{ "best_global_step": null, "best_metric": null, "best_model_checkpoint": null, "epoch": 1.0, "eval_steps": 500, "global_step": 100000, "is_hyper_param_search": false, "is_local_process_zero": true, "is_world_process_zero": true, "log_history": [ { "grad_norm": 1.0409797430038452, "learning_rate": 1.8e-07, "loss": 1.5859, "step": 10 }, { "grad_norm": 1.2581126689910889, "learning_rate": 3.8e-07, "loss": 1.5982, "step": 20 }, { "grad_norm": 1.123623251914978, "learning_rate": 5.8e-07, "loss": 1.5232, "step": 30 }, { "grad_norm": 1.387350082397461, "learning_rate": 7.8e-07, "loss": 1.5377, "step": 40 }, { "grad_norm": 1.086340308189392, "learning_rate": 9.8e-07, "loss": 1.5466, "step": 50 }, { "grad_norm": 1.1269241571426392, "learning_rate": 1.18e-06, "loss": 1.5195, "step": 60 }, { "grad_norm": 1.94477379322052, "learning_rate": 1.3800000000000001e-06, "loss": 1.4469, "step": 70 }, { "grad_norm": 1.9814035892486572, "learning_rate": 1.5800000000000003e-06, "loss": 1.3922, "step": 80 }, { "grad_norm": 1.3980494737625122, "learning_rate": 1.7800000000000001e-06, "loss": 1.2905, "step": 90 }, { "grad_norm": 1.7843081951141357, "learning_rate": 1.98e-06, "loss": 1.2706, "step": 100 }, { "grad_norm": 1.6025978326797485, "learning_rate": 2.1800000000000003e-06, "loss": 1.3049, "step": 110 }, { "grad_norm": 2.2889809608459473, "learning_rate": 2.38e-06, "loss": 1.2761, "step": 120 }, { "grad_norm": 1.8551321029663086, "learning_rate": 2.5800000000000003e-06, "loss": 1.1991, "step": 130 }, { "grad_norm": 1.8233875036239624, "learning_rate": 2.78e-06, "loss": 1.206, "step": 140 }, { "grad_norm": 1.8546581268310547, "learning_rate": 2.9800000000000003e-06, "loss": 1.138, "step": 150 }, { "grad_norm": 1.4078010320663452, "learning_rate": 3.1800000000000005e-06, "loss": 1.1453, "step": 160 }, { "grad_norm": 2.6628856658935547, "learning_rate": 3.38e-06, "loss": 1.1212, "step": 170 }, { "grad_norm": 1.5126729011535645, "learning_rate": 3.58e-06, "loss": 1.1239, "step": 180 }, { "grad_norm": 1.3946417570114136, "learning_rate": 3.7800000000000002e-06, "loss": 1.1193, "step": 190 }, { "grad_norm": 2.0726144313812256, "learning_rate": 3.98e-06, "loss": 1.1189, "step": 200 }, { "grad_norm": 1.638883113861084, "learning_rate": 4.18e-06, "loss": 1.1057, "step": 210 }, { "grad_norm": 1.5945985317230225, "learning_rate": 4.38e-06, "loss": 1.1181, "step": 220 }, { "grad_norm": 1.6846492290496826, "learning_rate": 4.58e-06, "loss": 1.1018, "step": 230 }, { "grad_norm": 2.0791494846343994, "learning_rate": 4.780000000000001e-06, "loss": 1.1188, "step": 240 }, { "grad_norm": 1.3614420890808105, "learning_rate": 4.98e-06, "loss": 1.1073, "step": 250 }, { "grad_norm": 2.409081220626831, "learning_rate": 5.18e-06, "loss": 1.0829, "step": 260 }, { "grad_norm": 2.072237253189087, "learning_rate": 5.38e-06, "loss": 1.0824, "step": 270 }, { "grad_norm": 1.6170815229415894, "learning_rate": 5.580000000000001e-06, "loss": 1.069, "step": 280 }, { "grad_norm": 1.9992694854736328, "learning_rate": 5.78e-06, "loss": 1.1109, "step": 290 }, { "grad_norm": 1.1730294227600098, "learning_rate": 5.98e-06, "loss": 1.0871, "step": 300 }, { "grad_norm": 1.6190649271011353, "learning_rate": 6.18e-06, "loss": 1.0976, "step": 310 }, { "grad_norm": 1.4961036443710327, "learning_rate": 6.38e-06, "loss": 1.0864, "step": 320 }, { "grad_norm": 1.4350250959396362, "learning_rate": 6.58e-06, "loss": 1.0786, "step": 330 }, { "grad_norm": 2.026902198791504, "learning_rate": 6.78e-06, "loss": 1.0793, "step": 340 }, { "grad_norm": 1.7435415983200073, "learning_rate": 6.98e-06, "loss": 1.0759, "step": 350 }, { "grad_norm": 1.5269943475723267, "learning_rate": 7.180000000000001e-06, "loss": 1.0868, "step": 360 }, { "grad_norm": 1.8267449140548706, "learning_rate": 7.3800000000000005e-06, "loss": 1.0706, "step": 370 }, { "grad_norm": 1.7098231315612793, "learning_rate": 7.580000000000001e-06, "loss": 1.0834, "step": 380 }, { "grad_norm": 1.608361005783081, "learning_rate": 7.78e-06, "loss": 1.0474, "step": 390 }, { "grad_norm": 1.8890362977981567, "learning_rate": 7.98e-06, "loss": 1.0687, "step": 400 }, { "grad_norm": 1.791141390800476, "learning_rate": 8.18e-06, "loss": 1.0465, "step": 410 }, { "grad_norm": 1.6148322820663452, "learning_rate": 8.380000000000001e-06, "loss": 1.0116, "step": 420 }, { "grad_norm": 2.2786214351654053, "learning_rate": 8.580000000000001e-06, "loss": 0.9833, "step": 430 }, { "grad_norm": 2.093506097793579, "learning_rate": 8.78e-06, "loss": 0.9915, "step": 440 }, { "grad_norm": 2.1449403762817383, "learning_rate": 8.98e-06, "loss": 0.9505, "step": 450 }, { "grad_norm": 2.5947985649108887, "learning_rate": 9.180000000000002e-06, "loss": 0.9396, "step": 460 }, { "grad_norm": 2.4674527645111084, "learning_rate": 9.38e-06, "loss": 0.9163, "step": 470 }, { "grad_norm": 3.6157233715057373, "learning_rate": 9.58e-06, "loss": 0.9145, "step": 480 }, { "grad_norm": 2.2148962020874023, "learning_rate": 9.78e-06, "loss": 0.8622, "step": 490 }, { "grad_norm": 2.216848611831665, "learning_rate": 9.980000000000001e-06, "loss": 0.8627, "step": 500 }, { "grad_norm": 2.869328737258911, "learning_rate": 1.018e-05, "loss": 0.8299, "step": 510 }, { "grad_norm": 2.4695138931274414, "learning_rate": 1.038e-05, "loss": 0.8004, "step": 520 }, { "grad_norm": 2.5674545764923096, "learning_rate": 1.058e-05, "loss": 0.7874, "step": 530 }, { "grad_norm": 2.7173240184783936, "learning_rate": 1.0780000000000002e-05, "loss": 0.7562, "step": 540 }, { "grad_norm": 2.5604238510131836, "learning_rate": 1.098e-05, "loss": 0.7483, "step": 550 }, { "grad_norm": 2.516826629638672, "learning_rate": 1.118e-05, "loss": 0.7426, "step": 560 }, { "grad_norm": 2.4411063194274902, "learning_rate": 1.1380000000000001e-05, "loss": 0.6778, "step": 570 }, { "grad_norm": 2.6552202701568604, "learning_rate": 1.1580000000000001e-05, "loss": 0.6446, "step": 580 }, { "grad_norm": 3.375182867050171, "learning_rate": 1.178e-05, "loss": 0.6382, "step": 590 }, { "grad_norm": 2.836557149887085, "learning_rate": 1.198e-05, "loss": 0.6127, "step": 600 }, { "grad_norm": 3.2284626960754395, "learning_rate": 1.2180000000000002e-05, "loss": 0.5914, "step": 610 }, { "grad_norm": 3.4015109539031982, "learning_rate": 1.238e-05, "loss": 0.5784, "step": 620 }, { "grad_norm": 2.8519301414489746, "learning_rate": 1.258e-05, "loss": 0.5693, "step": 630 }, { "grad_norm": 3.09889554977417, "learning_rate": 1.278e-05, "loss": 0.5369, "step": 640 }, { "grad_norm": 2.8688457012176514, "learning_rate": 1.2980000000000001e-05, "loss": 0.5156, "step": 650 }, { "grad_norm": 2.810347318649292, "learning_rate": 1.3180000000000001e-05, "loss": 0.4908, "step": 660 }, { "grad_norm": 2.695725202560425, "learning_rate": 1.338e-05, "loss": 0.4547, "step": 670 }, { "grad_norm": 3.6753761768341064, "learning_rate": 1.358e-05, "loss": 0.44, "step": 680 }, { "grad_norm": 3.490797519683838, "learning_rate": 1.3780000000000002e-05, "loss": 0.4552, "step": 690 }, { "grad_norm": 3.8565406799316406, "learning_rate": 1.3980000000000002e-05, "loss": 0.4703, "step": 700 }, { "grad_norm": 3.5381641387939453, "learning_rate": 1.4180000000000001e-05, "loss": 0.3846, "step": 710 }, { "grad_norm": 4.391994953155518, "learning_rate": 1.4380000000000001e-05, "loss": 0.4058, "step": 720 }, { "grad_norm": 3.9773595333099365, "learning_rate": 1.4580000000000003e-05, "loss": 0.3683, "step": 730 }, { "grad_norm": 3.0208687782287598, "learning_rate": 1.4779999999999999e-05, "loss": 0.3449, "step": 740 }, { "grad_norm": 3.7733609676361084, "learning_rate": 1.4979999999999999e-05, "loss": 0.3703, "step": 750 }, { "grad_norm": 3.314296245574951, "learning_rate": 1.518e-05, "loss": 0.3519, "step": 760 }, { "grad_norm": 2.823863983154297, "learning_rate": 1.538e-05, "loss": 0.346, "step": 770 }, { "grad_norm": 2.204200267791748, "learning_rate": 1.558e-05, "loss": 0.3546, "step": 780 }, { "grad_norm": 2.779160261154175, "learning_rate": 1.578e-05, "loss": 0.2901, "step": 790 }, { "grad_norm": 3.6979334354400635, "learning_rate": 1.598e-05, "loss": 0.29, "step": 800 }, { "grad_norm": 3.168806314468384, "learning_rate": 1.618e-05, "loss": 0.292, "step": 810 }, { "grad_norm": 2.7422420978546143, "learning_rate": 1.6380000000000002e-05, "loss": 0.2568, "step": 820 }, { "grad_norm": 2.9542760848999023, "learning_rate": 1.658e-05, "loss": 0.2765, "step": 830 }, { "grad_norm": 2.822622299194336, "learning_rate": 1.6780000000000002e-05, "loss": 0.2704, "step": 840 }, { "grad_norm": 2.677468776702881, "learning_rate": 1.698e-05, "loss": 0.2931, "step": 850 }, { "grad_norm": 3.2881972789764404, "learning_rate": 1.718e-05, "loss": 0.2409, "step": 860 }, { "grad_norm": 3.2362802028656006, "learning_rate": 1.7380000000000003e-05, "loss": 0.2747, "step": 870 }, { "grad_norm": 3.733360767364502, "learning_rate": 1.758e-05, "loss": 0.2543, "step": 880 }, { "grad_norm": 2.7417850494384766, "learning_rate": 1.7780000000000003e-05, "loss": 0.2453, "step": 890 }, { "grad_norm": 2.482605218887329, "learning_rate": 1.798e-05, "loss": 0.2435, "step": 900 }, { "grad_norm": 2.1488022804260254, "learning_rate": 1.818e-05, "loss": 0.2313, "step": 910 }, { "grad_norm": 2.02846360206604, "learning_rate": 1.838e-05, "loss": 0.2033, "step": 920 }, { "grad_norm": 2.6951394081115723, "learning_rate": 1.858e-05, "loss": 0.2191, "step": 930 }, { "grad_norm": 2.8674848079681396, "learning_rate": 1.878e-05, "loss": 0.2142, "step": 940 }, { "grad_norm": 2.8078970909118652, "learning_rate": 1.898e-05, "loss": 0.2094, "step": 950 }, { "grad_norm": 3.191218852996826, "learning_rate": 1.918e-05, "loss": 0.2344, "step": 960 }, { "grad_norm": 3.375678539276123, "learning_rate": 1.938e-05, "loss": 0.2439, "step": 970 }, { "grad_norm": 4.084616184234619, "learning_rate": 1.9580000000000002e-05, "loss": 0.2033, "step": 980 }, { "grad_norm": 3.1654646396636963, "learning_rate": 1.978e-05, "loss": 0.1879, "step": 990 }, { "grad_norm": 3.5396740436553955, "learning_rate": 1.9980000000000002e-05, "loss": 0.1866, "step": 1000 }, { "grad_norm": 2.2192375659942627, "learning_rate": 2.0180000000000003e-05, "loss": 0.2107, "step": 1010 }, { "grad_norm": 2.6795547008514404, "learning_rate": 2.038e-05, "loss": 0.2338, "step": 1020 }, { "grad_norm": 2.766584873199463, "learning_rate": 2.0580000000000003e-05, "loss": 0.1926, "step": 1030 }, { "grad_norm": 2.4533185958862305, "learning_rate": 2.078e-05, "loss": 0.2126, "step": 1040 }, { "grad_norm": 2.2557358741760254, "learning_rate": 2.098e-05, "loss": 0.2076, "step": 1050 }, { "grad_norm": 2.5905797481536865, "learning_rate": 2.118e-05, "loss": 0.2183, "step": 1060 }, { "grad_norm": 1.9323234558105469, "learning_rate": 2.138e-05, "loss": 0.2169, "step": 1070 }, { "grad_norm": 2.952728033065796, "learning_rate": 2.158e-05, "loss": 0.2268, "step": 1080 }, { "grad_norm": 1.8289939165115356, "learning_rate": 2.178e-05, "loss": 0.2008, "step": 1090 }, { "grad_norm": 2.895409345626831, "learning_rate": 2.198e-05, "loss": 0.1635, "step": 1100 }, { "grad_norm": 3.5399982929229736, "learning_rate": 2.218e-05, "loss": 0.2474, "step": 1110 }, { "grad_norm": 2.3324544429779053, "learning_rate": 2.2380000000000003e-05, "loss": 0.1651, "step": 1120 }, { "grad_norm": 1.917380928993225, "learning_rate": 2.258e-05, "loss": 0.198, "step": 1130 }, { "grad_norm": 3.1082985401153564, "learning_rate": 2.2780000000000002e-05, "loss": 0.2456, "step": 1140 }, { "grad_norm": 2.4794952869415283, "learning_rate": 2.298e-05, "loss": 0.2336, "step": 1150 }, { "grad_norm": 2.105470657348633, "learning_rate": 2.318e-05, "loss": 0.2037, "step": 1160 }, { "grad_norm": 2.061863660812378, "learning_rate": 2.3380000000000003e-05, "loss": 0.1858, "step": 1170 }, { "grad_norm": 2.1308810710906982, "learning_rate": 2.358e-05, "loss": 0.1707, "step": 1180 }, { "grad_norm": 2.8331387042999268, "learning_rate": 2.3780000000000003e-05, "loss": 0.1812, "step": 1190 }, { "grad_norm": 4.058098793029785, "learning_rate": 2.398e-05, "loss": 0.1961, "step": 1200 }, { "grad_norm": 4.051468372344971, "learning_rate": 2.418e-05, "loss": 0.2104, "step": 1210 }, { "grad_norm": 2.0009255409240723, "learning_rate": 2.438e-05, "loss": 0.1664, "step": 1220 }, { "grad_norm": 3.0845742225646973, "learning_rate": 2.4580000000000002e-05, "loss": 0.1816, "step": 1230 }, { "grad_norm": 1.8098511695861816, "learning_rate": 2.478e-05, "loss": 0.1899, "step": 1240 }, { "grad_norm": 3.151280403137207, "learning_rate": 2.498e-05, "loss": 0.186, "step": 1250 }, { "grad_norm": 2.911571502685547, "learning_rate": 2.5180000000000003e-05, "loss": 0.1891, "step": 1260 }, { "grad_norm": 2.1267905235290527, "learning_rate": 2.5380000000000004e-05, "loss": 0.1527, "step": 1270 }, { "grad_norm": 2.1608242988586426, "learning_rate": 2.5580000000000002e-05, "loss": 0.1864, "step": 1280 }, { "grad_norm": 1.5136322975158691, "learning_rate": 2.5779999999999997e-05, "loss": 0.1818, "step": 1290 }, { "grad_norm": 2.153066635131836, "learning_rate": 2.598e-05, "loss": 0.1651, "step": 1300 }, { "grad_norm": 2.201401948928833, "learning_rate": 2.618e-05, "loss": 0.1833, "step": 1310 }, { "grad_norm": 1.891803503036499, "learning_rate": 2.6379999999999998e-05, "loss": 0.1724, "step": 1320 }, { "grad_norm": 1.8677809238433838, "learning_rate": 2.658e-05, "loss": 0.1541, "step": 1330 }, { "grad_norm": 2.256333351135254, "learning_rate": 2.678e-05, "loss": 0.1466, "step": 1340 }, { "grad_norm": 1.9018926620483398, "learning_rate": 2.698e-05, "loss": 0.1888, "step": 1350 }, { "grad_norm": 2.4366331100463867, "learning_rate": 2.718e-05, "loss": 0.1668, "step": 1360 }, { "grad_norm": 3.1263256072998047, "learning_rate": 2.738e-05, "loss": 0.1592, "step": 1370 }, { "grad_norm": 2.187397003173828, "learning_rate": 2.758e-05, "loss": 0.1541, "step": 1380 }, { "grad_norm": 1.8762458562850952, "learning_rate": 2.778e-05, "loss": 0.1582, "step": 1390 }, { "grad_norm": 2.1082088947296143, "learning_rate": 2.798e-05, "loss": 0.1526, "step": 1400 }, { "grad_norm": 2.1983935832977295, "learning_rate": 2.818e-05, "loss": 0.1786, "step": 1410 }, { "grad_norm": 2.5814335346221924, "learning_rate": 2.8380000000000003e-05, "loss": 0.1942, "step": 1420 }, { "grad_norm": 1.4363963603973389, "learning_rate": 2.858e-05, "loss": 0.1372, "step": 1430 }, { "grad_norm": 2.031496047973633, "learning_rate": 2.8780000000000002e-05, "loss": 0.1739, "step": 1440 }, { "grad_norm": 1.9515199661254883, "learning_rate": 2.898e-05, "loss": 0.1788, "step": 1450 }, { "grad_norm": 1.6951018571853638, "learning_rate": 2.9180000000000002e-05, "loss": 0.1876, "step": 1460 }, { "grad_norm": 1.6178525686264038, "learning_rate": 2.9380000000000003e-05, "loss": 0.1714, "step": 1470 }, { "grad_norm": 1.5408958196640015, "learning_rate": 2.958e-05, "loss": 0.1472, "step": 1480 }, { "grad_norm": 1.828403115272522, "learning_rate": 2.9780000000000003e-05, "loss": 0.1539, "step": 1490 }, { "grad_norm": 1.4637744426727295, "learning_rate": 2.998e-05, "loss": 0.1745, "step": 1500 }, { "grad_norm": 1.6254668235778809, "learning_rate": 3.0180000000000002e-05, "loss": 0.1596, "step": 1510 }, { "grad_norm": 2.6991453170776367, "learning_rate": 3.0380000000000004e-05, "loss": 0.179, "step": 1520 }, { "grad_norm": 1.515657901763916, "learning_rate": 3.058e-05, "loss": 0.1658, "step": 1530 }, { "grad_norm": 2.0672099590301514, "learning_rate": 3.078e-05, "loss": 0.1699, "step": 1540 }, { "grad_norm": 2.022157669067383, "learning_rate": 3.0980000000000005e-05, "loss": 0.156, "step": 1550 }, { "grad_norm": 1.9997282028198242, "learning_rate": 3.118e-05, "loss": 0.1781, "step": 1560 }, { "grad_norm": 1.8578122854232788, "learning_rate": 3.138e-05, "loss": 0.1547, "step": 1570 }, { "grad_norm": 1.5381712913513184, "learning_rate": 3.1580000000000006e-05, "loss": 0.1698, "step": 1580 }, { "grad_norm": 1.819140911102295, "learning_rate": 3.1780000000000004e-05, "loss": 0.1778, "step": 1590 }, { "grad_norm": 1.9173951148986816, "learning_rate": 3.198e-05, "loss": 0.1607, "step": 1600 }, { "grad_norm": 1.7527467012405396, "learning_rate": 3.218e-05, "loss": 0.1382, "step": 1610 }, { "grad_norm": 2.31669545173645, "learning_rate": 3.238e-05, "loss": 0.1775, "step": 1620 }, { "grad_norm": 2.3204727172851562, "learning_rate": 3.2579999999999996e-05, "loss": 0.2116, "step": 1630 }, { "grad_norm": 1.6373883485794067, "learning_rate": 3.278e-05, "loss": 0.1566, "step": 1640 }, { "grad_norm": 1.704634666442871, "learning_rate": 3.298e-05, "loss": 0.1632, "step": 1650 }, { "grad_norm": 1.757416009902954, "learning_rate": 3.318e-05, "loss": 0.154, "step": 1660 }, { "grad_norm": 1.2687855958938599, "learning_rate": 3.338e-05, "loss": 0.1363, "step": 1670 }, { "grad_norm": 1.9311895370483398, "learning_rate": 3.358e-05, "loss": 0.1422, "step": 1680 }, { "grad_norm": 1.7194287776947021, "learning_rate": 3.378e-05, "loss": 0.1631, "step": 1690 }, { "grad_norm": 1.2714215517044067, "learning_rate": 3.398e-05, "loss": 0.1486, "step": 1700 }, { "grad_norm": 2.0845141410827637, "learning_rate": 3.418e-05, "loss": 0.1782, "step": 1710 }, { "grad_norm": 1.5767265558242798, "learning_rate": 3.438e-05, "loss": 0.1592, "step": 1720 }, { "grad_norm": 1.7762397527694702, "learning_rate": 3.4580000000000004e-05, "loss": 0.1665, "step": 1730 }, { "grad_norm": 1.5961068868637085, "learning_rate": 3.478e-05, "loss": 0.1471, "step": 1740 }, { "grad_norm": 1.8134907484054565, "learning_rate": 3.498e-05, "loss": 0.1676, "step": 1750 }, { "grad_norm": 1.9930537939071655, "learning_rate": 3.518e-05, "loss": 0.1446, "step": 1760 }, { "grad_norm": 1.5948872566223145, "learning_rate": 3.5380000000000003e-05, "loss": 0.14, "step": 1770 }, { "grad_norm": 1.8279650211334229, "learning_rate": 3.558e-05, "loss": 0.1618, "step": 1780 }, { "grad_norm": 2.115396022796631, "learning_rate": 3.578e-05, "loss": 0.1475, "step": 1790 }, { "grad_norm": 1.7060236930847168, "learning_rate": 3.5980000000000004e-05, "loss": 0.1536, "step": 1800 }, { "grad_norm": 1.702520728111267, "learning_rate": 3.618e-05, "loss": 0.1417, "step": 1810 }, { "grad_norm": 2.1642611026763916, "learning_rate": 3.638e-05, "loss": 0.1655, "step": 1820 }, { "grad_norm": 2.389721155166626, "learning_rate": 3.6580000000000006e-05, "loss": 0.1679, "step": 1830 }, { "grad_norm": 1.6136870384216309, "learning_rate": 3.6780000000000004e-05, "loss": 0.1569, "step": 1840 }, { "grad_norm": 1.3947687149047852, "learning_rate": 3.698e-05, "loss": 0.1822, "step": 1850 }, { "grad_norm": 1.3196953535079956, "learning_rate": 3.7180000000000007e-05, "loss": 0.1729, "step": 1860 }, { "grad_norm": 1.639669418334961, "learning_rate": 3.7380000000000005e-05, "loss": 0.1408, "step": 1870 }, { "grad_norm": 1.4793139696121216, "learning_rate": 3.758e-05, "loss": 0.1226, "step": 1880 }, { "grad_norm": 1.7060543298721313, "learning_rate": 3.778000000000001e-05, "loss": 0.1599, "step": 1890 }, { "grad_norm": 1.3625072240829468, "learning_rate": 3.7980000000000006e-05, "loss": 0.1474, "step": 1900 }, { "grad_norm": 1.8639649152755737, "learning_rate": 3.818e-05, "loss": 0.17, "step": 1910 }, { "grad_norm": 1.717461347579956, "learning_rate": 3.838e-05, "loss": 0.1459, "step": 1920 }, { "grad_norm": 1.4495171308517456, "learning_rate": 3.858e-05, "loss": 0.1529, "step": 1930 }, { "grad_norm": 2.0248630046844482, "learning_rate": 3.878e-05, "loss": 0.1354, "step": 1940 }, { "grad_norm": 1.3999882936477661, "learning_rate": 3.898e-05, "loss": 0.1354, "step": 1950 }, { "grad_norm": 1.5817592144012451, "learning_rate": 3.918e-05, "loss": 0.184, "step": 1960 }, { "grad_norm": 1.437929391860962, "learning_rate": 3.938e-05, "loss": 0.1364, "step": 1970 }, { "grad_norm": 1.6564701795578003, "learning_rate": 3.958e-05, "loss": 0.1726, "step": 1980 }, { "grad_norm": 1.6909083127975464, "learning_rate": 3.978e-05, "loss": 0.1315, "step": 1990 }, { "grad_norm": 1.3210620880126953, "learning_rate": 3.998e-05, "loss": 0.1504, "step": 2000 }, { "grad_norm": 1.677528738975525, "learning_rate": 4.018e-05, "loss": 0.1781, "step": 2010 }, { "grad_norm": 1.6938269138336182, "learning_rate": 4.038e-05, "loss": 0.1551, "step": 2020 }, { "grad_norm": 1.7732024192810059, "learning_rate": 4.058e-05, "loss": 0.1438, "step": 2030 }, { "grad_norm": 1.4802769422531128, "learning_rate": 4.078e-05, "loss": 0.1661, "step": 2040 }, { "grad_norm": 1.584979772567749, "learning_rate": 4.0980000000000004e-05, "loss": 0.1584, "step": 2050 }, { "grad_norm": 1.5896259546279907, "learning_rate": 4.118e-05, "loss": 0.1359, "step": 2060 }, { "grad_norm": 1.1761624813079834, "learning_rate": 4.138e-05, "loss": 0.1363, "step": 2070 }, { "grad_norm": 1.4788926839828491, "learning_rate": 4.1580000000000005e-05, "loss": 0.1386, "step": 2080 }, { "grad_norm": 1.2323838472366333, "learning_rate": 4.178e-05, "loss": 0.138, "step": 2090 }, { "grad_norm": 1.4220490455627441, "learning_rate": 4.198e-05, "loss": 0.141, "step": 2100 }, { "grad_norm": 1.4242217540740967, "learning_rate": 4.2180000000000006e-05, "loss": 0.1624, "step": 2110 }, { "grad_norm": 1.2128511667251587, "learning_rate": 4.2380000000000004e-05, "loss": 0.1329, "step": 2120 }, { "grad_norm": 1.0216522216796875, "learning_rate": 4.258e-05, "loss": 0.1367, "step": 2130 }, { "grad_norm": 1.3386441469192505, "learning_rate": 4.278e-05, "loss": 0.114, "step": 2140 }, { "grad_norm": 1.1598548889160156, "learning_rate": 4.2980000000000005e-05, "loss": 0.155, "step": 2150 }, { "grad_norm": 1.1854842901229858, "learning_rate": 4.318e-05, "loss": 0.1439, "step": 2160 }, { "grad_norm": 1.179848313331604, "learning_rate": 4.338e-05, "loss": 0.1489, "step": 2170 }, { "grad_norm": 1.5630700588226318, "learning_rate": 4.3580000000000006e-05, "loss": 0.146, "step": 2180 }, { "grad_norm": 1.5950483083724976, "learning_rate": 4.3780000000000004e-05, "loss": 0.1342, "step": 2190 }, { "grad_norm": 1.6282683610916138, "learning_rate": 4.398e-05, "loss": 0.1521, "step": 2200 }, { "grad_norm": 1.0409212112426758, "learning_rate": 4.418000000000001e-05, "loss": 0.1453, "step": 2210 }, { "grad_norm": 1.7522090673446655, "learning_rate": 4.438e-05, "loss": 0.1479, "step": 2220 }, { "grad_norm": 1.4171158075332642, "learning_rate": 4.458e-05, "loss": 0.1501, "step": 2230 }, { "grad_norm": 1.3113714456558228, "learning_rate": 4.478e-05, "loss": 0.1437, "step": 2240 }, { "grad_norm": 0.9338251352310181, "learning_rate": 4.498e-05, "loss": 0.1628, "step": 2250 }, { "grad_norm": 1.2317249774932861, "learning_rate": 4.518e-05, "loss": 0.1206, "step": 2260 }, { "grad_norm": 1.232877492904663, "learning_rate": 4.538e-05, "loss": 0.1303, "step": 2270 }, { "grad_norm": 1.5092926025390625, "learning_rate": 4.558e-05, "loss": 0.1397, "step": 2280 }, { "grad_norm": 1.5509443283081055, "learning_rate": 4.578e-05, "loss": 0.1648, "step": 2290 }, { "grad_norm": 1.3405483961105347, "learning_rate": 4.5980000000000004e-05, "loss": 0.1349, "step": 2300 }, { "grad_norm": 1.1892043352127075, "learning_rate": 4.618e-05, "loss": 0.1343, "step": 2310 }, { "grad_norm": 1.1421058177947998, "learning_rate": 4.638e-05, "loss": 0.1212, "step": 2320 }, { "grad_norm": 1.0781387090682983, "learning_rate": 4.6580000000000005e-05, "loss": 0.1581, "step": 2330 }, { "grad_norm": 1.2033944129943848, "learning_rate": 4.678e-05, "loss": 0.1243, "step": 2340 }, { "grad_norm": 1.3762532472610474, "learning_rate": 4.698e-05, "loss": 0.136, "step": 2350 }, { "grad_norm": 1.34773850440979, "learning_rate": 4.718e-05, "loss": 0.1437, "step": 2360 }, { "grad_norm": 1.1764341592788696, "learning_rate": 4.7380000000000004e-05, "loss": 0.145, "step": 2370 }, { "grad_norm": 1.2888580560684204, "learning_rate": 4.758e-05, "loss": 0.155, "step": 2380 }, { "grad_norm": 1.1333457231521606, "learning_rate": 4.778e-05, "loss": 0.1523, "step": 2390 }, { "grad_norm": 1.1395070552825928, "learning_rate": 4.7980000000000005e-05, "loss": 0.1456, "step": 2400 }, { "grad_norm": 1.002339482307434, "learning_rate": 4.818e-05, "loss": 0.1405, "step": 2410 }, { "grad_norm": 1.3793359994888306, "learning_rate": 4.838e-05, "loss": 0.1857, "step": 2420 }, { "grad_norm": 1.4205079078674316, "learning_rate": 4.8580000000000006e-05, "loss": 0.1488, "step": 2430 }, { "grad_norm": 1.3942800760269165, "learning_rate": 4.8780000000000004e-05, "loss": 0.154, "step": 2440 }, { "grad_norm": 1.3395607471466064, "learning_rate": 4.898e-05, "loss": 0.1174, "step": 2450 }, { "grad_norm": 1.0530065298080444, "learning_rate": 4.918000000000001e-05, "loss": 0.1536, "step": 2460 }, { "grad_norm": 1.4438434839248657, "learning_rate": 4.9380000000000005e-05, "loss": 0.1163, "step": 2470 }, { "grad_norm": 1.1892894506454468, "learning_rate": 4.958e-05, "loss": 0.1579, "step": 2480 }, { "grad_norm": 0.9605605602264404, "learning_rate": 4.978e-05, "loss": 0.1194, "step": 2490 }, { "grad_norm": 1.119078516960144, "learning_rate": 4.9980000000000006e-05, "loss": 0.1317, "step": 2500 }, { "grad_norm": 1.3199249505996704, "learning_rate": 5.0180000000000004e-05, "loss": 0.1235, "step": 2510 }, { "grad_norm": 1.1611462831497192, "learning_rate": 5.038e-05, "loss": 0.1379, "step": 2520 }, { "grad_norm": 0.8956231474876404, "learning_rate": 5.058000000000001e-05, "loss": 0.1239, "step": 2530 }, { "grad_norm": 1.392775297164917, "learning_rate": 5.0780000000000005e-05, "loss": 0.1395, "step": 2540 }, { "grad_norm": 1.0435150861740112, "learning_rate": 5.098e-05, "loss": 0.1444, "step": 2550 }, { "grad_norm": 1.0223997831344604, "learning_rate": 5.118000000000001e-05, "loss": 0.1377, "step": 2560 }, { "grad_norm": 1.2214759588241577, "learning_rate": 5.1380000000000006e-05, "loss": 0.1545, "step": 2570 }, { "grad_norm": 0.9126704931259155, "learning_rate": 5.1580000000000004e-05, "loss": 0.1552, "step": 2580 }, { "grad_norm": 0.9494376182556152, "learning_rate": 5.178000000000001e-05, "loss": 0.1284, "step": 2590 }, { "grad_norm": 1.4317152500152588, "learning_rate": 5.198000000000001e-05, "loss": 0.1361, "step": 2600 }, { "grad_norm": 1.2845513820648193, "learning_rate": 5.2180000000000005e-05, "loss": 0.1595, "step": 2610 }, { "grad_norm": 1.7960975170135498, "learning_rate": 5.238000000000001e-05, "loss": 0.126, "step": 2620 }, { "grad_norm": 1.3249255418777466, "learning_rate": 5.258000000000001e-05, "loss": 0.1295, "step": 2630 }, { "grad_norm": 1.1109740734100342, "learning_rate": 5.2780000000000006e-05, "loss": 0.1499, "step": 2640 }, { "grad_norm": 1.0858473777770996, "learning_rate": 5.2980000000000004e-05, "loss": 0.1438, "step": 2650 }, { "grad_norm": 1.2582902908325195, "learning_rate": 5.318000000000001e-05, "loss": 0.1658, "step": 2660 }, { "grad_norm": 0.9868754148483276, "learning_rate": 5.338000000000001e-05, "loss": 0.1532, "step": 2670 }, { "grad_norm": 1.42300546169281, "learning_rate": 5.3580000000000005e-05, "loss": 0.1705, "step": 2680 }, { "grad_norm": 1.3300366401672363, "learning_rate": 5.378e-05, "loss": 0.1312, "step": 2690 }, { "grad_norm": 0.9872952103614807, "learning_rate": 5.3979999999999995e-05, "loss": 0.1215, "step": 2700 }, { "grad_norm": 1.105914831161499, "learning_rate": 5.418e-05, "loss": 0.1471, "step": 2710 }, { "grad_norm": 1.2466647624969482, "learning_rate": 5.438e-05, "loss": 0.1576, "step": 2720 }, { "grad_norm": 1.0920727252960205, "learning_rate": 5.4579999999999996e-05, "loss": 0.1193, "step": 2730 }, { "grad_norm": 0.8931707739830017, "learning_rate": 5.478e-05, "loss": 0.1278, "step": 2740 }, { "grad_norm": 1.0870468616485596, "learning_rate": 5.498e-05, "loss": 0.1487, "step": 2750 }, { "grad_norm": 1.2180553674697876, "learning_rate": 5.518e-05, "loss": 0.1311, "step": 2760 }, { "grad_norm": 1.487365484237671, "learning_rate": 5.538e-05, "loss": 0.1442, "step": 2770 }, { "grad_norm": 0.8971457481384277, "learning_rate": 5.558e-05, "loss": 0.1512, "step": 2780 }, { "grad_norm": 1.1680463552474976, "learning_rate": 5.578e-05, "loss": 0.1442, "step": 2790 }, { "grad_norm": 0.9288302659988403, "learning_rate": 5.5979999999999996e-05, "loss": 0.1283, "step": 2800 }, { "grad_norm": 1.0919569730758667, "learning_rate": 5.618e-05, "loss": 0.1635, "step": 2810 }, { "grad_norm": 1.1164939403533936, "learning_rate": 5.638e-05, "loss": 0.1432, "step": 2820 }, { "grad_norm": 1.050858497619629, "learning_rate": 5.658e-05, "loss": 0.1196, "step": 2830 }, { "grad_norm": 1.2516911029815674, "learning_rate": 5.678e-05, "loss": 0.1452, "step": 2840 }, { "grad_norm": 1.1280914545059204, "learning_rate": 5.698e-05, "loss": 0.1326, "step": 2850 }, { "grad_norm": 0.8427284359931946, "learning_rate": 5.718e-05, "loss": 0.1287, "step": 2860 }, { "grad_norm": 1.0429826974868774, "learning_rate": 5.738e-05, "loss": 0.1345, "step": 2870 }, { "grad_norm": 0.858830451965332, "learning_rate": 5.758e-05, "loss": 0.1181, "step": 2880 }, { "grad_norm": 0.7517180442810059, "learning_rate": 5.778e-05, "loss": 0.1226, "step": 2890 }, { "grad_norm": 0.998399555683136, "learning_rate": 5.7980000000000004e-05, "loss": 0.1385, "step": 2900 }, { "grad_norm": 1.3531509637832642, "learning_rate": 5.818e-05, "loss": 0.1162, "step": 2910 }, { "grad_norm": 0.9095817804336548, "learning_rate": 5.838e-05, "loss": 0.1306, "step": 2920 }, { "grad_norm": 1.0745009183883667, "learning_rate": 5.858e-05, "loss": 0.1082, "step": 2930 }, { "grad_norm": 0.699069082736969, "learning_rate": 5.878e-05, "loss": 0.1157, "step": 2940 }, { "grad_norm": 1.1698421239852905, "learning_rate": 5.898e-05, "loss": 0.1235, "step": 2950 }, { "grad_norm": 0.8541887998580933, "learning_rate": 5.918e-05, "loss": 0.1545, "step": 2960 }, { "grad_norm": 1.2646267414093018, "learning_rate": 5.9380000000000004e-05, "loss": 0.1289, "step": 2970 }, { "grad_norm": 1.0175632238388062, "learning_rate": 5.958e-05, "loss": 0.1451, "step": 2980 }, { "grad_norm": 0.9919573664665222, "learning_rate": 5.978e-05, "loss": 0.1482, "step": 2990 }, { "grad_norm": 0.7799710035324097, "learning_rate": 5.9980000000000005e-05, "loss": 0.1327, "step": 3000 }, { "grad_norm": 0.9207127094268799, "learning_rate": 6.018e-05, "loss": 0.1151, "step": 3010 }, { "grad_norm": 0.888704776763916, "learning_rate": 6.038e-05, "loss": 0.1309, "step": 3020 }, { "grad_norm": 1.4224638938903809, "learning_rate": 6.0580000000000006e-05, "loss": 0.1258, "step": 3030 }, { "grad_norm": 0.7095049023628235, "learning_rate": 6.0780000000000004e-05, "loss": 0.1251, "step": 3040 }, { "grad_norm": 0.9506592154502869, "learning_rate": 6.098e-05, "loss": 0.1307, "step": 3050 }, { "grad_norm": 1.1474393606185913, "learning_rate": 6.118000000000001e-05, "loss": 0.1435, "step": 3060 }, { "grad_norm": 0.9955610632896423, "learning_rate": 6.138e-05, "loss": 0.1035, "step": 3070 }, { "grad_norm": 1.1093096733093262, "learning_rate": 6.158e-05, "loss": 0.1353, "step": 3080 }, { "grad_norm": 1.1862120628356934, "learning_rate": 6.178000000000001e-05, "loss": 0.1469, "step": 3090 }, { "grad_norm": 1.1100635528564453, "learning_rate": 6.198e-05, "loss": 0.1153, "step": 3100 }, { "grad_norm": 1.0270193815231323, "learning_rate": 6.218e-05, "loss": 0.1178, "step": 3110 }, { "grad_norm": 1.0717405080795288, "learning_rate": 6.238000000000001e-05, "loss": 0.1147, "step": 3120 }, { "grad_norm": 1.1138994693756104, "learning_rate": 6.258e-05, "loss": 0.1445, "step": 3130 }, { "grad_norm": 1.3675731420516968, "learning_rate": 6.278e-05, "loss": 0.1442, "step": 3140 }, { "grad_norm": 1.1413592100143433, "learning_rate": 6.298000000000001e-05, "loss": 0.1343, "step": 3150 }, { "grad_norm": 1.125429630279541, "learning_rate": 6.318e-05, "loss": 0.14, "step": 3160 }, { "grad_norm": 0.9342488646507263, "learning_rate": 6.338e-05, "loss": 0.1196, "step": 3170 }, { "grad_norm": 0.9332614541053772, "learning_rate": 6.358000000000001e-05, "loss": 0.1454, "step": 3180 }, { "grad_norm": 0.7652671337127686, "learning_rate": 6.378e-05, "loss": 0.1383, "step": 3190 }, { "grad_norm": 0.7512655854225159, "learning_rate": 6.398000000000001e-05, "loss": 0.1191, "step": 3200 }, { "grad_norm": 0.9811708927154541, "learning_rate": 6.418000000000001e-05, "loss": 0.1209, "step": 3210 }, { "grad_norm": 1.2933292388916016, "learning_rate": 6.438e-05, "loss": 0.1613, "step": 3220 }, { "grad_norm": 1.0854133367538452, "learning_rate": 6.458000000000001e-05, "loss": 0.1455, "step": 3230 }, { "grad_norm": 1.2243791818618774, "learning_rate": 6.478000000000001e-05, "loss": 0.1233, "step": 3240 }, { "grad_norm": 0.9004319906234741, "learning_rate": 6.498e-05, "loss": 0.1347, "step": 3250 }, { "grad_norm": 1.1962859630584717, "learning_rate": 6.518000000000001e-05, "loss": 0.1397, "step": 3260 }, { "grad_norm": 1.3618019819259644, "learning_rate": 6.538000000000001e-05, "loss": 0.1356, "step": 3270 }, { "grad_norm": 0.7826442718505859, "learning_rate": 6.558e-05, "loss": 0.1236, "step": 3280 }, { "grad_norm": 0.844795823097229, "learning_rate": 6.578000000000001e-05, "loss": 0.1275, "step": 3290 }, { "grad_norm": 0.8141269683837891, "learning_rate": 6.598e-05, "loss": 0.1185, "step": 3300 }, { "grad_norm": 0.7901903986930847, "learning_rate": 6.618e-05, "loss": 0.128, "step": 3310 }, { "grad_norm": 0.8900551795959473, "learning_rate": 6.638e-05, "loss": 0.1332, "step": 3320 }, { "grad_norm": 0.8581743836402893, "learning_rate": 6.658e-05, "loss": 0.1068, "step": 3330 }, { "grad_norm": 0.9678794741630554, "learning_rate": 6.678e-05, "loss": 0.1148, "step": 3340 }, { "grad_norm": 0.8838121294975281, "learning_rate": 6.698e-05, "loss": 0.1309, "step": 3350 }, { "grad_norm": 1.0079309940338135, "learning_rate": 6.718e-05, "loss": 0.1203, "step": 3360 }, { "grad_norm": 0.7593451738357544, "learning_rate": 6.738e-05, "loss": 0.1201, "step": 3370 }, { "grad_norm": 0.8682191371917725, "learning_rate": 6.758e-05, "loss": 0.1607, "step": 3380 }, { "grad_norm": 0.8251176476478577, "learning_rate": 6.778e-05, "loss": 0.1236, "step": 3390 }, { "grad_norm": 0.808404803276062, "learning_rate": 6.798e-05, "loss": 0.1326, "step": 3400 }, { "grad_norm": 1.1677780151367188, "learning_rate": 6.818e-05, "loss": 0.1164, "step": 3410 }, { "grad_norm": 0.996174693107605, "learning_rate": 6.838e-05, "loss": 0.131, "step": 3420 }, { "grad_norm": 0.7846856117248535, "learning_rate": 6.858e-05, "loss": 0.1126, "step": 3430 }, { "grad_norm": 0.6765000224113464, "learning_rate": 6.878e-05, "loss": 0.1343, "step": 3440 }, { "grad_norm": 0.9771571755409241, "learning_rate": 6.898e-05, "loss": 0.1273, "step": 3450 }, { "grad_norm": 0.5897001028060913, "learning_rate": 6.918e-05, "loss": 0.1039, "step": 3460 }, { "grad_norm": 0.8901366591453552, "learning_rate": 6.938e-05, "loss": 0.1406, "step": 3470 }, { "grad_norm": 0.9256598353385925, "learning_rate": 6.958e-05, "loss": 0.122, "step": 3480 }, { "grad_norm": 0.9345203638076782, "learning_rate": 6.978e-05, "loss": 0.1196, "step": 3490 }, { "grad_norm": 0.8335051536560059, "learning_rate": 6.998e-05, "loss": 0.0971, "step": 3500 }, { "grad_norm": 0.8773075342178345, "learning_rate": 7.018e-05, "loss": 0.1246, "step": 3510 }, { "grad_norm": 1.1611131429672241, "learning_rate": 7.038e-05, "loss": 0.1467, "step": 3520 }, { "grad_norm": 1.1719077825546265, "learning_rate": 7.058e-05, "loss": 0.1292, "step": 3530 }, { "grad_norm": 0.773768424987793, "learning_rate": 7.078e-05, "loss": 0.1092, "step": 3540 }, { "grad_norm": 0.837352991104126, "learning_rate": 7.098e-05, "loss": 0.1249, "step": 3550 }, { "grad_norm": 0.7235493063926697, "learning_rate": 7.118e-05, "loss": 0.1121, "step": 3560 }, { "grad_norm": 0.7879590392112732, "learning_rate": 7.138e-05, "loss": 0.137, "step": 3570 }, { "grad_norm": 0.8851912617683411, "learning_rate": 7.158e-05, "loss": 0.1343, "step": 3580 }, { "grad_norm": 0.7738597989082336, "learning_rate": 7.178000000000001e-05, "loss": 0.1238, "step": 3590 }, { "grad_norm": 0.9601605534553528, "learning_rate": 7.198e-05, "loss": 0.1106, "step": 3600 }, { "grad_norm": 0.7873350381851196, "learning_rate": 7.218e-05, "loss": 0.1263, "step": 3610 }, { "grad_norm": 0.871830940246582, "learning_rate": 7.238000000000001e-05, "loss": 0.1069, "step": 3620 }, { "grad_norm": 0.8942457437515259, "learning_rate": 7.258e-05, "loss": 0.0963, "step": 3630 }, { "grad_norm": 0.5401086211204529, "learning_rate": 7.278e-05, "loss": 0.1062, "step": 3640 }, { "grad_norm": 0.9698415398597717, "learning_rate": 7.298000000000001e-05, "loss": 0.1297, "step": 3650 }, { "grad_norm": 0.8692188858985901, "learning_rate": 7.318e-05, "loss": 0.1133, "step": 3660 }, { "grad_norm": 1.2223514318466187, "learning_rate": 7.338e-05, "loss": 0.1458, "step": 3670 }, { "grad_norm": 0.8857375383377075, "learning_rate": 7.358000000000001e-05, "loss": 0.1271, "step": 3680 }, { "grad_norm": 0.8813068866729736, "learning_rate": 7.378e-05, "loss": 0.1187, "step": 3690 }, { "grad_norm": 0.792212724685669, "learning_rate": 7.398e-05, "loss": 0.1071, "step": 3700 }, { "grad_norm": 0.7801458835601807, "learning_rate": 7.418000000000001e-05, "loss": 0.1211, "step": 3710 }, { "grad_norm": 0.7711488604545593, "learning_rate": 7.438e-05, "loss": 0.1137, "step": 3720 }, { "grad_norm": 1.3023813962936401, "learning_rate": 7.458000000000001e-05, "loss": 0.1114, "step": 3730 }, { "grad_norm": 1.0517629384994507, "learning_rate": 7.478e-05, "loss": 0.1249, "step": 3740 }, { "grad_norm": 0.8610154986381531, "learning_rate": 7.498e-05, "loss": 0.1261, "step": 3750 }, { "grad_norm": 0.8242989182472229, "learning_rate": 7.518000000000001e-05, "loss": 0.1068, "step": 3760 }, { "grad_norm": 1.0035909414291382, "learning_rate": 7.538e-05, "loss": 0.1206, "step": 3770 }, { "grad_norm": 0.8375622630119324, "learning_rate": 7.558e-05, "loss": 0.1124, "step": 3780 }, { "grad_norm": 0.9396575093269348, "learning_rate": 7.578000000000001e-05, "loss": 0.1315, "step": 3790 }, { "grad_norm": 1.0522489547729492, "learning_rate": 7.598e-05, "loss": 0.1662, "step": 3800 }, { "grad_norm": 0.964529812335968, "learning_rate": 7.618e-05, "loss": 0.113, "step": 3810 }, { "grad_norm": 1.0355490446090698, "learning_rate": 7.638000000000001e-05, "loss": 0.1245, "step": 3820 }, { "grad_norm": 1.155468225479126, "learning_rate": 7.658e-05, "loss": 0.1494, "step": 3830 }, { "grad_norm": 0.8513456583023071, "learning_rate": 7.678000000000001e-05, "loss": 0.1182, "step": 3840 }, { "grad_norm": 0.8916687965393066, "learning_rate": 7.698000000000001e-05, "loss": 0.1347, "step": 3850 }, { "grad_norm": 1.2717084884643555, "learning_rate": 7.718e-05, "loss": 0.1071, "step": 3860 }, { "grad_norm": 0.9127318859100342, "learning_rate": 7.738000000000001e-05, "loss": 0.118, "step": 3870 }, { "grad_norm": 0.8846708536148071, "learning_rate": 7.758000000000001e-05, "loss": 0.1225, "step": 3880 }, { "grad_norm": 0.9770024418830872, "learning_rate": 7.778e-05, "loss": 0.1328, "step": 3890 }, { "grad_norm": 0.9459984302520752, "learning_rate": 7.798000000000001e-05, "loss": 0.1126, "step": 3900 }, { "grad_norm": 0.966062068939209, "learning_rate": 7.818000000000001e-05, "loss": 0.1157, "step": 3910 }, { "grad_norm": 1.0541677474975586, "learning_rate": 7.838e-05, "loss": 0.1342, "step": 3920 }, { "grad_norm": 0.913090169429779, "learning_rate": 7.858000000000001e-05, "loss": 0.1073, "step": 3930 }, { "grad_norm": 0.8808351755142212, "learning_rate": 7.878e-05, "loss": 0.131, "step": 3940 }, { "grad_norm": 0.866256833076477, "learning_rate": 7.897999999999999e-05, "loss": 0.1194, "step": 3950 }, { "grad_norm": 0.8899617791175842, "learning_rate": 7.918e-05, "loss": 0.1336, "step": 3960 }, { "grad_norm": 0.8435922265052795, "learning_rate": 7.938e-05, "loss": 0.1061, "step": 3970 }, { "grad_norm": 0.986028254032135, "learning_rate": 7.958e-05, "loss": 0.1322, "step": 3980 }, { "grad_norm": 1.176267147064209, "learning_rate": 7.978e-05, "loss": 0.1303, "step": 3990 }, { "grad_norm": 0.934461236000061, "learning_rate": 7.998e-05, "loss": 0.1098, "step": 4000 }, { "grad_norm": 1.0001006126403809, "learning_rate": 8.018e-05, "loss": 0.1243, "step": 4010 }, { "grad_norm": 0.7348408102989197, "learning_rate": 8.038e-05, "loss": 0.1047, "step": 4020 }, { "grad_norm": 1.0022592544555664, "learning_rate": 8.058e-05, "loss": 0.1405, "step": 4030 }, { "grad_norm": 0.9987984299659729, "learning_rate": 8.078e-05, "loss": 0.1214, "step": 4040 }, { "grad_norm": 0.6801201701164246, "learning_rate": 8.098e-05, "loss": 0.1197, "step": 4050 }, { "grad_norm": 0.7645912766456604, "learning_rate": 8.118e-05, "loss": 0.1203, "step": 4060 }, { "grad_norm": 0.8591047525405884, "learning_rate": 8.138e-05, "loss": 0.1089, "step": 4070 }, { "grad_norm": 0.7731987833976746, "learning_rate": 8.158e-05, "loss": 0.1267, "step": 4080 }, { "grad_norm": 0.7653926014900208, "learning_rate": 8.178e-05, "loss": 0.1192, "step": 4090 }, { "grad_norm": 0.8614107370376587, "learning_rate": 8.198e-05, "loss": 0.1184, "step": 4100 }, { "grad_norm": 0.7911069989204407, "learning_rate": 8.218e-05, "loss": 0.1345, "step": 4110 }, { "grad_norm": 1.0130165815353394, "learning_rate": 8.238000000000001e-05, "loss": 0.1335, "step": 4120 }, { "grad_norm": 0.837378740310669, "learning_rate": 8.258e-05, "loss": 0.1234, "step": 4130 }, { "grad_norm": 0.8571679592132568, "learning_rate": 8.278e-05, "loss": 0.1314, "step": 4140 }, { "grad_norm": 1.0416197776794434, "learning_rate": 8.298000000000001e-05, "loss": 0.1455, "step": 4150 }, { "grad_norm": 0.8762593865394592, "learning_rate": 8.318e-05, "loss": 0.1069, "step": 4160 }, { "grad_norm": 0.985849916934967, "learning_rate": 8.338e-05, "loss": 0.1141, "step": 4170 }, { "grad_norm": 0.6650620102882385, "learning_rate": 8.358e-05, "loss": 0.0991, "step": 4180 }, { "grad_norm": 1.446082353591919, "learning_rate": 8.378e-05, "loss": 0.1108, "step": 4190 }, { "grad_norm": 0.6917089819908142, "learning_rate": 8.398e-05, "loss": 0.1114, "step": 4200 }, { "grad_norm": 0.8735893368721008, "learning_rate": 8.418e-05, "loss": 0.1207, "step": 4210 }, { "grad_norm": 1.3010250329971313, "learning_rate": 8.438e-05, "loss": 0.1423, "step": 4220 }, { "grad_norm": 1.1511226892471313, "learning_rate": 8.458e-05, "loss": 0.1539, "step": 4230 }, { "grad_norm": 0.9624840021133423, "learning_rate": 8.478e-05, "loss": 0.1261, "step": 4240 }, { "grad_norm": 0.8107709884643555, "learning_rate": 8.498e-05, "loss": 0.1315, "step": 4250 }, { "grad_norm": 0.8740515112876892, "learning_rate": 8.518000000000001e-05, "loss": 0.1353, "step": 4260 }, { "grad_norm": 0.9263874292373657, "learning_rate": 8.538e-05, "loss": 0.1281, "step": 4270 }, { "grad_norm": 0.851646900177002, "learning_rate": 8.558e-05, "loss": 0.0999, "step": 4280 }, { "grad_norm": 1.0607541799545288, "learning_rate": 8.578000000000001e-05, "loss": 0.117, "step": 4290 }, { "grad_norm": 0.8446372151374817, "learning_rate": 8.598e-05, "loss": 0.1081, "step": 4300 }, { "grad_norm": 0.9707340002059937, "learning_rate": 8.618e-05, "loss": 0.1326, "step": 4310 }, { "grad_norm": 1.2111458778381348, "learning_rate": 8.638000000000001e-05, "loss": 0.1344, "step": 4320 }, { "grad_norm": 0.6848767399787903, "learning_rate": 8.658e-05, "loss": 0.1099, "step": 4330 }, { "grad_norm": 0.8778305649757385, "learning_rate": 8.678e-05, "loss": 0.1218, "step": 4340 }, { "grad_norm": 0.7798698544502258, "learning_rate": 8.698000000000001e-05, "loss": 0.1166, "step": 4350 }, { "grad_norm": 0.7149578928947449, "learning_rate": 8.718e-05, "loss": 0.1098, "step": 4360 }, { "grad_norm": 1.030838131904602, "learning_rate": 8.738000000000001e-05, "loss": 0.1355, "step": 4370 }, { "grad_norm": 0.6932359933853149, "learning_rate": 8.758000000000001e-05, "loss": 0.104, "step": 4380 }, { "grad_norm": 0.5896483659744263, "learning_rate": 8.778e-05, "loss": 0.107, "step": 4390 }, { "grad_norm": 0.9357265830039978, "learning_rate": 8.798000000000001e-05, "loss": 0.1455, "step": 4400 }, { "grad_norm": 0.6802281737327576, "learning_rate": 8.818000000000001e-05, "loss": 0.1166, "step": 4410 }, { "grad_norm": 0.6559223532676697, "learning_rate": 8.838e-05, "loss": 0.1251, "step": 4420 }, { "grad_norm": 0.8331760764122009, "learning_rate": 8.858000000000001e-05, "loss": 0.1062, "step": 4430 }, { "grad_norm": 0.8242935538291931, "learning_rate": 8.878000000000001e-05, "loss": 0.1145, "step": 4440 }, { "grad_norm": 0.8720226287841797, "learning_rate": 8.898e-05, "loss": 0.1394, "step": 4450 }, { "grad_norm": 0.6743223667144775, "learning_rate": 8.918000000000001e-05, "loss": 0.1184, "step": 4460 }, { "grad_norm": 0.8100442290306091, "learning_rate": 8.938e-05, "loss": 0.1276, "step": 4470 }, { "grad_norm": 0.8308416604995728, "learning_rate": 8.958e-05, "loss": 0.1427, "step": 4480 }, { "grad_norm": 0.7757903933525085, "learning_rate": 8.978000000000001e-05, "loss": 0.124, "step": 4490 }, { "grad_norm": 0.9709296822547913, "learning_rate": 8.998e-05, "loss": 0.1238, "step": 4500 }, { "grad_norm": 0.7524985671043396, "learning_rate": 9.018000000000001e-05, "loss": 0.1421, "step": 4510 }, { "grad_norm": 1.0150562524795532, "learning_rate": 9.038000000000001e-05, "loss": 0.134, "step": 4520 }, { "grad_norm": 0.8431866765022278, "learning_rate": 9.058e-05, "loss": 0.1259, "step": 4530 }, { "grad_norm": 1.0178818702697754, "learning_rate": 9.078000000000001e-05, "loss": 0.1335, "step": 4540 }, { "grad_norm": 0.7534567713737488, "learning_rate": 9.098000000000001e-05, "loss": 0.15, "step": 4550 }, { "grad_norm": 0.7105011940002441, "learning_rate": 9.118e-05, "loss": 0.1131, "step": 4560 }, { "grad_norm": 0.8267039656639099, "learning_rate": 9.138e-05, "loss": 0.1493, "step": 4570 }, { "grad_norm": 1.514513373374939, "learning_rate": 9.158e-05, "loss": 0.1511, "step": 4580 }, { "grad_norm": 0.9252427220344543, "learning_rate": 9.178e-05, "loss": 0.121, "step": 4590 }, { "grad_norm": 0.8906248211860657, "learning_rate": 9.198e-05, "loss": 0.1311, "step": 4600 }, { "grad_norm": 0.7958332300186157, "learning_rate": 9.218e-05, "loss": 0.1279, "step": 4610 }, { "grad_norm": 1.0623115301132202, "learning_rate": 9.238e-05, "loss": 0.1067, "step": 4620 }, { "grad_norm": 0.872349739074707, "learning_rate": 9.258e-05, "loss": 0.1328, "step": 4630 }, { "grad_norm": 0.8973082304000854, "learning_rate": 9.278e-05, "loss": 0.1261, "step": 4640 }, { "grad_norm": 0.8251144289970398, "learning_rate": 9.298e-05, "loss": 0.1164, "step": 4650 }, { "grad_norm": 0.7264768481254578, "learning_rate": 9.318e-05, "loss": 0.1008, "step": 4660 }, { "grad_norm": 1.0089842081069946, "learning_rate": 9.338e-05, "loss": 0.1361, "step": 4670 }, { "grad_norm": 0.8482510447502136, "learning_rate": 9.358e-05, "loss": 0.1299, "step": 4680 }, { "grad_norm": 0.6215714812278748, "learning_rate": 9.378e-05, "loss": 0.1435, "step": 4690 }, { "grad_norm": 0.6949478387832642, "learning_rate": 9.398e-05, "loss": 0.1162, "step": 4700 }, { "grad_norm": 0.5986081957817078, "learning_rate": 9.418e-05, "loss": 0.1282, "step": 4710 }, { "grad_norm": 0.8446493148803711, "learning_rate": 9.438e-05, "loss": 0.1013, "step": 4720 }, { "grad_norm": 0.6176357269287109, "learning_rate": 9.458e-05, "loss": 0.1303, "step": 4730 }, { "grad_norm": 0.8615811467170715, "learning_rate": 9.478e-05, "loss": 0.1221, "step": 4740 }, { "grad_norm": 0.8793482184410095, "learning_rate": 9.498e-05, "loss": 0.1084, "step": 4750 }, { "grad_norm": 0.7766879200935364, "learning_rate": 9.518000000000001e-05, "loss": 0.1281, "step": 4760 }, { "grad_norm": 1.0039618015289307, "learning_rate": 9.538e-05, "loss": 0.1312, "step": 4770 }, { "grad_norm": 0.8904247283935547, "learning_rate": 9.558e-05, "loss": 0.1461, "step": 4780 }, { "grad_norm": 0.7712023854255676, "learning_rate": 9.578000000000001e-05, "loss": 0.1102, "step": 4790 }, { "grad_norm": 0.7423299551010132, "learning_rate": 9.598e-05, "loss": 0.1175, "step": 4800 }, { "grad_norm": 0.7368655204772949, "learning_rate": 9.618e-05, "loss": 0.1075, "step": 4810 }, { "grad_norm": 0.7001516819000244, "learning_rate": 9.638000000000001e-05, "loss": 0.1212, "step": 4820 }, { "grad_norm": 0.7217234373092651, "learning_rate": 9.658e-05, "loss": 0.1231, "step": 4830 }, { "grad_norm": 0.8470243811607361, "learning_rate": 9.678e-05, "loss": 0.1287, "step": 4840 }, { "grad_norm": 1.1480213403701782, "learning_rate": 9.698000000000001e-05, "loss": 0.1281, "step": 4850 }, { "grad_norm": 0.7256749272346497, "learning_rate": 9.718e-05, "loss": 0.1081, "step": 4860 }, { "grad_norm": 0.7787673473358154, "learning_rate": 9.738e-05, "loss": 0.1151, "step": 4870 }, { "grad_norm": 1.2163617610931396, "learning_rate": 9.758000000000001e-05, "loss": 0.1196, "step": 4880 }, { "grad_norm": 0.9713156223297119, "learning_rate": 9.778e-05, "loss": 0.1536, "step": 4890 }, { "grad_norm": 0.860348641872406, "learning_rate": 9.798000000000001e-05, "loss": 0.1241, "step": 4900 }, { "grad_norm": 0.8230133056640625, "learning_rate": 9.818000000000001e-05, "loss": 0.1171, "step": 4910 }, { "grad_norm": 0.6944698691368103, "learning_rate": 9.838e-05, "loss": 0.1253, "step": 4920 }, { "grad_norm": 0.8537330031394958, "learning_rate": 9.858000000000001e-05, "loss": 0.1393, "step": 4930 }, { "grad_norm": 0.6717764735221863, "learning_rate": 9.878e-05, "loss": 0.1224, "step": 4940 }, { "grad_norm": 0.8075504899024963, "learning_rate": 9.898e-05, "loss": 0.1413, "step": 4950 }, { "grad_norm": 0.7662265300750732, "learning_rate": 9.918000000000001e-05, "loss": 0.1274, "step": 4960 }, { "grad_norm": 0.7764124274253845, "learning_rate": 9.938e-05, "loss": 0.1339, "step": 4970 }, { "grad_norm": 0.7260814309120178, "learning_rate": 9.958e-05, "loss": 0.1331, "step": 4980 }, { "grad_norm": 0.6102144718170166, "learning_rate": 9.978000000000001e-05, "loss": 0.1253, "step": 4990 }, { "grad_norm": 0.6114370226860046, "learning_rate": 9.998e-05, "loss": 0.135, "step": 5000 }, { "grad_norm": 0.7573909759521484, "learning_rate": 9.999999778549045e-05, "loss": 0.1299, "step": 5010 }, { "grad_norm": 0.7158921957015991, "learning_rate": 9.999999013039593e-05, "loss": 0.13, "step": 5020 }, { "grad_norm": 0.8888089656829834, "learning_rate": 9.999997700737766e-05, "loss": 0.128, "step": 5030 }, { "grad_norm": 0.7359429597854614, "learning_rate": 9.999995841643709e-05, "loss": 0.1078, "step": 5040 }, { "grad_norm": 0.8966362476348877, "learning_rate": 9.999993435757623e-05, "loss": 0.1275, "step": 5050 }, { "grad_norm": 1.2378089427947998, "learning_rate": 9.999990483079773e-05, "loss": 0.1119, "step": 5060 }, { "grad_norm": 0.7970719933509827, "learning_rate": 9.999986983610481e-05, "loss": 0.1515, "step": 5070 }, { "grad_norm": 1.2411422729492188, "learning_rate": 9.99998293735013e-05, "loss": 0.1359, "step": 5080 }, { "grad_norm": 1.3488653898239136, "learning_rate": 9.999978344299161e-05, "loss": 0.1271, "step": 5090 }, { "grad_norm": 1.3479377031326294, "learning_rate": 9.99997320445808e-05, "loss": 0.1401, "step": 5100 }, { "grad_norm": 1.0768215656280518, "learning_rate": 9.999967517827444e-05, "loss": 0.1441, "step": 5110 }, { "grad_norm": 0.9719208478927612, "learning_rate": 9.999961284407879e-05, "loss": 0.1531, "step": 5120 }, { "grad_norm": 0.9853704571723938, "learning_rate": 9.999954504200067e-05, "loss": 0.1252, "step": 5130 }, { "grad_norm": 0.9406614899635315, "learning_rate": 9.999947177204744e-05, "loss": 0.1168, "step": 5140 }, { "grad_norm": 0.8559218049049377, "learning_rate": 9.999939303422718e-05, "loss": 0.0953, "step": 5150 }, { "grad_norm": 0.8365216851234436, "learning_rate": 9.999930882854847e-05, "loss": 0.1229, "step": 5160 }, { "grad_norm": 0.6910331845283508, "learning_rate": 9.999921915502051e-05, "loss": 0.126, "step": 5170 }, { "grad_norm": 0.7635325193405151, "learning_rate": 9.99991240136531e-05, "loss": 0.1303, "step": 5180 }, { "grad_norm": 0.7356813549995422, "learning_rate": 9.999902340445668e-05, "loss": 0.1543, "step": 5190 }, { "grad_norm": 0.7008518576622009, "learning_rate": 9.999891732744224e-05, "loss": 0.1086, "step": 5200 }, { "grad_norm": 0.7146685719490051, "learning_rate": 9.999880578262135e-05, "loss": 0.1164, "step": 5210 }, { "grad_norm": 0.768912136554718, "learning_rate": 9.999868877000624e-05, "loss": 0.124, "step": 5220 }, { "grad_norm": 0.6184179782867432, "learning_rate": 9.99985662896097e-05, "loss": 0.1121, "step": 5230 }, { "grad_norm": 0.9357815384864807, "learning_rate": 9.999843834144513e-05, "loss": 0.1225, "step": 5240 }, { "grad_norm": 0.8588082194328308, "learning_rate": 9.99983049255265e-05, "loss": 0.1451, "step": 5250 }, { "grad_norm": 0.9631198048591614, "learning_rate": 9.999816604186843e-05, "loss": 0.1365, "step": 5260 }, { "grad_norm": 0.7046023607254028, "learning_rate": 9.999802169048609e-05, "loss": 0.1153, "step": 5270 }, { "grad_norm": 0.7255672216415405, "learning_rate": 9.999787187139527e-05, "loss": 0.1076, "step": 5280 }, { "grad_norm": 0.7476485371589661, "learning_rate": 9.999771658461234e-05, "loss": 0.1169, "step": 5290 }, { "grad_norm": 1.049973487854004, "learning_rate": 9.999755583015431e-05, "loss": 0.1047, "step": 5300 }, { "grad_norm": 1.0252803564071655, "learning_rate": 9.999738960803874e-05, "loss": 0.1059, "step": 5310 }, { "grad_norm": 0.7473459243774414, "learning_rate": 9.99972179182838e-05, "loss": 0.1248, "step": 5320 }, { "grad_norm": 0.9531101584434509, "learning_rate": 9.99970407609083e-05, "loss": 0.1481, "step": 5330 }, { "grad_norm": 0.838178813457489, "learning_rate": 9.999685813593159e-05, "loss": 0.114, "step": 5340 }, { "grad_norm": 0.5288892388343811, "learning_rate": 9.999667004337362e-05, "loss": 0.1068, "step": 5350 }, { "grad_norm": 1.47739577293396, "learning_rate": 9.9996476483255e-05, "loss": 0.1415, "step": 5360 }, { "grad_norm": 0.7745621204376221, "learning_rate": 9.999627745559688e-05, "loss": 0.1152, "step": 5370 }, { "grad_norm": 0.7435718774795532, "learning_rate": 9.999607296042101e-05, "loss": 0.1117, "step": 5380 }, { "grad_norm": 0.8113773465156555, "learning_rate": 9.99958629977498e-05, "loss": 0.1059, "step": 5390 }, { "grad_norm": 0.743013322353363, "learning_rate": 9.999564756760615e-05, "loss": 0.1068, "step": 5400 }, { "grad_norm": 0.7056658864021301, "learning_rate": 9.999542667001366e-05, "loss": 0.0962, "step": 5410 }, { "grad_norm": 0.6076172590255737, "learning_rate": 9.999520030499647e-05, "loss": 0.1139, "step": 5420 }, { "grad_norm": 0.7764258980751038, "learning_rate": 9.999496847257936e-05, "loss": 0.0944, "step": 5430 }, { "grad_norm": 0.8192424178123474, "learning_rate": 9.999473117278764e-05, "loss": 0.1393, "step": 5440 }, { "grad_norm": 0.8556610941886902, "learning_rate": 9.999448840564731e-05, "loss": 0.0971, "step": 5450 }, { "grad_norm": 0.8086387515068054, "learning_rate": 9.999424017118488e-05, "loss": 0.1188, "step": 5460 }, { "grad_norm": 0.734250009059906, "learning_rate": 9.999398646942751e-05, "loss": 0.1041, "step": 5470 }, { "grad_norm": 0.7158524394035339, "learning_rate": 9.999372730040296e-05, "loss": 0.1018, "step": 5480 }, { "grad_norm": 0.6550337076187134, "learning_rate": 9.999346266413953e-05, "loss": 0.1288, "step": 5490 }, { "grad_norm": 0.8166854381561279, "learning_rate": 9.99931925606662e-05, "loss": 0.126, "step": 5500 }, { "grad_norm": 0.5734931230545044, "learning_rate": 9.99929169900125e-05, "loss": 0.1178, "step": 5510 }, { "grad_norm": 0.8101576566696167, "learning_rate": 9.999263595220855e-05, "loss": 0.105, "step": 5520 }, { "grad_norm": 0.965054988861084, "learning_rate": 9.99923494472851e-05, "loss": 0.1067, "step": 5530 }, { "grad_norm": 0.6408472657203674, "learning_rate": 9.999205747527348e-05, "loss": 0.112, "step": 5540 }, { "grad_norm": 0.6902292370796204, "learning_rate": 9.999176003620561e-05, "loss": 0.1175, "step": 5550 }, { "grad_norm": 0.6110274195671082, "learning_rate": 9.999145713011405e-05, "loss": 0.1137, "step": 5560 }, { "grad_norm": 0.8552706241607666, "learning_rate": 9.999114875703186e-05, "loss": 0.1267, "step": 5570 }, { "grad_norm": 0.9000657200813293, "learning_rate": 9.999083491699281e-05, "loss": 0.1299, "step": 5580 }, { "grad_norm": 0.8640192747116089, "learning_rate": 9.999051561003123e-05, "loss": 0.1303, "step": 5590 }, { "grad_norm": 0.8957227468490601, "learning_rate": 9.999019083618202e-05, "loss": 0.1311, "step": 5600 }, { "grad_norm": 1.0016535520553589, "learning_rate": 9.99898605954807e-05, "loss": 0.1131, "step": 5610 }, { "grad_norm": 0.9602082371711731, "learning_rate": 9.998952488796338e-05, "loss": 0.1049, "step": 5620 }, { "grad_norm": 0.8423422574996948, "learning_rate": 9.998918371366676e-05, "loss": 0.1061, "step": 5630 }, { "grad_norm": 1.093881368637085, "learning_rate": 9.99888370726282e-05, "loss": 0.1193, "step": 5640 }, { "grad_norm": 0.7913331389427185, "learning_rate": 9.998848496488556e-05, "loss": 0.1472, "step": 5650 }, { "grad_norm": 0.8573336005210876, "learning_rate": 9.998812739047736e-05, "loss": 0.106, "step": 5660 }, { "grad_norm": 0.7752903699874878, "learning_rate": 9.99877643494427e-05, "loss": 0.1169, "step": 5670 }, { "grad_norm": 0.8122528791427612, "learning_rate": 9.998739584182128e-05, "loss": 0.1302, "step": 5680 }, { "grad_norm": 1.0640912055969238, "learning_rate": 9.998702186765342e-05, "loss": 0.1068, "step": 5690 }, { "grad_norm": 0.5642983913421631, "learning_rate": 9.998664242698e-05, "loss": 0.1032, "step": 5700 }, { "grad_norm": 0.798561155796051, "learning_rate": 9.998625751984251e-05, "loss": 0.1125, "step": 5710 }, { "grad_norm": 0.6605877876281738, "learning_rate": 9.998586714628307e-05, "loss": 0.1037, "step": 5720 }, { "grad_norm": 0.6862432360649109, "learning_rate": 9.998547130634432e-05, "loss": 0.1263, "step": 5730 }, { "grad_norm": 0.5263802409172058, "learning_rate": 9.99850700000696e-05, "loss": 0.133, "step": 5740 }, { "grad_norm": 0.6787314414978027, "learning_rate": 9.998466322750278e-05, "loss": 0.1147, "step": 5750 }, { "grad_norm": 0.6793038845062256, "learning_rate": 9.998425098868834e-05, "loss": 0.1085, "step": 5760 }, { "grad_norm": 1.062788963317871, "learning_rate": 9.998383328367136e-05, "loss": 0.1306, "step": 5770 }, { "grad_norm": 0.8745782971382141, "learning_rate": 9.99834101124975e-05, "loss": 0.1315, "step": 5780 }, { "grad_norm": 0.731928825378418, "learning_rate": 9.998298147521309e-05, "loss": 0.1114, "step": 5790 }, { "grad_norm": 1.1528372764587402, "learning_rate": 9.998254737186496e-05, "loss": 0.1376, "step": 5800 }, { "grad_norm": 1.2470701932907104, "learning_rate": 9.99821078025006e-05, "loss": 0.1101, "step": 5810 }, { "grad_norm": 0.7445257306098938, "learning_rate": 9.998166276716807e-05, "loss": 0.1108, "step": 5820 }, { "grad_norm": 0.7191401720046997, "learning_rate": 9.998121226591606e-05, "loss": 0.1106, "step": 5830 }, { "grad_norm": 0.8718814253807068, "learning_rate": 9.998075629879382e-05, "loss": 0.1117, "step": 5840 }, { "grad_norm": 0.8172667026519775, "learning_rate": 9.99802948658512e-05, "loss": 0.0954, "step": 5850 }, { "grad_norm": 0.7819321751594543, "learning_rate": 9.99798279671387e-05, "loss": 0.1338, "step": 5860 }, { "grad_norm": 0.7331072092056274, "learning_rate": 9.997935560270734e-05, "loss": 0.1087, "step": 5870 }, { "grad_norm": 0.7412842512130737, "learning_rate": 9.997887777260879e-05, "loss": 0.121, "step": 5880 }, { "grad_norm": 0.739230215549469, "learning_rate": 9.997839447689532e-05, "loss": 0.1252, "step": 5890 }, { "grad_norm": 0.5789482593536377, "learning_rate": 9.997790571561978e-05, "loss": 0.1276, "step": 5900 }, { "grad_norm": 0.7792046070098877, "learning_rate": 9.99774114888356e-05, "loss": 0.1363, "step": 5910 }, { "grad_norm": 0.6896626949310303, "learning_rate": 9.997691179659684e-05, "loss": 0.1321, "step": 5920 }, { "grad_norm": 1.0430337190628052, "learning_rate": 9.997640663895815e-05, "loss": 0.1133, "step": 5930 }, { "grad_norm": 0.7786833047866821, "learning_rate": 9.997589601597477e-05, "loss": 0.1217, "step": 5940 }, { "grad_norm": 0.8281332850456238, "learning_rate": 9.997537992770252e-05, "loss": 0.1087, "step": 5950 }, { "grad_norm": 0.981611967086792, "learning_rate": 9.997485837419788e-05, "loss": 0.1211, "step": 5960 }, { "grad_norm": 0.8540297746658325, "learning_rate": 9.997433135551786e-05, "loss": 0.1224, "step": 5970 }, { "grad_norm": 0.8553882837295532, "learning_rate": 9.997379887172009e-05, "loss": 0.1125, "step": 5980 }, { "grad_norm": 0.9102415442466736, "learning_rate": 9.997326092286281e-05, "loss": 0.1321, "step": 5990 }, { "grad_norm": 0.6817457675933838, "learning_rate": 9.997271750900486e-05, "loss": 0.1516, "step": 6000 }, { "grad_norm": 0.7803475260734558, "learning_rate": 9.997216863020565e-05, "loss": 0.1438, "step": 6010 }, { "grad_norm": 0.7019057273864746, "learning_rate": 9.99716142865252e-05, "loss": 0.1382, "step": 6020 }, { "grad_norm": 0.7616969347000122, "learning_rate": 9.997105447802415e-05, "loss": 0.1238, "step": 6030 }, { "grad_norm": 0.7439306974411011, "learning_rate": 9.997048920476373e-05, "loss": 0.0995, "step": 6040 }, { "grad_norm": 0.8937216401100159, "learning_rate": 9.996991846680572e-05, "loss": 0.126, "step": 6050 }, { "grad_norm": 0.6791895031929016, "learning_rate": 9.996934226421257e-05, "loss": 0.1205, "step": 6060 }, { "grad_norm": 0.8414584994316101, "learning_rate": 9.996876059704726e-05, "loss": 0.1289, "step": 6070 }, { "grad_norm": 0.7037203311920166, "learning_rate": 9.996817346537343e-05, "loss": 0.1331, "step": 6080 }, { "grad_norm": 0.7345109581947327, "learning_rate": 9.996758086925526e-05, "loss": 0.1246, "step": 6090 }, { "grad_norm": 0.6681037545204163, "learning_rate": 9.996698280875759e-05, "loss": 0.1296, "step": 6100 }, { "grad_norm": 0.7232068181037903, "learning_rate": 9.99663792839458e-05, "loss": 0.0969, "step": 6110 }, { "grad_norm": 0.7821516394615173, "learning_rate": 9.99657702948859e-05, "loss": 0.119, "step": 6120 }, { "grad_norm": 0.7185366153717041, "learning_rate": 9.996515584164448e-05, "loss": 0.1061, "step": 6130 }, { "grad_norm": 0.7849397659301758, "learning_rate": 9.996453592428873e-05, "loss": 0.1158, "step": 6140 }, { "grad_norm": 0.8559790849685669, "learning_rate": 9.996391054288646e-05, "loss": 0.1473, "step": 6150 }, { "grad_norm": 0.6878917813301086, "learning_rate": 9.996327969750605e-05, "loss": 0.1393, "step": 6160 }, { "grad_norm": 0.7820389866828918, "learning_rate": 9.996264338821649e-05, "loss": 0.1184, "step": 6170 }, { "grad_norm": 0.6461158990859985, "learning_rate": 9.996200161508735e-05, "loss": 0.1082, "step": 6180 }, { "grad_norm": 0.9175824522972107, "learning_rate": 9.996135437818885e-05, "loss": 0.1306, "step": 6190 }, { "grad_norm": 0.5364166498184204, "learning_rate": 9.996070167759175e-05, "loss": 0.1051, "step": 6200 }, { "grad_norm": 0.8251135349273682, "learning_rate": 9.996004351336743e-05, "loss": 0.128, "step": 6210 }, { "grad_norm": 0.9957047700881958, "learning_rate": 9.995937988558785e-05, "loss": 0.0938, "step": 6220 }, { "grad_norm": 0.8324277400970459, "learning_rate": 9.995871079432561e-05, "loss": 0.119, "step": 6230 }, { "grad_norm": 0.8229281306266785, "learning_rate": 9.995803623965389e-05, "loss": 0.1345, "step": 6240 }, { "grad_norm": 0.6054319143295288, "learning_rate": 9.995735622164641e-05, "loss": 0.1158, "step": 6250 }, { "grad_norm": 0.6926694512367249, "learning_rate": 9.995667074037758e-05, "loss": 0.094, "step": 6260 }, { "grad_norm": 0.9311651587486267, "learning_rate": 9.995597979592232e-05, "loss": 0.1082, "step": 6270 }, { "grad_norm": 0.8369887471199036, "learning_rate": 9.995528338835625e-05, "loss": 0.1209, "step": 6280 }, { "grad_norm": 0.757278323173523, "learning_rate": 9.995458151775547e-05, "loss": 0.1046, "step": 6290 }, { "grad_norm": 0.6920607686042786, "learning_rate": 9.995387418419677e-05, "loss": 0.1026, "step": 6300 }, { "grad_norm": 0.6975706815719604, "learning_rate": 9.99531613877575e-05, "loss": 0.1, "step": 6310 }, { "grad_norm": 0.4872110188007355, "learning_rate": 9.995244312851559e-05, "loss": 0.0979, "step": 6320 }, { "grad_norm": 0.7396378517150879, "learning_rate": 9.995171940654961e-05, "loss": 0.1196, "step": 6330 }, { "grad_norm": 0.8781450390815735, "learning_rate": 9.995099022193871e-05, "loss": 0.1166, "step": 6340 }, { "grad_norm": 0.7170438766479492, "learning_rate": 9.995025557476261e-05, "loss": 0.1069, "step": 6350 }, { "grad_norm": 0.6667071580886841, "learning_rate": 9.994951546510165e-05, "loss": 0.107, "step": 6360 }, { "grad_norm": 0.4899360239505768, "learning_rate": 9.994876989303679e-05, "loss": 0.1178, "step": 6370 }, { "grad_norm": 0.8592397570610046, "learning_rate": 9.994801885864955e-05, "loss": 0.1085, "step": 6380 }, { "grad_norm": 1.1556916236877441, "learning_rate": 9.994726236202205e-05, "loss": 0.1286, "step": 6390 }, { "grad_norm": 0.8111037611961365, "learning_rate": 9.994650040323704e-05, "loss": 0.1251, "step": 6400 }, { "grad_norm": 0.5720699429512024, "learning_rate": 9.994573298237784e-05, "loss": 0.0891, "step": 6410 }, { "grad_norm": 0.7233849167823792, "learning_rate": 9.994496009952837e-05, "loss": 0.1184, "step": 6420 }, { "grad_norm": 0.6643378138542175, "learning_rate": 9.994418175477316e-05, "loss": 0.1222, "step": 6430 }, { "grad_norm": 0.6363747119903564, "learning_rate": 9.994339794819733e-05, "loss": 0.1152, "step": 6440 }, { "grad_norm": 0.803952693939209, "learning_rate": 9.994260867988658e-05, "loss": 0.1089, "step": 6450 }, { "grad_norm": 0.8416454195976257, "learning_rate": 9.994181394992723e-05, "loss": 0.1086, "step": 6460 }, { "grad_norm": 0.7636498212814331, "learning_rate": 9.994101375840618e-05, "loss": 0.1157, "step": 6470 }, { "grad_norm": 1.1395292282104492, "learning_rate": 9.994020810541098e-05, "loss": 0.1078, "step": 6480 }, { "grad_norm": 1.0761419534683228, "learning_rate": 9.99393969910297e-05, "loss": 0.1341, "step": 6490 }, { "grad_norm": 0.757580578327179, "learning_rate": 9.993858041535104e-05, "loss": 0.1024, "step": 6500 }, { "grad_norm": 0.6491345167160034, "learning_rate": 9.99377583784643e-05, "loss": 0.0956, "step": 6510 }, { "grad_norm": 0.9685371518135071, "learning_rate": 9.993693088045939e-05, "loss": 0.1095, "step": 6520 }, { "grad_norm": 0.541993260383606, "learning_rate": 9.99360979214268e-05, "loss": 0.1108, "step": 6530 }, { "grad_norm": 0.8824560642242432, "learning_rate": 9.99352595014576e-05, "loss": 0.1065, "step": 6540 }, { "grad_norm": 0.9028459191322327, "learning_rate": 9.993441562064354e-05, "loss": 0.1011, "step": 6550 }, { "grad_norm": 0.8878694176673889, "learning_rate": 9.993356627907685e-05, "loss": 0.1119, "step": 6560 }, { "grad_norm": 0.8327578902244568, "learning_rate": 9.99327114768504e-05, "loss": 0.1244, "step": 6570 }, { "grad_norm": 0.7200326919555664, "learning_rate": 9.99318512140577e-05, "loss": 0.095, "step": 6580 }, { "grad_norm": 0.8144255876541138, "learning_rate": 9.993098549079284e-05, "loss": 0.1289, "step": 6590 }, { "grad_norm": 0.7347866892814636, "learning_rate": 9.993011430715047e-05, "loss": 0.1164, "step": 6600 }, { "grad_norm": 0.5736420750617981, "learning_rate": 9.992923766322586e-05, "loss": 0.0987, "step": 6610 }, { "grad_norm": 0.6244783401489258, "learning_rate": 9.99283555591149e-05, "loss": 0.0929, "step": 6620 }, { "grad_norm": 0.5681979656219482, "learning_rate": 9.992746799491404e-05, "loss": 0.1116, "step": 6630 }, { "grad_norm": 0.786650538444519, "learning_rate": 9.992657497072033e-05, "loss": 0.1069, "step": 6640 }, { "grad_norm": 0.6826533079147339, "learning_rate": 9.992567648663147e-05, "loss": 0.1394, "step": 6650 }, { "grad_norm": 0.5396339297294617, "learning_rate": 9.992477254274568e-05, "loss": 0.0963, "step": 6660 }, { "grad_norm": 0.6588963270187378, "learning_rate": 9.992386313916183e-05, "loss": 0.1005, "step": 6670 }, { "grad_norm": 0.9031718373298645, "learning_rate": 9.992294827597934e-05, "loss": 0.1396, "step": 6680 }, { "grad_norm": 0.6935091614723206, "learning_rate": 9.992202795329831e-05, "loss": 0.119, "step": 6690 }, { "grad_norm": 0.7908576726913452, "learning_rate": 9.992110217121936e-05, "loss": 0.0989, "step": 6700 }, { "grad_norm": 0.7283032536506653, "learning_rate": 9.992017092984372e-05, "loss": 0.1246, "step": 6710 }, { "grad_norm": 0.6253425478935242, "learning_rate": 9.991923422927326e-05, "loss": 0.1137, "step": 6720 }, { "grad_norm": 0.9388003349304199, "learning_rate": 9.991829206961037e-05, "loss": 0.1225, "step": 6730 }, { "grad_norm": 0.8617849349975586, "learning_rate": 9.991734445095813e-05, "loss": 0.1348, "step": 6740 }, { "grad_norm": 0.6791841983795166, "learning_rate": 9.991639137342015e-05, "loss": 0.0879, "step": 6750 }, { "grad_norm": 0.5580704808235168, "learning_rate": 9.991543283710064e-05, "loss": 0.0988, "step": 6760 }, { "grad_norm": 0.7125455737113953, "learning_rate": 9.991446884210445e-05, "loss": 0.0945, "step": 6770 }, { "grad_norm": 0.8367588520050049, "learning_rate": 9.9913499388537e-05, "loss": 0.1199, "step": 6780 }, { "grad_norm": 0.8475788831710815, "learning_rate": 9.99125244765043e-05, "loss": 0.1019, "step": 6790 }, { "grad_norm": 0.9946908950805664, "learning_rate": 9.991154410611296e-05, "loss": 0.1243, "step": 6800 }, { "grad_norm": 0.6808988451957703, "learning_rate": 9.99105582774702e-05, "loss": 0.0883, "step": 6810 }, { "grad_norm": 0.7812274694442749, "learning_rate": 9.990956699068384e-05, "loss": 0.1105, "step": 6820 }, { "grad_norm": 0.628201425075531, "learning_rate": 9.990857024586224e-05, "loss": 0.0956, "step": 6830 }, { "grad_norm": 0.6207464933395386, "learning_rate": 9.990756804311446e-05, "loss": 0.1074, "step": 6840 }, { "grad_norm": 0.6096328496932983, "learning_rate": 9.990656038255006e-05, "loss": 0.1031, "step": 6850 }, { "grad_norm": 0.9315399527549744, "learning_rate": 9.990554726427926e-05, "loss": 0.0964, "step": 6860 }, { "grad_norm": 1.3488364219665527, "learning_rate": 9.990452868841284e-05, "loss": 0.1466, "step": 6870 }, { "grad_norm": 0.6599928140640259, "learning_rate": 9.99035046550622e-05, "loss": 0.1326, "step": 6880 }, { "grad_norm": 0.5851873755455017, "learning_rate": 9.99024751643393e-05, "loss": 0.1145, "step": 6890 }, { "grad_norm": 0.7156887650489807, "learning_rate": 9.990144021635677e-05, "loss": 0.1131, "step": 6900 }, { "grad_norm": 1.0950959920883179, "learning_rate": 9.990039981122775e-05, "loss": 0.0927, "step": 6910 }, { "grad_norm": 0.735602617263794, "learning_rate": 9.989935394906602e-05, "loss": 0.1232, "step": 6920 }, { "grad_norm": 0.8182206749916077, "learning_rate": 9.989830262998598e-05, "loss": 0.13, "step": 6930 }, { "grad_norm": 0.7915406823158264, "learning_rate": 9.989724585410259e-05, "loss": 0.1071, "step": 6940 }, { "grad_norm": 0.6824585795402527, "learning_rate": 9.989618362153139e-05, "loss": 0.1079, "step": 6950 }, { "grad_norm": 0.7255579233169556, "learning_rate": 9.989511593238859e-05, "loss": 0.1073, "step": 6960 }, { "grad_norm": 0.8127313256263733, "learning_rate": 9.98940427867909e-05, "loss": 0.1258, "step": 6970 }, { "grad_norm": 0.7854556441307068, "learning_rate": 9.989296418485573e-05, "loss": 0.1074, "step": 6980 }, { "grad_norm": 0.5172517895698547, "learning_rate": 9.989188012670101e-05, "loss": 0.0884, "step": 6990 }, { "grad_norm": 0.7951402068138123, "learning_rate": 9.989079061244528e-05, "loss": 0.0949, "step": 7000 }, { "grad_norm": 0.6600506901741028, "learning_rate": 9.988969564220769e-05, "loss": 0.1244, "step": 7010 }, { "grad_norm": 0.8216346502304077, "learning_rate": 9.988859521610801e-05, "loss": 0.1081, "step": 7020 }, { "grad_norm": 0.6828606724739075, "learning_rate": 9.988748933426656e-05, "loss": 0.1326, "step": 7030 }, { "grad_norm": 0.6151379942893982, "learning_rate": 9.988637799680428e-05, "loss": 0.1128, "step": 7040 }, { "grad_norm": 0.8064102530479431, "learning_rate": 9.98852612038427e-05, "loss": 0.0876, "step": 7050 }, { "grad_norm": 0.6821476817131042, "learning_rate": 9.988413895550397e-05, "loss": 0.0894, "step": 7060 }, { "grad_norm": 0.7960046529769897, "learning_rate": 9.98830112519108e-05, "loss": 0.1025, "step": 7070 }, { "grad_norm": 0.6301599144935608, "learning_rate": 9.98818780931865e-05, "loss": 0.1236, "step": 7080 }, { "grad_norm": 0.977385938167572, "learning_rate": 9.988073947945502e-05, "loss": 0.0986, "step": 7090 }, { "grad_norm": 0.8952676057815552, "learning_rate": 9.987959541084087e-05, "loss": 0.1143, "step": 7100 }, { "grad_norm": 0.7005480527877808, "learning_rate": 9.987844588746915e-05, "loss": 0.1072, "step": 7110 }, { "grad_norm": 1.2688026428222656, "learning_rate": 9.987729090946558e-05, "loss": 0.1022, "step": 7120 }, { "grad_norm": 0.5785859227180481, "learning_rate": 9.987613047695647e-05, "loss": 0.0963, "step": 7130 }, { "grad_norm": 1.3878778219223022, "learning_rate": 9.987496459006871e-05, "loss": 0.1108, "step": 7140 }, { "grad_norm": 0.8099382519721985, "learning_rate": 9.987379324892982e-05, "loss": 0.1348, "step": 7150 }, { "grad_norm": 0.9759450554847717, "learning_rate": 9.987261645366788e-05, "loss": 0.1116, "step": 7160 }, { "grad_norm": 0.5518325567245483, "learning_rate": 9.987143420441158e-05, "loss": 0.1136, "step": 7170 }, { "grad_norm": 0.9999750256538391, "learning_rate": 9.987024650129022e-05, "loss": 0.1174, "step": 7180 }, { "grad_norm": 0.5780439972877502, "learning_rate": 9.986905334443368e-05, "loss": 0.1165, "step": 7190 }, { "grad_norm": 0.8324973583221436, "learning_rate": 9.986785473397245e-05, "loss": 0.1249, "step": 7200 }, { "grad_norm": 0.6116830706596375, "learning_rate": 9.98666506700376e-05, "loss": 0.1121, "step": 7210 }, { "grad_norm": 0.8213176131248474, "learning_rate": 9.986544115276081e-05, "loss": 0.1107, "step": 7220 }, { "grad_norm": 0.942426323890686, "learning_rate": 9.986422618227433e-05, "loss": 0.1381, "step": 7230 }, { "grad_norm": 0.8043067455291748, "learning_rate": 9.986300575871106e-05, "loss": 0.1114, "step": 7240 }, { "grad_norm": 0.9888722896575928, "learning_rate": 9.986177988220444e-05, "loss": 0.1138, "step": 7250 }, { "grad_norm": 0.7964550256729126, "learning_rate": 9.986054855288856e-05, "loss": 0.1007, "step": 7260 }, { "grad_norm": 0.5764176249504089, "learning_rate": 9.985931177089802e-05, "loss": 0.1127, "step": 7270 }, { "grad_norm": 0.8323530554771423, "learning_rate": 9.985806953636814e-05, "loss": 0.1151, "step": 7280 }, { "grad_norm": 0.7879937887191772, "learning_rate": 9.985682184943471e-05, "loss": 0.1151, "step": 7290 }, { "grad_norm": 0.6031246185302734, "learning_rate": 9.98555687102342e-05, "loss": 0.1002, "step": 7300 }, { "grad_norm": 0.6770017147064209, "learning_rate": 9.985431011890367e-05, "loss": 0.1046, "step": 7310 }, { "grad_norm": 0.7862909436225891, "learning_rate": 9.985304607558075e-05, "loss": 0.1086, "step": 7320 }, { "grad_norm": 0.8050504922866821, "learning_rate": 9.985177658040364e-05, "loss": 0.1152, "step": 7330 }, { "grad_norm": 0.67024165391922, "learning_rate": 9.985050163351119e-05, "loss": 0.1153, "step": 7340 }, { "grad_norm": 0.848564863204956, "learning_rate": 9.984922123504286e-05, "loss": 0.0866, "step": 7350 }, { "grad_norm": 0.8265659213066101, "learning_rate": 9.984793538513862e-05, "loss": 0.0991, "step": 7360 }, { "grad_norm": 0.5680902004241943, "learning_rate": 9.984664408393912e-05, "loss": 0.1113, "step": 7370 }, { "grad_norm": 0.743098258972168, "learning_rate": 9.984534733158556e-05, "loss": 0.1002, "step": 7380 }, { "grad_norm": 0.8742406368255615, "learning_rate": 9.984404512821977e-05, "loss": 0.1218, "step": 7390 }, { "grad_norm": 0.8132476210594177, "learning_rate": 9.984273747398411e-05, "loss": 0.1128, "step": 7400 }, { "grad_norm": 0.6278675198554993, "learning_rate": 9.984142436902165e-05, "loss": 0.11, "step": 7410 }, { "grad_norm": 0.6144304871559143, "learning_rate": 9.984010581347596e-05, "loss": 0.1178, "step": 7420 }, { "grad_norm": 0.4917396008968353, "learning_rate": 9.983878180749121e-05, "loss": 0.1118, "step": 7430 }, { "grad_norm": 0.861224889755249, "learning_rate": 9.983745235121222e-05, "loss": 0.1015, "step": 7440 }, { "grad_norm": 0.9597683548927307, "learning_rate": 9.983611744478438e-05, "loss": 0.1009, "step": 7450 }, { "grad_norm": 0.8784281015396118, "learning_rate": 9.983477708835365e-05, "loss": 0.0913, "step": 7460 }, { "grad_norm": 1.0868514776229858, "learning_rate": 9.983343128206664e-05, "loss": 0.109, "step": 7470 }, { "grad_norm": 1.1725103855133057, "learning_rate": 9.983208002607049e-05, "loss": 0.1257, "step": 7480 }, { "grad_norm": 0.7920415997505188, "learning_rate": 9.9830723320513e-05, "loss": 0.1094, "step": 7490 }, { "grad_norm": 0.9432948231697083, "learning_rate": 9.982936116554254e-05, "loss": 0.1162, "step": 7500 }, { "grad_norm": 0.7163913249969482, "learning_rate": 9.982799356130803e-05, "loss": 0.0921, "step": 7510 }, { "grad_norm": 0.8028135299682617, "learning_rate": 9.982662050795908e-05, "loss": 0.0909, "step": 7520 }, { "grad_norm": 0.5800248980522156, "learning_rate": 9.982524200564583e-05, "loss": 0.0883, "step": 7530 }, { "grad_norm": 0.5274880528450012, "learning_rate": 9.982385805451901e-05, "loss": 0.0945, "step": 7540 }, { "grad_norm": 0.6574599742889404, "learning_rate": 9.982246865472998e-05, "loss": 0.0838, "step": 7550 }, { "grad_norm": 0.8814643025398254, "learning_rate": 9.982107380643069e-05, "loss": 0.1047, "step": 7560 }, { "grad_norm": 0.8162550926208496, "learning_rate": 9.981967350977368e-05, "loss": 0.1084, "step": 7570 }, { "grad_norm": 0.5611205697059631, "learning_rate": 9.981826776491208e-05, "loss": 0.0937, "step": 7580 }, { "grad_norm": 0.6246904730796814, "learning_rate": 9.98168565719996e-05, "loss": 0.114, "step": 7590 }, { "grad_norm": 0.623023271560669, "learning_rate": 9.98154399311906e-05, "loss": 0.0982, "step": 7600 }, { "grad_norm": 0.6922298669815063, "learning_rate": 9.981401784263997e-05, "loss": 0.1119, "step": 7610 }, { "grad_norm": 0.5456994771957397, "learning_rate": 9.981259030650326e-05, "loss": 0.1042, "step": 7620 }, { "grad_norm": 0.7205878496170044, "learning_rate": 9.981115732293655e-05, "loss": 0.0994, "step": 7630 }, { "grad_norm": 0.9959262609481812, "learning_rate": 9.980971889209659e-05, "loss": 0.0997, "step": 7640 }, { "grad_norm": 1.1316388845443726, "learning_rate": 9.980827501414064e-05, "loss": 0.0916, "step": 7650 }, { "grad_norm": 0.5285545587539673, "learning_rate": 9.980682568922663e-05, "loss": 0.0956, "step": 7660 }, { "grad_norm": 0.571725606918335, "learning_rate": 9.980537091751304e-05, "loss": 0.1083, "step": 7670 }, { "grad_norm": 0.8298535346984863, "learning_rate": 9.980391069915897e-05, "loss": 0.0964, "step": 7680 }, { "grad_norm": 0.607584536075592, "learning_rate": 9.98024450343241e-05, "loss": 0.1024, "step": 7690 }, { "grad_norm": 0.5957757830619812, "learning_rate": 9.980097392316872e-05, "loss": 0.0954, "step": 7700 }, { "grad_norm": 0.6403868198394775, "learning_rate": 9.97994973658537e-05, "loss": 0.1144, "step": 7710 }, { "grad_norm": 0.5858974456787109, "learning_rate": 9.979801536254054e-05, "loss": 0.1021, "step": 7720 }, { "grad_norm": 0.6165576577186584, "learning_rate": 9.979652791339127e-05, "loss": 0.1083, "step": 7730 }, { "grad_norm": 0.5677700042724609, "learning_rate": 9.97950350185686e-05, "loss": 0.0978, "step": 7740 }, { "grad_norm": 0.5730867385864258, "learning_rate": 9.979353667823574e-05, "loss": 0.1105, "step": 7750 }, { "grad_norm": 0.785773754119873, "learning_rate": 9.979203289255658e-05, "loss": 0.1139, "step": 7760 }, { "grad_norm": 0.6484279036521912, "learning_rate": 9.979052366169557e-05, "loss": 0.1114, "step": 7770 }, { "grad_norm": 0.7170355916023254, "learning_rate": 9.978900898581775e-05, "loss": 0.1097, "step": 7780 }, { "grad_norm": 0.5429760813713074, "learning_rate": 9.978748886508875e-05, "loss": 0.0965, "step": 7790 }, { "grad_norm": 0.7180110812187195, "learning_rate": 9.978596329967484e-05, "loss": 0.108, "step": 7800 }, { "grad_norm": 0.6950826644897461, "learning_rate": 9.978443228974284e-05, "loss": 0.1024, "step": 7810 }, { "grad_norm": 0.7251401543617249, "learning_rate": 9.978289583546015e-05, "loss": 0.0977, "step": 7820 }, { "grad_norm": 0.8709108829498291, "learning_rate": 9.978135393699484e-05, "loss": 0.1001, "step": 7830 }, { "grad_norm": 0.7028502821922302, "learning_rate": 9.977980659451548e-05, "loss": 0.1032, "step": 7840 }, { "grad_norm": 0.6359463930130005, "learning_rate": 9.977825380819135e-05, "loss": 0.0875, "step": 7850 }, { "grad_norm": 0.6452231407165527, "learning_rate": 9.97766955781922e-05, "loss": 0.0993, "step": 7860 }, { "grad_norm": 0.7716856002807617, "learning_rate": 9.977513190468848e-05, "loss": 0.0987, "step": 7870 }, { "grad_norm": 0.7294530272483826, "learning_rate": 9.977356278785116e-05, "loss": 0.1052, "step": 7880 }, { "grad_norm": 0.7912027835845947, "learning_rate": 9.977198822785184e-05, "loss": 0.1015, "step": 7890 }, { "grad_norm": 0.6944399476051331, "learning_rate": 9.977040822486273e-05, "loss": 0.1112, "step": 7900 }, { "grad_norm": 0.523735523223877, "learning_rate": 9.97688227790566e-05, "loss": 0.1197, "step": 7910 }, { "grad_norm": 0.6997634768486023, "learning_rate": 9.976723189060684e-05, "loss": 0.1312, "step": 7920 }, { "grad_norm": 0.7120993733406067, "learning_rate": 9.976563555968742e-05, "loss": 0.0907, "step": 7930 }, { "grad_norm": 0.6221331357955933, "learning_rate": 9.976403378647292e-05, "loss": 0.1017, "step": 7940 }, { "grad_norm": 1.529201626777649, "learning_rate": 9.97624265711385e-05, "loss": 0.0999, "step": 7950 }, { "grad_norm": 1.428389549255371, "learning_rate": 9.976081391385993e-05, "loss": 0.1125, "step": 7960 }, { "grad_norm": 0.7564881443977356, "learning_rate": 9.975919581481356e-05, "loss": 0.1001, "step": 7970 }, { "grad_norm": 0.8585585355758667, "learning_rate": 9.975757227417634e-05, "loss": 0.0929, "step": 7980 }, { "grad_norm": 0.6751505136489868, "learning_rate": 9.975594329212586e-05, "loss": 0.0919, "step": 7990 }, { "grad_norm": 0.8327727913856506, "learning_rate": 9.97543088688402e-05, "loss": 0.0979, "step": 8000 }, { "grad_norm": 0.8198424577713013, "learning_rate": 9.975266900449814e-05, "loss": 0.0839, "step": 8010 }, { "grad_norm": 1.08558189868927, "learning_rate": 9.975102369927898e-05, "loss": 0.1022, "step": 8020 }, { "grad_norm": 0.6531072854995728, "learning_rate": 9.974937295336269e-05, "loss": 0.1028, "step": 8030 }, { "grad_norm": 0.7134861946105957, "learning_rate": 9.974771676692975e-05, "loss": 0.0982, "step": 8040 }, { "grad_norm": 0.6714780330657959, "learning_rate": 9.974605514016131e-05, "loss": 0.102, "step": 8050 }, { "grad_norm": 0.6883942484855652, "learning_rate": 9.974438807323907e-05, "loss": 0.0985, "step": 8060 }, { "grad_norm": 0.666153609752655, "learning_rate": 9.974271556634535e-05, "loss": 0.0931, "step": 8070 }, { "grad_norm": 0.9543102383613586, "learning_rate": 9.974103761966302e-05, "loss": 0.1149, "step": 8080 }, { "grad_norm": 0.5815755128860474, "learning_rate": 9.973935423337563e-05, "loss": 0.1026, "step": 8090 }, { "grad_norm": 0.6485018134117126, "learning_rate": 9.973766540766722e-05, "loss": 0.1182, "step": 8100 }, { "grad_norm": 0.8159518241882324, "learning_rate": 9.97359711427225e-05, "loss": 0.1107, "step": 8110 }, { "grad_norm": 0.6313855648040771, "learning_rate": 9.973427143872677e-05, "loss": 0.1134, "step": 8120 }, { "grad_norm": 0.850178062915802, "learning_rate": 9.973256629586589e-05, "loss": 0.1151, "step": 8130 }, { "grad_norm": 0.6191561222076416, "learning_rate": 9.973085571432632e-05, "loss": 0.1147, "step": 8140 }, { "grad_norm": 0.7791539430618286, "learning_rate": 9.972913969429513e-05, "loss": 0.1114, "step": 8150 }, { "grad_norm": 0.5066817402839661, "learning_rate": 9.972741823596e-05, "loss": 0.0947, "step": 8160 }, { "grad_norm": 0.5814530253410339, "learning_rate": 9.972569133950917e-05, "loss": 0.1098, "step": 8170 }, { "grad_norm": 0.6568130850791931, "learning_rate": 9.972395900513151e-05, "loss": 0.085, "step": 8180 }, { "grad_norm": 0.5901515483856201, "learning_rate": 9.972222123301645e-05, "loss": 0.0988, "step": 8190 }, { "grad_norm": 0.8309993743896484, "learning_rate": 9.972047802335403e-05, "loss": 0.1014, "step": 8200 }, { "grad_norm": 0.4473201632499695, "learning_rate": 9.971872937633488e-05, "loss": 0.0822, "step": 8210 }, { "grad_norm": 0.7367064356803894, "learning_rate": 9.971697529215024e-05, "loss": 0.1111, "step": 8220 }, { "grad_norm": 0.6156007647514343, "learning_rate": 9.971521577099192e-05, "loss": 0.0935, "step": 8230 }, { "grad_norm": 0.5262095332145691, "learning_rate": 9.971345081305236e-05, "loss": 0.1019, "step": 8240 }, { "grad_norm": 0.6930136680603027, "learning_rate": 9.971168041852456e-05, "loss": 0.0923, "step": 8250 }, { "grad_norm": 0.6858784556388855, "learning_rate": 9.970990458760215e-05, "loss": 0.0785, "step": 8260 }, { "grad_norm": 0.5252299308776855, "learning_rate": 9.970812332047929e-05, "loss": 0.0991, "step": 8270 }, { "grad_norm": 0.5938407778739929, "learning_rate": 9.97063366173508e-05, "loss": 0.1064, "step": 8280 }, { "grad_norm": 0.7178565263748169, "learning_rate": 9.970454447841207e-05, "loss": 0.0985, "step": 8290 }, { "grad_norm": 0.5807419419288635, "learning_rate": 9.970274690385909e-05, "loss": 0.1047, "step": 8300 }, { "grad_norm": 0.5768489837646484, "learning_rate": 9.970094389388844e-05, "loss": 0.0914, "step": 8310 }, { "grad_norm": 0.592161238193512, "learning_rate": 9.969913544869728e-05, "loss": 0.0968, "step": 8320 }, { "grad_norm": 0.6527407169342041, "learning_rate": 9.96973215684834e-05, "loss": 0.0969, "step": 8330 }, { "grad_norm": 0.6663581728935242, "learning_rate": 9.969550225344513e-05, "loss": 0.0777, "step": 8340 }, { "grad_norm": 0.588917076587677, "learning_rate": 9.969367750378147e-05, "loss": 0.0987, "step": 8350 }, { "grad_norm": 0.6432686448097229, "learning_rate": 9.969184731969194e-05, "loss": 0.0889, "step": 8360 }, { "grad_norm": 0.6830676794052124, "learning_rate": 9.96900117013767e-05, "loss": 0.1185, "step": 8370 }, { "grad_norm": 0.6747348308563232, "learning_rate": 9.96881706490365e-05, "loss": 0.099, "step": 8380 }, { "grad_norm": 0.5713630318641663, "learning_rate": 9.968632416287265e-05, "loss": 0.0882, "step": 8390 }, { "grad_norm": 0.605571985244751, "learning_rate": 9.96844722430871e-05, "loss": 0.0848, "step": 8400 }, { "grad_norm": 0.8146535754203796, "learning_rate": 9.968261488988235e-05, "loss": 0.1044, "step": 8410 }, { "grad_norm": 0.6902555823326111, "learning_rate": 9.968075210346155e-05, "loss": 0.0873, "step": 8420 }, { "grad_norm": 0.8256837129592896, "learning_rate": 9.967888388402839e-05, "loss": 0.097, "step": 8430 }, { "grad_norm": 0.6604471802711487, "learning_rate": 9.967701023178717e-05, "loss": 0.0898, "step": 8440 }, { "grad_norm": 0.8641937971115112, "learning_rate": 9.967513114694282e-05, "loss": 0.1082, "step": 8450 }, { "grad_norm": 0.5108922123908997, "learning_rate": 9.967324662970079e-05, "loss": 0.0805, "step": 8460 }, { "grad_norm": 1.009664535522461, "learning_rate": 9.96713566802672e-05, "loss": 0.0992, "step": 8470 }, { "grad_norm": 0.6544330716133118, "learning_rate": 9.966946129884873e-05, "loss": 0.1002, "step": 8480 }, { "grad_norm": 0.736150860786438, "learning_rate": 9.966756048565265e-05, "loss": 0.1061, "step": 8490 }, { "grad_norm": 1.0628365278244019, "learning_rate": 9.966565424088681e-05, "loss": 0.1196, "step": 8500 }, { "grad_norm": 0.8461632132530212, "learning_rate": 9.96637425647597e-05, "loss": 0.1034, "step": 8510 }, { "grad_norm": 0.5547913312911987, "learning_rate": 9.966182545748038e-05, "loss": 0.0912, "step": 8520 }, { "grad_norm": 1.2567033767700195, "learning_rate": 9.96599029192585e-05, "loss": 0.0974, "step": 8530 }, { "grad_norm": 0.7255085110664368, "learning_rate": 9.965797495030428e-05, "loss": 0.1046, "step": 8540 }, { "grad_norm": 0.8249364495277405, "learning_rate": 9.96560415508286e-05, "loss": 0.0988, "step": 8550 }, { "grad_norm": 0.641168475151062, "learning_rate": 9.965410272104286e-05, "loss": 0.0946, "step": 8560 }, { "grad_norm": 0.6343355774879456, "learning_rate": 9.96521584611591e-05, "loss": 0.1258, "step": 8570 }, { "grad_norm": 0.7453083992004395, "learning_rate": 9.965020877138994e-05, "loss": 0.1105, "step": 8580 }, { "grad_norm": 0.5909978747367859, "learning_rate": 9.964825365194861e-05, "loss": 0.1064, "step": 8590 }, { "grad_norm": 0.6326723098754883, "learning_rate": 9.96462931030489e-05, "loss": 0.0915, "step": 8600 }, { "grad_norm": 0.6705448627471924, "learning_rate": 9.96443271249052e-05, "loss": 0.0887, "step": 8610 }, { "grad_norm": 0.5492733120918274, "learning_rate": 9.964235571773255e-05, "loss": 0.1136, "step": 8620 }, { "grad_norm": 0.7532504796981812, "learning_rate": 9.96403788817465e-05, "loss": 0.0967, "step": 8630 }, { "grad_norm": 0.6135904788970947, "learning_rate": 9.963839661716325e-05, "loss": 0.0945, "step": 8640 }, { "grad_norm": 0.6469835638999939, "learning_rate": 9.963640892419958e-05, "loss": 0.1023, "step": 8650 }, { "grad_norm": 0.7812805771827698, "learning_rate": 9.963441580307286e-05, "loss": 0.0855, "step": 8660 }, { "grad_norm": 0.59685879945755, "learning_rate": 9.963241725400104e-05, "loss": 0.0866, "step": 8670 }, { "grad_norm": 0.6290655732154846, "learning_rate": 9.963041327720271e-05, "loss": 0.0822, "step": 8680 }, { "grad_norm": 0.6089248657226562, "learning_rate": 9.962840387289697e-05, "loss": 0.098, "step": 8690 }, { "grad_norm": 0.7095808386802673, "learning_rate": 9.962638904130363e-05, "loss": 0.1022, "step": 8700 }, { "grad_norm": 0.6313075423240662, "learning_rate": 9.962436878264298e-05, "loss": 0.1136, "step": 8710 }, { "grad_norm": 0.5627358555793762, "learning_rate": 9.962234309713598e-05, "loss": 0.1108, "step": 8720 }, { "grad_norm": 0.6199731230735779, "learning_rate": 9.962031198500414e-05, "loss": 0.0981, "step": 8730 }, { "grad_norm": 0.6164750456809998, "learning_rate": 9.961827544646958e-05, "loss": 0.0961, "step": 8740 }, { "grad_norm": 0.6284675002098083, "learning_rate": 9.961623348175501e-05, "loss": 0.1072, "step": 8750 }, { "grad_norm": 0.5756822228431702, "learning_rate": 9.961418609108377e-05, "loss": 0.0828, "step": 8760 }, { "grad_norm": 0.6885284781455994, "learning_rate": 9.961213327467971e-05, "loss": 0.0989, "step": 8770 }, { "grad_norm": 0.6579473614692688, "learning_rate": 9.961007503276736e-05, "loss": 0.0955, "step": 8780 }, { "grad_norm": 0.7679935097694397, "learning_rate": 9.960801136557179e-05, "loss": 0.0858, "step": 8790 }, { "grad_norm": 0.6104416847229004, "learning_rate": 9.960594227331866e-05, "loss": 0.0816, "step": 8800 }, { "grad_norm": 0.5690643191337585, "learning_rate": 9.960386775623429e-05, "loss": 0.0882, "step": 8810 }, { "grad_norm": 0.6707371473312378, "learning_rate": 9.96017878145455e-05, "loss": 0.0958, "step": 8820 }, { "grad_norm": 0.750119149684906, "learning_rate": 9.959970244847977e-05, "loss": 0.0949, "step": 8830 }, { "grad_norm": 0.5760891437530518, "learning_rate": 9.959761165826518e-05, "loss": 0.0958, "step": 8840 }, { "grad_norm": 1.2138668298721313, "learning_rate": 9.959551544413033e-05, "loss": 0.0991, "step": 8850 }, { "grad_norm": 0.7122470736503601, "learning_rate": 9.959341380630448e-05, "loss": 0.0982, "step": 8860 }, { "grad_norm": 1.1674695014953613, "learning_rate": 9.959130674501746e-05, "loss": 0.1152, "step": 8870 }, { "grad_norm": 0.5339207053184509, "learning_rate": 9.958919426049968e-05, "loss": 0.1088, "step": 8880 }, { "grad_norm": 0.7170137763023376, "learning_rate": 9.958707635298219e-05, "loss": 0.0824, "step": 8890 }, { "grad_norm": 0.537982165813446, "learning_rate": 9.958495302269657e-05, "loss": 0.101, "step": 8900 }, { "grad_norm": 0.603482186794281, "learning_rate": 9.958282426987503e-05, "loss": 0.0929, "step": 8910 }, { "grad_norm": 0.8607549667358398, "learning_rate": 9.95806900947504e-05, "loss": 0.1181, "step": 8920 }, { "grad_norm": 0.9084377288818359, "learning_rate": 9.957855049755604e-05, "loss": 0.1049, "step": 8930 }, { "grad_norm": 0.5878090858459473, "learning_rate": 9.957640547852593e-05, "loss": 0.1127, "step": 8940 }, { "grad_norm": 0.5671360492706299, "learning_rate": 9.957425503789466e-05, "loss": 0.0966, "step": 8950 }, { "grad_norm": 0.5780009031295776, "learning_rate": 9.957209917589738e-05, "loss": 0.0848, "step": 8960 }, { "grad_norm": 0.6554332971572876, "learning_rate": 9.956993789276987e-05, "loss": 0.0889, "step": 8970 }, { "grad_norm": 0.5717939734458923, "learning_rate": 9.956777118874847e-05, "loss": 0.0957, "step": 8980 }, { "grad_norm": 0.8482783436775208, "learning_rate": 9.956559906407016e-05, "loss": 0.1242, "step": 8990 }, { "grad_norm": 0.5769920349121094, "learning_rate": 9.956342151897245e-05, "loss": 0.0998, "step": 9000 }, { "grad_norm": 0.6332108974456787, "learning_rate": 9.956123855369346e-05, "loss": 0.0857, "step": 9010 }, { "grad_norm": 0.9140949249267578, "learning_rate": 9.955905016847196e-05, "loss": 0.0886, "step": 9020 }, { "grad_norm": 0.7139384746551514, "learning_rate": 9.955685636354723e-05, "loss": 0.126, "step": 9030 }, { "grad_norm": 0.8133804202079773, "learning_rate": 9.95546571391592e-05, "loss": 0.1171, "step": 9040 }, { "grad_norm": 0.5297209620475769, "learning_rate": 9.955245249554837e-05, "loss": 0.0922, "step": 9050 }, { "grad_norm": 0.9399484992027283, "learning_rate": 9.955024243295582e-05, "loss": 0.1379, "step": 9060 }, { "grad_norm": 0.7135273814201355, "learning_rate": 9.954802695162328e-05, "loss": 0.1032, "step": 9070 }, { "grad_norm": 0.7993723154067993, "learning_rate": 9.954580605179302e-05, "loss": 0.0993, "step": 9080 }, { "grad_norm": 0.8193233609199524, "learning_rate": 9.954357973370788e-05, "loss": 0.1073, "step": 9090 }, { "grad_norm": 0.5273614525794983, "learning_rate": 9.954134799761135e-05, "loss": 0.1151, "step": 9100 }, { "grad_norm": 0.632396399974823, "learning_rate": 9.953911084374748e-05, "loss": 0.1067, "step": 9110 }, { "grad_norm": 0.7985724806785583, "learning_rate": 9.953686827236093e-05, "loss": 0.1105, "step": 9120 }, { "grad_norm": 0.5811571478843689, "learning_rate": 9.953462028369695e-05, "loss": 0.1189, "step": 9130 }, { "grad_norm": 0.6309804916381836, "learning_rate": 9.953236687800136e-05, "loss": 0.1008, "step": 9140 }, { "grad_norm": 0.49889814853668213, "learning_rate": 9.95301080555206e-05, "loss": 0.1163, "step": 9150 }, { "grad_norm": 0.4398837983608246, "learning_rate": 9.952784381650171e-05, "loss": 0.0847, "step": 9160 }, { "grad_norm": 0.7202220559120178, "learning_rate": 9.952557416119226e-05, "loss": 0.0908, "step": 9170 }, { "grad_norm": 0.664649486541748, "learning_rate": 9.95232990898405e-05, "loss": 0.0886, "step": 9180 }, { "grad_norm": 0.6199564337730408, "learning_rate": 9.95210186026952e-05, "loss": 0.0822, "step": 9190 }, { "grad_norm": 0.6235049962997437, "learning_rate": 9.951873270000576e-05, "loss": 0.1084, "step": 9200 }, { "grad_norm": 0.6454322934150696, "learning_rate": 9.951644138202216e-05, "loss": 0.1057, "step": 9210 }, { "grad_norm": 0.4410513639450073, "learning_rate": 9.951414464899498e-05, "loss": 0.0921, "step": 9220 }, { "grad_norm": 1.0365346670150757, "learning_rate": 9.951184250117538e-05, "loss": 0.0817, "step": 9230 }, { "grad_norm": 0.7110735774040222, "learning_rate": 9.950953493881513e-05, "loss": 0.0903, "step": 9240 }, { "grad_norm": 0.6652246713638306, "learning_rate": 9.950722196216658e-05, "loss": 0.0912, "step": 9250 }, { "grad_norm": 1.4935067892074585, "learning_rate": 9.950490357148265e-05, "loss": 0.0925, "step": 9260 }, { "grad_norm": 0.8027993440628052, "learning_rate": 9.950257976701692e-05, "loss": 0.1105, "step": 9270 }, { "grad_norm": 0.7835819125175476, "learning_rate": 9.950025054902348e-05, "loss": 0.0927, "step": 9280 }, { "grad_norm": 0.824580192565918, "learning_rate": 9.949791591775706e-05, "loss": 0.0882, "step": 9290 }, { "grad_norm": 0.6422032713890076, "learning_rate": 9.949557587347298e-05, "loss": 0.117, "step": 9300 }, { "grad_norm": 0.863757848739624, "learning_rate": 9.949323041642713e-05, "loss": 0.0937, "step": 9310 }, { "grad_norm": 0.8296629786491394, "learning_rate": 9.949087954687602e-05, "loss": 0.1138, "step": 9320 }, { "grad_norm": 0.8077664971351624, "learning_rate": 9.948852326507672e-05, "loss": 0.0987, "step": 9330 }, { "grad_norm": 0.6423782706260681, "learning_rate": 9.948616157128694e-05, "loss": 0.0871, "step": 9340 }, { "grad_norm": 0.5564231872558594, "learning_rate": 9.948379446576493e-05, "loss": 0.0818, "step": 9350 }, { "grad_norm": 0.5948541760444641, "learning_rate": 9.948142194876952e-05, "loss": 0.082, "step": 9360 }, { "grad_norm": 0.5351197719573975, "learning_rate": 9.947904402056024e-05, "loss": 0.0928, "step": 9370 }, { "grad_norm": 0.5828901529312134, "learning_rate": 9.947666068139708e-05, "loss": 0.1032, "step": 9380 }, { "grad_norm": 0.6311654448509216, "learning_rate": 9.947427193154071e-05, "loss": 0.0847, "step": 9390 }, { "grad_norm": 0.5558895468711853, "learning_rate": 9.947187777125233e-05, "loss": 0.1171, "step": 9400 }, { "grad_norm": 0.5113588571548462, "learning_rate": 9.946947820079377e-05, "loss": 0.0852, "step": 9410 }, { "grad_norm": 0.8850395083427429, "learning_rate": 9.946707322042747e-05, "loss": 0.09, "step": 9420 }, { "grad_norm": 0.700267493724823, "learning_rate": 9.94646628304164e-05, "loss": 0.0957, "step": 9430 }, { "grad_norm": 0.7684928178787231, "learning_rate": 9.946224703102418e-05, "loss": 0.1052, "step": 9440 }, { "grad_norm": 0.7233399748802185, "learning_rate": 9.945982582251498e-05, "loss": 0.0946, "step": 9450 }, { "grad_norm": 0.9910270571708679, "learning_rate": 9.94573992051536e-05, "loss": 0.0935, "step": 9460 }, { "grad_norm": 0.639900803565979, "learning_rate": 9.94549671792054e-05, "loss": 0.1094, "step": 9470 }, { "grad_norm": 0.7533113360404968, "learning_rate": 9.945252974493635e-05, "loss": 0.0961, "step": 9480 }, { "grad_norm": 0.5036046504974365, "learning_rate": 9.9450086902613e-05, "loss": 0.0968, "step": 9490 }, { "grad_norm": 0.5030291080474854, "learning_rate": 9.944763865250248e-05, "loss": 0.0911, "step": 9500 }, { "grad_norm": 0.6666432023048401, "learning_rate": 9.944518499487254e-05, "loss": 0.0803, "step": 9510 }, { "grad_norm": 0.707115650177002, "learning_rate": 9.944272592999151e-05, "loss": 0.1115, "step": 9520 }, { "grad_norm": 0.9917125105857849, "learning_rate": 9.94402614581283e-05, "loss": 0.1013, "step": 9530 }, { "grad_norm": 0.9923180341720581, "learning_rate": 9.943779157955244e-05, "loss": 0.1128, "step": 9540 }, { "grad_norm": 0.5614116787910461, "learning_rate": 9.943531629453403e-05, "loss": 0.099, "step": 9550 }, { "grad_norm": 0.6384596228599548, "learning_rate": 9.943283560334375e-05, "loss": 0.1013, "step": 9560 }, { "grad_norm": 0.6341807246208191, "learning_rate": 9.943034950625288e-05, "loss": 0.1022, "step": 9570 }, { "grad_norm": 0.7621190547943115, "learning_rate": 9.942785800353332e-05, "loss": 0.0988, "step": 9580 }, { "grad_norm": 0.6942609548568726, "learning_rate": 9.942536109545751e-05, "loss": 0.0848, "step": 9590 }, { "grad_norm": 0.6521564722061157, "learning_rate": 9.942285878229853e-05, "loss": 0.0851, "step": 9600 }, { "grad_norm": 0.7611565589904785, "learning_rate": 9.942035106433001e-05, "loss": 0.0856, "step": 9610 }, { "grad_norm": 0.6208741068840027, "learning_rate": 9.94178379418262e-05, "loss": 0.1011, "step": 9620 }, { "grad_norm": 0.919066309928894, "learning_rate": 9.941531941506194e-05, "loss": 0.0886, "step": 9630 }, { "grad_norm": 0.5778529644012451, "learning_rate": 9.941279548431263e-05, "loss": 0.0819, "step": 9640 }, { "grad_norm": 0.5328933596611023, "learning_rate": 9.941026614985431e-05, "loss": 0.1027, "step": 9650 }, { "grad_norm": 0.5442858934402466, "learning_rate": 9.940773141196357e-05, "loss": 0.0945, "step": 9660 }, { "grad_norm": 0.7060144543647766, "learning_rate": 9.94051912709176e-05, "loss": 0.0911, "step": 9670 }, { "grad_norm": 0.6398045420646667, "learning_rate": 9.940264572699421e-05, "loss": 0.0771, "step": 9680 }, { "grad_norm": 0.6228548288345337, "learning_rate": 9.940009478047174e-05, "loss": 0.0842, "step": 9690 }, { "grad_norm": 0.8562597632408142, "learning_rate": 9.939753843162918e-05, "loss": 0.079, "step": 9700 }, { "grad_norm": 0.7232752442359924, "learning_rate": 9.939497668074609e-05, "loss": 0.0742, "step": 9710 }, { "grad_norm": 0.7293767929077148, "learning_rate": 9.93924095281026e-05, "loss": 0.1, "step": 9720 }, { "grad_norm": 0.700738251209259, "learning_rate": 9.938983697397948e-05, "loss": 0.1017, "step": 9730 }, { "grad_norm": 0.7398735880851746, "learning_rate": 9.938725901865805e-05, "loss": 0.102, "step": 9740 }, { "grad_norm": 0.7037026882171631, "learning_rate": 9.93846756624202e-05, "loss": 0.0823, "step": 9750 }, { "grad_norm": 0.46460863947868347, "learning_rate": 9.938208690554849e-05, "loss": 0.092, "step": 9760 }, { "grad_norm": 0.684431791305542, "learning_rate": 9.9379492748326e-05, "loss": 0.096, "step": 9770 }, { "grad_norm": 0.6336464285850525, "learning_rate": 9.937689319103641e-05, "loss": 0.0868, "step": 9780 }, { "grad_norm": 0.6273999810218811, "learning_rate": 9.937428823396404e-05, "loss": 0.0861, "step": 9790 }, { "grad_norm": 0.5933031439781189, "learning_rate": 9.937167787739372e-05, "loss": 0.0689, "step": 9800 }, { "grad_norm": 0.6559424996376038, "learning_rate": 9.936906212161095e-05, "loss": 0.0817, "step": 9810 }, { "grad_norm": 0.6938247680664062, "learning_rate": 9.936644096690176e-05, "loss": 0.1029, "step": 9820 }, { "grad_norm": 0.5510004162788391, "learning_rate": 9.936381441355282e-05, "loss": 0.105, "step": 9830 }, { "grad_norm": 0.5518153309822083, "learning_rate": 9.936118246185136e-05, "loss": 0.1051, "step": 9840 }, { "grad_norm": 0.9793683886528015, "learning_rate": 9.935854511208518e-05, "loss": 0.1, "step": 9850 }, { "grad_norm": 0.5877102017402649, "learning_rate": 9.935590236454272e-05, "loss": 0.0874, "step": 9860 }, { "grad_norm": 0.6308822631835938, "learning_rate": 9.935325421951298e-05, "loss": 0.0806, "step": 9870 }, { "grad_norm": 1.0372501611709595, "learning_rate": 9.935060067728557e-05, "loss": 0.0846, "step": 9880 }, { "grad_norm": 0.5976530313491821, "learning_rate": 9.934794173815067e-05, "loss": 0.0743, "step": 9890 }, { "grad_norm": 0.531636118888855, "learning_rate": 9.934527740239906e-05, "loss": 0.0901, "step": 9900 }, { "grad_norm": 0.6538335680961609, "learning_rate": 9.934260767032209e-05, "loss": 0.0986, "step": 9910 }, { "grad_norm": 0.7140442728996277, "learning_rate": 9.933993254221172e-05, "loss": 0.0929, "step": 9920 }, { "grad_norm": 1.113791584968567, "learning_rate": 9.933725201836053e-05, "loss": 0.0853, "step": 9930 }, { "grad_norm": 0.8427344560623169, "learning_rate": 9.933456609906162e-05, "loss": 0.0877, "step": 9940 }, { "grad_norm": 0.7749441266059875, "learning_rate": 9.933187478460875e-05, "loss": 0.0859, "step": 9950 }, { "grad_norm": 0.5414090156555176, "learning_rate": 9.93291780752962e-05, "loss": 0.093, "step": 9960 }, { "grad_norm": 0.4570617377758026, "learning_rate": 9.932647597141893e-05, "loss": 0.0883, "step": 9970 }, { "grad_norm": 0.7797214984893799, "learning_rate": 9.932376847327239e-05, "loss": 0.0984, "step": 9980 }, { "grad_norm": 0.5189516544342041, "learning_rate": 9.932105558115268e-05, "loss": 0.1106, "step": 9990 }, { "grad_norm": 0.6782358288764954, "learning_rate": 9.931833729535651e-05, "loss": 0.0943, "step": 10000 }, { "grad_norm": 0.5391064882278442, "learning_rate": 9.931561361618111e-05, "loss": 0.0932, "step": 10010 }, { "grad_norm": 0.5278596878051758, "learning_rate": 9.931288454392435e-05, "loss": 0.09, "step": 10020 }, { "grad_norm": 0.6574098467826843, "learning_rate": 9.931015007888467e-05, "loss": 0.086, "step": 10030 }, { "grad_norm": 0.6044958829879761, "learning_rate": 9.930741022136112e-05, "loss": 0.0995, "step": 10040 }, { "grad_norm": 0.4474485516548157, "learning_rate": 9.930466497165333e-05, "loss": 0.0832, "step": 10050 }, { "grad_norm": 0.7187001705169678, "learning_rate": 9.93019143300615e-05, "loss": 0.1019, "step": 10060 }, { "grad_norm": 0.6538999080657959, "learning_rate": 9.929915829688644e-05, "loss": 0.0856, "step": 10070 }, { "grad_norm": 0.5378724932670593, "learning_rate": 9.929639687242955e-05, "loss": 0.0842, "step": 10080 }, { "grad_norm": 0.5459072589874268, "learning_rate": 9.929363005699281e-05, "loss": 0.096, "step": 10090 }, { "grad_norm": 0.44915932416915894, "learning_rate": 9.92908578508788e-05, "loss": 0.0868, "step": 10100 }, { "grad_norm": 0.5724955797195435, "learning_rate": 9.928808025439069e-05, "loss": 0.0915, "step": 10110 }, { "grad_norm": 0.6226571798324585, "learning_rate": 9.928529726783223e-05, "loss": 0.08, "step": 10120 }, { "grad_norm": 0.8073252439498901, "learning_rate": 9.928250889150774e-05, "loss": 0.0907, "step": 10130 }, { "grad_norm": 0.9165542125701904, "learning_rate": 9.92797151257222e-05, "loss": 0.0984, "step": 10140 }, { "grad_norm": 0.5997186899185181, "learning_rate": 9.927691597078108e-05, "loss": 0.0944, "step": 10150 }, { "grad_norm": 1.2177633047103882, "learning_rate": 9.927411142699053e-05, "loss": 0.0938, "step": 10160 }, { "grad_norm": 0.5542764067649841, "learning_rate": 9.927130149465725e-05, "loss": 0.0947, "step": 10170 }, { "grad_norm": 0.7047025561332703, "learning_rate": 9.92684861740885e-05, "loss": 0.1062, "step": 10180 }, { "grad_norm": 0.9518670439720154, "learning_rate": 9.926566546559217e-05, "loss": 0.1044, "step": 10190 }, { "grad_norm": 0.7038217186927795, "learning_rate": 9.926283936947673e-05, "loss": 0.0908, "step": 10200 }, { "grad_norm": 0.6359156370162964, "learning_rate": 9.926000788605126e-05, "loss": 0.0952, "step": 10210 }, { "grad_norm": 0.45783913135528564, "learning_rate": 9.92571710156254e-05, "loss": 0.0828, "step": 10220 }, { "grad_norm": 0.5749744772911072, "learning_rate": 9.925432875850936e-05, "loss": 0.0954, "step": 10230 }, { "grad_norm": 0.9174307584762573, "learning_rate": 9.925148111501396e-05, "loss": 0.0939, "step": 10240 }, { "grad_norm": 1.0594162940979004, "learning_rate": 9.924862808545066e-05, "loss": 0.0887, "step": 10250 }, { "grad_norm": 0.5721908807754517, "learning_rate": 9.924576967013141e-05, "loss": 0.0828, "step": 10260 }, { "grad_norm": 0.6074411273002625, "learning_rate": 9.924290586936887e-05, "loss": 0.091, "step": 10270 }, { "grad_norm": 0.8776545524597168, "learning_rate": 9.924003668347614e-05, "loss": 0.0944, "step": 10280 }, { "grad_norm": 1.0187650918960571, "learning_rate": 9.923716211276704e-05, "loss": 0.0889, "step": 10290 }, { "grad_norm": 0.668039858341217, "learning_rate": 9.923428215755594e-05, "loss": 0.0803, "step": 10300 }, { "grad_norm": 0.6327479481697083, "learning_rate": 9.923139681815775e-05, "loss": 0.0936, "step": 10310 }, { "grad_norm": 0.7615401744842529, "learning_rate": 9.922850609488801e-05, "loss": 0.0828, "step": 10320 }, { "grad_norm": 0.6007476449012756, "learning_rate": 9.922560998806287e-05, "loss": 0.0882, "step": 10330 }, { "grad_norm": 0.6967192888259888, "learning_rate": 9.922270849799905e-05, "loss": 0.0983, "step": 10340 }, { "grad_norm": 0.6335152387619019, "learning_rate": 9.92198016250138e-05, "loss": 0.1013, "step": 10350 }, { "grad_norm": 0.7048348188400269, "learning_rate": 9.921688936942506e-05, "loss": 0.0857, "step": 10360 }, { "grad_norm": 0.5258139371871948, "learning_rate": 9.921397173155129e-05, "loss": 0.0833, "step": 10370 }, { "grad_norm": 0.611756443977356, "learning_rate": 9.921104871171157e-05, "loss": 0.0813, "step": 10380 }, { "grad_norm": 0.3706061840057373, "learning_rate": 9.920812031022554e-05, "loss": 0.0899, "step": 10390 }, { "grad_norm": 0.5152764916419983, "learning_rate": 9.920518652741348e-05, "loss": 0.0891, "step": 10400 }, { "grad_norm": 0.8358342051506042, "learning_rate": 9.920224736359618e-05, "loss": 0.0982, "step": 10410 }, { "grad_norm": 0.5505240559577942, "learning_rate": 9.91993028190951e-05, "loss": 0.1092, "step": 10420 }, { "grad_norm": 0.7324361801147461, "learning_rate": 9.919635289423222e-05, "loss": 0.0929, "step": 10430 }, { "grad_norm": 0.5969918966293335, "learning_rate": 9.919339758933015e-05, "loss": 0.0812, "step": 10440 }, { "grad_norm": 0.6015221476554871, "learning_rate": 9.919043690471209e-05, "loss": 0.0866, "step": 10450 }, { "grad_norm": 0.6156935691833496, "learning_rate": 9.91874708407018e-05, "loss": 0.102, "step": 10460 }, { "grad_norm": 0.640999972820282, "learning_rate": 9.918449939762367e-05, "loss": 0.0979, "step": 10470 }, { "grad_norm": 0.8005288243293762, "learning_rate": 9.91815225758026e-05, "loss": 0.096, "step": 10480 }, { "grad_norm": 0.5702918171882629, "learning_rate": 9.917854037556419e-05, "loss": 0.0931, "step": 10490 }, { "grad_norm": 0.6629457473754883, "learning_rate": 9.917555279723454e-05, "loss": 0.0874, "step": 10500 }, { "grad_norm": 0.6890270709991455, "learning_rate": 9.917255984114036e-05, "loss": 0.1065, "step": 10510 }, { "grad_norm": 0.5115802884101868, "learning_rate": 9.916956150760896e-05, "loss": 0.0758, "step": 10520 }, { "grad_norm": 0.45529526472091675, "learning_rate": 9.916655779696826e-05, "loss": 0.0899, "step": 10530 }, { "grad_norm": 0.49347683787345886, "learning_rate": 9.916354870954671e-05, "loss": 0.0785, "step": 10540 }, { "grad_norm": 0.6091944575309753, "learning_rate": 9.91605342456734e-05, "loss": 0.0953, "step": 10550 }, { "grad_norm": 0.5795434713363647, "learning_rate": 9.915751440567795e-05, "loss": 0.0875, "step": 10560 }, { "grad_norm": 0.5721539855003357, "learning_rate": 9.915448918989066e-05, "loss": 0.0791, "step": 10570 }, { "grad_norm": 0.5850348472595215, "learning_rate": 9.915145859864232e-05, "loss": 0.0818, "step": 10580 }, { "grad_norm": 0.6871451139450073, "learning_rate": 9.914842263226437e-05, "loss": 0.1116, "step": 10590 }, { "grad_norm": 0.4824204444885254, "learning_rate": 9.914538129108882e-05, "loss": 0.1017, "step": 10600 }, { "grad_norm": 0.625619649887085, "learning_rate": 9.914233457544825e-05, "loss": 0.1235, "step": 10610 }, { "grad_norm": 0.7964378595352173, "learning_rate": 9.913928248567586e-05, "loss": 0.1238, "step": 10620 }, { "grad_norm": 0.6144936084747314, "learning_rate": 9.913622502210542e-05, "loss": 0.0858, "step": 10630 }, { "grad_norm": 0.6682107448577881, "learning_rate": 9.913316218507128e-05, "loss": 0.1167, "step": 10640 }, { "grad_norm": 0.6257179379463196, "learning_rate": 9.91300939749084e-05, "loss": 0.1019, "step": 10650 }, { "grad_norm": 0.595827043056488, "learning_rate": 9.91270203919523e-05, "loss": 0.1146, "step": 10660 }, { "grad_norm": 0.9626283645629883, "learning_rate": 9.912394143653912e-05, "loss": 0.0966, "step": 10670 }, { "grad_norm": 1.1544820070266724, "learning_rate": 9.912085710900555e-05, "loss": 0.0876, "step": 10680 }, { "grad_norm": 0.4768997132778168, "learning_rate": 9.911776740968892e-05, "loss": 0.0882, "step": 10690 }, { "grad_norm": 0.5927450060844421, "learning_rate": 9.911467233892709e-05, "loss": 0.11, "step": 10700 }, { "grad_norm": 0.5010989904403687, "learning_rate": 9.911157189705853e-05, "loss": 0.0984, "step": 10710 }, { "grad_norm": 0.6516096591949463, "learning_rate": 9.910846608442229e-05, "loss": 0.0872, "step": 10720 }, { "grad_norm": 0.618870198726654, "learning_rate": 9.910535490135805e-05, "loss": 0.0845, "step": 10730 }, { "grad_norm": 0.8473320007324219, "learning_rate": 9.910223834820603e-05, "loss": 0.0879, "step": 10740 }, { "grad_norm": 0.7336768507957458, "learning_rate": 9.909911642530703e-05, "loss": 0.0978, "step": 10750 }, { "grad_norm": 0.49818193912506104, "learning_rate": 9.909598913300249e-05, "loss": 0.107, "step": 10760 }, { "grad_norm": 0.7225478887557983, "learning_rate": 9.909285647163438e-05, "loss": 0.1105, "step": 10770 }, { "grad_norm": 0.7446126937866211, "learning_rate": 9.908971844154531e-05, "loss": 0.0907, "step": 10780 }, { "grad_norm": 0.5878486037254333, "learning_rate": 9.908657504307843e-05, "loss": 0.0803, "step": 10790 }, { "grad_norm": 0.7121114134788513, "learning_rate": 9.908342627657751e-05, "loss": 0.1007, "step": 10800 }, { "grad_norm": 0.8139199614524841, "learning_rate": 9.908027214238689e-05, "loss": 0.09, "step": 10810 }, { "grad_norm": 0.445389986038208, "learning_rate": 9.90771126408515e-05, "loss": 0.0969, "step": 10820 }, { "grad_norm": 0.6025968790054321, "learning_rate": 9.907394777231685e-05, "loss": 0.0988, "step": 10830 }, { "grad_norm": 0.5426611304283142, "learning_rate": 9.907077753712905e-05, "loss": 0.1008, "step": 10840 }, { "grad_norm": 1.0305262804031372, "learning_rate": 9.906760193563482e-05, "loss": 0.1166, "step": 10850 }, { "grad_norm": 0.9333426356315613, "learning_rate": 9.906442096818139e-05, "loss": 0.0941, "step": 10860 }, { "grad_norm": 0.7930182814598083, "learning_rate": 9.906123463511665e-05, "loss": 0.1127, "step": 10870 }, { "grad_norm": 0.6005622744560242, "learning_rate": 9.905804293678907e-05, "loss": 0.0957, "step": 10880 }, { "grad_norm": 0.7701706290245056, "learning_rate": 9.905484587354766e-05, "loss": 0.0927, "step": 10890 }, { "grad_norm": 0.51541668176651, "learning_rate": 9.905164344574205e-05, "loss": 0.0952, "step": 10900 }, { "grad_norm": 0.6233410239219666, "learning_rate": 9.904843565372248e-05, "loss": 0.094, "step": 10910 }, { "grad_norm": 0.5792451500892639, "learning_rate": 9.904522249783972e-05, "loss": 0.0985, "step": 10920 }, { "grad_norm": 0.6934570670127869, "learning_rate": 9.904200397844517e-05, "loss": 0.1157, "step": 10930 }, { "grad_norm": 0.6973375082015991, "learning_rate": 9.903878009589078e-05, "loss": 0.096, "step": 10940 }, { "grad_norm": 0.7716220617294312, "learning_rate": 9.903555085052915e-05, "loss": 0.0824, "step": 10950 }, { "grad_norm": 0.6970009803771973, "learning_rate": 9.903231624271338e-05, "loss": 0.0815, "step": 10960 }, { "grad_norm": 0.47545817494392395, "learning_rate": 9.902907627279724e-05, "loss": 0.0814, "step": 10970 }, { "grad_norm": 0.5835632085800171, "learning_rate": 9.902583094113504e-05, "loss": 0.0885, "step": 10980 }, { "grad_norm": 0.5681585669517517, "learning_rate": 9.902258024808168e-05, "loss": 0.0863, "step": 10990 }, { "grad_norm": 0.7206337451934814, "learning_rate": 9.901932419399264e-05, "loss": 0.0912, "step": 11000 }, { "grad_norm": 0.5538045167922974, "learning_rate": 9.9016062779224e-05, "loss": 0.1036, "step": 11010 }, { "grad_norm": 0.7317919135093689, "learning_rate": 9.901279600413242e-05, "loss": 0.0935, "step": 11020 }, { "grad_norm": 0.6891184449195862, "learning_rate": 9.900952386907518e-05, "loss": 0.0836, "step": 11030 }, { "grad_norm": 0.7600693106651306, "learning_rate": 9.90062463744101e-05, "loss": 0.0831, "step": 11040 }, { "grad_norm": 0.603670597076416, "learning_rate": 9.900296352049558e-05, "loss": 0.0752, "step": 11050 }, { "grad_norm": 0.5624920129776001, "learning_rate": 9.899967530769065e-05, "loss": 0.1005, "step": 11060 }, { "grad_norm": 0.6907076835632324, "learning_rate": 9.899638173635489e-05, "loss": 0.085, "step": 11070 }, { "grad_norm": 0.684603750705719, "learning_rate": 9.899308280684849e-05, "loss": 0.0818, "step": 11080 }, { "grad_norm": 0.5818989872932434, "learning_rate": 9.898977851953222e-05, "loss": 0.089, "step": 11090 }, { "grad_norm": 0.545263409614563, "learning_rate": 9.898646887476741e-05, "loss": 0.0983, "step": 11100 }, { "grad_norm": 0.7645264267921448, "learning_rate": 9.898315387291603e-05, "loss": 0.1006, "step": 11110 }, { "grad_norm": 0.7239577174186707, "learning_rate": 9.89798335143406e-05, "loss": 0.0742, "step": 11120 }, { "grad_norm": 0.5492663979530334, "learning_rate": 9.897650779940419e-05, "loss": 0.0959, "step": 11130 }, { "grad_norm": 0.42965054512023926, "learning_rate": 9.897317672847054e-05, "loss": 0.0871, "step": 11140 }, { "grad_norm": 0.7733812928199768, "learning_rate": 9.89698403019039e-05, "loss": 0.0998, "step": 11150 }, { "grad_norm": 0.702705442905426, "learning_rate": 9.896649852006917e-05, "loss": 0.0945, "step": 11160 }, { "grad_norm": 0.7633158564567566, "learning_rate": 9.896315138333177e-05, "loss": 0.0892, "step": 11170 }, { "grad_norm": 0.9347698092460632, "learning_rate": 9.895979889205774e-05, "loss": 0.0868, "step": 11180 }, { "grad_norm": 0.6355254650115967, "learning_rate": 9.895644104661372e-05, "loss": 0.0863, "step": 11190 }, { "grad_norm": 0.6248782873153687, "learning_rate": 9.895307784736691e-05, "loss": 0.0946, "step": 11200 }, { "grad_norm": 0.7546966075897217, "learning_rate": 9.894970929468512e-05, "loss": 0.0893, "step": 11210 }, { "grad_norm": 0.7024227380752563, "learning_rate": 9.89463353889367e-05, "loss": 0.0982, "step": 11220 }, { "grad_norm": 0.7054046988487244, "learning_rate": 9.894295613049065e-05, "loss": 0.1082, "step": 11230 }, { "grad_norm": 0.8200047016143799, "learning_rate": 9.893957151971649e-05, "loss": 0.107, "step": 11240 }, { "grad_norm": 0.7037248611450195, "learning_rate": 9.893618155698436e-05, "loss": 0.0883, "step": 11250 }, { "grad_norm": 0.6991463899612427, "learning_rate": 9.8932786242665e-05, "loss": 0.0822, "step": 11260 }, { "grad_norm": 0.5129389762878418, "learning_rate": 9.89293855771297e-05, "loss": 0.0988, "step": 11270 }, { "grad_norm": 0.7365736365318298, "learning_rate": 9.892597956075036e-05, "loss": 0.0984, "step": 11280 }, { "grad_norm": 0.5592517852783203, "learning_rate": 9.892256819389947e-05, "loss": 0.0818, "step": 11290 }, { "grad_norm": 0.616527259349823, "learning_rate": 9.891915147695006e-05, "loss": 0.0824, "step": 11300 }, { "grad_norm": 0.7448214888572693, "learning_rate": 9.891572941027577e-05, "loss": 0.0919, "step": 11310 }, { "grad_norm": 0.7621772885322571, "learning_rate": 9.89123019942509e-05, "loss": 0.0816, "step": 11320 }, { "grad_norm": 0.8349685668945312, "learning_rate": 9.89088692292502e-05, "loss": 0.0814, "step": 11330 }, { "grad_norm": 0.5605748295783997, "learning_rate": 9.89054311156491e-05, "loss": 0.0749, "step": 11340 }, { "grad_norm": 0.7286494374275208, "learning_rate": 9.890198765382357e-05, "loss": 0.1147, "step": 11350 }, { "grad_norm": 0.7758250832557678, "learning_rate": 9.889853884415021e-05, "loss": 0.0962, "step": 11360 }, { "grad_norm": 0.5775719881057739, "learning_rate": 9.889508468700614e-05, "loss": 0.0773, "step": 11370 }, { "grad_norm": 0.442965567111969, "learning_rate": 9.889162518276915e-05, "loss": 0.078, "step": 11380 }, { "grad_norm": 0.5345435738563538, "learning_rate": 9.888816033181752e-05, "loss": 0.0875, "step": 11390 }, { "grad_norm": 0.6884585618972778, "learning_rate": 9.888469013453018e-05, "loss": 0.107, "step": 11400 }, { "grad_norm": 0.9086427092552185, "learning_rate": 9.888121459128663e-05, "loss": 0.0962, "step": 11410 }, { "grad_norm": 0.7400583624839783, "learning_rate": 9.887773370246693e-05, "loss": 0.1048, "step": 11420 }, { "grad_norm": 0.5513676404953003, "learning_rate": 9.887424746845177e-05, "loss": 0.0909, "step": 11430 }, { "grad_norm": 0.5970160365104675, "learning_rate": 9.887075588962239e-05, "loss": 0.0904, "step": 11440 }, { "grad_norm": 0.4844469428062439, "learning_rate": 9.88672589663606e-05, "loss": 0.0997, "step": 11450 }, { "grad_norm": 0.6840354800224304, "learning_rate": 9.886375669904886e-05, "loss": 0.0957, "step": 11460 }, { "grad_norm": 0.6376686692237854, "learning_rate": 9.886024908807014e-05, "loss": 0.0865, "step": 11470 }, { "grad_norm": 1.034770131111145, "learning_rate": 9.885673613380806e-05, "loss": 0.0797, "step": 11480 }, { "grad_norm": 0.6044542193412781, "learning_rate": 9.885321783664676e-05, "loss": 0.0798, "step": 11490 }, { "grad_norm": 0.601992666721344, "learning_rate": 9.884969419697101e-05, "loss": 0.0841, "step": 11500 }, { "grad_norm": 0.5227922797203064, "learning_rate": 9.884616521516614e-05, "loss": 0.0855, "step": 11510 }, { "grad_norm": 0.5741783380508423, "learning_rate": 9.88426308916181e-05, "loss": 0.0932, "step": 11520 }, { "grad_norm": 0.5382272005081177, "learning_rate": 9.883909122671335e-05, "loss": 0.0795, "step": 11530 }, { "grad_norm": 0.5483359098434448, "learning_rate": 9.883554622083904e-05, "loss": 0.0786, "step": 11540 }, { "grad_norm": 0.675376832485199, "learning_rate": 9.88319958743828e-05, "loss": 0.0927, "step": 11550 }, { "grad_norm": 1.6595220565795898, "learning_rate": 9.882844018773291e-05, "loss": 0.0852, "step": 11560 }, { "grad_norm": 0.601567804813385, "learning_rate": 9.882487916127823e-05, "loss": 0.0963, "step": 11570 }, { "grad_norm": 0.6006075143814087, "learning_rate": 9.882131279540815e-05, "loss": 0.1141, "step": 11580 }, { "grad_norm": 0.7360178828239441, "learning_rate": 9.881774109051271e-05, "loss": 0.0993, "step": 11590 }, { "grad_norm": 0.6750975847244263, "learning_rate": 9.881416404698252e-05, "loss": 0.0929, "step": 11600 }, { "grad_norm": 0.5747734904289246, "learning_rate": 9.881058166520873e-05, "loss": 0.0957, "step": 11610 }, { "grad_norm": 0.7397626638412476, "learning_rate": 9.880699394558311e-05, "loss": 0.115, "step": 11620 }, { "grad_norm": 0.7284901142120361, "learning_rate": 9.880340088849801e-05, "loss": 0.1045, "step": 11630 }, { "grad_norm": 0.5653693079948425, "learning_rate": 9.879980249434637e-05, "loss": 0.0956, "step": 11640 }, { "grad_norm": 1.1581757068634033, "learning_rate": 9.879619876352168e-05, "loss": 0.0957, "step": 11650 }, { "grad_norm": 0.6582175493240356, "learning_rate": 9.879258969641809e-05, "loss": 0.097, "step": 11660 }, { "grad_norm": 0.7237258553504944, "learning_rate": 9.878897529343023e-05, "loss": 0.0846, "step": 11670 }, { "grad_norm": 0.7537770867347717, "learning_rate": 9.878535555495338e-05, "loss": 0.096, "step": 11680 }, { "grad_norm": 0.876156747341156, "learning_rate": 9.87817304813834e-05, "loss": 0.1056, "step": 11690 }, { "grad_norm": 0.4927794933319092, "learning_rate": 9.877810007311671e-05, "loss": 0.0933, "step": 11700 }, { "grad_norm": 0.9722226858139038, "learning_rate": 9.877446433055035e-05, "loss": 0.0912, "step": 11710 }, { "grad_norm": 0.6577757596969604, "learning_rate": 9.877082325408191e-05, "loss": 0.0938, "step": 11720 }, { "grad_norm": 0.9302864074707031, "learning_rate": 9.876717684410954e-05, "loss": 0.0735, "step": 11730 }, { "grad_norm": 0.7510039210319519, "learning_rate": 9.876352510103204e-05, "loss": 0.0869, "step": 11740 }, { "grad_norm": 0.6066937446594238, "learning_rate": 9.875986802524875e-05, "loss": 0.0745, "step": 11750 }, { "grad_norm": 0.5284473299980164, "learning_rate": 9.87562056171596e-05, "loss": 0.077, "step": 11760 }, { "grad_norm": 0.5534787774085999, "learning_rate": 9.875253787716511e-05, "loss": 0.0903, "step": 11770 }, { "grad_norm": 0.7668598890304565, "learning_rate": 9.874886480566637e-05, "loss": 0.1021, "step": 11780 }, { "grad_norm": 0.45569995045661926, "learning_rate": 9.874518640306507e-05, "loss": 0.081, "step": 11790 }, { "grad_norm": 0.5337285995483398, "learning_rate": 9.874150266976347e-05, "loss": 0.0926, "step": 11800 }, { "grad_norm": 0.5620971322059631, "learning_rate": 9.873781360616443e-05, "loss": 0.0969, "step": 11810 }, { "grad_norm": 0.8001932501792908, "learning_rate": 9.873411921267137e-05, "loss": 0.0847, "step": 11820 }, { "grad_norm": 0.5704538822174072, "learning_rate": 9.873041948968829e-05, "loss": 0.0906, "step": 11830 }, { "grad_norm": 0.5399866700172424, "learning_rate": 9.872671443761981e-05, "loss": 0.0796, "step": 11840 }, { "grad_norm": 0.5014262199401855, "learning_rate": 9.872300405687109e-05, "loss": 0.08, "step": 11850 }, { "grad_norm": 0.5930140614509583, "learning_rate": 9.871928834784792e-05, "loss": 0.0922, "step": 11860 }, { "grad_norm": 0.5077852010726929, "learning_rate": 9.871556731095661e-05, "loss": 0.0985, "step": 11870 }, { "grad_norm": 0.6745296716690063, "learning_rate": 9.871184094660411e-05, "loss": 0.0997, "step": 11880 }, { "grad_norm": 0.5646390318870544, "learning_rate": 9.870810925519791e-05, "loss": 0.0847, "step": 11890 }, { "grad_norm": 0.5115732550621033, "learning_rate": 9.870437223714612e-05, "loss": 0.0888, "step": 11900 }, { "grad_norm": 0.4844681918621063, "learning_rate": 9.87006298928574e-05, "loss": 0.088, "step": 11910 }, { "grad_norm": 0.5186592936515808, "learning_rate": 9.869688222274103e-05, "loss": 0.0734, "step": 11920 }, { "grad_norm": 0.6488363146781921, "learning_rate": 9.869312922720681e-05, "loss": 0.0864, "step": 11930 }, { "grad_norm": 0.44567784667015076, "learning_rate": 9.868937090666521e-05, "loss": 0.0841, "step": 11940 }, { "grad_norm": 0.45003408193588257, "learning_rate": 9.86856072615272e-05, "loss": 0.0718, "step": 11950 }, { "grad_norm": 0.6732233166694641, "learning_rate": 9.868183829220438e-05, "loss": 0.081, "step": 11960 }, { "grad_norm": 0.8994625210762024, "learning_rate": 9.867806399910893e-05, "loss": 0.0808, "step": 11970 }, { "grad_norm": 0.7633582949638367, "learning_rate": 9.867428438265356e-05, "loss": 0.0865, "step": 11980 }, { "grad_norm": 0.766821563243866, "learning_rate": 9.867049944325165e-05, "loss": 0.0953, "step": 11990 }, { "grad_norm": 0.784607470035553, "learning_rate": 9.86667091813171e-05, "loss": 0.1046, "step": 12000 }, { "grad_norm": 0.9471831321716309, "learning_rate": 9.866291359726438e-05, "loss": 0.086, "step": 12010 }, { "grad_norm": 0.5250786542892456, "learning_rate": 9.865911269150861e-05, "loss": 0.0934, "step": 12020 }, { "grad_norm": 0.47927042841911316, "learning_rate": 9.865530646446544e-05, "loss": 0.0788, "step": 12030 }, { "grad_norm": 0.5698023438453674, "learning_rate": 9.86514949165511e-05, "loss": 0.0933, "step": 12040 }, { "grad_norm": 0.5109734535217285, "learning_rate": 9.864767804818243e-05, "loss": 0.0852, "step": 12050 }, { "grad_norm": 0.7283066511154175, "learning_rate": 9.86438558597768e-05, "loss": 0.1009, "step": 12060 }, { "grad_norm": 0.5326876640319824, "learning_rate": 9.864002835175225e-05, "loss": 0.0944, "step": 12070 }, { "grad_norm": 0.6569128036499023, "learning_rate": 9.863619552452734e-05, "loss": 0.0756, "step": 12080 }, { "grad_norm": 0.691970705986023, "learning_rate": 9.863235737852119e-05, "loss": 0.0731, "step": 12090 }, { "grad_norm": 0.9805426001548767, "learning_rate": 9.862851391415356e-05, "loss": 0.0855, "step": 12100 }, { "grad_norm": 0.4859282374382019, "learning_rate": 9.862466513184477e-05, "loss": 0.0729, "step": 12110 }, { "grad_norm": 0.6378284692764282, "learning_rate": 9.86208110320157e-05, "loss": 0.0763, "step": 12120 }, { "grad_norm": 0.5409536957740784, "learning_rate": 9.861695161508784e-05, "loss": 0.0844, "step": 12130 }, { "grad_norm": 0.5743542313575745, "learning_rate": 9.861308688148324e-05, "loss": 0.0852, "step": 12140 }, { "grad_norm": 0.5265329480171204, "learning_rate": 9.860921683162455e-05, "loss": 0.0859, "step": 12150 }, { "grad_norm": 0.5716769099235535, "learning_rate": 9.860534146593499e-05, "loss": 0.079, "step": 12160 }, { "grad_norm": 0.6388117671012878, "learning_rate": 9.860146078483836e-05, "loss": 0.0839, "step": 12170 }, { "grad_norm": 0.68617844581604, "learning_rate": 9.859757478875905e-05, "loss": 0.0838, "step": 12180 }, { "grad_norm": 0.6014553308486938, "learning_rate": 9.859368347812204e-05, "loss": 0.0807, "step": 12190 }, { "grad_norm": 0.7333876490592957, "learning_rate": 9.858978685335285e-05, "loss": 0.0787, "step": 12200 }, { "grad_norm": 0.5736202597618103, "learning_rate": 9.858588491487763e-05, "loss": 0.1036, "step": 12210 }, { "grad_norm": 0.4692656993865967, "learning_rate": 9.858197766312308e-05, "loss": 0.0729, "step": 12220 }, { "grad_norm": 0.7126598954200745, "learning_rate": 9.857806509851649e-05, "loss": 0.0894, "step": 12230 }, { "grad_norm": 0.6037431359291077, "learning_rate": 9.857414722148574e-05, "loss": 0.0948, "step": 12240 }, { "grad_norm": 0.47358524799346924, "learning_rate": 9.857022403245928e-05, "loss": 0.1086, "step": 12250 }, { "grad_norm": 0.5880676507949829, "learning_rate": 9.856629553186615e-05, "loss": 0.1041, "step": 12260 }, { "grad_norm": 0.7245994806289673, "learning_rate": 9.856236172013595e-05, "loss": 0.0976, "step": 12270 }, { "grad_norm": 0.4170815646648407, "learning_rate": 9.85584225976989e-05, "loss": 0.0806, "step": 12280 }, { "grad_norm": 0.9039942026138306, "learning_rate": 9.855447816498575e-05, "loss": 0.1024, "step": 12290 }, { "grad_norm": 0.6899275183677673, "learning_rate": 9.855052842242787e-05, "loss": 0.0732, "step": 12300 }, { "grad_norm": 0.6975042223930359, "learning_rate": 9.85465733704572e-05, "loss": 0.0838, "step": 12310 }, { "grad_norm": 0.7056939005851746, "learning_rate": 9.854261300950624e-05, "loss": 0.1008, "step": 12320 }, { "grad_norm": 0.4810272455215454, "learning_rate": 9.853864734000813e-05, "loss": 0.0799, "step": 12330 }, { "grad_norm": 0.6817892789840698, "learning_rate": 9.85346763623965e-05, "loss": 0.1, "step": 12340 }, { "grad_norm": 0.5545966625213623, "learning_rate": 9.853070007710564e-05, "loss": 0.0847, "step": 12350 }, { "grad_norm": 0.8821640610694885, "learning_rate": 9.85267184845704e-05, "loss": 0.0723, "step": 12360 }, { "grad_norm": 0.6299097537994385, "learning_rate": 9.852273158522616e-05, "loss": 0.1086, "step": 12370 }, { "grad_norm": 0.4866696894168854, "learning_rate": 9.851873937950896e-05, "loss": 0.0751, "step": 12380 }, { "grad_norm": 0.7669113278388977, "learning_rate": 9.851474186785537e-05, "loss": 0.1014, "step": 12390 }, { "grad_norm": 0.5344756841659546, "learning_rate": 9.851073905070254e-05, "loss": 0.0849, "step": 12400 }, { "grad_norm": 1.0374774932861328, "learning_rate": 9.850673092848824e-05, "loss": 0.1003, "step": 12410 }, { "grad_norm": 0.5638212561607361, "learning_rate": 9.850271750165077e-05, "loss": 0.0733, "step": 12420 }, { "grad_norm": 1.7332415580749512, "learning_rate": 9.849869877062902e-05, "loss": 0.097, "step": 12430 }, { "grad_norm": 0.7051153182983398, "learning_rate": 9.849467473586252e-05, "loss": 0.1086, "step": 12440 }, { "grad_norm": 1.0755927562713623, "learning_rate": 9.849064539779127e-05, "loss": 0.0834, "step": 12450 }, { "grad_norm": 0.5359546542167664, "learning_rate": 9.848661075685594e-05, "loss": 0.0959, "step": 12460 }, { "grad_norm": 0.7489425539970398, "learning_rate": 9.848257081349778e-05, "loss": 0.0914, "step": 12470 }, { "grad_norm": 0.5672597289085388, "learning_rate": 9.847852556815856e-05, "loss": 0.0842, "step": 12480 }, { "grad_norm": 1.20534348487854, "learning_rate": 9.847447502128067e-05, "loss": 0.0849, "step": 12490 }, { "grad_norm": 0.5741618871688843, "learning_rate": 9.847041917330708e-05, "loss": 0.0841, "step": 12500 }, { "grad_norm": 0.6837983727455139, "learning_rate": 9.846635802468132e-05, "loss": 0.1141, "step": 12510 }, { "grad_norm": 0.5280585289001465, "learning_rate": 9.84622915758475e-05, "loss": 0.0799, "step": 12520 }, { "grad_norm": 0.5339528322219849, "learning_rate": 9.845821982725034e-05, "loss": 0.0783, "step": 12530 }, { "grad_norm": 0.5328267216682434, "learning_rate": 9.845414277933514e-05, "loss": 0.0899, "step": 12540 }, { "grad_norm": 0.675089418888092, "learning_rate": 9.845006043254771e-05, "loss": 0.0931, "step": 12550 }, { "grad_norm": 0.8504695296287537, "learning_rate": 9.844597278733451e-05, "loss": 0.0977, "step": 12560 }, { "grad_norm": 0.5808284282684326, "learning_rate": 9.844187984414259e-05, "loss": 0.0826, "step": 12570 }, { "grad_norm": 0.7397013902664185, "learning_rate": 9.84377816034195e-05, "loss": 0.0894, "step": 12580 }, { "grad_norm": 0.5105061531066895, "learning_rate": 9.843367806561345e-05, "loss": 0.0834, "step": 12590 }, { "grad_norm": 0.7662990689277649, "learning_rate": 9.842956923117317e-05, "loss": 0.0852, "step": 12600 }, { "grad_norm": 0.49413084983825684, "learning_rate": 9.842545510054802e-05, "loss": 0.0701, "step": 12610 }, { "grad_norm": 0.695483922958374, "learning_rate": 9.842133567418792e-05, "loss": 0.0738, "step": 12620 }, { "grad_norm": 0.4309834837913513, "learning_rate": 9.841721095254333e-05, "loss": 0.0772, "step": 12630 }, { "grad_norm": 0.4740822911262512, "learning_rate": 9.841308093606537e-05, "loss": 0.0828, "step": 12640 }, { "grad_norm": 0.546536922454834, "learning_rate": 9.840894562520565e-05, "loss": 0.0747, "step": 12650 }, { "grad_norm": 0.7811991572380066, "learning_rate": 9.840480502041642e-05, "loss": 0.0948, "step": 12660 }, { "grad_norm": 0.6354491114616394, "learning_rate": 9.840065912215049e-05, "loss": 0.0727, "step": 12670 }, { "grad_norm": 0.7838616371154785, "learning_rate": 9.839650793086124e-05, "loss": 0.0985, "step": 12680 }, { "grad_norm": 0.48885273933410645, "learning_rate": 9.839235144700265e-05, "loss": 0.0774, "step": 12690 }, { "grad_norm": 0.5276165008544922, "learning_rate": 9.838818967102926e-05, "loss": 0.0733, "step": 12700 }, { "grad_norm": 0.5244992971420288, "learning_rate": 9.83840226033962e-05, "loss": 0.0829, "step": 12710 }, { "grad_norm": 0.5339284539222717, "learning_rate": 9.837985024455918e-05, "loss": 0.0799, "step": 12720 }, { "grad_norm": 0.613282322883606, "learning_rate": 9.837567259497447e-05, "loss": 0.095, "step": 12730 }, { "grad_norm": 0.49110937118530273, "learning_rate": 9.837148965509894e-05, "loss": 0.0806, "step": 12740 }, { "grad_norm": 0.5509555339813232, "learning_rate": 9.836730142539001e-05, "loss": 0.0775, "step": 12750 }, { "grad_norm": 0.785368800163269, "learning_rate": 9.836310790630574e-05, "loss": 0.0992, "step": 12760 }, { "grad_norm": 0.6723958253860474, "learning_rate": 9.83589090983047e-05, "loss": 0.0894, "step": 12770 }, { "grad_norm": 0.6663535237312317, "learning_rate": 9.835470500184605e-05, "loss": 0.0825, "step": 12780 }, { "grad_norm": 0.5447731614112854, "learning_rate": 9.835049561738957e-05, "loss": 0.0839, "step": 12790 }, { "grad_norm": 0.5909860134124756, "learning_rate": 9.834628094539558e-05, "loss": 0.0687, "step": 12800 }, { "grad_norm": 0.6781706213951111, "learning_rate": 9.834206098632499e-05, "loss": 0.0874, "step": 12810 }, { "grad_norm": 0.6117033362388611, "learning_rate": 9.833783574063931e-05, "loss": 0.0967, "step": 12820 }, { "grad_norm": 0.6763530969619751, "learning_rate": 9.833360520880058e-05, "loss": 0.0921, "step": 12830 }, { "grad_norm": 0.5614015460014343, "learning_rate": 9.832936939127144e-05, "loss": 0.0767, "step": 12840 }, { "grad_norm": 0.4871125817298889, "learning_rate": 9.832512828851515e-05, "loss": 0.0883, "step": 12850 }, { "grad_norm": 0.7543162107467651, "learning_rate": 9.832088190099546e-05, "loss": 0.0817, "step": 12860 }, { "grad_norm": 0.6708624958992004, "learning_rate": 9.831663022917679e-05, "loss": 0.0916, "step": 12870 }, { "grad_norm": 0.6130321621894836, "learning_rate": 9.831237327352407e-05, "loss": 0.0737, "step": 12880 }, { "grad_norm": 0.4700535535812378, "learning_rate": 9.830811103450286e-05, "loss": 0.0764, "step": 12890 }, { "grad_norm": 0.6374039053916931, "learning_rate": 9.830384351257924e-05, "loss": 0.0803, "step": 12900 }, { "grad_norm": 0.9751349687576294, "learning_rate": 9.829957070821993e-05, "loss": 0.0858, "step": 12910 }, { "grad_norm": 0.6798281073570251, "learning_rate": 9.829529262189218e-05, "loss": 0.0753, "step": 12920 }, { "grad_norm": 0.6025866866111755, "learning_rate": 9.829100925406385e-05, "loss": 0.0921, "step": 12930 }, { "grad_norm": 0.5719643831253052, "learning_rate": 9.828672060520333e-05, "loss": 0.087, "step": 12940 }, { "grad_norm": 0.38822802901268005, "learning_rate": 9.828242667577966e-05, "loss": 0.0646, "step": 12950 }, { "grad_norm": 0.5379580855369568, "learning_rate": 9.82781274662624e-05, "loss": 0.0834, "step": 12960 }, { "grad_norm": 0.6533995270729065, "learning_rate": 9.82738229771217e-05, "loss": 0.0788, "step": 12970 }, { "grad_norm": 0.5394190549850464, "learning_rate": 9.826951320882829e-05, "loss": 0.0707, "step": 12980 }, { "grad_norm": 0.8133512139320374, "learning_rate": 9.826519816185351e-05, "loss": 0.0743, "step": 12990 }, { "grad_norm": 0.6731049418449402, "learning_rate": 9.826087783666921e-05, "loss": 0.0805, "step": 13000 }, { "grad_norm": 0.6244357824325562, "learning_rate": 9.825655223374787e-05, "loss": 0.1031, "step": 13010 }, { "grad_norm": 0.8119274377822876, "learning_rate": 9.825222135356253e-05, "loss": 0.1017, "step": 13020 }, { "grad_norm": 0.6713312268257141, "learning_rate": 9.82478851965868e-05, "loss": 0.1034, "step": 13030 }, { "grad_norm": 0.507358729839325, "learning_rate": 9.82435437632949e-05, "loss": 0.0727, "step": 13040 }, { "grad_norm": 0.5253500938415527, "learning_rate": 9.823919705416158e-05, "loss": 0.0696, "step": 13050 }, { "grad_norm": 0.8477048873901367, "learning_rate": 9.82348450696622e-05, "loss": 0.0987, "step": 13060 }, { "grad_norm": 0.5870533585548401, "learning_rate": 9.823048781027268e-05, "loss": 0.0967, "step": 13070 }, { "grad_norm": 0.8485523462295532, "learning_rate": 9.822612527646953e-05, "loss": 0.0834, "step": 13080 }, { "grad_norm": 0.7002829909324646, "learning_rate": 9.822175746872984e-05, "loss": 0.0836, "step": 13090 }, { "grad_norm": 0.99162358045578, "learning_rate": 9.821738438753123e-05, "loss": 0.0924, "step": 13100 }, { "grad_norm": 0.7908586263656616, "learning_rate": 9.821300603335196e-05, "loss": 0.0826, "step": 13110 }, { "grad_norm": 0.5548237562179565, "learning_rate": 9.820862240667085e-05, "loss": 0.0779, "step": 13120 }, { "grad_norm": 0.4645843207836151, "learning_rate": 9.820423350796726e-05, "loss": 0.0846, "step": 13130 }, { "grad_norm": 0.6327704787254333, "learning_rate": 9.819983933772118e-05, "loss": 0.0899, "step": 13140 }, { "grad_norm": 0.6996641755104065, "learning_rate": 9.819543989641314e-05, "loss": 0.09, "step": 13150 }, { "grad_norm": 0.47621625661849976, "learning_rate": 9.819103518452423e-05, "loss": 0.0743, "step": 13160 }, { "grad_norm": 0.8074049949645996, "learning_rate": 9.818662520253618e-05, "loss": 0.0797, "step": 13170 }, { "grad_norm": 0.5175573825836182, "learning_rate": 9.818220995093126e-05, "loss": 0.0759, "step": 13180 }, { "grad_norm": 1.0822869539260864, "learning_rate": 9.817778943019228e-05, "loss": 0.0928, "step": 13190 }, { "grad_norm": 0.6821334958076477, "learning_rate": 9.81733636408027e-05, "loss": 0.0943, "step": 13200 }, { "grad_norm": 0.5483928918838501, "learning_rate": 9.816893258324649e-05, "loss": 0.0691, "step": 13210 }, { "grad_norm": 0.5008249878883362, "learning_rate": 9.816449625800823e-05, "loss": 0.0778, "step": 13220 }, { "grad_norm": 0.7334038019180298, "learning_rate": 9.816005466557308e-05, "loss": 0.0934, "step": 13230 }, { "grad_norm": 0.6747305989265442, "learning_rate": 9.815560780642674e-05, "loss": 0.1014, "step": 13240 }, { "grad_norm": 0.4552426040172577, "learning_rate": 9.815115568105555e-05, "loss": 0.0721, "step": 13250 }, { "grad_norm": 0.8085271716117859, "learning_rate": 9.814669828994638e-05, "loss": 0.0945, "step": 13260 }, { "grad_norm": 0.6231966018676758, "learning_rate": 9.814223563358665e-05, "loss": 0.0771, "step": 13270 }, { "grad_norm": 0.5498697757720947, "learning_rate": 9.813776771246443e-05, "loss": 0.087, "step": 13280 }, { "grad_norm": 1.158875584602356, "learning_rate": 9.813329452706829e-05, "loss": 0.0851, "step": 13290 }, { "grad_norm": 0.7572137117385864, "learning_rate": 9.812881607788744e-05, "loss": 0.0813, "step": 13300 }, { "grad_norm": 0.5213229060173035, "learning_rate": 9.812433236541163e-05, "loss": 0.0848, "step": 13310 }, { "grad_norm": 0.6820774078369141, "learning_rate": 9.811984339013116e-05, "loss": 0.0932, "step": 13320 }, { "grad_norm": 0.43448200821876526, "learning_rate": 9.811534915253698e-05, "loss": 0.0844, "step": 13330 }, { "grad_norm": 0.6832419633865356, "learning_rate": 9.811084965312056e-05, "loss": 0.0935, "step": 13340 }, { "grad_norm": 0.6176707148551941, "learning_rate": 9.810634489237396e-05, "loss": 0.0756, "step": 13350 }, { "grad_norm": 0.5988872647285461, "learning_rate": 9.81018348707898e-05, "loss": 0.1017, "step": 13360 }, { "grad_norm": 0.6787335276603699, "learning_rate": 9.809731958886131e-05, "loss": 0.0856, "step": 13370 }, { "grad_norm": 0.7802730202674866, "learning_rate": 9.809279904708224e-05, "loss": 0.0926, "step": 13380 }, { "grad_norm": 0.5637101531028748, "learning_rate": 9.808827324594699e-05, "loss": 0.09, "step": 13390 }, { "grad_norm": 0.5298236608505249, "learning_rate": 9.808374218595046e-05, "loss": 0.0996, "step": 13400 }, { "grad_norm": 0.5982593894004822, "learning_rate": 9.80792058675882e-05, "loss": 0.0906, "step": 13410 }, { "grad_norm": 0.500018835067749, "learning_rate": 9.807466429135627e-05, "loss": 0.0829, "step": 13420 }, { "grad_norm": 0.42843490839004517, "learning_rate": 9.807011745775132e-05, "loss": 0.078, "step": 13430 }, { "grad_norm": 0.5604588389396667, "learning_rate": 9.806556536727061e-05, "loss": 0.0728, "step": 13440 }, { "grad_norm": 0.6319361925125122, "learning_rate": 9.806100802041193e-05, "loss": 0.0977, "step": 13450 }, { "grad_norm": 0.5365420579910278, "learning_rate": 9.805644541767368e-05, "loss": 0.0878, "step": 13460 }, { "grad_norm": 0.492648184299469, "learning_rate": 9.805187755955478e-05, "loss": 0.0828, "step": 13470 }, { "grad_norm": 0.46808549761772156, "learning_rate": 9.804730444655483e-05, "loss": 0.0862, "step": 13480 }, { "grad_norm": 0.4848829507827759, "learning_rate": 9.804272607917388e-05, "loss": 0.0906, "step": 13490 }, { "grad_norm": 0.44883349537849426, "learning_rate": 9.803814245791265e-05, "loss": 0.0682, "step": 13500 }, { "grad_norm": 0.7761044502258301, "learning_rate": 9.803355358327239e-05, "loss": 0.0891, "step": 13510 }, { "grad_norm": 0.6089494228363037, "learning_rate": 9.802895945575492e-05, "loss": 0.0908, "step": 13520 }, { "grad_norm": 0.7532863020896912, "learning_rate": 9.802436007586266e-05, "loss": 0.1023, "step": 13530 }, { "grad_norm": 0.5746718049049377, "learning_rate": 9.801975544409858e-05, "loss": 0.0784, "step": 13540 }, { "grad_norm": 0.9325221180915833, "learning_rate": 9.801514556096625e-05, "loss": 0.0863, "step": 13550 }, { "grad_norm": 0.802284300327301, "learning_rate": 9.801053042696977e-05, "loss": 0.0999, "step": 13560 }, { "grad_norm": 0.5754520893096924, "learning_rate": 9.800591004261388e-05, "loss": 0.084, "step": 13570 }, { "grad_norm": 0.7428523302078247, "learning_rate": 9.800128440840385e-05, "loss": 0.0955, "step": 13580 }, { "grad_norm": 0.6734499335289001, "learning_rate": 9.799665352484552e-05, "loss": 0.0936, "step": 13590 }, { "grad_norm": 0.6788382530212402, "learning_rate": 9.799201739244532e-05, "loss": 0.1055, "step": 13600 }, { "grad_norm": 0.6081355214118958, "learning_rate": 9.798737601171025e-05, "loss": 0.0818, "step": 13610 }, { "grad_norm": 0.90184485912323, "learning_rate": 9.79827293831479e-05, "loss": 0.1003, "step": 13620 }, { "grad_norm": 0.9339227676391602, "learning_rate": 9.797807750726638e-05, "loss": 0.086, "step": 13630 }, { "grad_norm": 0.6425230503082275, "learning_rate": 9.797342038457446e-05, "loss": 0.0885, "step": 13640 }, { "grad_norm": 0.7304977774620056, "learning_rate": 9.796875801558141e-05, "loss": 0.107, "step": 13650 }, { "grad_norm": 0.8900826573371887, "learning_rate": 9.79640904007971e-05, "loss": 0.0964, "step": 13660 }, { "grad_norm": 0.7446228861808777, "learning_rate": 9.795941754073199e-05, "loss": 0.0774, "step": 13670 }, { "grad_norm": 0.5981226563453674, "learning_rate": 9.795473943589705e-05, "loss": 0.0736, "step": 13680 }, { "grad_norm": 0.7523760795593262, "learning_rate": 9.795005608680394e-05, "loss": 0.0936, "step": 13690 }, { "grad_norm": 1.1383150815963745, "learning_rate": 9.794536749396477e-05, "loss": 0.095, "step": 13700 }, { "grad_norm": 0.5246384143829346, "learning_rate": 9.79406736578923e-05, "loss": 0.0869, "step": 13710 }, { "grad_norm": 0.691041111946106, "learning_rate": 9.793597457909984e-05, "loss": 0.0801, "step": 13720 }, { "grad_norm": 0.4322741627693176, "learning_rate": 9.793127025810127e-05, "loss": 0.0778, "step": 13730 }, { "grad_norm": 0.6231467723846436, "learning_rate": 9.792656069541104e-05, "loss": 0.0717, "step": 13740 }, { "grad_norm": 0.7701060175895691, "learning_rate": 9.79218458915442e-05, "loss": 0.0845, "step": 13750 }, { "grad_norm": 0.5630218982696533, "learning_rate": 9.791712584701634e-05, "loss": 0.0907, "step": 13760 }, { "grad_norm": 0.667775571346283, "learning_rate": 9.791240056234364e-05, "loss": 0.0889, "step": 13770 }, { "grad_norm": 0.5519061088562012, "learning_rate": 9.790767003804283e-05, "loss": 0.0938, "step": 13780 }, { "grad_norm": 0.5860850214958191, "learning_rate": 9.790293427463126e-05, "loss": 0.0748, "step": 13790 }, { "grad_norm": 0.5549200177192688, "learning_rate": 9.789819327262684e-05, "loss": 0.0764, "step": 13800 }, { "grad_norm": 0.5464663505554199, "learning_rate": 9.7893447032548e-05, "loss": 0.0849, "step": 13810 }, { "grad_norm": 0.877081036567688, "learning_rate": 9.78886955549138e-05, "loss": 0.0852, "step": 13820 }, { "grad_norm": 0.5237054824829102, "learning_rate": 9.788393884024387e-05, "loss": 0.0849, "step": 13830 }, { "grad_norm": 0.6160337328910828, "learning_rate": 9.787917688905836e-05, "loss": 0.0908, "step": 13840 }, { "grad_norm": 0.6209823489189148, "learning_rate": 9.787440970187807e-05, "loss": 0.0804, "step": 13850 }, { "grad_norm": 0.6589679718017578, "learning_rate": 9.786963727922429e-05, "loss": 0.0848, "step": 13860 }, { "grad_norm": 0.6978249549865723, "learning_rate": 9.786485962161897e-05, "loss": 0.0853, "step": 13870 }, { "grad_norm": 0.5203650593757629, "learning_rate": 9.786007672958455e-05, "loss": 0.09, "step": 13880 }, { "grad_norm": 0.5611609816551208, "learning_rate": 9.78552886036441e-05, "loss": 0.0943, "step": 13890 }, { "grad_norm": 0.5994705557823181, "learning_rate": 9.785049524432124e-05, "loss": 0.0687, "step": 13900 }, { "grad_norm": 0.8433824181556702, "learning_rate": 9.784569665214016e-05, "loss": 0.09, "step": 13910 }, { "grad_norm": 0.5586483478546143, "learning_rate": 9.784089282762563e-05, "loss": 0.0902, "step": 13920 }, { "grad_norm": 0.6780149936676025, "learning_rate": 9.7836083771303e-05, "loss": 0.0926, "step": 13930 }, { "grad_norm": 0.7307749390602112, "learning_rate": 9.783126948369817e-05, "loss": 0.0967, "step": 13940 }, { "grad_norm": 0.5153647065162659, "learning_rate": 9.78264499653376e-05, "loss": 0.0742, "step": 13950 }, { "grad_norm": 0.5652680993080139, "learning_rate": 9.782162521674838e-05, "loss": 0.0804, "step": 13960 }, { "grad_norm": 0.5956963300704956, "learning_rate": 9.781679523845812e-05, "loss": 0.0965, "step": 13970 }, { "grad_norm": 0.5443331003189087, "learning_rate": 9.781196003099502e-05, "loss": 0.0923, "step": 13980 }, { "grad_norm": 0.39997774362564087, "learning_rate": 9.780711959488786e-05, "loss": 0.0813, "step": 13990 }, { "grad_norm": 0.4886448383331299, "learning_rate": 9.780227393066599e-05, "loss": 0.0762, "step": 14000 }, { "grad_norm": 0.7942387461662292, "learning_rate": 9.77974230388593e-05, "loss": 0.1081, "step": 14010 }, { "grad_norm": 0.5317270159721375, "learning_rate": 9.779256691999829e-05, "loss": 0.1022, "step": 14020 }, { "grad_norm": 0.6410101056098938, "learning_rate": 9.778770557461403e-05, "loss": 0.1019, "step": 14030 }, { "grad_norm": 0.6217474937438965, "learning_rate": 9.778283900323812e-05, "loss": 0.073, "step": 14040 }, { "grad_norm": 0.7459864020347595, "learning_rate": 9.777796720640277e-05, "loss": 0.0882, "step": 14050 }, { "grad_norm": 0.7795644402503967, "learning_rate": 9.777309018464078e-05, "loss": 0.0768, "step": 14060 }, { "grad_norm": 0.6839827299118042, "learning_rate": 9.776820793848547e-05, "loss": 0.074, "step": 14070 }, { "grad_norm": 0.6303603649139404, "learning_rate": 9.776332046847075e-05, "loss": 0.0757, "step": 14080 }, { "grad_norm": 0.7859801650047302, "learning_rate": 9.775842777513111e-05, "loss": 0.0805, "step": 14090 }, { "grad_norm": 0.7069024443626404, "learning_rate": 9.775352985900163e-05, "loss": 0.1092, "step": 14100 }, { "grad_norm": 0.8743299841880798, "learning_rate": 9.774862672061791e-05, "loss": 0.1016, "step": 14110 }, { "grad_norm": 0.7238851189613342, "learning_rate": 9.774371836051616e-05, "loss": 0.0932, "step": 14120 }, { "grad_norm": 1.009328842163086, "learning_rate": 9.773880477923315e-05, "loss": 0.0936, "step": 14130 }, { "grad_norm": 0.5710124373435974, "learning_rate": 9.773388597730623e-05, "loss": 0.0891, "step": 14140 }, { "grad_norm": 0.7585248947143555, "learning_rate": 9.77289619552733e-05, "loss": 0.0769, "step": 14150 }, { "grad_norm": 0.8325570225715637, "learning_rate": 9.772403271367285e-05, "loss": 0.1059, "step": 14160 }, { "grad_norm": 0.8518211245536804, "learning_rate": 9.771909825304396e-05, "loss": 0.087, "step": 14170 }, { "grad_norm": 0.6228224039077759, "learning_rate": 9.771415857392619e-05, "loss": 0.0945, "step": 14180 }, { "grad_norm": 0.7187450528144836, "learning_rate": 9.770921367685978e-05, "loss": 0.1039, "step": 14190 }, { "grad_norm": 0.6198092699050903, "learning_rate": 9.770426356238551e-05, "loss": 0.0828, "step": 14200 }, { "grad_norm": 0.4688188135623932, "learning_rate": 9.769930823104469e-05, "loss": 0.0819, "step": 14210 }, { "grad_norm": 0.5468499064445496, "learning_rate": 9.769434768337926e-05, "loss": 0.0879, "step": 14220 }, { "grad_norm": 0.4972139000892639, "learning_rate": 9.768938191993164e-05, "loss": 0.0887, "step": 14230 }, { "grad_norm": 0.5697346329689026, "learning_rate": 9.768441094124494e-05, "loss": 0.0766, "step": 14240 }, { "grad_norm": 0.6074680685997009, "learning_rate": 9.767943474786275e-05, "loss": 0.09, "step": 14250 }, { "grad_norm": 0.43050506711006165, "learning_rate": 9.767445334032923e-05, "loss": 0.0645, "step": 14260 }, { "grad_norm": 0.9771966338157654, "learning_rate": 9.766946671918919e-05, "loss": 0.0933, "step": 14270 }, { "grad_norm": 0.6499735116958618, "learning_rate": 9.766447488498796e-05, "loss": 0.0737, "step": 14280 }, { "grad_norm": 0.47517573833465576, "learning_rate": 9.765947783827139e-05, "loss": 0.0782, "step": 14290 }, { "grad_norm": 0.651532769203186, "learning_rate": 9.765447557958599e-05, "loss": 0.0822, "step": 14300 }, { "grad_norm": 0.6377159357070923, "learning_rate": 9.764946810947879e-05, "loss": 0.0815, "step": 14310 }, { "grad_norm": 0.5045881867408752, "learning_rate": 9.764445542849738e-05, "loss": 0.0922, "step": 14320 }, { "grad_norm": 0.4415596127510071, "learning_rate": 9.763943753718998e-05, "loss": 0.0754, "step": 14330 }, { "grad_norm": 0.4223919212818146, "learning_rate": 9.76344144361053e-05, "loss": 0.076, "step": 14340 }, { "grad_norm": 0.7961145639419556, "learning_rate": 9.762938612579269e-05, "loss": 0.086, "step": 14350 }, { "grad_norm": 0.8604345321655273, "learning_rate": 9.762435260680202e-05, "loss": 0.0833, "step": 14360 }, { "grad_norm": 0.5311658978462219, "learning_rate": 9.761931387968373e-05, "loss": 0.0937, "step": 14370 }, { "grad_norm": 0.5594146847724915, "learning_rate": 9.76142699449889e-05, "loss": 0.0787, "step": 14380 }, { "grad_norm": 0.6234654784202576, "learning_rate": 9.760922080326908e-05, "loss": 0.0894, "step": 14390 }, { "grad_norm": 0.8555811047554016, "learning_rate": 9.760416645507644e-05, "loss": 0.0824, "step": 14400 }, { "grad_norm": 0.8853874802589417, "learning_rate": 9.759910690096375e-05, "loss": 0.0838, "step": 14410 }, { "grad_norm": 0.6144360303878784, "learning_rate": 9.759404214148429e-05, "loss": 0.0858, "step": 14420 }, { "grad_norm": 0.675664484500885, "learning_rate": 9.758897217719191e-05, "loss": 0.0726, "step": 14430 }, { "grad_norm": 0.7183471322059631, "learning_rate": 9.758389700864113e-05, "loss": 0.0708, "step": 14440 }, { "grad_norm": 0.9048227667808533, "learning_rate": 9.757881663638688e-05, "loss": 0.0858, "step": 14450 }, { "grad_norm": 0.5448217988014221, "learning_rate": 9.757373106098478e-05, "loss": 0.0827, "step": 14460 }, { "grad_norm": 0.596074640750885, "learning_rate": 9.756864028299097e-05, "loss": 0.0854, "step": 14470 }, { "grad_norm": 0.6800597310066223, "learning_rate": 9.75635443029622e-05, "loss": 0.0765, "step": 14480 }, { "grad_norm": 0.6458073258399963, "learning_rate": 9.755844312145572e-05, "loss": 0.0674, "step": 14490 }, { "grad_norm": 0.7548971772193909, "learning_rate": 9.755333673902941e-05, "loss": 0.1008, "step": 14500 }, { "grad_norm": 0.6188841462135315, "learning_rate": 9.75482251562417e-05, "loss": 0.0795, "step": 14510 }, { "grad_norm": 0.6535384654998779, "learning_rate": 9.754310837365155e-05, "loss": 0.0708, "step": 14520 }, { "grad_norm": 0.5691501498222351, "learning_rate": 9.753798639181856e-05, "loss": 0.0998, "step": 14530 }, { "grad_norm": 0.6452471613883972, "learning_rate": 9.753285921130286e-05, "loss": 0.0676, "step": 14540 }, { "grad_norm": 0.5701157450675964, "learning_rate": 9.752772683266512e-05, "loss": 0.0777, "step": 14550 }, { "grad_norm": 0.7374083399772644, "learning_rate": 9.752258925646665e-05, "loss": 0.0902, "step": 14560 }, { "grad_norm": 0.5082571506500244, "learning_rate": 9.751744648326926e-05, "loss": 0.0799, "step": 14570 }, { "grad_norm": 0.6070808172225952, "learning_rate": 9.751229851363536e-05, "loss": 0.0945, "step": 14580 }, { "grad_norm": 0.6577274203300476, "learning_rate": 9.750714534812793e-05, "loss": 0.1066, "step": 14590 }, { "grad_norm": 0.601830780506134, "learning_rate": 9.750198698731053e-05, "loss": 0.091, "step": 14600 }, { "grad_norm": 0.559055745601654, "learning_rate": 9.749682343174722e-05, "loss": 0.0835, "step": 14610 }, { "grad_norm": 0.5224106907844543, "learning_rate": 9.749165468200272e-05, "loss": 0.0769, "step": 14620 }, { "grad_norm": 0.45451581478118896, "learning_rate": 9.748648073864229e-05, "loss": 0.0812, "step": 14630 }, { "grad_norm": 0.7076655030250549, "learning_rate": 9.748130160223168e-05, "loss": 0.0801, "step": 14640 }, { "grad_norm": 0.7727342247962952, "learning_rate": 9.747611727333734e-05, "loss": 0.0887, "step": 14650 }, { "grad_norm": 0.7221342921257019, "learning_rate": 9.74709277525262e-05, "loss": 0.0834, "step": 14660 }, { "grad_norm": 0.7728449106216431, "learning_rate": 9.746573304036576e-05, "loss": 0.0877, "step": 14670 }, { "grad_norm": 0.7152071595191956, "learning_rate": 9.746053313742412e-05, "loss": 0.074, "step": 14680 }, { "grad_norm": 1.0927481651306152, "learning_rate": 9.745532804426994e-05, "loss": 0.0758, "step": 14690 }, { "grad_norm": 0.7894500494003296, "learning_rate": 9.745011776147242e-05, "loss": 0.0774, "step": 14700 }, { "grad_norm": 0.8356512188911438, "learning_rate": 9.744490228960138e-05, "loss": 0.0991, "step": 14710 }, { "grad_norm": 0.6496438384056091, "learning_rate": 9.743968162922713e-05, "loss": 0.0845, "step": 14720 }, { "grad_norm": 0.5755869746208191, "learning_rate": 9.743445578092064e-05, "loss": 0.0951, "step": 14730 }, { "grad_norm": 0.6240957975387573, "learning_rate": 9.742922474525338e-05, "loss": 0.1059, "step": 14740 }, { "grad_norm": 0.556950569152832, "learning_rate": 9.742398852279741e-05, "loss": 0.0732, "step": 14750 }, { "grad_norm": 0.659198522567749, "learning_rate": 9.741874711412535e-05, "loss": 0.1017, "step": 14760 }, { "grad_norm": 0.8886151909828186, "learning_rate": 9.741350051981042e-05, "loss": 0.0903, "step": 14770 }, { "grad_norm": 0.6106914281845093, "learning_rate": 9.740824874042633e-05, "loss": 0.0832, "step": 14780 }, { "grad_norm": 0.5758154988288879, "learning_rate": 9.740299177654746e-05, "loss": 0.0716, "step": 14790 }, { "grad_norm": 0.6969728469848633, "learning_rate": 9.739772962874867e-05, "loss": 0.0838, "step": 14800 }, { "grad_norm": 0.39408963918685913, "learning_rate": 9.739246229760541e-05, "loss": 0.0685, "step": 14810 }, { "grad_norm": 0.7631372809410095, "learning_rate": 9.738718978369376e-05, "loss": 0.0861, "step": 14820 }, { "grad_norm": 0.5077770352363586, "learning_rate": 9.738191208759025e-05, "loss": 0.0862, "step": 14830 }, { "grad_norm": 0.6603094339370728, "learning_rate": 9.73766292098721e-05, "loss": 0.1049, "step": 14840 }, { "grad_norm": 0.6362033486366272, "learning_rate": 9.737134115111699e-05, "loss": 0.1023, "step": 14850 }, { "grad_norm": 0.790863573551178, "learning_rate": 9.736604791190323e-05, "loss": 0.0844, "step": 14860 }, { "grad_norm": 0.7453370094299316, "learning_rate": 9.73607494928097e-05, "loss": 0.0706, "step": 14870 }, { "grad_norm": 0.5359207987785339, "learning_rate": 9.735544589441581e-05, "loss": 0.0857, "step": 14880 }, { "grad_norm": 0.6557624936103821, "learning_rate": 9.735013711730154e-05, "loss": 0.0858, "step": 14890 }, { "grad_norm": 0.6235408782958984, "learning_rate": 9.734482316204747e-05, "loss": 0.078, "step": 14900 }, { "grad_norm": 1.0847934484481812, "learning_rate": 9.733950402923473e-05, "loss": 0.0813, "step": 14910 }, { "grad_norm": 0.3818904161453247, "learning_rate": 9.7334179719445e-05, "loss": 0.0712, "step": 14920 }, { "grad_norm": 0.7349662184715271, "learning_rate": 9.732885023326053e-05, "loss": 0.0903, "step": 14930 }, { "grad_norm": 0.5520097613334656, "learning_rate": 9.732351557126418e-05, "loss": 0.0984, "step": 14940 }, { "grad_norm": 0.6251036524772644, "learning_rate": 9.731817573403929e-05, "loss": 0.0861, "step": 14950 }, { "grad_norm": 0.45252305269241333, "learning_rate": 9.731283072216985e-05, "loss": 0.0828, "step": 14960 }, { "grad_norm": 1.8573088645935059, "learning_rate": 9.730748053624039e-05, "loss": 0.0693, "step": 14970 }, { "grad_norm": 0.4637081027030945, "learning_rate": 9.730212517683598e-05, "loss": 0.0874, "step": 14980 }, { "grad_norm": 0.6761572957038879, "learning_rate": 9.729676464454228e-05, "loss": 0.079, "step": 14990 }, { "grad_norm": 0.42891013622283936, "learning_rate": 9.72913989399455e-05, "loss": 0.0911, "step": 15000 }, { "grad_norm": 0.5989745259284973, "learning_rate": 9.728602806363242e-05, "loss": 0.0703, "step": 15010 }, { "grad_norm": 0.5737553238868713, "learning_rate": 9.728065201619043e-05, "loss": 0.0834, "step": 15020 }, { "grad_norm": 0.6309412121772766, "learning_rate": 9.727527079820742e-05, "loss": 0.0794, "step": 15030 }, { "grad_norm": 0.3746660649776459, "learning_rate": 9.726988441027186e-05, "loss": 0.0915, "step": 15040 }, { "grad_norm": 1.4057035446166992, "learning_rate": 9.726449285297281e-05, "loss": 0.0941, "step": 15050 }, { "grad_norm": 0.8540138006210327, "learning_rate": 9.72590961268999e-05, "loss": 0.102, "step": 15060 }, { "grad_norm": 0.4609411954879761, "learning_rate": 9.725369423264328e-05, "loss": 0.0796, "step": 15070 }, { "grad_norm": 0.82408207654953, "learning_rate": 9.72482871707937e-05, "loss": 0.0912, "step": 15080 }, { "grad_norm": 0.7436610460281372, "learning_rate": 9.724287494194247e-05, "loss": 0.0909, "step": 15090 }, { "grad_norm": 0.3795028626918793, "learning_rate": 9.723745754668147e-05, "loss": 0.0922, "step": 15100 }, { "grad_norm": 0.6379377841949463, "learning_rate": 9.723203498560313e-05, "loss": 0.091, "step": 15110 }, { "grad_norm": 0.5246796011924744, "learning_rate": 9.722660725930046e-05, "loss": 0.0681, "step": 15120 }, { "grad_norm": 0.6385610103607178, "learning_rate": 9.722117436836702e-05, "loss": 0.077, "step": 15130 }, { "grad_norm": 0.5009950995445251, "learning_rate": 9.721573631339696e-05, "loss": 0.0839, "step": 15140 }, { "grad_norm": 0.916995644569397, "learning_rate": 9.721029309498494e-05, "loss": 0.1073, "step": 15150 }, { "grad_norm": 0.629764199256897, "learning_rate": 9.720484471372627e-05, "loss": 0.0922, "step": 15160 }, { "grad_norm": 0.616761326789856, "learning_rate": 9.719939117021673e-05, "loss": 0.085, "step": 15170 }, { "grad_norm": 0.6694615483283997, "learning_rate": 9.719393246505275e-05, "loss": 0.1017, "step": 15180 }, { "grad_norm": 0.5499993562698364, "learning_rate": 9.718846859883128e-05, "loss": 0.0802, "step": 15190 }, { "grad_norm": 1.064005970954895, "learning_rate": 9.718299957214982e-05, "loss": 0.0928, "step": 15200 }, { "grad_norm": 0.49354106187820435, "learning_rate": 9.717752538560646e-05, "loss": 0.0835, "step": 15210 }, { "grad_norm": 0.6278995275497437, "learning_rate": 9.717204603979986e-05, "loss": 0.0822, "step": 15220 }, { "grad_norm": 0.5516607761383057, "learning_rate": 9.716656153532922e-05, "loss": 0.0777, "step": 15230 }, { "grad_norm": 0.8947928547859192, "learning_rate": 9.716107187279434e-05, "loss": 0.0907, "step": 15240 }, { "grad_norm": 0.8358508348464966, "learning_rate": 9.715557705279555e-05, "loss": 0.1102, "step": 15250 }, { "grad_norm": 0.6883557438850403, "learning_rate": 9.715007707593372e-05, "loss": 0.1018, "step": 15260 }, { "grad_norm": 0.47640419006347656, "learning_rate": 9.714457194281036e-05, "loss": 0.079, "step": 15270 }, { "grad_norm": 0.6323399543762207, "learning_rate": 9.713906165402751e-05, "loss": 0.095, "step": 15280 }, { "grad_norm": 0.7476680874824524, "learning_rate": 9.713354621018774e-05, "loss": 0.1078, "step": 15290 }, { "grad_norm": 0.5924745798110962, "learning_rate": 9.712802561189422e-05, "loss": 0.0859, "step": 15300 }, { "grad_norm": 0.5588060617446899, "learning_rate": 9.712249985975069e-05, "loss": 0.103, "step": 15310 }, { "grad_norm": 0.48083755373954773, "learning_rate": 9.71169689543614e-05, "loss": 0.0727, "step": 15320 }, { "grad_norm": 0.6091481447219849, "learning_rate": 9.711143289633123e-05, "loss": 0.0728, "step": 15330 }, { "grad_norm": 0.5742486119270325, "learning_rate": 9.710589168626561e-05, "loss": 0.0748, "step": 15340 }, { "grad_norm": 0.5661290287971497, "learning_rate": 9.710034532477048e-05, "loss": 0.0837, "step": 15350 }, { "grad_norm": 0.661831259727478, "learning_rate": 9.709479381245239e-05, "loss": 0.0732, "step": 15360 }, { "grad_norm": 0.5077865123748779, "learning_rate": 9.708923714991847e-05, "loss": 0.0826, "step": 15370 }, { "grad_norm": 0.5321522355079651, "learning_rate": 9.708367533777638e-05, "loss": 0.0786, "step": 15380 }, { "grad_norm": 0.8090230822563171, "learning_rate": 9.707810837663431e-05, "loss": 0.0717, "step": 15390 }, { "grad_norm": 0.8442652225494385, "learning_rate": 9.707253626710113e-05, "loss": 0.0953, "step": 15400 }, { "grad_norm": 0.592659056186676, "learning_rate": 9.706695900978613e-05, "loss": 0.0694, "step": 15410 }, { "grad_norm": 0.672187864780426, "learning_rate": 9.706137660529926e-05, "loss": 0.0824, "step": 15420 }, { "grad_norm": 0.6265163421630859, "learning_rate": 9.705578905425101e-05, "loss": 0.1055, "step": 15430 }, { "grad_norm": 0.621193528175354, "learning_rate": 9.705019635725241e-05, "loss": 0.08, "step": 15440 }, { "grad_norm": 0.5447222590446472, "learning_rate": 9.704459851491508e-05, "loss": 0.0899, "step": 15450 }, { "grad_norm": 0.5113765597343445, "learning_rate": 9.703899552785118e-05, "loss": 0.072, "step": 15460 }, { "grad_norm": 0.5569828748703003, "learning_rate": 9.703338739667346e-05, "loss": 0.0746, "step": 15470 }, { "grad_norm": 0.521700382232666, "learning_rate": 9.70277741219952e-05, "loss": 0.0776, "step": 15480 }, { "grad_norm": 0.6812137365341187, "learning_rate": 9.702215570443027e-05, "loss": 0.0916, "step": 15490 }, { "grad_norm": 0.6802948117256165, "learning_rate": 9.701653214459309e-05, "loss": 0.0868, "step": 15500 }, { "grad_norm": 0.5658239126205444, "learning_rate": 9.701090344309865e-05, "loss": 0.086, "step": 15510 }, { "grad_norm": 0.8235664367675781, "learning_rate": 9.700526960056247e-05, "loss": 0.0739, "step": 15520 }, { "grad_norm": 0.6810824871063232, "learning_rate": 9.699963061760068e-05, "loss": 0.0652, "step": 15530 }, { "grad_norm": 0.6784184575080872, "learning_rate": 9.699398649482997e-05, "loss": 0.0807, "step": 15540 }, { "grad_norm": 0.7191452383995056, "learning_rate": 9.698833723286753e-05, "loss": 0.0655, "step": 15550 }, { "grad_norm": 0.47139284014701843, "learning_rate": 9.698268283233118e-05, "loss": 0.0889, "step": 15560 }, { "grad_norm": 0.8335251212120056, "learning_rate": 9.697702329383929e-05, "loss": 0.0872, "step": 15570 }, { "grad_norm": 0.6465653777122498, "learning_rate": 9.697135861801074e-05, "loss": 0.0916, "step": 15580 }, { "grad_norm": 0.4951867163181305, "learning_rate": 9.696568880546505e-05, "loss": 0.0807, "step": 15590 }, { "grad_norm": 0.5784424543380737, "learning_rate": 9.696001385682223e-05, "loss": 0.0803, "step": 15600 }, { "grad_norm": 0.5659007430076599, "learning_rate": 9.695433377270291e-05, "loss": 0.08, "step": 15610 }, { "grad_norm": 0.554498016834259, "learning_rate": 9.694864855372824e-05, "loss": 0.0734, "step": 15620 }, { "grad_norm": 0.7110916972160339, "learning_rate": 9.694295820051995e-05, "loss": 0.0754, "step": 15630 }, { "grad_norm": 0.7589677572250366, "learning_rate": 9.693726271370032e-05, "loss": 0.0723, "step": 15640 }, { "grad_norm": 0.5558983087539673, "learning_rate": 9.693156209389221e-05, "loss": 0.0775, "step": 15650 }, { "grad_norm": 0.5827017426490784, "learning_rate": 9.692585634171905e-05, "loss": 0.0746, "step": 15660 }, { "grad_norm": 0.6623978614807129, "learning_rate": 9.692014545780476e-05, "loss": 0.0809, "step": 15670 }, { "grad_norm": 0.5300402045249939, "learning_rate": 9.691442944277393e-05, "loss": 0.0718, "step": 15680 }, { "grad_norm": 0.5805031657218933, "learning_rate": 9.690870829725162e-05, "loss": 0.0805, "step": 15690 }, { "grad_norm": 0.6125296354293823, "learning_rate": 9.69029820218635e-05, "loss": 0.0861, "step": 15700 }, { "grad_norm": 0.4310109615325928, "learning_rate": 9.689725061723579e-05, "loss": 0.0887, "step": 15710 }, { "grad_norm": 0.6326513886451721, "learning_rate": 9.689151408399527e-05, "loss": 0.0859, "step": 15720 }, { "grad_norm": 0.5076085329055786, "learning_rate": 9.688577242276924e-05, "loss": 0.085, "step": 15730 }, { "grad_norm": 0.715472936630249, "learning_rate": 9.688002563418566e-05, "loss": 0.0936, "step": 15740 }, { "grad_norm": 0.6143381595611572, "learning_rate": 9.687427371887293e-05, "loss": 0.078, "step": 15750 }, { "grad_norm": 0.37656521797180176, "learning_rate": 9.686851667746012e-05, "loss": 0.0774, "step": 15760 }, { "grad_norm": 1.119653344154358, "learning_rate": 9.686275451057677e-05, "loss": 0.0736, "step": 15770 }, { "grad_norm": 0.7096685767173767, "learning_rate": 9.685698721885308e-05, "loss": 0.0771, "step": 15780 }, { "grad_norm": 0.577246904373169, "learning_rate": 9.68512148029197e-05, "loss": 0.0862, "step": 15790 }, { "grad_norm": 0.6338789463043213, "learning_rate": 9.684543726340791e-05, "loss": 0.0839, "step": 15800 }, { "grad_norm": 0.5719964504241943, "learning_rate": 9.683965460094952e-05, "loss": 0.0706, "step": 15810 }, { "grad_norm": 0.7383474707603455, "learning_rate": 9.683386681617694e-05, "loss": 0.0776, "step": 15820 }, { "grad_norm": 0.6527976393699646, "learning_rate": 9.68280739097231e-05, "loss": 0.0778, "step": 15830 }, { "grad_norm": 0.6051522493362427, "learning_rate": 9.682227588222148e-05, "loss": 0.0667, "step": 15840 }, { "grad_norm": 1.0793886184692383, "learning_rate": 9.681647273430618e-05, "loss": 0.0763, "step": 15850 }, { "grad_norm": 1.012118935585022, "learning_rate": 9.681066446661182e-05, "loss": 0.0804, "step": 15860 }, { "grad_norm": 0.70566725730896, "learning_rate": 9.680485107977357e-05, "loss": 0.0896, "step": 15870 }, { "grad_norm": 0.7151974439620972, "learning_rate": 9.679903257442716e-05, "loss": 0.0901, "step": 15880 }, { "grad_norm": 0.8010369539260864, "learning_rate": 9.679320895120891e-05, "loss": 0.0819, "step": 15890 }, { "grad_norm": 0.5439239740371704, "learning_rate": 9.67873802107557e-05, "loss": 0.0692, "step": 15900 }, { "grad_norm": 0.48423564434051514, "learning_rate": 9.67815463537049e-05, "loss": 0.0699, "step": 15910 }, { "grad_norm": 0.7387201189994812, "learning_rate": 9.677570738069457e-05, "loss": 0.0722, "step": 15920 }, { "grad_norm": 0.45531728863716125, "learning_rate": 9.676986329236318e-05, "loss": 0.0835, "step": 15930 }, { "grad_norm": 0.5451585650444031, "learning_rate": 9.676401408934987e-05, "loss": 0.0733, "step": 15940 }, { "grad_norm": 0.62251877784729, "learning_rate": 9.675815977229428e-05, "loss": 0.0764, "step": 15950 }, { "grad_norm": 0.5635558366775513, "learning_rate": 9.675230034183664e-05, "loss": 0.0752, "step": 15960 }, { "grad_norm": 0.886754035949707, "learning_rate": 9.674643579861773e-05, "loss": 0.0931, "step": 15970 }, { "grad_norm": 0.6498538851737976, "learning_rate": 9.674056614327886e-05, "loss": 0.0667, "step": 15980 }, { "grad_norm": 0.5647085309028625, "learning_rate": 9.673469137646198e-05, "loss": 0.0877, "step": 15990 }, { "grad_norm": 0.5376133918762207, "learning_rate": 9.67288114988095e-05, "loss": 0.0727, "step": 16000 }, { "grad_norm": 0.5811020135879517, "learning_rate": 9.672292651096447e-05, "loss": 0.0713, "step": 16010 }, { "grad_norm": 0.5484881401062012, "learning_rate": 9.671703641357042e-05, "loss": 0.0706, "step": 16020 }, { "grad_norm": 0.6107531785964966, "learning_rate": 9.67111412072715e-05, "loss": 0.0839, "step": 16030 }, { "grad_norm": 0.5986529588699341, "learning_rate": 9.670524089271242e-05, "loss": 0.0679, "step": 16040 }, { "grad_norm": 0.5312627553939819, "learning_rate": 9.669933547053842e-05, "loss": 0.0747, "step": 16050 }, { "grad_norm": 0.538051187992096, "learning_rate": 9.669342494139531e-05, "loss": 0.0802, "step": 16060 }, { "grad_norm": 0.6611968278884888, "learning_rate": 9.668750930592943e-05, "loss": 0.0702, "step": 16070 }, { "grad_norm": 0.47476938366889954, "learning_rate": 9.668158856478775e-05, "loss": 0.0698, "step": 16080 }, { "grad_norm": 0.8343870639801025, "learning_rate": 9.66756627186177e-05, "loss": 0.0828, "step": 16090 }, { "grad_norm": 0.5703035593032837, "learning_rate": 9.666973176806737e-05, "loss": 0.0773, "step": 16100 }, { "grad_norm": 0.567475438117981, "learning_rate": 9.666379571378534e-05, "loss": 0.0753, "step": 16110 }, { "grad_norm": 0.6599922180175781, "learning_rate": 9.665785455642076e-05, "loss": 0.0894, "step": 16120 }, { "grad_norm": 0.6716173887252808, "learning_rate": 9.665190829662337e-05, "loss": 0.0848, "step": 16130 }, { "grad_norm": 0.524003803730011, "learning_rate": 9.664595693504342e-05, "loss": 0.0683, "step": 16140 }, { "grad_norm": 0.5482262969017029, "learning_rate": 9.664000047233175e-05, "loss": 0.1162, "step": 16150 }, { "grad_norm": 0.5414419174194336, "learning_rate": 9.663403890913976e-05, "loss": 0.089, "step": 16160 }, { "grad_norm": 0.6948265433311462, "learning_rate": 9.662807224611938e-05, "loss": 0.0842, "step": 16170 }, { "grad_norm": 0.5458356142044067, "learning_rate": 9.662210048392311e-05, "loss": 0.0867, "step": 16180 }, { "grad_norm": 0.6933521628379822, "learning_rate": 9.661612362320405e-05, "loss": 0.0869, "step": 16190 }, { "grad_norm": 0.6356691122055054, "learning_rate": 9.661014166461579e-05, "loss": 0.0846, "step": 16200 }, { "grad_norm": 0.5886597633361816, "learning_rate": 9.66041546088125e-05, "loss": 0.092, "step": 16210 }, { "grad_norm": 0.7972354292869568, "learning_rate": 9.659816245644895e-05, "loss": 0.0785, "step": 16220 }, { "grad_norm": 1.0679594278335571, "learning_rate": 9.65921652081804e-05, "loss": 0.0839, "step": 16230 }, { "grad_norm": 0.511142373085022, "learning_rate": 9.658616286466271e-05, "loss": 0.0868, "step": 16240 }, { "grad_norm": 0.9673253893852234, "learning_rate": 9.65801554265523e-05, "loss": 0.0884, "step": 16250 }, { "grad_norm": 0.749812126159668, "learning_rate": 9.657414289450612e-05, "loss": 0.0806, "step": 16260 }, { "grad_norm": 0.6513935327529907, "learning_rate": 9.656812526918171e-05, "loss": 0.0901, "step": 16270 }, { "grad_norm": 0.5953586101531982, "learning_rate": 9.656210255123712e-05, "loss": 0.078, "step": 16280 }, { "grad_norm": 0.6863927841186523, "learning_rate": 9.6556074741331e-05, "loss": 0.0748, "step": 16290 }, { "grad_norm": 0.37114858627319336, "learning_rate": 9.655004184012256e-05, "loss": 0.0656, "step": 16300 }, { "grad_norm": 0.8889654278755188, "learning_rate": 9.654400384827152e-05, "loss": 0.0809, "step": 16310 }, { "grad_norm": 0.6578901410102844, "learning_rate": 9.653796076643818e-05, "loss": 0.0664, "step": 16320 }, { "grad_norm": 0.9391756057739258, "learning_rate": 9.653191259528344e-05, "loss": 0.0838, "step": 16330 }, { "grad_norm": 0.49389079213142395, "learning_rate": 9.65258593354687e-05, "loss": 0.0748, "step": 16340 }, { "grad_norm": 0.4361910820007324, "learning_rate": 9.651980098765591e-05, "loss": 0.0727, "step": 16350 }, { "grad_norm": 0.8534737825393677, "learning_rate": 9.651373755250765e-05, "loss": 0.0899, "step": 16360 }, { "grad_norm": 0.5035486221313477, "learning_rate": 9.650766903068697e-05, "loss": 0.0838, "step": 16370 }, { "grad_norm": 0.6657981872558594, "learning_rate": 9.650159542285753e-05, "loss": 0.0727, "step": 16380 }, { "grad_norm": 0.6771858930587769, "learning_rate": 9.649551672968353e-05, "loss": 0.0955, "step": 16390 }, { "grad_norm": 0.5992896556854248, "learning_rate": 9.648943295182973e-05, "loss": 0.077, "step": 16400 }, { "grad_norm": 0.5822808742523193, "learning_rate": 9.648334408996144e-05, "loss": 0.0727, "step": 16410 }, { "grad_norm": 0.5201818943023682, "learning_rate": 9.647725014474452e-05, "loss": 0.0772, "step": 16420 }, { "grad_norm": 0.49909520149230957, "learning_rate": 9.64711511168454e-05, "loss": 0.0843, "step": 16430 }, { "grad_norm": 0.39003464579582214, "learning_rate": 9.646504700693108e-05, "loss": 0.079, "step": 16440 }, { "grad_norm": 0.6554824709892273, "learning_rate": 9.645893781566907e-05, "loss": 0.0816, "step": 16450 }, { "grad_norm": 0.6294993162155151, "learning_rate": 9.645282354372744e-05, "loss": 0.0978, "step": 16460 }, { "grad_norm": 0.5790468454360962, "learning_rate": 9.644670419177491e-05, "loss": 0.0776, "step": 16470 }, { "grad_norm": 0.6229963898658752, "learning_rate": 9.644057976048062e-05, "loss": 0.0778, "step": 16480 }, { "grad_norm": 0.5797048211097717, "learning_rate": 9.643445025051435e-05, "loss": 0.0784, "step": 16490 }, { "grad_norm": 0.645988941192627, "learning_rate": 9.642831566254641e-05, "loss": 0.0692, "step": 16500 }, { "grad_norm": 0.5002651214599609, "learning_rate": 9.642217599724769e-05, "loss": 0.0688, "step": 16510 }, { "grad_norm": 0.4889354407787323, "learning_rate": 9.64160312552896e-05, "loss": 0.0723, "step": 16520 }, { "grad_norm": 0.6265323162078857, "learning_rate": 9.64098814373441e-05, "loss": 0.0669, "step": 16530 }, { "grad_norm": 0.6240803599357605, "learning_rate": 9.640372654408374e-05, "loss": 0.0667, "step": 16540 }, { "grad_norm": 0.41882237792015076, "learning_rate": 9.639756657618162e-05, "loss": 0.0801, "step": 16550 }, { "grad_norm": 0.639167308807373, "learning_rate": 9.639140153431138e-05, "loss": 0.0803, "step": 16560 }, { "grad_norm": 0.720615565776825, "learning_rate": 9.638523141914721e-05, "loss": 0.0875, "step": 16570 }, { "grad_norm": 0.533933162689209, "learning_rate": 9.637905623136388e-05, "loss": 0.0796, "step": 16580 }, { "grad_norm": 0.6343308091163635, "learning_rate": 9.637287597163669e-05, "loss": 0.0668, "step": 16590 }, { "grad_norm": 0.5884215235710144, "learning_rate": 9.63666906406415e-05, "loss": 0.0782, "step": 16600 }, { "grad_norm": 0.36582860350608826, "learning_rate": 9.636050023905473e-05, "loss": 0.0648, "step": 16610 }, { "grad_norm": 0.4887186884880066, "learning_rate": 9.635430476755336e-05, "loss": 0.066, "step": 16620 }, { "grad_norm": 0.7903050184249878, "learning_rate": 9.63481042268149e-05, "loss": 0.0829, "step": 16630 }, { "grad_norm": 0.6201320290565491, "learning_rate": 9.634189861751745e-05, "loss": 0.0845, "step": 16640 }, { "grad_norm": 0.482357382774353, "learning_rate": 9.633568794033967e-05, "loss": 0.0827, "step": 16650 }, { "grad_norm": 0.4920583963394165, "learning_rate": 9.63294721959607e-05, "loss": 0.0672, "step": 16660 }, { "grad_norm": 0.44920864701271057, "learning_rate": 9.63232513850603e-05, "loss": 0.0933, "step": 16670 }, { "grad_norm": 0.5179983377456665, "learning_rate": 9.631702550831878e-05, "loss": 0.0883, "step": 16680 }, { "grad_norm": 0.595863401889801, "learning_rate": 9.631079456641698e-05, "loss": 0.0811, "step": 16690 }, { "grad_norm": 0.7407206892967224, "learning_rate": 9.630455856003632e-05, "loss": 0.0971, "step": 16700 }, { "grad_norm": 0.6239014267921448, "learning_rate": 9.629831748985876e-05, "loss": 0.098, "step": 16710 }, { "grad_norm": 0.6806439161300659, "learning_rate": 9.629207135656679e-05, "loss": 0.0913, "step": 16720 }, { "grad_norm": 0.6328460574150085, "learning_rate": 9.628582016084353e-05, "loss": 0.0771, "step": 16730 }, { "grad_norm": 0.6754186153411865, "learning_rate": 9.627956390337254e-05, "loss": 0.0679, "step": 16740 }, { "grad_norm": 0.5350461602210999, "learning_rate": 9.627330258483802e-05, "loss": 0.0831, "step": 16750 }, { "grad_norm": 0.5665816068649292, "learning_rate": 9.62670362059247e-05, "loss": 0.069, "step": 16760 }, { "grad_norm": 0.5787496566772461, "learning_rate": 9.626076476731786e-05, "loss": 0.0827, "step": 16770 }, { "grad_norm": 0.6241004467010498, "learning_rate": 9.625448826970336e-05, "loss": 0.09, "step": 16780 }, { "grad_norm": 1.0873286724090576, "learning_rate": 9.624820671376755e-05, "loss": 0.0894, "step": 16790 }, { "grad_norm": 0.767466127872467, "learning_rate": 9.62419201001974e-05, "loss": 0.0901, "step": 16800 }, { "grad_norm": 0.6907799243927002, "learning_rate": 9.623562842968037e-05, "loss": 0.0731, "step": 16810 }, { "grad_norm": 0.6681950688362122, "learning_rate": 9.622933170290454e-05, "loss": 0.0806, "step": 16820 }, { "grad_norm": 0.49702972173690796, "learning_rate": 9.622302992055849e-05, "loss": 0.0713, "step": 16830 }, { "grad_norm": 0.40893247723579407, "learning_rate": 9.62167230833314e-05, "loss": 0.0657, "step": 16840 }, { "grad_norm": 0.6564500331878662, "learning_rate": 9.621041119191295e-05, "loss": 0.0686, "step": 16850 }, { "grad_norm": 1.035891056060791, "learning_rate": 9.620409424699342e-05, "loss": 0.0918, "step": 16860 }, { "grad_norm": 0.5602295398712158, "learning_rate": 9.619777224926359e-05, "loss": 0.0707, "step": 16870 }, { "grad_norm": 0.6468464732170105, "learning_rate": 9.619144519941485e-05, "loss": 0.0836, "step": 16880 }, { "grad_norm": 0.7193231582641602, "learning_rate": 9.618511309813912e-05, "loss": 0.0923, "step": 16890 }, { "grad_norm": 0.8387681841850281, "learning_rate": 9.617877594612886e-05, "loss": 0.0754, "step": 16900 }, { "grad_norm": 0.6215252876281738, "learning_rate": 9.617243374407707e-05, "loss": 0.0752, "step": 16910 }, { "grad_norm": 0.5381494164466858, "learning_rate": 9.616608649267736e-05, "loss": 0.0783, "step": 16920 }, { "grad_norm": 0.5669312477111816, "learning_rate": 9.615973419262385e-05, "loss": 0.0698, "step": 16930 }, { "grad_norm": 0.6631993651390076, "learning_rate": 9.615337684461119e-05, "loss": 0.0618, "step": 16940 }, { "grad_norm": 0.6834877729415894, "learning_rate": 9.614701444933465e-05, "loss": 0.079, "step": 16950 }, { "grad_norm": 0.7820590138435364, "learning_rate": 9.614064700748997e-05, "loss": 0.0837, "step": 16960 }, { "grad_norm": 0.595695972442627, "learning_rate": 9.613427451977352e-05, "loss": 0.0742, "step": 16970 }, { "grad_norm": 0.5367010235786438, "learning_rate": 9.612789698688216e-05, "loss": 0.0865, "step": 16980 }, { "grad_norm": 0.47453370690345764, "learning_rate": 9.612151440951334e-05, "loss": 0.0799, "step": 16990 }, { "grad_norm": 0.8479605317115784, "learning_rate": 9.611512678836506e-05, "loss": 0.0827, "step": 17000 }, { "grad_norm": 0.5238496661186218, "learning_rate": 9.610873412413584e-05, "loss": 0.0803, "step": 17010 }, { "grad_norm": 1.2420005798339844, "learning_rate": 9.610233641752476e-05, "loss": 0.0924, "step": 17020 }, { "grad_norm": 0.9216016530990601, "learning_rate": 9.609593366923151e-05, "loss": 0.079, "step": 17030 }, { "grad_norm": 0.624477744102478, "learning_rate": 9.608952587995625e-05, "loss": 0.0688, "step": 17040 }, { "grad_norm": 0.5512709617614746, "learning_rate": 9.608311305039972e-05, "loss": 0.0866, "step": 17050 }, { "grad_norm": 0.7026422619819641, "learning_rate": 9.607669518126326e-05, "loss": 0.0906, "step": 17060 }, { "grad_norm": 0.7369059324264526, "learning_rate": 9.607027227324866e-05, "loss": 0.0773, "step": 17070 }, { "grad_norm": 0.547974169254303, "learning_rate": 9.606384432705837e-05, "loss": 0.0754, "step": 17080 }, { "grad_norm": 0.7927464246749878, "learning_rate": 9.60574113433953e-05, "loss": 0.0922, "step": 17090 }, { "grad_norm": 0.46185538172721863, "learning_rate": 9.6050973322963e-05, "loss": 0.0777, "step": 17100 }, { "grad_norm": 0.7665438652038574, "learning_rate": 9.604453026646547e-05, "loss": 0.0762, "step": 17110 }, { "grad_norm": 0.5226749181747437, "learning_rate": 9.603808217460735e-05, "loss": 0.0727, "step": 17120 }, { "grad_norm": 0.5757685303688049, "learning_rate": 9.603162904809377e-05, "loss": 0.0685, "step": 17130 }, { "grad_norm": 0.754179835319519, "learning_rate": 9.602517088763045e-05, "loss": 0.0627, "step": 17140 }, { "grad_norm": 0.6425066590309143, "learning_rate": 9.601870769392365e-05, "loss": 0.0974, "step": 17150 }, { "grad_norm": 0.6470321416854858, "learning_rate": 9.601223946768017e-05, "loss": 0.0827, "step": 17160 }, { "grad_norm": 0.4915401041507721, "learning_rate": 9.600576620960734e-05, "loss": 0.0705, "step": 17170 }, { "grad_norm": 0.6065288782119751, "learning_rate": 9.599928792041308e-05, "loss": 0.0737, "step": 17180 }, { "grad_norm": 0.8109245896339417, "learning_rate": 9.599280460080587e-05, "loss": 0.0808, "step": 17190 }, { "grad_norm": 0.8171321153640747, "learning_rate": 9.59863162514947e-05, "loss": 0.0928, "step": 17200 }, { "grad_norm": 1.0549421310424805, "learning_rate": 9.597982287318911e-05, "loss": 0.0817, "step": 17210 }, { "grad_norm": 0.46921253204345703, "learning_rate": 9.597332446659923e-05, "loss": 0.0788, "step": 17220 }, { "grad_norm": 0.5683285593986511, "learning_rate": 9.59668210324357e-05, "loss": 0.0779, "step": 17230 }, { "grad_norm": 0.5306316614151001, "learning_rate": 9.596031257140974e-05, "loss": 0.063, "step": 17240 }, { "grad_norm": 0.4812130630016327, "learning_rate": 9.59537990842331e-05, "loss": 0.0883, "step": 17250 }, { "grad_norm": 0.47299614548683167, "learning_rate": 9.594728057161806e-05, "loss": 0.0586, "step": 17260 }, { "grad_norm": 0.6317301988601685, "learning_rate": 9.594075703427752e-05, "loss": 0.0653, "step": 17270 }, { "grad_norm": 0.38166818022727966, "learning_rate": 9.593422847292486e-05, "loss": 0.0638, "step": 17280 }, { "grad_norm": 0.5080211162567139, "learning_rate": 9.592769488827402e-05, "loss": 0.076, "step": 17290 }, { "grad_norm": 0.5249539017677307, "learning_rate": 9.592115628103952e-05, "loss": 0.0727, "step": 17300 }, { "grad_norm": 0.6103530526161194, "learning_rate": 9.591461265193643e-05, "loss": 0.0705, "step": 17310 }, { "grad_norm": 0.953325092792511, "learning_rate": 9.590806400168032e-05, "loss": 0.0658, "step": 17320 }, { "grad_norm": 0.4580405652523041, "learning_rate": 9.590151033098735e-05, "loss": 0.0635, "step": 17330 }, { "grad_norm": 0.5938311219215393, "learning_rate": 9.589495164057423e-05, "loss": 0.0733, "step": 17340 }, { "grad_norm": 0.5744596123695374, "learning_rate": 9.58883879311582e-05, "loss": 0.0673, "step": 17350 }, { "grad_norm": 0.5080183744430542, "learning_rate": 9.588181920345705e-05, "loss": 0.0745, "step": 17360 }, { "grad_norm": 0.6063559055328369, "learning_rate": 9.587524545818913e-05, "loss": 0.0797, "step": 17370 }, { "grad_norm": 0.5149218440055847, "learning_rate": 9.586866669607335e-05, "loss": 0.0683, "step": 17380 }, { "grad_norm": 0.6917893886566162, "learning_rate": 9.586208291782915e-05, "loss": 0.0805, "step": 17390 }, { "grad_norm": 0.5124993920326233, "learning_rate": 9.58554941241765e-05, "loss": 0.0748, "step": 17400 }, { "grad_norm": 0.5365190505981445, "learning_rate": 9.584890031583596e-05, "loss": 0.0645, "step": 17410 }, { "grad_norm": 0.49950674176216125, "learning_rate": 9.584230149352861e-05, "loss": 0.0703, "step": 17420 }, { "grad_norm": 0.8285687565803528, "learning_rate": 9.58356976579761e-05, "loss": 0.0715, "step": 17430 }, { "grad_norm": 0.5668394565582275, "learning_rate": 9.58290888099006e-05, "loss": 0.0744, "step": 17440 }, { "grad_norm": 0.5252701044082642, "learning_rate": 9.582247495002486e-05, "loss": 0.0778, "step": 17450 }, { "grad_norm": 0.6551718711853027, "learning_rate": 9.581585607907214e-05, "loss": 0.0811, "step": 17460 }, { "grad_norm": 2.8283963203430176, "learning_rate": 9.580923219776628e-05, "loss": 0.0782, "step": 17470 }, { "grad_norm": 0.5666705965995789, "learning_rate": 9.580260330683167e-05, "loss": 0.0663, "step": 17480 }, { "grad_norm": 0.6258137226104736, "learning_rate": 9.579596940699322e-05, "loss": 0.066, "step": 17490 }, { "grad_norm": 0.4845805764198303, "learning_rate": 9.578933049897643e-05, "loss": 0.063, "step": 17500 }, { "grad_norm": 0.5379893183708191, "learning_rate": 9.578268658350728e-05, "loss": 0.0741, "step": 17510 }, { "grad_norm": 0.552493155002594, "learning_rate": 9.577603766131235e-05, "loss": 0.0754, "step": 17520 }, { "grad_norm": 0.7119115591049194, "learning_rate": 9.576938373311878e-05, "loss": 0.0755, "step": 17530 }, { "grad_norm": 0.74110347032547, "learning_rate": 9.576272479965421e-05, "loss": 0.0781, "step": 17540 }, { "grad_norm": 0.6446050405502319, "learning_rate": 9.575606086164687e-05, "loss": 0.0749, "step": 17550 }, { "grad_norm": 0.8826566338539124, "learning_rate": 9.57493919198255e-05, "loss": 0.0907, "step": 17560 }, { "grad_norm": 0.6520911455154419, "learning_rate": 9.57427179749194e-05, "loss": 0.0966, "step": 17570 }, { "grad_norm": 0.5836178064346313, "learning_rate": 9.573603902765846e-05, "loss": 0.0794, "step": 17580 }, { "grad_norm": 0.538712739944458, "learning_rate": 9.572935507877304e-05, "loss": 0.0785, "step": 17590 }, { "grad_norm": 0.784792959690094, "learning_rate": 9.57226661289941e-05, "loss": 0.0892, "step": 17600 }, { "grad_norm": 0.6974707245826721, "learning_rate": 9.571597217905315e-05, "loss": 0.0819, "step": 17610 }, { "grad_norm": 1.081182599067688, "learning_rate": 9.57092732296822e-05, "loss": 0.0748, "step": 17620 }, { "grad_norm": 0.5792847871780396, "learning_rate": 9.570256928161385e-05, "loss": 0.0612, "step": 17630 }, { "grad_norm": 0.48397818207740784, "learning_rate": 9.569586033558126e-05, "loss": 0.0727, "step": 17640 }, { "grad_norm": 0.41472312808036804, "learning_rate": 9.568914639231807e-05, "loss": 0.0599, "step": 17650 }, { "grad_norm": 0.6768965721130371, "learning_rate": 9.568242745255852e-05, "loss": 0.0746, "step": 17660 }, { "grad_norm": 0.5412996411323547, "learning_rate": 9.567570351703739e-05, "loss": 0.08, "step": 17670 }, { "grad_norm": 0.7789647579193115, "learning_rate": 9.566897458649001e-05, "loss": 0.0846, "step": 17680 }, { "grad_norm": 0.5308372974395752, "learning_rate": 9.566224066165221e-05, "loss": 0.0911, "step": 17690 }, { "grad_norm": 0.6000989079475403, "learning_rate": 9.565550174326043e-05, "loss": 0.0886, "step": 17700 }, { "grad_norm": 0.520322322845459, "learning_rate": 9.564875783205162e-05, "loss": 0.0809, "step": 17710 }, { "grad_norm": 0.7185884714126587, "learning_rate": 9.564200892876328e-05, "loss": 0.081, "step": 17720 }, { "grad_norm": 0.8611742854118347, "learning_rate": 9.563525503413348e-05, "loss": 0.0755, "step": 17730 }, { "grad_norm": 0.6476524472236633, "learning_rate": 9.562849614890079e-05, "loss": 0.0851, "step": 17740 }, { "grad_norm": 0.5496971607208252, "learning_rate": 9.562173227380436e-05, "loss": 0.0896, "step": 17750 }, { "grad_norm": 0.45243680477142334, "learning_rate": 9.561496340958389e-05, "loss": 0.0747, "step": 17760 }, { "grad_norm": 0.634239912033081, "learning_rate": 9.560818955697959e-05, "loss": 0.0802, "step": 17770 }, { "grad_norm": 0.6241196990013123, "learning_rate": 9.560141071673228e-05, "loss": 0.0775, "step": 17780 }, { "grad_norm": 0.5144316554069519, "learning_rate": 9.559462688958323e-05, "loss": 0.0522, "step": 17790 }, { "grad_norm": 0.46039190888404846, "learning_rate": 9.558783807627434e-05, "loss": 0.0686, "step": 17800 }, { "grad_norm": 0.6353526711463928, "learning_rate": 9.558104427754801e-05, "loss": 0.0974, "step": 17810 }, { "grad_norm": 0.5202339291572571, "learning_rate": 9.557424549414722e-05, "loss": 0.0852, "step": 17820 }, { "grad_norm": 0.44429469108581543, "learning_rate": 9.556744172681546e-05, "loss": 0.0784, "step": 17830 }, { "grad_norm": 0.9539887309074402, "learning_rate": 9.556063297629677e-05, "loss": 0.0885, "step": 17840 }, { "grad_norm": 0.5785202383995056, "learning_rate": 9.555381924333578e-05, "loss": 0.0664, "step": 17850 }, { "grad_norm": 0.6651585102081299, "learning_rate": 9.554700052867758e-05, "loss": 0.0714, "step": 17860 }, { "grad_norm": 0.7977889776229858, "learning_rate": 9.554017683306789e-05, "loss": 0.0731, "step": 17870 }, { "grad_norm": 0.6524176001548767, "learning_rate": 9.553334815725294e-05, "loss": 0.0682, "step": 17880 }, { "grad_norm": 0.4362110495567322, "learning_rate": 9.552651450197949e-05, "loss": 0.0735, "step": 17890 }, { "grad_norm": 0.5677008032798767, "learning_rate": 9.551967586799486e-05, "loss": 0.0702, "step": 17900 }, { "grad_norm": 0.6366978883743286, "learning_rate": 9.551283225604692e-05, "loss": 0.076, "step": 17910 }, { "grad_norm": 0.7296777367591858, "learning_rate": 9.550598366688406e-05, "loss": 0.077, "step": 17920 }, { "grad_norm": 0.7126582264900208, "learning_rate": 9.549913010125526e-05, "loss": 0.0762, "step": 17930 }, { "grad_norm": 0.9645043015480042, "learning_rate": 9.549227155990999e-05, "loss": 0.0672, "step": 17940 }, { "grad_norm": 0.49512097239494324, "learning_rate": 9.548540804359828e-05, "loss": 0.08, "step": 17950 }, { "grad_norm": 0.8044940829277039, "learning_rate": 9.547853955307077e-05, "loss": 0.0862, "step": 17960 }, { "grad_norm": 0.4459740221500397, "learning_rate": 9.547166608907853e-05, "loss": 0.0681, "step": 17970 }, { "grad_norm": 0.5866318941116333, "learning_rate": 9.546478765237326e-05, "loss": 0.082, "step": 17980 }, { "grad_norm": 0.7160000205039978, "learning_rate": 9.545790424370715e-05, "loss": 0.0756, "step": 17990 }, { "grad_norm": 0.4179820418357849, "learning_rate": 9.5451015863833e-05, "loss": 0.0933, "step": 18000 }, { "grad_norm": 0.7416881322860718, "learning_rate": 9.544412251350408e-05, "loss": 0.0867, "step": 18010 }, { "grad_norm": 0.5776335000991821, "learning_rate": 9.543722419347422e-05, "loss": 0.0788, "step": 18020 }, { "grad_norm": 0.5171349048614502, "learning_rate": 9.543032090449788e-05, "loss": 0.0769, "step": 18030 }, { "grad_norm": 0.603001594543457, "learning_rate": 9.542341264732992e-05, "loss": 0.0956, "step": 18040 }, { "grad_norm": 0.6949207186698914, "learning_rate": 9.541649942272585e-05, "loss": 0.0841, "step": 18050 }, { "grad_norm": 0.5400875806808472, "learning_rate": 9.54095812314417e-05, "loss": 0.0665, "step": 18060 }, { "grad_norm": 0.5294807553291321, "learning_rate": 9.540265807423401e-05, "loss": 0.0672, "step": 18070 }, { "grad_norm": 0.6725385785102844, "learning_rate": 9.53957299518599e-05, "loss": 0.0771, "step": 18080 }, { "grad_norm": 0.36478906869888306, "learning_rate": 9.5388796865077e-05, "loss": 0.0695, "step": 18090 }, { "grad_norm": 0.4096502959728241, "learning_rate": 9.538185881464353e-05, "loss": 0.0862, "step": 18100 }, { "grad_norm": 0.7542049884796143, "learning_rate": 9.537491580131821e-05, "loss": 0.076, "step": 18110 }, { "grad_norm": 0.6750119924545288, "learning_rate": 9.53679678258603e-05, "loss": 0.0731, "step": 18120 }, { "grad_norm": 0.6535876989364624, "learning_rate": 9.536101488902966e-05, "loss": 0.0781, "step": 18130 }, { "grad_norm": 0.550834059715271, "learning_rate": 9.535405699158663e-05, "loss": 0.0722, "step": 18140 }, { "grad_norm": 0.5661576986312866, "learning_rate": 9.53470941342921e-05, "loss": 0.0618, "step": 18150 }, { "grad_norm": 0.3828941583633423, "learning_rate": 9.534012631790756e-05, "loss": 0.0652, "step": 18160 }, { "grad_norm": 0.6616635918617249, "learning_rate": 9.533315354319494e-05, "loss": 0.0733, "step": 18170 }, { "grad_norm": 0.45713716745376587, "learning_rate": 9.532617581091682e-05, "loss": 0.0715, "step": 18180 }, { "grad_norm": 0.5101895928382874, "learning_rate": 9.531919312183629e-05, "loss": 0.0687, "step": 18190 }, { "grad_norm": 0.48439663648605347, "learning_rate": 9.531220547671688e-05, "loss": 0.0634, "step": 18200 }, { "grad_norm": 0.654917299747467, "learning_rate": 9.530521287632285e-05, "loss": 0.078, "step": 18210 }, { "grad_norm": 0.54682457447052, "learning_rate": 9.529821532141884e-05, "loss": 0.0751, "step": 18220 }, { "grad_norm": 0.5792126059532166, "learning_rate": 9.52912128127701e-05, "loss": 0.0741, "step": 18230 }, { "grad_norm": 0.5387780070304871, "learning_rate": 9.528420535114244e-05, "loss": 0.0699, "step": 18240 }, { "grad_norm": 1.2897242307662964, "learning_rate": 9.527719293730215e-05, "loss": 0.0939, "step": 18250 }, { "grad_norm": 0.5099319219589233, "learning_rate": 9.527017557201611e-05, "loss": 0.0799, "step": 18260 }, { "grad_norm": 0.6361300349235535, "learning_rate": 9.526315325605176e-05, "loss": 0.0929, "step": 18270 }, { "grad_norm": 0.44502514600753784, "learning_rate": 9.525612599017699e-05, "loss": 0.061, "step": 18280 }, { "grad_norm": 0.562416136264801, "learning_rate": 9.524909377516033e-05, "loss": 0.0766, "step": 18290 }, { "grad_norm": 0.6675227880477905, "learning_rate": 9.524205661177081e-05, "loss": 0.0743, "step": 18300 }, { "grad_norm": 0.5669130086898804, "learning_rate": 9.523501450077801e-05, "loss": 0.0848, "step": 18310 }, { "grad_norm": 0.512948751449585, "learning_rate": 9.522796744295202e-05, "loss": 0.0773, "step": 18320 }, { "grad_norm": 0.8956956267356873, "learning_rate": 9.522091543906352e-05, "loss": 0.0771, "step": 18330 }, { "grad_norm": 0.8341231346130371, "learning_rate": 9.521385848988369e-05, "loss": 0.0697, "step": 18340 }, { "grad_norm": 0.5633924007415771, "learning_rate": 9.520679659618428e-05, "loss": 0.0677, "step": 18350 }, { "grad_norm": 0.4793887436389923, "learning_rate": 9.519972975873754e-05, "loss": 0.0772, "step": 18360 }, { "grad_norm": 0.6252787113189697, "learning_rate": 9.519265797831633e-05, "loss": 0.0721, "step": 18370 }, { "grad_norm": 0.4581109583377838, "learning_rate": 9.518558125569399e-05, "loss": 0.0718, "step": 18380 }, { "grad_norm": 0.6523916125297546, "learning_rate": 9.517849959164442e-05, "loss": 0.0725, "step": 18390 }, { "grad_norm": 0.5471792221069336, "learning_rate": 9.517141298694205e-05, "loss": 0.0902, "step": 18400 }, { "grad_norm": 0.4614265263080597, "learning_rate": 9.516432144236188e-05, "loss": 0.0738, "step": 18410 }, { "grad_norm": 0.49781033396720886, "learning_rate": 9.515722495867941e-05, "loss": 0.0777, "step": 18420 }, { "grad_norm": 0.8726680278778076, "learning_rate": 9.515012353667072e-05, "loss": 0.0739, "step": 18430 }, { "grad_norm": 1.0036354064941406, "learning_rate": 9.51430171771124e-05, "loss": 0.0707, "step": 18440 }, { "grad_norm": 0.5165112614631653, "learning_rate": 9.513590588078159e-05, "loss": 0.0849, "step": 18450 }, { "grad_norm": 0.7170679569244385, "learning_rate": 9.512878964845597e-05, "loss": 0.0882, "step": 18460 }, { "grad_norm": 0.7567163109779358, "learning_rate": 9.512166848091377e-05, "loss": 0.0909, "step": 18470 }, { "grad_norm": 0.46263280510902405, "learning_rate": 9.511454237893376e-05, "loss": 0.0834, "step": 18480 }, { "grad_norm": 0.6230337023735046, "learning_rate": 9.51074113432952e-05, "loss": 0.0689, "step": 18490 }, { "grad_norm": 0.6288004517555237, "learning_rate": 9.510027537477797e-05, "loss": 0.0871, "step": 18500 }, { "grad_norm": 0.6534849405288696, "learning_rate": 9.509313447416242e-05, "loss": 0.0956, "step": 18510 }, { "grad_norm": 0.7515078186988831, "learning_rate": 9.508598864222949e-05, "loss": 0.0861, "step": 18520 }, { "grad_norm": 0.6899498105049133, "learning_rate": 9.507883787976062e-05, "loss": 0.0734, "step": 18530 }, { "grad_norm": 0.6278219223022461, "learning_rate": 9.507168218753781e-05, "loss": 0.0701, "step": 18540 }, { "grad_norm": 0.5798496603965759, "learning_rate": 9.506452156634362e-05, "loss": 0.0781, "step": 18550 }, { "grad_norm": 0.511540949344635, "learning_rate": 9.505735601696109e-05, "loss": 0.0927, "step": 18560 }, { "grad_norm": 0.7117258906364441, "learning_rate": 9.505018554017385e-05, "loss": 0.0735, "step": 18570 }, { "grad_norm": 0.8313431739807129, "learning_rate": 9.504301013676604e-05, "loss": 0.0687, "step": 18580 }, { "grad_norm": 0.6923441886901855, "learning_rate": 9.503582980752238e-05, "loss": 0.0775, "step": 18590 }, { "grad_norm": 0.6376619338989258, "learning_rate": 9.502864455322809e-05, "loss": 0.0833, "step": 18600 }, { "grad_norm": 0.5186017751693726, "learning_rate": 9.502145437466891e-05, "loss": 0.0736, "step": 18610 }, { "grad_norm": 0.4536381959915161, "learning_rate": 9.501425927263116e-05, "loss": 0.087, "step": 18620 }, { "grad_norm": 0.690570592880249, "learning_rate": 9.500705924790172e-05, "loss": 0.0986, "step": 18630 }, { "grad_norm": 0.8167626261711121, "learning_rate": 9.499985430126794e-05, "loss": 0.0766, "step": 18640 }, { "grad_norm": 0.4773733913898468, "learning_rate": 9.499264443351775e-05, "loss": 0.072, "step": 18650 }, { "grad_norm": 0.5111244320869446, "learning_rate": 9.498542964543961e-05, "loss": 0.0728, "step": 18660 }, { "grad_norm": 0.635668933391571, "learning_rate": 9.497820993782252e-05, "loss": 0.0762, "step": 18670 }, { "grad_norm": 0.6126547455787659, "learning_rate": 9.497098531145601e-05, "loss": 0.0704, "step": 18680 }, { "grad_norm": 0.5094189047813416, "learning_rate": 9.496375576713017e-05, "loss": 0.0729, "step": 18690 }, { "grad_norm": 0.6145473718643188, "learning_rate": 9.49565213056356e-05, "loss": 0.0678, "step": 18700 }, { "grad_norm": 0.46812060475349426, "learning_rate": 9.494928192776342e-05, "loss": 0.0618, "step": 18710 }, { "grad_norm": 0.7081633806228638, "learning_rate": 9.494203763430538e-05, "loss": 0.084, "step": 18720 }, { "grad_norm": 0.6119225025177002, "learning_rate": 9.493478842605366e-05, "loss": 0.0983, "step": 18730 }, { "grad_norm": 2.1363742351531982, "learning_rate": 9.492753430380105e-05, "loss": 0.076, "step": 18740 }, { "grad_norm": 0.75874924659729, "learning_rate": 9.492027526834083e-05, "loss": 0.0808, "step": 18750 }, { "grad_norm": 0.7251395583152771, "learning_rate": 9.491301132046684e-05, "loss": 0.0912, "step": 18760 }, { "grad_norm": 0.5875306725502014, "learning_rate": 9.490574246097345e-05, "loss": 0.0931, "step": 18770 }, { "grad_norm": 0.49797090888023376, "learning_rate": 9.48984686906556e-05, "loss": 0.0847, "step": 18780 }, { "grad_norm": 0.5394578576087952, "learning_rate": 9.489119001030871e-05, "loss": 0.0637, "step": 18790 }, { "grad_norm": 0.5649099946022034, "learning_rate": 9.488390642072878e-05, "loss": 0.0867, "step": 18800 }, { "grad_norm": 0.4600967466831207, "learning_rate": 9.48766179227123e-05, "loss": 0.0788, "step": 18810 }, { "grad_norm": 0.918932318687439, "learning_rate": 9.486932451705636e-05, "loss": 0.0955, "step": 18820 }, { "grad_norm": 0.4269585907459259, "learning_rate": 9.486202620455857e-05, "loss": 0.0761, "step": 18830 }, { "grad_norm": 0.4600907862186432, "learning_rate": 9.485472298601704e-05, "loss": 0.0731, "step": 18840 }, { "grad_norm": 0.7324785590171814, "learning_rate": 9.484741486223043e-05, "loss": 0.0932, "step": 18850 }, { "grad_norm": 0.6017135977745056, "learning_rate": 9.484010183399797e-05, "loss": 0.0743, "step": 18860 }, { "grad_norm": 0.7425228953361511, "learning_rate": 9.483278390211938e-05, "loss": 0.0764, "step": 18870 }, { "grad_norm": 0.6236292123794556, "learning_rate": 9.482546106739496e-05, "loss": 0.0652, "step": 18880 }, { "grad_norm": 0.6469310522079468, "learning_rate": 9.48181333306255e-05, "loss": 0.0663, "step": 18890 }, { "grad_norm": 0.6015981435775757, "learning_rate": 9.481080069261237e-05, "loss": 0.0891, "step": 18900 }, { "grad_norm": 0.8127552270889282, "learning_rate": 9.480346315415745e-05, "loss": 0.0785, "step": 18910 }, { "grad_norm": 0.6668686866760254, "learning_rate": 9.479612071606314e-05, "loss": 0.0735, "step": 18920 }, { "grad_norm": 0.4603619873523712, "learning_rate": 9.478877337913244e-05, "loss": 0.074, "step": 18930 }, { "grad_norm": 0.5568247437477112, "learning_rate": 9.478142114416881e-05, "loss": 0.063, "step": 18940 }, { "grad_norm": 0.5155718326568604, "learning_rate": 9.47740640119763e-05, "loss": 0.0639, "step": 18950 }, { "grad_norm": 0.544976532459259, "learning_rate": 9.476670198335947e-05, "loss": 0.0818, "step": 18960 }, { "grad_norm": 0.519070565700531, "learning_rate": 9.47593350591234e-05, "loss": 0.0681, "step": 18970 }, { "grad_norm": 0.4566664397716522, "learning_rate": 9.475196324007376e-05, "loss": 0.0793, "step": 18980 }, { "grad_norm": 0.5001437664031982, "learning_rate": 9.474458652701669e-05, "loss": 0.0724, "step": 18990 }, { "grad_norm": 0.6990273594856262, "learning_rate": 9.473720492075892e-05, "loss": 0.0787, "step": 19000 }, { "grad_norm": 0.5787305235862732, "learning_rate": 9.472981842210768e-05, "loss": 0.0975, "step": 19010 }, { "grad_norm": 1.2192471027374268, "learning_rate": 9.472242703187074e-05, "loss": 0.0832, "step": 19020 }, { "grad_norm": 0.5032941102981567, "learning_rate": 9.471503075085643e-05, "loss": 0.0759, "step": 19030 }, { "grad_norm": 0.7650747299194336, "learning_rate": 9.470762957987359e-05, "loss": 0.0861, "step": 19040 }, { "grad_norm": 0.7179062366485596, "learning_rate": 9.470022351973158e-05, "loss": 0.0787, "step": 19050 }, { "grad_norm": 0.6609879732131958, "learning_rate": 9.469281257124034e-05, "loss": 0.0708, "step": 19060 }, { "grad_norm": 0.40562450885772705, "learning_rate": 9.46853967352103e-05, "loss": 0.0754, "step": 19070 }, { "grad_norm": 1.3017523288726807, "learning_rate": 9.467797601245246e-05, "loss": 0.0857, "step": 19080 }, { "grad_norm": 1.1051534414291382, "learning_rate": 9.467055040377834e-05, "loss": 0.0833, "step": 19090 }, { "grad_norm": 0.47751346230506897, "learning_rate": 9.466311990999999e-05, "loss": 0.0883, "step": 19100 }, { "grad_norm": 0.5369318723678589, "learning_rate": 9.465568453193e-05, "loss": 0.0702, "step": 19110 }, { "grad_norm": 0.4069730043411255, "learning_rate": 9.464824427038148e-05, "loss": 0.0609, "step": 19120 }, { "grad_norm": 0.6035214066505432, "learning_rate": 9.46407991261681e-05, "loss": 0.0913, "step": 19130 }, { "grad_norm": 0.6095861196517944, "learning_rate": 9.463334910010404e-05, "loss": 0.076, "step": 19140 }, { "grad_norm": 0.49531519412994385, "learning_rate": 9.462589419300403e-05, "loss": 0.0647, "step": 19150 }, { "grad_norm": 0.4731869697570801, "learning_rate": 9.461843440568333e-05, "loss": 0.0796, "step": 19160 }, { "grad_norm": 0.5022947192192078, "learning_rate": 9.461096973895773e-05, "loss": 0.0746, "step": 19170 }, { "grad_norm": 0.4811605215072632, "learning_rate": 9.460350019364355e-05, "loss": 0.0742, "step": 19180 }, { "grad_norm": 0.7249240875244141, "learning_rate": 9.459602577055764e-05, "loss": 0.0916, "step": 19190 }, { "grad_norm": 0.594458281993866, "learning_rate": 9.45885464705174e-05, "loss": 0.065, "step": 19200 }, { "grad_norm": 0.5757800340652466, "learning_rate": 9.458106229434076e-05, "loss": 0.0781, "step": 19210 }, { "grad_norm": 0.6777529716491699, "learning_rate": 9.457357324284617e-05, "loss": 0.0666, "step": 19220 }, { "grad_norm": 0.7558194994926453, "learning_rate": 9.456607931685262e-05, "loss": 0.086, "step": 19230 }, { "grad_norm": 0.48494061827659607, "learning_rate": 9.455858051717965e-05, "loss": 0.075, "step": 19240 }, { "grad_norm": 0.5529453158378601, "learning_rate": 9.45510768446473e-05, "loss": 0.0675, "step": 19250 }, { "grad_norm": 0.6761119961738586, "learning_rate": 9.454356830007618e-05, "loss": 0.0764, "step": 19260 }, { "grad_norm": 1.2180073261260986, "learning_rate": 9.45360548842874e-05, "loss": 0.0691, "step": 19270 }, { "grad_norm": 0.6565360426902771, "learning_rate": 9.452853659810261e-05, "loss": 0.0671, "step": 19280 }, { "grad_norm": 0.7710278034210205, "learning_rate": 9.452101344234401e-05, "loss": 0.0954, "step": 19290 }, { "grad_norm": 0.48478150367736816, "learning_rate": 9.451348541783431e-05, "loss": 0.0714, "step": 19300 }, { "grad_norm": 0.7182620763778687, "learning_rate": 9.450595252539678e-05, "loss": 0.0738, "step": 19310 }, { "grad_norm": 0.6393773555755615, "learning_rate": 9.449841476585518e-05, "loss": 0.0803, "step": 19320 }, { "grad_norm": 0.595012366771698, "learning_rate": 9.449087214003384e-05, "loss": 0.0898, "step": 19330 }, { "grad_norm": 0.7181703448295593, "learning_rate": 9.448332464875765e-05, "loss": 0.0991, "step": 19340 }, { "grad_norm": 0.778222918510437, "learning_rate": 9.447577229285192e-05, "loss": 0.0948, "step": 19350 }, { "grad_norm": 0.4378495514392853, "learning_rate": 9.446821507314261e-05, "loss": 0.0681, "step": 19360 }, { "grad_norm": 0.5960466861724854, "learning_rate": 9.446065299045617e-05, "loss": 0.0897, "step": 19370 }, { "grad_norm": 0.6374331712722778, "learning_rate": 9.445308604561955e-05, "loss": 0.0897, "step": 19380 }, { "grad_norm": 1.107153296470642, "learning_rate": 9.444551423946028e-05, "loss": 0.0655, "step": 19390 }, { "grad_norm": 0.7210219502449036, "learning_rate": 9.443793757280638e-05, "loss": 0.0788, "step": 19400 }, { "grad_norm": 0.5732666850090027, "learning_rate": 9.443035604648646e-05, "loss": 0.0786, "step": 19410 }, { "grad_norm": 0.5562695860862732, "learning_rate": 9.44227696613296e-05, "loss": 0.0715, "step": 19420 }, { "grad_norm": 0.8496845364570618, "learning_rate": 9.441517841816542e-05, "loss": 0.1062, "step": 19430 }, { "grad_norm": 0.5544634461402893, "learning_rate": 9.440758231782413e-05, "loss": 0.0757, "step": 19440 }, { "grad_norm": 0.4967092275619507, "learning_rate": 9.439998136113639e-05, "loss": 0.0719, "step": 19450 }, { "grad_norm": 0.5780096054077148, "learning_rate": 9.439237554893344e-05, "loss": 0.0764, "step": 19460 }, { "grad_norm": 0.5637192726135254, "learning_rate": 9.438476488204705e-05, "loss": 0.0773, "step": 19470 }, { "grad_norm": 0.4749239981174469, "learning_rate": 9.43771493613095e-05, "loss": 0.0639, "step": 19480 }, { "grad_norm": 0.7324615716934204, "learning_rate": 9.436952898755362e-05, "loss": 0.0891, "step": 19490 }, { "grad_norm": 0.6452179551124573, "learning_rate": 9.436190376161276e-05, "loss": 0.0762, "step": 19500 }, { "grad_norm": 0.6266588568687439, "learning_rate": 9.43542736843208e-05, "loss": 0.1033, "step": 19510 }, { "grad_norm": 0.7790524959564209, "learning_rate": 9.434663875651216e-05, "loss": 0.0764, "step": 19520 }, { "grad_norm": 0.5177575349807739, "learning_rate": 9.433899897902177e-05, "loss": 0.0798, "step": 19530 }, { "grad_norm": 0.5030081868171692, "learning_rate": 9.433135435268511e-05, "loss": 0.0571, "step": 19540 }, { "grad_norm": 0.45407992601394653, "learning_rate": 9.432370487833819e-05, "loss": 0.072, "step": 19550 }, { "grad_norm": 0.5541645884513855, "learning_rate": 9.431605055681756e-05, "loss": 0.0723, "step": 19560 }, { "grad_norm": 0.7096638679504395, "learning_rate": 9.430839138896026e-05, "loss": 0.0731, "step": 19570 }, { "grad_norm": 0.6288067102432251, "learning_rate": 9.43007273756039e-05, "loss": 0.0757, "step": 19580 }, { "grad_norm": 0.717628002166748, "learning_rate": 9.429305851758658e-05, "loss": 0.0818, "step": 19590 }, { "grad_norm": 0.6566044688224792, "learning_rate": 9.428538481574699e-05, "loss": 0.0837, "step": 19600 }, { "grad_norm": 0.5810635089874268, "learning_rate": 9.42777062709243e-05, "loss": 0.0697, "step": 19610 }, { "grad_norm": 0.3616519272327423, "learning_rate": 9.427002288395821e-05, "loss": 0.0664, "step": 19620 }, { "grad_norm": 0.5240603089332581, "learning_rate": 9.426233465568898e-05, "loss": 0.0714, "step": 19630 }, { "grad_norm": 0.5117831826210022, "learning_rate": 9.42546415869574e-05, "loss": 0.0638, "step": 19640 }, { "grad_norm": 0.3971326947212219, "learning_rate": 9.424694367860473e-05, "loss": 0.0575, "step": 19650 }, { "grad_norm": 0.5511816740036011, "learning_rate": 9.423924093147284e-05, "loss": 0.0637, "step": 19660 }, { "grad_norm": 0.41423869132995605, "learning_rate": 9.423153334640407e-05, "loss": 0.0653, "step": 19670 }, { "grad_norm": 1.0483266115188599, "learning_rate": 9.42238209242413e-05, "loss": 0.0599, "step": 19680 }, { "grad_norm": 0.6544947028160095, "learning_rate": 9.421610366582798e-05, "loss": 0.0842, "step": 19690 }, { "grad_norm": 0.6200827956199646, "learning_rate": 9.420838157200803e-05, "loss": 0.0708, "step": 19700 }, { "grad_norm": 0.46920856833457947, "learning_rate": 9.420065464362594e-05, "loss": 0.0857, "step": 19710 }, { "grad_norm": 0.5924236178398132, "learning_rate": 9.419292288152673e-05, "loss": 0.0676, "step": 19720 }, { "grad_norm": 0.5465502738952637, "learning_rate": 9.418518628655588e-05, "loss": 0.0725, "step": 19730 }, { "grad_norm": 0.5175719857215881, "learning_rate": 9.417744485955951e-05, "loss": 0.0666, "step": 19740 }, { "grad_norm": 0.4838004410266876, "learning_rate": 9.41696986013842e-05, "loss": 0.0645, "step": 19750 }, { "grad_norm": 0.6757290363311768, "learning_rate": 9.416194751287705e-05, "loss": 0.0701, "step": 19760 }, { "grad_norm": 0.5799652934074402, "learning_rate": 9.415419159488572e-05, "loss": 0.0604, "step": 19770 }, { "grad_norm": 0.9939319491386414, "learning_rate": 9.414643084825837e-05, "loss": 0.0742, "step": 19780 }, { "grad_norm": 0.6012560725212097, "learning_rate": 9.413866527384372e-05, "loss": 0.0754, "step": 19790 }, { "grad_norm": 0.5188357830047607, "learning_rate": 9.4130894872491e-05, "loss": 0.0715, "step": 19800 }, { "grad_norm": 0.6598438620567322, "learning_rate": 9.412311964504998e-05, "loss": 0.0792, "step": 19810 }, { "grad_norm": 0.537165641784668, "learning_rate": 9.411533959237091e-05, "loss": 0.0859, "step": 19820 }, { "grad_norm": 0.49519991874694824, "learning_rate": 9.410755471530464e-05, "loss": 0.054, "step": 19830 }, { "grad_norm": 0.6014136672019958, "learning_rate": 9.40997650147025e-05, "loss": 0.0743, "step": 19840 }, { "grad_norm": 0.6602025032043457, "learning_rate": 9.409197049141637e-05, "loss": 0.0862, "step": 19850 }, { "grad_norm": 0.595971405506134, "learning_rate": 9.408417114629863e-05, "loss": 0.0834, "step": 19860 }, { "grad_norm": 0.628408670425415, "learning_rate": 9.40763669802022e-05, "loss": 0.07, "step": 19870 }, { "grad_norm": 0.5784828066825867, "learning_rate": 9.406855799398056e-05, "loss": 0.0833, "step": 19880 }, { "grad_norm": 0.7732696533203125, "learning_rate": 9.406074418848767e-05, "loss": 0.0856, "step": 19890 }, { "grad_norm": 0.6935335397720337, "learning_rate": 9.405292556457805e-05, "loss": 0.0924, "step": 19900 }, { "grad_norm": 0.4742138385772705, "learning_rate": 9.404510212310671e-05, "loss": 0.0721, "step": 19910 }, { "grad_norm": 0.4972801208496094, "learning_rate": 9.403727386492924e-05, "loss": 0.0687, "step": 19920 }, { "grad_norm": 0.4328390061855316, "learning_rate": 9.40294407909017e-05, "loss": 0.0669, "step": 19930 }, { "grad_norm": 0.6165010333061218, "learning_rate": 9.40216029018807e-05, "loss": 0.0644, "step": 19940 }, { "grad_norm": 0.5265051126480103, "learning_rate": 9.401376019872338e-05, "loss": 0.0677, "step": 19950 }, { "grad_norm": 0.4763129651546478, "learning_rate": 9.400591268228746e-05, "loss": 0.0616, "step": 19960 }, { "grad_norm": 0.6514252424240112, "learning_rate": 9.399806035343106e-05, "loss": 0.0726, "step": 19970 }, { "grad_norm": 0.6105881333351135, "learning_rate": 9.399020321301294e-05, "loss": 0.0705, "step": 19980 }, { "grad_norm": 0.5821778178215027, "learning_rate": 9.398234126189234e-05, "loss": 0.0823, "step": 19990 }, { "grad_norm": 1.543766975402832, "learning_rate": 9.397447450092902e-05, "loss": 0.0844, "step": 20000 }, { "grad_norm": 0.5879276394844055, "learning_rate": 9.39666029309833e-05, "loss": 0.0691, "step": 20010 }, { "grad_norm": 0.8114721179008484, "learning_rate": 9.395872655291596e-05, "loss": 0.0674, "step": 20020 }, { "grad_norm": 0.5685903429985046, "learning_rate": 9.395084536758838e-05, "loss": 0.0669, "step": 20030 }, { "grad_norm": 0.6868470311164856, "learning_rate": 9.394295937586243e-05, "loss": 0.078, "step": 20040 }, { "grad_norm": 0.48384788632392883, "learning_rate": 9.393506857860052e-05, "loss": 0.0746, "step": 20050 }, { "grad_norm": 0.6233245730400085, "learning_rate": 9.392717297666555e-05, "loss": 0.0659, "step": 20060 }, { "grad_norm": 0.7760059237480164, "learning_rate": 9.391927257092101e-05, "loss": 0.0846, "step": 20070 }, { "grad_norm": 0.4200085699558258, "learning_rate": 9.391136736223085e-05, "loss": 0.0595, "step": 20080 }, { "grad_norm": 0.5539615154266357, "learning_rate": 9.390345735145956e-05, "loss": 0.0719, "step": 20090 }, { "grad_norm": 0.5819674134254456, "learning_rate": 9.389554253947219e-05, "loss": 0.0661, "step": 20100 }, { "grad_norm": 0.8059300780296326, "learning_rate": 9.388762292713428e-05, "loss": 0.0798, "step": 20110 }, { "grad_norm": 0.5359142422676086, "learning_rate": 9.38796985153119e-05, "loss": 0.0795, "step": 20120 }, { "grad_norm": 0.5013347864151001, "learning_rate": 9.387176930487169e-05, "loss": 0.073, "step": 20130 }, { "grad_norm": 0.9037903547286987, "learning_rate": 9.386383529668072e-05, "loss": 0.0758, "step": 20140 }, { "grad_norm": 0.40022289752960205, "learning_rate": 9.385589649160669e-05, "loss": 0.0777, "step": 20150 }, { "grad_norm": 0.5840714573860168, "learning_rate": 9.384795289051775e-05, "loss": 0.0873, "step": 20160 }, { "grad_norm": 0.5830475687980652, "learning_rate": 9.384000449428261e-05, "loss": 0.0793, "step": 20170 }, { "grad_norm": 0.6258863806724548, "learning_rate": 9.383205130377048e-05, "loss": 0.0757, "step": 20180 }, { "grad_norm": 0.5806873440742493, "learning_rate": 9.382409331985114e-05, "loss": 0.071, "step": 20190 }, { "grad_norm": 0.5820714831352234, "learning_rate": 9.381613054339482e-05, "loss": 0.065, "step": 20200 }, { "grad_norm": 0.5372800230979919, "learning_rate": 9.380816297527235e-05, "loss": 0.0623, "step": 20210 }, { "grad_norm": 0.5554584264755249, "learning_rate": 9.380019061635506e-05, "loss": 0.0775, "step": 20220 }, { "grad_norm": 0.4135645627975464, "learning_rate": 9.379221346751474e-05, "loss": 0.0798, "step": 20230 }, { "grad_norm": 0.6379081010818481, "learning_rate": 9.378423152962382e-05, "loss": 0.0703, "step": 20240 }, { "grad_norm": 0.5999014973640442, "learning_rate": 9.377624480355517e-05, "loss": 0.0767, "step": 20250 }, { "grad_norm": 0.4503636956214905, "learning_rate": 9.376825329018219e-05, "loss": 0.0766, "step": 20260 }, { "grad_norm": 0.3710686266422272, "learning_rate": 9.376025699037884e-05, "loss": 0.0719, "step": 20270 }, { "grad_norm": 0.46137720346450806, "learning_rate": 9.37522559050196e-05, "loss": 0.07, "step": 20280 }, { "grad_norm": 0.45367366075515747, "learning_rate": 9.37442500349794e-05, "loss": 0.0758, "step": 20290 }, { "grad_norm": 0.6261991262435913, "learning_rate": 9.373623938113381e-05, "loss": 0.0659, "step": 20300 }, { "grad_norm": 0.6365321278572083, "learning_rate": 9.372822394435883e-05, "loss": 0.0841, "step": 20310 }, { "grad_norm": 0.6527341604232788, "learning_rate": 9.372020372553102e-05, "loss": 0.0812, "step": 20320 }, { "grad_norm": 0.5593727827072144, "learning_rate": 9.371217872552746e-05, "loss": 0.0725, "step": 20330 }, { "grad_norm": 0.8056023120880127, "learning_rate": 9.370414894522576e-05, "loss": 0.0725, "step": 20340 }, { "grad_norm": 0.549432098865509, "learning_rate": 9.369611438550406e-05, "loss": 0.0694, "step": 20350 }, { "grad_norm": 0.4272562563419342, "learning_rate": 9.368807504724095e-05, "loss": 0.0642, "step": 20360 }, { "grad_norm": 0.5851807594299316, "learning_rate": 9.368003093131565e-05, "loss": 0.0724, "step": 20370 }, { "grad_norm": 0.5664780139923096, "learning_rate": 9.367198203860785e-05, "loss": 0.0823, "step": 20380 }, { "grad_norm": 0.5592412352561951, "learning_rate": 9.366392836999774e-05, "loss": 0.0614, "step": 20390 }, { "grad_norm": 0.6520680785179138, "learning_rate": 9.365586992636607e-05, "loss": 0.0624, "step": 20400 }, { "grad_norm": 0.5155981779098511, "learning_rate": 9.364780670859412e-05, "loss": 0.0592, "step": 20410 }, { "grad_norm": 0.3911503553390503, "learning_rate": 9.363973871756364e-05, "loss": 0.0657, "step": 20420 }, { "grad_norm": 0.8590943813323975, "learning_rate": 9.363166595415696e-05, "loss": 0.067, "step": 20430 }, { "grad_norm": 1.0354982614517212, "learning_rate": 9.362358841925686e-05, "loss": 0.079, "step": 20440 }, { "grad_norm": 0.4661560356616974, "learning_rate": 9.361550611374674e-05, "loss": 0.072, "step": 20450 }, { "grad_norm": 1.1730570793151855, "learning_rate": 9.360741903851043e-05, "loss": 0.0997, "step": 20460 }, { "grad_norm": 0.6340109705924988, "learning_rate": 9.359932719443236e-05, "loss": 0.0963, "step": 20470 }, { "grad_norm": 0.36014285683631897, "learning_rate": 9.35912305823974e-05, "loss": 0.0632, "step": 20480 }, { "grad_norm": 0.5126276016235352, "learning_rate": 9.358312920329101e-05, "loss": 0.0631, "step": 20490 }, { "grad_norm": 0.5462815761566162, "learning_rate": 9.357502305799914e-05, "loss": 0.0611, "step": 20500 }, { "grad_norm": 0.431607723236084, "learning_rate": 9.356691214740824e-05, "loss": 0.0551, "step": 20510 }, { "grad_norm": 0.5155755281448364, "learning_rate": 9.355879647240535e-05, "loss": 0.0622, "step": 20520 }, { "grad_norm": 1.061104416847229, "learning_rate": 9.355067603387798e-05, "loss": 0.0683, "step": 20530 }, { "grad_norm": 0.6490331292152405, "learning_rate": 9.354255083271412e-05, "loss": 0.076, "step": 20540 }, { "grad_norm": 0.5846430659294128, "learning_rate": 9.353442086980239e-05, "loss": 0.0785, "step": 20550 }, { "grad_norm": 0.6012140512466431, "learning_rate": 9.352628614603185e-05, "loss": 0.0747, "step": 20560 }, { "grad_norm": 0.7352594137191772, "learning_rate": 9.351814666229209e-05, "loss": 0.0721, "step": 20570 }, { "grad_norm": 0.8163196444511414, "learning_rate": 9.351000241947324e-05, "loss": 0.0713, "step": 20580 }, { "grad_norm": 0.5587918162345886, "learning_rate": 9.350185341846594e-05, "loss": 0.0796, "step": 20590 }, { "grad_norm": 0.8384374380111694, "learning_rate": 9.349369966016134e-05, "loss": 0.0807, "step": 20600 }, { "grad_norm": 0.8212686777114868, "learning_rate": 9.348554114545117e-05, "loss": 0.0966, "step": 20610 }, { "grad_norm": 0.5439570546150208, "learning_rate": 9.347737787522758e-05, "loss": 0.0762, "step": 20620 }, { "grad_norm": 0.5984358787536621, "learning_rate": 9.346920985038332e-05, "loss": 0.0869, "step": 20630 }, { "grad_norm": 0.5119311809539795, "learning_rate": 9.346103707181162e-05, "loss": 0.0698, "step": 20640 }, { "grad_norm": 0.517364501953125, "learning_rate": 9.345285954040626e-05, "loss": 0.0632, "step": 20650 }, { "grad_norm": 0.6724367737770081, "learning_rate": 9.34446772570615e-05, "loss": 0.0669, "step": 20660 }, { "grad_norm": 0.5676357746124268, "learning_rate": 9.343649022267214e-05, "loss": 0.0609, "step": 20670 }, { "grad_norm": 0.8689848184585571, "learning_rate": 9.342829843813353e-05, "loss": 0.0829, "step": 20680 }, { "grad_norm": 0.5399599075317383, "learning_rate": 9.342010190434149e-05, "loss": 0.062, "step": 20690 }, { "grad_norm": 0.5081528425216675, "learning_rate": 9.34119006221924e-05, "loss": 0.0689, "step": 20700 }, { "grad_norm": 0.4563450813293457, "learning_rate": 9.340369459258313e-05, "loss": 0.0584, "step": 20710 }, { "grad_norm": 0.5773863196372986, "learning_rate": 9.339548381641106e-05, "loss": 0.072, "step": 20720 }, { "grad_norm": 0.5137078166007996, "learning_rate": 9.338726829457413e-05, "loss": 0.0763, "step": 20730 }, { "grad_norm": 0.5292121767997742, "learning_rate": 9.337904802797078e-05, "loss": 0.0591, "step": 20740 }, { "grad_norm": 0.552216112613678, "learning_rate": 9.337082301749993e-05, "loss": 0.0712, "step": 20750 }, { "grad_norm": 0.5929949283599854, "learning_rate": 9.336259326406109e-05, "loss": 0.0735, "step": 20760 }, { "grad_norm": 0.7261190414428711, "learning_rate": 9.335435876855427e-05, "loss": 0.0835, "step": 20770 }, { "grad_norm": 0.5210738778114319, "learning_rate": 9.334611953187994e-05, "loss": 0.0815, "step": 20780 }, { "grad_norm": 0.4998683035373688, "learning_rate": 9.333787555493914e-05, "loss": 0.0671, "step": 20790 }, { "grad_norm": 0.5656333565711975, "learning_rate": 9.332962683863345e-05, "loss": 0.083, "step": 20800 }, { "grad_norm": 0.7115816473960876, "learning_rate": 9.332137338386489e-05, "loss": 0.063, "step": 20810 }, { "grad_norm": 0.5474618077278137, "learning_rate": 9.33131151915361e-05, "loss": 0.0792, "step": 20820 }, { "grad_norm": 0.6044096350669861, "learning_rate": 9.330485226255012e-05, "loss": 0.0666, "step": 20830 }, { "grad_norm": 0.561316192150116, "learning_rate": 9.329658459781061e-05, "loss": 0.074, "step": 20840 }, { "grad_norm": 0.6224427819252014, "learning_rate": 9.328831219822172e-05, "loss": 0.0757, "step": 20850 }, { "grad_norm": 0.5574579238891602, "learning_rate": 9.328003506468808e-05, "loss": 0.0782, "step": 20860 }, { "grad_norm": 0.5448927283287048, "learning_rate": 9.327175319811488e-05, "loss": 0.0708, "step": 20870 }, { "grad_norm": 0.5441024303436279, "learning_rate": 9.326346659940781e-05, "loss": 0.0677, "step": 20880 }, { "grad_norm": 0.5181921720504761, "learning_rate": 9.325517526947308e-05, "loss": 0.071, "step": 20890 }, { "grad_norm": 0.6062695384025574, "learning_rate": 9.32468792092174e-05, "loss": 0.0733, "step": 20900 }, { "grad_norm": 1.838491678237915, "learning_rate": 9.323857841954803e-05, "loss": 0.0828, "step": 20910 }, { "grad_norm": 0.7460097074508667, "learning_rate": 9.323027290137276e-05, "loss": 0.0752, "step": 20920 }, { "grad_norm": 0.6173479557037354, "learning_rate": 9.322196265559981e-05, "loss": 0.0866, "step": 20930 }, { "grad_norm": 0.41546815633773804, "learning_rate": 9.321364768313803e-05, "loss": 0.075, "step": 20940 }, { "grad_norm": 0.46494823694229126, "learning_rate": 9.32053279848967e-05, "loss": 0.0748, "step": 20950 }, { "grad_norm": 0.3723190426826477, "learning_rate": 9.319700356178567e-05, "loss": 0.0752, "step": 20960 }, { "grad_norm": 0.5236832499504089, "learning_rate": 9.318867441471527e-05, "loss": 0.0713, "step": 20970 }, { "grad_norm": 0.5201197862625122, "learning_rate": 9.318034054459637e-05, "loss": 0.0555, "step": 20980 }, { "grad_norm": 0.4973087012767792, "learning_rate": 9.317200195234034e-05, "loss": 0.0731, "step": 20990 }, { "grad_norm": 0.5250828266143799, "learning_rate": 9.316365863885909e-05, "loss": 0.0818, "step": 21000 }, { "grad_norm": 0.48121383786201477, "learning_rate": 9.315531060506502e-05, "loss": 0.08, "step": 21010 }, { "grad_norm": 0.6713453531265259, "learning_rate": 9.314695785187108e-05, "loss": 0.064, "step": 21020 }, { "grad_norm": 0.561259925365448, "learning_rate": 9.313860038019069e-05, "loss": 0.0636, "step": 21030 }, { "grad_norm": 0.6843715906143188, "learning_rate": 9.313023819093782e-05, "loss": 0.0818, "step": 21040 }, { "grad_norm": 0.5164893269538879, "learning_rate": 9.312187128502695e-05, "loss": 0.0723, "step": 21050 }, { "grad_norm": 0.4855506420135498, "learning_rate": 9.311349966337307e-05, "loss": 0.0723, "step": 21060 }, { "grad_norm": 0.5972747802734375, "learning_rate": 9.310512332689169e-05, "loss": 0.0725, "step": 21070 }, { "grad_norm": 0.41478198766708374, "learning_rate": 9.309674227649883e-05, "loss": 0.054, "step": 21080 }, { "grad_norm": 0.7555734515190125, "learning_rate": 9.308835651311103e-05, "loss": 0.0715, "step": 21090 }, { "grad_norm": 0.6853691339492798, "learning_rate": 9.307996603764533e-05, "loss": 0.0764, "step": 21100 }, { "grad_norm": 0.5320115089416504, "learning_rate": 9.307157085101932e-05, "loss": 0.0701, "step": 21110 }, { "grad_norm": 0.5735618472099304, "learning_rate": 9.306317095415109e-05, "loss": 0.0595, "step": 21120 }, { "grad_norm": 0.5024006962776184, "learning_rate": 9.305476634795922e-05, "loss": 0.0731, "step": 21130 }, { "grad_norm": 0.5181564688682556, "learning_rate": 9.304635703336284e-05, "loss": 0.0673, "step": 21140 }, { "grad_norm": 0.5131558179855347, "learning_rate": 9.303794301128157e-05, "loss": 0.0595, "step": 21150 }, { "grad_norm": 0.8209365010261536, "learning_rate": 9.302952428263555e-05, "loss": 0.0634, "step": 21160 }, { "grad_norm": 0.6629045009613037, "learning_rate": 9.302110084834545e-05, "loss": 0.0699, "step": 21170 }, { "grad_norm": 0.5944706201553345, "learning_rate": 9.301267270933245e-05, "loss": 0.0615, "step": 21180 }, { "grad_norm": 0.5140297412872314, "learning_rate": 9.300423986651823e-05, "loss": 0.0695, "step": 21190 }, { "grad_norm": 0.6472382545471191, "learning_rate": 9.299580232082501e-05, "loss": 0.0744, "step": 21200 }, { "grad_norm": 0.5310013890266418, "learning_rate": 9.298736007317547e-05, "loss": 0.0622, "step": 21210 }, { "grad_norm": 0.6755208969116211, "learning_rate": 9.297891312449288e-05, "loss": 0.0718, "step": 21220 }, { "grad_norm": 0.8290214538574219, "learning_rate": 9.297046147570094e-05, "loss": 0.0835, "step": 21230 }, { "grad_norm": 0.4395337998867035, "learning_rate": 9.296200512772396e-05, "loss": 0.0596, "step": 21240 }, { "grad_norm": 0.3053964078426361, "learning_rate": 9.295354408148668e-05, "loss": 0.0595, "step": 21250 }, { "grad_norm": 0.6148586273193359, "learning_rate": 9.294507833791441e-05, "loss": 0.0638, "step": 21260 }, { "grad_norm": 0.9550206065177917, "learning_rate": 9.293660789793295e-05, "loss": 0.0699, "step": 21270 }, { "grad_norm": 0.42598849534988403, "learning_rate": 9.292813276246858e-05, "loss": 0.0669, "step": 21280 }, { "grad_norm": 0.6080660223960876, "learning_rate": 9.291965293244816e-05, "loss": 0.0713, "step": 21290 }, { "grad_norm": 0.34548133611679077, "learning_rate": 9.291116840879904e-05, "loss": 0.0554, "step": 21300 }, { "grad_norm": 0.7282216548919678, "learning_rate": 9.290267919244904e-05, "loss": 0.071, "step": 21310 }, { "grad_norm": 0.9117546081542969, "learning_rate": 9.289418528432655e-05, "loss": 0.0658, "step": 21320 }, { "grad_norm": 0.6661195158958435, "learning_rate": 9.288568668536045e-05, "loss": 0.0933, "step": 21330 }, { "grad_norm": 0.4497717618942261, "learning_rate": 9.287718339648013e-05, "loss": 0.0626, "step": 21340 }, { "grad_norm": 0.6532623767852783, "learning_rate": 9.28686754186155e-05, "loss": 0.0675, "step": 21350 }, { "grad_norm": 0.5341510772705078, "learning_rate": 9.286016275269698e-05, "loss": 0.0895, "step": 21360 }, { "grad_norm": 0.5580918788909912, "learning_rate": 9.285164539965551e-05, "loss": 0.0745, "step": 21370 }, { "grad_norm": 0.4425353705883026, "learning_rate": 9.284312336042251e-05, "loss": 0.0652, "step": 21380 }, { "grad_norm": 0.4902373254299164, "learning_rate": 9.283459663592996e-05, "loss": 0.0737, "step": 21390 }, { "grad_norm": 0.601185142993927, "learning_rate": 9.282606522711033e-05, "loss": 0.0743, "step": 21400 }, { "grad_norm": 0.5536952614784241, "learning_rate": 9.281752913489657e-05, "loss": 0.0892, "step": 21410 }, { "grad_norm": 0.7331507802009583, "learning_rate": 9.280898836022222e-05, "loss": 0.07, "step": 21420 }, { "grad_norm": 0.45567992329597473, "learning_rate": 9.280044290402126e-05, "loss": 0.064, "step": 21430 }, { "grad_norm": 0.5982648134231567, "learning_rate": 9.279189276722821e-05, "loss": 0.065, "step": 21440 }, { "grad_norm": 0.4674210250377655, "learning_rate": 9.278333795077812e-05, "loss": 0.0778, "step": 21450 }, { "grad_norm": 0.5741408467292786, "learning_rate": 9.27747784556065e-05, "loss": 0.062, "step": 21460 }, { "grad_norm": 0.5433898568153381, "learning_rate": 9.276621428264942e-05, "loss": 0.0812, "step": 21470 }, { "grad_norm": 0.7514809966087341, "learning_rate": 9.275764543284345e-05, "loss": 0.0786, "step": 21480 }, { "grad_norm": 0.49295103549957275, "learning_rate": 9.274907190712566e-05, "loss": 0.0706, "step": 21490 }, { "grad_norm": 0.8236766457557678, "learning_rate": 9.274049370643363e-05, "loss": 0.0636, "step": 21500 }, { "grad_norm": 0.530345618724823, "learning_rate": 9.273191083170547e-05, "loss": 0.0761, "step": 21510 }, { "grad_norm": 0.5930851101875305, "learning_rate": 9.27233232838798e-05, "loss": 0.0648, "step": 21520 }, { "grad_norm": 0.5606374144554138, "learning_rate": 9.27147310638957e-05, "loss": 0.0717, "step": 21530 }, { "grad_norm": 0.5238965749740601, "learning_rate": 9.270613417269286e-05, "loss": 0.0772, "step": 21540 }, { "grad_norm": 0.5079407691955566, "learning_rate": 9.269753261121138e-05, "loss": 0.0725, "step": 21550 }, { "grad_norm": 0.3723219037055969, "learning_rate": 9.268892638039194e-05, "loss": 0.0756, "step": 21560 }, { "grad_norm": 0.5053583979606628, "learning_rate": 9.268031548117569e-05, "loss": 0.0765, "step": 21570 }, { "grad_norm": 0.46078401803970337, "learning_rate": 9.26716999145043e-05, "loss": 0.0564, "step": 21580 }, { "grad_norm": 0.5715039372444153, "learning_rate": 9.266307968131998e-05, "loss": 0.0612, "step": 21590 }, { "grad_norm": 0.5562132000923157, "learning_rate": 9.26544547825654e-05, "loss": 0.0602, "step": 21600 }, { "grad_norm": 0.4082188308238983, "learning_rate": 9.264582521918376e-05, "loss": 0.0635, "step": 21610 }, { "grad_norm": 0.5276612639427185, "learning_rate": 9.263719099211881e-05, "loss": 0.0566, "step": 21620 }, { "grad_norm": 0.4082368016242981, "learning_rate": 9.262855210231476e-05, "loss": 0.0792, "step": 21630 }, { "grad_norm": 0.40742602944374084, "learning_rate": 9.261990855071633e-05, "loss": 0.0641, "step": 21640 }, { "grad_norm": 0.5024327039718628, "learning_rate": 9.261126033826878e-05, "loss": 0.062, "step": 21650 }, { "grad_norm": 0.5870871543884277, "learning_rate": 9.260260746591786e-05, "loss": 0.0615, "step": 21660 }, { "grad_norm": 0.5986630916595459, "learning_rate": 9.259394993460985e-05, "loss": 0.0608, "step": 21670 }, { "grad_norm": 0.4096441864967346, "learning_rate": 9.258528774529151e-05, "loss": 0.0796, "step": 21680 }, { "grad_norm": 1.1347160339355469, "learning_rate": 9.257662089891013e-05, "loss": 0.0691, "step": 21690 }, { "grad_norm": 0.699593186378479, "learning_rate": 9.25679493964135e-05, "loss": 0.0776, "step": 21700 }, { "grad_norm": 0.5149637460708618, "learning_rate": 9.255927323874994e-05, "loss": 0.0617, "step": 21710 }, { "grad_norm": 0.568725049495697, "learning_rate": 9.255059242686822e-05, "loss": 0.0791, "step": 21720 }, { "grad_norm": 0.5004456043243408, "learning_rate": 9.254190696171769e-05, "loss": 0.0679, "step": 21730 }, { "grad_norm": 0.7526355981826782, "learning_rate": 9.25332168442482e-05, "loss": 0.0868, "step": 21740 }, { "grad_norm": 0.7879564762115479, "learning_rate": 9.252452207541004e-05, "loss": 0.0637, "step": 21750 }, { "grad_norm": 0.6988688707351685, "learning_rate": 9.251582265615409e-05, "loss": 0.0969, "step": 21760 }, { "grad_norm": 0.549152672290802, "learning_rate": 9.250711858743169e-05, "loss": 0.0786, "step": 21770 }, { "grad_norm": 0.4982289969921112, "learning_rate": 9.24984098701947e-05, "loss": 0.0922, "step": 21780 }, { "grad_norm": 0.47630831599235535, "learning_rate": 9.248969650539552e-05, "loss": 0.0742, "step": 21790 }, { "grad_norm": 0.7366946935653687, "learning_rate": 9.2480978493987e-05, "loss": 0.0717, "step": 21800 }, { "grad_norm": 0.6044705510139465, "learning_rate": 9.247225583692256e-05, "loss": 0.0658, "step": 21810 }, { "grad_norm": 0.503470778465271, "learning_rate": 9.246352853515607e-05, "loss": 0.068, "step": 21820 }, { "grad_norm": 1.059442162513733, "learning_rate": 9.245479658964194e-05, "loss": 0.0725, "step": 21830 }, { "grad_norm": 0.5681504011154175, "learning_rate": 9.244606000133507e-05, "loss": 0.079, "step": 21840 }, { "grad_norm": 0.4747937023639679, "learning_rate": 9.24373187711909e-05, "loss": 0.0714, "step": 21850 }, { "grad_norm": 0.454086571931839, "learning_rate": 9.242857290016537e-05, "loss": 0.0688, "step": 21860 }, { "grad_norm": 0.46564629673957825, "learning_rate": 9.241982238921488e-05, "loss": 0.0821, "step": 21870 }, { "grad_norm": 0.4233429729938507, "learning_rate": 9.24110672392964e-05, "loss": 0.0486, "step": 21880 }, { "grad_norm": 0.5259076356887817, "learning_rate": 9.240230745136737e-05, "loss": 0.0647, "step": 21890 }, { "grad_norm": 0.5998644828796387, "learning_rate": 9.239354302638575e-05, "loss": 0.0589, "step": 21900 }, { "grad_norm": 0.7221060395240784, "learning_rate": 9.238477396531e-05, "loss": 0.0701, "step": 21910 }, { "grad_norm": 1.6145492792129517, "learning_rate": 9.23760002690991e-05, "loss": 0.0685, "step": 21920 }, { "grad_norm": 0.5236430764198303, "learning_rate": 9.236722193871252e-05, "loss": 0.0564, "step": 21930 }, { "grad_norm": 0.44893303513526917, "learning_rate": 9.235843897511023e-05, "loss": 0.0702, "step": 21940 }, { "grad_norm": 0.5670191049575806, "learning_rate": 9.234965137925276e-05, "loss": 0.0503, "step": 21950 }, { "grad_norm": 0.5608376860618591, "learning_rate": 9.234085915210108e-05, "loss": 0.0605, "step": 21960 }, { "grad_norm": 0.5204911231994629, "learning_rate": 9.23320622946167e-05, "loss": 0.0587, "step": 21970 }, { "grad_norm": 0.47406211495399475, "learning_rate": 9.232326080776163e-05, "loss": 0.0488, "step": 21980 }, { "grad_norm": 0.7728177905082703, "learning_rate": 9.23144546924984e-05, "loss": 0.0566, "step": 21990 }, { "grad_norm": 0.4741659462451935, "learning_rate": 9.230564394979e-05, "loss": 0.0593, "step": 22000 }, { "grad_norm": 0.5631512403488159, "learning_rate": 9.22968285806e-05, "loss": 0.0547, "step": 22010 }, { "grad_norm": 0.5754753947257996, "learning_rate": 9.228800858589242e-05, "loss": 0.0694, "step": 22020 }, { "grad_norm": 0.7501879930496216, "learning_rate": 9.227918396663179e-05, "loss": 0.0666, "step": 22030 }, { "grad_norm": 0.5892961621284485, "learning_rate": 9.227035472378319e-05, "loss": 0.0614, "step": 22040 }, { "grad_norm": 1.112729549407959, "learning_rate": 9.226152085831213e-05, "loss": 0.0657, "step": 22050 }, { "grad_norm": 0.5867705941200256, "learning_rate": 9.22526823711847e-05, "loss": 0.0644, "step": 22060 }, { "grad_norm": 0.3920689821243286, "learning_rate": 9.224383926336745e-05, "loss": 0.0646, "step": 22070 }, { "grad_norm": 0.4479186236858368, "learning_rate": 9.223499153582744e-05, "loss": 0.0539, "step": 22080 }, { "grad_norm": 0.4385734796524048, "learning_rate": 9.222613918953226e-05, "loss": 0.0708, "step": 22090 }, { "grad_norm": 0.7884175777435303, "learning_rate": 9.221728222544999e-05, "loss": 0.0703, "step": 22100 }, { "grad_norm": 0.6346521973609924, "learning_rate": 9.22084206445492e-05, "loss": 0.0718, "step": 22110 }, { "grad_norm": 0.7828880548477173, "learning_rate": 9.2199554447799e-05, "loss": 0.0836, "step": 22120 }, { "grad_norm": 1.012251377105713, "learning_rate": 9.219068363616897e-05, "loss": 0.0718, "step": 22130 }, { "grad_norm": 0.47013628482818604, "learning_rate": 9.218180821062919e-05, "loss": 0.0701, "step": 22140 }, { "grad_norm": 0.467993825674057, "learning_rate": 9.21729281721503e-05, "loss": 0.0785, "step": 22150 }, { "grad_norm": 0.6486619114875793, "learning_rate": 9.216404352170339e-05, "loss": 0.0797, "step": 22160 }, { "grad_norm": 0.42502665519714355, "learning_rate": 9.215515426026007e-05, "loss": 0.0726, "step": 22170 }, { "grad_norm": 0.539286196231842, "learning_rate": 9.214626038879246e-05, "loss": 0.0745, "step": 22180 }, { "grad_norm": 0.5594690442085266, "learning_rate": 9.21373619082732e-05, "loss": 0.0701, "step": 22190 }, { "grad_norm": 1.2396167516708374, "learning_rate": 9.212845881967535e-05, "loss": 0.0732, "step": 22200 }, { "grad_norm": 0.582679033279419, "learning_rate": 9.211955112397262e-05, "loss": 0.0625, "step": 22210 }, { "grad_norm": 0.4383402466773987, "learning_rate": 9.211063882213909e-05, "loss": 0.0624, "step": 22220 }, { "grad_norm": 0.4904906749725342, "learning_rate": 9.210172191514942e-05, "loss": 0.0605, "step": 22230 }, { "grad_norm": 0.6283860206604004, "learning_rate": 9.209280040397874e-05, "loss": 0.0799, "step": 22240 }, { "grad_norm": 0.5365413427352905, "learning_rate": 9.208387428960268e-05, "loss": 0.056, "step": 22250 }, { "grad_norm": 0.524516761302948, "learning_rate": 9.20749435729974e-05, "loss": 0.0758, "step": 22260 }, { "grad_norm": 0.33470383286476135, "learning_rate": 9.206600825513957e-05, "loss": 0.067, "step": 22270 }, { "grad_norm": 1.04392671585083, "learning_rate": 9.20570683370063e-05, "loss": 0.081, "step": 22280 }, { "grad_norm": 0.6334083676338196, "learning_rate": 9.204812381957528e-05, "loss": 0.0654, "step": 22290 }, { "grad_norm": 0.5274370312690735, "learning_rate": 9.203917470382465e-05, "loss": 0.0762, "step": 22300 }, { "grad_norm": 0.5343247652053833, "learning_rate": 9.203022099073309e-05, "loss": 0.0608, "step": 22310 }, { "grad_norm": 0.44974371790885925, "learning_rate": 9.202126268127976e-05, "loss": 0.0679, "step": 22320 }, { "grad_norm": 0.624700665473938, "learning_rate": 9.20122997764443e-05, "loss": 0.0751, "step": 22330 }, { "grad_norm": 0.7540640830993652, "learning_rate": 9.200333227720692e-05, "loss": 0.0722, "step": 22340 }, { "grad_norm": 0.4972403049468994, "learning_rate": 9.199436018454826e-05, "loss": 0.0694, "step": 22350 }, { "grad_norm": 0.4796563684940338, "learning_rate": 9.198538349944952e-05, "loss": 0.0609, "step": 22360 }, { "grad_norm": 0.5308669209480286, "learning_rate": 9.197640222289234e-05, "loss": 0.0863, "step": 22370 }, { "grad_norm": 0.6976234316825867, "learning_rate": 9.196741635585895e-05, "loss": 0.0681, "step": 22380 }, { "grad_norm": 0.7886348366737366, "learning_rate": 9.195842589933199e-05, "loss": 0.067, "step": 22390 }, { "grad_norm": 0.5537570118904114, "learning_rate": 9.194943085429466e-05, "loss": 0.0712, "step": 22400 }, { "grad_norm": 0.5068917274475098, "learning_rate": 9.194043122173065e-05, "loss": 0.086, "step": 22410 }, { "grad_norm": 0.5093137621879578, "learning_rate": 9.193142700262413e-05, "loss": 0.0725, "step": 22420 }, { "grad_norm": 0.7309780120849609, "learning_rate": 9.192241819795979e-05, "loss": 0.0596, "step": 22430 }, { "grad_norm": 0.6875011920928955, "learning_rate": 9.191340480872284e-05, "loss": 0.0719, "step": 22440 }, { "grad_norm": 0.5650320053100586, "learning_rate": 9.190438683589895e-05, "loss": 0.0804, "step": 22450 }, { "grad_norm": 0.5595531463623047, "learning_rate": 9.189536428047432e-05, "loss": 0.0754, "step": 22460 }, { "grad_norm": 0.5130255222320557, "learning_rate": 9.188633714343564e-05, "loss": 0.0627, "step": 22470 }, { "grad_norm": 0.8665798902511597, "learning_rate": 9.18773054257701e-05, "loss": 0.0803, "step": 22480 }, { "grad_norm": 0.6371219754219055, "learning_rate": 9.18682691284654e-05, "loss": 0.0728, "step": 22490 }, { "grad_norm": 0.6416741013526917, "learning_rate": 9.185922825250974e-05, "loss": 0.0775, "step": 22500 }, { "grad_norm": 0.5031476616859436, "learning_rate": 9.185018279889181e-05, "loss": 0.0546, "step": 22510 }, { "grad_norm": 0.6732725501060486, "learning_rate": 9.184113276860082e-05, "loss": 0.0625, "step": 22520 }, { "grad_norm": 0.6155298948287964, "learning_rate": 9.183207816262645e-05, "loss": 0.0704, "step": 22530 }, { "grad_norm": 0.7559081315994263, "learning_rate": 9.182301898195891e-05, "loss": 0.0682, "step": 22540 }, { "grad_norm": 0.5433754324913025, "learning_rate": 9.181395522758889e-05, "loss": 0.085, "step": 22550 }, { "grad_norm": 0.6410033702850342, "learning_rate": 9.180488690050759e-05, "loss": 0.086, "step": 22560 }, { "grad_norm": 0.5966127514839172, "learning_rate": 9.179581400170671e-05, "loss": 0.0708, "step": 22570 }, { "grad_norm": 0.45391714572906494, "learning_rate": 9.178673653217845e-05, "loss": 0.0671, "step": 22580 }, { "grad_norm": 0.6219692826271057, "learning_rate": 9.177765449291551e-05, "loss": 0.0601, "step": 22590 }, { "grad_norm": 0.8179993629455566, "learning_rate": 9.176856788491109e-05, "loss": 0.0686, "step": 22600 }, { "grad_norm": 0.7260900735855103, "learning_rate": 9.175947670915887e-05, "loss": 0.0776, "step": 22610 }, { "grad_norm": 0.5219697952270508, "learning_rate": 9.175038096665309e-05, "loss": 0.0652, "step": 22620 }, { "grad_norm": 0.5248571634292603, "learning_rate": 9.17412806583884e-05, "loss": 0.058, "step": 22630 }, { "grad_norm": 0.7405717372894287, "learning_rate": 9.173217578536002e-05, "loss": 0.0814, "step": 22640 }, { "grad_norm": 0.6462997198104858, "learning_rate": 9.172306634856362e-05, "loss": 0.0668, "step": 22650 }, { "grad_norm": 0.5727830529212952, "learning_rate": 9.171395234899545e-05, "loss": 0.0825, "step": 22660 }, { "grad_norm": 0.41269150376319885, "learning_rate": 9.170483378765214e-05, "loss": 0.0636, "step": 22670 }, { "grad_norm": 0.5350953340530396, "learning_rate": 9.169571066553091e-05, "loss": 0.0702, "step": 22680 }, { "grad_norm": 0.4182334542274475, "learning_rate": 9.168658298362946e-05, "loss": 0.0692, "step": 22690 }, { "grad_norm": 0.4760415256023407, "learning_rate": 9.167745074294598e-05, "loss": 0.0737, "step": 22700 }, { "grad_norm": 0.5685505867004395, "learning_rate": 9.166831394447913e-05, "loss": 0.07, "step": 22710 }, { "grad_norm": 0.4229269027709961, "learning_rate": 9.165917258922812e-05, "loss": 0.0528, "step": 22720 }, { "grad_norm": 0.5574367642402649, "learning_rate": 9.165002667819262e-05, "loss": 0.0705, "step": 22730 }, { "grad_norm": 0.5810611844062805, "learning_rate": 9.164087621237282e-05, "loss": 0.059, "step": 22740 }, { "grad_norm": 0.6138317584991455, "learning_rate": 9.163172119276942e-05, "loss": 0.0739, "step": 22750 }, { "grad_norm": 0.5871634483337402, "learning_rate": 9.162256162038358e-05, "loss": 0.078, "step": 22760 }, { "grad_norm": 0.5677462220191956, "learning_rate": 9.161339749621698e-05, "loss": 0.0702, "step": 22770 }, { "grad_norm": 0.4729330241680145, "learning_rate": 9.160422882127177e-05, "loss": 0.0735, "step": 22780 }, { "grad_norm": 0.5561639070510864, "learning_rate": 9.159505559655069e-05, "loss": 0.0674, "step": 22790 }, { "grad_norm": 0.6210079193115234, "learning_rate": 9.158587782305684e-05, "loss": 0.0646, "step": 22800 }, { "grad_norm": 0.5874868631362915, "learning_rate": 9.157669550179391e-05, "loss": 0.077, "step": 22810 }, { "grad_norm": 0.573552668094635, "learning_rate": 9.156750863376609e-05, "loss": 0.0779, "step": 22820 }, { "grad_norm": 0.45820730924606323, "learning_rate": 9.155831721997801e-05, "loss": 0.0756, "step": 22830 }, { "grad_norm": 0.5132046937942505, "learning_rate": 9.154912126143484e-05, "loss": 0.0646, "step": 22840 }, { "grad_norm": 0.8800128698348999, "learning_rate": 9.153992075914224e-05, "loss": 0.0832, "step": 22850 }, { "grad_norm": 0.6940479278564453, "learning_rate": 9.153071571410635e-05, "loss": 0.0686, "step": 22860 }, { "grad_norm": 0.4512757956981659, "learning_rate": 9.152150612733384e-05, "loss": 0.0548, "step": 22870 }, { "grad_norm": 0.48576948046684265, "learning_rate": 9.151229199983184e-05, "loss": 0.0631, "step": 22880 }, { "grad_norm": 0.554044246673584, "learning_rate": 9.150307333260802e-05, "loss": 0.0613, "step": 22890 }, { "grad_norm": 0.5560354590415955, "learning_rate": 9.149385012667048e-05, "loss": 0.0745, "step": 22900 }, { "grad_norm": 0.987634539604187, "learning_rate": 9.148462238302788e-05, "loss": 0.0828, "step": 22910 }, { "grad_norm": 0.6343149542808533, "learning_rate": 9.147539010268936e-05, "loss": 0.0735, "step": 22920 }, { "grad_norm": 0.6666094660758972, "learning_rate": 9.14661532866645e-05, "loss": 0.0681, "step": 22930 }, { "grad_norm": 0.5228623747825623, "learning_rate": 9.145691193596348e-05, "loss": 0.0584, "step": 22940 }, { "grad_norm": 0.7657949924468994, "learning_rate": 9.144766605159691e-05, "loss": 0.0849, "step": 22950 }, { "grad_norm": 0.5533893704414368, "learning_rate": 9.14384156345759e-05, "loss": 0.0676, "step": 22960 }, { "grad_norm": 0.7620503902435303, "learning_rate": 9.142916068591204e-05, "loss": 0.0694, "step": 22970 }, { "grad_norm": 0.562843382358551, "learning_rate": 9.141990120661746e-05, "loss": 0.0665, "step": 22980 }, { "grad_norm": 0.4516483545303345, "learning_rate": 9.141063719770475e-05, "loss": 0.0643, "step": 22990 }, { "grad_norm": 0.6694170832633972, "learning_rate": 9.140136866018704e-05, "loss": 0.0688, "step": 23000 }, { "grad_norm": 0.5266687273979187, "learning_rate": 9.139209559507788e-05, "loss": 0.0547, "step": 23010 }, { "grad_norm": 0.5720780491828918, "learning_rate": 9.13828180033914e-05, "loss": 0.0694, "step": 23020 }, { "grad_norm": 0.6919931173324585, "learning_rate": 9.137353588614212e-05, "loss": 0.0719, "step": 23030 }, { "grad_norm": 0.40734240412712097, "learning_rate": 9.136424924434519e-05, "loss": 0.0655, "step": 23040 }, { "grad_norm": 1.5057355165481567, "learning_rate": 9.135495807901615e-05, "loss": 0.0745, "step": 23050 }, { "grad_norm": 0.42459458112716675, "learning_rate": 9.134566239117108e-05, "loss": 0.0724, "step": 23060 }, { "grad_norm": 0.5481026768684387, "learning_rate": 9.13363621818265e-05, "loss": 0.052, "step": 23070 }, { "grad_norm": 0.5222011208534241, "learning_rate": 9.132705745199953e-05, "loss": 0.0724, "step": 23080 }, { "grad_norm": 0.5846700072288513, "learning_rate": 9.131774820270768e-05, "loss": 0.0596, "step": 23090 }, { "grad_norm": 0.5168355703353882, "learning_rate": 9.130843443496901e-05, "loss": 0.0815, "step": 23100 }, { "grad_norm": 0.43572482466697693, "learning_rate": 9.129911614980206e-05, "loss": 0.0768, "step": 23110 }, { "grad_norm": 0.4996630847454071, "learning_rate": 9.128979334822584e-05, "loss": 0.0604, "step": 23120 }, { "grad_norm": 0.6192236542701721, "learning_rate": 9.128046603125992e-05, "loss": 0.0611, "step": 23130 }, { "grad_norm": 0.515626847743988, "learning_rate": 9.12711341999243e-05, "loss": 0.0554, "step": 23140 }, { "grad_norm": 0.5002483129501343, "learning_rate": 9.12617978552395e-05, "loss": 0.0873, "step": 23150 }, { "grad_norm": 0.541104257106781, "learning_rate": 9.12524569982265e-05, "loss": 0.0598, "step": 23160 }, { "grad_norm": 0.7115454077720642, "learning_rate": 9.124311162990684e-05, "loss": 0.0759, "step": 23170 }, { "grad_norm": 0.6067311763763428, "learning_rate": 9.12337617513025e-05, "loss": 0.0657, "step": 23180 }, { "grad_norm": 0.6320176720619202, "learning_rate": 9.122440736343596e-05, "loss": 0.0593, "step": 23190 }, { "grad_norm": 0.43723204731941223, "learning_rate": 9.12150484673302e-05, "loss": 0.0615, "step": 23200 }, { "grad_norm": 0.5632683634757996, "learning_rate": 9.120568506400873e-05, "loss": 0.0646, "step": 23210 }, { "grad_norm": 0.4072188138961792, "learning_rate": 9.119631715449548e-05, "loss": 0.0565, "step": 23220 }, { "grad_norm": 0.538736879825592, "learning_rate": 9.118694473981493e-05, "loss": 0.0614, "step": 23230 }, { "grad_norm": 0.7839851379394531, "learning_rate": 9.117756782099203e-05, "loss": 0.0746, "step": 23240 }, { "grad_norm": 1.6262575387954712, "learning_rate": 9.11681863990522e-05, "loss": 0.0727, "step": 23250 }, { "grad_norm": 0.6439855694770813, "learning_rate": 9.115880047502142e-05, "loss": 0.0598, "step": 23260 }, { "grad_norm": 0.8968387246131897, "learning_rate": 9.114941004992609e-05, "loss": 0.0797, "step": 23270 }, { "grad_norm": 0.5010428428649902, "learning_rate": 9.114001512479317e-05, "loss": 0.067, "step": 23280 }, { "grad_norm": 0.6193252801895142, "learning_rate": 9.113061570065003e-05, "loss": 0.0799, "step": 23290 }, { "grad_norm": 0.49044930934906006, "learning_rate": 9.112121177852459e-05, "loss": 0.0706, "step": 23300 }, { "grad_norm": 0.48737698793411255, "learning_rate": 9.111180335944527e-05, "loss": 0.0733, "step": 23310 }, { "grad_norm": 0.5778478384017944, "learning_rate": 9.110239044444093e-05, "loss": 0.0735, "step": 23320 }, { "grad_norm": 0.6039440035820007, "learning_rate": 9.109297303454099e-05, "loss": 0.0724, "step": 23330 }, { "grad_norm": 0.5192195177078247, "learning_rate": 9.108355113077526e-05, "loss": 0.0672, "step": 23340 }, { "grad_norm": 0.37270233035087585, "learning_rate": 9.107412473417419e-05, "loss": 0.0665, "step": 23350 }, { "grad_norm": 0.584049642086029, "learning_rate": 9.106469384576858e-05, "loss": 0.0679, "step": 23360 }, { "grad_norm": 2.287743330001831, "learning_rate": 9.105525846658978e-05, "loss": 0.0802, "step": 23370 }, { "grad_norm": 0.647791862487793, "learning_rate": 9.104581859766965e-05, "loss": 0.0863, "step": 23380 }, { "grad_norm": 0.5861358642578125, "learning_rate": 9.10363742400405e-05, "loss": 0.0783, "step": 23390 }, { "grad_norm": 0.984393835067749, "learning_rate": 9.102692539473518e-05, "loss": 0.0722, "step": 23400 }, { "grad_norm": 0.607352614402771, "learning_rate": 9.101747206278697e-05, "loss": 0.0797, "step": 23410 }, { "grad_norm": 0.6537714004516602, "learning_rate": 9.100801424522968e-05, "loss": 0.0723, "step": 23420 }, { "grad_norm": 0.42680373787879944, "learning_rate": 9.099855194309762e-05, "loss": 0.064, "step": 23430 }, { "grad_norm": 0.5044171810150146, "learning_rate": 9.098908515742554e-05, "loss": 0.0566, "step": 23440 }, { "grad_norm": 0.6212683916091919, "learning_rate": 9.097961388924873e-05, "loss": 0.0681, "step": 23450 }, { "grad_norm": 0.521245539188385, "learning_rate": 9.097013813960298e-05, "loss": 0.0532, "step": 23460 }, { "grad_norm": 1.2295583486557007, "learning_rate": 9.09606579095245e-05, "loss": 0.0656, "step": 23470 }, { "grad_norm": 0.5663278698921204, "learning_rate": 9.095117320005008e-05, "loss": 0.068, "step": 23480 }, { "grad_norm": 0.5937829613685608, "learning_rate": 9.094168401221691e-05, "loss": 0.0691, "step": 23490 }, { "grad_norm": 0.3752264380455017, "learning_rate": 9.093219034706273e-05, "loss": 0.0742, "step": 23500 }, { "grad_norm": 0.4606132209300995, "learning_rate": 9.092269220562577e-05, "loss": 0.0742, "step": 23510 }, { "grad_norm": 0.6044453978538513, "learning_rate": 9.09131895889447e-05, "loss": 0.0927, "step": 23520 }, { "grad_norm": 0.4018661677837372, "learning_rate": 9.090368249805873e-05, "loss": 0.0586, "step": 23530 }, { "grad_norm": 0.45586931705474854, "learning_rate": 9.089417093400754e-05, "loss": 0.0678, "step": 23540 }, { "grad_norm": 0.6118485331535339, "learning_rate": 9.088465489783131e-05, "loss": 0.0594, "step": 23550 }, { "grad_norm": 0.5826155543327332, "learning_rate": 9.087513439057068e-05, "loss": 0.0711, "step": 23560 }, { "grad_norm": 0.48470908403396606, "learning_rate": 9.08656094132668e-05, "loss": 0.0702, "step": 23570 }, { "grad_norm": 0.706407904624939, "learning_rate": 9.085607996696134e-05, "loss": 0.0867, "step": 23580 }, { "grad_norm": 0.5422961711883545, "learning_rate": 9.084654605269639e-05, "loss": 0.0768, "step": 23590 }, { "grad_norm": 0.530971884727478, "learning_rate": 9.083700767151457e-05, "loss": 0.0666, "step": 23600 }, { "grad_norm": 0.5694748759269714, "learning_rate": 9.082746482445898e-05, "loss": 0.0592, "step": 23610 }, { "grad_norm": 0.5334672927856445, "learning_rate": 9.081791751257325e-05, "loss": 0.0639, "step": 23620 }, { "grad_norm": 0.4208811819553375, "learning_rate": 9.080836573690142e-05, "loss": 0.0679, "step": 23630 }, { "grad_norm": 0.7098093628883362, "learning_rate": 9.079880949848805e-05, "loss": 0.084, "step": 23640 }, { "grad_norm": 0.682458221912384, "learning_rate": 9.078924879837822e-05, "loss": 0.0616, "step": 23650 }, { "grad_norm": 0.5095024704933167, "learning_rate": 9.077968363761747e-05, "loss": 0.0653, "step": 23660 }, { "grad_norm": 0.6446436643600464, "learning_rate": 9.077011401725182e-05, "loss": 0.0625, "step": 23670 }, { "grad_norm": 0.4679698944091797, "learning_rate": 9.07605399383278e-05, "loss": 0.0682, "step": 23680 }, { "grad_norm": 0.5505797863006592, "learning_rate": 9.075096140189243e-05, "loss": 0.0561, "step": 23690 }, { "grad_norm": 0.5123074650764465, "learning_rate": 9.074137840899318e-05, "loss": 0.0623, "step": 23700 }, { "grad_norm": 0.8811525106430054, "learning_rate": 9.073179096067804e-05, "loss": 0.0684, "step": 23710 }, { "grad_norm": 0.49224141240119934, "learning_rate": 9.072219905799549e-05, "loss": 0.065, "step": 23720 }, { "grad_norm": 0.6334220170974731, "learning_rate": 9.071260270199447e-05, "loss": 0.0703, "step": 23730 }, { "grad_norm": 0.39639604091644287, "learning_rate": 9.070300189372441e-05, "loss": 0.0617, "step": 23740 }, { "grad_norm": 0.5805243849754333, "learning_rate": 9.069339663423528e-05, "loss": 0.0584, "step": 23750 }, { "grad_norm": 0.5082097053527832, "learning_rate": 9.068378692457747e-05, "loss": 0.0663, "step": 23760 }, { "grad_norm": 0.4999041259288788, "learning_rate": 9.067417276580189e-05, "loss": 0.0611, "step": 23770 }, { "grad_norm": 0.6735391616821289, "learning_rate": 9.066455415895993e-05, "loss": 0.0591, "step": 23780 }, { "grad_norm": 0.5220180153846741, "learning_rate": 9.065493110510346e-05, "loss": 0.0639, "step": 23790 }, { "grad_norm": 0.6822410225868225, "learning_rate": 9.064530360528484e-05, "loss": 0.0602, "step": 23800 }, { "grad_norm": 0.962942898273468, "learning_rate": 9.063567166055695e-05, "loss": 0.0587, "step": 23810 }, { "grad_norm": 0.793026328086853, "learning_rate": 9.062603527197308e-05, "loss": 0.0727, "step": 23820 }, { "grad_norm": 0.45561566948890686, "learning_rate": 9.06163944405871e-05, "loss": 0.0763, "step": 23830 }, { "grad_norm": 0.9076520204544067, "learning_rate": 9.060674916745327e-05, "loss": 0.086, "step": 23840 }, { "grad_norm": 0.4287813901901245, "learning_rate": 9.05970994536264e-05, "loss": 0.0638, "step": 23850 }, { "grad_norm": 0.5171666145324707, "learning_rate": 9.05874453001618e-05, "loss": 0.059, "step": 23860 }, { "grad_norm": 0.7173047065734863, "learning_rate": 9.057778670811517e-05, "loss": 0.0704, "step": 23870 }, { "grad_norm": 0.6189417243003845, "learning_rate": 9.056812367854281e-05, "loss": 0.0587, "step": 23880 }, { "grad_norm": 0.49081993103027344, "learning_rate": 9.055845621250143e-05, "loss": 0.0614, "step": 23890 }, { "grad_norm": 0.5924930572509766, "learning_rate": 9.054878431104825e-05, "loss": 0.0697, "step": 23900 }, { "grad_norm": 0.6625182032585144, "learning_rate": 9.0539107975241e-05, "loss": 0.0807, "step": 23910 }, { "grad_norm": 0.5872881412506104, "learning_rate": 9.052942720613784e-05, "loss": 0.065, "step": 23920 }, { "grad_norm": 0.4461933374404907, "learning_rate": 9.051974200479745e-05, "loss": 0.0595, "step": 23930 }, { "grad_norm": 0.5917078852653503, "learning_rate": 9.051005237227901e-05, "loss": 0.0522, "step": 23940 }, { "grad_norm": 0.6249026656150818, "learning_rate": 9.050035830964215e-05, "loss": 0.062, "step": 23950 }, { "grad_norm": 0.5906643271446228, "learning_rate": 9.049065981794698e-05, "loss": 0.0814, "step": 23960 }, { "grad_norm": 0.5974211692810059, "learning_rate": 9.048095689825414e-05, "loss": 0.0632, "step": 23970 }, { "grad_norm": 0.4950424134731293, "learning_rate": 9.047124955162472e-05, "loss": 0.0583, "step": 23980 }, { "grad_norm": 0.4944208562374115, "learning_rate": 9.046153777912028e-05, "loss": 0.0638, "step": 23990 }, { "grad_norm": 0.7035951614379883, "learning_rate": 9.045182158180292e-05, "loss": 0.0589, "step": 24000 }, { "grad_norm": 0.4928056597709656, "learning_rate": 9.044210096073516e-05, "loss": 0.0513, "step": 24010 }, { "grad_norm": 0.44660016894340515, "learning_rate": 9.043237591698004e-05, "loss": 0.0526, "step": 24020 }, { "grad_norm": 0.7708948850631714, "learning_rate": 9.04226464516011e-05, "loss": 0.0682, "step": 24030 }, { "grad_norm": 0.620282769203186, "learning_rate": 9.041291256566229e-05, "loss": 0.0815, "step": 24040 }, { "grad_norm": 0.7534679770469666, "learning_rate": 9.040317426022814e-05, "loss": 0.0657, "step": 24050 }, { "grad_norm": 0.6449127197265625, "learning_rate": 9.03934315363636e-05, "loss": 0.0834, "step": 24060 }, { "grad_norm": 0.5581997036933899, "learning_rate": 9.038368439513409e-05, "loss": 0.0697, "step": 24070 }, { "grad_norm": 0.5815795063972473, "learning_rate": 9.03739328376056e-05, "loss": 0.0773, "step": 24080 }, { "grad_norm": 0.6127103567123413, "learning_rate": 9.036417686484451e-05, "loss": 0.0622, "step": 24090 }, { "grad_norm": 0.7814356684684753, "learning_rate": 9.035441647791773e-05, "loss": 0.0663, "step": 24100 }, { "grad_norm": 0.8943268060684204, "learning_rate": 9.034465167789263e-05, "loss": 0.0625, "step": 24110 }, { "grad_norm": 0.9184868931770325, "learning_rate": 9.033488246583706e-05, "loss": 0.0673, "step": 24120 }, { "grad_norm": 0.4254569411277771, "learning_rate": 9.032510884281941e-05, "loss": 0.0793, "step": 24130 }, { "grad_norm": 0.47275611758232117, "learning_rate": 9.031533080990848e-05, "loss": 0.0634, "step": 24140 }, { "grad_norm": 0.7538727521896362, "learning_rate": 9.030554836817358e-05, "loss": 0.0733, "step": 24150 }, { "grad_norm": 0.5429577231407166, "learning_rate": 9.029576151868451e-05, "loss": 0.0709, "step": 24160 }, { "grad_norm": 0.6102920174598694, "learning_rate": 9.028597026251155e-05, "loss": 0.0541, "step": 24170 }, { "grad_norm": 0.5679906606674194, "learning_rate": 9.027617460072547e-05, "loss": 0.0636, "step": 24180 }, { "grad_norm": 0.5817798972129822, "learning_rate": 9.026637453439745e-05, "loss": 0.0686, "step": 24190 }, { "grad_norm": 0.47499293088912964, "learning_rate": 9.025657006459927e-05, "loss": 0.0655, "step": 24200 }, { "grad_norm": 0.938252866268158, "learning_rate": 9.024676119240311e-05, "loss": 0.0651, "step": 24210 }, { "grad_norm": 0.5700410604476929, "learning_rate": 9.023694791888166e-05, "loss": 0.0648, "step": 24220 }, { "grad_norm": 0.4698275923728943, "learning_rate": 9.022713024510808e-05, "loss": 0.0763, "step": 24230 }, { "grad_norm": 0.6722338199615479, "learning_rate": 9.021730817215601e-05, "loss": 0.0593, "step": 24240 }, { "grad_norm": 0.6486609578132629, "learning_rate": 9.02074817010996e-05, "loss": 0.0779, "step": 24250 }, { "grad_norm": 0.497184157371521, "learning_rate": 9.019765083301342e-05, "loss": 0.0606, "step": 24260 }, { "grad_norm": 1.1413871049880981, "learning_rate": 9.01878155689726e-05, "loss": 0.0754, "step": 24270 }, { "grad_norm": 0.6220731735229492, "learning_rate": 9.017797591005268e-05, "loss": 0.0559, "step": 24280 }, { "grad_norm": 0.5254526138305664, "learning_rate": 9.016813185732972e-05, "loss": 0.0799, "step": 24290 }, { "grad_norm": 0.47200196981430054, "learning_rate": 9.015828341188027e-05, "loss": 0.078, "step": 24300 }, { "grad_norm": 0.7732537388801575, "learning_rate": 9.01484305747813e-05, "loss": 0.0662, "step": 24310 }, { "grad_norm": 0.6574952006340027, "learning_rate": 9.013857334711033e-05, "loss": 0.0861, "step": 24320 }, { "grad_norm": 0.548486053943634, "learning_rate": 9.012871172994534e-05, "loss": 0.0627, "step": 24330 }, { "grad_norm": 0.9049574732780457, "learning_rate": 9.011884572436476e-05, "loss": 0.0721, "step": 24340 }, { "grad_norm": 0.6941150426864624, "learning_rate": 9.010897533144754e-05, "loss": 0.0734, "step": 24350 }, { "grad_norm": 0.7117592096328735, "learning_rate": 9.009910055227306e-05, "loss": 0.0626, "step": 24360 }, { "grad_norm": 0.4992752969264984, "learning_rate": 9.008922138792124e-05, "loss": 0.0592, "step": 24370 }, { "grad_norm": 0.4654698967933655, "learning_rate": 9.007933783947244e-05, "loss": 0.0608, "step": 24380 }, { "grad_norm": 0.5429932475090027, "learning_rate": 9.006944990800752e-05, "loss": 0.0527, "step": 24390 }, { "grad_norm": 0.7354249358177185, "learning_rate": 9.005955759460779e-05, "loss": 0.0704, "step": 24400 }, { "grad_norm": 0.6060284972190857, "learning_rate": 9.004966090035508e-05, "loss": 0.0527, "step": 24410 }, { "grad_norm": 0.49298736453056335, "learning_rate": 9.003975982633166e-05, "loss": 0.0609, "step": 24420 }, { "grad_norm": 0.5926549434661865, "learning_rate": 9.00298543736203e-05, "loss": 0.0736, "step": 24430 }, { "grad_norm": 0.4528587758541107, "learning_rate": 9.001994454330427e-05, "loss": 0.0539, "step": 24440 }, { "grad_norm": 0.5784984230995178, "learning_rate": 9.001003033646727e-05, "loss": 0.0644, "step": 24450 }, { "grad_norm": 0.495909720659256, "learning_rate": 9.00001117541935e-05, "loss": 0.0578, "step": 24460 }, { "grad_norm": 0.5824598073959351, "learning_rate": 8.999018879756764e-05, "loss": 0.065, "step": 24470 }, { "grad_norm": 0.5632248520851135, "learning_rate": 8.998026146767487e-05, "loss": 0.0495, "step": 24480 }, { "grad_norm": 0.7231829762458801, "learning_rate": 8.99703297656008e-05, "loss": 0.0545, "step": 24490 }, { "grad_norm": 0.6498493552207947, "learning_rate": 8.996039369243156e-05, "loss": 0.0653, "step": 24500 }, { "grad_norm": 0.6988312005996704, "learning_rate": 8.995045324925378e-05, "loss": 0.0597, "step": 24510 }, { "grad_norm": 0.5157378911972046, "learning_rate": 8.994050843715448e-05, "loss": 0.075, "step": 24520 }, { "grad_norm": 0.42521432042121887, "learning_rate": 8.993055925722121e-05, "loss": 0.0636, "step": 24530 }, { "grad_norm": 0.5337198972702026, "learning_rate": 8.992060571054202e-05, "loss": 0.0681, "step": 24540 }, { "grad_norm": 0.47972744703292847, "learning_rate": 8.991064779820542e-05, "loss": 0.0623, "step": 24550 }, { "grad_norm": 0.5908670425415039, "learning_rate": 8.990068552130036e-05, "loss": 0.0572, "step": 24560 }, { "grad_norm": 0.4161723852157593, "learning_rate": 8.989071888091634e-05, "loss": 0.0621, "step": 24570 }, { "grad_norm": 0.5579099655151367, "learning_rate": 8.988074787814329e-05, "loss": 0.0595, "step": 24580 }, { "grad_norm": 0.7652625441551208, "learning_rate": 8.987077251407158e-05, "loss": 0.0568, "step": 24590 }, { "grad_norm": 1.2831089496612549, "learning_rate": 8.986079278979216e-05, "loss": 0.0682, "step": 24600 }, { "grad_norm": 0.6155043840408325, "learning_rate": 8.985080870639635e-05, "loss": 0.0855, "step": 24610 }, { "grad_norm": 0.8295047283172607, "learning_rate": 8.984082026497603e-05, "loss": 0.0611, "step": 24620 }, { "grad_norm": 0.6055135130882263, "learning_rate": 8.98308274666235e-05, "loss": 0.0787, "step": 24630 }, { "grad_norm": 0.5855976939201355, "learning_rate": 8.982083031243155e-05, "loss": 0.0696, "step": 24640 }, { "grad_norm": 0.8496087789535522, "learning_rate": 8.98108288034935e-05, "loss": 0.0618, "step": 24650 }, { "grad_norm": 0.6282922625541687, "learning_rate": 8.980082294090305e-05, "loss": 0.0647, "step": 24660 }, { "grad_norm": 1.3391228914260864, "learning_rate": 8.979081272575443e-05, "loss": 0.0709, "step": 24670 }, { "grad_norm": 0.7338660359382629, "learning_rate": 8.978079815914236e-05, "loss": 0.0701, "step": 24680 }, { "grad_norm": 0.5832288861274719, "learning_rate": 8.977077924216202e-05, "loss": 0.0754, "step": 24690 }, { "grad_norm": 0.8121653199195862, "learning_rate": 8.976075597590905e-05, "loss": 0.0846, "step": 24700 }, { "grad_norm": 0.4366995394229889, "learning_rate": 8.975072836147958e-05, "loss": 0.0504, "step": 24710 }, { "grad_norm": 0.6913661360740662, "learning_rate": 8.974069639997025e-05, "loss": 0.0663, "step": 24720 }, { "grad_norm": 0.4153931736946106, "learning_rate": 8.973066009247808e-05, "loss": 0.0693, "step": 24730 }, { "grad_norm": 0.33722686767578125, "learning_rate": 8.972061944010066e-05, "loss": 0.0592, "step": 24740 }, { "grad_norm": 0.48069894313812256, "learning_rate": 8.971057444393603e-05, "loss": 0.0617, "step": 24750 }, { "grad_norm": 0.6152902841567993, "learning_rate": 8.970052510508268e-05, "loss": 0.0635, "step": 24760 }, { "grad_norm": 0.5184821486473083, "learning_rate": 8.969047142463959e-05, "loss": 0.072, "step": 24770 }, { "grad_norm": 0.555732786655426, "learning_rate": 8.968041340370621e-05, "loss": 0.0632, "step": 24780 }, { "grad_norm": 0.3612757921218872, "learning_rate": 8.96703510433825e-05, "loss": 0.0663, "step": 24790 }, { "grad_norm": 0.4845631718635559, "learning_rate": 8.966028434476883e-05, "loss": 0.0571, "step": 24800 }, { "grad_norm": 0.6166475415229797, "learning_rate": 8.96502133089661e-05, "loss": 0.0611, "step": 24810 }, { "grad_norm": 0.5792697668075562, "learning_rate": 8.964013793707564e-05, "loss": 0.0561, "step": 24820 }, { "grad_norm": 0.5572596788406372, "learning_rate": 8.963005823019932e-05, "loss": 0.0746, "step": 24830 }, { "grad_norm": 0.5203433632850647, "learning_rate": 8.961997418943939e-05, "loss": 0.0575, "step": 24840 }, { "grad_norm": 0.6624910235404968, "learning_rate": 8.960988581589865e-05, "loss": 0.0668, "step": 24850 }, { "grad_norm": 0.6231163144111633, "learning_rate": 8.959979311068037e-05, "loss": 0.0578, "step": 24860 }, { "grad_norm": 0.5472943782806396, "learning_rate": 8.958969607488823e-05, "loss": 0.0754, "step": 24870 }, { "grad_norm": 0.8422539234161377, "learning_rate": 8.957959470962647e-05, "loss": 0.0618, "step": 24880 }, { "grad_norm": 0.9465091824531555, "learning_rate": 8.956948901599971e-05, "loss": 0.059, "step": 24890 }, { "grad_norm": 0.46746277809143066, "learning_rate": 8.955937899511315e-05, "loss": 0.0682, "step": 24900 }, { "grad_norm": 0.6542043089866638, "learning_rate": 8.954926464807238e-05, "loss": 0.0692, "step": 24910 }, { "grad_norm": 0.48984163999557495, "learning_rate": 8.953914597598347e-05, "loss": 0.0661, "step": 24920 }, { "grad_norm": 0.4230862855911255, "learning_rate": 8.952902297995303e-05, "loss": 0.0633, "step": 24930 }, { "grad_norm": 0.35068780183792114, "learning_rate": 8.951889566108804e-05, "loss": 0.0727, "step": 24940 }, { "grad_norm": 0.6093142628669739, "learning_rate": 8.950876402049606e-05, "loss": 0.0706, "step": 24950 }, { "grad_norm": 0.5485137104988098, "learning_rate": 8.949862805928504e-05, "loss": 0.0606, "step": 24960 }, { "grad_norm": 0.49884480237960815, "learning_rate": 8.948848777856343e-05, "loss": 0.0707, "step": 24970 }, { "grad_norm": 0.6215360760688782, "learning_rate": 8.947834317944017e-05, "loss": 0.0611, "step": 24980 }, { "grad_norm": 0.45561763644218445, "learning_rate": 8.946819426302466e-05, "loss": 0.071, "step": 24990 }, { "grad_norm": 0.5819719433784485, "learning_rate": 8.945804103042676e-05, "loss": 0.0598, "step": 25000 }, { "grad_norm": 0.6007626056671143, "learning_rate": 8.944788348275681e-05, "loss": 0.0608, "step": 25010 }, { "grad_norm": 0.5793912410736084, "learning_rate": 8.943772162112565e-05, "loss": 0.0646, "step": 25020 }, { "grad_norm": 0.568797767162323, "learning_rate": 8.942755544664454e-05, "loss": 0.0594, "step": 25030 }, { "grad_norm": 0.8671683073043823, "learning_rate": 8.941738496042525e-05, "loss": 0.0573, "step": 25040 }, { "grad_norm": 0.4955713152885437, "learning_rate": 8.940721016357999e-05, "loss": 0.0738, "step": 25050 }, { "grad_norm": 0.623817503452301, "learning_rate": 8.939703105722148e-05, "loss": 0.0727, "step": 25060 }, { "grad_norm": 0.6677738428115845, "learning_rate": 8.93868476424629e-05, "loss": 0.0554, "step": 25070 }, { "grad_norm": 0.5135695338249207, "learning_rate": 8.937665992041786e-05, "loss": 0.0671, "step": 25080 }, { "grad_norm": 0.6653926372528076, "learning_rate": 8.93664678922005e-05, "loss": 0.0622, "step": 25090 }, { "grad_norm": 0.6364065408706665, "learning_rate": 8.93562715589254e-05, "loss": 0.0561, "step": 25100 }, { "grad_norm": 0.4960215091705322, "learning_rate": 8.934607092170762e-05, "loss": 0.0606, "step": 25110 }, { "grad_norm": 0.49124881625175476, "learning_rate": 8.933586598166266e-05, "loss": 0.0715, "step": 25120 }, { "grad_norm": 0.4874342978000641, "learning_rate": 8.932565673990655e-05, "loss": 0.0714, "step": 25130 }, { "grad_norm": 0.49581286311149597, "learning_rate": 8.931544319755574e-05, "loss": 0.0542, "step": 25140 }, { "grad_norm": 0.4598248600959778, "learning_rate": 8.930522535572718e-05, "loss": 0.0633, "step": 25150 }, { "grad_norm": 0.4437303841114044, "learning_rate": 8.929500321553826e-05, "loss": 0.0486, "step": 25160 }, { "grad_norm": 0.7295082211494446, "learning_rate": 8.928477677810686e-05, "loss": 0.0664, "step": 25170 }, { "grad_norm": 0.5233712196350098, "learning_rate": 8.927454604455137e-05, "loss": 0.06, "step": 25180 }, { "grad_norm": 0.3700486719608307, "learning_rate": 8.926431101599053e-05, "loss": 0.0617, "step": 25190 }, { "grad_norm": 0.9271097779273987, "learning_rate": 8.925407169354369e-05, "loss": 0.0561, "step": 25200 }, { "grad_norm": 0.5363494157791138, "learning_rate": 8.92438280783306e-05, "loss": 0.08, "step": 25210 }, { "grad_norm": 0.5551268458366394, "learning_rate": 8.923358017147146e-05, "loss": 0.0761, "step": 25220 }, { "grad_norm": 0.7740959525108337, "learning_rate": 8.922332797408697e-05, "loss": 0.0718, "step": 25230 }, { "grad_norm": 0.4129914939403534, "learning_rate": 8.921307148729831e-05, "loss": 0.0564, "step": 25240 }, { "grad_norm": 0.500635027885437, "learning_rate": 8.920281071222712e-05, "loss": 0.0524, "step": 25250 }, { "grad_norm": 0.3723832368850708, "learning_rate": 8.919254564999548e-05, "loss": 0.0553, "step": 25260 }, { "grad_norm": 0.6903171539306641, "learning_rate": 8.918227630172598e-05, "loss": 0.0591, "step": 25270 }, { "grad_norm": 0.5211254358291626, "learning_rate": 8.917200266854165e-05, "loss": 0.0614, "step": 25280 }, { "grad_norm": 0.5307185649871826, "learning_rate": 8.9161724751566e-05, "loss": 0.0581, "step": 25290 }, { "grad_norm": 1.1518827676773071, "learning_rate": 8.915144255192302e-05, "loss": 0.0788, "step": 25300 }, { "grad_norm": 0.8283907771110535, "learning_rate": 8.914115607073714e-05, "loss": 0.0753, "step": 25310 }, { "grad_norm": 0.49895578622817993, "learning_rate": 8.913086530913327e-05, "loss": 0.0611, "step": 25320 }, { "grad_norm": 0.5468754768371582, "learning_rate": 8.912057026823681e-05, "loss": 0.074, "step": 25330 }, { "grad_norm": 0.4189812242984772, "learning_rate": 8.91102709491736e-05, "loss": 0.0671, "step": 25340 }, { "grad_norm": 0.6832775473594666, "learning_rate": 8.909996735306996e-05, "loss": 0.0659, "step": 25350 }, { "grad_norm": 0.72087162733078, "learning_rate": 8.908965948105268e-05, "loss": 0.0737, "step": 25360 }, { "grad_norm": 0.8581417798995972, "learning_rate": 8.907934733424901e-05, "loss": 0.0666, "step": 25370 }, { "grad_norm": 0.3874094486236572, "learning_rate": 8.906903091378666e-05, "loss": 0.0704, "step": 25380 }, { "grad_norm": 0.3995526432991028, "learning_rate": 8.905871022079384e-05, "loss": 0.0783, "step": 25390 }, { "grad_norm": 0.43339917063713074, "learning_rate": 8.90483852563992e-05, "loss": 0.07, "step": 25400 }, { "grad_norm": 0.5532582402229309, "learning_rate": 8.903805602173185e-05, "loss": 0.058, "step": 25410 }, { "grad_norm": 0.4686926603317261, "learning_rate": 8.902772251792137e-05, "loss": 0.0543, "step": 25420 }, { "grad_norm": 0.6347182393074036, "learning_rate": 8.901738474609786e-05, "loss": 0.069, "step": 25430 }, { "grad_norm": 0.38604021072387695, "learning_rate": 8.900704270739179e-05, "loss": 0.0734, "step": 25440 }, { "grad_norm": 0.5569580793380737, "learning_rate": 8.89966964029342e-05, "loss": 0.0872, "step": 25450 }, { "grad_norm": 0.6608636379241943, "learning_rate": 8.898634583385652e-05, "loss": 0.0603, "step": 25460 }, { "grad_norm": 0.7415279746055603, "learning_rate": 8.897599100129065e-05, "loss": 0.0671, "step": 25470 }, { "grad_norm": 0.5010066032409668, "learning_rate": 8.896563190636903e-05, "loss": 0.0565, "step": 25480 }, { "grad_norm": 0.8434252142906189, "learning_rate": 8.895526855022448e-05, "loss": 0.0669, "step": 25490 }, { "grad_norm": 0.7713882327079773, "learning_rate": 8.894490093399033e-05, "loss": 0.0707, "step": 25500 }, { "grad_norm": 0.4928089380264282, "learning_rate": 8.893452905880035e-05, "loss": 0.0683, "step": 25510 }, { "grad_norm": 0.6651431322097778, "learning_rate": 8.892415292578883e-05, "loss": 0.0663, "step": 25520 }, { "grad_norm": 0.31174230575561523, "learning_rate": 8.891377253609046e-05, "loss": 0.0596, "step": 25530 }, { "grad_norm": 0.8728663325309753, "learning_rate": 8.890338789084043e-05, "loss": 0.0656, "step": 25540 }, { "grad_norm": 0.7726324200630188, "learning_rate": 8.88929989911744e-05, "loss": 0.0552, "step": 25550 }, { "grad_norm": 0.731640100479126, "learning_rate": 8.888260583822847e-05, "loss": 0.0706, "step": 25560 }, { "grad_norm": 0.4731079936027527, "learning_rate": 8.887220843313921e-05, "loss": 0.0682, "step": 25570 }, { "grad_norm": 0.7215123772621155, "learning_rate": 8.88618067770437e-05, "loss": 0.0617, "step": 25580 }, { "grad_norm": 0.8510400652885437, "learning_rate": 8.885140087107942e-05, "loss": 0.0702, "step": 25590 }, { "grad_norm": 0.5377925634384155, "learning_rate": 8.884099071638436e-05, "loss": 0.0724, "step": 25600 }, { "grad_norm": 0.9723253846168518, "learning_rate": 8.883057631409695e-05, "loss": 0.0738, "step": 25610 }, { "grad_norm": 0.482155978679657, "learning_rate": 8.882015766535608e-05, "loss": 0.0607, "step": 25620 }, { "grad_norm": 0.6677339673042297, "learning_rate": 8.880973477130115e-05, "loss": 0.0601, "step": 25630 }, { "grad_norm": 0.6839582324028015, "learning_rate": 8.879930763307197e-05, "loss": 0.0495, "step": 25640 }, { "grad_norm": 0.5586569905281067, "learning_rate": 8.878887625180884e-05, "loss": 0.0618, "step": 25650 }, { "grad_norm": 0.42123302817344666, "learning_rate": 8.877844062865253e-05, "loss": 0.0437, "step": 25660 }, { "grad_norm": 0.6625422239303589, "learning_rate": 8.876800076474424e-05, "loss": 0.0567, "step": 25670 }, { "grad_norm": 0.5164692997932434, "learning_rate": 8.875755666122568e-05, "loss": 0.0498, "step": 25680 }, { "grad_norm": 0.5141403079032898, "learning_rate": 8.8747108319239e-05, "loss": 0.0687, "step": 25690 }, { "grad_norm": 1.857360601425171, "learning_rate": 8.87366557399268e-05, "loss": 0.057, "step": 25700 }, { "grad_norm": 1.1481943130493164, "learning_rate": 8.872619892443217e-05, "loss": 0.0648, "step": 25710 }, { "grad_norm": 0.5561198592185974, "learning_rate": 8.871573787389865e-05, "loss": 0.0597, "step": 25720 }, { "grad_norm": 0.6660863161087036, "learning_rate": 8.870527258947024e-05, "loss": 0.0791, "step": 25730 }, { "grad_norm": 0.4145081639289856, "learning_rate": 8.869480307229143e-05, "loss": 0.0583, "step": 25740 }, { "grad_norm": 0.5851934552192688, "learning_rate": 8.868432932350712e-05, "loss": 0.0619, "step": 25750 }, { "grad_norm": 0.4945329427719116, "learning_rate": 8.867385134426272e-05, "loss": 0.0648, "step": 25760 }, { "grad_norm": 0.3699434697628021, "learning_rate": 8.866336913570407e-05, "loss": 0.0539, "step": 25770 }, { "grad_norm": 0.35155293345451355, "learning_rate": 8.865288269897751e-05, "loss": 0.0481, "step": 25780 }, { "grad_norm": 0.5751003623008728, "learning_rate": 8.864239203522981e-05, "loss": 0.0558, "step": 25790 }, { "grad_norm": 0.4544244706630707, "learning_rate": 8.863189714560822e-05, "loss": 0.0674, "step": 25800 }, { "grad_norm": 0.7550662755966187, "learning_rate": 8.862139803126043e-05, "loss": 0.0608, "step": 25810 }, { "grad_norm": 1.2905713319778442, "learning_rate": 8.861089469333463e-05, "loss": 0.0668, "step": 25820 }, { "grad_norm": 0.5062763094902039, "learning_rate": 8.860038713297944e-05, "loss": 0.0473, "step": 25830 }, { "grad_norm": 0.5609181523323059, "learning_rate": 8.858987535134394e-05, "loss": 0.0644, "step": 25840 }, { "grad_norm": 0.5833131670951843, "learning_rate": 8.857935934957769e-05, "loss": 0.0701, "step": 25850 }, { "grad_norm": 0.44340595602989197, "learning_rate": 8.856883912883071e-05, "loss": 0.0551, "step": 25860 }, { "grad_norm": 0.4173352122306824, "learning_rate": 8.855831469025346e-05, "loss": 0.0552, "step": 25870 }, { "grad_norm": 0.8748902678489685, "learning_rate": 8.854778603499689e-05, "loss": 0.0856, "step": 25880 }, { "grad_norm": 0.834817111492157, "learning_rate": 8.85372531642124e-05, "loss": 0.0657, "step": 25890 }, { "grad_norm": 0.4992479681968689, "learning_rate": 8.852671607905185e-05, "loss": 0.0531, "step": 25900 }, { "grad_norm": 0.4436313211917877, "learning_rate": 8.851617478066754e-05, "loss": 0.0749, "step": 25910 }, { "grad_norm": 0.6739733219146729, "learning_rate": 8.850562927021227e-05, "loss": 0.0706, "step": 25920 }, { "grad_norm": 0.4819079041481018, "learning_rate": 8.849507954883928e-05, "loss": 0.0591, "step": 25930 }, { "grad_norm": 0.49863478541374207, "learning_rate": 8.848452561770226e-05, "loss": 0.0583, "step": 25940 }, { "grad_norm": 0.4187226891517639, "learning_rate": 8.847396747795538e-05, "loss": 0.0649, "step": 25950 }, { "grad_norm": 0.4822348356246948, "learning_rate": 8.846340513075327e-05, "loss": 0.0542, "step": 25960 }, { "grad_norm": 0.9844125509262085, "learning_rate": 8.845283857725099e-05, "loss": 0.0675, "step": 25970 }, { "grad_norm": 0.4925406873226166, "learning_rate": 8.844226781860409e-05, "loss": 0.0465, "step": 25980 }, { "grad_norm": 0.6199443340301514, "learning_rate": 8.84316928559686e-05, "loss": 0.0627, "step": 25990 }, { "grad_norm": 0.9584850668907166, "learning_rate": 8.842111369050094e-05, "loss": 0.077, "step": 26000 }, { "grad_norm": 0.527177095413208, "learning_rate": 8.841053032335808e-05, "loss": 0.0595, "step": 26010 }, { "grad_norm": 0.4413082003593445, "learning_rate": 8.839994275569735e-05, "loss": 0.0657, "step": 26020 }, { "grad_norm": 1.0008381605148315, "learning_rate": 8.838935098867662e-05, "loss": 0.0658, "step": 26030 }, { "grad_norm": 0.5908750295639038, "learning_rate": 8.837875502345418e-05, "loss": 0.0737, "step": 26040 }, { "grad_norm": 0.4407931864261627, "learning_rate": 8.83681548611888e-05, "loss": 0.0658, "step": 26050 }, { "grad_norm": 0.3858988583087921, "learning_rate": 8.835755050303969e-05, "loss": 0.0542, "step": 26060 }, { "grad_norm": 0.5217605233192444, "learning_rate": 8.834694195016653e-05, "loss": 0.0742, "step": 26070 }, { "grad_norm": 0.655817985534668, "learning_rate": 8.833632920372942e-05, "loss": 0.0828, "step": 26080 }, { "grad_norm": 0.902229905128479, "learning_rate": 8.832571226488903e-05, "loss": 0.0743, "step": 26090 }, { "grad_norm": 0.6051418781280518, "learning_rate": 8.831509113480634e-05, "loss": 0.061, "step": 26100 }, { "grad_norm": 0.5656899213790894, "learning_rate": 8.83044658146429e-05, "loss": 0.0713, "step": 26110 }, { "grad_norm": 0.4893331229686737, "learning_rate": 8.829383630556067e-05, "loss": 0.0623, "step": 26120 }, { "grad_norm": 0.46643123030662537, "learning_rate": 8.828320260872207e-05, "loss": 0.0601, "step": 26130 }, { "grad_norm": 0.607637882232666, "learning_rate": 8.827256472529e-05, "loss": 0.0609, "step": 26140 }, { "grad_norm": 0.4360467195510864, "learning_rate": 8.826192265642778e-05, "loss": 0.0576, "step": 26150 }, { "grad_norm": 0.43585097789764404, "learning_rate": 8.825127640329923e-05, "loss": 0.0552, "step": 26160 }, { "grad_norm": 0.4494618773460388, "learning_rate": 8.824062596706861e-05, "loss": 0.0557, "step": 26170 }, { "grad_norm": 0.6874494552612305, "learning_rate": 8.822997134890062e-05, "loss": 0.073, "step": 26180 }, { "grad_norm": 0.8345195055007935, "learning_rate": 8.821931254996044e-05, "loss": 0.0551, "step": 26190 }, { "grad_norm": 0.5893474221229553, "learning_rate": 8.82086495714137e-05, "loss": 0.0714, "step": 26200 }, { "grad_norm": 0.7177363634109497, "learning_rate": 8.81979824144265e-05, "loss": 0.073, "step": 26210 }, { "grad_norm": 0.5228496789932251, "learning_rate": 8.818731108016536e-05, "loss": 0.0653, "step": 26220 }, { "grad_norm": 0.43409234285354614, "learning_rate": 8.81766355697973e-05, "loss": 0.0569, "step": 26230 }, { "grad_norm": 1.0500004291534424, "learning_rate": 8.816595588448977e-05, "loss": 0.0591, "step": 26240 }, { "grad_norm": 0.7561540007591248, "learning_rate": 8.81552720254107e-05, "loss": 0.0786, "step": 26250 }, { "grad_norm": 0.5235586166381836, "learning_rate": 8.814458399372842e-05, "loss": 0.0728, "step": 26260 }, { "grad_norm": 0.8553668856620789, "learning_rate": 8.813389179061181e-05, "loss": 0.0639, "step": 26270 }, { "grad_norm": 0.5292639136314392, "learning_rate": 8.812319541723012e-05, "loss": 0.0716, "step": 26280 }, { "grad_norm": 0.48056209087371826, "learning_rate": 8.811249487475309e-05, "loss": 0.0687, "step": 26290 }, { "grad_norm": 0.8227010369300842, "learning_rate": 8.810179016435092e-05, "loss": 0.0655, "step": 26300 }, { "grad_norm": 0.8564942479133606, "learning_rate": 8.809108128719428e-05, "loss": 0.0848, "step": 26310 }, { "grad_norm": 0.9405698180198669, "learning_rate": 8.808036824445424e-05, "loss": 0.0764, "step": 26320 }, { "grad_norm": 0.4855426251888275, "learning_rate": 8.806965103730238e-05, "loss": 0.0712, "step": 26330 }, { "grad_norm": 0.49003899097442627, "learning_rate": 8.805892966691074e-05, "loss": 0.0608, "step": 26340 }, { "grad_norm": 0.6461734771728516, "learning_rate": 8.804820413445175e-05, "loss": 0.0651, "step": 26350 }, { "grad_norm": 0.5641816258430481, "learning_rate": 8.803747444109837e-05, "loss": 0.0677, "step": 26360 }, { "grad_norm": 0.6434492468833923, "learning_rate": 8.802674058802399e-05, "loss": 0.0498, "step": 26370 }, { "grad_norm": 0.4834686517715454, "learning_rate": 8.801600257640241e-05, "loss": 0.0576, "step": 26380 }, { "grad_norm": 1.0304443836212158, "learning_rate": 8.800526040740795e-05, "loss": 0.0669, "step": 26390 }, { "grad_norm": 0.6868031024932861, "learning_rate": 8.799451408221535e-05, "loss": 0.0657, "step": 26400 }, { "grad_norm": 0.5416727066040039, "learning_rate": 8.798376360199982e-05, "loss": 0.0488, "step": 26410 }, { "grad_norm": 0.5193591713905334, "learning_rate": 8.797300896793701e-05, "loss": 0.0529, "step": 26420 }, { "grad_norm": 0.6443980932235718, "learning_rate": 8.796225018120302e-05, "loss": 0.0646, "step": 26430 }, { "grad_norm": 0.4804973006248474, "learning_rate": 8.795148724297444e-05, "loss": 0.0585, "step": 26440 }, { "grad_norm": 0.637686014175415, "learning_rate": 8.794072015442825e-05, "loss": 0.0655, "step": 26450 }, { "grad_norm": 0.8004048466682434, "learning_rate": 8.792994891674198e-05, "loss": 0.0588, "step": 26460 }, { "grad_norm": 0.4522973597049713, "learning_rate": 8.79191735310935e-05, "loss": 0.0575, "step": 26470 }, { "grad_norm": 0.5347088575363159, "learning_rate": 8.790839399866122e-05, "loss": 0.0707, "step": 26480 }, { "grad_norm": 0.6541327238082886, "learning_rate": 8.789761032062397e-05, "loss": 0.0614, "step": 26490 }, { "grad_norm": 0.5388222932815552, "learning_rate": 8.788682249816103e-05, "loss": 0.06, "step": 26500 }, { "grad_norm": 0.4603804647922516, "learning_rate": 8.787603053245215e-05, "loss": 0.0723, "step": 26510 }, { "grad_norm": 0.686959445476532, "learning_rate": 8.78652344246775e-05, "loss": 0.0544, "step": 26520 }, { "grad_norm": 0.5342682600021362, "learning_rate": 8.785443417601776e-05, "loss": 0.0526, "step": 26530 }, { "grad_norm": 0.5271671414375305, "learning_rate": 8.784362978765401e-05, "loss": 0.0498, "step": 26540 }, { "grad_norm": 0.7444425225257874, "learning_rate": 8.783282126076779e-05, "loss": 0.0588, "step": 26550 }, { "grad_norm": 0.9874507188796997, "learning_rate": 8.782200859654112e-05, "loss": 0.0621, "step": 26560 }, { "grad_norm": 0.746283233165741, "learning_rate": 8.781119179615646e-05, "loss": 0.0602, "step": 26570 }, { "grad_norm": 0.5413995385169983, "learning_rate": 8.780037086079674e-05, "loss": 0.0604, "step": 26580 }, { "grad_norm": 0.5289431214332581, "learning_rate": 8.778954579164527e-05, "loss": 0.0633, "step": 26590 }, { "grad_norm": 0.5470707416534424, "learning_rate": 8.777871658988588e-05, "loss": 0.0571, "step": 26600 }, { "grad_norm": 0.45097121596336365, "learning_rate": 8.776788325670285e-05, "loss": 0.0611, "step": 26610 }, { "grad_norm": 0.7270565629005432, "learning_rate": 8.775704579328089e-05, "loss": 0.0627, "step": 26620 }, { "grad_norm": 0.9321972131729126, "learning_rate": 8.774620420080517e-05, "loss": 0.0596, "step": 26630 }, { "grad_norm": 0.5058234930038452, "learning_rate": 8.773535848046131e-05, "loss": 0.0542, "step": 26640 }, { "grad_norm": 0.5906752347946167, "learning_rate": 8.772450863343538e-05, "loss": 0.0613, "step": 26650 }, { "grad_norm": 0.5440535545349121, "learning_rate": 8.77136546609139e-05, "loss": 0.071, "step": 26660 }, { "grad_norm": 0.6691405773162842, "learning_rate": 8.770279656408385e-05, "loss": 0.0537, "step": 26670 }, { "grad_norm": 0.5304124355316162, "learning_rate": 8.769193434413265e-05, "loss": 0.0542, "step": 26680 }, { "grad_norm": 0.8154568672180176, "learning_rate": 8.76810680022482e-05, "loss": 0.0674, "step": 26690 }, { "grad_norm": 0.6596104502677917, "learning_rate": 8.767019753961878e-05, "loss": 0.0715, "step": 26700 }, { "grad_norm": 0.6070760488510132, "learning_rate": 8.765932295743321e-05, "loss": 0.0626, "step": 26710 }, { "grad_norm": 0.46112769842147827, "learning_rate": 8.764844425688068e-05, "loss": 0.0622, "step": 26720 }, { "grad_norm": 0.5496779680252075, "learning_rate": 8.763756143915092e-05, "loss": 0.0545, "step": 26730 }, { "grad_norm": 0.42560720443725586, "learning_rate": 8.7626674505434e-05, "loss": 0.054, "step": 26740 }, { "grad_norm": 0.4283939599990845, "learning_rate": 8.761578345692053e-05, "loss": 0.061, "step": 26750 }, { "grad_norm": 0.8769311308860779, "learning_rate": 8.760488829480156e-05, "loss": 0.0578, "step": 26760 }, { "grad_norm": 0.5756747126579285, "learning_rate": 8.759398902026854e-05, "loss": 0.0489, "step": 26770 }, { "grad_norm": 0.6537075042724609, "learning_rate": 8.758308563451339e-05, "loss": 0.0723, "step": 26780 }, { "grad_norm": 0.5949768424034119, "learning_rate": 8.75721781387285e-05, "loss": 0.0668, "step": 26790 }, { "grad_norm": 0.5843562483787537, "learning_rate": 8.75612665341067e-05, "loss": 0.0571, "step": 26800 }, { "grad_norm": 0.5284810066223145, "learning_rate": 8.755035082184126e-05, "loss": 0.0587, "step": 26810 }, { "grad_norm": 0.5501083135604858, "learning_rate": 8.753943100312592e-05, "loss": 0.0673, "step": 26820 }, { "grad_norm": 0.4211288094520569, "learning_rate": 8.752850707915484e-05, "loss": 0.0576, "step": 26830 }, { "grad_norm": 0.5645807385444641, "learning_rate": 8.751757905112264e-05, "loss": 0.0673, "step": 26840 }, { "grad_norm": 0.8043907284736633, "learning_rate": 8.75066469202244e-05, "loss": 0.0617, "step": 26850 }, { "grad_norm": 0.6287931203842163, "learning_rate": 8.749571068765567e-05, "loss": 0.0574, "step": 26860 }, { "grad_norm": 0.5796260833740234, "learning_rate": 8.748477035461238e-05, "loss": 0.0565, "step": 26870 }, { "grad_norm": 0.590823233127594, "learning_rate": 8.747382592229095e-05, "loss": 0.0618, "step": 26880 }, { "grad_norm": 0.43528467416763306, "learning_rate": 8.746287739188828e-05, "loss": 0.0715, "step": 26890 }, { "grad_norm": 0.5123223662376404, "learning_rate": 8.745192476460165e-05, "loss": 0.054, "step": 26900 }, { "grad_norm": 0.5085252523422241, "learning_rate": 8.744096804162882e-05, "loss": 0.0771, "step": 26910 }, { "grad_norm": 0.7344062328338623, "learning_rate": 8.743000722416804e-05, "loss": 0.0631, "step": 26920 }, { "grad_norm": 0.476353257894516, "learning_rate": 8.741904231341793e-05, "loss": 0.0542, "step": 26930 }, { "grad_norm": 0.7821011543273926, "learning_rate": 8.740807331057762e-05, "loss": 0.0665, "step": 26940 }, { "grad_norm": 0.508618175983429, "learning_rate": 8.739710021684667e-05, "loss": 0.0579, "step": 26950 }, { "grad_norm": 0.3995378613471985, "learning_rate": 8.738612303342503e-05, "loss": 0.0593, "step": 26960 }, { "grad_norm": 0.4748784303665161, "learning_rate": 8.73751417615132e-05, "loss": 0.0549, "step": 26970 }, { "grad_norm": 0.44203826785087585, "learning_rate": 8.736415640231208e-05, "loss": 0.0465, "step": 26980 }, { "grad_norm": 0.35493507981300354, "learning_rate": 8.735316695702297e-05, "loss": 0.0507, "step": 26990 }, { "grad_norm": 0.4397208094596863, "learning_rate": 8.734217342684769e-05, "loss": 0.0454, "step": 27000 }, { "grad_norm": 0.6211925148963928, "learning_rate": 8.733117581298847e-05, "loss": 0.0437, "step": 27010 }, { "grad_norm": 0.4349786341190338, "learning_rate": 8.732017411664796e-05, "loss": 0.0602, "step": 27020 }, { "grad_norm": 0.47412344813346863, "learning_rate": 8.730916833902936e-05, "loss": 0.0618, "step": 27030 }, { "grad_norm": 0.7184513807296753, "learning_rate": 8.729815848133618e-05, "loss": 0.0852, "step": 27040 }, { "grad_norm": 0.4333849251270294, "learning_rate": 8.728714454477247e-05, "loss": 0.0497, "step": 27050 }, { "grad_norm": 0.6230959892272949, "learning_rate": 8.727612653054269e-05, "loss": 0.0578, "step": 27060 }, { "grad_norm": 0.3453052341938019, "learning_rate": 8.726510443985176e-05, "loss": 0.0505, "step": 27070 }, { "grad_norm": 0.5466144680976868, "learning_rate": 8.725407827390503e-05, "loss": 0.0694, "step": 27080 }, { "grad_norm": 0.5730535984039307, "learning_rate": 8.724304803390833e-05, "loss": 0.0616, "step": 27090 }, { "grad_norm": 0.48326605558395386, "learning_rate": 8.723201372106788e-05, "loss": 0.0545, "step": 27100 }, { "grad_norm": 0.7617009282112122, "learning_rate": 8.722097533659038e-05, "loss": 0.0566, "step": 27110 }, { "grad_norm": 0.5672455430030823, "learning_rate": 8.720993288168299e-05, "loss": 0.0561, "step": 27120 }, { "grad_norm": 0.6624249815940857, "learning_rate": 8.719888635755327e-05, "loss": 0.0514, "step": 27130 }, { "grad_norm": 0.8266980051994324, "learning_rate": 8.718783576540928e-05, "loss": 0.0708, "step": 27140 }, { "grad_norm": 0.5580664873123169, "learning_rate": 8.717678110645948e-05, "loss": 0.066, "step": 27150 }, { "grad_norm": 0.4726871848106384, "learning_rate": 8.716572238191279e-05, "loss": 0.0635, "step": 27160 }, { "grad_norm": 0.6035423874855042, "learning_rate": 8.715465959297857e-05, "loss": 0.0678, "step": 27170 }, { "grad_norm": 0.7884936332702637, "learning_rate": 8.714359274086665e-05, "loss": 0.0885, "step": 27180 }, { "grad_norm": 0.5345973968505859, "learning_rate": 8.713252182678726e-05, "loss": 0.0628, "step": 27190 }, { "grad_norm": 0.5372947454452515, "learning_rate": 8.712144685195112e-05, "loss": 0.0799, "step": 27200 }, { "grad_norm": 0.4364323914051056, "learning_rate": 8.711036781756936e-05, "loss": 0.0677, "step": 27210 }, { "grad_norm": 0.6386190056800842, "learning_rate": 8.709928472485357e-05, "loss": 0.0661, "step": 27220 }, { "grad_norm": 0.6300086975097656, "learning_rate": 8.708819757501579e-05, "loss": 0.0573, "step": 27230 }, { "grad_norm": 0.44832345843315125, "learning_rate": 8.707710636926846e-05, "loss": 0.0541, "step": 27240 }, { "grad_norm": 0.4660182297229767, "learning_rate": 8.706601110882455e-05, "loss": 0.0728, "step": 27250 }, { "grad_norm": 0.6759119033813477, "learning_rate": 8.705491179489738e-05, "loss": 0.056, "step": 27260 }, { "grad_norm": 0.5850425958633423, "learning_rate": 8.704380842870077e-05, "loss": 0.0572, "step": 27270 }, { "grad_norm": 0.42804741859436035, "learning_rate": 8.703270101144895e-05, "loss": 0.0528, "step": 27280 }, { "grad_norm": 0.37175452709198, "learning_rate": 8.702158954435664e-05, "loss": 0.0547, "step": 27290 }, { "grad_norm": 0.4036599397659302, "learning_rate": 8.701047402863896e-05, "loss": 0.0516, "step": 27300 }, { "grad_norm": 0.9344568848609924, "learning_rate": 8.699935446551148e-05, "loss": 0.0619, "step": 27310 }, { "grad_norm": 0.5328969359397888, "learning_rate": 8.698823085619022e-05, "loss": 0.0656, "step": 27320 }, { "grad_norm": 0.5201311111450195, "learning_rate": 8.697710320189166e-05, "loss": 0.0476, "step": 27330 }, { "grad_norm": 0.6672635674476624, "learning_rate": 8.696597150383268e-05, "loss": 0.0588, "step": 27340 }, { "grad_norm": 2.73811411857605, "learning_rate": 8.695483576323063e-05, "loss": 0.0649, "step": 27350 }, { "grad_norm": 0.7997803092002869, "learning_rate": 8.69436959813033e-05, "loss": 0.0704, "step": 27360 }, { "grad_norm": 0.465445339679718, "learning_rate": 8.693255215926892e-05, "loss": 0.0628, "step": 27370 }, { "grad_norm": 0.48604530096054077, "learning_rate": 8.692140429834617e-05, "loss": 0.0675, "step": 27380 }, { "grad_norm": 0.6718078255653381, "learning_rate": 8.691025239975415e-05, "loss": 0.074, "step": 27390 }, { "grad_norm": 0.5432174205780029, "learning_rate": 8.689909646471243e-05, "loss": 0.0601, "step": 27400 }, { "grad_norm": 1.0208048820495605, "learning_rate": 8.688793649444099e-05, "loss": 0.08, "step": 27410 }, { "grad_norm": 0.6033051013946533, "learning_rate": 8.687677249016029e-05, "loss": 0.0639, "step": 27420 }, { "grad_norm": 0.7149170637130737, "learning_rate": 8.686560445309118e-05, "loss": 0.0628, "step": 27430 }, { "grad_norm": 0.7269973158836365, "learning_rate": 8.685443238445499e-05, "loss": 0.071, "step": 27440 }, { "grad_norm": 0.6846072673797607, "learning_rate": 8.68432562854735e-05, "loss": 0.0708, "step": 27450 }, { "grad_norm": 0.3861962854862213, "learning_rate": 8.683207615736887e-05, "loss": 0.0567, "step": 27460 }, { "grad_norm": 1.0023483037948608, "learning_rate": 8.682089200136379e-05, "loss": 0.074, "step": 27470 }, { "grad_norm": 0.7574734091758728, "learning_rate": 8.680970381868132e-05, "loss": 0.0758, "step": 27480 }, { "grad_norm": 0.5516501069068909, "learning_rate": 8.679851161054498e-05, "loss": 0.0504, "step": 27490 }, { "grad_norm": 0.5727289319038391, "learning_rate": 8.678731537817873e-05, "loss": 0.0706, "step": 27500 }, { "grad_norm": 0.720180332660675, "learning_rate": 8.677611512280697e-05, "loss": 0.0718, "step": 27510 }, { "grad_norm": 0.9301756620407104, "learning_rate": 8.676491084565457e-05, "loss": 0.0722, "step": 27520 }, { "grad_norm": 0.6005277037620544, "learning_rate": 8.675370254794678e-05, "loss": 0.0692, "step": 27530 }, { "grad_norm": 0.46317529678344727, "learning_rate": 8.674249023090935e-05, "loss": 0.0647, "step": 27540 }, { "grad_norm": 0.4644309878349304, "learning_rate": 8.673127389576843e-05, "loss": 0.0659, "step": 27550 }, { "grad_norm": 0.37335461378097534, "learning_rate": 8.67200535437506e-05, "loss": 0.0545, "step": 27560 }, { "grad_norm": 0.5817500948905945, "learning_rate": 8.670882917608296e-05, "loss": 0.0762, "step": 27570 }, { "grad_norm": 0.5804678201675415, "learning_rate": 8.669760079399292e-05, "loss": 0.0645, "step": 27580 }, { "grad_norm": 0.45251837372779846, "learning_rate": 8.668636839870845e-05, "loss": 0.0585, "step": 27590 }, { "grad_norm": 0.8791580200195312, "learning_rate": 8.667513199145789e-05, "loss": 0.0781, "step": 27600 }, { "grad_norm": 0.5959204435348511, "learning_rate": 8.666389157347002e-05, "loss": 0.0741, "step": 27610 }, { "grad_norm": 2.0401196479797363, "learning_rate": 8.66526471459741e-05, "loss": 0.076, "step": 27620 }, { "grad_norm": 0.8183550834655762, "learning_rate": 8.66413987101998e-05, "loss": 0.0642, "step": 27630 }, { "grad_norm": 0.5937092900276184, "learning_rate": 8.663014626737723e-05, "loss": 0.0639, "step": 27640 }, { "grad_norm": 0.5808686017990112, "learning_rate": 8.661888981873691e-05, "loss": 0.0709, "step": 27650 }, { "grad_norm": 0.4676949679851532, "learning_rate": 8.660762936550988e-05, "loss": 0.0699, "step": 27660 }, { "grad_norm": 1.5498906373977661, "learning_rate": 8.659636490892753e-05, "loss": 0.0687, "step": 27670 }, { "grad_norm": 0.461959570646286, "learning_rate": 8.658509645022174e-05, "loss": 0.0588, "step": 27680 }, { "grad_norm": 0.4408968389034271, "learning_rate": 8.657382399062481e-05, "loss": 0.0477, "step": 27690 }, { "grad_norm": 0.873640239238739, "learning_rate": 8.656254753136946e-05, "loss": 0.0615, "step": 27700 }, { "grad_norm": 1.1714857816696167, "learning_rate": 8.655126707368891e-05, "loss": 0.0474, "step": 27710 }, { "grad_norm": 0.5076577663421631, "learning_rate": 8.653998261881672e-05, "loss": 0.0629, "step": 27720 }, { "grad_norm": 0.42335832118988037, "learning_rate": 8.652869416798699e-05, "loss": 0.061, "step": 27730 }, { "grad_norm": 0.8453391790390015, "learning_rate": 8.651740172243417e-05, "loss": 0.0596, "step": 27740 }, { "grad_norm": 0.8429739475250244, "learning_rate": 8.65061052833932e-05, "loss": 0.0649, "step": 27750 }, { "grad_norm": 0.4993714392185211, "learning_rate": 8.649480485209945e-05, "loss": 0.0625, "step": 27760 }, { "grad_norm": 0.5068888068199158, "learning_rate": 8.64835004297887e-05, "loss": 0.0752, "step": 27770 }, { "grad_norm": 0.6187109351158142, "learning_rate": 8.64721920176972e-05, "loss": 0.0593, "step": 27780 }, { "grad_norm": 0.4290810525417328, "learning_rate": 8.646087961706164e-05, "loss": 0.0693, "step": 27790 }, { "grad_norm": 0.4847659766674042, "learning_rate": 8.644956322911908e-05, "loss": 0.0702, "step": 27800 }, { "grad_norm": 0.5276147127151489, "learning_rate": 8.643824285510709e-05, "loss": 0.0702, "step": 27810 }, { "grad_norm": 0.39992988109588623, "learning_rate": 8.642691849626364e-05, "loss": 0.0474, "step": 27820 }, { "grad_norm": 0.5175642371177673, "learning_rate": 8.641559015382717e-05, "loss": 0.0591, "step": 27830 }, { "grad_norm": 0.6382858753204346, "learning_rate": 8.640425782903649e-05, "loss": 0.0671, "step": 27840 }, { "grad_norm": 0.679652988910675, "learning_rate": 8.639292152313091e-05, "loss": 0.0641, "step": 27850 }, { "grad_norm": 0.7121847867965698, "learning_rate": 8.638158123735015e-05, "loss": 0.0603, "step": 27860 }, { "grad_norm": 0.42645010352134705, "learning_rate": 8.637023697293436e-05, "loss": 0.0546, "step": 27870 }, { "grad_norm": 0.7052738666534424, "learning_rate": 8.635888873112414e-05, "loss": 0.0737, "step": 27880 }, { "grad_norm": 0.4659373164176941, "learning_rate": 8.634753651316052e-05, "loss": 0.0686, "step": 27890 }, { "grad_norm": 0.5187537670135498, "learning_rate": 8.633618032028496e-05, "loss": 0.0606, "step": 27900 }, { "grad_norm": 0.6564949750900269, "learning_rate": 8.632482015373934e-05, "loss": 0.0648, "step": 27910 }, { "grad_norm": 0.4070889353752136, "learning_rate": 8.6313456014766e-05, "loss": 0.0685, "step": 27920 }, { "grad_norm": 0.7951181530952454, "learning_rate": 8.630208790460771e-05, "loss": 0.0585, "step": 27930 }, { "grad_norm": 0.3702634274959564, "learning_rate": 8.629071582450768e-05, "loss": 0.0617, "step": 27940 }, { "grad_norm": 0.8407825827598572, "learning_rate": 8.62793397757095e-05, "loss": 0.0709, "step": 27950 }, { "grad_norm": 0.5990816354751587, "learning_rate": 8.626795975945729e-05, "loss": 0.0616, "step": 27960 }, { "grad_norm": 0.4540656507015228, "learning_rate": 8.625657577699551e-05, "loss": 0.0539, "step": 27970 }, { "grad_norm": 0.5071150064468384, "learning_rate": 8.624518782956914e-05, "loss": 0.0544, "step": 27980 }, { "grad_norm": 0.5366997718811035, "learning_rate": 8.62337959184235e-05, "loss": 0.0652, "step": 27990 }, { "grad_norm": 0.5938019156455994, "learning_rate": 8.622240004480441e-05, "loss": 0.075, "step": 28000 }, { "grad_norm": 1.099975347518921, "learning_rate": 8.621100020995814e-05, "loss": 0.0668, "step": 28010 }, { "grad_norm": 0.6212335824966431, "learning_rate": 8.619959641513132e-05, "loss": 0.0746, "step": 28020 }, { "grad_norm": 0.7035363912582397, "learning_rate": 8.618818866157105e-05, "loss": 0.0712, "step": 28030 }, { "grad_norm": 1.2266768217086792, "learning_rate": 8.617677695052487e-05, "loss": 0.067, "step": 28040 }, { "grad_norm": 0.4971933662891388, "learning_rate": 8.616536128324078e-05, "loss": 0.0648, "step": 28050 }, { "grad_norm": 0.4428929388523102, "learning_rate": 8.615394166096712e-05, "loss": 0.0599, "step": 28060 }, { "grad_norm": 0.4264163076877594, "learning_rate": 8.614251808495279e-05, "loss": 0.0603, "step": 28070 }, { "grad_norm": 0.5007280707359314, "learning_rate": 8.6131090556447e-05, "loss": 0.0642, "step": 28080 }, { "grad_norm": 0.9794949293136597, "learning_rate": 8.611965907669947e-05, "loss": 0.0732, "step": 28090 }, { "grad_norm": 2.0102922916412354, "learning_rate": 8.610822364696034e-05, "loss": 0.074, "step": 28100 }, { "grad_norm": 0.5196062326431274, "learning_rate": 8.609678426848015e-05, "loss": 0.063, "step": 28110 }, { "grad_norm": 0.49377885460853577, "learning_rate": 8.60853409425099e-05, "loss": 0.0658, "step": 28120 }, { "grad_norm": 0.7420003414154053, "learning_rate": 8.607389367030104e-05, "loss": 0.0591, "step": 28130 }, { "grad_norm": 0.4635709226131439, "learning_rate": 8.606244245310538e-05, "loss": 0.0565, "step": 28140 }, { "grad_norm": 0.635688304901123, "learning_rate": 8.605098729217525e-05, "loss": 0.0626, "step": 28150 }, { "grad_norm": 0.4616979956626892, "learning_rate": 8.603952818876335e-05, "loss": 0.0625, "step": 28160 }, { "grad_norm": 0.4151225686073303, "learning_rate": 8.602806514412281e-05, "loss": 0.0564, "step": 28170 }, { "grad_norm": 0.589780867099762, "learning_rate": 8.601659815950726e-05, "loss": 0.0724, "step": 28180 }, { "grad_norm": 0.7487910389900208, "learning_rate": 8.600512723617067e-05, "loss": 0.0796, "step": 28190 }, { "grad_norm": 0.6176947355270386, "learning_rate": 8.59936523753675e-05, "loss": 0.0603, "step": 28200 }, { "grad_norm": 0.3507040739059448, "learning_rate": 8.598217357835264e-05, "loss": 0.0601, "step": 28210 }, { "grad_norm": 0.584114670753479, "learning_rate": 8.597069084638135e-05, "loss": 0.0676, "step": 28220 }, { "grad_norm": 0.5482330918312073, "learning_rate": 8.595920418070939e-05, "loss": 0.0547, "step": 28230 }, { "grad_norm": 0.6319543123245239, "learning_rate": 8.594771358259295e-05, "loss": 0.0578, "step": 28240 }, { "grad_norm": 1.7237036228179932, "learning_rate": 8.593621905328858e-05, "loss": 0.0692, "step": 28250 }, { "grad_norm": 0.4108985960483551, "learning_rate": 8.592472059405333e-05, "loss": 0.0617, "step": 28260 }, { "grad_norm": 0.6628097891807556, "learning_rate": 8.591321820614464e-05, "loss": 0.0599, "step": 28270 }, { "grad_norm": 0.5662732124328613, "learning_rate": 8.590171189082041e-05, "loss": 0.0763, "step": 28280 }, { "grad_norm": 0.519757091999054, "learning_rate": 8.589020164933894e-05, "loss": 0.068, "step": 28290 }, { "grad_norm": 0.9288551211357117, "learning_rate": 8.587868748295898e-05, "loss": 0.0599, "step": 28300 }, { "grad_norm": 0.5397670865058899, "learning_rate": 8.586716939293971e-05, "loss": 0.0549, "step": 28310 }, { "grad_norm": 0.4573623836040497, "learning_rate": 8.58556473805407e-05, "loss": 0.0554, "step": 28320 }, { "grad_norm": 0.5632128119468689, "learning_rate": 8.584412144702202e-05, "loss": 0.0624, "step": 28330 }, { "grad_norm": 0.4633280634880066, "learning_rate": 8.58325915936441e-05, "loss": 0.0602, "step": 28340 }, { "grad_norm": 0.6394897699356079, "learning_rate": 8.582105782166783e-05, "loss": 0.0548, "step": 28350 }, { "grad_norm": 0.45346319675445557, "learning_rate": 8.580952013235455e-05, "loss": 0.0576, "step": 28360 }, { "grad_norm": 1.1535089015960693, "learning_rate": 8.579797852696596e-05, "loss": 0.0584, "step": 28370 }, { "grad_norm": 0.6143168807029724, "learning_rate": 8.578643300676428e-05, "loss": 0.0732, "step": 28380 }, { "grad_norm": 0.4585890471935272, "learning_rate": 8.577488357301209e-05, "loss": 0.0731, "step": 28390 }, { "grad_norm": 0.6015957593917847, "learning_rate": 8.576333022697242e-05, "loss": 0.0542, "step": 28400 }, { "grad_norm": 1.3340020179748535, "learning_rate": 8.575177296990873e-05, "loss": 0.0531, "step": 28410 }, { "grad_norm": 0.8023906350135803, "learning_rate": 8.574021180308489e-05, "loss": 0.0774, "step": 28420 }, { "grad_norm": 0.6337206959724426, "learning_rate": 8.572864672776523e-05, "loss": 0.07, "step": 28430 }, { "grad_norm": 0.3707748353481293, "learning_rate": 8.571707774521447e-05, "loss": 0.0502, "step": 28440 }, { "grad_norm": 0.46233099699020386, "learning_rate": 8.57055048566978e-05, "loss": 0.0549, "step": 28450 }, { "grad_norm": 1.5544757843017578, "learning_rate": 8.569392806348078e-05, "loss": 0.0729, "step": 28460 }, { "grad_norm": 0.5087525248527527, "learning_rate": 8.568234736682947e-05, "loss": 0.0554, "step": 28470 }, { "grad_norm": 0.6262498497962952, "learning_rate": 8.567076276801029e-05, "loss": 0.0711, "step": 28480 }, { "grad_norm": 0.48280009627342224, "learning_rate": 8.565917426829013e-05, "loss": 0.0439, "step": 28490 }, { "grad_norm": 0.7629745006561279, "learning_rate": 8.564758186893628e-05, "loss": 0.0548, "step": 28500 }, { "grad_norm": 0.7993690371513367, "learning_rate": 8.563598557121649e-05, "loss": 0.0653, "step": 28510 }, { "grad_norm": 0.5578405857086182, "learning_rate": 8.562438537639888e-05, "loss": 0.0609, "step": 28520 }, { "grad_norm": 0.5657192468643188, "learning_rate": 8.561278128575206e-05, "loss": 0.0701, "step": 28530 }, { "grad_norm": 0.7642295360565186, "learning_rate": 8.5601173300545e-05, "loss": 0.0762, "step": 28540 }, { "grad_norm": 0.7473756074905396, "learning_rate": 8.558956142204717e-05, "loss": 0.0803, "step": 28550 }, { "grad_norm": 0.6838077902793884, "learning_rate": 8.55779456515284e-05, "loss": 0.0847, "step": 28560 }, { "grad_norm": 0.44661760330200195, "learning_rate": 8.556632599025898e-05, "loss": 0.0735, "step": 28570 }, { "grad_norm": 0.40429526567459106, "learning_rate": 8.555470243950964e-05, "loss": 0.0616, "step": 28580 }, { "grad_norm": 0.3799821138381958, "learning_rate": 8.554307500055148e-05, "loss": 0.0651, "step": 28590 }, { "grad_norm": 0.8485224843025208, "learning_rate": 8.553144367465609e-05, "loss": 0.054, "step": 28600 }, { "grad_norm": 0.512570321559906, "learning_rate": 8.551980846309544e-05, "loss": 0.071, "step": 28610 }, { "grad_norm": 0.6314694881439209, "learning_rate": 8.550816936714193e-05, "loss": 0.063, "step": 28620 }, { "grad_norm": 0.49429476261138916, "learning_rate": 8.549652638806841e-05, "loss": 0.0575, "step": 28630 }, { "grad_norm": 0.8001732230186462, "learning_rate": 8.548487952714812e-05, "loss": 0.0668, "step": 28640 }, { "grad_norm": 0.40588808059692383, "learning_rate": 8.547322878565478e-05, "loss": 0.0564, "step": 28650 }, { "grad_norm": 0.5233379602432251, "learning_rate": 8.546157416486245e-05, "loss": 0.0588, "step": 28660 }, { "grad_norm": 0.47286373376846313, "learning_rate": 8.54499156660457e-05, "loss": 0.0578, "step": 28670 }, { "grad_norm": 0.45679357647895813, "learning_rate": 8.543825329047947e-05, "loss": 0.0487, "step": 28680 }, { "grad_norm": 0.6751078963279724, "learning_rate": 8.542658703943913e-05, "loss": 0.0644, "step": 28690 }, { "grad_norm": 0.5161401629447937, "learning_rate": 8.541491691420051e-05, "loss": 0.061, "step": 28700 }, { "grad_norm": 0.6683601140975952, "learning_rate": 8.54032429160398e-05, "loss": 0.0548, "step": 28710 }, { "grad_norm": 0.5251818895339966, "learning_rate": 8.539156504623369e-05, "loss": 0.0644, "step": 28720 }, { "grad_norm": 0.45401883125305176, "learning_rate": 8.537988330605923e-05, "loss": 0.0799, "step": 28730 }, { "grad_norm": 0.5488929748535156, "learning_rate": 8.536819769679393e-05, "loss": 0.0549, "step": 28740 }, { "grad_norm": 0.5360538363456726, "learning_rate": 8.53565082197157e-05, "loss": 0.0557, "step": 28750 }, { "grad_norm": 0.5988972187042236, "learning_rate": 8.534481487610289e-05, "loss": 0.0795, "step": 28760 }, { "grad_norm": 0.54667067527771, "learning_rate": 8.533311766723428e-05, "loss": 0.0502, "step": 28770 }, { "grad_norm": 0.46123915910720825, "learning_rate": 8.532141659438901e-05, "loss": 0.0538, "step": 28780 }, { "grad_norm": 0.42090168595314026, "learning_rate": 8.530971165884675e-05, "loss": 0.0567, "step": 28790 }, { "grad_norm": 0.3656231760978699, "learning_rate": 8.529800286188752e-05, "loss": 0.0544, "step": 28800 }, { "grad_norm": 0.36026930809020996, "learning_rate": 8.528629020479175e-05, "loss": 0.061, "step": 28810 }, { "grad_norm": 0.9738221168518066, "learning_rate": 8.527457368884033e-05, "loss": 0.0612, "step": 28820 }, { "grad_norm": 0.5588860511779785, "learning_rate": 8.526285331531458e-05, "loss": 0.0513, "step": 28830 }, { "grad_norm": 0.512957751750946, "learning_rate": 8.525112908549621e-05, "loss": 0.047, "step": 28840 }, { "grad_norm": 0.6151920557022095, "learning_rate": 8.523940100066735e-05, "loss": 0.0694, "step": 28850 }, { "grad_norm": 0.35760068893432617, "learning_rate": 8.52276690621106e-05, "loss": 0.0435, "step": 28860 }, { "grad_norm": 0.9400164484977722, "learning_rate": 8.521593327110889e-05, "loss": 0.0619, "step": 28870 }, { "grad_norm": 0.536808967590332, "learning_rate": 8.520419362894569e-05, "loss": 0.0669, "step": 28880 }, { "grad_norm": 0.480652391910553, "learning_rate": 8.51924501369048e-05, "loss": 0.0461, "step": 28890 }, { "grad_norm": 0.6441835761070251, "learning_rate": 8.518070279627047e-05, "loss": 0.0647, "step": 28900 }, { "grad_norm": 1.1041008234024048, "learning_rate": 8.516895160832737e-05, "loss": 0.0599, "step": 28910 }, { "grad_norm": 0.5890645980834961, "learning_rate": 8.515719657436061e-05, "loss": 0.0537, "step": 28920 }, { "grad_norm": 0.7062212228775024, "learning_rate": 8.514543769565568e-05, "loss": 0.0589, "step": 28930 }, { "grad_norm": 1.3110913038253784, "learning_rate": 8.513367497349853e-05, "loss": 0.063, "step": 28940 }, { "grad_norm": 0.3999975621700287, "learning_rate": 8.51219084091755e-05, "loss": 0.0634, "step": 28950 }, { "grad_norm": 0.6611905097961426, "learning_rate": 8.511013800397338e-05, "loss": 0.0587, "step": 28960 }, { "grad_norm": 0.6883499026298523, "learning_rate": 8.509836375917937e-05, "loss": 0.0717, "step": 28970 }, { "grad_norm": 0.6587783098220825, "learning_rate": 8.508658567608104e-05, "loss": 0.0579, "step": 28980 }, { "grad_norm": 0.4869782626628876, "learning_rate": 8.507480375596647e-05, "loss": 0.0518, "step": 28990 }, { "grad_norm": 0.5414456129074097, "learning_rate": 8.506301800012408e-05, "loss": 0.0512, "step": 29000 }, { "grad_norm": 0.3914283812046051, "learning_rate": 8.505122840984278e-05, "loss": 0.0542, "step": 29010 }, { "grad_norm": 0.6051722168922424, "learning_rate": 8.503943498641182e-05, "loss": 0.0699, "step": 29020 }, { "grad_norm": 0.4954679310321808, "learning_rate": 8.502763773112095e-05, "loss": 0.0639, "step": 29030 }, { "grad_norm": 0.600144624710083, "learning_rate": 8.501583664526026e-05, "loss": 0.0582, "step": 29040 }, { "grad_norm": 0.5412831902503967, "learning_rate": 8.500403173012032e-05, "loss": 0.0645, "step": 29050 }, { "grad_norm": 0.4466613829135895, "learning_rate": 8.499222298699211e-05, "loss": 0.0707, "step": 29060 }, { "grad_norm": 0.48044252395629883, "learning_rate": 8.498041041716701e-05, "loss": 0.0586, "step": 29070 }, { "grad_norm": 0.6699829697608948, "learning_rate": 8.496859402193681e-05, "loss": 0.0597, "step": 29080 }, { "grad_norm": 0.5771399140357971, "learning_rate": 8.495677380259374e-05, "loss": 0.0554, "step": 29090 }, { "grad_norm": 0.44966328144073486, "learning_rate": 8.494494976043045e-05, "loss": 0.0573, "step": 29100 }, { "grad_norm": 0.4273490905761719, "learning_rate": 8.493312189673998e-05, "loss": 0.053, "step": 29110 }, { "grad_norm": 0.36817681789398193, "learning_rate": 8.492129021281584e-05, "loss": 0.0528, "step": 29120 }, { "grad_norm": 0.5765346884727478, "learning_rate": 8.490945470995188e-05, "loss": 0.0567, "step": 29130 }, { "grad_norm": 0.40052318572998047, "learning_rate": 8.489761538944247e-05, "loss": 0.0636, "step": 29140 }, { "grad_norm": 0.6834847927093506, "learning_rate": 8.48857722525823e-05, "loss": 0.0561, "step": 29150 }, { "grad_norm": 0.37314996123313904, "learning_rate": 8.487392530066652e-05, "loss": 0.0523, "step": 29160 }, { "grad_norm": 0.5547340512275696, "learning_rate": 8.486207453499069e-05, "loss": 0.0642, "step": 29170 }, { "grad_norm": 0.49837547540664673, "learning_rate": 8.485021995685082e-05, "loss": 0.0547, "step": 29180 }, { "grad_norm": 0.5006276369094849, "learning_rate": 8.483836156754328e-05, "loss": 0.0571, "step": 29190 }, { "grad_norm": 0.5257121324539185, "learning_rate": 8.482649936836491e-05, "loss": 0.0575, "step": 29200 }, { "grad_norm": 0.5393669009208679, "learning_rate": 8.481463336061293e-05, "loss": 0.0512, "step": 29210 }, { "grad_norm": 0.41403928399086, "learning_rate": 8.480276354558496e-05, "loss": 0.0804, "step": 29220 }, { "grad_norm": 1.2133731842041016, "learning_rate": 8.479088992457913e-05, "loss": 0.0705, "step": 29230 }, { "grad_norm": 0.4157496988773346, "learning_rate": 8.477901249889387e-05, "loss": 0.0664, "step": 29240 }, { "grad_norm": 0.5220253467559814, "learning_rate": 8.47671312698281e-05, "loss": 0.0499, "step": 29250 }, { "grad_norm": 0.8303579688072205, "learning_rate": 8.475524623868112e-05, "loss": 0.0666, "step": 29260 }, { "grad_norm": 0.4689617156982422, "learning_rate": 8.474335740675266e-05, "loss": 0.0572, "step": 29270 }, { "grad_norm": 0.49096986651420593, "learning_rate": 8.473146477534289e-05, "loss": 0.0638, "step": 29280 }, { "grad_norm": 0.7221724987030029, "learning_rate": 8.471956834575232e-05, "loss": 0.0614, "step": 29290 }, { "grad_norm": 0.6119093298912048, "learning_rate": 8.470766811928197e-05, "loss": 0.052, "step": 29300 }, { "grad_norm": 0.553091287612915, "learning_rate": 8.469576409723323e-05, "loss": 0.0522, "step": 29310 }, { "grad_norm": 0.5563251972198486, "learning_rate": 8.468385628090788e-05, "loss": 0.0552, "step": 29320 }, { "grad_norm": 0.6524003148078918, "learning_rate": 8.467194467160815e-05, "loss": 0.0589, "step": 29330 }, { "grad_norm": 1.079733967781067, "learning_rate": 8.466002927063667e-05, "loss": 0.0646, "step": 29340 }, { "grad_norm": 1.0809370279312134, "learning_rate": 8.464811007929651e-05, "loss": 0.0627, "step": 29350 }, { "grad_norm": 0.4810851812362671, "learning_rate": 8.463618709889114e-05, "loss": 0.0567, "step": 29360 }, { "grad_norm": 0.58122718334198, "learning_rate": 8.462426033072442e-05, "loss": 0.0746, "step": 29370 }, { "grad_norm": 0.5871580839157104, "learning_rate": 8.461232977610061e-05, "loss": 0.0628, "step": 29380 }, { "grad_norm": 0.6031127572059631, "learning_rate": 8.46003954363245e-05, "loss": 0.0543, "step": 29390 }, { "grad_norm": 0.5121972560882568, "learning_rate": 8.458845731270115e-05, "loss": 0.0633, "step": 29400 }, { "grad_norm": 0.5444295406341553, "learning_rate": 8.45765154065361e-05, "loss": 0.0597, "step": 29410 }, { "grad_norm": 0.2934027910232544, "learning_rate": 8.456456971913532e-05, "loss": 0.0505, "step": 29420 }, { "grad_norm": 0.6437637209892273, "learning_rate": 8.455262025180517e-05, "loss": 0.0643, "step": 29430 }, { "grad_norm": 0.4314122200012207, "learning_rate": 8.454066700585242e-05, "loss": 0.0487, "step": 29440 }, { "grad_norm": 0.6019929647445679, "learning_rate": 8.452870998258423e-05, "loss": 0.0687, "step": 29450 }, { "grad_norm": 0.3591826260089874, "learning_rate": 8.451674918330825e-05, "loss": 0.0547, "step": 29460 }, { "grad_norm": 0.5425375699996948, "learning_rate": 8.450478460933246e-05, "loss": 0.0672, "step": 29470 }, { "grad_norm": 0.4911608397960663, "learning_rate": 8.449281626196532e-05, "loss": 0.0655, "step": 29480 }, { "grad_norm": 0.5140109658241272, "learning_rate": 8.448084414251564e-05, "loss": 0.0668, "step": 29490 }, { "grad_norm": 0.49746838212013245, "learning_rate": 8.446886825229271e-05, "loss": 0.0527, "step": 29500 }, { "grad_norm": 0.622452437877655, "learning_rate": 8.445688859260615e-05, "loss": 0.0557, "step": 29510 }, { "grad_norm": 0.4766550660133362, "learning_rate": 8.444490516476606e-05, "loss": 0.0534, "step": 29520 }, { "grad_norm": 0.4592449963092804, "learning_rate": 8.443291797008293e-05, "loss": 0.062, "step": 29530 }, { "grad_norm": 0.39378273487091064, "learning_rate": 8.442092700986765e-05, "loss": 0.0657, "step": 29540 }, { "grad_norm": 0.35522809624671936, "learning_rate": 8.440893228543156e-05, "loss": 0.0538, "step": 29550 }, { "grad_norm": 0.8256114721298218, "learning_rate": 8.439693379808638e-05, "loss": 0.0516, "step": 29560 }, { "grad_norm": 0.505745530128479, "learning_rate": 8.43849315491442e-05, "loss": 0.0588, "step": 29570 }, { "grad_norm": 0.471416175365448, "learning_rate": 8.437292553991763e-05, "loss": 0.0523, "step": 29580 }, { "grad_norm": 0.8509436845779419, "learning_rate": 8.436091577171959e-05, "loss": 0.0745, "step": 29590 }, { "grad_norm": 0.6512067317962646, "learning_rate": 8.434890224586347e-05, "loss": 0.0684, "step": 29600 }, { "grad_norm": 0.4964480698108673, "learning_rate": 8.433688496366303e-05, "loss": 0.064, "step": 29610 }, { "grad_norm": 0.5177612900733948, "learning_rate": 8.432486392643248e-05, "loss": 0.0598, "step": 29620 }, { "grad_norm": 1.2780513763427734, "learning_rate": 8.431283913548643e-05, "loss": 0.0669, "step": 29630 }, { "grad_norm": 0.4937528669834137, "learning_rate": 8.430081059213985e-05, "loss": 0.059, "step": 29640 }, { "grad_norm": 0.41842636466026306, "learning_rate": 8.428877829770823e-05, "loss": 0.0579, "step": 29650 }, { "grad_norm": 0.44155657291412354, "learning_rate": 8.427674225350735e-05, "loss": 0.0639, "step": 29660 }, { "grad_norm": 0.41714128851890564, "learning_rate": 8.426470246085347e-05, "loss": 0.0594, "step": 29670 }, { "grad_norm": 0.42006921768188477, "learning_rate": 8.425265892106324e-05, "loss": 0.0628, "step": 29680 }, { "grad_norm": 0.3823021650314331, "learning_rate": 8.424061163545374e-05, "loss": 0.0659, "step": 29690 }, { "grad_norm": 0.5000033378601074, "learning_rate": 8.422856060534243e-05, "loss": 0.0539, "step": 29700 }, { "grad_norm": 0.5135219097137451, "learning_rate": 8.421650583204718e-05, "loss": 0.0529, "step": 29710 }, { "grad_norm": 0.5292919874191284, "learning_rate": 8.420444731688633e-05, "loss": 0.0589, "step": 29720 }, { "grad_norm": 0.44036659598350525, "learning_rate": 8.419238506117852e-05, "loss": 0.0794, "step": 29730 }, { "grad_norm": 0.6766067743301392, "learning_rate": 8.418031906624289e-05, "loss": 0.0571, "step": 29740 }, { "grad_norm": 0.5210286974906921, "learning_rate": 8.416824933339898e-05, "loss": 0.0642, "step": 29750 }, { "grad_norm": 0.6624630689620972, "learning_rate": 8.415617586396667e-05, "loss": 0.0502, "step": 29760 }, { "grad_norm": 0.7752684354782104, "learning_rate": 8.414409865926632e-05, "loss": 0.0643, "step": 29770 }, { "grad_norm": 0.4585416913032532, "learning_rate": 8.413201772061867e-05, "loss": 0.0574, "step": 29780 }, { "grad_norm": 0.504462718963623, "learning_rate": 8.411993304934488e-05, "loss": 0.0436, "step": 29790 }, { "grad_norm": 0.535199761390686, "learning_rate": 8.410784464676654e-05, "loss": 0.0569, "step": 29800 }, { "grad_norm": 0.37255802750587463, "learning_rate": 8.409575251420556e-05, "loss": 0.0553, "step": 29810 }, { "grad_norm": 0.6047162413597107, "learning_rate": 8.408365665298435e-05, "loss": 0.057, "step": 29820 }, { "grad_norm": 0.7427939176559448, "learning_rate": 8.40715570644257e-05, "loss": 0.0661, "step": 29830 }, { "grad_norm": 0.45048049092292786, "learning_rate": 8.40594537498528e-05, "loss": 0.0657, "step": 29840 }, { "grad_norm": 0.47616803646087646, "learning_rate": 8.404734671058924e-05, "loss": 0.0685, "step": 29850 }, { "grad_norm": 0.4858308434486389, "learning_rate": 8.403523594795902e-05, "loss": 0.064, "step": 29860 }, { "grad_norm": 0.6305108666419983, "learning_rate": 8.402312146328659e-05, "loss": 0.0584, "step": 29870 }, { "grad_norm": 0.7445534467697144, "learning_rate": 8.401100325789675e-05, "loss": 0.0589, "step": 29880 }, { "grad_norm": 0.6289721727371216, "learning_rate": 8.399888133311472e-05, "loss": 0.0579, "step": 29890 }, { "grad_norm": 0.49589529633522034, "learning_rate": 8.398675569026613e-05, "loss": 0.065, "step": 29900 }, { "grad_norm": 0.7454023361206055, "learning_rate": 8.397462633067705e-05, "loss": 0.0556, "step": 29910 }, { "grad_norm": 0.4876537322998047, "learning_rate": 8.396249325567392e-05, "loss": 0.0538, "step": 29920 }, { "grad_norm": 0.7923561930656433, "learning_rate": 8.395035646658357e-05, "loss": 0.0768, "step": 29930 }, { "grad_norm": 0.5903810262680054, "learning_rate": 8.39382159647333e-05, "loss": 0.0687, "step": 29940 }, { "grad_norm": 0.4649660587310791, "learning_rate": 8.392607175145075e-05, "loss": 0.0559, "step": 29950 }, { "grad_norm": 0.5522924065589905, "learning_rate": 8.3913923828064e-05, "loss": 0.0608, "step": 29960 }, { "grad_norm": 0.4945540130138397, "learning_rate": 8.390177219590152e-05, "loss": 0.0542, "step": 29970 }, { "grad_norm": 0.7893526554107666, "learning_rate": 8.388961685629222e-05, "loss": 0.0721, "step": 29980 }, { "grad_norm": 0.8801743984222412, "learning_rate": 8.387745781056536e-05, "loss": 0.0605, "step": 29990 }, { "grad_norm": 0.5337509512901306, "learning_rate": 8.386529506005065e-05, "loss": 0.0485, "step": 30000 }, { "grad_norm": 0.703195333480835, "learning_rate": 8.38531286060782e-05, "loss": 0.0559, "step": 30010 }, { "grad_norm": 0.8212457895278931, "learning_rate": 8.384095844997849e-05, "loss": 0.0525, "step": 30020 }, { "grad_norm": 0.5827746987342834, "learning_rate": 8.382878459308245e-05, "loss": 0.0485, "step": 30030 }, { "grad_norm": 0.503673791885376, "learning_rate": 8.381660703672138e-05, "loss": 0.0579, "step": 30040 }, { "grad_norm": 0.6063117980957031, "learning_rate": 8.380442578222702e-05, "loss": 0.0941, "step": 30050 }, { "grad_norm": 0.7677105665206909, "learning_rate": 8.379224083093146e-05, "loss": 0.0614, "step": 30060 }, { "grad_norm": 0.3947772979736328, "learning_rate": 8.378005218416727e-05, "loss": 0.0627, "step": 30070 }, { "grad_norm": 0.5742707252502441, "learning_rate": 8.376785984326735e-05, "loss": 0.069, "step": 30080 }, { "grad_norm": 0.8741599917411804, "learning_rate": 8.375566380956506e-05, "loss": 0.0689, "step": 30090 }, { "grad_norm": 0.9071758985519409, "learning_rate": 8.374346408439411e-05, "loss": 0.0677, "step": 30100 }, { "grad_norm": 0.4628143906593323, "learning_rate": 8.373126066908868e-05, "loss": 0.0537, "step": 30110 }, { "grad_norm": 0.7177779078483582, "learning_rate": 8.371905356498326e-05, "loss": 0.0637, "step": 30120 }, { "grad_norm": 0.4506618082523346, "learning_rate": 8.370684277341288e-05, "loss": 0.0536, "step": 30130 }, { "grad_norm": 0.7027643322944641, "learning_rate": 8.369462829571282e-05, "loss": 0.0582, "step": 30140 }, { "grad_norm": 1.0117309093475342, "learning_rate": 8.36824101332189e-05, "loss": 0.0539, "step": 30150 }, { "grad_norm": 0.42745718359947205, "learning_rate": 8.367018828726721e-05, "loss": 0.0518, "step": 30160 }, { "grad_norm": 0.49955788254737854, "learning_rate": 8.365796275919438e-05, "loss": 0.0687, "step": 30170 }, { "grad_norm": 0.6993494629859924, "learning_rate": 8.364573355033734e-05, "loss": 0.0579, "step": 30180 }, { "grad_norm": 0.6049877405166626, "learning_rate": 8.363350066203346e-05, "loss": 0.0489, "step": 30190 }, { "grad_norm": 0.7535688281059265, "learning_rate": 8.362126409562053e-05, "loss": 0.0579, "step": 30200 }, { "grad_norm": 0.5421367287635803, "learning_rate": 8.360902385243667e-05, "loss": 0.0475, "step": 30210 }, { "grad_norm": 0.47908392548561096, "learning_rate": 8.359677993382052e-05, "loss": 0.0505, "step": 30220 }, { "grad_norm": 0.3929991126060486, "learning_rate": 8.358453234111103e-05, "loss": 0.0556, "step": 30230 }, { "grad_norm": 2.5359907150268555, "learning_rate": 8.357228107564756e-05, "loss": 0.0594, "step": 30240 }, { "grad_norm": 0.642772376537323, "learning_rate": 8.356002613876993e-05, "loss": 0.0603, "step": 30250 }, { "grad_norm": 0.38519713282585144, "learning_rate": 8.35477675318183e-05, "loss": 0.0471, "step": 30260 }, { "grad_norm": 0.5462617874145508, "learning_rate": 8.353550525613323e-05, "loss": 0.0562, "step": 30270 }, { "grad_norm": 0.41797715425491333, "learning_rate": 8.352323931305572e-05, "loss": 0.0642, "step": 30280 }, { "grad_norm": 0.507734477519989, "learning_rate": 8.351096970392717e-05, "loss": 0.0635, "step": 30290 }, { "grad_norm": 0.3912416100502014, "learning_rate": 8.349869643008937e-05, "loss": 0.0496, "step": 30300 }, { "grad_norm": 0.3464157283306122, "learning_rate": 8.348641949288449e-05, "loss": 0.0596, "step": 30310 }, { "grad_norm": 0.5404161810874939, "learning_rate": 8.34741388936551e-05, "loss": 0.0657, "step": 30320 }, { "grad_norm": 0.5643109083175659, "learning_rate": 8.346185463374423e-05, "loss": 0.0667, "step": 30330 }, { "grad_norm": 0.6350741982460022, "learning_rate": 8.344956671449524e-05, "loss": 0.0623, "step": 30340 }, { "grad_norm": 0.5045971274375916, "learning_rate": 8.343727513725192e-05, "loss": 0.0603, "step": 30350 }, { "grad_norm": 0.6070746183395386, "learning_rate": 8.342497990335847e-05, "loss": 0.0621, "step": 30360 }, { "grad_norm": 0.8240295052528381, "learning_rate": 8.341268101415946e-05, "loss": 0.0686, "step": 30370 }, { "grad_norm": 0.46405941247940063, "learning_rate": 8.34003784709999e-05, "loss": 0.0578, "step": 30380 }, { "grad_norm": 0.35966959595680237, "learning_rate": 8.338807227522516e-05, "loss": 0.0545, "step": 30390 }, { "grad_norm": 0.8566376566886902, "learning_rate": 8.337576242818103e-05, "loss": 0.065, "step": 30400 }, { "grad_norm": 0.3688478171825409, "learning_rate": 8.336344893121372e-05, "loss": 0.0551, "step": 30410 }, { "grad_norm": 0.6384485363960266, "learning_rate": 8.335113178566977e-05, "loss": 0.0582, "step": 30420 }, { "grad_norm": 0.6032878160476685, "learning_rate": 8.33388109928962e-05, "loss": 0.0687, "step": 30430 }, { "grad_norm": 0.5389753580093384, "learning_rate": 8.33264865542404e-05, "loss": 0.0631, "step": 30440 }, { "grad_norm": 0.44004642963409424, "learning_rate": 8.331415847105013e-05, "loss": 0.0634, "step": 30450 }, { "grad_norm": 0.5389770269393921, "learning_rate": 8.330182674467357e-05, "loss": 0.0524, "step": 30460 }, { "grad_norm": 0.4896417260169983, "learning_rate": 8.32894913764593e-05, "loss": 0.0536, "step": 30470 }, { "grad_norm": 0.5783535838127136, "learning_rate": 8.327715236775633e-05, "loss": 0.0525, "step": 30480 }, { "grad_norm": 0.39595961570739746, "learning_rate": 8.326480971991398e-05, "loss": 0.0499, "step": 30490 }, { "grad_norm": 0.5318369269371033, "learning_rate": 8.325246343428206e-05, "loss": 0.0593, "step": 30500 }, { "grad_norm": 0.3562708795070648, "learning_rate": 8.324011351221072e-05, "loss": 0.0482, "step": 30510 }, { "grad_norm": 0.3770337402820587, "learning_rate": 8.322775995505057e-05, "loss": 0.0581, "step": 30520 }, { "grad_norm": 0.4844357371330261, "learning_rate": 8.321540276415254e-05, "loss": 0.0664, "step": 30530 }, { "grad_norm": 0.5530306696891785, "learning_rate": 8.320304194086798e-05, "loss": 0.0562, "step": 30540 }, { "grad_norm": 0.6505812406539917, "learning_rate": 8.31906774865487e-05, "loss": 0.0518, "step": 30550 }, { "grad_norm": 0.42251551151275635, "learning_rate": 8.317830940254682e-05, "loss": 0.0557, "step": 30560 }, { "grad_norm": 0.46304723620414734, "learning_rate": 8.316593769021491e-05, "loss": 0.0588, "step": 30570 }, { "grad_norm": 0.8624468445777893, "learning_rate": 8.315356235090592e-05, "loss": 0.0568, "step": 30580 }, { "grad_norm": 0.6705038547515869, "learning_rate": 8.314118338597319e-05, "loss": 0.0595, "step": 30590 }, { "grad_norm": 1.564041256904602, "learning_rate": 8.312880079677048e-05, "loss": 0.0581, "step": 30600 }, { "grad_norm": 0.5817983746528625, "learning_rate": 8.311641458465191e-05, "loss": 0.0671, "step": 30610 }, { "grad_norm": 0.5069672465324402, "learning_rate": 8.310402475097205e-05, "loss": 0.0646, "step": 30620 }, { "grad_norm": 0.4630220830440521, "learning_rate": 8.309163129708581e-05, "loss": 0.0499, "step": 30630 }, { "grad_norm": 0.4747750163078308, "learning_rate": 8.307923422434852e-05, "loss": 0.0547, "step": 30640 }, { "grad_norm": 0.8205863833427429, "learning_rate": 8.30668335341159e-05, "loss": 0.0699, "step": 30650 }, { "grad_norm": 0.5467492938041687, "learning_rate": 8.305442922774408e-05, "loss": 0.0657, "step": 30660 }, { "grad_norm": 0.715555727481842, "learning_rate": 8.304202130658959e-05, "loss": 0.0597, "step": 30670 }, { "grad_norm": 0.6891045570373535, "learning_rate": 8.302960977200931e-05, "loss": 0.054, "step": 30680 }, { "grad_norm": 0.7365955114364624, "learning_rate": 8.301719462536058e-05, "loss": 0.0599, "step": 30690 }, { "grad_norm": 0.5190304517745972, "learning_rate": 8.300477586800108e-05, "loss": 0.052, "step": 30700 }, { "grad_norm": 0.3699359595775604, "learning_rate": 8.299235350128892e-05, "loss": 0.0586, "step": 30710 }, { "grad_norm": 0.5338300466537476, "learning_rate": 8.297992752658258e-05, "loss": 0.0395, "step": 30720 }, { "grad_norm": 0.47332751750946045, "learning_rate": 8.296749794524094e-05, "loss": 0.0554, "step": 30730 }, { "grad_norm": 0.58188396692276, "learning_rate": 8.295506475862332e-05, "loss": 0.0434, "step": 30740 }, { "grad_norm": 0.42433294653892517, "learning_rate": 8.294262796808933e-05, "loss": 0.044, "step": 30750 }, { "grad_norm": 0.593316912651062, "learning_rate": 8.293018757499909e-05, "loss": 0.0544, "step": 30760 }, { "grad_norm": 0.3989294767379761, "learning_rate": 8.291774358071305e-05, "loss": 0.0564, "step": 30770 }, { "grad_norm": 1.3624463081359863, "learning_rate": 8.290529598659205e-05, "loss": 0.0657, "step": 30780 }, { "grad_norm": 0.5993204116821289, "learning_rate": 8.289284479399738e-05, "loss": 0.0578, "step": 30790 }, { "grad_norm": 0.5867372751235962, "learning_rate": 8.288039000429064e-05, "loss": 0.0563, "step": 30800 }, { "grad_norm": 0.6179808378219604, "learning_rate": 8.286793161883388e-05, "loss": 0.0605, "step": 30810 }, { "grad_norm": 0.43249014019966125, "learning_rate": 8.285546963898954e-05, "loss": 0.0589, "step": 30820 }, { "grad_norm": 0.6424086689949036, "learning_rate": 8.284300406612044e-05, "loss": 0.0612, "step": 30830 }, { "grad_norm": 0.652134120464325, "learning_rate": 8.283053490158978e-05, "loss": 0.0541, "step": 30840 }, { "grad_norm": 0.5771986842155457, "learning_rate": 8.28180621467612e-05, "loss": 0.0491, "step": 30850 }, { "grad_norm": 0.5059981346130371, "learning_rate": 8.280558580299867e-05, "loss": 0.0627, "step": 30860 }, { "grad_norm": 0.7338699102401733, "learning_rate": 8.279310587166661e-05, "loss": 0.0504, "step": 30870 }, { "grad_norm": 0.7490909695625305, "learning_rate": 8.278062235412978e-05, "loss": 0.0541, "step": 30880 }, { "grad_norm": 0.6800950765609741, "learning_rate": 8.276813525175339e-05, "loss": 0.0556, "step": 30890 }, { "grad_norm": 1.3230293989181519, "learning_rate": 8.2755644565903e-05, "loss": 0.0568, "step": 30900 }, { "grad_norm": 0.5608742237091064, "learning_rate": 8.274315029794454e-05, "loss": 0.0514, "step": 30910 }, { "grad_norm": 1.051353096961975, "learning_rate": 8.273065244924443e-05, "loss": 0.0672, "step": 30920 }, { "grad_norm": 1.465325951576233, "learning_rate": 8.271815102116936e-05, "loss": 0.0571, "step": 30930 }, { "grad_norm": 0.5275514125823975, "learning_rate": 8.270564601508648e-05, "loss": 0.0564, "step": 30940 }, { "grad_norm": 0.43292346596717834, "learning_rate": 8.269313743236333e-05, "loss": 0.0567, "step": 30950 }, { "grad_norm": 0.5185064673423767, "learning_rate": 8.268062527436783e-05, "loss": 0.0562, "step": 30960 }, { "grad_norm": 0.6024715304374695, "learning_rate": 8.266810954246828e-05, "loss": 0.0641, "step": 30970 }, { "grad_norm": 0.5609113574028015, "learning_rate": 8.265559023803338e-05, "loss": 0.0456, "step": 30980 }, { "grad_norm": 0.443105548620224, "learning_rate": 8.264306736243223e-05, "loss": 0.0505, "step": 30990 }, { "grad_norm": 0.7097318768501282, "learning_rate": 8.263054091703432e-05, "loss": 0.0634, "step": 31000 }, { "grad_norm": 0.4813219904899597, "learning_rate": 8.26180109032095e-05, "loss": 0.0498, "step": 31010 }, { "grad_norm": 0.612848699092865, "learning_rate": 8.260547732232807e-05, "loss": 0.0419, "step": 31020 }, { "grad_norm": 0.4589542746543884, "learning_rate": 8.259294017576063e-05, "loss": 0.0515, "step": 31030 }, { "grad_norm": 0.5150092840194702, "learning_rate": 8.258039946487828e-05, "loss": 0.048, "step": 31040 }, { "grad_norm": 0.420124888420105, "learning_rate": 8.256785519105241e-05, "loss": 0.0484, "step": 31050 }, { "grad_norm": 1.391716718673706, "learning_rate": 8.255530735565488e-05, "loss": 0.0538, "step": 31060 }, { "grad_norm": 0.3944701850414276, "learning_rate": 8.254275596005787e-05, "loss": 0.0419, "step": 31070 }, { "grad_norm": 0.46569857001304626, "learning_rate": 8.2530201005634e-05, "loss": 0.0699, "step": 31080 }, { "grad_norm": 0.7092862725257874, "learning_rate": 8.251764249375626e-05, "loss": 0.0645, "step": 31090 }, { "grad_norm": 0.6241666078567505, "learning_rate": 8.250508042579803e-05, "loss": 0.0562, "step": 31100 }, { "grad_norm": 1.900354266166687, "learning_rate": 8.249251480313307e-05, "loss": 0.0672, "step": 31110 }, { "grad_norm": 0.501530110836029, "learning_rate": 8.247994562713555e-05, "loss": 0.0567, "step": 31120 }, { "grad_norm": 0.5290302038192749, "learning_rate": 8.246737289917998e-05, "loss": 0.059, "step": 31130 }, { "grad_norm": 0.5867272615432739, "learning_rate": 8.245479662064135e-05, "loss": 0.0529, "step": 31140 }, { "grad_norm": 0.4644547402858734, "learning_rate": 8.244221679289496e-05, "loss": 0.0533, "step": 31150 }, { "grad_norm": 0.7093636989593506, "learning_rate": 8.242963341731652e-05, "loss": 0.0514, "step": 31160 }, { "grad_norm": 0.9275241494178772, "learning_rate": 8.24170464952821e-05, "loss": 0.0499, "step": 31170 }, { "grad_norm": 0.5062581896781921, "learning_rate": 8.240445602816824e-05, "loss": 0.0526, "step": 31180 }, { "grad_norm": 2.267211437225342, "learning_rate": 8.239186201735179e-05, "loss": 0.0578, "step": 31190 }, { "grad_norm": 0.4709128141403198, "learning_rate": 8.237926446420998e-05, "loss": 0.0463, "step": 31200 }, { "grad_norm": 0.5556013584136963, "learning_rate": 8.236666337012052e-05, "loss": 0.052, "step": 31210 }, { "grad_norm": 0.7634420394897461, "learning_rate": 8.23540587364614e-05, "loss": 0.0666, "step": 31220 }, { "grad_norm": 0.3161231279373169, "learning_rate": 8.234145056461107e-05, "loss": 0.053, "step": 31230 }, { "grad_norm": 0.4793567955493927, "learning_rate": 8.232883885594832e-05, "loss": 0.066, "step": 31240 }, { "grad_norm": 0.5114241242408752, "learning_rate": 8.231622361185236e-05, "loss": 0.0658, "step": 31250 }, { "grad_norm": 0.5601114630699158, "learning_rate": 8.230360483370278e-05, "loss": 0.078, "step": 31260 }, { "grad_norm": 0.4387335777282715, "learning_rate": 8.229098252287953e-05, "loss": 0.0569, "step": 31270 }, { "grad_norm": 0.41711148619651794, "learning_rate": 8.2278356680763e-05, "loss": 0.067, "step": 31280 }, { "grad_norm": 0.9403130412101746, "learning_rate": 8.22657273087339e-05, "loss": 0.0676, "step": 31290 }, { "grad_norm": 0.6826196908950806, "learning_rate": 8.225309440817336e-05, "loss": 0.0704, "step": 31300 }, { "grad_norm": 0.6558838486671448, "learning_rate": 8.224045798046293e-05, "loss": 0.0876, "step": 31310 }, { "grad_norm": 0.42934444546699524, "learning_rate": 8.22278180269845e-05, "loss": 0.0501, "step": 31320 }, { "grad_norm": 0.630704402923584, "learning_rate": 8.221517454912031e-05, "loss": 0.0554, "step": 31330 }, { "grad_norm": 0.7532792091369629, "learning_rate": 8.220252754825308e-05, "loss": 0.047, "step": 31340 }, { "grad_norm": 0.4221154451370239, "learning_rate": 8.218987702576586e-05, "loss": 0.0514, "step": 31350 }, { "grad_norm": 0.3519574999809265, "learning_rate": 8.217722298304207e-05, "loss": 0.0564, "step": 31360 }, { "grad_norm": 0.4690796136856079, "learning_rate": 8.216456542146557e-05, "loss": 0.0517, "step": 31370 }, { "grad_norm": 0.6331527829170227, "learning_rate": 8.215190434242055e-05, "loss": 0.0571, "step": 31380 }, { "grad_norm": 0.5978357195854187, "learning_rate": 8.213923974729161e-05, "loss": 0.0592, "step": 31390 }, { "grad_norm": 0.45894017815589905, "learning_rate": 8.212657163746373e-05, "loss": 0.0475, "step": 31400 }, { "grad_norm": 0.5722034573554993, "learning_rate": 8.211390001432227e-05, "loss": 0.0609, "step": 31410 }, { "grad_norm": 0.500774621963501, "learning_rate": 8.210122487925297e-05, "loss": 0.0563, "step": 31420 }, { "grad_norm": 0.5359107255935669, "learning_rate": 8.208854623364202e-05, "loss": 0.0624, "step": 31430 }, { "grad_norm": 0.35428473353385925, "learning_rate": 8.207586407887589e-05, "loss": 0.045, "step": 31440 }, { "grad_norm": 0.6686760187149048, "learning_rate": 8.206317841634148e-05, "loss": 0.0462, "step": 31450 }, { "grad_norm": 0.36852461099624634, "learning_rate": 8.205048924742609e-05, "loss": 0.0583, "step": 31460 }, { "grad_norm": 0.7850769758224487, "learning_rate": 8.203779657351738e-05, "loss": 0.0543, "step": 31470 }, { "grad_norm": 0.49976763129234314, "learning_rate": 8.20251003960034e-05, "loss": 0.0563, "step": 31480 }, { "grad_norm": 0.5526910424232483, "learning_rate": 8.201240071627258e-05, "loss": 0.0552, "step": 31490 }, { "grad_norm": 0.40279388427734375, "learning_rate": 8.199969753571377e-05, "loss": 0.0569, "step": 31500 }, { "grad_norm": 0.5538421869277954, "learning_rate": 8.198699085571615e-05, "loss": 0.0547, "step": 31510 }, { "grad_norm": 0.5062416195869446, "learning_rate": 8.197428067766928e-05, "loss": 0.0598, "step": 31520 }, { "grad_norm": 0.4784538149833679, "learning_rate": 8.196156700296316e-05, "loss": 0.0667, "step": 31530 }, { "grad_norm": 0.6149478554725647, "learning_rate": 8.194884983298814e-05, "loss": 0.0519, "step": 31540 }, { "grad_norm": 0.4721738398075104, "learning_rate": 8.193612916913491e-05, "loss": 0.0481, "step": 31550 }, { "grad_norm": 2.4026267528533936, "learning_rate": 8.192340501279463e-05, "loss": 0.0573, "step": 31560 }, { "grad_norm": 0.4572485387325287, "learning_rate": 8.191067736535876e-05, "loss": 0.0597, "step": 31570 }, { "grad_norm": 0.5851448774337769, "learning_rate": 8.18979462282192e-05, "loss": 0.0623, "step": 31580 }, { "grad_norm": 0.43497392535209656, "learning_rate": 8.188521160276819e-05, "loss": 0.0596, "step": 31590 }, { "grad_norm": 0.47981444001197815, "learning_rate": 8.187247349039837e-05, "loss": 0.0563, "step": 31600 }, { "grad_norm": 0.4051361083984375, "learning_rate": 8.185973189250278e-05, "loss": 0.066, "step": 31610 }, { "grad_norm": 0.5417718887329102, "learning_rate": 8.184698681047482e-05, "loss": 0.069, "step": 31620 }, { "grad_norm": 0.611240804195404, "learning_rate": 8.183423824570827e-05, "loss": 0.0585, "step": 31630 }, { "grad_norm": 0.7296437621116638, "learning_rate": 8.182148619959725e-05, "loss": 0.0648, "step": 31640 }, { "grad_norm": 0.5919944643974304, "learning_rate": 8.180873067353636e-05, "loss": 0.062, "step": 31650 }, { "grad_norm": 0.3513795733451843, "learning_rate": 8.179597166892052e-05, "loss": 0.0591, "step": 31660 }, { "grad_norm": 0.4432789981365204, "learning_rate": 8.178320918714501e-05, "loss": 0.0605, "step": 31670 }, { "grad_norm": 0.8973275423049927, "learning_rate": 8.177044322960554e-05, "loss": 0.064, "step": 31680 }, { "grad_norm": 0.6147710084915161, "learning_rate": 8.175767379769816e-05, "loss": 0.0602, "step": 31690 }, { "grad_norm": 0.5598738193511963, "learning_rate": 8.174490089281932e-05, "loss": 0.0648, "step": 31700 }, { "grad_norm": 0.5183092355728149, "learning_rate": 8.173212451636584e-05, "loss": 0.0674, "step": 31710 }, { "grad_norm": 0.5757068991661072, "learning_rate": 8.171934466973493e-05, "loss": 0.0755, "step": 31720 }, { "grad_norm": 0.48832109570503235, "learning_rate": 8.170656135432418e-05, "loss": 0.0499, "step": 31730 }, { "grad_norm": 1.4850153923034668, "learning_rate": 8.169377457153155e-05, "loss": 0.0716, "step": 31740 }, { "grad_norm": 0.4682040810585022, "learning_rate": 8.168098432275539e-05, "loss": 0.0637, "step": 31750 }, { "grad_norm": 0.6227017641067505, "learning_rate": 8.166819060939442e-05, "loss": 0.065, "step": 31760 }, { "grad_norm": 0.8553740382194519, "learning_rate": 8.165539343284772e-05, "loss": 0.0733, "step": 31770 }, { "grad_norm": 0.42863932251930237, "learning_rate": 8.16425927945148e-05, "loss": 0.0668, "step": 31780 }, { "grad_norm": 0.6168850660324097, "learning_rate": 8.162978869579551e-05, "loss": 0.0603, "step": 31790 }, { "grad_norm": 0.4557853639125824, "learning_rate": 8.161698113809007e-05, "loss": 0.0504, "step": 31800 }, { "grad_norm": 0.48003530502319336, "learning_rate": 8.160417012279911e-05, "loss": 0.0664, "step": 31810 }, { "grad_norm": 0.44703763723373413, "learning_rate": 8.159135565132363e-05, "loss": 0.0627, "step": 31820 }, { "grad_norm": 0.5534792542457581, "learning_rate": 8.157853772506498e-05, "loss": 0.0739, "step": 31830 }, { "grad_norm": 0.5518645644187927, "learning_rate": 8.156571634542494e-05, "loss": 0.0555, "step": 31840 }, { "grad_norm": 1.137537956237793, "learning_rate": 8.15528915138056e-05, "loss": 0.0493, "step": 31850 }, { "grad_norm": 0.8176870942115784, "learning_rate": 8.154006323160949e-05, "loss": 0.061, "step": 31860 }, { "grad_norm": 0.5370773077011108, "learning_rate": 8.152723150023949e-05, "loss": 0.063, "step": 31870 }, { "grad_norm": 0.6565002799034119, "learning_rate": 8.151439632109886e-05, "loss": 0.0642, "step": 31880 }, { "grad_norm": 0.5519813895225525, "learning_rate": 8.150155769559122e-05, "loss": 0.0696, "step": 31890 }, { "grad_norm": 0.6446585059165955, "learning_rate": 8.148871562512058e-05, "loss": 0.0538, "step": 31900 }, { "grad_norm": 0.4375988245010376, "learning_rate": 8.147587011109136e-05, "loss": 0.0718, "step": 31910 }, { "grad_norm": 0.5348199605941772, "learning_rate": 8.14630211549083e-05, "loss": 0.076, "step": 31920 }, { "grad_norm": 0.4943409562110901, "learning_rate": 8.145016875797655e-05, "loss": 0.0478, "step": 31930 }, { "grad_norm": 0.4132174849510193, "learning_rate": 8.143731292170164e-05, "loss": 0.0638, "step": 31940 }, { "grad_norm": 0.5458995699882507, "learning_rate": 8.142445364748944e-05, "loss": 0.0627, "step": 31950 }, { "grad_norm": 0.5410832762718201, "learning_rate": 8.141159093674624e-05, "loss": 0.0687, "step": 31960 }, { "grad_norm": 0.5589916110038757, "learning_rate": 8.139872479087869e-05, "loss": 0.0539, "step": 31970 }, { "grad_norm": 0.37152203917503357, "learning_rate": 8.13858552112938e-05, "loss": 0.0454, "step": 31980 }, { "grad_norm": 0.5071442723274231, "learning_rate": 8.137298219939895e-05, "loss": 0.0709, "step": 31990 }, { "grad_norm": 0.546244204044342, "learning_rate": 8.136010575660196e-05, "loss": 0.0633, "step": 32000 }, { "grad_norm": 0.5315930247306824, "learning_rate": 8.134722588431095e-05, "loss": 0.0643, "step": 32010 }, { "grad_norm": 0.9215197563171387, "learning_rate": 8.133434258393444e-05, "loss": 0.0712, "step": 32020 }, { "grad_norm": 0.4644748568534851, "learning_rate": 8.132145585688134e-05, "loss": 0.0581, "step": 32030 }, { "grad_norm": 0.6146690845489502, "learning_rate": 8.130856570456093e-05, "loss": 0.0643, "step": 32040 }, { "grad_norm": 1.0678473711013794, "learning_rate": 8.129567212838283e-05, "loss": 0.0582, "step": 32050 }, { "grad_norm": 0.5378435850143433, "learning_rate": 8.128277512975708e-05, "loss": 0.0649, "step": 32060 }, { "grad_norm": 0.6570029854774475, "learning_rate": 8.126987471009408e-05, "loss": 0.0672, "step": 32070 }, { "grad_norm": 0.45306867361068726, "learning_rate": 8.125697087080459e-05, "loss": 0.0483, "step": 32080 }, { "grad_norm": 0.8132404685020447, "learning_rate": 8.124406361329976e-05, "loss": 0.0578, "step": 32090 }, { "grad_norm": 0.6118680238723755, "learning_rate": 8.123115293899111e-05, "loss": 0.0516, "step": 32100 }, { "grad_norm": 0.6025282740592957, "learning_rate": 8.121823884929055e-05, "loss": 0.0552, "step": 32110 }, { "grad_norm": 0.5604836344718933, "learning_rate": 8.12053213456103e-05, "loss": 0.0412, "step": 32120 }, { "grad_norm": 0.5155856013298035, "learning_rate": 8.119240042936303e-05, "loss": 0.0428, "step": 32130 }, { "grad_norm": 0.34456706047058105, "learning_rate": 8.117947610196175e-05, "loss": 0.0485, "step": 32140 }, { "grad_norm": 0.6545716524124146, "learning_rate": 8.116654836481982e-05, "loss": 0.0476, "step": 32150 }, { "grad_norm": 0.4451481103897095, "learning_rate": 8.115361721935106e-05, "loss": 0.0592, "step": 32160 }, { "grad_norm": 0.5511645078659058, "learning_rate": 8.114068266696953e-05, "loss": 0.0592, "step": 32170 }, { "grad_norm": 0.38683515787124634, "learning_rate": 8.112774470908978e-05, "loss": 0.0426, "step": 32180 }, { "grad_norm": 0.31009846925735474, "learning_rate": 8.111480334712665e-05, "loss": 0.0466, "step": 32190 }, { "grad_norm": 0.575729250907898, "learning_rate": 8.110185858249542e-05, "loss": 0.0488, "step": 32200 }, { "grad_norm": 0.6073375344276428, "learning_rate": 8.108891041661168e-05, "loss": 0.0505, "step": 32210 }, { "grad_norm": 0.4756520092487335, "learning_rate": 8.107595885089146e-05, "loss": 0.0712, "step": 32220 }, { "grad_norm": 0.5173146724700928, "learning_rate": 8.106300388675108e-05, "loss": 0.0583, "step": 32230 }, { "grad_norm": 0.5415176749229431, "learning_rate": 8.105004552560731e-05, "loss": 0.0642, "step": 32240 }, { "grad_norm": 0.6258095502853394, "learning_rate": 8.103708376887724e-05, "loss": 0.0593, "step": 32250 }, { "grad_norm": 0.69330894947052, "learning_rate": 8.102411861797836e-05, "loss": 0.0533, "step": 32260 }, { "grad_norm": 0.7802801728248596, "learning_rate": 8.101115007432851e-05, "loss": 0.0511, "step": 32270 }, { "grad_norm": 0.7373355031013489, "learning_rate": 8.09981781393459e-05, "loss": 0.0456, "step": 32280 }, { "grad_norm": 0.8578933477401733, "learning_rate": 8.098520281444915e-05, "loss": 0.0523, "step": 32290 }, { "grad_norm": 0.6478957533836365, "learning_rate": 8.09722241010572e-05, "loss": 0.0489, "step": 32300 }, { "grad_norm": 0.3691602945327759, "learning_rate": 8.095924200058939e-05, "loss": 0.0625, "step": 32310 }, { "grad_norm": 0.5086363554000854, "learning_rate": 8.094625651446541e-05, "loss": 0.0507, "step": 32320 }, { "grad_norm": 0.39169424772262573, "learning_rate": 8.093326764410536e-05, "loss": 0.0634, "step": 32330 }, { "grad_norm": 0.5032260417938232, "learning_rate": 8.092027539092966e-05, "loss": 0.063, "step": 32340 }, { "grad_norm": 0.7045883536338806, "learning_rate": 8.090727975635913e-05, "loss": 0.0598, "step": 32350 }, { "grad_norm": 0.8410192728042603, "learning_rate": 8.089428074181497e-05, "loss": 0.0601, "step": 32360 }, { "grad_norm": 0.532767653465271, "learning_rate": 8.088127834871871e-05, "loss": 0.046, "step": 32370 }, { "grad_norm": 1.3956830501556396, "learning_rate": 8.086827257849226e-05, "loss": 0.0641, "step": 32380 }, { "grad_norm": 0.39647185802459717, "learning_rate": 8.085526343255795e-05, "loss": 0.049, "step": 32390 }, { "grad_norm": 0.7134770750999451, "learning_rate": 8.084225091233842e-05, "loss": 0.065, "step": 32400 }, { "grad_norm": 0.6366854906082153, "learning_rate": 8.082923501925668e-05, "loss": 0.0473, "step": 32410 }, { "grad_norm": 0.6895820498466492, "learning_rate": 8.081621575473617e-05, "loss": 0.0616, "step": 32420 }, { "grad_norm": 0.4876752495765686, "learning_rate": 8.080319312020064e-05, "loss": 0.0601, "step": 32430 }, { "grad_norm": 0.337382048368454, "learning_rate": 8.079016711707421e-05, "loss": 0.0546, "step": 32440 }, { "grad_norm": 0.5309956669807434, "learning_rate": 8.077713774678139e-05, "loss": 0.0563, "step": 32450 }, { "grad_norm": 0.5416870713233948, "learning_rate": 8.076410501074707e-05, "loss": 0.0483, "step": 32460 }, { "grad_norm": 0.5253667831420898, "learning_rate": 8.075106891039647e-05, "loss": 0.0606, "step": 32470 }, { "grad_norm": 0.5467075705528259, "learning_rate": 8.073802944715523e-05, "loss": 0.0656, "step": 32480 }, { "grad_norm": 0.6856020092964172, "learning_rate": 8.072498662244929e-05, "loss": 0.0592, "step": 32490 }, { "grad_norm": 0.5752717852592468, "learning_rate": 8.0711940437705e-05, "loss": 0.057, "step": 32500 }, { "grad_norm": 0.7397072315216064, "learning_rate": 8.06988908943491e-05, "loss": 0.0584, "step": 32510 }, { "grad_norm": 0.4981459379196167, "learning_rate": 8.068583799380863e-05, "loss": 0.0513, "step": 32520 }, { "grad_norm": 0.6736944317817688, "learning_rate": 8.067278173751104e-05, "loss": 0.0537, "step": 32530 }, { "grad_norm": 0.6384661793708801, "learning_rate": 8.065972212688417e-05, "loss": 0.0586, "step": 32540 }, { "grad_norm": 0.6209923028945923, "learning_rate": 8.064665916335618e-05, "loss": 0.0643, "step": 32550 }, { "grad_norm": 0.4164469540119171, "learning_rate": 8.063359284835564e-05, "loss": 0.0629, "step": 32560 }, { "grad_norm": 0.4626109004020691, "learning_rate": 8.062052318331142e-05, "loss": 0.0651, "step": 32570 }, { "grad_norm": 0.5172839164733887, "learning_rate": 8.060745016965283e-05, "loss": 0.0595, "step": 32580 }, { "grad_norm": 0.38527706265449524, "learning_rate": 8.059437380880952e-05, "loss": 0.0541, "step": 32590 }, { "grad_norm": 0.5611371397972107, "learning_rate": 8.058129410221146e-05, "loss": 0.0484, "step": 32600 }, { "grad_norm": 0.6336182355880737, "learning_rate": 8.056821105128908e-05, "loss": 0.0589, "step": 32610 }, { "grad_norm": 0.4731268584728241, "learning_rate": 8.05551246574731e-05, "loss": 0.0511, "step": 32620 }, { "grad_norm": 0.5026045441627502, "learning_rate": 8.05420349221946e-05, "loss": 0.0551, "step": 32630 }, { "grad_norm": 0.7428796291351318, "learning_rate": 8.05289418468851e-05, "loss": 0.0405, "step": 32640 }, { "grad_norm": 0.4042642116546631, "learning_rate": 8.051584543297642e-05, "loss": 0.0511, "step": 32650 }, { "grad_norm": 0.5114787817001343, "learning_rate": 8.050274568190074e-05, "loss": 0.0604, "step": 32660 }, { "grad_norm": 0.6599262952804565, "learning_rate": 8.048964259509067e-05, "loss": 0.058, "step": 32670 }, { "grad_norm": 0.6107605695724487, "learning_rate": 8.047653617397914e-05, "loss": 0.061, "step": 32680 }, { "grad_norm": 0.5773916244506836, "learning_rate": 8.046342641999941e-05, "loss": 0.0509, "step": 32690 }, { "grad_norm": 0.4794400632381439, "learning_rate": 8.045031333458517e-05, "loss": 0.0518, "step": 32700 }, { "grad_norm": 0.7335516810417175, "learning_rate": 8.043719691917047e-05, "loss": 0.06, "step": 32710 }, { "grad_norm": 0.6169756054878235, "learning_rate": 8.042407717518966e-05, "loss": 0.0498, "step": 32720 }, { "grad_norm": 0.7023661732673645, "learning_rate": 8.041095410407751e-05, "loss": 0.056, "step": 32730 }, { "grad_norm": 0.5540162920951843, "learning_rate": 8.039782770726913e-05, "loss": 0.051, "step": 32740 }, { "grad_norm": 0.3431716859340668, "learning_rate": 8.038469798620004e-05, "loss": 0.0503, "step": 32750 }, { "grad_norm": 0.4560040235519409, "learning_rate": 8.037156494230604e-05, "loss": 0.0565, "step": 32760 }, { "grad_norm": 0.5948464274406433, "learning_rate": 8.035842857702338e-05, "loss": 0.0447, "step": 32770 }, { "grad_norm": 0.5005390644073486, "learning_rate": 8.03452888917886e-05, "loss": 0.0586, "step": 32780 }, { "grad_norm": 0.418530136346817, "learning_rate": 8.033214588803866e-05, "loss": 0.0515, "step": 32790 }, { "grad_norm": 0.3598281145095825, "learning_rate": 8.031899956721083e-05, "loss": 0.0499, "step": 32800 }, { "grad_norm": 0.5497947335243225, "learning_rate": 8.030584993074282e-05, "loss": 0.0542, "step": 32810 }, { "grad_norm": 0.42138439416885376, "learning_rate": 8.02926969800726e-05, "loss": 0.0556, "step": 32820 }, { "grad_norm": 0.39869454503059387, "learning_rate": 8.027954071663859e-05, "loss": 0.0567, "step": 32830 }, { "grad_norm": 0.6311778426170349, "learning_rate": 8.026638114187954e-05, "loss": 0.0534, "step": 32840 }, { "grad_norm": 0.5232261419296265, "learning_rate": 8.025321825723456e-05, "loss": 0.056, "step": 32850 }, { "grad_norm": 0.4361083209514618, "learning_rate": 8.02400520641431e-05, "loss": 0.0686, "step": 32860 }, { "grad_norm": 0.35251450538635254, "learning_rate": 8.022688256404501e-05, "loss": 0.0519, "step": 32870 }, { "grad_norm": 0.6114368438720703, "learning_rate": 8.02137097583805e-05, "loss": 0.0546, "step": 32880 }, { "grad_norm": 0.5790903568267822, "learning_rate": 8.02005336485901e-05, "loss": 0.0565, "step": 32890 }, { "grad_norm": 0.320106565952301, "learning_rate": 8.018735423611476e-05, "loss": 0.0702, "step": 32900 }, { "grad_norm": 1.1173374652862549, "learning_rate": 8.017417152239574e-05, "loss": 0.0513, "step": 32910 }, { "grad_norm": 0.90611332654953, "learning_rate": 8.01609855088747e-05, "loss": 0.0562, "step": 32920 }, { "grad_norm": 0.4881840944290161, "learning_rate": 8.014779619699362e-05, "loss": 0.0587, "step": 32930 }, { "grad_norm": 0.486138254404068, "learning_rate": 8.013460358819489e-05, "loss": 0.0637, "step": 32940 }, { "grad_norm": 0.6727988123893738, "learning_rate": 8.01214076839212e-05, "loss": 0.0639, "step": 32950 }, { "grad_norm": 0.41168689727783203, "learning_rate": 8.010820848561565e-05, "loss": 0.05, "step": 32960 }, { "grad_norm": 0.5009121298789978, "learning_rate": 8.009500599472171e-05, "loss": 0.0443, "step": 32970 }, { "grad_norm": 0.48327159881591797, "learning_rate": 8.008180021268314e-05, "loss": 0.0501, "step": 32980 }, { "grad_norm": 0.4672057032585144, "learning_rate": 8.006859114094414e-05, "loss": 0.0483, "step": 32990 }, { "grad_norm": 0.5240311622619629, "learning_rate": 8.005537878094921e-05, "loss": 0.0698, "step": 33000 }, { "grad_norm": 0.4615548253059387, "learning_rate": 8.004216313414323e-05, "loss": 0.0624, "step": 33010 }, { "grad_norm": 0.29334744811058044, "learning_rate": 8.002894420197149e-05, "loss": 0.0513, "step": 33020 }, { "grad_norm": 0.41258350014686584, "learning_rate": 8.001572198587954e-05, "loss": 0.0488, "step": 33030 }, { "grad_norm": 1.013415813446045, "learning_rate": 8.000249648731338e-05, "loss": 0.0559, "step": 33040 }, { "grad_norm": 0.5649740099906921, "learning_rate": 7.998926770771928e-05, "loss": 0.0587, "step": 33050 }, { "grad_norm": 0.4246481955051422, "learning_rate": 7.997603564854397e-05, "loss": 0.0653, "step": 33060 }, { "grad_norm": 0.5100321769714355, "learning_rate": 7.996280031123448e-05, "loss": 0.0478, "step": 33070 }, { "grad_norm": 0.41748547554016113, "learning_rate": 7.994956169723818e-05, "loss": 0.0605, "step": 33080 }, { "grad_norm": 0.5194527506828308, "learning_rate": 7.993631980800285e-05, "loss": 0.047, "step": 33090 }, { "grad_norm": 0.6065375208854675, "learning_rate": 7.992307464497659e-05, "loss": 0.0595, "step": 33100 }, { "grad_norm": 0.38654249906539917, "learning_rate": 7.990982620960787e-05, "loss": 0.0426, "step": 33110 }, { "grad_norm": 0.4412902295589447, "learning_rate": 7.989657450334554e-05, "loss": 0.0549, "step": 33120 }, { "grad_norm": 0.5587765574455261, "learning_rate": 7.988331952763877e-05, "loss": 0.0456, "step": 33130 }, { "grad_norm": 0.5423702001571655, "learning_rate": 7.987006128393709e-05, "loss": 0.05, "step": 33140 }, { "grad_norm": 0.625078558921814, "learning_rate": 7.985679977369043e-05, "loss": 0.0475, "step": 33150 }, { "grad_norm": 0.46207737922668457, "learning_rate": 7.984353499834902e-05, "loss": 0.0474, "step": 33160 }, { "grad_norm": 0.6377201080322266, "learning_rate": 7.983026695936351e-05, "loss": 0.0556, "step": 33170 }, { "grad_norm": 0.8067448735237122, "learning_rate": 7.981699565818486e-05, "loss": 0.048, "step": 33180 }, { "grad_norm": 0.6717074513435364, "learning_rate": 7.980372109626437e-05, "loss": 0.0548, "step": 33190 }, { "grad_norm": 0.5296837091445923, "learning_rate": 7.979044327505375e-05, "loss": 0.0467, "step": 33200 }, { "grad_norm": 0.5378164052963257, "learning_rate": 7.977716219600506e-05, "loss": 0.0665, "step": 33210 }, { "grad_norm": 0.7176183462142944, "learning_rate": 7.97638778605707e-05, "loss": 0.0516, "step": 33220 }, { "grad_norm": 0.983171820640564, "learning_rate": 7.975059027020338e-05, "loss": 0.0718, "step": 33230 }, { "grad_norm": 0.6838605999946594, "learning_rate": 7.973729942635623e-05, "loss": 0.054, "step": 33240 }, { "grad_norm": 0.43385934829711914, "learning_rate": 7.972400533048273e-05, "loss": 0.0558, "step": 33250 }, { "grad_norm": 0.5761569738388062, "learning_rate": 7.97107079840367e-05, "loss": 0.0572, "step": 33260 }, { "grad_norm": 1.2641273736953735, "learning_rate": 7.969740738847231e-05, "loss": 0.0546, "step": 33270 }, { "grad_norm": 0.6123698949813843, "learning_rate": 7.968410354524411e-05, "loss": 0.0557, "step": 33280 }, { "grad_norm": 0.3832891285419464, "learning_rate": 7.967079645580697e-05, "loss": 0.0439, "step": 33290 }, { "grad_norm": 0.4946748912334442, "learning_rate": 7.965748612161612e-05, "loss": 0.0552, "step": 33300 }, { "grad_norm": 0.5298900008201599, "learning_rate": 7.96441725441272e-05, "loss": 0.073, "step": 33310 }, { "grad_norm": 0.673668622970581, "learning_rate": 7.963085572479614e-05, "loss": 0.0622, "step": 33320 }, { "grad_norm": 0.4515853226184845, "learning_rate": 7.961753566507924e-05, "loss": 0.0493, "step": 33330 }, { "grad_norm": 0.5672628879547119, "learning_rate": 7.960421236643316e-05, "loss": 0.0617, "step": 33340 }, { "grad_norm": 0.6577356457710266, "learning_rate": 7.959088583031496e-05, "loss": 0.0496, "step": 33350 }, { "grad_norm": 0.3697133958339691, "learning_rate": 7.957755605818194e-05, "loss": 0.0493, "step": 33360 }, { "grad_norm": 0.6080673336982727, "learning_rate": 7.956422305149185e-05, "loss": 0.0675, "step": 33370 }, { "grad_norm": 0.6742610335350037, "learning_rate": 7.95508868117028e-05, "loss": 0.0604, "step": 33380 }, { "grad_norm": 0.5318353772163391, "learning_rate": 7.953754734027318e-05, "loss": 0.0551, "step": 33390 }, { "grad_norm": 0.3566771149635315, "learning_rate": 7.952420463866182e-05, "loss": 0.0525, "step": 33400 }, { "grad_norm": 0.7921951413154602, "learning_rate": 7.951085870832782e-05, "loss": 0.0673, "step": 33410 }, { "grad_norm": 0.6229443550109863, "learning_rate": 7.949750955073067e-05, "loss": 0.054, "step": 33420 }, { "grad_norm": 0.7318341135978699, "learning_rate": 7.948415716733022e-05, "loss": 0.0647, "step": 33430 }, { "grad_norm": 0.5061448812484741, "learning_rate": 7.947080155958669e-05, "loss": 0.0599, "step": 33440 }, { "grad_norm": 0.5115580558776855, "learning_rate": 7.94574427289606e-05, "loss": 0.0501, "step": 33450 }, { "grad_norm": 0.636763334274292, "learning_rate": 7.944408067691284e-05, "loss": 0.0553, "step": 33460 }, { "grad_norm": 0.5407623052597046, "learning_rate": 7.943071540490473e-05, "loss": 0.0536, "step": 33470 }, { "grad_norm": 0.5656747817993164, "learning_rate": 7.94173469143978e-05, "loss": 0.0625, "step": 33480 }, { "grad_norm": 0.3554905951023102, "learning_rate": 7.940397520685406e-05, "loss": 0.0568, "step": 33490 }, { "grad_norm": 0.754442036151886, "learning_rate": 7.939060028373577e-05, "loss": 0.0707, "step": 33500 }, { "grad_norm": 0.7458467483520508, "learning_rate": 7.937722214650565e-05, "loss": 0.0755, "step": 33510 }, { "grad_norm": 0.40745413303375244, "learning_rate": 7.936384079662666e-05, "loss": 0.0613, "step": 33520 }, { "grad_norm": 0.49526283144950867, "learning_rate": 7.93504562355622e-05, "loss": 0.0565, "step": 33530 }, { "grad_norm": 0.7678223252296448, "learning_rate": 7.933706846477599e-05, "loss": 0.0719, "step": 33540 }, { "grad_norm": 0.6205379962921143, "learning_rate": 7.932367748573206e-05, "loss": 0.0619, "step": 33550 }, { "grad_norm": 0.3889004588127136, "learning_rate": 7.931028329989485e-05, "loss": 0.0608, "step": 33560 }, { "grad_norm": 0.611573338508606, "learning_rate": 7.929688590872913e-05, "loss": 0.061, "step": 33570 }, { "grad_norm": 0.3019229471683502, "learning_rate": 7.928348531370003e-05, "loss": 0.0379, "step": 33580 }, { "grad_norm": 0.45700809359550476, "learning_rate": 7.927008151627297e-05, "loss": 0.0662, "step": 33590 }, { "grad_norm": 0.44447994232177734, "learning_rate": 7.925667451791383e-05, "loss": 0.0546, "step": 33600 }, { "grad_norm": 0.6367312073707581, "learning_rate": 7.924326432008874e-05, "loss": 0.0598, "step": 33610 }, { "grad_norm": 0.43855708837509155, "learning_rate": 7.922985092426422e-05, "loss": 0.0608, "step": 33620 }, { "grad_norm": 0.4845365583896637, "learning_rate": 7.921643433190717e-05, "loss": 0.0626, "step": 33630 }, { "grad_norm": 0.5196780562400818, "learning_rate": 7.920301454448478e-05, "loss": 0.0743, "step": 33640 }, { "grad_norm": 0.3590977191925049, "learning_rate": 7.918959156346461e-05, "loss": 0.0478, "step": 33650 }, { "grad_norm": 0.5404797196388245, "learning_rate": 7.91761653903146e-05, "loss": 0.0508, "step": 33660 }, { "grad_norm": 0.6281169652938843, "learning_rate": 7.916273602650302e-05, "loss": 0.0515, "step": 33670 }, { "grad_norm": 0.42436137795448303, "learning_rate": 7.914930347349847e-05, "loss": 0.0483, "step": 33680 }, { "grad_norm": 0.6051806211471558, "learning_rate": 7.913586773276992e-05, "loss": 0.0577, "step": 33690 }, { "grad_norm": 0.561359167098999, "learning_rate": 7.912242880578667e-05, "loss": 0.0465, "step": 33700 }, { "grad_norm": 0.6133887767791748, "learning_rate": 7.910898669401839e-05, "loss": 0.0503, "step": 33710 }, { "grad_norm": 0.4381754398345947, "learning_rate": 7.909554139893511e-05, "loss": 0.0489, "step": 33720 }, { "grad_norm": 0.3299892246723175, "learning_rate": 7.908209292200715e-05, "loss": 0.041, "step": 33730 }, { "grad_norm": 0.6358283162117004, "learning_rate": 7.906864126470523e-05, "loss": 0.0471, "step": 33740 }, { "grad_norm": 1.0330579280853271, "learning_rate": 7.905518642850041e-05, "loss": 0.0485, "step": 33750 }, { "grad_norm": 0.5132498145103455, "learning_rate": 7.904172841486409e-05, "loss": 0.0565, "step": 33760 }, { "grad_norm": 0.6998270750045776, "learning_rate": 7.902826722526801e-05, "loss": 0.0536, "step": 33770 }, { "grad_norm": 0.34727150201797485, "learning_rate": 7.901480286118427e-05, "loss": 0.0526, "step": 33780 }, { "grad_norm": 0.3947597146034241, "learning_rate": 7.900133532408531e-05, "loss": 0.0465, "step": 33790 }, { "grad_norm": 0.626898467540741, "learning_rate": 7.898786461544395e-05, "loss": 0.0521, "step": 33800 }, { "grad_norm": 0.714269757270813, "learning_rate": 7.897439073673325e-05, "loss": 0.0665, "step": 33810 }, { "grad_norm": 0.4134159982204437, "learning_rate": 7.896091368942677e-05, "loss": 0.0639, "step": 33820 }, { "grad_norm": 0.8179986476898193, "learning_rate": 7.894743347499832e-05, "loss": 0.0568, "step": 33830 }, { "grad_norm": 0.3963102400302887, "learning_rate": 7.893395009492203e-05, "loss": 0.0584, "step": 33840 }, { "grad_norm": 0.408935010433197, "learning_rate": 7.892046355067248e-05, "loss": 0.0545, "step": 33850 }, { "grad_norm": 0.8870239853858948, "learning_rate": 7.890697384372451e-05, "loss": 0.0604, "step": 33860 }, { "grad_norm": 0.4390539228916168, "learning_rate": 7.889348097555336e-05, "loss": 0.042, "step": 33870 }, { "grad_norm": 0.5974496006965637, "learning_rate": 7.887998494763455e-05, "loss": 0.0556, "step": 33880 }, { "grad_norm": 0.4869737923145294, "learning_rate": 7.886648576144404e-05, "loss": 0.0551, "step": 33890 }, { "grad_norm": 0.5880168080329895, "learning_rate": 7.885298341845802e-05, "loss": 0.0568, "step": 33900 }, { "grad_norm": 0.6369588971138, "learning_rate": 7.883947792015311e-05, "loss": 0.0532, "step": 33910 }, { "grad_norm": 0.7146074175834656, "learning_rate": 7.882596926800628e-05, "loss": 0.0794, "step": 33920 }, { "grad_norm": 0.4884423613548279, "learning_rate": 7.881245746349477e-05, "loss": 0.066, "step": 33930 }, { "grad_norm": 0.41868579387664795, "learning_rate": 7.879894250809623e-05, "loss": 0.0545, "step": 33940 }, { "grad_norm": 0.5308130979537964, "learning_rate": 7.878542440328865e-05, "loss": 0.0708, "step": 33950 }, { "grad_norm": 0.4855727255344391, "learning_rate": 7.877190315055031e-05, "loss": 0.0438, "step": 33960 }, { "grad_norm": 0.54804527759552, "learning_rate": 7.875837875135991e-05, "loss": 0.066, "step": 33970 }, { "grad_norm": 0.6979466676712036, "learning_rate": 7.874485120719646e-05, "loss": 0.0627, "step": 33980 }, { "grad_norm": 0.32896122336387634, "learning_rate": 7.873132051953928e-05, "loss": 0.0551, "step": 33990 }, { "grad_norm": 0.3909085988998413, "learning_rate": 7.87177866898681e-05, "loss": 0.0557, "step": 34000 }, { "grad_norm": 0.2958482801914215, "learning_rate": 7.870424971966294e-05, "loss": 0.0574, "step": 34010 }, { "grad_norm": 1.077874779701233, "learning_rate": 7.869070961040419e-05, "loss": 0.0444, "step": 34020 }, { "grad_norm": 0.602302074432373, "learning_rate": 7.867716636357257e-05, "loss": 0.0547, "step": 34030 }, { "grad_norm": 0.6620128154754639, "learning_rate": 7.866361998064915e-05, "loss": 0.0601, "step": 34040 }, { "grad_norm": 0.41694679856300354, "learning_rate": 7.865007046311534e-05, "loss": 0.0482, "step": 34050 }, { "grad_norm": 0.5194876790046692, "learning_rate": 7.86365178124529e-05, "loss": 0.0523, "step": 34060 }, { "grad_norm": 0.5275111794471741, "learning_rate": 7.862296203014394e-05, "loss": 0.0518, "step": 34070 }, { "grad_norm": 0.4731702506542206, "learning_rate": 7.860940311767088e-05, "loss": 0.0508, "step": 34080 }, { "grad_norm": 0.4840991199016571, "learning_rate": 7.85958410765165e-05, "loss": 0.0508, "step": 34090 }, { "grad_norm": 0.5626075863838196, "learning_rate": 7.858227590816394e-05, "loss": 0.0567, "step": 34100 }, { "grad_norm": 0.4244909882545471, "learning_rate": 7.856870761409664e-05, "loss": 0.0617, "step": 34110 }, { "grad_norm": 0.6206321120262146, "learning_rate": 7.855513619579846e-05, "loss": 0.0557, "step": 34120 }, { "grad_norm": 0.49238666892051697, "learning_rate": 7.85415616547535e-05, "loss": 0.0644, "step": 34130 }, { "grad_norm": 0.434863418340683, "learning_rate": 7.852798399244627e-05, "loss": 0.0432, "step": 34140 }, { "grad_norm": 0.523005485534668, "learning_rate": 7.851440321036161e-05, "loss": 0.0509, "step": 34150 }, { "grad_norm": 0.5055211782455444, "learning_rate": 7.850081930998468e-05, "loss": 0.0581, "step": 34160 }, { "grad_norm": 0.5674814581871033, "learning_rate": 7.8487232292801e-05, "loss": 0.0598, "step": 34170 }, { "grad_norm": 0.542902946472168, "learning_rate": 7.847364216029642e-05, "loss": 0.0653, "step": 34180 }, { "grad_norm": 0.5978408455848694, "learning_rate": 7.846004891395716e-05, "loss": 0.0531, "step": 34190 }, { "grad_norm": 0.6010017991065979, "learning_rate": 7.844645255526972e-05, "loss": 0.0515, "step": 34200 }, { "grad_norm": 0.4054762125015259, "learning_rate": 7.843285308572101e-05, "loss": 0.0532, "step": 34210 }, { "grad_norm": 0.5167464017868042, "learning_rate": 7.841925050679823e-05, "loss": 0.0419, "step": 34220 }, { "grad_norm": 0.6300672888755798, "learning_rate": 7.840564481998895e-05, "loss": 0.0556, "step": 34230 }, { "grad_norm": 0.5813959836959839, "learning_rate": 7.839203602678105e-05, "loss": 0.0584, "step": 34240 }, { "grad_norm": 0.569561243057251, "learning_rate": 7.837842412866279e-05, "loss": 0.0655, "step": 34250 }, { "grad_norm": 0.5865454077720642, "learning_rate": 7.836480912712272e-05, "loss": 0.0617, "step": 34260 }, { "grad_norm": 0.662101686000824, "learning_rate": 7.835119102364976e-05, "loss": 0.0646, "step": 34270 }, { "grad_norm": 0.4589686691761017, "learning_rate": 7.83375698197332e-05, "loss": 0.0544, "step": 34280 }, { "grad_norm": 0.46954792737960815, "learning_rate": 7.83239455168626e-05, "loss": 0.0542, "step": 34290 }, { "grad_norm": 0.5208432674407959, "learning_rate": 7.83103181165279e-05, "loss": 0.0587, "step": 34300 }, { "grad_norm": 0.4916101396083832, "learning_rate": 7.829668762021937e-05, "loss": 0.0627, "step": 34310 }, { "grad_norm": 0.6287617683410645, "learning_rate": 7.828305402942764e-05, "loss": 0.0519, "step": 34320 }, { "grad_norm": 1.1326286792755127, "learning_rate": 7.826941734564363e-05, "loss": 0.0474, "step": 34330 }, { "grad_norm": 0.49279195070266724, "learning_rate": 7.825577757035865e-05, "loss": 0.0493, "step": 34340 }, { "grad_norm": 0.32560858130455017, "learning_rate": 7.824213470506431e-05, "loss": 0.0485, "step": 34350 }, { "grad_norm": 0.7542957663536072, "learning_rate": 7.822848875125257e-05, "loss": 0.0547, "step": 34360 }, { "grad_norm": 0.5448285937309265, "learning_rate": 7.821483971041576e-05, "loss": 0.0696, "step": 34370 }, { "grad_norm": 2.1555697917938232, "learning_rate": 7.820118758404649e-05, "loss": 0.0562, "step": 34380 }, { "grad_norm": 0.6478011608123779, "learning_rate": 7.818753237363776e-05, "loss": 0.0598, "step": 34390 }, { "grad_norm": 0.5765230059623718, "learning_rate": 7.817387408068286e-05, "loss": 0.062, "step": 34400 }, { "grad_norm": 0.47872328758239746, "learning_rate": 7.816021270667544e-05, "loss": 0.0611, "step": 34410 }, { "grad_norm": 0.546735942363739, "learning_rate": 7.81465482531095e-05, "loss": 0.0597, "step": 34420 }, { "grad_norm": 0.6030944585800171, "learning_rate": 7.813288072147938e-05, "loss": 0.0701, "step": 34430 }, { "grad_norm": 0.6260472536087036, "learning_rate": 7.811921011327972e-05, "loss": 0.0688, "step": 34440 }, { "grad_norm": 0.3326903283596039, "learning_rate": 7.810553643000549e-05, "loss": 0.051, "step": 34450 }, { "grad_norm": 0.7531945109367371, "learning_rate": 7.809185967315206e-05, "loss": 0.0561, "step": 34460 }, { "grad_norm": 0.5582968592643738, "learning_rate": 7.80781798442151e-05, "loss": 0.0575, "step": 34470 }, { "grad_norm": 1.302431344985962, "learning_rate": 7.806449694469061e-05, "loss": 0.0475, "step": 34480 }, { "grad_norm": 0.5225622653961182, "learning_rate": 7.805081097607492e-05, "loss": 0.0648, "step": 34490 }, { "grad_norm": 0.48296239972114563, "learning_rate": 7.803712193986474e-05, "loss": 0.0582, "step": 34500 }, { "grad_norm": 0.6077894568443298, "learning_rate": 7.802342983755702e-05, "loss": 0.0581, "step": 34510 }, { "grad_norm": 0.7401895523071289, "learning_rate": 7.800973467064918e-05, "loss": 0.0622, "step": 34520 }, { "grad_norm": 0.5325337648391724, "learning_rate": 7.799603644063884e-05, "loss": 0.0562, "step": 34530 }, { "grad_norm": 0.4725465178489685, "learning_rate": 7.798233514902405e-05, "loss": 0.0501, "step": 34540 }, { "grad_norm": 0.2982257008552551, "learning_rate": 7.796863079730318e-05, "loss": 0.0464, "step": 34550 }, { "grad_norm": 0.36147168278694153, "learning_rate": 7.795492338697488e-05, "loss": 0.0486, "step": 34560 }, { "grad_norm": 0.49240440130233765, "learning_rate": 7.794121291953819e-05, "loss": 0.0654, "step": 34570 }, { "grad_norm": 0.5191237330436707, "learning_rate": 7.792749939649246e-05, "loss": 0.0616, "step": 34580 }, { "grad_norm": 0.393393874168396, "learning_rate": 7.79137828193374e-05, "loss": 0.0405, "step": 34590 }, { "grad_norm": 0.7201977372169495, "learning_rate": 7.790006318957301e-05, "loss": 0.0655, "step": 34600 }, { "grad_norm": 0.4312441647052765, "learning_rate": 7.788634050869965e-05, "loss": 0.051, "step": 34610 }, { "grad_norm": 0.5545254945755005, "learning_rate": 7.787261477821803e-05, "loss": 0.057, "step": 34620 }, { "grad_norm": 0.5752192735671997, "learning_rate": 7.785888599962916e-05, "loss": 0.0518, "step": 34630 }, { "grad_norm": 0.807388424873352, "learning_rate": 7.784515417443439e-05, "loss": 0.0511, "step": 34640 }, { "grad_norm": 0.44958165287971497, "learning_rate": 7.783141930413545e-05, "loss": 0.0518, "step": 34650 }, { "grad_norm": 0.48540815711021423, "learning_rate": 7.78176813902343e-05, "loss": 0.0556, "step": 34660 }, { "grad_norm": 0.511771023273468, "learning_rate": 7.780394043423336e-05, "loss": 0.0624, "step": 34670 }, { "grad_norm": 0.6768555641174316, "learning_rate": 7.77901964376353e-05, "loss": 0.0677, "step": 34680 }, { "grad_norm": 0.5554772019386292, "learning_rate": 7.777644940194316e-05, "loss": 0.0536, "step": 34690 }, { "grad_norm": 0.5681753754615784, "learning_rate": 7.776269932866023e-05, "loss": 0.0551, "step": 34700 }, { "grad_norm": 0.3576594889163971, "learning_rate": 7.774894621929026e-05, "loss": 0.0504, "step": 34710 }, { "grad_norm": 0.4043031632900238, "learning_rate": 7.773519007533725e-05, "loss": 0.0632, "step": 34720 }, { "grad_norm": 0.5250842571258545, "learning_rate": 7.772143089830556e-05, "loss": 0.0491, "step": 34730 }, { "grad_norm": 0.3973662853240967, "learning_rate": 7.770766868969985e-05, "loss": 0.0589, "step": 34740 }, { "grad_norm": 0.3779512643814087, "learning_rate": 7.769390345102518e-05, "loss": 0.0594, "step": 34750 }, { "grad_norm": 0.6267340183258057, "learning_rate": 7.768013518378683e-05, "loss": 0.0533, "step": 34760 }, { "grad_norm": 0.41145744919776917, "learning_rate": 7.766636388949053e-05, "loss": 0.0443, "step": 34770 }, { "grad_norm": 0.33680033683776855, "learning_rate": 7.765258956964229e-05, "loss": 0.0695, "step": 34780 }, { "grad_norm": 0.7299748063087463, "learning_rate": 7.76388122257484e-05, "loss": 0.0612, "step": 34790 }, { "grad_norm": 0.3692269027233124, "learning_rate": 7.762503185931558e-05, "loss": 0.0506, "step": 34800 }, { "grad_norm": 0.4057476222515106, "learning_rate": 7.76112484718508e-05, "loss": 0.0546, "step": 34810 }, { "grad_norm": 0.4681050777435303, "learning_rate": 7.75974620648614e-05, "loss": 0.0625, "step": 34820 }, { "grad_norm": 0.4789089262485504, "learning_rate": 7.758367263985503e-05, "loss": 0.0454, "step": 34830 }, { "grad_norm": 0.597967803478241, "learning_rate": 7.75698801983397e-05, "loss": 0.0504, "step": 34840 }, { "grad_norm": 0.5186925530433655, "learning_rate": 7.755608474182372e-05, "loss": 0.0565, "step": 34850 }, { "grad_norm": 0.7059793472290039, "learning_rate": 7.754228627181574e-05, "loss": 0.0586, "step": 34860 }, { "grad_norm": 0.4183121621608734, "learning_rate": 7.752848478982476e-05, "loss": 0.0535, "step": 34870 }, { "grad_norm": 0.5038336515426636, "learning_rate": 7.751468029736006e-05, "loss": 0.0544, "step": 34880 }, { "grad_norm": 0.618839681148529, "learning_rate": 7.750087279593129e-05, "loss": 0.0588, "step": 34890 }, { "grad_norm": 0.5226964950561523, "learning_rate": 7.748706228704843e-05, "loss": 0.0535, "step": 34900 }, { "grad_norm": 0.5714001059532166, "learning_rate": 7.747324877222176e-05, "loss": 0.0562, "step": 34910 }, { "grad_norm": 0.5023396611213684, "learning_rate": 7.745943225296188e-05, "loss": 0.0726, "step": 34920 }, { "grad_norm": 0.6897778511047363, "learning_rate": 7.744561273077981e-05, "loss": 0.0427, "step": 34930 }, { "grad_norm": 0.3638670742511749, "learning_rate": 7.743179020718678e-05, "loss": 0.0487, "step": 34940 }, { "grad_norm": 0.495978981256485, "learning_rate": 7.741796468369443e-05, "loss": 0.0533, "step": 34950 }, { "grad_norm": 0.6167340874671936, "learning_rate": 7.740413616181466e-05, "loss": 0.0636, "step": 34960 }, { "grad_norm": 0.5201922059059143, "learning_rate": 7.739030464305978e-05, "loss": 0.0537, "step": 34970 }, { "grad_norm": 0.9670349955558777, "learning_rate": 7.737647012894235e-05, "loss": 0.0616, "step": 34980 }, { "grad_norm": 0.5608505010604858, "learning_rate": 7.736263262097532e-05, "loss": 0.062, "step": 34990 }, { "grad_norm": 0.5125638246536255, "learning_rate": 7.734879212067192e-05, "loss": 0.0547, "step": 35000 }, { "grad_norm": 0.37535953521728516, "learning_rate": 7.733494862954573e-05, "loss": 0.0557, "step": 35010 }, { "grad_norm": 0.4473631680011749, "learning_rate": 7.732110214911066e-05, "loss": 0.0497, "step": 35020 }, { "grad_norm": 0.7377324104309082, "learning_rate": 7.730725268088094e-05, "loss": 0.0644, "step": 35030 }, { "grad_norm": 0.6033360958099365, "learning_rate": 7.729340022637112e-05, "loss": 0.0508, "step": 35040 }, { "grad_norm": 0.44046372175216675, "learning_rate": 7.727954478709607e-05, "loss": 0.054, "step": 35050 }, { "grad_norm": 0.6981978416442871, "learning_rate": 7.726568636457103e-05, "loss": 0.0581, "step": 35060 }, { "grad_norm": 0.6432263255119324, "learning_rate": 7.725182496031153e-05, "loss": 0.067, "step": 35070 }, { "grad_norm": 0.5714514255523682, "learning_rate": 7.723796057583341e-05, "loss": 0.0749, "step": 35080 }, { "grad_norm": 0.4803330600261688, "learning_rate": 7.722409321265287e-05, "loss": 0.06, "step": 35090 }, { "grad_norm": 0.4768200218677521, "learning_rate": 7.721022287228645e-05, "loss": 0.0622, "step": 35100 }, { "grad_norm": 0.548204243183136, "learning_rate": 7.719634955625094e-05, "loss": 0.05, "step": 35110 }, { "grad_norm": 0.497917503118515, "learning_rate": 7.718247326606355e-05, "loss": 0.0402, "step": 35120 }, { "grad_norm": 0.4588577449321747, "learning_rate": 7.716859400324175e-05, "loss": 0.0476, "step": 35130 }, { "grad_norm": 0.5698361396789551, "learning_rate": 7.715471176930335e-05, "loss": 0.0807, "step": 35140 }, { "grad_norm": 0.48911744356155396, "learning_rate": 7.714082656576651e-05, "loss": 0.0455, "step": 35150 }, { "grad_norm": 0.642151415348053, "learning_rate": 7.712693839414968e-05, "loss": 0.0536, "step": 35160 }, { "grad_norm": 0.5833160877227783, "learning_rate": 7.711304725597164e-05, "loss": 0.0499, "step": 35170 }, { "grad_norm": 0.45099857449531555, "learning_rate": 7.709915315275151e-05, "loss": 0.0571, "step": 35180 }, { "grad_norm": 0.4645712375640869, "learning_rate": 7.708525608600876e-05, "loss": 0.0495, "step": 35190 }, { "grad_norm": 0.4604650139808655, "learning_rate": 7.707135605726311e-05, "loss": 0.0574, "step": 35200 }, { "grad_norm": 0.37219688296318054, "learning_rate": 7.705745306803466e-05, "loss": 0.08, "step": 35210 }, { "grad_norm": 0.6254631876945496, "learning_rate": 7.704354711984383e-05, "loss": 0.045, "step": 35220 }, { "grad_norm": 0.4204258918762207, "learning_rate": 7.702963821421133e-05, "loss": 0.0445, "step": 35230 }, { "grad_norm": 0.6001231670379639, "learning_rate": 7.701572635265826e-05, "loss": 0.0522, "step": 35240 }, { "grad_norm": 1.0541974306106567, "learning_rate": 7.700181153670596e-05, "loss": 0.0457, "step": 35250 }, { "grad_norm": 0.5326266884803772, "learning_rate": 7.698789376787616e-05, "loss": 0.06, "step": 35260 }, { "grad_norm": 1.791809320449829, "learning_rate": 7.697397304769085e-05, "loss": 0.0585, "step": 35270 }, { "grad_norm": 0.5947899222373962, "learning_rate": 7.696004937767241e-05, "loss": 0.0643, "step": 35280 }, { "grad_norm": 0.7496709227561951, "learning_rate": 7.694612275934352e-05, "loss": 0.0534, "step": 35290 }, { "grad_norm": 0.3468971848487854, "learning_rate": 7.693219319422714e-05, "loss": 0.0595, "step": 35300 }, { "grad_norm": 0.5024220943450928, "learning_rate": 7.69182606838466e-05, "loss": 0.0601, "step": 35310 }, { "grad_norm": 0.5093151330947876, "learning_rate": 7.690432522972558e-05, "loss": 0.0612, "step": 35320 }, { "grad_norm": 0.48454391956329346, "learning_rate": 7.689038683338796e-05, "loss": 0.0542, "step": 35330 }, { "grad_norm": 0.6141108274459839, "learning_rate": 7.687644549635808e-05, "loss": 0.0502, "step": 35340 }, { "grad_norm": 0.5871599316596985, "learning_rate": 7.686250122016053e-05, "loss": 0.0451, "step": 35350 }, { "grad_norm": 0.5968740582466125, "learning_rate": 7.684855400632022e-05, "loss": 0.0574, "step": 35360 }, { "grad_norm": 0.4241528809070587, "learning_rate": 7.683460385636243e-05, "loss": 0.0407, "step": 35370 }, { "grad_norm": 0.41504967212677, "learning_rate": 7.68206507718127e-05, "loss": 0.0489, "step": 35380 }, { "grad_norm": 0.6725618243217468, "learning_rate": 7.680669475419693e-05, "loss": 0.0652, "step": 35390 }, { "grad_norm": 0.5624213814735413, "learning_rate": 7.679273580504132e-05, "loss": 0.0437, "step": 35400 }, { "grad_norm": 0.39574119448661804, "learning_rate": 7.67787739258724e-05, "loss": 0.0471, "step": 35410 }, { "grad_norm": 0.40935325622558594, "learning_rate": 7.676480911821705e-05, "loss": 0.0597, "step": 35420 }, { "grad_norm": 0.7549253106117249, "learning_rate": 7.675084138360238e-05, "loss": 0.0547, "step": 35430 }, { "grad_norm": 0.542364776134491, "learning_rate": 7.673687072355592e-05, "loss": 0.0561, "step": 35440 }, { "grad_norm": 0.5490579009056091, "learning_rate": 7.67228971396055e-05, "loss": 0.0453, "step": 35450 }, { "grad_norm": 0.4969443380832672, "learning_rate": 7.670892063327922e-05, "loss": 0.0486, "step": 35460 }, { "grad_norm": 0.35368868708610535, "learning_rate": 7.669494120610552e-05, "loss": 0.0409, "step": 35470 }, { "grad_norm": 0.3255596160888672, "learning_rate": 7.66809588596132e-05, "loss": 0.0439, "step": 35480 }, { "grad_norm": 1.9078178405761719, "learning_rate": 7.666697359533132e-05, "loss": 0.0548, "step": 35490 }, { "grad_norm": 0.7222241163253784, "learning_rate": 7.665298541478932e-05, "loss": 0.0526, "step": 35500 }, { "grad_norm": 0.3785415291786194, "learning_rate": 7.66389943195169e-05, "loss": 0.0529, "step": 35510 }, { "grad_norm": 0.5124257802963257, "learning_rate": 7.66250003110441e-05, "loss": 0.0567, "step": 35520 }, { "grad_norm": 0.46611836552619934, "learning_rate": 7.66110033909013e-05, "loss": 0.0551, "step": 35530 }, { "grad_norm": 0.5906326174736023, "learning_rate": 7.659700356061918e-05, "loss": 0.0559, "step": 35540 }, { "grad_norm": 0.5665162801742554, "learning_rate": 7.658300082172875e-05, "loss": 0.0466, "step": 35550 }, { "grad_norm": 0.5139150619506836, "learning_rate": 7.65689951757613e-05, "loss": 0.0506, "step": 35560 }, { "grad_norm": 0.4567998945713043, "learning_rate": 7.655498662424849e-05, "loss": 0.0523, "step": 35570 }, { "grad_norm": 0.8549838662147522, "learning_rate": 7.654097516872227e-05, "loss": 0.051, "step": 35580 }, { "grad_norm": 0.6899686455726624, "learning_rate": 7.65269608107149e-05, "loss": 0.0586, "step": 35590 }, { "grad_norm": 0.4236510694026947, "learning_rate": 7.6512943551759e-05, "loss": 0.0669, "step": 35600 }, { "grad_norm": 0.6085930466651917, "learning_rate": 7.649892339338743e-05, "loss": 0.0476, "step": 35610 }, { "grad_norm": 1.5856324434280396, "learning_rate": 7.648490033713344e-05, "loss": 0.0367, "step": 35620 }, { "grad_norm": 0.5296205878257751, "learning_rate": 7.647087438453058e-05, "loss": 0.0531, "step": 35630 }, { "grad_norm": 0.48682650923728943, "learning_rate": 7.64568455371127e-05, "loss": 0.0582, "step": 35640 }, { "grad_norm": 0.4976653456687927, "learning_rate": 7.644281379641396e-05, "loss": 0.0424, "step": 35650 }, { "grad_norm": 0.43104612827301025, "learning_rate": 7.642877916396887e-05, "loss": 0.0451, "step": 35660 }, { "grad_norm": 0.44655609130859375, "learning_rate": 7.641474164131221e-05, "loss": 0.0554, "step": 35670 }, { "grad_norm": 0.5257101655006409, "learning_rate": 7.640070122997913e-05, "loss": 0.042, "step": 35680 }, { "grad_norm": 0.6597589254379272, "learning_rate": 7.638665793150505e-05, "loss": 0.0376, "step": 35690 }, { "grad_norm": 0.4800916612148285, "learning_rate": 7.637261174742574e-05, "loss": 0.046, "step": 35700 }, { "grad_norm": 0.427717000246048, "learning_rate": 7.635856267927725e-05, "loss": 0.0405, "step": 35710 }, { "grad_norm": 0.4110257625579834, "learning_rate": 7.634451072859597e-05, "loss": 0.0482, "step": 35720 }, { "grad_norm": 0.4800284504890442, "learning_rate": 7.633045589691863e-05, "loss": 0.0396, "step": 35730 }, { "grad_norm": 0.3501284718513489, "learning_rate": 7.63163981857822e-05, "loss": 0.0472, "step": 35740 }, { "grad_norm": 0.6180920600891113, "learning_rate": 7.630233759672403e-05, "loss": 0.0471, "step": 35750 }, { "grad_norm": 0.5817120671272278, "learning_rate": 7.628827413128177e-05, "loss": 0.0658, "step": 35760 }, { "grad_norm": 1.4696764945983887, "learning_rate": 7.627420779099338e-05, "loss": 0.0545, "step": 35770 }, { "grad_norm": 0.770193338394165, "learning_rate": 7.626013857739711e-05, "loss": 0.0497, "step": 35780 }, { "grad_norm": 0.3969866633415222, "learning_rate": 7.624606649203158e-05, "loss": 0.0423, "step": 35790 }, { "grad_norm": 0.5691788792610168, "learning_rate": 7.623199153643569e-05, "loss": 0.0443, "step": 35800 }, { "grad_norm": 0.6607389450073242, "learning_rate": 7.621791371214863e-05, "loss": 0.0548, "step": 35810 }, { "grad_norm": 0.41425445675849915, "learning_rate": 7.620383302070995e-05, "loss": 0.0463, "step": 35820 }, { "grad_norm": 0.7159019112586975, "learning_rate": 7.61897494636595e-05, "loss": 0.0425, "step": 35830 }, { "grad_norm": 0.8121739029884338, "learning_rate": 7.617566304253739e-05, "loss": 0.0479, "step": 35840 }, { "grad_norm": 0.5120505094528198, "learning_rate": 7.616157375888416e-05, "loss": 0.0442, "step": 35850 }, { "grad_norm": 0.5607699751853943, "learning_rate": 7.614748161424053e-05, "loss": 0.039, "step": 35860 }, { "grad_norm": 0.6371276378631592, "learning_rate": 7.613338661014763e-05, "loss": 0.0534, "step": 35870 }, { "grad_norm": 0.5257290005683899, "learning_rate": 7.611928874814686e-05, "loss": 0.0514, "step": 35880 }, { "grad_norm": 0.5055220723152161, "learning_rate": 7.610518802977996e-05, "loss": 0.0488, "step": 35890 }, { "grad_norm": 0.5477265119552612, "learning_rate": 7.609108445658893e-05, "loss": 0.0493, "step": 35900 }, { "grad_norm": 0.5140045285224915, "learning_rate": 7.607697803011612e-05, "loss": 0.0581, "step": 35910 }, { "grad_norm": 0.6731801629066467, "learning_rate": 7.606286875190421e-05, "loss": 0.0529, "step": 35920 }, { "grad_norm": 0.3751882314682007, "learning_rate": 7.604875662349617e-05, "loss": 0.0416, "step": 35930 }, { "grad_norm": 0.4361484944820404, "learning_rate": 7.603464164643525e-05, "loss": 0.0552, "step": 35940 }, { "grad_norm": 0.5236194729804993, "learning_rate": 7.602052382226507e-05, "loss": 0.052, "step": 35950 }, { "grad_norm": 0.4893972873687744, "learning_rate": 7.600640315252954e-05, "loss": 0.0519, "step": 35960 }, { "grad_norm": 0.619289755821228, "learning_rate": 7.599227963877284e-05, "loss": 0.067, "step": 35970 }, { "grad_norm": 0.6647900342941284, "learning_rate": 7.597815328253954e-05, "loss": 0.0687, "step": 35980 }, { "grad_norm": 0.8723238706588745, "learning_rate": 7.596402408537443e-05, "loss": 0.0656, "step": 35990 }, { "grad_norm": 0.7333725690841675, "learning_rate": 7.59498920488227e-05, "loss": 0.0676, "step": 36000 }, { "grad_norm": 0.42415642738342285, "learning_rate": 7.593575717442979e-05, "loss": 0.0466, "step": 36010 }, { "grad_norm": 0.3641573488712311, "learning_rate": 7.592161946374147e-05, "loss": 0.0546, "step": 36020 }, { "grad_norm": 0.3701939582824707, "learning_rate": 7.590747891830381e-05, "loss": 0.0495, "step": 36030 }, { "grad_norm": 0.6835311055183411, "learning_rate": 7.58933355396632e-05, "loss": 0.0482, "step": 36040 }, { "grad_norm": 0.5288687944412231, "learning_rate": 7.587918932936636e-05, "loss": 0.0538, "step": 36050 }, { "grad_norm": 0.6171461343765259, "learning_rate": 7.586504028896028e-05, "loss": 0.0587, "step": 36060 }, { "grad_norm": 0.44839319586753845, "learning_rate": 7.585088841999228e-05, "loss": 0.049, "step": 36070 }, { "grad_norm": 0.6653961539268494, "learning_rate": 7.583673372400999e-05, "loss": 0.0464, "step": 36080 }, { "grad_norm": 0.6767017841339111, "learning_rate": 7.582257620256134e-05, "loss": 0.0567, "step": 36090 }, { "grad_norm": 0.5091124773025513, "learning_rate": 7.580841585719458e-05, "loss": 0.0544, "step": 36100 }, { "grad_norm": 0.6771222352981567, "learning_rate": 7.579425268945825e-05, "loss": 0.0503, "step": 36110 }, { "grad_norm": 0.6260941028594971, "learning_rate": 7.578008670090127e-05, "loss": 0.0401, "step": 36120 }, { "grad_norm": 0.28886523842811584, "learning_rate": 7.576591789307272e-05, "loss": 0.0473, "step": 36130 }, { "grad_norm": 0.48334023356437683, "learning_rate": 7.575174626752216e-05, "loss": 0.0376, "step": 36140 }, { "grad_norm": 1.3142188787460327, "learning_rate": 7.573757182579934e-05, "loss": 0.0562, "step": 36150 }, { "grad_norm": 0.5671991109848022, "learning_rate": 7.572339456945435e-05, "loss": 0.0509, "step": 36160 }, { "grad_norm": 0.5130692720413208, "learning_rate": 7.570921450003762e-05, "loss": 0.0604, "step": 36170 }, { "grad_norm": 0.5642116665840149, "learning_rate": 7.569503161909983e-05, "loss": 0.0501, "step": 36180 }, { "grad_norm": 0.5630170702934265, "learning_rate": 7.568084592819202e-05, "loss": 0.0594, "step": 36190 }, { "grad_norm": 0.6236729621887207, "learning_rate": 7.566665742886551e-05, "loss": 0.0597, "step": 36200 }, { "grad_norm": 1.0526586771011353, "learning_rate": 7.565246612267194e-05, "loss": 0.0566, "step": 36210 }, { "grad_norm": 0.7165299654006958, "learning_rate": 7.563827201116325e-05, "loss": 0.0616, "step": 36220 }, { "grad_norm": 0.7943265438079834, "learning_rate": 7.562407509589164e-05, "loss": 0.0556, "step": 36230 }, { "grad_norm": 0.3930373787879944, "learning_rate": 7.560987537840973e-05, "loss": 0.057, "step": 36240 }, { "grad_norm": 0.39164987206459045, "learning_rate": 7.559567286027036e-05, "loss": 0.0403, "step": 36250 }, { "grad_norm": 0.5448435544967651, "learning_rate": 7.558146754302668e-05, "loss": 0.0577, "step": 36260 }, { "grad_norm": 0.7300060391426086, "learning_rate": 7.556725942823217e-05, "loss": 0.048, "step": 36270 }, { "grad_norm": 0.4711647629737854, "learning_rate": 7.555304851744061e-05, "loss": 0.0544, "step": 36280 }, { "grad_norm": 1.064404845237732, "learning_rate": 7.553883481220608e-05, "loss": 0.0703, "step": 36290 }, { "grad_norm": 0.41455936431884766, "learning_rate": 7.552461831408298e-05, "loss": 0.0527, "step": 36300 }, { "grad_norm": 0.7390123009681702, "learning_rate": 7.551039902462599e-05, "loss": 0.0571, "step": 36310 }, { "grad_norm": 0.45569291710853577, "learning_rate": 7.549617694539011e-05, "loss": 0.0399, "step": 36320 }, { "grad_norm": 0.5460058450698853, "learning_rate": 7.548195207793066e-05, "loss": 0.0533, "step": 36330 }, { "grad_norm": 0.6040600538253784, "learning_rate": 7.546772442380323e-05, "loss": 0.0467, "step": 36340 }, { "grad_norm": 0.7749884724617004, "learning_rate": 7.545349398456376e-05, "loss": 0.0662, "step": 36350 }, { "grad_norm": 0.32403117418289185, "learning_rate": 7.543926076176845e-05, "loss": 0.0635, "step": 36360 }, { "grad_norm": 0.46156659722328186, "learning_rate": 7.542502475697385e-05, "loss": 0.0497, "step": 36370 }, { "grad_norm": 0.6178463697433472, "learning_rate": 7.541078597173675e-05, "loss": 0.0741, "step": 36380 }, { "grad_norm": 0.6985850930213928, "learning_rate": 7.539654440761431e-05, "loss": 0.0666, "step": 36390 }, { "grad_norm": 0.5240857601165771, "learning_rate": 7.538230006616395e-05, "loss": 0.0572, "step": 36400 }, { "grad_norm": 0.31790852546691895, "learning_rate": 7.536805294894342e-05, "loss": 0.0674, "step": 36410 }, { "grad_norm": 0.8111591935157776, "learning_rate": 7.535380305751076e-05, "loss": 0.0493, "step": 36420 }, { "grad_norm": 0.6318939328193665, "learning_rate": 7.533955039342431e-05, "loss": 0.0581, "step": 36430 }, { "grad_norm": 1.9120739698410034, "learning_rate": 7.532529495824274e-05, "loss": 0.0576, "step": 36440 }, { "grad_norm": 0.46464455127716064, "learning_rate": 7.531103675352497e-05, "loss": 0.0473, "step": 36450 }, { "grad_norm": 0.48697003722190857, "learning_rate": 7.529677578083029e-05, "loss": 0.0529, "step": 36460 }, { "grad_norm": 0.6865927577018738, "learning_rate": 7.528251204171823e-05, "loss": 0.0485, "step": 36470 }, { "grad_norm": 0.4371841549873352, "learning_rate": 7.526824553774867e-05, "loss": 0.0632, "step": 36480 }, { "grad_norm": 0.6191602945327759, "learning_rate": 7.525397627048178e-05, "loss": 0.0605, "step": 36490 }, { "grad_norm": 0.5925769209861755, "learning_rate": 7.523970424147802e-05, "loss": 0.0528, "step": 36500 }, { "grad_norm": 0.37842831015586853, "learning_rate": 7.522542945229813e-05, "loss": 0.0479, "step": 36510 }, { "grad_norm": 0.9167138934135437, "learning_rate": 7.521115190450321e-05, "loss": 0.0582, "step": 36520 }, { "grad_norm": 0.8936524391174316, "learning_rate": 7.519687159965462e-05, "loss": 0.055, "step": 36530 }, { "grad_norm": 0.9054574966430664, "learning_rate": 7.518258853931403e-05, "loss": 0.0558, "step": 36540 }, { "grad_norm": 0.4872986078262329, "learning_rate": 7.516830272504342e-05, "loss": 0.0423, "step": 36550 }, { "grad_norm": 0.524600088596344, "learning_rate": 7.515401415840509e-05, "loss": 0.0645, "step": 36560 }, { "grad_norm": 0.3593958020210266, "learning_rate": 7.513972284096155e-05, "loss": 0.0502, "step": 36570 }, { "grad_norm": 0.582795262336731, "learning_rate": 7.512542877427574e-05, "loss": 0.0603, "step": 36580 }, { "grad_norm": 0.5697261095046997, "learning_rate": 7.51111319599108e-05, "loss": 0.0516, "step": 36590 }, { "grad_norm": 1.2739176750183105, "learning_rate": 7.509683239943024e-05, "loss": 0.0643, "step": 36600 }, { "grad_norm": 0.4621325433254242, "learning_rate": 7.50825300943978e-05, "loss": 0.0549, "step": 36610 }, { "grad_norm": 0.5323300361633301, "learning_rate": 7.50682250463776e-05, "loss": 0.0586, "step": 36620 }, { "grad_norm": 0.7654542326927185, "learning_rate": 7.5053917256934e-05, "loss": 0.0543, "step": 36630 }, { "grad_norm": 0.4042944312095642, "learning_rate": 7.503960672763166e-05, "loss": 0.0511, "step": 36640 }, { "grad_norm": 0.3640722334384918, "learning_rate": 7.502529346003559e-05, "loss": 0.0506, "step": 36650 }, { "grad_norm": 0.4322577118873596, "learning_rate": 7.501097745571106e-05, "loss": 0.0559, "step": 36660 }, { "grad_norm": 0.4404369294643402, "learning_rate": 7.499665871622361e-05, "loss": 0.0591, "step": 36670 }, { "grad_norm": 0.676115870475769, "learning_rate": 7.498233724313917e-05, "loss": 0.059, "step": 36680 }, { "grad_norm": 0.4355824291706085, "learning_rate": 7.496801303802387e-05, "loss": 0.0588, "step": 36690 }, { "grad_norm": 0.6398289799690247, "learning_rate": 7.495368610244423e-05, "loss": 0.0546, "step": 36700 }, { "grad_norm": 0.5473085045814514, "learning_rate": 7.493935643796697e-05, "loss": 0.0601, "step": 36710 }, { "grad_norm": 0.6026062369346619, "learning_rate": 7.49250240461592e-05, "loss": 0.0532, "step": 36720 }, { "grad_norm": 0.513648271560669, "learning_rate": 7.491068892858826e-05, "loss": 0.0531, "step": 36730 }, { "grad_norm": 0.4814600944519043, "learning_rate": 7.489635108682184e-05, "loss": 0.0491, "step": 36740 }, { "grad_norm": 0.6278595924377441, "learning_rate": 7.48820105224279e-05, "loss": 0.0481, "step": 36750 }, { "grad_norm": 0.529974639415741, "learning_rate": 7.48676672369747e-05, "loss": 0.0542, "step": 36760 }, { "grad_norm": 0.6997731924057007, "learning_rate": 7.485332123203079e-05, "loss": 0.0582, "step": 36770 }, { "grad_norm": 0.44955816864967346, "learning_rate": 7.483897250916504e-05, "loss": 0.0563, "step": 36780 }, { "grad_norm": 0.9871544241905212, "learning_rate": 7.48246210699466e-05, "loss": 0.0557, "step": 36790 }, { "grad_norm": 0.6123858690261841, "learning_rate": 7.481026691594492e-05, "loss": 0.0477, "step": 36800 }, { "grad_norm": 0.6334468722343445, "learning_rate": 7.479591004872975e-05, "loss": 0.0491, "step": 36810 }, { "grad_norm": 0.4952561557292938, "learning_rate": 7.478155046987114e-05, "loss": 0.0573, "step": 36820 }, { "grad_norm": 0.5893656015396118, "learning_rate": 7.476718818093944e-05, "loss": 0.0556, "step": 36830 }, { "grad_norm": 0.39421331882476807, "learning_rate": 7.475282318350527e-05, "loss": 0.0526, "step": 36840 }, { "grad_norm": 0.5662328004837036, "learning_rate": 7.473845547913957e-05, "loss": 0.0385, "step": 36850 }, { "grad_norm": 0.4879280924797058, "learning_rate": 7.472408506941357e-05, "loss": 0.0669, "step": 36860 }, { "grad_norm": 0.4611769914627075, "learning_rate": 7.470971195589881e-05, "loss": 0.0482, "step": 36870 }, { "grad_norm": 0.9318267703056335, "learning_rate": 7.46953361401671e-05, "loss": 0.0611, "step": 36880 }, { "grad_norm": 0.5028858184814453, "learning_rate": 7.468095762379055e-05, "loss": 0.0433, "step": 36890 }, { "grad_norm": 0.5016772747039795, "learning_rate": 7.466657640834158e-05, "loss": 0.056, "step": 36900 }, { "grad_norm": 0.47211435437202454, "learning_rate": 7.46521924953929e-05, "loss": 0.0504, "step": 36910 }, { "grad_norm": 0.5197139382362366, "learning_rate": 7.463780588651752e-05, "loss": 0.0505, "step": 36920 }, { "grad_norm": 0.45322999358177185, "learning_rate": 7.462341658328873e-05, "loss": 0.0522, "step": 36930 }, { "grad_norm": 0.4209339916706085, "learning_rate": 7.460902458728012e-05, "loss": 0.0582, "step": 36940 }, { "grad_norm": 0.47030067443847656, "learning_rate": 7.459462990006558e-05, "loss": 0.0545, "step": 36950 }, { "grad_norm": 0.49870604276657104, "learning_rate": 7.458023252321929e-05, "loss": 0.0515, "step": 36960 }, { "grad_norm": 0.5815719366073608, "learning_rate": 7.456583245831573e-05, "loss": 0.0466, "step": 36970 }, { "grad_norm": 0.6514045596122742, "learning_rate": 7.455142970692968e-05, "loss": 0.0634, "step": 36980 }, { "grad_norm": 0.5245375633239746, "learning_rate": 7.453702427063615e-05, "loss": 0.0528, "step": 36990 }, { "grad_norm": 0.6029703617095947, "learning_rate": 7.452261615101057e-05, "loss": 0.0522, "step": 37000 }, { "grad_norm": 0.40533575415611267, "learning_rate": 7.450820534962856e-05, "loss": 0.0507, "step": 37010 }, { "grad_norm": 1.2338669300079346, "learning_rate": 7.449379186806604e-05, "loss": 0.0506, "step": 37020 }, { "grad_norm": 0.7019270658493042, "learning_rate": 7.44793757078993e-05, "loss": 0.0595, "step": 37030 }, { "grad_norm": 0.6204167604446411, "learning_rate": 7.44649568707048e-05, "loss": 0.058, "step": 37040 }, { "grad_norm": 0.5484369993209839, "learning_rate": 7.445053535805942e-05, "loss": 0.0582, "step": 37050 }, { "grad_norm": 0.5820252895355225, "learning_rate": 7.443611117154024e-05, "loss": 0.0462, "step": 37060 }, { "grad_norm": 1.506324052810669, "learning_rate": 7.442168431272469e-05, "loss": 0.0562, "step": 37070 }, { "grad_norm": 0.7065199613571167, "learning_rate": 7.440725478319044e-05, "loss": 0.0621, "step": 37080 }, { "grad_norm": 0.5556910634040833, "learning_rate": 7.439282258451553e-05, "loss": 0.0549, "step": 37090 }, { "grad_norm": 0.4645139276981354, "learning_rate": 7.43783877182782e-05, "loss": 0.0527, "step": 37100 }, { "grad_norm": 0.40180328488349915, "learning_rate": 7.436395018605707e-05, "loss": 0.0514, "step": 37110 }, { "grad_norm": 0.4046365022659302, "learning_rate": 7.434950998943094e-05, "loss": 0.0451, "step": 37120 }, { "grad_norm": 0.41917333006858826, "learning_rate": 7.433506712997902e-05, "loss": 0.0602, "step": 37130 }, { "grad_norm": 0.6250825524330139, "learning_rate": 7.432062160928077e-05, "loss": 0.0725, "step": 37140 }, { "grad_norm": 0.47206375002861023, "learning_rate": 7.430617342891588e-05, "loss": 0.0515, "step": 37150 }, { "grad_norm": 0.463835746049881, "learning_rate": 7.429172259046442e-05, "loss": 0.0509, "step": 37160 }, { "grad_norm": 0.6185234785079956, "learning_rate": 7.42772690955067e-05, "loss": 0.0487, "step": 37170 }, { "grad_norm": 0.5692654252052307, "learning_rate": 7.426281294562332e-05, "loss": 0.0656, "step": 37180 }, { "grad_norm": 0.37844356894493103, "learning_rate": 7.424835414239519e-05, "loss": 0.0507, "step": 37190 }, { "grad_norm": 0.40014320611953735, "learning_rate": 7.423389268740352e-05, "loss": 0.0467, "step": 37200 }, { "grad_norm": 1.0003896951675415, "learning_rate": 7.42194285822298e-05, "loss": 0.0511, "step": 37210 }, { "grad_norm": 0.6260820031166077, "learning_rate": 7.420496182845576e-05, "loss": 0.0541, "step": 37220 }, { "grad_norm": 0.5268984436988831, "learning_rate": 7.419049242766348e-05, "loss": 0.0524, "step": 37230 }, { "grad_norm": 0.5843773484230042, "learning_rate": 7.417602038143533e-05, "loss": 0.0619, "step": 37240 }, { "grad_norm": 0.7445458769798279, "learning_rate": 7.416154569135393e-05, "loss": 0.0655, "step": 37250 }, { "grad_norm": 0.8502593040466309, "learning_rate": 7.414706835900222e-05, "loss": 0.0515, "step": 37260 }, { "grad_norm": 0.4768882691860199, "learning_rate": 7.413258838596345e-05, "loss": 0.0536, "step": 37270 }, { "grad_norm": 0.536484956741333, "learning_rate": 7.411810577382106e-05, "loss": 0.0545, "step": 37280 }, { "grad_norm": 0.5598533153533936, "learning_rate": 7.410362052415889e-05, "loss": 0.0523, "step": 37290 }, { "grad_norm": 0.4911198019981384, "learning_rate": 7.408913263856102e-05, "loss": 0.0461, "step": 37300 }, { "grad_norm": 0.4851668179035187, "learning_rate": 7.407464211861183e-05, "loss": 0.0528, "step": 37310 }, { "grad_norm": 0.4310857951641083, "learning_rate": 7.406014896589597e-05, "loss": 0.0525, "step": 37320 }, { "grad_norm": 0.3713977038860321, "learning_rate": 7.404565318199838e-05, "loss": 0.0473, "step": 37330 }, { "grad_norm": 0.4431585371494293, "learning_rate": 7.403115476850432e-05, "loss": 0.0684, "step": 37340 }, { "grad_norm": 0.41841405630111694, "learning_rate": 7.401665372699932e-05, "loss": 0.0521, "step": 37350 }, { "grad_norm": 0.5329870581626892, "learning_rate": 7.400215005906916e-05, "loss": 0.0552, "step": 37360 }, { "grad_norm": 0.6897110342979431, "learning_rate": 7.398764376629996e-05, "loss": 0.0547, "step": 37370 }, { "grad_norm": 0.5395128726959229, "learning_rate": 7.39731348502781e-05, "loss": 0.0569, "step": 37380 }, { "grad_norm": 0.6659893989562988, "learning_rate": 7.395862331259028e-05, "loss": 0.049, "step": 37390 }, { "grad_norm": 0.5119902491569519, "learning_rate": 7.394410915482341e-05, "loss": 0.0522, "step": 37400 }, { "grad_norm": 0.4714316725730896, "learning_rate": 7.392959237856476e-05, "loss": 0.0528, "step": 37410 }, { "grad_norm": 0.5067607164382935, "learning_rate": 7.391507298540187e-05, "loss": 0.0528, "step": 37420 }, { "grad_norm": 0.5521047711372375, "learning_rate": 7.390055097692257e-05, "loss": 0.0423, "step": 37430 }, { "grad_norm": 0.4916192889213562, "learning_rate": 7.388602635471494e-05, "loss": 0.0634, "step": 37440 }, { "grad_norm": 0.5805094242095947, "learning_rate": 7.387149912036738e-05, "loss": 0.0594, "step": 37450 }, { "grad_norm": 0.42540332674980164, "learning_rate": 7.385696927546854e-05, "loss": 0.0527, "step": 37460 }, { "grad_norm": 0.5595620274543762, "learning_rate": 7.384243682160743e-05, "loss": 0.0789, "step": 37470 }, { "grad_norm": 0.6144847273826599, "learning_rate": 7.382790176037329e-05, "loss": 0.0507, "step": 37480 }, { "grad_norm": 0.5935637354850769, "learning_rate": 7.381336409335562e-05, "loss": 0.0533, "step": 37490 }, { "grad_norm": 0.5960146188735962, "learning_rate": 7.379882382214426e-05, "loss": 0.0591, "step": 37500 }, { "grad_norm": 0.36193013191223145, "learning_rate": 7.378428094832931e-05, "loss": 0.0531, "step": 37510 }, { "grad_norm": 0.5699031352996826, "learning_rate": 7.376973547350114e-05, "loss": 0.0651, "step": 37520 }, { "grad_norm": 0.6828030347824097, "learning_rate": 7.375518739925046e-05, "loss": 0.051, "step": 37530 }, { "grad_norm": 0.5370264053344727, "learning_rate": 7.374063672716818e-05, "loss": 0.0462, "step": 37540 }, { "grad_norm": 0.38202813267707825, "learning_rate": 7.372608345884558e-05, "loss": 0.0497, "step": 37550 }, { "grad_norm": 0.7462555170059204, "learning_rate": 7.371152759587415e-05, "loss": 0.051, "step": 37560 }, { "grad_norm": 0.5411761403083801, "learning_rate": 7.369696913984572e-05, "loss": 0.0438, "step": 37570 }, { "grad_norm": 0.31627514958381653, "learning_rate": 7.368240809235237e-05, "loss": 0.0518, "step": 37580 }, { "grad_norm": 0.4789049029350281, "learning_rate": 7.366784445498649e-05, "loss": 0.0498, "step": 37590 }, { "grad_norm": 0.3115285336971283, "learning_rate": 7.36532782293407e-05, "loss": 0.0448, "step": 37600 }, { "grad_norm": 0.613832950592041, "learning_rate": 7.363870941700797e-05, "loss": 0.0525, "step": 37610 }, { "grad_norm": 0.5064272284507751, "learning_rate": 7.362413801958152e-05, "loss": 0.049, "step": 37620 }, { "grad_norm": 0.5409935712814331, "learning_rate": 7.360956403865486e-05, "loss": 0.0548, "step": 37630 }, { "grad_norm": 0.4934152364730835, "learning_rate": 7.359498747582177e-05, "loss": 0.0547, "step": 37640 }, { "grad_norm": 0.5050078630447388, "learning_rate": 7.358040833267634e-05, "loss": 0.0414, "step": 37650 }, { "grad_norm": 0.5473909378051758, "learning_rate": 7.356582661081288e-05, "loss": 0.0561, "step": 37660 }, { "grad_norm": 0.3858382999897003, "learning_rate": 7.355124231182607e-05, "loss": 0.0356, "step": 37670 }, { "grad_norm": 0.4628368318080902, "learning_rate": 7.353665543731083e-05, "loss": 0.0445, "step": 37680 }, { "grad_norm": 0.5083881616592407, "learning_rate": 7.352206598886229e-05, "loss": 0.0456, "step": 37690 }, { "grad_norm": 0.8398236036300659, "learning_rate": 7.350747396807601e-05, "loss": 0.0489, "step": 37700 }, { "grad_norm": 1.0433536767959595, "learning_rate": 7.349287937654772e-05, "loss": 0.0525, "step": 37710 }, { "grad_norm": 0.48512083292007446, "learning_rate": 7.347828221587345e-05, "loss": 0.0633, "step": 37720 }, { "grad_norm": 0.44809970259666443, "learning_rate": 7.346368248764955e-05, "loss": 0.05, "step": 37730 }, { "grad_norm": 0.7011911869049072, "learning_rate": 7.34490801934726e-05, "loss": 0.0458, "step": 37740 }, { "grad_norm": 0.35974690318107605, "learning_rate": 7.343447533493947e-05, "loss": 0.0657, "step": 37750 }, { "grad_norm": 0.41498205065727234, "learning_rate": 7.341986791364739e-05, "loss": 0.0713, "step": 37760 }, { "grad_norm": 0.33674827218055725, "learning_rate": 7.340525793119374e-05, "loss": 0.0454, "step": 37770 }, { "grad_norm": 0.5147663950920105, "learning_rate": 7.33906453891763e-05, "loss": 0.0466, "step": 37780 }, { "grad_norm": 0.8515644669532776, "learning_rate": 7.337603028919303e-05, "loss": 0.0571, "step": 37790 }, { "grad_norm": 0.30036652088165283, "learning_rate": 7.336141263284225e-05, "loss": 0.0483, "step": 37800 }, { "grad_norm": 0.5663512349128723, "learning_rate": 7.334679242172249e-05, "loss": 0.0507, "step": 37810 }, { "grad_norm": 0.5203085541725159, "learning_rate": 7.33321696574326e-05, "loss": 0.0463, "step": 37820 }, { "grad_norm": 0.6733338832855225, "learning_rate": 7.331754434157174e-05, "loss": 0.0443, "step": 37830 }, { "grad_norm": 0.4051860570907593, "learning_rate": 7.330291647573928e-05, "loss": 0.0465, "step": 37840 }, { "grad_norm": 0.5353267192840576, "learning_rate": 7.32882860615349e-05, "loss": 0.0407, "step": 37850 }, { "grad_norm": 0.8808449506759644, "learning_rate": 7.327365310055858e-05, "loss": 0.0503, "step": 37860 }, { "grad_norm": 0.519769549369812, "learning_rate": 7.325901759441055e-05, "loss": 0.0557, "step": 37870 }, { "grad_norm": 0.490689218044281, "learning_rate": 7.324437954469132e-05, "loss": 0.0464, "step": 37880 }, { "grad_norm": 0.4858494699001312, "learning_rate": 7.32297389530017e-05, "loss": 0.0437, "step": 37890 }, { "grad_norm": 0.866683840751648, "learning_rate": 7.321509582094274e-05, "loss": 0.0483, "step": 37900 }, { "grad_norm": 0.2869051694869995, "learning_rate": 7.320045015011582e-05, "loss": 0.0502, "step": 37910 }, { "grad_norm": 0.39866986870765686, "learning_rate": 7.318580194212253e-05, "loss": 0.0517, "step": 37920 }, { "grad_norm": 0.3703018128871918, "learning_rate": 7.317115119856483e-05, "loss": 0.0448, "step": 37930 }, { "grad_norm": 0.6475452780723572, "learning_rate": 7.315649792104485e-05, "loss": 0.0429, "step": 37940 }, { "grad_norm": 0.6177724003791809, "learning_rate": 7.314184211116507e-05, "loss": 0.0581, "step": 37950 }, { "grad_norm": 0.45874646306037903, "learning_rate": 7.312718377052823e-05, "loss": 0.0549, "step": 37960 }, { "grad_norm": 0.5449071526527405, "learning_rate": 7.311252290073737e-05, "loss": 0.0511, "step": 37970 }, { "grad_norm": 0.43030211329460144, "learning_rate": 7.309785950339572e-05, "loss": 0.0496, "step": 37980 }, { "grad_norm": 0.6401694416999817, "learning_rate": 7.308319358010692e-05, "loss": 0.0526, "step": 37990 }, { "grad_norm": 0.5605688691139221, "learning_rate": 7.306852513247474e-05, "loss": 0.0718, "step": 38000 }, { "grad_norm": 0.6310630440711975, "learning_rate": 7.305385416210336e-05, "loss": 0.0573, "step": 38010 }, { "grad_norm": 0.6176328063011169, "learning_rate": 7.303918067059711e-05, "loss": 0.0383, "step": 38020 }, { "grad_norm": 0.37311288714408875, "learning_rate": 7.302450465956075e-05, "loss": 0.0465, "step": 38030 }, { "grad_norm": 0.33169716596603394, "learning_rate": 7.300982613059914e-05, "loss": 0.0434, "step": 38040 }, { "grad_norm": 0.4842066466808319, "learning_rate": 7.299514508531757e-05, "loss": 0.062, "step": 38050 }, { "grad_norm": 0.4126241207122803, "learning_rate": 7.298046152532148e-05, "loss": 0.0504, "step": 38060 }, { "grad_norm": 0.7672654986381531, "learning_rate": 7.296577545221668e-05, "loss": 0.0465, "step": 38070 }, { "grad_norm": 1.0179613828659058, "learning_rate": 7.29510868676092e-05, "loss": 0.0594, "step": 38080 }, { "grad_norm": 0.5653968453407288, "learning_rate": 7.293639577310538e-05, "loss": 0.046, "step": 38090 }, { "grad_norm": 0.5550754070281982, "learning_rate": 7.292170217031179e-05, "loss": 0.0568, "step": 38100 }, { "grad_norm": 0.4937535524368286, "learning_rate": 7.290700606083532e-05, "loss": 0.0604, "step": 38110 }, { "grad_norm": 1.6560224294662476, "learning_rate": 7.28923074462831e-05, "loss": 0.0524, "step": 38120 }, { "grad_norm": 0.5209316611289978, "learning_rate": 7.287760632826255e-05, "loss": 0.0567, "step": 38130 }, { "grad_norm": 0.5577260255813599, "learning_rate": 7.286290270838139e-05, "loss": 0.061, "step": 38140 }, { "grad_norm": 0.668835461139679, "learning_rate": 7.284819658824756e-05, "loss": 0.0449, "step": 38150 }, { "grad_norm": 0.372176855802536, "learning_rate": 7.283348796946929e-05, "loss": 0.0606, "step": 38160 }, { "grad_norm": 0.4657711684703827, "learning_rate": 7.28187768536551e-05, "loss": 0.0482, "step": 38170 }, { "grad_norm": 0.7520318031311035, "learning_rate": 7.280406324241379e-05, "loss": 0.0562, "step": 38180 }, { "grad_norm": 0.3682289719581604, "learning_rate": 7.278934713735441e-05, "loss": 0.0594, "step": 38190 }, { "grad_norm": 0.5504017472267151, "learning_rate": 7.277462854008629e-05, "loss": 0.055, "step": 38200 }, { "grad_norm": 1.175667643547058, "learning_rate": 7.275990745221904e-05, "loss": 0.0525, "step": 38210 }, { "grad_norm": 0.6016707420349121, "learning_rate": 7.274518387536252e-05, "loss": 0.0558, "step": 38220 }, { "grad_norm": 0.6385751366615295, "learning_rate": 7.273045781112689e-05, "loss": 0.0592, "step": 38230 }, { "grad_norm": 0.3496890366077423, "learning_rate": 7.271572926112257e-05, "loss": 0.0478, "step": 38240 }, { "grad_norm": 0.5089292526245117, "learning_rate": 7.270099822696024e-05, "loss": 0.0612, "step": 38250 }, { "grad_norm": 0.2865902781486511, "learning_rate": 7.26862647102509e-05, "loss": 0.0517, "step": 38260 }, { "grad_norm": 0.3616912364959717, "learning_rate": 7.267152871260573e-05, "loss": 0.0525, "step": 38270 }, { "grad_norm": 0.5218363404273987, "learning_rate": 7.265679023563629e-05, "loss": 0.0582, "step": 38280 }, { "grad_norm": 0.893489420413971, "learning_rate": 7.264204928095433e-05, "loss": 0.0557, "step": 38290 }, { "grad_norm": 0.5060569643974304, "learning_rate": 7.262730585017188e-05, "loss": 0.051, "step": 38300 }, { "grad_norm": 0.3518390357494354, "learning_rate": 7.261255994490131e-05, "loss": 0.0441, "step": 38310 }, { "grad_norm": 0.4345034062862396, "learning_rate": 7.259781156675517e-05, "loss": 0.0584, "step": 38320 }, { "grad_norm": 0.5888918042182922, "learning_rate": 7.258306071734633e-05, "loss": 0.0454, "step": 38330 }, { "grad_norm": 0.47876015305519104, "learning_rate": 7.256830739828792e-05, "loss": 0.0538, "step": 38340 }, { "grad_norm": 0.5111622214317322, "learning_rate": 7.255355161119336e-05, "loss": 0.042, "step": 38350 }, { "grad_norm": 2.132859468460083, "learning_rate": 7.253879335767628e-05, "loss": 0.0491, "step": 38360 }, { "grad_norm": 0.47530597448349, "learning_rate": 7.252403263935067e-05, "loss": 0.0628, "step": 38370 }, { "grad_norm": 0.4855649471282959, "learning_rate": 7.25092694578307e-05, "loss": 0.0584, "step": 38380 }, { "grad_norm": 0.5921289920806885, "learning_rate": 7.249450381473087e-05, "loss": 0.057, "step": 38390 }, { "grad_norm": 0.4019990861415863, "learning_rate": 7.247973571166593e-05, "loss": 0.056, "step": 38400 }, { "grad_norm": 0.6114804148674011, "learning_rate": 7.246496515025089e-05, "loss": 0.0457, "step": 38410 }, { "grad_norm": 0.8705347180366516, "learning_rate": 7.245019213210101e-05, "loss": 0.0486, "step": 38420 }, { "grad_norm": 0.3127351999282837, "learning_rate": 7.24354166588319e-05, "loss": 0.0473, "step": 38430 }, { "grad_norm": 0.6131412982940674, "learning_rate": 7.242063873205935e-05, "loss": 0.0608, "step": 38440 }, { "grad_norm": 0.49448102712631226, "learning_rate": 7.240585835339946e-05, "loss": 0.06, "step": 38450 }, { "grad_norm": 0.43417030572891235, "learning_rate": 7.239107552446857e-05, "loss": 0.0576, "step": 38460 }, { "grad_norm": 0.7712007164955139, "learning_rate": 7.237629024688334e-05, "loss": 0.0534, "step": 38470 }, { "grad_norm": 0.5484076142311096, "learning_rate": 7.236150252226064e-05, "loss": 0.0586, "step": 38480 }, { "grad_norm": 0.2670334577560425, "learning_rate": 7.234671235221765e-05, "loss": 0.0605, "step": 38490 }, { "grad_norm": 0.8236950039863586, "learning_rate": 7.233191973837179e-05, "loss": 0.059, "step": 38500 }, { "grad_norm": 0.4153132140636444, "learning_rate": 7.231712468234078e-05, "loss": 0.0508, "step": 38510 }, { "grad_norm": 0.5154777765274048, "learning_rate": 7.230232718574254e-05, "loss": 0.0556, "step": 38520 }, { "grad_norm": 0.5355886220932007, "learning_rate": 7.228752725019535e-05, "loss": 0.0547, "step": 38530 }, { "grad_norm": 0.4015209972858429, "learning_rate": 7.227272487731769e-05, "loss": 0.0524, "step": 38540 }, { "grad_norm": 0.5794918537139893, "learning_rate": 7.225792006872831e-05, "loss": 0.056, "step": 38550 }, { "grad_norm": 0.6300376653671265, "learning_rate": 7.224311282604628e-05, "loss": 0.0526, "step": 38560 }, { "grad_norm": 0.8241471648216248, "learning_rate": 7.222830315089085e-05, "loss": 0.0668, "step": 38570 }, { "grad_norm": 0.8024306893348694, "learning_rate": 7.22134910448816e-05, "loss": 0.0536, "step": 38580 }, { "grad_norm": 0.47051000595092773, "learning_rate": 7.219867650963839e-05, "loss": 0.0454, "step": 38590 }, { "grad_norm": 0.4126133918762207, "learning_rate": 7.21838595467813e-05, "loss": 0.038, "step": 38600 }, { "grad_norm": 0.4732851982116699, "learning_rate": 7.216904015793067e-05, "loss": 0.054, "step": 38610 }, { "grad_norm": 0.6443623304367065, "learning_rate": 7.215421834470713e-05, "loss": 0.0472, "step": 38620 }, { "grad_norm": 0.559288501739502, "learning_rate": 7.213939410873158e-05, "loss": 0.0579, "step": 38630 }, { "grad_norm": 0.525383710861206, "learning_rate": 7.212456745162519e-05, "loss": 0.0561, "step": 38640 }, { "grad_norm": 0.5504887700080872, "learning_rate": 7.210973837500937e-05, "loss": 0.0595, "step": 38650 }, { "grad_norm": 0.4868478775024414, "learning_rate": 7.209490688050578e-05, "loss": 0.0469, "step": 38660 }, { "grad_norm": 0.7638612985610962, "learning_rate": 7.208007296973641e-05, "loss": 0.048, "step": 38670 }, { "grad_norm": 0.47973689436912537, "learning_rate": 7.206523664432345e-05, "loss": 0.0657, "step": 38680 }, { "grad_norm": 0.6391460299491882, "learning_rate": 7.205039790588939e-05, "loss": 0.0467, "step": 38690 }, { "grad_norm": 0.5239123106002808, "learning_rate": 7.203555675605697e-05, "loss": 0.0458, "step": 38700 }, { "grad_norm": 0.5361315608024597, "learning_rate": 7.202071319644917e-05, "loss": 0.0506, "step": 38710 }, { "grad_norm": 0.5959471464157104, "learning_rate": 7.200586722868932e-05, "loss": 0.0492, "step": 38720 }, { "grad_norm": 0.5238564014434814, "learning_rate": 7.199101885440088e-05, "loss": 0.0552, "step": 38730 }, { "grad_norm": 0.5554095506668091, "learning_rate": 7.197616807520771e-05, "loss": 0.0543, "step": 38740 }, { "grad_norm": 0.6166921257972717, "learning_rate": 7.196131489273381e-05, "loss": 0.0408, "step": 38750 }, { "grad_norm": 0.35762184858322144, "learning_rate": 7.194645930860355e-05, "loss": 0.042, "step": 38760 }, { "grad_norm": 0.5786793231964111, "learning_rate": 7.19316013244415e-05, "loss": 0.0511, "step": 38770 }, { "grad_norm": 0.5401131510734558, "learning_rate": 7.191674094187248e-05, "loss": 0.0537, "step": 38780 }, { "grad_norm": 0.7400352358818054, "learning_rate": 7.190187816252165e-05, "loss": 0.0434, "step": 38790 }, { "grad_norm": 0.8943634033203125, "learning_rate": 7.188701298801435e-05, "loss": 0.0585, "step": 38800 }, { "grad_norm": 0.5518742203712463, "learning_rate": 7.18721454199762e-05, "loss": 0.0521, "step": 38810 }, { "grad_norm": 0.7078374028205872, "learning_rate": 7.185727546003312e-05, "loss": 0.0462, "step": 38820 }, { "grad_norm": 0.4787794053554535, "learning_rate": 7.184240310981126e-05, "loss": 0.0472, "step": 38830 }, { "grad_norm": 0.5728646516799927, "learning_rate": 7.182752837093704e-05, "loss": 0.0562, "step": 38840 }, { "grad_norm": 0.5516157150268555, "learning_rate": 7.181265124503711e-05, "loss": 0.0626, "step": 38850 }, { "grad_norm": 0.5640326142311096, "learning_rate": 7.179777173373847e-05, "loss": 0.0546, "step": 38860 }, { "grad_norm": 0.6975908875465393, "learning_rate": 7.178288983866826e-05, "loss": 0.046, "step": 38870 }, { "grad_norm": 0.5974347591400146, "learning_rate": 7.176800556145397e-05, "loss": 0.0563, "step": 38880 }, { "grad_norm": 0.5869406461715698, "learning_rate": 7.175311890372334e-05, "loss": 0.0578, "step": 38890 }, { "grad_norm": 0.6042467951774597, "learning_rate": 7.17382298671043e-05, "loss": 0.0542, "step": 38900 }, { "grad_norm": 0.6424061059951782, "learning_rate": 7.172333845322515e-05, "loss": 0.0489, "step": 38910 }, { "grad_norm": 0.34955263137817383, "learning_rate": 7.170844466371436e-05, "loss": 0.061, "step": 38920 }, { "grad_norm": 0.49347344040870667, "learning_rate": 7.16935485002007e-05, "loss": 0.049, "step": 38930 }, { "grad_norm": 0.835045337677002, "learning_rate": 7.167864996431319e-05, "loss": 0.0594, "step": 38940 }, { "grad_norm": 0.9483333826065063, "learning_rate": 7.166374905768111e-05, "loss": 0.0569, "step": 38950 }, { "grad_norm": 0.4647408723831177, "learning_rate": 7.164884578193401e-05, "loss": 0.043, "step": 38960 }, { "grad_norm": 0.5080413818359375, "learning_rate": 7.163394013870168e-05, "loss": 0.0523, "step": 38970 }, { "grad_norm": 0.9400589466094971, "learning_rate": 7.161903212961419e-05, "loss": 0.0633, "step": 38980 }, { "grad_norm": 0.750450849533081, "learning_rate": 7.160412175630183e-05, "loss": 0.0482, "step": 38990 }, { "grad_norm": 0.5581291913986206, "learning_rate": 7.158920902039521e-05, "loss": 0.0436, "step": 39000 }, { "grad_norm": 1.2155044078826904, "learning_rate": 7.157429392352514e-05, "loss": 0.0593, "step": 39010 }, { "grad_norm": 0.7309526801109314, "learning_rate": 7.155937646732275e-05, "loss": 0.047, "step": 39020 }, { "grad_norm": 0.6609289646148682, "learning_rate": 7.154445665341933e-05, "loss": 0.0498, "step": 39030 }, { "grad_norm": 0.6001650094985962, "learning_rate": 7.152953448344654e-05, "loss": 0.0461, "step": 39040 }, { "grad_norm": 0.9292283654212952, "learning_rate": 7.151460995903624e-05, "loss": 0.058, "step": 39050 }, { "grad_norm": 0.5730419158935547, "learning_rate": 7.149968308182052e-05, "loss": 0.0531, "step": 39060 }, { "grad_norm": 0.481840044260025, "learning_rate": 7.14847538534318e-05, "loss": 0.053, "step": 39070 }, { "grad_norm": 0.6106027960777283, "learning_rate": 7.14698222755027e-05, "loss": 0.0441, "step": 39080 }, { "grad_norm": 0.6834585666656494, "learning_rate": 7.14548883496661e-05, "loss": 0.0471, "step": 39090 }, { "grad_norm": 0.6760976910591125, "learning_rate": 7.143995207755517e-05, "loss": 0.0542, "step": 39100 }, { "grad_norm": 0.7354968190193176, "learning_rate": 7.142501346080333e-05, "loss": 0.0415, "step": 39110 }, { "grad_norm": 0.632839560508728, "learning_rate": 7.141007250104421e-05, "loss": 0.052, "step": 39120 }, { "grad_norm": 0.8104535341262817, "learning_rate": 7.139512919991176e-05, "loss": 0.0554, "step": 39130 }, { "grad_norm": 0.7043783068656921, "learning_rate": 7.138018355904014e-05, "loss": 0.0485, "step": 39140 }, { "grad_norm": 0.7201046347618103, "learning_rate": 7.13652355800638e-05, "loss": 0.0708, "step": 39150 }, { "grad_norm": 0.8631027936935425, "learning_rate": 7.135028526461743e-05, "loss": 0.049, "step": 39160 }, { "grad_norm": 0.6291368007659912, "learning_rate": 7.133533261433594e-05, "loss": 0.0433, "step": 39170 }, { "grad_norm": 0.45153218507766724, "learning_rate": 7.132037763085457e-05, "loss": 0.0523, "step": 39180 }, { "grad_norm": 0.534344494342804, "learning_rate": 7.130542031580875e-05, "loss": 0.0425, "step": 39190 }, { "grad_norm": 0.4246244430541992, "learning_rate": 7.12904606708342e-05, "loss": 0.0547, "step": 39200 }, { "grad_norm": 0.4466071128845215, "learning_rate": 7.127549869756687e-05, "loss": 0.0588, "step": 39210 }, { "grad_norm": 0.3772035539150238, "learning_rate": 7.126053439764299e-05, "loss": 0.0503, "step": 39220 }, { "grad_norm": 0.41033095121383667, "learning_rate": 7.124556777269904e-05, "loss": 0.0552, "step": 39230 }, { "grad_norm": 0.797640860080719, "learning_rate": 7.123059882437174e-05, "loss": 0.0607, "step": 39240 }, { "grad_norm": 0.6802436709403992, "learning_rate": 7.121562755429807e-05, "loss": 0.049, "step": 39250 }, { "grad_norm": 0.7943371534347534, "learning_rate": 7.120065396411528e-05, "loss": 0.0477, "step": 39260 }, { "grad_norm": 0.5105482935905457, "learning_rate": 7.118567805546084e-05, "loss": 0.0456, "step": 39270 }, { "grad_norm": 0.6670964956283569, "learning_rate": 7.117069982997248e-05, "loss": 0.053, "step": 39280 }, { "grad_norm": 0.42575496435165405, "learning_rate": 7.115571928928825e-05, "loss": 0.0659, "step": 39290 }, { "grad_norm": 0.5613719820976257, "learning_rate": 7.114073643504635e-05, "loss": 0.0492, "step": 39300 }, { "grad_norm": 0.6059369444847107, "learning_rate": 7.11257512688853e-05, "loss": 0.0464, "step": 39310 }, { "grad_norm": 0.4466642439365387, "learning_rate": 7.111076379244384e-05, "loss": 0.0458, "step": 39320 }, { "grad_norm": 0.6894257664680481, "learning_rate": 7.109577400736101e-05, "loss": 0.0571, "step": 39330 }, { "grad_norm": 1.6858359575271606, "learning_rate": 7.108078191527605e-05, "loss": 0.0481, "step": 39340 }, { "grad_norm": 0.5727655291557312, "learning_rate": 7.106578751782847e-05, "loss": 0.0554, "step": 39350 }, { "grad_norm": 0.4040776491165161, "learning_rate": 7.105079081665803e-05, "loss": 0.0509, "step": 39360 }, { "grad_norm": 0.501721203327179, "learning_rate": 7.103579181340476e-05, "loss": 0.065, "step": 39370 }, { "grad_norm": 1.100218415260315, "learning_rate": 7.102079050970893e-05, "loss": 0.0529, "step": 39380 }, { "grad_norm": 0.36873114109039307, "learning_rate": 7.100578690721105e-05, "loss": 0.0556, "step": 39390 }, { "grad_norm": 0.5571656227111816, "learning_rate": 7.09907810075519e-05, "loss": 0.0602, "step": 39400 }, { "grad_norm": 0.4600946605205536, "learning_rate": 7.097577281237249e-05, "loss": 0.0595, "step": 39410 }, { "grad_norm": 0.7476599812507629, "learning_rate": 7.09607623233141e-05, "loss": 0.0443, "step": 39420 }, { "grad_norm": 0.5413403511047363, "learning_rate": 7.094574954201827e-05, "loss": 0.051, "step": 39430 }, { "grad_norm": 0.514276921749115, "learning_rate": 7.093073447012675e-05, "loss": 0.0531, "step": 39440 }, { "grad_norm": 0.7047935128211975, "learning_rate": 7.09157171092816e-05, "loss": 0.0523, "step": 39450 }, { "grad_norm": 0.4411696195602417, "learning_rate": 7.090069746112504e-05, "loss": 0.0516, "step": 39460 }, { "grad_norm": 0.5556467175483704, "learning_rate": 7.088567552729965e-05, "loss": 0.0446, "step": 39470 }, { "grad_norm": 0.4022117555141449, "learning_rate": 7.087065130944818e-05, "loss": 0.0557, "step": 39480 }, { "grad_norm": 0.42110803723335266, "learning_rate": 7.085562480921366e-05, "loss": 0.0558, "step": 39490 }, { "grad_norm": 0.577798068523407, "learning_rate": 7.084059602823937e-05, "loss": 0.0461, "step": 39500 }, { "grad_norm": 0.47364169359207153, "learning_rate": 7.082556496816882e-05, "loss": 0.047, "step": 39510 }, { "grad_norm": 0.4261304438114166, "learning_rate": 7.081053163064582e-05, "loss": 0.0626, "step": 39520 }, { "grad_norm": 0.4595625400543213, "learning_rate": 7.079549601731434e-05, "loss": 0.0417, "step": 39530 }, { "grad_norm": 0.5130418539047241, "learning_rate": 7.07804581298187e-05, "loss": 0.0562, "step": 39540 }, { "grad_norm": 0.513236939907074, "learning_rate": 7.07654179698034e-05, "loss": 0.0677, "step": 39550 }, { "grad_norm": 0.5683504343032837, "learning_rate": 7.075037553891321e-05, "loss": 0.0596, "step": 39560 }, { "grad_norm": 0.42077386379241943, "learning_rate": 7.073533083879315e-05, "loss": 0.0503, "step": 39570 }, { "grad_norm": 1.032562017440796, "learning_rate": 7.072028387108849e-05, "loss": 0.0535, "step": 39580 }, { "grad_norm": 0.5153048634529114, "learning_rate": 7.070523463744472e-05, "loss": 0.0511, "step": 39590 }, { "grad_norm": 0.35249680280685425, "learning_rate": 7.069018313950763e-05, "loss": 0.0475, "step": 39600 }, { "grad_norm": 0.49428853392601013, "learning_rate": 7.067512937892322e-05, "loss": 0.0403, "step": 39610 }, { "grad_norm": 0.44208335876464844, "learning_rate": 7.066007335733775e-05, "loss": 0.039, "step": 39620 }, { "grad_norm": 0.4805048704147339, "learning_rate": 7.064501507639772e-05, "loss": 0.0465, "step": 39630 }, { "grad_norm": 0.38602834939956665, "learning_rate": 7.062995453774987e-05, "loss": 0.0576, "step": 39640 }, { "grad_norm": 0.36560797691345215, "learning_rate": 7.061489174304121e-05, "loss": 0.0612, "step": 39650 }, { "grad_norm": 0.4725657105445862, "learning_rate": 7.0599826693919e-05, "loss": 0.0399, "step": 39660 }, { "grad_norm": 0.6241511106491089, "learning_rate": 7.05847593920307e-05, "loss": 0.0504, "step": 39670 }, { "grad_norm": 0.579116940498352, "learning_rate": 7.056968983902406e-05, "loss": 0.0422, "step": 39680 }, { "grad_norm": 0.3925558924674988, "learning_rate": 7.055461803654706e-05, "loss": 0.0426, "step": 39690 }, { "grad_norm": 0.316773384809494, "learning_rate": 7.053954398624794e-05, "loss": 0.0478, "step": 39700 }, { "grad_norm": 0.41179534792900085, "learning_rate": 7.052446768977518e-05, "loss": 0.0416, "step": 39710 }, { "grad_norm": 0.48172780871391296, "learning_rate": 7.050938914877748e-05, "loss": 0.0573, "step": 39720 }, { "grad_norm": 0.5713452696800232, "learning_rate": 7.049430836490382e-05, "loss": 0.0659, "step": 39730 }, { "grad_norm": 0.3932208716869354, "learning_rate": 7.047922533980342e-05, "loss": 0.0488, "step": 39740 }, { "grad_norm": 0.5115440487861633, "learning_rate": 7.046414007512571e-05, "loss": 0.047, "step": 39750 }, { "grad_norm": 1.231539249420166, "learning_rate": 7.044905257252042e-05, "loss": 0.0453, "step": 39760 }, { "grad_norm": 0.807392954826355, "learning_rate": 7.04339628336375e-05, "loss": 0.0577, "step": 39770 }, { "grad_norm": 0.7979562282562256, "learning_rate": 7.041887086012711e-05, "loss": 0.0448, "step": 39780 }, { "grad_norm": 0.8454252481460571, "learning_rate": 7.040377665363969e-05, "loss": 0.0484, "step": 39790 }, { "grad_norm": 0.4011366367340088, "learning_rate": 7.038868021582594e-05, "loss": 0.0522, "step": 39800 }, { "grad_norm": 0.37076669931411743, "learning_rate": 7.037358154833679e-05, "loss": 0.0473, "step": 39810 }, { "grad_norm": 0.3899856507778168, "learning_rate": 7.035848065282339e-05, "loss": 0.0482, "step": 39820 }, { "grad_norm": 0.6107420325279236, "learning_rate": 7.034337753093714e-05, "loss": 0.0444, "step": 39830 }, { "grad_norm": 0.4391191601753235, "learning_rate": 7.032827218432972e-05, "loss": 0.0476, "step": 39840 }, { "grad_norm": 0.41997015476226807, "learning_rate": 7.031316461465302e-05, "loss": 0.0428, "step": 39850 }, { "grad_norm": 0.5390275716781616, "learning_rate": 7.029805482355915e-05, "loss": 0.045, "step": 39860 }, { "grad_norm": 0.4861912131309509, "learning_rate": 7.028294281270055e-05, "loss": 0.0421, "step": 39870 }, { "grad_norm": 0.33431458473205566, "learning_rate": 7.026782858372982e-05, "loss": 0.0496, "step": 39880 }, { "grad_norm": 0.40015578269958496, "learning_rate": 7.025271213829982e-05, "loss": 0.0495, "step": 39890 }, { "grad_norm": 0.40185853838920593, "learning_rate": 7.023759347806366e-05, "loss": 0.0396, "step": 39900 }, { "grad_norm": 0.6771048903465271, "learning_rate": 7.022247260467472e-05, "loss": 0.0471, "step": 39910 }, { "grad_norm": 0.4657834470272064, "learning_rate": 7.020734951978658e-05, "loss": 0.0507, "step": 39920 }, { "grad_norm": 0.5936537384986877, "learning_rate": 7.019222422505307e-05, "loss": 0.0446, "step": 39930 }, { "grad_norm": 0.6152610778808594, "learning_rate": 7.017709672212827e-05, "loss": 0.058, "step": 39940 }, { "grad_norm": 0.5628200769424438, "learning_rate": 7.016196701266652e-05, "loss": 0.0583, "step": 39950 }, { "grad_norm": 0.5672861933708191, "learning_rate": 7.014683509832238e-05, "loss": 0.0495, "step": 39960 }, { "grad_norm": 0.439902663230896, "learning_rate": 7.013170098075063e-05, "loss": 0.0553, "step": 39970 }, { "grad_norm": 0.48277953267097473, "learning_rate": 7.011656466160632e-05, "loss": 0.0469, "step": 39980 }, { "grad_norm": 0.43924275040626526, "learning_rate": 7.010142614254475e-05, "loss": 0.0442, "step": 39990 }, { "grad_norm": 0.42572247982025146, "learning_rate": 7.008628542522147e-05, "loss": 0.0542, "step": 40000 }, { "grad_norm": 0.44596201181411743, "learning_rate": 7.007114251129218e-05, "loss": 0.0487, "step": 40010 }, { "grad_norm": 0.8143712878227234, "learning_rate": 7.005599740241293e-05, "loss": 0.0378, "step": 40020 }, { "grad_norm": 0.5833597779273987, "learning_rate": 7.004085010023996e-05, "loss": 0.0514, "step": 40030 }, { "grad_norm": 0.42946889996528625, "learning_rate": 7.002570060642976e-05, "loss": 0.0422, "step": 40040 }, { "grad_norm": 0.6362008452415466, "learning_rate": 7.001054892263903e-05, "loss": 0.0394, "step": 40050 }, { "grad_norm": 0.6694809198379517, "learning_rate": 6.999539505052477e-05, "loss": 0.0619, "step": 40060 }, { "grad_norm": 0.4400162994861603, "learning_rate": 6.998023899174418e-05, "loss": 0.0472, "step": 40070 }, { "grad_norm": 0.46347132325172424, "learning_rate": 6.996508074795467e-05, "loss": 0.0497, "step": 40080 }, { "grad_norm": 0.5515957474708557, "learning_rate": 6.994992032081396e-05, "loss": 0.0597, "step": 40090 }, { "grad_norm": 0.6269599795341492, "learning_rate": 6.993475771197995e-05, "loss": 0.0495, "step": 40100 }, { "grad_norm": 0.5471302270889282, "learning_rate": 6.991959292311082e-05, "loss": 0.0482, "step": 40110 }, { "grad_norm": 0.6366288661956787, "learning_rate": 6.990442595586495e-05, "loss": 0.0532, "step": 40120 }, { "grad_norm": 0.48713740706443787, "learning_rate": 6.988925681190098e-05, "loss": 0.0518, "step": 40130 }, { "grad_norm": 0.5884585380554199, "learning_rate": 6.987408549287778e-05, "loss": 0.052, "step": 40140 }, { "grad_norm": 0.6997694969177246, "learning_rate": 6.985891200045449e-05, "loss": 0.0569, "step": 40150 }, { "grad_norm": 0.5380082726478577, "learning_rate": 6.984373633629045e-05, "loss": 0.0612, "step": 40160 }, { "grad_norm": 0.43396663665771484, "learning_rate": 6.98285585020452e-05, "loss": 0.0415, "step": 40170 }, { "grad_norm": 0.41096702218055725, "learning_rate": 6.981337849937864e-05, "loss": 0.0498, "step": 40180 }, { "grad_norm": 0.6225002408027649, "learning_rate": 6.979819632995078e-05, "loss": 0.0519, "step": 40190 }, { "grad_norm": 0.4709635078907013, "learning_rate": 6.978301199542193e-05, "loss": 0.0404, "step": 40200 }, { "grad_norm": 0.7183271050453186, "learning_rate": 6.976782549745263e-05, "loss": 0.0574, "step": 40210 }, { "grad_norm": 0.6642780900001526, "learning_rate": 6.975263683770367e-05, "loss": 0.0509, "step": 40220 }, { "grad_norm": 0.5786572694778442, "learning_rate": 6.973744601783604e-05, "loss": 0.0459, "step": 40230 }, { "grad_norm": 0.6028708219528198, "learning_rate": 6.972225303951097e-05, "loss": 0.0489, "step": 40240 }, { "grad_norm": 0.522469699382782, "learning_rate": 6.970705790438998e-05, "loss": 0.0369, "step": 40250 }, { "grad_norm": 0.5147098898887634, "learning_rate": 6.969186061413477e-05, "loss": 0.0421, "step": 40260 }, { "grad_norm": 0.3446636199951172, "learning_rate": 6.967666117040727e-05, "loss": 0.0563, "step": 40270 }, { "grad_norm": 0.5534230470657349, "learning_rate": 6.966145957486972e-05, "loss": 0.0476, "step": 40280 }, { "grad_norm": 0.6544829607009888, "learning_rate": 6.964625582918449e-05, "loss": 0.0526, "step": 40290 }, { "grad_norm": 0.5967515110969543, "learning_rate": 6.963104993501425e-05, "loss": 0.0474, "step": 40300 }, { "grad_norm": 0.8583747148513794, "learning_rate": 6.961584189402192e-05, "loss": 0.0479, "step": 40310 }, { "grad_norm": 0.5665953159332275, "learning_rate": 6.96006317078706e-05, "loss": 0.0372, "step": 40320 }, { "grad_norm": 1.2164466381072998, "learning_rate": 6.958541937822367e-05, "loss": 0.0385, "step": 40330 }, { "grad_norm": 0.4334373474121094, "learning_rate": 6.957020490674473e-05, "loss": 0.0471, "step": 40340 }, { "grad_norm": 0.37983718514442444, "learning_rate": 6.95549882950976e-05, "loss": 0.0456, "step": 40350 }, { "grad_norm": 0.6154529452323914, "learning_rate": 6.953976954494635e-05, "loss": 0.0413, "step": 40360 }, { "grad_norm": 0.5573663711547852, "learning_rate": 6.952454865795528e-05, "loss": 0.0591, "step": 40370 }, { "grad_norm": 0.5637920498847961, "learning_rate": 6.950932563578892e-05, "loss": 0.0559, "step": 40380 }, { "grad_norm": 0.5038485527038574, "learning_rate": 6.949410048011206e-05, "loss": 0.0485, "step": 40390 }, { "grad_norm": 0.4028622806072235, "learning_rate": 6.947887319258966e-05, "loss": 0.0539, "step": 40400 }, { "grad_norm": 0.39531517028808594, "learning_rate": 6.946364377488696e-05, "loss": 0.0602, "step": 40410 }, { "grad_norm": 0.49381229281425476, "learning_rate": 6.944841222866947e-05, "loss": 0.0592, "step": 40420 }, { "grad_norm": 0.7713856101036072, "learning_rate": 6.943317855560284e-05, "loss": 0.0587, "step": 40430 }, { "grad_norm": 0.7081847786903381, "learning_rate": 6.941794275735302e-05, "loss": 0.0529, "step": 40440 }, { "grad_norm": 0.34339964389801025, "learning_rate": 6.94027048355862e-05, "loss": 0.0375, "step": 40450 }, { "grad_norm": 0.4216466546058655, "learning_rate": 6.938746479196871e-05, "loss": 0.0509, "step": 40460 }, { "grad_norm": 0.5470547080039978, "learning_rate": 6.937222262816724e-05, "loss": 0.0439, "step": 40470 }, { "grad_norm": 0.4329724907875061, "learning_rate": 6.935697834584865e-05, "loss": 0.0501, "step": 40480 }, { "grad_norm": 0.5833931565284729, "learning_rate": 6.934173194667998e-05, "loss": 0.0497, "step": 40490 }, { "grad_norm": 0.7499781847000122, "learning_rate": 6.93264834323286e-05, "loss": 0.0591, "step": 40500 }, { "grad_norm": 0.2707918584346771, "learning_rate": 6.931123280446206e-05, "loss": 0.0497, "step": 40510 }, { "grad_norm": 0.7621815204620361, "learning_rate": 6.929598006474811e-05, "loss": 0.0504, "step": 40520 }, { "grad_norm": 0.2859082818031311, "learning_rate": 6.92807252148548e-05, "loss": 0.046, "step": 40530 }, { "grad_norm": 0.390031099319458, "learning_rate": 6.926546825645037e-05, "loss": 0.0354, "step": 40540 }, { "grad_norm": 0.2708735764026642, "learning_rate": 6.92502091912033e-05, "loss": 0.0443, "step": 40550 }, { "grad_norm": 0.3638336658477783, "learning_rate": 6.923494802078229e-05, "loss": 0.0507, "step": 40560 }, { "grad_norm": 0.9175235629081726, "learning_rate": 6.921968474685628e-05, "loss": 0.0384, "step": 40570 }, { "grad_norm": 0.5792329907417297, "learning_rate": 6.920441937109445e-05, "loss": 0.0696, "step": 40580 }, { "grad_norm": 2.97039532661438, "learning_rate": 6.91891518951662e-05, "loss": 0.0612, "step": 40590 }, { "grad_norm": 1.18952214717865, "learning_rate": 6.917388232074114e-05, "loss": 0.0571, "step": 40600 }, { "grad_norm": 0.4839693009853363, "learning_rate": 6.915861064948914e-05, "loss": 0.0579, "step": 40610 }, { "grad_norm": 0.3436949849128723, "learning_rate": 6.914333688308025e-05, "loss": 0.043, "step": 40620 }, { "grad_norm": 0.35235074162483215, "learning_rate": 6.912806102318487e-05, "loss": 0.0446, "step": 40630 }, { "grad_norm": 0.3249760568141937, "learning_rate": 6.911278307147347e-05, "loss": 0.0474, "step": 40640 }, { "grad_norm": 0.38420844078063965, "learning_rate": 6.909750302961684e-05, "loss": 0.05, "step": 40650 }, { "grad_norm": 0.44725385308265686, "learning_rate": 6.9082220899286e-05, "loss": 0.0492, "step": 40660 }, { "grad_norm": 0.48744407296180725, "learning_rate": 6.906693668215219e-05, "loss": 0.0591, "step": 40670 }, { "grad_norm": 0.5755507349967957, "learning_rate": 6.905165037988683e-05, "loss": 0.0494, "step": 40680 }, { "grad_norm": 0.6500880718231201, "learning_rate": 6.903636199416164e-05, "loss": 0.0459, "step": 40690 }, { "grad_norm": 0.6028441190719604, "learning_rate": 6.902107152664851e-05, "loss": 0.047, "step": 40700 }, { "grad_norm": 0.5760281085968018, "learning_rate": 6.90057789790196e-05, "loss": 0.0506, "step": 40710 }, { "grad_norm": 0.43323615193367004, "learning_rate": 6.899048435294728e-05, "loss": 0.0542, "step": 40720 }, { "grad_norm": 0.4996045231819153, "learning_rate": 6.897518765010415e-05, "loss": 0.0547, "step": 40730 }, { "grad_norm": 0.3667576313018799, "learning_rate": 6.895988887216302e-05, "loss": 0.053, "step": 40740 }, { "grad_norm": 0.41264602541923523, "learning_rate": 6.894458802079694e-05, "loss": 0.0561, "step": 40750 }, { "grad_norm": 0.7088901400566101, "learning_rate": 6.892928509767922e-05, "loss": 0.0423, "step": 40760 }, { "grad_norm": 0.3088286817073822, "learning_rate": 6.891398010448333e-05, "loss": 0.0429, "step": 40770 }, { "grad_norm": 0.5164387226104736, "learning_rate": 6.8898673042883e-05, "loss": 0.0642, "step": 40780 }, { "grad_norm": 0.5321810245513916, "learning_rate": 6.888336391455222e-05, "loss": 0.0497, "step": 40790 }, { "grad_norm": 0.36452481150627136, "learning_rate": 6.886805272116513e-05, "loss": 0.0518, "step": 40800 }, { "grad_norm": 0.5156601667404175, "learning_rate": 6.885273946439617e-05, "loss": 0.0521, "step": 40810 }, { "grad_norm": 0.8569020628929138, "learning_rate": 6.883742414591998e-05, "loss": 0.0657, "step": 40820 }, { "grad_norm": 0.5718869566917419, "learning_rate": 6.88221067674114e-05, "loss": 0.0481, "step": 40830 }, { "grad_norm": 0.5315548181533813, "learning_rate": 6.88067873305455e-05, "loss": 0.0577, "step": 40840 }, { "grad_norm": 0.7012627124786377, "learning_rate": 6.879146583699765e-05, "loss": 0.0545, "step": 40850 }, { "grad_norm": 0.35028988122940063, "learning_rate": 6.877614228844334e-05, "loss": 0.0488, "step": 40860 }, { "grad_norm": 0.5669193863868713, "learning_rate": 6.876081668655832e-05, "loss": 0.049, "step": 40870 }, { "grad_norm": 0.448506236076355, "learning_rate": 6.874548903301861e-05, "loss": 0.0586, "step": 40880 }, { "grad_norm": 0.46083012223243713, "learning_rate": 6.873015932950039e-05, "loss": 0.0453, "step": 40890 }, { "grad_norm": 0.5248012542724609, "learning_rate": 6.871482757768012e-05, "loss": 0.0536, "step": 40900 }, { "grad_norm": 0.4642413258552551, "learning_rate": 6.869949377923442e-05, "loss": 0.0515, "step": 40910 }, { "grad_norm": 0.7122989892959595, "learning_rate": 6.868415793584022e-05, "loss": 0.0549, "step": 40920 }, { "grad_norm": 0.580053985118866, "learning_rate": 6.86688200491746e-05, "loss": 0.043, "step": 40930 }, { "grad_norm": 0.4061126410961151, "learning_rate": 6.865348012091488e-05, "loss": 0.0463, "step": 40940 }, { "grad_norm": 0.407226026058197, "learning_rate": 6.86381381527386e-05, "loss": 0.0491, "step": 40950 }, { "grad_norm": 0.4323979914188385, "learning_rate": 6.862279414632359e-05, "loss": 0.0483, "step": 40960 }, { "grad_norm": 0.49605730175971985, "learning_rate": 6.860744810334778e-05, "loss": 0.048, "step": 40970 }, { "grad_norm": 0.6394315361976624, "learning_rate": 6.859210002548943e-05, "loss": 0.0441, "step": 40980 }, { "grad_norm": 0.46724075078964233, "learning_rate": 6.8576749914427e-05, "loss": 0.0429, "step": 40990 }, { "grad_norm": 0.6018202304840088, "learning_rate": 6.85613977718391e-05, "loss": 0.0524, "step": 41000 }, { "grad_norm": 1.7261815071105957, "learning_rate": 6.854604359940464e-05, "loss": 0.0439, "step": 41010 }, { "grad_norm": 0.5387897491455078, "learning_rate": 6.853068739880276e-05, "loss": 0.0469, "step": 41020 }, { "grad_norm": 0.6564906239509583, "learning_rate": 6.851532917171276e-05, "loss": 0.047, "step": 41030 }, { "grad_norm": 0.6308006048202515, "learning_rate": 6.849996891981418e-05, "loss": 0.0641, "step": 41040 }, { "grad_norm": 1.2599220275878906, "learning_rate": 6.848460664478684e-05, "loss": 0.0495, "step": 41050 }, { "grad_norm": 0.5339293479919434, "learning_rate": 6.846924234831069e-05, "loss": 0.0361, "step": 41060 }, { "grad_norm": 0.4924677312374115, "learning_rate": 6.845387603206597e-05, "loss": 0.055, "step": 41070 }, { "grad_norm": 0.4273013174533844, "learning_rate": 6.843850769773312e-05, "loss": 0.0466, "step": 41080 }, { "grad_norm": 0.45053088665008545, "learning_rate": 6.842313734699278e-05, "loss": 0.0551, "step": 41090 }, { "grad_norm": 0.5418626666069031, "learning_rate": 6.840776498152584e-05, "loss": 0.0419, "step": 41100 }, { "grad_norm": 0.5862970948219299, "learning_rate": 6.83923906030134e-05, "loss": 0.0418, "step": 41110 }, { "grad_norm": 0.7104102969169617, "learning_rate": 6.837701421313677e-05, "loss": 0.0641, "step": 41120 }, { "grad_norm": 0.6381798386573792, "learning_rate": 6.83616358135775e-05, "loss": 0.0517, "step": 41130 }, { "grad_norm": 0.8991769552230835, "learning_rate": 6.834625540601734e-05, "loss": 0.0619, "step": 41140 }, { "grad_norm": 0.5123526453971863, "learning_rate": 6.833087299213829e-05, "loss": 0.0422, "step": 41150 }, { "grad_norm": 0.7348756790161133, "learning_rate": 6.83154885736225e-05, "loss": 0.0621, "step": 41160 }, { "grad_norm": 0.548491358757019, "learning_rate": 6.830010215215244e-05, "loss": 0.0541, "step": 41170 }, { "grad_norm": 0.8688469529151917, "learning_rate": 6.828471372941072e-05, "loss": 0.0533, "step": 41180 }, { "grad_norm": 0.6096020936965942, "learning_rate": 6.826932330708018e-05, "loss": 0.0515, "step": 41190 }, { "grad_norm": 0.41713637113571167, "learning_rate": 6.825393088684393e-05, "loss": 0.0437, "step": 41200 }, { "grad_norm": 0.4491086006164551, "learning_rate": 6.823853647038524e-05, "loss": 0.0449, "step": 41210 }, { "grad_norm": 0.3834781050682068, "learning_rate": 6.822314005938762e-05, "loss": 0.0471, "step": 41220 }, { "grad_norm": 0.62248295545578, "learning_rate": 6.820774165553479e-05, "loss": 0.0552, "step": 41230 }, { "grad_norm": 0.3489638566970825, "learning_rate": 6.819234126051072e-05, "loss": 0.0408, "step": 41240 }, { "grad_norm": 0.5175421833992004, "learning_rate": 6.817693887599956e-05, "loss": 0.054, "step": 41250 }, { "grad_norm": 0.5863044857978821, "learning_rate": 6.816153450368568e-05, "loss": 0.0611, "step": 41260 }, { "grad_norm": 0.3519512116909027, "learning_rate": 6.814612814525372e-05, "loss": 0.0517, "step": 41270 }, { "grad_norm": 0.5869372487068176, "learning_rate": 6.813071980238846e-05, "loss": 0.0457, "step": 41280 }, { "grad_norm": 0.6953763961791992, "learning_rate": 6.811530947677492e-05, "loss": 0.0505, "step": 41290 }, { "grad_norm": 0.4148716628551483, "learning_rate": 6.809989717009839e-05, "loss": 0.0448, "step": 41300 }, { "grad_norm": 0.581041693687439, "learning_rate": 6.808448288404431e-05, "loss": 0.0665, "step": 41310 }, { "grad_norm": 0.6032953858375549, "learning_rate": 6.806906662029838e-05, "loss": 0.0551, "step": 41320 }, { "grad_norm": 2.247333526611328, "learning_rate": 6.805364838054647e-05, "loss": 0.049, "step": 41330 }, { "grad_norm": 0.4503243565559387, "learning_rate": 6.803822816647474e-05, "loss": 0.0358, "step": 41340 }, { "grad_norm": 0.5907599329948425, "learning_rate": 6.802280597976949e-05, "loss": 0.0467, "step": 41350 }, { "grad_norm": 1.3560080528259277, "learning_rate": 6.800738182211727e-05, "loss": 0.0443, "step": 41360 }, { "grad_norm": 0.5177980065345764, "learning_rate": 6.799195569520485e-05, "loss": 0.0435, "step": 41370 }, { "grad_norm": 0.8117746710777283, "learning_rate": 6.79765276007192e-05, "loss": 0.0506, "step": 41380 }, { "grad_norm": 0.4023725688457489, "learning_rate": 6.796109754034753e-05, "loss": 0.0538, "step": 41390 }, { "grad_norm": 0.6239342093467712, "learning_rate": 6.794566551577724e-05, "loss": 0.0488, "step": 41400 }, { "grad_norm": 0.6014485955238342, "learning_rate": 6.793023152869594e-05, "loss": 0.0426, "step": 41410 }, { "grad_norm": 0.6404609084129333, "learning_rate": 6.791479558079147e-05, "loss": 0.0475, "step": 41420 }, { "grad_norm": 0.3128334879875183, "learning_rate": 6.789935767375191e-05, "loss": 0.0473, "step": 41430 }, { "grad_norm": 0.38478773832321167, "learning_rate": 6.78839178092655e-05, "loss": 0.0349, "step": 41440 }, { "grad_norm": 0.47756969928741455, "learning_rate": 6.786847598902072e-05, "loss": 0.0591, "step": 41450 }, { "grad_norm": 0.41502541303634644, "learning_rate": 6.785303221470627e-05, "loss": 0.0438, "step": 41460 }, { "grad_norm": 0.4293210506439209, "learning_rate": 6.783758648801108e-05, "loss": 0.0414, "step": 41470 }, { "grad_norm": 0.8639335036277771, "learning_rate": 6.782213881062422e-05, "loss": 0.0536, "step": 41480 }, { "grad_norm": 0.4257786273956299, "learning_rate": 6.780668918423508e-05, "loss": 0.0469, "step": 41490 }, { "grad_norm": 0.6512575745582581, "learning_rate": 6.779123761053317e-05, "loss": 0.0387, "step": 41500 }, { "grad_norm": 0.4075067639350891, "learning_rate": 6.777578409120826e-05, "loss": 0.0539, "step": 41510 }, { "grad_norm": 0.7265503406524658, "learning_rate": 6.776032862795033e-05, "loss": 0.0521, "step": 41520 }, { "grad_norm": 0.6389572620391846, "learning_rate": 6.774487122244956e-05, "loss": 0.0486, "step": 41530 }, { "grad_norm": 0.40463778376579285, "learning_rate": 6.772941187639637e-05, "loss": 0.0361, "step": 41540 }, { "grad_norm": 0.825369656085968, "learning_rate": 6.771395059148134e-05, "loss": 0.0506, "step": 41550 }, { "grad_norm": 0.45304396748542786, "learning_rate": 6.76984873693953e-05, "loss": 0.0496, "step": 41560 }, { "grad_norm": 0.637457013130188, "learning_rate": 6.768302221182931e-05, "loss": 0.0609, "step": 41570 }, { "grad_norm": 0.7574993371963501, "learning_rate": 6.766755512047457e-05, "loss": 0.0641, "step": 41580 }, { "grad_norm": 0.43504804372787476, "learning_rate": 6.765208609702259e-05, "loss": 0.0491, "step": 41590 }, { "grad_norm": 0.45982179045677185, "learning_rate": 6.763661514316499e-05, "loss": 0.0415, "step": 41600 }, { "grad_norm": 0.5586440563201904, "learning_rate": 6.76211422605937e-05, "loss": 0.0556, "step": 41610 }, { "grad_norm": 0.4943467080593109, "learning_rate": 6.760566745100076e-05, "loss": 0.0456, "step": 41620 }, { "grad_norm": 1.7729343175888062, "learning_rate": 6.75901907160785e-05, "loss": 0.0554, "step": 41630 }, { "grad_norm": 0.4876921772956848, "learning_rate": 6.757471205751943e-05, "loss": 0.0452, "step": 41640 }, { "grad_norm": 0.5827934741973877, "learning_rate": 6.75592314770163e-05, "loss": 0.0515, "step": 41650 }, { "grad_norm": 0.45357435941696167, "learning_rate": 6.7543748976262e-05, "loss": 0.0574, "step": 41660 }, { "grad_norm": 0.2784107029438019, "learning_rate": 6.752826455694968e-05, "loss": 0.0532, "step": 41670 }, { "grad_norm": 0.7233455181121826, "learning_rate": 6.751277822077271e-05, "loss": 0.0494, "step": 41680 }, { "grad_norm": 0.307358056306839, "learning_rate": 6.749728996942465e-05, "loss": 0.0466, "step": 41690 }, { "grad_norm": 0.294115275144577, "learning_rate": 6.748179980459924e-05, "loss": 0.0415, "step": 41700 }, { "grad_norm": 0.4032871425151825, "learning_rate": 6.746630772799052e-05, "loss": 0.0396, "step": 41710 }, { "grad_norm": 0.3779587745666504, "learning_rate": 6.745081374129262e-05, "loss": 0.0462, "step": 41720 }, { "grad_norm": 0.4799720048904419, "learning_rate": 6.743531784619998e-05, "loss": 0.0574, "step": 41730 }, { "grad_norm": 0.6716283559799194, "learning_rate": 6.741982004440719e-05, "loss": 0.0539, "step": 41740 }, { "grad_norm": 0.5256617665290833, "learning_rate": 6.740432033760907e-05, "loss": 0.044, "step": 41750 }, { "grad_norm": 1.3732578754425049, "learning_rate": 6.738881872750066e-05, "loss": 0.0475, "step": 41760 }, { "grad_norm": 0.4653773605823517, "learning_rate": 6.737331521577715e-05, "loss": 0.0515, "step": 41770 }, { "grad_norm": 0.5856157541275024, "learning_rate": 6.735780980413403e-05, "loss": 0.0543, "step": 41780 }, { "grad_norm": 0.5309376120567322, "learning_rate": 6.734230249426692e-05, "loss": 0.0592, "step": 41790 }, { "grad_norm": 0.6090170741081238, "learning_rate": 6.732679328787168e-05, "loss": 0.0498, "step": 41800 }, { "grad_norm": 0.30082786083221436, "learning_rate": 6.731128218664438e-05, "loss": 0.0385, "step": 41810 }, { "grad_norm": 0.40513184666633606, "learning_rate": 6.729576919228129e-05, "loss": 0.0371, "step": 41820 }, { "grad_norm": 0.47863131761550903, "learning_rate": 6.728025430647888e-05, "loss": 0.042, "step": 41830 }, { "grad_norm": 0.559073269367218, "learning_rate": 6.726473753093383e-05, "loss": 0.0446, "step": 41840 }, { "grad_norm": 0.5796008110046387, "learning_rate": 6.724921886734305e-05, "loss": 0.0474, "step": 41850 }, { "grad_norm": 0.4678763151168823, "learning_rate": 6.723369831740363e-05, "loss": 0.0582, "step": 41860 }, { "grad_norm": 0.3091990053653717, "learning_rate": 6.721817588281288e-05, "loss": 0.0448, "step": 41870 }, { "grad_norm": 0.8702201247215271, "learning_rate": 6.720265156526828e-05, "loss": 0.0476, "step": 41880 }, { "grad_norm": 0.9681878685951233, "learning_rate": 6.718712536646758e-05, "loss": 0.0528, "step": 41890 }, { "grad_norm": 0.6767218708992004, "learning_rate": 6.71715972881087e-05, "loss": 0.0573, "step": 41900 }, { "grad_norm": 0.46367910504341125, "learning_rate": 6.715606733188974e-05, "loss": 0.048, "step": 41910 }, { "grad_norm": 0.5765076875686646, "learning_rate": 6.714053549950909e-05, "loss": 0.052, "step": 41920 }, { "grad_norm": 0.7274192571640015, "learning_rate": 6.71250017926652e-05, "loss": 0.0471, "step": 41930 }, { "grad_norm": 0.852276623249054, "learning_rate": 6.71094662130569e-05, "loss": 0.043, "step": 41940 }, { "grad_norm": 1.8632532358169556, "learning_rate": 6.709392876238307e-05, "loss": 0.0468, "step": 41950 }, { "grad_norm": 0.4202445447444916, "learning_rate": 6.707838944234289e-05, "loss": 0.0614, "step": 41960 }, { "grad_norm": 0.44172126054763794, "learning_rate": 6.706284825463574e-05, "loss": 0.0453, "step": 41970 }, { "grad_norm": 0.5642414093017578, "learning_rate": 6.704730520096114e-05, "loss": 0.0446, "step": 41980 }, { "grad_norm": 0.5233088731765747, "learning_rate": 6.703176028301888e-05, "loss": 0.0425, "step": 41990 }, { "grad_norm": 0.37714263796806335, "learning_rate": 6.701621350250892e-05, "loss": 0.0431, "step": 42000 }, { "grad_norm": 1.4833728075027466, "learning_rate": 6.700066486113144e-05, "loss": 0.0448, "step": 42010 }, { "grad_norm": 0.4232681095600128, "learning_rate": 6.69851143605868e-05, "loss": 0.0502, "step": 42020 }, { "grad_norm": 0.3801524043083191, "learning_rate": 6.696956200257561e-05, "loss": 0.0471, "step": 42030 }, { "grad_norm": 0.361463338136673, "learning_rate": 6.695400778879863e-05, "loss": 0.0515, "step": 42040 }, { "grad_norm": 0.9066968560218811, "learning_rate": 6.693845172095683e-05, "loss": 0.0642, "step": 42050 }, { "grad_norm": 0.46255865693092346, "learning_rate": 6.692289380075142e-05, "loss": 0.0489, "step": 42060 }, { "grad_norm": 0.8026439547538757, "learning_rate": 6.690733402988379e-05, "loss": 0.0418, "step": 42070 }, { "grad_norm": 0.6499160528182983, "learning_rate": 6.689177241005553e-05, "loss": 0.0567, "step": 42080 }, { "grad_norm": 0.5294286012649536, "learning_rate": 6.687620894296844e-05, "loss": 0.0497, "step": 42090 }, { "grad_norm": 0.5032377243041992, "learning_rate": 6.686064363032451e-05, "loss": 0.0504, "step": 42100 }, { "grad_norm": 0.6279973387718201, "learning_rate": 6.684507647382596e-05, "loss": 0.0488, "step": 42110 }, { "grad_norm": 0.47295042872428894, "learning_rate": 6.682950747517513e-05, "loss": 0.0536, "step": 42120 }, { "grad_norm": 0.5109544992446899, "learning_rate": 6.68139366360747e-05, "loss": 0.0464, "step": 42130 }, { "grad_norm": 0.5334512591362, "learning_rate": 6.679836395822744e-05, "loss": 0.0405, "step": 42140 }, { "grad_norm": 0.41577407717704773, "learning_rate": 6.678278944333633e-05, "loss": 0.0387, "step": 42150 }, { "grad_norm": 0.6562377214431763, "learning_rate": 6.676721309310462e-05, "loss": 0.0388, "step": 42160 }, { "grad_norm": 0.6535090208053589, "learning_rate": 6.67516349092357e-05, "loss": 0.0493, "step": 42170 }, { "grad_norm": 0.3391488790512085, "learning_rate": 6.673605489343315e-05, "loss": 0.0405, "step": 42180 }, { "grad_norm": 0.5062974095344543, "learning_rate": 6.672047304740081e-05, "loss": 0.0418, "step": 42190 }, { "grad_norm": 0.4717523157596588, "learning_rate": 6.670488937284268e-05, "loss": 0.0443, "step": 42200 }, { "grad_norm": 0.5523738265037537, "learning_rate": 6.668930387146294e-05, "loss": 0.0473, "step": 42210 }, { "grad_norm": 0.4183468520641327, "learning_rate": 6.667371654496605e-05, "loss": 0.0447, "step": 42220 }, { "grad_norm": 0.46027565002441406, "learning_rate": 6.665812739505659e-05, "loss": 0.0501, "step": 42230 }, { "grad_norm": 0.455608069896698, "learning_rate": 6.664253642343935e-05, "loss": 0.0539, "step": 42240 }, { "grad_norm": 0.5985912084579468, "learning_rate": 6.662694363181935e-05, "loss": 0.0442, "step": 42250 }, { "grad_norm": 0.4907185137271881, "learning_rate": 6.66113490219018e-05, "loss": 0.0529, "step": 42260 }, { "grad_norm": 0.37135130167007446, "learning_rate": 6.659575259539207e-05, "loss": 0.0375, "step": 42270 }, { "grad_norm": 0.5399948954582214, "learning_rate": 6.65801543539958e-05, "loss": 0.0495, "step": 42280 }, { "grad_norm": 0.5647937059402466, "learning_rate": 6.65645542994188e-05, "loss": 0.0446, "step": 42290 }, { "grad_norm": 0.3042766749858856, "learning_rate": 6.654895243336702e-05, "loss": 0.0326, "step": 42300 }, { "grad_norm": 0.43343880772590637, "learning_rate": 6.65333487575467e-05, "loss": 0.0345, "step": 42310 }, { "grad_norm": 0.3235417604446411, "learning_rate": 6.651774327366422e-05, "loss": 0.039, "step": 42320 }, { "grad_norm": 0.6108757853507996, "learning_rate": 6.650213598342613e-05, "loss": 0.0527, "step": 42330 }, { "grad_norm": 0.4486720860004425, "learning_rate": 6.648652688853931e-05, "loss": 0.0521, "step": 42340 }, { "grad_norm": 0.556214451789856, "learning_rate": 6.647091599071066e-05, "loss": 0.057, "step": 42350 }, { "grad_norm": 0.44094449281692505, "learning_rate": 6.645530329164743e-05, "loss": 0.0576, "step": 42360 }, { "grad_norm": 0.38742485642433167, "learning_rate": 6.643968879305694e-05, "loss": 0.0462, "step": 42370 }, { "grad_norm": 0.42482125759124756, "learning_rate": 6.642407249664683e-05, "loss": 0.0395, "step": 42380 }, { "grad_norm": 0.5913671851158142, "learning_rate": 6.640845440412483e-05, "loss": 0.0474, "step": 42390 }, { "grad_norm": 0.4470035135746002, "learning_rate": 6.639283451719893e-05, "loss": 0.0445, "step": 42400 }, { "grad_norm": 0.6866008043289185, "learning_rate": 6.637721283757729e-05, "loss": 0.0422, "step": 42410 }, { "grad_norm": 1.0306240320205688, "learning_rate": 6.636158936696828e-05, "loss": 0.0471, "step": 42420 }, { "grad_norm": 0.7928833365440369, "learning_rate": 6.634596410708047e-05, "loss": 0.0501, "step": 42430 }, { "grad_norm": 0.624254584312439, "learning_rate": 6.633033705962257e-05, "loss": 0.0497, "step": 42440 }, { "grad_norm": 0.6891784071922302, "learning_rate": 6.631470822630359e-05, "loss": 0.05, "step": 42450 }, { "grad_norm": 0.8416476845741272, "learning_rate": 6.629907760883263e-05, "loss": 0.0566, "step": 42460 }, { "grad_norm": 0.531444787979126, "learning_rate": 6.628344520891905e-05, "loss": 0.0448, "step": 42470 }, { "grad_norm": 0.3837335705757141, "learning_rate": 6.62678110282724e-05, "loss": 0.0445, "step": 42480 }, { "grad_norm": 0.4446106255054474, "learning_rate": 6.625217506860237e-05, "loss": 0.0544, "step": 42490 }, { "grad_norm": 0.31364715099334717, "learning_rate": 6.623653733161892e-05, "loss": 0.0521, "step": 42500 }, { "grad_norm": 0.4545065760612488, "learning_rate": 6.622089781903217e-05, "loss": 0.05, "step": 42510 }, { "grad_norm": 0.5033398866653442, "learning_rate": 6.620525653255241e-05, "loss": 0.0458, "step": 42520 }, { "grad_norm": 0.5897138118743896, "learning_rate": 6.618961347389015e-05, "loss": 0.0402, "step": 42530 }, { "grad_norm": 0.34999603033065796, "learning_rate": 6.617396864475613e-05, "loss": 0.0375, "step": 42540 }, { "grad_norm": 0.44761568307876587, "learning_rate": 6.61583220468612e-05, "loss": 0.0392, "step": 42550 }, { "grad_norm": 0.3239976465702057, "learning_rate": 6.614267368191645e-05, "loss": 0.0405, "step": 42560 }, { "grad_norm": 0.6322676539421082, "learning_rate": 6.61270235516332e-05, "loss": 0.0456, "step": 42570 }, { "grad_norm": 0.2856235206127167, "learning_rate": 6.611137165772287e-05, "loss": 0.0364, "step": 42580 }, { "grad_norm": 0.4417538046836853, "learning_rate": 6.60957180018972e-05, "loss": 0.0511, "step": 42590 }, { "grad_norm": 0.45046862959861755, "learning_rate": 6.608006258586797e-05, "loss": 0.0405, "step": 42600 }, { "grad_norm": 0.29479965567588806, "learning_rate": 6.606440541134731e-05, "loss": 0.0427, "step": 42610 }, { "grad_norm": 0.31643831729888916, "learning_rate": 6.60487464800474e-05, "loss": 0.0534, "step": 42620 }, { "grad_norm": 0.6900262832641602, "learning_rate": 6.603308579368071e-05, "loss": 0.0564, "step": 42630 }, { "grad_norm": 0.5856921672821045, "learning_rate": 6.601742335395987e-05, "loss": 0.0483, "step": 42640 }, { "grad_norm": 0.8680187463760376, "learning_rate": 6.600175916259769e-05, "loss": 0.0486, "step": 42650 }, { "grad_norm": 0.4592784345149994, "learning_rate": 6.598609322130718e-05, "loss": 0.039, "step": 42660 }, { "grad_norm": 0.4948892593383789, "learning_rate": 6.597042553180155e-05, "loss": 0.0365, "step": 42670 }, { "grad_norm": 0.5226889252662659, "learning_rate": 6.59547560957942e-05, "loss": 0.0483, "step": 42680 }, { "grad_norm": 0.4460212290287018, "learning_rate": 6.59390849149987e-05, "loss": 0.0536, "step": 42690 }, { "grad_norm": 0.5337955951690674, "learning_rate": 6.592341199112886e-05, "loss": 0.0431, "step": 42700 }, { "grad_norm": 0.3904826045036316, "learning_rate": 6.590773732589861e-05, "loss": 0.0423, "step": 42710 }, { "grad_norm": 0.48650288581848145, "learning_rate": 6.589206092102213e-05, "loss": 0.0514, "step": 42720 }, { "grad_norm": 0.25950780510902405, "learning_rate": 6.587638277821375e-05, "loss": 0.0332, "step": 42730 }, { "grad_norm": 0.42923274636268616, "learning_rate": 6.586070289918803e-05, "loss": 0.0421, "step": 42740 }, { "grad_norm": 0.5586493015289307, "learning_rate": 6.584502128565968e-05, "loss": 0.0465, "step": 42750 }, { "grad_norm": 0.42897629737854004, "learning_rate": 6.582933793934363e-05, "loss": 0.0489, "step": 42760 }, { "grad_norm": 0.40837764739990234, "learning_rate": 6.581365286195498e-05, "loss": 0.0426, "step": 42770 }, { "grad_norm": 0.32762718200683594, "learning_rate": 6.579796605520903e-05, "loss": 0.0414, "step": 42780 }, { "grad_norm": 0.471667617559433, "learning_rate": 6.578227752082127e-05, "loss": 0.0506, "step": 42790 }, { "grad_norm": 0.517170250415802, "learning_rate": 6.576658726050735e-05, "loss": 0.043, "step": 42800 }, { "grad_norm": 0.29408445954322815, "learning_rate": 6.575089527598316e-05, "loss": 0.033, "step": 42810 }, { "grad_norm": 0.32001233100891113, "learning_rate": 6.573520156896476e-05, "loss": 0.0486, "step": 42820 }, { "grad_norm": 0.45322608947753906, "learning_rate": 6.571950614116835e-05, "loss": 0.0393, "step": 42830 }, { "grad_norm": 0.5635151863098145, "learning_rate": 6.570380899431039e-05, "loss": 0.0612, "step": 42840 }, { "grad_norm": 0.6794895529747009, "learning_rate": 6.568811013010749e-05, "loss": 0.0466, "step": 42850 }, { "grad_norm": 0.44204002618789673, "learning_rate": 6.567240955027645e-05, "loss": 0.0523, "step": 42860 }, { "grad_norm": 0.47942087054252625, "learning_rate": 6.565670725653427e-05, "loss": 0.0349, "step": 42870 }, { "grad_norm": 0.4137970507144928, "learning_rate": 6.564100325059812e-05, "loss": 0.0391, "step": 42880 }, { "grad_norm": 0.44881680607795715, "learning_rate": 6.562529753418538e-05, "loss": 0.0407, "step": 42890 }, { "grad_norm": 0.4803714156150818, "learning_rate": 6.56095901090136e-05, "loss": 0.0428, "step": 42900 }, { "grad_norm": 1.03565514087677, "learning_rate": 6.55938809768005e-05, "loss": 0.0399, "step": 42910 }, { "grad_norm": 0.6565727591514587, "learning_rate": 6.557817013926402e-05, "loss": 0.0483, "step": 42920 }, { "grad_norm": 0.47132089734077454, "learning_rate": 6.556245759812227e-05, "loss": 0.0524, "step": 42930 }, { "grad_norm": 0.5990739464759827, "learning_rate": 6.554674335509357e-05, "loss": 0.0466, "step": 42940 }, { "grad_norm": 0.37333762645721436, "learning_rate": 6.553102741189638e-05, "loss": 0.0377, "step": 42950 }, { "grad_norm": 0.40394991636276245, "learning_rate": 6.55153097702494e-05, "loss": 0.0503, "step": 42960 }, { "grad_norm": 0.46514657139778137, "learning_rate": 6.549959043187146e-05, "loss": 0.0462, "step": 42970 }, { "grad_norm": 0.9419865012168884, "learning_rate": 6.54838693984816e-05, "loss": 0.0467, "step": 42980 }, { "grad_norm": 0.6330208778381348, "learning_rate": 6.546814667179909e-05, "loss": 0.046, "step": 42990 }, { "grad_norm": 0.7476987242698669, "learning_rate": 6.545242225354328e-05, "loss": 0.0573, "step": 43000 }, { "grad_norm": 0.500153660774231, "learning_rate": 6.543669614543383e-05, "loss": 0.0548, "step": 43010 }, { "grad_norm": 0.516575276851654, "learning_rate": 6.542096834919049e-05, "loss": 0.0517, "step": 43020 }, { "grad_norm": 0.89825439453125, "learning_rate": 6.540523886653324e-05, "loss": 0.0473, "step": 43030 }, { "grad_norm": 0.3669498562812805, "learning_rate": 6.538950769918222e-05, "loss": 0.0441, "step": 43040 }, { "grad_norm": 0.5801528692245483, "learning_rate": 6.537377484885779e-05, "loss": 0.0379, "step": 43050 }, { "grad_norm": 0.7625153660774231, "learning_rate": 6.535804031728045e-05, "loss": 0.0475, "step": 43060 }, { "grad_norm": 0.4279743432998657, "learning_rate": 6.534230410617092e-05, "loss": 0.0525, "step": 43070 }, { "grad_norm": 0.7683597803115845, "learning_rate": 6.532656621725007e-05, "loss": 0.0488, "step": 43080 }, { "grad_norm": 0.5180913805961609, "learning_rate": 6.531082665223898e-05, "loss": 0.0527, "step": 43090 }, { "grad_norm": 0.3370841443538666, "learning_rate": 6.529508541285889e-05, "loss": 0.0512, "step": 43100 }, { "grad_norm": 0.6702005863189697, "learning_rate": 6.527934250083128e-05, "loss": 0.0608, "step": 43110 }, { "grad_norm": 0.42812442779541016, "learning_rate": 6.526359791787775e-05, "loss": 0.0406, "step": 43120 }, { "grad_norm": 0.32840800285339355, "learning_rate": 6.524785166572009e-05, "loss": 0.0438, "step": 43130 }, { "grad_norm": 0.4801940321922302, "learning_rate": 6.523210374608029e-05, "loss": 0.0474, "step": 43140 }, { "grad_norm": 0.7641382217407227, "learning_rate": 6.521635416068054e-05, "loss": 0.0412, "step": 43150 }, { "grad_norm": 0.3808894455432892, "learning_rate": 6.520060291124317e-05, "loss": 0.0665, "step": 43160 }, { "grad_norm": 0.29289576411247253, "learning_rate": 6.518484999949072e-05, "loss": 0.0473, "step": 43170 }, { "grad_norm": 0.4267284572124481, "learning_rate": 6.516909542714591e-05, "loss": 0.0482, "step": 43180 }, { "grad_norm": 0.6390981078147888, "learning_rate": 6.515333919593161e-05, "loss": 0.0455, "step": 43190 }, { "grad_norm": 0.43766850233078003, "learning_rate": 6.513758130757094e-05, "loss": 0.0455, "step": 43200 }, { "grad_norm": 0.4717312455177307, "learning_rate": 6.512182176378713e-05, "loss": 0.0513, "step": 43210 }, { "grad_norm": 0.5542662739753723, "learning_rate": 6.510606056630362e-05, "loss": 0.0454, "step": 43220 }, { "grad_norm": 0.4049275815486908, "learning_rate": 6.509029771684405e-05, "loss": 0.0426, "step": 43230 }, { "grad_norm": 0.7903389930725098, "learning_rate": 6.50745332171322e-05, "loss": 0.0486, "step": 43240 }, { "grad_norm": 0.617761492729187, "learning_rate": 6.505876706889207e-05, "loss": 0.0629, "step": 43250 }, { "grad_norm": 0.40032774209976196, "learning_rate": 6.504299927384781e-05, "loss": 0.05, "step": 43260 }, { "grad_norm": 0.6342506408691406, "learning_rate": 6.502722983372378e-05, "loss": 0.0434, "step": 43270 }, { "grad_norm": 0.3550586402416229, "learning_rate": 6.501145875024446e-05, "loss": 0.0492, "step": 43280 }, { "grad_norm": 0.5308623313903809, "learning_rate": 6.499568602513459e-05, "loss": 0.0501, "step": 43290 }, { "grad_norm": 0.4627743363380432, "learning_rate": 6.497991166011903e-05, "loss": 0.0392, "step": 43300 }, { "grad_norm": 0.3203416168689728, "learning_rate": 6.49641356569229e-05, "loss": 0.0375, "step": 43310 }, { "grad_norm": 0.41798919439315796, "learning_rate": 6.494835801727135e-05, "loss": 0.0424, "step": 43320 }, { "grad_norm": 0.4837464690208435, "learning_rate": 6.493257874288986e-05, "loss": 0.0439, "step": 43330 }, { "grad_norm": 0.4572618305683136, "learning_rate": 6.491679783550399e-05, "loss": 0.0474, "step": 43340 }, { "grad_norm": 0.4782850444316864, "learning_rate": 6.490101529683957e-05, "loss": 0.0427, "step": 43350 }, { "grad_norm": 0.2933390736579895, "learning_rate": 6.488523112862248e-05, "loss": 0.0476, "step": 43360 }, { "grad_norm": 0.4126524329185486, "learning_rate": 6.486944533257891e-05, "loss": 0.0504, "step": 43370 }, { "grad_norm": 0.578149139881134, "learning_rate": 6.485365791043517e-05, "loss": 0.0512, "step": 43380 }, { "grad_norm": 0.6329920887947083, "learning_rate": 6.483786886391772e-05, "loss": 0.0509, "step": 43390 }, { "grad_norm": 0.3979750871658325, "learning_rate": 6.482207819475323e-05, "loss": 0.0382, "step": 43400 }, { "grad_norm": 0.4470186233520508, "learning_rate": 6.480628590466857e-05, "loss": 0.0323, "step": 43410 }, { "grad_norm": 0.7287557721138, "learning_rate": 6.479049199539074e-05, "loss": 0.0479, "step": 43420 }, { "grad_norm": 0.45798128843307495, "learning_rate": 6.477469646864693e-05, "loss": 0.0475, "step": 43430 }, { "grad_norm": 0.41297653317451477, "learning_rate": 6.475889932616454e-05, "loss": 0.0469, "step": 43440 }, { "grad_norm": 0.4419213533401489, "learning_rate": 6.474310056967111e-05, "loss": 0.0493, "step": 43450 }, { "grad_norm": 0.500687301158905, "learning_rate": 6.472730020089437e-05, "loss": 0.0618, "step": 43460 }, { "grad_norm": 0.4078689515590668, "learning_rate": 6.471149822156222e-05, "loss": 0.0429, "step": 43470 }, { "grad_norm": 0.8185064196586609, "learning_rate": 6.469569463340275e-05, "loss": 0.0377, "step": 43480 }, { "grad_norm": 0.4825393557548523, "learning_rate": 6.467988943814418e-05, "loss": 0.0471, "step": 43490 }, { "grad_norm": 0.5702275633811951, "learning_rate": 6.4664082637515e-05, "loss": 0.0423, "step": 43500 }, { "grad_norm": 0.5194264650344849, "learning_rate": 6.464827423324377e-05, "loss": 0.0478, "step": 43510 }, { "grad_norm": 0.3020467460155487, "learning_rate": 6.463246422705931e-05, "loss": 0.0419, "step": 43520 }, { "grad_norm": 0.4833356440067291, "learning_rate": 6.461665262069055e-05, "loss": 0.0539, "step": 43530 }, { "grad_norm": 0.2719831168651581, "learning_rate": 6.460083941586665e-05, "loss": 0.0456, "step": 43540 }, { "grad_norm": 0.6787594556808472, "learning_rate": 6.45850246143169e-05, "loss": 0.0402, "step": 43550 }, { "grad_norm": 0.6517743468284607, "learning_rate": 6.456920821777077e-05, "loss": 0.0459, "step": 43560 }, { "grad_norm": 0.4935820400714874, "learning_rate": 6.455339022795796e-05, "loss": 0.0512, "step": 43570 }, { "grad_norm": 0.5555604696273804, "learning_rate": 6.453757064660825e-05, "loss": 0.0415, "step": 43580 }, { "grad_norm": 0.4517861604690552, "learning_rate": 6.452174947545169e-05, "loss": 0.0304, "step": 43590 }, { "grad_norm": 0.3993522524833679, "learning_rate": 6.450592671621842e-05, "loss": 0.0402, "step": 43600 }, { "grad_norm": 0.6087890863418579, "learning_rate": 6.449010237063882e-05, "loss": 0.0448, "step": 43610 }, { "grad_norm": 0.4547211825847626, "learning_rate": 6.447427644044343e-05, "loss": 0.0408, "step": 43620 }, { "grad_norm": 0.48303624987602234, "learning_rate": 6.44584489273629e-05, "loss": 0.0422, "step": 43630 }, { "grad_norm": 0.4665326774120331, "learning_rate": 6.444261983312817e-05, "loss": 0.0406, "step": 43640 }, { "grad_norm": 0.522079586982727, "learning_rate": 6.442678915947023e-05, "loss": 0.0449, "step": 43650 }, { "grad_norm": 0.4051268398761749, "learning_rate": 6.441095690812032e-05, "loss": 0.0405, "step": 43660 }, { "grad_norm": 0.5572646260261536, "learning_rate": 6.439512308080982e-05, "loss": 0.0429, "step": 43670 }, { "grad_norm": 0.2723163664340973, "learning_rate": 6.43792876792703e-05, "loss": 0.0436, "step": 43680 }, { "grad_norm": 0.32988837361335754, "learning_rate": 6.43634507052335e-05, "loss": 0.046, "step": 43690 }, { "grad_norm": 0.7722760438919067, "learning_rate": 6.434761216043135e-05, "loss": 0.051, "step": 43700 }, { "grad_norm": 1.1117886304855347, "learning_rate": 6.433177204659589e-05, "loss": 0.0528, "step": 43710 }, { "grad_norm": 0.48023656010627747, "learning_rate": 6.431593036545938e-05, "loss": 0.0426, "step": 43720 }, { "grad_norm": 0.5890105962753296, "learning_rate": 6.430008711875426e-05, "loss": 0.0359, "step": 43730 }, { "grad_norm": 1.2753547430038452, "learning_rate": 6.42842423082131e-05, "loss": 0.0488, "step": 43740 }, { "grad_norm": 0.502998948097229, "learning_rate": 6.426839593556868e-05, "loss": 0.0353, "step": 43750 }, { "grad_norm": 0.413847953081131, "learning_rate": 6.425254800255392e-05, "loss": 0.0446, "step": 43760 }, { "grad_norm": 0.4010234773159027, "learning_rate": 6.423669851090196e-05, "loss": 0.0513, "step": 43770 }, { "grad_norm": 0.49099719524383545, "learning_rate": 6.422084746234604e-05, "loss": 0.0368, "step": 43780 }, { "grad_norm": 0.5462241768836975, "learning_rate": 6.420499485861965e-05, "loss": 0.032, "step": 43790 }, { "grad_norm": 0.4307573437690735, "learning_rate": 6.418914070145636e-05, "loss": 0.0554, "step": 43800 }, { "grad_norm": 0.5573440194129944, "learning_rate": 6.417328499258996e-05, "loss": 0.0411, "step": 43810 }, { "grad_norm": 0.5251418948173523, "learning_rate": 6.415742773375446e-05, "loss": 0.0411, "step": 43820 }, { "grad_norm": 0.3708331882953644, "learning_rate": 6.414156892668393e-05, "loss": 0.0413, "step": 43830 }, { "grad_norm": 0.8293174505233765, "learning_rate": 6.412570857311267e-05, "loss": 0.0531, "step": 43840 }, { "grad_norm": 0.4451683461666107, "learning_rate": 6.410984667477518e-05, "loss": 0.0476, "step": 43850 }, { "grad_norm": 0.5277367234230042, "learning_rate": 6.409398323340607e-05, "loss": 0.0393, "step": 43860 }, { "grad_norm": 0.3298776149749756, "learning_rate": 6.407811825074012e-05, "loss": 0.0446, "step": 43870 }, { "grad_norm": 0.49312108755111694, "learning_rate": 6.406225172851234e-05, "loss": 0.0436, "step": 43880 }, { "grad_norm": 0.7189653515815735, "learning_rate": 6.404638366845786e-05, "loss": 0.047, "step": 43890 }, { "grad_norm": 0.4783216416835785, "learning_rate": 6.403051407231196e-05, "loss": 0.0495, "step": 43900 }, { "grad_norm": 0.7337607145309448, "learning_rate": 6.401464294181016e-05, "loss": 0.0481, "step": 43910 }, { "grad_norm": 0.46342015266418457, "learning_rate": 6.399877027868808e-05, "loss": 0.0417, "step": 43920 }, { "grad_norm": 0.4931900203227997, "learning_rate": 6.39828960846815e-05, "loss": 0.0462, "step": 43930 }, { "grad_norm": 0.4921559989452362, "learning_rate": 6.396702036152644e-05, "loss": 0.0352, "step": 43940 }, { "grad_norm": 0.4306677281856537, "learning_rate": 6.395114311095905e-05, "loss": 0.0504, "step": 43950 }, { "grad_norm": 0.4370947480201721, "learning_rate": 6.393526433471559e-05, "loss": 0.0436, "step": 43960 }, { "grad_norm": 0.41597017645835876, "learning_rate": 6.391938403453262e-05, "loss": 0.054, "step": 43970 }, { "grad_norm": 0.7000291347503662, "learning_rate": 6.390350221214671e-05, "loss": 0.054, "step": 43980 }, { "grad_norm": 0.9594206809997559, "learning_rate": 6.38876188692947e-05, "loss": 0.0607, "step": 43990 }, { "grad_norm": 0.8622502684593201, "learning_rate": 6.387173400771358e-05, "loss": 0.0477, "step": 44000 }, { "grad_norm": 0.48492953181266785, "learning_rate": 6.385584762914047e-05, "loss": 0.0433, "step": 44010 }, { "grad_norm": 0.330018550157547, "learning_rate": 6.38399597353127e-05, "loss": 0.0418, "step": 44020 }, { "grad_norm": 0.43930289149284363, "learning_rate": 6.382407032796775e-05, "loss": 0.0369, "step": 44030 }, { "grad_norm": 0.4145018458366394, "learning_rate": 6.380817940884325e-05, "loss": 0.0404, "step": 44040 }, { "grad_norm": 1.176414966583252, "learning_rate": 6.379228697967702e-05, "loss": 0.0494, "step": 44050 }, { "grad_norm": 0.4134065806865692, "learning_rate": 6.3776393042207e-05, "loss": 0.0429, "step": 44060 }, { "grad_norm": 0.44532421231269836, "learning_rate": 6.376049759817137e-05, "loss": 0.0482, "step": 44070 }, { "grad_norm": 0.3441155254840851, "learning_rate": 6.37446006493084e-05, "loss": 0.0498, "step": 44080 }, { "grad_norm": 0.5924969911575317, "learning_rate": 6.372870219735656e-05, "loss": 0.0568, "step": 44090 }, { "grad_norm": 0.7733631730079651, "learning_rate": 6.371280224405452e-05, "loss": 0.0453, "step": 44100 }, { "grad_norm": 0.5445915460586548, "learning_rate": 6.369690079114101e-05, "loss": 0.0432, "step": 44110 }, { "grad_norm": 2.4957451820373535, "learning_rate": 6.368099784035504e-05, "loss": 0.051, "step": 44120 }, { "grad_norm": 0.5280031561851501, "learning_rate": 6.366509339343572e-05, "loss": 0.0461, "step": 44130 }, { "grad_norm": 0.484849750995636, "learning_rate": 6.364918745212232e-05, "loss": 0.0436, "step": 44140 }, { "grad_norm": 0.5406026244163513, "learning_rate": 6.363328001815431e-05, "loss": 0.0409, "step": 44150 }, { "grad_norm": 0.46348094940185547, "learning_rate": 6.361737109327128e-05, "loss": 0.047, "step": 44160 }, { "grad_norm": 0.3230610489845276, "learning_rate": 6.360146067921301e-05, "loss": 0.0406, "step": 44170 }, { "grad_norm": 0.8853610754013062, "learning_rate": 6.35855487777195e-05, "loss": 0.0598, "step": 44180 }, { "grad_norm": 0.3015620708465576, "learning_rate": 6.356963539053076e-05, "loss": 0.0541, "step": 44190 }, { "grad_norm": 0.5067349672317505, "learning_rate": 6.355372051938711e-05, "loss": 0.0513, "step": 44200 }, { "grad_norm": 0.6237277388572693, "learning_rate": 6.353780416602894e-05, "loss": 0.043, "step": 44210 }, { "grad_norm": 0.41211315989494324, "learning_rate": 6.352188633219689e-05, "loss": 0.0402, "step": 44220 }, { "grad_norm": 0.5055344104766846, "learning_rate": 6.350596701963166e-05, "loss": 0.0469, "step": 44230 }, { "grad_norm": 0.578175961971283, "learning_rate": 6.349004623007419e-05, "loss": 0.0395, "step": 44240 }, { "grad_norm": 0.46632930636405945, "learning_rate": 6.347412396526555e-05, "loss": 0.0458, "step": 44250 }, { "grad_norm": 0.47617238759994507, "learning_rate": 6.345820022694696e-05, "loss": 0.0591, "step": 44260 }, { "grad_norm": 0.516084611415863, "learning_rate": 6.344227501685984e-05, "loss": 0.0438, "step": 44270 }, { "grad_norm": 0.7944274544715881, "learning_rate": 6.342634833674572e-05, "loss": 0.0483, "step": 44280 }, { "grad_norm": 0.6347430944442749, "learning_rate": 6.341042018834635e-05, "loss": 0.046, "step": 44290 }, { "grad_norm": 0.5103676915168762, "learning_rate": 6.339449057340359e-05, "loss": 0.043, "step": 44300 }, { "grad_norm": 0.317843496799469, "learning_rate": 6.337855949365945e-05, "loss": 0.0477, "step": 44310 }, { "grad_norm": 0.9671811461448669, "learning_rate": 6.336262695085619e-05, "loss": 0.0465, "step": 44320 }, { "grad_norm": 0.6419798731803894, "learning_rate": 6.334669294673612e-05, "loss": 0.0487, "step": 44330 }, { "grad_norm": 0.394119530916214, "learning_rate": 6.333075748304179e-05, "loss": 0.051, "step": 44340 }, { "grad_norm": 0.6597244739532471, "learning_rate": 6.331482056151585e-05, "loss": 0.0471, "step": 44350 }, { "grad_norm": 0.5168085098266602, "learning_rate": 6.329888218390117e-05, "loss": 0.0474, "step": 44360 }, { "grad_norm": 0.4452980160713196, "learning_rate": 6.328294235194072e-05, "loss": 0.0426, "step": 44370 }, { "grad_norm": 0.41411134600639343, "learning_rate": 6.326700106737765e-05, "loss": 0.0456, "step": 44380 }, { "grad_norm": 0.4299584627151489, "learning_rate": 6.325105833195531e-05, "loss": 0.0491, "step": 44390 }, { "grad_norm": 0.5252317190170288, "learning_rate": 6.323511414741715e-05, "loss": 0.042, "step": 44400 }, { "grad_norm": 0.5383924245834351, "learning_rate": 6.321916851550678e-05, "loss": 0.0611, "step": 44410 }, { "grad_norm": 1.0573145151138306, "learning_rate": 6.320322143796806e-05, "loss": 0.0654, "step": 44420 }, { "grad_norm": 0.36554279923439026, "learning_rate": 6.318727291654488e-05, "loss": 0.0473, "step": 44430 }, { "grad_norm": 0.5506241321563721, "learning_rate": 6.317132295298134e-05, "loss": 0.0385, "step": 44440 }, { "grad_norm": 0.33408501744270325, "learning_rate": 6.315537154902173e-05, "loss": 0.0379, "step": 44450 }, { "grad_norm": 0.9598536491394043, "learning_rate": 6.313941870641048e-05, "loss": 0.0478, "step": 44460 }, { "grad_norm": 0.47614458203315735, "learning_rate": 6.312346442689214e-05, "loss": 0.0563, "step": 44470 }, { "grad_norm": 1.0313963890075684, "learning_rate": 6.310750871221147e-05, "loss": 0.0443, "step": 44480 }, { "grad_norm": 0.43670642375946045, "learning_rate": 6.309155156411335e-05, "loss": 0.0421, "step": 44490 }, { "grad_norm": 0.36849507689476013, "learning_rate": 6.307559298434284e-05, "loss": 0.0418, "step": 44500 }, { "grad_norm": 0.4431203007698059, "learning_rate": 6.305963297464512e-05, "loss": 0.0411, "step": 44510 }, { "grad_norm": 0.4615723490715027, "learning_rate": 6.304367153676561e-05, "loss": 0.0378, "step": 44520 }, { "grad_norm": 0.5812388062477112, "learning_rate": 6.302770867244978e-05, "loss": 0.0418, "step": 44530 }, { "grad_norm": 0.6616706252098083, "learning_rate": 6.301174438344328e-05, "loss": 0.0463, "step": 44540 }, { "grad_norm": 0.5413603186607361, "learning_rate": 6.299577867149202e-05, "loss": 0.057, "step": 44550 }, { "grad_norm": 0.41317734122276306, "learning_rate": 6.297981153834192e-05, "loss": 0.043, "step": 44560 }, { "grad_norm": 0.3298666477203369, "learning_rate": 6.296384298573916e-05, "loss": 0.0577, "step": 44570 }, { "grad_norm": 0.4710899293422699, "learning_rate": 6.294787301543001e-05, "loss": 0.0695, "step": 44580 }, { "grad_norm": 0.43983784317970276, "learning_rate": 6.293190162916095e-05, "loss": 0.0439, "step": 44590 }, { "grad_norm": 0.403348833322525, "learning_rate": 6.291592882867855e-05, "loss": 0.0453, "step": 44600 }, { "grad_norm": 1.100657343864441, "learning_rate": 6.28999546157296e-05, "loss": 0.0469, "step": 44610 }, { "grad_norm": 0.38947898149490356, "learning_rate": 6.288397899206102e-05, "loss": 0.0426, "step": 44620 }, { "grad_norm": 0.7825836539268494, "learning_rate": 6.286800195941984e-05, "loss": 0.046, "step": 44630 }, { "grad_norm": 0.4163595139980316, "learning_rate": 6.285202351955334e-05, "loss": 0.0403, "step": 44640 }, { "grad_norm": 0.6983447670936584, "learning_rate": 6.283604367420887e-05, "loss": 0.0534, "step": 44650 }, { "grad_norm": 0.3817395567893982, "learning_rate": 6.282006242513394e-05, "loss": 0.0445, "step": 44660 }, { "grad_norm": 0.5060514807701111, "learning_rate": 6.280407977407628e-05, "loss": 0.0464, "step": 44670 }, { "grad_norm": 0.5500967502593994, "learning_rate": 6.27880957227837e-05, "loss": 0.0517, "step": 44680 }, { "grad_norm": 0.47624292969703674, "learning_rate": 6.27721102730042e-05, "loss": 0.0526, "step": 44690 }, { "grad_norm": 0.4940814673900604, "learning_rate": 6.275612342648592e-05, "loss": 0.0507, "step": 44700 }, { "grad_norm": 0.639575183391571, "learning_rate": 6.274013518497716e-05, "loss": 0.0489, "step": 44710 }, { "grad_norm": 0.3634425103664398, "learning_rate": 6.272414555022636e-05, "loss": 0.0404, "step": 44720 }, { "grad_norm": 0.5122812390327454, "learning_rate": 6.270815452398215e-05, "loss": 0.0461, "step": 44730 }, { "grad_norm": 0.7466108798980713, "learning_rate": 6.269216210799326e-05, "loss": 0.046, "step": 44740 }, { "grad_norm": 2.1299140453338623, "learning_rate": 6.26761683040086e-05, "loss": 0.056, "step": 44750 }, { "grad_norm": 0.3974746763706207, "learning_rate": 6.266017311377723e-05, "loss": 0.0498, "step": 44760 }, { "grad_norm": 0.46296724677085876, "learning_rate": 6.264417653904839e-05, "loss": 0.06, "step": 44770 }, { "grad_norm": 0.5915144085884094, "learning_rate": 6.262817858157139e-05, "loss": 0.0521, "step": 44780 }, { "grad_norm": 0.5624372959136963, "learning_rate": 6.261217924309576e-05, "loss": 0.0636, "step": 44790 }, { "grad_norm": 0.6441972255706787, "learning_rate": 6.259617852537118e-05, "loss": 0.0526, "step": 44800 }, { "grad_norm": 0.5759759545326233, "learning_rate": 6.258017643014747e-05, "loss": 0.0482, "step": 44810 }, { "grad_norm": 0.6025789380073547, "learning_rate": 6.256417295917456e-05, "loss": 0.0578, "step": 44820 }, { "grad_norm": 0.524732232093811, "learning_rate": 6.254816811420258e-05, "loss": 0.0684, "step": 44830 }, { "grad_norm": 0.7115709781646729, "learning_rate": 6.253216189698183e-05, "loss": 0.0499, "step": 44840 }, { "grad_norm": 0.3178532123565674, "learning_rate": 6.251615430926267e-05, "loss": 0.0382, "step": 44850 }, { "grad_norm": 0.46252012252807617, "learning_rate": 6.25001453527957e-05, "loss": 0.045, "step": 44860 }, { "grad_norm": 0.535433292388916, "learning_rate": 6.248413502933164e-05, "loss": 0.046, "step": 44870 }, { "grad_norm": 0.7071460485458374, "learning_rate": 6.246812334062133e-05, "loss": 0.0425, "step": 44880 }, { "grad_norm": 0.5320231318473816, "learning_rate": 6.245211028841579e-05, "loss": 0.0573, "step": 44890 }, { "grad_norm": 0.5114229917526245, "learning_rate": 6.24360958744662e-05, "loss": 0.045, "step": 44900 }, { "grad_norm": 0.4044419825077057, "learning_rate": 6.242008010052387e-05, "loss": 0.0482, "step": 44910 }, { "grad_norm": 0.5858936905860901, "learning_rate": 6.240406296834024e-05, "loss": 0.0535, "step": 44920 }, { "grad_norm": 0.41233953833580017, "learning_rate": 6.238804447966694e-05, "loss": 0.0488, "step": 44930 }, { "grad_norm": 0.6110755801200867, "learning_rate": 6.237202463625573e-05, "loss": 0.0385, "step": 44940 }, { "grad_norm": 0.4144178330898285, "learning_rate": 6.235600343985848e-05, "loss": 0.0472, "step": 44950 }, { "grad_norm": 1.1042191982269287, "learning_rate": 6.233998089222729e-05, "loss": 0.0422, "step": 44960 }, { "grad_norm": 0.5140751600265503, "learning_rate": 6.232395699511433e-05, "loss": 0.0472, "step": 44970 }, { "grad_norm": 0.4388456642627716, "learning_rate": 6.230793175027199e-05, "loss": 0.0618, "step": 44980 }, { "grad_norm": 0.5410519242286682, "learning_rate": 6.22919051594527e-05, "loss": 0.0458, "step": 44990 }, { "grad_norm": 0.8097476363182068, "learning_rate": 6.227587722440917e-05, "loss": 0.0586, "step": 45000 }, { "grad_norm": 0.33252087235450745, "learning_rate": 6.225984794689414e-05, "loss": 0.0418, "step": 45010 }, { "grad_norm": 0.4151979982852936, "learning_rate": 6.22438173286606e-05, "loss": 0.0445, "step": 45020 }, { "grad_norm": 0.48876434564590454, "learning_rate": 6.222778537146159e-05, "loss": 0.043, "step": 45030 }, { "grad_norm": 0.3814552128314972, "learning_rate": 6.221175207705037e-05, "loss": 0.0447, "step": 45040 }, { "grad_norm": 0.6537590622901917, "learning_rate": 6.219571744718027e-05, "loss": 0.0421, "step": 45050 }, { "grad_norm": 0.34347447752952576, "learning_rate": 6.217968148360487e-05, "loss": 0.0449, "step": 45060 }, { "grad_norm": 0.3304353356361389, "learning_rate": 6.216364418807782e-05, "loss": 0.0497, "step": 45070 }, { "grad_norm": 0.5265771746635437, "learning_rate": 6.214760556235292e-05, "loss": 0.0437, "step": 45080 }, { "grad_norm": 0.4410964250564575, "learning_rate": 6.213156560818414e-05, "loss": 0.04, "step": 45090 }, { "grad_norm": 0.5083997845649719, "learning_rate": 6.211552432732559e-05, "loss": 0.0431, "step": 45100 }, { "grad_norm": 0.5802305936813354, "learning_rate": 6.20994817215315e-05, "loss": 0.0484, "step": 45110 }, { "grad_norm": 0.4620799422264099, "learning_rate": 6.208343779255628e-05, "loss": 0.0405, "step": 45120 }, { "grad_norm": 0.6391850113868713, "learning_rate": 6.206739254215449e-05, "loss": 0.0539, "step": 45130 }, { "grad_norm": 0.6347276568412781, "learning_rate": 6.205134597208077e-05, "loss": 0.0597, "step": 45140 }, { "grad_norm": 0.48067519068717957, "learning_rate": 6.203529808409e-05, "loss": 0.0508, "step": 45150 }, { "grad_norm": 0.45242840051651, "learning_rate": 6.201924887993712e-05, "loss": 0.0489, "step": 45160 }, { "grad_norm": 0.541465163230896, "learning_rate": 6.200319836137725e-05, "loss": 0.0503, "step": 45170 }, { "grad_norm": 0.34840524196624756, "learning_rate": 6.198714653016565e-05, "loss": 0.0453, "step": 45180 }, { "grad_norm": 0.45836785435676575, "learning_rate": 6.197109338805774e-05, "loss": 0.0474, "step": 45190 }, { "grad_norm": 0.5643194913864136, "learning_rate": 6.195503893680903e-05, "loss": 0.048, "step": 45200 }, { "grad_norm": 0.4009495973587036, "learning_rate": 6.193898317817524e-05, "loss": 0.0414, "step": 45210 }, { "grad_norm": 0.8265069723129272, "learning_rate": 6.192292611391221e-05, "loss": 0.0372, "step": 45220 }, { "grad_norm": 0.2758135199546814, "learning_rate": 6.190686774577591e-05, "loss": 0.0377, "step": 45230 }, { "grad_norm": 0.2808265686035156, "learning_rate": 6.189080807552245e-05, "loss": 0.0489, "step": 45240 }, { "grad_norm": 0.8085670471191406, "learning_rate": 6.187474710490809e-05, "loss": 0.0456, "step": 45250 }, { "grad_norm": 0.6255159974098206, "learning_rate": 6.185868483568926e-05, "loss": 0.0453, "step": 45260 }, { "grad_norm": 0.4606318771839142, "learning_rate": 6.184262126962245e-05, "loss": 0.0452, "step": 45270 }, { "grad_norm": 0.550923228263855, "learning_rate": 6.182655640846442e-05, "loss": 0.0332, "step": 45280 }, { "grad_norm": 0.4225478768348694, "learning_rate": 6.181049025397196e-05, "loss": 0.0474, "step": 45290 }, { "grad_norm": 0.3387550115585327, "learning_rate": 6.179442280790202e-05, "loss": 0.0487, "step": 45300 }, { "grad_norm": 0.5289576649665833, "learning_rate": 6.177835407201174e-05, "loss": 0.0616, "step": 45310 }, { "grad_norm": 0.7425777316093445, "learning_rate": 6.176228404805839e-05, "loss": 0.0462, "step": 45320 }, { "grad_norm": 0.5533152222633362, "learning_rate": 6.174621273779932e-05, "loss": 0.0482, "step": 45330 }, { "grad_norm": 0.4945146143436432, "learning_rate": 6.17301401429921e-05, "loss": 0.0459, "step": 45340 }, { "grad_norm": 0.5106241106987, "learning_rate": 6.171406626539436e-05, "loss": 0.0439, "step": 45350 }, { "grad_norm": 0.7493733167648315, "learning_rate": 6.169799110676398e-05, "loss": 0.0657, "step": 45360 }, { "grad_norm": 0.4674302935600281, "learning_rate": 6.168191466885885e-05, "loss": 0.0418, "step": 45370 }, { "grad_norm": 0.4123746454715729, "learning_rate": 6.166583695343711e-05, "loss": 0.0449, "step": 45380 }, { "grad_norm": 0.3898713290691376, "learning_rate": 6.164975796225698e-05, "loss": 0.05, "step": 45390 }, { "grad_norm": 1.5036598443984985, "learning_rate": 6.163367769707683e-05, "loss": 0.0373, "step": 45400 }, { "grad_norm": 0.569703996181488, "learning_rate": 6.161759615965519e-05, "loss": 0.0511, "step": 45410 }, { "grad_norm": 0.3989284634590149, "learning_rate": 6.16015133517507e-05, "loss": 0.0383, "step": 45420 }, { "grad_norm": 0.634453535079956, "learning_rate": 6.158542927512214e-05, "loss": 0.0383, "step": 45430 }, { "grad_norm": 0.4379845857620239, "learning_rate": 6.156934393152846e-05, "loss": 0.0417, "step": 45440 }, { "grad_norm": 0.3675931394100189, "learning_rate": 6.15532573227287e-05, "loss": 0.0555, "step": 45450 }, { "grad_norm": 0.6430561542510986, "learning_rate": 6.153716945048212e-05, "loss": 0.0591, "step": 45460 }, { "grad_norm": 0.8023918271064758, "learning_rate": 6.152108031654802e-05, "loss": 0.0582, "step": 45470 }, { "grad_norm": 0.8582981824874878, "learning_rate": 6.15049899226859e-05, "loss": 0.0473, "step": 45480 }, { "grad_norm": 0.39302676916122437, "learning_rate": 6.148889827065537e-05, "loss": 0.0495, "step": 45490 }, { "grad_norm": 0.42895981669425964, "learning_rate": 6.147280536221622e-05, "loss": 0.0462, "step": 45500 }, { "grad_norm": 0.7643282413482666, "learning_rate": 6.145671119912832e-05, "loss": 0.0442, "step": 45510 }, { "grad_norm": 0.4884565770626068, "learning_rate": 6.144061578315169e-05, "loss": 0.0368, "step": 45520 }, { "grad_norm": 0.38519930839538574, "learning_rate": 6.142451911604654e-05, "loss": 0.0458, "step": 45530 }, { "grad_norm": 0.41384217143058777, "learning_rate": 6.140842119957315e-05, "loss": 0.0554, "step": 45540 }, { "grad_norm": 0.4678124487400055, "learning_rate": 6.139232203549197e-05, "loss": 0.0403, "step": 45550 }, { "grad_norm": 0.5469533205032349, "learning_rate": 6.137622162556357e-05, "loss": 0.0491, "step": 45560 }, { "grad_norm": 0.4430740177631378, "learning_rate": 6.13601199715487e-05, "loss": 0.0418, "step": 45570 }, { "grad_norm": 0.48296380043029785, "learning_rate": 6.134401707520816e-05, "loss": 0.049, "step": 45580 }, { "grad_norm": 0.30508649349212646, "learning_rate": 6.132791293830298e-05, "loss": 0.0388, "step": 45590 }, { "grad_norm": 0.4815385639667511, "learning_rate": 6.131180756259428e-05, "loss": 0.0451, "step": 45600 }, { "grad_norm": 0.5402326583862305, "learning_rate": 6.129570094984331e-05, "loss": 0.0317, "step": 45610 }, { "grad_norm": 0.6574417352676392, "learning_rate": 6.127959310181145e-05, "loss": 0.0513, "step": 45620 }, { "grad_norm": 0.5331013798713684, "learning_rate": 6.126348402026026e-05, "loss": 0.05, "step": 45630 }, { "grad_norm": 0.473596453666687, "learning_rate": 6.12473737069514e-05, "loss": 0.036, "step": 45640 }, { "grad_norm": 0.6890078186988831, "learning_rate": 6.123126216364665e-05, "loss": 0.0515, "step": 45650 }, { "grad_norm": 0.6509345769882202, "learning_rate": 6.121514939210797e-05, "loss": 0.0389, "step": 45660 }, { "grad_norm": 0.7017287015914917, "learning_rate": 6.119903539409741e-05, "loss": 0.0524, "step": 45670 }, { "grad_norm": 0.3471664786338806, "learning_rate": 6.118292017137716e-05, "loss": 0.0429, "step": 45680 }, { "grad_norm": 0.8322047591209412, "learning_rate": 6.116680372570959e-05, "loss": 0.0526, "step": 45690 }, { "grad_norm": 0.38778701424598694, "learning_rate": 6.115068605885713e-05, "loss": 0.0329, "step": 45700 }, { "grad_norm": 0.363751620054245, "learning_rate": 6.113456717258243e-05, "loss": 0.0517, "step": 45710 }, { "grad_norm": 0.34100744128227234, "learning_rate": 6.11184470686482e-05, "loss": 0.0434, "step": 45720 }, { "grad_norm": 0.5338833332061768, "learning_rate": 6.110232574881731e-05, "loss": 0.0429, "step": 45730 }, { "grad_norm": 1.17531418800354, "learning_rate": 6.108620321485277e-05, "loss": 0.0471, "step": 45740 }, { "grad_norm": 0.6104838252067566, "learning_rate": 6.107007946851773e-05, "loss": 0.044, "step": 45750 }, { "grad_norm": 0.25776708126068115, "learning_rate": 6.105395451157542e-05, "loss": 0.0475, "step": 45760 }, { "grad_norm": 0.4156385064125061, "learning_rate": 6.103782834578928e-05, "loss": 0.0408, "step": 45770 }, { "grad_norm": 0.32689550518989563, "learning_rate": 6.102170097292281e-05, "loss": 0.0432, "step": 45780 }, { "grad_norm": 0.49365484714508057, "learning_rate": 6.1005572394739716e-05, "loss": 0.0517, "step": 45790 }, { "grad_norm": 0.5934664607048035, "learning_rate": 6.0989442613003765e-05, "loss": 0.0321, "step": 45800 }, { "grad_norm": 1.0204546451568604, "learning_rate": 6.097331162947888e-05, "loss": 0.0478, "step": 45810 }, { "grad_norm": 0.508949339389801, "learning_rate": 6.095717944592914e-05, "loss": 0.0504, "step": 45820 }, { "grad_norm": 0.7636938095092773, "learning_rate": 6.094104606411873e-05, "loss": 0.057, "step": 45830 }, { "grad_norm": 0.527838945388794, "learning_rate": 6.0924911485811966e-05, "loss": 0.0515, "step": 45840 }, { "grad_norm": 0.7294138073921204, "learning_rate": 6.090877571277331e-05, "loss": 0.054, "step": 45850 }, { "grad_norm": 0.35612422227859497, "learning_rate": 6.089263874676736e-05, "loss": 0.0449, "step": 45860 }, { "grad_norm": 0.8612220883369446, "learning_rate": 6.0876500589558796e-05, "loss": 0.0431, "step": 45870 }, { "grad_norm": 1.425920009613037, "learning_rate": 6.086036124291248e-05, "loss": 0.0455, "step": 45880 }, { "grad_norm": 0.5095338821411133, "learning_rate": 6.084422070859339e-05, "loss": 0.0471, "step": 45890 }, { "grad_norm": 0.6015581488609314, "learning_rate": 6.082807898836663e-05, "loss": 0.0458, "step": 45900 }, { "grad_norm": 0.3974273204803467, "learning_rate": 6.081193608399742e-05, "loss": 0.043, "step": 45910 }, { "grad_norm": 0.4422559142112732, "learning_rate": 6.0795791997251164e-05, "loss": 0.0473, "step": 45920 }, { "grad_norm": 0.49121761322021484, "learning_rate": 6.0779646729893294e-05, "loss": 0.0391, "step": 45930 }, { "grad_norm": 0.3903568685054779, "learning_rate": 6.0763500283689476e-05, "loss": 0.0317, "step": 45940 }, { "grad_norm": 0.4161006212234497, "learning_rate": 6.0747352660405455e-05, "loss": 0.0435, "step": 45950 }, { "grad_norm": 0.44030362367630005, "learning_rate": 6.073120386180709e-05, "loss": 0.0435, "step": 45960 }, { "grad_norm": 0.6790411472320557, "learning_rate": 6.0715053889660425e-05, "loss": 0.0415, "step": 45970 }, { "grad_norm": 0.43015435338020325, "learning_rate": 6.069890274573157e-05, "loss": 0.0435, "step": 45980 }, { "grad_norm": 0.427778035402298, "learning_rate": 6.068275043178679e-05, "loss": 0.0585, "step": 45990 }, { "grad_norm": 0.6704531311988831, "learning_rate": 6.066659694959248e-05, "loss": 0.0412, "step": 46000 }, { "grad_norm": 0.33696791529655457, "learning_rate": 6.0650442300915176e-05, "loss": 0.0446, "step": 46010 }, { "grad_norm": 0.40279531478881836, "learning_rate": 6.063428648752152e-05, "loss": 0.0479, "step": 46020 }, { "grad_norm": 0.5341557860374451, "learning_rate": 6.0618129511178266e-05, "loss": 0.0429, "step": 46030 }, { "grad_norm": 0.2971374988555908, "learning_rate": 6.060197137365234e-05, "loss": 0.0527, "step": 46040 }, { "grad_norm": 0.4765755236148834, "learning_rate": 6.058581207671077e-05, "loss": 0.0397, "step": 46050 }, { "grad_norm": 0.3328109681606293, "learning_rate": 6.056965162212072e-05, "loss": 0.0479, "step": 46060 }, { "grad_norm": 0.3916610777378082, "learning_rate": 6.055349001164943e-05, "loss": 0.0415, "step": 46070 }, { "grad_norm": 0.39432886242866516, "learning_rate": 6.0537327247064347e-05, "loss": 0.0479, "step": 46080 }, { "grad_norm": 0.7707289457321167, "learning_rate": 6.0521163330133014e-05, "loss": 0.0497, "step": 46090 }, { "grad_norm": 0.5472941994667053, "learning_rate": 6.050499826262306e-05, "loss": 0.0416, "step": 46100 }, { "grad_norm": 0.5170687437057495, "learning_rate": 6.0488832046302294e-05, "loss": 0.0611, "step": 46110 }, { "grad_norm": 0.25391724705696106, "learning_rate": 6.0472664682938626e-05, "loss": 0.0487, "step": 46120 }, { "grad_norm": 0.3088100254535675, "learning_rate": 6.045649617430009e-05, "loss": 0.0486, "step": 46130 }, { "grad_norm": 0.3899429738521576, "learning_rate": 6.0440326522154866e-05, "loss": 0.0413, "step": 46140 }, { "grad_norm": 0.4978284537792206, "learning_rate": 6.0424155728271224e-05, "loss": 0.0476, "step": 46150 }, { "grad_norm": 0.29448193311691284, "learning_rate": 6.040798379441758e-05, "loss": 0.0484, "step": 46160 }, { "grad_norm": 0.288124144077301, "learning_rate": 6.0391810722362485e-05, "loss": 0.0386, "step": 46170 }, { "grad_norm": 0.38257864117622375, "learning_rate": 6.037563651387458e-05, "loss": 0.0375, "step": 46180 }, { "grad_norm": 0.5616655349731445, "learning_rate": 6.0359461170722666e-05, "loss": 0.0427, "step": 46190 }, { "grad_norm": 0.3975408673286438, "learning_rate": 6.034328469467566e-05, "loss": 0.0411, "step": 46200 }, { "grad_norm": 0.44657063484191895, "learning_rate": 6.0327107087502596e-05, "loss": 0.0413, "step": 46210 }, { "grad_norm": 0.47621437907218933, "learning_rate": 6.031092835097262e-05, "loss": 0.0484, "step": 46220 }, { "grad_norm": 0.31683820486068726, "learning_rate": 6.0294748486855024e-05, "loss": 0.0408, "step": 46230 }, { "grad_norm": 0.704038679599762, "learning_rate": 6.0278567496919216e-05, "loss": 0.057, "step": 46240 }, { "grad_norm": 0.4559658467769623, "learning_rate": 6.026238538293472e-05, "loss": 0.048, "step": 46250 }, { "grad_norm": 0.4026772975921631, "learning_rate": 6.024620214667118e-05, "loss": 0.0444, "step": 46260 }, { "grad_norm": 0.40657052397727966, "learning_rate": 6.0230017789898384e-05, "loss": 0.0397, "step": 46270 }, { "grad_norm": 0.4039594233036041, "learning_rate": 6.021383231438622e-05, "loss": 0.0502, "step": 46280 }, { "grad_norm": 0.4515346884727478, "learning_rate": 6.0197645721904704e-05, "loss": 0.042, "step": 46290 }, { "grad_norm": 0.44575369358062744, "learning_rate": 6.0181458014224e-05, "loss": 0.051, "step": 46300 }, { "grad_norm": 0.6980546116828918, "learning_rate": 6.016526919311435e-05, "loss": 0.0454, "step": 46310 }, { "grad_norm": 0.37750619649887085, "learning_rate": 6.014907926034613e-05, "loss": 0.0528, "step": 46320 }, { "grad_norm": 0.6299592852592468, "learning_rate": 6.0132888217689875e-05, "loss": 0.0426, "step": 46330 }, { "grad_norm": 0.6590644121170044, "learning_rate": 6.01166960669162e-05, "loss": 0.052, "step": 46340 }, { "grad_norm": 0.48970144987106323, "learning_rate": 6.0100502809795845e-05, "loss": 0.0459, "step": 46350 }, { "grad_norm": 0.3222176730632782, "learning_rate": 6.00843084480997e-05, "loss": 0.0431, "step": 46360 }, { "grad_norm": 0.3064524233341217, "learning_rate": 6.0068112983598736e-05, "loss": 0.0388, "step": 46370 }, { "grad_norm": 0.383060097694397, "learning_rate": 6.005191641806407e-05, "loss": 0.0407, "step": 46380 }, { "grad_norm": 0.5218490958213806, "learning_rate": 6.003571875326694e-05, "loss": 0.0428, "step": 46390 }, { "grad_norm": 0.4287131726741791, "learning_rate": 6.00195199909787e-05, "loss": 0.0427, "step": 46400 }, { "grad_norm": 0.5282581448554993, "learning_rate": 6.000332013297082e-05, "loss": 0.0472, "step": 46410 }, { "grad_norm": 0.3326645493507385, "learning_rate": 5.998711918101487e-05, "loss": 0.0453, "step": 46420 }, { "grad_norm": 0.42736929655075073, "learning_rate": 5.997091713688261e-05, "loss": 0.0512, "step": 46430 }, { "grad_norm": 0.3661309480667114, "learning_rate": 5.9954714002345836e-05, "loss": 0.04, "step": 46440 }, { "grad_norm": 0.7098157405853271, "learning_rate": 5.993850977917649e-05, "loss": 0.0483, "step": 46450 }, { "grad_norm": 0.4380768835544586, "learning_rate": 5.992230446914667e-05, "loss": 0.0413, "step": 46460 }, { "grad_norm": 0.37986433506011963, "learning_rate": 5.990609807402855e-05, "loss": 0.0552, "step": 46470 }, { "grad_norm": 0.6397569179534912, "learning_rate": 5.988989059559443e-05, "loss": 0.0509, "step": 46480 }, { "grad_norm": 0.46843671798706055, "learning_rate": 5.987368203561675e-05, "loss": 0.0535, "step": 46490 }, { "grad_norm": 0.7386255860328674, "learning_rate": 5.9857472395868055e-05, "loss": 0.0474, "step": 46500 }, { "grad_norm": 0.6825923919677734, "learning_rate": 5.9841261678120983e-05, "loss": 0.0493, "step": 46510 }, { "grad_norm": 0.29971569776535034, "learning_rate": 5.982504988414834e-05, "loss": 0.0362, "step": 46520 }, { "grad_norm": 0.5070285201072693, "learning_rate": 5.9808837015723015e-05, "loss": 0.0568, "step": 46530 }, { "grad_norm": 0.3749014735221863, "learning_rate": 5.9792623074618016e-05, "loss": 0.039, "step": 46540 }, { "grad_norm": 0.4116363227367401, "learning_rate": 5.977640806260648e-05, "loss": 0.0441, "step": 46550 }, { "grad_norm": 0.6201049089431763, "learning_rate": 5.976019198146166e-05, "loss": 0.0501, "step": 46560 }, { "grad_norm": 1.8178094625473022, "learning_rate": 5.974397483295692e-05, "loss": 0.0433, "step": 46570 }, { "grad_norm": 0.4342486560344696, "learning_rate": 5.972775661886575e-05, "loss": 0.0455, "step": 46580 }, { "grad_norm": 0.6348972916603088, "learning_rate": 5.971153734096173e-05, "loss": 0.0465, "step": 46590 }, { "grad_norm": 0.4317099452018738, "learning_rate": 5.969531700101859e-05, "loss": 0.0406, "step": 46600 }, { "grad_norm": 1.0667163133621216, "learning_rate": 5.9679095600810155e-05, "loss": 0.0461, "step": 46610 }, { "grad_norm": 0.5719022154808044, "learning_rate": 5.9662873142110384e-05, "loss": 0.0388, "step": 46620 }, { "grad_norm": 0.5970348715782166, "learning_rate": 5.964664962669333e-05, "loss": 0.048, "step": 46630 }, { "grad_norm": 0.40333524346351624, "learning_rate": 5.9630425056333186e-05, "loss": 0.046, "step": 46640 }, { "grad_norm": 0.2685006558895111, "learning_rate": 5.961419943280422e-05, "loss": 0.0475, "step": 46650 }, { "grad_norm": 0.4026000201702118, "learning_rate": 5.959797275788087e-05, "loss": 0.0378, "step": 46660 }, { "grad_norm": 0.4376264810562134, "learning_rate": 5.958174503333765e-05, "loss": 0.0544, "step": 46670 }, { "grad_norm": 0.4486438035964966, "learning_rate": 5.9565516260949195e-05, "loss": 0.0582, "step": 46680 }, { "grad_norm": 0.8763489127159119, "learning_rate": 5.9549286442490273e-05, "loss": 0.0534, "step": 46690 }, { "grad_norm": 0.5374938249588013, "learning_rate": 5.953305557973572e-05, "loss": 0.0551, "step": 46700 }, { "grad_norm": 0.7013370990753174, "learning_rate": 5.9516823674460564e-05, "loss": 0.0477, "step": 46710 }, { "grad_norm": 0.6488234400749207, "learning_rate": 5.9500590728439875e-05, "loss": 0.0423, "step": 46720 }, { "grad_norm": 0.36801066994667053, "learning_rate": 5.948435674344886e-05, "loss": 0.0468, "step": 46730 }, { "grad_norm": 1.7648756504058838, "learning_rate": 5.946812172126285e-05, "loss": 0.0429, "step": 46740 }, { "grad_norm": 0.8404155969619751, "learning_rate": 5.945188566365729e-05, "loss": 0.0343, "step": 46750 }, { "grad_norm": 1.304136872291565, "learning_rate": 5.943564857240773e-05, "loss": 0.0476, "step": 46760 }, { "grad_norm": 0.6348478198051453, "learning_rate": 5.941941044928981e-05, "loss": 0.0406, "step": 46770 }, { "grad_norm": 0.8793948888778687, "learning_rate": 5.940317129607935e-05, "loss": 0.0457, "step": 46780 }, { "grad_norm": 0.4706290662288666, "learning_rate": 5.9386931114552204e-05, "loss": 0.0548, "step": 46790 }, { "grad_norm": 0.3976486623287201, "learning_rate": 5.937068990648438e-05, "loss": 0.0463, "step": 46800 }, { "grad_norm": 0.6780657172203064, "learning_rate": 5.935444767365199e-05, "loss": 0.0671, "step": 46810 }, { "grad_norm": 0.5461853742599487, "learning_rate": 5.933820441783129e-05, "loss": 0.0551, "step": 46820 }, { "grad_norm": 0.4047617018222809, "learning_rate": 5.932196014079857e-05, "loss": 0.0482, "step": 46830 }, { "grad_norm": 0.32208582758903503, "learning_rate": 5.930571484433032e-05, "loss": 0.0353, "step": 46840 }, { "grad_norm": 0.340403288602829, "learning_rate": 5.928946853020309e-05, "loss": 0.0386, "step": 46850 }, { "grad_norm": 0.29578834772109985, "learning_rate": 5.927322120019352e-05, "loss": 0.0353, "step": 46860 }, { "grad_norm": 0.8078397512435913, "learning_rate": 5.925697285607844e-05, "loss": 0.0557, "step": 46870 }, { "grad_norm": 0.5895037651062012, "learning_rate": 5.924072349963472e-05, "loss": 0.0435, "step": 46880 }, { "grad_norm": 0.3593572974205017, "learning_rate": 5.922447313263938e-05, "loss": 0.037, "step": 46890 }, { "grad_norm": 0.37394076585769653, "learning_rate": 5.920822175686952e-05, "loss": 0.0462, "step": 46900 }, { "grad_norm": 0.5219018459320068, "learning_rate": 5.919196937410239e-05, "loss": 0.0538, "step": 46910 }, { "grad_norm": 0.4786173105239868, "learning_rate": 5.917571598611529e-05, "loss": 0.0428, "step": 46920 }, { "grad_norm": 0.6782716512680054, "learning_rate": 5.9159461594685686e-05, "loss": 0.0479, "step": 46930 }, { "grad_norm": 0.4322122037410736, "learning_rate": 5.914320620159114e-05, "loss": 0.0439, "step": 46940 }, { "grad_norm": 0.6865296363830566, "learning_rate": 5.912694980860931e-05, "loss": 0.0507, "step": 46950 }, { "grad_norm": 0.45332589745521545, "learning_rate": 5.9110692417517964e-05, "loss": 0.0358, "step": 46960 }, { "grad_norm": 0.4451228082180023, "learning_rate": 5.9094434030095e-05, "loss": 0.0547, "step": 46970 }, { "grad_norm": 0.3452111780643463, "learning_rate": 5.907817464811841e-05, "loss": 0.0419, "step": 46980 }, { "grad_norm": 0.3444940149784088, "learning_rate": 5.906191427336627e-05, "loss": 0.0489, "step": 46990 }, { "grad_norm": 0.856482982635498, "learning_rate": 5.9045652907616835e-05, "loss": 0.0505, "step": 47000 }, { "grad_norm": 0.5326331257820129, "learning_rate": 5.902939055264838e-05, "loss": 0.0487, "step": 47010 }, { "grad_norm": 0.41257163882255554, "learning_rate": 5.901312721023935e-05, "loss": 0.0411, "step": 47020 }, { "grad_norm": 0.3399612307548523, "learning_rate": 5.8996862882168294e-05, "loss": 0.0404, "step": 47030 }, { "grad_norm": 0.47882163524627686, "learning_rate": 5.8980597570213826e-05, "loss": 0.0474, "step": 47040 }, { "grad_norm": 0.7844752073287964, "learning_rate": 5.896433127615471e-05, "loss": 0.0623, "step": 47050 }, { "grad_norm": 0.5378547310829163, "learning_rate": 5.894806400176981e-05, "loss": 0.0372, "step": 47060 }, { "grad_norm": 0.8038886189460754, "learning_rate": 5.893179574883808e-05, "loss": 0.0435, "step": 47070 }, { "grad_norm": 0.48912185430526733, "learning_rate": 5.8915526519138585e-05, "loss": 0.047, "step": 47080 }, { "grad_norm": 1.0518652200698853, "learning_rate": 5.889925631445053e-05, "loss": 0.0545, "step": 47090 }, { "grad_norm": 0.5351830720901489, "learning_rate": 5.888298513655318e-05, "loss": 0.0372, "step": 47100 }, { "grad_norm": 0.440170556306839, "learning_rate": 5.886671298722592e-05, "loss": 0.0468, "step": 47110 }, { "grad_norm": 0.2953525185585022, "learning_rate": 5.885043986824828e-05, "loss": 0.0395, "step": 47120 }, { "grad_norm": 0.8816118836402893, "learning_rate": 5.883416578139982e-05, "loss": 0.0392, "step": 47130 }, { "grad_norm": 0.42044714093208313, "learning_rate": 5.881789072846029e-05, "loss": 0.0448, "step": 47140 }, { "grad_norm": 0.45868921279907227, "learning_rate": 5.8801614711209474e-05, "loss": 0.0342, "step": 47150 }, { "grad_norm": 0.4480786919593811, "learning_rate": 5.878533773142732e-05, "loss": 0.0436, "step": 47160 }, { "grad_norm": 0.6075236201286316, "learning_rate": 5.8769059790893834e-05, "loss": 0.046, "step": 47170 }, { "grad_norm": 0.4612235724925995, "learning_rate": 5.8752780891389145e-05, "loss": 0.0451, "step": 47180 }, { "grad_norm": 0.34475862979888916, "learning_rate": 5.873650103469351e-05, "loss": 0.0435, "step": 47190 }, { "grad_norm": 0.4539240598678589, "learning_rate": 5.872022022258725e-05, "loss": 0.0335, "step": 47200 }, { "grad_norm": 0.5951141715049744, "learning_rate": 5.8703938456850814e-05, "loss": 0.0391, "step": 47210 }, { "grad_norm": 0.46456602215766907, "learning_rate": 5.868765573926477e-05, "loss": 0.0505, "step": 47220 }, { "grad_norm": 0.4224106967449188, "learning_rate": 5.867137207160974e-05, "loss": 0.0535, "step": 47230 }, { "grad_norm": 0.5405718088150024, "learning_rate": 5.86550874556665e-05, "loss": 0.0487, "step": 47240 }, { "grad_norm": 0.4297504127025604, "learning_rate": 5.863880189321592e-05, "loss": 0.0384, "step": 47250 }, { "grad_norm": 0.6570985913276672, "learning_rate": 5.862251538603894e-05, "loss": 0.0376, "step": 47260 }, { "grad_norm": 0.3926331102848053, "learning_rate": 5.8606227935916656e-05, "loss": 0.0337, "step": 47270 }, { "grad_norm": 0.4373575448989868, "learning_rate": 5.858993954463021e-05, "loss": 0.0435, "step": 47280 }, { "grad_norm": 0.5329977869987488, "learning_rate": 5.8573650213960896e-05, "loss": 0.0494, "step": 47290 }, { "grad_norm": 0.3712256848812103, "learning_rate": 5.8557359945690084e-05, "loss": 0.0451, "step": 47300 }, { "grad_norm": 0.354236364364624, "learning_rate": 5.8541068741599246e-05, "loss": 0.0427, "step": 47310 }, { "grad_norm": 0.42435240745544434, "learning_rate": 5.8524776603469985e-05, "loss": 0.0364, "step": 47320 }, { "grad_norm": 0.484692245721817, "learning_rate": 5.850848353308397e-05, "loss": 0.0416, "step": 47330 }, { "grad_norm": 0.7778406739234924, "learning_rate": 5.849218953222297e-05, "loss": 0.0333, "step": 47340 }, { "grad_norm": 0.415128231048584, "learning_rate": 5.847589460266891e-05, "loss": 0.0487, "step": 47350 }, { "grad_norm": 0.48514896631240845, "learning_rate": 5.845959874620375e-05, "loss": 0.046, "step": 47360 }, { "grad_norm": 0.37667468190193176, "learning_rate": 5.84433019646096e-05, "loss": 0.0469, "step": 47370 }, { "grad_norm": 0.4294966161251068, "learning_rate": 5.842700425966863e-05, "loss": 0.0441, "step": 47380 }, { "grad_norm": 0.42790138721466064, "learning_rate": 5.841070563316315e-05, "loss": 0.0445, "step": 47390 }, { "grad_norm": 0.43570321798324585, "learning_rate": 5.839440608687554e-05, "loss": 0.0488, "step": 47400 }, { "grad_norm": 0.43023547530174255, "learning_rate": 5.837810562258831e-05, "loss": 0.0361, "step": 47410 }, { "grad_norm": 0.480292946100235, "learning_rate": 5.836180424208405e-05, "loss": 0.0653, "step": 47420 }, { "grad_norm": 0.38623785972595215, "learning_rate": 5.834550194714543e-05, "loss": 0.0437, "step": 47430 }, { "grad_norm": 0.5309616923332214, "learning_rate": 5.832919873955528e-05, "loss": 0.0631, "step": 47440 }, { "grad_norm": 0.5214033722877502, "learning_rate": 5.831289462109648e-05, "loss": 0.0439, "step": 47450 }, { "grad_norm": 0.3257804811000824, "learning_rate": 5.829658959355202e-05, "loss": 0.0459, "step": 47460 }, { "grad_norm": 0.5359699726104736, "learning_rate": 5.8280283658704995e-05, "loss": 0.0511, "step": 47470 }, { "grad_norm": 0.5205581784248352, "learning_rate": 5.826397681833859e-05, "loss": 0.0429, "step": 47480 }, { "grad_norm": 0.5245789885520935, "learning_rate": 5.824766907423612e-05, "loss": 0.0395, "step": 47490 }, { "grad_norm": 0.44048404693603516, "learning_rate": 5.8231360428180945e-05, "loss": 0.0465, "step": 47500 }, { "grad_norm": 0.3780224323272705, "learning_rate": 5.821505088195658e-05, "loss": 0.0415, "step": 47510 }, { "grad_norm": 0.5578895211219788, "learning_rate": 5.819874043734661e-05, "loss": 0.0502, "step": 47520 }, { "grad_norm": 0.47435227036476135, "learning_rate": 5.8182429096134695e-05, "loss": 0.0434, "step": 47530 }, { "grad_norm": 0.8399900794029236, "learning_rate": 5.816611686010465e-05, "loss": 0.0455, "step": 47540 }, { "grad_norm": 0.4534755349159241, "learning_rate": 5.814980373104033e-05, "loss": 0.0458, "step": 47550 }, { "grad_norm": 0.5740112662315369, "learning_rate": 5.813348971072572e-05, "loss": 0.0575, "step": 47560 }, { "grad_norm": 0.41583922505378723, "learning_rate": 5.811717480094492e-05, "loss": 0.0419, "step": 47570 }, { "grad_norm": 0.5216782689094543, "learning_rate": 5.810085900348209e-05, "loss": 0.0536, "step": 47580 }, { "grad_norm": 0.47793281078338623, "learning_rate": 5.8084542320121483e-05, "loss": 0.0429, "step": 47590 }, { "grad_norm": 0.44496965408325195, "learning_rate": 5.8068224752647497e-05, "loss": 0.0523, "step": 47600 }, { "grad_norm": 0.396843284368515, "learning_rate": 5.805190630284457e-05, "loss": 0.0299, "step": 47610 }, { "grad_norm": 0.5737031698226929, "learning_rate": 5.803558697249729e-05, "loss": 0.0463, "step": 47620 }, { "grad_norm": 0.4955446720123291, "learning_rate": 5.8019266763390276e-05, "loss": 0.0497, "step": 47630 }, { "grad_norm": 0.3305622935295105, "learning_rate": 5.800294567730833e-05, "loss": 0.0393, "step": 47640 }, { "grad_norm": 0.40150687098503113, "learning_rate": 5.798662371603628e-05, "loss": 0.0448, "step": 47650 }, { "grad_norm": 0.567929208278656, "learning_rate": 5.797030088135904e-05, "loss": 0.0484, "step": 47660 }, { "grad_norm": 0.39679810404777527, "learning_rate": 5.7953977175061705e-05, "loss": 0.0433, "step": 47670 }, { "grad_norm": 0.6723015308380127, "learning_rate": 5.793765259892938e-05, "loss": 0.0585, "step": 47680 }, { "grad_norm": 0.6882417798042297, "learning_rate": 5.792132715474729e-05, "loss": 0.0411, "step": 47690 }, { "grad_norm": 0.5568235516548157, "learning_rate": 5.790500084430078e-05, "loss": 0.0351, "step": 47700 }, { "grad_norm": 0.329571932554245, "learning_rate": 5.7888673669375274e-05, "loss": 0.038, "step": 47710 }, { "grad_norm": 0.4076901972293854, "learning_rate": 5.787234563175625e-05, "loss": 0.0404, "step": 47720 }, { "grad_norm": 0.5941630601882935, "learning_rate": 5.7856016733229355e-05, "loss": 0.0446, "step": 47730 }, { "grad_norm": 0.4004770815372467, "learning_rate": 5.7839686975580297e-05, "loss": 0.0482, "step": 47740 }, { "grad_norm": 0.33234140276908875, "learning_rate": 5.782335636059484e-05, "loss": 0.041, "step": 47750 }, { "grad_norm": 0.5331122875213623, "learning_rate": 5.780702489005889e-05, "loss": 0.0428, "step": 47760 }, { "grad_norm": 0.3826586604118347, "learning_rate": 5.779069256575845e-05, "loss": 0.048, "step": 47770 }, { "grad_norm": 1.2625528573989868, "learning_rate": 5.7774359389479574e-05, "loss": 0.0346, "step": 47780 }, { "grad_norm": 0.5347055792808533, "learning_rate": 5.775802536300845e-05, "loss": 0.0541, "step": 47790 }, { "grad_norm": 0.5060232281684875, "learning_rate": 5.774169048813134e-05, "loss": 0.0406, "step": 47800 }, { "grad_norm": 0.4107586741447449, "learning_rate": 5.77253547666346e-05, "loss": 0.0365, "step": 47810 }, { "grad_norm": 0.6640077233314514, "learning_rate": 5.770901820030465e-05, "loss": 0.0452, "step": 47820 }, { "grad_norm": 0.426227867603302, "learning_rate": 5.769268079092809e-05, "loss": 0.0437, "step": 47830 }, { "grad_norm": 0.2853057086467743, "learning_rate": 5.767634254029151e-05, "loss": 0.0422, "step": 47840 }, { "grad_norm": 0.5874415636062622, "learning_rate": 5.7660003450181655e-05, "loss": 0.0704, "step": 47850 }, { "grad_norm": 0.5853018760681152, "learning_rate": 5.764366352238534e-05, "loss": 0.0457, "step": 47860 }, { "grad_norm": 0.9346079230308533, "learning_rate": 5.7627322758689474e-05, "loss": 0.0365, "step": 47870 }, { "grad_norm": 0.6395016312599182, "learning_rate": 5.761098116088105e-05, "loss": 0.0426, "step": 47880 }, { "grad_norm": 0.4218047261238098, "learning_rate": 5.759463873074717e-05, "loss": 0.0402, "step": 47890 }, { "grad_norm": 0.5547118186950684, "learning_rate": 5.757829547007504e-05, "loss": 0.0396, "step": 47900 }, { "grad_norm": 0.6220501065254211, "learning_rate": 5.756195138065189e-05, "loss": 0.0421, "step": 47910 }, { "grad_norm": 0.4035327732563019, "learning_rate": 5.754560646426511e-05, "loss": 0.0478, "step": 47920 }, { "grad_norm": 0.37004056572914124, "learning_rate": 5.752926072270216e-05, "loss": 0.0356, "step": 47930 }, { "grad_norm": 0.48076844215393066, "learning_rate": 5.7512914157750563e-05, "loss": 0.046, "step": 47940 }, { "grad_norm": 0.3383843004703522, "learning_rate": 5.749656677119798e-05, "loss": 0.0267, "step": 47950 }, { "grad_norm": 0.468753457069397, "learning_rate": 5.7480218564832125e-05, "loss": 0.0439, "step": 47960 }, { "grad_norm": 0.390465646982193, "learning_rate": 5.746386954044082e-05, "loss": 0.0327, "step": 47970 }, { "grad_norm": 0.4985044002532959, "learning_rate": 5.744751969981195e-05, "loss": 0.0356, "step": 47980 }, { "grad_norm": 0.48157671093940735, "learning_rate": 5.7431169044733526e-05, "loss": 0.0505, "step": 47990 }, { "grad_norm": 0.5120133757591248, "learning_rate": 5.741481757699364e-05, "loss": 0.0352, "step": 48000 }, { "grad_norm": 0.40800338983535767, "learning_rate": 5.7398465298380434e-05, "loss": 0.0461, "step": 48010 }, { "grad_norm": 0.5345004200935364, "learning_rate": 5.7382112210682193e-05, "loss": 0.0603, "step": 48020 }, { "grad_norm": 0.6805002689361572, "learning_rate": 5.7365758315687266e-05, "loss": 0.0532, "step": 48030 }, { "grad_norm": 0.585664689540863, "learning_rate": 5.734940361518407e-05, "loss": 0.0524, "step": 48040 }, { "grad_norm": 0.42027971148490906, "learning_rate": 5.733304811096116e-05, "loss": 0.0501, "step": 48050 }, { "grad_norm": 0.317063570022583, "learning_rate": 5.731669180480713e-05, "loss": 0.0553, "step": 48060 }, { "grad_norm": 0.6502457857131958, "learning_rate": 5.730033469851067e-05, "loss": 0.0557, "step": 48070 }, { "grad_norm": 1.6799800395965576, "learning_rate": 5.7283976793860607e-05, "loss": 0.0507, "step": 48080 }, { "grad_norm": 0.5022237300872803, "learning_rate": 5.726761809264577e-05, "loss": 0.0381, "step": 48090 }, { "grad_norm": 0.4658142030239105, "learning_rate": 5.7251258596655155e-05, "loss": 0.0402, "step": 48100 }, { "grad_norm": 0.4197881817817688, "learning_rate": 5.72348983076778e-05, "loss": 0.0422, "step": 48110 }, { "grad_norm": 0.5256479978561401, "learning_rate": 5.7218537227502854e-05, "loss": 0.0441, "step": 48120 }, { "grad_norm": 0.3272363245487213, "learning_rate": 5.720217535791951e-05, "loss": 0.0309, "step": 48130 }, { "grad_norm": 0.5285094380378723, "learning_rate": 5.718581270071711e-05, "loss": 0.0424, "step": 48140 }, { "grad_norm": 0.5800421237945557, "learning_rate": 5.716944925768505e-05, "loss": 0.0355, "step": 48150 }, { "grad_norm": 0.46296852827072144, "learning_rate": 5.7153085030612786e-05, "loss": 0.0397, "step": 48160 }, { "grad_norm": 0.5352773666381836, "learning_rate": 5.713672002128989e-05, "loss": 0.0543, "step": 48170 }, { "grad_norm": 0.3651444613933563, "learning_rate": 5.712035423150602e-05, "loss": 0.0509, "step": 48180 }, { "grad_norm": 0.4593591094017029, "learning_rate": 5.710398766305094e-05, "loss": 0.0486, "step": 48190 }, { "grad_norm": 0.6016251444816589, "learning_rate": 5.708762031771442e-05, "loss": 0.0462, "step": 48200 }, { "grad_norm": 0.4406954050064087, "learning_rate": 5.707125219728642e-05, "loss": 0.0348, "step": 48210 }, { "grad_norm": 0.4672647714614868, "learning_rate": 5.7054883303556905e-05, "loss": 0.0359, "step": 48220 }, { "grad_norm": 0.5412846803665161, "learning_rate": 5.703851363831595e-05, "loss": 0.0491, "step": 48230 }, { "grad_norm": 0.4782774746417999, "learning_rate": 5.702214320335374e-05, "loss": 0.0531, "step": 48240 }, { "grad_norm": 0.38679149746894836, "learning_rate": 5.7005772000460514e-05, "loss": 0.0473, "step": 48250 }, { "grad_norm": 0.5168710947036743, "learning_rate": 5.698940003142656e-05, "loss": 0.0413, "step": 48260 }, { "grad_norm": 0.43914294242858887, "learning_rate": 5.697302729804236e-05, "loss": 0.0462, "step": 48270 }, { "grad_norm": 0.5147686004638672, "learning_rate": 5.695665380209837e-05, "loss": 0.043, "step": 48280 }, { "grad_norm": 0.7191098928451538, "learning_rate": 5.6940279545385165e-05, "loss": 0.0339, "step": 48290 }, { "grad_norm": 0.7335869073867798, "learning_rate": 5.692390452969344e-05, "loss": 0.0449, "step": 48300 }, { "grad_norm": 1.3694548606872559, "learning_rate": 5.690752875681392e-05, "loss": 0.0478, "step": 48310 }, { "grad_norm": 0.6292743682861328, "learning_rate": 5.6891152228537435e-05, "loss": 0.046, "step": 48320 }, { "grad_norm": 0.9299307465553284, "learning_rate": 5.687477494665492e-05, "loss": 0.0385, "step": 48330 }, { "grad_norm": 0.38756194710731506, "learning_rate": 5.685839691295733e-05, "loss": 0.0404, "step": 48340 }, { "grad_norm": 0.4522903859615326, "learning_rate": 5.6842018129235786e-05, "loss": 0.0464, "step": 48350 }, { "grad_norm": 1.0384981632232666, "learning_rate": 5.6825638597281404e-05, "loss": 0.0504, "step": 48360 }, { "grad_norm": 0.33504655957221985, "learning_rate": 5.680925831888546e-05, "loss": 0.0571, "step": 48370 }, { "grad_norm": 0.35854727029800415, "learning_rate": 5.6792877295839274e-05, "loss": 0.0446, "step": 48380 }, { "grad_norm": 0.5285090804100037, "learning_rate": 5.6776495529934224e-05, "loss": 0.0428, "step": 48390 }, { "grad_norm": 0.7401323318481445, "learning_rate": 5.6760113022961824e-05, "loss": 0.0367, "step": 48400 }, { "grad_norm": 0.46878165006637573, "learning_rate": 5.6743729776713617e-05, "loss": 0.0418, "step": 48410 }, { "grad_norm": 0.7106292247772217, "learning_rate": 5.672734579298126e-05, "loss": 0.036, "step": 48420 }, { "grad_norm": 0.7206700444221497, "learning_rate": 5.671096107355649e-05, "loss": 0.0411, "step": 48430 }, { "grad_norm": 0.37596747279167175, "learning_rate": 5.66945756202311e-05, "loss": 0.0361, "step": 48440 }, { "grad_norm": 0.6250073313713074, "learning_rate": 5.667818943479699e-05, "loss": 0.0467, "step": 48450 }, { "grad_norm": 0.5033596754074097, "learning_rate": 5.666180251904612e-05, "loss": 0.0436, "step": 48460 }, { "grad_norm": 0.578272819519043, "learning_rate": 5.6645414874770555e-05, "loss": 0.0349, "step": 48470 }, { "grad_norm": 0.6197903752326965, "learning_rate": 5.66290265037624e-05, "loss": 0.0493, "step": 48480 }, { "grad_norm": 1.146960735321045, "learning_rate": 5.661263740781386e-05, "loss": 0.0357, "step": 48490 }, { "grad_norm": 0.71073979139328, "learning_rate": 5.6596247588717254e-05, "loss": 0.049, "step": 48500 }, { "grad_norm": 0.5476256012916565, "learning_rate": 5.6579857048264926e-05, "loss": 0.0433, "step": 48510 }, { "grad_norm": 0.43963897228240967, "learning_rate": 5.6563465788249314e-05, "loss": 0.0394, "step": 48520 }, { "grad_norm": 0.4251996576786041, "learning_rate": 5.6547073810462956e-05, "loss": 0.0542, "step": 48530 }, { "grad_norm": 0.4514663815498352, "learning_rate": 5.653068111669846e-05, "loss": 0.0402, "step": 48540 }, { "grad_norm": 0.3772260546684265, "learning_rate": 5.651428770874848e-05, "loss": 0.0494, "step": 48550 }, { "grad_norm": 0.383563369512558, "learning_rate": 5.64978935884058e-05, "loss": 0.0437, "step": 48560 }, { "grad_norm": 0.3920772969722748, "learning_rate": 5.6481498757463244e-05, "loss": 0.0458, "step": 48570 }, { "grad_norm": 0.4474146366119385, "learning_rate": 5.646510321771373e-05, "loss": 0.0483, "step": 48580 }, { "grad_norm": 0.6648979187011719, "learning_rate": 5.644870697095024e-05, "loss": 0.045, "step": 48590 }, { "grad_norm": 0.4027436077594757, "learning_rate": 5.643231001896586e-05, "loss": 0.0367, "step": 48600 }, { "grad_norm": 0.3020753562450409, "learning_rate": 5.6415912363553726e-05, "loss": 0.0589, "step": 48610 }, { "grad_norm": 0.4922792613506317, "learning_rate": 5.639951400650706e-05, "loss": 0.0433, "step": 48620 }, { "grad_norm": 0.40961962938308716, "learning_rate": 5.6383114949619165e-05, "loss": 0.034, "step": 48630 }, { "grad_norm": 0.645460844039917, "learning_rate": 5.636671519468342e-05, "loss": 0.0485, "step": 48640 }, { "grad_norm": 0.38984429836273193, "learning_rate": 5.635031474349327e-05, "loss": 0.0418, "step": 48650 }, { "grad_norm": 0.47356516122817993, "learning_rate": 5.6333913597842246e-05, "loss": 0.0368, "step": 48660 }, { "grad_norm": 0.9928450584411621, "learning_rate": 5.6317511759523955e-05, "loss": 0.0502, "step": 48670 }, { "grad_norm": 0.6559474468231201, "learning_rate": 5.630110923033207e-05, "loss": 0.0456, "step": 48680 }, { "grad_norm": 0.387824684381485, "learning_rate": 5.628470601206036e-05, "loss": 0.0449, "step": 48690 }, { "grad_norm": 0.7590100169181824, "learning_rate": 5.626830210650266e-05, "loss": 0.0463, "step": 48700 }, { "grad_norm": 0.4352564215660095, "learning_rate": 5.625189751545285e-05, "loss": 0.0541, "step": 48710 }, { "grad_norm": 0.648452639579773, "learning_rate": 5.6235492240704936e-05, "loss": 0.0416, "step": 48720 }, { "grad_norm": 0.3948046863079071, "learning_rate": 5.621908628405296e-05, "loss": 0.0479, "step": 48730 }, { "grad_norm": 0.6127821207046509, "learning_rate": 5.620267964729106e-05, "loss": 0.0418, "step": 48740 }, { "grad_norm": 0.32876038551330566, "learning_rate": 5.618627233221344e-05, "loss": 0.0397, "step": 48750 }, { "grad_norm": 0.7992706298828125, "learning_rate": 5.616986434061438e-05, "loss": 0.0553, "step": 48760 }, { "grad_norm": 0.6604421138763428, "learning_rate": 5.615345567428822e-05, "loss": 0.0439, "step": 48770 }, { "grad_norm": 0.5783385038375854, "learning_rate": 5.613704633502941e-05, "loss": 0.0335, "step": 48780 }, { "grad_norm": 0.5262214541435242, "learning_rate": 5.612063632463245e-05, "loss": 0.0464, "step": 48790 }, { "grad_norm": 0.5916292071342468, "learning_rate": 5.610422564489188e-05, "loss": 0.0506, "step": 48800 }, { "grad_norm": 0.7007689476013184, "learning_rate": 5.608781429760239e-05, "loss": 0.0417, "step": 48810 }, { "grad_norm": 0.39947423338890076, "learning_rate": 5.607140228455866e-05, "loss": 0.0431, "step": 48820 }, { "grad_norm": 0.8661858439445496, "learning_rate": 5.605498960755553e-05, "loss": 0.0491, "step": 48830 }, { "grad_norm": 0.4928153157234192, "learning_rate": 5.603857626838782e-05, "loss": 0.0473, "step": 48840 }, { "grad_norm": 0.5983972549438477, "learning_rate": 5.60221622688505e-05, "loss": 0.0398, "step": 48850 }, { "grad_norm": 0.4834091365337372, "learning_rate": 5.6005747610738565e-05, "loss": 0.0382, "step": 48860 }, { "grad_norm": 0.37042975425720215, "learning_rate": 5.59893322958471e-05, "loss": 0.0375, "step": 48870 }, { "grad_norm": 0.5395445227622986, "learning_rate": 5.5972916325971256e-05, "loss": 0.0482, "step": 48880 }, { "grad_norm": 0.46833178400993347, "learning_rate": 5.595649970290628e-05, "loss": 0.0368, "step": 48890 }, { "grad_norm": 0.37448650598526, "learning_rate": 5.5940082428447426e-05, "loss": 0.0516, "step": 48900 }, { "grad_norm": 0.48355621099472046, "learning_rate": 5.592366450439012e-05, "loss": 0.0517, "step": 48910 }, { "grad_norm": 0.5800949931144714, "learning_rate": 5.590724593252975e-05, "loss": 0.0496, "step": 48920 }, { "grad_norm": 0.31678780913352966, "learning_rate": 5.589082671466184e-05, "loss": 0.0405, "step": 48930 }, { "grad_norm": 0.4230719208717346, "learning_rate": 5.587440685258199e-05, "loss": 0.0425, "step": 48940 }, { "grad_norm": 0.55695641040802, "learning_rate": 5.585798634808583e-05, "loss": 0.0359, "step": 48950 }, { "grad_norm": 0.4729233682155609, "learning_rate": 5.584156520296909e-05, "loss": 0.0386, "step": 48960 }, { "grad_norm": 0.6413134336471558, "learning_rate": 5.582514341902757e-05, "loss": 0.0448, "step": 48970 }, { "grad_norm": 0.48399120569229126, "learning_rate": 5.580872099805713e-05, "loss": 0.0439, "step": 48980 }, { "grad_norm": 0.3555474281311035, "learning_rate": 5.5792297941853674e-05, "loss": 0.0438, "step": 48990 }, { "grad_norm": 0.41983762383461, "learning_rate": 5.5775874252213247e-05, "loss": 0.0514, "step": 49000 }, { "grad_norm": 0.6380889415740967, "learning_rate": 5.575944993093189e-05, "loss": 0.0459, "step": 49010 }, { "grad_norm": 0.5059818625450134, "learning_rate": 5.574302497980574e-05, "loss": 0.0419, "step": 49020 }, { "grad_norm": 0.45391157269477844, "learning_rate": 5.5726599400631005e-05, "loss": 0.0446, "step": 49030 }, { "grad_norm": 0.4906439185142517, "learning_rate": 5.571017319520399e-05, "loss": 0.0488, "step": 49040 }, { "grad_norm": 0.38599732518196106, "learning_rate": 5.5693746365320984e-05, "loss": 0.0426, "step": 49050 }, { "grad_norm": 0.45784705877304077, "learning_rate": 5.567731891277848e-05, "loss": 0.0528, "step": 49060 }, { "grad_norm": 0.49389034509658813, "learning_rate": 5.566089083937288e-05, "loss": 0.0474, "step": 49070 }, { "grad_norm": 0.5624223351478577, "learning_rate": 5.564446214690079e-05, "loss": 0.0447, "step": 49080 }, { "grad_norm": 0.429593950510025, "learning_rate": 5.56280328371588e-05, "loss": 0.0404, "step": 49090 }, { "grad_norm": 0.41914957761764526, "learning_rate": 5.5611602911943596e-05, "loss": 0.0334, "step": 49100 }, { "grad_norm": 0.3101005256175995, "learning_rate": 5.559517237305194e-05, "loss": 0.0413, "step": 49110 }, { "grad_norm": 0.5993688106536865, "learning_rate": 5.557874122228064e-05, "loss": 0.0488, "step": 49120 }, { "grad_norm": 0.6335099339485168, "learning_rate": 5.55623094614266e-05, "loss": 0.0406, "step": 49130 }, { "grad_norm": 0.5988418459892273, "learning_rate": 5.5545877092286743e-05, "loss": 0.0486, "step": 49140 }, { "grad_norm": 0.4687717854976654, "learning_rate": 5.552944411665811e-05, "loss": 0.0369, "step": 49150 }, { "grad_norm": 0.4383751153945923, "learning_rate": 5.5513010536337795e-05, "loss": 0.0519, "step": 49160 }, { "grad_norm": 0.5281736254692078, "learning_rate": 5.5496576353122934e-05, "loss": 0.0425, "step": 49170 }, { "grad_norm": 0.5721704363822937, "learning_rate": 5.548014156881074e-05, "loss": 0.038, "step": 49180 }, { "grad_norm": 0.6003644466400146, "learning_rate": 5.5463706185198494e-05, "loss": 0.054, "step": 49190 }, { "grad_norm": 0.30030423402786255, "learning_rate": 5.544727020408358e-05, "loss": 0.0418, "step": 49200 }, { "grad_norm": 0.569137692451477, "learning_rate": 5.5430833627263366e-05, "loss": 0.0377, "step": 49210 }, { "grad_norm": 0.37143728137016296, "learning_rate": 5.5414396456535354e-05, "loss": 0.0373, "step": 49220 }, { "grad_norm": 0.4314470589160919, "learning_rate": 5.539795869369709e-05, "loss": 0.0408, "step": 49230 }, { "grad_norm": 0.5202564597129822, "learning_rate": 5.538152034054618e-05, "loss": 0.0401, "step": 49240 }, { "grad_norm": 0.5332887768745422, "learning_rate": 5.536508139888028e-05, "loss": 0.0412, "step": 49250 }, { "grad_norm": 0.5853897929191589, "learning_rate": 5.534864187049716e-05, "loss": 0.0332, "step": 49260 }, { "grad_norm": 0.21644583344459534, "learning_rate": 5.5332201757194615e-05, "loss": 0.0469, "step": 49270 }, { "grad_norm": 0.4727979004383087, "learning_rate": 5.531576106077049e-05, "loss": 0.038, "step": 49280 }, { "grad_norm": 0.4673824608325958, "learning_rate": 5.529931978302272e-05, "loss": 0.0343, "step": 49290 }, { "grad_norm": 0.25965750217437744, "learning_rate": 5.528287792574932e-05, "loss": 0.0385, "step": 49300 }, { "grad_norm": 0.36896827816963196, "learning_rate": 5.5266435490748324e-05, "loss": 0.0347, "step": 49310 }, { "grad_norm": 0.4634647071361542, "learning_rate": 5.524999247981787e-05, "loss": 0.0407, "step": 49320 }, { "grad_norm": 0.3685705065727234, "learning_rate": 5.523354889475613e-05, "loss": 0.0432, "step": 49330 }, { "grad_norm": 0.5296825766563416, "learning_rate": 5.521710473736134e-05, "loss": 0.042, "step": 49340 }, { "grad_norm": 0.477875292301178, "learning_rate": 5.5200660009431826e-05, "loss": 0.039, "step": 49350 }, { "grad_norm": 0.7025798559188843, "learning_rate": 5.518421471276596e-05, "loss": 0.0396, "step": 49360 }, { "grad_norm": 0.46534207463264465, "learning_rate": 5.5167768849162174e-05, "loss": 0.0399, "step": 49370 }, { "grad_norm": 0.3548068702220917, "learning_rate": 5.515132242041893e-05, "loss": 0.0399, "step": 49380 }, { "grad_norm": 0.3651830852031708, "learning_rate": 5.513487542833483e-05, "loss": 0.0387, "step": 49390 }, { "grad_norm": 0.5441117882728577, "learning_rate": 5.5118427874708466e-05, "loss": 0.0327, "step": 49400 }, { "grad_norm": 0.3028390407562256, "learning_rate": 5.510197976133853e-05, "loss": 0.0327, "step": 49410 }, { "grad_norm": 0.6089686155319214, "learning_rate": 5.508553109002376e-05, "loss": 0.0385, "step": 49420 }, { "grad_norm": 0.380744069814682, "learning_rate": 5.5069081862562957e-05, "loss": 0.0304, "step": 49430 }, { "grad_norm": 0.3068240284919739, "learning_rate": 5.5052632080754965e-05, "loss": 0.0352, "step": 49440 }, { "grad_norm": 0.4606984555721283, "learning_rate": 5.503618174639874e-05, "loss": 0.0411, "step": 49450 }, { "grad_norm": 0.8565353155136108, "learning_rate": 5.501973086129325e-05, "loss": 0.0468, "step": 49460 }, { "grad_norm": 0.3062950670719147, "learning_rate": 5.500327942723753e-05, "loss": 0.0307, "step": 49470 }, { "grad_norm": 0.49581605195999146, "learning_rate": 5.498682744603071e-05, "loss": 0.0466, "step": 49480 }, { "grad_norm": 0.5622185468673706, "learning_rate": 5.497037491947194e-05, "loss": 0.0413, "step": 49490 }, { "grad_norm": 0.45363423228263855, "learning_rate": 5.4953921849360424e-05, "loss": 0.045, "step": 49500 }, { "grad_norm": 0.41966113448143005, "learning_rate": 5.493746823749547e-05, "loss": 0.0484, "step": 49510 }, { "grad_norm": 0.5341507196426392, "learning_rate": 5.49210140856764e-05, "loss": 0.0428, "step": 49520 }, { "grad_norm": 0.38294827938079834, "learning_rate": 5.4904559395702647e-05, "loss": 0.0459, "step": 49530 }, { "grad_norm": 0.4182850420475006, "learning_rate": 5.488810416937364e-05, "loss": 0.0366, "step": 49540 }, { "grad_norm": 0.36688387393951416, "learning_rate": 5.4871648408488926e-05, "loss": 0.0389, "step": 49550 }, { "grad_norm": 0.39283451437950134, "learning_rate": 5.485519211484807e-05, "loss": 0.0294, "step": 49560 }, { "grad_norm": 1.4357717037200928, "learning_rate": 5.4838735290250676e-05, "loss": 0.0439, "step": 49570 }, { "grad_norm": 0.443563848733902, "learning_rate": 5.482227793649649e-05, "loss": 0.0407, "step": 49580 }, { "grad_norm": 0.3963256776332855, "learning_rate": 5.480582005538524e-05, "loss": 0.0382, "step": 49590 }, { "grad_norm": 0.4220190942287445, "learning_rate": 5.478936164871671e-05, "loss": 0.0419, "step": 49600 }, { "grad_norm": 0.498636394739151, "learning_rate": 5.477290271829081e-05, "loss": 0.0492, "step": 49610 }, { "grad_norm": 0.521899402141571, "learning_rate": 5.475644326590744e-05, "loss": 0.0351, "step": 49620 }, { "grad_norm": 0.410793274641037, "learning_rate": 5.473998329336658e-05, "loss": 0.0408, "step": 49630 }, { "grad_norm": 0.45768487453460693, "learning_rate": 5.4723522802468286e-05, "loss": 0.0348, "step": 49640 }, { "grad_norm": 0.6767413020133972, "learning_rate": 5.470706179501264e-05, "loss": 0.0359, "step": 49650 }, { "grad_norm": 0.48106804490089417, "learning_rate": 5.4690600272799776e-05, "loss": 0.0457, "step": 49660 }, { "grad_norm": 0.5256520509719849, "learning_rate": 5.467413823762993e-05, "loss": 0.0485, "step": 49670 }, { "grad_norm": 0.4631390869617462, "learning_rate": 5.465767569130335e-05, "loss": 0.0414, "step": 49680 }, { "grad_norm": 0.919440507888794, "learning_rate": 5.464121263562036e-05, "loss": 0.0419, "step": 49690 }, { "grad_norm": 0.3054601550102234, "learning_rate": 5.4624749072381343e-05, "loss": 0.0359, "step": 49700 }, { "grad_norm": 0.5532187223434448, "learning_rate": 5.460828500338672e-05, "loss": 0.0426, "step": 49710 }, { "grad_norm": 0.5327528119087219, "learning_rate": 5.459182043043698e-05, "loss": 0.0482, "step": 49720 }, { "grad_norm": 0.30494338274002075, "learning_rate": 5.457535535533265e-05, "loss": 0.0392, "step": 49730 }, { "grad_norm": 0.47106966376304626, "learning_rate": 5.4558889779874334e-05, "loss": 0.0463, "step": 49740 }, { "grad_norm": 0.5694872736930847, "learning_rate": 5.454242370586269e-05, "loss": 0.0382, "step": 49750 }, { "grad_norm": 0.47301211953163147, "learning_rate": 5.452595713509843e-05, "loss": 0.0383, "step": 49760 }, { "grad_norm": 0.40529194474220276, "learning_rate": 5.4509490069382274e-05, "loss": 0.0502, "step": 49770 }, { "grad_norm": 0.35457706451416016, "learning_rate": 5.449302251051509e-05, "loss": 0.0313, "step": 49780 }, { "grad_norm": 0.47815054655075073, "learning_rate": 5.44765544602977e-05, "loss": 0.0512, "step": 49790 }, { "grad_norm": 0.309627890586853, "learning_rate": 5.446008592053107e-05, "loss": 0.0379, "step": 49800 }, { "grad_norm": 0.4335368573665619, "learning_rate": 5.4443616893016135e-05, "loss": 0.039, "step": 49810 }, { "grad_norm": 0.778838574886322, "learning_rate": 5.4427147379553934e-05, "loss": 0.0467, "step": 49820 }, { "grad_norm": 0.6599646806716919, "learning_rate": 5.441067738194556e-05, "loss": 0.0382, "step": 49830 }, { "grad_norm": 0.39281660318374634, "learning_rate": 5.439420690199214e-05, "loss": 0.0494, "step": 49840 }, { "grad_norm": 0.4459367096424103, "learning_rate": 5.437773594149484e-05, "loss": 0.0387, "step": 49850 }, { "grad_norm": 0.5059236288070679, "learning_rate": 5.436126450225495e-05, "loss": 0.0413, "step": 49860 }, { "grad_norm": 0.43506157398223877, "learning_rate": 5.434479258607371e-05, "loss": 0.0414, "step": 49870 }, { "grad_norm": 0.3933638036251068, "learning_rate": 5.43283201947525e-05, "loss": 0.0396, "step": 49880 }, { "grad_norm": 0.5292896628379822, "learning_rate": 5.431184733009268e-05, "loss": 0.0514, "step": 49890 }, { "grad_norm": 0.42521408200263977, "learning_rate": 5.4295373993895736e-05, "loss": 0.0414, "step": 49900 }, { "grad_norm": 0.4251049757003784, "learning_rate": 5.4278900187963157e-05, "loss": 0.0468, "step": 49910 }, { "grad_norm": 0.38153061270713806, "learning_rate": 5.426242591409646e-05, "loss": 0.0381, "step": 49920 }, { "grad_norm": 0.3665234446525574, "learning_rate": 5.42459511740973e-05, "loss": 0.0365, "step": 49930 }, { "grad_norm": 0.450845330953598, "learning_rate": 5.422947596976729e-05, "loss": 0.0367, "step": 49940 }, { "grad_norm": 0.5203783512115479, "learning_rate": 5.4213000302908134e-05, "loss": 0.0458, "step": 49950 }, { "grad_norm": 0.4469897150993347, "learning_rate": 5.419652417532162e-05, "loss": 0.0364, "step": 49960 }, { "grad_norm": 0.5270485281944275, "learning_rate": 5.4180047588809534e-05, "loss": 0.0395, "step": 49970 }, { "grad_norm": 0.4060845971107483, "learning_rate": 5.4163570545173704e-05, "loss": 0.038, "step": 49980 }, { "grad_norm": 0.5322386026382446, "learning_rate": 5.414709304621608e-05, "loss": 0.0363, "step": 49990 }, { "grad_norm": 0.3563616871833801, "learning_rate": 5.413061509373858e-05, "loss": 0.05, "step": 50000 }, { "grad_norm": 0.4681228995323181, "learning_rate": 5.411413668954324e-05, "loss": 0.0428, "step": 50010 }, { "grad_norm": 0.4461292624473572, "learning_rate": 5.4097657835432083e-05, "loss": 0.0395, "step": 50020 }, { "grad_norm": 0.41944316029548645, "learning_rate": 5.408117853320723e-05, "loss": 0.0436, "step": 50030 }, { "grad_norm": 0.4243741035461426, "learning_rate": 5.406469878467084e-05, "loss": 0.0329, "step": 50040 }, { "grad_norm": 0.5204684734344482, "learning_rate": 5.404821859162509e-05, "loss": 0.045, "step": 50050 }, { "grad_norm": 0.33325129747390747, "learning_rate": 5.403173795587225e-05, "loss": 0.0388, "step": 50060 }, { "grad_norm": 0.5556039810180664, "learning_rate": 5.4015256879214606e-05, "loss": 0.0496, "step": 50070 }, { "grad_norm": 0.30666980147361755, "learning_rate": 5.39987753634545e-05, "loss": 0.0395, "step": 50080 }, { "grad_norm": 0.28286805748939514, "learning_rate": 5.3982293410394335e-05, "loss": 0.0435, "step": 50090 }, { "grad_norm": 0.3506399393081665, "learning_rate": 5.396581102183655e-05, "loss": 0.0413, "step": 50100 }, { "grad_norm": 0.5382381677627563, "learning_rate": 5.3949328199583634e-05, "loss": 0.0378, "step": 50110 }, { "grad_norm": 0.505147397518158, "learning_rate": 5.393284494543813e-05, "loss": 0.0518, "step": 50120 }, { "grad_norm": 0.3422546684741974, "learning_rate": 5.391636126120262e-05, "loss": 0.037, "step": 50130 }, { "grad_norm": 0.44292524456977844, "learning_rate": 5.389987714867971e-05, "loss": 0.0377, "step": 50140 }, { "grad_norm": 0.5532006025314331, "learning_rate": 5.3883392609672114e-05, "loss": 0.0539, "step": 50150 }, { "grad_norm": 0.3554825484752655, "learning_rate": 5.3866907645982545e-05, "loss": 0.0438, "step": 50160 }, { "grad_norm": 0.6127126812934875, "learning_rate": 5.385042225941375e-05, "loss": 0.0453, "step": 50170 }, { "grad_norm": 0.3885565996170044, "learning_rate": 5.3833936451768585e-05, "loss": 0.0438, "step": 50180 }, { "grad_norm": 0.4678890109062195, "learning_rate": 5.3817450224849884e-05, "loss": 0.0428, "step": 50190 }, { "grad_norm": 0.5925439596176147, "learning_rate": 5.380096358046056e-05, "loss": 0.0598, "step": 50200 }, { "grad_norm": 0.4141339957714081, "learning_rate": 5.378447652040359e-05, "loss": 0.0397, "step": 50210 }, { "grad_norm": 0.47753849625587463, "learning_rate": 5.376798904648195e-05, "loss": 0.0383, "step": 50220 }, { "grad_norm": 0.6321824789047241, "learning_rate": 5.375150116049869e-05, "loss": 0.0417, "step": 50230 }, { "grad_norm": 0.3650400638580322, "learning_rate": 5.373501286425691e-05, "loss": 0.0317, "step": 50240 }, { "grad_norm": 0.3459610342979431, "learning_rate": 5.3718524159559726e-05, "loss": 0.0438, "step": 50250 }, { "grad_norm": 0.4332207143306732, "learning_rate": 5.370203504821034e-05, "loss": 0.0463, "step": 50260 }, { "grad_norm": 0.2844315469264984, "learning_rate": 5.368554553201196e-05, "loss": 0.0433, "step": 50270 }, { "grad_norm": 0.35873517394065857, "learning_rate": 5.3669055612767874e-05, "loss": 0.0543, "step": 50280 }, { "grad_norm": 0.5274669528007507, "learning_rate": 5.3652565292281376e-05, "loss": 0.0412, "step": 50290 }, { "grad_norm": 0.5371641516685486, "learning_rate": 5.3636074572355824e-05, "loss": 0.0397, "step": 50300 }, { "grad_norm": 0.4749206602573395, "learning_rate": 5.361958345479463e-05, "loss": 0.0394, "step": 50310 }, { "grad_norm": 0.2991260588169098, "learning_rate": 5.360309194140123e-05, "loss": 0.0445, "step": 50320 }, { "grad_norm": 0.45105957984924316, "learning_rate": 5.3586600033979106e-05, "loss": 0.0344, "step": 50330 }, { "grad_norm": 0.4293876588344574, "learning_rate": 5.357010773433181e-05, "loss": 0.0392, "step": 50340 }, { "grad_norm": 0.4765579104423523, "learning_rate": 5.3553615044262895e-05, "loss": 0.0428, "step": 50350 }, { "grad_norm": 0.35759735107421875, "learning_rate": 5.3537121965575964e-05, "loss": 0.0387, "step": 50360 }, { "grad_norm": 0.3599848449230194, "learning_rate": 5.352062850007471e-05, "loss": 0.0501, "step": 50370 }, { "grad_norm": 0.2645256221294403, "learning_rate": 5.350413464956282e-05, "loss": 0.0392, "step": 50380 }, { "grad_norm": 0.5486769676208496, "learning_rate": 5.348764041584403e-05, "loss": 0.0447, "step": 50390 }, { "grad_norm": 0.3586057126522064, "learning_rate": 5.347114580072212e-05, "loss": 0.0415, "step": 50400 }, { "grad_norm": 0.45016300678253174, "learning_rate": 5.345465080600093e-05, "loss": 0.0466, "step": 50410 }, { "grad_norm": 0.3715272545814514, "learning_rate": 5.3438155433484326e-05, "loss": 0.0418, "step": 50420 }, { "grad_norm": 0.764342188835144, "learning_rate": 5.3421659684976197e-05, "loss": 0.0395, "step": 50430 }, { "grad_norm": 0.8219375014305115, "learning_rate": 5.340516356228052e-05, "loss": 0.0377, "step": 50440 }, { "grad_norm": 0.3988596796989441, "learning_rate": 5.338866706720128e-05, "loss": 0.0429, "step": 50450 }, { "grad_norm": 0.3062116801738739, "learning_rate": 5.337217020154249e-05, "loss": 0.0397, "step": 50460 }, { "grad_norm": 0.5387414693832397, "learning_rate": 5.335567296710825e-05, "loss": 0.0421, "step": 50470 }, { "grad_norm": 0.5186859369277954, "learning_rate": 5.333917536570265e-05, "loss": 0.0396, "step": 50480 }, { "grad_norm": 0.41867417097091675, "learning_rate": 5.332267739912986e-05, "loss": 0.0404, "step": 50490 }, { "grad_norm": 0.4608740210533142, "learning_rate": 5.330617906919405e-05, "loss": 0.0459, "step": 50500 }, { "grad_norm": 0.35173770785331726, "learning_rate": 5.328968037769949e-05, "loss": 0.0401, "step": 50510 }, { "grad_norm": 0.3628459572792053, "learning_rate": 5.327318132645042e-05, "loss": 0.046, "step": 50520 }, { "grad_norm": 0.4990937411785126, "learning_rate": 5.325668191725116e-05, "loss": 0.0389, "step": 50530 }, { "grad_norm": 0.5249059200286865, "learning_rate": 5.324018215190606e-05, "loss": 0.0421, "step": 50540 }, { "grad_norm": 0.46982523798942566, "learning_rate": 5.3223682032219515e-05, "loss": 0.0447, "step": 50550 }, { "grad_norm": 0.30002132058143616, "learning_rate": 5.320718155999595e-05, "loss": 0.0473, "step": 50560 }, { "grad_norm": 0.4828188717365265, "learning_rate": 5.3190680737039835e-05, "loss": 0.0522, "step": 50570 }, { "grad_norm": 0.4123810827732086, "learning_rate": 5.317417956515567e-05, "loss": 0.0405, "step": 50580 }, { "grad_norm": 0.42113712430000305, "learning_rate": 5.315767804614798e-05, "loss": 0.0396, "step": 50590 }, { "grad_norm": 0.47497060894966125, "learning_rate": 5.3141176181821395e-05, "loss": 0.0487, "step": 50600 }, { "grad_norm": 0.39382168650627136, "learning_rate": 5.31246739739805e-05, "loss": 0.0309, "step": 50610 }, { "grad_norm": 0.4392203092575073, "learning_rate": 5.310817142442994e-05, "loss": 0.0434, "step": 50620 }, { "grad_norm": 1.039960265159607, "learning_rate": 5.309166853497445e-05, "loss": 0.0418, "step": 50630 }, { "grad_norm": 0.504460334777832, "learning_rate": 5.307516530741873e-05, "loss": 0.043, "step": 50640 }, { "grad_norm": 0.3844486474990845, "learning_rate": 5.305866174356754e-05, "loss": 0.0395, "step": 50650 }, { "grad_norm": 0.31292539834976196, "learning_rate": 5.304215784522571e-05, "loss": 0.0355, "step": 50660 }, { "grad_norm": 0.34709909558296204, "learning_rate": 5.302565361419808e-05, "loss": 0.0533, "step": 50670 }, { "grad_norm": 0.450510710477829, "learning_rate": 5.3009149052289507e-05, "loss": 0.0383, "step": 50680 }, { "grad_norm": 0.4088190793991089, "learning_rate": 5.299264416130493e-05, "loss": 0.0371, "step": 50690 }, { "grad_norm": 0.3913087844848633, "learning_rate": 5.297613894304928e-05, "loss": 0.0413, "step": 50700 }, { "grad_norm": 0.581973671913147, "learning_rate": 5.2959633399327534e-05, "loss": 0.0343, "step": 50710 }, { "grad_norm": 0.5541800260543823, "learning_rate": 5.294312753194476e-05, "loss": 0.0345, "step": 50720 }, { "grad_norm": 0.3780294358730316, "learning_rate": 5.292662134270596e-05, "loss": 0.0427, "step": 50730 }, { "grad_norm": 0.6559217572212219, "learning_rate": 5.291011483341626e-05, "loss": 0.0449, "step": 50740 }, { "grad_norm": 0.750505268573761, "learning_rate": 5.2893608005880767e-05, "loss": 0.0551, "step": 50750 }, { "grad_norm": 0.35325390100479126, "learning_rate": 5.287710086190467e-05, "loss": 0.0423, "step": 50760 }, { "grad_norm": 0.45233821868896484, "learning_rate": 5.286059340329314e-05, "loss": 0.0395, "step": 50770 }, { "grad_norm": 0.4684089422225952, "learning_rate": 5.284408563185141e-05, "loss": 0.0451, "step": 50780 }, { "grad_norm": 0.5695167183876038, "learning_rate": 5.282757754938476e-05, "loss": 0.043, "step": 50790 }, { "grad_norm": 0.35464656352996826, "learning_rate": 5.281106915769849e-05, "loss": 0.0446, "step": 50800 }, { "grad_norm": 0.4402725100517273, "learning_rate": 5.2794560458597897e-05, "loss": 0.0435, "step": 50810 }, { "grad_norm": 0.38709455728530884, "learning_rate": 5.27780514538884e-05, "loss": 0.0372, "step": 50820 }, { "grad_norm": 1.136262059211731, "learning_rate": 5.2761542145375365e-05, "loss": 0.0453, "step": 50830 }, { "grad_norm": 0.5046416521072388, "learning_rate": 5.274503253486421e-05, "loss": 0.0405, "step": 50840 }, { "grad_norm": 0.41476818919181824, "learning_rate": 5.272852262416046e-05, "loss": 0.0377, "step": 50850 }, { "grad_norm": 0.39015305042266846, "learning_rate": 5.2712012415069555e-05, "loss": 0.0452, "step": 50860 }, { "grad_norm": 0.5554942488670349, "learning_rate": 5.269550190939705e-05, "loss": 0.047, "step": 50870 }, { "grad_norm": 0.6780744791030884, "learning_rate": 5.267899110894852e-05, "loss": 0.0421, "step": 50880 }, { "grad_norm": 0.3279956579208374, "learning_rate": 5.266248001552955e-05, "loss": 0.0445, "step": 50890 }, { "grad_norm": 0.39878493547439575, "learning_rate": 5.264596863094575e-05, "loss": 0.0471, "step": 50900 }, { "grad_norm": 0.768897533416748, "learning_rate": 5.262945695700282e-05, "loss": 0.0465, "step": 50910 }, { "grad_norm": 0.29980283975601196, "learning_rate": 5.261294499550643e-05, "loss": 0.0365, "step": 50920 }, { "grad_norm": 0.5058324337005615, "learning_rate": 5.25964327482623e-05, "loss": 0.0368, "step": 50930 }, { "grad_norm": 0.4815191328525543, "learning_rate": 5.257992021707617e-05, "loss": 0.0421, "step": 50940 }, { "grad_norm": 0.3789729177951813, "learning_rate": 5.256340740375387e-05, "loss": 0.0444, "step": 50950 }, { "grad_norm": 0.43167436122894287, "learning_rate": 5.254689431010117e-05, "loss": 0.05, "step": 50960 }, { "grad_norm": 0.4243313670158386, "learning_rate": 5.253038093792395e-05, "loss": 0.0339, "step": 50970 }, { "grad_norm": 0.40912967920303345, "learning_rate": 5.251386728902806e-05, "loss": 0.041, "step": 50980 }, { "grad_norm": 0.5248383283615112, "learning_rate": 5.2497353365219446e-05, "loss": 0.0417, "step": 50990 }, { "grad_norm": 0.34628626704216003, "learning_rate": 5.2480839168304e-05, "loss": 0.0373, "step": 51000 }, { "grad_norm": 0.3656860888004303, "learning_rate": 5.246432470008773e-05, "loss": 0.0376, "step": 51010 }, { "grad_norm": 0.33790794014930725, "learning_rate": 5.2447809962376616e-05, "loss": 0.0424, "step": 51020 }, { "grad_norm": 0.8207845687866211, "learning_rate": 5.243129495697666e-05, "loss": 0.0342, "step": 51030 }, { "grad_norm": 0.27955564856529236, "learning_rate": 5.241477968569397e-05, "loss": 0.0452, "step": 51040 }, { "grad_norm": 0.5159374475479126, "learning_rate": 5.23982641503346e-05, "loss": 0.0493, "step": 51050 }, { "grad_norm": 0.459383487701416, "learning_rate": 5.238174835270465e-05, "loss": 0.0519, "step": 51060 }, { "grad_norm": 0.5397163033485413, "learning_rate": 5.23652322946103e-05, "loss": 0.0389, "step": 51070 }, { "grad_norm": 0.24024996161460876, "learning_rate": 5.234871597785771e-05, "loss": 0.0338, "step": 51080 }, { "grad_norm": 1.165590524673462, "learning_rate": 5.233219940425306e-05, "loss": 0.0503, "step": 51090 }, { "grad_norm": 0.7421615719795227, "learning_rate": 5.2315682575602576e-05, "loss": 0.064, "step": 51100 }, { "grad_norm": 0.690802276134491, "learning_rate": 5.229916549371254e-05, "loss": 0.0361, "step": 51110 }, { "grad_norm": 0.36253640055656433, "learning_rate": 5.228264816038924e-05, "loss": 0.035, "step": 51120 }, { "grad_norm": 0.8024328351020813, "learning_rate": 5.2266130577438935e-05, "loss": 0.0377, "step": 51130 }, { "grad_norm": 0.36698493361473083, "learning_rate": 5.224961274666802e-05, "loss": 0.0403, "step": 51140 }, { "grad_norm": 0.6637650728225708, "learning_rate": 5.223309466988283e-05, "loss": 0.0487, "step": 51150 }, { "grad_norm": 0.42854321002960205, "learning_rate": 5.221657634888976e-05, "loss": 0.0477, "step": 51160 }, { "grad_norm": 0.4962902069091797, "learning_rate": 5.2200057785495235e-05, "loss": 0.0575, "step": 51170 }, { "grad_norm": 0.4808899164199829, "learning_rate": 5.2183538981505706e-05, "loss": 0.0438, "step": 51180 }, { "grad_norm": 0.3827960193157196, "learning_rate": 5.216701993872762e-05, "loss": 0.0375, "step": 51190 }, { "grad_norm": 0.4158838391304016, "learning_rate": 5.215050065896752e-05, "loss": 0.055, "step": 51200 }, { "grad_norm": 0.5956647396087646, "learning_rate": 5.2133981144031875e-05, "loss": 0.0631, "step": 51210 }, { "grad_norm": 0.5073121786117554, "learning_rate": 5.211746139572727e-05, "loss": 0.0445, "step": 51220 }, { "grad_norm": 0.32080739736557007, "learning_rate": 5.210094141586026e-05, "loss": 0.0429, "step": 51230 }, { "grad_norm": 0.41983741521835327, "learning_rate": 5.2084421206237476e-05, "loss": 0.0451, "step": 51240 }, { "grad_norm": 0.4470382630825043, "learning_rate": 5.20679007686655e-05, "loss": 0.0468, "step": 51250 }, { "grad_norm": 0.4071372449398041, "learning_rate": 5.205138010495102e-05, "loss": 0.0465, "step": 51260 }, { "grad_norm": 0.45826148986816406, "learning_rate": 5.20348592169007e-05, "loss": 0.0439, "step": 51270 }, { "grad_norm": 0.7889461517333984, "learning_rate": 5.201833810632123e-05, "loss": 0.0388, "step": 51280 }, { "grad_norm": 0.7935069799423218, "learning_rate": 5.200181677501933e-05, "loss": 0.0488, "step": 51290 }, { "grad_norm": 0.46992409229278564, "learning_rate": 5.1985295224801777e-05, "loss": 0.0423, "step": 51300 }, { "grad_norm": 0.4663240611553192, "learning_rate": 5.196877345747531e-05, "loss": 0.0448, "step": 51310 }, { "grad_norm": 0.44673049449920654, "learning_rate": 5.195225147484674e-05, "loss": 0.0498, "step": 51320 }, { "grad_norm": 0.6562634706497192, "learning_rate": 5.193572927872288e-05, "loss": 0.0469, "step": 51330 }, { "grad_norm": 0.5177887678146362, "learning_rate": 5.191920687091059e-05, "loss": 0.0422, "step": 51340 }, { "grad_norm": 0.5006603598594666, "learning_rate": 5.1902684253216695e-05, "loss": 0.0391, "step": 51350 }, { "grad_norm": 0.3904896080493927, "learning_rate": 5.1886161427448134e-05, "loss": 0.0418, "step": 51360 }, { "grad_norm": 0.314704030752182, "learning_rate": 5.186963839541179e-05, "loss": 0.0372, "step": 51370 }, { "grad_norm": 0.49274593591690063, "learning_rate": 5.185311515891459e-05, "loss": 0.0315, "step": 51380 }, { "grad_norm": 0.4844326972961426, "learning_rate": 5.1836591719763516e-05, "loss": 0.0369, "step": 51390 }, { "grad_norm": 0.43342325091362, "learning_rate": 5.182006807976554e-05, "loss": 0.0443, "step": 51400 }, { "grad_norm": 0.23696282505989075, "learning_rate": 5.180354424072763e-05, "loss": 0.0349, "step": 51410 }, { "grad_norm": 0.5642008781433105, "learning_rate": 5.178702020445686e-05, "loss": 0.0441, "step": 51420 }, { "grad_norm": 0.2343389391899109, "learning_rate": 5.177049597276024e-05, "loss": 0.0386, "step": 51430 }, { "grad_norm": 0.44833680987358093, "learning_rate": 5.175397154744482e-05, "loss": 0.0426, "step": 51440 }, { "grad_norm": 0.4244012236595154, "learning_rate": 5.1737446930317736e-05, "loss": 0.0312, "step": 51450 }, { "grad_norm": 0.604742705821991, "learning_rate": 5.172092212318604e-05, "loss": 0.05, "step": 51460 }, { "grad_norm": 0.28721022605895996, "learning_rate": 5.170439712785692e-05, "loss": 0.0456, "step": 51470 }, { "grad_norm": 0.5257278084754944, "learning_rate": 5.1687871946137467e-05, "loss": 0.0357, "step": 51480 }, { "grad_norm": 0.41731658577919006, "learning_rate": 5.16713465798349e-05, "loss": 0.0402, "step": 51490 }, { "grad_norm": 0.3809053301811218, "learning_rate": 5.165482103075637e-05, "loss": 0.0475, "step": 51500 }, { "grad_norm": 0.393372118473053, "learning_rate": 5.163829530070909e-05, "loss": 0.0323, "step": 51510 }, { "grad_norm": 0.5430035591125488, "learning_rate": 5.162176939150032e-05, "loss": 0.045, "step": 51520 }, { "grad_norm": 0.6359468698501587, "learning_rate": 5.160524330493728e-05, "loss": 0.0394, "step": 51530 }, { "grad_norm": 0.30649587512016296, "learning_rate": 5.158871704282724e-05, "loss": 0.0368, "step": 51540 }, { "grad_norm": 0.4339819848537445, "learning_rate": 5.15721906069775e-05, "loss": 0.0474, "step": 51550 }, { "grad_norm": 0.45359858870506287, "learning_rate": 5.155566399919536e-05, "loss": 0.0482, "step": 51560 }, { "grad_norm": 0.5168799757957458, "learning_rate": 5.153913722128813e-05, "loss": 0.0428, "step": 51570 }, { "grad_norm": 0.6899242401123047, "learning_rate": 5.1522610275063176e-05, "loss": 0.0419, "step": 51580 }, { "grad_norm": 0.4009862244129181, "learning_rate": 5.150608316232786e-05, "loss": 0.0387, "step": 51590 }, { "grad_norm": 0.5086601972579956, "learning_rate": 5.148955588488955e-05, "loss": 0.0366, "step": 51600 }, { "grad_norm": 0.3678295612335205, "learning_rate": 5.147302844455564e-05, "loss": 0.0415, "step": 51610 }, { "grad_norm": 0.38681715726852417, "learning_rate": 5.145650084313356e-05, "loss": 0.0472, "step": 51620 }, { "grad_norm": 0.3890168368816376, "learning_rate": 5.1439973082430734e-05, "loss": 0.0424, "step": 51630 }, { "grad_norm": 0.34188300371170044, "learning_rate": 5.142344516425461e-05, "loss": 0.0398, "step": 51640 }, { "grad_norm": 0.3625163435935974, "learning_rate": 5.1406917090412673e-05, "loss": 0.0476, "step": 51650 }, { "grad_norm": 0.49732258915901184, "learning_rate": 5.1390388862712404e-05, "loss": 0.0421, "step": 51660 }, { "grad_norm": 0.696510374546051, "learning_rate": 5.137386048296129e-05, "loss": 0.0464, "step": 51670 }, { "grad_norm": 0.5486563444137573, "learning_rate": 5.135733195296686e-05, "loss": 0.052, "step": 51680 }, { "grad_norm": 0.5196937918663025, "learning_rate": 5.134080327453664e-05, "loss": 0.0458, "step": 51690 }, { "grad_norm": 0.2727988362312317, "learning_rate": 5.1324274449478204e-05, "loss": 0.0395, "step": 51700 }, { "grad_norm": 0.27570655941963196, "learning_rate": 5.130774547959909e-05, "loss": 0.0457, "step": 51710 }, { "grad_norm": 0.5670202970504761, "learning_rate": 5.129121636670692e-05, "loss": 0.0419, "step": 51720 }, { "grad_norm": 0.3790082633495331, "learning_rate": 5.1274687112609246e-05, "loss": 0.0283, "step": 51730 }, { "grad_norm": 0.4095708131790161, "learning_rate": 5.125815771911372e-05, "loss": 0.0429, "step": 51740 }, { "grad_norm": 0.9456385970115662, "learning_rate": 5.1241628188027956e-05, "loss": 0.0407, "step": 51750 }, { "grad_norm": 0.5655332207679749, "learning_rate": 5.1225098521159585e-05, "loss": 0.0408, "step": 51760 }, { "grad_norm": 0.432439923286438, "learning_rate": 5.12085687203163e-05, "loss": 0.0338, "step": 51770 }, { "grad_norm": 0.6319143176078796, "learning_rate": 5.119203878730575e-05, "loss": 0.0377, "step": 51780 }, { "grad_norm": 0.7705896496772766, "learning_rate": 5.117550872393564e-05, "loss": 0.0442, "step": 51790 }, { "grad_norm": 0.5284075140953064, "learning_rate": 5.1158978532013645e-05, "loss": 0.0481, "step": 51800 }, { "grad_norm": 0.42729511857032776, "learning_rate": 5.114244821334751e-05, "loss": 0.0469, "step": 51810 }, { "grad_norm": 0.47476422786712646, "learning_rate": 5.112591776974497e-05, "loss": 0.0356, "step": 51820 }, { "grad_norm": 0.4085111618041992, "learning_rate": 5.110938720301374e-05, "loss": 0.0371, "step": 51830 }, { "grad_norm": 0.21603184938430786, "learning_rate": 5.1092856514961595e-05, "loss": 0.0439, "step": 51840 }, { "grad_norm": 0.932982325553894, "learning_rate": 5.107632570739631e-05, "loss": 0.0498, "step": 51850 }, { "grad_norm": 0.504852831363678, "learning_rate": 5.1059794782125645e-05, "loss": 0.0498, "step": 51860 }, { "grad_norm": 0.5705344080924988, "learning_rate": 5.104326374095744e-05, "loss": 0.0544, "step": 51870 }, { "grad_norm": 0.377479612827301, "learning_rate": 5.102673258569947e-05, "loss": 0.0402, "step": 51880 }, { "grad_norm": 0.46192291378974915, "learning_rate": 5.101020131815956e-05, "loss": 0.0477, "step": 51890 }, { "grad_norm": 0.5565617680549622, "learning_rate": 5.099366994014556e-05, "loss": 0.0514, "step": 51900 }, { "grad_norm": 0.514733612537384, "learning_rate": 5.097713845346532e-05, "loss": 0.046, "step": 51910 }, { "grad_norm": 0.45284947752952576, "learning_rate": 5.096060685992668e-05, "loss": 0.041, "step": 51920 }, { "grad_norm": 0.7745794057846069, "learning_rate": 5.0944075161337514e-05, "loss": 0.0528, "step": 51930 }, { "grad_norm": 0.4576669931411743, "learning_rate": 5.09275433595057e-05, "loss": 0.044, "step": 51940 }, { "grad_norm": 0.6931127309799194, "learning_rate": 5.0911011456239157e-05, "loss": 0.0398, "step": 51950 }, { "grad_norm": 0.5753313302993774, "learning_rate": 5.089447945334574e-05, "loss": 0.0438, "step": 51960 }, { "grad_norm": 0.4318084418773651, "learning_rate": 5.087794735263343e-05, "loss": 0.0371, "step": 51970 }, { "grad_norm": 0.5233256220817566, "learning_rate": 5.086141515591011e-05, "loss": 0.04, "step": 51980 }, { "grad_norm": 0.39488285779953003, "learning_rate": 5.084488286498371e-05, "loss": 0.0414, "step": 51990 }, { "grad_norm": 0.3546738028526306, "learning_rate": 5.08283504816622e-05, "loss": 0.0579, "step": 52000 }, { "grad_norm": 0.4481624364852905, "learning_rate": 5.081181800775353e-05, "loss": 0.0571, "step": 52010 }, { "grad_norm": 0.302346408367157, "learning_rate": 5.079528544506565e-05, "loss": 0.0449, "step": 52020 }, { "grad_norm": 0.43540558218955994, "learning_rate": 5.077875279540657e-05, "loss": 0.0425, "step": 52030 }, { "grad_norm": 0.3535500168800354, "learning_rate": 5.076222006058424e-05, "loss": 0.0482, "step": 52040 }, { "grad_norm": 0.47374704480171204, "learning_rate": 5.074568724240667e-05, "loss": 0.035, "step": 52050 }, { "grad_norm": 0.6117521524429321, "learning_rate": 5.072915434268187e-05, "loss": 0.0412, "step": 52060 }, { "grad_norm": 0.4458077847957611, "learning_rate": 5.071262136321785e-05, "loss": 0.0429, "step": 52070 }, { "grad_norm": 0.5405920147895813, "learning_rate": 5.069608830582262e-05, "loss": 0.0479, "step": 52080 }, { "grad_norm": 0.37497279047966003, "learning_rate": 5.067955517230423e-05, "loss": 0.0377, "step": 52090 }, { "grad_norm": 0.4632803797721863, "learning_rate": 5.06630219644707e-05, "loss": 0.0562, "step": 52100 }, { "grad_norm": 0.6211149096488953, "learning_rate": 5.064648868413008e-05, "loss": 0.0381, "step": 52110 }, { "grad_norm": 0.5241510272026062, "learning_rate": 5.062995533309044e-05, "loss": 0.0423, "step": 52120 }, { "grad_norm": 0.29629647731781006, "learning_rate": 5.061342191315983e-05, "loss": 0.0427, "step": 52130 }, { "grad_norm": 0.4104979634284973, "learning_rate": 5.0596888426146325e-05, "loss": 0.0531, "step": 52140 }, { "grad_norm": 0.5809335708618164, "learning_rate": 5.058035487385798e-05, "loss": 0.035, "step": 52150 }, { "grad_norm": 0.4417819380760193, "learning_rate": 5.0563821258102915e-05, "loss": 0.0442, "step": 52160 }, { "grad_norm": 0.4943191409111023, "learning_rate": 5.05472875806892e-05, "loss": 0.0419, "step": 52170 }, { "grad_norm": 0.4226415753364563, "learning_rate": 5.053075384342494e-05, "loss": 0.0444, "step": 52180 }, { "grad_norm": 0.367682546377182, "learning_rate": 5.0514220048118234e-05, "loss": 0.042, "step": 52190 }, { "grad_norm": 1.1691572666168213, "learning_rate": 5.049768619657721e-05, "loss": 0.0518, "step": 52200 }, { "grad_norm": 0.33374765515327454, "learning_rate": 5.048115229060996e-05, "loss": 0.039, "step": 52210 }, { "grad_norm": 0.6613821387290955, "learning_rate": 5.046461833202462e-05, "loss": 0.0436, "step": 52220 }, { "grad_norm": 0.2917162775993347, "learning_rate": 5.044808432262932e-05, "loss": 0.0476, "step": 52230 }, { "grad_norm": 0.3076675236225128, "learning_rate": 5.043155026423219e-05, "loss": 0.0387, "step": 52240 }, { "grad_norm": 0.2943373918533325, "learning_rate": 5.041501615864137e-05, "loss": 0.0355, "step": 52250 }, { "grad_norm": 0.31100162863731384, "learning_rate": 5.039848200766502e-05, "loss": 0.0336, "step": 52260 }, { "grad_norm": 0.2877359092235565, "learning_rate": 5.038194781311125e-05, "loss": 0.0423, "step": 52270 }, { "grad_norm": 0.3266395628452301, "learning_rate": 5.0365413576788264e-05, "loss": 0.0342, "step": 52280 }, { "grad_norm": 0.4084060788154602, "learning_rate": 5.0348879300504194e-05, "loss": 0.04, "step": 52290 }, { "grad_norm": 0.4028346538543701, "learning_rate": 5.033234498606721e-05, "loss": 0.0338, "step": 52300 }, { "grad_norm": 0.42145663499832153, "learning_rate": 5.0315810635285476e-05, "loss": 0.0411, "step": 52310 }, { "grad_norm": 0.5713338255882263, "learning_rate": 5.0299276249967164e-05, "loss": 0.0376, "step": 52320 }, { "grad_norm": 0.30062153935432434, "learning_rate": 5.0282741831920454e-05, "loss": 0.0519, "step": 52330 }, { "grad_norm": 0.42128226161003113, "learning_rate": 5.026620738295351e-05, "loss": 0.0457, "step": 52340 }, { "grad_norm": 0.4418591260910034, "learning_rate": 5.024967290487455e-05, "loss": 0.0332, "step": 52350 }, { "grad_norm": 0.8788148164749146, "learning_rate": 5.023313839949172e-05, "loss": 0.0424, "step": 52360 }, { "grad_norm": 0.4395011067390442, "learning_rate": 5.0216603868613224e-05, "loss": 0.0414, "step": 52370 }, { "grad_norm": 0.312959223985672, "learning_rate": 5.020006931404726e-05, "loss": 0.046, "step": 52380 }, { "grad_norm": 0.40759673714637756, "learning_rate": 5.0183534737602014e-05, "loss": 0.0258, "step": 52390 }, { "grad_norm": 0.3107728362083435, "learning_rate": 5.0167000141085686e-05, "loss": 0.034, "step": 52400 }, { "grad_norm": 0.24014900624752045, "learning_rate": 5.015046552630647e-05, "loss": 0.0326, "step": 52410 }, { "grad_norm": 0.41964221000671387, "learning_rate": 5.013393089507258e-05, "loss": 0.0334, "step": 52420 }, { "grad_norm": 0.4203827679157257, "learning_rate": 5.011739624919219e-05, "loss": 0.0418, "step": 52430 }, { "grad_norm": 0.780480146408081, "learning_rate": 5.0100861590473545e-05, "loss": 0.0479, "step": 52440 }, { "grad_norm": 0.4442591369152069, "learning_rate": 5.008432692072482e-05, "loss": 0.0343, "step": 52450 }, { "grad_norm": 0.5844644904136658, "learning_rate": 5.006779224175423e-05, "loss": 0.0561, "step": 52460 }, { "grad_norm": 0.41198956966400146, "learning_rate": 5.005125755536998e-05, "loss": 0.0384, "step": 52470 }, { "grad_norm": 0.4518933594226837, "learning_rate": 5.003472286338029e-05, "loss": 0.0403, "step": 52480 }, { "grad_norm": 0.3322794437408447, "learning_rate": 5.001818816759335e-05, "loss": 0.0547, "step": 52490 }, { "grad_norm": 0.43045204877853394, "learning_rate": 5.000165346981738e-05, "loss": 0.0368, "step": 52500 }, { "grad_norm": 0.3569653332233429, "learning_rate": 4.9985118771860595e-05, "loss": 0.0423, "step": 52510 }, { "grad_norm": 0.41880059242248535, "learning_rate": 4.996858407553119e-05, "loss": 0.0344, "step": 52520 }, { "grad_norm": 0.45382243394851685, "learning_rate": 4.99520493826374e-05, "loss": 0.0349, "step": 52530 }, { "grad_norm": 0.4331464171409607, "learning_rate": 4.9935514694987384e-05, "loss": 0.0405, "step": 52540 }, { "grad_norm": 0.7274436354637146, "learning_rate": 4.991898001438941e-05, "loss": 0.0448, "step": 52550 }, { "grad_norm": 0.5987768769264221, "learning_rate": 4.990244534265166e-05, "loss": 0.0423, "step": 52560 }, { "grad_norm": 0.7569208145141602, "learning_rate": 4.988591068158234e-05, "loss": 0.0398, "step": 52570 }, { "grad_norm": 0.4273989200592041, "learning_rate": 4.986937603298965e-05, "loss": 0.0339, "step": 52580 }, { "grad_norm": 0.5354576110839844, "learning_rate": 4.985284139868181e-05, "loss": 0.0384, "step": 52590 }, { "grad_norm": 0.4882069528102875, "learning_rate": 4.9836306780467e-05, "loss": 0.047, "step": 52600 }, { "grad_norm": 0.4821690320968628, "learning_rate": 4.981977218015343e-05, "loss": 0.037, "step": 52610 }, { "grad_norm": 0.5286755561828613, "learning_rate": 4.980323759954932e-05, "loss": 0.047, "step": 52620 }, { "grad_norm": 0.5465155839920044, "learning_rate": 4.978670304046284e-05, "loss": 0.0433, "step": 52630 }, { "grad_norm": 0.36179426312446594, "learning_rate": 4.9770168504702195e-05, "loss": 0.0472, "step": 52640 }, { "grad_norm": 0.46982645988464355, "learning_rate": 4.975363399407559e-05, "loss": 0.0435, "step": 52650 }, { "grad_norm": 0.5248092412948608, "learning_rate": 4.9737099510391194e-05, "loss": 0.0502, "step": 52660 }, { "grad_norm": 0.5850440263748169, "learning_rate": 4.972056505545719e-05, "loss": 0.0504, "step": 52670 }, { "grad_norm": 0.35130417346954346, "learning_rate": 4.9704030631081775e-05, "loss": 0.037, "step": 52680 }, { "grad_norm": 0.2685585021972656, "learning_rate": 4.968749623907313e-05, "loss": 0.0352, "step": 52690 }, { "grad_norm": 0.6074530482292175, "learning_rate": 4.967096188123943e-05, "loss": 0.043, "step": 52700 }, { "grad_norm": 0.30859634280204773, "learning_rate": 4.965442755938884e-05, "loss": 0.0298, "step": 52710 }, { "grad_norm": 0.5633233785629272, "learning_rate": 4.963789327532954e-05, "loss": 0.0371, "step": 52720 }, { "grad_norm": 0.475201815366745, "learning_rate": 4.962135903086968e-05, "loss": 0.0429, "step": 52730 }, { "grad_norm": 0.3192192018032074, "learning_rate": 4.960482482781741e-05, "loss": 0.0327, "step": 52740 }, { "grad_norm": 0.460071325302124, "learning_rate": 4.958829066798092e-05, "loss": 0.0353, "step": 52750 }, { "grad_norm": 0.5311809778213501, "learning_rate": 4.957175655316835e-05, "loss": 0.0403, "step": 52760 }, { "grad_norm": 0.5084406733512878, "learning_rate": 4.9555222485187834e-05, "loss": 0.046, "step": 52770 }, { "grad_norm": 0.45025527477264404, "learning_rate": 4.953868846584752e-05, "loss": 0.0368, "step": 52780 }, { "grad_norm": 0.6097113490104675, "learning_rate": 4.9522154496955544e-05, "loss": 0.049, "step": 52790 }, { "grad_norm": 0.3613065481185913, "learning_rate": 4.950562058032001e-05, "loss": 0.046, "step": 52800 }, { "grad_norm": 0.4580700993537903, "learning_rate": 4.948908671774909e-05, "loss": 0.047, "step": 52810 }, { "grad_norm": 0.61191326379776, "learning_rate": 4.947255291105087e-05, "loss": 0.045, "step": 52820 }, { "grad_norm": 0.7391625642776489, "learning_rate": 4.945601916203348e-05, "loss": 0.0396, "step": 52830 }, { "grad_norm": 0.662566602230072, "learning_rate": 4.9439485472505016e-05, "loss": 0.0384, "step": 52840 }, { "grad_norm": 0.34021809697151184, "learning_rate": 4.942295184427356e-05, "loss": 0.0442, "step": 52850 }, { "grad_norm": 0.37031450867652893, "learning_rate": 4.9406418279147234e-05, "loss": 0.0462, "step": 52860 }, { "grad_norm": 0.4745168387889862, "learning_rate": 4.9389884778934116e-05, "loss": 0.0398, "step": 52870 }, { "grad_norm": 0.6148914694786072, "learning_rate": 4.937335134544228e-05, "loss": 0.0404, "step": 52880 }, { "grad_norm": 0.3712761402130127, "learning_rate": 4.935681798047979e-05, "loss": 0.0579, "step": 52890 }, { "grad_norm": 0.4097561240196228, "learning_rate": 4.934028468585473e-05, "loss": 0.0334, "step": 52900 }, { "grad_norm": 1.0056778192520142, "learning_rate": 4.932375146337514e-05, "loss": 0.0411, "step": 52910 }, { "grad_norm": 0.8209702372550964, "learning_rate": 4.930721831484908e-05, "loss": 0.0464, "step": 52920 }, { "grad_norm": 0.6264739632606506, "learning_rate": 4.929068524208456e-05, "loss": 0.0496, "step": 52930 }, { "grad_norm": 0.5330182909965515, "learning_rate": 4.927415224688965e-05, "loss": 0.0495, "step": 52940 }, { "grad_norm": 0.381322979927063, "learning_rate": 4.9257619331072375e-05, "loss": 0.0429, "step": 52950 }, { "grad_norm": 0.5839164853096008, "learning_rate": 4.924108649644072e-05, "loss": 0.0488, "step": 52960 }, { "grad_norm": 0.3276512920856476, "learning_rate": 4.922455374480271e-05, "loss": 0.0502, "step": 52970 }, { "grad_norm": 0.4570222496986389, "learning_rate": 4.920802107796634e-05, "loss": 0.0373, "step": 52980 }, { "grad_norm": 0.5425445437431335, "learning_rate": 4.9191488497739564e-05, "loss": 0.0415, "step": 52990 }, { "grad_norm": 0.426982581615448, "learning_rate": 4.917495600593042e-05, "loss": 0.0428, "step": 53000 }, { "grad_norm": 0.4536266028881073, "learning_rate": 4.915842360434684e-05, "loss": 0.0399, "step": 53010 }, { "grad_norm": 0.7726253271102905, "learning_rate": 4.914189129479679e-05, "loss": 0.0523, "step": 53020 }, { "grad_norm": 0.39773938059806824, "learning_rate": 4.912535907908822e-05, "loss": 0.0465, "step": 53030 }, { "grad_norm": 0.7171463370323181, "learning_rate": 4.910882695902907e-05, "loss": 0.0472, "step": 53040 }, { "grad_norm": 0.6873939633369446, "learning_rate": 4.9092294936427234e-05, "loss": 0.0395, "step": 53050 }, { "grad_norm": 0.42970192432403564, "learning_rate": 4.9075763013090685e-05, "loss": 0.0368, "step": 53060 }, { "grad_norm": 0.4191995859146118, "learning_rate": 4.9059231190827305e-05, "loss": 0.0388, "step": 53070 }, { "grad_norm": 0.35670915246009827, "learning_rate": 4.904269947144499e-05, "loss": 0.0479, "step": 53080 }, { "grad_norm": 0.42809155583381653, "learning_rate": 4.902616785675161e-05, "loss": 0.0422, "step": 53090 }, { "grad_norm": 0.6077048182487488, "learning_rate": 4.900963634855505e-05, "loss": 0.0427, "step": 53100 }, { "grad_norm": 0.4642721116542816, "learning_rate": 4.8993104948663173e-05, "loss": 0.0335, "step": 53110 }, { "grad_norm": 0.3734625577926636, "learning_rate": 4.8976573658883823e-05, "loss": 0.0318, "step": 53120 }, { "grad_norm": 0.3729228377342224, "learning_rate": 4.896004248102484e-05, "loss": 0.0393, "step": 53130 }, { "grad_norm": 0.3968259394168854, "learning_rate": 4.8943511416894045e-05, "loss": 0.0448, "step": 53140 }, { "grad_norm": 0.5143992900848389, "learning_rate": 4.892698046829926e-05, "loss": 0.038, "step": 53150 }, { "grad_norm": 0.646393895149231, "learning_rate": 4.8910449637048275e-05, "loss": 0.0579, "step": 53160 }, { "grad_norm": 0.5419618487358093, "learning_rate": 4.8893918924948884e-05, "loss": 0.0423, "step": 53170 }, { "grad_norm": 0.39903080463409424, "learning_rate": 4.8877388333808835e-05, "loss": 0.0432, "step": 53180 }, { "grad_norm": 0.3458136320114136, "learning_rate": 4.8860857865435925e-05, "loss": 0.0414, "step": 53190 }, { "grad_norm": 0.5246516466140747, "learning_rate": 4.884432752163789e-05, "loss": 0.0353, "step": 53200 }, { "grad_norm": 0.7391046285629272, "learning_rate": 4.8827797304222466e-05, "loss": 0.0452, "step": 53210 }, { "grad_norm": 0.7240548133850098, "learning_rate": 4.8811267214997366e-05, "loss": 0.0373, "step": 53220 }, { "grad_norm": 0.4452851116657257, "learning_rate": 4.879473725577029e-05, "loss": 0.0354, "step": 53230 }, { "grad_norm": 0.8443415760993958, "learning_rate": 4.877820742834891e-05, "loss": 0.0421, "step": 53240 }, { "grad_norm": 0.4896622896194458, "learning_rate": 4.8761677734540956e-05, "loss": 0.0384, "step": 53250 }, { "grad_norm": 0.6309734582901001, "learning_rate": 4.874514817615407e-05, "loss": 0.0445, "step": 53260 }, { "grad_norm": 0.31231924891471863, "learning_rate": 4.872861875499588e-05, "loss": 0.0375, "step": 53270 }, { "grad_norm": 0.5580090880393982, "learning_rate": 4.871208947287404e-05, "loss": 0.0404, "step": 53280 }, { "grad_norm": 0.4800029695034027, "learning_rate": 4.869556033159615e-05, "loss": 0.0378, "step": 53290 }, { "grad_norm": 0.2793809175491333, "learning_rate": 4.8679031332969835e-05, "loss": 0.0399, "step": 53300 }, { "grad_norm": 0.3058148920536041, "learning_rate": 4.8662502478802635e-05, "loss": 0.0496, "step": 53310 }, { "grad_norm": 0.4028564393520355, "learning_rate": 4.8645973770902176e-05, "loss": 0.0398, "step": 53320 }, { "grad_norm": 0.4591578543186188, "learning_rate": 4.862944521107599e-05, "loss": 0.0305, "step": 53330 }, { "grad_norm": 0.600223958492279, "learning_rate": 4.861291680113162e-05, "loss": 0.0379, "step": 53340 }, { "grad_norm": 0.4021143913269043, "learning_rate": 4.8596388542876563e-05, "loss": 0.0371, "step": 53350 }, { "grad_norm": 1.071933388710022, "learning_rate": 4.857986043811836e-05, "loss": 0.0421, "step": 53360 }, { "grad_norm": 0.5396203994750977, "learning_rate": 4.856333248866448e-05, "loss": 0.0364, "step": 53370 }, { "grad_norm": 0.43080025911331177, "learning_rate": 4.85468046963224e-05, "loss": 0.0365, "step": 53380 }, { "grad_norm": 0.47643762826919556, "learning_rate": 4.853027706289957e-05, "loss": 0.0373, "step": 53390 }, { "grad_norm": 0.5027225017547607, "learning_rate": 4.851374959020344e-05, "loss": 0.0479, "step": 53400 }, { "grad_norm": 0.6332242488861084, "learning_rate": 4.849722228004141e-05, "loss": 0.0402, "step": 53410 }, { "grad_norm": 0.4999402165412903, "learning_rate": 4.848069513422091e-05, "loss": 0.0448, "step": 53420 }, { "grad_norm": 0.4729500412940979, "learning_rate": 4.8464168154549294e-05, "loss": 0.039, "step": 53430 }, { "grad_norm": 0.44764524698257446, "learning_rate": 4.844764134283392e-05, "loss": 0.0387, "step": 53440 }, { "grad_norm": 0.44143348932266235, "learning_rate": 4.843111470088218e-05, "loss": 0.0515, "step": 53450 }, { "grad_norm": 0.4279416501522064, "learning_rate": 4.841458823050137e-05, "loss": 0.0416, "step": 53460 }, { "grad_norm": 1.204991340637207, "learning_rate": 4.8398061933498816e-05, "loss": 0.0417, "step": 53470 }, { "grad_norm": 0.8304101824760437, "learning_rate": 4.83815358116818e-05, "loss": 0.0415, "step": 53480 }, { "grad_norm": 0.3941468298435211, "learning_rate": 4.8365009866857584e-05, "loss": 0.0433, "step": 53490 }, { "grad_norm": 0.32472750544548035, "learning_rate": 4.834848410083342e-05, "loss": 0.0396, "step": 53500 }, { "grad_norm": 0.45464348793029785, "learning_rate": 4.833195851541657e-05, "loss": 0.0388, "step": 53510 }, { "grad_norm": 0.3974180519580841, "learning_rate": 4.831543311241423e-05, "loss": 0.0385, "step": 53520 }, { "grad_norm": 0.8973613381385803, "learning_rate": 4.829890789363359e-05, "loss": 0.0503, "step": 53530 }, { "grad_norm": 0.43125492334365845, "learning_rate": 4.828238286088182e-05, "loss": 0.0351, "step": 53540 }, { "grad_norm": 0.413534551858902, "learning_rate": 4.826585801596609e-05, "loss": 0.0413, "step": 53550 }, { "grad_norm": 0.48475465178489685, "learning_rate": 4.8249333360693486e-05, "loss": 0.0393, "step": 53560 }, { "grad_norm": 0.47073283791542053, "learning_rate": 4.8232808896871185e-05, "loss": 0.0382, "step": 53570 }, { "grad_norm": 0.44847849011421204, "learning_rate": 4.821628462630624e-05, "loss": 0.0421, "step": 53580 }, { "grad_norm": 0.4063057601451874, "learning_rate": 4.8199760550805735e-05, "loss": 0.0379, "step": 53590 }, { "grad_norm": 0.5304486751556396, "learning_rate": 4.818323667217669e-05, "loss": 0.0415, "step": 53600 }, { "grad_norm": 0.44818389415740967, "learning_rate": 4.8166712992226165e-05, "loss": 0.0419, "step": 53610 }, { "grad_norm": 0.45094653964042664, "learning_rate": 4.8150189512761154e-05, "loss": 0.0504, "step": 53620 }, { "grad_norm": 0.32494476437568665, "learning_rate": 4.813366623558862e-05, "loss": 0.0372, "step": 53630 }, { "grad_norm": 0.5880461931228638, "learning_rate": 4.811714316251554e-05, "loss": 0.0425, "step": 53640 }, { "grad_norm": 0.3413636386394501, "learning_rate": 4.810062029534886e-05, "loss": 0.0347, "step": 53650 }, { "grad_norm": 0.29811739921569824, "learning_rate": 4.8084097635895505e-05, "loss": 0.0361, "step": 53660 }, { "grad_norm": 0.5258942246437073, "learning_rate": 4.8067575185962335e-05, "loss": 0.0448, "step": 53670 }, { "grad_norm": 0.5949018597602844, "learning_rate": 4.805105294735625e-05, "loss": 0.0344, "step": 53680 }, { "grad_norm": 0.7465265989303589, "learning_rate": 4.803453092188406e-05, "loss": 0.0443, "step": 53690 }, { "grad_norm": 0.7793271541595459, "learning_rate": 4.801800911135263e-05, "loss": 0.0416, "step": 53700 }, { "grad_norm": 0.364011287689209, "learning_rate": 4.8001487517568744e-05, "loss": 0.0436, "step": 53710 }, { "grad_norm": 0.5138776898384094, "learning_rate": 4.798496614233918e-05, "loss": 0.0461, "step": 53720 }, { "grad_norm": 0.3866844177246094, "learning_rate": 4.796844498747069e-05, "loss": 0.0548, "step": 53730 }, { "grad_norm": 0.40969687700271606, "learning_rate": 4.795192405477e-05, "loss": 0.0332, "step": 53740 }, { "grad_norm": 0.5259324312210083, "learning_rate": 4.79354033460438e-05, "loss": 0.0437, "step": 53750 }, { "grad_norm": 0.8096996545791626, "learning_rate": 4.79188828630988e-05, "loss": 0.05, "step": 53760 }, { "grad_norm": 0.5140699744224548, "learning_rate": 4.790236260774165e-05, "loss": 0.0458, "step": 53770 }, { "grad_norm": 0.5935071110725403, "learning_rate": 4.788584258177896e-05, "loss": 0.043, "step": 53780 }, { "grad_norm": 0.3463380038738251, "learning_rate": 4.7869322787017356e-05, "loss": 0.0385, "step": 53790 }, { "grad_norm": 0.3955892324447632, "learning_rate": 4.78528032252634e-05, "loss": 0.04, "step": 53800 }, { "grad_norm": 0.4110639691352844, "learning_rate": 4.783628389832366e-05, "loss": 0.0474, "step": 53810 }, { "grad_norm": 0.5701038241386414, "learning_rate": 4.7819764808004644e-05, "loss": 0.0495, "step": 53820 }, { "grad_norm": 0.5534960031509399, "learning_rate": 4.780324595611289e-05, "loss": 0.0342, "step": 53830 }, { "grad_norm": 0.6101047396659851, "learning_rate": 4.7786727344454854e-05, "loss": 0.0477, "step": 53840 }, { "grad_norm": 0.35599076747894287, "learning_rate": 4.777020897483697e-05, "loss": 0.0432, "step": 53850 }, { "grad_norm": 0.5063543915748596, "learning_rate": 4.775369084906569e-05, "loss": 0.035, "step": 53860 }, { "grad_norm": 0.5165577530860901, "learning_rate": 4.773717296894742e-05, "loss": 0.0553, "step": 53870 }, { "grad_norm": 0.5052487254142761, "learning_rate": 4.772065533628848e-05, "loss": 0.0478, "step": 53880 }, { "grad_norm": 0.5959643125534058, "learning_rate": 4.7704137952895265e-05, "loss": 0.0355, "step": 53890 }, { "grad_norm": 0.4434382915496826, "learning_rate": 4.768762082057407e-05, "loss": 0.0283, "step": 53900 }, { "grad_norm": 0.5476958155632019, "learning_rate": 4.7671103941131186e-05, "loss": 0.0401, "step": 53910 }, { "grad_norm": 0.36420756578445435, "learning_rate": 4.765458731637287e-05, "loss": 0.0355, "step": 53920 }, { "grad_norm": 0.39876583218574524, "learning_rate": 4.763807094810537e-05, "loss": 0.0441, "step": 53930 }, { "grad_norm": 0.6084120869636536, "learning_rate": 4.762155483813487e-05, "loss": 0.0473, "step": 53940 }, { "grad_norm": 0.6794766783714294, "learning_rate": 4.760503898826754e-05, "loss": 0.0447, "step": 53950 }, { "grad_norm": 0.35499364137649536, "learning_rate": 4.758852340030957e-05, "loss": 0.0574, "step": 53960 }, { "grad_norm": 0.42752084136009216, "learning_rate": 4.757200807606705e-05, "loss": 0.0389, "step": 53970 }, { "grad_norm": 0.4598905146121979, "learning_rate": 4.755549301734607e-05, "loss": 0.0364, "step": 53980 }, { "grad_norm": 0.4049941599369049, "learning_rate": 4.753897822595271e-05, "loss": 0.0525, "step": 53990 }, { "grad_norm": 1.0604400634765625, "learning_rate": 4.7522463703692994e-05, "loss": 0.0362, "step": 54000 }, { "grad_norm": 0.39782196283340454, "learning_rate": 4.75059494523729e-05, "loss": 0.0382, "step": 54010 }, { "grad_norm": 0.6020158529281616, "learning_rate": 4.748943547379843e-05, "loss": 0.0401, "step": 54020 }, { "grad_norm": 0.3105575740337372, "learning_rate": 4.7472921769775535e-05, "loss": 0.0521, "step": 54030 }, { "grad_norm": 0.530462920665741, "learning_rate": 4.745640834211011e-05, "loss": 0.0482, "step": 54040 }, { "grad_norm": 0.30717623233795166, "learning_rate": 4.7439895192608054e-05, "loss": 0.0381, "step": 54050 }, { "grad_norm": 0.4923345744609833, "learning_rate": 4.7423382323075186e-05, "loss": 0.0375, "step": 54060 }, { "grad_norm": 0.4866253137588501, "learning_rate": 4.740686973531737e-05, "loss": 0.0332, "step": 54070 }, { "grad_norm": 0.5480494499206543, "learning_rate": 4.7390357431140375e-05, "loss": 0.0433, "step": 54080 }, { "grad_norm": 3.4518139362335205, "learning_rate": 4.737384541234998e-05, "loss": 0.0489, "step": 54090 }, { "grad_norm": 0.39775702357292175, "learning_rate": 4.735733368075189e-05, "loss": 0.035, "step": 54100 }, { "grad_norm": 0.525385320186615, "learning_rate": 4.734082223815182e-05, "loss": 0.0429, "step": 54110 }, { "grad_norm": 0.312761127948761, "learning_rate": 4.7324311086355436e-05, "loss": 0.0278, "step": 54120 }, { "grad_norm": 0.8536919951438904, "learning_rate": 4.730780022716836e-05, "loss": 0.0422, "step": 54130 }, { "grad_norm": 0.37393006682395935, "learning_rate": 4.729128966239619e-05, "loss": 0.0327, "step": 54140 }, { "grad_norm": 0.5563544034957886, "learning_rate": 4.7274779393844536e-05, "loss": 0.0356, "step": 54150 }, { "grad_norm": 1.0200135707855225, "learning_rate": 4.725826942331891e-05, "loss": 0.055, "step": 54160 }, { "grad_norm": 0.49330461025238037, "learning_rate": 4.724175975262481e-05, "loss": 0.0396, "step": 54170 }, { "grad_norm": 0.5527102947235107, "learning_rate": 4.722525038356772e-05, "loss": 0.0431, "step": 54180 }, { "grad_norm": 0.41047805547714233, "learning_rate": 4.7208741317953075e-05, "loss": 0.0412, "step": 54190 }, { "grad_norm": 0.3040512502193451, "learning_rate": 4.7192232557586264e-05, "loss": 0.0412, "step": 54200 }, { "grad_norm": 0.6352671980857849, "learning_rate": 4.717572410427269e-05, "loss": 0.036, "step": 54210 }, { "grad_norm": 0.46769142150878906, "learning_rate": 4.71592159598177e-05, "loss": 0.0475, "step": 54220 }, { "grad_norm": 0.5553522706031799, "learning_rate": 4.714270812602657e-05, "loss": 0.0403, "step": 54230 }, { "grad_norm": 0.37703824043273926, "learning_rate": 4.712620060470458e-05, "loss": 0.0543, "step": 54240 }, { "grad_norm": 0.5202730298042297, "learning_rate": 4.7109693397656986e-05, "loss": 0.0434, "step": 54250 }, { "grad_norm": 0.4092676341533661, "learning_rate": 4.709318650668894e-05, "loss": 0.035, "step": 54260 }, { "grad_norm": 0.5785564184188843, "learning_rate": 4.7076679933605676e-05, "loss": 0.0416, "step": 54270 }, { "grad_norm": 0.5404103398323059, "learning_rate": 4.70601736802123e-05, "loss": 0.0378, "step": 54280 }, { "grad_norm": 0.39490097761154175, "learning_rate": 4.704366774831391e-05, "loss": 0.0523, "step": 54290 }, { "grad_norm": 0.552700400352478, "learning_rate": 4.702716213971557e-05, "loss": 0.0386, "step": 54300 }, { "grad_norm": 0.863682210445404, "learning_rate": 4.70106568562223e-05, "loss": 0.0507, "step": 54310 }, { "grad_norm": 0.42526349425315857, "learning_rate": 4.6994151899639105e-05, "loss": 0.0386, "step": 54320 }, { "grad_norm": 0.37401726841926575, "learning_rate": 4.697764727177093e-05, "loss": 0.051, "step": 54330 }, { "grad_norm": 0.4401160180568695, "learning_rate": 4.696114297442272e-05, "loss": 0.0419, "step": 54340 }, { "grad_norm": 0.37251609563827515, "learning_rate": 4.694463900939933e-05, "loss": 0.0359, "step": 54350 }, { "grad_norm": 0.36614659428596497, "learning_rate": 4.692813537850564e-05, "loss": 0.0414, "step": 54360 }, { "grad_norm": 0.4442945420742035, "learning_rate": 4.6911632083546445e-05, "loss": 0.0364, "step": 54370 }, { "grad_norm": 0.3653682768344879, "learning_rate": 4.6895129126326525e-05, "loss": 0.0442, "step": 54380 }, { "grad_norm": 0.5831429362297058, "learning_rate": 4.687862650865059e-05, "loss": 0.0412, "step": 54390 }, { "grad_norm": 0.4134852886199951, "learning_rate": 4.686212423232339e-05, "loss": 0.0325, "step": 54400 }, { "grad_norm": 0.3598484396934509, "learning_rate": 4.6845622299149565e-05, "loss": 0.0395, "step": 54410 }, { "grad_norm": 0.438735693693161, "learning_rate": 4.682912071093374e-05, "loss": 0.0387, "step": 54420 }, { "grad_norm": 0.34525856375694275, "learning_rate": 4.681261946948052e-05, "loss": 0.0401, "step": 54430 }, { "grad_norm": 0.600696861743927, "learning_rate": 4.679611857659443e-05, "loss": 0.0466, "step": 54440 }, { "grad_norm": 0.36385321617126465, "learning_rate": 4.6779618034079975e-05, "loss": 0.0422, "step": 54450 }, { "grad_norm": 0.3413242697715759, "learning_rate": 4.6763117843741675e-05, "loss": 0.0412, "step": 54460 }, { "grad_norm": 0.5346511006355286, "learning_rate": 4.674661800738394e-05, "loss": 0.0374, "step": 54470 }, { "grad_norm": 0.48113852739334106, "learning_rate": 4.673011852681117e-05, "loss": 0.0397, "step": 54480 }, { "grad_norm": 0.3974219560623169, "learning_rate": 4.671361940382772e-05, "loss": 0.0424, "step": 54490 }, { "grad_norm": 0.5726926326751709, "learning_rate": 4.669712064023791e-05, "loss": 0.0464, "step": 54500 }, { "grad_norm": 0.4510670304298401, "learning_rate": 4.668062223784602e-05, "loss": 0.0374, "step": 54510 }, { "grad_norm": 0.29202142357826233, "learning_rate": 4.6664124198456274e-05, "loss": 0.0353, "step": 54520 }, { "grad_norm": 0.4175335764884949, "learning_rate": 4.6647626523872905e-05, "loss": 0.0389, "step": 54530 }, { "grad_norm": 0.46546271443367004, "learning_rate": 4.6631129215900066e-05, "loss": 0.0344, "step": 54540 }, { "grad_norm": 0.552858829498291, "learning_rate": 4.661463227634187e-05, "loss": 0.0552, "step": 54550 }, { "grad_norm": 0.41148659586906433, "learning_rate": 4.659813570700238e-05, "loss": 0.0306, "step": 54560 }, { "grad_norm": 0.46565577387809753, "learning_rate": 4.658163950968568e-05, "loss": 0.0356, "step": 54570 }, { "grad_norm": 0.6219919919967651, "learning_rate": 4.6565143686195726e-05, "loss": 0.0446, "step": 54580 }, { "grad_norm": 0.5061758160591125, "learning_rate": 4.654864823833649e-05, "loss": 0.0355, "step": 54590 }, { "grad_norm": 0.6548724174499512, "learning_rate": 4.65321531679119e-05, "loss": 0.0403, "step": 54600 }, { "grad_norm": 0.3500862419605255, "learning_rate": 4.6515658476725834e-05, "loss": 0.0453, "step": 54610 }, { "grad_norm": 0.391737163066864, "learning_rate": 4.649916416658212e-05, "loss": 0.0432, "step": 54620 }, { "grad_norm": 0.7054447531700134, "learning_rate": 4.648267023928455e-05, "loss": 0.0382, "step": 54630 }, { "grad_norm": 0.5376734733581543, "learning_rate": 4.646617669663687e-05, "loss": 0.0391, "step": 54640 }, { "grad_norm": 0.49304479360580444, "learning_rate": 4.644968354044277e-05, "loss": 0.0392, "step": 54650 }, { "grad_norm": 0.5070409178733826, "learning_rate": 4.643319077250596e-05, "loss": 0.0358, "step": 54660 }, { "grad_norm": 1.073015809059143, "learning_rate": 4.6416698394630055e-05, "loss": 0.0391, "step": 54670 }, { "grad_norm": 1.1194463968276978, "learning_rate": 4.640020640861862e-05, "loss": 0.04, "step": 54680 }, { "grad_norm": 0.39669349789619446, "learning_rate": 4.6383714816275195e-05, "loss": 0.0477, "step": 54690 }, { "grad_norm": 0.41753795742988586, "learning_rate": 4.636722361940329e-05, "loss": 0.0366, "step": 54700 }, { "grad_norm": 0.37265267968177795, "learning_rate": 4.635073281980632e-05, "loss": 0.0417, "step": 54710 }, { "grad_norm": 0.39683911204338074, "learning_rate": 4.633424241928775e-05, "loss": 0.0634, "step": 54720 }, { "grad_norm": 0.40240299701690674, "learning_rate": 4.6317752419650914e-05, "loss": 0.0345, "step": 54730 }, { "grad_norm": 0.6035622358322144, "learning_rate": 4.630126282269914e-05, "loss": 0.0507, "step": 54740 }, { "grad_norm": 0.39628395438194275, "learning_rate": 4.628477363023569e-05, "loss": 0.0374, "step": 54750 }, { "grad_norm": 0.43448588252067566, "learning_rate": 4.626828484406383e-05, "loss": 0.0404, "step": 54760 }, { "grad_norm": 0.6232145428657532, "learning_rate": 4.625179646598669e-05, "loss": 0.0445, "step": 54770 }, { "grad_norm": 0.23288525640964508, "learning_rate": 4.6235308497807487e-05, "loss": 0.0375, "step": 54780 }, { "grad_norm": 0.45788365602493286, "learning_rate": 4.6218820941329276e-05, "loss": 0.0375, "step": 54790 }, { "grad_norm": 0.4110517203807831, "learning_rate": 4.620233379835513e-05, "loss": 0.0336, "step": 54800 }, { "grad_norm": 0.6775460243225098, "learning_rate": 4.618584707068804e-05, "loss": 0.0389, "step": 54810 }, { "grad_norm": 0.5657482743263245, "learning_rate": 4.616936076013099e-05, "loss": 0.0446, "step": 54820 }, { "grad_norm": 0.5904064774513245, "learning_rate": 4.615287486848688e-05, "loss": 0.0401, "step": 54830 }, { "grad_norm": 1.0825766324996948, "learning_rate": 4.6136389397558584e-05, "loss": 0.0518, "step": 54840 }, { "grad_norm": 0.4392966330051422, "learning_rate": 4.6119904349148935e-05, "loss": 0.0294, "step": 54850 }, { "grad_norm": 0.6089013814926147, "learning_rate": 4.6103419725060726e-05, "loss": 0.0364, "step": 54860 }, { "grad_norm": 0.49716299772262573, "learning_rate": 4.608693552709667e-05, "loss": 0.0293, "step": 54870 }, { "grad_norm": 0.5270337462425232, "learning_rate": 4.607045175705947e-05, "loss": 0.0424, "step": 54880 }, { "grad_norm": 0.45417505502700806, "learning_rate": 4.605396841675176e-05, "loss": 0.0363, "step": 54890 }, { "grad_norm": 0.5396014451980591, "learning_rate": 4.6037485507976094e-05, "loss": 0.0397, "step": 54900 }, { "grad_norm": 0.4384508430957794, "learning_rate": 4.602100303253509e-05, "loss": 0.0372, "step": 54910 }, { "grad_norm": 0.8829135298728943, "learning_rate": 4.600452099223122e-05, "loss": 0.0506, "step": 54920 }, { "grad_norm": 0.4236641824245453, "learning_rate": 4.598803938886691e-05, "loss": 0.0426, "step": 54930 }, { "grad_norm": 0.3358765244483948, "learning_rate": 4.597155822424459e-05, "loss": 0.0399, "step": 54940 }, { "grad_norm": 1.249976634979248, "learning_rate": 4.59550775001666e-05, "loss": 0.0543, "step": 54950 }, { "grad_norm": 0.47962623834609985, "learning_rate": 4.5938597218435234e-05, "loss": 0.0415, "step": 54960 }, { "grad_norm": 0.4188843369483948, "learning_rate": 4.592211738085279e-05, "loss": 0.0588, "step": 54970 }, { "grad_norm": 0.4106219410896301, "learning_rate": 4.590563798922146e-05, "loss": 0.0392, "step": 54980 }, { "grad_norm": 0.42804840207099915, "learning_rate": 4.5889159045343404e-05, "loss": 0.0518, "step": 54990 }, { "grad_norm": 0.40267816185951233, "learning_rate": 4.5872680551020734e-05, "loss": 0.0411, "step": 55000 }, { "grad_norm": 0.42795249819755554, "learning_rate": 4.585620250805552e-05, "loss": 0.0408, "step": 55010 }, { "grad_norm": 0.47461000084877014, "learning_rate": 4.583972491824975e-05, "loss": 0.0324, "step": 55020 }, { "grad_norm": 0.5027675628662109, "learning_rate": 4.582324778340541e-05, "loss": 0.0459, "step": 55030 }, { "grad_norm": 0.42608895897865295, "learning_rate": 4.5806771105324424e-05, "loss": 0.0454, "step": 55040 }, { "grad_norm": 0.7429755330085754, "learning_rate": 4.579029488580865e-05, "loss": 0.0436, "step": 55050 }, { "grad_norm": 0.4539829194545746, "learning_rate": 4.5773819126659876e-05, "loss": 0.0287, "step": 55060 }, { "grad_norm": 0.38698610663414, "learning_rate": 4.575734382967991e-05, "loss": 0.0365, "step": 55070 }, { "grad_norm": 0.4304763376712799, "learning_rate": 4.574086899667044e-05, "loss": 0.0368, "step": 55080 }, { "grad_norm": 0.4465525150299072, "learning_rate": 4.5724394629433106e-05, "loss": 0.0406, "step": 55090 }, { "grad_norm": 0.32261043787002563, "learning_rate": 4.570792072976958e-05, "loss": 0.0374, "step": 55100 }, { "grad_norm": 0.4038515090942383, "learning_rate": 4.569144729948138e-05, "loss": 0.0417, "step": 55110 }, { "grad_norm": 0.4408808946609497, "learning_rate": 4.5674974340370033e-05, "loss": 0.0315, "step": 55120 }, { "grad_norm": 0.4278935492038727, "learning_rate": 4.565850185423699e-05, "loss": 0.0408, "step": 55130 }, { "grad_norm": 0.7052352428436279, "learning_rate": 4.564202984288365e-05, "loss": 0.0555, "step": 55140 }, { "grad_norm": 0.4322502613067627, "learning_rate": 4.5625558308111356e-05, "loss": 0.0356, "step": 55150 }, { "grad_norm": 0.47361794114112854, "learning_rate": 4.560908725172145e-05, "loss": 0.0492, "step": 55160 }, { "grad_norm": 0.38810455799102783, "learning_rate": 4.559261667551516e-05, "loss": 0.0386, "step": 55170 }, { "grad_norm": 0.45031076669692993, "learning_rate": 4.5576146581293685e-05, "loss": 0.0424, "step": 55180 }, { "grad_norm": 0.6530593037605286, "learning_rate": 4.5559676970858164e-05, "loss": 0.0399, "step": 55190 }, { "grad_norm": 0.6682553291320801, "learning_rate": 4.5543207846009705e-05, "loss": 0.0387, "step": 55200 }, { "grad_norm": 0.4171220064163208, "learning_rate": 4.552673920854933e-05, "loss": 0.0465, "step": 55210 }, { "grad_norm": 0.47292381525039673, "learning_rate": 4.5510271060278006e-05, "loss": 0.0387, "step": 55220 }, { "grad_norm": 0.39031991362571716, "learning_rate": 4.5493803402996704e-05, "loss": 0.0428, "step": 55230 }, { "grad_norm": 0.34694626927375793, "learning_rate": 4.5477336238506295e-05, "loss": 0.0378, "step": 55240 }, { "grad_norm": 0.34974661469459534, "learning_rate": 4.546086956860759e-05, "loss": 0.0435, "step": 55250 }, { "grad_norm": 0.41747167706489563, "learning_rate": 4.5444403395101374e-05, "loss": 0.0296, "step": 55260 }, { "grad_norm": 0.4061773419380188, "learning_rate": 4.542793771978834e-05, "loss": 0.0324, "step": 55270 }, { "grad_norm": 0.5866023302078247, "learning_rate": 4.541147254446916e-05, "loss": 0.0462, "step": 55280 }, { "grad_norm": 0.3784621059894562, "learning_rate": 4.539500787094445e-05, "loss": 0.0396, "step": 55290 }, { "grad_norm": 0.5797656774520874, "learning_rate": 4.537854370101475e-05, "loss": 0.0429, "step": 55300 }, { "grad_norm": 0.31369033455848694, "learning_rate": 4.5362080036480566e-05, "loss": 0.0344, "step": 55310 }, { "grad_norm": 0.3052166700363159, "learning_rate": 4.534561687914235e-05, "loss": 0.0312, "step": 55320 }, { "grad_norm": 0.9081755876541138, "learning_rate": 4.5329154230800466e-05, "loss": 0.0416, "step": 55330 }, { "grad_norm": 0.33087751269340515, "learning_rate": 4.531269209325525e-05, "loss": 0.035, "step": 55340 }, { "grad_norm": 0.44331660866737366, "learning_rate": 4.529623046830696e-05, "loss": 0.032, "step": 55350 }, { "grad_norm": 0.5860433578491211, "learning_rate": 4.5279769357755866e-05, "loss": 0.0477, "step": 55360 }, { "grad_norm": 0.41905149817466736, "learning_rate": 4.5263308763402084e-05, "loss": 0.0395, "step": 55370 }, { "grad_norm": 0.8935702443122864, "learning_rate": 4.524684868704574e-05, "loss": 0.0343, "step": 55380 }, { "grad_norm": 0.4787873923778534, "learning_rate": 4.523038913048688e-05, "loss": 0.0461, "step": 55390 }, { "grad_norm": 0.48725879192352295, "learning_rate": 4.521393009552549e-05, "loss": 0.0518, "step": 55400 }, { "grad_norm": 0.34563592076301575, "learning_rate": 4.5197471583961484e-05, "loss": 0.0406, "step": 55410 }, { "grad_norm": 0.4880340099334717, "learning_rate": 4.51810135975948e-05, "loss": 0.0374, "step": 55420 }, { "grad_norm": 0.18034517765045166, "learning_rate": 4.516455613822521e-05, "loss": 0.0356, "step": 55430 }, { "grad_norm": 0.2961021363735199, "learning_rate": 4.514809920765249e-05, "loss": 0.0394, "step": 55440 }, { "grad_norm": 0.4937346279621124, "learning_rate": 4.5131642807676356e-05, "loss": 0.0314, "step": 55450 }, { "grad_norm": 1.3359256982803345, "learning_rate": 4.511518694009644e-05, "loss": 0.0451, "step": 55460 }, { "grad_norm": 0.49397847056388855, "learning_rate": 4.509873160671231e-05, "loss": 0.0428, "step": 55470 }, { "grad_norm": 0.6855730414390564, "learning_rate": 4.508227680932356e-05, "loss": 0.0351, "step": 55480 }, { "grad_norm": 0.38455915451049805, "learning_rate": 4.5065822549729605e-05, "loss": 0.0397, "step": 55490 }, { "grad_norm": 0.34321197867393494, "learning_rate": 4.504936882972989e-05, "loss": 0.0342, "step": 55500 }, { "grad_norm": 0.48472410440444946, "learning_rate": 4.503291565112374e-05, "loss": 0.029, "step": 55510 }, { "grad_norm": 0.5274242758750916, "learning_rate": 4.501646301571048e-05, "loss": 0.0377, "step": 55520 }, { "grad_norm": 0.43482205271720886, "learning_rate": 4.500001092528933e-05, "loss": 0.0412, "step": 55530 }, { "grad_norm": 0.2579323947429657, "learning_rate": 4.4983559381659455e-05, "loss": 0.0345, "step": 55540 }, { "grad_norm": 0.4272782802581787, "learning_rate": 4.496710838661999e-05, "loss": 0.0338, "step": 55550 }, { "grad_norm": 0.3075295686721802, "learning_rate": 4.495065794196999e-05, "loss": 0.0406, "step": 55560 }, { "grad_norm": 0.34617528319358826, "learning_rate": 4.4934208049508455e-05, "loss": 0.0329, "step": 55570 }, { "grad_norm": 0.4181792736053467, "learning_rate": 4.4917758711034303e-05, "loss": 0.0363, "step": 55580 }, { "grad_norm": 0.4181690812110901, "learning_rate": 4.490130992834642e-05, "loss": 0.0417, "step": 55590 }, { "grad_norm": 0.6063802242279053, "learning_rate": 4.488486170324359e-05, "loss": 0.0499, "step": 55600 }, { "grad_norm": 0.42993810772895813, "learning_rate": 4.486841403752463e-05, "loss": 0.0355, "step": 55610 }, { "grad_norm": 0.30683091282844543, "learning_rate": 4.4851966932988174e-05, "loss": 0.0289, "step": 55620 }, { "grad_norm": 0.5960800051689148, "learning_rate": 4.4835520391432894e-05, "loss": 0.0412, "step": 55630 }, { "grad_norm": 0.43723878264427185, "learning_rate": 4.481907441465733e-05, "loss": 0.0404, "step": 55640 }, { "grad_norm": 0.393065482378006, "learning_rate": 4.480262900446e-05, "loss": 0.0392, "step": 55650 }, { "grad_norm": 0.4183643162250519, "learning_rate": 4.478618416263933e-05, "loss": 0.0405, "step": 55660 }, { "grad_norm": 0.4139253497123718, "learning_rate": 4.476973989099373e-05, "loss": 0.0388, "step": 55670 }, { "grad_norm": 1.6364479064941406, "learning_rate": 4.475329619132153e-05, "loss": 0.0554, "step": 55680 }, { "grad_norm": 0.5437601208686829, "learning_rate": 4.4736853065420956e-05, "loss": 0.0514, "step": 55690 }, { "grad_norm": 0.3509170413017273, "learning_rate": 4.472041051509023e-05, "loss": 0.0429, "step": 55700 }, { "grad_norm": 0.3686159551143646, "learning_rate": 4.470396854212746e-05, "loss": 0.0435, "step": 55710 }, { "grad_norm": 0.3850451111793518, "learning_rate": 4.468752714833072e-05, "loss": 0.0462, "step": 55720 }, { "grad_norm": 0.3266075849533081, "learning_rate": 4.467108633549803e-05, "loss": 0.0395, "step": 55730 }, { "grad_norm": 0.5299540162086487, "learning_rate": 4.4654646105427335e-05, "loss": 0.0568, "step": 55740 }, { "grad_norm": 0.4299376606941223, "learning_rate": 4.463820645991651e-05, "loss": 0.0348, "step": 55750 }, { "grad_norm": 0.4273071885108948, "learning_rate": 4.462176740076334e-05, "loss": 0.0346, "step": 55760 }, { "grad_norm": 0.7542853951454163, "learning_rate": 4.4605328929765624e-05, "loss": 0.0474, "step": 55770 }, { "grad_norm": 0.6155291795730591, "learning_rate": 4.458889104872102e-05, "loss": 0.0385, "step": 55780 }, { "grad_norm": 0.7293035387992859, "learning_rate": 4.457245375942715e-05, "loss": 0.0502, "step": 55790 }, { "grad_norm": 0.418934166431427, "learning_rate": 4.455601706368157e-05, "loss": 0.0382, "step": 55800 }, { "grad_norm": 0.6152178049087524, "learning_rate": 4.453958096328179e-05, "loss": 0.044, "step": 55810 }, { "grad_norm": 0.6848716139793396, "learning_rate": 4.452314546002523e-05, "loss": 0.0412, "step": 55820 }, { "grad_norm": 0.34242719411849976, "learning_rate": 4.4506710555709246e-05, "loss": 0.0452, "step": 55830 }, { "grad_norm": 0.5511244535446167, "learning_rate": 4.449027625213114e-05, "loss": 0.0335, "step": 55840 }, { "grad_norm": 0.34454837441444397, "learning_rate": 4.4473842551088135e-05, "loss": 0.0343, "step": 55850 }, { "grad_norm": 0.5003082752227783, "learning_rate": 4.445740945437738e-05, "loss": 0.032, "step": 55860 }, { "grad_norm": 0.5176052451133728, "learning_rate": 4.444097696379602e-05, "loss": 0.0377, "step": 55870 }, { "grad_norm": 0.40396374464035034, "learning_rate": 4.442454508114106e-05, "loss": 0.0348, "step": 55880 }, { "grad_norm": 0.47845152020454407, "learning_rate": 4.4408113808209465e-05, "loss": 0.0421, "step": 55890 }, { "grad_norm": 0.4996029734611511, "learning_rate": 4.439168314679813e-05, "loss": 0.0354, "step": 55900 }, { "grad_norm": 0.3013947308063507, "learning_rate": 4.43752530987039e-05, "loss": 0.0352, "step": 55910 }, { "grad_norm": 0.6174671649932861, "learning_rate": 4.4358823665723515e-05, "loss": 0.0367, "step": 55920 }, { "grad_norm": 0.6576476693153381, "learning_rate": 4.434239484965371e-05, "loss": 0.0423, "step": 55930 }, { "grad_norm": 0.31441184878349304, "learning_rate": 4.4325966652291103e-05, "loss": 0.0421, "step": 55940 }, { "grad_norm": 0.3524423837661743, "learning_rate": 4.430953907543225e-05, "loss": 0.0309, "step": 55950 }, { "grad_norm": 0.42267534136772156, "learning_rate": 4.4293112120873645e-05, "loss": 0.0309, "step": 55960 }, { "grad_norm": 0.5962513089179993, "learning_rate": 4.427668579041172e-05, "loss": 0.0319, "step": 55970 }, { "grad_norm": 0.43785184621810913, "learning_rate": 4.4260260085842816e-05, "loss": 0.0386, "step": 55980 }, { "grad_norm": 0.35073786973953247, "learning_rate": 4.4243835008963256e-05, "loss": 0.0335, "step": 55990 }, { "grad_norm": 0.5235639810562134, "learning_rate": 4.422741056156925e-05, "loss": 0.0312, "step": 56000 }, { "grad_norm": 0.32995885610580444, "learning_rate": 4.421098674545693e-05, "loss": 0.0334, "step": 56010 }, { "grad_norm": 0.38688868284225464, "learning_rate": 4.419456356242241e-05, "loss": 0.0458, "step": 56020 }, { "grad_norm": 0.43374744057655334, "learning_rate": 4.4178141014261685e-05, "loss": 0.0357, "step": 56030 }, { "grad_norm": 1.0819910764694214, "learning_rate": 4.416171910277071e-05, "loss": 0.0395, "step": 56040 }, { "grad_norm": 0.5101876854896545, "learning_rate": 4.414529782974533e-05, "loss": 0.0339, "step": 56050 }, { "grad_norm": 0.29843834042549133, "learning_rate": 4.41288771969814e-05, "loss": 0.0341, "step": 56060 }, { "grad_norm": 0.5476035475730896, "learning_rate": 4.411245720627464e-05, "loss": 0.0333, "step": 56070 }, { "grad_norm": 1.2669991254806519, "learning_rate": 4.40960378594207e-05, "loss": 0.0418, "step": 56080 }, { "grad_norm": 0.716314435005188, "learning_rate": 4.407961915821519e-05, "loss": 0.0391, "step": 56090 }, { "grad_norm": 0.369843453168869, "learning_rate": 4.406320110445362e-05, "loss": 0.0353, "step": 56100 }, { "grad_norm": 0.45228156447410583, "learning_rate": 4.404678369993144e-05, "loss": 0.032, "step": 56110 }, { "grad_norm": 0.5542625188827515, "learning_rate": 4.403036694644406e-05, "loss": 0.0418, "step": 56120 }, { "grad_norm": 0.4549596309661865, "learning_rate": 4.4013950845786764e-05, "loss": 0.0409, "step": 56130 }, { "grad_norm": 0.5257827639579773, "learning_rate": 4.399753539975482e-05, "loss": 0.0492, "step": 56140 }, { "grad_norm": 0.3243251442909241, "learning_rate": 4.398112061014337e-05, "loss": 0.033, "step": 56150 }, { "grad_norm": 0.29185593128204346, "learning_rate": 4.396470647874753e-05, "loss": 0.032, "step": 56160 }, { "grad_norm": 0.558811366558075, "learning_rate": 4.394829300736229e-05, "loss": 0.0314, "step": 56170 }, { "grad_norm": 0.43843796849250793, "learning_rate": 4.393188019778265e-05, "loss": 0.0359, "step": 56180 }, { "grad_norm": 0.35318201780319214, "learning_rate": 4.391546805180347e-05, "loss": 0.0347, "step": 56190 }, { "grad_norm": 0.6360548138618469, "learning_rate": 4.389905657121955e-05, "loss": 0.033, "step": 56200 }, { "grad_norm": 0.7560192942619324, "learning_rate": 4.3882645757825635e-05, "loss": 0.035, "step": 56210 }, { "grad_norm": 0.3799571990966797, "learning_rate": 4.386623561341637e-05, "loss": 0.0323, "step": 56220 }, { "grad_norm": 0.45828261971473694, "learning_rate": 4.384982613978637e-05, "loss": 0.0393, "step": 56230 }, { "grad_norm": 0.5651963949203491, "learning_rate": 4.383341733873012e-05, "loss": 0.0352, "step": 56240 }, { "grad_norm": 0.44681769609451294, "learning_rate": 4.381700921204209e-05, "loss": 0.0434, "step": 56250 }, { "grad_norm": 0.5307766795158386, "learning_rate": 4.380060176151663e-05, "loss": 0.0337, "step": 56260 }, { "grad_norm": 0.5257053971290588, "learning_rate": 4.378419498894805e-05, "loss": 0.0431, "step": 56270 }, { "grad_norm": 0.48008397221565247, "learning_rate": 4.376778889613056e-05, "loss": 0.0428, "step": 56280 }, { "grad_norm": 0.7138111591339111, "learning_rate": 4.375138348485831e-05, "loss": 0.0467, "step": 56290 }, { "grad_norm": 0.3627437353134155, "learning_rate": 4.3734978756925346e-05, "loss": 0.0295, "step": 56300 }, { "grad_norm": 0.436492919921875, "learning_rate": 4.3718574714125706e-05, "loss": 0.0387, "step": 56310 }, { "grad_norm": 0.438638836145401, "learning_rate": 4.370217135825329e-05, "loss": 0.0349, "step": 56320 }, { "grad_norm": 0.30791175365448, "learning_rate": 4.368576869110194e-05, "loss": 0.0403, "step": 56330 }, { "grad_norm": 0.4157433807849884, "learning_rate": 4.366936671446544e-05, "loss": 0.0332, "step": 56340 }, { "grad_norm": 0.31863757967948914, "learning_rate": 4.3652965430137474e-05, "loss": 0.0405, "step": 56350 }, { "grad_norm": 0.3038724958896637, "learning_rate": 4.3636564839911646e-05, "loss": 0.0307, "step": 56360 }, { "grad_norm": 0.37723734974861145, "learning_rate": 4.3620164945581545e-05, "loss": 0.0323, "step": 56370 }, { "grad_norm": 0.41029298305511475, "learning_rate": 4.360376574894061e-05, "loss": 0.0417, "step": 56380 }, { "grad_norm": 0.31380802392959595, "learning_rate": 4.358736725178224e-05, "loss": 0.035, "step": 56390 }, { "grad_norm": 0.4988052546977997, "learning_rate": 4.357096945589974e-05, "loss": 0.044, "step": 56400 }, { "grad_norm": 0.3839464783668518, "learning_rate": 4.355457236308636e-05, "loss": 0.0305, "step": 56410 }, { "grad_norm": 0.5046128630638123, "learning_rate": 4.353817597513526e-05, "loss": 0.0376, "step": 56420 }, { "grad_norm": 0.40528541803359985, "learning_rate": 4.352178029383948e-05, "loss": 0.041, "step": 56430 }, { "grad_norm": 0.4887567162513733, "learning_rate": 4.3505385320992105e-05, "loss": 0.0374, "step": 56440 }, { "grad_norm": 0.43785515427589417, "learning_rate": 4.348899105838602e-05, "loss": 0.0382, "step": 56450 }, { "grad_norm": 0.5380211472511292, "learning_rate": 4.3472597507814087e-05, "loss": 0.0377, "step": 56460 }, { "grad_norm": 0.4244048297405243, "learning_rate": 4.3456204671069066e-05, "loss": 0.041, "step": 56470 }, { "grad_norm": 0.32519251108169556, "learning_rate": 4.343981254994367e-05, "loss": 0.0285, "step": 56480 }, { "grad_norm": 0.5703268647193909, "learning_rate": 4.34234211462305e-05, "loss": 0.0427, "step": 56490 }, { "grad_norm": 0.3858642876148224, "learning_rate": 4.340703046172213e-05, "loss": 0.0346, "step": 56500 }, { "grad_norm": 0.40000686049461365, "learning_rate": 4.339064049821097e-05, "loss": 0.0407, "step": 56510 }, { "grad_norm": 0.5465078353881836, "learning_rate": 4.3374251257489446e-05, "loss": 0.0417, "step": 56520 }, { "grad_norm": 0.7374443411827087, "learning_rate": 4.335786274134984e-05, "loss": 0.055, "step": 56530 }, { "grad_norm": 0.5648030042648315, "learning_rate": 4.334147495158438e-05, "loss": 0.0444, "step": 56540 }, { "grad_norm": 0.5905149579048157, "learning_rate": 4.33250878899852e-05, "loss": 0.0422, "step": 56550 }, { "grad_norm": 0.45807498693466187, "learning_rate": 4.330870155834437e-05, "loss": 0.0468, "step": 56560 }, { "grad_norm": 0.44547510147094727, "learning_rate": 4.3292315958453884e-05, "loss": 0.037, "step": 56570 }, { "grad_norm": 0.34636420011520386, "learning_rate": 4.3275931092105646e-05, "loss": 0.0374, "step": 56580 }, { "grad_norm": 0.29469946026802063, "learning_rate": 4.325954696109147e-05, "loss": 0.0367, "step": 56590 }, { "grad_norm": 0.5620444416999817, "learning_rate": 4.32431635672031e-05, "loss": 0.0501, "step": 56600 }, { "grad_norm": 0.733909010887146, "learning_rate": 4.3226780912232214e-05, "loss": 0.0403, "step": 56610 }, { "grad_norm": 0.353615939617157, "learning_rate": 4.321039899797036e-05, "loss": 0.0379, "step": 56620 }, { "grad_norm": 0.34288838505744934, "learning_rate": 4.319401782620908e-05, "loss": 0.0331, "step": 56630 }, { "grad_norm": 0.31567391753196716, "learning_rate": 4.317763739873978e-05, "loss": 0.0326, "step": 56640 }, { "grad_norm": 0.5207674503326416, "learning_rate": 4.31612577173538e-05, "loss": 0.0372, "step": 56650 }, { "grad_norm": 0.5649055242538452, "learning_rate": 4.314487878384239e-05, "loss": 0.0453, "step": 56660 }, { "grad_norm": 0.5854834318161011, "learning_rate": 4.3128500599996736e-05, "loss": 0.0317, "step": 56670 }, { "grad_norm": 0.33826151490211487, "learning_rate": 4.31121231676079e-05, "loss": 0.0433, "step": 56680 }, { "grad_norm": 0.49247968196868896, "learning_rate": 4.309574648846694e-05, "loss": 0.0422, "step": 56690 }, { "grad_norm": 0.5865444540977478, "learning_rate": 4.3079370564364755e-05, "loss": 0.046, "step": 56700 }, { "grad_norm": 0.30224451422691345, "learning_rate": 4.306299539709221e-05, "loss": 0.0346, "step": 56710 }, { "grad_norm": 0.3071497082710266, "learning_rate": 4.3046620988440046e-05, "loss": 0.0403, "step": 56720 }, { "grad_norm": 0.5690569281578064, "learning_rate": 4.303024734019897e-05, "loss": 0.0311, "step": 56730 }, { "grad_norm": 0.4891969561576843, "learning_rate": 4.3013874454159565e-05, "loss": 0.0427, "step": 56740 }, { "grad_norm": 0.6262803077697754, "learning_rate": 4.299750233211233e-05, "loss": 0.0542, "step": 56750 }, { "grad_norm": 0.4931578040122986, "learning_rate": 4.2981130975847715e-05, "loss": 0.0382, "step": 56760 }, { "grad_norm": 0.5431190729141235, "learning_rate": 4.2964760387156076e-05, "loss": 0.0399, "step": 56770 }, { "grad_norm": 0.44150376319885254, "learning_rate": 4.294839056782767e-05, "loss": 0.035, "step": 56780 }, { "grad_norm": 1.624319076538086, "learning_rate": 4.293202151965266e-05, "loss": 0.0524, "step": 56790 }, { "grad_norm": 0.4927184581756592, "learning_rate": 4.291565324442115e-05, "loss": 0.0367, "step": 56800 }, { "grad_norm": 0.44851186871528625, "learning_rate": 4.2899285743923135e-05, "loss": 0.0456, "step": 56810 }, { "grad_norm": 0.30199578404426575, "learning_rate": 4.288291901994857e-05, "loss": 0.0336, "step": 56820 }, { "grad_norm": 0.5592645406723022, "learning_rate": 4.286655307428729e-05, "loss": 0.0355, "step": 56830 }, { "grad_norm": 0.3833967447280884, "learning_rate": 4.285018790872903e-05, "loss": 0.0492, "step": 56840 }, { "grad_norm": 0.47641807794570923, "learning_rate": 4.283382352506347e-05, "loss": 0.0456, "step": 56850 }, { "grad_norm": 0.360260546207428, "learning_rate": 4.2817459925080205e-05, "loss": 0.0543, "step": 56860 }, { "grad_norm": 0.452221542596817, "learning_rate": 4.28010971105687e-05, "loss": 0.042, "step": 56870 }, { "grad_norm": 1.352655053138733, "learning_rate": 4.278473508331841e-05, "loss": 0.0407, "step": 56880 }, { "grad_norm": 0.45251789689064026, "learning_rate": 4.276837384511864e-05, "loss": 0.0445, "step": 56890 }, { "grad_norm": 0.4666394591331482, "learning_rate": 4.275201339775864e-05, "loss": 0.0386, "step": 56900 }, { "grad_norm": 0.3560999929904938, "learning_rate": 4.273565374302756e-05, "loss": 0.0454, "step": 56910 }, { "grad_norm": 0.496169775724411, "learning_rate": 4.2719294882714454e-05, "loss": 0.0353, "step": 56920 }, { "grad_norm": 0.2829429507255554, "learning_rate": 4.270293681860831e-05, "loss": 0.0402, "step": 56930 }, { "grad_norm": 0.5378888249397278, "learning_rate": 4.268657955249802e-05, "loss": 0.0386, "step": 56940 }, { "grad_norm": 0.4405052959918976, "learning_rate": 4.267022308617241e-05, "loss": 0.0293, "step": 56950 }, { "grad_norm": 0.35797518491744995, "learning_rate": 4.265386742142018e-05, "loss": 0.0396, "step": 56960 }, { "grad_norm": 0.8398118615150452, "learning_rate": 4.263751256002996e-05, "loss": 0.0376, "step": 56970 }, { "grad_norm": 0.5368621349334717, "learning_rate": 4.26211585037903e-05, "loss": 0.0427, "step": 56980 }, { "grad_norm": 0.47479891777038574, "learning_rate": 4.2604805254489656e-05, "loss": 0.0422, "step": 56990 }, { "grad_norm": 0.6645020842552185, "learning_rate": 4.258845281391638e-05, "loss": 0.0401, "step": 57000 }, { "grad_norm": 0.266613245010376, "learning_rate": 4.257210118385876e-05, "loss": 0.0376, "step": 57010 }, { "grad_norm": 0.5034677982330322, "learning_rate": 4.2555750366104994e-05, "loss": 0.0446, "step": 57020 }, { "grad_norm": 0.3346105217933655, "learning_rate": 4.2539400362443184e-05, "loss": 0.0409, "step": 57030 }, { "grad_norm": 0.6747689843177795, "learning_rate": 4.252305117466133e-05, "loss": 0.0434, "step": 57040 }, { "grad_norm": 0.5436294674873352, "learning_rate": 4.2506702804547356e-05, "loss": 0.0373, "step": 57050 }, { "grad_norm": 0.7110199928283691, "learning_rate": 4.249035525388907e-05, "loss": 0.0366, "step": 57060 }, { "grad_norm": 0.6129137873649597, "learning_rate": 4.247400852447427e-05, "loss": 0.0395, "step": 57070 }, { "grad_norm": 0.32945480942726135, "learning_rate": 4.245766261809059e-05, "loss": 0.0347, "step": 57080 }, { "grad_norm": 0.4782121181488037, "learning_rate": 4.244131753652558e-05, "loss": 0.0369, "step": 57090 }, { "grad_norm": 0.3639507591724396, "learning_rate": 4.242497328156672e-05, "loss": 0.0327, "step": 57100 }, { "grad_norm": 0.48583856225013733, "learning_rate": 4.240862985500139e-05, "loss": 0.0487, "step": 57110 }, { "grad_norm": 0.3404214382171631, "learning_rate": 4.2392287258616894e-05, "loss": 0.0314, "step": 57120 }, { "grad_norm": 0.24628044664859772, "learning_rate": 4.2375945494200396e-05, "loss": 0.0326, "step": 57130 }, { "grad_norm": 0.3746589124202728, "learning_rate": 4.2359604563539054e-05, "loss": 0.0462, "step": 57140 }, { "grad_norm": 0.41177478432655334, "learning_rate": 4.2343264468419885e-05, "loss": 0.0408, "step": 57150 }, { "grad_norm": 0.2682482600212097, "learning_rate": 4.232692521062979e-05, "loss": 0.0273, "step": 57160 }, { "grad_norm": 0.6065890789031982, "learning_rate": 4.231058679195562e-05, "loss": 0.0446, "step": 57170 }, { "grad_norm": 0.3088012635707855, "learning_rate": 4.229424921418411e-05, "loss": 0.0386, "step": 57180 }, { "grad_norm": 0.4271482229232788, "learning_rate": 4.2277912479101916e-05, "loss": 0.0404, "step": 57190 }, { "grad_norm": 0.8573782444000244, "learning_rate": 4.226157658849562e-05, "loss": 0.0429, "step": 57200 }, { "grad_norm": 0.41459009051322937, "learning_rate": 4.2245241544151664e-05, "loss": 0.0363, "step": 57210 }, { "grad_norm": 0.48939380049705505, "learning_rate": 4.2228907347856425e-05, "loss": 0.053, "step": 57220 }, { "grad_norm": 0.4074645936489105, "learning_rate": 4.2212574001396214e-05, "loss": 0.0421, "step": 57230 }, { "grad_norm": 0.4655267596244812, "learning_rate": 4.21962415065572e-05, "loss": 0.0392, "step": 57240 }, { "grad_norm": 0.3436918556690216, "learning_rate": 4.217990986512548e-05, "loss": 0.0392, "step": 57250 }, { "grad_norm": 0.3456002473831177, "learning_rate": 4.216357907888704e-05, "loss": 0.0414, "step": 57260 }, { "grad_norm": 0.6117721199989319, "learning_rate": 4.2147249149627824e-05, "loss": 0.0394, "step": 57270 }, { "grad_norm": 0.44506949186325073, "learning_rate": 4.213092007913364e-05, "loss": 0.0423, "step": 57280 }, { "grad_norm": 0.2088681161403656, "learning_rate": 4.2114591869190205e-05, "loss": 0.0486, "step": 57290 }, { "grad_norm": 0.2569322884082794, "learning_rate": 4.209826452158315e-05, "loss": 0.045, "step": 57300 }, { "grad_norm": 0.4864844083786011, "learning_rate": 4.2081938038098e-05, "loss": 0.0405, "step": 57310 }, { "grad_norm": 0.4961419105529785, "learning_rate": 4.2065612420520185e-05, "loss": 0.0404, "step": 57320 }, { "grad_norm": 0.466753751039505, "learning_rate": 4.204928767063509e-05, "loss": 0.0419, "step": 57330 }, { "grad_norm": 0.32392024993896484, "learning_rate": 4.2032963790227944e-05, "loss": 0.0385, "step": 57340 }, { "grad_norm": 0.3428501486778259, "learning_rate": 4.2016640781083887e-05, "loss": 0.0338, "step": 57350 }, { "grad_norm": 0.461290568113327, "learning_rate": 4.2000318644988004e-05, "loss": 0.0546, "step": 57360 }, { "grad_norm": 0.4238690435886383, "learning_rate": 4.198399738372524e-05, "loss": 0.034, "step": 57370 }, { "grad_norm": 0.38806259632110596, "learning_rate": 4.196767699908045e-05, "loss": 0.0325, "step": 57380 }, { "grad_norm": 0.41902315616607666, "learning_rate": 4.195135749283845e-05, "loss": 0.0395, "step": 57390 }, { "grad_norm": 0.6908210515975952, "learning_rate": 4.1935038866783894e-05, "loss": 0.0351, "step": 57400 }, { "grad_norm": 0.3302880823612213, "learning_rate": 4.1918721122701355e-05, "loss": 0.0318, "step": 57410 }, { "grad_norm": 0.5859477519989014, "learning_rate": 4.190240426237533e-05, "loss": 0.048, "step": 57420 }, { "grad_norm": 0.26641178131103516, "learning_rate": 4.188608828759019e-05, "loss": 0.0359, "step": 57430 }, { "grad_norm": 0.43499597907066345, "learning_rate": 4.1869773200130246e-05, "loss": 0.0625, "step": 57440 }, { "grad_norm": 0.48205479979515076, "learning_rate": 4.185345900177967e-05, "loss": 0.0448, "step": 57450 }, { "grad_norm": 0.35627609491348267, "learning_rate": 4.183714569432258e-05, "loss": 0.0384, "step": 57460 }, { "grad_norm": 0.36290016770362854, "learning_rate": 4.182083327954295e-05, "loss": 0.0366, "step": 57470 }, { "grad_norm": 0.47584855556488037, "learning_rate": 4.1804521759224714e-05, "loss": 0.0356, "step": 57480 }, { "grad_norm": 0.28326359391212463, "learning_rate": 4.178821113515165e-05, "loss": 0.0456, "step": 57490 }, { "grad_norm": 0.43534383177757263, "learning_rate": 4.1771901409107476e-05, "loss": 0.047, "step": 57500 }, { "grad_norm": 0.49471932649612427, "learning_rate": 4.175559258287577e-05, "loss": 0.0354, "step": 57510 }, { "grad_norm": 0.426054984331131, "learning_rate": 4.173928465824008e-05, "loss": 0.033, "step": 57520 }, { "grad_norm": 0.3548782169818878, "learning_rate": 4.17229776369838e-05, "loss": 0.0395, "step": 57530 }, { "grad_norm": 0.29302874207496643, "learning_rate": 4.170667152089025e-05, "loss": 0.0422, "step": 57540 }, { "grad_norm": 0.31805434823036194, "learning_rate": 4.1690366311742626e-05, "loss": 0.0263, "step": 57550 }, { "grad_norm": 0.4016324281692505, "learning_rate": 4.1674062011324055e-05, "loss": 0.0273, "step": 57560 }, { "grad_norm": 0.4376792311668396, "learning_rate": 4.165775862141752e-05, "loss": 0.0332, "step": 57570 }, { "grad_norm": 0.49753138422966003, "learning_rate": 4.164145614380598e-05, "loss": 0.0318, "step": 57580 }, { "grad_norm": 0.3530893325805664, "learning_rate": 4.162515458027223e-05, "loss": 0.0359, "step": 57590 }, { "grad_norm": 1.5965678691864014, "learning_rate": 4.160885393259899e-05, "loss": 0.0396, "step": 57600 }, { "grad_norm": 0.4897496998310089, "learning_rate": 4.159255420256886e-05, "loss": 0.043, "step": 57610 }, { "grad_norm": 0.43808236718177795, "learning_rate": 4.157625539196436e-05, "loss": 0.0359, "step": 57620 }, { "grad_norm": 0.44894757866859436, "learning_rate": 4.155995750256791e-05, "loss": 0.047, "step": 57630 }, { "grad_norm": 0.37713780999183655, "learning_rate": 4.15436605361618e-05, "loss": 0.0344, "step": 57640 }, { "grad_norm": 0.48236024379730225, "learning_rate": 4.152736449452827e-05, "loss": 0.0497, "step": 57650 }, { "grad_norm": 0.3581143617630005, "learning_rate": 4.151106937944943e-05, "loss": 0.031, "step": 57660 }, { "grad_norm": 0.34244200587272644, "learning_rate": 4.149477519270727e-05, "loss": 0.0301, "step": 57670 }, { "grad_norm": 0.48387449979782104, "learning_rate": 4.147848193608371e-05, "loss": 0.0474, "step": 57680 }, { "grad_norm": 1.2794232368469238, "learning_rate": 4.146218961136056e-05, "loss": 0.039, "step": 57690 }, { "grad_norm": 0.44986361265182495, "learning_rate": 4.14458982203195e-05, "loss": 0.0465, "step": 57700 }, { "grad_norm": 0.309137225151062, "learning_rate": 4.1429607764742165e-05, "loss": 0.0437, "step": 57710 }, { "grad_norm": 0.46565356850624084, "learning_rate": 4.141331824641003e-05, "loss": 0.0465, "step": 57720 }, { "grad_norm": 0.4532824158668518, "learning_rate": 4.139702966710452e-05, "loss": 0.048, "step": 57730 }, { "grad_norm": 0.377545565366745, "learning_rate": 4.13807420286069e-05, "loss": 0.0444, "step": 57740 }, { "grad_norm": 0.3482200801372528, "learning_rate": 4.136445533269837e-05, "loss": 0.038, "step": 57750 }, { "grad_norm": 0.3473886549472809, "learning_rate": 4.134816958116003e-05, "loss": 0.0306, "step": 57760 }, { "grad_norm": 0.6281755566596985, "learning_rate": 4.133188477577283e-05, "loss": 0.0508, "step": 57770 }, { "grad_norm": 0.5145444869995117, "learning_rate": 4.131560091831771e-05, "loss": 0.0351, "step": 57780 }, { "grad_norm": 0.6529936194419861, "learning_rate": 4.1299318010575405e-05, "loss": 0.036, "step": 57790 }, { "grad_norm": 0.48098042607307434, "learning_rate": 4.128303605432661e-05, "loss": 0.0588, "step": 57800 }, { "grad_norm": 0.4991486668586731, "learning_rate": 4.126675505135188e-05, "loss": 0.0373, "step": 57810 }, { "grad_norm": 0.5726073384284973, "learning_rate": 4.125047500343169e-05, "loss": 0.0371, "step": 57820 }, { "grad_norm": 0.5020180344581604, "learning_rate": 4.1234195912346376e-05, "loss": 0.0367, "step": 57830 }, { "grad_norm": 1.1542662382125854, "learning_rate": 4.1217917779876235e-05, "loss": 0.0423, "step": 57840 }, { "grad_norm": 0.43676456809043884, "learning_rate": 4.120164060780141e-05, "loss": 0.0453, "step": 57850 }, { "grad_norm": 0.38703790307044983, "learning_rate": 4.1185364397901935e-05, "loss": 0.0408, "step": 57860 }, { "grad_norm": 0.3903384804725647, "learning_rate": 4.116908915195776e-05, "loss": 0.032, "step": 57870 }, { "grad_norm": 0.4827577471733093, "learning_rate": 4.115281487174872e-05, "loss": 0.034, "step": 57880 }, { "grad_norm": 0.7315165996551514, "learning_rate": 4.113654155905452e-05, "loss": 0.0404, "step": 57890 }, { "grad_norm": 0.34891241788864136, "learning_rate": 4.112026921565483e-05, "loss": 0.0363, "step": 57900 }, { "grad_norm": 0.5571714639663696, "learning_rate": 4.1103997843329147e-05, "loss": 0.0353, "step": 57910 }, { "grad_norm": 0.4006617069244385, "learning_rate": 4.108772744385689e-05, "loss": 0.0501, "step": 57920 }, { "grad_norm": 0.3327639102935791, "learning_rate": 4.1071458019017346e-05, "loss": 0.0364, "step": 57930 }, { "grad_norm": 0.3807273209095001, "learning_rate": 4.1055189570589734e-05, "loss": 0.0461, "step": 57940 }, { "grad_norm": 0.4285677671432495, "learning_rate": 4.103892210035315e-05, "loss": 0.0315, "step": 57950 }, { "grad_norm": 1.1433296203613281, "learning_rate": 4.102265561008656e-05, "loss": 0.0405, "step": 57960 }, { "grad_norm": 0.3805805742740631, "learning_rate": 4.100639010156885e-05, "loss": 0.0316, "step": 57970 }, { "grad_norm": 0.44329729676246643, "learning_rate": 4.099012557657882e-05, "loss": 0.0373, "step": 57980 }, { "grad_norm": 0.6418752074241638, "learning_rate": 4.09738620368951e-05, "loss": 0.0453, "step": 57990 }, { "grad_norm": 0.35181939601898193, "learning_rate": 4.095759948429626e-05, "loss": 0.0322, "step": 58000 }, { "grad_norm": 0.48795944452285767, "learning_rate": 4.094133792056075e-05, "loss": 0.0442, "step": 58010 }, { "grad_norm": 0.38510704040527344, "learning_rate": 4.092507734746688e-05, "loss": 0.0409, "step": 58020 }, { "grad_norm": 0.4575027823448181, "learning_rate": 4.090881776679293e-05, "loss": 0.035, "step": 58030 }, { "grad_norm": 0.5280795097351074, "learning_rate": 4.0892559180316994e-05, "loss": 0.0408, "step": 58040 }, { "grad_norm": 0.5488776564598083, "learning_rate": 4.08763015898171e-05, "loss": 0.0314, "step": 58050 }, { "grad_norm": 0.5215383172035217, "learning_rate": 4.0860044997071145e-05, "loss": 0.0439, "step": 58060 }, { "grad_norm": 0.42492127418518066, "learning_rate": 4.084378940385693e-05, "loss": 0.037, "step": 58070 }, { "grad_norm": 0.5232763290405273, "learning_rate": 4.082753481195211e-05, "loss": 0.0397, "step": 58080 }, { "grad_norm": 0.46306127309799194, "learning_rate": 4.0811281223134324e-05, "loss": 0.0359, "step": 58090 }, { "grad_norm": 0.32645928859710693, "learning_rate": 4.0795028639181e-05, "loss": 0.0461, "step": 58100 }, { "grad_norm": 0.38600173592567444, "learning_rate": 4.077877706186951e-05, "loss": 0.0327, "step": 58110 }, { "grad_norm": 0.424019992351532, "learning_rate": 4.076252649297709e-05, "loss": 0.0355, "step": 58120 }, { "grad_norm": 0.4589746296405792, "learning_rate": 4.07462769342809e-05, "loss": 0.0376, "step": 58130 }, { "grad_norm": 0.5310263633728027, "learning_rate": 4.073002838755793e-05, "loss": 0.0387, "step": 58140 }, { "grad_norm": 0.4284641444683075, "learning_rate": 4.0713780854585115e-05, "loss": 0.0313, "step": 58150 }, { "grad_norm": 0.48425421118736267, "learning_rate": 4.069753433713927e-05, "loss": 0.0486, "step": 58160 }, { "grad_norm": 0.48664391040802, "learning_rate": 4.068128883699709e-05, "loss": 0.0377, "step": 58170 }, { "grad_norm": 0.6423162221908569, "learning_rate": 4.0665044355935155e-05, "loss": 0.0422, "step": 58180 }, { "grad_norm": 0.37384146451950073, "learning_rate": 4.064880089572993e-05, "loss": 0.0357, "step": 58190 }, { "grad_norm": 0.5002710223197937, "learning_rate": 4.063255845815779e-05, "loss": 0.0453, "step": 58200 }, { "grad_norm": 0.3257937431335449, "learning_rate": 4.0616317044994945e-05, "loss": 0.0354, "step": 58210 }, { "grad_norm": 0.4571438729763031, "learning_rate": 4.0600076658017585e-05, "loss": 0.0289, "step": 58220 }, { "grad_norm": 0.331897497177124, "learning_rate": 4.0583837299001706e-05, "loss": 0.0304, "step": 58230 }, { "grad_norm": 0.3632098436355591, "learning_rate": 4.056759896972322e-05, "loss": 0.0339, "step": 58240 }, { "grad_norm": 0.324451744556427, "learning_rate": 4.055136167195793e-05, "loss": 0.0402, "step": 58250 }, { "grad_norm": 0.3424450755119324, "learning_rate": 4.053512540748152e-05, "loss": 0.0354, "step": 58260 }, { "grad_norm": 0.5817524790763855, "learning_rate": 4.051889017806955e-05, "loss": 0.0438, "step": 58270 }, { "grad_norm": 0.6378970742225647, "learning_rate": 4.0502655985497516e-05, "loss": 0.0381, "step": 58280 }, { "grad_norm": 0.37293320894241333, "learning_rate": 4.048642283154074e-05, "loss": 0.0454, "step": 58290 }, { "grad_norm": 0.3504335582256317, "learning_rate": 4.0470190717974464e-05, "loss": 0.0457, "step": 58300 }, { "grad_norm": 0.5371382832527161, "learning_rate": 4.045395964657381e-05, "loss": 0.0352, "step": 58310 }, { "grad_norm": 0.6147612929344177, "learning_rate": 4.043772961911376e-05, "loss": 0.0418, "step": 58320 }, { "grad_norm": 0.49785444140434265, "learning_rate": 4.042150063736924e-05, "loss": 0.0353, "step": 58330 }, { "grad_norm": 0.5537617206573486, "learning_rate": 4.0405272703114976e-05, "loss": 0.0395, "step": 58340 }, { "grad_norm": 0.5452893376350403, "learning_rate": 4.038904581812569e-05, "loss": 0.036, "step": 58350 }, { "grad_norm": 0.35976481437683105, "learning_rate": 4.03728199841759e-05, "loss": 0.0297, "step": 58360 }, { "grad_norm": 0.2926620841026306, "learning_rate": 4.0356595203040044e-05, "loss": 0.038, "step": 58370 }, { "grad_norm": 0.33761370182037354, "learning_rate": 4.0340371476492426e-05, "loss": 0.0428, "step": 58380 }, { "grad_norm": 0.4617091715335846, "learning_rate": 4.032414880630727e-05, "loss": 0.0407, "step": 58390 }, { "grad_norm": 0.4243026673793793, "learning_rate": 4.0307927194258646e-05, "loss": 0.0317, "step": 58400 }, { "grad_norm": 0.34344369173049927, "learning_rate": 4.029170664212054e-05, "loss": 0.0358, "step": 58410 }, { "grad_norm": 0.3728729486465454, "learning_rate": 4.0275487151666794e-05, "loss": 0.0411, "step": 58420 }, { "grad_norm": 0.4748590886592865, "learning_rate": 4.025926872467116e-05, "loss": 0.0399, "step": 58430 }, { "grad_norm": 0.4847734868526459, "learning_rate": 4.024305136290725e-05, "loss": 0.036, "step": 58440 }, { "grad_norm": 0.3356042802333832, "learning_rate": 4.022683506814858e-05, "loss": 0.0401, "step": 58450 }, { "grad_norm": 0.4704173803329468, "learning_rate": 4.021061984216853e-05, "loss": 0.0395, "step": 58460 }, { "grad_norm": 0.26225972175598145, "learning_rate": 4.019440568674035e-05, "loss": 0.028, "step": 58470 }, { "grad_norm": 0.8062153458595276, "learning_rate": 4.017819260363724e-05, "loss": 0.0387, "step": 58480 }, { "grad_norm": 0.4447593688964844, "learning_rate": 4.016198059463223e-05, "loss": 0.05, "step": 58490 }, { "grad_norm": 0.2457939237356186, "learning_rate": 4.0145769661498233e-05, "loss": 0.0397, "step": 58500 }, { "grad_norm": 1.5276496410369873, "learning_rate": 4.012955980600804e-05, "loss": 0.0399, "step": 58510 }, { "grad_norm": 0.5189074873924255, "learning_rate": 4.0113351029934355e-05, "loss": 0.0422, "step": 58520 }, { "grad_norm": 0.39866819977760315, "learning_rate": 4.0097143335049714e-05, "loss": 0.0381, "step": 58530 }, { "grad_norm": 0.4058789610862732, "learning_rate": 4.0080936723126606e-05, "loss": 0.0297, "step": 58540 }, { "grad_norm": 0.3826952874660492, "learning_rate": 4.006473119593735e-05, "loss": 0.0387, "step": 58550 }, { "grad_norm": 0.4767889082431793, "learning_rate": 4.004852675525415e-05, "loss": 0.0444, "step": 58560 }, { "grad_norm": 0.4267693758010864, "learning_rate": 4.00323234028491e-05, "loss": 0.035, "step": 58570 }, { "grad_norm": 0.42019301652908325, "learning_rate": 4.0016121140494186e-05, "loss": 0.0426, "step": 58580 }, { "grad_norm": 0.4296271502971649, "learning_rate": 3.999991996996122e-05, "loss": 0.0345, "step": 58590 }, { "grad_norm": 0.4209338426589966, "learning_rate": 3.9983719893022e-05, "loss": 0.034, "step": 58600 }, { "grad_norm": 0.5670241713523865, "learning_rate": 3.996752091144812e-05, "loss": 0.0388, "step": 58610 }, { "grad_norm": 0.4855457842350006, "learning_rate": 3.9951323027011055e-05, "loss": 0.0363, "step": 58620 }, { "grad_norm": 0.31853845715522766, "learning_rate": 3.993512624148219e-05, "loss": 0.0412, "step": 58630 }, { "grad_norm": 0.35744431614875793, "learning_rate": 3.99189305566328e-05, "loss": 0.0394, "step": 58640 }, { "grad_norm": 0.27148059010505676, "learning_rate": 3.990273597423401e-05, "loss": 0.0422, "step": 58650 }, { "grad_norm": 0.3505784571170807, "learning_rate": 3.988654249605681e-05, "loss": 0.0355, "step": 58660 }, { "grad_norm": 0.5400404930114746, "learning_rate": 3.987035012387213e-05, "loss": 0.0425, "step": 58670 }, { "grad_norm": 0.5081378221511841, "learning_rate": 3.9854158859450735e-05, "loss": 0.0325, "step": 58680 }, { "grad_norm": 0.3490731120109558, "learning_rate": 3.983796870456328e-05, "loss": 0.0395, "step": 58690 }, { "grad_norm": 0.1531730592250824, "learning_rate": 3.982177966098028e-05, "loss": 0.0273, "step": 58700 }, { "grad_norm": 0.34034258127212524, "learning_rate": 3.980559173047216e-05, "loss": 0.0505, "step": 58710 }, { "grad_norm": 0.4506327509880066, "learning_rate": 3.9789404914809176e-05, "loss": 0.0402, "step": 58720 }, { "grad_norm": 0.3577331304550171, "learning_rate": 3.9773219215761544e-05, "loss": 0.0372, "step": 58730 }, { "grad_norm": 0.24397948384284973, "learning_rate": 3.975703463509928e-05, "loss": 0.036, "step": 58740 }, { "grad_norm": 0.5305646657943726, "learning_rate": 3.974085117459231e-05, "loss": 0.0371, "step": 58750 }, { "grad_norm": 0.38850367069244385, "learning_rate": 3.972466883601043e-05, "loss": 0.0449, "step": 58760 }, { "grad_norm": 0.24567966163158417, "learning_rate": 3.9708487621123316e-05, "loss": 0.0288, "step": 58770 }, { "grad_norm": 0.4314492642879486, "learning_rate": 3.969230753170051e-05, "loss": 0.0433, "step": 58780 }, { "grad_norm": 0.6751468181610107, "learning_rate": 3.967612856951146e-05, "loss": 0.0357, "step": 58790 }, { "grad_norm": 0.5829444527626038, "learning_rate": 3.9659950736325476e-05, "loss": 0.0353, "step": 58800 }, { "grad_norm": 0.7202253341674805, "learning_rate": 3.964377403391174e-05, "loss": 0.0707, "step": 58810 }, { "grad_norm": 0.43407246470451355, "learning_rate": 3.9627598464039303e-05, "loss": 0.0579, "step": 58820 }, { "grad_norm": 0.32371464371681213, "learning_rate": 3.9611424028477096e-05, "loss": 0.0411, "step": 58830 }, { "grad_norm": 0.7014611959457397, "learning_rate": 3.959525072899394e-05, "loss": 0.0423, "step": 58840 }, { "grad_norm": 0.4115981459617615, "learning_rate": 3.9579078567358506e-05, "loss": 0.0304, "step": 58850 }, { "grad_norm": 0.4676778316497803, "learning_rate": 3.956290754533939e-05, "loss": 0.0391, "step": 58860 }, { "grad_norm": 0.404496967792511, "learning_rate": 3.954673766470501e-05, "loss": 0.0479, "step": 58870 }, { "grad_norm": 0.3164180815219879, "learning_rate": 3.9530568927223674e-05, "loss": 0.0408, "step": 58880 }, { "grad_norm": 0.3331127166748047, "learning_rate": 3.95144013346636e-05, "loss": 0.0413, "step": 58890 }, { "grad_norm": 0.2895355820655823, "learning_rate": 3.949823488879282e-05, "loss": 0.0334, "step": 58900 }, { "grad_norm": 0.4000466763973236, "learning_rate": 3.9482069591379276e-05, "loss": 0.0433, "step": 58910 }, { "grad_norm": 0.4646075665950775, "learning_rate": 3.9465905444190784e-05, "loss": 0.0417, "step": 58920 }, { "grad_norm": 0.2912916839122772, "learning_rate": 3.944974244899505e-05, "loss": 0.0466, "step": 58930 }, { "grad_norm": 0.3551655113697052, "learning_rate": 3.9433580607559616e-05, "loss": 0.0401, "step": 58940 }, { "grad_norm": 0.323809415102005, "learning_rate": 3.941741992165193e-05, "loss": 0.0422, "step": 58950 }, { "grad_norm": 0.4606492221355438, "learning_rate": 3.940126039303928e-05, "loss": 0.0381, "step": 58960 }, { "grad_norm": 0.38545453548431396, "learning_rate": 3.938510202348886e-05, "loss": 0.0383, "step": 58970 }, { "grad_norm": 0.6982215642929077, "learning_rate": 3.93689448147677e-05, "loss": 0.0356, "step": 58980 }, { "grad_norm": 0.359162300825119, "learning_rate": 3.935278876864278e-05, "loss": 0.0321, "step": 58990 }, { "grad_norm": 0.4233788847923279, "learning_rate": 3.933663388688087e-05, "loss": 0.0419, "step": 59000 }, { "grad_norm": 0.5757928490638733, "learning_rate": 3.9320480171248645e-05, "loss": 0.0349, "step": 59010 }, { "grad_norm": 0.6728726625442505, "learning_rate": 3.930432762351265e-05, "loss": 0.0523, "step": 59020 }, { "grad_norm": 0.4638859033584595, "learning_rate": 3.928817624543931e-05, "loss": 0.0454, "step": 59030 }, { "grad_norm": 0.4681670367717743, "learning_rate": 3.9272026038794885e-05, "loss": 0.0347, "step": 59040 }, { "grad_norm": 0.5707354545593262, "learning_rate": 3.9255877005345586e-05, "loss": 0.0416, "step": 59050 }, { "grad_norm": 0.6634946465492249, "learning_rate": 3.923972914685743e-05, "loss": 0.0393, "step": 59060 }, { "grad_norm": 0.27898329496383667, "learning_rate": 3.922358246509631e-05, "loss": 0.0383, "step": 59070 }, { "grad_norm": 0.3449069857597351, "learning_rate": 3.920743696182801e-05, "loss": 0.0427, "step": 59080 }, { "grad_norm": 0.3584041893482208, "learning_rate": 3.9191292638818164e-05, "loss": 0.0401, "step": 59090 }, { "grad_norm": 0.7849502563476562, "learning_rate": 3.9175149497832316e-05, "loss": 0.0401, "step": 59100 }, { "grad_norm": 0.30390340089797974, "learning_rate": 3.915900754063584e-05, "loss": 0.043, "step": 59110 }, { "grad_norm": 0.46031343936920166, "learning_rate": 3.9142866768994014e-05, "loss": 0.0455, "step": 59120 }, { "grad_norm": 0.3858221769332886, "learning_rate": 3.912672718467193e-05, "loss": 0.0425, "step": 59130 }, { "grad_norm": 0.43234023451805115, "learning_rate": 3.911058878943463e-05, "loss": 0.0361, "step": 59140 }, { "grad_norm": 0.3259426951408386, "learning_rate": 3.9094451585046974e-05, "loss": 0.0322, "step": 59150 }, { "grad_norm": 0.3292646110057831, "learning_rate": 3.90783155732737e-05, "loss": 0.0345, "step": 59160 }, { "grad_norm": 0.39144396781921387, "learning_rate": 3.9062180755879373e-05, "loss": 0.0334, "step": 59170 }, { "grad_norm": 0.46188050508499146, "learning_rate": 3.904604713462855e-05, "loss": 0.0376, "step": 59180 }, { "grad_norm": 0.2918568551540375, "learning_rate": 3.902991471128554e-05, "loss": 0.0244, "step": 59190 }, { "grad_norm": 0.3136512041091919, "learning_rate": 3.901378348761456e-05, "loss": 0.028, "step": 59200 }, { "grad_norm": 0.5486901998519897, "learning_rate": 3.899765346537971e-05, "loss": 0.0356, "step": 59210 }, { "grad_norm": 0.3477281630039215, "learning_rate": 3.898152464634493e-05, "loss": 0.0353, "step": 59220 }, { "grad_norm": 0.34924495220184326, "learning_rate": 3.896539703227402e-05, "loss": 0.0337, "step": 59230 }, { "grad_norm": 0.43555784225463867, "learning_rate": 3.894927062493073e-05, "loss": 0.039, "step": 59240 }, { "grad_norm": 0.6044049263000488, "learning_rate": 3.893314542607859e-05, "loss": 0.0395, "step": 59250 }, { "grad_norm": 0.42849859595298767, "learning_rate": 3.891702143748103e-05, "loss": 0.0299, "step": 59260 }, { "grad_norm": 0.4931219816207886, "learning_rate": 3.890089866090133e-05, "loss": 0.0347, "step": 59270 }, { "grad_norm": 0.3673819899559021, "learning_rate": 3.888477709810267e-05, "loss": 0.0353, "step": 59280 }, { "grad_norm": 0.3561343848705292, "learning_rate": 3.886865675084806e-05, "loss": 0.0308, "step": 59290 }, { "grad_norm": 0.8282831311225891, "learning_rate": 3.885253762090043e-05, "loss": 0.034, "step": 59300 }, { "grad_norm": 0.3881959021091461, "learning_rate": 3.8836419710022534e-05, "loss": 0.0396, "step": 59310 }, { "grad_norm": 0.4955131411552429, "learning_rate": 3.882030301997698e-05, "loss": 0.0378, "step": 59320 }, { "grad_norm": 0.5837017893791199, "learning_rate": 3.880418755252629e-05, "loss": 0.0354, "step": 59330 }, { "grad_norm": 0.6451125741004944, "learning_rate": 3.87880733094328e-05, "loss": 0.0411, "step": 59340 }, { "grad_norm": 0.573147714138031, "learning_rate": 3.877196029245877e-05, "loss": 0.0303, "step": 59350 }, { "grad_norm": 0.5619629621505737, "learning_rate": 3.875584850336627e-05, "loss": 0.0424, "step": 59360 }, { "grad_norm": 0.34111258387565613, "learning_rate": 3.873973794391728e-05, "loss": 0.0315, "step": 59370 }, { "grad_norm": 0.40533390641212463, "learning_rate": 3.872362861587361e-05, "loss": 0.029, "step": 59380 }, { "grad_norm": 0.3598225712776184, "learning_rate": 3.870752052099698e-05, "loss": 0.0374, "step": 59390 }, { "grad_norm": 0.6258692741394043, "learning_rate": 3.8691413661048916e-05, "loss": 0.036, "step": 59400 }, { "grad_norm": 0.9285376071929932, "learning_rate": 3.867530803779085e-05, "loss": 0.0419, "step": 59410 }, { "grad_norm": 0.473249614238739, "learning_rate": 3.8659203652984054e-05, "loss": 0.0469, "step": 59420 }, { "grad_norm": 0.3897344470024109, "learning_rate": 3.864310050838972e-05, "loss": 0.036, "step": 59430 }, { "grad_norm": 0.32890892028808594, "learning_rate": 3.8626998605768835e-05, "loss": 0.0334, "step": 59440 }, { "grad_norm": 0.4440842866897583, "learning_rate": 3.8610897946882294e-05, "loss": 0.0463, "step": 59450 }, { "grad_norm": 0.48896709084510803, "learning_rate": 3.859479853349082e-05, "loss": 0.0432, "step": 59460 }, { "grad_norm": 0.5712453126907349, "learning_rate": 3.857870036735503e-05, "loss": 0.0376, "step": 59470 }, { "grad_norm": 0.42807018756866455, "learning_rate": 3.856260345023538e-05, "loss": 0.0449, "step": 59480 }, { "grad_norm": 0.34219396114349365, "learning_rate": 3.854650778389224e-05, "loss": 0.0306, "step": 59490 }, { "grad_norm": 0.43843287229537964, "learning_rate": 3.8530413370085786e-05, "loss": 0.0344, "step": 59500 }, { "grad_norm": 0.2947680354118347, "learning_rate": 3.851432021057608e-05, "loss": 0.0334, "step": 59510 }, { "grad_norm": 0.3257923424243927, "learning_rate": 3.849822830712306e-05, "loss": 0.0408, "step": 59520 }, { "grad_norm": 0.49753648042678833, "learning_rate": 3.848213766148649e-05, "loss": 0.0342, "step": 59530 }, { "grad_norm": 0.5400131940841675, "learning_rate": 3.846604827542602e-05, "loss": 0.031, "step": 59540 }, { "grad_norm": 0.3654334545135498, "learning_rate": 3.844996015070117e-05, "loss": 0.042, "step": 59550 }, { "grad_norm": 0.4419250786304474, "learning_rate": 3.843387328907132e-05, "loss": 0.0303, "step": 59560 }, { "grad_norm": 0.3925723433494568, "learning_rate": 3.8417787692295696e-05, "loss": 0.0482, "step": 59570 }, { "grad_norm": 0.30711138248443604, "learning_rate": 3.84017033621334e-05, "loss": 0.0357, "step": 59580 }, { "grad_norm": 0.32082387804985046, "learning_rate": 3.838562030034338e-05, "loss": 0.0294, "step": 59590 }, { "grad_norm": 0.48652634024620056, "learning_rate": 3.836953850868447e-05, "loss": 0.042, "step": 59600 }, { "grad_norm": 0.4954763948917389, "learning_rate": 3.835345798891533e-05, "loss": 0.0365, "step": 59610 }, { "grad_norm": 0.4375741183757782, "learning_rate": 3.833737874279452e-05, "loss": 0.036, "step": 59620 }, { "grad_norm": 0.43520471453666687, "learning_rate": 3.8321300772080433e-05, "loss": 0.0274, "step": 59630 }, { "grad_norm": 0.4742206931114197, "learning_rate": 3.8305224078531345e-05, "loss": 0.0367, "step": 59640 }, { "grad_norm": 0.570855438709259, "learning_rate": 3.828914866390537e-05, "loss": 0.0488, "step": 59650 }, { "grad_norm": 0.43107038736343384, "learning_rate": 3.827307452996048e-05, "loss": 0.0398, "step": 59660 }, { "grad_norm": 0.3763960301876068, "learning_rate": 3.825700167845454e-05, "loss": 0.0455, "step": 59670 }, { "grad_norm": 0.4909546375274658, "learning_rate": 3.8240930111145214e-05, "loss": 0.0446, "step": 59680 }, { "grad_norm": 0.4221886694431305, "learning_rate": 3.8224859829790116e-05, "loss": 0.0308, "step": 59690 }, { "grad_norm": 0.3651794493198395, "learning_rate": 3.820879083614664e-05, "loss": 0.045, "step": 59700 }, { "grad_norm": 0.5235683917999268, "learning_rate": 3.819272313197208e-05, "loss": 0.0479, "step": 59710 }, { "grad_norm": 0.600163459777832, "learning_rate": 3.817665671902355e-05, "loss": 0.0332, "step": 59720 }, { "grad_norm": 0.3657418191432953, "learning_rate": 3.816059159905808e-05, "loss": 0.0406, "step": 59730 }, { "grad_norm": 0.4752562940120697, "learning_rate": 3.8144527773832476e-05, "loss": 0.0498, "step": 59740 }, { "grad_norm": 0.4129181504249573, "learning_rate": 3.812846524510352e-05, "loss": 0.0287, "step": 59750 }, { "grad_norm": 0.4639974534511566, "learning_rate": 3.811240401462775e-05, "loss": 0.0455, "step": 59760 }, { "grad_norm": 0.4142265319824219, "learning_rate": 3.809634408416162e-05, "loss": 0.0363, "step": 59770 }, { "grad_norm": 0.49433693289756775, "learning_rate": 3.8080285455461394e-05, "loss": 0.0351, "step": 59780 }, { "grad_norm": 0.32350388169288635, "learning_rate": 3.806422813028323e-05, "loss": 0.0239, "step": 59790 }, { "grad_norm": 0.5347476601600647, "learning_rate": 3.8048172110383114e-05, "loss": 0.03, "step": 59800 }, { "grad_norm": 0.5176880359649658, "learning_rate": 3.8032117397516944e-05, "loss": 0.0405, "step": 59810 }, { "grad_norm": 0.5830410718917847, "learning_rate": 3.801606399344042e-05, "loss": 0.0411, "step": 59820 }, { "grad_norm": 0.24083389341831207, "learning_rate": 3.800001189990913e-05, "loss": 0.0289, "step": 59830 }, { "grad_norm": 0.5282776355743408, "learning_rate": 3.798396111867847e-05, "loss": 0.0361, "step": 59840 }, { "grad_norm": 0.6064653992652893, "learning_rate": 3.796791165150378e-05, "loss": 0.0366, "step": 59850 }, { "grad_norm": 0.46308478713035583, "learning_rate": 3.795186350014018e-05, "loss": 0.0421, "step": 59860 }, { "grad_norm": 0.29511559009552, "learning_rate": 3.793581666634266e-05, "loss": 0.036, "step": 59870 }, { "grad_norm": 0.538579523563385, "learning_rate": 3.791977115186609e-05, "loss": 0.0296, "step": 59880 }, { "grad_norm": 0.2975570559501648, "learning_rate": 3.79037269584652e-05, "loss": 0.0306, "step": 59890 }, { "grad_norm": 0.31910157203674316, "learning_rate": 3.7887684087894545e-05, "loss": 0.0366, "step": 59900 }, { "grad_norm": 0.4067416191101074, "learning_rate": 3.787164254190854e-05, "loss": 0.0337, "step": 59910 }, { "grad_norm": 0.35795050859451294, "learning_rate": 3.7855602322261473e-05, "loss": 0.0261, "step": 59920 }, { "grad_norm": 0.5497759580612183, "learning_rate": 3.783956343070746e-05, "loss": 0.044, "step": 59930 }, { "grad_norm": 0.5115810632705688, "learning_rate": 3.7823525869000526e-05, "loss": 0.0331, "step": 59940 }, { "grad_norm": 0.4679676592350006, "learning_rate": 3.78074896388945e-05, "loss": 0.033, "step": 59950 }, { "grad_norm": 0.43441042304039, "learning_rate": 3.7791454742143075e-05, "loss": 0.0357, "step": 59960 }, { "grad_norm": 0.45982617139816284, "learning_rate": 3.7775421180499795e-05, "loss": 0.0326, "step": 59970 }, { "grad_norm": 0.22211015224456787, "learning_rate": 3.7759388955718086e-05, "loss": 0.0394, "step": 59980 }, { "grad_norm": 0.33415716886520386, "learning_rate": 3.774335806955117e-05, "loss": 0.0388, "step": 59990 }, { "grad_norm": 0.9444299340248108, "learning_rate": 3.7727328523752215e-05, "loss": 0.041, "step": 60000 }, { "grad_norm": 0.4221494197845459, "learning_rate": 3.771130032007415e-05, "loss": 0.0422, "step": 60010 }, { "grad_norm": 0.34779971837997437, "learning_rate": 3.769527346026982e-05, "loss": 0.0275, "step": 60020 }, { "grad_norm": 0.35056424140930176, "learning_rate": 3.767924794609188e-05, "loss": 0.0331, "step": 60030 }, { "grad_norm": 0.4892027974128723, "learning_rate": 3.7663223779292854e-05, "loss": 0.0414, "step": 60040 }, { "grad_norm": 0.3786712884902954, "learning_rate": 3.764720096162512e-05, "loss": 0.0396, "step": 60050 }, { "grad_norm": 0.5582638382911682, "learning_rate": 3.763117949484092e-05, "loss": 0.0335, "step": 60060 }, { "grad_norm": 0.6247429847717285, "learning_rate": 3.7615159380692344e-05, "loss": 0.0407, "step": 60070 }, { "grad_norm": 0.6219003796577454, "learning_rate": 3.759914062093132e-05, "loss": 0.0278, "step": 60080 }, { "grad_norm": 0.38234537839889526, "learning_rate": 3.7583123217309615e-05, "loss": 0.0309, "step": 60090 }, { "grad_norm": 0.47783756256103516, "learning_rate": 3.7567107171578904e-05, "loss": 0.0434, "step": 60100 }, { "grad_norm": 0.7648560404777527, "learning_rate": 3.755109248549066e-05, "loss": 0.0345, "step": 60110 }, { "grad_norm": 0.4633147120475769, "learning_rate": 3.75350791607962e-05, "loss": 0.0349, "step": 60120 }, { "grad_norm": 0.7109164595603943, "learning_rate": 3.751906719924676e-05, "loss": 0.044, "step": 60130 }, { "grad_norm": 0.5144267678260803, "learning_rate": 3.750305660259337e-05, "loss": 0.0341, "step": 60140 }, { "grad_norm": 0.3486626148223877, "learning_rate": 3.748704737258693e-05, "loss": 0.0326, "step": 60150 }, { "grad_norm": 0.3171592056751251, "learning_rate": 3.747103951097816e-05, "loss": 0.0509, "step": 60160 }, { "grad_norm": 0.566907525062561, "learning_rate": 3.745503301951768e-05, "loss": 0.0366, "step": 60170 }, { "grad_norm": 0.4869747459888458, "learning_rate": 3.743902789995592e-05, "loss": 0.0378, "step": 60180 }, { "grad_norm": 0.4934009611606598, "learning_rate": 3.7423024154043183e-05, "loss": 0.032, "step": 60190 }, { "grad_norm": 0.49607113003730774, "learning_rate": 3.740702178352963e-05, "loss": 0.0279, "step": 60200 }, { "grad_norm": 0.309532105922699, "learning_rate": 3.739102079016523e-05, "loss": 0.0344, "step": 60210 }, { "grad_norm": 0.291456937789917, "learning_rate": 3.7375021175699846e-05, "loss": 0.0247, "step": 60220 }, { "grad_norm": 0.3929385542869568, "learning_rate": 3.735902294188317e-05, "loss": 0.0347, "step": 60230 }, { "grad_norm": 0.5602061748504639, "learning_rate": 3.734302609046473e-05, "loss": 0.0351, "step": 60240 }, { "grad_norm": 0.4366852939128876, "learning_rate": 3.732703062319391e-05, "loss": 0.0333, "step": 60250 }, { "grad_norm": 1.253164291381836, "learning_rate": 3.731103654181999e-05, "loss": 0.0421, "step": 60260 }, { "grad_norm": 0.43754443526268005, "learning_rate": 3.729504384809203e-05, "loss": 0.0303, "step": 60270 }, { "grad_norm": 0.49975526332855225, "learning_rate": 3.727905254375898e-05, "loss": 0.0386, "step": 60280 }, { "grad_norm": 0.4043356478214264, "learning_rate": 3.7263062630569614e-05, "loss": 0.0335, "step": 60290 }, { "grad_norm": 0.49578970670700073, "learning_rate": 3.724707411027256e-05, "loss": 0.0367, "step": 60300 }, { "grad_norm": 0.6411076784133911, "learning_rate": 3.723108698461631e-05, "loss": 0.0317, "step": 60310 }, { "grad_norm": 0.5510215163230896, "learning_rate": 3.72151012553492e-05, "loss": 0.0407, "step": 60320 }, { "grad_norm": 0.43657928705215454, "learning_rate": 3.719911692421939e-05, "loss": 0.0246, "step": 60330 }, { "grad_norm": 0.33852294087409973, "learning_rate": 3.71831339929749e-05, "loss": 0.0358, "step": 60340 }, { "grad_norm": 0.5604891777038574, "learning_rate": 3.716715246336362e-05, "loss": 0.0386, "step": 60350 }, { "grad_norm": 0.38633304834365845, "learning_rate": 3.715117233713324e-05, "loss": 0.0365, "step": 60360 }, { "grad_norm": 0.36185014247894287, "learning_rate": 3.713519361603135e-05, "loss": 0.0394, "step": 60370 }, { "grad_norm": 0.43232864141464233, "learning_rate": 3.711921630180532e-05, "loss": 0.041, "step": 60380 }, { "grad_norm": 0.5617576837539673, "learning_rate": 3.710324039620245e-05, "loss": 0.0307, "step": 60390 }, { "grad_norm": 0.4663505256175995, "learning_rate": 3.708726590096982e-05, "loss": 0.034, "step": 60400 }, { "grad_norm": 0.2660687267780304, "learning_rate": 3.707129281785437e-05, "loss": 0.0334, "step": 60410 }, { "grad_norm": 0.513825535774231, "learning_rate": 3.705532114860291e-05, "loss": 0.0344, "step": 60420 }, { "grad_norm": 0.5695502758026123, "learning_rate": 3.703935089496207e-05, "loss": 0.0575, "step": 60430 }, { "grad_norm": 0.44049936532974243, "learning_rate": 3.702338205867829e-05, "loss": 0.0344, "step": 60440 }, { "grad_norm": 0.4365310072898865, "learning_rate": 3.7007414641497976e-05, "loss": 0.0399, "step": 60450 }, { "grad_norm": 0.3482464551925659, "learning_rate": 3.699144864516726e-05, "loss": 0.0336, "step": 60460 }, { "grad_norm": 0.49017006158828735, "learning_rate": 3.6975484071432153e-05, "loss": 0.034, "step": 60470 }, { "grad_norm": 0.3645802140235901, "learning_rate": 3.695952092203854e-05, "loss": 0.0404, "step": 60480 }, { "grad_norm": 0.45570600032806396, "learning_rate": 3.69435591987321e-05, "loss": 0.0436, "step": 60490 }, { "grad_norm": 0.4540647864341736, "learning_rate": 3.6927598903258374e-05, "loss": 0.0306, "step": 60500 }, { "grad_norm": 0.423673152923584, "learning_rate": 3.691164003736281e-05, "loss": 0.026, "step": 60510 }, { "grad_norm": 0.37699002027511597, "learning_rate": 3.689568260279061e-05, "loss": 0.0425, "step": 60520 }, { "grad_norm": 0.45568886399269104, "learning_rate": 3.687972660128686e-05, "loss": 0.0304, "step": 60530 }, { "grad_norm": 0.437387079000473, "learning_rate": 3.686377203459648e-05, "loss": 0.0369, "step": 60540 }, { "grad_norm": 0.31004807353019714, "learning_rate": 3.6847818904464226e-05, "loss": 0.0367, "step": 60550 }, { "grad_norm": 0.41880539059638977, "learning_rate": 3.683186721263474e-05, "loss": 0.0401, "step": 60560 }, { "grad_norm": 0.4312590956687927, "learning_rate": 3.6815916960852435e-05, "loss": 0.0355, "step": 60570 }, { "grad_norm": 0.35199490189552307, "learning_rate": 3.679996815086165e-05, "loss": 0.0351, "step": 60580 }, { "grad_norm": 0.39667049050331116, "learning_rate": 3.678402078440649e-05, "loss": 0.0302, "step": 60590 }, { "grad_norm": 0.46586740016937256, "learning_rate": 3.676807486323096e-05, "loss": 0.0342, "step": 60600 }, { "grad_norm": 0.5230329036712646, "learning_rate": 3.6752130389078864e-05, "loss": 0.0373, "step": 60610 }, { "grad_norm": 0.4278852045536041, "learning_rate": 3.673618736369388e-05, "loss": 0.0305, "step": 60620 }, { "grad_norm": 0.3563329577445984, "learning_rate": 3.6720245788819474e-05, "loss": 0.0401, "step": 60630 }, { "grad_norm": 0.4197363257408142, "learning_rate": 3.6704305666199044e-05, "loss": 0.037, "step": 60640 }, { "grad_norm": 0.5879217982292175, "learning_rate": 3.668836699757576e-05, "loss": 0.0324, "step": 60650 }, { "grad_norm": 0.6918243765830994, "learning_rate": 3.667242978469265e-05, "loss": 0.0376, "step": 60660 }, { "grad_norm": 0.7411224842071533, "learning_rate": 3.6656494029292575e-05, "loss": 0.0425, "step": 60670 }, { "grad_norm": 0.42985019087791443, "learning_rate": 3.664055973311825e-05, "loss": 0.0367, "step": 60680 }, { "grad_norm": 0.5157648921012878, "learning_rate": 3.662462689791221e-05, "loss": 0.0329, "step": 60690 }, { "grad_norm": 0.33935967087745667, "learning_rate": 3.660869552541689e-05, "loss": 0.0368, "step": 60700 }, { "grad_norm": 0.9177976250648499, "learning_rate": 3.659276561737448e-05, "loss": 0.0364, "step": 60710 }, { "grad_norm": 0.3757515847682953, "learning_rate": 3.657683717552708e-05, "loss": 0.0394, "step": 60720 }, { "grad_norm": 0.7144133448600769, "learning_rate": 3.6560910201616574e-05, "loss": 0.0291, "step": 60730 }, { "grad_norm": 0.3684503734111786, "learning_rate": 3.654498469738472e-05, "loss": 0.0389, "step": 60740 }, { "grad_norm": 0.5540156364440918, "learning_rate": 3.6529060664573126e-05, "loss": 0.0357, "step": 60750 }, { "grad_norm": 0.25529253482818604, "learning_rate": 3.6513138104923176e-05, "loss": 0.0322, "step": 60760 }, { "grad_norm": 0.4426506459712982, "learning_rate": 3.6497217020176176e-05, "loss": 0.0307, "step": 60770 }, { "grad_norm": 0.543505847454071, "learning_rate": 3.648129741207323e-05, "loss": 0.0419, "step": 60780 }, { "grad_norm": 0.24270781874656677, "learning_rate": 3.646537928235527e-05, "loss": 0.0311, "step": 60790 }, { "grad_norm": 0.5106292963027954, "learning_rate": 3.6449462632763075e-05, "loss": 0.0359, "step": 60800 }, { "grad_norm": 0.6244081258773804, "learning_rate": 3.643354746503729e-05, "loss": 0.0397, "step": 60810 }, { "grad_norm": 0.5749831199645996, "learning_rate": 3.6417633780918345e-05, "loss": 0.0376, "step": 60820 }, { "grad_norm": 0.42614033818244934, "learning_rate": 3.640172158214655e-05, "loss": 0.0354, "step": 60830 }, { "grad_norm": 0.6260753273963928, "learning_rate": 3.6385810870462035e-05, "loss": 0.0276, "step": 60840 }, { "grad_norm": 0.3266044557094574, "learning_rate": 3.636990164760479e-05, "loss": 0.0316, "step": 60850 }, { "grad_norm": 0.4230920970439911, "learning_rate": 3.6353993915314596e-05, "loss": 0.035, "step": 60860 }, { "grad_norm": 0.3715735077857971, "learning_rate": 3.6338087675331124e-05, "loss": 0.0274, "step": 60870 }, { "grad_norm": 0.2951878309249878, "learning_rate": 3.632218292939383e-05, "loss": 0.0291, "step": 60880 }, { "grad_norm": 0.4152297377586365, "learning_rate": 3.6306279679242033e-05, "loss": 0.0345, "step": 60890 }, { "grad_norm": 0.3821801543235779, "learning_rate": 3.629037792661492e-05, "loss": 0.0434, "step": 60900 }, { "grad_norm": 0.3781338632106781, "learning_rate": 3.627447767325147e-05, "loss": 0.0342, "step": 60910 }, { "grad_norm": 0.5661157369613647, "learning_rate": 3.62585789208905e-05, "loss": 0.0307, "step": 60920 }, { "grad_norm": 0.4458519220352173, "learning_rate": 3.624268167127068e-05, "loss": 0.0486, "step": 60930 }, { "grad_norm": 0.31833598017692566, "learning_rate": 3.622678592613051e-05, "loss": 0.0356, "step": 60940 }, { "grad_norm": 0.6112306118011475, "learning_rate": 3.6210891687208304e-05, "loss": 0.0391, "step": 60950 }, { "grad_norm": 0.5242297053337097, "learning_rate": 3.619499895624228e-05, "loss": 0.0346, "step": 60960 }, { "grad_norm": 0.4453774392604828, "learning_rate": 3.61791077349704e-05, "loss": 0.0413, "step": 60970 }, { "grad_norm": 0.4953906834125519, "learning_rate": 3.616321802513053e-05, "loss": 0.0338, "step": 60980 }, { "grad_norm": 0.48139944672584534, "learning_rate": 3.6147329828460325e-05, "loss": 0.0374, "step": 60990 }, { "grad_norm": 0.513826310634613, "learning_rate": 3.6131443146697306e-05, "loss": 0.0342, "step": 61000 }, { "grad_norm": 0.49081265926361084, "learning_rate": 3.611555798157879e-05, "loss": 0.0291, "step": 61010 }, { "grad_norm": 0.601929783821106, "learning_rate": 3.6099674334842e-05, "loss": 0.0373, "step": 61020 }, { "grad_norm": 0.48683881759643555, "learning_rate": 3.6083792208223934e-05, "loss": 0.0393, "step": 61030 }, { "grad_norm": 0.4248100519180298, "learning_rate": 3.606791160346142e-05, "loss": 0.035, "step": 61040 }, { "grad_norm": 1.0553826093673706, "learning_rate": 3.6052032522291134e-05, "loss": 0.0385, "step": 61050 }, { "grad_norm": 0.6338725090026855, "learning_rate": 3.603615496644962e-05, "loss": 0.0489, "step": 61060 }, { "grad_norm": 0.5360622406005859, "learning_rate": 3.60202789376732e-05, "loss": 0.0371, "step": 61070 }, { "grad_norm": 0.45958611369132996, "learning_rate": 3.6004404437698036e-05, "loss": 0.0383, "step": 61080 }, { "grad_norm": 0.5490479469299316, "learning_rate": 3.5988531468260184e-05, "loss": 0.0335, "step": 61090 }, { "grad_norm": 0.474237859249115, "learning_rate": 3.597266003109546e-05, "loss": 0.0304, "step": 61100 }, { "grad_norm": 0.3400878310203552, "learning_rate": 3.5956790127939557e-05, "loss": 0.0341, "step": 61110 }, { "grad_norm": 0.7474790811538696, "learning_rate": 3.594092176052796e-05, "loss": 0.036, "step": 61120 }, { "grad_norm": 0.459459125995636, "learning_rate": 3.592505493059603e-05, "loss": 0.0324, "step": 61130 }, { "grad_norm": 0.359074205160141, "learning_rate": 3.5909189639878905e-05, "loss": 0.0406, "step": 61140 }, { "grad_norm": 0.31859737634658813, "learning_rate": 3.5893325890111626e-05, "loss": 0.0298, "step": 61150 }, { "grad_norm": 0.44732367992401123, "learning_rate": 3.5877463683029034e-05, "loss": 0.0424, "step": 61160 }, { "grad_norm": 0.2624373733997345, "learning_rate": 3.586160302036578e-05, "loss": 0.0375, "step": 61170 }, { "grad_norm": 0.38497787714004517, "learning_rate": 3.584574390385636e-05, "loss": 0.0309, "step": 61180 }, { "grad_norm": 0.268226683139801, "learning_rate": 3.582988633523511e-05, "loss": 0.0404, "step": 61190 }, { "grad_norm": 0.32032135128974915, "learning_rate": 3.581403031623616e-05, "loss": 0.0408, "step": 61200 }, { "grad_norm": 0.38715675473213196, "learning_rate": 3.579817584859354e-05, "loss": 0.0327, "step": 61210 }, { "grad_norm": 0.5298393368721008, "learning_rate": 3.578232293404107e-05, "loss": 0.0371, "step": 61220 }, { "grad_norm": 0.6018581986427307, "learning_rate": 3.5766471574312385e-05, "loss": 0.0318, "step": 61230 }, { "grad_norm": 0.3989261984825134, "learning_rate": 3.5750621771140964e-05, "loss": 0.0383, "step": 61240 }, { "grad_norm": 0.3652370274066925, "learning_rate": 3.5734773526260115e-05, "loss": 0.033, "step": 61250 }, { "grad_norm": 0.40128248929977417, "learning_rate": 3.5718926841402996e-05, "loss": 0.0373, "step": 61260 }, { "grad_norm": 0.32692909240722656, "learning_rate": 3.5703081718302555e-05, "loss": 0.0385, "step": 61270 }, { "grad_norm": 0.4203283190727234, "learning_rate": 3.568723815869162e-05, "loss": 0.0464, "step": 61280 }, { "grad_norm": 0.33931654691696167, "learning_rate": 3.5671396164302777e-05, "loss": 0.0546, "step": 61290 }, { "grad_norm": 0.4538567066192627, "learning_rate": 3.565555573686853e-05, "loss": 0.0321, "step": 61300 }, { "grad_norm": 0.511030912399292, "learning_rate": 3.563971687812113e-05, "loss": 0.0336, "step": 61310 }, { "grad_norm": 0.3850589692592621, "learning_rate": 3.5623879589792706e-05, "loss": 0.0345, "step": 61320 }, { "grad_norm": 0.2749774158000946, "learning_rate": 3.5608043873615185e-05, "loss": 0.0442, "step": 61330 }, { "grad_norm": 0.533682644367218, "learning_rate": 3.559220973132035e-05, "loss": 0.0296, "step": 61340 }, { "grad_norm": 0.4163014590740204, "learning_rate": 3.557637716463981e-05, "loss": 0.0337, "step": 61350 }, { "grad_norm": 0.623999297618866, "learning_rate": 3.556054617530497e-05, "loss": 0.036, "step": 61360 }, { "grad_norm": 0.4934895634651184, "learning_rate": 3.5544716765047106e-05, "loss": 0.0259, "step": 61370 }, { "grad_norm": 0.4361474812030792, "learning_rate": 3.5528888935597274e-05, "loss": 0.0359, "step": 61380 }, { "grad_norm": 0.6436002850532532, "learning_rate": 3.551306268868636e-05, "loss": 0.0328, "step": 61390 }, { "grad_norm": 0.28571373224258423, "learning_rate": 3.5497238026045166e-05, "loss": 0.034, "step": 61400 }, { "grad_norm": 0.32803264260292053, "learning_rate": 3.548141494940421e-05, "loss": 0.0308, "step": 61410 }, { "grad_norm": 0.5868821740150452, "learning_rate": 3.546559346049388e-05, "loss": 0.0315, "step": 61420 }, { "grad_norm": 0.34337446093559265, "learning_rate": 3.544977356104441e-05, "loss": 0.0332, "step": 61430 }, { "grad_norm": 0.3522690236568451, "learning_rate": 3.543395525278582e-05, "loss": 0.0344, "step": 61440 }, { "grad_norm": 0.4329942464828491, "learning_rate": 3.5418138537447985e-05, "loss": 0.0374, "step": 61450 }, { "grad_norm": 0.2644868791103363, "learning_rate": 3.5402323416760574e-05, "loss": 0.027, "step": 61460 }, { "grad_norm": 0.32194310426712036, "learning_rate": 3.538650989245316e-05, "loss": 0.0337, "step": 61470 }, { "grad_norm": 0.5974770188331604, "learning_rate": 3.537069796625504e-05, "loss": 0.0323, "step": 61480 }, { "grad_norm": 0.45177173614501953, "learning_rate": 3.53548876398954e-05, "loss": 0.0341, "step": 61490 }, { "grad_norm": 0.49459829926490784, "learning_rate": 3.5339078915103214e-05, "loss": 0.0334, "step": 61500 }, { "grad_norm": 0.48938119411468506, "learning_rate": 3.532327179360733e-05, "loss": 0.0422, "step": 61510 }, { "grad_norm": 0.3740880787372589, "learning_rate": 3.530746627713636e-05, "loss": 0.0373, "step": 61520 }, { "grad_norm": 0.35969847440719604, "learning_rate": 3.5291662367418805e-05, "loss": 0.0284, "step": 61530 }, { "grad_norm": 0.41236740350723267, "learning_rate": 3.527586006618293e-05, "loss": 0.0445, "step": 61540 }, { "grad_norm": 0.8188874125480652, "learning_rate": 3.526005937515687e-05, "loss": 0.0539, "step": 61550 }, { "grad_norm": 0.3633195161819458, "learning_rate": 3.524426029606856e-05, "loss": 0.0525, "step": 61560 }, { "grad_norm": 0.32963111996650696, "learning_rate": 3.5228462830645746e-05, "loss": 0.0392, "step": 61570 }, { "grad_norm": 0.5766000151634216, "learning_rate": 3.5212666980616036e-05, "loss": 0.0366, "step": 61580 }, { "grad_norm": 0.35565051436424255, "learning_rate": 3.519687274770681e-05, "loss": 0.0397, "step": 61590 }, { "grad_norm": 0.31364893913269043, "learning_rate": 3.518108013364535e-05, "loss": 0.0555, "step": 61600 }, { "grad_norm": 0.39079785346984863, "learning_rate": 3.516528914015868e-05, "loss": 0.0449, "step": 61610 }, { "grad_norm": 0.6645960211753845, "learning_rate": 3.5149499768973694e-05, "loss": 0.0436, "step": 61620 }, { "grad_norm": 0.16715103387832642, "learning_rate": 3.513371202181708e-05, "loss": 0.0351, "step": 61630 }, { "grad_norm": 0.6809031963348389, "learning_rate": 3.511792590041537e-05, "loss": 0.0362, "step": 61640 }, { "grad_norm": 0.42423397302627563, "learning_rate": 3.510214140649488e-05, "loss": 0.0419, "step": 61650 }, { "grad_norm": 0.34630101919174194, "learning_rate": 3.508635854178184e-05, "loss": 0.0404, "step": 61660 }, { "grad_norm": 0.6951030492782593, "learning_rate": 3.50705773080022e-05, "loss": 0.0424, "step": 61670 }, { "grad_norm": 0.3968123495578766, "learning_rate": 3.505479770688178e-05, "loss": 0.034, "step": 61680 }, { "grad_norm": 0.3618234395980835, "learning_rate": 3.503901974014622e-05, "loss": 0.0246, "step": 61690 }, { "grad_norm": 0.3555692136287689, "learning_rate": 3.502324340952095e-05, "loss": 0.0363, "step": 61700 }, { "grad_norm": 0.4526013731956482, "learning_rate": 3.500746871673126e-05, "loss": 0.0372, "step": 61710 }, { "grad_norm": 0.4220525324344635, "learning_rate": 3.499169566350227e-05, "loss": 0.0425, "step": 61720 }, { "grad_norm": 0.3704451024532318, "learning_rate": 3.4975924251558876e-05, "loss": 0.03, "step": 61730 }, { "grad_norm": 0.27648091316223145, "learning_rate": 3.496015448262582e-05, "loss": 0.0373, "step": 61740 }, { "grad_norm": 0.6804414987564087, "learning_rate": 3.494438635842764e-05, "loss": 0.0397, "step": 61750 }, { "grad_norm": 0.3198260962963104, "learning_rate": 3.492861988068874e-05, "loss": 0.0329, "step": 61760 }, { "grad_norm": 0.31373974680900574, "learning_rate": 3.491285505113332e-05, "loss": 0.0407, "step": 61770 }, { "grad_norm": 0.32299724221229553, "learning_rate": 3.4897091871485376e-05, "loss": 0.0476, "step": 61780 }, { "grad_norm": 0.3400510549545288, "learning_rate": 3.488133034346875e-05, "loss": 0.0316, "step": 61790 }, { "grad_norm": 0.7161152362823486, "learning_rate": 3.486557046880713e-05, "loss": 0.0333, "step": 61800 }, { "grad_norm": 0.5316948294639587, "learning_rate": 3.484981224922395e-05, "loss": 0.0413, "step": 61810 }, { "grad_norm": 0.4795542359352112, "learning_rate": 3.4834055686442544e-05, "loss": 0.0262, "step": 61820 }, { "grad_norm": 0.41607406735420227, "learning_rate": 3.4818300782186e-05, "loss": 0.0294, "step": 61830 }, { "grad_norm": 1.0136605501174927, "learning_rate": 3.4802547538177236e-05, "loss": 0.0346, "step": 61840 }, { "grad_norm": 0.3933970630168915, "learning_rate": 3.478679595613905e-05, "loss": 0.0334, "step": 61850 }, { "grad_norm": 0.34627512097358704, "learning_rate": 3.4771046037793975e-05, "loss": 0.0315, "step": 61860 }, { "grad_norm": 0.2486761212348938, "learning_rate": 3.4755297784864416e-05, "loss": 0.0372, "step": 61870 }, { "grad_norm": 0.41947659850120544, "learning_rate": 3.473955119907257e-05, "loss": 0.031, "step": 61880 }, { "grad_norm": 0.46883729100227356, "learning_rate": 3.472380628214046e-05, "loss": 0.0335, "step": 61890 }, { "grad_norm": 0.5127220749855042, "learning_rate": 3.4708063035789906e-05, "loss": 0.0334, "step": 61900 }, { "grad_norm": 0.44621941447257996, "learning_rate": 3.4692321461742614e-05, "loss": 0.0373, "step": 61910 }, { "grad_norm": 0.24624809622764587, "learning_rate": 3.4676581561720023e-05, "loss": 0.0305, "step": 61920 }, { "grad_norm": 1.3535611629486084, "learning_rate": 3.466084333744344e-05, "loss": 0.0366, "step": 61930 }, { "grad_norm": 0.39008787274360657, "learning_rate": 3.4645106790633964e-05, "loss": 0.039, "step": 61940 }, { "grad_norm": 0.5358604192733765, "learning_rate": 3.4629371923012526e-05, "loss": 0.0411, "step": 61950 }, { "grad_norm": 0.35706621408462524, "learning_rate": 3.4613638736299856e-05, "loss": 0.0364, "step": 61960 }, { "grad_norm": 0.45974063873291016, "learning_rate": 3.459790723221652e-05, "loss": 0.0269, "step": 61970 }, { "grad_norm": 0.7975365519523621, "learning_rate": 3.45821774124829e-05, "loss": 0.0381, "step": 61980 }, { "grad_norm": 0.3747932016849518, "learning_rate": 3.456644927881918e-05, "loss": 0.0323, "step": 61990 }, { "grad_norm": 0.47597992420196533, "learning_rate": 3.4550722832945354e-05, "loss": 0.0327, "step": 62000 }, { "grad_norm": 0.36569520831108093, "learning_rate": 3.453499807658126e-05, "loss": 0.0278, "step": 62010 }, { "grad_norm": 0.3776477575302124, "learning_rate": 3.451927501144653e-05, "loss": 0.0352, "step": 62020 }, { "grad_norm": 0.43243303894996643, "learning_rate": 3.450355363926059e-05, "loss": 0.0366, "step": 62030 }, { "grad_norm": 0.47386014461517334, "learning_rate": 3.448783396174273e-05, "loss": 0.0343, "step": 62040 }, { "grad_norm": 0.4267338514328003, "learning_rate": 3.4472115980612034e-05, "loss": 0.0365, "step": 62050 }, { "grad_norm": 0.3658028244972229, "learning_rate": 3.4456399697587396e-05, "loss": 0.0339, "step": 62060 }, { "grad_norm": 0.4786306917667389, "learning_rate": 3.4440685114387506e-05, "loss": 0.042, "step": 62070 }, { "grad_norm": 0.44578826427459717, "learning_rate": 3.442497223273091e-05, "loss": 0.0315, "step": 62080 }, { "grad_norm": 0.35425078868865967, "learning_rate": 3.4409261054335926e-05, "loss": 0.0463, "step": 62090 }, { "grad_norm": 0.3922865688800812, "learning_rate": 3.439355158092069e-05, "loss": 0.0481, "step": 62100 }, { "grad_norm": 0.7461777925491333, "learning_rate": 3.437784381420322e-05, "loss": 0.0448, "step": 62110 }, { "grad_norm": 0.43142029643058777, "learning_rate": 3.436213775590126e-05, "loss": 0.0299, "step": 62120 }, { "grad_norm": 0.28118231892585754, "learning_rate": 3.4346433407732405e-05, "loss": 0.036, "step": 62130 }, { "grad_norm": 0.4507592022418976, "learning_rate": 3.433073077141405e-05, "loss": 0.0248, "step": 62140 }, { "grad_norm": 0.8754996657371521, "learning_rate": 3.431502984866343e-05, "loss": 0.0402, "step": 62150 }, { "grad_norm": 0.4941065013408661, "learning_rate": 3.429933064119753e-05, "loss": 0.0559, "step": 62160 }, { "grad_norm": 0.5160372853279114, "learning_rate": 3.4283633150733255e-05, "loss": 0.0384, "step": 62170 }, { "grad_norm": 0.41109320521354675, "learning_rate": 3.4267937378987225e-05, "loss": 0.036, "step": 62180 }, { "grad_norm": 0.5972324013710022, "learning_rate": 3.425224332767592e-05, "loss": 0.0364, "step": 62190 }, { "grad_norm": 0.4665434658527374, "learning_rate": 3.4236550998515595e-05, "loss": 0.0444, "step": 62200 }, { "grad_norm": 0.4922274947166443, "learning_rate": 3.422086039322235e-05, "loss": 0.0354, "step": 62210 }, { "grad_norm": 0.49134817719459534, "learning_rate": 3.420517151351209e-05, "loss": 0.0341, "step": 62220 }, { "grad_norm": 0.33839669823646545, "learning_rate": 3.418948436110053e-05, "loss": 0.0388, "step": 62230 }, { "grad_norm": 0.4451746344566345, "learning_rate": 3.417379893770319e-05, "loss": 0.0435, "step": 62240 }, { "grad_norm": 0.337336003780365, "learning_rate": 3.41581152450354e-05, "loss": 0.0267, "step": 62250 }, { "grad_norm": 0.23190803825855255, "learning_rate": 3.414243328481231e-05, "loss": 0.0366, "step": 62260 }, { "grad_norm": 0.2588885426521301, "learning_rate": 3.412675305874888e-05, "loss": 0.0364, "step": 62270 }, { "grad_norm": 0.4353732168674469, "learning_rate": 3.4111074568559876e-05, "loss": 0.0297, "step": 62280 }, { "grad_norm": 0.5789223909378052, "learning_rate": 3.4095397815959846e-05, "loss": 0.0455, "step": 62290 }, { "grad_norm": 0.9292134642601013, "learning_rate": 3.407972280266322e-05, "loss": 0.0377, "step": 62300 }, { "grad_norm": 0.7777686715126038, "learning_rate": 3.4064049530384166e-05, "loss": 0.0395, "step": 62310 }, { "grad_norm": 0.5237143039703369, "learning_rate": 3.40483780008367e-05, "loss": 0.0375, "step": 62320 }, { "grad_norm": 0.36316409707069397, "learning_rate": 3.4032708215734635e-05, "loss": 0.0411, "step": 62330 }, { "grad_norm": 0.4620721936225891, "learning_rate": 3.4017040176791584e-05, "loss": 0.0386, "step": 62340 }, { "grad_norm": 1.4035141468048096, "learning_rate": 3.4001373885720976e-05, "loss": 0.0424, "step": 62350 }, { "grad_norm": 0.33273929357528687, "learning_rate": 3.398570934423608e-05, "loss": 0.0418, "step": 62360 }, { "grad_norm": 0.3670840859413147, "learning_rate": 3.397004655404994e-05, "loss": 0.0427, "step": 62370 }, { "grad_norm": 0.44344645738601685, "learning_rate": 3.395438551687541e-05, "loss": 0.0351, "step": 62380 }, { "grad_norm": 0.4892243444919586, "learning_rate": 3.393872623442515e-05, "loss": 0.0348, "step": 62390 }, { "grad_norm": 0.3676440119743347, "learning_rate": 3.392306870841164e-05, "loss": 0.0354, "step": 62400 }, { "grad_norm": 0.4431360363960266, "learning_rate": 3.390741294054715e-05, "loss": 0.0292, "step": 62410 }, { "grad_norm": 0.9193086624145508, "learning_rate": 3.38917589325438e-05, "loss": 0.0335, "step": 62420 }, { "grad_norm": 0.649553656578064, "learning_rate": 3.387610668611349e-05, "loss": 0.0512, "step": 62430 }, { "grad_norm": 0.6175101399421692, "learning_rate": 3.38604562029679e-05, "loss": 0.0406, "step": 62440 }, { "grad_norm": 0.6336150169372559, "learning_rate": 3.384480748481855e-05, "loss": 0.0407, "step": 62450 }, { "grad_norm": 0.34142404794692993, "learning_rate": 3.382916053337676e-05, "loss": 0.0343, "step": 62460 }, { "grad_norm": 0.2608465850353241, "learning_rate": 3.381351535035366e-05, "loss": 0.0381, "step": 62470 }, { "grad_norm": 0.44606223702430725, "learning_rate": 3.379787193746019e-05, "loss": 0.0313, "step": 62480 }, { "grad_norm": 0.360969603061676, "learning_rate": 3.378223029640709e-05, "loss": 0.0361, "step": 62490 }, { "grad_norm": 0.490496963262558, "learning_rate": 3.376659042890489e-05, "loss": 0.0349, "step": 62500 }, { "grad_norm": 0.2672935128211975, "learning_rate": 3.375095233666397e-05, "loss": 0.0313, "step": 62510 }, { "grad_norm": 0.2909899055957794, "learning_rate": 3.373531602139447e-05, "loss": 0.0297, "step": 62520 }, { "grad_norm": 0.49576056003570557, "learning_rate": 3.371968148480635e-05, "loss": 0.0366, "step": 62530 }, { "grad_norm": 0.4863124489784241, "learning_rate": 3.370404872860937e-05, "loss": 0.0297, "step": 62540 }, { "grad_norm": 0.27068501710891724, "learning_rate": 3.3688417754513145e-05, "loss": 0.0397, "step": 62550 }, { "grad_norm": 0.3241139054298401, "learning_rate": 3.367278856422703e-05, "loss": 0.0322, "step": 62560 }, { "grad_norm": 0.5764908194541931, "learning_rate": 3.365716115946021e-05, "loss": 0.0413, "step": 62570 }, { "grad_norm": 0.6525961756706238, "learning_rate": 3.3641535541921666e-05, "loss": 0.063, "step": 62580 }, { "grad_norm": 0.31615743041038513, "learning_rate": 3.36259117133202e-05, "loss": 0.0309, "step": 62590 }, { "grad_norm": 0.33906739950180054, "learning_rate": 3.36102896753644e-05, "loss": 0.0331, "step": 62600 }, { "grad_norm": 0.396929532289505, "learning_rate": 3.35946694297627e-05, "loss": 0.0343, "step": 62610 }, { "grad_norm": 0.3684979975223541, "learning_rate": 3.357905097822328e-05, "loss": 0.0486, "step": 62620 }, { "grad_norm": 0.337698757648468, "learning_rate": 3.356343432245416e-05, "loss": 0.0336, "step": 62630 }, { "grad_norm": 0.5147272348403931, "learning_rate": 3.354781946416315e-05, "loss": 0.0511, "step": 62640 }, { "grad_norm": 0.5943593382835388, "learning_rate": 3.3532206405057864e-05, "loss": 0.038, "step": 62650 }, { "grad_norm": 0.6020749807357788, "learning_rate": 3.351659514684573e-05, "loss": 0.0348, "step": 62660 }, { "grad_norm": 0.47274452447891235, "learning_rate": 3.350098569123395e-05, "loss": 0.0344, "step": 62670 }, { "grad_norm": 0.3348906338214874, "learning_rate": 3.348537803992958e-05, "loss": 0.0293, "step": 62680 }, { "grad_norm": 0.5145540237426758, "learning_rate": 3.346977219463945e-05, "loss": 0.037, "step": 62690 }, { "grad_norm": 0.4074576497077942, "learning_rate": 3.345416815707017e-05, "loss": 0.0337, "step": 62700 }, { "grad_norm": 0.48304617404937744, "learning_rate": 3.3438565928928185e-05, "loss": 0.0374, "step": 62710 }, { "grad_norm": 0.4401107728481293, "learning_rate": 3.3422965511919736e-05, "loss": 0.0281, "step": 62720 }, { "grad_norm": 0.3828999102115631, "learning_rate": 3.3407366907750846e-05, "loss": 0.0361, "step": 62730 }, { "grad_norm": 0.5266227722167969, "learning_rate": 3.339177011812737e-05, "loss": 0.0446, "step": 62740 }, { "grad_norm": 0.4148421585559845, "learning_rate": 3.3376175144754935e-05, "loss": 0.0327, "step": 62750 }, { "grad_norm": 0.42309796810150146, "learning_rate": 3.3360581989339004e-05, "loss": 0.0387, "step": 62760 }, { "grad_norm": 0.21491819620132446, "learning_rate": 3.3344990653584804e-05, "loss": 0.0353, "step": 62770 }, { "grad_norm": 0.2925396263599396, "learning_rate": 3.332940113919739e-05, "loss": 0.0353, "step": 62780 }, { "grad_norm": 0.4138442873954773, "learning_rate": 3.331381344788161e-05, "loss": 0.0272, "step": 62790 }, { "grad_norm": 0.44431155920028687, "learning_rate": 3.3298227581342076e-05, "loss": 0.0314, "step": 62800 }, { "grad_norm": 0.42955440282821655, "learning_rate": 3.328264354128329e-05, "loss": 0.0464, "step": 62810 }, { "grad_norm": 0.7823039889335632, "learning_rate": 3.3267061329409465e-05, "loss": 0.0393, "step": 62820 }, { "grad_norm": 0.3558126986026764, "learning_rate": 3.325148094742466e-05, "loss": 0.0302, "step": 62830 }, { "grad_norm": 1.563272476196289, "learning_rate": 3.323590239703271e-05, "loss": 0.033, "step": 62840 }, { "grad_norm": 0.38238954544067383, "learning_rate": 3.322032567993728e-05, "loss": 0.0404, "step": 62850 }, { "grad_norm": 0.38405200839042664, "learning_rate": 3.3204750797841774e-05, "loss": 0.0395, "step": 62860 }, { "grad_norm": 0.21653319895267487, "learning_rate": 3.31891777524495e-05, "loss": 0.0393, "step": 62870 }, { "grad_norm": 0.34610071778297424, "learning_rate": 3.317360654546347e-05, "loss": 0.0359, "step": 62880 }, { "grad_norm": 0.3737500309944153, "learning_rate": 3.3158037178586534e-05, "loss": 0.0361, "step": 62890 }, { "grad_norm": 0.35498809814453125, "learning_rate": 3.3142469653521316e-05, "loss": 0.0346, "step": 62900 }, { "grad_norm": 0.39831551909446716, "learning_rate": 3.3126903971970284e-05, "loss": 0.0341, "step": 62910 }, { "grad_norm": 0.43315622210502625, "learning_rate": 3.311134013563564e-05, "loss": 0.0345, "step": 62920 }, { "grad_norm": 0.4128789007663727, "learning_rate": 3.309577814621947e-05, "loss": 0.0286, "step": 62930 }, { "grad_norm": 0.29998257756233215, "learning_rate": 3.308021800542359e-05, "loss": 0.0318, "step": 62940 }, { "grad_norm": 0.3740278482437134, "learning_rate": 3.306465971494963e-05, "loss": 0.0374, "step": 62950 }, { "grad_norm": 0.5609492659568787, "learning_rate": 3.304910327649901e-05, "loss": 0.0413, "step": 62960 }, { "grad_norm": 0.3731367290019989, "learning_rate": 3.303354869177297e-05, "loss": 0.0336, "step": 62970 }, { "grad_norm": 0.4471966326236725, "learning_rate": 3.301799596247255e-05, "loss": 0.031, "step": 62980 }, { "grad_norm": 0.2807613015174866, "learning_rate": 3.300244509029854e-05, "loss": 0.0278, "step": 62990 }, { "grad_norm": 0.2945302724838257, "learning_rate": 3.2986896076951585e-05, "loss": 0.0289, "step": 63000 }, { "grad_norm": 0.45707401633262634, "learning_rate": 3.297134892413211e-05, "loss": 0.0421, "step": 63010 }, { "grad_norm": 0.3914331793785095, "learning_rate": 3.295580363354031e-05, "loss": 0.0367, "step": 63020 }, { "grad_norm": 0.4892044961452484, "learning_rate": 3.2940260206876194e-05, "loss": 0.0292, "step": 63030 }, { "grad_norm": 0.37613654136657715, "learning_rate": 3.292471864583958e-05, "loss": 0.0371, "step": 63040 }, { "grad_norm": 0.4454127550125122, "learning_rate": 3.290917895213004e-05, "loss": 0.0346, "step": 63050 }, { "grad_norm": 0.3830949068069458, "learning_rate": 3.289364112744701e-05, "loss": 0.0425, "step": 63060 }, { "grad_norm": 0.3346475064754486, "learning_rate": 3.287810517348967e-05, "loss": 0.0404, "step": 63070 }, { "grad_norm": 0.3897174894809723, "learning_rate": 3.2862571091957014e-05, "loss": 0.0364, "step": 63080 }, { "grad_norm": 0.2925833463668823, "learning_rate": 3.284703888454781e-05, "loss": 0.0326, "step": 63090 }, { "grad_norm": 0.5141297578811646, "learning_rate": 3.2831508552960646e-05, "loss": 0.0341, "step": 63100 }, { "grad_norm": 0.6633208394050598, "learning_rate": 3.281598009889387e-05, "loss": 0.0384, "step": 63110 }, { "grad_norm": 0.3606787323951721, "learning_rate": 3.28004535240457e-05, "loss": 0.0477, "step": 63120 }, { "grad_norm": 0.7204068899154663, "learning_rate": 3.278492883011407e-05, "loss": 0.0357, "step": 63130 }, { "grad_norm": 0.2883754372596741, "learning_rate": 3.276940601879675e-05, "loss": 0.0367, "step": 63140 }, { "grad_norm": 0.41865968704223633, "learning_rate": 3.275388509179127e-05, "loss": 0.0422, "step": 63150 }, { "grad_norm": 0.49025633931159973, "learning_rate": 3.2738366050794985e-05, "loss": 0.0412, "step": 63160 }, { "grad_norm": 0.41895732283592224, "learning_rate": 3.2722848897505034e-05, "loss": 0.0393, "step": 63170 }, { "grad_norm": 0.42250847816467285, "learning_rate": 3.2707333633618356e-05, "loss": 0.0278, "step": 63180 }, { "grad_norm": 0.31792017817497253, "learning_rate": 3.269182026083167e-05, "loss": 0.0314, "step": 63190 }, { "grad_norm": 0.3450860381126404, "learning_rate": 3.2676308780841516e-05, "loss": 0.035, "step": 63200 }, { "grad_norm": 0.3095358610153198, "learning_rate": 3.266079919534416e-05, "loss": 0.0335, "step": 63210 }, { "grad_norm": 0.49380213022232056, "learning_rate": 3.264529150603575e-05, "loss": 0.0309, "step": 63220 }, { "grad_norm": 0.358186811208725, "learning_rate": 3.262978571461217e-05, "loss": 0.0295, "step": 63230 }, { "grad_norm": 0.5537402629852295, "learning_rate": 3.26142818227691e-05, "loss": 0.0348, "step": 63240 }, { "grad_norm": 0.48331618309020996, "learning_rate": 3.259877983220203e-05, "loss": 0.0458, "step": 63250 }, { "grad_norm": 0.6178048253059387, "learning_rate": 3.2583279744606246e-05, "loss": 0.0346, "step": 63260 }, { "grad_norm": 0.2542179524898529, "learning_rate": 3.25677815616768e-05, "loss": 0.0277, "step": 63270 }, { "grad_norm": 0.3472786843776703, "learning_rate": 3.255228528510856e-05, "loss": 0.0444, "step": 63280 }, { "grad_norm": 0.2991025447845459, "learning_rate": 3.253679091659617e-05, "loss": 0.0271, "step": 63290 }, { "grad_norm": 0.7006130218505859, "learning_rate": 3.252129845783404e-05, "loss": 0.0291, "step": 63300 }, { "grad_norm": 0.2906831204891205, "learning_rate": 3.2505807910516456e-05, "loss": 0.0275, "step": 63310 }, { "grad_norm": 0.55418461561203, "learning_rate": 3.249031927633742e-05, "loss": 0.0496, "step": 63320 }, { "grad_norm": 0.34972310066223145, "learning_rate": 3.2474832556990737e-05, "loss": 0.0363, "step": 63330 }, { "grad_norm": 0.3550606071949005, "learning_rate": 3.245934775417002e-05, "loss": 0.0298, "step": 63340 }, { "grad_norm": 0.41917502880096436, "learning_rate": 3.244386486956866e-05, "loss": 0.037, "step": 63350 }, { "grad_norm": 0.3999388515949249, "learning_rate": 3.2428383904879846e-05, "loss": 0.0298, "step": 63360 }, { "grad_norm": 0.4217562973499298, "learning_rate": 3.241290486179652e-05, "loss": 0.049, "step": 63370 }, { "grad_norm": 0.4310065805912018, "learning_rate": 3.239742774201151e-05, "loss": 0.0264, "step": 63380 }, { "grad_norm": 0.3319675326347351, "learning_rate": 3.238195254721733e-05, "loss": 0.0316, "step": 63390 }, { "grad_norm": 0.637768566608429, "learning_rate": 3.236647927910633e-05, "loss": 0.0382, "step": 63400 }, { "grad_norm": 1.3452167510986328, "learning_rate": 3.2351007939370646e-05, "loss": 0.0391, "step": 63410 }, { "grad_norm": 0.49027177691459656, "learning_rate": 3.2335538529702184e-05, "loss": 0.0348, "step": 63420 }, { "grad_norm": 0.3349463939666748, "learning_rate": 3.232007105179268e-05, "loss": 0.0384, "step": 63430 }, { "grad_norm": 0.4129241704940796, "learning_rate": 3.230460550733363e-05, "loss": 0.0248, "step": 63440 }, { "grad_norm": 0.22699402272701263, "learning_rate": 3.2289141898016315e-05, "loss": 0.0286, "step": 63450 }, { "grad_norm": 0.3523741066455841, "learning_rate": 3.227368022553181e-05, "loss": 0.0377, "step": 63460 }, { "grad_norm": 0.5228717923164368, "learning_rate": 3.225822049157099e-05, "loss": 0.0403, "step": 63470 }, { "grad_norm": 0.6248700022697449, "learning_rate": 3.2242762697824494e-05, "loss": 0.0295, "step": 63480 }, { "grad_norm": 0.4539717733860016, "learning_rate": 3.222730684598279e-05, "loss": 0.0407, "step": 63490 }, { "grad_norm": 0.3951616585254669, "learning_rate": 3.221185293773605e-05, "loss": 0.0319, "step": 63500 }, { "grad_norm": 0.5635619163513184, "learning_rate": 3.2196400974774354e-05, "loss": 0.0298, "step": 63510 }, { "grad_norm": 0.3636937737464905, "learning_rate": 3.2180950958787485e-05, "loss": 0.0449, "step": 63520 }, { "grad_norm": 0.25264719128608704, "learning_rate": 3.2165502891465024e-05, "loss": 0.0353, "step": 63530 }, { "grad_norm": 0.7454338669776917, "learning_rate": 3.215005677449636e-05, "loss": 0.0361, "step": 63540 }, { "grad_norm": 0.3915889859199524, "learning_rate": 3.2134612609570644e-05, "loss": 0.0261, "step": 63550 }, { "grad_norm": 0.43057626485824585, "learning_rate": 3.211917039837682e-05, "loss": 0.0301, "step": 63560 }, { "grad_norm": 0.29146426916122437, "learning_rate": 3.210373014260365e-05, "loss": 0.0415, "step": 63570 }, { "grad_norm": 0.4677302837371826, "learning_rate": 3.208829184393965e-05, "loss": 0.0318, "step": 63580 }, { "grad_norm": 0.44249871373176575, "learning_rate": 3.207285550407312e-05, "loss": 0.0363, "step": 63590 }, { "grad_norm": 0.36303451657295227, "learning_rate": 3.205742112469217e-05, "loss": 0.0351, "step": 63600 }, { "grad_norm": 0.5081745386123657, "learning_rate": 3.204198870748465e-05, "loss": 0.0469, "step": 63610 }, { "grad_norm": 0.5186863541603088, "learning_rate": 3.202655825413824e-05, "loss": 0.0386, "step": 63620 }, { "grad_norm": 0.41425687074661255, "learning_rate": 3.2011129766340416e-05, "loss": 0.0437, "step": 63630 }, { "grad_norm": 0.2564457356929779, "learning_rate": 3.1995703245778384e-05, "loss": 0.0232, "step": 63640 }, { "grad_norm": 0.35225242376327515, "learning_rate": 3.1980278694139186e-05, "loss": 0.0324, "step": 63650 }, { "grad_norm": 0.44774991273880005, "learning_rate": 3.196485611310962e-05, "loss": 0.0328, "step": 63660 }, { "grad_norm": 0.5874190330505371, "learning_rate": 3.1949435504376257e-05, "loss": 0.0358, "step": 63670 }, { "grad_norm": 0.666580080986023, "learning_rate": 3.1934016869625505e-05, "loss": 0.0272, "step": 63680 }, { "grad_norm": 0.9409549236297607, "learning_rate": 3.191860021054349e-05, "loss": 0.0464, "step": 63690 }, { "grad_norm": 0.45337581634521484, "learning_rate": 3.190318552881618e-05, "loss": 0.0535, "step": 63700 }, { "grad_norm": 0.6370641589164734, "learning_rate": 3.1887772826129285e-05, "loss": 0.0313, "step": 63710 }, { "grad_norm": 0.2699822783470154, "learning_rate": 3.1872362104168335e-05, "loss": 0.0453, "step": 63720 }, { "grad_norm": 0.5612276196479797, "learning_rate": 3.185695336461861e-05, "loss": 0.0354, "step": 63730 }, { "grad_norm": 0.3766520321369171, "learning_rate": 3.184154660916519e-05, "loss": 0.0381, "step": 63740 }, { "grad_norm": 0.44796717166900635, "learning_rate": 3.1826141839492906e-05, "loss": 0.0319, "step": 63750 }, { "grad_norm": 0.42804720997810364, "learning_rate": 3.181073905728644e-05, "loss": 0.0308, "step": 63760 }, { "grad_norm": 0.5137006640434265, "learning_rate": 3.1795338264230214e-05, "loss": 0.0254, "step": 63770 }, { "grad_norm": 0.4084632396697998, "learning_rate": 3.1779939462008426e-05, "loss": 0.0323, "step": 63780 }, { "grad_norm": 0.5024541020393372, "learning_rate": 3.1764542652305066e-05, "loss": 0.0295, "step": 63790 }, { "grad_norm": 0.3618088960647583, "learning_rate": 3.17491478368039e-05, "loss": 0.0347, "step": 63800 }, { "grad_norm": 0.5867774486541748, "learning_rate": 3.173375501718847e-05, "loss": 0.0362, "step": 63810 }, { "grad_norm": 0.2941642701625824, "learning_rate": 3.171836419514215e-05, "loss": 0.0285, "step": 63820 }, { "grad_norm": 0.5584089756011963, "learning_rate": 3.170297537234803e-05, "loss": 0.0423, "step": 63830 }, { "grad_norm": 0.25259867310523987, "learning_rate": 3.168758855048901e-05, "loss": 0.0311, "step": 63840 }, { "grad_norm": 0.31222906708717346, "learning_rate": 3.1672203731247785e-05, "loss": 0.0271, "step": 63850 }, { "grad_norm": 0.4175209701061249, "learning_rate": 3.16568209163068e-05, "loss": 0.0363, "step": 63860 }, { "grad_norm": 0.4928315579891205, "learning_rate": 3.1641440107348295e-05, "loss": 0.0386, "step": 63870 }, { "grad_norm": 0.36814063787460327, "learning_rate": 3.162606130605429e-05, "loss": 0.031, "step": 63880 }, { "grad_norm": 0.35844704508781433, "learning_rate": 3.16106845141066e-05, "loss": 0.0255, "step": 63890 }, { "grad_norm": 0.4880867600440979, "learning_rate": 3.159530973318682e-05, "loss": 0.0287, "step": 63900 }, { "grad_norm": 0.36011096835136414, "learning_rate": 3.1579936964976286e-05, "loss": 0.0334, "step": 63910 }, { "grad_norm": 0.43495574593544006, "learning_rate": 3.156456621115615e-05, "loss": 0.0341, "step": 63920 }, { "grad_norm": 0.7470972537994385, "learning_rate": 3.154919747340734e-05, "loss": 0.0381, "step": 63930 }, { "grad_norm": 0.6409381628036499, "learning_rate": 3.153383075341054e-05, "loss": 0.0283, "step": 63940 }, { "grad_norm": 0.34403830766677856, "learning_rate": 3.1518466052846264e-05, "loss": 0.0318, "step": 63950 }, { "grad_norm": 0.382510781288147, "learning_rate": 3.150310337339474e-05, "loss": 0.0301, "step": 63960 }, { "grad_norm": 0.33180752396583557, "learning_rate": 3.148774271673603e-05, "loss": 0.0317, "step": 63970 }, { "grad_norm": 0.5764511823654175, "learning_rate": 3.147238408454995e-05, "loss": 0.0415, "step": 63980 }, { "grad_norm": 0.25966230034828186, "learning_rate": 3.1457027478516086e-05, "loss": 0.0264, "step": 63990 }, { "grad_norm": 0.2811738848686218, "learning_rate": 3.144167290031382e-05, "loss": 0.0346, "step": 64000 }, { "grad_norm": 0.5347445011138916, "learning_rate": 3.142632035162227e-05, "loss": 0.0403, "step": 64010 }, { "grad_norm": 0.41646334528923035, "learning_rate": 3.141096983412044e-05, "loss": 0.0286, "step": 64020 }, { "grad_norm": 0.2851617634296417, "learning_rate": 3.139562134948699e-05, "loss": 0.0298, "step": 64030 }, { "grad_norm": 0.4787053167819977, "learning_rate": 3.1380274899400424e-05, "loss": 0.0319, "step": 64040 }, { "grad_norm": 0.4510607123374939, "learning_rate": 3.1364930485539e-05, "loss": 0.0252, "step": 64050 }, { "grad_norm": 0.25899773836135864, "learning_rate": 3.134958810958076e-05, "loss": 0.0293, "step": 64060 }, { "grad_norm": 0.3415001928806305, "learning_rate": 3.13342477732035e-05, "loss": 0.0339, "step": 64070 }, { "grad_norm": 0.34929022192955017, "learning_rate": 3.131890947808487e-05, "loss": 0.0276, "step": 64080 }, { "grad_norm": 0.417582243680954, "learning_rate": 3.13035732259022e-05, "loss": 0.0336, "step": 64090 }, { "grad_norm": 0.32502320408821106, "learning_rate": 3.1288239018332656e-05, "loss": 0.0308, "step": 64100 }, { "grad_norm": 0.2911205291748047, "learning_rate": 3.1272906857053164e-05, "loss": 0.027, "step": 64110 }, { "grad_norm": 0.4062340557575226, "learning_rate": 3.125757674374042e-05, "loss": 0.0289, "step": 64120 }, { "grad_norm": 0.3559616506099701, "learning_rate": 3.124224868007088e-05, "loss": 0.0411, "step": 64130 }, { "grad_norm": 0.39115601778030396, "learning_rate": 3.1226922667720854e-05, "loss": 0.0336, "step": 64140 }, { "grad_norm": 0.5773534178733826, "learning_rate": 3.121159870836633e-05, "loss": 0.0369, "step": 64150 }, { "grad_norm": 0.3005763590335846, "learning_rate": 3.119627680368312e-05, "loss": 0.0334, "step": 64160 }, { "grad_norm": 0.47553345561027527, "learning_rate": 3.1180956955346816e-05, "loss": 0.0351, "step": 64170 }, { "grad_norm": 0.3163488209247589, "learning_rate": 3.116563916503276e-05, "loss": 0.035, "step": 64180 }, { "grad_norm": 0.20904433727264404, "learning_rate": 3.1150323434416095e-05, "loss": 0.0428, "step": 64190 }, { "grad_norm": 0.3955534100532532, "learning_rate": 3.1135009765171696e-05, "loss": 0.0326, "step": 64200 }, { "grad_norm": 0.44176241755485535, "learning_rate": 3.111969815897429e-05, "loss": 0.0287, "step": 64210 }, { "grad_norm": 0.34046489000320435, "learning_rate": 3.1104388617498295e-05, "loss": 0.0386, "step": 64220 }, { "grad_norm": 0.40348735451698303, "learning_rate": 3.1089081142417966e-05, "loss": 0.0358, "step": 64230 }, { "grad_norm": 0.3836784362792969, "learning_rate": 3.1073775735407285e-05, "loss": 0.0302, "step": 64240 }, { "grad_norm": 0.5293631553649902, "learning_rate": 3.1058472398140035e-05, "loss": 0.0368, "step": 64250 }, { "grad_norm": 0.5112830400466919, "learning_rate": 3.1043171132289736e-05, "loss": 0.0316, "step": 64260 }, { "grad_norm": 0.3187989592552185, "learning_rate": 3.1027871939529774e-05, "loss": 0.0246, "step": 64270 }, { "grad_norm": 0.5954037308692932, "learning_rate": 3.1012574821533214e-05, "loss": 0.0515, "step": 64280 }, { "grad_norm": 0.623009443283081, "learning_rate": 3.099727977997291e-05, "loss": 0.0405, "step": 64290 }, { "grad_norm": 0.2700721025466919, "learning_rate": 3.098198681652154e-05, "loss": 0.0447, "step": 64300 }, { "grad_norm": 0.2921293079853058, "learning_rate": 3.0966695932851475e-05, "loss": 0.0249, "step": 64310 }, { "grad_norm": 0.47037628293037415, "learning_rate": 3.0951407130634926e-05, "loss": 0.0389, "step": 64320 }, { "grad_norm": 0.4238707721233368, "learning_rate": 3.0936120411543864e-05, "loss": 0.032, "step": 64330 }, { "grad_norm": 0.46124500036239624, "learning_rate": 3.0920835777250015e-05, "loss": 0.0346, "step": 64340 }, { "grad_norm": 0.605132520198822, "learning_rate": 3.090555322942488e-05, "loss": 0.0285, "step": 64350 }, { "grad_norm": 0.31774213910102844, "learning_rate": 3.089027276973973e-05, "loss": 0.0338, "step": 64360 }, { "grad_norm": 0.29848483204841614, "learning_rate": 3.0874994399865615e-05, "loss": 0.0327, "step": 64370 }, { "grad_norm": 0.4400979280471802, "learning_rate": 3.085971812147337e-05, "loss": 0.0344, "step": 64380 }, { "grad_norm": 0.35354143381118774, "learning_rate": 3.0844443936233557e-05, "loss": 0.0398, "step": 64390 }, { "grad_norm": 0.4067327678203583, "learning_rate": 3.082917184581657e-05, "loss": 0.0275, "step": 64400 }, { "grad_norm": 0.4233700633049011, "learning_rate": 3.0813901851892513e-05, "loss": 0.0327, "step": 64410 }, { "grad_norm": 0.2391459196805954, "learning_rate": 3.0798633956131316e-05, "loss": 0.0313, "step": 64420 }, { "grad_norm": 0.17335541546344757, "learning_rate": 3.078336816020264e-05, "loss": 0.029, "step": 64430 }, { "grad_norm": 0.40630045533180237, "learning_rate": 3.0768104465775935e-05, "loss": 0.0326, "step": 64440 }, { "grad_norm": 0.48973000049591064, "learning_rate": 3.0752842874520396e-05, "loss": 0.0296, "step": 64450 }, { "grad_norm": 0.5033080577850342, "learning_rate": 3.073758338810503e-05, "loss": 0.0311, "step": 64460 }, { "grad_norm": 0.46461576223373413, "learning_rate": 3.072232600819859e-05, "loss": 0.0349, "step": 64470 }, { "grad_norm": 0.3205452263355255, "learning_rate": 3.070707073646959e-05, "loss": 0.0411, "step": 64480 }, { "grad_norm": 0.3723786175251007, "learning_rate": 3.069181757458633e-05, "loss": 0.0379, "step": 64490 }, { "grad_norm": 0.5113110542297363, "learning_rate": 3.067656652421687e-05, "loss": 0.0315, "step": 64500 }, { "grad_norm": 0.26034143567085266, "learning_rate": 3.0661317587029025e-05, "loss": 0.0314, "step": 64510 }, { "grad_norm": 0.5203694701194763, "learning_rate": 3.064607076469043e-05, "loss": 0.0237, "step": 64520 }, { "grad_norm": 0.5237347483634949, "learning_rate": 3.063082605886845e-05, "loss": 0.0432, "step": 64530 }, { "grad_norm": 0.37284231185913086, "learning_rate": 3.061558347123019e-05, "loss": 0.041, "step": 64540 }, { "grad_norm": 0.38436266779899597, "learning_rate": 3.060034300344259e-05, "loss": 0.0325, "step": 64550 }, { "grad_norm": 1.3417574167251587, "learning_rate": 3.058510465717231e-05, "loss": 0.0421, "step": 64560 }, { "grad_norm": 0.32794857025146484, "learning_rate": 3.0569868434085804e-05, "loss": 0.0269, "step": 64570 }, { "grad_norm": 0.3381042778491974, "learning_rate": 3.055463433584924e-05, "loss": 0.0253, "step": 64580 }, { "grad_norm": 0.4327017366886139, "learning_rate": 3.0539402364128656e-05, "loss": 0.0305, "step": 64590 }, { "grad_norm": 0.31404033303260803, "learning_rate": 3.052417252058977e-05, "loss": 0.0429, "step": 64600 }, { "grad_norm": 0.4858573377132416, "learning_rate": 3.0508944806898092e-05, "loss": 0.0416, "step": 64610 }, { "grad_norm": 0.42257705330848694, "learning_rate": 3.04937192247189e-05, "loss": 0.0367, "step": 64620 }, { "grad_norm": 0.42203712463378906, "learning_rate": 3.047849577571725e-05, "loss": 0.0409, "step": 64630 }, { "grad_norm": 0.4102749526500702, "learning_rate": 3.046327446155795e-05, "loss": 0.0285, "step": 64640 }, { "grad_norm": 0.3387761116027832, "learning_rate": 3.0448055283905587e-05, "loss": 0.0271, "step": 64650 }, { "grad_norm": 0.29807525873184204, "learning_rate": 3.0432838244424494e-05, "loss": 0.0288, "step": 64660 }, { "grad_norm": 0.3321416974067688, "learning_rate": 3.04176233447788e-05, "loss": 0.0259, "step": 64670 }, { "grad_norm": 0.2807154357433319, "learning_rate": 3.040241058663238e-05, "loss": 0.0254, "step": 64680 }, { "grad_norm": 0.5540557503700256, "learning_rate": 3.038719997164887e-05, "loss": 0.0386, "step": 64690 }, { "grad_norm": 0.3195435404777527, "learning_rate": 3.0371991501491683e-05, "loss": 0.0363, "step": 64700 }, { "grad_norm": 0.4052387773990631, "learning_rate": 3.0356785177823975e-05, "loss": 0.0356, "step": 64710 }, { "grad_norm": 0.23063044250011444, "learning_rate": 3.0341581002308726e-05, "loss": 0.0327, "step": 64720 }, { "grad_norm": 0.4969042241573334, "learning_rate": 3.0326378976608627e-05, "loss": 0.0298, "step": 64730 }, { "grad_norm": 0.466804563999176, "learning_rate": 3.0311179102386146e-05, "loss": 0.0465, "step": 64740 }, { "grad_norm": 0.6146173477172852, "learning_rate": 3.0295981381303513e-05, "loss": 0.0451, "step": 64750 }, { "grad_norm": 0.4035640358924866, "learning_rate": 3.0280785815022727e-05, "loss": 0.031, "step": 64760 }, { "grad_norm": 0.3828338086605072, "learning_rate": 3.0265592405205544e-05, "loss": 0.0338, "step": 64770 }, { "grad_norm": 0.4780218005180359, "learning_rate": 3.0250401153513517e-05, "loss": 0.039, "step": 64780 }, { "grad_norm": 1.6659313440322876, "learning_rate": 3.0235212061607936e-05, "loss": 0.0473, "step": 64790 }, { "grad_norm": 0.22866640985012054, "learning_rate": 3.0220025131149842e-05, "loss": 0.0323, "step": 64800 }, { "grad_norm": 0.28119587898254395, "learning_rate": 3.020484036380007e-05, "loss": 0.0272, "step": 64810 }, { "grad_norm": 0.3173642158508301, "learning_rate": 3.0189657761219188e-05, "loss": 0.0361, "step": 64820 }, { "grad_norm": 0.2303028255701065, "learning_rate": 3.017447732506753e-05, "loss": 0.0338, "step": 64830 }, { "grad_norm": 0.5670080184936523, "learning_rate": 3.015929905700524e-05, "loss": 0.0281, "step": 64840 }, { "grad_norm": 0.5865874886512756, "learning_rate": 3.014412295869218e-05, "loss": 0.0316, "step": 64850 }, { "grad_norm": 0.5212361216545105, "learning_rate": 3.0128949031787977e-05, "loss": 0.0378, "step": 64860 }, { "grad_norm": 0.23642690479755402, "learning_rate": 3.0113777277952022e-05, "loss": 0.0445, "step": 64870 }, { "grad_norm": 0.5996255874633789, "learning_rate": 3.0098607698843496e-05, "loss": 0.0353, "step": 64880 }, { "grad_norm": 0.9334957003593445, "learning_rate": 3.0083440296121312e-05, "loss": 0.0382, "step": 64890 }, { "grad_norm": 0.32597655057907104, "learning_rate": 3.0068275071444142e-05, "loss": 0.0251, "step": 64900 }, { "grad_norm": 0.3809918463230133, "learning_rate": 3.0053112026470442e-05, "loss": 0.0391, "step": 64910 }, { "grad_norm": 0.48543331027030945, "learning_rate": 3.0037951162858434e-05, "loss": 0.0355, "step": 64920 }, { "grad_norm": 0.343077152967453, "learning_rate": 3.002279248226607e-05, "loss": 0.0331, "step": 64930 }, { "grad_norm": 0.339309960603714, "learning_rate": 3.000763598635109e-05, "loss": 0.0355, "step": 64940 }, { "grad_norm": 0.43760135769844055, "learning_rate": 2.999248167677098e-05, "loss": 0.0343, "step": 64950 }, { "grad_norm": 0.35077592730522156, "learning_rate": 2.9977329555182976e-05, "loss": 0.0428, "step": 64960 }, { "grad_norm": 0.31689584255218506, "learning_rate": 2.9962179623244124e-05, "loss": 0.0337, "step": 64970 }, { "grad_norm": 0.400683730840683, "learning_rate": 2.9947031882611175e-05, "loss": 0.0304, "step": 64980 }, { "grad_norm": 0.46115541458129883, "learning_rate": 2.9931886334940685e-05, "loss": 0.0314, "step": 64990 }, { "grad_norm": 0.2593476474285126, "learning_rate": 2.9916742981888923e-05, "loss": 0.0289, "step": 65000 }, { "grad_norm": 0.49389809370040894, "learning_rate": 2.990160182511196e-05, "loss": 0.0424, "step": 65010 }, { "grad_norm": 0.4038500189781189, "learning_rate": 2.9886462866265574e-05, "loss": 0.0333, "step": 65020 }, { "grad_norm": 0.40082091093063354, "learning_rate": 2.9871326107005392e-05, "loss": 0.0318, "step": 65030 }, { "grad_norm": 0.40432292222976685, "learning_rate": 2.9856191548986735e-05, "loss": 0.0363, "step": 65040 }, { "grad_norm": 0.3396952450275421, "learning_rate": 2.9841059193864672e-05, "loss": 0.0319, "step": 65050 }, { "grad_norm": 0.387368768453598, "learning_rate": 2.982592904329407e-05, "loss": 0.0303, "step": 65060 }, { "grad_norm": 0.18159574270248413, "learning_rate": 2.981080109892954e-05, "loss": 0.0249, "step": 65070 }, { "grad_norm": 0.24828140437602997, "learning_rate": 2.9795675362425434e-05, "loss": 0.0273, "step": 65080 }, { "grad_norm": 0.36824387311935425, "learning_rate": 2.97805518354359e-05, "loss": 0.0276, "step": 65090 }, { "grad_norm": 0.3819996416568756, "learning_rate": 2.9765430519614825e-05, "loss": 0.0277, "step": 65100 }, { "grad_norm": 0.5570974946022034, "learning_rate": 2.975031141661584e-05, "loss": 0.0337, "step": 65110 }, { "grad_norm": 0.362484335899353, "learning_rate": 2.9735194528092347e-05, "loss": 0.0304, "step": 65120 }, { "grad_norm": 0.375328004360199, "learning_rate": 2.9720079855697525e-05, "loss": 0.0403, "step": 65130 }, { "grad_norm": 0.5525858998298645, "learning_rate": 2.970496740108427e-05, "loss": 0.042, "step": 65140 }, { "grad_norm": 0.3572808802127838, "learning_rate": 2.968985716590527e-05, "loss": 0.0279, "step": 65150 }, { "grad_norm": 0.45639386773109436, "learning_rate": 2.9674749151812953e-05, "loss": 0.0277, "step": 65160 }, { "grad_norm": 0.48208439350128174, "learning_rate": 2.965964336045951e-05, "loss": 0.0337, "step": 65170 }, { "grad_norm": 0.4263366162776947, "learning_rate": 2.9644539793496894e-05, "loss": 0.0313, "step": 65180 }, { "grad_norm": 0.4212328791618347, "learning_rate": 2.9629438452576807e-05, "loss": 0.0389, "step": 65190 }, { "grad_norm": 0.419570654630661, "learning_rate": 2.9614339339350698e-05, "loss": 0.0314, "step": 65200 }, { "grad_norm": 0.4233689606189728, "learning_rate": 2.959924245546979e-05, "loss": 0.0279, "step": 65210 }, { "grad_norm": 0.4178139567375183, "learning_rate": 2.9584147802585038e-05, "loss": 0.0408, "step": 65220 }, { "grad_norm": 0.46093544363975525, "learning_rate": 2.956905538234721e-05, "loss": 0.037, "step": 65230 }, { "grad_norm": 0.3197627663612366, "learning_rate": 2.9553965196406764e-05, "loss": 0.0284, "step": 65240 }, { "grad_norm": 0.3470955789089203, "learning_rate": 2.9538877246413943e-05, "loss": 0.0349, "step": 65250 }, { "grad_norm": 0.4392237663269043, "learning_rate": 2.952379153401874e-05, "loss": 0.0356, "step": 65260 }, { "grad_norm": 1.1580111980438232, "learning_rate": 2.9508708060870914e-05, "loss": 0.0344, "step": 65270 }, { "grad_norm": 0.3855421245098114, "learning_rate": 2.9493626828619948e-05, "loss": 0.0323, "step": 65280 }, { "grad_norm": 0.35809019207954407, "learning_rate": 2.9478547838915137e-05, "loss": 0.0288, "step": 65290 }, { "grad_norm": 0.42641136050224304, "learning_rate": 2.946347109340548e-05, "loss": 0.0362, "step": 65300 }, { "grad_norm": 0.542538583278656, "learning_rate": 2.944839659373976e-05, "loss": 0.0318, "step": 65310 }, { "grad_norm": 0.5956649780273438, "learning_rate": 2.943332434156647e-05, "loss": 0.0347, "step": 65320 }, { "grad_norm": 0.45229217410087585, "learning_rate": 2.9418254338533912e-05, "loss": 0.0397, "step": 65330 }, { "grad_norm": 0.40261220932006836, "learning_rate": 2.940318658629011e-05, "loss": 0.0335, "step": 65340 }, { "grad_norm": 0.40330469608306885, "learning_rate": 2.938812108648287e-05, "loss": 0.0395, "step": 65350 }, { "grad_norm": 0.3072671592235565, "learning_rate": 2.937305784075972e-05, "loss": 0.0292, "step": 65360 }, { "grad_norm": 0.5377143621444702, "learning_rate": 2.9357996850767937e-05, "loss": 0.0286, "step": 65370 }, { "grad_norm": 0.43114298582077026, "learning_rate": 2.9342938118154595e-05, "loss": 0.027, "step": 65380 }, { "grad_norm": 0.32694146037101746, "learning_rate": 2.9327881644566485e-05, "loss": 0.0458, "step": 65390 }, { "grad_norm": 0.4441976547241211, "learning_rate": 2.9312827431650157e-05, "loss": 0.0419, "step": 65400 }, { "grad_norm": 0.27354946732521057, "learning_rate": 2.92977754810519e-05, "loss": 0.0426, "step": 65410 }, { "grad_norm": 0.36466675996780396, "learning_rate": 2.928272579441781e-05, "loss": 0.036, "step": 65420 }, { "grad_norm": 0.3627941906452179, "learning_rate": 2.9267678373393676e-05, "loss": 0.0343, "step": 65430 }, { "grad_norm": 0.2707851529121399, "learning_rate": 2.925263321962507e-05, "loss": 0.0351, "step": 65440 }, { "grad_norm": 1.0240628719329834, "learning_rate": 2.9237590334757292e-05, "loss": 0.0349, "step": 65450 }, { "grad_norm": 0.44666799902915955, "learning_rate": 2.922254972043543e-05, "loss": 0.0359, "step": 65460 }, { "grad_norm": 0.32693496346473694, "learning_rate": 2.9207511378304265e-05, "loss": 0.0319, "step": 65470 }, { "grad_norm": 0.5098916292190552, "learning_rate": 2.9192475310008428e-05, "loss": 0.0451, "step": 65480 }, { "grad_norm": 0.830379843711853, "learning_rate": 2.917744151719219e-05, "loss": 0.0272, "step": 65490 }, { "grad_norm": 0.5850321054458618, "learning_rate": 2.916241000149964e-05, "loss": 0.0328, "step": 65500 }, { "grad_norm": 0.34762275218963623, "learning_rate": 2.9147380764574622e-05, "loss": 0.0292, "step": 65510 }, { "grad_norm": 0.46044662594795227, "learning_rate": 2.9132353808060674e-05, "loss": 0.0433, "step": 65520 }, { "grad_norm": 0.35199177265167236, "learning_rate": 2.9117329133601134e-05, "loss": 0.0317, "step": 65530 }, { "grad_norm": 0.3717697858810425, "learning_rate": 2.9102306742839093e-05, "loss": 0.045, "step": 65540 }, { "grad_norm": 0.37603309750556946, "learning_rate": 2.9087286637417375e-05, "loss": 0.0323, "step": 65550 }, { "grad_norm": 0.49357718229293823, "learning_rate": 2.9072268818978537e-05, "loss": 0.0371, "step": 65560 }, { "grad_norm": 0.4309873878955841, "learning_rate": 2.905725328916493e-05, "loss": 0.0299, "step": 65570 }, { "grad_norm": 0.6462588906288147, "learning_rate": 2.9042240049618608e-05, "loss": 0.0403, "step": 65580 }, { "grad_norm": 0.5203374624252319, "learning_rate": 2.9027229101981413e-05, "loss": 0.0325, "step": 65590 }, { "grad_norm": 0.3648344576358795, "learning_rate": 2.901222044789488e-05, "loss": 0.0371, "step": 65600 }, { "grad_norm": 0.607903242111206, "learning_rate": 2.89972140890004e-05, "loss": 0.036, "step": 65610 }, { "grad_norm": 0.46063655614852905, "learning_rate": 2.8982210026938988e-05, "loss": 0.0315, "step": 65620 }, { "grad_norm": 0.5683043599128723, "learning_rate": 2.896720826335151e-05, "loss": 0.0493, "step": 65630 }, { "grad_norm": 0.6257821321487427, "learning_rate": 2.8952208799878493e-05, "loss": 0.0373, "step": 65640 }, { "grad_norm": 0.4080794155597687, "learning_rate": 2.89372116381603e-05, "loss": 0.0258, "step": 65650 }, { "grad_norm": 0.22735555469989777, "learning_rate": 2.8922216779836932e-05, "loss": 0.0321, "step": 65660 }, { "grad_norm": 0.6282811164855957, "learning_rate": 2.8907224226548285e-05, "loss": 0.0258, "step": 65670 }, { "grad_norm": 0.5349422097206116, "learning_rate": 2.8892233979933868e-05, "loss": 0.0352, "step": 65680 }, { "grad_norm": 0.6065640449523926, "learning_rate": 2.8877246041633026e-05, "loss": 0.0387, "step": 65690 }, { "grad_norm": 0.48007678985595703, "learning_rate": 2.886226041328478e-05, "loss": 0.0255, "step": 65700 }, { "grad_norm": 0.26596859097480774, "learning_rate": 2.8847277096527957e-05, "loss": 0.0306, "step": 65710 }, { "grad_norm": 0.3986119031906128, "learning_rate": 2.8832296093001105e-05, "loss": 0.0249, "step": 65720 }, { "grad_norm": 0.3435240387916565, "learning_rate": 2.8817317404342547e-05, "loss": 0.0319, "step": 65730 }, { "grad_norm": 0.3881832957267761, "learning_rate": 2.880234103219029e-05, "loss": 0.0375, "step": 65740 }, { "grad_norm": 0.3866982161998749, "learning_rate": 2.878736697818215e-05, "loss": 0.0328, "step": 65750 }, { "grad_norm": 0.9032342433929443, "learning_rate": 2.877239524395568e-05, "loss": 0.0342, "step": 65760 }, { "grad_norm": 0.2519402503967285, "learning_rate": 2.8757425831148134e-05, "loss": 0.0323, "step": 65770 }, { "grad_norm": 0.4218209981918335, "learning_rate": 2.874245874139658e-05, "loss": 0.0327, "step": 65780 }, { "grad_norm": 0.24063052237033844, "learning_rate": 2.8727493976337737e-05, "loss": 0.0265, "step": 65790 }, { "grad_norm": 0.6111605167388916, "learning_rate": 2.8712531537608218e-05, "loss": 0.0301, "step": 65800 }, { "grad_norm": 0.4546998143196106, "learning_rate": 2.8697571426844217e-05, "loss": 0.042, "step": 65810 }, { "grad_norm": 0.42784181237220764, "learning_rate": 2.86826136456818e-05, "loss": 0.0324, "step": 65820 }, { "grad_norm": 0.44950348138809204, "learning_rate": 2.8667658195756685e-05, "loss": 0.0344, "step": 65830 }, { "grad_norm": 0.3899919092655182, "learning_rate": 2.8652705078704412e-05, "loss": 0.0334, "step": 65840 }, { "grad_norm": 0.32189786434173584, "learning_rate": 2.863775429616018e-05, "loss": 0.036, "step": 65850 }, { "grad_norm": 0.8438621759414673, "learning_rate": 2.862280584975906e-05, "loss": 0.0379, "step": 65860 }, { "grad_norm": 0.631893515586853, "learning_rate": 2.8607859741135735e-05, "loss": 0.0408, "step": 65870 }, { "grad_norm": 0.35017693042755127, "learning_rate": 2.8592915971924723e-05, "loss": 0.046, "step": 65880 }, { "grad_norm": 0.32119160890579224, "learning_rate": 2.8577974543760212e-05, "loss": 0.0321, "step": 65890 }, { "grad_norm": 0.5244695544242859, "learning_rate": 2.8563035458276222e-05, "loss": 0.0356, "step": 65900 }, { "grad_norm": 0.21728022396564484, "learning_rate": 2.8548098717106424e-05, "loss": 0.0344, "step": 65910 }, { "grad_norm": 0.45042887330055237, "learning_rate": 2.85331643218843e-05, "loss": 0.0307, "step": 65920 }, { "grad_norm": 0.3875656723976135, "learning_rate": 2.8518232274243046e-05, "loss": 0.0383, "step": 65930 }, { "grad_norm": 0.485730916261673, "learning_rate": 2.8503302575815638e-05, "loss": 0.0472, "step": 65940 }, { "grad_norm": 0.514458954334259, "learning_rate": 2.8488375228234716e-05, "loss": 0.0382, "step": 65950 }, { "grad_norm": 0.4266414940357208, "learning_rate": 2.8473450233132738e-05, "loss": 0.0407, "step": 65960 }, { "grad_norm": 0.3211320638656616, "learning_rate": 2.8458527592141898e-05, "loss": 0.0359, "step": 65970 }, { "grad_norm": 0.34165045619010925, "learning_rate": 2.8443607306894073e-05, "loss": 0.0398, "step": 65980 }, { "grad_norm": 0.2545975148677826, "learning_rate": 2.8428689379020946e-05, "loss": 0.0327, "step": 65990 }, { "grad_norm": 0.39346471428871155, "learning_rate": 2.841377381015391e-05, "loss": 0.0332, "step": 66000 }, { "grad_norm": 0.3357897698879242, "learning_rate": 2.8398860601924144e-05, "loss": 0.0262, "step": 66010 }, { "grad_norm": 0.3220025897026062, "learning_rate": 2.838394975596248e-05, "loss": 0.0323, "step": 66020 }, { "grad_norm": 0.45957890152931213, "learning_rate": 2.83690412738996e-05, "loss": 0.0371, "step": 66030 }, { "grad_norm": 0.32862964272499084, "learning_rate": 2.835413515736579e-05, "loss": 0.0389, "step": 66040 }, { "grad_norm": 0.4483116567134857, "learning_rate": 2.8339231407991267e-05, "loss": 0.0341, "step": 66050 }, { "grad_norm": 0.987456738948822, "learning_rate": 2.832433002740581e-05, "loss": 0.0302, "step": 66060 }, { "grad_norm": 0.3411128520965576, "learning_rate": 2.8309431017239047e-05, "loss": 0.0289, "step": 66070 }, { "grad_norm": 0.3777001202106476, "learning_rate": 2.829453437912028e-05, "loss": 0.0359, "step": 66080 }, { "grad_norm": 0.6318925619125366, "learning_rate": 2.8279640114678617e-05, "loss": 0.034, "step": 66090 }, { "grad_norm": 0.6357904672622681, "learning_rate": 2.8264748225542836e-05, "loss": 0.0395, "step": 66100 }, { "grad_norm": 0.2558440864086151, "learning_rate": 2.8249858713341516e-05, "loss": 0.0292, "step": 66110 }, { "grad_norm": 0.3242788016796112, "learning_rate": 2.823497157970294e-05, "loss": 0.0337, "step": 66120 }, { "grad_norm": 0.5962123870849609, "learning_rate": 2.8220086826255164e-05, "loss": 0.029, "step": 66130 }, { "grad_norm": 0.5395588278770447, "learning_rate": 2.8205204454625934e-05, "loss": 0.0405, "step": 66140 }, { "grad_norm": 0.45704343914985657, "learning_rate": 2.8190324466442798e-05, "loss": 0.0335, "step": 66150 }, { "grad_norm": 0.5065634846687317, "learning_rate": 2.8175446863332965e-05, "loss": 0.0298, "step": 66160 }, { "grad_norm": 0.47732308506965637, "learning_rate": 2.816057164692345e-05, "loss": 0.032, "step": 66170 }, { "grad_norm": 0.36243653297424316, "learning_rate": 2.8145698818840992e-05, "loss": 0.0382, "step": 66180 }, { "grad_norm": 0.3538331091403961, "learning_rate": 2.8130828380712065e-05, "loss": 0.0381, "step": 66190 }, { "grad_norm": 0.3379247784614563, "learning_rate": 2.811596033416285e-05, "loss": 0.0297, "step": 66200 }, { "grad_norm": 0.4090403616428375, "learning_rate": 2.8101094680819307e-05, "loss": 0.0336, "step": 66210 }, { "grad_norm": 1.0450079441070557, "learning_rate": 2.808623142230715e-05, "loss": 0.0329, "step": 66220 }, { "grad_norm": 0.9336413741111755, "learning_rate": 2.8071370560251757e-05, "loss": 0.0345, "step": 66230 }, { "grad_norm": 0.5356612801551819, "learning_rate": 2.805651209627831e-05, "loss": 0.0322, "step": 66240 }, { "grad_norm": 0.3703577220439911, "learning_rate": 2.804165603201171e-05, "loss": 0.0294, "step": 66250 }, { "grad_norm": 0.448406457901001, "learning_rate": 2.802680236907661e-05, "loss": 0.0361, "step": 66260 }, { "grad_norm": 0.24049392342567444, "learning_rate": 2.8011951109097344e-05, "loss": 0.0309, "step": 66270 }, { "grad_norm": 0.4415223002433777, "learning_rate": 2.7997102253698072e-05, "loss": 0.0365, "step": 66280 }, { "grad_norm": 0.39717602729797363, "learning_rate": 2.7982255804502588e-05, "loss": 0.0344, "step": 66290 }, { "grad_norm": 0.25445792078971863, "learning_rate": 2.7967411763134516e-05, "loss": 0.0416, "step": 66300 }, { "grad_norm": 0.45408937335014343, "learning_rate": 2.7952570131217166e-05, "loss": 0.0336, "step": 66310 }, { "grad_norm": 0.49466627836227417, "learning_rate": 2.7937730910373618e-05, "loss": 0.032, "step": 66320 }, { "grad_norm": 0.22918309271335602, "learning_rate": 2.7922894102226628e-05, "loss": 0.0304, "step": 66330 }, { "grad_norm": 0.31590667366981506, "learning_rate": 2.790805970839876e-05, "loss": 0.0299, "step": 66340 }, { "grad_norm": 0.27180206775665283, "learning_rate": 2.789322773051225e-05, "loss": 0.0304, "step": 66350 }, { "grad_norm": 0.343093603849411, "learning_rate": 2.7878398170189124e-05, "loss": 0.0317, "step": 66360 }, { "grad_norm": 0.37073346972465515, "learning_rate": 2.7863571029051112e-05, "loss": 0.0386, "step": 66370 }, { "grad_norm": 0.5138739943504333, "learning_rate": 2.784874630871971e-05, "loss": 0.0397, "step": 66380 }, { "grad_norm": 0.5615423321723938, "learning_rate": 2.7833924010816086e-05, "loss": 0.0348, "step": 66390 }, { "grad_norm": 0.1976107954978943, "learning_rate": 2.7819104136961226e-05, "loss": 0.0357, "step": 66400 }, { "grad_norm": 0.31035664677619934, "learning_rate": 2.780428668877577e-05, "loss": 0.0303, "step": 66410 }, { "grad_norm": 0.3104930818080902, "learning_rate": 2.7789471667880152e-05, "loss": 0.0296, "step": 66420 }, { "grad_norm": 0.566737174987793, "learning_rate": 2.7774659075894523e-05, "loss": 0.0439, "step": 66430 }, { "grad_norm": 0.38741010427474976, "learning_rate": 2.7759848914438778e-05, "loss": 0.0387, "step": 66440 }, { "grad_norm": 0.4817698001861572, "learning_rate": 2.7745041185132496e-05, "loss": 0.0354, "step": 66450 }, { "grad_norm": 0.5032477378845215, "learning_rate": 2.773023588959506e-05, "loss": 0.0418, "step": 66460 }, { "grad_norm": 0.420368492603302, "learning_rate": 2.7715433029445555e-05, "loss": 0.037, "step": 66470 }, { "grad_norm": 0.35686057806015015, "learning_rate": 2.7700632606302766e-05, "loss": 0.0298, "step": 66480 }, { "grad_norm": 0.44227346777915955, "learning_rate": 2.768583462178528e-05, "loss": 0.0362, "step": 66490 }, { "grad_norm": 0.34617507457733154, "learning_rate": 2.767103907751136e-05, "loss": 0.0326, "step": 66500 }, { "grad_norm": 0.6772667765617371, "learning_rate": 2.7656245975099065e-05, "loss": 0.0319, "step": 66510 }, { "grad_norm": 0.298080176115036, "learning_rate": 2.7641455316166093e-05, "loss": 0.0304, "step": 66520 }, { "grad_norm": 0.2288309931755066, "learning_rate": 2.7626667102329973e-05, "loss": 0.0333, "step": 66530 }, { "grad_norm": 0.33523041009902954, "learning_rate": 2.761188133520788e-05, "loss": 0.0307, "step": 66540 }, { "grad_norm": 0.655425488948822, "learning_rate": 2.7597098016416784e-05, "loss": 0.0349, "step": 66550 }, { "grad_norm": 0.5564056038856506, "learning_rate": 2.7582317147573368e-05, "loss": 0.0374, "step": 66560 }, { "grad_norm": 0.3383106589317322, "learning_rate": 2.7567538730294064e-05, "loss": 0.0352, "step": 66570 }, { "grad_norm": 0.45507752895355225, "learning_rate": 2.7552762766194973e-05, "loss": 0.0376, "step": 66580 }, { "grad_norm": 0.30940404534339905, "learning_rate": 2.753798925689202e-05, "loss": 0.0345, "step": 66590 }, { "grad_norm": 0.5030980706214905, "learning_rate": 2.7523218204000774e-05, "loss": 0.0316, "step": 66600 }, { "grad_norm": 0.32802796363830566, "learning_rate": 2.7508449609136612e-05, "loss": 0.0296, "step": 66610 }, { "grad_norm": 0.6088284254074097, "learning_rate": 2.749368347391454e-05, "loss": 0.0391, "step": 66620 }, { "grad_norm": 0.4425630569458008, "learning_rate": 2.7478919799949443e-05, "loss": 0.0311, "step": 66630 }, { "grad_norm": 0.37777066230773926, "learning_rate": 2.7464158588855803e-05, "loss": 0.0287, "step": 66640 }, { "grad_norm": 0.5411084294319153, "learning_rate": 2.7449399842247918e-05, "loss": 0.0318, "step": 66650 }, { "grad_norm": 0.3634944260120392, "learning_rate": 2.7434643561739736e-05, "loss": 0.0265, "step": 66660 }, { "grad_norm": 0.5447437167167664, "learning_rate": 2.7419889748945016e-05, "loss": 0.0257, "step": 66670 }, { "grad_norm": 0.2742719054222107, "learning_rate": 2.7405138405477203e-05, "loss": 0.0342, "step": 66680 }, { "grad_norm": 0.30971235036849976, "learning_rate": 2.73903895329495e-05, "loss": 0.0309, "step": 66690 }, { "grad_norm": 0.521195113658905, "learning_rate": 2.737564313297478e-05, "loss": 0.037, "step": 66700 }, { "grad_norm": 0.303972452878952, "learning_rate": 2.736089920716572e-05, "loss": 0.0266, "step": 66710 }, { "grad_norm": 0.3500533699989319, "learning_rate": 2.7346157757134706e-05, "loss": 0.0355, "step": 66720 }, { "grad_norm": 0.6018097400665283, "learning_rate": 2.733141878449379e-05, "loss": 0.0322, "step": 66730 }, { "grad_norm": 0.2833598256111145, "learning_rate": 2.7316682290854844e-05, "loss": 0.0343, "step": 66740 }, { "grad_norm": 0.5064399838447571, "learning_rate": 2.7301948277829414e-05, "loss": 0.0365, "step": 66750 }, { "grad_norm": 0.5149379372596741, "learning_rate": 2.7287216747028814e-05, "loss": 0.0327, "step": 66760 }, { "grad_norm": 0.3350323438644409, "learning_rate": 2.7272487700064025e-05, "loss": 0.0294, "step": 66770 }, { "grad_norm": 0.3924224376678467, "learning_rate": 2.725776113854583e-05, "loss": 0.0369, "step": 66780 }, { "grad_norm": 0.3560190796852112, "learning_rate": 2.7243037064084664e-05, "loss": 0.029, "step": 66790 }, { "grad_norm": 0.6786896586418152, "learning_rate": 2.7228315478290766e-05, "loss": 0.0393, "step": 66800 }, { "grad_norm": 0.27161023020744324, "learning_rate": 2.721359638277402e-05, "loss": 0.0313, "step": 66810 }, { "grad_norm": 0.3502667546272278, "learning_rate": 2.7198879779144148e-05, "loss": 0.032, "step": 66820 }, { "grad_norm": 0.28199535608291626, "learning_rate": 2.7184165669010485e-05, "loss": 0.0305, "step": 66830 }, { "grad_norm": 0.37446457147598267, "learning_rate": 2.716945405398218e-05, "loss": 0.0356, "step": 66840 }, { "grad_norm": 0.3693915903568268, "learning_rate": 2.715474493566803e-05, "loss": 0.0291, "step": 66850 }, { "grad_norm": 0.4144318699836731, "learning_rate": 2.7140038315676652e-05, "loss": 0.0281, "step": 66860 }, { "grad_norm": 0.4105609059333801, "learning_rate": 2.7125334195616275e-05, "loss": 0.0498, "step": 66870 }, { "grad_norm": 0.49256017804145813, "learning_rate": 2.7110632577094997e-05, "loss": 0.0389, "step": 66880 }, { "grad_norm": 0.345562219619751, "learning_rate": 2.709593346172051e-05, "loss": 0.0356, "step": 66890 }, { "grad_norm": 0.5802539587020874, "learning_rate": 2.708123685110032e-05, "loss": 0.0316, "step": 66900 }, { "grad_norm": 0.29301613569259644, "learning_rate": 2.7066542746841588e-05, "loss": 0.0215, "step": 66910 }, { "grad_norm": 0.7492597699165344, "learning_rate": 2.7051851150551267e-05, "loss": 0.0282, "step": 66920 }, { "grad_norm": 0.3570857644081116, "learning_rate": 2.7037162063835996e-05, "loss": 0.0319, "step": 66930 }, { "grad_norm": 0.6168721914291382, "learning_rate": 2.702247548830218e-05, "loss": 0.0325, "step": 66940 }, { "grad_norm": 0.2529233694076538, "learning_rate": 2.7007791425555883e-05, "loss": 0.0236, "step": 66950 }, { "grad_norm": 0.291273295879364, "learning_rate": 2.6993109877202945e-05, "loss": 0.0347, "step": 66960 }, { "grad_norm": 0.3746768534183502, "learning_rate": 2.6978430844848935e-05, "loss": 0.035, "step": 66970 }, { "grad_norm": 0.24704621732234955, "learning_rate": 2.69637543300991e-05, "loss": 0.028, "step": 66980 }, { "grad_norm": 0.30526137351989746, "learning_rate": 2.694908033455848e-05, "loss": 0.0304, "step": 66990 }, { "grad_norm": 0.31940463185310364, "learning_rate": 2.6934408859831733e-05, "loss": 0.0389, "step": 67000 }, { "grad_norm": 0.4241073429584503, "learning_rate": 2.69197399075234e-05, "loss": 0.0372, "step": 67010 }, { "grad_norm": 0.3837438225746155, "learning_rate": 2.6905073479237584e-05, "loss": 0.0275, "step": 67020 }, { "grad_norm": 0.4403733015060425, "learning_rate": 2.6890409576578235e-05, "loss": 0.0336, "step": 67030 }, { "grad_norm": 0.6337973475456238, "learning_rate": 2.687574820114892e-05, "loss": 0.0347, "step": 67040 }, { "grad_norm": 0.5028713345527649, "learning_rate": 2.686108935455305e-05, "loss": 0.0387, "step": 67050 }, { "grad_norm": 0.4528617560863495, "learning_rate": 2.6846433038393603e-05, "loss": 0.0361, "step": 67060 }, { "grad_norm": 0.5786473155021667, "learning_rate": 2.6831779254273482e-05, "loss": 0.0334, "step": 67070 }, { "grad_norm": 0.3743865489959717, "learning_rate": 2.6817128003795122e-05, "loss": 0.0289, "step": 67080 }, { "grad_norm": 0.4025677740573883, "learning_rate": 2.680247928856081e-05, "loss": 0.0293, "step": 67090 }, { "grad_norm": 0.4460088014602661, "learning_rate": 2.6787833110172466e-05, "loss": 0.0385, "step": 67100 }, { "grad_norm": 1.3205844163894653, "learning_rate": 2.6773189470231807e-05, "loss": 0.0339, "step": 67110 }, { "grad_norm": 0.43459075689315796, "learning_rate": 2.675854837034022e-05, "loss": 0.0388, "step": 67120 }, { "grad_norm": 0.4120723605155945, "learning_rate": 2.6743909812098833e-05, "loss": 0.0287, "step": 67130 }, { "grad_norm": 0.28345930576324463, "learning_rate": 2.67292737971085e-05, "loss": 0.0344, "step": 67140 }, { "grad_norm": 0.828737735748291, "learning_rate": 2.671464032696982e-05, "loss": 0.0339, "step": 67150 }, { "grad_norm": 0.5545713305473328, "learning_rate": 2.6700009403283055e-05, "loss": 0.0342, "step": 67160 }, { "grad_norm": 0.4716249406337738, "learning_rate": 2.6685381027648214e-05, "loss": 0.0368, "step": 67170 }, { "grad_norm": 0.5200282335281372, "learning_rate": 2.6670755201665086e-05, "loss": 0.0474, "step": 67180 }, { "grad_norm": 0.39223960041999817, "learning_rate": 2.6656131926933066e-05, "loss": 0.0434, "step": 67190 }, { "grad_norm": 0.4403271973133087, "learning_rate": 2.664151120505136e-05, "loss": 0.0306, "step": 67200 }, { "grad_norm": 0.32840538024902344, "learning_rate": 2.6626893037618873e-05, "loss": 0.0315, "step": 67210 }, { "grad_norm": 0.4960401654243469, "learning_rate": 2.6612277426234243e-05, "loss": 0.0486, "step": 67220 }, { "grad_norm": 0.2932750880718231, "learning_rate": 2.659766437249577e-05, "loss": 0.0396, "step": 67230 }, { "grad_norm": 0.4796620309352875, "learning_rate": 2.6583053878001557e-05, "loss": 0.0312, "step": 67240 }, { "grad_norm": 0.399431437253952, "learning_rate": 2.6568445944349323e-05, "loss": 0.0328, "step": 67250 }, { "grad_norm": 0.43265244364738464, "learning_rate": 2.6553840573136657e-05, "loss": 0.0331, "step": 67260 }, { "grad_norm": 0.7153671383857727, "learning_rate": 2.6539237765960706e-05, "loss": 0.0305, "step": 67270 }, { "grad_norm": 0.7974071502685547, "learning_rate": 2.6524637524418465e-05, "loss": 0.0429, "step": 67280 }, { "grad_norm": 0.407361775636673, "learning_rate": 2.651003985010655e-05, "loss": 0.026, "step": 67290 }, { "grad_norm": 0.49149227142333984, "learning_rate": 2.649544474462138e-05, "loss": 0.0397, "step": 67300 }, { "grad_norm": 0.45831847190856934, "learning_rate": 2.6480852209559015e-05, "loss": 0.0302, "step": 67310 }, { "grad_norm": 0.6410447359085083, "learning_rate": 2.6466262246515296e-05, "loss": 0.0299, "step": 67320 }, { "grad_norm": 0.34368863701820374, "learning_rate": 2.6451674857085746e-05, "loss": 0.0325, "step": 67330 }, { "grad_norm": 0.341235488653183, "learning_rate": 2.6437090042865655e-05, "loss": 0.0408, "step": 67340 }, { "grad_norm": 0.2811747193336487, "learning_rate": 2.6422507805449947e-05, "loss": 0.0302, "step": 67350 }, { "grad_norm": 0.6039906144142151, "learning_rate": 2.6407928146433357e-05, "loss": 0.0368, "step": 67360 }, { "grad_norm": 0.3760591149330139, "learning_rate": 2.639335106741026e-05, "loss": 0.0324, "step": 67370 }, { "grad_norm": 0.3942543864250183, "learning_rate": 2.6378776569974796e-05, "loss": 0.0368, "step": 67380 }, { "grad_norm": 0.36801591515541077, "learning_rate": 2.6364204655720814e-05, "loss": 0.0295, "step": 67390 }, { "grad_norm": 0.4188639521598816, "learning_rate": 2.6349635326241894e-05, "loss": 0.0365, "step": 67400 }, { "grad_norm": 0.7370200157165527, "learning_rate": 2.633506858313127e-05, "loss": 0.0337, "step": 67410 }, { "grad_norm": 0.5191206932067871, "learning_rate": 2.6320504427981975e-05, "loss": 0.0286, "step": 67420 }, { "grad_norm": 0.38040393590927124, "learning_rate": 2.6305942862386734e-05, "loss": 0.0341, "step": 67430 }, { "grad_norm": 0.5149023532867432, "learning_rate": 2.629138388793793e-05, "loss": 0.0294, "step": 67440 }, { "grad_norm": 0.41391700506210327, "learning_rate": 2.6276827506227743e-05, "loss": 0.0376, "step": 67450 }, { "grad_norm": 0.38212284445762634, "learning_rate": 2.6262273718848037e-05, "loss": 0.0296, "step": 67460 }, { "grad_norm": 0.45309779047966003, "learning_rate": 2.6247722527390407e-05, "loss": 0.029, "step": 67470 }, { "grad_norm": 0.2522594630718231, "learning_rate": 2.6233173933446104e-05, "loss": 0.0377, "step": 67480 }, { "grad_norm": 0.5345180630683899, "learning_rate": 2.6218627938606188e-05, "loss": 0.0352, "step": 67490 }, { "grad_norm": 0.309999018907547, "learning_rate": 2.620408454446135e-05, "loss": 0.0296, "step": 67500 }, { "grad_norm": 0.42911669611930847, "learning_rate": 2.6189543752602046e-05, "loss": 0.0294, "step": 67510 }, { "grad_norm": 0.4345991313457489, "learning_rate": 2.6175005564618444e-05, "loss": 0.0318, "step": 67520 }, { "grad_norm": 0.3311856985092163, "learning_rate": 2.6160469982100428e-05, "loss": 0.0369, "step": 67530 }, { "grad_norm": 0.7079976797103882, "learning_rate": 2.6145937006637555e-05, "loss": 0.0534, "step": 67540 }, { "grad_norm": 0.3832126557826996, "learning_rate": 2.6131406639819166e-05, "loss": 0.0353, "step": 67550 }, { "grad_norm": 0.41876348853111267, "learning_rate": 2.611687888323424e-05, "loss": 0.03, "step": 67560 }, { "grad_norm": 0.36622437834739685, "learning_rate": 2.610235373847153e-05, "loss": 0.0412, "step": 67570 }, { "grad_norm": 0.47291550040245056, "learning_rate": 2.6087831207119494e-05, "loss": 0.0377, "step": 67580 }, { "grad_norm": 0.45470115542411804, "learning_rate": 2.60733112907663e-05, "loss": 0.0404, "step": 67590 }, { "grad_norm": 0.37759968638420105, "learning_rate": 2.6058793990999786e-05, "loss": 0.0381, "step": 67600 }, { "grad_norm": 0.5614479780197144, "learning_rate": 2.6044279309407593e-05, "loss": 0.0307, "step": 67610 }, { "grad_norm": 0.451555460691452, "learning_rate": 2.6029767247576976e-05, "loss": 0.0398, "step": 67620 }, { "grad_norm": 0.30325424671173096, "learning_rate": 2.601525780709497e-05, "loss": 0.0299, "step": 67630 }, { "grad_norm": 0.5267826914787292, "learning_rate": 2.6000750989548318e-05, "loss": 0.0409, "step": 67640 }, { "grad_norm": 0.4520646929740906, "learning_rate": 2.5986246796523462e-05, "loss": 0.0323, "step": 67650 }, { "grad_norm": 0.30092522501945496, "learning_rate": 2.5971745229606546e-05, "loss": 0.0324, "step": 67660 }, { "grad_norm": 0.5479021668434143, "learning_rate": 2.5957246290383442e-05, "loss": 0.0434, "step": 67670 }, { "grad_norm": 0.35241830348968506, "learning_rate": 2.594274998043975e-05, "loss": 0.04, "step": 67680 }, { "grad_norm": 0.5866302847862244, "learning_rate": 2.5928256301360733e-05, "loss": 0.0288, "step": 67690 }, { "grad_norm": 0.38585150241851807, "learning_rate": 2.591376525473142e-05, "loss": 0.0259, "step": 67700 }, { "grad_norm": 0.41895967721939087, "learning_rate": 2.5899276842136523e-05, "loss": 0.0323, "step": 67710 }, { "grad_norm": 0.38536104559898376, "learning_rate": 2.5884791065160495e-05, "loss": 0.0306, "step": 67720 }, { "grad_norm": 0.3992888033390045, "learning_rate": 2.5870307925387436e-05, "loss": 0.0299, "step": 67730 }, { "grad_norm": 1.0495190620422363, "learning_rate": 2.585582742440125e-05, "loss": 0.0247, "step": 67740 }, { "grad_norm": 0.4580164849758148, "learning_rate": 2.5841349563785455e-05, "loss": 0.0304, "step": 67750 }, { "grad_norm": 0.1661113053560257, "learning_rate": 2.5826874345123355e-05, "loss": 0.0301, "step": 67760 }, { "grad_norm": 0.4032401442527771, "learning_rate": 2.581240176999792e-05, "loss": 0.0285, "step": 67770 }, { "grad_norm": 0.6881360411643982, "learning_rate": 2.579793183999189e-05, "loss": 0.0329, "step": 67780 }, { "grad_norm": 0.5793899297714233, "learning_rate": 2.5783464556687627e-05, "loss": 0.036, "step": 67790 }, { "grad_norm": 0.7609584927558899, "learning_rate": 2.576899992166729e-05, "loss": 0.0375, "step": 67800 }, { "grad_norm": 0.3886268734931946, "learning_rate": 2.5754537936512668e-05, "loss": 0.0281, "step": 67810 }, { "grad_norm": 0.6203373074531555, "learning_rate": 2.574007860280535e-05, "loss": 0.0502, "step": 67820 }, { "grad_norm": 0.2814624607563019, "learning_rate": 2.5725621922126518e-05, "loss": 0.0296, "step": 67830 }, { "grad_norm": 0.429175466299057, "learning_rate": 2.571116789605722e-05, "loss": 0.0338, "step": 67840 }, { "grad_norm": 0.5040152668952942, "learning_rate": 2.5696716526178054e-05, "loss": 0.0426, "step": 67850 }, { "grad_norm": 0.5970534682273865, "learning_rate": 2.568226781406946e-05, "loss": 0.0441, "step": 67860 }, { "grad_norm": 0.463778555393219, "learning_rate": 2.5667821761311474e-05, "loss": 0.0407, "step": 67870 }, { "grad_norm": 0.3100944459438324, "learning_rate": 2.5653378369483917e-05, "loss": 0.0284, "step": 67880 }, { "grad_norm": 0.5186207890510559, "learning_rate": 2.5638937640166292e-05, "loss": 0.029, "step": 67890 }, { "grad_norm": 0.3820953369140625, "learning_rate": 2.5624499574937842e-05, "loss": 0.0329, "step": 67900 }, { "grad_norm": 0.2843179404735565, "learning_rate": 2.5610064175377456e-05, "loss": 0.0327, "step": 67910 }, { "grad_norm": 0.37795183062553406, "learning_rate": 2.5595631443063782e-05, "loss": 0.0308, "step": 67920 }, { "grad_norm": 0.46669507026672363, "learning_rate": 2.5581201379575177e-05, "loss": 0.0286, "step": 67930 }, { "grad_norm": 0.4970901906490326, "learning_rate": 2.5566773986489655e-05, "loss": 0.0324, "step": 67940 }, { "grad_norm": 0.4335514307022095, "learning_rate": 2.5552349265385e-05, "loss": 0.0475, "step": 67950 }, { "grad_norm": 0.37874317169189453, "learning_rate": 2.5537927217838675e-05, "loss": 0.0272, "step": 67960 }, { "grad_norm": 0.3570016026496887, "learning_rate": 2.5523507845427864e-05, "loss": 0.0341, "step": 67970 }, { "grad_norm": 0.4076637327671051, "learning_rate": 2.5509091149729413e-05, "loss": 0.0319, "step": 67980 }, { "grad_norm": 0.35509660840034485, "learning_rate": 2.5494677132319962e-05, "loss": 0.0318, "step": 67990 }, { "grad_norm": 0.44256871938705444, "learning_rate": 2.548026579477575e-05, "loss": 0.0334, "step": 68000 }, { "grad_norm": 0.470886766910553, "learning_rate": 2.5465857138672834e-05, "loss": 0.0318, "step": 68010 }, { "grad_norm": 0.31914952397346497, "learning_rate": 2.5451451165586848e-05, "loss": 0.038, "step": 68020 }, { "grad_norm": 0.44449955224990845, "learning_rate": 2.543704787709329e-05, "loss": 0.029, "step": 68030 }, { "grad_norm": 0.48529067635536194, "learning_rate": 2.542264727476723e-05, "loss": 0.0331, "step": 68040 }, { "grad_norm": 0.47466230392456055, "learning_rate": 2.5408249360183535e-05, "loss": 0.0404, "step": 68050 }, { "grad_norm": 0.34558266401290894, "learning_rate": 2.5393854134916694e-05, "loss": 0.0255, "step": 68060 }, { "grad_norm": 0.5211125612258911, "learning_rate": 2.5379461600540993e-05, "loss": 0.0295, "step": 68070 }, { "grad_norm": 0.2985466718673706, "learning_rate": 2.5365071758630317e-05, "loss": 0.0455, "step": 68080 }, { "grad_norm": 0.3741890490055084, "learning_rate": 2.535068461075839e-05, "loss": 0.0305, "step": 68090 }, { "grad_norm": 0.35301604866981506, "learning_rate": 2.5336300158498516e-05, "loss": 0.0391, "step": 68100 }, { "grad_norm": 0.7402101159095764, "learning_rate": 2.5321918403423793e-05, "loss": 0.0297, "step": 68110 }, { "grad_norm": 0.38309770822525024, "learning_rate": 2.5307539347106957e-05, "loss": 0.0308, "step": 68120 }, { "grad_norm": 0.3852204978466034, "learning_rate": 2.529316299112049e-05, "loss": 0.0325, "step": 68130 }, { "grad_norm": 0.6060024499893188, "learning_rate": 2.5278789337036567e-05, "loss": 0.033, "step": 68140 }, { "grad_norm": 0.5778374075889587, "learning_rate": 2.52644183864271e-05, "loss": 0.0352, "step": 68150 }, { "grad_norm": 0.38965532183647156, "learning_rate": 2.525005014086363e-05, "loss": 0.0326, "step": 68160 }, { "grad_norm": 0.7563886642456055, "learning_rate": 2.5235684601917465e-05, "loss": 0.0386, "step": 68170 }, { "grad_norm": 0.45890331268310547, "learning_rate": 2.5221321771159618e-05, "loss": 0.0301, "step": 68180 }, { "grad_norm": 0.34841597080230713, "learning_rate": 2.5206961650160743e-05, "loss": 0.0315, "step": 68190 }, { "grad_norm": 0.2373328059911728, "learning_rate": 2.519260424049129e-05, "loss": 0.0352, "step": 68200 }, { "grad_norm": 0.25333964824676514, "learning_rate": 2.5178249543721304e-05, "loss": 0.0372, "step": 68210 }, { "grad_norm": 0.3260090947151184, "learning_rate": 2.5163897561420668e-05, "loss": 0.0279, "step": 68220 }, { "grad_norm": 0.5704978704452515, "learning_rate": 2.514954829515883e-05, "loss": 0.037, "step": 68230 }, { "grad_norm": 0.31074944138526917, "learning_rate": 2.513520174650505e-05, "loss": 0.0316, "step": 68240 }, { "grad_norm": 0.3144662380218506, "learning_rate": 2.5120857917028207e-05, "loss": 0.0278, "step": 68250 }, { "grad_norm": 0.4535314738750458, "learning_rate": 2.5106516808296953e-05, "loss": 0.024, "step": 68260 }, { "grad_norm": 0.28294071555137634, "learning_rate": 2.5092178421879552e-05, "loss": 0.0304, "step": 68270 }, { "grad_norm": 0.671770453453064, "learning_rate": 2.5077842759344116e-05, "loss": 0.0403, "step": 68280 }, { "grad_norm": 0.42709881067276, "learning_rate": 2.5063509822258304e-05, "loss": 0.0372, "step": 68290 }, { "grad_norm": 0.8179237246513367, "learning_rate": 2.5049179612189576e-05, "loss": 0.0372, "step": 68300 }, { "grad_norm": 0.45322519540786743, "learning_rate": 2.5034852130705043e-05, "loss": 0.0431, "step": 68310 }, { "grad_norm": 0.49642717838287354, "learning_rate": 2.5020527379371572e-05, "loss": 0.0401, "step": 68320 }, { "grad_norm": 0.4751732349395752, "learning_rate": 2.500620535975564e-05, "loss": 0.0302, "step": 68330 }, { "grad_norm": 0.44029173254966736, "learning_rate": 2.4991886073423515e-05, "loss": 0.0383, "step": 68340 }, { "grad_norm": 0.23137108981609344, "learning_rate": 2.4977569521941134e-05, "loss": 0.0276, "step": 68350 }, { "grad_norm": 0.4310547113418579, "learning_rate": 2.4963255706874143e-05, "loss": 0.028, "step": 68360 }, { "grad_norm": 0.39548176527023315, "learning_rate": 2.4948944629787853e-05, "loss": 0.0363, "step": 68370 }, { "grad_norm": 0.5096831917762756, "learning_rate": 2.493463629224731e-05, "loss": 0.0413, "step": 68380 }, { "grad_norm": 0.27896514534950256, "learning_rate": 2.492033069581728e-05, "loss": 0.0322, "step": 68390 }, { "grad_norm": 0.8431175351142883, "learning_rate": 2.4906027842062137e-05, "loss": 0.0368, "step": 68400 }, { "grad_norm": 0.2675110399723053, "learning_rate": 2.48917277325461e-05, "loss": 0.0316, "step": 68410 }, { "grad_norm": 0.3167223036289215, "learning_rate": 2.487743036883295e-05, "loss": 0.0329, "step": 68420 }, { "grad_norm": 0.29629403352737427, "learning_rate": 2.486313575248626e-05, "loss": 0.0276, "step": 68430 }, { "grad_norm": 0.36533230543136597, "learning_rate": 2.4848843885069235e-05, "loss": 0.0381, "step": 68440 }, { "grad_norm": 0.3203701972961426, "learning_rate": 2.483455476814484e-05, "loss": 0.0279, "step": 68450 }, { "grad_norm": 0.444421648979187, "learning_rate": 2.4820268403275672e-05, "loss": 0.0346, "step": 68460 }, { "grad_norm": 0.45546582341194153, "learning_rate": 2.480598479202413e-05, "loss": 0.0341, "step": 68470 }, { "grad_norm": 0.2951976954936981, "learning_rate": 2.4791703935952193e-05, "loss": 0.0314, "step": 68480 }, { "grad_norm": 0.48623955249786377, "learning_rate": 2.477742583662163e-05, "loss": 0.0362, "step": 68490 }, { "grad_norm": 0.28143396973609924, "learning_rate": 2.476315049559384e-05, "loss": 0.0346, "step": 68500 }, { "grad_norm": 0.6409662365913391, "learning_rate": 2.4748877914429987e-05, "loss": 0.042, "step": 68510 }, { "grad_norm": 0.2885066270828247, "learning_rate": 2.4734608094690857e-05, "loss": 0.0401, "step": 68520 }, { "grad_norm": 0.23822921514511108, "learning_rate": 2.4720341037937006e-05, "loss": 0.0449, "step": 68530 }, { "grad_norm": 0.2994224429130554, "learning_rate": 2.4706076745728645e-05, "loss": 0.042, "step": 68540 }, { "grad_norm": 0.35948339104652405, "learning_rate": 2.469181521962573e-05, "loss": 0.0315, "step": 68550 }, { "grad_norm": 0.5160295367240906, "learning_rate": 2.467755646118783e-05, "loss": 0.0253, "step": 68560 }, { "grad_norm": 0.317654550075531, "learning_rate": 2.46633004719743e-05, "loss": 0.0316, "step": 68570 }, { "grad_norm": 0.32083621621131897, "learning_rate": 2.4649047253544123e-05, "loss": 0.0297, "step": 68580 }, { "grad_norm": 0.28702816367149353, "learning_rate": 2.4634796807456016e-05, "loss": 0.0247, "step": 68590 }, { "grad_norm": 0.4514797329902649, "learning_rate": 2.46205491352684e-05, "loss": 0.0375, "step": 68600 }, { "grad_norm": 0.408602237701416, "learning_rate": 2.4606304238539375e-05, "loss": 0.0296, "step": 68610 }, { "grad_norm": 0.8081472516059875, "learning_rate": 2.4592062118826753e-05, "loss": 0.0321, "step": 68620 }, { "grad_norm": 0.47983452677726746, "learning_rate": 2.4577822777688004e-05, "loss": 0.0237, "step": 68630 }, { "grad_norm": 0.5540101528167725, "learning_rate": 2.4563586216680347e-05, "loss": 0.0381, "step": 68640 }, { "grad_norm": 0.5321470499038696, "learning_rate": 2.4549352437360622e-05, "loss": 0.0267, "step": 68650 }, { "grad_norm": 0.4440300166606903, "learning_rate": 2.4535121441285493e-05, "loss": 0.0376, "step": 68660 }, { "grad_norm": 0.4947969615459442, "learning_rate": 2.4520893230011172e-05, "loss": 0.0272, "step": 68670 }, { "grad_norm": 0.28778672218322754, "learning_rate": 2.450666780509368e-05, "loss": 0.0317, "step": 68680 }, { "grad_norm": 0.5198530554771423, "learning_rate": 2.4492445168088656e-05, "loss": 0.0421, "step": 68690 }, { "grad_norm": 0.3217993378639221, "learning_rate": 2.447822532055149e-05, "loss": 0.0363, "step": 68700 }, { "grad_norm": 0.5171933174133301, "learning_rate": 2.4464008264037212e-05, "loss": 0.0317, "step": 68710 }, { "grad_norm": 0.5080624222755432, "learning_rate": 2.4449794000100605e-05, "loss": 0.0351, "step": 68720 }, { "grad_norm": 0.47545406222343445, "learning_rate": 2.4435582530296108e-05, "loss": 0.0319, "step": 68730 }, { "grad_norm": 0.2281218320131302, "learning_rate": 2.4421373856177887e-05, "loss": 0.0283, "step": 68740 }, { "grad_norm": 0.3638085722923279, "learning_rate": 2.440716797929975e-05, "loss": 0.0295, "step": 68750 }, { "grad_norm": 0.3815038800239563, "learning_rate": 2.439296490121526e-05, "loss": 0.0224, "step": 68760 }, { "grad_norm": 0.42007866501808167, "learning_rate": 2.4378764623477608e-05, "loss": 0.0276, "step": 68770 }, { "grad_norm": 0.3553641438484192, "learning_rate": 2.436456714763974e-05, "loss": 0.0277, "step": 68780 }, { "grad_norm": 0.3434520959854126, "learning_rate": 2.4350372475254262e-05, "loss": 0.0333, "step": 68790 }, { "grad_norm": 0.28545308113098145, "learning_rate": 2.433618060787351e-05, "loss": 0.0349, "step": 68800 }, { "grad_norm": 0.3281495273113251, "learning_rate": 2.432199154704944e-05, "loss": 0.0429, "step": 68810 }, { "grad_norm": 0.4493880867958069, "learning_rate": 2.430780529433377e-05, "loss": 0.0351, "step": 68820 }, { "grad_norm": 0.3857964277267456, "learning_rate": 2.42936218512779e-05, "loss": 0.0362, "step": 68830 }, { "grad_norm": 0.332573801279068, "learning_rate": 2.427944121943288e-05, "loss": 0.023, "step": 68840 }, { "grad_norm": 0.3797268271446228, "learning_rate": 2.4265263400349497e-05, "loss": 0.0308, "step": 68850 }, { "grad_norm": 0.599071741104126, "learning_rate": 2.4251088395578213e-05, "loss": 0.0314, "step": 68860 }, { "grad_norm": 0.37924349308013916, "learning_rate": 2.4236916206669214e-05, "loss": 0.0372, "step": 68870 }, { "grad_norm": 0.5065199136734009, "learning_rate": 2.4222746835172304e-05, "loss": 0.0323, "step": 68880 }, { "grad_norm": 0.3917907774448395, "learning_rate": 2.4208580282637065e-05, "loss": 0.0318, "step": 68890 }, { "grad_norm": 0.42922645807266235, "learning_rate": 2.4194416550612687e-05, "loss": 0.0325, "step": 68900 }, { "grad_norm": 0.20426015555858612, "learning_rate": 2.418025564064812e-05, "loss": 0.0309, "step": 68910 }, { "grad_norm": 0.6130631566047668, "learning_rate": 2.4166097554291978e-05, "loss": 0.029, "step": 68920 }, { "grad_norm": 0.5982769131660461, "learning_rate": 2.4151942293092583e-05, "loss": 0.0313, "step": 68930 }, { "grad_norm": 0.13293878734111786, "learning_rate": 2.41377898585979e-05, "loss": 0.027, "step": 68940 }, { "grad_norm": 0.47247183322906494, "learning_rate": 2.4123640252355652e-05, "loss": 0.0327, "step": 68950 }, { "grad_norm": 0.34224674105644226, "learning_rate": 2.410949347591318e-05, "loss": 0.0328, "step": 68960 }, { "grad_norm": 0.47071361541748047, "learning_rate": 2.4095349530817578e-05, "loss": 0.0525, "step": 68970 }, { "grad_norm": 0.38512855768203735, "learning_rate": 2.4081208418615603e-05, "loss": 0.0291, "step": 68980 }, { "grad_norm": 0.44481518864631653, "learning_rate": 2.4067070140853725e-05, "loss": 0.0367, "step": 68990 }, { "grad_norm": 0.3214735984802246, "learning_rate": 2.405293469907805e-05, "loss": 0.0407, "step": 69000 }, { "grad_norm": 0.8410633206367493, "learning_rate": 2.4038802094834446e-05, "loss": 0.0336, "step": 69010 }, { "grad_norm": 0.31165367364883423, "learning_rate": 2.4024672329668395e-05, "loss": 0.0332, "step": 69020 }, { "grad_norm": 0.36311423778533936, "learning_rate": 2.4010545405125125e-05, "loss": 0.0225, "step": 69030 }, { "grad_norm": 0.4647122919559479, "learning_rate": 2.3996421322749528e-05, "loss": 0.0399, "step": 69040 }, { "grad_norm": 0.399953693151474, "learning_rate": 2.3982300084086224e-05, "loss": 0.0301, "step": 69050 }, { "grad_norm": 0.5829337239265442, "learning_rate": 2.396818169067944e-05, "loss": 0.0339, "step": 69060 }, { "grad_norm": 0.4016403555870056, "learning_rate": 2.3954066144073178e-05, "loss": 0.0281, "step": 69070 }, { "grad_norm": 0.6353239417076111, "learning_rate": 2.39399534458111e-05, "loss": 0.0321, "step": 69080 }, { "grad_norm": 0.3058718144893646, "learning_rate": 2.392584359743651e-05, "loss": 0.0301, "step": 69090 }, { "grad_norm": 0.35514935851097107, "learning_rate": 2.3911736600492463e-05, "loss": 0.0222, "step": 69100 }, { "grad_norm": 0.4206056296825409, "learning_rate": 2.3897632456521672e-05, "loss": 0.0283, "step": 69110 }, { "grad_norm": 0.5929557085037231, "learning_rate": 2.388353116706658e-05, "loss": 0.0409, "step": 69120 }, { "grad_norm": 0.3234335482120514, "learning_rate": 2.3869432733669232e-05, "loss": 0.0304, "step": 69130 }, { "grad_norm": 0.33211418986320496, "learning_rate": 2.385533715787145e-05, "loss": 0.0386, "step": 69140 }, { "grad_norm": 0.29767701029777527, "learning_rate": 2.3841244441214673e-05, "loss": 0.0263, "step": 69150 }, { "grad_norm": 0.572011411190033, "learning_rate": 2.3827154585240073e-05, "loss": 0.0287, "step": 69160 }, { "grad_norm": 0.41523370146751404, "learning_rate": 2.3813067591488498e-05, "loss": 0.0444, "step": 69170 }, { "grad_norm": 0.4015130400657654, "learning_rate": 2.3798983461500494e-05, "loss": 0.0457, "step": 69180 }, { "grad_norm": 0.8017426133155823, "learning_rate": 2.3784902196816256e-05, "loss": 0.0393, "step": 69190 }, { "grad_norm": 0.47110122442245483, "learning_rate": 2.3770823798975716e-05, "loss": 0.0345, "step": 69200 }, { "grad_norm": 0.4945499002933502, "learning_rate": 2.3756748269518424e-05, "loss": 0.03, "step": 69210 }, { "grad_norm": 0.3690513074398041, "learning_rate": 2.374267560998371e-05, "loss": 0.043, "step": 69220 }, { "grad_norm": 0.36057183146476746, "learning_rate": 2.3728605821910477e-05, "loss": 0.0336, "step": 69230 }, { "grad_norm": 0.5247197151184082, "learning_rate": 2.3714538906837453e-05, "loss": 0.0432, "step": 69240 }, { "grad_norm": 0.39859750866889954, "learning_rate": 2.3700474866302913e-05, "loss": 0.0294, "step": 69250 }, { "grad_norm": 0.3301790654659271, "learning_rate": 2.3686413701844917e-05, "loss": 0.0282, "step": 69260 }, { "grad_norm": 0.3567858934402466, "learning_rate": 2.367235541500114e-05, "loss": 0.0301, "step": 69270 }, { "grad_norm": 0.23195883631706238, "learning_rate": 2.3658300007308993e-05, "loss": 0.031, "step": 69280 }, { "grad_norm": 0.9186543226242065, "learning_rate": 2.364424748030555e-05, "loss": 0.0268, "step": 69290 }, { "grad_norm": 0.7388376593589783, "learning_rate": 2.3630197835527595e-05, "loss": 0.0359, "step": 69300 }, { "grad_norm": 0.33755698800086975, "learning_rate": 2.3616151074511546e-05, "loss": 0.0392, "step": 69310 }, { "grad_norm": 0.33823689818382263, "learning_rate": 2.3602107198793545e-05, "loss": 0.0403, "step": 69320 }, { "grad_norm": 0.3517979383468628, "learning_rate": 2.3588066209909427e-05, "loss": 0.0289, "step": 69330 }, { "grad_norm": 0.5191157460212708, "learning_rate": 2.3574028109394662e-05, "loss": 0.0327, "step": 69340 }, { "grad_norm": 0.36289843916893005, "learning_rate": 2.3559992898784454e-05, "loss": 0.0282, "step": 69350 }, { "grad_norm": 0.30383914709091187, "learning_rate": 2.354596057961367e-05, "loss": 0.027, "step": 69360 }, { "grad_norm": 0.37681514024734497, "learning_rate": 2.353193115341688e-05, "loss": 0.0387, "step": 69370 }, { "grad_norm": 0.3177138864994049, "learning_rate": 2.3517904621728294e-05, "loss": 0.0259, "step": 69380 }, { "grad_norm": 0.3181685507297516, "learning_rate": 2.3503880986081855e-05, "loss": 0.0319, "step": 69390 }, { "grad_norm": 0.38756850361824036, "learning_rate": 2.3489860248011136e-05, "loss": 0.0302, "step": 69400 }, { "grad_norm": 0.431939959526062, "learning_rate": 2.3475842409049464e-05, "loss": 0.0329, "step": 69410 }, { "grad_norm": 0.4410700500011444, "learning_rate": 2.346182747072975e-05, "loss": 0.038, "step": 69420 }, { "grad_norm": 0.29879993200302124, "learning_rate": 2.3447815434584718e-05, "loss": 0.0239, "step": 69430 }, { "grad_norm": 0.3294229209423065, "learning_rate": 2.3433806302146656e-05, "loss": 0.0291, "step": 69440 }, { "grad_norm": 0.6051582098007202, "learning_rate": 2.3419800074947612e-05, "loss": 0.0416, "step": 69450 }, { "grad_norm": 0.47337090969085693, "learning_rate": 2.3405796754519248e-05, "loss": 0.0434, "step": 69460 }, { "grad_norm": 0.40286344289779663, "learning_rate": 2.3391796342392992e-05, "loss": 0.041, "step": 69470 }, { "grad_norm": 0.4153488874435425, "learning_rate": 2.337779884009984e-05, "loss": 0.0322, "step": 69480 }, { "grad_norm": 0.32448190450668335, "learning_rate": 2.3363804249170624e-05, "loss": 0.0345, "step": 69490 }, { "grad_norm": 0.453551709651947, "learning_rate": 2.3349812571135703e-05, "loss": 0.0305, "step": 69500 }, { "grad_norm": 0.3360893726348877, "learning_rate": 2.333582380752523e-05, "loss": 0.0256, "step": 69510 }, { "grad_norm": 0.29232144355773926, "learning_rate": 2.3321837959868963e-05, "loss": 0.0311, "step": 69520 }, { "grad_norm": 0.3637884557247162, "learning_rate": 2.3307855029696383e-05, "loss": 0.0285, "step": 69530 }, { "grad_norm": 0.3916666507720947, "learning_rate": 2.3293875018536648e-05, "loss": 0.0287, "step": 69540 }, { "grad_norm": 0.48383113741874695, "learning_rate": 2.327989792791861e-05, "loss": 0.0316, "step": 69550 }, { "grad_norm": 0.5215501189231873, "learning_rate": 2.326592375937074e-05, "loss": 0.0294, "step": 69560 }, { "grad_norm": 0.45872819423675537, "learning_rate": 2.3251952514421248e-05, "loss": 0.0282, "step": 69570 }, { "grad_norm": 0.41318246722221375, "learning_rate": 2.323798419459804e-05, "loss": 0.0277, "step": 69580 }, { "grad_norm": 0.39589041471481323, "learning_rate": 2.3224018801428625e-05, "loss": 0.0305, "step": 69590 }, { "grad_norm": 0.44450873136520386, "learning_rate": 2.3210056336440273e-05, "loss": 0.0292, "step": 69600 }, { "grad_norm": 0.2899734377861023, "learning_rate": 2.3196096801159844e-05, "loss": 0.0278, "step": 69610 }, { "grad_norm": 0.4698033332824707, "learning_rate": 2.3182140197114012e-05, "loss": 0.0323, "step": 69620 }, { "grad_norm": 0.4444134533405304, "learning_rate": 2.3168186525828985e-05, "loss": 0.0403, "step": 69630 }, { "grad_norm": 0.3409711718559265, "learning_rate": 2.3154235788830757e-05, "loss": 0.0265, "step": 69640 }, { "grad_norm": 0.6360771059989929, "learning_rate": 2.3140287987644927e-05, "loss": 0.0386, "step": 69650 }, { "grad_norm": 0.2633388936519623, "learning_rate": 2.3126343123796833e-05, "loss": 0.035, "step": 69660 }, { "grad_norm": 0.35946962237358093, "learning_rate": 2.3112401198811418e-05, "loss": 0.03, "step": 69670 }, { "grad_norm": 0.4088451564311981, "learning_rate": 2.3098462214213418e-05, "loss": 0.0376, "step": 69680 }, { "grad_norm": 0.26776689291000366, "learning_rate": 2.308452617152712e-05, "loss": 0.0302, "step": 69690 }, { "grad_norm": 0.5906872153282166, "learning_rate": 2.3070593072276603e-05, "loss": 0.0371, "step": 69700 }, { "grad_norm": 0.5050833225250244, "learning_rate": 2.3056662917985518e-05, "loss": 0.0296, "step": 69710 }, { "grad_norm": 0.32284367084503174, "learning_rate": 2.3042735710177283e-05, "loss": 0.0387, "step": 69720 }, { "grad_norm": 0.23102717101573944, "learning_rate": 2.302881145037492e-05, "loss": 0.0261, "step": 69730 }, { "grad_norm": 0.2965736985206604, "learning_rate": 2.3014890140101192e-05, "loss": 0.0289, "step": 69740 }, { "grad_norm": 0.28930386900901794, "learning_rate": 2.30009717808785e-05, "loss": 0.0258, "step": 69750 }, { "grad_norm": 0.4636370539665222, "learning_rate": 2.2987056374228967e-05, "loss": 0.0336, "step": 69760 }, { "grad_norm": 0.5431278944015503, "learning_rate": 2.297314392167431e-05, "loss": 0.0267, "step": 69770 }, { "grad_norm": 0.46392661333084106, "learning_rate": 2.295923442473601e-05, "loss": 0.0281, "step": 69780 }, { "grad_norm": 0.36069250106811523, "learning_rate": 2.294532788493519e-05, "loss": 0.0324, "step": 69790 }, { "grad_norm": 0.4436236619949341, "learning_rate": 2.293142430379262e-05, "loss": 0.0346, "step": 69800 }, { "grad_norm": 0.5416415929794312, "learning_rate": 2.291752368282879e-05, "loss": 0.0361, "step": 69810 }, { "grad_norm": 0.3412292003631592, "learning_rate": 2.290362602356385e-05, "loss": 0.0319, "step": 69820 }, { "grad_norm": 0.314603328704834, "learning_rate": 2.288973132751765e-05, "loss": 0.0311, "step": 69830 }, { "grad_norm": 0.40565362572669983, "learning_rate": 2.287583959620965e-05, "loss": 0.0282, "step": 69840 }, { "grad_norm": 0.18781134486198425, "learning_rate": 2.2861950831159074e-05, "loss": 0.0361, "step": 69850 }, { "grad_norm": 0.3685792088508606, "learning_rate": 2.284806503388472e-05, "loss": 0.0341, "step": 69860 }, { "grad_norm": 0.3774217665195465, "learning_rate": 2.283418220590518e-05, "loss": 0.0269, "step": 69870 }, { "grad_norm": 0.6775707602500916, "learning_rate": 2.282030234873862e-05, "loss": 0.0354, "step": 69880 }, { "grad_norm": 0.9394699931144714, "learning_rate": 2.2806425463902948e-05, "loss": 0.0292, "step": 69890 }, { "grad_norm": 0.5134579539299011, "learning_rate": 2.279255155291568e-05, "loss": 0.0401, "step": 69900 }, { "grad_norm": 0.3091559410095215, "learning_rate": 2.277868061729409e-05, "loss": 0.0313, "step": 69910 }, { "grad_norm": 0.5046651363372803, "learning_rate": 2.276481265855504e-05, "loss": 0.0276, "step": 69920 }, { "grad_norm": 0.5971065759658813, "learning_rate": 2.2750947678215128e-05, "loss": 0.0332, "step": 69930 }, { "grad_norm": 0.6953181028366089, "learning_rate": 2.2737085677790616e-05, "loss": 0.0303, "step": 69940 }, { "grad_norm": 0.425963431596756, "learning_rate": 2.2723226658797443e-05, "loss": 0.0315, "step": 69950 }, { "grad_norm": 0.3226844370365143, "learning_rate": 2.270937062275117e-05, "loss": 0.0316, "step": 69960 }, { "grad_norm": 0.27005425095558167, "learning_rate": 2.2695517571167113e-05, "loss": 0.025, "step": 69970 }, { "grad_norm": 0.23860317468643188, "learning_rate": 2.268166750556019e-05, "loss": 0.032, "step": 69980 }, { "grad_norm": 0.2846045196056366, "learning_rate": 2.266782042744503e-05, "loss": 0.0196, "step": 69990 }, { "grad_norm": 0.5973275303840637, "learning_rate": 2.2653976338335935e-05, "loss": 0.0331, "step": 70000 }, { "grad_norm": 0.54105144739151, "learning_rate": 2.2640135239746895e-05, "loss": 0.0333, "step": 70010 }, { "grad_norm": 0.512098491191864, "learning_rate": 2.2626297133191515e-05, "loss": 0.0266, "step": 70020 }, { "grad_norm": 0.42422229051589966, "learning_rate": 2.261246202018312e-05, "loss": 0.0283, "step": 70030 }, { "grad_norm": 0.2565741539001465, "learning_rate": 2.2598629902234718e-05, "loss": 0.0269, "step": 70040 }, { "grad_norm": 0.4947255551815033, "learning_rate": 2.258480078085894e-05, "loss": 0.0331, "step": 70050 }, { "grad_norm": 0.7345004081726074, "learning_rate": 2.2570974657568126e-05, "loss": 0.04, "step": 70060 }, { "grad_norm": 0.2803264260292053, "learning_rate": 2.2557151533874286e-05, "loss": 0.0279, "step": 70070 }, { "grad_norm": 0.4135409891605377, "learning_rate": 2.2543331411289116e-05, "loss": 0.0266, "step": 70080 }, { "grad_norm": 0.4925782084465027, "learning_rate": 2.252951429132392e-05, "loss": 0.0367, "step": 70090 }, { "grad_norm": 0.2953474819660187, "learning_rate": 2.2515700175489762e-05, "loss": 0.0255, "step": 70100 }, { "grad_norm": 0.39727795124053955, "learning_rate": 2.2501889065297288e-05, "loss": 0.023, "step": 70110 }, { "grad_norm": 0.34385818243026733, "learning_rate": 2.2488080962256887e-05, "loss": 0.029, "step": 70120 }, { "grad_norm": 0.4651019275188446, "learning_rate": 2.2474275867878585e-05, "loss": 0.041, "step": 70130 }, { "grad_norm": 0.41904738545417786, "learning_rate": 2.246047378367211e-05, "loss": 0.0425, "step": 70140 }, { "grad_norm": 0.3569616973400116, "learning_rate": 2.2446674711146798e-05, "loss": 0.0287, "step": 70150 }, { "grad_norm": 0.35655438899993896, "learning_rate": 2.2432878651811734e-05, "loss": 0.036, "step": 70160 }, { "grad_norm": 0.34398216009140015, "learning_rate": 2.2419085607175594e-05, "loss": 0.0402, "step": 70170 }, { "grad_norm": 0.3994804620742798, "learning_rate": 2.240529557874678e-05, "loss": 0.0311, "step": 70180 }, { "grad_norm": 0.5084874033927917, "learning_rate": 2.2391508568033358e-05, "loss": 0.0325, "step": 70190 }, { "grad_norm": 0.394252210855484, "learning_rate": 2.237772457654307e-05, "loss": 0.0349, "step": 70200 }, { "grad_norm": 0.7158527970314026, "learning_rate": 2.236394360578327e-05, "loss": 0.0375, "step": 70210 }, { "grad_norm": 0.681697428226471, "learning_rate": 2.2350165657261073e-05, "loss": 0.0247, "step": 70220 }, { "grad_norm": 0.47145020961761475, "learning_rate": 2.2336390732483164e-05, "loss": 0.0385, "step": 70230 }, { "grad_norm": 0.2792620062828064, "learning_rate": 2.2322618832955976e-05, "loss": 0.0423, "step": 70240 }, { "grad_norm": 0.3708510398864746, "learning_rate": 2.2308849960185578e-05, "loss": 0.0307, "step": 70250 }, { "grad_norm": 0.25534874200820923, "learning_rate": 2.229508411567774e-05, "loss": 0.0287, "step": 70260 }, { "grad_norm": 0.46838125586509705, "learning_rate": 2.2281321300937823e-05, "loss": 0.0367, "step": 70270 }, { "grad_norm": 0.3138556480407715, "learning_rate": 2.2267561517470937e-05, "loss": 0.0317, "step": 70280 }, { "grad_norm": 0.551501452922821, "learning_rate": 2.2253804766781845e-05, "loss": 0.0349, "step": 70290 }, { "grad_norm": 0.5924550294876099, "learning_rate": 2.2240051050374933e-05, "loss": 0.0274, "step": 70300 }, { "grad_norm": 0.81731778383255, "learning_rate": 2.22263003697543e-05, "loss": 0.0356, "step": 70310 }, { "grad_norm": 0.2863861322402954, "learning_rate": 2.2212552726423702e-05, "loss": 0.0411, "step": 70320 }, { "grad_norm": 0.3360997438430786, "learning_rate": 2.2198808121886578e-05, "loss": 0.0269, "step": 70330 }, { "grad_norm": 0.18886303901672363, "learning_rate": 2.218506655764598e-05, "loss": 0.032, "step": 70340 }, { "grad_norm": 1.169495701789856, "learning_rate": 2.2171328035204704e-05, "loss": 0.0275, "step": 70350 }, { "grad_norm": 0.2493487000465393, "learning_rate": 2.2157592556065133e-05, "loss": 0.0256, "step": 70360 }, { "grad_norm": 0.4150797128677368, "learning_rate": 2.2143860121729382e-05, "loss": 0.0313, "step": 70370 }, { "grad_norm": 0.27472126483917236, "learning_rate": 2.2130130733699206e-05, "loss": 0.0266, "step": 70380 }, { "grad_norm": 0.43379834294319153, "learning_rate": 2.2116404393476052e-05, "loss": 0.034, "step": 70390 }, { "grad_norm": 0.46186909079551697, "learning_rate": 2.2102681102560967e-05, "loss": 0.0246, "step": 70400 }, { "grad_norm": 0.7519331574440002, "learning_rate": 2.208896086245476e-05, "loss": 0.0294, "step": 70410 }, { "grad_norm": 0.44876301288604736, "learning_rate": 2.2075243674657808e-05, "loss": 0.0371, "step": 70420 }, { "grad_norm": 0.3551342189311981, "learning_rate": 2.206152954067024e-05, "loss": 0.0288, "step": 70430 }, { "grad_norm": 0.4670572280883789, "learning_rate": 2.2047818461991758e-05, "loss": 0.0345, "step": 70440 }, { "grad_norm": 0.452245831489563, "learning_rate": 2.2034110440121864e-05, "loss": 0.0373, "step": 70450 }, { "grad_norm": 0.36063334345817566, "learning_rate": 2.2020405476559585e-05, "loss": 0.0292, "step": 70460 }, { "grad_norm": 0.3088937997817993, "learning_rate": 2.2006703572803722e-05, "loss": 0.0345, "step": 70470 }, { "grad_norm": 0.2572701573371887, "learning_rate": 2.1993004730352646e-05, "loss": 0.0268, "step": 70480 }, { "grad_norm": 0.25599968433380127, "learning_rate": 2.1979308950704464e-05, "loss": 0.025, "step": 70490 }, { "grad_norm": 0.27067315578460693, "learning_rate": 2.1965616235356923e-05, "loss": 0.0353, "step": 70500 }, { "grad_norm": 0.36695417761802673, "learning_rate": 2.1951926585807465e-05, "loss": 0.0253, "step": 70510 }, { "grad_norm": 0.2651981711387634, "learning_rate": 2.193824000355313e-05, "loss": 0.0359, "step": 70520 }, { "grad_norm": 0.6874729990959167, "learning_rate": 2.192455649009067e-05, "loss": 0.0371, "step": 70530 }, { "grad_norm": 0.4434303045272827, "learning_rate": 2.1910876046916523e-05, "loss": 0.0328, "step": 70540 }, { "grad_norm": 0.24537931382656097, "learning_rate": 2.1897198675526725e-05, "loss": 0.0297, "step": 70550 }, { "grad_norm": 0.5481587052345276, "learning_rate": 2.1883524377417024e-05, "loss": 0.0297, "step": 70560 }, { "grad_norm": 0.39839571714401245, "learning_rate": 2.1869853154082825e-05, "loss": 0.0307, "step": 70570 }, { "grad_norm": 0.27366384863853455, "learning_rate": 2.1856185007019215e-05, "loss": 0.029, "step": 70580 }, { "grad_norm": 0.45400530099868774, "learning_rate": 2.1842519937720874e-05, "loss": 0.0402, "step": 70590 }, { "grad_norm": 0.4694805443286896, "learning_rate": 2.182885794768224e-05, "loss": 0.0368, "step": 70600 }, { "grad_norm": 0.3999501168727875, "learning_rate": 2.1815199038397326e-05, "loss": 0.0364, "step": 70610 }, { "grad_norm": 0.4827049970626831, "learning_rate": 2.1801543211359893e-05, "loss": 0.0331, "step": 70620 }, { "grad_norm": 0.4542768895626068, "learning_rate": 2.178789046806326e-05, "loss": 0.0306, "step": 70630 }, { "grad_norm": 0.987542450428009, "learning_rate": 2.1774240810000547e-05, "loss": 0.0296, "step": 70640 }, { "grad_norm": 0.35057756304740906, "learning_rate": 2.17605942386644e-05, "loss": 0.0357, "step": 70650 }, { "grad_norm": 0.37308141589164734, "learning_rate": 2.174695075554723e-05, "loss": 0.0416, "step": 70660 }, { "grad_norm": 0.4943369925022125, "learning_rate": 2.173331036214103e-05, "loss": 0.043, "step": 70670 }, { "grad_norm": 0.3872893750667572, "learning_rate": 2.171967305993753e-05, "loss": 0.0294, "step": 70680 }, { "grad_norm": 0.32615920901298523, "learning_rate": 2.1706038850428024e-05, "loss": 0.0308, "step": 70690 }, { "grad_norm": 0.4420587420463562, "learning_rate": 2.169240773510361e-05, "loss": 0.0235, "step": 70700 }, { "grad_norm": 0.29804232716560364, "learning_rate": 2.1678779715454906e-05, "loss": 0.0325, "step": 70710 }, { "grad_norm": 0.3803504407405853, "learning_rate": 2.166515479297229e-05, "loss": 0.0258, "step": 70720 }, { "grad_norm": 0.527521550655365, "learning_rate": 2.1651532969145732e-05, "loss": 0.0353, "step": 70730 }, { "grad_norm": 0.22485226392745972, "learning_rate": 2.1637914245464903e-05, "loss": 0.0276, "step": 70740 }, { "grad_norm": 0.4177837669849396, "learning_rate": 2.1624298623419133e-05, "loss": 0.0289, "step": 70750 }, { "grad_norm": 0.62575763463974, "learning_rate": 2.1610686104497413e-05, "loss": 0.0343, "step": 70760 }, { "grad_norm": 0.6731492280960083, "learning_rate": 2.1597076690188366e-05, "loss": 0.0393, "step": 70770 }, { "grad_norm": 0.4670650064945221, "learning_rate": 2.1583470381980315e-05, "loss": 0.0453, "step": 70780 }, { "grad_norm": 0.2891165018081665, "learning_rate": 2.156986718136123e-05, "loss": 0.0295, "step": 70790 }, { "grad_norm": 0.3664488196372986, "learning_rate": 2.155626708981871e-05, "loss": 0.0305, "step": 70800 }, { "grad_norm": 0.31562331318855286, "learning_rate": 2.1542670108840075e-05, "loss": 0.0349, "step": 70810 }, { "grad_norm": 0.3842737674713135, "learning_rate": 2.152907623991221e-05, "loss": 0.0384, "step": 70820 }, { "grad_norm": 0.26475027203559875, "learning_rate": 2.1515485484521802e-05, "loss": 0.0275, "step": 70830 }, { "grad_norm": 0.29690003395080566, "learning_rate": 2.150189784415506e-05, "loss": 0.0339, "step": 70840 }, { "grad_norm": 0.42865461111068726, "learning_rate": 2.148831332029794e-05, "loss": 0.0265, "step": 70850 }, { "grad_norm": 0.19368129968643188, "learning_rate": 2.147473191443599e-05, "loss": 0.0272, "step": 70860 }, { "grad_norm": 0.29911309480667114, "learning_rate": 2.1461153628054492e-05, "loss": 0.0289, "step": 70870 }, { "grad_norm": 0.6008530259132385, "learning_rate": 2.1447578462638284e-05, "loss": 0.032, "step": 70880 }, { "grad_norm": 0.4117524325847626, "learning_rate": 2.143400641967201e-05, "loss": 0.0253, "step": 70890 }, { "grad_norm": 0.22853372991085052, "learning_rate": 2.1420437500639823e-05, "loss": 0.0243, "step": 70900 }, { "grad_norm": 0.38791537284851074, "learning_rate": 2.1406871707025646e-05, "loss": 0.0291, "step": 70910 }, { "grad_norm": 0.5629974603652954, "learning_rate": 2.1393309040312963e-05, "loss": 0.0385, "step": 70920 }, { "grad_norm": 0.33205682039260864, "learning_rate": 2.137974950198502e-05, "loss": 0.036, "step": 70930 }, { "grad_norm": 0.2805483937263489, "learning_rate": 2.1366193093524617e-05, "loss": 0.0249, "step": 70940 }, { "grad_norm": 0.7816116213798523, "learning_rate": 2.1352639816414288e-05, "loss": 0.0309, "step": 70950 }, { "grad_norm": 0.2765555679798126, "learning_rate": 2.1339089672136202e-05, "loss": 0.033, "step": 70960 }, { "grad_norm": 0.3040235936641693, "learning_rate": 2.1325542662172193e-05, "loss": 0.0383, "step": 70970 }, { "grad_norm": 0.30131295323371887, "learning_rate": 2.1311998788003712e-05, "loss": 0.0266, "step": 70980 }, { "grad_norm": 0.6849955916404724, "learning_rate": 2.1298458051111907e-05, "loss": 0.0409, "step": 70990 }, { "grad_norm": 0.2287452667951584, "learning_rate": 2.1284920452977602e-05, "loss": 0.0265, "step": 71000 }, { "grad_norm": 0.24951568245887756, "learning_rate": 2.12713859950812e-05, "loss": 0.0259, "step": 71010 }, { "grad_norm": 0.42117446660995483, "learning_rate": 2.1257854678902832e-05, "loss": 0.0248, "step": 71020 }, { "grad_norm": 0.45593535900115967, "learning_rate": 2.124432650592227e-05, "loss": 0.0369, "step": 71030 }, { "grad_norm": 0.5897288918495178, "learning_rate": 2.1230801477618937e-05, "loss": 0.0349, "step": 71040 }, { "grad_norm": 0.36447834968566895, "learning_rate": 2.121727959547189e-05, "loss": 0.0334, "step": 71050 }, { "grad_norm": 0.5490701794624329, "learning_rate": 2.1203760860959886e-05, "loss": 0.0323, "step": 71060 }, { "grad_norm": 0.5245676636695862, "learning_rate": 2.119024527556127e-05, "loss": 0.0436, "step": 71070 }, { "grad_norm": 0.38453546166419983, "learning_rate": 2.117673284075415e-05, "loss": 0.0244, "step": 71080 }, { "grad_norm": 0.7213836908340454, "learning_rate": 2.1163223558016165e-05, "loss": 0.0325, "step": 71090 }, { "grad_norm": 0.3620568811893463, "learning_rate": 2.1149717428824722e-05, "loss": 0.0335, "step": 71100 }, { "grad_norm": 0.535908043384552, "learning_rate": 2.1136214454656784e-05, "loss": 0.0334, "step": 71110 }, { "grad_norm": 0.6564407348632812, "learning_rate": 2.1122714636989056e-05, "loss": 0.0401, "step": 71120 }, { "grad_norm": 0.8727279305458069, "learning_rate": 2.1109217977297823e-05, "loss": 0.0325, "step": 71130 }, { "grad_norm": 1.9734190702438354, "learning_rate": 2.1095724477059077e-05, "loss": 0.0281, "step": 71140 }, { "grad_norm": 0.19895076751708984, "learning_rate": 2.1082234137748434e-05, "loss": 0.0226, "step": 71150 }, { "grad_norm": 0.391414999961853, "learning_rate": 2.106874696084122e-05, "loss": 0.0337, "step": 71160 }, { "grad_norm": 0.42153453826904297, "learning_rate": 2.1055262947812315e-05, "loss": 0.0328, "step": 71170 }, { "grad_norm": 0.41657719016075134, "learning_rate": 2.104178210013636e-05, "loss": 0.0434, "step": 71180 }, { "grad_norm": 0.946567714214325, "learning_rate": 2.1028304419287554e-05, "loss": 0.0372, "step": 71190 }, { "grad_norm": 0.4108660817146301, "learning_rate": 2.101482990673982e-05, "loss": 0.0322, "step": 71200 }, { "grad_norm": 0.3860214650630951, "learning_rate": 2.1001358563966707e-05, "loss": 0.035, "step": 71210 }, { "grad_norm": 0.34888628125190735, "learning_rate": 2.098789039244144e-05, "loss": 0.0291, "step": 71220 }, { "grad_norm": 0.29275545477867126, "learning_rate": 2.097442539363685e-05, "loss": 0.0322, "step": 71230 }, { "grad_norm": 0.26020577549934387, "learning_rate": 2.0960963569025448e-05, "loss": 0.0322, "step": 71240 }, { "grad_norm": 0.1715526431798935, "learning_rate": 2.0947504920079435e-05, "loss": 0.0259, "step": 71250 }, { "grad_norm": 0.33657437562942505, "learning_rate": 2.093404944827059e-05, "loss": 0.0318, "step": 71260 }, { "grad_norm": 0.37163376808166504, "learning_rate": 2.0920597155070394e-05, "loss": 0.0291, "step": 71270 }, { "grad_norm": 0.6983038783073425, "learning_rate": 2.090714804194997e-05, "loss": 0.0474, "step": 71280 }, { "grad_norm": 0.29922279715538025, "learning_rate": 2.0893702110380114e-05, "loss": 0.0275, "step": 71290 }, { "grad_norm": 0.4047894775867462, "learning_rate": 2.0880259361831216e-05, "loss": 0.0275, "step": 71300 }, { "grad_norm": 0.3841918408870697, "learning_rate": 2.0866819797773397e-05, "loss": 0.0291, "step": 71310 }, { "grad_norm": 0.3197802007198334, "learning_rate": 2.085338341967634e-05, "loss": 0.0304, "step": 71320 }, { "grad_norm": 0.3137357234954834, "learning_rate": 2.083995022900946e-05, "loss": 0.0303, "step": 71330 }, { "grad_norm": 0.4132225811481476, "learning_rate": 2.0826520227241768e-05, "loss": 0.0326, "step": 71340 }, { "grad_norm": 0.4272098243236542, "learning_rate": 2.081309341584199e-05, "loss": 0.0323, "step": 71350 }, { "grad_norm": 0.3473767042160034, "learning_rate": 2.0799669796278413e-05, "loss": 0.0295, "step": 71360 }, { "grad_norm": 0.2883583605289459, "learning_rate": 2.078624937001906e-05, "loss": 0.0371, "step": 71370 }, { "grad_norm": 0.45997413992881775, "learning_rate": 2.0772832138531544e-05, "loss": 0.0526, "step": 71380 }, { "grad_norm": 0.3969917893409729, "learning_rate": 2.0759418103283158e-05, "loss": 0.0329, "step": 71390 }, { "grad_norm": 1.081392526626587, "learning_rate": 2.0746007265740846e-05, "loss": 0.0274, "step": 71400 }, { "grad_norm": 0.33050480484962463, "learning_rate": 2.0732599627371215e-05, "loss": 0.0322, "step": 71410 }, { "grad_norm": 0.42459025979042053, "learning_rate": 2.0719195189640468e-05, "loss": 0.0317, "step": 71420 }, { "grad_norm": 0.37885528802871704, "learning_rate": 2.0705793954014528e-05, "loss": 0.0266, "step": 71430 }, { "grad_norm": 0.3371731638908386, "learning_rate": 2.06923959219589e-05, "loss": 0.027, "step": 71440 }, { "grad_norm": 0.3922233581542969, "learning_rate": 2.0679001094938793e-05, "loss": 0.027, "step": 71450 }, { "grad_norm": 0.45017924904823303, "learning_rate": 2.066560947441904e-05, "loss": 0.0354, "step": 71460 }, { "grad_norm": 0.33544597029685974, "learning_rate": 2.0652221061864152e-05, "loss": 0.0219, "step": 71470 }, { "grad_norm": 0.37512093782424927, "learning_rate": 2.0638835858738226e-05, "loss": 0.0364, "step": 71480 }, { "grad_norm": 0.46878430247306824, "learning_rate": 2.0625453866505068e-05, "loss": 0.0294, "step": 71490 }, { "grad_norm": 0.25782546401023865, "learning_rate": 2.0612075086628123e-05, "loss": 0.028, "step": 71500 }, { "grad_norm": 0.37455642223358154, "learning_rate": 2.059869952057045e-05, "loss": 0.0283, "step": 71510 }, { "grad_norm": 0.7447521090507507, "learning_rate": 2.0585327169794793e-05, "loss": 0.0312, "step": 71520 }, { "grad_norm": 0.36973172426223755, "learning_rate": 2.0571958035763534e-05, "loss": 0.0298, "step": 71530 }, { "grad_norm": 0.6656071543693542, "learning_rate": 2.0558592119938715e-05, "loss": 0.0336, "step": 71540 }, { "grad_norm": 0.3372868001461029, "learning_rate": 2.054522942378198e-05, "loss": 0.0271, "step": 71550 }, { "grad_norm": 0.8336663246154785, "learning_rate": 2.053186994875469e-05, "loss": 0.0339, "step": 71560 }, { "grad_norm": 0.3778628408908844, "learning_rate": 2.0518513696317775e-05, "loss": 0.0361, "step": 71570 }, { "grad_norm": 0.3529946506023407, "learning_rate": 2.050516066793188e-05, "loss": 0.0312, "step": 71580 }, { "grad_norm": 0.2916669249534607, "learning_rate": 2.0491810865057276e-05, "loss": 0.0255, "step": 71590 }, { "grad_norm": 0.4494417607784271, "learning_rate": 2.0478464289153882e-05, "loss": 0.0352, "step": 71600 }, { "grad_norm": 0.978377640247345, "learning_rate": 2.0465120941681232e-05, "loss": 0.0354, "step": 71610 }, { "grad_norm": 0.40398141741752625, "learning_rate": 2.0451780824098566e-05, "loss": 0.0254, "step": 71620 }, { "grad_norm": 0.27943310141563416, "learning_rate": 2.043844393786471e-05, "loss": 0.0271, "step": 71630 }, { "grad_norm": 0.8332996368408203, "learning_rate": 2.0425110284438193e-05, "loss": 0.0307, "step": 71640 }, { "grad_norm": 0.4694962501525879, "learning_rate": 2.041177986527712e-05, "loss": 0.0328, "step": 71650 }, { "grad_norm": 0.5231326818466187, "learning_rate": 2.0398452681839347e-05, "loss": 0.0269, "step": 71660 }, { "grad_norm": 0.4225291311740875, "learning_rate": 2.0385128735582272e-05, "loss": 0.0429, "step": 71670 }, { "grad_norm": 0.3030240535736084, "learning_rate": 2.0371808027963003e-05, "loss": 0.0228, "step": 71680 }, { "grad_norm": 0.48819640278816223, "learning_rate": 2.035849056043825e-05, "loss": 0.0406, "step": 71690 }, { "grad_norm": 0.2357865273952484, "learning_rate": 2.03451763344644e-05, "loss": 0.0245, "step": 71700 }, { "grad_norm": 0.35757333040237427, "learning_rate": 2.0331865351497482e-05, "loss": 0.0316, "step": 71710 }, { "grad_norm": 0.42241716384887695, "learning_rate": 2.0318557612993185e-05, "loss": 0.0287, "step": 71720 }, { "grad_norm": 0.4690009355545044, "learning_rate": 2.0305253120406782e-05, "loss": 0.033, "step": 71730 }, { "grad_norm": 0.25304287672042847, "learning_rate": 2.029195187519325e-05, "loss": 0.0253, "step": 71740 }, { "grad_norm": 0.2718474566936493, "learning_rate": 2.0278653878807217e-05, "loss": 0.0309, "step": 71750 }, { "grad_norm": 0.301334947347641, "learning_rate": 2.0265359132702892e-05, "loss": 0.038, "step": 71760 }, { "grad_norm": 0.7865338921546936, "learning_rate": 2.0252067638334182e-05, "loss": 0.0297, "step": 71770 }, { "grad_norm": 0.44182270765304565, "learning_rate": 2.023877939715464e-05, "loss": 0.0359, "step": 71780 }, { "grad_norm": 0.3267475366592407, "learning_rate": 2.0225494410617448e-05, "loss": 0.0312, "step": 71790 }, { "grad_norm": 0.3175681531429291, "learning_rate": 2.02122126801754e-05, "loss": 0.0344, "step": 71800 }, { "grad_norm": 0.30303671956062317, "learning_rate": 2.0198934207281018e-05, "loss": 0.0315, "step": 71810 }, { "grad_norm": 0.28697481751441956, "learning_rate": 2.0185658993386357e-05, "loss": 0.0261, "step": 71820 }, { "grad_norm": 0.4995671808719635, "learning_rate": 2.0172387039943226e-05, "loss": 0.0289, "step": 71830 }, { "grad_norm": 0.25136685371398926, "learning_rate": 2.0159118348402968e-05, "loss": 0.0346, "step": 71840 }, { "grad_norm": 0.5481020212173462, "learning_rate": 2.0145852920216697e-05, "loss": 0.0309, "step": 71850 }, { "grad_norm": 0.632993757724762, "learning_rate": 2.0132590756835045e-05, "loss": 0.031, "step": 71860 }, { "grad_norm": 0.3403113782405853, "learning_rate": 2.0119331859708386e-05, "loss": 0.0432, "step": 71870 }, { "grad_norm": 0.3035496771335602, "learning_rate": 2.0106076230286654e-05, "loss": 0.0301, "step": 71880 }, { "grad_norm": 0.3243783712387085, "learning_rate": 2.0092823870019488e-05, "loss": 0.0298, "step": 71890 }, { "grad_norm": 0.26513057947158813, "learning_rate": 2.0079574780356115e-05, "loss": 0.0247, "step": 71900 }, { "grad_norm": 0.36802390217781067, "learning_rate": 2.0066328962745486e-05, "loss": 0.032, "step": 71910 }, { "grad_norm": 0.5251895189285278, "learning_rate": 2.00530864186361e-05, "loss": 0.0254, "step": 71920 }, { "grad_norm": 0.4382905066013336, "learning_rate": 2.003984714947618e-05, "loss": 0.0393, "step": 71930 }, { "grad_norm": 0.4603225886821747, "learning_rate": 2.0026611156713515e-05, "loss": 0.0294, "step": 71940 }, { "grad_norm": 0.565665602684021, "learning_rate": 2.0013378441795584e-05, "loss": 0.0305, "step": 71950 }, { "grad_norm": 0.3706866502761841, "learning_rate": 2.0000149006169506e-05, "loss": 0.0263, "step": 71960 }, { "grad_norm": 0.32549184560775757, "learning_rate": 1.9986922851282048e-05, "loss": 0.0325, "step": 71970 }, { "grad_norm": 0.30935153365135193, "learning_rate": 1.997369997857956e-05, "loss": 0.0322, "step": 71980 }, { "grad_norm": 0.4877069890499115, "learning_rate": 1.9960480389508103e-05, "loss": 0.0294, "step": 71990 }, { "grad_norm": 0.34813401103019714, "learning_rate": 1.9947264085513363e-05, "loss": 0.0464, "step": 72000 }, { "grad_norm": 0.46594640612602234, "learning_rate": 1.993405106804062e-05, "loss": 0.0299, "step": 72010 }, { "grad_norm": 0.8388652205467224, "learning_rate": 1.9920841338534867e-05, "loss": 0.0384, "step": 72020 }, { "grad_norm": 0.39192265272140503, "learning_rate": 1.9907634898440646e-05, "loss": 0.0257, "step": 72030 }, { "grad_norm": 0.3936428725719452, "learning_rate": 1.989443174920227e-05, "loss": 0.0327, "step": 72040 }, { "grad_norm": 0.42728137969970703, "learning_rate": 1.988123189226355e-05, "loss": 0.0422, "step": 72050 }, { "grad_norm": 0.35644859075546265, "learning_rate": 1.986803532906805e-05, "loss": 0.0367, "step": 72060 }, { "grad_norm": 0.49838170409202576, "learning_rate": 1.985484206105889e-05, "loss": 0.0316, "step": 72070 }, { "grad_norm": 0.37748998403549194, "learning_rate": 1.9841652089678887e-05, "loss": 0.0252, "step": 72080 }, { "grad_norm": 0.4014721214771271, "learning_rate": 1.9828465416370435e-05, "loss": 0.0383, "step": 72090 }, { "grad_norm": 0.41693010926246643, "learning_rate": 1.9815282042575685e-05, "loss": 0.0327, "step": 72100 }, { "grad_norm": 0.15630845725536346, "learning_rate": 1.9802101969736285e-05, "loss": 0.0362, "step": 72110 }, { "grad_norm": 0.5081511735916138, "learning_rate": 1.9788925199293633e-05, "loss": 0.0354, "step": 72120 }, { "grad_norm": 0.36926668882369995, "learning_rate": 1.977575173268868e-05, "loss": 0.0308, "step": 72130 }, { "grad_norm": 0.40110617876052856, "learning_rate": 1.9762581571362094e-05, "loss": 0.0395, "step": 72140 }, { "grad_norm": 0.32511022686958313, "learning_rate": 1.9749414716754112e-05, "loss": 0.0317, "step": 72150 }, { "grad_norm": 0.318558007478714, "learning_rate": 1.973625117030465e-05, "loss": 0.0286, "step": 72160 }, { "grad_norm": 1.3760454654693604, "learning_rate": 1.972309093345326e-05, "loss": 0.0276, "step": 72170 }, { "grad_norm": 0.5578183531761169, "learning_rate": 1.970993400763915e-05, "loss": 0.0343, "step": 72180 }, { "grad_norm": 0.4059162735939026, "learning_rate": 1.969678039430109e-05, "loss": 0.0382, "step": 72190 }, { "grad_norm": 0.49679917097091675, "learning_rate": 1.9683630094877564e-05, "loss": 0.0297, "step": 72200 }, { "grad_norm": 0.36456504464149475, "learning_rate": 1.9670483110806696e-05, "loss": 0.031, "step": 72210 }, { "grad_norm": 1.0065901279449463, "learning_rate": 1.9657339443526172e-05, "loss": 0.037, "step": 72220 }, { "grad_norm": 0.2590215802192688, "learning_rate": 1.964419909447339e-05, "loss": 0.0353, "step": 72230 }, { "grad_norm": 0.3396596610546112, "learning_rate": 1.963106206508536e-05, "loss": 0.0328, "step": 72240 }, { "grad_norm": 0.2811385691165924, "learning_rate": 1.9617928356798738e-05, "loss": 0.0264, "step": 72250 }, { "grad_norm": 0.33961474895477295, "learning_rate": 1.960479797104977e-05, "loss": 0.03, "step": 72260 }, { "grad_norm": 0.8610458970069885, "learning_rate": 1.9591670909274418e-05, "loss": 0.0291, "step": 72270 }, { "grad_norm": 0.2184806615114212, "learning_rate": 1.9578547172908184e-05, "loss": 0.0364, "step": 72280 }, { "grad_norm": 0.47915318608283997, "learning_rate": 1.9565426763386325e-05, "loss": 0.0297, "step": 72290 }, { "grad_norm": 0.4360857903957367, "learning_rate": 1.9552309682143622e-05, "loss": 0.0267, "step": 72300 }, { "grad_norm": 0.2799319326877594, "learning_rate": 1.9539195930614574e-05, "loss": 0.0274, "step": 72310 }, { "grad_norm": 0.7290898561477661, "learning_rate": 1.952608551023324e-05, "loss": 0.0306, "step": 72320 }, { "grad_norm": 0.4732518196105957, "learning_rate": 1.9512978422433397e-05, "loss": 0.0297, "step": 72330 }, { "grad_norm": 0.26508665084838867, "learning_rate": 1.9499874668648376e-05, "loss": 0.0244, "step": 72340 }, { "grad_norm": 0.33763498067855835, "learning_rate": 1.9486774250311213e-05, "loss": 0.0275, "step": 72350 }, { "grad_norm": 0.41006460785865784, "learning_rate": 1.9473677168854537e-05, "loss": 0.0332, "step": 72360 }, { "grad_norm": 0.32736241817474365, "learning_rate": 1.946058342571065e-05, "loss": 0.0355, "step": 72370 }, { "grad_norm": 0.2838347256183624, "learning_rate": 1.9447493022311424e-05, "loss": 0.0314, "step": 72380 }, { "grad_norm": 0.4871814250946045, "learning_rate": 1.9434405960088442e-05, "loss": 0.0269, "step": 72390 }, { "grad_norm": 0.3339018225669861, "learning_rate": 1.9421322240472857e-05, "loss": 0.0278, "step": 72400 }, { "grad_norm": 0.387844443321228, "learning_rate": 1.9408241864895494e-05, "loss": 0.03, "step": 72410 }, { "grad_norm": 0.5791569352149963, "learning_rate": 1.939516483478681e-05, "loss": 0.0274, "step": 72420 }, { "grad_norm": 0.7013188004493713, "learning_rate": 1.9382091151576902e-05, "loss": 0.0277, "step": 72430 }, { "grad_norm": 0.31239748001098633, "learning_rate": 1.9369020816695454e-05, "loss": 0.0251, "step": 72440 }, { "grad_norm": 0.23780560493469238, "learning_rate": 1.935595383157184e-05, "loss": 0.0274, "step": 72450 }, { "grad_norm": 0.47548753023147583, "learning_rate": 1.9342890197635057e-05, "loss": 0.0234, "step": 72460 }, { "grad_norm": 0.3328304886817932, "learning_rate": 1.9329829916313685e-05, "loss": 0.0352, "step": 72470 }, { "grad_norm": 0.5417466163635254, "learning_rate": 1.9316772989036013e-05, "loss": 0.0331, "step": 72480 }, { "grad_norm": 0.40095454454421997, "learning_rate": 1.930371941722991e-05, "loss": 0.0276, "step": 72490 }, { "grad_norm": 0.25200343132019043, "learning_rate": 1.929066920232292e-05, "loss": 0.0269, "step": 72500 }, { "grad_norm": 0.4859490394592285, "learning_rate": 1.9277622345742152e-05, "loss": 0.0367, "step": 72510 }, { "grad_norm": 0.46319136023521423, "learning_rate": 1.926457884891444e-05, "loss": 0.0338, "step": 72520 }, { "grad_norm": 0.38110843300819397, "learning_rate": 1.9251538713266147e-05, "loss": 0.0267, "step": 72530 }, { "grad_norm": 0.48345887660980225, "learning_rate": 1.9238501940223358e-05, "loss": 0.0331, "step": 72540 }, { "grad_norm": 0.5869225859642029, "learning_rate": 1.922546853121175e-05, "loss": 0.0369, "step": 72550 }, { "grad_norm": 0.25420039892196655, "learning_rate": 1.9212438487656647e-05, "loss": 0.0279, "step": 72560 }, { "grad_norm": 0.19849368929862976, "learning_rate": 1.9199411810982965e-05, "loss": 0.0293, "step": 72570 }, { "grad_norm": 0.21919575333595276, "learning_rate": 1.918638850261532e-05, "loss": 0.0336, "step": 72580 }, { "grad_norm": 0.3129495680332184, "learning_rate": 1.9173368563977885e-05, "loss": 0.0316, "step": 72590 }, { "grad_norm": 0.4154224097728729, "learning_rate": 1.9160351996494513e-05, "loss": 0.0329, "step": 72600 }, { "grad_norm": 0.316373348236084, "learning_rate": 1.9147338801588677e-05, "loss": 0.0299, "step": 72610 }, { "grad_norm": 0.5825164318084717, "learning_rate": 1.913432898068351e-05, "loss": 0.0369, "step": 72620 }, { "grad_norm": 0.32449808716773987, "learning_rate": 1.912132253520169e-05, "loss": 0.0396, "step": 72630 }, { "grad_norm": 0.4869816303253174, "learning_rate": 1.910831946656564e-05, "loss": 0.0435, "step": 72640 }, { "grad_norm": 0.4267214238643646, "learning_rate": 1.909531977619731e-05, "loss": 0.0438, "step": 72650 }, { "grad_norm": 0.5301716923713684, "learning_rate": 1.908232346551834e-05, "loss": 0.0255, "step": 72660 }, { "grad_norm": 0.49330440163612366, "learning_rate": 1.9069330535949993e-05, "loss": 0.0395, "step": 72670 }, { "grad_norm": 0.526377260684967, "learning_rate": 1.9056340988913167e-05, "loss": 0.0252, "step": 72680 }, { "grad_norm": 0.5670815110206604, "learning_rate": 1.9043354825828357e-05, "loss": 0.0346, "step": 72690 }, { "grad_norm": 0.3193839192390442, "learning_rate": 1.9030372048115724e-05, "loss": 0.0339, "step": 72700 }, { "grad_norm": 0.4234316647052765, "learning_rate": 1.9017392657195056e-05, "loss": 0.0394, "step": 72710 }, { "grad_norm": 0.4903032183647156, "learning_rate": 1.9004416654485724e-05, "loss": 0.035, "step": 72720 }, { "grad_norm": 0.43850255012512207, "learning_rate": 1.8991444041406785e-05, "loss": 0.0374, "step": 72730 }, { "grad_norm": 0.5180771350860596, "learning_rate": 1.8978474819376912e-05, "loss": 0.0336, "step": 72740 }, { "grad_norm": 0.3924383223056793, "learning_rate": 1.8965508989814406e-05, "loss": 0.0312, "step": 72750 }, { "grad_norm": 0.5959470868110657, "learning_rate": 1.8952546554137158e-05, "loss": 0.0329, "step": 72760 }, { "grad_norm": 0.29593729972839355, "learning_rate": 1.8939587513762757e-05, "loss": 0.0246, "step": 72770 }, { "grad_norm": 0.40592294931411743, "learning_rate": 1.892663187010835e-05, "loss": 0.0274, "step": 72780 }, { "grad_norm": 0.4963058531284332, "learning_rate": 1.891367962459077e-05, "loss": 0.0296, "step": 72790 }, { "grad_norm": 0.33644983172416687, "learning_rate": 1.890073077862644e-05, "loss": 0.0305, "step": 72800 }, { "grad_norm": 0.38150423765182495, "learning_rate": 1.888778533363146e-05, "loss": 0.0285, "step": 72810 }, { "grad_norm": 0.2767121493816376, "learning_rate": 1.8874843291021478e-05, "loss": 0.0258, "step": 72820 }, { "grad_norm": 0.3375595211982727, "learning_rate": 1.8861904652211853e-05, "loss": 0.0307, "step": 72830 }, { "grad_norm": 0.4628118872642517, "learning_rate": 1.8848969418617507e-05, "loss": 0.0296, "step": 72840 }, { "grad_norm": 0.4225011169910431, "learning_rate": 1.8836037591653044e-05, "loss": 0.0308, "step": 72850 }, { "grad_norm": 0.5130951404571533, "learning_rate": 1.882310917273262e-05, "loss": 0.0314, "step": 72860 }, { "grad_norm": 1.8982570171356201, "learning_rate": 1.881018416327014e-05, "loss": 0.0381, "step": 72870 }, { "grad_norm": 0.3010018467903137, "learning_rate": 1.8797262564679014e-05, "loss": 0.0352, "step": 72880 }, { "grad_norm": 0.42005646228790283, "learning_rate": 1.8784344378372355e-05, "loss": 0.032, "step": 72890 }, { "grad_norm": 0.36589762568473816, "learning_rate": 1.8771429605762843e-05, "loss": 0.0311, "step": 72900 }, { "grad_norm": 0.36190155148506165, "learning_rate": 1.8758518248262835e-05, "loss": 0.03, "step": 72910 }, { "grad_norm": 0.36698371171951294, "learning_rate": 1.8745610307284305e-05, "loss": 0.0378, "step": 72920 }, { "grad_norm": 0.556293249130249, "learning_rate": 1.8732705784238862e-05, "loss": 0.029, "step": 72930 }, { "grad_norm": 0.436867356300354, "learning_rate": 1.8719804680537682e-05, "loss": 0.034, "step": 72940 }, { "grad_norm": 0.42335835099220276, "learning_rate": 1.870690699759164e-05, "loss": 0.0334, "step": 72950 }, { "grad_norm": 0.2583240568637848, "learning_rate": 1.8694012736811213e-05, "loss": 0.0302, "step": 72960 }, { "grad_norm": 0.4262329936027527, "learning_rate": 1.8681121899606463e-05, "loss": 0.0285, "step": 72970 }, { "grad_norm": 0.31966009736061096, "learning_rate": 1.8668234487387144e-05, "loss": 0.0295, "step": 72980 }, { "grad_norm": 0.19918890297412872, "learning_rate": 1.8655350501562592e-05, "loss": 0.0253, "step": 72990 }, { "grad_norm": 0.3478478789329529, "learning_rate": 1.8642469943541796e-05, "loss": 0.0379, "step": 73000 }, { "grad_norm": 0.2688131332397461, "learning_rate": 1.8629592814733328e-05, "loss": 0.0297, "step": 73010 }, { "grad_norm": 0.39968401193618774, "learning_rate": 1.8616719116545438e-05, "loss": 0.0384, "step": 73020 }, { "grad_norm": 0.28600096702575684, "learning_rate": 1.8603848850385952e-05, "loss": 0.0264, "step": 73030 }, { "grad_norm": 0.4483305811882019, "learning_rate": 1.859098201766236e-05, "loss": 0.0249, "step": 73040 }, { "grad_norm": 0.342649906873703, "learning_rate": 1.8578118619781726e-05, "loss": 0.0265, "step": 73050 }, { "grad_norm": 0.22334051132202148, "learning_rate": 1.856525865815083e-05, "loss": 0.0258, "step": 73060 }, { "grad_norm": 0.25230100750923157, "learning_rate": 1.8552402134175966e-05, "loss": 0.0264, "step": 73070 }, { "grad_norm": 0.3203323483467102, "learning_rate": 1.8539549049263143e-05, "loss": 0.0272, "step": 73080 }, { "grad_norm": 0.25185370445251465, "learning_rate": 1.8526699404817917e-05, "loss": 0.0264, "step": 73090 }, { "grad_norm": 0.6006633639335632, "learning_rate": 1.8513853202245546e-05, "loss": 0.0305, "step": 73100 }, { "grad_norm": 0.34284740686416626, "learning_rate": 1.850101044295081e-05, "loss": 0.0304, "step": 73110 }, { "grad_norm": 0.43133777379989624, "learning_rate": 1.848817112833825e-05, "loss": 0.0208, "step": 73120 }, { "grad_norm": 1.16267728805542, "learning_rate": 1.8475335259811905e-05, "loss": 0.03, "step": 73130 }, { "grad_norm": 0.45657485723495483, "learning_rate": 1.8462502838775508e-05, "loss": 0.0369, "step": 73140 }, { "grad_norm": 0.35477811098098755, "learning_rate": 1.8449673866632378e-05, "loss": 0.0348, "step": 73150 }, { "grad_norm": 0.27742284536361694, "learning_rate": 1.8436848344785473e-05, "loss": 0.0343, "step": 73160 }, { "grad_norm": 0.3324050009250641, "learning_rate": 1.8424026274637378e-05, "loss": 0.032, "step": 73170 }, { "grad_norm": 0.29872193932533264, "learning_rate": 1.8411207657590312e-05, "loss": 0.0324, "step": 73180 }, { "grad_norm": 0.2055845409631729, "learning_rate": 1.839839249504607e-05, "loss": 0.0277, "step": 73190 }, { "grad_norm": 0.37773558497428894, "learning_rate": 1.8385580788406108e-05, "loss": 0.0345, "step": 73200 }, { "grad_norm": 0.3047175407409668, "learning_rate": 1.837277253907152e-05, "loss": 0.0334, "step": 73210 }, { "grad_norm": 0.3658122718334198, "learning_rate": 1.8359967748442957e-05, "loss": 0.029, "step": 73220 }, { "grad_norm": 0.416568398475647, "learning_rate": 1.8347166417920774e-05, "loss": 0.026, "step": 73230 }, { "grad_norm": 0.47000131011009216, "learning_rate": 1.833436854890484e-05, "loss": 0.039, "step": 73240 }, { "grad_norm": 0.4300559461116791, "learning_rate": 1.8321574142794796e-05, "loss": 0.0245, "step": 73250 }, { "grad_norm": 0.37326985597610474, "learning_rate": 1.8308783200989748e-05, "loss": 0.027, "step": 73260 }, { "grad_norm": 0.3007744252681732, "learning_rate": 1.829599572488855e-05, "loss": 0.028, "step": 73270 }, { "grad_norm": 0.23859167098999023, "learning_rate": 1.828321171588957e-05, "loss": 0.0186, "step": 73280 }, { "grad_norm": 0.5033634305000305, "learning_rate": 1.827043117539089e-05, "loss": 0.0327, "step": 73290 }, { "grad_norm": 0.8209884166717529, "learning_rate": 1.8257654104790116e-05, "loss": 0.0288, "step": 73300 }, { "grad_norm": 0.6234678030014038, "learning_rate": 1.82448805054846e-05, "loss": 0.0354, "step": 73310 }, { "grad_norm": 0.5404093265533447, "learning_rate": 1.823211037887119e-05, "loss": 0.0376, "step": 73320 }, { "grad_norm": 0.378484845161438, "learning_rate": 1.821934372634645e-05, "loss": 0.0297, "step": 73330 }, { "grad_norm": 0.3936889171600342, "learning_rate": 1.8206580549306472e-05, "loss": 0.0265, "step": 73340 }, { "grad_norm": 0.22994250059127808, "learning_rate": 1.819382084914707e-05, "loss": 0.0312, "step": 73350 }, { "grad_norm": 0.3030118942260742, "learning_rate": 1.8181064627263577e-05, "loss": 0.024, "step": 73360 }, { "grad_norm": 0.4699181318283081, "learning_rate": 1.8168311885051022e-05, "loss": 0.036, "step": 73370 }, { "grad_norm": 0.3429650068283081, "learning_rate": 1.815556262390402e-05, "loss": 0.0226, "step": 73380 }, { "grad_norm": 0.28019002079963684, "learning_rate": 1.8142816845216826e-05, "loss": 0.0325, "step": 73390 }, { "grad_norm": 0.31660234928131104, "learning_rate": 1.8130074550383276e-05, "loss": 0.0265, "step": 73400 }, { "grad_norm": 0.8684329986572266, "learning_rate": 1.8117335740796848e-05, "loss": 0.0307, "step": 73410 }, { "grad_norm": 0.5560134053230286, "learning_rate": 1.810460041785067e-05, "loss": 0.0232, "step": 73420 }, { "grad_norm": 0.764583170413971, "learning_rate": 1.809186858293742e-05, "loss": 0.0394, "step": 73430 }, { "grad_norm": 0.4072398841381073, "learning_rate": 1.8079140237449448e-05, "loss": 0.0215, "step": 73440 }, { "grad_norm": 0.34404778480529785, "learning_rate": 1.8066415382778707e-05, "loss": 0.0337, "step": 73450 }, { "grad_norm": 0.2674297094345093, "learning_rate": 1.8053694020316786e-05, "loss": 0.0231, "step": 73460 }, { "grad_norm": 0.35217025876045227, "learning_rate": 1.8040976151454837e-05, "loss": 0.0314, "step": 73470 }, { "grad_norm": 0.3562016487121582, "learning_rate": 1.8028261777583704e-05, "loss": 0.0427, "step": 73480 }, { "grad_norm": 0.42638346552848816, "learning_rate": 1.8015550900093758e-05, "loss": 0.0328, "step": 73490 }, { "grad_norm": 0.5220339298248291, "learning_rate": 1.800284352037512e-05, "loss": 0.0253, "step": 73500 }, { "grad_norm": 0.40237122774124146, "learning_rate": 1.799013963981738e-05, "loss": 0.0357, "step": 73510 }, { "grad_norm": 0.48564469814300537, "learning_rate": 1.797743925980987e-05, "loss": 0.0297, "step": 73520 }, { "grad_norm": 0.4202057421207428, "learning_rate": 1.7964742381741435e-05, "loss": 0.0298, "step": 73530 }, { "grad_norm": 0.4663769006729126, "learning_rate": 1.7952049007000627e-05, "loss": 0.0326, "step": 73540 }, { "grad_norm": 0.5523834228515625, "learning_rate": 1.7939359136975535e-05, "loss": 0.0275, "step": 73550 }, { "grad_norm": 0.30828621983528137, "learning_rate": 1.7926672773053932e-05, "loss": 0.0303, "step": 73560 }, { "grad_norm": 0.5665034651756287, "learning_rate": 1.791398991662317e-05, "loss": 0.0299, "step": 73570 }, { "grad_norm": 0.24040520191192627, "learning_rate": 1.7901310569070244e-05, "loss": 0.0315, "step": 73580 }, { "grad_norm": 0.3741043508052826, "learning_rate": 1.7888634731781724e-05, "loss": 0.026, "step": 73590 }, { "grad_norm": 1.6047091484069824, "learning_rate": 1.7875962406143844e-05, "loss": 0.0258, "step": 73600 }, { "grad_norm": 0.5212081074714661, "learning_rate": 1.7863293593542403e-05, "loss": 0.0414, "step": 73610 }, { "grad_norm": 0.29402682185173035, "learning_rate": 1.7850628295362853e-05, "loss": 0.0306, "step": 73620 }, { "grad_norm": 0.7088867425918579, "learning_rate": 1.7837966512990263e-05, "loss": 0.0345, "step": 73630 }, { "grad_norm": 0.3548084497451782, "learning_rate": 1.782530824780931e-05, "loss": 0.0263, "step": 73640 }, { "grad_norm": 0.4591069221496582, "learning_rate": 1.7812653501204263e-05, "loss": 0.0344, "step": 73650 }, { "grad_norm": 0.38832664489746094, "learning_rate": 1.780000227455903e-05, "loss": 0.0371, "step": 73660 }, { "grad_norm": 0.4981134533882141, "learning_rate": 1.778735456925716e-05, "loss": 0.0328, "step": 73670 }, { "grad_norm": 0.24581463634967804, "learning_rate": 1.7774710386681737e-05, "loss": 0.0272, "step": 73680 }, { "grad_norm": 0.27759891748428345, "learning_rate": 1.7762069728215545e-05, "loss": 0.0364, "step": 73690 }, { "grad_norm": 0.4077339172363281, "learning_rate": 1.7749432595240932e-05, "loss": 0.0285, "step": 73700 }, { "grad_norm": 0.4213072955608368, "learning_rate": 1.773679898913991e-05, "loss": 0.0398, "step": 73710 }, { "grad_norm": 0.511566698551178, "learning_rate": 1.7724168911294015e-05, "loss": 0.0318, "step": 73720 }, { "grad_norm": 0.34963399171829224, "learning_rate": 1.7711542363084504e-05, "loss": 0.0245, "step": 73730 }, { "grad_norm": 0.2472471296787262, "learning_rate": 1.7698919345892158e-05, "loss": 0.0236, "step": 73740 }, { "grad_norm": 0.3603404462337494, "learning_rate": 1.768629986109743e-05, "loss": 0.0355, "step": 73750 }, { "grad_norm": 0.4748944342136383, "learning_rate": 1.7673683910080357e-05, "loss": 0.0262, "step": 73760 }, { "grad_norm": 0.22137877345085144, "learning_rate": 1.7661071494220638e-05, "loss": 0.0264, "step": 73770 }, { "grad_norm": 0.19817984104156494, "learning_rate": 1.7648462614897498e-05, "loss": 0.0262, "step": 73780 }, { "grad_norm": 0.45561549067497253, "learning_rate": 1.7635857273489863e-05, "loss": 0.025, "step": 73790 }, { "grad_norm": 0.3477630615234375, "learning_rate": 1.7623255471376198e-05, "loss": 0.026, "step": 73800 }, { "grad_norm": 0.4515931010246277, "learning_rate": 1.7610657209934646e-05, "loss": 0.0362, "step": 73810 }, { "grad_norm": 0.2939259111881256, "learning_rate": 1.7598062490542915e-05, "loss": 0.0293, "step": 73820 }, { "grad_norm": 0.34378376603126526, "learning_rate": 1.7585471314578377e-05, "loss": 0.0383, "step": 73830 }, { "grad_norm": 0.5992156267166138, "learning_rate": 1.7572883683417945e-05, "loss": 0.0335, "step": 73840 }, { "grad_norm": 0.4807659983634949, "learning_rate": 1.7560299598438212e-05, "loss": 0.0337, "step": 73850 }, { "grad_norm": 0.28541624546051025, "learning_rate": 1.7547719061015334e-05, "loss": 0.026, "step": 73860 }, { "grad_norm": 0.550731360912323, "learning_rate": 1.75351420725251e-05, "loss": 0.024, "step": 73870 }, { "grad_norm": 0.7483330368995667, "learning_rate": 1.7522568634342923e-05, "loss": 0.0342, "step": 73880 }, { "grad_norm": 1.5985019207000732, "learning_rate": 1.7509998747843825e-05, "loss": 0.0249, "step": 73890 }, { "grad_norm": 0.42597001791000366, "learning_rate": 1.7497432414402403e-05, "loss": 0.033, "step": 73900 }, { "grad_norm": 0.42589691281318665, "learning_rate": 1.7484869635392903e-05, "loss": 0.0285, "step": 73910 }, { "grad_norm": 0.3295291066169739, "learning_rate": 1.747231041218919e-05, "loss": 0.0238, "step": 73920 }, { "grad_norm": 0.2694220542907715, "learning_rate": 1.7459754746164692e-05, "loss": 0.0278, "step": 73930 }, { "grad_norm": 0.36556124687194824, "learning_rate": 1.744720263869249e-05, "loss": 0.0305, "step": 73940 }, { "grad_norm": 0.42922839522361755, "learning_rate": 1.7434654091145263e-05, "loss": 0.0331, "step": 73950 }, { "grad_norm": 0.36755257844924927, "learning_rate": 1.742210910489532e-05, "loss": 0.0256, "step": 73960 }, { "grad_norm": 0.18379081785678864, "learning_rate": 1.7409567681314532e-05, "loss": 0.0273, "step": 73970 }, { "grad_norm": 0.33768489956855774, "learning_rate": 1.7397029821774434e-05, "loss": 0.0285, "step": 73980 }, { "grad_norm": 0.5385494828224182, "learning_rate": 1.7384495527646126e-05, "loss": 0.0331, "step": 73990 }, { "grad_norm": 0.39520952105522156, "learning_rate": 1.737196480030035e-05, "loss": 0.0225, "step": 74000 }, { "grad_norm": 0.26772043108940125, "learning_rate": 1.7359437641107447e-05, "loss": 0.0286, "step": 74010 }, { "grad_norm": 0.32464852929115295, "learning_rate": 1.7346914051437387e-05, "loss": 0.0266, "step": 74020 }, { "grad_norm": 0.34431901574134827, "learning_rate": 1.733439403265969e-05, "loss": 0.0313, "step": 74030 }, { "grad_norm": 0.3227193355560303, "learning_rate": 1.732187758614357e-05, "loss": 0.0417, "step": 74040 }, { "grad_norm": 0.4052950143814087, "learning_rate": 1.730936471325777e-05, "loss": 0.0261, "step": 74050 }, { "grad_norm": 0.24368344247341156, "learning_rate": 1.7296855415370718e-05, "loss": 0.0315, "step": 74060 }, { "grad_norm": 0.3239632248878479, "learning_rate": 1.728434969385035e-05, "loss": 0.0289, "step": 74070 }, { "grad_norm": 0.36875003576278687, "learning_rate": 1.7271847550064346e-05, "loss": 0.0283, "step": 74080 }, { "grad_norm": 0.724443793296814, "learning_rate": 1.725934898537987e-05, "loss": 0.034, "step": 74090 }, { "grad_norm": 0.37661343812942505, "learning_rate": 1.7246854001163783e-05, "loss": 0.023, "step": 74100 }, { "grad_norm": 0.5595333576202393, "learning_rate": 1.7234362598782476e-05, "loss": 0.0297, "step": 74110 }, { "grad_norm": 0.7499368786811829, "learning_rate": 1.722187477960202e-05, "loss": 0.0289, "step": 74120 }, { "grad_norm": 0.5058386325836182, "learning_rate": 1.7209390544988057e-05, "loss": 0.0283, "step": 74130 }, { "grad_norm": 0.4135226607322693, "learning_rate": 1.7196909896305856e-05, "loss": 0.0384, "step": 74140 }, { "grad_norm": 0.2570382058620453, "learning_rate": 1.718443283492026e-05, "loss": 0.0301, "step": 74150 }, { "grad_norm": 0.33335238695144653, "learning_rate": 1.7171959362195746e-05, "loss": 0.0306, "step": 74160 }, { "grad_norm": 0.27267181873321533, "learning_rate": 1.715948947949642e-05, "loss": 0.0293, "step": 74170 }, { "grad_norm": 0.34880051016807556, "learning_rate": 1.714702318818593e-05, "loss": 0.0264, "step": 74180 }, { "grad_norm": 0.4192104935646057, "learning_rate": 1.7134560489627592e-05, "loss": 0.0224, "step": 74190 }, { "grad_norm": 0.30335527658462524, "learning_rate": 1.712210138518431e-05, "loss": 0.0242, "step": 74200 }, { "grad_norm": 0.2924554944038391, "learning_rate": 1.71096458762186e-05, "loss": 0.0375, "step": 74210 }, { "grad_norm": 0.414643257856369, "learning_rate": 1.7097193964092552e-05, "loss": 0.0341, "step": 74220 }, { "grad_norm": 0.721264660358429, "learning_rate": 1.708474565016792e-05, "loss": 0.0268, "step": 74230 }, { "grad_norm": 0.335592120885849, "learning_rate": 1.7072300935806e-05, "loss": 0.035, "step": 74240 }, { "grad_norm": 0.5648607015609741, "learning_rate": 1.705985982236776e-05, "loss": 0.027, "step": 74250 }, { "grad_norm": 0.4258667528629303, "learning_rate": 1.7047422311213697e-05, "loss": 0.0306, "step": 74260 }, { "grad_norm": 0.43023043870925903, "learning_rate": 1.703498840370401e-05, "loss": 0.0409, "step": 74270 }, { "grad_norm": 0.341823011636734, "learning_rate": 1.702255810119842e-05, "loss": 0.029, "step": 74280 }, { "grad_norm": 0.32178178429603577, "learning_rate": 1.7010131405056305e-05, "loss": 0.0225, "step": 74290 }, { "grad_norm": 0.3589054346084595, "learning_rate": 1.6997708316636613e-05, "loss": 0.032, "step": 74300 }, { "grad_norm": 0.3023988902568817, "learning_rate": 1.6985288837297932e-05, "loss": 0.0306, "step": 74310 }, { "grad_norm": 0.3546866178512573, "learning_rate": 1.6972872968398394e-05, "loss": 0.0283, "step": 74320 }, { "grad_norm": 0.29060232639312744, "learning_rate": 1.6960460711295845e-05, "loss": 0.0279, "step": 74330 }, { "grad_norm": 0.28547796607017517, "learning_rate": 1.6948052067347623e-05, "loss": 0.0261, "step": 74340 }, { "grad_norm": 0.5156837701797485, "learning_rate": 1.693564703791075e-05, "loss": 0.034, "step": 74350 }, { "grad_norm": 0.2695721387863159, "learning_rate": 1.6923245624341793e-05, "loss": 0.0286, "step": 74360 }, { "grad_norm": 0.3041239082813263, "learning_rate": 1.6910847827996962e-05, "loss": 0.0216, "step": 74370 }, { "grad_norm": 0.3946686089038849, "learning_rate": 1.689845365023206e-05, "loss": 0.0296, "step": 74380 }, { "grad_norm": 0.1781851053237915, "learning_rate": 1.6886063092402505e-05, "loss": 0.029, "step": 74390 }, { "grad_norm": 0.594343364238739, "learning_rate": 1.6873676155863317e-05, "loss": 0.0408, "step": 74400 }, { "grad_norm": 0.2718176245689392, "learning_rate": 1.6861292841969085e-05, "loss": 0.0327, "step": 74410 }, { "grad_norm": 0.5082507729530334, "learning_rate": 1.684891315207406e-05, "loss": 0.0324, "step": 74420 }, { "grad_norm": 0.3263865113258362, "learning_rate": 1.6836537087532034e-05, "loss": 0.0289, "step": 74430 }, { "grad_norm": 0.41579508781433105, "learning_rate": 1.6824164649696465e-05, "loss": 0.0422, "step": 74440 }, { "grad_norm": 0.6528142690658569, "learning_rate": 1.681179583992034e-05, "loss": 0.0253, "step": 74450 }, { "grad_norm": 0.41987094283103943, "learning_rate": 1.679943065955636e-05, "loss": 0.0308, "step": 74460 }, { "grad_norm": 1.1361783742904663, "learning_rate": 1.678706910995671e-05, "loss": 0.0327, "step": 74470 }, { "grad_norm": 0.24998325109481812, "learning_rate": 1.6774711192473262e-05, "loss": 0.0315, "step": 74480 }, { "grad_norm": 0.33582019805908203, "learning_rate": 1.676235690845743e-05, "loss": 0.0208, "step": 74490 }, { "grad_norm": 0.6202698945999146, "learning_rate": 1.6750006259260285e-05, "loss": 0.0358, "step": 74500 }, { "grad_norm": 0.34858548641204834, "learning_rate": 1.673765924623244e-05, "loss": 0.0328, "step": 74510 }, { "grad_norm": 0.48178017139434814, "learning_rate": 1.67253158707242e-05, "loss": 0.0246, "step": 74520 }, { "grad_norm": 0.31217527389526367, "learning_rate": 1.671297613408537e-05, "loss": 0.0255, "step": 74530 }, { "grad_norm": 0.48501908779144287, "learning_rate": 1.6700640037665444e-05, "loss": 0.0239, "step": 74540 }, { "grad_norm": 0.2431401014328003, "learning_rate": 1.668830758281344e-05, "loss": 0.0279, "step": 74550 }, { "grad_norm": 0.31560009717941284, "learning_rate": 1.6675978770878052e-05, "loss": 0.0265, "step": 74560 }, { "grad_norm": 0.27182936668395996, "learning_rate": 1.6663653603207506e-05, "loss": 0.0243, "step": 74570 }, { "grad_norm": 0.4082339107990265, "learning_rate": 1.6651332081149685e-05, "loss": 0.0306, "step": 74580 }, { "grad_norm": 0.5647366642951965, "learning_rate": 1.6639014206052046e-05, "loss": 0.0265, "step": 74590 }, { "grad_norm": 0.30270737409591675, "learning_rate": 1.6626699979261657e-05, "loss": 0.0267, "step": 74600 }, { "grad_norm": 0.3400479555130005, "learning_rate": 1.6614389402125197e-05, "loss": 0.0372, "step": 74610 }, { "grad_norm": 0.34136179089546204, "learning_rate": 1.66020824759889e-05, "loss": 0.0281, "step": 74620 }, { "grad_norm": 0.36219513416290283, "learning_rate": 1.6589779202198675e-05, "loss": 0.0308, "step": 74630 }, { "grad_norm": 0.5581791996955872, "learning_rate": 1.6577479582099924e-05, "loss": 0.0281, "step": 74640 }, { "grad_norm": 0.41913142800331116, "learning_rate": 1.6565183617037788e-05, "loss": 0.0379, "step": 74650 }, { "grad_norm": 0.4640980362892151, "learning_rate": 1.6552891308356894e-05, "loss": 0.0327, "step": 74660 }, { "grad_norm": 0.4239279329776764, "learning_rate": 1.6540602657401526e-05, "loss": 0.0402, "step": 74670 }, { "grad_norm": 0.2568897008895874, "learning_rate": 1.6528317665515537e-05, "loss": 0.0358, "step": 74680 }, { "grad_norm": 0.3518223464488983, "learning_rate": 1.6516036334042416e-05, "loss": 0.0373, "step": 74690 }, { "grad_norm": 0.26637470722198486, "learning_rate": 1.650375866432519e-05, "loss": 0.037, "step": 74700 }, { "grad_norm": 0.8810614943504333, "learning_rate": 1.649148465770659e-05, "loss": 0.0294, "step": 74710 }, { "grad_norm": 0.36193332076072693, "learning_rate": 1.647921431552883e-05, "loss": 0.0313, "step": 74720 }, { "grad_norm": 0.3767733871936798, "learning_rate": 1.646694763913381e-05, "loss": 0.0369, "step": 74730 }, { "grad_norm": 0.3131988048553467, "learning_rate": 1.6454684629862966e-05, "loss": 0.0307, "step": 74740 }, { "grad_norm": 0.19317956268787384, "learning_rate": 1.6442425289057388e-05, "loss": 0.0236, "step": 74750 }, { "grad_norm": 0.3585883378982544, "learning_rate": 1.6430169618057718e-05, "loss": 0.024, "step": 74760 }, { "grad_norm": 0.2789248824119568, "learning_rate": 1.641791761820422e-05, "loss": 0.0298, "step": 74770 }, { "grad_norm": 0.41143667697906494, "learning_rate": 1.640566929083676e-05, "loss": 0.0336, "step": 74780 }, { "grad_norm": 0.4739137887954712, "learning_rate": 1.6393424637294823e-05, "loss": 0.0348, "step": 74790 }, { "grad_norm": 0.276248961687088, "learning_rate": 1.638118365891742e-05, "loss": 0.0261, "step": 74800 }, { "grad_norm": 0.43505921959877014, "learning_rate": 1.636894635704323e-05, "loss": 0.0339, "step": 74810 }, { "grad_norm": 0.3368985950946808, "learning_rate": 1.6356712733010517e-05, "loss": 0.0231, "step": 74820 }, { "grad_norm": 0.8535666465759277, "learning_rate": 1.6344482788157096e-05, "loss": 0.028, "step": 74830 }, { "grad_norm": 0.29553622007369995, "learning_rate": 1.6332256523820444e-05, "loss": 0.0237, "step": 74840 }, { "grad_norm": 0.6556148529052734, "learning_rate": 1.6320033941337597e-05, "loss": 0.0379, "step": 74850 }, { "grad_norm": 0.3227466642856598, "learning_rate": 1.6307815042045216e-05, "loss": 0.024, "step": 74860 }, { "grad_norm": 0.4076339602470398, "learning_rate": 1.6295599827279513e-05, "loss": 0.0272, "step": 74870 }, { "grad_norm": 0.5446411967277527, "learning_rate": 1.6283388298376356e-05, "loss": 0.0236, "step": 74880 }, { "grad_norm": 0.7381402850151062, "learning_rate": 1.6271180456671125e-05, "loss": 0.0325, "step": 74890 }, { "grad_norm": 0.3663153648376465, "learning_rate": 1.625897630349893e-05, "loss": 0.0273, "step": 74900 }, { "grad_norm": 0.36987021565437317, "learning_rate": 1.6246775840194335e-05, "loss": 0.0253, "step": 74910 }, { "grad_norm": 0.18749812245368958, "learning_rate": 1.6234579068091604e-05, "loss": 0.0216, "step": 74920 }, { "grad_norm": 0.3164849579334259, "learning_rate": 1.6222385988524524e-05, "loss": 0.0324, "step": 74930 }, { "grad_norm": 0.21536549925804138, "learning_rate": 1.6210196602826544e-05, "loss": 0.0304, "step": 74940 }, { "grad_norm": 0.3862525224685669, "learning_rate": 1.6198010912330646e-05, "loss": 0.0257, "step": 74950 }, { "grad_norm": 0.3928215503692627, "learning_rate": 1.6185828918369456e-05, "loss": 0.035, "step": 74960 }, { "grad_norm": 0.26767879724502563, "learning_rate": 1.617365062227517e-05, "loss": 0.0241, "step": 74970 }, { "grad_norm": 0.2702634036540985, "learning_rate": 1.616147602537962e-05, "loss": 0.0459, "step": 74980 }, { "grad_norm": 0.3359052240848541, "learning_rate": 1.614930512901415e-05, "loss": 0.0292, "step": 74990 }, { "grad_norm": 0.6512497067451477, "learning_rate": 1.6137137934509804e-05, "loss": 0.0263, "step": 75000 }, { "grad_norm": 1.0056363344192505, "learning_rate": 1.6124974443197123e-05, "loss": 0.0268, "step": 75010 }, { "grad_norm": 0.3940291404724121, "learning_rate": 1.6112814656406306e-05, "loss": 0.0328, "step": 75020 }, { "grad_norm": 0.4275767505168915, "learning_rate": 1.6100658575467125e-05, "loss": 0.0282, "step": 75030 }, { "grad_norm": 0.3518041968345642, "learning_rate": 1.6088506201708974e-05, "loss": 0.0251, "step": 75040 }, { "grad_norm": 0.5196321606636047, "learning_rate": 1.6076357536460785e-05, "loss": 0.028, "step": 75050 }, { "grad_norm": 0.2837931513786316, "learning_rate": 1.606421258105113e-05, "loss": 0.03, "step": 75060 }, { "grad_norm": 0.5374297499656677, "learning_rate": 1.6052071336808183e-05, "loss": 0.0338, "step": 75070 }, { "grad_norm": 0.404930979013443, "learning_rate": 1.6039933805059658e-05, "loss": 0.0325, "step": 75080 }, { "grad_norm": 0.346658855676651, "learning_rate": 1.6027799987132915e-05, "loss": 0.0345, "step": 75090 }, { "grad_norm": 0.41254159808158875, "learning_rate": 1.6015669884354888e-05, "loss": 0.0384, "step": 75100 }, { "grad_norm": 0.4274809658527374, "learning_rate": 1.6003543498052125e-05, "loss": 0.0343, "step": 75110 }, { "grad_norm": 0.3838164210319519, "learning_rate": 1.5991420829550715e-05, "loss": 0.032, "step": 75120 }, { "grad_norm": 0.2611476182937622, "learning_rate": 1.5979301880176405e-05, "loss": 0.0402, "step": 75130 }, { "grad_norm": 0.37433087825775146, "learning_rate": 1.5967186651254485e-05, "loss": 0.0298, "step": 75140 }, { "grad_norm": 0.7465399503707886, "learning_rate": 1.5955075144109853e-05, "loss": 0.0293, "step": 75150 }, { "grad_norm": 0.6568738222122192, "learning_rate": 1.594296736006703e-05, "loss": 0.028, "step": 75160 }, { "grad_norm": 0.5100151300430298, "learning_rate": 1.59308633004501e-05, "loss": 0.0375, "step": 75170 }, { "grad_norm": 0.3509614169597626, "learning_rate": 1.5918762966582724e-05, "loss": 0.0328, "step": 75180 }, { "grad_norm": 0.6928947567939758, "learning_rate": 1.5906666359788203e-05, "loss": 0.0287, "step": 75190 }, { "grad_norm": 0.2455766797065735, "learning_rate": 1.589457348138938e-05, "loss": 0.0277, "step": 75200 }, { "grad_norm": 0.7478924989700317, "learning_rate": 1.588248433270872e-05, "loss": 0.0296, "step": 75210 }, { "grad_norm": 0.46223440766334534, "learning_rate": 1.587039891506828e-05, "loss": 0.0278, "step": 75220 }, { "grad_norm": 0.316121369600296, "learning_rate": 1.5858317229789715e-05, "loss": 0.0263, "step": 75230 }, { "grad_norm": 0.3890731632709503, "learning_rate": 1.584623927819423e-05, "loss": 0.0305, "step": 75240 }, { "grad_norm": 0.5670844912528992, "learning_rate": 1.583416506160269e-05, "loss": 0.0395, "step": 75250 }, { "grad_norm": 0.3681960701942444, "learning_rate": 1.5822094581335472e-05, "loss": 0.0272, "step": 75260 }, { "grad_norm": 0.2317257523536682, "learning_rate": 1.5810027838712604e-05, "loss": 0.0277, "step": 75270 }, { "grad_norm": 0.3696562945842743, "learning_rate": 1.5797964835053687e-05, "loss": 0.0354, "step": 75280 }, { "grad_norm": 0.2625918984413147, "learning_rate": 1.5785905571677933e-05, "loss": 0.0277, "step": 75290 }, { "grad_norm": 0.4135451316833496, "learning_rate": 1.5773850049904086e-05, "loss": 0.0282, "step": 75300 }, { "grad_norm": 0.28020966053009033, "learning_rate": 1.5761798271050543e-05, "loss": 0.0246, "step": 75310 }, { "grad_norm": 0.3591625690460205, "learning_rate": 1.5749750236435277e-05, "loss": 0.0297, "step": 75320 }, { "grad_norm": 0.5008553862571716, "learning_rate": 1.5737705947375815e-05, "loss": 0.0279, "step": 75330 }, { "grad_norm": 0.5505737066268921, "learning_rate": 1.5725665405189322e-05, "loss": 0.03, "step": 75340 }, { "grad_norm": 0.2615404427051544, "learning_rate": 1.571362861119253e-05, "loss": 0.0276, "step": 75350 }, { "grad_norm": 0.5635613799095154, "learning_rate": 1.570159556670177e-05, "loss": 0.0225, "step": 75360 }, { "grad_norm": 0.1935034841299057, "learning_rate": 1.5689566273032946e-05, "loss": 0.0243, "step": 75370 }, { "grad_norm": 0.3502713739871979, "learning_rate": 1.5677540731501583e-05, "loss": 0.0271, "step": 75380 }, { "grad_norm": 0.36098402738571167, "learning_rate": 1.5665518943422746e-05, "loss": 0.0296, "step": 75390 }, { "grad_norm": 0.41242730617523193, "learning_rate": 1.5653500910111134e-05, "loss": 0.0334, "step": 75400 }, { "grad_norm": 0.48924925923347473, "learning_rate": 1.5641486632881023e-05, "loss": 0.0287, "step": 75410 }, { "grad_norm": 0.6187868118286133, "learning_rate": 1.56294761130463e-05, "loss": 0.0324, "step": 75420 }, { "grad_norm": 0.36473554372787476, "learning_rate": 1.5617469351920373e-05, "loss": 0.0312, "step": 75430 }, { "grad_norm": 0.4811413288116455, "learning_rate": 1.5605466350816315e-05, "loss": 0.0346, "step": 75440 }, { "grad_norm": 0.47765180468559265, "learning_rate": 1.559346711104674e-05, "loss": 0.0291, "step": 75450 }, { "grad_norm": 0.3567972481250763, "learning_rate": 1.5581471633923883e-05, "loss": 0.0352, "step": 75460 }, { "grad_norm": 0.4677409529685974, "learning_rate": 1.5569479920759505e-05, "loss": 0.0309, "step": 75470 }, { "grad_norm": 0.39266765117645264, "learning_rate": 1.5557491972865072e-05, "loss": 0.0295, "step": 75480 }, { "grad_norm": 0.2668016254901886, "learning_rate": 1.554550779155152e-05, "loss": 0.0243, "step": 75490 }, { "grad_norm": 0.3600478172302246, "learning_rate": 1.5533527378129453e-05, "loss": 0.0235, "step": 75500 }, { "grad_norm": 0.40277916193008423, "learning_rate": 1.552155073390899e-05, "loss": 0.0358, "step": 75510 }, { "grad_norm": 1.6709710359573364, "learning_rate": 1.5509577860199903e-05, "loss": 0.0395, "step": 75520 }, { "grad_norm": 0.2574656307697296, "learning_rate": 1.5497608758311525e-05, "loss": 0.0268, "step": 75530 }, { "grad_norm": 0.4248432219028473, "learning_rate": 1.5485643429552797e-05, "loss": 0.0345, "step": 75540 }, { "grad_norm": 0.278115451335907, "learning_rate": 1.5473681875232197e-05, "loss": 0.0277, "step": 75550 }, { "grad_norm": 0.4200096130371094, "learning_rate": 1.5461724096657837e-05, "loss": 0.0273, "step": 75560 }, { "grad_norm": 0.3923583924770355, "learning_rate": 1.5449770095137416e-05, "loss": 0.024, "step": 75570 }, { "grad_norm": 0.2907438576221466, "learning_rate": 1.5437819871978178e-05, "loss": 0.0255, "step": 75580 }, { "grad_norm": 0.5682215094566345, "learning_rate": 1.5425873428486992e-05, "loss": 0.0298, "step": 75590 }, { "grad_norm": 0.35627490282058716, "learning_rate": 1.5413930765970312e-05, "loss": 0.0299, "step": 75600 }, { "grad_norm": 0.41051891446113586, "learning_rate": 1.5401991885734175e-05, "loss": 0.0274, "step": 75610 }, { "grad_norm": 0.5058444142341614, "learning_rate": 1.5390056789084174e-05, "loss": 0.0251, "step": 75620 }, { "grad_norm": 0.6604346036911011, "learning_rate": 1.5378125477325545e-05, "loss": 0.026, "step": 75630 }, { "grad_norm": 0.3060239255428314, "learning_rate": 1.5366197951763046e-05, "loss": 0.0262, "step": 75640 }, { "grad_norm": 0.6627296805381775, "learning_rate": 1.5354274213701075e-05, "loss": 0.0349, "step": 75650 }, { "grad_norm": 0.30355677008628845, "learning_rate": 1.5342354264443566e-05, "loss": 0.0287, "step": 75660 }, { "grad_norm": 0.9076786041259766, "learning_rate": 1.533043810529411e-05, "loss": 0.0302, "step": 75670 }, { "grad_norm": 0.3927645981311798, "learning_rate": 1.5318525737555807e-05, "loss": 0.0315, "step": 75680 }, { "grad_norm": 0.31148090958595276, "learning_rate": 1.53066171625314e-05, "loss": 0.0288, "step": 75690 }, { "grad_norm": 0.47804921865463257, "learning_rate": 1.529471238152317e-05, "loss": 0.0307, "step": 75700 }, { "grad_norm": 0.41944876313209534, "learning_rate": 1.5282811395833023e-05, "loss": 0.0308, "step": 75710 }, { "grad_norm": 0.4417389929294586, "learning_rate": 1.5270914206762397e-05, "loss": 0.0288, "step": 75720 }, { "grad_norm": 0.5116798877716064, "learning_rate": 1.525902081561241e-05, "loss": 0.0283, "step": 75730 }, { "grad_norm": 0.3515859842300415, "learning_rate": 1.5247131223683664e-05, "loss": 0.0338, "step": 75740 }, { "grad_norm": 0.3605997860431671, "learning_rate": 1.5235245432276408e-05, "loss": 0.0299, "step": 75750 }, { "grad_norm": 0.25732728838920593, "learning_rate": 1.5223363442690424e-05, "loss": 0.029, "step": 75760 }, { "grad_norm": 0.26533082127571106, "learning_rate": 1.5211485256225133e-05, "loss": 0.0296, "step": 75770 }, { "grad_norm": 0.40117359161376953, "learning_rate": 1.519961087417951e-05, "loss": 0.0242, "step": 75780 }, { "grad_norm": 0.467086523771286, "learning_rate": 1.5187740297852137e-05, "loss": 0.0292, "step": 75790 }, { "grad_norm": 0.30003970861434937, "learning_rate": 1.517587352854113e-05, "loss": 0.0209, "step": 75800 }, { "grad_norm": 0.757287323474884, "learning_rate": 1.516401056754423e-05, "loss": 0.0286, "step": 75810 }, { "grad_norm": 0.41797390580177307, "learning_rate": 1.5152151416158783e-05, "loss": 0.0411, "step": 75820 }, { "grad_norm": 0.3382749855518341, "learning_rate": 1.5140296075681642e-05, "loss": 0.0343, "step": 75830 }, { "grad_norm": 0.34040430188179016, "learning_rate": 1.5128444547409331e-05, "loss": 0.0297, "step": 75840 }, { "grad_norm": 0.5055113434791565, "learning_rate": 1.5116596832637864e-05, "loss": 0.0257, "step": 75850 }, { "grad_norm": 0.39872831106185913, "learning_rate": 1.5104752932662952e-05, "loss": 0.0354, "step": 75860 }, { "grad_norm": 0.5097109079360962, "learning_rate": 1.5092912848779778e-05, "loss": 0.032, "step": 75870 }, { "grad_norm": 0.37901535630226135, "learning_rate": 1.5081076582283194e-05, "loss": 0.0275, "step": 75880 }, { "grad_norm": 0.30848991870880127, "learning_rate": 1.5069244134467553e-05, "loss": 0.0292, "step": 75890 }, { "grad_norm": 0.8770999908447266, "learning_rate": 1.5057415506626876e-05, "loss": 0.0279, "step": 75900 }, { "grad_norm": 0.48006123304367065, "learning_rate": 1.5045590700054668e-05, "loss": 0.0307, "step": 75910 }, { "grad_norm": 0.32015103101730347, "learning_rate": 1.5033769716044144e-05, "loss": 0.0308, "step": 75920 }, { "grad_norm": 0.4689593017101288, "learning_rate": 1.5021952555887975e-05, "loss": 0.0221, "step": 75930 }, { "grad_norm": 0.3047982454299927, "learning_rate": 1.50101392208785e-05, "loss": 0.0287, "step": 75940 }, { "grad_norm": 0.4520728886127472, "learning_rate": 1.4998329712307579e-05, "loss": 0.0238, "step": 75950 }, { "grad_norm": 0.4269315302371979, "learning_rate": 1.4986524031466708e-05, "loss": 0.0315, "step": 75960 }, { "grad_norm": 0.3419525921344757, "learning_rate": 1.4974722179646916e-05, "loss": 0.0194, "step": 75970 }, { "grad_norm": 0.4527052640914917, "learning_rate": 1.4962924158138847e-05, "loss": 0.0341, "step": 75980 }, { "grad_norm": 0.25149449706077576, "learning_rate": 1.4951129968232714e-05, "loss": 0.0243, "step": 75990 }, { "grad_norm": 0.47801467776298523, "learning_rate": 1.4939339611218329e-05, "loss": 0.0304, "step": 76000 }, { "grad_norm": 0.3280363976955414, "learning_rate": 1.4927553088385032e-05, "loss": 0.0386, "step": 76010 }, { "grad_norm": 0.37870168685913086, "learning_rate": 1.4915770401021801e-05, "loss": 0.0274, "step": 76020 }, { "grad_norm": 0.357511967420578, "learning_rate": 1.4903991550417185e-05, "loss": 0.0329, "step": 76030 }, { "grad_norm": 0.5153192281723022, "learning_rate": 1.4892216537859277e-05, "loss": 0.0326, "step": 76040 }, { "grad_norm": 0.4937123656272888, "learning_rate": 1.4880445364635775e-05, "loss": 0.0323, "step": 76050 }, { "grad_norm": 0.48461446166038513, "learning_rate": 1.4868678032033973e-05, "loss": 0.0313, "step": 76060 }, { "grad_norm": 0.3026285767555237, "learning_rate": 1.4856914541340739e-05, "loss": 0.0302, "step": 76070 }, { "grad_norm": 0.746262788772583, "learning_rate": 1.4845154893842472e-05, "loss": 0.0285, "step": 76080 }, { "grad_norm": 0.5247007012367249, "learning_rate": 1.4833399090825234e-05, "loss": 0.0363, "step": 76090 }, { "grad_norm": 0.5836156606674194, "learning_rate": 1.4821647133574562e-05, "loss": 0.033, "step": 76100 }, { "grad_norm": 0.33099275827407837, "learning_rate": 1.4809899023375706e-05, "loss": 0.0268, "step": 76110 }, { "grad_norm": 0.4347139596939087, "learning_rate": 1.4798154761513366e-05, "loss": 0.0302, "step": 76120 }, { "grad_norm": 0.39860066771507263, "learning_rate": 1.4786414349271916e-05, "loss": 0.0374, "step": 76130 }, { "grad_norm": 0.5238243937492371, "learning_rate": 1.4774677787935226e-05, "loss": 0.0379, "step": 76140 }, { "grad_norm": 0.3496837317943573, "learning_rate": 1.4762945078786838e-05, "loss": 0.0236, "step": 76150 }, { "grad_norm": 0.5235558152198792, "learning_rate": 1.4751216223109777e-05, "loss": 0.0301, "step": 76160 }, { "grad_norm": 0.6263989806175232, "learning_rate": 1.4739491222186714e-05, "loss": 0.0335, "step": 76170 }, { "grad_norm": 0.5669523477554321, "learning_rate": 1.4727770077299879e-05, "loss": 0.0278, "step": 76180 }, { "grad_norm": 0.2685241103172302, "learning_rate": 1.4716052789731094e-05, "loss": 0.0205, "step": 76190 }, { "grad_norm": 0.7156597375869751, "learning_rate": 1.470433936076171e-05, "loss": 0.0334, "step": 76200 }, { "grad_norm": 0.22977237403392792, "learning_rate": 1.4692629791672719e-05, "loss": 0.0279, "step": 76210 }, { "grad_norm": 0.963640570640564, "learning_rate": 1.468092408374464e-05, "loss": 0.0233, "step": 76220 }, { "grad_norm": 0.6097389459609985, "learning_rate": 1.4669222238257602e-05, "loss": 0.0343, "step": 76230 }, { "grad_norm": 0.3377591669559479, "learning_rate": 1.4657524256491301e-05, "loss": 0.0242, "step": 76240 }, { "grad_norm": 0.4242747724056244, "learning_rate": 1.464583013972502e-05, "loss": 0.0302, "step": 76250 }, { "grad_norm": 0.36585158109664917, "learning_rate": 1.4634139889237591e-05, "loss": 0.0328, "step": 76260 }, { "grad_norm": 0.8545400500297546, "learning_rate": 1.4622453506307448e-05, "loss": 0.0315, "step": 76270 }, { "grad_norm": 0.4896695911884308, "learning_rate": 1.461077099221262e-05, "loss": 0.0338, "step": 76280 }, { "grad_norm": 0.2020551711320877, "learning_rate": 1.4599092348230652e-05, "loss": 0.0314, "step": 76290 }, { "grad_norm": 0.28191718459129333, "learning_rate": 1.4587417575638724e-05, "loss": 0.0249, "step": 76300 }, { "grad_norm": 0.8422520160675049, "learning_rate": 1.4575746675713563e-05, "loss": 0.0237, "step": 76310 }, { "grad_norm": 0.3458305895328522, "learning_rate": 1.4564079649731504e-05, "loss": 0.0335, "step": 76320 }, { "grad_norm": 0.5751326680183411, "learning_rate": 1.4552416498968402e-05, "loss": 0.0301, "step": 76330 }, { "grad_norm": 0.4013494849205017, "learning_rate": 1.4540757224699747e-05, "loss": 0.0279, "step": 76340 }, { "grad_norm": 0.5243530869483948, "learning_rate": 1.452910182820056e-05, "loss": 0.0312, "step": 76350 }, { "grad_norm": 0.4287698268890381, "learning_rate": 1.4517450310745468e-05, "loss": 0.0287, "step": 76360 }, { "grad_norm": 0.5731102824211121, "learning_rate": 1.450580267360866e-05, "loss": 0.03, "step": 76370 }, { "grad_norm": 0.306612491607666, "learning_rate": 1.4494158918063922e-05, "loss": 0.0254, "step": 76380 }, { "grad_norm": 0.4481618404388428, "learning_rate": 1.4482519045384568e-05, "loss": 0.0252, "step": 76390 }, { "grad_norm": 0.47484520077705383, "learning_rate": 1.4470883056843548e-05, "loss": 0.0257, "step": 76400 }, { "grad_norm": 0.382159948348999, "learning_rate": 1.4459250953713322e-05, "loss": 0.0316, "step": 76410 }, { "grad_norm": 0.39360031485557556, "learning_rate": 1.444762273726598e-05, "loss": 0.0301, "step": 76420 }, { "grad_norm": 0.6749203205108643, "learning_rate": 1.443599840877316e-05, "loss": 0.0283, "step": 76430 }, { "grad_norm": 0.34529903531074524, "learning_rate": 1.4424377969506103e-05, "loss": 0.0277, "step": 76440 }, { "grad_norm": 0.4291284680366516, "learning_rate": 1.4412761420735565e-05, "loss": 0.0255, "step": 76450 }, { "grad_norm": 0.31499290466308594, "learning_rate": 1.4401148763731953e-05, "loss": 0.0311, "step": 76460 }, { "grad_norm": 0.25665634870529175, "learning_rate": 1.4389539999765173e-05, "loss": 0.0241, "step": 76470 }, { "grad_norm": 0.5659424066543579, "learning_rate": 1.4377935130104758e-05, "loss": 0.0252, "step": 76480 }, { "grad_norm": 0.3292278051376343, "learning_rate": 1.4366334156019801e-05, "loss": 0.0336, "step": 76490 }, { "grad_norm": 0.3018965721130371, "learning_rate": 1.435473707877898e-05, "loss": 0.0265, "step": 76500 }, { "grad_norm": 0.41335776448249817, "learning_rate": 1.4343143899650508e-05, "loss": 0.0287, "step": 76510 }, { "grad_norm": 0.46753519773483276, "learning_rate": 1.4331554619902204e-05, "loss": 0.0304, "step": 76520 }, { "grad_norm": 0.38689306378364563, "learning_rate": 1.4319969240801484e-05, "loss": 0.0233, "step": 76530 }, { "grad_norm": 0.41053202748298645, "learning_rate": 1.430838776361526e-05, "loss": 0.0307, "step": 76540 }, { "grad_norm": 0.28148728609085083, "learning_rate": 1.4296810189610093e-05, "loss": 0.0291, "step": 76550 }, { "grad_norm": 0.2733190953731537, "learning_rate": 1.4285236520052087e-05, "loss": 0.0252, "step": 76560 }, { "grad_norm": 0.3915256857872009, "learning_rate": 1.4273666756206933e-05, "loss": 0.0236, "step": 76570 }, { "grad_norm": 0.2762221097946167, "learning_rate": 1.4262100899339848e-05, "loss": 0.0287, "step": 76580 }, { "grad_norm": 0.4234415590763092, "learning_rate": 1.4250538950715696e-05, "loss": 0.0261, "step": 76590 }, { "grad_norm": 0.4071279466152191, "learning_rate": 1.4238980911598842e-05, "loss": 0.0357, "step": 76600 }, { "grad_norm": 0.48478978872299194, "learning_rate": 1.4227426783253267e-05, "loss": 0.0274, "step": 76610 }, { "grad_norm": 0.5268417000770569, "learning_rate": 1.4215876566942521e-05, "loss": 0.0271, "step": 76620 }, { "grad_norm": 0.2706408202648163, "learning_rate": 1.4204330263929721e-05, "loss": 0.021, "step": 76630 }, { "grad_norm": 0.32023394107818604, "learning_rate": 1.4192787875477537e-05, "loss": 0.0339, "step": 76640 }, { "grad_norm": 0.21864397823810577, "learning_rate": 1.4181249402848246e-05, "loss": 0.0264, "step": 76650 }, { "grad_norm": 0.3439190983772278, "learning_rate": 1.4169714847303656e-05, "loss": 0.0288, "step": 76660 }, { "grad_norm": 0.7517613768577576, "learning_rate": 1.4158184210105197e-05, "loss": 0.0314, "step": 76670 }, { "grad_norm": 0.25727880001068115, "learning_rate": 1.4146657492513787e-05, "loss": 0.0293, "step": 76680 }, { "grad_norm": 0.2451629936695099, "learning_rate": 1.4135134695790047e-05, "loss": 0.0216, "step": 76690 }, { "grad_norm": 0.498878538608551, "learning_rate": 1.4123615821194037e-05, "loss": 0.0316, "step": 76700 }, { "grad_norm": 0.4945101737976074, "learning_rate": 1.4112100869985473e-05, "loss": 0.0259, "step": 76710 }, { "grad_norm": 0.24882982671260834, "learning_rate": 1.4100589843423589e-05, "loss": 0.0273, "step": 76720 }, { "grad_norm": 0.35656610131263733, "learning_rate": 1.408908274276723e-05, "loss": 0.0241, "step": 76730 }, { "grad_norm": 0.24893248081207275, "learning_rate": 1.4077579569274786e-05, "loss": 0.0324, "step": 76740 }, { "grad_norm": 0.36403369903564453, "learning_rate": 1.4066080324204245e-05, "loss": 0.0311, "step": 76750 }, { "grad_norm": 0.30959996581077576, "learning_rate": 1.405458500881312e-05, "loss": 0.0281, "step": 76760 }, { "grad_norm": 0.26293814182281494, "learning_rate": 1.4043093624358538e-05, "loss": 0.0267, "step": 76770 }, { "grad_norm": 0.7166528701782227, "learning_rate": 1.4031606172097189e-05, "loss": 0.0427, "step": 76780 }, { "grad_norm": 0.3552318811416626, "learning_rate": 1.40201226532853e-05, "loss": 0.0382, "step": 76790 }, { "grad_norm": 0.3182511031627655, "learning_rate": 1.4008643069178701e-05, "loss": 0.0255, "step": 76800 }, { "grad_norm": 0.35256293416023254, "learning_rate": 1.399716742103278e-05, "loss": 0.0314, "step": 76810 }, { "grad_norm": 0.38967081904411316, "learning_rate": 1.398569571010252e-05, "loss": 0.0281, "step": 76820 }, { "grad_norm": 0.44765904545783997, "learning_rate": 1.3974227937642414e-05, "loss": 0.0413, "step": 76830 }, { "grad_norm": 0.33835500478744507, "learning_rate": 1.3962764104906595e-05, "loss": 0.0258, "step": 76840 }, { "grad_norm": 0.2690318822860718, "learning_rate": 1.3951304213148696e-05, "loss": 0.0329, "step": 76850 }, { "grad_norm": 0.31651803851127625, "learning_rate": 1.3939848263621979e-05, "loss": 0.0276, "step": 76860 }, { "grad_norm": 0.6606650948524475, "learning_rate": 1.3928396257579219e-05, "loss": 0.0337, "step": 76870 }, { "grad_norm": 0.27191296219825745, "learning_rate": 1.3916948196272833e-05, "loss": 0.033, "step": 76880 }, { "grad_norm": 0.3265579640865326, "learning_rate": 1.3905504080954734e-05, "loss": 0.0424, "step": 76890 }, { "grad_norm": 0.3691324293613434, "learning_rate": 1.3894063912876454e-05, "loss": 0.0334, "step": 76900 }, { "grad_norm": 0.3501952886581421, "learning_rate": 1.3882627693289046e-05, "loss": 0.0268, "step": 76910 }, { "grad_norm": 0.4946606457233429, "learning_rate": 1.3871195423443183e-05, "loss": 0.0278, "step": 76920 }, { "grad_norm": 0.4077419340610504, "learning_rate": 1.3859767104589044e-05, "loss": 0.0244, "step": 76930 }, { "grad_norm": 0.4957638382911682, "learning_rate": 1.3848342737976467e-05, "loss": 0.0293, "step": 76940 }, { "grad_norm": 0.2572936415672302, "learning_rate": 1.3836922324854757e-05, "loss": 0.0315, "step": 76950 }, { "grad_norm": 0.5292333960533142, "learning_rate": 1.382550586647287e-05, "loss": 0.0367, "step": 76960 }, { "grad_norm": 0.26996248960494995, "learning_rate": 1.3814093364079257e-05, "loss": 0.0281, "step": 76970 }, { "grad_norm": 0.40886810421943665, "learning_rate": 1.3802684818921991e-05, "loss": 0.0333, "step": 76980 }, { "grad_norm": 0.4071793258190155, "learning_rate": 1.3791280232248687e-05, "loss": 0.0261, "step": 76990 }, { "grad_norm": 0.30350932478904724, "learning_rate": 1.3779879605306555e-05, "loss": 0.0276, "step": 77000 }, { "grad_norm": 0.4372849464416504, "learning_rate": 1.376848293934232e-05, "loss": 0.0401, "step": 77010 }, { "grad_norm": 0.4749375581741333, "learning_rate": 1.3757090235602316e-05, "loss": 0.028, "step": 77020 }, { "grad_norm": 0.3824637532234192, "learning_rate": 1.3745701495332447e-05, "loss": 0.0318, "step": 77030 }, { "grad_norm": 0.30873867869377136, "learning_rate": 1.3734316719778145e-05, "loss": 0.0248, "step": 77040 }, { "grad_norm": 0.4124964475631714, "learning_rate": 1.3722935910184454e-05, "loss": 0.025, "step": 77050 }, { "grad_norm": 0.2962840795516968, "learning_rate": 1.3711559067795921e-05, "loss": 0.0289, "step": 77060 }, { "grad_norm": 0.2650771141052246, "learning_rate": 1.3700186193856756e-05, "loss": 0.0291, "step": 77070 }, { "grad_norm": 0.5797641277313232, "learning_rate": 1.368881728961064e-05, "loss": 0.0284, "step": 77080 }, { "grad_norm": 0.28369152545928955, "learning_rate": 1.3677452356300885e-05, "loss": 0.0263, "step": 77090 }, { "grad_norm": 0.47400322556495667, "learning_rate": 1.3666091395170321e-05, "loss": 0.0271, "step": 77100 }, { "grad_norm": 0.4446870982646942, "learning_rate": 1.3654734407461384e-05, "loss": 0.0326, "step": 77110 }, { "grad_norm": 0.444802850484848, "learning_rate": 1.3643381394416022e-05, "loss": 0.0248, "step": 77120 }, { "grad_norm": 0.3469296097755432, "learning_rate": 1.3632032357275838e-05, "loss": 0.0348, "step": 77130 }, { "grad_norm": 0.3201819360256195, "learning_rate": 1.3620687297281898e-05, "loss": 0.0287, "step": 77140 }, { "grad_norm": 0.8277168869972229, "learning_rate": 1.3609346215674917e-05, "loss": 0.0286, "step": 77150 }, { "grad_norm": 0.5681495666503906, "learning_rate": 1.3598009113695109e-05, "loss": 0.0359, "step": 77160 }, { "grad_norm": 0.2930770814418793, "learning_rate": 1.358667599258231e-05, "loss": 0.0262, "step": 77170 }, { "grad_norm": 0.2922840118408203, "learning_rate": 1.357534685357586e-05, "loss": 0.0284, "step": 77180 }, { "grad_norm": 0.5167263746261597, "learning_rate": 1.3564021697914715e-05, "loss": 0.0363, "step": 77190 }, { "grad_norm": 0.3612908124923706, "learning_rate": 1.3552700526837381e-05, "loss": 0.0334, "step": 77200 }, { "grad_norm": 0.3582973778247833, "learning_rate": 1.3541383341581925e-05, "loss": 0.0245, "step": 77210 }, { "grad_norm": 0.2654057741165161, "learning_rate": 1.3530070143385965e-05, "loss": 0.0214, "step": 77220 }, { "grad_norm": 0.27241888642311096, "learning_rate": 1.3518760933486702e-05, "loss": 0.0223, "step": 77230 }, { "grad_norm": 0.2860138416290283, "learning_rate": 1.3507455713120908e-05, "loss": 0.0261, "step": 77240 }, { "grad_norm": 0.1322336345911026, "learning_rate": 1.3496154483524876e-05, "loss": 0.0288, "step": 77250 }, { "grad_norm": 0.45930176973342896, "learning_rate": 1.3484857245934506e-05, "loss": 0.0317, "step": 77260 }, { "grad_norm": 0.275181382894516, "learning_rate": 1.3473564001585248e-05, "loss": 0.0266, "step": 77270 }, { "grad_norm": 0.31208908557891846, "learning_rate": 1.3462274751712129e-05, "loss": 0.0315, "step": 77280 }, { "grad_norm": 0.3226473927497864, "learning_rate": 1.3450989497549698e-05, "loss": 0.0255, "step": 77290 }, { "grad_norm": 0.5774602890014648, "learning_rate": 1.3439708240332115e-05, "loss": 0.0342, "step": 77300 }, { "grad_norm": 0.33885636925697327, "learning_rate": 1.3428430981293043e-05, "loss": 0.0302, "step": 77310 }, { "grad_norm": 0.7439107298851013, "learning_rate": 1.3417157721665808e-05, "loss": 0.0303, "step": 77320 }, { "grad_norm": 0.4312320649623871, "learning_rate": 1.3405888462683186e-05, "loss": 0.0255, "step": 77330 }, { "grad_norm": 0.2958572804927826, "learning_rate": 1.3394623205577595e-05, "loss": 0.0307, "step": 77340 }, { "grad_norm": 1.060199499130249, "learning_rate": 1.3383361951580963e-05, "loss": 0.0307, "step": 77350 }, { "grad_norm": 0.6145480275154114, "learning_rate": 1.337210470192483e-05, "loss": 0.0278, "step": 77360 }, { "grad_norm": 0.3508690297603607, "learning_rate": 1.3360851457840245e-05, "loss": 0.0279, "step": 77370 }, { "grad_norm": 0.27279821038246155, "learning_rate": 1.3349602220557856e-05, "loss": 0.0257, "step": 77380 }, { "grad_norm": 0.4281516373157501, "learning_rate": 1.3338356991307865e-05, "loss": 0.0285, "step": 77390 }, { "grad_norm": 0.4965902864933014, "learning_rate": 1.3327115771320048e-05, "loss": 0.0333, "step": 77400 }, { "grad_norm": 0.20103828608989716, "learning_rate": 1.3315878561823697e-05, "loss": 0.0229, "step": 77410 }, { "grad_norm": 0.8733779191970825, "learning_rate": 1.330464536404773e-05, "loss": 0.0262, "step": 77420 }, { "grad_norm": 0.25906258821487427, "learning_rate": 1.3293416179220563e-05, "loss": 0.0281, "step": 77430 }, { "grad_norm": 0.4989452064037323, "learning_rate": 1.328219100857021e-05, "loss": 0.0282, "step": 77440 }, { "grad_norm": 0.3335617780685425, "learning_rate": 1.327096985332425e-05, "loss": 0.0251, "step": 77450 }, { "grad_norm": 0.4239751100540161, "learning_rate": 1.3259752714709816e-05, "loss": 0.0255, "step": 77460 }, { "grad_norm": 0.6939396858215332, "learning_rate": 1.3248539593953569e-05, "loss": 0.0369, "step": 77470 }, { "grad_norm": 0.4257345497608185, "learning_rate": 1.323733049228178e-05, "loss": 0.0385, "step": 77480 }, { "grad_norm": 0.39790546894073486, "learning_rate": 1.3226125410920276e-05, "loss": 0.0309, "step": 77490 }, { "grad_norm": 0.35381996631622314, "learning_rate": 1.3214924351094388e-05, "loss": 0.028, "step": 77500 }, { "grad_norm": 0.3638036549091339, "learning_rate": 1.3203727314029069e-05, "loss": 0.0282, "step": 77510 }, { "grad_norm": 0.45897531509399414, "learning_rate": 1.3192534300948811e-05, "loss": 0.0332, "step": 77520 }, { "grad_norm": 0.45150405168533325, "learning_rate": 1.3181345313077675e-05, "loss": 0.0264, "step": 77530 }, { "grad_norm": 0.6688516139984131, "learning_rate": 1.3170160351639243e-05, "loss": 0.0264, "step": 77540 }, { "grad_norm": 0.3616350591182709, "learning_rate": 1.3158979417856721e-05, "loss": 0.039, "step": 77550 }, { "grad_norm": 0.5017328858375549, "learning_rate": 1.31478025129528e-05, "loss": 0.0373, "step": 77560 }, { "grad_norm": 0.47292080521583557, "learning_rate": 1.3136629638149794e-05, "loss": 0.0276, "step": 77570 }, { "grad_norm": 0.4812183678150177, "learning_rate": 1.3125460794669547e-05, "loss": 0.0282, "step": 77580 }, { "grad_norm": 0.6166753172874451, "learning_rate": 1.3114295983733483e-05, "loss": 0.0232, "step": 77590 }, { "grad_norm": 0.42020124197006226, "learning_rate": 1.3103135206562534e-05, "loss": 0.0295, "step": 77600 }, { "grad_norm": 1.0256140232086182, "learning_rate": 1.3091978464377264e-05, "loss": 0.0336, "step": 77610 }, { "grad_norm": 0.2722657322883606, "learning_rate": 1.3080825758397724e-05, "loss": 0.0295, "step": 77620 }, { "grad_norm": 0.5880249738693237, "learning_rate": 1.3069677089843568e-05, "loss": 0.0299, "step": 77630 }, { "grad_norm": 0.33791738748550415, "learning_rate": 1.305853245993401e-05, "loss": 0.0317, "step": 77640 }, { "grad_norm": 0.21428515017032623, "learning_rate": 1.3047391869887809e-05, "loss": 0.0308, "step": 77650 }, { "grad_norm": 0.3580016493797302, "learning_rate": 1.3036255320923263e-05, "loss": 0.0291, "step": 77660 }, { "grad_norm": 0.42466723918914795, "learning_rate": 1.3025122814258273e-05, "loss": 0.0249, "step": 77670 }, { "grad_norm": 0.5158302187919617, "learning_rate": 1.3013994351110249e-05, "loss": 0.0314, "step": 77680 }, { "grad_norm": 0.28411969542503357, "learning_rate": 1.3002869932696194e-05, "loss": 0.0244, "step": 77690 }, { "grad_norm": 0.40409472584724426, "learning_rate": 1.2991749560232664e-05, "loss": 0.0357, "step": 77700 }, { "grad_norm": 0.32123181223869324, "learning_rate": 1.2980633234935768e-05, "loss": 0.0242, "step": 77710 }, { "grad_norm": 0.4591493010520935, "learning_rate": 1.2969520958021153e-05, "loss": 0.0304, "step": 77720 }, { "grad_norm": 0.3715592622756958, "learning_rate": 1.295841273070405e-05, "loss": 0.0307, "step": 77730 }, { "grad_norm": 0.496728777885437, "learning_rate": 1.2947308554199255e-05, "loss": 0.0275, "step": 77740 }, { "grad_norm": 0.5543130040168762, "learning_rate": 1.2936208429721075e-05, "loss": 0.0351, "step": 77750 }, { "grad_norm": 0.4612656533718109, "learning_rate": 1.2925112358483416e-05, "loss": 0.028, "step": 77760 }, { "grad_norm": 0.3633943498134613, "learning_rate": 1.2914020341699728e-05, "loss": 0.0269, "step": 77770 }, { "grad_norm": 0.34533846378326416, "learning_rate": 1.2902932380583032e-05, "loss": 0.0296, "step": 77780 }, { "grad_norm": 0.45875078439712524, "learning_rate": 1.2891848476345863e-05, "loss": 0.0322, "step": 77790 }, { "grad_norm": 0.4405179023742676, "learning_rate": 1.2880768630200368e-05, "loss": 0.0309, "step": 77800 }, { "grad_norm": 0.46374818682670593, "learning_rate": 1.286969284335819e-05, "loss": 0.0243, "step": 77810 }, { "grad_norm": 0.1440954953432083, "learning_rate": 1.2858621117030578e-05, "loss": 0.0245, "step": 77820 }, { "grad_norm": 0.43672096729278564, "learning_rate": 1.2847553452428324e-05, "loss": 0.0214, "step": 77830 }, { "grad_norm": 0.2999340295791626, "learning_rate": 1.283648985076178e-05, "loss": 0.0201, "step": 77840 }, { "grad_norm": 0.3403584063053131, "learning_rate": 1.2825430313240817e-05, "loss": 0.0284, "step": 77850 }, { "grad_norm": 0.5523242354393005, "learning_rate": 1.281437484107491e-05, "loss": 0.0326, "step": 77860 }, { "grad_norm": 0.3816065490245819, "learning_rate": 1.2803323435473058e-05, "loss": 0.0298, "step": 77870 }, { "grad_norm": 0.3206588625907898, "learning_rate": 1.2792276097643841e-05, "loss": 0.0268, "step": 77880 }, { "grad_norm": 0.36402449011802673, "learning_rate": 1.2781232828795342e-05, "loss": 0.029, "step": 77890 }, { "grad_norm": 0.3128381371498108, "learning_rate": 1.2770193630135292e-05, "loss": 0.0369, "step": 77900 }, { "grad_norm": 0.40556439757347107, "learning_rate": 1.2759158502870877e-05, "loss": 0.0316, "step": 77910 }, { "grad_norm": 0.4706805348396301, "learning_rate": 1.2748127448208913e-05, "loss": 0.028, "step": 77920 }, { "grad_norm": 0.269500195980072, "learning_rate": 1.2737100467355706e-05, "loss": 0.0328, "step": 77930 }, { "grad_norm": 0.5369123220443726, "learning_rate": 1.2726077561517169e-05, "loss": 0.0359, "step": 77940 }, { "grad_norm": 0.30652037262916565, "learning_rate": 1.2715058731898755e-05, "loss": 0.0243, "step": 77950 }, { "grad_norm": 0.40674889087677, "learning_rate": 1.2704043979705471e-05, "loss": 0.0353, "step": 77960 }, { "grad_norm": 0.6596917510032654, "learning_rate": 1.269303330614185e-05, "loss": 0.0311, "step": 77970 }, { "grad_norm": 1.6734285354614258, "learning_rate": 1.2682026712412015e-05, "loss": 0.0278, "step": 77980 }, { "grad_norm": 0.34981104731559753, "learning_rate": 1.2671024199719644e-05, "loss": 0.0251, "step": 77990 }, { "grad_norm": 0.4496646821498871, "learning_rate": 1.2660025769267936e-05, "loss": 0.0272, "step": 78000 }, { "grad_norm": 0.4503731429576874, "learning_rate": 1.2649031422259667e-05, "loss": 0.0263, "step": 78010 }, { "grad_norm": 0.3137314021587372, "learning_rate": 1.2638041159897163e-05, "loss": 0.0278, "step": 78020 }, { "grad_norm": 0.3702201247215271, "learning_rate": 1.262705498338232e-05, "loss": 0.0258, "step": 78030 }, { "grad_norm": 0.4532356858253479, "learning_rate": 1.2616072893916542e-05, "loss": 0.0368, "step": 78040 }, { "grad_norm": 0.3773995637893677, "learning_rate": 1.2605094892700836e-05, "loss": 0.0262, "step": 78050 }, { "grad_norm": 0.31201624870300293, "learning_rate": 1.259412098093572e-05, "loss": 0.0244, "step": 78060 }, { "grad_norm": 0.28425121307373047, "learning_rate": 1.258315115982131e-05, "loss": 0.0223, "step": 78070 }, { "grad_norm": 0.4375668466091156, "learning_rate": 1.2572185430557203e-05, "loss": 0.0246, "step": 78080 }, { "grad_norm": 0.5897138118743896, "learning_rate": 1.2561223794342659e-05, "loss": 0.0274, "step": 78090 }, { "grad_norm": 0.36754563450813293, "learning_rate": 1.255026625237637e-05, "loss": 0.0333, "step": 78100 }, { "grad_norm": 0.40670594573020935, "learning_rate": 1.2539312805856678e-05, "loss": 0.0354, "step": 78110 }, { "grad_norm": 0.24941757321357727, "learning_rate": 1.25283634559814e-05, "loss": 0.0314, "step": 78120 }, { "grad_norm": 0.3106899857521057, "learning_rate": 1.2517418203947972e-05, "loss": 0.0271, "step": 78130 }, { "grad_norm": 0.34578412771224976, "learning_rate": 1.2506477050953308e-05, "loss": 0.026, "step": 78140 }, { "grad_norm": 0.312421977519989, "learning_rate": 1.2495539998193972e-05, "loss": 0.03, "step": 78150 }, { "grad_norm": 0.3808020055294037, "learning_rate": 1.2484607046865975e-05, "loss": 0.0305, "step": 78160 }, { "grad_norm": 0.35720282793045044, "learning_rate": 1.2473678198164968e-05, "loss": 0.026, "step": 78170 }, { "grad_norm": 0.6242600679397583, "learning_rate": 1.2462753453286075e-05, "loss": 0.0298, "step": 78180 }, { "grad_norm": 0.2788771688938141, "learning_rate": 1.2451832813424031e-05, "loss": 0.044, "step": 78190 }, { "grad_norm": 0.664330005645752, "learning_rate": 1.2440916279773095e-05, "loss": 0.03, "step": 78200 }, { "grad_norm": 0.3620133697986603, "learning_rate": 1.2430003853527101e-05, "loss": 0.0308, "step": 78210 }, { "grad_norm": 0.669853687286377, "learning_rate": 1.2419095535879383e-05, "loss": 0.0376, "step": 78220 }, { "grad_norm": 0.25461262464523315, "learning_rate": 1.2408191328022878e-05, "loss": 0.0315, "step": 78230 }, { "grad_norm": 0.49712592363357544, "learning_rate": 1.2397291231150066e-05, "loss": 0.0274, "step": 78240 }, { "grad_norm": 0.4322856664657593, "learning_rate": 1.2386395246452937e-05, "loss": 0.0275, "step": 78250 }, { "grad_norm": 0.27369463443756104, "learning_rate": 1.2375503375123083e-05, "loss": 0.0289, "step": 78260 }, { "grad_norm": 0.23964695632457733, "learning_rate": 1.2364615618351583e-05, "loss": 0.0369, "step": 78270 }, { "grad_norm": 0.4285459816455841, "learning_rate": 1.2353731977329169e-05, "loss": 0.0253, "step": 78280 }, { "grad_norm": 0.2343485951423645, "learning_rate": 1.2342852453246007e-05, "loss": 0.0235, "step": 78290 }, { "grad_norm": 0.27143287658691406, "learning_rate": 1.2331977047291898e-05, "loss": 0.0307, "step": 78300 }, { "grad_norm": 0.3100900948047638, "learning_rate": 1.2321105760656137e-05, "loss": 0.0301, "step": 78310 }, { "grad_norm": 0.23117336630821228, "learning_rate": 1.2310238594527613e-05, "loss": 0.0272, "step": 78320 }, { "grad_norm": 0.4067240357398987, "learning_rate": 1.2299375550094693e-05, "loss": 0.0294, "step": 78330 }, { "grad_norm": 0.3583577573299408, "learning_rate": 1.2288516628545421e-05, "loss": 0.0253, "step": 78340 }, { "grad_norm": 0.22739821672439575, "learning_rate": 1.2277661831067255e-05, "loss": 0.0222, "step": 78350 }, { "grad_norm": 0.28640004992485046, "learning_rate": 1.2266811158847286e-05, "loss": 0.031, "step": 78360 }, { "grad_norm": 0.28911110758781433, "learning_rate": 1.2255964613072107e-05, "loss": 0.0199, "step": 78370 }, { "grad_norm": 0.2721096873283386, "learning_rate": 1.2245122194927905e-05, "loss": 0.0203, "step": 78380 }, { "grad_norm": 0.2862822413444519, "learning_rate": 1.2234283905600364e-05, "loss": 0.0293, "step": 78390 }, { "grad_norm": 0.5724888443946838, "learning_rate": 1.222344974627475e-05, "loss": 0.0311, "step": 78400 }, { "grad_norm": 0.4169324040412903, "learning_rate": 1.2212619718135875e-05, "loss": 0.0287, "step": 78410 }, { "grad_norm": 0.33377814292907715, "learning_rate": 1.2201793822368118e-05, "loss": 0.0262, "step": 78420 }, { "grad_norm": 0.3499316871166229, "learning_rate": 1.2190972060155337e-05, "loss": 0.022, "step": 78430 }, { "grad_norm": 0.3153296411037445, "learning_rate": 1.2180154432681013e-05, "loss": 0.024, "step": 78440 }, { "grad_norm": 0.32476067543029785, "learning_rate": 1.2169340941128148e-05, "loss": 0.0293, "step": 78450 }, { "grad_norm": 0.46573376655578613, "learning_rate": 1.2158531586679267e-05, "loss": 0.0265, "step": 78460 }, { "grad_norm": 0.4202146828174591, "learning_rate": 1.2147726370516477e-05, "loss": 0.0282, "step": 78470 }, { "grad_norm": 0.16862307488918304, "learning_rate": 1.2136925293821422e-05, "loss": 0.0348, "step": 78480 }, { "grad_norm": 0.6139562726020813, "learning_rate": 1.2126128357775306e-05, "loss": 0.0347, "step": 78490 }, { "grad_norm": 0.48047301173210144, "learning_rate": 1.2115335563558838e-05, "loss": 0.0294, "step": 78500 }, { "grad_norm": 0.1948167085647583, "learning_rate": 1.2104546912352327e-05, "loss": 0.0256, "step": 78510 }, { "grad_norm": 0.3057428002357483, "learning_rate": 1.2093762405335557e-05, "loss": 0.026, "step": 78520 }, { "grad_norm": 0.5340549349784851, "learning_rate": 1.2082982043687974e-05, "loss": 0.0271, "step": 78530 }, { "grad_norm": 0.22229188680648804, "learning_rate": 1.207220582858845e-05, "loss": 0.0304, "step": 78540 }, { "grad_norm": 0.446159303188324, "learning_rate": 1.206143376121549e-05, "loss": 0.0257, "step": 78550 }, { "grad_norm": 0.35744744539260864, "learning_rate": 1.2050665842747078e-05, "loss": 0.0257, "step": 78560 }, { "grad_norm": 0.4986947178840637, "learning_rate": 1.2039902074360804e-05, "loss": 0.0227, "step": 78570 }, { "grad_norm": 0.3952382504940033, "learning_rate": 1.2029142457233755e-05, "loss": 0.0304, "step": 78580 }, { "grad_norm": 0.451357364654541, "learning_rate": 1.2018386992542601e-05, "loss": 0.0357, "step": 78590 }, { "grad_norm": 0.1769479662179947, "learning_rate": 1.2007635681463542e-05, "loss": 0.0247, "step": 78600 }, { "grad_norm": 0.2228330373764038, "learning_rate": 1.1996888525172334e-05, "loss": 0.027, "step": 78610 }, { "grad_norm": 0.40095818042755127, "learning_rate": 1.198614552484425e-05, "loss": 0.0227, "step": 78620 }, { "grad_norm": 0.355772465467453, "learning_rate": 1.197540668165415e-05, "loss": 0.0261, "step": 78630 }, { "grad_norm": 0.2830396294593811, "learning_rate": 1.1964671996776395e-05, "loss": 0.0327, "step": 78640 }, { "grad_norm": 0.4233264625072479, "learning_rate": 1.1953941471384922e-05, "loss": 0.0339, "step": 78650 }, { "grad_norm": 0.3419806659221649, "learning_rate": 1.1943215106653205e-05, "loss": 0.0255, "step": 78660 }, { "grad_norm": 0.5429581999778748, "learning_rate": 1.193249290375429e-05, "loss": 0.0257, "step": 78670 }, { "grad_norm": 0.35648006200790405, "learning_rate": 1.1921774863860696e-05, "loss": 0.0304, "step": 78680 }, { "grad_norm": 0.7925136089324951, "learning_rate": 1.1911060988144556e-05, "loss": 0.0328, "step": 78690 }, { "grad_norm": 0.36453238129615784, "learning_rate": 1.1900351277777533e-05, "loss": 0.0274, "step": 78700 }, { "grad_norm": 0.5317931771278381, "learning_rate": 1.1889645733930798e-05, "loss": 0.0307, "step": 78710 }, { "grad_norm": 0.5770493745803833, "learning_rate": 1.1878944357775107e-05, "loss": 0.0337, "step": 78720 }, { "grad_norm": 0.2661687731742859, "learning_rate": 1.1868247150480749e-05, "loss": 0.0242, "step": 78730 }, { "grad_norm": 0.760674774646759, "learning_rate": 1.1857554113217567e-05, "loss": 0.0291, "step": 78740 }, { "grad_norm": 0.2821599841117859, "learning_rate": 1.1846865247154903e-05, "loss": 0.0274, "step": 78750 }, { "grad_norm": 0.1948518007993698, "learning_rate": 1.1836180553461711e-05, "loss": 0.0289, "step": 78760 }, { "grad_norm": 0.4056166112422943, "learning_rate": 1.1825500033306419e-05, "loss": 0.0355, "step": 78770 }, { "grad_norm": 0.31571102142333984, "learning_rate": 1.1814823687857052e-05, "loss": 0.035, "step": 78780 }, { "grad_norm": 0.30400288105010986, "learning_rate": 1.1804151518281153e-05, "loss": 0.0253, "step": 78790 }, { "grad_norm": 0.367574542760849, "learning_rate": 1.179348352574583e-05, "loss": 0.044, "step": 78800 }, { "grad_norm": 0.3838047385215759, "learning_rate": 1.1782819711417697e-05, "loss": 0.0295, "step": 78810 }, { "grad_norm": 0.3060041069984436, "learning_rate": 1.1772160076462951e-05, "loss": 0.0287, "step": 78820 }, { "grad_norm": 1.1376628875732422, "learning_rate": 1.1761504622047293e-05, "loss": 0.0358, "step": 78830 }, { "grad_norm": 0.24123844504356384, "learning_rate": 1.1750853349335999e-05, "loss": 0.0216, "step": 78840 }, { "grad_norm": 0.34198859333992004, "learning_rate": 1.1740206259493874e-05, "loss": 0.0282, "step": 78850 }, { "grad_norm": 0.3912513554096222, "learning_rate": 1.1729563353685286e-05, "loss": 0.0221, "step": 78860 }, { "grad_norm": 0.36205026507377625, "learning_rate": 1.1718924633074091e-05, "loss": 0.0309, "step": 78870 }, { "grad_norm": 0.22196437418460846, "learning_rate": 1.1708290098823765e-05, "loss": 0.0277, "step": 78880 }, { "grad_norm": 0.4748781621456146, "learning_rate": 1.1697659752097245e-05, "loss": 0.0311, "step": 78890 }, { "grad_norm": 0.38078850507736206, "learning_rate": 1.1687033594057068e-05, "loss": 0.0419, "step": 78900 }, { "grad_norm": 0.3503747284412384, "learning_rate": 1.1676411625865297e-05, "loss": 0.0362, "step": 78910 }, { "grad_norm": 0.6096788048744202, "learning_rate": 1.1665793848683542e-05, "loss": 0.0318, "step": 78920 }, { "grad_norm": 0.6647472381591797, "learning_rate": 1.1655180263672927e-05, "loss": 0.043, "step": 78930 }, { "grad_norm": 0.8649329543113708, "learning_rate": 1.1644570871994142e-05, "loss": 0.0288, "step": 78940 }, { "grad_norm": 0.5070624351501465, "learning_rate": 1.163396567480744e-05, "loss": 0.0308, "step": 78950 }, { "grad_norm": 0.5792957544326782, "learning_rate": 1.1623364673272552e-05, "loss": 0.033, "step": 78960 }, { "grad_norm": 0.40154770016670227, "learning_rate": 1.1612767868548807e-05, "loss": 0.0274, "step": 78970 }, { "grad_norm": 0.40002790093421936, "learning_rate": 1.1602175261795057e-05, "loss": 0.0352, "step": 78980 }, { "grad_norm": 0.34645453095436096, "learning_rate": 1.1591586854169696e-05, "loss": 0.033, "step": 78990 }, { "grad_norm": 0.2593190371990204, "learning_rate": 1.158100264683064e-05, "loss": 0.0236, "step": 79000 }, { "grad_norm": 0.513519287109375, "learning_rate": 1.1570422640935386e-05, "loss": 0.0324, "step": 79010 }, { "grad_norm": 0.29345688223838806, "learning_rate": 1.155984683764092e-05, "loss": 0.0364, "step": 79020 }, { "grad_norm": 0.3721565008163452, "learning_rate": 1.1549275238103812e-05, "loss": 0.0299, "step": 79030 }, { "grad_norm": 0.28917476534843445, "learning_rate": 1.1538707843480152e-05, "loss": 0.0311, "step": 79040 }, { "grad_norm": 0.29607897996902466, "learning_rate": 1.1528144654925593e-05, "loss": 0.0327, "step": 79050 }, { "grad_norm": 0.2982860505580902, "learning_rate": 1.1517585673595277e-05, "loss": 0.029, "step": 79060 }, { "grad_norm": 0.25551044940948486, "learning_rate": 1.150703090064395e-05, "loss": 0.0259, "step": 79070 }, { "grad_norm": 0.36336827278137207, "learning_rate": 1.1496480337225835e-05, "loss": 0.032, "step": 79080 }, { "grad_norm": 0.29569247364997864, "learning_rate": 1.1485933984494751e-05, "loss": 0.0347, "step": 79090 }, { "grad_norm": 0.36046910285949707, "learning_rate": 1.1475391843604e-05, "loss": 0.0288, "step": 79100 }, { "grad_norm": 0.2754097282886505, "learning_rate": 1.1464853915706503e-05, "loss": 0.0349, "step": 79110 }, { "grad_norm": 0.8002909421920776, "learning_rate": 1.1454320201954626e-05, "loss": 0.0309, "step": 79120 }, { "grad_norm": 0.2295936644077301, "learning_rate": 1.1443790703500356e-05, "loss": 0.0293, "step": 79130 }, { "grad_norm": 0.3954106867313385, "learning_rate": 1.1433265421495154e-05, "loss": 0.0259, "step": 79140 }, { "grad_norm": 0.3108268678188324, "learning_rate": 1.1422744357090065e-05, "loss": 0.0331, "step": 79150 }, { "grad_norm": 0.4240327477455139, "learning_rate": 1.1412227511435647e-05, "loss": 0.0297, "step": 79160 }, { "grad_norm": 0.3834955096244812, "learning_rate": 1.1401714885682025e-05, "loss": 0.0232, "step": 79170 }, { "grad_norm": 0.7419404983520508, "learning_rate": 1.1391206480978823e-05, "loss": 0.0251, "step": 79180 }, { "grad_norm": 0.475026398897171, "learning_rate": 1.1380702298475227e-05, "loss": 0.0313, "step": 79190 }, { "grad_norm": 0.35524049401283264, "learning_rate": 1.1370202339319985e-05, "loss": 0.0271, "step": 79200 }, { "grad_norm": 0.45263123512268066, "learning_rate": 1.1359706604661319e-05, "loss": 0.0318, "step": 79210 }, { "grad_norm": 0.5651861429214478, "learning_rate": 1.1349215095647042e-05, "loss": 0.0263, "step": 79220 }, { "grad_norm": 0.3542783856391907, "learning_rate": 1.1338727813424493e-05, "loss": 0.0252, "step": 79230 }, { "grad_norm": 0.33636167645454407, "learning_rate": 1.132824475914056e-05, "loss": 0.0251, "step": 79240 }, { "grad_norm": 0.7520842552185059, "learning_rate": 1.131776593394162e-05, "loss": 0.0226, "step": 79250 }, { "grad_norm": 1.3547793626785278, "learning_rate": 1.1307291338973657e-05, "loss": 0.0312, "step": 79260 }, { "grad_norm": 0.23743309080600739, "learning_rate": 1.1296820975382121e-05, "loss": 0.0275, "step": 79270 }, { "grad_norm": 0.22675935924053192, "learning_rate": 1.1286354844312074e-05, "loss": 0.0344, "step": 79280 }, { "grad_norm": 0.23479469120502472, "learning_rate": 1.1275892946908023e-05, "loss": 0.0269, "step": 79290 }, { "grad_norm": 0.3621223270893097, "learning_rate": 1.126543528431413e-05, "loss": 0.0429, "step": 79300 }, { "grad_norm": 0.5744609236717224, "learning_rate": 1.125498185767398e-05, "loss": 0.0292, "step": 79310 }, { "grad_norm": 0.1940276324748993, "learning_rate": 1.1244532668130781e-05, "loss": 0.0261, "step": 79320 }, { "grad_norm": 0.2371009737253189, "learning_rate": 1.1234087716827203e-05, "loss": 0.0279, "step": 79330 }, { "grad_norm": 0.34920617938041687, "learning_rate": 1.1223647004905524e-05, "loss": 0.0364, "step": 79340 }, { "grad_norm": 0.34323716163635254, "learning_rate": 1.1213210533507485e-05, "loss": 0.0352, "step": 79350 }, { "grad_norm": 0.421639621257782, "learning_rate": 1.1202778303774447e-05, "loss": 0.0226, "step": 79360 }, { "grad_norm": 0.6965311169624329, "learning_rate": 1.1192350316847234e-05, "loss": 0.0288, "step": 79370 }, { "grad_norm": 0.2989729046821594, "learning_rate": 1.1181926573866258e-05, "loss": 0.0292, "step": 79380 }, { "grad_norm": 0.3233281672000885, "learning_rate": 1.1171507075971416e-05, "loss": 0.028, "step": 79390 }, { "grad_norm": 0.2933317720890045, "learning_rate": 1.1161091824302177e-05, "loss": 0.0288, "step": 79400 }, { "grad_norm": 0.39345598220825195, "learning_rate": 1.115068081999755e-05, "loss": 0.0289, "step": 79410 }, { "grad_norm": 0.45789584517478943, "learning_rate": 1.1140274064196071e-05, "loss": 0.0265, "step": 79420 }, { "grad_norm": 0.3842420279979706, "learning_rate": 1.112987155803578e-05, "loss": 0.03, "step": 79430 }, { "grad_norm": 0.49471959471702576, "learning_rate": 1.1119473302654298e-05, "loss": 0.0299, "step": 79440 }, { "grad_norm": 1.1510999202728271, "learning_rate": 1.1109079299188769e-05, "loss": 0.0292, "step": 79450 }, { "grad_norm": 0.3575133681297302, "learning_rate": 1.1098689548775847e-05, "loss": 0.0283, "step": 79460 }, { "grad_norm": 0.4376008212566376, "learning_rate": 1.1088304052551762e-05, "loss": 0.0261, "step": 79470 }, { "grad_norm": 0.33580461144447327, "learning_rate": 1.1077922811652215e-05, "loss": 0.0334, "step": 79480 }, { "grad_norm": 0.5799779295921326, "learning_rate": 1.1067545827212534e-05, "loss": 0.031, "step": 79490 }, { "grad_norm": 0.5514068603515625, "learning_rate": 1.1057173100367496e-05, "loss": 0.0235, "step": 79500 }, { "grad_norm": 0.39127326011657715, "learning_rate": 1.1046804632251474e-05, "loss": 0.0296, "step": 79510 }, { "grad_norm": 0.31100428104400635, "learning_rate": 1.1036440423998318e-05, "loss": 0.0211, "step": 79520 }, { "grad_norm": 0.3834841847419739, "learning_rate": 1.1026080476741469e-05, "loss": 0.0328, "step": 79530 }, { "grad_norm": 0.3807423412799835, "learning_rate": 1.1015724791613836e-05, "loss": 0.0205, "step": 79540 }, { "grad_norm": 0.474298894405365, "learning_rate": 1.1005373369747951e-05, "loss": 0.0335, "step": 79550 }, { "grad_norm": 0.27694249153137207, "learning_rate": 1.09950262122758e-05, "loss": 0.0273, "step": 79560 }, { "grad_norm": 0.44702428579330444, "learning_rate": 1.0984683320328948e-05, "loss": 0.0278, "step": 79570 }, { "grad_norm": 0.27341964840888977, "learning_rate": 1.0974344695038458e-05, "loss": 0.028, "step": 79580 }, { "grad_norm": 0.28465574979782104, "learning_rate": 1.0964010337534968e-05, "loss": 0.0313, "step": 79590 }, { "grad_norm": 0.3917592465877533, "learning_rate": 1.0953680248948611e-05, "loss": 0.0262, "step": 79600 }, { "grad_norm": 0.2751794159412384, "learning_rate": 1.0943354430409069e-05, "loss": 0.0203, "step": 79610 }, { "grad_norm": 0.7095345258712769, "learning_rate": 1.0933032883045574e-05, "loss": 0.0248, "step": 79620 }, { "grad_norm": 0.4503999650478363, "learning_rate": 1.0922715607986872e-05, "loss": 0.035, "step": 79630 }, { "grad_norm": 0.4065204858779907, "learning_rate": 1.0912402606361234e-05, "loss": 0.0247, "step": 79640 }, { "grad_norm": 0.44527173042297363, "learning_rate": 1.090209387929647e-05, "loss": 0.0185, "step": 79650 }, { "grad_norm": 0.5725123882293701, "learning_rate": 1.0891789427919957e-05, "loss": 0.0269, "step": 79660 }, { "grad_norm": 0.22158898413181305, "learning_rate": 1.088148925335854e-05, "loss": 0.0222, "step": 79670 }, { "grad_norm": 0.4479297995567322, "learning_rate": 1.087119335673864e-05, "loss": 0.0305, "step": 79680 }, { "grad_norm": 0.2743763029575348, "learning_rate": 1.0860901739186208e-05, "loss": 0.0216, "step": 79690 }, { "grad_norm": 0.39208152890205383, "learning_rate": 1.085061440182673e-05, "loss": 0.0272, "step": 79700 }, { "grad_norm": 0.5010873675346375, "learning_rate": 1.0840331345785193e-05, "loss": 0.0269, "step": 79710 }, { "grad_norm": 0.6126912832260132, "learning_rate": 1.083005257218615e-05, "loss": 0.0293, "step": 79720 }, { "grad_norm": 0.3733692169189453, "learning_rate": 1.0819778082153642e-05, "loss": 0.0231, "step": 79730 }, { "grad_norm": 0.6986539959907532, "learning_rate": 1.0809507876811325e-05, "loss": 0.0323, "step": 79740 }, { "grad_norm": 0.2728273868560791, "learning_rate": 1.0799241957282292e-05, "loss": 0.0326, "step": 79750 }, { "grad_norm": 0.319487065076828, "learning_rate": 1.078898032468923e-05, "loss": 0.0269, "step": 79760 }, { "grad_norm": 0.3646259605884552, "learning_rate": 1.0778722980154315e-05, "loss": 0.0315, "step": 79770 }, { "grad_norm": 0.3529636859893799, "learning_rate": 1.0768469924799302e-05, "loss": 0.0256, "step": 79780 }, { "grad_norm": 0.5997932553291321, "learning_rate": 1.0758221159745419e-05, "loss": 0.0388, "step": 79790 }, { "grad_norm": 0.32049083709716797, "learning_rate": 1.074797668611347e-05, "loss": 0.0236, "step": 79800 }, { "grad_norm": 0.20647019147872925, "learning_rate": 1.0737736505023777e-05, "loss": 0.0229, "step": 79810 }, { "grad_norm": 0.9786174893379211, "learning_rate": 1.0727500617596203e-05, "loss": 0.0305, "step": 79820 }, { "grad_norm": 0.6343668699264526, "learning_rate": 1.0717269024950104e-05, "loss": 0.0276, "step": 79830 }, { "grad_norm": 0.22746187448501587, "learning_rate": 1.070704172820442e-05, "loss": 0.0275, "step": 79840 }, { "grad_norm": 0.3085644841194153, "learning_rate": 1.0696818728477559e-05, "loss": 0.0282, "step": 79850 }, { "grad_norm": 0.21066182851791382, "learning_rate": 1.0686600026887511e-05, "loss": 0.0227, "step": 79860 }, { "grad_norm": 0.37667492032051086, "learning_rate": 1.0676385624551777e-05, "loss": 0.0349, "step": 79870 }, { "grad_norm": 0.7284114360809326, "learning_rate": 1.0666175522587402e-05, "loss": 0.0341, "step": 79880 }, { "grad_norm": 0.4070889949798584, "learning_rate": 1.0655969722110925e-05, "loss": 0.0227, "step": 79890 }, { "grad_norm": 0.6144634485244751, "learning_rate": 1.0645768224238439e-05, "loss": 0.0226, "step": 79900 }, { "grad_norm": 0.3243190348148346, "learning_rate": 1.063557103008559e-05, "loss": 0.0316, "step": 79910 }, { "grad_norm": 0.2466018944978714, "learning_rate": 1.0625378140767494e-05, "loss": 0.0221, "step": 79920 }, { "grad_norm": 0.2843840420246124, "learning_rate": 1.0615189557398846e-05, "loss": 0.0244, "step": 79930 }, { "grad_norm": 1.0937786102294922, "learning_rate": 1.060500528109386e-05, "loss": 0.0313, "step": 79940 }, { "grad_norm": 0.2686738967895508, "learning_rate": 1.0594825312966272e-05, "loss": 0.0264, "step": 79950 }, { "grad_norm": 0.5810763835906982, "learning_rate": 1.0584649654129336e-05, "loss": 0.0321, "step": 79960 }, { "grad_norm": 0.33893296122550964, "learning_rate": 1.0574478305695863e-05, "loss": 0.0233, "step": 79970 }, { "grad_norm": 0.3280101418495178, "learning_rate": 1.0564311268778154e-05, "loss": 0.0245, "step": 79980 }, { "grad_norm": 0.5076180100440979, "learning_rate": 1.0554148544488073e-05, "loss": 0.0247, "step": 79990 }, { "grad_norm": 0.46136894822120667, "learning_rate": 1.0543990133937004e-05, "loss": 0.0292, "step": 80000 }, { "grad_norm": 0.5961599349975586, "learning_rate": 1.0533836038235868e-05, "loss": 0.0275, "step": 80010 }, { "grad_norm": 0.30235710740089417, "learning_rate": 1.052368625849507e-05, "loss": 0.0265, "step": 80020 }, { "grad_norm": 0.6733760237693787, "learning_rate": 1.05135407958246e-05, "loss": 0.027, "step": 80030 }, { "grad_norm": 0.19611896574497223, "learning_rate": 1.0503399651333934e-05, "loss": 0.0312, "step": 80040 }, { "grad_norm": 0.29764100909233093, "learning_rate": 1.04932628261321e-05, "loss": 0.0237, "step": 80050 }, { "grad_norm": 0.7404834032058716, "learning_rate": 1.0483130321327644e-05, "loss": 0.0297, "step": 80060 }, { "grad_norm": 0.22296613454818726, "learning_rate": 1.0473002138028654e-05, "loss": 0.0221, "step": 80070 }, { "grad_norm": 0.46119898557662964, "learning_rate": 1.0462878277342713e-05, "loss": 0.0273, "step": 80080 }, { "grad_norm": 0.2971404790878296, "learning_rate": 1.0452758740376972e-05, "loss": 0.0266, "step": 80090 }, { "grad_norm": 0.41712576150894165, "learning_rate": 1.0442643528238067e-05, "loss": 0.0215, "step": 80100 }, { "grad_norm": 0.4098515510559082, "learning_rate": 1.0432532642032188e-05, "loss": 0.0238, "step": 80110 }, { "grad_norm": 0.25505971908569336, "learning_rate": 1.042242608286505e-05, "loss": 0.0284, "step": 80120 }, { "grad_norm": 0.5322816967964172, "learning_rate": 1.041232385184191e-05, "loss": 0.0249, "step": 80130 }, { "grad_norm": 0.3586719036102295, "learning_rate": 1.0402225950067495e-05, "loss": 0.0264, "step": 80140 }, { "grad_norm": 0.3046956956386566, "learning_rate": 1.0392132378646119e-05, "loss": 0.032, "step": 80150 }, { "grad_norm": 0.41126707196235657, "learning_rate": 1.0382043138681607e-05, "loss": 0.0371, "step": 80160 }, { "grad_norm": 0.8627371788024902, "learning_rate": 1.0371958231277284e-05, "loss": 0.0236, "step": 80170 }, { "grad_norm": 0.4060733914375305, "learning_rate": 1.0361877657536024e-05, "loss": 0.0289, "step": 80180 }, { "grad_norm": 0.580207884311676, "learning_rate": 1.035180141856023e-05, "loss": 0.0296, "step": 80190 }, { "grad_norm": 0.36200350522994995, "learning_rate": 1.0341729515451836e-05, "loss": 0.0302, "step": 80200 }, { "grad_norm": 0.34713608026504517, "learning_rate": 1.0331661949312265e-05, "loss": 0.0235, "step": 80210 }, { "grad_norm": 0.3837696313858032, "learning_rate": 1.0321598721242514e-05, "loss": 0.0287, "step": 80220 }, { "grad_norm": 0.6469347476959229, "learning_rate": 1.0311539832343064e-05, "loss": 0.0299, "step": 80230 }, { "grad_norm": 0.3774886131286621, "learning_rate": 1.0301485283713942e-05, "loss": 0.0252, "step": 80240 }, { "grad_norm": 0.5465601682662964, "learning_rate": 1.0291435076454703e-05, "loss": 0.0317, "step": 80250 }, { "grad_norm": 0.48766010999679565, "learning_rate": 1.028138921166444e-05, "loss": 0.0289, "step": 80260 }, { "grad_norm": 0.5446478128433228, "learning_rate": 1.0271347690441719e-05, "loss": 0.0275, "step": 80270 }, { "grad_norm": 0.3333132266998291, "learning_rate": 1.0261310513884698e-05, "loss": 0.0372, "step": 80280 }, { "grad_norm": 0.30053645372390747, "learning_rate": 1.0251277683090998e-05, "loss": 0.0287, "step": 80290 }, { "grad_norm": 0.5095464587211609, "learning_rate": 1.0241249199157826e-05, "loss": 0.0321, "step": 80300 }, { "grad_norm": 0.4012162387371063, "learning_rate": 1.0231225063181837e-05, "loss": 0.0219, "step": 80310 }, { "grad_norm": 0.4438728094100952, "learning_rate": 1.0221205276259316e-05, "loss": 0.0276, "step": 80320 }, { "grad_norm": 0.4765984117984772, "learning_rate": 1.0211189839485958e-05, "loss": 0.0285, "step": 80330 }, { "grad_norm": 0.5374252200126648, "learning_rate": 1.0201178753957075e-05, "loss": 0.0242, "step": 80340 }, { "grad_norm": 0.4423403739929199, "learning_rate": 1.0191172020767432e-05, "loss": 0.0298, "step": 80350 }, { "grad_norm": 0.3209631145000458, "learning_rate": 1.0181169641011362e-05, "loss": 0.0279, "step": 80360 }, { "grad_norm": 0.38862013816833496, "learning_rate": 1.0171171615782721e-05, "loss": 0.0273, "step": 80370 }, { "grad_norm": 0.37126195430755615, "learning_rate": 1.0161177946174866e-05, "loss": 0.0383, "step": 80380 }, { "grad_norm": 0.26340818405151367, "learning_rate": 1.0151188633280706e-05, "loss": 0.0271, "step": 80390 }, { "grad_norm": 0.43383389711380005, "learning_rate": 1.0141203678192635e-05, "loss": 0.0281, "step": 80400 }, { "grad_norm": 0.45303308963775635, "learning_rate": 1.0131223082002612e-05, "loss": 0.0364, "step": 80410 }, { "grad_norm": 0.505443811416626, "learning_rate": 1.012124684580208e-05, "loss": 0.0272, "step": 80420 }, { "grad_norm": 0.23591847717761993, "learning_rate": 1.0111274970682039e-05, "loss": 0.0206, "step": 80430 }, { "grad_norm": 0.32616478204727173, "learning_rate": 1.010130745773299e-05, "loss": 0.0218, "step": 80440 }, { "grad_norm": 0.4886721968650818, "learning_rate": 1.0091344308044986e-05, "loss": 0.0214, "step": 80450 }, { "grad_norm": 0.6621180176734924, "learning_rate": 1.0081385522707554e-05, "loss": 0.0235, "step": 80460 }, { "grad_norm": 0.4262053966522217, "learning_rate": 1.0071431102809803e-05, "loss": 0.0297, "step": 80470 }, { "grad_norm": 0.2724291682243347, "learning_rate": 1.0061481049440297e-05, "loss": 0.0289, "step": 80480 }, { "grad_norm": 0.29116353392601013, "learning_rate": 1.0051535363687187e-05, "loss": 0.0169, "step": 80490 }, { "grad_norm": 0.39395955204963684, "learning_rate": 1.0041594046638087e-05, "loss": 0.0246, "step": 80500 }, { "grad_norm": 0.4077853262424469, "learning_rate": 1.0031657099380204e-05, "loss": 0.0285, "step": 80510 }, { "grad_norm": 0.5278229713439941, "learning_rate": 1.0021724523000204e-05, "loss": 0.0278, "step": 80520 }, { "grad_norm": 0.3471246361732483, "learning_rate": 1.001179631858431e-05, "loss": 0.0282, "step": 80530 }, { "grad_norm": 0.4733841121196747, "learning_rate": 1.0001872487218245e-05, "loss": 0.0242, "step": 80540 }, { "grad_norm": 0.23007726669311523, "learning_rate": 9.991953029987272e-06, "loss": 0.0233, "step": 80550 }, { "grad_norm": 0.44711053371429443, "learning_rate": 9.982037947976147e-06, "loss": 0.033, "step": 80560 }, { "grad_norm": 0.7149284482002258, "learning_rate": 9.972127242269208e-06, "loss": 0.0287, "step": 80570 }, { "grad_norm": 0.22399036586284637, "learning_rate": 9.962220913950238e-06, "loss": 0.0332, "step": 80580 }, { "grad_norm": 0.36548247933387756, "learning_rate": 9.952318964102591e-06, "loss": 0.0336, "step": 80590 }, { "grad_norm": 0.663730263710022, "learning_rate": 9.942421393809148e-06, "loss": 0.0281, "step": 80600 }, { "grad_norm": 0.3372008502483368, "learning_rate": 9.932528204152264e-06, "loss": 0.0223, "step": 80610 }, { "grad_norm": 0.2531547546386719, "learning_rate": 9.922639396213856e-06, "loss": 0.0462, "step": 80620 }, { "grad_norm": 0.5018501877784729, "learning_rate": 9.912754971075344e-06, "loss": 0.0232, "step": 80630 }, { "grad_norm": 0.7740896344184875, "learning_rate": 9.902874929817696e-06, "loss": 0.0306, "step": 80640 }, { "grad_norm": 0.2795960307121277, "learning_rate": 9.892999273521342e-06, "loss": 0.0255, "step": 80650 }, { "grad_norm": 0.30545440316200256, "learning_rate": 9.883128003266302e-06, "loss": 0.0258, "step": 80660 }, { "grad_norm": 0.44769486784935, "learning_rate": 9.87326112013206e-06, "loss": 0.0269, "step": 80670 }, { "grad_norm": 0.8335839509963989, "learning_rate": 9.863398625197662e-06, "loss": 0.0251, "step": 80680 }, { "grad_norm": 0.7068095207214355, "learning_rate": 9.853540519541615e-06, "loss": 0.0332, "step": 80690 }, { "grad_norm": 0.31679466366767883, "learning_rate": 9.843686804242053e-06, "loss": 0.0282, "step": 80700 }, { "grad_norm": 0.4069739580154419, "learning_rate": 9.833837480376506e-06, "loss": 0.0311, "step": 80710 }, { "grad_norm": 0.24627962708473206, "learning_rate": 9.82399254902212e-06, "loss": 0.0394, "step": 80720 }, { "grad_norm": 0.46053504943847656, "learning_rate": 9.814152011255494e-06, "loss": 0.0227, "step": 80730 }, { "grad_norm": 0.43579918146133423, "learning_rate": 9.804315868152797e-06, "loss": 0.027, "step": 80740 }, { "grad_norm": 0.447064608335495, "learning_rate": 9.79448412078966e-06, "loss": 0.0254, "step": 80750 }, { "grad_norm": 0.3554401695728302, "learning_rate": 9.784656770241312e-06, "loss": 0.0285, "step": 80760 }, { "grad_norm": 0.4410672187805176, "learning_rate": 9.77483381758243e-06, "loss": 0.0199, "step": 80770 }, { "grad_norm": 0.6145389080047607, "learning_rate": 9.76501526388725e-06, "loss": 0.0315, "step": 80780 }, { "grad_norm": 0.24167825281620026, "learning_rate": 9.7552011102295e-06, "loss": 0.0504, "step": 80790 }, { "grad_norm": 0.2827105224132538, "learning_rate": 9.745391357682449e-06, "loss": 0.034, "step": 80800 }, { "grad_norm": 0.2072351574897766, "learning_rate": 9.735586007318875e-06, "loss": 0.0366, "step": 80810 }, { "grad_norm": 0.24257594347000122, "learning_rate": 9.725785060211096e-06, "loss": 0.0279, "step": 80820 }, { "grad_norm": 0.5181000232696533, "learning_rate": 9.715988517430896e-06, "loss": 0.0283, "step": 80830 }, { "grad_norm": 0.4433134198188782, "learning_rate": 9.706196380049625e-06, "loss": 0.0179, "step": 80840 }, { "grad_norm": 0.2618207633495331, "learning_rate": 9.69640864913815e-06, "loss": 0.0261, "step": 80850 }, { "grad_norm": 0.4033956825733185, "learning_rate": 9.686625325766818e-06, "loss": 0.0213, "step": 80860 }, { "grad_norm": 0.3679892420768738, "learning_rate": 9.67684641100554e-06, "loss": 0.029, "step": 80870 }, { "grad_norm": 0.37898340821266174, "learning_rate": 9.667071905923692e-06, "loss": 0.0313, "step": 80880 }, { "grad_norm": 0.7006948590278625, "learning_rate": 9.657301811590242e-06, "loss": 0.0351, "step": 80890 }, { "grad_norm": 0.3302038013935089, "learning_rate": 9.647536129073603e-06, "loss": 0.0241, "step": 80900 }, { "grad_norm": 0.42797133326530457, "learning_rate": 9.637774859441751e-06, "loss": 0.0248, "step": 80910 }, { "grad_norm": 1.0019086599349976, "learning_rate": 9.628018003762152e-06, "loss": 0.0324, "step": 80920 }, { "grad_norm": 2.224425792694092, "learning_rate": 9.618265563101813e-06, "loss": 0.0336, "step": 80930 }, { "grad_norm": 0.37723180651664734, "learning_rate": 9.608517538527211e-06, "loss": 0.0376, "step": 80940 }, { "grad_norm": 0.3901399075984955, "learning_rate": 9.598773931104431e-06, "loss": 0.0329, "step": 80950 }, { "grad_norm": 0.719394326210022, "learning_rate": 9.589034741898983e-06, "loss": 0.029, "step": 80960 }, { "grad_norm": 0.3758350908756256, "learning_rate": 9.579299971975947e-06, "loss": 0.0265, "step": 80970 }, { "grad_norm": 0.2876198887825012, "learning_rate": 9.569569622399882e-06, "loss": 0.0223, "step": 80980 }, { "grad_norm": 0.3129623234272003, "learning_rate": 9.559843694234916e-06, "loss": 0.0176, "step": 80990 }, { "grad_norm": 0.43986645340919495, "learning_rate": 9.550122188544625e-06, "loss": 0.0284, "step": 81000 }, { "grad_norm": 0.5059196352958679, "learning_rate": 9.540405106392152e-06, "loss": 0.0258, "step": 81010 }, { "grad_norm": 0.22504876554012299, "learning_rate": 9.53069244884015e-06, "loss": 0.0241, "step": 81020 }, { "grad_norm": 0.28437671065330505, "learning_rate": 9.52098421695079e-06, "loss": 0.0261, "step": 81030 }, { "grad_norm": 0.5405462384223938, "learning_rate": 9.511280411785728e-06, "loss": 0.0275, "step": 81040 }, { "grad_norm": 0.23251482844352722, "learning_rate": 9.50158103440616e-06, "loss": 0.0268, "step": 81050 }, { "grad_norm": 0.3123440146446228, "learning_rate": 9.491886085872815e-06, "loss": 0.0227, "step": 81060 }, { "grad_norm": 0.23776593804359436, "learning_rate": 9.482195567245895e-06, "loss": 0.0312, "step": 81070 }, { "grad_norm": 0.2950146496295929, "learning_rate": 9.472509479585151e-06, "loss": 0.0311, "step": 81080 }, { "grad_norm": 0.5425861477851868, "learning_rate": 9.462827823949833e-06, "loss": 0.0291, "step": 81090 }, { "grad_norm": 0.7469035983085632, "learning_rate": 9.453150601398735e-06, "loss": 0.0251, "step": 81100 }, { "grad_norm": 0.33589327335357666, "learning_rate": 9.44347781299011e-06, "loss": 0.0258, "step": 81110 }, { "grad_norm": 0.3494504690170288, "learning_rate": 9.43380945978179e-06, "loss": 0.0302, "step": 81120 }, { "grad_norm": 0.737774670124054, "learning_rate": 9.424145542831047e-06, "loss": 0.0286, "step": 81130 }, { "grad_norm": 0.2453487664461136, "learning_rate": 9.414486063194778e-06, "loss": 0.0266, "step": 81140 }, { "grad_norm": 0.7175225019454956, "learning_rate": 9.404831021929273e-06, "loss": 0.0301, "step": 81150 }, { "grad_norm": 0.3818569779396057, "learning_rate": 9.395180420090427e-06, "loss": 0.0311, "step": 81160 }, { "grad_norm": 0.5043789148330688, "learning_rate": 9.385534258733591e-06, "loss": 0.0276, "step": 81170 }, { "grad_norm": 0.28978654742240906, "learning_rate": 9.375892538913677e-06, "loss": 0.0377, "step": 81180 }, { "grad_norm": 0.37847796082496643, "learning_rate": 9.36625526168507e-06, "loss": 0.0263, "step": 81190 }, { "grad_norm": 0.34060755372047424, "learning_rate": 9.356622428101686e-06, "loss": 0.035, "step": 81200 }, { "grad_norm": 0.3639647364616394, "learning_rate": 9.346994039216972e-06, "loss": 0.0258, "step": 81210 }, { "grad_norm": 0.2702743411064148, "learning_rate": 9.337370096083886e-06, "loss": 0.0209, "step": 81220 }, { "grad_norm": 0.44637683033943176, "learning_rate": 9.327750599754854e-06, "loss": 0.0333, "step": 81230 }, { "grad_norm": 0.3805011808872223, "learning_rate": 9.318135551281875e-06, "loss": 0.0374, "step": 81240 }, { "grad_norm": 0.3110862076282501, "learning_rate": 9.308524951716424e-06, "loss": 0.0282, "step": 81250 }, { "grad_norm": 0.31137615442276, "learning_rate": 9.298918802109497e-06, "loss": 0.0301, "step": 81260 }, { "grad_norm": 0.519045352935791, "learning_rate": 9.289317103511625e-06, "loss": 0.0233, "step": 81270 }, { "grad_norm": 0.30203941464424133, "learning_rate": 9.279719856972835e-06, "loss": 0.0258, "step": 81280 }, { "grad_norm": 0.4245211184024811, "learning_rate": 9.270127063542645e-06, "loss": 0.0306, "step": 81290 }, { "grad_norm": 0.44981902837753296, "learning_rate": 9.260538724270124e-06, "loss": 0.0331, "step": 81300 }, { "grad_norm": 0.4462563991546631, "learning_rate": 9.250954840203852e-06, "loss": 0.0245, "step": 81310 }, { "grad_norm": 0.24951094388961792, "learning_rate": 9.241375412391878e-06, "loss": 0.0246, "step": 81320 }, { "grad_norm": 0.4591754972934723, "learning_rate": 9.23180044188181e-06, "loss": 0.0303, "step": 81330 }, { "grad_norm": 0.4230632781982422, "learning_rate": 9.222229929720755e-06, "loss": 0.026, "step": 81340 }, { "grad_norm": 0.3335472345352173, "learning_rate": 9.212663876955336e-06, "loss": 0.0375, "step": 81350 }, { "grad_norm": 0.38738423585891724, "learning_rate": 9.20310228463166e-06, "loss": 0.0335, "step": 81360 }, { "grad_norm": 0.3584310710430145, "learning_rate": 9.19354515379539e-06, "loss": 0.0228, "step": 81370 }, { "grad_norm": 0.7001392841339111, "learning_rate": 9.183992485491666e-06, "loss": 0.0309, "step": 81380 }, { "grad_norm": 0.31697139143943787, "learning_rate": 9.174444280765154e-06, "loss": 0.0258, "step": 81390 }, { "grad_norm": 0.5078420042991638, "learning_rate": 9.164900540660032e-06, "loss": 0.0308, "step": 81400 }, { "grad_norm": 0.27413037419319153, "learning_rate": 9.155361266220008e-06, "loss": 0.0236, "step": 81410 }, { "grad_norm": 0.5141788721084595, "learning_rate": 9.145826458488254e-06, "loss": 0.0236, "step": 81420 }, { "grad_norm": 0.2273004800081253, "learning_rate": 9.136296118507509e-06, "loss": 0.0277, "step": 81430 }, { "grad_norm": 0.3432285189628601, "learning_rate": 9.126770247319966e-06, "loss": 0.0295, "step": 81440 }, { "grad_norm": 0.5095502138137817, "learning_rate": 9.11724884596738e-06, "loss": 0.0211, "step": 81450 }, { "grad_norm": 0.4103270471096039, "learning_rate": 9.107731915490996e-06, "loss": 0.0217, "step": 81460 }, { "grad_norm": 0.17176465690135956, "learning_rate": 9.098219456931579e-06, "loss": 0.0253, "step": 81470 }, { "grad_norm": 0.33918240666389465, "learning_rate": 9.088711471329376e-06, "loss": 0.0332, "step": 81480 }, { "grad_norm": 0.4188065230846405, "learning_rate": 9.079207959724184e-06, "loss": 0.0308, "step": 81490 }, { "grad_norm": 0.42065900564193726, "learning_rate": 9.069708923155279e-06, "loss": 0.0304, "step": 81500 }, { "grad_norm": 0.33001571893692017, "learning_rate": 9.060214362661468e-06, "loss": 0.0293, "step": 81510 }, { "grad_norm": 0.263317346572876, "learning_rate": 9.050724279281064e-06, "loss": 0.0273, "step": 81520 }, { "grad_norm": 0.40109631419181824, "learning_rate": 9.041238674051894e-06, "loss": 0.0303, "step": 81530 }, { "grad_norm": 0.5537015199661255, "learning_rate": 9.031757548011277e-06, "loss": 0.0378, "step": 81540 }, { "grad_norm": 0.41065433621406555, "learning_rate": 9.022280902196051e-06, "loss": 0.0295, "step": 81550 }, { "grad_norm": 0.2630983591079712, "learning_rate": 9.012808737642592e-06, "loss": 0.0278, "step": 81560 }, { "grad_norm": 0.32648104429244995, "learning_rate": 9.003341055386739e-06, "loss": 0.0285, "step": 81570 }, { "grad_norm": 0.22187954187393188, "learning_rate": 8.993877856463862e-06, "loss": 0.0255, "step": 81580 }, { "grad_norm": 0.22826631367206573, "learning_rate": 8.984419141908856e-06, "loss": 0.026, "step": 81590 }, { "grad_norm": 0.2222990095615387, "learning_rate": 8.974964912756117e-06, "loss": 0.0257, "step": 81600 }, { "grad_norm": 0.4877168834209442, "learning_rate": 8.965515170039523e-06, "loss": 0.0288, "step": 81610 }, { "grad_norm": 0.2555592656135559, "learning_rate": 8.956069914792514e-06, "loss": 0.0258, "step": 81620 }, { "grad_norm": 0.41925248503685, "learning_rate": 8.946629148047969e-06, "loss": 0.0315, "step": 81630 }, { "grad_norm": 0.37349677085876465, "learning_rate": 8.93719287083834e-06, "loss": 0.0327, "step": 81640 }, { "grad_norm": 0.8057994246482849, "learning_rate": 8.927761084195569e-06, "loss": 0.0335, "step": 81650 }, { "grad_norm": 0.39286547899246216, "learning_rate": 8.918333789151096e-06, "loss": 0.0277, "step": 81660 }, { "grad_norm": 0.2899935245513916, "learning_rate": 8.90891098673587e-06, "loss": 0.0343, "step": 81670 }, { "grad_norm": 0.15667566657066345, "learning_rate": 8.89949267798037e-06, "loss": 0.0254, "step": 81680 }, { "grad_norm": 0.2738984227180481, "learning_rate": 8.890078863914542e-06, "loss": 0.0225, "step": 81690 }, { "grad_norm": 0.2795386016368866, "learning_rate": 8.880669545567894e-06, "loss": 0.0218, "step": 81700 }, { "grad_norm": 0.2274991273880005, "learning_rate": 8.871264723969381e-06, "loss": 0.0259, "step": 81710 }, { "grad_norm": 0.3555561304092407, "learning_rate": 8.86186440014754e-06, "loss": 0.0281, "step": 81720 }, { "grad_norm": 0.3427934944629669, "learning_rate": 8.852468575130351e-06, "loss": 0.0285, "step": 81730 }, { "grad_norm": 0.3527452051639557, "learning_rate": 8.843077249945347e-06, "loss": 0.0235, "step": 81740 }, { "grad_norm": 0.28948652744293213, "learning_rate": 8.833690425619523e-06, "loss": 0.0278, "step": 81750 }, { "grad_norm": 0.2654774487018585, "learning_rate": 8.82430810317942e-06, "loss": 0.0311, "step": 81760 }, { "grad_norm": 0.4329588711261749, "learning_rate": 8.814930283651074e-06, "loss": 0.0273, "step": 81770 }, { "grad_norm": 0.2586699426174164, "learning_rate": 8.805556968060048e-06, "loss": 0.0273, "step": 81780 }, { "grad_norm": 0.3534785509109497, "learning_rate": 8.796188157431362e-06, "loss": 0.029, "step": 81790 }, { "grad_norm": 0.7695794105529785, "learning_rate": 8.786823852789595e-06, "loss": 0.0277, "step": 81800 }, { "grad_norm": 0.7363278269767761, "learning_rate": 8.777464055158813e-06, "loss": 0.0282, "step": 81810 }, { "grad_norm": 0.2535410225391388, "learning_rate": 8.768108765562578e-06, "loss": 0.0321, "step": 81820 }, { "grad_norm": 0.27383729815483093, "learning_rate": 8.758757985023975e-06, "loss": 0.0278, "step": 81830 }, { "grad_norm": 0.739067018032074, "learning_rate": 8.749411714565598e-06, "loss": 0.0198, "step": 81840 }, { "grad_norm": 0.5316398739814758, "learning_rate": 8.74006995520954e-06, "loss": 0.0369, "step": 81850 }, { "grad_norm": 0.42868557572364807, "learning_rate": 8.730732707977397e-06, "loss": 0.0243, "step": 81860 }, { "grad_norm": 0.507622480392456, "learning_rate": 8.721399973890281e-06, "loss": 0.0279, "step": 81870 }, { "grad_norm": 0.3063526749610901, "learning_rate": 8.712071753968792e-06, "loss": 0.0287, "step": 81880 }, { "grad_norm": 0.6119656562805176, "learning_rate": 8.702748049233072e-06, "loss": 0.0264, "step": 81890 }, { "grad_norm": 0.4229621887207031, "learning_rate": 8.693428860702718e-06, "loss": 0.0277, "step": 81900 }, { "grad_norm": 0.3866230547428131, "learning_rate": 8.684114189396898e-06, "loss": 0.0265, "step": 81910 }, { "grad_norm": 0.22900719940662384, "learning_rate": 8.674804036334222e-06, "loss": 0.0243, "step": 81920 }, { "grad_norm": 0.33745381236076355, "learning_rate": 8.665498402532857e-06, "loss": 0.0192, "step": 81930 }, { "grad_norm": 0.6967089772224426, "learning_rate": 8.656197289010432e-06, "loss": 0.0305, "step": 81940 }, { "grad_norm": 0.3155810534954071, "learning_rate": 8.646900696784122e-06, "loss": 0.0189, "step": 81950 }, { "grad_norm": 0.4508722722530365, "learning_rate": 8.637608626870553e-06, "loss": 0.0239, "step": 81960 }, { "grad_norm": 0.3660464584827423, "learning_rate": 8.628321080285944e-06, "loss": 0.0223, "step": 81970 }, { "grad_norm": 0.749505341053009, "learning_rate": 8.619038058045925e-06, "loss": 0.0351, "step": 81980 }, { "grad_norm": 0.32596349716186523, "learning_rate": 8.609759561165704e-06, "loss": 0.0259, "step": 81990 }, { "grad_norm": 0.2571917474269867, "learning_rate": 8.600485590659934e-06, "loss": 0.0308, "step": 82000 }, { "grad_norm": 0.2954556941986084, "learning_rate": 8.591216147542813e-06, "loss": 0.0294, "step": 82010 }, { "grad_norm": 0.24028918147087097, "learning_rate": 8.581951232828045e-06, "loss": 0.0379, "step": 82020 }, { "grad_norm": 0.5079107284545898, "learning_rate": 8.572690847528825e-06, "loss": 0.0278, "step": 82030 }, { "grad_norm": 0.19379498064517975, "learning_rate": 8.563434992657838e-06, "loss": 0.0284, "step": 82040 }, { "grad_norm": 0.47793176770210266, "learning_rate": 8.554183669227307e-06, "loss": 0.0274, "step": 82050 }, { "grad_norm": 0.25083908438682556, "learning_rate": 8.544936878248937e-06, "loss": 0.0326, "step": 82060 }, { "grad_norm": 0.43568453192710876, "learning_rate": 8.535694620733942e-06, "loss": 0.0305, "step": 82070 }, { "grad_norm": 0.35855937004089355, "learning_rate": 8.526456897693053e-06, "loss": 0.0249, "step": 82080 }, { "grad_norm": 0.31827986240386963, "learning_rate": 8.517223710136456e-06, "loss": 0.03, "step": 82090 }, { "grad_norm": 0.6150932908058167, "learning_rate": 8.507995059073931e-06, "loss": 0.0306, "step": 82100 }, { "grad_norm": 0.4017561674118042, "learning_rate": 8.498770945514673e-06, "loss": 0.0318, "step": 82110 }, { "grad_norm": 0.2571374475955963, "learning_rate": 8.489551370467436e-06, "loss": 0.0243, "step": 82120 }, { "grad_norm": 0.3888149857521057, "learning_rate": 8.480336334940442e-06, "loss": 0.0337, "step": 82130 }, { "grad_norm": 0.32909709215164185, "learning_rate": 8.471125839941458e-06, "loss": 0.0242, "step": 82140 }, { "grad_norm": 0.32634007930755615, "learning_rate": 8.461919886477682e-06, "loss": 0.0281, "step": 82150 }, { "grad_norm": 0.5720276832580566, "learning_rate": 8.452718475555927e-06, "loss": 0.0238, "step": 82160 }, { "grad_norm": 0.29922789335250854, "learning_rate": 8.443521608182398e-06, "loss": 0.0235, "step": 82170 }, { "grad_norm": 0.23889875411987305, "learning_rate": 8.43432928536288e-06, "loss": 0.0324, "step": 82180 }, { "grad_norm": 0.41363704204559326, "learning_rate": 8.425141508102607e-06, "loss": 0.0286, "step": 82190 }, { "grad_norm": 0.1815686821937561, "learning_rate": 8.415958277406365e-06, "loss": 0.0235, "step": 82200 }, { "grad_norm": 0.3815115988254547, "learning_rate": 8.406779594278392e-06, "loss": 0.0291, "step": 82210 }, { "grad_norm": 0.4457797706127167, "learning_rate": 8.397605459722468e-06, "loss": 0.0233, "step": 82220 }, { "grad_norm": 0.39693012833595276, "learning_rate": 8.388435874741863e-06, "loss": 0.0301, "step": 82230 }, { "grad_norm": 0.3629454970359802, "learning_rate": 8.37927084033936e-06, "loss": 0.0262, "step": 82240 }, { "grad_norm": 0.6192207336425781, "learning_rate": 8.370110357517208e-06, "loss": 0.0314, "step": 82250 }, { "grad_norm": 0.5998253226280212, "learning_rate": 8.360954427277201e-06, "loss": 0.0346, "step": 82260 }, { "grad_norm": 0.13621632754802704, "learning_rate": 8.351803050620627e-06, "loss": 0.0255, "step": 82270 }, { "grad_norm": 0.3725055754184723, "learning_rate": 8.34265622854824e-06, "loss": 0.0253, "step": 82280 }, { "grad_norm": 0.37290170788764954, "learning_rate": 8.333513962060335e-06, "loss": 0.0408, "step": 82290 }, { "grad_norm": 0.4176991283893585, "learning_rate": 8.324376252156701e-06, "loss": 0.0212, "step": 82300 }, { "grad_norm": 0.35421696305274963, "learning_rate": 8.315243099836633e-06, "loss": 0.0227, "step": 82310 }, { "grad_norm": 0.35659971833229065, "learning_rate": 8.3061145060989e-06, "loss": 0.0412, "step": 82320 }, { "grad_norm": 0.25365984439849854, "learning_rate": 8.296990471941806e-06, "loss": 0.0311, "step": 82330 }, { "grad_norm": 0.6529493927955627, "learning_rate": 8.287870998363112e-06, "loss": 0.0282, "step": 82340 }, { "grad_norm": 0.3082449734210968, "learning_rate": 8.278756086360156e-06, "loss": 0.0264, "step": 82350 }, { "grad_norm": 0.5883233547210693, "learning_rate": 8.269645736929693e-06, "loss": 0.0218, "step": 82360 }, { "grad_norm": 0.3003804683685303, "learning_rate": 8.260539951068048e-06, "loss": 0.0244, "step": 82370 }, { "grad_norm": 0.412112832069397, "learning_rate": 8.251438729770983e-06, "loss": 0.0359, "step": 82380 }, { "grad_norm": 0.4514470398426056, "learning_rate": 8.242342074033826e-06, "loss": 0.0293, "step": 82390 }, { "grad_norm": 0.39474040269851685, "learning_rate": 8.233249984851344e-06, "loss": 0.0278, "step": 82400 }, { "grad_norm": 0.482444703578949, "learning_rate": 8.224162463217849e-06, "loss": 0.0291, "step": 82410 }, { "grad_norm": 0.5022172331809998, "learning_rate": 8.215079510127139e-06, "loss": 0.0246, "step": 82420 }, { "grad_norm": 0.6194384694099426, "learning_rate": 8.206001126572521e-06, "loss": 0.0356, "step": 82430 }, { "grad_norm": 0.4554058909416199, "learning_rate": 8.196927313546776e-06, "loss": 0.032, "step": 82440 }, { "grad_norm": 0.32157081365585327, "learning_rate": 8.18785807204222e-06, "loss": 0.0296, "step": 82450 }, { "grad_norm": 0.38356953859329224, "learning_rate": 8.178793403050627e-06, "loss": 0.023, "step": 82460 }, { "grad_norm": 0.4378214180469513, "learning_rate": 8.16973330756331e-06, "loss": 0.028, "step": 82470 }, { "grad_norm": 0.47388163208961487, "learning_rate": 8.16067778657107e-06, "loss": 0.0234, "step": 82480 }, { "grad_norm": 0.43806686997413635, "learning_rate": 8.151626841064214e-06, "loss": 0.0225, "step": 82490 }, { "grad_norm": 0.5298512578010559, "learning_rate": 8.142580472032518e-06, "loss": 0.0248, "step": 82500 }, { "grad_norm": 0.3430221974849701, "learning_rate": 8.133538680465285e-06, "loss": 0.0254, "step": 82510 }, { "grad_norm": 0.5313065052032471, "learning_rate": 8.124501467351336e-06, "loss": 0.0306, "step": 82520 }, { "grad_norm": 0.37380677461624146, "learning_rate": 8.11546883367893e-06, "loss": 0.0256, "step": 82530 }, { "grad_norm": 0.30806806683540344, "learning_rate": 8.106440780435882e-06, "loss": 0.0268, "step": 82540 }, { "grad_norm": 0.3655749559402466, "learning_rate": 8.097417308609485e-06, "loss": 0.0268, "step": 82550 }, { "grad_norm": 0.3396992087364197, "learning_rate": 8.088398419186544e-06, "loss": 0.028, "step": 82560 }, { "grad_norm": 0.579191267490387, "learning_rate": 8.079384113153321e-06, "loss": 0.0271, "step": 82570 }, { "grad_norm": 0.26025551557540894, "learning_rate": 8.070374391495644e-06, "loss": 0.0228, "step": 82580 }, { "grad_norm": 0.33992472290992737, "learning_rate": 8.061369255198764e-06, "loss": 0.0212, "step": 82590 }, { "grad_norm": 0.5015311241149902, "learning_rate": 8.052368705247493e-06, "loss": 0.0317, "step": 82600 }, { "grad_norm": 0.5795083045959473, "learning_rate": 8.043372742626114e-06, "loss": 0.0399, "step": 82610 }, { "grad_norm": 0.31676211953163147, "learning_rate": 8.034381368318422e-06, "loss": 0.0263, "step": 82620 }, { "grad_norm": 0.38785526156425476, "learning_rate": 8.025394583307677e-06, "loss": 0.0286, "step": 82630 }, { "grad_norm": 1.0850646495819092, "learning_rate": 8.016412388576683e-06, "loss": 0.038, "step": 82640 }, { "grad_norm": 0.34049689769744873, "learning_rate": 8.007434785107699e-06, "loss": 0.0231, "step": 82650 }, { "grad_norm": 0.37040629982948303, "learning_rate": 7.998461773882509e-06, "loss": 0.0304, "step": 82660 }, { "grad_norm": 0.2766658067703247, "learning_rate": 7.989493355882393e-06, "loss": 0.0281, "step": 82670 }, { "grad_norm": 0.3915557265281677, "learning_rate": 7.980529532088137e-06, "loss": 0.0253, "step": 82680 }, { "grad_norm": 0.4819720983505249, "learning_rate": 7.97157030347998e-06, "loss": 0.0275, "step": 82690 }, { "grad_norm": 0.41519391536712646, "learning_rate": 7.96261567103772e-06, "loss": 0.0255, "step": 82700 }, { "grad_norm": 0.3466413915157318, "learning_rate": 7.953665635740593e-06, "loss": 0.0231, "step": 82710 }, { "grad_norm": 0.36707258224487305, "learning_rate": 7.94472019856738e-06, "loss": 0.0213, "step": 82720 }, { "grad_norm": 0.6462419033050537, "learning_rate": 7.935779360496337e-06, "loss": 0.0269, "step": 82730 }, { "grad_norm": 0.3523681163787842, "learning_rate": 7.92684312250523e-06, "loss": 0.0321, "step": 82740 }, { "grad_norm": 0.30518588423728943, "learning_rate": 7.917911485571289e-06, "loss": 0.0272, "step": 82750 }, { "grad_norm": 0.6951737403869629, "learning_rate": 7.90898445067128e-06, "loss": 0.0282, "step": 82760 }, { "grad_norm": 0.398580402135849, "learning_rate": 7.900062018781462e-06, "loss": 0.0269, "step": 82770 }, { "grad_norm": 0.3811752498149872, "learning_rate": 7.891144190877553e-06, "loss": 0.0271, "step": 82780 }, { "grad_norm": 0.19071969389915466, "learning_rate": 7.882230967934806e-06, "loss": 0.0208, "step": 82790 }, { "grad_norm": 0.38604113459587097, "learning_rate": 7.873322350927959e-06, "loss": 0.0265, "step": 82800 }, { "grad_norm": 0.6946394443511963, "learning_rate": 7.864418340831253e-06, "loss": 0.0245, "step": 82810 }, { "grad_norm": 0.3993660807609558, "learning_rate": 7.855518938618395e-06, "loss": 0.042, "step": 82820 }, { "grad_norm": 0.27025601267814636, "learning_rate": 7.846624145262633e-06, "loss": 0.0276, "step": 82830 }, { "grad_norm": 0.3484328091144562, "learning_rate": 7.837733961736672e-06, "loss": 0.0303, "step": 82840 }, { "grad_norm": 0.2209881693124771, "learning_rate": 7.828848389012727e-06, "loss": 0.024, "step": 82850 }, { "grad_norm": 0.2806564271450043, "learning_rate": 7.819967428062524e-06, "loss": 0.0291, "step": 82860 }, { "grad_norm": 0.4730947017669678, "learning_rate": 7.811091079857274e-06, "loss": 0.0356, "step": 82870 }, { "grad_norm": 0.37531429529190063, "learning_rate": 7.802219345367667e-06, "loss": 0.0212, "step": 82880 }, { "grad_norm": 0.6099961996078491, "learning_rate": 7.793352225563916e-06, "loss": 0.0306, "step": 82890 }, { "grad_norm": 0.4178319573402405, "learning_rate": 7.7844897214157e-06, "loss": 0.0256, "step": 82900 }, { "grad_norm": 0.3271365463733673, "learning_rate": 7.775631833892228e-06, "loss": 0.0347, "step": 82910 }, { "grad_norm": 0.3185563087463379, "learning_rate": 7.76677856396215e-06, "loss": 0.0186, "step": 82920 }, { "grad_norm": 0.4223393201828003, "learning_rate": 7.757929912593697e-06, "loss": 0.0277, "step": 82930 }, { "grad_norm": 0.47446170449256897, "learning_rate": 7.749085880754509e-06, "loss": 0.0287, "step": 82940 }, { "grad_norm": 0.28975844383239746, "learning_rate": 7.740246469411777e-06, "loss": 0.0371, "step": 82950 }, { "grad_norm": 0.3300287425518036, "learning_rate": 7.731411679532142e-06, "loss": 0.0308, "step": 82960 }, { "grad_norm": 0.29634764790534973, "learning_rate": 7.722581512081779e-06, "loss": 0.0377, "step": 82970 }, { "grad_norm": 0.3880462646484375, "learning_rate": 7.713755968026331e-06, "loss": 0.024, "step": 82980 }, { "grad_norm": 0.5385858416557312, "learning_rate": 7.704935048330969e-06, "loss": 0.0295, "step": 82990 }, { "grad_norm": 0.47873246669769287, "learning_rate": 7.696118753960313e-06, "loss": 0.0283, "step": 83000 }, { "grad_norm": 0.24916791915893555, "learning_rate": 7.687307085878503e-06, "loss": 0.0265, "step": 83010 }, { "grad_norm": 0.2375405877828598, "learning_rate": 7.678500045049186e-06, "loss": 0.0215, "step": 83020 }, { "grad_norm": 0.6482518315315247, "learning_rate": 7.669697632435463e-06, "loss": 0.0272, "step": 83030 }, { "grad_norm": 0.6394987106323242, "learning_rate": 7.660899848999963e-06, "loss": 0.0254, "step": 83040 }, { "grad_norm": 0.3211475610733032, "learning_rate": 7.652106695704803e-06, "loss": 0.0288, "step": 83050 }, { "grad_norm": 0.6530542373657227, "learning_rate": 7.643318173511598e-06, "loss": 0.0255, "step": 83060 }, { "grad_norm": 0.3700374960899353, "learning_rate": 7.634534283381423e-06, "loss": 0.0275, "step": 83070 }, { "grad_norm": 0.27837812900543213, "learning_rate": 7.625755026274889e-06, "loss": 0.0226, "step": 83080 }, { "grad_norm": 0.5706931948661804, "learning_rate": 7.616980403152069e-06, "loss": 0.0323, "step": 83090 }, { "grad_norm": 0.6442668437957764, "learning_rate": 7.608210414972561e-06, "loss": 0.0282, "step": 83100 }, { "grad_norm": 0.3462331295013428, "learning_rate": 7.599445062695404e-06, "loss": 0.0265, "step": 83110 }, { "grad_norm": 0.49054792523384094, "learning_rate": 7.590684347279214e-06, "loss": 0.0271, "step": 83120 }, { "grad_norm": 0.17238514125347137, "learning_rate": 7.581928269682004e-06, "loss": 0.02, "step": 83130 }, { "grad_norm": 0.19450655579566956, "learning_rate": 7.573176830861361e-06, "loss": 0.0247, "step": 83140 }, { "grad_norm": 0.3485090732574463, "learning_rate": 7.5644300317743e-06, "loss": 0.0263, "step": 83150 }, { "grad_norm": 0.3783577084541321, "learning_rate": 7.555687873377376e-06, "loss": 0.0381, "step": 83160 }, { "grad_norm": 0.5783967971801758, "learning_rate": 7.546950356626592e-06, "loss": 0.0345, "step": 83170 }, { "grad_norm": 0.30318978428840637, "learning_rate": 7.538217482477516e-06, "loss": 0.0312, "step": 83180 }, { "grad_norm": 0.39180275797843933, "learning_rate": 7.529489251885119e-06, "loss": 0.0272, "step": 83190 }, { "grad_norm": 0.4835474193096161, "learning_rate": 7.520765665803941e-06, "loss": 0.0382, "step": 83200 }, { "grad_norm": 0.37672901153564453, "learning_rate": 7.512046725187949e-06, "loss": 0.0264, "step": 83210 }, { "grad_norm": 0.2683439552783966, "learning_rate": 7.503332430990645e-06, "loss": 0.0253, "step": 83220 }, { "grad_norm": 0.395114004611969, "learning_rate": 7.494622784165017e-06, "loss": 0.0302, "step": 83230 }, { "grad_norm": 0.26696184277534485, "learning_rate": 7.485917785663543e-06, "loss": 0.0308, "step": 83240 }, { "grad_norm": 0.36736801266670227, "learning_rate": 7.477217436438172e-06, "loss": 0.0203, "step": 83250 }, { "grad_norm": 0.2520473301410675, "learning_rate": 7.46852173744037e-06, "loss": 0.0183, "step": 83260 }, { "grad_norm": 0.39587652683258057, "learning_rate": 7.4598306896211004e-06, "loss": 0.0377, "step": 83270 }, { "grad_norm": 0.21055006980895996, "learning_rate": 7.451144293930773e-06, "loss": 0.0233, "step": 83280 }, { "grad_norm": 0.3197919726371765, "learning_rate": 7.4424625513193455e-06, "loss": 0.0292, "step": 83290 }, { "grad_norm": 0.26764190196990967, "learning_rate": 7.433785462736209e-06, "loss": 0.0277, "step": 83300 }, { "grad_norm": 0.2676788866519928, "learning_rate": 7.425113029130315e-06, "loss": 0.0215, "step": 83310 }, { "grad_norm": 0.18478791415691376, "learning_rate": 7.416445251450044e-06, "loss": 0.0303, "step": 83320 }, { "grad_norm": 1.0728838443756104, "learning_rate": 7.4077821306433035e-06, "loss": 0.0319, "step": 83330 }, { "grad_norm": 0.20326562225818634, "learning_rate": 7.399123667657465e-06, "loss": 0.0217, "step": 83340 }, { "grad_norm": 0.43390387296676636, "learning_rate": 7.390469863439426e-06, "loss": 0.022, "step": 83350 }, { "grad_norm": 0.43382373452186584, "learning_rate": 7.381820718935512e-06, "loss": 0.0292, "step": 83360 }, { "grad_norm": 0.34720221161842346, "learning_rate": 7.373176235091633e-06, "loss": 0.0274, "step": 83370 }, { "grad_norm": 0.37820354104042053, "learning_rate": 7.364536412853101e-06, "loss": 0.0253, "step": 83380 }, { "grad_norm": 0.910317063331604, "learning_rate": 7.355901253164776e-06, "loss": 0.027, "step": 83390 }, { "grad_norm": 0.252390056848526, "learning_rate": 7.34727075697097e-06, "loss": 0.018, "step": 83400 }, { "grad_norm": 0.252666711807251, "learning_rate": 7.3386449252155095e-06, "loss": 0.0244, "step": 83410 }, { "grad_norm": 0.6453959941864014, "learning_rate": 7.330023758841692e-06, "loss": 0.0344, "step": 83420 }, { "grad_norm": 0.7301104068756104, "learning_rate": 7.321407258792323e-06, "loss": 0.0199, "step": 83430 }, { "grad_norm": 0.3190467953681946, "learning_rate": 7.312795426009694e-06, "loss": 0.018, "step": 83440 }, { "grad_norm": 0.6802884340286255, "learning_rate": 7.304188261435585e-06, "loss": 0.0254, "step": 83450 }, { "grad_norm": 0.1726628839969635, "learning_rate": 7.295585766011242e-06, "loss": 0.0225, "step": 83460 }, { "grad_norm": 0.41292521357536316, "learning_rate": 7.286987940677436e-06, "loss": 0.0323, "step": 83470 }, { "grad_norm": 0.45868608355522156, "learning_rate": 7.278394786374426e-06, "loss": 0.0252, "step": 83480 }, { "grad_norm": 0.18585903942584991, "learning_rate": 7.269806304041915e-06, "loss": 0.0296, "step": 83490 }, { "grad_norm": 0.4865724742412567, "learning_rate": 7.261222494619147e-06, "loss": 0.0236, "step": 83500 }, { "grad_norm": 0.5343415141105652, "learning_rate": 7.252643359044825e-06, "loss": 0.0301, "step": 83510 }, { "grad_norm": 0.25694093108177185, "learning_rate": 7.244068898257172e-06, "loss": 0.0207, "step": 83520 }, { "grad_norm": 0.27466195821762085, "learning_rate": 7.2354991131938495e-06, "loss": 0.0255, "step": 83530 }, { "grad_norm": 0.41276898980140686, "learning_rate": 7.2269340047920574e-06, "loss": 0.0249, "step": 83540 }, { "grad_norm": 0.7588006854057312, "learning_rate": 7.218373573988435e-06, "loss": 0.0243, "step": 83550 }, { "grad_norm": 0.41623085737228394, "learning_rate": 7.209817821719178e-06, "loss": 0.0353, "step": 83560 }, { "grad_norm": 0.5738614201545715, "learning_rate": 7.201266748919894e-06, "loss": 0.0329, "step": 83570 }, { "grad_norm": 0.3220290243625641, "learning_rate": 7.19272035652575e-06, "loss": 0.0227, "step": 83580 }, { "grad_norm": 0.2771970331668854, "learning_rate": 7.184178645471334e-06, "loss": 0.0264, "step": 83590 }, { "grad_norm": 0.30322742462158203, "learning_rate": 7.175641616690775e-06, "loss": 0.0254, "step": 83600 }, { "grad_norm": 0.45228445529937744, "learning_rate": 7.167109271117661e-06, "loss": 0.0265, "step": 83610 }, { "grad_norm": 0.4528460204601288, "learning_rate": 7.158581609685072e-06, "loss": 0.03, "step": 83620 }, { "grad_norm": 0.26976943016052246, "learning_rate": 7.150058633325591e-06, "loss": 0.0238, "step": 83630 }, { "grad_norm": 0.3359675705432892, "learning_rate": 7.141540342971287e-06, "loss": 0.0342, "step": 83640 }, { "grad_norm": 0.23579354584217072, "learning_rate": 7.133026739553677e-06, "loss": 0.0225, "step": 83650 }, { "grad_norm": 0.2761346995830536, "learning_rate": 7.124517824003829e-06, "loss": 0.0324, "step": 83660 }, { "grad_norm": 0.31248003244400024, "learning_rate": 7.116013597252236e-06, "loss": 0.026, "step": 83670 }, { "grad_norm": 0.36424627900123596, "learning_rate": 7.107514060228921e-06, "loss": 0.0335, "step": 83680 }, { "grad_norm": 0.4160434305667877, "learning_rate": 7.099019213863378e-06, "loss": 0.0332, "step": 83690 }, { "grad_norm": 0.3982929587364197, "learning_rate": 7.090529059084611e-06, "loss": 0.0213, "step": 83700 }, { "grad_norm": 0.25668907165527344, "learning_rate": 7.082043596821058e-06, "loss": 0.0197, "step": 83710 }, { "grad_norm": 0.3650873005390167, "learning_rate": 7.07356282800069e-06, "loss": 0.029, "step": 83720 }, { "grad_norm": 0.4869849681854248, "learning_rate": 7.0650867535509645e-06, "loss": 0.0194, "step": 83730 }, { "grad_norm": 0.36785146594047546, "learning_rate": 7.0566153743987895e-06, "loss": 0.0245, "step": 83740 }, { "grad_norm": 0.24732431769371033, "learning_rate": 7.04814869147059e-06, "loss": 0.0359, "step": 83750 }, { "grad_norm": 0.3258189260959625, "learning_rate": 7.0396867056922725e-06, "loss": 0.0246, "step": 83760 }, { "grad_norm": 0.27337655425071716, "learning_rate": 7.031229417989244e-06, "loss": 0.0267, "step": 83770 }, { "grad_norm": 0.3102266490459442, "learning_rate": 7.022776829286348e-06, "loss": 0.0295, "step": 83780 }, { "grad_norm": 0.4058206081390381, "learning_rate": 7.014328940507975e-06, "loss": 0.0295, "step": 83790 }, { "grad_norm": 0.27000948786735535, "learning_rate": 7.005885752577945e-06, "loss": 0.0287, "step": 83800 }, { "grad_norm": 0.44897186756134033, "learning_rate": 6.997447266419615e-06, "loss": 0.0254, "step": 83810 }, { "grad_norm": 0.33049914240837097, "learning_rate": 6.989013482955797e-06, "loss": 0.0292, "step": 83820 }, { "grad_norm": 0.5102536678314209, "learning_rate": 6.980584403108808e-06, "loss": 0.0288, "step": 83830 }, { "grad_norm": 0.37970978021621704, "learning_rate": 6.972160027800417e-06, "loss": 0.0264, "step": 83840 }, { "grad_norm": 0.44333699345588684, "learning_rate": 6.963740357951931e-06, "loss": 0.0296, "step": 83850 }, { "grad_norm": 0.7072513699531555, "learning_rate": 6.95532539448408e-06, "loss": 0.0299, "step": 83860 }, { "grad_norm": 0.4421795904636383, "learning_rate": 6.946915138317128e-06, "loss": 0.0302, "step": 83870 }, { "grad_norm": 0.1420988142490387, "learning_rate": 6.9385095903708065e-06, "loss": 0.0251, "step": 83880 }, { "grad_norm": 0.3586302697658539, "learning_rate": 6.930108751564346e-06, "loss": 0.0279, "step": 83890 }, { "grad_norm": 0.3740071952342987, "learning_rate": 6.921712622816428e-06, "loss": 0.027, "step": 83900 }, { "grad_norm": 0.398154616355896, "learning_rate": 6.913321205045259e-06, "loss": 0.0399, "step": 83910 }, { "grad_norm": 0.3691486716270447, "learning_rate": 6.904934499168491e-06, "loss": 0.0207, "step": 83920 }, { "grad_norm": 0.2415822297334671, "learning_rate": 6.896552506103299e-06, "loss": 0.0292, "step": 83930 }, { "grad_norm": 0.41694480180740356, "learning_rate": 6.888175226766313e-06, "loss": 0.0434, "step": 83940 }, { "grad_norm": 0.21298819780349731, "learning_rate": 6.879802662073686e-06, "loss": 0.0275, "step": 83950 }, { "grad_norm": 0.382038414478302, "learning_rate": 6.871434812940991e-06, "loss": 0.0411, "step": 83960 }, { "grad_norm": 0.35824739933013916, "learning_rate": 6.863071680283345e-06, "loss": 0.0205, "step": 83970 }, { "grad_norm": 1.1235896348953247, "learning_rate": 6.854713265015339e-06, "loss": 0.0509, "step": 83980 }, { "grad_norm": 0.5251803994178772, "learning_rate": 6.8463595680510025e-06, "loss": 0.0303, "step": 83990 }, { "grad_norm": 0.4772549867630005, "learning_rate": 6.838010590303911e-06, "loss": 0.0379, "step": 84000 }, { "grad_norm": 0.5172386169433594, "learning_rate": 6.829666332687085e-06, "loss": 0.0252, "step": 84010 }, { "grad_norm": 0.44290003180503845, "learning_rate": 6.821326796113059e-06, "loss": 0.0254, "step": 84020 }, { "grad_norm": 0.4011157155036926, "learning_rate": 6.812991981493799e-06, "loss": 0.0254, "step": 84030 }, { "grad_norm": 0.4300481975078583, "learning_rate": 6.804661889740821e-06, "loss": 0.0348, "step": 84040 }, { "grad_norm": 0.4477279484272003, "learning_rate": 6.7963365217650625e-06, "loss": 0.031, "step": 84050 }, { "grad_norm": 0.37477290630340576, "learning_rate": 6.788015878476983e-06, "loss": 0.0281, "step": 84060 }, { "grad_norm": 0.5249582529067993, "learning_rate": 6.779699960786523e-06, "loss": 0.031, "step": 84070 }, { "grad_norm": 0.224956676363945, "learning_rate": 6.771388769603104e-06, "loss": 0.0231, "step": 84080 }, { "grad_norm": 0.3249043822288513, "learning_rate": 6.763082305835605e-06, "loss": 0.0468, "step": 84090 }, { "grad_norm": 0.6808210611343384, "learning_rate": 6.754780570392433e-06, "loss": 0.0251, "step": 84100 }, { "grad_norm": 0.4324050843715668, "learning_rate": 6.7464835641814215e-06, "loss": 0.0247, "step": 84110 }, { "grad_norm": 0.14732739329338074, "learning_rate": 6.7381912881099515e-06, "loss": 0.0254, "step": 84120 }, { "grad_norm": 0.6675639152526855, "learning_rate": 6.72990374308482e-06, "loss": 0.03, "step": 84130 }, { "grad_norm": 0.5725083351135254, "learning_rate": 6.721620930012379e-06, "loss": 0.0266, "step": 84140 }, { "grad_norm": 0.3240968585014343, "learning_rate": 6.713342849798393e-06, "loss": 0.0263, "step": 84150 }, { "grad_norm": 1.2618539333343506, "learning_rate": 6.705069503348166e-06, "loss": 0.0306, "step": 84160 }, { "grad_norm": 0.33675649762153625, "learning_rate": 6.696800891566435e-06, "loss": 0.0257, "step": 84170 }, { "grad_norm": 0.21495382487773895, "learning_rate": 6.688537015357449e-06, "loss": 0.028, "step": 84180 }, { "grad_norm": 0.5223483443260193, "learning_rate": 6.680277875624941e-06, "loss": 0.0259, "step": 84190 }, { "grad_norm": 0.4722476303577423, "learning_rate": 6.672023473272121e-06, "loss": 0.0292, "step": 84200 }, { "grad_norm": 0.5380601286888123, "learning_rate": 6.663773809201667e-06, "loss": 0.0379, "step": 84210 }, { "grad_norm": 0.29756543040275574, "learning_rate": 6.655528884315748e-06, "loss": 0.0245, "step": 84220 }, { "grad_norm": 0.33038273453712463, "learning_rate": 6.647288699516036e-06, "loss": 0.0331, "step": 84230 }, { "grad_norm": 0.4172345697879791, "learning_rate": 6.639053255703642e-06, "loss": 0.0347, "step": 84240 }, { "grad_norm": 0.3771533966064453, "learning_rate": 6.6308225537791925e-06, "loss": 0.0271, "step": 84250 }, { "grad_norm": 0.29420891404151917, "learning_rate": 6.622596594642777e-06, "loss": 0.024, "step": 84260 }, { "grad_norm": 0.6345682144165039, "learning_rate": 6.614375379193999e-06, "loss": 0.023, "step": 84270 }, { "grad_norm": 0.5295936465263367, "learning_rate": 6.606158908331889e-06, "loss": 0.0242, "step": 84280 }, { "grad_norm": 0.46713462471961975, "learning_rate": 6.597947182955005e-06, "loss": 0.0312, "step": 84290 }, { "grad_norm": 0.2660883963108063, "learning_rate": 6.589740203961359e-06, "loss": 0.0336, "step": 84300 }, { "grad_norm": 0.6186065077781677, "learning_rate": 6.581537972248464e-06, "loss": 0.0337, "step": 84310 }, { "grad_norm": 0.7439056038856506, "learning_rate": 6.573340488713275e-06, "loss": 0.0286, "step": 84320 }, { "grad_norm": 0.21235042810440063, "learning_rate": 6.565147754252304e-06, "loss": 0.0313, "step": 84330 }, { "grad_norm": 0.41232526302337646, "learning_rate": 6.556959769761462e-06, "loss": 0.0332, "step": 84340 }, { "grad_norm": 0.627074658870697, "learning_rate": 6.548776536136192e-06, "loss": 0.0266, "step": 84350 }, { "grad_norm": 0.6872920989990234, "learning_rate": 6.54059805427138e-06, "loss": 0.027, "step": 84360 }, { "grad_norm": 0.2999778091907501, "learning_rate": 6.532424325061437e-06, "loss": 0.0345, "step": 84370 }, { "grad_norm": 0.33289581537246704, "learning_rate": 6.5242553494001935e-06, "loss": 0.0272, "step": 84380 }, { "grad_norm": 0.28468507528305054, "learning_rate": 6.516091128181046e-06, "loss": 0.0255, "step": 84390 }, { "grad_norm": 0.42476633191108704, "learning_rate": 6.507931662296785e-06, "loss": 0.0282, "step": 84400 }, { "grad_norm": 0.46910426020622253, "learning_rate": 6.4997769526397354e-06, "loss": 0.0254, "step": 84410 }, { "grad_norm": 0.4107767641544342, "learning_rate": 6.491627000101663e-06, "loss": 0.0327, "step": 84420 }, { "grad_norm": 0.47350287437438965, "learning_rate": 6.4834818055738545e-06, "loss": 0.0349, "step": 84430 }, { "grad_norm": 0.3820982575416565, "learning_rate": 6.4753413699470465e-06, "loss": 0.0197, "step": 84440 }, { "grad_norm": 0.5501038432121277, "learning_rate": 6.467205694111478e-06, "loss": 0.024, "step": 84450 }, { "grad_norm": 0.3305761516094208, "learning_rate": 6.459074778956836e-06, "loss": 0.0229, "step": 84460 }, { "grad_norm": 0.21515746414661407, "learning_rate": 6.450948625372311e-06, "loss": 0.0232, "step": 84470 }, { "grad_norm": 0.5712670087814331, "learning_rate": 6.442827234246585e-06, "loss": 0.0247, "step": 84480 }, { "grad_norm": 0.41581133008003235, "learning_rate": 6.43471060646777e-06, "loss": 0.0245, "step": 84490 }, { "grad_norm": 0.1692933738231659, "learning_rate": 6.426598742923512e-06, "loss": 0.0235, "step": 84500 }, { "grad_norm": 0.9126815795898438, "learning_rate": 6.418491644500885e-06, "loss": 0.0336, "step": 84510 }, { "grad_norm": 0.6092769503593445, "learning_rate": 6.410389312086512e-06, "loss": 0.0291, "step": 84520 }, { "grad_norm": 0.6932017803192139, "learning_rate": 6.4022917465664135e-06, "loss": 0.032, "step": 84530 }, { "grad_norm": 0.48476213216781616, "learning_rate": 6.3941989488261575e-06, "loss": 0.0284, "step": 84540 }, { "grad_norm": 0.29908642172813416, "learning_rate": 6.386110919750726e-06, "loss": 0.0278, "step": 84550 }, { "grad_norm": 0.22436700761318207, "learning_rate": 6.378027660224651e-06, "loss": 0.0226, "step": 84560 }, { "grad_norm": 0.1793423593044281, "learning_rate": 6.369949171131856e-06, "loss": 0.0202, "step": 84570 }, { "grad_norm": 0.4098973870277405, "learning_rate": 6.361875453355848e-06, "loss": 0.0225, "step": 84580 }, { "grad_norm": 0.4027869403362274, "learning_rate": 6.353806507779525e-06, "loss": 0.0228, "step": 84590 }, { "grad_norm": 0.4321717619895935, "learning_rate": 6.3457423352853044e-06, "loss": 0.0272, "step": 84600 }, { "grad_norm": 0.36836373805999756, "learning_rate": 6.337682936755062e-06, "loss": 0.031, "step": 84610 }, { "grad_norm": 0.2780262529850006, "learning_rate": 6.329628313070179e-06, "loss": 0.034, "step": 84620 }, { "grad_norm": 0.403402179479599, "learning_rate": 6.321578465111477e-06, "loss": 0.0292, "step": 84630 }, { "grad_norm": 0.5595877170562744, "learning_rate": 6.31353339375928e-06, "loss": 0.0262, "step": 84640 }, { "grad_norm": 0.41595977544784546, "learning_rate": 6.305493099893395e-06, "loss": 0.0214, "step": 84650 }, { "grad_norm": 0.42033129930496216, "learning_rate": 6.297457584393096e-06, "loss": 0.0306, "step": 84660 }, { "grad_norm": 0.20755237340927124, "learning_rate": 6.289426848137126e-06, "loss": 0.032, "step": 84670 }, { "grad_norm": 0.39570382237434387, "learning_rate": 6.281400892003713e-06, "loss": 0.0235, "step": 84680 }, { "grad_norm": 0.3365909159183502, "learning_rate": 6.273379716870581e-06, "loss": 0.024, "step": 84690 }, { "grad_norm": 0.6908467411994934, "learning_rate": 6.265363323614887e-06, "loss": 0.0296, "step": 84700 }, { "grad_norm": 0.688447892665863, "learning_rate": 6.257351713113308e-06, "loss": 0.0219, "step": 84710 }, { "grad_norm": 0.4034862518310547, "learning_rate": 6.249344886241981e-06, "loss": 0.0298, "step": 84720 }, { "grad_norm": 0.29370713233947754, "learning_rate": 6.2413428438765295e-06, "loss": 0.021, "step": 84730 }, { "grad_norm": 0.315935879945755, "learning_rate": 6.233345586892026e-06, "loss": 0.0222, "step": 84740 }, { "grad_norm": 0.25859662890434265, "learning_rate": 6.225353116163052e-06, "loss": 0.0246, "step": 84750 }, { "grad_norm": 0.33039259910583496, "learning_rate": 6.217365432563627e-06, "loss": 0.0278, "step": 84760 }, { "grad_norm": 0.3067808151245117, "learning_rate": 6.209382536967318e-06, "loss": 0.0221, "step": 84770 }, { "grad_norm": 0.33995434641838074, "learning_rate": 6.201404430247082e-06, "loss": 0.0295, "step": 84780 }, { "grad_norm": 0.43239906430244446, "learning_rate": 6.1934311132754165e-06, "loss": 0.0227, "step": 84790 }, { "grad_norm": 0.9103959202766418, "learning_rate": 6.185462586924257e-06, "loss": 0.0282, "step": 84800 }, { "grad_norm": 0.36024948954582214, "learning_rate": 6.177498852065039e-06, "loss": 0.0263, "step": 84810 }, { "grad_norm": 0.3004308044910431, "learning_rate": 6.169539909568655e-06, "loss": 0.0281, "step": 84820 }, { "grad_norm": 0.4246000051498413, "learning_rate": 6.161585760305488e-06, "loss": 0.0345, "step": 84830 }, { "grad_norm": 0.33504652976989746, "learning_rate": 6.153636405145391e-06, "loss": 0.0243, "step": 84840 }, { "grad_norm": 0.4538115859031677, "learning_rate": 6.145691844957702e-06, "loss": 0.0264, "step": 84850 }, { "grad_norm": 0.39879798889160156, "learning_rate": 6.1377520806112145e-06, "loss": 0.0248, "step": 84860 }, { "grad_norm": 0.283104807138443, "learning_rate": 6.129817112974218e-06, "loss": 0.0227, "step": 84870 }, { "grad_norm": 0.40974095463752747, "learning_rate": 6.121886942914457e-06, "loss": 0.0233, "step": 84880 }, { "grad_norm": 0.4162563681602478, "learning_rate": 6.11396157129917e-06, "loss": 0.0257, "step": 84890 }, { "grad_norm": 0.501423716545105, "learning_rate": 6.106040998995066e-06, "loss": 0.0326, "step": 84900 }, { "grad_norm": 0.343872994184494, "learning_rate": 6.098125226868329e-06, "loss": 0.03, "step": 84910 }, { "grad_norm": 0.33816564083099365, "learning_rate": 6.090214255784604e-06, "loss": 0.0287, "step": 84920 }, { "grad_norm": 0.4311319589614868, "learning_rate": 6.0823080866090346e-06, "loss": 0.0357, "step": 84930 }, { "grad_norm": 0.515042781829834, "learning_rate": 6.074406720206233e-06, "loss": 0.0262, "step": 84940 }, { "grad_norm": 0.3998149335384369, "learning_rate": 6.066510157440259e-06, "loss": 0.0238, "step": 84950 }, { "grad_norm": 0.42011919617652893, "learning_rate": 6.058618399174676e-06, "loss": 0.0325, "step": 84960 }, { "grad_norm": 0.37420445680618286, "learning_rate": 6.050731446272528e-06, "loss": 0.0393, "step": 84970 }, { "grad_norm": 0.40938663482666016, "learning_rate": 6.042849299596321e-06, "loss": 0.0279, "step": 84980 }, { "grad_norm": 0.40807682275772095, "learning_rate": 6.034971960008013e-06, "loss": 0.0293, "step": 84990 }, { "grad_norm": 0.7204404473304749, "learning_rate": 6.027099428369082e-06, "loss": 0.0256, "step": 85000 }, { "grad_norm": 0.3902572989463806, "learning_rate": 6.019231705540435e-06, "loss": 0.0278, "step": 85010 }, { "grad_norm": 1.1529868841171265, "learning_rate": 6.011368792382488e-06, "loss": 0.0341, "step": 85020 }, { "grad_norm": 0.2960279583930969, "learning_rate": 6.00351068975511e-06, "loss": 0.0282, "step": 85030 }, { "grad_norm": 1.1002967357635498, "learning_rate": 5.9956573985176674e-06, "loss": 0.0355, "step": 85040 }, { "grad_norm": 0.3805321753025055, "learning_rate": 5.987808919528959e-06, "loss": 0.0359, "step": 85050 }, { "grad_norm": 0.5025382041931152, "learning_rate": 5.979965253647307e-06, "loss": 0.0395, "step": 85060 }, { "grad_norm": 0.4198302626609802, "learning_rate": 5.9721264017304655e-06, "loss": 0.0331, "step": 85070 }, { "grad_norm": 0.3737753927707672, "learning_rate": 5.964292364635682e-06, "loss": 0.0304, "step": 85080 }, { "grad_norm": 0.2845873236656189, "learning_rate": 5.956463143219682e-06, "loss": 0.0284, "step": 85090 }, { "grad_norm": 0.7252874374389648, "learning_rate": 5.948638738338663e-06, "loss": 0.0302, "step": 85100 }, { "grad_norm": 0.40414950251579285, "learning_rate": 5.940819150848271e-06, "loss": 0.0264, "step": 85110 }, { "grad_norm": 0.26273348927497864, "learning_rate": 5.933004381603663e-06, "loss": 0.0214, "step": 85120 }, { "grad_norm": 0.3484521508216858, "learning_rate": 5.92519443145943e-06, "loss": 0.021, "step": 85130 }, { "grad_norm": 0.22910498082637787, "learning_rate": 5.917389301269671e-06, "loss": 0.0189, "step": 85140 }, { "grad_norm": 0.39961332082748413, "learning_rate": 5.9095889918879435e-06, "loss": 0.027, "step": 85150 }, { "grad_norm": 0.5814225673675537, "learning_rate": 5.901793504167286e-06, "loss": 0.0206, "step": 85160 }, { "grad_norm": 1.0723378658294678, "learning_rate": 5.894002838960172e-06, "loss": 0.0283, "step": 85170 }, { "grad_norm": 0.5171754956245422, "learning_rate": 5.886216997118604e-06, "loss": 0.0237, "step": 85180 }, { "grad_norm": 0.2719458341598511, "learning_rate": 5.878435979494023e-06, "loss": 0.0285, "step": 85190 }, { "grad_norm": 0.37491336464881897, "learning_rate": 5.870659786937344e-06, "loss": 0.03, "step": 85200 }, { "grad_norm": 0.44238224625587463, "learning_rate": 5.862888420298962e-06, "loss": 0.0229, "step": 85210 }, { "grad_norm": 0.5088148713111877, "learning_rate": 5.855121880428743e-06, "loss": 0.0373, "step": 85220 }, { "grad_norm": 0.3360508382320404, "learning_rate": 5.84736016817603e-06, "loss": 0.0266, "step": 85230 }, { "grad_norm": 0.26823997497558594, "learning_rate": 5.839603284389622e-06, "loss": 0.0199, "step": 85240 }, { "grad_norm": 0.32716768980026245, "learning_rate": 5.831851229917812e-06, "loss": 0.0238, "step": 85250 }, { "grad_norm": 0.4198160767555237, "learning_rate": 5.824104005608338e-06, "loss": 0.0275, "step": 85260 }, { "grad_norm": 0.35206398367881775, "learning_rate": 5.81636161230843e-06, "loss": 0.022, "step": 85270 }, { "grad_norm": 0.38154083490371704, "learning_rate": 5.808624050864791e-06, "loss": 0.0268, "step": 85280 }, { "grad_norm": 0.3224077522754669, "learning_rate": 5.800891322123592e-06, "loss": 0.0236, "step": 85290 }, { "grad_norm": 0.5941565036773682, "learning_rate": 5.793163426930454e-06, "loss": 0.0237, "step": 85300 }, { "grad_norm": 0.32106950879096985, "learning_rate": 5.785440366130512e-06, "loss": 0.0263, "step": 85310 }, { "grad_norm": 0.2529623210430145, "learning_rate": 5.777722140568326e-06, "loss": 0.0284, "step": 85320 }, { "grad_norm": 0.3336092233657837, "learning_rate": 5.7700087510879685e-06, "loss": 0.0242, "step": 85330 }, { "grad_norm": 0.5989705920219421, "learning_rate": 5.762300198532933e-06, "loss": 0.0259, "step": 85340 }, { "grad_norm": 0.6482734680175781, "learning_rate": 5.7545964837462606e-06, "loss": 0.0289, "step": 85350 }, { "grad_norm": 0.2699604630470276, "learning_rate": 5.746897607570384e-06, "loss": 0.0269, "step": 85360 }, { "grad_norm": 0.25095734000205994, "learning_rate": 5.7392035708472626e-06, "loss": 0.0192, "step": 85370 }, { "grad_norm": 0.3934548795223236, "learning_rate": 5.731514374418278e-06, "loss": 0.0267, "step": 85380 }, { "grad_norm": 0.49664899706840515, "learning_rate": 5.7238300191243325e-06, "loss": 0.0344, "step": 85390 }, { "grad_norm": 0.43007615208625793, "learning_rate": 5.716150505805762e-06, "loss": 0.0224, "step": 85400 }, { "grad_norm": 0.4496994614601135, "learning_rate": 5.708475835302407e-06, "loss": 0.0243, "step": 85410 }, { "grad_norm": 0.3944118618965149, "learning_rate": 5.70080600845353e-06, "loss": 0.0338, "step": 85420 }, { "grad_norm": 0.24771133065223694, "learning_rate": 5.693141026097909e-06, "loss": 0.0261, "step": 85430 }, { "grad_norm": 0.6012254953384399, "learning_rate": 5.6854808890737745e-06, "loss": 0.0287, "step": 85440 }, { "grad_norm": 0.286996066570282, "learning_rate": 5.6778255982188225e-06, "loss": 0.0266, "step": 85450 }, { "grad_norm": 0.44975095987319946, "learning_rate": 5.670175154370217e-06, "loss": 0.026, "step": 85460 }, { "grad_norm": 0.3123537302017212, "learning_rate": 5.662529558364616e-06, "loss": 0.0259, "step": 85470 }, { "grad_norm": 0.7885947227478027, "learning_rate": 5.654888811038128e-06, "loss": 0.0272, "step": 85480 }, { "grad_norm": 0.4084224998950958, "learning_rate": 5.647252913226314e-06, "loss": 0.0237, "step": 85490 }, { "grad_norm": 0.3576928973197937, "learning_rate": 5.639621865764255e-06, "loss": 0.0272, "step": 85500 }, { "grad_norm": 0.4673376679420471, "learning_rate": 5.63199566948644e-06, "loss": 0.0241, "step": 85510 }, { "grad_norm": 0.3501766622066498, "learning_rate": 5.6243743252268835e-06, "loss": 0.0292, "step": 85520 }, { "grad_norm": 0.40189921855926514, "learning_rate": 5.616757833819009e-06, "loss": 0.0249, "step": 85530 }, { "grad_norm": 0.4015659987926483, "learning_rate": 5.609146196095794e-06, "loss": 0.0228, "step": 85540 }, { "grad_norm": 0.42090165615081787, "learning_rate": 5.6015394128896005e-06, "loss": 0.023, "step": 85550 }, { "grad_norm": 0.37572181224823, "learning_rate": 5.593937485032308e-06, "loss": 0.0229, "step": 85560 }, { "grad_norm": 0.3236808478832245, "learning_rate": 5.586340413355246e-06, "loss": 0.0249, "step": 85570 }, { "grad_norm": 0.2748909294605255, "learning_rate": 5.578748198689227e-06, "loss": 0.0248, "step": 85580 }, { "grad_norm": 0.4719879627227783, "learning_rate": 5.571160841864504e-06, "loss": 0.0274, "step": 85590 }, { "grad_norm": 0.47842153906822205, "learning_rate": 5.563578343710846e-06, "loss": 0.0215, "step": 85600 }, { "grad_norm": 0.3680194020271301, "learning_rate": 5.556000705057446e-06, "loss": 0.0214, "step": 85610 }, { "grad_norm": 0.26982900500297546, "learning_rate": 5.548427926732996e-06, "loss": 0.0304, "step": 85620 }, { "grad_norm": 0.6970932483673096, "learning_rate": 5.540860009565629e-06, "loss": 0.0178, "step": 85630 }, { "grad_norm": 0.28486713767051697, "learning_rate": 5.533296954382966e-06, "loss": 0.0238, "step": 85640 }, { "grad_norm": 0.1788114607334137, "learning_rate": 5.525738762012095e-06, "loss": 0.0276, "step": 85650 }, { "grad_norm": 0.46963804960250854, "learning_rate": 5.518185433279577e-06, "loss": 0.0318, "step": 85660 }, { "grad_norm": 0.5979738235473633, "learning_rate": 5.510636969011407e-06, "loss": 0.036, "step": 85670 }, { "grad_norm": 0.16402462124824524, "learning_rate": 5.503093370033096e-06, "loss": 0.0217, "step": 85680 }, { "grad_norm": 0.6420791149139404, "learning_rate": 5.495554637169603e-06, "loss": 0.0211, "step": 85690 }, { "grad_norm": 0.3198685944080353, "learning_rate": 5.488020771245328e-06, "loss": 0.024, "step": 85700 }, { "grad_norm": 0.2958528399467468, "learning_rate": 5.480491773084195e-06, "loss": 0.0269, "step": 85710 }, { "grad_norm": 0.351046621799469, "learning_rate": 5.472967643509525e-06, "loss": 0.0264, "step": 85720 }, { "grad_norm": 0.45029416680336, "learning_rate": 5.46544838334419e-06, "loss": 0.026, "step": 85730 }, { "grad_norm": 0.27972570061683655, "learning_rate": 5.4579339934104525e-06, "loss": 0.0285, "step": 85740 }, { "grad_norm": 0.3450999855995178, "learning_rate": 5.450424474530091e-06, "loss": 0.039, "step": 85750 }, { "grad_norm": 0.36209529638290405, "learning_rate": 5.442919827524323e-06, "loss": 0.0268, "step": 85760 }, { "grad_norm": 0.3501811921596527, "learning_rate": 5.435420053213863e-06, "loss": 0.0275, "step": 85770 }, { "grad_norm": 0.31292104721069336, "learning_rate": 5.427925152418845e-06, "loss": 0.0375, "step": 85780 }, { "grad_norm": 0.4052993357181549, "learning_rate": 5.420435125958934e-06, "loss": 0.0268, "step": 85790 }, { "grad_norm": 0.38419193029403687, "learning_rate": 5.412949974653209e-06, "loss": 0.0239, "step": 85800 }, { "grad_norm": 0.19207292795181274, "learning_rate": 5.405469699320248e-06, "loss": 0.0234, "step": 85810 }, { "grad_norm": 0.43102410435676575, "learning_rate": 5.397994300778064e-06, "loss": 0.0198, "step": 85820 }, { "grad_norm": 1.3254402875900269, "learning_rate": 5.39052377984417e-06, "loss": 0.0289, "step": 85830 }, { "grad_norm": 0.4096222221851349, "learning_rate": 5.383058137335523e-06, "loss": 0.0282, "step": 85840 }, { "grad_norm": 0.3094348609447479, "learning_rate": 5.375597374068553e-06, "loss": 0.0321, "step": 85850 }, { "grad_norm": 0.40844374895095825, "learning_rate": 5.368141490859163e-06, "loss": 0.028, "step": 85860 }, { "grad_norm": 0.3992556631565094, "learning_rate": 5.360690488522729e-06, "loss": 0.0309, "step": 85870 }, { "grad_norm": 0.2552291452884674, "learning_rate": 5.353244367874055e-06, "loss": 0.0322, "step": 85880 }, { "grad_norm": 0.40267425775527954, "learning_rate": 5.345803129727456e-06, "loss": 0.0366, "step": 85890 }, { "grad_norm": 0.35070374608039856, "learning_rate": 5.338366774896697e-06, "loss": 0.0185, "step": 85900 }, { "grad_norm": 0.5552952885627747, "learning_rate": 5.330935304194995e-06, "loss": 0.0303, "step": 85910 }, { "grad_norm": 0.7532849311828613, "learning_rate": 5.323508718435044e-06, "loss": 0.033, "step": 85920 }, { "grad_norm": 1.0180912017822266, "learning_rate": 5.316087018429011e-06, "loss": 0.026, "step": 85930 }, { "grad_norm": 0.3223066031932831, "learning_rate": 5.30867020498853e-06, "loss": 0.0199, "step": 85940 }, { "grad_norm": 0.5012175440788269, "learning_rate": 5.30125827892467e-06, "loss": 0.0243, "step": 85950 }, { "grad_norm": 0.5983884930610657, "learning_rate": 5.293851241048015e-06, "loss": 0.032, "step": 85960 }, { "grad_norm": 0.28414830565452576, "learning_rate": 5.286449092168555e-06, "loss": 0.0294, "step": 85970 }, { "grad_norm": 0.5690802931785583, "learning_rate": 5.2790518330958105e-06, "loss": 0.0339, "step": 85980 }, { "grad_norm": 0.32767099142074585, "learning_rate": 5.271659464638717e-06, "loss": 0.0275, "step": 85990 }, { "grad_norm": 0.2941473424434662, "learning_rate": 5.264271987605701e-06, "loss": 0.0238, "step": 86000 }, { "grad_norm": 0.30177417397499084, "learning_rate": 5.256889402804632e-06, "loss": 0.0247, "step": 86010 }, { "grad_norm": 0.30800938606262207, "learning_rate": 5.249511711042881e-06, "loss": 0.0289, "step": 86020 }, { "grad_norm": 0.3958551287651062, "learning_rate": 5.242138913127237e-06, "loss": 0.0207, "step": 86030 }, { "grad_norm": 0.7067539691925049, "learning_rate": 5.234771009863987e-06, "loss": 0.0242, "step": 86040 }, { "grad_norm": 0.6403993368148804, "learning_rate": 5.227408002058881e-06, "loss": 0.0258, "step": 86050 }, { "grad_norm": 0.3794499933719635, "learning_rate": 5.220049890517126e-06, "loss": 0.0326, "step": 86060 }, { "grad_norm": 0.702499270439148, "learning_rate": 5.212696676043377e-06, "loss": 0.0314, "step": 86070 }, { "grad_norm": 0.420693576335907, "learning_rate": 5.205348359441798e-06, "loss": 0.0266, "step": 86080 }, { "grad_norm": 0.4233816862106323, "learning_rate": 5.198004941515966e-06, "loss": 0.0302, "step": 86090 }, { "grad_norm": 0.435133159160614, "learning_rate": 5.190666423068952e-06, "loss": 0.0228, "step": 86100 }, { "grad_norm": 0.5071933269500732, "learning_rate": 5.183332804903284e-06, "loss": 0.0227, "step": 86110 }, { "grad_norm": 0.5171573162078857, "learning_rate": 5.176004087820979e-06, "loss": 0.0332, "step": 86120 }, { "grad_norm": 0.4361225366592407, "learning_rate": 5.168680272623455e-06, "loss": 0.0252, "step": 86130 }, { "grad_norm": 0.28963252902030945, "learning_rate": 5.161361360111661e-06, "loss": 0.0264, "step": 86140 }, { "grad_norm": 0.28235262632369995, "learning_rate": 5.154047351085984e-06, "loss": 0.022, "step": 86150 }, { "grad_norm": 0.7905383110046387, "learning_rate": 5.1467382463462385e-06, "loss": 0.0347, "step": 86160 }, { "grad_norm": 0.34393319487571716, "learning_rate": 5.139434046691788e-06, "loss": 0.0246, "step": 86170 }, { "grad_norm": 0.2859663963317871, "learning_rate": 5.13213475292137e-06, "loss": 0.0201, "step": 86180 }, { "grad_norm": 0.2440730333328247, "learning_rate": 5.12484036583325e-06, "loss": 0.0232, "step": 86190 }, { "grad_norm": 0.320250004529953, "learning_rate": 5.117550886225109e-06, "loss": 0.0216, "step": 86200 }, { "grad_norm": 0.7190495729446411, "learning_rate": 5.11026631489413e-06, "loss": 0.0267, "step": 86210 }, { "grad_norm": 0.47432392835617065, "learning_rate": 5.102986652636926e-06, "loss": 0.0291, "step": 86220 }, { "grad_norm": 0.24352607131004333, "learning_rate": 5.095711900249606e-06, "loss": 0.0261, "step": 86230 }, { "grad_norm": 0.4126196801662445, "learning_rate": 5.088442058527715e-06, "loss": 0.0231, "step": 86240 }, { "grad_norm": 0.4142136871814728, "learning_rate": 5.08117712826629e-06, "loss": 0.0232, "step": 86250 }, { "grad_norm": 0.34133413434028625, "learning_rate": 5.073917110259785e-06, "loss": 0.0298, "step": 86260 }, { "grad_norm": 0.2735835313796997, "learning_rate": 5.066662005302175e-06, "loss": 0.0186, "step": 86270 }, { "grad_norm": 0.6194241046905518, "learning_rate": 5.059411814186838e-06, "loss": 0.0273, "step": 86280 }, { "grad_norm": 0.4120456576347351, "learning_rate": 5.05216653770666e-06, "loss": 0.0344, "step": 86290 }, { "grad_norm": 0.34345218539237976, "learning_rate": 5.044926176653969e-06, "loss": 0.0268, "step": 86300 }, { "grad_norm": 0.3203481435775757, "learning_rate": 5.037690731820571e-06, "loss": 0.021, "step": 86310 }, { "grad_norm": 0.267591655254364, "learning_rate": 5.030460203997711e-06, "loss": 0.0244, "step": 86320 }, { "grad_norm": 0.37473928928375244, "learning_rate": 5.0232345939761174e-06, "loss": 0.0225, "step": 86330 }, { "grad_norm": 0.3715893626213074, "learning_rate": 5.016013902545957e-06, "loss": 0.0305, "step": 86340 }, { "grad_norm": 0.5884615778923035, "learning_rate": 5.008798130496889e-06, "loss": 0.0303, "step": 86350 }, { "grad_norm": 0.48843875527381897, "learning_rate": 5.001587278618008e-06, "loss": 0.0323, "step": 86360 }, { "grad_norm": 0.46216753125190735, "learning_rate": 4.994381347697891e-06, "loss": 0.0205, "step": 86370 }, { "grad_norm": 0.3215785026550293, "learning_rate": 4.987180338524572e-06, "loss": 0.0244, "step": 86380 }, { "grad_norm": 0.3106445372104645, "learning_rate": 4.9799842518855275e-06, "loss": 0.0272, "step": 86390 }, { "grad_norm": 0.520401120185852, "learning_rate": 4.9727930885677285e-06, "loss": 0.0273, "step": 86400 }, { "grad_norm": 0.29369574785232544, "learning_rate": 4.965606849357573e-06, "loss": 0.0293, "step": 86410 }, { "grad_norm": 0.5535122752189636, "learning_rate": 4.9584255350409445e-06, "loss": 0.0256, "step": 86420 }, { "grad_norm": 0.231251522898674, "learning_rate": 4.951249146403175e-06, "loss": 0.024, "step": 86430 }, { "grad_norm": 0.3230290114879608, "learning_rate": 4.9440776842290815e-06, "loss": 0.0244, "step": 86440 }, { "grad_norm": 1.1430391073226929, "learning_rate": 4.936911149302903e-06, "loss": 0.0356, "step": 86450 }, { "grad_norm": 0.4803885519504547, "learning_rate": 4.929749542408379e-06, "loss": 0.0243, "step": 86460 }, { "grad_norm": 0.3577582538127899, "learning_rate": 4.922592864328673e-06, "loss": 0.0188, "step": 86470 }, { "grad_norm": 0.7111492156982422, "learning_rate": 4.915441115846436e-06, "loss": 0.0392, "step": 86480 }, { "grad_norm": 0.3929140865802765, "learning_rate": 4.908294297743782e-06, "loss": 0.0336, "step": 86490 }, { "grad_norm": 0.5310772061347961, "learning_rate": 4.901152410802273e-06, "loss": 0.0312, "step": 86500 }, { "grad_norm": 0.586074709892273, "learning_rate": 4.894015455802919e-06, "loss": 0.0253, "step": 86510 }, { "grad_norm": 0.3270707130432129, "learning_rate": 4.8868834335262346e-06, "loss": 0.0236, "step": 86520 }, { "grad_norm": 0.41342678666114807, "learning_rate": 4.87975634475214e-06, "loss": 0.0264, "step": 86530 }, { "grad_norm": 0.21239100396633148, "learning_rate": 4.8726341902600614e-06, "loss": 0.0283, "step": 86540 }, { "grad_norm": 0.27315542101860046, "learning_rate": 4.865516970828843e-06, "loss": 0.0231, "step": 86550 }, { "grad_norm": 0.48716092109680176, "learning_rate": 4.85840468723685e-06, "loss": 0.0252, "step": 86560 }, { "grad_norm": 0.2541520893573761, "learning_rate": 4.85129734026184e-06, "loss": 0.0268, "step": 86570 }, { "grad_norm": 0.2689129114151001, "learning_rate": 4.844194930681073e-06, "loss": 0.0174, "step": 86580 }, { "grad_norm": 0.23967495560646057, "learning_rate": 4.837097459271267e-06, "loss": 0.0298, "step": 86590 }, { "grad_norm": 0.5168798565864563, "learning_rate": 4.830004926808573e-06, "loss": 0.0224, "step": 86600 }, { "grad_norm": 0.5125444531440735, "learning_rate": 4.822917334068627e-06, "loss": 0.0238, "step": 86610 }, { "grad_norm": 1.120004653930664, "learning_rate": 4.81583468182652e-06, "loss": 0.0191, "step": 86620 }, { "grad_norm": 0.21736018359661102, "learning_rate": 4.808756970856809e-06, "loss": 0.0328, "step": 86630 }, { "grad_norm": 0.29617488384246826, "learning_rate": 4.801684201933476e-06, "loss": 0.0256, "step": 86640 }, { "grad_norm": 0.1905381828546524, "learning_rate": 4.794616375830019e-06, "loss": 0.022, "step": 86650 }, { "grad_norm": 0.6491953730583191, "learning_rate": 4.787553493319341e-06, "loss": 0.035, "step": 86660 }, { "grad_norm": 0.19089120626449585, "learning_rate": 4.780495555173836e-06, "loss": 0.0253, "step": 86670 }, { "grad_norm": 0.33234283328056335, "learning_rate": 4.7734425621653475e-06, "loss": 0.0269, "step": 86680 }, { "grad_norm": 0.3175787329673767, "learning_rate": 4.766394515065198e-06, "loss": 0.0219, "step": 86690 }, { "grad_norm": 0.18989014625549316, "learning_rate": 4.75935141464412e-06, "loss": 0.0254, "step": 86700 }, { "grad_norm": 0.3470418453216553, "learning_rate": 4.75231326167237e-06, "loss": 0.021, "step": 86710 }, { "grad_norm": 0.3134746551513672, "learning_rate": 4.745280056919599e-06, "loss": 0.0209, "step": 86720 }, { "grad_norm": 0.32581448554992676, "learning_rate": 4.738251801154975e-06, "loss": 0.0237, "step": 86730 }, { "grad_norm": 0.28928324580192566, "learning_rate": 4.731228495147061e-06, "loss": 0.0281, "step": 86740 }, { "grad_norm": 0.42207252979278564, "learning_rate": 4.7242101396639595e-06, "loss": 0.0298, "step": 86750 }, { "grad_norm": 0.22702988982200623, "learning_rate": 4.717196735473151e-06, "loss": 0.0241, "step": 86760 }, { "grad_norm": 0.40615108609199524, "learning_rate": 4.710188283341643e-06, "loss": 0.0273, "step": 86770 }, { "grad_norm": 0.3213616609573364, "learning_rate": 4.703184784035835e-06, "loss": 0.0381, "step": 86780 }, { "grad_norm": 0.3229941129684448, "learning_rate": 4.696186238321642e-06, "loss": 0.025, "step": 86790 }, { "grad_norm": 0.37245893478393555, "learning_rate": 4.689192646964408e-06, "loss": 0.0279, "step": 86800 }, { "grad_norm": 0.5249111652374268, "learning_rate": 4.682204010728952e-06, "loss": 0.0283, "step": 86810 }, { "grad_norm": 0.23431260883808136, "learning_rate": 4.67522033037952e-06, "loss": 0.0208, "step": 86820 }, { "grad_norm": 0.33043357729911804, "learning_rate": 4.668241606679846e-06, "loss": 0.0314, "step": 86830 }, { "grad_norm": 0.3961677551269531, "learning_rate": 4.66126784039313e-06, "loss": 0.0239, "step": 86840 }, { "grad_norm": 0.2918400168418884, "learning_rate": 4.6542990322819836e-06, "loss": 0.0283, "step": 86850 }, { "grad_norm": 0.1981908679008484, "learning_rate": 4.64733518310852e-06, "loss": 0.0249, "step": 86860 }, { "grad_norm": 0.4251110851764679, "learning_rate": 4.640376293634291e-06, "loss": 0.0168, "step": 86870 }, { "grad_norm": 0.38972970843315125, "learning_rate": 4.633422364620327e-06, "loss": 0.0306, "step": 86880 }, { "grad_norm": 0.6129607558250427, "learning_rate": 4.6264733968270745e-06, "loss": 0.0339, "step": 86890 }, { "grad_norm": 0.4181627035140991, "learning_rate": 4.619529391014482e-06, "loss": 0.0314, "step": 86900 }, { "grad_norm": 0.23652882874011993, "learning_rate": 4.61259034794192e-06, "loss": 0.0294, "step": 86910 }, { "grad_norm": 0.6204208135604858, "learning_rate": 4.605656268368247e-06, "loss": 0.0294, "step": 86920 }, { "grad_norm": 0.2305627465248108, "learning_rate": 4.598727153051735e-06, "loss": 0.0304, "step": 86930 }, { "grad_norm": 0.2953154742717743, "learning_rate": 4.591803002750178e-06, "loss": 0.0229, "step": 86940 }, { "grad_norm": 0.3555179238319397, "learning_rate": 4.58488381822077e-06, "loss": 0.0234, "step": 86950 }, { "grad_norm": 1.116361141204834, "learning_rate": 4.5779696002201946e-06, "loss": 0.0272, "step": 86960 }, { "grad_norm": 0.32194784283638, "learning_rate": 4.571060349504563e-06, "loss": 0.0255, "step": 86970 }, { "grad_norm": 0.31860506534576416, "learning_rate": 4.564156066829478e-06, "loss": 0.0232, "step": 86980 }, { "grad_norm": 0.3375367224216461, "learning_rate": 4.557256752949957e-06, "loss": 0.0311, "step": 86990 }, { "grad_norm": 0.4466029405593872, "learning_rate": 4.550362408620534e-06, "loss": 0.0279, "step": 87000 }, { "grad_norm": 0.18106189370155334, "learning_rate": 4.543473034595136e-06, "loss": 0.0267, "step": 87010 }, { "grad_norm": 0.6279038190841675, "learning_rate": 4.536588631627192e-06, "loss": 0.0254, "step": 87020 }, { "grad_norm": 0.5831702947616577, "learning_rate": 4.5297092004695565e-06, "loss": 0.028, "step": 87030 }, { "grad_norm": 0.2879873514175415, "learning_rate": 4.522834741874555e-06, "loss": 0.0232, "step": 87040 }, { "grad_norm": 0.3752816319465637, "learning_rate": 4.515965256593974e-06, "loss": 0.0247, "step": 87050 }, { "grad_norm": 0.5838091373443604, "learning_rate": 4.509100745379058e-06, "loss": 0.0316, "step": 87060 }, { "grad_norm": 0.312002569437027, "learning_rate": 4.502241208980479e-06, "loss": 0.025, "step": 87070 }, { "grad_norm": 0.8070008158683777, "learning_rate": 4.495386648148398e-06, "loss": 0.0259, "step": 87080 }, { "grad_norm": 0.3349057734012604, "learning_rate": 4.488537063632431e-06, "loss": 0.0245, "step": 87090 }, { "grad_norm": 0.4659658372402191, "learning_rate": 4.481692456181608e-06, "loss": 0.0236, "step": 87100 }, { "grad_norm": 0.5823246836662292, "learning_rate": 4.474852826544479e-06, "loss": 0.0258, "step": 87110 }, { "grad_norm": 0.4146362543106079, "learning_rate": 4.468018175468974e-06, "loss": 0.0288, "step": 87120 }, { "grad_norm": 0.3221823275089264, "learning_rate": 4.461188503702568e-06, "loss": 0.0401, "step": 87130 }, { "grad_norm": 1.7476378679275513, "learning_rate": 4.454363811992107e-06, "loss": 0.034, "step": 87140 }, { "grad_norm": 0.339229017496109, "learning_rate": 4.447544101083956e-06, "loss": 0.0283, "step": 87150 }, { "grad_norm": 0.4622066020965576, "learning_rate": 4.440729371723884e-06, "loss": 0.031, "step": 87160 }, { "grad_norm": 0.23190784454345703, "learning_rate": 4.433919624657162e-06, "loss": 0.0212, "step": 87170 }, { "grad_norm": 0.3240543603897095, "learning_rate": 4.427114860628462e-06, "loss": 0.026, "step": 87180 }, { "grad_norm": 0.2751222848892212, "learning_rate": 4.420315080381982e-06, "loss": 0.0258, "step": 87190 }, { "grad_norm": 0.3407225012779236, "learning_rate": 4.4135202846613045e-06, "loss": 0.0313, "step": 87200 }, { "grad_norm": 0.47014251351356506, "learning_rate": 4.406730474209525e-06, "loss": 0.0261, "step": 87210 }, { "grad_norm": 0.3753294050693512, "learning_rate": 4.399945649769138e-06, "loss": 0.0294, "step": 87220 }, { "grad_norm": 0.40211090445518494, "learning_rate": 4.393165812082151e-06, "loss": 0.025, "step": 87230 }, { "grad_norm": 0.3363155424594879, "learning_rate": 4.386390961889969e-06, "loss": 0.0405, "step": 87240 }, { "grad_norm": 0.35095399618148804, "learning_rate": 4.3796210999334895e-06, "loss": 0.0251, "step": 87250 }, { "grad_norm": 0.35593581199645996, "learning_rate": 4.372856226953065e-06, "loss": 0.0257, "step": 87260 }, { "grad_norm": 0.2716502547264099, "learning_rate": 4.366096343688486e-06, "loss": 0.026, "step": 87270 }, { "grad_norm": 0.38499510288238525, "learning_rate": 4.359341450878995e-06, "loss": 0.0212, "step": 87280 }, { "grad_norm": 0.46297982335090637, "learning_rate": 4.352591549263302e-06, "loss": 0.0285, "step": 87290 }, { "grad_norm": 0.2713642716407776, "learning_rate": 4.345846639579582e-06, "loss": 0.0276, "step": 87300 }, { "grad_norm": 0.38213229179382324, "learning_rate": 4.339106722565417e-06, "loss": 0.0233, "step": 87310 }, { "grad_norm": 0.4641241431236267, "learning_rate": 4.332371798957896e-06, "loss": 0.0327, "step": 87320 }, { "grad_norm": 0.44378989934921265, "learning_rate": 4.3256418694935295e-06, "loss": 0.0319, "step": 87330 }, { "grad_norm": 0.25610753893852234, "learning_rate": 4.3189169349083125e-06, "loss": 0.0232, "step": 87340 }, { "grad_norm": 0.24911746382713318, "learning_rate": 4.312196995937645e-06, "loss": 0.0236, "step": 87350 }, { "grad_norm": 0.44267573952674866, "learning_rate": 4.305482053316434e-06, "loss": 0.0278, "step": 87360 }, { "grad_norm": 0.3814820945262909, "learning_rate": 4.2987721077789925e-06, "loss": 0.0253, "step": 87370 }, { "grad_norm": 0.24649876356124878, "learning_rate": 4.2920671600591325e-06, "loss": 0.0247, "step": 87380 }, { "grad_norm": 0.5010753870010376, "learning_rate": 4.285367210890084e-06, "loss": 0.026, "step": 87390 }, { "grad_norm": 0.2944866418838501, "learning_rate": 4.278672261004551e-06, "loss": 0.029, "step": 87400 }, { "grad_norm": 0.41526293754577637, "learning_rate": 4.2719823111346704e-06, "loss": 0.0353, "step": 87410 }, { "grad_norm": 0.3964945673942566, "learning_rate": 4.265297362012061e-06, "loss": 0.0207, "step": 87420 }, { "grad_norm": 0.41067659854888916, "learning_rate": 4.258617414367766e-06, "loss": 0.0253, "step": 87430 }, { "grad_norm": 0.2772176265716553, "learning_rate": 4.251942468932296e-06, "loss": 0.0255, "step": 87440 }, { "grad_norm": 0.3781396150588989, "learning_rate": 4.2452725264356175e-06, "loss": 0.0229, "step": 87450 }, { "grad_norm": 0.9195890426635742, "learning_rate": 4.2386075876071575e-06, "loss": 0.0231, "step": 87460 }, { "grad_norm": 0.2825111746788025, "learning_rate": 4.231947653175755e-06, "loss": 0.0194, "step": 87470 }, { "grad_norm": 0.6206338405609131, "learning_rate": 4.225292723869761e-06, "loss": 0.0347, "step": 87480 }, { "grad_norm": 0.2768358588218689, "learning_rate": 4.218642800416916e-06, "loss": 0.0243, "step": 87490 }, { "grad_norm": 0.5216966271400452, "learning_rate": 4.211997883544472e-06, "loss": 0.016, "step": 87500 }, { "grad_norm": 0.21200689673423767, "learning_rate": 4.205357973979096e-06, "loss": 0.0219, "step": 87510 }, { "grad_norm": 0.28056108951568604, "learning_rate": 4.198723072446925e-06, "loss": 0.0255, "step": 87520 }, { "grad_norm": 0.3542073965072632, "learning_rate": 4.192093179673534e-06, "loss": 0.0237, "step": 87530 }, { "grad_norm": 0.29675331711769104, "learning_rate": 4.185468296383959e-06, "loss": 0.025, "step": 87540 }, { "grad_norm": 0.33437857031822205, "learning_rate": 4.1788484233027005e-06, "loss": 0.0235, "step": 87550 }, { "grad_norm": 0.5099217891693115, "learning_rate": 4.172233561153677e-06, "loss": 0.0257, "step": 87560 }, { "grad_norm": 0.3799522817134857, "learning_rate": 4.165623710660288e-06, "loss": 0.0243, "step": 87570 }, { "grad_norm": 0.37472042441368103, "learning_rate": 4.1590188725453835e-06, "loss": 0.0273, "step": 87580 }, { "grad_norm": 0.3340625762939453, "learning_rate": 4.152419047531259e-06, "loss": 0.0257, "step": 87590 }, { "grad_norm": 0.3833712637424469, "learning_rate": 4.145824236339651e-06, "loss": 0.0234, "step": 87600 }, { "grad_norm": 0.2031858116388321, "learning_rate": 4.1392344396917695e-06, "loss": 0.0193, "step": 87610 }, { "grad_norm": 0.26103371381759644, "learning_rate": 4.132649658308252e-06, "loss": 0.0265, "step": 87620 }, { "grad_norm": 0.4044608473777771, "learning_rate": 4.126069892909213e-06, "loss": 0.0215, "step": 87630 }, { "grad_norm": 0.2150769680738449, "learning_rate": 4.119495144214191e-06, "loss": 0.0258, "step": 87640 }, { "grad_norm": 0.2580558657646179, "learning_rate": 4.112925412942214e-06, "loss": 0.0215, "step": 87650 }, { "grad_norm": 0.3931869566440582, "learning_rate": 4.106360699811712e-06, "loss": 0.0229, "step": 87660 }, { "grad_norm": 0.27968117594718933, "learning_rate": 4.099801005540616e-06, "loss": 0.0301, "step": 87670 }, { "grad_norm": 0.254190593957901, "learning_rate": 4.093246330846256e-06, "loss": 0.0235, "step": 87680 }, { "grad_norm": 0.39556825160980225, "learning_rate": 4.086696676445467e-06, "loss": 0.0266, "step": 87690 }, { "grad_norm": 0.31109756231307983, "learning_rate": 4.0801520430544914e-06, "loss": 0.0225, "step": 87700 }, { "grad_norm": 0.3013036251068115, "learning_rate": 4.073612431389062e-06, "loss": 0.0293, "step": 87710 }, { "grad_norm": 0.301888108253479, "learning_rate": 4.067077842164319e-06, "loss": 0.0264, "step": 87720 }, { "grad_norm": 0.47736871242523193, "learning_rate": 4.0605482760948945e-06, "loss": 0.03, "step": 87730 }, { "grad_norm": 0.323047935962677, "learning_rate": 4.054023733894829e-06, "loss": 0.0327, "step": 87740 }, { "grad_norm": 0.3334044814109802, "learning_rate": 4.047504216277653e-06, "loss": 0.0257, "step": 87750 }, { "grad_norm": 0.431484580039978, "learning_rate": 4.040989723956323e-06, "loss": 0.0307, "step": 87760 }, { "grad_norm": 0.5132182240486145, "learning_rate": 4.03448025764327e-06, "loss": 0.034, "step": 87770 }, { "grad_norm": 0.2969397306442261, "learning_rate": 4.027975818050339e-06, "loss": 0.0272, "step": 87780 }, { "grad_norm": 1.8932794332504272, "learning_rate": 4.021476405888858e-06, "loss": 0.0293, "step": 87790 }, { "grad_norm": 0.5581536889076233, "learning_rate": 4.01498202186959e-06, "loss": 0.0265, "step": 87800 }, { "grad_norm": 0.40001749992370605, "learning_rate": 4.008492666702751e-06, "loss": 0.0242, "step": 87810 }, { "grad_norm": 0.3673208951950073, "learning_rate": 4.002008341098001e-06, "loss": 0.0277, "step": 87820 }, { "grad_norm": 0.2727344334125519, "learning_rate": 3.995529045764462e-06, "loss": 0.0192, "step": 87830 }, { "grad_norm": 0.23907162249088287, "learning_rate": 3.989054781410712e-06, "loss": 0.0244, "step": 87840 }, { "grad_norm": 0.30260801315307617, "learning_rate": 3.982585548744744e-06, "loss": 0.0212, "step": 87850 }, { "grad_norm": 0.21009017527103424, "learning_rate": 3.9761213484740435e-06, "loss": 0.0204, "step": 87860 }, { "grad_norm": 0.3581054210662842, "learning_rate": 3.9696621813055055e-06, "loss": 0.0182, "step": 87870 }, { "grad_norm": 0.25105834007263184, "learning_rate": 3.9632080479455044e-06, "loss": 0.0256, "step": 87880 }, { "grad_norm": 0.5552303194999695, "learning_rate": 3.956758949099859e-06, "loss": 0.03, "step": 87890 }, { "grad_norm": 0.3931501507759094, "learning_rate": 3.9503148854738385e-06, "loss": 0.0352, "step": 87900 }, { "grad_norm": 0.4677280783653259, "learning_rate": 3.943875857772139e-06, "loss": 0.041, "step": 87910 }, { "grad_norm": 0.3378680646419525, "learning_rate": 3.9374418666989375e-06, "loss": 0.0271, "step": 87920 }, { "grad_norm": 0.21638600528240204, "learning_rate": 3.931012912957838e-06, "loss": 0.0235, "step": 87930 }, { "grad_norm": 0.8216220140457153, "learning_rate": 3.9245889972519045e-06, "loss": 0.0259, "step": 87940 }, { "grad_norm": 0.22782059013843536, "learning_rate": 3.9181701202836265e-06, "loss": 0.0256, "step": 87950 }, { "grad_norm": 0.24870087206363678, "learning_rate": 3.911756282755003e-06, "loss": 0.0239, "step": 87960 }, { "grad_norm": 0.24639618396759033, "learning_rate": 3.905347485367411e-06, "loss": 0.025, "step": 87970 }, { "grad_norm": 0.2141178697347641, "learning_rate": 3.898943728821725e-06, "loss": 0.0188, "step": 87980 }, { "grad_norm": 0.5120988488197327, "learning_rate": 3.892545013818227e-06, "loss": 0.0272, "step": 87990 }, { "grad_norm": 0.38992130756378174, "learning_rate": 3.886151341056687e-06, "loss": 0.0241, "step": 88000 }, { "grad_norm": 0.5022687911987305, "learning_rate": 3.8797627112363066e-06, "loss": 0.0315, "step": 88010 }, { "grad_norm": 0.293211966753006, "learning_rate": 3.873379125055743e-06, "loss": 0.0175, "step": 88020 }, { "grad_norm": 0.4694760739803314, "learning_rate": 3.867000583213082e-06, "loss": 0.0261, "step": 88030 }, { "grad_norm": 0.4444483816623688, "learning_rate": 3.860627086405871e-06, "loss": 0.0292, "step": 88040 }, { "grad_norm": 0.42450401186943054, "learning_rate": 3.854258635331126e-06, "loss": 0.0268, "step": 88050 }, { "grad_norm": 0.30272093415260315, "learning_rate": 3.8478952306852665e-06, "loss": 0.0311, "step": 88060 }, { "grad_norm": 0.33422383666038513, "learning_rate": 3.841536873164198e-06, "loss": 0.0216, "step": 88070 }, { "grad_norm": 0.7426660060882568, "learning_rate": 3.8351835634632626e-06, "loss": 0.0266, "step": 88080 }, { "grad_norm": 0.3381359279155731, "learning_rate": 3.828835302277245e-06, "loss": 0.0315, "step": 88090 }, { "grad_norm": 0.36918455362319946, "learning_rate": 3.822492090300378e-06, "loss": 0.0284, "step": 88100 }, { "grad_norm": 0.38413873314857483, "learning_rate": 3.8161539282263566e-06, "loss": 0.0264, "step": 88110 }, { "grad_norm": 0.38815784454345703, "learning_rate": 3.8098208167482995e-06, "loss": 0.0237, "step": 88120 }, { "grad_norm": 0.27255213260650635, "learning_rate": 3.803492756558796e-06, "loss": 0.0247, "step": 88130 }, { "grad_norm": 0.4035804569721222, "learning_rate": 3.797169748349855e-06, "loss": 0.0272, "step": 88140 }, { "grad_norm": 0.34039580821990967, "learning_rate": 3.790851792812977e-06, "loss": 0.021, "step": 88150 }, { "grad_norm": 0.1468328833580017, "learning_rate": 3.784538890639072e-06, "loss": 0.0186, "step": 88160 }, { "grad_norm": 0.17402738332748413, "learning_rate": 3.7782310425185153e-06, "loss": 0.0278, "step": 88170 }, { "grad_norm": 0.22658003866672516, "learning_rate": 3.77192824914111e-06, "loss": 0.0221, "step": 88180 }, { "grad_norm": 0.29319071769714355, "learning_rate": 3.7656305111961333e-06, "loss": 0.0267, "step": 88190 }, { "grad_norm": 0.3483060300350189, "learning_rate": 3.7593378293722782e-06, "loss": 0.0266, "step": 88200 }, { "grad_norm": 0.4810851812362671, "learning_rate": 3.7530502043577333e-06, "loss": 0.0266, "step": 88210 }, { "grad_norm": 0.6181169152259827, "learning_rate": 3.746767636840076e-06, "loss": 0.0281, "step": 88220 }, { "grad_norm": 0.23708932101726532, "learning_rate": 3.74049012750638e-06, "loss": 0.0281, "step": 88230 }, { "grad_norm": 0.4293220639228821, "learning_rate": 3.734217677043128e-06, "loss": 0.0238, "step": 88240 }, { "grad_norm": 0.6260327100753784, "learning_rate": 3.727950286136267e-06, "loss": 0.0246, "step": 88250 }, { "grad_norm": 0.3578684329986572, "learning_rate": 3.7216879554711925e-06, "loss": 0.0256, "step": 88260 }, { "grad_norm": 0.5291414856910706, "learning_rate": 3.715430685732757e-06, "loss": 0.0259, "step": 88270 }, { "grad_norm": 0.7064237594604492, "learning_rate": 3.7091784776052243e-06, "loss": 0.0209, "step": 88280 }, { "grad_norm": 0.4221409261226654, "learning_rate": 3.7029313317723356e-06, "loss": 0.0215, "step": 88290 }, { "grad_norm": 0.21007394790649414, "learning_rate": 3.6966892489172787e-06, "loss": 0.0306, "step": 88300 }, { "grad_norm": 0.31590941548347473, "learning_rate": 3.6904522297226575e-06, "loss": 0.0243, "step": 88310 }, { "grad_norm": 0.20549950003623962, "learning_rate": 3.6842202748705647e-06, "loss": 0.0248, "step": 88320 }, { "grad_norm": 0.37608277797698975, "learning_rate": 3.677993385042494e-06, "loss": 0.0254, "step": 88330 }, { "grad_norm": 0.5825985074043274, "learning_rate": 3.67177156091944e-06, "loss": 0.0265, "step": 88340 }, { "grad_norm": 0.5354771614074707, "learning_rate": 3.6655548031817866e-06, "loss": 0.0225, "step": 88350 }, { "grad_norm": 0.44689005613327026, "learning_rate": 3.6593431125094e-06, "loss": 0.0289, "step": 88360 }, { "grad_norm": 0.4095403254032135, "learning_rate": 3.6531364895815766e-06, "loss": 0.0212, "step": 88370 }, { "grad_norm": 0.5638881921768188, "learning_rate": 3.6469349350770677e-06, "loss": 0.028, "step": 88380 }, { "grad_norm": 0.2806222438812256, "learning_rate": 3.6407384496740525e-06, "loss": 0.0275, "step": 88390 }, { "grad_norm": 0.5108308792114258, "learning_rate": 3.6345470340501954e-06, "loss": 0.0394, "step": 88400 }, { "grad_norm": 0.27062493562698364, "learning_rate": 3.6283606888825537e-06, "loss": 0.0202, "step": 88410 }, { "grad_norm": 0.4179452657699585, "learning_rate": 3.622179414847676e-06, "loss": 0.0287, "step": 88420 }, { "grad_norm": 0.707833468914032, "learning_rate": 3.616003212621527e-06, "loss": 0.0314, "step": 88430 }, { "grad_norm": 0.371009886264801, "learning_rate": 3.609832082879533e-06, "loss": 0.0258, "step": 88440 }, { "grad_norm": 0.17029057443141937, "learning_rate": 3.6036660262965485e-06, "loss": 0.0273, "step": 88450 }, { "grad_norm": 0.27418026328086853, "learning_rate": 3.597505043546895e-06, "loss": 0.0222, "step": 88460 }, { "grad_norm": 0.31274211406707764, "learning_rate": 3.5913491353043227e-06, "loss": 0.02, "step": 88470 }, { "grad_norm": 0.15313571691513062, "learning_rate": 3.5851983022420375e-06, "loss": 0.0299, "step": 88480 }, { "grad_norm": 0.40057989954948425, "learning_rate": 3.5790525450326797e-06, "loss": 0.0221, "step": 88490 }, { "grad_norm": 0.39672666788101196, "learning_rate": 3.572911864348344e-06, "loss": 0.0247, "step": 88500 }, { "grad_norm": 0.4912656545639038, "learning_rate": 3.5667762608605717e-06, "loss": 0.0265, "step": 88510 }, { "grad_norm": 0.40326982736587524, "learning_rate": 3.5606457352403255e-06, "loss": 0.0258, "step": 88520 }, { "grad_norm": 0.3933296203613281, "learning_rate": 3.554520288158042e-06, "loss": 0.0195, "step": 88530 }, { "grad_norm": 0.35853007435798645, "learning_rate": 3.5483999202835903e-06, "loss": 0.0225, "step": 88540 }, { "grad_norm": 0.39386260509490967, "learning_rate": 3.5422846322862913e-06, "loss": 0.0326, "step": 88550 }, { "grad_norm": 0.29013189673423767, "learning_rate": 3.5361744248348925e-06, "loss": 0.0228, "step": 88560 }, { "grad_norm": 0.30408304929733276, "learning_rate": 3.5300692985976047e-06, "loss": 0.0243, "step": 88570 }, { "grad_norm": 0.635138750076294, "learning_rate": 3.52396925424206e-06, "loss": 0.0212, "step": 88580 }, { "grad_norm": 0.46082067489624023, "learning_rate": 3.5178742924353747e-06, "loss": 0.0283, "step": 88590 }, { "grad_norm": 0.329384982585907, "learning_rate": 3.511784413844066e-06, "loss": 0.024, "step": 88600 }, { "grad_norm": 0.36936038732528687, "learning_rate": 3.5056996191341287e-06, "loss": 0.0271, "step": 88610 }, { "grad_norm": 0.5361416339874268, "learning_rate": 3.4996199089709692e-06, "loss": 0.0332, "step": 88620 }, { "grad_norm": 0.2124929577112198, "learning_rate": 3.4935452840194725e-06, "loss": 0.0251, "step": 88630 }, { "grad_norm": 0.5354697704315186, "learning_rate": 3.48747574494393e-06, "loss": 0.0273, "step": 88640 }, { "grad_norm": 0.39040127396583557, "learning_rate": 3.4814112924081154e-06, "loss": 0.0233, "step": 88650 }, { "grad_norm": 0.47249656915664673, "learning_rate": 3.475351927075221e-06, "loss": 0.0272, "step": 88660 }, { "grad_norm": 0.2171301692724228, "learning_rate": 3.469297649607894e-06, "loss": 0.0197, "step": 88670 }, { "grad_norm": 0.40284740924835205, "learning_rate": 3.463248460668217e-06, "loss": 0.0259, "step": 88680 }, { "grad_norm": 0.45974138379096985, "learning_rate": 3.457204360917726e-06, "loss": 0.0304, "step": 88690 }, { "grad_norm": 0.23368148505687714, "learning_rate": 3.4511653510173827e-06, "loss": 0.0221, "step": 88700 }, { "grad_norm": 0.22964933514595032, "learning_rate": 3.4451314316276074e-06, "loss": 0.0239, "step": 88710 }, { "grad_norm": 0.47651341557502747, "learning_rate": 3.4391026034082674e-06, "loss": 0.0298, "step": 88720 }, { "grad_norm": 0.2441319227218628, "learning_rate": 3.433078867018674e-06, "loss": 0.027, "step": 88730 }, { "grad_norm": 0.5517461895942688, "learning_rate": 3.427060223117551e-06, "loss": 0.0257, "step": 88740 }, { "grad_norm": 0.30932939052581787, "learning_rate": 3.4210466723631042e-06, "loss": 0.0238, "step": 88750 }, { "grad_norm": 0.47148507833480835, "learning_rate": 3.415038215412969e-06, "loss": 0.0227, "step": 88760 }, { "grad_norm": 0.416934609413147, "learning_rate": 3.4090348529242088e-06, "loss": 0.0242, "step": 88770 }, { "grad_norm": 0.4961385726928711, "learning_rate": 3.4030365855533487e-06, "loss": 0.0338, "step": 88780 }, { "grad_norm": 0.4917009770870209, "learning_rate": 3.397043413956347e-06, "loss": 0.0317, "step": 88790 }, { "grad_norm": 0.3821239173412323, "learning_rate": 3.3910553387886237e-06, "loss": 0.0224, "step": 88800 }, { "grad_norm": 0.3634127378463745, "learning_rate": 3.3850723607049996e-06, "loss": 0.028, "step": 88810 }, { "grad_norm": 0.568214476108551, "learning_rate": 3.3790944803597902e-06, "loss": 0.0225, "step": 88820 }, { "grad_norm": 0.3383716940879822, "learning_rate": 3.373121698406706e-06, "loss": 0.0297, "step": 88830 }, { "grad_norm": 0.450282484292984, "learning_rate": 3.3671540154989302e-06, "loss": 0.0241, "step": 88840 }, { "grad_norm": 0.3229287564754486, "learning_rate": 3.361191432289079e-06, "loss": 0.0278, "step": 88850 }, { "grad_norm": 0.39218980073928833, "learning_rate": 3.3552339494292207e-06, "loss": 0.0251, "step": 88860 }, { "grad_norm": 0.42834988236427307, "learning_rate": 3.349281567570839e-06, "loss": 0.0291, "step": 88870 }, { "grad_norm": 0.5648703575134277, "learning_rate": 3.3433342873648965e-06, "loss": 0.0344, "step": 88880 }, { "grad_norm": 0.308478444814682, "learning_rate": 3.337392109461762e-06, "loss": 0.0245, "step": 88890 }, { "grad_norm": 0.290551096200943, "learning_rate": 3.331455034511266e-06, "loss": 0.024, "step": 88900 }, { "grad_norm": 0.20992659032344818, "learning_rate": 3.325523063162689e-06, "loss": 0.0323, "step": 88910 }, { "grad_norm": 0.48595404624938965, "learning_rate": 3.3195961960647393e-06, "loss": 0.0232, "step": 88920 }, { "grad_norm": 0.27421456575393677, "learning_rate": 3.31367443386556e-06, "loss": 0.0212, "step": 88930 }, { "grad_norm": 0.27491965889930725, "learning_rate": 3.3077577772127556e-06, "loss": 0.025, "step": 88940 }, { "grad_norm": 0.34625235199928284, "learning_rate": 3.301846226753358e-06, "loss": 0.029, "step": 88950 }, { "grad_norm": 0.49580109119415283, "learning_rate": 3.2959397831338444e-06, "loss": 0.0245, "step": 88960 }, { "grad_norm": 0.34280189871788025, "learning_rate": 3.2900384470001432e-06, "loss": 0.0276, "step": 88970 }, { "grad_norm": 0.29186809062957764, "learning_rate": 3.28414221899761e-06, "loss": 0.0218, "step": 88980 }, { "grad_norm": 0.43242162466049194, "learning_rate": 3.278251099771046e-06, "loss": 0.0239, "step": 88990 }, { "grad_norm": 0.4949292838573456, "learning_rate": 3.2723650899646906e-06, "loss": 0.0283, "step": 89000 }, { "grad_norm": 0.42292502522468567, "learning_rate": 3.2664841902222466e-06, "loss": 0.0227, "step": 89010 }, { "grad_norm": 0.2644345164299011, "learning_rate": 3.2606084011868208e-06, "loss": 0.0304, "step": 89020 }, { "grad_norm": 0.26249322295188904, "learning_rate": 3.254737723500989e-06, "loss": 0.0229, "step": 89030 }, { "grad_norm": 0.443394273519516, "learning_rate": 3.2488721578067595e-06, "loss": 0.0238, "step": 89040 }, { "grad_norm": 0.30483120679855347, "learning_rate": 3.243011704745591e-06, "loss": 0.0277, "step": 89050 }, { "grad_norm": 0.4500083923339844, "learning_rate": 3.237156364958355e-06, "loss": 0.0257, "step": 89060 }, { "grad_norm": 0.2905125916004181, "learning_rate": 3.231306139085394e-06, "loss": 0.0247, "step": 89070 }, { "grad_norm": 0.44035273790359497, "learning_rate": 3.2254610277664797e-06, "loss": 0.0317, "step": 89080 }, { "grad_norm": 0.37238258123397827, "learning_rate": 3.2196210316408183e-06, "loss": 0.027, "step": 89090 }, { "grad_norm": 0.35368812084198, "learning_rate": 3.2137861513470644e-06, "loss": 0.0195, "step": 89100 }, { "grad_norm": 0.2741815447807312, "learning_rate": 3.2079563875233252e-06, "loss": 0.0233, "step": 89110 }, { "grad_norm": 0.2333061397075653, "learning_rate": 3.2021317408071184e-06, "loss": 0.0194, "step": 89120 }, { "grad_norm": 0.2736615836620331, "learning_rate": 3.196312211835434e-06, "loss": 0.028, "step": 89130 }, { "grad_norm": 0.42275768518447876, "learning_rate": 3.1904978012446686e-06, "loss": 0.0185, "step": 89140 }, { "grad_norm": 0.2741609811782837, "learning_rate": 3.184688509670691e-06, "loss": 0.0221, "step": 89150 }, { "grad_norm": 0.2452990561723709, "learning_rate": 3.178884337748783e-06, "loss": 0.0207, "step": 89160 }, { "grad_norm": 0.388028085231781, "learning_rate": 3.173085286113697e-06, "loss": 0.0255, "step": 89170 }, { "grad_norm": 0.2985742390155792, "learning_rate": 3.1672913553995986e-06, "loss": 0.0318, "step": 89180 }, { "grad_norm": 0.17029640078544617, "learning_rate": 3.1615025462401138e-06, "loss": 0.0185, "step": 89190 }, { "grad_norm": 0.3961735963821411, "learning_rate": 3.1557188592682808e-06, "loss": 0.0227, "step": 89200 }, { "grad_norm": 0.48988664150238037, "learning_rate": 3.1499402951166045e-06, "loss": 0.0189, "step": 89210 }, { "grad_norm": 0.2999882400035858, "learning_rate": 3.144166854417019e-06, "loss": 0.0221, "step": 89220 }, { "grad_norm": 0.5654280185699463, "learning_rate": 3.1383985378009073e-06, "loss": 0.0266, "step": 89230 }, { "grad_norm": 0.39635714888572693, "learning_rate": 3.1326353458990654e-06, "loss": 0.0282, "step": 89240 }, { "grad_norm": 0.3873276710510254, "learning_rate": 3.126877279341761e-06, "loss": 0.0202, "step": 89250 }, { "grad_norm": 0.3221455216407776, "learning_rate": 3.1211243387586964e-06, "loss": 0.0264, "step": 89260 }, { "grad_norm": 0.48376405239105225, "learning_rate": 3.115376524778979e-06, "loss": 0.0245, "step": 89270 }, { "grad_norm": 0.20979490876197815, "learning_rate": 3.109633838031195e-06, "loss": 0.0316, "step": 89280 }, { "grad_norm": 0.23749451339244843, "learning_rate": 3.103896279143359e-06, "loss": 0.0277, "step": 89290 }, { "grad_norm": 0.5671253800392151, "learning_rate": 3.098163848742924e-06, "loss": 0.0358, "step": 89300 }, { "grad_norm": 0.4775794744491577, "learning_rate": 3.0924365474567675e-06, "loss": 0.029, "step": 89310 }, { "grad_norm": 0.324032187461853, "learning_rate": 3.086714375911237e-06, "loss": 0.0249, "step": 89320 }, { "grad_norm": 0.3478989899158478, "learning_rate": 3.0809973347320775e-06, "loss": 0.0297, "step": 89330 }, { "grad_norm": 0.3279467821121216, "learning_rate": 3.075285424544516e-06, "loss": 0.029, "step": 89340 }, { "grad_norm": 0.5102726221084595, "learning_rate": 3.069578645973176e-06, "loss": 0.0244, "step": 89350 }, { "grad_norm": 0.3499177396297455, "learning_rate": 3.0638769996421735e-06, "loss": 0.0323, "step": 89360 }, { "grad_norm": 0.517933189868927, "learning_rate": 3.058180486175005e-06, "loss": 0.0306, "step": 89370 }, { "grad_norm": 0.32754674553871155, "learning_rate": 3.052489106194645e-06, "loss": 0.0263, "step": 89380 }, { "grad_norm": 0.5846385359764099, "learning_rate": 3.0468028603234946e-06, "loss": 0.0253, "step": 89390 }, { "grad_norm": 0.3376467525959015, "learning_rate": 3.04112174918339e-06, "loss": 0.02, "step": 89400 }, { "grad_norm": 0.6566863656044006, "learning_rate": 3.0354457733955954e-06, "loss": 0.0237, "step": 89410 }, { "grad_norm": 0.2932496666908264, "learning_rate": 3.029774933580859e-06, "loss": 0.0228, "step": 89420 }, { "grad_norm": 0.4818095862865448, "learning_rate": 3.0241092303593065e-06, "loss": 0.0258, "step": 89430 }, { "grad_norm": 0.2774663269519806, "learning_rate": 3.018448664350554e-06, "loss": 0.0221, "step": 89440 }, { "grad_norm": 0.29262951016426086, "learning_rate": 3.0127932361736055e-06, "loss": 0.0253, "step": 89450 }, { "grad_norm": 0.7028328776359558, "learning_rate": 3.00714294644695e-06, "loss": 0.0271, "step": 89460 }, { "grad_norm": 0.4421159625053406, "learning_rate": 3.0014977957884828e-06, "loss": 0.0261, "step": 89470 }, { "grad_norm": 0.31454548239707947, "learning_rate": 2.99585778481557e-06, "loss": 0.0268, "step": 89480 }, { "grad_norm": 0.5898545980453491, "learning_rate": 2.990222914144969e-06, "loss": 0.0315, "step": 89490 }, { "grad_norm": 0.7007471323013306, "learning_rate": 2.984593184392914e-06, "loss": 0.0293, "step": 89500 }, { "grad_norm": 0.2701758146286011, "learning_rate": 2.978968596175069e-06, "loss": 0.0267, "step": 89510 }, { "grad_norm": 1.2803207635879517, "learning_rate": 2.9733491501065137e-06, "loss": 0.0254, "step": 89520 }, { "grad_norm": 0.3524912893772125, "learning_rate": 2.9677348468018005e-06, "loss": 0.0277, "step": 89530 }, { "grad_norm": 0.21850286424160004, "learning_rate": 2.9621256868748726e-06, "loss": 0.0222, "step": 89540 }, { "grad_norm": 0.1874208152294159, "learning_rate": 2.956521670939183e-06, "loss": 0.0255, "step": 89550 }, { "grad_norm": 0.2323344498872757, "learning_rate": 2.950922799607542e-06, "loss": 0.0237, "step": 89560 }, { "grad_norm": 0.7338835000991821, "learning_rate": 2.9453290734922535e-06, "loss": 0.0207, "step": 89570 }, { "grad_norm": 0.3008083701133728, "learning_rate": 2.939740493205029e-06, "loss": 0.0362, "step": 89580 }, { "grad_norm": 0.2046620398759842, "learning_rate": 2.9341570593570345e-06, "loss": 0.0254, "step": 89590 }, { "grad_norm": 0.38104483485221863, "learning_rate": 2.928578772558843e-06, "loss": 0.0221, "step": 89600 }, { "grad_norm": 0.5557200312614441, "learning_rate": 2.923005633420528e-06, "loss": 0.0279, "step": 89610 }, { "grad_norm": 0.31915661692619324, "learning_rate": 2.9174376425515237e-06, "loss": 0.0259, "step": 89620 }, { "grad_norm": 0.6942251324653625, "learning_rate": 2.91187480056076e-06, "loss": 0.0294, "step": 89630 }, { "grad_norm": 0.4360610842704773, "learning_rate": 2.9063171080565664e-06, "loss": 0.0263, "step": 89640 }, { "grad_norm": 0.4466162919998169, "learning_rate": 2.900764565646741e-06, "loss": 0.0221, "step": 89650 }, { "grad_norm": 0.3040023446083069, "learning_rate": 2.895217173938475e-06, "loss": 0.0303, "step": 89660 }, { "grad_norm": 0.30472397804260254, "learning_rate": 2.8896749335384445e-06, "loss": 0.0298, "step": 89670 }, { "grad_norm": 0.24239027500152588, "learning_rate": 2.884137845052731e-06, "loss": 0.0227, "step": 89680 }, { "grad_norm": 0.35917991399765015, "learning_rate": 2.8786059090868734e-06, "loss": 0.0224, "step": 89690 }, { "grad_norm": 0.7792726755142212, "learning_rate": 2.87307912624582e-06, "loss": 0.0361, "step": 89700 }, { "grad_norm": 0.23841442167758942, "learning_rate": 2.867557497133977e-06, "loss": 0.0264, "step": 89710 }, { "grad_norm": 0.3002931475639343, "learning_rate": 2.862041022355194e-06, "loss": 0.0269, "step": 89720 }, { "grad_norm": 0.35712990164756775, "learning_rate": 2.8565297025127226e-06, "loss": 0.0281, "step": 89730 }, { "grad_norm": 0.3110673427581787, "learning_rate": 2.8510235382092853e-06, "loss": 0.0197, "step": 89740 }, { "grad_norm": 0.34246617555618286, "learning_rate": 2.845522530047029e-06, "loss": 0.0282, "step": 89750 }, { "grad_norm": 0.5680418014526367, "learning_rate": 2.8400266786275387e-06, "loss": 0.0305, "step": 89760 }, { "grad_norm": 0.545680046081543, "learning_rate": 2.834535984551817e-06, "loss": 0.0245, "step": 89770 }, { "grad_norm": 0.24627277255058289, "learning_rate": 2.8290504484203397e-06, "loss": 0.033, "step": 89780 }, { "grad_norm": 0.5162210464477539, "learning_rate": 2.8235700708329705e-06, "loss": 0.0278, "step": 89790 }, { "grad_norm": 0.5286935567855835, "learning_rate": 2.818094852389058e-06, "loss": 0.0247, "step": 89800 }, { "grad_norm": 0.25945043563842773, "learning_rate": 2.8126247936873516e-06, "loss": 0.0265, "step": 89810 }, { "grad_norm": 0.3876613676548004, "learning_rate": 2.8071598953260614e-06, "loss": 0.0249, "step": 89820 }, { "grad_norm": 0.4014628529548645, "learning_rate": 2.8017001579028035e-06, "loss": 0.0257, "step": 89830 }, { "grad_norm": 0.33801454305648804, "learning_rate": 2.7962455820146617e-06, "loss": 0.0346, "step": 89840 }, { "grad_norm": 0.24719269573688507, "learning_rate": 2.7907961682581253e-06, "loss": 0.0277, "step": 89850 }, { "grad_norm": 0.4669499099254608, "learning_rate": 2.7853519172291453e-06, "loss": 0.0302, "step": 89860 }, { "grad_norm": 0.33056434988975525, "learning_rate": 2.7799128295230904e-06, "loss": 0.0259, "step": 89870 }, { "grad_norm": 0.25313910841941833, "learning_rate": 2.774478905734784e-06, "loss": 0.0268, "step": 89880 }, { "grad_norm": 0.5293201208114624, "learning_rate": 2.7690501464584516e-06, "loss": 0.0248, "step": 89890 }, { "grad_norm": 0.9434283971786499, "learning_rate": 2.763626552287796e-06, "loss": 0.0212, "step": 89900 }, { "grad_norm": 0.3575618267059326, "learning_rate": 2.7582081238159083e-06, "loss": 0.0284, "step": 89910 }, { "grad_norm": 0.46051347255706787, "learning_rate": 2.7527948616353548e-06, "loss": 0.0269, "step": 89920 }, { "grad_norm": 0.38546982407569885, "learning_rate": 2.7473867663381226e-06, "loss": 0.0243, "step": 89930 }, { "grad_norm": 0.28129151463508606, "learning_rate": 2.7419838385156383e-06, "loss": 0.0238, "step": 89940 }, { "grad_norm": 0.53251051902771, "learning_rate": 2.7365860787587407e-06, "loss": 0.0303, "step": 89950 }, { "grad_norm": 0.6115577816963196, "learning_rate": 2.7311934876577295e-06, "loss": 0.0319, "step": 89960 }, { "grad_norm": 0.3081280291080475, "learning_rate": 2.725806065802339e-06, "loss": 0.0225, "step": 89970 }, { "grad_norm": 0.5044970512390137, "learning_rate": 2.720423813781714e-06, "loss": 0.0311, "step": 89980 }, { "grad_norm": 0.2750837206840515, "learning_rate": 2.715046732184462e-06, "loss": 0.0198, "step": 89990 }, { "grad_norm": 0.42114052176475525, "learning_rate": 2.7096748215986013e-06, "loss": 0.0221, "step": 90000 }, { "grad_norm": 0.8014368414878845, "learning_rate": 2.704308082611612e-06, "loss": 0.0253, "step": 90010 }, { "grad_norm": 0.3015703558921814, "learning_rate": 2.698946515810374e-06, "loss": 0.0247, "step": 90020 }, { "grad_norm": 0.5305701494216919, "learning_rate": 2.6935901217812367e-06, "loss": 0.0257, "step": 90030 }, { "grad_norm": 0.42906472086906433, "learning_rate": 2.688238901109952e-06, "loss": 0.0292, "step": 90040 }, { "grad_norm": 0.7749747037887573, "learning_rate": 2.68289285438173e-06, "loss": 0.0339, "step": 90050 }, { "grad_norm": 0.35160326957702637, "learning_rate": 2.6775519821812033e-06, "loss": 0.0262, "step": 90060 }, { "grad_norm": 0.32405996322631836, "learning_rate": 2.672216285092449e-06, "loss": 0.0285, "step": 90070 }, { "grad_norm": 0.4082207977771759, "learning_rate": 2.666885763698962e-06, "loss": 0.0229, "step": 90080 }, { "grad_norm": 0.3470044732093811, "learning_rate": 2.661560418583692e-06, "loss": 0.0344, "step": 90090 }, { "grad_norm": 0.22189848124980927, "learning_rate": 2.656240250328995e-06, "loss": 0.0218, "step": 90100 }, { "grad_norm": 0.3742346465587616, "learning_rate": 2.650925259516679e-06, "loss": 0.0339, "step": 90110 }, { "grad_norm": 0.1881193071603775, "learning_rate": 2.645615446727995e-06, "loss": 0.0237, "step": 90120 }, { "grad_norm": 0.36583781242370605, "learning_rate": 2.6403108125436115e-06, "loss": 0.0299, "step": 90130 }, { "grad_norm": 0.5729526877403259, "learning_rate": 2.6350113575436264e-06, "loss": 0.0265, "step": 90140 }, { "grad_norm": 0.2941334545612335, "learning_rate": 2.629717082307598e-06, "loss": 0.0277, "step": 90150 }, { "grad_norm": 0.4044271409511566, "learning_rate": 2.6244279874144793e-06, "loss": 0.0211, "step": 90160 }, { "grad_norm": 0.23768441379070282, "learning_rate": 2.6191440734426862e-06, "loss": 0.0251, "step": 90170 }, { "grad_norm": 0.4224392771720886, "learning_rate": 2.613865340970062e-06, "loss": 0.025, "step": 90180 }, { "grad_norm": 0.3688738942146301, "learning_rate": 2.608591790573889e-06, "loss": 0.0263, "step": 90190 }, { "grad_norm": 0.20515574514865875, "learning_rate": 2.603323422830861e-06, "loss": 0.0235, "step": 90200 }, { "grad_norm": 0.41442832350730896, "learning_rate": 2.5980602383171183e-06, "loss": 0.0304, "step": 90210 }, { "grad_norm": 0.397306889295578, "learning_rate": 2.5928022376082496e-06, "loss": 0.0307, "step": 90220 }, { "grad_norm": 0.26147589087486267, "learning_rate": 2.5875494212792505e-06, "loss": 0.0229, "step": 90230 }, { "grad_norm": 0.349581778049469, "learning_rate": 2.5823017899045565e-06, "loss": 0.0259, "step": 90240 }, { "grad_norm": 0.3761482238769531, "learning_rate": 2.577059344058047e-06, "loss": 0.0213, "step": 90250 }, { "grad_norm": 0.2968042492866516, "learning_rate": 2.5718220843130415e-06, "loss": 0.0256, "step": 90260 }, { "grad_norm": 0.7305581569671631, "learning_rate": 2.5665900112422535e-06, "loss": 0.0244, "step": 90270 }, { "grad_norm": 0.39288273453712463, "learning_rate": 2.561363125417876e-06, "loss": 0.0351, "step": 90280 }, { "grad_norm": 0.26209765672683716, "learning_rate": 2.556141427411496e-06, "loss": 0.018, "step": 90290 }, { "grad_norm": 0.6414013504981995, "learning_rate": 2.550924917794156e-06, "loss": 0.0261, "step": 90300 }, { "grad_norm": 0.26070743799209595, "learning_rate": 2.545713597136329e-06, "loss": 0.0181, "step": 90310 }, { "grad_norm": 0.23247231543064117, "learning_rate": 2.5405074660079298e-06, "loss": 0.0205, "step": 90320 }, { "grad_norm": 0.3484978973865509, "learning_rate": 2.5353065249782647e-06, "loss": 0.0375, "step": 90330 }, { "grad_norm": 0.4875461161136627, "learning_rate": 2.530110774616129e-06, "loss": 0.028, "step": 90340 }, { "grad_norm": 0.44933077692985535, "learning_rate": 2.524920215489701e-06, "loss": 0.0241, "step": 90350 }, { "grad_norm": 0.32386189699172974, "learning_rate": 2.519734848166627e-06, "loss": 0.0252, "step": 90360 }, { "grad_norm": 0.24387231469154358, "learning_rate": 2.514554673213948e-06, "loss": 0.0267, "step": 90370 }, { "grad_norm": 0.3413762152194977, "learning_rate": 2.5093796911981936e-06, "loss": 0.0196, "step": 90380 }, { "grad_norm": 0.30759695172309875, "learning_rate": 2.5042099026852672e-06, "loss": 0.0331, "step": 90390 }, { "grad_norm": 0.37518221139907837, "learning_rate": 2.499045308240544e-06, "loss": 0.0202, "step": 90400 }, { "grad_norm": 0.2865212857723236, "learning_rate": 2.4938859084288003e-06, "loss": 0.0217, "step": 90410 }, { "grad_norm": 0.2605617344379425, "learning_rate": 2.4887317038142734e-06, "loss": 0.0237, "step": 90420 }, { "grad_norm": 0.32579582929611206, "learning_rate": 2.4835826949606124e-06, "loss": 0.02, "step": 90430 }, { "grad_norm": 0.6435278654098511, "learning_rate": 2.4784388824309167e-06, "loss": 0.0318, "step": 90440 }, { "grad_norm": 0.40148332715034485, "learning_rate": 2.473300266787687e-06, "loss": 0.0242, "step": 90450 }, { "grad_norm": 0.537218451499939, "learning_rate": 2.46816684859289e-06, "loss": 0.027, "step": 90460 }, { "grad_norm": 0.4095841348171234, "learning_rate": 2.4630386284079043e-06, "loss": 0.0201, "step": 90470 }, { "grad_norm": 0.15811821818351746, "learning_rate": 2.4579156067935427e-06, "loss": 0.0264, "step": 90480 }, { "grad_norm": 0.3212384283542633, "learning_rate": 2.452797784310046e-06, "loss": 0.027, "step": 90490 }, { "grad_norm": 0.47272464632987976, "learning_rate": 2.447685161517105e-06, "loss": 0.0256, "step": 90500 }, { "grad_norm": 0.7969879508018494, "learning_rate": 2.4425777389738224e-06, "loss": 0.0253, "step": 90510 }, { "grad_norm": 0.33650365471839905, "learning_rate": 2.43747551723873e-06, "loss": 0.0177, "step": 90520 }, { "grad_norm": 0.21593572199344635, "learning_rate": 2.4323784968698136e-06, "loss": 0.0363, "step": 90530 }, { "grad_norm": 0.42198997735977173, "learning_rate": 2.4272866784244565e-06, "loss": 0.028, "step": 90540 }, { "grad_norm": 0.24462158977985382, "learning_rate": 2.422200062459518e-06, "loss": 0.0226, "step": 90550 }, { "grad_norm": 0.7675283551216125, "learning_rate": 2.41711864953123e-06, "loss": 0.0325, "step": 90560 }, { "grad_norm": 0.3905070424079895, "learning_rate": 2.412042440195322e-06, "loss": 0.0282, "step": 90570 }, { "grad_norm": 0.741083562374115, "learning_rate": 2.4069714350068993e-06, "loss": 0.0309, "step": 90580 }, { "grad_norm": 0.4565221965312958, "learning_rate": 2.4019056345205294e-06, "loss": 0.0237, "step": 90590 }, { "grad_norm": 0.42036086320877075, "learning_rate": 2.396845039290191e-06, "loss": 0.0321, "step": 90600 }, { "grad_norm": 0.24997444450855255, "learning_rate": 2.3917896498693194e-06, "loss": 0.0302, "step": 90610 }, { "grad_norm": 0.5065740346908569, "learning_rate": 2.3867394668107335e-06, "loss": 0.0323, "step": 90620 }, { "grad_norm": 0.16280066967010498, "learning_rate": 2.3816944906667527e-06, "loss": 0.0244, "step": 90630 }, { "grad_norm": 0.665678858757019, "learning_rate": 2.3766547219890634e-06, "loss": 0.0243, "step": 90640 }, { "grad_norm": 0.24682940542697906, "learning_rate": 2.371620161328819e-06, "loss": 0.0232, "step": 90650 }, { "grad_norm": 0.4445699155330658, "learning_rate": 2.366590809236574e-06, "loss": 0.0284, "step": 90660 }, { "grad_norm": 0.5652115941047668, "learning_rate": 2.3615666662623494e-06, "loss": 0.0247, "step": 90670 }, { "grad_norm": 0.5446616411209106, "learning_rate": 2.3565477329555664e-06, "loss": 0.0227, "step": 90680 }, { "grad_norm": 0.513521671295166, "learning_rate": 2.3515340098650975e-06, "loss": 0.0276, "step": 90690 }, { "grad_norm": 0.523646354675293, "learning_rate": 2.3465254975392258e-06, "loss": 0.0297, "step": 90700 }, { "grad_norm": 0.620549738407135, "learning_rate": 2.3415221965256805e-06, "loss": 0.0414, "step": 90710 }, { "grad_norm": 0.3760809898376465, "learning_rate": 2.3365241073716237e-06, "loss": 0.023, "step": 90720 }, { "grad_norm": 0.339871346950531, "learning_rate": 2.3315312306236183e-06, "loss": 0.0278, "step": 90730 }, { "grad_norm": 0.32411423325538635, "learning_rate": 2.326543566827699e-06, "loss": 0.0306, "step": 90740 }, { "grad_norm": 0.4331545829772949, "learning_rate": 2.3215611165292862e-06, "loss": 0.0291, "step": 90750 }, { "grad_norm": 0.3351925313472748, "learning_rate": 2.3165838802732765e-06, "loss": 0.0237, "step": 90760 }, { "grad_norm": 0.35403773188591003, "learning_rate": 2.311611858603957e-06, "loss": 0.0279, "step": 90770 }, { "grad_norm": 0.6002885699272156, "learning_rate": 2.3066450520650763e-06, "loss": 0.0316, "step": 90780 }, { "grad_norm": 0.2643463909626007, "learning_rate": 2.3016834611997773e-06, "loss": 0.0349, "step": 90790 }, { "grad_norm": 0.3000139594078064, "learning_rate": 2.296727086550665e-06, "loss": 0.0283, "step": 90800 }, { "grad_norm": 0.4894501566886902, "learning_rate": 2.2917759286597507e-06, "loss": 0.0291, "step": 90810 }, { "grad_norm": 0.373782753944397, "learning_rate": 2.2868299880685006e-06, "loss": 0.0341, "step": 90820 }, { "grad_norm": 0.7920611500740051, "learning_rate": 2.2818892653177827e-06, "loss": 0.0238, "step": 90830 }, { "grad_norm": 0.39055147767066956, "learning_rate": 2.2769537609479196e-06, "loss": 0.0219, "step": 90840 }, { "grad_norm": 0.3913981318473816, "learning_rate": 2.2720234754986357e-06, "loss": 0.0244, "step": 90850 }, { "grad_norm": 0.2108776718378067, "learning_rate": 2.2670984095091108e-06, "loss": 0.0243, "step": 90860 }, { "grad_norm": 0.44260987639427185, "learning_rate": 2.262178563517936e-06, "loss": 0.0321, "step": 90870 }, { "grad_norm": 0.3490503132343292, "learning_rate": 2.257263938063137e-06, "loss": 0.0284, "step": 90880 }, { "grad_norm": 1.2073537111282349, "learning_rate": 2.2523545336821715e-06, "loss": 0.0363, "step": 90890 }, { "grad_norm": 0.2678612470626831, "learning_rate": 2.247450350911939e-06, "loss": 0.0286, "step": 90900 }, { "grad_norm": 0.3148072063922882, "learning_rate": 2.2425513902887328e-06, "loss": 0.0215, "step": 90910 }, { "grad_norm": 0.3950432240962982, "learning_rate": 2.2376576523483015e-06, "loss": 0.0191, "step": 90920 }, { "grad_norm": 0.5724196434020996, "learning_rate": 2.232769137625829e-06, "loss": 0.0305, "step": 90930 }, { "grad_norm": 0.28093886375427246, "learning_rate": 2.227885846655903e-06, "loss": 0.0376, "step": 90940 }, { "grad_norm": 1.4249022006988525, "learning_rate": 2.2230077799725523e-06, "loss": 0.0312, "step": 90950 }, { "grad_norm": 0.19754987955093384, "learning_rate": 2.2181349381092396e-06, "loss": 0.0227, "step": 90960 }, { "grad_norm": 0.4835871458053589, "learning_rate": 2.2132673215988552e-06, "loss": 0.0253, "step": 90970 }, { "grad_norm": 0.41035789251327515, "learning_rate": 2.2084049309737065e-06, "loss": 0.0277, "step": 90980 }, { "grad_norm": 0.4895707368850708, "learning_rate": 2.2035477667655513e-06, "loss": 0.0214, "step": 90990 }, { "grad_norm": 0.27813151478767395, "learning_rate": 2.1986958295055315e-06, "loss": 0.0262, "step": 91000 }, { "grad_norm": 0.6595102548599243, "learning_rate": 2.193849119724284e-06, "loss": 0.0345, "step": 91010 }, { "grad_norm": 0.42379212379455566, "learning_rate": 2.1890076379518175e-06, "loss": 0.0253, "step": 91020 }, { "grad_norm": 0.7050286531448364, "learning_rate": 2.1841713847175925e-06, "loss": 0.0242, "step": 91030 }, { "grad_norm": 0.281444787979126, "learning_rate": 2.1793403605504968e-06, "loss": 0.0244, "step": 91040 }, { "grad_norm": 0.3344489336013794, "learning_rate": 2.1745145659788413e-06, "loss": 0.0327, "step": 91050 }, { "grad_norm": 0.9663748145103455, "learning_rate": 2.1696940015303645e-06, "loss": 0.0305, "step": 91060 }, { "grad_norm": 1.0411587953567505, "learning_rate": 2.1648786677322397e-06, "loss": 0.0356, "step": 91070 }, { "grad_norm": 0.27771487832069397, "learning_rate": 2.1600685651110673e-06, "loss": 0.026, "step": 91080 }, { "grad_norm": 0.3653985559940338, "learning_rate": 2.1552636941928717e-06, "loss": 0.0267, "step": 91090 }, { "grad_norm": 1.3389053344726562, "learning_rate": 2.1504640555031043e-06, "loss": 0.0247, "step": 91100 }, { "grad_norm": 0.5519673824310303, "learning_rate": 2.1456696495666506e-06, "loss": 0.0267, "step": 91110 }, { "grad_norm": 0.39176425337791443, "learning_rate": 2.1408804769078084e-06, "loss": 0.0267, "step": 91120 }, { "grad_norm": 0.8903425335884094, "learning_rate": 2.136096538050325e-06, "loss": 0.0277, "step": 91130 }, { "grad_norm": 0.41337552666664124, "learning_rate": 2.13131783351736e-06, "loss": 0.0323, "step": 91140 }, { "grad_norm": 0.3448583483695984, "learning_rate": 2.126544363831512e-06, "loss": 0.028, "step": 91150 }, { "grad_norm": 0.32223019003868103, "learning_rate": 2.121776129514791e-06, "loss": 0.0185, "step": 91160 }, { "grad_norm": 0.22156725823879242, "learning_rate": 2.1170131310886523e-06, "loss": 0.0221, "step": 91170 }, { "grad_norm": 0.4532339870929718, "learning_rate": 2.112255369073968e-06, "loss": 0.0286, "step": 91180 }, { "grad_norm": 0.4307742714881897, "learning_rate": 2.1075028439910328e-06, "loss": 0.0237, "step": 91190 }, { "grad_norm": 0.6015547513961792, "learning_rate": 2.1027555563595813e-06, "loss": 0.0307, "step": 91200 }, { "grad_norm": 0.48261967301368713, "learning_rate": 2.0980135066987706e-06, "loss": 0.0231, "step": 91210 }, { "grad_norm": 0.36741235852241516, "learning_rate": 2.093276695527191e-06, "loss": 0.0384, "step": 91220 }, { "grad_norm": 0.35847610235214233, "learning_rate": 2.0885451233628396e-06, "loss": 0.0288, "step": 91230 }, { "grad_norm": 0.22126661241054535, "learning_rate": 2.0838187907231633e-06, "loss": 0.028, "step": 91240 }, { "grad_norm": 0.21637463569641113, "learning_rate": 2.0790976981250153e-06, "loss": 0.0261, "step": 91250 }, { "grad_norm": 0.41821739077568054, "learning_rate": 2.0743818460847e-06, "loss": 0.0241, "step": 91260 }, { "grad_norm": 0.39035192131996155, "learning_rate": 2.069671235117926e-06, "loss": 0.0335, "step": 91270 }, { "grad_norm": 0.5574941635131836, "learning_rate": 2.0649658657398484e-06, "loss": 0.0257, "step": 91280 }, { "grad_norm": 0.2718367874622345, "learning_rate": 2.060265738465034e-06, "loss": 0.0216, "step": 91290 }, { "grad_norm": 0.39916226267814636, "learning_rate": 2.055570853807487e-06, "loss": 0.0254, "step": 91300 }, { "grad_norm": 0.43350711464881897, "learning_rate": 2.0508812122806196e-06, "loss": 0.0218, "step": 91310 }, { "grad_norm": 0.40989673137664795, "learning_rate": 2.046196814397294e-06, "loss": 0.026, "step": 91320 }, { "grad_norm": 0.3290049135684967, "learning_rate": 2.0415176606697896e-06, "loss": 0.0233, "step": 91330 }, { "grad_norm": 0.44652265310287476, "learning_rate": 2.0368437516098126e-06, "loss": 0.0256, "step": 91340 }, { "grad_norm": 0.4239092767238617, "learning_rate": 2.032175087728483e-06, "loss": 0.0271, "step": 91350 }, { "grad_norm": 0.49230796098709106, "learning_rate": 2.027511669536375e-06, "loss": 0.0292, "step": 91360 }, { "grad_norm": 0.46180611848831177, "learning_rate": 2.0228534975434643e-06, "loss": 0.0314, "step": 91370 }, { "grad_norm": 0.325155109167099, "learning_rate": 2.0182005722591548e-06, "loss": 0.0223, "step": 91380 }, { "grad_norm": 0.3978234529495239, "learning_rate": 2.0135528941922944e-06, "loss": 0.0264, "step": 91390 }, { "grad_norm": 0.3271626830101013, "learning_rate": 2.008910463851149e-06, "loss": 0.0238, "step": 91400 }, { "grad_norm": 0.39167487621307373, "learning_rate": 2.0042732817433963e-06, "loss": 0.0228, "step": 91410 }, { "grad_norm": 0.3974152207374573, "learning_rate": 1.999641348376158e-06, "loss": 0.0225, "step": 91420 }, { "grad_norm": 0.6166508197784424, "learning_rate": 1.9950146642559787e-06, "loss": 0.028, "step": 91430 }, { "grad_norm": 0.2758197486400604, "learning_rate": 1.990393229888815e-06, "loss": 0.0201, "step": 91440 }, { "grad_norm": 0.3585439622402191, "learning_rate": 1.9857770457800684e-06, "loss": 0.0254, "step": 91450 }, { "grad_norm": 0.24907627701759338, "learning_rate": 1.9811661124345517e-06, "loss": 0.0243, "step": 91460 }, { "grad_norm": 0.3837602138519287, "learning_rate": 1.976560430356522e-06, "loss": 0.0167, "step": 91470 }, { "grad_norm": 0.5762721300125122, "learning_rate": 1.9719600000496383e-06, "loss": 0.0271, "step": 91480 }, { "grad_norm": 0.3224914073944092, "learning_rate": 1.967364822017004e-06, "loss": 0.0301, "step": 91490 }, { "grad_norm": 0.4045662581920624, "learning_rate": 1.962774896761127e-06, "loss": 0.0234, "step": 91500 }, { "grad_norm": 0.20608767867088318, "learning_rate": 1.958190224783968e-06, "loss": 0.02, "step": 91510 }, { "grad_norm": 0.26276835799217224, "learning_rate": 1.9536108065868975e-06, "loss": 0.0264, "step": 91520 }, { "grad_norm": 0.26272138953208923, "learning_rate": 1.949036642670715e-06, "loss": 0.0264, "step": 91530 }, { "grad_norm": 0.448070764541626, "learning_rate": 1.9444677335356377e-06, "loss": 0.0288, "step": 91540 }, { "grad_norm": 0.740105390548706, "learning_rate": 1.939904079681326e-06, "loss": 0.0378, "step": 91550 }, { "grad_norm": 0.33046653866767883, "learning_rate": 1.935345681606837e-06, "loss": 0.0245, "step": 91560 }, { "grad_norm": 0.34999552369117737, "learning_rate": 1.930792539810694e-06, "loss": 0.024, "step": 91570 }, { "grad_norm": 0.7611908316612244, "learning_rate": 1.9262446547907877e-06, "loss": 0.0225, "step": 91580 }, { "grad_norm": 0.2919017970561981, "learning_rate": 1.921702027044503e-06, "loss": 0.0281, "step": 91590 }, { "grad_norm": 0.42649000883102417, "learning_rate": 1.917164657068593e-06, "loss": 0.0259, "step": 91600 }, { "grad_norm": 0.3522695302963257, "learning_rate": 1.912632545359272e-06, "loss": 0.032, "step": 91610 }, { "grad_norm": 0.20413044095039368, "learning_rate": 1.9081056924121488e-06, "loss": 0.0258, "step": 91620 }, { "grad_norm": 0.28906628489494324, "learning_rate": 1.9035840987222887e-06, "loss": 0.0297, "step": 91630 }, { "grad_norm": 0.21218544244766235, "learning_rate": 1.8990677647841515e-06, "loss": 0.0288, "step": 91640 }, { "grad_norm": 0.36151036620140076, "learning_rate": 1.8945566910916534e-06, "loss": 0.0399, "step": 91650 }, { "grad_norm": 0.3364596962928772, "learning_rate": 1.8900508781381054e-06, "loss": 0.0241, "step": 91660 }, { "grad_norm": 0.3534305691719055, "learning_rate": 1.8855503264162633e-06, "loss": 0.0185, "step": 91670 }, { "grad_norm": 0.34081268310546875, "learning_rate": 1.8810550364182998e-06, "loss": 0.0238, "step": 91680 }, { "grad_norm": 0.2738969624042511, "learning_rate": 1.8765650086358112e-06, "loss": 0.0236, "step": 91690 }, { "grad_norm": 0.4085618555545807, "learning_rate": 1.8720802435598151e-06, "loss": 0.0261, "step": 91700 }, { "grad_norm": 0.21009500324726105, "learning_rate": 1.8676007416807694e-06, "loss": 0.0223, "step": 91710 }, { "grad_norm": 0.32455572485923767, "learning_rate": 1.8631265034885437e-06, "loss": 0.0289, "step": 91720 }, { "grad_norm": 0.33478161692619324, "learning_rate": 1.858657529472424e-06, "loss": 0.0205, "step": 91730 }, { "grad_norm": 0.3605532646179199, "learning_rate": 1.8541938201211473e-06, "loss": 0.0182, "step": 91740 }, { "grad_norm": 0.3224130868911743, "learning_rate": 1.8497353759228398e-06, "loss": 0.0193, "step": 91750 }, { "grad_norm": 0.41493064165115356, "learning_rate": 1.8452821973650835e-06, "loss": 0.0255, "step": 91760 }, { "grad_norm": 0.2135973423719406, "learning_rate": 1.840834284934856e-06, "loss": 0.0239, "step": 91770 }, { "grad_norm": 0.30579355359077454, "learning_rate": 1.8363916391185954e-06, "loss": 0.0279, "step": 91780 }, { "grad_norm": 0.27964162826538086, "learning_rate": 1.8319542604021245e-06, "loss": 0.0255, "step": 91790 }, { "grad_norm": 0.42118147015571594, "learning_rate": 1.8275221492707273e-06, "loss": 0.026, "step": 91800 }, { "grad_norm": 0.4649750590324402, "learning_rate": 1.8230953062090717e-06, "loss": 0.0303, "step": 91810 }, { "grad_norm": 0.3983708918094635, "learning_rate": 1.8186737317012926e-06, "loss": 0.0202, "step": 91820 }, { "grad_norm": 0.3360854685306549, "learning_rate": 1.8142574262308977e-06, "loss": 0.0277, "step": 91830 }, { "grad_norm": 0.2641827464103699, "learning_rate": 1.8098463902808727e-06, "loss": 0.0223, "step": 91840 }, { "grad_norm": 0.6149513721466064, "learning_rate": 1.805440624333593e-06, "loss": 0.0242, "step": 91850 }, { "grad_norm": 0.27814415097236633, "learning_rate": 1.8010401288708734e-06, "loss": 0.0304, "step": 91860 }, { "grad_norm": 0.27542540431022644, "learning_rate": 1.7966449043739343e-06, "loss": 0.0225, "step": 91870 }, { "grad_norm": 0.30839815735816956, "learning_rate": 1.79225495132343e-06, "loss": 0.0209, "step": 91880 }, { "grad_norm": 0.5088947415351868, "learning_rate": 1.7878702701994487e-06, "loss": 0.0251, "step": 91890 }, { "grad_norm": 0.42839765548706055, "learning_rate": 1.783490861481496e-06, "loss": 0.0209, "step": 91900 }, { "grad_norm": 0.28070399165153503, "learning_rate": 1.7791167256484832e-06, "loss": 0.0219, "step": 91910 }, { "grad_norm": 0.26997578144073486, "learning_rate": 1.774747863178766e-06, "loss": 0.0271, "step": 91920 }, { "grad_norm": 0.27271732687950134, "learning_rate": 1.7703842745501186e-06, "loss": 0.0168, "step": 91930 }, { "grad_norm": 0.48409396409988403, "learning_rate": 1.7660259602397366e-06, "loss": 0.0286, "step": 91940 }, { "grad_norm": 0.232802152633667, "learning_rate": 1.7616729207242388e-06, "loss": 0.0308, "step": 91950 }, { "grad_norm": 0.6082252264022827, "learning_rate": 1.7573251564796556e-06, "loss": 0.0293, "step": 91960 }, { "grad_norm": 0.4482095539569855, "learning_rate": 1.7529826679814676e-06, "loss": 0.028, "step": 91970 }, { "grad_norm": 0.5163283348083496, "learning_rate": 1.7486454557045562e-06, "loss": 0.025, "step": 91980 }, { "grad_norm": 0.33774784207344055, "learning_rate": 1.7443135201232364e-06, "loss": 0.019, "step": 91990 }, { "grad_norm": 0.707669734954834, "learning_rate": 1.7399868617112347e-06, "loss": 0.0349, "step": 92000 }, { "grad_norm": 0.699016809463501, "learning_rate": 1.7356654809417171e-06, "loss": 0.0231, "step": 92010 }, { "grad_norm": 0.2828196585178375, "learning_rate": 1.7313493782872448e-06, "loss": 0.0241, "step": 92020 }, { "grad_norm": 0.3558337390422821, "learning_rate": 1.7270385542198453e-06, "loss": 0.0235, "step": 92030 }, { "grad_norm": 0.5619684457778931, "learning_rate": 1.7227330092109307e-06, "loss": 0.0232, "step": 92040 }, { "grad_norm": 0.2907407581806183, "learning_rate": 1.718432743731352e-06, "loss": 0.0244, "step": 92050 }, { "grad_norm": 0.3540368974208832, "learning_rate": 1.7141377582513718e-06, "loss": 0.028, "step": 92060 }, { "grad_norm": 0.4005925953388214, "learning_rate": 1.709848053240698e-06, "loss": 0.0253, "step": 92070 }, { "grad_norm": 0.4193316698074341, "learning_rate": 1.7055636291684328e-06, "loss": 0.0164, "step": 92080 }, { "grad_norm": 0.32362791895866394, "learning_rate": 1.7012844865031185e-06, "loss": 0.0229, "step": 92090 }, { "grad_norm": 0.28525376319885254, "learning_rate": 1.6970106257127138e-06, "loss": 0.0294, "step": 92100 }, { "grad_norm": 0.34612584114074707, "learning_rate": 1.6927420472646117e-06, "loss": 0.0228, "step": 92110 }, { "grad_norm": 0.31839659810066223, "learning_rate": 1.6884787516256106e-06, "loss": 0.027, "step": 92120 }, { "grad_norm": 0.580487072467804, "learning_rate": 1.6842207392619325e-06, "loss": 0.0231, "step": 92130 }, { "grad_norm": 0.33601775765419006, "learning_rate": 1.679968010639238e-06, "loss": 0.0288, "step": 92140 }, { "grad_norm": 0.46837717294692993, "learning_rate": 1.6757205662225829e-06, "loss": 0.0274, "step": 92150 }, { "grad_norm": 0.4155714213848114, "learning_rate": 1.6714784064764844e-06, "loss": 0.0247, "step": 92160 }, { "grad_norm": 0.224973663687706, "learning_rate": 1.6672415318648383e-06, "loss": 0.0232, "step": 92170 }, { "grad_norm": 0.6118292808532715, "learning_rate": 1.6630099428510016e-06, "loss": 0.0352, "step": 92180 }, { "grad_norm": 0.436266154050827, "learning_rate": 1.6587836398977153e-06, "loss": 0.0273, "step": 92190 }, { "grad_norm": 0.20472444593906403, "learning_rate": 1.6545626234671762e-06, "loss": 0.0243, "step": 92200 }, { "grad_norm": 0.4509761333465576, "learning_rate": 1.650346894020971e-06, "loss": 0.027, "step": 92210 }, { "grad_norm": 0.4112194776535034, "learning_rate": 1.6461364520201472e-06, "loss": 0.0262, "step": 92220 }, { "grad_norm": 0.5213640332221985, "learning_rate": 1.6419312979251366e-06, "loss": 0.0386, "step": 92230 }, { "grad_norm": 0.39468157291412354, "learning_rate": 1.6377314321958214e-06, "loss": 0.026, "step": 92240 }, { "grad_norm": 0.28628554940223694, "learning_rate": 1.6335368552914787e-06, "loss": 0.0243, "step": 92250 }, { "grad_norm": 0.29885345697402954, "learning_rate": 1.6293475676708358e-06, "loss": 0.0243, "step": 92260 }, { "grad_norm": 0.4501666724681854, "learning_rate": 1.6251635697920097e-06, "loss": 0.0289, "step": 92270 }, { "grad_norm": 0.5675349831581116, "learning_rate": 1.6209848621125624e-06, "loss": 0.0307, "step": 92280 }, { "grad_norm": 0.4942682683467865, "learning_rate": 1.6168114450894778e-06, "loss": 0.0286, "step": 92290 }, { "grad_norm": 0.27258816361427307, "learning_rate": 1.612643319179158e-06, "loss": 0.0262, "step": 92300 }, { "grad_norm": 0.42870455980300903, "learning_rate": 1.6084804848374046e-06, "loss": 0.0299, "step": 92310 }, { "grad_norm": 0.21577270328998566, "learning_rate": 1.6043229425194817e-06, "loss": 0.0257, "step": 92320 }, { "grad_norm": 0.2504420280456543, "learning_rate": 1.6001706926800308e-06, "loss": 0.0232, "step": 92330 }, { "grad_norm": 0.3300975561141968, "learning_rate": 1.5960237357731444e-06, "loss": 0.0197, "step": 92340 }, { "grad_norm": 0.46295061707496643, "learning_rate": 1.5918820722523264e-06, "loss": 0.0305, "step": 92350 }, { "grad_norm": 0.30238261818885803, "learning_rate": 1.587745702570509e-06, "loss": 0.0291, "step": 92360 }, { "grad_norm": 0.26930001378059387, "learning_rate": 1.5836146271800356e-06, "loss": 0.0286, "step": 92370 }, { "grad_norm": 0.43067023158073425, "learning_rate": 1.5794888465326728e-06, "loss": 0.0234, "step": 92380 }, { "grad_norm": 0.23176665604114532, "learning_rate": 1.5753683610796155e-06, "loss": 0.0281, "step": 92390 }, { "grad_norm": 0.2865979075431824, "learning_rate": 1.5712531712714528e-06, "loss": 0.0275, "step": 92400 }, { "grad_norm": 0.25953471660614014, "learning_rate": 1.5671432775582474e-06, "loss": 0.0273, "step": 92410 }, { "grad_norm": 0.28133153915405273, "learning_rate": 1.563038680389428e-06, "loss": 0.0291, "step": 92420 }, { "grad_norm": 0.5763248801231384, "learning_rate": 1.5589393802138808e-06, "loss": 0.0338, "step": 92430 }, { "grad_norm": 0.34102311730384827, "learning_rate": 1.554845377479891e-06, "loss": 0.0302, "step": 92440 }, { "grad_norm": 0.30863675475120544, "learning_rate": 1.550756672635173e-06, "loss": 0.0299, "step": 92450 }, { "grad_norm": 0.4162975251674652, "learning_rate": 1.5466732661268635e-06, "loss": 0.0264, "step": 92460 }, { "grad_norm": 0.5364974737167358, "learning_rate": 1.5425951584015163e-06, "loss": 0.0301, "step": 92470 }, { "grad_norm": 0.21931099891662598, "learning_rate": 1.5385223499051026e-06, "loss": 0.0279, "step": 92480 }, { "grad_norm": 0.451036661863327, "learning_rate": 1.5344548410830328e-06, "loss": 0.0203, "step": 92490 }, { "grad_norm": 0.3285973072052002, "learning_rate": 1.5303926323801121e-06, "loss": 0.0226, "step": 92500 }, { "grad_norm": 0.30945754051208496, "learning_rate": 1.5263357242405795e-06, "loss": 0.0176, "step": 92510 }, { "grad_norm": 0.342035174369812, "learning_rate": 1.5222841171080915e-06, "loss": 0.0295, "step": 92520 }, { "grad_norm": 0.3328424096107483, "learning_rate": 1.5182378114257268e-06, "loss": 0.0248, "step": 92530 }, { "grad_norm": 0.37576451897621155, "learning_rate": 1.5141968076359869e-06, "loss": 0.0283, "step": 92540 }, { "grad_norm": 0.43134716153144836, "learning_rate": 1.5101611061807908e-06, "loss": 0.0258, "step": 92550 }, { "grad_norm": 0.2765932083129883, "learning_rate": 1.506130707501463e-06, "loss": 0.023, "step": 92560 }, { "grad_norm": 0.6150661706924438, "learning_rate": 1.5021056120387734e-06, "loss": 0.0336, "step": 92570 }, { "grad_norm": 0.3905757665634155, "learning_rate": 1.4980858202329084e-06, "loss": 0.0258, "step": 92580 }, { "grad_norm": 0.2999533414840698, "learning_rate": 1.4940713325234445e-06, "loss": 0.0286, "step": 92590 }, { "grad_norm": 0.15365943312644958, "learning_rate": 1.4900621493494137e-06, "loss": 0.0219, "step": 92600 }, { "grad_norm": 0.4572235643863678, "learning_rate": 1.4860582711492544e-06, "loss": 0.0255, "step": 92610 }, { "grad_norm": 0.401927649974823, "learning_rate": 1.4820596983608325e-06, "loss": 0.0254, "step": 92620 }, { "grad_norm": 0.29179462790489197, "learning_rate": 1.4780664314214044e-06, "loss": 0.0238, "step": 92630 }, { "grad_norm": 0.4380654990673065, "learning_rate": 1.474078470767687e-06, "loss": 0.0285, "step": 92640 }, { "grad_norm": 0.23496639728546143, "learning_rate": 1.470095816835787e-06, "loss": 0.0278, "step": 92650 }, { "grad_norm": 0.471505731344223, "learning_rate": 1.4661184700612395e-06, "loss": 0.0292, "step": 92660 }, { "grad_norm": 0.22997896373271942, "learning_rate": 1.4621464308790133e-06, "loss": 0.0246, "step": 92670 }, { "grad_norm": 0.3515903949737549, "learning_rate": 1.4581796997234775e-06, "loss": 0.0281, "step": 92680 }, { "grad_norm": 0.2052280455827713, "learning_rate": 1.4542182770284297e-06, "loss": 0.0241, "step": 92690 }, { "grad_norm": 0.3737287223339081, "learning_rate": 1.450262163227084e-06, "loss": 0.0246, "step": 92700 }, { "grad_norm": 0.27862322330474854, "learning_rate": 1.4463113587520671e-06, "loss": 0.0301, "step": 92710 }, { "grad_norm": 0.20201259851455688, "learning_rate": 1.4423658640354443e-06, "loss": 0.0274, "step": 92720 }, { "grad_norm": 0.6196247935295105, "learning_rate": 1.438425679508687e-06, "loss": 0.0288, "step": 92730 }, { "grad_norm": 0.46061283349990845, "learning_rate": 1.4344908056026895e-06, "loss": 0.0251, "step": 92740 }, { "grad_norm": 0.19067955017089844, "learning_rate": 1.4305612427477521e-06, "loss": 0.0201, "step": 92750 }, { "grad_norm": 0.2851134240627289, "learning_rate": 1.42663699137362e-06, "loss": 0.0195, "step": 92760 }, { "grad_norm": 0.27701881527900696, "learning_rate": 1.4227180519094386e-06, "loss": 0.0332, "step": 92770 }, { "grad_norm": 0.4294099509716034, "learning_rate": 1.4188044247837706e-06, "loss": 0.0235, "step": 92780 }, { "grad_norm": 0.3444995582103729, "learning_rate": 1.4148961104246128e-06, "loss": 0.026, "step": 92790 }, { "grad_norm": 0.6565189361572266, "learning_rate": 1.4109931092593731e-06, "loss": 0.0179, "step": 92800 }, { "grad_norm": 0.5337018370628357, "learning_rate": 1.4070954217148658e-06, "loss": 0.0314, "step": 92810 }, { "grad_norm": 0.43501120805740356, "learning_rate": 1.4032030482173497e-06, "loss": 0.035, "step": 92820 }, { "grad_norm": 0.32628482580184937, "learning_rate": 1.3993159891924901e-06, "loss": 0.0271, "step": 92830 }, { "grad_norm": 0.46516311168670654, "learning_rate": 1.3954342450653524e-06, "loss": 0.0343, "step": 92840 }, { "grad_norm": 0.3749347925186157, "learning_rate": 1.3915578162604582e-06, "loss": 0.0258, "step": 92850 }, { "grad_norm": 0.41583019495010376, "learning_rate": 1.3876867032017126e-06, "loss": 0.0328, "step": 92860 }, { "grad_norm": 0.46874016523361206, "learning_rate": 1.383820906312472e-06, "loss": 0.0266, "step": 92870 }, { "grad_norm": 0.2737843692302704, "learning_rate": 1.3799604260154752e-06, "loss": 0.021, "step": 92880 }, { "grad_norm": 0.48103541135787964, "learning_rate": 1.376105262732913e-06, "loss": 0.0339, "step": 92890 }, { "grad_norm": 0.27976298332214355, "learning_rate": 1.37225541688637e-06, "loss": 0.035, "step": 92900 }, { "grad_norm": 0.33970192074775696, "learning_rate": 1.3684108888968594e-06, "loss": 0.03, "step": 92910 }, { "grad_norm": 0.3433670401573181, "learning_rate": 1.364571679184823e-06, "loss": 0.0284, "step": 92920 }, { "grad_norm": 0.5305892825126648, "learning_rate": 1.3607377881701077e-06, "loss": 0.0353, "step": 92930 }, { "grad_norm": 0.30505990982055664, "learning_rate": 1.3569092162719787e-06, "loss": 0.0238, "step": 92940 }, { "grad_norm": 0.325191855430603, "learning_rate": 1.353085963909123e-06, "loss": 0.0242, "step": 92950 }, { "grad_norm": 0.2735266089439392, "learning_rate": 1.3492680314996454e-06, "loss": 0.0243, "step": 92960 }, { "grad_norm": 0.44781631231307983, "learning_rate": 1.345455419461078e-06, "loss": 0.0284, "step": 92970 }, { "grad_norm": 0.20801450312137604, "learning_rate": 1.3416481282103431e-06, "loss": 0.0261, "step": 92980 }, { "grad_norm": 0.33571988344192505, "learning_rate": 1.337846158163819e-06, "loss": 0.0301, "step": 92990 }, { "grad_norm": 0.25114813446998596, "learning_rate": 1.334049509737273e-06, "loss": 0.0274, "step": 93000 }, { "grad_norm": 0.25317659974098206, "learning_rate": 1.3302581833459116e-06, "loss": 0.024, "step": 93010 }, { "grad_norm": 0.24613720178604126, "learning_rate": 1.3264721794043366e-06, "loss": 0.0245, "step": 93020 }, { "grad_norm": 0.4846266210079193, "learning_rate": 1.3226914983265836e-06, "loss": 0.0345, "step": 93030 }, { "grad_norm": 0.4920099973678589, "learning_rate": 1.3189161405260997e-06, "loss": 0.0325, "step": 93040 }, { "grad_norm": 0.4105075001716614, "learning_rate": 1.3151461064157656e-06, "loss": 0.0196, "step": 93050 }, { "grad_norm": 0.6416963934898376, "learning_rate": 1.3113813964078459e-06, "loss": 0.0276, "step": 93060 }, { "grad_norm": 0.2641638517379761, "learning_rate": 1.3076220109140614e-06, "loss": 0.0279, "step": 93070 }, { "grad_norm": 0.3804064989089966, "learning_rate": 1.3038679503455275e-06, "loss": 0.0254, "step": 93080 }, { "grad_norm": 0.387282133102417, "learning_rate": 1.3001192151127773e-06, "loss": 0.0274, "step": 93090 }, { "grad_norm": 0.2227860391139984, "learning_rate": 1.2963758056257657e-06, "loss": 0.0251, "step": 93100 }, { "grad_norm": 0.3098656237125397, "learning_rate": 1.2926377222938768e-06, "loss": 0.0181, "step": 93110 }, { "grad_norm": 0.6202396750450134, "learning_rate": 1.2889049655259e-06, "loss": 0.0337, "step": 93120 }, { "grad_norm": 0.40851789712905884, "learning_rate": 1.2851775357300312e-06, "loss": 0.0419, "step": 93130 }, { "grad_norm": 0.34734442830085754, "learning_rate": 1.281455433313916e-06, "loss": 0.0287, "step": 93140 }, { "grad_norm": 0.3180912435054779, "learning_rate": 1.2777386586845796e-06, "loss": 0.0217, "step": 93150 }, { "grad_norm": 0.4607674777507782, "learning_rate": 1.2740272122484964e-06, "loss": 0.0344, "step": 93160 }, { "grad_norm": 0.3721303939819336, "learning_rate": 1.2703210944115308e-06, "loss": 0.0245, "step": 93170 }, { "grad_norm": 0.48921364545822144, "learning_rate": 1.2666203055789916e-06, "loss": 0.0212, "step": 93180 }, { "grad_norm": 0.37943217158317566, "learning_rate": 1.2629248461555832e-06, "loss": 0.0262, "step": 93190 }, { "grad_norm": 0.3227826952934265, "learning_rate": 1.2592347165454431e-06, "loss": 0.0239, "step": 93200 }, { "grad_norm": 0.3854634165763855, "learning_rate": 1.2555499171521102e-06, "loss": 0.0336, "step": 93210 }, { "grad_norm": 0.19612354040145874, "learning_rate": 1.251870448378556e-06, "loss": 0.0238, "step": 93220 }, { "grad_norm": 0.3311674892902374, "learning_rate": 1.2481963106271478e-06, "loss": 0.0312, "step": 93230 }, { "grad_norm": 0.3433874845504761, "learning_rate": 1.2445275042997084e-06, "loss": 0.029, "step": 93240 }, { "grad_norm": 0.29131007194519043, "learning_rate": 1.2408640297974283e-06, "loss": 0.0243, "step": 93250 }, { "grad_norm": 0.3370644152164459, "learning_rate": 1.2372058875209535e-06, "loss": 0.0226, "step": 93260 }, { "grad_norm": 0.24680423736572266, "learning_rate": 1.2335530778703252e-06, "loss": 0.0233, "step": 93270 }, { "grad_norm": 0.4036667048931122, "learning_rate": 1.2299056012450182e-06, "loss": 0.0262, "step": 93280 }, { "grad_norm": 0.42451003193855286, "learning_rate": 1.2262634580439026e-06, "loss": 0.0287, "step": 93290 }, { "grad_norm": 0.5596786141395569, "learning_rate": 1.2226266486652926e-06, "loss": 0.0333, "step": 93300 }, { "grad_norm": 0.36067160964012146, "learning_rate": 1.2189951735068927e-06, "loss": 0.0264, "step": 93310 }, { "grad_norm": 0.8462899327278137, "learning_rate": 1.2153690329658407e-06, "loss": 0.0244, "step": 93320 }, { "grad_norm": 0.34893152117729187, "learning_rate": 1.2117482274386859e-06, "loss": 0.0295, "step": 93330 }, { "grad_norm": 0.3982904255390167, "learning_rate": 1.2081327573213953e-06, "loss": 0.0222, "step": 93340 }, { "grad_norm": 0.509300172328949, "learning_rate": 1.2045226230093465e-06, "loss": 0.0304, "step": 93350 }, { "grad_norm": 0.34967631101608276, "learning_rate": 1.2009178248973296e-06, "loss": 0.0263, "step": 93360 }, { "grad_norm": 0.40795812010765076, "learning_rate": 1.1973183633795849e-06, "loss": 0.0233, "step": 93370 }, { "grad_norm": 0.9556606411933899, "learning_rate": 1.193724238849725e-06, "loss": 0.0502, "step": 93380 }, { "grad_norm": 0.4246804714202881, "learning_rate": 1.190135451700808e-06, "loss": 0.0274, "step": 93390 }, { "grad_norm": 0.9065282344818115, "learning_rate": 1.186552002325292e-06, "loss": 0.0317, "step": 93400 }, { "grad_norm": 0.37567734718322754, "learning_rate": 1.1829738911150578e-06, "loss": 0.0198, "step": 93410 }, { "grad_norm": 0.5221681594848633, "learning_rate": 1.1794011184613985e-06, "loss": 0.0242, "step": 93420 }, { "grad_norm": 0.4096412658691406, "learning_rate": 1.17583368475504e-06, "loss": 0.0292, "step": 93430 }, { "grad_norm": 0.5414028167724609, "learning_rate": 1.1722715903860982e-06, "loss": 0.0295, "step": 93440 }, { "grad_norm": 0.2906462252140045, "learning_rate": 1.168714835744128e-06, "loss": 0.0259, "step": 93450 }, { "grad_norm": 0.2972148060798645, "learning_rate": 1.165163421218085e-06, "loss": 0.0342, "step": 93460 }, { "grad_norm": 0.5808201432228088, "learning_rate": 1.1616173471963477e-06, "loss": 0.0348, "step": 93470 }, { "grad_norm": 0.4052245616912842, "learning_rate": 1.1580766140667109e-06, "loss": 0.019, "step": 93480 }, { "grad_norm": 0.3110804557800293, "learning_rate": 1.1545412222163764e-06, "loss": 0.0258, "step": 93490 }, { "grad_norm": 0.32091212272644043, "learning_rate": 1.1510111720319795e-06, "loss": 0.0271, "step": 93500 }, { "grad_norm": 0.30177825689315796, "learning_rate": 1.1474864638995663e-06, "loss": 0.0273, "step": 93510 }, { "grad_norm": 0.4748592674732208, "learning_rate": 1.1439670982045735e-06, "loss": 0.0236, "step": 93520 }, { "grad_norm": 0.7005595564842224, "learning_rate": 1.1404530753318877e-06, "loss": 0.0356, "step": 93530 }, { "grad_norm": 0.1417529582977295, "learning_rate": 1.136944395665801e-06, "loss": 0.0325, "step": 93540 }, { "grad_norm": 0.5835989117622375, "learning_rate": 1.1334410595900014e-06, "loss": 0.0309, "step": 93550 }, { "grad_norm": 0.44654640555381775, "learning_rate": 1.129943067487621e-06, "loss": 0.0344, "step": 93560 }, { "grad_norm": 0.4005892872810364, "learning_rate": 1.126450419741193e-06, "loss": 0.0206, "step": 93570 }, { "grad_norm": 0.3980730473995209, "learning_rate": 1.1229631167326671e-06, "loss": 0.0255, "step": 93580 }, { "grad_norm": 0.957216203212738, "learning_rate": 1.1194811588434052e-06, "loss": 0.0234, "step": 93590 }, { "grad_norm": 0.37911421060562134, "learning_rate": 1.1160045464542023e-06, "loss": 0.0258, "step": 93600 }, { "grad_norm": 0.3460009694099426, "learning_rate": 1.1125332799452325e-06, "loss": 0.0369, "step": 93610 }, { "grad_norm": 0.3091280460357666, "learning_rate": 1.109067359696131e-06, "loss": 0.0208, "step": 93620 }, { "grad_norm": 1.2059893608093262, "learning_rate": 1.1056067860859166e-06, "loss": 0.0237, "step": 93630 }, { "grad_norm": 0.206151083111763, "learning_rate": 1.102151559493031e-06, "loss": 0.0242, "step": 93640 }, { "grad_norm": 0.5343543887138367, "learning_rate": 1.0987016802953332e-06, "loss": 0.0292, "step": 93650 }, { "grad_norm": 0.31746307015419006, "learning_rate": 1.0952571488701048e-06, "loss": 0.0286, "step": 93660 }, { "grad_norm": 0.414637953042984, "learning_rate": 1.0918179655940163e-06, "loss": 0.0321, "step": 93670 }, { "grad_norm": 0.5608611106872559, "learning_rate": 1.088384130843184e-06, "loss": 0.0256, "step": 93680 }, { "grad_norm": 0.48239246010780334, "learning_rate": 1.0849556449931296e-06, "loss": 0.0293, "step": 93690 }, { "grad_norm": 0.3333298861980438, "learning_rate": 1.0815325084187866e-06, "loss": 0.0278, "step": 93700 }, { "grad_norm": 0.8469745516777039, "learning_rate": 1.0781147214944943e-06, "loss": 0.0259, "step": 93710 }, { "grad_norm": 0.47120410203933716, "learning_rate": 1.0747022845940313e-06, "loss": 0.0257, "step": 93720 }, { "grad_norm": 0.36119475960731506, "learning_rate": 1.0712951980905605e-06, "loss": 0.0242, "step": 93730 }, { "grad_norm": 0.40423551201820374, "learning_rate": 1.0678934623566839e-06, "loss": 0.0338, "step": 93740 }, { "grad_norm": 0.3555378019809723, "learning_rate": 1.0644970777644091e-06, "loss": 0.0236, "step": 93750 }, { "grad_norm": 0.44273608922958374, "learning_rate": 1.0611060446851673e-06, "loss": 0.0262, "step": 93760 }, { "grad_norm": 0.3873556852340698, "learning_rate": 1.0577203634897836e-06, "loss": 0.0279, "step": 93770 }, { "grad_norm": 0.3602323532104492, "learning_rate": 1.0543400345485178e-06, "loss": 0.0264, "step": 93780 }, { "grad_norm": 0.28812113404273987, "learning_rate": 1.0509650582310404e-06, "loss": 0.0269, "step": 93790 }, { "grad_norm": 0.9775578379631042, "learning_rate": 1.0475954349064287e-06, "loss": 0.029, "step": 93800 }, { "grad_norm": 0.3161856532096863, "learning_rate": 1.0442311649431825e-06, "loss": 0.0252, "step": 93810 }, { "grad_norm": 0.27288618683815, "learning_rate": 1.0408722487092071e-06, "loss": 0.0364, "step": 93820 }, { "grad_norm": 0.26622092723846436, "learning_rate": 1.0375186865718424e-06, "loss": 0.0202, "step": 93830 }, { "grad_norm": 0.32115912437438965, "learning_rate": 1.0341704788978113e-06, "loss": 0.0279, "step": 93840 }, { "grad_norm": 0.39930373430252075, "learning_rate": 1.0308276260532879e-06, "loss": 0.0255, "step": 93850 }, { "grad_norm": 0.4751571714878082, "learning_rate": 1.0274901284038241e-06, "loss": 0.0208, "step": 93860 }, { "grad_norm": 1.214453101158142, "learning_rate": 1.0241579863144114e-06, "loss": 0.0289, "step": 93870 }, { "grad_norm": 0.35136276483535767, "learning_rate": 1.0208312001494413e-06, "loss": 0.0226, "step": 93880 }, { "grad_norm": 0.557377815246582, "learning_rate": 1.0175097702727453e-06, "loss": 0.0244, "step": 93890 }, { "grad_norm": 1.0176982879638672, "learning_rate": 1.0141936970475275e-06, "loss": 0.0257, "step": 93900 }, { "grad_norm": 0.2320384979248047, "learning_rate": 1.0108829808364417e-06, "loss": 0.0268, "step": 93910 }, { "grad_norm": 0.2564489543437958, "learning_rate": 1.0075776220015376e-06, "loss": 0.0267, "step": 93920 }, { "grad_norm": 0.3661787211894989, "learning_rate": 1.0042776209042816e-06, "loss": 0.0263, "step": 93930 }, { "grad_norm": 0.8631666898727417, "learning_rate": 1.000982977905568e-06, "loss": 0.0335, "step": 93940 }, { "grad_norm": 0.5470583438873291, "learning_rate": 9.976936933656867e-07, "loss": 0.0291, "step": 93950 }, { "grad_norm": 0.3409070670604706, "learning_rate": 9.944097676443442e-07, "loss": 0.0334, "step": 93960 }, { "grad_norm": 0.5671562552452087, "learning_rate": 9.911312011006757e-07, "loss": 0.0355, "step": 93970 }, { "grad_norm": 0.5869988799095154, "learning_rate": 9.878579940932109e-07, "loss": 0.026, "step": 93980 }, { "grad_norm": 0.4758791923522949, "learning_rate": 9.84590146979908e-07, "loss": 0.0268, "step": 93990 }, { "grad_norm": 0.19712117314338684, "learning_rate": 9.813276601181365e-07, "loss": 0.024, "step": 94000 }, { "grad_norm": 0.9165584444999695, "learning_rate": 9.780705338646724e-07, "loss": 0.0252, "step": 94010 }, { "grad_norm": 0.2666676342487335, "learning_rate": 9.748187685757083e-07, "loss": 0.0292, "step": 94020 }, { "grad_norm": 0.5997382402420044, "learning_rate": 9.715723646068543e-07, "loss": 0.0284, "step": 94030 }, { "grad_norm": 0.365304172039032, "learning_rate": 9.68331322313143e-07, "loss": 0.0278, "step": 94040 }, { "grad_norm": 0.2796950936317444, "learning_rate": 9.650956420489909e-07, "loss": 0.0281, "step": 94050 }, { "grad_norm": 0.3740481734275818, "learning_rate": 9.618653241682595e-07, "loss": 0.0285, "step": 94060 }, { "grad_norm": 1.2839231491088867, "learning_rate": 9.586403690242053e-07, "loss": 0.0306, "step": 94070 }, { "grad_norm": 0.32505032420158386, "learning_rate": 9.554207769695122e-07, "loss": 0.0241, "step": 94080 }, { "grad_norm": 0.3069664239883423, "learning_rate": 9.522065483562659e-07, "loss": 0.0335, "step": 94090 }, { "grad_norm": 0.41071221232414246, "learning_rate": 9.489976835359681e-07, "loss": 0.0257, "step": 94100 }, { "grad_norm": 0.43601179122924805, "learning_rate": 9.457941828595329e-07, "loss": 0.031, "step": 94110 }, { "grad_norm": 0.48752352595329285, "learning_rate": 9.425960466772965e-07, "loss": 0.0267, "step": 94120 }, { "grad_norm": 0.5461004376411438, "learning_rate": 9.394032753390014e-07, "loss": 0.0324, "step": 94130 }, { "grad_norm": 0.30344727635383606, "learning_rate": 9.362158691938017e-07, "loss": 0.022, "step": 94140 }, { "grad_norm": 0.5059201121330261, "learning_rate": 9.330338285902685e-07, "loss": 0.0232, "step": 94150 }, { "grad_norm": 0.25622814893722534, "learning_rate": 9.298571538763845e-07, "loss": 0.0273, "step": 94160 }, { "grad_norm": 0.4009518623352051, "learning_rate": 9.266858453995441e-07, "loss": 0.0202, "step": 94170 }, { "grad_norm": 0.5746121406555176, "learning_rate": 9.235199035065645e-07, "loss": 0.0212, "step": 94180 }, { "grad_norm": 0.1662023514509201, "learning_rate": 9.203593285436574e-07, "loss": 0.0271, "step": 94190 }, { "grad_norm": 0.42370063066482544, "learning_rate": 9.172041208564686e-07, "loss": 0.0248, "step": 94200 }, { "grad_norm": 0.45531192421913147, "learning_rate": 9.140542807900332e-07, "loss": 0.0226, "step": 94210 }, { "grad_norm": 0.32040125131607056, "learning_rate": 9.109098086888313e-07, "loss": 0.0264, "step": 94220 }, { "grad_norm": 0.28628501296043396, "learning_rate": 9.077707048967265e-07, "loss": 0.0316, "step": 94230 }, { "grad_norm": 0.5059992074966431, "learning_rate": 9.046369697570057e-07, "loss": 0.023, "step": 94240 }, { "grad_norm": 0.27044203877449036, "learning_rate": 9.015086036123721e-07, "loss": 0.0183, "step": 94250 }, { "grad_norm": 0.46988558769226074, "learning_rate": 8.983856068049413e-07, "loss": 0.0258, "step": 94260 }, { "grad_norm": 0.5560125708580017, "learning_rate": 8.952679796762398e-07, "loss": 0.0309, "step": 94270 }, { "grad_norm": 0.2776983380317688, "learning_rate": 8.921557225671951e-07, "loss": 0.0216, "step": 94280 }, { "grad_norm": 0.48534369468688965, "learning_rate": 8.890488358181792e-07, "loss": 0.0304, "step": 94290 }, { "grad_norm": 0.31395256519317627, "learning_rate": 8.859473197689372e-07, "loss": 0.029, "step": 94300 }, { "grad_norm": 0.46844255924224854, "learning_rate": 8.828511747586588e-07, "loss": 0.0244, "step": 94310 }, { "grad_norm": 0.3495234251022339, "learning_rate": 8.797604011259287e-07, "loss": 0.0316, "step": 94320 }, { "grad_norm": 0.39656752347946167, "learning_rate": 8.766749992087487e-07, "loss": 0.0236, "step": 94330 }, { "grad_norm": 0.4135306179523468, "learning_rate": 8.735949693445378e-07, "loss": 0.0213, "step": 94340 }, { "grad_norm": 0.5353272557258606, "learning_rate": 8.705203118701211e-07, "loss": 0.0216, "step": 94350 }, { "grad_norm": 0.2797512710094452, "learning_rate": 8.674510271217295e-07, "loss": 0.0276, "step": 94360 }, { "grad_norm": 1.1492552757263184, "learning_rate": 8.643871154350336e-07, "loss": 0.031, "step": 94370 }, { "grad_norm": 0.276889830827713, "learning_rate": 8.613285771450818e-07, "loss": 0.0255, "step": 94380 }, { "grad_norm": 0.2161533236503601, "learning_rate": 8.582754125863623e-07, "loss": 0.0265, "step": 94390 }, { "grad_norm": 0.30445459485054016, "learning_rate": 8.55227622092758e-07, "loss": 0.0233, "step": 94400 }, { "grad_norm": 0.32051700353622437, "learning_rate": 8.521852059975743e-07, "loss": 0.0206, "step": 94410 }, { "grad_norm": 0.29695212841033936, "learning_rate": 8.491481646335231e-07, "loss": 0.0248, "step": 94420 }, { "grad_norm": 0.41941535472869873, "learning_rate": 8.46116498332733e-07, "loss": 0.0229, "step": 94430 }, { "grad_norm": 0.35003650188446045, "learning_rate": 8.430902074267333e-07, "loss": 0.0326, "step": 94440 }, { "grad_norm": 1.934948205947876, "learning_rate": 8.400692922464925e-07, "loss": 0.031, "step": 94450 }, { "grad_norm": 0.5153554081916809, "learning_rate": 8.370537531223577e-07, "loss": 0.0259, "step": 94460 }, { "grad_norm": 0.4175833761692047, "learning_rate": 8.340435903841092e-07, "loss": 0.0257, "step": 94470 }, { "grad_norm": 0.28425467014312744, "learning_rate": 8.310388043609341e-07, "loss": 0.0228, "step": 94480 }, { "grad_norm": 0.31532078981399536, "learning_rate": 8.280393953814303e-07, "loss": 0.0288, "step": 94490 }, { "grad_norm": 0.5497093200683594, "learning_rate": 8.250453637736077e-07, "loss": 0.0273, "step": 94500 }, { "grad_norm": 0.38006147742271423, "learning_rate": 8.22056709864899e-07, "loss": 0.0226, "step": 94510 }, { "grad_norm": 0.29821282625198364, "learning_rate": 8.190734339821204e-07, "loss": 0.0257, "step": 94520 }, { "grad_norm": 0.46670353412628174, "learning_rate": 8.160955364515333e-07, "loss": 0.0241, "step": 94530 }, { "grad_norm": 0.32811182737350464, "learning_rate": 8.131230175987936e-07, "loss": 0.0207, "step": 94540 }, { "grad_norm": 0.39846888184547424, "learning_rate": 8.101558777489693e-07, "loss": 0.0288, "step": 94550 }, { "grad_norm": 0.6471047401428223, "learning_rate": 8.071941172265396e-07, "loss": 0.0248, "step": 94560 }, { "grad_norm": 0.2898285984992981, "learning_rate": 8.042377363553955e-07, "loss": 0.0277, "step": 94570 }, { "grad_norm": 0.3452385365962982, "learning_rate": 8.01286735458856e-07, "loss": 0.0341, "step": 94580 }, { "grad_norm": 0.30869174003601074, "learning_rate": 7.983411148596298e-07, "loss": 0.0246, "step": 94590 }, { "grad_norm": 0.31662264466285706, "learning_rate": 7.954008748798425e-07, "loss": 0.0273, "step": 94600 }, { "grad_norm": 0.4621870815753937, "learning_rate": 7.924660158410369e-07, "loss": 0.0229, "step": 94610 }, { "grad_norm": 0.6202028393745422, "learning_rate": 7.895365380641728e-07, "loss": 0.0343, "step": 94620 }, { "grad_norm": 0.3793763816356659, "learning_rate": 7.866124418695941e-07, "loss": 0.0272, "step": 94630 }, { "grad_norm": 0.25032567977905273, "learning_rate": 7.83693727577095e-07, "loss": 0.0366, "step": 94640 }, { "grad_norm": 0.3970358967781067, "learning_rate": 7.807803955058535e-07, "loss": 0.0266, "step": 94650 }, { "grad_norm": 0.34594443440437317, "learning_rate": 7.778724459744702e-07, "loss": 0.0202, "step": 94660 }, { "grad_norm": 0.6869308948516846, "learning_rate": 7.749698793009464e-07, "loss": 0.0325, "step": 94670 }, { "grad_norm": 0.36548754572868347, "learning_rate": 7.720726958027169e-07, "loss": 0.027, "step": 94680 }, { "grad_norm": 0.37138110399246216, "learning_rate": 7.691808957965951e-07, "loss": 0.0211, "step": 94690 }, { "grad_norm": 0.3577113747596741, "learning_rate": 7.662944795988336e-07, "loss": 0.0269, "step": 94700 }, { "grad_norm": 0.27377453446388245, "learning_rate": 7.634134475250854e-07, "loss": 0.0224, "step": 94710 }, { "grad_norm": 0.29338932037353516, "learning_rate": 7.605377998904262e-07, "loss": 0.0285, "step": 94720 }, { "grad_norm": 0.40092962980270386, "learning_rate": 7.5766753700931e-07, "loss": 0.0253, "step": 94730 }, { "grad_norm": 0.31469079852104187, "learning_rate": 7.54802659195647e-07, "loss": 0.0221, "step": 94740 }, { "grad_norm": 0.381178617477417, "learning_rate": 7.519431667627253e-07, "loss": 0.0253, "step": 94750 }, { "grad_norm": 0.49371856451034546, "learning_rate": 7.490890600232503e-07, "loss": 0.0324, "step": 94760 }, { "grad_norm": 0.6110184192657471, "learning_rate": 7.462403392893502e-07, "loss": 0.0268, "step": 94770 }, { "grad_norm": 0.3853584825992584, "learning_rate": 7.433970048725536e-07, "loss": 0.0183, "step": 94780 }, { "grad_norm": 0.29178640246391296, "learning_rate": 7.405590570838062e-07, "loss": 0.0255, "step": 94790 }, { "grad_norm": 0.14097945392131805, "learning_rate": 7.377264962334596e-07, "loss": 0.0205, "step": 94800 }, { "grad_norm": 0.3299608528614044, "learning_rate": 7.348993226312773e-07, "loss": 0.0196, "step": 94810 }, { "grad_norm": 0.3501763939857483, "learning_rate": 7.32077536586434e-07, "loss": 0.0271, "step": 94820 }, { "grad_norm": 0.39802148938179016, "learning_rate": 7.292611384075221e-07, "loss": 0.024, "step": 94830 }, { "grad_norm": 0.2493920624256134, "learning_rate": 7.264501284025338e-07, "loss": 0.0305, "step": 94840 }, { "grad_norm": 0.5006265044212341, "learning_rate": 7.236445068788789e-07, "loss": 0.0313, "step": 94850 }, { "grad_norm": 0.2688175439834595, "learning_rate": 7.20844274143373e-07, "loss": 0.0244, "step": 94860 }, { "grad_norm": 0.28927308320999146, "learning_rate": 7.180494305022489e-07, "loss": 0.0275, "step": 94870 }, { "grad_norm": 0.4668547213077545, "learning_rate": 7.152599762611456e-07, "loss": 0.0312, "step": 94880 }, { "grad_norm": 0.3305668830871582, "learning_rate": 7.124759117251079e-07, "loss": 0.05, "step": 94890 }, { "grad_norm": 0.4257667660713196, "learning_rate": 7.096972371986033e-07, "loss": 0.0252, "step": 94900 }, { "grad_norm": 0.49986034631729126, "learning_rate": 7.069239529855054e-07, "loss": 0.0316, "step": 94910 }, { "grad_norm": 0.32243025302886963, "learning_rate": 7.041560593890939e-07, "loss": 0.0211, "step": 94920 }, { "grad_norm": 0.31602540612220764, "learning_rate": 7.013935567120599e-07, "loss": 0.0246, "step": 94930 }, { "grad_norm": 0.36478883028030396, "learning_rate": 6.986364452565009e-07, "loss": 0.018, "step": 94940 }, { "grad_norm": 0.6074734330177307, "learning_rate": 6.95884725323942e-07, "loss": 0.0257, "step": 94950 }, { "grad_norm": 0.25266149640083313, "learning_rate": 6.931383972152982e-07, "loss": 0.0234, "step": 94960 }, { "grad_norm": 0.28904664516448975, "learning_rate": 6.903974612309127e-07, "loss": 0.0231, "step": 94970 }, { "grad_norm": 0.2813645899295807, "learning_rate": 6.876619176705179e-07, "loss": 0.026, "step": 94980 }, { "grad_norm": 0.34961557388305664, "learning_rate": 6.8493176683328e-07, "loss": 0.0268, "step": 94990 }, { "grad_norm": 0.3892488181591034, "learning_rate": 6.822070090177657e-07, "loss": 0.0262, "step": 95000 }, { "grad_norm": 0.27866649627685547, "learning_rate": 6.794876445219367e-07, "loss": 0.0278, "step": 95010 }, { "grad_norm": 0.41679635643959045, "learning_rate": 6.767736736431885e-07, "loss": 0.0292, "step": 95020 }, { "grad_norm": 0.40310803055763245, "learning_rate": 6.740650966783113e-07, "loss": 0.0272, "step": 95030 }, { "grad_norm": 0.3385263681411743, "learning_rate": 6.713619139235239e-07, "loss": 0.0216, "step": 95040 }, { "grad_norm": 0.28705060482025146, "learning_rate": 6.686641256744286e-07, "loss": 0.0237, "step": 95050 }, { "grad_norm": 0.22787539660930634, "learning_rate": 6.659717322260562e-07, "loss": 0.0326, "step": 95060 }, { "grad_norm": 0.3356461226940155, "learning_rate": 6.632847338728377e-07, "loss": 0.0281, "step": 95070 }, { "grad_norm": 0.32031381130218506, "learning_rate": 6.60603130908627e-07, "loss": 0.0202, "step": 95080 }, { "grad_norm": 0.41480085253715515, "learning_rate": 6.579269236266783e-07, "loss": 0.0297, "step": 95090 }, { "grad_norm": 0.3251006305217743, "learning_rate": 6.552561123196632e-07, "loss": 0.0237, "step": 95100 }, { "grad_norm": 0.33827829360961914, "learning_rate": 6.525906972796425e-07, "loss": 0.0295, "step": 95110 }, { "grad_norm": 0.31336095929145813, "learning_rate": 6.499306787981108e-07, "loss": 0.0299, "step": 95120 }, { "grad_norm": 0.3935805559158325, "learning_rate": 6.472760571659631e-07, "loss": 0.03, "step": 95130 }, { "grad_norm": 0.63776695728302, "learning_rate": 6.446268326735061e-07, "loss": 0.0287, "step": 95140 }, { "grad_norm": 0.21017727255821228, "learning_rate": 6.41983005610458e-07, "loss": 0.0234, "step": 95150 }, { "grad_norm": 0.21487759053707123, "learning_rate": 6.393445762659378e-07, "loss": 0.03, "step": 95160 }, { "grad_norm": 0.345430463552475, "learning_rate": 6.36711544928481e-07, "loss": 0.0243, "step": 95170 }, { "grad_norm": 0.4487111568450928, "learning_rate": 6.340839118860298e-07, "loss": 0.0294, "step": 95180 }, { "grad_norm": 0.8967283368110657, "learning_rate": 6.314616774259485e-07, "loss": 0.0227, "step": 95190 }, { "grad_norm": 0.3074501156806946, "learning_rate": 6.288448418349858e-07, "loss": 0.0237, "step": 95200 }, { "grad_norm": 0.29630613327026367, "learning_rate": 6.262334053993235e-07, "loss": 0.0271, "step": 95210 }, { "grad_norm": 0.34636926651000977, "learning_rate": 6.236273684045501e-07, "loss": 0.0214, "step": 95220 }, { "grad_norm": 0.37166786193847656, "learning_rate": 6.210267311356488e-07, "loss": 0.0267, "step": 95230 }, { "grad_norm": 0.3938354551792145, "learning_rate": 6.184314938770197e-07, "loss": 0.0272, "step": 95240 }, { "grad_norm": 0.4940677881240845, "learning_rate": 6.158416569124858e-07, "loss": 0.0226, "step": 95250 }, { "grad_norm": 0.34483087062835693, "learning_rate": 6.132572205252541e-07, "loss": 0.0191, "step": 95260 }, { "grad_norm": 0.40268874168395996, "learning_rate": 6.106781849979648e-07, "loss": 0.0249, "step": 95270 }, { "grad_norm": 0.44217047095298767, "learning_rate": 6.081045506126482e-07, "loss": 0.0221, "step": 95280 }, { "grad_norm": 0.2671356201171875, "learning_rate": 6.055363176507678e-07, "loss": 0.0267, "step": 95290 }, { "grad_norm": 0.3988375663757324, "learning_rate": 6.029734863931602e-07, "loss": 0.0272, "step": 95300 }, { "grad_norm": 0.3453664779663086, "learning_rate": 6.004160571201179e-07, "loss": 0.0373, "step": 95310 }, { "grad_norm": 0.3541446924209595, "learning_rate": 5.978640301112947e-07, "loss": 0.0264, "step": 95320 }, { "grad_norm": 0.36620521545410156, "learning_rate": 5.9531740564579e-07, "loss": 0.0282, "step": 95330 }, { "grad_norm": 0.44558486342430115, "learning_rate": 5.927761840020918e-07, "loss": 0.0233, "step": 95340 }, { "grad_norm": 0.23196491599082947, "learning_rate": 5.902403654581113e-07, "loss": 0.0237, "step": 95350 }, { "grad_norm": 0.34647995233535767, "learning_rate": 5.877099502911542e-07, "loss": 0.0311, "step": 95360 }, { "grad_norm": 0.365551620721817, "learning_rate": 5.851849387779496e-07, "loss": 0.0349, "step": 95370 }, { "grad_norm": 0.5002036690711975, "learning_rate": 5.826653311946262e-07, "loss": 0.0278, "step": 95380 }, { "grad_norm": 0.30579331517219543, "learning_rate": 5.801511278167249e-07, "loss": 0.0192, "step": 95390 }, { "grad_norm": 0.24401108920574188, "learning_rate": 5.776423289191868e-07, "loss": 0.0279, "step": 95400 }, { "grad_norm": 0.47581565380096436, "learning_rate": 5.751389347763869e-07, "loss": 0.0288, "step": 95410 }, { "grad_norm": 0.41434958577156067, "learning_rate": 5.726409456620841e-07, "loss": 0.0272, "step": 95420 }, { "grad_norm": 0.24715998768806458, "learning_rate": 5.701483618494541e-07, "loss": 0.0199, "step": 95430 }, { "grad_norm": 0.2394144982099533, "learning_rate": 5.67661183611079e-07, "loss": 0.0237, "step": 95440 }, { "grad_norm": 0.2465592473745346, "learning_rate": 5.651794112189579e-07, "loss": 0.0222, "step": 95450 }, { "grad_norm": 0.4288153350353241, "learning_rate": 5.627030449444903e-07, "loss": 0.0256, "step": 95460 }, { "grad_norm": 0.2996326684951782, "learning_rate": 5.60232085058493e-07, "loss": 0.0259, "step": 95470 }, { "grad_norm": 0.3782860338687897, "learning_rate": 5.577665318311776e-07, "loss": 0.0336, "step": 95480 }, { "grad_norm": 0.2799379527568817, "learning_rate": 5.553063855321838e-07, "loss": 0.0251, "step": 95490 }, { "grad_norm": 0.32134008407592773, "learning_rate": 5.528516464305467e-07, "loss": 0.0294, "step": 95500 }, { "grad_norm": 0.37239575386047363, "learning_rate": 5.504023147947068e-07, "loss": 0.0261, "step": 95510 }, { "grad_norm": 0.42631375789642334, "learning_rate": 5.479583908925279e-07, "loss": 0.0356, "step": 95520 }, { "grad_norm": 0.37570229172706604, "learning_rate": 5.455198749912626e-07, "loss": 0.0271, "step": 95530 }, { "grad_norm": 0.33692029118537903, "learning_rate": 5.430867673575979e-07, "loss": 0.0345, "step": 95540 }, { "grad_norm": 0.5114927291870117, "learning_rate": 5.406590682575985e-07, "loss": 0.023, "step": 95550 }, { "grad_norm": 0.25165313482284546, "learning_rate": 5.382367779567687e-07, "loss": 0.0252, "step": 95560 }, { "grad_norm": 0.4183574616909027, "learning_rate": 5.358198967200023e-07, "loss": 0.0225, "step": 95570 }, { "grad_norm": 0.4051879644393921, "learning_rate": 5.334084248115989e-07, "loss": 0.0329, "step": 95580 }, { "grad_norm": 0.29636895656585693, "learning_rate": 5.310023624952754e-07, "loss": 0.0219, "step": 95590 }, { "grad_norm": 0.4155772030353546, "learning_rate": 5.286017100341656e-07, "loss": 0.0229, "step": 95600 }, { "grad_norm": 0.2960456907749176, "learning_rate": 5.262064676907874e-07, "loss": 0.0323, "step": 95610 }, { "grad_norm": 0.3138124644756317, "learning_rate": 5.238166357270924e-07, "loss": 0.0234, "step": 95620 }, { "grad_norm": 0.4267098903656006, "learning_rate": 5.214322144044215e-07, "loss": 0.0205, "step": 95630 }, { "grad_norm": 0.35046517848968506, "learning_rate": 5.190532039835327e-07, "loss": 0.0408, "step": 95640 }, { "grad_norm": 0.6929893493652344, "learning_rate": 5.166796047245903e-07, "loss": 0.0318, "step": 95650 }, { "grad_norm": 0.34383511543273926, "learning_rate": 5.143114168871699e-07, "loss": 0.0273, "step": 95660 }, { "grad_norm": 0.4970656931400299, "learning_rate": 5.119486407302532e-07, "loss": 0.0339, "step": 95670 }, { "grad_norm": 0.2880610525608063, "learning_rate": 5.095912765122335e-07, "loss": 0.0257, "step": 95680 }, { "grad_norm": 0.7054330110549927, "learning_rate": 5.07239324490899e-07, "loss": 0.0318, "step": 95690 }, { "grad_norm": 0.3917529284954071, "learning_rate": 5.048927849234608e-07, "loss": 0.0286, "step": 95700 }, { "grad_norm": 0.4247596561908722, "learning_rate": 5.025516580665358e-07, "loss": 0.0348, "step": 95710 }, { "grad_norm": 0.44630593061447144, "learning_rate": 5.002159441761412e-07, "loss": 0.0212, "step": 95720 }, { "grad_norm": 0.29913559556007385, "learning_rate": 4.978856435077062e-07, "loss": 0.0296, "step": 95730 }, { "grad_norm": 0.7175484895706177, "learning_rate": 4.955607563160769e-07, "loss": 0.0233, "step": 95740 }, { "grad_norm": 0.4519861042499542, "learning_rate": 4.932412828554944e-07, "loss": 0.0259, "step": 95750 }, { "grad_norm": 0.20382539927959442, "learning_rate": 4.909272233796114e-07, "loss": 0.0214, "step": 95760 }, { "grad_norm": 0.3173462748527527, "learning_rate": 4.88618578141492e-07, "loss": 0.0289, "step": 95770 }, { "grad_norm": 0.36275357007980347, "learning_rate": 4.863153473936011e-07, "loss": 0.0289, "step": 95780 }, { "grad_norm": 0.32048535346984863, "learning_rate": 4.840175313878259e-07, "loss": 0.0209, "step": 95790 }, { "grad_norm": 0.2092469185590744, "learning_rate": 4.817251303754433e-07, "loss": 0.0235, "step": 95800 }, { "grad_norm": 0.3840689957141876, "learning_rate": 4.794381446071583e-07, "loss": 0.0226, "step": 95810 }, { "grad_norm": 0.29065099358558655, "learning_rate": 4.771565743330597e-07, "loss": 0.0278, "step": 95820 }, { "grad_norm": 0.2542378902435303, "learning_rate": 4.7488041980266463e-07, "loss": 0.0302, "step": 95830 }, { "grad_norm": 0.4770827293395996, "learning_rate": 4.726096812648795e-07, "loss": 0.033, "step": 95840 }, { "grad_norm": 0.23145879805088043, "learning_rate": 4.703443589680445e-07, "loss": 0.0232, "step": 95850 }, { "grad_norm": 0.35925304889678955, "learning_rate": 4.6808445315987826e-07, "loss": 0.0286, "step": 95860 }, { "grad_norm": 0.44731584191322327, "learning_rate": 4.658299640875274e-07, "loss": 0.0239, "step": 95870 }, { "grad_norm": 0.308440625667572, "learning_rate": 4.635808919975393e-07, "loss": 0.0209, "step": 95880 }, { "grad_norm": 0.3561733663082123, "learning_rate": 4.6133723713586706e-07, "loss": 0.0273, "step": 95890 }, { "grad_norm": 0.41444769501686096, "learning_rate": 4.590989997478756e-07, "loss": 0.0236, "step": 95900 }, { "grad_norm": 0.3077757656574249, "learning_rate": 4.5686618007833027e-07, "loss": 0.028, "step": 95910 }, { "grad_norm": 0.5798391103744507, "learning_rate": 4.54638778371419e-07, "loss": 0.0253, "step": 95920 }, { "grad_norm": 0.341291218996048, "learning_rate": 4.524167948707192e-07, "loss": 0.0303, "step": 95930 }, { "grad_norm": 0.4716741442680359, "learning_rate": 4.5020022981921986e-07, "loss": 0.0257, "step": 95940 }, { "grad_norm": 0.42355436086654663, "learning_rate": 4.47989083459327e-07, "loss": 0.0323, "step": 95950 }, { "grad_norm": 0.3593897521495819, "learning_rate": 4.457833560328528e-07, "loss": 0.0276, "step": 95960 }, { "grad_norm": 0.3551463186740875, "learning_rate": 4.4358304778100434e-07, "loss": 0.0218, "step": 95970 }, { "grad_norm": 0.2595745325088501, "learning_rate": 4.413881589444058e-07, "loss": 0.0285, "step": 95980 }, { "grad_norm": 0.5150793790817261, "learning_rate": 4.391986897630873e-07, "loss": 0.0252, "step": 95990 }, { "grad_norm": 0.4286114573478699, "learning_rate": 4.370146404764908e-07, "loss": 0.0315, "step": 96000 }, { "grad_norm": 0.22378437221050262, "learning_rate": 4.3483601132345285e-07, "loss": 0.0324, "step": 96010 }, { "grad_norm": 0.6819047331809998, "learning_rate": 4.32662802542233e-07, "loss": 0.0278, "step": 96020 }, { "grad_norm": 0.24302279949188232, "learning_rate": 4.3049501437047444e-07, "loss": 0.0257, "step": 96030 }, { "grad_norm": 0.5187196135520935, "learning_rate": 4.283326470452653e-07, "loss": 0.0325, "step": 96040 }, { "grad_norm": 0.21639402210712433, "learning_rate": 4.2617570080306647e-07, "loss": 0.0219, "step": 96050 }, { "grad_norm": 0.8967024087905884, "learning_rate": 4.2402417587975585e-07, "loss": 0.0289, "step": 96060 }, { "grad_norm": 0.5149555206298828, "learning_rate": 4.218780725106286e-07, "loss": 0.0241, "step": 96070 }, { "grad_norm": 0.3863741457462311, "learning_rate": 4.197373909303748e-07, "loss": 0.0283, "step": 96080 }, { "grad_norm": 0.362711638212204, "learning_rate": 4.17602131373096e-07, "loss": 0.0236, "step": 96090 }, { "grad_norm": 0.5283002853393555, "learning_rate": 4.154722940723055e-07, "loss": 0.0322, "step": 96100 }, { "grad_norm": 0.24391470849514008, "learning_rate": 4.1334787926091136e-07, "loss": 0.0249, "step": 96110 }, { "grad_norm": 0.41756388545036316, "learning_rate": 4.1122888717124444e-07, "loss": 0.027, "step": 96120 }, { "grad_norm": 0.1786799132823944, "learning_rate": 4.091153180350249e-07, "loss": 0.0225, "step": 96130 }, { "grad_norm": 0.46990033984184265, "learning_rate": 4.0700717208340125e-07, "loss": 0.0314, "step": 96140 }, { "grad_norm": 0.5432407259941101, "learning_rate": 4.049044495469112e-07, "loss": 0.0251, "step": 96150 }, { "grad_norm": 0.7243479490280151, "learning_rate": 4.0280715065549866e-07, "loss": 0.0308, "step": 96160 }, { "grad_norm": 0.6559025049209595, "learning_rate": 4.0071527563853575e-07, "loss": 0.0303, "step": 96170 }, { "grad_norm": 0.7377772331237793, "learning_rate": 3.9862882472477823e-07, "loss": 0.0229, "step": 96180 }, { "grad_norm": 0.6318160891532898, "learning_rate": 3.9654779814239375e-07, "loss": 0.031, "step": 96190 }, { "grad_norm": 0.3063926696777344, "learning_rate": 3.944721961189668e-07, "loss": 0.0245, "step": 96200 }, { "grad_norm": 0.3776380121707916, "learning_rate": 3.9240201888147697e-07, "loss": 0.0293, "step": 96210 }, { "grad_norm": 0.2579241693019867, "learning_rate": 3.9033726665632096e-07, "loss": 0.0317, "step": 96220 }, { "grad_norm": 0.4132010042667389, "learning_rate": 3.882779396692959e-07, "loss": 0.0239, "step": 96230 }, { "grad_norm": 0.3166300356388092, "learning_rate": 3.8622403814559947e-07, "loss": 0.0255, "step": 96240 }, { "grad_norm": 0.3613591492176056, "learning_rate": 3.841755623098575e-07, "loss": 0.0215, "step": 96250 }, { "grad_norm": 0.5436669588088989, "learning_rate": 3.821325123860742e-07, "loss": 0.0259, "step": 96260 }, { "grad_norm": 0.8909134268760681, "learning_rate": 3.800948885976818e-07, "loss": 0.0248, "step": 96270 }, { "grad_norm": 0.23695218563079834, "learning_rate": 3.780626911675078e-07, "loss": 0.0265, "step": 96280 }, { "grad_norm": 0.24085989594459534, "learning_rate": 3.76035920317791e-07, "loss": 0.0301, "step": 96290 }, { "grad_norm": 0.6009212732315063, "learning_rate": 3.740145762701819e-07, "loss": 0.0246, "step": 96300 }, { "grad_norm": 0.40708616375923157, "learning_rate": 3.7199865924572585e-07, "loss": 0.0232, "step": 96310 }, { "grad_norm": 0.5599310398101807, "learning_rate": 3.699881694648799e-07, "loss": 0.0167, "step": 96320 }, { "grad_norm": 0.5281565189361572, "learning_rate": 3.679831071475126e-07, "loss": 0.029, "step": 96330 }, { "grad_norm": 0.2034941166639328, "learning_rate": 3.65983472512893e-07, "loss": 0.021, "step": 96340 }, { "grad_norm": 0.3163149356842041, "learning_rate": 3.6398926577969616e-07, "loss": 0.021, "step": 96350 }, { "grad_norm": 0.4198472499847412, "learning_rate": 3.620004871660032e-07, "loss": 0.0226, "step": 96360 }, { "grad_norm": 0.27619895339012146, "learning_rate": 3.600171368893124e-07, "loss": 0.0194, "step": 96370 }, { "grad_norm": 0.2940114438533783, "learning_rate": 3.580392151665113e-07, "loss": 0.0289, "step": 96380 }, { "grad_norm": 0.3215871751308441, "learning_rate": 3.5606672221391024e-07, "loss": 0.0182, "step": 96390 }, { "grad_norm": 0.5055205821990967, "learning_rate": 3.540996582472089e-07, "loss": 0.0207, "step": 96400 }, { "grad_norm": 0.2938767671585083, "learning_rate": 3.521380234815297e-07, "loss": 0.0263, "step": 96410 }, { "grad_norm": 0.3996656537055969, "learning_rate": 3.5018181813138985e-07, "loss": 0.0202, "step": 96420 }, { "grad_norm": 0.5604433417320251, "learning_rate": 3.482310424107238e-07, "loss": 0.0244, "step": 96430 }, { "grad_norm": 0.3300129175186157, "learning_rate": 3.4628569653285537e-07, "loss": 0.024, "step": 96440 }, { "grad_norm": 0.26078200340270996, "learning_rate": 3.44345780710531e-07, "loss": 0.027, "step": 96450 }, { "grad_norm": 0.5995227098464966, "learning_rate": 3.424112951558978e-07, "loss": 0.0396, "step": 96460 }, { "grad_norm": 0.2555792033672333, "learning_rate": 3.40482240080503e-07, "loss": 0.0205, "step": 96470 }, { "grad_norm": 0.24345771968364716, "learning_rate": 3.3855861569531133e-07, "loss": 0.0265, "step": 96480 }, { "grad_norm": 1.1216198205947876, "learning_rate": 3.3664042221068226e-07, "loss": 0.028, "step": 96490 }, { "grad_norm": 0.467538058757782, "learning_rate": 3.347276598363924e-07, "loss": 0.0271, "step": 96500 }, { "grad_norm": 0.4494251012802124, "learning_rate": 3.328203287816134e-07, "loss": 0.0257, "step": 96510 }, { "grad_norm": 0.3832084536552429, "learning_rate": 3.3091842925493386e-07, "loss": 0.027, "step": 96520 }, { "grad_norm": 0.6148515939712524, "learning_rate": 3.2902196146433196e-07, "loss": 0.0247, "step": 96530 }, { "grad_norm": 0.4434150755405426, "learning_rate": 3.2713092561720836e-07, "loss": 0.0282, "step": 96540 }, { "grad_norm": 0.7141839861869812, "learning_rate": 3.2524532192036996e-07, "loss": 0.0275, "step": 96550 }, { "grad_norm": 0.189281165599823, "learning_rate": 3.2336515058001836e-07, "loss": 0.0266, "step": 96560 }, { "grad_norm": 0.4208400249481201, "learning_rate": 3.2149041180176143e-07, "loss": 0.0216, "step": 96570 }, { "grad_norm": 0.34294772148132324, "learning_rate": 3.1962110579062954e-07, "loss": 0.019, "step": 96580 }, { "grad_norm": 0.3244350552558899, "learning_rate": 3.177572327510314e-07, "loss": 0.0239, "step": 96590 }, { "grad_norm": 0.25501272082328796, "learning_rate": 3.158987928868151e-07, "loss": 0.0322, "step": 96600 }, { "grad_norm": 0.5877757668495178, "learning_rate": 3.1404578640120144e-07, "loss": 0.0255, "step": 96610 }, { "grad_norm": 0.3368484675884247, "learning_rate": 3.1219821349683933e-07, "loss": 0.021, "step": 96620 }, { "grad_norm": 0.3146895170211792, "learning_rate": 3.103560743757783e-07, "loss": 0.0242, "step": 96630 }, { "grad_norm": 0.34889140725135803, "learning_rate": 3.0851936923946834e-07, "loss": 0.0238, "step": 96640 }, { "grad_norm": 0.3845488429069519, "learning_rate": 3.06688098288771e-07, "loss": 0.0177, "step": 96650 }, { "grad_norm": 0.377054899930954, "learning_rate": 3.0486226172394825e-07, "loss": 0.0242, "step": 96660 }, { "grad_norm": 0.2777472734451294, "learning_rate": 3.030418597446738e-07, "loss": 0.0212, "step": 96670 }, { "grad_norm": 0.2570859491825104, "learning_rate": 3.012268925500272e-07, "loss": 0.0315, "step": 96680 }, { "grad_norm": 0.46981698274612427, "learning_rate": 2.994173603384831e-07, "loss": 0.0224, "step": 96690 }, { "grad_norm": 0.3245205581188202, "learning_rate": 2.9761326330793317e-07, "loss": 0.0304, "step": 96700 }, { "grad_norm": 0.37934643030166626, "learning_rate": 2.958146016556751e-07, "loss": 0.026, "step": 96710 }, { "grad_norm": 0.18143339455127716, "learning_rate": 2.94021375578396e-07, "loss": 0.0283, "step": 96720 }, { "grad_norm": 0.8400502800941467, "learning_rate": 2.9223358527221135e-07, "loss": 0.0313, "step": 96730 }, { "grad_norm": 0.3064885437488556, "learning_rate": 2.904512309326257e-07, "loss": 0.0241, "step": 96740 }, { "grad_norm": 0.3363606631755829, "learning_rate": 2.8867431275455545e-07, "loss": 0.0312, "step": 96750 }, { "grad_norm": 0.33534693717956543, "learning_rate": 2.86902830932323e-07, "loss": 0.0324, "step": 96760 }, { "grad_norm": 0.35270339250564575, "learning_rate": 2.851367856596565e-07, "loss": 0.0287, "step": 96770 }, { "grad_norm": 0.3056977391242981, "learning_rate": 2.83376177129685e-07, "loss": 0.0223, "step": 96780 }, { "grad_norm": 0.7875928282737732, "learning_rate": 2.8162100553494884e-07, "loss": 0.0322, "step": 96790 }, { "grad_norm": 0.3682122528553009, "learning_rate": 2.798712710673834e-07, "loss": 0.0222, "step": 96800 }, { "grad_norm": 0.367739737033844, "learning_rate": 2.7812697391834676e-07, "loss": 0.0273, "step": 96810 }, { "grad_norm": 0.27990466356277466, "learning_rate": 2.7638811427858627e-07, "loss": 0.025, "step": 96820 }, { "grad_norm": 0.3195177912712097, "learning_rate": 2.746546923382665e-07, "loss": 0.0284, "step": 96830 }, { "grad_norm": 0.47460007667541504, "learning_rate": 2.729267082869469e-07, "loss": 0.0291, "step": 96840 }, { "grad_norm": 0.4488050043582916, "learning_rate": 2.712041623135986e-07, "loss": 0.0324, "step": 96850 }, { "grad_norm": 0.33439502120018005, "learning_rate": 2.694870546065931e-07, "loss": 0.0339, "step": 96860 }, { "grad_norm": 0.5557937026023865, "learning_rate": 2.677753853537246e-07, "loss": 0.0318, "step": 96870 }, { "grad_norm": 0.3855382800102234, "learning_rate": 2.660691547421601e-07, "loss": 0.0214, "step": 96880 }, { "grad_norm": 0.3807697594165802, "learning_rate": 2.6436836295850033e-07, "loss": 0.0284, "step": 96890 }, { "grad_norm": 0.4037681519985199, "learning_rate": 2.6267301018874093e-07, "loss": 0.0251, "step": 96900 }, { "grad_norm": 0.18969640135765076, "learning_rate": 2.6098309661828355e-07, "loss": 0.0244, "step": 96910 }, { "grad_norm": 0.4977286756038666, "learning_rate": 2.592986224319305e-07, "loss": 0.0214, "step": 96920 }, { "grad_norm": 0.5727968811988831, "learning_rate": 2.576195878139009e-07, "loss": 0.0227, "step": 96930 }, { "grad_norm": 0.43772852420806885, "learning_rate": 2.5594599294780364e-07, "loss": 0.0213, "step": 96940 }, { "grad_norm": 0.22748760879039764, "learning_rate": 2.542778380166644e-07, "loss": 0.0281, "step": 96950 }, { "grad_norm": 0.24202223122119904, "learning_rate": 2.526151232029095e-07, "loss": 0.0309, "step": 96960 }, { "grad_norm": 0.8111929297447205, "learning_rate": 2.5095784868836567e-07, "loss": 0.0274, "step": 96970 }, { "grad_norm": 0.3942182958126068, "learning_rate": 2.493060146542825e-07, "loss": 0.0247, "step": 96980 }, { "grad_norm": 0.34624817967414856, "learning_rate": 2.476596212812876e-07, "loss": 0.0195, "step": 96990 }, { "grad_norm": 0.3171914219856262, "learning_rate": 2.4601866874943703e-07, "loss": 0.0238, "step": 97000 }, { "grad_norm": 0.37356048822402954, "learning_rate": 2.4438315723818163e-07, "loss": 0.0255, "step": 97010 }, { "grad_norm": 0.6085114479064941, "learning_rate": 2.4275308692637833e-07, "loss": 0.0264, "step": 97020 }, { "grad_norm": 0.3441590368747711, "learning_rate": 2.411284579922901e-07, "loss": 0.0216, "step": 97030 }, { "grad_norm": 0.2636563777923584, "learning_rate": 2.395092706135804e-07, "loss": 0.0238, "step": 97040 }, { "grad_norm": 0.2729128301143646, "learning_rate": 2.3789552496731872e-07, "loss": 0.0312, "step": 97050 }, { "grad_norm": 0.46984586119651794, "learning_rate": 2.3628722122999158e-07, "loss": 0.0207, "step": 97060 }, { "grad_norm": 0.5843489766120911, "learning_rate": 2.3468435957747503e-07, "loss": 0.0229, "step": 97070 }, { "grad_norm": 0.18897831439971924, "learning_rate": 2.330869401850566e-07, "loss": 0.0277, "step": 97080 }, { "grad_norm": 0.36242252588272095, "learning_rate": 2.3149496322742437e-07, "loss": 0.021, "step": 97090 }, { "grad_norm": 0.5203165411949158, "learning_rate": 2.2990842887868346e-07, "loss": 0.0234, "step": 97100 }, { "grad_norm": 0.2385951280593872, "learning_rate": 2.2832733731232292e-07, "loss": 0.024, "step": 97110 }, { "grad_norm": 0.29877740144729614, "learning_rate": 2.2675168870124886e-07, "loss": 0.0209, "step": 97120 }, { "grad_norm": 0.3612743318080902, "learning_rate": 2.2518148321778455e-07, "loss": 0.0248, "step": 97130 }, { "grad_norm": 0.39357802271842957, "learning_rate": 2.2361672103363706e-07, "loss": 0.0264, "step": 97140 }, { "grad_norm": 0.3976353108882904, "learning_rate": 2.220574023199251e-07, "loss": 0.0263, "step": 97150 }, { "grad_norm": 0.44789987802505493, "learning_rate": 2.2050352724717894e-07, "loss": 0.0246, "step": 97160 }, { "grad_norm": 0.43574801087379456, "learning_rate": 2.1895509598532372e-07, "loss": 0.0252, "step": 97170 }, { "grad_norm": 0.48785826563835144, "learning_rate": 2.1741210870369066e-07, "loss": 0.0277, "step": 97180 }, { "grad_norm": 0.4818914830684662, "learning_rate": 2.1587456557102814e-07, "loss": 0.0226, "step": 97190 }, { "grad_norm": 0.18456795811653137, "learning_rate": 2.143424667554683e-07, "loss": 0.034, "step": 97200 }, { "grad_norm": 0.29916608333587646, "learning_rate": 2.128158124245716e-07, "loss": 0.0307, "step": 97210 }, { "grad_norm": 0.23059259355068207, "learning_rate": 2.1129460274527668e-07, "loss": 0.0216, "step": 97220 }, { "grad_norm": 0.34556207060813904, "learning_rate": 2.0977883788395046e-07, "loss": 0.0276, "step": 97230 }, { "grad_norm": 0.7136864066123962, "learning_rate": 2.0826851800634928e-07, "loss": 0.0203, "step": 97240 }, { "grad_norm": 0.42073023319244385, "learning_rate": 2.0676364327764653e-07, "loss": 0.0323, "step": 97250 }, { "grad_norm": 0.7920930981636047, "learning_rate": 2.0526421386240502e-07, "loss": 0.0268, "step": 97260 }, { "grad_norm": 0.30415767431259155, "learning_rate": 2.0377022992461025e-07, "loss": 0.033, "step": 97270 }, { "grad_norm": 0.641381561756134, "learning_rate": 2.0228169162763156e-07, "loss": 0.0304, "step": 97280 }, { "grad_norm": 0.8485695123672485, "learning_rate": 2.0079859913425536e-07, "loss": 0.0253, "step": 97290 }, { "grad_norm": 0.3194569945335388, "learning_rate": 1.9932095260667417e-07, "loss": 0.0329, "step": 97300 }, { "grad_norm": 0.5073105096817017, "learning_rate": 1.9784875220648092e-07, "loss": 0.0274, "step": 97310 }, { "grad_norm": 0.23017723858356476, "learning_rate": 1.9638199809466907e-07, "loss": 0.0232, "step": 97320 }, { "grad_norm": 0.4516426622867584, "learning_rate": 1.9492069043164364e-07, "loss": 0.027, "step": 97330 }, { "grad_norm": 0.4395572245121002, "learning_rate": 1.9346482937721567e-07, "loss": 0.0234, "step": 97340 }, { "grad_norm": 0.43798649311065674, "learning_rate": 1.920144150905856e-07, "loss": 0.0271, "step": 97350 }, { "grad_norm": 0.40327635407447815, "learning_rate": 1.9056944773037655e-07, "loss": 0.0278, "step": 97360 }, { "grad_norm": 0.3996405601501465, "learning_rate": 1.8912992745460102e-07, "loss": 0.0243, "step": 97370 }, { "grad_norm": 0.11117930710315704, "learning_rate": 1.876958544206886e-07, "loss": 0.0207, "step": 97380 }, { "grad_norm": 0.7255097031593323, "learning_rate": 1.8626722878546942e-07, "loss": 0.031, "step": 97390 }, { "grad_norm": 0.316311776638031, "learning_rate": 1.8484405070516851e-07, "loss": 0.0236, "step": 97400 }, { "grad_norm": 0.32810211181640625, "learning_rate": 1.8342632033542805e-07, "loss": 0.0297, "step": 97410 }, { "grad_norm": 0.2969033420085907, "learning_rate": 1.8201403783129066e-07, "loss": 0.0206, "step": 97420 }, { "grad_norm": 0.5419394969940186, "learning_rate": 1.806072033471884e-07, "loss": 0.0273, "step": 97430 }, { "grad_norm": 0.49280208349227905, "learning_rate": 1.79205817036987e-07, "loss": 0.0285, "step": 97440 }, { "grad_norm": 0.4495903253555298, "learning_rate": 1.7780987905393065e-07, "loss": 0.0201, "step": 97450 }, { "grad_norm": 0.2451295256614685, "learning_rate": 1.764193895506805e-07, "loss": 0.0244, "step": 97460 }, { "grad_norm": 0.8598994016647339, "learning_rate": 1.7503434867929824e-07, "loss": 0.0246, "step": 97470 }, { "grad_norm": 1.0854160785675049, "learning_rate": 1.7365475659125165e-07, "loss": 0.0227, "step": 97480 }, { "grad_norm": 0.18539738655090332, "learning_rate": 1.7228061343740332e-07, "loss": 0.025, "step": 97490 }, { "grad_norm": 0.34128326177597046, "learning_rate": 1.7091191936803308e-07, "loss": 0.0249, "step": 97500 }, { "grad_norm": 0.16647732257843018, "learning_rate": 1.6954867453281563e-07, "loss": 0.0256, "step": 97510 }, { "grad_norm": 0.27672678232192993, "learning_rate": 1.6819087908084285e-07, "loss": 0.0266, "step": 97520 }, { "grad_norm": 0.34802931547164917, "learning_rate": 1.6683853316058485e-07, "loss": 0.0245, "step": 97530 }, { "grad_norm": 0.4693892300128937, "learning_rate": 1.6549163691995106e-07, "loss": 0.0252, "step": 97540 }, { "grad_norm": 0.23466579616069794, "learning_rate": 1.641501905062237e-07, "loss": 0.0326, "step": 97550 }, { "grad_norm": 0.23514190316200256, "learning_rate": 1.628141940661021e-07, "loss": 0.0276, "step": 97560 }, { "grad_norm": 0.41091418266296387, "learning_rate": 1.6148364774569158e-07, "loss": 0.0229, "step": 97570 }, { "grad_norm": 0.2516670525074005, "learning_rate": 1.6015855169050355e-07, "loss": 0.0402, "step": 97580 }, { "grad_norm": 0.503913402557373, "learning_rate": 1.5883890604543873e-07, "loss": 0.0293, "step": 97590 }, { "grad_norm": 0.38822081685066223, "learning_rate": 1.5752471095482057e-07, "loss": 0.0254, "step": 97600 }, { "grad_norm": 0.502564013004303, "learning_rate": 1.5621596656235638e-07, "loss": 0.0273, "step": 97610 }, { "grad_norm": 0.5356746912002563, "learning_rate": 1.5491267301117608e-07, "loss": 0.0261, "step": 97620 }, { "grad_norm": 0.22189682722091675, "learning_rate": 1.536148304438101e-07, "loss": 0.0245, "step": 97630 }, { "grad_norm": 0.1924888640642166, "learning_rate": 1.5232243900217823e-07, "loss": 0.0307, "step": 97640 }, { "grad_norm": 0.3249762952327728, "learning_rate": 1.51035498827623e-07, "loss": 0.0225, "step": 97650 }, { "grad_norm": 0.617504358291626, "learning_rate": 1.497540100608763e-07, "loss": 0.0311, "step": 97660 }, { "grad_norm": 0.40649697184562683, "learning_rate": 1.4847797284208153e-07, "loss": 0.0234, "step": 97670 }, { "grad_norm": 0.4593870937824249, "learning_rate": 1.472073873107882e-07, "loss": 0.0384, "step": 97680 }, { "grad_norm": 0.20116904377937317, "learning_rate": 1.459422536059407e-07, "loss": 0.0294, "step": 97690 }, { "grad_norm": 0.62967449426651, "learning_rate": 1.4468257186589508e-07, "loss": 0.0218, "step": 97700 }, { "grad_norm": 0.42973506450653076, "learning_rate": 1.4342834222840217e-07, "loss": 0.0243, "step": 97710 }, { "grad_norm": 0.3022543787956238, "learning_rate": 1.4217956483063566e-07, "loss": 0.0228, "step": 97720 }, { "grad_norm": 0.35592398047447205, "learning_rate": 1.409362398091474e-07, "loss": 0.0217, "step": 97730 }, { "grad_norm": 0.3394923210144043, "learning_rate": 1.3969836729990638e-07, "loss": 0.0263, "step": 97740 }, { "grad_norm": 0.3274185061454773, "learning_rate": 1.384659474382932e-07, "loss": 0.0276, "step": 97750 }, { "grad_norm": 0.3450966477394104, "learning_rate": 1.372389803590779e-07, "loss": 0.0265, "step": 97760 }, { "grad_norm": 0.5098999738693237, "learning_rate": 1.360174661964364e-07, "loss": 0.0261, "step": 97770 }, { "grad_norm": 0.5054552555084229, "learning_rate": 1.3480140508396188e-07, "loss": 0.0268, "step": 97780 }, { "grad_norm": 0.3470705449581146, "learning_rate": 1.335907971546313e-07, "loss": 0.0263, "step": 97790 }, { "grad_norm": 0.3417893052101135, "learning_rate": 1.3238564254083875e-07, "loss": 0.0357, "step": 97800 }, { "grad_norm": 0.22515058517456055, "learning_rate": 1.3118594137437323e-07, "loss": 0.0227, "step": 97810 }, { "grad_norm": 0.266010046005249, "learning_rate": 1.2999169378644094e-07, "loss": 0.0283, "step": 97820 }, { "grad_norm": 0.33294346928596497, "learning_rate": 1.2880289990763738e-07, "loss": 0.0296, "step": 97830 }, { "grad_norm": 0.29963505268096924, "learning_rate": 1.2761955986796968e-07, "loss": 0.0224, "step": 97840 }, { "grad_norm": 0.19416001439094543, "learning_rate": 1.264416737968399e-07, "loss": 0.0273, "step": 97850 }, { "grad_norm": 0.2446938455104828, "learning_rate": 1.2526924182307275e-07, "loss": 0.0203, "step": 97860 }, { "grad_norm": 0.7748965620994568, "learning_rate": 1.2410226407487124e-07, "loss": 0.035, "step": 97870 }, { "grad_norm": 0.3355356752872467, "learning_rate": 1.229407406798555e-07, "loss": 0.0341, "step": 97880 }, { "grad_norm": 0.26816490292549133, "learning_rate": 1.2178467176505726e-07, "loss": 0.0247, "step": 97890 }, { "grad_norm": 0.40209364891052246, "learning_rate": 1.2063405745689205e-07, "loss": 0.03, "step": 97900 }, { "grad_norm": 0.34826725721359253, "learning_rate": 1.1948889788119812e-07, "loss": 0.0225, "step": 97910 }, { "grad_norm": 0.3869851529598236, "learning_rate": 1.1834919316320303e-07, "loss": 0.0262, "step": 97920 }, { "grad_norm": 0.31863144040107727, "learning_rate": 1.1721494342754048e-07, "loss": 0.0308, "step": 97930 }, { "grad_norm": 0.35252416133880615, "learning_rate": 1.1608614879825563e-07, "loss": 0.023, "step": 97940 }, { "grad_norm": 0.297149658203125, "learning_rate": 1.1496280939879422e-07, "loss": 0.0226, "step": 97950 }, { "grad_norm": 0.40759992599487305, "learning_rate": 1.1384492535199687e-07, "loss": 0.0214, "step": 97960 }, { "grad_norm": 0.24163547158241272, "learning_rate": 1.1273249678011578e-07, "loss": 0.0224, "step": 97970 }, { "grad_norm": 0.4740840196609497, "learning_rate": 1.1162552380480362e-07, "loss": 0.0254, "step": 97980 }, { "grad_norm": 0.2294376939535141, "learning_rate": 1.1052400654711359e-07, "loss": 0.0219, "step": 97990 }, { "grad_norm": 0.6372721195220947, "learning_rate": 1.0942794512751597e-07, "loss": 0.0273, "step": 98000 }, { "grad_norm": 0.41437986493110657, "learning_rate": 1.0833733966587045e-07, "loss": 0.0242, "step": 98010 }, { "grad_norm": 0.4895530045032501, "learning_rate": 1.0725219028143718e-07, "loss": 0.0296, "step": 98020 }, { "grad_norm": 0.2580574154853821, "learning_rate": 1.0617249709289345e-07, "loss": 0.0183, "step": 98030 }, { "grad_norm": 0.5417991876602173, "learning_rate": 1.0509826021831703e-07, "loss": 0.0295, "step": 98040 }, { "grad_norm": 0.3499250113964081, "learning_rate": 1.0402947977517508e-07, "loss": 0.0287, "step": 98050 }, { "grad_norm": 0.4760885238647461, "learning_rate": 1.0296615588035185e-07, "loss": 0.0293, "step": 98060 }, { "grad_norm": 0.42895960807800293, "learning_rate": 1.0190828865012659e-07, "loss": 0.0231, "step": 98070 }, { "grad_norm": 0.37468284368515015, "learning_rate": 1.008558782001956e-07, "loss": 0.0232, "step": 98080 }, { "grad_norm": 0.4084659516811371, "learning_rate": 9.980892464563906e-08, "loss": 0.0291, "step": 98090 }, { "grad_norm": 0.18362566828727722, "learning_rate": 9.876742810095985e-08, "loss": 0.0203, "step": 98100 }, { "grad_norm": 0.28530314564704895, "learning_rate": 9.77313886800446e-08, "loss": 0.0263, "step": 98110 }, { "grad_norm": 0.5638617277145386, "learning_rate": 9.670080649619717e-08, "loss": 0.0279, "step": 98120 }, { "grad_norm": 0.3714289963245392, "learning_rate": 9.567568166212737e-08, "loss": 0.0381, "step": 98130 }, { "grad_norm": 0.5737436413764954, "learning_rate": 9.465601428992888e-08, "loss": 0.0316, "step": 98140 }, { "grad_norm": 0.35189491510391235, "learning_rate": 9.364180449111803e-08, "loss": 0.0288, "step": 98150 }, { "grad_norm": 0.3184433579444885, "learning_rate": 9.263305237661169e-08, "loss": 0.0247, "step": 98160 }, { "grad_norm": 0.2657302916049957, "learning_rate": 9.162975805671603e-08, "loss": 0.026, "step": 98170 }, { "grad_norm": 0.2873421311378479, "learning_rate": 9.063192164115442e-08, "loss": 0.0209, "step": 98180 }, { "grad_norm": 0.31173864006996155, "learning_rate": 8.963954323904511e-08, "loss": 0.0189, "step": 98190 }, { "grad_norm": 0.40922167897224426, "learning_rate": 8.865262295891796e-08, "loss": 0.0283, "step": 98200 }, { "grad_norm": 0.49581435322761536, "learning_rate": 8.767116090870331e-08, "loss": 0.0272, "step": 98210 }, { "grad_norm": 0.37281325459480286, "learning_rate": 8.669515719572086e-08, "loss": 0.031, "step": 98220 }, { "grad_norm": 1.1252752542495728, "learning_rate": 8.572461192671855e-08, "loss": 0.028, "step": 98230 }, { "grad_norm": 0.5504010915756226, "learning_rate": 8.475952520782815e-08, "loss": 0.0222, "step": 98240 }, { "grad_norm": 0.33250024914741516, "learning_rate": 8.379989714458747e-08, "loss": 0.0224, "step": 98250 }, { "grad_norm": 0.4424354135990143, "learning_rate": 8.284572784194034e-08, "loss": 0.0225, "step": 98260 }, { "grad_norm": 0.4134570360183716, "learning_rate": 8.189701740424216e-08, "loss": 0.0269, "step": 98270 }, { "grad_norm": 0.7559043765068054, "learning_rate": 8.095376593522663e-08, "loss": 0.0314, "step": 98280 }, { "grad_norm": 0.3726666569709778, "learning_rate": 8.001597353806123e-08, "loss": 0.0227, "step": 98290 }, { "grad_norm": 0.413841187953949, "learning_rate": 7.90836403152917e-08, "loss": 0.0294, "step": 98300 }, { "grad_norm": 0.4350787401199341, "learning_rate": 7.815676636888092e-08, "loss": 0.024, "step": 98310 }, { "grad_norm": 0.6669147610664368, "learning_rate": 7.723535180019226e-08, "loss": 0.0299, "step": 98320 }, { "grad_norm": 0.33737894892692566, "learning_rate": 7.6319396709984e-08, "loss": 0.028, "step": 98330 }, { "grad_norm": 0.4217052161693573, "learning_rate": 7.540890119843158e-08, "loss": 0.0274, "step": 98340 }, { "grad_norm": 0.28283044695854187, "learning_rate": 7.450386536509979e-08, "loss": 0.0242, "step": 98350 }, { "grad_norm": 0.24162617325782776, "learning_rate": 7.360428930895947e-08, "loss": 0.0347, "step": 98360 }, { "grad_norm": 0.4330427646636963, "learning_rate": 7.271017312839302e-08, "loss": 0.03, "step": 98370 }, { "grad_norm": 0.40331119298934937, "learning_rate": 7.182151692118333e-08, "loss": 0.0211, "step": 98380 }, { "grad_norm": 0.38837435841560364, "learning_rate": 7.093832078449713e-08, "loss": 0.0262, "step": 98390 }, { "grad_norm": 0.2626640498638153, "learning_rate": 7.006058481493493e-08, "loss": 0.0226, "step": 98400 }, { "grad_norm": 0.4457036256790161, "learning_rate": 6.918830910847552e-08, "loss": 0.0312, "step": 98410 }, { "grad_norm": 0.3827590048313141, "learning_rate": 6.832149376052033e-08, "loss": 0.0236, "step": 98420 }, { "grad_norm": 0.30891743302345276, "learning_rate": 6.746013886584912e-08, "loss": 0.0315, "step": 98430 }, { "grad_norm": 0.24634581804275513, "learning_rate": 6.660424451866431e-08, "loss": 0.0168, "step": 98440 }, { "grad_norm": 0.3102988302707672, "learning_rate": 6.575381081256327e-08, "loss": 0.0263, "step": 98450 }, { "grad_norm": 0.6022870540618896, "learning_rate": 6.49088378405549e-08, "loss": 0.0279, "step": 98460 }, { "grad_norm": 0.35984906554222107, "learning_rate": 6.406932569503755e-08, "loss": 0.0238, "step": 98470 }, { "grad_norm": 0.3834472894668579, "learning_rate": 6.323527446782662e-08, "loss": 0.0238, "step": 98480 }, { "grad_norm": 0.34995532035827637, "learning_rate": 6.240668425012142e-08, "loss": 0.029, "step": 98490 }, { "grad_norm": 0.2671752870082855, "learning_rate": 6.158355513254388e-08, "loss": 0.0201, "step": 98500 }, { "grad_norm": 0.31849461793899536, "learning_rate": 6.076588720510534e-08, "loss": 0.0241, "step": 98510 }, { "grad_norm": 0.513603687286377, "learning_rate": 5.995368055722316e-08, "loss": 0.0262, "step": 98520 }, { "grad_norm": 0.3355852961540222, "learning_rate": 5.914693527773185e-08, "loss": 0.0352, "step": 98530 }, { "grad_norm": 0.31726783514022827, "learning_rate": 5.834565145483861e-08, "loss": 0.0221, "step": 98540 }, { "grad_norm": 0.3301301896572113, "learning_rate": 5.754982917618446e-08, "loss": 0.0228, "step": 98550 }, { "grad_norm": 0.29492756724357605, "learning_rate": 5.675946852878866e-08, "loss": 0.0319, "step": 98560 }, { "grad_norm": 0.3421443700790405, "learning_rate": 5.59745695990932e-08, "loss": 0.0232, "step": 98570 }, { "grad_norm": 0.31354957818984985, "learning_rate": 5.51951324729294e-08, "loss": 0.0265, "step": 98580 }, { "grad_norm": 0.24335472285747528, "learning_rate": 5.4421157235529093e-08, "loss": 0.0258, "step": 98590 }, { "grad_norm": 0.4034356474876404, "learning_rate": 5.365264397154679e-08, "loss": 0.0256, "step": 98600 }, { "grad_norm": 0.3419667184352875, "learning_rate": 5.288959276501526e-08, "loss": 0.0315, "step": 98610 }, { "grad_norm": 0.5755404233932495, "learning_rate": 5.2132003699378876e-08, "loss": 0.0276, "step": 98620 }, { "grad_norm": 0.2004508674144745, "learning_rate": 5.137987685749912e-08, "loss": 0.0235, "step": 98630 }, { "grad_norm": 0.3845483362674713, "learning_rate": 5.0633212321610227e-08, "loss": 0.0275, "step": 98640 }, { "grad_norm": 0.5566402077674866, "learning_rate": 4.989201017338019e-08, "loss": 0.0274, "step": 98650 }, { "grad_norm": 0.5118895769119263, "learning_rate": 4.9156270493860846e-08, "loss": 0.0227, "step": 98660 }, { "grad_norm": 0.4260329604148865, "learning_rate": 4.842599336351561e-08, "loss": 0.0239, "step": 98670 }, { "grad_norm": 0.34912288188934326, "learning_rate": 4.7701178862197274e-08, "loss": 0.022, "step": 98680 }, { "grad_norm": 0.481466144323349, "learning_rate": 4.698182706918131e-08, "loss": 0.0196, "step": 98690 }, { "grad_norm": 0.5558066368103027, "learning_rate": 4.6267938063121466e-08, "loss": 0.0257, "step": 98700 }, { "grad_norm": 0.2635539174079895, "learning_rate": 4.555951192209973e-08, "loss": 0.0248, "step": 98710 }, { "grad_norm": 0.28948095440864563, "learning_rate": 4.485654872358747e-08, "loss": 0.0202, "step": 98720 }, { "grad_norm": 0.40872102975845337, "learning_rate": 4.415904854445097e-08, "loss": 0.0262, "step": 98730 }, { "grad_norm": 0.4619457721710205, "learning_rate": 4.34670114609792e-08, "loss": 0.0288, "step": 98740 }, { "grad_norm": 0.5134365558624268, "learning_rate": 4.278043754884498e-08, "loss": 0.0188, "step": 98750 }, { "grad_norm": 0.43724769353866577, "learning_rate": 4.209932688312712e-08, "loss": 0.0307, "step": 98760 }, { "grad_norm": 0.3615778386592865, "learning_rate": 4.14236795383216e-08, "loss": 0.03, "step": 98770 }, { "grad_norm": 0.2937861382961273, "learning_rate": 4.075349558830821e-08, "loss": 0.0244, "step": 98780 }, { "grad_norm": 0.24123308062553406, "learning_rate": 4.008877510638387e-08, "loss": 0.0129, "step": 98790 }, { "grad_norm": 0.22416894137859344, "learning_rate": 3.942951816523488e-08, "loss": 0.0238, "step": 98800 }, { "grad_norm": 0.427947074174881, "learning_rate": 3.8775724836959125e-08, "loss": 0.0337, "step": 98810 }, { "grad_norm": 0.3004651963710785, "learning_rate": 3.812739519305497e-08, "loss": 0.0215, "step": 98820 }, { "grad_norm": 0.31517452001571655, "learning_rate": 3.748452930442126e-08, "loss": 0.0308, "step": 98830 }, { "grad_norm": 0.34835076332092285, "learning_rate": 3.6847127241362855e-08, "loss": 0.0248, "step": 98840 }, { "grad_norm": 0.3963736891746521, "learning_rate": 3.621518907358512e-08, "loss": 0.0356, "step": 98850 }, { "grad_norm": 0.4520653188228607, "learning_rate": 3.558871487019388e-08, "loss": 0.0292, "step": 98860 }, { "grad_norm": 0.29323244094848633, "learning_rate": 3.4967704699701007e-08, "loss": 0.0257, "step": 98870 }, { "grad_norm": 0.3337026536464691, "learning_rate": 3.435215863001884e-08, "loss": 0.0239, "step": 98880 }, { "grad_norm": 0.34747767448425293, "learning_rate": 3.374207672846019e-08, "loss": 0.0302, "step": 98890 }, { "grad_norm": 0.3633190989494324, "learning_rate": 3.313745906174392e-08, "loss": 0.0234, "step": 98900 }, { "grad_norm": 0.38802269101142883, "learning_rate": 3.253830569599492e-08, "loss": 0.0221, "step": 98910 }, { "grad_norm": 0.3130766451358795, "learning_rate": 3.1944616696727436e-08, "loss": 0.0209, "step": 98920 }, { "grad_norm": 0.5450541973114014, "learning_rate": 3.135639212887287e-08, "loss": 0.0285, "step": 98930 }, { "grad_norm": 0.7912887930870056, "learning_rate": 3.077363205675754e-08, "loss": 0.0251, "step": 98940 }, { "grad_norm": 0.22514387965202332, "learning_rate": 3.0196336544113804e-08, "loss": 0.0215, "step": 98950 }, { "grad_norm": 0.4752424359321594, "learning_rate": 2.9624505654063383e-08, "loss": 0.0293, "step": 98960 }, { "grad_norm": 0.3080839514732361, "learning_rate": 2.90581394491507e-08, "loss": 0.0269, "step": 98970 }, { "grad_norm": 0.3525433838367462, "learning_rate": 2.8497237991309545e-08, "loss": 0.0288, "step": 98980 }, { "grad_norm": 0.4178753197193146, "learning_rate": 2.7941801341879735e-08, "loss": 0.026, "step": 98990 }, { "grad_norm": 0.6242159008979797, "learning_rate": 2.7391829561601578e-08, "loss": 0.0305, "step": 99000 }, { "grad_norm": 1.2629023790359497, "learning_rate": 2.6847322710621403e-08, "loss": 0.0212, "step": 99010 }, { "grad_norm": 0.3928625285625458, "learning_rate": 2.6308280848486022e-08, "loss": 0.0237, "step": 99020 }, { "grad_norm": 0.2914687693119049, "learning_rate": 2.5774704034137177e-08, "loss": 0.0265, "step": 99030 }, { "grad_norm": 0.2626413404941559, "learning_rate": 2.524659232593929e-08, "loss": 0.0265, "step": 99040 }, { "grad_norm": 0.2291116714477539, "learning_rate": 2.472394578163506e-08, "loss": 0.0254, "step": 99050 }, { "grad_norm": 0.245641827583313, "learning_rate": 2.4206764458378772e-08, "loss": 0.0196, "step": 99060 }, { "grad_norm": 0.3498840034008026, "learning_rate": 2.3695048412736285e-08, "loss": 0.0234, "step": 99070 }, { "grad_norm": 0.30870378017425537, "learning_rate": 2.3188797700668395e-08, "loss": 0.0249, "step": 99080 }, { "grad_norm": 0.38515955209732056, "learning_rate": 2.268801237753082e-08, "loss": 0.0189, "step": 99090 }, { "grad_norm": 0.3285263776779175, "learning_rate": 2.2192692498090862e-08, "loss": 0.0202, "step": 99100 }, { "grad_norm": 0.47499969601631165, "learning_rate": 2.170283811652185e-08, "loss": 0.0303, "step": 99110 }, { "grad_norm": 0.37506815791130066, "learning_rate": 2.12184492863865e-08, "loss": 0.029, "step": 99120 }, { "grad_norm": 0.40578189492225647, "learning_rate": 2.073952606066465e-08, "loss": 0.0263, "step": 99130 }, { "grad_norm": 0.23484894633293152, "learning_rate": 2.026606849171997e-08, "loss": 0.0226, "step": 99140 }, { "grad_norm": 0.31730106472969055, "learning_rate": 1.9798076631333262e-08, "loss": 0.0301, "step": 99150 }, { "grad_norm": 0.3518393933773041, "learning_rate": 1.933555053069136e-08, "loss": 0.0237, "step": 99160 }, { "grad_norm": 0.30399319529533386, "learning_rate": 1.887849024036492e-08, "loss": 0.023, "step": 99170 }, { "grad_norm": 0.8721805810928345, "learning_rate": 1.8426895810341736e-08, "loss": 0.0299, "step": 99180 }, { "grad_norm": 0.39373311400413513, "learning_rate": 1.798076729000453e-08, "loss": 0.0213, "step": 99190 }, { "grad_norm": 0.23869064450263977, "learning_rate": 1.75401047281476e-08, "loss": 0.0197, "step": 99200 }, { "grad_norm": 0.275142103433609, "learning_rate": 1.7104908172954626e-08, "loss": 0.024, "step": 99210 }, { "grad_norm": 0.2594561278820038, "learning_rate": 1.6675177672020868e-08, "loss": 0.0213, "step": 99220 }, { "grad_norm": 0.8350881934165955, "learning_rate": 1.6250913272342073e-08, "loss": 0.0301, "step": 99230 }, { "grad_norm": 0.3247620761394501, "learning_rate": 1.5832115020314453e-08, "loss": 0.0281, "step": 99240 }, { "grad_norm": 0.2721465229988098, "learning_rate": 1.5418782961734712e-08, "loss": 0.0256, "step": 99250 }, { "grad_norm": 0.4445924758911133, "learning_rate": 1.501091714181113e-08, "loss": 0.0243, "step": 99260 }, { "grad_norm": 0.41919663548469543, "learning_rate": 1.460851760513582e-08, "loss": 0.0285, "step": 99270 }, { "grad_norm": 0.3542095422744751, "learning_rate": 1.4211584395723566e-08, "loss": 0.0311, "step": 99280 }, { "grad_norm": 0.25720879435539246, "learning_rate": 1.382011755697854e-08, "loss": 0.0175, "step": 99290 }, { "grad_norm": 0.31498584151268005, "learning_rate": 1.3434117131716494e-08, "loss": 0.0177, "step": 99300 }, { "grad_norm": 0.42307695746421814, "learning_rate": 1.3053583162142557e-08, "loss": 0.0239, "step": 99310 }, { "grad_norm": 0.15656185150146484, "learning_rate": 1.2678515689873438e-08, "loss": 0.0276, "step": 99320 }, { "grad_norm": 0.13319070637226105, "learning_rate": 1.2308914755931877e-08, "loss": 0.0239, "step": 99330 }, { "grad_norm": 0.3515844941139221, "learning_rate": 1.1944780400729993e-08, "loss": 0.025, "step": 99340 }, { "grad_norm": 0.6046210527420044, "learning_rate": 1.1586112664085935e-08, "loss": 0.0212, "step": 99350 }, { "grad_norm": 0.44671764969825745, "learning_rate": 1.1232911585234984e-08, "loss": 0.0239, "step": 99360 }, { "grad_norm": 0.35286518931388855, "learning_rate": 1.0885177202790697e-08, "loss": 0.0293, "step": 99370 }, { "grad_norm": 0.33105042576789856, "learning_rate": 1.0542909554789315e-08, "loss": 0.0257, "step": 99380 }, { "grad_norm": 0.2661292254924774, "learning_rate": 1.0206108678656456e-08, "loss": 0.022, "step": 99390 }, { "grad_norm": 0.5432494282722473, "learning_rate": 9.874774611223769e-09, "loss": 0.0318, "step": 99400 }, { "grad_norm": 0.25176170468330383, "learning_rate": 9.548907388728933e-09, "loss": 0.0191, "step": 99410 }, { "grad_norm": 0.857023298740387, "learning_rate": 9.228507046804557e-09, "loss": 0.0301, "step": 99420 }, { "grad_norm": 0.17502151429653168, "learning_rate": 8.91357362048928e-09, "loss": 0.0363, "step": 99430 }, { "grad_norm": 0.5001162886619568, "learning_rate": 8.604107144227769e-09, "loss": 0.0246, "step": 99440 }, { "grad_norm": 0.19750584661960602, "learning_rate": 8.300107651859623e-09, "loss": 0.0194, "step": 99450 }, { "grad_norm": 1.0706442594528198, "learning_rate": 8.001575176630472e-09, "loss": 0.0305, "step": 99460 }, { "grad_norm": 0.33169305324554443, "learning_rate": 7.708509751186422e-09, "loss": 0.018, "step": 99470 }, { "grad_norm": 0.3352932035923004, "learning_rate": 7.420911407579611e-09, "loss": 0.023, "step": 99480 }, { "grad_norm": 0.5980112552642822, "learning_rate": 7.13878017725711e-09, "loss": 0.016, "step": 99490 }, { "grad_norm": 0.3315909206867218, "learning_rate": 6.8621160910720125e-09, "loss": 0.0302, "step": 99500 }, { "grad_norm": 0.5127311944961548, "learning_rate": 6.59091917928345e-09, "loss": 0.0292, "step": 99510 }, { "grad_norm": 0.6216202974319458, "learning_rate": 6.32518947155103e-09, "loss": 0.0275, "step": 99520 }, { "grad_norm": 0.37046289443969727, "learning_rate": 6.064926996929288e-09, "loss": 0.0219, "step": 99530 }, { "grad_norm": 0.27568450570106506, "learning_rate": 5.810131783884343e-09, "loss": 0.0233, "step": 99540 }, { "grad_norm": 0.27610763907432556, "learning_rate": 5.5608038602772415e-09, "loss": 0.0283, "step": 99550 }, { "grad_norm": 0.4278223216533661, "learning_rate": 5.316943253375062e-09, "loss": 0.0215, "step": 99560 }, { "grad_norm": 0.26289454102516174, "learning_rate": 5.07854998984536e-09, "loss": 0.0237, "step": 99570 }, { "grad_norm": 0.5225492119789124, "learning_rate": 4.845624095756174e-09, "loss": 0.0284, "step": 99580 }, { "grad_norm": 0.6601206660270691, "learning_rate": 4.618165596587121e-09, "loss": 0.0229, "step": 99590 }, { "grad_norm": 0.5779791474342346, "learning_rate": 4.396174517207197e-09, "loss": 0.0257, "step": 99600 }, { "grad_norm": 0.37955033779144287, "learning_rate": 4.179650881896979e-09, "loss": 0.0296, "step": 99610 }, { "grad_norm": 0.19435939192771912, "learning_rate": 3.968594714331975e-09, "loss": 0.0266, "step": 99620 }, { "grad_norm": 0.38692161440849304, "learning_rate": 3.763006037593719e-09, "loss": 0.0254, "step": 99630 }, { "grad_norm": 0.3415839672088623, "learning_rate": 3.5628848741586786e-09, "loss": 0.0196, "step": 99640 }, { "grad_norm": 0.3775966167449951, "learning_rate": 3.368231245926001e-09, "loss": 0.0269, "step": 99650 }, { "grad_norm": 0.3510943055152893, "learning_rate": 3.179045174167561e-09, "loss": 0.0295, "step": 99660 }, { "grad_norm": 0.3907158076763153, "learning_rate": 2.9953266795834657e-09, "loss": 0.0229, "step": 99670 }, { "grad_norm": 0.2691093385219574, "learning_rate": 2.817075782263201e-09, "loss": 0.031, "step": 99680 }, { "grad_norm": 0.46580982208251953, "learning_rate": 2.6442925016967325e-09, "loss": 0.0231, "step": 99690 }, { "grad_norm": 0.4485938847064972, "learning_rate": 2.4769768567800555e-09, "loss": 0.0271, "step": 99700 }, { "grad_norm": 0.5490649938583374, "learning_rate": 2.315128865809646e-09, "loss": 0.0268, "step": 99710 }, { "grad_norm": 0.9233383536338806, "learning_rate": 2.1587485464880097e-09, "loss": 0.029, "step": 99720 }, { "grad_norm": 0.3134004473686218, "learning_rate": 2.0078359159125813e-09, "loss": 0.0318, "step": 99730 }, { "grad_norm": 0.230943962931633, "learning_rate": 1.8623909905923776e-09, "loss": 0.0264, "step": 99740 }, { "grad_norm": 0.3696526288986206, "learning_rate": 1.7224137864257916e-09, "loss": 0.0248, "step": 99750 }, { "grad_norm": 0.9992347955703735, "learning_rate": 1.5879043187283505e-09, "loss": 0.0302, "step": 99760 }, { "grad_norm": 0.31577205657958984, "learning_rate": 1.458862602204958e-09, "loss": 0.0188, "step": 99770 }, { "grad_norm": 0.40674111247062683, "learning_rate": 1.3352886509720996e-09, "loss": 0.0254, "step": 99780 }, { "grad_norm": 0.6890877485275269, "learning_rate": 1.2171824785356389e-09, "loss": 0.0349, "step": 99790 }, { "grad_norm": 0.39679262042045593, "learning_rate": 1.1045440978185718e-09, "loss": 0.0245, "step": 99800 }, { "grad_norm": 0.2753223776817322, "learning_rate": 9.973735211332714e-10, "loss": 0.0263, "step": 99810 }, { "grad_norm": 0.42352935671806335, "learning_rate": 8.956707602036929e-10, "loss": 0.0252, "step": 99820 }, { "grad_norm": 0.33068007230758667, "learning_rate": 7.994358261542712e-10, "loss": 0.0252, "step": 99830 }, { "grad_norm": 0.28333547711372375, "learning_rate": 7.086687295043692e-10, "loss": 0.018, "step": 99840 }, { "grad_norm": 0.8704488277435303, "learning_rate": 6.23369480179381e-10, "loss": 0.022, "step": 99850 }, { "grad_norm": 0.4725516736507416, "learning_rate": 5.43538087510731e-10, "loss": 0.0334, "step": 99860 }, { "grad_norm": 0.6949617862701416, "learning_rate": 4.691745602303233e-10, "loss": 0.027, "step": 99870 }, { "grad_norm": 0.2117680162191391, "learning_rate": 4.0027890646499034e-10, "loss": 0.037, "step": 99880 }, { "grad_norm": 0.3038209080696106, "learning_rate": 3.368511337531466e-10, "loss": 0.0219, "step": 99890 }, { "grad_norm": 0.48742565512657166, "learning_rate": 2.788912490281348e-10, "loss": 0.0296, "step": 99900 }, { "grad_norm": 0.250200480222702, "learning_rate": 2.2639925863487954e-10, "loss": 0.0216, "step": 99910 }, { "grad_norm": 0.3350486755371094, "learning_rate": 1.7937516830213163e-10, "loss": 0.0235, "step": 99920 }, { "grad_norm": 0.39374423027038574, "learning_rate": 1.3781898318687702e-10, "loss": 0.0266, "step": 99930 }, { "grad_norm": 0.4458976686000824, "learning_rate": 1.0173070781882566e-10, "loss": 0.0271, "step": 99940 }, { "grad_norm": 0.29050955176353455, "learning_rate": 7.111034615592261e-11, "loss": 0.0247, "step": 99950 }, { "grad_norm": 0.33720943331718445, "learning_rate": 4.595790153993918e-11, "loss": 0.0229, "step": 99960 }, { "grad_norm": 0.41245824098587036, "learning_rate": 2.6273376729779585e-11, "loss": 0.0259, "step": 99970 }, { "grad_norm": 0.3569336533546448, "learning_rate": 1.2056773868174276e-11, "loss": 0.0252, "step": 99980 }, { "grad_norm": 0.26046258211135864, "learning_rate": 3.3080945149865926e-12, "loss": 0.02, "step": 99990 }, { "grad_norm": 0.4230760633945465, "learning_rate": 2.7339625008337977e-14, "loss": 0.0216, "step": 100000 } ], "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": true }, "attributes": {} } }, "total_flos": 0.0, "train_batch_size": 8, "trial_name": null, "trial_params": null }