "step": 36460 }, { "grad_norm": 0.7475279569625854, "learning_rate": 7.528251204171823e-05, "loss": 0.0623, "step": 36470 }, { "grad_norm": 0.6130141615867615, "learning_rate": 7.526824553774867e-05, "loss": 0.0549, "step": 36480 }, { "grad_norm": 0.5072158575057983, "learning_rate": 7.525397627048178e-05, "loss": 0.0559, "step": 36490 }, { "grad_norm": 0.7003971934318542, "learning_rate": 7.523970424147802e-05, "loss": 0.0516, "step": 36500 }, { "grad_norm": 0.41908204555511475, "learning_rate": 7.522542945229813e-05, "loss": 0.0424, "step": 36510 }, { "grad_norm": 0.5722483992576599, "learning_rate": 7.521115190450321e-05, "loss": 0.0504, "step": 36520 }, { "grad_norm": 0.6093616485595703, "learning_rate": 7.519687159965462e-05, "loss": 0.0543, "step": 36530 }, { "grad_norm": 0.6947941780090332, "learning_rate": 7.518258853931403e-05, "loss": 0.0542, "step": 36540 }, { "grad_norm": 0.31634238362312317, "learning_rate": 7.516830272504342e-05, "loss": 0.0435, "step": 36550 }, { "grad_norm": 0.5841870903968811, "learning_rate": 7.515401415840509e-05, "loss": 0.0719, "step": 36560 }, { "grad_norm": 0.863702118396759, "learning_rate": 7.513972284096155e-05, "loss": 0.0551, "step": 36570 }, { "grad_norm": 0.409074068069458, "learning_rate": 7.512542877427574e-05, "loss": 0.0612, "step": 36580 }, { "grad_norm": 0.37252548336982727, "learning_rate": 7.51111319599108e-05, "loss": 0.0557, "step": 36590 }, { "grad_norm": 0.5513882040977478, "learning_rate": 7.509683239943024e-05, "loss": 0.0622, "step": 36600 }, { "grad_norm": 0.628462016582489, "learning_rate": 7.50825300943978e-05, "loss": 0.0656, "step": 36610 }, { "grad_norm": 0.7996413111686707, "learning_rate": 7.50682250463776e-05, "loss": 0.0546, "step": 36620 }, { "grad_norm": 0.6150386333465576, "learning_rate": 7.5053917256934e-05, "loss": 0.052, "step": 36630 }, { "grad_norm": 0.8109657168388367, "learning_rate": 7.503960672763166e-05, "loss": 0.0638, "step": 36640 }, { "grad_norm": 0.5160093903541565, "learning_rate": 7.502529346003559e-05, "loss": 0.0628, "step": 36650 }, { "grad_norm": 0.4883386790752411, "learning_rate": 7.501097745571106e-05, "loss": 0.0589, "step": 36660 }, { "grad_norm": 0.40163370966911316, "learning_rate": 7.499665871622361e-05, "loss": 0.0655, "step": 36670 }, { "grad_norm": 0.9660546779632568, "learning_rate": 7.498233724313917e-05, "loss": 0.0664, "step": 36680 }, { "grad_norm": 0.45934346318244934, "learning_rate": 7.496801303802387e-05, "loss": 0.066, "step": 36690 }, { "grad_norm": 0.5664721131324768, "learning_rate": 7.495368610244423e-05, "loss": 0.0613, "step": 36700 }, { "grad_norm": 0.8729487061500549, "learning_rate": 7.493935643796697e-05, "loss": 0.0669, "step": 36710 }, { "grad_norm": 0.5596315264701843, "learning_rate": 7.49250240461592e-05, "loss": 0.0602, "step": 36720 }, { "grad_norm": 0.57123863697052, "learning_rate": 7.491068892858826e-05, "loss": 0.0639, "step": 36730 }, { "grad_norm": 0.6304845213890076, "learning_rate": 7.489635108682184e-05, "loss": 0.059, "step": 36740 }, { "grad_norm": 0.482115775346756, "learning_rate": 7.48820105224279e-05, "loss": 0.0526, "step": 36750 }, { "grad_norm": 0.7923217415809631, "learning_rate": 7.48676672369747e-05, "loss": 0.0662, "step": 36760 }, { "grad_norm": 0.6549747586250305, "learning_rate": 7.485332123203079e-05, "loss": 0.0654, "step": 36770 }, { "grad_norm": 0.47897112369537354, "learning_rate": 7.483897250916504e-05, "loss": 0.0669, "step": 36780 }, { "grad_norm": 1.1281527280807495, "learning_rate": 7.48246210699466e-05, "loss": 0.0629, "step": 36790 }, { "grad_norm": 0.8895730376243591, "learning_rate": 7.481026691594492e-05, "loss": 0.0601, "step": 36800 }, { "grad_norm": 0.5123453736305237, "learning_rate": 7.479591004872975e-05, "loss": 0.0473, "step": 36810 }, { "grad_norm": 0.44674399495124817, "learning_rate": 7.478155046987114e-05, "loss": 0.0602, "step": 36820 }, { "grad_norm": 0.4455936849117279, "learning_rate": 7.476718818093944e-05, "loss": 0.0605, "step": 36830 }, { "grad_norm": 0.5963395237922668, "learning_rate": 7.475282318350527e-05, "loss": 0.0586, "step": 36840 }, { "grad_norm": 0.48645997047424316, "learning_rate": 7.473845547913957e-05, "loss": 0.0429, "step": 36850 }, { "grad_norm": 0.6416488289833069, "learning_rate": 7.472408506941357e-05, "loss": 0.0789, "step": 36860 }, { "grad_norm": 0.46354031562805176, "learning_rate": 7.470971195589881e-05, "loss": 0.0551, "step": 36870 }, { "grad_norm": 0.7818049788475037, "learning_rate": 7.46953361401671e-05, "loss": 0.0722, "step": 36880 }, { "grad_norm": 0.740871787071228, "learning_rate": 7.468095762379055e-05, "loss": 0.0502, "step": 36890 }, { "grad_norm": 0.4346321225166321, "learning_rate": 7.466657640834158e-05, "loss": 0.0718, "step": 36900 }, { "grad_norm": 0.6962811350822449, "learning_rate": 7.46521924953929e-05, "loss": 0.054, "step": 36910 }, { "grad_norm": 0.5284515619277954, "learning_rate": 7.463780588651752e-05, "loss": 0.0521, "step": 36920 }, { "grad_norm": 1.2346336841583252, "learning_rate": 7.462341658328873e-05, "loss": 0.0612, "step": 36930 }, { "grad_norm": 0.46303150057792664, "learning_rate": 7.460902458728012e-05, "loss": 0.0554, "step": 36940 }, { "grad_norm": 0.8682425618171692, "learning_rate": 7.459462990006558e-05, "loss": 0.0501, "step": 36950 }, { "grad_norm": 0.7628907561302185, "learning_rate": 7.458023252321929e-05, "loss": 0.0546, "step": 36960 }, { "grad_norm": 0.6468732953071594, "learning_rate": 7.456583245831573e-05, "loss": 0.0494, "step": 36970 }, { "grad_norm": 0.6837670207023621, "learning_rate": 7.455142970692968e-05, "loss": 0.0811, "step": 36980 }, { "grad_norm": 0.874003529548645, "learning_rate": 7.453702427063615e-05, "loss": 0.0503, "step": 36990 }, { "grad_norm": 0.5595324039459229, "learning_rate": 7.452261615101057e-05, "loss": 0.0598, "step": 37000 }, { "grad_norm": 0.8048023581504822, "learning_rate": 7.450820534962856e-05, "loss": 0.0689, "step": 37010 }, { "grad_norm": 0.41262009739875793, "learning_rate": 7.449379186806604e-05, "loss": 0.0518, "step": 37020 }, { "grad_norm": 0.9311687350273132, "learning_rate": 7.44793757078993e-05, "loss": 0.062, "step": 37030 }, { "grad_norm": 0.5422584414482117, "learning_rate": 7.44649568707048e-05, "loss": 0.055, "step": 37040 }, { "grad_norm": 0.7782124876976013, "learning_rate": 7.445053535805942e-05, "loss": 0.0599, "step": 37050 }, { "grad_norm": 0.6052790880203247, "learning_rate": 7.443611117154024e-05, "loss": 0.0566, "step": 37060 }, { "grad_norm": 1.1608340740203857, "learning_rate": 7.442168431272469e-05, "loss": 0.054, "step": 37070 }, { "grad_norm": 0.6376600861549377, "learning_rate": 7.440725478319044e-05, "loss": 0.0601, "step": 37080 }, { "grad_norm": 0.7318310737609863, "learning_rate": 7.439282258451553e-05, "loss": 0.0607, "step": 37090 }, { "grad_norm": 0.5390441417694092, "learning_rate": 7.43783877182782e-05, "loss": 0.0554, "step": 37100 }, { "grad_norm": 0.531562864780426, "learning_rate": 7.436395018605707e-05, "loss": 0.049, "step": 37110 }, { "grad_norm": 0.4901030361652374, "learning_rate": 7.434950998943094e-05, "loss": 0.0485, "step": 37120 }, { "grad_norm": 0.597959041595459, "learning_rate": 7.433506712997902e-05, "loss": 0.0588, "step": 37130 }, { "grad_norm": 0.38004353642463684, "learning_rate": 7.432062160928077e-05, "loss": 0.0534, "step": 37140 }, { "grad_norm": 0.33529821038246155, "learning_rate": 7.430617342891588e-05, "loss": 0.0492, "step": 37150 }, { "grad_norm": 0.6086088418960571, "learning_rate": 7.429172259046442e-05, "loss": 0.0496, "step": 37160 }, { "grad_norm": 0.5643570423126221, "learning_rate": 7.42772690955067e-05, "loss": 0.0506, "step": 37170 }, { "grad_norm": 0.4972560703754425, "learning_rate": 7.426281294562332e-05, "loss": 0.0698, "step": 37180 }, { "grad_norm": 0.4099370837211609, "learning_rate": 7.424835414239519e-05, "loss": 0.0659, "step": 37190 }, { "grad_norm": 0.502923846244812, "learning_rate": 7.423389268740352e-05, "loss": 0.0538, "step": 37200 }, { "grad_norm": 0.8849736452102661, "learning_rate": 7.42194285822298e-05, "loss": 0.0489, "step": 37210 }, { "grad_norm": 0.34265169501304626, "learning_rate": 7.420496182845576e-05, "loss": 0.0594, "step": 37220 }, { "grad_norm": 0.45615440607070923, "learning_rate": 7.419049242766348e-05, "loss": 0.0626, "step": 37230 }, { "grad_norm": 0.7232270240783691, "learning_rate": 7.417602038143533e-05, "loss": 0.0624, "step": 37240 }, { "grad_norm": 0.7113361358642578, "learning_rate": 7.416154569135393e-05, "loss": 0.0676, "step": 37250 }, { "grad_norm": 0.6832366585731506, "learning_rate": 7.414706835900222e-05, "loss": 0.0473, "step": 37260 }, { "grad_norm": 0.7530160546302795, "learning_rate": 7.413258838596345e-05, "loss": 0.0549, "step": 37270 }, { "grad_norm": 0.7415152788162231, "learning_rate": 7.411810577382106e-05, "loss": 0.0588, "step": 37280 }, { "grad_norm": 0.54848313331604, "learning_rate": 7.410362052415889e-05, "loss": 0.0592, "step": 37290 }, { "grad_norm": 0.8616908192634583, "learning_rate": 7.408913263856102e-05, "loss": 0.0653, "step": 37300 }, { "grad_norm": 0.6951572895050049, "learning_rate": 7.407464211861183e-05, "loss": 0.0704, "step": 37310 }, { "grad_norm": 0.5551887154579163, "learning_rate": 7.406014896589597e-05, "loss": 0.0564, "step": 37320 }, { "grad_norm": 0.6127187013626099, "learning_rate": 7.404565318199838e-05, "loss": 0.0519, "step": 37330 }, { "grad_norm": 0.455620139837265, "learning_rate": 7.403115476850432e-05, "loss": 0.0696, "step": 37340 }, { "grad_norm": 0.4645118713378906, "learning_rate": 7.401665372699932e-05, "loss": 0.0572, "step": 37350 }, { "grad_norm": 1.3451855182647705, "learning_rate": 7.400215005906916e-05, "loss": 0.0674, "step": 37360 }, { "grad_norm": 0.7505593299865723, "learning_rate": 7.398764376629996e-05, "loss": 0.0636, "step": 37370 }, { "grad_norm": 0.4643048942089081, "learning_rate": 7.39731348502781e-05, "loss": 0.0697, "step": 37380 }, { "grad_norm": 0.6181807518005371, "learning_rate": 7.395862331259028e-05, "loss": 0.0595, "step": 37390 }, { "grad_norm": 0.9082328081130981, "learning_rate": 7.394410915482341e-05, "loss": 0.0647, "step": 37400 }, { "grad_norm": 0.4360048174858093, "learning_rate": 7.392959237856476e-05, "loss": 0.0582, "step": 37410 }, { "grad_norm": 0.7483494877815247, "learning_rate": 7.391507298540187e-05, "loss": 0.054, "step": 37420 }, { "grad_norm": 0.4291745126247406, "learning_rate": 7.390055097692257e-05, "loss": 0.0453, "step": 37430 }, { "grad_norm": 0.620229184627533, "learning_rate": 7.388602635471494e-05, "loss": 0.0624, "step": 37440 }, { "grad_norm": 0.6045175790786743, "learning_rate": 7.387149912036738e-05, "loss": 0.0657, "step": 37450 }, { "grad_norm": 0.5006904602050781, "learning_rate": 7.385696927546854e-05, "loss": 0.0533, "step": 37460 }, { "grad_norm": 0.5788230299949646, "learning_rate": 7.384243682160743e-05, "loss": 0.079, "step": 37470 }, { "grad_norm": 0.6955205202102661, "learning_rate": 7.382790176037329e-05, "loss": 0.0559, "step": 37480 }, { "grad_norm": 0.6109376549720764, "learning_rate": 7.381336409335562e-05, "loss": 0.065, "step": 37490 }, { "grad_norm": 0.9434003829956055, "learning_rate": 7.379882382214426e-05, "loss": 0.061, "step": 37500 }, { "grad_norm": 0.3400997519493103, "learning_rate": 7.378428094832931e-05, "loss": 0.0592, "step": 37510 }, { "grad_norm": 0.7589488625526428, "learning_rate": 7.376973547350114e-05, "loss": 0.061, "step": 37520 }, { "grad_norm": 0.540004312992096, "learning_rate": 7.375518739925046e-05, "loss": 0.0608, "step": 37530 }, { "grad_norm": 0.5040999054908752, "learning_rate": 7.374063672716818e-05, "loss": 0.0481, "step": 37540 }, { "grad_norm": 0.6583433151245117, "learning_rate": 7.372608345884558e-05, "loss": 0.0635, "step": 37550 }, { "grad_norm": 0.8430037498474121, "learning_rate": 7.371152759587415e-05, "loss": 0.0649, "step": 37560 }, { "grad_norm": 0.5804653167724609, "learning_rate": 7.369696913984572e-05, "loss": 0.0466, "step": 37570 }, { "grad_norm": 0.5427790880203247, "learning_rate": 7.368240809235237e-05, "loss": 0.0565, "step": 37580 }, { "grad_norm": 0.8643754124641418, "learning_rate": 7.366784445498649e-05, "loss": 0.0694, "step": 37590 }, { "grad_norm": 0.4011841416358948, "learning_rate": 7.36532782293407e-05, "loss": 0.0502, "step": 37600 }, { "grad_norm": 0.4416337013244629, "learning_rate": 7.363870941700797e-05, "loss": 0.0595, "step": 37610 }, { "grad_norm": 0.5016890168190002, "learning_rate": 7.362413801958152e-05, "loss": 0.0497, "step": 37620 }, { "grad_norm": 0.49767258763313293, "learning_rate": 7.360956403865486e-05, "loss": 0.058, "step": 37630 }, { "grad_norm": 0.4915306568145752, "learning_rate": 7.359498747582177e-05, "loss": 0.0614, "step": 37640 }, { "grad_norm": 0.49019983410835266, "learning_rate": 7.358040833267634e-05, "loss": 0.0529, "step": 37650 }, { "grad_norm": 0.4944135546684265, "learning_rate": 7.356582661081288e-05, "loss": 0.0582, "step": 37660 }, { "grad_norm": 0.5751136541366577, "learning_rate": 7.355124231182607e-05, "loss": 0.0417, "step": 37670 }, { "grad_norm": 0.6617022752761841, "learning_rate": 7.353665543731083e-05, "loss": 0.0483, "step": 37680 }, { "grad_norm": 0.42877236008644104, "learning_rate": 7.352206598886229e-05, "loss": 0.0526, "step": 37690 }, { "grad_norm": 0.4471248388290405, "learning_rate": 7.350747396807601e-05, "loss": 0.0538, "step": 37700 }, { "grad_norm": 0.6449949741363525, "learning_rate": 7.349287937654772e-05, "loss": 0.0579, "step": 37710 }, { "grad_norm": 0.393903523683548, "learning_rate": 7.347828221587345e-05, "loss": 0.0566, "step": 37720 }, { "grad_norm": 0.4620904326438904, "learning_rate": 7.346368248764955e-05, "loss": 0.0554, "step": 37730 }, { "grad_norm": 0.48202869296073914, "learning_rate": 7.34490801934726e-05, "loss": 0.0556, "step": 37740 }, { "grad_norm": 0.5100775957107544, "learning_rate": 7.343447533493947e-05, "loss": 0.0673, "step": 37750 }, { "grad_norm": 0.4179231524467468, "learning_rate": 7.341986791364739e-05, "loss": 0.0779, "step": 37760 }, { "grad_norm": 0.4321447014808655, "learning_rate": 7.340525793119374e-05, "loss": 0.0476, "step": 37770 }, { "grad_norm": 0.622501015663147, "learning_rate": 7.33906453891763e-05, "loss": 0.0577, "step": 37780 }, { "grad_norm": 0.9656192660331726, "learning_rate": 7.337603028919303e-05, "loss": 0.054, "step": 37790 }, { "grad_norm": 1.1650463342666626, "learning_rate": 7.336141263284225e-05, "loss": 0.0587, "step": 37800 }, { "grad_norm": 0.5183529853820801, "learning_rate": 7.334679242172249e-05, "loss": 0.0593, "step": 37810 }, { "grad_norm": 0.5595707297325134, "learning_rate": 7.33321696574326e-05, "loss": 0.0569, "step": 37820 }, { "grad_norm": 0.423820823431015, "learning_rate": 7.331754434157174e-05, "loss": 0.0543, "step": 37830 }, { "grad_norm": 0.49628594517707825, "learning_rate": 7.330291647573928e-05, "loss": 0.065, "step": 37840 }, { "grad_norm": 0.7145684361457825, "learning_rate": 7.32882860615349e-05, "loss": 0.0496, "step": 37850 }, { "grad_norm": 0.5145943760871887, "learning_rate": 7.327365310055858e-05, "loss": 0.0529, "step": 37860 }, { "grad_norm": 0.4773262143135071, "learning_rate": 7.325901759441055e-05, "loss": 0.0609, "step": 37870 }, { "grad_norm": 0.4886135756969452, "learning_rate": 7.324437954469132e-05, "loss": 0.049, "step": 37880 }, { "grad_norm": 0.978922426700592, "learning_rate": 7.32297389530017e-05, "loss": 0.0565, "step": 37890 }, { "grad_norm": 0.38826707005500793, "learning_rate": 7.321509582094274e-05, "loss": 0.0556, "step": 37900 }, { "grad_norm": 0.4083472192287445, "learning_rate": 7.320045015011582e-05, "loss": 0.0642, "step": 37910 }, { "grad_norm": 0.35005372762680054, "learning_rate": 7.318580194212253e-05, "loss": 0.0583, "step": 37920 }, { "grad_norm": 0.44940808415412903, "learning_rate": 7.317115119856483e-05, "loss": 0.0592, "step": 37930 }, { "grad_norm": 1.0409064292907715, "learning_rate": 7.315649792104485e-05, "loss": 0.0546, "step": 37940 }, { "grad_norm": 1.2181930541992188, "learning_rate": 7.314184211116507e-05, "loss": 0.0623, "step": 37950 }, { "grad_norm": 0.6357723474502563, "learning_rate": 7.312718377052823e-05, "loss": 0.053, "step": 37960 }, { "grad_norm": 0.6575264930725098, "learning_rate": 7.311252290073737e-05, "loss": 0.0542, "step": 37970 }, { "grad_norm": 0.4791549742221832, "learning_rate": 7.309785950339572e-05, "loss": 0.0566, "step": 37980 }, { "grad_norm": 0.5426809191703796, "learning_rate": 7.308319358010692e-05, "loss": 0.0547, "step": 37990 }, { "grad_norm": 0.4691433906555176, "learning_rate": 7.306852513247474e-05, "loss": 0.0746, "step": 38000 }, { "grad_norm": 0.7210711240768433, "learning_rate": 7.305385416210336e-05, "loss": 0.065, "step": 38010 }, { "grad_norm": 0.4124312698841095, "learning_rate": 7.303918067059711e-05, "loss": 0.0423, "step": 38020 }, { "grad_norm": 0.33637943863868713, "learning_rate": 7.302450465956075e-05, "loss": 0.0502, "step": 38030 }, { "grad_norm": 0.5400804281234741, "learning_rate": 7.300982613059914e-05, "loss": 0.0525, "step": 38040 }, { "grad_norm": 0.6573511362075806, "learning_rate": 7.299514508531757e-05, "loss": 0.0633, "step": 38050 }, { "grad_norm": 0.498218834400177, "learning_rate": 7.298046152532148e-05, "loss": 0.0536, "step": 38060 }, { "grad_norm": 1.057845115661621, "learning_rate": 7.296577545221668e-05, "loss": 0.0567, "step": 38070 }, { "grad_norm": 0.5756397843360901, "learning_rate": 7.29510868676092e-05, "loss": 0.0584, "step": 38080 }, { "grad_norm": 0.5985024571418762, "learning_rate": 7.293639577310538e-05, "loss": 0.0557, "step": 38090 }, { "grad_norm": 0.5231657028198242, "learning_rate": 7.292170217031179e-05, "loss": 0.0658, "step": 38100 }, { "grad_norm": 0.6127088665962219, "learning_rate": 7.290700606083532e-05, "loss": 0.0628, "step": 38110 }, { "grad_norm": 0.4709593951702118, "learning_rate": 7.28923074462831e-05, "loss": 0.0549, "step": 38120 }, { "grad_norm": 0.7278882265090942, "learning_rate": 7.287760632826255e-05, "loss": 0.0542, "step": 38130 }, { "grad_norm": 0.55739825963974, "learning_rate": 7.286290270838139e-05, "loss": 0.0655, "step": 38140 }, { "grad_norm": 0.7331674098968506, "learning_rate": 7.284819658824756e-05, "loss": 0.0448, "step": 38150 }, { "grad_norm": 0.5405963659286499, "learning_rate": 7.283348796946929e-05, "loss": 0.0534, "step": 38160 }, { "grad_norm": 0.48246270418167114, "learning_rate": 7.28187768536551e-05, "loss": 0.0534, "step": 38170 }, { "grad_norm": 0.6272900104522705, "learning_rate": 7.280406324241379e-05, "loss": 0.061, "step": 38180 }, { "grad_norm": 0.3905348777770996, "learning_rate": 7.278934713735441e-05, "loss": 0.0619, "step": 38190 }, { "grad_norm": 0.48071667551994324, "learning_rate": 7.277462854008629e-05, "loss": 0.0594, "step": 38200 }, { "grad_norm": 0.7946183085441589, "learning_rate": 7.275990745221904e-05, "loss": 0.0537, "step": 38210 }, { "grad_norm": 0.4108094573020935, "learning_rate": 7.274518387536252e-05, "loss": 0.0564, "step": 38220 }, { "grad_norm": 0.7394025325775146, "learning_rate": 7.273045781112689e-05, "loss": 0.06, "step": 38230 }, { "grad_norm": 0.5846706032752991, "learning_rate": 7.271572926112257e-05, "loss": 0.0568, "step": 38240 }, { "grad_norm": 0.6044721603393555, "learning_rate": 7.270099822696024e-05, "loss": 0.0548, "step": 38250 }, { "grad_norm": 0.4380296766757965, "learning_rate": 7.26862647102509e-05, "loss": 0.0538, "step": 38260 }, { "grad_norm": 0.41703712940216064, "learning_rate": 7.267152871260573e-05, "loss": 0.0523, "step": 38270 }, { "grad_norm": 1.1945438385009766, "learning_rate": 7.265679023563629e-05, "loss": 0.0729, "step": 38280 }, { "grad_norm": 0.647471010684967, "learning_rate": 7.264204928095433e-05, "loss": 0.066, "step": 38290 }, { "grad_norm": 0.5451099872589111, "learning_rate": 7.262730585017188e-05, "loss": 0.0598, "step": 38300 }, { "grad_norm": 0.40477344393730164, "learning_rate": 7.261255994490131e-05, "loss": 0.0589, "step": 38310 }, { "grad_norm": 0.4945862293243408, "learning_rate": 7.259781156675517e-05, "loss": 0.0597, "step": 38320 }, { "grad_norm": 1.5107413530349731, "learning_rate": 7.258306071734633e-05, "loss": 0.0615, "step": 38330 }, { "grad_norm": 0.4882907271385193, "learning_rate": 7.256830739828792e-05, "loss": 0.0613, "step": 38340 }, { "grad_norm": 0.6386990547180176, "learning_rate": 7.255355161119336e-05, "loss": 0.0465, "step": 38350 }, { "grad_norm": 0.6370879411697388, "learning_rate": 7.253879335767628e-05, "loss": 0.0482, "step": 38360 }, { "grad_norm": 0.619713306427002, "learning_rate": 7.252403263935067e-05, "loss": 0.0724, "step": 38370 }, { "grad_norm": 0.4655056893825531, "learning_rate": 7.25092694578307e-05, "loss": 0.067, "step": 38380 }, { "grad_norm": 0.6911669373512268, "learning_rate": 7.249450381473087e-05, "loss": 0.0683, "step": 38390 }, { "grad_norm": 0.8566727638244629, "learning_rate": 7.247973571166593e-05, "loss": 0.0696, "step": 38400 }, { "grad_norm": 0.611555814743042, "learning_rate": 7.246496515025089e-05, "loss": 0.059, "step": 38410 }, { "grad_norm": 0.6764767169952393, "learning_rate": 7.245019213210101e-05, "loss": 0.0589, "step": 38420 }, { "grad_norm": 0.30598756670951843, "learning_rate": 7.24354166588319e-05, "loss": 0.0611, "step": 38430 }, { "grad_norm": 0.530139148235321, "learning_rate": 7.242063873205935e-05, "loss": 0.0701, "step": 38440 }, { "grad_norm": 0.6520147323608398, "learning_rate": 7.240585835339946e-05, "loss": 0.061, "step": 38450 }, { "grad_norm": 0.8126518130302429, "learning_rate": 7.239107552446857e-05, "loss": 0.065, "step": 38460 }, { "grad_norm": 0.529926598072052, "learning_rate": 7.237629024688334e-05, "loss": 0.0585, "step": 38470 }, { "grad_norm": 0.608873188495636, "learning_rate": 7.236150252226064e-05, "loss": 0.0563, "step": 38480 }, { "grad_norm": 0.7331547737121582, "learning_rate": 7.234671235221765e-05, "loss": 0.0581, "step": 38490 }, { "grad_norm": 0.576574981212616, "learning_rate": 7.233191973837179e-05, "loss": 0.0616, "step": 38500 }, { "grad_norm": 0.7757172584533691, "learning_rate": 7.231712468234078e-05, "loss": 0.0592, "step": 38510 }, { "grad_norm": 0.5626431703567505, "learning_rate": 7.230232718574254e-05, "loss": 0.0643, "step": 38520 }, { "grad_norm": 0.30727261304855347, "learning_rate": 7.228752725019535e-05, "loss": 0.0588, "step": 38530 }, { "grad_norm": 0.35696840286254883, "learning_rate": 7.227272487731769e-05, "loss": 0.0518, "step": 38540 }, { "grad_norm": 0.47845935821533203, "learning_rate": 7.225792006872831e-05, "loss": 0.0492, "step": 38550 }, { "grad_norm": 0.8118934631347656, "learning_rate": 7.224311282604628e-05, "loss": 0.0518, "step": 38560 }, { "grad_norm": 0.7996848225593567, "learning_rate": 7.222830315089085e-05, "loss": 0.0569, "step": 38570 }, { "grad_norm": 0.688176155090332, "learning_rate": 7.22134910448816e-05, "loss": 0.0593, "step": 38580 }, { "grad_norm": 1.4984643459320068, "learning_rate": 7.219867650963839e-05, "loss": 0.0576, "step": 38590 }, { "grad_norm": 0.46143481135368347, "learning_rate": 7.21838595467813e-05, "loss": 0.048, "step": 38600 }, { "grad_norm": 0.3644261360168457, "learning_rate": 7.216904015793067e-05, "loss": 0.0559, "step": 38610 }, { "grad_norm": 0.4428068697452545, "learning_rate": 7.215421834470713e-05, "loss": 0.0515, "step": 38620 }, { "grad_norm": 0.48649197816848755, "learning_rate": 7.213939410873158e-05, "loss": 0.0561, "step": 38630 }, { "grad_norm": 1.480393409729004, "learning_rate": 7.212456745162519e-05, "loss": 0.0552, "step": 38640 }, { "grad_norm": 0.6477633118629456, "learning_rate": 7.210973837500937e-05, "loss": 0.0606, "step": 38650 }, { "grad_norm": 0.562311053276062, "learning_rate": 7.209490688050578e-05, "loss": 0.0517, "step": 38660 }, { "grad_norm": 0.5948933959007263, "learning_rate": 7.208007296973641e-05, "loss": 0.0555, "step": 38670 }, { "grad_norm": 1.3059773445129395, "learning_rate": 7.206523664432345e-05, "loss": 0.0758, "step": 38680 }, { "grad_norm": 0.9889471530914307, "learning_rate": 7.205039790588939e-05, "loss": 0.0399, "step": 38690 }, { "grad_norm": 0.4799160063266754, "learning_rate": 7.203555675605697e-05, "loss": 0.0517, "step": 38700 }, { "grad_norm": 0.6126844882965088, "learning_rate": 7.202071319644917e-05, "loss": 0.067, "step": 38710 }, { "grad_norm": 0.5216339230537415, "learning_rate": 7.200586722868932e-05, "loss": 0.0487, "step": 38720 }, { "grad_norm": 0.4968605637550354, "learning_rate": 7.199101885440088e-05, "loss": 0.0634, "step": 38730 }, { "grad_norm": 0.5808975696563721, "learning_rate": 7.197616807520771e-05, "loss": 0.0593, "step": 38740 }, { "grad_norm": 0.6403872966766357, "learning_rate": 7.196131489273381e-05, "loss": 0.0504, "step": 38750 }, { "grad_norm": 0.4665919542312622, "learning_rate": 7.194645930860355e-05, "loss": 0.0494, "step": 38760 }, { "grad_norm": 0.7499824166297913, "learning_rate": 7.19316013244415e-05, "loss": 0.0503, "step": 38770 }, { "grad_norm": 0.4961950480937958, "learning_rate": 7.191674094187248e-05, "loss": 0.0617, "step": 38780 }, { "grad_norm": 0.549068033695221, "learning_rate": 7.190187816252165e-05, "loss": 0.043, "step": 38790 }, { "grad_norm": 0.6481949090957642, "learning_rate": 7.188701298801435e-05, "loss": 0.0607, "step": 38800 }, { "grad_norm": 0.5697227120399475, "learning_rate": 7.18721454199762e-05, "loss": 0.0501, "step": 38810 }, { "grad_norm": 0.8731909394264221, "learning_rate": 7.185727546003312e-05, "loss": 0.0571, "step": 38820 }, { "grad_norm": 0.441727876663208, "learning_rate": 7.184240310981126e-05, "loss": 0.0463, "step": 38830 }, { "grad_norm": 0.4767805337905884, "learning_rate": 7.182752837093704e-05, "loss": 0.0613, "step": 38840 }, { "grad_norm": 0.6053096652030945, "learning_rate": 7.181265124503711e-05, "loss": 0.0643, "step": 38850 }, { "grad_norm": 0.508648693561554, "learning_rate": 7.179777173373847e-05, "loss": 0.048, "step": 38860 }, { "grad_norm": 0.5445845723152161, "learning_rate": 7.178288983866826e-05, "loss": 0.0444, "step": 38870 }, { "grad_norm": 0.450284481048584, "learning_rate": 7.176800556145397e-05, "loss": 0.0581, "step": 38880 }, { "grad_norm": 0.5127061009407043, "learning_rate": 7.175311890372334e-05, "loss": 0.054, "step": 38890 }, { "grad_norm": 0.6797904968261719, "learning_rate": 7.17382298671043e-05, "loss": 0.0539, "step": 38900 }, { "grad_norm": 0.6659561991691589, "learning_rate": 7.172333845322515e-05, "loss": 0.0557, "step": 38910 }, { "grad_norm": 0.33932754397392273, "learning_rate": 7.170844466371436e-05, "loss": 0.0521, "step": 38920 }, { "grad_norm": 0.5603283643722534, "learning_rate": 7.16935485002007e-05, "loss": 0.052, "step": 38930 }, { "grad_norm": 0.6661241054534912, "learning_rate": 7.167864996431319e-05, "loss": 0.0682, "step": 38940 }, { "grad_norm": 0.46087169647216797, "learning_rate": 7.166374905768111e-05, "loss": 0.0652, "step": 38950 }, { "grad_norm": 0.5991817712783813, "learning_rate": 7.164884578193401e-05, "loss": 0.0481, "step": 38960 }, { "grad_norm": 0.5547136068344116, "learning_rate": 7.163394013870168e-05, "loss": 0.0533, "step": 38970 }, { "grad_norm": 0.8634788393974304, "learning_rate": 7.161903212961419e-05, "loss": 0.0581, "step": 38980 }, { "grad_norm": 0.6129604578018188, "learning_rate": 7.160412175630183e-05, "loss": 0.0583, "step": 38990 }, { "grad_norm": 0.3154158592224121, "learning_rate": 7.158920902039521e-05, "loss": 0.0496, "step": 39000 }, { "grad_norm": 0.6200711727142334, "learning_rate": 7.157429392352514e-05, "loss": 0.0514, "step": 39010 }, { "grad_norm": 0.6844416260719299, "learning_rate": 7.155937646732275e-05, "loss": 0.0502, "step": 39020 }, { "grad_norm": 0.5423294305801392, "learning_rate": 7.154445665341933e-05, "loss": 0.0496, "step": 39030 }, { "grad_norm": 0.5836457014083862, "learning_rate": 7.152953448344654e-05, "loss": 0.0471, "step": 39040 }, { "grad_norm": 1.4269473552703857, "learning_rate": 7.151460995903624e-05, "loss": 0.0517, "step": 39050 }, { "grad_norm": 1.3975614309310913, "learning_rate": 7.149968308182052e-05, "loss": 0.0528, "step": 39060 }, { "grad_norm": 0.6530159115791321, "learning_rate": 7.14847538534318e-05, "loss": 0.0591, "step": 39070 }, { "grad_norm": 0.5523810386657715, "learning_rate": 7.14698222755027e-05, "loss": 0.0546, "step": 39080 }, { "grad_norm": 0.48837506771087646, "learning_rate": 7.14548883496661e-05, "loss": 0.0504, "step": 39090 }, { "grad_norm": 0.9735642075538635, "learning_rate": 7.143995207755517e-05, "loss": 0.0584, "step": 39100 }, { "grad_norm": 0.686022937297821, "learning_rate": 7.142501346080333e-05, "loss": 0.0585, "step": 39110 }, { "grad_norm": 0.8939307332038879, "learning_rate": 7.141007250104421e-05, "loss": 0.0575, "step": 39120 }, { "grad_norm": 0.6534618735313416, "learning_rate": 7.139512919991176e-05, "loss": 0.0547, "step": 39130 }, { "grad_norm": 1.2731199264526367, "learning_rate": 7.138018355904014e-05, "loss": 0.0588, "step": 39140 }, { "grad_norm": 0.7656754851341248, "learning_rate": 7.13652355800638e-05, "loss": 0.0766, "step": 39150 }, { "grad_norm": 0.616855263710022, "learning_rate": 7.135028526461743e-05, "loss": 0.0569, "step": 39160 }, { "grad_norm": 0.9388359785079956, "learning_rate": 7.133533261433594e-05, "loss": 0.0623, "step": 39170 }, { "grad_norm": 0.6182593703269958, "learning_rate": 7.132037763085457e-05, "loss": 0.0609, "step": 39180 }, { "grad_norm": 0.5167092680931091, "learning_rate": 7.130542031580875e-05, "loss": 0.0577, "step": 39190 }, { "grad_norm": 0.6342797875404358, "learning_rate": 7.12904606708342e-05, "loss": 0.0571, "step": 39200 }, { "grad_norm": 0.5777418613433838, "learning_rate": 7.127549869756687e-05, "loss": 0.0656, "step": 39210 }, { "grad_norm": 0.4702322781085968, "learning_rate": 7.126053439764299e-05, "loss": 0.0528, "step": 39220 }, { "grad_norm": 0.39780664443969727, "learning_rate": 7.124556777269904e-05, "loss": 0.0601, "step": 39230 }, { "grad_norm": 0.7934636473655701, "learning_rate": 7.123059882437174e-05, "loss": 0.0679, "step": 39240 }, { "grad_norm": 0.7869357466697693, "learning_rate": 7.121562755429807e-05, "loss": 0.056, "step": 39250 }, { "grad_norm": 0.38501620292663574, "learning_rate": 7.120065396411528e-05, "loss": 0.0543, "step": 39260 }, { "grad_norm": 0.4513856768608093, "learning_rate": 7.118567805546084e-05, "loss": 0.0528, "step": 39270 }, { "grad_norm": 0.6041397452354431, "learning_rate": 7.117069982997248e-05, "loss": 0.0557, "step": 39280 }, { "grad_norm": 0.4354332983493805, "learning_rate": 7.115571928928825e-05, "loss": 0.0777, "step": 39290 }, { "grad_norm": 0.6612229347229004, "learning_rate": 7.114073643504635e-05, "loss": 0.0527, "step": 39300 }, { "grad_norm": 0.7320302724838257, "learning_rate": 7.11257512688853e-05, "loss": 0.0564, "step": 39310 }, { "grad_norm": 0.48231932520866394, "learning_rate": 7.111076379244384e-05, "loss": 0.0412, "step": 39320 }, { "grad_norm": 0.7452765703201294, "learning_rate": 7.109577400736101e-05, "loss": 0.0565, "step": 39330 }, { "grad_norm": 0.47998037934303284, "learning_rate": 7.108078191527605e-05, "loss": 0.0588, "step": 39340 }, { "grad_norm": 0.5009056329727173, "learning_rate": 7.106578751782847e-05, "loss": 0.0616, "step": 39350 }, { "grad_norm": 0.6561254858970642, "learning_rate": 7.105079081665803e-05, "loss": 0.0519, "step": 39360 }, { "grad_norm": 0.5154011249542236, "learning_rate": 7.103579181340476e-05, "loss": 0.0644, "step": 39370 }, { "grad_norm": 1.6073485612869263, "learning_rate": 7.102079050970893e-05, "loss": 0.055, "step": 39380 }, { "grad_norm": 0.4102848470211029, "learning_rate": 7.100578690721105e-05, "loss": 0.059, "step": 39390 }, { "grad_norm": 0.47881656885147095, "learning_rate": 7.09907810075519e-05, "loss": 0.0563, "step": 39400 }, { "grad_norm": 0.5193561911582947, "learning_rate": 7.097577281237249e-05, "loss": 0.0608, "step": 39410 }, { "grad_norm": 0.9411050081253052, "learning_rate": 7.09607623233141e-05, "loss": 0.0398, "step": 39420 }, { "grad_norm": 0.7092535495758057, "learning_rate": 7.094574954201827e-05, "loss": 0.0581, "step": 39430 }, { "grad_norm": 0.5183830261230469, "learning_rate": 7.093073447012675e-05, "loss": 0.0583, "step": 39440 }, { "grad_norm": 0.6444512605667114, "learning_rate": 7.09157171092816e-05, "loss": 0.0518, "step": 39450 }, { "grad_norm": 0.6776873469352722, "learning_rate": 7.090069746112504e-05, "loss": 0.0601, "step": 39460 }, { "grad_norm": 0.7746493816375732, "learning_rate": 7.088567552729965e-05, "loss": 0.0497, "step": 39470 }, { "grad_norm": 0.5551180839538574, "learning_rate": 7.087065130944818e-05, "loss": 0.0603, "step": 39480 }, { "grad_norm": 0.5539163947105408, "learning_rate": 7.085562480921366e-05, "loss": 0.0634, "step": 39490 }, { "grad_norm": 0.5624970197677612, "learning_rate": 7.084059602823937e-05, "loss": 0.0583, "step": 39500 }, { "grad_norm": 0.7516185641288757, "learning_rate": 7.082556496816882e-05, "loss": 0.0565, "step": 39510 }, { "grad_norm": 0.6317214965820312, "learning_rate": 7.081053163064582e-05, "loss": 0.0683, "step": 39520 }, { "grad_norm": 0.788623034954071, "learning_rate": 7.079549601731434e-05, "loss": 0.058, "step": 39530 }, { "grad_norm": 0.6527904868125916, "learning_rate": 7.07804581298187e-05, "loss": 0.0561, "step": 39540 }, { "grad_norm": 0.6043756008148193, "learning_rate": 7.07654179698034e-05, "loss": 0.0636, "step": 39550 }, { "grad_norm": 0.8166341781616211, "learning_rate": 7.075037553891321e-05, "loss": 0.068, "step": 39560 }, { "grad_norm": 0.5103135108947754, "learning_rate": 7.073533083879315e-05, "loss": 0.0515, "step": 39570 }, { "grad_norm": 0.5172662138938904, "learning_rate": 7.072028387108849e-05, "loss": 0.0519, "step": 39580 }, { "grad_norm": 0.423788845539093, "learning_rate": 7.070523463744472e-05, "loss": 0.0631, "step": 39590 }, { "grad_norm": 0.34134799242019653, "learning_rate": 7.069018313950763e-05, "loss": 0.0536, "step": 39600 }, { "grad_norm": 0.5526277422904968, "learning_rate": 7.067512937892322e-05, "loss": 0.054, "step": 39610 }, { "grad_norm": 0.5354201793670654, "learning_rate": 7.066007335733775e-05, "loss": 0.0448, "step": 39620 }, { "grad_norm": 0.6038770079612732, "learning_rate": 7.064501507639772e-05, "loss": 0.0571, "step": 39630 }, { "grad_norm": 0.465716153383255, "learning_rate": 7.062995453774987e-05, "loss": 0.0685, "step": 39640 }, { "grad_norm": 0.47870442271232605, "learning_rate": 7.061489174304121e-05, "loss": 0.0733, "step": 39650 }, { "grad_norm": 0.33170872926712036, "learning_rate": 7.0599826693919e-05, "loss": 0.0548, "step": 39660 }, { "grad_norm": 0.6260983347892761, "learning_rate": 7.05847593920307e-05, "loss": 0.0563, "step": 39670 }, { "grad_norm": 1.114215612411499, "learning_rate": 7.056968983902406e-05, "loss": 0.0495, "step": 39680 }, { "grad_norm": 0.5650123357772827, "learning_rate": 7.055461803654706e-05, "loss": 0.0528, "step": 39690 }, { "grad_norm": 0.8493620157241821, "learning_rate": 7.053954398624794e-05, "loss": 0.0491, "step": 39700 }, { "grad_norm": 0.4254522919654846, "learning_rate": 7.052446768977518e-05, "loss": 0.0549, "step": 39710 }, { "grad_norm": 0.421359658241272, "learning_rate": 7.050938914877748e-05, "loss": 0.0529, "step": 39720 }, { "grad_norm": 0.5049706697463989, "learning_rate": 7.049430836490382e-05, "loss": 0.0692, "step": 39730 }, { "grad_norm": 0.4648221731185913, "learning_rate": 7.047922533980342e-05, "loss": 0.0519, "step": 39740 }, { "grad_norm": 0.8659952282905579, "learning_rate": 7.046414007512571e-05, "loss": 0.056, "step": 39750 }, { "grad_norm": 1.009719729423523, "learning_rate": 7.044905257252042e-05, "loss": 0.0513, "step": 39760 }, { "grad_norm": 0.7537956237792969, "learning_rate": 7.04339628336375e-05, "loss": 0.0625, "step": 39770 }, { "grad_norm": 0.4342609941959381, "learning_rate": 7.041887086012711e-05, "loss": 0.0499, "step": 39780 }, { "grad_norm": 0.7994714379310608, "learning_rate": 7.040377665363969e-05, "loss": 0.056, "step": 39790 }, { "grad_norm": 1.2022144794464111, "learning_rate": 7.038868021582594e-05, "loss": 0.0578, "step": 39800 }, { "grad_norm": 0.48175179958343506, "learning_rate": 7.037358154833679e-05, "loss": 0.0502, "step": 39810 }, { "grad_norm": 1.6275471448898315, "learning_rate": 7.035848065282339e-05, "loss": 0.0629, "step": 39820 }, { "grad_norm": 0.4126169979572296, "learning_rate": 7.034337753093714e-05, "loss": 0.0519, "step": 39830 }, { "grad_norm": 0.5136366486549377, "learning_rate": 7.032827218432972e-05, "loss": 0.0582, "step": 39840 }, { "grad_norm": 0.5578762888908386, "learning_rate": 7.031316461465302e-05, "loss": 0.0686, "step": 39850 }, { "grad_norm": 0.6362845301628113, "learning_rate": 7.029805482355915e-05, "loss": 0.0572, "step": 39860 }, { "grad_norm": 0.5834507346153259, "learning_rate": 7.028294281270055e-05, "loss": 0.0456, "step": 39870 }, { "grad_norm": 0.5296321511268616, "learning_rate": 7.026782858372982e-05, "loss": 0.0508, "step": 39880 }, { "grad_norm": 0.6062473058700562, "learning_rate": 7.025271213829982e-05, "loss": 0.0565, "step": 39890 }, { "grad_norm": 0.5326155424118042, "learning_rate": 7.023759347806366e-05, "loss": 0.0511, "step": 39900 }, { "grad_norm": 1.690826416015625, "learning_rate": 7.022247260467472e-05, "loss": 0.0608, "step": 39910 }, { "grad_norm": 0.5557600855827332, "learning_rate": 7.020734951978658e-05, "loss": 0.0559, "step": 39920 }, { "grad_norm": 0.6533111929893494, "learning_rate": 7.019222422505307e-05, "loss": 0.0616, "step": 39930 }, { "grad_norm": 0.5964019894599915, "learning_rate": 7.017709672212827e-05, "loss": 0.0624, "step": 39940 }, { "grad_norm": 0.5996084213256836, "learning_rate": 7.016196701266652e-05, "loss": 0.0633, "step": 39950 }, { "grad_norm": 0.7993064522743225, "learning_rate": 7.014683509832238e-05, "loss": 0.0542, "step": 39960 }, { "grad_norm": 0.7364700436592102, "learning_rate": 7.013170098075063e-05, "loss": 0.0647, "step": 39970 }, { "grad_norm": 0.47997593879699707, "learning_rate": 7.011656466160632e-05, "loss": 0.056, "step": 39980 }, { "grad_norm": 0.49469125270843506, "learning_rate": 7.010142614254475e-05, "loss": 0.0481, "step": 39990 }, { "grad_norm": 0.41336819529533386, "learning_rate": 7.008628542522147e-05, "loss": 0.0579, "step": 40000 }, { "grad_norm": 0.5951947569847107, "learning_rate": 7.007114251129218e-05, "loss": 0.053, "step": 40010 }, { "grad_norm": 0.7484080791473389, "learning_rate": 7.005599740241293e-05, "loss": 0.0418, "step": 40020 }, { "grad_norm": 0.6078843474388123, "learning_rate": 7.004085010023996e-05, "loss": 0.0603, "step": 40030 }, { "grad_norm": 0.7891576886177063, "learning_rate": 7.002570060642976e-05, "loss": 0.0479, "step": 40040 }, { "grad_norm": 0.9247652888298035, "learning_rate": 7.001054892263903e-05, "loss": 0.0456, "step": 40050 }, { "grad_norm": 0.5485419631004333, "learning_rate": 6.999539505052477e-05, "loss": 0.0625, "step": 40060 }, { "grad_norm": 0.5240855813026428, "learning_rate": 6.998023899174418e-05, "loss": 0.0438, "step": 40070 }, { "grad_norm": 0.5000883340835571, "learning_rate": 6.996508074795467e-05, "loss": 0.0519, "step": 40080 }, { "grad_norm": 0.5877660512924194, "learning_rate": 6.994992032081396e-05, "loss": 0.0641, "step": 40090 }, { "grad_norm": 0.5281187295913696, "learning_rate": 6.993475771197995e-05, "loss": 0.0585, "step": 40100 }, { "grad_norm": 0.6277592778205872, "learning_rate": 6.991959292311082e-05, "loss": 0.05, "step": 40110 }, { "grad_norm": 0.4865998923778534, "learning_rate": 6.990442595586495e-05, "loss": 0.0581, "step": 40120 }, { "grad_norm": 1.0631088018417358, "learning_rate": 6.988925681190098e-05, "loss": 0.0697, "step": 40130 }, { "grad_norm": 0.7026680707931519, "learning_rate": 6.987408549287778e-05, "loss": 0.0714, "step": 40140 }, { "grad_norm": 0.8409196734428406, "learning_rate": 6.985891200045449e-05, "loss": 0.0706, "step": 40150 }, { "grad_norm": 0.638158917427063, "learning_rate": 6.984373633629045e-05, "loss": 0.0715, "step": 40160 }, { "grad_norm": 1.6720067262649536, "learning_rate": 6.98285585020452e-05, "loss": 0.0625, "step": 40170 }, { "grad_norm": 0.4131622314453125, "learning_rate": 6.981337849937864e-05, "loss": 0.0578, "step": 40180 }, { "grad_norm": 0.9256375432014465, "learning_rate": 6.979819632995078e-05, "loss": 0.0746, "step": 40190 }, { "grad_norm": 0.39996209740638733, "learning_rate": 6.978301199542193e-05, "loss": 0.0415, "step": 40200 }, { "grad_norm": 0.4902171194553375, "learning_rate": 6.976782549745263e-05, "loss": 0.0546, "step": 40210 }, { "grad_norm": 0.4580894112586975, "learning_rate": 6.975263683770367e-05, "loss": 0.0686, "step": 40220 }, { "grad_norm": 0.4636147618293762, "learning_rate": 6.973744601783604e-05, "loss": 0.0618, "step": 40230 }, { "grad_norm": 0.5612168312072754, "learning_rate": 6.972225303951097e-05, "loss": 0.0554, "step": 40240 }, { "grad_norm": 0.47948771715164185, "learning_rate": 6.970705790438998e-05, "loss": 0.0505, "step": 40250 }, { "grad_norm": 0.844555139541626, "learning_rate": 6.969186061413477e-05, "loss": 0.0577, "step": 40260 }, { "grad_norm": 0.6007015705108643, "learning_rate": 6.967666117040727e-05, "loss": 0.0629, "step": 40270 }, { "grad_norm": 0.4732908308506012, "learning_rate": 6.966145957486972e-05, "loss": 0.0538, "step": 40280 }, { "grad_norm": 0.4997885227203369, "learning_rate": 6.964625582918449e-05, "loss": 0.0581, "step": 40290 }, { "grad_norm": 0.39856064319610596, "learning_rate": 6.963104993501425e-05, "loss": 0.0539, "step": 40300 }, { "grad_norm": 0.5516108274459839, "learning_rate": 6.961584189402192e-05, "loss": 0.0419, "step": 40310 }, { "grad_norm": 0.6970996260643005, "learning_rate": 6.96006317078706e-05, "loss": 0.0422, "step": 40320 }, { "grad_norm": 0.7950048446655273, "learning_rate": 6.958541937822367e-05, "loss": 0.0499, "step": 40330 }, { "grad_norm": 0.5306769609451294, "learning_rate": 6.957020490674473e-05, "loss": 0.057, "step": 40340 }, { "grad_norm": 0.4931422472000122, "learning_rate": 6.95549882950976e-05, "loss": 0.0521, "step": 40350 }, { "grad_norm": 0.6388999819755554, "learning_rate": 6.953976954494635e-05, "loss": 0.0396, "step": 40360 }, { "grad_norm": 0.5807745456695557, "learning_rate": 6.952454865795528e-05, "loss": 0.0508, "step": 40370 }, { "grad_norm": 0.905439019203186, "learning_rate": 6.950932563578892e-05, "loss": 0.056, "step": 40380 }, { "grad_norm": 0.5561270713806152, "learning_rate": 6.949410048011206e-05, "loss": 0.0536, "step": 40390 }, { "grad_norm": 0.5356329679489136, "learning_rate": 6.947887319258966e-05, "loss": 0.0774, "step": 40400 }, { "grad_norm": 0.5925329923629761, "learning_rate": 6.946364377488696e-05, "loss": 0.0579, "step": 40410 }, { "grad_norm": 0.9383106231689453, "learning_rate": 6.944841222866947e-05, "loss": 0.0734, "step": 40420 }, { "grad_norm": 0.4838906228542328, "learning_rate": 6.943317855560284e-05, "loss": 0.0615, "step": 40430 }, { "grad_norm": 0.9866542816162109, "learning_rate": 6.941794275735302e-05, "loss": 0.0519, "step": 40440 }, { "grad_norm": 0.47048628330230713, "learning_rate": 6.94027048355862e-05, "loss": 0.0467, "step": 40450 }, { "grad_norm": 0.9423754215240479, "learning_rate": 6.938746479196871e-05, "loss": 0.0589, "step": 40460 }, { "grad_norm": 0.7328784465789795, "learning_rate": 6.937222262816724e-05, "loss": 0.0493, "step": 40470 }, { "grad_norm": 0.25478649139404297, "learning_rate": 6.935697834584865e-05, "loss": 0.0481, "step": 40480 }, { "grad_norm": 0.5389764904975891, "learning_rate": 6.934173194667998e-05, "loss": 0.0509, "step": 40490 }, { "grad_norm": 0.6061449646949768, "learning_rate": 6.93264834323286e-05, "loss": 0.0594, "step": 40500 }, { "grad_norm": 1.19834303855896, "learning_rate": 6.931123280446206e-05, "loss": 0.0595, "step": 40510 }, { "grad_norm": 0.5271507501602173, "learning_rate": 6.929598006474811e-05, "loss": 0.0514, "step": 40520 }, { "grad_norm": 0.3537713587284088, "learning_rate": 6.92807252148548e-05, "loss": 0.0716, "step": 40530 }, { "grad_norm": 0.6415311098098755, "learning_rate": 6.926546825645037e-05, "loss": 0.0458, "step": 40540 }, { "grad_norm": 0.46468374133110046, "learning_rate": 6.92502091912033e-05, "loss": 0.0467, "step": 40550 }, { "grad_norm": 0.536281406879425, "learning_rate": 6.923494802078229e-05, "loss": 0.0548, "step": 40560 }, { "grad_norm": 0.5576459765434265, "learning_rate": 6.921968474685628e-05, "loss": 0.05, "step": 40570 }, { "grad_norm": 0.6102668046951294, "learning_rate": 6.920441937109445e-05, "loss": 0.0763, "step": 40580 }, { "grad_norm": 0.4386284649372101, "learning_rate": 6.91891518951662e-05, "loss": 0.0492, "step": 40590 }, { "grad_norm": 0.6398252248764038, "learning_rate": 6.917388232074114e-05, "loss": 0.0491, "step": 40600 }, { "grad_norm": 0.5858902931213379, "learning_rate": 6.915861064948914e-05, "loss": 0.0608, "step": 40610 }, { "grad_norm": 0.3635979890823364, "learning_rate": 6.914333688308025e-05, "loss": 0.0416, "step": 40620 }, { "grad_norm": 0.4705953001976013, "learning_rate": 6.912806102318487e-05, "loss": 0.045, "step": 40630 }, { "grad_norm": 0.7601523995399475, "learning_rate": 6.911278307147347e-05, "loss": 0.0565, "step": 40640 }, { "grad_norm": 0.3579935133457184, "learning_rate": 6.909750302961684e-05, "loss": 0.0529, "step": 40650 }, { "grad_norm": 2.1296703815460205, "learning_rate": 6.9082220899286e-05, "loss": 0.0642, "step": 40660 }, { "grad_norm": 1.1260703802108765, "learning_rate": 6.906693668215219e-05, "loss": 0.0772, "step": 40670 }, { "grad_norm": 0.5170148611068726, "learning_rate": 6.905165037988683e-05, "loss": 0.0484, "step": 40680 }, { "grad_norm": 0.8654814958572388, "learning_rate": 6.903636199416164e-05, "loss": 0.0574, "step": 40690 }, { "grad_norm": 0.4879842698574066, "learning_rate": 6.902107152664851e-05, "loss": 0.0549, "step": 40700 }, { "grad_norm": 0.5968148112297058, "learning_rate": 6.90057789790196e-05, "loss": 0.0531, "step": 40710 }, { "grad_norm": 0.4247645139694214, "learning_rate": 6.899048435294728e-05, "loss": 0.0606, "step": 40720 }, { "grad_norm": 0.5540153980255127, "learning_rate": 6.897518765010415e-05, "loss": 0.055, "step": 40730 }, { "grad_norm": 0.4238070249557495, "learning_rate": 6.895988887216302e-05, "loss": 0.0631, "step": 40740 }, { "grad_norm": 0.5825506448745728, "learning_rate": 6.894458802079694e-05, "loss": 0.0598, "step": 40750 }, { "grad_norm": 0.6931936740875244, "learning_rate": 6.892928509767922e-05, "loss": 0.0483, "step": 40760 }, { "grad_norm": 0.629880964756012, "learning_rate": 6.891398010448333e-05, "loss": 0.0566, "step": 40770 }, { "grad_norm": 0.7573308944702148, "learning_rate": 6.8898673042883e-05, "loss": 0.0705, "step": 40780 }, { "grad_norm": 0.5599467158317566, "learning_rate": 6.888336391455222e-05, "loss": 0.057, "step": 40790 }, { "grad_norm": 0.45653215050697327, "learning_rate": 6.886805272116513e-05, "loss": 0.052, "step": 40800 }, { "grad_norm": 0.6825516819953918, "learning_rate": 6.885273946439617e-05, "loss": 0.0717, "step": 40810 }, { "grad_norm": 0.8051032423973083, "learning_rate": 6.883742414591998e-05, "loss": 0.0694, "step": 40820 }, { "grad_norm": 0.7320258021354675, "learning_rate": 6.88221067674114e-05, "loss": 0.0526, "step": 40830 }, { "grad_norm": 1.5086817741394043, "learning_rate": 6.88067873305455e-05, "loss": 0.0621, "step": 40840 }, { "grad_norm": 0.41725772619247437, "learning_rate": 6.879146583699765e-05, "loss": 0.054, "step": 40850 }, { "grad_norm": 0.42771604657173157, "learning_rate": 6.877614228844334e-05, "loss": 0.0517, "step": 40860 }, { "grad_norm": 0.7422286868095398, "learning_rate": 6.876081668655832e-05, "loss": 0.0518, "step": 40870 }, { "grad_norm": 0.5310335755348206, "learning_rate": 6.874548903301861e-05, "loss": 0.0609, "step": 40880 }, { "grad_norm": 0.4388831853866577, "learning_rate": 6.873015932950039e-05, "loss": 0.049, "step": 40890 }, { "grad_norm": 1.0076884031295776, "learning_rate": 6.871482757768012e-05, "loss": 0.0569, "step": 40900 }, { "grad_norm": 0.5297008752822876, "learning_rate": 6.869949377923442e-05, "loss": 0.0634, "step": 40910 }, { "grad_norm": 0.6370620727539062, "learning_rate": 6.868415793584022e-05, "loss": 0.0617, "step": 40920 }, { "grad_norm": 0.42474880814552307, "learning_rate": 6.86688200491746e-05, "loss": 0.0472, "step": 40930 }, { "grad_norm": 0.3398754894733429, "learning_rate": 6.865348012091488e-05, "loss": 0.0495, "step": 40940 }, { "grad_norm": 0.41080915927886963, "learning_rate": 6.86381381527386e-05, "loss": 0.0506, "step": 40950 }, { "grad_norm": 0.6695840358734131, "learning_rate": 6.862279414632359e-05, "loss": 0.0526, "step": 40960 }, { "grad_norm": 0.5214352607727051, "learning_rate": 6.860744810334778e-05, "loss": 0.0472, "step": 40970 }, { "grad_norm": 0.49516648054122925, "learning_rate": 6.859210002548943e-05, "loss": 0.0542, "step": 40980 }, { "grad_norm": 0.5701211094856262, "learning_rate": 6.8576749914427e-05, "loss": 0.0521, "step": 40990 }, { "grad_norm": 0.694182276725769, "learning_rate": 6.85613977718391e-05, "loss": 0.0528, "step": 41000 }, { "grad_norm": 0.7780691385269165, "learning_rate": 6.854604359940464e-05, "loss": 0.0522, "step": 41010 }, { "grad_norm": 1.0944254398345947, "learning_rate": 6.853068739880276e-05, "loss": 0.0607, "step": 41020 }, { "grad_norm": 0.532897412776947, "learning_rate": 6.851532917171276e-05, "loss": 0.0562, "step": 41030 }, { "grad_norm": 0.583747923374176, "learning_rate": 6.849996891981418e-05, "loss": 0.0676, "step": 41040 }, { "grad_norm": 0.8510700464248657, "learning_rate": 6.848460664478684e-05, "loss": 0.0639, "step": 41050 }, { "grad_norm": 0.6803213953971863, "learning_rate": 6.846924234831069e-05, "loss": 0.0562, "step": 41060 }, { "grad_norm": 0.5741508603096008, "learning_rate": 6.845387603206597e-05, "loss": 0.0663, "step": 41070 }, { "grad_norm": 0.3925703167915344, "learning_rate": 6.843850769773312e-05, "loss": 0.0472, "step": 41080 }, { "grad_norm": 0.5178133845329285, "learning_rate": 6.842313734699278e-05, "loss": 0.0617, "step": 41090 }, { "grad_norm": 0.5295180082321167, "learning_rate": 6.840776498152584e-05, "loss": 0.055, "step": 41100 }, { "grad_norm": 0.6502736806869507, "learning_rate": 6.83923906030134e-05, "loss": 0.0433, "step": 41110 }, { "grad_norm": 1.1277333498001099, "learning_rate": 6.837701421313677e-05, "loss": 0.0701, "step": 41120 }, { "grad_norm": 0.6544389128684998, "learning_rate": 6.83616358135775e-05, "loss": 0.0579, "step": 41130 }, { "grad_norm": 0.8582653403282166, "learning_rate": 6.834625540601734e-05, "loss": 0.0736, "step": 41140 }, { "grad_norm": 0.663052499294281, "learning_rate": 6.833087299213829e-05, "loss": 0.0473, "step": 41150 }, { "grad_norm": 0.6838635206222534, "learning_rate": 6.83154885736225e-05, "loss": 0.0617, "step": 41160 }, { "grad_norm": 0.5210504531860352, "learning_rate": 6.830010215215244e-05, "loss": 0.0594, "step": 41170 }, { "grad_norm": 0.6521650552749634, "learning_rate": 6.828471372941072e-05, "loss": 0.0601, "step": 41180 }, { "grad_norm": 0.6222001314163208, "learning_rate": 6.826932330708018e-05, "loss": 0.046, "step": 41190 }, { "grad_norm": 0.6148068308830261, "learning_rate": 6.825393088684393e-05, "loss": 0.0503, "step": 41200 }, { "grad_norm": 0.5869789719581604, "learning_rate": 6.823853647038524e-05, "loss": 0.0504, "step": 41210 }, { "grad_norm": 0.337240606546402, "learning_rate": 6.822314005938762e-05, "loss": 0.0461, "step": 41220 }, { "grad_norm": 1.3436282873153687, "learning_rate": 6.820774165553479e-05, "loss": 0.0683, "step": 41230 }, { "grad_norm": 1.5405031442642212, "learning_rate": 6.819234126051072e-05, "loss": 0.0482, "step": 41240 }, { "grad_norm": 0.4461405575275421, "learning_rate": 6.817693887599956e-05, "loss": 0.0667, "step": 41250 }, { "grad_norm": 0.38100969791412354, "learning_rate": 6.816153450368568e-05, "loss": 0.062, "step": 41260 }, { "grad_norm": 0.4322863221168518, "learning_rate": 6.814612814525372e-05, "loss": 0.0612, "step": 41270 }, { "grad_norm": 0.6463562250137329, "learning_rate": 6.813071980238846e-05, "loss": 0.0557, "step": 41280 }, { "grad_norm": 0.5146586894989014, "learning_rate": 6.811530947677492e-05, "loss": 0.0544, "step": 41290 }, { "grad_norm": 0.49679896235466003, "learning_rate": 6.809989717009839e-05, "loss": 0.0523, "step": 41300 }, { "grad_norm": 0.6125856637954712, "learning_rate": 6.808448288404431e-05, "loss": 0.0653, "step": 41310 }, { "grad_norm": 0.49731290340423584, "learning_rate": 6.806906662029838e-05, "loss": 0.0571, "step": 41320 }, { "grad_norm": 1.8831777572631836, "learning_rate": 6.805364838054647e-05, "loss": 0.0683, "step": 41330 }, { "grad_norm": 0.49538952112197876, "learning_rate": 6.803822816647474e-05, "loss": 0.044, "step": 41340 }, { "grad_norm": 0.5646185278892517, "learning_rate": 6.802280597976949e-05, "loss": 0.0528, "step": 41350 }, { "grad_norm": 0.5420350432395935, "learning_rate": 6.800738182211727e-05, "loss": 0.0499, "step": 41360 }, { "grad_norm": 1.1964130401611328, "learning_rate": 6.799195569520485e-05, "loss": 0.0521, "step": 41370 }, { "grad_norm": 0.6294392943382263, "learning_rate": 6.79765276007192e-05, "loss": 0.0573, "step": 41380 }, { "grad_norm": 0.72060227394104, "learning_rate": 6.796109754034753e-05, "loss": 0.0523, "step": 41390 }, { "grad_norm": 0.5446924567222595, "learning_rate": 6.794566551577724e-05, "loss": 0.0518, "step": 41400 }, { "grad_norm": 1.696578860282898, "learning_rate": 6.793023152869594e-05, "loss": 0.0538, "step": 41410 }, { "grad_norm": 0.8625355362892151, "learning_rate": 6.791479558079147e-05, "loss": 0.0593, "step": 41420 }, { "grad_norm": 0.3519066870212555, "learning_rate": 6.789935767375191e-05, "loss": 0.0567, "step": 41430 }, { "grad_norm": 0.38546183705329895, "learning_rate": 6.78839178092655e-05, "loss": 0.0376, "step": 41440 }, { "grad_norm": 0.5115665197372437, "learning_rate": 6.786847598902072e-05, "loss": 0.0592, "step": 41450 }, { "grad_norm": 0.5637213587760925, "learning_rate": 6.785303221470627e-05, "loss": 0.0603, "step": 41460 }, { "grad_norm": 0.5094977021217346, "learning_rate": 6.783758648801108e-05, "loss": 0.044, "step": 41470 }, { "grad_norm": 0.6776844263076782, "learning_rate": 6.782213881062422e-05, "loss": 0.0573, "step": 41480 }, { "grad_norm": 0.6701709032058716, "learning_rate": 6.780668918423508e-05, "loss": 0.0497, "step": 41490 }, { "grad_norm": 1.087773084640503, "learning_rate": 6.779123761053317e-05, "loss": 0.0504, "step": 41500 }, { "grad_norm": 0.38600850105285645, "learning_rate": 6.777578409120826e-05, "loss": 0.0656, "step": 41510 }, { "grad_norm": 0.6046774983406067, "learning_rate": 6.776032862795033e-05, "loss": 0.0663, "step": 41520 }, { "grad_norm": 0.7912942171096802, "learning_rate": 6.774487122244956e-05, "loss": 0.059, "step": 41530 }, { "grad_norm": 1.3999202251434326, "learning_rate": 6.772941187639637e-05, "loss": 0.0487, "step": 41540 }, { "grad_norm": 0.43954750895500183, "learning_rate": 6.771395059148134e-05, "loss": 0.0654, "step": 41550 }, { "grad_norm": 0.5816747546195984, "learning_rate": 6.76984873693953e-05, "loss": 0.059, "step": 41560 }, { "grad_norm": 0.6718076467514038, "learning_rate": 6.768302221182931e-05, "loss": 0.062, "step": 41570 }, { "grad_norm": 0.6694862246513367, "learning_rate": 6.766755512047457e-05, "loss": 0.0653, "step": 41580 }, { "grad_norm": 0.6040195822715759, "learning_rate": 6.765208609702259e-05, "loss": 0.0635, "step": 41590 }, { "grad_norm": 0.6392684578895569, "learning_rate": 6.763661514316499e-05, "loss": 0.0511, "step": 41600 }, { "grad_norm": 0.32691875100135803, "learning_rate": 6.76211422605937e-05, "loss": 0.0531, "step": 41610 }, { "grad_norm": 0.5028720498085022, "learning_rate": 6.760566745100076e-05, "loss": 0.0523, "step": 41620 }, { "grad_norm": 0.9377782940864563, "learning_rate": 6.75901907160785e-05, "loss": 0.0509, "step": 41630 }, { "grad_norm": 0.2676413059234619, "learning_rate": 6.757471205751943e-05, "loss": 0.0416, "step": 41640 }, { "grad_norm": 0.5212651491165161, "learning_rate": 6.75592314770163e-05, "loss": 0.0575, "step": 41650 }, { "grad_norm": 0.6087678670883179, "learning_rate": 6.7543748976262e-05, "loss": 0.0621, "step": 41660 }, { "grad_norm": 0.5135816931724548, "learning_rate": 6.752826455694968e-05, "loss": 0.0592, "step": 41670 }, { "grad_norm": 0.5883086323738098, "learning_rate": 6.751277822077271e-05, "loss": 0.0511, "step": 41680 }, { "grad_norm": 0.3478078246116638, "learning_rate": 6.749728996942465e-05, "loss": 0.0439, "step": 41690 }, { "grad_norm": 0.33514803647994995, "learning_rate": 6.748179980459924e-05, "loss": 0.049, "step": 41700 }, { "grad_norm": 0.38818028569221497, "learning_rate": 6.746630772799052e-05, "loss": 0.0431, "step": 41710 }, { "grad_norm": 0.5637561082839966, "learning_rate": 6.745081374129262e-05, "loss": 0.0576, "step": 41720 }, { "grad_norm": 0.8858032822608948, "learning_rate": 6.743531784619998e-05, "loss": 0.063, "step": 41730 }, { "grad_norm": 0.662124514579773, "learning_rate": 6.741982004440719e-05, "loss": 0.0586, "step": 41740 }, { "grad_norm": 0.6354872584342957, "learning_rate": 6.740432033760907e-05, "loss": 0.0579, "step": 41750 }, { "grad_norm": 0.502484142780304, "learning_rate": 6.738881872750066e-05, "loss": 0.0507, "step": 41760 }, { "grad_norm": 0.43153002858161926, "learning_rate": 6.737331521577715e-05, "loss": 0.0625, "step": 41770 }, { "grad_norm": 0.6851864457130432, "learning_rate": 6.735780980413403e-05, "loss": 0.0531, "step": 41780 }, { "grad_norm": 0.6224169135093689, "learning_rate": 6.734230249426692e-05, "loss": 0.0575, "step": 41790 }, { "grad_norm": 0.7544516324996948, "learning_rate": 6.732679328787168e-05, "loss": 0.0609, "step": 41800 }, { "grad_norm": 0.3502284586429596, "learning_rate": 6.731128218664438e-05, "loss": 0.0427, "step": 41810 }, { "grad_norm": 0.4780668020248413, "learning_rate": 6.729576919228129e-05, "loss": 0.0431, "step": 41820 }, { "grad_norm": 0.5298231244087219, "learning_rate": 6.728025430647888e-05, "loss": 0.0457, "step": 41830 }, { "grad_norm": 0.5005666613578796, "learning_rate": 6.726473753093383e-05, "loss": 0.0544, "step": 41840 }, { "grad_norm": 0.527558445930481, "learning_rate": 6.724921886734305e-05, "loss": 0.0451, "step": 41850 }, { "grad_norm": 0.6413397192955017, "learning_rate": 6.723369831740363e-05, "loss": 0.0622, "step": 41860 }, { "grad_norm": 0.3288251757621765, "learning_rate": 6.721817588281288e-05, "loss": 0.0553, "step": 41870 }, { "grad_norm": 0.5646262168884277, "learning_rate": 6.720265156526828e-05, "loss": 0.0474, "step": 41880 }, { "grad_norm": 0.9005760550498962, "learning_rate": 6.718712536646758e-05, "loss": 0.046, "step": 41890 }, { "grad_norm": 1.0797640085220337, "learning_rate": 6.71715972881087e-05, "loss": 0.0537, "step": 41900 }, { "grad_norm": 0.4116118848323822, "learning_rate": 6.715606733188974e-05, "loss": 0.0482, "step": 41910 }, { "grad_norm": 0.3905658423900604, "learning_rate": 6.714053549950909e-05, "loss": 0.0512, "step": 41920 }, { "grad_norm": 0.7454348802566528, "learning_rate": 6.71250017926652e-05, "loss": 0.0584, "step": 41930 }, { "grad_norm": 0.5735909938812256, "learning_rate": 6.71094662130569e-05, "loss": 0.0498, "step": 41940 }, { "grad_norm": 0.8890023827552795, "learning_rate": 6.709392876238307e-05, "loss": 0.0557, "step": 41950 }, { "grad_norm": 0.9315004348754883, "learning_rate": 6.707838944234289e-05, "loss": 0.0627, "step": 41960 }, { "grad_norm": 0.5469741225242615, "learning_rate": 6.706284825463574e-05, "loss": 0.0477, "step": 41970 }, { "grad_norm": 0.5571537613868713, "learning_rate": 6.704730520096114e-05, "loss": 0.0502, "step": 41980 }, { "grad_norm": 0.6351865530014038, "learning_rate": 6.703176028301888e-05, "loss": 0.0516, "step": 41990 }, { "grad_norm": 0.5015770792961121, "learning_rate": 6.701621350250892e-05, "loss": 0.0554, "step": 42000 }, { "grad_norm": 0.9326778054237366, "learning_rate": 6.700066486113144e-05, "loss": 0.0478, "step": 42010 }, { "grad_norm": 0.5406900644302368, "learning_rate": 6.69851143605868e-05, "loss": 0.0565, "step": 42020 }, { "grad_norm": 0.49130162596702576, "learning_rate": 6.696956200257561e-05, "loss": 0.046, "step": 42030 }, { "grad_norm": 0.4261930286884308, "learning_rate": 6.695400778879863e-05, "loss": 0.0577, "step": 42040 }, { "grad_norm": 0.8853426575660706, "learning_rate": 6.693845172095683e-05, "loss": 0.0687, "step": 42050 }, { "grad_norm": 0.47890233993530273, "learning_rate": 6.692289380075142e-05, "loss": 0.0509, "step": 42060 }, { "grad_norm": 0.5684167146682739, "learning_rate": 6.690733402988379e-05, "loss": 0.0482, "step": 42070 }, { "grad_norm": 0.41325512528419495, "learning_rate": 6.689177241005553e-05, "loss": 0.0551, "step": 42080 }, { "grad_norm": 0.8491553068161011, "learning_rate": 6.687620894296844e-05, "loss": 0.0516, "step": 42090 }, { "grad_norm": 0.6071139574050903, "learning_rate": 6.686064363032451e-05, "loss": 0.0586, "step": 42100 }, { "grad_norm": 0.9924774765968323, "learning_rate": 6.684507647382596e-05, "loss": 0.0553, "step": 42110 }, { "grad_norm": 0.706654965877533, "learning_rate": 6.682950747517513e-05, "loss": 0.0617, "step": 42120 }, { "grad_norm": 0.8945185542106628, "learning_rate": 6.68139366360747e-05, "loss": 0.0652, "step": 42130 }, { "grad_norm": 0.5021058917045593, "learning_rate": 6.679836395822744e-05, "loss": 0.0505, "step": 42140 }, { "grad_norm": 0.7070976495742798, "learning_rate": 6.678278944333633e-05, "loss": 0.0489, "step": 42150 }, { "grad_norm": 0.7109405398368835, "learning_rate": 6.676721309310462e-05, "loss": 0.0481, "step": 42160 }, { "grad_norm": 0.6492284536361694, "learning_rate": 6.67516349092357e-05, "loss": 0.0601, "step": 42170 }, { "grad_norm": 0.37813976407051086, "learning_rate": 6.673605489343315e-05, "loss": 0.0428, "step": 42180 }, { "grad_norm": 0.5507404208183289, "learning_rate": 6.672047304740081e-05, "loss": 0.0482, "step": 42190 }, { "grad_norm": 0.8865291476249695, "learning_rate": 6.670488937284268e-05, "loss": 0.0443, "step": 42200 }, { "grad_norm": 0.6899557709693909, "learning_rate": 6.668930387146294e-05, "loss": 0.0499, "step": 42210 }, { "grad_norm": 1.3765652179718018, "learning_rate": 6.667371654496605e-05, "loss": 0.0548, "step": 42220 }, { "grad_norm": 0.574351966381073, "learning_rate": 6.665812739505659e-05, "loss": 0.0633, "step": 42230 }, { "grad_norm": 0.642322301864624, "learning_rate": 6.664253642343935e-05, "loss": 0.0579, "step": 42240 }, { "grad_norm": 0.8268353939056396, "learning_rate": 6.662694363181935e-05, "loss": 0.0487, "step": 42250 }, { "grad_norm": 0.6086351275444031, "learning_rate": 6.66113490219018e-05, "loss": 0.0639, "step": 42260 }, { "grad_norm": 0.473092257976532, "learning_rate": 6.659575259539207e-05, "loss": 0.0521, "step": 42270 }, { "grad_norm": 0.7086127996444702, "learning_rate": 6.65801543539958e-05, "loss": 0.0598, "step": 42280 }, { "grad_norm": 0.6391440033912659, "learning_rate": 6.65645542994188e-05, "loss": 0.05, "step": 42290 }, { "grad_norm": 0.5097687244415283, "learning_rate": 6.654895243336702e-05, "loss": 0.0486, "step": 42300 }, { "grad_norm": 0.5357710123062134, "learning_rate": 6.65333487575467e-05, "loss": 0.0396, "step": 42310 }, { "grad_norm": 0.6099088788032532, "learning_rate": 6.651774327366422e-05, "loss": 0.0461, "step": 42320 }, { "grad_norm": 0.57790607213974, "learning_rate": 6.650213598342613e-05, "loss": 0.0479, "step": 42330 }, { "grad_norm": 0.36490872502326965, "learning_rate": 6.648652688853931e-05, "loss": 0.0518, "step": 42340 }, { "grad_norm": 0.8354405760765076, "learning_rate": 6.647091599071066e-05, "loss": 0.0543, "step": 42350 }, { "grad_norm": 0.39951929450035095, "learning_rate": 6.645530329164743e-05, "loss": 0.0463, "step": 42360 }, { "grad_norm": 0.4294380247592926, "learning_rate": 6.643968879305694e-05, "loss": 0.0463, "step": 42370 }, { "grad_norm": 0.4999144673347473, "learning_rate": 6.642407249664683e-05, "loss": 0.0519, "step": 42380 }, { "grad_norm": 0.5623279809951782, "learning_rate": 6.640845440412483e-05, "loss": 0.0625, "step": 42390 }, { "grad_norm": 1.2075142860412598, "learning_rate": 6.639283451719893e-05, "loss": 0.0512, "step": 42400 }, { "grad_norm": 0.49376383423805237, "learning_rate": 6.637721283757729e-05, "loss": 0.0397, "step": 42410 }, { "grad_norm": 0.44997286796569824, "learning_rate": 6.636158936696828e-05, "loss": 0.0477, "step": 42420 }, { "grad_norm": 0.757218062877655, "learning_rate": 6.634596410708047e-05, "loss": 0.063, "step": 42430 }, { "grad_norm": 0.4900311231613159, "learning_rate": 6.633033705962257e-05, "loss": 0.0507, "step": 42440 }, { "grad_norm": 0.5882779359817505, "learning_rate": 6.631470822630359e-05, "loss": 0.0573, "step": 42450 }, { "grad_norm": 1.068205714225769, "learning_rate": 6.629907760883263e-05, "loss": 0.054, "step": 42460 }, { "grad_norm": 0.4785248339176178, "learning_rate": 6.628344520891905e-05, "loss": 0.0466, "step": 42470 }, { "grad_norm": 0.47936877608299255, "learning_rate": 6.62678110282724e-05, "loss": 0.0362, "step": 42480 }, { "grad_norm": 0.7035666108131409, "learning_rate": 6.625217506860237e-05, "loss": 0.0545, "step": 42490 }, { "grad_norm": 0.37630173563957214, "learning_rate": 6.623653733161892e-05, "loss": 0.0676, "step": 42500 }, { "grad_norm": 0.4872184097766876, "learning_rate": 6.622089781903217e-05, "loss": 0.0566, "step": 42510 }, { "grad_norm": 0.43526121973991394, "learning_rate": 6.620525653255241e-05, "loss": 0.052, "step": 42520 }, { "grad_norm": 0.5702306032180786, "learning_rate": 6.618961347389015e-05, "loss": 0.0557, "step": 42530 }, { "grad_norm": 0.6168506145477295, "learning_rate": 6.617396864475613e-05, "loss": 0.0527, "step": 42540 }, { "grad_norm": 0.49343976378440857, "learning_rate": 6.61583220468612e-05, "loss": 0.0464, "step": 42550 }, { "grad_norm": 0.5791578888893127, "learning_rate": 6.614267368191645e-05, "loss": 0.0506, "step": 42560 }, { "grad_norm": 0.42346203327178955, "learning_rate": 6.61270235516332e-05, "loss": 0.0566, "step": 42570 }, { "grad_norm": 0.4235984683036804, "learning_rate": 6.611137165772287e-05, "loss": 0.0414, "step": 42580 }, { "grad_norm": 0.43083298206329346, "learning_rate": 6.60957180018972e-05, "loss": 0.0536, "step": 42590 }, { "grad_norm": 0.45222848653793335, "learning_rate": 6.608006258586797e-05, "loss": 0.0454, "step": 42600 }, { "grad_norm": 0.7140845656394958, "learning_rate": 6.606440541134731e-05, "loss": 0.0581, "step": 42610 }, { "grad_norm": 0.7278788089752197, "learning_rate": 6.60487464800474e-05, "loss": 0.0657, "step": 42620 }, { "grad_norm": 0.6254517436027527, "learning_rate": 6.603308579368071e-05, "loss": 0.0588, "step": 42630 }, { "grad_norm": 0.8476150035858154, "learning_rate": 6.601742335395987e-05, "loss": 0.0569, "step": 42640 }, { "grad_norm": 0.5752871036529541, "learning_rate": 6.600175916259769e-05, "loss": 0.058, "step": 42650 }, { "grad_norm": 0.5283028483390808, "learning_rate": 6.598609322130718e-05, "loss": 0.0446, "step": 42660 }, { "grad_norm": 0.7304559946060181, "learning_rate": 6.597042553180155e-05, "loss": 0.0471, "step": 42670 }, { "grad_norm": 0.5389836430549622, "learning_rate": 6.59547560957942e-05, "loss": 0.05, "step": 42680 }, { "grad_norm": 2.662811756134033, "learning_rate": 6.59390849149987e-05, "loss": 0.0628, "step": 42690 }, { "grad_norm": 0.7891979217529297, "learning_rate": 6.592341199112886e-05, "loss": 0.0453, "step": 42700 }, { "grad_norm": 0.45006227493286133, "learning_rate": 6.590773732589861e-05, "loss": 0.0564, "step": 42710 }, { "grad_norm": 0.49278363585472107, "learning_rate": 6.589206092102213e-05, "loss": 0.0592, "step": 42720 }, { "grad_norm": 0.4751761853694916, "learning_rate": 6.587638277821375e-05, "loss": 0.0445, "step": 42730 }, { "grad_norm": 0.34946197271347046, "learning_rate": 6.586070289918803e-05, "loss": 0.0511, "step": 42740 }, { "grad_norm": 0.5009725689888, "learning_rate": 6.584502128565968e-05, "loss": 0.0554, "step": 42750 }, { "grad_norm": 0.4729049503803253, "learning_rate": 6.582933793934363e-05, "loss": 0.0557, "step": 42760 }, { "grad_norm": 0.5397858023643494, "learning_rate": 6.581365286195498e-05, "loss": 0.0444, "step": 42770 }, { "grad_norm": 0.5174250602722168, "learning_rate": 6.579796605520903e-05, "loss": 0.0502, "step": 42780 }, { "grad_norm": 0.6038115620613098, "learning_rate": 6.578227752082127e-05, "loss": 0.0617, "step": 42790 }, { "grad_norm": 0.4047485291957855, "learning_rate": 6.576658726050735e-05, "loss": 0.0442, "step": 42800 }, { "grad_norm": 0.7657809853553772, "learning_rate": 6.575089527598316e-05, "loss": 0.048, "step": 42810 }, { "grad_norm": 0.5021292567253113, "learning_rate": 6.573520156896476e-05, "loss": 0.0566, "step": 42820 }, { "grad_norm": 0.5435962080955505, "learning_rate": 6.571950614116835e-05, "loss": 0.0522, "step": 42830 }, { "grad_norm": 0.9752776026725769, "learning_rate": 6.570380899431039e-05, "loss": 0.0722, "step": 42840 }, { "grad_norm": 1.0413743257522583, "learning_rate": 6.568811013010749e-05, "loss": 0.0533, "step": 42850 }, { "grad_norm": 0.5219783186912537, "learning_rate": 6.567240955027645e-05, "loss": 0.0539, "step": 42860 }, { "grad_norm": 0.6568132042884827, "learning_rate": 6.565670725653427e-05, "loss": 0.0367, "step": 42870 }, { "grad_norm": 0.3305703103542328, "learning_rate": 6.564100325059812e-05, "loss": 0.0423, "step": 42880 }, { "grad_norm": 0.5472635626792908, "learning_rate": 6.562529753418538e-05, "loss": 0.0467, "step": 42890 }, { "grad_norm": 0.40038084983825684, "learning_rate": 6.56095901090136e-05, "loss": 0.0433, "step": 42900 }, { "grad_norm": 1.3222684860229492, "learning_rate": 6.55938809768005e-05, "loss": 0.0486, "step": 42910 }, { "grad_norm": 1.1877264976501465, "learning_rate": 6.557817013926402e-05, "loss": 0.0496, "step": 42920 }, { "grad_norm": 0.4647696018218994, "learning_rate": 6.556245759812227e-05, "loss": 0.0531, "step": 42930 }, { "grad_norm": 0.7935359477996826, "learning_rate": 6.554674335509357e-05, "loss": 0.0588, "step": 42940 }, { "grad_norm": 0.41774794459342957, "learning_rate": 6.553102741189638e-05, "loss": 0.0427, "step": 42950 }, { "grad_norm": 0.823005199432373, "learning_rate": 6.55153097702494e-05, "loss": 0.0554, "step": 42960 }, { "grad_norm": 0.5627454519271851, "learning_rate": 6.549959043187146e-05, "loss": 0.0557, "step": 42970 }, { "grad_norm": 0.860534131526947, "learning_rate": 6.54838693984816e-05, "loss": 0.0607, "step": 42980 }, { "grad_norm": 1.0200539827346802, "learning_rate": 6.546814667179909e-05, "loss": 0.0543, "step": 42990 }, { "grad_norm": 0.6279343366622925, "learning_rate": 6.545242225354328e-05, "loss": 0.057, "step": 43000 }, { "grad_norm": 0.5514112710952759, "learning_rate": 6.543669614543383e-05, "loss": 0.0596, "step": 43010 }, { "grad_norm": 0.6407552361488342, "learning_rate": 6.542096834919049e-05, "loss": 0.0582, "step": 43020 }, { "grad_norm": 1.0312353372573853, "learning_rate": 6.540523886653324e-05, "loss": 0.0522, "step": 43030 }, { "grad_norm": 2.132932662963867, "learning_rate": 6.538950769918222e-05, "loss": 0.0566, "step": 43040 }, { "grad_norm": 0.48794063925743103, "learning_rate": 6.537377484885779e-05, "loss": 0.0483, "step": 43050 }, { "grad_norm": 0.5459537506103516, "learning_rate": 6.535804031728045e-05, "loss": 0.0498, "step": 43060 }, { "grad_norm": 0.6344431638717651, "learning_rate": 6.534230410617092e-05, "loss": 0.052, "step": 43070 }, { "grad_norm": 0.8790818452835083, "learning_rate": 6.532656621725007e-05, "loss": 0.0558, "step": 43080 }, { "grad_norm": 0.5521367192268372, "learning_rate": 6.531082665223898e-05, "loss": 0.0633, "step": 43090 }, { "grad_norm": 0.3779416084289551, "learning_rate": 6.529508541285889e-05, "loss": 0.0518, "step": 43100 }, { "grad_norm": 0.9213665723800659, "learning_rate": 6.527934250083128e-05, "loss": 0.058, "step": 43110 }, { "grad_norm": 0.6749935150146484, "learning_rate": 6.526359791787775e-05, "loss": 0.0549, "step": 43120 }, { "grad_norm": 0.3601195216178894, "learning_rate": 6.524785166572009e-05, "loss": 0.0453, "step": 43130 }, { "grad_norm": 0.6409222483634949, "learning_rate": 6.523210374608029e-05, "loss": 0.0574, "step": 43140 }, { "grad_norm": 0.39405277371406555, "learning_rate": 6.521635416068054e-05, "loss": 0.0516, "step": 43150 }, { "grad_norm": 0.5737469792366028, "learning_rate": 6.520060291124317e-05, "loss": 0.0697, "step": 43160 }, { "grad_norm": 0.37725353240966797, "learning_rate": 6.518484999949072e-05, "loss": 0.0498, "step": 43170 }, { "grad_norm": 1.2287170886993408, "learning_rate": 6.516909542714591e-05, "loss": 0.0533, "step": 43180 }, { "grad_norm": 0.7436401844024658, "learning_rate": 6.515333919593161e-05, "loss": 0.0424, "step": 43190 }, { "grad_norm": 0.5835062861442566, "learning_rate": 6.513758130757094e-05, "loss": 0.0476, "step": 43200 }, { "grad_norm": 0.8377931714057922, "learning_rate": 6.512182176378713e-05, "loss": 0.0511, "step": 43210 }, { "grad_norm": 0.6789774298667908, "learning_rate": 6.510606056630362e-05, "loss": 0.0483, "step": 43220 }, { "grad_norm": 0.562764048576355, "learning_rate": 6.509029771684405e-05, "loss": 0.0482, "step": 43230 }, { "grad_norm": 0.6148091554641724, "learning_rate": 6.50745332171322e-05, "loss": 0.0591, "step": 43240 }, { "grad_norm": 0.441041499376297, "learning_rate": 6.505876706889207e-05, "loss": 0.0605, "step": 43250 }, { "grad_norm": 0.420606404542923, "learning_rate": 6.504299927384781e-05, "loss": 0.0481, "step": 43260 }, { "grad_norm": 1.7416967153549194, "learning_rate": 6.502722983372378e-05, "loss": 0.0527, "step": 43270 }, { "grad_norm": 0.7739654183387756, "learning_rate": 6.501145875024446e-05, "loss": 0.0562, "step": 43280 }, { "grad_norm": 0.5318948030471802, "learning_rate": 6.499568602513459e-05, "loss": 0.0517, "step": 43290 }, { "grad_norm": 0.4523927867412567, "learning_rate": 6.497991166011903e-05, "loss": 0.0525, "step": 43300 }, { "grad_norm": 0.49237045645713806, "learning_rate": 6.49641356569229e-05, "loss": 0.0495, "step": 43310 }, { "grad_norm": 0.5142504572868347, "learning_rate": 6.494835801727135e-05, "loss": 0.0539, "step": 43320 }, { "grad_norm": 0.5873032212257385, "learning_rate": 6.493257874288986e-05, "loss": 0.0638, "step": 43330 }, { "grad_norm": 0.3848322033882141, "learning_rate": 6.491679783550399e-05, "loss": 0.0546, "step": 43340 }, { "grad_norm": 0.5632919073104858, "learning_rate": 6.490101529683957e-05, "loss": 0.0474, "step": 43350 }, { "grad_norm": 0.28316038846969604, "learning_rate": 6.488523112862248e-05, "loss": 0.0491, "step": 43360 }, { "grad_norm": 0.4681561291217804, "learning_rate": 6.486944533257891e-05, "loss": 0.0524, "step": 43370 }, { "grad_norm": 0.574478268623352, "learning_rate": 6.485365791043517e-05, "loss": 0.0568, "step": 43380 }, { "grad_norm": 0.4460206627845764, "learning_rate": 6.483786886391772e-05, "loss": 0.0527, "step": 43390 }, { "grad_norm": 0.3891501724720001, "learning_rate": 6.482207819475323e-05, "loss": 0.0473, "step": 43400 }, { "grad_norm": 0.41951778531074524, "learning_rate": 6.480628590466857e-05, "loss": 0.0456, "step": 43410 }, { "grad_norm": 0.3521157205104828, "learning_rate": 6.479049199539074e-05, "loss": 0.0404, "step": 43420 }, { "grad_norm": 0.49539363384246826, "learning_rate": 6.477469646864693e-05, "loss": 0.0565, "step": 43430 }, { "grad_norm": 0.6400699019432068, "learning_rate": 6.475889932616454e-05, "loss": 0.0544, "step": 43440 }, { "grad_norm": 0.6025251150131226, "learning_rate": 6.474310056967111e-05, "loss": 0.0528, "step": 43450 }, { "grad_norm": 0.7483038902282715, "learning_rate": 6.472730020089437e-05, "loss": 0.0689, "step": 43460 }, { "grad_norm": 0.5705162286758423, "learning_rate": 6.471149822156222e-05, "loss": 0.0546, "step": 43470 }, { "grad_norm": 0.646806001663208, "learning_rate": 6.469569463340275e-05, "loss": 0.0455, "step": 43480 }, { "grad_norm": 0.4915297031402588, "learning_rate": 6.467988943814418e-05, "loss": 0.0545, "step": 43490 }, { "grad_norm": 0.49983590841293335, "learning_rate": 6.4664082637515e-05, "loss": 0.0466, "step": 43500 }, { "grad_norm": 0.5515999794006348, "learning_rate": 6.464827423324377e-05, "loss": 0.0511, "step": 43510 }, { "grad_norm": 0.4725261330604553, "learning_rate": 6.463246422705931e-05, "loss": 0.0475, "step": 43520 }, { "grad_norm": 0.6015138030052185, "learning_rate": 6.461665262069055e-05, "loss": 0.0615, "step": 43530 }, { "grad_norm": 0.5462106466293335, "learning_rate": 6.460083941586665e-05, "loss": 0.0495, "step": 43540 }, { "grad_norm": 0.5447742342948914, "learning_rate": 6.45850246143169e-05, "loss": 0.0537, "step": 43550 }, { "grad_norm": 0.6599940657615662, "learning_rate": 6.456920821777077e-05, "loss": 0.0531, "step": 43560 }, { "grad_norm": 0.5985670685768127, "learning_rate": 6.455339022795796e-05, "loss": 0.0643, "step": 43570 }, { "grad_norm": 1.142398715019226, "learning_rate": 6.453757064660825e-05, "loss": 0.047, "step": 43580 }, { "grad_norm": 0.43256673216819763, "learning_rate": 6.452174947545169e-05, "loss": 0.0361, "step": 43590 }, { "grad_norm": 0.4454262852668762, "learning_rate": 6.450592671621842e-05, "loss": 0.0433, "step": 43600 }, { "grad_norm": 0.8396552205085754, "learning_rate": 6.449010237063882e-05, "loss": 0.057, "step": 43610 }, { "grad_norm": 0.5560001730918884, "learning_rate": 6.447427644044343e-05, "loss": 0.0636, "step": 43620 }, { "grad_norm": 0.42694851756095886, "learning_rate": 6.44584489273629e-05, "loss": 0.0585, "step": 43630 }, { "grad_norm": 0.8823162913322449, "learning_rate": 6.444261983312817e-05, "loss": 0.0522, "step": 43640 }, { "grad_norm": 0.8301737904548645, "learning_rate": 6.442678915947023e-05, "loss": 0.061, "step": 43650 }, { "grad_norm": 0.6371743083000183, "learning_rate": 6.441095690812032e-05, "loss": 0.0487, "step": 43660 }, { "grad_norm": 0.4650670289993286, "learning_rate": 6.439512308080982e-05, "loss": 0.0522, "step": 43670 }, { "grad_norm": 0.36116844415664673, "learning_rate": 6.43792876792703e-05, "loss": 0.0502, "step": 43680 }, { "grad_norm": 0.5252641439437866, "learning_rate": 6.43634507052335e-05, "loss": 0.0497, "step": 43690 }, { "grad_norm": 0.6060448288917542, "learning_rate": 6.434761216043135e-05, "loss": 0.0549, "step": 43700 }, { "grad_norm": 0.5295920372009277, "learning_rate": 6.433177204659589e-05, "loss": 0.0496, "step": 43710 }, { "grad_norm": 0.5664638876914978, "learning_rate": 6.431593036545938e-05, "loss": 0.0482, "step": 43720 }, { "grad_norm": 0.6126886606216431, "learning_rate": 6.430008711875426e-05, "loss": 0.0396, "step": 43730 }, { "grad_norm": 0.7836576104164124, "learning_rate": 6.42842423082131e-05, "loss": 0.0578, "step": 43740 }, { "grad_norm": 0.8561025857925415, "learning_rate": 6.426839593556868e-05, "loss": 0.0451, "step": 43750 }, { "grad_norm": 0.336801141500473, "learning_rate": 6.425254800255392e-05, "loss": 0.0505, "step": 43760 }, { "grad_norm": 0.5496200919151306, "learning_rate": 6.423669851090196e-05, "loss": 0.05, "step": 43770 }, { "grad_norm": 0.5843899250030518, "learning_rate": 6.422084746234604e-05, "loss": 0.0438, "step": 43780 }, { "grad_norm": 0.6042582392692566, "learning_rate": 6.420499485861965e-05, "loss": 0.0386, "step": 43790 }, { "grad_norm": 0.5285853147506714, "learning_rate": 6.418914070145636e-05, "loss": 0.0584, "step": 43800 }, { "grad_norm": 0.570112943649292, "learning_rate": 6.417328499258996e-05, "loss": 0.0402, "step": 43810 }, { "grad_norm": 0.9312210083007812, "learning_rate": 6.415742773375446e-05, "loss": 0.046, "step": 43820 }, { "grad_norm": 0.3927834630012512, "learning_rate": 6.414156892668393e-05, "loss": 0.05, "step": 43830 }, { "grad_norm": 0.728809654712677, "learning_rate": 6.412570857311267e-05, "loss": 0.0537, "step": 43840 }, { "grad_norm": 0.46326255798339844, "learning_rate": 6.410984667477518e-05, "loss": 0.0527, "step": 43850 }, { "grad_norm": 0.6049686074256897, "learning_rate": 6.409398323340607e-05, "loss": 0.0458, "step": 43860 }, { "grad_norm": 0.6866533756256104, "learning_rate": 6.407811825074012e-05, "loss": 0.0573, "step": 43870 }, { "grad_norm": 0.47971975803375244, "learning_rate": 6.406225172851234e-05, "loss": 0.0483, "step": 43880 }, { "grad_norm": 0.4511978030204773, "learning_rate": 6.404638366845786e-05, "loss": 0.054, "step": 43890 }, { "grad_norm": 0.6922564506530762, "learning_rate": 6.403051407231196e-05, "loss": 0.0546, "step": 43900 }, { "grad_norm": 0.7204299569129944, "learning_rate": 6.401464294181016e-05, "loss": 0.0545, "step": 43910 }, { "grad_norm": 0.6440069079399109, "learning_rate": 6.399877027868808e-05, "loss": 0.0474, "step": 43920 }, { "grad_norm": 0.7170639634132385, "learning_rate": 6.39828960846815e-05, "loss": 0.0573, "step": 43930 }, { "grad_norm": 0.43604299426078796, "learning_rate": 6.396702036152644e-05, "loss": 0.0436, "step": 43940 }, { "grad_norm": 0.3674425780773163, "learning_rate": 6.395114311095905e-05, "loss": 0.0625, "step": 43950 }, { "grad_norm": 0.4189017713069916, "learning_rate": 6.393526433471559e-05, "loss": 0.0503, "step": 43960 }, { "grad_norm": 0.4544481933116913, "learning_rate": 6.391938403453262e-05, "loss": 0.066, "step": 43970 }, { "grad_norm": 0.5793918371200562, "learning_rate": 6.390350221214671e-05, "loss": 0.0563, "step": 43980 }, { "grad_norm": 1.0389615297317505, "learning_rate": 6.38876188692947e-05, "loss": 0.061, "step": 43990 }, { "grad_norm": 1.5043447017669678, "learning_rate": 6.387173400771358e-05, "loss": 0.0622, "step": 44000 }, { "grad_norm": 0.42072826623916626, "learning_rate": 6.385584762914047e-05, "loss": 0.0565, "step": 44010 }, { "grad_norm": 0.3281506597995758, "learning_rate": 6.38399597353127e-05, "loss": 0.0517, "step": 44020 }, { "grad_norm": 0.8028948307037354, "learning_rate": 6.382407032796775e-05, "loss": 0.0428, "step": 44030 }, { "grad_norm": 0.35023432970046997, "learning_rate": 6.380817940884325e-05, "loss": 0.0467, "step": 44040 }, { "grad_norm": 2.643874168395996, "learning_rate": 6.379228697967702e-05, "loss": 0.0537, "step": 44050 }, { "grad_norm": 0.501479983329773, "learning_rate": 6.3776393042207e-05, "loss": 0.0492, "step": 44060 }, { "grad_norm": 0.8146124482154846, "learning_rate": 6.376049759817137e-05, "loss": 0.0489, "step": 44070 }, { "grad_norm": 0.6032668948173523, "learning_rate": 6.37446006493084e-05, "loss": 0.066, "step": 44080 }, { "grad_norm": 0.43078821897506714, "learning_rate": 6.372870219735656e-05, "loss": 0.0609, "step": 44090 }, { "grad_norm": 0.6675812602043152, "learning_rate": 6.371280224405452e-05, "loss": 0.0575, "step": 44100 }, { "grad_norm": 0.539654016494751, "learning_rate": 6.369690079114101e-05, "loss": 0.046, "step": 44110 }, { "grad_norm": 0.6326085329055786, "learning_rate": 6.368099784035504e-05, "loss": 0.0506, "step": 44120 }, { "grad_norm": 0.7041642665863037, "learning_rate": 6.366509339343572e-05, "loss": 0.0471, "step": 44130 }, { "grad_norm": 0.6067276000976562, "learning_rate": 6.364918745212232e-05, "loss": 0.0459, "step": 44140 }, { "grad_norm": 0.48043742775917053, "learning_rate": 6.363328001815431e-05, "loss": 0.0491, "step": 44150 }, { "grad_norm": 0.733562707901001, "learning_rate": 6.361737109327128e-05, "loss": 0.0503, "step": 44160 }, { "grad_norm": 0.5322730541229248, "learning_rate": 6.360146067921301e-05, "loss": 0.0526, "step": 44170 }, { "grad_norm": 0.8082640767097473, "learning_rate": 6.35855487777195e-05, "loss": 0.0548, "step": 44180 }, { "grad_norm": 0.3805755376815796, "learning_rate": 6.356963539053076e-05, "loss": 0.0485, "step": 44190 }, { "grad_norm": 0.3920753598213196, "learning_rate": 6.355372051938711e-05, "loss": 0.0399, "step": 44200 }, { "grad_norm": 0.6346223950386047, "learning_rate": 6.353780416602894e-05, "loss": 0.0469, "step": 44210 }, { "grad_norm": 0.569652259349823, "learning_rate": 6.352188633219689e-05, "loss": 0.0462, "step": 44220 }, { "grad_norm": 0.4164990782737732, "learning_rate": 6.350596701963166e-05, "loss": 0.0385, "step": 44230 }, { "grad_norm": 0.5282664895057678, "learning_rate": 6.349004623007419e-05, "loss": 0.0465, "step": 44240 }, { "grad_norm": 0.4117696285247803, "learning_rate": 6.347412396526555e-05, "loss": 0.0501, "step": 44250 }, { "grad_norm": 0.7302660942077637, "learning_rate": 6.345820022694696e-05, "loss": 0.0623, "step": 44260 }, { "grad_norm": 1.173402190208435, "learning_rate": 6.344227501685984e-05, "loss": 0.0521, "step": 44270 }, { "grad_norm": 0.6828538179397583, "learning_rate": 6.342634833674572e-05, "loss": 0.0499, "step": 44280 }, { "grad_norm": 0.752922773361206, "learning_rate": 6.341042018834635e-05, "loss": 0.0621, "step": 44290 }, { "grad_norm": 0.44578230381011963, "learning_rate": 6.339449057340359e-05, "loss": 0.0525, "step": 44300 }, { "grad_norm": 0.389651358127594, "learning_rate": 6.337855949365945e-05, "loss": 0.0515, "step": 44310 }, { "grad_norm": 0.7859710454940796, "learning_rate": 6.336262695085619e-05, "loss": 0.0513, "step": 44320 }, { "grad_norm": 0.5844464302062988, "learning_rate": 6.334669294673612e-05, "loss": 0.0514, "step": 44330 }, { "grad_norm": 0.9593633413314819, "learning_rate": 6.333075748304179e-05, "loss": 0.0549, "step": 44340 }, { "grad_norm": 0.6959928274154663, "learning_rate": 6.331482056151585e-05, "loss": 0.0564, "step": 44350 }, { "grad_norm": 0.7872714996337891, "learning_rate": 6.329888218390117e-05, "loss": 0.0581, "step": 44360 }, { "grad_norm": 0.6526308059692383, "learning_rate": 6.328294235194072e-05, "loss": 0.0485, "step": 44370 }, { "grad_norm": 0.5357775688171387, "learning_rate": 6.326700106737765e-05, "loss": 0.0544, "step": 44380 }, { "grad_norm": 0.3946286737918854, "learning_rate": 6.325105833195531e-05, "loss": 0.0532, "step": 44390 }, { "grad_norm": 0.7624372839927673, "learning_rate": 6.323511414741715e-05, "loss": 0.0458, "step": 44400 }, { "grad_norm": 0.5822306275367737, "learning_rate": 6.321916851550678e-05, "loss": 0.0649, "step": 44410 }, { "grad_norm": 1.1697604656219482, "learning_rate": 6.320322143796806e-05, "loss": 0.0678, "step": 44420 }, { "grad_norm": 0.5599773526191711, "learning_rate": 6.318727291654488e-05, "loss": 0.0468, "step": 44430 }, { "grad_norm": 0.566753625869751, "learning_rate": 6.317132295298134e-05, "loss": 0.0515, "step": 44440 }, { "grad_norm": 0.5614622831344604, "learning_rate": 6.315537154902173e-05, "loss": 0.0494, "step": 44450 }, { "grad_norm": 0.9598678946495056, "learning_rate": 6.313941870641048e-05, "loss": 0.0487, "step": 44460 }, { "grad_norm": 0.38544896245002747, "learning_rate": 6.312346442689214e-05, "loss": 0.0538, "step": 44470 }, { "grad_norm": 0.6837027072906494, "learning_rate": 6.310750871221147e-05, "loss": 0.0418, "step": 44480 }, { "grad_norm": 0.5766477584838867, "learning_rate": 6.309155156411335e-05, "loss": 0.0462, "step": 44490 }, { "grad_norm": 1.4232847690582275, "learning_rate": 6.307559298434284e-05, "loss": 0.0481, "step": 44500 }, { "grad_norm": 0.31260478496551514, "learning_rate": 6.305963297464512e-05, "loss": 0.0545, "step": 44510 }, { "grad_norm": 0.7231863737106323, "learning_rate": 6.304367153676561e-05, "loss": 0.0507, "step": 44520 }, { "grad_norm": 0.6377111077308655, "learning_rate": 6.302770867244978e-05, "loss": 0.0522, "step": 44530 }, { "grad_norm": 0.6255235075950623, "learning_rate": 6.301174438344328e-05, "loss": 0.0477, "step": 44540 }, { "grad_norm": 0.5639519691467285, "learning_rate": 6.299577867149202e-05, "loss": 0.0595, "step": 44550 }, { "grad_norm": 0.34435757994651794, "learning_rate": 6.297981153834192e-05, "loss": 0.0524, "step": 44560 }, { "grad_norm": 1.0732336044311523, "learning_rate": 6.296384298573916e-05, "loss": 0.0553, "step": 44570 }, { "grad_norm": 0.5077340602874756, "learning_rate": 6.294787301543001e-05, "loss": 0.0742, "step": 44580 }, { "grad_norm": 0.7140036225318909, "learning_rate": 6.293190162916095e-05, "loss": 0.0504, "step": 44590 }, { "grad_norm": 0.8611962795257568, "learning_rate": 6.291592882867855e-05, "loss": 0.0502, "step": 44600 }, { "grad_norm": 0.53465336561203, "learning_rate": 6.28999546157296e-05, "loss": 0.044, "step": 44610 }, { "grad_norm": 0.4570600092411041, "learning_rate": 6.288397899206102e-05, "loss": 0.0484, "step": 44620 }, { "grad_norm": 0.7000143527984619, "learning_rate": 6.286800195941984e-05, "loss": 0.0567, "step": 44630 }, { "grad_norm": 0.477398157119751, "learning_rate": 6.285202351955334e-05, "loss": 0.0453, "step": 44640 }, { "grad_norm": 0.4812745153903961, "learning_rate": 6.283604367420887e-05, "loss": 0.0579, "step": 44650 }, { "grad_norm": 0.8656771183013916, "learning_rate": 6.282006242513394e-05, "loss": 0.0552, "step": 44660 }, { "grad_norm": 0.641389787197113, "learning_rate": 6.280407977407628e-05, "loss": 0.0558, "step": 44670 }, { "grad_norm": 0.5862332582473755, "learning_rate": 6.27880957227837e-05, "loss": 0.0497, "step": 44680 }, { "grad_norm": 0.5583957433700562, "learning_rate": 6.27721102730042e-05, "loss": 0.0525, "step": 44690 }, { "grad_norm": 0.8054150938987732, "learning_rate": 6.275612342648592e-05, "loss": 0.0564, "step": 44700 }, { "grad_norm": 0.5582514405250549, "learning_rate": 6.274013518497716e-05, "loss": 0.0546, "step": 44710 }, { "grad_norm": 0.4776754379272461, "learning_rate": 6.272414555022636e-05, "loss": 0.0424, "step": 44720 }, { "grad_norm": 0.5899377465248108, "learning_rate": 6.270815452398215e-05, "loss": 0.0479, "step": 44730 }, { "grad_norm": 0.5290055274963379, "learning_rate": 6.269216210799326e-05, "loss": 0.0406, "step": 44740 }, { "grad_norm": 0.5084717273712158, "learning_rate": 6.26761683040086e-05, "loss": 0.0544, "step": 44750 }, { "grad_norm": 0.5070968866348267, "learning_rate": 6.266017311377723e-05, "loss": 0.0483, "step": 44760 }, { "grad_norm": 0.40514299273490906, "learning_rate": 6.264417653904839e-05, "loss": 0.0608, "step": 44770 }, { "grad_norm": 0.4507980942726135, "learning_rate": 6.262817858157139e-05, "loss": 0.0529, "step": 44780 }, { "grad_norm": 0.6444348692893982, "learning_rate": 6.261217924309576e-05, "loss": 0.072, "step": 44790 }, { "grad_norm": 0.5334988832473755, "learning_rate": 6.259617852537118e-05, "loss": 0.0534, "step": 44800 }, { "grad_norm": 0.6235567331314087, "learning_rate": 6.258017643014747e-05, "loss": 0.0514, "step": 44810 }, { "grad_norm": 0.8667639493942261, "learning_rate": 6.256417295917456e-05, "loss": 0.0601, "step": 44820 }, { "grad_norm": 0.5012351870536804, "learning_rate": 6.254816811420258e-05, "loss": 0.0635, "step": 44830 }, { "grad_norm": 0.7470853328704834, "learning_rate": 6.253216189698183e-05, "loss": 0.0473, "step": 44840 }, { "grad_norm": 0.2557268440723419, "learning_rate": 6.251615430926267e-05, "loss": 0.0411, "step": 44850 }, { "grad_norm": 0.6084647178649902, "learning_rate": 6.25001453527957e-05, "loss": 0.0486, "step": 44860 }, { "grad_norm": 0.44235292077064514, "learning_rate": 6.248413502933164e-05, "loss": 0.0531, "step": 44870 }, { "grad_norm": 0.79328453540802, "learning_rate": 6.246812334062133e-05, "loss": 0.0483, "step": 44880 }, { "grad_norm": 0.6966454386711121, "learning_rate": 6.245211028841579e-05, "loss": 0.0561, "step": 44890 }, { "grad_norm": 0.4720425009727478, "learning_rate": 6.24360958744662e-05, "loss": 0.0477, "step": 44900 }, { "grad_norm": 0.5055557489395142, "learning_rate": 6.242008010052387e-05, "loss": 0.0595, "step": 44910 }, { "grad_norm": 0.5443801879882812, "learning_rate": 6.240406296834024e-05, "loss": 0.0513, "step": 44920 }, { "grad_norm": 0.47159671783447266, "learning_rate": 6.238804447966694e-05, "loss": 0.0474, "step": 44930 }, { "grad_norm": 0.48796993494033813, "learning_rate": 6.237202463625573e-05, "loss": 0.0376, "step": 44940 }, { "grad_norm": 0.37227654457092285, "learning_rate": 6.235600343985848e-05, "loss": 0.0467, "step": 44950 }, { "grad_norm": 0.7182309031486511, "learning_rate": 6.233998089222729e-05, "loss": 0.0418, "step": 44960 }, { "grad_norm": 0.39634406566619873, "learning_rate": 6.232395699511433e-05, "loss": 0.0551, "step": 44970 }, { "grad_norm": 0.6838513612747192, "learning_rate": 6.230793175027199e-05, "loss": 0.0493, "step": 44980 }, { "grad_norm": 0.5795310139656067, "learning_rate": 6.22919051594527e-05, "loss": 0.0471, "step": 44990 }, { "grad_norm": 0.6008448600769043, "learning_rate": 6.227587722440917e-05, "loss": 0.0548, "step": 45000 }, { "grad_norm": 0.3207743465900421, "learning_rate": 6.225984794689414e-05, "loss": 0.0333, "step": 45010 }, { "grad_norm": 1.335193157196045, "learning_rate": 6.22438173286606e-05, "loss": 0.0698, "step": 45020 }, { "grad_norm": 0.4737381041049957, "learning_rate": 6.222778537146159e-05, "loss": 0.0489, "step": 45030 }, { "grad_norm": 0.4195203483104706, "learning_rate": 6.221175207705037e-05, "loss": 0.0537, "step": 45040 }, { "grad_norm": 1.1631869077682495, "learning_rate": 6.219571744718027e-05, "loss": 0.0549, "step": 45050 }, { "grad_norm": 0.42863109707832336, "learning_rate": 6.217968148360487e-05, "loss": 0.0507, "step": 45060 }, { "grad_norm": 0.7119256854057312, "learning_rate": 6.216364418807782e-05, "loss": 0.0538, "step": 45070 }, { "grad_norm": 0.5284764170646667, "learning_rate": 6.214760556235292e-05, "loss": 0.0458, "step": 45080 }, { "grad_norm": 0.5426310896873474, "learning_rate": 6.213156560818414e-05, "loss": 0.0424, "step": 45090 }, { "grad_norm": 0.5038391947746277, "learning_rate": 6.211552432732559e-05, "loss": 0.0574, "step": 45100 }, { "grad_norm": 0.7084044218063354, "learning_rate": 6.20994817215315e-05, "loss": 0.0581, "step": 45110 }, { "grad_norm": 0.4377574920654297, "learning_rate": 6.208343779255628e-05, "loss": 0.0538, "step": 45120 }, { "grad_norm": 0.7718461155891418, "learning_rate": 6.206739254215449e-05, "loss": 0.0576, "step": 45130 }, { "grad_norm": 0.7370709180831909, "learning_rate": 6.205134597208077e-05, "loss": 0.0668, "step": 45140 }, { "grad_norm": 0.6964086294174194, "learning_rate": 6.203529808409e-05, "loss": 0.0492, "step": 45150 }, { "grad_norm": 0.5533784031867981, "learning_rate": 6.201924887993712e-05, "loss": 0.0564, "step": 45160 }, { "grad_norm": 0.6803911924362183, "learning_rate": 6.200319836137725e-05, "loss": 0.0509, "step": 45170 }, { "grad_norm": 0.470845490694046, "learning_rate": 6.198714653016565e-05, "loss": 0.043, "step": 45180 }, { "grad_norm": 0.4473813772201538, "learning_rate": 6.197109338805774e-05, "loss": 0.0504, "step": 45190 }, { "grad_norm": 0.6557478308677673, "learning_rate": 6.195503893680903e-05, "loss": 0.0531, "step": 45200 }, { "grad_norm": 0.44119781255722046, "learning_rate": 6.193898317817524e-05, "loss": 0.0515, "step": 45210 }, { "grad_norm": 0.6574156880378723, "learning_rate": 6.192292611391221e-05, "loss": 0.046, "step": 45220 }, { "grad_norm": 0.7543601989746094, "learning_rate": 6.190686774577591e-05, "loss": 0.052, "step": 45230 }, { "grad_norm": 0.5827010869979858, "learning_rate": 6.189080807552245e-05, "loss": 0.0628, "step": 45240 }, { "grad_norm": 0.6385674476623535, "learning_rate": 6.187474710490809e-05, "loss": 0.0489, "step": 45250 }, { "grad_norm": 0.3889425992965698, "learning_rate": 6.185868483568926e-05, "loss": 0.0465, "step": 45260 }, { "grad_norm": 0.5910928249359131, "learning_rate": 6.184262126962245e-05, "loss": 0.0561, "step": 45270 }, { "grad_norm": 0.5460061430931091, "learning_rate": 6.182655640846442e-05, "loss": 0.0342, "step": 45280 }, { "grad_norm": 0.4560983180999756, "learning_rate": 6.181049025397196e-05, "loss": 0.0488, "step": 45290 }, { "grad_norm": 0.47738561034202576, "learning_rate": 6.179442280790202e-05, "loss": 0.0545, "step": 45300 }, { "grad_norm": 0.5326017737388611, "learning_rate": 6.177835407201174e-05, "loss": 0.059, "step": 45310 }, { "grad_norm": 0.7711637616157532, "learning_rate": 6.176228404805839e-05, "loss": 0.0503, "step": 45320 }, { "grad_norm": 0.5331929326057434, "learning_rate": 6.174621273779932e-05, "loss": 0.0469, "step": 45330 }, { "grad_norm": 0.5564383864402771, "learning_rate": 6.17301401429921e-05, "loss": 0.046, "step": 45340 }, { "grad_norm": 0.40059950947761536, "learning_rate": 6.171406626539436e-05, "loss": 0.0538, "step": 45350 }, { "grad_norm": 0.8129579424858093, "learning_rate": 6.169799110676398e-05, "loss": 0.0668, "step": 45360 }, { "grad_norm": 0.5175986289978027, "learning_rate": 6.168191466885885e-05, "loss": 0.0488, "step": 45370 }, { "grad_norm": 0.7928370237350464, "learning_rate": 6.166583695343711e-05, "loss": 0.0491, "step": 45380 }, { "grad_norm": 0.38928425312042236, "learning_rate": 6.164975796225698e-05, "loss": 0.0604, "step": 45390 }, { "grad_norm": 0.4645492732524872, "learning_rate": 6.163367769707683e-05, "loss": 0.0432, "step": 45400 }, { "grad_norm": 0.6139646768569946, "learning_rate": 6.161759615965519e-05, "loss": 0.0559, "step": 45410 }, { "grad_norm": 0.5452105402946472, "learning_rate": 6.16015133517507e-05, "loss": 0.0526, "step": 45420 }, { "grad_norm": 0.43253105878829956, "learning_rate": 6.158542927512214e-05, "loss": 0.039, "step": 45430 }, { "grad_norm": 0.5815310478210449, "learning_rate": 6.156934393152846e-05, "loss": 0.0488, "step": 45440 }, { "grad_norm": 0.466679185628891, "learning_rate": 6.15532573227287e-05, "loss": 0.0581, "step": 45450 }, { "grad_norm": 0.632652223110199, "learning_rate": 6.153716945048212e-05, "loss": 0.0595, "step": 45460 }, { "grad_norm": 0.6808396577835083, "learning_rate": 6.152108031654802e-05, "loss": 0.0583, "step": 45470 }, { "grad_norm": 0.33529672026634216, "learning_rate": 6.15049899226859e-05, "loss": 0.0468, "step": 45480 }, { "grad_norm": 0.9698951244354248, "learning_rate": 6.148889827065537e-05, "loss": 0.0491, "step": 45490 }, { "grad_norm": 0.4942110776901245, "learning_rate": 6.147280536221622e-05, "loss": 0.0551, "step": 45500 }, { "grad_norm": 0.6688922047615051, "learning_rate": 6.145671119912832e-05, "loss": 0.0493, "step": 45510 }, { "grad_norm": 0.5377026200294495, "learning_rate": 6.144061578315169e-05, "loss": 0.0479, "step": 45520 }, { "grad_norm": 0.4800390303134918, "learning_rate": 6.142451911604654e-05, "loss": 0.0594, "step": 45530 }, { "grad_norm": 0.5443520545959473, "learning_rate": 6.140842119957315e-05, "loss": 0.0606, "step": 45540 }, { "grad_norm": 0.8292084336280823, "learning_rate": 6.139232203549197e-05, "loss": 0.0488, "step": 45550 }, { "grad_norm": 0.6051720976829529, "learning_rate": 6.137622162556357e-05, "loss": 0.0573, "step": 45560 }, { "grad_norm": 0.36977657675743103, "learning_rate": 6.13601199715487e-05, "loss": 0.0396, "step": 45570 }, { "grad_norm": 0.4357800781726837, "learning_rate": 6.134401707520816e-05, "loss": 0.0597, "step": 45580 }, { "grad_norm": 0.33326399326324463, "learning_rate": 6.132791293830298e-05, "loss": 0.0437, "step": 45590 }, { "grad_norm": 0.40714237093925476, "learning_rate": 6.131180756259428e-05, "loss": 0.0478, "step": 45600 }, { "grad_norm": 0.4569256603717804, "learning_rate": 6.129570094984331e-05, "loss": 0.0384, "step": 45610 }, { "grad_norm": 0.7721378803253174, "learning_rate": 6.127959310181145e-05, "loss": 0.0566, "step": 45620 }, { "grad_norm": 0.5149717926979065, "learning_rate": 6.126348402026026e-05, "loss": 0.0479, "step": 45630 }, { "grad_norm": 0.7116760611534119, "learning_rate": 6.12473737069514e-05, "loss": 0.0481, "step": 45640 }, { "grad_norm": 0.6352697610855103, "learning_rate": 6.123126216364665e-05, "loss": 0.0495, "step": 45650 }, { "grad_norm": 0.690462589263916, "learning_rate": 6.121514939210797e-05, "loss": 0.0387, "step": 45660 }, { "grad_norm": 0.4659680426120758, "learning_rate": 6.119903539409741e-05, "loss": 0.0536, "step": 45670 }, { "grad_norm": 0.4766596853733063, "learning_rate": 6.118292017137716e-05, "loss": 0.043, "step": 45680 }, { "grad_norm": 0.47174331545829773, "learning_rate": 6.116680372570959e-05, "loss": 0.0593, "step": 45690 }, { "grad_norm": 0.5676721334457397, "learning_rate": 6.115068605885713e-05, "loss": 0.0345, "step": 45700 }, { "grad_norm": 0.511194109916687, "learning_rate": 6.113456717258243e-05, "loss": 0.0592, "step": 45710 }, { "grad_norm": 0.40013426542282104, "learning_rate": 6.11184470686482e-05, "loss": 0.0482, "step": 45720 }, { "grad_norm": 0.5910576581954956, "learning_rate": 6.110232574881731e-05, "loss": 0.0479, "step": 45730 }, { "grad_norm": 0.8486358523368835, "learning_rate": 6.108620321485277e-05, "loss": 0.0644, "step": 45740 }, { "grad_norm": 0.4780968427658081, "learning_rate": 6.107007946851773e-05, "loss": 0.0497, "step": 45750 }, { "grad_norm": 0.617534339427948, "learning_rate": 6.105395451157542e-05, "loss": 0.0519, "step": 45760 }, { "grad_norm": 0.3793008327484131, "learning_rate": 6.103782834578928e-05, "loss": 0.0425, "step": 45770 }, { "grad_norm": 0.49471956491470337, "learning_rate": 6.102170097292281e-05, "loss": 0.051, "step": 45780 }, { "grad_norm": 0.8581316471099854, "learning_rate": 6.1005572394739716e-05, "loss": 0.0589, "step": 45790 }, { "grad_norm": 0.7369313836097717, "learning_rate": 6.0989442613003765e-05, "loss": 0.0386, "step": 45800 }, { "grad_norm": 0.44512927532196045, "learning_rate": 6.097331162947888e-05, "loss": 0.0467, "step": 45810 }, { "grad_norm": 0.5797634124755859, "learning_rate": 6.095717944592914e-05, "loss": 0.0578, "step": 45820 }, { "grad_norm": 0.8493303060531616, "learning_rate": 6.094104606411873e-05, "loss": 0.0559, "step": 45830 }, { "grad_norm": 0.6271822452545166, "learning_rate": 6.0924911485811966e-05, "loss": 0.0586, "step": 45840 }, { "grad_norm": 0.7170917391777039, "learning_rate": 6.090877571277331e-05, "loss": 0.0595, "step": 45850 }, { "grad_norm": 0.39402076601982117, "learning_rate": 6.089263874676736e-05, "loss": 0.047, "step": 45860 }, { "grad_norm": 0.7048584818840027, "learning_rate": 6.0876500589558796e-05, "loss": 0.0542, "step": 45870 }, { "grad_norm": 0.8457115292549133, "learning_rate": 6.086036124291248e-05, "loss": 0.0566, "step": 45880 }, { "grad_norm": 0.5617847442626953, "learning_rate": 6.084422070859339e-05, "loss": 0.0498, "step": 45890 }, { "grad_norm": 0.8140738606452942, "learning_rate": 6.082807898836663e-05, "loss": 0.0512, "step": 45900 }, { "grad_norm": 0.52864009141922, "learning_rate": 6.081193608399742e-05, "loss": 0.044, "step": 45910 }, { "grad_norm": 0.44458943605422974, "learning_rate": 6.0795791997251164e-05, "loss": 0.0418, "step": 45920 }, { "grad_norm": 0.6251254677772522, "learning_rate": 6.0779646729893294e-05, "loss": 0.0429, "step": 45930 }, { "grad_norm": 0.43162474036216736, "learning_rate": 6.0763500283689476e-05, "loss": 0.0376, "step": 45940 }, { "grad_norm": 0.523688554763794, "learning_rate": 6.0747352660405455e-05, "loss": 0.0515, "step": 45950 }, { "grad_norm": 0.5625415444374084, "learning_rate": 6.073120386180709e-05, "loss": 0.0527, "step": 45960 }, { "grad_norm": 0.4552106559276581, "learning_rate": 6.0715053889660425e-05, "loss": 0.0572, "step": 45970 }, { "grad_norm": 0.6087052822113037, "learning_rate": 6.069890274573157e-05, "loss": 0.0422, "step": 45980 }, { "grad_norm": 0.5673790574073792, "learning_rate": 6.068275043178679e-05, "loss": 0.0618, "step": 45990 }, { "grad_norm": 0.9235381484031677, "learning_rate": 6.066659694959248e-05, "loss": 0.0471, "step": 46000 }, { "grad_norm": 0.3006986379623413, "learning_rate": 6.0650442300915176e-05, "loss": 0.0468, "step": 46010 }, { "grad_norm": 0.44786718487739563, "learning_rate": 6.063428648752152e-05, "loss": 0.0616, "step": 46020 }, { "grad_norm": 0.5301766395568848, "learning_rate": 6.0618129511178266e-05, "loss": 0.0419, "step": 46030 }, { "grad_norm": 0.33807268738746643, "learning_rate": 6.060197137365234e-05, "loss": 0.06, "step": 46040 }, { "grad_norm": 0.6161512732505798, "learning_rate": 6.058581207671077e-05, "loss": 0.0424, "step": 46050 }, { "grad_norm": 0.6484296917915344, "learning_rate": 6.056965162212072e-05, "loss": 0.0602, "step": 46060 }, { "grad_norm": 0.38856300711631775, "learning_rate": 6.055349001164943e-05, "loss": 0.0488, "step": 46070 }, { "grad_norm": 0.31932154297828674, "learning_rate": 6.0537327247064347e-05, "loss": 0.0529, "step": 46080 }, { "grad_norm": 0.6340541243553162, "learning_rate": 6.0521163330133014e-05, "loss": 0.0606, "step": 46090 }, { "grad_norm": 0.39215540885925293, "learning_rate": 6.050499826262306e-05, "loss": 0.0477, "step": 46100 }, { "grad_norm": 0.7552198171615601, "learning_rate": 6.0488832046302294e-05, "loss": 0.0552, "step": 46110 }, { "grad_norm": 0.4206775426864624, "learning_rate": 6.0472664682938626e-05, "loss": 0.0482, "step": 46120 }, { "grad_norm": 0.49948930740356445, "learning_rate": 6.045649617430009e-05, "loss": 0.0483, "step": 46130 }, { "grad_norm": 0.5999391078948975, "learning_rate": 6.0440326522154866e-05, "loss": 0.0525, "step": 46140 }, { "grad_norm": 0.4735353887081146, "learning_rate": 6.0424155728271224e-05, "loss": 0.0498, "step": 46150 }, { "grad_norm": 0.3837420344352722, "learning_rate": 6.040798379441758e-05, "loss": 0.0557, "step": 46160 }, { "grad_norm": 0.5188344717025757, "learning_rate": 6.0391810722362485e-05, "loss": 0.0517, "step": 46170 }, { "grad_norm": 0.6456087827682495, "learning_rate": 6.037563651387458e-05, "loss": 0.0446, "step": 46180 }, { "grad_norm": 0.521536111831665, "learning_rate": 6.0359461170722666e-05, "loss": 0.0441, "step": 46190 }, { "grad_norm": 0.7108014225959778, "learning_rate": 6.034328469467566e-05, "loss": 0.0457, "step": 46200 }, { "grad_norm": 0.6589436531066895, "learning_rate": 6.0327107087502596e-05, "loss": 0.0477, "step": 46210 }, { "grad_norm": 0.6163837313652039, "learning_rate": 6.031092835097262e-05, "loss": 0.049, "step": 46220 }, { "grad_norm": 0.4047697186470032, "learning_rate": 6.0294748486855024e-05, "loss": 0.0454, "step": 46230 }, { "grad_norm": 0.40835970640182495, "learning_rate": 6.0278567496919216e-05, "loss": 0.0559, "step": 46240 }, { "grad_norm": 0.339262455701828, "learning_rate": 6.026238538293472e-05, "loss": 0.0478, "step": 46250 }, { "grad_norm": 0.4205763339996338, "learning_rate": 6.024620214667118e-05, "loss": 0.0487, "step": 46260 }, { "grad_norm": 0.5236865878105164, "learning_rate": 6.0230017789898384e-05, "loss": 0.0447, "step": 46270 }, { "grad_norm": 0.38341304659843445, "learning_rate": 6.021383231438622e-05, "loss": 0.0456, "step": 46280 }, { "grad_norm": 0.40747711062431335, "learning_rate": 6.0197645721904704e-05, "loss": 0.0448, "step": 46290 }, { "grad_norm": 0.6274895668029785, "learning_rate": 6.0181458014224e-05, "loss": 0.0481, "step": 46300 }, { "grad_norm": 0.7816572189331055, "learning_rate": 6.016526919311435e-05, "loss": 0.0515, "step": 46310 }, { "grad_norm": 0.34196460247039795, "learning_rate": 6.014907926034613e-05, "loss": 0.0425, "step": 46320 }, { "grad_norm": 0.6460970640182495, "learning_rate": 6.0132888217689875e-05, "loss": 0.0498, "step": 46330 }, { "grad_norm": 0.4905092120170593, "learning_rate": 6.01166960669162e-05, "loss": 0.051, "step": 46340 }, { "grad_norm": 1.9214118719100952, "learning_rate": 6.0100502809795845e-05, "loss": 0.0527, "step": 46350 }, { "grad_norm": 0.3639918863773346, "learning_rate": 6.00843084480997e-05, "loss": 0.0508, "step": 46360 }, { "grad_norm": 0.4537084996700287, "learning_rate": 6.0068112983598736e-05, "loss": 0.0453, "step": 46370 }, { "grad_norm": 0.4926251471042633, "learning_rate": 6.005191641806407e-05, "loss": 0.044, "step": 46380 }, { "grad_norm": 0.6177276968955994, "learning_rate": 6.003571875326694e-05, "loss": 0.0462, "step": 46390 }, { "grad_norm": 0.5298627614974976, "learning_rate": 6.00195199909787e-05, "loss": 0.0482, "step": 46400 }, { "grad_norm": 0.5759105086326599, "learning_rate": 6.000332013297082e-05, "loss": 0.0463, "step": 46410 }, { "grad_norm": 0.49437958002090454, "learning_rate": 5.998711918101487e-05, "loss": 0.0466, "step": 46420 }, { "grad_norm": 0.4043077230453491, "learning_rate": 5.997091713688261e-05, "loss": 0.0503, "step": 46430 }, { "grad_norm": 0.33779239654541016, "learning_rate": 5.9954714002345836e-05, "loss": 0.0468, "step": 46440 }, { "grad_norm": 0.5626050233840942, "learning_rate": 5.993850977917649e-05, "loss": 0.0496, "step": 46450 }, { "grad_norm": 0.5039499998092651, "learning_rate": 5.992230446914667e-05, "loss": 0.0437, "step": 46460 }, { "grad_norm": 0.3919086754322052, "learning_rate": 5.990609807402855e-05, "loss": 0.0506, "step": 46470 }, { "grad_norm": 0.8143527507781982, "learning_rate": 5.988989059559443e-05, "loss": 0.0547, "step": 46480 }, { "grad_norm": 0.5843640565872192, "learning_rate": 5.987368203561675e-05, "loss": 0.0481, "step": 46490 }, { "grad_norm": 0.5019986629486084, "learning_rate": 5.9857472395868055e-05, "loss": 0.0566, "step": 46500 }, { "grad_norm": 0.5841085314750671, "learning_rate": 5.9841261678120983e-05, "loss": 0.0522, "step": 46510 }, { "grad_norm": 0.5312080383300781, "learning_rate": 5.982504988414834e-05, "loss": 0.0374, "step": 46520 }, { "grad_norm": 0.48634999990463257, "learning_rate": 5.9808837015723015e-05, "loss": 0.0649, "step": 46530 }, { "grad_norm": 0.412195086479187, "learning_rate": 5.9792623074618016e-05, "loss": 0.045, "step": 46540 }, { "grad_norm": 0.6836538314819336, "learning_rate": 5.977640806260648e-05, "loss": 0.0492, "step": 46550 }, { "grad_norm": 0.43875637650489807, "learning_rate": 5.976019198146166e-05, "loss": 0.0419, "step": 46560 }, { "grad_norm": 0.6260179281234741, "learning_rate": 5.974397483295692e-05, "loss": 0.0436, "step": 46570 }, { "grad_norm": 0.5314152240753174, "learning_rate": 5.972775661886575e-05, "loss": 0.0418, "step": 46580 }, { "grad_norm": 0.5142025947570801, "learning_rate": 5.971153734096173e-05, "loss": 0.0501, "step": 46590 }, { "grad_norm": 0.5354870557785034, "learning_rate": 5.969531700101859e-05, "loss": 0.0396, "step": 46600 }, { "grad_norm": 0.613262951374054, "learning_rate": 5.9679095600810155e-05, "loss": 0.0535, "step": 46610 }, { "grad_norm": 0.5121066570281982, "learning_rate": 5.9662873142110384e-05, "loss": 0.0426, "step": 46620 }, { "grad_norm": 0.754867672920227, "learning_rate": 5.964664962669333e-05, "loss": 0.0458, "step": 46630 }, { "grad_norm": 0.5073996186256409, "learning_rate": 5.9630425056333186e-05, "loss": 0.0514, "step": 46640 }, { "grad_norm": 0.3649841547012329, "learning_rate": 5.961419943280422e-05, "loss": 0.0625, "step": 46650 }, { "grad_norm": 0.5043334364891052, "learning_rate": 5.959797275788087e-05, "loss": 0.0414, "step": 46660 }, { "grad_norm": 0.7454565763473511, "learning_rate": 5.958174503333765e-05, "loss": 0.0586, "step": 46670 }, { "grad_norm": 0.4220435917377472, "learning_rate": 5.9565516260949195e-05, "loss": 0.0569, "step": 46680 }, { "grad_norm": 0.9195899963378906, "learning_rate": 5.9549286442490273e-05, "loss": 0.0499, "step": 46690 }, { "grad_norm": 0.5129812359809875, "learning_rate": 5.953305557973572e-05, "loss": 0.0476, "step": 46700 }, { "grad_norm": 0.7647511959075928, "learning_rate": 5.9516823674460564e-05, "loss": 0.0462, "step": 46710 }, { "grad_norm": 0.5771504640579224, "learning_rate": 5.9500590728439875e-05, "loss": 0.0493, "step": 46720 }, { "grad_norm": 0.4457084536552429, "learning_rate": 5.948435674344886e-05, "loss": 0.0532, "step": 46730 }, { "grad_norm": 0.8086717128753662, "learning_rate": 5.946812172126285e-05, "loss": 0.0525, "step": 46740 }, { "grad_norm": 0.40929892659187317, "learning_rate": 5.945188566365729e-05, "loss": 0.043, "step": 46750 }, { "grad_norm": 0.4696193039417267, "learning_rate": 5.943564857240773e-05, "loss": 0.0481, "step": 46760 }, { "grad_norm": 1.6939828395843506, "learning_rate": 5.941941044928981e-05, "loss": 0.0545, "step": 46770 }, { "grad_norm": 0.508932888507843, "learning_rate": 5.940317129607935e-05, "loss": 0.046, "step": 46780 }, { "grad_norm": 0.4717613458633423, "learning_rate": 5.9386931114552204e-05, "loss": 0.045, "step": 46790 }, { "grad_norm": 0.7411723732948303, "learning_rate": 5.937068990648438e-05, "loss": 0.0472, "step": 46800 }, { "grad_norm": 0.5764237642288208, "learning_rate": 5.935444767365199e-05, "loss": 0.0701, "step": 46810 }, { "grad_norm": 0.45124346017837524, "learning_rate": 5.933820441783129e-05, "loss": 0.056, "step": 46820 }, { "grad_norm": 0.6477968692779541, "learning_rate": 5.932196014079857e-05, "loss": 0.0505, "step": 46830 }, { "grad_norm": 0.39917632937431335, "learning_rate": 5.930571484433032e-05, "loss": 0.0364, "step": 46840 }, { "grad_norm": 0.4650338590145111, "learning_rate": 5.928946853020309e-05, "loss": 0.0473, "step": 46850 }, { "grad_norm": 0.400249183177948, "learning_rate": 5.927322120019352e-05, "loss": 0.0459, "step": 46860 }, { "grad_norm": 1.2371467351913452, "learning_rate": 5.925697285607844e-05, "loss": 0.0639, "step": 46870 }, { "grad_norm": 1.1920363903045654, "learning_rate": 5.924072349963472e-05, "loss": 0.0572, "step": 46880 }, { "grad_norm": 0.48342350125312805, "learning_rate": 5.922447313263938e-05, "loss": 0.0431, "step": 46890 }, { "grad_norm": 0.38047856092453003, "learning_rate": 5.920822175686952e-05, "loss": 0.0525, "step": 46900 }, { "grad_norm": 0.517864465713501, "learning_rate": 5.919196937410239e-05, "loss": 0.0514, "step": 46910 }, { "grad_norm": 0.5173258781433105, "learning_rate": 5.917571598611529e-05, "loss": 0.046, "step": 46920 }, { "grad_norm": 0.659260094165802, "learning_rate": 5.9159461594685686e-05, "loss": 0.0501, "step": 46930 }, { "grad_norm": 0.5167750120162964, "learning_rate": 5.914320620159114e-05, "loss": 0.0477, "step": 46940 }, { "grad_norm": 0.5069207549095154, "learning_rate": 5.912694980860931e-05, "loss": 0.0527, "step": 46950 }, { "grad_norm": 0.5727458596229553, "learning_rate": 5.9110692417517964e-05, "loss": 0.0489, "step": 46960 }, { "grad_norm": 0.5649262070655823, "learning_rate": 5.9094434030095e-05, "loss": 0.0608, "step": 46970 }, { "grad_norm": 0.4222100079059601, "learning_rate": 5.907817464811841e-05, "loss": 0.0415, "step": 46980 }, { "grad_norm": 0.5202469229698181, "learning_rate": 5.906191427336627e-05, "loss": 0.0489, "step": 46990 }, { "grad_norm": 0.7549607157707214, "learning_rate": 5.9045652907616835e-05, "loss": 0.0488, "step": 47000 }, { "grad_norm": 0.589414119720459, "learning_rate": 5.902939055264838e-05, "loss": 0.0474, "step": 47010 }, { "grad_norm": 0.6693695187568665, "learning_rate": 5.901312721023935e-05, "loss": 0.0507, "step": 47020 }, { "grad_norm": 0.3759493827819824, "learning_rate": 5.8996862882168294e-05, "loss": 0.0403, "step": 47030 }, { "grad_norm": 0.42782431840896606, "learning_rate": 5.8980597570213826e-05, "loss": 0.0579, "step": 47040 }, { "grad_norm": 1.0822288990020752, "learning_rate": 5.896433127615471e-05, "loss": 0.0575, "step": 47050 }, { "grad_norm": 0.6104793548583984, "learning_rate": 5.894806400176981e-05, "loss": 0.0369, "step": 47060 }, { "grad_norm": 0.6161625385284424, "learning_rate": 5.893179574883808e-05, "loss": 0.0482, "step": 47070 }, { "grad_norm": 0.41539397835731506, "learning_rate": 5.8915526519138585e-05, "loss": 0.0547, "step": 47080 }, { "grad_norm": 0.6151688694953918, "learning_rate": 5.889925631445053e-05, "loss": 0.0589, "step": 47090 }, { "grad_norm": 0.39310726523399353, "learning_rate": 5.888298513655318e-05, "loss": 0.0362, "step": 47100 }, { "grad_norm": 0.391282856464386, "learning_rate": 5.886671298722592e-05, "loss": 0.0448, "step": 47110 }, { "grad_norm": 0.4467553496360779, "learning_rate": 5.885043986824828e-05, "loss": 0.0452, "step": 47120 }, { "grad_norm": 0.8035224080085754, "learning_rate": 5.883416578139982e-05, "loss": 0.0542, "step": 47130 }, { "grad_norm": 0.456717312335968, "learning_rate": 5.881789072846029e-05, "loss": 0.0498, "step": 47140 }, { "grad_norm": 0.5999552011489868, "learning_rate": 5.8801614711209474e-05, "loss": 0.0418, "step": 47150 }, { "grad_norm": 0.48175543546676636, "learning_rate": 5.878533773142732e-05, "loss": 0.0451, "step": 47160 }, { "grad_norm": 0.47871291637420654, "learning_rate": 5.8769059790893834e-05, "loss": 0.0514, "step": 47170 }, { "grad_norm": 0.7704599499702454, "learning_rate": 5.8752780891389145e-05, "loss": 0.0662, "step": 47180 }, { "grad_norm": 0.7056052684783936, "learning_rate": 5.873650103469351e-05, "loss": 0.0465, "step": 47190 }, { "grad_norm": 0.38537198305130005, "learning_rate": 5.872022022258725e-05, "loss": 0.0359, "step": 47200 }, { "grad_norm": 0.329194039106369, "learning_rate": 5.8703938456850814e-05, "loss": 0.0371, "step": 47210 }, { "grad_norm": 0.5394628047943115, "learning_rate": 5.868765573926477e-05, "loss": 0.055, "step": 47220 }, { "grad_norm": 0.48304620385169983, "learning_rate": 5.867137207160974e-05, "loss": 0.0652, "step": 47230 }, { "grad_norm": 0.5269725322723389, "learning_rate": 5.86550874556665e-05, "loss": 0.0502, "step": 47240 }, { "grad_norm": 1.6226459741592407, "learning_rate": 5.863880189321592e-05, "loss": 0.0456, "step": 47250 }, { "grad_norm": 0.632146954536438, "learning_rate": 5.862251538603894e-05, "loss": 0.0408, "step": 47260 }, { "grad_norm": 0.38393542170524597, "learning_rate": 5.8606227935916656e-05, "loss": 0.0436, "step": 47270 }, { "grad_norm": 0.4360278844833374, "learning_rate": 5.858993954463021e-05, "loss": 0.0492, "step": 47280 }, { "grad_norm": 0.4179830551147461, "learning_rate": 5.8573650213960896e-05, "loss": 0.0492, "step": 47290 }, { "grad_norm": 0.783272385597229, "learning_rate": 5.8557359945690084e-05, "loss": 0.0493, "step": 47300 }, { "grad_norm": 0.5286440253257751, "learning_rate": 5.8541068741599246e-05, "loss": 0.0564, "step": 47310 }, { "grad_norm": 0.6180658340454102, "learning_rate": 5.8524776603469985e-05, "loss": 0.043, "step": 47320 }, { "grad_norm": 0.4915822446346283, "learning_rate": 5.850848353308397e-05, "loss": 0.0438, "step": 47330 }, { "grad_norm": 1.0153392553329468, "learning_rate": 5.849218953222297e-05, "loss": 0.0585, "step": 47340 }, { "grad_norm": 0.4504231810569763, "learning_rate": 5.847589460266891e-05, "loss": 0.0449, "step": 47350 }, { "grad_norm": 1.3249738216400146, "learning_rate": 5.845959874620375e-05, "loss": 0.0502, "step": 47360 }, { "grad_norm": 0.44791221618652344, "learning_rate": 5.84433019646096e-05, "loss": 0.0534, "step": 47370 }, { "grad_norm": 0.9890669584274292, "learning_rate": 5.842700425966863e-05, "loss": 0.0525, "step": 47380 }, { "grad_norm": 0.42916253209114075, "learning_rate": 5.841070563316315e-05, "loss": 0.0495, "step": 47390 }, { "grad_norm": 1.0880444049835205, "learning_rate": 5.839440608687554e-05, "loss": 0.0615, "step": 47400 }, { "grad_norm": 0.6529765129089355, "learning_rate": 5.837810562258831e-05, "loss": 0.0456, "step": 47410 }, { "grad_norm": 0.6422637701034546, "learning_rate": 5.836180424208405e-05, "loss": 0.0685, "step": 47420 }, { "grad_norm": 0.4834481179714203, "learning_rate": 5.834550194714543e-05, "loss": 0.0511, "step": 47430 }, { "grad_norm": 0.44466090202331543, "learning_rate": 5.832919873955528e-05, "loss": 0.0672, "step": 47440 }, { "grad_norm": 0.7740647792816162, "learning_rate": 5.831289462109648e-05, "loss": 0.0617, "step": 47450 }, { "grad_norm": 0.26869693398475647, "learning_rate": 5.829658959355202e-05, "loss": 0.0486, "step": 47460 }, { "grad_norm": 0.47137612104415894, "learning_rate": 5.8280283658704995e-05, "loss": 0.059, "step": 47470 }, { "grad_norm": 0.5092771053314209, "learning_rate": 5.826397681833859e-05, "loss": 0.0572, "step": 47480 }, { "grad_norm": 0.36123642325401306, "learning_rate": 5.824766907423612e-05, "loss": 0.0475, "step": 47490 }, { "grad_norm": 0.5201128125190735, "learning_rate": 5.8231360428180945e-05, "loss": 0.0537, "step": 47500 }, { "grad_norm": 0.45334601402282715, "learning_rate": 5.821505088195658e-05, "loss": 0.0529, "step": 47510 }, { "grad_norm": 0.5043615698814392, "learning_rate": 5.819874043734661e-05, "loss": 0.0579, "step": 47520 }, { "grad_norm": 0.6223980188369751, "learning_rate": 5.8182429096134695e-05, "loss": 0.0464, "step": 47530 }, { "grad_norm": 0.6779683232307434, "learning_rate": 5.816611686010465e-05, "loss": 0.0466, "step": 47540 }, { "grad_norm": 0.41114047169685364, "learning_rate": 5.814980373104033e-05, "loss": 0.0533, "step": 47550 }, { "grad_norm": 0.4901444613933563, "learning_rate": 5.813348971072572e-05, "loss": 0.062, "step": 47560 }, { "grad_norm": 0.3608442544937134, "learning_rate": 5.811717480094492e-05, "loss": 0.0468, "step": 47570 }, { "grad_norm": 0.4612104594707489, "learning_rate": 5.810085900348209e-05, "loss": 0.0479, "step": 47580 }, { "grad_norm": 1.208434820175171, "learning_rate": 5.8084542320121483e-05, "loss": 0.0525, "step": 47590 }, { "grad_norm": 0.6521031260490417, "learning_rate": 5.8068224752647497e-05, "loss": 0.0646, "step": 47600 }, { "grad_norm": 0.5056250095367432, "learning_rate": 5.805190630284457e-05, "loss": 0.0385, "step": 47610 }, { "grad_norm": 0.5495983958244324, "learning_rate": 5.803558697249729e-05, "loss": 0.0473, "step": 47620 }, { "grad_norm": 0.4426336884498596, "learning_rate": 5.8019266763390276e-05, "loss": 0.0625, "step": 47630 }, { "grad_norm": 0.6191259622573853, "learning_rate": 5.800294567730833e-05, "loss": 0.0405, "step": 47640 }, { "grad_norm": 0.4934898912906647, "learning_rate": 5.798662371603628e-05, "loss": 0.0497, "step": 47650 }, { "grad_norm": 0.4767102301120758, "learning_rate": 5.797030088135904e-05, "loss": 0.0554, "step": 47660 }, { "grad_norm": 0.3921271562576294, "learning_rate": 5.7953977175061705e-05, "loss": 0.0424, "step": 47670 }, { "grad_norm": 0.8961835503578186, "learning_rate": 5.793765259892938e-05, "loss": 0.058, "step": 47680 }, { "grad_norm": 0.31836220622062683, "learning_rate": 5.792132715474729e-05, "loss": 0.041, "step": 47690 }, { "grad_norm": 0.6527819037437439, "learning_rate": 5.790500084430078e-05, "loss": 0.0424, "step": 47700 }, { "grad_norm": 0.36555415391921997, "learning_rate": 5.7888673669375274e-05, "loss": 0.0423, "step": 47710 }, { "grad_norm": 0.3516921401023865, "learning_rate": 5.787234563175625e-05, "loss": 0.0439, "step": 47720 }, { "grad_norm": 0.7432467341423035, "learning_rate": 5.7856016733229355e-05, "loss": 0.0563, "step": 47730 }, { "grad_norm": 0.5744590759277344, "learning_rate": 5.7839686975580297e-05, "loss": 0.0596, "step": 47740 }, { "grad_norm": 0.4306148886680603, "learning_rate": 5.782335636059484e-05, "loss": 0.0442, "step": 47750 }, { "grad_norm": 0.49110084772109985, "learning_rate": 5.780702489005889e-05, "loss": 0.0467, "step": 47760 }, { "grad_norm": 0.383148193359375, "learning_rate": 5.779069256575845e-05, "loss": 0.05, "step": 47770 }, { "grad_norm": 0.45610475540161133, "learning_rate": 5.7774359389479574e-05, "loss": 0.0372, "step": 47780 }, { "grad_norm": 0.7402877807617188, "learning_rate": 5.775802536300845e-05, "loss": 0.0576, "step": 47790 }, { "grad_norm": 0.4570108950138092, "learning_rate": 5.774169048813134e-05, "loss": 0.0489, "step": 47800 }, { "grad_norm": 0.5407254099845886, "learning_rate": 5.77253547666346e-05, "loss": 0.0468, "step": 47810 }, { "grad_norm": 0.8566112518310547, "learning_rate": 5.770901820030465e-05, "loss": 0.0555, "step": 47820 }, { "grad_norm": 0.5266805291175842, "learning_rate": 5.769268079092809e-05, "loss": 0.0537, "step": 47830 }, { "grad_norm": 0.35844188928604126, "learning_rate": 5.767634254029151e-05, "loss": 0.0376, "step": 47840 }, { "grad_norm": 0.671869695186615, "learning_rate": 5.7660003450181655e-05, "loss": 0.0729, "step": 47850 }, { "grad_norm": 1.4816621541976929, "learning_rate": 5.764366352238534e-05, "loss": 0.0446, "step": 47860 }, { "grad_norm": 0.7085952758789062, "learning_rate": 5.7627322758689474e-05, "loss": 0.0418, "step": 47870 }, { "grad_norm": 0.5624690651893616, "learning_rate": 5.761098116088105e-05, "loss": 0.0526, "step": 47880 }, { "grad_norm": 0.6839287281036377, "learning_rate": 5.759463873074717e-05, "loss": 0.0428, "step": 47890 }, { "grad_norm": 0.5997526049613953, "learning_rate": 5.757829547007504e-05, "loss": 0.0474, "step": 47900 }, { "grad_norm": 0.6579053401947021, "learning_rate": 5.756195138065189e-05, "loss": 0.0456, "step": 47910 }, { "grad_norm": 0.5049791932106018, "learning_rate": 5.754560646426511e-05, "loss": 0.0516, "step": 47920 }, { "grad_norm": 0.41995200514793396, "learning_rate": 5.752926072270216e-05, "loss": 0.0448, "step": 47930 }, { "grad_norm": 0.6600032448768616, "learning_rate": 5.7512914157750563e-05, "loss": 0.0414, "step": 47940 }, { "grad_norm": 0.3878406584262848, "learning_rate": 5.749656677119798e-05, "loss": 0.037, "step": 47950 }, { "grad_norm": 0.6742215752601624, "learning_rate": 5.7480218564832125e-05, "loss": 0.0585, "step": 47960 }, { "grad_norm": 0.8408018946647644, "learning_rate": 5.746386954044082e-05, "loss": 0.0396, "step": 47970 }, { "grad_norm": 0.9774906635284424, "learning_rate": 5.744751969981195e-05, "loss": 0.0481, "step": 47980 }, { "grad_norm": 0.5958298444747925, "learning_rate": 5.7431169044733526e-05, "loss": 0.0533, "step": 47990 }, { "grad_norm": 0.8645719885826111, "learning_rate": 5.741481757699364e-05, "loss": 0.0465, "step": 48000 }, { "grad_norm": 0.365068644285202, "learning_rate": 5.7398465298380434e-05, "loss": 0.06, "step": 48010 }, { "grad_norm": 0.46514785289764404, "learning_rate": 5.7382112210682193e-05, "loss": 0.0673, "step": 48020 }, { "grad_norm": 0.676576554775238, "learning_rate": 5.7365758315687266e-05, "loss": 0.0473, "step": 48030 }, { "grad_norm": 0.560009241104126, "learning_rate": 5.734940361518407e-05, "loss": 0.0563, "step": 48040 }, { "grad_norm": 0.5513941049575806, "learning_rate": 5.733304811096116e-05, "loss": 0.0524, "step": 48050 }, { "grad_norm": 0.4539825916290283, "learning_rate": 5.731669180480713e-05, "loss": 0.0555, "step": 48060 }, { "grad_norm": 0.5599998235702515, "learning_rate": 5.730033469851067e-05, "loss": 0.0603, "step": 48070 }, { "grad_norm": 0.7215896248817444, "learning_rate": 5.7283976793860607e-05, "loss": 0.0577, "step": 48080 }, { "grad_norm": 0.34312233328819275, "learning_rate": 5.726761809264577e-05, "loss": 0.0414, "step": 48090 }, { "grad_norm": 0.9906890988349915, "learning_rate": 5.7251258596655155e-05, "loss": 0.0456, "step": 48100 }, { "grad_norm": 0.7209709286689758, "learning_rate": 5.72348983076778e-05, "loss": 0.05, "step": 48110 }, { "grad_norm": 1.9210914373397827, "learning_rate": 5.7218537227502854e-05, "loss": 0.0483, "step": 48120 }, { "grad_norm": 0.3807750642299652, "learning_rate": 5.720217535791951e-05, "loss": 0.0415, "step": 48130 }, { "grad_norm": 0.8001039028167725, "learning_rate": 5.718581270071711e-05, "loss": 0.0641, "step": 48140 }, { "grad_norm": 0.4549729526042938, "learning_rate": 5.716944925768505e-05, "loss": 0.0468, "step": 48150 }, { "grad_norm": 0.411735862493515, "learning_rate": 5.7153085030612786e-05, "loss": 0.0417, "step": 48160 }, { "grad_norm": 0.3724707067012787, "learning_rate": 5.713672002128989e-05, "loss": 0.0625, "step": 48170 }, { "grad_norm": 0.5091912150382996, "learning_rate": 5.712035423150602e-05, "loss": 0.0509, "step": 48180 }, { "grad_norm": 0.4318329989910126, "learning_rate": 5.710398766305094e-05, "loss": 0.0452, "step": 48190 }, { "grad_norm": 0.5125555396080017, "learning_rate": 5.708762031771442e-05, "loss": 0.0408, "step": 48200 }, { "grad_norm": 0.6097803711891174, "learning_rate": 5.707125219728642e-05, "loss": 0.0463, "step": 48210 }, { "grad_norm": 0.4446793496608734, "learning_rate": 5.7054883303556905e-05, "loss": 0.0526, "step": 48220 }, { "grad_norm": 0.5382695198059082, "learning_rate": 5.703851363831595e-05, "loss": 0.0516, "step": 48230 }, { "grad_norm": 0.4958469569683075, "learning_rate": 5.702214320335374e-05, "loss": 0.059, "step": 48240 }, { "grad_norm": 0.5451736450195312, "learning_rate": 5.7005772000460514e-05, "loss": 0.048, "step": 48250 }, { "grad_norm": 0.531630277633667, "learning_rate": 5.698940003142656e-05, "loss": 0.048, "step": 48260 }, { "grad_norm": 0.41111132502555847, "learning_rate": 5.697302729804236e-05, "loss": 0.0595, "step": 48270 }, { "grad_norm": 0.84466153383255, "learning_rate": 5.695665380209837e-05, "loss": 0.0552, "step": 48280 }, { "grad_norm": 0.3836187422275543, "learning_rate": 5.6940279545385165e-05, "loss": 0.0407, "step": 48290 }, { "grad_norm": 1.2367804050445557, "learning_rate": 5.692390452969344e-05, "loss": 0.0524, "step": 48300 }, { "grad_norm": 0.7825073599815369, "learning_rate": 5.690752875681392e-05, "loss": 0.0449, "step": 48310 }, { "grad_norm": 0.3729950487613678, "learning_rate": 5.6891152228537435e-05, "loss": 0.0465, "step": 48320 }, { "grad_norm": 0.5997197031974792, "learning_rate": 5.687477494665492e-05, "loss": 0.0434, "step": 48330 }, { "grad_norm": 0.48160696029663086, "learning_rate": 5.685839691295733e-05, "loss": 0.0459, "step": 48340 }, { "grad_norm": 0.5074182152748108, "learning_rate": 5.6842018129235786e-05, "loss": 0.0458, "step": 48350 }, { "grad_norm": 0.8842090964317322, "learning_rate": 5.6825638597281404e-05, "loss": 0.0368, "step": 48360 }, { "grad_norm": 0.5022940039634705, "learning_rate": 5.680925831888546e-05, "loss": 0.0621, "step": 48370 }, { "grad_norm": 0.44211286306381226, "learning_rate": 5.6792877295839274e-05, "loss": 0.0496, "step": 48380 }, { "grad_norm": 0.6680053472518921, "learning_rate": 5.6776495529934224e-05, "loss": 0.0484, "step": 48390 }, { "grad_norm": 0.4611109793186188, "learning_rate": 5.6760113022961824e-05, "loss": 0.0457, "step": 48400 }, { "grad_norm": 0.7036463022232056, "learning_rate": 5.6743729776713617e-05, "loss": 0.054, "step": 48410 }, { "grad_norm": 0.6748472452163696, "learning_rate": 5.672734579298126e-05, "loss": 0.0472, "step": 48420 }, { "grad_norm": 0.5506162047386169, "learning_rate": 5.671096107355649e-05, "loss": 0.0511, "step": 48430 }, { "grad_norm": 0.4147465229034424, "learning_rate": 5.66945756202311e-05, "loss": 0.0424, "step": 48440 }, { "grad_norm": 0.47986266016960144, "learning_rate": 5.667818943479699e-05, "loss": 0.0602, "step": 48450 }, { "grad_norm": 0.5578479170799255, "learning_rate": 5.666180251904612e-05, "loss": 0.0536, "step": 48460 }, { "grad_norm": 0.46732401847839355, "learning_rate": 5.6645414874770555e-05, "loss": 0.0426, "step": 48470 }, { "grad_norm": 0.7261407375335693, "learning_rate": 5.66290265037624e-05, "loss": 0.0606, "step": 48480 }, { "grad_norm": 0.4551587998867035, "learning_rate": 5.661263740781386e-05, "loss": 0.0387, "step": 48490 }, { "grad_norm": 0.7516099810600281, "learning_rate": 5.6596247588717254e-05, "loss": 0.0489, "step": 48500 }, { "grad_norm": 0.8388445377349854, "learning_rate": 5.6579857048264926e-05, "loss": 0.0457, "step": 48510 }, { "grad_norm": 0.6540353298187256, "learning_rate": 5.6563465788249314e-05, "loss": 0.045, "step": 48520 }, { "grad_norm": 0.5334793329238892, "learning_rate": 5.6547073810462956e-05, "loss": 0.0475, "step": 48530 }, { "grad_norm": 0.5529042482376099, "learning_rate": 5.653068111669846e-05, "loss": 0.0461, "step": 48540 }, { "grad_norm": 0.3700396716594696, "learning_rate": 5.651428770874848e-05, "loss": 0.0534, "step": 48550 }, { "grad_norm": 0.43248337507247925, "learning_rate": 5.64978935884058e-05, "loss": 0.0428, "step": 48560 }, { "grad_norm": 0.5497766137123108, "learning_rate": 5.6481498757463244e-05, "loss": 0.0551, "step": 48570 }, { "grad_norm": 0.41056397557258606, "learning_rate": 5.646510321771373e-05, "loss": 0.05, "step": 48580 }, { "grad_norm": 0.6331035494804382, "learning_rate": 5.644870697095024e-05, "loss": 0.048, "step": 48590 }, { "grad_norm": 0.6750892400741577, "learning_rate": 5.643231001896586e-05, "loss": 0.0538, "step": 48600 }, { "grad_norm": 0.48631271719932556, "learning_rate": 5.6415912363553726e-05, "loss": 0.0754, "step": 48610 }, { "grad_norm": 0.9862855672836304, "learning_rate": 5.639951400650706e-05, "loss": 0.0577, "step": 48620 }, { "grad_norm": 0.6576645374298096, "learning_rate": 5.6383114949619165e-05, "loss": 0.0497, "step": 48630 }, { "grad_norm": 1.2207579612731934, "learning_rate": 5.636671519468342e-05, "loss": 0.0573, "step": 48640 }, { "grad_norm": 0.5996956825256348, "learning_rate": 5.635031474349327e-05, "loss": 0.0457, "step": 48650 }, { "grad_norm": 0.4863170087337494, "learning_rate": 5.6333913597842246e-05, "loss": 0.0373, "step": 48660 }, { "grad_norm": 0.6213166117668152, "learning_rate": 5.6317511759523955e-05, "loss": 0.0591, "step": 48670 }, { "grad_norm": 0.4208948016166687, "learning_rate": 5.630110923033207e-05, "loss": 0.0454, "step": 48680 }, { "grad_norm": 0.6109558939933777, "learning_rate": 5.628470601206036e-05, "loss": 0.0477, "step": 48690 }, { "grad_norm": 0.8811514377593994, "learning_rate": 5.626830210650266e-05, "loss": 0.0478, "step": 48700 }, { "grad_norm": 0.5046055316925049, "learning_rate": 5.625189751545285e-05, "loss": 0.0551, "step": 48710 }, { "grad_norm": 0.33309945464134216, "learning_rate": 5.6235492240704936e-05, "loss": 0.0385, "step": 48720 }, { "grad_norm": 0.5117056965827942, "learning_rate": 5.621908628405296e-05, "loss": 0.0476, "step": 48730 }, { "grad_norm": 0.8565239906311035, "learning_rate": 5.620267964729106e-05, "loss": 0.0488, "step": 48740 }, { "grad_norm": 0.3929729461669922, "learning_rate": 5.618627233221344e-05, "loss": 0.0395, "step": 48750 }, { "grad_norm": 1.2609225511550903, "learning_rate": 5.616986434061438e-05, "loss": 0.0552, "step": 48760 }, { "grad_norm": 1.2037290334701538, "learning_rate": 5.615345567428822e-05, "loss": 0.0466, "step": 48770 }, { "grad_norm": 0.3301422595977783, "learning_rate": 5.613704633502941e-05, "loss": 0.0358, "step": 48780 }, { "grad_norm": 0.6114118695259094, "learning_rate": 5.612063632463245e-05, "loss": 0.0482, "step": 48790 }, { "grad_norm": 0.38874804973602295, "learning_rate": 5.610422564489188e-05, "loss": 0.0456, "step": 48800 }, { "grad_norm": 0.5253026485443115, "learning_rate": 5.608781429760239e-05, "loss": 0.045, "step": 48810 }, { "grad_norm": 0.43387046456336975, "learning_rate": 5.607140228455866e-05, "loss": 0.0386, "step": 48820 }, { "grad_norm": 0.4759795665740967, "learning_rate": 5.605498960755553e-05, "loss": 0.0584, "step": 48830 }, { "grad_norm": 0.6330108642578125, "learning_rate": 5.603857626838782e-05, "loss": 0.0471, "step": 48840 }, { "grad_norm": 0.5746000409126282, "learning_rate": 5.60221622688505e-05, "loss": 0.0448, "step": 48850 }, { "grad_norm": 0.7983056306838989, "learning_rate": 5.6005747610738565e-05, "loss": 0.043, "step": 48860 }, { "grad_norm": 0.3688174784183502, "learning_rate": 5.59893322958471e-05, "loss": 0.0425, "step": 48870 }, { "grad_norm": 0.6210774183273315, "learning_rate": 5.5972916325971256e-05, "loss": 0.0537, "step": 48880 }, { "grad_norm": 0.41882938146591187, "learning_rate": 5.595649970290628e-05, "loss": 0.0548, "step": 48890 }, { "grad_norm": 0.39094266295433044, "learning_rate": 5.5940082428447426e-05, "loss": 0.0489, "step": 48900 }, { "grad_norm": 0.6554269790649414, "learning_rate": 5.592366450439012e-05, "loss": 0.053, "step": 48910 }, { "grad_norm": 0.6459417343139648, "learning_rate": 5.590724593252975e-05, "loss": 0.0513, "step": 48920 }, { "grad_norm": 0.3924335539340973, "learning_rate": 5.589082671466184e-05, "loss": 0.0432, "step": 48930 }, { "grad_norm": 0.43761399388313293, "learning_rate": 5.587440685258199e-05, "loss": 0.0522, "step": 48940 }, { "grad_norm": 0.5510258674621582, "learning_rate": 5.585798634808583e-05, "loss": 0.0452, "step": 48950 }, { "grad_norm": 0.5934953093528748, "learning_rate": 5.584156520296909e-05, "loss": 0.0469, "step": 48960 }, { "grad_norm": 0.6365788578987122, "learning_rate": 5.582514341902757e-05, "loss": 0.0474, "step": 48970 }, { "grad_norm": 0.44677719473838806, "learning_rate": 5.580872099805713e-05, "loss": 0.0417, "step": 48980 }, { "grad_norm": 0.38645824790000916, "learning_rate": 5.5792297941853674e-05, "loss": 0.0476, "step": 48990 }, { "grad_norm": 0.4894968867301941, "learning_rate": 5.5775874252213247e-05, "loss": 0.0478, "step": 49000 }, { "grad_norm": 0.49810507893562317, "learning_rate": 5.575944993093189e-05, "loss": 0.054, "step": 49010 }, { "grad_norm": 0.42960813641548157, "learning_rate": 5.574302497980574e-05, "loss": 0.0416, "step": 49020 }, { "grad_norm": 0.7651437520980835, "learning_rate": 5.5726599400631005e-05, "loss": 0.051, "step": 49030 }, { "grad_norm": 0.460671067237854, "learning_rate": 5.571017319520399e-05, "loss": 0.0558, "step": 49040 }, { "grad_norm": 0.47567346692085266, "learning_rate": 5.5693746365320984e-05, "loss": 0.0428, "step": 49050 }, { "grad_norm": 0.7476441264152527, "learning_rate": 5.567731891277848e-05, "loss": 0.0545, "step": 49060 }, { "grad_norm": 0.9362293481826782, "learning_rate": 5.566089083937288e-05, "loss": 0.0618, "step": 49070 }, { "grad_norm": 0.6302391886711121, "learning_rate": 5.564446214690079e-05, "loss": 0.0558, "step": 49080 }, { "grad_norm": 0.8626688122749329, "learning_rate": 5.56280328371588e-05, "loss": 0.0676, "step": 49090 }, { "grad_norm": 0.485465943813324, "learning_rate": 5.5611602911943596e-05, "loss": 0.0426, "step": 49100 }, { "grad_norm": 0.8149463534355164, "learning_rate": 5.559517237305194e-05, "loss": 0.0518, "step": 49110 }, { "grad_norm": 0.6332792639732361, "learning_rate": 5.557874122228064e-05, "loss": 0.0511, "step": 49120 }, { "grad_norm": 0.6312366724014282, "learning_rate": 5.55623094614266e-05, "loss": 0.0545, "step": 49130 }, { "grad_norm": 0.5423977971076965, "learning_rate": 5.5545877092286743e-05, "loss": 0.055, "step": 49140 }, { "grad_norm": 0.5526838302612305, "learning_rate": 5.552944411665811e-05, "loss": 0.0393, "step": 49150 }, { "grad_norm": 0.4837694764137268, "learning_rate": 5.5513010536337795e-05, "loss": 0.0541, "step": 49160 }, { "grad_norm": 0.5087741017341614, "learning_rate": 5.5496576353122934e-05, "loss": 0.0508, "step": 49170 }, { "grad_norm": 0.5616195201873779, "learning_rate": 5.548014156881074e-05, "loss": 0.0598, "step": 49180 }, { "grad_norm": 0.4608246386051178, "learning_rate": 5.5463706185198494e-05, "loss": 0.0508, "step": 49190 }, { "grad_norm": 0.6202324628829956, "learning_rate": 5.544727020408358e-05, "loss": 0.0511, "step": 49200 }, { "grad_norm": 0.8300656080245972, "learning_rate": 5.5430833627263366e-05, "loss": 0.038, "step": 49210 }, { "grad_norm": 0.39962396025657654, "learning_rate": 5.5414396456535354e-05, "loss": 0.0468, "step": 49220 }, { "grad_norm": 0.5045738816261292, "learning_rate": 5.539795869369709e-05, "loss": 0.0387, "step": 49230 }, { "grad_norm": 0.709884524345398, "learning_rate": 5.538152034054618e-05, "loss": 0.0444, "step": 49240 }, { "grad_norm": 0.4823417663574219, "learning_rate": 5.536508139888028e-05, "loss": 0.0491, "step": 49250 }, { "grad_norm": 0.47797971963882446, "learning_rate": 5.534864187049716e-05, "loss": 0.0412, "step": 49260 }, { "grad_norm": 0.38311395049095154, "learning_rate": 5.5332201757194615e-05, "loss": 0.0451, "step": 49270 }, { "grad_norm": 0.3309313654899597, "learning_rate": 5.531576106077049e-05, "loss": 0.0407, "step": 49280 }, { "grad_norm": 0.6465871930122375, "learning_rate": 5.529931978302272e-05, "loss": 0.0359, "step": 49290 }, { "grad_norm": 0.4196827709674835, "learning_rate": 5.528287792574932e-05, "loss": 0.0397, "step": 49300 }, { "grad_norm": 0.4522743821144104, "learning_rate": 5.5266435490748324e-05, "loss": 0.0416, "step": 49310 }, { "grad_norm": 0.3774966299533844, "learning_rate": 5.524999247981787e-05, "loss": 0.0438, "step": 49320 }, { "grad_norm": 0.6003715991973877, "learning_rate": 5.523354889475613e-05, "loss": 0.0463, "step": 49330 }, { "grad_norm": 0.5032938122749329, "learning_rate": 5.521710473736134e-05, "loss": 0.0495, "step": 49340 }, { "grad_norm": 0.5076174736022949, "learning_rate": 5.5200660009431826e-05, "loss": 0.0406, "step": 49350 }, { "grad_norm": 0.5193919539451599, "learning_rate": 5.518421471276596e-05, "loss": 0.0412, "step": 49360 }, { "grad_norm": 0.4764234125614166, "learning_rate": 5.5167768849162174e-05, "loss": 0.0362, "step": 49370 }, { "grad_norm": 0.5770221948623657, "learning_rate": 5.515132242041893e-05, "loss": 0.0455, "step": 49380 }, { "grad_norm": 0.4794163405895233, "learning_rate": 5.513487542833483e-05, "loss": 0.0485, "step": 49390 }, { "grad_norm": 0.6369404792785645, "learning_rate": 5.5118427874708466e-05, "loss": 0.0483, "step": 49400 }, { "grad_norm": 0.6561362743377686, "learning_rate": 5.510197976133853e-05, "loss": 0.0421, "step": 49410 }, { "grad_norm": 0.674812912940979, "learning_rate": 5.508553109002376e-05, "loss": 0.0424, "step": 49420 }, { "grad_norm": 0.435167521238327, "learning_rate": 5.5069081862562957e-05, "loss": 0.0442, "step": 49430 }, { "grad_norm": 0.40769919753074646, "learning_rate": 5.5052632080754965e-05, "loss": 0.0432, "step": 49440 }, { "grad_norm": 0.5492900013923645, "learning_rate": 5.503618174639874e-05, "loss": 0.0508, "step": 49450 }, { "grad_norm": 0.5882378220558167, "learning_rate": 5.501973086129325e-05, "loss": 0.0457, "step": 49460 }, { "grad_norm": 0.4103412330150604, "learning_rate": 5.500327942723753e-05, "loss": 0.0392, "step": 49470 }, { "grad_norm": 0.38318443298339844, "learning_rate": 5.498682744603071e-05, "loss": 0.0452, "step": 49480 }, { "grad_norm": 1.1158112287521362, "learning_rate": 5.497037491947194e-05, "loss": 0.0478, "step": 49490 }, { "grad_norm": 0.5422366857528687, "learning_rate": 5.4953921849360424e-05, "loss": 0.0552, "step": 49500 }, { "grad_norm": 1.2614976167678833, "learning_rate": 5.493746823749547e-05, "loss": 0.048, "step": 49510 }, { "grad_norm": 0.8221474885940552, "learning_rate": 5.49210140856764e-05, "loss": 0.0506, "step": 49520 }, { "grad_norm": 0.3022589683532715, "learning_rate": 5.4904559395702647e-05, "loss": 0.0454, "step": 49530 }, { "grad_norm": 0.36380529403686523, "learning_rate": 5.488810416937364e-05, "loss": 0.0467, "step": 49540 }, { "grad_norm": 0.4134448170661926, "learning_rate": 5.4871648408488926e-05, "loss": 0.0505, "step": 49550 }, { "grad_norm": 0.9109108448028564, "learning_rate": 5.485519211484807e-05, "loss": 0.0457, "step": 49560 }, { "grad_norm": 0.5881561636924744, "learning_rate": 5.4838735290250676e-05, "loss": 0.0506, "step": 49570 }, { "grad_norm": 0.48839282989501953, "learning_rate": 5.482227793649649e-05, "loss": 0.0499, "step": 49580 }, { "grad_norm": 0.4555743634700775, "learning_rate": 5.480582005538524e-05, "loss": 0.0461, "step": 49590 }, { "grad_norm": 0.3730389177799225, "learning_rate": 5.478936164871671e-05, "loss": 0.0481, "step": 49600 }, { "grad_norm": 0.4707052409648895, "learning_rate": 5.477290271829081e-05, "loss": 0.0491, "step": 49610 }, { "grad_norm": 0.44381141662597656, "learning_rate": 5.475644326590744e-05, "loss": 0.0363, "step": 49620 }, { "grad_norm": 0.4322269558906555, "learning_rate": 5.473998329336658e-05, "loss": 0.0481, "step": 49630 }, { "grad_norm": 0.41045835614204407, "learning_rate": 5.4723522802468286e-05, "loss": 0.0426, "step": 49640 }, { "grad_norm": 0.8382574915885925, "learning_rate": 5.470706179501264e-05, "loss": 0.0474, "step": 49650 }, { "grad_norm": 0.5750255584716797, "learning_rate": 5.4690600272799776e-05, "loss": 0.0507, "step": 49660 }, { "grad_norm": 0.6426969170570374, "learning_rate": 5.467413823762993e-05, "loss": 0.0603, "step": 49670 }, { "grad_norm": 0.646193265914917, "learning_rate": 5.465767569130335e-05, "loss": 0.0501, "step": 49680 }, { "grad_norm": 0.5721970796585083, "learning_rate": 5.464121263562036e-05, "loss": 0.0493, "step": 49690 }, { "grad_norm": 0.44297048449516296, "learning_rate": 5.4624749072381343e-05, "loss": 0.0477, "step": 49700 }, { "grad_norm": 0.31419774889945984, "learning_rate": 5.460828500338672e-05, "loss": 0.0437, "step": 49710 }, { "grad_norm": 0.48117512464523315, "learning_rate": 5.459182043043698e-05, "loss": 0.0495, "step": 49720 }, { "grad_norm": 0.41595977544784546, "learning_rate": 5.457535535533265e-05, "loss": 0.0446, "step": 49730 }, { "grad_norm": 0.47305697202682495, "learning_rate": 5.4558889779874334e-05, "loss": 0.0531, "step": 49740 }, { "grad_norm": 0.7877486348152161, "learning_rate": 5.454242370586269e-05, "loss": 0.0559, "step": 49750 }, { "grad_norm": 0.481297105550766, "learning_rate": 5.452595713509843e-05, "loss": 0.0398, "step": 49760 }, { "grad_norm": 0.47672298550605774, "learning_rate": 5.4509490069382274e-05, "loss": 0.0473, "step": 49770 }, { "grad_norm": 0.4418010115623474, "learning_rate": 5.449302251051509e-05, "loss": 0.0389, "step": 49780 }, { "grad_norm": 0.27207499742507935, "learning_rate": 5.44765544602977e-05, "loss": 0.0545, "step": 49790 }, { "grad_norm": 0.27604034543037415, "learning_rate": 5.446008592053107e-05, "loss": 0.0449, "step": 49800 }, { "grad_norm": 0.37538623809814453, "learning_rate": 5.4443616893016135e-05, "loss": 0.0437, "step": 49810 }, { "grad_norm": 0.5581329464912415, "learning_rate": 5.4427147379553934e-05, "loss": 0.0546, "step": 49820 }, { "grad_norm": 0.4172918498516083, "learning_rate": 5.441067738194556e-05, "loss": 0.0432, "step": 49830 }, { "grad_norm": 0.4389480650424957, "learning_rate": 5.439420690199214e-05, "loss": 0.0533, "step": 49840 }, { "grad_norm": 0.3905056118965149, "learning_rate": 5.437773594149484e-05, "loss": 0.0415, "step": 49850 }, { "grad_norm": 0.30529841780662537, "learning_rate": 5.436126450225495e-05, "loss": 0.0521, "step": 49860 }, { "grad_norm": 0.40584176778793335, "learning_rate": 5.434479258607371e-05, "loss": 0.0371, "step": 49870 }, { "grad_norm": 0.4604199528694153, "learning_rate": 5.43283201947525e-05, "loss": 0.053, "step": 49880 }, { "grad_norm": 0.4109826385974884, "learning_rate": 5.431184733009268e-05, "loss": 0.0524, "step": 49890 }, { "grad_norm": 0.35533609986305237, "learning_rate": 5.4295373993895736e-05, "loss": 0.0503, "step": 49900 }, { "grad_norm": 0.4209405779838562, "learning_rate": 5.4278900187963157e-05, "loss": 0.0497, "step": 49910 }, { "grad_norm": 0.5112942457199097, "learning_rate": 5.426242591409646e-05, "loss": 0.0458, "step": 49920 }, { "grad_norm": 0.3331012427806854, "learning_rate": 5.42459511740973e-05, "loss": 0.0532, "step": 49930 }, { "grad_norm": 0.5582677721977234, "learning_rate": 5.422947596976729e-05, "loss": 0.0451, "step": 49940 }, { "grad_norm": 0.5960338115692139, "learning_rate": 5.4213000302908134e-05, "loss": 0.0464, "step": 49950 }, { "grad_norm": 0.3320437967777252, "learning_rate": 5.419652417532162e-05, "loss": 0.0415, "step": 49960 }, { "grad_norm": 0.5893663763999939, "learning_rate": 5.4180047588809534e-05, "loss": 0.0389, "step": 49970 }, { "grad_norm": 0.8263242244720459, "learning_rate": 5.4163570545173704e-05, "loss": 0.0446, "step": 49980 }, { "grad_norm": 0.6558340787887573, "learning_rate": 5.414709304621608e-05, "loss": 0.0474, "step": 49990 }, { "grad_norm": 0.4394862651824951, "learning_rate": 5.413061509373858e-05, "loss": 0.0478, "step": 50000 }, { "grad_norm": 1.9786291122436523, "learning_rate": 5.411413668954324e-05, "loss": 0.0533, "step": 50010 }, { "grad_norm": 0.47168534994125366, "learning_rate": 5.4097657835432083e-05, "loss": 0.046, "step": 50020 }, { "grad_norm": 0.4052119255065918, "learning_rate": 5.408117853320723e-05, "loss": 0.0485, "step": 50030 }, { "grad_norm": 0.5892705917358398, "learning_rate": 5.406469878467084e-05, "loss": 0.038, "step": 50040 }, { "grad_norm": 1.0271941423416138, "learning_rate": 5.404821859162509e-05, "loss": 0.052, "step": 50050 }, { "grad_norm": 0.560914933681488, "learning_rate": 5.403173795587225e-05, "loss": 0.0447, "step": 50060 }, { "grad_norm": 0.5866584777832031, "learning_rate": 5.4015256879214606e-05, "loss": 0.0495, "step": 50070 }, { "grad_norm": 0.38572508096694946, "learning_rate": 5.39987753634545e-05, "loss": 0.0422, "step": 50080 }, { "grad_norm": 0.27223676443099976, "learning_rate": 5.3982293410394335e-05, "loss": 0.0518, "step": 50090 }, { "grad_norm": 0.47110724449157715, "learning_rate": 5.396581102183655e-05, "loss": 0.0443, "step": 50100 }, { "grad_norm": 0.5943149924278259, "learning_rate": 5.3949328199583634e-05, "loss": 0.0431, "step": 50110 }, { "grad_norm": 0.4075859487056732, "learning_rate": 5.393284494543813e-05, "loss": 0.0521, "step": 50120 }, { "grad_norm": 0.4248519539833069, "learning_rate": 5.391636126120262e-05, "loss": 0.0396, "step": 50130 }, { "grad_norm": 0.43973761796951294, "learning_rate": 5.389987714867971e-05, "loss": 0.0443, "step": 50140 }, { "grad_norm": 0.46829742193222046, "learning_rate": 5.3883392609672114e-05, "loss": 0.0472, "step": 50150 }, { "grad_norm": 0.3866610825061798, "learning_rate": 5.3866907645982545e-05, "loss": 0.0444, "step": 50160 }, { "grad_norm": 1.288065791130066, "learning_rate": 5.385042225941375e-05, "loss": 0.0484, "step": 50170 }, { "grad_norm": 0.4432472288608551, "learning_rate": 5.3833936451768585e-05, "loss": 0.0476, "step": 50180 }, { "grad_norm": 0.532293975353241, "learning_rate": 5.3817450224849884e-05, "loss": 0.0501, "step": 50190 }, { "grad_norm": 0.6161267757415771, "learning_rate": 5.380096358046056e-05, "loss": 0.0564, "step": 50200 }, { "grad_norm": 0.26690924167633057, "learning_rate": 5.378447652040359e-05, "loss": 0.0446, "step": 50210 }, { "grad_norm": 0.42937779426574707, "learning_rate": 5.376798904648195e-05, "loss": 0.0374, "step": 50220 }, { "grad_norm": 0.5154770612716675, "learning_rate": 5.375150116049869e-05, "loss": 0.0406, "step": 50230 }, { "grad_norm": 0.60673588514328, "learning_rate": 5.373501286425691e-05, "loss": 0.035, "step": 50240 }, { "grad_norm": 0.7793914675712585, "learning_rate": 5.3718524159559726e-05, "loss": 0.0415, "step": 50250 }, { "grad_norm": 0.46942174434661865, "learning_rate": 5.370203504821034e-05, "loss": 0.0492, "step": 50260 }, { "grad_norm": 0.4358541965484619, "learning_rate": 5.368554553201196e-05, "loss": 0.0527, "step": 50270 }, { "grad_norm": 0.5512257814407349, "learning_rate": 5.3669055612767874e-05, "loss": 0.0561, "step": 50280 }, { "grad_norm": 0.6213289499282837, "learning_rate": 5.3652565292281376e-05, "loss": 0.0444, "step": 50290 }, { "grad_norm": 0.48868638277053833, "learning_rate": 5.3636074572355824e-05, "loss": 0.0495, "step": 50300 }, { "grad_norm": 0.46185049414634705, "learning_rate": 5.361958345479463e-05, "loss": 0.0484, "step": 50310 }, { "grad_norm": 0.3660484254360199, "learning_rate": 5.360309194140123e-05, "loss": 0.0422, "step": 50320 }, { "grad_norm": 0.7896373867988586, "learning_rate": 5.3586600033979106e-05, "loss": 0.0442, "step": 50330 }, { "grad_norm": 1.2559552192687988, "learning_rate": 5.357010773433181e-05, "loss": 0.0497, "step": 50340 }, { "grad_norm": 0.4918675720691681, "learning_rate": 5.3553615044262895e-05, "loss": 0.0355, "step": 50350 }, { "grad_norm": 0.447292685508728, "learning_rate": 5.3537121965575964e-05, "loss": 0.0454, "step": 50360 }, { "grad_norm": 1.231067419052124, "learning_rate": 5.352062850007471e-05, "loss": 0.0642, "step": 50370 }, { "grad_norm": 0.4701552093029022, "learning_rate": 5.350413464956282e-05, "loss": 0.0416, "step": 50380 }, { "grad_norm": 0.6325267553329468, "learning_rate": 5.348764041584403e-05, "loss": 0.0527, "step": 50390 }, { "grad_norm": 0.5611899495124817, "learning_rate": 5.347114580072212e-05, "loss": 0.0442, "step": 50400 }, { "grad_norm": 0.6074778437614441, "learning_rate": 5.345465080600093e-05, "loss": 0.0434, "step": 50410 }, { "grad_norm": 0.4659513533115387, "learning_rate": 5.3438155433484326e-05, "loss": 0.0493, "step": 50420 }, { "grad_norm": 0.567805826663971, "learning_rate": 5.3421659684976197e-05, "loss": 0.0494, "step": 50430 }, { "grad_norm": 0.7635312080383301, "learning_rate": 5.340516356228052e-05, "loss": 0.0486, "step": 50440 }, { "grad_norm": 0.399356871843338, "learning_rate": 5.338866706720128e-05, "loss": 0.0498, "step": 50450 }, { "grad_norm": 0.5188507437705994, "learning_rate": 5.337217020154249e-05, "loss": 0.0507, "step": 50460 }, { "grad_norm": 0.5662838220596313, "learning_rate": 5.335567296710825e-05, "loss": 0.0435, "step": 50470 }, { "grad_norm": 0.5752311944961548, "learning_rate": 5.333917536570265e-05, "loss": 0.0433, "step": 50480 }, { "grad_norm": 0.5180765986442566, "learning_rate": 5.332267739912986e-05, "loss": 0.0473, "step": 50490 }, { "grad_norm": 0.3738529682159424, "learning_rate": 5.330617906919405e-05, "loss": 0.0534, "step": 50500 }, { "grad_norm": 0.3709796369075775, "learning_rate": 5.328968037769949e-05, "loss": 0.0437, "step": 50510 }, { "grad_norm": 0.4072938561439514, "learning_rate": 5.327318132645042e-05, "loss": 0.0482, "step": 50520 }, { "grad_norm": 0.4413878321647644, "learning_rate": 5.325668191725116e-05, "loss": 0.0397, "step": 50530 }, { "grad_norm": 0.48613664507865906, "learning_rate": 5.324018215190606e-05, "loss": 0.0507, "step": 50540 }, { "grad_norm": 0.5752778649330139, "learning_rate": 5.3223682032219515e-05, "loss": 0.0432, "step": 50550 }, { "grad_norm": 0.3499477803707123, "learning_rate": 5.320718155999595e-05, "loss": 0.0485, "step": 50560 }, { "grad_norm": 0.5121092200279236, "learning_rate": 5.3190680737039835e-05, "loss": 0.0615, "step": 50570 }, { "grad_norm": 0.3577420115470886, "learning_rate": 5.317417956515567e-05, "loss": 0.0396, "step": 50580 }, { "grad_norm": 0.42882534861564636, "learning_rate": 5.315767804614798e-05, "loss": 0.0474, "step": 50590 }, { "grad_norm": 0.3575896918773651, "learning_rate": 5.3141176181821395e-05, "loss": 0.0461, "step": 50600 }, { "grad_norm": 0.34406155347824097, "learning_rate": 5.31246739739805e-05, "loss": 0.0422, "step": 50610 }, { "grad_norm": 0.4597219228744507, "learning_rate": 5.310817142442994e-05, "loss": 0.056, "step": 50620 }, { "grad_norm": 0.8164836168289185, "learning_rate": 5.309166853497445e-05, "loss": 0.0498, "step": 50630 }, { "grad_norm": 0.501401960849762, "learning_rate": 5.307516530741873e-05, "loss": 0.0511, "step": 50640 }, { "grad_norm": 0.6397583484649658, "learning_rate": 5.305866174356754e-05, "loss": 0.0502, "step": 50650 }, { "grad_norm": 0.5288936495780945, "learning_rate": 5.304215784522571e-05, "loss": 0.0399, "step": 50660 }, { "grad_norm": 0.6733275055885315, "learning_rate": 5.302565361419808e-05, "loss": 0.0584, "step": 50670 }, { "grad_norm": 1.9220706224441528, "learning_rate": 5.3009149052289507e-05, "loss": 0.0475, "step": 50680 }, { "grad_norm": 0.3902234435081482, "learning_rate": 5.299264416130493e-05, "loss": 0.0509, "step": 50690 }, { "grad_norm": 0.5033369064331055, "learning_rate": 5.297613894304928e-05, "loss": 0.0444, "step": 50700 }, { "grad_norm": 0.6404257416725159, "learning_rate": 5.2959633399327534e-05, "loss": 0.0416, "step": 50710 }, { "grad_norm": 0.49606263637542725, "learning_rate": 5.294312753194476e-05, "loss": 0.0382, "step": 50720 }, { "grad_norm": 0.5364190936088562, "learning_rate": 5.292662134270596e-05, "loss": 0.0489, "step": 50730 }, { "grad_norm": 0.4069853127002716, "learning_rate": 5.291011483341626e-05, "loss": 0.0474, "step": 50740 }, { "grad_norm": 0.7341957688331604, "learning_rate": 5.2893608005880767e-05, "loss": 0.0544, "step": 50750 }, { "grad_norm": 0.3752872347831726, "learning_rate": 5.287710086190467e-05, "loss": 0.0421, "step": 50760 }, { "grad_norm": 0.37658804655075073, "learning_rate": 5.286059340329314e-05, "loss": 0.0506, "step": 50770 }, { "grad_norm": 0.7039373517036438, "learning_rate": 5.284408563185141e-05, "loss": 0.0527, "step": 50780 }, { "grad_norm": 0.5036616921424866, "learning_rate": 5.282757754938476e-05, "loss": 0.0476, "step": 50790 }, { "grad_norm": 0.5251761078834534, "learning_rate": 5.281106915769849e-05, "loss": 0.0546, "step": 50800 }, { "grad_norm": 0.6735659241676331, "learning_rate": 5.2794560458597897e-05, "loss": 0.0507, "step": 50810 }, { "grad_norm": 0.45249271392822266, "learning_rate": 5.27780514538884e-05, "loss": 0.0404, "step": 50820 }, { "grad_norm": 0.5599123239517212, "learning_rate": 5.2761542145375365e-05, "loss": 0.0527, "step": 50830 }, { "grad_norm": 1.0838373899459839, "learning_rate": 5.274503253486421e-05, "loss": 0.0435, "step": 50840 }, { "grad_norm": 0.40046143531799316, "learning_rate": 5.272852262416046e-05, "loss": 0.0474, "step": 50850 }, { "grad_norm": 0.8017778396606445, "learning_rate": 5.2712012415069555e-05, "loss": 0.0493, "step": 50860 }, { "grad_norm": 0.6170330047607422, "learning_rate": 5.269550190939705e-05, "loss": 0.0488, "step": 50870 }, { "grad_norm": 0.9621348977088928, "learning_rate": 5.267899110894852e-05, "loss": 0.0393, "step": 50880 }, { "grad_norm": 0.3454309403896332, "learning_rate": 5.266248001552955e-05, "loss": 0.0428, "step": 50890 }, { "grad_norm": 0.39577755331993103, "learning_rate": 5.264596863094575e-05, "loss": 0.0489, "step": 50900 }, { "grad_norm": 0.9749054312705994, "learning_rate": 5.262945695700282e-05, "loss": 0.0433, "step": 50910 }, { "grad_norm": 1.024625539779663, "learning_rate": 5.261294499550643e-05, "loss": 0.0431, "step": 50920 }, { "grad_norm": 0.49809104204177856, "learning_rate": 5.25964327482623e-05, "loss": 0.0407, "step": 50930 }, { "grad_norm": 0.6985495090484619, "learning_rate": 5.257992021707617e-05, "loss": 0.0439, "step": 50940 }, { "grad_norm": 2.100318193435669, "learning_rate": 5.256340740375387e-05, "loss": 0.0469, "step": 50950 }, { "grad_norm": 0.40298622846603394, "learning_rate": 5.254689431010117e-05, "loss": 0.0477, "step": 50960 }, { "grad_norm": 0.38957974314689636, "learning_rate": 5.253038093792395e-05, "loss": 0.0378, "step": 50970 }, { "grad_norm": 0.40536317229270935, "learning_rate": 5.251386728902806e-05, "loss": 0.0464, "step": 50980 }, { "grad_norm": 0.4364190101623535, "learning_rate": 5.2497353365219446e-05, "loss": 0.0556, "step": 50990 }, { "grad_norm": 0.39047810435295105, "learning_rate": 5.2480839168304e-05, "loss": 0.0409, "step": 51000 }, { "grad_norm": 0.6603979468345642, "learning_rate": 5.246432470008773e-05, "loss": 0.0423, "step": 51010 }, { "grad_norm": 0.6167837381362915, "learning_rate": 5.2447809962376616e-05, "loss": 0.0475, "step": 51020 }, { "grad_norm": 0.5288113355636597, "learning_rate": 5.243129495697666e-05, "loss": 0.0365, "step": 51030 }, { "grad_norm": 0.9511082172393799, "learning_rate": 5.241477968569397e-05, "loss": 0.0623, "step": 51040 }, { "grad_norm": 0.5618813037872314, "learning_rate": 5.23982641503346e-05, "loss": 0.0443, "step": 51050 }, { "grad_norm": 0.5105093717575073, "learning_rate": 5.238174835270465e-05, "loss": 0.0441, "step": 51060 }, { "grad_norm": 0.4066494405269623, "learning_rate": 5.23652322946103e-05, "loss": 0.0375, "step": 51070 }, { "grad_norm": 0.44762787222862244, "learning_rate": 5.234871597785771e-05, "loss": 0.0372, "step": 51080 }, { "grad_norm": 0.8283163905143738, "learning_rate": 5.233219940425306e-05, "loss": 0.0529, "step": 51090 }, { "grad_norm": 1.204703688621521, "learning_rate": 5.2315682575602576e-05, "loss": 0.0732, "step": 51100 }, { "grad_norm": 0.49133995175361633, "learning_rate": 5.229916549371254e-05, "loss": 0.0417, "step": 51110 }, { "grad_norm": 0.3781753182411194, "learning_rate": 5.228264816038924e-05, "loss": 0.0469, "step": 51120 }, { "grad_norm": 0.6811866760253906, "learning_rate": 5.2266130577438935e-05, "loss": 0.0449, "step": 51130 }, { "grad_norm": 0.5489951372146606, "learning_rate": 5.224961274666802e-05, "loss": 0.0386, "step": 51140 }, { "grad_norm": 0.710637092590332, "learning_rate": 5.223309466988283e-05, "loss": 0.0472, "step": 51150 }, { "grad_norm": 0.44571077823638916, "learning_rate": 5.221657634888976e-05, "loss": 0.061, "step": 51160 }, { "grad_norm": 0.5188539624214172, "learning_rate": 5.2200057785495235e-05, "loss": 0.0717, "step": 51170 }, { "grad_norm": 0.8766458034515381, "learning_rate": 5.2183538981505706e-05, "loss": 0.0477, "step": 51180 }, { "grad_norm": 0.3412640690803528, "learning_rate": 5.216701993872762e-05, "loss": 0.0427, "step": 51190 }, { "grad_norm": 0.4759259819984436, "learning_rate": 5.215050065896752e-05, "loss": 0.0494, "step": 51200 }, { "grad_norm": 0.4190877377986908, "learning_rate": 5.2133981144031875e-05, "loss": 0.0617, "step": 51210 }, { "grad_norm": 0.8325234651565552, "learning_rate": 5.211746139572727e-05, "loss": 0.055, "step": 51220 }, { "grad_norm": 0.32145076990127563, "learning_rate": 5.210094141586026e-05, "loss": 0.0405, "step": 51230 }, { "grad_norm": 0.6811079382896423, "learning_rate": 5.2084421206237476e-05, "loss": 0.0588, "step": 51240 }, { "grad_norm": 0.506424069404602, "learning_rate": 5.20679007686655e-05, "loss": 0.0536, "step": 51250 }, { "grad_norm": 0.43688708543777466, "learning_rate": 5.205138010495102e-05, "loss": 0.0515, "step": 51260 }, { "grad_norm": 0.4656848609447479, "learning_rate": 5.20348592169007e-05, "loss": 0.046, "step": 51270 }, { "grad_norm": 0.7806984782218933, "learning_rate": 5.201833810632123e-05, "loss": 0.0423, "step": 51280 }, { "grad_norm": 0.5172237157821655, "learning_rate": 5.200181677501933e-05, "loss": 0.0525, "step": 51290 }, { "grad_norm": 0.4092831611633301, "learning_rate": 5.1985295224801777e-05, "loss": 0.0521, "step": 51300 }, { "grad_norm": 0.48152071237564087, "learning_rate": 5.196877345747531e-05, "loss": 0.0565, "step": 51310 }, { "grad_norm": 0.5587053894996643, "learning_rate": 5.195225147484674e-05, "loss": 0.0525, "step": 51320 }, { "grad_norm": 0.7702241539955139, "learning_rate": 5.193572927872288e-05, "loss": 0.0516, "step": 51330 }, { "grad_norm": 0.4929315745830536, "learning_rate": 5.191920687091059e-05, "loss": 0.0465, "step": 51340 }, { "grad_norm": 0.3271571099758148, "learning_rate": 5.1902684253216695e-05, "loss": 0.0423, "step": 51350 }, { "grad_norm": 0.4998215436935425, "learning_rate": 5.1886161427448134e-05, "loss": 0.0401, "step": 51360 }, { "grad_norm": 0.3553745150566101, "learning_rate": 5.186963839541179e-05, "loss": 0.0384, "step": 51370 }, { "grad_norm": 0.5171698331832886, "learning_rate": 5.185311515891459e-05, "loss": 0.0368, "step": 51380 }, { "grad_norm": 0.8279030919075012, "learning_rate": 5.1836591719763516e-05, "loss": 0.0416, "step": 51390 }, { "grad_norm": 0.3717763423919678, "learning_rate": 5.182006807976554e-05, "loss": 0.0511, "step": 51400 }, { "grad_norm": 0.3312853276729584, "learning_rate": 5.180354424072763e-05, "loss": 0.0376, "step": 51410 }, { "grad_norm": 0.5290469527244568, "learning_rate": 5.178702020445686e-05, "loss": 0.0523, "step": 51420 }, { "grad_norm": 0.5415771007537842, "learning_rate": 5.177049597276024e-05, "loss": 0.0512, "step": 51430 }, { "grad_norm": 0.4651666283607483, "learning_rate": 5.175397154744482e-05, "loss": 0.0467, "step": 51440 }, { "grad_norm": 0.4929555356502533, "learning_rate": 5.1737446930317736e-05, "loss": 0.0392, "step": 51450 }, { "grad_norm": 0.6422504782676697, "learning_rate": 5.172092212318604e-05, "loss": 0.0517, "step": 51460 }, { "grad_norm": 0.4282398223876953, "learning_rate": 5.170439712785692e-05, "loss": 0.0504, "step": 51470 }, { "grad_norm": 0.5015827417373657, "learning_rate": 5.1687871946137467e-05, "loss": 0.0479, "step": 51480 }, { "grad_norm": 0.3152088224887848, "learning_rate": 5.16713465798349e-05, "loss": 0.0444, "step": 51490 }, { "grad_norm": 0.5591263771057129, "learning_rate": 5.165482103075637e-05, "loss": 0.0586, "step": 51500 }, { "grad_norm": 0.42751801013946533, "learning_rate": 5.163829530070909e-05, "loss": 0.0393, "step": 51510 }, { "grad_norm": 0.5060638189315796, "learning_rate": 5.162176939150032e-05, "loss": 0.0435, "step": 51520 }, { "grad_norm": 1.122382640838623, "learning_rate": 5.160524330493728e-05, "loss": 0.0495, "step": 51530 }, { "grad_norm": 0.3356051445007324, "learning_rate": 5.158871704282724e-05, "loss": 0.0437, "step": 51540 }, { "grad_norm": 0.6512438058853149, "learning_rate": 5.15721906069775e-05, "loss": 0.0538, "step": 51550 }, { "grad_norm": 0.4418378472328186, "learning_rate": 5.155566399919536e-05, "loss": 0.0405, "step": 51560 }, { "grad_norm": 0.5681630373001099, "learning_rate": 5.153913722128813e-05, "loss": 0.0436, "step": 51570 }, { "grad_norm": 0.7643696069717407, "learning_rate": 5.1522610275063176e-05, "loss": 0.0489, "step": 51580 }, { "grad_norm": 0.3864164650440216, "learning_rate": 5.150608316232786e-05, "loss": 0.0402, "step": 51590 }, { "grad_norm": 0.6263785362243652, "learning_rate": 5.148955588488955e-05, "loss": 0.0365, "step": 51600 }, { "grad_norm": 0.44293609261512756, "learning_rate": 5.147302844455564e-05, "loss": 0.0394, "step": 51610 }, { "grad_norm": 0.44454681873321533, "learning_rate": 5.145650084313356e-05, "loss": 0.0458, "step": 51620 }, { "grad_norm": 0.4914166033267975, "learning_rate": 5.1439973082430734e-05, "loss": 0.0446, "step": 51630 }, { "grad_norm": 0.5772936344146729, "learning_rate": 5.142344516425461e-05, "loss": 0.0384, "step": 51640 }, { "grad_norm": 0.2935831546783447, "learning_rate": 5.1406917090412673e-05, "loss": 0.0525, "step": 51650 }, { "grad_norm": 0.4563322961330414, "learning_rate": 5.1390388862712404e-05, "loss": 0.0398, "step": 51660 }, { "grad_norm": 0.5850939750671387, "learning_rate": 5.137386048296129e-05, "loss": 0.0437, "step": 51670 }, { "grad_norm": 0.46462926268577576, "learning_rate": 5.135733195296686e-05, "loss": 0.0508, "step": 51680 }, { "grad_norm": 0.35713469982147217, "learning_rate": 5.134080327453664e-05, "loss": 0.0431, "step": 51690 }, { "grad_norm": 0.7079458832740784, "learning_rate": 5.1324274449478204e-05, "loss": 0.0453, "step": 51700 }, { "grad_norm": 0.31880101561546326, "learning_rate": 5.130774547959909e-05, "loss": 0.0459, "step": 51710 }, { "grad_norm": 0.3861430585384369, "learning_rate": 5.129121636670692e-05, "loss": 0.0419, "step": 51720 }, { "grad_norm": 0.21028077602386475, "learning_rate": 5.1274687112609246e-05, "loss": 0.0279, "step": 51730 }, { "grad_norm": 0.39554816484451294, "learning_rate": 5.125815771911372e-05, "loss": 0.0421, "step": 51740 }, { "grad_norm": 0.8099057674407959, "learning_rate": 5.1241628188027956e-05, "loss": 0.039, "step": 51750 }, { "grad_norm": 0.6304005980491638, "learning_rate": 5.1225098521159585e-05, "loss": 0.0386, "step": 51760 }, { "grad_norm": 0.7392862439155579, "learning_rate": 5.12085687203163e-05, "loss": 0.0396, "step": 51770 }, { "grad_norm": 0.48256057500839233, "learning_rate": 5.119203878730575e-05, "loss": 0.0461, "step": 51780 }, { "grad_norm": 0.49033844470977783, "learning_rate": 5.117550872393564e-05, "loss": 0.0484, "step": 51790 }, { "grad_norm": 0.46729743480682373, "learning_rate": 5.1158978532013645e-05, "loss": 0.0536, "step": 51800 }, { "grad_norm": 0.39775893092155457, "learning_rate": 5.114244821334751e-05, "loss": 0.049, "step": 51810 }, { "grad_norm": 0.24550461769104004, "learning_rate": 5.112591776974497e-05, "loss": 0.0402, "step": 51820 }, { "grad_norm": 0.3942744731903076, "learning_rate": 5.110938720301374e-05, "loss": 0.0384, "step": 51830 }, { "grad_norm": 0.26225417852401733, "learning_rate": 5.1092856514961595e-05, "loss": 0.0454, "step": 51840 }, { "grad_norm": 0.4382496178150177, "learning_rate": 5.107632570739631e-05, "loss": 0.0438, "step": 51850 }, { "grad_norm": 0.379434198141098, "learning_rate": 5.1059794782125645e-05, "loss": 0.0538, "step": 51860 }, { "grad_norm": 0.8104698657989502, "learning_rate": 5.104326374095744e-05, "loss": 0.0562, "step": 51870 }, { "grad_norm": 0.32841962575912476, "learning_rate": 5.102673258569947e-05, "loss": 0.0446, "step": 51880 }, { "grad_norm": 0.49063840508461, "learning_rate": 5.101020131815956e-05, "loss": 0.051, "step": 51890 }, { "grad_norm": 0.6984083652496338, "learning_rate": 5.099366994014556e-05, "loss": 0.0427, "step": 51900 }, { "grad_norm": 0.5479761958122253, "learning_rate": 5.097713845346532e-05, "loss": 0.046, "step": 51910 }, { "grad_norm": 0.6219354271888733, "learning_rate": 5.096060685992668e-05, "loss": 0.0353, "step": 51920 }, { "grad_norm": 0.5602088570594788, "learning_rate": 5.0944075161337514e-05, "loss": 0.0573, "step": 51930 }, { "grad_norm": 0.5516945123672485, "learning_rate": 5.09275433595057e-05, "loss": 0.0453, "step": 51940 }, { "grad_norm": 1.5280406475067139, "learning_rate": 5.0911011456239157e-05, "loss": 0.048, "step": 51950 }, { "grad_norm": 0.6644101738929749, "learning_rate": 5.089447945334574e-05, "loss": 0.045, "step": 51960 }, { "grad_norm": 0.453324556350708, "learning_rate": 5.087794735263343e-05, "loss": 0.0389, "step": 51970 }, { "grad_norm": 0.3738752007484436, "learning_rate": 5.086141515591011e-05, "loss": 0.0458, "step": 51980 }, { "grad_norm": 0.49821004271507263, "learning_rate": 5.084488286498371e-05, "loss": 0.0417, "step": 51990 }, { "grad_norm": 0.42062851786613464, "learning_rate": 5.08283504816622e-05, "loss": 0.0611, "step": 52000 }, { "grad_norm": 0.46483999490737915, "learning_rate": 5.081181800775353e-05, "loss": 0.0487, "step": 52010 }, { "grad_norm": 0.3475448489189148, "learning_rate": 5.079528544506565e-05, "loss": 0.0463, "step": 52020 }, { "grad_norm": 0.4656320810317993, "learning_rate": 5.077875279540657e-05, "loss": 0.0398, "step": 52030 }, { "grad_norm": 0.31590911746025085, "learning_rate": 5.076222006058424e-05, "loss": 0.0476, "step": 52040 }, { "grad_norm": 0.5372300744056702, "learning_rate": 5.074568724240667e-05, "loss": 0.0338, "step": 52050 }, { "grad_norm": 0.41090336441993713, "learning_rate": 5.072915434268187e-05, "loss": 0.043, "step": 52060 }, { "grad_norm": 0.35806038975715637, "learning_rate": 5.071262136321785e-05, "loss": 0.0406, "step": 52070 }, { "grad_norm": 0.4528855085372925, "learning_rate": 5.069608830582262e-05, "loss": 0.0511, "step": 52080 }, { "grad_norm": 0.232716366648674, "learning_rate": 5.067955517230423e-05, "loss": 0.0468, "step": 52090 }, { "grad_norm": 0.3848732113838196, "learning_rate": 5.06630219644707e-05, "loss": 0.0521, "step": 52100 }, { "grad_norm": 0.5380024313926697, "learning_rate": 5.064648868413008e-05, "loss": 0.0336, "step": 52110 }, { "grad_norm": 0.43149882555007935, "learning_rate": 5.062995533309044e-05, "loss": 0.0415, "step": 52120 }, { "grad_norm": 0.46005168557167053, "learning_rate": 5.061342191315983e-05, "loss": 0.0578, "step": 52130 }, { "grad_norm": 0.6030945181846619, "learning_rate": 5.0596888426146325e-05, "loss": 0.0441, "step": 52140 }, { "grad_norm": 0.5383235216140747, "learning_rate": 5.058035487385798e-05, "loss": 0.0402, "step": 52150 }, { "grad_norm": 1.154018759727478, "learning_rate": 5.0563821258102915e-05, "loss": 0.046, "step": 52160 }, { "grad_norm": 1.3271980285644531, "learning_rate": 5.05472875806892e-05, "loss": 0.0559, "step": 52170 }, { "grad_norm": 1.3344502449035645, "learning_rate": 5.053075384342494e-05, "loss": 0.0508, "step": 52180 }, { "grad_norm": 0.42450442910194397, "learning_rate": 5.0514220048118234e-05, "loss": 0.0514, "step": 52190 }, { "grad_norm": 1.0412086248397827, "learning_rate": 5.049768619657721e-05, "loss": 0.0561, "step": 52200 }, { "grad_norm": 0.3342089354991913, "learning_rate": 5.048115229060996e-05, "loss": 0.0421, "step": 52210 }, { "grad_norm": 0.849395751953125, "learning_rate": 5.046461833202462e-05, "loss": 0.0488, "step": 52220 }, { "grad_norm": 0.28714632987976074, "learning_rate": 5.044808432262932e-05, "loss": 0.051, "step": 52230 }, { "grad_norm": 0.36225825548171997, "learning_rate": 5.043155026423219e-05, "loss": 0.0435, "step": 52240 }, { "grad_norm": 0.5052968263626099, "learning_rate": 5.041501615864137e-05, "loss": 0.0395, "step": 52250 }, { "grad_norm": 0.3006386160850525, "learning_rate": 5.039848200766502e-05, "loss": 0.0342, "step": 52260 }, { "grad_norm": 0.5830371975898743, "learning_rate": 5.038194781311125e-05, "loss": 0.0498, "step": 52270 }, { "grad_norm": 0.4332548975944519, "learning_rate": 5.0365413576788264e-05, "loss": 0.0339, "step": 52280 }, { "grad_norm": 0.5158843994140625, "learning_rate": 5.0348879300504194e-05, "loss": 0.0412, "step": 52290 }, { "grad_norm": 0.3849404752254486, "learning_rate": 5.033234498606721e-05, "loss": 0.0403, "step": 52300 }, { "grad_norm": 0.65132075548172, "learning_rate": 5.0315810635285476e-05, "loss": 0.0563, "step": 52310 }, { "grad_norm": 0.4902178645133972, "learning_rate": 5.0299276249967164e-05, "loss": 0.0451, "step": 52320 }, { "grad_norm": 0.5845174789428711, "learning_rate": 5.0282741831920454e-05, "loss": 0.0543, "step": 52330 }, { "grad_norm": 0.5366929769515991, "learning_rate": 5.026620738295351e-05, "loss": 0.043, "step": 52340 }, { "grad_norm": 0.4094056189060211, "learning_rate": 5.024967290487455e-05, "loss": 0.0386, "step": 52350 }, { "grad_norm": 0.5805431604385376, "learning_rate": 5.023313839949172e-05, "loss": 0.0479, "step": 52360 }, { "grad_norm": 0.5566660165786743, "learning_rate": 5.0216603868613224e-05, "loss": 0.0463, "step": 52370 }, { "grad_norm": 0.6353172063827515, "learning_rate": 5.020006931404726e-05, "loss": 0.0531, "step": 52380 }, { "grad_norm": 0.521867573261261, "learning_rate": 5.0183534737602014e-05, "loss": 0.0325, "step": 52390 }, { "grad_norm": 0.4943092465400696, "learning_rate": 5.0167000141085686e-05, "loss": 0.038, "step": 52400 }, { "grad_norm": 0.3530133366584778, "learning_rate": 5.015046552630647e-05, "loss": 0.0398, "step": 52410 }, { "grad_norm": 0.49821433424949646, "learning_rate": 5.013393089507258e-05, "loss": 0.0409, "step": 52420 }, { "grad_norm": 0.5833649039268494, "learning_rate": 5.011739624919219e-05, "loss": 0.0447, "step": 52430 }, { "grad_norm": 0.5208232998847961, "learning_rate": 5.0100861590473545e-05, "loss": 0.0496, "step": 52440 }, { "grad_norm": 0.5934762358665466, "learning_rate": 5.008432692072482e-05, "loss": 0.0356, "step": 52450 }, { "grad_norm": 1.2533912658691406, "learning_rate": 5.006779224175423e-05, "loss": 0.0667, "step": 52460 }, { "grad_norm": 0.5379055142402649, "learning_rate": 5.005125755536998e-05, "loss": 0.0504, "step": 52470 }, { "grad_norm": 1.1280500888824463, "learning_rate": 5.003472286338029e-05, "loss": 0.0531, "step": 52480 }, { "grad_norm": 0.3338945806026459, "learning_rate": 5.001818816759335e-05, "loss": 0.0606, "step": 52490 }, { "grad_norm": 0.5814914703369141, "learning_rate": 5.000165346981738e-05, "loss": 0.0446, "step": 52500 }, { "grad_norm": 0.6036858558654785, "learning_rate": 4.9985118771860595e-05, "loss": 0.0418, "step": 52510 }, { "grad_norm": 0.8871707320213318, "learning_rate": 4.996858407553119e-05, "loss": 0.0417, "step": 52520 }, { "grad_norm": 0.3243319094181061, "learning_rate": 4.99520493826374e-05, "loss": 0.0487, "step": 52530 }, { "grad_norm": 0.4949703812599182, "learning_rate": 4.9935514694987384e-05, "loss": 0.0555, "step": 52540 }, { "grad_norm": 0.6712106466293335, "learning_rate": 4.991898001438941e-05, "loss": 0.0429, "step": 52550 }, { "grad_norm": 1.6503803730010986, "learning_rate": 4.990244534265166e-05, "loss": 0.0445, "step": 52560 }, { "grad_norm": 0.7378571629524231, "learning_rate": 4.988591068158234e-05, "loss": 0.0539, "step": 52570 }, { "grad_norm": 0.5659905672073364, "learning_rate": 4.986937603298965e-05, "loss": 0.0449, "step": 52580 }, { "grad_norm": 0.5689627528190613, "learning_rate": 4.985284139868181e-05, "loss": 0.0482, "step": 52590 }, { "grad_norm": 0.415208637714386, "learning_rate": 4.9836306780467e-05, "loss": 0.045, "step": 52600 }, { "grad_norm": 0.7054372429847717, "learning_rate": 4.981977218015343e-05, "loss": 0.0438, "step": 52610 }, { "grad_norm": 0.62521892786026, "learning_rate": 4.980323759954932e-05, "loss": 0.0459, "step": 52620 }, { "grad_norm": 0.5210055112838745, "learning_rate": 4.978670304046284e-05, "loss": 0.0436, "step": 52630 }, { "grad_norm": 0.4859367907047272, "learning_rate": 4.9770168504702195e-05, "loss": 0.0438, "step": 52640 }, { "grad_norm": 0.6210041046142578, "learning_rate": 4.975363399407559e-05, "loss": 0.0541, "step": 52650 }, { "grad_norm": 0.8004753589630127, "learning_rate": 4.9737099510391194e-05, "loss": 0.0505, "step": 52660 }, { "grad_norm": 0.42213642597198486, "learning_rate": 4.972056505545719e-05, "loss": 0.0473, "step": 52670 }, { "grad_norm": 0.49680763483047485, "learning_rate": 4.9704030631081775e-05, "loss": 0.039, "step": 52680 }, { "grad_norm": 0.3265058994293213, "learning_rate": 4.968749623907313e-05, "loss": 0.033, "step": 52690 }, { "grad_norm": 0.41713666915893555, "learning_rate": 4.967096188123943e-05, "loss": 0.0498, "step": 52700 }, { "grad_norm": 0.5775282979011536, "learning_rate": 4.965442755938884e-05, "loss": 0.0424, "step": 52710 }, { "grad_norm": 0.5177456140518188, "learning_rate": 4.963789327532954e-05, "loss": 0.0473, "step": 52720 }, { "grad_norm": 0.6046038866043091, "learning_rate": 4.962135903086968e-05, "loss": 0.0507, "step": 52730 }, { "grad_norm": 0.45843327045440674, "learning_rate": 4.960482482781741e-05, "loss": 0.0409, "step": 52740 }, { "grad_norm": 0.5522348284721375, "learning_rate": 4.958829066798092e-05, "loss": 0.0486, "step": 52750 }, { "grad_norm": 0.43148958683013916, "learning_rate": 4.957175655316835e-05, "loss": 0.0465, "step": 52760 }, { "grad_norm": 0.43899428844451904, "learning_rate": 4.9555222485187834e-05, "loss": 0.0474, "step": 52770 }, { "grad_norm": 0.3908233046531677, "learning_rate": 4.953868846584752e-05, "loss": 0.0389, "step": 52780 }, { "grad_norm": 0.5018739700317383, "learning_rate": 4.9522154496955544e-05, "loss": 0.0457, "step": 52790 }, { "grad_norm": 0.3464469313621521, "learning_rate": 4.950562058032001e-05, "loss": 0.0531, "step": 52800 }, { "grad_norm": 0.5617718696594238, "learning_rate": 4.948908671774909e-05, "loss": 0.0515, "step": 52810 }, { "grad_norm": 0.45278823375701904, "learning_rate": 4.947255291105087e-05, "loss": 0.0478, "step": 52820 }, { "grad_norm": 0.6049468517303467, "learning_rate": 4.945601916203348e-05, "loss": 0.0508, "step": 52830 }, { "grad_norm": 0.45557066798210144, "learning_rate": 4.9439485472505016e-05, "loss": 0.0458, "step": 52840 }, { "grad_norm": 1.9841288328170776, "learning_rate": 4.942295184427356e-05, "loss": 0.0511, "step": 52850 }, { "grad_norm": 0.5272172093391418, "learning_rate": 4.9406418279147234e-05, "loss": 0.043, "step": 52860 }, { "grad_norm": 0.604644238948822, "learning_rate": 4.9389884778934116e-05, "loss": 0.0445, "step": 52870 }, { "grad_norm": 0.6440103650093079, "learning_rate": 4.937335134544228e-05, "loss": 0.0518, "step": 52880 }, { "grad_norm": 0.3113623559474945, "learning_rate": 4.935681798047979e-05, "loss": 0.0572, "step": 52890 }, { "grad_norm": 0.3520124852657318, "learning_rate": 4.934028468585473e-05, "loss": 0.0468, "step": 52900 }, { "grad_norm": 0.5198375582695007, "learning_rate": 4.932375146337514e-05, "loss": 0.0512, "step": 52910 }, { "grad_norm": 0.4067387282848358, "learning_rate": 4.930721831484908e-05, "loss": 0.0457, "step": 52920 }, { "grad_norm": 0.7573917508125305, "learning_rate": 4.929068524208456e-05, "loss": 0.0514, "step": 52930 }, { "grad_norm": 0.4207363426685333, "learning_rate": 4.927415224688965e-05, "loss": 0.0506, "step": 52940 }, { "grad_norm": 0.34591442346572876, "learning_rate": 4.9257619331072375e-05, "loss": 0.048, "step": 52950 }, { "grad_norm": 0.7089989185333252, "learning_rate": 4.924108649644072e-05, "loss": 0.0524, "step": 52960 }, { "grad_norm": 0.40176278352737427, "learning_rate": 4.922455374480271e-05, "loss": 0.0576, "step": 52970 }, { "grad_norm": 0.684051513671875, "learning_rate": 4.920802107796634e-05, "loss": 0.0475, "step": 52980 }, { "grad_norm": 0.6541571021080017, "learning_rate": 4.9191488497739564e-05, "loss": 0.0561, "step": 52990 }, { "grad_norm": 0.6021079421043396, "learning_rate": 4.917495600593042e-05, "loss": 0.0516, "step": 53000 }, { "grad_norm": 0.47683748602867126, "learning_rate": 4.915842360434684e-05, "loss": 0.0443, "step": 53010 }, { "grad_norm": 1.1153922080993652, "learning_rate": 4.914189129479679e-05, "loss": 0.0564, "step": 53020 }, { "grad_norm": 0.3586083948612213, "learning_rate": 4.912535907908822e-05, "loss": 0.0468, "step": 53030 }, { "grad_norm": 0.8006194829940796, "learning_rate": 4.910882695902907e-05, "loss": 0.0532, "step": 53040 }, { "grad_norm": 0.6183185577392578, "learning_rate": 4.9092294936427234e-05, "loss": 0.0447, "step": 53050 }, { "grad_norm": 0.43206724524497986, "learning_rate": 4.9075763013090685e-05, "loss": 0.0429, "step": 53060 }, { "grad_norm": 0.538983166217804, "learning_rate": 4.9059231190827305e-05, "loss": 0.0382, "step": 53070 }, { "grad_norm": 0.4734025299549103, "learning_rate": 4.904269947144499e-05, "loss": 0.0543, "step": 53080 }, { "grad_norm": 0.6289510130882263, "learning_rate": 4.902616785675161e-05, "loss": 0.0503, "step": 53090 }, { "grad_norm": 0.46622344851493835, "learning_rate": 4.900963634855505e-05, "loss": 0.0497, "step": 53100 }, { "grad_norm": 0.45476752519607544, "learning_rate": 4.8993104948663173e-05, "loss": 0.0347, "step": 53110 }, { "grad_norm": 0.41491246223449707, "learning_rate": 4.8976573658883823e-05, "loss": 0.0399, "step": 53120 }, { "grad_norm": 0.486449271440506, "learning_rate": 4.896004248102484e-05, "loss": 0.0464, "step": 53130 }, { "grad_norm": 0.46419262886047363, "learning_rate": 4.8943511416894045e-05, "loss": 0.053, "step": 53140 }, { "grad_norm": 0.5038174986839294, "learning_rate": 4.892698046829926e-05, "loss": 0.0349, "step": 53150 }, { "grad_norm": 1.0636693239212036, "learning_rate": 4.8910449637048275e-05, "loss": 0.053, "step": 53160 }, { "grad_norm": 0.5727435350418091, "learning_rate": 4.8893918924948884e-05, "loss": 0.0465, "step": 53170 }, { "grad_norm": 0.5150883793830872, "learning_rate": 4.8877388333808835e-05, "loss": 0.0482, "step": 53180 }, { "grad_norm": 0.3610430955886841, "learning_rate": 4.8860857865435925e-05, "loss": 0.0442, "step": 53190 }, { "grad_norm": 0.4526786506175995, "learning_rate": 4.884432752163789e-05, "loss": 0.0445, "step": 53200 }, { "grad_norm": 0.4352532625198364, "learning_rate": 4.8827797304222466e-05, "loss": 0.0438, "step": 53210 }, { "grad_norm": 1.1028136014938354, "learning_rate": 4.8811267214997366e-05, "loss": 0.0415, "step": 53220 }, { "grad_norm": 0.4359525442123413, "learning_rate": 4.879473725577029e-05, "loss": 0.0521, "step": 53230 }, { "grad_norm": 0.5788222551345825, "learning_rate": 4.877820742834891e-05, "loss": 0.0404, "step": 53240 }, { "grad_norm": 0.5913876295089722, "learning_rate": 4.8761677734540956e-05, "loss": 0.0458, "step": 53250 }, { "grad_norm": 0.8564956188201904, "learning_rate": 4.874514817615407e-05, "loss": 0.0444, "step": 53260 }, { "grad_norm": 1.3639099597930908, "learning_rate": 4.872861875499588e-05, "loss": 0.054, "step": 53270 }, { "grad_norm": 0.3911714553833008, "learning_rate": 4.871208947287404e-05, "loss": 0.0494, "step": 53280 }, { "grad_norm": 0.770940363407135, "learning_rate": 4.869556033159615e-05, "loss": 0.0556, "step": 53290 }, { "grad_norm": 0.4116729199886322, "learning_rate": 4.8679031332969835e-05, "loss": 0.0405, "step": 53300 }, { "grad_norm": 0.42497438192367554, "learning_rate": 4.8662502478802635e-05, "loss": 0.0662, "step": 53310 }, { "grad_norm": 0.5247918963432312, "learning_rate": 4.8645973770902176e-05, "loss": 0.0413, "step": 53320 }, { "grad_norm": 0.48706722259521484, "learning_rate": 4.862944521107599e-05, "loss": 0.0393, "step": 53330 }, { "grad_norm": 0.9126066565513611, "learning_rate": 4.861291680113162e-05, "loss": 0.0501, "step": 53340 }, { "grad_norm": 0.465613454580307, "learning_rate": 4.8596388542876563e-05, "loss": 0.0436, "step": 53350 }, { "grad_norm": 0.6063438057899475, "learning_rate": 4.857986043811836e-05, "loss": 0.0487, "step": 53360 }, { "grad_norm": 0.5163609385490417, "learning_rate": 4.856333248866448e-05, "loss": 0.0427, "step": 53370 }, { "grad_norm": 0.797654926776886, "learning_rate": 4.85468046963224e-05, "loss": 0.0405, "step": 53380 }, { "grad_norm": 0.5181553959846497, "learning_rate": 4.853027706289957e-05, "loss": 0.0385, "step": 53390 }, { "grad_norm": 0.5019146800041199, "learning_rate": 4.851374959020344e-05, "loss": 0.0402, "step": 53400 }, { "grad_norm": 0.362856924533844, "learning_rate": 4.849722228004141e-05, "loss": 0.0409, "step": 53410 }, { "grad_norm": 1.0674209594726562, "learning_rate": 4.848069513422091e-05, "loss": 0.0541, "step": 53420 }, { "grad_norm": 0.39719635248184204, "learning_rate": 4.8464168154549294e-05, "loss": 0.0449, "step": 53430 }, { "grad_norm": 0.8136705160140991, "learning_rate": 4.844764134283392e-05, "loss": 0.0454, "step": 53440 }, { "grad_norm": 3.2758188247680664, "learning_rate": 4.843111470088218e-05, "loss": 0.0596, "step": 53450 }, { "grad_norm": 0.7182644009590149, "learning_rate": 4.841458823050137e-05, "loss": 0.0428, "step": 53460 }, { "grad_norm": 0.7694621086120605, "learning_rate": 4.8398061933498816e-05, "loss": 0.0392, "step": 53470 }, { "grad_norm": 0.7505732178688049, "learning_rate": 4.83815358116818e-05, "loss": 0.0402, "step": 53480 }, { "grad_norm": 0.49238911271095276, "learning_rate": 4.8365009866857584e-05, "loss": 0.0433, "step": 53490 }, { "grad_norm": 1.0391449928283691, "learning_rate": 4.834848410083342e-05, "loss": 0.0404, "step": 53500 }, { "grad_norm": 1.4204462766647339, "learning_rate": 4.833195851541657e-05, "loss": 0.0559, "step": 53510 }, { "grad_norm": 0.3446641266345978, "learning_rate": 4.831543311241423e-05, "loss": 0.0449, "step": 53520 }, { "grad_norm": 0.7212533354759216, "learning_rate": 4.829890789363359e-05, "loss": 0.0615, "step": 53530 }, { "grad_norm": 0.4075397551059723, "learning_rate": 4.828238286088182e-05, "loss": 0.0442, "step": 53540 }, { "grad_norm": 0.3683910667896271, "learning_rate": 4.826585801596609e-05, "loss": 0.042, "step": 53550 }, { "grad_norm": 0.5662336945533752, "learning_rate": 4.8249333360693486e-05, "loss": 0.0426, "step": 53560 }, { "grad_norm": 0.44021156430244446, "learning_rate": 4.8232808896871185e-05, "loss": 0.0379, "step": 53570 }, { "grad_norm": 0.38929814100265503, "learning_rate": 4.821628462630624e-05, "loss": 0.0409, "step": 53580 }, { "grad_norm": 0.5128087997436523, "learning_rate": 4.8199760550805735e-05, "loss": 0.0374, "step": 53590 }, { "grad_norm": 0.7352339625358582, "learning_rate": 4.818323667217669e-05, "loss": 0.0448, "step": 53600 }, { "grad_norm": 0.5391306281089783, "learning_rate": 4.8166712992226165e-05, "loss": 0.0452, "step": 53610 }, { "grad_norm": 0.5225163102149963, "learning_rate": 4.8150189512761154e-05, "loss": 0.0503, "step": 53620 }, { "grad_norm": 2.1982357501983643, "learning_rate": 4.813366623558862e-05, "loss": 0.0474, "step": 53630 }, { "grad_norm": 0.5737112164497375, "learning_rate": 4.811714316251554e-05, "loss": 0.0435, "step": 53640 }, { "grad_norm": 0.8690634965896606, "learning_rate": 4.810062029534886e-05, "loss": 0.0418, "step": 53650 }, { "grad_norm": 0.4315897226333618, "learning_rate": 4.8084097635895505e-05, "loss": 0.0467, "step": 53660 }, { "grad_norm": 0.5705453157424927, "learning_rate": 4.8067575185962335e-05, "loss": 0.0442, "step": 53670 }, { "grad_norm": 0.4339613616466522, "learning_rate": 4.805105294735625e-05, "loss": 0.0419, "step": 53680 }, { "grad_norm": 0.5326846241950989, "learning_rate": 4.803453092188406e-05, "loss": 0.0437, "step": 53690 }, { "grad_norm": 0.7843128442764282, "learning_rate": 4.801800911135263e-05, "loss": 0.0463, "step": 53700 }, { "grad_norm": 0.6738852262496948, "learning_rate": 4.8001487517568744e-05, "loss": 0.0513, "step": 53710 }, { "grad_norm": 0.553874135017395, "learning_rate": 4.798496614233918e-05, "loss": 0.055, "step": 53720 }, { "grad_norm": 0.41456353664398193, "learning_rate": 4.796844498747069e-05, "loss": 0.0592, "step": 53730 }, { "grad_norm": 0.5851441025733948, "learning_rate": 4.795192405477e-05, "loss": 0.0404, "step": 53740 }, { "grad_norm": 0.7530317306518555, "learning_rate": 4.79354033460438e-05, "loss": 0.0515, "step": 53750 }, { "grad_norm": 0.4183143675327301, "learning_rate": 4.79188828630988e-05, "loss": 0.0488, "step": 53760 }, { "grad_norm": 0.568049430847168, "learning_rate": 4.790236260774165e-05, "loss": 0.044, "step": 53770 }, { "grad_norm": 0.46788468956947327, "learning_rate": 4.788584258177896e-05, "loss": 0.0446, "step": 53780 }, { "grad_norm": 0.46069401502609253, "learning_rate": 4.7869322787017356e-05, "loss": 0.0396, "step": 53790 }, { "grad_norm": 0.4480098783969879, "learning_rate": 4.78528032252634e-05, "loss": 0.0485, "step": 53800 }, { "grad_norm": 0.4589523673057556, "learning_rate": 4.783628389832366e-05, "loss": 0.0433, "step": 53810 }, { "grad_norm": 1.0519095659255981, "learning_rate": 4.7819764808004644e-05, "loss": 0.0598, "step": 53820 }, { "grad_norm": 0.6506483554840088, "learning_rate": 4.780324595611289e-05, "loss": 0.0449, "step": 53830 }, { "grad_norm": 0.4809948801994324, "learning_rate": 4.7786727344454854e-05, "loss": 0.0529, "step": 53840 }, { "grad_norm": 0.5846717953681946, "learning_rate": 4.777020897483697e-05, "loss": 0.0511, "step": 53850 }, { "grad_norm": 0.4555545151233673, "learning_rate": 4.775369084906569e-05, "loss": 0.0368, "step": 53860 }, { "grad_norm": 0.4681496322154999, "learning_rate": 4.773717296894742e-05, "loss": 0.0538, "step": 53870 }, { "grad_norm": 0.3994450867176056, "learning_rate": 4.772065533628848e-05, "loss": 0.0493, "step": 53880 }, { "grad_norm": 0.43002283573150635, "learning_rate": 4.7704137952895265e-05, "loss": 0.0449, "step": 53890 }, { "grad_norm": 0.5308976769447327, "learning_rate": 4.768762082057407e-05, "loss": 0.0321, "step": 53900 }, { "grad_norm": 0.41687849164009094, "learning_rate": 4.7671103941131186e-05, "loss": 0.0434, "step": 53910 }, { "grad_norm": 0.3960007131099701, "learning_rate": 4.765458731637287e-05, "loss": 0.0424, "step": 53920 }, { "grad_norm": 0.4007890224456787, "learning_rate": 4.763807094810537e-05, "loss": 0.0517, "step": 53930 }, { "grad_norm": 0.6237417459487915, "learning_rate": 4.762155483813487e-05, "loss": 0.0551, "step": 53940 }, { "grad_norm": 0.5995973944664001, "learning_rate": 4.760503898826754e-05, "loss": 0.0561, "step": 53950 }, { "grad_norm": 0.6662456393241882, "learning_rate": 4.758852340030957e-05, "loss": 0.0546, "step": 53960 }, { "grad_norm": 0.770463764667511, "learning_rate": 4.757200807606705e-05, "loss": 0.0467, "step": 53970 }, { "grad_norm": 0.7805770039558411, "learning_rate": 4.755549301734607e-05, "loss": 0.0435, "step": 53980 }, { "grad_norm": 0.39095190167427063, "learning_rate": 4.753897822595271e-05, "loss": 0.0612, "step": 53990 }, { "grad_norm": 0.855279803276062, "learning_rate": 4.7522463703692994e-05, "loss": 0.0407, "step": 54000 }, { "grad_norm": 0.32418861985206604, "learning_rate": 4.75059494523729e-05, "loss": 0.0395, "step": 54010 }, { "grad_norm": 0.3205382227897644, "learning_rate": 4.748943547379843e-05, "loss": 0.044, "step": 54020 }, { "grad_norm": 0.31580376625061035, "learning_rate": 4.7472921769775535e-05, "loss": 0.0545, "step": 54030 }, { "grad_norm": 0.406917929649353, "learning_rate": 4.745640834211011e-05, "loss": 0.0612, "step": 54040 }, { "grad_norm": 0.34323492646217346, "learning_rate": 4.7439895192608054e-05, "loss": 0.0434, "step": 54050 }, { "grad_norm": 0.4782826006412506, "learning_rate": 4.7423382323075186e-05, "loss": 0.0472, "step": 54060 }, { "grad_norm": 0.6591819524765015, "learning_rate": 4.740686973531737e-05, "loss": 0.0443, "step": 54070 }, { "grad_norm": 0.6715533137321472, "learning_rate": 4.7390357431140375e-05, "loss": 0.0488, "step": 54080 }, { "grad_norm": 0.4732391834259033, "learning_rate": 4.737384541234998e-05, "loss": 0.0534, "step": 54090 }, { "grad_norm": 0.6924636363983154, "learning_rate": 4.735733368075189e-05, "loss": 0.0345, "step": 54100 }, { "grad_norm": 0.5658735632896423, "learning_rate": 4.734082223815182e-05, "loss": 0.0457, "step": 54110 }, { "grad_norm": 0.38629764318466187, "learning_rate": 4.7324311086355436e-05, "loss": 0.0444, "step": 54120 }, { "grad_norm": 0.40097951889038086, "learning_rate": 4.730780022716836e-05, "loss": 0.0435, "step": 54130 }, { "grad_norm": 0.5071962475776672, "learning_rate": 4.729128966239619e-05, "loss": 0.0376, "step": 54140 }, { "grad_norm": 0.6819288730621338, "learning_rate": 4.7274779393844536e-05, "loss": 0.037, "step": 54150 }, { "grad_norm": 1.074892520904541, "learning_rate": 4.725826942331891e-05, "loss": 0.0611, "step": 54160 }, { "grad_norm": 0.699609100818634, "learning_rate": 4.724175975262481e-05, "loss": 0.0506, "step": 54170 }, { "grad_norm": 0.5813665986061096, "learning_rate": 4.722525038356772e-05, "loss": 0.0408, "step": 54180 }, { "grad_norm": 0.47240084409713745, "learning_rate": 4.7208741317953075e-05, "loss": 0.0471, "step": 54190 }, { "grad_norm": 0.41508182883262634, "learning_rate": 4.7192232557586264e-05, "loss": 0.0482, "step": 54200 }, { "grad_norm": 0.8261206746101379, "learning_rate": 4.717572410427269e-05, "loss": 0.0443, "step": 54210 }, { "grad_norm": 0.7132400274276733, "learning_rate": 4.71592159598177e-05, "loss": 0.0494, "step": 54220 }, { "grad_norm": 0.4900439381599426, "learning_rate": 4.714270812602657e-05, "loss": 0.0379, "step": 54230 }, { "grad_norm": 0.34502163529396057, "learning_rate": 4.712620060470458e-05, "loss": 0.059, "step": 54240 }, { "grad_norm": 0.732952892780304, "learning_rate": 4.7109693397656986e-05, "loss": 0.0568, "step": 54250 }, { "grad_norm": 0.7123873829841614, "learning_rate": 4.709318650668894e-05, "loss": 0.0419, "step": 54260 }, { "grad_norm": 0.6469003558158875, "learning_rate": 4.7076679933605676e-05, "loss": 0.048, "step": 54270 }, { "grad_norm": 0.47581028938293457, "learning_rate": 4.70601736802123e-05, "loss": 0.0395, "step": 54280 }, { "grad_norm": 0.7781259417533875, "learning_rate": 4.704366774831391e-05, "loss": 0.0559, "step": 54290 }, { "grad_norm": 0.5611518025398254, "learning_rate": 4.702716213971557e-05, "loss": 0.0442, "step": 54300 }, { "grad_norm": 0.9813752770423889, "learning_rate": 4.70106568562223e-05, "loss": 0.0497, "step": 54310 }, { "grad_norm": 0.9891679286956787, "learning_rate": 4.6994151899639105e-05, "loss": 0.0453, "step": 54320 }, { "grad_norm": 0.5311829447746277, "learning_rate": 4.697764727177093e-05, "loss": 0.0661, "step": 54330 }, { "grad_norm": 0.587888240814209, "learning_rate": 4.696114297442272e-05, "loss": 0.048, "step": 54340 }, { "grad_norm": 0.4791751503944397, "learning_rate": 4.694463900939933e-05, "loss": 0.0383, "step": 54350 }, { "grad_norm": 0.4205947518348694, "learning_rate": 4.692813537850564e-05, "loss": 0.0446, "step": 54360 }, { "grad_norm": 0.2544257938861847, "learning_rate": 4.6911632083546445e-05, "loss": 0.034, "step": 54370 }, { "grad_norm": 0.4315017759799957, "learning_rate": 4.6895129126326525e-05, "loss": 0.0483, "step": 54380 }, { "grad_norm": 0.42303380370140076, "learning_rate": 4.687862650865059e-05, "loss": 0.041, "step": 54390 }, { "grad_norm": 0.46288952231407166, "learning_rate": 4.686212423232339e-05, "loss": 0.0355, "step": 54400 }, { "grad_norm": 0.41182664036750793, "learning_rate": 4.6845622299149565e-05, "loss": 0.0473, "step": 54410 }, { "grad_norm": 0.3959178328514099, "learning_rate": 4.682912071093374e-05, "loss": 0.0374, "step": 54420 }, { "grad_norm": 0.7453820705413818, "learning_rate": 4.681261946948052e-05, "loss": 0.0503, "step": 54430 }, { "grad_norm": 0.4799118638038635, "learning_rate": 4.679611857659443e-05, "loss": 0.0451, "step": 54440 }, { "grad_norm": 0.3296149969100952, "learning_rate": 4.6779618034079975e-05, "loss": 0.0487, "step": 54450 }, { "grad_norm": 0.516074538230896, "learning_rate": 4.6763117843741675e-05, "loss": 0.0524, "step": 54460 }, { "grad_norm": 0.5294572114944458, "learning_rate": 4.674661800738394e-05, "loss": 0.0419, "step": 54470 }, { "grad_norm": 0.40602508187294006, "learning_rate": 4.673011852681117e-05, "loss": 0.0402, "step": 54480 }, { "grad_norm": 0.36953404545783997, "learning_rate": 4.671361940382772e-05, "loss": 0.051, "step": 54490 }, { "grad_norm": 0.5113062262535095, "learning_rate": 4.669712064023791e-05, "loss": 0.0472, "step": 54500 }, { "grad_norm": 0.5328537821769714, "learning_rate": 4.668062223784602e-05, "loss": 0.0381, "step": 54510 }, { "grad_norm": 0.5036441087722778, "learning_rate": 4.6664124198456274e-05, "loss": 0.0455, "step": 54520 }, { "grad_norm": 0.4359135925769806, "learning_rate": 4.6647626523872905e-05, "loss": 0.0454, "step": 54530 }, { "grad_norm": 0.2984607517719269, "learning_rate": 4.6631129215900066e-05, "loss": 0.0443, "step": 54540 }, { "grad_norm": 0.6527952551841736, "learning_rate": 4.661463227634187e-05, "loss": 0.0653, "step": 54550 }, { "grad_norm": 0.5525022745132446, "learning_rate": 4.659813570700238e-05, "loss": 0.036, "step": 54560 }, { "grad_norm": 0.5768859386444092, "learning_rate": 4.658163950968568e-05, "loss": 0.0374, "step": 54570 }, { "grad_norm": 0.4982897639274597, "learning_rate": 4.6565143686195726e-05, "loss": 0.0473, "step": 54580 }, { "grad_norm": 0.6351163387298584, "learning_rate": 4.654864823833649e-05, "loss": 0.0398, "step": 54590 }, { "grad_norm": 0.563077986240387, "learning_rate": 4.65321531679119e-05, "loss": 0.0437, "step": 54600 }, { "grad_norm": 0.4166448712348938, "learning_rate": 4.6515658476725834e-05, "loss": 0.0518, "step": 54610 }, { "grad_norm": 0.5346288681030273, "learning_rate": 4.649916416658212e-05, "loss": 0.0495, "step": 54620 }, { "grad_norm": 0.8948080539703369, "learning_rate": 4.648267023928455e-05, "loss": 0.0466, "step": 54630 }, { "grad_norm": 0.6182506680488586, "learning_rate": 4.646617669663687e-05, "loss": 0.0396, "step": 54640 }, { "grad_norm": 0.5873215198516846, "learning_rate": 4.644968354044277e-05, "loss": 0.0384, "step": 54650 }, { "grad_norm": 0.8231598138809204, "learning_rate": 4.643319077250596e-05, "loss": 0.0485, "step": 54660 }, { "grad_norm": 0.8998904228210449, "learning_rate": 4.6416698394630055e-05, "loss": 0.0371, "step": 54670 }, { "grad_norm": 0.6547792553901672, "learning_rate": 4.640020640861862e-05, "loss": 0.0422, "step": 54680 }, { "grad_norm": 0.3756164610385895, "learning_rate": 4.6383714816275195e-05, "loss": 0.0429, "step": 54690 }, { "grad_norm": 0.413487046957016, "learning_rate": 4.636722361940329e-05, "loss": 0.0407, "step": 54700 }, { "grad_norm": 2.1406757831573486, "learning_rate": 4.635073281980632e-05, "loss": 0.0488, "step": 54710 }, { "grad_norm": 0.3415277600288391, "learning_rate": 4.633424241928775e-05, "loss": 0.055, "step": 54720 }, { "grad_norm": 0.4725012183189392, "learning_rate": 4.6317752419650914e-05, "loss": 0.0404, "step": 54730 }, { "grad_norm": 0.5414153933525085, "learning_rate": 4.630126282269914e-05, "loss": 0.0505, "step": 54740 }, { "grad_norm": 1.559499740600586, "learning_rate": 4.628477363023569e-05, "loss": 0.0412, "step": 54750 }, { "grad_norm": 0.4176596403121948, "learning_rate": 4.626828484406383e-05, "loss": 0.0439, "step": 54760 }, { "grad_norm": 0.537227988243103, "learning_rate": 4.625179646598669e-05, "loss": 0.0519, "step": 54770 }, { "grad_norm": 0.4989771544933319, "learning_rate": 4.6235308497807487e-05, "loss": 0.0438, "step": 54780 }, { "grad_norm": 0.6150333285331726, "learning_rate": 4.6218820941329276e-05, "loss": 0.0549, "step": 54790 }, { "grad_norm": 0.32141271233558655, "learning_rate": 4.620233379835513e-05, "loss": 0.0365, "step": 54800 }, { "grad_norm": 0.4521290361881256, "learning_rate": 4.618584707068804e-05, "loss": 0.0466, "step": 54810 }, { "grad_norm": 0.4225204586982727, "learning_rate": 4.616936076013099e-05, "loss": 0.0496, "step": 54820 }, { "grad_norm": 0.5400198698043823, "learning_rate": 4.615287486848688e-05, "loss": 0.0462, "step": 54830 }, { "grad_norm": 0.7948580384254456, "learning_rate": 4.6136389397558584e-05, "loss": 0.0537, "step": 54840 }, { "grad_norm": 0.6250433325767517, "learning_rate": 4.6119904349148935e-05, "loss": 0.0349, "step": 54850 }, { "grad_norm": 0.883251965045929, "learning_rate": 4.6103419725060726e-05, "loss": 0.0386, "step": 54860 }, { "grad_norm": 0.45150843262672424, "learning_rate": 4.608693552709667e-05, "loss": 0.0291, "step": 54870 }, { "grad_norm": 0.58339923620224, "learning_rate": 4.607045175705947e-05, "loss": 0.0447, "step": 54880 }, { "grad_norm": 0.49060890078544617, "learning_rate": 4.605396841675176e-05, "loss": 0.0401, "step": 54890 }, { "grad_norm": 0.571864664554596, "learning_rate": 4.6037485507976094e-05, "loss": 0.0455, "step": 54900 }, { "grad_norm": 0.38054317235946655, "learning_rate": 4.602100303253509e-05, "loss": 0.0432, "step": 54910 }, { "grad_norm": 0.4883393943309784, "learning_rate": 4.600452099223122e-05, "loss": 0.0588, "step": 54920 }, { "grad_norm": 0.33761143684387207, "learning_rate": 4.598803938886691e-05, "loss": 0.0398, "step": 54930 }, { "grad_norm": 0.4033181369304657, "learning_rate": 4.597155822424459e-05, "loss": 0.0373, "step": 54940 }, { "grad_norm": 0.42883795499801636, "learning_rate": 4.59550775001666e-05, "loss": 0.0499, "step": 54950 }, { "grad_norm": 0.3764042854309082, "learning_rate": 4.5938597218435234e-05, "loss": 0.0456, "step": 54960 }, { "grad_norm": 0.600566029548645, "learning_rate": 4.592211738085279e-05, "loss": 0.0604, "step": 54970 }, { "grad_norm": 0.5754021406173706, "learning_rate": 4.590563798922146e-05, "loss": 0.0366, "step": 54980 }, { "grad_norm": 0.4760842025279999, "learning_rate": 4.5889159045343404e-05, "loss": 0.0505, "step": 54990 }, { "grad_norm": 0.684790313243866, "learning_rate": 4.5872680551020734e-05, "loss": 0.0404, "step": 55000 }, { "grad_norm": 0.446686327457428, "learning_rate": 4.585620250805552e-05, "loss": 0.0342, "step": 55010 }, { "grad_norm": 0.7436028718948364, "learning_rate": 4.583972491824975e-05, "loss": 0.0378, "step": 55020 }, { "grad_norm": 0.840121865272522, "learning_rate": 4.582324778340541e-05, "loss": 0.0477, "step": 55030 }, { "grad_norm": 0.5050724744796753, "learning_rate": 4.5806771105324424e-05, "loss": 0.0502, "step": 55040 }, { "grad_norm": 0.7508479356765747, "learning_rate": 4.579029488580865e-05, "loss": 0.0478, "step": 55050 }, { "grad_norm": 0.3723532557487488, "learning_rate": 4.5773819126659876e-05, "loss": 0.0331, "step": 55060 }, { "grad_norm": 0.6905053853988647, "learning_rate": 4.575734382967991e-05, "loss": 0.0425, "step": 55070 }, { "grad_norm": 0.447725772857666, "learning_rate": 4.574086899667044e-05, "loss": 0.0426, "step": 55080 }, { "grad_norm": 0.7413299679756165, "learning_rate": 4.5724394629433106e-05, "loss": 0.0422, "step": 55090 }, { "grad_norm": 0.3366217315196991, "learning_rate": 4.570792072976958e-05, "loss": 0.0429, "step": 55100 }, { "grad_norm": 0.564115583896637, "learning_rate": 4.569144729948138e-05, "loss": 0.0504, "step": 55110 }, { "grad_norm": 0.4956109821796417, "learning_rate": 4.5674974340370033e-05, "loss": 0.0378, "step": 55120 }, { "grad_norm": 0.4468443691730499, "learning_rate": 4.565850185423699e-05, "loss": 0.0477, "step": 55130 }, { "grad_norm": 0.517864465713501, "learning_rate": 4.564202984288365e-05, "loss": 0.0499, "step": 55140 }, { "grad_norm": 0.8576467633247375, "learning_rate": 4.5625558308111356e-05, "loss": 0.0411, "step": 55150 }, { "grad_norm": 0.37825772166252136, "learning_rate": 4.560908725172145e-05, "loss": 0.0463, "step": 55160 }, { "grad_norm": 0.457916259765625, "learning_rate": 4.559261667551516e-05, "loss": 0.0479, "step": 55170 }, { "grad_norm": 0.3781305253505707, "learning_rate": 4.5576146581293685e-05, "loss": 0.0522, "step": 55180 }, { "grad_norm": 0.6810226440429688, "learning_rate": 4.5559676970858164e-05, "loss": 0.0481, "step": 55190 }, { "grad_norm": 0.661027193069458, "learning_rate": 4.5543207846009705e-05, "loss": 0.0514, "step": 55200 }, { "grad_norm": 0.5950120091438293, "learning_rate": 4.552673920854933e-05, "loss": 0.0568, "step": 55210 }, { "grad_norm": 0.7883816957473755, "learning_rate": 4.5510271060278006e-05, "loss": 0.043, "step": 55220 }, { "grad_norm": 0.4166015088558197, "learning_rate": 4.5493803402996704e-05, "loss": 0.0518, "step": 55230 }, { "grad_norm": 0.26747220754623413, "learning_rate": 4.5477336238506295e-05, "loss": 0.0421, "step": 55240 }, { "grad_norm": 0.39774906635284424, "learning_rate": 4.546086956860759e-05, "loss": 0.0442, "step": 55250 }, { "grad_norm": 0.6484290957450867, "learning_rate": 4.5444403395101374e-05, "loss": 0.0411, "step": 55260 }, { "grad_norm": 0.7823782563209534, "learning_rate": 4.542793771978834e-05, "loss": 0.0404, "step": 55270 }, { "grad_norm": 0.6431869268417358, "learning_rate": 4.541147254446916e-05, "loss": 0.0486, "step": 55280 }, { "grad_norm": 0.4523971676826477, "learning_rate": 4.539500787094445e-05, "loss": 0.0474, "step": 55290 }, { "grad_norm": 0.45985960960388184, "learning_rate": 4.537854370101475e-05, "loss": 0.0476, "step": 55300 }, { "grad_norm": 0.334567129611969, "learning_rate": 4.5362080036480566e-05, "loss": 0.0391, "step": 55310 }, { "grad_norm": 0.5044422745704651, "learning_rate": 4.534561687914235e-05, "loss": 0.0387, "step": 55320 }, { "grad_norm": 0.48220884799957275, "learning_rate": 4.5329154230800466e-05, "loss": 0.0447, "step": 55330 }, { "grad_norm": 0.3804926574230194, "learning_rate": 4.531269209325525e-05, "loss": 0.0369, "step": 55340 }, { "grad_norm": 0.5474656224250793, "learning_rate": 4.529623046830696e-05, "loss": 0.0344, "step": 55350 }, { "grad_norm": 0.7110009789466858, "learning_rate": 4.5279769357755866e-05, "loss": 0.0488, "step": 55360 }, { "grad_norm": 0.690524160861969, "learning_rate": 4.5263308763402084e-05, "loss": 0.0387, "step": 55370 }, { "grad_norm": 0.25871574878692627, "learning_rate": 4.524684868704574e-05, "loss": 0.0453, "step": 55380 }, { "grad_norm": 0.5691101551055908, "learning_rate": 4.523038913048688e-05, "loss": 0.0444, "step": 55390 }, { "grad_norm": 0.47522762417793274, "learning_rate": 4.521393009552549e-05, "loss": 0.0407, "step": 55400 }, { "grad_norm": 0.39427077770233154, "learning_rate": 4.5197471583961484e-05, "loss": 0.0475, "step": 55410 }, { "grad_norm": 0.30220794677734375, "learning_rate": 4.51810135975948e-05, "loss": 0.0429, "step": 55420 }, { "grad_norm": 0.38889703154563904, "learning_rate": 4.516455613822521e-05, "loss": 0.0397, "step": 55430 }, { "grad_norm": 0.33073925971984863, "learning_rate": 4.514809920765249e-05, "loss": 0.0462, "step": 55440 }, { "grad_norm": 0.5304275155067444, "learning_rate": 4.5131642807676356e-05, "loss": 0.0399, "step": 55450 }, { "grad_norm": 0.884427547454834, "learning_rate": 4.511518694009644e-05, "loss": 0.0562, "step": 55460 }, { "grad_norm": 0.6306154727935791, "learning_rate": 4.509873160671231e-05, "loss": 0.0488, "step": 55470 }, { "grad_norm": 0.5634375810623169, "learning_rate": 4.508227680932356e-05, "loss": 0.0384, "step": 55480 }, { "grad_norm": 1.1916974782943726, "learning_rate": 4.5065822549729605e-05, "loss": 0.0414, "step": 55490 }, { "grad_norm": 0.39914190769195557, "learning_rate": 4.504936882972989e-05, "loss": 0.0372, "step": 55500 }, { "grad_norm": 0.4054110050201416, "learning_rate": 4.503291565112374e-05, "loss": 0.0303, "step": 55510 }, { "grad_norm": 0.4285624027252197, "learning_rate": 4.501646301571048e-05, "loss": 0.0456, "step": 55520 }, { "grad_norm": 0.5981781482696533, "learning_rate": 4.500001092528933e-05, "loss": 0.0492, "step": 55530 }, { "grad_norm": 0.4008307158946991, "learning_rate": 4.4983559381659455e-05, "loss": 0.0387, "step": 55540 }, { "grad_norm": 0.3914217948913574, "learning_rate": 4.496710838661999e-05, "loss": 0.0436, "step": 55550 }, { "grad_norm": 0.42945316433906555, "learning_rate": 4.495065794196999e-05, "loss": 0.0454, "step": 55560 }, { "grad_norm": 0.5529891848564148, "learning_rate": 4.4934208049508455e-05, "loss": 0.0376, "step": 55570 }, { "grad_norm": 0.50519198179245, "learning_rate": 4.4917758711034303e-05, "loss": 0.0377, "step": 55580 }, { "grad_norm": 0.44309183955192566, "learning_rate": 4.490130992834642e-05, "loss": 0.0434, "step": 55590 }, { "grad_norm": 0.6430642008781433, "learning_rate": 4.488486170324359e-05, "loss": 0.0659, "step": 55600 }, { "grad_norm": 0.7367786169052124, "learning_rate": 4.486841403752463e-05, "loss": 0.0354, "step": 55610 }, { "grad_norm": 0.44498926401138306, "learning_rate": 4.4851966932988174e-05, "loss": 0.0316, "step": 55620 }, { "grad_norm": 0.6620075702667236, "learning_rate": 4.4835520391432894e-05, "loss": 0.0485, "step": 55630 }, { "grad_norm": 0.4345439076423645, "learning_rate": 4.481907441465733e-05, "loss": 0.0497, "step": 55640 }, { "grad_norm": 0.5656676292419434, "learning_rate": 4.480262900446e-05, "loss": 0.0469, "step": 55650 }, { "grad_norm": 0.3762684166431427, "learning_rate": 4.478618416263933e-05, "loss": 0.05, "step": 55660 }, { "grad_norm": 0.38635626435279846, "learning_rate": 4.476973989099373e-05, "loss": 0.042, "step": 55670 }, { "grad_norm": 0.42150890827178955, "learning_rate": 4.475329619132153e-05, "loss": 0.0534, "step": 55680 }, { "grad_norm": 0.5657723546028137, "learning_rate": 4.4736853065420956e-05, "loss": 0.0567, "step": 55690 }, { "grad_norm": 0.6386726498603821, "learning_rate": 4.472041051509023e-05, "loss": 0.0427, "step": 55700 }, { "grad_norm": 0.37054237723350525, "learning_rate": 4.470396854212746e-05, "loss": 0.0433, "step": 55710 }, { "grad_norm": 0.6930151581764221, "learning_rate": 4.468752714833072e-05, "loss": 0.0476, "step": 55720 }, { "grad_norm": 0.27498218417167664, "learning_rate": 4.467108633549803e-05, "loss": 0.0381, "step": 55730 }, { "grad_norm": 0.46091845631599426, "learning_rate": 4.4654646105427335e-05, "loss": 0.0606, "step": 55740 }, { "grad_norm": 0.44717937707901, "learning_rate": 4.463820645991651e-05, "loss": 0.0401, "step": 55750 }, { "grad_norm": 0.3538439869880676, "learning_rate": 4.462176740076334e-05, "loss": 0.0335, "step": 55760 }, { "grad_norm": 0.40252143144607544, "learning_rate": 4.4605328929765624e-05, "loss": 0.0565, "step": 55770 }, { "grad_norm": 0.41488930583000183, "learning_rate": 4.458889104872102e-05, "loss": 0.0384, "step": 55780 }, { "grad_norm": 0.6321847438812256, "learning_rate": 4.457245375942715e-05, "loss": 0.047, "step": 55790 }, { "grad_norm": 0.3997161388397217, "learning_rate": 4.455601706368157e-05, "loss": 0.0477, "step": 55800 }, { "grad_norm": 0.6964524984359741, "learning_rate": 4.453958096328179e-05, "loss": 0.0387, "step": 55810 }, { "grad_norm": 0.6150414347648621, "learning_rate": 4.452314546002523e-05, "loss": 0.0434, "step": 55820 }, { "grad_norm": 0.35338491201400757, "learning_rate": 4.4506710555709246e-05, "loss": 0.0527, "step": 55830 }, { "grad_norm": 0.458169162273407, "learning_rate": 4.449027625213114e-05, "loss": 0.0406, "step": 55840 }, { "grad_norm": 0.685653030872345, "learning_rate": 4.4473842551088135e-05, "loss": 0.0435, "step": 55850 }, { "grad_norm": 0.6411053538322449, "learning_rate": 4.445740945437738e-05, "loss": 0.0429, "step": 55860 }, { "grad_norm": 0.39928266406059265, "learning_rate": 4.444097696379602e-05, "loss": 0.0416, "step": 55870 }, { "grad_norm": 0.38313519954681396, "learning_rate": 4.442454508114106e-05, "loss": 0.0398, "step": 55880 }, { "grad_norm": 0.9392313957214355, "learning_rate": 4.4408113808209465e-05, "loss": 0.0493, "step": 55890 }, { "grad_norm": 0.4267013370990753, "learning_rate": 4.439168314679813e-05, "loss": 0.044, "step": 55900 }, { "grad_norm": 1.776320457458496, "learning_rate": 4.43752530987039e-05, "loss": 0.0424, "step": 55910 }, { "grad_norm": 0.4680209457874298, "learning_rate": 4.4358823665723515e-05, "loss": 0.0363, "step": 55920 }, { "grad_norm": 0.6710439324378967, "learning_rate": 4.434239484965371e-05, "loss": 0.0499, "step": 55930 }, { "grad_norm": 0.3779048025608063, "learning_rate": 4.4325966652291103e-05, "loss": 0.0495, "step": 55940 }, { "grad_norm": 0.556693971157074, "learning_rate": 4.430953907543225e-05, "loss": 0.0383, "step": 55950 }, { "grad_norm": 0.4181585907936096, "learning_rate": 4.4293112120873645e-05, "loss": 0.0324, "step": 55960 }, { "grad_norm": 0.3287923038005829, "learning_rate": 4.427668579041172e-05, "loss": 0.0342, "step": 55970 }, { "grad_norm": 0.44788405299186707, "learning_rate": 4.4260260085842816e-05, "loss": 0.0413, "step": 55980 }, { "grad_norm": 0.6086676120758057, "learning_rate": 4.4243835008963256e-05, "loss": 0.0398, "step": 55990 }, { "grad_norm": 0.33980295062065125, "learning_rate": 4.422741056156925e-05, "loss": 0.04, "step": 56000 }, { "grad_norm": 0.43625110387802124, "learning_rate": 4.421098674545693e-05, "loss": 0.0314, "step": 56010 }, { "grad_norm": 0.4895833134651184, "learning_rate": 4.419456356242241e-05, "loss": 0.043, "step": 56020 }, { "grad_norm": 0.3793869614601135, "learning_rate": 4.4178141014261685e-05, "loss": 0.0392, "step": 56030 }, { "grad_norm": 0.508904755115509, "learning_rate": 4.416171910277071e-05, "loss": 0.0435, "step": 56040 }, { "grad_norm": 0.5548250079154968, "learning_rate": 4.414529782974533e-05, "loss": 0.0372, "step": 56050 }, { "grad_norm": 0.2295214682817459, "learning_rate": 4.41288771969814e-05, "loss": 0.037, "step": 56060 }, { "grad_norm": 0.5291264653205872, "learning_rate": 4.411245720627464e-05, "loss": 0.0409, "step": 56070 }, { "grad_norm": 0.7990670204162598, "learning_rate": 4.40960378594207e-05, "loss": 0.0393, "step": 56080 }, { "grad_norm": 1.079327940940857, "learning_rate": 4.407961915821519e-05, "loss": 0.0478, "step": 56090 }, { "grad_norm": 0.26742494106292725, "learning_rate": 4.406320110445362e-05, "loss": 0.0401, "step": 56100 }, { "grad_norm": 0.4789997339248657, "learning_rate": 4.404678369993144e-05, "loss": 0.0368, "step": 56110 }, { "grad_norm": 1.1349425315856934, "learning_rate": 4.403036694644406e-05, "loss": 0.0517, "step": 56120 }, { "grad_norm": 0.42528775334358215, "learning_rate": 4.4013950845786764e-05, "loss": 0.0473, "step": 56130 }, { "grad_norm": 0.39725056290626526, "learning_rate": 4.399753539975482e-05, "loss": 0.0528, "step": 56140 }, { "grad_norm": 0.3865397572517395, "learning_rate": 4.398112061014337e-05, "loss": 0.0389, "step": 56150 }, { "grad_norm": 0.336995929479599, "learning_rate": 4.396470647874753e-05, "loss": 0.0431, "step": 56160 }, { "grad_norm": 0.7294092178344727, "learning_rate": 4.394829300736229e-05, "loss": 0.0357, "step": 56170 }, { "grad_norm": 0.4851996898651123, "learning_rate": 4.393188019778265e-05, "loss": 0.046, "step": 56180 }, { "grad_norm": 0.29706427454948425, "learning_rate": 4.391546805180347e-05, "loss": 0.0456, "step": 56190 }, { "grad_norm": 0.9175385236740112, "learning_rate": 4.389905657121955e-05, "loss": 0.0428, "step": 56200 }, { "grad_norm": 0.5457308888435364, "learning_rate": 4.3882645757825635e-05, "loss": 0.043, "step": 56210 }, { "grad_norm": 0.41047927737236023, "learning_rate": 4.386623561341637e-05, "loss": 0.04, "step": 56220 }, { "grad_norm": 1.2766482830047607, "learning_rate": 4.384982613978637e-05, "loss": 0.0457, "step": 56230 }, { "grad_norm": 0.5019412040710449, "learning_rate": 4.383341733873012e-05, "loss": 0.0423, "step": 56240 }, { "grad_norm": 0.4624094069004059, "learning_rate": 4.381700921204209e-05, "loss": 0.0385, "step": 56250 }, { "grad_norm": 0.9021411538124084, "learning_rate": 4.380060176151663e-05, "loss": 0.0359, "step": 56260 }, { "grad_norm": 0.36778825521469116, "learning_rate": 4.378419498894805e-05, "loss": 0.048, "step": 56270 }, { "grad_norm": 0.40519291162490845, "learning_rate": 4.376778889613056e-05, "loss": 0.0443, "step": 56280 }, { "grad_norm": 0.39591333270072937, "learning_rate": 4.375138348485831e-05, "loss": 0.0523, "step": 56290 }, { "grad_norm": 0.39180922508239746, "learning_rate": 4.3734978756925346e-05, "loss": 0.0403, "step": 56300 }, { "grad_norm": 0.490201860666275, "learning_rate": 4.3718574714125706e-05, "loss": 0.049, "step": 56310 }, { "grad_norm": 0.5798774361610413, "learning_rate": 4.370217135825329e-05, "loss": 0.0456, "step": 56320 }, { "grad_norm": 0.7064782381057739, "learning_rate": 4.368576869110194e-05, "loss": 0.047, "step": 56330 }, { "grad_norm": 0.40981096029281616, "learning_rate": 4.366936671446544e-05, "loss": 0.0426, "step": 56340 }, { "grad_norm": 0.43913155794143677, "learning_rate": 4.3652965430137474e-05, "loss": 0.044, "step": 56350 }, { "grad_norm": 0.4285159707069397, "learning_rate": 4.3636564839911646e-05, "loss": 0.0335, "step": 56360 }, { "grad_norm": 0.38142338395118713, "learning_rate": 4.3620164945581545e-05, "loss": 0.0389, "step": 56370 }, { "grad_norm": 0.8348338007926941, "learning_rate": 4.360376574894061e-05, "loss": 0.0449, "step": 56380 }, { "grad_norm": 0.2458961457014084, "learning_rate": 4.358736725178224e-05, "loss": 0.0396, "step": 56390 }, { "grad_norm": 0.5392361879348755, "learning_rate": 4.357096945589974e-05, "loss": 0.0488, "step": 56400 }, { "grad_norm": 0.438478946685791, "learning_rate": 4.355457236308636e-05, "loss": 0.0405, "step": 56410 }, { "grad_norm": 0.4973950982093811, "learning_rate": 4.353817597513526e-05, "loss": 0.0379, "step": 56420 }, { "grad_norm": 0.5200930237770081, "learning_rate": 4.352178029383948e-05, "loss": 0.0332, "step": 56430 }, { "grad_norm": 0.5486636161804199, "learning_rate": 4.3505385320992105e-05, "loss": 0.0447, "step": 56440 }, { "grad_norm": 0.41152095794677734, "learning_rate": 4.348899105838602e-05, "loss": 0.0356, "step": 56450 }, { "grad_norm": 0.42142027616500854, "learning_rate": 4.3472597507814087e-05, "loss": 0.0388, "step": 56460 }, { "grad_norm": 0.649494469165802, "learning_rate": 4.3456204671069066e-05, "loss": 0.0469, "step": 56470 }, { "grad_norm": 0.3961481750011444, "learning_rate": 4.343981254994367e-05, "loss": 0.0354, "step": 56480 }, { "grad_norm": 0.4350820481777191, "learning_rate": 4.34234211462305e-05, "loss": 0.0405, "step": 56490 }, { "grad_norm": 0.48370596766471863, "learning_rate": 4.340703046172213e-05, "loss": 0.0405, "step": 56500 }, { "grad_norm": 0.37645867466926575, "learning_rate": 4.339064049821097e-05, "loss": 0.0531, "step": 56510 }, { "grad_norm": 0.3641960918903351, "learning_rate": 4.3374251257489446e-05, "loss": 0.0462, "step": 56520 }, { "grad_norm": 1.0867246389389038, "learning_rate": 4.335786274134984e-05, "loss": 0.0482, "step": 56530 }, { "grad_norm": 0.7909472584724426, "learning_rate": 4.334147495158438e-05, "loss": 0.0447, "step": 56540 }, { "grad_norm": 0.5299379825592041, "learning_rate": 4.33250878899852e-05, "loss": 0.0482, "step": 56550 }, { "grad_norm": 0.4181661903858185, "learning_rate": 4.330870155834437e-05, "loss": 0.0403, "step": 56560 }, { "grad_norm": 0.6458278298377991, "learning_rate": 4.3292315958453884e-05, "loss": 0.0434, "step": 56570 }, { "grad_norm": 1.5824373960494995, "learning_rate": 4.3275931092105646e-05, "loss": 0.042, "step": 56580 }, { "grad_norm": 0.44135114550590515, "learning_rate": 4.325954696109147e-05, "loss": 0.0409, "step": 56590 }, { "grad_norm": 0.5727918148040771, "learning_rate": 4.32431635672031e-05, "loss": 0.0521, "step": 56600 }, { "grad_norm": 0.5332741141319275, "learning_rate": 4.3226780912232214e-05, "loss": 0.0404, "step": 56610 }, { "grad_norm": 0.29144397377967834, "learning_rate": 4.321039899797036e-05, "loss": 0.0426, "step": 56620 }, { "grad_norm": 0.4633723497390747, "learning_rate": 4.319401782620908e-05, "loss": 0.0414, "step": 56630 }, { "grad_norm": 0.336722195148468, "learning_rate": 4.317763739873978e-05, "loss": 0.0365, "step": 56640 }, { "grad_norm": 0.5082413554191589, "learning_rate": 4.31612577173538e-05, "loss": 0.0408, "step": 56650 }, { "grad_norm": 0.28993871808052063, "learning_rate": 4.314487878384239e-05, "loss": 0.044, "step": 56660 }, { "grad_norm": 0.6705368161201477, "learning_rate": 4.3128500599996736e-05, "loss": 0.0503, "step": 56670 }, { "grad_norm": 0.3524230718612671, "learning_rate": 4.31121231676079e-05, "loss": 0.0396, "step": 56680 }, { "grad_norm": 1.3538137674331665, "learning_rate": 4.309574648846694e-05, "loss": 0.0472, "step": 56690 }, { "grad_norm": 0.797331690788269, "learning_rate": 4.3079370564364755e-05, "loss": 0.0548, "step": 56700 }, { "grad_norm": 0.36452624201774597, "learning_rate": 4.306299539709221e-05, "loss": 0.043, "step": 56710 }, { "grad_norm": 0.37097522616386414, "learning_rate": 4.3046620988440046e-05, "loss": 0.041, "step": 56720 }, { "grad_norm": 0.4763125777244568, "learning_rate": 4.303024734019897e-05, "loss": 0.0519, "step": 56730 }, { "grad_norm": 0.4805646240711212, "learning_rate": 4.3013874454159565e-05, "loss": 0.0432, "step": 56740 }, { "grad_norm": 0.9640550017356873, "learning_rate": 4.299750233211233e-05, "loss": 0.0534, "step": 56750 }, { "grad_norm": 0.42185455560684204, "learning_rate": 4.2981130975847715e-05, "loss": 0.039, "step": 56760 }, { "grad_norm": 0.48579853773117065, "learning_rate": 4.2964760387156076e-05, "loss": 0.043, "step": 56770 }, { "grad_norm": 0.4488954246044159, "learning_rate": 4.294839056782767e-05, "loss": 0.0438, "step": 56780 }, { "grad_norm": 2.086817979812622, "learning_rate": 4.293202151965266e-05, "loss": 0.0604, "step": 56790 }, { "grad_norm": 0.374008446931839, "learning_rate": 4.291565324442115e-05, "loss": 0.0384, "step": 56800 }, { "grad_norm": 0.3506811559200287, "learning_rate": 4.2899285743923135e-05, "loss": 0.0455, "step": 56810 }, { "grad_norm": 0.3095397651195526, "learning_rate": 4.288291901994857e-05, "loss": 0.0342, "step": 56820 }, { "grad_norm": 0.4761693477630615, "learning_rate": 4.286655307428729e-05, "loss": 0.0364, "step": 56830 }, { "grad_norm": 0.44548165798187256, "learning_rate": 4.285018790872903e-05, "loss": 0.0516, "step": 56840 }, { "grad_norm": 0.6110193729400635, "learning_rate": 4.283382352506347e-05, "loss": 0.0536, "step": 56850 }, { "grad_norm": 0.36664679646492004, "learning_rate": 4.2817459925080205e-05, "loss": 0.0526, "step": 56860 }, { "grad_norm": 0.46009311079978943, "learning_rate": 4.28010971105687e-05, "loss": 0.0492, "step": 56870 }, { "grad_norm": 0.5867972373962402, "learning_rate": 4.278473508331841e-05, "loss": 0.0442, "step": 56880 }, { "grad_norm": 0.5797456502914429, "learning_rate": 4.276837384511864e-05, "loss": 0.0478, "step": 56890 }, { "grad_norm": 0.44578930735588074, "learning_rate": 4.275201339775864e-05, "loss": 0.0405, "step": 56900 }, { "grad_norm": 0.4795222878456116, "learning_rate": 4.273565374302756e-05, "loss": 0.044, "step": 56910 }, { "grad_norm": 0.6822552680969238, "learning_rate": 4.2719294882714454e-05, "loss": 0.0387, "step": 56920 }, { "grad_norm": 0.7057689428329468, "learning_rate": 4.270293681860831e-05, "loss": 0.0426, "step": 56930 }, { "grad_norm": 1.0341154336929321, "learning_rate": 4.268657955249802e-05, "loss": 0.0392, "step": 56940 }, { "grad_norm": 0.5974655151367188, "learning_rate": 4.267022308617241e-05, "loss": 0.0349, "step": 56950 }, { "grad_norm": 0.9060894250869751, "learning_rate": 4.265386742142018e-05, "loss": 0.0503, "step": 56960 }, { "grad_norm": 0.5837939381599426, "learning_rate": 4.263751256002996e-05, "loss": 0.0402, "step": 56970 }, { "grad_norm": 0.5832939743995667, "learning_rate": 4.26211585037903e-05, "loss": 0.0512, "step": 56980 }, { "grad_norm": 0.5402244329452515, "learning_rate": 4.2604805254489656e-05, "loss": 0.0513, "step": 56990 }, { "grad_norm": 1.2459875345230103, "learning_rate": 4.258845281391638e-05, "loss": 0.0377, "step": 57000 }, { "grad_norm": 0.47119951248168945, "learning_rate": 4.257210118385876e-05, "loss": 0.0458, "step": 57010 }, { "grad_norm": 0.5834695100784302, "learning_rate": 4.2555750366104994e-05, "loss": 0.0443, "step": 57020 }, { "grad_norm": 0.41398072242736816, "learning_rate": 4.2539400362443184e-05, "loss": 0.039, "step": 57030 }, { "grad_norm": 0.5245487093925476, "learning_rate": 4.252305117466133e-05, "loss": 0.0516, "step": 57040 }, { "grad_norm": 0.49533599615097046, "learning_rate": 4.2506702804547356e-05, "loss": 0.0438, "step": 57050 }, { "grad_norm": 0.6787128448486328, "learning_rate": 4.249035525388907e-05, "loss": 0.0396, "step": 57060 }, { "grad_norm": 0.9706096053123474, "learning_rate": 4.247400852447427e-05, "loss": 0.049, "step": 57070 }, { "grad_norm": 0.5405175089836121, "learning_rate": 4.245766261809059e-05, "loss": 0.0455, "step": 57080 }, { "grad_norm": 0.44075635075569153, "learning_rate": 4.244131753652558e-05, "loss": 0.0392, "step": 57090 }, { "grad_norm": 0.3588103652000427, "learning_rate": 4.242497328156672e-05, "loss": 0.0377, "step": 57100 }, { "grad_norm": 0.44053348898887634, "learning_rate": 4.240862985500139e-05, "loss": 0.0465, "step": 57110 }, { "grad_norm": 0.38664186000823975, "learning_rate": 4.2392287258616894e-05, "loss": 0.0373, "step": 57120 }, { "grad_norm": 0.3479814827442169, "learning_rate": 4.2375945494200396e-05, "loss": 0.036, "step": 57130 }, { "grad_norm": 0.46177002787590027, "learning_rate": 4.2359604563539054e-05, "loss": 0.0505, "step": 57140 }, { "grad_norm": 0.33809781074523926, "learning_rate": 4.2343264468419885e-05, "loss": 0.04, "step": 57150 }, { "grad_norm": 0.5268710255622864, "learning_rate": 4.232692521062979e-05, "loss": 0.0303, "step": 57160 }, { "grad_norm": 0.7125335931777954, "learning_rate": 4.231058679195562e-05, "loss": 0.0351, "step": 57170 }, { "grad_norm": 0.5269103646278381, "learning_rate": 4.229424921418411e-05, "loss": 0.0437, "step": 57180 }, { "grad_norm": 0.49478185176849365, "learning_rate": 4.2277912479101916e-05, "loss": 0.0499, "step": 57190 }, { "grad_norm": 0.5693857669830322, "learning_rate": 4.226157658849562e-05, "loss": 0.0402, "step": 57200 }, { "grad_norm": 0.3563341796398163, "learning_rate": 4.2245241544151664e-05, "loss": 0.0409, "step": 57210 }, { "grad_norm": 0.5468988418579102, "learning_rate": 4.2228907347856425e-05, "loss": 0.0547, "step": 57220 }, { "grad_norm": 0.6963083148002625, "learning_rate": 4.2212574001396214e-05, "loss": 0.0456, "step": 57230 }, { "grad_norm": 0.4254370927810669, "learning_rate": 4.21962415065572e-05, "loss": 0.0389, "step": 57240 }, { "grad_norm": 0.5204071998596191, "learning_rate": 4.217990986512548e-05, "loss": 0.0407, "step": 57250 }, { "grad_norm": 0.5521912574768066, "learning_rate": 4.216357907888704e-05, "loss": 0.0418, "step": 57260 }, { "grad_norm": 0.625553548336029, "learning_rate": 4.2147249149627824e-05, "loss": 0.0442, "step": 57270 }, { "grad_norm": 0.3641912341117859, "learning_rate": 4.213092007913364e-05, "loss": 0.045, "step": 57280 }, { "grad_norm": 0.26930201053619385, "learning_rate": 4.2114591869190205e-05, "loss": 0.0444, "step": 57290 }, { "grad_norm": 1.199590802192688, "learning_rate": 4.209826452158315e-05, "loss": 0.0488, "step": 57300 }, { "grad_norm": 0.6137425899505615, "learning_rate": 4.2081938038098e-05, "loss": 0.043, "step": 57310 }, { "grad_norm": 0.4852061867713928, "learning_rate": 4.2065612420520185e-05, "loss": 0.0358, "step": 57320 }, { "grad_norm": 0.2674347460269928, "learning_rate": 4.204928767063509e-05, "loss": 0.0429, "step": 57330 }, { "grad_norm": 0.34681960940361023, "learning_rate": 4.2032963790227944e-05, "loss": 0.0422, "step": 57340 }, { "grad_norm": 0.29472410678863525, "learning_rate": 4.2016640781083887e-05, "loss": 0.0357, "step": 57350 }, { "grad_norm": 0.38672155141830444, "learning_rate": 4.2000318644988004e-05, "loss": 0.0568, "step": 57360 }, { "grad_norm": 0.9354803562164307, "learning_rate": 4.198399738372524e-05, "loss": 0.0381, "step": 57370 }, { "grad_norm": 0.43275460600852966, "learning_rate": 4.196767699908045e-05, "loss": 0.0375, "step": 57380 }, { "grad_norm": 0.5717988610267639, "learning_rate": 4.195135749283845e-05, "loss": 0.0415, "step": 57390 }, { "grad_norm": 0.39142605662345886, "learning_rate": 4.1935038866783894e-05, "loss": 0.0389, "step": 57400 }, { "grad_norm": 0.3463146388530731, "learning_rate": 4.1918721122701355e-05, "loss": 0.0365, "step": 57410 }, { "grad_norm": 0.6729097366333008, "learning_rate": 4.190240426237533e-05, "loss": 0.053, "step": 57420 }, { "grad_norm": 0.36239105463027954, "learning_rate": 4.188608828759019e-05, "loss": 0.0396, "step": 57430 }, { "grad_norm": 0.49410945177078247, "learning_rate": 4.1869773200130246e-05, "loss": 0.0695, "step": 57440 }, { "grad_norm": 0.6017895936965942, "learning_rate": 4.185345900177967e-05, "loss": 0.0503, "step": 57450 }, { "grad_norm": 0.32816073298454285, "learning_rate": 4.183714569432258e-05, "loss": 0.0464, "step": 57460 }, { "grad_norm": 0.4456772804260254, "learning_rate": 4.182083327954295e-05, "loss": 0.0389, "step": 57470 }, { "grad_norm": 0.37324613332748413, "learning_rate": 4.1804521759224714e-05, "loss": 0.0416, "step": 57480 }, { "grad_norm": 0.32324355840682983, "learning_rate": 4.178821113515165e-05, "loss": 0.0418, "step": 57490 }, { "grad_norm": 0.42490699887275696, "learning_rate": 4.1771901409107476e-05, "loss": 0.0465, "step": 57500 }, { "grad_norm": 0.47574445605278015, "learning_rate": 4.175559258287577e-05, "loss": 0.0394, "step": 57510 }, { "grad_norm": 0.8373574018478394, "learning_rate": 4.173928465824008e-05, "loss": 0.038, "step": 57520 }, { "grad_norm": 0.44686460494995117, "learning_rate": 4.17229776369838e-05, "loss": 0.0438, "step": 57530 }, { "grad_norm": 0.5604062676429749, "learning_rate": 4.170667152089025e-05, "loss": 0.0502, "step": 57540 }, { "grad_norm": 0.42635464668273926, "learning_rate": 4.1690366311742626e-05, "loss": 0.0389, "step": 57550 }, { "grad_norm": 0.6883972883224487, "learning_rate": 4.1674062011324055e-05, "loss": 0.0409, "step": 57560 }, { "grad_norm": 0.3011237382888794, "learning_rate": 4.165775862141752e-05, "loss": 0.0333, "step": 57570 }, { "grad_norm": 0.7401235103607178, "learning_rate": 4.164145614380598e-05, "loss": 0.036, "step": 57580 }, { "grad_norm": 0.3432281017303467, "learning_rate": 4.162515458027223e-05, "loss": 0.0357, "step": 57590 }, { "grad_norm": 0.6163048148155212, "learning_rate": 4.160885393259899e-05, "loss": 0.0494, "step": 57600 }, { "grad_norm": 0.3972168564796448, "learning_rate": 4.159255420256886e-05, "loss": 0.0467, "step": 57610 }, { "grad_norm": 0.5334274768829346, "learning_rate": 4.157625539196436e-05, "loss": 0.0392, "step": 57620 }, { "grad_norm": 0.36962515115737915, "learning_rate": 4.155995750256791e-05, "loss": 0.0433, "step": 57630 }, { "grad_norm": 0.3617318868637085, "learning_rate": 4.15436605361618e-05, "loss": 0.0484, "step": 57640 }, { "grad_norm": 0.49260151386260986, "learning_rate": 4.152736449452827e-05, "loss": 0.0477, "step": 57650 }, { "grad_norm": 0.43343523144721985, "learning_rate": 4.151106937944943e-05, "loss": 0.0447, "step": 57660 }, { "grad_norm": 0.40746980905532837, "learning_rate": 4.149477519270727e-05, "loss": 0.0364, "step": 57670 }, { "grad_norm": 0.8852056860923767, "learning_rate": 4.147848193608371e-05, "loss": 0.0532, "step": 57680 }, { "grad_norm": 0.4505899250507355, "learning_rate": 4.146218961136056e-05, "loss": 0.0402, "step": 57690 }, { "grad_norm": 0.303195983171463, "learning_rate": 4.14458982203195e-05, "loss": 0.0476, "step": 57700 }, { "grad_norm": 0.35403555631637573, "learning_rate": 4.1429607764742165e-05, "loss": 0.0477, "step": 57710 }, { "grad_norm": 0.388899564743042, "learning_rate": 4.141331824641003e-05, "loss": 0.0485, "step": 57720 }, { "grad_norm": 0.5691407322883606, "learning_rate": 4.139702966710452e-05, "loss": 0.057, "step": 57730 }, { "grad_norm": 0.35572460293769836, "learning_rate": 4.13807420286069e-05, "loss": 0.045, "step": 57740 }, { "grad_norm": 0.47436580061912537, "learning_rate": 4.136445533269837e-05, "loss": 0.0481, "step": 57750 }, { "grad_norm": 0.32365700602531433, "learning_rate": 4.134816958116003e-05, "loss": 0.033, "step": 57760 }, { "grad_norm": 0.6032266616821289, "learning_rate": 4.133188477577283e-05, "loss": 0.0431, "step": 57770 }, { "grad_norm": 0.47431814670562744, "learning_rate": 4.131560091831771e-05, "loss": 0.0375, "step": 57780 }, { "grad_norm": 0.6495683789253235, "learning_rate": 4.1299318010575405e-05, "loss": 0.0349, "step": 57790 }, { "grad_norm": 1.2200407981872559, "learning_rate": 4.128303605432661e-05, "loss": 0.0676, "step": 57800 }, { "grad_norm": 0.5275098085403442, "learning_rate": 4.126675505135188e-05, "loss": 0.0405, "step": 57810 }, { "grad_norm": 0.7128943800926208, "learning_rate": 4.125047500343169e-05, "loss": 0.0387, "step": 57820 }, { "grad_norm": 0.6629016399383545, "learning_rate": 4.1234195912346376e-05, "loss": 0.0416, "step": 57830 }, { "grad_norm": 0.6502124071121216, "learning_rate": 4.1217917779876235e-05, "loss": 0.0463, "step": 57840 }, { "grad_norm": 0.3736846148967743, "learning_rate": 4.120164060780141e-05, "loss": 0.0496, "step": 57850 }, { "grad_norm": 0.4891628324985504, "learning_rate": 4.1185364397901935e-05, "loss": 0.0449, "step": 57860 }, { "grad_norm": 0.3539184033870697, "learning_rate": 4.116908915195776e-05, "loss": 0.0397, "step": 57870 }, { "grad_norm": 0.5481688380241394, "learning_rate": 4.115281487174872e-05, "loss": 0.038, "step": 57880 }, { "grad_norm": 0.9608185291290283, "learning_rate": 4.113654155905452e-05, "loss": 0.0386, "step": 57890 }, { "grad_norm": 0.4929003417491913, "learning_rate": 4.112026921565483e-05, "loss": 0.0445, "step": 57900 }, { "grad_norm": 0.5276167392730713, "learning_rate": 4.1103997843329147e-05, "loss": 0.0344, "step": 57910 }, { "grad_norm": 0.5060994029045105, "learning_rate": 4.108772744385689e-05, "loss": 0.0562, "step": 57920 }, { "grad_norm": 0.5455513000488281, "learning_rate": 4.1071458019017346e-05, "loss": 0.0348, "step": 57930 }, { "grad_norm": 0.4544530212879181, "learning_rate": 4.1055189570589734e-05, "loss": 0.0434, "step": 57940 }, { "grad_norm": 0.48604053258895874, "learning_rate": 4.103892210035315e-05, "loss": 0.0317, "step": 57950 }, { "grad_norm": 0.6757197380065918, "learning_rate": 4.102265561008656e-05, "loss": 0.0458, "step": 57960 }, { "grad_norm": 0.539827287197113, "learning_rate": 4.100639010156885e-05, "loss": 0.0345, "step": 57970 }, { "grad_norm": 0.4467882513999939, "learning_rate": 4.099012557657882e-05, "loss": 0.0427, "step": 57980 }, { "grad_norm": 0.6373007893562317, "learning_rate": 4.09738620368951e-05, "loss": 0.0461, "step": 57990 }, { "grad_norm": 0.2836066484451294, "learning_rate": 4.095759948429626e-05, "loss": 0.0354, "step": 58000 }, { "grad_norm": 0.5239490866661072, "learning_rate": 4.094133792056075e-05, "loss": 0.0459, "step": 58010 }, { "grad_norm": 0.30355340242385864, "learning_rate": 4.092507734746688e-05, "loss": 0.0496, "step": 58020 }, { "grad_norm": 0.5357921123504639, "learning_rate": 4.090881776679293e-05, "loss": 0.0376, "step": 58030 }, { "grad_norm": 0.4277862012386322, "learning_rate": 4.0892559180316994e-05, "loss": 0.0404, "step": 58040 }, { "grad_norm": 0.5389938354492188, "learning_rate": 4.08763015898171e-05, "loss": 0.0416, "step": 58050 }, { "grad_norm": 0.4725273847579956, "learning_rate": 4.0860044997071145e-05, "loss": 0.0453, "step": 58060 }, { "grad_norm": 0.6645432710647583, "learning_rate": 4.084378940385693e-05, "loss": 0.0445, "step": 58070 }, { "grad_norm": 0.522980272769928, "learning_rate": 4.082753481195211e-05, "loss": 0.0412, "step": 58080 }, { "grad_norm": 0.42364367842674255, "learning_rate": 4.0811281223134324e-05, "loss": 0.0396, "step": 58090 }, { "grad_norm": 1.2732911109924316, "learning_rate": 4.0795028639181e-05, "loss": 0.0488, "step": 58100 }, { "grad_norm": 0.7638841867446899, "learning_rate": 4.077877706186951e-05, "loss": 0.0475, "step": 58110 }, { "grad_norm": 0.5463421940803528, "learning_rate": 4.076252649297709e-05, "loss": 0.0424, "step": 58120 }, { "grad_norm": 0.3792135417461395, "learning_rate": 4.07462769342809e-05, "loss": 0.0417, "step": 58130 }, { "grad_norm": 0.5185579061508179, "learning_rate": 4.073002838755793e-05, "loss": 0.0489, "step": 58140 }, { "grad_norm": 0.6141214966773987, "learning_rate": 4.0713780854585115e-05, "loss": 0.0476, "step": 58150 }, { "grad_norm": 0.4687246084213257, "learning_rate": 4.069753433713927e-05, "loss": 0.0544, "step": 58160 }, { "grad_norm": 0.8651167154312134, "learning_rate": 4.068128883699709e-05, "loss": 0.0363, "step": 58170 }, { "grad_norm": 0.5963608026504517, "learning_rate": 4.0665044355935155e-05, "loss": 0.039, "step": 58180 }, { "grad_norm": 0.3484349250793457, "learning_rate": 4.064880089572993e-05, "loss": 0.0466, "step": 58190 }, { "grad_norm": 0.5311936140060425, "learning_rate": 4.063255845815779e-05, "loss": 0.0438, "step": 58200 }, { "grad_norm": 0.3602257966995239, "learning_rate": 4.0616317044994945e-05, "loss": 0.0425, "step": 58210 }, { "grad_norm": 0.3938644826412201, "learning_rate": 4.0600076658017585e-05, "loss": 0.0341, "step": 58220 }, { "grad_norm": 0.43306395411491394, "learning_rate": 4.0583837299001706e-05, "loss": 0.039, "step": 58230 }, { "grad_norm": 0.4898452162742615, "learning_rate": 4.056759896972322e-05, "loss": 0.0391, "step": 58240 }, { "grad_norm": 0.5201796293258667, "learning_rate": 4.055136167195793e-05, "loss": 0.0492, "step": 58250 }, { "grad_norm": 0.36911168694496155, "learning_rate": 4.053512540748152e-05, "loss": 0.044, "step": 58260 }, { "grad_norm": 0.9928928017616272, "learning_rate": 4.051889017806955e-05, "loss": 0.064, "step": 58270 }, { "grad_norm": 0.6774256229400635, "learning_rate": 4.0502655985497516e-05, "loss": 0.0461, "step": 58280 }, { "grad_norm": 0.7996504902839661, "learning_rate": 4.048642283154074e-05, "loss": 0.0546, "step": 58290 }, { "grad_norm": 0.3459218740463257, "learning_rate": 4.0470190717974464e-05, "loss": 0.0451, "step": 58300 }, { "grad_norm": 0.3746657967567444, "learning_rate": 4.045395964657381e-05, "loss": 0.0482, "step": 58310 }, { "grad_norm": 0.6459255218505859, "learning_rate": 4.043772961911376e-05, "loss": 0.05, "step": 58320 }, { "grad_norm": 0.46662718057632446, "learning_rate": 4.042150063736924e-05, "loss": 0.0438, "step": 58330 }, { "grad_norm": 0.5667155385017395, "learning_rate": 4.0405272703114976e-05, "loss": 0.0397, "step": 58340 }, { "grad_norm": 0.8770315647125244, "learning_rate": 4.038904581812569e-05, "loss": 0.0454, "step": 58350 }, { "grad_norm": 0.47806236147880554, "learning_rate": 4.03728199841759e-05, "loss": 0.0392, "step": 58360 }, { "grad_norm": 0.4197111129760742, "learning_rate": 4.0356595203040044e-05, "loss": 0.0463, "step": 58370 }, { "grad_norm": 0.35457131266593933, "learning_rate": 4.0340371476492426e-05, "loss": 0.0432, "step": 58380 }, { "grad_norm": 0.8195424675941467, "learning_rate": 4.032414880630727e-05, "loss": 0.0392, "step": 58390 }, { "grad_norm": 0.4727485477924347, "learning_rate": 4.0307927194258646e-05, "loss": 0.0368, "step": 58400 }, { "grad_norm": 0.5703104734420776, "learning_rate": 4.029170664212054e-05, "loss": 0.042, "step": 58410 }, { "grad_norm": 0.4666248857975006, "learning_rate": 4.0275487151666794e-05, "loss": 0.0528, "step": 58420 }, { "grad_norm": 0.44791191816329956, "learning_rate": 4.025926872467116e-05, "loss": 0.0462, "step": 58430 }, { "grad_norm": 0.44173383712768555, "learning_rate": 4.024305136290725e-05, "loss": 0.0527, "step": 58440 }, { "grad_norm": 0.3364259600639343, "learning_rate": 4.022683506814858e-05, "loss": 0.0479, "step": 58450 }, { "grad_norm": 0.6345527768135071, "learning_rate": 4.021061984216853e-05, "loss": 0.0453, "step": 58460 }, { "grad_norm": 0.5990741848945618, "learning_rate": 4.019440568674035e-05, "loss": 0.0372, "step": 58470 }, { "grad_norm": 0.8149361610412598, "learning_rate": 4.017819260363724e-05, "loss": 0.0381, "step": 58480 }, { "grad_norm": 0.451519638299942, "learning_rate": 4.016198059463223e-05, "loss": 0.0562, "step": 58490 }, { "grad_norm": 0.2674528956413269, "learning_rate": 4.0145769661498233e-05, "loss": 0.0373, "step": 58500 }, { "grad_norm": 0.7748318910598755, "learning_rate": 4.012955980600804e-05, "loss": 0.0383, "step": 58510 }, { "grad_norm": 0.8286656737327576, "learning_rate": 4.0113351029934355e-05, "loss": 0.0445, "step": 58520 }, { "grad_norm": 0.31925252079963684, "learning_rate": 4.0097143335049714e-05, "loss": 0.0405, "step": 58530 }, { "grad_norm": 1.1040256023406982, "learning_rate": 4.0080936723126606e-05, "loss": 0.0342, "step": 58540 }, { "grad_norm": 0.3674704432487488, "learning_rate": 4.006473119593735e-05, "loss": 0.04, "step": 58550 }, { "grad_norm": 0.8422746062278748, "learning_rate": 4.004852675525415e-05, "loss": 0.0435, "step": 58560 }, { "grad_norm": 0.4481930434703827, "learning_rate": 4.00323234028491e-05, "loss": 0.0413, "step": 58570 }, { "grad_norm": 0.44227588176727295, "learning_rate": 4.0016121140494186e-05, "loss": 0.0419, "step": 58580 }, { "grad_norm": 0.5210809707641602, "learning_rate": 3.999991996996122e-05, "loss": 0.0418, "step": 58590 }, { "grad_norm": 0.494739830493927, "learning_rate": 3.9983719893022e-05, "loss": 0.0356, "step": 58600 }, { "grad_norm": 0.5082895159721375, "learning_rate": 3.996752091144812e-05, "loss": 0.0426, "step": 58610 }, { "grad_norm": 0.4555918574333191, "learning_rate": 3.9951323027011055e-05, "loss": 0.0413, "step": 58620 }, { "grad_norm": 0.41755351424217224, "learning_rate": 3.993512624148219e-05, "loss": 0.04, "step": 58630 }, { "grad_norm": 0.4151427149772644, "learning_rate": 3.99189305566328e-05, "loss": 0.0396, "step": 58640 }, { "grad_norm": 0.48099324107170105, "learning_rate": 3.990273597423401e-05, "loss": 0.0488, "step": 58650 }, { "grad_norm": 0.4067959189414978, "learning_rate": 3.988654249605681e-05, "loss": 0.0475, "step": 58660 }, { "grad_norm": 1.0146645307540894, "learning_rate": 3.987035012387213e-05, "loss": 0.0436, "step": 58670 }, { "grad_norm": 0.5582222938537598, "learning_rate": 3.9854158859450735e-05, "loss": 0.0413, "step": 58680 }, { "grad_norm": 0.40004006028175354, "learning_rate": 3.983796870456328e-05, "loss": 0.0454, "step": 58690 }, { "grad_norm": 0.34295111894607544, "learning_rate": 3.982177966098028e-05, "loss": 0.0359, "step": 58700 }, { "grad_norm": 0.5099659562110901, "learning_rate": 3.980559173047216e-05, "loss": 0.0515, "step": 58710 }, { "grad_norm": 0.5960479974746704, "learning_rate": 3.9789404914809176e-05, "loss": 0.0467, "step": 58720 }, { "grad_norm": 0.4221060872077942, "learning_rate": 3.9773219215761544e-05, "loss": 0.044, "step": 58730 }, { "grad_norm": 0.31494060158729553, "learning_rate": 3.975703463509928e-05, "loss": 0.0346, "step": 58740 }, { "grad_norm": 0.8550887107849121, "learning_rate": 3.974085117459231e-05, "loss": 0.0433, "step": 58750 }, { "grad_norm": 0.4069550931453705, "learning_rate": 3.972466883601043e-05, "loss": 0.0465, "step": 58760 }, { "grad_norm": 0.32553642988204956, "learning_rate": 3.9708487621123316e-05, "loss": 0.0305, "step": 58770 }, { "grad_norm": 0.48233762383461, "learning_rate": 3.969230753170051e-05, "loss": 0.0525, "step": 58780 }, { "grad_norm": 0.8071359992027283, "learning_rate": 3.967612856951146e-05, "loss": 0.0369, "step": 58790 }, { "grad_norm": 0.6443824768066406, "learning_rate": 3.9659950736325476e-05, "loss": 0.0413, "step": 58800 }, { "grad_norm": 1.3523352146148682, "learning_rate": 3.964377403391174e-05, "loss": 0.078, "step": 58810 }, { "grad_norm": 0.6209058165550232, "learning_rate": 3.9627598464039303e-05, "loss": 0.0638, "step": 58820 }, { "grad_norm": 0.38647544384002686, "learning_rate": 3.9611424028477096e-05, "loss": 0.036, "step": 58830 }, { "grad_norm": 0.610429048538208, "learning_rate": 3.959525072899394e-05, "loss": 0.0479, "step": 58840 }, { "grad_norm": 0.3937910199165344, "learning_rate": 3.9579078567358506e-05, "loss": 0.0364, "step": 58850 }, { "grad_norm": 0.5715333223342896, "learning_rate": 3.956290754533939e-05, "loss": 0.0408, "step": 58860 }, { "grad_norm": 0.4518420398235321, "learning_rate": 3.954673766470501e-05, "loss": 0.042, "step": 58870 }, { "grad_norm": 0.4952488839626312, "learning_rate": 3.9530568927223674e-05, "loss": 0.0394, "step": 58880 }, { "grad_norm": 0.3523675501346588, "learning_rate": 3.95144013346636e-05, "loss": 0.0508, "step": 58890 }, { "grad_norm": 0.7400800585746765, "learning_rate": 3.949823488879282e-05, "loss": 0.0327, "step": 58900 }, { "grad_norm": 0.3611091375350952, "learning_rate": 3.9482069591379276e-05, "loss": 0.0454, "step": 58910 }, { "grad_norm": 0.38603898882865906, "learning_rate": 3.9465905444190784e-05, "loss": 0.034, "step": 58920 }, { "grad_norm": 0.33409926295280457, "learning_rate": 3.944974244899505e-05, "loss": 0.0501, "step": 58930 }, { "grad_norm": 0.643771231174469, "learning_rate": 3.9433580607559616e-05, "loss": 0.0426, "step": 58940 }, { "grad_norm": 0.36627402901649475, "learning_rate": 3.941741992165193e-05, "loss": 0.0388, "step": 58950 }, { "grad_norm": 0.5974752306938171, "learning_rate": 3.940126039303928e-05, "loss": 0.0425, "step": 58960 }, { "grad_norm": 0.4051295816898346, "learning_rate": 3.938510202348886e-05, "loss": 0.0437, "step": 58970 }, { "grad_norm": 0.47949686646461487, "learning_rate": 3.93689448147677e-05, "loss": 0.0411, "step": 58980 }, { "grad_norm": 0.3738384246826172, "learning_rate": 3.935278876864278e-05, "loss": 0.0402, "step": 58990 }, { "grad_norm": 0.45851778984069824, "learning_rate": 3.933663388688087e-05, "loss": 0.0471, "step": 59000 }, { "grad_norm": 0.5359424948692322, "learning_rate": 3.9320480171248645e-05, "loss": 0.0403, "step": 59010 }, { "grad_norm": 0.7170045375823975, "learning_rate": 3.930432762351265e-05, "loss": 0.046, "step": 59020 }, { "grad_norm": 0.46758657693862915, "learning_rate": 3.928817624543931e-05, "loss": 0.044, "step": 59030 }, { "grad_norm": 0.31890395283699036, "learning_rate": 3.9272026038794885e-05, "loss": 0.0329, "step": 59040 }, { "grad_norm": 0.5402324795722961, "learning_rate": 3.9255877005345586e-05, "loss": 0.0436, "step": 59050 }, { "grad_norm": 0.5512139201164246, "learning_rate": 3.923972914685743e-05, "loss": 0.0464, "step": 59060 }, { "grad_norm": 0.41582736372947693, "learning_rate": 3.922358246509631e-05, "loss": 0.0375, "step": 59070 }, { "grad_norm": 0.5115801095962524, "learning_rate": 3.920743696182801e-05, "loss": 0.0473, "step": 59080 }, { "grad_norm": 0.3907909095287323, "learning_rate": 3.9191292638818164e-05, "loss": 0.0417, "step": 59090 }, { "grad_norm": 0.687910795211792, "learning_rate": 3.9175149497832316e-05, "loss": 0.0391, "step": 59100 }, { "grad_norm": 0.7161940932273865, "learning_rate": 3.915900754063584e-05, "loss": 0.0491, "step": 59110 }, { "grad_norm": 0.4197298288345337, "learning_rate": 3.9142866768994014e-05, "loss": 0.0382, "step": 59120 }, { "grad_norm": 0.42071962356567383, "learning_rate": 3.912672718467193e-05, "loss": 0.0475, "step": 59130 }, { "grad_norm": 0.4187988042831421, "learning_rate": 3.911058878943463e-05, "loss": 0.0426, "step": 59140 }, { "grad_norm": 0.6755034327507019, "learning_rate": 3.9094451585046974e-05, "loss": 0.034, "step": 59150 }, { "grad_norm": 0.3566739857196808, "learning_rate": 3.90783155732737e-05, "loss": 0.0387, "step": 59160 }, { "grad_norm": 0.5499334931373596, "learning_rate": 3.9062180755879373e-05, "loss": 0.039, "step": 59170 }, { "grad_norm": 0.8529960513114929, "learning_rate": 3.904604713462855e-05, "loss": 0.05, "step": 59180 }, { "grad_norm": 0.35340410470962524, "learning_rate": 3.902991471128554e-05, "loss": 0.0269, "step": 59190 }, { "grad_norm": 0.22471418976783752, "learning_rate": 3.901378348761456e-05, "loss": 0.0344, "step": 59200 }, { "grad_norm": 0.35610321164131165, "learning_rate": 3.899765346537971e-05, "loss": 0.0295, "step": 59210 }, { "grad_norm": 0.3376423418521881, "learning_rate": 3.898152464634493e-05, "loss": 0.0343, "step": 59220 }, { "grad_norm": 0.31693172454833984, "learning_rate": 3.896539703227402e-05, "loss": 0.0371, "step": 59230 }, { "grad_norm": 0.8087718486785889, "learning_rate": 3.894927062493073e-05, "loss": 0.0473, "step": 59240 }, { "grad_norm": 0.6040167808532715, "learning_rate": 3.893314542607859e-05, "loss": 0.0468, "step": 59250 }, { "grad_norm": 0.4519032835960388, "learning_rate": 3.891702143748103e-05, "loss": 0.0434, "step": 59260 }, { "grad_norm": 0.40920206904411316, "learning_rate": 3.890089866090133e-05, "loss": 0.0416, "step": 59270 }, { "grad_norm": 0.47325196862220764, "learning_rate": 3.888477709810267e-05, "loss": 0.0364, "step": 59280 }, { "grad_norm": 0.47991493344306946, "learning_rate": 3.886865675084806e-05, "loss": 0.0445, "step": 59290 }, { "grad_norm": 0.467601478099823, "learning_rate": 3.885253762090043e-05, "loss": 0.0461, "step": 59300 }, { "grad_norm": 0.5266045331954956, "learning_rate": 3.8836419710022534e-05, "loss": 0.0523, "step": 59310 }, { "grad_norm": 0.5623378753662109, "learning_rate": 3.882030301997698e-05, "loss": 0.0445, "step": 59320 }, { "grad_norm": 0.4980355501174927, "learning_rate": 3.880418755252629e-05, "loss": 0.0351, "step": 59330 }, { "grad_norm": 0.2861303389072418, "learning_rate": 3.87880733094328e-05, "loss": 0.0513, "step": 59340 }, { "grad_norm": 0.6820107102394104, "learning_rate": 3.877196029245877e-05, "loss": 0.0373, "step": 59350 }, { "grad_norm": 0.47686490416526794, "learning_rate": 3.875584850336627e-05, "loss": 0.0559, "step": 59360 }, { "grad_norm": 0.3933711051940918, "learning_rate": 3.873973794391728e-05, "loss": 0.037, "step": 59370 }, { "grad_norm": 0.3417309522628784, "learning_rate": 3.872362861587361e-05, "loss": 0.0315, "step": 59380 }, { "grad_norm": 0.4610939025878906, "learning_rate": 3.870752052099698e-05, "loss": 0.0443, "step": 59390 }, { "grad_norm": 0.44523555040359497, "learning_rate": 3.8691413661048916e-05, "loss": 0.0499, "step": 59400 }, { "grad_norm": 1.1008292436599731, "learning_rate": 3.867530803779085e-05, "loss": 0.0436, "step": 59410 }, { "grad_norm": 0.31826192140579224, "learning_rate": 3.8659203652984054e-05, "loss": 0.0395, "step": 59420 }, { "grad_norm": 0.41039928793907166, "learning_rate": 3.864310050838972e-05, "loss": 0.0423, "step": 59430 }, { "grad_norm": 0.3159467279911041, "learning_rate": 3.8626998605768835e-05, "loss": 0.0413, "step": 59440 }, { "grad_norm": 0.5093991756439209, "learning_rate": 3.8610897946882294e-05, "loss": 0.0456, "step": 59450 }, { "grad_norm": 0.5965248346328735, "learning_rate": 3.859479853349082e-05, "loss": 0.0453, "step": 59460 }, { "grad_norm": 0.9736264944076538, "learning_rate": 3.857870036735503e-05, "loss": 0.0412, "step": 59470 }, { "grad_norm": 0.790847897529602, "learning_rate": 3.856260345023538e-05, "loss": 0.0522, "step": 59480 }, { "grad_norm": 0.3870275616645813, "learning_rate": 3.854650778389224e-05, "loss": 0.0353, "step": 59490 }, { "grad_norm": 0.4672962427139282, "learning_rate": 3.8530413370085786e-05, "loss": 0.0349, "step": 59500 }, { "grad_norm": 0.3705674409866333, "learning_rate": 3.851432021057608e-05, "loss": 0.0393, "step": 59510 }, { "grad_norm": 0.4442751705646515, "learning_rate": 3.849822830712306e-05, "loss": 0.0425, "step": 59520 }, { "grad_norm": 0.5080272555351257, "learning_rate": 3.848213766148649e-05, "loss": 0.0409, "step": 59530 }, { "grad_norm": 0.6270425319671631, "learning_rate": 3.846604827542602e-05, "loss": 0.0345, "step": 59540 }, { "grad_norm": 0.5783002972602844, "learning_rate": 3.844996015070117e-05, "loss": 0.0418, "step": 59550 }, { "grad_norm": 0.33946526050567627, "learning_rate": 3.843387328907132e-05, "loss": 0.033, "step": 59560 }, { "grad_norm": 0.45118090510368347, "learning_rate": 3.8417787692295696e-05, "loss": 0.05, "step": 59570 }, { "grad_norm": 0.6074469685554504, "learning_rate": 3.84017033621334e-05, "loss": 0.0409, "step": 59580 }, { "grad_norm": 0.2171657830476761, "learning_rate": 3.838562030034338e-05, "loss": 0.0337, "step": 59590 }, { "grad_norm": 0.4809408485889435, "learning_rate": 3.836953850868447e-05, "loss": 0.0352, "step": 59600 }, { "grad_norm": 0.3811325132846832, "learning_rate": 3.835345798891533e-05, "loss": 0.0337, "step": 59610 }, { "grad_norm": 0.43079304695129395, "learning_rate": 3.833737874279452e-05, "loss": 0.0385, "step": 59620 }, { "grad_norm": 0.47059378027915955, "learning_rate": 3.8321300772080433e-05, "loss": 0.0288, "step": 59630 }, { "grad_norm": 0.6988294720649719, "learning_rate": 3.8305224078531345e-05, "loss": 0.0418, "step": 59640 }, { "grad_norm": 0.5821465253829956, "learning_rate": 3.828914866390537e-05, "loss": 0.0518, "step": 59650 }, { "grad_norm": 0.3990427255630493, "learning_rate": 3.827307452996048e-05, "loss": 0.0389, "step": 59660 }, { "grad_norm": 0.7241541743278503, "learning_rate": 3.825700167845454e-05, "loss": 0.0548, "step": 59670 }, { "grad_norm": 0.36575719714164734, "learning_rate": 3.8240930111145214e-05, "loss": 0.0499, "step": 59680 }, { "grad_norm": 0.4236461818218231, "learning_rate": 3.8224859829790116e-05, "loss": 0.0377, "step": 59690 }, { "grad_norm": 0.458583265542984, "learning_rate": 3.820879083614664e-05, "loss": 0.0509, "step": 59700 }, { "grad_norm": 0.5716450810432434, "learning_rate": 3.819272313197208e-05, "loss": 0.0438, "step": 59710 }, { "grad_norm": 0.5007369518280029, "learning_rate": 3.817665671902355e-05, "loss": 0.04, "step": 59720 }, { "grad_norm": 0.4878799617290497, "learning_rate": 3.816059159905808e-05, "loss": 0.0518, "step": 59730 }, { "grad_norm": 0.6031111478805542, "learning_rate": 3.8144527773832476e-05, "loss": 0.0429, "step": 59740 }, { "grad_norm": 0.5363399982452393, "learning_rate": 3.812846524510352e-05, "loss": 0.0285, "step": 59750 }, { "grad_norm": 0.4822993278503418, "learning_rate": 3.811240401462775e-05, "loss": 0.0458, "step": 59760 }, { "grad_norm": 0.5350284576416016, "learning_rate": 3.809634408416162e-05, "loss": 0.0403, "step": 59770 }, { "grad_norm": 0.4677117168903351, "learning_rate": 3.8080285455461394e-05, "loss": 0.0421, "step": 59780 }, { "grad_norm": 0.5680740475654602, "learning_rate": 3.806422813028323e-05, "loss": 0.0326, "step": 59790 }, { "grad_norm": 0.4621516764163971, "learning_rate": 3.8048172110383114e-05, "loss": 0.0351, "step": 59800 }, { "grad_norm": 0.3454470932483673, "learning_rate": 3.8032117397516944e-05, "loss": 0.0393, "step": 59810 }, { "grad_norm": 0.4662776589393616, "learning_rate": 3.801606399344042e-05, "loss": 0.0469, "step": 59820 }, { "grad_norm": 0.3536212742328644, "learning_rate": 3.800001189990913e-05, "loss": 0.0447, "step": 59830 }, { "grad_norm": 0.580406129360199, "learning_rate": 3.798396111867847e-05, "loss": 0.0472, "step": 59840 }, { "grad_norm": 0.678582489490509, "learning_rate": 3.796791165150378e-05, "loss": 0.0383, "step": 59850 }, { "grad_norm": 0.6094356179237366, "learning_rate": 3.795186350014018e-05, "loss": 0.0414, "step": 59860 }, { "grad_norm": 0.32416045665740967, "learning_rate": 3.793581666634266e-05, "loss": 0.0427, "step": 59870 }, { "grad_norm": 0.721973717212677, "learning_rate": 3.791977115186609e-05, "loss": 0.0436, "step": 59880 }, { "grad_norm": 0.3636070787906647, "learning_rate": 3.79037269584652e-05, "loss": 0.0318, "step": 59890 }, { "grad_norm": 0.5660223364830017, "learning_rate": 3.7887684087894545e-05, "loss": 0.0415, "step": 59900 }, { "grad_norm": 0.301870197057724, "learning_rate": 3.787164254190854e-05, "loss": 0.0377, "step": 59910 }, { "grad_norm": 0.4014618396759033, "learning_rate": 3.7855602322261473e-05, "loss": 0.0404, "step": 59920 }, { "grad_norm": 0.30756044387817383, "learning_rate": 3.783956343070746e-05, "loss": 0.0527, "step": 59930 }, { "grad_norm": 1.522948145866394, "learning_rate": 3.7823525869000526e-05, "loss": 0.0418, "step": 59940 }, { "grad_norm": 0.6177307963371277, "learning_rate": 3.78074896388945e-05, "loss": 0.0403, "step": 59950 }, { "grad_norm": 0.643233060836792, "learning_rate": 3.7791454742143075e-05, "loss": 0.0495, "step": 59960 }, { "grad_norm": 0.5189913511276245, "learning_rate": 3.7775421180499795e-05, "loss": 0.037, "step": 59970 }, { "grad_norm": 0.31299665570259094, "learning_rate": 3.7759388955718086e-05, "loss": 0.0423, "step": 59980 }, { "grad_norm": 0.3067038655281067, "learning_rate": 3.774335806955117e-05, "loss": 0.0407, "step": 59990 }, { "grad_norm": 0.432115375995636, "learning_rate": 3.7727328523752215e-05, "loss": 0.0391, "step": 60000 }, { "grad_norm": 0.4193793535232544, "learning_rate": 3.771130032007415e-05, "loss": 0.0387, "step": 60010 }, { "grad_norm": 0.5558561086654663, "learning_rate": 3.769527346026982e-05, "loss": 0.0326, "step": 60020 }, { "grad_norm": 0.4762921929359436, "learning_rate": 3.767924794609188e-05, "loss": 0.0412, "step": 60030 }, { "grad_norm": 0.5884876251220703, "learning_rate": 3.7663223779292854e-05, "loss": 0.0456, "step": 60040 }, { "grad_norm": 0.3905280828475952, "learning_rate": 3.764720096162512e-05, "loss": 0.0434, "step": 60050 }, { "grad_norm": 0.5663606524467468, "learning_rate": 3.763117949484092e-05, "loss": 0.0367, "step": 60060 }, { "grad_norm": 0.6778584122657776, "learning_rate": 3.7615159380692344e-05, "loss": 0.0427, "step": 60070 }, { "grad_norm": 0.39801836013793945, "learning_rate": 3.759914062093132e-05, "loss": 0.0299, "step": 60080 }, { "grad_norm": 0.434017151594162, "learning_rate": 3.7583123217309615e-05, "loss": 0.0379, "step": 60090 }, { "grad_norm": 0.4100404381752014, "learning_rate": 3.7567107171578904e-05, "loss": 0.0495, "step": 60100 }, { "grad_norm": 0.7570252418518066, "learning_rate": 3.755109248549066e-05, "loss": 0.0466, "step": 60110 }, { "grad_norm": 0.4575861394405365, "learning_rate": 3.75350791607962e-05, "loss": 0.0348, "step": 60120 }, { "grad_norm": 0.7603968977928162, "learning_rate": 3.751906719924676e-05, "loss": 0.0428, "step": 60130 }, { "grad_norm": 0.402896910905838, "learning_rate": 3.750305660259337e-05, "loss": 0.037, "step": 60140 }, { "grad_norm": 0.3515462577342987, "learning_rate": 3.748704737258693e-05, "loss": 0.0375, "step": 60150 }, { "grad_norm": 0.3799891173839569, "learning_rate": 3.747103951097816e-05, "loss": 0.0554, "step": 60160 }, { "grad_norm": 0.5820413827896118, "learning_rate": 3.745503301951768e-05, "loss": 0.0441, "step": 60170 }, { "grad_norm": 0.7874883413314819, "learning_rate": 3.743902789995592e-05, "loss": 0.0438, "step": 60180 }, { "grad_norm": 0.4764115512371063, "learning_rate": 3.7423024154043183e-05, "loss": 0.0329, "step": 60190 }, { "grad_norm": 0.6478261947631836, "learning_rate": 3.740702178352963e-05, "loss": 0.035, "step": 60200 }, { "grad_norm": 1.5983048677444458, "learning_rate": 3.739102079016523e-05, "loss": 0.0441, "step": 60210 }, { "grad_norm": 0.3217673599720001, "learning_rate": 3.7375021175699846e-05, "loss": 0.0289, "step": 60220 }, { "grad_norm": 0.3783838748931885, "learning_rate": 3.735902294188317e-05, "loss": 0.0386, "step": 60230 }, { "grad_norm": 0.5382925868034363, "learning_rate": 3.734302609046473e-05, "loss": 0.0377, "step": 60240 }, { "grad_norm": 0.5434248447418213, "learning_rate": 3.732703062319391e-05, "loss": 0.0465, "step": 60250 }, { "grad_norm": 1.2339831590652466, "learning_rate": 3.731103654181999e-05, "loss": 0.0531, "step": 60260 }, { "grad_norm": 0.6208419799804688, "learning_rate": 3.729504384809203e-05, "loss": 0.0391, "step": 60270 }, { "grad_norm": 0.3928523361682892, "learning_rate": 3.727905254375898e-05, "loss": 0.0345, "step": 60280 }, { "grad_norm": 0.385294646024704, "learning_rate": 3.7263062630569614e-05, "loss": 0.0371, "step": 60290 }, { "grad_norm": 0.7864543199539185, "learning_rate": 3.724707411027256e-05, "loss": 0.0384, "step": 60300 }, { "grad_norm": 0.502978503704071, "learning_rate": 3.723108698461631e-05, "loss": 0.0335, "step": 60310 }, { "grad_norm": 0.38415640592575073, "learning_rate": 3.72151012553492e-05, "loss": 0.044, "step": 60320 }, { "grad_norm": 0.32947322726249695, "learning_rate": 3.719911692421939e-05, "loss": 0.0273, "step": 60330 }, { "grad_norm": 1.0000625848770142, "learning_rate": 3.71831339929749e-05, "loss": 0.0383, "step": 60340 }, { "grad_norm": 0.3989925682544708, "learning_rate": 3.716715246336362e-05, "loss": 0.0374, "step": 60350 }, { "grad_norm": 0.32409217953681946, "learning_rate": 3.715117233713324e-05, "loss": 0.0308, "step": 60360 }, { "grad_norm": 0.37257930636405945, "learning_rate": 3.713519361603135e-05, "loss": 0.0408, "step": 60370 }, { "grad_norm": 0.4408928453922272, "learning_rate": 3.711921630180532e-05, "loss": 0.045, "step": 60380 }, { "grad_norm": 0.36547568440437317, "learning_rate": 3.710324039620245e-05, "loss": 0.0376, "step": 60390 }, { "grad_norm": 0.5318403840065002, "learning_rate": 3.708726590096982e-05, "loss": 0.0414, "step": 60400 }, { "grad_norm": 0.4528782367706299, "learning_rate": 3.707129281785437e-05, "loss": 0.0362, "step": 60410 }, { "grad_norm": 0.6567926406860352, "learning_rate": 3.705532114860291e-05, "loss": 0.0437, "step": 60420 }, { "grad_norm": 0.5049139261245728, "learning_rate": 3.703935089496207e-05, "loss": 0.0597, "step": 60430 }, { "grad_norm": 0.4057851731777191, "learning_rate": 3.702338205867829e-05, "loss": 0.0307, "step": 60440 }, { "grad_norm": 0.42662909626960754, "learning_rate": 3.7007414641497976e-05, "loss": 0.0442, "step": 60450 }, { "grad_norm": 0.4922066330909729, "learning_rate": 3.699144864516726e-05, "loss": 0.0346, "step": 60460 }, { "grad_norm": 0.5469229221343994, "learning_rate": 3.6975484071432153e-05, "loss": 0.0367, "step": 60470 }, { "grad_norm": 0.4991283416748047, "learning_rate": 3.695952092203854e-05, "loss": 0.0483, "step": 60480 }, { "grad_norm": 0.5189239382743835, "learning_rate": 3.69435591987321e-05, "loss": 0.0492, "step": 60490 }, { "grad_norm": 0.7436213493347168, "learning_rate": 3.6927598903258374e-05, "loss": 0.0428, "step": 60500 }, { "grad_norm": 1.4647351503372192, "learning_rate": 3.691164003736281e-05, "loss": 0.0324, "step": 60510 }, { "grad_norm": 0.3821192979812622, "learning_rate": 3.689568260279061e-05, "loss": 0.039, "step": 60520 }, { "grad_norm": 0.3623894155025482, "learning_rate": 3.687972660128686e-05, "loss": 0.0429, "step": 60530 }, { "grad_norm": 0.6020617485046387, "learning_rate": 3.686377203459648e-05, "loss": 0.0463, "step": 60540 }, { "grad_norm": 0.4292658865451813, "learning_rate": 3.6847818904464226e-05, "loss": 0.0397, "step": 60550 }, { "grad_norm": 0.49489259719848633, "learning_rate": 3.683186721263474e-05, "loss": 0.0379, "step": 60560 }, { "grad_norm": 0.46914142370224, "learning_rate": 3.6815916960852435e-05, "loss": 0.0473, "step": 60570 }, { "grad_norm": 0.7816373109817505, "learning_rate": 3.679996815086165e-05, "loss": 0.0411, "step": 60580 }, { "grad_norm": 0.45290425419807434, "learning_rate": 3.678402078440649e-05, "loss": 0.0434, "step": 60590 }, { "grad_norm": 0.4204246997833252, "learning_rate": 3.676807486323096e-05, "loss": 0.0416, "step": 60600 }, { "grad_norm": 0.5029093027114868, "learning_rate": 3.6752130389078864e-05, "loss": 0.0409, "step": 60610 }, { "grad_norm": 0.5220350623130798, "learning_rate": 3.673618736369388e-05, "loss": 0.0342, "step": 60620 }, { "grad_norm": 0.3680122196674347, "learning_rate": 3.6720245788819474e-05, "loss": 0.0372, "step": 60630 }, { "grad_norm": 0.7011322975158691, "learning_rate": 3.6704305666199044e-05, "loss": 0.049, "step": 60640 }, { "grad_norm": 0.8429995179176331, "learning_rate": 3.668836699757576e-05, "loss": 0.0358, "step": 60650 }, { "grad_norm": 0.8324047327041626, "learning_rate": 3.667242978469265e-05, "loss": 0.044, "step": 60660 }, { "grad_norm": 0.6039121747016907, "learning_rate": 3.6656494029292575e-05, "loss": 0.0485, "step": 60670 }, { "grad_norm": 0.5432902574539185, "learning_rate": 3.664055973311825e-05, "loss": 0.0389, "step": 60680 }, { "grad_norm": 0.4051414132118225, "learning_rate": 3.662462689791221e-05, "loss": 0.0351, "step": 60690 }, { "grad_norm": 0.3911128640174866, "learning_rate": 3.660869552541689e-05, "loss": 0.0319, "step": 60700 }, { "grad_norm": 1.3857356309890747, "learning_rate": 3.659276561737448e-05, "loss": 0.04, "step": 60710 }, { "grad_norm": 0.3503914475440979, "learning_rate": 3.657683717552708e-05, "loss": 0.0398, "step": 60720 }, { "grad_norm": 0.7190766334533691, "learning_rate": 3.6560910201616574e-05, "loss": 0.0322, "step": 60730 }, { "grad_norm": 0.3873461186885834, "learning_rate": 3.654498469738472e-05, "loss": 0.037, "step": 60740 }, { "grad_norm": 0.597050130367279, "learning_rate": 3.6529060664573126e-05, "loss": 0.0452, "step": 60750 }, { "grad_norm": 0.43460091948509216, "learning_rate": 3.6513138104923176e-05, "loss": 0.0423, "step": 60760 }, { "grad_norm": 0.43131932616233826, "learning_rate": 3.6497217020176176e-05, "loss": 0.037, "step": 60770 }, { "grad_norm": 0.4703100025653839, "learning_rate": 3.648129741207323e-05, "loss": 0.0425, "step": 60780 }, { "grad_norm": 0.30232781171798706, "learning_rate": 3.646537928235527e-05, "loss": 0.0339, "step": 60790 }, { "grad_norm": 0.5108146667480469, "learning_rate": 3.6449462632763075e-05, "loss": 0.0398, "step": 60800 }, { "grad_norm": 0.6240545511245728, "learning_rate": 3.643354746503729e-05, "loss": 0.037, "step": 60810 }, { "grad_norm": 0.8595520257949829, "learning_rate": 3.6417633780918345e-05, "loss": 0.0309, "step": 60820 }, { "grad_norm": 0.6092502474784851, "learning_rate": 3.640172158214655e-05, "loss": 0.0453, "step": 60830 }, { "grad_norm": 0.6824521422386169, "learning_rate": 3.6385810870462035e-05, "loss": 0.0311, "step": 60840 }, { "grad_norm": 0.5751546621322632, "learning_rate": 3.636990164760479e-05, "loss": 0.0334, "step": 60850 }, { "grad_norm": 0.5635437965393066, "learning_rate": 3.6353993915314596e-05, "loss": 0.0381, "step": 60860 }, { "grad_norm": 0.7084916830062866, "learning_rate": 3.6338087675331124e-05, "loss": 0.036, "step": 60870 }, { "grad_norm": 0.44740718603134155, "learning_rate": 3.632218292939383e-05, "loss": 0.0313, "step": 60880 }, { "grad_norm": 0.4956822991371155, "learning_rate": 3.6306279679242033e-05, "loss": 0.033, "step": 60890 }, { "grad_norm": 0.44477593898773193, "learning_rate": 3.629037792661492e-05, "loss": 0.0498, "step": 60900 }, { "grad_norm": 0.5511406064033508, "learning_rate": 3.627447767325147e-05, "loss": 0.0431, "step": 60910 }, { "grad_norm": 0.69029301404953, "learning_rate": 3.62585789208905e-05, "loss": 0.0371, "step": 60920 }, { "grad_norm": 0.41285908222198486, "learning_rate": 3.624268167127068e-05, "loss": 0.0546, "step": 60930 }, { "grad_norm": 0.36896681785583496, "learning_rate": 3.622678592613051e-05, "loss": 0.0383, "step": 60940 }, { "grad_norm": 0.7527371048927307, "learning_rate": 3.6210891687208304e-05, "loss": 0.0411, "step": 60950 }, { "grad_norm": 0.843264639377594, "learning_rate": 3.619499895624228e-05, "loss": 0.0349, "step": 60960 }, { "grad_norm": 0.4272044897079468, "learning_rate": 3.61791077349704e-05, "loss": 0.0411, "step": 60970 }, { "grad_norm": 0.5168490409851074, "learning_rate": 3.616321802513053e-05, "loss": 0.0342, "step": 60980 }, { "grad_norm": 0.7046256065368652, "learning_rate": 3.6147329828460325e-05, "loss": 0.0488, "step": 60990 }, { "grad_norm": 0.8237344622612, "learning_rate": 3.6131443146697306e-05, "loss": 0.0398, "step": 61000 }, { "grad_norm": 0.6070957183837891, "learning_rate": 3.611555798157879e-05, "loss": 0.0368, "step": 61010 }, { "grad_norm": 0.5379728674888611, "learning_rate": 3.6099674334842e-05, "loss": 0.0436, "step": 61020 }, { "grad_norm": 0.42867058515548706, "learning_rate": 3.6083792208223934e-05, "loss": 0.0542, "step": 61030 }, { "grad_norm": 0.3518476188182831, "learning_rate": 3.606791160346142e-05, "loss": 0.0434, "step": 61040 }, { "grad_norm": 0.24162551760673523, "learning_rate": 3.6052032522291134e-05, "loss": 0.038, "step": 61050 }, { "grad_norm": 0.7365055680274963, "learning_rate": 3.603615496644962e-05, "loss": 0.0516, "step": 61060 }, { "grad_norm": 0.6113377213478088, "learning_rate": 3.60202789376732e-05, "loss": 0.0436, "step": 61070 }, { "grad_norm": 0.470741868019104, "learning_rate": 3.6004404437698036e-05, "loss": 0.0417, "step": 61080 }, { "grad_norm": 0.4333055019378662, "learning_rate": 3.5988531468260184e-05, "loss": 0.0392, "step": 61090 }, { "grad_norm": 0.7146710753440857, "learning_rate": 3.597266003109546e-05, "loss": 0.0394, "step": 61100 }, { "grad_norm": 0.4404386579990387, "learning_rate": 3.5956790127939557e-05, "loss": 0.0427, "step": 61110 }, { "grad_norm": 0.47749483585357666, "learning_rate": 3.594092176052796e-05, "loss": 0.0406, "step": 61120 }, { "grad_norm": 0.32547295093536377, "learning_rate": 3.592505493059603e-05, "loss": 0.0415, "step": 61130 }, { "grad_norm": 0.45761367678642273, "learning_rate": 3.5909189639878905e-05, "loss": 0.0453, "step": 61140 }, { "grad_norm": 0.3276096284389496, "learning_rate": 3.5893325890111626e-05, "loss": 0.0358, "step": 61150 }, { "grad_norm": 0.5596778988838196, "learning_rate": 3.5877463683029034e-05, "loss": 0.047, "step": 61160 }, { "grad_norm": 0.29930055141448975, "learning_rate": 3.586160302036578e-05, "loss": 0.0389, "step": 61170 }, { "grad_norm": 0.2937772274017334, "learning_rate": 3.584574390385636e-05, "loss": 0.0318, "step": 61180 }, { "grad_norm": 0.3028520941734314, "learning_rate": 3.582988633523511e-05, "loss": 0.0453, "step": 61190 }, { "grad_norm": 0.4098735451698303, "learning_rate": 3.581403031623616e-05, "loss": 0.0419, "step": 61200 }, { "grad_norm": 0.5306563973426819, "learning_rate": 3.579817584859354e-05, "loss": 0.0396, "step": 61210 }, { "grad_norm": 0.507095456123352, "learning_rate": 3.578232293404107e-05, "loss": 0.0369, "step": 61220 }, { "grad_norm": 0.6521900296211243, "learning_rate": 3.5766471574312385e-05, "loss": 0.036, "step": 61230 }, { "grad_norm": 0.46852391958236694, "learning_rate": 3.5750621771140964e-05, "loss": 0.0459, "step": 61240 }, { "grad_norm": 0.5628983974456787, "learning_rate": 3.5734773526260115e-05, "loss": 0.0372, "step": 61250 }, { "grad_norm": 0.4036934971809387, "learning_rate": 3.5718926841402996e-05, "loss": 0.0393, "step": 61260 }, { "grad_norm": 0.49412402510643005, "learning_rate": 3.5703081718302555e-05, "loss": 0.0457, "step": 61270 }, { "grad_norm": 0.47532570362091064, "learning_rate": 3.568723815869162e-05, "loss": 0.0436, "step": 61280 }, { "grad_norm": 0.3047861158847809, "learning_rate": 3.5671396164302777e-05, "loss": 0.0627, "step": 61290 }, { "grad_norm": 0.8021506071090698, "learning_rate": 3.565555573686853e-05, "loss": 0.0362, "step": 61300 }, { "grad_norm": 1.3470724821090698, "learning_rate": 3.563971687812113e-05, "loss": 0.037, "step": 61310 }, { "grad_norm": 0.872588574886322, "learning_rate": 3.5623879589792706e-05, "loss": 0.044, "step": 61320 }, { "grad_norm": 0.25775718688964844, "learning_rate": 3.5608043873615185e-05, "loss": 0.0468, "step": 61330 }, { "grad_norm": 0.39234641194343567, "learning_rate": 3.559220973132035e-05, "loss": 0.0334, "step": 61340 }, { "grad_norm": 0.450467586517334, "learning_rate": 3.557637716463981e-05, "loss": 0.044, "step": 61350 }, { "grad_norm": 0.7363805174827576, "learning_rate": 3.556054617530497e-05, "loss": 0.0418, "step": 61360 }, { "grad_norm": 0.39055749773979187, "learning_rate": 3.5544716765047106e-05, "loss": 0.0384, "step": 61370 }, { "grad_norm": 0.520939826965332, "learning_rate": 3.5528888935597274e-05, "loss": 0.0378, "step": 61380 }, { "grad_norm": 0.6062943935394287, "learning_rate": 3.551306268868636e-05, "loss": 0.0403, "step": 61390 }, { "grad_norm": 0.32962730526924133, "learning_rate": 3.5497238026045166e-05, "loss": 0.0371, "step": 61400 }, { "grad_norm": 0.43840664625167847, "learning_rate": 3.548141494940421e-05, "loss": 0.0418, "step": 61410 }, { "grad_norm": 0.7416201233863831, "learning_rate": 3.546559346049388e-05, "loss": 0.0431, "step": 61420 }, { "grad_norm": 0.29762500524520874, "learning_rate": 3.544977356104441e-05, "loss": 0.0361, "step": 61430 }, { "grad_norm": 0.3616940975189209, "learning_rate": 3.543395525278582e-05, "loss": 0.0417, "step": 61440 }, { "grad_norm": 0.48806944489479065, "learning_rate": 3.5418138537447985e-05, "loss": 0.0385, "step": 61450 }, { "grad_norm": 0.3219585716724396, "learning_rate": 3.5402323416760574e-05, "loss": 0.0342, "step": 61460 }, { "grad_norm": 0.26936450600624084, "learning_rate": 3.538650989245316e-05, "loss": 0.0318, "step": 61470 }, { "grad_norm": 0.7039703726768494, "learning_rate": 3.537069796625504e-05, "loss": 0.0331, "step": 61480 }, { "grad_norm": 0.5010887980461121, "learning_rate": 3.53548876398954e-05, "loss": 0.0404, "step": 61490 }, { "grad_norm": 0.5402124524116516, "learning_rate": 3.5339078915103214e-05, "loss": 0.0377, "step": 61500 }, { "grad_norm": 0.4365193843841553, "learning_rate": 3.532327179360733e-05, "loss": 0.048, "step": 61510 }, { "grad_norm": 0.4468770921230316, "learning_rate": 3.530746627713636e-05, "loss": 0.0335, "step": 61520 }, { "grad_norm": 0.44542959332466125, "learning_rate": 3.5291662367418805e-05, "loss": 0.0415, "step": 61530 }, { "grad_norm": 0.5476759076118469, "learning_rate": 3.527586006618293e-05, "loss": 0.0547, "step": 61540 }, { "grad_norm": 1.3625303506851196, "learning_rate": 3.526005937515687e-05, "loss": 0.0528, "step": 61550 }, { "grad_norm": 0.40066707134246826, "learning_rate": 3.524426029606856e-05, "loss": 0.0523, "step": 61560 }, { "grad_norm": 0.33517053723335266, "learning_rate": 3.5228462830645746e-05, "loss": 0.0389, "step": 61570 }, { "grad_norm": 0.5607145428657532, "learning_rate": 3.5212666980616036e-05, "loss": 0.038, "step": 61580 }, { "grad_norm": 0.48195523023605347, "learning_rate": 3.519687274770681e-05, "loss": 0.0493, "step": 61590 }, { "grad_norm": 0.36929574608802795, "learning_rate": 3.518108013364535e-05, "loss": 0.0547, "step": 61600 }, { "grad_norm": 0.5465283393859863, "learning_rate": 3.516528914015868e-05, "loss": 0.0443, "step": 61610 }, { "grad_norm": 0.47684481739997864, "learning_rate": 3.5149499768973694e-05, "loss": 0.0523, "step": 61620 }, { "grad_norm": 0.2877468764781952, "learning_rate": 3.513371202181708e-05, "loss": 0.0438, "step": 61630 }, { "grad_norm": 0.5840166211128235, "learning_rate": 3.511792590041537e-05, "loss": 0.0406, "step": 61640 }, { "grad_norm": 0.7650293111801147, "learning_rate": 3.510214140649488e-05, "loss": 0.0485, "step": 61650 }, { "grad_norm": 0.43142327666282654, "learning_rate": 3.508635854178184e-05, "loss": 0.0485, "step": 61660 }, { "grad_norm": 0.9470416307449341, "learning_rate": 3.50705773080022e-05, "loss": 0.0453, "step": 61670 }, { "grad_norm": 0.5317044854164124, "learning_rate": 3.505479770688178e-05, "loss": 0.0401, "step": 61680 }, { "grad_norm": 0.3499860167503357, "learning_rate": 3.503901974014622e-05, "loss": 0.0305, "step": 61690 }, { "grad_norm": 0.5329395532608032, "learning_rate": 3.502324340952095e-05, "loss": 0.0415, "step": 61700 }, { "grad_norm": 0.3032453656196594, "learning_rate": 3.500746871673126e-05, "loss": 0.0502, "step": 61710 }, { "grad_norm": 0.40255025029182434, "learning_rate": 3.499169566350227e-05, "loss": 0.0488, "step": 61720 }, { "grad_norm": 0.7677717208862305, "learning_rate": 3.4975924251558876e-05, "loss": 0.0355, "step": 61730 }, { "grad_norm": 0.4428887665271759, "learning_rate": 3.496015448262582e-05, "loss": 0.0406, "step": 61740 }, { "grad_norm": 0.6823180317878723, "learning_rate": 3.494438635842764e-05, "loss": 0.042, "step": 61750 }, { "grad_norm": 0.2623981237411499, "learning_rate": 3.492861988068874e-05, "loss": 0.0357, "step": 61760 }, { "grad_norm": 0.28876736760139465, "learning_rate": 3.491285505113332e-05, "loss": 0.0545, "step": 61770 }, { "grad_norm": 1.0864931344985962, "learning_rate": 3.4897091871485376e-05, "loss": 0.054, "step": 61780 }, { "grad_norm": 0.4172060489654541, "learning_rate": 3.488133034346875e-05, "loss": 0.0335, "step": 61790 }, { "grad_norm": 0.7425265908241272, "learning_rate": 3.486557046880713e-05, "loss": 0.0414, "step": 61800 }, { "grad_norm": 0.5118045210838318, "learning_rate": 3.484981224922395e-05, "loss": 0.0488, "step": 61810 }, { "grad_norm": 0.3507651686668396, "learning_rate": 3.4834055686442544e-05, "loss": 0.027, "step": 61820 }, { "grad_norm": 0.9221596121788025, "learning_rate": 3.4818300782186e-05, "loss": 0.0399, "step": 61830 }, { "grad_norm": 0.3856106996536255, "learning_rate": 3.4802547538177236e-05, "loss": 0.0327, "step": 61840 }, { "grad_norm": 0.7639452219009399, "learning_rate": 3.478679595613905e-05, "loss": 0.0355, "step": 61850 }, { "grad_norm": 0.3235023617744446, "learning_rate": 3.4771046037793975e-05, "loss": 0.038, "step": 61860 }, { "grad_norm": 0.3303067088127136, "learning_rate": 3.4755297784864416e-05, "loss": 0.0324, "step": 61870 }, { "grad_norm": 0.6025225520133972, "learning_rate": 3.473955119907257e-05, "loss": 0.038, "step": 61880 }, { "grad_norm": 0.3863365948200226, "learning_rate": 3.472380628214046e-05, "loss": 0.0354, "step": 61890 }, { "grad_norm": 0.4017904996871948, "learning_rate": 3.4708063035789906e-05, "loss": 0.0386, "step": 61900 }, { "grad_norm": 0.4244008958339691, "learning_rate": 3.4692321461742614e-05, "loss": 0.0401, "step": 61910 }, { "grad_norm": 0.31298449635505676, "learning_rate": 3.4676581561720023e-05, "loss": 0.0366, "step": 61920 }, { "grad_norm": 0.49869436025619507, "learning_rate": 3.466084333744344e-05, "loss": 0.0342, "step": 61930 }, { "grad_norm": 0.7373018860816956, "learning_rate": 3.4645106790633964e-05, "loss": 0.046, "step": 61940 }, { "grad_norm": 0.6534608006477356, "learning_rate": 3.4629371923012526e-05, "loss": 0.0497, "step": 61950 }, { "grad_norm": 0.43705156445503235, "learning_rate": 3.4613638736299856e-05, "loss": 0.0401, "step": 61960 }, { "grad_norm": 0.3196304738521576, "learning_rate": 3.459790723221652e-05, "loss": 0.0274, "step": 61970 }, { "grad_norm": 0.783573567867279, "learning_rate": 3.45821774124829e-05, "loss": 0.0474, "step": 61980 }, { "grad_norm": 0.46724218130111694, "learning_rate": 3.456644927881918e-05, "loss": 0.0409, "step": 61990 }, { "grad_norm": 0.3753223121166229, "learning_rate": 3.4550722832945354e-05, "loss": 0.0446, "step": 62000 }, { "grad_norm": 0.4005770981311798, "learning_rate": 3.453499807658126e-05, "loss": 0.0416, "step": 62010 }, { "grad_norm": 0.32771924138069153, "learning_rate": 3.451927501144653e-05, "loss": 0.0322, "step": 62020 }, { "grad_norm": 0.5118669867515564, "learning_rate": 3.450355363926059e-05, "loss": 0.0624, "step": 62030 }, { "grad_norm": 0.5883913636207581, "learning_rate": 3.448783396174273e-05, "loss": 0.0469, "step": 62040 }, { "grad_norm": 0.647215723991394, "learning_rate": 3.4472115980612034e-05, "loss": 0.0389, "step": 62050 }, { "grad_norm": 1.4027056694030762, "learning_rate": 3.4456399697587396e-05, "loss": 0.0405, "step": 62060 }, { "grad_norm": 0.44021663069725037, "learning_rate": 3.4440685114387506e-05, "loss": 0.0481, "step": 62070 }, { "grad_norm": 0.5637475252151489, "learning_rate": 3.442497223273091e-05, "loss": 0.0369, "step": 62080 }, { "grad_norm": 0.7135449647903442, "learning_rate": 3.4409261054335926e-05, "loss": 0.046, "step": 62090 }, { "grad_norm": 0.4629006087779999, "learning_rate": 3.439355158092069e-05, "loss": 0.0511, "step": 62100 }, { "grad_norm": 0.3967985212802887, "learning_rate": 3.437784381420322e-05, "loss": 0.0446, "step": 62110 }, { "grad_norm": 0.3045403063297272, "learning_rate": 3.436213775590126e-05, "loss": 0.0298, "step": 62120 }, { "grad_norm": 0.44400840997695923, "learning_rate": 3.4346433407732405e-05, "loss": 0.0383, "step": 62130 }, { "grad_norm": 0.3957447409629822, "learning_rate": 3.433073077141405e-05, "loss": 0.0258, "step": 62140 }, { "grad_norm": 0.6180782318115234, "learning_rate": 3.431502984866343e-05, "loss": 0.0434, "step": 62150 }, { "grad_norm": 0.3980492055416107, "learning_rate": 3.429933064119753e-05, "loss": 0.0576, "step": 62160 }, { "grad_norm": 1.0075230598449707, "learning_rate": 3.4283633150733255e-05, "loss": 0.0412, "step": 62170 }, { "grad_norm": 0.5325610041618347, "learning_rate": 3.4267937378987225e-05, "loss": 0.0476, "step": 62180 }, { "grad_norm": 0.6641397476196289, "learning_rate": 3.425224332767592e-05, "loss": 0.0369, "step": 62190 }, { "grad_norm": 0.7528337240219116, "learning_rate": 3.4236550998515595e-05, "loss": 0.0519, "step": 62200 }, { "grad_norm": 0.6056404113769531, "learning_rate": 3.422086039322235e-05, "loss": 0.0402, "step": 62210 }, { "grad_norm": 0.4173104166984558, "learning_rate": 3.420517151351209e-05, "loss": 0.0457, "step": 62220 }, { "grad_norm": 0.4384901821613312, "learning_rate": 3.418948436110053e-05, "loss": 0.0407, "step": 62230 }, { "grad_norm": 0.39518022537231445, "learning_rate": 3.417379893770319e-05, "loss": 0.0405, "step": 62240 }, { "grad_norm": 0.4904075562953949, "learning_rate": 3.41581152450354e-05, "loss": 0.0327, "step": 62250 }, { "grad_norm": 0.5845404863357544, "learning_rate": 3.414243328481231e-05, "loss": 0.0431, "step": 62260 }, { "grad_norm": 0.3965940475463867, "learning_rate": 3.412675305874888e-05, "loss": 0.0404, "step": 62270 }, { "grad_norm": 0.6068743467330933, "learning_rate": 3.4111074568559876e-05, "loss": 0.0377, "step": 62280 }, { "grad_norm": 0.45056286454200745, "learning_rate": 3.4095397815959846e-05, "loss": 0.0472, "step": 62290 }, { "grad_norm": 0.6137012243270874, "learning_rate": 3.407972280266322e-05, "loss": 0.0417, "step": 62300 }, { "grad_norm": 0.3188716769218445, "learning_rate": 3.4064049530384166e-05, "loss": 0.0398, "step": 62310 }, { "grad_norm": 0.9732717275619507, "learning_rate": 3.40483780008367e-05, "loss": 0.0378, "step": 62320 }, { "grad_norm": 0.29535308480262756, "learning_rate": 3.4032708215734635e-05, "loss": 0.0391, "step": 62330 }, { "grad_norm": 0.7376766204833984, "learning_rate": 3.4017040176791584e-05, "loss": 0.0404, "step": 62340 }, { "grad_norm": 0.501335859298706, "learning_rate": 3.4001373885720976e-05, "loss": 0.0511, "step": 62350 }, { "grad_norm": 0.371369868516922, "learning_rate": 3.398570934423608e-05, "loss": 0.0392, "step": 62360 }, { "grad_norm": 0.4943073093891144, "learning_rate": 3.397004655404994e-05, "loss": 0.052, "step": 62370 }, { "grad_norm": 0.36844199895858765, "learning_rate": 3.395438551687541e-05, "loss": 0.046, "step": 62380 }, { "grad_norm": 0.6038527488708496, "learning_rate": 3.393872623442515e-05, "loss": 0.0405, "step": 62390 }, { "grad_norm": 0.3220958411693573, "learning_rate": 3.392306870841164e-05, "loss": 0.0439, "step": 62400 }, { "grad_norm": 0.47300243377685547, "learning_rate": 3.390741294054715e-05, "loss": 0.0381, "step": 62410 }, { "grad_norm": 0.6071593761444092, "learning_rate": 3.38917589325438e-05, "loss": 0.0373, "step": 62420 }, { "grad_norm": 1.032541275024414, "learning_rate": 3.387610668611349e-05, "loss": 0.0508, "step": 62430 }, { "grad_norm": 0.7160457968711853, "learning_rate": 3.38604562029679e-05, "loss": 0.0484, "step": 62440 }, { "grad_norm": 0.4817902445793152, "learning_rate": 3.384480748481855e-05, "loss": 0.0397, "step": 62450 }, { "grad_norm": 0.3429311215877533, "learning_rate": 3.382916053337676e-05, "loss": 0.0382, "step": 62460 }, { "grad_norm": 0.3656647801399231, "learning_rate": 3.381351535035366e-05, "loss": 0.0473, "step": 62470 }, { "grad_norm": 0.5972380042076111, "learning_rate": 3.379787193746019e-05, "loss": 0.0449, "step": 62480 }, { "grad_norm": 0.4116555154323578, "learning_rate": 3.378223029640709e-05, "loss": 0.0396, "step": 62490 }, { "grad_norm": 0.4468621611595154, "learning_rate": 3.376659042890489e-05, "loss": 0.0432, "step": 62500 }, { "grad_norm": 0.45199933648109436, "learning_rate": 3.375095233666397e-05, "loss": 0.0427, "step": 62510 }, { "grad_norm": 0.6031293869018555, "learning_rate": 3.373531602139447e-05, "loss": 0.0383, "step": 62520 }, { "grad_norm": 0.8365116715431213, "learning_rate": 3.371968148480635e-05, "loss": 0.0405, "step": 62530 }, { "grad_norm": 1.2255386114120483, "learning_rate": 3.370404872860937e-05, "loss": 0.0429, "step": 62540 }, { "grad_norm": 0.417407751083374, "learning_rate": 3.3688417754513145e-05, "loss": 0.0406, "step": 62550 }, { "grad_norm": 0.381343811750412, "learning_rate": 3.367278856422703e-05, "loss": 0.0366, "step": 62560 }, { "grad_norm": 0.8815986514091492, "learning_rate": 3.365716115946021e-05, "loss": 0.0533, "step": 62570 }, { "grad_norm": 0.5957632660865784, "learning_rate": 3.3641535541921666e-05, "loss": 0.0555, "step": 62580 }, { "grad_norm": 0.6659436821937561, "learning_rate": 3.36259117133202e-05, "loss": 0.0439, "step": 62590 }, { "grad_norm": 0.4837273061275482, "learning_rate": 3.36102896753644e-05, "loss": 0.0319, "step": 62600 }, { "grad_norm": 0.5022212862968445, "learning_rate": 3.35946694297627e-05, "loss": 0.031, "step": 62610 }, { "grad_norm": 0.3226469159126282, "learning_rate": 3.357905097822328e-05, "loss": 0.0431, "step": 62620 }, { "grad_norm": 1.146855354309082, "learning_rate": 3.356343432245416e-05, "loss": 0.0415, "step": 62630 }, { "grad_norm": 0.7080380320549011, "learning_rate": 3.354781946416315e-05, "loss": 0.0595, "step": 62640 }, { "grad_norm": 0.5424690842628479, "learning_rate": 3.3532206405057864e-05, "loss": 0.0472, "step": 62650 }, { "grad_norm": 0.6527087688446045, "learning_rate": 3.351659514684573e-05, "loss": 0.0412, "step": 62660 }, { "grad_norm": 0.5808143019676208, "learning_rate": 3.350098569123395e-05, "loss": 0.0356, "step": 62670 }, { "grad_norm": 0.46431776881217957, "learning_rate": 3.348537803992958e-05, "loss": 0.0347, "step": 62680 }, { "grad_norm": 0.34827011823654175, "learning_rate": 3.346977219463945e-05, "loss": 0.0381, "step": 62690 }, { "grad_norm": 0.6140247583389282, "learning_rate": 3.345416815707017e-05, "loss": 0.0342, "step": 62700 }, { "grad_norm": 0.9017196297645569, "learning_rate": 3.3438565928928185e-05, "loss": 0.0503, "step": 62710 }, { "grad_norm": 0.6419437527656555, "learning_rate": 3.3422965511919736e-05, "loss": 0.0398, "step": 62720 }, { "grad_norm": 0.41653668880462646, "learning_rate": 3.3407366907750846e-05, "loss": 0.0398, "step": 62730 }, { "grad_norm": 0.6322270631790161, "learning_rate": 3.339177011812737e-05, "loss": 0.0504, "step": 62740 }, { "grad_norm": 0.4445699155330658, "learning_rate": 3.3376175144754935e-05, "loss": 0.0349, "step": 62750 }, { "grad_norm": 0.5315884947776794, "learning_rate": 3.3360581989339004e-05, "loss": 0.0395, "step": 62760 }, { "grad_norm": 0.23535093665122986, "learning_rate": 3.3344990653584804e-05, "loss": 0.0368, "step": 62770 }, { "grad_norm": 0.4162096083164215, "learning_rate": 3.332940113919739e-05, "loss": 0.0465, "step": 62780 }, { "grad_norm": 0.49716755747795105, "learning_rate": 3.331381344788161e-05, "loss": 0.0296, "step": 62790 }, { "grad_norm": 0.5527153611183167, "learning_rate": 3.3298227581342076e-05, "loss": 0.0408, "step": 62800 }, { "grad_norm": 0.6587285995483398, "learning_rate": 3.328264354128329e-05, "loss": 0.0487, "step": 62810 }, { "grad_norm": 0.7138145565986633, "learning_rate": 3.3267061329409465e-05, "loss": 0.0433, "step": 62820 }, { "grad_norm": 0.3621358275413513, "learning_rate": 3.325148094742466e-05, "loss": 0.042, "step": 62830 }, { "grad_norm": 0.4040829837322235, "learning_rate": 3.323590239703271e-05, "loss": 0.031, "step": 62840 }, { "grad_norm": 0.37011510133743286, "learning_rate": 3.322032567993728e-05, "loss": 0.0361, "step": 62850 }, { "grad_norm": 0.6297501921653748, "learning_rate": 3.3204750797841774e-05, "loss": 0.0397, "step": 62860 }, { "grad_norm": 0.31420549750328064, "learning_rate": 3.31891777524495e-05, "loss": 0.0367, "step": 62870 }, { "grad_norm": 0.9025535583496094, "learning_rate": 3.317360654546347e-05, "loss": 0.039, "step": 62880 }, { "grad_norm": 0.458645224571228, "learning_rate": 3.3158037178586534e-05, "loss": 0.0481, "step": 62890 }, { "grad_norm": 0.42540687322616577, "learning_rate": 3.3142469653521316e-05, "loss": 0.0374, "step": 62900 }, { "grad_norm": 0.9729782342910767, "learning_rate": 3.3126903971970284e-05, "loss": 0.0497, "step": 62910 }, { "grad_norm": 0.759310245513916, "learning_rate": 3.311134013563564e-05, "loss": 0.0481, "step": 62920 }, { "grad_norm": 0.7574853301048279, "learning_rate": 3.309577814621947e-05, "loss": 0.036, "step": 62930 }, { "grad_norm": 1.509188175201416, "learning_rate": 3.308021800542359e-05, "loss": 0.0361, "step": 62940 }, { "grad_norm": 0.4636932611465454, "learning_rate": 3.306465971494963e-05, "loss": 0.0372, "step": 62950 }, { "grad_norm": 0.9967347383499146, "learning_rate": 3.304910327649901e-05, "loss": 0.0441, "step": 62960 }, { "grad_norm": 0.3472293019294739, "learning_rate": 3.303354869177297e-05, "loss": 0.0356, "step": 62970 }, { "grad_norm": 0.4645636975765228, "learning_rate": 3.301799596247255e-05, "loss": 0.0404, "step": 62980 }, { "grad_norm": 0.3299257755279541, "learning_rate": 3.300244509029854e-05, "loss": 0.0319, "step": 62990 }, { "grad_norm": 0.47459784150123596, "learning_rate": 3.2986896076951585e-05, "loss": 0.0369, "step": 63000 }, { "grad_norm": 0.4487970471382141, "learning_rate": 3.297134892413211e-05, "loss": 0.0506, "step": 63010 }, { "grad_norm": 0.5865408182144165, "learning_rate": 3.295580363354031e-05, "loss": 0.052, "step": 63020 }, { "grad_norm": 0.5806403756141663, "learning_rate": 3.2940260206876194e-05, "loss": 0.0375, "step": 63030 }, { "grad_norm": 0.4900549054145813, "learning_rate": 3.292471864583958e-05, "loss": 0.0453, "step": 63040 }, { "grad_norm": 0.4244046211242676, "learning_rate": 3.290917895213004e-05, "loss": 0.0366, "step": 63050 }, { "grad_norm": 0.44658350944519043, "learning_rate": 3.289364112744701e-05, "loss": 0.0432, "step": 63060 }, { "grad_norm": 0.3350087106227875, "learning_rate": 3.287810517348967e-05, "loss": 0.0409, "step": 63070 }, { "grad_norm": 0.36881014704704285, "learning_rate": 3.2862571091957014e-05, "loss": 0.0396, "step": 63080 }, { "grad_norm": 0.3645998239517212, "learning_rate": 3.284703888454781e-05, "loss": 0.0342, "step": 63090 }, { "grad_norm": 0.5186560153961182, "learning_rate": 3.2831508552960646e-05, "loss": 0.0274, "step": 63100 }, { "grad_norm": 0.6521294116973877, "learning_rate": 3.281598009889387e-05, "loss": 0.0432, "step": 63110 }, { "grad_norm": 0.6459099054336548, "learning_rate": 3.28004535240457e-05, "loss": 0.0495, "step": 63120 }, { "grad_norm": 0.5584574937820435, "learning_rate": 3.278492883011407e-05, "loss": 0.0404, "step": 63130 }, { "grad_norm": 0.32259267568588257, "learning_rate": 3.276940601879675e-05, "loss": 0.0432, "step": 63140 }, { "grad_norm": 0.6660867929458618, "learning_rate": 3.275388509179127e-05, "loss": 0.0482, "step": 63150 }, { "grad_norm": 0.5554379224777222, "learning_rate": 3.2738366050794985e-05, "loss": 0.0501, "step": 63160 }, { "grad_norm": 0.38314035534858704, "learning_rate": 3.2722848897505034e-05, "loss": 0.0366, "step": 63170 }, { "grad_norm": 0.5198185443878174, "learning_rate": 3.2707333633618356e-05, "loss": 0.0365, "step": 63180 }, { "grad_norm": 0.3571878671646118, "learning_rate": 3.269182026083167e-05, "loss": 0.0371, "step": 63190 }, { "grad_norm": 0.36884376406669617, "learning_rate": 3.2676308780841516e-05, "loss": 0.0413, "step": 63200 }, { "grad_norm": 0.3615906238555908, "learning_rate": 3.266079919534416e-05, "loss": 0.0397, "step": 63210 }, { "grad_norm": 0.35056957602500916, "learning_rate": 3.264529150603575e-05, "loss": 0.0346, "step": 63220 }, { "grad_norm": 0.5335418581962585, "learning_rate": 3.262978571461217e-05, "loss": 0.0403, "step": 63230 }, { "grad_norm": 0.589489758014679, "learning_rate": 3.26142818227691e-05, "loss": 0.0493, "step": 63240 }, { "grad_norm": 0.9395914673805237, "learning_rate": 3.259877983220203e-05, "loss": 0.0447, "step": 63250 }, { "grad_norm": 0.5748582482337952, "learning_rate": 3.2583279744606246e-05, "loss": 0.0428, "step": 63260 }, { "grad_norm": 0.48370885848999023, "learning_rate": 3.25677815616768e-05, "loss": 0.027, "step": 63270 }, { "grad_norm": 0.5782707929611206, "learning_rate": 3.255228528510856e-05, "loss": 0.0495, "step": 63280 }, { "grad_norm": 0.6184280514717102, "learning_rate": 3.253679091659617e-05, "loss": 0.038, "step": 63290 }, { "grad_norm": 0.6744414567947388, "learning_rate": 3.252129845783404e-05, "loss": 0.0402, "step": 63300 }, { "grad_norm": 0.2969801723957062, "learning_rate": 3.2505807910516456e-05, "loss": 0.0406, "step": 63310 }, { "grad_norm": 1.0443321466445923, "learning_rate": 3.249031927633742e-05, "loss": 0.0497, "step": 63320 }, { "grad_norm": 0.36299943923950195, "learning_rate": 3.2474832556990737e-05, "loss": 0.033, "step": 63330 }, { "grad_norm": 0.47129589319229126, "learning_rate": 3.245934775417002e-05, "loss": 0.0363, "step": 63340 }, { "grad_norm": 0.8969942331314087, "learning_rate": 3.244386486956866e-05, "loss": 0.0398, "step": 63350 }, { "grad_norm": 0.5738767981529236, "learning_rate": 3.2428383904879846e-05, "loss": 0.0461, "step": 63360 }, { "grad_norm": 0.4501820504665375, "learning_rate": 3.241290486179652e-05, "loss": 0.0562, "step": 63370 }, { "grad_norm": 0.44405725598335266, "learning_rate": 3.239742774201151e-05, "loss": 0.0341, "step": 63380 }, { "grad_norm": 0.34422144293785095, "learning_rate": 3.238195254721733e-05, "loss": 0.0333, "step": 63390 }, { "grad_norm": 0.515276312828064, "learning_rate": 3.236647927910633e-05, "loss": 0.0419, "step": 63400 }, { "grad_norm": 0.17712517082691193, "learning_rate": 3.2351007939370646e-05, "loss": 0.0327, "step": 63410 }, { "grad_norm": 0.6776489019393921, "learning_rate": 3.2335538529702184e-05, "loss": 0.0415, "step": 63420 }, { "grad_norm": 0.7760145664215088, "learning_rate": 3.232007105179268e-05, "loss": 0.0412, "step": 63430 }, { "grad_norm": 0.6374441981315613, "learning_rate": 3.230460550733363e-05, "loss": 0.0286, "step": 63440 }, { "grad_norm": 0.26038289070129395, "learning_rate": 3.2289141898016315e-05, "loss": 0.0341, "step": 63450 }, { "grad_norm": 0.4214502274990082, "learning_rate": 3.227368022553181e-05, "loss": 0.0357, "step": 63460 }, { "grad_norm": 0.45170485973358154, "learning_rate": 3.225822049157099e-05, "loss": 0.0358, "step": 63470 }, { "grad_norm": 0.46746599674224854, "learning_rate": 3.2242762697824494e-05, "loss": 0.0317, "step": 63480 }, { "grad_norm": 1.653268575668335, "learning_rate": 3.222730684598279e-05, "loss": 0.0455, "step": 63490 }, { "grad_norm": 0.4239266514778137, "learning_rate": 3.221185293773605e-05, "loss": 0.0331, "step": 63500 }, { "grad_norm": 0.3645574748516083, "learning_rate": 3.2196400974774354e-05, "loss": 0.0311, "step": 63510 }, { "grad_norm": 0.8672304153442383, "learning_rate": 3.2180950958787485e-05, "loss": 0.0409, "step": 63520 }, { "grad_norm": 0.2905944287776947, "learning_rate": 3.2165502891465024e-05, "loss": 0.0322, "step": 63530 }, { "grad_norm": 0.7136133313179016, "learning_rate": 3.215005677449636e-05, "loss": 0.0371, "step": 63540 }, { "grad_norm": 0.805601179599762, "learning_rate": 3.2134612609570644e-05, "loss": 0.0317, "step": 63550 }, { "grad_norm": 0.32484757900238037, "learning_rate": 3.211917039837682e-05, "loss": 0.035, "step": 63560 }, { "grad_norm": 0.4797491729259491, "learning_rate": 3.210373014260365e-05, "loss": 0.0526, "step": 63570 }, { "grad_norm": 0.4028753638267517, "learning_rate": 3.208829184393965e-05, "loss": 0.0369, "step": 63580 }, { "grad_norm": 0.6653838157653809, "learning_rate": 3.207285550407312e-05, "loss": 0.0509, "step": 63590 }, { "grad_norm": 0.42524024844169617, "learning_rate": 3.205742112469217e-05, "loss": 0.0382, "step": 63600 }, { "grad_norm": 0.5140153765678406, "learning_rate": 3.204198870748465e-05, "loss": 0.0448, "step": 63610 }, { "grad_norm": 0.5384199619293213, "learning_rate": 3.202655825413824e-05, "loss": 0.0382, "step": 63620 }, { "grad_norm": 0.4996872842311859, "learning_rate": 3.2011129766340416e-05, "loss": 0.056, "step": 63630 }, { "grad_norm": 0.38020968437194824, "learning_rate": 3.1995703245778384e-05, "loss": 0.0288, "step": 63640 }, { "grad_norm": 0.37765657901763916, "learning_rate": 3.1980278694139186e-05, "loss": 0.033, "step": 63650 }, { "grad_norm": 0.537706196308136, "learning_rate": 3.196485611310962e-05, "loss": 0.0387, "step": 63660 }, { "grad_norm": 0.7202037572860718, "learning_rate": 3.1949435504376257e-05, "loss": 0.0435, "step": 63670 }, { "grad_norm": 0.6619460582733154, "learning_rate": 3.1934016869625505e-05, "loss": 0.0315, "step": 63680 }, { "grad_norm": 0.5443788170814514, "learning_rate": 3.191860021054349e-05, "loss": 0.0514, "step": 63690 }, { "grad_norm": 0.5390627384185791, "learning_rate": 3.190318552881618e-05, "loss": 0.0486, "step": 63700 }, { "grad_norm": 0.836303174495697, "learning_rate": 3.1887772826129285e-05, "loss": 0.0391, "step": 63710 }, { "grad_norm": 0.5413007140159607, "learning_rate": 3.1872362104168335e-05, "loss": 0.0428, "step": 63720 }, { "grad_norm": 0.5439905524253845, "learning_rate": 3.185695336461861e-05, "loss": 0.0416, "step": 63730 }, { "grad_norm": 0.6772120594978333, "learning_rate": 3.184154660916519e-05, "loss": 0.0349, "step": 63740 }, { "grad_norm": 0.4778219163417816, "learning_rate": 3.1826141839492906e-05, "loss": 0.0357, "step": 63750 }, { "grad_norm": 0.7189950942993164, "learning_rate": 3.181073905728644e-05, "loss": 0.0338, "step": 63760 }, { "grad_norm": 0.4634392261505127, "learning_rate": 3.1795338264230214e-05, "loss": 0.0254, "step": 63770 }, { "grad_norm": 0.5546732544898987, "learning_rate": 3.1779939462008426e-05, "loss": 0.0414, "step": 63780 }, { "grad_norm": 0.4628002345561981, "learning_rate": 3.1764542652305066e-05, "loss": 0.0328, "step": 63790 }, { "grad_norm": 0.435183048248291, "learning_rate": 3.17491478368039e-05, "loss": 0.0456, "step": 63800 }, { "grad_norm": 0.724111795425415, "learning_rate": 3.173375501718847e-05, "loss": 0.0372, "step": 63810 }, { "grad_norm": 0.5954081416130066, "learning_rate": 3.171836419514215e-05, "loss": 0.0315, "step": 63820 }, { "grad_norm": 0.6336634159088135, "learning_rate": 3.170297537234803e-05, "loss": 0.0518, "step": 63830 }, { "grad_norm": 0.5383710265159607, "learning_rate": 3.168758855048901e-05, "loss": 0.0331, "step": 63840 }, { "grad_norm": 0.393852561712265, "learning_rate": 3.1672203731247785e-05, "loss": 0.0402, "step": 63850 }, { "grad_norm": 0.45649147033691406, "learning_rate": 3.16568209163068e-05, "loss": 0.0478, "step": 63860 }, { "grad_norm": 0.34062471985816956, "learning_rate": 3.1641440107348295e-05, "loss": 0.043, "step": 63870 }, { "grad_norm": 0.4706365764141083, "learning_rate": 3.162606130605429e-05, "loss": 0.0333, "step": 63880 }, { "grad_norm": 0.3584689199924469, "learning_rate": 3.16106845141066e-05, "loss": 0.0343, "step": 63890 }, { "grad_norm": 0.454367071390152, "learning_rate": 3.159530973318682e-05, "loss": 0.044, "step": 63900 }, { "grad_norm": 0.42255282402038574, "learning_rate": 3.1579936964976286e-05, "loss": 0.039, "step": 63910 }, { "grad_norm": 0.39874032139778137, "learning_rate": 3.156456621115615e-05, "loss": 0.0451, "step": 63920 }, { "grad_norm": 0.5963199138641357, "learning_rate": 3.154919747340734e-05, "loss": 0.0337, "step": 63930 }, { "grad_norm": 0.6209023594856262, "learning_rate": 3.153383075341054e-05, "loss": 0.0387, "step": 63940 }, { "grad_norm": 0.5413435101509094, "learning_rate": 3.1518466052846264e-05, "loss": 0.0328, "step": 63950 }, { "grad_norm": 0.5931860208511353, "learning_rate": 3.150310337339474e-05, "loss": 0.0394, "step": 63960 }, { "grad_norm": 0.563567578792572, "learning_rate": 3.148774271673603e-05, "loss": 0.0336, "step": 63970 }, { "grad_norm": 0.43428441882133484, "learning_rate": 3.147238408454995e-05, "loss": 0.0456, "step": 63980 }, { "grad_norm": 0.3563743233680725, "learning_rate": 3.1457027478516086e-05, "loss": 0.0378, "step": 63990 }, { "grad_norm": 0.4653841555118561, "learning_rate": 3.144167290031382e-05, "loss": 0.0417, "step": 64000 }, { "grad_norm": 0.9091822504997253, "learning_rate": 3.142632035162227e-05, "loss": 0.0422, "step": 64010 }, { "grad_norm": 0.6289766430854797, "learning_rate": 3.141096983412044e-05, "loss": 0.0308, "step": 64020 }, { "grad_norm": 0.3622378408908844, "learning_rate": 3.139562134948699e-05, "loss": 0.0375, "step": 64030 }, { "grad_norm": 1.38420832157135, "learning_rate": 3.1380274899400424e-05, "loss": 0.0402, "step": 64040 }, { "grad_norm": 0.38535064458847046, "learning_rate": 3.1364930485539e-05, "loss": 0.0373, "step": 64050 }, { "grad_norm": 0.7315963506698608, "learning_rate": 3.134958810958076e-05, "loss": 0.0336, "step": 64060 }, { "grad_norm": 0.4642851948738098, "learning_rate": 3.13342477732035e-05, "loss": 0.0392, "step": 64070 }, { "grad_norm": 1.3834782838821411, "learning_rate": 3.131890947808487e-05, "loss": 0.0306, "step": 64080 }, { "grad_norm": 0.3662092089653015, "learning_rate": 3.13035732259022e-05, "loss": 0.0454, "step": 64090 }, { "grad_norm": 0.40941792726516724, "learning_rate": 3.1288239018332656e-05, "loss": 0.0321, "step": 64100 }, { "grad_norm": 0.39847448468208313, "learning_rate": 3.1272906857053164e-05, "loss": 0.0357, "step": 64110 }, { "grad_norm": 0.5933202505111694, "learning_rate": 3.125757674374042e-05, "loss": 0.0361, "step": 64120 }, { "grad_norm": 0.32675448060035706, "learning_rate": 3.124224868007088e-05, "loss": 0.0402, "step": 64130 }, { "grad_norm": 0.40800777077674866, "learning_rate": 3.1226922667720854e-05, "loss": 0.0326, "step": 64140 }, { "grad_norm": 0.6735436916351318, "learning_rate": 3.121159870836633e-05, "loss": 0.045, "step": 64150 }, { "grad_norm": 0.4979449212551117, "learning_rate": 3.119627680368312e-05, "loss": 0.0362, "step": 64160 }, { "grad_norm": 0.3612469732761383, "learning_rate": 3.1180956955346816e-05, "loss": 0.0375, "step": 64170 }, { "grad_norm": 0.6394248008728027, "learning_rate": 3.116563916503276e-05, "loss": 0.0392, "step": 64180 }, { "grad_norm": 0.30111929774284363, "learning_rate": 3.1150323434416095e-05, "loss": 0.0413, "step": 64190 }, { "grad_norm": 0.32185685634613037, "learning_rate": 3.1135009765171696e-05, "loss": 0.0381, "step": 64200 }, { "grad_norm": 0.37800511717796326, "learning_rate": 3.111969815897429e-05, "loss": 0.039, "step": 64210 }, { "grad_norm": 0.3540898561477661, "learning_rate": 3.1104388617498295e-05, "loss": 0.0363, "step": 64220 }, { "grad_norm": 0.365016907453537, "learning_rate": 3.1089081142417966e-05, "loss": 0.0444, "step": 64230 }, { "grad_norm": 0.3689930737018585, "learning_rate": 3.1073775735407285e-05, "loss": 0.0404, "step": 64240 }, { "grad_norm": 0.5387905836105347, "learning_rate": 3.1058472398140035e-05, "loss": 0.0414, "step": 64250 }, { "grad_norm": 0.42955517768859863, "learning_rate": 3.1043171132289736e-05, "loss": 0.0343, "step": 64260 }, { "grad_norm": 0.3435881733894348, "learning_rate": 3.1027871939529774e-05, "loss": 0.0296, "step": 64270 }, { "grad_norm": 0.6880974173545837, "learning_rate": 3.1012574821533214e-05, "loss": 0.0476, "step": 64280 }, { "grad_norm": 1.4094853401184082, "learning_rate": 3.099727977997291e-05, "loss": 0.0409, "step": 64290 }, { "grad_norm": 0.5061857104301453, "learning_rate": 3.098198681652154e-05, "loss": 0.0485, "step": 64300 }, { "grad_norm": 0.4082726538181305, "learning_rate": 3.0966695932851475e-05, "loss": 0.0335, "step": 64310 }, { "grad_norm": 0.35448989272117615, "learning_rate": 3.0951407130634926e-05, "loss": 0.0387, "step": 64320 }, { "grad_norm": 0.4388807415962219, "learning_rate": 3.0936120411543864e-05, "loss": 0.0355, "step": 64330 }, { "grad_norm": 0.6351879239082336, "learning_rate": 3.0920835777250015e-05, "loss": 0.0407, "step": 64340 }, { "grad_norm": 0.6030561923980713, "learning_rate": 3.090555322942488e-05, "loss": 0.0484, "step": 64350 }, { "grad_norm": 0.5409356951713562, "learning_rate": 3.089027276973973e-05, "loss": 0.0391, "step": 64360 }, { "grad_norm": 0.3700060546398163, "learning_rate": 3.0874994399865615e-05, "loss": 0.0316, "step": 64370 }, { "grad_norm": 0.48203131556510925, "learning_rate": 3.085971812147337e-05, "loss": 0.0349, "step": 64380 }, { "grad_norm": 0.3871021270751953, "learning_rate": 3.0844443936233557e-05, "loss": 0.0412, "step": 64390 }, { "grad_norm": 0.3792259991168976, "learning_rate": 3.082917184581657e-05, "loss": 0.0415, "step": 64400 }, { "grad_norm": 0.45611780881881714, "learning_rate": 3.0813901851892513e-05, "loss": 0.0374, "step": 64410 }, { "grad_norm": 0.7729665040969849, "learning_rate": 3.0798633956131316e-05, "loss": 0.0429, "step": 64420 }, { "grad_norm": 0.22209012508392334, "learning_rate": 3.078336816020264e-05, "loss": 0.0316, "step": 64430 }, { "grad_norm": 0.5632045865058899, "learning_rate": 3.0768104465775935e-05, "loss": 0.0404, "step": 64440 }, { "grad_norm": 0.5867853760719299, "learning_rate": 3.0752842874520396e-05, "loss": 0.0346, "step": 64450 }, { "grad_norm": 0.5915814638137817, "learning_rate": 3.073758338810503e-05, "loss": 0.0374, "step": 64460 }, { "grad_norm": 0.723997175693512, "learning_rate": 3.072232600819859e-05, "loss": 0.0378, "step": 64470 }, { "grad_norm": 0.4131598472595215, "learning_rate": 3.070707073646959e-05, "loss": 0.041, "step": 64480 }, { "grad_norm": 0.3627701699733734, "learning_rate": 3.069181757458633e-05, "loss": 0.0416, "step": 64490 }, { "grad_norm": 0.360867977142334, "learning_rate": 3.067656652421687e-05, "loss": 0.0329, "step": 64500 }, { "grad_norm": 0.2532239556312561, "learning_rate": 3.0661317587029025e-05, "loss": 0.0384, "step": 64510 }, { "grad_norm": 0.6522480249404907, "learning_rate": 3.064607076469043e-05, "loss": 0.0319, "step": 64520 }, { "grad_norm": 0.452938050031662, "learning_rate": 3.063082605886845e-05, "loss": 0.0414, "step": 64530 }, { "grad_norm": 0.4535786807537079, "learning_rate": 3.061558347123019e-05, "loss": 0.0437, "step": 64540 }, { "grad_norm": 0.2929413616657257, "learning_rate": 3.060034300344259e-05, "loss": 0.0392, "step": 64550 }, { "grad_norm": 0.43754810094833374, "learning_rate": 3.058510465717231e-05, "loss": 0.0448, "step": 64560 }, { "grad_norm": 0.34348803758621216, "learning_rate": 3.0569868434085804e-05, "loss": 0.0293, "step": 64570 }, { "grad_norm": 0.31259968876838684, "learning_rate": 3.055463433584924e-05, "loss": 0.0269, "step": 64580 }, { "grad_norm": 0.5310232639312744, "learning_rate": 3.0539402364128656e-05, "loss": 0.0343, "step": 64590 }, { "grad_norm": 0.7061344981193542, "learning_rate": 3.052417252058977e-05, "loss": 0.0438, "step": 64600 }, { "grad_norm": 0.47467154264450073, "learning_rate": 3.0508944806898092e-05, "loss": 0.0452, "step": 64610 }, { "grad_norm": 0.39675185084342957, "learning_rate": 3.04937192247189e-05, "loss": 0.0541, "step": 64620 }, { "grad_norm": 1.3183610439300537, "learning_rate": 3.047849577571725e-05, "loss": 0.0456, "step": 64630 }, { "grad_norm": 0.8107433319091797, "learning_rate": 3.046327446155795e-05, "loss": 0.0389, "step": 64640 }, { "grad_norm": 0.37201616168022156, "learning_rate": 3.0448055283905587e-05, "loss": 0.0308, "step": 64650 }, { "grad_norm": 0.39955243468284607, "learning_rate": 3.0432838244424494e-05, "loss": 0.035, "step": 64660 }, { "grad_norm": 0.4433991611003876, "learning_rate": 3.04176233447788e-05, "loss": 0.028, "step": 64670 }, { "grad_norm": 0.28048843145370483, "learning_rate": 3.040241058663238e-05, "loss": 0.031, "step": 64680 }, { "grad_norm": 0.3649843633174896, "learning_rate": 3.038719997164887e-05, "loss": 0.0467, "step": 64690 }, { "grad_norm": 0.30739280581474304, "learning_rate": 3.0371991501491683e-05, "loss": 0.0378, "step": 64700 }, { "grad_norm": 0.4890843331813812, "learning_rate": 3.0356785177823975e-05, "loss": 0.0354, "step": 64710 }, { "grad_norm": 0.3368557095527649, "learning_rate": 3.0341581002308726e-05, "loss": 0.0332, "step": 64720 }, { "grad_norm": 0.4710488021373749, "learning_rate": 3.0326378976608627e-05, "loss": 0.0351, "step": 64730 }, { "grad_norm": 0.4012064039707184, "learning_rate": 3.0311179102386146e-05, "loss": 0.0536, "step": 64740 }, { "grad_norm": 0.6129899621009827, "learning_rate": 3.0295981381303513e-05, "loss": 0.0392, "step": 64750 }, { "grad_norm": 0.43059808015823364, "learning_rate": 3.0280785815022727e-05, "loss": 0.036, "step": 64760 }, { "grad_norm": 0.4033389687538147, "learning_rate": 3.0265592405205544e-05, "loss": 0.036, "step": 64770 }, { "grad_norm": 0.5111687183380127, "learning_rate": 3.0250401153513517e-05, "loss": 0.0328, "step": 64780 }, { "grad_norm": 0.6818541288375854, "learning_rate": 3.0235212061607936e-05, "loss": 0.0519, "step": 64790 }, { "grad_norm": 0.3444018065929413, "learning_rate": 3.0220025131149842e-05, "loss": 0.033, "step": 64800 }, { "grad_norm": 0.3090386688709259, "learning_rate": 3.020484036380007e-05, "loss": 0.0329, "step": 64810 }, { "grad_norm": 0.4114309549331665, "learning_rate": 3.0189657761219188e-05, "loss": 0.0421, "step": 64820 }, { "grad_norm": 0.2987445294857025, "learning_rate": 3.017447732506753e-05, "loss": 0.0427, "step": 64830 }, { "grad_norm": 0.45161497592926025, "learning_rate": 3.015929905700524e-05, "loss": 0.0342, "step": 64840 }, { "grad_norm": 0.3316754996776581, "learning_rate": 3.014412295869218e-05, "loss": 0.034, "step": 64850 }, { "grad_norm": 0.6775216460227966, "learning_rate": 3.0128949031787977e-05, "loss": 0.0392, "step": 64860 }, { "grad_norm": 0.37840911746025085, "learning_rate": 3.0113777277952022e-05, "loss": 0.0438, "step": 64870 }, { "grad_norm": 0.5841498970985413, "learning_rate": 3.0098607698843496e-05, "loss": 0.038, "step": 64880 }, { "grad_norm": 0.49397656321525574, "learning_rate": 3.0083440296121312e-05, "loss": 0.0365, "step": 64890 }, { "grad_norm": 0.37404492497444153, "learning_rate": 3.0068275071444142e-05, "loss": 0.0296, "step": 64900 }, { "grad_norm": 0.4924781918525696, "learning_rate": 3.0053112026470442e-05, "loss": 0.0441, "step": 64910 }, { "grad_norm": 0.39377376437187195, "learning_rate": 3.0037951162858434e-05, "loss": 0.032, "step": 64920 }, { "grad_norm": 0.29718339443206787, "learning_rate": 3.002279248226607e-05, "loss": 0.0316, "step": 64930 }, { "grad_norm": 0.43223026394844055, "learning_rate": 3.000763598635109e-05, "loss": 0.0419, "step": 64940 }, { "grad_norm": 0.8882981538772583, "learning_rate": 2.999248167677098e-05, "loss": 0.0484, "step": 64950 }, { "grad_norm": 0.4818163812160492, "learning_rate": 2.9977329555182976e-05, "loss": 0.0452, "step": 64960 }, { "grad_norm": 0.4632363021373749, "learning_rate": 2.9962179623244124e-05, "loss": 0.041, "step": 64970 }, { "grad_norm": 0.5335662961006165, "learning_rate": 2.9947031882611175e-05, "loss": 0.0354, "step": 64980 }, { "grad_norm": 0.4842219352722168, "learning_rate": 2.9931886334940685e-05, "loss": 0.0353, "step": 64990 }, { "grad_norm": 0.43591466546058655, "learning_rate": 2.9916742981888923e-05, "loss": 0.038, "step": 65000 }, { "grad_norm": 0.5368339419364929, "learning_rate": 2.990160182511196e-05, "loss": 0.0444, "step": 65010 }, { "grad_norm": 1.1659945249557495, "learning_rate": 2.9886462866265574e-05, "loss": 0.0423, "step": 65020 }, { "grad_norm": 0.4106976389884949, "learning_rate": 2.9871326107005392e-05, "loss": 0.0325, "step": 65030 }, { "grad_norm": 0.6151836514472961, "learning_rate": 2.9856191548986735e-05, "loss": 0.0456, "step": 65040 }, { "grad_norm": 0.28170111775398254, "learning_rate": 2.9841059193864672e-05, "loss": 0.0329, "step": 65050 }, { "grad_norm": 0.3537669777870178, "learning_rate": 2.982592904329407e-05, "loss": 0.0328, "step": 65060 }, { "grad_norm": 0.19748465716838837, "learning_rate": 2.981080109892954e-05, "loss": 0.0278, "step": 65070 }, { "grad_norm": 1.2785600423812866, "learning_rate": 2.9795675362425434e-05, "loss": 0.0403, "step": 65080 }, { "grad_norm": 0.43823978304862976, "learning_rate": 2.97805518354359e-05, "loss": 0.0303, "step": 65090 }, { "grad_norm": 0.842454731464386, "learning_rate": 2.9765430519614825e-05, "loss": 0.0333, "step": 65100 }, { "grad_norm": 0.6258570551872253, "learning_rate": 2.975031141661584e-05, "loss": 0.0343, "step": 65110 }, { "grad_norm": 0.37342506647109985, "learning_rate": 2.9735194528092347e-05, "loss": 0.0369, "step": 65120 }, { "grad_norm": 0.4734342694282532, "learning_rate": 2.9720079855697525e-05, "loss": 0.0527, "step": 65130 }, { "grad_norm": 1.117408037185669, "learning_rate": 2.970496740108427e-05, "loss": 0.0529, "step": 65140 }, { "grad_norm": 0.4069578945636749, "learning_rate": 2.968985716590527e-05, "loss": 0.0313, "step": 65150 }, { "grad_norm": 0.3348253071308136, "learning_rate": 2.9674749151812953e-05, "loss": 0.0354, "step": 65160 }, { "grad_norm": 0.5147761106491089, "learning_rate": 2.965964336045951e-05, "loss": 0.0365, "step": 65170 }, { "grad_norm": 0.38023069500923157, "learning_rate": 2.9644539793496894e-05, "loss": 0.0342, "step": 65180 }, { "grad_norm": 0.6541862487792969, "learning_rate": 2.9629438452576807e-05, "loss": 0.0374, "step": 65190 }, { "grad_norm": 0.5482854843139648, "learning_rate": 2.9614339339350698e-05, "loss": 0.0311, "step": 65200 }, { "grad_norm": 1.1861616373062134, "learning_rate": 2.959924245546979e-05, "loss": 0.0354, "step": 65210 }, { "grad_norm": 0.5257419347763062, "learning_rate": 2.9584147802585038e-05, "loss": 0.0524, "step": 65220 }, { "grad_norm": 0.44240275025367737, "learning_rate": 2.956905538234721e-05, "loss": 0.0398, "step": 65230 }, { "grad_norm": 0.4982464611530304, "learning_rate": 2.9553965196406764e-05, "loss": 0.0408, "step": 65240 }, { "grad_norm": 0.47957316040992737, "learning_rate": 2.9538877246413943e-05, "loss": 0.0318, "step": 65250 }, { "grad_norm": 0.4186058044433594, "learning_rate": 2.952379153401874e-05, "loss": 0.0395, "step": 65260 }, { "grad_norm": 0.4279170334339142, "learning_rate": 2.9508708060870914e-05, "loss": 0.0328, "step": 65270 }, { "grad_norm": 0.3504939675331116, "learning_rate": 2.9493626828619948e-05, "loss": 0.031, "step": 65280 }, { "grad_norm": 0.3474327325820923, "learning_rate": 2.9478547838915137e-05, "loss": 0.0294, "step": 65290 }, { "grad_norm": 0.983529806137085, "learning_rate": 2.946347109340548e-05, "loss": 0.0419, "step": 65300 }, { "grad_norm": 0.5109795331954956, "learning_rate": 2.944839659373976e-05, "loss": 0.0344, "step": 65310 }, { "grad_norm": 0.5121685862541199, "learning_rate": 2.943332434156647e-05, "loss": 0.0388, "step": 65320 }, { "grad_norm": 0.5749241709709167, "learning_rate": 2.9418254338533912e-05, "loss": 0.043, "step": 65330 }, { "grad_norm": 1.0236250162124634, "learning_rate": 2.940318658629011e-05, "loss": 0.0475, "step": 65340 }, { "grad_norm": 0.5884863138198853, "learning_rate": 2.938812108648287e-05, "loss": 0.0443, "step": 65350 }, { "grad_norm": 0.33869433403015137, "learning_rate": 2.937305784075972e-05, "loss": 0.0341, "step": 65360 }, { "grad_norm": 0.6550202965736389, "learning_rate": 2.9357996850767937e-05, "loss": 0.0315, "step": 65370 }, { "grad_norm": 0.5287904143333435, "learning_rate": 2.9342938118154595e-05, "loss": 0.0319, "step": 65380 }, { "grad_norm": 0.9264327883720398, "learning_rate": 2.9327881644566485e-05, "loss": 0.0507, "step": 65390 }, { "grad_norm": 0.41963890194892883, "learning_rate": 2.9312827431650157e-05, "loss": 0.0335, "step": 65400 }, { "grad_norm": 0.22007444500923157, "learning_rate": 2.92977754810519e-05, "loss": 0.0538, "step": 65410 }, { "grad_norm": 0.8442291021347046, "learning_rate": 2.928272579441781e-05, "loss": 0.0407, "step": 65420 }, { "grad_norm": 0.39019501209259033, "learning_rate": 2.9267678373393676e-05, "loss": 0.0498, "step": 65430 }, { "grad_norm": 0.3723510205745697, "learning_rate": 2.925263321962507e-05, "loss": 0.0431, "step": 65440 }, { "grad_norm": 0.8280491828918457, "learning_rate": 2.9237590334757292e-05, "loss": 0.0399, "step": 65450 }, { "grad_norm": 0.5086879134178162, "learning_rate": 2.922254972043543e-05, "loss": 0.0414, "step": 65460 }, { "grad_norm": 1.2219111919403076, "learning_rate": 2.9207511378304265e-05, "loss": 0.0412, "step": 65470 }, { "grad_norm": 0.6596665978431702, "learning_rate": 2.9192475310008428e-05, "loss": 0.0472, "step": 65480 }, { "grad_norm": 0.4037916362285614, "learning_rate": 2.917744151719219e-05, "loss": 0.0381, "step": 65490 }, { "grad_norm": 2.2862727642059326, "learning_rate": 2.916241000149964e-05, "loss": 0.0409, "step": 65500 }, { "grad_norm": 0.3821549415588379, "learning_rate": 2.9147380764574622e-05, "loss": 0.0314, "step": 65510 }, { "grad_norm": 0.4965914487838745, "learning_rate": 2.9132353808060674e-05, "loss": 0.0508, "step": 65520 }, { "grad_norm": 0.41280120611190796, "learning_rate": 2.9117329133601134e-05, "loss": 0.0368, "step": 65530 }, { "grad_norm": 0.3395861089229584, "learning_rate": 2.9102306742839093e-05, "loss": 0.0346, "step": 65540 }, { "grad_norm": 0.2999188303947449, "learning_rate": 2.9087286637417375e-05, "loss": 0.0345, "step": 65550 }, { "grad_norm": 0.6648188829421997, "learning_rate": 2.9072268818978537e-05, "loss": 0.0492, "step": 65560 }, { "grad_norm": 0.3463060259819031, "learning_rate": 2.905725328916493e-05, "loss": 0.035, "step": 65570 }, { "grad_norm": 0.6516644358634949, "learning_rate": 2.9042240049618608e-05, "loss": 0.0418, "step": 65580 }, { "grad_norm": 0.5256273150444031, "learning_rate": 2.9027229101981413e-05, "loss": 0.0357, "step": 65590 }, { "grad_norm": 0.32042422890663147, "learning_rate": 2.901222044789488e-05, "loss": 0.0421, "step": 65600 }, { "grad_norm": 0.3818958103656769, "learning_rate": 2.89972140890004e-05, "loss": 0.0335, "step": 65610 }, { "grad_norm": 0.5906689167022705, "learning_rate": 2.8982210026938988e-05, "loss": 0.0422, "step": 65620 }, { "grad_norm": 1.0488146543502808, "learning_rate": 2.896720826335151e-05, "loss": 0.0542, "step": 65630 }, { "grad_norm": 0.44027116894721985, "learning_rate": 2.8952208799878493e-05, "loss": 0.0411, "step": 65640 }, { "grad_norm": 0.4858795404434204, "learning_rate": 2.89372116381603e-05, "loss": 0.0354, "step": 65650 }, { "grad_norm": 0.6702636480331421, "learning_rate": 2.8922216779836932e-05, "loss": 0.041, "step": 65660 }, { "grad_norm": 0.6319762468338013, "learning_rate": 2.8907224226548285e-05, "loss": 0.0301, "step": 65670 }, { "grad_norm": 0.46274420619010925, "learning_rate": 2.8892233979933868e-05, "loss": 0.0348, "step": 65680 }, { "grad_norm": 0.959877610206604, "learning_rate": 2.8877246041633026e-05, "loss": 0.0394, "step": 65690 }, { "grad_norm": 0.728302001953125, "learning_rate": 2.886226041328478e-05, "loss": 0.0396, "step": 65700 }, { "grad_norm": 0.3781941533088684, "learning_rate": 2.8847277096527957e-05, "loss": 0.0317, "step": 65710 }, { "grad_norm": 0.5040633082389832, "learning_rate": 2.8832296093001105e-05, "loss": 0.0289, "step": 65720 }, { "grad_norm": 0.3175525665283203, "learning_rate": 2.8817317404342547e-05, "loss": 0.0334, "step": 65730 }, { "grad_norm": 0.5209565162658691, "learning_rate": 2.880234103219029e-05, "loss": 0.0438, "step": 65740 }, { "grad_norm": 0.7490935921669006, "learning_rate": 2.878736697818215e-05, "loss": 0.033, "step": 65750 }, { "grad_norm": 0.5315729379653931, "learning_rate": 2.877239524395568e-05, "loss": 0.0351, "step": 65760 }, { "grad_norm": 0.3387298882007599, "learning_rate": 2.8757425831148134e-05, "loss": 0.0386, "step": 65770 }, { "grad_norm": 0.4420487582683563, "learning_rate": 2.874245874139658e-05, "loss": 0.0401, "step": 65780 }, { "grad_norm": 0.39842087030410767, "learning_rate": 2.8727493976337737e-05, "loss": 0.025, "step": 65790 }, { "grad_norm": 0.5430203676223755, "learning_rate": 2.8712531537608218e-05, "loss": 0.0319, "step": 65800 }, { "grad_norm": 0.4659813642501831, "learning_rate": 2.8697571426844217e-05, "loss": 0.0415, "step": 65810 }, { "grad_norm": 0.5493913292884827, "learning_rate": 2.86826136456818e-05, "loss": 0.0366, "step": 65820 }, { "grad_norm": 0.33026012778282166, "learning_rate": 2.8667658195756685e-05, "loss": 0.0432, "step": 65830 }, { "grad_norm": 1.4769295454025269, "learning_rate": 2.8652705078704412e-05, "loss": 0.0452, "step": 65840 }, { "grad_norm": 0.40642136335372925, "learning_rate": 2.863775429616018e-05, "loss": 0.0478, "step": 65850 }, { "grad_norm": 0.41055983304977417, "learning_rate": 2.862280584975906e-05, "loss": 0.0334, "step": 65860 }, { "grad_norm": 0.3636328876018524, "learning_rate": 2.8607859741135735e-05, "loss": 0.0411, "step": 65870 }, { "grad_norm": 0.35887572169303894, "learning_rate": 2.8592915971924723e-05, "loss": 0.0479, "step": 65880 }, { "grad_norm": 0.4948078691959381, "learning_rate": 2.8577974543760212e-05, "loss": 0.0422, "step": 65890 }, { "grad_norm": 0.5666177868843079, "learning_rate": 2.8563035458276222e-05, "loss": 0.0375, "step": 65900 }, { "grad_norm": 0.3399089574813843, "learning_rate": 2.8548098717106424e-05, "loss": 0.0388, "step": 65910 }, { "grad_norm": 0.4330960810184479, "learning_rate": 2.85331643218843e-05, "loss": 0.039, "step": 65920 }, { "grad_norm": 2.052797555923462, "learning_rate": 2.8518232274243046e-05, "loss": 0.0453, "step": 65930 }, { "grad_norm": 1.2409663200378418, "learning_rate": 2.8503302575815638e-05, "loss": 0.0575, "step": 65940 }, { "grad_norm": 0.3315475285053253, "learning_rate": 2.8488375228234716e-05, "loss": 0.0394, "step": 65950 }, { "grad_norm": 0.5160521268844604, "learning_rate": 2.8473450233132738e-05, "loss": 0.0398, "step": 65960 }, { "grad_norm": 0.29627904295921326, "learning_rate": 2.8458527592141898e-05, "loss": 0.0342, "step": 65970 }, { "grad_norm": 0.3939135670661926, "learning_rate": 2.8443607306894073e-05, "loss": 0.0448, "step": 65980 }, { "grad_norm": 0.3668997585773468, "learning_rate": 2.8428689379020946e-05, "loss": 0.0381, "step": 65990 }, { "grad_norm": 0.30838167667388916, "learning_rate": 2.841377381015391e-05, "loss": 0.0439, "step": 66000 }, { "grad_norm": 0.3425573408603668, "learning_rate": 2.8398860601924144e-05, "loss": 0.0324, "step": 66010 }, { "grad_norm": 0.39793217182159424, "learning_rate": 2.838394975596248e-05, "loss": 0.0325, "step": 66020 }, { "grad_norm": 0.5958703756332397, "learning_rate": 2.83690412738996e-05, "loss": 0.046, "step": 66030 }, { "grad_norm": 0.34763455390930176, "learning_rate": 2.835413515736579e-05, "loss": 0.0411, "step": 66040 }, { "grad_norm": 0.46611288189888, "learning_rate": 2.8339231407991267e-05, "loss": 0.0355, "step": 66050 }, { "grad_norm": 0.6980740427970886, "learning_rate": 2.832433002740581e-05, "loss": 0.0329, "step": 66060 }, { "grad_norm": 0.7301799058914185, "learning_rate": 2.8309431017239047e-05, "loss": 0.0296, "step": 66070 }, { "grad_norm": 0.3763127326965332, "learning_rate": 2.829453437912028e-05, "loss": 0.0377, "step": 66080 }, { "grad_norm": 0.3741230368614197, "learning_rate": 2.8279640114678617e-05, "loss": 0.0326, "step": 66090 }, { "grad_norm": 0.6485029458999634, "learning_rate": 2.8264748225542836e-05, "loss": 0.0445, "step": 66100 }, { "grad_norm": 0.38118642568588257, "learning_rate": 2.8249858713341516e-05, "loss": 0.0329, "step": 66110 }, { "grad_norm": 0.7106660008430481, "learning_rate": 2.823497157970294e-05, "loss": 0.0453, "step": 66120 }, { "grad_norm": 0.32931506633758545, "learning_rate": 2.8220086826255164e-05, "loss": 0.03, "step": 66130 }, { "grad_norm": 0.5235068202018738, "learning_rate": 2.8205204454625934e-05, "loss": 0.0435, "step": 66140 }, { "grad_norm": 0.5339564085006714, "learning_rate": 2.8190324466442798e-05, "loss": 0.0297, "step": 66150 }, { "grad_norm": 0.7386385202407837, "learning_rate": 2.8175446863332965e-05, "loss": 0.0352, "step": 66160 }, { "grad_norm": 0.5450489521026611, "learning_rate": 2.816057164692345e-05, "loss": 0.0405, "step": 66170 }, { "grad_norm": 0.23976710438728333, "learning_rate": 2.8145698818840992e-05, "loss": 0.0444, "step": 66180 }, { "grad_norm": 0.5707858800888062, "learning_rate": 2.8130828380712065e-05, "loss": 0.0387, "step": 66190 }, { "grad_norm": 0.8961787223815918, "learning_rate": 2.811596033416285e-05, "loss": 0.0374, "step": 66200 }, { "grad_norm": 0.6878200173377991, "learning_rate": 2.8101094680819307e-05, "loss": 0.0411, "step": 66210 }, { "grad_norm": 0.6608654260635376, "learning_rate": 2.808623142230715e-05, "loss": 0.0322, "step": 66220 }, { "grad_norm": 0.45817723870277405, "learning_rate": 2.8071370560251757e-05, "loss": 0.0303, "step": 66230 }, { "grad_norm": 0.9022242426872253, "learning_rate": 2.805651209627831e-05, "loss": 0.0436, "step": 66240 }, { "grad_norm": 0.5031852126121521, "learning_rate": 2.804165603201171e-05, "loss": 0.0359, "step": 66250 }, { "grad_norm": 0.46928685903549194, "learning_rate": 2.802680236907661e-05, "loss": 0.0324, "step": 66260 }, { "grad_norm": 0.41173163056373596, "learning_rate": 2.8011951109097344e-05, "loss": 0.0322, "step": 66270 }, { "grad_norm": 1.2175309658050537, "learning_rate": 2.7997102253698072e-05, "loss": 0.0373, "step": 66280 }, { "grad_norm": 0.3828069269657135, "learning_rate": 2.7982255804502588e-05, "loss": 0.0373, "step": 66290 }, { "grad_norm": 0.4891728162765503, "learning_rate": 2.7967411763134516e-05, "loss": 0.0417, "step": 66300 }, { "grad_norm": 0.46715760231018066, "learning_rate": 2.7952570131217166e-05, "loss": 0.0316, "step": 66310 }, { "grad_norm": 0.4221089482307434, "learning_rate": 2.7937730910373618e-05, "loss": 0.0379, "step": 66320 }, { "grad_norm": 0.2565990686416626, "learning_rate": 2.7922894102226628e-05, "loss": 0.0299, "step": 66330 }, { "grad_norm": 0.39437422156333923, "learning_rate": 2.790805970839876e-05, "loss": 0.0318, "step": 66340 }, { "grad_norm": 0.4480251967906952, "learning_rate": 2.789322773051225e-05, "loss": 0.0256, "step": 66350 }, { "grad_norm": 0.6006066203117371, "learning_rate": 2.7878398170189124e-05, "loss": 0.0344, "step": 66360 }, { "grad_norm": 0.472250372171402, "learning_rate": 2.7863571029051112e-05, "loss": 0.0384, "step": 66370 }, { "grad_norm": 0.3511877954006195, "learning_rate": 2.784874630871971e-05, "loss": 0.0396, "step": 66380 }, { "grad_norm": 0.794318675994873, "learning_rate": 2.7833924010816086e-05, "loss": 0.0357, "step": 66390 }, { "grad_norm": 0.41810768842697144, "learning_rate": 2.7819104136961226e-05, "loss": 0.0378, "step": 66400 }, { "grad_norm": 0.24036988615989685, "learning_rate": 2.780428668877577e-05, "loss": 0.039, "step": 66410 }, { "grad_norm": 0.54201340675354, "learning_rate": 2.7789471667880152e-05, "loss": 0.0439, "step": 66420 }, { "grad_norm": 0.45553267002105713, "learning_rate": 2.7774659075894523e-05, "loss": 0.0421, "step": 66430 }, { "grad_norm": 0.5818895101547241, "learning_rate": 2.7759848914438778e-05, "loss": 0.0525, "step": 66440 }, { "grad_norm": 0.4422329366207123, "learning_rate": 2.7745041185132496e-05, "loss": 0.047, "step": 66450 }, { "grad_norm": 0.5569695830345154, "learning_rate": 2.773023588959506e-05, "loss": 0.0418, "step": 66460 }, { "grad_norm": 0.4298098683357239, "learning_rate": 2.7715433029445555e-05, "loss": 0.0384, "step": 66470 }, { "grad_norm": 0.46766188740730286, "learning_rate": 2.7700632606302766e-05, "loss": 0.0341, "step": 66480 }, { "grad_norm": 0.540093719959259, "learning_rate": 2.768583462178528e-05, "loss": 0.0408, "step": 66490 }, { "grad_norm": 0.666773796081543, "learning_rate": 2.767103907751136e-05, "loss": 0.035, "step": 66500 }, { "grad_norm": 0.7313019633293152, "learning_rate": 2.7656245975099065e-05, "loss": 0.0359, "step": 66510 }, { "grad_norm": 0.6214317083358765, "learning_rate": 2.7641455316166093e-05, "loss": 0.0429, "step": 66520 }, { "grad_norm": 0.3806876242160797, "learning_rate": 2.7626667102329973e-05, "loss": 0.0366, "step": 66530 }, { "grad_norm": 0.37035509943962097, "learning_rate": 2.761188133520788e-05, "loss": 0.0341, "step": 66540 }, { "grad_norm": 0.9586029052734375, "learning_rate": 2.7597098016416784e-05, "loss": 0.0428, "step": 66550 }, { "grad_norm": 0.5277143716812134, "learning_rate": 2.7582317147573368e-05, "loss": 0.0335, "step": 66560 }, { "grad_norm": 0.7872401475906372, "learning_rate": 2.7567538730294064e-05, "loss": 0.0451, "step": 66570 }, { "grad_norm": 0.8324618935585022, "learning_rate": 2.7552762766194973e-05, "loss": 0.0444, "step": 66580 }, { "grad_norm": 0.46097058057785034, "learning_rate": 2.753798925689202e-05, "loss": 0.0385, "step": 66590 }, { "grad_norm": 1.2723941802978516, "learning_rate": 2.7523218204000774e-05, "loss": 0.0375, "step": 66600 }, { "grad_norm": 0.3767600357532501, "learning_rate": 2.7508449609136612e-05, "loss": 0.0302, "step": 66610 }, { "grad_norm": 0.5256187319755554, "learning_rate": 2.749368347391454e-05, "loss": 0.0517, "step": 66620 }, { "grad_norm": 1.300622582435608, "learning_rate": 2.7478919799949443e-05, "loss": 0.0464, "step": 66630 }, { "grad_norm": 0.4123672544956207, "learning_rate": 2.7464158588855803e-05, "loss": 0.039, "step": 66640 }, { "grad_norm": 0.49023905396461487, "learning_rate": 2.7449399842247918e-05, "loss": 0.0306, "step": 66650 }, { "grad_norm": 0.42186981439590454, "learning_rate": 2.7434643561739736e-05, "loss": 0.0368, "step": 66660 }, { "grad_norm": 0.5124906897544861, "learning_rate": 2.7419889748945016e-05, "loss": 0.0306, "step": 66670 }, { "grad_norm": 0.3442326486110687, "learning_rate": 2.7405138405477203e-05, "loss": 0.0429, "step": 66680 }, { "grad_norm": 0.3321817219257355, "learning_rate": 2.73903895329495e-05, "loss": 0.045, "step": 66690 }, { "grad_norm": 0.6675534248352051, "learning_rate": 2.737564313297478e-05, "loss": 0.0409, "step": 66700 }, { "grad_norm": 0.387445330619812, "learning_rate": 2.736089920716572e-05, "loss": 0.0397, "step": 66710 }, { "grad_norm": 0.45939743518829346, "learning_rate": 2.7346157757134706e-05, "loss": 0.0369, "step": 66720 }, { "grad_norm": 1.3375476598739624, "learning_rate": 2.733141878449379e-05, "loss": 0.0356, "step": 66730 }, { "grad_norm": 0.26766860485076904, "learning_rate": 2.7316682290854844e-05, "loss": 0.0405, "step": 66740 }, { "grad_norm": 0.4037444293498993, "learning_rate": 2.7301948277829414e-05, "loss": 0.04, "step": 66750 }, { "grad_norm": 0.4853421151638031, "learning_rate": 2.7287216747028814e-05, "loss": 0.0398, "step": 66760 }, { "grad_norm": 0.535270631313324, "learning_rate": 2.7272487700064025e-05, "loss": 0.0382, "step": 66770 }, { "grad_norm": 0.48024317622184753, "learning_rate": 2.725776113854583e-05, "loss": 0.0472, "step": 66780 }, { "grad_norm": 0.3836852014064789, "learning_rate": 2.7243037064084664e-05, "loss": 0.0337, "step": 66790 }, { "grad_norm": 0.4811525344848633, "learning_rate": 2.7228315478290766e-05, "loss": 0.0463, "step": 66800 }, { "grad_norm": 0.35273081064224243, "learning_rate": 2.721359638277402e-05, "loss": 0.0358, "step": 66810 }, { "grad_norm": 0.5295715928077698, "learning_rate": 2.7198879779144148e-05, "loss": 0.0296, "step": 66820 }, { "grad_norm": 0.3483120799064636, "learning_rate": 2.7184165669010485e-05, "loss": 0.0354, "step": 66830 }, { "grad_norm": 0.6118720769882202, "learning_rate": 2.716945405398218e-05, "loss": 0.0347, "step": 66840 }, { "grad_norm": 0.4897468686103821, "learning_rate": 2.715474493566803e-05, "loss": 0.0334, "step": 66850 }, { "grad_norm": 0.4943317174911499, "learning_rate": 2.7140038315676652e-05, "loss": 0.0277, "step": 66860 }, { "grad_norm": 0.3213736414909363, "learning_rate": 2.7125334195616275e-05, "loss": 0.0611, "step": 66870 }, { "grad_norm": 0.43276381492614746, "learning_rate": 2.7110632577094997e-05, "loss": 0.039, "step": 66880 }, { "grad_norm": 0.4012767970561981, "learning_rate": 2.709593346172051e-05, "loss": 0.0415, "step": 66890 }, { "grad_norm": 0.4235313832759857, "learning_rate": 2.708123685110032e-05, "loss": 0.034, "step": 66900 }, { "grad_norm": 0.3772922456264496, "learning_rate": 2.7066542746841588e-05, "loss": 0.0228, "step": 66910 }, { "grad_norm": 0.4674576222896576, "learning_rate": 2.7051851150551267e-05, "loss": 0.037, "step": 66920 }, { "grad_norm": 0.3969009220600128, "learning_rate": 2.7037162063835996e-05, "loss": 0.0328, "step": 66930 }, { "grad_norm": 0.7144719362258911, "learning_rate": 2.702247548830218e-05, "loss": 0.042, "step": 66940 }, { "grad_norm": 0.836223840713501, "learning_rate": 2.7007791425555883e-05, "loss": 0.0361, "step": 66950 }, { "grad_norm": 0.2902778685092926, "learning_rate": 2.6993109877202945e-05, "loss": 0.0348, "step": 66960 }, { "grad_norm": 0.47573307156562805, "learning_rate": 2.6978430844848935e-05, "loss": 0.0381, "step": 66970 }, { "grad_norm": 0.5216289162635803, "learning_rate": 2.69637543300991e-05, "loss": 0.0346, "step": 66980 }, { "grad_norm": 0.3325619697570801, "learning_rate": 2.694908033455848e-05, "loss": 0.0362, "step": 66990 }, { "grad_norm": 0.44630908966064453, "learning_rate": 2.6934408859831733e-05, "loss": 0.0412, "step": 67000 }, { "grad_norm": 0.5356078147888184, "learning_rate": 2.69197399075234e-05, "loss": 0.0428, "step": 67010 }, { "grad_norm": 0.6415482759475708, "learning_rate": 2.6905073479237584e-05, "loss": 0.0383, "step": 67020 }, { "grad_norm": 0.6297237277030945, "learning_rate": 2.6890409576578235e-05, "loss": 0.0455, "step": 67030 }, { "grad_norm": 1.21418035030365, "learning_rate": 2.687574820114892e-05, "loss": 0.0369, "step": 67040 }, { "grad_norm": 0.6225712895393372, "learning_rate": 2.686108935455305e-05, "loss": 0.042, "step": 67050 }, { "grad_norm": 0.5107100009918213, "learning_rate": 2.6846433038393603e-05, "loss": 0.0383, "step": 67060 }, { "grad_norm": 0.48718082904815674, "learning_rate": 2.6831779254273482e-05, "loss": 0.035, "step": 67070 }, { "grad_norm": 0.8878793120384216, "learning_rate": 2.6817128003795122e-05, "loss": 0.032, "step": 67080 }, { "grad_norm": 0.41419005393981934, "learning_rate": 2.680247928856081e-05, "loss": 0.0331, "step": 67090 }, { "grad_norm": 0.47895944118499756, "learning_rate": 2.6787833110172466e-05, "loss": 0.0416, "step": 67100 }, { "grad_norm": 0.8787256479263306, "learning_rate": 2.6773189470231807e-05, "loss": 0.0329, "step": 67110 }, { "grad_norm": 0.5736607313156128, "learning_rate": 2.675854837034022e-05, "loss": 0.0373, "step": 67120 }, { "grad_norm": 0.3917074203491211, "learning_rate": 2.6743909812098833e-05, "loss": 0.0347, "step": 67130 }, { "grad_norm": 0.27113738656044006, "learning_rate": 2.67292737971085e-05, "loss": 0.0395, "step": 67140 }, { "grad_norm": 0.6338295936584473, "learning_rate": 2.671464032696982e-05, "loss": 0.0392, "step": 67150 }, { "grad_norm": 0.45263463258743286, "learning_rate": 2.6700009403283055e-05, "loss": 0.0421, "step": 67160 }, { "grad_norm": 0.5808085799217224, "learning_rate": 2.6685381027648214e-05, "loss": 0.0387, "step": 67170 }, { "grad_norm": 0.500919759273529, "learning_rate": 2.6670755201665086e-05, "loss": 0.0503, "step": 67180 }, { "grad_norm": 0.47015342116355896, "learning_rate": 2.6656131926933066e-05, "loss": 0.0413, "step": 67190 }, { "grad_norm": 0.727434515953064, "learning_rate": 2.664151120505136e-05, "loss": 0.036, "step": 67200 }, { "grad_norm": 0.3700834810733795, "learning_rate": 2.6626893037618873e-05, "loss": 0.0337, "step": 67210 }, { "grad_norm": 0.5669971704483032, "learning_rate": 2.6612277426234243e-05, "loss": 0.0595, "step": 67220 }, { "grad_norm": 0.556877076625824, "learning_rate": 2.659766437249577e-05, "loss": 0.0535, "step": 67230 }, { "grad_norm": 0.498449444770813, "learning_rate": 2.6583053878001557e-05, "loss": 0.0369, "step": 67240 }, { "grad_norm": 0.7413736581802368, "learning_rate": 2.6568445944349323e-05, "loss": 0.0434, "step": 67250 }, { "grad_norm": 0.5274490714073181, "learning_rate": 2.6553840573136657e-05, "loss": 0.0363, "step": 67260 }, { "grad_norm": 0.5459637641906738, "learning_rate": 2.6539237765960706e-05, "loss": 0.0394, "step": 67270 }, { "grad_norm": 1.169614553451538, "learning_rate": 2.6524637524418465e-05, "loss": 0.0433, "step": 67280 }, { "grad_norm": 0.2801419496536255, "learning_rate": 2.651003985010655e-05, "loss": 0.0306, "step": 67290 }, { "grad_norm": 0.4521235525608063, "learning_rate": 2.649544474462138e-05, "loss": 0.0417, "step": 67300 }, { "grad_norm": 0.42985066771507263, "learning_rate": 2.6480852209559015e-05, "loss": 0.0405, "step": 67310 }, { "grad_norm": 0.5326300263404846, "learning_rate": 2.6466262246515296e-05, "loss": 0.0346, "step": 67320 }, { "grad_norm": 0.4614045023918152, "learning_rate": 2.6451674857085746e-05, "loss": 0.0403, "step": 67330 }, { "grad_norm": 0.7685893774032593, "learning_rate": 2.6437090042865655e-05, "loss": 0.0504, "step": 67340 }, { "grad_norm": 0.25530534982681274, "learning_rate": 2.6422507805449947e-05, "loss": 0.0349, "step": 67350 }, { "grad_norm": 0.37572169303894043, "learning_rate": 2.6407928146433357e-05, "loss": 0.0355, "step": 67360 }, { "grad_norm": 1.046558141708374, "learning_rate": 2.639335106741026e-05, "loss": 0.0351, "step": 67370 }, { "grad_norm": 0.42743584513664246, "learning_rate": 2.6378776569974796e-05, "loss": 0.0435, "step": 67380 }, { "grad_norm": 10.313703536987305, "learning_rate": 2.6364204655720814e-05, "loss": 0.0339, "step": 67390 }, { "grad_norm": 0.47739377617836, "learning_rate": 2.6349635326241894e-05, "loss": 0.036, "step": 67400 }, { "grad_norm": 0.48226475715637207, "learning_rate": 2.633506858313127e-05, "loss": 0.0408, "step": 67410 }, { "grad_norm": 0.2713472843170166, "learning_rate": 2.6320504427981975e-05, "loss": 0.0353, "step": 67420 }, { "grad_norm": 0.3189360201358795, "learning_rate": 2.6305942862386734e-05, "loss": 0.0334, "step": 67430 }, { "grad_norm": 0.7987059950828552, "learning_rate": 2.629138388793793e-05, "loss": 0.0347, "step": 67440 }, { "grad_norm": 0.44236645102500916, "learning_rate": 2.6276827506227743e-05, "loss": 0.036, "step": 67450 }, { "grad_norm": 0.4075104594230652, "learning_rate": 2.6262273718848037e-05, "loss": 0.0311, "step": 67460 }, { "grad_norm": 0.6443784236907959, "learning_rate": 2.6247722527390407e-05, "loss": 0.0365, "step": 67470 }, { "grad_norm": 0.47139275074005127, "learning_rate": 2.6233173933446104e-05, "loss": 0.0432, "step": 67480 }, { "grad_norm": 0.7602839469909668, "learning_rate": 2.6218627938606188e-05, "loss": 0.0381, "step": 67490 }, { "grad_norm": 0.3462536036968231, "learning_rate": 2.620408454446135e-05, "loss": 0.0368, "step": 67500 }, { "grad_norm": 0.2440725713968277, "learning_rate": 2.6189543752602046e-05, "loss": 0.0315, "step": 67510 }, { "grad_norm": 0.573536217212677, "learning_rate": 2.6175005564618444e-05, "loss": 0.0392, "step": 67520 }, { "grad_norm": 0.4390859305858612, "learning_rate": 2.6160469982100428e-05, "loss": 0.0392, "step": 67530 }, { "grad_norm": 0.5913974642753601, "learning_rate": 2.6145937006637555e-05, "loss": 0.0594, "step": 67540 }, { "grad_norm": 0.3385263681411743, "learning_rate": 2.6131406639819166e-05, "loss": 0.0379, "step": 67550 }, { "grad_norm": 0.2412824034690857, "learning_rate": 2.611687888323424e-05, "loss": 0.0299, "step": 67560 }, { "grad_norm": 0.5412945747375488, "learning_rate": 2.610235373847153e-05, "loss": 0.0382, "step": 67570 }, { "grad_norm": 0.6068228483200073, "learning_rate": 2.6087831207119494e-05, "loss": 0.0404, "step": 67580 }, { "grad_norm": 0.5810843706130981, "learning_rate": 2.60733112907663e-05, "loss": 0.0461, "step": 67590 }, { "grad_norm": 0.5652711987495422, "learning_rate": 2.6058793990999786e-05, "loss": 0.0461, "step": 67600 }, { "grad_norm": 0.4525730311870575, "learning_rate": 2.6044279309407593e-05, "loss": 0.0382, "step": 67610 }, { "grad_norm": 0.43341726064682007, "learning_rate": 2.6029767247576976e-05, "loss": 0.0442, "step": 67620 }, { "grad_norm": 0.4174865186214447, "learning_rate": 2.601525780709497e-05, "loss": 0.034, "step": 67630 }, { "grad_norm": 0.5304103493690491, "learning_rate": 2.6000750989548318e-05, "loss": 0.0355, "step": 67640 }, { "grad_norm": 0.47804224491119385, "learning_rate": 2.5986246796523462e-05, "loss": 0.0283, "step": 67650 }, { "grad_norm": 0.33628225326538086, "learning_rate": 2.5971745229606546e-05, "loss": 0.0347, "step": 67660 }, { "grad_norm": 0.4941173791885376, "learning_rate": 2.5957246290383442e-05, "loss": 0.0411, "step": 67670 }, { "grad_norm": 0.7318733334541321, "learning_rate": 2.594274998043975e-05, "loss": 0.0391, "step": 67680 }, { "grad_norm": 0.3172076642513275, "learning_rate": 2.5928256301360733e-05, "loss": 0.0345, "step": 67690 }, { "grad_norm": 1.4363906383514404, "learning_rate": 2.591376525473142e-05, "loss": 0.0298, "step": 67700 }, { "grad_norm": 0.41272488236427307, "learning_rate": 2.5899276842136523e-05, "loss": 0.0393, "step": 67710 }, { "grad_norm": 0.23211701214313507, "learning_rate": 2.5884791065160495e-05, "loss": 0.0419, "step": 67720 }, { "grad_norm": 0.5240070223808289, "learning_rate": 2.5870307925387436e-05, "loss": 0.0331, "step": 67730 }, { "grad_norm": 0.5677525401115417, "learning_rate": 2.585582742440125e-05, "loss": 0.0323, "step": 67740 }, { "grad_norm": 0.49151840806007385, "learning_rate": 2.5841349563785455e-05, "loss": 0.0423, "step": 67750 }, { "grad_norm": 0.1630176603794098, "learning_rate": 2.5826874345123355e-05, "loss": 0.0295, "step": 67760 }, { "grad_norm": 0.3231022357940674, "learning_rate": 2.581240176999792e-05, "loss": 0.0316, "step": 67770 }, { "grad_norm": 0.6035572290420532, "learning_rate": 2.579793183999189e-05, "loss": 0.0382, "step": 67780 }, { "grad_norm": 0.5519751906394958, "learning_rate": 2.5783464556687627e-05, "loss": 0.0415, "step": 67790 }, { "grad_norm": 0.4940846860408783, "learning_rate": 2.576899992166729e-05, "loss": 0.0339, "step": 67800 }, { "grad_norm": 0.329460084438324, "learning_rate": 2.5754537936512668e-05, "loss": 0.0295, "step": 67810 }, { "grad_norm": 0.7235486507415771, "learning_rate": 2.574007860280535e-05, "loss": 0.049, "step": 67820 }, { "grad_norm": 0.341430127620697, "learning_rate": 2.5725621922126518e-05, "loss": 0.0295, "step": 67830 }, { "grad_norm": 0.4084756374359131, "learning_rate": 2.571116789605722e-05, "loss": 0.0357, "step": 67840 }, { "grad_norm": 0.2691735327243805, "learning_rate": 2.5696716526178054e-05, "loss": 0.042, "step": 67850 }, { "grad_norm": 0.6257413029670715, "learning_rate": 2.568226781406946e-05, "loss": 0.0399, "step": 67860 }, { "grad_norm": 0.40802642703056335, "learning_rate": 2.5667821761311474e-05, "loss": 0.0537, "step": 67870 }, { "grad_norm": 0.6126371026039124, "learning_rate": 2.5653378369483917e-05, "loss": 0.0291, "step": 67880 }, { "grad_norm": 0.4450187087059021, "learning_rate": 2.5638937640166292e-05, "loss": 0.0327, "step": 67890 }, { "grad_norm": 0.3724660575389862, "learning_rate": 2.5624499574937842e-05, "loss": 0.0309, "step": 67900 }, { "grad_norm": 0.3345942795276642, "learning_rate": 2.5610064175377456e-05, "loss": 0.0366, "step": 67910 }, { "grad_norm": 0.3767077326774597, "learning_rate": 2.5595631443063782e-05, "loss": 0.037, "step": 67920 }, { "grad_norm": 0.35128235816955566, "learning_rate": 2.5581201379575177e-05, "loss": 0.0315, "step": 67930 }, { "grad_norm": 0.4224465489387512, "learning_rate": 2.5566773986489655e-05, "loss": 0.0391, "step": 67940 }, { "grad_norm": 0.4624235928058624, "learning_rate": 2.5552349265385e-05, "loss": 0.0515, "step": 67950 }, { "grad_norm": 0.5281395316123962, "learning_rate": 2.5537927217838675e-05, "loss": 0.0283, "step": 67960 }, { "grad_norm": 0.6379057765007019, "learning_rate": 2.5523507845427864e-05, "loss": 0.0405, "step": 67970 }, { "grad_norm": 0.7210142016410828, "learning_rate": 2.5509091149729413e-05, "loss": 0.0362, "step": 67980 }, { "grad_norm": 0.4529575705528259, "learning_rate": 2.5494677132319962e-05, "loss": 0.0368, "step": 67990 }, { "grad_norm": 0.5067100524902344, "learning_rate": 2.548026579477575e-05, "loss": 0.0415, "step": 68000 }, { "grad_norm": 0.4669864773750305, "learning_rate": 2.5465857138672834e-05, "loss": 0.0329, "step": 68010 }, { "grad_norm": 0.4438662528991699, "learning_rate": 2.5451451165586848e-05, "loss": 0.0427, "step": 68020 }, { "grad_norm": 0.353190541267395, "learning_rate": 2.543704787709329e-05, "loss": 0.0338, "step": 68030 }, { "grad_norm": 0.5293075442314148, "learning_rate": 2.542264727476723e-05, "loss": 0.0289, "step": 68040 }, { "grad_norm": 0.4031767249107361, "learning_rate": 2.5408249360183535e-05, "loss": 0.0456, "step": 68050 }, { "grad_norm": 0.31486764550209045, "learning_rate": 2.5393854134916694e-05, "loss": 0.0306, "step": 68060 }, { "grad_norm": 0.5187243819236755, "learning_rate": 2.5379461600540993e-05, "loss": 0.0407, "step": 68070 }, { "grad_norm": 0.3544539213180542, "learning_rate": 2.5365071758630317e-05, "loss": 0.0282, "step": 68080 }, { "grad_norm": 0.3760485053062439, "learning_rate": 2.535068461075839e-05, "loss": 0.0312, "step": 68090 }, { "grad_norm": 0.3766729533672333, "learning_rate": 2.5336300158498516e-05, "loss": 0.0371, "step": 68100 }, { "grad_norm": 1.4945967197418213, "learning_rate": 2.5321918403423793e-05, "loss": 0.0405, "step": 68110 }, { "grad_norm": 0.4207919239997864, "learning_rate": 2.5307539347106957e-05, "loss": 0.0346, "step": 68120 }, { "grad_norm": 0.42719560861587524, "learning_rate": 2.529316299112049e-05, "loss": 0.0423, "step": 68130 }, { "grad_norm": 0.5600806474685669, "learning_rate": 2.5278789337036567e-05, "loss": 0.031, "step": 68140 }, { "grad_norm": 0.46071091294288635, "learning_rate": 2.52644183864271e-05, "loss": 0.0342, "step": 68150 }, { "grad_norm": 0.3353048264980316, "learning_rate": 2.525005014086363e-05, "loss": 0.0405, "step": 68160 }, { "grad_norm": 0.5400593876838684, "learning_rate": 2.5235684601917465e-05, "loss": 0.0471, "step": 68170 }, { "grad_norm": 0.5048929452896118, "learning_rate": 2.5221321771159618e-05, "loss": 0.04, "step": 68180 }, { "grad_norm": 0.7402087450027466, "learning_rate": 2.5206961650160743e-05, "loss": 0.0421, "step": 68190 }, { "grad_norm": 0.2901172339916229, "learning_rate": 2.519260424049129e-05, "loss": 0.0383, "step": 68200 }, { "grad_norm": 0.38426050543785095, "learning_rate": 2.5178249543721304e-05, "loss": 0.0446, "step": 68210 }, { "grad_norm": 0.307662695646286, "learning_rate": 2.5163897561420668e-05, "loss": 0.0336, "step": 68220 }, { "grad_norm": 0.6865987181663513, "learning_rate": 2.514954829515883e-05, "loss": 0.0481, "step": 68230 }, { "grad_norm": 0.6481029391288757, "learning_rate": 2.513520174650505e-05, "loss": 0.037, "step": 68240 }, { "grad_norm": 0.5223102569580078, "learning_rate": 2.5120857917028207e-05, "loss": 0.0331, "step": 68250 }, { "grad_norm": 0.6117689609527588, "learning_rate": 2.5106516808296953e-05, "loss": 0.0279, "step": 68260 }, { "grad_norm": 0.4388372004032135, "learning_rate": 2.5092178421879552e-05, "loss": 0.0416, "step": 68270 }, { "grad_norm": 0.746080756187439, "learning_rate": 2.5077842759344116e-05, "loss": 0.0549, "step": 68280 }, { "grad_norm": 0.4639204442501068, "learning_rate": 2.5063509822258304e-05, "loss": 0.0379, "step": 68290 }, { "grad_norm": 0.8107840418815613, "learning_rate": 2.5049179612189576e-05, "loss": 0.034, "step": 68300 }, { "grad_norm": 0.5659812092781067, "learning_rate": 2.5034852130705043e-05, "loss": 0.0428, "step": 68310 }, { "grad_norm": 0.480378121137619, "learning_rate": 2.5020527379371572e-05, "loss": 0.0451, "step": 68320 }, { "grad_norm": 0.3296324908733368, "learning_rate": 2.500620535975564e-05, "loss": 0.0312, "step": 68330 }, { "grad_norm": 0.46577906608581543, "learning_rate": 2.4991886073423515e-05, "loss": 0.0402, "step": 68340 }, { "grad_norm": 0.2721540331840515, "learning_rate": 2.4977569521941134e-05, "loss": 0.031, "step": 68350 }, { "grad_norm": 0.44674843549728394, "learning_rate": 2.4963255706874143e-05, "loss": 0.038, "step": 68360 }, { "grad_norm": 0.4966926872730255, "learning_rate": 2.4948944629787853e-05, "loss": 0.037, "step": 68370 }, { "grad_norm": 0.6057369112968445, "learning_rate": 2.493463629224731e-05, "loss": 0.0423, "step": 68380 }, { "grad_norm": 0.4812600910663605, "learning_rate": 2.492033069581728e-05, "loss": 0.0416, "step": 68390 }, { "grad_norm": 0.6011479496955872, "learning_rate": 2.4906027842062137e-05, "loss": 0.0337, "step": 68400 }, { "grad_norm": 0.4249691963195801, "learning_rate": 2.48917277325461e-05, "loss": 0.0389, "step": 68410 }, { "grad_norm": 0.3657802939414978, "learning_rate": 2.487743036883295e-05, "loss": 0.0331, "step": 68420 }, { "grad_norm": 0.2839213013648987, "learning_rate": 2.486313575248626e-05, "loss": 0.0292, "step": 68430 }, { "grad_norm": 0.4324285686016083, "learning_rate": 2.4848843885069235e-05, "loss": 0.0347, "step": 68440 }, { "grad_norm": 0.5065982341766357, "learning_rate": 2.483455476814484e-05, "loss": 0.0343, "step": 68450 }, { "grad_norm": 0.41390007734298706, "learning_rate": 2.4820268403275672e-05, "loss": 0.0336, "step": 68460 }, { "grad_norm": 0.5317227840423584, "learning_rate": 2.480598479202413e-05, "loss": 0.0399, "step": 68470 }, { "grad_norm": 0.8452409505844116, "learning_rate": 2.4791703935952193e-05, "loss": 0.0433, "step": 68480 }, { "grad_norm": 0.4916956126689911, "learning_rate": 2.477742583662163e-05, "loss": 0.0394, "step": 68490 }, { "grad_norm": 0.5065693259239197, "learning_rate": 2.476315049559384e-05, "loss": 0.0342, "step": 68500 }, { "grad_norm": 0.8748573064804077, "learning_rate": 2.4748877914429987e-05, "loss": 0.0449, "step": 68510 }, { "grad_norm": 0.3713635802268982, "learning_rate": 2.4734608094690857e-05, "loss": 0.0408, "step": 68520 }, { "grad_norm": 0.27555862069129944, "learning_rate": 2.4720341037937006e-05, "loss": 0.0482, "step": 68530 }, { "grad_norm": 0.40852004289627075, "learning_rate": 2.4706076745728645e-05, "loss": 0.0468, "step": 68540 }, { "grad_norm": 0.38166263699531555, "learning_rate": 2.469181521962573e-05, "loss": 0.0406, "step": 68550 }, { "grad_norm": 0.7229251861572266, "learning_rate": 2.467755646118783e-05, "loss": 0.0301, "step": 68560 }, { "grad_norm": 0.6749442219734192, "learning_rate": 2.46633004719743e-05, "loss": 0.0404, "step": 68570 }, { "grad_norm": 0.6617091298103333, "learning_rate": 2.4649047253544123e-05, "loss": 0.0435, "step": 68580 }, { "grad_norm": 0.24774327874183655, "learning_rate": 2.4634796807456016e-05, "loss": 0.0318, "step": 68590 }, { "grad_norm": 0.33530765771865845, "learning_rate": 2.46205491352684e-05, "loss": 0.0396, "step": 68600 }, { "grad_norm": 0.3130866289138794, "learning_rate": 2.4606304238539375e-05, "loss": 0.0317, "step": 68610 }, { "grad_norm": 1.0308692455291748, "learning_rate": 2.4592062118826753e-05, "loss": 0.0369, "step": 68620 }, { "grad_norm": 0.4794594943523407, "learning_rate": 2.4577822777688004e-05, "loss": 0.0252, "step": 68630 }, { "grad_norm": 0.5185052752494812, "learning_rate": 2.4563586216680347e-05, "loss": 0.0383, "step": 68640 }, { "grad_norm": 0.34883302450180054, "learning_rate": 2.4549352437360622e-05, "loss": 0.0279, "step": 68650 }, { "grad_norm": 0.48553746938705444, "learning_rate": 2.4535121441285493e-05, "loss": 0.0422, "step": 68660 }, { "grad_norm": 1.074216604232788, "learning_rate": 2.4520893230011172e-05, "loss": 0.0367, "step": 68670 }, { "grad_norm": 0.21282275021076202, "learning_rate": 2.450666780509368e-05, "loss": 0.0342, "step": 68680 }, { "grad_norm": 0.37342241406440735, "learning_rate": 2.4492445168088656e-05, "loss": 0.035, "step": 68690 }, { "grad_norm": 0.4201303720474243, "learning_rate": 2.447822532055149e-05, "loss": 0.0438, "step": 68700 }, { "grad_norm": 0.7971006035804749, "learning_rate": 2.4464008264037212e-05, "loss": 0.039, "step": 68710 }, { "grad_norm": 0.5062553286552429, "learning_rate": 2.4449794000100605e-05, "loss": 0.0444, "step": 68720 }, { "grad_norm": 0.27989083528518677, "learning_rate": 2.4435582530296108e-05, "loss": 0.0337, "step": 68730 }, { "grad_norm": 0.5992090106010437, "learning_rate": 2.4421373856177887e-05, "loss": 0.0351, "step": 68740 }, { "grad_norm": 0.4656940698623657, "learning_rate": 2.440716797929975e-05, "loss": 0.0312, "step": 68750 }, { "grad_norm": 0.46611297130584717, "learning_rate": 2.439296490121526e-05, "loss": 0.0284, "step": 68760 }, { "grad_norm": 0.4184914827346802, "learning_rate": 2.4378764623477608e-05, "loss": 0.0313, "step": 68770 }, { "grad_norm": 0.4572141766548157, "learning_rate": 2.436456714763974e-05, "loss": 0.0326, "step": 68780 }, { "grad_norm": 0.4277198016643524, "learning_rate": 2.4350372475254262e-05, "loss": 0.0407, "step": 68790 }, { "grad_norm": 0.5017381310462952, "learning_rate": 2.433618060787351e-05, "loss": 0.0382, "step": 68800 }, { "grad_norm": 0.3903234004974365, "learning_rate": 2.432199154704944e-05, "loss": 0.0386, "step": 68810 }, { "grad_norm": 0.45592808723449707, "learning_rate": 2.430780529433377e-05, "loss": 0.0387, "step": 68820 }, { "grad_norm": 0.3948748707771301, "learning_rate": 2.42936218512779e-05, "loss": 0.0368, "step": 68830 }, { "grad_norm": 0.33994656801223755, "learning_rate": 2.427944121943288e-05, "loss": 0.0302, "step": 68840 }, { "grad_norm": 0.4065758287906647, "learning_rate": 2.4265263400349497e-05, "loss": 0.0386, "step": 68850 }, { "grad_norm": 0.5646661520004272, "learning_rate": 2.4251088395578213e-05, "loss": 0.0369, "step": 68860 }, { "grad_norm": 0.5917996168136597, "learning_rate": 2.4236916206669214e-05, "loss": 0.0439, "step": 68870 }, { "grad_norm": 0.582026481628418, "learning_rate": 2.4222746835172304e-05, "loss": 0.0437, "step": 68880 }, { "grad_norm": 0.5097042322158813, "learning_rate": 2.4208580282637065e-05, "loss": 0.0421, "step": 68890 }, { "grad_norm": 0.5972045063972473, "learning_rate": 2.4194416550612687e-05, "loss": 0.0321, "step": 68900 }, { "grad_norm": 0.8566651344299316, "learning_rate": 2.418025564064812e-05, "loss": 0.0348, "step": 68910 }, { "grad_norm": 0.6422281861305237, "learning_rate": 2.4166097554291978e-05, "loss": 0.0287, "step": 68920 }, { "grad_norm": 0.46820464730262756, "learning_rate": 2.4151942293092583e-05, "loss": 0.0323, "step": 68930 }, { "grad_norm": 0.2722328305244446, "learning_rate": 2.41377898585979e-05, "loss": 0.0283, "step": 68940 }, { "grad_norm": 0.5023215413093567, "learning_rate": 2.4123640252355652e-05, "loss": 0.0409, "step": 68950 }, { "grad_norm": 0.323487788438797, "learning_rate": 2.410949347591318e-05, "loss": 0.044, "step": 68960 }, { "grad_norm": 0.6828168630599976, "learning_rate": 2.4095349530817578e-05, "loss": 0.0544, "step": 68970 }, { "grad_norm": 0.3638501465320587, "learning_rate": 2.4081208418615603e-05, "loss": 0.0377, "step": 68980 }, { "grad_norm": 0.5667725801467896, "learning_rate": 2.4067070140853725e-05, "loss": 0.0479, "step": 68990 }, { "grad_norm": 0.3086825907230377, "learning_rate": 2.405293469907805e-05, "loss": 0.0433, "step": 69000 }, { "grad_norm": 0.494308739900589, "learning_rate": 2.4038802094834446e-05, "loss": 0.0337, "step": 69010 }, { "grad_norm": 0.5004290342330933, "learning_rate": 2.4024672329668395e-05, "loss": 0.0371, "step": 69020 }, { "grad_norm": 0.4601580500602722, "learning_rate": 2.4010545405125125e-05, "loss": 0.0296, "step": 69030 }, { "grad_norm": 0.6324974298477173, "learning_rate": 2.3996421322749528e-05, "loss": 0.0394, "step": 69040 }, { "grad_norm": 0.47499510645866394, "learning_rate": 2.3982300084086224e-05, "loss": 0.036, "step": 69050 }, { "grad_norm": 0.7283183932304382, "learning_rate": 2.396818169067944e-05, "loss": 0.0337, "step": 69060 }, { "grad_norm": 0.6788845062255859, "learning_rate": 2.3954066144073178e-05, "loss": 0.0381, "step": 69070 }, { "grad_norm": 0.8489265441894531, "learning_rate": 2.39399534458111e-05, "loss": 0.043, "step": 69080 }, { "grad_norm": 0.43953973054885864, "learning_rate": 2.392584359743651e-05, "loss": 0.0492, "step": 69090 }, { "grad_norm": 0.40646204352378845, "learning_rate": 2.3911736600492463e-05, "loss": 0.0266, "step": 69100 }, { "grad_norm": 0.5500551462173462, "learning_rate": 2.3897632456521672e-05, "loss": 0.0354, "step": 69110 }, { "grad_norm": 0.4031338393688202, "learning_rate": 2.388353116706658e-05, "loss": 0.0465, "step": 69120 }, { "grad_norm": 0.5467380285263062, "learning_rate": 2.3869432733669232e-05, "loss": 0.0388, "step": 69130 }, { "grad_norm": 0.39345595240592957, "learning_rate": 2.385533715787145e-05, "loss": 0.0431, "step": 69140 }, { "grad_norm": 0.3362426459789276, "learning_rate": 2.3841244441214673e-05, "loss": 0.0316, "step": 69150 }, { "grad_norm": 0.4773119390010834, "learning_rate": 2.3827154585240073e-05, "loss": 0.0373, "step": 69160 }, { "grad_norm": 0.5837053656578064, "learning_rate": 2.3813067591488498e-05, "loss": 0.0492, "step": 69170 }, { "grad_norm": 0.5581355094909668, "learning_rate": 2.3798983461500494e-05, "loss": 0.0463, "step": 69180 }, { "grad_norm": 1.0050950050354004, "learning_rate": 2.3784902196816256e-05, "loss": 0.0447, "step": 69190 }, { "grad_norm": 0.39206159114837646, "learning_rate": 2.3770823798975716e-05, "loss": 0.0364, "step": 69200 }, { "grad_norm": 0.43946337699890137, "learning_rate": 2.3756748269518424e-05, "loss": 0.0408, "step": 69210 }, { "grad_norm": 0.4204000234603882, "learning_rate": 2.374267560998371e-05, "loss": 0.0435, "step": 69220 }, { "grad_norm": 0.391367644071579, "learning_rate": 2.3728605821910477e-05, "loss": 0.0358, "step": 69230 }, { "grad_norm": 0.6042565703392029, "learning_rate": 2.3714538906837453e-05, "loss": 0.0459, "step": 69240 }, { "grad_norm": 0.508579671382904, "learning_rate": 2.3700474866302913e-05, "loss": 0.037, "step": 69250 }, { "grad_norm": 0.29590070247650146, "learning_rate": 2.3686413701844917e-05, "loss": 0.0344, "step": 69260 }, { "grad_norm": 0.39383774995803833, "learning_rate": 2.367235541500114e-05, "loss": 0.0332, "step": 69270 }, { "grad_norm": 0.2545414865016937, "learning_rate": 2.3658300007308993e-05, "loss": 0.0344, "step": 69280 }, { "grad_norm": 0.5636623501777649, "learning_rate": 2.364424748030555e-05, "loss": 0.0279, "step": 69290 }, { "grad_norm": 0.4440714716911316, "learning_rate": 2.3630197835527595e-05, "loss": 0.0429, "step": 69300 }, { "grad_norm": 0.3140801191329956, "learning_rate": 2.3616151074511546e-05, "loss": 0.0417, "step": 69310 }, { "grad_norm": 0.3824937641620636, "learning_rate": 2.3602107198793545e-05, "loss": 0.0374, "step": 69320 }, { "grad_norm": 0.448546439409256, "learning_rate": 2.3588066209909427e-05, "loss": 0.0359, "step": 69330 }, { "grad_norm": 0.6360067129135132, "learning_rate": 2.3574028109394662e-05, "loss": 0.0382, "step": 69340 }, { "grad_norm": 0.32021430134773254, "learning_rate": 2.3559992898784454e-05, "loss": 0.0375, "step": 69350 }, { "grad_norm": 0.39855024218559265, "learning_rate": 2.354596057961367e-05, "loss": 0.0307, "step": 69360 }, { "grad_norm": 0.34443747997283936, "learning_rate": 2.353193115341688e-05, "loss": 0.0388, "step": 69370 }, { "grad_norm": 0.3368326723575592, "learning_rate": 2.3517904621728294e-05, "loss": 0.0298, "step": 69380 }, { "grad_norm": 0.3259260058403015, "learning_rate": 2.3503880986081855e-05, "loss": 0.0353, "step": 69390 }, { "grad_norm": 0.4798264503479004, "learning_rate": 2.3489860248011136e-05, "loss": 0.0362, "step": 69400 }, { "grad_norm": 0.28999578952789307, "learning_rate": 2.3475842409049464e-05, "loss": 0.0372, "step": 69410 }, { "grad_norm": 0.7269556522369385, "learning_rate": 2.346182747072975e-05, "loss": 0.0454, "step": 69420 }, { "grad_norm": 0.39051780104637146, "learning_rate": 2.3447815434584718e-05, "loss": 0.031, "step": 69430 }, { "grad_norm": 0.8507480025291443, "learning_rate": 2.3433806302146656e-05, "loss": 0.0295, "step": 69440 }, { "grad_norm": 0.5239812135696411, "learning_rate": 2.3419800074947612e-05, "loss": 0.0413, "step": 69450 }, { "grad_norm": 0.5248023271560669, "learning_rate": 2.3405796754519248e-05, "loss": 0.044, "step": 69460 }, { "grad_norm": 0.6425431966781616, "learning_rate": 2.3391796342392992e-05, "loss": 0.0447, "step": 69470 }, { "grad_norm": 0.41635170578956604, "learning_rate": 2.337779884009984e-05, "loss": 0.0373, "step": 69480 }, { "grad_norm": 0.36771029233932495, "learning_rate": 2.3363804249170624e-05, "loss": 0.0358, "step": 69490 }, { "grad_norm": 1.1163207292556763, "learning_rate": 2.3349812571135703e-05, "loss": 0.0355, "step": 69500 }, { "grad_norm": 0.3972122073173523, "learning_rate": 2.333582380752523e-05, "loss": 0.028, "step": 69510 }, { "grad_norm": 0.3804510533809662, "learning_rate": 2.3321837959868963e-05, "loss": 0.0306, "step": 69520 }, { "grad_norm": 0.5281643271446228, "learning_rate": 2.3307855029696383e-05, "loss": 0.0308, "step": 69530 }, { "grad_norm": 0.7453190088272095, "learning_rate": 2.3293875018536648e-05, "loss": 0.0391, "step": 69540 }, { "grad_norm": 0.40518009662628174, "learning_rate": 2.327989792791861e-05, "loss": 0.0322, "step": 69550 }, { "grad_norm": 0.7523431777954102, "learning_rate": 2.326592375937074e-05, "loss": 0.0349, "step": 69560 }, { "grad_norm": 0.43585047125816345, "learning_rate": 2.3251952514421248e-05, "loss": 0.0321, "step": 69570 }, { "grad_norm": 0.5078245401382446, "learning_rate": 2.323798419459804e-05, "loss": 0.0369, "step": 69580 }, { "grad_norm": 0.4210977554321289, "learning_rate": 2.3224018801428625e-05, "loss": 0.0292, "step": 69590 }, { "grad_norm": 0.4586764872074127, "learning_rate": 2.3210056336440273e-05, "loss": 0.0285, "step": 69600 }, { "grad_norm": 0.4209192097187042, "learning_rate": 2.3196096801159844e-05, "loss": 0.0308, "step": 69610 }, { "grad_norm": 0.4194570481777191, "learning_rate": 2.3182140197114012e-05, "loss": 0.0375, "step": 69620 }, { "grad_norm": 0.44417816400527954, "learning_rate": 2.3168186525828985e-05, "loss": 0.0443, "step": 69630 }, { "grad_norm": 0.4069049060344696, "learning_rate": 2.3154235788830757e-05, "loss": 0.0314, "step": 69640 }, { "grad_norm": 0.43355825543403625, "learning_rate": 2.3140287987644927e-05, "loss": 0.0386, "step": 69650 }, { "grad_norm": 0.27501749992370605, "learning_rate": 2.3126343123796833e-05, "loss": 0.0306, "step": 69660 }, { "grad_norm": 0.4783511459827423, "learning_rate": 2.3112401198811418e-05, "loss": 0.0327, "step": 69670 }, { "grad_norm": 0.4775148332118988, "learning_rate": 2.3098462214213418e-05, "loss": 0.0382, "step": 69680 }, { "grad_norm": 0.5232803225517273, "learning_rate": 2.308452617152712e-05, "loss": 0.0367, "step": 69690 }, { "grad_norm": 0.6170586943626404, "learning_rate": 2.3070593072276603e-05, "loss": 0.0394, "step": 69700 }, { "grad_norm": 0.442111998796463, "learning_rate": 2.3056662917985518e-05, "loss": 0.0312, "step": 69710 }, { "grad_norm": 0.470766544342041, "learning_rate": 2.3042735710177283e-05, "loss": 0.0421, "step": 69720 }, { "grad_norm": 0.3756084740161896, "learning_rate": 2.302881145037492e-05, "loss": 0.0294, "step": 69730 }, { "grad_norm": 0.245224267244339, "learning_rate": 2.3014890140101192e-05, "loss": 0.0329, "step": 69740 }, { "grad_norm": 0.3537794351577759, "learning_rate": 2.30009717808785e-05, "loss": 0.028, "step": 69750 }, { "grad_norm": 0.33624541759490967, "learning_rate": 2.2987056374228967e-05, "loss": 0.031, "step": 69760 }, { "grad_norm": 0.3212878108024597, "learning_rate": 2.297314392167431e-05, "loss": 0.0401, "step": 69770 }, { "grad_norm": 0.410384863615036, "learning_rate": 2.295923442473601e-05, "loss": 0.0356, "step": 69780 }, { "grad_norm": 0.4752084016799927, "learning_rate": 2.294532788493519e-05, "loss": 0.0308, "step": 69790 }, { "grad_norm": 0.60956209897995, "learning_rate": 2.293142430379262e-05, "loss": 0.0338, "step": 69800 }, { "grad_norm": 0.7173731327056885, "learning_rate": 2.291752368282879e-05, "loss": 0.0379, "step": 69810 }, { "grad_norm": 0.3472878932952881, "learning_rate": 2.290362602356385e-05, "loss": 0.0321, "step": 69820 }, { "grad_norm": 0.6128770709037781, "learning_rate": 2.288973132751765e-05, "loss": 0.0421, "step": 69830 }, { "grad_norm": 0.4140855073928833, "learning_rate": 2.287583959620965e-05, "loss": 0.0403, "step": 69840 }, { "grad_norm": 0.34707263112068176, "learning_rate": 2.2861950831159074e-05, "loss": 0.0303, "step": 69850 }, { "grad_norm": 0.8183268308639526, "learning_rate": 2.284806503388472e-05, "loss": 0.0326, "step": 69860 }, { "grad_norm": 0.46082550287246704, "learning_rate": 2.283418220590518e-05, "loss": 0.0322, "step": 69870 }, { "grad_norm": 0.49825942516326904, "learning_rate": 2.282030234873862e-05, "loss": 0.0426, "step": 69880 }, { "grad_norm": 0.6245404481887817, "learning_rate": 2.2806425463902948e-05, "loss": 0.0236, "step": 69890 }, { "grad_norm": 0.5183817744255066, "learning_rate": 2.279255155291568e-05, "loss": 0.0364, "step": 69900 }, { "grad_norm": 0.3164271116256714, "learning_rate": 2.277868061729409e-05, "loss": 0.0263, "step": 69910 }, { "grad_norm": 0.6316968202590942, "learning_rate": 2.276481265855504e-05, "loss": 0.0297, "step": 69920 }, { "grad_norm": 0.387500137090683, "learning_rate": 2.2750947678215128e-05, "loss": 0.0322, "step": 69930 }, { "grad_norm": 0.8821156620979309, "learning_rate": 2.2737085677790616e-05, "loss": 0.0299, "step": 69940 }, { "grad_norm": 0.3866143524646759, "learning_rate": 2.2723226658797443e-05, "loss": 0.0321, "step": 69950 }, { "grad_norm": 0.3535332679748535, "learning_rate": 2.270937062275117e-05, "loss": 0.0302, "step": 69960 }, { "grad_norm": 0.25471171736717224, "learning_rate": 2.2695517571167113e-05, "loss": 0.0291, "step": 69970 }, { "grad_norm": 0.82001793384552, "learning_rate": 2.268166750556019e-05, "loss": 0.0378, "step": 69980 }, { "grad_norm": 0.267586886882782, "learning_rate": 2.266782042744503e-05, "loss": 0.0273, "step": 69990 }, { "grad_norm": 0.5569981336593628, "learning_rate": 2.2653976338335935e-05, "loss": 0.04, "step": 70000 }, { "grad_norm": 0.5114941000938416, "learning_rate": 2.2640135239746895e-05, "loss": 0.0397, "step": 70010 }, { "grad_norm": 0.583283007144928, "learning_rate": 2.2626297133191515e-05, "loss": 0.0338, "step": 70020 }, { "grad_norm": 0.6069692373275757, "learning_rate": 2.261246202018312e-05, "loss": 0.0345, "step": 70030 }, { "grad_norm": 0.31468477845191956, "learning_rate": 2.2598629902234718e-05, "loss": 0.0281, "step": 70040 }, { "grad_norm": 0.3384776711463928, "learning_rate": 2.258480078085894e-05, "loss": 0.035, "step": 70050 }, { "grad_norm": 0.9871075749397278, "learning_rate": 2.2570974657568126e-05, "loss": 0.0386, "step": 70060 }, { "grad_norm": 0.3207789957523346, "learning_rate": 2.2557151533874286e-05, "loss": 0.0378, "step": 70070 }, { "grad_norm": 0.4055212140083313, "learning_rate": 2.2543331411289116e-05, "loss": 0.0314, "step": 70080 }, { "grad_norm": 0.33414456248283386, "learning_rate": 2.252951429132392e-05, "loss": 0.046, "step": 70090 }, { "grad_norm": 0.3985218405723572, "learning_rate": 2.2515700175489762e-05, "loss": 0.0427, "step": 70100 }, { "grad_norm": 0.42663395404815674, "learning_rate": 2.2501889065297288e-05, "loss": 0.0316, "step": 70110 }, { "grad_norm": 0.32865768671035767, "learning_rate": 2.2488080962256887e-05, "loss": 0.0357, "step": 70120 }, { "grad_norm": 0.4630272388458252, "learning_rate": 2.2474275867878585e-05, "loss": 0.0484, "step": 70130 }, { "grad_norm": 0.4981573522090912, "learning_rate": 2.246047378367211e-05, "loss": 0.05, "step": 70140 }, { "grad_norm": 0.30203425884246826, "learning_rate": 2.2446674711146798e-05, "loss": 0.0305, "step": 70150 }, { "grad_norm": 0.8157743215560913, "learning_rate": 2.2432878651811734e-05, "loss": 0.0345, "step": 70160 }, { "grad_norm": 0.35467076301574707, "learning_rate": 2.2419085607175594e-05, "loss": 0.0392, "step": 70170 }, { "grad_norm": 0.37111082673072815, "learning_rate": 2.240529557874678e-05, "loss": 0.0333, "step": 70180 }, { "grad_norm": 0.4948335886001587, "learning_rate": 2.2391508568033358e-05, "loss": 0.0387, "step": 70190 }, { "grad_norm": 0.388833224773407, "learning_rate": 2.237772457654307e-05, "loss": 0.0412, "step": 70200 }, { "grad_norm": 0.37485820055007935, "learning_rate": 2.236394360578327e-05, "loss": 0.0396, "step": 70210 }, { "grad_norm": 0.41574546694755554, "learning_rate": 2.2350165657261073e-05, "loss": 0.0409, "step": 70220 }, { "grad_norm": 0.5350884199142456, "learning_rate": 2.2336390732483164e-05, "loss": 0.0388, "step": 70230 }, { "grad_norm": 0.542129397392273, "learning_rate": 2.2322618832955976e-05, "loss": 0.052, "step": 70240 }, { "grad_norm": 0.6740113496780396, "learning_rate": 2.2308849960185578e-05, "loss": 0.037, "step": 70250 }, { "grad_norm": 0.36261066794395447, "learning_rate": 2.229508411567774e-05, "loss": 0.0345, "step": 70260 }, { "grad_norm": 2.2230403423309326, "learning_rate": 2.2281321300937823e-05, "loss": 0.0461, "step": 70270 }, { "grad_norm": 0.3357667922973633, "learning_rate": 2.2267561517470937e-05, "loss": 0.0322, "step": 70280 }, { "grad_norm": 0.6896462440490723, "learning_rate": 2.2253804766781845e-05, "loss": 0.031, "step": 70290 }, { "grad_norm": 0.5019999146461487, "learning_rate": 2.2240051050374933e-05, "loss": 0.0392, "step": 70300 }, { "grad_norm": 0.484717458486557, "learning_rate": 2.22263003697543e-05, "loss": 0.0355, "step": 70310 }, { "grad_norm": 0.39283108711242676, "learning_rate": 2.2212552726423702e-05, "loss": 0.0339, "step": 70320 }, { "grad_norm": 0.37217819690704346, "learning_rate": 2.2198808121886578e-05, "loss": 0.0372, "step": 70330 }, { "grad_norm": 0.36512696743011475, "learning_rate": 2.218506655764598e-05, "loss": 0.0347, "step": 70340 }, { "grad_norm": 1.3123024702072144, "learning_rate": 2.2171328035204704e-05, "loss": 0.03, "step": 70350 }, { "grad_norm": 0.2955555319786072, "learning_rate": 2.2157592556065133e-05, "loss": 0.0317, "step": 70360 }, { "grad_norm": 0.3560616672039032, "learning_rate": 2.2143860121729382e-05, "loss": 0.0277, "step": 70370 }, { "grad_norm": 0.31469589471817017, "learning_rate": 2.2130130733699206e-05, "loss": 0.0362, "step": 70380 }, { "grad_norm": 0.6328287124633789, "learning_rate": 2.2116404393476052e-05, "loss": 0.0393, "step": 70390 }, { "grad_norm": 0.9981919527053833, "learning_rate": 2.2102681102560967e-05, "loss": 0.0345, "step": 70400 }, { "grad_norm": 0.906822919845581, "learning_rate": 2.208896086245476e-05, "loss": 0.0303, "step": 70410 }, { "grad_norm": 0.4313388466835022, "learning_rate": 2.2075243674657808e-05, "loss": 0.0358, "step": 70420 }, { "grad_norm": 0.359631210565567, "learning_rate": 2.206152954067024e-05, "loss": 0.0281, "step": 70430 }, { "grad_norm": 0.28485989570617676, "learning_rate": 2.2047818461991758e-05, "loss": 0.0369, "step": 70440 }, { "grad_norm": 0.3938944637775421, "learning_rate": 2.2034110440121864e-05, "loss": 0.0334, "step": 70450 }, { "grad_norm": 0.40115389227867126, "learning_rate": 2.2020405476559585e-05, "loss": 0.0326, "step": 70460 }, { "grad_norm": 0.46256712079048157, "learning_rate": 2.2006703572803722e-05, "loss": 0.0384, "step": 70470 }, { "grad_norm": 0.4957091510295868, "learning_rate": 2.1993004730352646e-05, "loss": 0.0388, "step": 70480 }, { "grad_norm": 0.30168595910072327, "learning_rate": 2.1979308950704464e-05, "loss": 0.0264, "step": 70490 }, { "grad_norm": 0.33850783109664917, "learning_rate": 2.1965616235356923e-05, "loss": 0.0425, "step": 70500 }, { "grad_norm": 0.5134997963905334, "learning_rate": 2.1951926585807465e-05, "loss": 0.0313, "step": 70510 }, { "grad_norm": 0.28896138072013855, "learning_rate": 2.193824000355313e-05, "loss": 0.0359, "step": 70520 }, { "grad_norm": 0.58396977186203, "learning_rate": 2.192455649009067e-05, "loss": 0.0423, "step": 70530 }, { "grad_norm": 0.47127681970596313, "learning_rate": 2.1910876046916523e-05, "loss": 0.038, "step": 70540 }, { "grad_norm": 0.18836185336112976, "learning_rate": 2.1897198675526725e-05, "loss": 0.0296, "step": 70550 }, { "grad_norm": 0.4479982256889343, "learning_rate": 2.1883524377417024e-05, "loss": 0.033, "step": 70560 }, { "grad_norm": 0.44332221150398254, "learning_rate": 2.1869853154082825e-05, "loss": 0.034, "step": 70570 }, { "grad_norm": 0.48687177896499634, "learning_rate": 2.1856185007019215e-05, "loss": 0.0373, "step": 70580 }, { "grad_norm": 0.33567729592323303, "learning_rate": 2.1842519937720874e-05, "loss": 0.0384, "step": 70590 }, { "grad_norm": 0.5269050002098083, "learning_rate": 2.182885794768224e-05, "loss": 0.0523, "step": 70600 }, { "grad_norm": 0.45301300287246704, "learning_rate": 2.1815199038397326e-05, "loss": 0.0408, "step": 70610 }, { "grad_norm": 0.9765554666519165, "learning_rate": 2.1801543211359893e-05, "loss": 0.0351, "step": 70620 }, { "grad_norm": 0.3546133041381836, "learning_rate": 2.178789046806326e-05, "loss": 0.0361, "step": 70630 }, { "grad_norm": 0.8463224768638611, "learning_rate": 2.1774240810000547e-05, "loss": 0.0325, "step": 70640 }, { "grad_norm": 0.4605032205581665, "learning_rate": 2.17605942386644e-05, "loss": 0.0463, "step": 70650 }, { "grad_norm": 0.4014095962047577, "learning_rate": 2.174695075554723e-05, "loss": 0.0464, "step": 70660 }, { "grad_norm": 0.5269991159439087, "learning_rate": 2.173331036214103e-05, "loss": 0.0444, "step": 70670 }, { "grad_norm": 0.724331796169281, "learning_rate": 2.171967305993753e-05, "loss": 0.0324, "step": 70680 }, { "grad_norm": 0.4759090840816498, "learning_rate": 2.1706038850428024e-05, "loss": 0.0371, "step": 70690 }, { "grad_norm": 0.5228233337402344, "learning_rate": 2.169240773510361e-05, "loss": 0.0262, "step": 70700 }, { "grad_norm": 0.41196009516716003, "learning_rate": 2.1678779715454906e-05, "loss": 0.0398, "step": 70710 }, { "grad_norm": 0.36685410141944885, "learning_rate": 2.166515479297229e-05, "loss": 0.0339, "step": 70720 }, { "grad_norm": 0.40464985370635986, "learning_rate": 2.1651532969145732e-05, "loss": 0.0309, "step": 70730 }, { "grad_norm": 0.26622244715690613, "learning_rate": 2.1637914245464903e-05, "loss": 0.0391, "step": 70740 }, { "grad_norm": 0.4363147020339966, "learning_rate": 2.1624298623419133e-05, "loss": 0.0317, "step": 70750 }, { "grad_norm": 0.6916325092315674, "learning_rate": 2.1610686104497413e-05, "loss": 0.0395, "step": 70760 }, { "grad_norm": 0.6469757556915283, "learning_rate": 2.1597076690188366e-05, "loss": 0.0362, "step": 70770 }, { "grad_norm": 0.41655251383781433, "learning_rate": 2.1583470381980315e-05, "loss": 0.0384, "step": 70780 }, { "grad_norm": 0.44957974553108215, "learning_rate": 2.156986718136123e-05, "loss": 0.0388, "step": 70790 }, { "grad_norm": 0.27418848872184753, "learning_rate": 2.155626708981871e-05, "loss": 0.0306, "step": 70800 }, { "grad_norm": 0.6313270926475525, "learning_rate": 2.1542670108840075e-05, "loss": 0.0369, "step": 70810 }, { "grad_norm": 0.5055649876594543, "learning_rate": 2.152907623991221e-05, "loss": 0.0374, "step": 70820 }, { "grad_norm": 0.5075629949569702, "learning_rate": 2.1515485484521802e-05, "loss": 0.0318, "step": 70830 }, { "grad_norm": 0.31677016615867615, "learning_rate": 2.150189784415506e-05, "loss": 0.0309, "step": 70840 }, { "grad_norm": 0.334768146276474, "learning_rate": 2.148831332029794e-05, "loss": 0.0292, "step": 70850 }, { "grad_norm": 0.40380144119262695, "learning_rate": 2.147473191443599e-05, "loss": 0.0307, "step": 70860 }, { "grad_norm": 0.2421848177909851, "learning_rate": 2.1461153628054492e-05, "loss": 0.0274, "step": 70870 }, { "grad_norm": 0.31131473183631897, "learning_rate": 2.1447578462638284e-05, "loss": 0.039, "step": 70880 }, { "grad_norm": 0.489764004945755, "learning_rate": 2.143400641967201e-05, "loss": 0.026, "step": 70890 }, { "grad_norm": 0.3184635043144226, "learning_rate": 2.1420437500639823e-05, "loss": 0.0281, "step": 70900 }, { "grad_norm": 1.3675453662872314, "learning_rate": 2.1406871707025646e-05, "loss": 0.0387, "step": 70910 }, { "grad_norm": 0.5056453347206116, "learning_rate": 2.1393309040312963e-05, "loss": 0.0525, "step": 70920 }, { "grad_norm": 0.38446518778800964, "learning_rate": 2.137974950198502e-05, "loss": 0.0345, "step": 70930 }, { "grad_norm": 0.30568116903305054, "learning_rate": 2.1366193093524617e-05, "loss": 0.0267, "step": 70940 }, { "grad_norm": 0.5279858708381653, "learning_rate": 2.1352639816414288e-05, "loss": 0.0368, "step": 70950 }, { "grad_norm": 0.33996447920799255, "learning_rate": 2.1339089672136202e-05, "loss": 0.0338, "step": 70960 }, { "grad_norm": 0.5417195558547974, "learning_rate": 2.1325542662172193e-05, "loss": 0.0331, "step": 70970 }, { "grad_norm": 0.387283056974411, "learning_rate": 2.1311998788003712e-05, "loss": 0.0272, "step": 70980 }, { "grad_norm": 0.3808833360671997, "learning_rate": 2.1298458051111907e-05, "loss": 0.0391, "step": 70990 }, { "grad_norm": 0.39984428882598877, "learning_rate": 2.1284920452977602e-05, "loss": 0.0289, "step": 71000 }, { "grad_norm": 0.49446266889572144, "learning_rate": 2.12713859950812e-05, "loss": 0.0342, "step": 71010 }, { "grad_norm": 0.5007971525192261, "learning_rate": 2.1257854678902832e-05, "loss": 0.0345, "step": 71020 }, { "grad_norm": 0.5114337801933289, "learning_rate": 2.124432650592227e-05, "loss": 0.0421, "step": 71030 }, { "grad_norm": 0.7474220395088196, "learning_rate": 2.1230801477618937e-05, "loss": 0.0411, "step": 71040 }, { "grad_norm": 0.4548342823982239, "learning_rate": 2.121727959547189e-05, "loss": 0.0353, "step": 71050 }, { "grad_norm": 0.2921213209629059, "learning_rate": 2.1203760860959886e-05, "loss": 0.0294, "step": 71060 }, { "grad_norm": 0.5348349809646606, "learning_rate": 2.119024527556127e-05, "loss": 0.0409, "step": 71070 }, { "grad_norm": 0.5978463888168335, "learning_rate": 2.117673284075415e-05, "loss": 0.0274, "step": 71080 }, { "grad_norm": 0.7347913980484009, "learning_rate": 2.1163223558016165e-05, "loss": 0.0271, "step": 71090 }, { "grad_norm": 0.3667486310005188, "learning_rate": 2.1149717428824722e-05, "loss": 0.0348, "step": 71100 }, { "grad_norm": 0.5603308081626892, "learning_rate": 2.1136214454656784e-05, "loss": 0.0401, "step": 71110 }, { "grad_norm": 1.3628596067428589, "learning_rate": 2.1122714636989056e-05, "loss": 0.0387, "step": 71120 }, { "grad_norm": 1.641710877418518, "learning_rate": 2.1109217977297823e-05, "loss": 0.0401, "step": 71130 }, { "grad_norm": 0.39203721284866333, "learning_rate": 2.1095724477059077e-05, "loss": 0.0271, "step": 71140 }, { "grad_norm": 0.3840814232826233, "learning_rate": 2.1082234137748434e-05, "loss": 0.0244, "step": 71150 }, { "grad_norm": 0.3940206468105316, "learning_rate": 2.106874696084122e-05, "loss": 0.04, "step": 71160 }, { "grad_norm": 0.41718244552612305, "learning_rate": 2.1055262947812315e-05, "loss": 0.0384, "step": 71170 }, { "grad_norm": 0.30486345291137695, "learning_rate": 2.104178210013636e-05, "loss": 0.0523, "step": 71180 }, { "grad_norm": 0.7556307911872864, "learning_rate": 2.1028304419287554e-05, "loss": 0.0393, "step": 71190 }, { "grad_norm": 0.8425025939941406, "learning_rate": 2.101482990673982e-05, "loss": 0.0347, "step": 71200 }, { "grad_norm": 0.36061012744903564, "learning_rate": 2.1001358563966707e-05, "loss": 0.0385, "step": 71210 }, { "grad_norm": 0.5708689093589783, "learning_rate": 2.098789039244144e-05, "loss": 0.031, "step": 71220 }, { "grad_norm": 0.29577475786209106, "learning_rate": 2.097442539363685e-05, "loss": 0.0305, "step": 71230 }, { "grad_norm": 0.38896578550338745, "learning_rate": 2.0960963569025448e-05, "loss": 0.0306, "step": 71240 }, { "grad_norm": 0.28344830870628357, "learning_rate": 2.0947504920079435e-05, "loss": 0.0271, "step": 71250 }, { "grad_norm": 0.38108956813812256, "learning_rate": 2.093404944827059e-05, "loss": 0.0308, "step": 71260 }, { "grad_norm": 0.44279181957244873, "learning_rate": 2.0920597155070394e-05, "loss": 0.0361, "step": 71270 }, { "grad_norm": 0.6223527193069458, "learning_rate": 2.090714804194997e-05, "loss": 0.0439, "step": 71280 }, { "grad_norm": 0.19806207716464996, "learning_rate": 2.0893702110380114e-05, "loss": 0.0301, "step": 71290 }, { "grad_norm": 0.5894851684570312, "learning_rate": 2.0880259361831216e-05, "loss": 0.0349, "step": 71300 }, { "grad_norm": 0.4144997298717499, "learning_rate": 2.0866819797773397e-05, "loss": 0.0312, "step": 71310 }, { "grad_norm": 0.33855298161506653, "learning_rate": 2.085338341967634e-05, "loss": 0.0321, "step": 71320 }, { "grad_norm": 0.45589733123779297, "learning_rate": 2.083995022900946e-05, "loss": 0.0322, "step": 71330 }, { "grad_norm": 0.4586436152458191, "learning_rate": 2.0826520227241768e-05, "loss": 0.0344, "step": 71340 }, { "grad_norm": 0.5382913947105408, "learning_rate": 2.081309341584199e-05, "loss": 0.0355, "step": 71350 }, { "grad_norm": 0.40761658549308777, "learning_rate": 2.0799669796278413e-05, "loss": 0.0303, "step": 71360 }, { "grad_norm": 0.513237476348877, "learning_rate": 2.078624937001906e-05, "loss": 0.0439, "step": 71370 }, { "grad_norm": 1.673985242843628, "learning_rate": 2.0772832138531544e-05, "loss": 0.0638, "step": 71380 }, { "grad_norm": 0.3524630665779114, "learning_rate": 2.0759418103283158e-05, "loss": 0.031, "step": 71390 }, { "grad_norm": 1.0384421348571777, "learning_rate": 2.0746007265740846e-05, "loss": 0.0314, "step": 71400 }, { "grad_norm": 0.6521735787391663, "learning_rate": 2.0732599627371215e-05, "loss": 0.0454, "step": 71410 }, { "grad_norm": 0.3225102722644806, "learning_rate": 2.0719195189640468e-05, "loss": 0.0282, "step": 71420 }, { "grad_norm": 0.692538321018219, "learning_rate": 2.0705793954014528e-05, "loss": 0.0379, "step": 71430 }, { "grad_norm": 0.386958509683609, "learning_rate": 2.06923959219589e-05, "loss": 0.028, "step": 71440 }, { "grad_norm": 0.6951014399528503, "learning_rate": 2.0679001094938793e-05, "loss": 0.0299, "step": 71450 }, { "grad_norm": 0.373680979013443, "learning_rate": 2.066560947441904e-05, "loss": 0.0353, "step": 71460 }, { "grad_norm": 0.392782598733902, "learning_rate": 2.0652221061864152e-05, "loss": 0.0262, "step": 71470 }, { "grad_norm": 0.7395935654640198, "learning_rate": 2.0638835858738226e-05, "loss": 0.0393, "step": 71480 }, { "grad_norm": 0.30996766686439514, "learning_rate": 2.0625453866505068e-05, "loss": 0.0312, "step": 71490 }, { "grad_norm": 0.3075636029243469, "learning_rate": 2.0612075086628123e-05, "loss": 0.0339, "step": 71500 }, { "grad_norm": 0.4700322151184082, "learning_rate": 2.059869952057045e-05, "loss": 0.0322, "step": 71510 }, { "grad_norm": 0.4886191785335541, "learning_rate": 2.0585327169794793e-05, "loss": 0.0374, "step": 71520 }, { "grad_norm": 0.4641951024532318, "learning_rate": 2.0571958035763534e-05, "loss": 0.0321, "step": 71530 }, { "grad_norm": 0.46684253215789795, "learning_rate": 2.0558592119938715e-05, "loss": 0.0388, "step": 71540 }, { "grad_norm": 0.233891099691391, "learning_rate": 2.054522942378198e-05, "loss": 0.028, "step": 71550 }, { "grad_norm": 0.48466262221336365, "learning_rate": 2.053186994875469e-05, "loss": 0.0353, "step": 71560 }, { "grad_norm": 0.42399969696998596, "learning_rate": 2.0518513696317775e-05, "loss": 0.0414, "step": 71570 }, { "grad_norm": 0.39720386266708374, "learning_rate": 2.050516066793188e-05, "loss": 0.0325, "step": 71580 }, { "grad_norm": 0.29518815875053406, "learning_rate": 2.0491810865057276e-05, "loss": 0.0309, "step": 71590 }, { "grad_norm": 0.536261796951294, "learning_rate": 2.0478464289153882e-05, "loss": 0.0362, "step": 71600 }, { "grad_norm": 0.6239326000213623, "learning_rate": 2.0465120941681232e-05, "loss": 0.0333, "step": 71610 }, { "grad_norm": 0.35185062885284424, "learning_rate": 2.0451780824098566e-05, "loss": 0.0319, "step": 71620 }, { "grad_norm": 0.3147449791431427, "learning_rate": 2.043844393786471e-05, "loss": 0.0321, "step": 71630 }, { "grad_norm": 1.2257721424102783, "learning_rate": 2.0425110284438193e-05, "loss": 0.0342, "step": 71640 }, { "grad_norm": 0.31187543272972107, "learning_rate": 2.041177986527712e-05, "loss": 0.0328, "step": 71650 }, { "grad_norm": 0.7742713689804077, "learning_rate": 2.0398452681839347e-05, "loss": 0.0292, "step": 71660 }, { "grad_norm": 0.4043390452861786, "learning_rate": 2.0385128735582272e-05, "loss": 0.0432, "step": 71670 }, { "grad_norm": 0.3502851128578186, "learning_rate": 2.0371808027963003e-05, "loss": 0.0318, "step": 71680 }, { "grad_norm": 0.39404311776161194, "learning_rate": 2.035849056043825e-05, "loss": 0.0363, "step": 71690 }, { "grad_norm": 0.6755574941635132, "learning_rate": 2.03451763344644e-05, "loss": 0.0341, "step": 71700 }, { "grad_norm": 0.6138083338737488, "learning_rate": 2.0331865351497482e-05, "loss": 0.0381, "step": 71710 }, { "grad_norm": 0.588363766670227, "learning_rate": 2.0318557612993185e-05, "loss": 0.0368, "step": 71720 }, { "grad_norm": 0.38869035243988037, "learning_rate": 2.0305253120406782e-05, "loss": 0.0308, "step": 71730 }, { "grad_norm": 0.3781786859035492, "learning_rate": 2.029195187519325e-05, "loss": 0.0289, "step": 71740 }, { "grad_norm": 0.36781230568885803, "learning_rate": 2.0278653878807217e-05, "loss": 0.0299, "step": 71750 }, { "grad_norm": 0.28379759192466736, "learning_rate": 2.0265359132702892e-05, "loss": 0.0407, "step": 71760 }, { "grad_norm": 0.4466920495033264, "learning_rate": 2.0252067638334182e-05, "loss": 0.0292, "step": 71770 }, { "grad_norm": 0.4684968888759613, "learning_rate": 2.023877939715464e-05, "loss": 0.0452, "step": 71780 }, { "grad_norm": 0.4782375395298004, "learning_rate": 2.0225494410617448e-05, "loss": 0.0391, "step": 71790 }, { "grad_norm": 0.28568440675735474, "learning_rate": 2.02122126801754e-05, "loss": 0.0389, "step": 71800 }, { "grad_norm": 0.34631332755088806, "learning_rate": 2.0198934207281018e-05, "loss": 0.0405, "step": 71810 }, { "grad_norm": 0.5382400751113892, "learning_rate": 2.0185658993386357e-05, "loss": 0.034, "step": 71820 }, { "grad_norm": 0.513760507106781, "learning_rate": 2.0172387039943226e-05, "loss": 0.0361, "step": 71830 }, { "grad_norm": 0.2231728434562683, "learning_rate": 2.0159118348402968e-05, "loss": 0.038, "step": 71840 }, { "grad_norm": 0.30809450149536133, "learning_rate": 2.0145852920216697e-05, "loss": 0.0318, "step": 71850 }, { "grad_norm": 0.5665779113769531, "learning_rate": 2.0132590756835045e-05, "loss": 0.0339, "step": 71860 }, { "grad_norm": 0.3602038621902466, "learning_rate": 2.0119331859708386e-05, "loss": 0.0459, "step": 71870 }, { "grad_norm": 0.7365388870239258, "learning_rate": 2.0106076230286654e-05, "loss": 0.0329, "step": 71880 }, { "grad_norm": 0.32931193709373474, "learning_rate": 2.0092823870019488e-05, "loss": 0.0306, "step": 71890 }, { "grad_norm": 0.31125250458717346, "learning_rate": 2.0079574780356115e-05, "loss": 0.041, "step": 71900 }, { "grad_norm": 0.486192911863327, "learning_rate": 2.0066328962745486e-05, "loss": 0.0358, "step": 71910 }, { "grad_norm": 0.29353368282318115, "learning_rate": 2.00530864186361e-05, "loss": 0.0275, "step": 71920 }, { "grad_norm": 0.4356212913990021, "learning_rate": 2.003984714947618e-05, "loss": 0.0454, "step": 71930 }, { "grad_norm": 0.3521300256252289, "learning_rate": 2.0026611156713515e-05, "loss": 0.0405, "step": 71940 }, { "grad_norm": 0.38481926918029785, "learning_rate": 2.0013378441795584e-05, "loss": 0.0348, "step": 71950 }, { "grad_norm": 0.3905644416809082, "learning_rate": 2.0000149006169506e-05, "loss": 0.0285, "step": 71960 }, { "grad_norm": 0.4547870457172394, "learning_rate": 1.9986922851282048e-05, "loss": 0.035, "step": 71970 }, { "grad_norm": 0.35336244106292725, "learning_rate": 1.997369997857956e-05, "loss": 0.0362, "step": 71980 }, { "grad_norm": 0.49213907122612, "learning_rate": 1.9960480389508103e-05, "loss": 0.0332, "step": 71990 }, { "grad_norm": 0.4424944818019867, "learning_rate": 1.9947264085513363e-05, "loss": 0.0521, "step": 72000 }, { "grad_norm": 0.3831472396850586, "learning_rate": 1.993405106804062e-05, "loss": 0.0361, "step": 72010 }, { "grad_norm": 0.614057183265686, "learning_rate": 1.9920841338534867e-05, "loss": 0.0429, "step": 72020 }, { "grad_norm": 0.8867473006248474, "learning_rate": 1.9907634898440646e-05, "loss": 0.0293, "step": 72030 }, { "grad_norm": 0.3366120159626007, "learning_rate": 1.989443174920227e-05, "loss": 0.0309, "step": 72040 }, { "grad_norm": 0.46598559617996216, "learning_rate": 1.988123189226355e-05, "loss": 0.0455, "step": 72050 }, { "grad_norm": 0.8287079334259033, "learning_rate": 1.986803532906805e-05, "loss": 0.0373, "step": 72060 }, { "grad_norm": 0.5262616276741028, "learning_rate": 1.985484206105889e-05, "loss": 0.0318, "step": 72070 }, { "grad_norm": 0.600207507610321, "learning_rate": 1.9841652089678887e-05, "loss": 0.0311, "step": 72080 }, { "grad_norm": 0.2944346070289612, "learning_rate": 1.9828465416370435e-05, "loss": 0.0467, "step": 72090 }, { "grad_norm": 0.5383476614952087, "learning_rate": 1.9815282042575685e-05, "loss": 0.0341, "step": 72100 }, { "grad_norm": 0.24414347112178802, "learning_rate": 1.9802101969736285e-05, "loss": 0.0434, "step": 72110 }, { "grad_norm": 0.37440672516822815, "learning_rate": 1.9788925199293633e-05, "loss": 0.036, "step": 72120 }, { "grad_norm": 0.31787002086639404, "learning_rate": 1.977575173268868e-05, "loss": 0.0305, "step": 72130 }, { "grad_norm": 0.5678937435150146, "learning_rate": 1.9762581571362094e-05, "loss": 0.0404, "step": 72140 }, { "grad_norm": 0.47625309228897095, "learning_rate": 1.9749414716754112e-05, "loss": 0.0377, "step": 72150 }, { "grad_norm": 0.2723256051540375, "learning_rate": 1.973625117030465e-05, "loss": 0.0327, "step": 72160 }, { "grad_norm": 0.47877031564712524, "learning_rate": 1.972309093345326e-05, "loss": 0.0304, "step": 72170 }, { "grad_norm": 0.46901971101760864, "learning_rate": 1.970993400763915e-05, "loss": 0.0377, "step": 72180 }, { "grad_norm": 0.9029970169067383, "learning_rate": 1.969678039430109e-05, "loss": 0.0439, "step": 72190 }, { "grad_norm": 0.504963755607605, "learning_rate": 1.9683630094877564e-05, "loss": 0.029, "step": 72200 }, { "grad_norm": 0.3443323075771332, "learning_rate": 1.9670483110806696e-05, "loss": 0.0354, "step": 72210 }, { "grad_norm": 0.47534921765327454, "learning_rate": 1.9657339443526172e-05, "loss": 0.0378, "step": 72220 }, { "grad_norm": 0.21765993535518646, "learning_rate": 1.964419909447339e-05, "loss": 0.0381, "step": 72230 }, { "grad_norm": 0.493293821811676, "learning_rate": 1.963106206508536e-05, "loss": 0.0359, "step": 72240 }, { "grad_norm": 0.3516150712966919, "learning_rate": 1.9617928356798738e-05, "loss": 0.0328, "step": 72250 }, { "grad_norm": 0.3414579927921295, "learning_rate": 1.960479797104977e-05, "loss": 0.0334, "step": 72260 }, { "grad_norm": 0.6581740379333496, "learning_rate": 1.9591670909274418e-05, "loss": 0.038, "step": 72270 }, { "grad_norm": 0.25356659293174744, "learning_rate": 1.9578547172908184e-05, "loss": 0.0344, "step": 72280 }, { "grad_norm": 0.7760594487190247, "learning_rate": 1.9565426763386325e-05, "loss": 0.0332, "step": 72290 }, { "grad_norm": 0.3597918748855591, "learning_rate": 1.9552309682143622e-05, "loss": 0.0363, "step": 72300 }, { "grad_norm": 0.25820043683052063, "learning_rate": 1.9539195930614574e-05, "loss": 0.0344, "step": 72310 }, { "grad_norm": 0.7014407515525818, "learning_rate": 1.952608551023324e-05, "loss": 0.0396, "step": 72320 }, { "grad_norm": 0.7975139021873474, "learning_rate": 1.9512978422433397e-05, "loss": 0.0344, "step": 72330 }, { "grad_norm": 0.2896582782268524, "learning_rate": 1.9499874668648376e-05, "loss": 0.0271, "step": 72340 }, { "grad_norm": 0.4231259822845459, "learning_rate": 1.9486774250311213e-05, "loss": 0.0292, "step": 72350 }, { "grad_norm": 0.6621841192245483, "learning_rate": 1.9473677168854537e-05, "loss": 0.0474, "step": 72360 }, { "grad_norm": 0.38701432943344116, "learning_rate": 1.946058342571065e-05, "loss": 0.0401, "step": 72370 }, { "grad_norm": 0.32566025853157043, "learning_rate": 1.9447493022311424e-05, "loss": 0.0332, "step": 72380 }, { "grad_norm": 0.6979164481163025, "learning_rate": 1.9434405960088442e-05, "loss": 0.0317, "step": 72390 }, { "grad_norm": 0.35883983969688416, "learning_rate": 1.9421322240472857e-05, "loss": 0.0271, "step": 72400 }, { "grad_norm": 0.8120486736297607, "learning_rate": 1.9408241864895494e-05, "loss": 0.0327, "step": 72410 }, { "grad_norm": 0.5213786959648132, "learning_rate": 1.939516483478681e-05, "loss": 0.033, "step": 72420 }, { "grad_norm": 1.1591707468032837, "learning_rate": 1.9382091151576902e-05, "loss": 0.036, "step": 72430 }, { "grad_norm": 0.3768864572048187, "learning_rate": 1.9369020816695454e-05, "loss": 0.0273, "step": 72440 }, { "grad_norm": 0.28974390029907227, "learning_rate": 1.935595383157184e-05, "loss": 0.0304, "step": 72450 }, { "grad_norm": 0.4711373448371887, "learning_rate": 1.9342890197635057e-05, "loss": 0.0271, "step": 72460 }, { "grad_norm": 0.36069780588150024, "learning_rate": 1.9329829916313685e-05, "loss": 0.0331, "step": 72470 }, { "grad_norm": 0.34751495718955994, "learning_rate": 1.9316772989036013e-05, "loss": 0.0382, "step": 72480 }, { "grad_norm": 0.3945673704147339, "learning_rate": 1.930371941722991e-05, "loss": 0.0353, "step": 72490 }, { "grad_norm": 0.3896017074584961, "learning_rate": 1.929066920232292e-05, "loss": 0.0305, "step": 72500 }, { "grad_norm": 0.5003896355628967, "learning_rate": 1.9277622345742152e-05, "loss": 0.0358, "step": 72510 }, { "grad_norm": 0.3390439450740814, "learning_rate": 1.926457884891444e-05, "loss": 0.0296, "step": 72520 }, { "grad_norm": 0.6469379663467407, "learning_rate": 1.9251538713266147e-05, "loss": 0.0368, "step": 72530 }, { "grad_norm": 0.43732622265815735, "learning_rate": 1.9238501940223358e-05, "loss": 0.0316, "step": 72540 }, { "grad_norm": 1.0566165447235107, "learning_rate": 1.922546853121175e-05, "loss": 0.041, "step": 72550 }, { "grad_norm": 0.40772774815559387, "learning_rate": 1.9212438487656647e-05, "loss": 0.0358, "step": 72560 }, { "grad_norm": 0.28651368618011475, "learning_rate": 1.9199411810982965e-05, "loss": 0.032, "step": 72570 }, { "grad_norm": 0.260123610496521, "learning_rate": 1.918638850261532e-05, "loss": 0.0316, "step": 72580 }, { "grad_norm": 0.4598642587661743, "learning_rate": 1.9173368563977885e-05, "loss": 0.0347, "step": 72590 }, { "grad_norm": 0.28873148560523987, "learning_rate": 1.9160351996494513e-05, "loss": 0.0341, "step": 72600 }, { "grad_norm": 0.7064077258110046, "learning_rate": 1.9147338801588677e-05, "loss": 0.0346, "step": 72610 }, { "grad_norm": 0.5851560831069946, "learning_rate": 1.913432898068351e-05, "loss": 0.0376, "step": 72620 }, { "grad_norm": 0.3694738745689392, "learning_rate": 1.912132253520169e-05, "loss": 0.0426, "step": 72630 }, { "grad_norm": 1.0941829681396484, "learning_rate": 1.910831946656564e-05, "loss": 0.0417, "step": 72640 }, { "grad_norm": 0.40475037693977356, "learning_rate": 1.909531977619731e-05, "loss": 0.0483, "step": 72650 }, { "grad_norm": 0.40489858388900757, "learning_rate": 1.908232346551834e-05, "loss": 0.0277, "step": 72660 }, { "grad_norm": 0.4822537899017334, "learning_rate": 1.9069330535949993e-05, "loss": 0.0433, "step": 72670 }, { "grad_norm": 0.4823302626609802, "learning_rate": 1.9056340988913167e-05, "loss": 0.0246, "step": 72680 }, { "grad_norm": 0.46951034665107727, "learning_rate": 1.9043354825828357e-05, "loss": 0.0364, "step": 72690 }, { "grad_norm": 0.19209706783294678, "learning_rate": 1.9030372048115724e-05, "loss": 0.0317, "step": 72700 }, { "grad_norm": 0.5013521313667297, "learning_rate": 1.9017392657195056e-05, "loss": 0.0377, "step": 72710 }, { "grad_norm": 0.3160460889339447, "learning_rate": 1.9004416654485724e-05, "loss": 0.0363, "step": 72720 }, { "grad_norm": 0.38136565685272217, "learning_rate": 1.8991444041406785e-05, "loss": 0.0384, "step": 72730 }, { "grad_norm": 0.4659259617328644, "learning_rate": 1.8978474819376912e-05, "loss": 0.0419, "step": 72740 }, { "grad_norm": 0.35807695984840393, "learning_rate": 1.8965508989814406e-05, "loss": 0.0357, "step": 72750 }, { "grad_norm": 0.3838290274143219, "learning_rate": 1.8952546554137158e-05, "loss": 0.0353, "step": 72760 }, { "grad_norm": 0.38350939750671387, "learning_rate": 1.8939587513762757e-05, "loss": 0.0264, "step": 72770 }, { "grad_norm": 0.3300534188747406, "learning_rate": 1.892663187010835e-05, "loss": 0.0247, "step": 72780 }, { "grad_norm": 0.3413306474685669, "learning_rate": 1.891367962459077e-05, "loss": 0.0325, "step": 72790 }, { "grad_norm": 0.7149813771247864, "learning_rate": 1.890073077862644e-05, "loss": 0.0441, "step": 72800 }, { "grad_norm": 0.6892805695533752, "learning_rate": 1.888778533363146e-05, "loss": 0.0278, "step": 72810 }, { "grad_norm": 0.6527498960494995, "learning_rate": 1.8874843291021478e-05, "loss": 0.0345, "step": 72820 }, { "grad_norm": 0.39501333236694336, "learning_rate": 1.8861904652211853e-05, "loss": 0.0319, "step": 72830 }, { "grad_norm": 0.45528703927993774, "learning_rate": 1.8848969418617507e-05, "loss": 0.0321, "step": 72840 }, { "grad_norm": 0.4246956408023834, "learning_rate": 1.8836037591653044e-05, "loss": 0.0371, "step": 72850 }, { "grad_norm": 0.7974666953086853, "learning_rate": 1.882310917273262e-05, "loss": 0.0392, "step": 72860 }, { "grad_norm": 0.453708678483963, "learning_rate": 1.881018416327014e-05, "loss": 0.0371, "step": 72870 }, { "grad_norm": 0.4242897927761078, "learning_rate": 1.8797262564679014e-05, "loss": 0.0361, "step": 72880 }, { "grad_norm": 0.503818154335022, "learning_rate": 1.8784344378372355e-05, "loss": 0.0346, "step": 72890 }, { "grad_norm": 0.3219132423400879, "learning_rate": 1.8771429605762843e-05, "loss": 0.0322, "step": 72900 }, { "grad_norm": 0.378891259431839, "learning_rate": 1.8758518248262835e-05, "loss": 0.0368, "step": 72910 }, { "grad_norm": 0.3760147988796234, "learning_rate": 1.8745610307284305e-05, "loss": 0.0448, "step": 72920 }, { "grad_norm": 0.37158092856407166, "learning_rate": 1.8732705784238862e-05, "loss": 0.0336, "step": 72930 }, { "grad_norm": 0.4765073359012604, "learning_rate": 1.8719804680537682e-05, "loss": 0.0377, "step": 72940 }, { "grad_norm": 0.8139141201972961, "learning_rate": 1.870690699759164e-05, "loss": 0.0315, "step": 72950 }, { "grad_norm": 0.4185659885406494, "learning_rate": 1.8694012736811213e-05, "loss": 0.0393, "step": 72960 }, { "grad_norm": 0.49207398295402527, "learning_rate": 1.8681121899606463e-05, "loss": 0.0279, "step": 72970 }, { "grad_norm": 0.323960542678833, "learning_rate": 1.8668234487387144e-05, "loss": 0.0365, "step": 72980 }, { "grad_norm": 0.3488905131816864, "learning_rate": 1.8655350501562592e-05, "loss": 0.036, "step": 72990 }, { "grad_norm": 0.537280797958374, "learning_rate": 1.8642469943541796e-05, "loss": 0.042, "step": 73000 }, { "grad_norm": 0.3222426772117615, "learning_rate": 1.8629592814733328e-05, "loss": 0.0309, "step": 73010 }, { "grad_norm": 0.9320775866508484, "learning_rate": 1.8616719116545438e-05, "loss": 0.0468, "step": 73020 }, { "grad_norm": 0.2811151146888733, "learning_rate": 1.8603848850385952e-05, "loss": 0.0289, "step": 73030 }, { "grad_norm": 0.6599915027618408, "learning_rate": 1.859098201766236e-05, "loss": 0.0338, "step": 73040 }, { "grad_norm": 0.334415078163147, "learning_rate": 1.8578118619781726e-05, "loss": 0.0279, "step": 73050 }, { "grad_norm": 0.19771699607372284, "learning_rate": 1.856525865815083e-05, "loss": 0.0235, "step": 73060 }, { "grad_norm": 0.3422929048538208, "learning_rate": 1.8552402134175966e-05, "loss": 0.0286, "step": 73070 }, { "grad_norm": 0.4742758870124817, "learning_rate": 1.8539549049263143e-05, "loss": 0.034, "step": 73080 }, { "grad_norm": 0.292050838470459, "learning_rate": 1.8526699404817917e-05, "loss": 0.0361, "step": 73090 }, { "grad_norm": 0.3647323250770569, "learning_rate": 1.8513853202245546e-05, "loss": 0.032, "step": 73100 }, { "grad_norm": 0.4705539345741272, "learning_rate": 1.850101044295081e-05, "loss": 0.0348, "step": 73110 }, { "grad_norm": 0.44013580679893494, "learning_rate": 1.848817112833825e-05, "loss": 0.0224, "step": 73120 }, { "grad_norm": 0.5439109206199646, "learning_rate": 1.8475335259811905e-05, "loss": 0.0291, "step": 73130 }, { "grad_norm": 0.6444814205169678, "learning_rate": 1.8462502838775508e-05, "loss": 0.0403, "step": 73140 }, { "grad_norm": 0.3485417068004608, "learning_rate": 1.8449673866632378e-05, "loss": 0.0395, "step": 73150 }, { "grad_norm": 0.29978540539741516, "learning_rate": 1.8436848344785473e-05, "loss": 0.0404, "step": 73160 }, { "grad_norm": 0.38845160603523254, "learning_rate": 1.8424026274637378e-05, "loss": 0.0311, "step": 73170 }, { "grad_norm": 0.5260446071624756, "learning_rate": 1.8411207657590312e-05, "loss": 0.0429, "step": 73180 }, { "grad_norm": 1.3895649909973145, "learning_rate": 1.839839249504607e-05, "loss": 0.0401, "step": 73190 }, { "grad_norm": 0.3677808344364166, "learning_rate": 1.8385580788406108e-05, "loss": 0.0368, "step": 73200 }, { "grad_norm": 0.5542430281639099, "learning_rate": 1.837277253907152e-05, "loss": 0.0358, "step": 73210 }, { "grad_norm": 0.3941532373428345, "learning_rate": 1.8359967748442957e-05, "loss": 0.0341, "step": 73220 }, { "grad_norm": 0.3973962664604187, "learning_rate": 1.8347166417920774e-05, "loss": 0.0314, "step": 73230 }, { "grad_norm": 0.751543402671814, "learning_rate": 1.833436854890484e-05, "loss": 0.0373, "step": 73240 }, { "grad_norm": 0.4905005097389221, "learning_rate": 1.8321574142794796e-05, "loss": 0.031, "step": 73250 }, { "grad_norm": 0.6701771020889282, "learning_rate": 1.8308783200989748e-05, "loss": 0.0277, "step": 73260 }, { "grad_norm": 0.589974582195282, "learning_rate": 1.829599572488855e-05, "loss": 0.0328, "step": 73270 }, { "grad_norm": 0.20103995501995087, "learning_rate": 1.828321171588957e-05, "loss": 0.0346, "step": 73280 }, { "grad_norm": 0.7462528944015503, "learning_rate": 1.827043117539089e-05, "loss": 0.0409, "step": 73290 }, { "grad_norm": 0.17440791428089142, "learning_rate": 1.8257654104790116e-05, "loss": 0.024, "step": 73300 }, { "grad_norm": 0.7571423649787903, "learning_rate": 1.82448805054846e-05, "loss": 0.0344, "step": 73310 }, { "grad_norm": 0.4332348704338074, "learning_rate": 1.823211037887119e-05, "loss": 0.0442, "step": 73320 }, { "grad_norm": 1.9032572507858276, "learning_rate": 1.821934372634645e-05, "loss": 0.0418, "step": 73330 }, { "grad_norm": 0.3950361907482147, "learning_rate": 1.8206580549306472e-05, "loss": 0.0275, "step": 73340 }, { "grad_norm": 0.5134320855140686, "learning_rate": 1.819382084914707e-05, "loss": 0.0332, "step": 73350 }, { "grad_norm": 0.36903008818626404, "learning_rate": 1.8181064627263577e-05, "loss": 0.0256, "step": 73360 }, { "grad_norm": 0.4373856484889984, "learning_rate": 1.8168311885051022e-05, "loss": 0.0345, "step": 73370 }, { "grad_norm": 0.4652567207813263, "learning_rate": 1.815556262390402e-05, "loss": 0.0305, "step": 73380 }, { "grad_norm": 0.4082932770252228, "learning_rate": 1.8142816845216826e-05, "loss": 0.035, "step": 73390 }, { "grad_norm": 0.3985322415828705, "learning_rate": 1.8130074550383276e-05, "loss": 0.0316, "step": 73400 }, { "grad_norm": 0.8476848602294922, "learning_rate": 1.8117335740796848e-05, "loss": 0.0273, "step": 73410 }, { "grad_norm": 0.5796045064926147, "learning_rate": 1.810460041785067e-05, "loss": 0.0326, "step": 73420 }, { "grad_norm": 1.3196808099746704, "learning_rate": 1.809186858293742e-05, "loss": 0.0467, "step": 73430 }, { "grad_norm": 0.4956863820552826, "learning_rate": 1.8079140237449448e-05, "loss": 0.0245, "step": 73440 }, { "grad_norm": 0.539194643497467, "learning_rate": 1.8066415382778707e-05, "loss": 0.0354, "step": 73450 }, { "grad_norm": 0.9699665307998657, "learning_rate": 1.8053694020316786e-05, "loss": 0.0355, "step": 73460 }, { "grad_norm": 0.3704849183559418, "learning_rate": 1.8040976151454837e-05, "loss": 0.0353, "step": 73470 }, { "grad_norm": 0.5647715926170349, "learning_rate": 1.8028261777583704e-05, "loss": 0.0444, "step": 73480 }, { "grad_norm": 0.39748188853263855, "learning_rate": 1.8015550900093758e-05, "loss": 0.0365, "step": 73490 }, { "grad_norm": 0.5000801682472229, "learning_rate": 1.800284352037512e-05, "loss": 0.0332, "step": 73500 }, { "grad_norm": 0.29768726229667664, "learning_rate": 1.799013963981738e-05, "loss": 0.0396, "step": 73510 }, { "grad_norm": 0.4463832974433899, "learning_rate": 1.797743925980987e-05, "loss": 0.0305, "step": 73520 }, { "grad_norm": 0.53299480676651, "learning_rate": 1.7964742381741435e-05, "loss": 0.038, "step": 73530 }, { "grad_norm": 0.5181142091751099, "learning_rate": 1.7952049007000627e-05, "loss": 0.0432, "step": 73540 }, { "grad_norm": 0.5869644284248352, "learning_rate": 1.7939359136975535e-05, "loss": 0.0295, "step": 73550 }, { "grad_norm": 0.45942696928977966, "learning_rate": 1.7926672773053932e-05, "loss": 0.0337, "step": 73560 }, { "grad_norm": 0.41214412450790405, "learning_rate": 1.791398991662317e-05, "loss": 0.0339, "step": 73570 }, { "grad_norm": 0.2743268609046936, "learning_rate": 1.7901310569070244e-05, "loss": 0.0337, "step": 73580 }, { "grad_norm": 0.5498406291007996, "learning_rate": 1.7888634731781724e-05, "loss": 0.0301, "step": 73590 }, { "grad_norm": 1.6291885375976562, "learning_rate": 1.7875962406143844e-05, "loss": 0.0291, "step": 73600 }, { "grad_norm": 0.3900437653064728, "learning_rate": 1.7863293593542403e-05, "loss": 0.0411, "step": 73610 }, { "grad_norm": 0.4358263611793518, "learning_rate": 1.7850628295362853e-05, "loss": 0.0272, "step": 73620 }, { "grad_norm": 0.3699612021446228, "learning_rate": 1.7837966512990263e-05, "loss": 0.0365, "step": 73630 }, { "grad_norm": 0.35090672969818115, "learning_rate": 1.782530824780931e-05, "loss": 0.0335, "step": 73640 }, { "grad_norm": 0.3447674810886383, "learning_rate": 1.7812653501204263e-05, "loss": 0.0324, "step": 73650 }, { "grad_norm": 0.4064137637615204, "learning_rate": 1.780000227455903e-05, "loss": 0.036, "step": 73660 }, { "grad_norm": 0.6501221060752869, "learning_rate": 1.778735456925716e-05, "loss": 0.04, "step": 73670 }, { "grad_norm": 0.26303496956825256, "learning_rate": 1.7774710386681737e-05, "loss": 0.0264, "step": 73680 }, { "grad_norm": 0.3540661036968231, "learning_rate": 1.7762069728215545e-05, "loss": 0.0404, "step": 73690 }, { "grad_norm": 0.38221627473831177, "learning_rate": 1.7749432595240932e-05, "loss": 0.0364, "step": 73700 }, { "grad_norm": 0.45959874987602234, "learning_rate": 1.773679898913991e-05, "loss": 0.0405, "step": 73710 }, { "grad_norm": 0.5770509243011475, "learning_rate": 1.7724168911294015e-05, "loss": 0.0419, "step": 73720 }, { "grad_norm": 0.5458205938339233, "learning_rate": 1.7711542363084504e-05, "loss": 0.0309, "step": 73730 }, { "grad_norm": 0.3473704159259796, "learning_rate": 1.7698919345892158e-05, "loss": 0.0258, "step": 73740 }, { "grad_norm": 0.3067312240600586, "learning_rate": 1.768629986109743e-05, "loss": 0.0378, "step": 73750 }, { "grad_norm": 0.5530743598937988, "learning_rate": 1.7673683910080357e-05, "loss": 0.0305, "step": 73760 }, { "grad_norm": 0.2718523442745209, "learning_rate": 1.7661071494220638e-05, "loss": 0.0352, "step": 73770 }, { "grad_norm": 0.25562065839767456, "learning_rate": 1.7648462614897498e-05, "loss": 0.0288, "step": 73780 }, { "grad_norm": 0.6455155611038208, "learning_rate": 1.7635857273489863e-05, "loss": 0.0283, "step": 73790 }, { "grad_norm": 0.31185176968574524, "learning_rate": 1.7623255471376198e-05, "loss": 0.0335, "step": 73800 }, { "grad_norm": 0.42986950278282166, "learning_rate": 1.7610657209934646e-05, "loss": 0.0441, "step": 73810 }, { "grad_norm": 0.2716847360134125, "learning_rate": 1.7598062490542915e-05, "loss": 0.0361, "step": 73820 }, { "grad_norm": 0.35827556252479553, "learning_rate": 1.7585471314578377e-05, "loss": 0.0313, "step": 73830 }, { "grad_norm": 0.41869932413101196, "learning_rate": 1.7572883683417945e-05, "loss": 0.0419, "step": 73840 }, { "grad_norm": 0.4570982754230499, "learning_rate": 1.7560299598438212e-05, "loss": 0.0422, "step": 73850 }, { "grad_norm": 0.32383039593696594, "learning_rate": 1.7547719061015334e-05, "loss": 0.0308, "step": 73860 }, { "grad_norm": 0.2621538043022156, "learning_rate": 1.75351420725251e-05, "loss": 0.0266, "step": 73870 }, { "grad_norm": 0.3404111862182617, "learning_rate": 1.7522568634342923e-05, "loss": 0.0302, "step": 73880 }, { "grad_norm": 0.9330703616142273, "learning_rate": 1.7509998747843825e-05, "loss": 0.0264, "step": 73890 }, { "grad_norm": 0.35354456305503845, "learning_rate": 1.7497432414402403e-05, "loss": 0.0299, "step": 73900 }, { "grad_norm": 0.6943604350090027, "learning_rate": 1.7484869635392903e-05, "loss": 0.0282, "step": 73910 }, { "grad_norm": 0.3977605402469635, "learning_rate": 1.747231041218919e-05, "loss": 0.0272, "step": 73920 }, { "grad_norm": 0.418277382850647, "learning_rate": 1.7459754746164692e-05, "loss": 0.0326, "step": 73930 }, { "grad_norm": 1.0691490173339844, "learning_rate": 1.744720263869249e-05, "loss": 0.0392, "step": 73940 }, { "grad_norm": 0.6658172011375427, "learning_rate": 1.7434654091145263e-05, "loss": 0.0315, "step": 73950 }, { "grad_norm": 0.4257597029209137, "learning_rate": 1.742210910489532e-05, "loss": 0.0272, "step": 73960 }, { "grad_norm": 0.24850569665431976, "learning_rate": 1.7409567681314532e-05, "loss": 0.0311, "step": 73970 }, { "grad_norm": 0.2618447542190552, "learning_rate": 1.7397029821774434e-05, "loss": 0.0365, "step": 73980 }, { "grad_norm": 0.429889053106308, "learning_rate": 1.7384495527646126e-05, "loss": 0.0336, "step": 73990 }, { "grad_norm": 0.44261276721954346, "learning_rate": 1.737196480030035e-05, "loss": 0.0357, "step": 74000 }, { "grad_norm": 0.370377779006958, "learning_rate": 1.7359437641107447e-05, "loss": 0.0342, "step": 74010 }, { "grad_norm": 0.4356900751590729, "learning_rate": 1.7346914051437387e-05, "loss": 0.0284, "step": 74020 }, { "grad_norm": 0.42039954662323, "learning_rate": 1.733439403265969e-05, "loss": 0.0283, "step": 74030 }, { "grad_norm": 0.32452380657196045, "learning_rate": 1.732187758614357e-05, "loss": 0.042, "step": 74040 }, { "grad_norm": 0.43690040707588196, "learning_rate": 1.730936471325777e-05, "loss": 0.0337, "step": 74050 }, { "grad_norm": 0.4933975040912628, "learning_rate": 1.7296855415370718e-05, "loss": 0.0287, "step": 74060 }, { "grad_norm": 0.5671006441116333, "learning_rate": 1.728434969385035e-05, "loss": 0.0292, "step": 74070 }, { "grad_norm": 1.191896915435791, "learning_rate": 1.7271847550064346e-05, "loss": 0.0333, "step": 74080 }, { "grad_norm": 0.6014240384101868, "learning_rate": 1.725934898537987e-05, "loss": 0.0359, "step": 74090 }, { "grad_norm": 0.5775192975997925, "learning_rate": 1.7246854001163783e-05, "loss": 0.0258, "step": 74100 }, { "grad_norm": 0.4807402193546295, "learning_rate": 1.7234362598782476e-05, "loss": 0.0323, "step": 74110 }, { "grad_norm": 1.0185505151748657, "learning_rate": 1.722187477960202e-05, "loss": 0.0385, "step": 74120 }, { "grad_norm": 1.4254150390625, "learning_rate": 1.7209390544988057e-05, "loss": 0.0411, "step": 74130 }, { "grad_norm": 0.3809785842895508, "learning_rate": 1.7196909896305856e-05, "loss": 0.0434, "step": 74140 }, { "grad_norm": 0.35986262559890747, "learning_rate": 1.718443283492026e-05, "loss": 0.0337, "step": 74150 }, { "grad_norm": 0.4086296856403351, "learning_rate": 1.7171959362195746e-05, "loss": 0.0372, "step": 74160 }, { "grad_norm": 0.26084214448928833, "learning_rate": 1.715948947949642e-05, "loss": 0.0304, "step": 74170 }, { "grad_norm": 0.2998597323894501, "learning_rate": 1.714702318818593e-05, "loss": 0.0341, "step": 74180 }, { "grad_norm": 0.2627529799938202, "learning_rate": 1.7134560489627592e-05, "loss": 0.0282, "step": 74190 }, { "grad_norm": 0.32968318462371826, "learning_rate": 1.712210138518431e-05, "loss": 0.0264, "step": 74200 }, { "grad_norm": 0.40194427967071533, "learning_rate": 1.71096458762186e-05, "loss": 0.0403, "step": 74210 }, { "grad_norm": 0.525042712688446, "learning_rate": 1.7097193964092552e-05, "loss": 0.045, "step": 74220 }, { "grad_norm": 0.5178354382514954, "learning_rate": 1.708474565016792e-05, "loss": 0.0352, "step": 74230 }, { "grad_norm": 0.4006744623184204, "learning_rate": 1.7072300935806e-05, "loss": 0.0363, "step": 74240 }, { "grad_norm": 0.47601598501205444, "learning_rate": 1.705985982236776e-05, "loss": 0.0342, "step": 74250 }, { "grad_norm": 0.5947557091712952, "learning_rate": 1.7047422311213697e-05, "loss": 0.0274, "step": 74260 }, { "grad_norm": 0.5139418840408325, "learning_rate": 1.703498840370401e-05, "loss": 0.038, "step": 74270 }, { "grad_norm": 0.4939559996128082, "learning_rate": 1.702255810119842e-05, "loss": 0.0288, "step": 74280 }, { "grad_norm": 0.49152451753616333, "learning_rate": 1.7010131405056305e-05, "loss": 0.03, "step": 74290 }, { "grad_norm": 0.3966412842273712, "learning_rate": 1.6997708316636613e-05, "loss": 0.0336, "step": 74300 }, { "grad_norm": 0.36611583828926086, "learning_rate": 1.6985288837297932e-05, "loss": 0.0336, "step": 74310 }, { "grad_norm": 0.6203241944313049, "learning_rate": 1.6972872968398394e-05, "loss": 0.0354, "step": 74320 }, { "grad_norm": 0.4243573248386383, "learning_rate": 1.6960460711295845e-05, "loss": 0.0337, "step": 74330 }, { "grad_norm": 0.34721502661705017, "learning_rate": 1.6948052067347623e-05, "loss": 0.0268, "step": 74340 }, { "grad_norm": 0.7130040526390076, "learning_rate": 1.693564703791075e-05, "loss": 0.0317, "step": 74350 }, { "grad_norm": 0.25661700963974, "learning_rate": 1.6923245624341793e-05, "loss": 0.0373, "step": 74360 }, { "grad_norm": 0.3261047303676605, "learning_rate": 1.6910847827996962e-05, "loss": 0.0275, "step": 74370 }, { "grad_norm": 0.38495421409606934, "learning_rate": 1.689845365023206e-05, "loss": 0.0267, "step": 74380 }, { "grad_norm": 0.18481089174747467, "learning_rate": 1.6886063092402505e-05, "loss": 0.0311, "step": 74390 }, { "grad_norm": 0.5357906818389893, "learning_rate": 1.6873676155863317e-05, "loss": 0.0412, "step": 74400 }, { "grad_norm": 0.2502936124801636, "learning_rate": 1.6861292841969085e-05, "loss": 0.0291, "step": 74410 }, { "grad_norm": 0.553282618522644, "learning_rate": 1.684891315207406e-05, "loss": 0.032, "step": 74420 }, { "grad_norm": 0.408047080039978, "learning_rate": 1.6836537087532034e-05, "loss": 0.0343, "step": 74430 }, { "grad_norm": 0.31239843368530273, "learning_rate": 1.6824164649696465e-05, "loss": 0.0349, "step": 74440 }, { "grad_norm": 0.4770658314228058, "learning_rate": 1.681179583992034e-05, "loss": 0.0271, "step": 74450 }, { "grad_norm": 0.4448018968105316, "learning_rate": 1.679943065955636e-05, "loss": 0.027, "step": 74460 }, { "grad_norm": 1.4702783823013306, "learning_rate": 1.678706910995671e-05, "loss": 0.0293, "step": 74470 }, { "grad_norm": 0.6060961484909058, "learning_rate": 1.6774711192473262e-05, "loss": 0.0329, "step": 74480 }, { "grad_norm": 0.3633757531642914, "learning_rate": 1.676235690845743e-05, "loss": 0.0258, "step": 74490 }, { "grad_norm": 0.6655406951904297, "learning_rate": 1.6750006259260285e-05, "loss": 0.0362, "step": 74500 }, { "grad_norm": 0.5356758236885071, "learning_rate": 1.673765924623244e-05, "loss": 0.0402, "step": 74510 }, { "grad_norm": 0.4003126323223114, "learning_rate": 1.67253158707242e-05, "loss": 0.0251, "step": 74520 }, { "grad_norm": 0.510413408279419, "learning_rate": 1.671297613408537e-05, "loss": 0.0316, "step": 74530 }, { "grad_norm": 0.806593656539917, "learning_rate": 1.6700640037665444e-05, "loss": 0.0366, "step": 74540 }, { "grad_norm": 0.40611687302589417, "learning_rate": 1.668830758281344e-05, "loss": 0.0338, "step": 74550 }, { "grad_norm": 0.3738502562046051, "learning_rate": 1.6675978770878052e-05, "loss": 0.0334, "step": 74560 }, { "grad_norm": 0.27968829870224, "learning_rate": 1.6663653603207506e-05, "loss": 0.028, "step": 74570 }, { "grad_norm": 0.5458336472511292, "learning_rate": 1.6651332081149685e-05, "loss": 0.0419, "step": 74580 }, { "grad_norm": 0.5939068794250488, "learning_rate": 1.6639014206052046e-05, "loss": 0.0295, "step": 74590 }, { "grad_norm": 0.3272991180419922, "learning_rate": 1.6626699979261657e-05, "loss": 0.0258, "step": 74600 }, { "grad_norm": 0.46365416049957275, "learning_rate": 1.6614389402125197e-05, "loss": 0.0429, "step": 74610 }, { "grad_norm": 0.34042298793792725, "learning_rate": 1.66020824759889e-05, "loss": 0.0268, "step": 74620 }, { "grad_norm": 0.4474911689758301, "learning_rate": 1.6589779202198675e-05, "loss": 0.0281, "step": 74630 }, { "grad_norm": 0.580708384513855, "learning_rate": 1.6577479582099924e-05, "loss": 0.038, "step": 74640 }, { "grad_norm": 0.6751823425292969, "learning_rate": 1.6565183617037788e-05, "loss": 0.04, "step": 74650 }, { "grad_norm": 0.45768582820892334, "learning_rate": 1.6552891308356894e-05, "loss": 0.0306, "step": 74660 }, { "grad_norm": 0.694661021232605, "learning_rate": 1.6540602657401526e-05, "loss": 0.0403, "step": 74670 }, { "grad_norm": 0.3661320209503174, "learning_rate": 1.6528317665515537e-05, "loss": 0.0432, "step": 74680 }, { "grad_norm": 0.8396532535552979, "learning_rate": 1.6516036334042416e-05, "loss": 0.0503, "step": 74690 }, { "grad_norm": 0.4266652762889862, "learning_rate": 1.650375866432519e-05, "loss": 0.0483, "step": 74700 }, { "grad_norm": 0.4020584225654602, "learning_rate": 1.649148465770659e-05, "loss": 0.0345, "step": 74710 }, { "grad_norm": 0.8800705075263977, "learning_rate": 1.647921431552883e-05, "loss": 0.0427, "step": 74720 }, { "grad_norm": 1.340224027633667, "learning_rate": 1.646694763913381e-05, "loss": 0.0423, "step": 74730 }, { "grad_norm": 0.5012605786323547, "learning_rate": 1.6454684629862966e-05, "loss": 0.0348, "step": 74740 }, { "grad_norm": 1.4241021871566772, "learning_rate": 1.6442425289057388e-05, "loss": 0.0317, "step": 74750 }, { "grad_norm": 0.3741244077682495, "learning_rate": 1.6430169618057718e-05, "loss": 0.0329, "step": 74760 }, { "grad_norm": 0.21351805329322815, "learning_rate": 1.641791761820422e-05, "loss": 0.0291, "step": 74770 }, { "grad_norm": 0.4832835793495178, "learning_rate": 1.640566929083676e-05, "loss": 0.0338, "step": 74780 }, { "grad_norm": 0.579054057598114, "learning_rate": 1.6393424637294823e-05, "loss": 0.0396, "step": 74790 }, { "grad_norm": 0.3436720371246338, "learning_rate": 1.638118365891742e-05, "loss": 0.0272, "step": 74800 }, { "grad_norm": 0.714670717716217, "learning_rate": 1.636894635704323e-05, "loss": 0.0403, "step": 74810 }, { "grad_norm": 0.3253423571586609, "learning_rate": 1.6356712733010517e-05, "loss": 0.0307, "step": 74820 }, { "grad_norm": 1.8727161884307861, "learning_rate": 1.6344482788157096e-05, "loss": 0.0288, "step": 74830 }, { "grad_norm": 0.3518538773059845, "learning_rate": 1.6332256523820444e-05, "loss": 0.0277, "step": 74840 }, { "grad_norm": 0.8520132303237915, "learning_rate": 1.6320033941337597e-05, "loss": 0.045, "step": 74850 }, { "grad_norm": 0.4680301547050476, "learning_rate": 1.6307815042045216e-05, "loss": 0.0315, "step": 74860 }, { "grad_norm": 0.49575960636138916, "learning_rate": 1.6295599827279513e-05, "loss": 0.0296, "step": 74870 }, { "grad_norm": 0.5241394639015198, "learning_rate": 1.6283388298376356e-05, "loss": 0.0332, "step": 74880 }, { "grad_norm": 0.34035542607307434, "learning_rate": 1.6271180456671125e-05, "loss": 0.0301, "step": 74890 }, { "grad_norm": 0.34240883588790894, "learning_rate": 1.625897630349893e-05, "loss": 0.0242, "step": 74900 }, { "grad_norm": 0.40871578454971313, "learning_rate": 1.6246775840194335e-05, "loss": 0.0309, "step": 74910 }, { "grad_norm": 0.2842159867286682, "learning_rate": 1.6234579068091604e-05, "loss": 0.03, "step": 74920 }, { "grad_norm": 0.5615611672401428, "learning_rate": 1.6222385988524524e-05, "loss": 0.0333, "step": 74930 }, { "grad_norm": 0.318488746881485, "learning_rate": 1.6210196602826544e-05, "loss": 0.0325, "step": 74940 }, { "grad_norm": 0.27972713112831116, "learning_rate": 1.6198010912330646e-05, "loss": 0.0337, "step": 74950 }, { "grad_norm": 0.41070422530174255, "learning_rate": 1.6185828918369456e-05, "loss": 0.0355, "step": 74960 }, { "grad_norm": 0.7455722689628601, "learning_rate": 1.617365062227517e-05, "loss": 0.0391, "step": 74970 }, { "grad_norm": 0.24167253077030182, "learning_rate": 1.616147602537962e-05, "loss": 0.0398, "step": 74980 }, { "grad_norm": 0.32473236322402954, "learning_rate": 1.614930512901415e-05, "loss": 0.0332, "step": 74990 }, { "grad_norm": 0.4556882381439209, "learning_rate": 1.6137137934509804e-05, "loss": 0.0255, "step": 75000 }, { "grad_norm": 0.3427698016166687, "learning_rate": 1.6124974443197123e-05, "loss": 0.0327, "step": 75010 }, { "grad_norm": 0.31058549880981445, "learning_rate": 1.6112814656406306e-05, "loss": 0.0362, "step": 75020 }, { "grad_norm": 0.36713671684265137, "learning_rate": 1.6100658575467125e-05, "loss": 0.0367, "step": 75030 }, { "grad_norm": 0.4658273160457611, "learning_rate": 1.6088506201708974e-05, "loss": 0.0274, "step": 75040 }, { "grad_norm": 0.6352308988571167, "learning_rate": 1.6076357536460785e-05, "loss": 0.0397, "step": 75050 }, { "grad_norm": 0.4689602255821228, "learning_rate": 1.606421258105113e-05, "loss": 0.0309, "step": 75060 }, { "grad_norm": 0.38248157501220703, "learning_rate": 1.6052071336808183e-05, "loss": 0.0367, "step": 75070 }, { "grad_norm": 0.30388471484184265, "learning_rate": 1.6039933805059658e-05, "loss": 0.038, "step": 75080 }, { "grad_norm": 0.38352587819099426, "learning_rate": 1.6027799987132915e-05, "loss": 0.0334, "step": 75090 }, { "grad_norm": 0.38366612792015076, "learning_rate": 1.6015669884354888e-05, "loss": 0.0449, "step": 75100 }, { "grad_norm": 0.49445968866348267, "learning_rate": 1.6003543498052125e-05, "loss": 0.0414, "step": 75110 }, { "grad_norm": 0.4669809639453888, "learning_rate": 1.5991420829550715e-05, "loss": 0.0328, "step": 75120 }, { "grad_norm": 1.6453746557235718, "learning_rate": 1.5979301880176405e-05, "loss": 0.0411, "step": 75130 }, { "grad_norm": 0.31088489294052124, "learning_rate": 1.5967186651254485e-05, "loss": 0.034, "step": 75140 }, { "grad_norm": 0.5172330737113953, "learning_rate": 1.5955075144109853e-05, "loss": 0.0306, "step": 75150 }, { "grad_norm": 0.32613420486450195, "learning_rate": 1.594296736006703e-05, "loss": 0.0313, "step": 75160 }, { "grad_norm": 0.24497026205062866, "learning_rate": 1.59308633004501e-05, "loss": 0.037, "step": 75170 }, { "grad_norm": 0.4711659252643585, "learning_rate": 1.5918762966582724e-05, "loss": 0.0386, "step": 75180 }, { "grad_norm": 0.42288124561309814, "learning_rate": 1.5906666359788203e-05, "loss": 0.0316, "step": 75190 }, { "grad_norm": 0.2898693382740021, "learning_rate": 1.589457348138938e-05, "loss": 0.0358, "step": 75200 }, { "grad_norm": 0.49322956800460815, "learning_rate": 1.588248433270872e-05, "loss": 0.0345, "step": 75210 }, { "grad_norm": 0.4872850179672241, "learning_rate": 1.587039891506828e-05, "loss": 0.032, "step": 75220 }, { "grad_norm": 0.27914512157440186, "learning_rate": 1.5858317229789715e-05, "loss": 0.0332, "step": 75230 }, { "grad_norm": 0.6022593975067139, "learning_rate": 1.584623927819423e-05, "loss": 0.0283, "step": 75240 }, { "grad_norm": 0.7672948241233826, "learning_rate": 1.583416506160269e-05, "loss": 0.0417, "step": 75250 }, { "grad_norm": 0.4028777480125427, "learning_rate": 1.5822094581335472e-05, "loss": 0.0414, "step": 75260 }, { "grad_norm": 0.3101207911968231, "learning_rate": 1.5810027838712604e-05, "loss": 0.0369, "step": 75270 }, { "grad_norm": 0.7471645474433899, "learning_rate": 1.5797964835053687e-05, "loss": 0.0318, "step": 75280 }, { "grad_norm": 0.34088724851608276, "learning_rate": 1.5785905571677933e-05, "loss": 0.0327, "step": 75290 }, { "grad_norm": 0.5885730385780334, "learning_rate": 1.5773850049904086e-05, "loss": 0.0358, "step": 75300 }, { "grad_norm": 0.6780279874801636, "learning_rate": 1.5761798271050543e-05, "loss": 0.0377, "step": 75310 }, { "grad_norm": 0.4357964098453522, "learning_rate": 1.5749750236435277e-05, "loss": 0.0312, "step": 75320 }, { "grad_norm": 0.634122371673584, "learning_rate": 1.5737705947375815e-05, "loss": 0.0402, "step": 75330 }, { "grad_norm": 0.49911999702453613, "learning_rate": 1.5725665405189322e-05, "loss": 0.034, "step": 75340 }, { "grad_norm": 0.45384126901626587, "learning_rate": 1.571362861119253e-05, "loss": 0.0375, "step": 75350 }, { "grad_norm": 0.29489612579345703, "learning_rate": 1.570159556670177e-05, "loss": 0.0328, "step": 75360 }, { "grad_norm": 0.2760561406612396, "learning_rate": 1.5689566273032946e-05, "loss": 0.0374, "step": 75370 }, { "grad_norm": 0.35600191354751587, "learning_rate": 1.5677540731501583e-05, "loss": 0.0465, "step": 75380 }, { "grad_norm": 0.4597538709640503, "learning_rate": 1.5665518943422746e-05, "loss": 0.0335, "step": 75390 }, { "grad_norm": 0.3653445243835449, "learning_rate": 1.5653500910111134e-05, "loss": 0.034, "step": 75400 }, { "grad_norm": 0.6004595756530762, "learning_rate": 1.5641486632881023e-05, "loss": 0.0347, "step": 75410 }, { "grad_norm": 0.36530041694641113, "learning_rate": 1.56294761130463e-05, "loss": 0.0305, "step": 75420 }, { "grad_norm": 0.4087637960910797, "learning_rate": 1.5617469351920373e-05, "loss": 0.0362, "step": 75430 }, { "grad_norm": 0.35782018303871155, "learning_rate": 1.5605466350816315e-05, "loss": 0.0331, "step": 75440 }, { "grad_norm": 0.39143311977386475, "learning_rate": 1.559346711104674e-05, "loss": 0.033, "step": 75450 }, { "grad_norm": 0.45463326573371887, "learning_rate": 1.5581471633923883e-05, "loss": 0.0374, "step": 75460 }, { "grad_norm": 0.46097907423973083, "learning_rate": 1.5569479920759505e-05, "loss": 0.0349, "step": 75470 }, { "grad_norm": 0.45878633856773376, "learning_rate": 1.5557491972865072e-05, "loss": 0.0335, "step": 75480 }, { "grad_norm": 0.22070297598838806, "learning_rate": 1.554550779155152e-05, "loss": 0.0229, "step": 75490 }, { "grad_norm": 0.34183424711227417, "learning_rate": 1.5533527378129453e-05, "loss": 0.0261, "step": 75500 }, { "grad_norm": 0.3921383023262024, "learning_rate": 1.552155073390899e-05, "loss": 0.0345, "step": 75510 }, { "grad_norm": 0.8531547784805298, "learning_rate": 1.5509577860199903e-05, "loss": 0.0466, "step": 75520 }, { "grad_norm": 0.25215375423431396, "learning_rate": 1.5497608758311525e-05, "loss": 0.0376, "step": 75530 }, { "grad_norm": 0.4334830343723297, "learning_rate": 1.5485643429552797e-05, "loss": 0.0342, "step": 75540 }, { "grad_norm": 0.45237547159194946, "learning_rate": 1.5473681875232197e-05, "loss": 0.0364, "step": 75550 }, { "grad_norm": 0.8830284476280212, "learning_rate": 1.5461724096657837e-05, "loss": 0.0328, "step": 75560 }, { "grad_norm": 0.37143969535827637, "learning_rate": 1.5449770095137416e-05, "loss": 0.032, "step": 75570 }, { "grad_norm": 0.4888467788696289, "learning_rate": 1.5437819871978178e-05, "loss": 0.0301, "step": 75580 }, { "grad_norm": 0.6484808325767517, "learning_rate": 1.5425873428486992e-05, "loss": 0.0359, "step": 75590 }, { "grad_norm": 0.2651945948600769, "learning_rate": 1.5413930765970312e-05, "loss": 0.034, "step": 75600 }, { "grad_norm": 0.44969332218170166, "learning_rate": 1.5401991885734175e-05, "loss": 0.0368, "step": 75610 }, { "grad_norm": 0.42854538559913635, "learning_rate": 1.5390056789084174e-05, "loss": 0.0285, "step": 75620 }, { "grad_norm": 0.6623419523239136, "learning_rate": 1.5378125477325545e-05, "loss": 0.0278, "step": 75630 }, { "grad_norm": 0.29232165217399597, "learning_rate": 1.5366197951763046e-05, "loss": 0.0287, "step": 75640 }, { "grad_norm": 0.6735997200012207, "learning_rate": 1.5354274213701075e-05, "loss": 0.0356, "step": 75650 }, { "grad_norm": 0.32670608162879944, "learning_rate": 1.5342354264443566e-05, "loss": 0.032, "step": 75660 }, { "grad_norm": 0.560443103313446, "learning_rate": 1.533043810529411e-05, "loss": 0.035, "step": 75670 }, { "grad_norm": 0.3905884027481079, "learning_rate": 1.5318525737555807e-05, "loss": 0.041, "step": 75680 }, { "grad_norm": 0.613964855670929, "learning_rate": 1.53066171625314e-05, "loss": 0.0366, "step": 75690 }, { "grad_norm": 0.6572368144989014, "learning_rate": 1.529471238152317e-05, "loss": 0.0334, "step": 75700 }, { "grad_norm": 0.452227383852005, "learning_rate": 1.5282811395833023e-05, "loss": 0.0322, "step": 75710 }, { "grad_norm": 0.46605074405670166, "learning_rate": 1.5270914206762397e-05, "loss": 0.0325, "step": 75720 }, { "grad_norm": 0.39844685792922974, "learning_rate": 1.525902081561241e-05, "loss": 0.0297, "step": 75730 }, { "grad_norm": 0.36139410734176636, "learning_rate": 1.5247131223683664e-05, "loss": 0.0354, "step": 75740 }, { "grad_norm": 0.5245230793952942, "learning_rate": 1.5235245432276408e-05, "loss": 0.0289, "step": 75750 }, { "grad_norm": 0.4002263844013214, "learning_rate": 1.5223363442690424e-05, "loss": 0.0335, "step": 75760 }, { "grad_norm": 0.3439808487892151, "learning_rate": 1.5211485256225133e-05, "loss": 0.0292, "step": 75770 }, { "grad_norm": 0.6594204902648926, "learning_rate": 1.519961087417951e-05, "loss": 0.0252, "step": 75780 }, { "grad_norm": 0.29046592116355896, "learning_rate": 1.5187740297852137e-05, "loss": 0.0337, "step": 75790 }, { "grad_norm": 0.2479471117258072, "learning_rate": 1.517587352854113e-05, "loss": 0.0252, "step": 75800 }, { "grad_norm": 0.865129828453064, "learning_rate": 1.516401056754423e-05, "loss": 0.0355, "step": 75810 }, { "grad_norm": 0.4966765344142914, "learning_rate": 1.5152151416158783e-05, "loss": 0.0357, "step": 75820 }, { "grad_norm": 0.35039022564888, "learning_rate": 1.5140296075681642e-05, "loss": 0.0355, "step": 75830 }, { "grad_norm": 0.4242498278617859, "learning_rate": 1.5128444547409331e-05, "loss": 0.0289, "step": 75840 }, { "grad_norm": 0.6538329124450684, "learning_rate": 1.5116596832637864e-05, "loss": 0.0282, "step": 75850 }, { "grad_norm": 0.409799188375473, "learning_rate": 1.5104752932662952e-05, "loss": 0.0375, "step": 75860 }, { "grad_norm": 0.5462835431098938, "learning_rate": 1.5092912848779778e-05, "loss": 0.0366, "step": 75870 }, { "grad_norm": 0.5164987444877625, "learning_rate": 1.5081076582283194e-05, "loss": 0.0305, "step": 75880 }, { "grad_norm": 0.262474924325943, "learning_rate": 1.5069244134467553e-05, "loss": 0.0262, "step": 75890 }, { "grad_norm": 0.32201799750328064, "learning_rate": 1.5057415506626876e-05, "loss": 0.0352, "step": 75900 }, { "grad_norm": 0.43631991744041443, "learning_rate": 1.5045590700054668e-05, "loss": 0.0324, "step": 75910 }, { "grad_norm": 0.4854012727737427, "learning_rate": 1.5033769716044144e-05, "loss": 0.0268, "step": 75920 }, { "grad_norm": 1.1286855936050415, "learning_rate": 1.5021952555887975e-05, "loss": 0.0291, "step": 75930 }, { "grad_norm": 0.27411332726478577, "learning_rate": 1.50101392208785e-05, "loss": 0.0305, "step": 75940 }, { "grad_norm": 0.484207421541214, "learning_rate": 1.4998329712307579e-05, "loss": 0.0251, "step": 75950 }, { "grad_norm": 0.5045138001441956, "learning_rate": 1.4986524031466708e-05, "loss": 0.0351, "step": 75960 }, { "grad_norm": 0.42353254556655884, "learning_rate": 1.4974722179646916e-05, "loss": 0.0217, "step": 75970 }, { "grad_norm": 0.5997898578643799, "learning_rate": 1.4962924158138847e-05, "loss": 0.0409, "step": 75980 }, { "grad_norm": 0.3995409309864044, "learning_rate": 1.4951129968232714e-05, "loss": 0.0307, "step": 75990 }, { "grad_norm": 0.739828884601593, "learning_rate": 1.4939339611218329e-05, "loss": 0.0338, "step": 76000 }, { "grad_norm": 0.4311608374118805, "learning_rate": 1.4927553088385032e-05, "loss": 0.0327, "step": 76010 }, { "grad_norm": 0.4265004098415375, "learning_rate": 1.4915770401021801e-05, "loss": 0.0265, "step": 76020 }, { "grad_norm": 0.5060151219367981, "learning_rate": 1.4903991550417185e-05, "loss": 0.0466, "step": 76030 }, { "grad_norm": 0.3951425850391388, "learning_rate": 1.4892216537859277e-05, "loss": 0.0392, "step": 76040 }, { "grad_norm": 0.6581159234046936, "learning_rate": 1.4880445364635775e-05, "loss": 0.0306, "step": 76050 }, { "grad_norm": 0.4676905870437622, "learning_rate": 1.4868678032033973e-05, "loss": 0.0368, "step": 76060 }, { "grad_norm": 0.4521365165710449, "learning_rate": 1.4856914541340739e-05, "loss": 0.0298, "step": 76070 }, { "grad_norm": 0.35682761669158936, "learning_rate": 1.4845154893842472e-05, "loss": 0.0317, "step": 76080 }, { "grad_norm": 0.5322655439376831, "learning_rate": 1.4833399090825234e-05, "loss": 0.0408, "step": 76090 }, { "grad_norm": 0.5821700692176819, "learning_rate": 1.4821647133574562e-05, "loss": 0.0402, "step": 76100 }, { "grad_norm": 0.5846085548400879, "learning_rate": 1.4809899023375706e-05, "loss": 0.0297, "step": 76110 }, { "grad_norm": 0.6051536202430725, "learning_rate": 1.4798154761513366e-05, "loss": 0.0382, "step": 76120 }, { "grad_norm": 0.3955062925815582, "learning_rate": 1.4786414349271916e-05, "loss": 0.0332, "step": 76130 }, { "grad_norm": 2.6760458946228027, "learning_rate": 1.4774677787935226e-05, "loss": 0.0336, "step": 76140 }, { "grad_norm": 0.5683445930480957, "learning_rate": 1.4762945078786838e-05, "loss": 0.0313, "step": 76150 }, { "grad_norm": 0.4133862853050232, "learning_rate": 1.4751216223109777e-05, "loss": 0.0265, "step": 76160 }, { "grad_norm": 0.5787840485572815, "learning_rate": 1.4739491222186714e-05, "loss": 0.0353, "step": 76170 }, { "grad_norm": 0.6045151352882385, "learning_rate": 1.4727770077299879e-05, "loss": 0.0354, "step": 76180 }, { "grad_norm": 0.2757772207260132, "learning_rate": 1.4716052789731094e-05, "loss": 0.0339, "step": 76190 }, { "grad_norm": 0.6096463203430176, "learning_rate": 1.470433936076171e-05, "loss": 0.0327, "step": 76200 }, { "grad_norm": 0.3877055048942566, "learning_rate": 1.4692629791672719e-05, "loss": 0.0275, "step": 76210 }, { "grad_norm": 0.8028006553649902, "learning_rate": 1.468092408374464e-05, "loss": 0.0314, "step": 76220 }, { "grad_norm": 0.6807332634925842, "learning_rate": 1.4669222238257602e-05, "loss": 0.045, "step": 76230 }, { "grad_norm": 0.399600625038147, "learning_rate": 1.4657524256491301e-05, "loss": 0.0285, "step": 76240 }, { "grad_norm": 0.4298838973045349, "learning_rate": 1.464583013972502e-05, "loss": 0.0281, "step": 76250 }, { "grad_norm": 0.2919982373714447, "learning_rate": 1.4634139889237591e-05, "loss": 0.0295, "step": 76260 }, { "grad_norm": 0.356456995010376, "learning_rate": 1.4622453506307448e-05, "loss": 0.0386, "step": 76270 }, { "grad_norm": 0.49721601605415344, "learning_rate": 1.461077099221262e-05, "loss": 0.03, "step": 76280 }, { "grad_norm": 1.3032543659210205, "learning_rate": 1.4599092348230652e-05, "loss": 0.0331, "step": 76290 }, { "grad_norm": 0.5610826015472412, "learning_rate": 1.4587417575638724e-05, "loss": 0.0269, "step": 76300 }, { "grad_norm": 0.6714617013931274, "learning_rate": 1.4575746675713563e-05, "loss": 0.0307, "step": 76310 }, { "grad_norm": 0.5076737999916077, "learning_rate": 1.4564079649731504e-05, "loss": 0.0334, "step": 76320 }, { "grad_norm": 0.43639594316482544, "learning_rate": 1.4552416498968402e-05, "loss": 0.0294, "step": 76330 }, { "grad_norm": 0.5661795139312744, "learning_rate": 1.4540757224699747e-05, "loss": 0.0261, "step": 76340 }, { "grad_norm": 0.30854862928390503, "learning_rate": 1.452910182820056e-05, "loss": 0.0338, "step": 76350 }, { "grad_norm": 0.49826428294181824, "learning_rate": 1.4517450310745468e-05, "loss": 0.0343, "step": 76360 }, { "grad_norm": 0.37970972061157227, "learning_rate": 1.450580267360866e-05, "loss": 0.0292, "step": 76370 }, { "grad_norm": 0.22619791328907013, "learning_rate": 1.4494158918063922e-05, "loss": 0.0297, "step": 76380 }, { "grad_norm": 0.48052966594696045, "learning_rate": 1.4482519045384568e-05, "loss": 0.0286, "step": 76390 }, { "grad_norm": 0.4707590341567993, "learning_rate": 1.4470883056843548e-05, "loss": 0.032, "step": 76400 }, { "grad_norm": 0.47092780470848083, "learning_rate": 1.4459250953713322e-05, "loss": 0.0367, "step": 76410 }, { "grad_norm": 0.5230299830436707, "learning_rate": 1.444762273726598e-05, "loss": 0.0324, "step": 76420 }, { "grad_norm": 0.3415740430355072, "learning_rate": 1.443599840877316e-05, "loss": 0.0334, "step": 76430 }, { "grad_norm": 0.35706955194473267, "learning_rate": 1.4424377969506103e-05, "loss": 0.0315, "step": 76440 }, { "grad_norm": 0.6553839445114136, "learning_rate": 1.4412761420735565e-05, "loss": 0.0368, "step": 76450 }, { "grad_norm": 0.6645786762237549, "learning_rate": 1.4401148763731953e-05, "loss": 0.036, "step": 76460 }, { "grad_norm": 0.5478073954582214, "learning_rate": 1.4389539999765173e-05, "loss": 0.0318, "step": 76470 }, { "grad_norm": 0.4872435927391052, "learning_rate": 1.4377935130104758e-05, "loss": 0.031, "step": 76480 }, { "grad_norm": 0.3597814738750458, "learning_rate": 1.4366334156019801e-05, "loss": 0.0348, "step": 76490 }, { "grad_norm": 0.43039247393608093, "learning_rate": 1.435473707877898e-05, "loss": 0.0341, "step": 76500 }, { "grad_norm": 0.3827939033508301, "learning_rate": 1.4343143899650508e-05, "loss": 0.035, "step": 76510 }, { "grad_norm": 0.338187038898468, "learning_rate": 1.4331554619902204e-05, "loss": 0.033, "step": 76520 }, { "grad_norm": 0.5721644759178162, "learning_rate": 1.4319969240801484e-05, "loss": 0.0264, "step": 76530 }, { "grad_norm": 0.4096975326538086, "learning_rate": 1.430838776361526e-05, "loss": 0.0383, "step": 76540 }, { "grad_norm": 0.4121900200843811, "learning_rate": 1.4296810189610093e-05, "loss": 0.0401, "step": 76550 }, { "grad_norm": 0.336464524269104, "learning_rate": 1.4285236520052087e-05, "loss": 0.0297, "step": 76560 }, { "grad_norm": 0.37732914090156555, "learning_rate": 1.4273666756206933e-05, "loss": 0.0302, "step": 76570 }, { "grad_norm": 0.36496564745903015, "learning_rate": 1.4262100899339848e-05, "loss": 0.0332, "step": 76580 }, { "grad_norm": 0.29608190059661865, "learning_rate": 1.4250538950715696e-05, "loss": 0.0275, "step": 76590 }, { "grad_norm": 0.5355689525604248, "learning_rate": 1.4238980911598842e-05, "loss": 0.0424, "step": 76600 }, { "grad_norm": 0.31758132576942444, "learning_rate": 1.4227426783253267e-05, "loss": 0.0318, "step": 76610 }, { "grad_norm": 0.45606815814971924, "learning_rate": 1.4215876566942521e-05, "loss": 0.0336, "step": 76620 }, { "grad_norm": 0.6381782293319702, "learning_rate": 1.4204330263929721e-05, "loss": 0.0265, "step": 76630 }, { "grad_norm": 0.45418089628219604, "learning_rate": 1.4192787875477537e-05, "loss": 0.0421, "step": 76640 }, { "grad_norm": 0.1810838133096695, "learning_rate": 1.4181249402848246e-05, "loss": 0.0305, "step": 76650 }, { "grad_norm": 0.34558573365211487, "learning_rate": 1.4169714847303656e-05, "loss": 0.0298, "step": 76660 }, { "grad_norm": 0.7744959592819214, "learning_rate": 1.4158184210105197e-05, "loss": 0.037, "step": 76670 }, { "grad_norm": 0.5292722582817078, "learning_rate": 1.4146657492513787e-05, "loss": 0.0322, "step": 76680 }, { "grad_norm": 0.6756730675697327, "learning_rate": 1.4135134695790047e-05, "loss": 0.0281, "step": 76690 }, { "grad_norm": 0.5571401119232178, "learning_rate": 1.4123615821194037e-05, "loss": 0.0384, "step": 76700 }, { "grad_norm": 0.5362396836280823, "learning_rate": 1.4112100869985473e-05, "loss": 0.0327, "step": 76710 }, { "grad_norm": 0.30477067828178406, "learning_rate": 1.4100589843423589e-05, "loss": 0.033, "step": 76720 }, { "grad_norm": 0.3482462167739868, "learning_rate": 1.408908274276723e-05, "loss": 0.0354, "step": 76730 }, { "grad_norm": 0.6534432172775269, "learning_rate": 1.4077579569274786e-05, "loss": 0.0315, "step": 76740 }, { "grad_norm": 0.4572693407535553, "learning_rate": 1.4066080324204245e-05, "loss": 0.0278, "step": 76750 }, { "grad_norm": 0.48532307147979736, "learning_rate": 1.405458500881312e-05, "loss": 0.0372, "step": 76760 }, { "grad_norm": 0.3836328983306885, "learning_rate": 1.4043093624358538e-05, "loss": 0.0303, "step": 76770 }, { "grad_norm": 0.34578177332878113, "learning_rate": 1.4031606172097189e-05, "loss": 0.0486, "step": 76780 }, { "grad_norm": 0.39875897765159607, "learning_rate": 1.40201226532853e-05, "loss": 0.0409, "step": 76790 }, { "grad_norm": 0.33008089661598206, "learning_rate": 1.4008643069178701e-05, "loss": 0.0282, "step": 76800 }, { "grad_norm": 0.3745155930519104, "learning_rate": 1.399716742103278e-05, "loss": 0.0313, "step": 76810 }, { "grad_norm": 0.4425872266292572, "learning_rate": 1.398569571010252e-05, "loss": 0.0311, "step": 76820 }, { "grad_norm": 0.6151439547538757, "learning_rate": 1.3974227937642414e-05, "loss": 0.038, "step": 76830 }, { "grad_norm": 0.4532724916934967, "learning_rate": 1.3962764104906595e-05, "loss": 0.0237, "step": 76840 }, { "grad_norm": 0.26856884360313416, "learning_rate": 1.3951304213148696e-05, "loss": 0.0357, "step": 76850 }, { "grad_norm": 0.3631068170070648, "learning_rate": 1.3939848263621979e-05, "loss": 0.0294, "step": 76860 }, { "grad_norm": 0.797009289264679, "learning_rate": 1.3928396257579219e-05, "loss": 0.0429, "step": 76870 }, { "grad_norm": 0.48168861865997314, "learning_rate": 1.3916948196272833e-05, "loss": 0.0394, "step": 76880 }, { "grad_norm": 0.5173290967941284, "learning_rate": 1.3905504080954734e-05, "loss": 0.0439, "step": 76890 }, { "grad_norm": 0.33452120423316956, "learning_rate": 1.3894063912876454e-05, "loss": 0.036, "step": 76900 }, { "grad_norm": 0.45621591806411743, "learning_rate": 1.3882627693289046e-05, "loss": 0.0278, "step": 76910 }, { "grad_norm": 0.3690461218357086, "learning_rate": 1.3871195423443183e-05, "loss": 0.0313, "step": 76920 }, { "grad_norm": 0.42421138286590576, "learning_rate": 1.3859767104589044e-05, "loss": 0.0334, "step": 76930 }, { "grad_norm": 0.6871045231819153, "learning_rate": 1.3848342737976467e-05, "loss": 0.0312, "step": 76940 }, { "grad_norm": 0.44586285948753357, "learning_rate": 1.3836922324854757e-05, "loss": 0.0334, "step": 76950 }, { "grad_norm": 0.9035670757293701, "learning_rate": 1.382550586647287e-05, "loss": 0.0392, "step": 76960 }, { "grad_norm": 0.25472792983055115, "learning_rate": 1.3814093364079257e-05, "loss": 0.033, "step": 76970 }, { "grad_norm": 0.5979515910148621, "learning_rate": 1.3802684818921991e-05, "loss": 0.0389, "step": 76980 }, { "grad_norm": 0.2618972659111023, "learning_rate": 1.3791280232248687e-05, "loss": 0.0358, "step": 76990 }, { "grad_norm": 0.3553757071495056, "learning_rate": 1.3779879605306555e-05, "loss": 0.0338, "step": 77000 }, { "grad_norm": 0.48795831203460693, "learning_rate": 1.376848293934232e-05, "loss": 0.0328, "step": 77010 }, { "grad_norm": 0.3228665888309479, "learning_rate": 1.3757090235602316e-05, "loss": 0.0274, "step": 77020 }, { "grad_norm": 0.33580100536346436, "learning_rate": 1.3745701495332447e-05, "loss": 0.0332, "step": 77030 }, { "grad_norm": 0.5201590657234192, "learning_rate": 1.3734316719778145e-05, "loss": 0.0265, "step": 77040 }, { "grad_norm": 0.29040849208831787, "learning_rate": 1.3722935910184454e-05, "loss": 0.0279, "step": 77050 }, { "grad_norm": 0.46015432476997375, "learning_rate": 1.3711559067795921e-05, "loss": 0.031, "step": 77060 }, { "grad_norm": 0.3340958058834076, "learning_rate": 1.3700186193856756e-05, "loss": 0.0402, "step": 77070 }, { "grad_norm": 0.5293184518814087, "learning_rate": 1.368881728961064e-05, "loss": 0.0297, "step": 77080 }, { "grad_norm": 0.35731184482574463, "learning_rate": 1.3677452356300885e-05, "loss": 0.0306, "step": 77090 }, { "grad_norm": 0.5542365312576294, "learning_rate": 1.3666091395170321e-05, "loss": 0.0344, "step": 77100 }, { "grad_norm": 0.554040789604187, "learning_rate": 1.3654734407461384e-05, "loss": 0.0341, "step": 77110 }, { "grad_norm": 0.6162955164909363, "learning_rate": 1.3643381394416022e-05, "loss": 0.0283, "step": 77120 }, { "grad_norm": 0.6061355471611023, "learning_rate": 1.3632032357275838e-05, "loss": 0.0338, "step": 77130 }, { "grad_norm": 0.3317607343196869, "learning_rate": 1.3620687297281898e-05, "loss": 0.0296, "step": 77140 }, { "grad_norm": 1.920413613319397, "learning_rate": 1.3609346215674917e-05, "loss": 0.0266, "step": 77150 }, { "grad_norm": 0.6042835116386414, "learning_rate": 1.3598009113695109e-05, "loss": 0.0369, "step": 77160 }, { "grad_norm": 0.29872915148735046, "learning_rate": 1.358667599258231e-05, "loss": 0.0305, "step": 77170 }, { "grad_norm": 0.4255705773830414, "learning_rate": 1.357534685357586e-05, "loss": 0.0341, "step": 77180 }, { "grad_norm": 0.5788257122039795, "learning_rate": 1.3564021697914715e-05, "loss": 0.0325, "step": 77190 }, { "grad_norm": 0.3585653007030487, "learning_rate": 1.3552700526837381e-05, "loss": 0.0307, "step": 77200 }, { "grad_norm": 0.32371389865875244, "learning_rate": 1.3541383341581925e-05, "loss": 0.0271, "step": 77210 }, { "grad_norm": 0.5961920022964478, "learning_rate": 1.3530070143385965e-05, "loss": 0.0282, "step": 77220 }, { "grad_norm": 0.34420421719551086, "learning_rate": 1.3518760933486702e-05, "loss": 0.0227, "step": 77230 }, { "grad_norm": 0.3896526098251343, "learning_rate": 1.3507455713120908e-05, "loss": 0.0311, "step": 77240 }, { "grad_norm": 0.2841308116912842, "learning_rate": 1.3496154483524876e-05, "loss": 0.0275, "step": 77250 }, { "grad_norm": 0.8145843744277954, "learning_rate": 1.3484857245934506e-05, "loss": 0.0346, "step": 77260 }, { "grad_norm": 0.2758523225784302, "learning_rate": 1.3473564001585248e-05, "loss": 0.0363, "step": 77270 }, { "grad_norm": 0.3693941831588745, "learning_rate": 1.3462274751712129e-05, "loss": 0.0426, "step": 77280 }, { "grad_norm": 0.3675427734851837, "learning_rate": 1.3450989497549698e-05, "loss": 0.0321, "step": 77290 }, { "grad_norm": 0.5647257566452026, "learning_rate": 1.3439708240332115e-05, "loss": 0.0393, "step": 77300 }, { "grad_norm": 0.47629299759864807, "learning_rate": 1.3428430981293043e-05, "loss": 0.0328, "step": 77310 }, { "grad_norm": 0.37775304913520813, "learning_rate": 1.3417157721665808e-05, "loss": 0.0376, "step": 77320 }, { "grad_norm": 0.3519171178340912, "learning_rate": 1.3405888462683186e-05, "loss": 0.0256, "step": 77330 }, { "grad_norm": 0.509892463684082, "learning_rate": 1.3394623205577595e-05, "loss": 0.0339, "step": 77340 }, { "grad_norm": 0.6479668617248535, "learning_rate": 1.3383361951580963e-05, "loss": 0.0301, "step": 77350 }, { "grad_norm": 0.5500020980834961, "learning_rate": 1.337210470192483e-05, "loss": 0.0343, "step": 77360 }, { "grad_norm": 0.4525749683380127, "learning_rate": 1.3360851457840245e-05, "loss": 0.0352, "step": 77370 }, { "grad_norm": 0.2478107511997223, "learning_rate": 1.3349602220557856e-05, "loss": 0.026, "step": 77380 }, { "grad_norm": 0.536191463470459, "learning_rate": 1.3338356991307865e-05, "loss": 0.0416, "step": 77390 }, { "grad_norm": 0.6260436177253723, "learning_rate": 1.3327115771320048e-05, "loss": 0.0343, "step": 77400 }, { "grad_norm": 0.4229280650615692, "learning_rate": 1.3315878561823697e-05, "loss": 0.0256, "step": 77410 }, { "grad_norm": 0.7904686331748962, "learning_rate": 1.330464536404773e-05, "loss": 0.0307, "step": 77420 }, { "grad_norm": 0.413868248462677, "learning_rate": 1.3293416179220563e-05, "loss": 0.0354, "step": 77430 }, { "grad_norm": 0.7841487526893616, "learning_rate": 1.328219100857021e-05, "loss": 0.03, "step": 77440 }, { "grad_norm": 0.4522055387496948, "learning_rate": 1.327096985332425e-05, "loss": 0.0335, "step": 77450 }, { "grad_norm": 0.6018089652061462, "learning_rate": 1.3259752714709816e-05, "loss": 0.0326, "step": 77460 }, { "grad_norm": 0.686632513999939, "learning_rate": 1.3248539593953569e-05, "loss": 0.0309, "step": 77470 }, { "grad_norm": 0.6187466382980347, "learning_rate": 1.323733049228178e-05, "loss": 0.0362, "step": 77480 }, { "grad_norm": 0.5952764749526978, "learning_rate": 1.3226125410920276e-05, "loss": 0.0362, "step": 77490 }, { "grad_norm": 0.9015006422996521, "learning_rate": 1.3214924351094388e-05, "loss": 0.032, "step": 77500 }, { "grad_norm": 0.3380313813686371, "learning_rate": 1.3203727314029069e-05, "loss": 0.0284, "step": 77510 }, { "grad_norm": 0.5533334016799927, "learning_rate": 1.3192534300948811e-05, "loss": 0.0357, "step": 77520 }, { "grad_norm": 0.4280312955379486, "learning_rate": 1.3181345313077675e-05, "loss": 0.0393, "step": 77530 }, { "grad_norm": 0.6050316691398621, "learning_rate": 1.3170160351639243e-05, "loss": 0.0357, "step": 77540 }, { "grad_norm": 0.5066415667533875, "learning_rate": 1.3158979417856721e-05, "loss": 0.0397, "step": 77550 }, { "grad_norm": 0.4882713854312897, "learning_rate": 1.31478025129528e-05, "loss": 0.0335, "step": 77560 }, { "grad_norm": 0.3611270487308502, "learning_rate": 1.3136629638149794e-05, "loss": 0.03, "step": 77570 }, { "grad_norm": 0.48512953519821167, "learning_rate": 1.3125460794669547e-05, "loss": 0.032, "step": 77580 }, { "grad_norm": 0.49754366278648376, "learning_rate": 1.3114295983733483e-05, "loss": 0.0306, "step": 77590 }, { "grad_norm": 0.4296010434627533, "learning_rate": 1.3103135206562534e-05, "loss": 0.0421, "step": 77600 }, { "grad_norm": 0.6422790288925171, "learning_rate": 1.3091978464377264e-05, "loss": 0.0383, "step": 77610 }, { "grad_norm": 0.2987949848175049, "learning_rate": 1.3080825758397724e-05, "loss": 0.0323, "step": 77620 }, { "grad_norm": 0.8288151025772095, "learning_rate": 1.3069677089843568e-05, "loss": 0.0327, "step": 77630 }, { "grad_norm": 0.509022057056427, "learning_rate": 1.305853245993401e-05, "loss": 0.0346, "step": 77640 }, { "grad_norm": 0.41463741660118103, "learning_rate": 1.3047391869887809e-05, "loss": 0.0333, "step": 77650 }, { "grad_norm": 0.4841182827949524, "learning_rate": 1.3036255320923263e-05, "loss": 0.0294, "step": 77660 }, { "grad_norm": 0.45375779271125793, "learning_rate": 1.3025122814258273e-05, "loss": 0.0269, "step": 77670 }, { "grad_norm": 0.811795711517334, "learning_rate": 1.3013994351110249e-05, "loss": 0.0355, "step": 77680 }, { "grad_norm": 0.41060805320739746, "learning_rate": 1.3002869932696194e-05, "loss": 0.0324, "step": 77690 }, { "grad_norm": 0.5073956251144409, "learning_rate": 1.2991749560232664e-05, "loss": 0.0354, "step": 77700 }, { "grad_norm": 0.33499598503112793, "learning_rate": 1.2980633234935768e-05, "loss": 0.0215, "step": 77710 }, { "grad_norm": 0.4253293573856354, "learning_rate": 1.2969520958021153e-05, "loss": 0.0342, "step": 77720 }, { "grad_norm": 0.3548421859741211, "learning_rate": 1.295841273070405e-05, "loss": 0.0267, "step": 77730 }, { "grad_norm": 0.4088135063648224, "learning_rate": 1.2947308554199255e-05, "loss": 0.0327, "step": 77740 }, { "grad_norm": 0.5612087249755859, "learning_rate": 1.2936208429721075e-05, "loss": 0.041, "step": 77750 }, { "grad_norm": 0.2702343165874481, "learning_rate": 1.2925112358483416e-05, "loss": 0.0267, "step": 77760 }, { "grad_norm": 0.3576396703720093, "learning_rate": 1.2914020341699728e-05, "loss": 0.0354, "step": 77770 }, { "grad_norm": 0.39162319898605347, "learning_rate": 1.2902932380583032e-05, "loss": 0.0291, "step": 77780 }, { "grad_norm": 0.5235079526901245, "learning_rate": 1.2891848476345863e-05, "loss": 0.035, "step": 77790 }, { "grad_norm": 0.47700467705726624, "learning_rate": 1.2880768630200368e-05, "loss": 0.0335, "step": 77800 }, { "grad_norm": 0.444651335477829, "learning_rate": 1.286969284335819e-05, "loss": 0.0264, "step": 77810 }, { "grad_norm": 0.45515134930610657, "learning_rate": 1.2858621117030578e-05, "loss": 0.0277, "step": 77820 }, { "grad_norm": 0.43890315294265747, "learning_rate": 1.2847553452428324e-05, "loss": 0.0238, "step": 77830 }, { "grad_norm": 0.26283422112464905, "learning_rate": 1.283648985076178e-05, "loss": 0.0224, "step": 77840 }, { "grad_norm": 0.6494321227073669, "learning_rate": 1.2825430313240817e-05, "loss": 0.0401, "step": 77850 }, { "grad_norm": 0.62565678358078, "learning_rate": 1.281437484107491e-05, "loss": 0.0325, "step": 77860 }, { "grad_norm": 0.29552891850471497, "learning_rate": 1.2803323435473058e-05, "loss": 0.0296, "step": 77870 }, { "grad_norm": 0.4073113203048706, "learning_rate": 1.2792276097643841e-05, "loss": 0.0293, "step": 77880 }, { "grad_norm": 0.5514587163925171, "learning_rate": 1.2781232828795342e-05, "loss": 0.0313, "step": 77890 }, { "grad_norm": 0.36359676718711853, "learning_rate": 1.2770193630135292e-05, "loss": 0.0379, "step": 77900 }, { "grad_norm": 1.1828418970108032, "learning_rate": 1.2759158502870877e-05, "loss": 0.0371, "step": 77910 }, { "grad_norm": 0.4132232666015625, "learning_rate": 1.2748127448208913e-05, "loss": 0.0287, "step": 77920 }, { "grad_norm": 0.25045567750930786, "learning_rate": 1.2737100467355706e-05, "loss": 0.0394, "step": 77930 }, { "grad_norm": 0.6187223792076111, "learning_rate": 1.2726077561517169e-05, "loss": 0.0402, "step": 77940 }, { "grad_norm": 0.3023138642311096, "learning_rate": 1.2715058731898755e-05, "loss": 0.0306, "step": 77950 }, { "grad_norm": 0.5053488612174988, "learning_rate": 1.2704043979705471e-05, "loss": 0.0426, "step": 77960 }, { "grad_norm": 0.6433627605438232, "learning_rate": 1.269303330614185e-05, "loss": 0.0343, "step": 77970 }, { "grad_norm": 0.8935604691505432, "learning_rate": 1.2682026712412015e-05, "loss": 0.0347, "step": 77980 }, { "grad_norm": 0.37399911880493164, "learning_rate": 1.2671024199719644e-05, "loss": 0.0259, "step": 77990 }, { "grad_norm": 0.6473186612129211, "learning_rate": 1.2660025769267936e-05, "loss": 0.0336, "step": 78000 }, { "grad_norm": 0.7915492653846741, "learning_rate": 1.2649031422259667e-05, "loss": 0.0316, "step": 78010 }, { "grad_norm": 0.5697062015533447, "learning_rate": 1.2638041159897163e-05, "loss": 0.0307, "step": 78020 }, { "grad_norm": 0.21520639955997467, "learning_rate": 1.262705498338232e-05, "loss": 0.036, "step": 78030 }, { "grad_norm": 0.3033694624900818, "learning_rate": 1.2616072893916542e-05, "loss": 0.0397, "step": 78040 }, { "grad_norm": 0.28836095333099365, "learning_rate": 1.2605094892700836e-05, "loss": 0.0337, "step": 78050 }, { "grad_norm": 0.3144679665565491, "learning_rate": 1.259412098093572e-05, "loss": 0.0318, "step": 78060 }, { "grad_norm": 0.3422110378742218, "learning_rate": 1.258315115982131e-05, "loss": 0.0247, "step": 78070 }, { "grad_norm": 0.6869938969612122, "learning_rate": 1.2572185430557203e-05, "loss": 0.0289, "step": 78080 }, { "grad_norm": 0.8351108431816101, "learning_rate": 1.2561223794342659e-05, "loss": 0.0297, "step": 78090 }, { "grad_norm": 0.3805723488330841, "learning_rate": 1.255026625237637e-05, "loss": 0.0375, "step": 78100 }, { "grad_norm": 0.4553017020225525, "learning_rate": 1.2539312805856678e-05, "loss": 0.0403, "step": 78110 }, { "grad_norm": 0.5167204737663269, "learning_rate": 1.25283634559814e-05, "loss": 0.0337, "step": 78120 }, { "grad_norm": 0.17247937619686127, "learning_rate": 1.2517418203947972e-05, "loss": 0.0352, "step": 78130 }, { "grad_norm": 0.380243718624115, "learning_rate": 1.2506477050953308e-05, "loss": 0.0275, "step": 78140 }, { "grad_norm": 0.3520427644252777, "learning_rate": 1.2495539998193972e-05, "loss": 0.0341, "step": 78150 }, { "grad_norm": 0.34612512588500977, "learning_rate": 1.2484607046865975e-05, "loss": 0.03, "step": 78160 }, { "grad_norm": 0.40625542402267456, "learning_rate": 1.2473678198164968e-05, "loss": 0.0279, "step": 78170 }, { "grad_norm": 0.47977277636528015, "learning_rate": 1.2462753453286075e-05, "loss": 0.0327, "step": 78180 }, { "grad_norm": 0.2990202009677887, "learning_rate": 1.2451832813424031e-05, "loss": 0.0519, "step": 78190 }, { "grad_norm": 0.5213788151741028, "learning_rate": 1.2440916279773095e-05, "loss": 0.0311, "step": 78200 }, { "grad_norm": 0.31512725353240967, "learning_rate": 1.2430003853527101e-05, "loss": 0.0355, "step": 78210 }, { "grad_norm": 0.6303650140762329, "learning_rate": 1.2419095535879383e-05, "loss": 0.0482, "step": 78220 }, { "grad_norm": 0.2545967102050781, "learning_rate": 1.2408191328022878e-05, "loss": 0.0334, "step": 78230 }, { "grad_norm": 0.49653974175453186, "learning_rate": 1.2397291231150066e-05, "loss": 0.0305, "step": 78240 }, { "grad_norm": 0.8593483567237854, "learning_rate": 1.2386395246452937e-05, "loss": 0.0364, "step": 78250 }, { "grad_norm": 0.31446099281311035, "learning_rate": 1.2375503375123083e-05, "loss": 0.0336, "step": 78260 }, { "grad_norm": 0.2060113400220871, "learning_rate": 1.2364615618351583e-05, "loss": 0.0428, "step": 78270 }, { "grad_norm": 0.4033711850643158, "learning_rate": 1.2353731977329169e-05, "loss": 0.0308, "step": 78280 }, { "grad_norm": 0.4171411693096161, "learning_rate": 1.2342852453246007e-05, "loss": 0.0221, "step": 78290 }, { "grad_norm": 0.5310978889465332, "learning_rate": 1.2331977047291898e-05, "loss": 0.0332, "step": 78300 }, { "grad_norm": 0.6540131568908691, "learning_rate": 1.2321105760656137e-05, "loss": 0.0352, "step": 78310 }, { "grad_norm": 0.28820985555648804, "learning_rate": 1.2310238594527613e-05, "loss": 0.0338, "step": 78320 }, { "grad_norm": 0.5167069435119629, "learning_rate": 1.2299375550094693e-05, "loss": 0.0297, "step": 78330 }, { "grad_norm": 0.6006011366844177, "learning_rate": 1.2288516628545421e-05, "loss": 0.0308, "step": 78340 }, { "grad_norm": 0.3303450047969818, "learning_rate": 1.2277661831067255e-05, "loss": 0.0325, "step": 78350 }, { "grad_norm": 0.23221522569656372, "learning_rate": 1.2266811158847286e-05, "loss": 0.0317, "step": 78360 }, { "grad_norm": 0.3747505843639374, "learning_rate": 1.2255964613072107e-05, "loss": 0.0231, "step": 78370 }, { "grad_norm": 0.4453483521938324, "learning_rate": 1.2245122194927905e-05, "loss": 0.0244, "step": 78380 }, { "grad_norm": 0.40427154302597046, "learning_rate": 1.2234283905600364e-05, "loss": 0.0357, "step": 78390 }, { "grad_norm": 0.7495762705802917, "learning_rate": 1.222344974627475e-05, "loss": 0.0427, "step": 78400 }, { "grad_norm": 0.3677256107330322, "learning_rate": 1.2212619718135875e-05, "loss": 0.0302, "step": 78410 }, { "grad_norm": 0.39191704988479614, "learning_rate": 1.2201793822368118e-05, "loss": 0.0396, "step": 78420 }, { "grad_norm": 0.506240725517273, "learning_rate": 1.2190972060155337e-05, "loss": 0.0276, "step": 78430 }, { "grad_norm": 0.30190563201904297, "learning_rate": 1.2180154432681013e-05, "loss": 0.0299, "step": 78440 }, { "grad_norm": 0.36546415090560913, "learning_rate": 1.2169340941128148e-05, "loss": 0.0333, "step": 78450 }, { "grad_norm": 0.22500744462013245, "learning_rate": 1.2158531586679267e-05, "loss": 0.0279, "step": 78460 }, { "grad_norm": 0.7919117212295532, "learning_rate": 1.2147726370516477e-05, "loss": 0.0306, "step": 78470 }, { "grad_norm": 0.26468896865844727, "learning_rate": 1.2136925293821422e-05, "loss": 0.0403, "step": 78480 }, { "grad_norm": 0.48787444829940796, "learning_rate": 1.2126128357775306e-05, "loss": 0.0335, "step": 78490 }, { "grad_norm": 0.6335440874099731, "learning_rate": 1.2115335563558838e-05, "loss": 0.0341, "step": 78500 }, { "grad_norm": 0.3275967836380005, "learning_rate": 1.2104546912352327e-05, "loss": 0.0353, "step": 78510 }, { "grad_norm": 0.4355471134185791, "learning_rate": 1.2093762405335557e-05, "loss": 0.0251, "step": 78520 }, { "grad_norm": 0.5374628305435181, "learning_rate": 1.2082982043687974e-05, "loss": 0.0303, "step": 78530 }, { "grad_norm": 0.66145920753479, "learning_rate": 1.207220582858845e-05, "loss": 0.0328, "step": 78540 }, { "grad_norm": 0.5848878026008606, "learning_rate": 1.206143376121549e-05, "loss": 0.0276, "step": 78550 }, { "grad_norm": 0.8555722236633301, "learning_rate": 1.2050665842747078e-05, "loss": 0.0247, "step": 78560 }, { "grad_norm": 0.4951240122318268, "learning_rate": 1.2039902074360804e-05, "loss": 0.0257, "step": 78570 }, { "grad_norm": 0.5863065719604492, "learning_rate": 1.2029142457233755e-05, "loss": 0.035, "step": 78580 }, { "grad_norm": 0.6344230771064758, "learning_rate": 1.2018386992542601e-05, "loss": 0.0346, "step": 78590 }, { "grad_norm": 0.3101809024810791, "learning_rate": 1.2007635681463542e-05, "loss": 0.0353, "step": 78600 }, { "grad_norm": 0.21806596219539642, "learning_rate": 1.1996888525172334e-05, "loss": 0.0312, "step": 78610 }, { "grad_norm": 0.3650397062301636, "learning_rate": 1.198614552484425e-05, "loss": 0.0254, "step": 78620 }, { "grad_norm": 0.4356447756290436, "learning_rate": 1.197540668165415e-05, "loss": 0.0307, "step": 78630 }, { "grad_norm": 0.5659685134887695, "learning_rate": 1.1964671996776395e-05, "loss": 0.043, "step": 78640 }, { "grad_norm": 0.5399960875511169, "learning_rate": 1.1953941471384922e-05, "loss": 0.0443, "step": 78650 }, { "grad_norm": 0.5638548135757446, "learning_rate": 1.1943215106653205e-05, "loss": 0.0339, "step": 78660 }, { "grad_norm": 0.38648343086242676, "learning_rate": 1.193249290375429e-05, "loss": 0.0326, "step": 78670 }, { "grad_norm": 0.42279645800590515, "learning_rate": 1.1921774863860696e-05, "loss": 0.0354, "step": 78680 }, { "grad_norm": 0.5171492099761963, "learning_rate": 1.1911060988144556e-05, "loss": 0.0371, "step": 78690 }, { "grad_norm": 0.3774077296257019, "learning_rate": 1.1900351277777533e-05, "loss": 0.0305, "step": 78700 }, { "grad_norm": 0.5082809329032898, "learning_rate": 1.1889645733930798e-05, "loss": 0.0311, "step": 78710 }, { "grad_norm": 0.7067633867263794, "learning_rate": 1.1878944357775107e-05, "loss": 0.0316, "step": 78720 }, { "grad_norm": 1.0013864040374756, "learning_rate": 1.1868247150480749e-05, "loss": 0.0226, "step": 78730 }, { "grad_norm": 0.6842032670974731, "learning_rate": 1.1857554113217567e-05, "loss": 0.0297, "step": 78740 }, { "grad_norm": 0.3754989802837372, "learning_rate": 1.1846865247154903e-05, "loss": 0.0337, "step": 78750 }, { "grad_norm": 0.3587980568408966, "learning_rate": 1.1836180553461711e-05, "loss": 0.0318, "step": 78760 }, { "grad_norm": 0.4695170521736145, "learning_rate": 1.1825500033306419e-05, "loss": 0.0442, "step": 78770 }, { "grad_norm": 0.498220294713974, "learning_rate": 1.1814823687857052e-05, "loss": 0.0355, "step": 78780 }, { "grad_norm": 0.564440131187439, "learning_rate": 1.1804151518281153e-05, "loss": 0.0336, "step": 78790 }, { "grad_norm": 0.8423082232475281, "learning_rate": 1.179348352574583e-05, "loss": 0.0413, "step": 78800 }, { "grad_norm": 0.5296828150749207, "learning_rate": 1.1782819711417697e-05, "loss": 0.0337, "step": 78810 }, { "grad_norm": 0.35289230942726135, "learning_rate": 1.1772160076462951e-05, "loss": 0.0235, "step": 78820 }, { "grad_norm": 0.44161832332611084, "learning_rate": 1.1761504622047293e-05, "loss": 0.0371, "step": 78830 }, { "grad_norm": 0.3486612141132355, "learning_rate": 1.1750853349335999e-05, "loss": 0.0434, "step": 78840 }, { "grad_norm": 0.5427452325820923, "learning_rate": 1.1740206259493874e-05, "loss": 0.0368, "step": 78850 }, { "grad_norm": 0.33599674701690674, "learning_rate": 1.1729563353685286e-05, "loss": 0.0274, "step": 78860 }, { "grad_norm": 0.6855083703994751, "learning_rate": 1.1718924633074091e-05, "loss": 0.0328, "step": 78870 }, { "grad_norm": 0.4846790134906769, "learning_rate": 1.1708290098823765e-05, "loss": 0.0311, "step": 78880 }, { "grad_norm": 0.46744608879089355, "learning_rate": 1.1697659752097245e-05, "loss": 0.0315, "step": 78890 }, { "grad_norm": 0.46046194434165955, "learning_rate": 1.1687033594057068e-05, "loss": 0.0429, "step": 78900 }, { "grad_norm": 0.35940179228782654, "learning_rate": 1.1676411625865297e-05, "loss": 0.0392, "step": 78910 }, { "grad_norm": 0.5246855616569519, "learning_rate": 1.1665793848683542e-05, "loss": 0.0331, "step": 78920 }, { "grad_norm": 0.6162341237068176, "learning_rate": 1.1655180263672927e-05, "loss": 0.0469, "step": 78930 }, { "grad_norm": 0.4029148519039154, "learning_rate": 1.1644570871994142e-05, "loss": 0.0325, "step": 78940 }, { "grad_norm": 0.6119742393493652, "learning_rate": 1.163396567480744e-05, "loss": 0.03, "step": 78950 }, { "grad_norm": 0.4541960060596466, "learning_rate": 1.1623364673272552e-05, "loss": 0.0349, "step": 78960 }, { "grad_norm": 0.3293306231498718, "learning_rate": 1.1612767868548807e-05, "loss": 0.0305, "step": 78970 }, { "grad_norm": 0.46771901845932007, "learning_rate": 1.1602175261795057e-05, "loss": 0.032, "step": 78980 }, { "grad_norm": 0.4146507978439331, "learning_rate": 1.1591586854169696e-05, "loss": 0.0382, "step": 78990 }, { "grad_norm": 0.29760435223579407, "learning_rate": 1.158100264683064e-05, "loss": 0.0333, "step": 79000 }, { "grad_norm": 0.42827871441841125, "learning_rate": 1.1570422640935386e-05, "loss": 0.0378, "step": 79010 }, { "grad_norm": 0.431388258934021, "learning_rate": 1.155984683764092e-05, "loss": 0.04, "step": 79020 }, { "grad_norm": 0.7076418995857239, "learning_rate": 1.1549275238103812e-05, "loss": 0.0321, "step": 79030 }, { "grad_norm": 0.29599642753601074, "learning_rate": 1.1538707843480152e-05, "loss": 0.0336, "step": 79040 }, { "grad_norm": 0.3788738250732422, "learning_rate": 1.1528144654925593e-05, "loss": 0.0399, "step": 79050 }, { "grad_norm": 0.4011062681674957, "learning_rate": 1.1517585673595277e-05, "loss": 0.0291, "step": 79060 }, { "grad_norm": 0.32831940054893494, "learning_rate": 1.150703090064395e-05, "loss": 0.0357, "step": 79070 }, { "grad_norm": 0.32650649547576904, "learning_rate": 1.1496480337225835e-05, "loss": 0.0298, "step": 79080 }, { "grad_norm": 0.2866392135620117, "learning_rate": 1.1485933984494751e-05, "loss": 0.0263, "step": 79090 }, { "grad_norm": 0.5338444709777832, "learning_rate": 1.1475391843604e-05, "loss": 0.03, "step": 79100 }, { "grad_norm": 0.5776775479316711, "learning_rate": 1.1464853915706503e-05, "loss": 0.0384, "step": 79110 }, { "grad_norm": 0.4885120093822479, "learning_rate": 1.1454320201954626e-05, "loss": 0.0297, "step": 79120 }, { "grad_norm": 0.41432857513427734, "learning_rate": 1.1443790703500356e-05, "loss": 0.0341, "step": 79130 }, { "grad_norm": 0.3596750497817993, "learning_rate": 1.1433265421495154e-05, "loss": 0.0351, "step": 79140 }, { "grad_norm": 0.36974024772644043, "learning_rate": 1.1422744357090065e-05, "loss": 0.039, "step": 79150 }, { "grad_norm": 0.6647486686706543, "learning_rate": 1.1412227511435647e-05, "loss": 0.0374, "step": 79160 }, { "grad_norm": 0.3541578948497772, "learning_rate": 1.1401714885682025e-05, "loss": 0.0286, "step": 79170 }, { "grad_norm": 0.39646902680397034, "learning_rate": 1.1391206480978823e-05, "loss": 0.0255, "step": 79180 }, { "grad_norm": 0.2889341413974762, "learning_rate": 1.1380702298475227e-05, "loss": 0.0364, "step": 79190 }, { "grad_norm": 0.3698030114173889, "learning_rate": 1.1370202339319985e-05, "loss": 0.0289, "step": 79200 }, { "grad_norm": 0.35067218542099, "learning_rate": 1.1359706604661319e-05, "loss": 0.0462, "step": 79210 }, { "grad_norm": 0.572064220905304, "learning_rate": 1.1349215095647042e-05, "loss": 0.044, "step": 79220 }, { "grad_norm": 0.3900774419307709, "learning_rate": 1.1338727813424493e-05, "loss": 0.0264, "step": 79230 }, { "grad_norm": 0.30108433961868286, "learning_rate": 1.132824475914056e-05, "loss": 0.0282, "step": 79240 }, { "grad_norm": 0.5456246733665466, "learning_rate": 1.131776593394162e-05, "loss": 0.0259, "step": 79250 }, { "grad_norm": 0.7376134991645813, "learning_rate": 1.1307291338973657e-05, "loss": 0.0396, "step": 79260 }, { "grad_norm": 0.2307346612215042, "learning_rate": 1.1296820975382121e-05, "loss": 0.0322, "step": 79270 }, { "grad_norm": 0.2102033942937851, "learning_rate": 1.1286354844312074e-05, "loss": 0.0402, "step": 79280 }, { "grad_norm": 0.5201425552368164, "learning_rate": 1.1275892946908023e-05, "loss": 0.0348, "step": 79290 }, { "grad_norm": 0.6632148027420044, "learning_rate": 1.126543528431413e-05, "loss": 0.0471, "step": 79300 }, { "grad_norm": 0.5526479482650757, "learning_rate": 1.125498185767398e-05, "loss": 0.0297, "step": 79310 }, { "grad_norm": 0.2548527121543884, "learning_rate": 1.1244532668130781e-05, "loss": 0.0429, "step": 79320 }, { "grad_norm": 0.29552486538887024, "learning_rate": 1.1234087716827203e-05, "loss": 0.0341, "step": 79330 }, { "grad_norm": 0.5005876421928406, "learning_rate": 1.1223647004905524e-05, "loss": 0.0481, "step": 79340 }, { "grad_norm": 0.9385947585105896, "learning_rate": 1.1213210533507485e-05, "loss": 0.0471, "step": 79350 }, { "grad_norm": 0.32162022590637207, "learning_rate": 1.1202778303774447e-05, "loss": 0.0262, "step": 79360 }, { "grad_norm": 0.9411239624023438, "learning_rate": 1.1192350316847234e-05, "loss": 0.0332, "step": 79370 }, { "grad_norm": 0.5116127133369446, "learning_rate": 1.1181926573866258e-05, "loss": 0.0383, "step": 79380 }, { "grad_norm": 0.41177037358283997, "learning_rate": 1.1171507075971416e-05, "loss": 0.0414, "step": 79390 }, { "grad_norm": 0.2960212528705597, "learning_rate": 1.1161091824302177e-05, "loss": 0.0296, "step": 79400 }, { "grad_norm": 0.8129019737243652, "learning_rate": 1.115068081999755e-05, "loss": 0.0391, "step": 79410 }, { "grad_norm": 0.4931844174861908, "learning_rate": 1.1140274064196071e-05, "loss": 0.0438, "step": 79420 }, { "grad_norm": 0.36128267645835876, "learning_rate": 1.112987155803578e-05, "loss": 0.0305, "step": 79430 }, { "grad_norm": 0.8114184141159058, "learning_rate": 1.1119473302654298e-05, "loss": 0.0359, "step": 79440 }, { "grad_norm": 2.5111351013183594, "learning_rate": 1.1109079299188769e-05, "loss": 0.037, "step": 79450 }, { "grad_norm": 0.39203768968582153, "learning_rate": 1.1098689548775847e-05, "loss": 0.0375, "step": 79460 }, { "grad_norm": 0.5700613856315613, "learning_rate": 1.1088304052551762e-05, "loss": 0.0324, "step": 79470 }, { "grad_norm": 0.3128742575645447, "learning_rate": 1.1077922811652215e-05, "loss": 0.0248, "step": 79480 }, { "grad_norm": 0.6524490714073181, "learning_rate": 1.1067545827212534e-05, "loss": 0.0303, "step": 79490 }, { "grad_norm": 0.8123583197593689, "learning_rate": 1.1057173100367496e-05, "loss": 0.0285, "step": 79500 }, { "grad_norm": 0.5137064456939697, "learning_rate": 1.1046804632251474e-05, "loss": 0.0322, "step": 79510 }, { "grad_norm": 0.8580268621444702, "learning_rate": 1.1036440423998318e-05, "loss": 0.0244, "step": 79520 }, { "grad_norm": 0.6150388717651367, "learning_rate": 1.1026080476741469e-05, "loss": 0.0345, "step": 79530 }, { "grad_norm": 0.4149690568447113, "learning_rate": 1.1015724791613836e-05, "loss": 0.0257, "step": 79540 }, { "grad_norm": 1.3356636762619019, "learning_rate": 1.1005373369747951e-05, "loss": 0.0434, "step": 79550 }, { "grad_norm": 0.25797200202941895, "learning_rate": 1.09950262122758e-05, "loss": 0.0277, "step": 79560 }, { "grad_norm": 0.5605296492576599, "learning_rate": 1.0984683320328948e-05, "loss": 0.0313, "step": 79570 }, { "grad_norm": 0.2802084982395172, "learning_rate": 1.0974344695038458e-05, "loss": 0.031, "step": 79580 }, { "grad_norm": 0.34407177567481995, "learning_rate": 1.0964010337534968e-05, "loss": 0.0354, "step": 79590 }, { "grad_norm": 0.5548902153968811, "learning_rate": 1.0953680248948611e-05, "loss": 0.0367, "step": 79600 }, { "grad_norm": 0.3380110561847687, "learning_rate": 1.0943354430409069e-05, "loss": 0.0289, "step": 79610 }, { "grad_norm": 0.5874832272529602, "learning_rate": 1.0933032883045574e-05, "loss": 0.0299, "step": 79620 }, { "grad_norm": 0.5652338862419128, "learning_rate": 1.0922715607986872e-05, "loss": 0.0458, "step": 79630 }, { "grad_norm": 0.26995953917503357, "learning_rate": 1.0912402606361234e-05, "loss": 0.0313, "step": 79640 }, { "grad_norm": 0.3473871350288391, "learning_rate": 1.090209387929647e-05, "loss": 0.0221, "step": 79650 }, { "grad_norm": 0.5531457662582397, "learning_rate": 1.0891789427919957e-05, "loss": 0.0278, "step": 79660 }, { "grad_norm": 0.2611813247203827, "learning_rate": 1.088148925335854e-05, "loss": 0.025, "step": 79670 }, { "grad_norm": 0.48055821657180786, "learning_rate": 1.087119335673864e-05, "loss": 0.0385, "step": 79680 }, { "grad_norm": 0.38403722643852234, "learning_rate": 1.0860901739186208e-05, "loss": 0.0271, "step": 79690 }, { "grad_norm": 0.3192010521888733, "learning_rate": 1.085061440182673e-05, "loss": 0.0279, "step": 79700 }, { "grad_norm": 0.303718626499176, "learning_rate": 1.0840331345785193e-05, "loss": 0.0289, "step": 79710 }, { "grad_norm": 0.34651675820350647, "learning_rate": 1.083005257218615e-05, "loss": 0.0287, "step": 79720 }, { "grad_norm": 0.3949637711048126, "learning_rate": 1.0819778082153642e-05, "loss": 0.0268, "step": 79730 }, { "grad_norm": 0.361463338136673, "learning_rate": 1.0809507876811325e-05, "loss": 0.0351, "step": 79740 }, { "grad_norm": 0.2837567627429962, "learning_rate": 1.0799241957282292e-05, "loss": 0.0346, "step": 79750 }, { "grad_norm": 0.42971640825271606, "learning_rate": 1.078898032468923e-05, "loss": 0.0308, "step": 79760 }, { "grad_norm": 0.40272435545921326, "learning_rate": 1.0778722980154315e-05, "loss": 0.0333, "step": 79770 }, { "grad_norm": 0.3770044147968292, "learning_rate": 1.0768469924799302e-05, "loss": 0.0273, "step": 79780 }, { "grad_norm": 0.6009833216667175, "learning_rate": 1.0758221159745419e-05, "loss": 0.0396, "step": 79790 }, { "grad_norm": 0.5503982305526733, "learning_rate": 1.074797668611347e-05, "loss": 0.0287, "step": 79800 }, { "grad_norm": 0.2710034251213074, "learning_rate": 1.0737736505023777e-05, "loss": 0.0303, "step": 79810 }, { "grad_norm": 1.205475926399231, "learning_rate": 1.0727500617596203e-05, "loss": 0.0328, "step": 79820 }, { "grad_norm": 0.7882890701293945, "learning_rate": 1.0717269024950104e-05, "loss": 0.0298, "step": 79830 }, { "grad_norm": 0.34288281202316284, "learning_rate": 1.070704172820442e-05, "loss": 0.0354, "step": 79840 }, { "grad_norm": 0.25620198249816895, "learning_rate": 1.0696818728477559e-05, "loss": 0.0294, "step": 79850 }, { "grad_norm": 0.2410871982574463, "learning_rate": 1.0686600026887511e-05, "loss": 0.0303, "step": 79860 }, { "grad_norm": 0.505828320980072, "learning_rate": 1.0676385624551777e-05, "loss": 0.0374, "step": 79870 }, { "grad_norm": 0.6184489727020264, "learning_rate": 1.0666175522587402e-05, "loss": 0.0418, "step": 79880 }, { "grad_norm": 0.2861912250518799, "learning_rate": 1.0655969722110925e-05, "loss": 0.028, "step": 79890 }, { "grad_norm": 0.5905826091766357, "learning_rate": 1.0645768224238439e-05, "loss": 0.0293, "step": 79900 }, { "grad_norm": 0.4477255940437317, "learning_rate": 1.063557103008559e-05, "loss": 0.0315, "step": 79910 }, { "grad_norm": 0.23644348978996277, "learning_rate": 1.0625378140767494e-05, "loss": 0.0263, "step": 79920 }, { "grad_norm": 0.4517808258533478, "learning_rate": 1.0615189557398846e-05, "loss": 0.024, "step": 79930 }, { "grad_norm": 0.36116480827331543, "learning_rate": 1.060500528109386e-05, "loss": 0.0329, "step": 79940 }, { "grad_norm": 0.2530815899372101, "learning_rate": 1.0594825312966272e-05, "loss": 0.0307, "step": 79950 }, { "grad_norm": 0.3950920104980469, "learning_rate": 1.0584649654129336e-05, "loss": 0.0365, "step": 79960 }, { "grad_norm": 0.2938721477985382, "learning_rate": 1.0574478305695863e-05, "loss": 0.0188, "step": 79970 }, { "grad_norm": 0.37080124020576477, "learning_rate": 1.0564311268778154e-05, "loss": 0.0365, "step": 79980 }, { "grad_norm": 0.4548410177230835, "learning_rate": 1.0554148544488073e-05, "loss": 0.0323, "step": 79990 }, { "grad_norm": 0.7066830396652222, "learning_rate": 1.0543990133937004e-05, "loss": 0.0349, "step": 80000 } ], "logging_steps": 10, "max_steps": 100000, "num_input_tokens_seen": 0, "num_train_epochs": 9223372036854775807, "save_steps": 20000, "stateful_callbacks": { "TrainerControl": { "args": { "should_epoch_stop": false, "should_evaluate": false, "should_log": false, "should_save": true, "should_training_stop": false }, "attributes": {} } }, "total_flos": 0.0, "train_batch_size": 8, "trial_name": null, "trial_params": null }