{ "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": 0.8464847207069397, "learning_rate": 1.8e-07, "loss": 1.4985, "step": 10 }, { "grad_norm": 1.054180383682251, "learning_rate": 3.8e-07, "loss": 1.5032, "step": 20 }, { "grad_norm": 1.0063644647598267, "learning_rate": 5.8e-07, "loss": 1.4998, "step": 30 }, { "grad_norm": 0.8441745042800903, "learning_rate": 7.8e-07, "loss": 1.4806, "step": 40 }, { "grad_norm": 0.8750828504562378, "learning_rate": 9.8e-07, "loss": 1.5083, "step": 50 }, { "grad_norm": 1.1510131359100342, "learning_rate": 1.18e-06, "loss": 1.5183, "step": 60 }, { "grad_norm": 2.160525321960449, "learning_rate": 1.3800000000000001e-06, "loss": 1.399, "step": 70 }, { "grad_norm": 2.4494149684906006, "learning_rate": 1.5800000000000003e-06, "loss": 1.3499, "step": 80 }, { "grad_norm": 1.7925329208374023, "learning_rate": 1.7800000000000001e-06, "loss": 1.291, "step": 90 }, { "grad_norm": 1.3948293924331665, "learning_rate": 1.98e-06, "loss": 1.2958, "step": 100 }, { "grad_norm": 1.6880111694335938, "learning_rate": 2.1800000000000003e-06, "loss": 1.2703, "step": 110 }, { "grad_norm": 2.130403518676758, "learning_rate": 2.38e-06, "loss": 1.2323, "step": 120 }, { "grad_norm": 1.3792985677719116, "learning_rate": 2.5800000000000003e-06, "loss": 1.2101, "step": 130 }, { "grad_norm": 2.1650993824005127, "learning_rate": 2.78e-06, "loss": 1.1658, "step": 140 }, { "grad_norm": 1.7764437198638916, "learning_rate": 2.9800000000000003e-06, "loss": 1.194, "step": 150 }, { "grad_norm": 1.901289463043213, "learning_rate": 3.1800000000000005e-06, "loss": 1.1568, "step": 160 }, { "grad_norm": 1.3630796670913696, "learning_rate": 3.38e-06, "loss": 1.1482, "step": 170 }, { "grad_norm": 2.095909595489502, "learning_rate": 3.58e-06, "loss": 1.1755, "step": 180 }, { "grad_norm": 1.9114879369735718, "learning_rate": 3.7800000000000002e-06, "loss": 1.1654, "step": 190 }, { "grad_norm": 6.440182685852051, "learning_rate": 3.98e-06, "loss": 1.1521, "step": 200 }, { "grad_norm": 1.6306313276290894, "learning_rate": 4.18e-06, "loss": 1.1464, "step": 210 }, { "grad_norm": 5.229728698730469, "learning_rate": 4.38e-06, "loss": 1.1705, "step": 220 }, { "grad_norm": 1.8705716133117676, "learning_rate": 4.58e-06, "loss": 1.1503, "step": 230 }, { "grad_norm": 1.3275644779205322, "learning_rate": 4.780000000000001e-06, "loss": 1.1284, "step": 240 }, { "grad_norm": 1.2458494901657104, "learning_rate": 4.98e-06, "loss": 1.1075, "step": 250 }, { "grad_norm": 1.198982834815979, "learning_rate": 5.18e-06, "loss": 1.1178, "step": 260 }, { "grad_norm": 1.8580105304718018, "learning_rate": 5.38e-06, "loss": 1.1401, "step": 270 }, { "grad_norm": 1.388753890991211, "learning_rate": 5.580000000000001e-06, "loss": 1.1423, "step": 280 }, { "grad_norm": 1.9064571857452393, "learning_rate": 5.78e-06, "loss": 1.1465, "step": 290 }, { "grad_norm": 1.5941218137741089, "learning_rate": 5.98e-06, "loss": 1.1132, "step": 300 }, { "grad_norm": 1.7689636945724487, "learning_rate": 6.18e-06, "loss": 1.1362, "step": 310 }, { "grad_norm": 2.2115235328674316, "learning_rate": 6.38e-06, "loss": 1.1142, "step": 320 }, { "grad_norm": 1.7677314281463623, "learning_rate": 6.58e-06, "loss": 1.0925, "step": 330 }, { "grad_norm": 4.405050277709961, "learning_rate": 6.78e-06, "loss": 1.1217, "step": 340 }, { "grad_norm": 5.294124126434326, "learning_rate": 6.98e-06, "loss": 1.1033, "step": 350 }, { "grad_norm": 1.7716467380523682, "learning_rate": 7.180000000000001e-06, "loss": 1.1108, "step": 360 }, { "grad_norm": 1.3554683923721313, "learning_rate": 7.3800000000000005e-06, "loss": 1.1217, "step": 370 }, { "grad_norm": 1.6257944107055664, "learning_rate": 7.580000000000001e-06, "loss": 1.1104, "step": 380 }, { "grad_norm": 2.04974102973938, "learning_rate": 7.78e-06, "loss": 1.1085, "step": 390 }, { "grad_norm": 1.6043713092803955, "learning_rate": 7.98e-06, "loss": 1.0869, "step": 400 }, { "grad_norm": 1.2561283111572266, "learning_rate": 8.18e-06, "loss": 1.1088, "step": 410 }, { "grad_norm": 1.5179990530014038, "learning_rate": 8.380000000000001e-06, "loss": 1.1058, "step": 420 }, { "grad_norm": 2.098391056060791, "learning_rate": 8.580000000000001e-06, "loss": 1.1091, "step": 430 }, { "grad_norm": 1.4530293941497803, "learning_rate": 8.78e-06, "loss": 1.0744, "step": 440 }, { "grad_norm": 1.690532922744751, "learning_rate": 8.98e-06, "loss": 1.0692, "step": 450 }, { "grad_norm": 1.3673982620239258, "learning_rate": 9.180000000000002e-06, "loss": 1.0862, "step": 460 }, { "grad_norm": 1.7667092084884644, "learning_rate": 9.38e-06, "loss": 1.0889, "step": 470 }, { "grad_norm": 2.4466946125030518, "learning_rate": 9.58e-06, "loss": 1.0947, "step": 480 }, { "grad_norm": 1.4869695901870728, "learning_rate": 9.78e-06, "loss": 1.0826, "step": 490 }, { "grad_norm": 1.4189809560775757, "learning_rate": 9.980000000000001e-06, "loss": 1.0688, "step": 500 }, { "grad_norm": 1.9000593423843384, "learning_rate": 1.018e-05, "loss": 1.0677, "step": 510 }, { "grad_norm": 1.7268728017807007, "learning_rate": 1.038e-05, "loss": 1.0886, "step": 520 }, { "grad_norm": 1.47995126247406, "learning_rate": 1.058e-05, "loss": 1.0782, "step": 530 }, { "grad_norm": 1.5974761247634888, "learning_rate": 1.0780000000000002e-05, "loss": 1.0628, "step": 540 }, { "grad_norm": 1.3457727432250977, "learning_rate": 1.098e-05, "loss": 1.0682, "step": 550 }, { "grad_norm": 1.5447564125061035, "learning_rate": 1.118e-05, "loss": 1.0618, "step": 560 }, { "grad_norm": 1.488935947418213, "learning_rate": 1.1380000000000001e-05, "loss": 1.0511, "step": 570 }, { "grad_norm": 3.154773235321045, "learning_rate": 1.1580000000000001e-05, "loss": 1.0308, "step": 580 }, { "grad_norm": 1.9619965553283691, "learning_rate": 1.178e-05, "loss": 0.9919, "step": 590 }, { "grad_norm": 3.526035785675049, "learning_rate": 1.198e-05, "loss": 0.9811, "step": 600 }, { "grad_norm": 1.6965219974517822, "learning_rate": 1.2180000000000002e-05, "loss": 0.9795, "step": 610 }, { "grad_norm": 4.75495719909668, "learning_rate": 1.238e-05, "loss": 0.9691, "step": 620 }, { "grad_norm": 1.678918719291687, "learning_rate": 1.258e-05, "loss": 0.9471, "step": 630 }, { "grad_norm": 8.04729175567627, "learning_rate": 1.278e-05, "loss": 0.9115, "step": 640 }, { "grad_norm": 11.15231704711914, "learning_rate": 1.2980000000000001e-05, "loss": 0.8912, "step": 650 }, { "grad_norm": 2.7929611206054688, "learning_rate": 1.3180000000000001e-05, "loss": 0.891, "step": 660 }, { "grad_norm": 1.8924009799957275, "learning_rate": 1.338e-05, "loss": 0.8337, "step": 670 }, { "grad_norm": 2.0691111087799072, "learning_rate": 1.358e-05, "loss": 0.8129, "step": 680 }, { "grad_norm": 2.387789487838745, "learning_rate": 1.3780000000000002e-05, "loss": 0.8065, "step": 690 }, { "grad_norm": 2.815542697906494, "learning_rate": 1.3980000000000002e-05, "loss": 0.8069, "step": 700 }, { "grad_norm": 3.1250264644622803, "learning_rate": 1.4180000000000001e-05, "loss": 0.7347, "step": 710 }, { "grad_norm": 3.046013593673706, "learning_rate": 1.4380000000000001e-05, "loss": 0.7434, "step": 720 }, { "grad_norm": 2.461061716079712, "learning_rate": 1.4580000000000003e-05, "loss": 0.6974, "step": 730 }, { "grad_norm": 1.946713924407959, "learning_rate": 1.4779999999999999e-05, "loss": 0.6549, "step": 740 }, { "grad_norm": 2.5496644973754883, "learning_rate": 1.4979999999999999e-05, "loss": 0.6525, "step": 750 }, { "grad_norm": 2.907802104949951, "learning_rate": 1.518e-05, "loss": 0.6437, "step": 760 }, { "grad_norm": 2.2145323753356934, "learning_rate": 1.538e-05, "loss": 0.6226, "step": 770 }, { "grad_norm": 2.3674404621124268, "learning_rate": 1.558e-05, "loss": 0.6259, "step": 780 }, { "grad_norm": 2.3556861877441406, "learning_rate": 1.578e-05, "loss": 0.5664, "step": 790 }, { "grad_norm": 3.5584850311279297, "learning_rate": 1.598e-05, "loss": 0.5516, "step": 800 }, { "grad_norm": 3.0080909729003906, "learning_rate": 1.618e-05, "loss": 0.5328, "step": 810 }, { "grad_norm": 2.1218254566192627, "learning_rate": 1.6380000000000002e-05, "loss": 0.5171, "step": 820 }, { "grad_norm": 2.34804105758667, "learning_rate": 1.658e-05, "loss": 0.5263, "step": 830 }, { "grad_norm": 3.3104326725006104, "learning_rate": 1.6780000000000002e-05, "loss": 0.5022, "step": 840 }, { "grad_norm": 3.6980347633361816, "learning_rate": 1.698e-05, "loss": 0.502, "step": 850 }, { "grad_norm": 2.9864091873168945, "learning_rate": 1.718e-05, "loss": 0.4714, "step": 860 }, { "grad_norm": 2.3037071228027344, "learning_rate": 1.7380000000000003e-05, "loss": 0.4787, "step": 870 }, { "grad_norm": 3.2855396270751953, "learning_rate": 1.758e-05, "loss": 0.4447, "step": 880 }, { "grad_norm": 2.5812811851501465, "learning_rate": 1.7780000000000003e-05, "loss": 0.4325, "step": 890 }, { "grad_norm": 2.234724521636963, "learning_rate": 1.798e-05, "loss": 0.4243, "step": 900 }, { "grad_norm": 2.9969539642333984, "learning_rate": 1.818e-05, "loss": 0.3996, "step": 910 }, { "grad_norm": 3.620634078979492, "learning_rate": 1.838e-05, "loss": 0.3522, "step": 920 }, { "grad_norm": 2.9550716876983643, "learning_rate": 1.858e-05, "loss": 0.3672, "step": 930 }, { "grad_norm": 2.8491363525390625, "learning_rate": 1.878e-05, "loss": 0.3518, "step": 940 }, { "grad_norm": 3.061596393585205, "learning_rate": 1.898e-05, "loss": 0.337, "step": 950 }, { "grad_norm": 3.132007598876953, "learning_rate": 1.918e-05, "loss": 0.3204, "step": 960 }, { "grad_norm": 2.755042791366577, "learning_rate": 1.938e-05, "loss": 0.3512, "step": 970 }, { "grad_norm": 2.054859161376953, "learning_rate": 1.9580000000000002e-05, "loss": 0.2972, "step": 980 }, { "grad_norm": 2.97625994682312, "learning_rate": 1.978e-05, "loss": 0.2621, "step": 990 }, { "grad_norm": 3.11704421043396, "learning_rate": 1.9980000000000002e-05, "loss": 0.2736, "step": 1000 }, { "grad_norm": 2.266841173171997, "learning_rate": 2.0180000000000003e-05, "loss": 0.246, "step": 1010 }, { "grad_norm": 2.812222480773926, "learning_rate": 2.038e-05, "loss": 0.2762, "step": 1020 }, { "grad_norm": 2.743760347366333, "learning_rate": 2.0580000000000003e-05, "loss": 0.2093, "step": 1030 }, { "grad_norm": 2.4753851890563965, "learning_rate": 2.078e-05, "loss": 0.2296, "step": 1040 }, { "grad_norm": 3.355809211730957, "learning_rate": 2.098e-05, "loss": 0.2306, "step": 1050 }, { "grad_norm": 1.8589327335357666, "learning_rate": 2.118e-05, "loss": 0.2421, "step": 1060 }, { "grad_norm": 3.7372043132781982, "learning_rate": 2.138e-05, "loss": 0.263, "step": 1070 }, { "grad_norm": 3.015549421310425, "learning_rate": 2.158e-05, "loss": 0.262, "step": 1080 }, { "grad_norm": 3.3100428581237793, "learning_rate": 2.178e-05, "loss": 0.2479, "step": 1090 }, { "grad_norm": 2.3396921157836914, "learning_rate": 2.198e-05, "loss": 0.1934, "step": 1100 }, { "grad_norm": 2.4416465759277344, "learning_rate": 2.218e-05, "loss": 0.2677, "step": 1110 }, { "grad_norm": 2.049943447113037, "learning_rate": 2.2380000000000003e-05, "loss": 0.1823, "step": 1120 }, { "grad_norm": 2.027369976043701, "learning_rate": 2.258e-05, "loss": 0.2178, "step": 1130 }, { "grad_norm": 3.267463207244873, "learning_rate": 2.2780000000000002e-05, "loss": 0.2694, "step": 1140 }, { "grad_norm": 2.7576065063476562, "learning_rate": 2.298e-05, "loss": 0.2511, "step": 1150 }, { "grad_norm": 3.6810054779052734, "learning_rate": 2.318e-05, "loss": 0.2158, "step": 1160 }, { "grad_norm": 2.6482138633728027, "learning_rate": 2.3380000000000003e-05, "loss": 0.2176, "step": 1170 }, { "grad_norm": 2.3779971599578857, "learning_rate": 2.358e-05, "loss": 0.1833, "step": 1180 }, { "grad_norm": 2.3454740047454834, "learning_rate": 2.3780000000000003e-05, "loss": 0.2057, "step": 1190 }, { "grad_norm": 3.1935665607452393, "learning_rate": 2.398e-05, "loss": 0.2234, "step": 1200 }, { "grad_norm": 3.9271671772003174, "learning_rate": 2.418e-05, "loss": 0.2416, "step": 1210 }, { "grad_norm": 2.4799983501434326, "learning_rate": 2.438e-05, "loss": 0.2071, "step": 1220 }, { "grad_norm": 2.7889060974121094, "learning_rate": 2.4580000000000002e-05, "loss": 0.2125, "step": 1230 }, { "grad_norm": 1.8250389099121094, "learning_rate": 2.478e-05, "loss": 0.2029, "step": 1240 }, { "grad_norm": 2.5654561519622803, "learning_rate": 2.498e-05, "loss": 0.1933, "step": 1250 }, { "grad_norm": 2.497488021850586, "learning_rate": 2.5180000000000003e-05, "loss": 0.1846, "step": 1260 }, { "grad_norm": 2.457164764404297, "learning_rate": 2.5380000000000004e-05, "loss": 0.1605, "step": 1270 }, { "grad_norm": 1.3478517532348633, "learning_rate": 2.5580000000000002e-05, "loss": 0.2025, "step": 1280 }, { "grad_norm": 1.389578104019165, "learning_rate": 2.5779999999999997e-05, "loss": 0.2122, "step": 1290 }, { "grad_norm": 1.6912145614624023, "learning_rate": 2.598e-05, "loss": 0.1781, "step": 1300 }, { "grad_norm": 1.9606235027313232, "learning_rate": 2.618e-05, "loss": 0.1771, "step": 1310 }, { "grad_norm": 1.8594911098480225, "learning_rate": 2.6379999999999998e-05, "loss": 0.1861, "step": 1320 }, { "grad_norm": 1.7909629344940186, "learning_rate": 2.658e-05, "loss": 0.1671, "step": 1330 }, { "grad_norm": 2.4841365814208984, "learning_rate": 2.678e-05, "loss": 0.1703, "step": 1340 }, { "grad_norm": 3.4602773189544678, "learning_rate": 2.698e-05, "loss": 0.2224, "step": 1350 }, { "grad_norm": 3.1080129146575928, "learning_rate": 2.718e-05, "loss": 0.2012, "step": 1360 }, { "grad_norm": 2.28251314163208, "learning_rate": 2.738e-05, "loss": 0.1609, "step": 1370 }, { "grad_norm": 2.0572996139526367, "learning_rate": 2.758e-05, "loss": 0.1619, "step": 1380 }, { "grad_norm": 2.5376811027526855, "learning_rate": 2.778e-05, "loss": 0.1887, "step": 1390 }, { "grad_norm": 1.7168525457382202, "learning_rate": 2.798e-05, "loss": 0.1614, "step": 1400 }, { "grad_norm": 1.7297608852386475, "learning_rate": 2.818e-05, "loss": 0.1741, "step": 1410 }, { "grad_norm": 1.73287034034729, "learning_rate": 2.8380000000000003e-05, "loss": 0.1843, "step": 1420 }, { "grad_norm": 2.252885341644287, "learning_rate": 2.858e-05, "loss": 0.161, "step": 1430 }, { "grad_norm": 1.7528536319732666, "learning_rate": 2.8780000000000002e-05, "loss": 0.1719, "step": 1440 }, { "grad_norm": 2.225951671600342, "learning_rate": 2.898e-05, "loss": 0.1876, "step": 1450 }, { "grad_norm": 1.2541762590408325, "learning_rate": 2.9180000000000002e-05, "loss": 0.1746, "step": 1460 }, { "grad_norm": 1.5567556619644165, "learning_rate": 2.9380000000000003e-05, "loss": 0.1808, "step": 1470 }, { "grad_norm": 2.1847963333129883, "learning_rate": 2.958e-05, "loss": 0.1595, "step": 1480 }, { "grad_norm": 1.9323586225509644, "learning_rate": 2.9780000000000003e-05, "loss": 0.1616, "step": 1490 }, { "grad_norm": 1.3349732160568237, "learning_rate": 2.998e-05, "loss": 0.1628, "step": 1500 }, { "grad_norm": 1.222111701965332, "learning_rate": 3.0180000000000002e-05, "loss": 0.1449, "step": 1510 }, { "grad_norm": 2.083432912826538, "learning_rate": 3.0380000000000004e-05, "loss": 0.163, "step": 1520 }, { "grad_norm": 1.2809672355651855, "learning_rate": 3.058e-05, "loss": 0.1733, "step": 1530 }, { "grad_norm": 2.0132434368133545, "learning_rate": 3.078e-05, "loss": 0.1677, "step": 1540 }, { "grad_norm": 1.8756369352340698, "learning_rate": 3.0980000000000005e-05, "loss": 0.1577, "step": 1550 }, { "grad_norm": 1.5455114841461182, "learning_rate": 3.118e-05, "loss": 0.1776, "step": 1560 }, { "grad_norm": 1.5924005508422852, "learning_rate": 3.138e-05, "loss": 0.1427, "step": 1570 }, { "grad_norm": 2.033231496810913, "learning_rate": 3.1580000000000006e-05, "loss": 0.1602, "step": 1580 }, { "grad_norm": 1.72951340675354, "learning_rate": 3.1780000000000004e-05, "loss": 0.1684, "step": 1590 }, { "grad_norm": 1.9148659706115723, "learning_rate": 3.198e-05, "loss": 0.1441, "step": 1600 }, { "grad_norm": 1.609736442565918, "learning_rate": 3.218e-05, "loss": 0.1596, "step": 1610 }, { "grad_norm": 1.5207281112670898, "learning_rate": 3.238e-05, "loss": 0.1687, "step": 1620 }, { "grad_norm": 1.63938570022583, "learning_rate": 3.2579999999999996e-05, "loss": 0.1896, "step": 1630 }, { "grad_norm": 1.9251595735549927, "learning_rate": 3.278e-05, "loss": 0.1581, "step": 1640 }, { "grad_norm": 1.4724160432815552, "learning_rate": 3.298e-05, "loss": 0.1673, "step": 1650 }, { "grad_norm": 1.7391561269760132, "learning_rate": 3.318e-05, "loss": 0.1676, "step": 1660 }, { "grad_norm": 1.641829490661621, "learning_rate": 3.338e-05, "loss": 0.1316, "step": 1670 }, { "grad_norm": 1.6889485120773315, "learning_rate": 3.358e-05, "loss": 0.1463, "step": 1680 }, { "grad_norm": 1.7081063985824585, "learning_rate": 3.378e-05, "loss": 0.1471, "step": 1690 }, { "grad_norm": 2.3906006813049316, "learning_rate": 3.398e-05, "loss": 0.1744, "step": 1700 }, { "grad_norm": 2.0532920360565186, "learning_rate": 3.418e-05, "loss": 0.1831, "step": 1710 }, { "grad_norm": 1.6177265644073486, "learning_rate": 3.438e-05, "loss": 0.1742, "step": 1720 }, { "grad_norm": 1.6027008295059204, "learning_rate": 3.4580000000000004e-05, "loss": 0.1613, "step": 1730 }, { "grad_norm": 1.788079857826233, "learning_rate": 3.478e-05, "loss": 0.1434, "step": 1740 }, { "grad_norm": 2.3123486042022705, "learning_rate": 3.498e-05, "loss": 0.1607, "step": 1750 }, { "grad_norm": 1.588165521621704, "learning_rate": 3.518e-05, "loss": 0.1523, "step": 1760 }, { "grad_norm": 1.5312891006469727, "learning_rate": 3.5380000000000003e-05, "loss": 0.1523, "step": 1770 }, { "grad_norm": 1.5345933437347412, "learning_rate": 3.558e-05, "loss": 0.1707, "step": 1780 }, { "grad_norm": 2.2131741046905518, "learning_rate": 3.578e-05, "loss": 0.1595, "step": 1790 }, { "grad_norm": 1.3439041376113892, "learning_rate": 3.5980000000000004e-05, "loss": 0.1523, "step": 1800 }, { "grad_norm": 1.415622591972351, "learning_rate": 3.618e-05, "loss": 0.154, "step": 1810 }, { "grad_norm": 1.9201874732971191, "learning_rate": 3.638e-05, "loss": 0.1653, "step": 1820 }, { "grad_norm": 2.5167155265808105, "learning_rate": 3.6580000000000006e-05, "loss": 0.1563, "step": 1830 }, { "grad_norm": 1.5734045505523682, "learning_rate": 3.6780000000000004e-05, "loss": 0.1741, "step": 1840 }, { "grad_norm": 1.799631953239441, "learning_rate": 3.698e-05, "loss": 0.1773, "step": 1850 }, { "grad_norm": 1.5582793951034546, "learning_rate": 3.7180000000000007e-05, "loss": 0.1744, "step": 1860 }, { "grad_norm": 1.1879222393035889, "learning_rate": 3.7380000000000005e-05, "loss": 0.1427, "step": 1870 }, { "grad_norm": 2.0146141052246094, "learning_rate": 3.758e-05, "loss": 0.1366, "step": 1880 }, { "grad_norm": 1.354150652885437, "learning_rate": 3.778000000000001e-05, "loss": 0.1503, "step": 1890 }, { "grad_norm": 1.3200602531433105, "learning_rate": 3.7980000000000006e-05, "loss": 0.1412, "step": 1900 }, { "grad_norm": 2.3848421573638916, "learning_rate": 3.818e-05, "loss": 0.1877, "step": 1910 }, { "grad_norm": 1.4737353324890137, "learning_rate": 3.838e-05, "loss": 0.1595, "step": 1920 }, { "grad_norm": 1.263453722000122, "learning_rate": 3.858e-05, "loss": 0.1394, "step": 1930 }, { "grad_norm": 1.9549065828323364, "learning_rate": 3.878e-05, "loss": 0.1592, "step": 1940 }, { "grad_norm": 2.2550175189971924, "learning_rate": 3.898e-05, "loss": 0.1452, "step": 1950 }, { "grad_norm": 1.901489496231079, "learning_rate": 3.918e-05, "loss": 0.1816, "step": 1960 }, { "grad_norm": 1.4062530994415283, "learning_rate": 3.938e-05, "loss": 0.1289, "step": 1970 }, { "grad_norm": 1.1941865682601929, "learning_rate": 3.958e-05, "loss": 0.162, "step": 1980 }, { "grad_norm": 1.4222673177719116, "learning_rate": 3.978e-05, "loss": 0.1324, "step": 1990 }, { "grad_norm": 1.3182713985443115, "learning_rate": 3.998e-05, "loss": 0.1652, "step": 2000 }, { "grad_norm": 1.1659132242202759, "learning_rate": 4.018e-05, "loss": 0.184, "step": 2010 }, { "grad_norm": 1.5052553415298462, "learning_rate": 4.038e-05, "loss": 0.1489, "step": 2020 }, { "grad_norm": 1.3025153875350952, "learning_rate": 4.058e-05, "loss": 0.1288, "step": 2030 }, { "grad_norm": 1.293699860572815, "learning_rate": 4.078e-05, "loss": 0.1488, "step": 2040 }, { "grad_norm": 1.2826334238052368, "learning_rate": 4.0980000000000004e-05, "loss": 0.144, "step": 2050 }, { "grad_norm": 1.439159631729126, "learning_rate": 4.118e-05, "loss": 0.1508, "step": 2060 }, { "grad_norm": 1.3388177156448364, "learning_rate": 4.138e-05, "loss": 0.1343, "step": 2070 }, { "grad_norm": 1.5834600925445557, "learning_rate": 4.1580000000000005e-05, "loss": 0.1355, "step": 2080 }, { "grad_norm": 1.2623538970947266, "learning_rate": 4.178e-05, "loss": 0.1511, "step": 2090 }, { "grad_norm": 1.3258960247039795, "learning_rate": 4.198e-05, "loss": 0.1438, "step": 2100 }, { "grad_norm": 1.410923957824707, "learning_rate": 4.2180000000000006e-05, "loss": 0.1699, "step": 2110 }, { "grad_norm": 1.2242172956466675, "learning_rate": 4.2380000000000004e-05, "loss": 0.1321, "step": 2120 }, { "grad_norm": 1.1105581521987915, "learning_rate": 4.258e-05, "loss": 0.1384, "step": 2130 }, { "grad_norm": 0.8846121430397034, "learning_rate": 4.278e-05, "loss": 0.1225, "step": 2140 }, { "grad_norm": 1.2703771591186523, "learning_rate": 4.2980000000000005e-05, "loss": 0.1692, "step": 2150 }, { "grad_norm": 1.1009869575500488, "learning_rate": 4.318e-05, "loss": 0.1326, "step": 2160 }, { "grad_norm": 1.5546493530273438, "learning_rate": 4.338e-05, "loss": 0.138, "step": 2170 }, { "grad_norm": 1.3081358671188354, "learning_rate": 4.3580000000000006e-05, "loss": 0.1553, "step": 2180 }, { "grad_norm": 1.1655915975570679, "learning_rate": 4.3780000000000004e-05, "loss": 0.1393, "step": 2190 }, { "grad_norm": 1.0659396648406982, "learning_rate": 4.398e-05, "loss": 0.1289, "step": 2200 }, { "grad_norm": 1.0222684144973755, "learning_rate": 4.418000000000001e-05, "loss": 0.1369, "step": 2210 }, { "grad_norm": 1.2713292837142944, "learning_rate": 4.438e-05, "loss": 0.1408, "step": 2220 }, { "grad_norm": 1.607108473777771, "learning_rate": 4.458e-05, "loss": 0.1309, "step": 2230 }, { "grad_norm": 0.9565442800521851, "learning_rate": 4.478e-05, "loss": 0.1307, "step": 2240 }, { "grad_norm": 1.227523684501648, "learning_rate": 4.498e-05, "loss": 0.1441, "step": 2250 }, { "grad_norm": 1.112364649772644, "learning_rate": 4.518e-05, "loss": 0.1269, "step": 2260 }, { "grad_norm": 1.3739880323410034, "learning_rate": 4.538e-05, "loss": 0.1369, "step": 2270 }, { "grad_norm": 1.4287141561508179, "learning_rate": 4.558e-05, "loss": 0.1298, "step": 2280 }, { "grad_norm": 1.4427670240402222, "learning_rate": 4.578e-05, "loss": 0.1655, "step": 2290 }, { "grad_norm": 1.2694076299667358, "learning_rate": 4.5980000000000004e-05, "loss": 0.1338, "step": 2300 }, { "grad_norm": 0.9776490926742554, "learning_rate": 4.618e-05, "loss": 0.1268, "step": 2310 }, { "grad_norm": 1.3730971813201904, "learning_rate": 4.638e-05, "loss": 0.1257, "step": 2320 }, { "grad_norm": 1.1651381254196167, "learning_rate": 4.6580000000000005e-05, "loss": 0.1682, "step": 2330 }, { "grad_norm": 0.8320552706718445, "learning_rate": 4.678e-05, "loss": 0.1382, "step": 2340 }, { "grad_norm": 1.102421760559082, "learning_rate": 4.698e-05, "loss": 0.1415, "step": 2350 }, { "grad_norm": 1.6081587076187134, "learning_rate": 4.718e-05, "loss": 0.1508, "step": 2360 }, { "grad_norm": 0.8851668834686279, "learning_rate": 4.7380000000000004e-05, "loss": 0.1516, "step": 2370 }, { "grad_norm": 0.918836236000061, "learning_rate": 4.758e-05, "loss": 0.1442, "step": 2380 }, { "grad_norm": 1.2458869218826294, "learning_rate": 4.778e-05, "loss": 0.1603, "step": 2390 }, { "grad_norm": 0.9778954982757568, "learning_rate": 4.7980000000000005e-05, "loss": 0.1384, "step": 2400 }, { "grad_norm": 1.3220839500427246, "learning_rate": 4.818e-05, "loss": 0.1438, "step": 2410 }, { "grad_norm": 1.1186658143997192, "learning_rate": 4.838e-05, "loss": 0.1864, "step": 2420 }, { "grad_norm": 1.4317970275878906, "learning_rate": 4.8580000000000006e-05, "loss": 0.1415, "step": 2430 }, { "grad_norm": 1.4380428791046143, "learning_rate": 4.8780000000000004e-05, "loss": 0.1455, "step": 2440 }, { "grad_norm": 1.2117576599121094, "learning_rate": 4.898e-05, "loss": 0.1275, "step": 2450 }, { "grad_norm": 1.2224642038345337, "learning_rate": 4.918000000000001e-05, "loss": 0.1401, "step": 2460 }, { "grad_norm": 1.1302728652954102, "learning_rate": 4.9380000000000005e-05, "loss": 0.1371, "step": 2470 }, { "grad_norm": 0.9531450271606445, "learning_rate": 4.958e-05, "loss": 0.1586, "step": 2480 }, { "grad_norm": 1.2441776990890503, "learning_rate": 4.978e-05, "loss": 0.1546, "step": 2490 }, { "grad_norm": 0.8465694785118103, "learning_rate": 4.9980000000000006e-05, "loss": 0.1239, "step": 2500 }, { "grad_norm": 1.0055274963378906, "learning_rate": 5.0180000000000004e-05, "loss": 0.123, "step": 2510 }, { "grad_norm": 1.0733528137207031, "learning_rate": 5.038e-05, "loss": 0.1302, "step": 2520 }, { "grad_norm": 1.48505437374115, "learning_rate": 5.058000000000001e-05, "loss": 0.1198, "step": 2530 }, { "grad_norm": 1.0169817209243774, "learning_rate": 5.0780000000000005e-05, "loss": 0.1331, "step": 2540 }, { "grad_norm": 0.7386301159858704, "learning_rate": 5.098e-05, "loss": 0.1332, "step": 2550 }, { "grad_norm": 1.186721920967102, "learning_rate": 5.118000000000001e-05, "loss": 0.1589, "step": 2560 }, { "grad_norm": 1.2242825031280518, "learning_rate": 5.1380000000000006e-05, "loss": 0.1556, "step": 2570 }, { "grad_norm": 1.1624842882156372, "learning_rate": 5.1580000000000004e-05, "loss": 0.1529, "step": 2580 }, { "grad_norm": 1.0451737642288208, "learning_rate": 5.178000000000001e-05, "loss": 0.1199, "step": 2590 }, { "grad_norm": 0.9340122938156128, "learning_rate": 5.198000000000001e-05, "loss": 0.1549, "step": 2600 }, { "grad_norm": 0.9167302250862122, "learning_rate": 5.2180000000000005e-05, "loss": 0.1409, "step": 2610 }, { "grad_norm": 1.1286225318908691, "learning_rate": 5.238000000000001e-05, "loss": 0.1291, "step": 2620 }, { "grad_norm": 1.2213175296783447, "learning_rate": 5.258000000000001e-05, "loss": 0.1393, "step": 2630 }, { "grad_norm": 0.9978955984115601, "learning_rate": 5.2780000000000006e-05, "loss": 0.1505, "step": 2640 }, { "grad_norm": 0.8362855911254883, "learning_rate": 5.2980000000000004e-05, "loss": 0.1358, "step": 2650 }, { "grad_norm": 1.1066304445266724, "learning_rate": 5.318000000000001e-05, "loss": 0.1345, "step": 2660 }, { "grad_norm": 0.8679432272911072, "learning_rate": 5.338000000000001e-05, "loss": 0.1627, "step": 2670 }, { "grad_norm": 1.2029210329055786, "learning_rate": 5.3580000000000005e-05, "loss": 0.1778, "step": 2680 }, { "grad_norm": 0.650617778301239, "learning_rate": 5.378e-05, "loss": 0.1371, "step": 2690 }, { "grad_norm": 1.0780823230743408, "learning_rate": 5.3979999999999995e-05, "loss": 0.1206, "step": 2700 }, { "grad_norm": 1.1895666122436523, "learning_rate": 5.418e-05, "loss": 0.136, "step": 2710 }, { "grad_norm": 1.7175955772399902, "learning_rate": 5.438e-05, "loss": 0.1572, "step": 2720 }, { "grad_norm": 1.4107229709625244, "learning_rate": 5.4579999999999996e-05, "loss": 0.1335, "step": 2730 }, { "grad_norm": 1.166433572769165, "learning_rate": 5.478e-05, "loss": 0.136, "step": 2740 }, { "grad_norm": 0.9347928166389465, "learning_rate": 5.498e-05, "loss": 0.1516, "step": 2750 }, { "grad_norm": 1.208484411239624, "learning_rate": 5.518e-05, "loss": 0.13, "step": 2760 }, { "grad_norm": 1.146512508392334, "learning_rate": 5.538e-05, "loss": 0.1395, "step": 2770 }, { "grad_norm": 1.0666559934616089, "learning_rate": 5.558e-05, "loss": 0.1382, "step": 2780 }, { "grad_norm": 1.1187649965286255, "learning_rate": 5.578e-05, "loss": 0.135, "step": 2790 }, { "grad_norm": 1.034802794456482, "learning_rate": 5.5979999999999996e-05, "loss": 0.1112, "step": 2800 }, { "grad_norm": 0.9549369812011719, "learning_rate": 5.618e-05, "loss": 0.1571, "step": 2810 }, { "grad_norm": 1.0294355154037476, "learning_rate": 5.638e-05, "loss": 0.1343, "step": 2820 }, { "grad_norm": 1.0849475860595703, "learning_rate": 5.658e-05, "loss": 0.1263, "step": 2830 }, { "grad_norm": 0.8792015910148621, "learning_rate": 5.678e-05, "loss": 0.1586, "step": 2840 }, { "grad_norm": 1.1274044513702393, "learning_rate": 5.698e-05, "loss": 0.1405, "step": 2850 }, { "grad_norm": 1.1565700769424438, "learning_rate": 5.718e-05, "loss": 0.1354, "step": 2860 }, { "grad_norm": 0.8071972727775574, "learning_rate": 5.738e-05, "loss": 0.1164, "step": 2870 }, { "grad_norm": 0.9591960906982422, "learning_rate": 5.758e-05, "loss": 0.1095, "step": 2880 }, { "grad_norm": 0.9878796339035034, "learning_rate": 5.778e-05, "loss": 0.1312, "step": 2890 }, { "grad_norm": 0.8330735564231873, "learning_rate": 5.7980000000000004e-05, "loss": 0.1236, "step": 2900 }, { "grad_norm": 1.1143181324005127, "learning_rate": 5.818e-05, "loss": 0.1248, "step": 2910 }, { "grad_norm": 0.8297885060310364, "learning_rate": 5.838e-05, "loss": 0.1352, "step": 2920 }, { "grad_norm": 0.9068672060966492, "learning_rate": 5.858e-05, "loss": 0.1081, "step": 2930 }, { "grad_norm": 0.8303351402282715, "learning_rate": 5.878e-05, "loss": 0.1218, "step": 2940 }, { "grad_norm": 0.7112345099449158, "learning_rate": 5.898e-05, "loss": 0.1031, "step": 2950 }, { "grad_norm": 0.570334792137146, "learning_rate": 5.918e-05, "loss": 0.1277, "step": 2960 }, { "grad_norm": 0.9131639003753662, "learning_rate": 5.9380000000000004e-05, "loss": 0.1086, "step": 2970 }, { "grad_norm": 1.1094890832901, "learning_rate": 5.958e-05, "loss": 0.125, "step": 2980 }, { "grad_norm": 1.0515364408493042, "learning_rate": 5.978e-05, "loss": 0.1349, "step": 2990 }, { "grad_norm": 0.8320118188858032, "learning_rate": 5.9980000000000005e-05, "loss": 0.1331, "step": 3000 }, { "grad_norm": 0.7234342098236084, "learning_rate": 6.018e-05, "loss": 0.1152, "step": 3010 }, { "grad_norm": 0.8307440280914307, "learning_rate": 6.038e-05, "loss": 0.1128, "step": 3020 }, { "grad_norm": 0.9209387898445129, "learning_rate": 6.0580000000000006e-05, "loss": 0.1196, "step": 3030 }, { "grad_norm": 0.9921720623970032, "learning_rate": 6.0780000000000004e-05, "loss": 0.1197, "step": 3040 }, { "grad_norm": 1.1020227670669556, "learning_rate": 6.098e-05, "loss": 0.1264, "step": 3050 }, { "grad_norm": 0.8432523608207703, "learning_rate": 6.118000000000001e-05, "loss": 0.1246, "step": 3060 }, { "grad_norm": 0.9514641165733337, "learning_rate": 6.138e-05, "loss": 0.103, "step": 3070 }, { "grad_norm": 0.7290985584259033, "learning_rate": 6.158e-05, "loss": 0.1241, "step": 3080 }, { "grad_norm": 1.0609800815582275, "learning_rate": 6.178000000000001e-05, "loss": 0.143, "step": 3090 }, { "grad_norm": 1.106446623802185, "learning_rate": 6.198e-05, "loss": 0.1149, "step": 3100 }, { "grad_norm": 0.8228296637535095, "learning_rate": 6.218e-05, "loss": 0.1193, "step": 3110 }, { "grad_norm": 0.926352322101593, "learning_rate": 6.238000000000001e-05, "loss": 0.1121, "step": 3120 }, { "grad_norm": 0.7906055450439453, "learning_rate": 6.258e-05, "loss": 0.1098, "step": 3130 }, { "grad_norm": 0.7207959294319153, "learning_rate": 6.278e-05, "loss": 0.1244, "step": 3140 }, { "grad_norm": 1.0375984907150269, "learning_rate": 6.298000000000001e-05, "loss": 0.1174, "step": 3150 }, { "grad_norm": 0.7370395064353943, "learning_rate": 6.318e-05, "loss": 0.1176, "step": 3160 }, { "grad_norm": 0.829399049282074, "learning_rate": 6.338e-05, "loss": 0.1102, "step": 3170 }, { "grad_norm": 0.673848569393158, "learning_rate": 6.358000000000001e-05, "loss": 0.1303, "step": 3180 }, { "grad_norm": 1.024038314819336, "learning_rate": 6.378e-05, "loss": 0.1271, "step": 3190 }, { "grad_norm": 0.9954149127006531, "learning_rate": 6.398000000000001e-05, "loss": 0.1292, "step": 3200 }, { "grad_norm": 1.231956958770752, "learning_rate": 6.418000000000001e-05, "loss": 0.116, "step": 3210 }, { "grad_norm": 0.9846222400665283, "learning_rate": 6.438e-05, "loss": 0.1574, "step": 3220 }, { "grad_norm": 0.8542965650558472, "learning_rate": 6.458000000000001e-05, "loss": 0.1329, "step": 3230 }, { "grad_norm": 1.254880428314209, "learning_rate": 6.478000000000001e-05, "loss": 0.1195, "step": 3240 }, { "grad_norm": 0.8810468316078186, "learning_rate": 6.498e-05, "loss": 0.1428, "step": 3250 }, { "grad_norm": 0.8934184908866882, "learning_rate": 6.518000000000001e-05, "loss": 0.1246, "step": 3260 }, { "grad_norm": 1.4766439199447632, "learning_rate": 6.538000000000001e-05, "loss": 0.1278, "step": 3270 }, { "grad_norm": 0.8836305141448975, "learning_rate": 6.558e-05, "loss": 0.115, "step": 3280 }, { "grad_norm": 1.1804264783859253, "learning_rate": 6.578000000000001e-05, "loss": 0.1226, "step": 3290 }, { "grad_norm": 0.8295596837997437, "learning_rate": 6.598e-05, "loss": 0.1141, "step": 3300 }, { "grad_norm": 0.7507311701774597, "learning_rate": 6.618e-05, "loss": 0.1185, "step": 3310 }, { "grad_norm": 0.8252259492874146, "learning_rate": 6.638e-05, "loss": 0.1278, "step": 3320 }, { "grad_norm": 0.5866500735282898, "learning_rate": 6.658e-05, "loss": 0.1261, "step": 3330 }, { "grad_norm": 0.7191810011863708, "learning_rate": 6.678e-05, "loss": 0.134, "step": 3340 }, { "grad_norm": 0.7447108030319214, "learning_rate": 6.698e-05, "loss": 0.1139, "step": 3350 }, { "grad_norm": 0.8154744505882263, "learning_rate": 6.718e-05, "loss": 0.1178, "step": 3360 }, { "grad_norm": 0.9433508515357971, "learning_rate": 6.738e-05, "loss": 0.1238, "step": 3370 }, { "grad_norm": 0.7532247304916382, "learning_rate": 6.758e-05, "loss": 0.1643, "step": 3380 }, { "grad_norm": 0.7859284281730652, "learning_rate": 6.778e-05, "loss": 0.106, "step": 3390 }, { "grad_norm": 0.9460292458534241, "learning_rate": 6.798e-05, "loss": 0.119, "step": 3400 }, { "grad_norm": 1.3354878425598145, "learning_rate": 6.818e-05, "loss": 0.1209, "step": 3410 }, { "grad_norm": 0.8185003995895386, "learning_rate": 6.838e-05, "loss": 0.1063, "step": 3420 }, { "grad_norm": 1.0446351766586304, "learning_rate": 6.858e-05, "loss": 0.1159, "step": 3430 }, { "grad_norm": 0.8946515321731567, "learning_rate": 6.878e-05, "loss": 0.1379, "step": 3440 }, { "grad_norm": 1.1144753694534302, "learning_rate": 6.898e-05, "loss": 0.1345, "step": 3450 }, { "grad_norm": 0.9335242509841919, "learning_rate": 6.918e-05, "loss": 0.1246, "step": 3460 }, { "grad_norm": 0.6235050559043884, "learning_rate": 6.938e-05, "loss": 0.1336, "step": 3470 }, { "grad_norm": 1.0498002767562866, "learning_rate": 6.958e-05, "loss": 0.1342, "step": 3480 }, { "grad_norm": 0.8993150591850281, "learning_rate": 6.978e-05, "loss": 0.123, "step": 3490 }, { "grad_norm": 0.9908478856086731, "learning_rate": 6.998e-05, "loss": 0.0945, "step": 3500 }, { "grad_norm": 0.8289916515350342, "learning_rate": 7.018e-05, "loss": 0.1189, "step": 3510 }, { "grad_norm": 0.9478346109390259, "learning_rate": 7.038e-05, "loss": 0.1142, "step": 3520 }, { "grad_norm": 1.187625527381897, "learning_rate": 7.058e-05, "loss": 0.1289, "step": 3530 }, { "grad_norm": 0.9518083930015564, "learning_rate": 7.078e-05, "loss": 0.1299, "step": 3540 }, { "grad_norm": 0.7316707372665405, "learning_rate": 7.098e-05, "loss": 0.122, "step": 3550 }, { "grad_norm": 0.7866323590278625, "learning_rate": 7.118e-05, "loss": 0.0926, "step": 3560 }, { "grad_norm": 1.1045398712158203, "learning_rate": 7.138e-05, "loss": 0.1211, "step": 3570 }, { "grad_norm": 0.8704642653465271, "learning_rate": 7.158e-05, "loss": 0.1401, "step": 3580 }, { "grad_norm": 0.8099538683891296, "learning_rate": 7.178000000000001e-05, "loss": 0.1296, "step": 3590 }, { "grad_norm": 0.794862687587738, "learning_rate": 7.198e-05, "loss": 0.1148, "step": 3600 }, { "grad_norm": 0.6869994401931763, "learning_rate": 7.218e-05, "loss": 0.106, "step": 3610 }, { "grad_norm": 0.7379189729690552, "learning_rate": 7.238000000000001e-05, "loss": 0.1125, "step": 3620 }, { "grad_norm": 0.7695256471633911, "learning_rate": 7.258e-05, "loss": 0.0987, "step": 3630 }, { "grad_norm": 0.831959068775177, "learning_rate": 7.278e-05, "loss": 0.1169, "step": 3640 }, { "grad_norm": 0.7626714706420898, "learning_rate": 7.298000000000001e-05, "loss": 0.1124, "step": 3650 }, { "grad_norm": 1.082428216934204, "learning_rate": 7.318e-05, "loss": 0.1159, "step": 3660 }, { "grad_norm": 1.033618450164795, "learning_rate": 7.338e-05, "loss": 0.131, "step": 3670 }, { "grad_norm": 0.658189594745636, "learning_rate": 7.358000000000001e-05, "loss": 0.1086, "step": 3680 }, { "grad_norm": 0.8319560289382935, "learning_rate": 7.378e-05, "loss": 0.1193, "step": 3690 }, { "grad_norm": 1.0438131093978882, "learning_rate": 7.398e-05, "loss": 0.1145, "step": 3700 }, { "grad_norm": 0.9053070545196533, "learning_rate": 7.418000000000001e-05, "loss": 0.123, "step": 3710 }, { "grad_norm": 0.7303109169006348, "learning_rate": 7.438e-05, "loss": 0.1173, "step": 3720 }, { "grad_norm": 0.9428359270095825, "learning_rate": 7.458000000000001e-05, "loss": 0.1068, "step": 3730 }, { "grad_norm": 0.704149603843689, "learning_rate": 7.478e-05, "loss": 0.1123, "step": 3740 }, { "grad_norm": 0.7761657238006592, "learning_rate": 7.498e-05, "loss": 0.1269, "step": 3750 }, { "grad_norm": 0.7709400653839111, "learning_rate": 7.518000000000001e-05, "loss": 0.1104, "step": 3760 }, { "grad_norm": 0.9271396994590759, "learning_rate": 7.538e-05, "loss": 0.1239, "step": 3770 }, { "grad_norm": 0.8370038866996765, "learning_rate": 7.558e-05, "loss": 0.1216, "step": 3780 }, { "grad_norm": 0.6458547711372375, "learning_rate": 7.578000000000001e-05, "loss": 0.1033, "step": 3790 }, { "grad_norm": 0.8977019190788269, "learning_rate": 7.598e-05, "loss": 0.1344, "step": 3800 }, { "grad_norm": 0.765487015247345, "learning_rate": 7.618e-05, "loss": 0.1022, "step": 3810 }, { "grad_norm": 0.7146403789520264, "learning_rate": 7.638000000000001e-05, "loss": 0.1226, "step": 3820 }, { "grad_norm": 0.7465777397155762, "learning_rate": 7.658e-05, "loss": 0.1099, "step": 3830 }, { "grad_norm": 0.7486196756362915, "learning_rate": 7.678000000000001e-05, "loss": 0.1045, "step": 3840 }, { "grad_norm": 0.76885986328125, "learning_rate": 7.698000000000001e-05, "loss": 0.1236, "step": 3850 }, { "grad_norm": 1.5189528465270996, "learning_rate": 7.718e-05, "loss": 0.1081, "step": 3860 }, { "grad_norm": 1.4639289379119873, "learning_rate": 7.738000000000001e-05, "loss": 0.1137, "step": 3870 }, { "grad_norm": 0.7902600765228271, "learning_rate": 7.758000000000001e-05, "loss": 0.097, "step": 3880 }, { "grad_norm": 0.7208539247512817, "learning_rate": 7.778e-05, "loss": 0.1139, "step": 3890 }, { "grad_norm": 0.6460484266281128, "learning_rate": 7.798000000000001e-05, "loss": 0.1021, "step": 3900 }, { "grad_norm": 0.8676468133926392, "learning_rate": 7.818000000000001e-05, "loss": 0.118, "step": 3910 }, { "grad_norm": 1.034490942955017, "learning_rate": 7.838e-05, "loss": 0.1211, "step": 3920 }, { "grad_norm": 0.6468971967697144, "learning_rate": 7.858000000000001e-05, "loss": 0.0976, "step": 3930 }, { "grad_norm": 0.6959784030914307, "learning_rate": 7.878e-05, "loss": 0.1241, "step": 3940 }, { "grad_norm": 0.8283549547195435, "learning_rate": 7.897999999999999e-05, "loss": 0.1224, "step": 3950 }, { "grad_norm": 0.8549661636352539, "learning_rate": 7.918e-05, "loss": 0.1205, "step": 3960 }, { "grad_norm": 0.867145299911499, "learning_rate": 7.938e-05, "loss": 0.1074, "step": 3970 }, { "grad_norm": 1.072058081626892, "learning_rate": 7.958e-05, "loss": 0.1282, "step": 3980 }, { "grad_norm": 0.9086289405822754, "learning_rate": 7.978e-05, "loss": 0.1296, "step": 3990 }, { "grad_norm": 0.8169048428535461, "learning_rate": 7.998e-05, "loss": 0.1221, "step": 4000 }, { "grad_norm": 0.607533872127533, "learning_rate": 8.018e-05, "loss": 0.1138, "step": 4010 }, { "grad_norm": 0.806968629360199, "learning_rate": 8.038e-05, "loss": 0.1247, "step": 4020 }, { "grad_norm": 0.8597993850708008, "learning_rate": 8.058e-05, "loss": 0.1081, "step": 4030 }, { "grad_norm": 0.6444070935249329, "learning_rate": 8.078e-05, "loss": 0.0826, "step": 4040 }, { "grad_norm": 0.6254380345344543, "learning_rate": 8.098e-05, "loss": 0.1117, "step": 4050 }, { "grad_norm": 0.9512288570404053, "learning_rate": 8.118e-05, "loss": 0.1032, "step": 4060 }, { "grad_norm": 0.8021355867385864, "learning_rate": 8.138e-05, "loss": 0.1087, "step": 4070 }, { "grad_norm": 0.9134974479675293, "learning_rate": 8.158e-05, "loss": 0.1127, "step": 4080 }, { "grad_norm": 1.1473115682601929, "learning_rate": 8.178e-05, "loss": 0.1118, "step": 4090 }, { "grad_norm": 0.9211744666099548, "learning_rate": 8.198e-05, "loss": 0.1056, "step": 4100 }, { "grad_norm": 0.7614268064498901, "learning_rate": 8.218e-05, "loss": 0.1223, "step": 4110 }, { "grad_norm": 0.9371539354324341, "learning_rate": 8.238000000000001e-05, "loss": 0.1262, "step": 4120 }, { "grad_norm": 0.5152413845062256, "learning_rate": 8.258e-05, "loss": 0.1011, "step": 4130 }, { "grad_norm": 0.663923978805542, "learning_rate": 8.278e-05, "loss": 0.1132, "step": 4140 }, { "grad_norm": 0.6277439594268799, "learning_rate": 8.298000000000001e-05, "loss": 0.1318, "step": 4150 }, { "grad_norm": 0.6391077637672424, "learning_rate": 8.318e-05, "loss": 0.1081, "step": 4160 }, { "grad_norm": 1.018549919128418, "learning_rate": 8.338e-05, "loss": 0.1101, "step": 4170 }, { "grad_norm": 0.9553570747375488, "learning_rate": 8.358e-05, "loss": 0.1072, "step": 4180 }, { "grad_norm": 0.7684000134468079, "learning_rate": 8.378e-05, "loss": 0.1145, "step": 4190 }, { "grad_norm": 1.177592396736145, "learning_rate": 8.398e-05, "loss": 0.1163, "step": 4200 }, { "grad_norm": 1.1847755908966064, "learning_rate": 8.418e-05, "loss": 0.1149, "step": 4210 }, { "grad_norm": 0.9419437050819397, "learning_rate": 8.438e-05, "loss": 0.1352, "step": 4220 }, { "grad_norm": 0.7381961345672607, "learning_rate": 8.458e-05, "loss": 0.1478, "step": 4230 }, { "grad_norm": 0.7458416819572449, "learning_rate": 8.478e-05, "loss": 0.1315, "step": 4240 }, { "grad_norm": 0.8436911106109619, "learning_rate": 8.498e-05, "loss": 0.1297, "step": 4250 }, { "grad_norm": 1.098670482635498, "learning_rate": 8.518000000000001e-05, "loss": 0.1345, "step": 4260 }, { "grad_norm": 0.7706539630889893, "learning_rate": 8.538e-05, "loss": 0.1236, "step": 4270 }, { "grad_norm": 0.6520305871963501, "learning_rate": 8.558e-05, "loss": 0.0962, "step": 4280 }, { "grad_norm": 0.6856263875961304, "learning_rate": 8.578000000000001e-05, "loss": 0.1022, "step": 4290 }, { "grad_norm": 0.7555966377258301, "learning_rate": 8.598e-05, "loss": 0.0896, "step": 4300 }, { "grad_norm": 0.8859468102455139, "learning_rate": 8.618e-05, "loss": 0.1025, "step": 4310 }, { "grad_norm": 0.9309375882148743, "learning_rate": 8.638000000000001e-05, "loss": 0.1388, "step": 4320 }, { "grad_norm": 0.6511387825012207, "learning_rate": 8.658e-05, "loss": 0.1043, "step": 4330 }, { "grad_norm": 0.9176981449127197, "learning_rate": 8.678e-05, "loss": 0.0854, "step": 4340 }, { "grad_norm": 0.6177654266357422, "learning_rate": 8.698000000000001e-05, "loss": 0.1046, "step": 4350 }, { "grad_norm": 0.7828007340431213, "learning_rate": 8.718e-05, "loss": 0.098, "step": 4360 }, { "grad_norm": 0.7929646968841553, "learning_rate": 8.738000000000001e-05, "loss": 0.1014, "step": 4370 }, { "grad_norm": 0.6308605670928955, "learning_rate": 8.758000000000001e-05, "loss": 0.1162, "step": 4380 }, { "grad_norm": 0.9195981025695801, "learning_rate": 8.778e-05, "loss": 0.0997, "step": 4390 }, { "grad_norm": 0.7626798152923584, "learning_rate": 8.798000000000001e-05, "loss": 0.1175, "step": 4400 }, { "grad_norm": 0.7040757536888123, "learning_rate": 8.818000000000001e-05, "loss": 0.1189, "step": 4410 }, { "grad_norm": 0.6952450275421143, "learning_rate": 8.838e-05, "loss": 0.1265, "step": 4420 }, { "grad_norm": 0.855869710445404, "learning_rate": 8.858000000000001e-05, "loss": 0.1194, "step": 4430 }, { "grad_norm": 0.7785576581954956, "learning_rate": 8.878000000000001e-05, "loss": 0.1126, "step": 4440 }, { "grad_norm": 0.7592102289199829, "learning_rate": 8.898e-05, "loss": 0.1369, "step": 4450 }, { "grad_norm": 0.716750979423523, "learning_rate": 8.918000000000001e-05, "loss": 0.1138, "step": 4460 }, { "grad_norm": 0.657835066318512, "learning_rate": 8.938e-05, "loss": 0.1052, "step": 4470 }, { "grad_norm": 0.6910808086395264, "learning_rate": 8.958e-05, "loss": 0.1146, "step": 4480 }, { "grad_norm": 1.0847851037979126, "learning_rate": 8.978000000000001e-05, "loss": 0.119, "step": 4490 }, { "grad_norm": 0.8688377141952515, "learning_rate": 8.998e-05, "loss": 0.1241, "step": 4500 }, { "grad_norm": 0.652010977268219, "learning_rate": 9.018000000000001e-05, "loss": 0.1274, "step": 4510 }, { "grad_norm": 0.7409290671348572, "learning_rate": 9.038000000000001e-05, "loss": 0.111, "step": 4520 }, { "grad_norm": 0.7363811731338501, "learning_rate": 9.058e-05, "loss": 0.0978, "step": 4530 }, { "grad_norm": 0.9112172722816467, "learning_rate": 9.078000000000001e-05, "loss": 0.1294, "step": 4540 }, { "grad_norm": 0.8589863777160645, "learning_rate": 9.098000000000001e-05, "loss": 0.1245, "step": 4550 }, { "grad_norm": 0.7281624674797058, "learning_rate": 9.118e-05, "loss": 0.1081, "step": 4560 }, { "grad_norm": 1.0299357175827026, "learning_rate": 9.138e-05, "loss": 0.1393, "step": 4570 }, { "grad_norm": 0.99617999792099, "learning_rate": 9.158e-05, "loss": 0.1233, "step": 4580 }, { "grad_norm": 0.7418128252029419, "learning_rate": 9.178e-05, "loss": 0.1161, "step": 4590 }, { "grad_norm": 1.20258367061615, "learning_rate": 9.198e-05, "loss": 0.1198, "step": 4600 }, { "grad_norm": 0.8616735339164734, "learning_rate": 9.218e-05, "loss": 0.1187, "step": 4610 }, { "grad_norm": 0.7446396350860596, "learning_rate": 9.238e-05, "loss": 0.1102, "step": 4620 }, { "grad_norm": 0.8031488060951233, "learning_rate": 9.258e-05, "loss": 0.1235, "step": 4630 }, { "grad_norm": 0.9036995768547058, "learning_rate": 9.278e-05, "loss": 0.1274, "step": 4640 }, { "grad_norm": 0.6776604056358337, "learning_rate": 9.298e-05, "loss": 0.0984, "step": 4650 }, { "grad_norm": 0.8313411474227905, "learning_rate": 9.318e-05, "loss": 0.1148, "step": 4660 }, { "grad_norm": 0.982352614402771, "learning_rate": 9.338e-05, "loss": 0.1104, "step": 4670 }, { "grad_norm": 0.4500960409641266, "learning_rate": 9.358e-05, "loss": 0.1001, "step": 4680 }, { "grad_norm": 0.737207293510437, "learning_rate": 9.378e-05, "loss": 0.1184, "step": 4690 }, { "grad_norm": 0.7405228018760681, "learning_rate": 9.398e-05, "loss": 0.1195, "step": 4700 }, { "grad_norm": 0.7431883811950684, "learning_rate": 9.418e-05, "loss": 0.1097, "step": 4710 }, { "grad_norm": 0.5968936681747437, "learning_rate": 9.438e-05, "loss": 0.1025, "step": 4720 }, { "grad_norm": 0.5638996362686157, "learning_rate": 9.458e-05, "loss": 0.1177, "step": 4730 }, { "grad_norm": 0.7234206795692444, "learning_rate": 9.478e-05, "loss": 0.1081, "step": 4740 }, { "grad_norm": 2.16925048828125, "learning_rate": 9.498e-05, "loss": 0.107, "step": 4750 }, { "grad_norm": 1.098538875579834, "learning_rate": 9.518000000000001e-05, "loss": 0.131, "step": 4760 }, { "grad_norm": 0.7695603966712952, "learning_rate": 9.538e-05, "loss": 0.1052, "step": 4770 }, { "grad_norm": 0.9862655401229858, "learning_rate": 9.558e-05, "loss": 0.1274, "step": 4780 }, { "grad_norm": 0.7250051498413086, "learning_rate": 9.578000000000001e-05, "loss": 0.1028, "step": 4790 }, { "grad_norm": 0.5943331122398376, "learning_rate": 9.598e-05, "loss": 0.1017, "step": 4800 }, { "grad_norm": 0.7662895917892456, "learning_rate": 9.618e-05, "loss": 0.1206, "step": 4810 }, { "grad_norm": 0.7862272262573242, "learning_rate": 9.638000000000001e-05, "loss": 0.1223, "step": 4820 }, { "grad_norm": 0.9420836567878723, "learning_rate": 9.658e-05, "loss": 0.1256, "step": 4830 }, { "grad_norm": 0.7111907601356506, "learning_rate": 9.678e-05, "loss": 0.1097, "step": 4840 }, { "grad_norm": 0.7357295751571655, "learning_rate": 9.698000000000001e-05, "loss": 0.1077, "step": 4850 }, { "grad_norm": 0.674422025680542, "learning_rate": 9.718e-05, "loss": 0.0951, "step": 4860 }, { "grad_norm": 0.6723672747612, "learning_rate": 9.738e-05, "loss": 0.1164, "step": 4870 }, { "grad_norm": 0.7506921887397766, "learning_rate": 9.758000000000001e-05, "loss": 0.103, "step": 4880 }, { "grad_norm": 0.7662144303321838, "learning_rate": 9.778e-05, "loss": 0.1304, "step": 4890 }, { "grad_norm": 0.832770049571991, "learning_rate": 9.798000000000001e-05, "loss": 0.1007, "step": 4900 }, { "grad_norm": 1.6698192358016968, "learning_rate": 9.818000000000001e-05, "loss": 0.1044, "step": 4910 }, { "grad_norm": 0.5577151775360107, "learning_rate": 9.838e-05, "loss": 0.1159, "step": 4920 }, { "grad_norm": 0.8275358080863953, "learning_rate": 9.858000000000001e-05, "loss": 0.1173, "step": 4930 }, { "grad_norm": 0.5582678318023682, "learning_rate": 9.878e-05, "loss": 0.1067, "step": 4940 }, { "grad_norm": 0.8612533211708069, "learning_rate": 9.898e-05, "loss": 0.1269, "step": 4950 }, { "grad_norm": 0.6250671744346619, "learning_rate": 9.918000000000001e-05, "loss": 0.1081, "step": 4960 }, { "grad_norm": 0.6971936225891113, "learning_rate": 9.938e-05, "loss": 0.1257, "step": 4970 }, { "grad_norm": 0.6219602823257446, "learning_rate": 9.958e-05, "loss": 0.1112, "step": 4980 }, { "grad_norm": 0.8126499652862549, "learning_rate": 9.978000000000001e-05, "loss": 0.1112, "step": 4990 }, { "grad_norm": 0.7673567533493042, "learning_rate": 9.998e-05, "loss": 0.1232, "step": 5000 }, { "grad_norm": 0.571673572063446, "learning_rate": 9.999999778549045e-05, "loss": 0.124, "step": 5010 }, { "grad_norm": 1.0504188537597656, "learning_rate": 9.999999013039593e-05, "loss": 0.1156, "step": 5020 }, { "grad_norm": 0.7017810940742493, "learning_rate": 9.999997700737766e-05, "loss": 0.0969, "step": 5030 }, { "grad_norm": 0.868200421333313, "learning_rate": 9.999995841643709e-05, "loss": 0.1004, "step": 5040 }, { "grad_norm": 0.625031590461731, "learning_rate": 9.999993435757623e-05, "loss": 0.0977, "step": 5050 }, { "grad_norm": 0.6012261509895325, "learning_rate": 9.999990483079773e-05, "loss": 0.0946, "step": 5060 }, { "grad_norm": 1.0141037702560425, "learning_rate": 9.999986983610481e-05, "loss": 0.1358, "step": 5070 }, { "grad_norm": 0.9312992691993713, "learning_rate": 9.99998293735013e-05, "loss": 0.1245, "step": 5080 }, { "grad_norm": 0.8650181889533997, "learning_rate": 9.999978344299161e-05, "loss": 0.1172, "step": 5090 }, { "grad_norm": 0.7659963369369507, "learning_rate": 9.99997320445808e-05, "loss": 0.1171, "step": 5100 }, { "grad_norm": 0.7148801684379578, "learning_rate": 9.999967517827444e-05, "loss": 0.1209, "step": 5110 }, { "grad_norm": 0.7079012393951416, "learning_rate": 9.999961284407879e-05, "loss": 0.1397, "step": 5120 }, { "grad_norm": 0.6899038553237915, "learning_rate": 9.999954504200067e-05, "loss": 0.1121, "step": 5130 }, { "grad_norm": 0.753409743309021, "learning_rate": 9.999947177204744e-05, "loss": 0.1298, "step": 5140 }, { "grad_norm": 0.4882716238498688, "learning_rate": 9.999939303422718e-05, "loss": 0.1027, "step": 5150 }, { "grad_norm": 0.6429000496864319, "learning_rate": 9.999930882854847e-05, "loss": 0.1377, "step": 5160 }, { "grad_norm": 0.6824862360954285, "learning_rate": 9.999921915502051e-05, "loss": 0.1146, "step": 5170 }, { "grad_norm": 1.1304340362548828, "learning_rate": 9.99991240136531e-05, "loss": 0.112, "step": 5180 }, { "grad_norm": 0.7057089805603027, "learning_rate": 9.999902340445668e-05, "loss": 0.1396, "step": 5190 }, { "grad_norm": 0.6216955184936523, "learning_rate": 9.999891732744224e-05, "loss": 0.0966, "step": 5200 }, { "grad_norm": 0.9719898700714111, "learning_rate": 9.999880578262135e-05, "loss": 0.1194, "step": 5210 }, { "grad_norm": 0.869681179523468, "learning_rate": 9.999868877000624e-05, "loss": 0.1072, "step": 5220 }, { "grad_norm": 0.6613479852676392, "learning_rate": 9.99985662896097e-05, "loss": 0.0999, "step": 5230 }, { "grad_norm": 0.7876036763191223, "learning_rate": 9.999843834144513e-05, "loss": 0.1128, "step": 5240 }, { "grad_norm": 0.873076856136322, "learning_rate": 9.99983049255265e-05, "loss": 0.1312, "step": 5250 }, { "grad_norm": 0.6760768294334412, "learning_rate": 9.999816604186843e-05, "loss": 0.107, "step": 5260 }, { "grad_norm": 0.5783974528312683, "learning_rate": 9.999802169048609e-05, "loss": 0.1, "step": 5270 }, { "grad_norm": 0.7899062037467957, "learning_rate": 9.999787187139527e-05, "loss": 0.1225, "step": 5280 }, { "grad_norm": 0.6718594431877136, "learning_rate": 9.999771658461234e-05, "loss": 0.1118, "step": 5290 }, { "grad_norm": 0.7585247159004211, "learning_rate": 9.999755583015431e-05, "loss": 0.1019, "step": 5300 }, { "grad_norm": 0.5409383773803711, "learning_rate": 9.999738960803874e-05, "loss": 0.11, "step": 5310 }, { "grad_norm": 0.6448890566825867, "learning_rate": 9.99972179182838e-05, "loss": 0.1046, "step": 5320 }, { "grad_norm": 0.7619364857673645, "learning_rate": 9.99970407609083e-05, "loss": 0.1243, "step": 5330 }, { "grad_norm": 0.7527822852134705, "learning_rate": 9.999685813593159e-05, "loss": 0.1095, "step": 5340 }, { "grad_norm": 0.730817437171936, "learning_rate": 9.999667004337362e-05, "loss": 0.0981, "step": 5350 }, { "grad_norm": 0.6539983153343201, "learning_rate": 9.9996476483255e-05, "loss": 0.0964, "step": 5360 }, { "grad_norm": 0.8117627501487732, "learning_rate": 9.999627745559688e-05, "loss": 0.1066, "step": 5370 }, { "grad_norm": 0.5323205590248108, "learning_rate": 9.999607296042101e-05, "loss": 0.1033, "step": 5380 }, { "grad_norm": 0.9408429265022278, "learning_rate": 9.99958629977498e-05, "loss": 0.1172, "step": 5390 }, { "grad_norm": 0.8160943984985352, "learning_rate": 9.999564756760615e-05, "loss": 0.1466, "step": 5400 }, { "grad_norm": 0.7519421577453613, "learning_rate": 9.999542667001366e-05, "loss": 0.1, "step": 5410 }, { "grad_norm": 1.1171889305114746, "learning_rate": 9.999520030499647e-05, "loss": 0.1053, "step": 5420 }, { "grad_norm": 0.6684105396270752, "learning_rate": 9.999496847257936e-05, "loss": 0.1083, "step": 5430 }, { "grad_norm": 1.0621122121810913, "learning_rate": 9.999473117278764e-05, "loss": 0.1199, "step": 5440 }, { "grad_norm": 0.6110079884529114, "learning_rate": 9.999448840564731e-05, "loss": 0.0817, "step": 5450 }, { "grad_norm": 0.6806934475898743, "learning_rate": 9.999424017118488e-05, "loss": 0.089, "step": 5460 }, { "grad_norm": 0.8684186935424805, "learning_rate": 9.999398646942751e-05, "loss": 0.0981, "step": 5470 }, { "grad_norm": 0.6622333526611328, "learning_rate": 9.999372730040296e-05, "loss": 0.096, "step": 5480 }, { "grad_norm": 0.7584202885627747, "learning_rate": 9.999346266413953e-05, "loss": 0.1095, "step": 5490 }, { "grad_norm": 0.6396030783653259, "learning_rate": 9.99931925606662e-05, "loss": 0.1078, "step": 5500 }, { "grad_norm": 0.6654953360557556, "learning_rate": 9.99929169900125e-05, "loss": 0.1071, "step": 5510 }, { "grad_norm": 0.7010968327522278, "learning_rate": 9.999263595220855e-05, "loss": 0.0918, "step": 5520 }, { "grad_norm": 0.7411254048347473, "learning_rate": 9.99923494472851e-05, "loss": 0.0989, "step": 5530 }, { "grad_norm": 0.8328588008880615, "learning_rate": 9.999205747527348e-05, "loss": 0.0913, "step": 5540 }, { "grad_norm": 0.5926578640937805, "learning_rate": 9.999176003620561e-05, "loss": 0.0967, "step": 5550 }, { "grad_norm": 0.7793147563934326, "learning_rate": 9.999145713011405e-05, "loss": 0.0919, "step": 5560 }, { "grad_norm": 0.7192371487617493, "learning_rate": 9.999114875703186e-05, "loss": 0.1104, "step": 5570 }, { "grad_norm": 0.6669001579284668, "learning_rate": 9.999083491699281e-05, "loss": 0.1104, "step": 5580 }, { "grad_norm": 0.5262667536735535, "learning_rate": 9.999051561003123e-05, "loss": 0.0947, "step": 5590 }, { "grad_norm": 0.7820863723754883, "learning_rate": 9.999019083618202e-05, "loss": 0.1028, "step": 5600 }, { "grad_norm": 0.8021803498268127, "learning_rate": 9.99898605954807e-05, "loss": 0.1007, "step": 5610 }, { "grad_norm": 0.6757989525794983, "learning_rate": 9.998952488796338e-05, "loss": 0.092, "step": 5620 }, { "grad_norm": 0.5470777153968811, "learning_rate": 9.998918371366676e-05, "loss": 0.1033, "step": 5630 }, { "grad_norm": 0.675439715385437, "learning_rate": 9.99888370726282e-05, "loss": 0.1074, "step": 5640 }, { "grad_norm": 0.5893768072128296, "learning_rate": 9.998848496488556e-05, "loss": 0.1168, "step": 5650 }, { "grad_norm": 0.6934617161750793, "learning_rate": 9.998812739047736e-05, "loss": 0.1028, "step": 5660 }, { "grad_norm": 0.8610920906066895, "learning_rate": 9.99877643494427e-05, "loss": 0.0956, "step": 5670 }, { "grad_norm": 0.673172116279602, "learning_rate": 9.998739584182128e-05, "loss": 0.1033, "step": 5680 }, { "grad_norm": 0.7415969371795654, "learning_rate": 9.998702186765342e-05, "loss": 0.0996, "step": 5690 }, { "grad_norm": 0.5156886577606201, "learning_rate": 9.998664242698e-05, "loss": 0.1008, "step": 5700 }, { "grad_norm": 0.4922043979167938, "learning_rate": 9.998625751984251e-05, "loss": 0.0986, "step": 5710 }, { "grad_norm": 0.6182453632354736, "learning_rate": 9.998586714628307e-05, "loss": 0.0874, "step": 5720 }, { "grad_norm": 0.5580442547798157, "learning_rate": 9.998547130634432e-05, "loss": 0.1117, "step": 5730 }, { "grad_norm": 0.6278534531593323, "learning_rate": 9.99850700000696e-05, "loss": 0.1151, "step": 5740 }, { "grad_norm": 0.5946013927459717, "learning_rate": 9.998466322750278e-05, "loss": 0.0892, "step": 5750 }, { "grad_norm": 0.6295024156570435, "learning_rate": 9.998425098868834e-05, "loss": 0.1001, "step": 5760 }, { "grad_norm": 0.965300977230072, "learning_rate": 9.998383328367136e-05, "loss": 0.1009, "step": 5770 }, { "grad_norm": 0.8904846906661987, "learning_rate": 9.99834101124975e-05, "loss": 0.117, "step": 5780 }, { "grad_norm": 0.6652921438217163, "learning_rate": 9.998298147521309e-05, "loss": 0.0983, "step": 5790 }, { "grad_norm": 0.5645515322685242, "learning_rate": 9.998254737186496e-05, "loss": 0.1092, "step": 5800 }, { "grad_norm": 0.6095924973487854, "learning_rate": 9.99821078025006e-05, "loss": 0.0908, "step": 5810 }, { "grad_norm": 0.5557326674461365, "learning_rate": 9.998166276716807e-05, "loss": 0.0905, "step": 5820 }, { "grad_norm": 0.6554915904998779, "learning_rate": 9.998121226591606e-05, "loss": 0.1134, "step": 5830 }, { "grad_norm": 1.0518323183059692, "learning_rate": 9.998075629879382e-05, "loss": 0.0973, "step": 5840 }, { "grad_norm": 0.6105421781539917, "learning_rate": 9.99802948658512e-05, "loss": 0.0837, "step": 5850 }, { "grad_norm": 0.6071467995643616, "learning_rate": 9.99798279671387e-05, "loss": 0.1074, "step": 5860 }, { "grad_norm": 1.2553514242172241, "learning_rate": 9.997935560270734e-05, "loss": 0.1007, "step": 5870 }, { "grad_norm": 0.6012411117553711, "learning_rate": 9.997887777260879e-05, "loss": 0.1059, "step": 5880 }, { "grad_norm": 0.880332887172699, "learning_rate": 9.997839447689532e-05, "loss": 0.1057, "step": 5890 }, { "grad_norm": 0.5926403403282166, "learning_rate": 9.997790571561978e-05, "loss": 0.1142, "step": 5900 }, { "grad_norm": 0.6205403208732605, "learning_rate": 9.99774114888356e-05, "loss": 0.096, "step": 5910 }, { "grad_norm": 0.685496985912323, "learning_rate": 9.997691179659684e-05, "loss": 0.099, "step": 5920 }, { "grad_norm": 0.5288262367248535, "learning_rate": 9.997640663895815e-05, "loss": 0.0882, "step": 5930 }, { "grad_norm": 0.5793743133544922, "learning_rate": 9.997589601597477e-05, "loss": 0.0981, "step": 5940 }, { "grad_norm": 0.6262168884277344, "learning_rate": 9.997537992770252e-05, "loss": 0.0898, "step": 5950 }, { "grad_norm": 0.8627931475639343, "learning_rate": 9.997485837419788e-05, "loss": 0.0848, "step": 5960 }, { "grad_norm": 0.6946116089820862, "learning_rate": 9.997433135551786e-05, "loss": 0.0931, "step": 5970 }, { "grad_norm": 0.6595926880836487, "learning_rate": 9.997379887172009e-05, "loss": 0.1129, "step": 5980 }, { "grad_norm": 0.7903900146484375, "learning_rate": 9.997326092286281e-05, "loss": 0.1208, "step": 5990 }, { "grad_norm": 0.552056610584259, "learning_rate": 9.997271750900486e-05, "loss": 0.1332, "step": 6000 }, { "grad_norm": 0.5679735541343689, "learning_rate": 9.997216863020565e-05, "loss": 0.1008, "step": 6010 }, { "grad_norm": 0.582317590713501, "learning_rate": 9.99716142865252e-05, "loss": 0.0899, "step": 6020 }, { "grad_norm": 0.6429145932197571, "learning_rate": 9.997105447802415e-05, "loss": 0.0934, "step": 6030 }, { "grad_norm": 0.5398716926574707, "learning_rate": 9.997048920476373e-05, "loss": 0.0883, "step": 6040 }, { "grad_norm": 0.7352452278137207, "learning_rate": 9.996991846680572e-05, "loss": 0.1055, "step": 6050 }, { "grad_norm": 0.6637707352638245, "learning_rate": 9.996934226421257e-05, "loss": 0.1024, "step": 6060 }, { "grad_norm": 0.6340194344520569, "learning_rate": 9.996876059704726e-05, "loss": 0.1063, "step": 6070 }, { "grad_norm": 0.4884272515773773, "learning_rate": 9.996817346537343e-05, "loss": 0.1038, "step": 6080 }, { "grad_norm": 0.8415223360061646, "learning_rate": 9.996758086925526e-05, "loss": 0.1212, "step": 6090 }, { "grad_norm": 0.5946312546730042, "learning_rate": 9.996698280875759e-05, "loss": 0.1279, "step": 6100 }, { "grad_norm": 0.45061516761779785, "learning_rate": 9.99663792839458e-05, "loss": 0.0827, "step": 6110 }, { "grad_norm": 0.6249215602874756, "learning_rate": 9.99657702948859e-05, "loss": 0.1096, "step": 6120 }, { "grad_norm": 0.5249542593955994, "learning_rate": 9.996515584164448e-05, "loss": 0.0831, "step": 6130 }, { "grad_norm": 0.5691341161727905, "learning_rate": 9.996453592428873e-05, "loss": 0.1086, "step": 6140 }, { "grad_norm": 0.744260847568512, "learning_rate": 9.996391054288646e-05, "loss": 0.1334, "step": 6150 }, { "grad_norm": 0.6014875173568726, "learning_rate": 9.996327969750605e-05, "loss": 0.1228, "step": 6160 }, { "grad_norm": 0.748761773109436, "learning_rate": 9.996264338821649e-05, "loss": 0.1214, "step": 6170 }, { "grad_norm": 0.6651579141616821, "learning_rate": 9.996200161508735e-05, "loss": 0.1066, "step": 6180 }, { "grad_norm": 0.6225765943527222, "learning_rate": 9.996135437818885e-05, "loss": 0.1184, "step": 6190 }, { "grad_norm": 0.6369983553886414, "learning_rate": 9.996070167759175e-05, "loss": 0.0963, "step": 6200 }, { "grad_norm": 0.7307928800582886, "learning_rate": 9.996004351336743e-05, "loss": 0.1189, "step": 6210 }, { "grad_norm": 0.664073646068573, "learning_rate": 9.995937988558785e-05, "loss": 0.1, "step": 6220 }, { "grad_norm": 0.6946779489517212, "learning_rate": 9.995871079432561e-05, "loss": 0.1048, "step": 6230 }, { "grad_norm": 0.6386467218399048, "learning_rate": 9.995803623965389e-05, "loss": 0.1311, "step": 6240 }, { "grad_norm": 0.5615909695625305, "learning_rate": 9.995735622164641e-05, "loss": 0.1086, "step": 6250 }, { "grad_norm": 0.7328569293022156, "learning_rate": 9.995667074037758e-05, "loss": 0.0874, "step": 6260 }, { "grad_norm": 0.699407696723938, "learning_rate": 9.995597979592232e-05, "loss": 0.1052, "step": 6270 }, { "grad_norm": 0.506592333316803, "learning_rate": 9.995528338835625e-05, "loss": 0.0949, "step": 6280 }, { "grad_norm": 0.7981855273246765, "learning_rate": 9.995458151775547e-05, "loss": 0.0987, "step": 6290 }, { "grad_norm": 0.5718139410018921, "learning_rate": 9.995387418419677e-05, "loss": 0.1123, "step": 6300 }, { "grad_norm": 0.6734912991523743, "learning_rate": 9.99531613877575e-05, "loss": 0.116, "step": 6310 }, { "grad_norm": 0.6704980134963989, "learning_rate": 9.995244312851559e-05, "loss": 0.0968, "step": 6320 }, { "grad_norm": 0.712530255317688, "learning_rate": 9.995171940654961e-05, "loss": 0.1, "step": 6330 }, { "grad_norm": 0.7532194256782532, "learning_rate": 9.995099022193871e-05, "loss": 0.1108, "step": 6340 }, { "grad_norm": 0.613387405872345, "learning_rate": 9.995025557476261e-05, "loss": 0.093, "step": 6350 }, { "grad_norm": 0.8446703553199768, "learning_rate": 9.994951546510165e-05, "loss": 0.1042, "step": 6360 }, { "grad_norm": 0.6087359189987183, "learning_rate": 9.994876989303679e-05, "loss": 0.1041, "step": 6370 }, { "grad_norm": 0.6726990342140198, "learning_rate": 9.994801885864955e-05, "loss": 0.0944, "step": 6380 }, { "grad_norm": 0.6045747399330139, "learning_rate": 9.994726236202205e-05, "loss": 0.0963, "step": 6390 }, { "grad_norm": 0.6117050647735596, "learning_rate": 9.994650040323704e-05, "loss": 0.1143, "step": 6400 }, { "grad_norm": 0.9047026634216309, "learning_rate": 9.994573298237784e-05, "loss": 0.0872, "step": 6410 }, { "grad_norm": 0.6937757730484009, "learning_rate": 9.994496009952837e-05, "loss": 0.1012, "step": 6420 }, { "grad_norm": 0.5322405695915222, "learning_rate": 9.994418175477316e-05, "loss": 0.0982, "step": 6430 }, { "grad_norm": 0.5128275752067566, "learning_rate": 9.994339794819733e-05, "loss": 0.0919, "step": 6440 }, { "grad_norm": 0.6530842781066895, "learning_rate": 9.994260867988658e-05, "loss": 0.0856, "step": 6450 }, { "grad_norm": 0.6003933548927307, "learning_rate": 9.994181394992723e-05, "loss": 0.0936, "step": 6460 }, { "grad_norm": 0.4531678855419159, "learning_rate": 9.994101375840618e-05, "loss": 0.0862, "step": 6470 }, { "grad_norm": 0.604589581489563, "learning_rate": 9.994020810541098e-05, "loss": 0.0766, "step": 6480 }, { "grad_norm": 0.6459479331970215, "learning_rate": 9.99393969910297e-05, "loss": 0.1006, "step": 6490 }, { "grad_norm": 0.5562050342559814, "learning_rate": 9.993858041535104e-05, "loss": 0.08, "step": 6500 }, { "grad_norm": 0.7187528014183044, "learning_rate": 9.99377583784643e-05, "loss": 0.0799, "step": 6510 }, { "grad_norm": 0.5628030300140381, "learning_rate": 9.993693088045939e-05, "loss": 0.1006, "step": 6520 }, { "grad_norm": 0.9009335041046143, "learning_rate": 9.99360979214268e-05, "loss": 0.0938, "step": 6530 }, { "grad_norm": 0.6691103577613831, "learning_rate": 9.99352595014576e-05, "loss": 0.105, "step": 6540 }, { "grad_norm": 0.7672010064125061, "learning_rate": 9.993441562064354e-05, "loss": 0.0993, "step": 6550 }, { "grad_norm": 0.45356839895248413, "learning_rate": 9.993356627907685e-05, "loss": 0.0932, "step": 6560 }, { "grad_norm": 0.7190225124359131, "learning_rate": 9.99327114768504e-05, "loss": 0.1033, "step": 6570 }, { "grad_norm": 0.8044756650924683, "learning_rate": 9.99318512140577e-05, "loss": 0.0882, "step": 6580 }, { "grad_norm": 0.7842721939086914, "learning_rate": 9.993098549079284e-05, "loss": 0.1014, "step": 6590 }, { "grad_norm": 0.6256402730941772, "learning_rate": 9.993011430715047e-05, "loss": 0.1049, "step": 6600 }, { "grad_norm": 0.5213868021965027, "learning_rate": 9.992923766322586e-05, "loss": 0.0814, "step": 6610 }, { "grad_norm": 0.5658170580863953, "learning_rate": 9.99283555591149e-05, "loss": 0.0752, "step": 6620 }, { "grad_norm": 0.7389668226242065, "learning_rate": 9.992746799491404e-05, "loss": 0.0998, "step": 6630 }, { "grad_norm": 0.5209572315216064, "learning_rate": 9.992657497072033e-05, "loss": 0.0808, "step": 6640 }, { "grad_norm": 0.6544954180717468, "learning_rate": 9.992567648663147e-05, "loss": 0.1048, "step": 6650 }, { "grad_norm": 0.6839659214019775, "learning_rate": 9.992477254274568e-05, "loss": 0.0912, "step": 6660 }, { "grad_norm": 0.6790168285369873, "learning_rate": 9.992386313916183e-05, "loss": 0.0951, "step": 6670 }, { "grad_norm": 0.6363322138786316, "learning_rate": 9.992294827597934e-05, "loss": 0.1106, "step": 6680 }, { "grad_norm": 0.611175537109375, "learning_rate": 9.992202795329831e-05, "loss": 0.1093, "step": 6690 }, { "grad_norm": 0.6597739458084106, "learning_rate": 9.992110217121936e-05, "loss": 0.1003, "step": 6700 }, { "grad_norm": 0.7732040286064148, "learning_rate": 9.992017092984372e-05, "loss": 0.1031, "step": 6710 }, { "grad_norm": 0.7767745852470398, "learning_rate": 9.991923422927326e-05, "loss": 0.1165, "step": 6720 }, { "grad_norm": 0.5492589473724365, "learning_rate": 9.991829206961037e-05, "loss": 0.0936, "step": 6730 }, { "grad_norm": 0.7316416501998901, "learning_rate": 9.991734445095813e-05, "loss": 0.1255, "step": 6740 }, { "grad_norm": 0.6716341972351074, "learning_rate": 9.991639137342015e-05, "loss": 0.1005, "step": 6750 }, { "grad_norm": 0.6477833390235901, "learning_rate": 9.991543283710064e-05, "loss": 0.0938, "step": 6760 }, { "grad_norm": 0.6821715235710144, "learning_rate": 9.991446884210445e-05, "loss": 0.0868, "step": 6770 }, { "grad_norm": 0.5952050685882568, "learning_rate": 9.9913499388537e-05, "loss": 0.0875, "step": 6780 }, { "grad_norm": 0.46557652950286865, "learning_rate": 9.99125244765043e-05, "loss": 0.0792, "step": 6790 }, { "grad_norm": 0.7065438628196716, "learning_rate": 9.991154410611296e-05, "loss": 0.0998, "step": 6800 }, { "grad_norm": 0.8418972492218018, "learning_rate": 9.99105582774702e-05, "loss": 0.0947, "step": 6810 }, { "grad_norm": 0.6722028851509094, "learning_rate": 9.990956699068384e-05, "loss": 0.1055, "step": 6820 }, { "grad_norm": 0.6242336630821228, "learning_rate": 9.990857024586224e-05, "loss": 0.0922, "step": 6830 }, { "grad_norm": 0.46954989433288574, "learning_rate": 9.990756804311446e-05, "loss": 0.0968, "step": 6840 }, { "grad_norm": 0.5034964680671692, "learning_rate": 9.990656038255006e-05, "loss": 0.1005, "step": 6850 }, { "grad_norm": 0.5462385416030884, "learning_rate": 9.990554726427926e-05, "loss": 0.091, "step": 6860 }, { "grad_norm": 0.6190842986106873, "learning_rate": 9.990452868841284e-05, "loss": 0.0991, "step": 6870 }, { "grad_norm": 0.7405328750610352, "learning_rate": 9.99035046550622e-05, "loss": 0.0987, "step": 6880 }, { "grad_norm": 0.5715514421463013, "learning_rate": 9.99024751643393e-05, "loss": 0.092, "step": 6890 }, { "grad_norm": 0.6506269574165344, "learning_rate": 9.990144021635677e-05, "loss": 0.101, "step": 6900 }, { "grad_norm": 0.7333269119262695, "learning_rate": 9.990039981122775e-05, "loss": 0.082, "step": 6910 }, { "grad_norm": 0.5328932404518127, "learning_rate": 9.989935394906602e-05, "loss": 0.087, "step": 6920 }, { "grad_norm": 0.6102286577224731, "learning_rate": 9.989830262998598e-05, "loss": 0.0913, "step": 6930 }, { "grad_norm": 0.6444092392921448, "learning_rate": 9.989724585410259e-05, "loss": 0.0821, "step": 6940 }, { "grad_norm": 0.5598540902137756, "learning_rate": 9.989618362153139e-05, "loss": 0.0792, "step": 6950 }, { "grad_norm": 0.5981823205947876, "learning_rate": 9.989511593238859e-05, "loss": 0.0926, "step": 6960 }, { "grad_norm": 0.5058404207229614, "learning_rate": 9.98940427867909e-05, "loss": 0.0964, "step": 6970 }, { "grad_norm": 0.6949524283409119, "learning_rate": 9.989296418485573e-05, "loss": 0.1015, "step": 6980 }, { "grad_norm": 0.7824429273605347, "learning_rate": 9.989188012670101e-05, "loss": 0.093, "step": 6990 }, { "grad_norm": 0.6345438361167908, "learning_rate": 9.989079061244528e-05, "loss": 0.0894, "step": 7000 }, { "grad_norm": 0.5742673277854919, "learning_rate": 9.988969564220769e-05, "loss": 0.0947, "step": 7010 }, { "grad_norm": 0.5603479146957397, "learning_rate": 9.988859521610801e-05, "loss": 0.0792, "step": 7020 }, { "grad_norm": 0.6069443821907043, "learning_rate": 9.988748933426656e-05, "loss": 0.0988, "step": 7030 }, { "grad_norm": 0.5717985033988953, "learning_rate": 9.988637799680428e-05, "loss": 0.086, "step": 7040 }, { "grad_norm": 0.5817700624465942, "learning_rate": 9.98852612038427e-05, "loss": 0.0749, "step": 7050 }, { "grad_norm": 0.5049903988838196, "learning_rate": 9.988413895550397e-05, "loss": 0.0863, "step": 7060 }, { "grad_norm": 0.4425566494464874, "learning_rate": 9.98830112519108e-05, "loss": 0.0956, "step": 7070 }, { "grad_norm": 0.5840396881103516, "learning_rate": 9.98818780931865e-05, "loss": 0.1057, "step": 7080 }, { "grad_norm": 0.6256061792373657, "learning_rate": 9.988073947945502e-05, "loss": 0.0922, "step": 7090 }, { "grad_norm": 0.7041241526603699, "learning_rate": 9.987959541084087e-05, "loss": 0.0821, "step": 7100 }, { "grad_norm": 0.6402825713157654, "learning_rate": 9.987844588746915e-05, "loss": 0.0962, "step": 7110 }, { "grad_norm": 0.5722049474716187, "learning_rate": 9.987729090946558e-05, "loss": 0.0796, "step": 7120 }, { "grad_norm": 0.5556443333625793, "learning_rate": 9.987613047695647e-05, "loss": 0.082, "step": 7130 }, { "grad_norm": 0.45223385095596313, "learning_rate": 9.987496459006871e-05, "loss": 0.106, "step": 7140 }, { "grad_norm": 0.43291375041007996, "learning_rate": 9.987379324892982e-05, "loss": 0.0959, "step": 7150 }, { "grad_norm": 0.7554875612258911, "learning_rate": 9.987261645366788e-05, "loss": 0.0892, "step": 7160 }, { "grad_norm": 0.528753936290741, "learning_rate": 9.987143420441158e-05, "loss": 0.082, "step": 7170 }, { "grad_norm": 0.5725471377372742, "learning_rate": 9.987024650129022e-05, "loss": 0.0749, "step": 7180 }, { "grad_norm": 0.439857542514801, "learning_rate": 9.986905334443368e-05, "loss": 0.0938, "step": 7190 }, { "grad_norm": 0.6234448552131653, "learning_rate": 9.986785473397245e-05, "loss": 0.0862, "step": 7200 }, { "grad_norm": 0.7863122820854187, "learning_rate": 9.98666506700376e-05, "loss": 0.0859, "step": 7210 }, { "grad_norm": 0.8754162788391113, "learning_rate": 9.986544115276081e-05, "loss": 0.0876, "step": 7220 }, { "grad_norm": 1.1517103910446167, "learning_rate": 9.986422618227433e-05, "loss": 0.0919, "step": 7230 }, { "grad_norm": 0.7141802906990051, "learning_rate": 9.986300575871106e-05, "loss": 0.1039, "step": 7240 }, { "grad_norm": 0.585526704788208, "learning_rate": 9.986177988220444e-05, "loss": 0.0823, "step": 7250 }, { "grad_norm": 0.558699369430542, "learning_rate": 9.986054855288856e-05, "loss": 0.0979, "step": 7260 }, { "grad_norm": 0.47092384099960327, "learning_rate": 9.985931177089802e-05, "loss": 0.0881, "step": 7270 }, { "grad_norm": 0.48922544717788696, "learning_rate": 9.985806953636814e-05, "loss": 0.0804, "step": 7280 }, { "grad_norm": 0.5936744809150696, "learning_rate": 9.985682184943471e-05, "loss": 0.0832, "step": 7290 }, { "grad_norm": 0.6779588460922241, "learning_rate": 9.98555687102342e-05, "loss": 0.0879, "step": 7300 }, { "grad_norm": 0.46367159485816956, "learning_rate": 9.985431011890367e-05, "loss": 0.0861, "step": 7310 }, { "grad_norm": 0.5895957946777344, "learning_rate": 9.985304607558075e-05, "loss": 0.0839, "step": 7320 }, { "grad_norm": 0.7567394375801086, "learning_rate": 9.985177658040364e-05, "loss": 0.0879, "step": 7330 }, { "grad_norm": 0.46615034341812134, "learning_rate": 9.985050163351119e-05, "loss": 0.0881, "step": 7340 }, { "grad_norm": 0.473219633102417, "learning_rate": 9.984922123504286e-05, "loss": 0.0961, "step": 7350 }, { "grad_norm": 0.7762292623519897, "learning_rate": 9.984793538513862e-05, "loss": 0.0985, "step": 7360 }, { "grad_norm": 0.6828744411468506, "learning_rate": 9.984664408393912e-05, "loss": 0.1036, "step": 7370 }, { "grad_norm": 0.5774598717689514, "learning_rate": 9.984534733158556e-05, "loss": 0.1001, "step": 7380 }, { "grad_norm": 0.6548837423324585, "learning_rate": 9.984404512821977e-05, "loss": 0.0948, "step": 7390 }, { "grad_norm": 0.7003796100616455, "learning_rate": 9.984273747398411e-05, "loss": 0.0964, "step": 7400 }, { "grad_norm": 0.6236117482185364, "learning_rate": 9.984142436902165e-05, "loss": 0.0922, "step": 7410 }, { "grad_norm": 0.6642976403236389, "learning_rate": 9.984010581347596e-05, "loss": 0.1166, "step": 7420 }, { "grad_norm": 0.5233953595161438, "learning_rate": 9.983878180749121e-05, "loss": 0.0944, "step": 7430 }, { "grad_norm": 0.48562124371528625, "learning_rate": 9.983745235121222e-05, "loss": 0.0877, "step": 7440 }, { "grad_norm": 0.9721300005912781, "learning_rate": 9.983611744478438e-05, "loss": 0.0884, "step": 7450 }, { "grad_norm": 0.7351133823394775, "learning_rate": 9.983477708835365e-05, "loss": 0.096, "step": 7460 }, { "grad_norm": 0.8834327459335327, "learning_rate": 9.983343128206664e-05, "loss": 0.1052, "step": 7470 }, { "grad_norm": 1.6350648403167725, "learning_rate": 9.983208002607049e-05, "loss": 0.1152, "step": 7480 }, { "grad_norm": 1.2093350887298584, "learning_rate": 9.9830723320513e-05, "loss": 0.1189, "step": 7490 }, { "grad_norm": 0.7292709946632385, "learning_rate": 9.982936116554254e-05, "loss": 0.1073, "step": 7500 }, { "grad_norm": 0.686250627040863, "learning_rate": 9.982799356130803e-05, "loss": 0.1041, "step": 7510 }, { "grad_norm": 0.6559327840805054, "learning_rate": 9.982662050795908e-05, "loss": 0.0929, "step": 7520 }, { "grad_norm": 0.8003918528556824, "learning_rate": 9.982524200564583e-05, "loss": 0.091, "step": 7530 }, { "grad_norm": 0.9432471990585327, "learning_rate": 9.982385805451901e-05, "loss": 0.1028, "step": 7540 }, { "grad_norm": 0.6471906304359436, "learning_rate": 9.982246865472998e-05, "loss": 0.0846, "step": 7550 }, { "grad_norm": 0.5283576846122742, "learning_rate": 9.982107380643069e-05, "loss": 0.0964, "step": 7560 }, { "grad_norm": 0.42896926403045654, "learning_rate": 9.981967350977368e-05, "loss": 0.1116, "step": 7570 }, { "grad_norm": 0.6741037964820862, "learning_rate": 9.981826776491208e-05, "loss": 0.1135, "step": 7580 }, { "grad_norm": 0.6610816121101379, "learning_rate": 9.98168565719996e-05, "loss": 0.097, "step": 7590 }, { "grad_norm": 0.6739612221717834, "learning_rate": 9.98154399311906e-05, "loss": 0.1066, "step": 7600 }, { "grad_norm": 0.6314200162887573, "learning_rate": 9.981401784263997e-05, "loss": 0.1006, "step": 7610 }, { "grad_norm": 0.46295303106307983, "learning_rate": 9.981259030650326e-05, "loss": 0.1023, "step": 7620 }, { "grad_norm": 0.8510385155677795, "learning_rate": 9.981115732293655e-05, "loss": 0.0882, "step": 7630 }, { "grad_norm": 0.5693060755729675, "learning_rate": 9.980971889209659e-05, "loss": 0.0886, "step": 7640 }, { "grad_norm": 0.5053068995475769, "learning_rate": 9.980827501414064e-05, "loss": 0.0784, "step": 7650 }, { "grad_norm": 0.8164373636245728, "learning_rate": 9.980682568922663e-05, "loss": 0.0935, "step": 7660 }, { "grad_norm": 0.8609005808830261, "learning_rate": 9.980537091751304e-05, "loss": 0.1006, "step": 7670 }, { "grad_norm": 0.5309481620788574, "learning_rate": 9.980391069915897e-05, "loss": 0.0992, "step": 7680 }, { "grad_norm": 0.6371475458145142, "learning_rate": 9.98024450343241e-05, "loss": 0.0902, "step": 7690 }, { "grad_norm": 0.8495932221412659, "learning_rate": 9.980097392316872e-05, "loss": 0.0863, "step": 7700 }, { "grad_norm": 0.633476197719574, "learning_rate": 9.97994973658537e-05, "loss": 0.1096, "step": 7710 }, { "grad_norm": 0.7507010102272034, "learning_rate": 9.979801536254054e-05, "loss": 0.1, "step": 7720 }, { "grad_norm": 0.5972937941551208, "learning_rate": 9.979652791339127e-05, "loss": 0.0904, "step": 7730 }, { "grad_norm": 0.5872116684913635, "learning_rate": 9.97950350185686e-05, "loss": 0.0949, "step": 7740 }, { "grad_norm": 0.5933674573898315, "learning_rate": 9.979353667823574e-05, "loss": 0.0805, "step": 7750 }, { "grad_norm": 0.5908458828926086, "learning_rate": 9.979203289255658e-05, "loss": 0.0957, "step": 7760 }, { "grad_norm": 0.5762913227081299, "learning_rate": 9.979052366169557e-05, "loss": 0.1002, "step": 7770 }, { "grad_norm": 0.5593842267990112, "learning_rate": 9.978900898581775e-05, "loss": 0.1044, "step": 7780 }, { "grad_norm": 0.822117030620575, "learning_rate": 9.978748886508875e-05, "loss": 0.1165, "step": 7790 }, { "grad_norm": 0.6309576630592346, "learning_rate": 9.978596329967484e-05, "loss": 0.0941, "step": 7800 }, { "grad_norm": 0.6345176100730896, "learning_rate": 9.978443228974284e-05, "loss": 0.0825, "step": 7810 }, { "grad_norm": 0.6218083500862122, "learning_rate": 9.978289583546015e-05, "loss": 0.0777, "step": 7820 }, { "grad_norm": 0.6032912731170654, "learning_rate": 9.978135393699484e-05, "loss": 0.108, "step": 7830 }, { "grad_norm": 0.6971540451049805, "learning_rate": 9.977980659451548e-05, "loss": 0.1002, "step": 7840 }, { "grad_norm": 0.7098552584648132, "learning_rate": 9.977825380819135e-05, "loss": 0.1028, "step": 7850 }, { "grad_norm": 0.5880926251411438, "learning_rate": 9.97766955781922e-05, "loss": 0.0947, "step": 7860 }, { "grad_norm": 0.6247857809066772, "learning_rate": 9.977513190468848e-05, "loss": 0.1099, "step": 7870 }, { "grad_norm": 0.5872806906700134, "learning_rate": 9.977356278785116e-05, "loss": 0.0996, "step": 7880 }, { "grad_norm": 0.5779925584793091, "learning_rate": 9.977198822785184e-05, "loss": 0.0895, "step": 7890 }, { "grad_norm": 0.5156219005584717, "learning_rate": 9.977040822486273e-05, "loss": 0.1015, "step": 7900 }, { "grad_norm": 0.5538572669029236, "learning_rate": 9.97688227790566e-05, "loss": 0.0995, "step": 7910 }, { "grad_norm": 0.6293849349021912, "learning_rate": 9.976723189060684e-05, "loss": 0.1278, "step": 7920 }, { "grad_norm": 0.4663901925086975, "learning_rate": 9.976563555968742e-05, "loss": 0.0929, "step": 7930 }, { "grad_norm": 0.5087030529975891, "learning_rate": 9.976403378647292e-05, "loss": 0.0905, "step": 7940 }, { "grad_norm": 0.6534384489059448, "learning_rate": 9.97624265711385e-05, "loss": 0.0944, "step": 7950 }, { "grad_norm": 0.8637900352478027, "learning_rate": 9.976081391385993e-05, "loss": 0.083, "step": 7960 }, { "grad_norm": 0.6322811841964722, "learning_rate": 9.975919581481356e-05, "loss": 0.0925, "step": 7970 }, { "grad_norm": 0.479742169380188, "learning_rate": 9.975757227417634e-05, "loss": 0.0835, "step": 7980 }, { "grad_norm": 0.657014787197113, "learning_rate": 9.975594329212586e-05, "loss": 0.0837, "step": 7990 }, { "grad_norm": 0.6533064246177673, "learning_rate": 9.97543088688402e-05, "loss": 0.087, "step": 8000 }, { "grad_norm": 0.5715410113334656, "learning_rate": 9.975266900449814e-05, "loss": 0.0889, "step": 8010 }, { "grad_norm": 0.6719250679016113, "learning_rate": 9.975102369927898e-05, "loss": 0.0802, "step": 8020 }, { "grad_norm": 0.5059524774551392, "learning_rate": 9.974937295336269e-05, "loss": 0.0839, "step": 8030 }, { "grad_norm": 0.7160252332687378, "learning_rate": 9.974771676692975e-05, "loss": 0.0806, "step": 8040 }, { "grad_norm": 0.6278270483016968, "learning_rate": 9.974605514016131e-05, "loss": 0.0789, "step": 8050 }, { "grad_norm": 0.49910345673561096, "learning_rate": 9.974438807323907e-05, "loss": 0.0818, "step": 8060 }, { "grad_norm": 0.5275735855102539, "learning_rate": 9.974271556634535e-05, "loss": 0.0759, "step": 8070 }, { "grad_norm": 0.6295331716537476, "learning_rate": 9.974103761966302e-05, "loss": 0.1032, "step": 8080 }, { "grad_norm": 0.537682294845581, "learning_rate": 9.973935423337563e-05, "loss": 0.094, "step": 8090 }, { "grad_norm": 0.5466257333755493, "learning_rate": 9.973766540766722e-05, "loss": 0.0957, "step": 8100 }, { "grad_norm": 0.7586444020271301, "learning_rate": 9.97359711427225e-05, "loss": 0.094, "step": 8110 }, { "grad_norm": 0.6739636659622192, "learning_rate": 9.973427143872677e-05, "loss": 0.103, "step": 8120 }, { "grad_norm": 0.7197970747947693, "learning_rate": 9.973256629586589e-05, "loss": 0.11, "step": 8130 }, { "grad_norm": 0.6140158176422119, "learning_rate": 9.973085571432632e-05, "loss": 0.1121, "step": 8140 }, { "grad_norm": 0.4605438709259033, "learning_rate": 9.972913969429513e-05, "loss": 0.0852, "step": 8150 }, { "grad_norm": 0.5958495140075684, "learning_rate": 9.972741823596e-05, "loss": 0.1008, "step": 8160 }, { "grad_norm": 0.5503720045089722, "learning_rate": 9.972569133950917e-05, "loss": 0.0939, "step": 8170 }, { "grad_norm": 0.37126782536506653, "learning_rate": 9.972395900513151e-05, "loss": 0.0801, "step": 8180 }, { "grad_norm": 0.6265949010848999, "learning_rate": 9.972222123301645e-05, "loss": 0.099, "step": 8190 }, { "grad_norm": 0.6566608548164368, "learning_rate": 9.972047802335403e-05, "loss": 0.0847, "step": 8200 }, { "grad_norm": 0.5813020467758179, "learning_rate": 9.971872937633488e-05, "loss": 0.0868, "step": 8210 }, { "grad_norm": 0.8228596448898315, "learning_rate": 9.971697529215024e-05, "loss": 0.0827, "step": 8220 }, { "grad_norm": 0.6258443593978882, "learning_rate": 9.971521577099192e-05, "loss": 0.0866, "step": 8230 }, { "grad_norm": 0.4830636978149414, "learning_rate": 9.971345081305236e-05, "loss": 0.0897, "step": 8240 }, { "grad_norm": 0.6605749130249023, "learning_rate": 9.971168041852456e-05, "loss": 0.0975, "step": 8250 }, { "grad_norm": 0.5807731747627258, "learning_rate": 9.970990458760215e-05, "loss": 0.0936, "step": 8260 }, { "grad_norm": 0.585645318031311, "learning_rate": 9.970812332047929e-05, "loss": 0.0769, "step": 8270 }, { "grad_norm": 0.5284704566001892, "learning_rate": 9.97063366173508e-05, "loss": 0.0933, "step": 8280 }, { "grad_norm": 0.6122317910194397, "learning_rate": 9.970454447841207e-05, "loss": 0.092, "step": 8290 }, { "grad_norm": 0.6643040776252747, "learning_rate": 9.970274690385909e-05, "loss": 0.1011, "step": 8300 }, { "grad_norm": 0.906521201133728, "learning_rate": 9.970094389388844e-05, "loss": 0.1052, "step": 8310 }, { "grad_norm": 0.8811122179031372, "learning_rate": 9.969913544869728e-05, "loss": 0.0996, "step": 8320 }, { "grad_norm": 0.6487271189689636, "learning_rate": 9.96973215684834e-05, "loss": 0.0955, "step": 8330 }, { "grad_norm": 0.6385776400566101, "learning_rate": 9.969550225344513e-05, "loss": 0.0831, "step": 8340 }, { "grad_norm": 0.6224812865257263, "learning_rate": 9.969367750378147e-05, "loss": 0.102, "step": 8350 }, { "grad_norm": 0.9452023506164551, "learning_rate": 9.969184731969194e-05, "loss": 0.0834, "step": 8360 }, { "grad_norm": 0.6463735699653625, "learning_rate": 9.96900117013767e-05, "loss": 0.0842, "step": 8370 }, { "grad_norm": 0.6627678275108337, "learning_rate": 9.96881706490365e-05, "loss": 0.0922, "step": 8380 }, { "grad_norm": 0.5583174824714661, "learning_rate": 9.968632416287265e-05, "loss": 0.0846, "step": 8390 }, { "grad_norm": 0.42793625593185425, "learning_rate": 9.96844722430871e-05, "loss": 0.0814, "step": 8400 }, { "grad_norm": 0.6076598167419434, "learning_rate": 9.968261488988235e-05, "loss": 0.0874, "step": 8410 }, { "grad_norm": 0.36135774850845337, "learning_rate": 9.968075210346155e-05, "loss": 0.0902, "step": 8420 }, { "grad_norm": 0.7143969535827637, "learning_rate": 9.967888388402839e-05, "loss": 0.1012, "step": 8430 }, { "grad_norm": 0.500761866569519, "learning_rate": 9.967701023178717e-05, "loss": 0.0936, "step": 8440 }, { "grad_norm": 0.509012758731842, "learning_rate": 9.967513114694282e-05, "loss": 0.08, "step": 8450 }, { "grad_norm": 0.5424039363861084, "learning_rate": 9.967324662970079e-05, "loss": 0.0776, "step": 8460 }, { "grad_norm": 0.6973658800125122, "learning_rate": 9.96713566802672e-05, "loss": 0.0746, "step": 8470 }, { "grad_norm": 0.4733758270740509, "learning_rate": 9.966946129884873e-05, "loss": 0.0863, "step": 8480 }, { "grad_norm": 0.5508481860160828, "learning_rate": 9.966756048565265e-05, "loss": 0.0811, "step": 8490 }, { "grad_norm": 0.6284762024879456, "learning_rate": 9.966565424088681e-05, "loss": 0.0978, "step": 8500 }, { "grad_norm": 0.611441969871521, "learning_rate": 9.96637425647597e-05, "loss": 0.0807, "step": 8510 }, { "grad_norm": 0.5389959216117859, "learning_rate": 9.966182545748038e-05, "loss": 0.088, "step": 8520 }, { "grad_norm": 0.49964818358421326, "learning_rate": 9.96599029192585e-05, "loss": 0.0894, "step": 8530 }, { "grad_norm": 0.5434533357620239, "learning_rate": 9.965797495030428e-05, "loss": 0.0841, "step": 8540 }, { "grad_norm": 0.6658257246017456, "learning_rate": 9.96560415508286e-05, "loss": 0.0842, "step": 8550 }, { "grad_norm": 0.48628559708595276, "learning_rate": 9.965410272104286e-05, "loss": 0.0872, "step": 8560 }, { "grad_norm": 0.744028627872467, "learning_rate": 9.96521584611591e-05, "loss": 0.1188, "step": 8570 }, { "grad_norm": 0.6700451374053955, "learning_rate": 9.965020877138994e-05, "loss": 0.1041, "step": 8580 }, { "grad_norm": 0.4791503846645355, "learning_rate": 9.964825365194861e-05, "loss": 0.0954, "step": 8590 }, { "grad_norm": 0.6344658136367798, "learning_rate": 9.96462931030489e-05, "loss": 0.0712, "step": 8600 }, { "grad_norm": 0.5308001041412354, "learning_rate": 9.96443271249052e-05, "loss": 0.0715, "step": 8610 }, { "grad_norm": 0.7477872967720032, "learning_rate": 9.964235571773255e-05, "loss": 0.0919, "step": 8620 }, { "grad_norm": 0.6497483849525452, "learning_rate": 9.96403788817465e-05, "loss": 0.0911, "step": 8630 }, { "grad_norm": 0.5263763070106506, "learning_rate": 9.963839661716325e-05, "loss": 0.0854, "step": 8640 }, { "grad_norm": 0.876334547996521, "learning_rate": 9.963640892419958e-05, "loss": 0.0876, "step": 8650 }, { "grad_norm": 0.8078854084014893, "learning_rate": 9.963441580307286e-05, "loss": 0.0885, "step": 8660 }, { "grad_norm": 0.5878669023513794, "learning_rate": 9.963241725400104e-05, "loss": 0.0855, "step": 8670 }, { "grad_norm": 0.8087043762207031, "learning_rate": 9.963041327720271e-05, "loss": 0.0864, "step": 8680 }, { "grad_norm": 0.519523561000824, "learning_rate": 9.962840387289697e-05, "loss": 0.083, "step": 8690 }, { "grad_norm": 0.5021026134490967, "learning_rate": 9.962638904130363e-05, "loss": 0.0819, "step": 8700 }, { "grad_norm": 0.5075709223747253, "learning_rate": 9.962436878264298e-05, "loss": 0.0925, "step": 8710 }, { "grad_norm": 0.765192449092865, "learning_rate": 9.962234309713598e-05, "loss": 0.1041, "step": 8720 }, { "grad_norm": 0.7451958060264587, "learning_rate": 9.962031198500414e-05, "loss": 0.1008, "step": 8730 }, { "grad_norm": 0.5206911563873291, "learning_rate": 9.961827544646958e-05, "loss": 0.0828, "step": 8740 }, { "grad_norm": 0.6317855715751648, "learning_rate": 9.961623348175501e-05, "loss": 0.0937, "step": 8750 }, { "grad_norm": 1.224263072013855, "learning_rate": 9.961418609108377e-05, "loss": 0.0976, "step": 8760 }, { "grad_norm": 0.5676149129867554, "learning_rate": 9.961213327467971e-05, "loss": 0.0867, "step": 8770 }, { "grad_norm": 0.5696203708648682, "learning_rate": 9.961007503276736e-05, "loss": 0.0889, "step": 8780 }, { "grad_norm": 1.9527502059936523, "learning_rate": 9.960801136557179e-05, "loss": 0.073, "step": 8790 }, { "grad_norm": 0.44497236609458923, "learning_rate": 9.960594227331866e-05, "loss": 0.076, "step": 8800 }, { "grad_norm": 0.5620980858802795, "learning_rate": 9.960386775623429e-05, "loss": 0.0898, "step": 8810 }, { "grad_norm": 0.6256392598152161, "learning_rate": 9.96017878145455e-05, "loss": 0.0977, "step": 8820 }, { "grad_norm": 0.6145468950271606, "learning_rate": 9.959970244847977e-05, "loss": 0.107, "step": 8830 }, { "grad_norm": 0.9680429100990295, "learning_rate": 9.959761165826518e-05, "loss": 0.1032, "step": 8840 }, { "grad_norm": 0.5816349983215332, "learning_rate": 9.959551544413033e-05, "loss": 0.0967, "step": 8850 }, { "grad_norm": 0.5254432559013367, "learning_rate": 9.959341380630448e-05, "loss": 0.091, "step": 8860 }, { "grad_norm": 0.8596649169921875, "learning_rate": 9.959130674501746e-05, "loss": 0.1022, "step": 8870 }, { "grad_norm": 1.1022839546203613, "learning_rate": 9.958919426049968e-05, "loss": 0.0983, "step": 8880 }, { "grad_norm": 0.5949689745903015, "learning_rate": 9.958707635298219e-05, "loss": 0.0987, "step": 8890 }, { "grad_norm": 0.555210292339325, "learning_rate": 9.958495302269657e-05, "loss": 0.0902, "step": 8900 }, { "grad_norm": 0.6119750738143921, "learning_rate": 9.958282426987503e-05, "loss": 0.0906, "step": 8910 }, { "grad_norm": 0.631633996963501, "learning_rate": 9.95806900947504e-05, "loss": 0.1034, "step": 8920 }, { "grad_norm": 0.6859337687492371, "learning_rate": 9.957855049755604e-05, "loss": 0.0789, "step": 8930 }, { "grad_norm": 0.5083649754524231, "learning_rate": 9.957640547852593e-05, "loss": 0.0858, "step": 8940 }, { "grad_norm": 0.5550873279571533, "learning_rate": 9.957425503789466e-05, "loss": 0.0837, "step": 8950 }, { "grad_norm": 0.6154430508613586, "learning_rate": 9.957209917589738e-05, "loss": 0.0744, "step": 8960 }, { "grad_norm": 0.7694313526153564, "learning_rate": 9.956993789276987e-05, "loss": 0.0725, "step": 8970 }, { "grad_norm": 0.5354666709899902, "learning_rate": 9.956777118874847e-05, "loss": 0.0794, "step": 8980 }, { "grad_norm": 0.8294225931167603, "learning_rate": 9.956559906407016e-05, "loss": 0.108, "step": 8990 }, { "grad_norm": 0.8610910773277283, "learning_rate": 9.956342151897245e-05, "loss": 0.0839, "step": 9000 }, { "grad_norm": 0.6283301115036011, "learning_rate": 9.956123855369346e-05, "loss": 0.0899, "step": 9010 }, { "grad_norm": 0.7831689715385437, "learning_rate": 9.955905016847196e-05, "loss": 0.0959, "step": 9020 }, { "grad_norm": 0.74435955286026, "learning_rate": 9.955685636354723e-05, "loss": 0.1005, "step": 9030 }, { "grad_norm": 0.6143853068351746, "learning_rate": 9.95546571391592e-05, "loss": 0.0982, "step": 9040 }, { "grad_norm": 0.6173670291900635, "learning_rate": 9.955245249554837e-05, "loss": 0.0765, "step": 9050 }, { "grad_norm": 0.8999009132385254, "learning_rate": 9.955024243295582e-05, "loss": 0.1004, "step": 9060 }, { "grad_norm": 0.5042170882225037, "learning_rate": 9.954802695162328e-05, "loss": 0.0728, "step": 9070 }, { "grad_norm": 0.5412552952766418, "learning_rate": 9.954580605179302e-05, "loss": 0.093, "step": 9080 }, { "grad_norm": 0.6816517114639282, "learning_rate": 9.954357973370788e-05, "loss": 0.0854, "step": 9090 }, { "grad_norm": 0.6152529120445251, "learning_rate": 9.954134799761135e-05, "loss": 0.0985, "step": 9100 }, { "grad_norm": 0.5686957836151123, "learning_rate": 9.953911084374748e-05, "loss": 0.0849, "step": 9110 }, { "grad_norm": 0.5237109661102295, "learning_rate": 9.953686827236093e-05, "loss": 0.0812, "step": 9120 }, { "grad_norm": 0.6048642992973328, "learning_rate": 9.953462028369695e-05, "loss": 0.0889, "step": 9130 }, { "grad_norm": 0.4534448981285095, "learning_rate": 9.953236687800136e-05, "loss": 0.0792, "step": 9140 }, { "grad_norm": 0.5696825981140137, "learning_rate": 9.95301080555206e-05, "loss": 0.095, "step": 9150 }, { "grad_norm": 0.7312986850738525, "learning_rate": 9.952784381650171e-05, "loss": 0.0764, "step": 9160 }, { "grad_norm": 0.5395380854606628, "learning_rate": 9.952557416119226e-05, "loss": 0.0791, "step": 9170 }, { "grad_norm": 0.431795597076416, "learning_rate": 9.95232990898405e-05, "loss": 0.0745, "step": 9180 }, { "grad_norm": 0.4831680953502655, "learning_rate": 9.95210186026952e-05, "loss": 0.0713, "step": 9190 }, { "grad_norm": 0.486198365688324, "learning_rate": 9.951873270000576e-05, "loss": 0.0775, "step": 9200 }, { "grad_norm": 0.4675604999065399, "learning_rate": 9.951644138202216e-05, "loss": 0.0993, "step": 9210 }, { "grad_norm": 0.841275155544281, "learning_rate": 9.951414464899498e-05, "loss": 0.1058, "step": 9220 }, { "grad_norm": 0.6240523457527161, "learning_rate": 9.951184250117538e-05, "loss": 0.101, "step": 9230 }, { "grad_norm": 0.5407705307006836, "learning_rate": 9.950953493881513e-05, "loss": 0.0947, "step": 9240 }, { "grad_norm": 0.6073134541511536, "learning_rate": 9.950722196216658e-05, "loss": 0.084, "step": 9250 }, { "grad_norm": 0.5614350438117981, "learning_rate": 9.950490357148265e-05, "loss": 0.078, "step": 9260 }, { "grad_norm": 0.6530522108078003, "learning_rate": 9.950257976701692e-05, "loss": 0.0785, "step": 9270 }, { "grad_norm": 0.5687869191169739, "learning_rate": 9.950025054902348e-05, "loss": 0.0975, "step": 9280 }, { "grad_norm": 0.7237759828567505, "learning_rate": 9.949791591775706e-05, "loss": 0.0902, "step": 9290 }, { "grad_norm": 0.5582401156425476, "learning_rate": 9.949557587347298e-05, "loss": 0.096, "step": 9300 }, { "grad_norm": 0.4719853103160858, "learning_rate": 9.949323041642713e-05, "loss": 0.0618, "step": 9310 }, { "grad_norm": 0.4870290756225586, "learning_rate": 9.949087954687602e-05, "loss": 0.1017, "step": 9320 }, { "grad_norm": 0.6176109313964844, "learning_rate": 9.948852326507672e-05, "loss": 0.0829, "step": 9330 }, { "grad_norm": 0.46457308530807495, "learning_rate": 9.948616157128694e-05, "loss": 0.0771, "step": 9340 }, { "grad_norm": 0.5580139756202698, "learning_rate": 9.948379446576493e-05, "loss": 0.0734, "step": 9350 }, { "grad_norm": 0.49550214409828186, "learning_rate": 9.948142194876952e-05, "loss": 0.0747, "step": 9360 }, { "grad_norm": 0.6195956468582153, "learning_rate": 9.947904402056024e-05, "loss": 0.0942, "step": 9370 }, { "grad_norm": 0.48746567964553833, "learning_rate": 9.947666068139708e-05, "loss": 0.0744, "step": 9380 }, { "grad_norm": 0.5440317988395691, "learning_rate": 9.947427193154071e-05, "loss": 0.0784, "step": 9390 }, { "grad_norm": 0.4671891927719116, "learning_rate": 9.947187777125233e-05, "loss": 0.0937, "step": 9400 }, { "grad_norm": 0.7194022536277771, "learning_rate": 9.946947820079377e-05, "loss": 0.0935, "step": 9410 }, { "grad_norm": 0.6202794909477234, "learning_rate": 9.946707322042747e-05, "loss": 0.084, "step": 9420 }, { "grad_norm": 0.4537772238254547, "learning_rate": 9.94646628304164e-05, "loss": 0.072, "step": 9430 }, { "grad_norm": 0.5070168375968933, "learning_rate": 9.946224703102418e-05, "loss": 0.0701, "step": 9440 }, { "grad_norm": 0.6199833154678345, "learning_rate": 9.945982582251498e-05, "loss": 0.0803, "step": 9450 }, { "grad_norm": 0.584141194820404, "learning_rate": 9.94573992051536e-05, "loss": 0.0763, "step": 9460 }, { "grad_norm": 0.5403038263320923, "learning_rate": 9.94549671792054e-05, "loss": 0.1139, "step": 9470 }, { "grad_norm": 0.5483202934265137, "learning_rate": 9.945252974493635e-05, "loss": 0.0783, "step": 9480 }, { "grad_norm": 0.48296475410461426, "learning_rate": 9.9450086902613e-05, "loss": 0.0761, "step": 9490 }, { "grad_norm": 0.4274666905403137, "learning_rate": 9.944763865250248e-05, "loss": 0.0701, "step": 9500 }, { "grad_norm": 0.5141958594322205, "learning_rate": 9.944518499487254e-05, "loss": 0.078, "step": 9510 }, { "grad_norm": 0.5613479614257812, "learning_rate": 9.944272592999151e-05, "loss": 0.0828, "step": 9520 }, { "grad_norm": 0.40101635456085205, "learning_rate": 9.94402614581283e-05, "loss": 0.0815, "step": 9530 }, { "grad_norm": 0.7508925795555115, "learning_rate": 9.943779157955244e-05, "loss": 0.084, "step": 9540 }, { "grad_norm": 0.5659773349761963, "learning_rate": 9.943531629453403e-05, "loss": 0.0826, "step": 9550 }, { "grad_norm": 0.5688828825950623, "learning_rate": 9.943283560334375e-05, "loss": 0.0775, "step": 9560 }, { "grad_norm": 0.5252583622932434, "learning_rate": 9.943034950625288e-05, "loss": 0.0944, "step": 9570 }, { "grad_norm": 0.5408691763877869, "learning_rate": 9.942785800353332e-05, "loss": 0.0909, "step": 9580 }, { "grad_norm": 0.7914308309555054, "learning_rate": 9.942536109545751e-05, "loss": 0.0973, "step": 9590 }, { "grad_norm": 0.6199980974197388, "learning_rate": 9.942285878229853e-05, "loss": 0.0867, "step": 9600 }, { "grad_norm": 0.5067251324653625, "learning_rate": 9.942035106433001e-05, "loss": 0.0849, "step": 9610 }, { "grad_norm": 0.46741899847984314, "learning_rate": 9.94178379418262e-05, "loss": 0.0948, "step": 9620 }, { "grad_norm": 0.5513319373130798, "learning_rate": 9.941531941506194e-05, "loss": 0.0856, "step": 9630 }, { "grad_norm": 0.7660220861434937, "learning_rate": 9.941279548431263e-05, "loss": 0.0732, "step": 9640 }, { "grad_norm": 0.5228189826011658, "learning_rate": 9.941026614985431e-05, "loss": 0.0811, "step": 9650 }, { "grad_norm": 0.5666506886482239, "learning_rate": 9.940773141196357e-05, "loss": 0.0751, "step": 9660 }, { "grad_norm": 0.572827935218811, "learning_rate": 9.94051912709176e-05, "loss": 0.0762, "step": 9670 }, { "grad_norm": 0.4288659691810608, "learning_rate": 9.940264572699421e-05, "loss": 0.0775, "step": 9680 }, { "grad_norm": 0.5054352283477783, "learning_rate": 9.940009478047174e-05, "loss": 0.0906, "step": 9690 }, { "grad_norm": 0.8149074912071228, "learning_rate": 9.939753843162918e-05, "loss": 0.0792, "step": 9700 }, { "grad_norm": 0.47952762246131897, "learning_rate": 9.939497668074609e-05, "loss": 0.0795, "step": 9710 }, { "grad_norm": 0.5648713707923889, "learning_rate": 9.93924095281026e-05, "loss": 0.0775, "step": 9720 }, { "grad_norm": 0.6069598197937012, "learning_rate": 9.938983697397948e-05, "loss": 0.0906, "step": 9730 }, { "grad_norm": 0.6495863795280457, "learning_rate": 9.938725901865805e-05, "loss": 0.0812, "step": 9740 }, { "grad_norm": 0.859237790107727, "learning_rate": 9.93846756624202e-05, "loss": 0.0932, "step": 9750 }, { "grad_norm": 0.9048582315444946, "learning_rate": 9.938208690554849e-05, "loss": 0.1142, "step": 9760 }, { "grad_norm": 0.6140553951263428, "learning_rate": 9.9379492748326e-05, "loss": 0.086, "step": 9770 }, { "grad_norm": 0.6148127317428589, "learning_rate": 9.937689319103641e-05, "loss": 0.0857, "step": 9780 }, { "grad_norm": 0.6768519282341003, "learning_rate": 9.937428823396404e-05, "loss": 0.0757, "step": 9790 }, { "grad_norm": 0.5174987316131592, "learning_rate": 9.937167787739372e-05, "loss": 0.087, "step": 9800 }, { "grad_norm": 0.590381383895874, "learning_rate": 9.936906212161095e-05, "loss": 0.0892, "step": 9810 }, { "grad_norm": 0.479815274477005, "learning_rate": 9.936644096690176e-05, "loss": 0.0825, "step": 9820 }, { "grad_norm": 0.5346323847770691, "learning_rate": 9.936381441355282e-05, "loss": 0.0818, "step": 9830 }, { "grad_norm": 0.46879714727401733, "learning_rate": 9.936118246185136e-05, "loss": 0.0895, "step": 9840 }, { "grad_norm": 0.6286042332649231, "learning_rate": 9.935854511208518e-05, "loss": 0.088, "step": 9850 }, { "grad_norm": 0.6407456994056702, "learning_rate": 9.935590236454272e-05, "loss": 0.0787, "step": 9860 }, { "grad_norm": 0.42070508003234863, "learning_rate": 9.935325421951298e-05, "loss": 0.0774, "step": 9870 }, { "grad_norm": 0.5569301843643188, "learning_rate": 9.935060067728557e-05, "loss": 0.0844, "step": 9880 }, { "grad_norm": 0.48179855942726135, "learning_rate": 9.934794173815067e-05, "loss": 0.0655, "step": 9890 }, { "grad_norm": 0.5715886950492859, "learning_rate": 9.934527740239906e-05, "loss": 0.0851, "step": 9900 }, { "grad_norm": 0.7177139520645142, "learning_rate": 9.934260767032209e-05, "loss": 0.0776, "step": 9910 }, { "grad_norm": 0.6537452936172485, "learning_rate": 9.933993254221172e-05, "loss": 0.0942, "step": 9920 }, { "grad_norm": 0.6452089548110962, "learning_rate": 9.933725201836053e-05, "loss": 0.094, "step": 9930 }, { "grad_norm": 0.6496132016181946, "learning_rate": 9.933456609906162e-05, "loss": 0.0927, "step": 9940 }, { "grad_norm": 0.6298534870147705, "learning_rate": 9.933187478460875e-05, "loss": 0.0827, "step": 9950 }, { "grad_norm": 0.5706610679626465, "learning_rate": 9.93291780752962e-05, "loss": 0.0762, "step": 9960 }, { "grad_norm": 0.4550118148326874, "learning_rate": 9.932647597141893e-05, "loss": 0.0911, "step": 9970 }, { "grad_norm": 0.4146193563938141, "learning_rate": 9.932376847327239e-05, "loss": 0.0784, "step": 9980 }, { "grad_norm": 0.6639746427536011, "learning_rate": 9.932105558115268e-05, "loss": 0.0839, "step": 9990 }, { "grad_norm": 0.7016950845718384, "learning_rate": 9.931833729535651e-05, "loss": 0.089, "step": 10000 }, { "grad_norm": 0.5899638533592224, "learning_rate": 9.931561361618111e-05, "loss": 0.0901, "step": 10010 }, { "grad_norm": 0.5620181560516357, "learning_rate": 9.931288454392435e-05, "loss": 0.0899, "step": 10020 }, { "grad_norm": 0.5864284038543701, "learning_rate": 9.931015007888467e-05, "loss": 0.0842, "step": 10030 }, { "grad_norm": 0.6108407378196716, "learning_rate": 9.930741022136112e-05, "loss": 0.0983, "step": 10040 }, { "grad_norm": 0.5652732849121094, "learning_rate": 9.930466497165333e-05, "loss": 0.0886, "step": 10050 }, { "grad_norm": 0.56644606590271, "learning_rate": 9.93019143300615e-05, "loss": 0.0886, "step": 10060 }, { "grad_norm": 0.7374792695045471, "learning_rate": 9.929915829688644e-05, "loss": 0.0683, "step": 10070 }, { "grad_norm": 0.3845058083534241, "learning_rate": 9.929639687242955e-05, "loss": 0.0792, "step": 10080 }, { "grad_norm": 0.5078858733177185, "learning_rate": 9.929363005699281e-05, "loss": 0.074, "step": 10090 }, { "grad_norm": 0.5324487090110779, "learning_rate": 9.92908578508788e-05, "loss": 0.0786, "step": 10100 }, { "grad_norm": 0.8212671279907227, "learning_rate": 9.928808025439069e-05, "loss": 0.0808, "step": 10110 }, { "grad_norm": 1.0980452299118042, "learning_rate": 9.928529726783223e-05, "loss": 0.0768, "step": 10120 }, { "grad_norm": 0.607117772102356, "learning_rate": 9.928250889150774e-05, "loss": 0.0895, "step": 10130 }, { "grad_norm": 0.5918005108833313, "learning_rate": 9.92797151257222e-05, "loss": 0.0741, "step": 10140 }, { "grad_norm": 0.5398584604263306, "learning_rate": 9.927691597078108e-05, "loss": 0.092, "step": 10150 }, { "grad_norm": 0.5423286557197571, "learning_rate": 9.927411142699053e-05, "loss": 0.084, "step": 10160 }, { "grad_norm": 0.6650620102882385, "learning_rate": 9.927130149465725e-05, "loss": 0.0797, "step": 10170 }, { "grad_norm": 0.5455688834190369, "learning_rate": 9.92684861740885e-05, "loss": 0.101, "step": 10180 }, { "grad_norm": 0.595747709274292, "learning_rate": 9.926566546559217e-05, "loss": 0.1011, "step": 10190 }, { "grad_norm": 0.5477171540260315, "learning_rate": 9.926283936947673e-05, "loss": 0.0845, "step": 10200 }, { "grad_norm": 0.53519606590271, "learning_rate": 9.926000788605126e-05, "loss": 0.0825, "step": 10210 }, { "grad_norm": 0.3989523649215698, "learning_rate": 9.92571710156254e-05, "loss": 0.0778, "step": 10220 }, { "grad_norm": 0.9273155927658081, "learning_rate": 9.925432875850936e-05, "loss": 0.0792, "step": 10230 }, { "grad_norm": 0.6357234716415405, "learning_rate": 9.925148111501396e-05, "loss": 0.083, "step": 10240 }, { "grad_norm": 1.9356865882873535, "learning_rate": 9.924862808545066e-05, "loss": 0.0749, "step": 10250 }, { "grad_norm": 0.4829665720462799, "learning_rate": 9.924576967013141e-05, "loss": 0.0787, "step": 10260 }, { "grad_norm": 0.4919174313545227, "learning_rate": 9.924290586936887e-05, "loss": 0.0773, "step": 10270 }, { "grad_norm": 0.6054056882858276, "learning_rate": 9.924003668347614e-05, "loss": 0.0984, "step": 10280 }, { "grad_norm": 0.5025943517684937, "learning_rate": 9.923716211276704e-05, "loss": 0.1004, "step": 10290 }, { "grad_norm": 0.7721869945526123, "learning_rate": 9.923428215755594e-05, "loss": 0.0829, "step": 10300 }, { "grad_norm": 1.3913660049438477, "learning_rate": 9.923139681815775e-05, "loss": 0.0774, "step": 10310 }, { "grad_norm": 0.4788392186164856, "learning_rate": 9.922850609488801e-05, "loss": 0.0817, "step": 10320 }, { "grad_norm": 0.9696179628372192, "learning_rate": 9.922560998806287e-05, "loss": 0.088, "step": 10330 }, { "grad_norm": 0.6690558791160583, "learning_rate": 9.922270849799905e-05, "loss": 0.0882, "step": 10340 }, { "grad_norm": 0.57146817445755, "learning_rate": 9.92198016250138e-05, "loss": 0.0843, "step": 10350 }, { "grad_norm": 0.47604694962501526, "learning_rate": 9.921688936942506e-05, "loss": 0.0795, "step": 10360 }, { "grad_norm": 0.5503217577934265, "learning_rate": 9.921397173155129e-05, "loss": 0.0675, "step": 10370 }, { "grad_norm": 0.7751238942146301, "learning_rate": 9.921104871171157e-05, "loss": 0.0766, "step": 10380 }, { "grad_norm": 0.6876411437988281, "learning_rate": 9.920812031022554e-05, "loss": 0.0811, "step": 10390 }, { "grad_norm": 0.8728201389312744, "learning_rate": 9.920518652741348e-05, "loss": 0.0833, "step": 10400 }, { "grad_norm": 0.48823490738868713, "learning_rate": 9.920224736359618e-05, "loss": 0.0964, "step": 10410 }, { "grad_norm": 0.7297697067260742, "learning_rate": 9.91993028190951e-05, "loss": 0.0806, "step": 10420 }, { "grad_norm": 0.7151899933815002, "learning_rate": 9.919635289423222e-05, "loss": 0.0789, "step": 10430 }, { "grad_norm": 0.5224092602729797, "learning_rate": 9.919339758933015e-05, "loss": 0.0707, "step": 10440 }, { "grad_norm": 1.0945719480514526, "learning_rate": 9.919043690471209e-05, "loss": 0.0915, "step": 10450 }, { "grad_norm": 0.6053642630577087, "learning_rate": 9.91874708407018e-05, "loss": 0.0909, "step": 10460 }, { "grad_norm": 0.446267306804657, "learning_rate": 9.918449939762367e-05, "loss": 0.0787, "step": 10470 }, { "grad_norm": 0.8508643507957458, "learning_rate": 9.91815225758026e-05, "loss": 0.081, "step": 10480 }, { "grad_norm": 0.5563758015632629, "learning_rate": 9.917854037556419e-05, "loss": 0.0789, "step": 10490 }, { "grad_norm": 0.518271267414093, "learning_rate": 9.917555279723454e-05, "loss": 0.0783, "step": 10500 }, { "grad_norm": 0.5000117421150208, "learning_rate": 9.917255984114036e-05, "loss": 0.0892, "step": 10510 }, { "grad_norm": 0.5143613219261169, "learning_rate": 9.916956150760896e-05, "loss": 0.0687, "step": 10520 }, { "grad_norm": 0.38491174578666687, "learning_rate": 9.916655779696826e-05, "loss": 0.0788, "step": 10530 }, { "grad_norm": 0.5304978489875793, "learning_rate": 9.916354870954671e-05, "loss": 0.0743, "step": 10540 }, { "grad_norm": 0.4240629971027374, "learning_rate": 9.91605342456734e-05, "loss": 0.0774, "step": 10550 }, { "grad_norm": 0.6660621762275696, "learning_rate": 9.915751440567795e-05, "loss": 0.0796, "step": 10560 }, { "grad_norm": 0.6962225437164307, "learning_rate": 9.915448918989066e-05, "loss": 0.0824, "step": 10570 }, { "grad_norm": 0.43996578454971313, "learning_rate": 9.915145859864232e-05, "loss": 0.1005, "step": 10580 }, { "grad_norm": 0.5916410088539124, "learning_rate": 9.914842263226437e-05, "loss": 0.0943, "step": 10590 }, { "grad_norm": 0.36248913407325745, "learning_rate": 9.914538129108882e-05, "loss": 0.0922, "step": 10600 }, { "grad_norm": 0.4734724164009094, "learning_rate": 9.914233457544825e-05, "loss": 0.1064, "step": 10610 }, { "grad_norm": 0.6500482559204102, "learning_rate": 9.913928248567586e-05, "loss": 0.1002, "step": 10620 }, { "grad_norm": 0.6526547074317932, "learning_rate": 9.913622502210542e-05, "loss": 0.0716, "step": 10630 }, { "grad_norm": 0.8153764009475708, "learning_rate": 9.913316218507128e-05, "loss": 0.0845, "step": 10640 }, { "grad_norm": 0.5265912413597107, "learning_rate": 9.91300939749084e-05, "loss": 0.0788, "step": 10650 }, { "grad_norm": 0.49132129549980164, "learning_rate": 9.91270203919523e-05, "loss": 0.0835, "step": 10660 }, { "grad_norm": 0.6046527624130249, "learning_rate": 9.912394143653912e-05, "loss": 0.0687, "step": 10670 }, { "grad_norm": 0.5204525589942932, "learning_rate": 9.912085710900555e-05, "loss": 0.0611, "step": 10680 }, { "grad_norm": 0.6887679696083069, "learning_rate": 9.911776740968892e-05, "loss": 0.0709, "step": 10690 }, { "grad_norm": 0.51556396484375, "learning_rate": 9.911467233892709e-05, "loss": 0.1022, "step": 10700 }, { "grad_norm": 0.39621880650520325, "learning_rate": 9.911157189705853e-05, "loss": 0.0904, "step": 10710 }, { "grad_norm": 0.6908466219902039, "learning_rate": 9.910846608442229e-05, "loss": 0.0779, "step": 10720 }, { "grad_norm": 0.5226860642433167, "learning_rate": 9.910535490135805e-05, "loss": 0.0681, "step": 10730 }, { "grad_norm": 0.938178539276123, "learning_rate": 9.910223834820603e-05, "loss": 0.0847, "step": 10740 }, { "grad_norm": 0.6361491680145264, "learning_rate": 9.909911642530703e-05, "loss": 0.0826, "step": 10750 }, { "grad_norm": 0.5330131649971008, "learning_rate": 9.909598913300249e-05, "loss": 0.0974, "step": 10760 }, { "grad_norm": 0.6860202550888062, "learning_rate": 9.909285647163438e-05, "loss": 0.0913, "step": 10770 }, { "grad_norm": 0.5809422731399536, "learning_rate": 9.908971844154531e-05, "loss": 0.0907, "step": 10780 }, { "grad_norm": 0.882177472114563, "learning_rate": 9.908657504307843e-05, "loss": 0.0888, "step": 10790 }, { "grad_norm": 0.4813498258590698, "learning_rate": 9.908342627657751e-05, "loss": 0.0893, "step": 10800 }, { "grad_norm": 0.5748324990272522, "learning_rate": 9.908027214238689e-05, "loss": 0.0789, "step": 10810 }, { "grad_norm": 0.5980732440948486, "learning_rate": 9.90771126408515e-05, "loss": 0.0873, "step": 10820 }, { "grad_norm": 0.789328932762146, "learning_rate": 9.907394777231685e-05, "loss": 0.085, "step": 10830 }, { "grad_norm": 0.8565263748168945, "learning_rate": 9.907077753712905e-05, "loss": 0.0886, "step": 10840 }, { "grad_norm": 0.7767583727836609, "learning_rate": 9.906760193563482e-05, "loss": 0.0967, "step": 10850 }, { "grad_norm": 0.5076245665550232, "learning_rate": 9.906442096818139e-05, "loss": 0.0718, "step": 10860 }, { "grad_norm": 0.906876802444458, "learning_rate": 9.906123463511665e-05, "loss": 0.0855, "step": 10870 }, { "grad_norm": 0.5745806694030762, "learning_rate": 9.905804293678907e-05, "loss": 0.0955, "step": 10880 }, { "grad_norm": 0.5841134190559387, "learning_rate": 9.905484587354766e-05, "loss": 0.0898, "step": 10890 }, { "grad_norm": 0.5298270583152771, "learning_rate": 9.905164344574205e-05, "loss": 0.083, "step": 10900 }, { "grad_norm": 0.5234941840171814, "learning_rate": 9.904843565372248e-05, "loss": 0.0839, "step": 10910 }, { "grad_norm": 0.630058765411377, "learning_rate": 9.904522249783972e-05, "loss": 0.0825, "step": 10920 }, { "grad_norm": 0.5797252655029297, "learning_rate": 9.904200397844517e-05, "loss": 0.0982, "step": 10930 }, { "grad_norm": 0.6448240876197815, "learning_rate": 9.903878009589078e-05, "loss": 0.0919, "step": 10940 }, { "grad_norm": 0.5455141067504883, "learning_rate": 9.903555085052915e-05, "loss": 0.078, "step": 10950 }, { "grad_norm": 0.4630672335624695, "learning_rate": 9.903231624271338e-05, "loss": 0.0703, "step": 10960 }, { "grad_norm": 0.5502226948738098, "learning_rate": 9.902907627279724e-05, "loss": 0.0842, "step": 10970 }, { "grad_norm": 0.5343467593193054, "learning_rate": 9.902583094113504e-05, "loss": 0.0752, "step": 10980 }, { "grad_norm": 0.49039292335510254, "learning_rate": 9.902258024808168e-05, "loss": 0.0805, "step": 10990 }, { "grad_norm": 1.1393020153045654, "learning_rate": 9.901932419399264e-05, "loss": 0.0819, "step": 11000 }, { "grad_norm": 0.5140417814254761, "learning_rate": 9.9016062779224e-05, "loss": 0.0768, "step": 11010 }, { "grad_norm": 0.5552489161491394, "learning_rate": 9.901279600413242e-05, "loss": 0.07, "step": 11020 }, { "grad_norm": 0.6391453146934509, "learning_rate": 9.900952386907518e-05, "loss": 0.0808, "step": 11030 }, { "grad_norm": 0.7574442028999329, "learning_rate": 9.90062463744101e-05, "loss": 0.0774, "step": 11040 }, { "grad_norm": 0.6024361252784729, "learning_rate": 9.900296352049558e-05, "loss": 0.0687, "step": 11050 }, { "grad_norm": 0.5179422497749329, "learning_rate": 9.899967530769065e-05, "loss": 0.0826, "step": 11060 }, { "grad_norm": 0.5430834889411926, "learning_rate": 9.899638173635489e-05, "loss": 0.0664, "step": 11070 }, { "grad_norm": 0.4740569293498993, "learning_rate": 9.899308280684849e-05, "loss": 0.0722, "step": 11080 }, { "grad_norm": 0.7829235196113586, "learning_rate": 9.898977851953222e-05, "loss": 0.0731, "step": 11090 }, { "grad_norm": 1.336442232131958, "learning_rate": 9.898646887476741e-05, "loss": 0.1026, "step": 11100 }, { "grad_norm": 0.7024753093719482, "learning_rate": 9.898315387291603e-05, "loss": 0.0889, "step": 11110 }, { "grad_norm": 0.5648981332778931, "learning_rate": 9.89798335143406e-05, "loss": 0.0735, "step": 11120 }, { "grad_norm": 1.0287418365478516, "learning_rate": 9.897650779940419e-05, "loss": 0.0904, "step": 11130 }, { "grad_norm": 0.679236114025116, "learning_rate": 9.897317672847054e-05, "loss": 0.0799, "step": 11140 }, { "grad_norm": 0.4004500210285187, "learning_rate": 9.89698403019039e-05, "loss": 0.0726, "step": 11150 }, { "grad_norm": 0.6263647079467773, "learning_rate": 9.896649852006917e-05, "loss": 0.0804, "step": 11160 }, { "grad_norm": 0.5893408060073853, "learning_rate": 9.896315138333177e-05, "loss": 0.0794, "step": 11170 }, { "grad_norm": 0.5356624126434326, "learning_rate": 9.895979889205774e-05, "loss": 0.0743, "step": 11180 }, { "grad_norm": 0.5748914480209351, "learning_rate": 9.895644104661372e-05, "loss": 0.0804, "step": 11190 }, { "grad_norm": 0.5571421980857849, "learning_rate": 9.895307784736691e-05, "loss": 0.0772, "step": 11200 }, { "grad_norm": 0.6274701356887817, "learning_rate": 9.894970929468512e-05, "loss": 0.0797, "step": 11210 }, { "grad_norm": 0.5064911246299744, "learning_rate": 9.89463353889367e-05, "loss": 0.0788, "step": 11220 }, { "grad_norm": 0.6358806490898132, "learning_rate": 9.894295613049065e-05, "loss": 0.1034, "step": 11230 }, { "grad_norm": 0.5119845271110535, "learning_rate": 9.893957151971649e-05, "loss": 0.0797, "step": 11240 }, { "grad_norm": 0.4959370195865631, "learning_rate": 9.893618155698436e-05, "loss": 0.0698, "step": 11250 }, { "grad_norm": 0.5536757111549377, "learning_rate": 9.8932786242665e-05, "loss": 0.066, "step": 11260 }, { "grad_norm": 0.41316667199134827, "learning_rate": 9.89293855771297e-05, "loss": 0.0625, "step": 11270 }, { "grad_norm": 0.6403241157531738, "learning_rate": 9.892597956075036e-05, "loss": 0.0873, "step": 11280 }, { "grad_norm": 0.4412843883037567, "learning_rate": 9.892256819389947e-05, "loss": 0.0672, "step": 11290 }, { "grad_norm": 0.5781263113021851, "learning_rate": 9.891915147695006e-05, "loss": 0.0649, "step": 11300 }, { "grad_norm": 0.5156962871551514, "learning_rate": 9.891572941027577e-05, "loss": 0.0819, "step": 11310 }, { "grad_norm": 0.5169123411178589, "learning_rate": 9.89123019942509e-05, "loss": 0.0727, "step": 11320 }, { "grad_norm": 0.4578074812889099, "learning_rate": 9.89088692292502e-05, "loss": 0.0721, "step": 11330 }, { "grad_norm": 0.5630961060523987, "learning_rate": 9.89054311156491e-05, "loss": 0.0731, "step": 11340 }, { "grad_norm": 0.45422568917274475, "learning_rate": 9.890198765382357e-05, "loss": 0.0974, "step": 11350 }, { "grad_norm": 0.6459624171257019, "learning_rate": 9.889853884415021e-05, "loss": 0.0748, "step": 11360 }, { "grad_norm": 0.5301288962364197, "learning_rate": 9.889508468700614e-05, "loss": 0.0698, "step": 11370 }, { "grad_norm": 0.4160058796405792, "learning_rate": 9.889162518276915e-05, "loss": 0.0763, "step": 11380 }, { "grad_norm": 0.5253047347068787, "learning_rate": 9.888816033181752e-05, "loss": 0.0717, "step": 11390 }, { "grad_norm": 0.5715725421905518, "learning_rate": 9.888469013453018e-05, "loss": 0.0853, "step": 11400 }, { "grad_norm": 0.5873308181762695, "learning_rate": 9.888121459128663e-05, "loss": 0.0751, "step": 11410 }, { "grad_norm": 0.5795129537582397, "learning_rate": 9.887773370246693e-05, "loss": 0.0913, "step": 11420 }, { "grad_norm": 0.7295461297035217, "learning_rate": 9.887424746845177e-05, "loss": 0.093, "step": 11430 }, { "grad_norm": 0.7256125807762146, "learning_rate": 9.887075588962239e-05, "loss": 0.0781, "step": 11440 }, { "grad_norm": 0.6100373268127441, "learning_rate": 9.88672589663606e-05, "loss": 0.0815, "step": 11450 }, { "grad_norm": 0.5657457709312439, "learning_rate": 9.886375669904886e-05, "loss": 0.0895, "step": 11460 }, { "grad_norm": 0.7509289979934692, "learning_rate": 9.886024908807014e-05, "loss": 0.0749, "step": 11470 }, { "grad_norm": 0.630613386631012, "learning_rate": 9.885673613380806e-05, "loss": 0.0786, "step": 11480 }, { "grad_norm": 0.5222012400627136, "learning_rate": 9.885321783664676e-05, "loss": 0.0878, "step": 11490 }, { "grad_norm": 0.475776344537735, "learning_rate": 9.884969419697101e-05, "loss": 0.072, "step": 11500 }, { "grad_norm": 0.45581111311912537, "learning_rate": 9.884616521516614e-05, "loss": 0.0779, "step": 11510 }, { "grad_norm": 0.3896608352661133, "learning_rate": 9.88426308916181e-05, "loss": 0.0798, "step": 11520 }, { "grad_norm": 0.5628349781036377, "learning_rate": 9.883909122671335e-05, "loss": 0.0683, "step": 11530 }, { "grad_norm": 0.4942054748535156, "learning_rate": 9.883554622083904e-05, "loss": 0.0702, "step": 11540 }, { "grad_norm": 0.5827838778495789, "learning_rate": 9.88319958743828e-05, "loss": 0.0745, "step": 11550 }, { "grad_norm": 0.6358640789985657, "learning_rate": 9.882844018773291e-05, "loss": 0.072, "step": 11560 }, { "grad_norm": 0.5835574269294739, "learning_rate": 9.882487916127823e-05, "loss": 0.0716, "step": 11570 }, { "grad_norm": 0.5689127445220947, "learning_rate": 9.882131279540815e-05, "loss": 0.0775, "step": 11580 }, { "grad_norm": 0.5413917303085327, "learning_rate": 9.881774109051271e-05, "loss": 0.0717, "step": 11590 }, { "grad_norm": 0.4506465792655945, "learning_rate": 9.881416404698252e-05, "loss": 0.0687, "step": 11600 }, { "grad_norm": 0.39434629678726196, "learning_rate": 9.881058166520873e-05, "loss": 0.0844, "step": 11610 }, { "grad_norm": 0.6298508048057556, "learning_rate": 9.880699394558311e-05, "loss": 0.0913, "step": 11620 }, { "grad_norm": 0.6297813057899475, "learning_rate": 9.880340088849801e-05, "loss": 0.1055, "step": 11630 }, { "grad_norm": 0.5221153497695923, "learning_rate": 9.879980249434637e-05, "loss": 0.0678, "step": 11640 }, { "grad_norm": 0.644862949848175, "learning_rate": 9.879619876352168e-05, "loss": 0.0754, "step": 11650 }, { "grad_norm": 0.498611718416214, "learning_rate": 9.879258969641809e-05, "loss": 0.0721, "step": 11660 }, { "grad_norm": 0.4187641739845276, "learning_rate": 9.878897529343023e-05, "loss": 0.0713, "step": 11670 }, { "grad_norm": 0.459998220205307, "learning_rate": 9.878535555495338e-05, "loss": 0.0775, "step": 11680 }, { "grad_norm": 0.6437307000160217, "learning_rate": 9.87817304813834e-05, "loss": 0.0909, "step": 11690 }, { "grad_norm": 0.5704212784767151, "learning_rate": 9.877810007311671e-05, "loss": 0.0782, "step": 11700 }, { "grad_norm": 0.5618981122970581, "learning_rate": 9.877446433055035e-05, "loss": 0.0723, "step": 11710 }, { "grad_norm": 0.67950838804245, "learning_rate": 9.877082325408191e-05, "loss": 0.0959, "step": 11720 }, { "grad_norm": 1.217835545539856, "learning_rate": 9.876717684410954e-05, "loss": 0.0722, "step": 11730 }, { "grad_norm": 1.708711862564087, "learning_rate": 9.876352510103204e-05, "loss": 0.0918, "step": 11740 }, { "grad_norm": 0.6173080205917358, "learning_rate": 9.875986802524875e-05, "loss": 0.0835, "step": 11750 }, { "grad_norm": 0.4977591633796692, "learning_rate": 9.87562056171596e-05, "loss": 0.0829, "step": 11760 }, { "grad_norm": 0.8709372282028198, "learning_rate": 9.875253787716511e-05, "loss": 0.0773, "step": 11770 }, { "grad_norm": 0.6895302534103394, "learning_rate": 9.874886480566637e-05, "loss": 0.0762, "step": 11780 }, { "grad_norm": 0.47415292263031006, "learning_rate": 9.874518640306507e-05, "loss": 0.0723, "step": 11790 }, { "grad_norm": 0.47539281845092773, "learning_rate": 9.874150266976347e-05, "loss": 0.0717, "step": 11800 }, { "grad_norm": 0.3950406312942505, "learning_rate": 9.873781360616443e-05, "loss": 0.082, "step": 11810 }, { "grad_norm": 0.7181929349899292, "learning_rate": 9.873411921267137e-05, "loss": 0.0757, "step": 11820 }, { "grad_norm": 0.657210111618042, "learning_rate": 9.873041948968829e-05, "loss": 0.0896, "step": 11830 }, { "grad_norm": 0.5692567825317383, "learning_rate": 9.872671443761981e-05, "loss": 0.0676, "step": 11840 }, { "grad_norm": 0.5480948686599731, "learning_rate": 9.872300405687109e-05, "loss": 0.1015, "step": 11850 }, { "grad_norm": 0.3888741731643677, "learning_rate": 9.871928834784792e-05, "loss": 0.0809, "step": 11860 }, { "grad_norm": 0.5422941446304321, "learning_rate": 9.871556731095661e-05, "loss": 0.07, "step": 11870 }, { "grad_norm": 0.6868987679481506, "learning_rate": 9.871184094660411e-05, "loss": 0.1089, "step": 11880 }, { "grad_norm": 0.5424380302429199, "learning_rate": 9.870810925519791e-05, "loss": 0.0824, "step": 11890 }, { "grad_norm": 0.6060078740119934, "learning_rate": 9.870437223714612e-05, "loss": 0.0876, "step": 11900 }, { "grad_norm": 0.49363213777542114, "learning_rate": 9.87006298928574e-05, "loss": 0.0876, "step": 11910 }, { "grad_norm": 0.7147068381309509, "learning_rate": 9.869688222274103e-05, "loss": 0.0777, "step": 11920 }, { "grad_norm": 0.6184251308441162, "learning_rate": 9.869312922720681e-05, "loss": 0.0803, "step": 11930 }, { "grad_norm": 0.7328113913536072, "learning_rate": 9.868937090666521e-05, "loss": 0.0724, "step": 11940 }, { "grad_norm": 0.8590551018714905, "learning_rate": 9.86856072615272e-05, "loss": 0.0599, "step": 11950 }, { "grad_norm": 0.5318983197212219, "learning_rate": 9.868183829220438e-05, "loss": 0.0778, "step": 11960 }, { "grad_norm": 1.0610337257385254, "learning_rate": 9.867806399910893e-05, "loss": 0.0902, "step": 11970 }, { "grad_norm": 0.6968745589256287, "learning_rate": 9.867428438265356e-05, "loss": 0.0811, "step": 11980 }, { "grad_norm": 0.35123568773269653, "learning_rate": 9.867049944325165e-05, "loss": 0.0806, "step": 11990 }, { "grad_norm": 0.5755341649055481, "learning_rate": 9.86667091813171e-05, "loss": 0.0978, "step": 12000 }, { "grad_norm": 0.7186183929443359, "learning_rate": 9.866291359726438e-05, "loss": 0.0847, "step": 12010 }, { "grad_norm": 0.423641562461853, "learning_rate": 9.865911269150861e-05, "loss": 0.0955, "step": 12020 }, { "grad_norm": 0.7056972980499268, "learning_rate": 9.865530646446544e-05, "loss": 0.0858, "step": 12030 }, { "grad_norm": 0.7500908374786377, "learning_rate": 9.86514949165511e-05, "loss": 0.0862, "step": 12040 }, { "grad_norm": 0.5049628615379333, "learning_rate": 9.864767804818243e-05, "loss": 0.0824, "step": 12050 }, { "grad_norm": 0.702847957611084, "learning_rate": 9.86438558597768e-05, "loss": 0.0777, "step": 12060 }, { "grad_norm": 0.5156053900718689, "learning_rate": 9.864002835175225e-05, "loss": 0.102, "step": 12070 }, { "grad_norm": 0.7792355418205261, "learning_rate": 9.863619552452734e-05, "loss": 0.0718, "step": 12080 }, { "grad_norm": 0.7836655378341675, "learning_rate": 9.863235737852119e-05, "loss": 0.0838, "step": 12090 }, { "grad_norm": 0.515976071357727, "learning_rate": 9.862851391415356e-05, "loss": 0.0805, "step": 12100 }, { "grad_norm": 0.6690770983695984, "learning_rate": 9.862466513184477e-05, "loss": 0.0808, "step": 12110 }, { "grad_norm": 0.601223349571228, "learning_rate": 9.86208110320157e-05, "loss": 0.0749, "step": 12120 }, { "grad_norm": 0.588204562664032, "learning_rate": 9.861695161508784e-05, "loss": 0.0761, "step": 12130 }, { "grad_norm": 1.609395980834961, "learning_rate": 9.861308688148324e-05, "loss": 0.0703, "step": 12140 }, { "grad_norm": 0.4795415997505188, "learning_rate": 9.860921683162455e-05, "loss": 0.0761, "step": 12150 }, { "grad_norm": 0.47556790709495544, "learning_rate": 9.860534146593499e-05, "loss": 0.0682, "step": 12160 }, { "grad_norm": 0.3935212790966034, "learning_rate": 9.860146078483836e-05, "loss": 0.0694, "step": 12170 }, { "grad_norm": 0.6759193539619446, "learning_rate": 9.859757478875905e-05, "loss": 0.09, "step": 12180 }, { "grad_norm": 0.4969331622123718, "learning_rate": 9.859368347812204e-05, "loss": 0.0653, "step": 12190 }, { "grad_norm": 0.5054757595062256, "learning_rate": 9.858978685335285e-05, "loss": 0.0836, "step": 12200 }, { "grad_norm": 0.5641167759895325, "learning_rate": 9.858588491487763e-05, "loss": 0.0995, "step": 12210 }, { "grad_norm": 0.7139019966125488, "learning_rate": 9.858197766312308e-05, "loss": 0.0656, "step": 12220 }, { "grad_norm": 0.5430561304092407, "learning_rate": 9.857806509851649e-05, "loss": 0.076, "step": 12230 }, { "grad_norm": 0.45142099261283875, "learning_rate": 9.857414722148574e-05, "loss": 0.065, "step": 12240 }, { "grad_norm": 0.4267752170562744, "learning_rate": 9.857022403245928e-05, "loss": 0.0761, "step": 12250 }, { "grad_norm": 0.6345266699790955, "learning_rate": 9.856629553186615e-05, "loss": 0.09, "step": 12260 }, { "grad_norm": 0.6531481146812439, "learning_rate": 9.856236172013595e-05, "loss": 0.0726, "step": 12270 }, { "grad_norm": 0.6847506761550903, "learning_rate": 9.85584225976989e-05, "loss": 0.0608, "step": 12280 }, { "grad_norm": 0.6904215216636658, "learning_rate": 9.855447816498575e-05, "loss": 0.0695, "step": 12290 }, { "grad_norm": 0.5556085109710693, "learning_rate": 9.855052842242787e-05, "loss": 0.0625, "step": 12300 }, { "grad_norm": 0.45534682273864746, "learning_rate": 9.85465733704572e-05, "loss": 0.0599, "step": 12310 }, { "grad_norm": 0.7702237963676453, "learning_rate": 9.854261300950624e-05, "loss": 0.0771, "step": 12320 }, { "grad_norm": 0.5669450759887695, "learning_rate": 9.853864734000813e-05, "loss": 0.075, "step": 12330 }, { "grad_norm": 0.5915327072143555, "learning_rate": 9.85346763623965e-05, "loss": 0.0783, "step": 12340 }, { "grad_norm": 0.6184828281402588, "learning_rate": 9.853070007710564e-05, "loss": 0.083, "step": 12350 }, { "grad_norm": 0.50102698802948, "learning_rate": 9.85267184845704e-05, "loss": 0.0869, "step": 12360 }, { "grad_norm": 0.49679702520370483, "learning_rate": 9.852273158522616e-05, "loss": 0.0611, "step": 12370 }, { "grad_norm": 0.512780487537384, "learning_rate": 9.851873937950896e-05, "loss": 0.0696, "step": 12380 }, { "grad_norm": 0.4891941547393799, "learning_rate": 9.851474186785537e-05, "loss": 0.0744, "step": 12390 }, { "grad_norm": 0.536969780921936, "learning_rate": 9.851073905070254e-05, "loss": 0.0855, "step": 12400 }, { "grad_norm": 0.7352870106697083, "learning_rate": 9.850673092848824e-05, "loss": 0.0776, "step": 12410 }, { "grad_norm": 0.4803907573223114, "learning_rate": 9.850271750165077e-05, "loss": 0.068, "step": 12420 }, { "grad_norm": 0.6254199743270874, "learning_rate": 9.849869877062902e-05, "loss": 0.082, "step": 12430 }, { "grad_norm": 0.4349352717399597, "learning_rate": 9.849467473586252e-05, "loss": 0.0813, "step": 12440 }, { "grad_norm": 0.4448443353176117, "learning_rate": 9.849064539779127e-05, "loss": 0.0653, "step": 12450 }, { "grad_norm": 0.5427790880203247, "learning_rate": 9.848661075685594e-05, "loss": 0.079, "step": 12460 }, { "grad_norm": 0.7491257786750793, "learning_rate": 9.848257081349778e-05, "loss": 0.0745, "step": 12470 }, { "grad_norm": 0.45396533608436584, "learning_rate": 9.847852556815856e-05, "loss": 0.0607, "step": 12480 }, { "grad_norm": 0.594918966293335, "learning_rate": 9.847447502128067e-05, "loss": 0.0728, "step": 12490 }, { "grad_norm": 0.5973802804946899, "learning_rate": 9.847041917330708e-05, "loss": 0.075, "step": 12500 }, { "grad_norm": 0.5872713327407837, "learning_rate": 9.846635802468132e-05, "loss": 0.0837, "step": 12510 }, { "grad_norm": 0.4665496051311493, "learning_rate": 9.84622915758475e-05, "loss": 0.0609, "step": 12520 }, { "grad_norm": 0.3695456385612488, "learning_rate": 9.845821982725034e-05, "loss": 0.0714, "step": 12530 }, { "grad_norm": 0.49616023898124695, "learning_rate": 9.845414277933514e-05, "loss": 0.0902, "step": 12540 }, { "grad_norm": 0.5404226183891296, "learning_rate": 9.845006043254771e-05, "loss": 0.0791, "step": 12550 }, { "grad_norm": 0.6364492774009705, "learning_rate": 9.844597278733451e-05, "loss": 0.0773, "step": 12560 }, { "grad_norm": 0.5531954765319824, "learning_rate": 9.844187984414259e-05, "loss": 0.0752, "step": 12570 }, { "grad_norm": 0.4851355254650116, "learning_rate": 9.84377816034195e-05, "loss": 0.0764, "step": 12580 }, { "grad_norm": 0.41711434721946716, "learning_rate": 9.843367806561345e-05, "loss": 0.0777, "step": 12590 }, { "grad_norm": 0.5073497295379639, "learning_rate": 9.842956923117317e-05, "loss": 0.0695, "step": 12600 }, { "grad_norm": 0.4292295575141907, "learning_rate": 9.842545510054802e-05, "loss": 0.0619, "step": 12610 }, { "grad_norm": 0.5060393810272217, "learning_rate": 9.842133567418792e-05, "loss": 0.0684, "step": 12620 }, { "grad_norm": 0.4899902939796448, "learning_rate": 9.841721095254333e-05, "loss": 0.0651, "step": 12630 }, { "grad_norm": 0.6252016425132751, "learning_rate": 9.841308093606537e-05, "loss": 0.0744, "step": 12640 }, { "grad_norm": 0.406350314617157, "learning_rate": 9.840894562520565e-05, "loss": 0.0697, "step": 12650 }, { "grad_norm": 0.4791690707206726, "learning_rate": 9.840480502041642e-05, "loss": 0.0821, "step": 12660 }, { "grad_norm": 0.4558219611644745, "learning_rate": 9.840065912215049e-05, "loss": 0.0658, "step": 12670 }, { "grad_norm": 0.5390247106552124, "learning_rate": 9.839650793086124e-05, "loss": 0.0843, "step": 12680 }, { "grad_norm": 0.539884626865387, "learning_rate": 9.839235144700265e-05, "loss": 0.0752, "step": 12690 }, { "grad_norm": 0.5432820320129395, "learning_rate": 9.838818967102926e-05, "loss": 0.0736, "step": 12700 }, { "grad_norm": 0.42961567640304565, "learning_rate": 9.83840226033962e-05, "loss": 0.0713, "step": 12710 }, { "grad_norm": 0.5297060608863831, "learning_rate": 9.837985024455918e-05, "loss": 0.0787, "step": 12720 }, { "grad_norm": 0.5784377455711365, "learning_rate": 9.837567259497447e-05, "loss": 0.0627, "step": 12730 }, { "grad_norm": 0.9706358313560486, "learning_rate": 9.837148965509894e-05, "loss": 0.0808, "step": 12740 }, { "grad_norm": 0.6515229940414429, "learning_rate": 9.836730142539001e-05, "loss": 0.0745, "step": 12750 }, { "grad_norm": 0.595328152179718, "learning_rate": 9.836310790630574e-05, "loss": 0.0897, "step": 12760 }, { "grad_norm": 0.9633748531341553, "learning_rate": 9.83589090983047e-05, "loss": 0.0884, "step": 12770 }, { "grad_norm": 0.4823974668979645, "learning_rate": 9.835470500184605e-05, "loss": 0.0794, "step": 12780 }, { "grad_norm": 0.5930135250091553, "learning_rate": 9.835049561738957e-05, "loss": 0.0829, "step": 12790 }, { "grad_norm": 0.6038619875907898, "learning_rate": 9.834628094539558e-05, "loss": 0.0795, "step": 12800 }, { "grad_norm": 0.4351971745491028, "learning_rate": 9.834206098632499e-05, "loss": 0.073, "step": 12810 }, { "grad_norm": 0.6855603456497192, "learning_rate": 9.833783574063931e-05, "loss": 0.0976, "step": 12820 }, { "grad_norm": 0.7239429950714111, "learning_rate": 9.833360520880058e-05, "loss": 0.0867, "step": 12830 }, { "grad_norm": 0.573098361492157, "learning_rate": 9.832936939127144e-05, "loss": 0.0703, "step": 12840 }, { "grad_norm": 0.6491153240203857, "learning_rate": 9.832512828851515e-05, "loss": 0.0726, "step": 12850 }, { "grad_norm": 0.5629779100418091, "learning_rate": 9.832088190099546e-05, "loss": 0.0669, "step": 12860 }, { "grad_norm": 0.7387993335723877, "learning_rate": 9.831663022917679e-05, "loss": 0.0723, "step": 12870 }, { "grad_norm": 0.5757182836532593, "learning_rate": 9.831237327352407e-05, "loss": 0.0736, "step": 12880 }, { "grad_norm": 0.562229573726654, "learning_rate": 9.830811103450286e-05, "loss": 0.0728, "step": 12890 }, { "grad_norm": 0.4365726113319397, "learning_rate": 9.830384351257924e-05, "loss": 0.0879, "step": 12900 }, { "grad_norm": 1.2864347696304321, "learning_rate": 9.829957070821993e-05, "loss": 0.0721, "step": 12910 }, { "grad_norm": 0.6187273859977722, "learning_rate": 9.829529262189218e-05, "loss": 0.0721, "step": 12920 }, { "grad_norm": 0.4502071738243103, "learning_rate": 9.829100925406385e-05, "loss": 0.0903, "step": 12930 }, { "grad_norm": 0.5398077368736267, "learning_rate": 9.828672060520333e-05, "loss": 0.0815, "step": 12940 }, { "grad_norm": 0.631338357925415, "learning_rate": 9.828242667577966e-05, "loss": 0.0697, "step": 12950 }, { "grad_norm": 0.6461072564125061, "learning_rate": 9.82781274662624e-05, "loss": 0.0755, "step": 12960 }, { "grad_norm": 0.542168915271759, "learning_rate": 9.82738229771217e-05, "loss": 0.0842, "step": 12970 }, { "grad_norm": 0.5836749076843262, "learning_rate": 9.826951320882829e-05, "loss": 0.0719, "step": 12980 }, { "grad_norm": 0.7636890411376953, "learning_rate": 9.826519816185351e-05, "loss": 0.0928, "step": 12990 }, { "grad_norm": 0.5290622115135193, "learning_rate": 9.826087783666921e-05, "loss": 0.0764, "step": 13000 }, { "grad_norm": 0.5906341075897217, "learning_rate": 9.825655223374787e-05, "loss": 0.0953, "step": 13010 }, { "grad_norm": 0.5253537893295288, "learning_rate": 9.825222135356253e-05, "loss": 0.0851, "step": 13020 }, { "grad_norm": 0.4353395104408264, "learning_rate": 9.82478851965868e-05, "loss": 0.0797, "step": 13030 }, { "grad_norm": 0.6381043195724487, "learning_rate": 9.82435437632949e-05, "loss": 0.0735, "step": 13040 }, { "grad_norm": 0.6919745206832886, "learning_rate": 9.823919705416158e-05, "loss": 0.0601, "step": 13050 }, { "grad_norm": 0.6777269840240479, "learning_rate": 9.82348450696622e-05, "loss": 0.0762, "step": 13060 }, { "grad_norm": 0.5669152736663818, "learning_rate": 9.823048781027268e-05, "loss": 0.1038, "step": 13070 }, { "grad_norm": 0.5196255445480347, "learning_rate": 9.822612527646953e-05, "loss": 0.0827, "step": 13080 }, { "grad_norm": 0.5433371663093567, "learning_rate": 9.822175746872984e-05, "loss": 0.0924, "step": 13090 }, { "grad_norm": 0.498661607503891, "learning_rate": 9.821738438753123e-05, "loss": 0.0742, "step": 13100 }, { "grad_norm": 0.5330330729484558, "learning_rate": 9.821300603335196e-05, "loss": 0.0715, "step": 13110 }, { "grad_norm": 0.5794325470924377, "learning_rate": 9.820862240667085e-05, "loss": 0.0802, "step": 13120 }, { "grad_norm": 0.4789789021015167, "learning_rate": 9.820423350796726e-05, "loss": 0.0873, "step": 13130 }, { "grad_norm": 0.5500144958496094, "learning_rate": 9.819983933772118e-05, "loss": 0.0928, "step": 13140 }, { "grad_norm": 0.5096775889396667, "learning_rate": 9.819543989641314e-05, "loss": 0.0717, "step": 13150 }, { "grad_norm": 0.7274949550628662, "learning_rate": 9.819103518452423e-05, "loss": 0.0726, "step": 13160 }, { "grad_norm": 0.4821552038192749, "learning_rate": 9.818662520253618e-05, "loss": 0.0747, "step": 13170 }, { "grad_norm": 0.5996457934379578, "learning_rate": 9.818220995093126e-05, "loss": 0.076, "step": 13180 }, { "grad_norm": 0.4428446590900421, "learning_rate": 9.817778943019228e-05, "loss": 0.0642, "step": 13190 }, { "grad_norm": 0.579748809337616, "learning_rate": 9.81733636408027e-05, "loss": 0.0825, "step": 13200 }, { "grad_norm": 0.5319037437438965, "learning_rate": 9.816893258324649e-05, "loss": 0.0735, "step": 13210 }, { "grad_norm": 0.3806709945201874, "learning_rate": 9.816449625800823e-05, "loss": 0.0677, "step": 13220 }, { "grad_norm": 0.5735424160957336, "learning_rate": 9.816005466557308e-05, "loss": 0.0848, "step": 13230 }, { "grad_norm": 0.6226041913032532, "learning_rate": 9.815560780642674e-05, "loss": 0.0787, "step": 13240 }, { "grad_norm": 0.4704526960849762, "learning_rate": 9.815115568105555e-05, "loss": 0.0685, "step": 13250 }, { "grad_norm": 0.6217973232269287, "learning_rate": 9.814669828994638e-05, "loss": 0.0752, "step": 13260 }, { "grad_norm": 0.4788847863674164, "learning_rate": 9.814223563358665e-05, "loss": 0.061, "step": 13270 }, { "grad_norm": 0.6806874871253967, "learning_rate": 9.813776771246443e-05, "loss": 0.0662, "step": 13280 }, { "grad_norm": 0.36248934268951416, "learning_rate": 9.813329452706829e-05, "loss": 0.0648, "step": 13290 }, { "grad_norm": 0.5812167525291443, "learning_rate": 9.812881607788744e-05, "loss": 0.0645, "step": 13300 }, { "grad_norm": 0.5074732899665833, "learning_rate": 9.812433236541163e-05, "loss": 0.0887, "step": 13310 }, { "grad_norm": 0.6852806210517883, "learning_rate": 9.811984339013116e-05, "loss": 0.0904, "step": 13320 }, { "grad_norm": 0.4731719493865967, "learning_rate": 9.811534915253698e-05, "loss": 0.0698, "step": 13330 }, { "grad_norm": 0.49721089005470276, "learning_rate": 9.811084965312056e-05, "loss": 0.072, "step": 13340 }, { "grad_norm": 0.6207499504089355, "learning_rate": 9.810634489237396e-05, "loss": 0.0638, "step": 13350 }, { "grad_norm": 0.7469450235366821, "learning_rate": 9.81018348707898e-05, "loss": 0.0828, "step": 13360 }, { "grad_norm": 0.6989936828613281, "learning_rate": 9.809731958886131e-05, "loss": 0.0804, "step": 13370 }, { "grad_norm": 0.620055615901947, "learning_rate": 9.809279904708224e-05, "loss": 0.081, "step": 13380 }, { "grad_norm": 0.5486478805541992, "learning_rate": 9.808827324594699e-05, "loss": 0.0747, "step": 13390 }, { "grad_norm": 0.3409775495529175, "learning_rate": 9.808374218595046e-05, "loss": 0.0836, "step": 13400 }, { "grad_norm": 0.5290846228599548, "learning_rate": 9.80792058675882e-05, "loss": 0.0682, "step": 13410 }, { "grad_norm": 0.43967360258102417, "learning_rate": 9.807466429135627e-05, "loss": 0.0631, "step": 13420 }, { "grad_norm": 0.6327794194221497, "learning_rate": 9.807011745775132e-05, "loss": 0.071, "step": 13430 }, { "grad_norm": 0.5471040606498718, "learning_rate": 9.806556536727061e-05, "loss": 0.0734, "step": 13440 }, { "grad_norm": 0.5299858450889587, "learning_rate": 9.806100802041193e-05, "loss": 0.0845, "step": 13450 }, { "grad_norm": 0.6439458131790161, "learning_rate": 9.805644541767368e-05, "loss": 0.078, "step": 13460 }, { "grad_norm": 0.5737595558166504, "learning_rate": 9.805187755955478e-05, "loss": 0.0877, "step": 13470 }, { "grad_norm": 0.630183219909668, "learning_rate": 9.804730444655483e-05, "loss": 0.0813, "step": 13480 }, { "grad_norm": 0.6070468425750732, "learning_rate": 9.804272607917388e-05, "loss": 0.0828, "step": 13490 }, { "grad_norm": 0.47908279299736023, "learning_rate": 9.803814245791265e-05, "loss": 0.0687, "step": 13500 }, { "grad_norm": 0.5326957702636719, "learning_rate": 9.803355358327239e-05, "loss": 0.0849, "step": 13510 }, { "grad_norm": 0.36869147419929504, "learning_rate": 9.802895945575492e-05, "loss": 0.0837, "step": 13520 }, { "grad_norm": 0.6292535662651062, "learning_rate": 9.802436007586266e-05, "loss": 0.0794, "step": 13530 }, { "grad_norm": 0.5967928767204285, "learning_rate": 9.801975544409858e-05, "loss": 0.0844, "step": 13540 }, { "grad_norm": 0.4865572154521942, "learning_rate": 9.801514556096625e-05, "loss": 0.0784, "step": 13550 }, { "grad_norm": 0.5708318948745728, "learning_rate": 9.801053042696977e-05, "loss": 0.0854, "step": 13560 }, { "grad_norm": 0.5017790794372559, "learning_rate": 9.800591004261388e-05, "loss": 0.0713, "step": 13570 }, { "grad_norm": 0.4848058819770813, "learning_rate": 9.800128440840385e-05, "loss": 0.0787, "step": 13580 }, { "grad_norm": 0.5729997158050537, "learning_rate": 9.799665352484552e-05, "loss": 0.0951, "step": 13590 }, { "grad_norm": 0.5006541609764099, "learning_rate": 9.799201739244532e-05, "loss": 0.0878, "step": 13600 }, { "grad_norm": 0.7567723393440247, "learning_rate": 9.798737601171025e-05, "loss": 0.0785, "step": 13610 }, { "grad_norm": 0.7868842482566833, "learning_rate": 9.79827293831479e-05, "loss": 0.0946, "step": 13620 }, { "grad_norm": 0.5851988792419434, "learning_rate": 9.797807750726638e-05, "loss": 0.0654, "step": 13630 }, { "grad_norm": 0.487657368183136, "learning_rate": 9.797342038457446e-05, "loss": 0.0661, "step": 13640 }, { "grad_norm": 0.6399727463722229, "learning_rate": 9.796875801558141e-05, "loss": 0.0693, "step": 13650 }, { "grad_norm": 0.616977870464325, "learning_rate": 9.79640904007971e-05, "loss": 0.0881, "step": 13660 }, { "grad_norm": 0.6867347359657288, "learning_rate": 9.795941754073199e-05, "loss": 0.0773, "step": 13670 }, { "grad_norm": 0.4292350709438324, "learning_rate": 9.795473943589705e-05, "loss": 0.067, "step": 13680 }, { "grad_norm": 0.6631706357002258, "learning_rate": 9.795005608680394e-05, "loss": 0.0686, "step": 13690 }, { "grad_norm": 0.45714443922042847, "learning_rate": 9.794536749396477e-05, "loss": 0.0787, "step": 13700 }, { "grad_norm": 0.466202050447464, "learning_rate": 9.79406736578923e-05, "loss": 0.058, "step": 13710 }, { "grad_norm": 0.5827690362930298, "learning_rate": 9.793597457909984e-05, "loss": 0.0794, "step": 13720 }, { "grad_norm": 0.6889392733573914, "learning_rate": 9.793127025810127e-05, "loss": 0.0661, "step": 13730 }, { "grad_norm": 0.4960564374923706, "learning_rate": 9.792656069541104e-05, "loss": 0.0766, "step": 13740 }, { "grad_norm": 0.4531016945838928, "learning_rate": 9.79218458915442e-05, "loss": 0.0648, "step": 13750 }, { "grad_norm": 0.5182006359100342, "learning_rate": 9.791712584701634e-05, "loss": 0.0843, "step": 13760 }, { "grad_norm": 0.6105155348777771, "learning_rate": 9.791240056234364e-05, "loss": 0.0741, "step": 13770 }, { "grad_norm": 0.4919947683811188, "learning_rate": 9.790767003804283e-05, "loss": 0.0686, "step": 13780 }, { "grad_norm": 0.6065329313278198, "learning_rate": 9.790293427463126e-05, "loss": 0.0671, "step": 13790 }, { "grad_norm": 0.4782351851463318, "learning_rate": 9.789819327262684e-05, "loss": 0.0617, "step": 13800 }, { "grad_norm": 0.573773205280304, "learning_rate": 9.7893447032548e-05, "loss": 0.0761, "step": 13810 }, { "grad_norm": 0.491957426071167, "learning_rate": 9.78886955549138e-05, "loss": 0.0806, "step": 13820 }, { "grad_norm": 0.8361251950263977, "learning_rate": 9.788393884024387e-05, "loss": 0.0775, "step": 13830 }, { "grad_norm": 0.8143141269683838, "learning_rate": 9.787917688905836e-05, "loss": 0.0854, "step": 13840 }, { "grad_norm": 0.7790778875350952, "learning_rate": 9.787440970187807e-05, "loss": 0.0713, "step": 13850 }, { "grad_norm": 0.4006319046020508, "learning_rate": 9.786963727922429e-05, "loss": 0.0688, "step": 13860 }, { "grad_norm": 0.6993169188499451, "learning_rate": 9.786485962161897e-05, "loss": 0.0787, "step": 13870 }, { "grad_norm": 0.4108664095401764, "learning_rate": 9.786007672958455e-05, "loss": 0.0587, "step": 13880 }, { "grad_norm": 0.6787965297698975, "learning_rate": 9.78552886036441e-05, "loss": 0.0891, "step": 13890 }, { "grad_norm": 0.43815702199935913, "learning_rate": 9.785049524432124e-05, "loss": 0.0602, "step": 13900 }, { "grad_norm": 0.4991980791091919, "learning_rate": 9.784569665214016e-05, "loss": 0.0573, "step": 13910 }, { "grad_norm": 0.7330456972122192, "learning_rate": 9.784089282762563e-05, "loss": 0.0791, "step": 13920 }, { "grad_norm": 0.6439687013626099, "learning_rate": 9.7836083771303e-05, "loss": 0.0669, "step": 13930 }, { "grad_norm": 0.5741353631019592, "learning_rate": 9.783126948369817e-05, "loss": 0.0745, "step": 13940 }, { "grad_norm": 0.5436040759086609, "learning_rate": 9.78264499653376e-05, "loss": 0.0628, "step": 13950 }, { "grad_norm": 0.36061981320381165, "learning_rate": 9.782162521674838e-05, "loss": 0.0562, "step": 13960 }, { "grad_norm": 0.3855211138725281, "learning_rate": 9.781679523845812e-05, "loss": 0.0865, "step": 13970 }, { "grad_norm": 0.4706871807575226, "learning_rate": 9.781196003099502e-05, "loss": 0.0692, "step": 13980 }, { "grad_norm": 0.5846919417381287, "learning_rate": 9.780711959488786e-05, "loss": 0.0686, "step": 13990 }, { "grad_norm": 0.5890252590179443, "learning_rate": 9.780227393066599e-05, "loss": 0.07, "step": 14000 }, { "grad_norm": 0.44215962290763855, "learning_rate": 9.77974230388593e-05, "loss": 0.0727, "step": 14010 }, { "grad_norm": 0.41900649666786194, "learning_rate": 9.779256691999829e-05, "loss": 0.076, "step": 14020 }, { "grad_norm": 0.7619921565055847, "learning_rate": 9.778770557461403e-05, "loss": 0.0833, "step": 14030 }, { "grad_norm": 0.6022079586982727, "learning_rate": 9.778283900323812e-05, "loss": 0.0657, "step": 14040 }, { "grad_norm": 0.4499014914035797, "learning_rate": 9.777796720640277e-05, "loss": 0.058, "step": 14050 }, { "grad_norm": 0.4090447723865509, "learning_rate": 9.777309018464078e-05, "loss": 0.0774, "step": 14060 }, { "grad_norm": 0.4693705439567566, "learning_rate": 9.776820793848547e-05, "loss": 0.0693, "step": 14070 }, { "grad_norm": 0.800218403339386, "learning_rate": 9.776332046847075e-05, "loss": 0.0654, "step": 14080 }, { "grad_norm": 0.4710763990879059, "learning_rate": 9.775842777513111e-05, "loss": 0.0703, "step": 14090 }, { "grad_norm": 0.33448460698127747, "learning_rate": 9.775352985900163e-05, "loss": 0.0771, "step": 14100 }, { "grad_norm": 0.4524190127849579, "learning_rate": 9.774862672061791e-05, "loss": 0.0807, "step": 14110 }, { "grad_norm": 0.4843241572380066, "learning_rate": 9.774371836051616e-05, "loss": 0.0766, "step": 14120 }, { "grad_norm": 0.5384321808815002, "learning_rate": 9.773880477923315e-05, "loss": 0.0747, "step": 14130 }, { "grad_norm": 0.5420536994934082, "learning_rate": 9.773388597730623e-05, "loss": 0.0796, "step": 14140 }, { "grad_norm": 0.49807009100914, "learning_rate": 9.77289619552733e-05, "loss": 0.0679, "step": 14150 }, { "grad_norm": 0.5239222645759583, "learning_rate": 9.772403271367285e-05, "loss": 0.0932, "step": 14160 }, { "grad_norm": 0.8430804014205933, "learning_rate": 9.771909825304396e-05, "loss": 0.0703, "step": 14170 }, { "grad_norm": 0.4618360102176666, "learning_rate": 9.771415857392619e-05, "loss": 0.0779, "step": 14180 }, { "grad_norm": 0.6998855471611023, "learning_rate": 9.770921367685978e-05, "loss": 0.0947, "step": 14190 }, { "grad_norm": 0.7474045753479004, "learning_rate": 9.770426356238551e-05, "loss": 0.0782, "step": 14200 }, { "grad_norm": 0.48530375957489014, "learning_rate": 9.769930823104469e-05, "loss": 0.0744, "step": 14210 }, { "grad_norm": 0.5442929267883301, "learning_rate": 9.769434768337926e-05, "loss": 0.0751, "step": 14220 }, { "grad_norm": 0.5016579627990723, "learning_rate": 9.768938191993164e-05, "loss": 0.0747, "step": 14230 }, { "grad_norm": 0.6522194743156433, "learning_rate": 9.768441094124494e-05, "loss": 0.0645, "step": 14240 }, { "grad_norm": 0.3965444564819336, "learning_rate": 9.767943474786275e-05, "loss": 0.0753, "step": 14250 }, { "grad_norm": 0.47319209575653076, "learning_rate": 9.767445334032923e-05, "loss": 0.0599, "step": 14260 }, { "grad_norm": 0.5333196520805359, "learning_rate": 9.766946671918919e-05, "loss": 0.0819, "step": 14270 }, { "grad_norm": 0.5734333992004395, "learning_rate": 9.766447488498796e-05, "loss": 0.0717, "step": 14280 }, { "grad_norm": 0.6006932854652405, "learning_rate": 9.765947783827139e-05, "loss": 0.0688, "step": 14290 }, { "grad_norm": 1.1924039125442505, "learning_rate": 9.765447557958599e-05, "loss": 0.0864, "step": 14300 }, { "grad_norm": 0.522357165813446, "learning_rate": 9.764946810947879e-05, "loss": 0.0719, "step": 14310 }, { "grad_norm": 0.4209396243095398, "learning_rate": 9.764445542849738e-05, "loss": 0.0785, "step": 14320 }, { "grad_norm": 0.6772891879081726, "learning_rate": 9.763943753718998e-05, "loss": 0.068, "step": 14330 }, { "grad_norm": 0.6110374927520752, "learning_rate": 9.76344144361053e-05, "loss": 0.0748, "step": 14340 }, { "grad_norm": 0.7279322147369385, "learning_rate": 9.762938612579269e-05, "loss": 0.0731, "step": 14350 }, { "grad_norm": 0.5074071884155273, "learning_rate": 9.762435260680202e-05, "loss": 0.074, "step": 14360 }, { "grad_norm": 0.9235722422599792, "learning_rate": 9.761931387968373e-05, "loss": 0.079, "step": 14370 }, { "grad_norm": 0.4857359528541565, "learning_rate": 9.76142699449889e-05, "loss": 0.0732, "step": 14380 }, { "grad_norm": 0.5619069337844849, "learning_rate": 9.760922080326908e-05, "loss": 0.0824, "step": 14390 }, { "grad_norm": 0.7266397476196289, "learning_rate": 9.760416645507644e-05, "loss": 0.0741, "step": 14400 }, { "grad_norm": 0.5790905356407166, "learning_rate": 9.759910690096375e-05, "loss": 0.0745, "step": 14410 }, { "grad_norm": 0.6105474829673767, "learning_rate": 9.759404214148429e-05, "loss": 0.0739, "step": 14420 }, { "grad_norm": 0.39097654819488525, "learning_rate": 9.758897217719191e-05, "loss": 0.0739, "step": 14430 }, { "grad_norm": 0.5080263614654541, "learning_rate": 9.758389700864113e-05, "loss": 0.0653, "step": 14440 }, { "grad_norm": 0.5531677603721619, "learning_rate": 9.757881663638688e-05, "loss": 0.0667, "step": 14450 }, { "grad_norm": 0.4652726948261261, "learning_rate": 9.757373106098478e-05, "loss": 0.0673, "step": 14460 }, { "grad_norm": 0.45812687277793884, "learning_rate": 9.756864028299097e-05, "loss": 0.0781, "step": 14470 }, { "grad_norm": 0.45285752415657043, "learning_rate": 9.75635443029622e-05, "loss": 0.068, "step": 14480 }, { "grad_norm": 0.4998152256011963, "learning_rate": 9.755844312145572e-05, "loss": 0.0742, "step": 14490 }, { "grad_norm": 0.5164082646369934, "learning_rate": 9.755333673902941e-05, "loss": 0.0734, "step": 14500 }, { "grad_norm": 0.5931414365768433, "learning_rate": 9.75482251562417e-05, "loss": 0.083, "step": 14510 }, { "grad_norm": 0.47123849391937256, "learning_rate": 9.754310837365155e-05, "loss": 0.0715, "step": 14520 }, { "grad_norm": 0.45103609561920166, "learning_rate": 9.753798639181856e-05, "loss": 0.1005, "step": 14530 }, { "grad_norm": 0.6345214247703552, "learning_rate": 9.753285921130286e-05, "loss": 0.0778, "step": 14540 }, { "grad_norm": 0.48955559730529785, "learning_rate": 9.752772683266512e-05, "loss": 0.0739, "step": 14550 }, { "grad_norm": 0.699445903301239, "learning_rate": 9.752258925646665e-05, "loss": 0.0848, "step": 14560 }, { "grad_norm": 0.4686570167541504, "learning_rate": 9.751744648326926e-05, "loss": 0.0713, "step": 14570 }, { "grad_norm": 0.6389544010162354, "learning_rate": 9.751229851363536e-05, "loss": 0.0845, "step": 14580 }, { "grad_norm": 0.4949929118156433, "learning_rate": 9.750714534812793e-05, "loss": 0.0696, "step": 14590 }, { "grad_norm": 0.8234508037567139, "learning_rate": 9.750198698731053e-05, "loss": 0.0922, "step": 14600 }, { "grad_norm": 1.0231306552886963, "learning_rate": 9.749682343174722e-05, "loss": 0.0787, "step": 14610 }, { "grad_norm": 0.5136340260505676, "learning_rate": 9.749165468200272e-05, "loss": 0.0724, "step": 14620 }, { "grad_norm": 0.5492709875106812, "learning_rate": 9.748648073864229e-05, "loss": 0.0654, "step": 14630 }, { "grad_norm": 0.4770876467227936, "learning_rate": 9.748130160223168e-05, "loss": 0.0704, "step": 14640 }, { "grad_norm": 0.6930038332939148, "learning_rate": 9.747611727333734e-05, "loss": 0.0668, "step": 14650 }, { "grad_norm": 0.45779141783714294, "learning_rate": 9.74709277525262e-05, "loss": 0.0738, "step": 14660 }, { "grad_norm": 0.9312993288040161, "learning_rate": 9.746573304036576e-05, "loss": 0.0873, "step": 14670 }, { "grad_norm": 1.1958178281784058, "learning_rate": 9.746053313742412e-05, "loss": 0.0698, "step": 14680 }, { "grad_norm": 0.45898106694221497, "learning_rate": 9.745532804426994e-05, "loss": 0.078, "step": 14690 }, { "grad_norm": 0.4430961310863495, "learning_rate": 9.745011776147242e-05, "loss": 0.0707, "step": 14700 }, { "grad_norm": 0.7039397358894348, "learning_rate": 9.744490228960138e-05, "loss": 0.0788, "step": 14710 }, { "grad_norm": 0.4299646317958832, "learning_rate": 9.743968162922713e-05, "loss": 0.0718, "step": 14720 }, { "grad_norm": 3.5491762161254883, "learning_rate": 9.743445578092064e-05, "loss": 0.0954, "step": 14730 }, { "grad_norm": 0.4489254653453827, "learning_rate": 9.742922474525338e-05, "loss": 0.1032, "step": 14740 }, { "grad_norm": 0.5233479738235474, "learning_rate": 9.742398852279741e-05, "loss": 0.0848, "step": 14750 }, { "grad_norm": 0.6704705357551575, "learning_rate": 9.741874711412535e-05, "loss": 0.0796, "step": 14760 }, { "grad_norm": 0.6547485589981079, "learning_rate": 9.741350051981042e-05, "loss": 0.0792, "step": 14770 }, { "grad_norm": 0.5232723355293274, "learning_rate": 9.740824874042633e-05, "loss": 0.0764, "step": 14780 }, { "grad_norm": 0.5346686244010925, "learning_rate": 9.740299177654746e-05, "loss": 0.0656, "step": 14790 }, { "grad_norm": 0.6000982522964478, "learning_rate": 9.739772962874867e-05, "loss": 0.0742, "step": 14800 }, { "grad_norm": 0.5696518421173096, "learning_rate": 9.739246229760541e-05, "loss": 0.0678, "step": 14810 }, { "grad_norm": 0.5506489276885986, "learning_rate": 9.738718978369376e-05, "loss": 0.0767, "step": 14820 }, { "grad_norm": 0.6869252324104309, "learning_rate": 9.738191208759025e-05, "loss": 0.0843, "step": 14830 }, { "grad_norm": 0.46502506732940674, "learning_rate": 9.73766292098721e-05, "loss": 0.0921, "step": 14840 }, { "grad_norm": 0.5337228178977966, "learning_rate": 9.737134115111699e-05, "loss": 0.0872, "step": 14850 }, { "grad_norm": 0.41099175810813904, "learning_rate": 9.736604791190323e-05, "loss": 0.0704, "step": 14860 }, { "grad_norm": 0.3418247103691101, "learning_rate": 9.73607494928097e-05, "loss": 0.0619, "step": 14870 }, { "grad_norm": 0.4552450180053711, "learning_rate": 9.735544589441581e-05, "loss": 0.0716, "step": 14880 }, { "grad_norm": 0.4306298494338989, "learning_rate": 9.735013711730154e-05, "loss": 0.0695, "step": 14890 }, { "grad_norm": 0.5773239731788635, "learning_rate": 9.734482316204747e-05, "loss": 0.0675, "step": 14900 }, { "grad_norm": 0.7119613885879517, "learning_rate": 9.733950402923473e-05, "loss": 0.0711, "step": 14910 }, { "grad_norm": 0.39098605513572693, "learning_rate": 9.7334179719445e-05, "loss": 0.0602, "step": 14920 }, { "grad_norm": 0.6815020442008972, "learning_rate": 9.732885023326053e-05, "loss": 0.0785, "step": 14930 }, { "grad_norm": 0.4825390577316284, "learning_rate": 9.732351557126418e-05, "loss": 0.0797, "step": 14940 }, { "grad_norm": 0.5751298666000366, "learning_rate": 9.731817573403929e-05, "loss": 0.0816, "step": 14950 }, { "grad_norm": 0.5657252073287964, "learning_rate": 9.731283072216985e-05, "loss": 0.0879, "step": 14960 }, { "grad_norm": 0.7368891835212708, "learning_rate": 9.730748053624039e-05, "loss": 0.0757, "step": 14970 }, { "grad_norm": 0.4262913465499878, "learning_rate": 9.730212517683598e-05, "loss": 0.0696, "step": 14980 }, { "grad_norm": 0.6402279138565063, "learning_rate": 9.729676464454228e-05, "loss": 0.0756, "step": 14990 }, { "grad_norm": 0.6089183688163757, "learning_rate": 9.72913989399455e-05, "loss": 0.0772, "step": 15000 }, { "grad_norm": 0.47334718704223633, "learning_rate": 9.728602806363242e-05, "loss": 0.0615, "step": 15010 }, { "grad_norm": 0.7348153591156006, "learning_rate": 9.728065201619043e-05, "loss": 0.0676, "step": 15020 }, { "grad_norm": 0.5456428527832031, "learning_rate": 9.727527079820742e-05, "loss": 0.0691, "step": 15030 }, { "grad_norm": 0.6294335722923279, "learning_rate": 9.726988441027186e-05, "loss": 0.0751, "step": 15040 }, { "grad_norm": 0.4619937241077423, "learning_rate": 9.726449285297281e-05, "loss": 0.0792, "step": 15050 }, { "grad_norm": 0.7406962513923645, "learning_rate": 9.72590961268999e-05, "loss": 0.0762, "step": 15060 }, { "grad_norm": 0.7234615087509155, "learning_rate": 9.725369423264328e-05, "loss": 0.0718, "step": 15070 }, { "grad_norm": 0.3690895438194275, "learning_rate": 9.72482871707937e-05, "loss": 0.0763, "step": 15080 }, { "grad_norm": 0.49230337142944336, "learning_rate": 9.724287494194247e-05, "loss": 0.074, "step": 15090 }, { "grad_norm": 0.6786134243011475, "learning_rate": 9.723745754668147e-05, "loss": 0.0626, "step": 15100 }, { "grad_norm": 0.5720148682594299, "learning_rate": 9.723203498560313e-05, "loss": 0.0692, "step": 15110 }, { "grad_norm": 0.48764729499816895, "learning_rate": 9.722660725930046e-05, "loss": 0.0583, "step": 15120 }, { "grad_norm": 0.6329078078269958, "learning_rate": 9.722117436836702e-05, "loss": 0.0657, "step": 15130 }, { "grad_norm": 0.5986809134483337, "learning_rate": 9.721573631339696e-05, "loss": 0.057, "step": 15140 }, { "grad_norm": 0.5179982781410217, "learning_rate": 9.721029309498494e-05, "loss": 0.0689, "step": 15150 }, { "grad_norm": 0.6920948028564453, "learning_rate": 9.720484471372627e-05, "loss": 0.0664, "step": 15160 }, { "grad_norm": 0.5048090219497681, "learning_rate": 9.719939117021673e-05, "loss": 0.0713, "step": 15170 }, { "grad_norm": 0.5907307863235474, "learning_rate": 9.719393246505275e-05, "loss": 0.0712, "step": 15180 }, { "grad_norm": 0.541496992111206, "learning_rate": 9.718846859883128e-05, "loss": 0.068, "step": 15190 }, { "grad_norm": 0.5465351939201355, "learning_rate": 9.718299957214982e-05, "loss": 0.0837, "step": 15200 }, { "grad_norm": 0.7846565842628479, "learning_rate": 9.717752538560646e-05, "loss": 0.074, "step": 15210 }, { "grad_norm": 0.6107269525527954, "learning_rate": 9.717204603979986e-05, "loss": 0.067, "step": 15220 }, { "grad_norm": 0.5067793130874634, "learning_rate": 9.716656153532922e-05, "loss": 0.0646, "step": 15230 }, { "grad_norm": 0.5482295155525208, "learning_rate": 9.716107187279434e-05, "loss": 0.0623, "step": 15240 }, { "grad_norm": 0.9483293294906616, "learning_rate": 9.715557705279555e-05, "loss": 0.0843, "step": 15250 }, { "grad_norm": 0.4769880473613739, "learning_rate": 9.715007707593372e-05, "loss": 0.0718, "step": 15260 }, { "grad_norm": 0.801807165145874, "learning_rate": 9.714457194281036e-05, "loss": 0.0664, "step": 15270 }, { "grad_norm": 0.8985933661460876, "learning_rate": 9.713906165402751e-05, "loss": 0.0874, "step": 15280 }, { "grad_norm": 0.5008364319801331, "learning_rate": 9.713354621018774e-05, "loss": 0.0942, "step": 15290 }, { "grad_norm": 0.5687165856361389, "learning_rate": 9.712802561189422e-05, "loss": 0.0718, "step": 15300 }, { "grad_norm": 0.6319238543510437, "learning_rate": 9.712249985975069e-05, "loss": 0.0812, "step": 15310 }, { "grad_norm": 0.9873611927032471, "learning_rate": 9.71169689543614e-05, "loss": 0.0752, "step": 15320 }, { "grad_norm": 0.6009294390678406, "learning_rate": 9.711143289633123e-05, "loss": 0.0696, "step": 15330 }, { "grad_norm": 0.46998342871665955, "learning_rate": 9.710589168626561e-05, "loss": 0.0866, "step": 15340 }, { "grad_norm": 0.5264092087745667, "learning_rate": 9.710034532477048e-05, "loss": 0.0871, "step": 15350 }, { "grad_norm": 0.48826873302459717, "learning_rate": 9.709479381245239e-05, "loss": 0.0715, "step": 15360 }, { "grad_norm": 0.5643842220306396, "learning_rate": 9.708923714991847e-05, "loss": 0.072, "step": 15370 }, { "grad_norm": 0.5494413375854492, "learning_rate": 9.708367533777638e-05, "loss": 0.0733, "step": 15380 }, { "grad_norm": 0.5508524775505066, "learning_rate": 9.707810837663431e-05, "loss": 0.0679, "step": 15390 }, { "grad_norm": 0.36473318934440613, "learning_rate": 9.707253626710113e-05, "loss": 0.061, "step": 15400 }, { "grad_norm": 0.34637704491615295, "learning_rate": 9.706695900978613e-05, "loss": 0.0598, "step": 15410 }, { "grad_norm": 0.5263386964797974, "learning_rate": 9.706137660529926e-05, "loss": 0.0745, "step": 15420 }, { "grad_norm": 0.6394739747047424, "learning_rate": 9.705578905425101e-05, "loss": 0.0882, "step": 15430 }, { "grad_norm": 0.47641614079475403, "learning_rate": 9.705019635725241e-05, "loss": 0.0645, "step": 15440 }, { "grad_norm": 0.37260493636131287, "learning_rate": 9.704459851491508e-05, "loss": 0.0648, "step": 15450 }, { "grad_norm": 0.5600638389587402, "learning_rate": 9.703899552785118e-05, "loss": 0.0628, "step": 15460 }, { "grad_norm": 0.9494820833206177, "learning_rate": 9.703338739667346e-05, "loss": 0.0622, "step": 15470 }, { "grad_norm": 0.4898974597454071, "learning_rate": 9.70277741219952e-05, "loss": 0.0719, "step": 15480 }, { "grad_norm": 0.49783897399902344, "learning_rate": 9.702215570443027e-05, "loss": 0.0774, "step": 15490 }, { "grad_norm": 0.38654860854148865, "learning_rate": 9.701653214459309e-05, "loss": 0.0708, "step": 15500 }, { "grad_norm": 0.5247926115989685, "learning_rate": 9.701090344309865e-05, "loss": 0.079, "step": 15510 }, { "grad_norm": 0.7810113430023193, "learning_rate": 9.700526960056247e-05, "loss": 0.0761, "step": 15520 }, { "grad_norm": 0.4405263066291809, "learning_rate": 9.699963061760068e-05, "loss": 0.06, "step": 15530 }, { "grad_norm": 0.5190252661705017, "learning_rate": 9.699398649482997e-05, "loss": 0.0686, "step": 15540 }, { "grad_norm": 0.4706776738166809, "learning_rate": 9.698833723286753e-05, "loss": 0.061, "step": 15550 }, { "grad_norm": 0.5405471324920654, "learning_rate": 9.698268283233118e-05, "loss": 0.08, "step": 15560 }, { "grad_norm": 0.5507063865661621, "learning_rate": 9.697702329383929e-05, "loss": 0.0699, "step": 15570 }, { "grad_norm": 0.48683393001556396, "learning_rate": 9.697135861801074e-05, "loss": 0.074, "step": 15580 }, { "grad_norm": 0.5608963370323181, "learning_rate": 9.696568880546505e-05, "loss": 0.0547, "step": 15590 }, { "grad_norm": 0.5039353370666504, "learning_rate": 9.696001385682223e-05, "loss": 0.0672, "step": 15600 }, { "grad_norm": 0.4180893003940582, "learning_rate": 9.695433377270291e-05, "loss": 0.067, "step": 15610 }, { "grad_norm": 0.3595694899559021, "learning_rate": 9.694864855372824e-05, "loss": 0.0607, "step": 15620 }, { "grad_norm": 0.5771121382713318, "learning_rate": 9.694295820051995e-05, "loss": 0.0649, "step": 15630 }, { "grad_norm": 0.3749651610851288, "learning_rate": 9.693726271370032e-05, "loss": 0.0557, "step": 15640 }, { "grad_norm": 0.5545381903648376, "learning_rate": 9.693156209389221e-05, "loss": 0.0566, "step": 15650 }, { "grad_norm": 0.5134329795837402, "learning_rate": 9.692585634171905e-05, "loss": 0.0688, "step": 15660 }, { "grad_norm": 0.45968401432037354, "learning_rate": 9.692014545780476e-05, "loss": 0.0665, "step": 15670 }, { "grad_norm": 0.4060109257698059, "learning_rate": 9.691442944277393e-05, "loss": 0.0483, "step": 15680 }, { "grad_norm": 0.41127917170524597, "learning_rate": 9.690870829725162e-05, "loss": 0.0548, "step": 15690 }, { "grad_norm": 0.7100592255592346, "learning_rate": 9.69029820218635e-05, "loss": 0.0606, "step": 15700 }, { "grad_norm": 0.6626259684562683, "learning_rate": 9.689725061723579e-05, "loss": 0.0706, "step": 15710 }, { "grad_norm": 1.2747795581817627, "learning_rate": 9.689151408399527e-05, "loss": 0.0589, "step": 15720 }, { "grad_norm": 0.5705156326293945, "learning_rate": 9.688577242276924e-05, "loss": 0.0878, "step": 15730 }, { "grad_norm": 0.6760439276695251, "learning_rate": 9.688002563418566e-05, "loss": 0.0724, "step": 15740 }, { "grad_norm": 0.47373661398887634, "learning_rate": 9.687427371887293e-05, "loss": 0.0791, "step": 15750 }, { "grad_norm": 0.6513746976852417, "learning_rate": 9.686851667746012e-05, "loss": 0.0816, "step": 15760 }, { "grad_norm": 0.5739185810089111, "learning_rate": 9.686275451057677e-05, "loss": 0.0726, "step": 15770 }, { "grad_norm": 0.5640231966972351, "learning_rate": 9.685698721885308e-05, "loss": 0.0723, "step": 15780 }, { "grad_norm": 0.5330219268798828, "learning_rate": 9.68512148029197e-05, "loss": 0.0754, "step": 15790 }, { "grad_norm": 0.8483518362045288, "learning_rate": 9.684543726340791e-05, "loss": 0.0775, "step": 15800 }, { "grad_norm": 0.64500492811203, "learning_rate": 9.683965460094952e-05, "loss": 0.0772, "step": 15810 }, { "grad_norm": 0.3626444339752197, "learning_rate": 9.683386681617694e-05, "loss": 0.0733, "step": 15820 }, { "grad_norm": 0.4478566646575928, "learning_rate": 9.68280739097231e-05, "loss": 0.0735, "step": 15830 }, { "grad_norm": 0.5756970643997192, "learning_rate": 9.682227588222148e-05, "loss": 0.0695, "step": 15840 }, { "grad_norm": 0.5863731503486633, "learning_rate": 9.681647273430618e-05, "loss": 0.0633, "step": 15850 }, { "grad_norm": 0.46198493242263794, "learning_rate": 9.681066446661182e-05, "loss": 0.0624, "step": 15860 }, { "grad_norm": 0.4584331810474396, "learning_rate": 9.680485107977357e-05, "loss": 0.0727, "step": 15870 }, { "grad_norm": 0.48875635862350464, "learning_rate": 9.679903257442716e-05, "loss": 0.0746, "step": 15880 }, { "grad_norm": 0.43604111671447754, "learning_rate": 9.679320895120891e-05, "loss": 0.0635, "step": 15890 }, { "grad_norm": 0.39273330569267273, "learning_rate": 9.67873802107557e-05, "loss": 0.06, "step": 15900 }, { "grad_norm": 0.5335736870765686, "learning_rate": 9.67815463537049e-05, "loss": 0.0676, "step": 15910 }, { "grad_norm": 0.7653653621673584, "learning_rate": 9.677570738069457e-05, "loss": 0.0793, "step": 15920 }, { "grad_norm": 1.3132401704788208, "learning_rate": 9.676986329236318e-05, "loss": 0.0668, "step": 15930 }, { "grad_norm": 0.5551922917366028, "learning_rate": 9.676401408934987e-05, "loss": 0.0731, "step": 15940 }, { "grad_norm": 0.4306107759475708, "learning_rate": 9.675815977229428e-05, "loss": 0.0659, "step": 15950 }, { "grad_norm": 0.4778654873371124, "learning_rate": 9.675230034183664e-05, "loss": 0.0702, "step": 15960 }, { "grad_norm": 0.5945552587509155, "learning_rate": 9.674643579861773e-05, "loss": 0.0756, "step": 15970 }, { "grad_norm": 1.0460649728775024, "learning_rate": 9.674056614327886e-05, "loss": 0.0584, "step": 15980 }, { "grad_norm": 0.5573794841766357, "learning_rate": 9.673469137646198e-05, "loss": 0.0723, "step": 15990 }, { "grad_norm": 0.4695611894130707, "learning_rate": 9.67288114988095e-05, "loss": 0.0648, "step": 16000 }, { "grad_norm": 0.7116871476173401, "learning_rate": 9.672292651096447e-05, "loss": 0.0707, "step": 16010 }, { "grad_norm": 0.4624581038951874, "learning_rate": 9.671703641357042e-05, "loss": 0.0686, "step": 16020 }, { "grad_norm": 0.4715109169483185, "learning_rate": 9.67111412072715e-05, "loss": 0.073, "step": 16030 }, { "grad_norm": 0.43602338433265686, "learning_rate": 9.670524089271242e-05, "loss": 0.0575, "step": 16040 }, { "grad_norm": 0.5334305167198181, "learning_rate": 9.669933547053842e-05, "loss": 0.0567, "step": 16050 }, { "grad_norm": 0.465320348739624, "learning_rate": 9.669342494139531e-05, "loss": 0.0654, "step": 16060 }, { "grad_norm": 0.3965696394443512, "learning_rate": 9.668750930592943e-05, "loss": 0.0543, "step": 16070 }, { "grad_norm": 0.49167364835739136, "learning_rate": 9.668158856478775e-05, "loss": 0.0617, "step": 16080 }, { "grad_norm": 0.4966554045677185, "learning_rate": 9.66756627186177e-05, "loss": 0.0736, "step": 16090 }, { "grad_norm": 0.4933612048625946, "learning_rate": 9.666973176806737e-05, "loss": 0.0712, "step": 16100 }, { "grad_norm": 0.7078776359558105, "learning_rate": 9.666379571378534e-05, "loss": 0.0747, "step": 16110 }, { "grad_norm": 0.5713638663291931, "learning_rate": 9.665785455642076e-05, "loss": 0.0876, "step": 16120 }, { "grad_norm": 0.7107515931129456, "learning_rate": 9.665190829662337e-05, "loss": 0.0728, "step": 16130 }, { "grad_norm": 0.6691153049468994, "learning_rate": 9.664595693504342e-05, "loss": 0.0683, "step": 16140 }, { "grad_norm": 0.5544379353523254, "learning_rate": 9.664000047233175e-05, "loss": 0.0732, "step": 16150 }, { "grad_norm": 0.3817453682422638, "learning_rate": 9.663403890913976e-05, "loss": 0.0655, "step": 16160 }, { "grad_norm": 0.6377798914909363, "learning_rate": 9.662807224611938e-05, "loss": 0.0693, "step": 16170 }, { "grad_norm": 0.45731931924819946, "learning_rate": 9.662210048392311e-05, "loss": 0.0582, "step": 16180 }, { "grad_norm": 0.5143126249313354, "learning_rate": 9.661612362320405e-05, "loss": 0.0665, "step": 16190 }, { "grad_norm": 0.465679407119751, "learning_rate": 9.661014166461579e-05, "loss": 0.0687, "step": 16200 }, { "grad_norm": 0.539967954158783, "learning_rate": 9.66041546088125e-05, "loss": 0.0787, "step": 16210 }, { "grad_norm": 0.4111764430999756, "learning_rate": 9.659816245644895e-05, "loss": 0.0583, "step": 16220 }, { "grad_norm": 0.5679903626441956, "learning_rate": 9.65921652081804e-05, "loss": 0.0656, "step": 16230 }, { "grad_norm": 0.6125041842460632, "learning_rate": 9.658616286466271e-05, "loss": 0.0715, "step": 16240 }, { "grad_norm": 0.5275124907493591, "learning_rate": 9.65801554265523e-05, "loss": 0.0642, "step": 16250 }, { "grad_norm": 0.588887095451355, "learning_rate": 9.657414289450612e-05, "loss": 0.0718, "step": 16260 }, { "grad_norm": 0.5471784472465515, "learning_rate": 9.656812526918171e-05, "loss": 0.0654, "step": 16270 }, { "grad_norm": 0.45912450551986694, "learning_rate": 9.656210255123712e-05, "loss": 0.0751, "step": 16280 }, { "grad_norm": 2.0246894359588623, "learning_rate": 9.6556074741331e-05, "loss": 0.0701, "step": 16290 }, { "grad_norm": 1.0709949731826782, "learning_rate": 9.655004184012256e-05, "loss": 0.085, "step": 16300 }, { "grad_norm": 0.6020469665527344, "learning_rate": 9.654400384827152e-05, "loss": 0.1033, "step": 16310 }, { "grad_norm": 0.5083732604980469, "learning_rate": 9.653796076643818e-05, "loss": 0.0738, "step": 16320 }, { "grad_norm": 0.4648799002170563, "learning_rate": 9.653191259528344e-05, "loss": 0.0741, "step": 16330 }, { "grad_norm": 0.6976860761642456, "learning_rate": 9.65258593354687e-05, "loss": 0.0606, "step": 16340 }, { "grad_norm": 0.49402377009391785, "learning_rate": 9.651980098765591e-05, "loss": 0.0705, "step": 16350 }, { "grad_norm": 0.47042518854141235, "learning_rate": 9.651373755250765e-05, "loss": 0.0653, "step": 16360 }, { "grad_norm": 0.5118752717971802, "learning_rate": 9.650766903068697e-05, "loss": 0.072, "step": 16370 }, { "grad_norm": 0.4791579246520996, "learning_rate": 9.650159542285753e-05, "loss": 0.0624, "step": 16380 }, { "grad_norm": 0.4036940336227417, "learning_rate": 9.649551672968353e-05, "loss": 0.0715, "step": 16390 }, { "grad_norm": 0.6375586986541748, "learning_rate": 9.648943295182973e-05, "loss": 0.0727, "step": 16400 }, { "grad_norm": 0.5176889300346375, "learning_rate": 9.648334408996144e-05, "loss": 0.0782, "step": 16410 }, { "grad_norm": 0.697277307510376, "learning_rate": 9.647725014474452e-05, "loss": 0.0773, "step": 16420 }, { "grad_norm": 0.5383924245834351, "learning_rate": 9.64711511168454e-05, "loss": 0.0744, "step": 16430 }, { "grad_norm": 0.4934294521808624, "learning_rate": 9.646504700693108e-05, "loss": 0.0657, "step": 16440 }, { "grad_norm": 0.693514883518219, "learning_rate": 9.645893781566907e-05, "loss": 0.0606, "step": 16450 }, { "grad_norm": 0.614876925945282, "learning_rate": 9.645282354372744e-05, "loss": 0.0639, "step": 16460 }, { "grad_norm": 0.7283938527107239, "learning_rate": 9.644670419177491e-05, "loss": 0.0615, "step": 16470 }, { "grad_norm": 0.6203370690345764, "learning_rate": 9.644057976048062e-05, "loss": 0.0706, "step": 16480 }, { "grad_norm": 0.4732312858104706, "learning_rate": 9.643445025051435e-05, "loss": 0.0637, "step": 16490 }, { "grad_norm": 0.6799542307853699, "learning_rate": 9.642831566254641e-05, "loss": 0.0729, "step": 16500 }, { "grad_norm": 1.1437608003616333, "learning_rate": 9.642217599724769e-05, "loss": 0.0822, "step": 16510 }, { "grad_norm": 0.5479095578193665, "learning_rate": 9.64160312552896e-05, "loss": 0.0794, "step": 16520 }, { "grad_norm": 0.5769115686416626, "learning_rate": 9.64098814373441e-05, "loss": 0.0715, "step": 16530 }, { "grad_norm": 0.6210693717002869, "learning_rate": 9.640372654408374e-05, "loss": 0.0595, "step": 16540 }, { "grad_norm": 0.46328964829444885, "learning_rate": 9.639756657618162e-05, "loss": 0.0627, "step": 16550 }, { "grad_norm": 0.48436883091926575, "learning_rate": 9.639140153431138e-05, "loss": 0.0706, "step": 16560 }, { "grad_norm": 0.5712242722511292, "learning_rate": 9.638523141914721e-05, "loss": 0.0656, "step": 16570 }, { "grad_norm": 0.47568032145500183, "learning_rate": 9.637905623136388e-05, "loss": 0.0762, "step": 16580 }, { "grad_norm": 0.6214029788970947, "learning_rate": 9.637287597163669e-05, "loss": 0.0494, "step": 16590 }, { "grad_norm": 0.4812009036540985, "learning_rate": 9.63666906406415e-05, "loss": 0.0599, "step": 16600 }, { "grad_norm": 0.6153444647789001, "learning_rate": 9.636050023905473e-05, "loss": 0.0605, "step": 16610 }, { "grad_norm": 0.4594849646091461, "learning_rate": 9.635430476755336e-05, "loss": 0.0602, "step": 16620 }, { "grad_norm": 0.5607725977897644, "learning_rate": 9.63481042268149e-05, "loss": 0.0577, "step": 16630 }, { "grad_norm": 0.5533176064491272, "learning_rate": 9.634189861751745e-05, "loss": 0.0704, "step": 16640 }, { "grad_norm": 0.5158842206001282, "learning_rate": 9.633568794033967e-05, "loss": 0.0692, "step": 16650 }, { "grad_norm": 0.4619842767715454, "learning_rate": 9.63294721959607e-05, "loss": 0.0594, "step": 16660 }, { "grad_norm": 0.48537853360176086, "learning_rate": 9.63232513850603e-05, "loss": 0.0849, "step": 16670 }, { "grad_norm": 0.382118821144104, "learning_rate": 9.631702550831878e-05, "loss": 0.0741, "step": 16680 }, { "grad_norm": 0.4831131398677826, "learning_rate": 9.631079456641698e-05, "loss": 0.0622, "step": 16690 }, { "grad_norm": 0.481490820646286, "learning_rate": 9.630455856003632e-05, "loss": 0.0622, "step": 16700 }, { "grad_norm": 0.6214474439620972, "learning_rate": 9.629831748985876e-05, "loss": 0.0681, "step": 16710 }, { "grad_norm": 0.5496878623962402, "learning_rate": 9.629207135656679e-05, "loss": 0.0777, "step": 16720 }, { "grad_norm": 0.543828547000885, "learning_rate": 9.628582016084353e-05, "loss": 0.0661, "step": 16730 }, { "grad_norm": 0.3819250762462616, "learning_rate": 9.627956390337254e-05, "loss": 0.0598, "step": 16740 }, { "grad_norm": 0.38556623458862305, "learning_rate": 9.627330258483802e-05, "loss": 0.076, "step": 16750 }, { "grad_norm": 0.7155166864395142, "learning_rate": 9.62670362059247e-05, "loss": 0.0571, "step": 16760 }, { "grad_norm": 0.43798086047172546, "learning_rate": 9.626076476731786e-05, "loss": 0.0599, "step": 16770 }, { "grad_norm": 0.683295726776123, "learning_rate": 9.625448826970336e-05, "loss": 0.0676, "step": 16780 }, { "grad_norm": 0.5893052816390991, "learning_rate": 9.624820671376755e-05, "loss": 0.06, "step": 16790 }, { "grad_norm": 0.5173915028572083, "learning_rate": 9.62419201001974e-05, "loss": 0.0682, "step": 16800 }, { "grad_norm": 0.4592340290546417, "learning_rate": 9.623562842968037e-05, "loss": 0.065, "step": 16810 }, { "grad_norm": 0.3901031017303467, "learning_rate": 9.622933170290454e-05, "loss": 0.0574, "step": 16820 }, { "grad_norm": 0.7422803640365601, "learning_rate": 9.622302992055849e-05, "loss": 0.0605, "step": 16830 }, { "grad_norm": 0.44039297103881836, "learning_rate": 9.62167230833314e-05, "loss": 0.0668, "step": 16840 }, { "grad_norm": 0.44890615344047546, "learning_rate": 9.621041119191295e-05, "loss": 0.0584, "step": 16850 }, { "grad_norm": 0.557006299495697, "learning_rate": 9.620409424699342e-05, "loss": 0.0697, "step": 16860 }, { "grad_norm": 0.5106359720230103, "learning_rate": 9.619777224926359e-05, "loss": 0.0623, "step": 16870 }, { "grad_norm": 0.5804237723350525, "learning_rate": 9.619144519941485e-05, "loss": 0.0636, "step": 16880 }, { "grad_norm": 0.5938372015953064, "learning_rate": 9.618511309813912e-05, "loss": 0.0692, "step": 16890 }, { "grad_norm": 0.4251546263694763, "learning_rate": 9.617877594612886e-05, "loss": 0.0659, "step": 16900 }, { "grad_norm": 0.5476889610290527, "learning_rate": 9.617243374407707e-05, "loss": 0.0644, "step": 16910 }, { "grad_norm": 0.7825525403022766, "learning_rate": 9.616608649267736e-05, "loss": 0.0768, "step": 16920 }, { "grad_norm": 0.8246567845344543, "learning_rate": 9.615973419262385e-05, "loss": 0.0779, "step": 16930 }, { "grad_norm": 0.4320276379585266, "learning_rate": 9.615337684461119e-05, "loss": 0.0784, "step": 16940 }, { "grad_norm": 0.5507753491401672, "learning_rate": 9.614701444933465e-05, "loss": 0.0888, "step": 16950 }, { "grad_norm": 0.4703153669834137, "learning_rate": 9.614064700748997e-05, "loss": 0.0687, "step": 16960 }, { "grad_norm": 0.6093554496765137, "learning_rate": 9.613427451977352e-05, "loss": 0.0777, "step": 16970 }, { "grad_norm": 0.44436097145080566, "learning_rate": 9.612789698688216e-05, "loss": 0.0717, "step": 16980 }, { "grad_norm": 0.6333308815956116, "learning_rate": 9.612151440951334e-05, "loss": 0.0787, "step": 16990 }, { "grad_norm": 0.5419270992279053, "learning_rate": 9.611512678836506e-05, "loss": 0.0659, "step": 17000 }, { "grad_norm": 0.41292303800582886, "learning_rate": 9.610873412413584e-05, "loss": 0.0675, "step": 17010 }, { "grad_norm": 0.4913501441478729, "learning_rate": 9.610233641752476e-05, "loss": 0.0799, "step": 17020 }, { "grad_norm": 0.5770123600959778, "learning_rate": 9.609593366923151e-05, "loss": 0.0703, "step": 17030 }, { "grad_norm": 0.6127603650093079, "learning_rate": 9.608952587995625e-05, "loss": 0.061, "step": 17040 }, { "grad_norm": 0.4895453453063965, "learning_rate": 9.608311305039972e-05, "loss": 0.0574, "step": 17050 }, { "grad_norm": 0.6770351529121399, "learning_rate": 9.607669518126326e-05, "loss": 0.0778, "step": 17060 }, { "grad_norm": 0.5304007530212402, "learning_rate": 9.607027227324866e-05, "loss": 0.0617, "step": 17070 }, { "grad_norm": 0.6564772129058838, "learning_rate": 9.606384432705837e-05, "loss": 0.062, "step": 17080 }, { "grad_norm": 0.5713914632797241, "learning_rate": 9.60574113433953e-05, "loss": 0.0658, "step": 17090 }, { "grad_norm": 0.4178912937641144, "learning_rate": 9.6050973322963e-05, "loss": 0.0704, "step": 17100 }, { "grad_norm": 0.5216736197471619, "learning_rate": 9.604453026646547e-05, "loss": 0.0695, "step": 17110 }, { "grad_norm": 0.5257477164268494, "learning_rate": 9.603808217460735e-05, "loss": 0.0638, "step": 17120 }, { "grad_norm": 0.5681430101394653, "learning_rate": 9.603162904809377e-05, "loss": 0.0658, "step": 17130 }, { "grad_norm": 0.5733679533004761, "learning_rate": 9.602517088763045e-05, "loss": 0.0524, "step": 17140 }, { "grad_norm": 0.6261345744132996, "learning_rate": 9.601870769392365e-05, "loss": 0.0652, "step": 17150 }, { "grad_norm": 0.4288036823272705, "learning_rate": 9.601223946768017e-05, "loss": 0.0576, "step": 17160 }, { "grad_norm": 0.6139827966690063, "learning_rate": 9.600576620960734e-05, "loss": 0.0622, "step": 17170 }, { "grad_norm": 0.5649517774581909, "learning_rate": 9.599928792041308e-05, "loss": 0.0555, "step": 17180 }, { "grad_norm": 0.4458653926849365, "learning_rate": 9.599280460080587e-05, "loss": 0.0493, "step": 17190 }, { "grad_norm": 0.6101009249687195, "learning_rate": 9.59863162514947e-05, "loss": 0.0708, "step": 17200 }, { "grad_norm": 0.4895186424255371, "learning_rate": 9.597982287318911e-05, "loss": 0.0585, "step": 17210 }, { "grad_norm": 0.5775108337402344, "learning_rate": 9.597332446659923e-05, "loss": 0.0557, "step": 17220 }, { "grad_norm": 0.5076477527618408, "learning_rate": 9.59668210324357e-05, "loss": 0.0594, "step": 17230 }, { "grad_norm": 0.5221977829933167, "learning_rate": 9.596031257140974e-05, "loss": 0.0599, "step": 17240 }, { "grad_norm": 0.5110391974449158, "learning_rate": 9.59537990842331e-05, "loss": 0.0789, "step": 17250 }, { "grad_norm": 0.44729989767074585, "learning_rate": 9.594728057161806e-05, "loss": 0.0645, "step": 17260 }, { "grad_norm": 0.6780222654342651, "learning_rate": 9.594075703427752e-05, "loss": 0.057, "step": 17270 }, { "grad_norm": 0.5152426958084106, "learning_rate": 9.593422847292486e-05, "loss": 0.0544, "step": 17280 }, { "grad_norm": 0.39927738904953003, "learning_rate": 9.592769488827402e-05, "loss": 0.0628, "step": 17290 }, { "grad_norm": 0.38461989164352417, "learning_rate": 9.592115628103952e-05, "loss": 0.0563, "step": 17300 }, { "grad_norm": 0.46248090267181396, "learning_rate": 9.591461265193643e-05, "loss": 0.0557, "step": 17310 }, { "grad_norm": 0.4561895728111267, "learning_rate": 9.590806400168032e-05, "loss": 0.0584, "step": 17320 }, { "grad_norm": 0.36912962794303894, "learning_rate": 9.590151033098735e-05, "loss": 0.06, "step": 17330 }, { "grad_norm": 0.5095506310462952, "learning_rate": 9.589495164057423e-05, "loss": 0.0574, "step": 17340 }, { "grad_norm": 0.4816640615463257, "learning_rate": 9.58883879311582e-05, "loss": 0.0596, "step": 17350 }, { "grad_norm": 0.5107567310333252, "learning_rate": 9.588181920345705e-05, "loss": 0.0755, "step": 17360 }, { "grad_norm": 0.5557198524475098, "learning_rate": 9.587524545818913e-05, "loss": 0.0764, "step": 17370 }, { "grad_norm": 0.5479634404182434, "learning_rate": 9.586866669607335e-05, "loss": 0.0584, "step": 17380 }, { "grad_norm": 0.40054449439048767, "learning_rate": 9.586208291782915e-05, "loss": 0.0658, "step": 17390 }, { "grad_norm": 0.6525931358337402, "learning_rate": 9.58554941241765e-05, "loss": 0.0735, "step": 17400 }, { "grad_norm": 0.6402416825294495, "learning_rate": 9.584890031583596e-05, "loss": 0.0571, "step": 17410 }, { "grad_norm": 0.7773352265357971, "learning_rate": 9.584230149352861e-05, "loss": 0.0652, "step": 17420 }, { "grad_norm": 0.6460899710655212, "learning_rate": 9.58356976579761e-05, "loss": 0.0588, "step": 17430 }, { "grad_norm": 0.7602895498275757, "learning_rate": 9.58290888099006e-05, "loss": 0.0689, "step": 17440 }, { "grad_norm": 0.694516658782959, "learning_rate": 9.582247495002486e-05, "loss": 0.0573, "step": 17450 }, { "grad_norm": 0.6157718896865845, "learning_rate": 9.581585607907214e-05, "loss": 0.0626, "step": 17460 }, { "grad_norm": 0.40429285168647766, "learning_rate": 9.580923219776628e-05, "loss": 0.065, "step": 17470 }, { "grad_norm": 0.5559362173080444, "learning_rate": 9.580260330683167e-05, "loss": 0.0655, "step": 17480 }, { "grad_norm": 0.42054519057273865, "learning_rate": 9.579596940699322e-05, "loss": 0.0616, "step": 17490 }, { "grad_norm": 0.5445975065231323, "learning_rate": 9.578933049897643e-05, "loss": 0.0652, "step": 17500 }, { "grad_norm": 0.6396329402923584, "learning_rate": 9.578268658350728e-05, "loss": 0.0693, "step": 17510 }, { "grad_norm": 0.44916555285453796, "learning_rate": 9.577603766131235e-05, "loss": 0.07, "step": 17520 }, { "grad_norm": 0.5494825839996338, "learning_rate": 9.576938373311878e-05, "loss": 0.0687, "step": 17530 }, { "grad_norm": 0.5166257619857788, "learning_rate": 9.576272479965421e-05, "loss": 0.0728, "step": 17540 }, { "grad_norm": 0.5141533017158508, "learning_rate": 9.575606086164687e-05, "loss": 0.0568, "step": 17550 }, { "grad_norm": 0.5987950563430786, "learning_rate": 9.57493919198255e-05, "loss": 0.0789, "step": 17560 }, { "grad_norm": 0.46588411927223206, "learning_rate": 9.57427179749194e-05, "loss": 0.0734, "step": 17570 }, { "grad_norm": 0.43021437525749207, "learning_rate": 9.573603902765846e-05, "loss": 0.0704, "step": 17580 }, { "grad_norm": 0.5506337285041809, "learning_rate": 9.572935507877304e-05, "loss": 0.0729, "step": 17590 }, { "grad_norm": 0.5658084750175476, "learning_rate": 9.57226661289941e-05, "loss": 0.0837, "step": 17600 }, { "grad_norm": 0.6390411257743835, "learning_rate": 9.571597217905315e-05, "loss": 0.0742, "step": 17610 }, { "grad_norm": 0.4069778621196747, "learning_rate": 9.57092732296822e-05, "loss": 0.0605, "step": 17620 }, { "grad_norm": 0.6769593954086304, "learning_rate": 9.570256928161385e-05, "loss": 0.0625, "step": 17630 }, { "grad_norm": 0.4896377623081207, "learning_rate": 9.569586033558126e-05, "loss": 0.0616, "step": 17640 }, { "grad_norm": 0.47911536693573, "learning_rate": 9.568914639231807e-05, "loss": 0.0658, "step": 17650 }, { "grad_norm": 0.605923056602478, "learning_rate": 9.568242745255852e-05, "loss": 0.0767, "step": 17660 }, { "grad_norm": 0.6716811060905457, "learning_rate": 9.567570351703739e-05, "loss": 0.0825, "step": 17670 }, { "grad_norm": 0.7221586108207703, "learning_rate": 9.566897458649001e-05, "loss": 0.0717, "step": 17680 }, { "grad_norm": 0.4474989175796509, "learning_rate": 9.566224066165221e-05, "loss": 0.0806, "step": 17690 }, { "grad_norm": 0.5838717222213745, "learning_rate": 9.565550174326043e-05, "loss": 0.0789, "step": 17700 }, { "grad_norm": 0.5352205634117126, "learning_rate": 9.564875783205162e-05, "loss": 0.0682, "step": 17710 }, { "grad_norm": 0.9849647283554077, "learning_rate": 9.564200892876328e-05, "loss": 0.0713, "step": 17720 }, { "grad_norm": 0.797589123249054, "learning_rate": 9.563525503413348e-05, "loss": 0.0702, "step": 17730 }, { "grad_norm": 0.6280205249786377, "learning_rate": 9.562849614890079e-05, "loss": 0.0752, "step": 17740 }, { "grad_norm": 0.5618021488189697, "learning_rate": 9.562173227380436e-05, "loss": 0.089, "step": 17750 }, { "grad_norm": 0.49406906962394714, "learning_rate": 9.561496340958389e-05, "loss": 0.0754, "step": 17760 }, { "grad_norm": 0.4547071158885956, "learning_rate": 9.560818955697959e-05, "loss": 0.069, "step": 17770 }, { "grad_norm": 0.5997897982597351, "learning_rate": 9.560141071673228e-05, "loss": 0.0833, "step": 17780 }, { "grad_norm": 0.5008649826049805, "learning_rate": 9.559462688958323e-05, "loss": 0.0536, "step": 17790 }, { "grad_norm": 1.2894517183303833, "learning_rate": 9.558783807627434e-05, "loss": 0.0633, "step": 17800 }, { "grad_norm": 0.5692579746246338, "learning_rate": 9.558104427754801e-05, "loss": 0.075, "step": 17810 }, { "grad_norm": 0.7757157683372498, "learning_rate": 9.557424549414722e-05, "loss": 0.0753, "step": 17820 }, { "grad_norm": 0.46664249897003174, "learning_rate": 9.556744172681546e-05, "loss": 0.0673, "step": 17830 }, { "grad_norm": 0.6030845642089844, "learning_rate": 9.556063297629677e-05, "loss": 0.073, "step": 17840 }, { "grad_norm": 0.6072911024093628, "learning_rate": 9.555381924333578e-05, "loss": 0.0631, "step": 17850 }, { "grad_norm": 0.5432445406913757, "learning_rate": 9.554700052867758e-05, "loss": 0.068, "step": 17860 }, { "grad_norm": 0.559074342250824, "learning_rate": 9.554017683306789e-05, "loss": 0.0618, "step": 17870 }, { "grad_norm": 0.5144839882850647, "learning_rate": 9.553334815725294e-05, "loss": 0.0541, "step": 17880 }, { "grad_norm": 0.5923844575881958, "learning_rate": 9.552651450197949e-05, "loss": 0.0709, "step": 17890 }, { "grad_norm": 0.522574245929718, "learning_rate": 9.551967586799486e-05, "loss": 0.061, "step": 17900 }, { "grad_norm": 0.5789140462875366, "learning_rate": 9.551283225604692e-05, "loss": 0.0711, "step": 17910 }, { "grad_norm": 0.45910122990608215, "learning_rate": 9.550598366688406e-05, "loss": 0.0709, "step": 17920 }, { "grad_norm": 0.6221543550491333, "learning_rate": 9.549913010125526e-05, "loss": 0.0605, "step": 17930 }, { "grad_norm": 0.5521449446678162, "learning_rate": 9.549227155990999e-05, "loss": 0.0711, "step": 17940 }, { "grad_norm": 0.5843616127967834, "learning_rate": 9.548540804359828e-05, "loss": 0.0928, "step": 17950 }, { "grad_norm": 0.5819032192230225, "learning_rate": 9.547853955307077e-05, "loss": 0.0684, "step": 17960 }, { "grad_norm": 0.35236144065856934, "learning_rate": 9.547166608907853e-05, "loss": 0.0561, "step": 17970 }, { "grad_norm": 0.35964256525039673, "learning_rate": 9.546478765237326e-05, "loss": 0.0733, "step": 17980 }, { "grad_norm": 0.6246028542518616, "learning_rate": 9.545790424370715e-05, "loss": 0.0671, "step": 17990 }, { "grad_norm": 0.36593130230903625, "learning_rate": 9.5451015863833e-05, "loss": 0.0623, "step": 18000 }, { "grad_norm": 0.42368054389953613, "learning_rate": 9.544412251350408e-05, "loss": 0.0762, "step": 18010 }, { "grad_norm": 0.585513710975647, "learning_rate": 9.543722419347422e-05, "loss": 0.0697, "step": 18020 }, { "grad_norm": 0.8051398992538452, "learning_rate": 9.543032090449788e-05, "loss": 0.0688, "step": 18030 }, { "grad_norm": 0.36918023228645325, "learning_rate": 9.542341264732992e-05, "loss": 0.071, "step": 18040 }, { "grad_norm": 0.5870610475540161, "learning_rate": 9.541649942272585e-05, "loss": 0.0712, "step": 18050 }, { "grad_norm": 0.5167108774185181, "learning_rate": 9.54095812314417e-05, "loss": 0.0661, "step": 18060 }, { "grad_norm": 0.8431577086448669, "learning_rate": 9.540265807423401e-05, "loss": 0.0603, "step": 18070 }, { "grad_norm": 0.4089096784591675, "learning_rate": 9.53957299518599e-05, "loss": 0.0596, "step": 18080 }, { "grad_norm": 0.8624567985534668, "learning_rate": 9.5388796865077e-05, "loss": 0.0672, "step": 18090 }, { "grad_norm": 0.4953261911869049, "learning_rate": 9.538185881464353e-05, "loss": 0.0702, "step": 18100 }, { "grad_norm": 0.4973703622817993, "learning_rate": 9.537491580131821e-05, "loss": 0.0748, "step": 18110 }, { "grad_norm": 0.5584858655929565, "learning_rate": 9.53679678258603e-05, "loss": 0.0684, "step": 18120 }, { "grad_norm": 0.7686853408813477, "learning_rate": 9.536101488902966e-05, "loss": 0.0755, "step": 18130 }, { "grad_norm": 0.5510821342468262, "learning_rate": 9.535405699158663e-05, "loss": 0.0499, "step": 18140 }, { "grad_norm": 0.41433537006378174, "learning_rate": 9.53470941342921e-05, "loss": 0.0531, "step": 18150 }, { "grad_norm": 0.8419488668441772, "learning_rate": 9.534012631790756e-05, "loss": 0.0667, "step": 18160 }, { "grad_norm": 0.5240404605865479, "learning_rate": 9.533315354319494e-05, "loss": 0.0588, "step": 18170 }, { "grad_norm": 0.4898126721382141, "learning_rate": 9.532617581091682e-05, "loss": 0.0549, "step": 18180 }, { "grad_norm": 0.560073971748352, "learning_rate": 9.531919312183629e-05, "loss": 0.0726, "step": 18190 }, { "grad_norm": 0.38751617074012756, "learning_rate": 9.531220547671688e-05, "loss": 0.0479, "step": 18200 }, { "grad_norm": 0.4646923243999481, "learning_rate": 9.530521287632285e-05, "loss": 0.074, "step": 18210 }, { "grad_norm": 0.5490145683288574, "learning_rate": 9.529821532141884e-05, "loss": 0.0515, "step": 18220 }, { "grad_norm": 0.45959049463272095, "learning_rate": 9.52912128127701e-05, "loss": 0.0619, "step": 18230 }, { "grad_norm": 1.94789457321167, "learning_rate": 9.528420535114244e-05, "loss": 0.0524, "step": 18240 }, { "grad_norm": 0.6063447594642639, "learning_rate": 9.527719293730215e-05, "loss": 0.0743, "step": 18250 }, { "grad_norm": 0.37814682722091675, "learning_rate": 9.527017557201611e-05, "loss": 0.0552, "step": 18260 }, { "grad_norm": 0.5214531421661377, "learning_rate": 9.526315325605176e-05, "loss": 0.0715, "step": 18270 }, { "grad_norm": 1.5684208869934082, "learning_rate": 9.525612599017699e-05, "loss": 0.0582, "step": 18280 }, { "grad_norm": 0.5079652070999146, "learning_rate": 9.524909377516033e-05, "loss": 0.0752, "step": 18290 }, { "grad_norm": 0.4383428394794464, "learning_rate": 9.524205661177081e-05, "loss": 0.0496, "step": 18300 }, { "grad_norm": 0.44954872131347656, "learning_rate": 9.523501450077801e-05, "loss": 0.061, "step": 18310 }, { "grad_norm": 0.4486777186393738, "learning_rate": 9.522796744295202e-05, "loss": 0.0604, "step": 18320 }, { "grad_norm": 0.39192888140678406, "learning_rate": 9.522091543906352e-05, "loss": 0.0566, "step": 18330 }, { "grad_norm": 0.6302684545516968, "learning_rate": 9.521385848988369e-05, "loss": 0.0678, "step": 18340 }, { "grad_norm": 0.5196202993392944, "learning_rate": 9.520679659618428e-05, "loss": 0.0632, "step": 18350 }, { "grad_norm": 0.42934033274650574, "learning_rate": 9.519972975873754e-05, "loss": 0.0717, "step": 18360 }, { "grad_norm": 0.6321066617965698, "learning_rate": 9.519265797831633e-05, "loss": 0.0702, "step": 18370 }, { "grad_norm": 0.5139192938804626, "learning_rate": 9.518558125569399e-05, "loss": 0.0714, "step": 18380 }, { "grad_norm": 0.3787201941013336, "learning_rate": 9.517849959164442e-05, "loss": 0.0504, "step": 18390 }, { "grad_norm": 1.8154231309890747, "learning_rate": 9.517141298694205e-05, "loss": 0.0719, "step": 18400 }, { "grad_norm": 0.8535369634628296, "learning_rate": 9.516432144236188e-05, "loss": 0.0754, "step": 18410 }, { "grad_norm": 0.8103001117706299, "learning_rate": 9.515722495867941e-05, "loss": 0.0684, "step": 18420 }, { "grad_norm": 0.34972041845321655, "learning_rate": 9.515012353667072e-05, "loss": 0.0646, "step": 18430 }, { "grad_norm": 0.4287622272968292, "learning_rate": 9.51430171771124e-05, "loss": 0.0604, "step": 18440 }, { "grad_norm": 0.5768088102340698, "learning_rate": 9.513590588078159e-05, "loss": 0.0736, "step": 18450 }, { "grad_norm": 0.569801926612854, "learning_rate": 9.512878964845597e-05, "loss": 0.0645, "step": 18460 }, { "grad_norm": 0.6571255922317505, "learning_rate": 9.512166848091377e-05, "loss": 0.0761, "step": 18470 }, { "grad_norm": 0.6570623517036438, "learning_rate": 9.511454237893376e-05, "loss": 0.0735, "step": 18480 }, { "grad_norm": 0.4749930500984192, "learning_rate": 9.51074113432952e-05, "loss": 0.062, "step": 18490 }, { "grad_norm": 0.851772129535675, "learning_rate": 9.510027537477797e-05, "loss": 0.0873, "step": 18500 }, { "grad_norm": 0.6061285138130188, "learning_rate": 9.509313447416242e-05, "loss": 0.0784, "step": 18510 }, { "grad_norm": 0.6180872321128845, "learning_rate": 9.508598864222949e-05, "loss": 0.0646, "step": 18520 }, { "grad_norm": 0.5485802888870239, "learning_rate": 9.507883787976062e-05, "loss": 0.0773, "step": 18530 }, { "grad_norm": 0.6800968647003174, "learning_rate": 9.507168218753781e-05, "loss": 0.0594, "step": 18540 }, { "grad_norm": 0.40281903743743896, "learning_rate": 9.506452156634362e-05, "loss": 0.071, "step": 18550 }, { "grad_norm": 0.49766671657562256, "learning_rate": 9.505735601696109e-05, "loss": 0.0867, "step": 18560 }, { "grad_norm": 0.567372739315033, "learning_rate": 9.505018554017385e-05, "loss": 0.0705, "step": 18570 }, { "grad_norm": 0.5021393299102783, "learning_rate": 9.504301013676604e-05, "loss": 0.066, "step": 18580 }, { "grad_norm": 0.45382338762283325, "learning_rate": 9.503582980752238e-05, "loss": 0.063, "step": 18590 }, { "grad_norm": 0.623603880405426, "learning_rate": 9.502864455322809e-05, "loss": 0.0817, "step": 18600 }, { "grad_norm": 0.8105921149253845, "learning_rate": 9.502145437466891e-05, "loss": 0.0629, "step": 18610 }, { "grad_norm": 0.3803757131099701, "learning_rate": 9.501425927263116e-05, "loss": 0.0637, "step": 18620 }, { "grad_norm": 0.6902218461036682, "learning_rate": 9.500705924790172e-05, "loss": 0.0814, "step": 18630 }, { "grad_norm": 0.45382168889045715, "learning_rate": 9.499985430126794e-05, "loss": 0.0698, "step": 18640 }, { "grad_norm": 0.6883152723312378, "learning_rate": 9.499264443351775e-05, "loss": 0.0778, "step": 18650 }, { "grad_norm": 0.4301798641681671, "learning_rate": 9.498542964543961e-05, "loss": 0.0673, "step": 18660 }, { "grad_norm": 0.4441806972026825, "learning_rate": 9.497820993782252e-05, "loss": 0.0662, "step": 18670 }, { "grad_norm": 0.5713391900062561, "learning_rate": 9.497098531145601e-05, "loss": 0.0611, "step": 18680 }, { "grad_norm": 0.6695278882980347, "learning_rate": 9.496375576713017e-05, "loss": 0.0647, "step": 18690 }, { "grad_norm": 0.4610508680343628, "learning_rate": 9.49565213056356e-05, "loss": 0.0502, "step": 18700 }, { "grad_norm": 0.5010209083557129, "learning_rate": 9.494928192776342e-05, "loss": 0.061, "step": 18710 }, { "grad_norm": 0.6972229480743408, "learning_rate": 9.494203763430538e-05, "loss": 0.0555, "step": 18720 }, { "grad_norm": 1.3238043785095215, "learning_rate": 9.493478842605366e-05, "loss": 0.091, "step": 18730 }, { "grad_norm": 0.4631696045398712, "learning_rate": 9.492753430380105e-05, "loss": 0.0611, "step": 18740 }, { "grad_norm": 0.4670812785625458, "learning_rate": 9.492027526834083e-05, "loss": 0.0646, "step": 18750 }, { "grad_norm": 0.5645702481269836, "learning_rate": 9.491301132046684e-05, "loss": 0.0571, "step": 18760 }, { "grad_norm": 0.47606295347213745, "learning_rate": 9.490574246097345e-05, "loss": 0.0849, "step": 18770 }, { "grad_norm": 0.7798668146133423, "learning_rate": 9.48984686906556e-05, "loss": 0.0802, "step": 18780 }, { "grad_norm": 0.8538448214530945, "learning_rate": 9.489119001030871e-05, "loss": 0.071, "step": 18790 }, { "grad_norm": 0.722083330154419, "learning_rate": 9.488390642072878e-05, "loss": 0.0661, "step": 18800 }, { "grad_norm": 0.47468382120132446, "learning_rate": 9.48766179227123e-05, "loss": 0.0662, "step": 18810 }, { "grad_norm": 1.0379589796066284, "learning_rate": 9.486932451705636e-05, "loss": 0.0691, "step": 18820 }, { "grad_norm": 0.7171428799629211, "learning_rate": 9.486202620455857e-05, "loss": 0.0691, "step": 18830 }, { "grad_norm": 0.5450888872146606, "learning_rate": 9.485472298601704e-05, "loss": 0.0576, "step": 18840 }, { "grad_norm": 0.41880175471305847, "learning_rate": 9.484741486223043e-05, "loss": 0.0638, "step": 18850 }, { "grad_norm": 0.42185866832733154, "learning_rate": 9.484010183399797e-05, "loss": 0.0685, "step": 18860 }, { "grad_norm": 0.6265957355499268, "learning_rate": 9.483278390211938e-05, "loss": 0.0633, "step": 18870 }, { "grad_norm": 0.5321903824806213, "learning_rate": 9.482546106739496e-05, "loss": 0.0576, "step": 18880 }, { "grad_norm": 0.481489360332489, "learning_rate": 9.48181333306255e-05, "loss": 0.0619, "step": 18890 }, { "grad_norm": 0.47237420082092285, "learning_rate": 9.481080069261237e-05, "loss": 0.0608, "step": 18900 }, { "grad_norm": 0.37268638610839844, "learning_rate": 9.480346315415745e-05, "loss": 0.0505, "step": 18910 }, { "grad_norm": 0.6458445191383362, "learning_rate": 9.479612071606314e-05, "loss": 0.0499, "step": 18920 }, { "grad_norm": 0.6995109915733337, "learning_rate": 9.478877337913244e-05, "loss": 0.0664, "step": 18930 }, { "grad_norm": 0.48693686723709106, "learning_rate": 9.478142114416881e-05, "loss": 0.0546, "step": 18940 }, { "grad_norm": 0.5988968014717102, "learning_rate": 9.47740640119763e-05, "loss": 0.0555, "step": 18950 }, { "grad_norm": 0.4973340630531311, "learning_rate": 9.476670198335947e-05, "loss": 0.0607, "step": 18960 }, { "grad_norm": 0.5158348083496094, "learning_rate": 9.47593350591234e-05, "loss": 0.0636, "step": 18970 }, { "grad_norm": 0.4168627858161926, "learning_rate": 9.475196324007376e-05, "loss": 0.0551, "step": 18980 }, { "grad_norm": 0.6336456537246704, "learning_rate": 9.474458652701669e-05, "loss": 0.0612, "step": 18990 }, { "grad_norm": 0.4918913245201111, "learning_rate": 9.473720492075892e-05, "loss": 0.0651, "step": 19000 }, { "grad_norm": 1.0353407859802246, "learning_rate": 9.472981842210768e-05, "loss": 0.0644, "step": 19010 }, { "grad_norm": 0.6236046552658081, "learning_rate": 9.472242703187074e-05, "loss": 0.0632, "step": 19020 }, { "grad_norm": 0.45927852392196655, "learning_rate": 9.471503075085643e-05, "loss": 0.0586, "step": 19030 }, { "grad_norm": 1.2072765827178955, "learning_rate": 9.470762957987359e-05, "loss": 0.0764, "step": 19040 }, { "grad_norm": 0.35165026783943176, "learning_rate": 9.470022351973158e-05, "loss": 0.0659, "step": 19050 }, { "grad_norm": 0.5542214512825012, "learning_rate": 9.469281257124034e-05, "loss": 0.0762, "step": 19060 }, { "grad_norm": 0.4336659908294678, "learning_rate": 9.46853967352103e-05, "loss": 0.0752, "step": 19070 }, { "grad_norm": 0.5987039804458618, "learning_rate": 9.467797601245246e-05, "loss": 0.0584, "step": 19080 }, { "grad_norm": 0.5351722836494446, "learning_rate": 9.467055040377834e-05, "loss": 0.076, "step": 19090 }, { "grad_norm": 0.38676056265830994, "learning_rate": 9.466311990999999e-05, "loss": 0.0652, "step": 19100 }, { "grad_norm": 0.5724105834960938, "learning_rate": 9.465568453193e-05, "loss": 0.0659, "step": 19110 }, { "grad_norm": 0.5133010149002075, "learning_rate": 9.464824427038148e-05, "loss": 0.0652, "step": 19120 }, { "grad_norm": 0.6758784055709839, "learning_rate": 9.46407991261681e-05, "loss": 0.0748, "step": 19130 }, { "grad_norm": 0.4267110526561737, "learning_rate": 9.463334910010404e-05, "loss": 0.0649, "step": 19140 }, { "grad_norm": 0.43283650279045105, "learning_rate": 9.462589419300403e-05, "loss": 0.0533, "step": 19150 }, { "grad_norm": 0.5286108255386353, "learning_rate": 9.461843440568333e-05, "loss": 0.064, "step": 19160 }, { "grad_norm": 0.48597919940948486, "learning_rate": 9.461096973895773e-05, "loss": 0.0563, "step": 19170 }, { "grad_norm": 0.5338952541351318, "learning_rate": 9.460350019364355e-05, "loss": 0.0787, "step": 19180 }, { "grad_norm": 0.47780683636665344, "learning_rate": 9.459602577055764e-05, "loss": 0.0693, "step": 19190 }, { "grad_norm": 0.4437192380428314, "learning_rate": 9.45885464705174e-05, "loss": 0.0564, "step": 19200 }, { "grad_norm": 0.5930082201957703, "learning_rate": 9.458106229434076e-05, "loss": 0.0574, "step": 19210 }, { "grad_norm": 0.5574308633804321, "learning_rate": 9.457357324284617e-05, "loss": 0.0639, "step": 19220 }, { "grad_norm": 0.468139111995697, "learning_rate": 9.456607931685262e-05, "loss": 0.0615, "step": 19230 }, { "grad_norm": 0.3920302987098694, "learning_rate": 9.455858051717965e-05, "loss": 0.0558, "step": 19240 }, { "grad_norm": 0.4164775609970093, "learning_rate": 9.45510768446473e-05, "loss": 0.061, "step": 19250 }, { "grad_norm": 0.4291519224643707, "learning_rate": 9.454356830007618e-05, "loss": 0.0527, "step": 19260 }, { "grad_norm": 0.6313633918762207, "learning_rate": 9.45360548842874e-05, "loss": 0.0494, "step": 19270 }, { "grad_norm": 0.5041594505310059, "learning_rate": 9.452853659810261e-05, "loss": 0.0591, "step": 19280 }, { "grad_norm": 0.6175519824028015, "learning_rate": 9.452101344234401e-05, "loss": 0.0869, "step": 19290 }, { "grad_norm": 0.5526174306869507, "learning_rate": 9.451348541783431e-05, "loss": 0.0691, "step": 19300 }, { "grad_norm": 0.7405081987380981, "learning_rate": 9.450595252539678e-05, "loss": 0.0591, "step": 19310 }, { "grad_norm": 0.5868982672691345, "learning_rate": 9.449841476585518e-05, "loss": 0.0874, "step": 19320 }, { "grad_norm": 0.6882375478744507, "learning_rate": 9.449087214003384e-05, "loss": 0.0698, "step": 19330 }, { "grad_norm": 0.5632055401802063, "learning_rate": 9.448332464875765e-05, "loss": 0.0822, "step": 19340 }, { "grad_norm": 0.6098043918609619, "learning_rate": 9.447577229285192e-05, "loss": 0.0773, "step": 19350 }, { "grad_norm": 0.5745555758476257, "learning_rate": 9.446821507314261e-05, "loss": 0.0668, "step": 19360 }, { "grad_norm": 1.6641944646835327, "learning_rate": 9.446065299045617e-05, "loss": 0.0604, "step": 19370 }, { "grad_norm": 0.5729604959487915, "learning_rate": 9.445308604561955e-05, "loss": 0.0761, "step": 19380 }, { "grad_norm": 0.47447842359542847, "learning_rate": 9.444551423946028e-05, "loss": 0.0562, "step": 19390 }, { "grad_norm": 0.5390896201133728, "learning_rate": 9.443793757280638e-05, "loss": 0.0574, "step": 19400 }, { "grad_norm": 0.5328306555747986, "learning_rate": 9.443035604648646e-05, "loss": 0.061, "step": 19410 }, { "grad_norm": 0.7002894282341003, "learning_rate": 9.44227696613296e-05, "loss": 0.0675, "step": 19420 }, { "grad_norm": 0.5750247836112976, "learning_rate": 9.441517841816542e-05, "loss": 0.0823, "step": 19430 }, { "grad_norm": 0.42535489797592163, "learning_rate": 9.440758231782413e-05, "loss": 0.0682, "step": 19440 }, { "grad_norm": 0.6101638078689575, "learning_rate": 9.439998136113639e-05, "loss": 0.0705, "step": 19450 }, { "grad_norm": 0.4791085422039032, "learning_rate": 9.439237554893344e-05, "loss": 0.0587, "step": 19460 }, { "grad_norm": 0.4053283929824829, "learning_rate": 9.438476488204705e-05, "loss": 0.0646, "step": 19470 }, { "grad_norm": 0.43341585993766785, "learning_rate": 9.43771493613095e-05, "loss": 0.0561, "step": 19480 }, { "grad_norm": 0.5018873810768127, "learning_rate": 9.436952898755362e-05, "loss": 0.0786, "step": 19490 }, { "grad_norm": 0.4934431314468384, "learning_rate": 9.436190376161276e-05, "loss": 0.0559, "step": 19500 }, { "grad_norm": 0.6016120910644531, "learning_rate": 9.43542736843208e-05, "loss": 0.0802, "step": 19510 }, { "grad_norm": 0.38679468631744385, "learning_rate": 9.434663875651216e-05, "loss": 0.0497, "step": 19520 }, { "grad_norm": 0.3753456771373749, "learning_rate": 9.433899897902177e-05, "loss": 0.0571, "step": 19530 }, { "grad_norm": 0.46986204385757446, "learning_rate": 9.433135435268511e-05, "loss": 0.0682, "step": 19540 }, { "grad_norm": 0.5077071785926819, "learning_rate": 9.432370487833819e-05, "loss": 0.0609, "step": 19550 }, { "grad_norm": 0.6483305096626282, "learning_rate": 9.431605055681756e-05, "loss": 0.0771, "step": 19560 }, { "grad_norm": 0.48875170946121216, "learning_rate": 9.430839138896026e-05, "loss": 0.0635, "step": 19570 }, { "grad_norm": 0.7464920878410339, "learning_rate": 9.43007273756039e-05, "loss": 0.0812, "step": 19580 }, { "grad_norm": 0.48392805457115173, "learning_rate": 9.429305851758658e-05, "loss": 0.0607, "step": 19590 }, { "grad_norm": 0.5312497615814209, "learning_rate": 9.428538481574699e-05, "loss": 0.0726, "step": 19600 }, { "grad_norm": 0.6853419542312622, "learning_rate": 9.42777062709243e-05, "loss": 0.059, "step": 19610 }, { "grad_norm": 0.4680754244327545, "learning_rate": 9.427002288395821e-05, "loss": 0.0575, "step": 19620 }, { "grad_norm": 0.5809340476989746, "learning_rate": 9.426233465568898e-05, "loss": 0.0587, "step": 19630 }, { "grad_norm": 0.3576730489730835, "learning_rate": 9.42546415869574e-05, "loss": 0.0501, "step": 19640 }, { "grad_norm": 0.43489184975624084, "learning_rate": 9.424694367860473e-05, "loss": 0.057, "step": 19650 }, { "grad_norm": 0.46944376826286316, "learning_rate": 9.423924093147284e-05, "loss": 0.0621, "step": 19660 }, { "grad_norm": 0.4818020164966583, "learning_rate": 9.423153334640407e-05, "loss": 0.0658, "step": 19670 }, { "grad_norm": 0.6134549975395203, "learning_rate": 9.42238209242413e-05, "loss": 0.0564, "step": 19680 }, { "grad_norm": 0.5947306752204895, "learning_rate": 9.421610366582798e-05, "loss": 0.0713, "step": 19690 }, { "grad_norm": 0.43696507811546326, "learning_rate": 9.420838157200803e-05, "loss": 0.0699, "step": 19700 }, { "grad_norm": 0.41524410247802734, "learning_rate": 9.420065464362594e-05, "loss": 0.0707, "step": 19710 }, { "grad_norm": 0.496994286775589, "learning_rate": 9.419292288152673e-05, "loss": 0.0633, "step": 19720 }, { "grad_norm": 0.5383461713790894, "learning_rate": 9.418518628655588e-05, "loss": 0.0559, "step": 19730 }, { "grad_norm": 0.5151389837265015, "learning_rate": 9.417744485955951e-05, "loss": 0.0728, "step": 19740 }, { "grad_norm": 0.6752846240997314, "learning_rate": 9.41696986013842e-05, "loss": 0.07, "step": 19750 }, { "grad_norm": 0.5710800886154175, "learning_rate": 9.416194751287705e-05, "loss": 0.0797, "step": 19760 }, { "grad_norm": 0.5007146596908569, "learning_rate": 9.415419159488572e-05, "loss": 0.0585, "step": 19770 }, { "grad_norm": 0.6285337209701538, "learning_rate": 9.414643084825837e-05, "loss": 0.0707, "step": 19780 }, { "grad_norm": 0.4920511841773987, "learning_rate": 9.413866527384372e-05, "loss": 0.0753, "step": 19790 }, { "grad_norm": 0.4047528803348541, "learning_rate": 9.4130894872491e-05, "loss": 0.0537, "step": 19800 }, { "grad_norm": 0.472974956035614, "learning_rate": 9.412311964504998e-05, "loss": 0.0622, "step": 19810 }, { "grad_norm": 0.6763337850570679, "learning_rate": 9.411533959237091e-05, "loss": 0.0578, "step": 19820 }, { "grad_norm": 0.469371497631073, "learning_rate": 9.410755471530464e-05, "loss": 0.0574, "step": 19830 }, { "grad_norm": 0.508438766002655, "learning_rate": 9.40997650147025e-05, "loss": 0.0584, "step": 19840 }, { "grad_norm": 0.5454630255699158, "learning_rate": 9.409197049141637e-05, "loss": 0.0804, "step": 19850 }, { "grad_norm": 0.5596538782119751, "learning_rate": 9.408417114629863e-05, "loss": 0.0794, "step": 19860 }, { "grad_norm": 0.5243396162986755, "learning_rate": 9.40763669802022e-05, "loss": 0.0639, "step": 19870 }, { "grad_norm": 0.49746665358543396, "learning_rate": 9.406855799398056e-05, "loss": 0.0719, "step": 19880 }, { "grad_norm": 0.7293604016304016, "learning_rate": 9.406074418848767e-05, "loss": 0.0697, "step": 19890 }, { "grad_norm": 0.5737509727478027, "learning_rate": 9.405292556457805e-05, "loss": 0.0817, "step": 19900 }, { "grad_norm": 0.40201181173324585, "learning_rate": 9.404510212310671e-05, "loss": 0.0732, "step": 19910 }, { "grad_norm": 0.5233862400054932, "learning_rate": 9.403727386492924e-05, "loss": 0.0704, "step": 19920 }, { "grad_norm": 0.7154479622840881, "learning_rate": 9.40294407909017e-05, "loss": 0.0754, "step": 19930 }, { "grad_norm": 0.7265825867652893, "learning_rate": 9.40216029018807e-05, "loss": 0.0635, "step": 19940 }, { "grad_norm": 0.6422313451766968, "learning_rate": 9.401376019872338e-05, "loss": 0.0688, "step": 19950 }, { "grad_norm": 0.5171796083450317, "learning_rate": 9.400591268228746e-05, "loss": 0.0657, "step": 19960 }, { "grad_norm": 0.5671125054359436, "learning_rate": 9.399806035343106e-05, "loss": 0.0689, "step": 19970 }, { "grad_norm": 0.6414792537689209, "learning_rate": 9.399020321301294e-05, "loss": 0.0608, "step": 19980 }, { "grad_norm": 0.5278209447860718, "learning_rate": 9.398234126189234e-05, "loss": 0.0609, "step": 19990 }, { "grad_norm": 0.4623742997646332, "learning_rate": 9.397447450092902e-05, "loss": 0.056, "step": 20000 }, { "grad_norm": 0.6488063931465149, "learning_rate": 9.39666029309833e-05, "loss": 0.0576, "step": 20010 }, { "grad_norm": 0.7354999780654907, "learning_rate": 9.395872655291596e-05, "loss": 0.0678, "step": 20020 }, { "grad_norm": 0.46635571122169495, "learning_rate": 9.395084536758838e-05, "loss": 0.0609, "step": 20030 }, { "grad_norm": 0.7404280304908752, "learning_rate": 9.394295937586243e-05, "loss": 0.0717, "step": 20040 }, { "grad_norm": 0.5843643546104431, "learning_rate": 9.393506857860052e-05, "loss": 0.0572, "step": 20050 }, { "grad_norm": 0.34286972880363464, "learning_rate": 9.392717297666555e-05, "loss": 0.0503, "step": 20060 }, { "grad_norm": 0.6104995608329773, "learning_rate": 9.391927257092101e-05, "loss": 0.0616, "step": 20070 }, { "grad_norm": 0.49391135573387146, "learning_rate": 9.391136736223085e-05, "loss": 0.0546, "step": 20080 }, { "grad_norm": 0.42004847526550293, "learning_rate": 9.390345735145956e-05, "loss": 0.0625, "step": 20090 }, { "grad_norm": 0.6029525399208069, "learning_rate": 9.389554253947219e-05, "loss": 0.07, "step": 20100 }, { "grad_norm": 0.5610728859901428, "learning_rate": 9.388762292713428e-05, "loss": 0.0658, "step": 20110 }, { "grad_norm": 0.9196134209632874, "learning_rate": 9.38796985153119e-05, "loss": 0.0687, "step": 20120 }, { "grad_norm": 1.6163160800933838, "learning_rate": 9.387176930487169e-05, "loss": 0.0674, "step": 20130 }, { "grad_norm": 0.5101215243339539, "learning_rate": 9.386383529668072e-05, "loss": 0.0585, "step": 20140 }, { "grad_norm": 0.4198448061943054, "learning_rate": 9.385589649160669e-05, "loss": 0.0631, "step": 20150 }, { "grad_norm": 0.3415832817554474, "learning_rate": 9.384795289051775e-05, "loss": 0.077, "step": 20160 }, { "grad_norm": 0.4574646055698395, "learning_rate": 9.384000449428261e-05, "loss": 0.0644, "step": 20170 }, { "grad_norm": 0.4303615987300873, "learning_rate": 9.383205130377048e-05, "loss": 0.0752, "step": 20180 }, { "grad_norm": 0.39292919635772705, "learning_rate": 9.382409331985114e-05, "loss": 0.0533, "step": 20190 }, { "grad_norm": 0.5253565907478333, "learning_rate": 9.381613054339482e-05, "loss": 0.0687, "step": 20200 }, { "grad_norm": 0.5105947852134705, "learning_rate": 9.380816297527235e-05, "loss": 0.055, "step": 20210 }, { "grad_norm": 0.40764403343200684, "learning_rate": 9.380019061635506e-05, "loss": 0.0521, "step": 20220 }, { "grad_norm": 0.52288818359375, "learning_rate": 9.379221346751474e-05, "loss": 0.0845, "step": 20230 }, { "grad_norm": 0.5151060819625854, "learning_rate": 9.378423152962382e-05, "loss": 0.0555, "step": 20240 }, { "grad_norm": 0.5370897054672241, "learning_rate": 9.377624480355517e-05, "loss": 0.0608, "step": 20250 }, { "grad_norm": 0.534475564956665, "learning_rate": 9.376825329018219e-05, "loss": 0.0576, "step": 20260 }, { "grad_norm": 1.0376819372177124, "learning_rate": 9.376025699037884e-05, "loss": 0.0642, "step": 20270 }, { "grad_norm": 0.41463959217071533, "learning_rate": 9.37522559050196e-05, "loss": 0.0594, "step": 20280 }, { "grad_norm": 0.5529673099517822, "learning_rate": 9.37442500349794e-05, "loss": 0.0719, "step": 20290 }, { "grad_norm": 0.42249277234077454, "learning_rate": 9.373623938113381e-05, "loss": 0.0641, "step": 20300 }, { "grad_norm": 0.5605267286300659, "learning_rate": 9.372822394435883e-05, "loss": 0.0681, "step": 20310 }, { "grad_norm": 0.5199106931686401, "learning_rate": 9.372020372553102e-05, "loss": 0.063, "step": 20320 }, { "grad_norm": 0.5012014508247375, "learning_rate": 9.371217872552746e-05, "loss": 0.0633, "step": 20330 }, { "grad_norm": 0.48287808895111084, "learning_rate": 9.370414894522576e-05, "loss": 0.0636, "step": 20340 }, { "grad_norm": 0.4988301992416382, "learning_rate": 9.369611438550406e-05, "loss": 0.0658, "step": 20350 }, { "grad_norm": 0.4351753890514374, "learning_rate": 9.368807504724095e-05, "loss": 0.066, "step": 20360 }, { "grad_norm": 0.4795835018157959, "learning_rate": 9.368003093131565e-05, "loss": 0.0681, "step": 20370 }, { "grad_norm": 0.48667290806770325, "learning_rate": 9.367198203860785e-05, "loss": 0.0651, "step": 20380 }, { "grad_norm": 0.4577970802783966, "learning_rate": 9.366392836999774e-05, "loss": 0.0574, "step": 20390 }, { "grad_norm": 0.36657479405403137, "learning_rate": 9.365586992636607e-05, "loss": 0.0473, "step": 20400 }, { "grad_norm": 0.42450040578842163, "learning_rate": 9.364780670859412e-05, "loss": 0.0482, "step": 20410 }, { "grad_norm": 0.5717199444770813, "learning_rate": 9.363973871756364e-05, "loss": 0.0611, "step": 20420 }, { "grad_norm": 0.41257432103157043, "learning_rate": 9.363166595415696e-05, "loss": 0.0544, "step": 20430 }, { "grad_norm": 0.5181920528411865, "learning_rate": 9.362358841925686e-05, "loss": 0.0649, "step": 20440 }, { "grad_norm": 0.4053511619567871, "learning_rate": 9.361550611374674e-05, "loss": 0.0664, "step": 20450 }, { "grad_norm": 0.3767508566379547, "learning_rate": 9.360741903851043e-05, "loss": 0.0808, "step": 20460 }, { "grad_norm": 0.624267578125, "learning_rate": 9.359932719443236e-05, "loss": 0.0777, "step": 20470 }, { "grad_norm": 0.6695809960365295, "learning_rate": 9.35912305823974e-05, "loss": 0.0645, "step": 20480 }, { "grad_norm": 0.43433940410614014, "learning_rate": 9.358312920329101e-05, "loss": 0.0474, "step": 20490 }, { "grad_norm": 1.3941922187805176, "learning_rate": 9.357502305799914e-05, "loss": 0.0649, "step": 20500 }, { "grad_norm": 0.6009373664855957, "learning_rate": 9.356691214740824e-05, "loss": 0.0578, "step": 20510 }, { "grad_norm": 0.6202855110168457, "learning_rate": 9.355879647240535e-05, "loss": 0.0683, "step": 20520 }, { "grad_norm": 0.9825270175933838, "learning_rate": 9.355067603387798e-05, "loss": 0.0661, "step": 20530 }, { "grad_norm": 0.496248334646225, "learning_rate": 9.354255083271412e-05, "loss": 0.0535, "step": 20540 }, { "grad_norm": 0.4822152853012085, "learning_rate": 9.353442086980239e-05, "loss": 0.0674, "step": 20550 }, { "grad_norm": 0.6177304983139038, "learning_rate": 9.352628614603185e-05, "loss": 0.0604, "step": 20560 }, { "grad_norm": 0.5158217549324036, "learning_rate": 9.351814666229209e-05, "loss": 0.0569, "step": 20570 }, { "grad_norm": 0.5100889801979065, "learning_rate": 9.351000241947324e-05, "loss": 0.0703, "step": 20580 }, { "grad_norm": 0.4487951099872589, "learning_rate": 9.350185341846594e-05, "loss": 0.0652, "step": 20590 }, { "grad_norm": 0.7312614321708679, "learning_rate": 9.349369966016134e-05, "loss": 0.0657, "step": 20600 }, { "grad_norm": 0.5272960066795349, "learning_rate": 9.348554114545117e-05, "loss": 0.0754, "step": 20610 }, { "grad_norm": 1.309855341911316, "learning_rate": 9.347737787522758e-05, "loss": 0.0565, "step": 20620 }, { "grad_norm": 0.5698373913764954, "learning_rate": 9.346920985038332e-05, "loss": 0.0724, "step": 20630 }, { "grad_norm": 0.6450580358505249, "learning_rate": 9.346103707181162e-05, "loss": 0.0698, "step": 20640 }, { "grad_norm": 0.5560049414634705, "learning_rate": 9.345285954040626e-05, "loss": 0.0681, "step": 20650 }, { "grad_norm": 0.9217497110366821, "learning_rate": 9.34446772570615e-05, "loss": 0.0711, "step": 20660 }, { "grad_norm": 0.494261771440506, "learning_rate": 9.343649022267214e-05, "loss": 0.066, "step": 20670 }, { "grad_norm": 0.5657370686531067, "learning_rate": 9.342829843813353e-05, "loss": 0.063, "step": 20680 }, { "grad_norm": 0.65087890625, "learning_rate": 9.342010190434149e-05, "loss": 0.0642, "step": 20690 }, { "grad_norm": 0.5966586470603943, "learning_rate": 9.34119006221924e-05, "loss": 0.0684, "step": 20700 }, { "grad_norm": 0.4139447808265686, "learning_rate": 9.340369459258313e-05, "loss": 0.0555, "step": 20710 }, { "grad_norm": 0.39797350764274597, "learning_rate": 9.339548381641106e-05, "loss": 0.0621, "step": 20720 }, { "grad_norm": 0.6436700820922852, "learning_rate": 9.338726829457413e-05, "loss": 0.067, "step": 20730 }, { "grad_norm": 0.6189528703689575, "learning_rate": 9.337904802797078e-05, "loss": 0.0605, "step": 20740 }, { "grad_norm": 0.4828248918056488, "learning_rate": 9.337082301749993e-05, "loss": 0.0659, "step": 20750 }, { "grad_norm": 0.9381044507026672, "learning_rate": 9.336259326406109e-05, "loss": 0.0685, "step": 20760 }, { "grad_norm": 0.5581046938896179, "learning_rate": 9.335435876855427e-05, "loss": 0.0737, "step": 20770 }, { "grad_norm": 1.3865042924880981, "learning_rate": 9.334611953187994e-05, "loss": 0.0718, "step": 20780 }, { "grad_norm": 0.3996114134788513, "learning_rate": 9.333787555493914e-05, "loss": 0.0616, "step": 20790 }, { "grad_norm": 0.6223896145820618, "learning_rate": 9.332962683863345e-05, "loss": 0.0629, "step": 20800 }, { "grad_norm": 0.47925958037376404, "learning_rate": 9.332137338386489e-05, "loss": 0.0622, "step": 20810 }, { "grad_norm": 0.4954361021518707, "learning_rate": 9.33131151915361e-05, "loss": 0.0753, "step": 20820 }, { "grad_norm": 0.4471040368080139, "learning_rate": 9.330485226255012e-05, "loss": 0.061, "step": 20830 }, { "grad_norm": 0.44474658370018005, "learning_rate": 9.329658459781061e-05, "loss": 0.0703, "step": 20840 }, { "grad_norm": 0.40099024772644043, "learning_rate": 9.328831219822172e-05, "loss": 0.0621, "step": 20850 }, { "grad_norm": 0.4178962707519531, "learning_rate": 9.328003506468808e-05, "loss": 0.0696, "step": 20860 }, { "grad_norm": 0.4825117886066437, "learning_rate": 9.327175319811488e-05, "loss": 0.0648, "step": 20870 }, { "grad_norm": 0.45579496026039124, "learning_rate": 9.326346659940781e-05, "loss": 0.0601, "step": 20880 }, { "grad_norm": 0.5120090246200562, "learning_rate": 9.325517526947308e-05, "loss": 0.054, "step": 20890 }, { "grad_norm": 0.84293532371521, "learning_rate": 9.32468792092174e-05, "loss": 0.0831, "step": 20900 }, { "grad_norm": 0.49807998538017273, "learning_rate": 9.323857841954803e-05, "loss": 0.0632, "step": 20910 }, { "grad_norm": 0.3621980547904968, "learning_rate": 9.323027290137276e-05, "loss": 0.0598, "step": 20920 }, { "grad_norm": 0.2678508460521698, "learning_rate": 9.322196265559981e-05, "loss": 0.0633, "step": 20930 }, { "grad_norm": 0.7414566874504089, "learning_rate": 9.321364768313803e-05, "loss": 0.0587, "step": 20940 }, { "grad_norm": 0.5092113614082336, "learning_rate": 9.32053279848967e-05, "loss": 0.0623, "step": 20950 }, { "grad_norm": 0.6205677390098572, "learning_rate": 9.319700356178567e-05, "loss": 0.0641, "step": 20960 }, { "grad_norm": 0.6673940420150757, "learning_rate": 9.318867441471527e-05, "loss": 0.0529, "step": 20970 }, { "grad_norm": 0.5682013034820557, "learning_rate": 9.318034054459637e-05, "loss": 0.0541, "step": 20980 }, { "grad_norm": 0.7478047609329224, "learning_rate": 9.317200195234034e-05, "loss": 0.0551, "step": 20990 }, { "grad_norm": 0.4825708866119385, "learning_rate": 9.316365863885909e-05, "loss": 0.0625, "step": 21000 }, { "grad_norm": 0.45289841294288635, "learning_rate": 9.315531060506502e-05, "loss": 0.0713, "step": 21010 }, { "grad_norm": 0.37019863724708557, "learning_rate": 9.314695785187108e-05, "loss": 0.0583, "step": 21020 }, { "grad_norm": 0.864518940448761, "learning_rate": 9.313860038019069e-05, "loss": 0.0674, "step": 21030 }, { "grad_norm": 0.45875200629234314, "learning_rate": 9.313023819093782e-05, "loss": 0.0645, "step": 21040 }, { "grad_norm": 0.5703597068786621, "learning_rate": 9.312187128502695e-05, "loss": 0.0634, "step": 21050 }, { "grad_norm": 0.42809081077575684, "learning_rate": 9.311349966337307e-05, "loss": 0.073, "step": 21060 }, { "grad_norm": 0.5712168216705322, "learning_rate": 9.310512332689169e-05, "loss": 0.0752, "step": 21070 }, { "grad_norm": 0.4826411008834839, "learning_rate": 9.309674227649883e-05, "loss": 0.0647, "step": 21080 }, { "grad_norm": 0.42488136887550354, "learning_rate": 9.308835651311103e-05, "loss": 0.0668, "step": 21090 }, { "grad_norm": 0.4329875707626343, "learning_rate": 9.307996603764533e-05, "loss": 0.0684, "step": 21100 }, { "grad_norm": 0.7274289131164551, "learning_rate": 9.307157085101932e-05, "loss": 0.0651, "step": 21110 }, { "grad_norm": 0.4878891110420227, "learning_rate": 9.306317095415109e-05, "loss": 0.0651, "step": 21120 }, { "grad_norm": 0.40512320399284363, "learning_rate": 9.305476634795922e-05, "loss": 0.0667, "step": 21130 }, { "grad_norm": 0.5048794746398926, "learning_rate": 9.304635703336284e-05, "loss": 0.058, "step": 21140 }, { "grad_norm": 0.4427172541618347, "learning_rate": 9.303794301128157e-05, "loss": 0.051, "step": 21150 }, { "grad_norm": 0.4516366422176361, "learning_rate": 9.302952428263555e-05, "loss": 0.0688, "step": 21160 }, { "grad_norm": 0.5312151312828064, "learning_rate": 9.302110084834545e-05, "loss": 0.064, "step": 21170 }, { "grad_norm": 0.4296814501285553, "learning_rate": 9.301267270933245e-05, "loss": 0.059, "step": 21180 }, { "grad_norm": 1.0887309312820435, "learning_rate": 9.300423986651823e-05, "loss": 0.0652, "step": 21190 }, { "grad_norm": 0.4758943021297455, "learning_rate": 9.299580232082501e-05, "loss": 0.065, "step": 21200 }, { "grad_norm": 0.5386186838150024, "learning_rate": 9.298736007317547e-05, "loss": 0.0576, "step": 21210 }, { "grad_norm": 0.45610231161117554, "learning_rate": 9.297891312449288e-05, "loss": 0.068, "step": 21220 }, { "grad_norm": 0.5237026214599609, "learning_rate": 9.297046147570094e-05, "loss": 0.0672, "step": 21230 }, { "grad_norm": 0.48616600036621094, "learning_rate": 9.296200512772396e-05, "loss": 0.0624, "step": 21240 }, { "grad_norm": 0.47373393177986145, "learning_rate": 9.295354408148668e-05, "loss": 0.0607, "step": 21250 }, { "grad_norm": 0.4787628650665283, "learning_rate": 9.294507833791441e-05, "loss": 0.0697, "step": 21260 }, { "grad_norm": 0.43361425399780273, "learning_rate": 9.293660789793295e-05, "loss": 0.0641, "step": 21270 }, { "grad_norm": 0.4829561114311218, "learning_rate": 9.292813276246858e-05, "loss": 0.062, "step": 21280 }, { "grad_norm": 0.8170564770698547, "learning_rate": 9.291965293244816e-05, "loss": 0.0544, "step": 21290 }, { "grad_norm": 0.4724957346916199, "learning_rate": 9.291116840879904e-05, "loss": 0.0585, "step": 21300 }, { "grad_norm": 0.627755343914032, "learning_rate": 9.290267919244904e-05, "loss": 0.0597, "step": 21310 }, { "grad_norm": 0.37510958313941956, "learning_rate": 9.289418528432655e-05, "loss": 0.0579, "step": 21320 }, { "grad_norm": 0.49261581897735596, "learning_rate": 9.288568668536045e-05, "loss": 0.0831, "step": 21330 }, { "grad_norm": 0.4401322305202484, "learning_rate": 9.287718339648013e-05, "loss": 0.0632, "step": 21340 }, { "grad_norm": 0.40220335125923157, "learning_rate": 9.28686754186155e-05, "loss": 0.0658, "step": 21350 }, { "grad_norm": 0.5475693345069885, "learning_rate": 9.286016275269698e-05, "loss": 0.0777, "step": 21360 }, { "grad_norm": 0.5575567483901978, "learning_rate": 9.285164539965551e-05, "loss": 0.0583, "step": 21370 }, { "grad_norm": 0.5399059653282166, "learning_rate": 9.284312336042251e-05, "loss": 0.052, "step": 21380 }, { "grad_norm": 0.4213319718837738, "learning_rate": 9.283459663592996e-05, "loss": 0.0581, "step": 21390 }, { "grad_norm": 0.6664466857910156, "learning_rate": 9.282606522711033e-05, "loss": 0.0689, "step": 21400 }, { "grad_norm": 0.3074309825897217, "learning_rate": 9.281752913489657e-05, "loss": 0.0638, "step": 21410 }, { "grad_norm": 0.569466769695282, "learning_rate": 9.280898836022222e-05, "loss": 0.0702, "step": 21420 }, { "grad_norm": 0.5911204814910889, "learning_rate": 9.280044290402126e-05, "loss": 0.0677, "step": 21430 }, { "grad_norm": 0.48158711194992065, "learning_rate": 9.279189276722821e-05, "loss": 0.0689, "step": 21440 }, { "grad_norm": 0.8324434161186218, "learning_rate": 9.278333795077812e-05, "loss": 0.0679, "step": 21450 }, { "grad_norm": 0.5434072613716125, "learning_rate": 9.27747784556065e-05, "loss": 0.0563, "step": 21460 }, { "grad_norm": 0.4226687550544739, "learning_rate": 9.276621428264942e-05, "loss": 0.0706, "step": 21470 }, { "grad_norm": 0.5323660373687744, "learning_rate": 9.275764543284345e-05, "loss": 0.0522, "step": 21480 }, { "grad_norm": 0.4658588767051697, "learning_rate": 9.274907190712566e-05, "loss": 0.0659, "step": 21490 }, { "grad_norm": 0.47412317991256714, "learning_rate": 9.274049370643363e-05, "loss": 0.0673, "step": 21500 }, { "grad_norm": 0.39042356610298157, "learning_rate": 9.273191083170547e-05, "loss": 0.0685, "step": 21510 }, { "grad_norm": 0.5057073831558228, "learning_rate": 9.27233232838798e-05, "loss": 0.0595, "step": 21520 }, { "grad_norm": 0.5089465975761414, "learning_rate": 9.27147310638957e-05, "loss": 0.0569, "step": 21530 }, { "grad_norm": 0.5348525643348694, "learning_rate": 9.270613417269286e-05, "loss": 0.0629, "step": 21540 }, { "grad_norm": 0.495769202709198, "learning_rate": 9.269753261121138e-05, "loss": 0.0626, "step": 21550 }, { "grad_norm": 0.37024399638175964, "learning_rate": 9.268892638039194e-05, "loss": 0.0572, "step": 21560 }, { "grad_norm": 0.7011501789093018, "learning_rate": 9.268031548117569e-05, "loss": 0.07, "step": 21570 }, { "grad_norm": 0.4796820282936096, "learning_rate": 9.26716999145043e-05, "loss": 0.0529, "step": 21580 }, { "grad_norm": 0.45882782340049744, "learning_rate": 9.266307968131998e-05, "loss": 0.0657, "step": 21590 }, { "grad_norm": 0.42636480927467346, "learning_rate": 9.26544547825654e-05, "loss": 0.0537, "step": 21600 }, { "grad_norm": 0.3997771143913269, "learning_rate": 9.264582521918376e-05, "loss": 0.0709, "step": 21610 }, { "grad_norm": 0.6306847929954529, "learning_rate": 9.263719099211881e-05, "loss": 0.0591, "step": 21620 }, { "grad_norm": 0.4888899624347687, "learning_rate": 9.262855210231476e-05, "loss": 0.0657, "step": 21630 }, { "grad_norm": 0.6189544796943665, "learning_rate": 9.261990855071633e-05, "loss": 0.0608, "step": 21640 }, { "grad_norm": 0.3822519779205322, "learning_rate": 9.261126033826878e-05, "loss": 0.0569, "step": 21650 }, { "grad_norm": 0.4926575720310211, "learning_rate": 9.260260746591786e-05, "loss": 0.0532, "step": 21660 }, { "grad_norm": 0.4237489402294159, "learning_rate": 9.259394993460985e-05, "loss": 0.0522, "step": 21670 }, { "grad_norm": 0.5611758828163147, "learning_rate": 9.258528774529151e-05, "loss": 0.0479, "step": 21680 }, { "grad_norm": 0.6090332269668579, "learning_rate": 9.257662089891013e-05, "loss": 0.0531, "step": 21690 }, { "grad_norm": 0.5878511071205139, "learning_rate": 9.25679493964135e-05, "loss": 0.0626, "step": 21700 }, { "grad_norm": 0.4203527569770813, "learning_rate": 9.255927323874994e-05, "loss": 0.0579, "step": 21710 }, { "grad_norm": 0.6662154197692871, "learning_rate": 9.255059242686822e-05, "loss": 0.0838, "step": 21720 }, { "grad_norm": 0.555899977684021, "learning_rate": 9.254190696171769e-05, "loss": 0.0608, "step": 21730 }, { "grad_norm": 0.5143251419067383, "learning_rate": 9.25332168442482e-05, "loss": 0.0641, "step": 21740 }, { "grad_norm": 0.6476737260818481, "learning_rate": 9.252452207541004e-05, "loss": 0.0581, "step": 21750 }, { "grad_norm": 0.7539442777633667, "learning_rate": 9.251582265615409e-05, "loss": 0.0777, "step": 21760 }, { "grad_norm": 0.4334961771965027, "learning_rate": 9.250711858743169e-05, "loss": 0.0666, "step": 21770 }, { "grad_norm": 0.41286954283714294, "learning_rate": 9.24984098701947e-05, "loss": 0.0666, "step": 21780 }, { "grad_norm": 0.6001006364822388, "learning_rate": 9.248969650539552e-05, "loss": 0.06, "step": 21790 }, { "grad_norm": 0.4681745767593384, "learning_rate": 9.2480978493987e-05, "loss": 0.0597, "step": 21800 }, { "grad_norm": 0.3485301434993744, "learning_rate": 9.247225583692256e-05, "loss": 0.0542, "step": 21810 }, { "grad_norm": 0.8139595985412598, "learning_rate": 9.246352853515607e-05, "loss": 0.0592, "step": 21820 }, { "grad_norm": 0.6888283491134644, "learning_rate": 9.245479658964194e-05, "loss": 0.0574, "step": 21830 }, { "grad_norm": 0.4592781960964203, "learning_rate": 9.244606000133507e-05, "loss": 0.0694, "step": 21840 }, { "grad_norm": 0.47444188594818115, "learning_rate": 9.24373187711909e-05, "loss": 0.0518, "step": 21850 }, { "grad_norm": 0.482646107673645, "learning_rate": 9.242857290016537e-05, "loss": 0.0561, "step": 21860 }, { "grad_norm": 0.5051520466804504, "learning_rate": 9.241982238921488e-05, "loss": 0.0638, "step": 21870 }, { "grad_norm": 0.42507848143577576, "learning_rate": 9.24110672392964e-05, "loss": 0.0581, "step": 21880 }, { "grad_norm": 0.36943110823631287, "learning_rate": 9.240230745136737e-05, "loss": 0.0482, "step": 21890 }, { "grad_norm": 0.5017275810241699, "learning_rate": 9.239354302638575e-05, "loss": 0.0525, "step": 21900 }, { "grad_norm": 0.5116034150123596, "learning_rate": 9.238477396531e-05, "loss": 0.0646, "step": 21910 }, { "grad_norm": 0.5076795220375061, "learning_rate": 9.23760002690991e-05, "loss": 0.0647, "step": 21920 }, { "grad_norm": 0.5014586448669434, "learning_rate": 9.236722193871252e-05, "loss": 0.057, "step": 21930 }, { "grad_norm": 0.5580476522445679, "learning_rate": 9.235843897511023e-05, "loss": 0.058, "step": 21940 }, { "grad_norm": 0.4679100215435028, "learning_rate": 9.234965137925276e-05, "loss": 0.0494, "step": 21950 }, { "grad_norm": 0.7396606802940369, "learning_rate": 9.234085915210108e-05, "loss": 0.0568, "step": 21960 }, { "grad_norm": 0.8408891558647156, "learning_rate": 9.23320622946167e-05, "loss": 0.0565, "step": 21970 }, { "grad_norm": 0.5144662261009216, "learning_rate": 9.232326080776163e-05, "loss": 0.0715, "step": 21980 }, { "grad_norm": 0.530777633190155, "learning_rate": 9.23144546924984e-05, "loss": 0.0628, "step": 21990 }, { "grad_norm": 0.41041114926338196, "learning_rate": 9.230564394979e-05, "loss": 0.0637, "step": 22000 }, { "grad_norm": 0.5565873384475708, "learning_rate": 9.22968285806e-05, "loss": 0.056, "step": 22010 }, { "grad_norm": 0.4919584095478058, "learning_rate": 9.228800858589242e-05, "loss": 0.0606, "step": 22020 }, { "grad_norm": 0.35326677560806274, "learning_rate": 9.227918396663179e-05, "loss": 0.061, "step": 22030 }, { "grad_norm": 0.8725579977035522, "learning_rate": 9.227035472378319e-05, "loss": 0.0588, "step": 22040 }, { "grad_norm": 0.40065643191337585, "learning_rate": 9.226152085831213e-05, "loss": 0.0492, "step": 22050 }, { "grad_norm": 0.4875454306602478, "learning_rate": 9.22526823711847e-05, "loss": 0.0567, "step": 22060 }, { "grad_norm": 0.6126884818077087, "learning_rate": 9.224383926336745e-05, "loss": 0.0505, "step": 22070 }, { "grad_norm": 0.3741256296634674, "learning_rate": 9.223499153582744e-05, "loss": 0.0517, "step": 22080 }, { "grad_norm": 0.4015422761440277, "learning_rate": 9.222613918953226e-05, "loss": 0.0684, "step": 22090 }, { "grad_norm": 0.4101387858390808, "learning_rate": 9.221728222544999e-05, "loss": 0.0574, "step": 22100 }, { "grad_norm": 0.586704432964325, "learning_rate": 9.22084206445492e-05, "loss": 0.0656, "step": 22110 }, { "grad_norm": 0.40087538957595825, "learning_rate": 9.2199554447799e-05, "loss": 0.071, "step": 22120 }, { "grad_norm": 0.3654704988002777, "learning_rate": 9.219068363616897e-05, "loss": 0.0578, "step": 22130 }, { "grad_norm": 0.6392187476158142, "learning_rate": 9.218180821062919e-05, "loss": 0.0637, "step": 22140 }, { "grad_norm": 0.7573833465576172, "learning_rate": 9.21729281721503e-05, "loss": 0.0713, "step": 22150 }, { "grad_norm": 0.4149189889431, "learning_rate": 9.216404352170339e-05, "loss": 0.0586, "step": 22160 }, { "grad_norm": 0.4459666609764099, "learning_rate": 9.215515426026007e-05, "loss": 0.0571, "step": 22170 }, { "grad_norm": 0.3916003704071045, "learning_rate": 9.214626038879246e-05, "loss": 0.0467, "step": 22180 }, { "grad_norm": 0.42933589220046997, "learning_rate": 9.21373619082732e-05, "loss": 0.052, "step": 22190 }, { "grad_norm": 0.6219271421432495, "learning_rate": 9.212845881967535e-05, "loss": 0.0518, "step": 22200 }, { "grad_norm": 0.6190275549888611, "learning_rate": 9.211955112397262e-05, "loss": 0.0522, "step": 22210 }, { "grad_norm": 0.49873092770576477, "learning_rate": 9.211063882213909e-05, "loss": 0.0527, "step": 22220 }, { "grad_norm": 0.4786223769187927, "learning_rate": 9.210172191514942e-05, "loss": 0.042, "step": 22230 }, { "grad_norm": 0.7542403936386108, "learning_rate": 9.209280040397874e-05, "loss": 0.0557, "step": 22240 }, { "grad_norm": 0.4599112272262573, "learning_rate": 9.208387428960268e-05, "loss": 0.048, "step": 22250 }, { "grad_norm": 0.3955696225166321, "learning_rate": 9.20749435729974e-05, "loss": 0.0545, "step": 22260 }, { "grad_norm": 1.1406364440917969, "learning_rate": 9.206600825513957e-05, "loss": 0.047, "step": 22270 }, { "grad_norm": 0.4571143090724945, "learning_rate": 9.20570683370063e-05, "loss": 0.0518, "step": 22280 }, { "grad_norm": 0.5386435389518738, "learning_rate": 9.204812381957528e-05, "loss": 0.0504, "step": 22290 }, { "grad_norm": 0.6737300753593445, "learning_rate": 9.203917470382465e-05, "loss": 0.0517, "step": 22300 }, { "grad_norm": 0.4037025570869446, "learning_rate": 9.203022099073309e-05, "loss": 0.0507, "step": 22310 }, { "grad_norm": 0.6306372284889221, "learning_rate": 9.202126268127976e-05, "loss": 0.0564, "step": 22320 }, { "grad_norm": 0.6276096105575562, "learning_rate": 9.20122997764443e-05, "loss": 0.0604, "step": 22330 }, { "grad_norm": 1.0501753091812134, "learning_rate": 9.200333227720692e-05, "loss": 0.0714, "step": 22340 }, { "grad_norm": 0.37608882784843445, "learning_rate": 9.199436018454826e-05, "loss": 0.0584, "step": 22350 }, { "grad_norm": 0.4963574707508087, "learning_rate": 9.198538349944952e-05, "loss": 0.0541, "step": 22360 }, { "grad_norm": 0.5262537002563477, "learning_rate": 9.197640222289234e-05, "loss": 0.0658, "step": 22370 }, { "grad_norm": 0.9049615859985352, "learning_rate": 9.196741635585895e-05, "loss": 0.0651, "step": 22380 }, { "grad_norm": 0.6592603325843811, "learning_rate": 9.195842589933199e-05, "loss": 0.0583, "step": 22390 }, { "grad_norm": 0.7068859338760376, "learning_rate": 9.194943085429466e-05, "loss": 0.0631, "step": 22400 }, { "grad_norm": 0.38838034868240356, "learning_rate": 9.194043122173065e-05, "loss": 0.0673, "step": 22410 }, { "grad_norm": 0.4870874285697937, "learning_rate": 9.193142700262413e-05, "loss": 0.0599, "step": 22420 }, { "grad_norm": 0.4528581202030182, "learning_rate": 9.192241819795979e-05, "loss": 0.0563, "step": 22430 }, { "grad_norm": 0.9243950247764587, "learning_rate": 9.191340480872284e-05, "loss": 0.0626, "step": 22440 }, { "grad_norm": 0.6699384450912476, "learning_rate": 9.190438683589895e-05, "loss": 0.0629, "step": 22450 }, { "grad_norm": 0.535615086555481, "learning_rate": 9.189536428047432e-05, "loss": 0.0704, "step": 22460 }, { "grad_norm": 0.6720163226127625, "learning_rate": 9.188633714343564e-05, "loss": 0.0757, "step": 22470 }, { "grad_norm": 0.5854867696762085, "learning_rate": 9.18773054257701e-05, "loss": 0.0656, "step": 22480 }, { "grad_norm": 0.8709897994995117, "learning_rate": 9.18682691284654e-05, "loss": 0.0625, "step": 22490 }, { "grad_norm": 0.5681429505348206, "learning_rate": 9.185922825250974e-05, "loss": 0.0538, "step": 22500 }, { "grad_norm": 0.47012239694595337, "learning_rate": 9.185018279889181e-05, "loss": 0.0457, "step": 22510 }, { "grad_norm": 0.9350798726081848, "learning_rate": 9.184113276860082e-05, "loss": 0.0609, "step": 22520 }, { "grad_norm": 0.6553289890289307, "learning_rate": 9.183207816262645e-05, "loss": 0.0657, "step": 22530 }, { "grad_norm": 0.42865636944770813, "learning_rate": 9.182301898195891e-05, "loss": 0.0436, "step": 22540 }, { "grad_norm": 0.5368898510932922, "learning_rate": 9.181395522758889e-05, "loss": 0.0738, "step": 22550 }, { "grad_norm": 0.8704342842102051, "learning_rate": 9.180488690050759e-05, "loss": 0.0792, "step": 22560 }, { "grad_norm": 0.3769721984863281, "learning_rate": 9.179581400170671e-05, "loss": 0.0501, "step": 22570 }, { "grad_norm": 0.4954010844230652, "learning_rate": 9.178673653217845e-05, "loss": 0.0607, "step": 22580 }, { "grad_norm": 0.7132068276405334, "learning_rate": 9.177765449291551e-05, "loss": 0.0569, "step": 22590 }, { "grad_norm": 0.37047290802001953, "learning_rate": 9.176856788491109e-05, "loss": 0.0564, "step": 22600 }, { "grad_norm": 0.44738033413887024, "learning_rate": 9.175947670915887e-05, "loss": 0.0519, "step": 22610 }, { "grad_norm": 0.5872564911842346, "learning_rate": 9.175038096665309e-05, "loss": 0.0556, "step": 22620 }, { "grad_norm": 0.5185151100158691, "learning_rate": 9.17412806583884e-05, "loss": 0.0564, "step": 22630 }, { "grad_norm": 0.4256534278392792, "learning_rate": 9.173217578536002e-05, "loss": 0.0651, "step": 22640 }, { "grad_norm": 0.6153314709663391, "learning_rate": 9.172306634856362e-05, "loss": 0.0732, "step": 22650 }, { "grad_norm": 0.6187392473220825, "learning_rate": 9.171395234899545e-05, "loss": 0.0534, "step": 22660 }, { "grad_norm": 0.584439218044281, "learning_rate": 9.170483378765214e-05, "loss": 0.0537, "step": 22670 }, { "grad_norm": 0.43343353271484375, "learning_rate": 9.169571066553091e-05, "loss": 0.0552, "step": 22680 }, { "grad_norm": 0.5896680951118469, "learning_rate": 9.168658298362946e-05, "loss": 0.0629, "step": 22690 }, { "grad_norm": 0.5288482904434204, "learning_rate": 9.167745074294598e-05, "loss": 0.0521, "step": 22700 }, { "grad_norm": 0.56710284948349, "learning_rate": 9.166831394447913e-05, "loss": 0.066, "step": 22710 }, { "grad_norm": 0.5112552046775818, "learning_rate": 9.165917258922812e-05, "loss": 0.0567, "step": 22720 }, { "grad_norm": 0.47623205184936523, "learning_rate": 9.165002667819262e-05, "loss": 0.0596, "step": 22730 }, { "grad_norm": 0.44055479764938354, "learning_rate": 9.164087621237282e-05, "loss": 0.058, "step": 22740 }, { "grad_norm": 0.5781305432319641, "learning_rate": 9.163172119276942e-05, "loss": 0.0693, "step": 22750 }, { "grad_norm": 0.5706616044044495, "learning_rate": 9.162256162038358e-05, "loss": 0.0747, "step": 22760 }, { "grad_norm": 0.4950575828552246, "learning_rate": 9.161339749621698e-05, "loss": 0.0637, "step": 22770 }, { "grad_norm": 0.38617420196533203, "learning_rate": 9.160422882127177e-05, "loss": 0.0554, "step": 22780 }, { "grad_norm": 0.38296210765838623, "learning_rate": 9.159505559655069e-05, "loss": 0.0473, "step": 22790 }, { "grad_norm": 0.5652896165847778, "learning_rate": 9.158587782305684e-05, "loss": 0.0624, "step": 22800 }, { "grad_norm": 0.3810248076915741, "learning_rate": 9.157669550179391e-05, "loss": 0.0555, "step": 22810 }, { "grad_norm": 0.4241912066936493, "learning_rate": 9.156750863376609e-05, "loss": 0.0643, "step": 22820 }, { "grad_norm": 0.40485909581184387, "learning_rate": 9.155831721997801e-05, "loss": 0.0578, "step": 22830 }, { "grad_norm": 0.5051476359367371, "learning_rate": 9.154912126143484e-05, "loss": 0.0665, "step": 22840 }, { "grad_norm": 0.6334030628204346, "learning_rate": 9.153992075914224e-05, "loss": 0.0695, "step": 22850 }, { "grad_norm": 0.4154733121395111, "learning_rate": 9.153071571410635e-05, "loss": 0.0435, "step": 22860 }, { "grad_norm": 0.5731126666069031, "learning_rate": 9.152150612733384e-05, "loss": 0.054, "step": 22870 }, { "grad_norm": 0.5731433033943176, "learning_rate": 9.151229199983184e-05, "loss": 0.0683, "step": 22880 }, { "grad_norm": 0.5748174786567688, "learning_rate": 9.150307333260802e-05, "loss": 0.0584, "step": 22890 }, { "grad_norm": 0.5073225498199463, "learning_rate": 9.149385012667048e-05, "loss": 0.0744, "step": 22900 }, { "grad_norm": 0.6447618007659912, "learning_rate": 9.148462238302788e-05, "loss": 0.0675, "step": 22910 }, { "grad_norm": 0.4260808527469635, "learning_rate": 9.147539010268936e-05, "loss": 0.0624, "step": 22920 }, { "grad_norm": 1.1107683181762695, "learning_rate": 9.14661532866645e-05, "loss": 0.0616, "step": 22930 }, { "grad_norm": 1.019986867904663, "learning_rate": 9.145691193596348e-05, "loss": 0.0632, "step": 22940 }, { "grad_norm": 0.5521666407585144, "learning_rate": 9.144766605159691e-05, "loss": 0.0683, "step": 22950 }, { "grad_norm": 0.48208898305892944, "learning_rate": 9.14384156345759e-05, "loss": 0.0557, "step": 22960 }, { "grad_norm": 0.777580201625824, "learning_rate": 9.142916068591204e-05, "loss": 0.054, "step": 22970 }, { "grad_norm": 0.7654664516448975, "learning_rate": 9.141990120661746e-05, "loss": 0.065, "step": 22980 }, { "grad_norm": 0.45988014340400696, "learning_rate": 9.141063719770475e-05, "loss": 0.0607, "step": 22990 }, { "grad_norm": 0.3975081741809845, "learning_rate": 9.140136866018704e-05, "loss": 0.0521, "step": 23000 }, { "grad_norm": 0.4166940748691559, "learning_rate": 9.139209559507788e-05, "loss": 0.0526, "step": 23010 }, { "grad_norm": 0.5125858783721924, "learning_rate": 9.13828180033914e-05, "loss": 0.0629, "step": 23020 }, { "grad_norm": 0.4367315173149109, "learning_rate": 9.137353588614212e-05, "loss": 0.0574, "step": 23030 }, { "grad_norm": 0.5224470496177673, "learning_rate": 9.136424924434519e-05, "loss": 0.0621, "step": 23040 }, { "grad_norm": 0.8097521066665649, "learning_rate": 9.135495807901615e-05, "loss": 0.0659, "step": 23050 }, { "grad_norm": 0.36334314942359924, "learning_rate": 9.134566239117108e-05, "loss": 0.0495, "step": 23060 }, { "grad_norm": 0.5509983897209167, "learning_rate": 9.13363621818265e-05, "loss": 0.0534, "step": 23070 }, { "grad_norm": 0.541938066482544, "learning_rate": 9.132705745199953e-05, "loss": 0.0549, "step": 23080 }, { "grad_norm": 0.8038413524627686, "learning_rate": 9.131774820270768e-05, "loss": 0.0526, "step": 23090 }, { "grad_norm": 0.5434849262237549, "learning_rate": 9.130843443496901e-05, "loss": 0.0557, "step": 23100 }, { "grad_norm": 0.389290452003479, "learning_rate": 9.129911614980206e-05, "loss": 0.0646, "step": 23110 }, { "grad_norm": 0.6169844269752502, "learning_rate": 9.128979334822584e-05, "loss": 0.054, "step": 23120 }, { "grad_norm": 0.3474586009979248, "learning_rate": 9.128046603125992e-05, "loss": 0.0578, "step": 23130 }, { "grad_norm": 0.38697266578674316, "learning_rate": 9.12711341999243e-05, "loss": 0.0497, "step": 23140 }, { "grad_norm": 0.45605629682540894, "learning_rate": 9.12617978552395e-05, "loss": 0.0785, "step": 23150 }, { "grad_norm": 0.5577274560928345, "learning_rate": 9.12524569982265e-05, "loss": 0.0608, "step": 23160 }, { "grad_norm": 0.4372688829898834, "learning_rate": 9.124311162990684e-05, "loss": 0.0548, "step": 23170 }, { "grad_norm": 0.3736056685447693, "learning_rate": 9.12337617513025e-05, "loss": 0.0457, "step": 23180 }, { "grad_norm": 0.3673502802848816, "learning_rate": 9.122440736343596e-05, "loss": 0.0494, "step": 23190 }, { "grad_norm": 0.5683883428573608, "learning_rate": 9.12150484673302e-05, "loss": 0.0567, "step": 23200 }, { "grad_norm": 0.40901899337768555, "learning_rate": 9.120568506400873e-05, "loss": 0.0534, "step": 23210 }, { "grad_norm": 0.42885300517082214, "learning_rate": 9.119631715449548e-05, "loss": 0.0477, "step": 23220 }, { "grad_norm": 0.5165443420410156, "learning_rate": 9.118694473981493e-05, "loss": 0.0659, "step": 23230 }, { "grad_norm": 0.5450721383094788, "learning_rate": 9.117756782099203e-05, "loss": 0.0478, "step": 23240 }, { "grad_norm": 0.492827832698822, "learning_rate": 9.11681863990522e-05, "loss": 0.0502, "step": 23250 }, { "grad_norm": 0.5021347403526306, "learning_rate": 9.115880047502142e-05, "loss": 0.0606, "step": 23260 }, { "grad_norm": 0.6894471645355225, "learning_rate": 9.114941004992609e-05, "loss": 0.0649, "step": 23270 }, { "grad_norm": 0.8034186363220215, "learning_rate": 9.114001512479317e-05, "loss": 0.0551, "step": 23280 }, { "grad_norm": 0.4070020914077759, "learning_rate": 9.113061570065003e-05, "loss": 0.0668, "step": 23290 }, { "grad_norm": 0.5944883823394775, "learning_rate": 9.112121177852459e-05, "loss": 0.0636, "step": 23300 }, { "grad_norm": 0.5526825189590454, "learning_rate": 9.111180335944527e-05, "loss": 0.0543, "step": 23310 }, { "grad_norm": 0.5864605903625488, "learning_rate": 9.110239044444093e-05, "loss": 0.0539, "step": 23320 }, { "grad_norm": 0.3796246647834778, "learning_rate": 9.109297303454099e-05, "loss": 0.0569, "step": 23330 }, { "grad_norm": 0.45980945229530334, "learning_rate": 9.108355113077526e-05, "loss": 0.0597, "step": 23340 }, { "grad_norm": 0.4036484658718109, "learning_rate": 9.107412473417419e-05, "loss": 0.062, "step": 23350 }, { "grad_norm": 0.8056945204734802, "learning_rate": 9.106469384576858e-05, "loss": 0.0603, "step": 23360 }, { "grad_norm": 0.4020267724990845, "learning_rate": 9.105525846658978e-05, "loss": 0.0629, "step": 23370 }, { "grad_norm": 0.5432178974151611, "learning_rate": 9.104581859766965e-05, "loss": 0.0703, "step": 23380 }, { "grad_norm": 0.8376476764678955, "learning_rate": 9.10363742400405e-05, "loss": 0.0712, "step": 23390 }, { "grad_norm": 0.41794151067733765, "learning_rate": 9.102692539473518e-05, "loss": 0.0655, "step": 23400 }, { "grad_norm": 0.41161540150642395, "learning_rate": 9.101747206278697e-05, "loss": 0.0661, "step": 23410 }, { "grad_norm": 0.5559561848640442, "learning_rate": 9.100801424522968e-05, "loss": 0.0568, "step": 23420 }, { "grad_norm": 0.7929554581642151, "learning_rate": 9.099855194309762e-05, "loss": 0.067, "step": 23430 }, { "grad_norm": 0.6632627248764038, "learning_rate": 9.098908515742554e-05, "loss": 0.0578, "step": 23440 }, { "grad_norm": 0.47952741384506226, "learning_rate": 9.097961388924873e-05, "loss": 0.0541, "step": 23450 }, { "grad_norm": 0.44722238183021545, "learning_rate": 9.097013813960298e-05, "loss": 0.0668, "step": 23460 }, { "grad_norm": 0.49007704854011536, "learning_rate": 9.09606579095245e-05, "loss": 0.0588, "step": 23470 }, { "grad_norm": 0.4484451711177826, "learning_rate": 9.095117320005008e-05, "loss": 0.0553, "step": 23480 }, { "grad_norm": 0.5963743925094604, "learning_rate": 9.094168401221691e-05, "loss": 0.0682, "step": 23490 }, { "grad_norm": 0.4202639162540436, "learning_rate": 9.093219034706273e-05, "loss": 0.0698, "step": 23500 }, { "grad_norm": 0.6832425594329834, "learning_rate": 9.092269220562577e-05, "loss": 0.0673, "step": 23510 }, { "grad_norm": 0.42792704701423645, "learning_rate": 9.09131895889447e-05, "loss": 0.0894, "step": 23520 }, { "grad_norm": 0.686927080154419, "learning_rate": 9.090368249805873e-05, "loss": 0.0576, "step": 23530 }, { "grad_norm": 0.3524669408798218, "learning_rate": 9.089417093400754e-05, "loss": 0.0625, "step": 23540 }, { "grad_norm": 0.6564162969589233, "learning_rate": 9.088465489783131e-05, "loss": 0.0549, "step": 23550 }, { "grad_norm": 0.3509758412837982, "learning_rate": 9.087513439057068e-05, "loss": 0.0556, "step": 23560 }, { "grad_norm": 0.3598053455352783, "learning_rate": 9.08656094132668e-05, "loss": 0.0607, "step": 23570 }, { "grad_norm": 0.5746265649795532, "learning_rate": 9.085607996696134e-05, "loss": 0.0551, "step": 23580 }, { "grad_norm": 0.5307992696762085, "learning_rate": 9.084654605269639e-05, "loss": 0.059, "step": 23590 }, { "grad_norm": 0.5462053418159485, "learning_rate": 9.083700767151457e-05, "loss": 0.0685, "step": 23600 }, { "grad_norm": 0.606933057308197, "learning_rate": 9.082746482445898e-05, "loss": 0.0726, "step": 23610 }, { "grad_norm": 0.37105730175971985, "learning_rate": 9.081791751257325e-05, "loss": 0.0572, "step": 23620 }, { "grad_norm": 0.4889479875564575, "learning_rate": 9.080836573690142e-05, "loss": 0.056, "step": 23630 }, { "grad_norm": 0.5131807327270508, "learning_rate": 9.079880949848805e-05, "loss": 0.0691, "step": 23640 }, { "grad_norm": 0.6887387633323669, "learning_rate": 9.078924879837822e-05, "loss": 0.0693, "step": 23650 }, { "grad_norm": 0.5924184322357178, "learning_rate": 9.077968363761747e-05, "loss": 0.0844, "step": 23660 }, { "grad_norm": 0.47306928038597107, "learning_rate": 9.077011401725182e-05, "loss": 0.0639, "step": 23670 }, { "grad_norm": 0.42119574546813965, "learning_rate": 9.07605399383278e-05, "loss": 0.063, "step": 23680 }, { "grad_norm": 0.5630735158920288, "learning_rate": 9.075096140189243e-05, "loss": 0.057, "step": 23690 }, { "grad_norm": 0.5492733120918274, "learning_rate": 9.074137840899318e-05, "loss": 0.0607, "step": 23700 }, { "grad_norm": 0.6295628547668457, "learning_rate": 9.073179096067804e-05, "loss": 0.0511, "step": 23710 }, { "grad_norm": 0.480854332447052, "learning_rate": 9.072219905799549e-05, "loss": 0.0506, "step": 23720 }, { "grad_norm": 0.5419313311576843, "learning_rate": 9.071260270199447e-05, "loss": 0.0586, "step": 23730 }, { "grad_norm": 0.5297919511795044, "learning_rate": 9.070300189372441e-05, "loss": 0.0565, "step": 23740 }, { "grad_norm": 0.3417738974094391, "learning_rate": 9.069339663423528e-05, "loss": 0.0523, "step": 23750 }, { "grad_norm": 0.3535699248313904, "learning_rate": 9.068378692457747e-05, "loss": 0.0538, "step": 23760 }, { "grad_norm": 0.49326205253601074, "learning_rate": 9.067417276580189e-05, "loss": 0.0544, "step": 23770 }, { "grad_norm": 0.6101590991020203, "learning_rate": 9.066455415895993e-05, "loss": 0.0632, "step": 23780 }, { "grad_norm": 0.47226473689079285, "learning_rate": 9.065493110510346e-05, "loss": 0.0604, "step": 23790 }, { "grad_norm": 0.4428851306438446, "learning_rate": 9.064530360528484e-05, "loss": 0.066, "step": 23800 }, { "grad_norm": 0.6796203255653381, "learning_rate": 9.063567166055695e-05, "loss": 0.0606, "step": 23810 }, { "grad_norm": 0.4832100570201874, "learning_rate": 9.062603527197308e-05, "loss": 0.0578, "step": 23820 }, { "grad_norm": 0.5392678380012512, "learning_rate": 9.06163944405871e-05, "loss": 0.0561, "step": 23830 }, { "grad_norm": 0.5482923984527588, "learning_rate": 9.060674916745327e-05, "loss": 0.0689, "step": 23840 }, { "grad_norm": 0.4617955684661865, "learning_rate": 9.05970994536264e-05, "loss": 0.0503, "step": 23850 }, { "grad_norm": 0.5557423233985901, "learning_rate": 9.05874453001618e-05, "loss": 0.0608, "step": 23860 }, { "grad_norm": 0.5571051836013794, "learning_rate": 9.057778670811517e-05, "loss": 0.0557, "step": 23870 }, { "grad_norm": 0.4427214562892914, "learning_rate": 9.056812367854281e-05, "loss": 0.0667, "step": 23880 }, { "grad_norm": 0.5010796785354614, "learning_rate": 9.055845621250143e-05, "loss": 0.0624, "step": 23890 }, { "grad_norm": 0.4482139050960541, "learning_rate": 9.054878431104825e-05, "loss": 0.0479, "step": 23900 }, { "grad_norm": 0.4574988782405853, "learning_rate": 9.0539107975241e-05, "loss": 0.0493, "step": 23910 }, { "grad_norm": 0.8070902228355408, "learning_rate": 9.052942720613784e-05, "loss": 0.0704, "step": 23920 }, { "grad_norm": 0.6865751147270203, "learning_rate": 9.051974200479745e-05, "loss": 0.0525, "step": 23930 }, { "grad_norm": 0.5288773775100708, "learning_rate": 9.051005237227901e-05, "loss": 0.0536, "step": 23940 }, { "grad_norm": 0.4941781461238861, "learning_rate": 9.050035830964215e-05, "loss": 0.0482, "step": 23950 }, { "grad_norm": 0.5912821888923645, "learning_rate": 9.049065981794698e-05, "loss": 0.0708, "step": 23960 }, { "grad_norm": 0.36116763949394226, "learning_rate": 9.048095689825414e-05, "loss": 0.0555, "step": 23970 }, { "grad_norm": 0.5055028200149536, "learning_rate": 9.047124955162472e-05, "loss": 0.0615, "step": 23980 }, { "grad_norm": 0.5344076156616211, "learning_rate": 9.046153777912028e-05, "loss": 0.0682, "step": 23990 }, { "grad_norm": 0.3342629075050354, "learning_rate": 9.045182158180292e-05, "loss": 0.0587, "step": 24000 }, { "grad_norm": 0.4045271575450897, "learning_rate": 9.044210096073516e-05, "loss": 0.0467, "step": 24010 }, { "grad_norm": 0.4013771712779999, "learning_rate": 9.043237591698004e-05, "loss": 0.052, "step": 24020 }, { "grad_norm": 0.44168174266815186, "learning_rate": 9.04226464516011e-05, "loss": 0.0535, "step": 24030 }, { "grad_norm": 0.44962427020072937, "learning_rate": 9.041291256566229e-05, "loss": 0.0656, "step": 24040 }, { "grad_norm": 0.5023583173751831, "learning_rate": 9.040317426022814e-05, "loss": 0.069, "step": 24050 }, { "grad_norm": 0.45050132274627686, "learning_rate": 9.03934315363636e-05, "loss": 0.059, "step": 24060 }, { "grad_norm": 0.5879310965538025, "learning_rate": 9.038368439513409e-05, "loss": 0.0525, "step": 24070 }, { "grad_norm": 0.6897771954536438, "learning_rate": 9.03739328376056e-05, "loss": 0.0758, "step": 24080 }, { "grad_norm": 0.6156973838806152, "learning_rate": 9.036417686484451e-05, "loss": 0.0562, "step": 24090 }, { "grad_norm": 0.4094890356063843, "learning_rate": 9.035441647791773e-05, "loss": 0.0575, "step": 24100 }, { "grad_norm": 0.5641431212425232, "learning_rate": 9.034465167789263e-05, "loss": 0.0482, "step": 24110 }, { "grad_norm": 0.6107102036476135, "learning_rate": 9.033488246583706e-05, "loss": 0.0505, "step": 24120 }, { "grad_norm": 0.3463740348815918, "learning_rate": 9.032510884281941e-05, "loss": 0.0723, "step": 24130 }, { "grad_norm": 0.49101588129997253, "learning_rate": 9.031533080990848e-05, "loss": 0.0633, "step": 24140 }, { "grad_norm": 0.43644076585769653, "learning_rate": 9.030554836817358e-05, "loss": 0.0572, "step": 24150 }, { "grad_norm": 0.8915787935256958, "learning_rate": 9.029576151868451e-05, "loss": 0.0702, "step": 24160 }, { "grad_norm": 0.7338671684265137, "learning_rate": 9.028597026251155e-05, "loss": 0.0647, "step": 24170 }, { "grad_norm": 0.605664074420929, "learning_rate": 9.027617460072547e-05, "loss": 0.0703, "step": 24180 }, { "grad_norm": 0.45829930901527405, "learning_rate": 9.026637453439745e-05, "loss": 0.0688, "step": 24190 }, { "grad_norm": 0.47971445322036743, "learning_rate": 9.025657006459927e-05, "loss": 0.0576, "step": 24200 }, { "grad_norm": 0.47322726249694824, "learning_rate": 9.024676119240311e-05, "loss": 0.0678, "step": 24210 }, { "grad_norm": 0.4133901000022888, "learning_rate": 9.023694791888166e-05, "loss": 0.0668, "step": 24220 }, { "grad_norm": 0.4907687306404114, "learning_rate": 9.022713024510808e-05, "loss": 0.081, "step": 24230 }, { "grad_norm": 0.4918828010559082, "learning_rate": 9.021730817215601e-05, "loss": 0.0538, "step": 24240 }, { "grad_norm": 0.44276687502861023, "learning_rate": 9.02074817010996e-05, "loss": 0.0539, "step": 24250 }, { "grad_norm": 0.4124127924442291, "learning_rate": 9.019765083301342e-05, "loss": 0.0502, "step": 24260 }, { "grad_norm": 0.5291175246238708, "learning_rate": 9.01878155689726e-05, "loss": 0.0572, "step": 24270 }, { "grad_norm": 0.6149037480354309, "learning_rate": 9.017797591005268e-05, "loss": 0.055, "step": 24280 }, { "grad_norm": 0.47335031628608704, "learning_rate": 9.016813185732972e-05, "loss": 0.0664, "step": 24290 }, { "grad_norm": 0.6336020231246948, "learning_rate": 9.015828341188027e-05, "loss": 0.0684, "step": 24300 }, { "grad_norm": 0.498104989528656, "learning_rate": 9.01484305747813e-05, "loss": 0.0601, "step": 24310 }, { "grad_norm": 0.4315156042575836, "learning_rate": 9.013857334711033e-05, "loss": 0.0568, "step": 24320 }, { "grad_norm": 0.8537513017654419, "learning_rate": 9.012871172994534e-05, "loss": 0.0611, "step": 24330 }, { "grad_norm": 0.454451322555542, "learning_rate": 9.011884572436476e-05, "loss": 0.0658, "step": 24340 }, { "grad_norm": 0.35533350706100464, "learning_rate": 9.010897533144754e-05, "loss": 0.0726, "step": 24350 }, { "grad_norm": 0.5727822780609131, "learning_rate": 9.009910055227306e-05, "loss": 0.0602, "step": 24360 }, { "grad_norm": 0.45199447870254517, "learning_rate": 9.008922138792124e-05, "loss": 0.0491, "step": 24370 }, { "grad_norm": 0.6659074425697327, "learning_rate": 9.007933783947244e-05, "loss": 0.0482, "step": 24380 }, { "grad_norm": 0.6366320252418518, "learning_rate": 9.006944990800752e-05, "loss": 0.0583, "step": 24390 }, { "grad_norm": 1.1278595924377441, "learning_rate": 9.005955759460779e-05, "loss": 0.0588, "step": 24400 }, { "grad_norm": 0.42616865038871765, "learning_rate": 9.004966090035508e-05, "loss": 0.0589, "step": 24410 }, { "grad_norm": 0.511139452457428, "learning_rate": 9.003975982633166e-05, "loss": 0.0605, "step": 24420 }, { "grad_norm": 0.45041561126708984, "learning_rate": 9.00298543736203e-05, "loss": 0.0681, "step": 24430 }, { "grad_norm": 0.42211681604385376, "learning_rate": 9.001994454330427e-05, "loss": 0.057, "step": 24440 }, { "grad_norm": 0.49189648032188416, "learning_rate": 9.001003033646727e-05, "loss": 0.0591, "step": 24450 }, { "grad_norm": 0.6555505394935608, "learning_rate": 9.00001117541935e-05, "loss": 0.0563, "step": 24460 }, { "grad_norm": 0.5787897706031799, "learning_rate": 8.999018879756764e-05, "loss": 0.0675, "step": 24470 }, { "grad_norm": 0.5101579427719116, "learning_rate": 8.998026146767487e-05, "loss": 0.0523, "step": 24480 }, { "grad_norm": 0.4517718553543091, "learning_rate": 8.99703297656008e-05, "loss": 0.0553, "step": 24490 }, { "grad_norm": 0.47451329231262207, "learning_rate": 8.996039369243156e-05, "loss": 0.0557, "step": 24500 }, { "grad_norm": 0.4669279158115387, "learning_rate": 8.995045324925378e-05, "loss": 0.0454, "step": 24510 }, { "grad_norm": 0.5867713093757629, "learning_rate": 8.994050843715448e-05, "loss": 0.0631, "step": 24520 }, { "grad_norm": 0.5402282476425171, "learning_rate": 8.993055925722121e-05, "loss": 0.0511, "step": 24530 }, { "grad_norm": 0.5143417119979858, "learning_rate": 8.992060571054202e-05, "loss": 0.0541, "step": 24540 }, { "grad_norm": 0.43766072392463684, "learning_rate": 8.991064779820542e-05, "loss": 0.069, "step": 24550 }, { "grad_norm": 0.37858638167381287, "learning_rate": 8.990068552130036e-05, "loss": 0.0478, "step": 24560 }, { "grad_norm": 1.1186965703964233, "learning_rate": 8.989071888091634e-05, "loss": 0.0527, "step": 24570 }, { "grad_norm": 0.5364306569099426, "learning_rate": 8.988074787814329e-05, "loss": 0.0508, "step": 24580 }, { "grad_norm": 0.5144123435020447, "learning_rate": 8.987077251407158e-05, "loss": 0.0535, "step": 24590 }, { "grad_norm": 0.5058667659759521, "learning_rate": 8.986079278979216e-05, "loss": 0.0565, "step": 24600 }, { "grad_norm": 0.4314633011817932, "learning_rate": 8.985080870639635e-05, "loss": 0.0688, "step": 24610 }, { "grad_norm": 0.4141463041305542, "learning_rate": 8.984082026497603e-05, "loss": 0.0473, "step": 24620 }, { "grad_norm": 0.5067434906959534, "learning_rate": 8.98308274666235e-05, "loss": 0.0724, "step": 24630 }, { "grad_norm": 0.47306522727012634, "learning_rate": 8.982083031243155e-05, "loss": 0.0544, "step": 24640 }, { "grad_norm": 0.530584990978241, "learning_rate": 8.98108288034935e-05, "loss": 0.0618, "step": 24650 }, { "grad_norm": 0.5080496668815613, "learning_rate": 8.980082294090305e-05, "loss": 0.0577, "step": 24660 }, { "grad_norm": 0.3902592658996582, "learning_rate": 8.979081272575443e-05, "loss": 0.0596, "step": 24670 }, { "grad_norm": 0.40471115708351135, "learning_rate": 8.978079815914236e-05, "loss": 0.0619, "step": 24680 }, { "grad_norm": 0.43859076499938965, "learning_rate": 8.977077924216202e-05, "loss": 0.0576, "step": 24690 }, { "grad_norm": 0.46582645177841187, "learning_rate": 8.976075597590905e-05, "loss": 0.0577, "step": 24700 }, { "grad_norm": 0.4767081141471863, "learning_rate": 8.975072836147958e-05, "loss": 0.0526, "step": 24710 }, { "grad_norm": 0.7129901051521301, "learning_rate": 8.974069639997025e-05, "loss": 0.054, "step": 24720 }, { "grad_norm": 0.5284965634346008, "learning_rate": 8.973066009247808e-05, "loss": 0.051, "step": 24730 }, { "grad_norm": 0.5111680626869202, "learning_rate": 8.972061944010066e-05, "loss": 0.0599, "step": 24740 }, { "grad_norm": 0.29907482862472534, "learning_rate": 8.971057444393603e-05, "loss": 0.0529, "step": 24750 }, { "grad_norm": 0.3196185231208801, "learning_rate": 8.970052510508268e-05, "loss": 0.066, "step": 24760 }, { "grad_norm": 0.6168971657752991, "learning_rate": 8.969047142463959e-05, "loss": 0.0623, "step": 24770 }, { "grad_norm": 0.5342128276824951, "learning_rate": 8.968041340370621e-05, "loss": 0.0575, "step": 24780 }, { "grad_norm": 0.8271695971488953, "learning_rate": 8.96703510433825e-05, "loss": 0.0494, "step": 24790 }, { "grad_norm": 0.4083079993724823, "learning_rate": 8.966028434476883e-05, "loss": 0.0501, "step": 24800 }, { "grad_norm": 0.41576826572418213, "learning_rate": 8.96502133089661e-05, "loss": 0.0458, "step": 24810 }, { "grad_norm": 0.3660695254802704, "learning_rate": 8.964013793707564e-05, "loss": 0.0512, "step": 24820 }, { "grad_norm": 0.3706781268119812, "learning_rate": 8.963005823019932e-05, "loss": 0.0537, "step": 24830 }, { "grad_norm": 0.49236294627189636, "learning_rate": 8.961997418943939e-05, "loss": 0.0496, "step": 24840 }, { "grad_norm": 0.5325479507446289, "learning_rate": 8.960988581589865e-05, "loss": 0.0539, "step": 24850 }, { "grad_norm": 0.42707645893096924, "learning_rate": 8.959979311068037e-05, "loss": 0.051, "step": 24860 }, { "grad_norm": 0.7314972877502441, "learning_rate": 8.958969607488823e-05, "loss": 0.0835, "step": 24870 }, { "grad_norm": 0.4118112027645111, "learning_rate": 8.957959470962647e-05, "loss": 0.0487, "step": 24880 }, { "grad_norm": 0.4464663863182068, "learning_rate": 8.956948901599971e-05, "loss": 0.0478, "step": 24890 }, { "grad_norm": 0.9923187494277954, "learning_rate": 8.955937899511315e-05, "loss": 0.0568, "step": 24900 }, { "grad_norm": 0.6920223832130432, "learning_rate": 8.954926464807238e-05, "loss": 0.0484, "step": 24910 }, { "grad_norm": 0.4061046540737152, "learning_rate": 8.953914597598347e-05, "loss": 0.0482, "step": 24920 }, { "grad_norm": 0.26118960976600647, "learning_rate": 8.952902297995303e-05, "loss": 0.0594, "step": 24930 }, { "grad_norm": 0.5902453064918518, "learning_rate": 8.951889566108804e-05, "loss": 0.0545, "step": 24940 }, { "grad_norm": 0.4558807611465454, "learning_rate": 8.950876402049606e-05, "loss": 0.055, "step": 24950 }, { "grad_norm": 0.4227079749107361, "learning_rate": 8.949862805928504e-05, "loss": 0.0751, "step": 24960 }, { "grad_norm": 0.41735780239105225, "learning_rate": 8.948848777856343e-05, "loss": 0.07, "step": 24970 }, { "grad_norm": 0.8576886653900146, "learning_rate": 8.947834317944017e-05, "loss": 0.0558, "step": 24980 }, { "grad_norm": 0.4201032221317291, "learning_rate": 8.946819426302466e-05, "loss": 0.0659, "step": 24990 }, { "grad_norm": 0.48473453521728516, "learning_rate": 8.945804103042676e-05, "loss": 0.0538, "step": 25000 }, { "grad_norm": 0.455221563577652, "learning_rate": 8.944788348275681e-05, "loss": 0.0472, "step": 25010 }, { "grad_norm": 0.5857715010643005, "learning_rate": 8.943772162112565e-05, "loss": 0.0534, "step": 25020 }, { "grad_norm": 0.4751618206501007, "learning_rate": 8.942755544664454e-05, "loss": 0.0714, "step": 25030 }, { "grad_norm": 0.47026169300079346, "learning_rate": 8.941738496042525e-05, "loss": 0.0493, "step": 25040 }, { "grad_norm": 0.34213846921920776, "learning_rate": 8.940721016357999e-05, "loss": 0.0508, "step": 25050 }, { "grad_norm": 1.510669469833374, "learning_rate": 8.939703105722148e-05, "loss": 0.056, "step": 25060 }, { "grad_norm": 0.45953163504600525, "learning_rate": 8.93868476424629e-05, "loss": 0.0521, "step": 25070 }, { "grad_norm": 0.9045236110687256, "learning_rate": 8.937665992041786e-05, "loss": 0.068, "step": 25080 }, { "grad_norm": 0.5556012392044067, "learning_rate": 8.93664678922005e-05, "loss": 0.0609, "step": 25090 }, { "grad_norm": 0.3412863612174988, "learning_rate": 8.93562715589254e-05, "loss": 0.0545, "step": 25100 }, { "grad_norm": 0.48142388463020325, "learning_rate": 8.934607092170762e-05, "loss": 0.0601, "step": 25110 }, { "grad_norm": 0.4897625744342804, "learning_rate": 8.933586598166266e-05, "loss": 0.0633, "step": 25120 }, { "grad_norm": 0.9122121334075928, "learning_rate": 8.932565673990655e-05, "loss": 0.0649, "step": 25130 }, { "grad_norm": 0.4255865514278412, "learning_rate": 8.931544319755574e-05, "loss": 0.0572, "step": 25140 }, { "grad_norm": 0.3779549300670624, "learning_rate": 8.930522535572718e-05, "loss": 0.0447, "step": 25150 }, { "grad_norm": 0.4070340394973755, "learning_rate": 8.929500321553826e-05, "loss": 0.0457, "step": 25160 }, { "grad_norm": 0.4598815143108368, "learning_rate": 8.928477677810686e-05, "loss": 0.0494, "step": 25170 }, { "grad_norm": 0.3799469769001007, "learning_rate": 8.927454604455137e-05, "loss": 0.0614, "step": 25180 }, { "grad_norm": 0.4003397822380066, "learning_rate": 8.926431101599053e-05, "loss": 0.0446, "step": 25190 }, { "grad_norm": 0.6492235660552979, "learning_rate": 8.925407169354369e-05, "loss": 0.0526, "step": 25200 }, { "grad_norm": 0.3612429201602936, "learning_rate": 8.92438280783306e-05, "loss": 0.0584, "step": 25210 }, { "grad_norm": 0.44945579767227173, "learning_rate": 8.923358017147146e-05, "loss": 0.0438, "step": 25220 }, { "grad_norm": 0.3758869469165802, "learning_rate": 8.922332797408697e-05, "loss": 0.0532, "step": 25230 }, { "grad_norm": 0.38153740763664246, "learning_rate": 8.921307148729831e-05, "loss": 0.0431, "step": 25240 }, { "grad_norm": 0.5298619866371155, "learning_rate": 8.920281071222712e-05, "loss": 0.0481, "step": 25250 }, { "grad_norm": 0.5685141086578369, "learning_rate": 8.919254564999548e-05, "loss": 0.05, "step": 25260 }, { "grad_norm": 1.3662081956863403, "learning_rate": 8.918227630172598e-05, "loss": 0.0562, "step": 25270 }, { "grad_norm": 0.24514992535114288, "learning_rate": 8.917200266854165e-05, "loss": 0.0454, "step": 25280 }, { "grad_norm": 0.7423809766769409, "learning_rate": 8.9161724751566e-05, "loss": 0.0512, "step": 25290 }, { "grad_norm": 0.3836148977279663, "learning_rate": 8.915144255192302e-05, "loss": 0.0584, "step": 25300 }, { "grad_norm": 0.4933170974254608, "learning_rate": 8.914115607073714e-05, "loss": 0.0558, "step": 25310 }, { "grad_norm": 0.8853304386138916, "learning_rate": 8.913086530913327e-05, "loss": 0.0533, "step": 25320 }, { "grad_norm": 0.38992196321487427, "learning_rate": 8.912057026823681e-05, "loss": 0.0578, "step": 25330 }, { "grad_norm": 0.41540321707725525, "learning_rate": 8.91102709491736e-05, "loss": 0.0526, "step": 25340 }, { "grad_norm": 0.4788714349269867, "learning_rate": 8.909996735306996e-05, "loss": 0.0582, "step": 25350 }, { "grad_norm": 0.3770676255226135, "learning_rate": 8.908965948105268e-05, "loss": 0.0545, "step": 25360 }, { "grad_norm": 0.5351041555404663, "learning_rate": 8.907934733424901e-05, "loss": 0.0567, "step": 25370 }, { "grad_norm": 0.7511116862297058, "learning_rate": 8.906903091378666e-05, "loss": 0.077, "step": 25380 }, { "grad_norm": 0.5818199515342712, "learning_rate": 8.905871022079384e-05, "loss": 0.0525, "step": 25390 }, { "grad_norm": 0.38335373997688293, "learning_rate": 8.90483852563992e-05, "loss": 0.0513, "step": 25400 }, { "grad_norm": 0.4038885235786438, "learning_rate": 8.903805602173185e-05, "loss": 0.0481, "step": 25410 }, { "grad_norm": 0.6369397044181824, "learning_rate": 8.902772251792137e-05, "loss": 0.0459, "step": 25420 }, { "grad_norm": 0.5125622153282166, "learning_rate": 8.901738474609786e-05, "loss": 0.0701, "step": 25430 }, { "grad_norm": 0.7029269933700562, "learning_rate": 8.900704270739179e-05, "loss": 0.0526, "step": 25440 }, { "grad_norm": 0.4316338300704956, "learning_rate": 8.89966964029342e-05, "loss": 0.0626, "step": 25450 }, { "grad_norm": 0.6135818958282471, "learning_rate": 8.898634583385652e-05, "loss": 0.0502, "step": 25460 }, { "grad_norm": 0.45593857765197754, "learning_rate": 8.897599100129065e-05, "loss": 0.055, "step": 25470 }, { "grad_norm": 0.5733987092971802, "learning_rate": 8.896563190636903e-05, "loss": 0.0549, "step": 25480 }, { "grad_norm": 0.4300867021083832, "learning_rate": 8.895526855022448e-05, "loss": 0.0487, "step": 25490 }, { "grad_norm": 0.4735119044780731, "learning_rate": 8.894490093399033e-05, "loss": 0.0599, "step": 25500 }, { "grad_norm": 0.3627884089946747, "learning_rate": 8.893452905880035e-05, "loss": 0.0518, "step": 25510 }, { "grad_norm": 0.6712827682495117, "learning_rate": 8.892415292578883e-05, "loss": 0.0666, "step": 25520 }, { "grad_norm": 0.4961587190628052, "learning_rate": 8.891377253609046e-05, "loss": 0.0571, "step": 25530 }, { "grad_norm": 0.43177926540374756, "learning_rate": 8.890338789084043e-05, "loss": 0.0536, "step": 25540 }, { "grad_norm": 0.49721917510032654, "learning_rate": 8.88929989911744e-05, "loss": 0.0543, "step": 25550 }, { "grad_norm": 0.5026703476905823, "learning_rate": 8.888260583822847e-05, "loss": 0.0502, "step": 25560 }, { "grad_norm": 0.4913904666900635, "learning_rate": 8.887220843313921e-05, "loss": 0.0525, "step": 25570 }, { "grad_norm": 0.4762500822544098, "learning_rate": 8.88618067770437e-05, "loss": 0.0555, "step": 25580 }, { "grad_norm": 0.7887136340141296, "learning_rate": 8.885140087107942e-05, "loss": 0.0472, "step": 25590 }, { "grad_norm": 0.5991580486297607, "learning_rate": 8.884099071638436e-05, "loss": 0.0504, "step": 25600 }, { "grad_norm": 0.407505065202713, "learning_rate": 8.883057631409695e-05, "loss": 0.056, "step": 25610 }, { "grad_norm": 0.4391649067401886, "learning_rate": 8.882015766535608e-05, "loss": 0.0531, "step": 25620 }, { "grad_norm": 0.4659122824668884, "learning_rate": 8.880973477130115e-05, "loss": 0.0595, "step": 25630 }, { "grad_norm": 0.7703647613525391, "learning_rate": 8.879930763307197e-05, "loss": 0.0512, "step": 25640 }, { "grad_norm": 0.5747919678688049, "learning_rate": 8.878887625180884e-05, "loss": 0.0605, "step": 25650 }, { "grad_norm": 0.4280780851840973, "learning_rate": 8.877844062865253e-05, "loss": 0.0433, "step": 25660 }, { "grad_norm": 0.4417617917060852, "learning_rate": 8.876800076474424e-05, "loss": 0.0641, "step": 25670 }, { "grad_norm": 0.4254712164402008, "learning_rate": 8.875755666122568e-05, "loss": 0.062, "step": 25680 }, { "grad_norm": 0.6625084280967712, "learning_rate": 8.8747108319239e-05, "loss": 0.0585, "step": 25690 }, { "grad_norm": 0.41948190331459045, "learning_rate": 8.87366557399268e-05, "loss": 0.0541, "step": 25700 }, { "grad_norm": 0.678746223449707, "learning_rate": 8.872619892443217e-05, "loss": 0.0529, "step": 25710 }, { "grad_norm": 0.509438157081604, "learning_rate": 8.871573787389865e-05, "loss": 0.0585, "step": 25720 }, { "grad_norm": 0.7057428359985352, "learning_rate": 8.870527258947024e-05, "loss": 0.0636, "step": 25730 }, { "grad_norm": 0.5527837872505188, "learning_rate": 8.869480307229143e-05, "loss": 0.0626, "step": 25740 }, { "grad_norm": 0.5248028635978699, "learning_rate": 8.868432932350712e-05, "loss": 0.0679, "step": 25750 }, { "grad_norm": 0.4945908784866333, "learning_rate": 8.867385134426272e-05, "loss": 0.0531, "step": 25760 }, { "grad_norm": 0.7462959885597229, "learning_rate": 8.866336913570407e-05, "loss": 0.0551, "step": 25770 }, { "grad_norm": 0.38456133008003235, "learning_rate": 8.865288269897751e-05, "loss": 0.052, "step": 25780 }, { "grad_norm": 0.40084120631217957, "learning_rate": 8.864239203522981e-05, "loss": 0.0592, "step": 25790 }, { "grad_norm": 0.46490925550460815, "learning_rate": 8.863189714560822e-05, "loss": 0.0458, "step": 25800 }, { "grad_norm": 0.4915704131126404, "learning_rate": 8.862139803126043e-05, "loss": 0.0543, "step": 25810 }, { "grad_norm": 0.45209431648254395, "learning_rate": 8.861089469333463e-05, "loss": 0.0569, "step": 25820 }, { "grad_norm": 1.036157488822937, "learning_rate": 8.860038713297944e-05, "loss": 0.0691, "step": 25830 }, { "grad_norm": 0.6083539128303528, "learning_rate": 8.858987535134394e-05, "loss": 0.0571, "step": 25840 }, { "grad_norm": 0.5846062898635864, "learning_rate": 8.857935934957769e-05, "loss": 0.0665, "step": 25850 }, { "grad_norm": 0.848327100276947, "learning_rate": 8.856883912883071e-05, "loss": 0.0483, "step": 25860 }, { "grad_norm": 0.32785820960998535, "learning_rate": 8.855831469025346e-05, "loss": 0.0527, "step": 25870 }, { "grad_norm": 0.5031456351280212, "learning_rate": 8.854778603499689e-05, "loss": 0.0656, "step": 25880 }, { "grad_norm": 0.5360159277915955, "learning_rate": 8.85372531642124e-05, "loss": 0.0596, "step": 25890 }, { "grad_norm": 0.9948400259017944, "learning_rate": 8.852671607905185e-05, "loss": 0.0528, "step": 25900 }, { "grad_norm": 0.5821824669837952, "learning_rate": 8.851617478066754e-05, "loss": 0.0634, "step": 25910 }, { "grad_norm": 0.5018247961997986, "learning_rate": 8.850562927021227e-05, "loss": 0.0465, "step": 25920 }, { "grad_norm": 0.5018559098243713, "learning_rate": 8.849507954883928e-05, "loss": 0.0521, "step": 25930 }, { "grad_norm": 1.1003515720367432, "learning_rate": 8.848452561770226e-05, "loss": 0.0466, "step": 25940 }, { "grad_norm": 0.9826477766036987, "learning_rate": 8.847396747795538e-05, "loss": 0.0613, "step": 25950 }, { "grad_norm": 0.5756570100784302, "learning_rate": 8.846340513075327e-05, "loss": 0.0541, "step": 25960 }, { "grad_norm": 0.6100037097930908, "learning_rate": 8.845283857725099e-05, "loss": 0.0627, "step": 25970 }, { "grad_norm": 0.5798276662826538, "learning_rate": 8.844226781860409e-05, "loss": 0.0685, "step": 25980 }, { "grad_norm": 0.6523046493530273, "learning_rate": 8.84316928559686e-05, "loss": 0.0852, "step": 25990 }, { "grad_norm": 0.5956154465675354, "learning_rate": 8.842111369050094e-05, "loss": 0.0617, "step": 26000 }, { "grad_norm": 0.8198310732841492, "learning_rate": 8.841053032335808e-05, "loss": 0.0897, "step": 26010 }, { "grad_norm": 0.4909387230873108, "learning_rate": 8.839994275569735e-05, "loss": 0.0569, "step": 26020 }, { "grad_norm": 0.8160286545753479, "learning_rate": 8.838935098867662e-05, "loss": 0.0611, "step": 26030 }, { "grad_norm": 0.7126194834709167, "learning_rate": 8.837875502345418e-05, "loss": 0.0636, "step": 26040 }, { "grad_norm": 0.3803582191467285, "learning_rate": 8.83681548611888e-05, "loss": 0.0673, "step": 26050 }, { "grad_norm": 0.44844937324523926, "learning_rate": 8.835755050303969e-05, "loss": 0.0554, "step": 26060 }, { "grad_norm": 0.34068813920021057, "learning_rate": 8.834694195016653e-05, "loss": 0.0562, "step": 26070 }, { "grad_norm": 0.47512802481651306, "learning_rate": 8.833632920372942e-05, "loss": 0.0767, "step": 26080 }, { "grad_norm": 0.6842356324195862, "learning_rate": 8.832571226488903e-05, "loss": 0.0652, "step": 26090 }, { "grad_norm": 0.47300291061401367, "learning_rate": 8.831509113480634e-05, "loss": 0.0566, "step": 26100 }, { "grad_norm": 0.49296459555625916, "learning_rate": 8.83044658146429e-05, "loss": 0.0454, "step": 26110 }, { "grad_norm": 0.424103707075119, "learning_rate": 8.829383630556067e-05, "loss": 0.0531, "step": 26120 }, { "grad_norm": 0.32432499527931213, "learning_rate": 8.828320260872207e-05, "loss": 0.0489, "step": 26130 }, { "grad_norm": 0.4555779993534088, "learning_rate": 8.827256472529e-05, "loss": 0.0547, "step": 26140 }, { "grad_norm": 0.527288556098938, "learning_rate": 8.826192265642778e-05, "loss": 0.0605, "step": 26150 }, { "grad_norm": 0.3777262270450592, "learning_rate": 8.825127640329923e-05, "loss": 0.0517, "step": 26160 }, { "grad_norm": 0.6272415518760681, "learning_rate": 8.824062596706861e-05, "loss": 0.0486, "step": 26170 }, { "grad_norm": 0.4009823203086853, "learning_rate": 8.822997134890062e-05, "loss": 0.0461, "step": 26180 }, { "grad_norm": 0.7110051512718201, "learning_rate": 8.821931254996044e-05, "loss": 0.0483, "step": 26190 }, { "grad_norm": 0.566008985042572, "learning_rate": 8.82086495714137e-05, "loss": 0.0494, "step": 26200 }, { "grad_norm": 0.5076749920845032, "learning_rate": 8.81979824144265e-05, "loss": 0.0637, "step": 26210 }, { "grad_norm": 0.43486425280570984, "learning_rate": 8.818731108016536e-05, "loss": 0.0583, "step": 26220 }, { "grad_norm": 0.47426798939704895, "learning_rate": 8.81766355697973e-05, "loss": 0.0549, "step": 26230 }, { "grad_norm": 0.4888053238391876, "learning_rate": 8.816595588448977e-05, "loss": 0.0436, "step": 26240 }, { "grad_norm": 0.45004695653915405, "learning_rate": 8.81552720254107e-05, "loss": 0.05, "step": 26250 }, { "grad_norm": 0.5156064629554749, "learning_rate": 8.814458399372842e-05, "loss": 0.0577, "step": 26260 }, { "grad_norm": 0.48251065611839294, "learning_rate": 8.813389179061181e-05, "loss": 0.0435, "step": 26270 }, { "grad_norm": 0.3256992995738983, "learning_rate": 8.812319541723012e-05, "loss": 0.0615, "step": 26280 }, { "grad_norm": 0.38141024112701416, "learning_rate": 8.811249487475309e-05, "loss": 0.0648, "step": 26290 }, { "grad_norm": 0.44790348410606384, "learning_rate": 8.810179016435092e-05, "loss": 0.078, "step": 26300 }, { "grad_norm": 0.4209040105342865, "learning_rate": 8.809108128719428e-05, "loss": 0.0684, "step": 26310 }, { "grad_norm": 0.7163590788841248, "learning_rate": 8.808036824445424e-05, "loss": 0.0578, "step": 26320 }, { "grad_norm": 0.5849977135658264, "learning_rate": 8.806965103730238e-05, "loss": 0.053, "step": 26330 }, { "grad_norm": 0.6297500133514404, "learning_rate": 8.805892966691074e-05, "loss": 0.0622, "step": 26340 }, { "grad_norm": 0.4380365312099457, "learning_rate": 8.804820413445175e-05, "loss": 0.0602, "step": 26350 }, { "grad_norm": 0.6604151129722595, "learning_rate": 8.803747444109837e-05, "loss": 0.0572, "step": 26360 }, { "grad_norm": 0.748039960861206, "learning_rate": 8.802674058802399e-05, "loss": 0.0538, "step": 26370 }, { "grad_norm": 0.534291684627533, "learning_rate": 8.801600257640241e-05, "loss": 0.0641, "step": 26380 }, { "grad_norm": 0.4933188855648041, "learning_rate": 8.800526040740795e-05, "loss": 0.0548, "step": 26390 }, { "grad_norm": 0.4202742576599121, "learning_rate": 8.799451408221535e-05, "loss": 0.0529, "step": 26400 }, { "grad_norm": 0.378729909658432, "learning_rate": 8.798376360199982e-05, "loss": 0.0472, "step": 26410 }, { "grad_norm": 0.7181846499443054, "learning_rate": 8.797300896793701e-05, "loss": 0.0514, "step": 26420 }, { "grad_norm": 0.39014989137649536, "learning_rate": 8.796225018120302e-05, "loss": 0.0604, "step": 26430 }, { "grad_norm": 0.7863737940788269, "learning_rate": 8.795148724297444e-05, "loss": 0.0693, "step": 26440 }, { "grad_norm": 0.4849780201911926, "learning_rate": 8.794072015442825e-05, "loss": 0.0604, "step": 26450 }, { "grad_norm": 0.6014844179153442, "learning_rate": 8.792994891674198e-05, "loss": 0.0594, "step": 26460 }, { "grad_norm": 0.6078366637229919, "learning_rate": 8.79191735310935e-05, "loss": 0.0461, "step": 26470 }, { "grad_norm": 0.5507506132125854, "learning_rate": 8.790839399866122e-05, "loss": 0.0639, "step": 26480 }, { "grad_norm": 0.4741585850715637, "learning_rate": 8.789761032062397e-05, "loss": 0.0612, "step": 26490 }, { "grad_norm": 0.5281417965888977, "learning_rate": 8.788682249816103e-05, "loss": 0.061, "step": 26500 }, { "grad_norm": 0.30276739597320557, "learning_rate": 8.787603053245215e-05, "loss": 0.0587, "step": 26510 }, { "grad_norm": 0.4056788384914398, "learning_rate": 8.78652344246775e-05, "loss": 0.0528, "step": 26520 }, { "grad_norm": 0.5310745239257812, "learning_rate": 8.785443417601776e-05, "loss": 0.0497, "step": 26530 }, { "grad_norm": 0.55991530418396, "learning_rate": 8.784362978765401e-05, "loss": 0.0545, "step": 26540 }, { "grad_norm": 0.45904576778411865, "learning_rate": 8.783282126076779e-05, "loss": 0.0509, "step": 26550 }, { "grad_norm": 0.6189858913421631, "learning_rate": 8.782200859654112e-05, "loss": 0.0548, "step": 26560 }, { "grad_norm": 0.570124089717865, "learning_rate": 8.781119179615646e-05, "loss": 0.0442, "step": 26570 }, { "grad_norm": 0.6696491837501526, "learning_rate": 8.780037086079674e-05, "loss": 0.057, "step": 26580 }, { "grad_norm": 0.4974271059036255, "learning_rate": 8.778954579164527e-05, "loss": 0.0493, "step": 26590 }, { "grad_norm": 0.5474729537963867, "learning_rate": 8.777871658988588e-05, "loss": 0.0601, "step": 26600 }, { "grad_norm": 0.49031564593315125, "learning_rate": 8.776788325670285e-05, "loss": 0.0467, "step": 26610 }, { "grad_norm": 0.49542126059532166, "learning_rate": 8.775704579328089e-05, "loss": 0.0643, "step": 26620 }, { "grad_norm": 0.5141925811767578, "learning_rate": 8.774620420080517e-05, "loss": 0.0614, "step": 26630 }, { "grad_norm": 0.5336389541625977, "learning_rate": 8.773535848046131e-05, "loss": 0.0509, "step": 26640 }, { "grad_norm": 0.5662606954574585, "learning_rate": 8.772450863343538e-05, "loss": 0.0522, "step": 26650 }, { "grad_norm": 0.48908209800720215, "learning_rate": 8.77136546609139e-05, "loss": 0.0566, "step": 26660 }, { "grad_norm": 0.49862560629844666, "learning_rate": 8.770279656408385e-05, "loss": 0.0543, "step": 26670 }, { "grad_norm": 0.5137081146240234, "learning_rate": 8.769193434413265e-05, "loss": 0.0616, "step": 26680 }, { "grad_norm": 0.4489526152610779, "learning_rate": 8.76810680022482e-05, "loss": 0.0541, "step": 26690 }, { "grad_norm": 0.5151548385620117, "learning_rate": 8.767019753961878e-05, "loss": 0.0646, "step": 26700 }, { "grad_norm": 0.5330237150192261, "learning_rate": 8.765932295743321e-05, "loss": 0.0538, "step": 26710 }, { "grad_norm": 0.42258089780807495, "learning_rate": 8.764844425688068e-05, "loss": 0.0581, "step": 26720 }, { "grad_norm": 0.45816412568092346, "learning_rate": 8.763756143915092e-05, "loss": 0.0454, "step": 26730 }, { "grad_norm": 0.5414748191833496, "learning_rate": 8.7626674505434e-05, "loss": 0.0688, "step": 26740 }, { "grad_norm": 0.6020222902297974, "learning_rate": 8.761578345692053e-05, "loss": 0.0612, "step": 26750 }, { "grad_norm": 0.5363080501556396, "learning_rate": 8.760488829480156e-05, "loss": 0.0559, "step": 26760 }, { "grad_norm": 0.41072389483451843, "learning_rate": 8.759398902026854e-05, "loss": 0.0582, "step": 26770 }, { "grad_norm": 0.3794947564601898, "learning_rate": 8.758308563451339e-05, "loss": 0.0644, "step": 26780 }, { "grad_norm": 0.5081952810287476, "learning_rate": 8.75721781387285e-05, "loss": 0.0614, "step": 26790 }, { "grad_norm": 0.5343404412269592, "learning_rate": 8.75612665341067e-05, "loss": 0.0698, "step": 26800 }, { "grad_norm": 0.5215017795562744, "learning_rate": 8.755035082184126e-05, "loss": 0.0499, "step": 26810 }, { "grad_norm": 0.4608495533466339, "learning_rate": 8.753943100312592e-05, "loss": 0.0462, "step": 26820 }, { "grad_norm": 0.4696837365627289, "learning_rate": 8.752850707915484e-05, "loss": 0.0483, "step": 26830 }, { "grad_norm": 0.5479635000228882, "learning_rate": 8.751757905112264e-05, "loss": 0.0624, "step": 26840 }, { "grad_norm": 0.7710681557655334, "learning_rate": 8.75066469202244e-05, "loss": 0.0654, "step": 26850 }, { "grad_norm": 0.45858243107795715, "learning_rate": 8.749571068765567e-05, "loss": 0.063, "step": 26860 }, { "grad_norm": 0.42827367782592773, "learning_rate": 8.748477035461238e-05, "loss": 0.058, "step": 26870 }, { "grad_norm": 0.47750231623649597, "learning_rate": 8.747382592229095e-05, "loss": 0.0568, "step": 26880 }, { "grad_norm": 0.3838111460208893, "learning_rate": 8.746287739188828e-05, "loss": 0.0463, "step": 26890 }, { "grad_norm": 0.47581177949905396, "learning_rate": 8.745192476460165e-05, "loss": 0.0509, "step": 26900 }, { "grad_norm": 2.776719570159912, "learning_rate": 8.744096804162882e-05, "loss": 0.0606, "step": 26910 }, { "grad_norm": 0.6902446746826172, "learning_rate": 8.743000722416804e-05, "loss": 0.07, "step": 26920 }, { "grad_norm": 0.5632504820823669, "learning_rate": 8.741904231341793e-05, "loss": 0.0549, "step": 26930 }, { "grad_norm": 0.5871655344963074, "learning_rate": 8.740807331057762e-05, "loss": 0.0667, "step": 26940 }, { "grad_norm": 0.4401579201221466, "learning_rate": 8.739710021684667e-05, "loss": 0.0586, "step": 26950 }, { "grad_norm": 0.42811819911003113, "learning_rate": 8.738612303342503e-05, "loss": 0.0533, "step": 26960 }, { "grad_norm": 0.34098201990127563, "learning_rate": 8.73751417615132e-05, "loss": 0.0547, "step": 26970 }, { "grad_norm": 0.6459508538246155, "learning_rate": 8.736415640231208e-05, "loss": 0.0615, "step": 26980 }, { "grad_norm": 0.4502764940261841, "learning_rate": 8.735316695702297e-05, "loss": 0.055, "step": 26990 }, { "grad_norm": 0.7139972448348999, "learning_rate": 8.734217342684769e-05, "loss": 0.0489, "step": 27000 }, { "grad_norm": 0.4763008654117584, "learning_rate": 8.733117581298847e-05, "loss": 0.0416, "step": 27010 }, { "grad_norm": 0.6031872034072876, "learning_rate": 8.732017411664796e-05, "loss": 0.0514, "step": 27020 }, { "grad_norm": 0.5241274833679199, "learning_rate": 8.730916833902936e-05, "loss": 0.0446, "step": 27030 }, { "grad_norm": 0.5228323936462402, "learning_rate": 8.729815848133618e-05, "loss": 0.0708, "step": 27040 }, { "grad_norm": 0.43856242299079895, "learning_rate": 8.728714454477247e-05, "loss": 0.0461, "step": 27050 }, { "grad_norm": 0.6823486685752869, "learning_rate": 8.727612653054269e-05, "loss": 0.0582, "step": 27060 }, { "grad_norm": 0.48861435055732727, "learning_rate": 8.726510443985176e-05, "loss": 0.0531, "step": 27070 }, { "grad_norm": 0.475299596786499, "learning_rate": 8.725407827390503e-05, "loss": 0.0425, "step": 27080 }, { "grad_norm": 0.3199794888496399, "learning_rate": 8.724304803390833e-05, "loss": 0.055, "step": 27090 }, { "grad_norm": 0.49635323882102966, "learning_rate": 8.723201372106788e-05, "loss": 0.0544, "step": 27100 }, { "grad_norm": 0.5824825167655945, "learning_rate": 8.722097533659038e-05, "loss": 0.0637, "step": 27110 }, { "grad_norm": 0.4274412989616394, "learning_rate": 8.720993288168299e-05, "loss": 0.0555, "step": 27120 }, { "grad_norm": 0.5677741765975952, "learning_rate": 8.719888635755327e-05, "loss": 0.0464, "step": 27130 }, { "grad_norm": 0.3966921269893646, "learning_rate": 8.718783576540928e-05, "loss": 0.0574, "step": 27140 }, { "grad_norm": 0.48971742391586304, "learning_rate": 8.717678110645948e-05, "loss": 0.0511, "step": 27150 }, { "grad_norm": 0.30966469645500183, "learning_rate": 8.716572238191279e-05, "loss": 0.0502, "step": 27160 }, { "grad_norm": 0.3943493068218231, "learning_rate": 8.715465959297857e-05, "loss": 0.0652, "step": 27170 }, { "grad_norm": 0.7851887941360474, "learning_rate": 8.714359274086665e-05, "loss": 0.0631, "step": 27180 }, { "grad_norm": 0.4406510293483734, "learning_rate": 8.713252182678726e-05, "loss": 0.0579, "step": 27190 }, { "grad_norm": 0.33311909437179565, "learning_rate": 8.712144685195112e-05, "loss": 0.0558, "step": 27200 }, { "grad_norm": 0.7023317217826843, "learning_rate": 8.711036781756936e-05, "loss": 0.0561, "step": 27210 }, { "grad_norm": 0.40381836891174316, "learning_rate": 8.709928472485357e-05, "loss": 0.0473, "step": 27220 }, { "grad_norm": 0.4804566204547882, "learning_rate": 8.708819757501579e-05, "loss": 0.0513, "step": 27230 }, { "grad_norm": 0.6283192038536072, "learning_rate": 8.707710636926846e-05, "loss": 0.0516, "step": 27240 }, { "grad_norm": 0.6408250331878662, "learning_rate": 8.706601110882455e-05, "loss": 0.0628, "step": 27250 }, { "grad_norm": 0.4497360289096832, "learning_rate": 8.705491179489738e-05, "loss": 0.0447, "step": 27260 }, { "grad_norm": 0.49387338757514954, "learning_rate": 8.704380842870077e-05, "loss": 0.0431, "step": 27270 }, { "grad_norm": 2.131840705871582, "learning_rate": 8.703270101144895e-05, "loss": 0.0497, "step": 27280 }, { "grad_norm": 0.31713640689849854, "learning_rate": 8.702158954435664e-05, "loss": 0.0477, "step": 27290 }, { "grad_norm": 0.35646089911460876, "learning_rate": 8.701047402863896e-05, "loss": 0.0471, "step": 27300 }, { "grad_norm": 0.4990808367729187, "learning_rate": 8.699935446551148e-05, "loss": 0.0541, "step": 27310 }, { "grad_norm": 0.4088737368583679, "learning_rate": 8.698823085619022e-05, "loss": 0.0495, "step": 27320 }, { "grad_norm": 0.3756863474845886, "learning_rate": 8.697710320189166e-05, "loss": 0.0411, "step": 27330 }, { "grad_norm": 0.5051238536834717, "learning_rate": 8.696597150383268e-05, "loss": 0.0592, "step": 27340 }, { "grad_norm": 0.3761393129825592, "learning_rate": 8.695483576323063e-05, "loss": 0.0496, "step": 27350 }, { "grad_norm": 0.4960467517375946, "learning_rate": 8.69436959813033e-05, "loss": 0.057, "step": 27360 }, { "grad_norm": 0.2262049913406372, "learning_rate": 8.693255215926892e-05, "loss": 0.0504, "step": 27370 }, { "grad_norm": 0.4282448887825012, "learning_rate": 8.692140429834617e-05, "loss": 0.0447, "step": 27380 }, { "grad_norm": 0.40093690156936646, "learning_rate": 8.691025239975415e-05, "loss": 0.0565, "step": 27390 }, { "grad_norm": 0.43833059072494507, "learning_rate": 8.689909646471243e-05, "loss": 0.0497, "step": 27400 }, { "grad_norm": 0.6638022661209106, "learning_rate": 8.688793649444099e-05, "loss": 0.0694, "step": 27410 }, { "grad_norm": 0.5306519865989685, "learning_rate": 8.687677249016029e-05, "loss": 0.0548, "step": 27420 }, { "grad_norm": 0.6627102494239807, "learning_rate": 8.686560445309118e-05, "loss": 0.059, "step": 27430 }, { "grad_norm": 0.4847365617752075, "learning_rate": 8.685443238445499e-05, "loss": 0.0584, "step": 27440 }, { "grad_norm": 0.47509947419166565, "learning_rate": 8.68432562854735e-05, "loss": 0.057, "step": 27450 }, { "grad_norm": 0.4613494575023651, "learning_rate": 8.683207615736887e-05, "loss": 0.0532, "step": 27460 }, { "grad_norm": 0.5429264307022095, "learning_rate": 8.682089200136379e-05, "loss": 0.0692, "step": 27470 }, { "grad_norm": 0.7888994216918945, "learning_rate": 8.680970381868132e-05, "loss": 0.0547, "step": 27480 }, { "grad_norm": 0.37752944231033325, "learning_rate": 8.679851161054498e-05, "loss": 0.0504, "step": 27490 }, { "grad_norm": 0.6686398386955261, "learning_rate": 8.678731537817873e-05, "loss": 0.0582, "step": 27500 }, { "grad_norm": 0.5180788040161133, "learning_rate": 8.677611512280697e-05, "loss": 0.0505, "step": 27510 }, { "grad_norm": 0.4399212896823883, "learning_rate": 8.676491084565457e-05, "loss": 0.0545, "step": 27520 }, { "grad_norm": 0.39392757415771484, "learning_rate": 8.675370254794678e-05, "loss": 0.0526, "step": 27530 }, { "grad_norm": 0.3702433109283447, "learning_rate": 8.674249023090935e-05, "loss": 0.0516, "step": 27540 }, { "grad_norm": 0.44891393184661865, "learning_rate": 8.673127389576843e-05, "loss": 0.0567, "step": 27550 }, { "grad_norm": 0.3538278639316559, "learning_rate": 8.67200535437506e-05, "loss": 0.0584, "step": 27560 }, { "grad_norm": 0.5727540850639343, "learning_rate": 8.670882917608296e-05, "loss": 0.0527, "step": 27570 }, { "grad_norm": 0.32676568627357483, "learning_rate": 8.669760079399292e-05, "loss": 0.0473, "step": 27580 }, { "grad_norm": 0.5858384370803833, "learning_rate": 8.668636839870845e-05, "loss": 0.0456, "step": 27590 }, { "grad_norm": 0.5047476291656494, "learning_rate": 8.667513199145789e-05, "loss": 0.0539, "step": 27600 }, { "grad_norm": 0.49799585342407227, "learning_rate": 8.666389157347002e-05, "loss": 0.0636, "step": 27610 }, { "grad_norm": 0.5131428837776184, "learning_rate": 8.66526471459741e-05, "loss": 0.0532, "step": 27620 }, { "grad_norm": 0.4202728867530823, "learning_rate": 8.66413987101998e-05, "loss": 0.0593, "step": 27630 }, { "grad_norm": 0.47580307722091675, "learning_rate": 8.663014626737723e-05, "loss": 0.0608, "step": 27640 }, { "grad_norm": 0.5105908513069153, "learning_rate": 8.661888981873691e-05, "loss": 0.0537, "step": 27650 }, { "grad_norm": 0.45044809579849243, "learning_rate": 8.660762936550988e-05, "loss": 0.0629, "step": 27660 }, { "grad_norm": 0.9930222034454346, "learning_rate": 8.659636490892753e-05, "loss": 0.0681, "step": 27670 }, { "grad_norm": 0.6212034821510315, "learning_rate": 8.658509645022174e-05, "loss": 0.057, "step": 27680 }, { "grad_norm": 0.5745781064033508, "learning_rate": 8.657382399062481e-05, "loss": 0.0512, "step": 27690 }, { "grad_norm": 0.5772159099578857, "learning_rate": 8.656254753136946e-05, "loss": 0.0491, "step": 27700 }, { "grad_norm": 0.6830604076385498, "learning_rate": 8.655126707368891e-05, "loss": 0.048, "step": 27710 }, { "grad_norm": 0.5650069713592529, "learning_rate": 8.653998261881672e-05, "loss": 0.0484, "step": 27720 }, { "grad_norm": 0.5138360261917114, "learning_rate": 8.652869416798699e-05, "loss": 0.0486, "step": 27730 }, { "grad_norm": 0.622210681438446, "learning_rate": 8.651740172243417e-05, "loss": 0.0532, "step": 27740 }, { "grad_norm": 1.2594350576400757, "learning_rate": 8.65061052833932e-05, "loss": 0.0578, "step": 27750 }, { "grad_norm": 0.35334837436676025, "learning_rate": 8.649480485209945e-05, "loss": 0.0486, "step": 27760 }, { "grad_norm": 0.5777584314346313, "learning_rate": 8.64835004297887e-05, "loss": 0.0692, "step": 27770 }, { "grad_norm": 0.40653082728385925, "learning_rate": 8.64721920176972e-05, "loss": 0.0558, "step": 27780 }, { "grad_norm": 0.5168054103851318, "learning_rate": 8.646087961706164e-05, "loss": 0.0652, "step": 27790 }, { "grad_norm": 0.5140433311462402, "learning_rate": 8.644956322911908e-05, "loss": 0.0532, "step": 27800 }, { "grad_norm": 0.6890414357185364, "learning_rate": 8.643824285510709e-05, "loss": 0.0556, "step": 27810 }, { "grad_norm": 0.31884312629699707, "learning_rate": 8.642691849626364e-05, "loss": 0.041, "step": 27820 }, { "grad_norm": 0.4649348258972168, "learning_rate": 8.641559015382717e-05, "loss": 0.0689, "step": 27830 }, { "grad_norm": 0.3751985728740692, "learning_rate": 8.640425782903649e-05, "loss": 0.0565, "step": 27840 }, { "grad_norm": 0.5938826203346252, "learning_rate": 8.639292152313091e-05, "loss": 0.0588, "step": 27850 }, { "grad_norm": 0.598077654838562, "learning_rate": 8.638158123735015e-05, "loss": 0.054, "step": 27860 }, { "grad_norm": 1.0494657754898071, "learning_rate": 8.637023697293436e-05, "loss": 0.0509, "step": 27870 }, { "grad_norm": 0.5657238364219666, "learning_rate": 8.635888873112414e-05, "loss": 0.0732, "step": 27880 }, { "grad_norm": 0.7159402370452881, "learning_rate": 8.634753651316052e-05, "loss": 0.0613, "step": 27890 }, { "grad_norm": 0.534035861492157, "learning_rate": 8.633618032028496e-05, "loss": 0.0485, "step": 27900 }, { "grad_norm": 0.49968746304512024, "learning_rate": 8.632482015373934e-05, "loss": 0.0559, "step": 27910 }, { "grad_norm": 0.567660927772522, "learning_rate": 8.6313456014766e-05, "loss": 0.0665, "step": 27920 }, { "grad_norm": 0.47864359617233276, "learning_rate": 8.630208790460771e-05, "loss": 0.0543, "step": 27930 }, { "grad_norm": 0.32915452122688293, "learning_rate": 8.629071582450768e-05, "loss": 0.0577, "step": 27940 }, { "grad_norm": 0.395405650138855, "learning_rate": 8.62793397757095e-05, "loss": 0.0553, "step": 27950 }, { "grad_norm": 0.6357736587524414, "learning_rate": 8.626795975945729e-05, "loss": 0.0583, "step": 27960 }, { "grad_norm": 0.3761804699897766, "learning_rate": 8.625657577699551e-05, "loss": 0.0496, "step": 27970 }, { "grad_norm": 0.5588232278823853, "learning_rate": 8.624518782956914e-05, "loss": 0.0633, "step": 27980 }, { "grad_norm": 0.7850347757339478, "learning_rate": 8.62337959184235e-05, "loss": 0.0561, "step": 27990 }, { "grad_norm": 0.5736287236213684, "learning_rate": 8.622240004480441e-05, "loss": 0.0614, "step": 28000 }, { "grad_norm": 1.1392611265182495, "learning_rate": 8.621100020995814e-05, "loss": 0.0538, "step": 28010 }, { "grad_norm": 0.6931791305541992, "learning_rate": 8.619959641513132e-05, "loss": 0.0654, "step": 28020 }, { "grad_norm": 0.6366913318634033, "learning_rate": 8.618818866157105e-05, "loss": 0.0691, "step": 28030 }, { "grad_norm": 0.38066044449806213, "learning_rate": 8.617677695052487e-05, "loss": 0.0518, "step": 28040 }, { "grad_norm": 0.578686535358429, "learning_rate": 8.616536128324078e-05, "loss": 0.0456, "step": 28050 }, { "grad_norm": 0.3791287839412689, "learning_rate": 8.615394166096712e-05, "loss": 0.0514, "step": 28060 }, { "grad_norm": 0.47580963373184204, "learning_rate": 8.614251808495279e-05, "loss": 0.0425, "step": 28070 }, { "grad_norm": 0.4370402991771698, "learning_rate": 8.6131090556447e-05, "loss": 0.0479, "step": 28080 }, { "grad_norm": 0.7729889750480652, "learning_rate": 8.611965907669947e-05, "loss": 0.0544, "step": 28090 }, { "grad_norm": 0.46479669213294983, "learning_rate": 8.610822364696034e-05, "loss": 0.054, "step": 28100 }, { "grad_norm": 0.5298226475715637, "learning_rate": 8.609678426848015e-05, "loss": 0.0468, "step": 28110 }, { "grad_norm": 0.457540363073349, "learning_rate": 8.60853409425099e-05, "loss": 0.0538, "step": 28120 }, { "grad_norm": 0.4648648798465729, "learning_rate": 8.607389367030104e-05, "loss": 0.0462, "step": 28130 }, { "grad_norm": 0.5695264935493469, "learning_rate": 8.606244245310538e-05, "loss": 0.0448, "step": 28140 }, { "grad_norm": 0.5845697522163391, "learning_rate": 8.605098729217525e-05, "loss": 0.0653, "step": 28150 }, { "grad_norm": 0.3613917827606201, "learning_rate": 8.603952818876335e-05, "loss": 0.0516, "step": 28160 }, { "grad_norm": 0.3302253782749176, "learning_rate": 8.602806514412281e-05, "loss": 0.0417, "step": 28170 }, { "grad_norm": 0.46865901350975037, "learning_rate": 8.601659815950726e-05, "loss": 0.0737, "step": 28180 }, { "grad_norm": 0.8210723400115967, "learning_rate": 8.600512723617067e-05, "loss": 0.0819, "step": 28190 }, { "grad_norm": 0.6369470357894897, "learning_rate": 8.59936523753675e-05, "loss": 0.0643, "step": 28200 }, { "grad_norm": 0.4748069941997528, "learning_rate": 8.598217357835264e-05, "loss": 0.0482, "step": 28210 }, { "grad_norm": 0.4025554955005646, "learning_rate": 8.597069084638135e-05, "loss": 0.0515, "step": 28220 }, { "grad_norm": 0.42909571528434753, "learning_rate": 8.595920418070939e-05, "loss": 0.056, "step": 28230 }, { "grad_norm": 0.4351378083229065, "learning_rate": 8.594771358259295e-05, "loss": 0.0604, "step": 28240 }, { "grad_norm": 0.5166282057762146, "learning_rate": 8.593621905328858e-05, "loss": 0.0745, "step": 28250 }, { "grad_norm": 0.4008626341819763, "learning_rate": 8.592472059405333e-05, "loss": 0.0547, "step": 28260 }, { "grad_norm": 0.631970226764679, "learning_rate": 8.591321820614464e-05, "loss": 0.0562, "step": 28270 }, { "grad_norm": 0.4906548261642456, "learning_rate": 8.590171189082041e-05, "loss": 0.0566, "step": 28280 }, { "grad_norm": 0.5570477843284607, "learning_rate": 8.589020164933894e-05, "loss": 0.063, "step": 28290 }, { "grad_norm": 0.5043908357620239, "learning_rate": 8.587868748295898e-05, "loss": 0.0485, "step": 28300 }, { "grad_norm": 0.42851415276527405, "learning_rate": 8.586716939293971e-05, "loss": 0.0526, "step": 28310 }, { "grad_norm": 0.440585732460022, "learning_rate": 8.58556473805407e-05, "loss": 0.0495, "step": 28320 }, { "grad_norm": 0.4291463792324066, "learning_rate": 8.584412144702202e-05, "loss": 0.0572, "step": 28330 }, { "grad_norm": 0.3329418897628784, "learning_rate": 8.58325915936441e-05, "loss": 0.0488, "step": 28340 }, { "grad_norm": 0.4528796672821045, "learning_rate": 8.582105782166783e-05, "loss": 0.0462, "step": 28350 }, { "grad_norm": 0.4939219057559967, "learning_rate": 8.580952013235455e-05, "loss": 0.0483, "step": 28360 }, { "grad_norm": 0.41825157403945923, "learning_rate": 8.579797852696596e-05, "loss": 0.0582, "step": 28370 }, { "grad_norm": 0.6980377435684204, "learning_rate": 8.578643300676428e-05, "loss": 0.0504, "step": 28380 }, { "grad_norm": 0.7048837542533875, "learning_rate": 8.577488357301209e-05, "loss": 0.0573, "step": 28390 }, { "grad_norm": 0.4241499900817871, "learning_rate": 8.576333022697242e-05, "loss": 0.0464, "step": 28400 }, { "grad_norm": 0.5961443185806274, "learning_rate": 8.575177296990873e-05, "loss": 0.0486, "step": 28410 }, { "grad_norm": 0.39380350708961487, "learning_rate": 8.574021180308489e-05, "loss": 0.0533, "step": 28420 }, { "grad_norm": 0.5115619897842407, "learning_rate": 8.572864672776523e-05, "loss": 0.0524, "step": 28430 }, { "grad_norm": 0.36652055382728577, "learning_rate": 8.571707774521447e-05, "loss": 0.0463, "step": 28440 }, { "grad_norm": 0.6368506550788879, "learning_rate": 8.57055048566978e-05, "loss": 0.0417, "step": 28450 }, { "grad_norm": 0.4460197389125824, "learning_rate": 8.569392806348078e-05, "loss": 0.0496, "step": 28460 }, { "grad_norm": 0.47268640995025635, "learning_rate": 8.568234736682947e-05, "loss": 0.0521, "step": 28470 }, { "grad_norm": 0.6417900323867798, "learning_rate": 8.567076276801029e-05, "loss": 0.0662, "step": 28480 }, { "grad_norm": 0.4472847878932953, "learning_rate": 8.565917426829013e-05, "loss": 0.0443, "step": 28490 }, { "grad_norm": 0.7326557636260986, "learning_rate": 8.564758186893628e-05, "loss": 0.0496, "step": 28500 }, { "grad_norm": 0.6567228436470032, "learning_rate": 8.563598557121649e-05, "loss": 0.0447, "step": 28510 }, { "grad_norm": 0.6878859400749207, "learning_rate": 8.562438537639888e-05, "loss": 0.0555, "step": 28520 }, { "grad_norm": 0.6011951565742493, "learning_rate": 8.561278128575206e-05, "loss": 0.0499, "step": 28530 }, { "grad_norm": 0.5090128183364868, "learning_rate": 8.5601173300545e-05, "loss": 0.0516, "step": 28540 }, { "grad_norm": 0.5454370379447937, "learning_rate": 8.558956142204717e-05, "loss": 0.0569, "step": 28550 }, { "grad_norm": 0.48799237608909607, "learning_rate": 8.55779456515284e-05, "loss": 0.0581, "step": 28560 }, { "grad_norm": 0.4821561872959137, "learning_rate": 8.556632599025898e-05, "loss": 0.0646, "step": 28570 }, { "grad_norm": 0.36412298679351807, "learning_rate": 8.555470243950964e-05, "loss": 0.0515, "step": 28580 }, { "grad_norm": 0.46650129556655884, "learning_rate": 8.554307500055148e-05, "loss": 0.0464, "step": 28590 }, { "grad_norm": 0.5691989660263062, "learning_rate": 8.553144367465609e-05, "loss": 0.0484, "step": 28600 }, { "grad_norm": 0.741927444934845, "learning_rate": 8.551980846309544e-05, "loss": 0.0653, "step": 28610 }, { "grad_norm": 0.4190835654735565, "learning_rate": 8.550816936714193e-05, "loss": 0.0625, "step": 28620 }, { "grad_norm": 0.31027543544769287, "learning_rate": 8.549652638806841e-05, "loss": 0.0491, "step": 28630 }, { "grad_norm": 0.3506554663181305, "learning_rate": 8.548487952714812e-05, "loss": 0.0528, "step": 28640 }, { "grad_norm": 0.8817445635795593, "learning_rate": 8.547322878565478e-05, "loss": 0.0557, "step": 28650 }, { "grad_norm": 0.5268308520317078, "learning_rate": 8.546157416486245e-05, "loss": 0.057, "step": 28660 }, { "grad_norm": 0.5139079689979553, "learning_rate": 8.54499156660457e-05, "loss": 0.0513, "step": 28670 }, { "grad_norm": 0.7772884368896484, "learning_rate": 8.543825329047947e-05, "loss": 0.0452, "step": 28680 }, { "grad_norm": 1.1160285472869873, "learning_rate": 8.542658703943913e-05, "loss": 0.0642, "step": 28690 }, { "grad_norm": 0.47128087282180786, "learning_rate": 8.541491691420051e-05, "loss": 0.059, "step": 28700 }, { "grad_norm": 0.49384498596191406, "learning_rate": 8.54032429160398e-05, "loss": 0.0565, "step": 28710 }, { "grad_norm": 1.0445834398269653, "learning_rate": 8.539156504623369e-05, "loss": 0.0605, "step": 28720 }, { "grad_norm": 0.45879828929901123, "learning_rate": 8.537988330605923e-05, "loss": 0.0674, "step": 28730 }, { "grad_norm": 0.4134170114994049, "learning_rate": 8.536819769679393e-05, "loss": 0.0426, "step": 28740 }, { "grad_norm": 0.555166482925415, "learning_rate": 8.53565082197157e-05, "loss": 0.0495, "step": 28750 }, { "grad_norm": 0.4399087131023407, "learning_rate": 8.534481487610289e-05, "loss": 0.0617, "step": 28760 }, { "grad_norm": 0.7281087636947632, "learning_rate": 8.533311766723428e-05, "loss": 0.052, "step": 28770 }, { "grad_norm": 0.4271465837955475, "learning_rate": 8.532141659438901e-05, "loss": 0.0476, "step": 28780 }, { "grad_norm": 0.4022626578807831, "learning_rate": 8.530971165884675e-05, "loss": 0.0453, "step": 28790 }, { "grad_norm": 0.607791543006897, "learning_rate": 8.529800286188752e-05, "loss": 0.0468, "step": 28800 }, { "grad_norm": 0.48413750529289246, "learning_rate": 8.528629020479175e-05, "loss": 0.0485, "step": 28810 }, { "grad_norm": 0.4224163889884949, "learning_rate": 8.527457368884033e-05, "loss": 0.0542, "step": 28820 }, { "grad_norm": 0.5028003454208374, "learning_rate": 8.526285331531458e-05, "loss": 0.0461, "step": 28830 }, { "grad_norm": 0.5028203725814819, "learning_rate": 8.525112908549621e-05, "loss": 0.0412, "step": 28840 }, { "grad_norm": 0.4875197112560272, "learning_rate": 8.523940100066735e-05, "loss": 0.0564, "step": 28850 }, { "grad_norm": 0.46347641944885254, "learning_rate": 8.52276690621106e-05, "loss": 0.0527, "step": 28860 }, { "grad_norm": 0.36792507767677307, "learning_rate": 8.521593327110889e-05, "loss": 0.0463, "step": 28870 }, { "grad_norm": 0.5249746441841125, "learning_rate": 8.520419362894569e-05, "loss": 0.0579, "step": 28880 }, { "grad_norm": 0.5057438015937805, "learning_rate": 8.51924501369048e-05, "loss": 0.0424, "step": 28890 }, { "grad_norm": 0.4090816080570221, "learning_rate": 8.518070279627047e-05, "loss": 0.0502, "step": 28900 }, { "grad_norm": 0.6691562533378601, "learning_rate": 8.516895160832737e-05, "loss": 0.0524, "step": 28910 }, { "grad_norm": 0.455770879983902, "learning_rate": 8.515719657436061e-05, "loss": 0.0461, "step": 28920 }, { "grad_norm": 0.38586875796318054, "learning_rate": 8.514543769565568e-05, "loss": 0.039, "step": 28930 }, { "grad_norm": 0.5584458708763123, "learning_rate": 8.513367497349853e-05, "loss": 0.0511, "step": 28940 }, { "grad_norm": 0.373473197221756, "learning_rate": 8.51219084091755e-05, "loss": 0.0521, "step": 28950 }, { "grad_norm": 0.3850267231464386, "learning_rate": 8.511013800397338e-05, "loss": 0.0452, "step": 28960 }, { "grad_norm": 0.4892539978027344, "learning_rate": 8.509836375917937e-05, "loss": 0.0574, "step": 28970 }, { "grad_norm": 0.429104208946228, "learning_rate": 8.508658567608104e-05, "loss": 0.0522, "step": 28980 }, { "grad_norm": 0.414623498916626, "learning_rate": 8.507480375596647e-05, "loss": 0.045, "step": 28990 }, { "grad_norm": 0.5972381830215454, "learning_rate": 8.506301800012408e-05, "loss": 0.0559, "step": 29000 }, { "grad_norm": 0.5396526455879211, "learning_rate": 8.505122840984278e-05, "loss": 0.0449, "step": 29010 }, { "grad_norm": 0.5364084839820862, "learning_rate": 8.503943498641182e-05, "loss": 0.048, "step": 29020 }, { "grad_norm": 0.6894177794456482, "learning_rate": 8.502763773112095e-05, "loss": 0.0653, "step": 29030 }, { "grad_norm": 0.906873881816864, "learning_rate": 8.501583664526026e-05, "loss": 0.0584, "step": 29040 }, { "grad_norm": 0.46406152844429016, "learning_rate": 8.500403173012032e-05, "loss": 0.0553, "step": 29050 }, { "grad_norm": 0.4716106653213501, "learning_rate": 8.499222298699211e-05, "loss": 0.0542, "step": 29060 }, { "grad_norm": 0.43360036611557007, "learning_rate": 8.498041041716701e-05, "loss": 0.0573, "step": 29070 }, { "grad_norm": 0.4490923285484314, "learning_rate": 8.496859402193681e-05, "loss": 0.0456, "step": 29080 }, { "grad_norm": 0.3867252767086029, "learning_rate": 8.495677380259374e-05, "loss": 0.0539, "step": 29090 }, { "grad_norm": 0.5383354425430298, "learning_rate": 8.494494976043045e-05, "loss": 0.069, "step": 29100 }, { "grad_norm": 0.4381449818611145, "learning_rate": 8.493312189673998e-05, "loss": 0.0634, "step": 29110 }, { "grad_norm": 0.6419572830200195, "learning_rate": 8.492129021281584e-05, "loss": 0.0506, "step": 29120 }, { "grad_norm": 0.5429107546806335, "learning_rate": 8.490945470995188e-05, "loss": 0.055, "step": 29130 }, { "grad_norm": 0.5785916447639465, "learning_rate": 8.489761538944247e-05, "loss": 0.0516, "step": 29140 }, { "grad_norm": 0.614651620388031, "learning_rate": 8.48857722525823e-05, "loss": 0.053, "step": 29150 }, { "grad_norm": 0.4171825349330902, "learning_rate": 8.487392530066652e-05, "loss": 0.0617, "step": 29160 }, { "grad_norm": 0.6550410389900208, "learning_rate": 8.486207453499069e-05, "loss": 0.0671, "step": 29170 }, { "grad_norm": 0.6222624778747559, "learning_rate": 8.485021995685082e-05, "loss": 0.0532, "step": 29180 }, { "grad_norm": 0.4144599437713623, "learning_rate": 8.483836156754328e-05, "loss": 0.0547, "step": 29190 }, { "grad_norm": 0.3683219254016876, "learning_rate": 8.482649936836491e-05, "loss": 0.0538, "step": 29200 }, { "grad_norm": 0.434430867433548, "learning_rate": 8.481463336061293e-05, "loss": 0.0501, "step": 29210 }, { "grad_norm": 0.46298038959503174, "learning_rate": 8.480276354558496e-05, "loss": 0.0589, "step": 29220 }, { "grad_norm": 0.5856636762619019, "learning_rate": 8.479088992457913e-05, "loss": 0.0534, "step": 29230 }, { "grad_norm": 0.48042166233062744, "learning_rate": 8.477901249889387e-05, "loss": 0.0572, "step": 29240 }, { "grad_norm": 1.1985112428665161, "learning_rate": 8.47671312698281e-05, "loss": 0.039, "step": 29250 }, { "grad_norm": 0.34331566095352173, "learning_rate": 8.475524623868112e-05, "loss": 0.0467, "step": 29260 }, { "grad_norm": 0.5096166133880615, "learning_rate": 8.474335740675266e-05, "loss": 0.0572, "step": 29270 }, { "grad_norm": 0.5860372185707092, "learning_rate": 8.473146477534289e-05, "loss": 0.0517, "step": 29280 }, { "grad_norm": 0.5770675539970398, "learning_rate": 8.471956834575232e-05, "loss": 0.0531, "step": 29290 }, { "grad_norm": 0.8839508295059204, "learning_rate": 8.470766811928197e-05, "loss": 0.0577, "step": 29300 }, { "grad_norm": 0.46946772933006287, "learning_rate": 8.469576409723323e-05, "loss": 0.0541, "step": 29310 }, { "grad_norm": 0.4365536868572235, "learning_rate": 8.468385628090788e-05, "loss": 0.0437, "step": 29320 }, { "grad_norm": 0.4379693567752838, "learning_rate": 8.467194467160815e-05, "loss": 0.0509, "step": 29330 }, { "grad_norm": 0.6514460444450378, "learning_rate": 8.466002927063667e-05, "loss": 0.0524, "step": 29340 }, { "grad_norm": 0.4765911400318146, "learning_rate": 8.464811007929651e-05, "loss": 0.059, "step": 29350 }, { "grad_norm": 0.4555410146713257, "learning_rate": 8.463618709889114e-05, "loss": 0.0586, "step": 29360 }, { "grad_norm": 0.4246673583984375, "learning_rate": 8.462426033072442e-05, "loss": 0.0509, "step": 29370 }, { "grad_norm": 0.41257691383361816, "learning_rate": 8.461232977610061e-05, "loss": 0.0507, "step": 29380 }, { "grad_norm": 0.93895024061203, "learning_rate": 8.46003954363245e-05, "loss": 0.0488, "step": 29390 }, { "grad_norm": 0.5717213153839111, "learning_rate": 8.458845731270115e-05, "loss": 0.0558, "step": 29400 }, { "grad_norm": 0.45412030816078186, "learning_rate": 8.45765154065361e-05, "loss": 0.045, "step": 29410 }, { "grad_norm": 0.3272949457168579, "learning_rate": 8.456456971913532e-05, "loss": 0.0483, "step": 29420 }, { "grad_norm": 0.37770697474479675, "learning_rate": 8.455262025180517e-05, "loss": 0.049, "step": 29430 }, { "grad_norm": 0.43661224842071533, "learning_rate": 8.454066700585242e-05, "loss": 0.0523, "step": 29440 }, { "grad_norm": 0.45821842551231384, "learning_rate": 8.452870998258423e-05, "loss": 0.0384, "step": 29450 }, { "grad_norm": 0.3192391097545624, "learning_rate": 8.451674918330825e-05, "loss": 0.0423, "step": 29460 }, { "grad_norm": 0.5622149109840393, "learning_rate": 8.450478460933246e-05, "loss": 0.0526, "step": 29470 }, { "grad_norm": 0.511567234992981, "learning_rate": 8.449281626196532e-05, "loss": 0.0533, "step": 29480 }, { "grad_norm": 0.5653937458992004, "learning_rate": 8.448084414251564e-05, "loss": 0.0534, "step": 29490 }, { "grad_norm": 0.30269140005111694, "learning_rate": 8.446886825229271e-05, "loss": 0.0504, "step": 29500 }, { "grad_norm": 0.4468104839324951, "learning_rate": 8.445688859260615e-05, "loss": 0.0426, "step": 29510 }, { "grad_norm": 0.516721248626709, "learning_rate": 8.444490516476606e-05, "loss": 0.0457, "step": 29520 }, { "grad_norm": 0.3737834692001343, "learning_rate": 8.443291797008293e-05, "loss": 0.0543, "step": 29530 }, { "grad_norm": 0.40075474977493286, "learning_rate": 8.442092700986765e-05, "loss": 0.0562, "step": 29540 }, { "grad_norm": 0.3797427713871002, "learning_rate": 8.440893228543156e-05, "loss": 0.0553, "step": 29550 }, { "grad_norm": 0.5147292017936707, "learning_rate": 8.439693379808638e-05, "loss": 0.0443, "step": 29560 }, { "grad_norm": 0.4163926839828491, "learning_rate": 8.43849315491442e-05, "loss": 0.0525, "step": 29570 }, { "grad_norm": 0.7370895147323608, "learning_rate": 8.437292553991763e-05, "loss": 0.0568, "step": 29580 }, { "grad_norm": 0.43965327739715576, "learning_rate": 8.436091577171959e-05, "loss": 0.043, "step": 29590 }, { "grad_norm": 0.45111435651779175, "learning_rate": 8.434890224586347e-05, "loss": 0.0363, "step": 29600 }, { "grad_norm": 0.5085189938545227, "learning_rate": 8.433688496366303e-05, "loss": 0.0404, "step": 29610 }, { "grad_norm": 0.46174511313438416, "learning_rate": 8.432486392643248e-05, "loss": 0.0477, "step": 29620 }, { "grad_norm": 0.46904224157333374, "learning_rate": 8.431283913548643e-05, "loss": 0.0488, "step": 29630 }, { "grad_norm": 0.4975849986076355, "learning_rate": 8.430081059213985e-05, "loss": 0.0478, "step": 29640 }, { "grad_norm": 0.4234677851200104, "learning_rate": 8.428877829770823e-05, "loss": 0.0451, "step": 29650 }, { "grad_norm": 0.42509782314300537, "learning_rate": 8.427674225350735e-05, "loss": 0.0693, "step": 29660 }, { "grad_norm": 0.7976384162902832, "learning_rate": 8.426470246085347e-05, "loss": 0.054, "step": 29670 }, { "grad_norm": 0.4991353452205658, "learning_rate": 8.425265892106324e-05, "loss": 0.0462, "step": 29680 }, { "grad_norm": 0.4904203712940216, "learning_rate": 8.424061163545374e-05, "loss": 0.0509, "step": 29690 }, { "grad_norm": 0.4109917879104614, "learning_rate": 8.422856060534243e-05, "loss": 0.0535, "step": 29700 }, { "grad_norm": 0.6869656443595886, "learning_rate": 8.421650583204718e-05, "loss": 0.0664, "step": 29710 }, { "grad_norm": 1.1210801601409912, "learning_rate": 8.420444731688633e-05, "loss": 0.0495, "step": 29720 }, { "grad_norm": 0.4227503538131714, "learning_rate": 8.419238506117852e-05, "loss": 0.0647, "step": 29730 }, { "grad_norm": 0.5889717936515808, "learning_rate": 8.418031906624289e-05, "loss": 0.0516, "step": 29740 }, { "grad_norm": 0.49797049164772034, "learning_rate": 8.416824933339898e-05, "loss": 0.0496, "step": 29750 }, { "grad_norm": 0.32241663336753845, "learning_rate": 8.415617586396667e-05, "loss": 0.0401, "step": 29760 }, { "grad_norm": 0.2648395895957947, "learning_rate": 8.414409865926632e-05, "loss": 0.0503, "step": 29770 }, { "grad_norm": 0.34334129095077515, "learning_rate": 8.413201772061867e-05, "loss": 0.043, "step": 29780 }, { "grad_norm": 0.33254438638687134, "learning_rate": 8.411993304934488e-05, "loss": 0.043, "step": 29790 }, { "grad_norm": 0.9210594296455383, "learning_rate": 8.410784464676654e-05, "loss": 0.0508, "step": 29800 }, { "grad_norm": 0.32125988602638245, "learning_rate": 8.409575251420556e-05, "loss": 0.0454, "step": 29810 }, { "grad_norm": 0.3635045289993286, "learning_rate": 8.408365665298435e-05, "loss": 0.0561, "step": 29820 }, { "grad_norm": 0.7329310178756714, "learning_rate": 8.40715570644257e-05, "loss": 0.0639, "step": 29830 }, { "grad_norm": 0.8416972160339355, "learning_rate": 8.40594537498528e-05, "loss": 0.061, "step": 29840 }, { "grad_norm": 0.43702128529548645, "learning_rate": 8.404734671058924e-05, "loss": 0.0544, "step": 29850 }, { "grad_norm": 0.7463392019271851, "learning_rate": 8.403523594795902e-05, "loss": 0.0577, "step": 29860 }, { "grad_norm": 1.0293965339660645, "learning_rate": 8.402312146328659e-05, "loss": 0.0538, "step": 29870 }, { "grad_norm": 0.5950551629066467, "learning_rate": 8.401100325789675e-05, "loss": 0.0551, "step": 29880 }, { "grad_norm": 0.68710857629776, "learning_rate": 8.399888133311472e-05, "loss": 0.0375, "step": 29890 }, { "grad_norm": 0.6048406362533569, "learning_rate": 8.398675569026613e-05, "loss": 0.0634, "step": 29900 }, { "grad_norm": 0.5120473504066467, "learning_rate": 8.397462633067705e-05, "loss": 0.0469, "step": 29910 }, { "grad_norm": 0.5279939770698547, "learning_rate": 8.396249325567392e-05, "loss": 0.0412, "step": 29920 }, { "grad_norm": 0.3836723268032074, "learning_rate": 8.395035646658357e-05, "loss": 0.0585, "step": 29930 }, { "grad_norm": 0.3356046676635742, "learning_rate": 8.39382159647333e-05, "loss": 0.057, "step": 29940 }, { "grad_norm": 0.42212599515914917, "learning_rate": 8.392607175145075e-05, "loss": 0.0484, "step": 29950 }, { "grad_norm": 0.5075567364692688, "learning_rate": 8.3913923828064e-05, "loss": 0.0458, "step": 29960 }, { "grad_norm": 0.4859171211719513, "learning_rate": 8.390177219590152e-05, "loss": 0.0531, "step": 29970 }, { "grad_norm": 0.5711663961410522, "learning_rate": 8.388961685629222e-05, "loss": 0.0509, "step": 29980 }, { "grad_norm": 0.45250681042671204, "learning_rate": 8.387745781056536e-05, "loss": 0.0526, "step": 29990 }, { "grad_norm": 0.8829255104064941, "learning_rate": 8.386529506005065e-05, "loss": 0.0456, "step": 30000 }, { "grad_norm": 0.41064053773880005, "learning_rate": 8.38531286060782e-05, "loss": 0.048, "step": 30010 }, { "grad_norm": 1.5371898412704468, "learning_rate": 8.384095844997849e-05, "loss": 0.0504, "step": 30020 }, { "grad_norm": 0.49219027161598206, "learning_rate": 8.382878459308245e-05, "loss": 0.0525, "step": 30030 }, { "grad_norm": 0.7565158605575562, "learning_rate": 8.381660703672138e-05, "loss": 0.0641, "step": 30040 }, { "grad_norm": 0.6612686514854431, "learning_rate": 8.380442578222702e-05, "loss": 0.0714, "step": 30050 }, { "grad_norm": 0.3241236209869385, "learning_rate": 8.379224083093146e-05, "loss": 0.048, "step": 30060 }, { "grad_norm": 0.39237886667251587, "learning_rate": 8.378005218416727e-05, "loss": 0.0567, "step": 30070 }, { "grad_norm": 0.5772407054901123, "learning_rate": 8.376785984326735e-05, "loss": 0.0468, "step": 30080 }, { "grad_norm": 0.447674959897995, "learning_rate": 8.375566380956506e-05, "loss": 0.0488, "step": 30090 }, { "grad_norm": 0.5334745645523071, "learning_rate": 8.374346408439411e-05, "loss": 0.0579, "step": 30100 }, { "grad_norm": 0.6569058299064636, "learning_rate": 8.373126066908868e-05, "loss": 0.0609, "step": 30110 }, { "grad_norm": 0.5516849160194397, "learning_rate": 8.371905356498326e-05, "loss": 0.0553, "step": 30120 }, { "grad_norm": 0.41753119230270386, "learning_rate": 8.370684277341288e-05, "loss": 0.0473, "step": 30130 }, { "grad_norm": 0.6363565921783447, "learning_rate": 8.369462829571282e-05, "loss": 0.0514, "step": 30140 }, { "grad_norm": 0.5073820352554321, "learning_rate": 8.36824101332189e-05, "loss": 0.0469, "step": 30150 }, { "grad_norm": 0.41006115078926086, "learning_rate": 8.367018828726721e-05, "loss": 0.0528, "step": 30160 }, { "grad_norm": 0.4071946442127228, "learning_rate": 8.365796275919438e-05, "loss": 0.0521, "step": 30170 }, { "grad_norm": 0.43571794033050537, "learning_rate": 8.364573355033734e-05, "loss": 0.0517, "step": 30180 }, { "grad_norm": 0.7948068976402283, "learning_rate": 8.363350066203346e-05, "loss": 0.0592, "step": 30190 }, { "grad_norm": 0.422911673784256, "learning_rate": 8.362126409562053e-05, "loss": 0.0513, "step": 30200 }, { "grad_norm": 0.5310256481170654, "learning_rate": 8.360902385243667e-05, "loss": 0.0579, "step": 30210 }, { "grad_norm": 0.5226840972900391, "learning_rate": 8.359677993382052e-05, "loss": 0.049, "step": 30220 }, { "grad_norm": 0.6039405465126038, "learning_rate": 8.358453234111103e-05, "loss": 0.0603, "step": 30230 }, { "grad_norm": 0.4307592511177063, "learning_rate": 8.357228107564756e-05, "loss": 0.0618, "step": 30240 }, { "grad_norm": 0.535192608833313, "learning_rate": 8.356002613876993e-05, "loss": 0.0444, "step": 30250 }, { "grad_norm": 0.526844322681427, "learning_rate": 8.35477675318183e-05, "loss": 0.0457, "step": 30260 }, { "grad_norm": 0.43220606446266174, "learning_rate": 8.353550525613323e-05, "loss": 0.0582, "step": 30270 }, { "grad_norm": 0.5376718044281006, "learning_rate": 8.352323931305572e-05, "loss": 0.0606, "step": 30280 }, { "grad_norm": 0.3013681173324585, "learning_rate": 8.351096970392717e-05, "loss": 0.0442, "step": 30290 }, { "grad_norm": 0.4096112847328186, "learning_rate": 8.349869643008937e-05, "loss": 0.0466, "step": 30300 }, { "grad_norm": 0.4765174090862274, "learning_rate": 8.348641949288449e-05, "loss": 0.0573, "step": 30310 }, { "grad_norm": 0.38980942964553833, "learning_rate": 8.34741388936551e-05, "loss": 0.0501, "step": 30320 }, { "grad_norm": 0.6391616463661194, "learning_rate": 8.346185463374423e-05, "loss": 0.0515, "step": 30330 }, { "grad_norm": 0.43884900212287903, "learning_rate": 8.344956671449524e-05, "loss": 0.0508, "step": 30340 }, { "grad_norm": 0.335802286863327, "learning_rate": 8.343727513725192e-05, "loss": 0.0382, "step": 30350 }, { "grad_norm": 0.449452668428421, "learning_rate": 8.342497990335847e-05, "loss": 0.0529, "step": 30360 }, { "grad_norm": 0.5448156595230103, "learning_rate": 8.341268101415946e-05, "loss": 0.067, "step": 30370 }, { "grad_norm": 0.4641931354999542, "learning_rate": 8.34003784709999e-05, "loss": 0.0491, "step": 30380 }, { "grad_norm": 0.5283324718475342, "learning_rate": 8.338807227522516e-05, "loss": 0.0587, "step": 30390 }, { "grad_norm": 0.5973585844039917, "learning_rate": 8.337576242818103e-05, "loss": 0.0497, "step": 30400 }, { "grad_norm": 0.626746416091919, "learning_rate": 8.336344893121372e-05, "loss": 0.0517, "step": 30410 }, { "grad_norm": 0.5316136479377747, "learning_rate": 8.335113178566977e-05, "loss": 0.056, "step": 30420 }, { "grad_norm": 0.41074642539024353, "learning_rate": 8.33388109928962e-05, "loss": 0.0638, "step": 30430 }, { "grad_norm": 0.443188339471817, "learning_rate": 8.33264865542404e-05, "loss": 0.0487, "step": 30440 }, { "grad_norm": 0.3823726177215576, "learning_rate": 8.331415847105013e-05, "loss": 0.0401, "step": 30450 }, { "grad_norm": 0.5159330368041992, "learning_rate": 8.330182674467357e-05, "loss": 0.0506, "step": 30460 }, { "grad_norm": 0.46860814094543457, "learning_rate": 8.32894913764593e-05, "loss": 0.039, "step": 30470 }, { "grad_norm": 0.5641918778419495, "learning_rate": 8.327715236775633e-05, "loss": 0.05, "step": 30480 }, { "grad_norm": 0.3956620991230011, "learning_rate": 8.326480971991398e-05, "loss": 0.0476, "step": 30490 }, { "grad_norm": 0.365348756313324, "learning_rate": 8.325246343428206e-05, "loss": 0.0549, "step": 30500 }, { "grad_norm": 0.31574615836143494, "learning_rate": 8.324011351221072e-05, "loss": 0.0538, "step": 30510 }, { "grad_norm": 0.5639433860778809, "learning_rate": 8.322775995505057e-05, "loss": 0.0498, "step": 30520 }, { "grad_norm": 0.837437629699707, "learning_rate": 8.321540276415254e-05, "loss": 0.0542, "step": 30530 }, { "grad_norm": 0.7736146450042725, "learning_rate": 8.320304194086798e-05, "loss": 0.0544, "step": 30540 }, { "grad_norm": 0.4938196837902069, "learning_rate": 8.31906774865487e-05, "loss": 0.0483, "step": 30550 }, { "grad_norm": 0.45269203186035156, "learning_rate": 8.317830940254682e-05, "loss": 0.048, "step": 30560 }, { "grad_norm": 0.49889177083969116, "learning_rate": 8.316593769021491e-05, "loss": 0.0591, "step": 30570 }, { "grad_norm": 0.49942901730537415, "learning_rate": 8.315356235090592e-05, "loss": 0.0586, "step": 30580 }, { "grad_norm": 0.5138787627220154, "learning_rate": 8.314118338597319e-05, "loss": 0.0473, "step": 30590 }, { "grad_norm": 0.4488034248352051, "learning_rate": 8.312880079677048e-05, "loss": 0.0541, "step": 30600 }, { "grad_norm": 0.9206785559654236, "learning_rate": 8.311641458465191e-05, "loss": 0.0516, "step": 30610 }, { "grad_norm": 0.5042867064476013, "learning_rate": 8.310402475097205e-05, "loss": 0.0638, "step": 30620 }, { "grad_norm": 0.40208473801612854, "learning_rate": 8.309163129708581e-05, "loss": 0.0432, "step": 30630 }, { "grad_norm": 0.472099632024765, "learning_rate": 8.307923422434852e-05, "loss": 0.0404, "step": 30640 }, { "grad_norm": 0.7944864630699158, "learning_rate": 8.30668335341159e-05, "loss": 0.0537, "step": 30650 }, { "grad_norm": 0.4838303327560425, "learning_rate": 8.305442922774408e-05, "loss": 0.0548, "step": 30660 }, { "grad_norm": 0.2609967589378357, "learning_rate": 8.304202130658959e-05, "loss": 0.0471, "step": 30670 }, { "grad_norm": 0.47800615429878235, "learning_rate": 8.302960977200931e-05, "loss": 0.0488, "step": 30680 }, { "grad_norm": 0.6116860508918762, "learning_rate": 8.301719462536058e-05, "loss": 0.0499, "step": 30690 }, { "grad_norm": 0.45199528336524963, "learning_rate": 8.300477586800108e-05, "loss": 0.0452, "step": 30700 }, { "grad_norm": 0.43717020750045776, "learning_rate": 8.299235350128892e-05, "loss": 0.0486, "step": 30710 }, { "grad_norm": 0.46880993247032166, "learning_rate": 8.297992752658258e-05, "loss": 0.0406, "step": 30720 }, { "grad_norm": 0.7406615614891052, "learning_rate": 8.296749794524094e-05, "loss": 0.0631, "step": 30730 }, { "grad_norm": 0.5067640542984009, "learning_rate": 8.295506475862332e-05, "loss": 0.0521, "step": 30740 }, { "grad_norm": 0.5021148324012756, "learning_rate": 8.294262796808933e-05, "loss": 0.0497, "step": 30750 }, { "grad_norm": 0.48730626702308655, "learning_rate": 8.293018757499909e-05, "loss": 0.0497, "step": 30760 }, { "grad_norm": 0.44479745626449585, "learning_rate": 8.291774358071305e-05, "loss": 0.0438, "step": 30770 }, { "grad_norm": 0.4355941414833069, "learning_rate": 8.290529598659205e-05, "loss": 0.0469, "step": 30780 }, { "grad_norm": 0.5006017684936523, "learning_rate": 8.289284479399738e-05, "loss": 0.0501, "step": 30790 }, { "grad_norm": 0.3886869549751282, "learning_rate": 8.288039000429064e-05, "loss": 0.0556, "step": 30800 }, { "grad_norm": 0.5497540831565857, "learning_rate": 8.286793161883388e-05, "loss": 0.0458, "step": 30810 }, { "grad_norm": 0.49019619822502136, "learning_rate": 8.285546963898954e-05, "loss": 0.0438, "step": 30820 }, { "grad_norm": 0.3766087293624878, "learning_rate": 8.284300406612044e-05, "loss": 0.0476, "step": 30830 }, { "grad_norm": 0.5386318564414978, "learning_rate": 8.283053490158978e-05, "loss": 0.0464, "step": 30840 }, { "grad_norm": 0.3363097012042999, "learning_rate": 8.28180621467612e-05, "loss": 0.0445, "step": 30850 }, { "grad_norm": 0.3678731620311737, "learning_rate": 8.280558580299867e-05, "loss": 0.0597, "step": 30860 }, { "grad_norm": 0.4756242334842682, "learning_rate": 8.279310587166661e-05, "loss": 0.0489, "step": 30870 }, { "grad_norm": 0.5521650314331055, "learning_rate": 8.278062235412978e-05, "loss": 0.0468, "step": 30880 }, { "grad_norm": 0.5008777976036072, "learning_rate": 8.276813525175339e-05, "loss": 0.0537, "step": 30890 }, { "grad_norm": 0.4002068340778351, "learning_rate": 8.2755644565903e-05, "loss": 0.061, "step": 30900 }, { "grad_norm": 0.387643426656723, "learning_rate": 8.274315029794454e-05, "loss": 0.0561, "step": 30910 }, { "grad_norm": 0.8307895064353943, "learning_rate": 8.273065244924443e-05, "loss": 0.0542, "step": 30920 }, { "grad_norm": 0.545673668384552, "learning_rate": 8.271815102116936e-05, "loss": 0.0475, "step": 30930 }, { "grad_norm": 0.33917176723480225, "learning_rate": 8.270564601508648e-05, "loss": 0.0456, "step": 30940 }, { "grad_norm": 0.5267147421836853, "learning_rate": 8.269313743236333e-05, "loss": 0.0539, "step": 30950 }, { "grad_norm": 0.49618083238601685, "learning_rate": 8.268062527436783e-05, "loss": 0.0482, "step": 30960 }, { "grad_norm": 0.430153489112854, "learning_rate": 8.266810954246828e-05, "loss": 0.0545, "step": 30970 }, { "grad_norm": 0.41770029067993164, "learning_rate": 8.265559023803338e-05, "loss": 0.0484, "step": 30980 }, { "grad_norm": 0.46460819244384766, "learning_rate": 8.264306736243223e-05, "loss": 0.0505, "step": 30990 }, { "grad_norm": 0.5161412358283997, "learning_rate": 8.263054091703432e-05, "loss": 0.0678, "step": 31000 }, { "grad_norm": 0.43068572878837585, "learning_rate": 8.26180109032095e-05, "loss": 0.0559, "step": 31010 }, { "grad_norm": 0.5279104709625244, "learning_rate": 8.260547732232807e-05, "loss": 0.0446, "step": 31020 }, { "grad_norm": 0.5179184675216675, "learning_rate": 8.259294017576063e-05, "loss": 0.0466, "step": 31030 }, { "grad_norm": 0.491397887468338, "learning_rate": 8.258039946487828e-05, "loss": 0.0547, "step": 31040 }, { "grad_norm": 0.36485302448272705, "learning_rate": 8.256785519105241e-05, "loss": 0.0432, "step": 31050 }, { "grad_norm": 0.4256230890750885, "learning_rate": 8.255530735565488e-05, "loss": 0.0464, "step": 31060 }, { "grad_norm": 0.5215448141098022, "learning_rate": 8.254275596005787e-05, "loss": 0.0444, "step": 31070 }, { "grad_norm": 0.48486003279685974, "learning_rate": 8.2530201005634e-05, "loss": 0.0603, "step": 31080 }, { "grad_norm": 0.5453437566757202, "learning_rate": 8.251764249375626e-05, "loss": 0.0589, "step": 31090 }, { "grad_norm": 0.5119855403900146, "learning_rate": 8.250508042579803e-05, "loss": 0.0522, "step": 31100 }, { "grad_norm": 0.6101858019828796, "learning_rate": 8.249251480313307e-05, "loss": 0.0566, "step": 31110 }, { "grad_norm": 0.486985981464386, "learning_rate": 8.247994562713555e-05, "loss": 0.0456, "step": 31120 }, { "grad_norm": 0.38106679916381836, "learning_rate": 8.246737289917998e-05, "loss": 0.0467, "step": 31130 }, { "grad_norm": 0.4867584705352783, "learning_rate": 8.245479662064135e-05, "loss": 0.0479, "step": 31140 }, { "grad_norm": 0.5783746838569641, "learning_rate": 8.244221679289496e-05, "loss": 0.0431, "step": 31150 }, { "grad_norm": 0.3230699598789215, "learning_rate": 8.242963341731652e-05, "loss": 0.0572, "step": 31160 }, { "grad_norm": 0.3514266312122345, "learning_rate": 8.24170464952821e-05, "loss": 0.046, "step": 31170 }, { "grad_norm": 0.6090875864028931, "learning_rate": 8.240445602816824e-05, "loss": 0.0629, "step": 31180 }, { "grad_norm": 0.38741743564605713, "learning_rate": 8.239186201735179e-05, "loss": 0.0525, "step": 31190 }, { "grad_norm": 0.8042653203010559, "learning_rate": 8.237926446420998e-05, "loss": 0.0434, "step": 31200 }, { "grad_norm": 0.422624409198761, "learning_rate": 8.236666337012052e-05, "loss": 0.0463, "step": 31210 }, { "grad_norm": 0.5943599343299866, "learning_rate": 8.23540587364614e-05, "loss": 0.0554, "step": 31220 }, { "grad_norm": 0.7877777814865112, "learning_rate": 8.234145056461107e-05, "loss": 0.0604, "step": 31230 }, { "grad_norm": 0.5525192022323608, "learning_rate": 8.232883885594832e-05, "loss": 0.0529, "step": 31240 }, { "grad_norm": 0.5412662029266357, "learning_rate": 8.231622361185236e-05, "loss": 0.0447, "step": 31250 }, { "grad_norm": 0.45826229453086853, "learning_rate": 8.230360483370278e-05, "loss": 0.0686, "step": 31260 }, { "grad_norm": 0.6293337941169739, "learning_rate": 8.229098252287953e-05, "loss": 0.0453, "step": 31270 }, { "grad_norm": 0.34845399856567383, "learning_rate": 8.2278356680763e-05, "loss": 0.0553, "step": 31280 }, { "grad_norm": 0.4896032512187958, "learning_rate": 8.22657273087339e-05, "loss": 0.053, "step": 31290 }, { "grad_norm": 0.511482298374176, "learning_rate": 8.225309440817336e-05, "loss": 0.0562, "step": 31300 }, { "grad_norm": 0.38469359278678894, "learning_rate": 8.224045798046293e-05, "loss": 0.0642, "step": 31310 }, { "grad_norm": 0.421228289604187, "learning_rate": 8.22278180269845e-05, "loss": 0.0504, "step": 31320 }, { "grad_norm": 0.47150322794914246, "learning_rate": 8.221517454912031e-05, "loss": 0.0416, "step": 31330 }, { "grad_norm": 0.5176181197166443, "learning_rate": 8.220252754825308e-05, "loss": 0.0451, "step": 31340 }, { "grad_norm": 0.677291750907898, "learning_rate": 8.218987702576586e-05, "loss": 0.0457, "step": 31350 }, { "grad_norm": 0.28549981117248535, "learning_rate": 8.217722298304207e-05, "loss": 0.0567, "step": 31360 }, { "grad_norm": 0.4412584602832794, "learning_rate": 8.216456542146557e-05, "loss": 0.0617, "step": 31370 }, { "grad_norm": 0.5078198909759521, "learning_rate": 8.215190434242055e-05, "loss": 0.0461, "step": 31380 }, { "grad_norm": 0.505553662776947, "learning_rate": 8.213923974729161e-05, "loss": 0.0494, "step": 31390 }, { "grad_norm": 0.5574935078620911, "learning_rate": 8.212657163746373e-05, "loss": 0.0569, "step": 31400 }, { "grad_norm": 0.38780859112739563, "learning_rate": 8.211390001432227e-05, "loss": 0.045, "step": 31410 }, { "grad_norm": 0.4726540148258209, "learning_rate": 8.210122487925297e-05, "loss": 0.0465, "step": 31420 }, { "grad_norm": 0.4360799491405487, "learning_rate": 8.208854623364202e-05, "loss": 0.0512, "step": 31430 }, { "grad_norm": 0.8456200957298279, "learning_rate": 8.207586407887589e-05, "loss": 0.047, "step": 31440 }, { "grad_norm": 1.80648934841156, "learning_rate": 8.206317841634148e-05, "loss": 0.0416, "step": 31450 }, { "grad_norm": 0.43681564927101135, "learning_rate": 8.205048924742609e-05, "loss": 0.0632, "step": 31460 }, { "grad_norm": 0.7070340514183044, "learning_rate": 8.203779657351738e-05, "loss": 0.0587, "step": 31470 }, { "grad_norm": 0.3983926475048065, "learning_rate": 8.20251003960034e-05, "loss": 0.0585, "step": 31480 }, { "grad_norm": 0.2907378077507019, "learning_rate": 8.201240071627258e-05, "loss": 0.0497, "step": 31490 }, { "grad_norm": 0.45733654499053955, "learning_rate": 8.199969753571377e-05, "loss": 0.063, "step": 31500 }, { "grad_norm": 0.4502321779727936, "learning_rate": 8.198699085571615e-05, "loss": 0.0584, "step": 31510 }, { "grad_norm": 0.37437704205513, "learning_rate": 8.197428067766928e-05, "loss": 0.0457, "step": 31520 }, { "grad_norm": 0.4722115695476532, "learning_rate": 8.196156700296316e-05, "loss": 0.0571, "step": 31530 }, { "grad_norm": 0.5434902906417847, "learning_rate": 8.194884983298814e-05, "loss": 0.041, "step": 31540 }, { "grad_norm": 0.3923145830631256, "learning_rate": 8.193612916913491e-05, "loss": 0.0455, "step": 31550 }, { "grad_norm": 0.42396435141563416, "learning_rate": 8.192340501279463e-05, "loss": 0.0462, "step": 31560 }, { "grad_norm": 1.2399811744689941, "learning_rate": 8.191067736535876e-05, "loss": 0.0518, "step": 31570 }, { "grad_norm": 1.122887134552002, "learning_rate": 8.18979462282192e-05, "loss": 0.0395, "step": 31580 }, { "grad_norm": 0.47322341799736023, "learning_rate": 8.188521160276819e-05, "loss": 0.0466, "step": 31590 }, { "grad_norm": 1.2314003705978394, "learning_rate": 8.187247349039837e-05, "loss": 0.0486, "step": 31600 }, { "grad_norm": 1.1870739459991455, "learning_rate": 8.185973189250278e-05, "loss": 0.0608, "step": 31610 }, { "grad_norm": 0.49267733097076416, "learning_rate": 8.184698681047482e-05, "loss": 0.0536, "step": 31620 }, { "grad_norm": 0.6398318409919739, "learning_rate": 8.183423824570827e-05, "loss": 0.0472, "step": 31630 }, { "grad_norm": 0.44248121976852417, "learning_rate": 8.182148619959725e-05, "loss": 0.0519, "step": 31640 }, { "grad_norm": 0.3697686195373535, "learning_rate": 8.180873067353636e-05, "loss": 0.0516, "step": 31650 }, { "grad_norm": 0.41572386026382446, "learning_rate": 8.179597166892052e-05, "loss": 0.0543, "step": 31660 }, { "grad_norm": 0.5137704610824585, "learning_rate": 8.178320918714501e-05, "loss": 0.0502, "step": 31670 }, { "grad_norm": 0.37035176157951355, "learning_rate": 8.177044322960554e-05, "loss": 0.0543, "step": 31680 }, { "grad_norm": 0.4637630581855774, "learning_rate": 8.175767379769816e-05, "loss": 0.0686, "step": 31690 }, { "grad_norm": 0.6622240543365479, "learning_rate": 8.174490089281932e-05, "loss": 0.0513, "step": 31700 }, { "grad_norm": 0.6394286751747131, "learning_rate": 8.173212451636584e-05, "loss": 0.0523, "step": 31710 }, { "grad_norm": 0.5484076142311096, "learning_rate": 8.171934466973493e-05, "loss": 0.0502, "step": 31720 }, { "grad_norm": 0.43693363666534424, "learning_rate": 8.170656135432418e-05, "loss": 0.0483, "step": 31730 }, { "grad_norm": 0.5237790942192078, "learning_rate": 8.169377457153155e-05, "loss": 0.0558, "step": 31740 }, { "grad_norm": 0.3863604664802551, "learning_rate": 8.168098432275539e-05, "loss": 0.0448, "step": 31750 }, { "grad_norm": 0.42011141777038574, "learning_rate": 8.166819060939442e-05, "loss": 0.0577, "step": 31760 }, { "grad_norm": 0.4017992913722992, "learning_rate": 8.165539343284772e-05, "loss": 0.0607, "step": 31770 }, { "grad_norm": 0.5267631411552429, "learning_rate": 8.16425927945148e-05, "loss": 0.0538, "step": 31780 }, { "grad_norm": 0.35004711151123047, "learning_rate": 8.162978869579551e-05, "loss": 0.056, "step": 31790 }, { "grad_norm": 0.43562474846839905, "learning_rate": 8.161698113809007e-05, "loss": 0.0527, "step": 31800 }, { "grad_norm": 0.40368834137916565, "learning_rate": 8.160417012279911e-05, "loss": 0.0551, "step": 31810 }, { "grad_norm": 0.4687488377094269, "learning_rate": 8.159135565132363e-05, "loss": 0.0487, "step": 31820 }, { "grad_norm": 0.39629942178726196, "learning_rate": 8.157853772506498e-05, "loss": 0.0629, "step": 31830 }, { "grad_norm": 0.6226157546043396, "learning_rate": 8.156571634542494e-05, "loss": 0.0541, "step": 31840 }, { "grad_norm": 0.44183433055877686, "learning_rate": 8.15528915138056e-05, "loss": 0.0505, "step": 31850 }, { "grad_norm": 0.4143139123916626, "learning_rate": 8.154006323160949e-05, "loss": 0.0365, "step": 31860 }, { "grad_norm": 0.4311133623123169, "learning_rate": 8.152723150023949e-05, "loss": 0.0514, "step": 31870 }, { "grad_norm": 0.48788315057754517, "learning_rate": 8.151439632109886e-05, "loss": 0.0511, "step": 31880 }, { "grad_norm": 0.4571242034435272, "learning_rate": 8.150155769559122e-05, "loss": 0.0581, "step": 31890 }, { "grad_norm": 0.8949800133705139, "learning_rate": 8.148871562512058e-05, "loss": 0.0459, "step": 31900 }, { "grad_norm": 0.467087060213089, "learning_rate": 8.147587011109136e-05, "loss": 0.0552, "step": 31910 }, { "grad_norm": 0.4373834729194641, "learning_rate": 8.14630211549083e-05, "loss": 0.0635, "step": 31920 }, { "grad_norm": 0.256366491317749, "learning_rate": 8.145016875797655e-05, "loss": 0.0515, "step": 31930 }, { "grad_norm": 0.43918129801750183, "learning_rate": 8.143731292170164e-05, "loss": 0.0555, "step": 31940 }, { "grad_norm": 0.3801446259021759, "learning_rate": 8.142445364748944e-05, "loss": 0.059, "step": 31950 }, { "grad_norm": 0.4601556658744812, "learning_rate": 8.141159093674624e-05, "loss": 0.057, "step": 31960 }, { "grad_norm": 0.46013492345809937, "learning_rate": 8.139872479087869e-05, "loss": 0.0467, "step": 31970 }, { "grad_norm": 0.5964556932449341, "learning_rate": 8.13858552112938e-05, "loss": 0.045, "step": 31980 }, { "grad_norm": 0.41430914402008057, "learning_rate": 8.137298219939895e-05, "loss": 0.0661, "step": 31990 }, { "grad_norm": 0.3804260492324829, "learning_rate": 8.136010575660196e-05, "loss": 0.0359, "step": 32000 }, { "grad_norm": 0.5080418586730957, "learning_rate": 8.134722588431095e-05, "loss": 0.0538, "step": 32010 }, { "grad_norm": 0.7084540128707886, "learning_rate": 8.133434258393444e-05, "loss": 0.0558, "step": 32020 }, { "grad_norm": 0.5940992832183838, "learning_rate": 8.132145585688134e-05, "loss": 0.0507, "step": 32030 }, { "grad_norm": 0.8910017609596252, "learning_rate": 8.130856570456093e-05, "loss": 0.0697, "step": 32040 }, { "grad_norm": 0.6170821189880371, "learning_rate": 8.129567212838283e-05, "loss": 0.0566, "step": 32050 }, { "grad_norm": 0.500700056552887, "learning_rate": 8.128277512975708e-05, "loss": 0.0558, "step": 32060 }, { "grad_norm": 0.5320931077003479, "learning_rate": 8.126987471009408e-05, "loss": 0.0566, "step": 32070 }, { "grad_norm": 0.5725850462913513, "learning_rate": 8.125697087080459e-05, "loss": 0.0446, "step": 32080 }, { "grad_norm": 0.5587963461875916, "learning_rate": 8.124406361329976e-05, "loss": 0.072, "step": 32090 }, { "grad_norm": 0.46290454268455505, "learning_rate": 8.123115293899111e-05, "loss": 0.0627, "step": 32100 }, { "grad_norm": 0.48646846413612366, "learning_rate": 8.121823884929055e-05, "loss": 0.0554, "step": 32110 }, { "grad_norm": 0.47238364815711975, "learning_rate": 8.12053213456103e-05, "loss": 0.0467, "step": 32120 }, { "grad_norm": 0.42064136266708374, "learning_rate": 8.119240042936303e-05, "loss": 0.054, "step": 32130 }, { "grad_norm": 0.4062533974647522, "learning_rate": 8.117947610196175e-05, "loss": 0.0454, "step": 32140 }, { "grad_norm": 0.4474640190601349, "learning_rate": 8.116654836481982e-05, "loss": 0.0463, "step": 32150 }, { "grad_norm": 0.4403149485588074, "learning_rate": 8.115361721935106e-05, "loss": 0.0424, "step": 32160 }, { "grad_norm": 0.5792954564094543, "learning_rate": 8.114068266696953e-05, "loss": 0.0438, "step": 32170 }, { "grad_norm": 0.4940851330757141, "learning_rate": 8.112774470908978e-05, "loss": 0.0378, "step": 32180 }, { "grad_norm": 0.4299820363521576, "learning_rate": 8.111480334712665e-05, "loss": 0.0408, "step": 32190 }, { "grad_norm": 0.6214194893836975, "learning_rate": 8.110185858249542e-05, "loss": 0.0491, "step": 32200 }, { "grad_norm": 0.48705658316612244, "learning_rate": 8.108891041661168e-05, "loss": 0.0543, "step": 32210 }, { "grad_norm": 0.6263862252235413, "learning_rate": 8.107595885089146e-05, "loss": 0.0548, "step": 32220 }, { "grad_norm": 0.60976243019104, "learning_rate": 8.106300388675108e-05, "loss": 0.0413, "step": 32230 }, { "grad_norm": 0.4385759234428406, "learning_rate": 8.105004552560731e-05, "loss": 0.0547, "step": 32240 }, { "grad_norm": 1.6541790962219238, "learning_rate": 8.103708376887724e-05, "loss": 0.0516, "step": 32250 }, { "grad_norm": 0.394091933965683, "learning_rate": 8.102411861797836e-05, "loss": 0.0463, "step": 32260 }, { "grad_norm": 0.6810627579689026, "learning_rate": 8.101115007432851e-05, "loss": 0.0422, "step": 32270 }, { "grad_norm": 0.3567715585231781, "learning_rate": 8.09981781393459e-05, "loss": 0.0433, "step": 32280 }, { "grad_norm": 0.3431337773799896, "learning_rate": 8.098520281444915e-05, "loss": 0.0389, "step": 32290 }, { "grad_norm": 0.4563450217247009, "learning_rate": 8.09722241010572e-05, "loss": 0.0487, "step": 32300 }, { "grad_norm": 0.33714917302131653, "learning_rate": 8.095924200058939e-05, "loss": 0.0514, "step": 32310 }, { "grad_norm": 0.37626463174819946, "learning_rate": 8.094625651446541e-05, "loss": 0.0395, "step": 32320 }, { "grad_norm": 2.17051100730896, "learning_rate": 8.093326764410536e-05, "loss": 0.0459, "step": 32330 }, { "grad_norm": 0.47294020652770996, "learning_rate": 8.092027539092966e-05, "loss": 0.044, "step": 32340 }, { "grad_norm": 0.748420774936676, "learning_rate": 8.090727975635913e-05, "loss": 0.0435, "step": 32350 }, { "grad_norm": 0.6387124061584473, "learning_rate": 8.089428074181497e-05, "loss": 0.0558, "step": 32360 }, { "grad_norm": 0.48292452096939087, "learning_rate": 8.088127834871871e-05, "loss": 0.0447, "step": 32370 }, { "grad_norm": 0.6570122838020325, "learning_rate": 8.086827257849226e-05, "loss": 0.0548, "step": 32380 }, { "grad_norm": 0.346362829208374, "learning_rate": 8.085526343255795e-05, "loss": 0.0383, "step": 32390 }, { "grad_norm": 0.4626754820346832, "learning_rate": 8.084225091233842e-05, "loss": 0.0494, "step": 32400 }, { "grad_norm": 0.47100067138671875, "learning_rate": 8.082923501925668e-05, "loss": 0.0459, "step": 32410 }, { "grad_norm": 0.5356322526931763, "learning_rate": 8.081621575473617e-05, "loss": 0.0596, "step": 32420 }, { "grad_norm": 0.5600137114524841, "learning_rate": 8.080319312020064e-05, "loss": 0.0574, "step": 32430 }, { "grad_norm": 0.4797821342945099, "learning_rate": 8.079016711707421e-05, "loss": 0.0453, "step": 32440 }, { "grad_norm": 0.42338770627975464, "learning_rate": 8.077713774678139e-05, "loss": 0.051, "step": 32450 }, { "grad_norm": 0.46564146876335144, "learning_rate": 8.076410501074707e-05, "loss": 0.0442, "step": 32460 }, { "grad_norm": 0.475918710231781, "learning_rate": 8.075106891039647e-05, "loss": 0.0445, "step": 32470 }, { "grad_norm": 0.351923406124115, "learning_rate": 8.073802944715523e-05, "loss": 0.0462, "step": 32480 }, { "grad_norm": 0.4820878207683563, "learning_rate": 8.072498662244929e-05, "loss": 0.0518, "step": 32490 }, { "grad_norm": 0.34802642464637756, "learning_rate": 8.0711940437705e-05, "loss": 0.0486, "step": 32500 }, { "grad_norm": 0.6494655609130859, "learning_rate": 8.06988908943491e-05, "loss": 0.0457, "step": 32510 }, { "grad_norm": 0.357624888420105, "learning_rate": 8.068583799380863e-05, "loss": 0.0397, "step": 32520 }, { "grad_norm": 0.34042760729789734, "learning_rate": 8.067278173751104e-05, "loss": 0.0483, "step": 32530 }, { "grad_norm": 0.6182964444160461, "learning_rate": 8.065972212688417e-05, "loss": 0.0429, "step": 32540 }, { "grad_norm": 0.5038017630577087, "learning_rate": 8.064665916335618e-05, "loss": 0.0447, "step": 32550 }, { "grad_norm": 0.40738657116889954, "learning_rate": 8.063359284835564e-05, "loss": 0.0597, "step": 32560 }, { "grad_norm": 0.46191468834877014, "learning_rate": 8.062052318331142e-05, "loss": 0.0489, "step": 32570 }, { "grad_norm": 0.6236262917518616, "learning_rate": 8.060745016965283e-05, "loss": 0.044, "step": 32580 }, { "grad_norm": 0.324859082698822, "learning_rate": 8.059437380880952e-05, "loss": 0.0533, "step": 32590 }, { "grad_norm": 0.45872825384140015, "learning_rate": 8.058129410221146e-05, "loss": 0.0491, "step": 32600 }, { "grad_norm": 0.4263606667518616, "learning_rate": 8.056821105128908e-05, "loss": 0.0593, "step": 32610 }, { "grad_norm": 0.36469966173171997, "learning_rate": 8.05551246574731e-05, "loss": 0.0544, "step": 32620 }, { "grad_norm": 0.4475961923599243, "learning_rate": 8.05420349221946e-05, "loss": 0.0503, "step": 32630 }, { "grad_norm": 0.3863854706287384, "learning_rate": 8.05289418468851e-05, "loss": 0.0474, "step": 32640 }, { "grad_norm": 0.49379390478134155, "learning_rate": 8.051584543297642e-05, "loss": 0.0455, "step": 32650 }, { "grad_norm": 0.36377066373825073, "learning_rate": 8.050274568190074e-05, "loss": 0.0417, "step": 32660 }, { "grad_norm": 0.3054562509059906, "learning_rate": 8.048964259509067e-05, "loss": 0.0519, "step": 32670 }, { "grad_norm": 0.3875231146812439, "learning_rate": 8.047653617397914e-05, "loss": 0.0388, "step": 32680 }, { "grad_norm": 0.5133560299873352, "learning_rate": 8.046342641999941e-05, "loss": 0.0498, "step": 32690 }, { "grad_norm": 0.5040869116783142, "learning_rate": 8.045031333458517e-05, "loss": 0.0467, "step": 32700 }, { "grad_norm": 0.4641421437263489, "learning_rate": 8.043719691917047e-05, "loss": 0.0638, "step": 32710 }, { "grad_norm": 0.8265696167945862, "learning_rate": 8.042407717518966e-05, "loss": 0.0519, "step": 32720 }, { "grad_norm": 0.452671080827713, "learning_rate": 8.041095410407751e-05, "loss": 0.0444, "step": 32730 }, { "grad_norm": 0.5483545064926147, "learning_rate": 8.039782770726913e-05, "loss": 0.0571, "step": 32740 }, { "grad_norm": 0.2877472937107086, "learning_rate": 8.038469798620004e-05, "loss": 0.0423, "step": 32750 }, { "grad_norm": 0.4404021203517914, "learning_rate": 8.037156494230604e-05, "loss": 0.0437, "step": 32760 }, { "grad_norm": 0.5092678666114807, "learning_rate": 8.035842857702338e-05, "loss": 0.044, "step": 32770 }, { "grad_norm": 0.569448709487915, "learning_rate": 8.03452888917886e-05, "loss": 0.0437, "step": 32780 }, { "grad_norm": 0.3826753795146942, "learning_rate": 8.033214588803866e-05, "loss": 0.0377, "step": 32790 }, { "grad_norm": 0.41868826746940613, "learning_rate": 8.031899956721083e-05, "loss": 0.037, "step": 32800 }, { "grad_norm": 0.4718017280101776, "learning_rate": 8.030584993074282e-05, "loss": 0.0449, "step": 32810 }, { "grad_norm": 0.47011464834213257, "learning_rate": 8.02926969800726e-05, "loss": 0.0408, "step": 32820 }, { "grad_norm": 0.3673759996891022, "learning_rate": 8.027954071663859e-05, "loss": 0.0517, "step": 32830 }, { "grad_norm": 0.5209037661552429, "learning_rate": 8.026638114187954e-05, "loss": 0.0415, "step": 32840 }, { "grad_norm": 0.34132012724876404, "learning_rate": 8.025321825723456e-05, "loss": 0.049, "step": 32850 }, { "grad_norm": 0.6669528484344482, "learning_rate": 8.02400520641431e-05, "loss": 0.051, "step": 32860 }, { "grad_norm": 1.3482654094696045, "learning_rate": 8.022688256404501e-05, "loss": 0.0439, "step": 32870 }, { "grad_norm": 0.4483191967010498, "learning_rate": 8.02137097583805e-05, "loss": 0.0385, "step": 32880 }, { "grad_norm": 0.43122750520706177, "learning_rate": 8.02005336485901e-05, "loss": 0.0482, "step": 32890 }, { "grad_norm": 0.7663486003875732, "learning_rate": 8.018735423611476e-05, "loss": 0.0669, "step": 32900 }, { "grad_norm": 0.3560352325439453, "learning_rate": 8.017417152239574e-05, "loss": 0.0535, "step": 32910 }, { "grad_norm": 0.7608107924461365, "learning_rate": 8.01609855088747e-05, "loss": 0.0554, "step": 32920 }, { "grad_norm": 0.3949199914932251, "learning_rate": 8.014779619699362e-05, "loss": 0.0488, "step": 32930 }, { "grad_norm": 0.39590948820114136, "learning_rate": 8.013460358819489e-05, "loss": 0.0487, "step": 32940 }, { "grad_norm": 0.3662780821323395, "learning_rate": 8.01214076839212e-05, "loss": 0.0483, "step": 32950 }, { "grad_norm": 0.43181586265563965, "learning_rate": 8.010820848561565e-05, "loss": 0.0479, "step": 32960 }, { "grad_norm": 0.33737093210220337, "learning_rate": 8.009500599472171e-05, "loss": 0.0438, "step": 32970 }, { "grad_norm": 0.5812344551086426, "learning_rate": 8.008180021268314e-05, "loss": 0.038, "step": 32980 }, { "grad_norm": 0.40476781129837036, "learning_rate": 8.006859114094414e-05, "loss": 0.0365, "step": 32990 }, { "grad_norm": 0.5075405240058899, "learning_rate": 8.005537878094921e-05, "loss": 0.0458, "step": 33000 }, { "grad_norm": 0.4124268889427185, "learning_rate": 8.004216313414323e-05, "loss": 0.0455, "step": 33010 }, { "grad_norm": 0.3990432918071747, "learning_rate": 8.002894420197149e-05, "loss": 0.0405, "step": 33020 }, { "grad_norm": 0.41293513774871826, "learning_rate": 8.001572198587954e-05, "loss": 0.0472, "step": 33030 }, { "grad_norm": 0.6124399304389954, "learning_rate": 8.000249648731338e-05, "loss": 0.0446, "step": 33040 }, { "grad_norm": 0.5210483074188232, "learning_rate": 7.998926770771928e-05, "loss": 0.0602, "step": 33050 }, { "grad_norm": 0.5729021430015564, "learning_rate": 7.997603564854397e-05, "loss": 0.0543, "step": 33060 }, { "grad_norm": 0.5667002201080322, "learning_rate": 7.996280031123448e-05, "loss": 0.0578, "step": 33070 }, { "grad_norm": 0.7722586989402771, "learning_rate": 7.994956169723818e-05, "loss": 0.0669, "step": 33080 }, { "grad_norm": 0.6219077110290527, "learning_rate": 7.993631980800285e-05, "loss": 0.0393, "step": 33090 }, { "grad_norm": 0.6747680306434631, "learning_rate": 7.992307464497659e-05, "loss": 0.0572, "step": 33100 }, { "grad_norm": 0.7142345309257507, "learning_rate": 7.990982620960787e-05, "loss": 0.0425, "step": 33110 }, { "grad_norm": 0.5013749599456787, "learning_rate": 7.989657450334554e-05, "loss": 0.055, "step": 33120 }, { "grad_norm": 0.3917314410209656, "learning_rate": 7.988331952763877e-05, "loss": 0.0468, "step": 33130 }, { "grad_norm": 0.5947215557098389, "learning_rate": 7.987006128393709e-05, "loss": 0.0647, "step": 33140 }, { "grad_norm": 0.5866904258728027, "learning_rate": 7.985679977369043e-05, "loss": 0.0588, "step": 33150 }, { "grad_norm": 0.8228768706321716, "learning_rate": 7.984353499834902e-05, "loss": 0.0533, "step": 33160 }, { "grad_norm": 1.1051055192947388, "learning_rate": 7.983026695936351e-05, "loss": 0.0538, "step": 33170 }, { "grad_norm": 1.0102565288543701, "learning_rate": 7.981699565818486e-05, "loss": 0.0534, "step": 33180 }, { "grad_norm": 0.38971397280693054, "learning_rate": 7.980372109626437e-05, "loss": 0.041, "step": 33190 }, { "grad_norm": 0.3910974860191345, "learning_rate": 7.979044327505375e-05, "loss": 0.0509, "step": 33200 }, { "grad_norm": 0.5354548692703247, "learning_rate": 7.977716219600506e-05, "loss": 0.0538, "step": 33210 }, { "grad_norm": 0.4203084409236908, "learning_rate": 7.97638778605707e-05, "loss": 0.0515, "step": 33220 }, { "grad_norm": 0.3658675253391266, "learning_rate": 7.975059027020338e-05, "loss": 0.0587, "step": 33230 }, { "grad_norm": 0.3668314814567566, "learning_rate": 7.973729942635623e-05, "loss": 0.0493, "step": 33240 }, { "grad_norm": 0.4133210778236389, "learning_rate": 7.972400533048273e-05, "loss": 0.0537, "step": 33250 }, { "grad_norm": 0.4657071530818939, "learning_rate": 7.97107079840367e-05, "loss": 0.0467, "step": 33260 }, { "grad_norm": 0.4895913004875183, "learning_rate": 7.969740738847231e-05, "loss": 0.0506, "step": 33270 }, { "grad_norm": 0.5719534158706665, "learning_rate": 7.968410354524411e-05, "loss": 0.0528, "step": 33280 }, { "grad_norm": 0.40895259380340576, "learning_rate": 7.967079645580697e-05, "loss": 0.0488, "step": 33290 }, { "grad_norm": 0.3530847430229187, "learning_rate": 7.965748612161612e-05, "loss": 0.0398, "step": 33300 }, { "grad_norm": 0.6250398755073547, "learning_rate": 7.96441725441272e-05, "loss": 0.043, "step": 33310 }, { "grad_norm": 0.400055855512619, "learning_rate": 7.963085572479614e-05, "loss": 0.041, "step": 33320 }, { "grad_norm": 0.4699268937110901, "learning_rate": 7.961753566507924e-05, "loss": 0.0441, "step": 33330 }, { "grad_norm": 0.3645963966846466, "learning_rate": 7.960421236643316e-05, "loss": 0.0492, "step": 33340 }, { "grad_norm": 0.5913457274436951, "learning_rate": 7.959088583031496e-05, "loss": 0.05, "step": 33350 }, { "grad_norm": 0.43097755312919617, "learning_rate": 7.957755605818194e-05, "loss": 0.0481, "step": 33360 }, { "grad_norm": 0.4094424843788147, "learning_rate": 7.956422305149185e-05, "loss": 0.0501, "step": 33370 }, { "grad_norm": 0.49379971623420715, "learning_rate": 7.95508868117028e-05, "loss": 0.052, "step": 33380 }, { "grad_norm": 0.5422527194023132, "learning_rate": 7.953754734027318e-05, "loss": 0.0608, "step": 33390 }, { "grad_norm": 0.45573505759239197, "learning_rate": 7.952420463866182e-05, "loss": 0.0408, "step": 33400 }, { "grad_norm": 0.526559591293335, "learning_rate": 7.951085870832782e-05, "loss": 0.0568, "step": 33410 }, { "grad_norm": 0.3725185692310333, "learning_rate": 7.949750955073067e-05, "loss": 0.048, "step": 33420 }, { "grad_norm": 0.479224294424057, "learning_rate": 7.948415716733022e-05, "loss": 0.0557, "step": 33430 }, { "grad_norm": 0.441642701625824, "learning_rate": 7.947080155958669e-05, "loss": 0.0457, "step": 33440 }, { "grad_norm": 0.9477012157440186, "learning_rate": 7.94574427289606e-05, "loss": 0.0457, "step": 33450 }, { "grad_norm": 0.380191832780838, "learning_rate": 7.944408067691284e-05, "loss": 0.0481, "step": 33460 }, { "grad_norm": 0.37459251284599304, "learning_rate": 7.943071540490473e-05, "loss": 0.0398, "step": 33470 }, { "grad_norm": 0.7156181335449219, "learning_rate": 7.94173469143978e-05, "loss": 0.0456, "step": 33480 }, { "grad_norm": 0.3879058361053467, "learning_rate": 7.940397520685406e-05, "loss": 0.0421, "step": 33490 }, { "grad_norm": 0.3783032298088074, "learning_rate": 7.939060028373577e-05, "loss": 0.0467, "step": 33500 }, { "grad_norm": 0.4489656388759613, "learning_rate": 7.937722214650565e-05, "loss": 0.0567, "step": 33510 }, { "grad_norm": 0.2524546980857849, "learning_rate": 7.936384079662666e-05, "loss": 0.0342, "step": 33520 }, { "grad_norm": 0.4081370234489441, "learning_rate": 7.93504562355622e-05, "loss": 0.0536, "step": 33530 }, { "grad_norm": 0.8643641471862793, "learning_rate": 7.933706846477599e-05, "loss": 0.0516, "step": 33540 }, { "grad_norm": 0.4185478389263153, "learning_rate": 7.932367748573206e-05, "loss": 0.0593, "step": 33550 }, { "grad_norm": 0.5609970688819885, "learning_rate": 7.931028329989485e-05, "loss": 0.0646, "step": 33560 }, { "grad_norm": 0.5251219868659973, "learning_rate": 7.929688590872913e-05, "loss": 0.0609, "step": 33570 }, { "grad_norm": 0.5377345085144043, "learning_rate": 7.928348531370003e-05, "loss": 0.0411, "step": 33580 }, { "grad_norm": 0.3516751229763031, "learning_rate": 7.927008151627297e-05, "loss": 0.0385, "step": 33590 }, { "grad_norm": 0.3875887989997864, "learning_rate": 7.925667451791383e-05, "loss": 0.0448, "step": 33600 }, { "grad_norm": 0.7845546007156372, "learning_rate": 7.924326432008874e-05, "loss": 0.0575, "step": 33610 }, { "grad_norm": 0.4702932834625244, "learning_rate": 7.922985092426422e-05, "loss": 0.0454, "step": 33620 }, { "grad_norm": 0.512648344039917, "learning_rate": 7.921643433190717e-05, "loss": 0.0448, "step": 33630 }, { "grad_norm": 0.6541152596473694, "learning_rate": 7.920301454448478e-05, "loss": 0.052, "step": 33640 }, { "grad_norm": 0.3287505507469177, "learning_rate": 7.918959156346461e-05, "loss": 0.0339, "step": 33650 }, { "grad_norm": 2.8092243671417236, "learning_rate": 7.91761653903146e-05, "loss": 0.0474, "step": 33660 }, { "grad_norm": 0.36858516931533813, "learning_rate": 7.916273602650302e-05, "loss": 0.0473, "step": 33670 }, { "grad_norm": 0.29352471232414246, "learning_rate": 7.914930347349847e-05, "loss": 0.0421, "step": 33680 }, { "grad_norm": 0.33712872862815857, "learning_rate": 7.913586773276992e-05, "loss": 0.0521, "step": 33690 }, { "grad_norm": 0.45149919390678406, "learning_rate": 7.912242880578667e-05, "loss": 0.0438, "step": 33700 }, { "grad_norm": 0.40309080481529236, "learning_rate": 7.910898669401839e-05, "loss": 0.0465, "step": 33710 }, { "grad_norm": 0.3726237118244171, "learning_rate": 7.909554139893511e-05, "loss": 0.048, "step": 33720 }, { "grad_norm": 0.7679177522659302, "learning_rate": 7.908209292200715e-05, "loss": 0.0507, "step": 33730 }, { "grad_norm": 0.3435478210449219, "learning_rate": 7.906864126470523e-05, "loss": 0.0404, "step": 33740 }, { "grad_norm": 0.47970297932624817, "learning_rate": 7.905518642850041e-05, "loss": 0.0416, "step": 33750 }, { "grad_norm": 0.4872299134731293, "learning_rate": 7.904172841486409e-05, "loss": 0.0659, "step": 33760 }, { "grad_norm": 0.6823678612709045, "learning_rate": 7.902826722526801e-05, "loss": 0.0562, "step": 33770 }, { "grad_norm": 0.4116073250770569, "learning_rate": 7.901480286118427e-05, "loss": 0.0441, "step": 33780 }, { "grad_norm": 0.46848684549331665, "learning_rate": 7.900133532408531e-05, "loss": 0.0493, "step": 33790 }, { "grad_norm": 0.4153235852718353, "learning_rate": 7.898786461544395e-05, "loss": 0.0491, "step": 33800 }, { "grad_norm": 0.36700499057769775, "learning_rate": 7.897439073673325e-05, "loss": 0.0414, "step": 33810 }, { "grad_norm": 0.5072822570800781, "learning_rate": 7.896091368942677e-05, "loss": 0.049, "step": 33820 }, { "grad_norm": 0.8812363147735596, "learning_rate": 7.894743347499832e-05, "loss": 0.0509, "step": 33830 }, { "grad_norm": 0.4747236967086792, "learning_rate": 7.893395009492203e-05, "loss": 0.04, "step": 33840 }, { "grad_norm": 0.49135294556617737, "learning_rate": 7.892046355067248e-05, "loss": 0.0438, "step": 33850 }, { "grad_norm": 0.7122088670730591, "learning_rate": 7.890697384372451e-05, "loss": 0.0439, "step": 33860 }, { "grad_norm": 0.3939962387084961, "learning_rate": 7.889348097555336e-05, "loss": 0.0431, "step": 33870 }, { "grad_norm": 0.4971941113471985, "learning_rate": 7.887998494763455e-05, "loss": 0.0436, "step": 33880 }, { "grad_norm": 0.6405777931213379, "learning_rate": 7.886648576144404e-05, "loss": 0.0414, "step": 33890 }, { "grad_norm": 0.4292692542076111, "learning_rate": 7.885298341845802e-05, "loss": 0.0446, "step": 33900 }, { "grad_norm": 0.4284351170063019, "learning_rate": 7.883947792015311e-05, "loss": 0.0587, "step": 33910 }, { "grad_norm": 0.5962519645690918, "learning_rate": 7.882596926800628e-05, "loss": 0.07, "step": 33920 }, { "grad_norm": 1.9465324878692627, "learning_rate": 7.881245746349477e-05, "loss": 0.0523, "step": 33930 }, { "grad_norm": 0.718449592590332, "learning_rate": 7.879894250809623e-05, "loss": 0.0471, "step": 33940 }, { "grad_norm": 0.45663800835609436, "learning_rate": 7.878542440328865e-05, "loss": 0.0644, "step": 33950 }, { "grad_norm": 0.5257205367088318, "learning_rate": 7.877190315055031e-05, "loss": 0.0532, "step": 33960 }, { "grad_norm": 0.35130423307418823, "learning_rate": 7.875837875135991e-05, "loss": 0.0414, "step": 33970 }, { "grad_norm": 0.44061020016670227, "learning_rate": 7.874485120719646e-05, "loss": 0.034, "step": 33980 }, { "grad_norm": 0.5530441403388977, "learning_rate": 7.873132051953928e-05, "loss": 0.0448, "step": 33990 }, { "grad_norm": 0.8614196181297302, "learning_rate": 7.87177866898681e-05, "loss": 0.053, "step": 34000 }, { "grad_norm": 0.33389419317245483, "learning_rate": 7.870424971966294e-05, "loss": 0.0467, "step": 34010 }, { "grad_norm": 0.3725472390651703, "learning_rate": 7.869070961040419e-05, "loss": 0.0441, "step": 34020 }, { "grad_norm": 0.5576638579368591, "learning_rate": 7.867716636357257e-05, "loss": 0.0603, "step": 34030 }, { "grad_norm": 0.4248354732990265, "learning_rate": 7.866361998064915e-05, "loss": 0.0418, "step": 34040 }, { "grad_norm": 0.6979309320449829, "learning_rate": 7.865007046311534e-05, "loss": 0.0472, "step": 34050 }, { "grad_norm": 0.5639699697494507, "learning_rate": 7.86365178124529e-05, "loss": 0.0449, "step": 34060 }, { "grad_norm": 0.3092838227748871, "learning_rate": 7.862296203014394e-05, "loss": 0.0477, "step": 34070 }, { "grad_norm": 0.4106456935405731, "learning_rate": 7.860940311767088e-05, "loss": 0.0456, "step": 34080 }, { "grad_norm": 0.781898021697998, "learning_rate": 7.85958410765165e-05, "loss": 0.0476, "step": 34090 }, { "grad_norm": 0.4693025052547455, "learning_rate": 7.858227590816394e-05, "loss": 0.0417, "step": 34100 }, { "grad_norm": 0.45026862621307373, "learning_rate": 7.856870761409664e-05, "loss": 0.0485, "step": 34110 }, { "grad_norm": 0.6332845687866211, "learning_rate": 7.855513619579846e-05, "loss": 0.0442, "step": 34120 }, { "grad_norm": 0.37065792083740234, "learning_rate": 7.85415616547535e-05, "loss": 0.0472, "step": 34130 }, { "grad_norm": 0.7086060643196106, "learning_rate": 7.852798399244627e-05, "loss": 0.0422, "step": 34140 }, { "grad_norm": 0.6568113565444946, "learning_rate": 7.851440321036161e-05, "loss": 0.0478, "step": 34150 }, { "grad_norm": 0.5148226618766785, "learning_rate": 7.850081930998468e-05, "loss": 0.0483, "step": 34160 }, { "grad_norm": 0.46272626519203186, "learning_rate": 7.8487232292801e-05, "loss": 0.0489, "step": 34170 }, { "grad_norm": 0.5997464656829834, "learning_rate": 7.847364216029642e-05, "loss": 0.0463, "step": 34180 }, { "grad_norm": 0.5929998755455017, "learning_rate": 7.846004891395716e-05, "loss": 0.0538, "step": 34190 }, { "grad_norm": 0.46522992849349976, "learning_rate": 7.844645255526972e-05, "loss": 0.0415, "step": 34200 }, { "grad_norm": 0.9737408757209778, "learning_rate": 7.843285308572101e-05, "loss": 0.0448, "step": 34210 }, { "grad_norm": 0.4308841824531555, "learning_rate": 7.841925050679823e-05, "loss": 0.0409, "step": 34220 }, { "grad_norm": 0.5555514693260193, "learning_rate": 7.840564481998895e-05, "loss": 0.0468, "step": 34230 }, { "grad_norm": 0.6217373609542847, "learning_rate": 7.839203602678105e-05, "loss": 0.0487, "step": 34240 }, { "grad_norm": 0.343107134103775, "learning_rate": 7.837842412866279e-05, "loss": 0.0455, "step": 34250 }, { "grad_norm": 0.4156661927700043, "learning_rate": 7.836480912712272e-05, "loss": 0.0455, "step": 34260 }, { "grad_norm": 0.5540315508842468, "learning_rate": 7.835119102364976e-05, "loss": 0.057, "step": 34270 }, { "grad_norm": 0.38190004229545593, "learning_rate": 7.83375698197332e-05, "loss": 0.0456, "step": 34280 }, { "grad_norm": 0.42211222648620605, "learning_rate": 7.83239455168626e-05, "loss": 0.045, "step": 34290 }, { "grad_norm": 0.43379828333854675, "learning_rate": 7.83103181165279e-05, "loss": 0.0496, "step": 34300 }, { "grad_norm": 0.5189018249511719, "learning_rate": 7.829668762021937e-05, "loss": 0.0592, "step": 34310 }, { "grad_norm": 0.6911748647689819, "learning_rate": 7.828305402942764e-05, "loss": 0.0481, "step": 34320 }, { "grad_norm": 0.5933958292007446, "learning_rate": 7.826941734564363e-05, "loss": 0.0454, "step": 34330 }, { "grad_norm": 0.53706294298172, "learning_rate": 7.825577757035865e-05, "loss": 0.0457, "step": 34340 }, { "grad_norm": 0.7562934160232544, "learning_rate": 7.824213470506431e-05, "loss": 0.0628, "step": 34350 }, { "grad_norm": 0.48241981863975525, "learning_rate": 7.822848875125257e-05, "loss": 0.0533, "step": 34360 }, { "grad_norm": 0.4387836754322052, "learning_rate": 7.821483971041576e-05, "loss": 0.0578, "step": 34370 }, { "grad_norm": 0.5956125259399414, "learning_rate": 7.820118758404649e-05, "loss": 0.0537, "step": 34380 }, { "grad_norm": 0.28082314133644104, "learning_rate": 7.818753237363776e-05, "loss": 0.0463, "step": 34390 }, { "grad_norm": 0.9423462748527527, "learning_rate": 7.817387408068286e-05, "loss": 0.0524, "step": 34400 }, { "grad_norm": 0.3521486520767212, "learning_rate": 7.816021270667544e-05, "loss": 0.0468, "step": 34410 }, { "grad_norm": 0.3309849500656128, "learning_rate": 7.81465482531095e-05, "loss": 0.039, "step": 34420 }, { "grad_norm": 1.033035397529602, "learning_rate": 7.813288072147938e-05, "loss": 0.0479, "step": 34430 }, { "grad_norm": 0.40757831931114197, "learning_rate": 7.811921011327972e-05, "loss": 0.0533, "step": 34440 }, { "grad_norm": 0.32782530784606934, "learning_rate": 7.810553643000549e-05, "loss": 0.053, "step": 34450 }, { "grad_norm": 1.178110957145691, "learning_rate": 7.809185967315206e-05, "loss": 0.0398, "step": 34460 }, { "grad_norm": 0.4411434829235077, "learning_rate": 7.80781798442151e-05, "loss": 0.0498, "step": 34470 }, { "grad_norm": 0.7269233465194702, "learning_rate": 7.806449694469061e-05, "loss": 0.0528, "step": 34480 }, { "grad_norm": 0.3366507589817047, "learning_rate": 7.805081097607492e-05, "loss": 0.0543, "step": 34490 }, { "grad_norm": 0.45840129256248474, "learning_rate": 7.803712193986474e-05, "loss": 0.0384, "step": 34500 }, { "grad_norm": 0.5138755440711975, "learning_rate": 7.802342983755702e-05, "loss": 0.0405, "step": 34510 }, { "grad_norm": 0.709720253944397, "learning_rate": 7.800973467064918e-05, "loss": 0.0404, "step": 34520 }, { "grad_norm": 0.5515240430831909, "learning_rate": 7.799603644063884e-05, "loss": 0.0515, "step": 34530 }, { "grad_norm": 0.5372046828269958, "learning_rate": 7.798233514902405e-05, "loss": 0.0525, "step": 34540 }, { "grad_norm": 0.46029385924339294, "learning_rate": 7.796863079730318e-05, "loss": 0.0553, "step": 34550 }, { "grad_norm": 0.4428824186325073, "learning_rate": 7.795492338697488e-05, "loss": 0.0377, "step": 34560 }, { "grad_norm": 0.4515652060508728, "learning_rate": 7.794121291953819e-05, "loss": 0.0437, "step": 34570 }, { "grad_norm": 0.4352078139781952, "learning_rate": 7.792749939649246e-05, "loss": 0.0427, "step": 34580 }, { "grad_norm": 0.3671533167362213, "learning_rate": 7.79137828193374e-05, "loss": 0.049, "step": 34590 }, { "grad_norm": 0.37610870599746704, "learning_rate": 7.790006318957301e-05, "loss": 0.0531, "step": 34600 }, { "grad_norm": 0.4633308947086334, "learning_rate": 7.788634050869965e-05, "loss": 0.0486, "step": 34610 }, { "grad_norm": 1.8306641578674316, "learning_rate": 7.787261477821803e-05, "loss": 0.0441, "step": 34620 }, { "grad_norm": 0.5549707412719727, "learning_rate": 7.785888599962916e-05, "loss": 0.0453, "step": 34630 }, { "grad_norm": 0.5523878931999207, "learning_rate": 7.784515417443439e-05, "loss": 0.0468, "step": 34640 }, { "grad_norm": 0.5432084202766418, "learning_rate": 7.783141930413545e-05, "loss": 0.0486, "step": 34650 }, { "grad_norm": 0.4065263271331787, "learning_rate": 7.78176813902343e-05, "loss": 0.0421, "step": 34660 }, { "grad_norm": 0.6289668679237366, "learning_rate": 7.780394043423336e-05, "loss": 0.0473, "step": 34670 }, { "grad_norm": 0.5037851929664612, "learning_rate": 7.77901964376353e-05, "loss": 0.0496, "step": 34680 }, { "grad_norm": 0.7212119102478027, "learning_rate": 7.777644940194316e-05, "loss": 0.0595, "step": 34690 }, { "grad_norm": 0.43196773529052734, "learning_rate": 7.776269932866023e-05, "loss": 0.044, "step": 34700 }, { "grad_norm": 0.7475755214691162, "learning_rate": 7.774894621929026e-05, "loss": 0.0562, "step": 34710 }, { "grad_norm": 0.6518808007240295, "learning_rate": 7.773519007533725e-05, "loss": 0.0562, "step": 34720 }, { "grad_norm": 1.2705395221710205, "learning_rate": 7.772143089830556e-05, "loss": 0.0479, "step": 34730 }, { "grad_norm": 0.31568288803100586, "learning_rate": 7.770766868969985e-05, "loss": 0.0463, "step": 34740 }, { "grad_norm": 0.3704566955566406, "learning_rate": 7.769390345102518e-05, "loss": 0.047, "step": 34750 }, { "grad_norm": 0.5141006112098694, "learning_rate": 7.768013518378683e-05, "loss": 0.0612, "step": 34760 }, { "grad_norm": 0.49264851212501526, "learning_rate": 7.766636388949053e-05, "loss": 0.0337, "step": 34770 }, { "grad_norm": 0.29416871070861816, "learning_rate": 7.765258956964229e-05, "loss": 0.0625, "step": 34780 }, { "grad_norm": 0.7715259790420532, "learning_rate": 7.76388122257484e-05, "loss": 0.0434, "step": 34790 }, { "grad_norm": 0.5708476305007935, "learning_rate": 7.762503185931558e-05, "loss": 0.0391, "step": 34800 }, { "grad_norm": 0.3538724184036255, "learning_rate": 7.76112484718508e-05, "loss": 0.04, "step": 34810 }, { "grad_norm": 1.0407060384750366, "learning_rate": 7.75974620648614e-05, "loss": 0.0438, "step": 34820 }, { "grad_norm": 0.6759210228919983, "learning_rate": 7.758367263985503e-05, "loss": 0.0548, "step": 34830 }, { "grad_norm": 0.533420741558075, "learning_rate": 7.75698801983397e-05, "loss": 0.0502, "step": 34840 }, { "grad_norm": 0.6713451147079468, "learning_rate": 7.755608474182372e-05, "loss": 0.0509, "step": 34850 }, { "grad_norm": 0.621674656867981, "learning_rate": 7.754228627181574e-05, "loss": 0.0641, "step": 34860 }, { "grad_norm": 0.4424152374267578, "learning_rate": 7.752848478982476e-05, "loss": 0.0472, "step": 34870 }, { "grad_norm": 0.7124353647232056, "learning_rate": 7.751468029736006e-05, "loss": 0.0439, "step": 34880 }, { "grad_norm": 0.48801714181900024, "learning_rate": 7.750087279593129e-05, "loss": 0.0486, "step": 34890 }, { "grad_norm": 0.44765758514404297, "learning_rate": 7.748706228704843e-05, "loss": 0.0374, "step": 34900 }, { "grad_norm": 0.38310694694519043, "learning_rate": 7.747324877222176e-05, "loss": 0.0426, "step": 34910 }, { "grad_norm": 0.7190164923667908, "learning_rate": 7.745943225296188e-05, "loss": 0.0465, "step": 34920 }, { "grad_norm": 0.5386738181114197, "learning_rate": 7.744561273077981e-05, "loss": 0.0342, "step": 34930 }, { "grad_norm": 0.40208977460861206, "learning_rate": 7.743179020718678e-05, "loss": 0.0347, "step": 34940 }, { "grad_norm": 0.4458523392677307, "learning_rate": 7.741796468369443e-05, "loss": 0.0396, "step": 34950 }, { "grad_norm": 0.40804722905158997, "learning_rate": 7.740413616181466e-05, "loss": 0.049, "step": 34960 }, { "grad_norm": 1.4458389282226562, "learning_rate": 7.739030464305978e-05, "loss": 0.0522, "step": 34970 }, { "grad_norm": 0.69940584897995, "learning_rate": 7.737647012894235e-05, "loss": 0.0588, "step": 34980 }, { "grad_norm": 0.6433539390563965, "learning_rate": 7.736263262097532e-05, "loss": 0.0576, "step": 34990 }, { "grad_norm": 0.5143626928329468, "learning_rate": 7.734879212067192e-05, "loss": 0.042, "step": 35000 }, { "grad_norm": 0.5389835238456726, "learning_rate": 7.733494862954573e-05, "loss": 0.0524, "step": 35010 }, { "grad_norm": 0.4102475345134735, "learning_rate": 7.732110214911066e-05, "loss": 0.0407, "step": 35020 }, { "grad_norm": 0.7237799763679504, "learning_rate": 7.730725268088094e-05, "loss": 0.0543, "step": 35030 }, { "grad_norm": 0.54477459192276, "learning_rate": 7.729340022637112e-05, "loss": 0.0514, "step": 35040 }, { "grad_norm": 0.45505744218826294, "learning_rate": 7.727954478709607e-05, "loss": 0.0422, "step": 35050 }, { "grad_norm": 0.3201419413089752, "learning_rate": 7.726568636457103e-05, "loss": 0.0405, "step": 35060 }, { "grad_norm": 0.44119659066200256, "learning_rate": 7.725182496031153e-05, "loss": 0.0411, "step": 35070 }, { "grad_norm": 0.4456323981285095, "learning_rate": 7.723796057583341e-05, "loss": 0.0515, "step": 35080 }, { "grad_norm": 0.47753065824508667, "learning_rate": 7.722409321265287e-05, "loss": 0.0439, "step": 35090 }, { "grad_norm": 0.3937237560749054, "learning_rate": 7.721022287228645e-05, "loss": 0.0657, "step": 35100 }, { "grad_norm": 0.6620082855224609, "learning_rate": 7.719634955625094e-05, "loss": 0.0455, "step": 35110 }, { "grad_norm": 0.5369495749473572, "learning_rate": 7.718247326606355e-05, "loss": 0.044, "step": 35120 }, { "grad_norm": 0.4548152983188629, "learning_rate": 7.716859400324175e-05, "loss": 0.0414, "step": 35130 }, { "grad_norm": 0.4336451590061188, "learning_rate": 7.715471176930335e-05, "loss": 0.0724, "step": 35140 }, { "grad_norm": 0.6017077565193176, "learning_rate": 7.714082656576651e-05, "loss": 0.0437, "step": 35150 }, { "grad_norm": 0.3849048316478729, "learning_rate": 7.712693839414968e-05, "loss": 0.0426, "step": 35160 }, { "grad_norm": 2.355609655380249, "learning_rate": 7.711304725597164e-05, "loss": 0.0463, "step": 35170 }, { "grad_norm": 0.5531862378120422, "learning_rate": 7.709915315275151e-05, "loss": 0.0435, "step": 35180 }, { "grad_norm": 0.3557113707065582, "learning_rate": 7.708525608600876e-05, "loss": 0.0404, "step": 35190 }, { "grad_norm": 0.40591877698898315, "learning_rate": 7.707135605726311e-05, "loss": 0.0552, "step": 35200 }, { "grad_norm": 0.43230199813842773, "learning_rate": 7.705745306803466e-05, "loss": 0.07, "step": 35210 }, { "grad_norm": 0.34650492668151855, "learning_rate": 7.704354711984383e-05, "loss": 0.0425, "step": 35220 }, { "grad_norm": 0.44843727350234985, "learning_rate": 7.702963821421133e-05, "loss": 0.0449, "step": 35230 }, { "grad_norm": 0.26010000705718994, "learning_rate": 7.701572635265826e-05, "loss": 0.0463, "step": 35240 }, { "grad_norm": 0.3355829417705536, "learning_rate": 7.700181153670596e-05, "loss": 0.0337, "step": 35250 }, { "grad_norm": 0.49612513184547424, "learning_rate": 7.698789376787616e-05, "loss": 0.0386, "step": 35260 }, { "grad_norm": 0.40758273005485535, "learning_rate": 7.697397304769085e-05, "loss": 0.0503, "step": 35270 }, { "grad_norm": 0.44510921835899353, "learning_rate": 7.696004937767241e-05, "loss": 0.0553, "step": 35280 }, { "grad_norm": 0.43240076303482056, "learning_rate": 7.694612275934352e-05, "loss": 0.051, "step": 35290 }, { "grad_norm": 0.46026182174682617, "learning_rate": 7.693219319422714e-05, "loss": 0.0447, "step": 35300 }, { "grad_norm": 1.2860932350158691, "learning_rate": 7.69182606838466e-05, "loss": 0.0454, "step": 35310 }, { "grad_norm": 0.3362618386745453, "learning_rate": 7.690432522972558e-05, "loss": 0.0426, "step": 35320 }, { "grad_norm": 0.43004027009010315, "learning_rate": 7.689038683338796e-05, "loss": 0.0416, "step": 35330 }, { "grad_norm": 0.36883577704429626, "learning_rate": 7.687644549635808e-05, "loss": 0.0338, "step": 35340 }, { "grad_norm": 0.35750773549079895, "learning_rate": 7.686250122016053e-05, "loss": 0.0383, "step": 35350 }, { "grad_norm": 0.39656949043273926, "learning_rate": 7.684855400632022e-05, "loss": 0.0407, "step": 35360 }, { "grad_norm": 0.4387245774269104, "learning_rate": 7.683460385636243e-05, "loss": 0.0374, "step": 35370 }, { "grad_norm": 0.39503318071365356, "learning_rate": 7.68206507718127e-05, "loss": 0.0463, "step": 35380 }, { "grad_norm": 0.3331753611564636, "learning_rate": 7.680669475419693e-05, "loss": 0.0471, "step": 35390 }, { "grad_norm": 0.5340307950973511, "learning_rate": 7.679273580504132e-05, "loss": 0.0482, "step": 35400 }, { "grad_norm": 0.4296788275241852, "learning_rate": 7.67787739258724e-05, "loss": 0.0388, "step": 35410 }, { "grad_norm": 0.39093828201293945, "learning_rate": 7.676480911821705e-05, "loss": 0.0446, "step": 35420 }, { "grad_norm": 0.6470893621444702, "learning_rate": 7.675084138360238e-05, "loss": 0.0538, "step": 35430 }, { "grad_norm": 0.59540855884552, "learning_rate": 7.673687072355592e-05, "loss": 0.0504, "step": 35440 }, { "grad_norm": 0.5871368050575256, "learning_rate": 7.67228971396055e-05, "loss": 0.0503, "step": 35450 }, { "grad_norm": 0.4219570755958557, "learning_rate": 7.670892063327922e-05, "loss": 0.049, "step": 35460 }, { "grad_norm": 0.6185372471809387, "learning_rate": 7.669494120610552e-05, "loss": 0.0462, "step": 35470 }, { "grad_norm": 0.39121881127357483, "learning_rate": 7.66809588596132e-05, "loss": 0.0389, "step": 35480 }, { "grad_norm": 0.5610802173614502, "learning_rate": 7.666697359533132e-05, "loss": 0.0444, "step": 35490 }, { "grad_norm": 0.5031247735023499, "learning_rate": 7.665298541478932e-05, "loss": 0.051, "step": 35500 }, { "grad_norm": 0.42719778418540955, "learning_rate": 7.66389943195169e-05, "loss": 0.0443, "step": 35510 }, { "grad_norm": 0.38719743490219116, "learning_rate": 7.66250003110441e-05, "loss": 0.0597, "step": 35520 }, { "grad_norm": 0.7817380428314209, "learning_rate": 7.66110033909013e-05, "loss": 0.0427, "step": 35530 }, { "grad_norm": 0.3524607717990875, "learning_rate": 7.659700356061918e-05, "loss": 0.0396, "step": 35540 }, { "grad_norm": 0.39112135767936707, "learning_rate": 7.658300082172875e-05, "loss": 0.0337, "step": 35550 }, { "grad_norm": 0.640751302242279, "learning_rate": 7.65689951757613e-05, "loss": 0.0437, "step": 35560 }, { "grad_norm": 0.32938840985298157, "learning_rate": 7.655498662424849e-05, "loss": 0.0384, "step": 35570 }, { "grad_norm": 0.44630271196365356, "learning_rate": 7.654097516872227e-05, "loss": 0.0472, "step": 35580 }, { "grad_norm": 0.5394911170005798, "learning_rate": 7.65269608107149e-05, "loss": 0.0482, "step": 35590 }, { "grad_norm": 0.9913168549537659, "learning_rate": 7.6512943551759e-05, "loss": 0.0711, "step": 35600 }, { "grad_norm": 0.5634380578994751, "learning_rate": 7.649892339338743e-05, "loss": 0.0475, "step": 35610 }, { "grad_norm": 0.9037850499153137, "learning_rate": 7.648490033713344e-05, "loss": 0.0567, "step": 35620 }, { "grad_norm": 0.2556670308113098, "learning_rate": 7.647087438453058e-05, "loss": 0.0439, "step": 35630 }, { "grad_norm": 0.7018770575523376, "learning_rate": 7.64568455371127e-05, "loss": 0.0513, "step": 35640 }, { "grad_norm": 0.45014694333076477, "learning_rate": 7.644281379641396e-05, "loss": 0.0402, "step": 35650 }, { "grad_norm": 0.3938494026660919, "learning_rate": 7.642877916396887e-05, "loss": 0.0486, "step": 35660 }, { "grad_norm": 0.4590975046157837, "learning_rate": 7.641474164131221e-05, "loss": 0.0453, "step": 35670 }, { "grad_norm": 0.36927878856658936, "learning_rate": 7.640070122997913e-05, "loss": 0.0395, "step": 35680 }, { "grad_norm": 0.44435009360313416, "learning_rate": 7.638665793150505e-05, "loss": 0.0419, "step": 35690 }, { "grad_norm": 1.0092616081237793, "learning_rate": 7.637261174742574e-05, "loss": 0.0508, "step": 35700 }, { "grad_norm": 0.4725804626941681, "learning_rate": 7.635856267927725e-05, "loss": 0.0376, "step": 35710 }, { "grad_norm": 0.3738217353820801, "learning_rate": 7.634451072859597e-05, "loss": 0.0393, "step": 35720 }, { "grad_norm": 0.35246872901916504, "learning_rate": 7.633045589691863e-05, "loss": 0.0466, "step": 35730 }, { "grad_norm": 0.5757357478141785, "learning_rate": 7.63163981857822e-05, "loss": 0.04, "step": 35740 }, { "grad_norm": 0.42283353209495544, "learning_rate": 7.630233759672403e-05, "loss": 0.0424, "step": 35750 }, { "grad_norm": 0.49516624212265015, "learning_rate": 7.628827413128177e-05, "loss": 0.0442, "step": 35760 }, { "grad_norm": 0.5928220748901367, "learning_rate": 7.627420779099338e-05, "loss": 0.0453, "step": 35770 }, { "grad_norm": 0.5805806517601013, "learning_rate": 7.626013857739711e-05, "loss": 0.0496, "step": 35780 }, { "grad_norm": 0.3867015838623047, "learning_rate": 7.624606649203158e-05, "loss": 0.04, "step": 35790 }, { "grad_norm": 0.32555797696113586, "learning_rate": 7.623199153643569e-05, "loss": 0.049, "step": 35800 }, { "grad_norm": 0.32038530707359314, "learning_rate": 7.621791371214863e-05, "loss": 0.0478, "step": 35810 }, { "grad_norm": 0.481538325548172, "learning_rate": 7.620383302070995e-05, "loss": 0.0452, "step": 35820 }, { "grad_norm": 0.6334377527236938, "learning_rate": 7.61897494636595e-05, "loss": 0.0467, "step": 35830 }, { "grad_norm": 0.30631282925605774, "learning_rate": 7.617566304253739e-05, "loss": 0.0611, "step": 35840 }, { "grad_norm": 0.3828947842121124, "learning_rate": 7.616157375888416e-05, "loss": 0.0425, "step": 35850 }, { "grad_norm": 0.4905940294265747, "learning_rate": 7.614748161424053e-05, "loss": 0.0408, "step": 35860 }, { "grad_norm": 0.3348868787288666, "learning_rate": 7.613338661014763e-05, "loss": 0.0406, "step": 35870 }, { "grad_norm": 0.4186111092567444, "learning_rate": 7.611928874814686e-05, "loss": 0.0459, "step": 35880 }, { "grad_norm": 0.39141374826431274, "learning_rate": 7.610518802977996e-05, "loss": 0.0414, "step": 35890 }, { "grad_norm": 0.39705750346183777, "learning_rate": 7.609108445658893e-05, "loss": 0.0477, "step": 35900 }, { "grad_norm": 0.3845829665660858, "learning_rate": 7.607697803011612e-05, "loss": 0.0588, "step": 35910 }, { "grad_norm": 0.3397459387779236, "learning_rate": 7.606286875190421e-05, "loss": 0.0406, "step": 35920 }, { "grad_norm": 0.5465185046195984, "learning_rate": 7.604875662349617e-05, "loss": 0.0402, "step": 35930 }, { "grad_norm": 0.425983726978302, "learning_rate": 7.603464164643525e-05, "loss": 0.0522, "step": 35940 }, { "grad_norm": 0.611186146736145, "learning_rate": 7.602052382226507e-05, "loss": 0.0434, "step": 35950 }, { "grad_norm": 1.62742018699646, "learning_rate": 7.600640315252954e-05, "loss": 0.0453, "step": 35960 }, { "grad_norm": 0.4805527627468109, "learning_rate": 7.599227963877284e-05, "loss": 0.0465, "step": 35970 }, { "grad_norm": 0.5568427443504333, "learning_rate": 7.597815328253954e-05, "loss": 0.0429, "step": 35980 }, { "grad_norm": 0.4271458685398102, "learning_rate": 7.596402408537443e-05, "loss": 0.0418, "step": 35990 }, { "grad_norm": 0.40073567628860474, "learning_rate": 7.59498920488227e-05, "loss": 0.0581, "step": 36000 }, { "grad_norm": 0.6341190934181213, "learning_rate": 7.593575717442979e-05, "loss": 0.0612, "step": 36010 }, { "grad_norm": 0.5426791310310364, "learning_rate": 7.592161946374147e-05, "loss": 0.0449, "step": 36020 }, { "grad_norm": 1.0443588495254517, "learning_rate": 7.590747891830381e-05, "loss": 0.0446, "step": 36030 }, { "grad_norm": 0.4929226338863373, "learning_rate": 7.58933355396632e-05, "loss": 0.0408, "step": 36040 }, { "grad_norm": 0.5674669146537781, "learning_rate": 7.587918932936636e-05, "loss": 0.0415, "step": 36050 }, { "grad_norm": 0.3649580478668213, "learning_rate": 7.586504028896028e-05, "loss": 0.0371, "step": 36060 }, { "grad_norm": 0.3878961503505707, "learning_rate": 7.585088841999228e-05, "loss": 0.0465, "step": 36070 }, { "grad_norm": 1.2068167924880981, "learning_rate": 7.583673372400999e-05, "loss": 0.0501, "step": 36080 }, { "grad_norm": 0.4188520014286041, "learning_rate": 7.582257620256134e-05, "loss": 0.0382, "step": 36090 }, { "grad_norm": 0.5292372703552246, "learning_rate": 7.580841585719458e-05, "loss": 0.0533, "step": 36100 }, { "grad_norm": 1.1113801002502441, "learning_rate": 7.579425268945825e-05, "loss": 0.0413, "step": 36110 }, { "grad_norm": 0.5662904381752014, "learning_rate": 7.578008670090127e-05, "loss": 0.0367, "step": 36120 }, { "grad_norm": 0.35307812690734863, "learning_rate": 7.576591789307272e-05, "loss": 0.0396, "step": 36130 }, { "grad_norm": 0.38657528162002563, "learning_rate": 7.575174626752216e-05, "loss": 0.0432, "step": 36140 }, { "grad_norm": 0.6890565156936646, "learning_rate": 7.573757182579934e-05, "loss": 0.0512, "step": 36150 }, { "grad_norm": 0.32855838537216187, "learning_rate": 7.572339456945435e-05, "loss": 0.039, "step": 36160 }, { "grad_norm": 0.41126638650894165, "learning_rate": 7.570921450003762e-05, "loss": 0.0459, "step": 36170 }, { "grad_norm": 0.4501020014286041, "learning_rate": 7.569503161909983e-05, "loss": 0.0471, "step": 36180 }, { "grad_norm": 0.561722457408905, "learning_rate": 7.568084592819202e-05, "loss": 0.0485, "step": 36190 }, { "grad_norm": 0.36606088280677795, "learning_rate": 7.566665742886551e-05, "loss": 0.0437, "step": 36200 }, { "grad_norm": 0.6452077031135559, "learning_rate": 7.565246612267194e-05, "loss": 0.0543, "step": 36210 }, { "grad_norm": 0.39661550521850586, "learning_rate": 7.563827201116325e-05, "loss": 0.0414, "step": 36220 }, { "grad_norm": 0.4788404107093811, "learning_rate": 7.562407509589164e-05, "loss": 0.0457, "step": 36230 }, { "grad_norm": 0.3484184741973877, "learning_rate": 7.560987537840973e-05, "loss": 0.0423, "step": 36240 }, { "grad_norm": 0.3256628215312958, "learning_rate": 7.559567286027036e-05, "loss": 0.0433, "step": 36250 }, { "grad_norm": 0.5279515981674194, "learning_rate": 7.558146754302668e-05, "loss": 0.0491, "step": 36260 }, { "grad_norm": 0.35306045413017273, "learning_rate": 7.556725942823217e-05, "loss": 0.0446, "step": 36270 }, { "grad_norm": 0.44633275270462036, "learning_rate": 7.555304851744061e-05, "loss": 0.0492, "step": 36280 }, { "grad_norm": 0.4612831771373749, "learning_rate": 7.553883481220608e-05, "loss": 0.0602, "step": 36290 }, { "grad_norm": 0.6615542769432068, "learning_rate": 7.552461831408298e-05, "loss": 0.0436, "step": 36300 }, { "grad_norm": 0.44230568408966064, "learning_rate": 7.551039902462599e-05, "loss": 0.0463, "step": 36310 }, { "grad_norm": 0.3320310115814209, "learning_rate": 7.549617694539011e-05, "loss": 0.0402, "step": 36320 }, { "grad_norm": 0.39709606766700745, "learning_rate": 7.548195207793066e-05, "loss": 0.0522, "step": 36330 }, { "grad_norm": 0.44662708044052124, "learning_rate": 7.546772442380323e-05, "loss": 0.0503, "step": 36340 }, { "grad_norm": 0.37352755665779114, "learning_rate": 7.545349398456376e-05, "loss": 0.0436, "step": 36350 }, { "grad_norm": 0.48745065927505493, "learning_rate": 7.543926076176845e-05, "loss": 0.0533, "step": 36360 }, { "grad_norm": 0.5992931723594666, "learning_rate": 7.542502475697385e-05, "loss": 0.0481, "step": 36370 }, { "grad_norm": 0.4090680480003357, "learning_rate": 7.541078597173675e-05, "loss": 0.0525, "step": 36380 }, { "grad_norm": 0.40944772958755493, "learning_rate": 7.539654440761431e-05, "loss": 0.0372, "step": 36390 }, { "grad_norm": 0.47309279441833496, "learning_rate": 7.538230006616395e-05, "loss": 0.0481, "step": 36400 }, { "grad_norm": 0.3371293842792511, "learning_rate": 7.536805294894342e-05, "loss": 0.0436, "step": 36410 }, { "grad_norm": 0.5769063830375671, "learning_rate": 7.535380305751076e-05, "loss": 0.0509, "step": 36420 }, { "grad_norm": 0.5481953024864197, "learning_rate": 7.533955039342431e-05, "loss": 0.052, "step": 36430 }, { "grad_norm": 0.34347453713417053, "learning_rate": 7.532529495824274e-05, "loss": 0.046, "step": 36440 }, { "grad_norm": 0.4149847626686096, "learning_rate": 7.531103675352497e-05, "loss": 0.0486, "step": 36450 }, { "grad_norm": 0.39427050948143005, "learning_rate": 7.529677578083029e-05, "loss": 0.0492, "step": 36460 }, { "grad_norm": 0.4851242005825043, "learning_rate": 7.528251204171823e-05, "loss": 0.0325, "step": 36470 }, { "grad_norm": 0.4296781122684479, "learning_rate": 7.526824553774867e-05, "loss": 0.0448, "step": 36480 }, { "grad_norm": 0.29106011986732483, "learning_rate": 7.525397627048178e-05, "loss": 0.0447, "step": 36490 }, { "grad_norm": 0.7492936849594116, "learning_rate": 7.523970424147802e-05, "loss": 0.0367, "step": 36500 }, { "grad_norm": 0.7731438875198364, "learning_rate": 7.522542945229813e-05, "loss": 0.0427, "step": 36510 }, { "grad_norm": 0.31474772095680237, "learning_rate": 7.521115190450321e-05, "loss": 0.0448, "step": 36520 }, { "grad_norm": 0.3485431373119354, "learning_rate": 7.519687159965462e-05, "loss": 0.0401, "step": 36530 }, { "grad_norm": 0.40749895572662354, "learning_rate": 7.518258853931403e-05, "loss": 0.0365, "step": 36540 }, { "grad_norm": 0.44107142090797424, "learning_rate": 7.516830272504342e-05, "loss": 0.0453, "step": 36550 }, { "grad_norm": 0.5845305323600769, "learning_rate": 7.515401415840509e-05, "loss": 0.0507, "step": 36560 }, { "grad_norm": 0.5280250906944275, "learning_rate": 7.513972284096155e-05, "loss": 0.0388, "step": 36570 }, { "grad_norm": 0.6043676137924194, "learning_rate": 7.512542877427574e-05, "loss": 0.0563, "step": 36580 }, { "grad_norm": 0.4639056622982025, "learning_rate": 7.51111319599108e-05, "loss": 0.0585, "step": 36590 }, { "grad_norm": 0.5167993307113647, "learning_rate": 7.509683239943024e-05, "loss": 0.0478, "step": 36600 }, { "grad_norm": 0.5225712060928345, "learning_rate": 7.50825300943978e-05, "loss": 0.0469, "step": 36610 }, { "grad_norm": 0.5396139025688171, "learning_rate": 7.50682250463776e-05, "loss": 0.047, "step": 36620 }, { "grad_norm": 0.9694468975067139, "learning_rate": 7.5053917256934e-05, "loss": 0.0555, "step": 36630 }, { "grad_norm": 0.3868545591831207, "learning_rate": 7.503960672763166e-05, "loss": 0.0525, "step": 36640 }, { "grad_norm": 0.3897464871406555, "learning_rate": 7.502529346003559e-05, "loss": 0.0447, "step": 36650 }, { "grad_norm": 0.40775758028030396, "learning_rate": 7.501097745571106e-05, "loss": 0.0501, "step": 36660 }, { "grad_norm": 0.4634154736995697, "learning_rate": 7.499665871622361e-05, "loss": 0.0596, "step": 36670 }, { "grad_norm": 0.35588738322257996, "learning_rate": 7.498233724313917e-05, "loss": 0.0601, "step": 36680 }, { "grad_norm": 0.4314947724342346, "learning_rate": 7.496801303802387e-05, "loss": 0.047, "step": 36690 }, { "grad_norm": 0.40212565660476685, "learning_rate": 7.495368610244423e-05, "loss": 0.0434, "step": 36700 }, { "grad_norm": 0.4959561228752136, "learning_rate": 7.493935643796697e-05, "loss": 0.0412, "step": 36710 }, { "grad_norm": 0.4675660729408264, "learning_rate": 7.49250240461592e-05, "loss": 0.0449, "step": 36720 }, { "grad_norm": 0.41581711173057556, "learning_rate": 7.491068892858826e-05, "loss": 0.0428, "step": 36730 }, { "grad_norm": 0.3355027735233307, "learning_rate": 7.489635108682184e-05, "loss": 0.0401, "step": 36740 }, { "grad_norm": 0.4696669280529022, "learning_rate": 7.48820105224279e-05, "loss": 0.042, "step": 36750 }, { "grad_norm": 0.5284770131111145, "learning_rate": 7.48676672369747e-05, "loss": 0.043, "step": 36760 }, { "grad_norm": 0.4263695478439331, "learning_rate": 7.485332123203079e-05, "loss": 0.0413, "step": 36770 }, { "grad_norm": 0.2615886330604553, "learning_rate": 7.483897250916504e-05, "loss": 0.0462, "step": 36780 }, { "grad_norm": 0.418941855430603, "learning_rate": 7.48246210699466e-05, "loss": 0.0435, "step": 36790 }, { "grad_norm": 0.37916260957717896, "learning_rate": 7.481026691594492e-05, "loss": 0.0407, "step": 36800 }, { "grad_norm": 0.290301650762558, "learning_rate": 7.479591004872975e-05, "loss": 0.0466, "step": 36810 }, { "grad_norm": 0.5188556909561157, "learning_rate": 7.478155046987114e-05, "loss": 0.046, "step": 36820 }, { "grad_norm": 0.5039433240890503, "learning_rate": 7.476718818093944e-05, "loss": 0.0563, "step": 36830 }, { "grad_norm": 0.9051039218902588, "learning_rate": 7.475282318350527e-05, "loss": 0.0453, "step": 36840 }, { "grad_norm": 0.3809075355529785, "learning_rate": 7.473845547913957e-05, "loss": 0.0475, "step": 36850 }, { "grad_norm": 0.9391749501228333, "learning_rate": 7.472408506941357e-05, "loss": 0.0583, "step": 36860 }, { "grad_norm": 0.40658342838287354, "learning_rate": 7.470971195589881e-05, "loss": 0.0441, "step": 36870 }, { "grad_norm": 0.5240753293037415, "learning_rate": 7.46953361401671e-05, "loss": 0.0601, "step": 36880 }, { "grad_norm": 0.6370009779930115, "learning_rate": 7.468095762379055e-05, "loss": 0.0488, "step": 36890 }, { "grad_norm": 0.4580635130405426, "learning_rate": 7.466657640834158e-05, "loss": 0.0425, "step": 36900 }, { "grad_norm": 0.41636836528778076, "learning_rate": 7.46521924953929e-05, "loss": 0.0416, "step": 36910 }, { "grad_norm": 0.5001466870307922, "learning_rate": 7.463780588651752e-05, "loss": 0.043, "step": 36920 }, { "grad_norm": 1.03081214427948, "learning_rate": 7.462341658328873e-05, "loss": 0.0507, "step": 36930 }, { "grad_norm": 0.45853492617607117, "learning_rate": 7.460902458728012e-05, "loss": 0.0506, "step": 36940 }, { "grad_norm": 0.469999760389328, "learning_rate": 7.459462990006558e-05, "loss": 0.047, "step": 36950 }, { "grad_norm": 0.3776153326034546, "learning_rate": 7.458023252321929e-05, "loss": 0.0475, "step": 36960 }, { "grad_norm": 0.5772284269332886, "learning_rate": 7.456583245831573e-05, "loss": 0.0403, "step": 36970 }, { "grad_norm": 0.37670353055000305, "learning_rate": 7.455142970692968e-05, "loss": 0.0572, "step": 36980 }, { "grad_norm": 0.5110841989517212, "learning_rate": 7.453702427063615e-05, "loss": 0.0424, "step": 36990 }, { "grad_norm": 0.4208846092224121, "learning_rate": 7.452261615101057e-05, "loss": 0.0399, "step": 37000 }, { "grad_norm": 0.3908161222934723, "learning_rate": 7.450820534962856e-05, "loss": 0.0483, "step": 37010 }, { "grad_norm": 0.362376868724823, "learning_rate": 7.449379186806604e-05, "loss": 0.0455, "step": 37020 }, { "grad_norm": 0.596366822719574, "learning_rate": 7.44793757078993e-05, "loss": 0.0435, "step": 37030 }, { "grad_norm": 0.4462447464466095, "learning_rate": 7.44649568707048e-05, "loss": 0.0519, "step": 37040 }, { "grad_norm": 0.37302958965301514, "learning_rate": 7.445053535805942e-05, "loss": 0.0415, "step": 37050 }, { "grad_norm": 0.456971138715744, "learning_rate": 7.443611117154024e-05, "loss": 0.0421, "step": 37060 }, { "grad_norm": 0.3786298334598541, "learning_rate": 7.442168431272469e-05, "loss": 0.05, "step": 37070 }, { "grad_norm": 0.5028876662254333, "learning_rate": 7.440725478319044e-05, "loss": 0.0366, "step": 37080 }, { "grad_norm": 0.3832760453224182, "learning_rate": 7.439282258451553e-05, "loss": 0.0466, "step": 37090 }, { "grad_norm": 0.4010971188545227, "learning_rate": 7.43783877182782e-05, "loss": 0.0534, "step": 37100 }, { "grad_norm": 0.48167431354522705, "learning_rate": 7.436395018605707e-05, "loss": 0.0447, "step": 37110 }, { "grad_norm": 0.3215634822845459, "learning_rate": 7.434950998943094e-05, "loss": 0.038, "step": 37120 }, { "grad_norm": 0.5448696613311768, "learning_rate": 7.433506712997902e-05, "loss": 0.0544, "step": 37130 }, { "grad_norm": 0.3918777108192444, "learning_rate": 7.432062160928077e-05, "loss": 0.0374, "step": 37140 }, { "grad_norm": 0.32846927642822266, "learning_rate": 7.430617342891588e-05, "loss": 0.0376, "step": 37150 }, { "grad_norm": 0.49107232689857483, "learning_rate": 7.429172259046442e-05, "loss": 0.0484, "step": 37160 }, { "grad_norm": 0.38386112451553345, "learning_rate": 7.42772690955067e-05, "loss": 0.0369, "step": 37170 }, { "grad_norm": 0.4719637632369995, "learning_rate": 7.426281294562332e-05, "loss": 0.0502, "step": 37180 }, { "grad_norm": 0.4821794927120209, "learning_rate": 7.424835414239519e-05, "loss": 0.0441, "step": 37190 }, { "grad_norm": 0.6018915772438049, "learning_rate": 7.423389268740352e-05, "loss": 0.0396, "step": 37200 }, { "grad_norm": 0.3830341100692749, "learning_rate": 7.42194285822298e-05, "loss": 0.0395, "step": 37210 }, { "grad_norm": 0.46303024888038635, "learning_rate": 7.420496182845576e-05, "loss": 0.0393, "step": 37220 }, { "grad_norm": 0.34999728202819824, "learning_rate": 7.419049242766348e-05, "loss": 0.0476, "step": 37230 }, { "grad_norm": 0.5468374490737915, "learning_rate": 7.417602038143533e-05, "loss": 0.0504, "step": 37240 }, { "grad_norm": 0.6013652682304382, "learning_rate": 7.416154569135393e-05, "loss": 0.0431, "step": 37250 }, { "grad_norm": 0.37281447649002075, "learning_rate": 7.414706835900222e-05, "loss": 0.0417, "step": 37260 }, { "grad_norm": 0.5254356861114502, "learning_rate": 7.413258838596345e-05, "loss": 0.0428, "step": 37270 }, { "grad_norm": 0.3901149332523346, "learning_rate": 7.411810577382106e-05, "loss": 0.045, "step": 37280 }, { "grad_norm": 0.4940224587917328, "learning_rate": 7.410362052415889e-05, "loss": 0.0509, "step": 37290 }, { "grad_norm": 0.3523554503917694, "learning_rate": 7.408913263856102e-05, "loss": 0.0441, "step": 37300 }, { "grad_norm": 0.43197739124298096, "learning_rate": 7.407464211861183e-05, "loss": 0.0453, "step": 37310 }, { "grad_norm": 0.4963345229625702, "learning_rate": 7.406014896589597e-05, "loss": 0.06, "step": 37320 }, { "grad_norm": 0.4536018371582031, "learning_rate": 7.404565318199838e-05, "loss": 0.049, "step": 37330 }, { "grad_norm": 0.35210680961608887, "learning_rate": 7.403115476850432e-05, "loss": 0.0456, "step": 37340 }, { "grad_norm": 0.4842255711555481, "learning_rate": 7.401665372699932e-05, "loss": 0.0388, "step": 37350 }, { "grad_norm": 0.4526502788066864, "learning_rate": 7.400215005906916e-05, "loss": 0.0467, "step": 37360 }, { "grad_norm": 0.3986954092979431, "learning_rate": 7.398764376629996e-05, "loss": 0.0381, "step": 37370 }, { "grad_norm": 0.5663161277770996, "learning_rate": 7.39731348502781e-05, "loss": 0.0447, "step": 37380 }, { "grad_norm": 1.2986201047897339, "learning_rate": 7.395862331259028e-05, "loss": 0.0333, "step": 37390 }, { "grad_norm": 0.33474549651145935, "learning_rate": 7.394410915482341e-05, "loss": 0.0388, "step": 37400 }, { "grad_norm": 1.1980867385864258, "learning_rate": 7.392959237856476e-05, "loss": 0.0543, "step": 37410 }, { "grad_norm": 0.5134378671646118, "learning_rate": 7.391507298540187e-05, "loss": 0.0447, "step": 37420 }, { "grad_norm": 0.4735996127128601, "learning_rate": 7.390055097692257e-05, "loss": 0.037, "step": 37430 }, { "grad_norm": 0.5878653526306152, "learning_rate": 7.388602635471494e-05, "loss": 0.059, "step": 37440 }, { "grad_norm": 0.5941770076751709, "learning_rate": 7.387149912036738e-05, "loss": 0.0661, "step": 37450 }, { "grad_norm": 0.6577693223953247, "learning_rate": 7.385696927546854e-05, "loss": 0.0428, "step": 37460 }, { "grad_norm": 0.6160150766372681, "learning_rate": 7.384243682160743e-05, "loss": 0.0611, "step": 37470 }, { "grad_norm": 0.430791974067688, "learning_rate": 7.382790176037329e-05, "loss": 0.0454, "step": 37480 }, { "grad_norm": 0.5949933528900146, "learning_rate": 7.381336409335562e-05, "loss": 0.0527, "step": 37490 }, { "grad_norm": 0.5236151814460754, "learning_rate": 7.379882382214426e-05, "loss": 0.0507, "step": 37500 }, { "grad_norm": 0.3960215449333191, "learning_rate": 7.378428094832931e-05, "loss": 0.0423, "step": 37510 }, { "grad_norm": 0.3993217349052429, "learning_rate": 7.376973547350114e-05, "loss": 0.0514, "step": 37520 }, { "grad_norm": 0.4113960862159729, "learning_rate": 7.375518739925046e-05, "loss": 0.0417, "step": 37530 }, { "grad_norm": 0.5185235738754272, "learning_rate": 7.374063672716818e-05, "loss": 0.0394, "step": 37540 }, { "grad_norm": 0.43053922057151794, "learning_rate": 7.372608345884558e-05, "loss": 0.043, "step": 37550 }, { "grad_norm": 0.6057499647140503, "learning_rate": 7.371152759587415e-05, "loss": 0.0489, "step": 37560 }, { "grad_norm": 0.4029698073863983, "learning_rate": 7.369696913984572e-05, "loss": 0.0429, "step": 37570 }, { "grad_norm": 0.5262873768806458, "learning_rate": 7.368240809235237e-05, "loss": 0.0383, "step": 37580 }, { "grad_norm": 0.39451947808265686, "learning_rate": 7.366784445498649e-05, "loss": 0.0446, "step": 37590 }, { "grad_norm": 0.4015263020992279, "learning_rate": 7.36532782293407e-05, "loss": 0.0403, "step": 37600 }, { "grad_norm": 0.6250522136688232, "learning_rate": 7.363870941700797e-05, "loss": 0.0471, "step": 37610 }, { "grad_norm": 0.44881850481033325, "learning_rate": 7.362413801958152e-05, "loss": 0.0388, "step": 37620 }, { "grad_norm": 0.36659613251686096, "learning_rate": 7.360956403865486e-05, "loss": 0.0414, "step": 37630 }, { "grad_norm": 0.5217164754867554, "learning_rate": 7.359498747582177e-05, "loss": 0.0395, "step": 37640 }, { "grad_norm": 0.30807745456695557, "learning_rate": 7.358040833267634e-05, "loss": 0.0486, "step": 37650 }, { "grad_norm": 0.5409486293792725, "learning_rate": 7.356582661081288e-05, "loss": 0.0517, "step": 37660 }, { "grad_norm": 0.42128798365592957, "learning_rate": 7.355124231182607e-05, "loss": 0.0443, "step": 37670 }, { "grad_norm": 0.41469645500183105, "learning_rate": 7.353665543731083e-05, "loss": 0.0501, "step": 37680 }, { "grad_norm": 0.4145365059375763, "learning_rate": 7.352206598886229e-05, "loss": 0.0355, "step": 37690 }, { "grad_norm": 0.5858395099639893, "learning_rate": 7.350747396807601e-05, "loss": 0.0403, "step": 37700 }, { "grad_norm": 0.4219357669353485, "learning_rate": 7.349287937654772e-05, "loss": 0.0397, "step": 37710 }, { "grad_norm": 0.7968283891677856, "learning_rate": 7.347828221587345e-05, "loss": 0.0407, "step": 37720 }, { "grad_norm": 0.4072430431842804, "learning_rate": 7.346368248764955e-05, "loss": 0.0492, "step": 37730 }, { "grad_norm": 0.3669516444206238, "learning_rate": 7.34490801934726e-05, "loss": 0.0391, "step": 37740 }, { "grad_norm": 0.3952712118625641, "learning_rate": 7.343447533493947e-05, "loss": 0.0534, "step": 37750 }, { "grad_norm": 0.4824649393558502, "learning_rate": 7.341986791364739e-05, "loss": 0.073, "step": 37760 }, { "grad_norm": 0.479615181684494, "learning_rate": 7.340525793119374e-05, "loss": 0.0514, "step": 37770 }, { "grad_norm": 0.5453259348869324, "learning_rate": 7.33906453891763e-05, "loss": 0.0411, "step": 37780 }, { "grad_norm": 0.6431543231010437, "learning_rate": 7.337603028919303e-05, "loss": 0.0457, "step": 37790 }, { "grad_norm": 0.6270379424095154, "learning_rate": 7.336141263284225e-05, "loss": 0.0371, "step": 37800 }, { "grad_norm": 0.44296228885650635, "learning_rate": 7.334679242172249e-05, "loss": 0.0515, "step": 37810 }, { "grad_norm": 0.3776143193244934, "learning_rate": 7.33321696574326e-05, "loss": 0.0479, "step": 37820 }, { "grad_norm": 0.4544045925140381, "learning_rate": 7.331754434157174e-05, "loss": 0.044, "step": 37830 }, { "grad_norm": 0.48851290345191956, "learning_rate": 7.330291647573928e-05, "loss": 0.0505, "step": 37840 }, { "grad_norm": 0.7865851521492004, "learning_rate": 7.32882860615349e-05, "loss": 0.0471, "step": 37850 }, { "grad_norm": 0.4097907543182373, "learning_rate": 7.327365310055858e-05, "loss": 0.0471, "step": 37860 }, { "grad_norm": 0.38243231177330017, "learning_rate": 7.325901759441055e-05, "loss": 0.0396, "step": 37870 }, { "grad_norm": 0.3681526184082031, "learning_rate": 7.324437954469132e-05, "loss": 0.0511, "step": 37880 }, { "grad_norm": 0.5621645450592041, "learning_rate": 7.32297389530017e-05, "loss": 0.041, "step": 37890 }, { "grad_norm": 0.41393741965293884, "learning_rate": 7.321509582094274e-05, "loss": 0.0337, "step": 37900 }, { "grad_norm": 0.36056390404701233, "learning_rate": 7.320045015011582e-05, "loss": 0.043, "step": 37910 }, { "grad_norm": 0.3890555500984192, "learning_rate": 7.318580194212253e-05, "loss": 0.0547, "step": 37920 }, { "grad_norm": 0.44089120626449585, "learning_rate": 7.317115119856483e-05, "loss": 0.0571, "step": 37930 }, { "grad_norm": 0.7616183757781982, "learning_rate": 7.315649792104485e-05, "loss": 0.0442, "step": 37940 }, { "grad_norm": 0.5931852459907532, "learning_rate": 7.314184211116507e-05, "loss": 0.0506, "step": 37950 }, { "grad_norm": 0.4053160846233368, "learning_rate": 7.312718377052823e-05, "loss": 0.0453, "step": 37960 }, { "grad_norm": 0.2654139995574951, "learning_rate": 7.311252290073737e-05, "loss": 0.0381, "step": 37970 }, { "grad_norm": 0.5318470001220703, "learning_rate": 7.309785950339572e-05, "loss": 0.037, "step": 37980 }, { "grad_norm": 0.43515074253082275, "learning_rate": 7.308319358010692e-05, "loss": 0.0367, "step": 37990 }, { "grad_norm": 0.4300527274608612, "learning_rate": 7.306852513247474e-05, "loss": 0.0581, "step": 38000 }, { "grad_norm": 0.7034837603569031, "learning_rate": 7.305385416210336e-05, "loss": 0.0518, "step": 38010 }, { "grad_norm": 0.37210872769355774, "learning_rate": 7.303918067059711e-05, "loss": 0.037, "step": 38020 }, { "grad_norm": 0.7331014275550842, "learning_rate": 7.302450465956075e-05, "loss": 0.0443, "step": 38030 }, { "grad_norm": 0.5060908794403076, "learning_rate": 7.300982613059914e-05, "loss": 0.0438, "step": 38040 }, { "grad_norm": 0.36828723549842834, "learning_rate": 7.299514508531757e-05, "loss": 0.0533, "step": 38050 }, { "grad_norm": 0.580774188041687, "learning_rate": 7.298046152532148e-05, "loss": 0.0543, "step": 38060 }, { "grad_norm": 0.28430208563804626, "learning_rate": 7.296577545221668e-05, "loss": 0.0361, "step": 38070 }, { "grad_norm": 0.4021552503108978, "learning_rate": 7.29510868676092e-05, "loss": 0.0534, "step": 38080 }, { "grad_norm": 0.49474969506263733, "learning_rate": 7.293639577310538e-05, "loss": 0.0448, "step": 38090 }, { "grad_norm": 0.4512941241264343, "learning_rate": 7.292170217031179e-05, "loss": 0.0442, "step": 38100 }, { "grad_norm": 0.7207484841346741, "learning_rate": 7.290700606083532e-05, "loss": 0.0441, "step": 38110 }, { "grad_norm": 0.3438278138637543, "learning_rate": 7.28923074462831e-05, "loss": 0.0448, "step": 38120 }, { "grad_norm": 0.6711441278457642, "learning_rate": 7.287760632826255e-05, "loss": 0.0492, "step": 38130 }, { "grad_norm": 0.6909121870994568, "learning_rate": 7.286290270838139e-05, "loss": 0.0594, "step": 38140 }, { "grad_norm": 0.4179113805294037, "learning_rate": 7.284819658824756e-05, "loss": 0.0358, "step": 38150 }, { "grad_norm": 0.4061858355998993, "learning_rate": 7.283348796946929e-05, "loss": 0.0411, "step": 38160 }, { "grad_norm": 0.4806639552116394, "learning_rate": 7.28187768536551e-05, "loss": 0.0418, "step": 38170 }, { "grad_norm": 0.6257554888725281, "learning_rate": 7.280406324241379e-05, "loss": 0.0565, "step": 38180 }, { "grad_norm": 0.323868066072464, "learning_rate": 7.278934713735441e-05, "loss": 0.0427, "step": 38190 }, { "grad_norm": 0.3677937984466553, "learning_rate": 7.277462854008629e-05, "loss": 0.0424, "step": 38200 }, { "grad_norm": 0.6301807761192322, "learning_rate": 7.275990745221904e-05, "loss": 0.0415, "step": 38210 }, { "grad_norm": 0.5120449662208557, "learning_rate": 7.274518387536252e-05, "loss": 0.0445, "step": 38220 }, { "grad_norm": 0.4802100360393524, "learning_rate": 7.273045781112689e-05, "loss": 0.0454, "step": 38230 }, { "grad_norm": 0.3803972005844116, "learning_rate": 7.271572926112257e-05, "loss": 0.0456, "step": 38240 }, { "grad_norm": 0.42954424023628235, "learning_rate": 7.270099822696024e-05, "loss": 0.0457, "step": 38250 }, { "grad_norm": 0.4472440183162689, "learning_rate": 7.26862647102509e-05, "loss": 0.0345, "step": 38260 }, { "grad_norm": 0.45379385352134705, "learning_rate": 7.267152871260573e-05, "loss": 0.0523, "step": 38270 }, { "grad_norm": 0.571619987487793, "learning_rate": 7.265679023563629e-05, "loss": 0.0407, "step": 38280 }, { "grad_norm": 0.33745527267456055, "learning_rate": 7.264204928095433e-05, "loss": 0.0274, "step": 38290 }, { "grad_norm": 0.4754400849342346, "learning_rate": 7.262730585017188e-05, "loss": 0.0461, "step": 38300 }, { "grad_norm": 0.38621747493743896, "learning_rate": 7.261255994490131e-05, "loss": 0.0408, "step": 38310 }, { "grad_norm": 0.3168965280056, "learning_rate": 7.259781156675517e-05, "loss": 0.0512, "step": 38320 }, { "grad_norm": 0.48019012808799744, "learning_rate": 7.258306071734633e-05, "loss": 0.0425, "step": 38330 }, { "grad_norm": 0.41572725772857666, "learning_rate": 7.256830739828792e-05, "loss": 0.041, "step": 38340 }, { "grad_norm": 0.3690830171108246, "learning_rate": 7.255355161119336e-05, "loss": 0.0361, "step": 38350 }, { "grad_norm": 0.5012595057487488, "learning_rate": 7.253879335767628e-05, "loss": 0.0374, "step": 38360 }, { "grad_norm": 0.28575336933135986, "learning_rate": 7.252403263935067e-05, "loss": 0.0384, "step": 38370 }, { "grad_norm": 0.5312322378158569, "learning_rate": 7.25092694578307e-05, "loss": 0.0524, "step": 38380 }, { "grad_norm": 0.4179936349391937, "learning_rate": 7.249450381473087e-05, "loss": 0.044, "step": 38390 }, { "grad_norm": 0.43109846115112305, "learning_rate": 7.247973571166593e-05, "loss": 0.0413, "step": 38400 }, { "grad_norm": 0.5664229393005371, "learning_rate": 7.246496515025089e-05, "loss": 0.0441, "step": 38410 }, { "grad_norm": 0.6231862902641296, "learning_rate": 7.245019213210101e-05, "loss": 0.0485, "step": 38420 }, { "grad_norm": 0.3859821557998657, "learning_rate": 7.24354166588319e-05, "loss": 0.0539, "step": 38430 }, { "grad_norm": 0.5511589050292969, "learning_rate": 7.242063873205935e-05, "loss": 0.0503, "step": 38440 }, { "grad_norm": 0.8109700083732605, "learning_rate": 7.240585835339946e-05, "loss": 0.0551, "step": 38450 }, { "grad_norm": 0.30949196219444275, "learning_rate": 7.239107552446857e-05, "loss": 0.0471, "step": 38460 }, { "grad_norm": 0.7956220507621765, "learning_rate": 7.237629024688334e-05, "loss": 0.0389, "step": 38470 }, { "grad_norm": 0.6329715847969055, "learning_rate": 7.236150252226064e-05, "loss": 0.0523, "step": 38480 }, { "grad_norm": 0.5207690000534058, "learning_rate": 7.234671235221765e-05, "loss": 0.0535, "step": 38490 }, { "grad_norm": 0.4613608717918396, "learning_rate": 7.233191973837179e-05, "loss": 0.0461, "step": 38500 }, { "grad_norm": 0.3176043927669525, "learning_rate": 7.231712468234078e-05, "loss": 0.0482, "step": 38510 }, { "grad_norm": 0.625577986240387, "learning_rate": 7.230232718574254e-05, "loss": 0.0405, "step": 38520 }, { "grad_norm": 0.4931753873825073, "learning_rate": 7.228752725019535e-05, "loss": 0.0536, "step": 38530 }, { "grad_norm": 0.43583741784095764, "learning_rate": 7.227272487731769e-05, "loss": 0.0443, "step": 38540 }, { "grad_norm": 0.4339413046836853, "learning_rate": 7.225792006872831e-05, "loss": 0.0421, "step": 38550 }, { "grad_norm": 0.2742650508880615, "learning_rate": 7.224311282604628e-05, "loss": 0.0388, "step": 38560 }, { "grad_norm": 0.42666590213775635, "learning_rate": 7.222830315089085e-05, "loss": 0.0494, "step": 38570 }, { "grad_norm": 0.3695038855075836, "learning_rate": 7.22134910448816e-05, "loss": 0.0396, "step": 38580 }, { "grad_norm": 0.5789576768875122, "learning_rate": 7.219867650963839e-05, "loss": 0.0444, "step": 38590 }, { "grad_norm": 0.5245236754417419, "learning_rate": 7.21838595467813e-05, "loss": 0.0523, "step": 38600 }, { "grad_norm": 0.2588041424751282, "learning_rate": 7.216904015793067e-05, "loss": 0.0387, "step": 38610 }, { "grad_norm": 0.30311140418052673, "learning_rate": 7.215421834470713e-05, "loss": 0.0419, "step": 38620 }, { "grad_norm": 0.3990594446659088, "learning_rate": 7.213939410873158e-05, "loss": 0.0453, "step": 38630 }, { "grad_norm": 0.5429508090019226, "learning_rate": 7.212456745162519e-05, "loss": 0.0474, "step": 38640 }, { "grad_norm": 0.4491405487060547, "learning_rate": 7.210973837500937e-05, "loss": 0.0471, "step": 38650 }, { "grad_norm": 0.4354325532913208, "learning_rate": 7.209490688050578e-05, "loss": 0.0423, "step": 38660 }, { "grad_norm": 0.35397961735725403, "learning_rate": 7.208007296973641e-05, "loss": 0.0397, "step": 38670 }, { "grad_norm": 0.373508483171463, "learning_rate": 7.206523664432345e-05, "loss": 0.0507, "step": 38680 }, { "grad_norm": 0.5220803022384644, "learning_rate": 7.205039790588939e-05, "loss": 0.0407, "step": 38690 }, { "grad_norm": 0.5374025702476501, "learning_rate": 7.203555675605697e-05, "loss": 0.0507, "step": 38700 }, { "grad_norm": 0.5654780864715576, "learning_rate": 7.202071319644917e-05, "loss": 0.0572, "step": 38710 }, { "grad_norm": 0.37984949350357056, "learning_rate": 7.200586722868932e-05, "loss": 0.0466, "step": 38720 }, { "grad_norm": 0.3367711305618286, "learning_rate": 7.199101885440088e-05, "loss": 0.0348, "step": 38730 }, { "grad_norm": 0.5783627033233643, "learning_rate": 7.197616807520771e-05, "loss": 0.046, "step": 38740 }, { "grad_norm": 0.33717241883277893, "learning_rate": 7.196131489273381e-05, "loss": 0.0393, "step": 38750 }, { "grad_norm": 0.38010531663894653, "learning_rate": 7.194645930860355e-05, "loss": 0.0395, "step": 38760 }, { "grad_norm": 0.9470652937889099, "learning_rate": 7.19316013244415e-05, "loss": 0.0465, "step": 38770 }, { "grad_norm": 0.45243123173713684, "learning_rate": 7.191674094187248e-05, "loss": 0.0424, "step": 38780 }, { "grad_norm": 0.587733268737793, "learning_rate": 7.190187816252165e-05, "loss": 0.0425, "step": 38790 }, { "grad_norm": 0.8528493642807007, "learning_rate": 7.188701298801435e-05, "loss": 0.0348, "step": 38800 }, { "grad_norm": 0.48377496004104614, "learning_rate": 7.18721454199762e-05, "loss": 0.0518, "step": 38810 }, { "grad_norm": 0.3616032898426056, "learning_rate": 7.185727546003312e-05, "loss": 0.044, "step": 38820 }, { "grad_norm": 0.4595969617366791, "learning_rate": 7.184240310981126e-05, "loss": 0.0485, "step": 38830 }, { "grad_norm": 0.46101102232933044, "learning_rate": 7.182752837093704e-05, "loss": 0.042, "step": 38840 }, { "grad_norm": 0.34666475653648376, "learning_rate": 7.181265124503711e-05, "loss": 0.0649, "step": 38850 }, { "grad_norm": 0.23937758803367615, "learning_rate": 7.179777173373847e-05, "loss": 0.0379, "step": 38860 }, { "grad_norm": 0.3851086497306824, "learning_rate": 7.178288983866826e-05, "loss": 0.0433, "step": 38870 }, { "grad_norm": 1.1096291542053223, "learning_rate": 7.176800556145397e-05, "loss": 0.0487, "step": 38880 }, { "grad_norm": 0.3205771744251251, "learning_rate": 7.175311890372334e-05, "loss": 0.0484, "step": 38890 }, { "grad_norm": 1.3217047452926636, "learning_rate": 7.17382298671043e-05, "loss": 0.0442, "step": 38900 }, { "grad_norm": 0.5575039982795715, "learning_rate": 7.172333845322515e-05, "loss": 0.0527, "step": 38910 }, { "grad_norm": 0.3104797303676605, "learning_rate": 7.170844466371436e-05, "loss": 0.0472, "step": 38920 }, { "grad_norm": 0.3549070358276367, "learning_rate": 7.16935485002007e-05, "loss": 0.0377, "step": 38930 }, { "grad_norm": 0.30850765109062195, "learning_rate": 7.167864996431319e-05, "loss": 0.0485, "step": 38940 }, { "grad_norm": 0.5021174550056458, "learning_rate": 7.166374905768111e-05, "loss": 0.0448, "step": 38950 }, { "grad_norm": 0.512803316116333, "learning_rate": 7.164884578193401e-05, "loss": 0.0473, "step": 38960 }, { "grad_norm": 0.4775283932685852, "learning_rate": 7.163394013870168e-05, "loss": 0.0434, "step": 38970 }, { "grad_norm": 0.47348496317863464, "learning_rate": 7.161903212961419e-05, "loss": 0.0534, "step": 38980 }, { "grad_norm": 0.2757556736469269, "learning_rate": 7.160412175630183e-05, "loss": 0.0518, "step": 38990 }, { "grad_norm": 0.3788091838359833, "learning_rate": 7.158920902039521e-05, "loss": 0.032, "step": 39000 }, { "grad_norm": 0.5068143606185913, "learning_rate": 7.157429392352514e-05, "loss": 0.0337, "step": 39010 }, { "grad_norm": 0.602124810218811, "learning_rate": 7.155937646732275e-05, "loss": 0.0413, "step": 39020 }, { "grad_norm": 0.4795437455177307, "learning_rate": 7.154445665341933e-05, "loss": 0.0389, "step": 39030 }, { "grad_norm": 0.47476503252983093, "learning_rate": 7.152953448344654e-05, "loss": 0.0404, "step": 39040 }, { "grad_norm": 0.5172357559204102, "learning_rate": 7.151460995903624e-05, "loss": 0.0392, "step": 39050 }, { "grad_norm": 0.38453131914138794, "learning_rate": 7.149968308182052e-05, "loss": 0.0396, "step": 39060 }, { "grad_norm": 0.5264629125595093, "learning_rate": 7.14847538534318e-05, "loss": 0.0489, "step": 39070 }, { "grad_norm": 0.4404030740261078, "learning_rate": 7.14698222755027e-05, "loss": 0.038, "step": 39080 }, { "grad_norm": 0.43617329001426697, "learning_rate": 7.14548883496661e-05, "loss": 0.0481, "step": 39090 }, { "grad_norm": 0.36924687027931213, "learning_rate": 7.143995207755517e-05, "loss": 0.047, "step": 39100 }, { "grad_norm": 0.5206234455108643, "learning_rate": 7.142501346080333e-05, "loss": 0.0511, "step": 39110 }, { "grad_norm": 0.3985595405101776, "learning_rate": 7.141007250104421e-05, "loss": 0.0466, "step": 39120 }, { "grad_norm": 0.5992079377174377, "learning_rate": 7.139512919991176e-05, "loss": 0.056, "step": 39130 }, { "grad_norm": 0.45744094252586365, "learning_rate": 7.138018355904014e-05, "loss": 0.0467, "step": 39140 }, { "grad_norm": 0.5642319321632385, "learning_rate": 7.13652355800638e-05, "loss": 0.0536, "step": 39150 }, { "grad_norm": 0.3627586364746094, "learning_rate": 7.135028526461743e-05, "loss": 0.0453, "step": 39160 }, { "grad_norm": 0.4421320855617523, "learning_rate": 7.133533261433594e-05, "loss": 0.0399, "step": 39170 }, { "grad_norm": 0.6037509441375732, "learning_rate": 7.132037763085457e-05, "loss": 0.0409, "step": 39180 }, { "grad_norm": 0.6936784982681274, "learning_rate": 7.130542031580875e-05, "loss": 0.0495, "step": 39190 }, { "grad_norm": 0.6097484230995178, "learning_rate": 7.12904606708342e-05, "loss": 0.0403, "step": 39200 }, { "grad_norm": 0.43495723605155945, "learning_rate": 7.127549869756687e-05, "loss": 0.0471, "step": 39210 }, { "grad_norm": 0.3766065537929535, "learning_rate": 7.126053439764299e-05, "loss": 0.0391, "step": 39220 }, { "grad_norm": 0.39263081550598145, "learning_rate": 7.124556777269904e-05, "loss": 0.055, "step": 39230 }, { "grad_norm": 0.3221723139286041, "learning_rate": 7.123059882437174e-05, "loss": 0.0384, "step": 39240 }, { "grad_norm": 0.3437437117099762, "learning_rate": 7.121562755429807e-05, "loss": 0.0355, "step": 39250 }, { "grad_norm": 0.33576175570487976, "learning_rate": 7.120065396411528e-05, "loss": 0.0427, "step": 39260 }, { "grad_norm": 0.44010186195373535, "learning_rate": 7.118567805546084e-05, "loss": 0.0356, "step": 39270 }, { "grad_norm": 0.4550255239009857, "learning_rate": 7.117069982997248e-05, "loss": 0.0476, "step": 39280 }, { "grad_norm": 0.6386981010437012, "learning_rate": 7.115571928928825e-05, "loss": 0.0715, "step": 39290 }, { "grad_norm": 0.5104573369026184, "learning_rate": 7.114073643504635e-05, "loss": 0.0518, "step": 39300 }, { "grad_norm": 0.44978147745132446, "learning_rate": 7.11257512688853e-05, "loss": 0.0397, "step": 39310 }, { "grad_norm": 0.4946425259113312, "learning_rate": 7.111076379244384e-05, "loss": 0.0397, "step": 39320 }, { "grad_norm": 0.2729858160018921, "learning_rate": 7.109577400736101e-05, "loss": 0.0373, "step": 39330 }, { "grad_norm": 0.4118731617927551, "learning_rate": 7.108078191527605e-05, "loss": 0.0392, "step": 39340 }, { "grad_norm": 0.36603620648384094, "learning_rate": 7.106578751782847e-05, "loss": 0.0542, "step": 39350 }, { "grad_norm": 0.38453659415245056, "learning_rate": 7.105079081665803e-05, "loss": 0.0437, "step": 39360 }, { "grad_norm": 0.32539352774620056, "learning_rate": 7.103579181340476e-05, "loss": 0.0455, "step": 39370 }, { "grad_norm": 0.42068156599998474, "learning_rate": 7.102079050970893e-05, "loss": 0.0361, "step": 39380 }, { "grad_norm": 0.36763718724250793, "learning_rate": 7.100578690721105e-05, "loss": 0.0437, "step": 39390 }, { "grad_norm": 0.7725616097450256, "learning_rate": 7.09907810075519e-05, "loss": 0.0553, "step": 39400 }, { "grad_norm": 0.4312637746334076, "learning_rate": 7.097577281237249e-05, "loss": 0.0515, "step": 39410 }, { "grad_norm": 0.4109556972980499, "learning_rate": 7.09607623233141e-05, "loss": 0.0363, "step": 39420 }, { "grad_norm": 0.6649289727210999, "learning_rate": 7.094574954201827e-05, "loss": 0.038, "step": 39430 }, { "grad_norm": 0.7712973952293396, "learning_rate": 7.093073447012675e-05, "loss": 0.0575, "step": 39440 }, { "grad_norm": 0.5267809629440308, "learning_rate": 7.09157171092816e-05, "loss": 0.0394, "step": 39450 }, { "grad_norm": 0.3424154818058014, "learning_rate": 7.090069746112504e-05, "loss": 0.0392, "step": 39460 }, { "grad_norm": 0.5418708324432373, "learning_rate": 7.088567552729965e-05, "loss": 0.0412, "step": 39470 }, { "grad_norm": 0.34370073676109314, "learning_rate": 7.087065130944818e-05, "loss": 0.0456, "step": 39480 }, { "grad_norm": 0.3857472240924835, "learning_rate": 7.085562480921366e-05, "loss": 0.0523, "step": 39490 }, { "grad_norm": 0.31063002347946167, "learning_rate": 7.084059602823937e-05, "loss": 0.0435, "step": 39500 }, { "grad_norm": 0.3645537793636322, "learning_rate": 7.082556496816882e-05, "loss": 0.0381, "step": 39510 }, { "grad_norm": 0.4169541895389557, "learning_rate": 7.081053163064582e-05, "loss": 0.0388, "step": 39520 }, { "grad_norm": 0.4187989830970764, "learning_rate": 7.079549601731434e-05, "loss": 0.039, "step": 39530 }, { "grad_norm": 0.4574403464794159, "learning_rate": 7.07804581298187e-05, "loss": 0.0385, "step": 39540 }, { "grad_norm": 0.42949673533439636, "learning_rate": 7.07654179698034e-05, "loss": 0.0546, "step": 39550 }, { "grad_norm": 0.4483352303504944, "learning_rate": 7.075037553891321e-05, "loss": 0.0466, "step": 39560 }, { "grad_norm": 0.33242669701576233, "learning_rate": 7.073533083879315e-05, "loss": 0.0417, "step": 39570 }, { "grad_norm": 0.545911967754364, "learning_rate": 7.072028387108849e-05, "loss": 0.0423, "step": 39580 }, { "grad_norm": 0.30035001039505005, "learning_rate": 7.070523463744472e-05, "loss": 0.0471, "step": 39590 }, { "grad_norm": 0.40676024556159973, "learning_rate": 7.069018313950763e-05, "loss": 0.039, "step": 39600 }, { "grad_norm": 0.5054722428321838, "learning_rate": 7.067512937892322e-05, "loss": 0.0403, "step": 39610 }, { "grad_norm": 0.4904322326183319, "learning_rate": 7.066007335733775e-05, "loss": 0.0361, "step": 39620 }, { "grad_norm": 0.40164807438850403, "learning_rate": 7.064501507639772e-05, "loss": 0.0459, "step": 39630 }, { "grad_norm": 0.6748655438423157, "learning_rate": 7.062995453774987e-05, "loss": 0.042, "step": 39640 }, { "grad_norm": 0.3423067629337311, "learning_rate": 7.061489174304121e-05, "loss": 0.0599, "step": 39650 }, { "grad_norm": 0.6822856068611145, "learning_rate": 7.0599826693919e-05, "loss": 0.0315, "step": 39660 }, { "grad_norm": 0.5176382064819336, "learning_rate": 7.05847593920307e-05, "loss": 0.0481, "step": 39670 }, { "grad_norm": 0.8028169870376587, "learning_rate": 7.056968983902406e-05, "loss": 0.0419, "step": 39680 }, { "grad_norm": 0.3469596207141876, "learning_rate": 7.055461803654706e-05, "loss": 0.0372, "step": 39690 }, { "grad_norm": 0.42667168378829956, "learning_rate": 7.053954398624794e-05, "loss": 0.0324, "step": 39700 }, { "grad_norm": 0.2861235737800598, "learning_rate": 7.052446768977518e-05, "loss": 0.0425, "step": 39710 }, { "grad_norm": 0.6167629957199097, "learning_rate": 7.050938914877748e-05, "loss": 0.0417, "step": 39720 }, { "grad_norm": 0.45036014914512634, "learning_rate": 7.049430836490382e-05, "loss": 0.0591, "step": 39730 }, { "grad_norm": 0.4056069552898407, "learning_rate": 7.047922533980342e-05, "loss": 0.0415, "step": 39740 }, { "grad_norm": 0.452187716960907, "learning_rate": 7.046414007512571e-05, "loss": 0.0498, "step": 39750 }, { "grad_norm": 0.4350798726081848, "learning_rate": 7.044905257252042e-05, "loss": 0.0391, "step": 39760 }, { "grad_norm": 0.7011847496032715, "learning_rate": 7.04339628336375e-05, "loss": 0.0588, "step": 39770 }, { "grad_norm": 0.42145633697509766, "learning_rate": 7.041887086012711e-05, "loss": 0.0493, "step": 39780 }, { "grad_norm": 0.46278342604637146, "learning_rate": 7.040377665363969e-05, "loss": 0.0431, "step": 39790 }, { "grad_norm": 0.6150349974632263, "learning_rate": 7.038868021582594e-05, "loss": 0.0491, "step": 39800 }, { "grad_norm": 0.44565778970718384, "learning_rate": 7.037358154833679e-05, "loss": 0.0395, "step": 39810 }, { "grad_norm": 0.718988299369812, "learning_rate": 7.035848065282339e-05, "loss": 0.0483, "step": 39820 }, { "grad_norm": 0.385349839925766, "learning_rate": 7.034337753093714e-05, "loss": 0.0344, "step": 39830 }, { "grad_norm": 0.35942795872688293, "learning_rate": 7.032827218432972e-05, "loss": 0.0411, "step": 39840 }, { "grad_norm": 0.44057604670524597, "learning_rate": 7.031316461465302e-05, "loss": 0.0404, "step": 39850 }, { "grad_norm": 0.47339263558387756, "learning_rate": 7.029805482355915e-05, "loss": 0.037, "step": 39860 }, { "grad_norm": 0.6533347964286804, "learning_rate": 7.028294281270055e-05, "loss": 0.0484, "step": 39870 }, { "grad_norm": 0.34709417819976807, "learning_rate": 7.026782858372982e-05, "loss": 0.0526, "step": 39880 }, { "grad_norm": 0.3980366289615631, "learning_rate": 7.025271213829982e-05, "loss": 0.0449, "step": 39890 }, { "grad_norm": 0.5093982815742493, "learning_rate": 7.023759347806366e-05, "loss": 0.0353, "step": 39900 }, { "grad_norm": 0.4192695617675781, "learning_rate": 7.022247260467472e-05, "loss": 0.0438, "step": 39910 }, { "grad_norm": 0.4160941541194916, "learning_rate": 7.020734951978658e-05, "loss": 0.0487, "step": 39920 }, { "grad_norm": 0.49201712012290955, "learning_rate": 7.019222422505307e-05, "loss": 0.0476, "step": 39930 }, { "grad_norm": 0.31215986609458923, "learning_rate": 7.017709672212827e-05, "loss": 0.0359, "step": 39940 }, { "grad_norm": 0.47386112809181213, "learning_rate": 7.016196701266652e-05, "loss": 0.0401, "step": 39950 }, { "grad_norm": 0.27394139766693115, "learning_rate": 7.014683509832238e-05, "loss": 0.0421, "step": 39960 }, { "grad_norm": 0.4196154773235321, "learning_rate": 7.013170098075063e-05, "loss": 0.0476, "step": 39970 }, { "grad_norm": 0.41459348797798157, "learning_rate": 7.011656466160632e-05, "loss": 0.0439, "step": 39980 }, { "grad_norm": 0.3974737823009491, "learning_rate": 7.010142614254475e-05, "loss": 0.0423, "step": 39990 }, { "grad_norm": 0.39820596575737, "learning_rate": 7.008628542522147e-05, "loss": 0.0452, "step": 40000 }, { "grad_norm": 0.860914945602417, "learning_rate": 7.007114251129218e-05, "loss": 0.0436, "step": 40010 }, { "grad_norm": 0.5349844694137573, "learning_rate": 7.005599740241293e-05, "loss": 0.0431, "step": 40020 }, { "grad_norm": 0.5058090686798096, "learning_rate": 7.004085010023996e-05, "loss": 0.0433, "step": 40030 }, { "grad_norm": 1.407070279121399, "learning_rate": 7.002570060642976e-05, "loss": 0.0507, "step": 40040 }, { "grad_norm": 0.36762040853500366, "learning_rate": 7.001054892263903e-05, "loss": 0.0465, "step": 40050 }, { "grad_norm": 0.46597152948379517, "learning_rate": 6.999539505052477e-05, "loss": 0.0507, "step": 40060 }, { "grad_norm": 0.44720956683158875, "learning_rate": 6.998023899174418e-05, "loss": 0.041, "step": 40070 }, { "grad_norm": 0.2891176640987396, "learning_rate": 6.996508074795467e-05, "loss": 0.0377, "step": 40080 }, { "grad_norm": 0.41263511776924133, "learning_rate": 6.994992032081396e-05, "loss": 0.0504, "step": 40090 }, { "grad_norm": 0.353322833776474, "learning_rate": 6.993475771197995e-05, "loss": 0.036, "step": 40100 }, { "grad_norm": 0.45540958642959595, "learning_rate": 6.991959292311082e-05, "loss": 0.0353, "step": 40110 }, { "grad_norm": 0.3568907082080841, "learning_rate": 6.990442595586495e-05, "loss": 0.0439, "step": 40120 }, { "grad_norm": 0.3768884539604187, "learning_rate": 6.988925681190098e-05, "loss": 0.0438, "step": 40130 }, { "grad_norm": 0.4914277195930481, "learning_rate": 6.987408549287778e-05, "loss": 0.0435, "step": 40140 }, { "grad_norm": 1.5342762470245361, "learning_rate": 6.985891200045449e-05, "loss": 0.0447, "step": 40150 }, { "grad_norm": 0.6735498905181885, "learning_rate": 6.984373633629045e-05, "loss": 0.0469, "step": 40160 }, { "grad_norm": 0.4752374589443207, "learning_rate": 6.98285585020452e-05, "loss": 0.036, "step": 40170 }, { "grad_norm": 0.49828267097473145, "learning_rate": 6.981337849937864e-05, "loss": 0.0415, "step": 40180 }, { "grad_norm": 0.3336135745048523, "learning_rate": 6.979819632995078e-05, "loss": 0.0426, "step": 40190 }, { "grad_norm": 0.2803622782230377, "learning_rate": 6.978301199542193e-05, "loss": 0.0325, "step": 40200 }, { "grad_norm": 0.49845919013023376, "learning_rate": 6.976782549745263e-05, "loss": 0.0396, "step": 40210 }, { "grad_norm": 0.38415980339050293, "learning_rate": 6.975263683770367e-05, "loss": 0.0485, "step": 40220 }, { "grad_norm": 0.4813092350959778, "learning_rate": 6.973744601783604e-05, "loss": 0.0493, "step": 40230 }, { "grad_norm": 0.4365828335285187, "learning_rate": 6.972225303951097e-05, "loss": 0.0418, "step": 40240 }, { "grad_norm": 0.39549532532691956, "learning_rate": 6.970705790438998e-05, "loss": 0.0415, "step": 40250 }, { "grad_norm": 0.5320591330528259, "learning_rate": 6.969186061413477e-05, "loss": 0.0457, "step": 40260 }, { "grad_norm": 0.466134637594223, "learning_rate": 6.967666117040727e-05, "loss": 0.0506, "step": 40270 }, { "grad_norm": 0.43801847100257874, "learning_rate": 6.966145957486972e-05, "loss": 0.0439, "step": 40280 }, { "grad_norm": 0.462569922208786, "learning_rate": 6.964625582918449e-05, "loss": 0.0475, "step": 40290 }, { "grad_norm": 1.0604830980300903, "learning_rate": 6.963104993501425e-05, "loss": 0.0493, "step": 40300 }, { "grad_norm": 0.6839027404785156, "learning_rate": 6.961584189402192e-05, "loss": 0.0491, "step": 40310 }, { "grad_norm": 0.48888522386550903, "learning_rate": 6.96006317078706e-05, "loss": 0.0353, "step": 40320 }, { "grad_norm": 0.6151255369186401, "learning_rate": 6.958541937822367e-05, "loss": 0.0447, "step": 40330 }, { "grad_norm": 0.44421204924583435, "learning_rate": 6.957020490674473e-05, "loss": 0.0433, "step": 40340 }, { "grad_norm": 0.6373677253723145, "learning_rate": 6.95549882950976e-05, "loss": 0.0492, "step": 40350 }, { "grad_norm": 0.9982037544250488, "learning_rate": 6.953976954494635e-05, "loss": 0.0402, "step": 40360 }, { "grad_norm": 0.43828535079956055, "learning_rate": 6.952454865795528e-05, "loss": 0.0448, "step": 40370 }, { "grad_norm": 0.7471442818641663, "learning_rate": 6.950932563578892e-05, "loss": 0.0387, "step": 40380 }, { "grad_norm": 0.3848007619380951, "learning_rate": 6.949410048011206e-05, "loss": 0.0393, "step": 40390 }, { "grad_norm": 0.5584544539451599, "learning_rate": 6.947887319258966e-05, "loss": 0.0516, "step": 40400 }, { "grad_norm": 0.4152417778968811, "learning_rate": 6.946364377488696e-05, "loss": 0.0402, "step": 40410 }, { "grad_norm": 0.672662615776062, "learning_rate": 6.944841222866947e-05, "loss": 0.0424, "step": 40420 }, { "grad_norm": 0.4636322557926178, "learning_rate": 6.943317855560284e-05, "loss": 0.0521, "step": 40430 }, { "grad_norm": 0.42262741923332214, "learning_rate": 6.941794275735302e-05, "loss": 0.0377, "step": 40440 }, { "grad_norm": 0.5179104804992676, "learning_rate": 6.94027048355862e-05, "loss": 0.0386, "step": 40450 }, { "grad_norm": 0.6976966261863708, "learning_rate": 6.938746479196871e-05, "loss": 0.0541, "step": 40460 }, { "grad_norm": 0.9889299869537354, "learning_rate": 6.937222262816724e-05, "loss": 0.0404, "step": 40470 }, { "grad_norm": 0.515966534614563, "learning_rate": 6.935697834584865e-05, "loss": 0.0428, "step": 40480 }, { "grad_norm": 0.34586209058761597, "learning_rate": 6.934173194667998e-05, "loss": 0.0398, "step": 40490 }, { "grad_norm": 0.509283185005188, "learning_rate": 6.93264834323286e-05, "loss": 0.0413, "step": 40500 }, { "grad_norm": 0.5226027369499207, "learning_rate": 6.931123280446206e-05, "loss": 0.0371, "step": 40510 }, { "grad_norm": 0.5489265322685242, "learning_rate": 6.929598006474811e-05, "loss": 0.04, "step": 40520 }, { "grad_norm": 0.42164143919944763, "learning_rate": 6.92807252148548e-05, "loss": 0.0489, "step": 40530 }, { "grad_norm": 0.44145989418029785, "learning_rate": 6.926546825645037e-05, "loss": 0.0359, "step": 40540 }, { "grad_norm": 0.28630074858665466, "learning_rate": 6.92502091912033e-05, "loss": 0.0397, "step": 40550 }, { "grad_norm": 0.42295897006988525, "learning_rate": 6.923494802078229e-05, "loss": 0.0379, "step": 40560 }, { "grad_norm": 0.4462425410747528, "learning_rate": 6.921968474685628e-05, "loss": 0.0368, "step": 40570 }, { "grad_norm": 0.4667622148990631, "learning_rate": 6.920441937109445e-05, "loss": 0.0499, "step": 40580 }, { "grad_norm": 0.4900438189506531, "learning_rate": 6.91891518951662e-05, "loss": 0.0532, "step": 40590 }, { "grad_norm": 0.4813772737979889, "learning_rate": 6.917388232074114e-05, "loss": 0.0402, "step": 40600 }, { "grad_norm": 0.3112170696258545, "learning_rate": 6.915861064948914e-05, "loss": 0.0391, "step": 40610 }, { "grad_norm": 0.662825882434845, "learning_rate": 6.914333688308025e-05, "loss": 0.0417, "step": 40620 }, { "grad_norm": 0.6400831937789917, "learning_rate": 6.912806102318487e-05, "loss": 0.0376, "step": 40630 }, { "grad_norm": 0.5212370753288269, "learning_rate": 6.911278307147347e-05, "loss": 0.0456, "step": 40640 }, { "grad_norm": 0.4695931077003479, "learning_rate": 6.909750302961684e-05, "loss": 0.0417, "step": 40650 }, { "grad_norm": 0.5920886993408203, "learning_rate": 6.9082220899286e-05, "loss": 0.054, "step": 40660 }, { "grad_norm": 0.39217495918273926, "learning_rate": 6.906693668215219e-05, "loss": 0.0421, "step": 40670 }, { "grad_norm": 0.5218005776405334, "learning_rate": 6.905165037988683e-05, "loss": 0.0392, "step": 40680 }, { "grad_norm": 0.34085163474082947, "learning_rate": 6.903636199416164e-05, "loss": 0.0387, "step": 40690 }, { "grad_norm": 0.3353287875652313, "learning_rate": 6.902107152664851e-05, "loss": 0.0396, "step": 40700 }, { "grad_norm": 0.41524264216423035, "learning_rate": 6.90057789790196e-05, "loss": 0.0315, "step": 40710 }, { "grad_norm": 0.38081493973731995, "learning_rate": 6.899048435294728e-05, "loss": 0.0377, "step": 40720 }, { "grad_norm": 0.3575165271759033, "learning_rate": 6.897518765010415e-05, "loss": 0.0364, "step": 40730 }, { "grad_norm": 0.2978489398956299, "learning_rate": 6.895988887216302e-05, "loss": 0.0415, "step": 40740 }, { "grad_norm": 0.43555527925491333, "learning_rate": 6.894458802079694e-05, "loss": 0.0454, "step": 40750 }, { "grad_norm": 0.29082024097442627, "learning_rate": 6.892928509767922e-05, "loss": 0.0361, "step": 40760 }, { "grad_norm": 0.45109736919403076, "learning_rate": 6.891398010448333e-05, "loss": 0.0367, "step": 40770 }, { "grad_norm": 0.40937894582748413, "learning_rate": 6.8898673042883e-05, "loss": 0.0394, "step": 40780 }, { "grad_norm": 0.43719494342803955, "learning_rate": 6.888336391455222e-05, "loss": 0.0427, "step": 40790 }, { "grad_norm": 0.3014541268348694, "learning_rate": 6.886805272116513e-05, "loss": 0.0414, "step": 40800 }, { "grad_norm": 0.45011013746261597, "learning_rate": 6.885273946439617e-05, "loss": 0.0518, "step": 40810 }, { "grad_norm": 0.5436296463012695, "learning_rate": 6.883742414591998e-05, "loss": 0.0474, "step": 40820 }, { "grad_norm": 0.30676963925361633, "learning_rate": 6.88221067674114e-05, "loss": 0.0311, "step": 40830 }, { "grad_norm": 0.3683055639266968, "learning_rate": 6.88067873305455e-05, "loss": 0.0437, "step": 40840 }, { "grad_norm": 0.2501045763492584, "learning_rate": 6.879146583699765e-05, "loss": 0.0347, "step": 40850 }, { "grad_norm": 0.31053948402404785, "learning_rate": 6.877614228844334e-05, "loss": 0.0422, "step": 40860 }, { "grad_norm": 0.43976959586143494, "learning_rate": 6.876081668655832e-05, "loss": 0.0541, "step": 40870 }, { "grad_norm": 0.7596452832221985, "learning_rate": 6.874548903301861e-05, "loss": 0.048, "step": 40880 }, { "grad_norm": 0.3256349563598633, "learning_rate": 6.873015932950039e-05, "loss": 0.0455, "step": 40890 }, { "grad_norm": 0.2972891628742218, "learning_rate": 6.871482757768012e-05, "loss": 0.0536, "step": 40900 }, { "grad_norm": 0.7053359746932983, "learning_rate": 6.869949377923442e-05, "loss": 0.0471, "step": 40910 }, { "grad_norm": 1.325518012046814, "learning_rate": 6.868415793584022e-05, "loss": 0.0355, "step": 40920 }, { "grad_norm": 0.576802134513855, "learning_rate": 6.86688200491746e-05, "loss": 0.0385, "step": 40930 }, { "grad_norm": 0.47765687108039856, "learning_rate": 6.865348012091488e-05, "loss": 0.042, "step": 40940 }, { "grad_norm": 0.35075297951698303, "learning_rate": 6.86381381527386e-05, "loss": 0.0394, "step": 40950 }, { "grad_norm": 0.35927891731262207, "learning_rate": 6.862279414632359e-05, "loss": 0.0348, "step": 40960 }, { "grad_norm": 0.5827056765556335, "learning_rate": 6.860744810334778e-05, "loss": 0.043, "step": 40970 }, { "grad_norm": 0.3475539982318878, "learning_rate": 6.859210002548943e-05, "loss": 0.0373, "step": 40980 }, { "grad_norm": 0.565067708492279, "learning_rate": 6.8576749914427e-05, "loss": 0.0381, "step": 40990 }, { "grad_norm": 0.4027485251426697, "learning_rate": 6.85613977718391e-05, "loss": 0.0493, "step": 41000 }, { "grad_norm": 0.4244912564754486, "learning_rate": 6.854604359940464e-05, "loss": 0.0484, "step": 41010 }, { "grad_norm": 0.5452826619148254, "learning_rate": 6.853068739880276e-05, "loss": 0.0395, "step": 41020 }, { "grad_norm": 0.6115821599960327, "learning_rate": 6.851532917171276e-05, "loss": 0.0535, "step": 41030 }, { "grad_norm": 0.6456614136695862, "learning_rate": 6.849996891981418e-05, "loss": 0.0572, "step": 41040 }, { "grad_norm": 0.4224415719509125, "learning_rate": 6.848460664478684e-05, "loss": 0.0477, "step": 41050 }, { "grad_norm": 0.35485780239105225, "learning_rate": 6.846924234831069e-05, "loss": 0.0351, "step": 41060 }, { "grad_norm": 1.1044929027557373, "learning_rate": 6.845387603206597e-05, "loss": 0.0416, "step": 41070 }, { "grad_norm": 0.458880215883255, "learning_rate": 6.843850769773312e-05, "loss": 0.033, "step": 41080 }, { "grad_norm": 0.49388259649276733, "learning_rate": 6.842313734699278e-05, "loss": 0.0471, "step": 41090 }, { "grad_norm": 0.4974888265132904, "learning_rate": 6.840776498152584e-05, "loss": 0.0411, "step": 41100 }, { "grad_norm": 0.8113541007041931, "learning_rate": 6.83923906030134e-05, "loss": 0.0394, "step": 41110 }, { "grad_norm": 0.3225492835044861, "learning_rate": 6.837701421313677e-05, "loss": 0.0487, "step": 41120 }, { "grad_norm": 0.4365193843841553, "learning_rate": 6.83616358135775e-05, "loss": 0.0375, "step": 41130 }, { "grad_norm": 0.6339321136474609, "learning_rate": 6.834625540601734e-05, "loss": 0.0516, "step": 41140 }, { "grad_norm": 0.4736141860485077, "learning_rate": 6.833087299213829e-05, "loss": 0.0433, "step": 41150 }, { "grad_norm": 0.40632179379463196, "learning_rate": 6.83154885736225e-05, "loss": 0.0502, "step": 41160 }, { "grad_norm": 0.6455939412117004, "learning_rate": 6.830010215215244e-05, "loss": 0.0449, "step": 41170 }, { "grad_norm": 0.40384504199028015, "learning_rate": 6.828471372941072e-05, "loss": 0.0404, "step": 41180 }, { "grad_norm": 1.268881916999817, "learning_rate": 6.826932330708018e-05, "loss": 0.0491, "step": 41190 }, { "grad_norm": 0.3622455894947052, "learning_rate": 6.825393088684393e-05, "loss": 0.0284, "step": 41200 }, { "grad_norm": 0.4936639964580536, "learning_rate": 6.823853647038524e-05, "loss": 0.0328, "step": 41210 }, { "grad_norm": 0.5144369602203369, "learning_rate": 6.822314005938762e-05, "loss": 0.0385, "step": 41220 }, { "grad_norm": 0.5188654065132141, "learning_rate": 6.820774165553479e-05, "loss": 0.0451, "step": 41230 }, { "grad_norm": 0.6667060256004333, "learning_rate": 6.819234126051072e-05, "loss": 0.0398, "step": 41240 }, { "grad_norm": 0.3958682119846344, "learning_rate": 6.817693887599956e-05, "loss": 0.0451, "step": 41250 }, { "grad_norm": 0.502518355846405, "learning_rate": 6.816153450368568e-05, "loss": 0.0505, "step": 41260 }, { "grad_norm": 0.5900925993919373, "learning_rate": 6.814612814525372e-05, "loss": 0.0452, "step": 41270 }, { "grad_norm": 0.535014808177948, "learning_rate": 6.813071980238846e-05, "loss": 0.0361, "step": 41280 }, { "grad_norm": 0.5161029100418091, "learning_rate": 6.811530947677492e-05, "loss": 0.042, "step": 41290 }, { "grad_norm": 0.41044703125953674, "learning_rate": 6.809989717009839e-05, "loss": 0.0388, "step": 41300 }, { "grad_norm": 0.7564983367919922, "learning_rate": 6.808448288404431e-05, "loss": 0.0469, "step": 41310 }, { "grad_norm": 0.32315269112586975, "learning_rate": 6.806906662029838e-05, "loss": 0.0387, "step": 41320 }, { "grad_norm": 0.4458788335323334, "learning_rate": 6.805364838054647e-05, "loss": 0.0462, "step": 41330 }, { "grad_norm": 0.384529709815979, "learning_rate": 6.803822816647474e-05, "loss": 0.0487, "step": 41340 }, { "grad_norm": 0.8004761934280396, "learning_rate": 6.802280597976949e-05, "loss": 0.0445, "step": 41350 }, { "grad_norm": 0.39244845509529114, "learning_rate": 6.800738182211727e-05, "loss": 0.0416, "step": 41360 }, { "grad_norm": 0.790102481842041, "learning_rate": 6.799195569520485e-05, "loss": 0.0517, "step": 41370 }, { "grad_norm": 0.4545299708843231, "learning_rate": 6.79765276007192e-05, "loss": 0.0505, "step": 41380 }, { "grad_norm": 0.683147132396698, "learning_rate": 6.796109754034753e-05, "loss": 0.0433, "step": 41390 }, { "grad_norm": 0.5159928798675537, "learning_rate": 6.794566551577724e-05, "loss": 0.0389, "step": 41400 }, { "grad_norm": 0.44829222559928894, "learning_rate": 6.793023152869594e-05, "loss": 0.0391, "step": 41410 }, { "grad_norm": 0.9359946846961975, "learning_rate": 6.791479558079147e-05, "loss": 0.0453, "step": 41420 }, { "grad_norm": 0.984844982624054, "learning_rate": 6.789935767375191e-05, "loss": 0.0442, "step": 41430 }, { "grad_norm": 0.6316990852355957, "learning_rate": 6.78839178092655e-05, "loss": 0.0453, "step": 41440 }, { "grad_norm": 0.5825209021568298, "learning_rate": 6.786847598902072e-05, "loss": 0.0402, "step": 41450 }, { "grad_norm": 0.5025922060012817, "learning_rate": 6.785303221470627e-05, "loss": 0.0462, "step": 41460 }, { "grad_norm": 0.43380796909332275, "learning_rate": 6.783758648801108e-05, "loss": 0.0396, "step": 41470 }, { "grad_norm": 0.5916401147842407, "learning_rate": 6.782213881062422e-05, "loss": 0.0504, "step": 41480 }, { "grad_norm": 0.5326083898544312, "learning_rate": 6.780668918423508e-05, "loss": 0.0483, "step": 41490 }, { "grad_norm": 0.6547818183898926, "learning_rate": 6.779123761053317e-05, "loss": 0.0483, "step": 41500 }, { "grad_norm": 0.4983710050582886, "learning_rate": 6.777578409120826e-05, "loss": 0.0406, "step": 41510 }, { "grad_norm": 0.41277188062667847, "learning_rate": 6.776032862795033e-05, "loss": 0.0558, "step": 41520 }, { "grad_norm": 0.3648909330368042, "learning_rate": 6.774487122244956e-05, "loss": 0.0406, "step": 41530 }, { "grad_norm": 0.5468120574951172, "learning_rate": 6.772941187639637e-05, "loss": 0.0358, "step": 41540 }, { "grad_norm": 0.4057021141052246, "learning_rate": 6.771395059148134e-05, "loss": 0.0417, "step": 41550 }, { "grad_norm": 0.3501966595649719, "learning_rate": 6.76984873693953e-05, "loss": 0.0403, "step": 41560 }, { "grad_norm": 0.5459970235824585, "learning_rate": 6.768302221182931e-05, "loss": 0.0428, "step": 41570 }, { "grad_norm": 0.42299702763557434, "learning_rate": 6.766755512047457e-05, "loss": 0.0529, "step": 41580 }, { "grad_norm": 0.34439048171043396, "learning_rate": 6.765208609702259e-05, "loss": 0.0417, "step": 41590 }, { "grad_norm": 0.31361013650894165, "learning_rate": 6.763661514316499e-05, "loss": 0.0371, "step": 41600 }, { "grad_norm": 0.4408722519874573, "learning_rate": 6.76211422605937e-05, "loss": 0.04, "step": 41610 }, { "grad_norm": 0.6375316977500916, "learning_rate": 6.760566745100076e-05, "loss": 0.043, "step": 41620 }, { "grad_norm": 0.7072446942329407, "learning_rate": 6.75901907160785e-05, "loss": 0.0552, "step": 41630 }, { "grad_norm": 0.36413535475730896, "learning_rate": 6.757471205751943e-05, "loss": 0.0347, "step": 41640 }, { "grad_norm": 0.34886109828948975, "learning_rate": 6.75592314770163e-05, "loss": 0.0332, "step": 41650 }, { "grad_norm": 0.3993329703807831, "learning_rate": 6.7543748976262e-05, "loss": 0.0402, "step": 41660 }, { "grad_norm": 0.4017980694770813, "learning_rate": 6.752826455694968e-05, "loss": 0.0316, "step": 41670 }, { "grad_norm": 0.43193161487579346, "learning_rate": 6.751277822077271e-05, "loss": 0.0357, "step": 41680 }, { "grad_norm": 0.46066394448280334, "learning_rate": 6.749728996942465e-05, "loss": 0.0343, "step": 41690 }, { "grad_norm": 0.45991337299346924, "learning_rate": 6.748179980459924e-05, "loss": 0.0374, "step": 41700 }, { "grad_norm": 0.2609732151031494, "learning_rate": 6.746630772799052e-05, "loss": 0.0389, "step": 41710 }, { "grad_norm": 0.38033047318458557, "learning_rate": 6.745081374129262e-05, "loss": 0.039, "step": 41720 }, { "grad_norm": 0.5152207016944885, "learning_rate": 6.743531784619998e-05, "loss": 0.0531, "step": 41730 }, { "grad_norm": 0.4974544942378998, "learning_rate": 6.741982004440719e-05, "loss": 0.0416, "step": 41740 }, { "grad_norm": 0.40183013677597046, "learning_rate": 6.740432033760907e-05, "loss": 0.0496, "step": 41750 }, { "grad_norm": 0.3799397349357605, "learning_rate": 6.738881872750066e-05, "loss": 0.0367, "step": 41760 }, { "grad_norm": 0.6487741470336914, "learning_rate": 6.737331521577715e-05, "loss": 0.0409, "step": 41770 }, { "grad_norm": 0.3514498174190521, "learning_rate": 6.735780980413403e-05, "loss": 0.0402, "step": 41780 }, { "grad_norm": 0.6476510763168335, "learning_rate": 6.734230249426692e-05, "loss": 0.0469, "step": 41790 }, { "grad_norm": 0.4024946689605713, "learning_rate": 6.732679328787168e-05, "loss": 0.0375, "step": 41800 }, { "grad_norm": 0.3807837665081024, "learning_rate": 6.731128218664438e-05, "loss": 0.0354, "step": 41810 }, { "grad_norm": 0.5784426331520081, "learning_rate": 6.729576919228129e-05, "loss": 0.0293, "step": 41820 }, { "grad_norm": 0.6122447848320007, "learning_rate": 6.728025430647888e-05, "loss": 0.0314, "step": 41830 }, { "grad_norm": 0.5356582999229431, "learning_rate": 6.726473753093383e-05, "loss": 0.041, "step": 41840 }, { "grad_norm": 0.3571174144744873, "learning_rate": 6.724921886734305e-05, "loss": 0.0338, "step": 41850 }, { "grad_norm": 0.4304252862930298, "learning_rate": 6.723369831740363e-05, "loss": 0.0525, "step": 41860 }, { "grad_norm": 0.5392406582832336, "learning_rate": 6.721817588281288e-05, "loss": 0.0408, "step": 41870 }, { "grad_norm": 0.5052840709686279, "learning_rate": 6.720265156526828e-05, "loss": 0.0382, "step": 41880 }, { "grad_norm": 0.46654269099235535, "learning_rate": 6.718712536646758e-05, "loss": 0.0406, "step": 41890 }, { "grad_norm": 0.5348768830299377, "learning_rate": 6.71715972881087e-05, "loss": 0.0485, "step": 41900 }, { "grad_norm": 0.441315233707428, "learning_rate": 6.715606733188974e-05, "loss": 0.0407, "step": 41910 }, { "grad_norm": 0.43881168961524963, "learning_rate": 6.714053549950909e-05, "loss": 0.0383, "step": 41920 }, { "grad_norm": 0.4642486572265625, "learning_rate": 6.71250017926652e-05, "loss": 0.039, "step": 41930 }, { "grad_norm": 0.3412144184112549, "learning_rate": 6.71094662130569e-05, "loss": 0.0397, "step": 41940 }, { "grad_norm": 1.7054252624511719, "learning_rate": 6.709392876238307e-05, "loss": 0.0471, "step": 41950 }, { "grad_norm": 0.32985690236091614, "learning_rate": 6.707838944234289e-05, "loss": 0.0409, "step": 41960 }, { "grad_norm": 0.47934532165527344, "learning_rate": 6.706284825463574e-05, "loss": 0.0415, "step": 41970 }, { "grad_norm": 0.6050023436546326, "learning_rate": 6.704730520096114e-05, "loss": 0.041, "step": 41980 }, { "grad_norm": 0.4632384479045868, "learning_rate": 6.703176028301888e-05, "loss": 0.0388, "step": 41990 }, { "grad_norm": 0.6230427026748657, "learning_rate": 6.701621350250892e-05, "loss": 0.0382, "step": 42000 }, { "grad_norm": 0.7746738791465759, "learning_rate": 6.700066486113144e-05, "loss": 0.038, "step": 42010 }, { "grad_norm": 0.378841370344162, "learning_rate": 6.69851143605868e-05, "loss": 0.0341, "step": 42020 }, { "grad_norm": 0.6065164804458618, "learning_rate": 6.696956200257561e-05, "loss": 0.0407, "step": 42030 }, { "grad_norm": 0.37619179487228394, "learning_rate": 6.695400778879863e-05, "loss": 0.0535, "step": 42040 }, { "grad_norm": 0.5786355137825012, "learning_rate": 6.693845172095683e-05, "loss": 0.0585, "step": 42050 }, { "grad_norm": 0.4233734607696533, "learning_rate": 6.692289380075142e-05, "loss": 0.0427, "step": 42060 }, { "grad_norm": 0.35606950521469116, "learning_rate": 6.690733402988379e-05, "loss": 0.0422, "step": 42070 }, { "grad_norm": 0.33929184079170227, "learning_rate": 6.689177241005553e-05, "loss": 0.0386, "step": 42080 }, { "grad_norm": 0.49605223536491394, "learning_rate": 6.687620894296844e-05, "loss": 0.0421, "step": 42090 }, { "grad_norm": 0.49111050367355347, "learning_rate": 6.686064363032451e-05, "loss": 0.045, "step": 42100 }, { "grad_norm": 0.303889662027359, "learning_rate": 6.684507647382596e-05, "loss": 0.0386, "step": 42110 }, { "grad_norm": 0.4355742037296295, "learning_rate": 6.682950747517513e-05, "loss": 0.0422, "step": 42120 }, { "grad_norm": 0.4537895917892456, "learning_rate": 6.68139366360747e-05, "loss": 0.0444, "step": 42130 }, { "grad_norm": 0.3735072910785675, "learning_rate": 6.679836395822744e-05, "loss": 0.0286, "step": 42140 }, { "grad_norm": 0.5287324786186218, "learning_rate": 6.678278944333633e-05, "loss": 0.0465, "step": 42150 }, { "grad_norm": 0.3595447242259979, "learning_rate": 6.676721309310462e-05, "loss": 0.0378, "step": 42160 }, { "grad_norm": 0.34665045142173767, "learning_rate": 6.67516349092357e-05, "loss": 0.048, "step": 42170 }, { "grad_norm": 0.33841630816459656, "learning_rate": 6.673605489343315e-05, "loss": 0.0359, "step": 42180 }, { "grad_norm": 0.43400612473487854, "learning_rate": 6.672047304740081e-05, "loss": 0.0398, "step": 42190 }, { "grad_norm": 0.36970749497413635, "learning_rate": 6.670488937284268e-05, "loss": 0.0394, "step": 42200 }, { "grad_norm": 0.3595288097858429, "learning_rate": 6.668930387146294e-05, "loss": 0.0423, "step": 42210 }, { "grad_norm": 0.527576208114624, "learning_rate": 6.667371654496605e-05, "loss": 0.0442, "step": 42220 }, { "grad_norm": 1.0970675945281982, "learning_rate": 6.665812739505659e-05, "loss": 0.046, "step": 42230 }, { "grad_norm": 0.4998874366283417, "learning_rate": 6.664253642343935e-05, "loss": 0.0471, "step": 42240 }, { "grad_norm": 0.3634301424026489, "learning_rate": 6.662694363181935e-05, "loss": 0.0338, "step": 42250 }, { "grad_norm": 0.29481029510498047, "learning_rate": 6.66113490219018e-05, "loss": 0.0471, "step": 42260 }, { "grad_norm": 0.35906895995140076, "learning_rate": 6.659575259539207e-05, "loss": 0.0385, "step": 42270 }, { "grad_norm": 0.3063846528530121, "learning_rate": 6.65801543539958e-05, "loss": 0.0375, "step": 42280 }, { "grad_norm": 0.5819162726402283, "learning_rate": 6.65645542994188e-05, "loss": 0.0385, "step": 42290 }, { "grad_norm": 0.4110771417617798, "learning_rate": 6.654895243336702e-05, "loss": 0.0282, "step": 42300 }, { "grad_norm": 0.7346780300140381, "learning_rate": 6.65333487575467e-05, "loss": 0.0393, "step": 42310 }, { "grad_norm": 0.8185300230979919, "learning_rate": 6.651774327366422e-05, "loss": 0.0447, "step": 42320 }, { "grad_norm": 0.5472936034202576, "learning_rate": 6.650213598342613e-05, "loss": 0.0348, "step": 42330 }, { "grad_norm": 0.4816780686378479, "learning_rate": 6.648652688853931e-05, "loss": 0.0509, "step": 42340 }, { "grad_norm": 0.385118305683136, "learning_rate": 6.647091599071066e-05, "loss": 0.0462, "step": 42350 }, { "grad_norm": 0.5016862750053406, "learning_rate": 6.645530329164743e-05, "loss": 0.0379, "step": 42360 }, { "grad_norm": 0.5584526658058167, "learning_rate": 6.643968879305694e-05, "loss": 0.0514, "step": 42370 }, { "grad_norm": 0.4645717442035675, "learning_rate": 6.642407249664683e-05, "loss": 0.0414, "step": 42380 }, { "grad_norm": 0.5306795835494995, "learning_rate": 6.640845440412483e-05, "loss": 0.052, "step": 42390 }, { "grad_norm": 0.4890879690647125, "learning_rate": 6.639283451719893e-05, "loss": 0.0413, "step": 42400 }, { "grad_norm": 0.30219846963882446, "learning_rate": 6.637721283757729e-05, "loss": 0.0361, "step": 42410 }, { "grad_norm": 0.5384700298309326, "learning_rate": 6.636158936696828e-05, "loss": 0.0398, "step": 42420 }, { "grad_norm": 0.48287826776504517, "learning_rate": 6.634596410708047e-05, "loss": 0.0427, "step": 42430 }, { "grad_norm": 0.6779300570487976, "learning_rate": 6.633033705962257e-05, "loss": 0.0389, "step": 42440 }, { "grad_norm": 0.41114306449890137, "learning_rate": 6.631470822630359e-05, "loss": 0.0407, "step": 42450 }, { "grad_norm": 0.6607508063316345, "learning_rate": 6.629907760883263e-05, "loss": 0.0425, "step": 42460 }, { "grad_norm": 0.7310492396354675, "learning_rate": 6.628344520891905e-05, "loss": 0.0345, "step": 42470 }, { "grad_norm": 0.5004390478134155, "learning_rate": 6.62678110282724e-05, "loss": 0.0346, "step": 42480 }, { "grad_norm": 0.4367229640483856, "learning_rate": 6.625217506860237e-05, "loss": 0.0401, "step": 42490 }, { "grad_norm": 0.5327873229980469, "learning_rate": 6.623653733161892e-05, "loss": 0.039, "step": 42500 }, { "grad_norm": 0.7102595567703247, "learning_rate": 6.622089781903217e-05, "loss": 0.0417, "step": 42510 }, { "grad_norm": 2.4763832092285156, "learning_rate": 6.620525653255241e-05, "loss": 0.0471, "step": 42520 }, { "grad_norm": 0.43575534224510193, "learning_rate": 6.618961347389015e-05, "loss": 0.0363, "step": 42530 }, { "grad_norm": 0.8494817614555359, "learning_rate": 6.617396864475613e-05, "loss": 0.046, "step": 42540 }, { "grad_norm": 0.4175967574119568, "learning_rate": 6.61583220468612e-05, "loss": 0.0428, "step": 42550 }, { "grad_norm": 0.35888129472732544, "learning_rate": 6.614267368191645e-05, "loss": 0.0509, "step": 42560 }, { "grad_norm": 0.40657156705856323, "learning_rate": 6.61270235516332e-05, "loss": 0.0394, "step": 42570 }, { "grad_norm": 0.37133288383483887, "learning_rate": 6.611137165772287e-05, "loss": 0.0388, "step": 42580 }, { "grad_norm": 0.6147099733352661, "learning_rate": 6.60957180018972e-05, "loss": 0.0442, "step": 42590 }, { "grad_norm": 0.7972551584243774, "learning_rate": 6.608006258586797e-05, "loss": 0.0376, "step": 42600 }, { "grad_norm": 0.30629318952560425, "learning_rate": 6.606440541134731e-05, "loss": 0.0401, "step": 42610 }, { "grad_norm": 0.5778802633285522, "learning_rate": 6.60487464800474e-05, "loss": 0.0474, "step": 42620 }, { "grad_norm": 0.6284780502319336, "learning_rate": 6.603308579368071e-05, "loss": 0.0367, "step": 42630 }, { "grad_norm": 0.34454742074012756, "learning_rate": 6.601742335395987e-05, "loss": 0.0353, "step": 42640 }, { "grad_norm": 0.38487082719802856, "learning_rate": 6.600175916259769e-05, "loss": 0.0398, "step": 42650 }, { "grad_norm": 0.4417954087257385, "learning_rate": 6.598609322130718e-05, "loss": 0.0355, "step": 42660 }, { "grad_norm": 0.4455528259277344, "learning_rate": 6.597042553180155e-05, "loss": 0.0349, "step": 42670 }, { "grad_norm": 0.3977312445640564, "learning_rate": 6.59547560957942e-05, "loss": 0.0432, "step": 42680 }, { "grad_norm": 0.41988879442214966, "learning_rate": 6.59390849149987e-05, "loss": 0.0461, "step": 42690 }, { "grad_norm": 0.3663902282714844, "learning_rate": 6.592341199112886e-05, "loss": 0.0271, "step": 42700 }, { "grad_norm": 0.3687162697315216, "learning_rate": 6.590773732589861e-05, "loss": 0.0426, "step": 42710 }, { "grad_norm": 0.4804909825325012, "learning_rate": 6.589206092102213e-05, "loss": 0.0527, "step": 42720 }, { "grad_norm": 0.28969937562942505, "learning_rate": 6.587638277821375e-05, "loss": 0.0503, "step": 42730 }, { "grad_norm": 0.41869378089904785, "learning_rate": 6.586070289918803e-05, "loss": 0.0379, "step": 42740 }, { "grad_norm": 0.9996352195739746, "learning_rate": 6.584502128565968e-05, "loss": 0.0522, "step": 42750 }, { "grad_norm": 0.5527384877204895, "learning_rate": 6.582933793934363e-05, "loss": 0.0475, "step": 42760 }, { "grad_norm": 0.28314709663391113, "learning_rate": 6.581365286195498e-05, "loss": 0.0455, "step": 42770 }, { "grad_norm": 0.4140782356262207, "learning_rate": 6.579796605520903e-05, "loss": 0.0384, "step": 42780 }, { "grad_norm": 0.42323657870292664, "learning_rate": 6.578227752082127e-05, "loss": 0.0467, "step": 42790 }, { "grad_norm": 0.45310670137405396, "learning_rate": 6.576658726050735e-05, "loss": 0.0342, "step": 42800 }, { "grad_norm": 0.324983686208725, "learning_rate": 6.575089527598316e-05, "loss": 0.0389, "step": 42810 }, { "grad_norm": 0.4679068326950073, "learning_rate": 6.573520156896476e-05, "loss": 0.0428, "step": 42820 }, { "grad_norm": 0.5258707404136658, "learning_rate": 6.571950614116835e-05, "loss": 0.0436, "step": 42830 }, { "grad_norm": 1.3874647617340088, "learning_rate": 6.570380899431039e-05, "loss": 0.0545, "step": 42840 }, { "grad_norm": 0.4110795259475708, "learning_rate": 6.568811013010749e-05, "loss": 0.0424, "step": 42850 }, { "grad_norm": 0.8164290189743042, "learning_rate": 6.567240955027645e-05, "loss": 0.0407, "step": 42860 }, { "grad_norm": 1.3243646621704102, "learning_rate": 6.565670725653427e-05, "loss": 0.0419, "step": 42870 }, { "grad_norm": 0.5745415091514587, "learning_rate": 6.564100325059812e-05, "loss": 0.0353, "step": 42880 }, { "grad_norm": 0.5971854329109192, "learning_rate": 6.562529753418538e-05, "loss": 0.0447, "step": 42890 }, { "grad_norm": 0.3401833772659302, "learning_rate": 6.56095901090136e-05, "loss": 0.0417, "step": 42900 }, { "grad_norm": 0.7925502061843872, "learning_rate": 6.55938809768005e-05, "loss": 0.0375, "step": 42910 }, { "grad_norm": 0.4272993206977844, "learning_rate": 6.557817013926402e-05, "loss": 0.042, "step": 42920 }, { "grad_norm": 0.2844032049179077, "learning_rate": 6.556245759812227e-05, "loss": 0.0508, "step": 42930 }, { "grad_norm": 0.8751742243766785, "learning_rate": 6.554674335509357e-05, "loss": 0.0443, "step": 42940 }, { "grad_norm": 0.6362159252166748, "learning_rate": 6.553102741189638e-05, "loss": 0.035, "step": 42950 }, { "grad_norm": 0.34714698791503906, "learning_rate": 6.55153097702494e-05, "loss": 0.0366, "step": 42960 }, { "grad_norm": 0.393486887216568, "learning_rate": 6.549959043187146e-05, "loss": 0.0403, "step": 42970 }, { "grad_norm": 0.41269877552986145, "learning_rate": 6.54838693984816e-05, "loss": 0.036, "step": 42980 }, { "grad_norm": 0.4714987874031067, "learning_rate": 6.546814667179909e-05, "loss": 0.0358, "step": 42990 }, { "grad_norm": 0.6198506355285645, "learning_rate": 6.545242225354328e-05, "loss": 0.0454, "step": 43000 }, { "grad_norm": 0.44068285822868347, "learning_rate": 6.543669614543383e-05, "loss": 0.0413, "step": 43010 }, { "grad_norm": 0.41690710186958313, "learning_rate": 6.542096834919049e-05, "loss": 0.0405, "step": 43020 }, { "grad_norm": 0.40023815631866455, "learning_rate": 6.540523886653324e-05, "loss": 0.0411, "step": 43030 }, { "grad_norm": 0.4277292788028717, "learning_rate": 6.538950769918222e-05, "loss": 0.0379, "step": 43040 }, { "grad_norm": 0.4279629588127136, "learning_rate": 6.537377484885779e-05, "loss": 0.0391, "step": 43050 }, { "grad_norm": 0.39791205525398254, "learning_rate": 6.535804031728045e-05, "loss": 0.0469, "step": 43060 }, { "grad_norm": 0.46173614263534546, "learning_rate": 6.534230410617092e-05, "loss": 0.0399, "step": 43070 }, { "grad_norm": 0.3849620819091797, "learning_rate": 6.532656621725007e-05, "loss": 0.0371, "step": 43080 }, { "grad_norm": 0.3772151470184326, "learning_rate": 6.531082665223898e-05, "loss": 0.0387, "step": 43090 }, { "grad_norm": 0.5024232268333435, "learning_rate": 6.529508541285889e-05, "loss": 0.045, "step": 43100 }, { "grad_norm": 0.5209150314331055, "learning_rate": 6.527934250083128e-05, "loss": 0.0409, "step": 43110 }, { "grad_norm": 0.38338834047317505, "learning_rate": 6.526359791787775e-05, "loss": 0.0376, "step": 43120 }, { "grad_norm": 0.38099905848503113, "learning_rate": 6.524785166572009e-05, "loss": 0.0337, "step": 43130 }, { "grad_norm": 0.5142888426780701, "learning_rate": 6.523210374608029e-05, "loss": 0.0421, "step": 43140 }, { "grad_norm": 0.28917422890663147, "learning_rate": 6.521635416068054e-05, "loss": 0.032, "step": 43150 }, { "grad_norm": 0.35876837372779846, "learning_rate": 6.520060291124317e-05, "loss": 0.0505, "step": 43160 }, { "grad_norm": 0.7960715889930725, "learning_rate": 6.518484999949072e-05, "loss": 0.0466, "step": 43170 }, { "grad_norm": 0.6347602009773254, "learning_rate": 6.516909542714591e-05, "loss": 0.0453, "step": 43180 }, { "grad_norm": 0.4104277789592743, "learning_rate": 6.515333919593161e-05, "loss": 0.0423, "step": 43190 }, { "grad_norm": 0.6007084250450134, "learning_rate": 6.513758130757094e-05, "loss": 0.0417, "step": 43200 }, { "grad_norm": 0.3659750521183014, "learning_rate": 6.512182176378713e-05, "loss": 0.0389, "step": 43210 }, { "grad_norm": 0.5828446745872498, "learning_rate": 6.510606056630362e-05, "loss": 0.0381, "step": 43220 }, { "grad_norm": 0.34917986392974854, "learning_rate": 6.509029771684405e-05, "loss": 0.0433, "step": 43230 }, { "grad_norm": 0.44776639342308044, "learning_rate": 6.50745332171322e-05, "loss": 0.043, "step": 43240 }, { "grad_norm": 0.45552393794059753, "learning_rate": 6.505876706889207e-05, "loss": 0.0384, "step": 43250 }, { "grad_norm": 0.3739386796951294, "learning_rate": 6.504299927384781e-05, "loss": 0.0451, "step": 43260 }, { "grad_norm": 0.4520984888076782, "learning_rate": 6.502722983372378e-05, "loss": 0.0352, "step": 43270 }, { "grad_norm": 0.3776567280292511, "learning_rate": 6.501145875024446e-05, "loss": 0.0374, "step": 43280 }, { "grad_norm": 0.40674445033073425, "learning_rate": 6.499568602513459e-05, "loss": 0.0438, "step": 43290 }, { "grad_norm": 0.6794505715370178, "learning_rate": 6.497991166011903e-05, "loss": 0.0382, "step": 43300 }, { "grad_norm": 0.3791283369064331, "learning_rate": 6.49641356569229e-05, "loss": 0.039, "step": 43310 }, { "grad_norm": 0.49977415800094604, "learning_rate": 6.494835801727135e-05, "loss": 0.05, "step": 43320 }, { "grad_norm": 0.4643181264400482, "learning_rate": 6.493257874288986e-05, "loss": 0.0382, "step": 43330 }, { "grad_norm": 0.39083001017570496, "learning_rate": 6.491679783550399e-05, "loss": 0.044, "step": 43340 }, { "grad_norm": 0.7139122486114502, "learning_rate": 6.490101529683957e-05, "loss": 0.046, "step": 43350 }, { "grad_norm": 0.4795355796813965, "learning_rate": 6.488523112862248e-05, "loss": 0.0461, "step": 43360 }, { "grad_norm": 0.5815975069999695, "learning_rate": 6.486944533257891e-05, "loss": 0.0438, "step": 43370 }, { "grad_norm": 0.34867027401924133, "learning_rate": 6.485365791043517e-05, "loss": 0.0416, "step": 43380 }, { "grad_norm": 0.8535497188568115, "learning_rate": 6.483786886391772e-05, "loss": 0.041, "step": 43390 }, { "grad_norm": 0.47382277250289917, "learning_rate": 6.482207819475323e-05, "loss": 0.0377, "step": 43400 }, { "grad_norm": 0.5177022218704224, "learning_rate": 6.480628590466857e-05, "loss": 0.0427, "step": 43410 }, { "grad_norm": 0.5097840428352356, "learning_rate": 6.479049199539074e-05, "loss": 0.0355, "step": 43420 }, { "grad_norm": 0.417024701833725, "learning_rate": 6.477469646864693e-05, "loss": 0.043, "step": 43430 }, { "grad_norm": 0.501672089099884, "learning_rate": 6.475889932616454e-05, "loss": 0.0403, "step": 43440 }, { "grad_norm": 0.43323567509651184, "learning_rate": 6.474310056967111e-05, "loss": 0.0458, "step": 43450 }, { "grad_norm": 0.6718138456344604, "learning_rate": 6.472730020089437e-05, "loss": 0.0409, "step": 43460 }, { "grad_norm": 0.2826460301876068, "learning_rate": 6.471149822156222e-05, "loss": 0.0335, "step": 43470 }, { "grad_norm": 0.32946252822875977, "learning_rate": 6.469569463340275e-05, "loss": 0.0302, "step": 43480 }, { "grad_norm": 0.416144996881485, "learning_rate": 6.467988943814418e-05, "loss": 0.0334, "step": 43490 }, { "grad_norm": 0.5352477431297302, "learning_rate": 6.4664082637515e-05, "loss": 0.0435, "step": 43500 }, { "grad_norm": 0.4358389973640442, "learning_rate": 6.464827423324377e-05, "loss": 0.0422, "step": 43510 }, { "grad_norm": 0.5291310548782349, "learning_rate": 6.463246422705931e-05, "loss": 0.0392, "step": 43520 }, { "grad_norm": 0.406892329454422, "learning_rate": 6.461665262069055e-05, "loss": 0.044, "step": 43530 }, { "grad_norm": 1.4334359169006348, "learning_rate": 6.460083941586665e-05, "loss": 0.0388, "step": 43540 }, { "grad_norm": 0.6026831269264221, "learning_rate": 6.45850246143169e-05, "loss": 0.0397, "step": 43550 }, { "grad_norm": 0.4374249577522278, "learning_rate": 6.456920821777077e-05, "loss": 0.0445, "step": 43560 }, { "grad_norm": 0.49765917658805847, "learning_rate": 6.455339022795796e-05, "loss": 0.0468, "step": 43570 }, { "grad_norm": 0.3834744095802307, "learning_rate": 6.453757064660825e-05, "loss": 0.0417, "step": 43580 }, { "grad_norm": 0.3792761266231537, "learning_rate": 6.452174947545169e-05, "loss": 0.0292, "step": 43590 }, { "grad_norm": 0.4494844079017639, "learning_rate": 6.450592671621842e-05, "loss": 0.0422, "step": 43600 }, { "grad_norm": 0.4077546000480652, "learning_rate": 6.449010237063882e-05, "loss": 0.0433, "step": 43610 }, { "grad_norm": 0.465243399143219, "learning_rate": 6.447427644044343e-05, "loss": 0.0343, "step": 43620 }, { "grad_norm": 0.3931018114089966, "learning_rate": 6.44584489273629e-05, "loss": 0.0363, "step": 43630 }, { "grad_norm": 0.50096195936203, "learning_rate": 6.444261983312817e-05, "loss": 0.0332, "step": 43640 }, { "grad_norm": 0.3207021653652191, "learning_rate": 6.442678915947023e-05, "loss": 0.0377, "step": 43650 }, { "grad_norm": 0.2731122672557831, "learning_rate": 6.441095690812032e-05, "loss": 0.0348, "step": 43660 }, { "grad_norm": 0.2790701687335968, "learning_rate": 6.439512308080982e-05, "loss": 0.0359, "step": 43670 }, { "grad_norm": 0.5237201452255249, "learning_rate": 6.43792876792703e-05, "loss": 0.0344, "step": 43680 }, { "grad_norm": 0.4661445617675781, "learning_rate": 6.43634507052335e-05, "loss": 0.0471, "step": 43690 }, { "grad_norm": 0.5010504126548767, "learning_rate": 6.434761216043135e-05, "loss": 0.0442, "step": 43700 }, { "grad_norm": 0.4674485921859741, "learning_rate": 6.433177204659589e-05, "loss": 0.0394, "step": 43710 }, { "grad_norm": 0.372340589761734, "learning_rate": 6.431593036545938e-05, "loss": 0.0349, "step": 43720 }, { "grad_norm": 0.3424474000930786, "learning_rate": 6.430008711875426e-05, "loss": 0.0376, "step": 43730 }, { "grad_norm": 0.25233954191207886, "learning_rate": 6.42842423082131e-05, "loss": 0.0379, "step": 43740 }, { "grad_norm": 0.5164945125579834, "learning_rate": 6.426839593556868e-05, "loss": 0.0416, "step": 43750 }, { "grad_norm": 0.4023858308792114, "learning_rate": 6.425254800255392e-05, "loss": 0.0371, "step": 43760 }, { "grad_norm": 0.4887026250362396, "learning_rate": 6.423669851090196e-05, "loss": 0.0428, "step": 43770 }, { "grad_norm": 0.3796927332878113, "learning_rate": 6.422084746234604e-05, "loss": 0.042, "step": 43780 }, { "grad_norm": 0.9365013837814331, "learning_rate": 6.420499485861965e-05, "loss": 0.0331, "step": 43790 }, { "grad_norm": 0.5498682260513306, "learning_rate": 6.418914070145636e-05, "loss": 0.042, "step": 43800 }, { "grad_norm": 0.6071022152900696, "learning_rate": 6.417328499258996e-05, "loss": 0.0448, "step": 43810 }, { "grad_norm": 0.3220950961112976, "learning_rate": 6.415742773375446e-05, "loss": 0.0347, "step": 43820 }, { "grad_norm": 0.24343739449977875, "learning_rate": 6.414156892668393e-05, "loss": 0.036, "step": 43830 }, { "grad_norm": 0.4172031879425049, "learning_rate": 6.412570857311267e-05, "loss": 0.0403, "step": 43840 }, { "grad_norm": 0.29329001903533936, "learning_rate": 6.410984667477518e-05, "loss": 0.0427, "step": 43850 }, { "grad_norm": 0.35982632637023926, "learning_rate": 6.409398323340607e-05, "loss": 0.0353, "step": 43860 }, { "grad_norm": 0.35681986808776855, "learning_rate": 6.407811825074012e-05, "loss": 0.0372, "step": 43870 }, { "grad_norm": 0.5447524189949036, "learning_rate": 6.406225172851234e-05, "loss": 0.0461, "step": 43880 }, { "grad_norm": 0.41554322838783264, "learning_rate": 6.404638366845786e-05, "loss": 0.0445, "step": 43890 }, { "grad_norm": 0.4323384165763855, "learning_rate": 6.403051407231196e-05, "loss": 0.0488, "step": 43900 }, { "grad_norm": 0.6701781153678894, "learning_rate": 6.401464294181016e-05, "loss": 0.0432, "step": 43910 }, { "grad_norm": 0.39417895674705505, "learning_rate": 6.399877027868808e-05, "loss": 0.045, "step": 43920 }, { "grad_norm": 0.45756959915161133, "learning_rate": 6.39828960846815e-05, "loss": 0.0415, "step": 43930 }, { "grad_norm": 0.327845960855484, "learning_rate": 6.396702036152644e-05, "loss": 0.0397, "step": 43940 }, { "grad_norm": 0.640162467956543, "learning_rate": 6.395114311095905e-05, "loss": 0.0512, "step": 43950 }, { "grad_norm": 0.7202092409133911, "learning_rate": 6.393526433471559e-05, "loss": 0.0491, "step": 43960 }, { "grad_norm": 0.385871022939682, "learning_rate": 6.391938403453262e-05, "loss": 0.0482, "step": 43970 }, { "grad_norm": 0.4830302298069, "learning_rate": 6.390350221214671e-05, "loss": 0.0445, "step": 43980 }, { "grad_norm": 0.4854254722595215, "learning_rate": 6.38876188692947e-05, "loss": 0.0474, "step": 43990 }, { "grad_norm": 0.410247802734375, "learning_rate": 6.387173400771358e-05, "loss": 0.0426, "step": 44000 }, { "grad_norm": 0.5665075182914734, "learning_rate": 6.385584762914047e-05, "loss": 0.0387, "step": 44010 }, { "grad_norm": 0.5053203701972961, "learning_rate": 6.38399597353127e-05, "loss": 0.0348, "step": 44020 }, { "grad_norm": 0.7546053528785706, "learning_rate": 6.382407032796775e-05, "loss": 0.0394, "step": 44030 }, { "grad_norm": 0.7949835062026978, "learning_rate": 6.380817940884325e-05, "loss": 0.0415, "step": 44040 }, { "grad_norm": 0.3709847629070282, "learning_rate": 6.379228697967702e-05, "loss": 0.0391, "step": 44050 }, { "grad_norm": 0.4338897168636322, "learning_rate": 6.3776393042207e-05, "loss": 0.0364, "step": 44060 }, { "grad_norm": 0.38192257285118103, "learning_rate": 6.376049759817137e-05, "loss": 0.0315, "step": 44070 }, { "grad_norm": 0.6526222229003906, "learning_rate": 6.37446006493084e-05, "loss": 0.0489, "step": 44080 }, { "grad_norm": 0.6077436208724976, "learning_rate": 6.372870219735656e-05, "loss": 0.0466, "step": 44090 }, { "grad_norm": 0.42479729652404785, "learning_rate": 6.371280224405452e-05, "loss": 0.0471, "step": 44100 }, { "grad_norm": 0.39332789182662964, "learning_rate": 6.369690079114101e-05, "loss": 0.0356, "step": 44110 }, { "grad_norm": 0.5429085493087769, "learning_rate": 6.368099784035504e-05, "loss": 0.0363, "step": 44120 }, { "grad_norm": 0.5055682063102722, "learning_rate": 6.366509339343572e-05, "loss": 0.0308, "step": 44130 }, { "grad_norm": 0.5098738670349121, "learning_rate": 6.364918745212232e-05, "loss": 0.0371, "step": 44140 }, { "grad_norm": 0.4261857271194458, "learning_rate": 6.363328001815431e-05, "loss": 0.0363, "step": 44150 }, { "grad_norm": 0.5989701747894287, "learning_rate": 6.361737109327128e-05, "loss": 0.0387, "step": 44160 }, { "grad_norm": 0.4358741343021393, "learning_rate": 6.360146067921301e-05, "loss": 0.0411, "step": 44170 }, { "grad_norm": 0.45864200592041016, "learning_rate": 6.35855487777195e-05, "loss": 0.0431, "step": 44180 }, { "grad_norm": 0.573314368724823, "learning_rate": 6.356963539053076e-05, "loss": 0.0523, "step": 44190 }, { "grad_norm": 0.5159916281700134, "learning_rate": 6.355372051938711e-05, "loss": 0.0476, "step": 44200 }, { "grad_norm": 0.4562132954597473, "learning_rate": 6.353780416602894e-05, "loss": 0.038, "step": 44210 }, { "grad_norm": 0.626912534236908, "learning_rate": 6.352188633219689e-05, "loss": 0.0362, "step": 44220 }, { "grad_norm": 0.7598516941070557, "learning_rate": 6.350596701963166e-05, "loss": 0.034, "step": 44230 }, { "grad_norm": 0.6231699585914612, "learning_rate": 6.349004623007419e-05, "loss": 0.0445, "step": 44240 }, { "grad_norm": 0.4344988465309143, "learning_rate": 6.347412396526555e-05, "loss": 0.0369, "step": 44250 }, { "grad_norm": 0.47114434838294983, "learning_rate": 6.345820022694696e-05, "loss": 0.0488, "step": 44260 }, { "grad_norm": 0.7240598797798157, "learning_rate": 6.344227501685984e-05, "loss": 0.0387, "step": 44270 }, { "grad_norm": 0.4940245747566223, "learning_rate": 6.342634833674572e-05, "loss": 0.0344, "step": 44280 }, { "grad_norm": 0.48439157009124756, "learning_rate": 6.341042018834635e-05, "loss": 0.0497, "step": 44290 }, { "grad_norm": 0.4220053553581238, "learning_rate": 6.339449057340359e-05, "loss": 0.0343, "step": 44300 }, { "grad_norm": 0.382525771856308, "learning_rate": 6.337855949365945e-05, "loss": 0.0366, "step": 44310 }, { "grad_norm": 0.34790414571762085, "learning_rate": 6.336262695085619e-05, "loss": 0.0353, "step": 44320 }, { "grad_norm": 0.49495670199394226, "learning_rate": 6.334669294673612e-05, "loss": 0.0385, "step": 44330 }, { "grad_norm": 0.411767840385437, "learning_rate": 6.333075748304179e-05, "loss": 0.0476, "step": 44340 }, { "grad_norm": 0.36754104495048523, "learning_rate": 6.331482056151585e-05, "loss": 0.046, "step": 44350 }, { "grad_norm": 0.3080795705318451, "learning_rate": 6.329888218390117e-05, "loss": 0.0353, "step": 44360 }, { "grad_norm": 0.5261817574501038, "learning_rate": 6.328294235194072e-05, "loss": 0.039, "step": 44370 }, { "grad_norm": 0.4356713891029358, "learning_rate": 6.326700106737765e-05, "loss": 0.0451, "step": 44380 }, { "grad_norm": 0.4252820611000061, "learning_rate": 6.325105833195531e-05, "loss": 0.0374, "step": 44390 }, { "grad_norm": 0.4586776793003082, "learning_rate": 6.323511414741715e-05, "loss": 0.036, "step": 44400 }, { "grad_norm": 0.38265490531921387, "learning_rate": 6.321916851550678e-05, "loss": 0.0525, "step": 44410 }, { "grad_norm": 0.36266303062438965, "learning_rate": 6.320322143796806e-05, "loss": 0.0498, "step": 44420 }, { "grad_norm": 0.41146254539489746, "learning_rate": 6.318727291654488e-05, "loss": 0.0387, "step": 44430 }, { "grad_norm": 0.3589278757572174, "learning_rate": 6.317132295298134e-05, "loss": 0.0348, "step": 44440 }, { "grad_norm": 0.32234326004981995, "learning_rate": 6.315537154902173e-05, "loss": 0.0284, "step": 44450 }, { "grad_norm": 0.55597323179245, "learning_rate": 6.313941870641048e-05, "loss": 0.0394, "step": 44460 }, { "grad_norm": 0.3661746680736542, "learning_rate": 6.312346442689214e-05, "loss": 0.0414, "step": 44470 }, { "grad_norm": 0.4899676740169525, "learning_rate": 6.310750871221147e-05, "loss": 0.0371, "step": 44480 }, { "grad_norm": 1.4274559020996094, "learning_rate": 6.309155156411335e-05, "loss": 0.0319, "step": 44490 }, { "grad_norm": 0.8038927912712097, "learning_rate": 6.307559298434284e-05, "loss": 0.0403, "step": 44500 }, { "grad_norm": 0.5678172707557678, "learning_rate": 6.305963297464512e-05, "loss": 0.0513, "step": 44510 }, { "grad_norm": 0.5132642388343811, "learning_rate": 6.304367153676561e-05, "loss": 0.0458, "step": 44520 }, { "grad_norm": 0.7522934079170227, "learning_rate": 6.302770867244978e-05, "loss": 0.039, "step": 44530 }, { "grad_norm": 0.3329455852508545, "learning_rate": 6.301174438344328e-05, "loss": 0.0447, "step": 44540 }, { "grad_norm": 0.5219468474388123, "learning_rate": 6.299577867149202e-05, "loss": 0.0472, "step": 44550 }, { "grad_norm": 0.5158553123474121, "learning_rate": 6.297981153834192e-05, "loss": 0.0423, "step": 44560 }, { "grad_norm": 0.40548938512802124, "learning_rate": 6.296384298573916e-05, "loss": 0.0464, "step": 44570 }, { "grad_norm": 0.38065338134765625, "learning_rate": 6.294787301543001e-05, "loss": 0.0666, "step": 44580 }, { "grad_norm": 0.4677305221557617, "learning_rate": 6.293190162916095e-05, "loss": 0.0357, "step": 44590 }, { "grad_norm": 0.5218735337257385, "learning_rate": 6.291592882867855e-05, "loss": 0.0394, "step": 44600 }, { "grad_norm": 0.563856840133667, "learning_rate": 6.28999546157296e-05, "loss": 0.055, "step": 44610 }, { "grad_norm": 0.3532768189907074, "learning_rate": 6.288397899206102e-05, "loss": 0.0331, "step": 44620 }, { "grad_norm": 0.5285446643829346, "learning_rate": 6.286800195941984e-05, "loss": 0.0563, "step": 44630 }, { "grad_norm": 0.451962411403656, "learning_rate": 6.285202351955334e-05, "loss": 0.0409, "step": 44640 }, { "grad_norm": 0.5140925049781799, "learning_rate": 6.283604367420887e-05, "loss": 0.0473, "step": 44650 }, { "grad_norm": 0.5112884640693665, "learning_rate": 6.282006242513394e-05, "loss": 0.0406, "step": 44660 }, { "grad_norm": 0.4190581142902374, "learning_rate": 6.280407977407628e-05, "loss": 0.0468, "step": 44670 }, { "grad_norm": 0.3988383412361145, "learning_rate": 6.27880957227837e-05, "loss": 0.0397, "step": 44680 }, { "grad_norm": 0.4437159597873688, "learning_rate": 6.27721102730042e-05, "loss": 0.0432, "step": 44690 }, { "grad_norm": 0.4059925079345703, "learning_rate": 6.275612342648592e-05, "loss": 0.0411, "step": 44700 }, { "grad_norm": 0.8529931306838989, "learning_rate": 6.274013518497716e-05, "loss": 0.0411, "step": 44710 }, { "grad_norm": 0.48760509490966797, "learning_rate": 6.272414555022636e-05, "loss": 0.0382, "step": 44720 }, { "grad_norm": 0.5016079545021057, "learning_rate": 6.270815452398215e-05, "loss": 0.0432, "step": 44730 }, { "grad_norm": 0.31463921070098877, "learning_rate": 6.269216210799326e-05, "loss": 0.0559, "step": 44740 }, { "grad_norm": 0.40383341908454895, "learning_rate": 6.26761683040086e-05, "loss": 0.045, "step": 44750 }, { "grad_norm": 0.49610066413879395, "learning_rate": 6.266017311377723e-05, "loss": 0.0361, "step": 44760 }, { "grad_norm": 1.2608469724655151, "learning_rate": 6.264417653904839e-05, "loss": 0.0496, "step": 44770 }, { "grad_norm": 0.26354843378067017, "learning_rate": 6.262817858157139e-05, "loss": 0.0434, "step": 44780 }, { "grad_norm": 1.5204064846038818, "learning_rate": 6.261217924309576e-05, "loss": 0.051, "step": 44790 }, { "grad_norm": 0.5838943123817444, "learning_rate": 6.259617852537118e-05, "loss": 0.0395, "step": 44800 }, { "grad_norm": 0.7598920464515686, "learning_rate": 6.258017643014747e-05, "loss": 0.0454, "step": 44810 }, { "grad_norm": 0.4053135812282562, "learning_rate": 6.256417295917456e-05, "loss": 0.0536, "step": 44820 }, { "grad_norm": 0.6520607471466064, "learning_rate": 6.254816811420258e-05, "loss": 0.0422, "step": 44830 }, { "grad_norm": 0.2879360020160675, "learning_rate": 6.253216189698183e-05, "loss": 0.036, "step": 44840 }, { "grad_norm": 0.4467441141605377, "learning_rate": 6.251615430926267e-05, "loss": 0.0377, "step": 44850 }, { "grad_norm": 0.37825167179107666, "learning_rate": 6.25001453527957e-05, "loss": 0.0399, "step": 44860 }, { "grad_norm": 0.5359427332878113, "learning_rate": 6.248413502933164e-05, "loss": 0.0358, "step": 44870 }, { "grad_norm": 0.4008791148662567, "learning_rate": 6.246812334062133e-05, "loss": 0.0337, "step": 44880 }, { "grad_norm": 0.5880616903305054, "learning_rate": 6.245211028841579e-05, "loss": 0.0432, "step": 44890 }, { "grad_norm": 0.46915364265441895, "learning_rate": 6.24360958744662e-05, "loss": 0.0357, "step": 44900 }, { "grad_norm": 0.4719870090484619, "learning_rate": 6.242008010052387e-05, "loss": 0.0427, "step": 44910 }, { "grad_norm": 0.42955073714256287, "learning_rate": 6.240406296834024e-05, "loss": 0.0425, "step": 44920 }, { "grad_norm": 0.2633112370967865, "learning_rate": 6.238804447966694e-05, "loss": 0.0333, "step": 44930 }, { "grad_norm": 0.4438464641571045, "learning_rate": 6.237202463625573e-05, "loss": 0.0394, "step": 44940 }, { "grad_norm": 0.376086562871933, "learning_rate": 6.235600343985848e-05, "loss": 0.0399, "step": 44950 }, { "grad_norm": 0.4680398106575012, "learning_rate": 6.233998089222729e-05, "loss": 0.0328, "step": 44960 }, { "grad_norm": 0.41031262278556824, "learning_rate": 6.232395699511433e-05, "loss": 0.0477, "step": 44970 }, { "grad_norm": 0.2812710106372833, "learning_rate": 6.230793175027199e-05, "loss": 0.0429, "step": 44980 }, { "grad_norm": 0.9397768378257751, "learning_rate": 6.22919051594527e-05, "loss": 0.0308, "step": 44990 }, { "grad_norm": 0.4598744809627533, "learning_rate": 6.227587722440917e-05, "loss": 0.0544, "step": 45000 }, { "grad_norm": 0.40033039450645447, "learning_rate": 6.225984794689414e-05, "loss": 0.0314, "step": 45010 }, { "grad_norm": 0.5212623476982117, "learning_rate": 6.22438173286606e-05, "loss": 0.0413, "step": 45020 }, { "grad_norm": 0.276631623506546, "learning_rate": 6.222778537146159e-05, "loss": 0.0372, "step": 45030 }, { "grad_norm": 0.5950415730476379, "learning_rate": 6.221175207705037e-05, "loss": 0.0365, "step": 45040 }, { "grad_norm": 0.5367156863212585, "learning_rate": 6.219571744718027e-05, "loss": 0.039, "step": 45050 }, { "grad_norm": 0.4175492823123932, "learning_rate": 6.217968148360487e-05, "loss": 0.0363, "step": 45060 }, { "grad_norm": 0.3568282425403595, "learning_rate": 6.216364418807782e-05, "loss": 0.0405, "step": 45070 }, { "grad_norm": 0.46505117416381836, "learning_rate": 6.214760556235292e-05, "loss": 0.034, "step": 45080 }, { "grad_norm": 0.38217005133628845, "learning_rate": 6.213156560818414e-05, "loss": 0.0298, "step": 45090 }, { "grad_norm": 0.30538642406463623, "learning_rate": 6.211552432732559e-05, "loss": 0.0487, "step": 45100 }, { "grad_norm": 0.3846074342727661, "learning_rate": 6.20994817215315e-05, "loss": 0.0452, "step": 45110 }, { "grad_norm": 0.32025212049484253, "learning_rate": 6.208343779255628e-05, "loss": 0.0435, "step": 45120 }, { "grad_norm": 0.4213568866252899, "learning_rate": 6.206739254215449e-05, "loss": 0.0462, "step": 45130 }, { "grad_norm": 0.3847675323486328, "learning_rate": 6.205134597208077e-05, "loss": 0.0323, "step": 45140 }, { "grad_norm": 0.37386777997016907, "learning_rate": 6.203529808409e-05, "loss": 0.0378, "step": 45150 }, { "grad_norm": 0.5812889933586121, "learning_rate": 6.201924887993712e-05, "loss": 0.0493, "step": 45160 }, { "grad_norm": 0.39262762665748596, "learning_rate": 6.200319836137725e-05, "loss": 0.0422, "step": 45170 }, { "grad_norm": 0.22628040611743927, "learning_rate": 6.198714653016565e-05, "loss": 0.0405, "step": 45180 }, { "grad_norm": 0.5520639419555664, "learning_rate": 6.197109338805774e-05, "loss": 0.0431, "step": 45190 }, { "grad_norm": 0.335446298122406, "learning_rate": 6.195503893680903e-05, "loss": 0.0421, "step": 45200 }, { "grad_norm": 0.5315722227096558, "learning_rate": 6.193898317817524e-05, "loss": 0.0301, "step": 45210 }, { "grad_norm": 0.364370733499527, "learning_rate": 6.192292611391221e-05, "loss": 0.0338, "step": 45220 }, { "grad_norm": 0.5945119857788086, "learning_rate": 6.190686774577591e-05, "loss": 0.0439, "step": 45230 }, { "grad_norm": 0.45407167077064514, "learning_rate": 6.189080807552245e-05, "loss": 0.0429, "step": 45240 }, { "grad_norm": 0.608760416507721, "learning_rate": 6.187474710490809e-05, "loss": 0.042, "step": 45250 }, { "grad_norm": 0.389505535364151, "learning_rate": 6.185868483568926e-05, "loss": 0.0425, "step": 45260 }, { "grad_norm": 0.446431040763855, "learning_rate": 6.184262126962245e-05, "loss": 0.0352, "step": 45270 }, { "grad_norm": 0.4529285132884979, "learning_rate": 6.182655640846442e-05, "loss": 0.0309, "step": 45280 }, { "grad_norm": 0.5097358226776123, "learning_rate": 6.181049025397196e-05, "loss": 0.0396, "step": 45290 }, { "grad_norm": 0.43212243914604187, "learning_rate": 6.179442280790202e-05, "loss": 0.0505, "step": 45300 }, { "grad_norm": 0.35305026173591614, "learning_rate": 6.177835407201174e-05, "loss": 0.0368, "step": 45310 }, { "grad_norm": 0.3040350377559662, "learning_rate": 6.176228404805839e-05, "loss": 0.0345, "step": 45320 }, { "grad_norm": 0.382737398147583, "learning_rate": 6.174621273779932e-05, "loss": 0.0311, "step": 45330 }, { "grad_norm": 0.4796590507030487, "learning_rate": 6.17301401429921e-05, "loss": 0.0378, "step": 45340 }, { "grad_norm": 0.39400699734687805, "learning_rate": 6.171406626539436e-05, "loss": 0.0411, "step": 45350 }, { "grad_norm": 0.5570146441459656, "learning_rate": 6.169799110676398e-05, "loss": 0.0582, "step": 45360 }, { "grad_norm": 0.43276870250701904, "learning_rate": 6.168191466885885e-05, "loss": 0.0433, "step": 45370 }, { "grad_norm": 0.6463586091995239, "learning_rate": 6.166583695343711e-05, "loss": 0.0362, "step": 45380 }, { "grad_norm": 0.3922111988067627, "learning_rate": 6.164975796225698e-05, "loss": 0.0367, "step": 45390 }, { "grad_norm": 0.3745269775390625, "learning_rate": 6.163367769707683e-05, "loss": 0.0402, "step": 45400 }, { "grad_norm": 0.3021963834762573, "learning_rate": 6.161759615965519e-05, "loss": 0.0482, "step": 45410 }, { "grad_norm": 0.33571016788482666, "learning_rate": 6.16015133517507e-05, "loss": 0.0376, "step": 45420 }, { "grad_norm": 0.31890931725502014, "learning_rate": 6.158542927512214e-05, "loss": 0.0399, "step": 45430 }, { "grad_norm": 0.30416616797447205, "learning_rate": 6.156934393152846e-05, "loss": 0.0432, "step": 45440 }, { "grad_norm": 0.4262513518333435, "learning_rate": 6.15532573227287e-05, "loss": 0.0442, "step": 45450 }, { "grad_norm": 0.5389230251312256, "learning_rate": 6.153716945048212e-05, "loss": 0.0478, "step": 45460 }, { "grad_norm": 0.404948890209198, "learning_rate": 6.152108031654802e-05, "loss": 0.0511, "step": 45470 }, { "grad_norm": 0.41727033257484436, "learning_rate": 6.15049899226859e-05, "loss": 0.0444, "step": 45480 }, { "grad_norm": 0.8841757774353027, "learning_rate": 6.148889827065537e-05, "loss": 0.0334, "step": 45490 }, { "grad_norm": 0.41611218452453613, "learning_rate": 6.147280536221622e-05, "loss": 0.0322, "step": 45500 }, { "grad_norm": 0.530813992023468, "learning_rate": 6.145671119912832e-05, "loss": 0.0404, "step": 45510 }, { "grad_norm": 0.5674512982368469, "learning_rate": 6.144061578315169e-05, "loss": 0.0365, "step": 45520 }, { "grad_norm": 0.5403916239738464, "learning_rate": 6.142451911604654e-05, "loss": 0.0282, "step": 45530 }, { "grad_norm": 0.3830164670944214, "learning_rate": 6.140842119957315e-05, "loss": 0.0312, "step": 45540 }, { "grad_norm": 0.7737224698066711, "learning_rate": 6.139232203549197e-05, "loss": 0.0469, "step": 45550 }, { "grad_norm": 0.6038124561309814, "learning_rate": 6.137622162556357e-05, "loss": 0.0411, "step": 45560 }, { "grad_norm": 0.42811620235443115, "learning_rate": 6.13601199715487e-05, "loss": 0.0451, "step": 45570 }, { "grad_norm": 0.35055845975875854, "learning_rate": 6.134401707520816e-05, "loss": 0.0283, "step": 45580 }, { "grad_norm": 0.35074132680892944, "learning_rate": 6.132791293830298e-05, "loss": 0.034, "step": 45590 }, { "grad_norm": 0.28837206959724426, "learning_rate": 6.131180756259428e-05, "loss": 0.0387, "step": 45600 }, { "grad_norm": 0.4254339933395386, "learning_rate": 6.129570094984331e-05, "loss": 0.0369, "step": 45610 }, { "grad_norm": 1.415874719619751, "learning_rate": 6.127959310181145e-05, "loss": 0.0454, "step": 45620 }, { "grad_norm": 0.42064806818962097, "learning_rate": 6.126348402026026e-05, "loss": 0.0355, "step": 45630 }, { "grad_norm": 0.36811622977256775, "learning_rate": 6.12473737069514e-05, "loss": 0.0386, "step": 45640 }, { "grad_norm": 0.3020230829715729, "learning_rate": 6.123126216364665e-05, "loss": 0.0371, "step": 45650 }, { "grad_norm": 0.6843202710151672, "learning_rate": 6.121514939210797e-05, "loss": 0.0351, "step": 45660 }, { "grad_norm": 0.36551085114479065, "learning_rate": 6.119903539409741e-05, "loss": 0.0408, "step": 45670 }, { "grad_norm": 0.8379408121109009, "learning_rate": 6.118292017137716e-05, "loss": 0.0457, "step": 45680 }, { "grad_norm": 0.3403092622756958, "learning_rate": 6.116680372570959e-05, "loss": 0.0288, "step": 45690 }, { "grad_norm": 0.4035106599330902, "learning_rate": 6.115068605885713e-05, "loss": 0.0361, "step": 45700 }, { "grad_norm": 0.42884793877601624, "learning_rate": 6.113456717258243e-05, "loss": 0.0438, "step": 45710 }, { "grad_norm": 0.5338648557662964, "learning_rate": 6.11184470686482e-05, "loss": 0.039, "step": 45720 }, { "grad_norm": 0.3958989381790161, "learning_rate": 6.110232574881731e-05, "loss": 0.037, "step": 45730 }, { "grad_norm": 0.3391076922416687, "learning_rate": 6.108620321485277e-05, "loss": 0.0415, "step": 45740 }, { "grad_norm": 0.358471542596817, "learning_rate": 6.107007946851773e-05, "loss": 0.0266, "step": 45750 }, { "grad_norm": 0.2755391001701355, "learning_rate": 6.105395451157542e-05, "loss": 0.034, "step": 45760 }, { "grad_norm": 0.46740078926086426, "learning_rate": 6.103782834578928e-05, "loss": 0.0384, "step": 45770 }, { "grad_norm": 0.2662505507469177, "learning_rate": 6.102170097292281e-05, "loss": 0.0357, "step": 45780 }, { "grad_norm": 0.4413604140281677, "learning_rate": 6.1005572394739716e-05, "loss": 0.0383, "step": 45790 }, { "grad_norm": 0.3992461860179901, "learning_rate": 6.0989442613003765e-05, "loss": 0.043, "step": 45800 }, { "grad_norm": 0.40597814321517944, "learning_rate": 6.097331162947888e-05, "loss": 0.0425, "step": 45810 }, { "grad_norm": 0.4186197817325592, "learning_rate": 6.095717944592914e-05, "loss": 0.0446, "step": 45820 }, { "grad_norm": 0.5549944043159485, "learning_rate": 6.094104606411873e-05, "loss": 0.0542, "step": 45830 }, { "grad_norm": 0.6178275346755981, "learning_rate": 6.0924911485811966e-05, "loss": 0.0612, "step": 45840 }, { "grad_norm": 0.3925275504589081, "learning_rate": 6.090877571277331e-05, "loss": 0.0519, "step": 45850 }, { "grad_norm": 0.4451814889907837, "learning_rate": 6.089263874676736e-05, "loss": 0.0433, "step": 45860 }, { "grad_norm": 0.46344777941703796, "learning_rate": 6.0876500589558796e-05, "loss": 0.0367, "step": 45870 }, { "grad_norm": 0.92072594165802, "learning_rate": 6.086036124291248e-05, "loss": 0.0346, "step": 45880 }, { "grad_norm": 0.5793653130531311, "learning_rate": 6.084422070859339e-05, "loss": 0.0511, "step": 45890 }, { "grad_norm": 0.30479809641838074, "learning_rate": 6.082807898836663e-05, "loss": 0.0343, "step": 45900 }, { "grad_norm": 0.2643166780471802, "learning_rate": 6.081193608399742e-05, "loss": 0.0342, "step": 45910 }, { "grad_norm": 0.48410552740097046, "learning_rate": 6.0795791997251164e-05, "loss": 0.0359, "step": 45920 }, { "grad_norm": 0.43602901697158813, "learning_rate": 6.0779646729893294e-05, "loss": 0.0306, "step": 45930 }, { "grad_norm": 0.648554801940918, "learning_rate": 6.0763500283689476e-05, "loss": 0.0343, "step": 45940 }, { "grad_norm": 0.4988363981246948, "learning_rate": 6.0747352660405455e-05, "loss": 0.0381, "step": 45950 }, { "grad_norm": 0.43754148483276367, "learning_rate": 6.073120386180709e-05, "loss": 0.0376, "step": 45960 }, { "grad_norm": 0.3547087013721466, "learning_rate": 6.0715053889660425e-05, "loss": 0.0327, "step": 45970 }, { "grad_norm": 0.29226481914520264, "learning_rate": 6.069890274573157e-05, "loss": 0.0344, "step": 45980 }, { "grad_norm": 0.3107514977455139, "learning_rate": 6.068275043178679e-05, "loss": 0.0476, "step": 45990 }, { "grad_norm": 1.0258543491363525, "learning_rate": 6.066659694959248e-05, "loss": 0.0371, "step": 46000 }, { "grad_norm": 0.2959460914134979, "learning_rate": 6.0650442300915176e-05, "loss": 0.0368, "step": 46010 }, { "grad_norm": 0.3819555342197418, "learning_rate": 6.063428648752152e-05, "loss": 0.0332, "step": 46020 }, { "grad_norm": 0.4514642655849457, "learning_rate": 6.0618129511178266e-05, "loss": 0.0416, "step": 46030 }, { "grad_norm": 0.3590140640735626, "learning_rate": 6.060197137365234e-05, "loss": 0.0314, "step": 46040 }, { "grad_norm": 0.49317070841789246, "learning_rate": 6.058581207671077e-05, "loss": 0.0396, "step": 46050 }, { "grad_norm": 0.4210126996040344, "learning_rate": 6.056965162212072e-05, "loss": 0.0399, "step": 46060 }, { "grad_norm": 0.4789047837257385, "learning_rate": 6.055349001164943e-05, "loss": 0.0479, "step": 46070 }, { "grad_norm": 0.3900507688522339, "learning_rate": 6.0537327247064347e-05, "loss": 0.0373, "step": 46080 }, { "grad_norm": 0.526302695274353, "learning_rate": 6.0521163330133014e-05, "loss": 0.0446, "step": 46090 }, { "grad_norm": 0.36614176630973816, "learning_rate": 6.050499826262306e-05, "loss": 0.0382, "step": 46100 }, { "grad_norm": 0.26574069261550903, "learning_rate": 6.0488832046302294e-05, "loss": 0.0445, "step": 46110 }, { "grad_norm": 0.35191598534584045, "learning_rate": 6.0472664682938626e-05, "loss": 0.032, "step": 46120 }, { "grad_norm": 0.3660490810871124, "learning_rate": 6.045649617430009e-05, "loss": 0.032, "step": 46130 }, { "grad_norm": 0.5003899931907654, "learning_rate": 6.0440326522154866e-05, "loss": 0.0337, "step": 46140 }, { "grad_norm": 0.485769122838974, "learning_rate": 6.0424155728271224e-05, "loss": 0.0336, "step": 46150 }, { "grad_norm": 0.7005007266998291, "learning_rate": 6.040798379441758e-05, "loss": 0.033, "step": 46160 }, { "grad_norm": 0.6051247119903564, "learning_rate": 6.0391810722362485e-05, "loss": 0.0369, "step": 46170 }, { "grad_norm": 0.4653175473213196, "learning_rate": 6.037563651387458e-05, "loss": 0.0437, "step": 46180 }, { "grad_norm": 0.3861212432384491, "learning_rate": 6.0359461170722666e-05, "loss": 0.0404, "step": 46190 }, { "grad_norm": 0.37725380063056946, "learning_rate": 6.034328469467566e-05, "loss": 0.0445, "step": 46200 }, { "grad_norm": 0.8063200116157532, "learning_rate": 6.0327107087502596e-05, "loss": 0.0392, "step": 46210 }, { "grad_norm": 0.4098213315010071, "learning_rate": 6.031092835097262e-05, "loss": 0.042, "step": 46220 }, { "grad_norm": 0.6378278732299805, "learning_rate": 6.0294748486855024e-05, "loss": 0.04, "step": 46230 }, { "grad_norm": 0.33947890996932983, "learning_rate": 6.0278567496919216e-05, "loss": 0.0359, "step": 46240 }, { "grad_norm": 0.3768307864665985, "learning_rate": 6.026238538293472e-05, "loss": 0.0365, "step": 46250 }, { "grad_norm": 0.4702341854572296, "learning_rate": 6.024620214667118e-05, "loss": 0.044, "step": 46260 }, { "grad_norm": 0.4257024824619293, "learning_rate": 6.0230017789898384e-05, "loss": 0.0272, "step": 46270 }, { "grad_norm": 0.3063507378101349, "learning_rate": 6.021383231438622e-05, "loss": 0.0314, "step": 46280 }, { "grad_norm": 0.2922447621822357, "learning_rate": 6.0197645721904704e-05, "loss": 0.0359, "step": 46290 }, { "grad_norm": 0.5218875408172607, "learning_rate": 6.0181458014224e-05, "loss": 0.0425, "step": 46300 }, { "grad_norm": 0.6320068836212158, "learning_rate": 6.016526919311435e-05, "loss": 0.042, "step": 46310 }, { "grad_norm": 0.4735645353794098, "learning_rate": 6.014907926034613e-05, "loss": 0.0398, "step": 46320 }, { "grad_norm": 0.7997386455535889, "learning_rate": 6.0132888217689875e-05, "loss": 0.0349, "step": 46330 }, { "grad_norm": 0.5111150145530701, "learning_rate": 6.01166960669162e-05, "loss": 0.0506, "step": 46340 }, { "grad_norm": 0.3177053928375244, "learning_rate": 6.0100502809795845e-05, "loss": 0.0351, "step": 46350 }, { "grad_norm": 2.151235580444336, "learning_rate": 6.00843084480997e-05, "loss": 0.0525, "step": 46360 }, { "grad_norm": 0.8278418183326721, "learning_rate": 6.0068112983598736e-05, "loss": 0.0375, "step": 46370 }, { "grad_norm": 0.4166147708892822, "learning_rate": 6.005191641806407e-05, "loss": 0.0447, "step": 46380 }, { "grad_norm": 0.4901234209537506, "learning_rate": 6.003571875326694e-05, "loss": 0.0395, "step": 46390 }, { "grad_norm": 0.2719017267227173, "learning_rate": 6.00195199909787e-05, "loss": 0.034, "step": 46400 }, { "grad_norm": 0.38162097334861755, "learning_rate": 6.000332013297082e-05, "loss": 0.0318, "step": 46410 }, { "grad_norm": 0.46856364607810974, "learning_rate": 5.998711918101487e-05, "loss": 0.0402, "step": 46420 }, { "grad_norm": 0.4415574371814728, "learning_rate": 5.997091713688261e-05, "loss": 0.0353, "step": 46430 }, { "grad_norm": 0.4375356137752533, "learning_rate": 5.9954714002345836e-05, "loss": 0.0344, "step": 46440 }, { "grad_norm": 0.31199562549591064, "learning_rate": 5.993850977917649e-05, "loss": 0.0458, "step": 46450 }, { "grad_norm": 0.55405592918396, "learning_rate": 5.992230446914667e-05, "loss": 0.042, "step": 46460 }, { "grad_norm": 0.29327550530433655, "learning_rate": 5.990609807402855e-05, "loss": 0.0341, "step": 46470 }, { "grad_norm": 0.3311910927295685, "learning_rate": 5.988989059559443e-05, "loss": 0.0404, "step": 46480 }, { "grad_norm": 0.521793007850647, "learning_rate": 5.987368203561675e-05, "loss": 0.0384, "step": 46490 }, { "grad_norm": 0.5899183750152588, "learning_rate": 5.9857472395868055e-05, "loss": 0.0361, "step": 46500 }, { "grad_norm": 0.33440735936164856, "learning_rate": 5.9841261678120983e-05, "loss": 0.0313, "step": 46510 }, { "grad_norm": 0.3991660177707672, "learning_rate": 5.982504988414834e-05, "loss": 0.0311, "step": 46520 }, { "grad_norm": 0.5062770247459412, "learning_rate": 5.9808837015723015e-05, "loss": 0.054, "step": 46530 }, { "grad_norm": 0.8202290534973145, "learning_rate": 5.9792623074618016e-05, "loss": 0.0356, "step": 46540 }, { "grad_norm": 0.36190265417099, "learning_rate": 5.977640806260648e-05, "loss": 0.0371, "step": 46550 }, { "grad_norm": 0.29287105798721313, "learning_rate": 5.976019198146166e-05, "loss": 0.0239, "step": 46560 }, { "grad_norm": 0.2878519296646118, "learning_rate": 5.974397483295692e-05, "loss": 0.0283, "step": 46570 }, { "grad_norm": 0.7974435091018677, "learning_rate": 5.972775661886575e-05, "loss": 0.0381, "step": 46580 }, { "grad_norm": 0.3403075337409973, "learning_rate": 5.971153734096173e-05, "loss": 0.0382, "step": 46590 }, { "grad_norm": 0.21924753487110138, "learning_rate": 5.969531700101859e-05, "loss": 0.0317, "step": 46600 }, { "grad_norm": 0.39919373393058777, "learning_rate": 5.9679095600810155e-05, "loss": 0.0455, "step": 46610 }, { "grad_norm": 0.6109535098075867, "learning_rate": 5.9662873142110384e-05, "loss": 0.0375, "step": 46620 }, { "grad_norm": 0.4248720407485962, "learning_rate": 5.964664962669333e-05, "loss": 0.0399, "step": 46630 }, { "grad_norm": 0.423453688621521, "learning_rate": 5.9630425056333186e-05, "loss": 0.0387, "step": 46640 }, { "grad_norm": 0.43897518515586853, "learning_rate": 5.961419943280422e-05, "loss": 0.0436, "step": 46650 }, { "grad_norm": 0.3176755905151367, "learning_rate": 5.959797275788087e-05, "loss": 0.0325, "step": 46660 }, { "grad_norm": 0.34200242161750793, "learning_rate": 5.958174503333765e-05, "loss": 0.0371, "step": 46670 }, { "grad_norm": 0.47032830119132996, "learning_rate": 5.9565516260949195e-05, "loss": 0.0472, "step": 46680 }, { "grad_norm": 2.4509823322296143, "learning_rate": 5.9549286442490273e-05, "loss": 0.0416, "step": 46690 }, { "grad_norm": 0.6306917667388916, "learning_rate": 5.953305557973572e-05, "loss": 0.0397, "step": 46700 }, { "grad_norm": 0.9066749811172485, "learning_rate": 5.9516823674460564e-05, "loss": 0.0344, "step": 46710 }, { "grad_norm": 0.3148701786994934, "learning_rate": 5.9500590728439875e-05, "loss": 0.0417, "step": 46720 }, { "grad_norm": 0.2758888602256775, "learning_rate": 5.948435674344886e-05, "loss": 0.0338, "step": 46730 }, { "grad_norm": 0.3523917496204376, "learning_rate": 5.946812172126285e-05, "loss": 0.0268, "step": 46740 }, { "grad_norm": 0.306266725063324, "learning_rate": 5.945188566365729e-05, "loss": 0.0372, "step": 46750 }, { "grad_norm": 0.2679465115070343, "learning_rate": 5.943564857240773e-05, "loss": 0.0391, "step": 46760 }, { "grad_norm": 0.43442368507385254, "learning_rate": 5.941941044928981e-05, "loss": 0.031, "step": 46770 }, { "grad_norm": 0.5182177424430847, "learning_rate": 5.940317129607935e-05, "loss": 0.039, "step": 46780 }, { "grad_norm": 0.3699520230293274, "learning_rate": 5.9386931114552204e-05, "loss": 0.0319, "step": 46790 }, { "grad_norm": 0.4004666805267334, "learning_rate": 5.937068990648438e-05, "loss": 0.0398, "step": 46800 }, { "grad_norm": 0.5167508721351624, "learning_rate": 5.935444767365199e-05, "loss": 0.0524, "step": 46810 }, { "grad_norm": 0.36967170238494873, "learning_rate": 5.933820441783129e-05, "loss": 0.0417, "step": 46820 }, { "grad_norm": 0.30560579895973206, "learning_rate": 5.932196014079857e-05, "loss": 0.0449, "step": 46830 }, { "grad_norm": 0.4527992010116577, "learning_rate": 5.930571484433032e-05, "loss": 0.0296, "step": 46840 }, { "grad_norm": 0.1838182806968689, "learning_rate": 5.928946853020309e-05, "loss": 0.0304, "step": 46850 }, { "grad_norm": 0.40542274713516235, "learning_rate": 5.927322120019352e-05, "loss": 0.0362, "step": 46860 }, { "grad_norm": 1.9345418214797974, "learning_rate": 5.925697285607844e-05, "loss": 0.0396, "step": 46870 }, { "grad_norm": 0.44996482133865356, "learning_rate": 5.924072349963472e-05, "loss": 0.0435, "step": 46880 }, { "grad_norm": 0.3264370560646057, "learning_rate": 5.922447313263938e-05, "loss": 0.0407, "step": 46890 }, { "grad_norm": 0.3445589244365692, "learning_rate": 5.920822175686952e-05, "loss": 0.034, "step": 46900 }, { "grad_norm": 0.4044019877910614, "learning_rate": 5.919196937410239e-05, "loss": 0.0321, "step": 46910 }, { "grad_norm": 0.5782977342605591, "learning_rate": 5.917571598611529e-05, "loss": 0.0378, "step": 46920 }, { "grad_norm": 0.38585761189460754, "learning_rate": 5.9159461594685686e-05, "loss": 0.0324, "step": 46930 }, { "grad_norm": 0.5931575894355774, "learning_rate": 5.914320620159114e-05, "loss": 0.0458, "step": 46940 }, { "grad_norm": 0.4292575716972351, "learning_rate": 5.912694980860931e-05, "loss": 0.0317, "step": 46950 }, { "grad_norm": 0.5199536085128784, "learning_rate": 5.9110692417517964e-05, "loss": 0.0372, "step": 46960 }, { "grad_norm": 0.3839050531387329, "learning_rate": 5.9094434030095e-05, "loss": 0.0367, "step": 46970 }, { "grad_norm": 0.3378833532333374, "learning_rate": 5.907817464811841e-05, "loss": 0.0368, "step": 46980 }, { "grad_norm": 0.5341875553131104, "learning_rate": 5.906191427336627e-05, "loss": 0.0384, "step": 46990 }, { "grad_norm": 0.35817283391952515, "learning_rate": 5.9045652907616835e-05, "loss": 0.0328, "step": 47000 }, { "grad_norm": 0.5871265530586243, "learning_rate": 5.902939055264838e-05, "loss": 0.0389, "step": 47010 }, { "grad_norm": 0.3970034718513489, "learning_rate": 5.901312721023935e-05, "loss": 0.0324, "step": 47020 }, { "grad_norm": 0.4807403087615967, "learning_rate": 5.8996862882168294e-05, "loss": 0.0423, "step": 47030 }, { "grad_norm": 0.5246493816375732, "learning_rate": 5.8980597570213826e-05, "loss": 0.0501, "step": 47040 }, { "grad_norm": 0.5193317532539368, "learning_rate": 5.896433127615471e-05, "loss": 0.0337, "step": 47050 }, { "grad_norm": 0.3767804503440857, "learning_rate": 5.894806400176981e-05, "loss": 0.0341, "step": 47060 }, { "grad_norm": 0.2535938024520874, "learning_rate": 5.893179574883808e-05, "loss": 0.0338, "step": 47070 }, { "grad_norm": 0.45580995082855225, "learning_rate": 5.8915526519138585e-05, "loss": 0.0375, "step": 47080 }, { "grad_norm": 0.7329106330871582, "learning_rate": 5.889925631445053e-05, "loss": 0.0376, "step": 47090 }, { "grad_norm": 0.24893638491630554, "learning_rate": 5.888298513655318e-05, "loss": 0.0424, "step": 47100 }, { "grad_norm": 0.4078856408596039, "learning_rate": 5.886671298722592e-05, "loss": 0.0303, "step": 47110 }, { "grad_norm": 0.3263024389743805, "learning_rate": 5.885043986824828e-05, "loss": 0.0313, "step": 47120 }, { "grad_norm": 0.4918375611305237, "learning_rate": 5.883416578139982e-05, "loss": 0.0371, "step": 47130 }, { "grad_norm": 0.7220318913459778, "learning_rate": 5.881789072846029e-05, "loss": 0.0456, "step": 47140 }, { "grad_norm": 0.48395711183547974, "learning_rate": 5.8801614711209474e-05, "loss": 0.0252, "step": 47150 }, { "grad_norm": 0.409498929977417, "learning_rate": 5.878533773142732e-05, "loss": 0.0473, "step": 47160 }, { "grad_norm": 0.424907922744751, "learning_rate": 5.8769059790893834e-05, "loss": 0.0417, "step": 47170 }, { "grad_norm": 0.5717907547950745, "learning_rate": 5.8752780891389145e-05, "loss": 0.0389, "step": 47180 }, { "grad_norm": 0.3564455807209015, "learning_rate": 5.873650103469351e-05, "loss": 0.0386, "step": 47190 }, { "grad_norm": 0.6699652075767517, "learning_rate": 5.872022022258725e-05, "loss": 0.043, "step": 47200 }, { "grad_norm": 0.37154319882392883, "learning_rate": 5.8703938456850814e-05, "loss": 0.044, "step": 47210 }, { "grad_norm": 0.6314781904220581, "learning_rate": 5.868765573926477e-05, "loss": 0.0458, "step": 47220 }, { "grad_norm": 0.4459134340286255, "learning_rate": 5.867137207160974e-05, "loss": 0.0393, "step": 47230 }, { "grad_norm": 0.5262450575828552, "learning_rate": 5.86550874556665e-05, "loss": 0.0372, "step": 47240 }, { "grad_norm": 0.6384997367858887, "learning_rate": 5.863880189321592e-05, "loss": 0.0414, "step": 47250 }, { "grad_norm": 0.5377941727638245, "learning_rate": 5.862251538603894e-05, "loss": 0.0452, "step": 47260 }, { "grad_norm": 0.4159742593765259, "learning_rate": 5.8606227935916656e-05, "loss": 0.0393, "step": 47270 }, { "grad_norm": 0.3028666079044342, "learning_rate": 5.858993954463021e-05, "loss": 0.0318, "step": 47280 }, { "grad_norm": 0.7636559009552002, "learning_rate": 5.8573650213960896e-05, "loss": 0.0494, "step": 47290 }, { "grad_norm": 0.34134790301322937, "learning_rate": 5.8557359945690084e-05, "loss": 0.0329, "step": 47300 }, { "grad_norm": 0.560337245464325, "learning_rate": 5.8541068741599246e-05, "loss": 0.0365, "step": 47310 }, { "grad_norm": 0.36861565709114075, "learning_rate": 5.8524776603469985e-05, "loss": 0.0373, "step": 47320 }, { "grad_norm": 0.26235470175743103, "learning_rate": 5.850848353308397e-05, "loss": 0.0272, "step": 47330 }, { "grad_norm": 0.4819171726703644, "learning_rate": 5.849218953222297e-05, "loss": 0.0328, "step": 47340 }, { "grad_norm": 0.9066927433013916, "learning_rate": 5.847589460266891e-05, "loss": 0.0399, "step": 47350 }, { "grad_norm": 0.3749403655529022, "learning_rate": 5.845959874620375e-05, "loss": 0.0357, "step": 47360 }, { "grad_norm": 0.6166494488716125, "learning_rate": 5.84433019646096e-05, "loss": 0.0381, "step": 47370 }, { "grad_norm": 0.818290650844574, "learning_rate": 5.842700425966863e-05, "loss": 0.0419, "step": 47380 }, { "grad_norm": 0.7242274880409241, "learning_rate": 5.841070563316315e-05, "loss": 0.0366, "step": 47390 }, { "grad_norm": 0.47328320145606995, "learning_rate": 5.839440608687554e-05, "loss": 0.0401, "step": 47400 }, { "grad_norm": 0.32448649406433105, "learning_rate": 5.837810562258831e-05, "loss": 0.0316, "step": 47410 }, { "grad_norm": 0.5198994874954224, "learning_rate": 5.836180424208405e-05, "loss": 0.0492, "step": 47420 }, { "grad_norm": 0.4289017915725708, "learning_rate": 5.834550194714543e-05, "loss": 0.0338, "step": 47430 }, { "grad_norm": 0.34284016489982605, "learning_rate": 5.832919873955528e-05, "loss": 0.0533, "step": 47440 }, { "grad_norm": 0.40520718693733215, "learning_rate": 5.831289462109648e-05, "loss": 0.036, "step": 47450 }, { "grad_norm": 0.441293329000473, "learning_rate": 5.829658959355202e-05, "loss": 0.0516, "step": 47460 }, { "grad_norm": 0.41237175464630127, "learning_rate": 5.8280283658704995e-05, "loss": 0.0419, "step": 47470 }, { "grad_norm": 0.41051235795021057, "learning_rate": 5.826397681833859e-05, "loss": 0.0321, "step": 47480 }, { "grad_norm": 0.7968671917915344, "learning_rate": 5.824766907423612e-05, "loss": 0.0292, "step": 47490 }, { "grad_norm": 0.2822308838367462, "learning_rate": 5.8231360428180945e-05, "loss": 0.0365, "step": 47500 }, { "grad_norm": 0.3688790798187256, "learning_rate": 5.821505088195658e-05, "loss": 0.0349, "step": 47510 }, { "grad_norm": 0.3878319561481476, "learning_rate": 5.819874043734661e-05, "loss": 0.0414, "step": 47520 }, { "grad_norm": 0.5558347105979919, "learning_rate": 5.8182429096134695e-05, "loss": 0.0351, "step": 47530 }, { "grad_norm": 0.395559161901474, "learning_rate": 5.816611686010465e-05, "loss": 0.027, "step": 47540 }, { "grad_norm": 0.3994746804237366, "learning_rate": 5.814980373104033e-05, "loss": 0.0471, "step": 47550 }, { "grad_norm": 0.6522774696350098, "learning_rate": 5.813348971072572e-05, "loss": 0.0458, "step": 47560 }, { "grad_norm": 0.39023253321647644, "learning_rate": 5.811717480094492e-05, "loss": 0.0431, "step": 47570 }, { "grad_norm": 0.44093188643455505, "learning_rate": 5.810085900348209e-05, "loss": 0.0334, "step": 47580 }, { "grad_norm": 0.30922162532806396, "learning_rate": 5.8084542320121483e-05, "loss": 0.0448, "step": 47590 }, { "grad_norm": 0.4120861887931824, "learning_rate": 5.8068224752647497e-05, "loss": 0.049, "step": 47600 }, { "grad_norm": 0.27523958683013916, "learning_rate": 5.805190630284457e-05, "loss": 0.0399, "step": 47610 }, { "grad_norm": 0.5445250272750854, "learning_rate": 5.803558697249729e-05, "loss": 0.03, "step": 47620 }, { "grad_norm": 0.312734991312027, "learning_rate": 5.8019266763390276e-05, "loss": 0.0431, "step": 47630 }, { "grad_norm": 0.4739578068256378, "learning_rate": 5.800294567730833e-05, "loss": 0.0369, "step": 47640 }, { "grad_norm": 0.47452303767204285, "learning_rate": 5.798662371603628e-05, "loss": 0.0366, "step": 47650 }, { "grad_norm": 0.7846161127090454, "learning_rate": 5.797030088135904e-05, "loss": 0.0303, "step": 47660 }, { "grad_norm": 0.2541068494319916, "learning_rate": 5.7953977175061705e-05, "loss": 0.0336, "step": 47670 }, { "grad_norm": 2.479403257369995, "learning_rate": 5.793765259892938e-05, "loss": 0.0368, "step": 47680 }, { "grad_norm": 0.30580615997314453, "learning_rate": 5.792132715474729e-05, "loss": 0.0238, "step": 47690 }, { "grad_norm": 0.4077453911304474, "learning_rate": 5.790500084430078e-05, "loss": 0.0269, "step": 47700 }, { "grad_norm": 0.24406686425209045, "learning_rate": 5.7888673669375274e-05, "loss": 0.0373, "step": 47710 }, { "grad_norm": 0.6375353336334229, "learning_rate": 5.787234563175625e-05, "loss": 0.0281, "step": 47720 }, { "grad_norm": 0.3630277216434479, "learning_rate": 5.7856016733229355e-05, "loss": 0.0338, "step": 47730 }, { "grad_norm": 0.554499089717865, "learning_rate": 5.7839686975580297e-05, "loss": 0.0418, "step": 47740 }, { "grad_norm": 0.5332739949226379, "learning_rate": 5.782335636059484e-05, "loss": 0.0267, "step": 47750 }, { "grad_norm": 0.38786399364471436, "learning_rate": 5.780702489005889e-05, "loss": 0.0357, "step": 47760 }, { "grad_norm": 2.8232316970825195, "learning_rate": 5.779069256575845e-05, "loss": 0.0439, "step": 47770 }, { "grad_norm": 0.5554417371749878, "learning_rate": 5.7774359389479574e-05, "loss": 0.0476, "step": 47780 }, { "grad_norm": 0.8817017078399658, "learning_rate": 5.775802536300845e-05, "loss": 0.0481, "step": 47790 }, { "grad_norm": 0.2946726381778717, "learning_rate": 5.774169048813134e-05, "loss": 0.0513, "step": 47800 }, { "grad_norm": 0.41684049367904663, "learning_rate": 5.77253547666346e-05, "loss": 0.0353, "step": 47810 }, { "grad_norm": 0.5129040479660034, "learning_rate": 5.770901820030465e-05, "loss": 0.0403, "step": 47820 }, { "grad_norm": 0.3231360912322998, "learning_rate": 5.769268079092809e-05, "loss": 0.0379, "step": 47830 }, { "grad_norm": 0.41220420598983765, "learning_rate": 5.767634254029151e-05, "loss": 0.0341, "step": 47840 }, { "grad_norm": 0.4859035313129425, "learning_rate": 5.7660003450181655e-05, "loss": 0.0568, "step": 47850 }, { "grad_norm": 0.33764156699180603, "learning_rate": 5.764366352238534e-05, "loss": 0.024, "step": 47860 }, { "grad_norm": 0.2771601676940918, "learning_rate": 5.7627322758689474e-05, "loss": 0.0287, "step": 47870 }, { "grad_norm": 0.5487275123596191, "learning_rate": 5.761098116088105e-05, "loss": 0.0366, "step": 47880 }, { "grad_norm": 0.6549267768859863, "learning_rate": 5.759463873074717e-05, "loss": 0.0382, "step": 47890 }, { "grad_norm": 0.326351523399353, "learning_rate": 5.757829547007504e-05, "loss": 0.0423, "step": 47900 }, { "grad_norm": 0.5005651116371155, "learning_rate": 5.756195138065189e-05, "loss": 0.0376, "step": 47910 }, { "grad_norm": 0.3980076313018799, "learning_rate": 5.754560646426511e-05, "loss": 0.0429, "step": 47920 }, { "grad_norm": 0.5303959250450134, "learning_rate": 5.752926072270216e-05, "loss": 0.0446, "step": 47930 }, { "grad_norm": 0.7973228096961975, "learning_rate": 5.7512914157750563e-05, "loss": 0.0334, "step": 47940 }, { "grad_norm": 0.5170308947563171, "learning_rate": 5.749656677119798e-05, "loss": 0.0349, "step": 47950 }, { "grad_norm": 0.4964102506637573, "learning_rate": 5.7480218564832125e-05, "loss": 0.0441, "step": 47960 }, { "grad_norm": 0.6337239742279053, "learning_rate": 5.746386954044082e-05, "loss": 0.0458, "step": 47970 }, { "grad_norm": 0.317846417427063, "learning_rate": 5.744751969981195e-05, "loss": 0.0284, "step": 47980 }, { "grad_norm": 0.4239121675491333, "learning_rate": 5.7431169044733526e-05, "loss": 0.0489, "step": 47990 }, { "grad_norm": 0.31981465220451355, "learning_rate": 5.741481757699364e-05, "loss": 0.0318, "step": 48000 }, { "grad_norm": 0.30306798219680786, "learning_rate": 5.7398465298380434e-05, "loss": 0.0304, "step": 48010 }, { "grad_norm": 0.3134138286113739, "learning_rate": 5.7382112210682193e-05, "loss": 0.0474, "step": 48020 }, { "grad_norm": 0.41608932614326477, "learning_rate": 5.7365758315687266e-05, "loss": 0.0346, "step": 48030 }, { "grad_norm": 0.5294435620307922, "learning_rate": 5.734940361518407e-05, "loss": 0.0312, "step": 48040 }, { "grad_norm": 0.3169447183609009, "learning_rate": 5.733304811096116e-05, "loss": 0.0314, "step": 48050 }, { "grad_norm": 0.4835958778858185, "learning_rate": 5.731669180480713e-05, "loss": 0.0403, "step": 48060 }, { "grad_norm": 0.460553377866745, "learning_rate": 5.730033469851067e-05, "loss": 0.0468, "step": 48070 }, { "grad_norm": 0.5596362948417664, "learning_rate": 5.7283976793860607e-05, "loss": 0.0434, "step": 48080 }, { "grad_norm": 0.37357237935066223, "learning_rate": 5.726761809264577e-05, "loss": 0.0449, "step": 48090 }, { "grad_norm": 0.30823931097984314, "learning_rate": 5.7251258596655155e-05, "loss": 0.0373, "step": 48100 }, { "grad_norm": 0.3910451829433441, "learning_rate": 5.72348983076778e-05, "loss": 0.041, "step": 48110 }, { "grad_norm": 0.721116840839386, "learning_rate": 5.7218537227502854e-05, "loss": 0.0414, "step": 48120 }, { "grad_norm": 0.34590965509414673, "learning_rate": 5.720217535791951e-05, "loss": 0.0292, "step": 48130 }, { "grad_norm": 0.4032614827156067, "learning_rate": 5.718581270071711e-05, "loss": 0.0322, "step": 48140 }, { "grad_norm": 0.3715131878852844, "learning_rate": 5.716944925768505e-05, "loss": 0.038, "step": 48150 }, { "grad_norm": 0.35638752579689026, "learning_rate": 5.7153085030612786e-05, "loss": 0.0393, "step": 48160 }, { "grad_norm": 1.335254430770874, "learning_rate": 5.713672002128989e-05, "loss": 0.0391, "step": 48170 }, { "grad_norm": 0.3587425649166107, "learning_rate": 5.712035423150602e-05, "loss": 0.0284, "step": 48180 }, { "grad_norm": 0.3364633321762085, "learning_rate": 5.710398766305094e-05, "loss": 0.0343, "step": 48190 }, { "grad_norm": 0.3027963936328888, "learning_rate": 5.708762031771442e-05, "loss": 0.0381, "step": 48200 }, { "grad_norm": 0.7435903549194336, "learning_rate": 5.707125219728642e-05, "loss": 0.0408, "step": 48210 }, { "grad_norm": 0.3757971525192261, "learning_rate": 5.7054883303556905e-05, "loss": 0.0345, "step": 48220 }, { "grad_norm": 0.5088815689086914, "learning_rate": 5.703851363831595e-05, "loss": 0.039, "step": 48230 }, { "grad_norm": 0.6237242817878723, "learning_rate": 5.702214320335374e-05, "loss": 0.0507, "step": 48240 }, { "grad_norm": 0.2961750030517578, "learning_rate": 5.7005772000460514e-05, "loss": 0.0414, "step": 48250 }, { "grad_norm": 0.3741031289100647, "learning_rate": 5.698940003142656e-05, "loss": 0.034, "step": 48260 }, { "grad_norm": 0.360931396484375, "learning_rate": 5.697302729804236e-05, "loss": 0.0355, "step": 48270 }, { "grad_norm": 0.4666386842727661, "learning_rate": 5.695665380209837e-05, "loss": 0.0378, "step": 48280 }, { "grad_norm": 0.35742446780204773, "learning_rate": 5.6940279545385165e-05, "loss": 0.0319, "step": 48290 }, { "grad_norm": 0.3544636368751526, "learning_rate": 5.692390452969344e-05, "loss": 0.0348, "step": 48300 }, { "grad_norm": 0.3599748909473419, "learning_rate": 5.690752875681392e-05, "loss": 0.0375, "step": 48310 }, { "grad_norm": 0.5015711188316345, "learning_rate": 5.6891152228537435e-05, "loss": 0.0464, "step": 48320 }, { "grad_norm": 0.6044873595237732, "learning_rate": 5.687477494665492e-05, "loss": 0.0319, "step": 48330 }, { "grad_norm": 1.000949501991272, "learning_rate": 5.685839691295733e-05, "loss": 0.037, "step": 48340 }, { "grad_norm": 0.40397003293037415, "learning_rate": 5.6842018129235786e-05, "loss": 0.0412, "step": 48350 }, { "grad_norm": 0.5010393857955933, "learning_rate": 5.6825638597281404e-05, "loss": 0.0375, "step": 48360 }, { "grad_norm": 0.5586630702018738, "learning_rate": 5.680925831888546e-05, "loss": 0.0419, "step": 48370 }, { "grad_norm": 0.3473643958568573, "learning_rate": 5.6792877295839274e-05, "loss": 0.0448, "step": 48380 }, { "grad_norm": 0.3088631331920624, "learning_rate": 5.6776495529934224e-05, "loss": 0.0353, "step": 48390 }, { "grad_norm": 0.5454824566841125, "learning_rate": 5.6760113022961824e-05, "loss": 0.0315, "step": 48400 }, { "grad_norm": 0.37960684299468994, "learning_rate": 5.6743729776713617e-05, "loss": 0.0329, "step": 48410 }, { "grad_norm": 0.670695424079895, "learning_rate": 5.672734579298126e-05, "loss": 0.0314, "step": 48420 }, { "grad_norm": 0.6211812496185303, "learning_rate": 5.671096107355649e-05, "loss": 0.0342, "step": 48430 }, { "grad_norm": 0.3727768063545227, "learning_rate": 5.66945756202311e-05, "loss": 0.0468, "step": 48440 }, { "grad_norm": 0.4681057631969452, "learning_rate": 5.667818943479699e-05, "loss": 0.0334, "step": 48450 }, { "grad_norm": 0.3481092154979706, "learning_rate": 5.666180251904612e-05, "loss": 0.0501, "step": 48460 }, { "grad_norm": 0.3965296149253845, "learning_rate": 5.6645414874770555e-05, "loss": 0.0382, "step": 48470 }, { "grad_norm": 0.4931899607181549, "learning_rate": 5.66290265037624e-05, "loss": 0.0517, "step": 48480 }, { "grad_norm": 0.4912665784358978, "learning_rate": 5.661263740781386e-05, "loss": 0.0369, "step": 48490 }, { "grad_norm": 0.3296135365962982, "learning_rate": 5.6596247588717254e-05, "loss": 0.0265, "step": 48500 }, { "grad_norm": 0.46416860818862915, "learning_rate": 5.6579857048264926e-05, "loss": 0.0395, "step": 48510 }, { "grad_norm": 0.47016167640686035, "learning_rate": 5.6563465788249314e-05, "loss": 0.0413, "step": 48520 }, { "grad_norm": 0.47857534885406494, "learning_rate": 5.6547073810462956e-05, "loss": 0.0316, "step": 48530 }, { "grad_norm": 0.6852777004241943, "learning_rate": 5.653068111669846e-05, "loss": 0.0273, "step": 48540 }, { "grad_norm": 0.5131657719612122, "learning_rate": 5.651428770874848e-05, "loss": 0.0406, "step": 48550 }, { "grad_norm": 0.2401735633611679, "learning_rate": 5.64978935884058e-05, "loss": 0.0388, "step": 48560 }, { "grad_norm": 0.5376062393188477, "learning_rate": 5.6481498757463244e-05, "loss": 0.0462, "step": 48570 }, { "grad_norm": 0.5060554146766663, "learning_rate": 5.646510321771373e-05, "loss": 0.0411, "step": 48580 }, { "grad_norm": 0.5122165679931641, "learning_rate": 5.644870697095024e-05, "loss": 0.0384, "step": 48590 }, { "grad_norm": 0.4253578782081604, "learning_rate": 5.643231001896586e-05, "loss": 0.0351, "step": 48600 }, { "grad_norm": 0.474148690700531, "learning_rate": 5.6415912363553726e-05, "loss": 0.0669, "step": 48610 }, { "grad_norm": 0.49122098088264465, "learning_rate": 5.639951400650706e-05, "loss": 0.0378, "step": 48620 }, { "grad_norm": 0.36226198077201843, "learning_rate": 5.6383114949619165e-05, "loss": 0.0382, "step": 48630 }, { "grad_norm": 0.2564096450805664, "learning_rate": 5.636671519468342e-05, "loss": 0.0331, "step": 48640 }, { "grad_norm": 0.5756986737251282, "learning_rate": 5.635031474349327e-05, "loss": 0.0384, "step": 48650 }, { "grad_norm": 0.23346222937107086, "learning_rate": 5.6333913597842246e-05, "loss": 0.0357, "step": 48660 }, { "grad_norm": 1.4102643728256226, "learning_rate": 5.6317511759523955e-05, "loss": 0.0362, "step": 48670 }, { "grad_norm": 0.4294295012950897, "learning_rate": 5.630110923033207e-05, "loss": 0.0378, "step": 48680 }, { "grad_norm": 3.972734212875366, "learning_rate": 5.628470601206036e-05, "loss": 0.0405, "step": 48690 }, { "grad_norm": 1.0462812185287476, "learning_rate": 5.626830210650266e-05, "loss": 0.0348, "step": 48700 }, { "grad_norm": 0.36606138944625854, "learning_rate": 5.625189751545285e-05, "loss": 0.0421, "step": 48710 }, { "grad_norm": 0.4476599395275116, "learning_rate": 5.6235492240704936e-05, "loss": 0.0367, "step": 48720 }, { "grad_norm": 0.6687281131744385, "learning_rate": 5.621908628405296e-05, "loss": 0.0374, "step": 48730 }, { "grad_norm": 0.44308167695999146, "learning_rate": 5.620267964729106e-05, "loss": 0.0395, "step": 48740 }, { "grad_norm": 0.5548567771911621, "learning_rate": 5.618627233221344e-05, "loss": 0.026, "step": 48750 }, { "grad_norm": 0.44673848152160645, "learning_rate": 5.616986434061438e-05, "loss": 0.0396, "step": 48760 }, { "grad_norm": 2.541524887084961, "learning_rate": 5.615345567428822e-05, "loss": 0.0364, "step": 48770 }, { "grad_norm": 0.3705888092517853, "learning_rate": 5.613704633502941e-05, "loss": 0.038, "step": 48780 }, { "grad_norm": 0.42006176710128784, "learning_rate": 5.612063632463245e-05, "loss": 0.0283, "step": 48790 }, { "grad_norm": 0.744544506072998, "learning_rate": 5.610422564489188e-05, "loss": 0.0386, "step": 48800 }, { "grad_norm": 0.4958105981349945, "learning_rate": 5.608781429760239e-05, "loss": 0.0351, "step": 48810 }, { "grad_norm": 0.44973060488700867, "learning_rate": 5.607140228455866e-05, "loss": 0.0409, "step": 48820 }, { "grad_norm": 0.3402666449546814, "learning_rate": 5.605498960755553e-05, "loss": 0.0351, "step": 48830 }, { "grad_norm": 0.5180098414421082, "learning_rate": 5.603857626838782e-05, "loss": 0.0408, "step": 48840 }, { "grad_norm": 0.3517896831035614, "learning_rate": 5.60221622688505e-05, "loss": 0.0338, "step": 48850 }, { "grad_norm": 0.5613322257995605, "learning_rate": 5.6005747610738565e-05, "loss": 0.0408, "step": 48860 }, { "grad_norm": 0.3509279191493988, "learning_rate": 5.59893322958471e-05, "loss": 0.0377, "step": 48870 }, { "grad_norm": 0.4467126727104187, "learning_rate": 5.5972916325971256e-05, "loss": 0.0341, "step": 48880 }, { "grad_norm": 0.39659199118614197, "learning_rate": 5.595649970290628e-05, "loss": 0.0417, "step": 48890 }, { "grad_norm": 0.3081683814525604, "learning_rate": 5.5940082428447426e-05, "loss": 0.0371, "step": 48900 }, { "grad_norm": 0.3459798991680145, "learning_rate": 5.592366450439012e-05, "loss": 0.0443, "step": 48910 }, { "grad_norm": 0.6984117031097412, "learning_rate": 5.590724593252975e-05, "loss": 0.0349, "step": 48920 }, { "grad_norm": 0.3432863652706146, "learning_rate": 5.589082671466184e-05, "loss": 0.0305, "step": 48930 }, { "grad_norm": 0.4118565022945404, "learning_rate": 5.587440685258199e-05, "loss": 0.0369, "step": 48940 }, { "grad_norm": 0.37556377053260803, "learning_rate": 5.585798634808583e-05, "loss": 0.0373, "step": 48950 }, { "grad_norm": 0.6569643020629883, "learning_rate": 5.584156520296909e-05, "loss": 0.0389, "step": 48960 }, { "grad_norm": 0.5868387222290039, "learning_rate": 5.582514341902757e-05, "loss": 0.0358, "step": 48970 }, { "grad_norm": 0.5184380412101746, "learning_rate": 5.580872099805713e-05, "loss": 0.0431, "step": 48980 }, { "grad_norm": 0.35840749740600586, "learning_rate": 5.5792297941853674e-05, "loss": 0.0351, "step": 48990 }, { "grad_norm": 0.39182549715042114, "learning_rate": 5.5775874252213247e-05, "loss": 0.0411, "step": 49000 }, { "grad_norm": 0.8200304508209229, "learning_rate": 5.575944993093189e-05, "loss": 0.0449, "step": 49010 }, { "grad_norm": 0.44425857067108154, "learning_rate": 5.574302497980574e-05, "loss": 0.0278, "step": 49020 }, { "grad_norm": 0.616081953048706, "learning_rate": 5.5726599400631005e-05, "loss": 0.0374, "step": 49030 }, { "grad_norm": 0.32413724064826965, "learning_rate": 5.571017319520399e-05, "loss": 0.036, "step": 49040 }, { "grad_norm": 0.4956516623497009, "learning_rate": 5.5693746365320984e-05, "loss": 0.0339, "step": 49050 }, { "grad_norm": 0.2761881947517395, "learning_rate": 5.567731891277848e-05, "loss": 0.0387, "step": 49060 }, { "grad_norm": 0.4690741002559662, "learning_rate": 5.566089083937288e-05, "loss": 0.0345, "step": 49070 }, { "grad_norm": 0.448309063911438, "learning_rate": 5.564446214690079e-05, "loss": 0.0358, "step": 49080 }, { "grad_norm": 0.3706420660018921, "learning_rate": 5.56280328371588e-05, "loss": 0.0382, "step": 49090 }, { "grad_norm": 0.444076806306839, "learning_rate": 5.5611602911943596e-05, "loss": 0.0402, "step": 49100 }, { "grad_norm": 0.292833149433136, "learning_rate": 5.559517237305194e-05, "loss": 0.041, "step": 49110 }, { "grad_norm": 0.42667385935783386, "learning_rate": 5.557874122228064e-05, "loss": 0.0482, "step": 49120 }, { "grad_norm": 0.7492549419403076, "learning_rate": 5.55623094614266e-05, "loss": 0.0413, "step": 49130 }, { "grad_norm": 0.3476462960243225, "learning_rate": 5.5545877092286743e-05, "loss": 0.0323, "step": 49140 }, { "grad_norm": 0.5625907182693481, "learning_rate": 5.552944411665811e-05, "loss": 0.0352, "step": 49150 }, { "grad_norm": 0.45231515169143677, "learning_rate": 5.5513010536337795e-05, "loss": 0.0385, "step": 49160 }, { "grad_norm": 0.5030367970466614, "learning_rate": 5.5496576353122934e-05, "loss": 0.0405, "step": 49170 }, { "grad_norm": 0.32985860109329224, "learning_rate": 5.548014156881074e-05, "loss": 0.0375, "step": 49180 }, { "grad_norm": 0.5743913054466248, "learning_rate": 5.5463706185198494e-05, "loss": 0.0308, "step": 49190 }, { "grad_norm": 0.39567023515701294, "learning_rate": 5.544727020408358e-05, "loss": 0.0286, "step": 49200 }, { "grad_norm": 0.2956555187702179, "learning_rate": 5.5430833627263366e-05, "loss": 0.0334, "step": 49210 }, { "grad_norm": 0.5725975036621094, "learning_rate": 5.5414396456535354e-05, "loss": 0.0459, "step": 49220 }, { "grad_norm": 0.44554540514945984, "learning_rate": 5.539795869369709e-05, "loss": 0.0345, "step": 49230 }, { "grad_norm": 0.4284367859363556, "learning_rate": 5.538152034054618e-05, "loss": 0.0385, "step": 49240 }, { "grad_norm": 0.3727809488773346, "learning_rate": 5.536508139888028e-05, "loss": 0.0419, "step": 49250 }, { "grad_norm": 0.290511816740036, "learning_rate": 5.534864187049716e-05, "loss": 0.0331, "step": 49260 }, { "grad_norm": 0.26198163628578186, "learning_rate": 5.5332201757194615e-05, "loss": 0.0345, "step": 49270 }, { "grad_norm": 0.3100763261318207, "learning_rate": 5.531576106077049e-05, "loss": 0.0293, "step": 49280 }, { "grad_norm": 0.517230749130249, "learning_rate": 5.529931978302272e-05, "loss": 0.0342, "step": 49290 }, { "grad_norm": 0.2778962254524231, "learning_rate": 5.528287792574932e-05, "loss": 0.0271, "step": 49300 }, { "grad_norm": 0.42006734013557434, "learning_rate": 5.5266435490748324e-05, "loss": 0.031, "step": 49310 }, { "grad_norm": 1.1262154579162598, "learning_rate": 5.524999247981787e-05, "loss": 0.0397, "step": 49320 }, { "grad_norm": 0.4983370900154114, "learning_rate": 5.523354889475613e-05, "loss": 0.0317, "step": 49330 }, { "grad_norm": 0.32001665234565735, "learning_rate": 5.521710473736134e-05, "loss": 0.0363, "step": 49340 }, { "grad_norm": 0.5548318028450012, "learning_rate": 5.5200660009431826e-05, "loss": 0.0346, "step": 49350 }, { "grad_norm": 0.2541428804397583, "learning_rate": 5.518421471276596e-05, "loss": 0.036, "step": 49360 }, { "grad_norm": 0.25915852189064026, "learning_rate": 5.5167768849162174e-05, "loss": 0.0323, "step": 49370 }, { "grad_norm": 0.3380545377731323, "learning_rate": 5.515132242041893e-05, "loss": 0.0377, "step": 49380 }, { "grad_norm": 0.3741616904735565, "learning_rate": 5.513487542833483e-05, "loss": 0.0364, "step": 49390 }, { "grad_norm": 0.4204990267753601, "learning_rate": 5.5118427874708466e-05, "loss": 0.0469, "step": 49400 }, { "grad_norm": 0.4915277361869812, "learning_rate": 5.510197976133853e-05, "loss": 0.0326, "step": 49410 }, { "grad_norm": 0.4209434390068054, "learning_rate": 5.508553109002376e-05, "loss": 0.0299, "step": 49420 }, { "grad_norm": 0.5173854827880859, "learning_rate": 5.5069081862562957e-05, "loss": 0.0318, "step": 49430 }, { "grad_norm": 0.6833212971687317, "learning_rate": 5.5052632080754965e-05, "loss": 0.0324, "step": 49440 }, { "grad_norm": 0.2984492778778076, "learning_rate": 5.503618174639874e-05, "loss": 0.0348, "step": 49450 }, { "grad_norm": 0.48300033807754517, "learning_rate": 5.501973086129325e-05, "loss": 0.0354, "step": 49460 }, { "grad_norm": 0.7068521976470947, "learning_rate": 5.500327942723753e-05, "loss": 0.0356, "step": 49470 }, { "grad_norm": 0.5228576064109802, "learning_rate": 5.498682744603071e-05, "loss": 0.0315, "step": 49480 }, { "grad_norm": 0.37268683314323425, "learning_rate": 5.497037491947194e-05, "loss": 0.0336, "step": 49490 }, { "grad_norm": 0.3202272951602936, "learning_rate": 5.4953921849360424e-05, "loss": 0.0588, "step": 49500 }, { "grad_norm": 0.4388934373855591, "learning_rate": 5.493746823749547e-05, "loss": 0.0471, "step": 49510 }, { "grad_norm": 0.40154606103897095, "learning_rate": 5.49210140856764e-05, "loss": 0.0433, "step": 49520 }, { "grad_norm": 0.4503931701183319, "learning_rate": 5.4904559395702647e-05, "loss": 0.0362, "step": 49530 }, { "grad_norm": 0.46377032995224, "learning_rate": 5.488810416937364e-05, "loss": 0.0343, "step": 49540 }, { "grad_norm": 0.395691454410553, "learning_rate": 5.4871648408488926e-05, "loss": 0.0385, "step": 49550 }, { "grad_norm": 0.5234808921813965, "learning_rate": 5.485519211484807e-05, "loss": 0.0327, "step": 49560 }, { "grad_norm": 0.425454705953598, "learning_rate": 5.4838735290250676e-05, "loss": 0.0376, "step": 49570 }, { "grad_norm": 0.4170372486114502, "learning_rate": 5.482227793649649e-05, "loss": 0.0365, "step": 49580 }, { "grad_norm": 0.34641095995903015, "learning_rate": 5.480582005538524e-05, "loss": 0.0359, "step": 49590 }, { "grad_norm": 0.4814254343509674, "learning_rate": 5.478936164871671e-05, "loss": 0.0456, "step": 49600 }, { "grad_norm": 0.688025951385498, "learning_rate": 5.477290271829081e-05, "loss": 0.0429, "step": 49610 }, { "grad_norm": 0.3163251280784607, "learning_rate": 5.475644326590744e-05, "loss": 0.03, "step": 49620 }, { "grad_norm": 1.052183985710144, "learning_rate": 5.473998329336658e-05, "loss": 0.0436, "step": 49630 }, { "grad_norm": 0.6493785381317139, "learning_rate": 5.4723522802468286e-05, "loss": 0.0424, "step": 49640 }, { "grad_norm": 0.7451938390731812, "learning_rate": 5.470706179501264e-05, "loss": 0.043, "step": 49650 }, { "grad_norm": 0.29838937520980835, "learning_rate": 5.4690600272799776e-05, "loss": 0.0348, "step": 49660 }, { "grad_norm": 0.7581887245178223, "learning_rate": 5.467413823762993e-05, "loss": 0.0365, "step": 49670 }, { "grad_norm": 0.46814966201782227, "learning_rate": 5.465767569130335e-05, "loss": 0.0444, "step": 49680 }, { "grad_norm": 0.46951666474342346, "learning_rate": 5.464121263562036e-05, "loss": 0.0345, "step": 49690 }, { "grad_norm": 0.5753358602523804, "learning_rate": 5.4624749072381343e-05, "loss": 0.0392, "step": 49700 }, { "grad_norm": 0.4622049927711487, "learning_rate": 5.460828500338672e-05, "loss": 0.0365, "step": 49710 }, { "grad_norm": 2.0327279567718506, "learning_rate": 5.459182043043698e-05, "loss": 0.0464, "step": 49720 }, { "grad_norm": 0.4776695966720581, "learning_rate": 5.457535535533265e-05, "loss": 0.0388, "step": 49730 }, { "grad_norm": 0.3428497612476349, "learning_rate": 5.4558889779874334e-05, "loss": 0.0312, "step": 49740 }, { "grad_norm": 0.63494873046875, "learning_rate": 5.454242370586269e-05, "loss": 0.0417, "step": 49750 }, { "grad_norm": 0.48136305809020996, "learning_rate": 5.452595713509843e-05, "loss": 0.0406, "step": 49760 }, { "grad_norm": 0.3787743151187897, "learning_rate": 5.4509490069382274e-05, "loss": 0.0304, "step": 49770 }, { "grad_norm": 0.41985848546028137, "learning_rate": 5.449302251051509e-05, "loss": 0.0282, "step": 49780 }, { "grad_norm": 0.2954627275466919, "learning_rate": 5.44765544602977e-05, "loss": 0.0374, "step": 49790 }, { "grad_norm": 0.4535033404827118, "learning_rate": 5.446008592053107e-05, "loss": 0.0478, "step": 49800 }, { "grad_norm": 0.4961111545562744, "learning_rate": 5.4443616893016135e-05, "loss": 0.0359, "step": 49810 }, { "grad_norm": 0.34363076090812683, "learning_rate": 5.4427147379553934e-05, "loss": 0.0376, "step": 49820 }, { "grad_norm": 0.3509458303451538, "learning_rate": 5.441067738194556e-05, "loss": 0.0348, "step": 49830 }, { "grad_norm": 0.8922227025032043, "learning_rate": 5.439420690199214e-05, "loss": 0.0484, "step": 49840 }, { "grad_norm": 0.33179450035095215, "learning_rate": 5.437773594149484e-05, "loss": 0.0358, "step": 49850 }, { "grad_norm": 0.4592290222644806, "learning_rate": 5.436126450225495e-05, "loss": 0.0361, "step": 49860 }, { "grad_norm": 0.4414117634296417, "learning_rate": 5.434479258607371e-05, "loss": 0.0344, "step": 49870 }, { "grad_norm": 0.3249308168888092, "learning_rate": 5.43283201947525e-05, "loss": 0.0366, "step": 49880 }, { "grad_norm": 0.2460641860961914, "learning_rate": 5.431184733009268e-05, "loss": 0.0483, "step": 49890 }, { "grad_norm": 0.25005602836608887, "learning_rate": 5.4295373993895736e-05, "loss": 0.0361, "step": 49900 }, { "grad_norm": 0.6285763382911682, "learning_rate": 5.4278900187963157e-05, "loss": 0.0348, "step": 49910 }, { "grad_norm": 0.47507867217063904, "learning_rate": 5.426242591409646e-05, "loss": 0.04, "step": 49920 }, { "grad_norm": 0.38039979338645935, "learning_rate": 5.42459511740973e-05, "loss": 0.0283, "step": 49930 }, { "grad_norm": 0.4490739107131958, "learning_rate": 5.422947596976729e-05, "loss": 0.0337, "step": 49940 }, { "grad_norm": 0.4292283058166504, "learning_rate": 5.4213000302908134e-05, "loss": 0.0338, "step": 49950 }, { "grad_norm": 0.5529274940490723, "learning_rate": 5.419652417532162e-05, "loss": 0.0329, "step": 49960 }, { "grad_norm": 0.2780550420284271, "learning_rate": 5.4180047588809534e-05, "loss": 0.0288, "step": 49970 }, { "grad_norm": 0.5870646238327026, "learning_rate": 5.4163570545173704e-05, "loss": 0.0413, "step": 49980 }, { "grad_norm": 0.6127461791038513, "learning_rate": 5.414709304621608e-05, "loss": 0.0301, "step": 49990 }, { "grad_norm": 0.6707499027252197, "learning_rate": 5.413061509373858e-05, "loss": 0.0363, "step": 50000 }, { "grad_norm": 0.5106968879699707, "learning_rate": 5.411413668954324e-05, "loss": 0.0335, "step": 50010 }, { "grad_norm": 0.43481600284576416, "learning_rate": 5.4097657835432083e-05, "loss": 0.0346, "step": 50020 }, { "grad_norm": 0.5202882289886475, "learning_rate": 5.408117853320723e-05, "loss": 0.0285, "step": 50030 }, { "grad_norm": 0.3306635916233063, "learning_rate": 5.406469878467084e-05, "loss": 0.0392, "step": 50040 }, { "grad_norm": 0.4445941150188446, "learning_rate": 5.404821859162509e-05, "loss": 0.0445, "step": 50050 }, { "grad_norm": 0.361871600151062, "learning_rate": 5.403173795587225e-05, "loss": 0.0377, "step": 50060 }, { "grad_norm": 0.6716408729553223, "learning_rate": 5.4015256879214606e-05, "loss": 0.036, "step": 50070 }, { "grad_norm": 0.5276238918304443, "learning_rate": 5.39987753634545e-05, "loss": 0.0335, "step": 50080 }, { "grad_norm": 0.34134534001350403, "learning_rate": 5.3982293410394335e-05, "loss": 0.0241, "step": 50090 }, { "grad_norm": 0.3373989164829254, "learning_rate": 5.396581102183655e-05, "loss": 0.0367, "step": 50100 }, { "grad_norm": 0.40498989820480347, "learning_rate": 5.3949328199583634e-05, "loss": 0.0354, "step": 50110 }, { "grad_norm": 0.5354496240615845, "learning_rate": 5.393284494543813e-05, "loss": 0.0325, "step": 50120 }, { "grad_norm": 0.3412296175956726, "learning_rate": 5.391636126120262e-05, "loss": 0.0349, "step": 50130 }, { "grad_norm": 0.6194444298744202, "learning_rate": 5.389987714867971e-05, "loss": 0.0308, "step": 50140 }, { "grad_norm": 0.42223572731018066, "learning_rate": 5.3883392609672114e-05, "loss": 0.0335, "step": 50150 }, { "grad_norm": 0.511719286441803, "learning_rate": 5.3866907645982545e-05, "loss": 0.036, "step": 50160 }, { "grad_norm": 0.3622961938381195, "learning_rate": 5.385042225941375e-05, "loss": 0.035, "step": 50170 }, { "grad_norm": 0.5685595273971558, "learning_rate": 5.3833936451768585e-05, "loss": 0.0333, "step": 50180 }, { "grad_norm": 0.43857884407043457, "learning_rate": 5.3817450224849884e-05, "loss": 0.0358, "step": 50190 }, { "grad_norm": 0.511988639831543, "learning_rate": 5.380096358046056e-05, "loss": 0.0494, "step": 50200 }, { "grad_norm": 0.5605129599571228, "learning_rate": 5.378447652040359e-05, "loss": 0.0329, "step": 50210 }, { "grad_norm": 0.702711284160614, "learning_rate": 5.376798904648195e-05, "loss": 0.0411, "step": 50220 }, { "grad_norm": 0.8820772171020508, "learning_rate": 5.375150116049869e-05, "loss": 0.0356, "step": 50230 }, { "grad_norm": 0.5922644138336182, "learning_rate": 5.373501286425691e-05, "loss": 0.031, "step": 50240 }, { "grad_norm": 0.3777352273464203, "learning_rate": 5.3718524159559726e-05, "loss": 0.031, "step": 50250 }, { "grad_norm": 0.7038941383361816, "learning_rate": 5.370203504821034e-05, "loss": 0.0352, "step": 50260 }, { "grad_norm": 0.2894514203071594, "learning_rate": 5.368554553201196e-05, "loss": 0.0421, "step": 50270 }, { "grad_norm": 0.4062845706939697, "learning_rate": 5.3669055612767874e-05, "loss": 0.0485, "step": 50280 }, { "grad_norm": 0.34032782912254333, "learning_rate": 5.3652565292281376e-05, "loss": 0.0296, "step": 50290 }, { "grad_norm": 0.49586015939712524, "learning_rate": 5.3636074572355824e-05, "loss": 0.0349, "step": 50300 }, { "grad_norm": 0.3374764025211334, "learning_rate": 5.361958345479463e-05, "loss": 0.0324, "step": 50310 }, { "grad_norm": 0.486950546503067, "learning_rate": 5.360309194140123e-05, "loss": 0.0345, "step": 50320 }, { "grad_norm": 0.4668383300304413, "learning_rate": 5.3586600033979106e-05, "loss": 0.0335, "step": 50330 }, { "grad_norm": 0.41377413272857666, "learning_rate": 5.357010773433181e-05, "loss": 0.0288, "step": 50340 }, { "grad_norm": 0.3486332893371582, "learning_rate": 5.3553615044262895e-05, "loss": 0.0299, "step": 50350 }, { "grad_norm": 0.33548504114151, "learning_rate": 5.3537121965575964e-05, "loss": 0.033, "step": 50360 }, { "grad_norm": 0.45162445306777954, "learning_rate": 5.352062850007471e-05, "loss": 0.0374, "step": 50370 }, { "grad_norm": 0.5158198475837708, "learning_rate": 5.350413464956282e-05, "loss": 0.0305, "step": 50380 }, { "grad_norm": 0.4565374255180359, "learning_rate": 5.348764041584403e-05, "loss": 0.0362, "step": 50390 }, { "grad_norm": 0.44247159361839294, "learning_rate": 5.347114580072212e-05, "loss": 0.0354, "step": 50400 }, { "grad_norm": 0.45900192856788635, "learning_rate": 5.345465080600093e-05, "loss": 0.0319, "step": 50410 }, { "grad_norm": 0.3754757046699524, "learning_rate": 5.3438155433484326e-05, "loss": 0.0362, "step": 50420 }, { "grad_norm": 0.386280357837677, "learning_rate": 5.3421659684976197e-05, "loss": 0.0286, "step": 50430 }, { "grad_norm": 0.6561687588691711, "learning_rate": 5.340516356228052e-05, "loss": 0.0294, "step": 50440 }, { "grad_norm": 0.7336093783378601, "learning_rate": 5.338866706720128e-05, "loss": 0.0438, "step": 50450 }, { "grad_norm": 0.455841064453125, "learning_rate": 5.337217020154249e-05, "loss": 0.033, "step": 50460 }, { "grad_norm": 0.49840810894966125, "learning_rate": 5.335567296710825e-05, "loss": 0.0402, "step": 50470 }, { "grad_norm": 0.42953628301620483, "learning_rate": 5.333917536570265e-05, "loss": 0.031, "step": 50480 }, { "grad_norm": 0.6984949111938477, "learning_rate": 5.332267739912986e-05, "loss": 0.0347, "step": 50490 }, { "grad_norm": 0.3267395496368408, "learning_rate": 5.330617906919405e-05, "loss": 0.0464, "step": 50500 }, { "grad_norm": 0.4962023198604584, "learning_rate": 5.328968037769949e-05, "loss": 0.0398, "step": 50510 }, { "grad_norm": 0.2317313700914383, "learning_rate": 5.327318132645042e-05, "loss": 0.0396, "step": 50520 }, { "grad_norm": 0.4272361397743225, "learning_rate": 5.325668191725116e-05, "loss": 0.0322, "step": 50530 }, { "grad_norm": 0.3238653838634491, "learning_rate": 5.324018215190606e-05, "loss": 0.0257, "step": 50540 }, { "grad_norm": 0.4656999111175537, "learning_rate": 5.3223682032219515e-05, "loss": 0.0287, "step": 50550 }, { "grad_norm": 0.41136717796325684, "learning_rate": 5.320718155999595e-05, "loss": 0.038, "step": 50560 }, { "grad_norm": 0.4414677321910858, "learning_rate": 5.3190680737039835e-05, "loss": 0.0459, "step": 50570 }, { "grad_norm": 0.4144201874732971, "learning_rate": 5.317417956515567e-05, "loss": 0.0351, "step": 50580 }, { "grad_norm": 0.5151588320732117, "learning_rate": 5.315767804614798e-05, "loss": 0.0355, "step": 50590 }, { "grad_norm": 0.4878838062286377, "learning_rate": 5.3141176181821395e-05, "loss": 0.0372, "step": 50600 }, { "grad_norm": 0.5538903474807739, "learning_rate": 5.31246739739805e-05, "loss": 0.0331, "step": 50610 }, { "grad_norm": 0.5814516544342041, "learning_rate": 5.310817142442994e-05, "loss": 0.0441, "step": 50620 }, { "grad_norm": 0.39474520087242126, "learning_rate": 5.309166853497445e-05, "loss": 0.037, "step": 50630 }, { "grad_norm": 0.36283838748931885, "learning_rate": 5.307516530741873e-05, "loss": 0.0336, "step": 50640 }, { "grad_norm": 0.46624279022216797, "learning_rate": 5.305866174356754e-05, "loss": 0.0397, "step": 50650 }, { "grad_norm": 0.3665950298309326, "learning_rate": 5.304215784522571e-05, "loss": 0.0465, "step": 50660 }, { "grad_norm": 0.4991014003753662, "learning_rate": 5.302565361419808e-05, "loss": 0.0408, "step": 50670 }, { "grad_norm": 0.36799171566963196, "learning_rate": 5.3009149052289507e-05, "loss": 0.0457, "step": 50680 }, { "grad_norm": 0.35452377796173096, "learning_rate": 5.299264416130493e-05, "loss": 0.0405, "step": 50690 }, { "grad_norm": 0.5392374992370605, "learning_rate": 5.297613894304928e-05, "loss": 0.0302, "step": 50700 }, { "grad_norm": 0.353451669216156, "learning_rate": 5.2959633399327534e-05, "loss": 0.0303, "step": 50710 }, { "grad_norm": 0.5528209209442139, "learning_rate": 5.294312753194476e-05, "loss": 0.042, "step": 50720 }, { "grad_norm": 0.4050830900669098, "learning_rate": 5.292662134270596e-05, "loss": 0.0363, "step": 50730 }, { "grad_norm": 0.41280418634414673, "learning_rate": 5.291011483341626e-05, "loss": 0.0399, "step": 50740 }, { "grad_norm": 0.48244208097457886, "learning_rate": 5.2893608005880767e-05, "loss": 0.0418, "step": 50750 }, { "grad_norm": 0.3823542594909668, "learning_rate": 5.287710086190467e-05, "loss": 0.0379, "step": 50760 }, { "grad_norm": 0.5366970896720886, "learning_rate": 5.286059340329314e-05, "loss": 0.036, "step": 50770 }, { "grad_norm": 0.23036052286624908, "learning_rate": 5.284408563185141e-05, "loss": 0.0417, "step": 50780 }, { "grad_norm": 0.3637787699699402, "learning_rate": 5.282757754938476e-05, "loss": 0.0415, "step": 50790 }, { "grad_norm": 0.2893141210079193, "learning_rate": 5.281106915769849e-05, "loss": 0.0334, "step": 50800 }, { "grad_norm": 0.3416343331336975, "learning_rate": 5.2794560458597897e-05, "loss": 0.036, "step": 50810 }, { "grad_norm": 0.38380253314971924, "learning_rate": 5.27780514538884e-05, "loss": 0.0422, "step": 50820 }, { "grad_norm": 0.3601115345954895, "learning_rate": 5.2761542145375365e-05, "loss": 0.0423, "step": 50830 }, { "grad_norm": 0.4250929057598114, "learning_rate": 5.274503253486421e-05, "loss": 0.0358, "step": 50840 }, { "grad_norm": 0.5082903504371643, "learning_rate": 5.272852262416046e-05, "loss": 0.0354, "step": 50850 }, { "grad_norm": 0.5708844065666199, "learning_rate": 5.2712012415069555e-05, "loss": 0.035, "step": 50860 }, { "grad_norm": 0.49069952964782715, "learning_rate": 5.269550190939705e-05, "loss": 0.0393, "step": 50870 }, { "grad_norm": 0.4432521164417267, "learning_rate": 5.267899110894852e-05, "loss": 0.0322, "step": 50880 }, { "grad_norm": 0.40483391284942627, "learning_rate": 5.266248001552955e-05, "loss": 0.0385, "step": 50890 }, { "grad_norm": 0.6120926737785339, "learning_rate": 5.264596863094575e-05, "loss": 0.0336, "step": 50900 }, { "grad_norm": 0.42359521985054016, "learning_rate": 5.262945695700282e-05, "loss": 0.0302, "step": 50910 }, { "grad_norm": 0.4293774366378784, "learning_rate": 5.261294499550643e-05, "loss": 0.0449, "step": 50920 }, { "grad_norm": 0.34304818511009216, "learning_rate": 5.25964327482623e-05, "loss": 0.033, "step": 50930 }, { "grad_norm": 0.4895417392253876, "learning_rate": 5.257992021707617e-05, "loss": 0.0367, "step": 50940 }, { "grad_norm": 0.2822408974170685, "learning_rate": 5.256340740375387e-05, "loss": 0.0438, "step": 50950 }, { "grad_norm": 0.430412232875824, "learning_rate": 5.254689431010117e-05, "loss": 0.0343, "step": 50960 }, { "grad_norm": 0.23707979917526245, "learning_rate": 5.253038093792395e-05, "loss": 0.0323, "step": 50970 }, { "grad_norm": 0.4966181218624115, "learning_rate": 5.251386728902806e-05, "loss": 0.0362, "step": 50980 }, { "grad_norm": 0.37512555718421936, "learning_rate": 5.2497353365219446e-05, "loss": 0.035, "step": 50990 }, { "grad_norm": 0.42192527651786804, "learning_rate": 5.2480839168304e-05, "loss": 0.0334, "step": 51000 }, { "grad_norm": 0.8154318332672119, "learning_rate": 5.246432470008773e-05, "loss": 0.0381, "step": 51010 }, { "grad_norm": 0.3539314568042755, "learning_rate": 5.2447809962376616e-05, "loss": 0.035, "step": 51020 }, { "grad_norm": 0.43287959694862366, "learning_rate": 5.243129495697666e-05, "loss": 0.0325, "step": 51030 }, { "grad_norm": 0.506662130355835, "learning_rate": 5.241477968569397e-05, "loss": 0.0399, "step": 51040 }, { "grad_norm": 0.7024481892585754, "learning_rate": 5.23982641503346e-05, "loss": 0.0299, "step": 51050 }, { "grad_norm": 0.49458861351013184, "learning_rate": 5.238174835270465e-05, "loss": 0.0395, "step": 51060 }, { "grad_norm": 0.30916285514831543, "learning_rate": 5.23652322946103e-05, "loss": 0.0312, "step": 51070 }, { "grad_norm": 0.7792930603027344, "learning_rate": 5.234871597785771e-05, "loss": 0.0324, "step": 51080 }, { "grad_norm": 0.413878858089447, "learning_rate": 5.233219940425306e-05, "loss": 0.035, "step": 51090 }, { "grad_norm": 0.6281466484069824, "learning_rate": 5.2315682575602576e-05, "loss": 0.0554, "step": 51100 }, { "grad_norm": 0.9925255179405212, "learning_rate": 5.229916549371254e-05, "loss": 0.0282, "step": 51110 }, { "grad_norm": 0.5080657005310059, "learning_rate": 5.228264816038924e-05, "loss": 0.0325, "step": 51120 }, { "grad_norm": 0.3448835015296936, "learning_rate": 5.2266130577438935e-05, "loss": 0.0323, "step": 51130 }, { "grad_norm": 0.6637932658195496, "learning_rate": 5.224961274666802e-05, "loss": 0.0471, "step": 51140 }, { "grad_norm": 0.40457016229629517, "learning_rate": 5.223309466988283e-05, "loss": 0.0507, "step": 51150 }, { "grad_norm": 0.368723064661026, "learning_rate": 5.221657634888976e-05, "loss": 0.0419, "step": 51160 }, { "grad_norm": 0.37560364603996277, "learning_rate": 5.2200057785495235e-05, "loss": 0.0456, "step": 51170 }, { "grad_norm": 0.40517017245292664, "learning_rate": 5.2183538981505706e-05, "loss": 0.0365, "step": 51180 }, { "grad_norm": 0.6925089955329895, "learning_rate": 5.216701993872762e-05, "loss": 0.0321, "step": 51190 }, { "grad_norm": 0.2746182978153229, "learning_rate": 5.215050065896752e-05, "loss": 0.0405, "step": 51200 }, { "grad_norm": 0.3685210049152374, "learning_rate": 5.2133981144031875e-05, "loss": 0.0472, "step": 51210 }, { "grad_norm": 0.3667285740375519, "learning_rate": 5.211746139572727e-05, "loss": 0.0346, "step": 51220 }, { "grad_norm": 0.34894710779190063, "learning_rate": 5.210094141586026e-05, "loss": 0.0357, "step": 51230 }, { "grad_norm": 0.38346391916275024, "learning_rate": 5.2084421206237476e-05, "loss": 0.0388, "step": 51240 }, { "grad_norm": 0.4860418140888214, "learning_rate": 5.20679007686655e-05, "loss": 0.0447, "step": 51250 }, { "grad_norm": 0.29986312985420227, "learning_rate": 5.205138010495102e-05, "loss": 0.0311, "step": 51260 }, { "grad_norm": 0.5739316940307617, "learning_rate": 5.20348592169007e-05, "loss": 0.0445, "step": 51270 }, { "grad_norm": 0.9723648428916931, "learning_rate": 5.201833810632123e-05, "loss": 0.0387, "step": 51280 }, { "grad_norm": 0.31643784046173096, "learning_rate": 5.200181677501933e-05, "loss": 0.0349, "step": 51290 }, { "grad_norm": 0.3398360311985016, "learning_rate": 5.1985295224801777e-05, "loss": 0.0283, "step": 51300 }, { "grad_norm": 0.43555113673210144, "learning_rate": 5.196877345747531e-05, "loss": 0.0386, "step": 51310 }, { "grad_norm": 0.4391074478626251, "learning_rate": 5.195225147484674e-05, "loss": 0.0356, "step": 51320 }, { "grad_norm": 0.49951407313346863, "learning_rate": 5.193572927872288e-05, "loss": 0.0382, "step": 51330 }, { "grad_norm": 0.6242334246635437, "learning_rate": 5.191920687091059e-05, "loss": 0.0365, "step": 51340 }, { "grad_norm": 0.3485756516456604, "learning_rate": 5.1902684253216695e-05, "loss": 0.0392, "step": 51350 }, { "grad_norm": 0.391855388879776, "learning_rate": 5.1886161427448134e-05, "loss": 0.0378, "step": 51360 }, { "grad_norm": 0.5249121785163879, "learning_rate": 5.186963839541179e-05, "loss": 0.0285, "step": 51370 }, { "grad_norm": 0.2987469732761383, "learning_rate": 5.185311515891459e-05, "loss": 0.0316, "step": 51380 }, { "grad_norm": 0.42118385434150696, "learning_rate": 5.1836591719763516e-05, "loss": 0.037, "step": 51390 }, { "grad_norm": 0.5669984221458435, "learning_rate": 5.182006807976554e-05, "loss": 0.0313, "step": 51400 }, { "grad_norm": 0.4337039291858673, "learning_rate": 5.180354424072763e-05, "loss": 0.0305, "step": 51410 }, { "grad_norm": 0.43472927808761597, "learning_rate": 5.178702020445686e-05, "loss": 0.0375, "step": 51420 }, { "grad_norm": 0.3400382995605469, "learning_rate": 5.177049597276024e-05, "loss": 0.0322, "step": 51430 }, { "grad_norm": 0.624335765838623, "learning_rate": 5.175397154744482e-05, "loss": 0.032, "step": 51440 }, { "grad_norm": 0.40070873498916626, "learning_rate": 5.1737446930317736e-05, "loss": 0.0298, "step": 51450 }, { "grad_norm": 0.37883320450782776, "learning_rate": 5.172092212318604e-05, "loss": 0.0319, "step": 51460 }, { "grad_norm": 0.315047949552536, "learning_rate": 5.170439712785692e-05, "loss": 0.0361, "step": 51470 }, { "grad_norm": 0.45807892084121704, "learning_rate": 5.1687871946137467e-05, "loss": 0.0346, "step": 51480 }, { "grad_norm": 0.3822750151157379, "learning_rate": 5.16713465798349e-05, "loss": 0.0284, "step": 51490 }, { "grad_norm": 0.47147849202156067, "learning_rate": 5.165482103075637e-05, "loss": 0.0387, "step": 51500 }, { "grad_norm": 0.2824496924877167, "learning_rate": 5.163829530070909e-05, "loss": 0.0335, "step": 51510 }, { "grad_norm": 0.49034589529037476, "learning_rate": 5.162176939150032e-05, "loss": 0.042, "step": 51520 }, { "grad_norm": 0.4339931905269623, "learning_rate": 5.160524330493728e-05, "loss": 0.0423, "step": 51530 }, { "grad_norm": 0.3796665072441101, "learning_rate": 5.158871704282724e-05, "loss": 0.0288, "step": 51540 }, { "grad_norm": 0.3023165762424469, "learning_rate": 5.15721906069775e-05, "loss": 0.0317, "step": 51550 }, { "grad_norm": 0.3491252660751343, "learning_rate": 5.155566399919536e-05, "loss": 0.0287, "step": 51560 }, { "grad_norm": 0.30400770902633667, "learning_rate": 5.153913722128813e-05, "loss": 0.0363, "step": 51570 }, { "grad_norm": 0.3052174746990204, "learning_rate": 5.1522610275063176e-05, "loss": 0.0286, "step": 51580 }, { "grad_norm": 0.5902420282363892, "learning_rate": 5.150608316232786e-05, "loss": 0.0311, "step": 51590 }, { "grad_norm": 0.37968045473098755, "learning_rate": 5.148955588488955e-05, "loss": 0.0303, "step": 51600 }, { "grad_norm": 0.40067800879478455, "learning_rate": 5.147302844455564e-05, "loss": 0.0381, "step": 51610 }, { "grad_norm": 0.3814952075481415, "learning_rate": 5.145650084313356e-05, "loss": 0.035, "step": 51620 }, { "grad_norm": 1.119694471359253, "learning_rate": 5.1439973082430734e-05, "loss": 0.0398, "step": 51630 }, { "grad_norm": 0.31387004256248474, "learning_rate": 5.142344516425461e-05, "loss": 0.0277, "step": 51640 }, { "grad_norm": 0.41297051310539246, "learning_rate": 5.1406917090412673e-05, "loss": 0.041, "step": 51650 }, { "grad_norm": 0.8496735095977783, "learning_rate": 5.1390388862712404e-05, "loss": 0.0424, "step": 51660 }, { "grad_norm": 1.0490691661834717, "learning_rate": 5.137386048296129e-05, "loss": 0.0351, "step": 51670 }, { "grad_norm": 0.36535391211509705, "learning_rate": 5.135733195296686e-05, "loss": 0.0359, "step": 51680 }, { "grad_norm": 0.5236711502075195, "learning_rate": 5.134080327453664e-05, "loss": 0.0343, "step": 51690 }, { "grad_norm": 0.8045138716697693, "learning_rate": 5.1324274449478204e-05, "loss": 0.0319, "step": 51700 }, { "grad_norm": 0.32219189405441284, "learning_rate": 5.130774547959909e-05, "loss": 0.0319, "step": 51710 }, { "grad_norm": 0.32899683713912964, "learning_rate": 5.129121636670692e-05, "loss": 0.032, "step": 51720 }, { "grad_norm": 0.34320130944252014, "learning_rate": 5.1274687112609246e-05, "loss": 0.0347, "step": 51730 }, { "grad_norm": 0.3390030860900879, "learning_rate": 5.125815771911372e-05, "loss": 0.0306, "step": 51740 }, { "grad_norm": 0.3745097517967224, "learning_rate": 5.1241628188027956e-05, "loss": 0.0294, "step": 51750 }, { "grad_norm": 0.2282164990901947, "learning_rate": 5.1225098521159585e-05, "loss": 0.0249, "step": 51760 }, { "grad_norm": 0.45663365721702576, "learning_rate": 5.12085687203163e-05, "loss": 0.0323, "step": 51770 }, { "grad_norm": 1.1310145854949951, "learning_rate": 5.119203878730575e-05, "loss": 0.0356, "step": 51780 }, { "grad_norm": 0.36699914932250977, "learning_rate": 5.117550872393564e-05, "loss": 0.039, "step": 51790 }, { "grad_norm": 0.36948660016059875, "learning_rate": 5.1158978532013645e-05, "loss": 0.0329, "step": 51800 }, { "grad_norm": 0.37895098328590393, "learning_rate": 5.114244821334751e-05, "loss": 0.0318, "step": 51810 }, { "grad_norm": 0.33561620116233826, "learning_rate": 5.112591776974497e-05, "loss": 0.0258, "step": 51820 }, { "grad_norm": 0.31567832827568054, "learning_rate": 5.110938720301374e-05, "loss": 0.0316, "step": 51830 }, { "grad_norm": 0.49515998363494873, "learning_rate": 5.1092856514961595e-05, "loss": 0.0387, "step": 51840 }, { "grad_norm": 0.46297866106033325, "learning_rate": 5.107632570739631e-05, "loss": 0.0363, "step": 51850 }, { "grad_norm": 0.32320311665534973, "learning_rate": 5.1059794782125645e-05, "loss": 0.0326, "step": 51860 }, { "grad_norm": 0.6377895474433899, "learning_rate": 5.104326374095744e-05, "loss": 0.0447, "step": 51870 }, { "grad_norm": 0.3858388364315033, "learning_rate": 5.102673258569947e-05, "loss": 0.0371, "step": 51880 }, { "grad_norm": 0.3933717906475067, "learning_rate": 5.101020131815956e-05, "loss": 0.045, "step": 51890 }, { "grad_norm": 0.42243146896362305, "learning_rate": 5.099366994014556e-05, "loss": 0.0393, "step": 51900 }, { "grad_norm": 0.3735397160053253, "learning_rate": 5.097713845346532e-05, "loss": 0.0452, "step": 51910 }, { "grad_norm": 0.9407588243484497, "learning_rate": 5.096060685992668e-05, "loss": 0.0289, "step": 51920 }, { "grad_norm": 0.4908912777900696, "learning_rate": 5.0944075161337514e-05, "loss": 0.0363, "step": 51930 }, { "grad_norm": 0.26265788078308105, "learning_rate": 5.09275433595057e-05, "loss": 0.0404, "step": 51940 }, { "grad_norm": 0.33545589447021484, "learning_rate": 5.0911011456239157e-05, "loss": 0.0386, "step": 51950 }, { "grad_norm": 0.6467908620834351, "learning_rate": 5.089447945334574e-05, "loss": 0.0389, "step": 51960 }, { "grad_norm": 0.2694142460823059, "learning_rate": 5.087794735263343e-05, "loss": 0.0296, "step": 51970 }, { "grad_norm": 0.2670030891895294, "learning_rate": 5.086141515591011e-05, "loss": 0.0285, "step": 51980 }, { "grad_norm": 0.7715486884117126, "learning_rate": 5.084488286498371e-05, "loss": 0.039, "step": 51990 }, { "grad_norm": 0.38485899567604065, "learning_rate": 5.08283504816622e-05, "loss": 0.0392, "step": 52000 }, { "grad_norm": 0.4974948763847351, "learning_rate": 5.081181800775353e-05, "loss": 0.0417, "step": 52010 }, { "grad_norm": 0.36989736557006836, "learning_rate": 5.079528544506565e-05, "loss": 0.0328, "step": 52020 }, { "grad_norm": 0.25706321001052856, "learning_rate": 5.077875279540657e-05, "loss": 0.0291, "step": 52030 }, { "grad_norm": 0.41899722814559937, "learning_rate": 5.076222006058424e-05, "loss": 0.0349, "step": 52040 }, { "grad_norm": 0.4567413032054901, "learning_rate": 5.074568724240667e-05, "loss": 0.0324, "step": 52050 }, { "grad_norm": 0.27824410796165466, "learning_rate": 5.072915434268187e-05, "loss": 0.0379, "step": 52060 }, { "grad_norm": 0.49742165207862854, "learning_rate": 5.071262136321785e-05, "loss": 0.0409, "step": 52070 }, { "grad_norm": 0.4527653753757477, "learning_rate": 5.069608830582262e-05, "loss": 0.0308, "step": 52080 }, { "grad_norm": 0.40707629919052124, "learning_rate": 5.067955517230423e-05, "loss": 0.0332, "step": 52090 }, { "grad_norm": 0.36964356899261475, "learning_rate": 5.06630219644707e-05, "loss": 0.0375, "step": 52100 }, { "grad_norm": 0.4167741537094116, "learning_rate": 5.064648868413008e-05, "loss": 0.03, "step": 52110 }, { "grad_norm": 0.3140318691730499, "learning_rate": 5.062995533309044e-05, "loss": 0.0283, "step": 52120 }, { "grad_norm": 0.5394006371498108, "learning_rate": 5.061342191315983e-05, "loss": 0.0422, "step": 52130 }, { "grad_norm": 0.34631261229515076, "learning_rate": 5.0596888426146325e-05, "loss": 0.0366, "step": 52140 }, { "grad_norm": 0.400299072265625, "learning_rate": 5.058035487385798e-05, "loss": 0.0337, "step": 52150 }, { "grad_norm": 0.4635723829269409, "learning_rate": 5.0563821258102915e-05, "loss": 0.0315, "step": 52160 }, { "grad_norm": 0.5984185934066772, "learning_rate": 5.05472875806892e-05, "loss": 0.0371, "step": 52170 }, { "grad_norm": 0.3511272966861725, "learning_rate": 5.053075384342494e-05, "loss": 0.0363, "step": 52180 }, { "grad_norm": 0.2948077321052551, "learning_rate": 5.0514220048118234e-05, "loss": 0.034, "step": 52190 }, { "grad_norm": 0.527844250202179, "learning_rate": 5.049768619657721e-05, "loss": 0.0376, "step": 52200 }, { "grad_norm": 0.5760714411735535, "learning_rate": 5.048115229060996e-05, "loss": 0.0281, "step": 52210 }, { "grad_norm": 0.4026179015636444, "learning_rate": 5.046461833202462e-05, "loss": 0.0412, "step": 52220 }, { "grad_norm": 0.5411841869354248, "learning_rate": 5.044808432262932e-05, "loss": 0.051, "step": 52230 }, { "grad_norm": 0.32579442858695984, "learning_rate": 5.043155026423219e-05, "loss": 0.027, "step": 52240 }, { "grad_norm": 0.47387173771858215, "learning_rate": 5.041501615864137e-05, "loss": 0.034, "step": 52250 }, { "grad_norm": 0.3985666036605835, "learning_rate": 5.039848200766502e-05, "loss": 0.0298, "step": 52260 }, { "grad_norm": 0.4667665958404541, "learning_rate": 5.038194781311125e-05, "loss": 0.0287, "step": 52270 }, { "grad_norm": 0.3850798010826111, "learning_rate": 5.0365413576788264e-05, "loss": 0.0251, "step": 52280 }, { "grad_norm": 0.3165077865123749, "learning_rate": 5.0348879300504194e-05, "loss": 0.0341, "step": 52290 }, { "grad_norm": 0.2557438015937805, "learning_rate": 5.033234498606721e-05, "loss": 0.0324, "step": 52300 }, { "grad_norm": 0.6822537779808044, "learning_rate": 5.0315810635285476e-05, "loss": 0.0381, "step": 52310 }, { "grad_norm": 0.3814033567905426, "learning_rate": 5.0299276249967164e-05, "loss": 0.0265, "step": 52320 }, { "grad_norm": 1.559038519859314, "learning_rate": 5.0282741831920454e-05, "loss": 0.0523, "step": 52330 }, { "grad_norm": 0.4850858747959137, "learning_rate": 5.026620738295351e-05, "loss": 0.0364, "step": 52340 }, { "grad_norm": 0.2989536225795746, "learning_rate": 5.024967290487455e-05, "loss": 0.0307, "step": 52350 }, { "grad_norm": 0.3582487404346466, "learning_rate": 5.023313839949172e-05, "loss": 0.0385, "step": 52360 }, { "grad_norm": 0.3412993252277374, "learning_rate": 5.0216603868613224e-05, "loss": 0.0374, "step": 52370 }, { "grad_norm": 0.27697741985321045, "learning_rate": 5.020006931404726e-05, "loss": 0.0376, "step": 52380 }, { "grad_norm": 0.359650582075119, "learning_rate": 5.0183534737602014e-05, "loss": 0.0318, "step": 52390 }, { "grad_norm": 0.44466742873191833, "learning_rate": 5.0167000141085686e-05, "loss": 0.0361, "step": 52400 }, { "grad_norm": 0.424191951751709, "learning_rate": 5.015046552630647e-05, "loss": 0.038, "step": 52410 }, { "grad_norm": 0.4061553478240967, "learning_rate": 5.013393089507258e-05, "loss": 0.0373, "step": 52420 }, { "grad_norm": 0.6956772208213806, "learning_rate": 5.011739624919219e-05, "loss": 0.0368, "step": 52430 }, { "grad_norm": 0.3352659344673157, "learning_rate": 5.0100861590473545e-05, "loss": 0.0262, "step": 52440 }, { "grad_norm": 0.4587835371494293, "learning_rate": 5.008432692072482e-05, "loss": 0.0275, "step": 52450 }, { "grad_norm": 0.49423250555992126, "learning_rate": 5.006779224175423e-05, "loss": 0.0403, "step": 52460 }, { "grad_norm": 0.48506543040275574, "learning_rate": 5.005125755536998e-05, "loss": 0.0315, "step": 52470 }, { "grad_norm": 0.3518174886703491, "learning_rate": 5.003472286338029e-05, "loss": 0.0398, "step": 52480 }, { "grad_norm": 0.8192997574806213, "learning_rate": 5.001818816759335e-05, "loss": 0.0515, "step": 52490 }, { "grad_norm": 0.33301782608032227, "learning_rate": 5.000165346981738e-05, "loss": 0.0396, "step": 52500 }, { "grad_norm": 0.6287874579429626, "learning_rate": 4.9985118771860595e-05, "loss": 0.0384, "step": 52510 }, { "grad_norm": 0.5303353071212769, "learning_rate": 4.996858407553119e-05, "loss": 0.0314, "step": 52520 }, { "grad_norm": 0.37879279255867004, "learning_rate": 4.99520493826374e-05, "loss": 0.0334, "step": 52530 }, { "grad_norm": 0.4587230980396271, "learning_rate": 4.9935514694987384e-05, "loss": 0.0362, "step": 52540 }, { "grad_norm": 0.30664530396461487, "learning_rate": 4.991898001438941e-05, "loss": 0.0292, "step": 52550 }, { "grad_norm": 0.6594879627227783, "learning_rate": 4.990244534265166e-05, "loss": 0.0339, "step": 52560 }, { "grad_norm": 0.37412819266319275, "learning_rate": 4.988591068158234e-05, "loss": 0.0404, "step": 52570 }, { "grad_norm": 0.5206798911094666, "learning_rate": 4.986937603298965e-05, "loss": 0.0379, "step": 52580 }, { "grad_norm": 0.49121174216270447, "learning_rate": 4.985284139868181e-05, "loss": 0.0297, "step": 52590 }, { "grad_norm": 0.41014137864112854, "learning_rate": 4.9836306780467e-05, "loss": 0.0379, "step": 52600 }, { "grad_norm": 0.586728572845459, "learning_rate": 4.981977218015343e-05, "loss": 0.0349, "step": 52610 }, { "grad_norm": 0.45410701632499695, "learning_rate": 4.980323759954932e-05, "loss": 0.0289, "step": 52620 }, { "grad_norm": 0.7658073902130127, "learning_rate": 4.978670304046284e-05, "loss": 0.037, "step": 52630 }, { "grad_norm": 0.34701329469680786, "learning_rate": 4.9770168504702195e-05, "loss": 0.0291, "step": 52640 }, { "grad_norm": 0.6485552191734314, "learning_rate": 4.975363399407559e-05, "loss": 0.0335, "step": 52650 }, { "grad_norm": 0.39670005440711975, "learning_rate": 4.9737099510391194e-05, "loss": 0.0302, "step": 52660 }, { "grad_norm": 0.2538062334060669, "learning_rate": 4.972056505545719e-05, "loss": 0.0395, "step": 52670 }, { "grad_norm": 0.43231889605522156, "learning_rate": 4.9704030631081775e-05, "loss": 0.0276, "step": 52680 }, { "grad_norm": 0.424354612827301, "learning_rate": 4.968749623907313e-05, "loss": 0.0273, "step": 52690 }, { "grad_norm": 0.2546888291835785, "learning_rate": 4.967096188123943e-05, "loss": 0.0349, "step": 52700 }, { "grad_norm": 0.43032410740852356, "learning_rate": 4.965442755938884e-05, "loss": 0.0271, "step": 52710 }, { "grad_norm": 0.8043290376663208, "learning_rate": 4.963789327532954e-05, "loss": 0.0425, "step": 52720 }, { "grad_norm": 0.6021983623504639, "learning_rate": 4.962135903086968e-05, "loss": 0.0398, "step": 52730 }, { "grad_norm": 0.37225106358528137, "learning_rate": 4.960482482781741e-05, "loss": 0.0313, "step": 52740 }, { "grad_norm": 0.44205909967422485, "learning_rate": 4.958829066798092e-05, "loss": 0.0324, "step": 52750 }, { "grad_norm": 0.33211401104927063, "learning_rate": 4.957175655316835e-05, "loss": 0.0286, "step": 52760 }, { "grad_norm": 0.4448060989379883, "learning_rate": 4.9555222485187834e-05, "loss": 0.0356, "step": 52770 }, { "grad_norm": 0.4137182831764221, "learning_rate": 4.953868846584752e-05, "loss": 0.0302, "step": 52780 }, { "grad_norm": 0.32512110471725464, "learning_rate": 4.9522154496955544e-05, "loss": 0.0455, "step": 52790 }, { "grad_norm": 0.472522497177124, "learning_rate": 4.950562058032001e-05, "loss": 0.0345, "step": 52800 }, { "grad_norm": 0.30569276213645935, "learning_rate": 4.948908671774909e-05, "loss": 0.0387, "step": 52810 }, { "grad_norm": 0.31584492325782776, "learning_rate": 4.947255291105087e-05, "loss": 0.0299, "step": 52820 }, { "grad_norm": 0.40101760625839233, "learning_rate": 4.945601916203348e-05, "loss": 0.0337, "step": 52830 }, { "grad_norm": 0.48757603764533997, "learning_rate": 4.9439485472505016e-05, "loss": 0.0376, "step": 52840 }, { "grad_norm": 0.488808274269104, "learning_rate": 4.942295184427356e-05, "loss": 0.0349, "step": 52850 }, { "grad_norm": 0.3317393958568573, "learning_rate": 4.9406418279147234e-05, "loss": 0.0315, "step": 52860 }, { "grad_norm": 0.42293819785118103, "learning_rate": 4.9389884778934116e-05, "loss": 0.0457, "step": 52870 }, { "grad_norm": 0.6593171954154968, "learning_rate": 4.937335134544228e-05, "loss": 0.029, "step": 52880 }, { "grad_norm": 0.33709660172462463, "learning_rate": 4.935681798047979e-05, "loss": 0.0337, "step": 52890 }, { "grad_norm": 0.6290308833122253, "learning_rate": 4.934028468585473e-05, "loss": 0.042, "step": 52900 }, { "grad_norm": 0.2981448769569397, "learning_rate": 4.932375146337514e-05, "loss": 0.0322, "step": 52910 }, { "grad_norm": 0.46463391184806824, "learning_rate": 4.930721831484908e-05, "loss": 0.0271, "step": 52920 }, { "grad_norm": 0.7146937847137451, "learning_rate": 4.929068524208456e-05, "loss": 0.034, "step": 52930 }, { "grad_norm": 0.3518352210521698, "learning_rate": 4.927415224688965e-05, "loss": 0.0447, "step": 52940 }, { "grad_norm": 0.3974674344062805, "learning_rate": 4.9257619331072375e-05, "loss": 0.0524, "step": 52950 }, { "grad_norm": 0.5558618307113647, "learning_rate": 4.924108649644072e-05, "loss": 0.0225, "step": 52960 }, { "grad_norm": 0.3788207471370697, "learning_rate": 4.922455374480271e-05, "loss": 0.0585, "step": 52970 }, { "grad_norm": 0.5666531324386597, "learning_rate": 4.920802107796634e-05, "loss": 0.0375, "step": 52980 }, { "grad_norm": 0.6480003595352173, "learning_rate": 4.9191488497739564e-05, "loss": 0.0323, "step": 52990 }, { "grad_norm": 0.24425344169139862, "learning_rate": 4.917495600593042e-05, "loss": 0.0376, "step": 53000 }, { "grad_norm": 0.3508097529411316, "learning_rate": 4.915842360434684e-05, "loss": 0.0365, "step": 53010 }, { "grad_norm": 0.654055118560791, "learning_rate": 4.914189129479679e-05, "loss": 0.0497, "step": 53020 }, { "grad_norm": 0.31613513827323914, "learning_rate": 4.912535907908822e-05, "loss": 0.0375, "step": 53030 }, { "grad_norm": 0.4134383499622345, "learning_rate": 4.910882695902907e-05, "loss": 0.0418, "step": 53040 }, { "grad_norm": 0.4355514645576477, "learning_rate": 4.9092294936427234e-05, "loss": 0.0394, "step": 53050 }, { "grad_norm": 0.32986241579055786, "learning_rate": 4.9075763013090685e-05, "loss": 0.0279, "step": 53060 }, { "grad_norm": 0.7853244543075562, "learning_rate": 4.9059231190827305e-05, "loss": 0.0419, "step": 53070 }, { "grad_norm": 0.28869014978408813, "learning_rate": 4.904269947144499e-05, "loss": 0.0535, "step": 53080 }, { "grad_norm": 0.39605164527893066, "learning_rate": 4.902616785675161e-05, "loss": 0.0328, "step": 53090 }, { "grad_norm": 0.41309526562690735, "learning_rate": 4.900963634855505e-05, "loss": 0.0338, "step": 53100 }, { "grad_norm": 0.34168317914009094, "learning_rate": 4.8993104948663173e-05, "loss": 0.0347, "step": 53110 }, { "grad_norm": 0.3652855455875397, "learning_rate": 4.8976573658883823e-05, "loss": 0.0306, "step": 53120 }, { "grad_norm": 0.355569452047348, "learning_rate": 4.896004248102484e-05, "loss": 0.0282, "step": 53130 }, { "grad_norm": 0.4684121906757355, "learning_rate": 4.8943511416894045e-05, "loss": 0.0416, "step": 53140 }, { "grad_norm": 0.3212265372276306, "learning_rate": 4.892698046829926e-05, "loss": 0.0409, "step": 53150 }, { "grad_norm": 0.43730202317237854, "learning_rate": 4.8910449637048275e-05, "loss": 0.0458, "step": 53160 }, { "grad_norm": 0.49254685640335083, "learning_rate": 4.8893918924948884e-05, "loss": 0.0474, "step": 53170 }, { "grad_norm": 0.3208085596561432, "learning_rate": 4.8877388333808835e-05, "loss": 0.0386, "step": 53180 }, { "grad_norm": 0.3904944658279419, "learning_rate": 4.8860857865435925e-05, "loss": 0.0382, "step": 53190 }, { "grad_norm": 0.22668315470218658, "learning_rate": 4.884432752163789e-05, "loss": 0.0306, "step": 53200 }, { "grad_norm": 0.42863720655441284, "learning_rate": 4.8827797304222466e-05, "loss": 0.0307, "step": 53210 }, { "grad_norm": 0.43978622555732727, "learning_rate": 4.8811267214997366e-05, "loss": 0.0301, "step": 53220 }, { "grad_norm": 0.7020763158798218, "learning_rate": 4.879473725577029e-05, "loss": 0.0301, "step": 53230 }, { "grad_norm": 0.4760487675666809, "learning_rate": 4.877820742834891e-05, "loss": 0.0274, "step": 53240 }, { "grad_norm": 0.41484534740448, "learning_rate": 4.8761677734540956e-05, "loss": 0.0342, "step": 53250 }, { "grad_norm": 0.4813736379146576, "learning_rate": 4.874514817615407e-05, "loss": 0.0371, "step": 53260 }, { "grad_norm": 0.3356202244758606, "learning_rate": 4.872861875499588e-05, "loss": 0.0307, "step": 53270 }, { "grad_norm": 0.39022186398506165, "learning_rate": 4.871208947287404e-05, "loss": 0.0386, "step": 53280 }, { "grad_norm": 0.6032276153564453, "learning_rate": 4.869556033159615e-05, "loss": 0.037, "step": 53290 }, { "grad_norm": 0.49224212765693665, "learning_rate": 4.8679031332969835e-05, "loss": 0.0334, "step": 53300 }, { "grad_norm": 0.32002362608909607, "learning_rate": 4.8662502478802635e-05, "loss": 0.0438, "step": 53310 }, { "grad_norm": 0.3541264533996582, "learning_rate": 4.8645973770902176e-05, "loss": 0.0426, "step": 53320 }, { "grad_norm": 0.4186575412750244, "learning_rate": 4.862944521107599e-05, "loss": 0.0362, "step": 53330 }, { "grad_norm": 0.5638298392295837, "learning_rate": 4.861291680113162e-05, "loss": 0.0379, "step": 53340 }, { "grad_norm": 0.4817765951156616, "learning_rate": 4.8596388542876563e-05, "loss": 0.0277, "step": 53350 }, { "grad_norm": 0.30865731835365295, "learning_rate": 4.857986043811836e-05, "loss": 0.038, "step": 53360 }, { "grad_norm": 0.46393194794654846, "learning_rate": 4.856333248866448e-05, "loss": 0.0301, "step": 53370 }, { "grad_norm": 0.46120449900627136, "learning_rate": 4.85468046963224e-05, "loss": 0.0273, "step": 53380 }, { "grad_norm": 0.27294081449508667, "learning_rate": 4.853027706289957e-05, "loss": 0.0285, "step": 53390 }, { "grad_norm": 0.5217426419258118, "learning_rate": 4.851374959020344e-05, "loss": 0.032, "step": 53400 }, { "grad_norm": 0.4360075294971466, "learning_rate": 4.849722228004141e-05, "loss": 0.0388, "step": 53410 }, { "grad_norm": 0.7961679697036743, "learning_rate": 4.848069513422091e-05, "loss": 0.0328, "step": 53420 }, { "grad_norm": 0.24524495005607605, "learning_rate": 4.8464168154549294e-05, "loss": 0.0255, "step": 53430 }, { "grad_norm": 0.37859946489334106, "learning_rate": 4.844764134283392e-05, "loss": 0.0405, "step": 53440 }, { "grad_norm": 0.49884843826293945, "learning_rate": 4.843111470088218e-05, "loss": 0.039, "step": 53450 }, { "grad_norm": 0.23543232679367065, "learning_rate": 4.841458823050137e-05, "loss": 0.0345, "step": 53460 }, { "grad_norm": 0.3012489378452301, "learning_rate": 4.8398061933498816e-05, "loss": 0.0377, "step": 53470 }, { "grad_norm": 0.47448521852493286, "learning_rate": 4.83815358116818e-05, "loss": 0.033, "step": 53480 }, { "grad_norm": 0.3418886363506317, "learning_rate": 4.8365009866857584e-05, "loss": 0.0429, "step": 53490 }, { "grad_norm": 0.22219949960708618, "learning_rate": 4.834848410083342e-05, "loss": 0.023, "step": 53500 }, { "grad_norm": 3.6513447761535645, "learning_rate": 4.833195851541657e-05, "loss": 0.0393, "step": 53510 }, { "grad_norm": 0.4828304350376129, "learning_rate": 4.831543311241423e-05, "loss": 0.0344, "step": 53520 }, { "grad_norm": 0.5975543856620789, "learning_rate": 4.829890789363359e-05, "loss": 0.0469, "step": 53530 }, { "grad_norm": 0.6858503818511963, "learning_rate": 4.828238286088182e-05, "loss": 0.0288, "step": 53540 }, { "grad_norm": 0.3662776052951813, "learning_rate": 4.826585801596609e-05, "loss": 0.0334, "step": 53550 }, { "grad_norm": 0.4402564465999603, "learning_rate": 4.8249333360693486e-05, "loss": 0.0344, "step": 53560 }, { "grad_norm": 0.4233815371990204, "learning_rate": 4.8232808896871185e-05, "loss": 0.0353, "step": 53570 }, { "grad_norm": 0.32894131541252136, "learning_rate": 4.821628462630624e-05, "loss": 0.0264, "step": 53580 }, { "grad_norm": 0.48924770951271057, "learning_rate": 4.8199760550805735e-05, "loss": 0.0429, "step": 53590 }, { "grad_norm": 0.4030384123325348, "learning_rate": 4.818323667217669e-05, "loss": 0.0308, "step": 53600 }, { "grad_norm": 0.6860235929489136, "learning_rate": 4.8166712992226165e-05, "loss": 0.0315, "step": 53610 }, { "grad_norm": 0.390125572681427, "learning_rate": 4.8150189512761154e-05, "loss": 0.0327, "step": 53620 }, { "grad_norm": 0.47423040866851807, "learning_rate": 4.813366623558862e-05, "loss": 0.0437, "step": 53630 }, { "grad_norm": 0.3103533685207367, "learning_rate": 4.811714316251554e-05, "loss": 0.0304, "step": 53640 }, { "grad_norm": 0.4243469536304474, "learning_rate": 4.810062029534886e-05, "loss": 0.0365, "step": 53650 }, { "grad_norm": 0.5707099437713623, "learning_rate": 4.8084097635895505e-05, "loss": 0.0391, "step": 53660 }, { "grad_norm": 0.39347249269485474, "learning_rate": 4.8067575185962335e-05, "loss": 0.039, "step": 53670 }, { "grad_norm": 0.47950637340545654, "learning_rate": 4.805105294735625e-05, "loss": 0.0335, "step": 53680 }, { "grad_norm": 0.34900104999542236, "learning_rate": 4.803453092188406e-05, "loss": 0.0372, "step": 53690 }, { "grad_norm": 0.41766661405563354, "learning_rate": 4.801800911135263e-05, "loss": 0.0368, "step": 53700 }, { "grad_norm": 0.38827720284461975, "learning_rate": 4.8001487517568744e-05, "loss": 0.029, "step": 53710 }, { "grad_norm": 0.6153196692466736, "learning_rate": 4.798496614233918e-05, "loss": 0.0368, "step": 53720 }, { "grad_norm": 0.47953808307647705, "learning_rate": 4.796844498747069e-05, "loss": 0.0455, "step": 53730 }, { "grad_norm": 0.4261956214904785, "learning_rate": 4.795192405477e-05, "loss": 0.0422, "step": 53740 }, { "grad_norm": 0.4423324167728424, "learning_rate": 4.79354033460438e-05, "loss": 0.0298, "step": 53750 }, { "grad_norm": 0.3661709427833557, "learning_rate": 4.79188828630988e-05, "loss": 0.0441, "step": 53760 }, { "grad_norm": 0.39772337675094604, "learning_rate": 4.790236260774165e-05, "loss": 0.0309, "step": 53770 }, { "grad_norm": 0.36187446117401123, "learning_rate": 4.788584258177896e-05, "loss": 0.0309, "step": 53780 }, { "grad_norm": 0.6261236071586609, "learning_rate": 4.7869322787017356e-05, "loss": 0.0305, "step": 53790 }, { "grad_norm": 0.5466277599334717, "learning_rate": 4.78528032252634e-05, "loss": 0.0304, "step": 53800 }, { "grad_norm": 0.39388585090637207, "learning_rate": 4.783628389832366e-05, "loss": 0.035, "step": 53810 }, { "grad_norm": 0.4159057140350342, "learning_rate": 4.7819764808004644e-05, "loss": 0.0378, "step": 53820 }, { "grad_norm": 0.5240321159362793, "learning_rate": 4.780324595611289e-05, "loss": 0.0387, "step": 53830 }, { "grad_norm": 0.3798556923866272, "learning_rate": 4.7786727344454854e-05, "loss": 0.039, "step": 53840 }, { "grad_norm": 0.3066514730453491, "learning_rate": 4.777020897483697e-05, "loss": 0.0404, "step": 53850 }, { "grad_norm": 0.38531336188316345, "learning_rate": 4.775369084906569e-05, "loss": 0.0344, "step": 53860 }, { "grad_norm": 0.298631489276886, "learning_rate": 4.773717296894742e-05, "loss": 0.0353, "step": 53870 }, { "grad_norm": 0.4820915460586548, "learning_rate": 4.772065533628848e-05, "loss": 0.0355, "step": 53880 }, { "grad_norm": 0.3139346241950989, "learning_rate": 4.7704137952895265e-05, "loss": 0.0277, "step": 53890 }, { "grad_norm": 0.5205668807029724, "learning_rate": 4.768762082057407e-05, "loss": 0.0355, "step": 53900 }, { "grad_norm": 0.48315227031707764, "learning_rate": 4.7671103941131186e-05, "loss": 0.0374, "step": 53910 }, { "grad_norm": 0.4384879171848297, "learning_rate": 4.765458731637287e-05, "loss": 0.0328, "step": 53920 }, { "grad_norm": 0.8555459976196289, "learning_rate": 4.763807094810537e-05, "loss": 0.0384, "step": 53930 }, { "grad_norm": 0.30491432547569275, "learning_rate": 4.762155483813487e-05, "loss": 0.0363, "step": 53940 }, { "grad_norm": 0.6974449157714844, "learning_rate": 4.760503898826754e-05, "loss": 0.0441, "step": 53950 }, { "grad_norm": 0.2315499186515808, "learning_rate": 4.758852340030957e-05, "loss": 0.05, "step": 53960 }, { "grad_norm": 0.3999892771244049, "learning_rate": 4.757200807606705e-05, "loss": 0.0453, "step": 53970 }, { "grad_norm": 0.4836312234401703, "learning_rate": 4.755549301734607e-05, "loss": 0.0264, "step": 53980 }, { "grad_norm": 0.3640608787536621, "learning_rate": 4.753897822595271e-05, "loss": 0.0433, "step": 53990 }, { "grad_norm": 0.5510070323944092, "learning_rate": 4.7522463703692994e-05, "loss": 0.0306, "step": 54000 }, { "grad_norm": 0.4455724358558655, "learning_rate": 4.75059494523729e-05, "loss": 0.0372, "step": 54010 }, { "grad_norm": 0.26442888379096985, "learning_rate": 4.748943547379843e-05, "loss": 0.0379, "step": 54020 }, { "grad_norm": 0.29146328568458557, "learning_rate": 4.7472921769775535e-05, "loss": 0.0327, "step": 54030 }, { "grad_norm": 0.4661118686199188, "learning_rate": 4.745640834211011e-05, "loss": 0.0377, "step": 54040 }, { "grad_norm": 0.23938938975334167, "learning_rate": 4.7439895192608054e-05, "loss": 0.0339, "step": 54050 }, { "grad_norm": 0.4607292711734772, "learning_rate": 4.7423382323075186e-05, "loss": 0.0296, "step": 54060 }, { "grad_norm": 0.4411310851573944, "learning_rate": 4.740686973531737e-05, "loss": 0.0291, "step": 54070 }, { "grad_norm": 0.4666075110435486, "learning_rate": 4.7390357431140375e-05, "loss": 0.036, "step": 54080 }, { "grad_norm": 0.2862359583377838, "learning_rate": 4.737384541234998e-05, "loss": 0.0307, "step": 54090 }, { "grad_norm": 0.3521648645401001, "learning_rate": 4.735733368075189e-05, "loss": 0.0258, "step": 54100 }, { "grad_norm": 0.3681008815765381, "learning_rate": 4.734082223815182e-05, "loss": 0.0323, "step": 54110 }, { "grad_norm": 0.5935869216918945, "learning_rate": 4.7324311086355436e-05, "loss": 0.0358, "step": 54120 }, { "grad_norm": 0.5964924097061157, "learning_rate": 4.730780022716836e-05, "loss": 0.0344, "step": 54130 }, { "grad_norm": 0.33110475540161133, "learning_rate": 4.729128966239619e-05, "loss": 0.0272, "step": 54140 }, { "grad_norm": 0.5987585783004761, "learning_rate": 4.7274779393844536e-05, "loss": 0.0454, "step": 54150 }, { "grad_norm": 0.6989577412605286, "learning_rate": 4.725826942331891e-05, "loss": 0.0497, "step": 54160 }, { "grad_norm": 0.3293260633945465, "learning_rate": 4.724175975262481e-05, "loss": 0.0281, "step": 54170 }, { "grad_norm": 0.43204906582832336, "learning_rate": 4.722525038356772e-05, "loss": 0.0253, "step": 54180 }, { "grad_norm": 0.3087835907936096, "learning_rate": 4.7208741317953075e-05, "loss": 0.0301, "step": 54190 }, { "grad_norm": 0.3592235743999481, "learning_rate": 4.7192232557586264e-05, "loss": 0.0466, "step": 54200 }, { "grad_norm": 0.43868786096572876, "learning_rate": 4.717572410427269e-05, "loss": 0.033, "step": 54210 }, { "grad_norm": 0.3661310076713562, "learning_rate": 4.71592159598177e-05, "loss": 0.035, "step": 54220 }, { "grad_norm": 0.4711432158946991, "learning_rate": 4.714270812602657e-05, "loss": 0.0367, "step": 54230 }, { "grad_norm": 0.3742847144603729, "learning_rate": 4.712620060470458e-05, "loss": 0.0336, "step": 54240 }, { "grad_norm": 0.38631415367126465, "learning_rate": 4.7109693397656986e-05, "loss": 0.035, "step": 54250 }, { "grad_norm": 0.3422819674015045, "learning_rate": 4.709318650668894e-05, "loss": 0.0305, "step": 54260 }, { "grad_norm": 0.5429989099502563, "learning_rate": 4.7076679933605676e-05, "loss": 0.0308, "step": 54270 }, { "grad_norm": 0.41658660769462585, "learning_rate": 4.70601736802123e-05, "loss": 0.029, "step": 54280 }, { "grad_norm": 0.39474937319755554, "learning_rate": 4.704366774831391e-05, "loss": 0.0461, "step": 54290 }, { "grad_norm": 0.4082862138748169, "learning_rate": 4.702716213971557e-05, "loss": 0.0359, "step": 54300 }, { "grad_norm": 0.42967861890792847, "learning_rate": 4.70106568562223e-05, "loss": 0.0356, "step": 54310 }, { "grad_norm": 0.2856864929199219, "learning_rate": 4.6994151899639105e-05, "loss": 0.0318, "step": 54320 }, { "grad_norm": 0.34627237915992737, "learning_rate": 4.697764727177093e-05, "loss": 0.0381, "step": 54330 }, { "grad_norm": 0.4454251229763031, "learning_rate": 4.696114297442272e-05, "loss": 0.0398, "step": 54340 }, { "grad_norm": 0.3470882177352905, "learning_rate": 4.694463900939933e-05, "loss": 0.0303, "step": 54350 }, { "grad_norm": 0.3622733950614929, "learning_rate": 4.692813537850564e-05, "loss": 0.0311, "step": 54360 }, { "grad_norm": 0.2369624376296997, "learning_rate": 4.6911632083546445e-05, "loss": 0.0272, "step": 54370 }, { "grad_norm": 0.4172489047050476, "learning_rate": 4.6895129126326525e-05, "loss": 0.0398, "step": 54380 }, { "grad_norm": 0.351195365190506, "learning_rate": 4.687862650865059e-05, "loss": 0.0295, "step": 54390 }, { "grad_norm": 0.4027939736843109, "learning_rate": 4.686212423232339e-05, "loss": 0.0371, "step": 54400 }, { "grad_norm": 0.3082429766654968, "learning_rate": 4.6845622299149565e-05, "loss": 0.0279, "step": 54410 }, { "grad_norm": 0.31334763765335083, "learning_rate": 4.682912071093374e-05, "loss": 0.0328, "step": 54420 }, { "grad_norm": 0.3365725576877594, "learning_rate": 4.681261946948052e-05, "loss": 0.0288, "step": 54430 }, { "grad_norm": 0.2857356667518616, "learning_rate": 4.679611857659443e-05, "loss": 0.0383, "step": 54440 }, { "grad_norm": 0.40831539034843445, "learning_rate": 4.6779618034079975e-05, "loss": 0.0357, "step": 54450 }, { "grad_norm": 0.4174247682094574, "learning_rate": 4.6763117843741675e-05, "loss": 0.0382, "step": 54460 }, { "grad_norm": 0.280645489692688, "learning_rate": 4.674661800738394e-05, "loss": 0.0304, "step": 54470 }, { "grad_norm": 0.5107662677764893, "learning_rate": 4.673011852681117e-05, "loss": 0.0376, "step": 54480 }, { "grad_norm": 0.5454421639442444, "learning_rate": 4.671361940382772e-05, "loss": 0.0414, "step": 54490 }, { "grad_norm": 0.3535121977329254, "learning_rate": 4.669712064023791e-05, "loss": 0.0362, "step": 54500 }, { "grad_norm": 0.3246750831604004, "learning_rate": 4.668062223784602e-05, "loss": 0.0341, "step": 54510 }, { "grad_norm": 0.34564992785453796, "learning_rate": 4.6664124198456274e-05, "loss": 0.0339, "step": 54520 }, { "grad_norm": 0.5027449727058411, "learning_rate": 4.6647626523872905e-05, "loss": 0.0296, "step": 54530 }, { "grad_norm": 0.354627400636673, "learning_rate": 4.6631129215900066e-05, "loss": 0.0349, "step": 54540 }, { "grad_norm": 0.5702235698699951, "learning_rate": 4.661463227634187e-05, "loss": 0.0537, "step": 54550 }, { "grad_norm": 0.38976147770881653, "learning_rate": 4.659813570700238e-05, "loss": 0.0299, "step": 54560 }, { "grad_norm": 0.28106197714805603, "learning_rate": 4.658163950968568e-05, "loss": 0.0273, "step": 54570 }, { "grad_norm": 0.34700077772140503, "learning_rate": 4.6565143686195726e-05, "loss": 0.0318, "step": 54580 }, { "grad_norm": 0.48929956555366516, "learning_rate": 4.654864823833649e-05, "loss": 0.0304, "step": 54590 }, { "grad_norm": 0.3118898868560791, "learning_rate": 4.65321531679119e-05, "loss": 0.0297, "step": 54600 }, { "grad_norm": 0.4508977234363556, "learning_rate": 4.6515658476725834e-05, "loss": 0.0457, "step": 54610 }, { "grad_norm": 0.3652622103691101, "learning_rate": 4.649916416658212e-05, "loss": 0.0338, "step": 54620 }, { "grad_norm": 0.5109151005744934, "learning_rate": 4.648267023928455e-05, "loss": 0.0326, "step": 54630 }, { "grad_norm": 0.4844105839729309, "learning_rate": 4.646617669663687e-05, "loss": 0.0246, "step": 54640 }, { "grad_norm": 0.31849685311317444, "learning_rate": 4.644968354044277e-05, "loss": 0.0295, "step": 54650 }, { "grad_norm": 0.8548608422279358, "learning_rate": 4.643319077250596e-05, "loss": 0.0328, "step": 54660 }, { "grad_norm": 0.42634084820747375, "learning_rate": 4.6416698394630055e-05, "loss": 0.0278, "step": 54670 }, { "grad_norm": 0.3793862462043762, "learning_rate": 4.640020640861862e-05, "loss": 0.0331, "step": 54680 }, { "grad_norm": 0.3107777535915375, "learning_rate": 4.6383714816275195e-05, "loss": 0.0343, "step": 54690 }, { "grad_norm": 0.3450895845890045, "learning_rate": 4.636722361940329e-05, "loss": 0.0283, "step": 54700 }, { "grad_norm": 0.5483681559562683, "learning_rate": 4.635073281980632e-05, "loss": 0.035, "step": 54710 }, { "grad_norm": 0.4978562295436859, "learning_rate": 4.633424241928775e-05, "loss": 0.0323, "step": 54720 }, { "grad_norm": 0.28244006633758545, "learning_rate": 4.6317752419650914e-05, "loss": 0.028, "step": 54730 }, { "grad_norm": 0.40836483240127563, "learning_rate": 4.630126282269914e-05, "loss": 0.033, "step": 54740 }, { "grad_norm": 0.2608252763748169, "learning_rate": 4.628477363023569e-05, "loss": 0.0345, "step": 54750 }, { "grad_norm": 0.547185480594635, "learning_rate": 4.626828484406383e-05, "loss": 0.033, "step": 54760 }, { "grad_norm": 0.3479593098163605, "learning_rate": 4.625179646598669e-05, "loss": 0.0374, "step": 54770 }, { "grad_norm": 0.5058972239494324, "learning_rate": 4.6235308497807487e-05, "loss": 0.0337, "step": 54780 }, { "grad_norm": 0.32845067977905273, "learning_rate": 4.6218820941329276e-05, "loss": 0.0348, "step": 54790 }, { "grad_norm": 0.29076722264289856, "learning_rate": 4.620233379835513e-05, "loss": 0.0286, "step": 54800 }, { "grad_norm": 0.4472501575946808, "learning_rate": 4.618584707068804e-05, "loss": 0.0304, "step": 54810 }, { "grad_norm": 0.5455201864242554, "learning_rate": 4.616936076013099e-05, "loss": 0.0356, "step": 54820 }, { "grad_norm": 0.3917945623397827, "learning_rate": 4.615287486848688e-05, "loss": 0.043, "step": 54830 }, { "grad_norm": 0.5120567083358765, "learning_rate": 4.6136389397558584e-05, "loss": 0.0311, "step": 54840 }, { "grad_norm": 0.4194912016391754, "learning_rate": 4.6119904349148935e-05, "loss": 0.0331, "step": 54850 }, { "grad_norm": 0.33219093084335327, "learning_rate": 4.6103419725060726e-05, "loss": 0.0321, "step": 54860 }, { "grad_norm": 0.47100919485092163, "learning_rate": 4.608693552709667e-05, "loss": 0.0302, "step": 54870 }, { "grad_norm": 0.6091354489326477, "learning_rate": 4.607045175705947e-05, "loss": 0.0412, "step": 54880 }, { "grad_norm": 0.3171538710594177, "learning_rate": 4.605396841675176e-05, "loss": 0.0371, "step": 54890 }, { "grad_norm": 0.44694215059280396, "learning_rate": 4.6037485507976094e-05, "loss": 0.0358, "step": 54900 }, { "grad_norm": 0.6212148666381836, "learning_rate": 4.602100303253509e-05, "loss": 0.0374, "step": 54910 }, { "grad_norm": 0.4434320628643036, "learning_rate": 4.600452099223122e-05, "loss": 0.0352, "step": 54920 }, { "grad_norm": 0.31144487857818604, "learning_rate": 4.598803938886691e-05, "loss": 0.027, "step": 54930 }, { "grad_norm": 1.5969988107681274, "learning_rate": 4.597155822424459e-05, "loss": 0.0315, "step": 54940 }, { "grad_norm": 0.45108261704444885, "learning_rate": 4.59550775001666e-05, "loss": 0.0389, "step": 54950 }, { "grad_norm": 0.3660663068294525, "learning_rate": 4.5938597218435234e-05, "loss": 0.0426, "step": 54960 }, { "grad_norm": 0.32074230909347534, "learning_rate": 4.592211738085279e-05, "loss": 0.0404, "step": 54970 }, { "grad_norm": 0.34521839022636414, "learning_rate": 4.590563798922146e-05, "loss": 0.0239, "step": 54980 }, { "grad_norm": 0.5296350121498108, "learning_rate": 4.5889159045343404e-05, "loss": 0.0421, "step": 54990 }, { "grad_norm": 0.49466708302497864, "learning_rate": 4.5872680551020734e-05, "loss": 0.0407, "step": 55000 }, { "grad_norm": 0.5159314274787903, "learning_rate": 4.585620250805552e-05, "loss": 0.0292, "step": 55010 }, { "grad_norm": 0.6901553869247437, "learning_rate": 4.583972491824975e-05, "loss": 0.0327, "step": 55020 }, { "grad_norm": 0.6331256031990051, "learning_rate": 4.582324778340541e-05, "loss": 0.0373, "step": 55030 }, { "grad_norm": 0.2912967801094055, "learning_rate": 4.5806771105324424e-05, "loss": 0.0347, "step": 55040 }, { "grad_norm": 0.42655494809150696, "learning_rate": 4.579029488580865e-05, "loss": 0.0345, "step": 55050 }, { "grad_norm": 0.5412520170211792, "learning_rate": 4.5773819126659876e-05, "loss": 0.0332, "step": 55060 }, { "grad_norm": 0.47860047221183777, "learning_rate": 4.575734382967991e-05, "loss": 0.0307, "step": 55070 }, { "grad_norm": 0.3532177805900574, "learning_rate": 4.574086899667044e-05, "loss": 0.0294, "step": 55080 }, { "grad_norm": 0.4339825510978699, "learning_rate": 4.5724394629433106e-05, "loss": 0.032, "step": 55090 }, { "grad_norm": 0.4543589949607849, "learning_rate": 4.570792072976958e-05, "loss": 0.0321, "step": 55100 }, { "grad_norm": 0.24822980165481567, "learning_rate": 4.569144729948138e-05, "loss": 0.0283, "step": 55110 }, { "grad_norm": 0.5021560788154602, "learning_rate": 4.5674974340370033e-05, "loss": 0.0302, "step": 55120 }, { "grad_norm": 0.4104738235473633, "learning_rate": 4.565850185423699e-05, "loss": 0.0306, "step": 55130 }, { "grad_norm": 0.9425100088119507, "learning_rate": 4.564202984288365e-05, "loss": 0.0416, "step": 55140 }, { "grad_norm": 0.4278291165828705, "learning_rate": 4.5625558308111356e-05, "loss": 0.0297, "step": 55150 }, { "grad_norm": 0.9990520477294922, "learning_rate": 4.560908725172145e-05, "loss": 0.0379, "step": 55160 }, { "grad_norm": 0.3978555202484131, "learning_rate": 4.559261667551516e-05, "loss": 0.035, "step": 55170 }, { "grad_norm": 0.355020135641098, "learning_rate": 4.5576146581293685e-05, "loss": 0.0342, "step": 55180 }, { "grad_norm": 0.3863206207752228, "learning_rate": 4.5559676970858164e-05, "loss": 0.0312, "step": 55190 }, { "grad_norm": 0.5156752467155457, "learning_rate": 4.5543207846009705e-05, "loss": 0.0389, "step": 55200 }, { "grad_norm": 0.7557867169380188, "learning_rate": 4.552673920854933e-05, "loss": 0.0382, "step": 55210 }, { "grad_norm": 0.367523193359375, "learning_rate": 4.5510271060278006e-05, "loss": 0.0318, "step": 55220 }, { "grad_norm": 0.2501370310783386, "learning_rate": 4.5493803402996704e-05, "loss": 0.039, "step": 55230 }, { "grad_norm": 0.37393122911453247, "learning_rate": 4.5477336238506295e-05, "loss": 0.0278, "step": 55240 }, { "grad_norm": 0.23376479744911194, "learning_rate": 4.546086956860759e-05, "loss": 0.026, "step": 55250 }, { "grad_norm": 0.4510616362094879, "learning_rate": 4.5444403395101374e-05, "loss": 0.0283, "step": 55260 }, { "grad_norm": 0.44631025195121765, "learning_rate": 4.542793771978834e-05, "loss": 0.0424, "step": 55270 }, { "grad_norm": 0.5534515976905823, "learning_rate": 4.541147254446916e-05, "loss": 0.038, "step": 55280 }, { "grad_norm": 0.2773470878601074, "learning_rate": 4.539500787094445e-05, "loss": 0.0443, "step": 55290 }, { "grad_norm": 0.2979849576950073, "learning_rate": 4.537854370101475e-05, "loss": 0.0411, "step": 55300 }, { "grad_norm": 0.1998601108789444, "learning_rate": 4.5362080036480566e-05, "loss": 0.0292, "step": 55310 }, { "grad_norm": 0.32236403226852417, "learning_rate": 4.534561687914235e-05, "loss": 0.0351, "step": 55320 }, { "grad_norm": 0.48166677355766296, "learning_rate": 4.5329154230800466e-05, "loss": 0.037, "step": 55330 }, { "grad_norm": 0.2806689441204071, "learning_rate": 4.531269209325525e-05, "loss": 0.028, "step": 55340 }, { "grad_norm": 0.4191969633102417, "learning_rate": 4.529623046830696e-05, "loss": 0.0302, "step": 55350 }, { "grad_norm": 0.3860204219818115, "learning_rate": 4.5279769357755866e-05, "loss": 0.034, "step": 55360 }, { "grad_norm": 0.4023297131061554, "learning_rate": 4.5263308763402084e-05, "loss": 0.0223, "step": 55370 }, { "grad_norm": 0.4125747084617615, "learning_rate": 4.524684868704574e-05, "loss": 0.0304, "step": 55380 }, { "grad_norm": 0.6726717948913574, "learning_rate": 4.523038913048688e-05, "loss": 0.0291, "step": 55390 }, { "grad_norm": 0.4587404727935791, "learning_rate": 4.521393009552549e-05, "loss": 0.0343, "step": 55400 }, { "grad_norm": 0.7346574068069458, "learning_rate": 4.5197471583961484e-05, "loss": 0.0383, "step": 55410 }, { "grad_norm": 0.29599717259407043, "learning_rate": 4.51810135975948e-05, "loss": 0.0495, "step": 55420 }, { "grad_norm": 0.516943633556366, "learning_rate": 4.516455613822521e-05, "loss": 0.0427, "step": 55430 }, { "grad_norm": 0.5532903671264648, "learning_rate": 4.514809920765249e-05, "loss": 0.03, "step": 55440 }, { "grad_norm": 0.4755723178386688, "learning_rate": 4.5131642807676356e-05, "loss": 0.0366, "step": 55450 }, { "grad_norm": 0.4082454741001129, "learning_rate": 4.511518694009644e-05, "loss": 0.0332, "step": 55460 }, { "grad_norm": 0.46209684014320374, "learning_rate": 4.509873160671231e-05, "loss": 0.0354, "step": 55470 }, { "grad_norm": 0.9268755912780762, "learning_rate": 4.508227680932356e-05, "loss": 0.0417, "step": 55480 }, { "grad_norm": 0.382982462644577, "learning_rate": 4.5065822549729605e-05, "loss": 0.0439, "step": 55490 }, { "grad_norm": 0.44898250699043274, "learning_rate": 4.504936882972989e-05, "loss": 0.0319, "step": 55500 }, { "grad_norm": 0.4369097650051117, "learning_rate": 4.503291565112374e-05, "loss": 0.0324, "step": 55510 }, { "grad_norm": 0.2701673209667206, "learning_rate": 4.501646301571048e-05, "loss": 0.0348, "step": 55520 }, { "grad_norm": 0.2537263333797455, "learning_rate": 4.500001092528933e-05, "loss": 0.0341, "step": 55530 }, { "grad_norm": 0.5639099478721619, "learning_rate": 4.4983559381659455e-05, "loss": 0.0374, "step": 55540 }, { "grad_norm": 0.5415008664131165, "learning_rate": 4.496710838661999e-05, "loss": 0.0287, "step": 55550 }, { "grad_norm": 0.3591887056827545, "learning_rate": 4.495065794196999e-05, "loss": 0.0304, "step": 55560 }, { "grad_norm": 0.3329507112503052, "learning_rate": 4.4934208049508455e-05, "loss": 0.0333, "step": 55570 }, { "grad_norm": 0.41179460287094116, "learning_rate": 4.4917758711034303e-05, "loss": 0.0322, "step": 55580 }, { "grad_norm": 0.34657254815101624, "learning_rate": 4.490130992834642e-05, "loss": 0.0313, "step": 55590 }, { "grad_norm": 0.40674978494644165, "learning_rate": 4.488486170324359e-05, "loss": 0.0403, "step": 55600 }, { "grad_norm": 0.6571085453033447, "learning_rate": 4.486841403752463e-05, "loss": 0.0342, "step": 55610 }, { "grad_norm": 0.319711834192276, "learning_rate": 4.4851966932988174e-05, "loss": 0.0278, "step": 55620 }, { "grad_norm": 0.29846861958503723, "learning_rate": 4.4835520391432894e-05, "loss": 0.0303, "step": 55630 }, { "grad_norm": 0.30694058537483215, "learning_rate": 4.481907441465733e-05, "loss": 0.0415, "step": 55640 }, { "grad_norm": 0.4034634828567505, "learning_rate": 4.480262900446e-05, "loss": 0.0384, "step": 55650 }, { "grad_norm": 0.46671056747436523, "learning_rate": 4.478618416263933e-05, "loss": 0.0278, "step": 55660 }, { "grad_norm": 0.39583754539489746, "learning_rate": 4.476973989099373e-05, "loss": 0.0277, "step": 55670 }, { "grad_norm": 0.6000816822052002, "learning_rate": 4.475329619132153e-05, "loss": 0.0494, "step": 55680 }, { "grad_norm": 0.4289263188838959, "learning_rate": 4.4736853065420956e-05, "loss": 0.0357, "step": 55690 }, { "grad_norm": 0.37599679827690125, "learning_rate": 4.472041051509023e-05, "loss": 0.0269, "step": 55700 }, { "grad_norm": 0.3585725426673889, "learning_rate": 4.470396854212746e-05, "loss": 0.0262, "step": 55710 }, { "grad_norm": 0.3952234387397766, "learning_rate": 4.468752714833072e-05, "loss": 0.0255, "step": 55720 }, { "grad_norm": 0.9639552235603333, "learning_rate": 4.467108633549803e-05, "loss": 0.0262, "step": 55730 }, { "grad_norm": 0.2772463262081146, "learning_rate": 4.4654646105427335e-05, "loss": 0.0433, "step": 55740 }, { "grad_norm": 0.640876054763794, "learning_rate": 4.463820645991651e-05, "loss": 0.03, "step": 55750 }, { "grad_norm": 0.33975866436958313, "learning_rate": 4.462176740076334e-05, "loss": 0.0263, "step": 55760 }, { "grad_norm": 0.5702488422393799, "learning_rate": 4.4605328929765624e-05, "loss": 0.0447, "step": 55770 }, { "grad_norm": 0.34712523221969604, "learning_rate": 4.458889104872102e-05, "loss": 0.0281, "step": 55780 }, { "grad_norm": 0.4138750433921814, "learning_rate": 4.457245375942715e-05, "loss": 0.0425, "step": 55790 }, { "grad_norm": 0.3995422422885895, "learning_rate": 4.455601706368157e-05, "loss": 0.0339, "step": 55800 }, { "grad_norm": 0.46139615774154663, "learning_rate": 4.453958096328179e-05, "loss": 0.0346, "step": 55810 }, { "grad_norm": 0.4826284646987915, "learning_rate": 4.452314546002523e-05, "loss": 0.0354, "step": 55820 }, { "grad_norm": 0.28263041377067566, "learning_rate": 4.4506710555709246e-05, "loss": 0.0314, "step": 55830 }, { "grad_norm": 0.6840590238571167, "learning_rate": 4.449027625213114e-05, "loss": 0.0337, "step": 55840 }, { "grad_norm": 0.2978944182395935, "learning_rate": 4.4473842551088135e-05, "loss": 0.0307, "step": 55850 }, { "grad_norm": 0.612598180770874, "learning_rate": 4.445740945437738e-05, "loss": 0.0394, "step": 55860 }, { "grad_norm": 0.2581845819950104, "learning_rate": 4.444097696379602e-05, "loss": 0.0356, "step": 55870 }, { "grad_norm": 0.37877774238586426, "learning_rate": 4.442454508114106e-05, "loss": 0.0316, "step": 55880 }, { "grad_norm": 0.38874801993370056, "learning_rate": 4.4408113808209465e-05, "loss": 0.031, "step": 55890 }, { "grad_norm": 0.5428292751312256, "learning_rate": 4.439168314679813e-05, "loss": 0.0354, "step": 55900 }, { "grad_norm": 0.4208718538284302, "learning_rate": 4.43752530987039e-05, "loss": 0.0345, "step": 55910 }, { "grad_norm": 0.28895530104637146, "learning_rate": 4.4358823665723515e-05, "loss": 0.0264, "step": 55920 }, { "grad_norm": 0.5283138751983643, "learning_rate": 4.434239484965371e-05, "loss": 0.0328, "step": 55930 }, { "grad_norm": 0.5631781220436096, "learning_rate": 4.4325966652291103e-05, "loss": 0.0329, "step": 55940 }, { "grad_norm": 0.41489115357398987, "learning_rate": 4.430953907543225e-05, "loss": 0.0342, "step": 55950 }, { "grad_norm": 0.8786977529525757, "learning_rate": 4.4293112120873645e-05, "loss": 0.0338, "step": 55960 }, { "grad_norm": 0.28137895464897156, "learning_rate": 4.427668579041172e-05, "loss": 0.0274, "step": 55970 }, { "grad_norm": 0.46901726722717285, "learning_rate": 4.4260260085842816e-05, "loss": 0.0446, "step": 55980 }, { "grad_norm": 0.37207189202308655, "learning_rate": 4.4243835008963256e-05, "loss": 0.0317, "step": 55990 }, { "grad_norm": 0.7383219003677368, "learning_rate": 4.422741056156925e-05, "loss": 0.0425, "step": 56000 }, { "grad_norm": 0.24729470908641815, "learning_rate": 4.421098674545693e-05, "loss": 0.0308, "step": 56010 }, { "grad_norm": 0.7346018552780151, "learning_rate": 4.419456356242241e-05, "loss": 0.0382, "step": 56020 }, { "grad_norm": 0.24361220002174377, "learning_rate": 4.4178141014261685e-05, "loss": 0.034, "step": 56030 }, { "grad_norm": 0.49095216393470764, "learning_rate": 4.416171910277071e-05, "loss": 0.0325, "step": 56040 }, { "grad_norm": 0.4739778935909271, "learning_rate": 4.414529782974533e-05, "loss": 0.0312, "step": 56050 }, { "grad_norm": 0.9088736176490784, "learning_rate": 4.41288771969814e-05, "loss": 0.0325, "step": 56060 }, { "grad_norm": 0.5074094533920288, "learning_rate": 4.411245720627464e-05, "loss": 0.0373, "step": 56070 }, { "grad_norm": 0.45411109924316406, "learning_rate": 4.40960378594207e-05, "loss": 0.0366, "step": 56080 }, { "grad_norm": 0.275979220867157, "learning_rate": 4.407961915821519e-05, "loss": 0.0329, "step": 56090 }, { "grad_norm": 0.5341859459877014, "learning_rate": 4.406320110445362e-05, "loss": 0.0364, "step": 56100 }, { "grad_norm": 0.3099191188812256, "learning_rate": 4.404678369993144e-05, "loss": 0.0341, "step": 56110 }, { "grad_norm": 0.3665928244590759, "learning_rate": 4.403036694644406e-05, "loss": 0.0285, "step": 56120 }, { "grad_norm": 0.7026934027671814, "learning_rate": 4.4013950845786764e-05, "loss": 0.0353, "step": 56130 }, { "grad_norm": 0.597410261631012, "learning_rate": 4.399753539975482e-05, "loss": 0.0292, "step": 56140 }, { "grad_norm": 0.6238805055618286, "learning_rate": 4.398112061014337e-05, "loss": 0.0383, "step": 56150 }, { "grad_norm": 0.40622246265411377, "learning_rate": 4.396470647874753e-05, "loss": 0.0241, "step": 56160 }, { "grad_norm": 0.39983534812927246, "learning_rate": 4.394829300736229e-05, "loss": 0.038, "step": 56170 }, { "grad_norm": 0.30549636483192444, "learning_rate": 4.393188019778265e-05, "loss": 0.0265, "step": 56180 }, { "grad_norm": 0.4128950238227844, "learning_rate": 4.391546805180347e-05, "loss": 0.0267, "step": 56190 }, { "grad_norm": 0.2962628901004791, "learning_rate": 4.389905657121955e-05, "loss": 0.0361, "step": 56200 }, { "grad_norm": 0.4830148220062256, "learning_rate": 4.3882645757825635e-05, "loss": 0.0331, "step": 56210 }, { "grad_norm": 0.5275553464889526, "learning_rate": 4.386623561341637e-05, "loss": 0.0281, "step": 56220 }, { "grad_norm": 0.37628939747810364, "learning_rate": 4.384982613978637e-05, "loss": 0.0277, "step": 56230 }, { "grad_norm": 0.3947363495826721, "learning_rate": 4.383341733873012e-05, "loss": 0.0286, "step": 56240 }, { "grad_norm": 0.5744883418083191, "learning_rate": 4.381700921204209e-05, "loss": 0.0345, "step": 56250 }, { "grad_norm": 0.362226277589798, "learning_rate": 4.380060176151663e-05, "loss": 0.0223, "step": 56260 }, { "grad_norm": 0.363849401473999, "learning_rate": 4.378419498894805e-05, "loss": 0.036, "step": 56270 }, { "grad_norm": 0.392909437417984, "learning_rate": 4.376778889613056e-05, "loss": 0.0321, "step": 56280 }, { "grad_norm": 0.4749753475189209, "learning_rate": 4.375138348485831e-05, "loss": 0.0305, "step": 56290 }, { "grad_norm": 0.6325572729110718, "learning_rate": 4.3734978756925346e-05, "loss": 0.0289, "step": 56300 }, { "grad_norm": 0.30361172556877136, "learning_rate": 4.3718574714125706e-05, "loss": 0.0303, "step": 56310 }, { "grad_norm": 0.3398020267486572, "learning_rate": 4.370217135825329e-05, "loss": 0.0381, "step": 56320 }, { "grad_norm": 0.5084532499313354, "learning_rate": 4.368576869110194e-05, "loss": 0.0273, "step": 56330 }, { "grad_norm": 0.4132348299026489, "learning_rate": 4.366936671446544e-05, "loss": 0.0358, "step": 56340 }, { "grad_norm": 0.4573315680027008, "learning_rate": 4.3652965430137474e-05, "loss": 0.0363, "step": 56350 }, { "grad_norm": 0.7200540900230408, "learning_rate": 4.3636564839911646e-05, "loss": 0.0342, "step": 56360 }, { "grad_norm": 0.2257453203201294, "learning_rate": 4.3620164945581545e-05, "loss": 0.0264, "step": 56370 }, { "grad_norm": 0.4123348593711853, "learning_rate": 4.360376574894061e-05, "loss": 0.0281, "step": 56380 }, { "grad_norm": 0.9916888475418091, "learning_rate": 4.358736725178224e-05, "loss": 0.0373, "step": 56390 }, { "grad_norm": 0.7919682860374451, "learning_rate": 4.357096945589974e-05, "loss": 0.044, "step": 56400 }, { "grad_norm": 0.39586755633354187, "learning_rate": 4.355457236308636e-05, "loss": 0.035, "step": 56410 }, { "grad_norm": 0.37561240792274475, "learning_rate": 4.353817597513526e-05, "loss": 0.0265, "step": 56420 }, { "grad_norm": 0.3382346034049988, "learning_rate": 4.352178029383948e-05, "loss": 0.0246, "step": 56430 }, { "grad_norm": 0.39675605297088623, "learning_rate": 4.3505385320992105e-05, "loss": 0.0301, "step": 56440 }, { "grad_norm": 0.466605007648468, "learning_rate": 4.348899105838602e-05, "loss": 0.0271, "step": 56450 }, { "grad_norm": 0.3270680010318756, "learning_rate": 4.3472597507814087e-05, "loss": 0.0261, "step": 56460 }, { "grad_norm": 0.41562405228614807, "learning_rate": 4.3456204671069066e-05, "loss": 0.0281, "step": 56470 }, { "grad_norm": 0.21972502768039703, "learning_rate": 4.343981254994367e-05, "loss": 0.0337, "step": 56480 }, { "grad_norm": 0.4063805937767029, "learning_rate": 4.34234211462305e-05, "loss": 0.0313, "step": 56490 }, { "grad_norm": 0.48911431431770325, "learning_rate": 4.340703046172213e-05, "loss": 0.0402, "step": 56500 }, { "grad_norm": 0.3545535206794739, "learning_rate": 4.339064049821097e-05, "loss": 0.0349, "step": 56510 }, { "grad_norm": 0.3696560859680176, "learning_rate": 4.3374251257489446e-05, "loss": 0.0403, "step": 56520 }, { "grad_norm": 0.5067299008369446, "learning_rate": 4.335786274134984e-05, "loss": 0.0323, "step": 56530 }, { "grad_norm": 0.6066825985908508, "learning_rate": 4.334147495158438e-05, "loss": 0.0472, "step": 56540 }, { "grad_norm": 0.33208921551704407, "learning_rate": 4.33250878899852e-05, "loss": 0.0294, "step": 56550 }, { "grad_norm": 0.8303199410438538, "learning_rate": 4.330870155834437e-05, "loss": 0.0288, "step": 56560 }, { "grad_norm": 0.32676833868026733, "learning_rate": 4.3292315958453884e-05, "loss": 0.0261, "step": 56570 }, { "grad_norm": 0.2567271888256073, "learning_rate": 4.3275931092105646e-05, "loss": 0.0254, "step": 56580 }, { "grad_norm": 0.3058256208896637, "learning_rate": 4.325954696109147e-05, "loss": 0.0256, "step": 56590 }, { "grad_norm": 0.42261576652526855, "learning_rate": 4.32431635672031e-05, "loss": 0.0375, "step": 56600 }, { "grad_norm": 0.5557882785797119, "learning_rate": 4.3226780912232214e-05, "loss": 0.0451, "step": 56610 }, { "grad_norm": 0.49306654930114746, "learning_rate": 4.321039899797036e-05, "loss": 0.0291, "step": 56620 }, { "grad_norm": 1.2772902250289917, "learning_rate": 4.319401782620908e-05, "loss": 0.0291, "step": 56630 }, { "grad_norm": 0.35420843958854675, "learning_rate": 4.317763739873978e-05, "loss": 0.0457, "step": 56640 }, { "grad_norm": 0.37758463621139526, "learning_rate": 4.31612577173538e-05, "loss": 0.0311, "step": 56650 }, { "grad_norm": 0.36855629086494446, "learning_rate": 4.314487878384239e-05, "loss": 0.0365, "step": 56660 }, { "grad_norm": 0.5736621022224426, "learning_rate": 4.3128500599996736e-05, "loss": 0.034, "step": 56670 }, { "grad_norm": 0.5102661848068237, "learning_rate": 4.31121231676079e-05, "loss": 0.0318, "step": 56680 }, { "grad_norm": 0.3961482346057892, "learning_rate": 4.309574648846694e-05, "loss": 0.0283, "step": 56690 }, { "grad_norm": 0.4802846908569336, "learning_rate": 4.3079370564364755e-05, "loss": 0.0395, "step": 56700 }, { "grad_norm": 0.3787076473236084, "learning_rate": 4.306299539709221e-05, "loss": 0.0262, "step": 56710 }, { "grad_norm": 0.27363356947898865, "learning_rate": 4.3046620988440046e-05, "loss": 0.0328, "step": 56720 }, { "grad_norm": 0.3463147282600403, "learning_rate": 4.303024734019897e-05, "loss": 0.0282, "step": 56730 }, { "grad_norm": 0.3762844502925873, "learning_rate": 4.3013874454159565e-05, "loss": 0.0301, "step": 56740 }, { "grad_norm": 0.7336440086364746, "learning_rate": 4.299750233211233e-05, "loss": 0.0343, "step": 56750 }, { "grad_norm": 0.5443066358566284, "learning_rate": 4.2981130975847715e-05, "loss": 0.029, "step": 56760 }, { "grad_norm": 0.2530319392681122, "learning_rate": 4.2964760387156076e-05, "loss": 0.0256, "step": 56770 }, { "grad_norm": 0.27925896644592285, "learning_rate": 4.294839056782767e-05, "loss": 0.0279, "step": 56780 }, { "grad_norm": 0.35064154863357544, "learning_rate": 4.293202151965266e-05, "loss": 0.0306, "step": 56790 }, { "grad_norm": 0.9888091683387756, "learning_rate": 4.291565324442115e-05, "loss": 0.031, "step": 56800 }, { "grad_norm": 0.41037917137145996, "learning_rate": 4.2899285743923135e-05, "loss": 0.0283, "step": 56810 }, { "grad_norm": 0.35108882188796997, "learning_rate": 4.288291901994857e-05, "loss": 0.0261, "step": 56820 }, { "grad_norm": 0.3853716552257538, "learning_rate": 4.286655307428729e-05, "loss": 0.0316, "step": 56830 }, { "grad_norm": 0.33574989438056946, "learning_rate": 4.285018790872903e-05, "loss": 0.0316, "step": 56840 }, { "grad_norm": 0.43993043899536133, "learning_rate": 4.283382352506347e-05, "loss": 0.0359, "step": 56850 }, { "grad_norm": 0.3687466084957123, "learning_rate": 4.2817459925080205e-05, "loss": 0.0413, "step": 56860 }, { "grad_norm": 0.40215614438056946, "learning_rate": 4.28010971105687e-05, "loss": 0.0407, "step": 56870 }, { "grad_norm": 0.5060895085334778, "learning_rate": 4.278473508331841e-05, "loss": 0.0381, "step": 56880 }, { "grad_norm": 0.46902403235435486, "learning_rate": 4.276837384511864e-05, "loss": 0.0418, "step": 56890 }, { "grad_norm": 0.5699923038482666, "learning_rate": 4.275201339775864e-05, "loss": 0.0327, "step": 56900 }, { "grad_norm": 0.3900529742240906, "learning_rate": 4.273565374302756e-05, "loss": 0.0491, "step": 56910 }, { "grad_norm": 0.25980761647224426, "learning_rate": 4.2719294882714454e-05, "loss": 0.0268, "step": 56920 }, { "grad_norm": 0.7064586877822876, "learning_rate": 4.270293681860831e-05, "loss": 0.0371, "step": 56930 }, { "grad_norm": 0.389101505279541, "learning_rate": 4.268657955249802e-05, "loss": 0.024, "step": 56940 }, { "grad_norm": 0.31842413544654846, "learning_rate": 4.267022308617241e-05, "loss": 0.031, "step": 56950 }, { "grad_norm": 0.42442259192466736, "learning_rate": 4.265386742142018e-05, "loss": 0.0341, "step": 56960 }, { "grad_norm": 0.3769401013851166, "learning_rate": 4.263751256002996e-05, "loss": 0.0405, "step": 56970 }, { "grad_norm": 0.3478715717792511, "learning_rate": 4.26211585037903e-05, "loss": 0.0342, "step": 56980 }, { "grad_norm": 0.7575750946998596, "learning_rate": 4.2604805254489656e-05, "loss": 0.0386, "step": 56990 }, { "grad_norm": 0.5565524697303772, "learning_rate": 4.258845281391638e-05, "loss": 0.0333, "step": 57000 }, { "grad_norm": 0.3453887104988098, "learning_rate": 4.257210118385876e-05, "loss": 0.0441, "step": 57010 }, { "grad_norm": 0.42528438568115234, "learning_rate": 4.2555750366104994e-05, "loss": 0.0232, "step": 57020 }, { "grad_norm": 0.2532059848308563, "learning_rate": 4.2539400362443184e-05, "loss": 0.0253, "step": 57030 }, { "grad_norm": 0.6432675123214722, "learning_rate": 4.252305117466133e-05, "loss": 0.0343, "step": 57040 }, { "grad_norm": 0.296507865190506, "learning_rate": 4.2506702804547356e-05, "loss": 0.0282, "step": 57050 }, { "grad_norm": 0.7265031933784485, "learning_rate": 4.249035525388907e-05, "loss": 0.0359, "step": 57060 }, { "grad_norm": 0.3650255799293518, "learning_rate": 4.247400852447427e-05, "loss": 0.0296, "step": 57070 }, { "grad_norm": 0.21483993530273438, "learning_rate": 4.245766261809059e-05, "loss": 0.0245, "step": 57080 }, { "grad_norm": 0.3364298343658447, "learning_rate": 4.244131753652558e-05, "loss": 0.0343, "step": 57090 }, { "grad_norm": 0.5293449759483337, "learning_rate": 4.242497328156672e-05, "loss": 0.0293, "step": 57100 }, { "grad_norm": 0.4147661626338959, "learning_rate": 4.240862985500139e-05, "loss": 0.0323, "step": 57110 }, { "grad_norm": 0.25330299139022827, "learning_rate": 4.2392287258616894e-05, "loss": 0.0275, "step": 57120 }, { "grad_norm": 0.25124114751815796, "learning_rate": 4.2375945494200396e-05, "loss": 0.0272, "step": 57130 }, { "grad_norm": 0.5706856846809387, "learning_rate": 4.2359604563539054e-05, "loss": 0.0273, "step": 57140 }, { "grad_norm": 0.368037611246109, "learning_rate": 4.2343264468419885e-05, "loss": 0.0331, "step": 57150 }, { "grad_norm": 0.28165486454963684, "learning_rate": 4.232692521062979e-05, "loss": 0.0279, "step": 57160 }, { "grad_norm": 0.4407106935977936, "learning_rate": 4.231058679195562e-05, "loss": 0.0279, "step": 57170 }, { "grad_norm": 0.3491728901863098, "learning_rate": 4.229424921418411e-05, "loss": 0.0265, "step": 57180 }, { "grad_norm": 0.7686807513237, "learning_rate": 4.2277912479101916e-05, "loss": 0.0457, "step": 57190 }, { "grad_norm": 0.24053531885147095, "learning_rate": 4.226157658849562e-05, "loss": 0.0305, "step": 57200 }, { "grad_norm": 0.3692273199558258, "learning_rate": 4.2245241544151664e-05, "loss": 0.03, "step": 57210 }, { "grad_norm": 0.37270668148994446, "learning_rate": 4.2228907347856425e-05, "loss": 0.0318, "step": 57220 }, { "grad_norm": 0.37815120816230774, "learning_rate": 4.2212574001396214e-05, "loss": 0.0291, "step": 57230 }, { "grad_norm": 0.33348673582077026, "learning_rate": 4.21962415065572e-05, "loss": 0.0272, "step": 57240 }, { "grad_norm": 0.3811033070087433, "learning_rate": 4.217990986512548e-05, "loss": 0.032, "step": 57250 }, { "grad_norm": 0.32313844561576843, "learning_rate": 4.216357907888704e-05, "loss": 0.0379, "step": 57260 }, { "grad_norm": 0.35029488801956177, "learning_rate": 4.2147249149627824e-05, "loss": 0.0407, "step": 57270 }, { "grad_norm": 0.5281282663345337, "learning_rate": 4.213092007913364e-05, "loss": 0.0399, "step": 57280 }, { "grad_norm": 0.6978352069854736, "learning_rate": 4.2114591869190205e-05, "loss": 0.0332, "step": 57290 }, { "grad_norm": 0.2843961715698242, "learning_rate": 4.209826452158315e-05, "loss": 0.0329, "step": 57300 }, { "grad_norm": 0.47375279664993286, "learning_rate": 4.2081938038098e-05, "loss": 0.0392, "step": 57310 }, { "grad_norm": 0.4808506369590759, "learning_rate": 4.2065612420520185e-05, "loss": 0.0357, "step": 57320 }, { "grad_norm": 0.3971061110496521, "learning_rate": 4.204928767063509e-05, "loss": 0.0289, "step": 57330 }, { "grad_norm": 0.5545461773872375, "learning_rate": 4.2032963790227944e-05, "loss": 0.0337, "step": 57340 }, { "grad_norm": 0.36360421776771545, "learning_rate": 4.2016640781083887e-05, "loss": 0.0302, "step": 57350 }, { "grad_norm": 0.3634650707244873, "learning_rate": 4.2000318644988004e-05, "loss": 0.0478, "step": 57360 }, { "grad_norm": 0.5096963047981262, "learning_rate": 4.198399738372524e-05, "loss": 0.0283, "step": 57370 }, { "grad_norm": 1.1813275814056396, "learning_rate": 4.196767699908045e-05, "loss": 0.0332, "step": 57380 }, { "grad_norm": 0.5647363662719727, "learning_rate": 4.195135749283845e-05, "loss": 0.0278, "step": 57390 }, { "grad_norm": 1.250471830368042, "learning_rate": 4.1935038866783894e-05, "loss": 0.0245, "step": 57400 }, { "grad_norm": 0.24604564905166626, "learning_rate": 4.1918721122701355e-05, "loss": 0.0234, "step": 57410 }, { "grad_norm": 0.4182569682598114, "learning_rate": 4.190240426237533e-05, "loss": 0.0426, "step": 57420 }, { "grad_norm": 0.4128325879573822, "learning_rate": 4.188608828759019e-05, "loss": 0.0325, "step": 57430 }, { "grad_norm": 0.3097747564315796, "learning_rate": 4.1869773200130246e-05, "loss": 0.0415, "step": 57440 }, { "grad_norm": 0.6427181959152222, "learning_rate": 4.185345900177967e-05, "loss": 0.0456, "step": 57450 }, { "grad_norm": 0.40799424052238464, "learning_rate": 4.183714569432258e-05, "loss": 0.0299, "step": 57460 }, { "grad_norm": 0.48735207319259644, "learning_rate": 4.182083327954295e-05, "loss": 0.029, "step": 57470 }, { "grad_norm": 0.4787719249725342, "learning_rate": 4.1804521759224714e-05, "loss": 0.0351, "step": 57480 }, { "grad_norm": 0.33288154006004333, "learning_rate": 4.178821113515165e-05, "loss": 0.0327, "step": 57490 }, { "grad_norm": 0.3226923942565918, "learning_rate": 4.1771901409107476e-05, "loss": 0.0359, "step": 57500 }, { "grad_norm": 0.4061843156814575, "learning_rate": 4.175559258287577e-05, "loss": 0.0398, "step": 57510 }, { "grad_norm": 0.4657389223575592, "learning_rate": 4.173928465824008e-05, "loss": 0.0351, "step": 57520 }, { "grad_norm": 0.3175031244754791, "learning_rate": 4.17229776369838e-05, "loss": 0.0278, "step": 57530 }, { "grad_norm": 0.457106351852417, "learning_rate": 4.170667152089025e-05, "loss": 0.035, "step": 57540 }, { "grad_norm": 0.25948211550712585, "learning_rate": 4.1690366311742626e-05, "loss": 0.0301, "step": 57550 }, { "grad_norm": 0.8424015045166016, "learning_rate": 4.1674062011324055e-05, "loss": 0.0295, "step": 57560 }, { "grad_norm": 0.5106372833251953, "learning_rate": 4.165775862141752e-05, "loss": 0.0307, "step": 57570 }, { "grad_norm": 0.678632378578186, "learning_rate": 4.164145614380598e-05, "loss": 0.0274, "step": 57580 }, { "grad_norm": 0.3789062201976776, "learning_rate": 4.162515458027223e-05, "loss": 0.0379, "step": 57590 }, { "grad_norm": 1.8719152212142944, "learning_rate": 4.160885393259899e-05, "loss": 0.0422, "step": 57600 }, { "grad_norm": 0.3655911684036255, "learning_rate": 4.159255420256886e-05, "loss": 0.0291, "step": 57610 }, { "grad_norm": 0.3630937337875366, "learning_rate": 4.157625539196436e-05, "loss": 0.0399, "step": 57620 }, { "grad_norm": 0.40271231532096863, "learning_rate": 4.155995750256791e-05, "loss": 0.0386, "step": 57630 }, { "grad_norm": 0.3060082793235779, "learning_rate": 4.15436605361618e-05, "loss": 0.0251, "step": 57640 }, { "grad_norm": 0.3672734498977661, "learning_rate": 4.152736449452827e-05, "loss": 0.0357, "step": 57650 }, { "grad_norm": 0.5163816809654236, "learning_rate": 4.151106937944943e-05, "loss": 0.0335, "step": 57660 }, { "grad_norm": 0.3629853427410126, "learning_rate": 4.149477519270727e-05, "loss": 0.0305, "step": 57670 }, { "grad_norm": 0.2864585220813751, "learning_rate": 4.147848193608371e-05, "loss": 0.0285, "step": 57680 }, { "grad_norm": 0.42653822898864746, "learning_rate": 4.146218961136056e-05, "loss": 0.0281, "step": 57690 }, { "grad_norm": 0.28427112102508545, "learning_rate": 4.14458982203195e-05, "loss": 0.033, "step": 57700 }, { "grad_norm": 0.6844100952148438, "learning_rate": 4.1429607764742165e-05, "loss": 0.0304, "step": 57710 }, { "grad_norm": 0.2868390679359436, "learning_rate": 4.141331824641003e-05, "loss": 0.032, "step": 57720 }, { "grad_norm": 0.3951777517795563, "learning_rate": 4.139702966710452e-05, "loss": 0.0284, "step": 57730 }, { "grad_norm": 0.3766923248767853, "learning_rate": 4.13807420286069e-05, "loss": 0.0347, "step": 57740 }, { "grad_norm": 0.39939606189727783, "learning_rate": 4.136445533269837e-05, "loss": 0.0311, "step": 57750 }, { "grad_norm": 0.2209479957818985, "learning_rate": 4.134816958116003e-05, "loss": 0.0351, "step": 57760 }, { "grad_norm": 0.47325530648231506, "learning_rate": 4.133188477577283e-05, "loss": 0.027, "step": 57770 }, { "grad_norm": 0.3842267692089081, "learning_rate": 4.131560091831771e-05, "loss": 0.0346, "step": 57780 }, { "grad_norm": 0.3637964725494385, "learning_rate": 4.1299318010575405e-05, "loss": 0.026, "step": 57790 }, { "grad_norm": 0.3491794764995575, "learning_rate": 4.128303605432661e-05, "loss": 0.0551, "step": 57800 }, { "grad_norm": 0.39502260088920593, "learning_rate": 4.126675505135188e-05, "loss": 0.0293, "step": 57810 }, { "grad_norm": 0.4790922999382019, "learning_rate": 4.125047500343169e-05, "loss": 0.0308, "step": 57820 }, { "grad_norm": 0.2605883777141571, "learning_rate": 4.1234195912346376e-05, "loss": 0.0293, "step": 57830 }, { "grad_norm": 0.37237659096717834, "learning_rate": 4.1217917779876235e-05, "loss": 0.0278, "step": 57840 }, { "grad_norm": 0.27782225608825684, "learning_rate": 4.120164060780141e-05, "loss": 0.0288, "step": 57850 }, { "grad_norm": 0.6078958511352539, "learning_rate": 4.1185364397901935e-05, "loss": 0.0412, "step": 57860 }, { "grad_norm": 0.5475184321403503, "learning_rate": 4.116908915195776e-05, "loss": 0.0214, "step": 57870 }, { "grad_norm": 0.48540714383125305, "learning_rate": 4.115281487174872e-05, "loss": 0.0366, "step": 57880 }, { "grad_norm": 0.4072869122028351, "learning_rate": 4.113654155905452e-05, "loss": 0.0306, "step": 57890 }, { "grad_norm": 0.30729731917381287, "learning_rate": 4.112026921565483e-05, "loss": 0.0318, "step": 57900 }, { "grad_norm": 0.41116178035736084, "learning_rate": 4.1103997843329147e-05, "loss": 0.0297, "step": 57910 }, { "grad_norm": 0.5416795015335083, "learning_rate": 4.108772744385689e-05, "loss": 0.0413, "step": 57920 }, { "grad_norm": 0.39925891160964966, "learning_rate": 4.1071458019017346e-05, "loss": 0.0339, "step": 57930 }, { "grad_norm": 0.31941792368888855, "learning_rate": 4.1055189570589734e-05, "loss": 0.0252, "step": 57940 }, { "grad_norm": 0.435118705034256, "learning_rate": 4.103892210035315e-05, "loss": 0.0237, "step": 57950 }, { "grad_norm": 0.43419355154037476, "learning_rate": 4.102265561008656e-05, "loss": 0.0312, "step": 57960 }, { "grad_norm": 0.44561657309532166, "learning_rate": 4.100639010156885e-05, "loss": 0.028, "step": 57970 }, { "grad_norm": 0.319787859916687, "learning_rate": 4.099012557657882e-05, "loss": 0.0269, "step": 57980 }, { "grad_norm": 0.4781843423843384, "learning_rate": 4.09738620368951e-05, "loss": 0.0224, "step": 57990 }, { "grad_norm": 0.4024215042591095, "learning_rate": 4.095759948429626e-05, "loss": 0.0283, "step": 58000 }, { "grad_norm": 0.2471821904182434, "learning_rate": 4.094133792056075e-05, "loss": 0.0223, "step": 58010 }, { "grad_norm": 0.5452932119369507, "learning_rate": 4.092507734746688e-05, "loss": 0.0278, "step": 58020 }, { "grad_norm": 0.6246496438980103, "learning_rate": 4.090881776679293e-05, "loss": 0.0274, "step": 58030 }, { "grad_norm": 0.6241511106491089, "learning_rate": 4.0892559180316994e-05, "loss": 0.0282, "step": 58040 }, { "grad_norm": 0.24798256158828735, "learning_rate": 4.08763015898171e-05, "loss": 0.0223, "step": 58050 }, { "grad_norm": 0.725861132144928, "learning_rate": 4.0860044997071145e-05, "loss": 0.032, "step": 58060 }, { "grad_norm": 0.6109472513198853, "learning_rate": 4.084378940385693e-05, "loss": 0.0265, "step": 58070 }, { "grad_norm": 0.2870544195175171, "learning_rate": 4.082753481195211e-05, "loss": 0.0243, "step": 58080 }, { "grad_norm": 0.2793990671634674, "learning_rate": 4.0811281223134324e-05, "loss": 0.0303, "step": 58090 }, { "grad_norm": 0.23753425478935242, "learning_rate": 4.0795028639181e-05, "loss": 0.0331, "step": 58100 }, { "grad_norm": 0.505192756652832, "learning_rate": 4.077877706186951e-05, "loss": 0.0348, "step": 58110 }, { "grad_norm": 0.34562692046165466, "learning_rate": 4.076252649297709e-05, "loss": 0.0341, "step": 58120 }, { "grad_norm": 0.3896714150905609, "learning_rate": 4.07462769342809e-05, "loss": 0.029, "step": 58130 }, { "grad_norm": 0.41457071900367737, "learning_rate": 4.073002838755793e-05, "loss": 0.0347, "step": 58140 }, { "grad_norm": 0.2820736765861511, "learning_rate": 4.0713780854585115e-05, "loss": 0.0348, "step": 58150 }, { "grad_norm": 0.3707563281059265, "learning_rate": 4.069753433713927e-05, "loss": 0.03, "step": 58160 }, { "grad_norm": 0.5715282559394836, "learning_rate": 4.068128883699709e-05, "loss": 0.0307, "step": 58170 }, { "grad_norm": 0.39233216643333435, "learning_rate": 4.0665044355935155e-05, "loss": 0.0293, "step": 58180 }, { "grad_norm": 0.5905152559280396, "learning_rate": 4.064880089572993e-05, "loss": 0.0257, "step": 58190 }, { "grad_norm": 0.32870885729789734, "learning_rate": 4.063255845815779e-05, "loss": 0.0481, "step": 58200 }, { "grad_norm": 0.3866782486438751, "learning_rate": 4.0616317044994945e-05, "loss": 0.0278, "step": 58210 }, { "grad_norm": 0.30007627606391907, "learning_rate": 4.0600076658017585e-05, "loss": 0.0336, "step": 58220 }, { "grad_norm": 0.2722839415073395, "learning_rate": 4.0583837299001706e-05, "loss": 0.0375, "step": 58230 }, { "grad_norm": 0.27254050970077515, "learning_rate": 4.056759896972322e-05, "loss": 0.0304, "step": 58240 }, { "grad_norm": 0.33886203169822693, "learning_rate": 4.055136167195793e-05, "loss": 0.0326, "step": 58250 }, { "grad_norm": 0.6072843670845032, "learning_rate": 4.053512540748152e-05, "loss": 0.0307, "step": 58260 }, { "grad_norm": 0.49405160546302795, "learning_rate": 4.051889017806955e-05, "loss": 0.0345, "step": 58270 }, { "grad_norm": 0.8352574110031128, "learning_rate": 4.0502655985497516e-05, "loss": 0.028, "step": 58280 }, { "grad_norm": 0.3600448668003082, "learning_rate": 4.048642283154074e-05, "loss": 0.0361, "step": 58290 }, { "grad_norm": 0.2269892692565918, "learning_rate": 4.0470190717974464e-05, "loss": 0.0315, "step": 58300 }, { "grad_norm": 0.43217170238494873, "learning_rate": 4.045395964657381e-05, "loss": 0.032, "step": 58310 }, { "grad_norm": 0.37001851201057434, "learning_rate": 4.043772961911376e-05, "loss": 0.0224, "step": 58320 }, { "grad_norm": 0.34087467193603516, "learning_rate": 4.042150063736924e-05, "loss": 0.0335, "step": 58330 }, { "grad_norm": 0.4081982374191284, "learning_rate": 4.0405272703114976e-05, "loss": 0.0301, "step": 58340 }, { "grad_norm": 0.33283501863479614, "learning_rate": 4.038904581812569e-05, "loss": 0.0232, "step": 58350 }, { "grad_norm": 0.6360003352165222, "learning_rate": 4.03728199841759e-05, "loss": 0.0269, "step": 58360 }, { "grad_norm": 0.30896052718162537, "learning_rate": 4.0356595203040044e-05, "loss": 0.0338, "step": 58370 }, { "grad_norm": 0.46518346667289734, "learning_rate": 4.0340371476492426e-05, "loss": 0.022, "step": 58380 }, { "grad_norm": 0.7876062989234924, "learning_rate": 4.032414880630727e-05, "loss": 0.0322, "step": 58390 }, { "grad_norm": 0.23888437449932098, "learning_rate": 4.0307927194258646e-05, "loss": 0.0252, "step": 58400 }, { "grad_norm": 0.9243814945220947, "learning_rate": 4.029170664212054e-05, "loss": 0.0311, "step": 58410 }, { "grad_norm": 0.3738029897212982, "learning_rate": 4.0275487151666794e-05, "loss": 0.0317, "step": 58420 }, { "grad_norm": 0.46686097979545593, "learning_rate": 4.025926872467116e-05, "loss": 0.0319, "step": 58430 }, { "grad_norm": 0.4560399651527405, "learning_rate": 4.024305136290725e-05, "loss": 0.028, "step": 58440 }, { "grad_norm": 0.6723423600196838, "learning_rate": 4.022683506814858e-05, "loss": 0.0325, "step": 58450 }, { "grad_norm": 0.31905508041381836, "learning_rate": 4.021061984216853e-05, "loss": 0.0346, "step": 58460 }, { "grad_norm": 0.3239368200302124, "learning_rate": 4.019440568674035e-05, "loss": 0.0353, "step": 58470 }, { "grad_norm": 0.3779032826423645, "learning_rate": 4.017819260363724e-05, "loss": 0.038, "step": 58480 }, { "grad_norm": 0.31863221526145935, "learning_rate": 4.016198059463223e-05, "loss": 0.0394, "step": 58490 }, { "grad_norm": 0.283488005399704, "learning_rate": 4.0145769661498233e-05, "loss": 0.0264, "step": 58500 }, { "grad_norm": 0.3995877802371979, "learning_rate": 4.012955980600804e-05, "loss": 0.0323, "step": 58510 }, { "grad_norm": 0.36373937129974365, "learning_rate": 4.0113351029934355e-05, "loss": 0.0359, "step": 58520 }, { "grad_norm": 0.6074092388153076, "learning_rate": 4.0097143335049714e-05, "loss": 0.0309, "step": 58530 }, { "grad_norm": 0.29966121912002563, "learning_rate": 4.0080936723126606e-05, "loss": 0.0313, "step": 58540 }, { "grad_norm": 0.41139012575149536, "learning_rate": 4.006473119593735e-05, "loss": 0.0348, "step": 58550 }, { "grad_norm": 0.45340728759765625, "learning_rate": 4.004852675525415e-05, "loss": 0.0273, "step": 58560 }, { "grad_norm": 0.2276235669851303, "learning_rate": 4.00323234028491e-05, "loss": 0.0351, "step": 58570 }, { "grad_norm": 0.543928861618042, "learning_rate": 4.0016121140494186e-05, "loss": 0.0369, "step": 58580 }, { "grad_norm": 0.32228484749794006, "learning_rate": 3.999991996996122e-05, "loss": 0.0352, "step": 58590 }, { "grad_norm": 0.2651427388191223, "learning_rate": 3.9983719893022e-05, "loss": 0.0286, "step": 58600 }, { "grad_norm": 0.35660025477409363, "learning_rate": 3.996752091144812e-05, "loss": 0.0451, "step": 58610 }, { "grad_norm": 0.3008517026901245, "learning_rate": 3.9951323027011055e-05, "loss": 0.0336, "step": 58620 }, { "grad_norm": 0.3518969416618347, "learning_rate": 3.993512624148219e-05, "loss": 0.0305, "step": 58630 }, { "grad_norm": 0.3440016210079193, "learning_rate": 3.99189305566328e-05, "loss": 0.0274, "step": 58640 }, { "grad_norm": 0.35896897315979004, "learning_rate": 3.990273597423401e-05, "loss": 0.0289, "step": 58650 }, { "grad_norm": 0.33978986740112305, "learning_rate": 3.988654249605681e-05, "loss": 0.0288, "step": 58660 }, { "grad_norm": 0.572349488735199, "learning_rate": 3.987035012387213e-05, "loss": 0.0341, "step": 58670 }, { "grad_norm": 0.2247723937034607, "learning_rate": 3.9854158859450735e-05, "loss": 0.0304, "step": 58680 }, { "grad_norm": 0.459559828042984, "learning_rate": 3.983796870456328e-05, "loss": 0.0311, "step": 58690 }, { "grad_norm": 0.2520380914211273, "learning_rate": 3.982177966098028e-05, "loss": 0.0266, "step": 58700 }, { "grad_norm": 0.2942425310611725, "learning_rate": 3.980559173047216e-05, "loss": 0.0355, "step": 58710 }, { "grad_norm": 0.42574626207351685, "learning_rate": 3.9789404914809176e-05, "loss": 0.0418, "step": 58720 }, { "grad_norm": 0.28795504570007324, "learning_rate": 3.9773219215761544e-05, "loss": 0.0322, "step": 58730 }, { "grad_norm": 0.4476533830165863, "learning_rate": 3.975703463509928e-05, "loss": 0.035, "step": 58740 }, { "grad_norm": 0.3021102547645569, "learning_rate": 3.974085117459231e-05, "loss": 0.0279, "step": 58750 }, { "grad_norm": 0.35352444648742676, "learning_rate": 3.972466883601043e-05, "loss": 0.0329, "step": 58760 }, { "grad_norm": 0.4283352196216583, "learning_rate": 3.9708487621123316e-05, "loss": 0.0333, "step": 58770 }, { "grad_norm": 0.6994404792785645, "learning_rate": 3.969230753170051e-05, "loss": 0.0337, "step": 58780 }, { "grad_norm": 0.3118453919887543, "learning_rate": 3.967612856951146e-05, "loss": 0.0252, "step": 58790 }, { "grad_norm": 0.49445751309394836, "learning_rate": 3.9659950736325476e-05, "loss": 0.0287, "step": 58800 }, { "grad_norm": 0.5456562042236328, "learning_rate": 3.964377403391174e-05, "loss": 0.0581, "step": 58810 }, { "grad_norm": 0.3629055619239807, "learning_rate": 3.9627598464039303e-05, "loss": 0.0511, "step": 58820 }, { "grad_norm": 0.7490105032920837, "learning_rate": 3.9611424028477096e-05, "loss": 0.0258, "step": 58830 }, { "grad_norm": 0.3499581515789032, "learning_rate": 3.959525072899394e-05, "loss": 0.0373, "step": 58840 }, { "grad_norm": 0.3733901083469391, "learning_rate": 3.9579078567358506e-05, "loss": 0.0272, "step": 58850 }, { "grad_norm": 0.3629085421562195, "learning_rate": 3.956290754533939e-05, "loss": 0.0325, "step": 58860 }, { "grad_norm": 0.23675525188446045, "learning_rate": 3.954673766470501e-05, "loss": 0.0383, "step": 58870 }, { "grad_norm": 0.4265841543674469, "learning_rate": 3.9530568927223674e-05, "loss": 0.0299, "step": 58880 }, { "grad_norm": 0.25221967697143555, "learning_rate": 3.95144013346636e-05, "loss": 0.0375, "step": 58890 }, { "grad_norm": 0.2744952440261841, "learning_rate": 3.949823488879282e-05, "loss": 0.0221, "step": 58900 }, { "grad_norm": 0.3451370298862457, "learning_rate": 3.9482069591379276e-05, "loss": 0.0479, "step": 58910 }, { "grad_norm": 0.8835740089416504, "learning_rate": 3.9465905444190784e-05, "loss": 0.0354, "step": 58920 }, { "grad_norm": 0.3342963457107544, "learning_rate": 3.944974244899505e-05, "loss": 0.0302, "step": 58930 }, { "grad_norm": 0.37181809544563293, "learning_rate": 3.9433580607559616e-05, "loss": 0.0359, "step": 58940 }, { "grad_norm": 0.5349906086921692, "learning_rate": 3.941741992165193e-05, "loss": 0.0406, "step": 58950 }, { "grad_norm": 0.4418013393878937, "learning_rate": 3.940126039303928e-05, "loss": 0.0368, "step": 58960 }, { "grad_norm": 0.35302990674972534, "learning_rate": 3.938510202348886e-05, "loss": 0.0307, "step": 58970 }, { "grad_norm": 0.4261036813259125, "learning_rate": 3.93689448147677e-05, "loss": 0.032, "step": 58980 }, { "grad_norm": 0.2943914234638214, "learning_rate": 3.935278876864278e-05, "loss": 0.0375, "step": 58990 }, { "grad_norm": 0.3816532790660858, "learning_rate": 3.933663388688087e-05, "loss": 0.0466, "step": 59000 }, { "grad_norm": 0.4565812349319458, "learning_rate": 3.9320480171248645e-05, "loss": 0.033, "step": 59010 }, { "grad_norm": 0.6102982759475708, "learning_rate": 3.930432762351265e-05, "loss": 0.0412, "step": 59020 }, { "grad_norm": 0.4767790138721466, "learning_rate": 3.928817624543931e-05, "loss": 0.0327, "step": 59030 }, { "grad_norm": 0.3748553395271301, "learning_rate": 3.9272026038794885e-05, "loss": 0.03, "step": 59040 }, { "grad_norm": 0.4475882351398468, "learning_rate": 3.9255877005345586e-05, "loss": 0.0298, "step": 59050 }, { "grad_norm": 0.30117636919021606, "learning_rate": 3.923972914685743e-05, "loss": 0.0333, "step": 59060 }, { "grad_norm": 0.364406019449234, "learning_rate": 3.922358246509631e-05, "loss": 0.0423, "step": 59070 }, { "grad_norm": 0.675345242023468, "learning_rate": 3.920743696182801e-05, "loss": 0.0368, "step": 59080 }, { "grad_norm": 0.38932082056999207, "learning_rate": 3.9191292638818164e-05, "loss": 0.0362, "step": 59090 }, { "grad_norm": 0.28167012333869934, "learning_rate": 3.9175149497832316e-05, "loss": 0.03, "step": 59100 }, { "grad_norm": 0.24850738048553467, "learning_rate": 3.915900754063584e-05, "loss": 0.0294, "step": 59110 }, { "grad_norm": 0.3978976011276245, "learning_rate": 3.9142866768994014e-05, "loss": 0.027, "step": 59120 }, { "grad_norm": 0.32465195655822754, "learning_rate": 3.912672718467193e-05, "loss": 0.0364, "step": 59130 }, { "grad_norm": 0.47755059599876404, "learning_rate": 3.911058878943463e-05, "loss": 0.0303, "step": 59140 }, { "grad_norm": 0.45647573471069336, "learning_rate": 3.9094451585046974e-05, "loss": 0.0304, "step": 59150 }, { "grad_norm": 0.3056420087814331, "learning_rate": 3.90783155732737e-05, "loss": 0.0313, "step": 59160 }, { "grad_norm": 0.4208352267742157, "learning_rate": 3.9062180755879373e-05, "loss": 0.0353, "step": 59170 }, { "grad_norm": 0.6559253334999084, "learning_rate": 3.904604713462855e-05, "loss": 0.0325, "step": 59180 }, { "grad_norm": 0.3398822546005249, "learning_rate": 3.902991471128554e-05, "loss": 0.0268, "step": 59190 }, { "grad_norm": 0.240260511636734, "learning_rate": 3.901378348761456e-05, "loss": 0.0237, "step": 59200 }, { "grad_norm": 0.2614125907421112, "learning_rate": 3.899765346537971e-05, "loss": 0.0348, "step": 59210 }, { "grad_norm": 0.3782195746898651, "learning_rate": 3.898152464634493e-05, "loss": 0.0271, "step": 59220 }, { "grad_norm": 0.46891939640045166, "learning_rate": 3.896539703227402e-05, "loss": 0.0336, "step": 59230 }, { "grad_norm": 0.41724538803100586, "learning_rate": 3.894927062493073e-05, "loss": 0.0321, "step": 59240 }, { "grad_norm": 0.34628021717071533, "learning_rate": 3.893314542607859e-05, "loss": 0.0252, "step": 59250 }, { "grad_norm": 0.43370944261550903, "learning_rate": 3.891702143748103e-05, "loss": 0.0406, "step": 59260 }, { "grad_norm": 0.43902716040611267, "learning_rate": 3.890089866090133e-05, "loss": 0.029, "step": 59270 }, { "grad_norm": 0.5520307421684265, "learning_rate": 3.888477709810267e-05, "loss": 0.0313, "step": 59280 }, { "grad_norm": 0.2745589315891266, "learning_rate": 3.886865675084806e-05, "loss": 0.0361, "step": 59290 }, { "grad_norm": 0.4289856553077698, "learning_rate": 3.885253762090043e-05, "loss": 0.0252, "step": 59300 }, { "grad_norm": 0.3780837953090668, "learning_rate": 3.8836419710022534e-05, "loss": 0.0412, "step": 59310 }, { "grad_norm": 0.6961049437522888, "learning_rate": 3.882030301997698e-05, "loss": 0.0259, "step": 59320 }, { "grad_norm": 0.42503589391708374, "learning_rate": 3.880418755252629e-05, "loss": 0.0336, "step": 59330 }, { "grad_norm": 0.3991132080554962, "learning_rate": 3.87880733094328e-05, "loss": 0.0257, "step": 59340 }, { "grad_norm": 0.4447467029094696, "learning_rate": 3.877196029245877e-05, "loss": 0.0243, "step": 59350 }, { "grad_norm": 0.2583649754524231, "learning_rate": 3.875584850336627e-05, "loss": 0.0406, "step": 59360 }, { "grad_norm": 0.2847800552845001, "learning_rate": 3.873973794391728e-05, "loss": 0.0284, "step": 59370 }, { "grad_norm": 0.38765382766723633, "learning_rate": 3.872362861587361e-05, "loss": 0.027, "step": 59380 }, { "grad_norm": 0.47431492805480957, "learning_rate": 3.870752052099698e-05, "loss": 0.0262, "step": 59390 }, { "grad_norm": 0.43385329842567444, "learning_rate": 3.8691413661048916e-05, "loss": 0.0292, "step": 59400 }, { "grad_norm": 0.395985871553421, "learning_rate": 3.867530803779085e-05, "loss": 0.0239, "step": 59410 }, { "grad_norm": 0.28203102946281433, "learning_rate": 3.8659203652984054e-05, "loss": 0.0323, "step": 59420 }, { "grad_norm": 0.3454078733921051, "learning_rate": 3.864310050838972e-05, "loss": 0.0313, "step": 59430 }, { "grad_norm": 0.3963296413421631, "learning_rate": 3.8626998605768835e-05, "loss": 0.0272, "step": 59440 }, { "grad_norm": 0.26573407649993896, "learning_rate": 3.8610897946882294e-05, "loss": 0.0318, "step": 59450 }, { "grad_norm": 0.4007907807826996, "learning_rate": 3.859479853349082e-05, "loss": 0.0331, "step": 59460 }, { "grad_norm": 0.546303927898407, "learning_rate": 3.857870036735503e-05, "loss": 0.0268, "step": 59470 }, { "grad_norm": 0.38978564739227295, "learning_rate": 3.856260345023538e-05, "loss": 0.033, "step": 59480 }, { "grad_norm": 0.3861328363418579, "learning_rate": 3.854650778389224e-05, "loss": 0.0286, "step": 59490 }, { "grad_norm": 0.4004354774951935, "learning_rate": 3.8530413370085786e-05, "loss": 0.0262, "step": 59500 }, { "grad_norm": 0.33549052476882935, "learning_rate": 3.851432021057608e-05, "loss": 0.028, "step": 59510 }, { "grad_norm": 0.3432837426662445, "learning_rate": 3.849822830712306e-05, "loss": 0.0274, "step": 59520 }, { "grad_norm": 0.4096421003341675, "learning_rate": 3.848213766148649e-05, "loss": 0.0309, "step": 59530 }, { "grad_norm": 0.3579135239124298, "learning_rate": 3.846604827542602e-05, "loss": 0.0324, "step": 59540 }, { "grad_norm": 0.381165087223053, "learning_rate": 3.844996015070117e-05, "loss": 0.0344, "step": 59550 }, { "grad_norm": 0.3064161539077759, "learning_rate": 3.843387328907132e-05, "loss": 0.0241, "step": 59560 }, { "grad_norm": 0.46883490681648254, "learning_rate": 3.8417787692295696e-05, "loss": 0.0313, "step": 59570 }, { "grad_norm": 0.8477678298950195, "learning_rate": 3.84017033621334e-05, "loss": 0.0274, "step": 59580 }, { "grad_norm": 0.7799570560455322, "learning_rate": 3.838562030034338e-05, "loss": 0.0307, "step": 59590 }, { "grad_norm": 0.3018283247947693, "learning_rate": 3.836953850868447e-05, "loss": 0.0281, "step": 59600 }, { "grad_norm": 0.4386853277683258, "learning_rate": 3.835345798891533e-05, "loss": 0.0321, "step": 59610 }, { "grad_norm": 0.3433855473995209, "learning_rate": 3.833737874279452e-05, "loss": 0.0355, "step": 59620 }, { "grad_norm": 0.392602801322937, "learning_rate": 3.8321300772080433e-05, "loss": 0.025, "step": 59630 }, { "grad_norm": 0.383764386177063, "learning_rate": 3.8305224078531345e-05, "loss": 0.0261, "step": 59640 }, { "grad_norm": 0.6420131921768188, "learning_rate": 3.828914866390537e-05, "loss": 0.0404, "step": 59650 }, { "grad_norm": 0.384985089302063, "learning_rate": 3.827307452996048e-05, "loss": 0.036, "step": 59660 }, { "grad_norm": 0.25451427698135376, "learning_rate": 3.825700167845454e-05, "loss": 0.0423, "step": 59670 }, { "grad_norm": 0.5361944437026978, "learning_rate": 3.8240930111145214e-05, "loss": 0.0312, "step": 59680 }, { "grad_norm": 0.31061092019081116, "learning_rate": 3.8224859829790116e-05, "loss": 0.0257, "step": 59690 }, { "grad_norm": 0.3403680920600891, "learning_rate": 3.820879083614664e-05, "loss": 0.0375, "step": 59700 }, { "grad_norm": 0.3876376450061798, "learning_rate": 3.819272313197208e-05, "loss": 0.0355, "step": 59710 }, { "grad_norm": 0.2813655436038971, "learning_rate": 3.817665671902355e-05, "loss": 0.0301, "step": 59720 }, { "grad_norm": 0.2236107736825943, "learning_rate": 3.816059159905808e-05, "loss": 0.0297, "step": 59730 }, { "grad_norm": 0.3256121873855591, "learning_rate": 3.8144527773832476e-05, "loss": 0.0364, "step": 59740 }, { "grad_norm": 0.5630416870117188, "learning_rate": 3.812846524510352e-05, "loss": 0.0299, "step": 59750 }, { "grad_norm": 0.4249323010444641, "learning_rate": 3.811240401462775e-05, "loss": 0.0272, "step": 59760 }, { "grad_norm": 0.4535975158214569, "learning_rate": 3.809634408416162e-05, "loss": 0.0408, "step": 59770 }, { "grad_norm": 0.41103023290634155, "learning_rate": 3.8080285455461394e-05, "loss": 0.0333, "step": 59780 }, { "grad_norm": 0.34270215034484863, "learning_rate": 3.806422813028323e-05, "loss": 0.0369, "step": 59790 }, { "grad_norm": 0.6014274954795837, "learning_rate": 3.8048172110383114e-05, "loss": 0.03, "step": 59800 }, { "grad_norm": 0.5504427552223206, "learning_rate": 3.8032117397516944e-05, "loss": 0.0374, "step": 59810 }, { "grad_norm": 0.4556073248386383, "learning_rate": 3.801606399344042e-05, "loss": 0.0258, "step": 59820 }, { "grad_norm": 0.5342052578926086, "learning_rate": 3.800001189990913e-05, "loss": 0.0279, "step": 59830 }, { "grad_norm": 0.47390931844711304, "learning_rate": 3.798396111867847e-05, "loss": 0.0374, "step": 59840 }, { "grad_norm": 0.26612845063209534, "learning_rate": 3.796791165150378e-05, "loss": 0.0268, "step": 59850 }, { "grad_norm": 0.4898710548877716, "learning_rate": 3.795186350014018e-05, "loss": 0.0347, "step": 59860 }, { "grad_norm": 0.9080566167831421, "learning_rate": 3.793581666634266e-05, "loss": 0.0329, "step": 59870 }, { "grad_norm": 0.43180230259895325, "learning_rate": 3.791977115186609e-05, "loss": 0.0317, "step": 59880 }, { "grad_norm": 0.3380758464336395, "learning_rate": 3.79037269584652e-05, "loss": 0.0405, "step": 59890 }, { "grad_norm": 0.3071821630001068, "learning_rate": 3.7887684087894545e-05, "loss": 0.0423, "step": 59900 }, { "grad_norm": 0.35919153690338135, "learning_rate": 3.787164254190854e-05, "loss": 0.0362, "step": 59910 }, { "grad_norm": 1.263013243675232, "learning_rate": 3.7855602322261473e-05, "loss": 0.0349, "step": 59920 }, { "grad_norm": 0.40446558594703674, "learning_rate": 3.783956343070746e-05, "loss": 0.0385, "step": 59930 }, { "grad_norm": 0.37006083130836487, "learning_rate": 3.7823525869000526e-05, "loss": 0.0229, "step": 59940 }, { "grad_norm": 0.42532405257225037, "learning_rate": 3.78074896388945e-05, "loss": 0.0367, "step": 59950 }, { "grad_norm": 0.46749529242515564, "learning_rate": 3.7791454742143075e-05, "loss": 0.0283, "step": 59960 }, { "grad_norm": 0.5021992325782776, "learning_rate": 3.7775421180499795e-05, "loss": 0.0259, "step": 59970 }, { "grad_norm": 0.3336973488330841, "learning_rate": 3.7759388955718086e-05, "loss": 0.0284, "step": 59980 }, { "grad_norm": 0.4635908007621765, "learning_rate": 3.774335806955117e-05, "loss": 0.0256, "step": 59990 }, { "grad_norm": 0.526085615158081, "learning_rate": 3.7727328523752215e-05, "loss": 0.0336, "step": 60000 }, { "grad_norm": 0.2733405530452728, "learning_rate": 3.771130032007415e-05, "loss": 0.0321, "step": 60010 }, { "grad_norm": 0.35971418023109436, "learning_rate": 3.769527346026982e-05, "loss": 0.0259, "step": 60020 }, { "grad_norm": 0.27950793504714966, "learning_rate": 3.767924794609188e-05, "loss": 0.0268, "step": 60030 }, { "grad_norm": 0.37398627400398254, "learning_rate": 3.7663223779292854e-05, "loss": 0.0316, "step": 60040 }, { "grad_norm": 0.36676278710365295, "learning_rate": 3.764720096162512e-05, "loss": 0.0351, "step": 60050 }, { "grad_norm": 0.43408334255218506, "learning_rate": 3.763117949484092e-05, "loss": 0.0263, "step": 60060 }, { "grad_norm": 0.5061949491500854, "learning_rate": 3.7615159380692344e-05, "loss": 0.0264, "step": 60070 }, { "grad_norm": 0.36909016966819763, "learning_rate": 3.759914062093132e-05, "loss": 0.0288, "step": 60080 }, { "grad_norm": 0.43330153822898865, "learning_rate": 3.7583123217309615e-05, "loss": 0.0307, "step": 60090 }, { "grad_norm": 0.4732844829559326, "learning_rate": 3.7567107171578904e-05, "loss": 0.0346, "step": 60100 }, { "grad_norm": 0.5847465395927429, "learning_rate": 3.755109248549066e-05, "loss": 0.0374, "step": 60110 }, { "grad_norm": 0.3878706097602844, "learning_rate": 3.75350791607962e-05, "loss": 0.0247, "step": 60120 }, { "grad_norm": 0.4016667902469635, "learning_rate": 3.751906719924676e-05, "loss": 0.0247, "step": 60130 }, { "grad_norm": 0.3625797927379608, "learning_rate": 3.750305660259337e-05, "loss": 0.0339, "step": 60140 }, { "grad_norm": 0.3902626931667328, "learning_rate": 3.748704737258693e-05, "loss": 0.0322, "step": 60150 }, { "grad_norm": 0.51238614320755, "learning_rate": 3.747103951097816e-05, "loss": 0.0411, "step": 60160 }, { "grad_norm": 0.3257591128349304, "learning_rate": 3.745503301951768e-05, "loss": 0.0243, "step": 60170 }, { "grad_norm": 0.602842390537262, "learning_rate": 3.743902789995592e-05, "loss": 0.0349, "step": 60180 }, { "grad_norm": 0.41346508264541626, "learning_rate": 3.7423024154043183e-05, "loss": 0.0283, "step": 60190 }, { "grad_norm": 0.44495657086372375, "learning_rate": 3.740702178352963e-05, "loss": 0.0342, "step": 60200 }, { "grad_norm": 0.4584675431251526, "learning_rate": 3.739102079016523e-05, "loss": 0.0362, "step": 60210 }, { "grad_norm": 0.38577109575271606, "learning_rate": 3.7375021175699846e-05, "loss": 0.0274, "step": 60220 }, { "grad_norm": 0.42262205481529236, "learning_rate": 3.735902294188317e-05, "loss": 0.0324, "step": 60230 }, { "grad_norm": 0.5228291749954224, "learning_rate": 3.734302609046473e-05, "loss": 0.0225, "step": 60240 }, { "grad_norm": 0.6829743981361389, "learning_rate": 3.732703062319391e-05, "loss": 0.0288, "step": 60250 }, { "grad_norm": 0.45564988255500793, "learning_rate": 3.731103654181999e-05, "loss": 0.0377, "step": 60260 }, { "grad_norm": 0.5677486658096313, "learning_rate": 3.729504384809203e-05, "loss": 0.0335, "step": 60270 }, { "grad_norm": 0.39962923526763916, "learning_rate": 3.727905254375898e-05, "loss": 0.0306, "step": 60280 }, { "grad_norm": 0.3267369568347931, "learning_rate": 3.7263062630569614e-05, "loss": 0.0296, "step": 60290 }, { "grad_norm": 0.8638478517532349, "learning_rate": 3.724707411027256e-05, "loss": 0.023, "step": 60300 }, { "grad_norm": 0.47240087389945984, "learning_rate": 3.723108698461631e-05, "loss": 0.0288, "step": 60310 }, { "grad_norm": 0.2766284942626953, "learning_rate": 3.72151012553492e-05, "loss": 0.0339, "step": 60320 }, { "grad_norm": 0.3574378490447998, "learning_rate": 3.719911692421939e-05, "loss": 0.0309, "step": 60330 }, { "grad_norm": 0.4347565472126007, "learning_rate": 3.71831339929749e-05, "loss": 0.0352, "step": 60340 }, { "grad_norm": 0.3089313209056854, "learning_rate": 3.716715246336362e-05, "loss": 0.0292, "step": 60350 }, { "grad_norm": 0.27900683879852295, "learning_rate": 3.715117233713324e-05, "loss": 0.0202, "step": 60360 }, { "grad_norm": 0.4158841371536255, "learning_rate": 3.713519361603135e-05, "loss": 0.0276, "step": 60370 }, { "grad_norm": 0.43331044912338257, "learning_rate": 3.711921630180532e-05, "loss": 0.0296, "step": 60380 }, { "grad_norm": 0.5708861947059631, "learning_rate": 3.710324039620245e-05, "loss": 0.0265, "step": 60390 }, { "grad_norm": 0.8775218725204468, "learning_rate": 3.708726590096982e-05, "loss": 0.0245, "step": 60400 }, { "grad_norm": 0.29731467366218567, "learning_rate": 3.707129281785437e-05, "loss": 0.0305, "step": 60410 }, { "grad_norm": 0.49005189538002014, "learning_rate": 3.705532114860291e-05, "loss": 0.0313, "step": 60420 }, { "grad_norm": 0.6905903816223145, "learning_rate": 3.703935089496207e-05, "loss": 0.0398, "step": 60430 }, { "grad_norm": 0.4348736107349396, "learning_rate": 3.702338205867829e-05, "loss": 0.0311, "step": 60440 }, { "grad_norm": 0.41211897134780884, "learning_rate": 3.7007414641497976e-05, "loss": 0.0359, "step": 60450 }, { "grad_norm": 0.3051442503929138, "learning_rate": 3.699144864516726e-05, "loss": 0.0365, "step": 60460 }, { "grad_norm": 0.3617306649684906, "learning_rate": 3.6975484071432153e-05, "loss": 0.0429, "step": 60470 }, { "grad_norm": 0.3842702805995941, "learning_rate": 3.695952092203854e-05, "loss": 0.0246, "step": 60480 }, { "grad_norm": 0.6286901235580444, "learning_rate": 3.69435591987321e-05, "loss": 0.0418, "step": 60490 }, { "grad_norm": 0.8535678386688232, "learning_rate": 3.6927598903258374e-05, "loss": 0.0371, "step": 60500 }, { "grad_norm": 0.5501943826675415, "learning_rate": 3.691164003736281e-05, "loss": 0.0369, "step": 60510 }, { "grad_norm": 0.41738253831863403, "learning_rate": 3.689568260279061e-05, "loss": 0.0312, "step": 60520 }, { "grad_norm": 0.429686963558197, "learning_rate": 3.687972660128686e-05, "loss": 0.0273, "step": 60530 }, { "grad_norm": 0.4564892053604126, "learning_rate": 3.686377203459648e-05, "loss": 0.0289, "step": 60540 }, { "grad_norm": 0.525260329246521, "learning_rate": 3.6847818904464226e-05, "loss": 0.0372, "step": 60550 }, { "grad_norm": 0.32384541630744934, "learning_rate": 3.683186721263474e-05, "loss": 0.027, "step": 60560 }, { "grad_norm": 0.4910478889942169, "learning_rate": 3.6815916960852435e-05, "loss": 0.0394, "step": 60570 }, { "grad_norm": 0.24781182408332825, "learning_rate": 3.679996815086165e-05, "loss": 0.0315, "step": 60580 }, { "grad_norm": 0.3161913752555847, "learning_rate": 3.678402078440649e-05, "loss": 0.0267, "step": 60590 }, { "grad_norm": 0.21417343616485596, "learning_rate": 3.676807486323096e-05, "loss": 0.0302, "step": 60600 }, { "grad_norm": 0.4783322811126709, "learning_rate": 3.6752130389078864e-05, "loss": 0.0414, "step": 60610 }, { "grad_norm": 0.27395614981651306, "learning_rate": 3.673618736369388e-05, "loss": 0.0352, "step": 60620 }, { "grad_norm": 0.33352425694465637, "learning_rate": 3.6720245788819474e-05, "loss": 0.034, "step": 60630 }, { "grad_norm": 0.4187242388725281, "learning_rate": 3.6704305666199044e-05, "loss": 0.0255, "step": 60640 }, { "grad_norm": 0.2759321928024292, "learning_rate": 3.668836699757576e-05, "loss": 0.0255, "step": 60650 }, { "grad_norm": 0.4127470552921295, "learning_rate": 3.667242978469265e-05, "loss": 0.0224, "step": 60660 }, { "grad_norm": 0.9913913011550903, "learning_rate": 3.6656494029292575e-05, "loss": 0.0401, "step": 60670 }, { "grad_norm": 0.26815152168273926, "learning_rate": 3.664055973311825e-05, "loss": 0.0288, "step": 60680 }, { "grad_norm": 0.21917657554149628, "learning_rate": 3.662462689791221e-05, "loss": 0.0294, "step": 60690 }, { "grad_norm": 0.28203755617141724, "learning_rate": 3.660869552541689e-05, "loss": 0.0252, "step": 60700 }, { "grad_norm": 0.39010584354400635, "learning_rate": 3.659276561737448e-05, "loss": 0.0298, "step": 60710 }, { "grad_norm": 0.3965105712413788, "learning_rate": 3.657683717552708e-05, "loss": 0.0301, "step": 60720 }, { "grad_norm": 0.46101462841033936, "learning_rate": 3.6560910201616574e-05, "loss": 0.0273, "step": 60730 }, { "grad_norm": 0.5627101063728333, "learning_rate": 3.654498469738472e-05, "loss": 0.0333, "step": 60740 }, { "grad_norm": 0.48459839820861816, "learning_rate": 3.6529060664573126e-05, "loss": 0.0325, "step": 60750 }, { "grad_norm": 0.34464243054389954, "learning_rate": 3.6513138104923176e-05, "loss": 0.0289, "step": 60760 }, { "grad_norm": 0.49985024333000183, "learning_rate": 3.6497217020176176e-05, "loss": 0.0284, "step": 60770 }, { "grad_norm": 0.3461321294307709, "learning_rate": 3.648129741207323e-05, "loss": 0.0309, "step": 60780 }, { "grad_norm": 0.37033921480178833, "learning_rate": 3.646537928235527e-05, "loss": 0.0244, "step": 60790 }, { "grad_norm": 0.39004623889923096, "learning_rate": 3.6449462632763075e-05, "loss": 0.0306, "step": 60800 }, { "grad_norm": 0.5560392141342163, "learning_rate": 3.643354746503729e-05, "loss": 0.0292, "step": 60810 }, { "grad_norm": 0.3071853220462799, "learning_rate": 3.6417633780918345e-05, "loss": 0.0267, "step": 60820 }, { "grad_norm": 0.283172070980072, "learning_rate": 3.640172158214655e-05, "loss": 0.0285, "step": 60830 }, { "grad_norm": 0.71707683801651, "learning_rate": 3.6385810870462035e-05, "loss": 0.0309, "step": 60840 }, { "grad_norm": 0.3746010363101959, "learning_rate": 3.636990164760479e-05, "loss": 0.033, "step": 60850 }, { "grad_norm": 0.22468383610248566, "learning_rate": 3.6353993915314596e-05, "loss": 0.023, "step": 60860 }, { "grad_norm": 0.7843456268310547, "learning_rate": 3.6338087675331124e-05, "loss": 0.0345, "step": 60870 }, { "grad_norm": 0.35820552706718445, "learning_rate": 3.632218292939383e-05, "loss": 0.0387, "step": 60880 }, { "grad_norm": 0.31173646450042725, "learning_rate": 3.6306279679242033e-05, "loss": 0.0254, "step": 60890 }, { "grad_norm": 0.3718225359916687, "learning_rate": 3.629037792661492e-05, "loss": 0.0242, "step": 60900 }, { "grad_norm": 0.30270713567733765, "learning_rate": 3.627447767325147e-05, "loss": 0.0316, "step": 60910 }, { "grad_norm": 0.4914771020412445, "learning_rate": 3.62585789208905e-05, "loss": 0.0255, "step": 60920 }, { "grad_norm": 0.4200749695301056, "learning_rate": 3.624268167127068e-05, "loss": 0.0433, "step": 60930 }, { "grad_norm": 0.41094452142715454, "learning_rate": 3.622678592613051e-05, "loss": 0.0309, "step": 60940 }, { "grad_norm": 0.33967211842536926, "learning_rate": 3.6210891687208304e-05, "loss": 0.0271, "step": 60950 }, { "grad_norm": 0.32331082224845886, "learning_rate": 3.619499895624228e-05, "loss": 0.0296, "step": 60960 }, { "grad_norm": 0.47975119948387146, "learning_rate": 3.61791077349704e-05, "loss": 0.0324, "step": 60970 }, { "grad_norm": 0.35292771458625793, "learning_rate": 3.616321802513053e-05, "loss": 0.0405, "step": 60980 }, { "grad_norm": 0.4292606711387634, "learning_rate": 3.6147329828460325e-05, "loss": 0.0298, "step": 60990 }, { "grad_norm": 0.4152046740055084, "learning_rate": 3.6131443146697306e-05, "loss": 0.0328, "step": 61000 }, { "grad_norm": 0.2046593874692917, "learning_rate": 3.611555798157879e-05, "loss": 0.0271, "step": 61010 }, { "grad_norm": 0.39600852131843567, "learning_rate": 3.6099674334842e-05, "loss": 0.0269, "step": 61020 }, { "grad_norm": 0.30361777544021606, "learning_rate": 3.6083792208223934e-05, "loss": 0.0214, "step": 61030 }, { "grad_norm": 0.33167657256126404, "learning_rate": 3.606791160346142e-05, "loss": 0.0335, "step": 61040 }, { "grad_norm": 0.40023380517959595, "learning_rate": 3.6052032522291134e-05, "loss": 0.0277, "step": 61050 }, { "grad_norm": 0.5293272137641907, "learning_rate": 3.603615496644962e-05, "loss": 0.0408, "step": 61060 }, { "grad_norm": 0.4142746031284332, "learning_rate": 3.60202789376732e-05, "loss": 0.0268, "step": 61070 }, { "grad_norm": 0.32569620013237, "learning_rate": 3.6004404437698036e-05, "loss": 0.0263, "step": 61080 }, { "grad_norm": 0.4542677104473114, "learning_rate": 3.5988531468260184e-05, "loss": 0.0248, "step": 61090 }, { "grad_norm": 0.3876276910305023, "learning_rate": 3.597266003109546e-05, "loss": 0.0309, "step": 61100 }, { "grad_norm": 0.2742026448249817, "learning_rate": 3.5956790127939557e-05, "loss": 0.0305, "step": 61110 }, { "grad_norm": 0.5386349558830261, "learning_rate": 3.594092176052796e-05, "loss": 0.0276, "step": 61120 }, { "grad_norm": 0.35677435994148254, "learning_rate": 3.592505493059603e-05, "loss": 0.0363, "step": 61130 }, { "grad_norm": 0.5315008759498596, "learning_rate": 3.5909189639878905e-05, "loss": 0.0383, "step": 61140 }, { "grad_norm": 0.6496252417564392, "learning_rate": 3.5893325890111626e-05, "loss": 0.0267, "step": 61150 }, { "grad_norm": 0.2855297923088074, "learning_rate": 3.5877463683029034e-05, "loss": 0.0369, "step": 61160 }, { "grad_norm": 0.38219884037971497, "learning_rate": 3.586160302036578e-05, "loss": 0.0341, "step": 61170 }, { "grad_norm": 0.23821161687374115, "learning_rate": 3.584574390385636e-05, "loss": 0.0272, "step": 61180 }, { "grad_norm": 1.1584322452545166, "learning_rate": 3.582988633523511e-05, "loss": 0.0373, "step": 61190 }, { "grad_norm": 0.37331175804138184, "learning_rate": 3.581403031623616e-05, "loss": 0.0288, "step": 61200 }, { "grad_norm": 0.49056538939476013, "learning_rate": 3.579817584859354e-05, "loss": 0.0251, "step": 61210 }, { "grad_norm": 0.484225869178772, "learning_rate": 3.578232293404107e-05, "loss": 0.0302, "step": 61220 }, { "grad_norm": 0.8378182053565979, "learning_rate": 3.5766471574312385e-05, "loss": 0.0275, "step": 61230 }, { "grad_norm": 0.35283735394477844, "learning_rate": 3.5750621771140964e-05, "loss": 0.0347, "step": 61240 }, { "grad_norm": 0.33713802695274353, "learning_rate": 3.5734773526260115e-05, "loss": 0.0375, "step": 61250 }, { "grad_norm": 0.5313279628753662, "learning_rate": 3.5718926841402996e-05, "loss": 0.0326, "step": 61260 }, { "grad_norm": 0.43146419525146484, "learning_rate": 3.5703081718302555e-05, "loss": 0.0364, "step": 61270 }, { "grad_norm": 0.3056666851043701, "learning_rate": 3.568723815869162e-05, "loss": 0.0355, "step": 61280 }, { "grad_norm": 0.2910350561141968, "learning_rate": 3.5671396164302777e-05, "loss": 0.0357, "step": 61290 }, { "grad_norm": 0.48216500878334045, "learning_rate": 3.565555573686853e-05, "loss": 0.0396, "step": 61300 }, { "grad_norm": 0.4366825520992279, "learning_rate": 3.563971687812113e-05, "loss": 0.0269, "step": 61310 }, { "grad_norm": 0.35482221841812134, "learning_rate": 3.5623879589792706e-05, "loss": 0.0286, "step": 61320 }, { "grad_norm": 0.3901647925376892, "learning_rate": 3.5608043873615185e-05, "loss": 0.0419, "step": 61330 }, { "grad_norm": 0.2008858174085617, "learning_rate": 3.559220973132035e-05, "loss": 0.0294, "step": 61340 }, { "grad_norm": 0.35029762983322144, "learning_rate": 3.557637716463981e-05, "loss": 0.0258, "step": 61350 }, { "grad_norm": 0.3754078149795532, "learning_rate": 3.556054617530497e-05, "loss": 0.0269, "step": 61360 }, { "grad_norm": 0.21879102289676666, "learning_rate": 3.5544716765047106e-05, "loss": 0.0242, "step": 61370 }, { "grad_norm": 0.37524810433387756, "learning_rate": 3.5528888935597274e-05, "loss": 0.0333, "step": 61380 }, { "grad_norm": 0.22683599591255188, "learning_rate": 3.551306268868636e-05, "loss": 0.0326, "step": 61390 }, { "grad_norm": 0.3478700816631317, "learning_rate": 3.5497238026045166e-05, "loss": 0.0268, "step": 61400 }, { "grad_norm": 0.4430693984031677, "learning_rate": 3.548141494940421e-05, "loss": 0.029, "step": 61410 }, { "grad_norm": 0.40641888976097107, "learning_rate": 3.546559346049388e-05, "loss": 0.0272, "step": 61420 }, { "grad_norm": 0.3184100389480591, "learning_rate": 3.544977356104441e-05, "loss": 0.0227, "step": 61430 }, { "grad_norm": 0.3379160761833191, "learning_rate": 3.543395525278582e-05, "loss": 0.0344, "step": 61440 }, { "grad_norm": 0.26920950412750244, "learning_rate": 3.5418138537447985e-05, "loss": 0.0248, "step": 61450 }, { "grad_norm": 0.6991006135940552, "learning_rate": 3.5402323416760574e-05, "loss": 0.0262, "step": 61460 }, { "grad_norm": 0.7477411031723022, "learning_rate": 3.538650989245316e-05, "loss": 0.0301, "step": 61470 }, { "grad_norm": 0.3488060534000397, "learning_rate": 3.537069796625504e-05, "loss": 0.0233, "step": 61480 }, { "grad_norm": 0.2827448546886444, "learning_rate": 3.53548876398954e-05, "loss": 0.0198, "step": 61490 }, { "grad_norm": 0.3212963342666626, "learning_rate": 3.5339078915103214e-05, "loss": 0.0267, "step": 61500 }, { "grad_norm": 0.5026386976242065, "learning_rate": 3.532327179360733e-05, "loss": 0.0331, "step": 61510 }, { "grad_norm": 0.7130103707313538, "learning_rate": 3.530746627713636e-05, "loss": 0.0282, "step": 61520 }, { "grad_norm": 0.2917228639125824, "learning_rate": 3.5291662367418805e-05, "loss": 0.0268, "step": 61530 }, { "grad_norm": 0.3109686076641083, "learning_rate": 3.527586006618293e-05, "loss": 0.0354, "step": 61540 }, { "grad_norm": 1.0628389120101929, "learning_rate": 3.526005937515687e-05, "loss": 0.0424, "step": 61550 }, { "grad_norm": 0.4070896804332733, "learning_rate": 3.524426029606856e-05, "loss": 0.0373, "step": 61560 }, { "grad_norm": 0.452166885137558, "learning_rate": 3.5228462830645746e-05, "loss": 0.0342, "step": 61570 }, { "grad_norm": 0.49863219261169434, "learning_rate": 3.5212666980616036e-05, "loss": 0.0296, "step": 61580 }, { "grad_norm": 0.24215112626552582, "learning_rate": 3.519687274770681e-05, "loss": 0.0358, "step": 61590 }, { "grad_norm": 0.2505866587162018, "learning_rate": 3.518108013364535e-05, "loss": 0.0402, "step": 61600 }, { "grad_norm": 0.31951427459716797, "learning_rate": 3.516528914015868e-05, "loss": 0.0308, "step": 61610 }, { "grad_norm": 0.3294694721698761, "learning_rate": 3.5149499768973694e-05, "loss": 0.0258, "step": 61620 }, { "grad_norm": 0.6222355961799622, "learning_rate": 3.513371202181708e-05, "loss": 0.027, "step": 61630 }, { "grad_norm": 0.6926174759864807, "learning_rate": 3.511792590041537e-05, "loss": 0.0208, "step": 61640 }, { "grad_norm": 0.39873847365379333, "learning_rate": 3.510214140649488e-05, "loss": 0.037, "step": 61650 }, { "grad_norm": 0.36782145500183105, "learning_rate": 3.508635854178184e-05, "loss": 0.0419, "step": 61660 }, { "grad_norm": 0.3268624544143677, "learning_rate": 3.50705773080022e-05, "loss": 0.0321, "step": 61670 }, { "grad_norm": 0.36305946111679077, "learning_rate": 3.505479770688178e-05, "loss": 0.0267, "step": 61680 }, { "grad_norm": 0.6609141230583191, "learning_rate": 3.503901974014622e-05, "loss": 0.0376, "step": 61690 }, { "grad_norm": 0.32908347249031067, "learning_rate": 3.502324340952095e-05, "loss": 0.0342, "step": 61700 }, { "grad_norm": 0.4423414170742035, "learning_rate": 3.500746871673126e-05, "loss": 0.0337, "step": 61710 }, { "grad_norm": 0.35486677289009094, "learning_rate": 3.499169566350227e-05, "loss": 0.0307, "step": 61720 }, { "grad_norm": 0.30341821908950806, "learning_rate": 3.4975924251558876e-05, "loss": 0.028, "step": 61730 }, { "grad_norm": 0.3770216107368469, "learning_rate": 3.496015448262582e-05, "loss": 0.0341, "step": 61740 }, { "grad_norm": 0.5822699069976807, "learning_rate": 3.494438635842764e-05, "loss": 0.0228, "step": 61750 }, { "grad_norm": 0.29716435074806213, "learning_rate": 3.492861988068874e-05, "loss": 0.0287, "step": 61760 }, { "grad_norm": 0.46039074659347534, "learning_rate": 3.491285505113332e-05, "loss": 0.0378, "step": 61770 }, { "grad_norm": 0.25229334831237793, "learning_rate": 3.4897091871485376e-05, "loss": 0.032, "step": 61780 }, { "grad_norm": 0.25158464908599854, "learning_rate": 3.488133034346875e-05, "loss": 0.0266, "step": 61790 }, { "grad_norm": 0.39855828881263733, "learning_rate": 3.486557046880713e-05, "loss": 0.0293, "step": 61800 }, { "grad_norm": 0.6054248809814453, "learning_rate": 3.484981224922395e-05, "loss": 0.0341, "step": 61810 }, { "grad_norm": 0.4107932150363922, "learning_rate": 3.4834055686442544e-05, "loss": 0.03, "step": 61820 }, { "grad_norm": 0.38283097743988037, "learning_rate": 3.4818300782186e-05, "loss": 0.0285, "step": 61830 }, { "grad_norm": 0.22378411889076233, "learning_rate": 3.4802547538177236e-05, "loss": 0.0269, "step": 61840 }, { "grad_norm": 0.2917477786540985, "learning_rate": 3.478679595613905e-05, "loss": 0.0308, "step": 61850 }, { "grad_norm": 0.4627634584903717, "learning_rate": 3.4771046037793975e-05, "loss": 0.0369, "step": 61860 }, { "grad_norm": 0.9986240863800049, "learning_rate": 3.4755297784864416e-05, "loss": 0.0275, "step": 61870 }, { "grad_norm": 0.35997888445854187, "learning_rate": 3.473955119907257e-05, "loss": 0.0198, "step": 61880 }, { "grad_norm": 0.560450553894043, "learning_rate": 3.472380628214046e-05, "loss": 0.0291, "step": 61890 }, { "grad_norm": 0.36919018626213074, "learning_rate": 3.4708063035789906e-05, "loss": 0.0352, "step": 61900 }, { "grad_norm": 0.25130486488342285, "learning_rate": 3.4692321461742614e-05, "loss": 0.0321, "step": 61910 }, { "grad_norm": 0.3451940417289734, "learning_rate": 3.4676581561720023e-05, "loss": 0.0242, "step": 61920 }, { "grad_norm": 0.8483043313026428, "learning_rate": 3.466084333744344e-05, "loss": 0.0376, "step": 61930 }, { "grad_norm": 0.3125505745410919, "learning_rate": 3.4645106790633964e-05, "loss": 0.0371, "step": 61940 }, { "grad_norm": 0.23749694228172302, "learning_rate": 3.4629371923012526e-05, "loss": 0.0271, "step": 61950 }, { "grad_norm": 0.45477673411369324, "learning_rate": 3.4613638736299856e-05, "loss": 0.0317, "step": 61960 }, { "grad_norm": 0.3508300185203552, "learning_rate": 3.459790723221652e-05, "loss": 0.024, "step": 61970 }, { "grad_norm": 0.42988982796669006, "learning_rate": 3.45821774124829e-05, "loss": 0.027, "step": 61980 }, { "grad_norm": 0.6760603785514832, "learning_rate": 3.456644927881918e-05, "loss": 0.0277, "step": 61990 }, { "grad_norm": 0.2943008840084076, "learning_rate": 3.4550722832945354e-05, "loss": 0.0236, "step": 62000 }, { "grad_norm": 0.628438413143158, "learning_rate": 3.453499807658126e-05, "loss": 0.0251, "step": 62010 }, { "grad_norm": 0.2660258114337921, "learning_rate": 3.451927501144653e-05, "loss": 0.0264, "step": 62020 }, { "grad_norm": 0.2948867678642273, "learning_rate": 3.450355363926059e-05, "loss": 0.0281, "step": 62030 }, { "grad_norm": 0.4158478081226349, "learning_rate": 3.448783396174273e-05, "loss": 0.0357, "step": 62040 }, { "grad_norm": 0.369978129863739, "learning_rate": 3.4472115980612034e-05, "loss": 0.0323, "step": 62050 }, { "grad_norm": 0.28923821449279785, "learning_rate": 3.4456399697587396e-05, "loss": 0.0257, "step": 62060 }, { "grad_norm": 0.30413496494293213, "learning_rate": 3.4440685114387506e-05, "loss": 0.027, "step": 62070 }, { "grad_norm": 0.36944779753685, "learning_rate": 3.442497223273091e-05, "loss": 0.0315, "step": 62080 }, { "grad_norm": 0.35409533977508545, "learning_rate": 3.4409261054335926e-05, "loss": 0.0474, "step": 62090 }, { "grad_norm": 0.3715273141860962, "learning_rate": 3.439355158092069e-05, "loss": 0.0371, "step": 62100 }, { "grad_norm": 0.3692476153373718, "learning_rate": 3.437784381420322e-05, "loss": 0.0292, "step": 62110 }, { "grad_norm": 0.6121883988380432, "learning_rate": 3.436213775590126e-05, "loss": 0.022, "step": 62120 }, { "grad_norm": 0.45746520161628723, "learning_rate": 3.4346433407732405e-05, "loss": 0.026, "step": 62130 }, { "grad_norm": 0.2731863856315613, "learning_rate": 3.433073077141405e-05, "loss": 0.0271, "step": 62140 }, { "grad_norm": 1.1920082569122314, "learning_rate": 3.431502984866343e-05, "loss": 0.0385, "step": 62150 }, { "grad_norm": 0.5600683093070984, "learning_rate": 3.429933064119753e-05, "loss": 0.0405, "step": 62160 }, { "grad_norm": 1.1370742321014404, "learning_rate": 3.4283633150733255e-05, "loss": 0.0266, "step": 62170 }, { "grad_norm": 0.41365620493888855, "learning_rate": 3.4267937378987225e-05, "loss": 0.0332, "step": 62180 }, { "grad_norm": 0.29924875497817993, "learning_rate": 3.425224332767592e-05, "loss": 0.0228, "step": 62190 }, { "grad_norm": 0.5313100218772888, "learning_rate": 3.4236550998515595e-05, "loss": 0.0352, "step": 62200 }, { "grad_norm": 0.47016164660453796, "learning_rate": 3.422086039322235e-05, "loss": 0.0246, "step": 62210 }, { "grad_norm": 0.32428422570228577, "learning_rate": 3.420517151351209e-05, "loss": 0.0281, "step": 62220 }, { "grad_norm": 0.40880998969078064, "learning_rate": 3.418948436110053e-05, "loss": 0.0364, "step": 62230 }, { "grad_norm": 0.2767348289489746, "learning_rate": 3.417379893770319e-05, "loss": 0.0256, "step": 62240 }, { "grad_norm": 0.29988041520118713, "learning_rate": 3.41581152450354e-05, "loss": 0.0271, "step": 62250 }, { "grad_norm": 0.37867459654808044, "learning_rate": 3.414243328481231e-05, "loss": 0.0295, "step": 62260 }, { "grad_norm": 0.3269144296646118, "learning_rate": 3.412675305874888e-05, "loss": 0.0279, "step": 62270 }, { "grad_norm": 0.9193000197410583, "learning_rate": 3.4111074568559876e-05, "loss": 0.0365, "step": 62280 }, { "grad_norm": 0.5199177861213684, "learning_rate": 3.4095397815959846e-05, "loss": 0.0337, "step": 62290 }, { "grad_norm": 0.6087578535079956, "learning_rate": 3.407972280266322e-05, "loss": 0.0322, "step": 62300 }, { "grad_norm": 0.23680201172828674, "learning_rate": 3.4064049530384166e-05, "loss": 0.0289, "step": 62310 }, { "grad_norm": 0.27225780487060547, "learning_rate": 3.40483780008367e-05, "loss": 0.04, "step": 62320 }, { "grad_norm": 1.3927546739578247, "learning_rate": 3.4032708215734635e-05, "loss": 0.0289, "step": 62330 }, { "grad_norm": 0.4157269299030304, "learning_rate": 3.4017040176791584e-05, "loss": 0.0279, "step": 62340 }, { "grad_norm": 0.37861818075180054, "learning_rate": 3.4001373885720976e-05, "loss": 0.0248, "step": 62350 }, { "grad_norm": 0.367458313703537, "learning_rate": 3.398570934423608e-05, "loss": 0.0378, "step": 62360 }, { "grad_norm": 0.24931718409061432, "learning_rate": 3.397004655404994e-05, "loss": 0.0233, "step": 62370 }, { "grad_norm": 0.5578045845031738, "learning_rate": 3.395438551687541e-05, "loss": 0.0343, "step": 62380 }, { "grad_norm": 0.2625240981578827, "learning_rate": 3.393872623442515e-05, "loss": 0.0291, "step": 62390 }, { "grad_norm": 0.38210442662239075, "learning_rate": 3.392306870841164e-05, "loss": 0.0237, "step": 62400 }, { "grad_norm": 0.382556289434433, "learning_rate": 3.390741294054715e-05, "loss": 0.0319, "step": 62410 }, { "grad_norm": 0.37775254249572754, "learning_rate": 3.38917589325438e-05, "loss": 0.0232, "step": 62420 }, { "grad_norm": 0.21663245558738708, "learning_rate": 3.387610668611349e-05, "loss": 0.0292, "step": 62430 }, { "grad_norm": 0.28399422764778137, "learning_rate": 3.38604562029679e-05, "loss": 0.032, "step": 62440 }, { "grad_norm": 0.4085908830165863, "learning_rate": 3.384480748481855e-05, "loss": 0.0292, "step": 62450 }, { "grad_norm": 0.399014413356781, "learning_rate": 3.382916053337676e-05, "loss": 0.0257, "step": 62460 }, { "grad_norm": 0.3282289206981659, "learning_rate": 3.381351535035366e-05, "loss": 0.0301, "step": 62470 }, { "grad_norm": 0.33207273483276367, "learning_rate": 3.379787193746019e-05, "loss": 0.0311, "step": 62480 }, { "grad_norm": 0.2959323525428772, "learning_rate": 3.378223029640709e-05, "loss": 0.0271, "step": 62490 }, { "grad_norm": 0.36186525225639343, "learning_rate": 3.376659042890489e-05, "loss": 0.0276, "step": 62500 }, { "grad_norm": 0.2581906318664551, "learning_rate": 3.375095233666397e-05, "loss": 0.0255, "step": 62510 }, { "grad_norm": 0.21445141732692719, "learning_rate": 3.373531602139447e-05, "loss": 0.0249, "step": 62520 }, { "grad_norm": 0.4793718457221985, "learning_rate": 3.371968148480635e-05, "loss": 0.0282, "step": 62530 }, { "grad_norm": 0.38539403676986694, "learning_rate": 3.370404872860937e-05, "loss": 0.0272, "step": 62540 }, { "grad_norm": 0.6377567052841187, "learning_rate": 3.3688417754513145e-05, "loss": 0.0342, "step": 62550 }, { "grad_norm": 0.468827486038208, "learning_rate": 3.367278856422703e-05, "loss": 0.0291, "step": 62560 }, { "grad_norm": 0.32758158445358276, "learning_rate": 3.365716115946021e-05, "loss": 0.0289, "step": 62570 }, { "grad_norm": 0.586032509803772, "learning_rate": 3.3641535541921666e-05, "loss": 0.0436, "step": 62580 }, { "grad_norm": 0.24072451889514923, "learning_rate": 3.36259117133202e-05, "loss": 0.0359, "step": 62590 }, { "grad_norm": 0.3781239092350006, "learning_rate": 3.36102896753644e-05, "loss": 0.0262, "step": 62600 }, { "grad_norm": 0.45933809876441956, "learning_rate": 3.35946694297627e-05, "loss": 0.0312, "step": 62610 }, { "grad_norm": 0.32800835371017456, "learning_rate": 3.357905097822328e-05, "loss": 0.0258, "step": 62620 }, { "grad_norm": 0.6081997156143188, "learning_rate": 3.356343432245416e-05, "loss": 0.0337, "step": 62630 }, { "grad_norm": 0.3507574796676636, "learning_rate": 3.354781946416315e-05, "loss": 0.0461, "step": 62640 }, { "grad_norm": 0.3530886769294739, "learning_rate": 3.3532206405057864e-05, "loss": 0.0282, "step": 62650 }, { "grad_norm": 0.29994869232177734, "learning_rate": 3.351659514684573e-05, "loss": 0.029, "step": 62660 }, { "grad_norm": 0.5261679887771606, "learning_rate": 3.350098569123395e-05, "loss": 0.0273, "step": 62670 }, { "grad_norm": 0.41590961813926697, "learning_rate": 3.348537803992958e-05, "loss": 0.027, "step": 62680 }, { "grad_norm": 0.2304355502128601, "learning_rate": 3.346977219463945e-05, "loss": 0.0296, "step": 62690 }, { "grad_norm": 0.2603332996368408, "learning_rate": 3.345416815707017e-05, "loss": 0.0214, "step": 62700 }, { "grad_norm": 0.30011317133903503, "learning_rate": 3.3438565928928185e-05, "loss": 0.0237, "step": 62710 }, { "grad_norm": 0.22543957829475403, "learning_rate": 3.3422965511919736e-05, "loss": 0.0305, "step": 62720 }, { "grad_norm": 0.4237183928489685, "learning_rate": 3.3407366907750846e-05, "loss": 0.023, "step": 62730 }, { "grad_norm": 0.30226242542266846, "learning_rate": 3.339177011812737e-05, "loss": 0.0293, "step": 62740 }, { "grad_norm": 0.4462341070175171, "learning_rate": 3.3376175144754935e-05, "loss": 0.0378, "step": 62750 }, { "grad_norm": 0.30976492166519165, "learning_rate": 3.3360581989339004e-05, "loss": 0.0216, "step": 62760 }, { "grad_norm": 0.2844661474227905, "learning_rate": 3.3344990653584804e-05, "loss": 0.0258, "step": 62770 }, { "grad_norm": 0.3863868713378906, "learning_rate": 3.332940113919739e-05, "loss": 0.0307, "step": 62780 }, { "grad_norm": 0.273206502199173, "learning_rate": 3.331381344788161e-05, "loss": 0.0377, "step": 62790 }, { "grad_norm": 0.40490052103996277, "learning_rate": 3.3298227581342076e-05, "loss": 0.0352, "step": 62800 }, { "grad_norm": 0.41164660453796387, "learning_rate": 3.328264354128329e-05, "loss": 0.0392, "step": 62810 }, { "grad_norm": 0.39919644594192505, "learning_rate": 3.3267061329409465e-05, "loss": 0.0302, "step": 62820 }, { "grad_norm": 0.2589864134788513, "learning_rate": 3.325148094742466e-05, "loss": 0.0321, "step": 62830 }, { "grad_norm": 0.36073675751686096, "learning_rate": 3.323590239703271e-05, "loss": 0.0301, "step": 62840 }, { "grad_norm": 0.37807419896125793, "learning_rate": 3.322032567993728e-05, "loss": 0.0298, "step": 62850 }, { "grad_norm": 0.37069299817085266, "learning_rate": 3.3204750797841774e-05, "loss": 0.0266, "step": 62860 }, { "grad_norm": 0.2921024560928345, "learning_rate": 3.31891777524495e-05, "loss": 0.0303, "step": 62870 }, { "grad_norm": 0.6304243803024292, "learning_rate": 3.317360654546347e-05, "loss": 0.0352, "step": 62880 }, { "grad_norm": 0.5005564093589783, "learning_rate": 3.3158037178586534e-05, "loss": 0.0296, "step": 62890 }, { "grad_norm": 0.4556523561477661, "learning_rate": 3.3142469653521316e-05, "loss": 0.0341, "step": 62900 }, { "grad_norm": 0.2923244833946228, "learning_rate": 3.3126903971970284e-05, "loss": 0.0296, "step": 62910 }, { "grad_norm": 0.4616814851760864, "learning_rate": 3.311134013563564e-05, "loss": 0.025, "step": 62920 }, { "grad_norm": 0.23015689849853516, "learning_rate": 3.309577814621947e-05, "loss": 0.0287, "step": 62930 }, { "grad_norm": 0.4901183843612671, "learning_rate": 3.308021800542359e-05, "loss": 0.0239, "step": 62940 }, { "grad_norm": 0.36722734570503235, "learning_rate": 3.306465971494963e-05, "loss": 0.0291, "step": 62950 }, { "grad_norm": 0.3396478295326233, "learning_rate": 3.304910327649901e-05, "loss": 0.0279, "step": 62960 }, { "grad_norm": 0.25992393493652344, "learning_rate": 3.303354869177297e-05, "loss": 0.0212, "step": 62970 }, { "grad_norm": 0.2775655686855316, "learning_rate": 3.301799596247255e-05, "loss": 0.0227, "step": 62980 }, { "grad_norm": 0.2513357996940613, "learning_rate": 3.300244509029854e-05, "loss": 0.0271, "step": 62990 }, { "grad_norm": 0.3097909092903137, "learning_rate": 3.2986896076951585e-05, "loss": 0.0293, "step": 63000 }, { "grad_norm": 0.26756551861763, "learning_rate": 3.297134892413211e-05, "loss": 0.0243, "step": 63010 }, { "grad_norm": 0.29979372024536133, "learning_rate": 3.295580363354031e-05, "loss": 0.03, "step": 63020 }, { "grad_norm": 0.5480395555496216, "learning_rate": 3.2940260206876194e-05, "loss": 0.0384, "step": 63030 }, { "grad_norm": 0.26944097876548767, "learning_rate": 3.292471864583958e-05, "loss": 0.0267, "step": 63040 }, { "grad_norm": 0.4448608458042145, "learning_rate": 3.290917895213004e-05, "loss": 0.0297, "step": 63050 }, { "grad_norm": 0.40937939286231995, "learning_rate": 3.289364112744701e-05, "loss": 0.0302, "step": 63060 }, { "grad_norm": 0.3391706943511963, "learning_rate": 3.287810517348967e-05, "loss": 0.0348, "step": 63070 }, { "grad_norm": 0.5639208555221558, "learning_rate": 3.2862571091957014e-05, "loss": 0.0293, "step": 63080 }, { "grad_norm": 0.34655001759529114, "learning_rate": 3.284703888454781e-05, "loss": 0.0271, "step": 63090 }, { "grad_norm": 0.5328788757324219, "learning_rate": 3.2831508552960646e-05, "loss": 0.0341, "step": 63100 }, { "grad_norm": 0.2546689212322235, "learning_rate": 3.281598009889387e-05, "loss": 0.023, "step": 63110 }, { "grad_norm": 0.36871013045310974, "learning_rate": 3.28004535240457e-05, "loss": 0.0449, "step": 63120 }, { "grad_norm": 0.3414349853992462, "learning_rate": 3.278492883011407e-05, "loss": 0.031, "step": 63130 }, { "grad_norm": 0.41580435633659363, "learning_rate": 3.276940601879675e-05, "loss": 0.0239, "step": 63140 }, { "grad_norm": 0.8885312676429749, "learning_rate": 3.275388509179127e-05, "loss": 0.0305, "step": 63150 }, { "grad_norm": 0.3510245084762573, "learning_rate": 3.2738366050794985e-05, "loss": 0.0427, "step": 63160 }, { "grad_norm": 0.32553693652153015, "learning_rate": 3.2722848897505034e-05, "loss": 0.0254, "step": 63170 }, { "grad_norm": 0.47984007000923157, "learning_rate": 3.2707333633618356e-05, "loss": 0.0263, "step": 63180 }, { "grad_norm": 0.25656694173812866, "learning_rate": 3.269182026083167e-05, "loss": 0.0213, "step": 63190 }, { "grad_norm": 0.8688891530036926, "learning_rate": 3.2676308780841516e-05, "loss": 0.0393, "step": 63200 }, { "grad_norm": 0.5093346834182739, "learning_rate": 3.266079919534416e-05, "loss": 0.0274, "step": 63210 }, { "grad_norm": 0.423016220331192, "learning_rate": 3.264529150603575e-05, "loss": 0.029, "step": 63220 }, { "grad_norm": 0.3386208117008209, "learning_rate": 3.262978571461217e-05, "loss": 0.0294, "step": 63230 }, { "grad_norm": 0.39113369584083557, "learning_rate": 3.26142818227691e-05, "loss": 0.0297, "step": 63240 }, { "grad_norm": 0.36220234632492065, "learning_rate": 3.259877983220203e-05, "loss": 0.0317, "step": 63250 }, { "grad_norm": 0.2255190908908844, "learning_rate": 3.2583279744606246e-05, "loss": 0.0385, "step": 63260 }, { "grad_norm": 0.40183642506599426, "learning_rate": 3.25677815616768e-05, "loss": 0.0274, "step": 63270 }, { "grad_norm": 0.4057755172252655, "learning_rate": 3.255228528510856e-05, "loss": 0.0323, "step": 63280 }, { "grad_norm": 0.31672587990760803, "learning_rate": 3.253679091659617e-05, "loss": 0.0285, "step": 63290 }, { "grad_norm": 0.2858211398124695, "learning_rate": 3.252129845783404e-05, "loss": 0.0413, "step": 63300 }, { "grad_norm": 0.4435085356235504, "learning_rate": 3.2505807910516456e-05, "loss": 0.027, "step": 63310 }, { "grad_norm": 0.6157718300819397, "learning_rate": 3.249031927633742e-05, "loss": 0.028, "step": 63320 }, { "grad_norm": 0.35998988151550293, "learning_rate": 3.2474832556990737e-05, "loss": 0.0288, "step": 63330 }, { "grad_norm": 0.6790488362312317, "learning_rate": 3.245934775417002e-05, "loss": 0.032, "step": 63340 }, { "grad_norm": 0.4289059638977051, "learning_rate": 3.244386486956866e-05, "loss": 0.0292, "step": 63350 }, { "grad_norm": 0.2557070553302765, "learning_rate": 3.2428383904879846e-05, "loss": 0.0248, "step": 63360 }, { "grad_norm": 0.4206075668334961, "learning_rate": 3.241290486179652e-05, "loss": 0.0391, "step": 63370 }, { "grad_norm": 0.4364253282546997, "learning_rate": 3.239742774201151e-05, "loss": 0.0347, "step": 63380 }, { "grad_norm": 0.45518869161605835, "learning_rate": 3.238195254721733e-05, "loss": 0.0271, "step": 63390 }, { "grad_norm": 0.3882528841495514, "learning_rate": 3.236647927910633e-05, "loss": 0.0268, "step": 63400 }, { "grad_norm": 1.544481635093689, "learning_rate": 3.2351007939370646e-05, "loss": 0.0281, "step": 63410 }, { "grad_norm": 0.2933535575866699, "learning_rate": 3.2335538529702184e-05, "loss": 0.0236, "step": 63420 }, { "grad_norm": 0.4669279456138611, "learning_rate": 3.232007105179268e-05, "loss": 0.0274, "step": 63430 }, { "grad_norm": 1.1259772777557373, "learning_rate": 3.230460550733363e-05, "loss": 0.0279, "step": 63440 }, { "grad_norm": 0.38876211643218994, "learning_rate": 3.2289141898016315e-05, "loss": 0.031, "step": 63450 }, { "grad_norm": 0.23576703667640686, "learning_rate": 3.227368022553181e-05, "loss": 0.0376, "step": 63460 }, { "grad_norm": 0.5247352123260498, "learning_rate": 3.225822049157099e-05, "loss": 0.0301, "step": 63470 }, { "grad_norm": 0.6208071708679199, "learning_rate": 3.2242762697824494e-05, "loss": 0.0274, "step": 63480 }, { "grad_norm": 0.3777366876602173, "learning_rate": 3.222730684598279e-05, "loss": 0.0367, "step": 63490 }, { "grad_norm": 0.4447297155857086, "learning_rate": 3.221185293773605e-05, "loss": 0.0302, "step": 63500 }, { "grad_norm": 0.4451015591621399, "learning_rate": 3.2196400974774354e-05, "loss": 0.0257, "step": 63510 }, { "grad_norm": 0.49671390652656555, "learning_rate": 3.2180950958787485e-05, "loss": 0.0365, "step": 63520 }, { "grad_norm": 0.29696571826934814, "learning_rate": 3.2165502891465024e-05, "loss": 0.0394, "step": 63530 }, { "grad_norm": 0.3667440116405487, "learning_rate": 3.215005677449636e-05, "loss": 0.0239, "step": 63540 }, { "grad_norm": 0.2777314782142639, "learning_rate": 3.2134612609570644e-05, "loss": 0.0288, "step": 63550 }, { "grad_norm": 0.28219857811927795, "learning_rate": 3.211917039837682e-05, "loss": 0.0266, "step": 63560 }, { "grad_norm": 0.5341177582740784, "learning_rate": 3.210373014260365e-05, "loss": 0.0332, "step": 63570 }, { "grad_norm": 0.3341703712940216, "learning_rate": 3.208829184393965e-05, "loss": 0.0305, "step": 63580 }, { "grad_norm": 0.3692561089992523, "learning_rate": 3.207285550407312e-05, "loss": 0.0455, "step": 63590 }, { "grad_norm": 0.45449915528297424, "learning_rate": 3.205742112469217e-05, "loss": 0.0349, "step": 63600 }, { "grad_norm": 0.49289175868034363, "learning_rate": 3.204198870748465e-05, "loss": 0.0419, "step": 63610 }, { "grad_norm": 0.35901573300361633, "learning_rate": 3.202655825413824e-05, "loss": 0.0338, "step": 63620 }, { "grad_norm": 0.6516342163085938, "learning_rate": 3.2011129766340416e-05, "loss": 0.0408, "step": 63630 }, { "grad_norm": 0.3005634546279907, "learning_rate": 3.1995703245778384e-05, "loss": 0.0278, "step": 63640 }, { "grad_norm": 0.42877116799354553, "learning_rate": 3.1980278694139186e-05, "loss": 0.0326, "step": 63650 }, { "grad_norm": 0.41841062903404236, "learning_rate": 3.196485611310962e-05, "loss": 0.0334, "step": 63660 }, { "grad_norm": 0.3417663276195526, "learning_rate": 3.1949435504376257e-05, "loss": 0.0307, "step": 63670 }, { "grad_norm": 0.5366164445877075, "learning_rate": 3.1934016869625505e-05, "loss": 0.0331, "step": 63680 }, { "grad_norm": 0.42472878098487854, "learning_rate": 3.191860021054349e-05, "loss": 0.0378, "step": 63690 }, { "grad_norm": 0.33098217844963074, "learning_rate": 3.190318552881618e-05, "loss": 0.037, "step": 63700 }, { "grad_norm": 0.27448031306266785, "learning_rate": 3.1887772826129285e-05, "loss": 0.0243, "step": 63710 }, { "grad_norm": 0.2604908049106598, "learning_rate": 3.1872362104168335e-05, "loss": 0.0316, "step": 63720 }, { "grad_norm": 0.34124132990837097, "learning_rate": 3.185695336461861e-05, "loss": 0.0229, "step": 63730 }, { "grad_norm": 0.38178789615631104, "learning_rate": 3.184154660916519e-05, "loss": 0.0239, "step": 63740 }, { "grad_norm": 0.31715935468673706, "learning_rate": 3.1826141839492906e-05, "loss": 0.0266, "step": 63750 }, { "grad_norm": 0.5708016157150269, "learning_rate": 3.181073905728644e-05, "loss": 0.0267, "step": 63760 }, { "grad_norm": 0.3373091220855713, "learning_rate": 3.1795338264230214e-05, "loss": 0.0203, "step": 63770 }, { "grad_norm": 0.4875582754611969, "learning_rate": 3.1779939462008426e-05, "loss": 0.0277, "step": 63780 }, { "grad_norm": 0.29615843296051025, "learning_rate": 3.1764542652305066e-05, "loss": 0.0262, "step": 63790 }, { "grad_norm": 0.3663825988769531, "learning_rate": 3.17491478368039e-05, "loss": 0.0273, "step": 63800 }, { "grad_norm": 0.4796983003616333, "learning_rate": 3.173375501718847e-05, "loss": 0.035, "step": 63810 }, { "grad_norm": 0.532669723033905, "learning_rate": 3.171836419514215e-05, "loss": 0.0374, "step": 63820 }, { "grad_norm": 0.3467765748500824, "learning_rate": 3.170297537234803e-05, "loss": 0.0323, "step": 63830 }, { "grad_norm": 0.3383398950099945, "learning_rate": 3.168758855048901e-05, "loss": 0.0277, "step": 63840 }, { "grad_norm": 0.3729619085788727, "learning_rate": 3.1672203731247785e-05, "loss": 0.0286, "step": 63850 }, { "grad_norm": 0.22204801440238953, "learning_rate": 3.16568209163068e-05, "loss": 0.0275, "step": 63860 }, { "grad_norm": 0.28203025460243225, "learning_rate": 3.1641440107348295e-05, "loss": 0.0282, "step": 63870 }, { "grad_norm": 0.3994520306587219, "learning_rate": 3.162606130605429e-05, "loss": 0.0307, "step": 63880 }, { "grad_norm": 0.5559597015380859, "learning_rate": 3.16106845141066e-05, "loss": 0.031, "step": 63890 }, { "grad_norm": 0.39393308758735657, "learning_rate": 3.159530973318682e-05, "loss": 0.0289, "step": 63900 }, { "grad_norm": 0.35233157873153687, "learning_rate": 3.1579936964976286e-05, "loss": 0.0257, "step": 63910 }, { "grad_norm": 0.30925992131233215, "learning_rate": 3.156456621115615e-05, "loss": 0.0243, "step": 63920 }, { "grad_norm": 0.2887282967567444, "learning_rate": 3.154919747340734e-05, "loss": 0.0268, "step": 63930 }, { "grad_norm": 0.40057289600372314, "learning_rate": 3.153383075341054e-05, "loss": 0.0283, "step": 63940 }, { "grad_norm": 0.2843886613845825, "learning_rate": 3.1518466052846264e-05, "loss": 0.0276, "step": 63950 }, { "grad_norm": 0.5691540837287903, "learning_rate": 3.150310337339474e-05, "loss": 0.0299, "step": 63960 }, { "grad_norm": 0.3397000730037689, "learning_rate": 3.148774271673603e-05, "loss": 0.0208, "step": 63970 }, { "grad_norm": 0.3337855637073517, "learning_rate": 3.147238408454995e-05, "loss": 0.0272, "step": 63980 }, { "grad_norm": 0.2982301414012909, "learning_rate": 3.1457027478516086e-05, "loss": 0.0216, "step": 63990 }, { "grad_norm": 0.44334471225738525, "learning_rate": 3.144167290031382e-05, "loss": 0.0292, "step": 64000 }, { "grad_norm": 0.6972126364707947, "learning_rate": 3.142632035162227e-05, "loss": 0.0307, "step": 64010 }, { "grad_norm": 0.5232101082801819, "learning_rate": 3.141096983412044e-05, "loss": 0.0266, "step": 64020 }, { "grad_norm": 0.31957104802131653, "learning_rate": 3.139562134948699e-05, "loss": 0.024, "step": 64030 }, { "grad_norm": 0.2518694996833801, "learning_rate": 3.1380274899400424e-05, "loss": 0.0282, "step": 64040 }, { "grad_norm": 0.5854881405830383, "learning_rate": 3.1364930485539e-05, "loss": 0.0219, "step": 64050 }, { "grad_norm": 0.2863861620426178, "learning_rate": 3.134958810958076e-05, "loss": 0.0254, "step": 64060 }, { "grad_norm": 0.35753607749938965, "learning_rate": 3.13342477732035e-05, "loss": 0.0275, "step": 64070 }, { "grad_norm": 0.3137103021144867, "learning_rate": 3.131890947808487e-05, "loss": 0.0295, "step": 64080 }, { "grad_norm": 0.33366331458091736, "learning_rate": 3.13035732259022e-05, "loss": 0.0248, "step": 64090 }, { "grad_norm": 0.29336798191070557, "learning_rate": 3.1288239018332656e-05, "loss": 0.0267, "step": 64100 }, { "grad_norm": 0.3201591968536377, "learning_rate": 3.1272906857053164e-05, "loss": 0.031, "step": 64110 }, { "grad_norm": 0.2982841730117798, "learning_rate": 3.125757674374042e-05, "loss": 0.0257, "step": 64120 }, { "grad_norm": 0.4105183184146881, "learning_rate": 3.124224868007088e-05, "loss": 0.0273, "step": 64130 }, { "grad_norm": 0.32041993737220764, "learning_rate": 3.1226922667720854e-05, "loss": 0.0256, "step": 64140 }, { "grad_norm": 0.3672653138637543, "learning_rate": 3.121159870836633e-05, "loss": 0.0313, "step": 64150 }, { "grad_norm": 0.371500164270401, "learning_rate": 3.119627680368312e-05, "loss": 0.0241, "step": 64160 }, { "grad_norm": 0.3252183496952057, "learning_rate": 3.1180956955346816e-05, "loss": 0.0276, "step": 64170 }, { "grad_norm": 0.37044522166252136, "learning_rate": 3.116563916503276e-05, "loss": 0.0317, "step": 64180 }, { "grad_norm": 0.4825764298439026, "learning_rate": 3.1150323434416095e-05, "loss": 0.0296, "step": 64190 }, { "grad_norm": 0.5933237671852112, "learning_rate": 3.1135009765171696e-05, "loss": 0.0307, "step": 64200 }, { "grad_norm": 2.6666054725646973, "learning_rate": 3.111969815897429e-05, "loss": 0.0264, "step": 64210 }, { "grad_norm": 0.2653680145740509, "learning_rate": 3.1104388617498295e-05, "loss": 0.034, "step": 64220 }, { "grad_norm": 0.3475066125392914, "learning_rate": 3.1089081142417966e-05, "loss": 0.0315, "step": 64230 }, { "grad_norm": 0.2564437985420227, "learning_rate": 3.1073775735407285e-05, "loss": 0.0266, "step": 64240 }, { "grad_norm": 0.31189262866973877, "learning_rate": 3.1058472398140035e-05, "loss": 0.0264, "step": 64250 }, { "grad_norm": 0.3300604224205017, "learning_rate": 3.1043171132289736e-05, "loss": 0.0248, "step": 64260 }, { "grad_norm": 0.314653217792511, "learning_rate": 3.1027871939529774e-05, "loss": 0.0276, "step": 64270 }, { "grad_norm": 0.4542143940925598, "learning_rate": 3.1012574821533214e-05, "loss": 0.043, "step": 64280 }, { "grad_norm": 0.480839341878891, "learning_rate": 3.099727977997291e-05, "loss": 0.0219, "step": 64290 }, { "grad_norm": 0.3836398720741272, "learning_rate": 3.098198681652154e-05, "loss": 0.0287, "step": 64300 }, { "grad_norm": 0.2962093949317932, "learning_rate": 3.0966695932851475e-05, "loss": 0.0207, "step": 64310 }, { "grad_norm": 0.3558529317378998, "learning_rate": 3.0951407130634926e-05, "loss": 0.0342, "step": 64320 }, { "grad_norm": 0.3887556493282318, "learning_rate": 3.0936120411543864e-05, "loss": 0.029, "step": 64330 }, { "grad_norm": 0.39978110790252686, "learning_rate": 3.0920835777250015e-05, "loss": 0.0267, "step": 64340 }, { "grad_norm": 0.43677815794944763, "learning_rate": 3.090555322942488e-05, "loss": 0.0216, "step": 64350 }, { "grad_norm": 0.35455575585365295, "learning_rate": 3.089027276973973e-05, "loss": 0.0317, "step": 64360 }, { "grad_norm": 0.3121426999568939, "learning_rate": 3.0874994399865615e-05, "loss": 0.0301, "step": 64370 }, { "grad_norm": 0.39773812890052795, "learning_rate": 3.085971812147337e-05, "loss": 0.0438, "step": 64380 }, { "grad_norm": 0.30066511034965515, "learning_rate": 3.0844443936233557e-05, "loss": 0.029, "step": 64390 }, { "grad_norm": 0.2639737129211426, "learning_rate": 3.082917184581657e-05, "loss": 0.025, "step": 64400 }, { "grad_norm": 0.6028568744659424, "learning_rate": 3.0813901851892513e-05, "loss": 0.0237, "step": 64410 }, { "grad_norm": 0.395469605922699, "learning_rate": 3.0798633956131316e-05, "loss": 0.0349, "step": 64420 }, { "grad_norm": 0.39148420095443726, "learning_rate": 3.078336816020264e-05, "loss": 0.0314, "step": 64430 }, { "grad_norm": 0.3794926106929779, "learning_rate": 3.0768104465775935e-05, "loss": 0.0267, "step": 64440 }, { "grad_norm": 0.35783082246780396, "learning_rate": 3.0752842874520396e-05, "loss": 0.0329, "step": 64450 }, { "grad_norm": 0.3280891478061676, "learning_rate": 3.073758338810503e-05, "loss": 0.0258, "step": 64460 }, { "grad_norm": 0.38675251603126526, "learning_rate": 3.072232600819859e-05, "loss": 0.0273, "step": 64470 }, { "grad_norm": 2.072543144226074, "learning_rate": 3.070707073646959e-05, "loss": 0.0275, "step": 64480 }, { "grad_norm": 0.2782938778400421, "learning_rate": 3.069181757458633e-05, "loss": 0.0242, "step": 64490 }, { "grad_norm": 0.5815706849098206, "learning_rate": 3.067656652421687e-05, "loss": 0.0304, "step": 64500 }, { "grad_norm": 0.41547274589538574, "learning_rate": 3.0661317587029025e-05, "loss": 0.0319, "step": 64510 }, { "grad_norm": 0.33382004499435425, "learning_rate": 3.064607076469043e-05, "loss": 0.0333, "step": 64520 }, { "grad_norm": 0.586801290512085, "learning_rate": 3.063082605886845e-05, "loss": 0.0271, "step": 64530 }, { "grad_norm": 0.3342461585998535, "learning_rate": 3.061558347123019e-05, "loss": 0.0413, "step": 64540 }, { "grad_norm": 0.3233102560043335, "learning_rate": 3.060034300344259e-05, "loss": 0.0292, "step": 64550 }, { "grad_norm": 0.6150782108306885, "learning_rate": 3.058510465717231e-05, "loss": 0.0416, "step": 64560 }, { "grad_norm": 0.30277010798454285, "learning_rate": 3.0569868434085804e-05, "loss": 0.0318, "step": 64570 }, { "grad_norm": 0.48008623719215393, "learning_rate": 3.055463433584924e-05, "loss": 0.0301, "step": 64580 }, { "grad_norm": 0.6859458684921265, "learning_rate": 3.0539402364128656e-05, "loss": 0.0267, "step": 64590 }, { "grad_norm": 0.5017361640930176, "learning_rate": 3.052417252058977e-05, "loss": 0.0325, "step": 64600 }, { "grad_norm": 0.3766953647136688, "learning_rate": 3.0508944806898092e-05, "loss": 0.0296, "step": 64610 }, { "grad_norm": 0.61708664894104, "learning_rate": 3.04937192247189e-05, "loss": 0.0332, "step": 64620 }, { "grad_norm": 0.21483300626277924, "learning_rate": 3.047849577571725e-05, "loss": 0.0275, "step": 64630 }, { "grad_norm": 0.45477426052093506, "learning_rate": 3.046327446155795e-05, "loss": 0.0338, "step": 64640 }, { "grad_norm": 0.8563293814659119, "learning_rate": 3.0448055283905587e-05, "loss": 0.0259, "step": 64650 }, { "grad_norm": 0.3686477243900299, "learning_rate": 3.0432838244424494e-05, "loss": 0.0391, "step": 64660 }, { "grad_norm": 0.3801623582839966, "learning_rate": 3.04176233447788e-05, "loss": 0.0272, "step": 64670 }, { "grad_norm": 0.33057504892349243, "learning_rate": 3.040241058663238e-05, "loss": 0.0219, "step": 64680 }, { "grad_norm": 0.3682181239128113, "learning_rate": 3.038719997164887e-05, "loss": 0.0323, "step": 64690 }, { "grad_norm": 0.3271371126174927, "learning_rate": 3.0371991501491683e-05, "loss": 0.0332, "step": 64700 }, { "grad_norm": 0.6577286124229431, "learning_rate": 3.0356785177823975e-05, "loss": 0.0342, "step": 64710 }, { "grad_norm": 0.3470497131347656, "learning_rate": 3.0341581002308726e-05, "loss": 0.0316, "step": 64720 }, { "grad_norm": 1.1396863460540771, "learning_rate": 3.0326378976608627e-05, "loss": 0.0278, "step": 64730 }, { "grad_norm": 0.6853213310241699, "learning_rate": 3.0311179102386146e-05, "loss": 0.0344, "step": 64740 }, { "grad_norm": 0.28255411982536316, "learning_rate": 3.0295981381303513e-05, "loss": 0.0204, "step": 64750 }, { "grad_norm": 0.6631383299827576, "learning_rate": 3.0280785815022727e-05, "loss": 0.0253, "step": 64760 }, { "grad_norm": 0.28618568181991577, "learning_rate": 3.0265592405205544e-05, "loss": 0.0286, "step": 64770 }, { "grad_norm": 0.4779394268989563, "learning_rate": 3.0250401153513517e-05, "loss": 0.0333, "step": 64780 }, { "grad_norm": 0.8668689131736755, "learning_rate": 3.0235212061607936e-05, "loss": 0.0349, "step": 64790 }, { "grad_norm": 0.40490713715553284, "learning_rate": 3.0220025131149842e-05, "loss": 0.0273, "step": 64800 }, { "grad_norm": 1.3654881715774536, "learning_rate": 3.020484036380007e-05, "loss": 0.0262, "step": 64810 }, { "grad_norm": 1.7151118516921997, "learning_rate": 3.0189657761219188e-05, "loss": 0.0287, "step": 64820 }, { "grad_norm": 0.6837132573127747, "learning_rate": 3.017447732506753e-05, "loss": 0.0305, "step": 64830 }, { "grad_norm": 0.3258788287639618, "learning_rate": 3.015929905700524e-05, "loss": 0.0248, "step": 64840 }, { "grad_norm": 0.2024795264005661, "learning_rate": 3.014412295869218e-05, "loss": 0.02, "step": 64850 }, { "grad_norm": 0.31503593921661377, "learning_rate": 3.0128949031787977e-05, "loss": 0.0285, "step": 64860 }, { "grad_norm": 0.38639721274375916, "learning_rate": 3.0113777277952022e-05, "loss": 0.0345, "step": 64870 }, { "grad_norm": 0.41814568638801575, "learning_rate": 3.0098607698843496e-05, "loss": 0.0288, "step": 64880 }, { "grad_norm": 0.27678167819976807, "learning_rate": 3.0083440296121312e-05, "loss": 0.032, "step": 64890 }, { "grad_norm": 1.160821557044983, "learning_rate": 3.0068275071444142e-05, "loss": 0.0304, "step": 64900 }, { "grad_norm": 0.6668110489845276, "learning_rate": 3.0053112026470442e-05, "loss": 0.0401, "step": 64910 }, { "grad_norm": 0.22075103223323822, "learning_rate": 3.0037951162858434e-05, "loss": 0.0264, "step": 64920 }, { "grad_norm": 0.3513757884502411, "learning_rate": 3.002279248226607e-05, "loss": 0.0254, "step": 64930 }, { "grad_norm": 0.47595664858818054, "learning_rate": 3.000763598635109e-05, "loss": 0.0296, "step": 64940 }, { "grad_norm": 0.4254843294620514, "learning_rate": 2.999248167677098e-05, "loss": 0.0295, "step": 64950 }, { "grad_norm": 0.4792054295539856, "learning_rate": 2.9977329555182976e-05, "loss": 0.0351, "step": 64960 }, { "grad_norm": 0.3186979293823242, "learning_rate": 2.9962179623244124e-05, "loss": 0.0249, "step": 64970 }, { "grad_norm": 0.2905786335468292, "learning_rate": 2.9947031882611175e-05, "loss": 0.0334, "step": 64980 }, { "grad_norm": 0.47976154088974, "learning_rate": 2.9931886334940685e-05, "loss": 0.0263, "step": 64990 }, { "grad_norm": 0.29721927642822266, "learning_rate": 2.9916742981888923e-05, "loss": 0.0305, "step": 65000 }, { "grad_norm": 0.341660737991333, "learning_rate": 2.990160182511196e-05, "loss": 0.0348, "step": 65010 }, { "grad_norm": 0.4207155704498291, "learning_rate": 2.9886462866265574e-05, "loss": 0.0218, "step": 65020 }, { "grad_norm": 0.31160974502563477, "learning_rate": 2.9871326107005392e-05, "loss": 0.0283, "step": 65030 }, { "grad_norm": 0.43215465545654297, "learning_rate": 2.9856191548986735e-05, "loss": 0.0293, "step": 65040 }, { "grad_norm": 0.3498002886772156, "learning_rate": 2.9841059193864672e-05, "loss": 0.0252, "step": 65050 }, { "grad_norm": 0.21939292550086975, "learning_rate": 2.982592904329407e-05, "loss": 0.033, "step": 65060 }, { "grad_norm": 0.283344566822052, "learning_rate": 2.981080109892954e-05, "loss": 0.032, "step": 65070 }, { "grad_norm": 0.2154446244239807, "learning_rate": 2.9795675362425434e-05, "loss": 0.0265, "step": 65080 }, { "grad_norm": 0.20880572497844696, "learning_rate": 2.97805518354359e-05, "loss": 0.0286, "step": 65090 }, { "grad_norm": 0.3172016143798828, "learning_rate": 2.9765430519614825e-05, "loss": 0.018, "step": 65100 }, { "grad_norm": 0.4078013002872467, "learning_rate": 2.975031141661584e-05, "loss": 0.0291, "step": 65110 }, { "grad_norm": 0.4825853109359741, "learning_rate": 2.9735194528092347e-05, "loss": 0.0276, "step": 65120 }, { "grad_norm": 0.4202898144721985, "learning_rate": 2.9720079855697525e-05, "loss": 0.037, "step": 65130 }, { "grad_norm": 0.32113614678382874, "learning_rate": 2.970496740108427e-05, "loss": 0.0339, "step": 65140 }, { "grad_norm": 0.37022924423217773, "learning_rate": 2.968985716590527e-05, "loss": 0.0185, "step": 65150 }, { "grad_norm": 0.47459521889686584, "learning_rate": 2.9674749151812953e-05, "loss": 0.0272, "step": 65160 }, { "grad_norm": 0.3124427795410156, "learning_rate": 2.965964336045951e-05, "loss": 0.0257, "step": 65170 }, { "grad_norm": 0.2201365828514099, "learning_rate": 2.9644539793496894e-05, "loss": 0.0285, "step": 65180 }, { "grad_norm": 0.3010048568248749, "learning_rate": 2.9629438452576807e-05, "loss": 0.0292, "step": 65190 }, { "grad_norm": 0.3459872603416443, "learning_rate": 2.9614339339350698e-05, "loss": 0.0212, "step": 65200 }, { "grad_norm": 0.35714974999427795, "learning_rate": 2.959924245546979e-05, "loss": 0.0222, "step": 65210 }, { "grad_norm": 0.3701719045639038, "learning_rate": 2.9584147802585038e-05, "loss": 0.0364, "step": 65220 }, { "grad_norm": 0.39141708612442017, "learning_rate": 2.956905538234721e-05, "loss": 0.03, "step": 65230 }, { "grad_norm": 0.4742485284805298, "learning_rate": 2.9553965196406764e-05, "loss": 0.0284, "step": 65240 }, { "grad_norm": 0.3535346984863281, "learning_rate": 2.9538877246413943e-05, "loss": 0.0294, "step": 65250 }, { "grad_norm": 0.25182437896728516, "learning_rate": 2.952379153401874e-05, "loss": 0.0288, "step": 65260 }, { "grad_norm": 0.4194033145904541, "learning_rate": 2.9508708060870914e-05, "loss": 0.0306, "step": 65270 }, { "grad_norm": 0.3936109244823456, "learning_rate": 2.9493626828619948e-05, "loss": 0.0277, "step": 65280 }, { "grad_norm": 0.597221314907074, "learning_rate": 2.9478547838915137e-05, "loss": 0.0235, "step": 65290 }, { "grad_norm": 0.21324345469474792, "learning_rate": 2.946347109340548e-05, "loss": 0.0237, "step": 65300 }, { "grad_norm": 0.36276715993881226, "learning_rate": 2.944839659373976e-05, "loss": 0.0275, "step": 65310 }, { "grad_norm": 0.807558536529541, "learning_rate": 2.943332434156647e-05, "loss": 0.0268, "step": 65320 }, { "grad_norm": 0.5158652663230896, "learning_rate": 2.9418254338533912e-05, "loss": 0.0374, "step": 65330 }, { "grad_norm": 0.5285084247589111, "learning_rate": 2.940318658629011e-05, "loss": 0.032, "step": 65340 }, { "grad_norm": 0.4470376968383789, "learning_rate": 2.938812108648287e-05, "loss": 0.0248, "step": 65350 }, { "grad_norm": 0.4108945429325104, "learning_rate": 2.937305784075972e-05, "loss": 0.0259, "step": 65360 }, { "grad_norm": 0.6703172922134399, "learning_rate": 2.9357996850767937e-05, "loss": 0.0326, "step": 65370 }, { "grad_norm": 0.4038527309894562, "learning_rate": 2.9342938118154595e-05, "loss": 0.0233, "step": 65380 }, { "grad_norm": 0.37840384244918823, "learning_rate": 2.9327881644566485e-05, "loss": 0.0312, "step": 65390 }, { "grad_norm": 0.3811354637145996, "learning_rate": 2.9312827431650157e-05, "loss": 0.0298, "step": 65400 }, { "grad_norm": 0.20636942982673645, "learning_rate": 2.92977754810519e-05, "loss": 0.0238, "step": 65410 }, { "grad_norm": 0.3689786493778229, "learning_rate": 2.928272579441781e-05, "loss": 0.0256, "step": 65420 }, { "grad_norm": 0.20946073532104492, "learning_rate": 2.9267678373393676e-05, "loss": 0.0264, "step": 65430 }, { "grad_norm": 0.4639252722263336, "learning_rate": 2.925263321962507e-05, "loss": 0.0322, "step": 65440 }, { "grad_norm": 0.47012245655059814, "learning_rate": 2.9237590334757292e-05, "loss": 0.0251, "step": 65450 }, { "grad_norm": 0.8404554128646851, "learning_rate": 2.922254972043543e-05, "loss": 0.0381, "step": 65460 }, { "grad_norm": 0.30284109711647034, "learning_rate": 2.9207511378304265e-05, "loss": 0.0295, "step": 65470 }, { "grad_norm": 0.3503751754760742, "learning_rate": 2.9192475310008428e-05, "loss": 0.0314, "step": 65480 }, { "grad_norm": 0.2932352125644684, "learning_rate": 2.917744151719219e-05, "loss": 0.0242, "step": 65490 }, { "grad_norm": 0.5543427467346191, "learning_rate": 2.916241000149964e-05, "loss": 0.0364, "step": 65500 }, { "grad_norm": 0.3387201726436615, "learning_rate": 2.9147380764574622e-05, "loss": 0.0249, "step": 65510 }, { "grad_norm": 0.3411726951599121, "learning_rate": 2.9132353808060674e-05, "loss": 0.0388, "step": 65520 }, { "grad_norm": 0.3246491551399231, "learning_rate": 2.9117329133601134e-05, "loss": 0.0353, "step": 65530 }, { "grad_norm": 0.30772241950035095, "learning_rate": 2.9102306742839093e-05, "loss": 0.0239, "step": 65540 }, { "grad_norm": 0.2995922863483429, "learning_rate": 2.9087286637417375e-05, "loss": 0.0222, "step": 65550 }, { "grad_norm": 0.3489850163459778, "learning_rate": 2.9072268818978537e-05, "loss": 0.0342, "step": 65560 }, { "grad_norm": 0.6774522662162781, "learning_rate": 2.905725328916493e-05, "loss": 0.0266, "step": 65570 }, { "grad_norm": 0.3884853422641754, "learning_rate": 2.9042240049618608e-05, "loss": 0.0276, "step": 65580 }, { "grad_norm": 0.6545013189315796, "learning_rate": 2.9027229101981413e-05, "loss": 0.0261, "step": 65590 }, { "grad_norm": 0.37452420592308044, "learning_rate": 2.901222044789488e-05, "loss": 0.0286, "step": 65600 }, { "grad_norm": 0.32883524894714355, "learning_rate": 2.89972140890004e-05, "loss": 0.0288, "step": 65610 }, { "grad_norm": 0.43191859126091003, "learning_rate": 2.8982210026938988e-05, "loss": 0.0305, "step": 65620 }, { "grad_norm": 0.8240290880203247, "learning_rate": 2.896720826335151e-05, "loss": 0.0407, "step": 65630 }, { "grad_norm": 0.19427579641342163, "learning_rate": 2.8952208799878493e-05, "loss": 0.0238, "step": 65640 }, { "grad_norm": 0.39617422223091125, "learning_rate": 2.89372116381603e-05, "loss": 0.0281, "step": 65650 }, { "grad_norm": 0.386074960231781, "learning_rate": 2.8922216779836932e-05, "loss": 0.0243, "step": 65660 }, { "grad_norm": 0.3328129053115845, "learning_rate": 2.8907224226548285e-05, "loss": 0.0307, "step": 65670 }, { "grad_norm": 0.428925096988678, "learning_rate": 2.8892233979933868e-05, "loss": 0.0293, "step": 65680 }, { "grad_norm": 0.6700981259346008, "learning_rate": 2.8877246041633026e-05, "loss": 0.0266, "step": 65690 }, { "grad_norm": 0.5386039018630981, "learning_rate": 2.886226041328478e-05, "loss": 0.0263, "step": 65700 }, { "grad_norm": 0.3613501787185669, "learning_rate": 2.8847277096527957e-05, "loss": 0.0325, "step": 65710 }, { "grad_norm": 0.32947197556495667, "learning_rate": 2.8832296093001105e-05, "loss": 0.022, "step": 65720 }, { "grad_norm": 0.6094349026679993, "learning_rate": 2.8817317404342547e-05, "loss": 0.0281, "step": 65730 }, { "grad_norm": 0.28754645586013794, "learning_rate": 2.880234103219029e-05, "loss": 0.0269, "step": 65740 }, { "grad_norm": 0.3541359007358551, "learning_rate": 2.878736697818215e-05, "loss": 0.0262, "step": 65750 }, { "grad_norm": 0.25379064679145813, "learning_rate": 2.877239524395568e-05, "loss": 0.0244, "step": 65760 }, { "grad_norm": 0.44362953305244446, "learning_rate": 2.8757425831148134e-05, "loss": 0.0443, "step": 65770 }, { "grad_norm": 0.29129210114479065, "learning_rate": 2.874245874139658e-05, "loss": 0.031, "step": 65780 }, { "grad_norm": 0.16473174095153809, "learning_rate": 2.8727493976337737e-05, "loss": 0.02, "step": 65790 }, { "grad_norm": 0.3620336353778839, "learning_rate": 2.8712531537608218e-05, "loss": 0.0263, "step": 65800 }, { "grad_norm": 0.34579989314079285, "learning_rate": 2.8697571426844217e-05, "loss": 0.0263, "step": 65810 }, { "grad_norm": 0.9978516101837158, "learning_rate": 2.86826136456818e-05, "loss": 0.0334, "step": 65820 }, { "grad_norm": 0.5181480050086975, "learning_rate": 2.8667658195756685e-05, "loss": 0.0284, "step": 65830 }, { "grad_norm": 0.5085431337356567, "learning_rate": 2.8652705078704412e-05, "loss": 0.0373, "step": 65840 }, { "grad_norm": 1.084030032157898, "learning_rate": 2.863775429616018e-05, "loss": 0.0243, "step": 65850 }, { "grad_norm": 0.262877881526947, "learning_rate": 2.862280584975906e-05, "loss": 0.0246, "step": 65860 }, { "grad_norm": 0.5646742582321167, "learning_rate": 2.8607859741135735e-05, "loss": 0.0287, "step": 65870 }, { "grad_norm": 0.8965082168579102, "learning_rate": 2.8592915971924723e-05, "loss": 0.0403, "step": 65880 }, { "grad_norm": 0.3986814022064209, "learning_rate": 2.8577974543760212e-05, "loss": 0.0284, "step": 65890 }, { "grad_norm": 0.27443957328796387, "learning_rate": 2.8563035458276222e-05, "loss": 0.0205, "step": 65900 }, { "grad_norm": 0.4394734501838684, "learning_rate": 2.8548098717106424e-05, "loss": 0.0345, "step": 65910 }, { "grad_norm": 0.26664119958877563, "learning_rate": 2.85331643218843e-05, "loss": 0.0257, "step": 65920 }, { "grad_norm": 0.2470775991678238, "learning_rate": 2.8518232274243046e-05, "loss": 0.0261, "step": 65930 }, { "grad_norm": 0.5120760798454285, "learning_rate": 2.8503302575815638e-05, "loss": 0.044, "step": 65940 }, { "grad_norm": 0.454702764749527, "learning_rate": 2.8488375228234716e-05, "loss": 0.032, "step": 65950 }, { "grad_norm": 1.0026730298995972, "learning_rate": 2.8473450233132738e-05, "loss": 0.0325, "step": 65960 }, { "grad_norm": 0.3096012473106384, "learning_rate": 2.8458527592141898e-05, "loss": 0.02, "step": 65970 }, { "grad_norm": 0.42380812764167786, "learning_rate": 2.8443607306894073e-05, "loss": 0.0358, "step": 65980 }, { "grad_norm": 0.3491719365119934, "learning_rate": 2.8428689379020946e-05, "loss": 0.0334, "step": 65990 }, { "grad_norm": 0.330304354429245, "learning_rate": 2.841377381015391e-05, "loss": 0.0293, "step": 66000 }, { "grad_norm": 0.33270183205604553, "learning_rate": 2.8398860601924144e-05, "loss": 0.0245, "step": 66010 }, { "grad_norm": 0.4574386179447174, "learning_rate": 2.838394975596248e-05, "loss": 0.0211, "step": 66020 }, { "grad_norm": 0.3012312352657318, "learning_rate": 2.83690412738996e-05, "loss": 0.0311, "step": 66030 }, { "grad_norm": 0.36970844864845276, "learning_rate": 2.835413515736579e-05, "loss": 0.0406, "step": 66040 }, { "grad_norm": 0.3199850022792816, "learning_rate": 2.8339231407991267e-05, "loss": 0.0267, "step": 66050 }, { "grad_norm": 0.25431814789772034, "learning_rate": 2.832433002740581e-05, "loss": 0.0219, "step": 66060 }, { "grad_norm": 0.26833510398864746, "learning_rate": 2.8309431017239047e-05, "loss": 0.0259, "step": 66070 }, { "grad_norm": 0.40810924768447876, "learning_rate": 2.829453437912028e-05, "loss": 0.0279, "step": 66080 }, { "grad_norm": 0.25327327847480774, "learning_rate": 2.8279640114678617e-05, "loss": 0.0253, "step": 66090 }, { "grad_norm": 0.3813358247280121, "learning_rate": 2.8264748225542836e-05, "loss": 0.0274, "step": 66100 }, { "grad_norm": 0.48449939489364624, "learning_rate": 2.8249858713341516e-05, "loss": 0.0279, "step": 66110 }, { "grad_norm": 0.5574881434440613, "learning_rate": 2.823497157970294e-05, "loss": 0.0318, "step": 66120 }, { "grad_norm": 0.3358471989631653, "learning_rate": 2.8220086826255164e-05, "loss": 0.027, "step": 66130 }, { "grad_norm": 0.19090576469898224, "learning_rate": 2.8205204454625934e-05, "loss": 0.0271, "step": 66140 }, { "grad_norm": 0.39498573541641235, "learning_rate": 2.8190324466442798e-05, "loss": 0.025, "step": 66150 }, { "grad_norm": 0.31689730286598206, "learning_rate": 2.8175446863332965e-05, "loss": 0.025, "step": 66160 }, { "grad_norm": 0.3701039254665375, "learning_rate": 2.816057164692345e-05, "loss": 0.0237, "step": 66170 }, { "grad_norm": 0.6639032959938049, "learning_rate": 2.8145698818840992e-05, "loss": 0.0322, "step": 66180 }, { "grad_norm": 0.30893465876579285, "learning_rate": 2.8130828380712065e-05, "loss": 0.024, "step": 66190 }, { "grad_norm": 0.46200916171073914, "learning_rate": 2.811596033416285e-05, "loss": 0.0253, "step": 66200 }, { "grad_norm": 0.2438153773546219, "learning_rate": 2.8101094680819307e-05, "loss": 0.0274, "step": 66210 }, { "grad_norm": 0.31090402603149414, "learning_rate": 2.808623142230715e-05, "loss": 0.0393, "step": 66220 }, { "grad_norm": 0.4862464964389801, "learning_rate": 2.8071370560251757e-05, "loss": 0.026, "step": 66230 }, { "grad_norm": 0.4870448112487793, "learning_rate": 2.805651209627831e-05, "loss": 0.0243, "step": 66240 }, { "grad_norm": 0.41407227516174316, "learning_rate": 2.804165603201171e-05, "loss": 0.023, "step": 66250 }, { "grad_norm": 0.49976837635040283, "learning_rate": 2.802680236907661e-05, "loss": 0.0283, "step": 66260 }, { "grad_norm": 0.3707749545574188, "learning_rate": 2.8011951109097344e-05, "loss": 0.0256, "step": 66270 }, { "grad_norm": 0.4869915246963501, "learning_rate": 2.7997102253698072e-05, "loss": 0.0279, "step": 66280 }, { "grad_norm": 0.31895002722740173, "learning_rate": 2.7982255804502588e-05, "loss": 0.0215, "step": 66290 }, { "grad_norm": 0.3016749322414398, "learning_rate": 2.7967411763134516e-05, "loss": 0.0274, "step": 66300 }, { "grad_norm": 0.4328387975692749, "learning_rate": 2.7952570131217166e-05, "loss": 0.0278, "step": 66310 }, { "grad_norm": 0.44271761178970337, "learning_rate": 2.7937730910373618e-05, "loss": 0.0288, "step": 66320 }, { "grad_norm": 0.35212138295173645, "learning_rate": 2.7922894102226628e-05, "loss": 0.0216, "step": 66330 }, { "grad_norm": 0.3479640483856201, "learning_rate": 2.790805970839876e-05, "loss": 0.0276, "step": 66340 }, { "grad_norm": 0.3691343665122986, "learning_rate": 2.789322773051225e-05, "loss": 0.027, "step": 66350 }, { "grad_norm": 0.4306473731994629, "learning_rate": 2.7878398170189124e-05, "loss": 0.0247, "step": 66360 }, { "grad_norm": 0.28790268301963806, "learning_rate": 2.7863571029051112e-05, "loss": 0.0236, "step": 66370 }, { "grad_norm": 0.47671833634376526, "learning_rate": 2.784874630871971e-05, "loss": 0.0313, "step": 66380 }, { "grad_norm": 0.35081160068511963, "learning_rate": 2.7833924010816086e-05, "loss": 0.022, "step": 66390 }, { "grad_norm": 0.29921793937683105, "learning_rate": 2.7819104136961226e-05, "loss": 0.0301, "step": 66400 }, { "grad_norm": 0.25562214851379395, "learning_rate": 2.780428668877577e-05, "loss": 0.0207, "step": 66410 }, { "grad_norm": 0.6610012650489807, "learning_rate": 2.7789471667880152e-05, "loss": 0.0405, "step": 66420 }, { "grad_norm": 0.3055894672870636, "learning_rate": 2.7774659075894523e-05, "loss": 0.0354, "step": 66430 }, { "grad_norm": 0.46280527114868164, "learning_rate": 2.7759848914438778e-05, "loss": 0.029, "step": 66440 }, { "grad_norm": 0.7494480609893799, "learning_rate": 2.7745041185132496e-05, "loss": 0.022, "step": 66450 }, { "grad_norm": 0.43471595644950867, "learning_rate": 2.773023588959506e-05, "loss": 0.0282, "step": 66460 }, { "grad_norm": 0.2671894133090973, "learning_rate": 2.7715433029445555e-05, "loss": 0.0296, "step": 66470 }, { "grad_norm": 0.2637825906276703, "learning_rate": 2.7700632606302766e-05, "loss": 0.0239, "step": 66480 }, { "grad_norm": 0.5395865440368652, "learning_rate": 2.768583462178528e-05, "loss": 0.0296, "step": 66490 }, { "grad_norm": 0.6129273772239685, "learning_rate": 2.767103907751136e-05, "loss": 0.0215, "step": 66500 }, { "grad_norm": 0.8564392924308777, "learning_rate": 2.7656245975099065e-05, "loss": 0.0228, "step": 66510 }, { "grad_norm": 0.3473258316516876, "learning_rate": 2.7641455316166093e-05, "loss": 0.0298, "step": 66520 }, { "grad_norm": 0.30157554149627686, "learning_rate": 2.7626667102329973e-05, "loss": 0.0308, "step": 66530 }, { "grad_norm": 0.3002694547176361, "learning_rate": 2.761188133520788e-05, "loss": 0.0267, "step": 66540 }, { "grad_norm": 0.45535704493522644, "learning_rate": 2.7597098016416784e-05, "loss": 0.0283, "step": 66550 }, { "grad_norm": 0.37994104623794556, "learning_rate": 2.7582317147573368e-05, "loss": 0.0169, "step": 66560 }, { "grad_norm": 0.5752457976341248, "learning_rate": 2.7567538730294064e-05, "loss": 0.0379, "step": 66570 }, { "grad_norm": 0.340285062789917, "learning_rate": 2.7552762766194973e-05, "loss": 0.0337, "step": 66580 }, { "grad_norm": 0.3048340380191803, "learning_rate": 2.753798925689202e-05, "loss": 0.0354, "step": 66590 }, { "grad_norm": 0.4275359511375427, "learning_rate": 2.7523218204000774e-05, "loss": 0.0328, "step": 66600 }, { "grad_norm": 0.519482433795929, "learning_rate": 2.7508449609136612e-05, "loss": 0.0375, "step": 66610 }, { "grad_norm": 0.3070300221443176, "learning_rate": 2.749368347391454e-05, "loss": 0.0247, "step": 66620 }, { "grad_norm": 0.24573659896850586, "learning_rate": 2.7478919799949443e-05, "loss": 0.0307, "step": 66630 }, { "grad_norm": 0.3534572720527649, "learning_rate": 2.7464158588855803e-05, "loss": 0.0337, "step": 66640 }, { "grad_norm": 0.38742130994796753, "learning_rate": 2.7449399842247918e-05, "loss": 0.0325, "step": 66650 }, { "grad_norm": 0.48776230216026306, "learning_rate": 2.7434643561739736e-05, "loss": 0.0369, "step": 66660 }, { "grad_norm": 0.3529651463031769, "learning_rate": 2.7419889748945016e-05, "loss": 0.0282, "step": 66670 }, { "grad_norm": 0.2787994146347046, "learning_rate": 2.7405138405477203e-05, "loss": 0.0264, "step": 66680 }, { "grad_norm": 0.2844304144382477, "learning_rate": 2.73903895329495e-05, "loss": 0.0297, "step": 66690 }, { "grad_norm": 0.43284928798675537, "learning_rate": 2.737564313297478e-05, "loss": 0.0245, "step": 66700 }, { "grad_norm": 0.2459564208984375, "learning_rate": 2.736089920716572e-05, "loss": 0.0267, "step": 66710 }, { "grad_norm": 0.3359212875366211, "learning_rate": 2.7346157757134706e-05, "loss": 0.0255, "step": 66720 }, { "grad_norm": 0.28864601254463196, "learning_rate": 2.733141878449379e-05, "loss": 0.0168, "step": 66730 }, { "grad_norm": 0.17328980565071106, "learning_rate": 2.7316682290854844e-05, "loss": 0.0238, "step": 66740 }, { "grad_norm": 0.5199764966964722, "learning_rate": 2.7301948277829414e-05, "loss": 0.0364, "step": 66750 }, { "grad_norm": 0.3416590988636017, "learning_rate": 2.7287216747028814e-05, "loss": 0.0246, "step": 66760 }, { "grad_norm": 0.38633453845977783, "learning_rate": 2.7272487700064025e-05, "loss": 0.026, "step": 66770 }, { "grad_norm": 0.2536408007144928, "learning_rate": 2.725776113854583e-05, "loss": 0.0337, "step": 66780 }, { "grad_norm": 0.7267934083938599, "learning_rate": 2.7243037064084664e-05, "loss": 0.0327, "step": 66790 }, { "grad_norm": 0.5498008728027344, "learning_rate": 2.7228315478290766e-05, "loss": 0.0389, "step": 66800 }, { "grad_norm": 0.30080971121788025, "learning_rate": 2.721359638277402e-05, "loss": 0.0319, "step": 66810 }, { "grad_norm": 0.3593350946903229, "learning_rate": 2.7198879779144148e-05, "loss": 0.0287, "step": 66820 }, { "grad_norm": 0.3185965120792389, "learning_rate": 2.7184165669010485e-05, "loss": 0.0296, "step": 66830 }, { "grad_norm": 0.5022891163825989, "learning_rate": 2.716945405398218e-05, "loss": 0.0306, "step": 66840 }, { "grad_norm": 0.4022381603717804, "learning_rate": 2.715474493566803e-05, "loss": 0.0292, "step": 66850 }, { "grad_norm": 0.2916201949119568, "learning_rate": 2.7140038315676652e-05, "loss": 0.0247, "step": 66860 }, { "grad_norm": 0.3678755760192871, "learning_rate": 2.7125334195616275e-05, "loss": 0.0424, "step": 66870 }, { "grad_norm": 0.4349450170993805, "learning_rate": 2.7110632577094997e-05, "loss": 0.0298, "step": 66880 }, { "grad_norm": 0.2702813148498535, "learning_rate": 2.709593346172051e-05, "loss": 0.0317, "step": 66890 }, { "grad_norm": 0.26315826177597046, "learning_rate": 2.708123685110032e-05, "loss": 0.024, "step": 66900 }, { "grad_norm": 0.6398231983184814, "learning_rate": 2.7066542746841588e-05, "loss": 0.0282, "step": 66910 }, { "grad_norm": 0.29880380630493164, "learning_rate": 2.7051851150551267e-05, "loss": 0.0307, "step": 66920 }, { "grad_norm": 0.30711090564727783, "learning_rate": 2.7037162063835996e-05, "loss": 0.0256, "step": 66930 }, { "grad_norm": 0.5407742857933044, "learning_rate": 2.702247548830218e-05, "loss": 0.0221, "step": 66940 }, { "grad_norm": 0.2903349697589874, "learning_rate": 2.7007791425555883e-05, "loss": 0.0218, "step": 66950 }, { "grad_norm": 0.40774989128112793, "learning_rate": 2.6993109877202945e-05, "loss": 0.025, "step": 66960 }, { "grad_norm": 1.040334939956665, "learning_rate": 2.6978430844848935e-05, "loss": 0.02, "step": 66970 }, { "grad_norm": 0.41102340817451477, "learning_rate": 2.69637543300991e-05, "loss": 0.026, "step": 66980 }, { "grad_norm": 0.29057738184928894, "learning_rate": 2.694908033455848e-05, "loss": 0.0196, "step": 66990 }, { "grad_norm": 0.281472384929657, "learning_rate": 2.6934408859831733e-05, "loss": 0.0289, "step": 67000 }, { "grad_norm": 0.2993474304676056, "learning_rate": 2.69197399075234e-05, "loss": 0.0313, "step": 67010 }, { "grad_norm": 0.3395955264568329, "learning_rate": 2.6905073479237584e-05, "loss": 0.0272, "step": 67020 }, { "grad_norm": 0.2743203341960907, "learning_rate": 2.6890409576578235e-05, "loss": 0.0291, "step": 67030 }, { "grad_norm": 0.334563285112381, "learning_rate": 2.687574820114892e-05, "loss": 0.0231, "step": 67040 }, { "grad_norm": 0.4686749279499054, "learning_rate": 2.686108935455305e-05, "loss": 0.0343, "step": 67050 }, { "grad_norm": 0.3975426256656647, "learning_rate": 2.6846433038393603e-05, "loss": 0.0289, "step": 67060 }, { "grad_norm": 0.2609754502773285, "learning_rate": 2.6831779254273482e-05, "loss": 0.0209, "step": 67070 }, { "grad_norm": 0.4210594594478607, "learning_rate": 2.6817128003795122e-05, "loss": 0.0216, "step": 67080 }, { "grad_norm": 0.27360019087791443, "learning_rate": 2.680247928856081e-05, "loss": 0.0187, "step": 67090 }, { "grad_norm": 0.42413783073425293, "learning_rate": 2.6787833110172466e-05, "loss": 0.0221, "step": 67100 }, { "grad_norm": 0.24990050494670868, "learning_rate": 2.6773189470231807e-05, "loss": 0.0272, "step": 67110 }, { "grad_norm": 0.2673502266407013, "learning_rate": 2.675854837034022e-05, "loss": 0.0221, "step": 67120 }, { "grad_norm": 0.24376079440116882, "learning_rate": 2.6743909812098833e-05, "loss": 0.026, "step": 67130 }, { "grad_norm": 0.37445512413978577, "learning_rate": 2.67292737971085e-05, "loss": 0.034, "step": 67140 }, { "grad_norm": 0.2768924832344055, "learning_rate": 2.671464032696982e-05, "loss": 0.0195, "step": 67150 }, { "grad_norm": 0.295075923204422, "learning_rate": 2.6700009403283055e-05, "loss": 0.0248, "step": 67160 }, { "grad_norm": 0.37315142154693604, "learning_rate": 2.6685381027648214e-05, "loss": 0.0302, "step": 67170 }, { "grad_norm": 0.38079914450645447, "learning_rate": 2.6670755201665086e-05, "loss": 0.0381, "step": 67180 }, { "grad_norm": 0.32988855242729187, "learning_rate": 2.6656131926933066e-05, "loss": 0.0235, "step": 67190 }, { "grad_norm": 0.35721006989479065, "learning_rate": 2.664151120505136e-05, "loss": 0.0216, "step": 67200 }, { "grad_norm": 0.7711405158042908, "learning_rate": 2.6626893037618873e-05, "loss": 0.0316, "step": 67210 }, { "grad_norm": 0.36190658807754517, "learning_rate": 2.6612277426234243e-05, "loss": 0.041, "step": 67220 }, { "grad_norm": 0.3223669230937958, "learning_rate": 2.659766437249577e-05, "loss": 0.0346, "step": 67230 }, { "grad_norm": 0.2640567719936371, "learning_rate": 2.6583053878001557e-05, "loss": 0.0232, "step": 67240 }, { "grad_norm": 0.29466772079467773, "learning_rate": 2.6568445944349323e-05, "loss": 0.0269, "step": 67250 }, { "grad_norm": 0.3603203296661377, "learning_rate": 2.6553840573136657e-05, "loss": 0.0219, "step": 67260 }, { "grad_norm": 0.7161639928817749, "learning_rate": 2.6539237765960706e-05, "loss": 0.0316, "step": 67270 }, { "grad_norm": 0.23905403912067413, "learning_rate": 2.6524637524418465e-05, "loss": 0.0256, "step": 67280 }, { "grad_norm": 0.1878151297569275, "learning_rate": 2.651003985010655e-05, "loss": 0.0322, "step": 67290 }, { "grad_norm": 0.329926460981369, "learning_rate": 2.649544474462138e-05, "loss": 0.0276, "step": 67300 }, { "grad_norm": 0.3279963433742523, "learning_rate": 2.6480852209559015e-05, "loss": 0.0258, "step": 67310 }, { "grad_norm": 0.45117467641830444, "learning_rate": 2.6466262246515296e-05, "loss": 0.0338, "step": 67320 }, { "grad_norm": 0.2495144158601761, "learning_rate": 2.6451674857085746e-05, "loss": 0.0313, "step": 67330 }, { "grad_norm": 0.47376930713653564, "learning_rate": 2.6437090042865655e-05, "loss": 0.0347, "step": 67340 }, { "grad_norm": 0.48275530338287354, "learning_rate": 2.6422507805449947e-05, "loss": 0.0232, "step": 67350 }, { "grad_norm": 0.22813349962234497, "learning_rate": 2.6407928146433357e-05, "loss": 0.0319, "step": 67360 }, { "grad_norm": 0.7356252670288086, "learning_rate": 2.639335106741026e-05, "loss": 0.0259, "step": 67370 }, { "grad_norm": 0.28706055879592896, "learning_rate": 2.6378776569974796e-05, "loss": 0.0296, "step": 67380 }, { "grad_norm": 0.46990692615509033, "learning_rate": 2.6364204655720814e-05, "loss": 0.0205, "step": 67390 }, { "grad_norm": 0.33353573083877563, "learning_rate": 2.6349635326241894e-05, "loss": 0.0325, "step": 67400 }, { "grad_norm": 0.3516506254673004, "learning_rate": 2.633506858313127e-05, "loss": 0.0255, "step": 67410 }, { "grad_norm": 0.26285073161125183, "learning_rate": 2.6320504427981975e-05, "loss": 0.0377, "step": 67420 }, { "grad_norm": 0.5313519835472107, "learning_rate": 2.6305942862386734e-05, "loss": 0.0193, "step": 67430 }, { "grad_norm": 0.3966064155101776, "learning_rate": 2.629138388793793e-05, "loss": 0.0293, "step": 67440 }, { "grad_norm": 0.37407487630844116, "learning_rate": 2.6276827506227743e-05, "loss": 0.0306, "step": 67450 }, { "grad_norm": 0.32773932814598083, "learning_rate": 2.6262273718848037e-05, "loss": 0.0322, "step": 67460 }, { "grad_norm": 0.44961607456207275, "learning_rate": 2.6247722527390407e-05, "loss": 0.0277, "step": 67470 }, { "grad_norm": 0.2698918581008911, "learning_rate": 2.6233173933446104e-05, "loss": 0.0229, "step": 67480 }, { "grad_norm": 0.21941520273685455, "learning_rate": 2.6218627938606188e-05, "loss": 0.0356, "step": 67490 }, { "grad_norm": 0.4625452160835266, "learning_rate": 2.620408454446135e-05, "loss": 0.0226, "step": 67500 }, { "grad_norm": 0.29780349135398865, "learning_rate": 2.6189543752602046e-05, "loss": 0.0295, "step": 67510 }, { "grad_norm": 0.4570877254009247, "learning_rate": 2.6175005564618444e-05, "loss": 0.0222, "step": 67520 }, { "grad_norm": 0.35755741596221924, "learning_rate": 2.6160469982100428e-05, "loss": 0.0253, "step": 67530 }, { "grad_norm": 1.5026919841766357, "learning_rate": 2.6145937006637555e-05, "loss": 0.0279, "step": 67540 }, { "grad_norm": 0.6789805293083191, "learning_rate": 2.6131406639819166e-05, "loss": 0.0249, "step": 67550 }, { "grad_norm": 0.1777317076921463, "learning_rate": 2.611687888323424e-05, "loss": 0.0236, "step": 67560 }, { "grad_norm": 0.5529653429985046, "learning_rate": 2.610235373847153e-05, "loss": 0.0347, "step": 67570 }, { "grad_norm": 0.5925436019897461, "learning_rate": 2.6087831207119494e-05, "loss": 0.0429, "step": 67580 }, { "grad_norm": 0.1882578283548355, "learning_rate": 2.60733112907663e-05, "loss": 0.0384, "step": 67590 }, { "grad_norm": 0.29585596919059753, "learning_rate": 2.6058793990999786e-05, "loss": 0.0303, "step": 67600 }, { "grad_norm": 0.43811020255088806, "learning_rate": 2.6044279309407593e-05, "loss": 0.0258, "step": 67610 }, { "grad_norm": 0.40152642130851746, "learning_rate": 2.6029767247576976e-05, "loss": 0.027, "step": 67620 }, { "grad_norm": 0.33143186569213867, "learning_rate": 2.601525780709497e-05, "loss": 0.0266, "step": 67630 }, { "grad_norm": 0.32486867904663086, "learning_rate": 2.6000750989548318e-05, "loss": 0.0271, "step": 67640 }, { "grad_norm": 0.41877710819244385, "learning_rate": 2.5986246796523462e-05, "loss": 0.0245, "step": 67650 }, { "grad_norm": 0.42680999636650085, "learning_rate": 2.5971745229606546e-05, "loss": 0.0248, "step": 67660 }, { "grad_norm": 0.40727198123931885, "learning_rate": 2.5957246290383442e-05, "loss": 0.0304, "step": 67670 }, { "grad_norm": 0.3512989580631256, "learning_rate": 2.594274998043975e-05, "loss": 0.0252, "step": 67680 }, { "grad_norm": 0.8291856050491333, "learning_rate": 2.5928256301360733e-05, "loss": 0.0253, "step": 67690 }, { "grad_norm": 0.3658364415168762, "learning_rate": 2.591376525473142e-05, "loss": 0.0229, "step": 67700 }, { "grad_norm": 0.34031808376312256, "learning_rate": 2.5899276842136523e-05, "loss": 0.0261, "step": 67710 }, { "grad_norm": 0.4637172520160675, "learning_rate": 2.5884791065160495e-05, "loss": 0.0341, "step": 67720 }, { "grad_norm": 0.42101001739501953, "learning_rate": 2.5870307925387436e-05, "loss": 0.0324, "step": 67730 }, { "grad_norm": 0.3887019455432892, "learning_rate": 2.585582742440125e-05, "loss": 0.0311, "step": 67740 }, { "grad_norm": 0.3139304220676422, "learning_rate": 2.5841349563785455e-05, "loss": 0.0241, "step": 67750 }, { "grad_norm": 0.24058085680007935, "learning_rate": 2.5826874345123355e-05, "loss": 0.0329, "step": 67760 }, { "grad_norm": 0.4110965430736542, "learning_rate": 2.581240176999792e-05, "loss": 0.0264, "step": 67770 }, { "grad_norm": 0.5212697386741638, "learning_rate": 2.579793183999189e-05, "loss": 0.0317, "step": 67780 }, { "grad_norm": 0.35744160413742065, "learning_rate": 2.5783464556687627e-05, "loss": 0.0299, "step": 67790 }, { "grad_norm": 0.33823296427726746, "learning_rate": 2.576899992166729e-05, "loss": 0.022, "step": 67800 }, { "grad_norm": 0.21684782207012177, "learning_rate": 2.5754537936512668e-05, "loss": 0.0316, "step": 67810 }, { "grad_norm": 0.48200473189353943, "learning_rate": 2.574007860280535e-05, "loss": 0.0343, "step": 67820 }, { "grad_norm": 0.2802388072013855, "learning_rate": 2.5725621922126518e-05, "loss": 0.0243, "step": 67830 }, { "grad_norm": 0.4362418055534363, "learning_rate": 2.571116789605722e-05, "loss": 0.0209, "step": 67840 }, { "grad_norm": 0.48893171548843384, "learning_rate": 2.5696716526178054e-05, "loss": 0.0316, "step": 67850 }, { "grad_norm": 0.4215737581253052, "learning_rate": 2.568226781406946e-05, "loss": 0.0393, "step": 67860 }, { "grad_norm": 0.27636101841926575, "learning_rate": 2.5667821761311474e-05, "loss": 0.0292, "step": 67870 }, { "grad_norm": 0.6061708927154541, "learning_rate": 2.5653378369483917e-05, "loss": 0.0275, "step": 67880 }, { "grad_norm": 0.5673897862434387, "learning_rate": 2.5638937640166292e-05, "loss": 0.0338, "step": 67890 }, { "grad_norm": 0.2872399389743805, "learning_rate": 2.5624499574937842e-05, "loss": 0.0268, "step": 67900 }, { "grad_norm": 0.41301587224006653, "learning_rate": 2.5610064175377456e-05, "loss": 0.0305, "step": 67910 }, { "grad_norm": 0.48144736886024475, "learning_rate": 2.5595631443063782e-05, "loss": 0.0211, "step": 67920 }, { "grad_norm": 0.25021466612815857, "learning_rate": 2.5581201379575177e-05, "loss": 0.0249, "step": 67930 }, { "grad_norm": 0.32004061341285706, "learning_rate": 2.5566773986489655e-05, "loss": 0.0246, "step": 67940 }, { "grad_norm": 0.3099970519542694, "learning_rate": 2.5552349265385e-05, "loss": 0.0373, "step": 67950 }, { "grad_norm": 0.45527228713035583, "learning_rate": 2.5537927217838675e-05, "loss": 0.033, "step": 67960 }, { "grad_norm": 0.3201799988746643, "learning_rate": 2.5523507845427864e-05, "loss": 0.0295, "step": 67970 }, { "grad_norm": 0.4848775565624237, "learning_rate": 2.5509091149729413e-05, "loss": 0.0319, "step": 67980 }, { "grad_norm": 0.3778751790523529, "learning_rate": 2.5494677132319962e-05, "loss": 0.0326, "step": 67990 }, { "grad_norm": 0.35653984546661377, "learning_rate": 2.548026579477575e-05, "loss": 0.0385, "step": 68000 }, { "grad_norm": 0.36899322271347046, "learning_rate": 2.5465857138672834e-05, "loss": 0.0365, "step": 68010 }, { "grad_norm": 0.37848523259162903, "learning_rate": 2.5451451165586848e-05, "loss": 0.0324, "step": 68020 }, { "grad_norm": 0.26408734917640686, "learning_rate": 2.543704787709329e-05, "loss": 0.0213, "step": 68030 }, { "grad_norm": 0.40080034732818604, "learning_rate": 2.542264727476723e-05, "loss": 0.0253, "step": 68040 }, { "grad_norm": 0.354979008436203, "learning_rate": 2.5408249360183535e-05, "loss": 0.0396, "step": 68050 }, { "grad_norm": 0.21471644937992096, "learning_rate": 2.5393854134916694e-05, "loss": 0.0273, "step": 68060 }, { "grad_norm": 0.4463709592819214, "learning_rate": 2.5379461600540993e-05, "loss": 0.0323, "step": 68070 }, { "grad_norm": 0.2967548668384552, "learning_rate": 2.5365071758630317e-05, "loss": 0.0256, "step": 68080 }, { "grad_norm": 0.3892342746257782, "learning_rate": 2.535068461075839e-05, "loss": 0.0351, "step": 68090 }, { "grad_norm": 0.4043869376182556, "learning_rate": 2.5336300158498516e-05, "loss": 0.0326, "step": 68100 }, { "grad_norm": 0.5753110647201538, "learning_rate": 2.5321918403423793e-05, "loss": 0.0403, "step": 68110 }, { "grad_norm": 0.3526677191257477, "learning_rate": 2.5307539347106957e-05, "loss": 0.0244, "step": 68120 }, { "grad_norm": 0.3376169204711914, "learning_rate": 2.529316299112049e-05, "loss": 0.0287, "step": 68130 }, { "grad_norm": 0.367341548204422, "learning_rate": 2.5278789337036567e-05, "loss": 0.0206, "step": 68140 }, { "grad_norm": 0.3739863336086273, "learning_rate": 2.52644183864271e-05, "loss": 0.0243, "step": 68150 }, { "grad_norm": 0.34959444403648376, "learning_rate": 2.525005014086363e-05, "loss": 0.0245, "step": 68160 }, { "grad_norm": 0.515425443649292, "learning_rate": 2.5235684601917465e-05, "loss": 0.0291, "step": 68170 }, { "grad_norm": 1.390454649925232, "learning_rate": 2.5221321771159618e-05, "loss": 0.0256, "step": 68180 }, { "grad_norm": 0.27584174275398254, "learning_rate": 2.5206961650160743e-05, "loss": 0.0336, "step": 68190 }, { "grad_norm": 0.3503170311450958, "learning_rate": 2.519260424049129e-05, "loss": 0.0326, "step": 68200 }, { "grad_norm": 0.3521477282047272, "learning_rate": 2.5178249543721304e-05, "loss": 0.0313, "step": 68210 }, { "grad_norm": 0.20320972800254822, "learning_rate": 2.5163897561420668e-05, "loss": 0.0241, "step": 68220 }, { "grad_norm": 0.2671913206577301, "learning_rate": 2.514954829515883e-05, "loss": 0.0379, "step": 68230 }, { "grad_norm": 0.24537912011146545, "learning_rate": 2.513520174650505e-05, "loss": 0.0309, "step": 68240 }, { "grad_norm": 0.6457130908966064, "learning_rate": 2.5120857917028207e-05, "loss": 0.0279, "step": 68250 }, { "grad_norm": 0.2723018527030945, "learning_rate": 2.5106516808296953e-05, "loss": 0.0229, "step": 68260 }, { "grad_norm": 0.5930660367012024, "learning_rate": 2.5092178421879552e-05, "loss": 0.0286, "step": 68270 }, { "grad_norm": 0.390078067779541, "learning_rate": 2.5077842759344116e-05, "loss": 0.0419, "step": 68280 }, { "grad_norm": 0.3192746639251709, "learning_rate": 2.5063509822258304e-05, "loss": 0.0273, "step": 68290 }, { "grad_norm": 0.5915026068687439, "learning_rate": 2.5049179612189576e-05, "loss": 0.0324, "step": 68300 }, { "grad_norm": 0.3649630844593048, "learning_rate": 2.5034852130705043e-05, "loss": 0.0241, "step": 68310 }, { "grad_norm": 0.6848375797271729, "learning_rate": 2.5020527379371572e-05, "loss": 0.0338, "step": 68320 }, { "grad_norm": 0.3212249279022217, "learning_rate": 2.500620535975564e-05, "loss": 0.025, "step": 68330 }, { "grad_norm": 0.2389480173587799, "learning_rate": 2.4991886073423515e-05, "loss": 0.0254, "step": 68340 }, { "grad_norm": 0.3622489273548126, "learning_rate": 2.4977569521941134e-05, "loss": 0.022, "step": 68350 }, { "grad_norm": 0.32506516575813293, "learning_rate": 2.4963255706874143e-05, "loss": 0.0286, "step": 68360 }, { "grad_norm": 0.33794549107551575, "learning_rate": 2.4948944629787853e-05, "loss": 0.0321, "step": 68370 }, { "grad_norm": 0.31722989678382874, "learning_rate": 2.493463629224731e-05, "loss": 0.0244, "step": 68380 }, { "grad_norm": 0.24888868629932404, "learning_rate": 2.492033069581728e-05, "loss": 0.0255, "step": 68390 }, { "grad_norm": 0.17304034531116486, "learning_rate": 2.4906027842062137e-05, "loss": 0.0271, "step": 68400 }, { "grad_norm": 0.20838455855846405, "learning_rate": 2.48917277325461e-05, "loss": 0.0243, "step": 68410 }, { "grad_norm": 0.26049143075942993, "learning_rate": 2.487743036883295e-05, "loss": 0.0299, "step": 68420 }, { "grad_norm": 0.2537529766559601, "learning_rate": 2.486313575248626e-05, "loss": 0.0246, "step": 68430 }, { "grad_norm": 0.32699063420295715, "learning_rate": 2.4848843885069235e-05, "loss": 0.0223, "step": 68440 }, { "grad_norm": 0.26674890518188477, "learning_rate": 2.483455476814484e-05, "loss": 0.0262, "step": 68450 }, { "grad_norm": 0.41595837473869324, "learning_rate": 2.4820268403275672e-05, "loss": 0.0254, "step": 68460 }, { "grad_norm": 0.3665667772293091, "learning_rate": 2.480598479202413e-05, "loss": 0.0311, "step": 68470 }, { "grad_norm": 0.436117947101593, "learning_rate": 2.4791703935952193e-05, "loss": 0.0297, "step": 68480 }, { "grad_norm": 0.29330429434776306, "learning_rate": 2.477742583662163e-05, "loss": 0.0313, "step": 68490 }, { "grad_norm": 0.6190764307975769, "learning_rate": 2.476315049559384e-05, "loss": 0.0272, "step": 68500 }, { "grad_norm": 0.41566288471221924, "learning_rate": 2.4748877914429987e-05, "loss": 0.0246, "step": 68510 }, { "grad_norm": 0.39988577365875244, "learning_rate": 2.4734608094690857e-05, "loss": 0.0287, "step": 68520 }, { "grad_norm": 0.2561781406402588, "learning_rate": 2.4720341037937006e-05, "loss": 0.0362, "step": 68530 }, { "grad_norm": 0.30614861845970154, "learning_rate": 2.4706076745728645e-05, "loss": 0.0333, "step": 68540 }, { "grad_norm": 1.124974250793457, "learning_rate": 2.469181521962573e-05, "loss": 0.0352, "step": 68550 }, { "grad_norm": 0.24645079672336578, "learning_rate": 2.467755646118783e-05, "loss": 0.0204, "step": 68560 }, { "grad_norm": 0.5693639516830444, "learning_rate": 2.46633004719743e-05, "loss": 0.0262, "step": 68570 }, { "grad_norm": 0.4730927646160126, "learning_rate": 2.4649047253544123e-05, "loss": 0.0236, "step": 68580 }, { "grad_norm": 0.31497570872306824, "learning_rate": 2.4634796807456016e-05, "loss": 0.0331, "step": 68590 }, { "grad_norm": 0.3833460807800293, "learning_rate": 2.46205491352684e-05, "loss": 0.0274, "step": 68600 }, { "grad_norm": 0.5847966074943542, "learning_rate": 2.4606304238539375e-05, "loss": 0.0336, "step": 68610 }, { "grad_norm": 0.2631024420261383, "learning_rate": 2.4592062118826753e-05, "loss": 0.0255, "step": 68620 }, { "grad_norm": 0.3387587368488312, "learning_rate": 2.4577822777688004e-05, "loss": 0.0294, "step": 68630 }, { "grad_norm": 0.45502981543540955, "learning_rate": 2.4563586216680347e-05, "loss": 0.0255, "step": 68640 }, { "grad_norm": 0.24326395988464355, "learning_rate": 2.4549352437360622e-05, "loss": 0.0304, "step": 68650 }, { "grad_norm": 1.5499498844146729, "learning_rate": 2.4535121441285493e-05, "loss": 0.0346, "step": 68660 }, { "grad_norm": 0.8761243224143982, "learning_rate": 2.4520893230011172e-05, "loss": 0.0274, "step": 68670 }, { "grad_norm": 0.3394416272640228, "learning_rate": 2.450666780509368e-05, "loss": 0.0345, "step": 68680 }, { "grad_norm": 0.3613678812980652, "learning_rate": 2.4492445168088656e-05, "loss": 0.0278, "step": 68690 }, { "grad_norm": 0.2845844030380249, "learning_rate": 2.447822532055149e-05, "loss": 0.0263, "step": 68700 }, { "grad_norm": 0.3597909212112427, "learning_rate": 2.4464008264037212e-05, "loss": 0.0197, "step": 68710 }, { "grad_norm": 0.3121410310268402, "learning_rate": 2.4449794000100605e-05, "loss": 0.0237, "step": 68720 }, { "grad_norm": 0.6900610327720642, "learning_rate": 2.4435582530296108e-05, "loss": 0.0241, "step": 68730 }, { "grad_norm": 0.41994398832321167, "learning_rate": 2.4421373856177887e-05, "loss": 0.0249, "step": 68740 }, { "grad_norm": 0.3475397527217865, "learning_rate": 2.440716797929975e-05, "loss": 0.0322, "step": 68750 }, { "grad_norm": 0.22497817873954773, "learning_rate": 2.439296490121526e-05, "loss": 0.0209, "step": 68760 }, { "grad_norm": 0.32419347763061523, "learning_rate": 2.4378764623477608e-05, "loss": 0.0222, "step": 68770 }, { "grad_norm": 0.3841756582260132, "learning_rate": 2.436456714763974e-05, "loss": 0.0348, "step": 68780 }, { "grad_norm": 0.7830901145935059, "learning_rate": 2.4350372475254262e-05, "loss": 0.033, "step": 68790 }, { "grad_norm": 0.2690716087818146, "learning_rate": 2.433618060787351e-05, "loss": 0.0238, "step": 68800 }, { "grad_norm": 0.4944252371788025, "learning_rate": 2.432199154704944e-05, "loss": 0.0304, "step": 68810 }, { "grad_norm": 0.5508385896682739, "learning_rate": 2.430780529433377e-05, "loss": 0.0276, "step": 68820 }, { "grad_norm": 0.5581400990486145, "learning_rate": 2.42936218512779e-05, "loss": 0.0283, "step": 68830 }, { "grad_norm": 0.6002866625785828, "learning_rate": 2.427944121943288e-05, "loss": 0.0254, "step": 68840 }, { "grad_norm": 0.3364022374153137, "learning_rate": 2.4265263400349497e-05, "loss": 0.0296, "step": 68850 }, { "grad_norm": 0.49057304859161377, "learning_rate": 2.4251088395578213e-05, "loss": 0.0373, "step": 68860 }, { "grad_norm": 0.3847920596599579, "learning_rate": 2.4236916206669214e-05, "loss": 0.0282, "step": 68870 }, { "grad_norm": 0.3112742602825165, "learning_rate": 2.4222746835172304e-05, "loss": 0.0258, "step": 68880 }, { "grad_norm": 0.6735737323760986, "learning_rate": 2.4208580282637065e-05, "loss": 0.038, "step": 68890 }, { "grad_norm": 0.47224053740501404, "learning_rate": 2.4194416550612687e-05, "loss": 0.0292, "step": 68900 }, { "grad_norm": 0.2872398793697357, "learning_rate": 2.418025564064812e-05, "loss": 0.0222, "step": 68910 }, { "grad_norm": 0.5086055397987366, "learning_rate": 2.4166097554291978e-05, "loss": 0.0288, "step": 68920 }, { "grad_norm": 0.5648636817932129, "learning_rate": 2.4151942293092583e-05, "loss": 0.0358, "step": 68930 }, { "grad_norm": 0.2428974211215973, "learning_rate": 2.41377898585979e-05, "loss": 0.0235, "step": 68940 }, { "grad_norm": 0.49544769525527954, "learning_rate": 2.4123640252355652e-05, "loss": 0.0297, "step": 68950 }, { "grad_norm": 0.5782148241996765, "learning_rate": 2.410949347591318e-05, "loss": 0.0246, "step": 68960 }, { "grad_norm": 0.45042064785957336, "learning_rate": 2.4095349530817578e-05, "loss": 0.0452, "step": 68970 }, { "grad_norm": 0.3234488368034363, "learning_rate": 2.4081208418615603e-05, "loss": 0.0327, "step": 68980 }, { "grad_norm": 0.20158587396144867, "learning_rate": 2.4067070140853725e-05, "loss": 0.0223, "step": 68990 }, { "grad_norm": 0.6709163188934326, "learning_rate": 2.405293469907805e-05, "loss": 0.0225, "step": 69000 }, { "grad_norm": 0.29111218452453613, "learning_rate": 2.4038802094834446e-05, "loss": 0.0292, "step": 69010 }, { "grad_norm": 0.38262155652046204, "learning_rate": 2.4024672329668395e-05, "loss": 0.0295, "step": 69020 }, { "grad_norm": 0.36794301867485046, "learning_rate": 2.4010545405125125e-05, "loss": 0.0202, "step": 69030 }, { "grad_norm": 0.37434276938438416, "learning_rate": 2.3996421322749528e-05, "loss": 0.0284, "step": 69040 }, { "grad_norm": 0.5420308113098145, "learning_rate": 2.3982300084086224e-05, "loss": 0.0263, "step": 69050 }, { "grad_norm": 0.8366695642471313, "learning_rate": 2.396818169067944e-05, "loss": 0.0253, "step": 69060 }, { "grad_norm": 0.3090227246284485, "learning_rate": 2.3954066144073178e-05, "loss": 0.024, "step": 69070 }, { "grad_norm": 0.3823522627353668, "learning_rate": 2.39399534458111e-05, "loss": 0.0319, "step": 69080 }, { "grad_norm": 0.6834444999694824, "learning_rate": 2.392584359743651e-05, "loss": 0.028, "step": 69090 }, { "grad_norm": 0.21066173911094666, "learning_rate": 2.3911736600492463e-05, "loss": 0.0199, "step": 69100 }, { "grad_norm": 0.531861424446106, "learning_rate": 2.3897632456521672e-05, "loss": 0.0291, "step": 69110 }, { "grad_norm": 0.2127753347158432, "learning_rate": 2.388353116706658e-05, "loss": 0.0329, "step": 69120 }, { "grad_norm": 0.30656903982162476, "learning_rate": 2.3869432733669232e-05, "loss": 0.036, "step": 69130 }, { "grad_norm": 0.3163752257823944, "learning_rate": 2.385533715787145e-05, "loss": 0.0348, "step": 69140 }, { "grad_norm": 0.30642277002334595, "learning_rate": 2.3841244441214673e-05, "loss": 0.0283, "step": 69150 }, { "grad_norm": 0.7762805223464966, "learning_rate": 2.3827154585240073e-05, "loss": 0.0253, "step": 69160 }, { "grad_norm": 0.20172756910324097, "learning_rate": 2.3813067591488498e-05, "loss": 0.0373, "step": 69170 }, { "grad_norm": 0.3186088800430298, "learning_rate": 2.3798983461500494e-05, "loss": 0.0326, "step": 69180 }, { "grad_norm": 0.6941335797309875, "learning_rate": 2.3784902196816256e-05, "loss": 0.0278, "step": 69190 }, { "grad_norm": 0.2718953788280487, "learning_rate": 2.3770823798975716e-05, "loss": 0.0244, "step": 69200 }, { "grad_norm": 0.20448991656303406, "learning_rate": 2.3756748269518424e-05, "loss": 0.0191, "step": 69210 }, { "grad_norm": 0.26629218459129333, "learning_rate": 2.374267560998371e-05, "loss": 0.025, "step": 69220 }, { "grad_norm": 0.4867890179157257, "learning_rate": 2.3728605821910477e-05, "loss": 0.0267, "step": 69230 }, { "grad_norm": 0.446565717458725, "learning_rate": 2.3714538906837453e-05, "loss": 0.0352, "step": 69240 }, { "grad_norm": 0.3069387674331665, "learning_rate": 2.3700474866302913e-05, "loss": 0.0243, "step": 69250 }, { "grad_norm": 0.3012094795703888, "learning_rate": 2.3686413701844917e-05, "loss": 0.0224, "step": 69260 }, { "grad_norm": 0.31339481472969055, "learning_rate": 2.367235541500114e-05, "loss": 0.0347, "step": 69270 }, { "grad_norm": 0.47557100653648376, "learning_rate": 2.3658300007308993e-05, "loss": 0.023, "step": 69280 }, { "grad_norm": 0.23021776974201202, "learning_rate": 2.364424748030555e-05, "loss": 0.0172, "step": 69290 }, { "grad_norm": 0.26319876313209534, "learning_rate": 2.3630197835527595e-05, "loss": 0.0246, "step": 69300 }, { "grad_norm": 0.3163056969642639, "learning_rate": 2.3616151074511546e-05, "loss": 0.0328, "step": 69310 }, { "grad_norm": 0.26690688729286194, "learning_rate": 2.3602107198793545e-05, "loss": 0.0248, "step": 69320 }, { "grad_norm": 0.32452094554901123, "learning_rate": 2.3588066209909427e-05, "loss": 0.0226, "step": 69330 }, { "grad_norm": 0.401509165763855, "learning_rate": 2.3574028109394662e-05, "loss": 0.0348, "step": 69340 }, { "grad_norm": 0.30548134446144104, "learning_rate": 2.3559992898784454e-05, "loss": 0.0291, "step": 69350 }, { "grad_norm": 0.20987777411937714, "learning_rate": 2.354596057961367e-05, "loss": 0.021, "step": 69360 }, { "grad_norm": 0.37113845348358154, "learning_rate": 2.353193115341688e-05, "loss": 0.0354, "step": 69370 }, { "grad_norm": 0.466493159532547, "learning_rate": 2.3517904621728294e-05, "loss": 0.0316, "step": 69380 }, { "grad_norm": 0.218496173620224, "learning_rate": 2.3503880986081855e-05, "loss": 0.0295, "step": 69390 }, { "grad_norm": 0.23652638494968414, "learning_rate": 2.3489860248011136e-05, "loss": 0.0256, "step": 69400 }, { "grad_norm": 0.35414233803749084, "learning_rate": 2.3475842409049464e-05, "loss": 0.0278, "step": 69410 }, { "grad_norm": 0.322455495595932, "learning_rate": 2.346182747072975e-05, "loss": 0.0307, "step": 69420 }, { "grad_norm": 0.3950260579586029, "learning_rate": 2.3447815434584718e-05, "loss": 0.0355, "step": 69430 }, { "grad_norm": 0.34842434525489807, "learning_rate": 2.3433806302146656e-05, "loss": 0.0209, "step": 69440 }, { "grad_norm": 0.38053151965141296, "learning_rate": 2.3419800074947612e-05, "loss": 0.0282, "step": 69450 }, { "grad_norm": 0.4835226833820343, "learning_rate": 2.3405796754519248e-05, "loss": 0.0356, "step": 69460 }, { "grad_norm": 0.3774507939815521, "learning_rate": 2.3391796342392992e-05, "loss": 0.0256, "step": 69470 }, { "grad_norm": 1.0263946056365967, "learning_rate": 2.337779884009984e-05, "loss": 0.0327, "step": 69480 }, { "grad_norm": 0.23672880232334137, "learning_rate": 2.3363804249170624e-05, "loss": 0.0284, "step": 69490 }, { "grad_norm": 0.4223158657550812, "learning_rate": 2.3349812571135703e-05, "loss": 0.0233, "step": 69500 }, { "grad_norm": 0.5825425982475281, "learning_rate": 2.333582380752523e-05, "loss": 0.0264, "step": 69510 }, { "grad_norm": 0.4099825620651245, "learning_rate": 2.3321837959868963e-05, "loss": 0.0225, "step": 69520 }, { "grad_norm": 0.5482739806175232, "learning_rate": 2.3307855029696383e-05, "loss": 0.0236, "step": 69530 }, { "grad_norm": 0.48594868183135986, "learning_rate": 2.3293875018536648e-05, "loss": 0.0229, "step": 69540 }, { "grad_norm": 0.40251263976097107, "learning_rate": 2.327989792791861e-05, "loss": 0.0265, "step": 69550 }, { "grad_norm": 0.3281508684158325, "learning_rate": 2.326592375937074e-05, "loss": 0.0253, "step": 69560 }, { "grad_norm": 0.4940183758735657, "learning_rate": 2.3251952514421248e-05, "loss": 0.0237, "step": 69570 }, { "grad_norm": 0.5105725526809692, "learning_rate": 2.323798419459804e-05, "loss": 0.0255, "step": 69580 }, { "grad_norm": 0.4878372550010681, "learning_rate": 2.3224018801428625e-05, "loss": 0.0185, "step": 69590 }, { "grad_norm": 0.5369704365730286, "learning_rate": 2.3210056336440273e-05, "loss": 0.0332, "step": 69600 }, { "grad_norm": 0.21387220919132233, "learning_rate": 2.3196096801159844e-05, "loss": 0.0299, "step": 69610 }, { "grad_norm": 0.3594772517681122, "learning_rate": 2.3182140197114012e-05, "loss": 0.0264, "step": 69620 }, { "grad_norm": 0.4387483596801758, "learning_rate": 2.3168186525828985e-05, "loss": 0.0322, "step": 69630 }, { "grad_norm": 0.31506475806236267, "learning_rate": 2.3154235788830757e-05, "loss": 0.0251, "step": 69640 }, { "grad_norm": 1.4690871238708496, "learning_rate": 2.3140287987644927e-05, "loss": 0.0255, "step": 69650 }, { "grad_norm": 0.7635940313339233, "learning_rate": 2.3126343123796833e-05, "loss": 0.032, "step": 69660 }, { "grad_norm": 0.5324673056602478, "learning_rate": 2.3112401198811418e-05, "loss": 0.0345, "step": 69670 }, { "grad_norm": 0.5016292929649353, "learning_rate": 2.3098462214213418e-05, "loss": 0.034, "step": 69680 }, { "grad_norm": 0.24782663583755493, "learning_rate": 2.308452617152712e-05, "loss": 0.0228, "step": 69690 }, { "grad_norm": 0.3974475860595703, "learning_rate": 2.3070593072276603e-05, "loss": 0.0314, "step": 69700 }, { "grad_norm": 0.5965907573699951, "learning_rate": 2.3056662917985518e-05, "loss": 0.032, "step": 69710 }, { "grad_norm": 0.24757499992847443, "learning_rate": 2.3042735710177283e-05, "loss": 0.0346, "step": 69720 }, { "grad_norm": 0.5820891857147217, "learning_rate": 2.302881145037492e-05, "loss": 0.0269, "step": 69730 }, { "grad_norm": 0.25452980399131775, "learning_rate": 2.3014890140101192e-05, "loss": 0.0247, "step": 69740 }, { "grad_norm": 0.3355647921562195, "learning_rate": 2.30009717808785e-05, "loss": 0.0283, "step": 69750 }, { "grad_norm": 0.7436366081237793, "learning_rate": 2.2987056374228967e-05, "loss": 0.0246, "step": 69760 }, { "grad_norm": 0.4055073857307434, "learning_rate": 2.297314392167431e-05, "loss": 0.0254, "step": 69770 }, { "grad_norm": 0.36498603224754333, "learning_rate": 2.295923442473601e-05, "loss": 0.0246, "step": 69780 }, { "grad_norm": 0.28540143370628357, "learning_rate": 2.294532788493519e-05, "loss": 0.0273, "step": 69790 }, { "grad_norm": 0.28884974122047424, "learning_rate": 2.293142430379262e-05, "loss": 0.0252, "step": 69800 }, { "grad_norm": 0.34838002920150757, "learning_rate": 2.291752368282879e-05, "loss": 0.027, "step": 69810 }, { "grad_norm": 0.3172301948070526, "learning_rate": 2.290362602356385e-05, "loss": 0.0243, "step": 69820 }, { "grad_norm": 0.2920142710208893, "learning_rate": 2.288973132751765e-05, "loss": 0.022, "step": 69830 }, { "grad_norm": 0.48868563771247864, "learning_rate": 2.287583959620965e-05, "loss": 0.032, "step": 69840 }, { "grad_norm": 0.5545762777328491, "learning_rate": 2.2861950831159074e-05, "loss": 0.0276, "step": 69850 }, { "grad_norm": 0.30362871289253235, "learning_rate": 2.284806503388472e-05, "loss": 0.0297, "step": 69860 }, { "grad_norm": 0.34176748991012573, "learning_rate": 2.283418220590518e-05, "loss": 0.0342, "step": 69870 }, { "grad_norm": 0.3418788015842438, "learning_rate": 2.282030234873862e-05, "loss": 0.0239, "step": 69880 }, { "grad_norm": 0.3390693962574005, "learning_rate": 2.2806425463902948e-05, "loss": 0.0248, "step": 69890 }, { "grad_norm": 0.41172701120376587, "learning_rate": 2.279255155291568e-05, "loss": 0.0241, "step": 69900 }, { "grad_norm": 0.21045467257499695, "learning_rate": 2.277868061729409e-05, "loss": 0.0204, "step": 69910 }, { "grad_norm": 0.30386534333229065, "learning_rate": 2.276481265855504e-05, "loss": 0.0245, "step": 69920 }, { "grad_norm": 0.35136133432388306, "learning_rate": 2.2750947678215128e-05, "loss": 0.0321, "step": 69930 }, { "grad_norm": 0.3642174005508423, "learning_rate": 2.2737085677790616e-05, "loss": 0.0253, "step": 69940 }, { "grad_norm": 0.44882336258888245, "learning_rate": 2.2723226658797443e-05, "loss": 0.0198, "step": 69950 }, { "grad_norm": 0.45157933235168457, "learning_rate": 2.270937062275117e-05, "loss": 0.0238, "step": 69960 }, { "grad_norm": 0.3413858115673065, "learning_rate": 2.2695517571167113e-05, "loss": 0.0289, "step": 69970 }, { "grad_norm": 0.39654770493507385, "learning_rate": 2.268166750556019e-05, "loss": 0.0266, "step": 69980 }, { "grad_norm": 0.7693771123886108, "learning_rate": 2.266782042744503e-05, "loss": 0.0255, "step": 69990 }, { "grad_norm": 0.24434435367584229, "learning_rate": 2.2653976338335935e-05, "loss": 0.0206, "step": 70000 }, { "grad_norm": 0.585292398929596, "learning_rate": 2.2640135239746895e-05, "loss": 0.0283, "step": 70010 }, { "grad_norm": 0.2229156345129013, "learning_rate": 2.2626297133191515e-05, "loss": 0.0296, "step": 70020 }, { "grad_norm": 0.26375114917755127, "learning_rate": 2.261246202018312e-05, "loss": 0.0253, "step": 70030 }, { "grad_norm": 0.31257280707359314, "learning_rate": 2.2598629902234718e-05, "loss": 0.0303, "step": 70040 }, { "grad_norm": 0.35952872037887573, "learning_rate": 2.258480078085894e-05, "loss": 0.0286, "step": 70050 }, { "grad_norm": 0.44501540064811707, "learning_rate": 2.2570974657568126e-05, "loss": 0.0288, "step": 70060 }, { "grad_norm": 0.508615255355835, "learning_rate": 2.2557151533874286e-05, "loss": 0.0256, "step": 70070 }, { "grad_norm": 0.24104972183704376, "learning_rate": 2.2543331411289116e-05, "loss": 0.0175, "step": 70080 }, { "grad_norm": 0.3388156294822693, "learning_rate": 2.252951429132392e-05, "loss": 0.0286, "step": 70090 }, { "grad_norm": 0.3389478623867035, "learning_rate": 2.2515700175489762e-05, "loss": 0.0345, "step": 70100 }, { "grad_norm": 0.3245140612125397, "learning_rate": 2.2501889065297288e-05, "loss": 0.0308, "step": 70110 }, { "grad_norm": 0.4990209639072418, "learning_rate": 2.2488080962256887e-05, "loss": 0.032, "step": 70120 }, { "grad_norm": 0.4812091588973999, "learning_rate": 2.2474275867878585e-05, "loss": 0.0245, "step": 70130 }, { "grad_norm": 2.707324981689453, "learning_rate": 2.246047378367211e-05, "loss": 0.0297, "step": 70140 }, { "grad_norm": 0.32126426696777344, "learning_rate": 2.2446674711146798e-05, "loss": 0.0263, "step": 70150 }, { "grad_norm": 0.4482310712337494, "learning_rate": 2.2432878651811734e-05, "loss": 0.0221, "step": 70160 }, { "grad_norm": 0.3240586221218109, "learning_rate": 2.2419085607175594e-05, "loss": 0.0284, "step": 70170 }, { "grad_norm": 0.24954117834568024, "learning_rate": 2.240529557874678e-05, "loss": 0.036, "step": 70180 }, { "grad_norm": 0.3103595972061157, "learning_rate": 2.2391508568033358e-05, "loss": 0.0236, "step": 70190 }, { "grad_norm": 0.39552879333496094, "learning_rate": 2.237772457654307e-05, "loss": 0.0255, "step": 70200 }, { "grad_norm": 0.32398760318756104, "learning_rate": 2.236394360578327e-05, "loss": 0.037, "step": 70210 }, { "grad_norm": 0.885729968547821, "learning_rate": 2.2350165657261073e-05, "loss": 0.021, "step": 70220 }, { "grad_norm": 0.30487677454948425, "learning_rate": 2.2336390732483164e-05, "loss": 0.0288, "step": 70230 }, { "grad_norm": 0.5378754734992981, "learning_rate": 2.2322618832955976e-05, "loss": 0.0364, "step": 70240 }, { "grad_norm": 0.27846193313598633, "learning_rate": 2.2308849960185578e-05, "loss": 0.0209, "step": 70250 }, { "grad_norm": 0.5165155529975891, "learning_rate": 2.229508411567774e-05, "loss": 0.0288, "step": 70260 }, { "grad_norm": 0.22363725304603577, "learning_rate": 2.2281321300937823e-05, "loss": 0.0353, "step": 70270 }, { "grad_norm": 0.3099033832550049, "learning_rate": 2.2267561517470937e-05, "loss": 0.0205, "step": 70280 }, { "grad_norm": 0.5285758972167969, "learning_rate": 2.2253804766781845e-05, "loss": 0.0352, "step": 70290 }, { "grad_norm": 0.3774970769882202, "learning_rate": 2.2240051050374933e-05, "loss": 0.0247, "step": 70300 }, { "grad_norm": 0.3688333332538605, "learning_rate": 2.22263003697543e-05, "loss": 0.0276, "step": 70310 }, { "grad_norm": 0.31406083703041077, "learning_rate": 2.2212552726423702e-05, "loss": 0.0301, "step": 70320 }, { "grad_norm": 0.3798879086971283, "learning_rate": 2.2198808121886578e-05, "loss": 0.0234, "step": 70330 }, { "grad_norm": 0.28135988116264343, "learning_rate": 2.218506655764598e-05, "loss": 0.023, "step": 70340 }, { "grad_norm": 0.3476146459579468, "learning_rate": 2.2171328035204704e-05, "loss": 0.0235, "step": 70350 }, { "grad_norm": 0.2634429931640625, "learning_rate": 2.2157592556065133e-05, "loss": 0.0209, "step": 70360 }, { "grad_norm": 0.2859942615032196, "learning_rate": 2.2143860121729382e-05, "loss": 0.0256, "step": 70370 }, { "grad_norm": 0.3395737409591675, "learning_rate": 2.2130130733699206e-05, "loss": 0.0224, "step": 70380 }, { "grad_norm": 0.4333697557449341, "learning_rate": 2.2116404393476052e-05, "loss": 0.0259, "step": 70390 }, { "grad_norm": 0.5114978551864624, "learning_rate": 2.2102681102560967e-05, "loss": 0.0271, "step": 70400 }, { "grad_norm": 0.23587991297245026, "learning_rate": 2.208896086245476e-05, "loss": 0.0194, "step": 70410 }, { "grad_norm": 0.266747385263443, "learning_rate": 2.2075243674657808e-05, "loss": 0.0243, "step": 70420 }, { "grad_norm": 0.5067321062088013, "learning_rate": 2.206152954067024e-05, "loss": 0.0219, "step": 70430 }, { "grad_norm": 0.6532455682754517, "learning_rate": 2.2047818461991758e-05, "loss": 0.0316, "step": 70440 }, { "grad_norm": 0.572399914264679, "learning_rate": 2.2034110440121864e-05, "loss": 0.036, "step": 70450 }, { "grad_norm": 0.4730087220668793, "learning_rate": 2.2020405476559585e-05, "loss": 0.0272, "step": 70460 }, { "grad_norm": 0.484360933303833, "learning_rate": 2.2006703572803722e-05, "loss": 0.0343, "step": 70470 }, { "grad_norm": 0.330795556306839, "learning_rate": 2.1993004730352646e-05, "loss": 0.0289, "step": 70480 }, { "grad_norm": 0.45790088176727295, "learning_rate": 2.1979308950704464e-05, "loss": 0.0252, "step": 70490 }, { "grad_norm": 0.3617950975894928, "learning_rate": 2.1965616235356923e-05, "loss": 0.0265, "step": 70500 }, { "grad_norm": 0.32681846618652344, "learning_rate": 2.1951926585807465e-05, "loss": 0.0243, "step": 70510 }, { "grad_norm": 0.27981066703796387, "learning_rate": 2.193824000355313e-05, "loss": 0.0268, "step": 70520 }, { "grad_norm": 0.5326762199401855, "learning_rate": 2.192455649009067e-05, "loss": 0.0208, "step": 70530 }, { "grad_norm": 0.3012193739414215, "learning_rate": 2.1910876046916523e-05, "loss": 0.0244, "step": 70540 }, { "grad_norm": 0.2998046875, "learning_rate": 2.1897198675526725e-05, "loss": 0.0258, "step": 70550 }, { "grad_norm": 0.4899197518825531, "learning_rate": 2.1883524377417024e-05, "loss": 0.0276, "step": 70560 }, { "grad_norm": 0.38296473026275635, "learning_rate": 2.1869853154082825e-05, "loss": 0.025, "step": 70570 }, { "grad_norm": 0.17657387256622314, "learning_rate": 2.1856185007019215e-05, "loss": 0.0217, "step": 70580 }, { "grad_norm": 0.37741708755493164, "learning_rate": 2.1842519937720874e-05, "loss": 0.0355, "step": 70590 }, { "grad_norm": 0.5059241652488708, "learning_rate": 2.182885794768224e-05, "loss": 0.0306, "step": 70600 }, { "grad_norm": 0.3282897472381592, "learning_rate": 2.1815199038397326e-05, "loss": 0.0269, "step": 70610 }, { "grad_norm": 0.2977214455604553, "learning_rate": 2.1801543211359893e-05, "loss": 0.0329, "step": 70620 }, { "grad_norm": 0.5114621520042419, "learning_rate": 2.178789046806326e-05, "loss": 0.0273, "step": 70630 }, { "grad_norm": 0.49425700306892395, "learning_rate": 2.1774240810000547e-05, "loss": 0.0283, "step": 70640 }, { "grad_norm": 0.43310877680778503, "learning_rate": 2.17605942386644e-05, "loss": 0.0301, "step": 70650 }, { "grad_norm": 0.423780620098114, "learning_rate": 2.174695075554723e-05, "loss": 0.0366, "step": 70660 }, { "grad_norm": 0.43352073431015015, "learning_rate": 2.173331036214103e-05, "loss": 0.0365, "step": 70670 }, { "grad_norm": 0.2823050022125244, "learning_rate": 2.171967305993753e-05, "loss": 0.0238, "step": 70680 }, { "grad_norm": 0.33397141098976135, "learning_rate": 2.1706038850428024e-05, "loss": 0.022, "step": 70690 }, { "grad_norm": 0.39191094040870667, "learning_rate": 2.169240773510361e-05, "loss": 0.0217, "step": 70700 }, { "grad_norm": 0.2155178040266037, "learning_rate": 2.1678779715454906e-05, "loss": 0.0219, "step": 70710 }, { "grad_norm": 0.2759936451911926, "learning_rate": 2.166515479297229e-05, "loss": 0.028, "step": 70720 }, { "grad_norm": 0.2702253758907318, "learning_rate": 2.1651532969145732e-05, "loss": 0.0184, "step": 70730 }, { "grad_norm": 0.35538098216056824, "learning_rate": 2.1637914245464903e-05, "loss": 0.0307, "step": 70740 }, { "grad_norm": 0.3577924072742462, "learning_rate": 2.1624298623419133e-05, "loss": 0.031, "step": 70750 }, { "grad_norm": 0.43235015869140625, "learning_rate": 2.1610686104497413e-05, "loss": 0.0296, "step": 70760 }, { "grad_norm": 0.7163278460502625, "learning_rate": 2.1597076690188366e-05, "loss": 0.0349, "step": 70770 }, { "grad_norm": 0.26975271105766296, "learning_rate": 2.1583470381980315e-05, "loss": 0.0288, "step": 70780 }, { "grad_norm": 0.20575688779354095, "learning_rate": 2.156986718136123e-05, "loss": 0.0392, "step": 70790 }, { "grad_norm": 0.5085625052452087, "learning_rate": 2.155626708981871e-05, "loss": 0.0286, "step": 70800 }, { "grad_norm": 0.310684472322464, "learning_rate": 2.1542670108840075e-05, "loss": 0.0317, "step": 70810 }, { "grad_norm": 0.6434874534606934, "learning_rate": 2.152907623991221e-05, "loss": 0.0307, "step": 70820 }, { "grad_norm": 0.3620641827583313, "learning_rate": 2.1515485484521802e-05, "loss": 0.0267, "step": 70830 }, { "grad_norm": 0.6275088787078857, "learning_rate": 2.150189784415506e-05, "loss": 0.0264, "step": 70840 }, { "grad_norm": 0.8377845883369446, "learning_rate": 2.148831332029794e-05, "loss": 0.0293, "step": 70850 }, { "grad_norm": 0.2951663136482239, "learning_rate": 2.147473191443599e-05, "loss": 0.026, "step": 70860 }, { "grad_norm": 0.3120831251144409, "learning_rate": 2.1461153628054492e-05, "loss": 0.0268, "step": 70870 }, { "grad_norm": 0.43488624691963196, "learning_rate": 2.1447578462638284e-05, "loss": 0.0338, "step": 70880 }, { "grad_norm": 0.2275419533252716, "learning_rate": 2.143400641967201e-05, "loss": 0.0265, "step": 70890 }, { "grad_norm": 0.4373268783092499, "learning_rate": 2.1420437500639823e-05, "loss": 0.0252, "step": 70900 }, { "grad_norm": 0.21762649714946747, "learning_rate": 2.1406871707025646e-05, "loss": 0.032, "step": 70910 }, { "grad_norm": 0.34828490018844604, "learning_rate": 2.1393309040312963e-05, "loss": 0.0361, "step": 70920 }, { "grad_norm": 0.21460475027561188, "learning_rate": 2.137974950198502e-05, "loss": 0.0322, "step": 70930 }, { "grad_norm": 0.30198532342910767, "learning_rate": 2.1366193093524617e-05, "loss": 0.026, "step": 70940 }, { "grad_norm": 0.5401572585105896, "learning_rate": 2.1352639816414288e-05, "loss": 0.0206, "step": 70950 }, { "grad_norm": 0.3931776285171509, "learning_rate": 2.1339089672136202e-05, "loss": 0.0292, "step": 70960 }, { "grad_norm": 0.3488512635231018, "learning_rate": 2.1325542662172193e-05, "loss": 0.0243, "step": 70970 }, { "grad_norm": 0.2774413824081421, "learning_rate": 2.1311998788003712e-05, "loss": 0.0279, "step": 70980 }, { "grad_norm": 0.36585041880607605, "learning_rate": 2.1298458051111907e-05, "loss": 0.0259, "step": 70990 }, { "grad_norm": 0.3263642191886902, "learning_rate": 2.1284920452977602e-05, "loss": 0.0295, "step": 71000 }, { "grad_norm": 0.26723530888557434, "learning_rate": 2.12713859950812e-05, "loss": 0.0263, "step": 71010 }, { "grad_norm": 0.3309745788574219, "learning_rate": 2.1257854678902832e-05, "loss": 0.0366, "step": 71020 }, { "grad_norm": 0.3624345660209656, "learning_rate": 2.124432650592227e-05, "loss": 0.029, "step": 71030 }, { "grad_norm": 0.28921106457710266, "learning_rate": 2.1230801477618937e-05, "loss": 0.0254, "step": 71040 }, { "grad_norm": 0.6307228803634644, "learning_rate": 2.121727959547189e-05, "loss": 0.0351, "step": 71050 }, { "grad_norm": 0.3496832251548767, "learning_rate": 2.1203760860959886e-05, "loss": 0.0237, "step": 71060 }, { "grad_norm": 0.4230174422264099, "learning_rate": 2.119024527556127e-05, "loss": 0.0299, "step": 71070 }, { "grad_norm": 0.26170212030410767, "learning_rate": 2.117673284075415e-05, "loss": 0.0206, "step": 71080 }, { "grad_norm": 0.20370391011238098, "learning_rate": 2.1163223558016165e-05, "loss": 0.0206, "step": 71090 }, { "grad_norm": 0.3971250355243683, "learning_rate": 2.1149717428824722e-05, "loss": 0.0241, "step": 71100 }, { "grad_norm": 0.4229511022567749, "learning_rate": 2.1136214454656784e-05, "loss": 0.0295, "step": 71110 }, { "grad_norm": 0.3400741219520569, "learning_rate": 2.1122714636989056e-05, "loss": 0.0238, "step": 71120 }, { "grad_norm": 0.28721851110458374, "learning_rate": 2.1109217977297823e-05, "loss": 0.024, "step": 71130 }, { "grad_norm": 0.33767637610435486, "learning_rate": 2.1095724477059077e-05, "loss": 0.0228, "step": 71140 }, { "grad_norm": 0.47976458072662354, "learning_rate": 2.1082234137748434e-05, "loss": 0.0309, "step": 71150 }, { "grad_norm": 0.3080427348613739, "learning_rate": 2.106874696084122e-05, "loss": 0.0229, "step": 71160 }, { "grad_norm": 0.21624381840229034, "learning_rate": 2.1055262947812315e-05, "loss": 0.0222, "step": 71170 }, { "grad_norm": 0.3647078275680542, "learning_rate": 2.104178210013636e-05, "loss": 0.03, "step": 71180 }, { "grad_norm": 0.3854422867298126, "learning_rate": 2.1028304419287554e-05, "loss": 0.0382, "step": 71190 }, { "grad_norm": 0.35183295607566833, "learning_rate": 2.101482990673982e-05, "loss": 0.0242, "step": 71200 }, { "grad_norm": 0.4045516550540924, "learning_rate": 2.1001358563966707e-05, "loss": 0.0327, "step": 71210 }, { "grad_norm": 0.20931874215602875, "learning_rate": 2.098789039244144e-05, "loss": 0.0213, "step": 71220 }, { "grad_norm": 0.6550837755203247, "learning_rate": 2.097442539363685e-05, "loss": 0.0255, "step": 71230 }, { "grad_norm": 0.309360146522522, "learning_rate": 2.0960963569025448e-05, "loss": 0.0238, "step": 71240 }, { "grad_norm": 0.2935583293437958, "learning_rate": 2.0947504920079435e-05, "loss": 0.0225, "step": 71250 }, { "grad_norm": 0.2803201377391815, "learning_rate": 2.093404944827059e-05, "loss": 0.0235, "step": 71260 }, { "grad_norm": 0.20333144068717957, "learning_rate": 2.0920597155070394e-05, "loss": 0.025, "step": 71270 }, { "grad_norm": 0.379977822303772, "learning_rate": 2.090714804194997e-05, "loss": 0.0237, "step": 71280 }, { "grad_norm": 0.26420074701309204, "learning_rate": 2.0893702110380114e-05, "loss": 0.0192, "step": 71290 }, { "grad_norm": 0.24899423122406006, "learning_rate": 2.0880259361831216e-05, "loss": 0.0324, "step": 71300 }, { "grad_norm": 0.31585732102394104, "learning_rate": 2.0866819797773397e-05, "loss": 0.0244, "step": 71310 }, { "grad_norm": 0.4936915934085846, "learning_rate": 2.085338341967634e-05, "loss": 0.0305, "step": 71320 }, { "grad_norm": 0.4112372398376465, "learning_rate": 2.083995022900946e-05, "loss": 0.0226, "step": 71330 }, { "grad_norm": 0.35252758860588074, "learning_rate": 2.0826520227241768e-05, "loss": 0.0248, "step": 71340 }, { "grad_norm": 0.3404652774333954, "learning_rate": 2.081309341584199e-05, "loss": 0.0322, "step": 71350 }, { "grad_norm": 0.3820529580116272, "learning_rate": 2.0799669796278413e-05, "loss": 0.0329, "step": 71360 }, { "grad_norm": 0.355221688747406, "learning_rate": 2.078624937001906e-05, "loss": 0.0304, "step": 71370 }, { "grad_norm": 0.8541982173919678, "learning_rate": 2.0772832138531544e-05, "loss": 0.0505, "step": 71380 }, { "grad_norm": 0.45672184228897095, "learning_rate": 2.0759418103283158e-05, "loss": 0.0297, "step": 71390 }, { "grad_norm": 0.35106703639030457, "learning_rate": 2.0746007265740846e-05, "loss": 0.0291, "step": 71400 }, { "grad_norm": 0.28047651052474976, "learning_rate": 2.0732599627371215e-05, "loss": 0.0296, "step": 71410 }, { "grad_norm": 0.4395713210105896, "learning_rate": 2.0719195189640468e-05, "loss": 0.0308, "step": 71420 }, { "grad_norm": 0.1566210687160492, "learning_rate": 2.0705793954014528e-05, "loss": 0.0194, "step": 71430 }, { "grad_norm": 0.21872861683368683, "learning_rate": 2.06923959219589e-05, "loss": 0.0184, "step": 71440 }, { "grad_norm": 0.3981750011444092, "learning_rate": 2.0679001094938793e-05, "loss": 0.0288, "step": 71450 }, { "grad_norm": 0.29416656494140625, "learning_rate": 2.066560947441904e-05, "loss": 0.0215, "step": 71460 }, { "grad_norm": 0.4931856691837311, "learning_rate": 2.0652221061864152e-05, "loss": 0.0287, "step": 71470 }, { "grad_norm": 0.39765042066574097, "learning_rate": 2.0638835858738226e-05, "loss": 0.0268, "step": 71480 }, { "grad_norm": 0.5692580938339233, "learning_rate": 2.0625453866505068e-05, "loss": 0.033, "step": 71490 }, { "grad_norm": 0.34046536684036255, "learning_rate": 2.0612075086628123e-05, "loss": 0.0283, "step": 71500 }, { "grad_norm": 0.38929712772369385, "learning_rate": 2.059869952057045e-05, "loss": 0.0342, "step": 71510 }, { "grad_norm": 0.4965323805809021, "learning_rate": 2.0585327169794793e-05, "loss": 0.0351, "step": 71520 }, { "grad_norm": 0.29588449001312256, "learning_rate": 2.0571958035763534e-05, "loss": 0.0221, "step": 71530 }, { "grad_norm": 0.3236818015575409, "learning_rate": 2.0558592119938715e-05, "loss": 0.0218, "step": 71540 }, { "grad_norm": 0.5381591320037842, "learning_rate": 2.054522942378198e-05, "loss": 0.0216, "step": 71550 }, { "grad_norm": 0.48236513137817383, "learning_rate": 2.053186994875469e-05, "loss": 0.0271, "step": 71560 }, { "grad_norm": 0.43754392862319946, "learning_rate": 2.0518513696317775e-05, "loss": 0.0315, "step": 71570 }, { "grad_norm": 0.6736088991165161, "learning_rate": 2.050516066793188e-05, "loss": 0.0305, "step": 71580 }, { "grad_norm": 0.22011978924274445, "learning_rate": 2.0491810865057276e-05, "loss": 0.0223, "step": 71590 }, { "grad_norm": 0.40708261728286743, "learning_rate": 2.0478464289153882e-05, "loss": 0.0225, "step": 71600 }, { "grad_norm": 0.2700121998786926, "learning_rate": 2.0465120941681232e-05, "loss": 0.0252, "step": 71610 }, { "grad_norm": 0.3186335563659668, "learning_rate": 2.0451780824098566e-05, "loss": 0.0277, "step": 71620 }, { "grad_norm": 0.3112495243549347, "learning_rate": 2.043844393786471e-05, "loss": 0.0207, "step": 71630 }, { "grad_norm": 0.34990835189819336, "learning_rate": 2.0425110284438193e-05, "loss": 0.0221, "step": 71640 }, { "grad_norm": 0.3564479947090149, "learning_rate": 2.041177986527712e-05, "loss": 0.0195, "step": 71650 }, { "grad_norm": 0.3116355240345001, "learning_rate": 2.0398452681839347e-05, "loss": 0.0227, "step": 71660 }, { "grad_norm": 0.4378874897956848, "learning_rate": 2.0385128735582272e-05, "loss": 0.0433, "step": 71670 }, { "grad_norm": 0.2573919892311096, "learning_rate": 2.0371808027963003e-05, "loss": 0.0228, "step": 71680 }, { "grad_norm": 0.39156386256217957, "learning_rate": 2.035849056043825e-05, "loss": 0.033, "step": 71690 }, { "grad_norm": 0.524807870388031, "learning_rate": 2.03451763344644e-05, "loss": 0.0259, "step": 71700 }, { "grad_norm": 0.2564934194087982, "learning_rate": 2.0331865351497482e-05, "loss": 0.0215, "step": 71710 }, { "grad_norm": 0.456905335187912, "learning_rate": 2.0318557612993185e-05, "loss": 0.0264, "step": 71720 }, { "grad_norm": 0.33070623874664307, "learning_rate": 2.0305253120406782e-05, "loss": 0.0263, "step": 71730 }, { "grad_norm": 0.3409329950809479, "learning_rate": 2.029195187519325e-05, "loss": 0.0276, "step": 71740 }, { "grad_norm": 0.34957024455070496, "learning_rate": 2.0278653878807217e-05, "loss": 0.0268, "step": 71750 }, { "grad_norm": 0.5513356924057007, "learning_rate": 2.0265359132702892e-05, "loss": 0.0289, "step": 71760 }, { "grad_norm": 0.3377019762992859, "learning_rate": 2.0252067638334182e-05, "loss": 0.0316, "step": 71770 }, { "grad_norm": 0.33201152086257935, "learning_rate": 2.023877939715464e-05, "loss": 0.0237, "step": 71780 }, { "grad_norm": 0.41843923926353455, "learning_rate": 2.0225494410617448e-05, "loss": 0.0296, "step": 71790 }, { "grad_norm": 0.36472171545028687, "learning_rate": 2.02122126801754e-05, "loss": 0.0261, "step": 71800 }, { "grad_norm": 0.23874227702617645, "learning_rate": 2.0198934207281018e-05, "loss": 0.0239, "step": 71810 }, { "grad_norm": 0.3759511709213257, "learning_rate": 2.0185658993386357e-05, "loss": 0.0233, "step": 71820 }, { "grad_norm": 0.3748284876346588, "learning_rate": 2.0172387039943226e-05, "loss": 0.0278, "step": 71830 }, { "grad_norm": 0.30602332949638367, "learning_rate": 2.0159118348402968e-05, "loss": 0.0309, "step": 71840 }, { "grad_norm": 0.2635386884212494, "learning_rate": 2.0145852920216697e-05, "loss": 0.0224, "step": 71850 }, { "grad_norm": 0.5381132960319519, "learning_rate": 2.0132590756835045e-05, "loss": 0.0329, "step": 71860 }, { "grad_norm": 0.2936464846134186, "learning_rate": 2.0119331859708386e-05, "loss": 0.0324, "step": 71870 }, { "grad_norm": 0.5309588313102722, "learning_rate": 2.0106076230286654e-05, "loss": 0.0223, "step": 71880 }, { "grad_norm": 0.3450636863708496, "learning_rate": 2.0092823870019488e-05, "loss": 0.029, "step": 71890 }, { "grad_norm": 0.26331156492233276, "learning_rate": 2.0079574780356115e-05, "loss": 0.0254, "step": 71900 }, { "grad_norm": 0.38610097765922546, "learning_rate": 2.0066328962745486e-05, "loss": 0.0339, "step": 71910 }, { "grad_norm": 0.6915959119796753, "learning_rate": 2.00530864186361e-05, "loss": 0.0281, "step": 71920 }, { "grad_norm": 0.5660017132759094, "learning_rate": 2.003984714947618e-05, "loss": 0.0279, "step": 71930 }, { "grad_norm": 0.30638521909713745, "learning_rate": 2.0026611156713515e-05, "loss": 0.0291, "step": 71940 }, { "grad_norm": 0.3503197133541107, "learning_rate": 2.0013378441795584e-05, "loss": 0.0226, "step": 71950 }, { "grad_norm": 0.2900736927986145, "learning_rate": 2.0000149006169506e-05, "loss": 0.0261, "step": 71960 }, { "grad_norm": 0.4017716944217682, "learning_rate": 1.9986922851282048e-05, "loss": 0.0261, "step": 71970 }, { "grad_norm": 0.39148151874542236, "learning_rate": 1.997369997857956e-05, "loss": 0.0227, "step": 71980 }, { "grad_norm": 0.30140742659568787, "learning_rate": 1.9960480389508103e-05, "loss": 0.0294, "step": 71990 }, { "grad_norm": 0.4611875116825104, "learning_rate": 1.9947264085513363e-05, "loss": 0.0238, "step": 72000 }, { "grad_norm": 0.3336869478225708, "learning_rate": 1.993405106804062e-05, "loss": 0.0259, "step": 72010 }, { "grad_norm": 0.4942825138568878, "learning_rate": 1.9920841338534867e-05, "loss": 0.025, "step": 72020 }, { "grad_norm": 0.6446898579597473, "learning_rate": 1.9907634898440646e-05, "loss": 0.0276, "step": 72030 }, { "grad_norm": 0.26150423288345337, "learning_rate": 1.989443174920227e-05, "loss": 0.0217, "step": 72040 }, { "grad_norm": 0.44054868817329407, "learning_rate": 1.988123189226355e-05, "loss": 0.0418, "step": 72050 }, { "grad_norm": 0.5008106827735901, "learning_rate": 1.986803532906805e-05, "loss": 0.0336, "step": 72060 }, { "grad_norm": 0.19849568605422974, "learning_rate": 1.985484206105889e-05, "loss": 0.0214, "step": 72070 }, { "grad_norm": 0.3323899507522583, "learning_rate": 1.9841652089678887e-05, "loss": 0.0177, "step": 72080 }, { "grad_norm": 0.5456673502922058, "learning_rate": 1.9828465416370435e-05, "loss": 0.0334, "step": 72090 }, { "grad_norm": 0.37052565813064575, "learning_rate": 1.9815282042575685e-05, "loss": 0.0272, "step": 72100 }, { "grad_norm": 0.4239543378353119, "learning_rate": 1.9802101969736285e-05, "loss": 0.0315, "step": 72110 }, { "grad_norm": 0.4087217152118683, "learning_rate": 1.9788925199293633e-05, "loss": 0.0269, "step": 72120 }, { "grad_norm": 0.3611605167388916, "learning_rate": 1.977575173268868e-05, "loss": 0.0258, "step": 72130 }, { "grad_norm": 0.3238665461540222, "learning_rate": 1.9762581571362094e-05, "loss": 0.023, "step": 72140 }, { "grad_norm": 1.332629919052124, "learning_rate": 1.9749414716754112e-05, "loss": 0.035, "step": 72150 }, { "grad_norm": 0.39399152994155884, "learning_rate": 1.973625117030465e-05, "loss": 0.0296, "step": 72160 }, { "grad_norm": 0.33124464750289917, "learning_rate": 1.972309093345326e-05, "loss": 0.0215, "step": 72170 }, { "grad_norm": 0.2635563313961029, "learning_rate": 1.970993400763915e-05, "loss": 0.0295, "step": 72180 }, { "grad_norm": 0.22433188557624817, "learning_rate": 1.969678039430109e-05, "loss": 0.034, "step": 72190 }, { "grad_norm": 0.6711770296096802, "learning_rate": 1.9683630094877564e-05, "loss": 0.0265, "step": 72200 }, { "grad_norm": 0.23922976851463318, "learning_rate": 1.9670483110806696e-05, "loss": 0.0296, "step": 72210 }, { "grad_norm": 0.47305479645729065, "learning_rate": 1.9657339443526172e-05, "loss": 0.0341, "step": 72220 }, { "grad_norm": 0.3469267785549164, "learning_rate": 1.964419909447339e-05, "loss": 0.0245, "step": 72230 }, { "grad_norm": 0.40283867716789246, "learning_rate": 1.963106206508536e-05, "loss": 0.02, "step": 72240 }, { "grad_norm": 0.31713399291038513, "learning_rate": 1.9617928356798738e-05, "loss": 0.0224, "step": 72250 }, { "grad_norm": 0.2728672921657562, "learning_rate": 1.960479797104977e-05, "loss": 0.0298, "step": 72260 }, { "grad_norm": 0.293622225522995, "learning_rate": 1.9591670909274418e-05, "loss": 0.0225, "step": 72270 }, { "grad_norm": 0.3180742561817169, "learning_rate": 1.9578547172908184e-05, "loss": 0.0312, "step": 72280 }, { "grad_norm": 0.7657878398895264, "learning_rate": 1.9565426763386325e-05, "loss": 0.0215, "step": 72290 }, { "grad_norm": 0.1877269148826599, "learning_rate": 1.9552309682143622e-05, "loss": 0.0252, "step": 72300 }, { "grad_norm": 0.20702998340129852, "learning_rate": 1.9539195930614574e-05, "loss": 0.0229, "step": 72310 }, { "grad_norm": 0.21216343343257904, "learning_rate": 1.952608551023324e-05, "loss": 0.0225, "step": 72320 }, { "grad_norm": 0.6749818921089172, "learning_rate": 1.9512978422433397e-05, "loss": 0.0265, "step": 72330 }, { "grad_norm": 0.33424240350723267, "learning_rate": 1.9499874668648376e-05, "loss": 0.0173, "step": 72340 }, { "grad_norm": 0.4893006384372711, "learning_rate": 1.9486774250311213e-05, "loss": 0.0191, "step": 72350 }, { "grad_norm": 0.35023707151412964, "learning_rate": 1.9473677168854537e-05, "loss": 0.0278, "step": 72360 }, { "grad_norm": 1.0354743003845215, "learning_rate": 1.946058342571065e-05, "loss": 0.0341, "step": 72370 }, { "grad_norm": 0.2859097123146057, "learning_rate": 1.9447493022311424e-05, "loss": 0.0214, "step": 72380 }, { "grad_norm": 0.4337795674800873, "learning_rate": 1.9434405960088442e-05, "loss": 0.0285, "step": 72390 }, { "grad_norm": 0.2520604133605957, "learning_rate": 1.9421322240472857e-05, "loss": 0.0237, "step": 72400 }, { "grad_norm": 0.37320926785469055, "learning_rate": 1.9408241864895494e-05, "loss": 0.0248, "step": 72410 }, { "grad_norm": 0.3175516128540039, "learning_rate": 1.939516483478681e-05, "loss": 0.0246, "step": 72420 }, { "grad_norm": 2.4696319103240967, "learning_rate": 1.9382091151576902e-05, "loss": 0.042, "step": 72430 }, { "grad_norm": 0.5887700319290161, "learning_rate": 1.9369020816695454e-05, "loss": 0.0327, "step": 72440 }, { "grad_norm": 0.19960032403469086, "learning_rate": 1.935595383157184e-05, "loss": 0.0223, "step": 72450 }, { "grad_norm": 0.42675480246543884, "learning_rate": 1.9342890197635057e-05, "loss": 0.0209, "step": 72460 }, { "grad_norm": 0.3206101953983307, "learning_rate": 1.9329829916313685e-05, "loss": 0.0271, "step": 72470 }, { "grad_norm": 0.3338262736797333, "learning_rate": 1.9316772989036013e-05, "loss": 0.0282, "step": 72480 }, { "grad_norm": 0.29227522015571594, "learning_rate": 1.930371941722991e-05, "loss": 0.0155, "step": 72490 }, { "grad_norm": 0.3203674852848053, "learning_rate": 1.929066920232292e-05, "loss": 0.0264, "step": 72500 }, { "grad_norm": 0.24132534861564636, "learning_rate": 1.9277622345742152e-05, "loss": 0.0232, "step": 72510 }, { "grad_norm": 0.4069019854068756, "learning_rate": 1.926457884891444e-05, "loss": 0.0281, "step": 72520 }, { "grad_norm": 0.28766709566116333, "learning_rate": 1.9251538713266147e-05, "loss": 0.0259, "step": 72530 }, { "grad_norm": 0.505238950252533, "learning_rate": 1.9238501940223358e-05, "loss": 0.0269, "step": 72540 }, { "grad_norm": 0.4650963246822357, "learning_rate": 1.922546853121175e-05, "loss": 0.0338, "step": 72550 }, { "grad_norm": 0.30693572759628296, "learning_rate": 1.9212438487656647e-05, "loss": 0.0399, "step": 72560 }, { "grad_norm": 0.23426100611686707, "learning_rate": 1.9199411810982965e-05, "loss": 0.0258, "step": 72570 }, { "grad_norm": 0.26086339354515076, "learning_rate": 1.918638850261532e-05, "loss": 0.0171, "step": 72580 }, { "grad_norm": 0.8092524409294128, "learning_rate": 1.9173368563977885e-05, "loss": 0.0192, "step": 72590 }, { "grad_norm": 0.2518918812274933, "learning_rate": 1.9160351996494513e-05, "loss": 0.0234, "step": 72600 }, { "grad_norm": 0.38438650965690613, "learning_rate": 1.9147338801588677e-05, "loss": 0.0262, "step": 72610 }, { "grad_norm": 0.35301923751831055, "learning_rate": 1.913432898068351e-05, "loss": 0.0257, "step": 72620 }, { "grad_norm": 0.28534767031669617, "learning_rate": 1.912132253520169e-05, "loss": 0.0253, "step": 72630 }, { "grad_norm": 0.44058355689048767, "learning_rate": 1.910831946656564e-05, "loss": 0.0375, "step": 72640 }, { "grad_norm": 0.23981733620166779, "learning_rate": 1.909531977619731e-05, "loss": 0.0287, "step": 72650 }, { "grad_norm": 1.6216700077056885, "learning_rate": 1.908232346551834e-05, "loss": 0.0172, "step": 72660 }, { "grad_norm": 0.32130175828933716, "learning_rate": 1.9069330535949993e-05, "loss": 0.0249, "step": 72670 }, { "grad_norm": 0.4579578638076782, "learning_rate": 1.9056340988913167e-05, "loss": 0.0187, "step": 72680 }, { "grad_norm": 0.26040250062942505, "learning_rate": 1.9043354825828357e-05, "loss": 0.0228, "step": 72690 }, { "grad_norm": 1.4013727903366089, "learning_rate": 1.9030372048115724e-05, "loss": 0.0202, "step": 72700 }, { "grad_norm": 0.4080129861831665, "learning_rate": 1.9017392657195056e-05, "loss": 0.0338, "step": 72710 }, { "grad_norm": 0.5508304834365845, "learning_rate": 1.9004416654485724e-05, "loss": 0.0235, "step": 72720 }, { "grad_norm": 0.36317721009254456, "learning_rate": 1.8991444041406785e-05, "loss": 0.0255, "step": 72730 }, { "grad_norm": 0.29419445991516113, "learning_rate": 1.8978474819376912e-05, "loss": 0.031, "step": 72740 }, { "grad_norm": 0.7816488146781921, "learning_rate": 1.8965508989814406e-05, "loss": 0.033, "step": 72750 }, { "grad_norm": 0.2840067148208618, "learning_rate": 1.8952546554137158e-05, "loss": 0.0238, "step": 72760 }, { "grad_norm": 0.4041483998298645, "learning_rate": 1.8939587513762757e-05, "loss": 0.0253, "step": 72770 }, { "grad_norm": 0.30675411224365234, "learning_rate": 1.892663187010835e-05, "loss": 0.0228, "step": 72780 }, { "grad_norm": 0.34095001220703125, "learning_rate": 1.891367962459077e-05, "loss": 0.0284, "step": 72790 }, { "grad_norm": 0.29538872838020325, "learning_rate": 1.890073077862644e-05, "loss": 0.0218, "step": 72800 }, { "grad_norm": 0.3330189287662506, "learning_rate": 1.888778533363146e-05, "loss": 0.0271, "step": 72810 }, { "grad_norm": 0.5919203162193298, "learning_rate": 1.8874843291021478e-05, "loss": 0.0306, "step": 72820 }, { "grad_norm": 0.5669712424278259, "learning_rate": 1.8861904652211853e-05, "loss": 0.0231, "step": 72830 }, { "grad_norm": 0.37655389308929443, "learning_rate": 1.8848969418617507e-05, "loss": 0.0257, "step": 72840 }, { "grad_norm": 0.2982751429080963, "learning_rate": 1.8836037591653044e-05, "loss": 0.0283, "step": 72850 }, { "grad_norm": 0.37801995873451233, "learning_rate": 1.882310917273262e-05, "loss": 0.0256, "step": 72860 }, { "grad_norm": 0.7078540325164795, "learning_rate": 1.881018416327014e-05, "loss": 0.0251, "step": 72870 }, { "grad_norm": 0.2526775598526001, "learning_rate": 1.8797262564679014e-05, "loss": 0.0254, "step": 72880 }, { "grad_norm": 0.39468881487846375, "learning_rate": 1.8784344378372355e-05, "loss": 0.0204, "step": 72890 }, { "grad_norm": 0.21336770057678223, "learning_rate": 1.8771429605762843e-05, "loss": 0.0256, "step": 72900 }, { "grad_norm": 0.3034725785255432, "learning_rate": 1.8758518248262835e-05, "loss": 0.0273, "step": 72910 }, { "grad_norm": 0.8999172449111938, "learning_rate": 1.8745610307284305e-05, "loss": 0.0334, "step": 72920 }, { "grad_norm": 0.4473761022090912, "learning_rate": 1.8732705784238862e-05, "loss": 0.0233, "step": 72930 }, { "grad_norm": 0.2747601568698883, "learning_rate": 1.8719804680537682e-05, "loss": 0.0265, "step": 72940 }, { "grad_norm": 0.4122065007686615, "learning_rate": 1.870690699759164e-05, "loss": 0.0267, "step": 72950 }, { "grad_norm": 0.391583651304245, "learning_rate": 1.8694012736811213e-05, "loss": 0.0285, "step": 72960 }, { "grad_norm": 0.24544976651668549, "learning_rate": 1.8681121899606463e-05, "loss": 0.0274, "step": 72970 }, { "grad_norm": 0.14553426206111908, "learning_rate": 1.8668234487387144e-05, "loss": 0.0246, "step": 72980 }, { "grad_norm": 0.3103868365287781, "learning_rate": 1.8655350501562592e-05, "loss": 0.0249, "step": 72990 }, { "grad_norm": 0.2951084077358246, "learning_rate": 1.8642469943541796e-05, "loss": 0.0354, "step": 73000 }, { "grad_norm": 0.2608885169029236, "learning_rate": 1.8629592814733328e-05, "loss": 0.0237, "step": 73010 }, { "grad_norm": 0.2558518946170807, "learning_rate": 1.8616719116545438e-05, "loss": 0.0212, "step": 73020 }, { "grad_norm": 0.3586437702178955, "learning_rate": 1.8603848850385952e-05, "loss": 0.0214, "step": 73030 }, { "grad_norm": 0.2290414571762085, "learning_rate": 1.859098201766236e-05, "loss": 0.0225, "step": 73040 }, { "grad_norm": 0.2940036952495575, "learning_rate": 1.8578118619781726e-05, "loss": 0.0337, "step": 73050 }, { "grad_norm": 0.3526044189929962, "learning_rate": 1.856525865815083e-05, "loss": 0.025, "step": 73060 }, { "grad_norm": 0.23739033937454224, "learning_rate": 1.8552402134175966e-05, "loss": 0.017, "step": 73070 }, { "grad_norm": 0.2591753304004669, "learning_rate": 1.8539549049263143e-05, "loss": 0.0253, "step": 73080 }, { "grad_norm": 0.4479406774044037, "learning_rate": 1.8526699404817917e-05, "loss": 0.0334, "step": 73090 }, { "grad_norm": 0.5567077398300171, "learning_rate": 1.8513853202245546e-05, "loss": 0.0266, "step": 73100 }, { "grad_norm": 0.6258291602134705, "learning_rate": 1.850101044295081e-05, "loss": 0.0225, "step": 73110 }, { "grad_norm": 0.5978832840919495, "learning_rate": 1.848817112833825e-05, "loss": 0.0295, "step": 73120 }, { "grad_norm": 0.3682095408439636, "learning_rate": 1.8475335259811905e-05, "loss": 0.0235, "step": 73130 }, { "grad_norm": 0.4340178370475769, "learning_rate": 1.8462502838775508e-05, "loss": 0.0311, "step": 73140 }, { "grad_norm": 0.31181028485298157, "learning_rate": 1.8449673866632378e-05, "loss": 0.0297, "step": 73150 }, { "grad_norm": 0.16843216121196747, "learning_rate": 1.8436848344785473e-05, "loss": 0.0326, "step": 73160 }, { "grad_norm": 0.32234057784080505, "learning_rate": 1.8424026274637378e-05, "loss": 0.0224, "step": 73170 }, { "grad_norm": 0.3754141330718994, "learning_rate": 1.8411207657590312e-05, "loss": 0.024, "step": 73180 }, { "grad_norm": 0.4080677926540375, "learning_rate": 1.839839249504607e-05, "loss": 0.0305, "step": 73190 }, { "grad_norm": 0.3824409544467926, "learning_rate": 1.8385580788406108e-05, "loss": 0.0255, "step": 73200 }, { "grad_norm": 0.23964440822601318, "learning_rate": 1.837277253907152e-05, "loss": 0.0245, "step": 73210 }, { "grad_norm": 0.42720890045166016, "learning_rate": 1.8359967748442957e-05, "loss": 0.0266, "step": 73220 }, { "grad_norm": 0.30560484528541565, "learning_rate": 1.8347166417920774e-05, "loss": 0.0179, "step": 73230 }, { "grad_norm": 0.3585347831249237, "learning_rate": 1.833436854890484e-05, "loss": 0.0288, "step": 73240 }, { "grad_norm": 0.4506855905056, "learning_rate": 1.8321574142794796e-05, "loss": 0.0278, "step": 73250 }, { "grad_norm": 0.30457568168640137, "learning_rate": 1.8308783200989748e-05, "loss": 0.0234, "step": 73260 }, { "grad_norm": 0.330973356962204, "learning_rate": 1.829599572488855e-05, "loss": 0.02, "step": 73270 }, { "grad_norm": 0.2853878140449524, "learning_rate": 1.828321171588957e-05, "loss": 0.0216, "step": 73280 }, { "grad_norm": 0.6764575839042664, "learning_rate": 1.827043117539089e-05, "loss": 0.0318, "step": 73290 }, { "grad_norm": 0.5492085814476013, "learning_rate": 1.8257654104790116e-05, "loss": 0.0237, "step": 73300 }, { "grad_norm": 0.3059310019016266, "learning_rate": 1.82448805054846e-05, "loss": 0.0225, "step": 73310 }, { "grad_norm": 0.38937750458717346, "learning_rate": 1.823211037887119e-05, "loss": 0.0317, "step": 73320 }, { "grad_norm": 0.40738412737846375, "learning_rate": 1.821934372634645e-05, "loss": 0.0192, "step": 73330 }, { "grad_norm": 0.5192722082138062, "learning_rate": 1.8206580549306472e-05, "loss": 0.0206, "step": 73340 }, { "grad_norm": 0.1700410693883896, "learning_rate": 1.819382084914707e-05, "loss": 0.0166, "step": 73350 }, { "grad_norm": 0.2621970474720001, "learning_rate": 1.8181064627263577e-05, "loss": 0.0217, "step": 73360 }, { "grad_norm": 0.43672144412994385, "learning_rate": 1.8168311885051022e-05, "loss": 0.0321, "step": 73370 }, { "grad_norm": 0.2818705141544342, "learning_rate": 1.815556262390402e-05, "loss": 0.0214, "step": 73380 }, { "grad_norm": 0.4037059247493744, "learning_rate": 1.8142816845216826e-05, "loss": 0.0212, "step": 73390 }, { "grad_norm": 0.3454464077949524, "learning_rate": 1.8130074550383276e-05, "loss": 0.0192, "step": 73400 }, { "grad_norm": 0.31585752964019775, "learning_rate": 1.8117335740796848e-05, "loss": 0.0266, "step": 73410 }, { "grad_norm": 0.32705411314964294, "learning_rate": 1.810460041785067e-05, "loss": 0.022, "step": 73420 }, { "grad_norm": 0.23255880177021027, "learning_rate": 1.809186858293742e-05, "loss": 0.0384, "step": 73430 }, { "grad_norm": 0.3416517972946167, "learning_rate": 1.8079140237449448e-05, "loss": 0.0245, "step": 73440 }, { "grad_norm": 0.4158470034599304, "learning_rate": 1.8066415382778707e-05, "loss": 0.0259, "step": 73450 }, { "grad_norm": 0.2861178517341614, "learning_rate": 1.8053694020316786e-05, "loss": 0.0225, "step": 73460 }, { "grad_norm": 0.36993205547332764, "learning_rate": 1.8040976151454837e-05, "loss": 0.0247, "step": 73470 }, { "grad_norm": 0.37422212958335876, "learning_rate": 1.8028261777583704e-05, "loss": 0.0288, "step": 73480 }, { "grad_norm": 0.3755371570587158, "learning_rate": 1.8015550900093758e-05, "loss": 0.0184, "step": 73490 }, { "grad_norm": 0.3455173075199127, "learning_rate": 1.800284352037512e-05, "loss": 0.0347, "step": 73500 }, { "grad_norm": 0.28334668278694153, "learning_rate": 1.799013963981738e-05, "loss": 0.022, "step": 73510 }, { "grad_norm": 0.3926714062690735, "learning_rate": 1.797743925980987e-05, "loss": 0.03, "step": 73520 }, { "grad_norm": 0.4036639630794525, "learning_rate": 1.7964742381741435e-05, "loss": 0.0225, "step": 73530 }, { "grad_norm": 0.4085574746131897, "learning_rate": 1.7952049007000627e-05, "loss": 0.0235, "step": 73540 }, { "grad_norm": 0.39122509956359863, "learning_rate": 1.7939359136975535e-05, "loss": 0.0244, "step": 73550 }, { "grad_norm": 0.22433365881443024, "learning_rate": 1.7926672773053932e-05, "loss": 0.0264, "step": 73560 }, { "grad_norm": 0.36885011196136475, "learning_rate": 1.791398991662317e-05, "loss": 0.0366, "step": 73570 }, { "grad_norm": 0.2254299372434616, "learning_rate": 1.7901310569070244e-05, "loss": 0.0244, "step": 73580 }, { "grad_norm": 0.9381552934646606, "learning_rate": 1.7888634731781724e-05, "loss": 0.031, "step": 73590 }, { "grad_norm": 0.36766859889030457, "learning_rate": 1.7875962406143844e-05, "loss": 0.02, "step": 73600 }, { "grad_norm": 0.18234886229038239, "learning_rate": 1.7863293593542403e-05, "loss": 0.0351, "step": 73610 }, { "grad_norm": 0.32923200726509094, "learning_rate": 1.7850628295362853e-05, "loss": 0.0261, "step": 73620 }, { "grad_norm": 0.5387689471244812, "learning_rate": 1.7837966512990263e-05, "loss": 0.0327, "step": 73630 }, { "grad_norm": 0.34761297702789307, "learning_rate": 1.782530824780931e-05, "loss": 0.0253, "step": 73640 }, { "grad_norm": 0.37814030051231384, "learning_rate": 1.7812653501204263e-05, "loss": 0.0296, "step": 73650 }, { "grad_norm": 0.27833425998687744, "learning_rate": 1.780000227455903e-05, "loss": 0.0251, "step": 73660 }, { "grad_norm": 0.531846821308136, "learning_rate": 1.778735456925716e-05, "loss": 0.0267, "step": 73670 }, { "grad_norm": 0.4068792164325714, "learning_rate": 1.7774710386681737e-05, "loss": 0.0317, "step": 73680 }, { "grad_norm": 1.6513758897781372, "learning_rate": 1.7762069728215545e-05, "loss": 0.0268, "step": 73690 }, { "grad_norm": 0.4991103708744049, "learning_rate": 1.7749432595240932e-05, "loss": 0.0191, "step": 73700 }, { "grad_norm": 0.3609210252761841, "learning_rate": 1.773679898913991e-05, "loss": 0.0351, "step": 73710 }, { "grad_norm": 0.460634708404541, "learning_rate": 1.7724168911294015e-05, "loss": 0.0235, "step": 73720 }, { "grad_norm": 0.4446299374103546, "learning_rate": 1.7711542363084504e-05, "loss": 0.0292, "step": 73730 }, { "grad_norm": 0.3131772577762604, "learning_rate": 1.7698919345892158e-05, "loss": 0.0221, "step": 73740 }, { "grad_norm": 0.600624144077301, "learning_rate": 1.768629986109743e-05, "loss": 0.0203, "step": 73750 }, { "grad_norm": 0.4443891644477844, "learning_rate": 1.7673683910080357e-05, "loss": 0.0314, "step": 73760 }, { "grad_norm": 0.28065451979637146, "learning_rate": 1.7661071494220638e-05, "loss": 0.0216, "step": 73770 }, { "grad_norm": 0.2326432317495346, "learning_rate": 1.7648462614897498e-05, "loss": 0.0227, "step": 73780 }, { "grad_norm": 0.45985496044158936, "learning_rate": 1.7635857273489863e-05, "loss": 0.0313, "step": 73790 }, { "grad_norm": 0.375161737203598, "learning_rate": 1.7623255471376198e-05, "loss": 0.0295, "step": 73800 }, { "grad_norm": 0.14200888574123383, "learning_rate": 1.7610657209934646e-05, "loss": 0.0368, "step": 73810 }, { "grad_norm": 0.2952873706817627, "learning_rate": 1.7598062490542915e-05, "loss": 0.0302, "step": 73820 }, { "grad_norm": 0.26267096400260925, "learning_rate": 1.7585471314578377e-05, "loss": 0.0269, "step": 73830 }, { "grad_norm": 0.4047735333442688, "learning_rate": 1.7572883683417945e-05, "loss": 0.0287, "step": 73840 }, { "grad_norm": 0.2883511483669281, "learning_rate": 1.7560299598438212e-05, "loss": 0.0269, "step": 73850 }, { "grad_norm": 0.34265127778053284, "learning_rate": 1.7547719061015334e-05, "loss": 0.0306, "step": 73860 }, { "grad_norm": 0.2614176869392395, "learning_rate": 1.75351420725251e-05, "loss": 0.0232, "step": 73870 }, { "grad_norm": 0.27539220452308655, "learning_rate": 1.7522568634342923e-05, "loss": 0.0286, "step": 73880 }, { "grad_norm": 0.2998487055301666, "learning_rate": 1.7509998747843825e-05, "loss": 0.02, "step": 73890 }, { "grad_norm": 0.1825304925441742, "learning_rate": 1.7497432414402403e-05, "loss": 0.025, "step": 73900 }, { "grad_norm": 0.5067284107208252, "learning_rate": 1.7484869635392903e-05, "loss": 0.0219, "step": 73910 }, { "grad_norm": 0.32971611618995667, "learning_rate": 1.747231041218919e-05, "loss": 0.026, "step": 73920 }, { "grad_norm": 0.33928561210632324, "learning_rate": 1.7459754746164692e-05, "loss": 0.0187, "step": 73930 }, { "grad_norm": 0.3073003888130188, "learning_rate": 1.744720263869249e-05, "loss": 0.0295, "step": 73940 }, { "grad_norm": 0.26690247654914856, "learning_rate": 1.7434654091145263e-05, "loss": 0.0258, "step": 73950 }, { "grad_norm": 0.317942351102829, "learning_rate": 1.742210910489532e-05, "loss": 0.0323, "step": 73960 }, { "grad_norm": 0.33471572399139404, "learning_rate": 1.7409567681314532e-05, "loss": 0.0263, "step": 73970 }, { "grad_norm": 0.8324373960494995, "learning_rate": 1.7397029821774434e-05, "loss": 0.0247, "step": 73980 }, { "grad_norm": 0.4094406068325043, "learning_rate": 1.7384495527646126e-05, "loss": 0.0203, "step": 73990 }, { "grad_norm": 0.18406619131565094, "learning_rate": 1.737196480030035e-05, "loss": 0.0296, "step": 74000 }, { "grad_norm": 0.3692528009414673, "learning_rate": 1.7359437641107447e-05, "loss": 0.0245, "step": 74010 }, { "grad_norm": 0.333565354347229, "learning_rate": 1.7346914051437387e-05, "loss": 0.027, "step": 74020 }, { "grad_norm": 0.6271889805793762, "learning_rate": 1.733439403265969e-05, "loss": 0.0221, "step": 74030 }, { "grad_norm": 0.38821491599082947, "learning_rate": 1.732187758614357e-05, "loss": 0.0338, "step": 74040 }, { "grad_norm": 0.6276283860206604, "learning_rate": 1.730936471325777e-05, "loss": 0.0221, "step": 74050 }, { "grad_norm": 0.36215177178382874, "learning_rate": 1.7296855415370718e-05, "loss": 0.0318, "step": 74060 }, { "grad_norm": 0.4396142363548279, "learning_rate": 1.728434969385035e-05, "loss": 0.0207, "step": 74070 }, { "grad_norm": 0.2526659667491913, "learning_rate": 1.7271847550064346e-05, "loss": 0.0147, "step": 74080 }, { "grad_norm": 0.27109456062316895, "learning_rate": 1.725934898537987e-05, "loss": 0.0226, "step": 74090 }, { "grad_norm": 0.45123401284217834, "learning_rate": 1.7246854001163783e-05, "loss": 0.0233, "step": 74100 }, { "grad_norm": 0.8079216480255127, "learning_rate": 1.7234362598782476e-05, "loss": 0.0229, "step": 74110 }, { "grad_norm": 0.40895646810531616, "learning_rate": 1.722187477960202e-05, "loss": 0.0313, "step": 74120 }, { "grad_norm": 0.319911390542984, "learning_rate": 1.7209390544988057e-05, "loss": 0.0272, "step": 74130 }, { "grad_norm": 0.25586792826652527, "learning_rate": 1.7196909896305856e-05, "loss": 0.0422, "step": 74140 }, { "grad_norm": 0.21549037098884583, "learning_rate": 1.718443283492026e-05, "loss": 0.0273, "step": 74150 }, { "grad_norm": 0.27238595485687256, "learning_rate": 1.7171959362195746e-05, "loss": 0.0276, "step": 74160 }, { "grad_norm": 0.3298286497592926, "learning_rate": 1.715948947949642e-05, "loss": 0.0253, "step": 74170 }, { "grad_norm": 1.595070481300354, "learning_rate": 1.714702318818593e-05, "loss": 0.0277, "step": 74180 }, { "grad_norm": 0.8193837404251099, "learning_rate": 1.7134560489627592e-05, "loss": 0.0263, "step": 74190 }, { "grad_norm": 0.5214722752571106, "learning_rate": 1.712210138518431e-05, "loss": 0.0333, "step": 74200 }, { "grad_norm": 0.33665281534194946, "learning_rate": 1.71096458762186e-05, "loss": 0.0313, "step": 74210 }, { "grad_norm": 0.44848498702049255, "learning_rate": 1.7097193964092552e-05, "loss": 0.0263, "step": 74220 }, { "grad_norm": 0.2859601080417633, "learning_rate": 1.708474565016792e-05, "loss": 0.0252, "step": 74230 }, { "grad_norm": 0.29621630907058716, "learning_rate": 1.7072300935806e-05, "loss": 0.0243, "step": 74240 }, { "grad_norm": 0.35346904397010803, "learning_rate": 1.705985982236776e-05, "loss": 0.0194, "step": 74250 }, { "grad_norm": 0.18821491301059723, "learning_rate": 1.7047422311213697e-05, "loss": 0.021, "step": 74260 }, { "grad_norm": 0.35770389437675476, "learning_rate": 1.703498840370401e-05, "loss": 0.0311, "step": 74270 }, { "grad_norm": 0.477766215801239, "learning_rate": 1.702255810119842e-05, "loss": 0.0296, "step": 74280 }, { "grad_norm": 0.17124271392822266, "learning_rate": 1.7010131405056305e-05, "loss": 0.0221, "step": 74290 }, { "grad_norm": 0.3406882584095001, "learning_rate": 1.6997708316636613e-05, "loss": 0.0265, "step": 74300 }, { "grad_norm": 0.3020051419734955, "learning_rate": 1.6985288837297932e-05, "loss": 0.0295, "step": 74310 }, { "grad_norm": 0.3806108832359314, "learning_rate": 1.6972872968398394e-05, "loss": 0.0348, "step": 74320 }, { "grad_norm": 0.3857825994491577, "learning_rate": 1.6960460711295845e-05, "loss": 0.0223, "step": 74330 }, { "grad_norm": 0.9074849486351013, "learning_rate": 1.6948052067347623e-05, "loss": 0.0249, "step": 74340 }, { "grad_norm": 0.2555720806121826, "learning_rate": 1.693564703791075e-05, "loss": 0.0242, "step": 74350 }, { "grad_norm": 0.6122108697891235, "learning_rate": 1.6923245624341793e-05, "loss": 0.0214, "step": 74360 }, { "grad_norm": 2.1312108039855957, "learning_rate": 1.6910847827996962e-05, "loss": 0.0369, "step": 74370 }, { "grad_norm": 1.438517451286316, "learning_rate": 1.689845365023206e-05, "loss": 0.0214, "step": 74380 }, { "grad_norm": 0.541698694229126, "learning_rate": 1.6886063092402505e-05, "loss": 0.024, "step": 74390 }, { "grad_norm": 0.40495750308036804, "learning_rate": 1.6873676155863317e-05, "loss": 0.0237, "step": 74400 }, { "grad_norm": 0.4397534430027008, "learning_rate": 1.6861292841969085e-05, "loss": 0.0229, "step": 74410 }, { "grad_norm": 0.3310600817203522, "learning_rate": 1.684891315207406e-05, "loss": 0.0237, "step": 74420 }, { "grad_norm": 0.25689491629600525, "learning_rate": 1.6836537087532034e-05, "loss": 0.0241, "step": 74430 }, { "grad_norm": 0.2640324532985687, "learning_rate": 1.6824164649696465e-05, "loss": 0.0347, "step": 74440 }, { "grad_norm": 0.3783245086669922, "learning_rate": 1.681179583992034e-05, "loss": 0.0205, "step": 74450 }, { "grad_norm": 0.16622084379196167, "learning_rate": 1.679943065955636e-05, "loss": 0.02, "step": 74460 }, { "grad_norm": 0.20285047590732574, "learning_rate": 1.678706910995671e-05, "loss": 0.0207, "step": 74470 }, { "grad_norm": 0.2606956958770752, "learning_rate": 1.6774711192473262e-05, "loss": 0.0251, "step": 74480 }, { "grad_norm": 0.2851451635360718, "learning_rate": 1.676235690845743e-05, "loss": 0.0205, "step": 74490 }, { "grad_norm": 0.512169599533081, "learning_rate": 1.6750006259260285e-05, "loss": 0.0291, "step": 74500 }, { "grad_norm": 0.33937525749206543, "learning_rate": 1.673765924623244e-05, "loss": 0.0232, "step": 74510 }, { "grad_norm": 0.4351532757282257, "learning_rate": 1.67253158707242e-05, "loss": 0.0253, "step": 74520 }, { "grad_norm": 0.4009709358215332, "learning_rate": 1.671297613408537e-05, "loss": 0.0206, "step": 74530 }, { "grad_norm": 0.21713818609714508, "learning_rate": 1.6700640037665444e-05, "loss": 0.0154, "step": 74540 }, { "grad_norm": 0.2671957314014435, "learning_rate": 1.668830758281344e-05, "loss": 0.0261, "step": 74550 }, { "grad_norm": 0.9434975981712341, "learning_rate": 1.6675978770878052e-05, "loss": 0.0257, "step": 74560 }, { "grad_norm": 1.0870871543884277, "learning_rate": 1.6663653603207506e-05, "loss": 0.026, "step": 74570 }, { "grad_norm": 0.2634543180465698, "learning_rate": 1.6651332081149685e-05, "loss": 0.0231, "step": 74580 }, { "grad_norm": 1.8122657537460327, "learning_rate": 1.6639014206052046e-05, "loss": 0.0253, "step": 74590 }, { "grad_norm": 0.28309166431427, "learning_rate": 1.6626699979261657e-05, "loss": 0.025, "step": 74600 }, { "grad_norm": 1.0629055500030518, "learning_rate": 1.6614389402125197e-05, "loss": 0.0267, "step": 74610 }, { "grad_norm": 0.4855862557888031, "learning_rate": 1.66020824759889e-05, "loss": 0.0269, "step": 74620 }, { "grad_norm": 0.33970120549201965, "learning_rate": 1.6589779202198675e-05, "loss": 0.0215, "step": 74630 }, { "grad_norm": 0.3024961054325104, "learning_rate": 1.6577479582099924e-05, "loss": 0.0296, "step": 74640 }, { "grad_norm": 0.3769637942314148, "learning_rate": 1.6565183617037788e-05, "loss": 0.0362, "step": 74650 }, { "grad_norm": 0.42975473403930664, "learning_rate": 1.6552891308356894e-05, "loss": 0.027, "step": 74660 }, { "grad_norm": 0.36227282881736755, "learning_rate": 1.6540602657401526e-05, "loss": 0.0353, "step": 74670 }, { "grad_norm": 0.40469077229499817, "learning_rate": 1.6528317665515537e-05, "loss": 0.0297, "step": 74680 }, { "grad_norm": 0.22207669913768768, "learning_rate": 1.6516036334042416e-05, "loss": 0.03, "step": 74690 }, { "grad_norm": 0.26728031039237976, "learning_rate": 1.650375866432519e-05, "loss": 0.0268, "step": 74700 }, { "grad_norm": 0.2777063250541687, "learning_rate": 1.649148465770659e-05, "loss": 0.0239, "step": 74710 }, { "grad_norm": 0.3631337583065033, "learning_rate": 1.647921431552883e-05, "loss": 0.0149, "step": 74720 }, { "grad_norm": 0.4981934726238251, "learning_rate": 1.646694763913381e-05, "loss": 0.034, "step": 74730 }, { "grad_norm": 0.22490906715393066, "learning_rate": 1.6454684629862966e-05, "loss": 0.0192, "step": 74740 }, { "grad_norm": 0.3341204524040222, "learning_rate": 1.6442425289057388e-05, "loss": 0.0288, "step": 74750 }, { "grad_norm": 0.25434383749961853, "learning_rate": 1.6430169618057718e-05, "loss": 0.0235, "step": 74760 }, { "grad_norm": 0.4314160645008087, "learning_rate": 1.641791761820422e-05, "loss": 0.0249, "step": 74770 }, { "grad_norm": 0.31665340065956116, "learning_rate": 1.640566929083676e-05, "loss": 0.0297, "step": 74780 }, { "grad_norm": 0.3188595473766327, "learning_rate": 1.6393424637294823e-05, "loss": 0.0307, "step": 74790 }, { "grad_norm": 0.41619381308555603, "learning_rate": 1.638118365891742e-05, "loss": 0.029, "step": 74800 }, { "grad_norm": 0.518916666507721, "learning_rate": 1.636894635704323e-05, "loss": 0.0274, "step": 74810 }, { "grad_norm": 0.17708547413349152, "learning_rate": 1.6356712733010517e-05, "loss": 0.0226, "step": 74820 }, { "grad_norm": 0.8484772443771362, "learning_rate": 1.6344482788157096e-05, "loss": 0.0355, "step": 74830 }, { "grad_norm": 0.7408357858657837, "learning_rate": 1.6332256523820444e-05, "loss": 0.0219, "step": 74840 }, { "grad_norm": 0.3018658459186554, "learning_rate": 1.6320033941337597e-05, "loss": 0.0193, "step": 74850 }, { "grad_norm": 0.3511892557144165, "learning_rate": 1.6307815042045216e-05, "loss": 0.0255, "step": 74860 }, { "grad_norm": 0.3441629707813263, "learning_rate": 1.6295599827279513e-05, "loss": 0.0211, "step": 74870 }, { "grad_norm": 0.3264640271663666, "learning_rate": 1.6283388298376356e-05, "loss": 0.0243, "step": 74880 }, { "grad_norm": 0.33018913865089417, "learning_rate": 1.6271180456671125e-05, "loss": 0.0203, "step": 74890 }, { "grad_norm": 0.32139334082603455, "learning_rate": 1.625897630349893e-05, "loss": 0.0229, "step": 74900 }, { "grad_norm": 0.30269309878349304, "learning_rate": 1.6246775840194335e-05, "loss": 0.027, "step": 74910 }, { "grad_norm": 0.3996983468532562, "learning_rate": 1.6234579068091604e-05, "loss": 0.0227, "step": 74920 }, { "grad_norm": 0.31006449460983276, "learning_rate": 1.6222385988524524e-05, "loss": 0.0241, "step": 74930 }, { "grad_norm": 0.3715326488018036, "learning_rate": 1.6210196602826544e-05, "loss": 0.0197, "step": 74940 }, { "grad_norm": 0.3698221743106842, "learning_rate": 1.6198010912330646e-05, "loss": 0.0292, "step": 74950 }, { "grad_norm": 0.31654706597328186, "learning_rate": 1.6185828918369456e-05, "loss": 0.0318, "step": 74960 }, { "grad_norm": 0.33211013674736023, "learning_rate": 1.617365062227517e-05, "loss": 0.0197, "step": 74970 }, { "grad_norm": 0.7014352679252625, "learning_rate": 1.616147602537962e-05, "loss": 0.0298, "step": 74980 }, { "grad_norm": 0.19896799325942993, "learning_rate": 1.614930512901415e-05, "loss": 0.024, "step": 74990 }, { "grad_norm": 0.4170383810997009, "learning_rate": 1.6137137934509804e-05, "loss": 0.0175, "step": 75000 }, { "grad_norm": 0.6031068563461304, "learning_rate": 1.6124974443197123e-05, "loss": 0.0259, "step": 75010 }, { "grad_norm": 0.3158794641494751, "learning_rate": 1.6112814656406306e-05, "loss": 0.0218, "step": 75020 }, { "grad_norm": 0.16260728240013123, "learning_rate": 1.6100658575467125e-05, "loss": 0.0248, "step": 75030 }, { "grad_norm": 0.2850354313850403, "learning_rate": 1.6088506201708974e-05, "loss": 0.0301, "step": 75040 }, { "grad_norm": 0.2506636083126068, "learning_rate": 1.6076357536460785e-05, "loss": 0.024, "step": 75050 }, { "grad_norm": 0.2767799496650696, "learning_rate": 1.606421258105113e-05, "loss": 0.0222, "step": 75060 }, { "grad_norm": 0.41678085923194885, "learning_rate": 1.6052071336808183e-05, "loss": 0.0197, "step": 75070 }, { "grad_norm": 0.4108410179615021, "learning_rate": 1.6039933805059658e-05, "loss": 0.0305, "step": 75080 }, { "grad_norm": 1.2271686792373657, "learning_rate": 1.6027799987132915e-05, "loss": 0.0192, "step": 75090 }, { "grad_norm": 0.37815061211586, "learning_rate": 1.6015669884354888e-05, "loss": 0.0297, "step": 75100 }, { "grad_norm": 0.5451768636703491, "learning_rate": 1.6003543498052125e-05, "loss": 0.0292, "step": 75110 }, { "grad_norm": 0.40988266468048096, "learning_rate": 1.5991420829550715e-05, "loss": 0.025, "step": 75120 }, { "grad_norm": 0.38551241159439087, "learning_rate": 1.5979301880176405e-05, "loss": 0.0284, "step": 75130 }, { "grad_norm": 0.5641183257102966, "learning_rate": 1.5967186651254485e-05, "loss": 0.025, "step": 75140 }, { "grad_norm": 0.2200421392917633, "learning_rate": 1.5955075144109853e-05, "loss": 0.0195, "step": 75150 }, { "grad_norm": 0.35922765731811523, "learning_rate": 1.594296736006703e-05, "loss": 0.0342, "step": 75160 }, { "grad_norm": 0.4380887448787689, "learning_rate": 1.59308633004501e-05, "loss": 0.0407, "step": 75170 }, { "grad_norm": 0.30989548563957214, "learning_rate": 1.5918762966582724e-05, "loss": 0.0189, "step": 75180 }, { "grad_norm": 0.30110594630241394, "learning_rate": 1.5906666359788203e-05, "loss": 0.0182, "step": 75190 }, { "grad_norm": 0.3941165506839752, "learning_rate": 1.589457348138938e-05, "loss": 0.0239, "step": 75200 }, { "grad_norm": 0.6474146246910095, "learning_rate": 1.588248433270872e-05, "loss": 0.0264, "step": 75210 }, { "grad_norm": 0.33791735768318176, "learning_rate": 1.587039891506828e-05, "loss": 0.0286, "step": 75220 }, { "grad_norm": 0.23271308839321136, "learning_rate": 1.5858317229789715e-05, "loss": 0.0259, "step": 75230 }, { "grad_norm": 0.3040444850921631, "learning_rate": 1.584623927819423e-05, "loss": 0.0217, "step": 75240 }, { "grad_norm": 0.3161320984363556, "learning_rate": 1.583416506160269e-05, "loss": 0.0295, "step": 75250 }, { "grad_norm": 0.2284645140171051, "learning_rate": 1.5822094581335472e-05, "loss": 0.0226, "step": 75260 }, { "grad_norm": 0.26909253001213074, "learning_rate": 1.5810027838712604e-05, "loss": 0.0291, "step": 75270 }, { "grad_norm": 0.6212795376777649, "learning_rate": 1.5797964835053687e-05, "loss": 0.026, "step": 75280 }, { "grad_norm": 0.27693766355514526, "learning_rate": 1.5785905571677933e-05, "loss": 0.0305, "step": 75290 }, { "grad_norm": 0.47336339950561523, "learning_rate": 1.5773850049904086e-05, "loss": 0.03, "step": 75300 }, { "grad_norm": 0.3081341087818146, "learning_rate": 1.5761798271050543e-05, "loss": 0.0243, "step": 75310 }, { "grad_norm": 0.4215796887874603, "learning_rate": 1.5749750236435277e-05, "loss": 0.0252, "step": 75320 }, { "grad_norm": 0.3122490346431732, "learning_rate": 1.5737705947375815e-05, "loss": 0.03, "step": 75330 }, { "grad_norm": 0.36702677607536316, "learning_rate": 1.5725665405189322e-05, "loss": 0.0312, "step": 75340 }, { "grad_norm": 0.904097855091095, "learning_rate": 1.571362861119253e-05, "loss": 0.0224, "step": 75350 }, { "grad_norm": 0.34693577885627747, "learning_rate": 1.570159556670177e-05, "loss": 0.021, "step": 75360 }, { "grad_norm": 0.2238789200782776, "learning_rate": 1.5689566273032946e-05, "loss": 0.0225, "step": 75370 }, { "grad_norm": 0.475580096244812, "learning_rate": 1.5677540731501583e-05, "loss": 0.0211, "step": 75380 }, { "grad_norm": 0.4559524357318878, "learning_rate": 1.5665518943422746e-05, "loss": 0.0194, "step": 75390 }, { "grad_norm": 0.5096158385276794, "learning_rate": 1.5653500910111134e-05, "loss": 0.028, "step": 75400 }, { "grad_norm": 0.3456946313381195, "learning_rate": 1.5641486632881023e-05, "loss": 0.0239, "step": 75410 }, { "grad_norm": 0.3806454539299011, "learning_rate": 1.56294761130463e-05, "loss": 0.0265, "step": 75420 }, { "grad_norm": 0.346606969833374, "learning_rate": 1.5617469351920373e-05, "loss": 0.026, "step": 75430 }, { "grad_norm": 0.30333471298217773, "learning_rate": 1.5605466350816315e-05, "loss": 0.0318, "step": 75440 }, { "grad_norm": 0.262743204832077, "learning_rate": 1.559346711104674e-05, "loss": 0.0282, "step": 75450 }, { "grad_norm": 0.33484548330307007, "learning_rate": 1.5581471633923883e-05, "loss": 0.0276, "step": 75460 }, { "grad_norm": 0.4411848485469818, "learning_rate": 1.5569479920759505e-05, "loss": 0.0249, "step": 75470 }, { "grad_norm": 0.34591740369796753, "learning_rate": 1.5557491972865072e-05, "loss": 0.0205, "step": 75480 }, { "grad_norm": 0.3282424509525299, "learning_rate": 1.554550779155152e-05, "loss": 0.022, "step": 75490 }, { "grad_norm": 0.2652110159397125, "learning_rate": 1.5533527378129453e-05, "loss": 0.0242, "step": 75500 }, { "grad_norm": 0.38916754722595215, "learning_rate": 1.552155073390899e-05, "loss": 0.0235, "step": 75510 }, { "grad_norm": 0.4647594392299652, "learning_rate": 1.5509577860199903e-05, "loss": 0.0292, "step": 75520 }, { "grad_norm": 0.2015065997838974, "learning_rate": 1.5497608758311525e-05, "loss": 0.0176, "step": 75530 }, { "grad_norm": 0.4652765095233917, "learning_rate": 1.5485643429552797e-05, "loss": 0.0212, "step": 75540 }, { "grad_norm": 0.7054060697555542, "learning_rate": 1.5473681875232197e-05, "loss": 0.0258, "step": 75550 }, { "grad_norm": 0.559249222278595, "learning_rate": 1.5461724096657837e-05, "loss": 0.0182, "step": 75560 }, { "grad_norm": 0.45905303955078125, "learning_rate": 1.5449770095137416e-05, "loss": 0.0227, "step": 75570 }, { "grad_norm": 0.3057628571987152, "learning_rate": 1.5437819871978178e-05, "loss": 0.0232, "step": 75580 }, { "grad_norm": 0.5152026414871216, "learning_rate": 1.5425873428486992e-05, "loss": 0.0315, "step": 75590 }, { "grad_norm": 0.5448943376541138, "learning_rate": 1.5413930765970312e-05, "loss": 0.0253, "step": 75600 }, { "grad_norm": 0.3589089512825012, "learning_rate": 1.5401991885734175e-05, "loss": 0.026, "step": 75610 }, { "grad_norm": 0.28740477561950684, "learning_rate": 1.5390056789084174e-05, "loss": 0.0351, "step": 75620 }, { "grad_norm": 0.5237617492675781, "learning_rate": 1.5378125477325545e-05, "loss": 0.0218, "step": 75630 }, { "grad_norm": 0.3783079981803894, "learning_rate": 1.5366197951763046e-05, "loss": 0.0257, "step": 75640 }, { "grad_norm": 0.48501989245414734, "learning_rate": 1.5354274213701075e-05, "loss": 0.0242, "step": 75650 }, { "grad_norm": 1.4138530492782593, "learning_rate": 1.5342354264443566e-05, "loss": 0.0257, "step": 75660 }, { "grad_norm": 0.6047191023826599, "learning_rate": 1.533043810529411e-05, "loss": 0.0289, "step": 75670 }, { "grad_norm": 0.4259708523750305, "learning_rate": 1.5318525737555807e-05, "loss": 0.0347, "step": 75680 }, { "grad_norm": 0.4068228006362915, "learning_rate": 1.53066171625314e-05, "loss": 0.0328, "step": 75690 }, { "grad_norm": 0.3161562383174896, "learning_rate": 1.529471238152317e-05, "loss": 0.0314, "step": 75700 }, { "grad_norm": 0.31491464376449585, "learning_rate": 1.5282811395833023e-05, "loss": 0.0227, "step": 75710 }, { "grad_norm": 0.30797988176345825, "learning_rate": 1.5270914206762397e-05, "loss": 0.0217, "step": 75720 }, { "grad_norm": 0.7530409693717957, "learning_rate": 1.525902081561241e-05, "loss": 0.0227, "step": 75730 }, { "grad_norm": 0.36702993512153625, "learning_rate": 1.5247131223683664e-05, "loss": 0.0234, "step": 75740 }, { "grad_norm": 0.4053744673728943, "learning_rate": 1.5235245432276408e-05, "loss": 0.0217, "step": 75750 }, { "grad_norm": 0.3708547055721283, "learning_rate": 1.5223363442690424e-05, "loss": 0.0171, "step": 75760 }, { "grad_norm": 0.3993573784828186, "learning_rate": 1.5211485256225133e-05, "loss": 0.0244, "step": 75770 }, { "grad_norm": 0.39147529006004333, "learning_rate": 1.519961087417951e-05, "loss": 0.0285, "step": 75780 }, { "grad_norm": 0.3225889205932617, "learning_rate": 1.5187740297852137e-05, "loss": 0.0271, "step": 75790 }, { "grad_norm": 0.31639838218688965, "learning_rate": 1.517587352854113e-05, "loss": 0.0189, "step": 75800 }, { "grad_norm": 0.2988170385360718, "learning_rate": 1.516401056754423e-05, "loss": 0.0232, "step": 75810 }, { "grad_norm": 0.4103865325450897, "learning_rate": 1.5152151416158783e-05, "loss": 0.0324, "step": 75820 }, { "grad_norm": 0.5613588690757751, "learning_rate": 1.5140296075681642e-05, "loss": 0.0241, "step": 75830 }, { "grad_norm": 0.5398033261299133, "learning_rate": 1.5128444547409331e-05, "loss": 0.0326, "step": 75840 }, { "grad_norm": 0.28998008370399475, "learning_rate": 1.5116596832637864e-05, "loss": 0.0246, "step": 75850 }, { "grad_norm": 0.5511002540588379, "learning_rate": 1.5104752932662952e-05, "loss": 0.0272, "step": 75860 }, { "grad_norm": 0.4511701464653015, "learning_rate": 1.5092912848779778e-05, "loss": 0.0179, "step": 75870 }, { "grad_norm": 0.40395209193229675, "learning_rate": 1.5081076582283194e-05, "loss": 0.0223, "step": 75880 }, { "grad_norm": 0.29641780257225037, "learning_rate": 1.5069244134467553e-05, "loss": 0.0258, "step": 75890 }, { "grad_norm": 0.2784181237220764, "learning_rate": 1.5057415506626876e-05, "loss": 0.0244, "step": 75900 }, { "grad_norm": 0.39792129397392273, "learning_rate": 1.5045590700054668e-05, "loss": 0.0247, "step": 75910 }, { "grad_norm": 0.3785562515258789, "learning_rate": 1.5033769716044144e-05, "loss": 0.0186, "step": 75920 }, { "grad_norm": 0.3207397758960724, "learning_rate": 1.5021952555887975e-05, "loss": 0.0233, "step": 75930 }, { "grad_norm": 0.328625351190567, "learning_rate": 1.50101392208785e-05, "loss": 0.021, "step": 75940 }, { "grad_norm": 0.36145254969596863, "learning_rate": 1.4998329712307579e-05, "loss": 0.0215, "step": 75950 }, { "grad_norm": 0.3971088230609894, "learning_rate": 1.4986524031466708e-05, "loss": 0.0194, "step": 75960 }, { "grad_norm": 1.4487075805664062, "learning_rate": 1.4974722179646916e-05, "loss": 0.0214, "step": 75970 }, { "grad_norm": 0.393169105052948, "learning_rate": 1.4962924158138847e-05, "loss": 0.0268, "step": 75980 }, { "grad_norm": 0.33537769317626953, "learning_rate": 1.4951129968232714e-05, "loss": 0.0158, "step": 75990 }, { "grad_norm": 0.3930119276046753, "learning_rate": 1.4939339611218329e-05, "loss": 0.0215, "step": 76000 }, { "grad_norm": 0.480964332818985, "learning_rate": 1.4927553088385032e-05, "loss": 0.0246, "step": 76010 }, { "grad_norm": 0.41890549659729004, "learning_rate": 1.4915770401021801e-05, "loss": 0.0203, "step": 76020 }, { "grad_norm": 0.38364192843437195, "learning_rate": 1.4903991550417185e-05, "loss": 0.0218, "step": 76030 }, { "grad_norm": 0.5483102798461914, "learning_rate": 1.4892216537859277e-05, "loss": 0.0312, "step": 76040 }, { "grad_norm": 0.3653157651424408, "learning_rate": 1.4880445364635775e-05, "loss": 0.0202, "step": 76050 }, { "grad_norm": 0.5359356999397278, "learning_rate": 1.4868678032033973e-05, "loss": 0.0298, "step": 76060 }, { "grad_norm": 0.216228649020195, "learning_rate": 1.4856914541340739e-05, "loss": 0.0199, "step": 76070 }, { "grad_norm": 0.392536461353302, "learning_rate": 1.4845154893842472e-05, "loss": 0.0256, "step": 76080 }, { "grad_norm": 0.642133891582489, "learning_rate": 1.4833399090825234e-05, "loss": 0.0298, "step": 76090 }, { "grad_norm": 0.30420491099357605, "learning_rate": 1.4821647133574562e-05, "loss": 0.0252, "step": 76100 }, { "grad_norm": 0.40431344509124756, "learning_rate": 1.4809899023375706e-05, "loss": 0.0214, "step": 76110 }, { "grad_norm": 0.3359854817390442, "learning_rate": 1.4798154761513366e-05, "loss": 0.0222, "step": 76120 }, { "grad_norm": 0.5439251065254211, "learning_rate": 1.4786414349271916e-05, "loss": 0.0253, "step": 76130 }, { "grad_norm": 0.5068089365959167, "learning_rate": 1.4774677787935226e-05, "loss": 0.023, "step": 76140 }, { "grad_norm": 0.26560014486312866, "learning_rate": 1.4762945078786838e-05, "loss": 0.0267, "step": 76150 }, { "grad_norm": 0.27354297041893005, "learning_rate": 1.4751216223109777e-05, "loss": 0.0254, "step": 76160 }, { "grad_norm": 0.4816944897174835, "learning_rate": 1.4739491222186714e-05, "loss": 0.0227, "step": 76170 }, { "grad_norm": 0.23450565338134766, "learning_rate": 1.4727770077299879e-05, "loss": 0.029, "step": 76180 }, { "grad_norm": 0.322982519865036, "learning_rate": 1.4716052789731094e-05, "loss": 0.0263, "step": 76190 }, { "grad_norm": 0.3381555676460266, "learning_rate": 1.470433936076171e-05, "loss": 0.0173, "step": 76200 }, { "grad_norm": 0.31254813075065613, "learning_rate": 1.4692629791672719e-05, "loss": 0.0228, "step": 76210 }, { "grad_norm": 0.42732754349708557, "learning_rate": 1.468092408374464e-05, "loss": 0.0215, "step": 76220 }, { "grad_norm": 0.4018895924091339, "learning_rate": 1.4669222238257602e-05, "loss": 0.0239, "step": 76230 }, { "grad_norm": 0.22093446552753448, "learning_rate": 1.4657524256491301e-05, "loss": 0.0199, "step": 76240 }, { "grad_norm": 0.29441624879837036, "learning_rate": 1.464583013972502e-05, "loss": 0.0205, "step": 76250 }, { "grad_norm": 0.23840628564357758, "learning_rate": 1.4634139889237591e-05, "loss": 0.0163, "step": 76260 }, { "grad_norm": 0.2988588511943817, "learning_rate": 1.4622453506307448e-05, "loss": 0.0171, "step": 76270 }, { "grad_norm": 0.40105992555618286, "learning_rate": 1.461077099221262e-05, "loss": 0.0209, "step": 76280 }, { "grad_norm": 0.5155014991760254, "learning_rate": 1.4599092348230652e-05, "loss": 0.0244, "step": 76290 }, { "grad_norm": 0.4393501281738281, "learning_rate": 1.4587417575638724e-05, "loss": 0.0224, "step": 76300 }, { "grad_norm": 0.3202104866504669, "learning_rate": 1.4575746675713563e-05, "loss": 0.0259, "step": 76310 }, { "grad_norm": 0.2607320249080658, "learning_rate": 1.4564079649731504e-05, "loss": 0.0356, "step": 76320 }, { "grad_norm": 0.9918129444122314, "learning_rate": 1.4552416498968402e-05, "loss": 0.0221, "step": 76330 }, { "grad_norm": 0.25926312804222107, "learning_rate": 1.4540757224699747e-05, "loss": 0.0238, "step": 76340 }, { "grad_norm": 0.4398413598537445, "learning_rate": 1.452910182820056e-05, "loss": 0.021, "step": 76350 }, { "grad_norm": 0.33055952191352844, "learning_rate": 1.4517450310745468e-05, "loss": 0.0197, "step": 76360 }, { "grad_norm": 0.4155871868133545, "learning_rate": 1.450580267360866e-05, "loss": 0.024, "step": 76370 }, { "grad_norm": 0.6686175465583801, "learning_rate": 1.4494158918063922e-05, "loss": 0.0183, "step": 76380 }, { "grad_norm": 0.31712883710861206, "learning_rate": 1.4482519045384568e-05, "loss": 0.0273, "step": 76390 }, { "grad_norm": 0.27876362204551697, "learning_rate": 1.4470883056843548e-05, "loss": 0.0276, "step": 76400 }, { "grad_norm": 0.3251400291919708, "learning_rate": 1.4459250953713322e-05, "loss": 0.0226, "step": 76410 }, { "grad_norm": 0.24793395400047302, "learning_rate": 1.444762273726598e-05, "loss": 0.0237, "step": 76420 }, { "grad_norm": 1.3618011474609375, "learning_rate": 1.443599840877316e-05, "loss": 0.0213, "step": 76430 }, { "grad_norm": 0.2867167294025421, "learning_rate": 1.4424377969506103e-05, "loss": 0.0204, "step": 76440 }, { "grad_norm": 0.26938873529434204, "learning_rate": 1.4412761420735565e-05, "loss": 0.019, "step": 76450 }, { "grad_norm": 0.40755024552345276, "learning_rate": 1.4401148763731953e-05, "loss": 0.0274, "step": 76460 }, { "grad_norm": 0.2794654369354248, "learning_rate": 1.4389539999765173e-05, "loss": 0.0174, "step": 76470 }, { "grad_norm": 0.3358205556869507, "learning_rate": 1.4377935130104758e-05, "loss": 0.0217, "step": 76480 }, { "grad_norm": 0.7054564952850342, "learning_rate": 1.4366334156019801e-05, "loss": 0.0358, "step": 76490 }, { "grad_norm": 0.31618404388427734, "learning_rate": 1.435473707877898e-05, "loss": 0.02, "step": 76500 }, { "grad_norm": 0.25238513946533203, "learning_rate": 1.4343143899650508e-05, "loss": 0.0205, "step": 76510 }, { "grad_norm": 0.2757248878479004, "learning_rate": 1.4331554619902204e-05, "loss": 0.019, "step": 76520 }, { "grad_norm": 0.39870685338974, "learning_rate": 1.4319969240801484e-05, "loss": 0.0267, "step": 76530 }, { "grad_norm": 0.5100225210189819, "learning_rate": 1.430838776361526e-05, "loss": 0.0233, "step": 76540 }, { "grad_norm": 0.4710730016231537, "learning_rate": 1.4296810189610093e-05, "loss": 0.0305, "step": 76550 }, { "grad_norm": 0.33036619424819946, "learning_rate": 1.4285236520052087e-05, "loss": 0.0186, "step": 76560 }, { "grad_norm": 0.22362175583839417, "learning_rate": 1.4273666756206933e-05, "loss": 0.0289, "step": 76570 }, { "grad_norm": 0.2935297191143036, "learning_rate": 1.4262100899339848e-05, "loss": 0.0234, "step": 76580 }, { "grad_norm": 1.8205015659332275, "learning_rate": 1.4250538950715696e-05, "loss": 0.0294, "step": 76590 }, { "grad_norm": 0.31354641914367676, "learning_rate": 1.4238980911598842e-05, "loss": 0.0182, "step": 76600 }, { "grad_norm": 0.46174874901771545, "learning_rate": 1.4227426783253267e-05, "loss": 0.0225, "step": 76610 }, { "grad_norm": 0.7071247100830078, "learning_rate": 1.4215876566942521e-05, "loss": 0.0269, "step": 76620 }, { "grad_norm": 0.4610264003276825, "learning_rate": 1.4204330263929721e-05, "loss": 0.0205, "step": 76630 }, { "grad_norm": 0.32037079334259033, "learning_rate": 1.4192787875477537e-05, "loss": 0.0221, "step": 76640 }, { "grad_norm": 0.27172863483428955, "learning_rate": 1.4181249402848246e-05, "loss": 0.0216, "step": 76650 }, { "grad_norm": 0.26652440428733826, "learning_rate": 1.4169714847303656e-05, "loss": 0.0232, "step": 76660 }, { "grad_norm": 0.2291182428598404, "learning_rate": 1.4158184210105197e-05, "loss": 0.0318, "step": 76670 }, { "grad_norm": 0.28519853949546814, "learning_rate": 1.4146657492513787e-05, "loss": 0.0194, "step": 76680 }, { "grad_norm": 0.18510085344314575, "learning_rate": 1.4135134695790047e-05, "loss": 0.022, "step": 76690 }, { "grad_norm": 0.4820023775100708, "learning_rate": 1.4123615821194037e-05, "loss": 0.0249, "step": 76700 }, { "grad_norm": 0.3713941276073456, "learning_rate": 1.4112100869985473e-05, "loss": 0.019, "step": 76710 }, { "grad_norm": 0.7141150832176208, "learning_rate": 1.4100589843423589e-05, "loss": 0.024, "step": 76720 }, { "grad_norm": 1.3710800409317017, "learning_rate": 1.408908274276723e-05, "loss": 0.0233, "step": 76730 }, { "grad_norm": 1.786567211151123, "learning_rate": 1.4077579569274786e-05, "loss": 0.0265, "step": 76740 }, { "grad_norm": 0.37512925267219543, "learning_rate": 1.4066080324204245e-05, "loss": 0.0285, "step": 76750 }, { "grad_norm": 0.2852247953414917, "learning_rate": 1.405458500881312e-05, "loss": 0.0301, "step": 76760 }, { "grad_norm": 0.20446300506591797, "learning_rate": 1.4043093624358538e-05, "loss": 0.0247, "step": 76770 }, { "grad_norm": 0.3351881206035614, "learning_rate": 1.4031606172097189e-05, "loss": 0.0363, "step": 76780 }, { "grad_norm": 0.3207348883152008, "learning_rate": 1.40201226532853e-05, "loss": 0.0194, "step": 76790 }, { "grad_norm": 0.21805258095264435, "learning_rate": 1.4008643069178701e-05, "loss": 0.0209, "step": 76800 }, { "grad_norm": 0.48101165890693665, "learning_rate": 1.399716742103278e-05, "loss": 0.019, "step": 76810 }, { "grad_norm": 0.3510461747646332, "learning_rate": 1.398569571010252e-05, "loss": 0.0202, "step": 76820 }, { "grad_norm": 0.28397759795188904, "learning_rate": 1.3974227937642414e-05, "loss": 0.0259, "step": 76830 }, { "grad_norm": 0.2625250220298767, "learning_rate": 1.3962764104906595e-05, "loss": 0.0174, "step": 76840 }, { "grad_norm": 0.32761260867118835, "learning_rate": 1.3951304213148696e-05, "loss": 0.02, "step": 76850 }, { "grad_norm": 0.242067351937294, "learning_rate": 1.3939848263621979e-05, "loss": 0.0199, "step": 76860 }, { "grad_norm": 0.30190032720565796, "learning_rate": 1.3928396257579219e-05, "loss": 0.0229, "step": 76870 }, { "grad_norm": 0.3045836389064789, "learning_rate": 1.3916948196272833e-05, "loss": 0.0322, "step": 76880 }, { "grad_norm": 0.30384212732315063, "learning_rate": 1.3905504080954734e-05, "loss": 0.0301, "step": 76890 }, { "grad_norm": 0.806149959564209, "learning_rate": 1.3894063912876454e-05, "loss": 0.0209, "step": 76900 }, { "grad_norm": 0.22531934082508087, "learning_rate": 1.3882627693289046e-05, "loss": 0.0207, "step": 76910 }, { "grad_norm": 0.3416008949279785, "learning_rate": 1.3871195423443183e-05, "loss": 0.0287, "step": 76920 }, { "grad_norm": 0.2539273798465729, "learning_rate": 1.3859767104589044e-05, "loss": 0.0183, "step": 76930 }, { "grad_norm": 0.20058827102184296, "learning_rate": 1.3848342737976467e-05, "loss": 0.0182, "step": 76940 }, { "grad_norm": 0.5048577785491943, "learning_rate": 1.3836922324854757e-05, "loss": 0.0225, "step": 76950 }, { "grad_norm": 0.41502925753593445, "learning_rate": 1.382550586647287e-05, "loss": 0.0253, "step": 76960 }, { "grad_norm": 0.2884781062602997, "learning_rate": 1.3814093364079257e-05, "loss": 0.0203, "step": 76970 }, { "grad_norm": 0.411609947681427, "learning_rate": 1.3802684818921991e-05, "loss": 0.0279, "step": 76980 }, { "grad_norm": 0.33862584829330444, "learning_rate": 1.3791280232248687e-05, "loss": 0.0266, "step": 76990 }, { "grad_norm": 0.35239288210868835, "learning_rate": 1.3779879605306555e-05, "loss": 0.028, "step": 77000 }, { "grad_norm": 0.3224122226238251, "learning_rate": 1.376848293934232e-05, "loss": 0.0293, "step": 77010 }, { "grad_norm": 0.3845612704753876, "learning_rate": 1.3757090235602316e-05, "loss": 0.0282, "step": 77020 }, { "grad_norm": 0.23397095501422882, "learning_rate": 1.3745701495332447e-05, "loss": 0.0292, "step": 77030 }, { "grad_norm": 1.3824338912963867, "learning_rate": 1.3734316719778145e-05, "loss": 0.0224, "step": 77040 }, { "grad_norm": 0.34283575415611267, "learning_rate": 1.3722935910184454e-05, "loss": 0.0227, "step": 77050 }, { "grad_norm": 0.48351308703422546, "learning_rate": 1.3711559067795921e-05, "loss": 0.0242, "step": 77060 }, { "grad_norm": 0.2889125347137451, "learning_rate": 1.3700186193856756e-05, "loss": 0.0257, "step": 77070 }, { "grad_norm": 0.5038675665855408, "learning_rate": 1.368881728961064e-05, "loss": 0.017, "step": 77080 }, { "grad_norm": 0.19352594017982483, "learning_rate": 1.3677452356300885e-05, "loss": 0.0233, "step": 77090 }, { "grad_norm": 0.5629395246505737, "learning_rate": 1.3666091395170321e-05, "loss": 0.0335, "step": 77100 }, { "grad_norm": 0.32822057604789734, "learning_rate": 1.3654734407461384e-05, "loss": 0.0281, "step": 77110 }, { "grad_norm": 0.37120211124420166, "learning_rate": 1.3643381394416022e-05, "loss": 0.0235, "step": 77120 }, { "grad_norm": 0.407754510641098, "learning_rate": 1.3632032357275838e-05, "loss": 0.0272, "step": 77130 }, { "grad_norm": 0.4524548053741455, "learning_rate": 1.3620687297281898e-05, "loss": 0.0213, "step": 77140 }, { "grad_norm": 0.23628777265548706, "learning_rate": 1.3609346215674917e-05, "loss": 0.0233, "step": 77150 }, { "grad_norm": 0.4521312117576599, "learning_rate": 1.3598009113695109e-05, "loss": 0.0275, "step": 77160 }, { "grad_norm": 0.2998572885990143, "learning_rate": 1.358667599258231e-05, "loss": 0.0245, "step": 77170 }, { "grad_norm": 0.20885923504829407, "learning_rate": 1.357534685357586e-05, "loss": 0.0183, "step": 77180 }, { "grad_norm": 0.28121861815452576, "learning_rate": 1.3564021697914715e-05, "loss": 0.0199, "step": 77190 }, { "grad_norm": 0.22971919178962708, "learning_rate": 1.3552700526837381e-05, "loss": 0.0169, "step": 77200 }, { "grad_norm": 0.1848447024822235, "learning_rate": 1.3541383341581925e-05, "loss": 0.0217, "step": 77210 }, { "grad_norm": 0.4913995862007141, "learning_rate": 1.3530070143385965e-05, "loss": 0.0217, "step": 77220 }, { "grad_norm": 0.9234644174575806, "learning_rate": 1.3518760933486702e-05, "loss": 0.0218, "step": 77230 }, { "grad_norm": 0.2899872958660126, "learning_rate": 1.3507455713120908e-05, "loss": 0.0225, "step": 77240 }, { "grad_norm": 0.3212431073188782, "learning_rate": 1.3496154483524876e-05, "loss": 0.0221, "step": 77250 }, { "grad_norm": 0.2793463170528412, "learning_rate": 1.3484857245934506e-05, "loss": 0.0199, "step": 77260 }, { "grad_norm": 0.4124245047569275, "learning_rate": 1.3473564001585248e-05, "loss": 0.0205, "step": 77270 }, { "grad_norm": 0.4093720316886902, "learning_rate": 1.3462274751712129e-05, "loss": 0.0238, "step": 77280 }, { "grad_norm": 0.2740460932254791, "learning_rate": 1.3450989497549698e-05, "loss": 0.0224, "step": 77290 }, { "grad_norm": 0.44655120372772217, "learning_rate": 1.3439708240332115e-05, "loss": 0.0238, "step": 77300 }, { "grad_norm": 0.3762459456920624, "learning_rate": 1.3428430981293043e-05, "loss": 0.0201, "step": 77310 }, { "grad_norm": 0.32606473565101624, "learning_rate": 1.3417157721665808e-05, "loss": 0.0271, "step": 77320 }, { "grad_norm": 0.45456963777542114, "learning_rate": 1.3405888462683186e-05, "loss": 0.0181, "step": 77330 }, { "grad_norm": 0.3053956925868988, "learning_rate": 1.3394623205577595e-05, "loss": 0.0226, "step": 77340 }, { "grad_norm": 0.5868293046951294, "learning_rate": 1.3383361951580963e-05, "loss": 0.0261, "step": 77350 }, { "grad_norm": 0.3287004828453064, "learning_rate": 1.337210470192483e-05, "loss": 0.0208, "step": 77360 }, { "grad_norm": 0.7841672897338867, "learning_rate": 1.3360851457840245e-05, "loss": 0.0245, "step": 77370 }, { "grad_norm": 0.3309065103530884, "learning_rate": 1.3349602220557856e-05, "loss": 0.0226, "step": 77380 }, { "grad_norm": 0.43291181325912476, "learning_rate": 1.3338356991307865e-05, "loss": 0.0186, "step": 77390 }, { "grad_norm": 0.6663122177124023, "learning_rate": 1.3327115771320048e-05, "loss": 0.0301, "step": 77400 }, { "grad_norm": 0.23293758928775787, "learning_rate": 1.3315878561823697e-05, "loss": 0.0243, "step": 77410 }, { "grad_norm": 1.3031562566757202, "learning_rate": 1.330464536404773e-05, "loss": 0.0202, "step": 77420 }, { "grad_norm": 0.3970976769924164, "learning_rate": 1.3293416179220563e-05, "loss": 0.0236, "step": 77430 }, { "grad_norm": 0.3523072600364685, "learning_rate": 1.328219100857021e-05, "loss": 0.0252, "step": 77440 }, { "grad_norm": 0.4639642536640167, "learning_rate": 1.327096985332425e-05, "loss": 0.0281, "step": 77450 }, { "grad_norm": 0.9655867218971252, "learning_rate": 1.3259752714709816e-05, "loss": 0.0227, "step": 77460 }, { "grad_norm": 0.36288201808929443, "learning_rate": 1.3248539593953569e-05, "loss": 0.0227, "step": 77470 }, { "grad_norm": 0.2506873905658722, "learning_rate": 1.323733049228178e-05, "loss": 0.0246, "step": 77480 }, { "grad_norm": 0.3386775255203247, "learning_rate": 1.3226125410920276e-05, "loss": 0.0223, "step": 77490 }, { "grad_norm": 0.5124915838241577, "learning_rate": 1.3214924351094388e-05, "loss": 0.0188, "step": 77500 }, { "grad_norm": 0.29342496395111084, "learning_rate": 1.3203727314029069e-05, "loss": 0.021, "step": 77510 }, { "grad_norm": 0.6316903233528137, "learning_rate": 1.3192534300948811e-05, "loss": 0.0154, "step": 77520 }, { "grad_norm": 0.24636976420879364, "learning_rate": 1.3181345313077675e-05, "loss": 0.0198, "step": 77530 }, { "grad_norm": 0.23983773589134216, "learning_rate": 1.3170160351639243e-05, "loss": 0.0302, "step": 77540 }, { "grad_norm": 0.2713735103607178, "learning_rate": 1.3158979417856721e-05, "loss": 0.0325, "step": 77550 }, { "grad_norm": 0.9761229157447815, "learning_rate": 1.31478025129528e-05, "loss": 0.036, "step": 77560 }, { "grad_norm": 0.5749509930610657, "learning_rate": 1.3136629638149794e-05, "loss": 0.0293, "step": 77570 }, { "grad_norm": 0.2619943916797638, "learning_rate": 1.3125460794669547e-05, "loss": 0.0291, "step": 77580 }, { "grad_norm": 0.5603393912315369, "learning_rate": 1.3114295983733483e-05, "loss": 0.0261, "step": 77590 }, { "grad_norm": 0.3660691976547241, "learning_rate": 1.3103135206562534e-05, "loss": 0.0191, "step": 77600 }, { "grad_norm": 0.2846817076206207, "learning_rate": 1.3091978464377264e-05, "loss": 0.0275, "step": 77610 }, { "grad_norm": 0.3759514391422272, "learning_rate": 1.3080825758397724e-05, "loss": 0.0269, "step": 77620 }, { "grad_norm": 0.4575652778148651, "learning_rate": 1.3069677089843568e-05, "loss": 0.025, "step": 77630 }, { "grad_norm": 0.385153204202652, "learning_rate": 1.305853245993401e-05, "loss": 0.0343, "step": 77640 }, { "grad_norm": 0.29543232917785645, "learning_rate": 1.3047391869887809e-05, "loss": 0.0186, "step": 77650 }, { "grad_norm": 0.381303071975708, "learning_rate": 1.3036255320923263e-05, "loss": 0.0186, "step": 77660 }, { "grad_norm": 1.2670873403549194, "learning_rate": 1.3025122814258273e-05, "loss": 0.0215, "step": 77670 }, { "grad_norm": 0.44792526960372925, "learning_rate": 1.3013994351110249e-05, "loss": 0.0256, "step": 77680 }, { "grad_norm": 0.1792747676372528, "learning_rate": 1.3002869932696194e-05, "loss": 0.0182, "step": 77690 }, { "grad_norm": 0.27623671293258667, "learning_rate": 1.2991749560232664e-05, "loss": 0.0231, "step": 77700 }, { "grad_norm": 0.25962767004966736, "learning_rate": 1.2980633234935768e-05, "loss": 0.0286, "step": 77710 }, { "grad_norm": 0.35811078548431396, "learning_rate": 1.2969520958021153e-05, "loss": 0.0266, "step": 77720 }, { "grad_norm": 0.31588828563690186, "learning_rate": 1.295841273070405e-05, "loss": 0.0257, "step": 77730 }, { "grad_norm": 0.3791327476501465, "learning_rate": 1.2947308554199255e-05, "loss": 0.0189, "step": 77740 }, { "grad_norm": 1.0655407905578613, "learning_rate": 1.2936208429721075e-05, "loss": 0.0445, "step": 77750 }, { "grad_norm": 0.36214199662208557, "learning_rate": 1.2925112358483416e-05, "loss": 0.0263, "step": 77760 }, { "grad_norm": 0.20745712518692017, "learning_rate": 1.2914020341699728e-05, "loss": 0.0344, "step": 77770 }, { "grad_norm": 0.5956981778144836, "learning_rate": 1.2902932380583032e-05, "loss": 0.0246, "step": 77780 }, { "grad_norm": 0.4017282724380493, "learning_rate": 1.2891848476345863e-05, "loss": 0.0273, "step": 77790 }, { "grad_norm": 0.2616802155971527, "learning_rate": 1.2880768630200368e-05, "loss": 0.0232, "step": 77800 }, { "grad_norm": 0.3330584764480591, "learning_rate": 1.286969284335819e-05, "loss": 0.0252, "step": 77810 }, { "grad_norm": 0.3763558566570282, "learning_rate": 1.2858621117030578e-05, "loss": 0.0291, "step": 77820 }, { "grad_norm": 0.22759997844696045, "learning_rate": 1.2847553452428324e-05, "loss": 0.0215, "step": 77830 }, { "grad_norm": 0.32388097047805786, "learning_rate": 1.283648985076178e-05, "loss": 0.0214, "step": 77840 }, { "grad_norm": 0.3986338973045349, "learning_rate": 1.2825430313240817e-05, "loss": 0.0272, "step": 77850 }, { "grad_norm": 0.4607401192188263, "learning_rate": 1.281437484107491e-05, "loss": 0.0267, "step": 77860 }, { "grad_norm": 0.21091756224632263, "learning_rate": 1.2803323435473058e-05, "loss": 0.0245, "step": 77870 }, { "grad_norm": 0.2720035910606384, "learning_rate": 1.2792276097643841e-05, "loss": 0.0206, "step": 77880 }, { "grad_norm": 0.2788738012313843, "learning_rate": 1.2781232828795342e-05, "loss": 0.0211, "step": 77890 }, { "grad_norm": 0.30030524730682373, "learning_rate": 1.2770193630135292e-05, "loss": 0.0286, "step": 77900 }, { "grad_norm": 0.3019265830516815, "learning_rate": 1.2759158502870877e-05, "loss": 0.034, "step": 77910 }, { "grad_norm": 0.255338191986084, "learning_rate": 1.2748127448208913e-05, "loss": 0.0217, "step": 77920 }, { "grad_norm": 0.5604537725448608, "learning_rate": 1.2737100467355706e-05, "loss": 0.032, "step": 77930 }, { "grad_norm": 0.4193832278251648, "learning_rate": 1.2726077561517169e-05, "loss": 0.038, "step": 77940 }, { "grad_norm": 0.28932902216911316, "learning_rate": 1.2715058731898755e-05, "loss": 0.027, "step": 77950 }, { "grad_norm": 0.4224596917629242, "learning_rate": 1.2704043979705471e-05, "loss": 0.021, "step": 77960 }, { "grad_norm": 0.4564356505870819, "learning_rate": 1.269303330614185e-05, "loss": 0.0265, "step": 77970 }, { "grad_norm": 0.3069542348384857, "learning_rate": 1.2682026712412015e-05, "loss": 0.0292, "step": 77980 }, { "grad_norm": 0.5440033078193665, "learning_rate": 1.2671024199719644e-05, "loss": 0.0204, "step": 77990 }, { "grad_norm": 0.4474259614944458, "learning_rate": 1.2660025769267936e-05, "loss": 0.0267, "step": 78000 }, { "grad_norm": 0.33372920751571655, "learning_rate": 1.2649031422259667e-05, "loss": 0.0281, "step": 78010 }, { "grad_norm": 0.37974050641059875, "learning_rate": 1.2638041159897163e-05, "loss": 0.0226, "step": 78020 }, { "grad_norm": 0.5161775946617126, "learning_rate": 1.262705498338232e-05, "loss": 0.0246, "step": 78030 }, { "grad_norm": 0.7922442555427551, "learning_rate": 1.2616072893916542e-05, "loss": 0.0358, "step": 78040 }, { "grad_norm": 0.5073697566986084, "learning_rate": 1.2605094892700836e-05, "loss": 0.0292, "step": 78050 }, { "grad_norm": 0.2136813998222351, "learning_rate": 1.259412098093572e-05, "loss": 0.0195, "step": 78060 }, { "grad_norm": 0.3874508738517761, "learning_rate": 1.258315115982131e-05, "loss": 0.0249, "step": 78070 }, { "grad_norm": 0.3729530870914459, "learning_rate": 1.2572185430557203e-05, "loss": 0.0293, "step": 78080 }, { "grad_norm": 0.19484907388687134, "learning_rate": 1.2561223794342659e-05, "loss": 0.0233, "step": 78090 }, { "grad_norm": 0.4168320596218109, "learning_rate": 1.255026625237637e-05, "loss": 0.0262, "step": 78100 }, { "grad_norm": 0.28977951407432556, "learning_rate": 1.2539312805856678e-05, "loss": 0.0271, "step": 78110 }, { "grad_norm": 0.5049331784248352, "learning_rate": 1.25283634559814e-05, "loss": 0.0206, "step": 78120 }, { "grad_norm": 0.2595694363117218, "learning_rate": 1.2517418203947972e-05, "loss": 0.0193, "step": 78130 }, { "grad_norm": 0.3642145097255707, "learning_rate": 1.2506477050953308e-05, "loss": 0.0227, "step": 78140 }, { "grad_norm": 0.37532877922058105, "learning_rate": 1.2495539998193972e-05, "loss": 0.0229, "step": 78150 }, { "grad_norm": 0.6377831697463989, "learning_rate": 1.2484607046865975e-05, "loss": 0.0226, "step": 78160 }, { "grad_norm": 0.39182037115097046, "learning_rate": 1.2473678198164968e-05, "loss": 0.0275, "step": 78170 }, { "grad_norm": 0.4732513725757599, "learning_rate": 1.2462753453286075e-05, "loss": 0.033, "step": 78180 }, { "grad_norm": 0.390810489654541, "learning_rate": 1.2451832813424031e-05, "loss": 0.0456, "step": 78190 }, { "grad_norm": 0.29461848735809326, "learning_rate": 1.2440916279773095e-05, "loss": 0.0258, "step": 78200 }, { "grad_norm": 0.22497878968715668, "learning_rate": 1.2430003853527101e-05, "loss": 0.0197, "step": 78210 }, { "grad_norm": 0.5339065194129944, "learning_rate": 1.2419095535879383e-05, "loss": 0.0313, "step": 78220 }, { "grad_norm": 0.4055200517177582, "learning_rate": 1.2408191328022878e-05, "loss": 0.0175, "step": 78230 }, { "grad_norm": 0.46792006492614746, "learning_rate": 1.2397291231150066e-05, "loss": 0.028, "step": 78240 }, { "grad_norm": 1.871353268623352, "learning_rate": 1.2386395246452937e-05, "loss": 0.0256, "step": 78250 }, { "grad_norm": 0.32461974024772644, "learning_rate": 1.2375503375123083e-05, "loss": 0.0228, "step": 78260 }, { "grad_norm": 0.25056302547454834, "learning_rate": 1.2364615618351583e-05, "loss": 0.0278, "step": 78270 }, { "grad_norm": 0.41544848680496216, "learning_rate": 1.2353731977329169e-05, "loss": 0.0241, "step": 78280 }, { "grad_norm": 0.6670957803726196, "learning_rate": 1.2342852453246007e-05, "loss": 0.0242, "step": 78290 }, { "grad_norm": 0.48881375789642334, "learning_rate": 1.2331977047291898e-05, "loss": 0.0304, "step": 78300 }, { "grad_norm": 0.5001994967460632, "learning_rate": 1.2321105760656137e-05, "loss": 0.0201, "step": 78310 }, { "grad_norm": 0.4404761791229248, "learning_rate": 1.2310238594527613e-05, "loss": 0.0231, "step": 78320 }, { "grad_norm": 0.5668857097625732, "learning_rate": 1.2299375550094693e-05, "loss": 0.0219, "step": 78330 }, { "grad_norm": 0.6876158118247986, "learning_rate": 1.2288516628545421e-05, "loss": 0.0236, "step": 78340 }, { "grad_norm": 0.24201934039592743, "learning_rate": 1.2277661831067255e-05, "loss": 0.0288, "step": 78350 }, { "grad_norm": 0.8192808628082275, "learning_rate": 1.2266811158847286e-05, "loss": 0.0232, "step": 78360 }, { "grad_norm": 0.4443497061729431, "learning_rate": 1.2255964613072107e-05, "loss": 0.0173, "step": 78370 }, { "grad_norm": 0.2600700557231903, "learning_rate": 1.2245122194927905e-05, "loss": 0.0313, "step": 78380 }, { "grad_norm": 0.32015860080718994, "learning_rate": 1.2234283905600364e-05, "loss": 0.0242, "step": 78390 }, { "grad_norm": 1.988237977027893, "learning_rate": 1.222344974627475e-05, "loss": 0.0276, "step": 78400 }, { "grad_norm": 0.3754602074623108, "learning_rate": 1.2212619718135875e-05, "loss": 0.0267, "step": 78410 }, { "grad_norm": 0.3489510118961334, "learning_rate": 1.2201793822368118e-05, "loss": 0.0193, "step": 78420 }, { "grad_norm": 0.45116621255874634, "learning_rate": 1.2190972060155337e-05, "loss": 0.0227, "step": 78430 }, { "grad_norm": 0.40515148639678955, "learning_rate": 1.2180154432681013e-05, "loss": 0.02, "step": 78440 }, { "grad_norm": 0.4923275411128998, "learning_rate": 1.2169340941128148e-05, "loss": 0.0268, "step": 78450 }, { "grad_norm": 0.20118410885334015, "learning_rate": 1.2158531586679267e-05, "loss": 0.0229, "step": 78460 }, { "grad_norm": 0.31794965267181396, "learning_rate": 1.2147726370516477e-05, "loss": 0.0271, "step": 78470 }, { "grad_norm": 0.3599902093410492, "learning_rate": 1.2136925293821422e-05, "loss": 0.0232, "step": 78480 }, { "grad_norm": 0.5302683115005493, "learning_rate": 1.2126128357775306e-05, "loss": 0.0235, "step": 78490 }, { "grad_norm": 0.30995801091194153, "learning_rate": 1.2115335563558838e-05, "loss": 0.013, "step": 78500 }, { "grad_norm": 0.2741081118583679, "learning_rate": 1.2104546912352327e-05, "loss": 0.0224, "step": 78510 }, { "grad_norm": 0.3483131527900696, "learning_rate": 1.2093762405335557e-05, "loss": 0.0238, "step": 78520 }, { "grad_norm": 0.43491700291633606, "learning_rate": 1.2082982043687974e-05, "loss": 0.0187, "step": 78530 }, { "grad_norm": 0.3567349910736084, "learning_rate": 1.207220582858845e-05, "loss": 0.0155, "step": 78540 }, { "grad_norm": 0.21601763367652893, "learning_rate": 1.206143376121549e-05, "loss": 0.0194, "step": 78550 }, { "grad_norm": 0.38293471932411194, "learning_rate": 1.2050665842747078e-05, "loss": 0.0202, "step": 78560 }, { "grad_norm": 0.29158371686935425, "learning_rate": 1.2039902074360804e-05, "loss": 0.0194, "step": 78570 }, { "grad_norm": 1.1049432754516602, "learning_rate": 1.2029142457233755e-05, "loss": 0.0314, "step": 78580 }, { "grad_norm": 0.29980722069740295, "learning_rate": 1.2018386992542601e-05, "loss": 0.0178, "step": 78590 }, { "grad_norm": 0.3341280519962311, "learning_rate": 1.2007635681463542e-05, "loss": 0.0248, "step": 78600 }, { "grad_norm": 0.4618494510650635, "learning_rate": 1.1996888525172334e-05, "loss": 0.0268, "step": 78610 }, { "grad_norm": 0.36686697602272034, "learning_rate": 1.198614552484425e-05, "loss": 0.0306, "step": 78620 }, { "grad_norm": 0.37212514877319336, "learning_rate": 1.197540668165415e-05, "loss": 0.0211, "step": 78630 }, { "grad_norm": 0.3325721323490143, "learning_rate": 1.1964671996776395e-05, "loss": 0.0286, "step": 78640 }, { "grad_norm": 0.32178565859794617, "learning_rate": 1.1953941471384922e-05, "loss": 0.0319, "step": 78650 }, { "grad_norm": 0.3192351758480072, "learning_rate": 1.1943215106653205e-05, "loss": 0.0248, "step": 78660 }, { "grad_norm": 0.291914701461792, "learning_rate": 1.193249290375429e-05, "loss": 0.0207, "step": 78670 }, { "grad_norm": 0.34570959210395813, "learning_rate": 1.1921774863860696e-05, "loss": 0.025, "step": 78680 }, { "grad_norm": 0.4631493389606476, "learning_rate": 1.1911060988144556e-05, "loss": 0.0214, "step": 78690 }, { "grad_norm": 0.370242178440094, "learning_rate": 1.1900351277777533e-05, "loss": 0.0256, "step": 78700 }, { "grad_norm": 0.48939409852027893, "learning_rate": 1.1889645733930798e-05, "loss": 0.0224, "step": 78710 }, { "grad_norm": 0.2520069181919098, "learning_rate": 1.1878944357775107e-05, "loss": 0.0195, "step": 78720 }, { "grad_norm": 0.5354989767074585, "learning_rate": 1.1868247150480749e-05, "loss": 0.0214, "step": 78730 }, { "grad_norm": 0.24940747022628784, "learning_rate": 1.1857554113217567e-05, "loss": 0.0251, "step": 78740 }, { "grad_norm": 0.40830209851264954, "learning_rate": 1.1846865247154903e-05, "loss": 0.0228, "step": 78750 }, { "grad_norm": 0.20928987860679626, "learning_rate": 1.1836180553461711e-05, "loss": 0.024, "step": 78760 }, { "grad_norm": 0.41362619400024414, "learning_rate": 1.1825500033306419e-05, "loss": 0.0411, "step": 78770 }, { "grad_norm": 0.3440850079059601, "learning_rate": 1.1814823687857052e-05, "loss": 0.0217, "step": 78780 }, { "grad_norm": 0.45470595359802246, "learning_rate": 1.1804151518281153e-05, "loss": 0.0244, "step": 78790 }, { "grad_norm": 0.9058800339698792, "learning_rate": 1.179348352574583e-05, "loss": 0.0368, "step": 78800 }, { "grad_norm": 0.40688014030456543, "learning_rate": 1.1782819711417697e-05, "loss": 0.0246, "step": 78810 }, { "grad_norm": 0.272450715303421, "learning_rate": 1.1772160076462951e-05, "loss": 0.0178, "step": 78820 }, { "grad_norm": 0.4222138822078705, "learning_rate": 1.1761504622047293e-05, "loss": 0.024, "step": 78830 }, { "grad_norm": 0.4944295883178711, "learning_rate": 1.1750853349335999e-05, "loss": 0.0188, "step": 78840 }, { "grad_norm": 0.23528656363487244, "learning_rate": 1.1740206259493874e-05, "loss": 0.0245, "step": 78850 }, { "grad_norm": 0.4467511475086212, "learning_rate": 1.1729563353685286e-05, "loss": 0.0247, "step": 78860 }, { "grad_norm": 0.2519535422325134, "learning_rate": 1.1718924633074091e-05, "loss": 0.0209, "step": 78870 }, { "grad_norm": 0.17449966073036194, "learning_rate": 1.1708290098823765e-05, "loss": 0.0234, "step": 78880 }, { "grad_norm": 0.3217701315879822, "learning_rate": 1.1697659752097245e-05, "loss": 0.0182, "step": 78890 }, { "grad_norm": 0.3240037262439728, "learning_rate": 1.1687033594057068e-05, "loss": 0.046, "step": 78900 }, { "grad_norm": 0.30418917536735535, "learning_rate": 1.1676411625865297e-05, "loss": 0.0187, "step": 78910 }, { "grad_norm": 0.44882261753082275, "learning_rate": 1.1665793848683542e-05, "loss": 0.022, "step": 78920 }, { "grad_norm": 0.7963400483131409, "learning_rate": 1.1655180263672927e-05, "loss": 0.0334, "step": 78930 }, { "grad_norm": 0.3249349594116211, "learning_rate": 1.1644570871994142e-05, "loss": 0.0208, "step": 78940 }, { "grad_norm": 0.2824959456920624, "learning_rate": 1.163396567480744e-05, "loss": 0.0188, "step": 78950 }, { "grad_norm": 0.5550521016120911, "learning_rate": 1.1623364673272552e-05, "loss": 0.0224, "step": 78960 }, { "grad_norm": 0.2127155363559723, "learning_rate": 1.1612767868548807e-05, "loss": 0.0265, "step": 78970 }, { "grad_norm": 0.3392508327960968, "learning_rate": 1.1602175261795057e-05, "loss": 0.0248, "step": 78980 }, { "grad_norm": 0.31617817282676697, "learning_rate": 1.1591586854169696e-05, "loss": 0.0248, "step": 78990 }, { "grad_norm": 0.19087029993534088, "learning_rate": 1.158100264683064e-05, "loss": 0.0227, "step": 79000 }, { "grad_norm": 0.43171703815460205, "learning_rate": 1.1570422640935386e-05, "loss": 0.0276, "step": 79010 }, { "grad_norm": 0.2169903665781021, "learning_rate": 1.155984683764092e-05, "loss": 0.0295, "step": 79020 }, { "grad_norm": 0.3945656418800354, "learning_rate": 1.1549275238103812e-05, "loss": 0.0257, "step": 79030 }, { "grad_norm": 0.3452806770801544, "learning_rate": 1.1538707843480152e-05, "loss": 0.0255, "step": 79040 }, { "grad_norm": 0.38417157530784607, "learning_rate": 1.1528144654925593e-05, "loss": 0.0243, "step": 79050 }, { "grad_norm": 0.4556807577610016, "learning_rate": 1.1517585673595277e-05, "loss": 0.0292, "step": 79060 }, { "grad_norm": 0.3767537474632263, "learning_rate": 1.150703090064395e-05, "loss": 0.0258, "step": 79070 }, { "grad_norm": 0.3559460937976837, "learning_rate": 1.1496480337225835e-05, "loss": 0.0233, "step": 79080 }, { "grad_norm": 0.4683338701725006, "learning_rate": 1.1485933984494751e-05, "loss": 0.0199, "step": 79090 }, { "grad_norm": 0.3449169397354126, "learning_rate": 1.1475391843604e-05, "loss": 0.0224, "step": 79100 }, { "grad_norm": 0.39694252610206604, "learning_rate": 1.1464853915706503e-05, "loss": 0.0285, "step": 79110 }, { "grad_norm": 0.3523907959461212, "learning_rate": 1.1454320201954626e-05, "loss": 0.0219, "step": 79120 }, { "grad_norm": 0.19128096103668213, "learning_rate": 1.1443790703500356e-05, "loss": 0.0204, "step": 79130 }, { "grad_norm": 0.25838878750801086, "learning_rate": 1.1433265421495154e-05, "loss": 0.0251, "step": 79140 }, { "grad_norm": 0.38492968678474426, "learning_rate": 1.1422744357090065e-05, "loss": 0.0329, "step": 79150 }, { "grad_norm": 0.4103923439979553, "learning_rate": 1.1412227511435647e-05, "loss": 0.0274, "step": 79160 }, { "grad_norm": 0.3096124529838562, "learning_rate": 1.1401714885682025e-05, "loss": 0.022, "step": 79170 }, { "grad_norm": 0.3112947344779968, "learning_rate": 1.1391206480978823e-05, "loss": 0.0201, "step": 79180 }, { "grad_norm": 0.36422228813171387, "learning_rate": 1.1380702298475227e-05, "loss": 0.0284, "step": 79190 }, { "grad_norm": 0.3510231673717499, "learning_rate": 1.1370202339319985e-05, "loss": 0.0221, "step": 79200 }, { "grad_norm": 0.279012531042099, "learning_rate": 1.1359706604661319e-05, "loss": 0.0226, "step": 79210 }, { "grad_norm": 0.333079069852829, "learning_rate": 1.1349215095647042e-05, "loss": 0.0226, "step": 79220 }, { "grad_norm": 0.8566305637359619, "learning_rate": 1.1338727813424493e-05, "loss": 0.0199, "step": 79230 }, { "grad_norm": 0.21110780537128448, "learning_rate": 1.132824475914056e-05, "loss": 0.0211, "step": 79240 }, { "grad_norm": 0.7438730001449585, "learning_rate": 1.131776593394162e-05, "loss": 0.0282, "step": 79250 }, { "grad_norm": 0.387946754693985, "learning_rate": 1.1307291338973657e-05, "loss": 0.0306, "step": 79260 }, { "grad_norm": 0.5050097703933716, "learning_rate": 1.1296820975382121e-05, "loss": 0.0192, "step": 79270 }, { "grad_norm": 0.43885746598243713, "learning_rate": 1.1286354844312074e-05, "loss": 0.0263, "step": 79280 }, { "grad_norm": 0.41736799478530884, "learning_rate": 1.1275892946908023e-05, "loss": 0.0277, "step": 79290 }, { "grad_norm": 0.35387906432151794, "learning_rate": 1.126543528431413e-05, "loss": 0.0272, "step": 79300 }, { "grad_norm": 0.2595386803150177, "learning_rate": 1.125498185767398e-05, "loss": 0.0206, "step": 79310 }, { "grad_norm": 0.29451170563697815, "learning_rate": 1.1244532668130781e-05, "loss": 0.0207, "step": 79320 }, { "grad_norm": 0.353374183177948, "learning_rate": 1.1234087716827203e-05, "loss": 0.0311, "step": 79330 }, { "grad_norm": 0.6837678551673889, "learning_rate": 1.1223647004905524e-05, "loss": 0.028, "step": 79340 }, { "grad_norm": 0.7086607813835144, "learning_rate": 1.1213210533507485e-05, "loss": 0.023, "step": 79350 }, { "grad_norm": 0.26314908266067505, "learning_rate": 1.1202778303774447e-05, "loss": 0.0224, "step": 79360 }, { "grad_norm": 0.23759055137634277, "learning_rate": 1.1192350316847234e-05, "loss": 0.019, "step": 79370 }, { "grad_norm": 0.25034299492836, "learning_rate": 1.1181926573866258e-05, "loss": 0.0192, "step": 79380 }, { "grad_norm": 0.22272464632987976, "learning_rate": 1.1171507075971416e-05, "loss": 0.0315, "step": 79390 }, { "grad_norm": 0.31091606616973877, "learning_rate": 1.1161091824302177e-05, "loss": 0.0314, "step": 79400 }, { "grad_norm": 0.31601253151893616, "learning_rate": 1.115068081999755e-05, "loss": 0.0293, "step": 79410 }, { "grad_norm": 0.38614997267723083, "learning_rate": 1.1140274064196071e-05, "loss": 0.0284, "step": 79420 }, { "grad_norm": 0.45569828152656555, "learning_rate": 1.112987155803578e-05, "loss": 0.025, "step": 79430 }, { "grad_norm": 0.44867825508117676, "learning_rate": 1.1119473302654298e-05, "loss": 0.0257, "step": 79440 }, { "grad_norm": 0.2879219651222229, "learning_rate": 1.1109079299188769e-05, "loss": 0.0283, "step": 79450 }, { "grad_norm": 0.2529003918170929, "learning_rate": 1.1098689548775847e-05, "loss": 0.0169, "step": 79460 }, { "grad_norm": 0.44534575939178467, "learning_rate": 1.1088304052551762e-05, "loss": 0.0299, "step": 79470 }, { "grad_norm": 0.3203613758087158, "learning_rate": 1.1077922811652215e-05, "loss": 0.0215, "step": 79480 }, { "grad_norm": 0.2713952362537384, "learning_rate": 1.1067545827212534e-05, "loss": 0.0291, "step": 79490 }, { "grad_norm": 0.3486751616001129, "learning_rate": 1.1057173100367496e-05, "loss": 0.0262, "step": 79500 }, { "grad_norm": 0.34942862391471863, "learning_rate": 1.1046804632251474e-05, "loss": 0.018, "step": 79510 }, { "grad_norm": 0.3444461226463318, "learning_rate": 1.1036440423998318e-05, "loss": 0.0243, "step": 79520 }, { "grad_norm": 0.633256733417511, "learning_rate": 1.1026080476741469e-05, "loss": 0.033, "step": 79530 }, { "grad_norm": 0.49649086594581604, "learning_rate": 1.1015724791613836e-05, "loss": 0.027, "step": 79540 }, { "grad_norm": 0.9785105586051941, "learning_rate": 1.1005373369747951e-05, "loss": 0.031, "step": 79550 }, { "grad_norm": 0.1739947646856308, "learning_rate": 1.09950262122758e-05, "loss": 0.0216, "step": 79560 }, { "grad_norm": 0.5993320941925049, "learning_rate": 1.0984683320328948e-05, "loss": 0.0285, "step": 79570 }, { "grad_norm": 0.3680379092693329, "learning_rate": 1.0974344695038458e-05, "loss": 0.0184, "step": 79580 }, { "grad_norm": 0.4194203019142151, "learning_rate": 1.0964010337534968e-05, "loss": 0.0208, "step": 79590 }, { "grad_norm": 0.3572624623775482, "learning_rate": 1.0953680248948611e-05, "loss": 0.0214, "step": 79600 }, { "grad_norm": 0.2590457797050476, "learning_rate": 1.0943354430409069e-05, "loss": 0.0225, "step": 79610 }, { "grad_norm": 0.4937073290348053, "learning_rate": 1.0933032883045574e-05, "loss": 0.0191, "step": 79620 }, { "grad_norm": 0.4251152276992798, "learning_rate": 1.0922715607986872e-05, "loss": 0.0385, "step": 79630 }, { "grad_norm": 0.42869481444358826, "learning_rate": 1.0912402606361234e-05, "loss": 0.0212, "step": 79640 }, { "grad_norm": 0.4183851480484009, "learning_rate": 1.090209387929647e-05, "loss": 0.0268, "step": 79650 }, { "grad_norm": 0.24983960390090942, "learning_rate": 1.0891789427919957e-05, "loss": 0.0237, "step": 79660 }, { "grad_norm": 0.3813338875770569, "learning_rate": 1.088148925335854e-05, "loss": 0.0205, "step": 79670 }, { "grad_norm": 0.2714218199253082, "learning_rate": 1.087119335673864e-05, "loss": 0.0139, "step": 79680 }, { "grad_norm": 0.35387930274009705, "learning_rate": 1.0860901739186208e-05, "loss": 0.0298, "step": 79690 }, { "grad_norm": 0.2862967848777771, "learning_rate": 1.085061440182673e-05, "loss": 0.0218, "step": 79700 }, { "grad_norm": 0.21001730859279633, "learning_rate": 1.0840331345785193e-05, "loss": 0.023, "step": 79710 }, { "grad_norm": 0.47323861718177795, "learning_rate": 1.083005257218615e-05, "loss": 0.0219, "step": 79720 }, { "grad_norm": 0.43311768770217896, "learning_rate": 1.0819778082153642e-05, "loss": 0.0184, "step": 79730 }, { "grad_norm": 0.37860122323036194, "learning_rate": 1.0809507876811325e-05, "loss": 0.0293, "step": 79740 }, { "grad_norm": 0.3290208578109741, "learning_rate": 1.0799241957282292e-05, "loss": 0.0201, "step": 79750 }, { "grad_norm": 0.34875473380088806, "learning_rate": 1.078898032468923e-05, "loss": 0.021, "step": 79760 }, { "grad_norm": 0.30284321308135986, "learning_rate": 1.0778722980154315e-05, "loss": 0.0218, "step": 79770 }, { "grad_norm": 0.2888219654560089, "learning_rate": 1.0768469924799302e-05, "loss": 0.019, "step": 79780 }, { "grad_norm": 0.315410315990448, "learning_rate": 1.0758221159745419e-05, "loss": 0.0206, "step": 79790 }, { "grad_norm": 1.2628216743469238, "learning_rate": 1.074797668611347e-05, "loss": 0.0288, "step": 79800 }, { "grad_norm": 0.24958136677742004, "learning_rate": 1.0737736505023777e-05, "loss": 0.0209, "step": 79810 }, { "grad_norm": 0.3493783175945282, "learning_rate": 1.0727500617596203e-05, "loss": 0.0207, "step": 79820 }, { "grad_norm": 0.2683069109916687, "learning_rate": 1.0717269024950104e-05, "loss": 0.0314, "step": 79830 }, { "grad_norm": 0.1890832930803299, "learning_rate": 1.070704172820442e-05, "loss": 0.0317, "step": 79840 }, { "grad_norm": 2.3418853282928467, "learning_rate": 1.0696818728477559e-05, "loss": 0.0286, "step": 79850 }, { "grad_norm": 0.3685571551322937, "learning_rate": 1.0686600026887511e-05, "loss": 0.0195, "step": 79860 }, { "grad_norm": 0.4467039704322815, "learning_rate": 1.0676385624551777e-05, "loss": 0.0291, "step": 79870 }, { "grad_norm": 0.3474195897579193, "learning_rate": 1.0666175522587402e-05, "loss": 0.0316, "step": 79880 }, { "grad_norm": 0.22860005497932434, "learning_rate": 1.0655969722110925e-05, "loss": 0.0213, "step": 79890 }, { "grad_norm": 0.3787168562412262, "learning_rate": 1.0645768224238439e-05, "loss": 0.025, "step": 79900 }, { "grad_norm": 0.4321984052658081, "learning_rate": 1.063557103008559e-05, "loss": 0.0259, "step": 79910 }, { "grad_norm": 0.2783287763595581, "learning_rate": 1.0625378140767494e-05, "loss": 0.0231, "step": 79920 }, { "grad_norm": 0.26922547817230225, "learning_rate": 1.0615189557398846e-05, "loss": 0.0206, "step": 79930 }, { "grad_norm": 0.4273214638233185, "learning_rate": 1.060500528109386e-05, "loss": 0.0231, "step": 79940 }, { "grad_norm": 0.3474545180797577, "learning_rate": 1.0594825312966272e-05, "loss": 0.022, "step": 79950 }, { "grad_norm": 0.4868428111076355, "learning_rate": 1.0584649654129336e-05, "loss": 0.0334, "step": 79960 }, { "grad_norm": 0.3708491623401642, "learning_rate": 1.0574478305695863e-05, "loss": 0.0287, "step": 79970 }, { "grad_norm": 0.42972132563591003, "learning_rate": 1.0564311268778154e-05, "loss": 0.021, "step": 79980 }, { "grad_norm": 0.2707556486129761, "learning_rate": 1.0554148544488073e-05, "loss": 0.0298, "step": 79990 }, { "grad_norm": 0.2429172694683075, "learning_rate": 1.0543990133937004e-05, "loss": 0.0191, "step": 80000 }, { "grad_norm": 0.3790987730026245, "learning_rate": 1.0533836038235868e-05, "loss": 0.0309, "step": 80010 }, { "grad_norm": 0.262341171503067, "learning_rate": 1.052368625849507e-05, "loss": 0.0217, "step": 80020 }, { "grad_norm": 0.3717206120491028, "learning_rate": 1.05135407958246e-05, "loss": 0.0205, "step": 80030 }, { "grad_norm": 0.2377728819847107, "learning_rate": 1.0503399651333934e-05, "loss": 0.0252, "step": 80040 }, { "grad_norm": 0.41850927472114563, "learning_rate": 1.04932628261321e-05, "loss": 0.0219, "step": 80050 }, { "grad_norm": 0.4055512845516205, "learning_rate": 1.0483130321327644e-05, "loss": 0.0208, "step": 80060 }, { "grad_norm": 0.4744139015674591, "learning_rate": 1.0473002138028654e-05, "loss": 0.0233, "step": 80070 }, { "grad_norm": 0.34951311349868774, "learning_rate": 1.0462878277342713e-05, "loss": 0.0186, "step": 80080 }, { "grad_norm": 0.22976841032505035, "learning_rate": 1.0452758740376972e-05, "loss": 0.0248, "step": 80090 }, { "grad_norm": 0.34022071957588196, "learning_rate": 1.0442643528238067e-05, "loss": 0.0226, "step": 80100 }, { "grad_norm": 0.37196990847587585, "learning_rate": 1.0432532642032188e-05, "loss": 0.0249, "step": 80110 }, { "grad_norm": 0.27589529752731323, "learning_rate": 1.042242608286505e-05, "loss": 0.0208, "step": 80120 }, { "grad_norm": 0.4593421220779419, "learning_rate": 1.041232385184191e-05, "loss": 0.0262, "step": 80130 }, { "grad_norm": 0.22952157258987427, "learning_rate": 1.0402225950067495e-05, "loss": 0.018, "step": 80140 }, { "grad_norm": 0.5165368914604187, "learning_rate": 1.0392132378646119e-05, "loss": 0.0288, "step": 80150 }, { "grad_norm": 0.2610943019390106, "learning_rate": 1.0382043138681607e-05, "loss": 0.0235, "step": 80160 }, { "grad_norm": 0.2829466164112091, "learning_rate": 1.0371958231277284e-05, "loss": 0.0173, "step": 80170 }, { "grad_norm": 0.2669786512851715, "learning_rate": 1.0361877657536024e-05, "loss": 0.0233, "step": 80180 }, { "grad_norm": 0.3702408969402313, "learning_rate": 1.035180141856023e-05, "loss": 0.0216, "step": 80190 }, { "grad_norm": 0.4907791316509247, "learning_rate": 1.0341729515451836e-05, "loss": 0.0254, "step": 80200 }, { "grad_norm": 0.32420167326927185, "learning_rate": 1.0331661949312265e-05, "loss": 0.0232, "step": 80210 }, { "grad_norm": 0.3601701855659485, "learning_rate": 1.0321598721242514e-05, "loss": 0.0321, "step": 80220 }, { "grad_norm": 0.24695630371570587, "learning_rate": 1.0311539832343064e-05, "loss": 0.0196, "step": 80230 }, { "grad_norm": 0.47352275252342224, "learning_rate": 1.0301485283713942e-05, "loss": 0.019, "step": 80240 }, { "grad_norm": 0.4436083137989044, "learning_rate": 1.0291435076454703e-05, "loss": 0.0307, "step": 80250 }, { "grad_norm": 0.465804785490036, "learning_rate": 1.028138921166444e-05, "loss": 0.0205, "step": 80260 }, { "grad_norm": 0.26705625653266907, "learning_rate": 1.0271347690441719e-05, "loss": 0.0211, "step": 80270 }, { "grad_norm": 0.4320022761821747, "learning_rate": 1.0261310513884698e-05, "loss": 0.0293, "step": 80280 }, { "grad_norm": 0.25235068798065186, "learning_rate": 1.0251277683090998e-05, "loss": 0.0224, "step": 80290 }, { "grad_norm": 0.1896127313375473, "learning_rate": 1.0241249199157826e-05, "loss": 0.016, "step": 80300 }, { "grad_norm": 0.23338913917541504, "learning_rate": 1.0231225063181837e-05, "loss": 0.0214, "step": 80310 }, { "grad_norm": 0.36767658591270447, "learning_rate": 1.0221205276259316e-05, "loss": 0.0229, "step": 80320 }, { "grad_norm": 0.38910597562789917, "learning_rate": 1.0211189839485958e-05, "loss": 0.0189, "step": 80330 }, { "grad_norm": 0.437924861907959, "learning_rate": 1.0201178753957075e-05, "loss": 0.0209, "step": 80340 }, { "grad_norm": 0.26495200395584106, "learning_rate": 1.0191172020767432e-05, "loss": 0.0211, "step": 80350 }, { "grad_norm": 0.22947975993156433, "learning_rate": 1.0181169641011362e-05, "loss": 0.0326, "step": 80360 }, { "grad_norm": 0.4355228543281555, "learning_rate": 1.0171171615782721e-05, "loss": 0.0308, "step": 80370 }, { "grad_norm": 0.5178399085998535, "learning_rate": 1.0161177946174866e-05, "loss": 0.0235, "step": 80380 }, { "grad_norm": 0.3212535083293915, "learning_rate": 1.0151188633280706e-05, "loss": 0.0212, "step": 80390 }, { "grad_norm": 0.35233646631240845, "learning_rate": 1.0141203678192635e-05, "loss": 0.0237, "step": 80400 }, { "grad_norm": 0.6000458598136902, "learning_rate": 1.0131223082002612e-05, "loss": 0.0304, "step": 80410 }, { "grad_norm": 0.3826501965522766, "learning_rate": 1.012124684580208e-05, "loss": 0.0202, "step": 80420 }, { "grad_norm": 0.6367318034172058, "learning_rate": 1.0111274970682039e-05, "loss": 0.0227, "step": 80430 }, { "grad_norm": 1.3007633686065674, "learning_rate": 1.010130745773299e-05, "loss": 0.0263, "step": 80440 }, { "grad_norm": 0.503541111946106, "learning_rate": 1.0091344308044986e-05, "loss": 0.0242, "step": 80450 }, { "grad_norm": 0.7379726767539978, "learning_rate": 1.0081385522707554e-05, "loss": 0.0166, "step": 80460 }, { "grad_norm": 0.41295263171195984, "learning_rate": 1.0071431102809803e-05, "loss": 0.0201, "step": 80470 }, { "grad_norm": 0.6524457335472107, "learning_rate": 1.0061481049440297e-05, "loss": 0.0225, "step": 80480 }, { "grad_norm": 0.20012541115283966, "learning_rate": 1.0051535363687187e-05, "loss": 0.0264, "step": 80490 }, { "grad_norm": 0.3444124162197113, "learning_rate": 1.0041594046638087e-05, "loss": 0.0218, "step": 80500 }, { "grad_norm": 0.24946513772010803, "learning_rate": 1.0031657099380204e-05, "loss": 0.0162, "step": 80510 }, { "grad_norm": 0.2892499268054962, "learning_rate": 1.0021724523000204e-05, "loss": 0.024, "step": 80520 }, { "grad_norm": 0.29667261242866516, "learning_rate": 1.001179631858431e-05, "loss": 0.0257, "step": 80530 }, { "grad_norm": 0.36354249715805054, "learning_rate": 1.0001872487218245e-05, "loss": 0.0243, "step": 80540 }, { "grad_norm": 0.42363566160202026, "learning_rate": 9.991953029987272e-06, "loss": 0.021, "step": 80550 }, { "grad_norm": 0.42792803049087524, "learning_rate": 9.982037947976147e-06, "loss": 0.0227, "step": 80560 }, { "grad_norm": 0.4599318504333496, "learning_rate": 9.972127242269208e-06, "loss": 0.0247, "step": 80570 }, { "grad_norm": 0.19770140945911407, "learning_rate": 9.962220913950238e-06, "loss": 0.0303, "step": 80580 }, { "grad_norm": 0.20007595419883728, "learning_rate": 9.952318964102591e-06, "loss": 0.0294, "step": 80590 }, { "grad_norm": 0.28392940759658813, "learning_rate": 9.942421393809148e-06, "loss": 0.0248, "step": 80600 }, { "grad_norm": 0.5892558693885803, "learning_rate": 9.932528204152264e-06, "loss": 0.0166, "step": 80610 }, { "grad_norm": 0.8164382576942444, "learning_rate": 9.922639396213856e-06, "loss": 0.0338, "step": 80620 }, { "grad_norm": 0.6025303602218628, "learning_rate": 9.912754971075344e-06, "loss": 0.0254, "step": 80630 }, { "grad_norm": 0.3358917534351349, "learning_rate": 9.902874929817696e-06, "loss": 0.0266, "step": 80640 }, { "grad_norm": 0.3239719867706299, "learning_rate": 9.892999273521342e-06, "loss": 0.0202, "step": 80650 }, { "grad_norm": 0.362861692905426, "learning_rate": 9.883128003266302e-06, "loss": 0.019, "step": 80660 }, { "grad_norm": 0.25644662976264954, "learning_rate": 9.87326112013206e-06, "loss": 0.0184, "step": 80670 }, { "grad_norm": 0.30788373947143555, "learning_rate": 9.863398625197662e-06, "loss": 0.0248, "step": 80680 }, { "grad_norm": 0.45869171619415283, "learning_rate": 9.853540519541615e-06, "loss": 0.0219, "step": 80690 }, { "grad_norm": 0.8031682372093201, "learning_rate": 9.843686804242053e-06, "loss": 0.0368, "step": 80700 }, { "grad_norm": 0.5244068503379822, "learning_rate": 9.833837480376506e-06, "loss": 0.0234, "step": 80710 }, { "grad_norm": 0.2875879108905792, "learning_rate": 9.82399254902212e-06, "loss": 0.0233, "step": 80720 }, { "grad_norm": 0.33663368225097656, "learning_rate": 9.814152011255494e-06, "loss": 0.0202, "step": 80730 }, { "grad_norm": 0.13515189290046692, "learning_rate": 9.804315868152797e-06, "loss": 0.0192, "step": 80740 }, { "grad_norm": 0.4098990261554718, "learning_rate": 9.79448412078966e-06, "loss": 0.0194, "step": 80750 }, { "grad_norm": 0.2803281545639038, "learning_rate": 9.784656770241312e-06, "loss": 0.0231, "step": 80760 }, { "grad_norm": 0.15049217641353607, "learning_rate": 9.77483381758243e-06, "loss": 0.0193, "step": 80770 }, { "grad_norm": 0.36650457978248596, "learning_rate": 9.76501526388725e-06, "loss": 0.0256, "step": 80780 }, { "grad_norm": 0.2102443426847458, "learning_rate": 9.7552011102295e-06, "loss": 0.0359, "step": 80790 }, { "grad_norm": 0.31326982378959656, "learning_rate": 9.745391357682449e-06, "loss": 0.0228, "step": 80800 }, { "grad_norm": 0.8352639675140381, "learning_rate": 9.735586007318875e-06, "loss": 0.0258, "step": 80810 }, { "grad_norm": 0.2139214724302292, "learning_rate": 9.725785060211096e-06, "loss": 0.0167, "step": 80820 }, { "grad_norm": 0.21180127561092377, "learning_rate": 9.715988517430896e-06, "loss": 0.02, "step": 80830 }, { "grad_norm": 0.3351506292819977, "learning_rate": 9.706196380049625e-06, "loss": 0.0282, "step": 80840 }, { "grad_norm": 0.3453655540943146, "learning_rate": 9.69640864913815e-06, "loss": 0.0214, "step": 80850 }, { "grad_norm": 0.19837713241577148, "learning_rate": 9.686625325766818e-06, "loss": 0.0193, "step": 80860 }, { "grad_norm": 0.2767930030822754, "learning_rate": 9.67684641100554e-06, "loss": 0.022, "step": 80870 }, { "grad_norm": 1.2115753889083862, "learning_rate": 9.667071905923692e-06, "loss": 0.0265, "step": 80880 }, { "grad_norm": 0.536919355392456, "learning_rate": 9.657301811590242e-06, "loss": 0.027, "step": 80890 }, { "grad_norm": 0.3432914912700653, "learning_rate": 9.647536129073603e-06, "loss": 0.021, "step": 80900 }, { "grad_norm": 0.36888861656188965, "learning_rate": 9.637774859441751e-06, "loss": 0.0231, "step": 80910 }, { "grad_norm": 0.4259331524372101, "learning_rate": 9.628018003762152e-06, "loss": 0.0186, "step": 80920 }, { "grad_norm": 0.9541062712669373, "learning_rate": 9.618265563101813e-06, "loss": 0.0232, "step": 80930 }, { "grad_norm": 0.4869752824306488, "learning_rate": 9.608517538527211e-06, "loss": 0.0319, "step": 80940 }, { "grad_norm": 0.328850656747818, "learning_rate": 9.598773931104431e-06, "loss": 0.0212, "step": 80950 }, { "grad_norm": 0.6801923513412476, "learning_rate": 9.589034741898983e-06, "loss": 0.0254, "step": 80960 }, { "grad_norm": 0.3514783978462219, "learning_rate": 9.579299971975947e-06, "loss": 0.0214, "step": 80970 }, { "grad_norm": 0.5278856754302979, "learning_rate": 9.569569622399882e-06, "loss": 0.0167, "step": 80980 }, { "grad_norm": 0.5222499966621399, "learning_rate": 9.559843694234916e-06, "loss": 0.0191, "step": 80990 }, { "grad_norm": 0.54195237159729, "learning_rate": 9.550122188544625e-06, "loss": 0.0165, "step": 81000 }, { "grad_norm": 0.4063059985637665, "learning_rate": 9.540405106392152e-06, "loss": 0.026, "step": 81010 }, { "grad_norm": 0.34267768263816833, "learning_rate": 9.53069244884015e-06, "loss": 0.0252, "step": 81020 }, { "grad_norm": 0.2680736482143402, "learning_rate": 9.52098421695079e-06, "loss": 0.0242, "step": 81030 }, { "grad_norm": 0.3011215329170227, "learning_rate": 9.511280411785728e-06, "loss": 0.0185, "step": 81040 }, { "grad_norm": 0.5797920823097229, "learning_rate": 9.50158103440616e-06, "loss": 0.032, "step": 81050 }, { "grad_norm": 0.5410269498825073, "learning_rate": 9.491886085872815e-06, "loss": 0.0302, "step": 81060 }, { "grad_norm": 0.32193318009376526, "learning_rate": 9.482195567245895e-06, "loss": 0.0228, "step": 81070 }, { "grad_norm": 0.2960798144340515, "learning_rate": 9.472509479585151e-06, "loss": 0.0196, "step": 81080 }, { "grad_norm": 0.3081463873386383, "learning_rate": 9.462827823949833e-06, "loss": 0.0157, "step": 81090 }, { "grad_norm": 0.5625726580619812, "learning_rate": 9.453150601398735e-06, "loss": 0.0234, "step": 81100 }, { "grad_norm": 0.34219396114349365, "learning_rate": 9.44347781299011e-06, "loss": 0.0251, "step": 81110 }, { "grad_norm": 0.24341849982738495, "learning_rate": 9.43380945978179e-06, "loss": 0.02, "step": 81120 }, { "grad_norm": 0.30909264087677, "learning_rate": 9.424145542831047e-06, "loss": 0.0207, "step": 81130 }, { "grad_norm": 0.22099247574806213, "learning_rate": 9.414486063194778e-06, "loss": 0.0251, "step": 81140 }, { "grad_norm": 0.42849040031433105, "learning_rate": 9.404831021929273e-06, "loss": 0.0198, "step": 81150 }, { "grad_norm": 1.0913872718811035, "learning_rate": 9.395180420090427e-06, "loss": 0.0222, "step": 81160 }, { "grad_norm": 0.18723031878471375, "learning_rate": 9.385534258733591e-06, "loss": 0.0164, "step": 81170 }, { "grad_norm": 0.37708380818367004, "learning_rate": 9.375892538913677e-06, "loss": 0.0236, "step": 81180 }, { "grad_norm": 0.19105352461338043, "learning_rate": 9.36625526168507e-06, "loss": 0.0188, "step": 81190 }, { "grad_norm": 0.31371355056762695, "learning_rate": 9.356622428101686e-06, "loss": 0.0235, "step": 81200 }, { "grad_norm": 0.35792428255081177, "learning_rate": 9.346994039216972e-06, "loss": 0.0221, "step": 81210 }, { "grad_norm": 0.3539413809776306, "learning_rate": 9.337370096083886e-06, "loss": 0.0235, "step": 81220 }, { "grad_norm": 0.14962223172187805, "learning_rate": 9.327750599754854e-06, "loss": 0.0238, "step": 81230 }, { "grad_norm": 0.4440276324748993, "learning_rate": 9.318135551281875e-06, "loss": 0.0295, "step": 81240 }, { "grad_norm": 0.3439524471759796, "learning_rate": 9.308524951716424e-06, "loss": 0.0231, "step": 81250 }, { "grad_norm": 0.17117908596992493, "learning_rate": 9.298918802109497e-06, "loss": 0.0185, "step": 81260 }, { "grad_norm": 0.3588266670703888, "learning_rate": 9.289317103511625e-06, "loss": 0.0192, "step": 81270 }, { "grad_norm": 0.6578691601753235, "learning_rate": 9.279719856972835e-06, "loss": 0.0228, "step": 81280 }, { "grad_norm": 0.5548902750015259, "learning_rate": 9.270127063542645e-06, "loss": 0.0218, "step": 81290 }, { "grad_norm": 0.32870304584503174, "learning_rate": 9.260538724270124e-06, "loss": 0.0193, "step": 81300 }, { "grad_norm": 0.23706203699111938, "learning_rate": 9.250954840203852e-06, "loss": 0.0266, "step": 81310 }, { "grad_norm": 0.2804686427116394, "learning_rate": 9.241375412391878e-06, "loss": 0.0293, "step": 81320 }, { "grad_norm": 0.24086780846118927, "learning_rate": 9.23180044188181e-06, "loss": 0.0248, "step": 81330 }, { "grad_norm": 0.2472156435251236, "learning_rate": 9.222229929720755e-06, "loss": 0.0217, "step": 81340 }, { "grad_norm": 0.304354727268219, "learning_rate": 9.212663876955336e-06, "loss": 0.0385, "step": 81350 }, { "grad_norm": 0.41243061423301697, "learning_rate": 9.20310228463166e-06, "loss": 0.0269, "step": 81360 }, { "grad_norm": 0.33682525157928467, "learning_rate": 9.19354515379539e-06, "loss": 0.0201, "step": 81370 }, { "grad_norm": 0.25376519560813904, "learning_rate": 9.183992485491666e-06, "loss": 0.0208, "step": 81380 }, { "grad_norm": 0.1828087419271469, "learning_rate": 9.174444280765154e-06, "loss": 0.0244, "step": 81390 }, { "grad_norm": 0.7422009110450745, "learning_rate": 9.164900540660032e-06, "loss": 0.0254, "step": 81400 }, { "grad_norm": 0.25832074880599976, "learning_rate": 9.155361266220008e-06, "loss": 0.02, "step": 81410 }, { "grad_norm": 0.3254994750022888, "learning_rate": 9.145826458488254e-06, "loss": 0.0207, "step": 81420 }, { "grad_norm": 0.2961796224117279, "learning_rate": 9.136296118507509e-06, "loss": 0.0215, "step": 81430 }, { "grad_norm": 0.30718907713890076, "learning_rate": 9.126770247319966e-06, "loss": 0.0211, "step": 81440 }, { "grad_norm": 0.2984037697315216, "learning_rate": 9.11724884596738e-06, "loss": 0.0161, "step": 81450 }, { "grad_norm": 0.5188777446746826, "learning_rate": 9.107731915490996e-06, "loss": 0.0186, "step": 81460 }, { "grad_norm": 0.20791226625442505, "learning_rate": 9.098219456931579e-06, "loss": 0.0202, "step": 81470 }, { "grad_norm": 0.2763707637786865, "learning_rate": 9.088711471329376e-06, "loss": 0.0275, "step": 81480 }, { "grad_norm": 0.2047887146472931, "learning_rate": 9.079207959724184e-06, "loss": 0.0258, "step": 81490 }, { "grad_norm": 0.2895124852657318, "learning_rate": 9.069708923155279e-06, "loss": 0.0254, "step": 81500 }, { "grad_norm": 0.4635074734687805, "learning_rate": 9.060214362661468e-06, "loss": 0.02, "step": 81510 }, { "grad_norm": 0.32612794637680054, "learning_rate": 9.050724279281064e-06, "loss": 0.0216, "step": 81520 }, { "grad_norm": 0.3690911829471588, "learning_rate": 9.041238674051894e-06, "loss": 0.0206, "step": 81530 }, { "grad_norm": 0.3404921293258667, "learning_rate": 9.031757548011277e-06, "loss": 0.029, "step": 81540 }, { "grad_norm": 0.3293006718158722, "learning_rate": 9.022280902196051e-06, "loss": 0.0196, "step": 81550 }, { "grad_norm": 0.39008259773254395, "learning_rate": 9.012808737642592e-06, "loss": 0.0257, "step": 81560 }, { "grad_norm": 0.3306809663772583, "learning_rate": 9.003341055386739e-06, "loss": 0.0259, "step": 81570 }, { "grad_norm": 0.30891939997673035, "learning_rate": 8.993877856463862e-06, "loss": 0.0137, "step": 81580 }, { "grad_norm": 0.9435792565345764, "learning_rate": 8.984419141908856e-06, "loss": 0.0198, "step": 81590 }, { "grad_norm": 0.8529148101806641, "learning_rate": 8.974964912756117e-06, "loss": 0.0201, "step": 81600 }, { "grad_norm": 0.4446858763694763, "learning_rate": 8.965515170039523e-06, "loss": 0.0196, "step": 81610 }, { "grad_norm": 0.2550327479839325, "learning_rate": 8.956069914792514e-06, "loss": 0.022, "step": 81620 }, { "grad_norm": 1.8449194431304932, "learning_rate": 8.946629148047969e-06, "loss": 0.026, "step": 81630 }, { "grad_norm": 0.2129279375076294, "learning_rate": 8.93719287083834e-06, "loss": 0.0297, "step": 81640 }, { "grad_norm": 0.4818505346775055, "learning_rate": 8.927761084195569e-06, "loss": 0.0195, "step": 81650 }, { "grad_norm": 0.30730193853378296, "learning_rate": 8.918333789151096e-06, "loss": 0.0259, "step": 81660 }, { "grad_norm": 0.4602147340774536, "learning_rate": 8.90891098673587e-06, "loss": 0.026, "step": 81670 }, { "grad_norm": 0.3213822841644287, "learning_rate": 8.89949267798037e-06, "loss": 0.0328, "step": 81680 }, { "grad_norm": 0.49545520544052124, "learning_rate": 8.890078863914542e-06, "loss": 0.0321, "step": 81690 }, { "grad_norm": 0.30919718742370605, "learning_rate": 8.880669545567894e-06, "loss": 0.0209, "step": 81700 }, { "grad_norm": 0.21286725997924805, "learning_rate": 8.871264723969381e-06, "loss": 0.0256, "step": 81710 }, { "grad_norm": 0.30778658390045166, "learning_rate": 8.86186440014754e-06, "loss": 0.0202, "step": 81720 }, { "grad_norm": 0.24501994252204895, "learning_rate": 8.852468575130351e-06, "loss": 0.0298, "step": 81730 }, { "grad_norm": 0.65543532371521, "learning_rate": 8.843077249945347e-06, "loss": 0.0262, "step": 81740 }, { "grad_norm": 0.3554142415523529, "learning_rate": 8.833690425619523e-06, "loss": 0.0228, "step": 81750 }, { "grad_norm": 0.2420978993177414, "learning_rate": 8.82430810317942e-06, "loss": 0.0276, "step": 81760 }, { "grad_norm": 1.0065507888793945, "learning_rate": 8.814930283651074e-06, "loss": 0.022, "step": 81770 }, { "grad_norm": 0.28843143582344055, "learning_rate": 8.805556968060048e-06, "loss": 0.0214, "step": 81780 }, { "grad_norm": 0.3392723798751831, "learning_rate": 8.796188157431362e-06, "loss": 0.0305, "step": 81790 }, { "grad_norm": 0.23191703855991364, "learning_rate": 8.786823852789595e-06, "loss": 0.0249, "step": 81800 }, { "grad_norm": 1.1176106929779053, "learning_rate": 8.777464055158813e-06, "loss": 0.0239, "step": 81810 }, { "grad_norm": 0.23515863716602325, "learning_rate": 8.768108765562578e-06, "loss": 0.0243, "step": 81820 }, { "grad_norm": 0.4105947017669678, "learning_rate": 8.758757985023975e-06, "loss": 0.0222, "step": 81830 }, { "grad_norm": 0.2758994996547699, "learning_rate": 8.749411714565598e-06, "loss": 0.0204, "step": 81840 }, { "grad_norm": 0.17481958866119385, "learning_rate": 8.74006995520954e-06, "loss": 0.0293, "step": 81850 }, { "grad_norm": 0.2802119255065918, "learning_rate": 8.730732707977397e-06, "loss": 0.0204, "step": 81860 }, { "grad_norm": 0.3828364312648773, "learning_rate": 8.721399973890281e-06, "loss": 0.0228, "step": 81870 }, { "grad_norm": 0.4145788252353668, "learning_rate": 8.712071753968792e-06, "loss": 0.027, "step": 81880 }, { "grad_norm": 0.7374309301376343, "learning_rate": 8.702748049233072e-06, "loss": 0.0274, "step": 81890 }, { "grad_norm": 0.2524978816509247, "learning_rate": 8.693428860702718e-06, "loss": 0.0195, "step": 81900 }, { "grad_norm": 0.4651292562484741, "learning_rate": 8.684114189396898e-06, "loss": 0.0287, "step": 81910 }, { "grad_norm": 0.17024943232536316, "learning_rate": 8.674804036334222e-06, "loss": 0.0227, "step": 81920 }, { "grad_norm": 0.35547223687171936, "learning_rate": 8.665498402532857e-06, "loss": 0.0225, "step": 81930 }, { "grad_norm": 0.4355204999446869, "learning_rate": 8.656197289010432e-06, "loss": 0.0206, "step": 81940 }, { "grad_norm": 0.1897183060646057, "learning_rate": 8.646900696784122e-06, "loss": 0.0177, "step": 81950 }, { "grad_norm": 0.3678685426712036, "learning_rate": 8.637608626870553e-06, "loss": 0.0253, "step": 81960 }, { "grad_norm": 0.19227156043052673, "learning_rate": 8.628321080285944e-06, "loss": 0.0193, "step": 81970 }, { "grad_norm": 0.23729069530963898, "learning_rate": 8.619038058045925e-06, "loss": 0.0214, "step": 81980 }, { "grad_norm": 0.17121803760528564, "learning_rate": 8.609759561165704e-06, "loss": 0.0317, "step": 81990 }, { "grad_norm": 0.171129509806633, "learning_rate": 8.600485590659934e-06, "loss": 0.0211, "step": 82000 }, { "grad_norm": 0.175893634557724, "learning_rate": 8.591216147542813e-06, "loss": 0.0202, "step": 82010 }, { "grad_norm": 0.20683850347995758, "learning_rate": 8.581951232828045e-06, "loss": 0.0361, "step": 82020 }, { "grad_norm": 0.5374031066894531, "learning_rate": 8.572690847528825e-06, "loss": 0.0205, "step": 82030 }, { "grad_norm": 0.24212488532066345, "learning_rate": 8.563434992657838e-06, "loss": 0.0244, "step": 82040 }, { "grad_norm": 0.2250581681728363, "learning_rate": 8.554183669227307e-06, "loss": 0.0201, "step": 82050 }, { "grad_norm": 0.26420050859451294, "learning_rate": 8.544936878248937e-06, "loss": 0.0246, "step": 82060 }, { "grad_norm": 0.22651471197605133, "learning_rate": 8.535694620733942e-06, "loss": 0.0272, "step": 82070 }, { "grad_norm": 0.3342150151729584, "learning_rate": 8.526456897693053e-06, "loss": 0.0286, "step": 82080 }, { "grad_norm": 0.38299450278282166, "learning_rate": 8.517223710136456e-06, "loss": 0.0351, "step": 82090 }, { "grad_norm": 0.37383580207824707, "learning_rate": 8.507995059073931e-06, "loss": 0.0255, "step": 82100 }, { "grad_norm": 0.3077946901321411, "learning_rate": 8.498770945514673e-06, "loss": 0.0172, "step": 82110 }, { "grad_norm": 0.19492675364017487, "learning_rate": 8.489551370467436e-06, "loss": 0.0165, "step": 82120 }, { "grad_norm": 0.2773280739784241, "learning_rate": 8.480336334940442e-06, "loss": 0.0236, "step": 82130 }, { "grad_norm": 0.29570409655570984, "learning_rate": 8.471125839941458e-06, "loss": 0.0232, "step": 82140 }, { "grad_norm": 0.41229984164237976, "learning_rate": 8.461919886477682e-06, "loss": 0.0217, "step": 82150 }, { "grad_norm": 0.43847882747650146, "learning_rate": 8.452718475555927e-06, "loss": 0.0203, "step": 82160 }, { "grad_norm": 0.19182540476322174, "learning_rate": 8.443521608182398e-06, "loss": 0.0179, "step": 82170 }, { "grad_norm": 1.277091145515442, "learning_rate": 8.43432928536288e-06, "loss": 0.0177, "step": 82180 }, { "grad_norm": 0.3513268232345581, "learning_rate": 8.425141508102607e-06, "loss": 0.0294, "step": 82190 }, { "grad_norm": 0.5596058368682861, "learning_rate": 8.415958277406365e-06, "loss": 0.0257, "step": 82200 }, { "grad_norm": 0.31540679931640625, "learning_rate": 8.406779594278392e-06, "loss": 0.0214, "step": 82210 }, { "grad_norm": 0.4722341299057007, "learning_rate": 8.397605459722468e-06, "loss": 0.0285, "step": 82220 }, { "grad_norm": 0.4429866075515747, "learning_rate": 8.388435874741863e-06, "loss": 0.0295, "step": 82230 }, { "grad_norm": 0.27646422386169434, "learning_rate": 8.37927084033936e-06, "loss": 0.0185, "step": 82240 }, { "grad_norm": 0.295105516910553, "learning_rate": 8.370110357517208e-06, "loss": 0.024, "step": 82250 }, { "grad_norm": 0.3908773362636566, "learning_rate": 8.360954427277201e-06, "loss": 0.023, "step": 82260 }, { "grad_norm": 0.45534253120422363, "learning_rate": 8.351803050620627e-06, "loss": 0.0203, "step": 82270 }, { "grad_norm": 0.3949257731437683, "learning_rate": 8.34265622854824e-06, "loss": 0.0226, "step": 82280 }, { "grad_norm": 0.37424084544181824, "learning_rate": 8.333513962060335e-06, "loss": 0.0336, "step": 82290 }, { "grad_norm": 0.3053850531578064, "learning_rate": 8.324376252156701e-06, "loss": 0.027, "step": 82300 }, { "grad_norm": 0.19277650117874146, "learning_rate": 8.315243099836633e-06, "loss": 0.0235, "step": 82310 }, { "grad_norm": 0.1968526840209961, "learning_rate": 8.3061145060989e-06, "loss": 0.0218, "step": 82320 }, { "grad_norm": 0.2522926926612854, "learning_rate": 8.296990471941806e-06, "loss": 0.0184, "step": 82330 }, { "grad_norm": 0.2836960554122925, "learning_rate": 8.287870998363112e-06, "loss": 0.0216, "step": 82340 }, { "grad_norm": 0.2805722951889038, "learning_rate": 8.278756086360156e-06, "loss": 0.0191, "step": 82350 }, { "grad_norm": 0.1346452236175537, "learning_rate": 8.269645736929693e-06, "loss": 0.0211, "step": 82360 }, { "grad_norm": 0.3569132387638092, "learning_rate": 8.260539951068048e-06, "loss": 0.0252, "step": 82370 }, { "grad_norm": 0.5483736991882324, "learning_rate": 8.251438729770983e-06, "loss": 0.0342, "step": 82380 }, { "grad_norm": 0.35363849997520447, "learning_rate": 8.242342074033826e-06, "loss": 0.0161, "step": 82390 }, { "grad_norm": 0.3300602436065674, "learning_rate": 8.233249984851344e-06, "loss": 0.0305, "step": 82400 }, { "grad_norm": 0.4003780484199524, "learning_rate": 8.224162463217849e-06, "loss": 0.0231, "step": 82410 }, { "grad_norm": 0.4311531186103821, "learning_rate": 8.215079510127139e-06, "loss": 0.0232, "step": 82420 }, { "grad_norm": 0.2804718613624573, "learning_rate": 8.206001126572521e-06, "loss": 0.0239, "step": 82430 }, { "grad_norm": 0.34756794571876526, "learning_rate": 8.196927313546776e-06, "loss": 0.0214, "step": 82440 }, { "grad_norm": 0.38248082995414734, "learning_rate": 8.18785807204222e-06, "loss": 0.0152, "step": 82450 }, { "grad_norm": 0.18905316293239594, "learning_rate": 8.178793403050627e-06, "loss": 0.0216, "step": 82460 }, { "grad_norm": 0.32976004481315613, "learning_rate": 8.16973330756331e-06, "loss": 0.0217, "step": 82470 }, { "grad_norm": 0.42043939232826233, "learning_rate": 8.16067778657107e-06, "loss": 0.0201, "step": 82480 }, { "grad_norm": 0.2184305638074875, "learning_rate": 8.151626841064214e-06, "loss": 0.021, "step": 82490 }, { "grad_norm": 0.3423043191432953, "learning_rate": 8.142580472032518e-06, "loss": 0.0201, "step": 82500 }, { "grad_norm": 0.5460172891616821, "learning_rate": 8.133538680465285e-06, "loss": 0.0278, "step": 82510 }, { "grad_norm": 0.3269585371017456, "learning_rate": 8.124501467351336e-06, "loss": 0.029, "step": 82520 }, { "grad_norm": 0.22315800189971924, "learning_rate": 8.11546883367893e-06, "loss": 0.0191, "step": 82530 }, { "grad_norm": 0.5609590411186218, "learning_rate": 8.106440780435882e-06, "loss": 0.0211, "step": 82540 }, { "grad_norm": 0.7394909858703613, "learning_rate": 8.097417308609485e-06, "loss": 0.0184, "step": 82550 }, { "grad_norm": 0.319002628326416, "learning_rate": 8.088398419186544e-06, "loss": 0.0279, "step": 82560 }, { "grad_norm": 0.26437056064605713, "learning_rate": 8.079384113153321e-06, "loss": 0.0201, "step": 82570 }, { "grad_norm": 0.6581612229347229, "learning_rate": 8.070374391495644e-06, "loss": 0.0162, "step": 82580 }, { "grad_norm": 0.38329994678497314, "learning_rate": 8.061369255198764e-06, "loss": 0.0242, "step": 82590 }, { "grad_norm": 0.3192741870880127, "learning_rate": 8.052368705247493e-06, "loss": 0.0194, "step": 82600 }, { "grad_norm": 0.4061248004436493, "learning_rate": 8.043372742626114e-06, "loss": 0.0276, "step": 82610 }, { "grad_norm": 0.23200024664402008, "learning_rate": 8.034381368318422e-06, "loss": 0.0178, "step": 82620 }, { "grad_norm": 0.16224215924739838, "learning_rate": 8.025394583307677e-06, "loss": 0.0194, "step": 82630 }, { "grad_norm": 0.38873159885406494, "learning_rate": 8.016412388576683e-06, "loss": 0.026, "step": 82640 }, { "grad_norm": 0.3059386909008026, "learning_rate": 8.007434785107699e-06, "loss": 0.0198, "step": 82650 }, { "grad_norm": 2.720639705657959, "learning_rate": 7.998461773882509e-06, "loss": 0.0297, "step": 82660 }, { "grad_norm": 0.3274705410003662, "learning_rate": 7.989493355882393e-06, "loss": 0.025, "step": 82670 }, { "grad_norm": 0.3940947651863098, "learning_rate": 7.980529532088137e-06, "loss": 0.0205, "step": 82680 }, { "grad_norm": 0.2709861397743225, "learning_rate": 7.97157030347998e-06, "loss": 0.0195, "step": 82690 }, { "grad_norm": 0.9822914600372314, "learning_rate": 7.96261567103772e-06, "loss": 0.0215, "step": 82700 }, { "grad_norm": 0.18063777685165405, "learning_rate": 7.953665635740593e-06, "loss": 0.0237, "step": 82710 }, { "grad_norm": 0.6412606239318848, "learning_rate": 7.94472019856738e-06, "loss": 0.026, "step": 82720 }, { "grad_norm": 0.3961597681045532, "learning_rate": 7.935779360496337e-06, "loss": 0.0266, "step": 82730 }, { "grad_norm": 0.4148726463317871, "learning_rate": 7.92684312250523e-06, "loss": 0.0225, "step": 82740 }, { "grad_norm": 0.41120222210884094, "learning_rate": 7.917911485571289e-06, "loss": 0.0166, "step": 82750 }, { "grad_norm": 0.19743013381958008, "learning_rate": 7.90898445067128e-06, "loss": 0.0211, "step": 82760 }, { "grad_norm": 0.5237624645233154, "learning_rate": 7.900062018781462e-06, "loss": 0.0243, "step": 82770 }, { "grad_norm": 0.45750442147254944, "learning_rate": 7.891144190877553e-06, "loss": 0.0237, "step": 82780 }, { "grad_norm": 0.29993772506713867, "learning_rate": 7.882230967934806e-06, "loss": 0.0195, "step": 82790 }, { "grad_norm": 0.23017257452011108, "learning_rate": 7.873322350927959e-06, "loss": 0.0262, "step": 82800 }, { "grad_norm": 0.23880361020565033, "learning_rate": 7.864418340831253e-06, "loss": 0.0217, "step": 82810 }, { "grad_norm": 0.2777990698814392, "learning_rate": 7.855518938618395e-06, "loss": 0.0188, "step": 82820 }, { "grad_norm": 0.3705531060695648, "learning_rate": 7.846624145262633e-06, "loss": 0.0211, "step": 82830 }, { "grad_norm": 0.2602616548538208, "learning_rate": 7.837733961736672e-06, "loss": 0.0214, "step": 82840 }, { "grad_norm": 0.22302870452404022, "learning_rate": 7.828848389012727e-06, "loss": 0.025, "step": 82850 }, { "grad_norm": 0.2972079813480377, "learning_rate": 7.819967428062524e-06, "loss": 0.0237, "step": 82860 }, { "grad_norm": 0.25800248980522156, "learning_rate": 7.811091079857274e-06, "loss": 0.0207, "step": 82870 }, { "grad_norm": 0.25811994075775146, "learning_rate": 7.802219345367667e-06, "loss": 0.023, "step": 82880 }, { "grad_norm": 0.29697051644325256, "learning_rate": 7.793352225563916e-06, "loss": 0.0238, "step": 82890 }, { "grad_norm": 0.26441267132759094, "learning_rate": 7.7844897214157e-06, "loss": 0.0289, "step": 82900 }, { "grad_norm": 0.42817333340644836, "learning_rate": 7.775631833892228e-06, "loss": 0.0199, "step": 82910 }, { "grad_norm": 0.2634131610393524, "learning_rate": 7.76677856396215e-06, "loss": 0.0205, "step": 82920 }, { "grad_norm": 0.2919654846191406, "learning_rate": 7.757929912593697e-06, "loss": 0.0287, "step": 82930 }, { "grad_norm": 0.49127641320228577, "learning_rate": 7.749085880754509e-06, "loss": 0.0205, "step": 82940 }, { "grad_norm": 0.2978138029575348, "learning_rate": 7.740246469411777e-06, "loss": 0.026, "step": 82950 }, { "grad_norm": 0.6239769458770752, "learning_rate": 7.731411679532142e-06, "loss": 0.0221, "step": 82960 }, { "grad_norm": 0.2981143295764923, "learning_rate": 7.722581512081779e-06, "loss": 0.0244, "step": 82970 }, { "grad_norm": 0.33988338708877563, "learning_rate": 7.713755968026331e-06, "loss": 0.0196, "step": 82980 }, { "grad_norm": 0.24909336864948273, "learning_rate": 7.704935048330969e-06, "loss": 0.0238, "step": 82990 }, { "grad_norm": 0.44785332679748535, "learning_rate": 7.696118753960313e-06, "loss": 0.0186, "step": 83000 }, { "grad_norm": 0.2081882208585739, "learning_rate": 7.687307085878503e-06, "loss": 0.0182, "step": 83010 }, { "grad_norm": 0.5924224257469177, "learning_rate": 7.678500045049186e-06, "loss": 0.0207, "step": 83020 }, { "grad_norm": 0.498151570558548, "learning_rate": 7.669697632435463e-06, "loss": 0.0255, "step": 83030 }, { "grad_norm": 0.24540351331233978, "learning_rate": 7.660899848999963e-06, "loss": 0.0215, "step": 83040 }, { "grad_norm": 0.28212055563926697, "learning_rate": 7.652106695704803e-06, "loss": 0.0263, "step": 83050 }, { "grad_norm": 0.303908109664917, "learning_rate": 7.643318173511598e-06, "loss": 0.025, "step": 83060 }, { "grad_norm": 0.2111225426197052, "learning_rate": 7.634534283381423e-06, "loss": 0.0183, "step": 83070 }, { "grad_norm": 0.27743959426879883, "learning_rate": 7.625755026274889e-06, "loss": 0.0209, "step": 83080 }, { "grad_norm": 0.4671633243560791, "learning_rate": 7.616980403152069e-06, "loss": 0.0229, "step": 83090 }, { "grad_norm": 0.4079831540584564, "learning_rate": 7.608210414972561e-06, "loss": 0.0286, "step": 83100 }, { "grad_norm": 0.46496185660362244, "learning_rate": 7.599445062695404e-06, "loss": 0.036, "step": 83110 }, { "grad_norm": 0.24014724791049957, "learning_rate": 7.590684347279214e-06, "loss": 0.0243, "step": 83120 }, { "grad_norm": 0.26864683628082275, "learning_rate": 7.581928269682004e-06, "loss": 0.0272, "step": 83130 }, { "grad_norm": 0.3531283736228943, "learning_rate": 7.573176830861361e-06, "loss": 0.0188, "step": 83140 }, { "grad_norm": 0.2960503101348877, "learning_rate": 7.5644300317743e-06, "loss": 0.0202, "step": 83150 }, { "grad_norm": 0.352591335773468, "learning_rate": 7.555687873377376e-06, "loss": 0.0219, "step": 83160 }, { "grad_norm": 0.28992903232574463, "learning_rate": 7.546950356626592e-06, "loss": 0.0318, "step": 83170 }, { "grad_norm": 0.3766695261001587, "learning_rate": 7.538217482477516e-06, "loss": 0.0219, "step": 83180 }, { "grad_norm": 0.26587778329849243, "learning_rate": 7.529489251885119e-06, "loss": 0.0253, "step": 83190 }, { "grad_norm": 0.21835094690322876, "learning_rate": 7.520765665803941e-06, "loss": 0.0209, "step": 83200 }, { "grad_norm": 0.2500433921813965, "learning_rate": 7.512046725187949e-06, "loss": 0.025, "step": 83210 }, { "grad_norm": 0.24284671247005463, "learning_rate": 7.503332430990645e-06, "loss": 0.0176, "step": 83220 }, { "grad_norm": 0.25224602222442627, "learning_rate": 7.494622784165017e-06, "loss": 0.0244, "step": 83230 }, { "grad_norm": 0.34387287497520447, "learning_rate": 7.485917785663543e-06, "loss": 0.0213, "step": 83240 }, { "grad_norm": 0.46388518810272217, "learning_rate": 7.477217436438172e-06, "loss": 0.023, "step": 83250 }, { "grad_norm": 0.47095873951911926, "learning_rate": 7.46852173744037e-06, "loss": 0.0253, "step": 83260 }, { "grad_norm": 0.24902024865150452, "learning_rate": 7.4598306896211004e-06, "loss": 0.0261, "step": 83270 }, { "grad_norm": 0.4164895713329315, "learning_rate": 7.451144293930773e-06, "loss": 0.0271, "step": 83280 }, { "grad_norm": 0.3994908034801483, "learning_rate": 7.4424625513193455e-06, "loss": 0.0215, "step": 83290 }, { "grad_norm": 0.2801365256309509, "learning_rate": 7.433785462736209e-06, "loss": 0.0192, "step": 83300 }, { "grad_norm": 0.22341203689575195, "learning_rate": 7.425113029130315e-06, "loss": 0.0235, "step": 83310 }, { "grad_norm": 0.2798765301704407, "learning_rate": 7.416445251450044e-06, "loss": 0.0242, "step": 83320 }, { "grad_norm": 0.2584085166454315, "learning_rate": 7.4077821306433035e-06, "loss": 0.0217, "step": 83330 }, { "grad_norm": 0.9566287398338318, "learning_rate": 7.399123667657465e-06, "loss": 0.0222, "step": 83340 }, { "grad_norm": 0.7310293316841125, "learning_rate": 7.390469863439426e-06, "loss": 0.0308, "step": 83350 }, { "grad_norm": 0.23736263811588287, "learning_rate": 7.381820718935512e-06, "loss": 0.0285, "step": 83360 }, { "grad_norm": 0.5519867539405823, "learning_rate": 7.373176235091633e-06, "loss": 0.0341, "step": 83370 }, { "grad_norm": 0.3291529715061188, "learning_rate": 7.364536412853101e-06, "loss": 0.0206, "step": 83380 }, { "grad_norm": 0.37922343611717224, "learning_rate": 7.355901253164776e-06, "loss": 0.0211, "step": 83390 }, { "grad_norm": 0.41150808334350586, "learning_rate": 7.34727075697097e-06, "loss": 0.0281, "step": 83400 }, { "grad_norm": 0.41320982575416565, "learning_rate": 7.3386449252155095e-06, "loss": 0.0277, "step": 83410 }, { "grad_norm": 0.46629562973976135, "learning_rate": 7.330023758841692e-06, "loss": 0.0239, "step": 83420 }, { "grad_norm": 0.18108999729156494, "learning_rate": 7.321407258792323e-06, "loss": 0.0156, "step": 83430 }, { "grad_norm": 0.26140543818473816, "learning_rate": 7.312795426009694e-06, "loss": 0.0195, "step": 83440 }, { "grad_norm": 0.3272831439971924, "learning_rate": 7.304188261435585e-06, "loss": 0.0216, "step": 83450 }, { "grad_norm": 0.38742271065711975, "learning_rate": 7.295585766011242e-06, "loss": 0.0277, "step": 83460 }, { "grad_norm": 0.32194966077804565, "learning_rate": 7.286987940677436e-06, "loss": 0.0198, "step": 83470 }, { "grad_norm": 0.721415102481842, "learning_rate": 7.278394786374426e-06, "loss": 0.0213, "step": 83480 }, { "grad_norm": 0.17983898520469666, "learning_rate": 7.269806304041915e-06, "loss": 0.0162, "step": 83490 }, { "grad_norm": 0.29844552278518677, "learning_rate": 7.261222494619147e-06, "loss": 0.0157, "step": 83500 }, { "grad_norm": 0.5457531213760376, "learning_rate": 7.252643359044825e-06, "loss": 0.0298, "step": 83510 }, { "grad_norm": 0.3271256983280182, "learning_rate": 7.244068898257172e-06, "loss": 0.0237, "step": 83520 }, { "grad_norm": 0.34835565090179443, "learning_rate": 7.2354991131938495e-06, "loss": 0.0175, "step": 83530 }, { "grad_norm": 0.2776297628879547, "learning_rate": 7.2269340047920574e-06, "loss": 0.0232, "step": 83540 }, { "grad_norm": 0.7464707493782043, "learning_rate": 7.218373573988435e-06, "loss": 0.0215, "step": 83550 }, { "grad_norm": 0.23602992296218872, "learning_rate": 7.209817821719178e-06, "loss": 0.03, "step": 83560 }, { "grad_norm": 0.36429518461227417, "learning_rate": 7.201266748919894e-06, "loss": 0.0242, "step": 83570 }, { "grad_norm": 0.23021256923675537, "learning_rate": 7.19272035652575e-06, "loss": 0.0201, "step": 83580 }, { "grad_norm": 0.5261588096618652, "learning_rate": 7.184178645471334e-06, "loss": 0.019, "step": 83590 }, { "grad_norm": 0.23011666536331177, "learning_rate": 7.175641616690775e-06, "loss": 0.0226, "step": 83600 }, { "grad_norm": 0.4455009996891022, "learning_rate": 7.167109271117661e-06, "loss": 0.0294, "step": 83610 }, { "grad_norm": 0.3409719169139862, "learning_rate": 7.158581609685072e-06, "loss": 0.0224, "step": 83620 }, { "grad_norm": 0.20337431132793427, "learning_rate": 7.150058633325591e-06, "loss": 0.0238, "step": 83630 }, { "grad_norm": 0.39154043793678284, "learning_rate": 7.141540342971287e-06, "loss": 0.0254, "step": 83640 }, { "grad_norm": 0.3943447470664978, "learning_rate": 7.133026739553677e-06, "loss": 0.024, "step": 83650 }, { "grad_norm": 0.4782091975212097, "learning_rate": 7.124517824003829e-06, "loss": 0.0257, "step": 83660 }, { "grad_norm": 0.49616265296936035, "learning_rate": 7.116013597252236e-06, "loss": 0.0235, "step": 83670 }, { "grad_norm": 0.33745333552360535, "learning_rate": 7.107514060228921e-06, "loss": 0.0241, "step": 83680 }, { "grad_norm": 0.3432190418243408, "learning_rate": 7.099019213863378e-06, "loss": 0.0245, "step": 83690 }, { "grad_norm": 0.42011111974716187, "learning_rate": 7.090529059084611e-06, "loss": 0.0244, "step": 83700 }, { "grad_norm": 0.16440241038799286, "learning_rate": 7.082043596821058e-06, "loss": 0.015, "step": 83710 }, { "grad_norm": 0.25685828924179077, "learning_rate": 7.07356282800069e-06, "loss": 0.0308, "step": 83720 }, { "grad_norm": 0.24533292651176453, "learning_rate": 7.0650867535509645e-06, "loss": 0.0244, "step": 83730 }, { "grad_norm": 0.35714098811149597, "learning_rate": 7.0566153743987895e-06, "loss": 0.0187, "step": 83740 }, { "grad_norm": 0.38197147846221924, "learning_rate": 7.04814869147059e-06, "loss": 0.0283, "step": 83750 }, { "grad_norm": 0.9365959763526917, "learning_rate": 7.0396867056922725e-06, "loss": 0.0215, "step": 83760 }, { "grad_norm": 0.3648175597190857, "learning_rate": 7.031229417989244e-06, "loss": 0.0245, "step": 83770 }, { "grad_norm": 0.2661552131175995, "learning_rate": 7.022776829286348e-06, "loss": 0.0227, "step": 83780 }, { "grad_norm": 0.4156728982925415, "learning_rate": 7.014328940507975e-06, "loss": 0.0219, "step": 83790 }, { "grad_norm": 0.270181268453598, "learning_rate": 7.005885752577945e-06, "loss": 0.0267, "step": 83800 }, { "grad_norm": 0.3836013376712799, "learning_rate": 6.997447266419615e-06, "loss": 0.0228, "step": 83810 }, { "grad_norm": 0.363766610622406, "learning_rate": 6.989013482955797e-06, "loss": 0.0181, "step": 83820 }, { "grad_norm": 0.2954426109790802, "learning_rate": 6.980584403108808e-06, "loss": 0.025, "step": 83830 }, { "grad_norm": 0.3788638412952423, "learning_rate": 6.972160027800417e-06, "loss": 0.0196, "step": 83840 }, { "grad_norm": 0.26631084084510803, "learning_rate": 6.963740357951931e-06, "loss": 0.0208, "step": 83850 }, { "grad_norm": 0.2558153569698334, "learning_rate": 6.95532539448408e-06, "loss": 0.0276, "step": 83860 }, { "grad_norm": 0.36348727345466614, "learning_rate": 6.946915138317128e-06, "loss": 0.0259, "step": 83870 }, { "grad_norm": 0.130931094288826, "learning_rate": 6.9385095903708065e-06, "loss": 0.0221, "step": 83880 }, { "grad_norm": 0.4491550028324127, "learning_rate": 6.930108751564346e-06, "loss": 0.0272, "step": 83890 }, { "grad_norm": 0.322002112865448, "learning_rate": 6.921712622816428e-06, "loss": 0.0227, "step": 83900 }, { "grad_norm": 0.2748622000217438, "learning_rate": 6.913321205045259e-06, "loss": 0.0207, "step": 83910 }, { "grad_norm": 0.28192177414894104, "learning_rate": 6.904934499168491e-06, "loss": 0.0202, "step": 83920 }, { "grad_norm": 0.2578703761100769, "learning_rate": 6.896552506103299e-06, "loss": 0.0218, "step": 83930 }, { "grad_norm": 0.3445051312446594, "learning_rate": 6.888175226766313e-06, "loss": 0.0341, "step": 83940 }, { "grad_norm": 0.8629656434059143, "learning_rate": 6.879802662073686e-06, "loss": 0.0171, "step": 83950 }, { "grad_norm": 0.4967665672302246, "learning_rate": 6.871434812940991e-06, "loss": 0.0329, "step": 83960 }, { "grad_norm": 0.24132296442985535, "learning_rate": 6.863071680283345e-06, "loss": 0.0241, "step": 83970 }, { "grad_norm": 0.2575376033782959, "learning_rate": 6.854713265015339e-06, "loss": 0.0365, "step": 83980 }, { "grad_norm": 0.5501908659934998, "learning_rate": 6.8463595680510025e-06, "loss": 0.0274, "step": 83990 }, { "grad_norm": 0.5289474129676819, "learning_rate": 6.838010590303911e-06, "loss": 0.0207, "step": 84000 }, { "grad_norm": 0.5076455473899841, "learning_rate": 6.829666332687085e-06, "loss": 0.0195, "step": 84010 }, { "grad_norm": 0.2814621031284332, "learning_rate": 6.821326796113059e-06, "loss": 0.0189, "step": 84020 }, { "grad_norm": 0.2912577688694, "learning_rate": 6.812991981493799e-06, "loss": 0.0201, "step": 84030 }, { "grad_norm": 0.903913140296936, "learning_rate": 6.804661889740821e-06, "loss": 0.0228, "step": 84040 }, { "grad_norm": 0.2288278490304947, "learning_rate": 6.7963365217650625e-06, "loss": 0.0178, "step": 84050 }, { "grad_norm": 0.2964758574962616, "learning_rate": 6.788015878476983e-06, "loss": 0.0222, "step": 84060 }, { "grad_norm": 0.32907867431640625, "learning_rate": 6.779699960786523e-06, "loss": 0.0211, "step": 84070 }, { "grad_norm": 0.28915226459503174, "learning_rate": 6.771388769603104e-06, "loss": 0.0177, "step": 84080 }, { "grad_norm": 0.24399442970752716, "learning_rate": 6.763082305835605e-06, "loss": 0.0362, "step": 84090 }, { "grad_norm": 0.48110342025756836, "learning_rate": 6.754780570392433e-06, "loss": 0.0218, "step": 84100 }, { "grad_norm": 0.5775101184844971, "learning_rate": 6.7464835641814215e-06, "loss": 0.025, "step": 84110 }, { "grad_norm": 1.797719955444336, "learning_rate": 6.7381912881099515e-06, "loss": 0.0231, "step": 84120 }, { "grad_norm": 0.3260740339756012, "learning_rate": 6.72990374308482e-06, "loss": 0.0213, "step": 84130 }, { "grad_norm": 0.4269065260887146, "learning_rate": 6.721620930012379e-06, "loss": 0.0203, "step": 84140 }, { "grad_norm": 0.41885998845100403, "learning_rate": 6.713342849798393e-06, "loss": 0.0179, "step": 84150 }, { "grad_norm": 0.2622363269329071, "learning_rate": 6.705069503348166e-06, "loss": 0.0289, "step": 84160 }, { "grad_norm": 0.3105601370334625, "learning_rate": 6.696800891566435e-06, "loss": 0.0235, "step": 84170 }, { "grad_norm": 0.32496780157089233, "learning_rate": 6.688537015357449e-06, "loss": 0.0215, "step": 84180 }, { "grad_norm": 0.3725205659866333, "learning_rate": 6.680277875624941e-06, "loss": 0.0278, "step": 84190 }, { "grad_norm": 0.36271408200263977, "learning_rate": 6.672023473272121e-06, "loss": 0.0167, "step": 84200 }, { "grad_norm": 0.2664220929145813, "learning_rate": 6.663773809201667e-06, "loss": 0.0199, "step": 84210 }, { "grad_norm": 0.26140591502189636, "learning_rate": 6.655528884315748e-06, "loss": 0.0236, "step": 84220 }, { "grad_norm": 0.3482436239719391, "learning_rate": 6.647288699516036e-06, "loss": 0.0204, "step": 84230 }, { "grad_norm": 0.2782691419124603, "learning_rate": 6.639053255703642e-06, "loss": 0.0221, "step": 84240 }, { "grad_norm": 0.394121915102005, "learning_rate": 6.6308225537791925e-06, "loss": 0.0231, "step": 84250 }, { "grad_norm": 0.3028896152973175, "learning_rate": 6.622596594642777e-06, "loss": 0.0187, "step": 84260 }, { "grad_norm": 0.16823306679725647, "learning_rate": 6.614375379193999e-06, "loss": 0.0245, "step": 84270 }, { "grad_norm": 0.26449689269065857, "learning_rate": 6.606158908331889e-06, "loss": 0.021, "step": 84280 }, { "grad_norm": 0.28553271293640137, "learning_rate": 6.597947182955005e-06, "loss": 0.0212, "step": 84290 }, { "grad_norm": 0.3392663598060608, "learning_rate": 6.589740203961359e-06, "loss": 0.0295, "step": 84300 }, { "grad_norm": 0.47662702202796936, "learning_rate": 6.581537972248464e-06, "loss": 0.0323, "step": 84310 }, { "grad_norm": 0.264552503824234, "learning_rate": 6.573340488713275e-06, "loss": 0.0287, "step": 84320 }, { "grad_norm": 0.4231495261192322, "learning_rate": 6.565147754252304e-06, "loss": 0.0204, "step": 84330 }, { "grad_norm": 0.29657936096191406, "learning_rate": 6.556959769761462e-06, "loss": 0.0255, "step": 84340 }, { "grad_norm": 0.2725220322608948, "learning_rate": 6.548776536136192e-06, "loss": 0.0229, "step": 84350 }, { "grad_norm": 0.33178573846817017, "learning_rate": 6.54059805427138e-06, "loss": 0.0204, "step": 84360 }, { "grad_norm": 0.28129860758781433, "learning_rate": 6.532424325061437e-06, "loss": 0.0236, "step": 84370 }, { "grad_norm": 0.28586673736572266, "learning_rate": 6.5242553494001935e-06, "loss": 0.0249, "step": 84380 }, { "grad_norm": 0.2324811965227127, "learning_rate": 6.516091128181046e-06, "loss": 0.0228, "step": 84390 }, { "grad_norm": 0.25393667817115784, "learning_rate": 6.507931662296785e-06, "loss": 0.0167, "step": 84400 }, { "grad_norm": 0.3620581328868866, "learning_rate": 6.4997769526397354e-06, "loss": 0.0212, "step": 84410 }, { "grad_norm": 0.43896299600601196, "learning_rate": 6.491627000101663e-06, "loss": 0.0233, "step": 84420 }, { "grad_norm": 0.3495572507381439, "learning_rate": 6.4834818055738545e-06, "loss": 0.0231, "step": 84430 }, { "grad_norm": 0.3441839814186096, "learning_rate": 6.4753413699470465e-06, "loss": 0.0178, "step": 84440 }, { "grad_norm": 0.35737916827201843, "learning_rate": 6.467205694111478e-06, "loss": 0.0243, "step": 84450 }, { "grad_norm": 0.1938323676586151, "learning_rate": 6.459074778956836e-06, "loss": 0.0273, "step": 84460 }, { "grad_norm": 0.21413591504096985, "learning_rate": 6.450948625372311e-06, "loss": 0.0224, "step": 84470 }, { "grad_norm": 0.4441412389278412, "learning_rate": 6.442827234246585e-06, "loss": 0.026, "step": 84480 }, { "grad_norm": 0.15799067914485931, "learning_rate": 6.43471060646777e-06, "loss": 0.0126, "step": 84490 }, { "grad_norm": 0.33413705229759216, "learning_rate": 6.426598742923512e-06, "loss": 0.0199, "step": 84500 }, { "grad_norm": 0.7534624934196472, "learning_rate": 6.418491644500885e-06, "loss": 0.0272, "step": 84510 }, { "grad_norm": 0.2825048863887787, "learning_rate": 6.410389312086512e-06, "loss": 0.0184, "step": 84520 }, { "grad_norm": 0.2708951234817505, "learning_rate": 6.4022917465664135e-06, "loss": 0.0222, "step": 84530 }, { "grad_norm": 0.5121678113937378, "learning_rate": 6.3941989488261575e-06, "loss": 0.0271, "step": 84540 }, { "grad_norm": 0.25893092155456543, "learning_rate": 6.386110919750726e-06, "loss": 0.0222, "step": 84550 }, { "grad_norm": 0.29751333594322205, "learning_rate": 6.378027660224651e-06, "loss": 0.0154, "step": 84560 }, { "grad_norm": 0.2134166955947876, "learning_rate": 6.369949171131856e-06, "loss": 0.0167, "step": 84570 }, { "grad_norm": 0.34109029173851013, "learning_rate": 6.361875453355848e-06, "loss": 0.0223, "step": 84580 }, { "grad_norm": 0.2729124426841736, "learning_rate": 6.353806507779525e-06, "loss": 0.0238, "step": 84590 }, { "grad_norm": 0.2877342700958252, "learning_rate": 6.3457423352853044e-06, "loss": 0.0226, "step": 84600 }, { "grad_norm": 0.514703094959259, "learning_rate": 6.337682936755062e-06, "loss": 0.0234, "step": 84610 }, { "grad_norm": 0.31199905276298523, "learning_rate": 6.329628313070179e-06, "loss": 0.0275, "step": 84620 }, { "grad_norm": 0.3968361020088196, "learning_rate": 6.321578465111477e-06, "loss": 0.0207, "step": 84630 }, { "grad_norm": 0.4564247727394104, "learning_rate": 6.31353339375928e-06, "loss": 0.0205, "step": 84640 }, { "grad_norm": 0.3288005292415619, "learning_rate": 6.305493099893395e-06, "loss": 0.0226, "step": 84650 }, { "grad_norm": 0.5741580128669739, "learning_rate": 6.297457584393096e-06, "loss": 0.0193, "step": 84660 }, { "grad_norm": 0.1613607257604599, "learning_rate": 6.289426848137126e-06, "loss": 0.0264, "step": 84670 }, { "grad_norm": 0.24135050177574158, "learning_rate": 6.281400892003713e-06, "loss": 0.0153, "step": 84680 }, { "grad_norm": 0.3882618844509125, "learning_rate": 6.273379716870581e-06, "loss": 0.0374, "step": 84690 }, { "grad_norm": 0.20688576996326447, "learning_rate": 6.265363323614887e-06, "loss": 0.0256, "step": 84700 }, { "grad_norm": 0.4627731144428253, "learning_rate": 6.257351713113308e-06, "loss": 0.0192, "step": 84710 }, { "grad_norm": 0.7975443601608276, "learning_rate": 6.249344886241981e-06, "loss": 0.0173, "step": 84720 }, { "grad_norm": 0.2803198993206024, "learning_rate": 6.2413428438765295e-06, "loss": 0.02, "step": 84730 }, { "grad_norm": 0.3572321832180023, "learning_rate": 6.233345586892026e-06, "loss": 0.0257, "step": 84740 }, { "grad_norm": 0.23353272676467896, "learning_rate": 6.225353116163052e-06, "loss": 0.0216, "step": 84750 }, { "grad_norm": 0.3001082241535187, "learning_rate": 6.217365432563627e-06, "loss": 0.0272, "step": 84760 }, { "grad_norm": 0.26745370030403137, "learning_rate": 6.209382536967318e-06, "loss": 0.0162, "step": 84770 }, { "grad_norm": 0.3524214029312134, "learning_rate": 6.201404430247082e-06, "loss": 0.0206, "step": 84780 }, { "grad_norm": 1.3386675119400024, "learning_rate": 6.1934311132754165e-06, "loss": 0.0238, "step": 84790 }, { "grad_norm": 0.22964325547218323, "learning_rate": 6.185462586924257e-06, "loss": 0.0243, "step": 84800 }, { "grad_norm": 0.28986215591430664, "learning_rate": 6.177498852065039e-06, "loss": 0.0235, "step": 84810 }, { "grad_norm": 0.3929360508918762, "learning_rate": 6.169539909568655e-06, "loss": 0.0218, "step": 84820 }, { "grad_norm": 0.3479476869106293, "learning_rate": 6.161585760305488e-06, "loss": 0.0176, "step": 84830 }, { "grad_norm": 0.2797483205795288, "learning_rate": 6.153636405145391e-06, "loss": 0.0224, "step": 84840 }, { "grad_norm": 0.28052252531051636, "learning_rate": 6.145691844957702e-06, "loss": 0.0254, "step": 84850 }, { "grad_norm": 0.3329099118709564, "learning_rate": 6.1377520806112145e-06, "loss": 0.0244, "step": 84860 }, { "grad_norm": 0.5500240325927734, "learning_rate": 6.129817112974218e-06, "loss": 0.023, "step": 84870 }, { "grad_norm": 0.13961999118328094, "learning_rate": 6.121886942914457e-06, "loss": 0.0183, "step": 84880 }, { "grad_norm": 0.28293150663375854, "learning_rate": 6.11396157129917e-06, "loss": 0.0222, "step": 84890 }, { "grad_norm": 0.3533814251422882, "learning_rate": 6.106040998995066e-06, "loss": 0.0205, "step": 84900 }, { "grad_norm": 0.3464071452617645, "learning_rate": 6.098125226868329e-06, "loss": 0.0209, "step": 84910 }, { "grad_norm": 0.5518773198127747, "learning_rate": 6.090214255784604e-06, "loss": 0.0291, "step": 84920 }, { "grad_norm": 0.22285772860050201, "learning_rate": 6.0823080866090346e-06, "loss": 0.0249, "step": 84930 }, { "grad_norm": 0.29450270533561707, "learning_rate": 6.074406720206233e-06, "loss": 0.0248, "step": 84940 }, { "grad_norm": 0.6134210824966431, "learning_rate": 6.066510157440259e-06, "loss": 0.0209, "step": 84950 }, { "grad_norm": 0.2752879559993744, "learning_rate": 6.058618399174676e-06, "loss": 0.0249, "step": 84960 }, { "grad_norm": 0.30776646733283997, "learning_rate": 6.050731446272528e-06, "loss": 0.0281, "step": 84970 }, { "grad_norm": 0.385388046503067, "learning_rate": 6.042849299596321e-06, "loss": 0.0208, "step": 84980 }, { "grad_norm": 0.3618007004261017, "learning_rate": 6.034971960008013e-06, "loss": 0.0261, "step": 84990 }, { "grad_norm": 0.4569321572780609, "learning_rate": 6.027099428369082e-06, "loss": 0.026, "step": 85000 }, { "grad_norm": 0.40133363008499146, "learning_rate": 6.019231705540435e-06, "loss": 0.0198, "step": 85010 }, { "grad_norm": 0.4756852686405182, "learning_rate": 6.011368792382488e-06, "loss": 0.0264, "step": 85020 }, { "grad_norm": 0.4487701952457428, "learning_rate": 6.00351068975511e-06, "loss": 0.0316, "step": 85030 }, { "grad_norm": 0.33796200156211853, "learning_rate": 5.9956573985176674e-06, "loss": 0.0199, "step": 85040 }, { "grad_norm": 0.36741378903388977, "learning_rate": 5.987808919528959e-06, "loss": 0.0174, "step": 85050 }, { "grad_norm": 0.4296872317790985, "learning_rate": 5.979965253647307e-06, "loss": 0.0343, "step": 85060 }, { "grad_norm": 0.32021060585975647, "learning_rate": 5.9721264017304655e-06, "loss": 0.0222, "step": 85070 }, { "grad_norm": 0.2880363166332245, "learning_rate": 5.964292364635682e-06, "loss": 0.0246, "step": 85080 }, { "grad_norm": 0.22752849757671356, "learning_rate": 5.956463143219682e-06, "loss": 0.0258, "step": 85090 }, { "grad_norm": 0.3480771780014038, "learning_rate": 5.948638738338663e-06, "loss": 0.0273, "step": 85100 }, { "grad_norm": 0.27784639596939087, "learning_rate": 5.940819150848271e-06, "loss": 0.025, "step": 85110 }, { "grad_norm": 0.3470723330974579, "learning_rate": 5.933004381603663e-06, "loss": 0.0192, "step": 85120 }, { "grad_norm": 0.32808181643486023, "learning_rate": 5.92519443145943e-06, "loss": 0.0182, "step": 85130 }, { "grad_norm": 0.34930989146232605, "learning_rate": 5.917389301269671e-06, "loss": 0.0242, "step": 85140 }, { "grad_norm": 0.9054216742515564, "learning_rate": 5.9095889918879435e-06, "loss": 0.0255, "step": 85150 }, { "grad_norm": 0.41251492500305176, "learning_rate": 5.901793504167286e-06, "loss": 0.0285, "step": 85160 }, { "grad_norm": 0.32659244537353516, "learning_rate": 5.894002838960172e-06, "loss": 0.023, "step": 85170 }, { "grad_norm": 0.45287445187568665, "learning_rate": 5.886216997118604e-06, "loss": 0.0242, "step": 85180 }, { "grad_norm": 0.3258286118507385, "learning_rate": 5.878435979494023e-06, "loss": 0.021, "step": 85190 }, { "grad_norm": 0.3030046820640564, "learning_rate": 5.870659786937344e-06, "loss": 0.0204, "step": 85200 }, { "grad_norm": 0.3742607831954956, "learning_rate": 5.862888420298962e-06, "loss": 0.0247, "step": 85210 }, { "grad_norm": 0.42701396346092224, "learning_rate": 5.855121880428743e-06, "loss": 0.0139, "step": 85220 }, { "grad_norm": 0.2913735806941986, "learning_rate": 5.84736016817603e-06, "loss": 0.0236, "step": 85230 }, { "grad_norm": 0.3586409091949463, "learning_rate": 5.839603284389622e-06, "loss": 0.0276, "step": 85240 }, { "grad_norm": 0.34112685918807983, "learning_rate": 5.831851229917812e-06, "loss": 0.0227, "step": 85250 }, { "grad_norm": 0.377866268157959, "learning_rate": 5.824104005608338e-06, "loss": 0.0277, "step": 85260 }, { "grad_norm": 0.144601508975029, "learning_rate": 5.81636161230843e-06, "loss": 0.0241, "step": 85270 }, { "grad_norm": 0.4956502318382263, "learning_rate": 5.808624050864791e-06, "loss": 0.0297, "step": 85280 }, { "grad_norm": 0.5309878587722778, "learning_rate": 5.800891322123592e-06, "loss": 0.0259, "step": 85290 }, { "grad_norm": 0.21643657982349396, "learning_rate": 5.793163426930454e-06, "loss": 0.0179, "step": 85300 }, { "grad_norm": 0.23862433433532715, "learning_rate": 5.785440366130512e-06, "loss": 0.0225, "step": 85310 }, { "grad_norm": 0.301753968000412, "learning_rate": 5.777722140568326e-06, "loss": 0.0219, "step": 85320 }, { "grad_norm": 0.7906678318977356, "learning_rate": 5.7700087510879685e-06, "loss": 0.0262, "step": 85330 }, { "grad_norm": 0.4173702001571655, "learning_rate": 5.762300198532933e-06, "loss": 0.0204, "step": 85340 }, { "grad_norm": 0.2777049243450165, "learning_rate": 5.7545964837462606e-06, "loss": 0.0174, "step": 85350 }, { "grad_norm": 1.0539956092834473, "learning_rate": 5.746897607570384e-06, "loss": 0.0283, "step": 85360 }, { "grad_norm": 0.34293133020401, "learning_rate": 5.7392035708472626e-06, "loss": 0.0163, "step": 85370 }, { "grad_norm": 0.23992539942264557, "learning_rate": 5.731514374418278e-06, "loss": 0.0215, "step": 85380 }, { "grad_norm": 0.6478356719017029, "learning_rate": 5.7238300191243325e-06, "loss": 0.0228, "step": 85390 }, { "grad_norm": 0.34732890129089355, "learning_rate": 5.716150505805762e-06, "loss": 0.02, "step": 85400 }, { "grad_norm": 0.40814170241355896, "learning_rate": 5.708475835302407e-06, "loss": 0.0236, "step": 85410 }, { "grad_norm": 0.43771931529045105, "learning_rate": 5.70080600845353e-06, "loss": 0.0285, "step": 85420 }, { "grad_norm": 0.5295434594154358, "learning_rate": 5.693141026097909e-06, "loss": 0.0242, "step": 85430 }, { "grad_norm": 0.2944304347038269, "learning_rate": 5.6854808890737745e-06, "loss": 0.02, "step": 85440 }, { "grad_norm": 0.5316982269287109, "learning_rate": 5.6778255982188225e-06, "loss": 0.0173, "step": 85450 }, { "grad_norm": 0.3193190097808838, "learning_rate": 5.670175154370217e-06, "loss": 0.0225, "step": 85460 }, { "grad_norm": 0.3291122317314148, "learning_rate": 5.662529558364616e-06, "loss": 0.0225, "step": 85470 }, { "grad_norm": 0.1891283392906189, "learning_rate": 5.654888811038128e-06, "loss": 0.022, "step": 85480 }, { "grad_norm": 1.865163803100586, "learning_rate": 5.647252913226314e-06, "loss": 0.0231, "step": 85490 }, { "grad_norm": 0.34387651085853577, "learning_rate": 5.639621865764255e-06, "loss": 0.0275, "step": 85500 }, { "grad_norm": 0.3888598084449768, "learning_rate": 5.63199566948644e-06, "loss": 0.0287, "step": 85510 }, { "grad_norm": 0.49247613549232483, "learning_rate": 5.6243743252268835e-06, "loss": 0.0243, "step": 85520 }, { "grad_norm": 0.17847026884555817, "learning_rate": 5.616757833819009e-06, "loss": 0.0207, "step": 85530 }, { "grad_norm": 0.22441567480564117, "learning_rate": 5.609146196095794e-06, "loss": 0.0257, "step": 85540 }, { "grad_norm": 0.3336436450481415, "learning_rate": 5.6015394128896005e-06, "loss": 0.02, "step": 85550 }, { "grad_norm": 0.41732415556907654, "learning_rate": 5.593937485032308e-06, "loss": 0.023, "step": 85560 }, { "grad_norm": 0.3366491198539734, "learning_rate": 5.586340413355246e-06, "loss": 0.0267, "step": 85570 }, { "grad_norm": 0.2582058310508728, "learning_rate": 5.578748198689227e-06, "loss": 0.0259, "step": 85580 }, { "grad_norm": 0.27751973271369934, "learning_rate": 5.571160841864504e-06, "loss": 0.0173, "step": 85590 }, { "grad_norm": 0.3152622580528259, "learning_rate": 5.563578343710846e-06, "loss": 0.0227, "step": 85600 }, { "grad_norm": 0.34598997235298157, "learning_rate": 5.556000705057446e-06, "loss": 0.0171, "step": 85610 }, { "grad_norm": 0.368020623922348, "learning_rate": 5.548427926732996e-06, "loss": 0.0228, "step": 85620 }, { "grad_norm": 0.2441444844007492, "learning_rate": 5.540860009565629e-06, "loss": 0.0261, "step": 85630 }, { "grad_norm": 0.3398786187171936, "learning_rate": 5.533296954382966e-06, "loss": 0.0262, "step": 85640 }, { "grad_norm": 0.4055977463722229, "learning_rate": 5.525738762012095e-06, "loss": 0.0279, "step": 85650 }, { "grad_norm": 0.6720249056816101, "learning_rate": 5.518185433279577e-06, "loss": 0.0261, "step": 85660 }, { "grad_norm": 0.3427172601222992, "learning_rate": 5.510636969011407e-06, "loss": 0.0217, "step": 85670 }, { "grad_norm": 0.38526010513305664, "learning_rate": 5.503093370033096e-06, "loss": 0.0251, "step": 85680 }, { "grad_norm": 0.2755570113658905, "learning_rate": 5.495554637169603e-06, "loss": 0.0222, "step": 85690 }, { "grad_norm": 1.1580662727355957, "learning_rate": 5.488020771245328e-06, "loss": 0.0234, "step": 85700 }, { "grad_norm": 0.2520420253276825, "learning_rate": 5.480491773084195e-06, "loss": 0.0181, "step": 85710 }, { "grad_norm": 0.34678900241851807, "learning_rate": 5.472967643509525e-06, "loss": 0.0201, "step": 85720 }, { "grad_norm": 0.31972891092300415, "learning_rate": 5.46544838334419e-06, "loss": 0.0239, "step": 85730 }, { "grad_norm": 0.19899901747703552, "learning_rate": 5.4579339934104525e-06, "loss": 0.0375, "step": 85740 }, { "grad_norm": 0.29147034883499146, "learning_rate": 5.450424474530091e-06, "loss": 0.03, "step": 85750 }, { "grad_norm": 0.4025866985321045, "learning_rate": 5.442919827524323e-06, "loss": 0.0207, "step": 85760 }, { "grad_norm": 0.2667542099952698, "learning_rate": 5.435420053213863e-06, "loss": 0.0281, "step": 85770 }, { "grad_norm": 0.4203556776046753, "learning_rate": 5.427925152418845e-06, "loss": 0.0321, "step": 85780 }, { "grad_norm": 0.3764609396457672, "learning_rate": 5.420435125958934e-06, "loss": 0.023, "step": 85790 }, { "grad_norm": 0.27682727575302124, "learning_rate": 5.412949974653209e-06, "loss": 0.0298, "step": 85800 }, { "grad_norm": 0.5198631286621094, "learning_rate": 5.405469699320248e-06, "loss": 0.0241, "step": 85810 }, { "grad_norm": 0.2287948578596115, "learning_rate": 5.397994300778064e-06, "loss": 0.0235, "step": 85820 }, { "grad_norm": 0.31131452322006226, "learning_rate": 5.39052377984417e-06, "loss": 0.0261, "step": 85830 }, { "grad_norm": 0.3251105546951294, "learning_rate": 5.383058137335523e-06, "loss": 0.0234, "step": 85840 }, { "grad_norm": 0.3844369947910309, "learning_rate": 5.375597374068553e-06, "loss": 0.0252, "step": 85850 }, { "grad_norm": 0.3127816617488861, "learning_rate": 5.368141490859163e-06, "loss": 0.0218, "step": 85860 }, { "grad_norm": 1.001544713973999, "learning_rate": 5.360690488522729e-06, "loss": 0.0245, "step": 85870 }, { "grad_norm": 0.44381916522979736, "learning_rate": 5.353244367874055e-06, "loss": 0.0193, "step": 85880 }, { "grad_norm": 0.24954485893249512, "learning_rate": 5.345803129727456e-06, "loss": 0.0174, "step": 85890 }, { "grad_norm": 0.20002639293670654, "learning_rate": 5.338366774896697e-06, "loss": 0.0199, "step": 85900 }, { "grad_norm": 0.685148298740387, "learning_rate": 5.330935304194995e-06, "loss": 0.0287, "step": 85910 }, { "grad_norm": 0.390720933675766, "learning_rate": 5.323508718435044e-06, "loss": 0.0217, "step": 85920 }, { "grad_norm": 0.25110676884651184, "learning_rate": 5.316087018429011e-06, "loss": 0.0247, "step": 85930 }, { "grad_norm": 0.39133426547050476, "learning_rate": 5.30867020498853e-06, "loss": 0.0169, "step": 85940 }, { "grad_norm": 0.37572890520095825, "learning_rate": 5.30125827892467e-06, "loss": 0.0273, "step": 85950 }, { "grad_norm": 0.4473062753677368, "learning_rate": 5.293851241048015e-06, "loss": 0.0224, "step": 85960 }, { "grad_norm": 0.4812571108341217, "learning_rate": 5.286449092168555e-06, "loss": 0.0254, "step": 85970 }, { "grad_norm": 0.47740229964256287, "learning_rate": 5.2790518330958105e-06, "loss": 0.0244, "step": 85980 }, { "grad_norm": 0.2202879786491394, "learning_rate": 5.271659464638717e-06, "loss": 0.0222, "step": 85990 }, { "grad_norm": 0.211203932762146, "learning_rate": 5.264271987605701e-06, "loss": 0.0214, "step": 86000 }, { "grad_norm": 0.36977797746658325, "learning_rate": 5.256889402804632e-06, "loss": 0.0207, "step": 86010 }, { "grad_norm": 0.20626746118068695, "learning_rate": 5.249511711042881e-06, "loss": 0.0227, "step": 86020 }, { "grad_norm": 0.3349393904209137, "learning_rate": 5.242138913127237e-06, "loss": 0.0253, "step": 86030 }, { "grad_norm": 0.26270484924316406, "learning_rate": 5.234771009863987e-06, "loss": 0.0162, "step": 86040 }, { "grad_norm": 0.37693753838539124, "learning_rate": 5.227408002058881e-06, "loss": 0.0275, "step": 86050 }, { "grad_norm": 0.23346470296382904, "learning_rate": 5.220049890517126e-06, "loss": 0.0215, "step": 86060 }, { "grad_norm": 0.22030970454216003, "learning_rate": 5.212696676043377e-06, "loss": 0.0254, "step": 86070 }, { "grad_norm": 0.34144532680511475, "learning_rate": 5.205348359441798e-06, "loss": 0.0202, "step": 86080 }, { "grad_norm": 0.4289323687553406, "learning_rate": 5.198004941515966e-06, "loss": 0.0269, "step": 86090 }, { "grad_norm": 0.3629968762397766, "learning_rate": 5.190666423068952e-06, "loss": 0.0184, "step": 86100 }, { "grad_norm": 0.2835788428783417, "learning_rate": 5.183332804903284e-06, "loss": 0.0182, "step": 86110 }, { "grad_norm": 0.18722106516361237, "learning_rate": 5.176004087820979e-06, "loss": 0.0209, "step": 86120 }, { "grad_norm": 0.4089144766330719, "learning_rate": 5.168680272623455e-06, "loss": 0.0241, "step": 86130 }, { "grad_norm": 0.2223266065120697, "learning_rate": 5.161361360111661e-06, "loss": 0.023, "step": 86140 }, { "grad_norm": 0.39041754603385925, "learning_rate": 5.154047351085984e-06, "loss": 0.0198, "step": 86150 }, { "grad_norm": 0.25277942419052124, "learning_rate": 5.1467382463462385e-06, "loss": 0.0177, "step": 86160 }, { "grad_norm": 0.28594452142715454, "learning_rate": 5.139434046691788e-06, "loss": 0.0199, "step": 86170 }, { "grad_norm": 0.1844826489686966, "learning_rate": 5.13213475292137e-06, "loss": 0.0216, "step": 86180 }, { "grad_norm": 0.519698977470398, "learning_rate": 5.12484036583325e-06, "loss": 0.0276, "step": 86190 }, { "grad_norm": 0.37862786650657654, "learning_rate": 5.117550886225109e-06, "loss": 0.025, "step": 86200 }, { "grad_norm": 0.2069387286901474, "learning_rate": 5.11026631489413e-06, "loss": 0.0191, "step": 86210 }, { "grad_norm": 0.29515230655670166, "learning_rate": 5.102986652636926e-06, "loss": 0.0316, "step": 86220 }, { "grad_norm": 0.3192315101623535, "learning_rate": 5.095711900249606e-06, "loss": 0.0188, "step": 86230 }, { "grad_norm": 0.19680155813694, "learning_rate": 5.088442058527715e-06, "loss": 0.0206, "step": 86240 }, { "grad_norm": 0.26828381419181824, "learning_rate": 5.08117712826629e-06, "loss": 0.0302, "step": 86250 }, { "grad_norm": 0.3867356777191162, "learning_rate": 5.073917110259785e-06, "loss": 0.0309, "step": 86260 }, { "grad_norm": 0.44783473014831543, "learning_rate": 5.066662005302175e-06, "loss": 0.0196, "step": 86270 }, { "grad_norm": 4.460592746734619, "learning_rate": 5.059411814186838e-06, "loss": 0.0212, "step": 86280 }, { "grad_norm": 0.49615153670310974, "learning_rate": 5.05216653770666e-06, "loss": 0.0384, "step": 86290 }, { "grad_norm": 0.42762377858161926, "learning_rate": 5.044926176653969e-06, "loss": 0.0149, "step": 86300 }, { "grad_norm": 0.23154008388519287, "learning_rate": 5.037690731820571e-06, "loss": 0.0194, "step": 86310 }, { "grad_norm": 1.3232030868530273, "learning_rate": 5.030460203997711e-06, "loss": 0.017, "step": 86320 }, { "grad_norm": 0.24350687861442566, "learning_rate": 5.0232345939761174e-06, "loss": 0.0207, "step": 86330 }, { "grad_norm": 0.28311464190483093, "learning_rate": 5.016013902545957e-06, "loss": 0.0227, "step": 86340 }, { "grad_norm": 0.3394676744937897, "learning_rate": 5.008798130496889e-06, "loss": 0.0245, "step": 86350 }, { "grad_norm": 0.22918450832366943, "learning_rate": 5.001587278618008e-06, "loss": 0.018, "step": 86360 }, { "grad_norm": 0.695021390914917, "learning_rate": 4.994381347697891e-06, "loss": 0.0247, "step": 86370 }, { "grad_norm": 0.5226091146469116, "learning_rate": 4.987180338524572e-06, "loss": 0.0241, "step": 86380 }, { "grad_norm": 0.23878976702690125, "learning_rate": 4.9799842518855275e-06, "loss": 0.0172, "step": 86390 }, { "grad_norm": 1.1747862100601196, "learning_rate": 4.9727930885677285e-06, "loss": 0.0237, "step": 86400 }, { "grad_norm": 0.34906214475631714, "learning_rate": 4.965606849357573e-06, "loss": 0.0274, "step": 86410 }, { "grad_norm": 0.8509446978569031, "learning_rate": 4.9584255350409445e-06, "loss": 0.0192, "step": 86420 }, { "grad_norm": 0.2906666696071625, "learning_rate": 4.951249146403175e-06, "loss": 0.0218, "step": 86430 }, { "grad_norm": 0.4053439497947693, "learning_rate": 4.9440776842290815e-06, "loss": 0.0222, "step": 86440 }, { "grad_norm": 0.4665057361125946, "learning_rate": 4.936911149302903e-06, "loss": 0.0239, "step": 86450 }, { "grad_norm": 0.18675827980041504, "learning_rate": 4.929749542408379e-06, "loss": 0.0209, "step": 86460 }, { "grad_norm": 0.2874438762664795, "learning_rate": 4.922592864328673e-06, "loss": 0.0185, "step": 86470 }, { "grad_norm": 0.3379380702972412, "learning_rate": 4.915441115846436e-06, "loss": 0.0309, "step": 86480 }, { "grad_norm": 0.31617385149002075, "learning_rate": 4.908294297743782e-06, "loss": 0.028, "step": 86490 }, { "grad_norm": 0.2367715984582901, "learning_rate": 4.901152410802273e-06, "loss": 0.0201, "step": 86500 }, { "grad_norm": 0.37877586483955383, "learning_rate": 4.894015455802919e-06, "loss": 0.0176, "step": 86510 }, { "grad_norm": 0.3862878680229187, "learning_rate": 4.8868834335262346e-06, "loss": 0.0349, "step": 86520 }, { "grad_norm": 0.2691354751586914, "learning_rate": 4.87975634475214e-06, "loss": 0.0191, "step": 86530 }, { "grad_norm": 0.2594585716724396, "learning_rate": 4.8726341902600614e-06, "loss": 0.0228, "step": 86540 }, { "grad_norm": 0.22755314409732819, "learning_rate": 4.865516970828843e-06, "loss": 0.0217, "step": 86550 }, { "grad_norm": 0.41121459007263184, "learning_rate": 4.85840468723685e-06, "loss": 0.0255, "step": 86560 }, { "grad_norm": 0.23337918519973755, "learning_rate": 4.85129734026184e-06, "loss": 0.0165, "step": 86570 }, { "grad_norm": 0.3869476318359375, "learning_rate": 4.844194930681073e-06, "loss": 0.025, "step": 86580 }, { "grad_norm": 0.33243003487586975, "learning_rate": 4.837097459271267e-06, "loss": 0.0237, "step": 86590 }, { "grad_norm": 0.22369879484176636, "learning_rate": 4.830004926808573e-06, "loss": 0.0181, "step": 86600 }, { "grad_norm": 0.3307715952396393, "learning_rate": 4.822917334068627e-06, "loss": 0.021, "step": 86610 }, { "grad_norm": 0.4332295358181, "learning_rate": 4.81583468182652e-06, "loss": 0.0309, "step": 86620 }, { "grad_norm": 0.3584349453449249, "learning_rate": 4.808756970856809e-06, "loss": 0.0181, "step": 86630 }, { "grad_norm": 0.21790991723537445, "learning_rate": 4.801684201933476e-06, "loss": 0.0189, "step": 86640 }, { "grad_norm": 0.2581321597099304, "learning_rate": 4.794616375830019e-06, "loss": 0.0248, "step": 86650 }, { "grad_norm": 0.35066425800323486, "learning_rate": 4.787553493319341e-06, "loss": 0.0165, "step": 86660 }, { "grad_norm": 0.18039391934871674, "learning_rate": 4.780495555173836e-06, "loss": 0.0191, "step": 86670 }, { "grad_norm": 0.504952073097229, "learning_rate": 4.7734425621653475e-06, "loss": 0.0186, "step": 86680 }, { "grad_norm": 0.21638529002666473, "learning_rate": 4.766394515065198e-06, "loss": 0.0165, "step": 86690 }, { "grad_norm": 0.2897099256515503, "learning_rate": 4.75935141464412e-06, "loss": 0.0177, "step": 86700 }, { "grad_norm": 0.5637381672859192, "learning_rate": 4.75231326167237e-06, "loss": 0.027, "step": 86710 }, { "grad_norm": 0.2872350811958313, "learning_rate": 4.745280056919599e-06, "loss": 0.0193, "step": 86720 }, { "grad_norm": 0.3077760934829712, "learning_rate": 4.738251801154975e-06, "loss": 0.0224, "step": 86730 }, { "grad_norm": 0.5441223382949829, "learning_rate": 4.731228495147061e-06, "loss": 0.0234, "step": 86740 }, { "grad_norm": 0.43595245480537415, "learning_rate": 4.7242101396639595e-06, "loss": 0.0272, "step": 86750 }, { "grad_norm": 0.21119017899036407, "learning_rate": 4.717196735473151e-06, "loss": 0.0192, "step": 86760 }, { "grad_norm": 0.4817703366279602, "learning_rate": 4.710188283341643e-06, "loss": 0.0191, "step": 86770 }, { "grad_norm": 0.4387674629688263, "learning_rate": 4.703184784035835e-06, "loss": 0.0319, "step": 86780 }, { "grad_norm": 0.7945395112037659, "learning_rate": 4.696186238321642e-06, "loss": 0.0273, "step": 86790 }, { "grad_norm": 0.47462549805641174, "learning_rate": 4.689192646964408e-06, "loss": 0.0212, "step": 86800 }, { "grad_norm": 0.42080360651016235, "learning_rate": 4.682204010728952e-06, "loss": 0.0231, "step": 86810 }, { "grad_norm": 0.30458229780197144, "learning_rate": 4.67522033037952e-06, "loss": 0.0206, "step": 86820 }, { "grad_norm": 0.7908281087875366, "learning_rate": 4.668241606679846e-06, "loss": 0.0237, "step": 86830 }, { "grad_norm": 0.3034822940826416, "learning_rate": 4.66126784039313e-06, "loss": 0.0241, "step": 86840 }, { "grad_norm": 0.2988588511943817, "learning_rate": 4.6542990322819836e-06, "loss": 0.0282, "step": 86850 }, { "grad_norm": 0.25105956196784973, "learning_rate": 4.64733518310852e-06, "loss": 0.021, "step": 86860 }, { "grad_norm": 0.2800784111022949, "learning_rate": 4.640376293634291e-06, "loss": 0.017, "step": 86870 }, { "grad_norm": 0.30963122844696045, "learning_rate": 4.633422364620327e-06, "loss": 0.0239, "step": 86880 }, { "grad_norm": 0.41084766387939453, "learning_rate": 4.6264733968270745e-06, "loss": 0.0171, "step": 86890 }, { "grad_norm": 0.8887107968330383, "learning_rate": 4.619529391014482e-06, "loss": 0.0294, "step": 86900 }, { "grad_norm": 0.3242364823818207, "learning_rate": 4.61259034794192e-06, "loss": 0.0209, "step": 86910 }, { "grad_norm": 0.4209008812904358, "learning_rate": 4.605656268368247e-06, "loss": 0.017, "step": 86920 }, { "grad_norm": 0.3886342942714691, "learning_rate": 4.598727153051735e-06, "loss": 0.0218, "step": 86930 }, { "grad_norm": 0.33900365233421326, "learning_rate": 4.591803002750178e-06, "loss": 0.0188, "step": 86940 }, { "grad_norm": 0.18682809174060822, "learning_rate": 4.58488381822077e-06, "loss": 0.0174, "step": 86950 }, { "grad_norm": 0.23178409039974213, "learning_rate": 4.5779696002201946e-06, "loss": 0.0199, "step": 86960 }, { "grad_norm": 0.19798994064331055, "learning_rate": 4.571060349504563e-06, "loss": 0.0204, "step": 86970 }, { "grad_norm": 0.22365455329418182, "learning_rate": 4.564156066829478e-06, "loss": 0.0246, "step": 86980 }, { "grad_norm": 0.3917221426963806, "learning_rate": 4.557256752949957e-06, "loss": 0.0259, "step": 86990 }, { "grad_norm": 0.327108770608902, "learning_rate": 4.550362408620534e-06, "loss": 0.0189, "step": 87000 }, { "grad_norm": 0.12741994857788086, "learning_rate": 4.543473034595136e-06, "loss": 0.0219, "step": 87010 }, { "grad_norm": 0.27445611357688904, "learning_rate": 4.536588631627192e-06, "loss": 0.0194, "step": 87020 }, { "grad_norm": 0.3510470986366272, "learning_rate": 4.5297092004695565e-06, "loss": 0.0254, "step": 87030 }, { "grad_norm": 0.32374101877212524, "learning_rate": 4.522834741874555e-06, "loss": 0.0178, "step": 87040 }, { "grad_norm": 0.3095605671405792, "learning_rate": 4.515965256593974e-06, "loss": 0.021, "step": 87050 }, { "grad_norm": 0.639114260673523, "learning_rate": 4.509100745379058e-06, "loss": 0.023, "step": 87060 }, { "grad_norm": 0.32796013355255127, "learning_rate": 4.502241208980479e-06, "loss": 0.0244, "step": 87070 }, { "grad_norm": 0.25064948201179504, "learning_rate": 4.495386648148398e-06, "loss": 0.017, "step": 87080 }, { "grad_norm": 0.23435179889202118, "learning_rate": 4.488537063632431e-06, "loss": 0.0148, "step": 87090 }, { "grad_norm": 0.3286714553833008, "learning_rate": 4.481692456181608e-06, "loss": 0.0239, "step": 87100 }, { "grad_norm": 0.3753833770751953, "learning_rate": 4.474852826544479e-06, "loss": 0.0207, "step": 87110 }, { "grad_norm": 0.39271923899650574, "learning_rate": 4.468018175468974e-06, "loss": 0.0302, "step": 87120 }, { "grad_norm": 0.2085064947605133, "learning_rate": 4.461188503702568e-06, "loss": 0.0219, "step": 87130 }, { "grad_norm": 0.2826555073261261, "learning_rate": 4.454363811992107e-06, "loss": 0.0214, "step": 87140 }, { "grad_norm": 0.2866348624229431, "learning_rate": 4.447544101083956e-06, "loss": 0.0233, "step": 87150 }, { "grad_norm": 0.3657233417034149, "learning_rate": 4.440729371723884e-06, "loss": 0.0226, "step": 87160 }, { "grad_norm": 0.23559316992759705, "learning_rate": 4.433919624657162e-06, "loss": 0.0215, "step": 87170 }, { "grad_norm": 0.4731171131134033, "learning_rate": 4.427114860628462e-06, "loss": 0.0212, "step": 87180 }, { "grad_norm": 0.31288576126098633, "learning_rate": 4.420315080381982e-06, "loss": 0.0263, "step": 87190 }, { "grad_norm": 0.2859545946121216, "learning_rate": 4.4135202846613045e-06, "loss": 0.0337, "step": 87200 }, { "grad_norm": 0.4559788107872009, "learning_rate": 4.406730474209525e-06, "loss": 0.0207, "step": 87210 }, { "grad_norm": 0.5305960774421692, "learning_rate": 4.399945649769138e-06, "loss": 0.0297, "step": 87220 }, { "grad_norm": 0.7294443249702454, "learning_rate": 4.393165812082151e-06, "loss": 0.0277, "step": 87230 }, { "grad_norm": 0.21343372762203217, "learning_rate": 4.386390961889969e-06, "loss": 0.0247, "step": 87240 }, { "grad_norm": 0.2573778033256531, "learning_rate": 4.3796210999334895e-06, "loss": 0.021, "step": 87250 }, { "grad_norm": 0.1988816112279892, "learning_rate": 4.372856226953065e-06, "loss": 0.0161, "step": 87260 }, { "grad_norm": 0.2635120749473572, "learning_rate": 4.366096343688486e-06, "loss": 0.0154, "step": 87270 }, { "grad_norm": 0.30370938777923584, "learning_rate": 4.359341450878995e-06, "loss": 0.0226, "step": 87280 }, { "grad_norm": 0.5224593281745911, "learning_rate": 4.352591549263302e-06, "loss": 0.0207, "step": 87290 }, { "grad_norm": 0.5017786622047424, "learning_rate": 4.345846639579582e-06, "loss": 0.0257, "step": 87300 }, { "grad_norm": 0.9911690950393677, "learning_rate": 4.339106722565417e-06, "loss": 0.0173, "step": 87310 }, { "grad_norm": 0.5504854917526245, "learning_rate": 4.332371798957896e-06, "loss": 0.0207, "step": 87320 }, { "grad_norm": 0.21289311349391937, "learning_rate": 4.3256418694935295e-06, "loss": 0.0224, "step": 87330 }, { "grad_norm": 0.32565563917160034, "learning_rate": 4.3189169349083125e-06, "loss": 0.0221, "step": 87340 }, { "grad_norm": 0.27592235803604126, "learning_rate": 4.312196995937645e-06, "loss": 0.0241, "step": 87350 }, { "grad_norm": 0.35399922728538513, "learning_rate": 4.305482053316434e-06, "loss": 0.021, "step": 87360 }, { "grad_norm": 0.5843597054481506, "learning_rate": 4.2987721077789925e-06, "loss": 0.0232, "step": 87370 }, { "grad_norm": 0.34223949909210205, "learning_rate": 4.2920671600591325e-06, "loss": 0.0164, "step": 87380 }, { "grad_norm": 0.2915235757827759, "learning_rate": 4.285367210890084e-06, "loss": 0.0215, "step": 87390 }, { "grad_norm": 0.31526920199394226, "learning_rate": 4.278672261004551e-06, "loss": 0.0213, "step": 87400 }, { "grad_norm": 0.22383105754852295, "learning_rate": 4.2719823111346704e-06, "loss": 0.0158, "step": 87410 }, { "grad_norm": 0.2751379907131195, "learning_rate": 4.265297362012061e-06, "loss": 0.0163, "step": 87420 }, { "grad_norm": 0.35722002387046814, "learning_rate": 4.258617414367766e-06, "loss": 0.0227, "step": 87430 }, { "grad_norm": 1.3987956047058105, "learning_rate": 4.251942468932296e-06, "loss": 0.0268, "step": 87440 }, { "grad_norm": 0.2688013017177582, "learning_rate": 4.2452725264356175e-06, "loss": 0.0245, "step": 87450 }, { "grad_norm": 0.18832430243492126, "learning_rate": 4.2386075876071575e-06, "loss": 0.0218, "step": 87460 }, { "grad_norm": 0.3593518137931824, "learning_rate": 4.231947653175755e-06, "loss": 0.0179, "step": 87470 }, { "grad_norm": 0.3432241976261139, "learning_rate": 4.225292723869761e-06, "loss": 0.0226, "step": 87480 }, { "grad_norm": 0.22368952631950378, "learning_rate": 4.218642800416916e-06, "loss": 0.0205, "step": 87490 }, { "grad_norm": 0.43925416469573975, "learning_rate": 4.211997883544472e-06, "loss": 0.0234, "step": 87500 }, { "grad_norm": 0.32687875628471375, "learning_rate": 4.205357973979096e-06, "loss": 0.0203, "step": 87510 }, { "grad_norm": 0.33522769808769226, "learning_rate": 4.198723072446925e-06, "loss": 0.0222, "step": 87520 }, { "grad_norm": 0.16202975809574127, "learning_rate": 4.192093179673534e-06, "loss": 0.0213, "step": 87530 }, { "grad_norm": 0.29214411973953247, "learning_rate": 4.185468296383959e-06, "loss": 0.0291, "step": 87540 }, { "grad_norm": 0.5310460925102234, "learning_rate": 4.1788484233027005e-06, "loss": 0.027, "step": 87550 }, { "grad_norm": 0.45286184549331665, "learning_rate": 4.172233561153677e-06, "loss": 0.0253, "step": 87560 }, { "grad_norm": 0.24279101192951202, "learning_rate": 4.165623710660288e-06, "loss": 0.0185, "step": 87570 }, { "grad_norm": 0.1814364790916443, "learning_rate": 4.1590188725453835e-06, "loss": 0.0111, "step": 87580 }, { "grad_norm": 0.3159778118133545, "learning_rate": 4.152419047531259e-06, "loss": 0.0238, "step": 87590 }, { "grad_norm": 0.2868102192878723, "learning_rate": 4.145824236339651e-06, "loss": 0.0188, "step": 87600 }, { "grad_norm": 0.19551996886730194, "learning_rate": 4.1392344396917695e-06, "loss": 0.0276, "step": 87610 }, { "grad_norm": 0.4193054437637329, "learning_rate": 4.132649658308252e-06, "loss": 0.029, "step": 87620 }, { "grad_norm": 0.3079632520675659, "learning_rate": 4.126069892909213e-06, "loss": 0.0241, "step": 87630 }, { "grad_norm": 0.6529732346534729, "learning_rate": 4.119495144214191e-06, "loss": 0.0252, "step": 87640 }, { "grad_norm": 0.23616160452365875, "learning_rate": 4.112925412942214e-06, "loss": 0.0198, "step": 87650 }, { "grad_norm": 0.28696590662002563, "learning_rate": 4.106360699811712e-06, "loss": 0.0213, "step": 87660 }, { "grad_norm": 0.30381646752357483, "learning_rate": 4.099801005540616e-06, "loss": 0.0186, "step": 87670 }, { "grad_norm": 0.3274804651737213, "learning_rate": 4.093246330846256e-06, "loss": 0.0241, "step": 87680 }, { "grad_norm": 0.2802445888519287, "learning_rate": 4.086696676445467e-06, "loss": 0.0233, "step": 87690 }, { "grad_norm": 0.3062910735607147, "learning_rate": 4.0801520430544914e-06, "loss": 0.0211, "step": 87700 }, { "grad_norm": 0.39439868927001953, "learning_rate": 4.073612431389062e-06, "loss": 0.0277, "step": 87710 }, { "grad_norm": 0.5990837812423706, "learning_rate": 4.067077842164319e-06, "loss": 0.0198, "step": 87720 }, { "grad_norm": 0.28333133459091187, "learning_rate": 4.0605482760948945e-06, "loss": 0.0311, "step": 87730 }, { "grad_norm": 0.23702238500118256, "learning_rate": 4.054023733894829e-06, "loss": 0.0209, "step": 87740 }, { "grad_norm": 0.26778078079223633, "learning_rate": 4.047504216277653e-06, "loss": 0.0226, "step": 87750 }, { "grad_norm": 0.36992400884628296, "learning_rate": 4.040989723956323e-06, "loss": 0.0264, "step": 87760 }, { "grad_norm": 0.29872965812683105, "learning_rate": 4.03448025764327e-06, "loss": 0.024, "step": 87770 }, { "grad_norm": 0.40063464641571045, "learning_rate": 4.027975818050339e-06, "loss": 0.0224, "step": 87780 }, { "grad_norm": 0.3589308261871338, "learning_rate": 4.021476405888858e-06, "loss": 0.0238, "step": 87790 }, { "grad_norm": 0.30370625853538513, "learning_rate": 4.01498202186959e-06, "loss": 0.0224, "step": 87800 }, { "grad_norm": 0.37079885601997375, "learning_rate": 4.008492666702751e-06, "loss": 0.0305, "step": 87810 }, { "grad_norm": 0.3197768032550812, "learning_rate": 4.002008341098001e-06, "loss": 0.0214, "step": 87820 }, { "grad_norm": 1.0558876991271973, "learning_rate": 3.995529045764462e-06, "loss": 0.0227, "step": 87830 }, { "grad_norm": 0.2504558861255646, "learning_rate": 3.989054781410712e-06, "loss": 0.0224, "step": 87840 }, { "grad_norm": 0.26280808448791504, "learning_rate": 3.982585548744744e-06, "loss": 0.026, "step": 87850 }, { "grad_norm": 0.2616095244884491, "learning_rate": 3.9761213484740435e-06, "loss": 0.0212, "step": 87860 }, { "grad_norm": 0.2368241399526596, "learning_rate": 3.9696621813055055e-06, "loss": 0.023, "step": 87870 }, { "grad_norm": 0.43151625990867615, "learning_rate": 3.9632080479455044e-06, "loss": 0.0225, "step": 87880 }, { "grad_norm": 0.46252018213272095, "learning_rate": 3.956758949099859e-06, "loss": 0.0212, "step": 87890 }, { "grad_norm": 0.36274588108062744, "learning_rate": 3.9503148854738385e-06, "loss": 0.0182, "step": 87900 }, { "grad_norm": 0.4341520369052887, "learning_rate": 3.943875857772139e-06, "loss": 0.0349, "step": 87910 }, { "grad_norm": 0.29427623748779297, "learning_rate": 3.9374418666989375e-06, "loss": 0.0253, "step": 87920 }, { "grad_norm": 0.4147396981716156, "learning_rate": 3.931012912957838e-06, "loss": 0.0224, "step": 87930 }, { "grad_norm": 0.2416972517967224, "learning_rate": 3.9245889972519045e-06, "loss": 0.027, "step": 87940 }, { "grad_norm": 0.5284728407859802, "learning_rate": 3.9181701202836265e-06, "loss": 0.0259, "step": 87950 }, { "grad_norm": 1.6532057523727417, "learning_rate": 3.911756282755003e-06, "loss": 0.0281, "step": 87960 }, { "grad_norm": 0.33465468883514404, "learning_rate": 3.905347485367411e-06, "loss": 0.0201, "step": 87970 }, { "grad_norm": 0.3655734062194824, "learning_rate": 3.898943728821725e-06, "loss": 0.0212, "step": 87980 }, { "grad_norm": 1.0589160919189453, "learning_rate": 3.892545013818227e-06, "loss": 0.0186, "step": 87990 }, { "grad_norm": 0.4194317162036896, "learning_rate": 3.886151341056687e-06, "loss": 0.0192, "step": 88000 }, { "grad_norm": 0.33622708916664124, "learning_rate": 3.8797627112363066e-06, "loss": 0.0251, "step": 88010 }, { "grad_norm": 0.36564919352531433, "learning_rate": 3.873379125055743e-06, "loss": 0.0295, "step": 88020 }, { "grad_norm": 0.40160486102104187, "learning_rate": 3.867000583213082e-06, "loss": 0.0342, "step": 88030 }, { "grad_norm": 0.40740081667900085, "learning_rate": 3.860627086405871e-06, "loss": 0.0168, "step": 88040 }, { "grad_norm": 0.5783330798149109, "learning_rate": 3.854258635331126e-06, "loss": 0.0307, "step": 88050 }, { "grad_norm": 0.29020994901657104, "learning_rate": 3.8478952306852665e-06, "loss": 0.0259, "step": 88060 }, { "grad_norm": 0.3230631947517395, "learning_rate": 3.841536873164198e-06, "loss": 0.0199, "step": 88070 }, { "grad_norm": 0.2969048023223877, "learning_rate": 3.8351835634632626e-06, "loss": 0.0237, "step": 88080 }, { "grad_norm": 0.4268794357776642, "learning_rate": 3.828835302277245e-06, "loss": 0.0219, "step": 88090 }, { "grad_norm": 0.5909970998764038, "learning_rate": 3.822492090300378e-06, "loss": 0.0197, "step": 88100 }, { "grad_norm": 0.3735015094280243, "learning_rate": 3.8161539282263566e-06, "loss": 0.0164, "step": 88110 }, { "grad_norm": 0.3000566065311432, "learning_rate": 3.8098208167482995e-06, "loss": 0.0294, "step": 88120 }, { "grad_norm": 0.4897431433200836, "learning_rate": 3.803492756558796e-06, "loss": 0.0231, "step": 88130 }, { "grad_norm": 0.4405982494354248, "learning_rate": 3.797169748349855e-06, "loss": 0.0189, "step": 88140 }, { "grad_norm": 0.26453378796577454, "learning_rate": 3.790851792812977e-06, "loss": 0.0216, "step": 88150 }, { "grad_norm": 0.3713015913963318, "learning_rate": 3.784538890639072e-06, "loss": 0.0228, "step": 88160 }, { "grad_norm": 0.4922407865524292, "learning_rate": 3.7782310425185153e-06, "loss": 0.0212, "step": 88170 }, { "grad_norm": 0.4669059216976166, "learning_rate": 3.77192824914111e-06, "loss": 0.0257, "step": 88180 }, { "grad_norm": 0.4578544795513153, "learning_rate": 3.7656305111961333e-06, "loss": 0.0207, "step": 88190 }, { "grad_norm": 0.23208853602409363, "learning_rate": 3.7593378293722782e-06, "loss": 0.0208, "step": 88200 }, { "grad_norm": 2.581482410430908, "learning_rate": 3.7530502043577333e-06, "loss": 0.0171, "step": 88210 }, { "grad_norm": 0.38182440400123596, "learning_rate": 3.746767636840076e-06, "loss": 0.0218, "step": 88220 }, { "grad_norm": 0.18368889391422272, "learning_rate": 3.74049012750638e-06, "loss": 0.0163, "step": 88230 }, { "grad_norm": 0.43798649311065674, "learning_rate": 3.734217677043128e-06, "loss": 0.0242, "step": 88240 }, { "grad_norm": 0.32581913471221924, "learning_rate": 3.727950286136267e-06, "loss": 0.0127, "step": 88250 }, { "grad_norm": 0.3902742564678192, "learning_rate": 3.7216879554711925e-06, "loss": 0.0205, "step": 88260 }, { "grad_norm": 0.18893173336982727, "learning_rate": 3.715430685732757e-06, "loss": 0.0203, "step": 88270 }, { "grad_norm": 0.2511287331581116, "learning_rate": 3.7091784776052243e-06, "loss": 0.0229, "step": 88280 }, { "grad_norm": 0.45665067434310913, "learning_rate": 3.7029313317723356e-06, "loss": 0.0223, "step": 88290 }, { "grad_norm": 0.18550091981887817, "learning_rate": 3.6966892489172787e-06, "loss": 0.0201, "step": 88300 }, { "grad_norm": 0.27526941895484924, "learning_rate": 3.6904522297226575e-06, "loss": 0.0198, "step": 88310 }, { "grad_norm": 0.41215184330940247, "learning_rate": 3.6842202748705647e-06, "loss": 0.0202, "step": 88320 }, { "grad_norm": 0.3332326114177704, "learning_rate": 3.677993385042494e-06, "loss": 0.0241, "step": 88330 }, { "grad_norm": 0.18558967113494873, "learning_rate": 3.67177156091944e-06, "loss": 0.0212, "step": 88340 }, { "grad_norm": 0.2856270670890808, "learning_rate": 3.6655548031817866e-06, "loss": 0.0254, "step": 88350 }, { "grad_norm": 0.35578468441963196, "learning_rate": 3.6593431125094e-06, "loss": 0.0231, "step": 88360 }, { "grad_norm": 0.24624843895435333, "learning_rate": 3.6531364895815766e-06, "loss": 0.0268, "step": 88370 }, { "grad_norm": 0.3632616102695465, "learning_rate": 3.6469349350770677e-06, "loss": 0.0219, "step": 88380 }, { "grad_norm": 0.28676125407218933, "learning_rate": 3.6407384496740525e-06, "loss": 0.0225, "step": 88390 }, { "grad_norm": 0.7398428320884705, "learning_rate": 3.6345470340501954e-06, "loss": 0.0211, "step": 88400 }, { "grad_norm": 0.8475703001022339, "learning_rate": 3.6283606888825537e-06, "loss": 0.0239, "step": 88410 }, { "grad_norm": 0.25606608390808105, "learning_rate": 3.622179414847676e-06, "loss": 0.0192, "step": 88420 }, { "grad_norm": 0.6645585894584656, "learning_rate": 3.616003212621527e-06, "loss": 0.0266, "step": 88430 }, { "grad_norm": 0.1767847090959549, "learning_rate": 3.609832082879533e-06, "loss": 0.021, "step": 88440 }, { "grad_norm": 0.3456202745437622, "learning_rate": 3.6036660262965485e-06, "loss": 0.0297, "step": 88450 }, { "grad_norm": 0.40883108973503113, "learning_rate": 3.597505043546895e-06, "loss": 0.0186, "step": 88460 }, { "grad_norm": 0.45562806725502014, "learning_rate": 3.5913491353043227e-06, "loss": 0.0185, "step": 88470 }, { "grad_norm": 0.24950705468654633, "learning_rate": 3.5851983022420375e-06, "loss": 0.0217, "step": 88480 }, { "grad_norm": 0.3210699260234833, "learning_rate": 3.5790525450326797e-06, "loss": 0.0186, "step": 88490 }, { "grad_norm": 0.3035736680030823, "learning_rate": 3.572911864348344e-06, "loss": 0.0165, "step": 88500 }, { "grad_norm": 0.19086936116218567, "learning_rate": 3.5667762608605717e-06, "loss": 0.018, "step": 88510 }, { "grad_norm": 0.4571988880634308, "learning_rate": 3.5606457352403255e-06, "loss": 0.0174, "step": 88520 }, { "grad_norm": 0.221058189868927, "learning_rate": 3.554520288158042e-06, "loss": 0.0174, "step": 88530 }, { "grad_norm": 0.1683473438024521, "learning_rate": 3.5483999202835903e-06, "loss": 0.0246, "step": 88540 }, { "grad_norm": 0.3238094747066498, "learning_rate": 3.5422846322862913e-06, "loss": 0.0228, "step": 88550 }, { "grad_norm": 0.3105461597442627, "learning_rate": 3.5361744248348925e-06, "loss": 0.0165, "step": 88560 }, { "grad_norm": 0.3672339618206024, "learning_rate": 3.5300692985976047e-06, "loss": 0.0239, "step": 88570 }, { "grad_norm": 0.2590796649456024, "learning_rate": 3.52396925424206e-06, "loss": 0.0167, "step": 88580 }, { "grad_norm": 0.2295248806476593, "learning_rate": 3.5178742924353747e-06, "loss": 0.0195, "step": 88590 }, { "grad_norm": 0.3238043189048767, "learning_rate": 3.511784413844066e-06, "loss": 0.021, "step": 88600 }, { "grad_norm": 0.4855080842971802, "learning_rate": 3.5056996191341287e-06, "loss": 0.0205, "step": 88610 }, { "grad_norm": 0.38341760635375977, "learning_rate": 3.4996199089709692e-06, "loss": 0.0185, "step": 88620 }, { "grad_norm": 1.6323914527893066, "learning_rate": 3.4935452840194725e-06, "loss": 0.025, "step": 88630 }, { "grad_norm": 0.4110892713069916, "learning_rate": 3.48747574494393e-06, "loss": 0.0206, "step": 88640 }, { "grad_norm": 0.5302989482879639, "learning_rate": 3.4814112924081154e-06, "loss": 0.0204, "step": 88650 }, { "grad_norm": 0.3156740665435791, "learning_rate": 3.475351927075221e-06, "loss": 0.0238, "step": 88660 }, { "grad_norm": 0.4734020531177521, "learning_rate": 3.469297649607894e-06, "loss": 0.0204, "step": 88670 }, { "grad_norm": 0.14658822119235992, "learning_rate": 3.463248460668217e-06, "loss": 0.0177, "step": 88680 }, { "grad_norm": 0.6715136170387268, "learning_rate": 3.457204360917726e-06, "loss": 0.0342, "step": 88690 }, { "grad_norm": 0.2076283097267151, "learning_rate": 3.4511653510173827e-06, "loss": 0.0164, "step": 88700 }, { "grad_norm": 0.38609933853149414, "learning_rate": 3.4451314316276074e-06, "loss": 0.0263, "step": 88710 }, { "grad_norm": 0.461177259683609, "learning_rate": 3.4391026034082674e-06, "loss": 0.0297, "step": 88720 }, { "grad_norm": 0.41488996148109436, "learning_rate": 3.433078867018674e-06, "loss": 0.0209, "step": 88730 }, { "grad_norm": 0.3669340908527374, "learning_rate": 3.427060223117551e-06, "loss": 0.0217, "step": 88740 }, { "grad_norm": 0.3652999997138977, "learning_rate": 3.4210466723631042e-06, "loss": 0.0249, "step": 88750 }, { "grad_norm": 0.5733596086502075, "learning_rate": 3.415038215412969e-06, "loss": 0.0215, "step": 88760 }, { "grad_norm": 0.39691808819770813, "learning_rate": 3.4090348529242088e-06, "loss": 0.0235, "step": 88770 }, { "grad_norm": 0.6083050966262817, "learning_rate": 3.4030365855533487e-06, "loss": 0.0283, "step": 88780 }, { "grad_norm": 0.39601364731788635, "learning_rate": 3.397043413956347e-06, "loss": 0.0217, "step": 88790 }, { "grad_norm": 0.4312916398048401, "learning_rate": 3.3910553387886237e-06, "loss": 0.022, "step": 88800 }, { "grad_norm": 0.4851374328136444, "learning_rate": 3.3850723607049996e-06, "loss": 0.0272, "step": 88810 }, { "grad_norm": 0.6265157461166382, "learning_rate": 3.3790944803597902e-06, "loss": 0.0212, "step": 88820 }, { "grad_norm": 0.21362224221229553, "learning_rate": 3.373121698406706e-06, "loss": 0.0231, "step": 88830 }, { "grad_norm": 0.29474154114723206, "learning_rate": 3.3671540154989302e-06, "loss": 0.0181, "step": 88840 }, { "grad_norm": 0.361689954996109, "learning_rate": 3.361191432289079e-06, "loss": 0.02, "step": 88850 }, { "grad_norm": 0.3751175105571747, "learning_rate": 3.3552339494292207e-06, "loss": 0.0224, "step": 88860 }, { "grad_norm": 0.2482513189315796, "learning_rate": 3.349281567570839e-06, "loss": 0.0246, "step": 88870 }, { "grad_norm": 0.23883643746376038, "learning_rate": 3.3433342873648965e-06, "loss": 0.0226, "step": 88880 }, { "grad_norm": 0.3384586572647095, "learning_rate": 3.337392109461762e-06, "loss": 0.023, "step": 88890 }, { "grad_norm": 0.5529384613037109, "learning_rate": 3.331455034511266e-06, "loss": 0.0296, "step": 88900 }, { "grad_norm": 0.32227784395217896, "learning_rate": 3.325523063162689e-06, "loss": 0.0227, "step": 88910 }, { "grad_norm": 0.38367870450019836, "learning_rate": 3.3195961960647393e-06, "loss": 0.0172, "step": 88920 }, { "grad_norm": 0.2685835063457489, "learning_rate": 3.31367443386556e-06, "loss": 0.0217, "step": 88930 }, { "grad_norm": 0.325104296207428, "learning_rate": 3.3077577772127556e-06, "loss": 0.0187, "step": 88940 }, { "grad_norm": 0.42248982191085815, "learning_rate": 3.301846226753358e-06, "loss": 0.0281, "step": 88950 }, { "grad_norm": 0.27413588762283325, "learning_rate": 3.2959397831338444e-06, "loss": 0.0227, "step": 88960 }, { "grad_norm": 0.5348985195159912, "learning_rate": 3.2900384470001432e-06, "loss": 0.0244, "step": 88970 }, { "grad_norm": 0.4098385274410248, "learning_rate": 3.28414221899761e-06, "loss": 0.0203, "step": 88980 }, { "grad_norm": 0.2738470733165741, "learning_rate": 3.278251099771046e-06, "loss": 0.0215, "step": 88990 }, { "grad_norm": 0.8875157833099365, "learning_rate": 3.2723650899646906e-06, "loss": 0.0275, "step": 89000 }, { "grad_norm": 0.4780159592628479, "learning_rate": 3.2664841902222466e-06, "loss": 0.018, "step": 89010 }, { "grad_norm": 0.21018123626708984, "learning_rate": 3.2606084011868208e-06, "loss": 0.0238, "step": 89020 }, { "grad_norm": 0.9949271082878113, "learning_rate": 3.254737723500989e-06, "loss": 0.0184, "step": 89030 }, { "grad_norm": 0.4969741702079773, "learning_rate": 3.2488721578067595e-06, "loss": 0.0165, "step": 89040 }, { "grad_norm": 0.1693297177553177, "learning_rate": 3.243011704745591e-06, "loss": 0.0208, "step": 89050 }, { "grad_norm": 0.38954824209213257, "learning_rate": 3.237156364958355e-06, "loss": 0.0289, "step": 89060 }, { "grad_norm": 0.44425058364868164, "learning_rate": 3.231306139085394e-06, "loss": 0.0183, "step": 89070 }, { "grad_norm": 0.35903599858283997, "learning_rate": 3.2254610277664797e-06, "loss": 0.0167, "step": 89080 }, { "grad_norm": 0.32725709676742554, "learning_rate": 3.2196210316408183e-06, "loss": 0.0226, "step": 89090 }, { "grad_norm": 0.33546799421310425, "learning_rate": 3.2137861513470644e-06, "loss": 0.0237, "step": 89100 }, { "grad_norm": 0.27756258845329285, "learning_rate": 3.2079563875233252e-06, "loss": 0.0187, "step": 89110 }, { "grad_norm": 0.26821815967559814, "learning_rate": 3.2021317408071184e-06, "loss": 0.0182, "step": 89120 }, { "grad_norm": 0.5725843906402588, "learning_rate": 3.196312211835434e-06, "loss": 0.0195, "step": 89130 }, { "grad_norm": 0.28033262491226196, "learning_rate": 3.1904978012446686e-06, "loss": 0.0172, "step": 89140 }, { "grad_norm": 0.3058093786239624, "learning_rate": 3.184688509670691e-06, "loss": 0.0186, "step": 89150 }, { "grad_norm": 0.23782679438591003, "learning_rate": 3.178884337748783e-06, "loss": 0.0158, "step": 89160 }, { "grad_norm": 0.22628219425678253, "learning_rate": 3.173085286113697e-06, "loss": 0.0266, "step": 89170 }, { "grad_norm": 0.4280529022216797, "learning_rate": 3.1672913553995986e-06, "loss": 0.0196, "step": 89180 }, { "grad_norm": 0.2691478133201599, "learning_rate": 3.1615025462401138e-06, "loss": 0.0138, "step": 89190 }, { "grad_norm": 0.3078155219554901, "learning_rate": 3.1557188592682808e-06, "loss": 0.0215, "step": 89200 }, { "grad_norm": 0.21134549379348755, "learning_rate": 3.1499402951166045e-06, "loss": 0.0228, "step": 89210 }, { "grad_norm": 1.3455053567886353, "learning_rate": 3.144166854417019e-06, "loss": 0.0138, "step": 89220 }, { "grad_norm": 0.3522684574127197, "learning_rate": 3.1383985378009073e-06, "loss": 0.0177, "step": 89230 }, { "grad_norm": 0.25146692991256714, "learning_rate": 3.1326353458990654e-06, "loss": 0.018, "step": 89240 }, { "grad_norm": 0.6374611258506775, "learning_rate": 3.126877279341761e-06, "loss": 0.0177, "step": 89250 }, { "grad_norm": 0.2726072371006012, "learning_rate": 3.1211243387586964e-06, "loss": 0.0221, "step": 89260 }, { "grad_norm": 0.3546031713485718, "learning_rate": 3.115376524778979e-06, "loss": 0.0265, "step": 89270 }, { "grad_norm": 0.28289544582366943, "learning_rate": 3.109633838031195e-06, "loss": 0.0237, "step": 89280 }, { "grad_norm": 0.19246642291545868, "learning_rate": 3.103896279143359e-06, "loss": 0.0251, "step": 89290 }, { "grad_norm": 3.267343282699585, "learning_rate": 3.098163848742924e-06, "loss": 0.0248, "step": 89300 }, { "grad_norm": 0.26974934339523315, "learning_rate": 3.0924365474567675e-06, "loss": 0.0237, "step": 89310 }, { "grad_norm": 0.35034358501434326, "learning_rate": 3.086714375911237e-06, "loss": 0.0202, "step": 89320 }, { "grad_norm": 0.36963191628456116, "learning_rate": 3.0809973347320775e-06, "loss": 0.0239, "step": 89330 }, { "grad_norm": 0.31051522493362427, "learning_rate": 3.075285424544516e-06, "loss": 0.0228, "step": 89340 }, { "grad_norm": 0.3566654324531555, "learning_rate": 3.069578645973176e-06, "loss": 0.0278, "step": 89350 }, { "grad_norm": 0.31628531217575073, "learning_rate": 3.0638769996421735e-06, "loss": 0.0171, "step": 89360 }, { "grad_norm": 0.6733806729316711, "learning_rate": 3.058180486175005e-06, "loss": 0.0325, "step": 89370 }, { "grad_norm": 0.3460615873336792, "learning_rate": 3.052489106194645e-06, "loss": 0.0256, "step": 89380 }, { "grad_norm": 0.5457899570465088, "learning_rate": 3.0468028603234946e-06, "loss": 0.0189, "step": 89390 }, { "grad_norm": 0.3085246682167053, "learning_rate": 3.04112174918339e-06, "loss": 0.0227, "step": 89400 }, { "grad_norm": 0.31022581458091736, "learning_rate": 3.0354457733955954e-06, "loss": 0.0204, "step": 89410 }, { "grad_norm": 0.34468066692352295, "learning_rate": 3.029774933580859e-06, "loss": 0.0185, "step": 89420 }, { "grad_norm": 0.37889572978019714, "learning_rate": 3.0241092303593065e-06, "loss": 0.0261, "step": 89430 }, { "grad_norm": 0.19566166400909424, "learning_rate": 3.018448664350554e-06, "loss": 0.0162, "step": 89440 }, { "grad_norm": 0.4261704981327057, "learning_rate": 3.0127932361736055e-06, "loss": 0.0208, "step": 89450 }, { "grad_norm": 0.6513588428497314, "learning_rate": 3.00714294644695e-06, "loss": 0.0223, "step": 89460 }, { "grad_norm": 0.2093847692012787, "learning_rate": 3.0014977957884828e-06, "loss": 0.0209, "step": 89470 }, { "grad_norm": 0.18429745733737946, "learning_rate": 2.99585778481557e-06, "loss": 0.0201, "step": 89480 }, { "grad_norm": 0.3843257427215576, "learning_rate": 2.990222914144969e-06, "loss": 0.028, "step": 89490 }, { "grad_norm": 0.3334127366542816, "learning_rate": 2.984593184392914e-06, "loss": 0.0232, "step": 89500 }, { "grad_norm": 0.9754445552825928, "learning_rate": 2.978968596175069e-06, "loss": 0.0234, "step": 89510 }, { "grad_norm": 0.21710346639156342, "learning_rate": 2.9733491501065137e-06, "loss": 0.0215, "step": 89520 }, { "grad_norm": 0.30074983835220337, "learning_rate": 2.9677348468018005e-06, "loss": 0.0268, "step": 89530 }, { "grad_norm": 0.634329617023468, "learning_rate": 2.9621256868748726e-06, "loss": 0.0269, "step": 89540 }, { "grad_norm": 0.4051687717437744, "learning_rate": 2.956521670939183e-06, "loss": 0.0293, "step": 89550 }, { "grad_norm": 0.2393137663602829, "learning_rate": 2.950922799607542e-06, "loss": 0.0273, "step": 89560 }, { "grad_norm": 0.19542425870895386, "learning_rate": 2.9453290734922535e-06, "loss": 0.0256, "step": 89570 }, { "grad_norm": 0.49950358271598816, "learning_rate": 2.939740493205029e-06, "loss": 0.0372, "step": 89580 }, { "grad_norm": 0.2269536703824997, "learning_rate": 2.9341570593570345e-06, "loss": 0.0223, "step": 89590 }, { "grad_norm": 0.32314711809158325, "learning_rate": 2.928578772558843e-06, "loss": 0.0226, "step": 89600 }, { "grad_norm": 0.4564574360847473, "learning_rate": 2.923005633420528e-06, "loss": 0.0187, "step": 89610 }, { "grad_norm": 0.30072149634361267, "learning_rate": 2.9174376425515237e-06, "loss": 0.0249, "step": 89620 }, { "grad_norm": 0.6436334848403931, "learning_rate": 2.91187480056076e-06, "loss": 0.0224, "step": 89630 }, { "grad_norm": 0.2718408703804016, "learning_rate": 2.9063171080565664e-06, "loss": 0.0192, "step": 89640 }, { "grad_norm": 0.23600783944129944, "learning_rate": 2.900764565646741e-06, "loss": 0.0209, "step": 89650 }, { "grad_norm": 0.5973851680755615, "learning_rate": 2.895217173938475e-06, "loss": 0.0225, "step": 89660 }, { "grad_norm": 0.25816434621810913, "learning_rate": 2.8896749335384445e-06, "loss": 0.0172, "step": 89670 }, { "grad_norm": 0.42950502038002014, "learning_rate": 2.884137845052731e-06, "loss": 0.0279, "step": 89680 }, { "grad_norm": 0.3895989954471588, "learning_rate": 2.8786059090868734e-06, "loss": 0.0193, "step": 89690 }, { "grad_norm": 0.316724956035614, "learning_rate": 2.87307912624582e-06, "loss": 0.0198, "step": 89700 }, { "grad_norm": 0.254854679107666, "learning_rate": 2.867557497133977e-06, "loss": 0.0177, "step": 89710 }, { "grad_norm": 0.28057152032852173, "learning_rate": 2.862041022355194e-06, "loss": 0.0268, "step": 89720 }, { "grad_norm": 0.32265913486480713, "learning_rate": 2.8565297025127226e-06, "loss": 0.0241, "step": 89730 }, { "grad_norm": 0.2370363473892212, "learning_rate": 2.8510235382092853e-06, "loss": 0.0146, "step": 89740 }, { "grad_norm": 0.23401814699172974, "learning_rate": 2.845522530047029e-06, "loss": 0.0195, "step": 89750 }, { "grad_norm": 0.30935874581336975, "learning_rate": 2.8400266786275387e-06, "loss": 0.0196, "step": 89760 }, { "grad_norm": 0.47789040207862854, "learning_rate": 2.834535984551817e-06, "loss": 0.0259, "step": 89770 }, { "grad_norm": 0.2594662308692932, "learning_rate": 2.8290504484203397e-06, "loss": 0.0327, "step": 89780 }, { "grad_norm": 0.282816082239151, "learning_rate": 2.8235700708329705e-06, "loss": 0.0246, "step": 89790 }, { "grad_norm": 0.3637571930885315, "learning_rate": 2.818094852389058e-06, "loss": 0.0258, "step": 89800 }, { "grad_norm": 0.44642889499664307, "learning_rate": 2.8126247936873516e-06, "loss": 0.0264, "step": 89810 }, { "grad_norm": 0.27831658720970154, "learning_rate": 2.8071598953260614e-06, "loss": 0.0255, "step": 89820 }, { "grad_norm": 0.25209906697273254, "learning_rate": 2.8017001579028035e-06, "loss": 0.0258, "step": 89830 }, { "grad_norm": 0.25465628504753113, "learning_rate": 2.7962455820146617e-06, "loss": 0.0256, "step": 89840 }, { "grad_norm": 0.27883562445640564, "learning_rate": 2.7907961682581253e-06, "loss": 0.0164, "step": 89850 }, { "grad_norm": 1.04653799533844, "learning_rate": 2.7853519172291453e-06, "loss": 0.0247, "step": 89860 }, { "grad_norm": 0.21723386645317078, "learning_rate": 2.7799128295230904e-06, "loss": 0.0198, "step": 89870 }, { "grad_norm": 0.3448318541049957, "learning_rate": 2.774478905734784e-06, "loss": 0.0198, "step": 89880 }, { "grad_norm": 0.2862992584705353, "learning_rate": 2.7690501464584516e-06, "loss": 0.0201, "step": 89890 }, { "grad_norm": 0.5312470197677612, "learning_rate": 2.763626552287796e-06, "loss": 0.0243, "step": 89900 }, { "grad_norm": 0.2246425300836563, "learning_rate": 2.7582081238159083e-06, "loss": 0.0176, "step": 89910 }, { "grad_norm": 0.24031734466552734, "learning_rate": 2.7527948616353548e-06, "loss": 0.0224, "step": 89920 }, { "grad_norm": 0.19704633951187134, "learning_rate": 2.7473867663381226e-06, "loss": 0.0256, "step": 89930 }, { "grad_norm": 0.4984841048717499, "learning_rate": 2.7419838385156383e-06, "loss": 0.0204, "step": 89940 }, { "grad_norm": 0.37764057517051697, "learning_rate": 2.7365860787587407e-06, "loss": 0.0299, "step": 89950 }, { "grad_norm": 0.48376643657684326, "learning_rate": 2.7311934876577295e-06, "loss": 0.0241, "step": 89960 }, { "grad_norm": 0.28878679871559143, "learning_rate": 2.725806065802339e-06, "loss": 0.0234, "step": 89970 }, { "grad_norm": 0.6439412832260132, "learning_rate": 2.720423813781714e-06, "loss": 0.019, "step": 89980 }, { "grad_norm": 0.251670241355896, "learning_rate": 2.715046732184462e-06, "loss": 0.0184, "step": 89990 }, { "grad_norm": 0.31258994340896606, "learning_rate": 2.7096748215986013e-06, "loss": 0.0199, "step": 90000 }, { "grad_norm": 0.2582642734050751, "learning_rate": 2.704308082611612e-06, "loss": 0.0228, "step": 90010 }, { "grad_norm": 0.3938048183917999, "learning_rate": 2.698946515810374e-06, "loss": 0.0182, "step": 90020 }, { "grad_norm": 0.21339836716651917, "learning_rate": 2.6935901217812367e-06, "loss": 0.019, "step": 90030 }, { "grad_norm": 0.474916011095047, "learning_rate": 2.688238901109952e-06, "loss": 0.0215, "step": 90040 }, { "grad_norm": 0.3433132767677307, "learning_rate": 2.68289285438173e-06, "loss": 0.0202, "step": 90050 }, { "grad_norm": 1.7903392314910889, "learning_rate": 2.6775519821812033e-06, "loss": 0.0192, "step": 90060 }, { "grad_norm": 1.047564148902893, "learning_rate": 2.672216285092449e-06, "loss": 0.0215, "step": 90070 }, { "grad_norm": 0.2685430347919464, "learning_rate": 2.666885763698962e-06, "loss": 0.0179, "step": 90080 }, { "grad_norm": 0.20983773469924927, "learning_rate": 2.661560418583692e-06, "loss": 0.0152, "step": 90090 }, { "grad_norm": 0.32453885674476624, "learning_rate": 2.656240250328995e-06, "loss": 0.023, "step": 90100 }, { "grad_norm": 0.46158167719841003, "learning_rate": 2.650925259516679e-06, "loss": 0.0244, "step": 90110 }, { "grad_norm": 0.6190706491470337, "learning_rate": 2.645615446727995e-06, "loss": 0.0188, "step": 90120 }, { "grad_norm": 0.6409497261047363, "learning_rate": 2.6403108125436115e-06, "loss": 0.0197, "step": 90130 }, { "grad_norm": 1.2898786067962646, "learning_rate": 2.6350113575436264e-06, "loss": 0.0259, "step": 90140 }, { "grad_norm": 0.4551607370376587, "learning_rate": 2.629717082307598e-06, "loss": 0.0161, "step": 90150 }, { "grad_norm": 0.4209073781967163, "learning_rate": 2.6244279874144793e-06, "loss": 0.0231, "step": 90160 }, { "grad_norm": 0.4134924113750458, "learning_rate": 2.6191440734426862e-06, "loss": 0.0235, "step": 90170 }, { "grad_norm": 0.7867029905319214, "learning_rate": 2.613865340970062e-06, "loss": 0.0208, "step": 90180 }, { "grad_norm": 0.2307473123073578, "learning_rate": 2.608591790573889e-06, "loss": 0.0219, "step": 90190 }, { "grad_norm": 0.201446071267128, "learning_rate": 2.603323422830861e-06, "loss": 0.0198, "step": 90200 }, { "grad_norm": 0.2602895200252533, "learning_rate": 2.5980602383171183e-06, "loss": 0.0176, "step": 90210 }, { "grad_norm": 0.3217598497867584, "learning_rate": 2.5928022376082496e-06, "loss": 0.0228, "step": 90220 }, { "grad_norm": 0.46910378336906433, "learning_rate": 2.5875494212792505e-06, "loss": 0.036, "step": 90230 }, { "grad_norm": 1.3783518075942993, "learning_rate": 2.5823017899045565e-06, "loss": 0.0286, "step": 90240 }, { "grad_norm": 0.18772079050540924, "learning_rate": 2.577059344058047e-06, "loss": 0.0219, "step": 90250 }, { "grad_norm": 0.25119084119796753, "learning_rate": 2.5718220843130415e-06, "loss": 0.0185, "step": 90260 }, { "grad_norm": 0.3772468864917755, "learning_rate": 2.5665900112422535e-06, "loss": 0.0261, "step": 90270 }, { "grad_norm": 0.3368675708770752, "learning_rate": 2.561363125417876e-06, "loss": 0.0318, "step": 90280 }, { "grad_norm": 0.2899065613746643, "learning_rate": 2.556141427411496e-06, "loss": 0.0181, "step": 90290 }, { "grad_norm": 0.2767437994480133, "learning_rate": 2.550924917794156e-06, "loss": 0.0218, "step": 90300 }, { "grad_norm": 0.3873520791530609, "learning_rate": 2.545713597136329e-06, "loss": 0.028, "step": 90310 }, { "grad_norm": 0.1983659416437149, "learning_rate": 2.5405074660079298e-06, "loss": 0.0195, "step": 90320 }, { "grad_norm": 0.34857773780822754, "learning_rate": 2.5353065249782647e-06, "loss": 0.0202, "step": 90330 }, { "grad_norm": 0.45158103108406067, "learning_rate": 2.530110774616129e-06, "loss": 0.0327, "step": 90340 }, { "grad_norm": 0.2478100210428238, "learning_rate": 2.524920215489701e-06, "loss": 0.0186, "step": 90350 }, { "grad_norm": 0.20347870886325836, "learning_rate": 2.519734848166627e-06, "loss": 0.0168, "step": 90360 }, { "grad_norm": 0.5211716890335083, "learning_rate": 2.514554673213948e-06, "loss": 0.0223, "step": 90370 }, { "grad_norm": 0.44585198163986206, "learning_rate": 2.5093796911981936e-06, "loss": 0.0296, "step": 90380 }, { "grad_norm": 0.24346069991588593, "learning_rate": 2.5042099026852672e-06, "loss": 0.0319, "step": 90390 }, { "grad_norm": 0.27294617891311646, "learning_rate": 2.499045308240544e-06, "loss": 0.0198, "step": 90400 }, { "grad_norm": 0.2754809260368347, "learning_rate": 2.4938859084288003e-06, "loss": 0.0211, "step": 90410 }, { "grad_norm": 0.40697187185287476, "learning_rate": 2.4887317038142734e-06, "loss": 0.0233, "step": 90420 }, { "grad_norm": 0.26483139395713806, "learning_rate": 2.4835826949606124e-06, "loss": 0.0173, "step": 90430 }, { "grad_norm": 0.4007694721221924, "learning_rate": 2.4784388824309167e-06, "loss": 0.0275, "step": 90440 }, { "grad_norm": 0.36197763681411743, "learning_rate": 2.473300266787687e-06, "loss": 0.0214, "step": 90450 }, { "grad_norm": 0.3187837302684784, "learning_rate": 2.46816684859289e-06, "loss": 0.03, "step": 90460 }, { "grad_norm": 0.3058595657348633, "learning_rate": 2.4630386284079043e-06, "loss": 0.0231, "step": 90470 }, { "grad_norm": 0.4661146402359009, "learning_rate": 2.4579156067935427e-06, "loss": 0.029, "step": 90480 }, { "grad_norm": 0.2792145907878876, "learning_rate": 2.452797784310046e-06, "loss": 0.0204, "step": 90490 }, { "grad_norm": 1.9965119361877441, "learning_rate": 2.447685161517105e-06, "loss": 0.0248, "step": 90500 }, { "grad_norm": 0.23594503104686737, "learning_rate": 2.4425777389738224e-06, "loss": 0.0233, "step": 90510 }, { "grad_norm": 0.7052959203720093, "learning_rate": 2.43747551723873e-06, "loss": 0.0212, "step": 90520 }, { "grad_norm": 0.21811261773109436, "learning_rate": 2.4323784968698136e-06, "loss": 0.0255, "step": 90530 }, { "grad_norm": 0.38899165391921997, "learning_rate": 2.4272866784244565e-06, "loss": 0.0214, "step": 90540 }, { "grad_norm": 0.4617716670036316, "learning_rate": 2.422200062459518e-06, "loss": 0.0193, "step": 90550 }, { "grad_norm": 0.22724953293800354, "learning_rate": 2.41711864953123e-06, "loss": 0.0225, "step": 90560 }, { "grad_norm": 0.6777654886245728, "learning_rate": 2.412042440195322e-06, "loss": 0.031, "step": 90570 }, { "grad_norm": 0.45288020372390747, "learning_rate": 2.4069714350068993e-06, "loss": 0.0219, "step": 90580 }, { "grad_norm": 0.24324588477611542, "learning_rate": 2.4019056345205294e-06, "loss": 0.0243, "step": 90590 }, { "grad_norm": 0.3507422208786011, "learning_rate": 2.396845039290191e-06, "loss": 0.0153, "step": 90600 }, { "grad_norm": 1.4512118101119995, "learning_rate": 2.3917896498693194e-06, "loss": 0.0164, "step": 90610 }, { "grad_norm": 0.4230686128139496, "learning_rate": 2.3867394668107335e-06, "loss": 0.0184, "step": 90620 }, { "grad_norm": 0.512110710144043, "learning_rate": 2.3816944906667527e-06, "loss": 0.0158, "step": 90630 }, { "grad_norm": 0.3448067605495453, "learning_rate": 2.3766547219890634e-06, "loss": 0.0192, "step": 90640 }, { "grad_norm": 0.29559698700904846, "learning_rate": 2.371620161328819e-06, "loss": 0.032, "step": 90650 }, { "grad_norm": 0.7630044221878052, "learning_rate": 2.366590809236574e-06, "loss": 0.0229, "step": 90660 }, { "grad_norm": 0.2980910837650299, "learning_rate": 2.3615666662623494e-06, "loss": 0.0224, "step": 90670 }, { "grad_norm": 0.23262648284435272, "learning_rate": 2.3565477329555664e-06, "loss": 0.0205, "step": 90680 }, { "grad_norm": 0.20193317532539368, "learning_rate": 2.3515340098650975e-06, "loss": 0.0138, "step": 90690 }, { "grad_norm": 0.23499323427677155, "learning_rate": 2.3465254975392258e-06, "loss": 0.016, "step": 90700 }, { "grad_norm": 0.3620157539844513, "learning_rate": 2.3415221965256805e-06, "loss": 0.0281, "step": 90710 }, { "grad_norm": 0.29825639724731445, "learning_rate": 2.3365241073716237e-06, "loss": 0.0249, "step": 90720 }, { "grad_norm": 0.43967536091804504, "learning_rate": 2.3315312306236183e-06, "loss": 0.0276, "step": 90730 }, { "grad_norm": 0.19959942996501923, "learning_rate": 2.326543566827699e-06, "loss": 0.0224, "step": 90740 }, { "grad_norm": 0.722782552242279, "learning_rate": 2.3215611165292862e-06, "loss": 0.018, "step": 90750 }, { "grad_norm": 0.5590406656265259, "learning_rate": 2.3165838802732765e-06, "loss": 0.0219, "step": 90760 }, { "grad_norm": 0.4797528088092804, "learning_rate": 2.311611858603957e-06, "loss": 0.0223, "step": 90770 }, { "grad_norm": 0.7402040362358093, "learning_rate": 2.3066450520650763e-06, "loss": 0.0257, "step": 90780 }, { "grad_norm": 0.20095543563365936, "learning_rate": 2.3016834611997773e-06, "loss": 0.0304, "step": 90790 }, { "grad_norm": 0.27037784457206726, "learning_rate": 2.296727086550665e-06, "loss": 0.0235, "step": 90800 }, { "grad_norm": 0.26486197113990784, "learning_rate": 2.2917759286597507e-06, "loss": 0.0268, "step": 90810 }, { "grad_norm": 0.15266139805316925, "learning_rate": 2.2868299880685006e-06, "loss": 0.026, "step": 90820 }, { "grad_norm": 0.33546245098114014, "learning_rate": 2.2818892653177827e-06, "loss": 0.0242, "step": 90830 }, { "grad_norm": 0.30452924966812134, "learning_rate": 2.2769537609479196e-06, "loss": 0.0205, "step": 90840 }, { "grad_norm": 0.2504807114601135, "learning_rate": 2.2720234754986357e-06, "loss": 0.0214, "step": 90850 }, { "grad_norm": 0.39712104201316833, "learning_rate": 2.2670984095091108e-06, "loss": 0.0234, "step": 90860 }, { "grad_norm": 0.4854748547077179, "learning_rate": 2.262178563517936e-06, "loss": 0.0264, "step": 90870 }, { "grad_norm": 0.2261931449174881, "learning_rate": 2.257263938063137e-06, "loss": 0.0189, "step": 90880 }, { "grad_norm": 0.4079064428806305, "learning_rate": 2.2523545336821715e-06, "loss": 0.0245, "step": 90890 }, { "grad_norm": 0.1836652308702469, "learning_rate": 2.247450350911939e-06, "loss": 0.022, "step": 90900 }, { "grad_norm": 0.2122773677110672, "learning_rate": 2.2425513902887328e-06, "loss": 0.0181, "step": 90910 }, { "grad_norm": 0.5178730487823486, "learning_rate": 2.2376576523483015e-06, "loss": 0.0229, "step": 90920 }, { "grad_norm": 0.6877391338348389, "learning_rate": 2.232769137625829e-06, "loss": 0.0238, "step": 90930 }, { "grad_norm": 0.30274006724357605, "learning_rate": 2.227885846655903e-06, "loss": 0.0224, "step": 90940 }, { "grad_norm": 0.5709846019744873, "learning_rate": 2.2230077799725523e-06, "loss": 0.025, "step": 90950 }, { "grad_norm": 0.43848586082458496, "learning_rate": 2.2181349381092396e-06, "loss": 0.0264, "step": 90960 }, { "grad_norm": 0.30328309535980225, "learning_rate": 2.2132673215988552e-06, "loss": 0.021, "step": 90970 }, { "grad_norm": 0.3856034576892853, "learning_rate": 2.2084049309737065e-06, "loss": 0.0156, "step": 90980 }, { "grad_norm": 0.2790158987045288, "learning_rate": 2.2035477667655513e-06, "loss": 0.021, "step": 90990 }, { "grad_norm": 0.6804947853088379, "learning_rate": 2.1986958295055315e-06, "loss": 0.0211, "step": 91000 }, { "grad_norm": 0.27752766013145447, "learning_rate": 2.193849119724284e-06, "loss": 0.0227, "step": 91010 }, { "grad_norm": 0.6922517418861389, "learning_rate": 2.1890076379518175e-06, "loss": 0.0228, "step": 91020 }, { "grad_norm": 0.4480845332145691, "learning_rate": 2.1841713847175925e-06, "loss": 0.027, "step": 91030 }, { "grad_norm": 0.41589897871017456, "learning_rate": 2.1793403605504968e-06, "loss": 0.027, "step": 91040 }, { "grad_norm": 0.363031804561615, "learning_rate": 2.1745145659788413e-06, "loss": 0.0254, "step": 91050 }, { "grad_norm": 0.4071204364299774, "learning_rate": 2.1696940015303645e-06, "loss": 0.016, "step": 91060 }, { "grad_norm": 0.3573334217071533, "learning_rate": 2.1648786677322397e-06, "loss": 0.0287, "step": 91070 }, { "grad_norm": 0.5584710240364075, "learning_rate": 2.1600685651110673e-06, "loss": 0.0243, "step": 91080 }, { "grad_norm": 0.3879458010196686, "learning_rate": 2.1552636941928717e-06, "loss": 0.0276, "step": 91090 }, { "grad_norm": 0.4351278245449066, "learning_rate": 2.1504640555031043e-06, "loss": 0.0206, "step": 91100 }, { "grad_norm": 0.3946174383163452, "learning_rate": 2.1456696495666506e-06, "loss": 0.0237, "step": 91110 }, { "grad_norm": 0.3716668486595154, "learning_rate": 2.1408804769078084e-06, "loss": 0.0257, "step": 91120 }, { "grad_norm": 0.22825756669044495, "learning_rate": 2.136096538050325e-06, "loss": 0.0216, "step": 91130 }, { "grad_norm": 0.28518423438072205, "learning_rate": 2.13131783351736e-06, "loss": 0.0274, "step": 91140 }, { "grad_norm": 0.2911650836467743, "learning_rate": 2.126544363831512e-06, "loss": 0.0195, "step": 91150 }, { "grad_norm": 0.27205318212509155, "learning_rate": 2.121776129514791e-06, "loss": 0.0209, "step": 91160 }, { "grad_norm": 0.4301101863384247, "learning_rate": 2.1170131310886523e-06, "loss": 0.0171, "step": 91170 }, { "grad_norm": 0.2029721736907959, "learning_rate": 2.112255369073968e-06, "loss": 0.021, "step": 91180 }, { "grad_norm": 0.16527198255062103, "learning_rate": 2.1075028439910328e-06, "loss": 0.0158, "step": 91190 }, { "grad_norm": 0.2996843457221985, "learning_rate": 2.1027555563595813e-06, "loss": 0.0224, "step": 91200 }, { "grad_norm": 0.45641255378723145, "learning_rate": 2.0980135066987706e-06, "loss": 0.0195, "step": 91210 }, { "grad_norm": 0.22078843414783478, "learning_rate": 2.093276695527191e-06, "loss": 0.0227, "step": 91220 }, { "grad_norm": 0.5997380614280701, "learning_rate": 2.0885451233628396e-06, "loss": 0.0207, "step": 91230 }, { "grad_norm": 0.23064884543418884, "learning_rate": 2.0838187907231633e-06, "loss": 0.0277, "step": 91240 }, { "grad_norm": 0.3118920922279358, "learning_rate": 2.0790976981250153e-06, "loss": 0.0189, "step": 91250 }, { "grad_norm": 0.3446860909461975, "learning_rate": 2.0743818460847e-06, "loss": 0.0161, "step": 91260 }, { "grad_norm": 0.26675334572792053, "learning_rate": 2.069671235117926e-06, "loss": 0.0217, "step": 91270 }, { "grad_norm": 0.31490084528923035, "learning_rate": 2.0649658657398484e-06, "loss": 0.0255, "step": 91280 }, { "grad_norm": 0.2632286548614502, "learning_rate": 2.060265738465034e-06, "loss": 0.0195, "step": 91290 }, { "grad_norm": 0.21924401819705963, "learning_rate": 2.055570853807487e-06, "loss": 0.0168, "step": 91300 }, { "grad_norm": 0.3196547329425812, "learning_rate": 2.0508812122806196e-06, "loss": 0.016, "step": 91310 }, { "grad_norm": 0.2917501628398895, "learning_rate": 2.046196814397294e-06, "loss": 0.0215, "step": 91320 }, { "grad_norm": 0.29706400632858276, "learning_rate": 2.0415176606697896e-06, "loss": 0.0208, "step": 91330 }, { "grad_norm": 0.3211340010166168, "learning_rate": 2.0368437516098126e-06, "loss": 0.0268, "step": 91340 }, { "grad_norm": 0.37486231327056885, "learning_rate": 2.032175087728483e-06, "loss": 0.0265, "step": 91350 }, { "grad_norm": 0.24553710222244263, "learning_rate": 2.027511669536375e-06, "loss": 0.0213, "step": 91360 }, { "grad_norm": 0.26566392183303833, "learning_rate": 2.0228534975434643e-06, "loss": 0.0241, "step": 91370 }, { "grad_norm": 0.20516784489154816, "learning_rate": 2.0182005722591548e-06, "loss": 0.0175, "step": 91380 }, { "grad_norm": 0.36155417561531067, "learning_rate": 2.0135528941922944e-06, "loss": 0.0175, "step": 91390 }, { "grad_norm": 0.25290486216545105, "learning_rate": 2.008910463851149e-06, "loss": 0.0219, "step": 91400 }, { "grad_norm": 0.4401240050792694, "learning_rate": 2.0042732817433963e-06, "loss": 0.018, "step": 91410 }, { "grad_norm": 0.4925661087036133, "learning_rate": 1.999641348376158e-06, "loss": 0.025, "step": 91420 }, { "grad_norm": 0.33280763030052185, "learning_rate": 1.9950146642559787e-06, "loss": 0.0202, "step": 91430 }, { "grad_norm": 0.2270328849554062, "learning_rate": 1.990393229888815e-06, "loss": 0.0145, "step": 91440 }, { "grad_norm": 0.35594192147254944, "learning_rate": 1.9857770457800684e-06, "loss": 0.0233, "step": 91450 }, { "grad_norm": 0.23592078685760498, "learning_rate": 1.9811661124345517e-06, "loss": 0.022, "step": 91460 }, { "grad_norm": 0.4297475814819336, "learning_rate": 1.976560430356522e-06, "loss": 0.0197, "step": 91470 }, { "grad_norm": 0.4110124707221985, "learning_rate": 1.9719600000496383e-06, "loss": 0.0225, "step": 91480 }, { "grad_norm": 0.4862442910671234, "learning_rate": 1.967364822017004e-06, "loss": 0.0244, "step": 91490 }, { "grad_norm": 0.34500741958618164, "learning_rate": 1.962774896761127e-06, "loss": 0.0186, "step": 91500 }, { "grad_norm": 0.45838093757629395, "learning_rate": 1.958190224783968e-06, "loss": 0.0212, "step": 91510 }, { "grad_norm": 0.20483747124671936, "learning_rate": 1.9536108065868975e-06, "loss": 0.0193, "step": 91520 }, { "grad_norm": 0.326711505651474, "learning_rate": 1.949036642670715e-06, "loss": 0.0207, "step": 91530 }, { "grad_norm": 0.34430062770843506, "learning_rate": 1.9444677335356377e-06, "loss": 0.0188, "step": 91540 }, { "grad_norm": 0.2934858500957489, "learning_rate": 1.939904079681326e-06, "loss": 0.0225, "step": 91550 }, { "grad_norm": 0.3483625054359436, "learning_rate": 1.935345681606837e-06, "loss": 0.0195, "step": 91560 }, { "grad_norm": 0.3174113929271698, "learning_rate": 1.930792539810694e-06, "loss": 0.0182, "step": 91570 }, { "grad_norm": 0.3471103310585022, "learning_rate": 1.9262446547907877e-06, "loss": 0.0183, "step": 91580 }, { "grad_norm": 0.45250165462493896, "learning_rate": 1.921702027044503e-06, "loss": 0.0208, "step": 91590 }, { "grad_norm": 0.8849025368690491, "learning_rate": 1.917164657068593e-06, "loss": 0.0225, "step": 91600 }, { "grad_norm": 0.3019343614578247, "learning_rate": 1.912632545359272e-06, "loss": 0.0262, "step": 91610 }, { "grad_norm": 0.4702950119972229, "learning_rate": 1.9081056924121488e-06, "loss": 0.0228, "step": 91620 }, { "grad_norm": 0.25000107288360596, "learning_rate": 1.9035840987222887e-06, "loss": 0.0183, "step": 91630 }, { "grad_norm": 0.3639707863330841, "learning_rate": 1.8990677647841515e-06, "loss": 0.0226, "step": 91640 }, { "grad_norm": 0.2294761836528778, "learning_rate": 1.8945566910916534e-06, "loss": 0.0196, "step": 91650 }, { "grad_norm": 0.2645016610622406, "learning_rate": 1.8900508781381054e-06, "loss": 0.0199, "step": 91660 }, { "grad_norm": 0.41924649477005005, "learning_rate": 1.8855503264162633e-06, "loss": 0.0162, "step": 91670 }, { "grad_norm": 0.26547062397003174, "learning_rate": 1.8810550364182998e-06, "loss": 0.0203, "step": 91680 }, { "grad_norm": 0.6441366076469421, "learning_rate": 1.8765650086358112e-06, "loss": 0.0225, "step": 91690 }, { "grad_norm": 0.3821715712547302, "learning_rate": 1.8720802435598151e-06, "loss": 0.0252, "step": 91700 }, { "grad_norm": 0.7975829839706421, "learning_rate": 1.8676007416807694e-06, "loss": 0.0193, "step": 91710 }, { "grad_norm": 0.30171942710876465, "learning_rate": 1.8631265034885437e-06, "loss": 0.0251, "step": 91720 }, { "grad_norm": 0.4720565378665924, "learning_rate": 1.858657529472424e-06, "loss": 0.0232, "step": 91730 }, { "grad_norm": 0.853946328163147, "learning_rate": 1.8541938201211473e-06, "loss": 0.022, "step": 91740 }, { "grad_norm": 0.3516612648963928, "learning_rate": 1.8497353759228398e-06, "loss": 0.0182, "step": 91750 }, { "grad_norm": 0.253991037607193, "learning_rate": 1.8452821973650835e-06, "loss": 0.0206, "step": 91760 }, { "grad_norm": 0.25355666875839233, "learning_rate": 1.840834284934856e-06, "loss": 0.0213, "step": 91770 }, { "grad_norm": 0.4381563365459442, "learning_rate": 1.8363916391185954e-06, "loss": 0.0176, "step": 91780 }, { "grad_norm": 0.43231552839279175, "learning_rate": 1.8319542604021245e-06, "loss": 0.0289, "step": 91790 }, { "grad_norm": 2.1312103271484375, "learning_rate": 1.8275221492707273e-06, "loss": 0.0196, "step": 91800 }, { "grad_norm": 0.26066693663597107, "learning_rate": 1.8230953062090717e-06, "loss": 0.0206, "step": 91810 }, { "grad_norm": 0.21434292197227478, "learning_rate": 1.8186737317012926e-06, "loss": 0.023, "step": 91820 }, { "grad_norm": 0.44300487637519836, "learning_rate": 1.8142574262308977e-06, "loss": 0.0238, "step": 91830 }, { "grad_norm": 0.430966317653656, "learning_rate": 1.8098463902808727e-06, "loss": 0.0236, "step": 91840 }, { "grad_norm": 0.40361738204956055, "learning_rate": 1.805440624333593e-06, "loss": 0.0216, "step": 91850 }, { "grad_norm": 0.25472724437713623, "learning_rate": 1.8010401288708734e-06, "loss": 0.0165, "step": 91860 }, { "grad_norm": 0.27193301916122437, "learning_rate": 1.7966449043739343e-06, "loss": 0.0185, "step": 91870 }, { "grad_norm": 0.14214733242988586, "learning_rate": 1.79225495132343e-06, "loss": 0.0164, "step": 91880 }, { "grad_norm": 0.6251235604286194, "learning_rate": 1.7878702701994487e-06, "loss": 0.0313, "step": 91890 }, { "grad_norm": 0.37675145268440247, "learning_rate": 1.783490861481496e-06, "loss": 0.0218, "step": 91900 }, { "grad_norm": 0.21236787736415863, "learning_rate": 1.7791167256484832e-06, "loss": 0.0156, "step": 91910 }, { "grad_norm": 0.4901261329650879, "learning_rate": 1.774747863178766e-06, "loss": 0.0253, "step": 91920 }, { "grad_norm": 0.44130396842956543, "learning_rate": 1.7703842745501186e-06, "loss": 0.023, "step": 91930 }, { "grad_norm": 0.5492197275161743, "learning_rate": 1.7660259602397366e-06, "loss": 0.0236, "step": 91940 }, { "grad_norm": 0.20236167311668396, "learning_rate": 1.7616729207242388e-06, "loss": 0.0205, "step": 91950 }, { "grad_norm": 0.21375679969787598, "learning_rate": 1.7573251564796556e-06, "loss": 0.0169, "step": 91960 }, { "grad_norm": 3.5429840087890625, "learning_rate": 1.7529826679814676e-06, "loss": 0.021, "step": 91970 }, { "grad_norm": 0.24958589673042297, "learning_rate": 1.7486454557045562e-06, "loss": 0.0223, "step": 91980 }, { "grad_norm": 0.6080080270767212, "learning_rate": 1.7443135201232364e-06, "loss": 0.0228, "step": 91990 }, { "grad_norm": 0.39169034361839294, "learning_rate": 1.7399868617112347e-06, "loss": 0.0231, "step": 92000 }, { "grad_norm": 0.9340218901634216, "learning_rate": 1.7356654809417171e-06, "loss": 0.0181, "step": 92010 }, { "grad_norm": 0.2585786283016205, "learning_rate": 1.7313493782872448e-06, "loss": 0.0219, "step": 92020 }, { "grad_norm": 0.6703369617462158, "learning_rate": 1.7270385542198453e-06, "loss": 0.0249, "step": 92030 }, { "grad_norm": 0.2796778082847595, "learning_rate": 1.7227330092109307e-06, "loss": 0.0189, "step": 92040 }, { "grad_norm": 0.4821307063102722, "learning_rate": 1.718432743731352e-06, "loss": 0.0272, "step": 92050 }, { "grad_norm": 0.20356285572052002, "learning_rate": 1.7141377582513718e-06, "loss": 0.0156, "step": 92060 }, { "grad_norm": 0.27681881189346313, "learning_rate": 1.709848053240698e-06, "loss": 0.0206, "step": 92070 }, { "grad_norm": 0.31657302379608154, "learning_rate": 1.7055636291684328e-06, "loss": 0.0229, "step": 92080 }, { "grad_norm": 0.6199615597724915, "learning_rate": 1.7012844865031185e-06, "loss": 0.0302, "step": 92090 }, { "grad_norm": 0.3155001401901245, "learning_rate": 1.6970106257127138e-06, "loss": 0.0299, "step": 92100 }, { "grad_norm": 0.2680329978466034, "learning_rate": 1.6927420472646117e-06, "loss": 0.0202, "step": 92110 }, { "grad_norm": 0.45354923605918884, "learning_rate": 1.6884787516256106e-06, "loss": 0.0182, "step": 92120 }, { "grad_norm": 0.7837327718734741, "learning_rate": 1.6842207392619325e-06, "loss": 0.02, "step": 92130 }, { "grad_norm": 0.2507796287536621, "learning_rate": 1.679968010639238e-06, "loss": 0.0239, "step": 92140 }, { "grad_norm": 0.2429848164319992, "learning_rate": 1.6757205662225829e-06, "loss": 0.0241, "step": 92150 }, { "grad_norm": 0.2917273938655853, "learning_rate": 1.6714784064764844e-06, "loss": 0.0258, "step": 92160 }, { "grad_norm": 0.49317505955696106, "learning_rate": 1.6672415318648383e-06, "loss": 0.0245, "step": 92170 }, { "grad_norm": 0.4801299273967743, "learning_rate": 1.6630099428510016e-06, "loss": 0.0205, "step": 92180 }, { "grad_norm": 0.22807693481445312, "learning_rate": 1.6587836398977153e-06, "loss": 0.0279, "step": 92190 }, { "grad_norm": 0.6120078563690186, "learning_rate": 1.6545626234671762e-06, "loss": 0.0264, "step": 92200 }, { "grad_norm": 0.3115580976009369, "learning_rate": 1.650346894020971e-06, "loss": 0.024, "step": 92210 }, { "grad_norm": 0.24318227171897888, "learning_rate": 1.6461364520201472e-06, "loss": 0.0196, "step": 92220 }, { "grad_norm": 0.4005368649959564, "learning_rate": 1.6419312979251366e-06, "loss": 0.031, "step": 92230 }, { "grad_norm": 0.3066485524177551, "learning_rate": 1.6377314321958214e-06, "loss": 0.0215, "step": 92240 }, { "grad_norm": 0.5500648617744446, "learning_rate": 1.6335368552914787e-06, "loss": 0.0136, "step": 92250 }, { "grad_norm": 0.18773113191127777, "learning_rate": 1.6293475676708358e-06, "loss": 0.0271, "step": 92260 }, { "grad_norm": 0.3673819899559021, "learning_rate": 1.6251635697920097e-06, "loss": 0.0351, "step": 92270 }, { "grad_norm": 0.6193400025367737, "learning_rate": 1.6209848621125624e-06, "loss": 0.0283, "step": 92280 }, { "grad_norm": 0.3749433159828186, "learning_rate": 1.6168114450894778e-06, "loss": 0.0224, "step": 92290 }, { "grad_norm": 0.6087660193443298, "learning_rate": 1.612643319179158e-06, "loss": 0.0226, "step": 92300 }, { "grad_norm": 0.41420722007751465, "learning_rate": 1.6084804848374046e-06, "loss": 0.0236, "step": 92310 }, { "grad_norm": 0.30361083149909973, "learning_rate": 1.6043229425194817e-06, "loss": 0.0211, "step": 92320 }, { "grad_norm": 0.323785662651062, "learning_rate": 1.6001706926800308e-06, "loss": 0.0246, "step": 92330 }, { "grad_norm": 0.2881772220134735, "learning_rate": 1.5960237357731444e-06, "loss": 0.0156, "step": 92340 }, { "grad_norm": 0.3555395305156708, "learning_rate": 1.5918820722523264e-06, "loss": 0.0224, "step": 92350 }, { "grad_norm": 0.43814316391944885, "learning_rate": 1.587745702570509e-06, "loss": 0.0236, "step": 92360 }, { "grad_norm": 0.3404676616191864, "learning_rate": 1.5836146271800356e-06, "loss": 0.0167, "step": 92370 }, { "grad_norm": 0.4939047396183014, "learning_rate": 1.5794888465326728e-06, "loss": 0.0191, "step": 92380 }, { "grad_norm": 0.28368136286735535, "learning_rate": 1.5753683610796155e-06, "loss": 0.0198, "step": 92390 }, { "grad_norm": 0.2966870367527008, "learning_rate": 1.5712531712714528e-06, "loss": 0.0245, "step": 92400 }, { "grad_norm": 1.1172808408737183, "learning_rate": 1.5671432775582474e-06, "loss": 0.0221, "step": 92410 }, { "grad_norm": 0.2643037438392639, "learning_rate": 1.563038680389428e-06, "loss": 0.019, "step": 92420 }, { "grad_norm": 0.4216190278530121, "learning_rate": 1.5589393802138808e-06, "loss": 0.0215, "step": 92430 }, { "grad_norm": 0.9804202914237976, "learning_rate": 1.554845377479891e-06, "loss": 0.0237, "step": 92440 }, { "grad_norm": 0.28110742568969727, "learning_rate": 1.550756672635173e-06, "loss": 0.0256, "step": 92450 }, { "grad_norm": 1.1938073635101318, "learning_rate": 1.5466732661268635e-06, "loss": 0.0227, "step": 92460 }, { "grad_norm": 0.1838352382183075, "learning_rate": 1.5425951584015163e-06, "loss": 0.0157, "step": 92470 }, { "grad_norm": 0.29874464869499207, "learning_rate": 1.5385223499051026e-06, "loss": 0.0259, "step": 92480 }, { "grad_norm": 0.3089657127857208, "learning_rate": 1.5344548410830328e-06, "loss": 0.0203, "step": 92490 }, { "grad_norm": 0.17150530219078064, "learning_rate": 1.5303926323801121e-06, "loss": 0.0232, "step": 92500 }, { "grad_norm": 0.2884633243083954, "learning_rate": 1.5263357242405795e-06, "loss": 0.0144, "step": 92510 }, { "grad_norm": 0.32408517599105835, "learning_rate": 1.5222841171080915e-06, "loss": 0.0256, "step": 92520 }, { "grad_norm": 0.2649345397949219, "learning_rate": 1.5182378114257268e-06, "loss": 0.0185, "step": 92530 }, { "grad_norm": 0.1962997019290924, "learning_rate": 1.5141968076359869e-06, "loss": 0.0183, "step": 92540 }, { "grad_norm": 0.2910466492176056, "learning_rate": 1.5101611061807908e-06, "loss": 0.0247, "step": 92550 }, { "grad_norm": 0.25641921162605286, "learning_rate": 1.506130707501463e-06, "loss": 0.0183, "step": 92560 }, { "grad_norm": 0.44308480620384216, "learning_rate": 1.5021056120387734e-06, "loss": 0.0242, "step": 92570 }, { "grad_norm": 0.2766970098018646, "learning_rate": 1.4980858202329084e-06, "loss": 0.0208, "step": 92580 }, { "grad_norm": 0.23132865130901337, "learning_rate": 1.4940713325234445e-06, "loss": 0.0223, "step": 92590 }, { "grad_norm": 0.15520524978637695, "learning_rate": 1.4900621493494137e-06, "loss": 0.0156, "step": 92600 }, { "grad_norm": 0.4708911180496216, "learning_rate": 1.4860582711492544e-06, "loss": 0.023, "step": 92610 }, { "grad_norm": 0.3534536361694336, "learning_rate": 1.4820596983608325e-06, "loss": 0.0201, "step": 92620 }, { "grad_norm": 0.2504189610481262, "learning_rate": 1.4780664314214044e-06, "loss": 0.0241, "step": 92630 }, { "grad_norm": 0.28198274970054626, "learning_rate": 1.474078470767687e-06, "loss": 0.0224, "step": 92640 }, { "grad_norm": 0.37904566526412964, "learning_rate": 1.470095816835787e-06, "loss": 0.0211, "step": 92650 }, { "grad_norm": 0.279100239276886, "learning_rate": 1.4661184700612395e-06, "loss": 0.023, "step": 92660 }, { "grad_norm": 0.3908618688583374, "learning_rate": 1.4621464308790133e-06, "loss": 0.0272, "step": 92670 }, { "grad_norm": 0.24513834714889526, "learning_rate": 1.4581796997234775e-06, "loss": 0.0227, "step": 92680 }, { "grad_norm": 0.3467278480529785, "learning_rate": 1.4542182770284297e-06, "loss": 0.0206, "step": 92690 }, { "grad_norm": 0.2877286970615387, "learning_rate": 1.450262163227084e-06, "loss": 0.0183, "step": 92700 }, { "grad_norm": 0.3742713928222656, "learning_rate": 1.4463113587520671e-06, "loss": 0.0256, "step": 92710 }, { "grad_norm": 0.18993864953517914, "learning_rate": 1.4423658640354443e-06, "loss": 0.0215, "step": 92720 }, { "grad_norm": 0.27245616912841797, "learning_rate": 1.438425679508687e-06, "loss": 0.0287, "step": 92730 }, { "grad_norm": 0.42377692461013794, "learning_rate": 1.4344908056026895e-06, "loss": 0.0238, "step": 92740 }, { "grad_norm": 0.2091846466064453, "learning_rate": 1.4305612427477521e-06, "loss": 0.0212, "step": 92750 }, { "grad_norm": 0.5530288219451904, "learning_rate": 1.42663699137362e-06, "loss": 0.0206, "step": 92760 }, { "grad_norm": 0.17503595352172852, "learning_rate": 1.4227180519094386e-06, "loss": 0.0221, "step": 92770 }, { "grad_norm": 0.6863117814064026, "learning_rate": 1.4188044247837706e-06, "loss": 0.0245, "step": 92780 }, { "grad_norm": 0.2892013192176819, "learning_rate": 1.4148961104246128e-06, "loss": 0.0235, "step": 92790 }, { "grad_norm": 1.0849921703338623, "learning_rate": 1.4109931092593731e-06, "loss": 0.0191, "step": 92800 }, { "grad_norm": 0.5942007899284363, "learning_rate": 1.4070954217148658e-06, "loss": 0.0179, "step": 92810 }, { "grad_norm": 0.40850719809532166, "learning_rate": 1.4032030482173497e-06, "loss": 0.0273, "step": 92820 }, { "grad_norm": 0.3040021061897278, "learning_rate": 1.3993159891924901e-06, "loss": 0.0222, "step": 92830 }, { "grad_norm": 0.25636687874794006, "learning_rate": 1.3954342450653524e-06, "loss": 0.0281, "step": 92840 }, { "grad_norm": 0.3225899338722229, "learning_rate": 1.3915578162604582e-06, "loss": 0.016, "step": 92850 }, { "grad_norm": 0.13580483198165894, "learning_rate": 1.3876867032017126e-06, "loss": 0.0208, "step": 92860 }, { "grad_norm": 0.45053109526634216, "learning_rate": 1.383820906312472e-06, "loss": 0.0201, "step": 92870 }, { "grad_norm": 0.40827852487564087, "learning_rate": 1.3799604260154752e-06, "loss": 0.0226, "step": 92880 }, { "grad_norm": 0.3702143728733063, "learning_rate": 1.376105262732913e-06, "loss": 0.0336, "step": 92890 }, { "grad_norm": 0.46084004640579224, "learning_rate": 1.37225541688637e-06, "loss": 0.0227, "step": 92900 }, { "grad_norm": 0.34917694330215454, "learning_rate": 1.3684108888968594e-06, "loss": 0.0187, "step": 92910 }, { "grad_norm": 0.5902806520462036, "learning_rate": 1.364571679184823e-06, "loss": 0.0222, "step": 92920 }, { "grad_norm": 0.4081784784793854, "learning_rate": 1.3607377881701077e-06, "loss": 0.0299, "step": 92930 }, { "grad_norm": 0.24611712992191315, "learning_rate": 1.3569092162719787e-06, "loss": 0.0165, "step": 92940 }, { "grad_norm": 0.33305618166923523, "learning_rate": 1.353085963909123e-06, "loss": 0.0246, "step": 92950 }, { "grad_norm": 0.2750171422958374, "learning_rate": 1.3492680314996454e-06, "loss": 0.0231, "step": 92960 }, { "grad_norm": 0.852009117603302, "learning_rate": 1.345455419461078e-06, "loss": 0.0331, "step": 92970 }, { "grad_norm": 0.41974136233329773, "learning_rate": 1.3416481282103431e-06, "loss": 0.0295, "step": 92980 }, { "grad_norm": 0.3631308972835541, "learning_rate": 1.337846158163819e-06, "loss": 0.0196, "step": 92990 }, { "grad_norm": 0.4235677719116211, "learning_rate": 1.334049509737273e-06, "loss": 0.0278, "step": 93000 }, { "grad_norm": 0.30467477440834045, "learning_rate": 1.3302581833459116e-06, "loss": 0.0256, "step": 93010 }, { "grad_norm": 0.4868203401565552, "learning_rate": 1.3264721794043366e-06, "loss": 0.0169, "step": 93020 }, { "grad_norm": 0.23257605731487274, "learning_rate": 1.3226914983265836e-06, "loss": 0.025, "step": 93030 }, { "grad_norm": 0.3903437554836273, "learning_rate": 1.3189161405260997e-06, "loss": 0.0244, "step": 93040 }, { "grad_norm": 2.1234195232391357, "learning_rate": 1.3151461064157656e-06, "loss": 0.0243, "step": 93050 }, { "grad_norm": 0.2119276374578476, "learning_rate": 1.3113813964078459e-06, "loss": 0.0206, "step": 93060 }, { "grad_norm": 0.3437754809856415, "learning_rate": 1.3076220109140614e-06, "loss": 0.0235, "step": 93070 }, { "grad_norm": 0.17029951512813568, "learning_rate": 1.3038679503455275e-06, "loss": 0.0184, "step": 93080 }, { "grad_norm": 0.39732199907302856, "learning_rate": 1.3001192151127773e-06, "loss": 0.0303, "step": 93090 }, { "grad_norm": 0.14874805510044098, "learning_rate": 1.2963758056257657e-06, "loss": 0.0212, "step": 93100 }, { "grad_norm": 0.7070385813713074, "learning_rate": 1.2926377222938768e-06, "loss": 0.0227, "step": 93110 }, { "grad_norm": 0.4220486283302307, "learning_rate": 1.2889049655259e-06, "loss": 0.0353, "step": 93120 }, { "grad_norm": 0.3381383419036865, "learning_rate": 1.2851775357300312e-06, "loss": 0.0197, "step": 93130 }, { "grad_norm": 0.2771449387073517, "learning_rate": 1.281455433313916e-06, "loss": 0.0244, "step": 93140 }, { "grad_norm": 1.086598515510559, "learning_rate": 1.2777386586845796e-06, "loss": 0.0213, "step": 93150 }, { "grad_norm": 0.2411990612745285, "learning_rate": 1.2740272122484964e-06, "loss": 0.0243, "step": 93160 }, { "grad_norm": 0.4675484299659729, "learning_rate": 1.2703210944115308e-06, "loss": 0.0183, "step": 93170 }, { "grad_norm": 0.38513246178627014, "learning_rate": 1.2666203055789916e-06, "loss": 0.0192, "step": 93180 }, { "grad_norm": 0.2800614535808563, "learning_rate": 1.2629248461555832e-06, "loss": 0.0181, "step": 93190 }, { "grad_norm": 0.353718638420105, "learning_rate": 1.2592347165454431e-06, "loss": 0.0177, "step": 93200 }, { "grad_norm": 0.2185467928647995, "learning_rate": 1.2555499171521102e-06, "loss": 0.0197, "step": 93210 }, { "grad_norm": 0.23096239566802979, "learning_rate": 1.251870448378556e-06, "loss": 0.023, "step": 93220 }, { "grad_norm": 0.1910897046327591, "learning_rate": 1.2481963106271478e-06, "loss": 0.0223, "step": 93230 }, { "grad_norm": 0.9041979312896729, "learning_rate": 1.2445275042997084e-06, "loss": 0.024, "step": 93240 }, { "grad_norm": 0.32239213585853577, "learning_rate": 1.2408640297974283e-06, "loss": 0.0192, "step": 93250 }, { "grad_norm": 0.4032647907733917, "learning_rate": 1.2372058875209535e-06, "loss": 0.0237, "step": 93260 }, { "grad_norm": 0.30980831384658813, "learning_rate": 1.2335530778703252e-06, "loss": 0.017, "step": 93270 }, { "grad_norm": 0.3200336992740631, "learning_rate": 1.2299056012450182e-06, "loss": 0.0221, "step": 93280 }, { "grad_norm": 0.3619575500488281, "learning_rate": 1.2262634580439026e-06, "loss": 0.0194, "step": 93290 }, { "grad_norm": 0.3668782114982605, "learning_rate": 1.2226266486652926e-06, "loss": 0.023, "step": 93300 }, { "grad_norm": 0.2247762531042099, "learning_rate": 1.2189951735068927e-06, "loss": 0.0143, "step": 93310 }, { "grad_norm": 0.391368567943573, "learning_rate": 1.2153690329658407e-06, "loss": 0.0186, "step": 93320 }, { "grad_norm": 0.2293420135974884, "learning_rate": 1.2117482274386859e-06, "loss": 0.0265, "step": 93330 }, { "grad_norm": 0.19962039589881897, "learning_rate": 1.2081327573213953e-06, "loss": 0.0216, "step": 93340 }, { "grad_norm": 0.3711429536342621, "learning_rate": 1.2045226230093465e-06, "loss": 0.0228, "step": 93350 }, { "grad_norm": 0.33822035789489746, "learning_rate": 1.2009178248973296e-06, "loss": 0.0177, "step": 93360 }, { "grad_norm": 0.3431788980960846, "learning_rate": 1.1973183633795849e-06, "loss": 0.0226, "step": 93370 }, { "grad_norm": 0.306738942861557, "learning_rate": 1.193724238849725e-06, "loss": 0.0197, "step": 93380 }, { "grad_norm": 0.30473455786705017, "learning_rate": 1.190135451700808e-06, "loss": 0.0212, "step": 93390 }, { "grad_norm": 0.32743266224861145, "learning_rate": 1.186552002325292e-06, "loss": 0.0281, "step": 93400 }, { "grad_norm": 0.3566511273384094, "learning_rate": 1.1829738911150578e-06, "loss": 0.0215, "step": 93410 }, { "grad_norm": 0.25041067600250244, "learning_rate": 1.1794011184613985e-06, "loss": 0.0174, "step": 93420 }, { "grad_norm": 0.46629399061203003, "learning_rate": 1.17583368475504e-06, "loss": 0.0247, "step": 93430 }, { "grad_norm": 0.3759061396121979, "learning_rate": 1.1722715903860982e-06, "loss": 0.0202, "step": 93440 }, { "grad_norm": 0.28270554542541504, "learning_rate": 1.168714835744128e-06, "loss": 0.0163, "step": 93450 }, { "grad_norm": 0.40479767322540283, "learning_rate": 1.165163421218085e-06, "loss": 0.029, "step": 93460 }, { "grad_norm": 0.3531205654144287, "learning_rate": 1.1616173471963477e-06, "loss": 0.0219, "step": 93470 }, { "grad_norm": 0.343466579914093, "learning_rate": 1.1580766140667109e-06, "loss": 0.0186, "step": 93480 }, { "grad_norm": 0.16580888628959656, "learning_rate": 1.1545412222163764e-06, "loss": 0.0251, "step": 93490 }, { "grad_norm": 0.3167003095149994, "learning_rate": 1.1510111720319795e-06, "loss": 0.0246, "step": 93500 }, { "grad_norm": 0.40582555532455444, "learning_rate": 1.1474864638995663e-06, "loss": 0.017, "step": 93510 }, { "grad_norm": 0.2086162567138672, "learning_rate": 1.1439670982045735e-06, "loss": 0.0209, "step": 93520 }, { "grad_norm": 0.33455702662467957, "learning_rate": 1.1404530753318877e-06, "loss": 0.0303, "step": 93530 }, { "grad_norm": 0.2851213216781616, "learning_rate": 1.136944395665801e-06, "loss": 0.0171, "step": 93540 }, { "grad_norm": 0.42030102014541626, "learning_rate": 1.1334410595900014e-06, "loss": 0.016, "step": 93550 }, { "grad_norm": 0.6057742238044739, "learning_rate": 1.129943067487621e-06, "loss": 0.0282, "step": 93560 }, { "grad_norm": 0.2417818158864975, "learning_rate": 1.126450419741193e-06, "loss": 0.0277, "step": 93570 }, { "grad_norm": 0.4571234881877899, "learning_rate": 1.1229631167326671e-06, "loss": 0.0232, "step": 93580 }, { "grad_norm": 0.4219551980495453, "learning_rate": 1.1194811588434052e-06, "loss": 0.0193, "step": 93590 }, { "grad_norm": 0.4404623210430145, "learning_rate": 1.1160045464542023e-06, "loss": 0.0257, "step": 93600 }, { "grad_norm": 0.39936086535453796, "learning_rate": 1.1125332799452325e-06, "loss": 0.0269, "step": 93610 }, { "grad_norm": 0.26264867186546326, "learning_rate": 1.109067359696131e-06, "loss": 0.017, "step": 93620 }, { "grad_norm": 0.2616293430328369, "learning_rate": 1.1056067860859166e-06, "loss": 0.0217, "step": 93630 }, { "grad_norm": 0.5337089896202087, "learning_rate": 1.102151559493031e-06, "loss": 0.019, "step": 93640 }, { "grad_norm": 0.37424516677856445, "learning_rate": 1.0987016802953332e-06, "loss": 0.0203, "step": 93650 }, { "grad_norm": 0.30624115467071533, "learning_rate": 1.0952571488701048e-06, "loss": 0.0157, "step": 93660 }, { "grad_norm": 0.6773447394371033, "learning_rate": 1.0918179655940163e-06, "loss": 0.0265, "step": 93670 }, { "grad_norm": 0.3574538826942444, "learning_rate": 1.088384130843184e-06, "loss": 0.0215, "step": 93680 }, { "grad_norm": 0.3037741482257843, "learning_rate": 1.0849556449931296e-06, "loss": 0.0317, "step": 93690 }, { "grad_norm": 0.24827469885349274, "learning_rate": 1.0815325084187866e-06, "loss": 0.022, "step": 93700 }, { "grad_norm": 0.3887416422367096, "learning_rate": 1.0781147214944943e-06, "loss": 0.019, "step": 93710 }, { "grad_norm": 0.6741865277290344, "learning_rate": 1.0747022845940313e-06, "loss": 0.0137, "step": 93720 }, { "grad_norm": 0.4073214828968048, "learning_rate": 1.0712951980905605e-06, "loss": 0.0167, "step": 93730 }, { "grad_norm": 0.43472492694854736, "learning_rate": 1.0678934623566839e-06, "loss": 0.0219, "step": 93740 }, { "grad_norm": 0.38073527812957764, "learning_rate": 1.0644970777644091e-06, "loss": 0.0158, "step": 93750 }, { "grad_norm": 1.3213433027267456, "learning_rate": 1.0611060446851673e-06, "loss": 0.0167, "step": 93760 }, { "grad_norm": 0.45515626668930054, "learning_rate": 1.0577203634897836e-06, "loss": 0.0288, "step": 93770 }, { "grad_norm": 0.4726531505584717, "learning_rate": 1.0543400345485178e-06, "loss": 0.0229, "step": 93780 }, { "grad_norm": 0.33188924193382263, "learning_rate": 1.0509650582310404e-06, "loss": 0.0193, "step": 93790 }, { "grad_norm": 0.19289974868297577, "learning_rate": 1.0475954349064287e-06, "loss": 0.0197, "step": 93800 }, { "grad_norm": 0.2587697505950928, "learning_rate": 1.0442311649431825e-06, "loss": 0.0192, "step": 93810 }, { "grad_norm": 0.1867600828409195, "learning_rate": 1.0408722487092071e-06, "loss": 0.0224, "step": 93820 }, { "grad_norm": 0.4760687053203583, "learning_rate": 1.0375186865718424e-06, "loss": 0.0234, "step": 93830 }, { "grad_norm": 0.30735689401626587, "learning_rate": 1.0341704788978113e-06, "loss": 0.0224, "step": 93840 }, { "grad_norm": 0.7374832630157471, "learning_rate": 1.0308276260532879e-06, "loss": 0.0324, "step": 93850 }, { "grad_norm": 0.2958689033985138, "learning_rate": 1.0274901284038241e-06, "loss": 0.0224, "step": 93860 }, { "grad_norm": 0.23096820712089539, "learning_rate": 1.0241579863144114e-06, "loss": 0.0206, "step": 93870 }, { "grad_norm": 0.35683268308639526, "learning_rate": 1.0208312001494413e-06, "loss": 0.026, "step": 93880 }, { "grad_norm": 0.26233574748039246, "learning_rate": 1.0175097702727453e-06, "loss": 0.0282, "step": 93890 }, { "grad_norm": 0.41615021228790283, "learning_rate": 1.0141936970475275e-06, "loss": 0.024, "step": 93900 }, { "grad_norm": 0.39005598425865173, "learning_rate": 1.0108829808364417e-06, "loss": 0.0194, "step": 93910 }, { "grad_norm": 0.31744444370269775, "learning_rate": 1.0075776220015376e-06, "loss": 0.0228, "step": 93920 }, { "grad_norm": 0.3961411118507385, "learning_rate": 1.0042776209042816e-06, "loss": 0.0223, "step": 93930 }, { "grad_norm": 0.251995325088501, "learning_rate": 1.000982977905568e-06, "loss": 0.0182, "step": 93940 }, { "grad_norm": 0.2844749987125397, "learning_rate": 9.976936933656867e-07, "loss": 0.0181, "step": 93950 }, { "grad_norm": 0.4435306489467621, "learning_rate": 9.944097676443442e-07, "loss": 0.0197, "step": 93960 }, { "grad_norm": 0.4126073122024536, "learning_rate": 9.911312011006757e-07, "loss": 0.0163, "step": 93970 }, { "grad_norm": 0.38967177271842957, "learning_rate": 9.878579940932109e-07, "loss": 0.0225, "step": 93980 }, { "grad_norm": 0.2361895591020584, "learning_rate": 9.84590146979908e-07, "loss": 0.0156, "step": 93990 }, { "grad_norm": 0.1795324683189392, "learning_rate": 9.813276601181365e-07, "loss": 0.0188, "step": 94000 }, { "grad_norm": 0.4210875332355499, "learning_rate": 9.780705338646724e-07, "loss": 0.0208, "step": 94010 }, { "grad_norm": 0.3429895043373108, "learning_rate": 9.748187685757083e-07, "loss": 0.0227, "step": 94020 }, { "grad_norm": 0.7609484791755676, "learning_rate": 9.715723646068543e-07, "loss": 0.0154, "step": 94030 }, { "grad_norm": 0.8671625852584839, "learning_rate": 9.68331322313143e-07, "loss": 0.0172, "step": 94040 }, { "grad_norm": 0.26834577322006226, "learning_rate": 9.650956420489909e-07, "loss": 0.0195, "step": 94050 }, { "grad_norm": 0.5057996511459351, "learning_rate": 9.618653241682595e-07, "loss": 0.0195, "step": 94060 }, { "grad_norm": 0.5606022477149963, "learning_rate": 9.586403690242053e-07, "loss": 0.0257, "step": 94070 }, { "grad_norm": 0.3104914724826813, "learning_rate": 9.554207769695122e-07, "loss": 0.0195, "step": 94080 }, { "grad_norm": 0.533391535282135, "learning_rate": 9.522065483562659e-07, "loss": 0.0163, "step": 94090 }, { "grad_norm": 0.8248297572135925, "learning_rate": 9.489976835359681e-07, "loss": 0.0178, "step": 94100 }, { "grad_norm": 0.3607635796070099, "learning_rate": 9.457941828595329e-07, "loss": 0.0197, "step": 94110 }, { "grad_norm": 0.3030707538127899, "learning_rate": 9.425960466772965e-07, "loss": 0.0196, "step": 94120 }, { "grad_norm": 0.4630362093448639, "learning_rate": 9.394032753390014e-07, "loss": 0.0219, "step": 94130 }, { "grad_norm": 0.32883206009864807, "learning_rate": 9.362158691938017e-07, "loss": 0.0228, "step": 94140 }, { "grad_norm": 0.272992879152298, "learning_rate": 9.330338285902685e-07, "loss": 0.0178, "step": 94150 }, { "grad_norm": 0.1658896803855896, "learning_rate": 9.298571538763845e-07, "loss": 0.0201, "step": 94160 }, { "grad_norm": 0.19767138361930847, "learning_rate": 9.266858453995441e-07, "loss": 0.0209, "step": 94170 }, { "grad_norm": 0.2729325294494629, "learning_rate": 9.235199035065645e-07, "loss": 0.0213, "step": 94180 }, { "grad_norm": 0.22178639471530914, "learning_rate": 9.203593285436574e-07, "loss": 0.0176, "step": 94190 }, { "grad_norm": 0.37795788049697876, "learning_rate": 9.172041208564686e-07, "loss": 0.0199, "step": 94200 }, { "grad_norm": 0.174735426902771, "learning_rate": 9.140542807900332e-07, "loss": 0.0199, "step": 94210 }, { "grad_norm": 0.5405455827713013, "learning_rate": 9.109098086888313e-07, "loss": 0.02, "step": 94220 }, { "grad_norm": 0.3731677532196045, "learning_rate": 9.077707048967265e-07, "loss": 0.0217, "step": 94230 }, { "grad_norm": 0.44885241985321045, "learning_rate": 9.046369697570057e-07, "loss": 0.0244, "step": 94240 }, { "grad_norm": 0.3875609338283539, "learning_rate": 9.015086036123721e-07, "loss": 0.0237, "step": 94250 }, { "grad_norm": 0.286329060792923, "learning_rate": 8.983856068049413e-07, "loss": 0.0218, "step": 94260 }, { "grad_norm": 0.2608638107776642, "learning_rate": 8.952679796762398e-07, "loss": 0.0185, "step": 94270 }, { "grad_norm": 0.42371392250061035, "learning_rate": 8.921557225671951e-07, "loss": 0.0238, "step": 94280 }, { "grad_norm": 0.1806168109178543, "learning_rate": 8.890488358181792e-07, "loss": 0.0216, "step": 94290 }, { "grad_norm": 0.3388921022415161, "learning_rate": 8.859473197689372e-07, "loss": 0.014, "step": 94300 }, { "grad_norm": 0.42813676595687866, "learning_rate": 8.828511747586588e-07, "loss": 0.0243, "step": 94310 }, { "grad_norm": 0.6014139652252197, "learning_rate": 8.797604011259287e-07, "loss": 0.0219, "step": 94320 }, { "grad_norm": 0.7548184394836426, "learning_rate": 8.766749992087487e-07, "loss": 0.0222, "step": 94330 }, { "grad_norm": 0.28444570302963257, "learning_rate": 8.735949693445378e-07, "loss": 0.021, "step": 94340 }, { "grad_norm": 0.3449234664440155, "learning_rate": 8.705203118701211e-07, "loss": 0.0147, "step": 94350 }, { "grad_norm": 0.49097055196762085, "learning_rate": 8.674510271217295e-07, "loss": 0.0196, "step": 94360 }, { "grad_norm": 0.2083168923854828, "learning_rate": 8.643871154350336e-07, "loss": 0.0215, "step": 94370 }, { "grad_norm": 0.22048628330230713, "learning_rate": 8.613285771450818e-07, "loss": 0.0199, "step": 94380 }, { "grad_norm": 0.1705775260925293, "learning_rate": 8.582754125863623e-07, "loss": 0.0143, "step": 94390 }, { "grad_norm": 0.35420718789100647, "learning_rate": 8.55227622092758e-07, "loss": 0.021, "step": 94400 }, { "grad_norm": 0.18579189479351044, "learning_rate": 8.521852059975743e-07, "loss": 0.0251, "step": 94410 }, { "grad_norm": 0.4967942237854004, "learning_rate": 8.491481646335231e-07, "loss": 0.0198, "step": 94420 }, { "grad_norm": 0.4191635251045227, "learning_rate": 8.46116498332733e-07, "loss": 0.021, "step": 94430 }, { "grad_norm": 0.4037286639213562, "learning_rate": 8.430902074267333e-07, "loss": 0.0224, "step": 94440 }, { "grad_norm": 0.39242151379585266, "learning_rate": 8.400692922464925e-07, "loss": 0.0193, "step": 94450 }, { "grad_norm": 0.5389981269836426, "learning_rate": 8.370537531223577e-07, "loss": 0.0208, "step": 94460 }, { "grad_norm": 0.28952088952064514, "learning_rate": 8.340435903841092e-07, "loss": 0.016, "step": 94470 }, { "grad_norm": 0.2989135682582855, "learning_rate": 8.310388043609341e-07, "loss": 0.0192, "step": 94480 }, { "grad_norm": 0.28299933671951294, "learning_rate": 8.280393953814303e-07, "loss": 0.0175, "step": 94490 }, { "grad_norm": 0.4197347164154053, "learning_rate": 8.250453637736077e-07, "loss": 0.0189, "step": 94500 }, { "grad_norm": 0.17654432356357574, "learning_rate": 8.22056709864899e-07, "loss": 0.0243, "step": 94510 }, { "grad_norm": 0.3164117634296417, "learning_rate": 8.190734339821204e-07, "loss": 0.0182, "step": 94520 }, { "grad_norm": 0.30993983149528503, "learning_rate": 8.160955364515333e-07, "loss": 0.0197, "step": 94530 }, { "grad_norm": 0.32793301343917847, "learning_rate": 8.131230175987936e-07, "loss": 0.0248, "step": 94540 }, { "grad_norm": 0.3265828490257263, "learning_rate": 8.101558777489693e-07, "loss": 0.0254, "step": 94550 }, { "grad_norm": 0.31206032633781433, "learning_rate": 8.071941172265396e-07, "loss": 0.033, "step": 94560 }, { "grad_norm": 0.8999521732330322, "learning_rate": 8.042377363553955e-07, "loss": 0.0191, "step": 94570 }, { "grad_norm": 0.2604665160179138, "learning_rate": 8.01286735458856e-07, "loss": 0.021, "step": 94580 }, { "grad_norm": 0.337358295917511, "learning_rate": 7.983411148596298e-07, "loss": 0.0208, "step": 94590 }, { "grad_norm": 0.22988659143447876, "learning_rate": 7.954008748798425e-07, "loss": 0.0239, "step": 94600 }, { "grad_norm": 0.2671266496181488, "learning_rate": 7.924660158410369e-07, "loss": 0.027, "step": 94610 }, { "grad_norm": 0.35756394267082214, "learning_rate": 7.895365380641728e-07, "loss": 0.0276, "step": 94620 }, { "grad_norm": 0.32951590418815613, "learning_rate": 7.866124418695941e-07, "loss": 0.019, "step": 94630 }, { "grad_norm": 0.28281068801879883, "learning_rate": 7.83693727577095e-07, "loss": 0.0177, "step": 94640 }, { "grad_norm": 0.2964169383049011, "learning_rate": 7.807803955058535e-07, "loss": 0.0229, "step": 94650 }, { "grad_norm": 0.6562440991401672, "learning_rate": 7.778724459744702e-07, "loss": 0.0193, "step": 94660 }, { "grad_norm": 0.331030011177063, "learning_rate": 7.749698793009464e-07, "loss": 0.022, "step": 94670 }, { "grad_norm": 0.42884597182273865, "learning_rate": 7.720726958027169e-07, "loss": 0.0231, "step": 94680 }, { "grad_norm": 1.3002386093139648, "learning_rate": 7.691808957965951e-07, "loss": 0.0184, "step": 94690 }, { "grad_norm": 0.211996927857399, "learning_rate": 7.662944795988336e-07, "loss": 0.0281, "step": 94700 }, { "grad_norm": 0.23296694457530975, "learning_rate": 7.634134475250854e-07, "loss": 0.0238, "step": 94710 }, { "grad_norm": 0.16603928804397583, "learning_rate": 7.605377998904262e-07, "loss": 0.0232, "step": 94720 }, { "grad_norm": 0.25445982813835144, "learning_rate": 7.5766753700931e-07, "loss": 0.0149, "step": 94730 }, { "grad_norm": 1.3871275186538696, "learning_rate": 7.54802659195647e-07, "loss": 0.0164, "step": 94740 }, { "grad_norm": 0.3427852988243103, "learning_rate": 7.519431667627253e-07, "loss": 0.0185, "step": 94750 }, { "grad_norm": 0.3449583649635315, "learning_rate": 7.490890600232503e-07, "loss": 0.0216, "step": 94760 }, { "grad_norm": 1.280703067779541, "learning_rate": 7.462403392893502e-07, "loss": 0.0281, "step": 94770 }, { "grad_norm": 0.21702876687049866, "learning_rate": 7.433970048725536e-07, "loss": 0.0213, "step": 94780 }, { "grad_norm": 0.15856972336769104, "learning_rate": 7.405590570838062e-07, "loss": 0.0177, "step": 94790 }, { "grad_norm": 0.1479697972536087, "learning_rate": 7.377264962334596e-07, "loss": 0.0174, "step": 94800 }, { "grad_norm": 0.2561832368373871, "learning_rate": 7.348993226312773e-07, "loss": 0.0207, "step": 94810 }, { "grad_norm": 0.5008316040039062, "learning_rate": 7.32077536586434e-07, "loss": 0.0184, "step": 94820 }, { "grad_norm": 0.25665172934532166, "learning_rate": 7.292611384075221e-07, "loss": 0.0198, "step": 94830 }, { "grad_norm": 0.401284396648407, "learning_rate": 7.264501284025338e-07, "loss": 0.0171, "step": 94840 }, { "grad_norm": 0.3686458468437195, "learning_rate": 7.236445068788789e-07, "loss": 0.0239, "step": 94850 }, { "grad_norm": 0.9259354472160339, "learning_rate": 7.20844274143373e-07, "loss": 0.0218, "step": 94860 }, { "grad_norm": 0.2657887935638428, "learning_rate": 7.180494305022489e-07, "loss": 0.0248, "step": 94870 }, { "grad_norm": 0.3700140118598938, "learning_rate": 7.152599762611456e-07, "loss": 0.0213, "step": 94880 }, { "grad_norm": 0.6845199465751648, "learning_rate": 7.124759117251079e-07, "loss": 0.0452, "step": 94890 }, { "grad_norm": 0.24646688997745514, "learning_rate": 7.096972371986033e-07, "loss": 0.0146, "step": 94900 }, { "grad_norm": 0.3050723075866699, "learning_rate": 7.069239529855054e-07, "loss": 0.0273, "step": 94910 }, { "grad_norm": 0.5112259387969971, "learning_rate": 7.041560593890939e-07, "loss": 0.0211, "step": 94920 }, { "grad_norm": 0.21120893955230713, "learning_rate": 7.013935567120599e-07, "loss": 0.0196, "step": 94930 }, { "grad_norm": 1.3560882806777954, "learning_rate": 6.986364452565009e-07, "loss": 0.0186, "step": 94940 }, { "grad_norm": 1.0329840183258057, "learning_rate": 6.95884725323942e-07, "loss": 0.0203, "step": 94950 }, { "grad_norm": 0.36310887336730957, "learning_rate": 6.931383972152982e-07, "loss": 0.0227, "step": 94960 }, { "grad_norm": 0.27081045508384705, "learning_rate": 6.903974612309127e-07, "loss": 0.0222, "step": 94970 }, { "grad_norm": 1.3519022464752197, "learning_rate": 6.876619176705179e-07, "loss": 0.0246, "step": 94980 }, { "grad_norm": 0.46678662300109863, "learning_rate": 6.8493176683328e-07, "loss": 0.0202, "step": 94990 }, { "grad_norm": 0.1969621181488037, "learning_rate": 6.822070090177657e-07, "loss": 0.0264, "step": 95000 }, { "grad_norm": 0.3826504647731781, "learning_rate": 6.794876445219367e-07, "loss": 0.0237, "step": 95010 }, { "grad_norm": 0.33513569831848145, "learning_rate": 6.767736736431885e-07, "loss": 0.0276, "step": 95020 }, { "grad_norm": 0.4698536694049835, "learning_rate": 6.740650966783113e-07, "loss": 0.0268, "step": 95030 }, { "grad_norm": 0.316509872674942, "learning_rate": 6.713619139235239e-07, "loss": 0.02, "step": 95040 }, { "grad_norm": 0.49414992332458496, "learning_rate": 6.686641256744286e-07, "loss": 0.0242, "step": 95050 }, { "grad_norm": 0.2202584445476532, "learning_rate": 6.659717322260562e-07, "loss": 0.0269, "step": 95060 }, { "grad_norm": 0.21658822894096375, "learning_rate": 6.632847338728377e-07, "loss": 0.0247, "step": 95070 }, { "grad_norm": 0.22613583505153656, "learning_rate": 6.60603130908627e-07, "loss": 0.0186, "step": 95080 }, { "grad_norm": 0.7316327691078186, "learning_rate": 6.579269236266783e-07, "loss": 0.0204, "step": 95090 }, { "grad_norm": 0.33315756916999817, "learning_rate": 6.552561123196632e-07, "loss": 0.0221, "step": 95100 }, { "grad_norm": 0.1631304919719696, "learning_rate": 6.525906972796425e-07, "loss": 0.0214, "step": 95110 }, { "grad_norm": 0.43702036142349243, "learning_rate": 6.499306787981108e-07, "loss": 0.0198, "step": 95120 }, { "grad_norm": 0.436669260263443, "learning_rate": 6.472760571659631e-07, "loss": 0.0276, "step": 95130 }, { "grad_norm": 0.4113892912864685, "learning_rate": 6.446268326735061e-07, "loss": 0.0181, "step": 95140 }, { "grad_norm": 0.2759357690811157, "learning_rate": 6.41983005610458e-07, "loss": 0.0265, "step": 95150 }, { "grad_norm": 0.20380225777626038, "learning_rate": 6.393445762659378e-07, "loss": 0.0186, "step": 95160 }, { "grad_norm": 0.36284345388412476, "learning_rate": 6.36711544928481e-07, "loss": 0.0166, "step": 95170 }, { "grad_norm": 0.169364333152771, "learning_rate": 6.340839118860298e-07, "loss": 0.029, "step": 95180 }, { "grad_norm": 0.28208595514297485, "learning_rate": 6.314616774259485e-07, "loss": 0.0164, "step": 95190 }, { "grad_norm": 0.46477726101875305, "learning_rate": 6.288448418349858e-07, "loss": 0.023, "step": 95200 }, { "grad_norm": 0.6710760593414307, "learning_rate": 6.262334053993235e-07, "loss": 0.0224, "step": 95210 }, { "grad_norm": 0.4761621952056885, "learning_rate": 6.236273684045501e-07, "loss": 0.0152, "step": 95220 }, { "grad_norm": 0.3809766471385956, "learning_rate": 6.210267311356488e-07, "loss": 0.0191, "step": 95230 }, { "grad_norm": 0.3188821077346802, "learning_rate": 6.184314938770197e-07, "loss": 0.0236, "step": 95240 }, { "grad_norm": 0.3198756277561188, "learning_rate": 6.158416569124858e-07, "loss": 0.0138, "step": 95250 }, { "grad_norm": 1.7537797689437866, "learning_rate": 6.132572205252541e-07, "loss": 0.0226, "step": 95260 }, { "grad_norm": 0.29540327191352844, "learning_rate": 6.106781849979648e-07, "loss": 0.0259, "step": 95270 }, { "grad_norm": 0.47123122215270996, "learning_rate": 6.081045506126482e-07, "loss": 0.0259, "step": 95280 }, { "grad_norm": 0.3844277560710907, "learning_rate": 6.055363176507678e-07, "loss": 0.0233, "step": 95290 }, { "grad_norm": 0.865249752998352, "learning_rate": 6.029734863931602e-07, "loss": 0.0215, "step": 95300 }, { "grad_norm": 0.3000594675540924, "learning_rate": 6.004160571201179e-07, "loss": 0.0186, "step": 95310 }, { "grad_norm": 0.32554060220718384, "learning_rate": 5.978640301112947e-07, "loss": 0.0231, "step": 95320 }, { "grad_norm": 0.24072906374931335, "learning_rate": 5.9531740564579e-07, "loss": 0.0188, "step": 95330 }, { "grad_norm": 0.3088212013244629, "learning_rate": 5.927761840020918e-07, "loss": 0.0178, "step": 95340 }, { "grad_norm": 0.2588688135147095, "learning_rate": 5.902403654581113e-07, "loss": 0.0173, "step": 95350 }, { "grad_norm": 0.4788568615913391, "learning_rate": 5.877099502911542e-07, "loss": 0.0289, "step": 95360 }, { "grad_norm": 0.33261266350746155, "learning_rate": 5.851849387779496e-07, "loss": 0.0241, "step": 95370 }, { "grad_norm": 0.30251747369766235, "learning_rate": 5.826653311946262e-07, "loss": 0.0319, "step": 95380 }, { "grad_norm": 0.3008710443973541, "learning_rate": 5.801511278167249e-07, "loss": 0.0162, "step": 95390 }, { "grad_norm": 0.24147045612335205, "learning_rate": 5.776423289191868e-07, "loss": 0.0244, "step": 95400 }, { "grad_norm": 0.4261671304702759, "learning_rate": 5.751389347763869e-07, "loss": 0.0306, "step": 95410 }, { "grad_norm": 0.5076841711997986, "learning_rate": 5.726409456620841e-07, "loss": 0.0195, "step": 95420 }, { "grad_norm": 0.29030683636665344, "learning_rate": 5.701483618494541e-07, "loss": 0.0191, "step": 95430 }, { "grad_norm": 0.43830904364585876, "learning_rate": 5.67661183611079e-07, "loss": 0.0202, "step": 95440 }, { "grad_norm": 0.585025429725647, "learning_rate": 5.651794112189579e-07, "loss": 0.0247, "step": 95450 }, { "grad_norm": 0.28054890036582947, "learning_rate": 5.627030449444903e-07, "loss": 0.0251, "step": 95460 }, { "grad_norm": 0.26704391837120056, "learning_rate": 5.60232085058493e-07, "loss": 0.0213, "step": 95470 }, { "grad_norm": 0.4310707747936249, "learning_rate": 5.577665318311776e-07, "loss": 0.0225, "step": 95480 }, { "grad_norm": 0.5029534101486206, "learning_rate": 5.553063855321838e-07, "loss": 0.0221, "step": 95490 }, { "grad_norm": 0.3717751204967499, "learning_rate": 5.528516464305467e-07, "loss": 0.0214, "step": 95500 }, { "grad_norm": 0.43865591287612915, "learning_rate": 5.504023147947068e-07, "loss": 0.0202, "step": 95510 }, { "grad_norm": 0.6598334312438965, "learning_rate": 5.479583908925279e-07, "loss": 0.0382, "step": 95520 }, { "grad_norm": 0.35050129890441895, "learning_rate": 5.455198749912626e-07, "loss": 0.0304, "step": 95530 }, { "grad_norm": 0.3114273250102997, "learning_rate": 5.430867673575979e-07, "loss": 0.0263, "step": 95540 }, { "grad_norm": 0.25039753317832947, "learning_rate": 5.406590682575985e-07, "loss": 0.0153, "step": 95550 }, { "grad_norm": 0.4154767394065857, "learning_rate": 5.382367779567687e-07, "loss": 0.0231, "step": 95560 }, { "grad_norm": 0.2759352922439575, "learning_rate": 5.358198967200023e-07, "loss": 0.0184, "step": 95570 }, { "grad_norm": 0.365829199552536, "learning_rate": 5.334084248115989e-07, "loss": 0.0252, "step": 95580 }, { "grad_norm": 0.3610937297344208, "learning_rate": 5.310023624952754e-07, "loss": 0.0185, "step": 95590 }, { "grad_norm": 0.5421450734138489, "learning_rate": 5.286017100341656e-07, "loss": 0.0181, "step": 95600 }, { "grad_norm": 0.41527649760246277, "learning_rate": 5.262064676907874e-07, "loss": 0.0229, "step": 95610 }, { "grad_norm": 0.2029270976781845, "learning_rate": 5.238166357270924e-07, "loss": 0.0248, "step": 95620 }, { "grad_norm": 0.37650495767593384, "learning_rate": 5.214322144044215e-07, "loss": 0.0201, "step": 95630 }, { "grad_norm": 0.3228461742401123, "learning_rate": 5.190532039835327e-07, "loss": 0.025, "step": 95640 }, { "grad_norm": 0.42767855525016785, "learning_rate": 5.166796047245903e-07, "loss": 0.0188, "step": 95650 }, { "grad_norm": 0.599577009677887, "learning_rate": 5.143114168871699e-07, "loss": 0.0224, "step": 95660 }, { "grad_norm": 0.5322142839431763, "learning_rate": 5.119486407302532e-07, "loss": 0.0239, "step": 95670 }, { "grad_norm": 0.31767770648002625, "learning_rate": 5.095912765122335e-07, "loss": 0.0269, "step": 95680 }, { "grad_norm": 0.2786630690097809, "learning_rate": 5.07239324490899e-07, "loss": 0.0214, "step": 95690 }, { "grad_norm": 0.3951161503791809, "learning_rate": 5.048927849234608e-07, "loss": 0.0219, "step": 95700 }, { "grad_norm": 0.2582569420337677, "learning_rate": 5.025516580665358e-07, "loss": 0.0277, "step": 95710 }, { "grad_norm": 0.44402754306793213, "learning_rate": 5.002159441761412e-07, "loss": 0.0287, "step": 95720 }, { "grad_norm": 0.6056705117225647, "learning_rate": 4.978856435077062e-07, "loss": 0.0285, "step": 95730 }, { "grad_norm": 0.2677159011363983, "learning_rate": 4.955607563160769e-07, "loss": 0.0252, "step": 95740 }, { "grad_norm": 0.2807868719100952, "learning_rate": 4.932412828554944e-07, "loss": 0.0226, "step": 95750 }, { "grad_norm": 0.5874834656715393, "learning_rate": 4.909272233796114e-07, "loss": 0.026, "step": 95760 }, { "grad_norm": 0.6152199506759644, "learning_rate": 4.88618578141492e-07, "loss": 0.0177, "step": 95770 }, { "grad_norm": 0.19722014665603638, "learning_rate": 4.863153473936011e-07, "loss": 0.0243, "step": 95780 }, { "grad_norm": 0.4015940725803375, "learning_rate": 4.840175313878259e-07, "loss": 0.0182, "step": 95790 }, { "grad_norm": 0.3117103576660156, "learning_rate": 4.817251303754433e-07, "loss": 0.0159, "step": 95800 }, { "grad_norm": 0.218414306640625, "learning_rate": 4.794381446071583e-07, "loss": 0.017, "step": 95810 }, { "grad_norm": 0.3641885817050934, "learning_rate": 4.771565743330597e-07, "loss": 0.0243, "step": 95820 }, { "grad_norm": 0.3120744228363037, "learning_rate": 4.7488041980266463e-07, "loss": 0.0206, "step": 95830 }, { "grad_norm": 0.45659729838371277, "learning_rate": 4.726096812648795e-07, "loss": 0.0204, "step": 95840 }, { "grad_norm": 0.39717498421669006, "learning_rate": 4.703443589680445e-07, "loss": 0.019, "step": 95850 }, { "grad_norm": 0.39588743448257446, "learning_rate": 4.6808445315987826e-07, "loss": 0.0222, "step": 95860 }, { "grad_norm": 0.2654927968978882, "learning_rate": 4.658299640875274e-07, "loss": 0.0235, "step": 95870 }, { "grad_norm": 0.3380621671676636, "learning_rate": 4.635808919975393e-07, "loss": 0.0186, "step": 95880 }, { "grad_norm": 0.21512548625469208, "learning_rate": 4.6133723713586706e-07, "loss": 0.0241, "step": 95890 }, { "grad_norm": 0.34507250785827637, "learning_rate": 4.590989997478756e-07, "loss": 0.014, "step": 95900 }, { "grad_norm": 1.001327395439148, "learning_rate": 4.5686618007833027e-07, "loss": 0.0179, "step": 95910 }, { "grad_norm": 0.6232869029045105, "learning_rate": 4.54638778371419e-07, "loss": 0.0146, "step": 95920 }, { "grad_norm": 0.28033071756362915, "learning_rate": 4.524167948707192e-07, "loss": 0.0231, "step": 95930 }, { "grad_norm": 0.4469176232814789, "learning_rate": 4.5020022981921986e-07, "loss": 0.0125, "step": 95940 }, { "grad_norm": 0.36847546696662903, "learning_rate": 4.47989083459327e-07, "loss": 0.0292, "step": 95950 }, { "grad_norm": 0.30037766695022583, "learning_rate": 4.457833560328528e-07, "loss": 0.0197, "step": 95960 }, { "grad_norm": 0.3369538187980652, "learning_rate": 4.4358304778100434e-07, "loss": 0.0232, "step": 95970 }, { "grad_norm": 0.2897661328315735, "learning_rate": 4.413881589444058e-07, "loss": 0.0205, "step": 95980 }, { "grad_norm": 0.30546805262565613, "learning_rate": 4.391986897630873e-07, "loss": 0.0275, "step": 95990 }, { "grad_norm": 0.570052981376648, "learning_rate": 4.370146404764908e-07, "loss": 0.0253, "step": 96000 }, { "grad_norm": 0.3390023112297058, "learning_rate": 4.3483601132345285e-07, "loss": 0.0288, "step": 96010 }, { "grad_norm": 0.3998303711414337, "learning_rate": 4.32662802542233e-07, "loss": 0.0188, "step": 96020 }, { "grad_norm": 0.25407910346984863, "learning_rate": 4.3049501437047444e-07, "loss": 0.0204, "step": 96030 }, { "grad_norm": 0.28749409317970276, "learning_rate": 4.283326470452653e-07, "loss": 0.0215, "step": 96040 }, { "grad_norm": 0.47081446647644043, "learning_rate": 4.2617570080306647e-07, "loss": 0.0182, "step": 96050 }, { "grad_norm": 0.22813624143600464, "learning_rate": 4.2402417587975585e-07, "loss": 0.0179, "step": 96060 }, { "grad_norm": 0.20806682109832764, "learning_rate": 4.218780725106286e-07, "loss": 0.0181, "step": 96070 }, { "grad_norm": 0.6284893155097961, "learning_rate": 4.197373909303748e-07, "loss": 0.0347, "step": 96080 }, { "grad_norm": 0.8251147866249084, "learning_rate": 4.17602131373096e-07, "loss": 0.0287, "step": 96090 }, { "grad_norm": 0.7108100652694702, "learning_rate": 4.154722940723055e-07, "loss": 0.0375, "step": 96100 }, { "grad_norm": 0.2655598521232605, "learning_rate": 4.1334787926091136e-07, "loss": 0.0186, "step": 96110 }, { "grad_norm": 0.33716902136802673, "learning_rate": 4.1122888717124444e-07, "loss": 0.0164, "step": 96120 }, { "grad_norm": 0.48520544171333313, "learning_rate": 4.091153180350249e-07, "loss": 0.0245, "step": 96130 }, { "grad_norm": 0.7723462581634521, "learning_rate": 4.0700717208340125e-07, "loss": 0.0301, "step": 96140 }, { "grad_norm": 0.4155136048793793, "learning_rate": 4.049044495469112e-07, "loss": 0.0158, "step": 96150 }, { "grad_norm": 0.27885109186172485, "learning_rate": 4.0280715065549866e-07, "loss": 0.0174, "step": 96160 }, { "grad_norm": 1.834060788154602, "learning_rate": 4.0071527563853575e-07, "loss": 0.0144, "step": 96170 }, { "grad_norm": 0.38710975646972656, "learning_rate": 3.9862882472477823e-07, "loss": 0.0216, "step": 96180 }, { "grad_norm": 0.244527205824852, "learning_rate": 3.9654779814239375e-07, "loss": 0.021, "step": 96190 }, { "grad_norm": 0.5563559532165527, "learning_rate": 3.944721961189668e-07, "loss": 0.0167, "step": 96200 }, { "grad_norm": 0.6791391968727112, "learning_rate": 3.9240201888147697e-07, "loss": 0.0236, "step": 96210 }, { "grad_norm": 0.19356507062911987, "learning_rate": 3.9033726665632096e-07, "loss": 0.0176, "step": 96220 }, { "grad_norm": 0.425102174282074, "learning_rate": 3.882779396692959e-07, "loss": 0.0173, "step": 96230 }, { "grad_norm": 0.48035380244255066, "learning_rate": 3.8622403814559947e-07, "loss": 0.0276, "step": 96240 }, { "grad_norm": 0.2195846438407898, "learning_rate": 3.841755623098575e-07, "loss": 0.0152, "step": 96250 }, { "grad_norm": 0.6840006113052368, "learning_rate": 3.821325123860742e-07, "loss": 0.0191, "step": 96260 }, { "grad_norm": 0.44467487931251526, "learning_rate": 3.800948885976818e-07, "loss": 0.0187, "step": 96270 }, { "grad_norm": 0.24883350729942322, "learning_rate": 3.780626911675078e-07, "loss": 0.0228, "step": 96280 }, { "grad_norm": 0.25569769740104675, "learning_rate": 3.76035920317791e-07, "loss": 0.0257, "step": 96290 }, { "grad_norm": 0.23339970409870148, "learning_rate": 3.740145762701819e-07, "loss": 0.0221, "step": 96300 }, { "grad_norm": 0.33197689056396484, "learning_rate": 3.7199865924572585e-07, "loss": 0.0162, "step": 96310 }, { "grad_norm": 0.37037065625190735, "learning_rate": 3.699881694648799e-07, "loss": 0.0204, "step": 96320 }, { "grad_norm": 0.2831283509731293, "learning_rate": 3.679831071475126e-07, "loss": 0.0291, "step": 96330 }, { "grad_norm": 0.34609854221343994, "learning_rate": 3.65983472512893e-07, "loss": 0.0249, "step": 96340 }, { "grad_norm": 0.23315773904323578, "learning_rate": 3.6398926577969616e-07, "loss": 0.0186, "step": 96350 }, { "grad_norm": 0.24192047119140625, "learning_rate": 3.620004871660032e-07, "loss": 0.0321, "step": 96360 }, { "grad_norm": 0.25955143570899963, "learning_rate": 3.600171368893124e-07, "loss": 0.0189, "step": 96370 }, { "grad_norm": 0.45992937684059143, "learning_rate": 3.580392151665113e-07, "loss": 0.0331, "step": 96380 }, { "grad_norm": 0.19267216324806213, "learning_rate": 3.5606672221391024e-07, "loss": 0.0191, "step": 96390 }, { "grad_norm": 1.165046215057373, "learning_rate": 3.540996582472089e-07, "loss": 0.0153, "step": 96400 }, { "grad_norm": 0.5530420541763306, "learning_rate": 3.521380234815297e-07, "loss": 0.0252, "step": 96410 }, { "grad_norm": 0.15525056421756744, "learning_rate": 3.5018181813138985e-07, "loss": 0.0304, "step": 96420 }, { "grad_norm": 0.2815380394458771, "learning_rate": 3.482310424107238e-07, "loss": 0.0185, "step": 96430 }, { "grad_norm": 0.2610624134540558, "learning_rate": 3.4628569653285537e-07, "loss": 0.0191, "step": 96440 }, { "grad_norm": 0.5421361327171326, "learning_rate": 3.44345780710531e-07, "loss": 0.0247, "step": 96450 }, { "grad_norm": 0.5560587048530579, "learning_rate": 3.424112951558978e-07, "loss": 0.0237, "step": 96460 }, { "grad_norm": 0.25029289722442627, "learning_rate": 3.40482240080503e-07, "loss": 0.0222, "step": 96470 }, { "grad_norm": 0.20388305187225342, "learning_rate": 3.3855861569531133e-07, "loss": 0.0186, "step": 96480 }, { "grad_norm": 0.29052233695983887, "learning_rate": 3.3664042221068226e-07, "loss": 0.0163, "step": 96490 }, { "grad_norm": 0.3178784251213074, "learning_rate": 3.347276598363924e-07, "loss": 0.0225, "step": 96500 }, { "grad_norm": 0.45779404044151306, "learning_rate": 3.328203287816134e-07, "loss": 0.022, "step": 96510 }, { "grad_norm": 0.2549504339694977, "learning_rate": 3.3091842925493386e-07, "loss": 0.0244, "step": 96520 }, { "grad_norm": 0.44027483463287354, "learning_rate": 3.2902196146433196e-07, "loss": 0.0251, "step": 96530 }, { "grad_norm": 0.4661727249622345, "learning_rate": 3.2713092561720836e-07, "loss": 0.0249, "step": 96540 }, { "grad_norm": 0.40305349230766296, "learning_rate": 3.2524532192036996e-07, "loss": 0.021, "step": 96550 }, { "grad_norm": 0.6024436354637146, "learning_rate": 3.2336515058001836e-07, "loss": 0.0171, "step": 96560 }, { "grad_norm": 0.3727589547634125, "learning_rate": 3.2149041180176143e-07, "loss": 0.0226, "step": 96570 }, { "grad_norm": 0.21219956874847412, "learning_rate": 3.1962110579062954e-07, "loss": 0.0225, "step": 96580 }, { "grad_norm": 0.20495522022247314, "learning_rate": 3.177572327510314e-07, "loss": 0.0253, "step": 96590 }, { "grad_norm": 0.417906790971756, "learning_rate": 3.158987928868151e-07, "loss": 0.0198, "step": 96600 }, { "grad_norm": 0.21000346541404724, "learning_rate": 3.1404578640120144e-07, "loss": 0.0279, "step": 96610 }, { "grad_norm": 0.31336405873298645, "learning_rate": 3.1219821349683933e-07, "loss": 0.0203, "step": 96620 }, { "grad_norm": 0.3776780664920807, "learning_rate": 3.103560743757783e-07, "loss": 0.0295, "step": 96630 }, { "grad_norm": 0.318463534116745, "learning_rate": 3.0851936923946834e-07, "loss": 0.024, "step": 96640 }, { "grad_norm": 0.4170074164867401, "learning_rate": 3.06688098288771e-07, "loss": 0.0185, "step": 96650 }, { "grad_norm": 0.1923995465040207, "learning_rate": 3.0486226172394825e-07, "loss": 0.0225, "step": 96660 }, { "grad_norm": 0.31550461053848267, "learning_rate": 3.030418597446738e-07, "loss": 0.0167, "step": 96670 }, { "grad_norm": 1.5348423719406128, "learning_rate": 3.012268925500272e-07, "loss": 0.0232, "step": 96680 }, { "grad_norm": 0.3331559896469116, "learning_rate": 2.994173603384831e-07, "loss": 0.0221, "step": 96690 }, { "grad_norm": 0.46386438608169556, "learning_rate": 2.9761326330793317e-07, "loss": 0.0151, "step": 96700 }, { "grad_norm": 0.9751425385475159, "learning_rate": 2.958146016556751e-07, "loss": 0.0188, "step": 96710 }, { "grad_norm": 1.3189796209335327, "learning_rate": 2.94021375578396e-07, "loss": 0.0201, "step": 96720 }, { "grad_norm": 0.1899743527173996, "learning_rate": 2.9223358527221135e-07, "loss": 0.0219, "step": 96730 }, { "grad_norm": 0.25991272926330566, "learning_rate": 2.904512309326257e-07, "loss": 0.0239, "step": 96740 }, { "grad_norm": 0.27458709478378296, "learning_rate": 2.8867431275455545e-07, "loss": 0.0212, "step": 96750 }, { "grad_norm": 0.24857346713542938, "learning_rate": 2.86902830932323e-07, "loss": 0.0267, "step": 96760 }, { "grad_norm": 0.5630422830581665, "learning_rate": 2.851367856596565e-07, "loss": 0.0207, "step": 96770 }, { "grad_norm": 0.4417623281478882, "learning_rate": 2.83376177129685e-07, "loss": 0.0223, "step": 96780 }, { "grad_norm": 1.5302236080169678, "learning_rate": 2.8162100553494884e-07, "loss": 0.0164, "step": 96790 }, { "grad_norm": 0.449603796005249, "learning_rate": 2.798712710673834e-07, "loss": 0.0179, "step": 96800 }, { "grad_norm": 0.7758010625839233, "learning_rate": 2.7812697391834676e-07, "loss": 0.0217, "step": 96810 }, { "grad_norm": 0.32750189304351807, "learning_rate": 2.7638811427858627e-07, "loss": 0.0212, "step": 96820 }, { "grad_norm": 0.9153763651847839, "learning_rate": 2.746546923382665e-07, "loss": 0.0202, "step": 96830 }, { "grad_norm": 0.19730323553085327, "learning_rate": 2.729267082869469e-07, "loss": 0.0227, "step": 96840 }, { "grad_norm": 0.4845619797706604, "learning_rate": 2.712041623135986e-07, "loss": 0.0176, "step": 96850 }, { "grad_norm": 0.2524772584438324, "learning_rate": 2.694870546065931e-07, "loss": 0.0209, "step": 96860 }, { "grad_norm": 0.21393759548664093, "learning_rate": 2.677753853537246e-07, "loss": 0.018, "step": 96870 }, { "grad_norm": 0.532730758190155, "learning_rate": 2.660691547421601e-07, "loss": 0.0226, "step": 96880 }, { "grad_norm": 0.28149181604385376, "learning_rate": 2.6436836295850033e-07, "loss": 0.0191, "step": 96890 }, { "grad_norm": 0.213880255818367, "learning_rate": 2.6267301018874093e-07, "loss": 0.0177, "step": 96900 }, { "grad_norm": 0.4818271994590759, "learning_rate": 2.6098309661828355e-07, "loss": 0.031, "step": 96910 }, { "grad_norm": 0.17084871232509613, "learning_rate": 2.592986224319305e-07, "loss": 0.0181, "step": 96920 }, { "grad_norm": 0.2573096454143524, "learning_rate": 2.576195878139009e-07, "loss": 0.0159, "step": 96930 }, { "grad_norm": 0.2747814655303955, "learning_rate": 2.5594599294780364e-07, "loss": 0.0226, "step": 96940 }, { "grad_norm": 0.17973606288433075, "learning_rate": 2.542778380166644e-07, "loss": 0.0248, "step": 96950 }, { "grad_norm": 0.3067588210105896, "learning_rate": 2.526151232029095e-07, "loss": 0.0264, "step": 96960 }, { "grad_norm": 0.359640508890152, "learning_rate": 2.5095784868836567e-07, "loss": 0.0236, "step": 96970 }, { "grad_norm": 0.22783412039279938, "learning_rate": 2.493060146542825e-07, "loss": 0.0164, "step": 96980 }, { "grad_norm": 0.2615582346916199, "learning_rate": 2.476596212812876e-07, "loss": 0.0179, "step": 96990 }, { "grad_norm": 0.3341326117515564, "learning_rate": 2.4601866874943703e-07, "loss": 0.0198, "step": 97000 }, { "grad_norm": 0.2864551246166229, "learning_rate": 2.4438315723818163e-07, "loss": 0.0232, "step": 97010 }, { "grad_norm": 0.22251196205615997, "learning_rate": 2.4275308692637833e-07, "loss": 0.0194, "step": 97020 }, { "grad_norm": 0.43991854786872864, "learning_rate": 2.411284579922901e-07, "loss": 0.0209, "step": 97030 }, { "grad_norm": 0.41955113410949707, "learning_rate": 2.395092706135804e-07, "loss": 0.0213, "step": 97040 }, { "grad_norm": 0.28637441992759705, "learning_rate": 2.3789552496731872e-07, "loss": 0.0161, "step": 97050 }, { "grad_norm": 0.33880895376205444, "learning_rate": 2.3628722122999158e-07, "loss": 0.02, "step": 97060 }, { "grad_norm": 1.4966273307800293, "learning_rate": 2.3468435957747503e-07, "loss": 0.0171, "step": 97070 }, { "grad_norm": 0.24162322282791138, "learning_rate": 2.330869401850566e-07, "loss": 0.0185, "step": 97080 }, { "grad_norm": 0.31650489568710327, "learning_rate": 2.3149496322742437e-07, "loss": 0.0235, "step": 97090 }, { "grad_norm": 0.2902147173881531, "learning_rate": 2.2990842887868346e-07, "loss": 0.0231, "step": 97100 }, { "grad_norm": 0.2650167942047119, "learning_rate": 2.2832733731232292e-07, "loss": 0.0173, "step": 97110 }, { "grad_norm": 0.3986707329750061, "learning_rate": 2.2675168870124886e-07, "loss": 0.0259, "step": 97120 }, { "grad_norm": 0.7337808609008789, "learning_rate": 2.2518148321778455e-07, "loss": 0.0206, "step": 97130 }, { "grad_norm": 0.30438974499702454, "learning_rate": 2.2361672103363706e-07, "loss": 0.0328, "step": 97140 }, { "grad_norm": 0.4746466875076294, "learning_rate": 2.220574023199251e-07, "loss": 0.0285, "step": 97150 }, { "grad_norm": 0.3643520176410675, "learning_rate": 2.2050352724717894e-07, "loss": 0.0202, "step": 97160 }, { "grad_norm": 0.1508384644985199, "learning_rate": 2.1895509598532372e-07, "loss": 0.0151, "step": 97170 }, { "grad_norm": 0.24580635130405426, "learning_rate": 2.1741210870369066e-07, "loss": 0.0224, "step": 97180 }, { "grad_norm": 0.37641191482543945, "learning_rate": 2.1587456557102814e-07, "loss": 0.0301, "step": 97190 }, { "grad_norm": 0.6436935067176819, "learning_rate": 2.143424667554683e-07, "loss": 0.023, "step": 97200 }, { "grad_norm": 0.24571029841899872, "learning_rate": 2.128158124245716e-07, "loss": 0.0206, "step": 97210 }, { "grad_norm": 0.5118485689163208, "learning_rate": 2.1129460274527668e-07, "loss": 0.0167, "step": 97220 }, { "grad_norm": 0.21507993340492249, "learning_rate": 2.0977883788395046e-07, "loss": 0.023, "step": 97230 }, { "grad_norm": 0.26654040813446045, "learning_rate": 2.0826851800634928e-07, "loss": 0.0146, "step": 97240 }, { "grad_norm": 0.42891207337379456, "learning_rate": 2.0676364327764653e-07, "loss": 0.0244, "step": 97250 }, { "grad_norm": 0.3192689120769501, "learning_rate": 2.0526421386240502e-07, "loss": 0.0215, "step": 97260 }, { "grad_norm": 0.2560434341430664, "learning_rate": 2.0377022992461025e-07, "loss": 0.0245, "step": 97270 }, { "grad_norm": 0.3585522472858429, "learning_rate": 2.0228169162763156e-07, "loss": 0.0174, "step": 97280 }, { "grad_norm": 0.3515172004699707, "learning_rate": 2.0079859913425536e-07, "loss": 0.0196, "step": 97290 }, { "grad_norm": 0.36162734031677246, "learning_rate": 1.9932095260667417e-07, "loss": 0.0235, "step": 97300 }, { "grad_norm": 0.9579376578330994, "learning_rate": 1.9784875220648092e-07, "loss": 0.02, "step": 97310 }, { "grad_norm": 0.3446502089500427, "learning_rate": 1.9638199809466907e-07, "loss": 0.0211, "step": 97320 }, { "grad_norm": 0.368965744972229, "learning_rate": 1.9492069043164364e-07, "loss": 0.0214, "step": 97330 }, { "grad_norm": 0.3533765375614166, "learning_rate": 1.9346482937721567e-07, "loss": 0.0192, "step": 97340 }, { "grad_norm": 0.21614114940166473, "learning_rate": 1.920144150905856e-07, "loss": 0.0245, "step": 97350 }, { "grad_norm": 0.20032577216625214, "learning_rate": 1.9056944773037655e-07, "loss": 0.0231, "step": 97360 }, { "grad_norm": 0.26946204900741577, "learning_rate": 1.8912992745460102e-07, "loss": 0.022, "step": 97370 }, { "grad_norm": 0.22410069406032562, "learning_rate": 1.876958544206886e-07, "loss": 0.0246, "step": 97380 }, { "grad_norm": 0.45618316531181335, "learning_rate": 1.8626722878546942e-07, "loss": 0.0301, "step": 97390 }, { "grad_norm": 0.3778950273990631, "learning_rate": 1.8484405070516851e-07, "loss": 0.018, "step": 97400 }, { "grad_norm": 0.482026606798172, "learning_rate": 1.8342632033542805e-07, "loss": 0.0262, "step": 97410 }, { "grad_norm": 0.3596839904785156, "learning_rate": 1.8201403783129066e-07, "loss": 0.0256, "step": 97420 }, { "grad_norm": 0.6634079813957214, "learning_rate": 1.806072033471884e-07, "loss": 0.0221, "step": 97430 }, { "grad_norm": 0.30839967727661133, "learning_rate": 1.79205817036987e-07, "loss": 0.0204, "step": 97440 }, { "grad_norm": 0.2877484858036041, "learning_rate": 1.7780987905393065e-07, "loss": 0.0218, "step": 97450 }, { "grad_norm": 0.37608247995376587, "learning_rate": 1.764193895506805e-07, "loss": 0.0161, "step": 97460 }, { "grad_norm": 0.2682403326034546, "learning_rate": 1.7503434867929824e-07, "loss": 0.0249, "step": 97470 }, { "grad_norm": 0.3082340955734253, "learning_rate": 1.7365475659125165e-07, "loss": 0.0166, "step": 97480 }, { "grad_norm": 0.3170710504055023, "learning_rate": 1.7228061343740332e-07, "loss": 0.0243, "step": 97490 }, { "grad_norm": 0.16906343400478363, "learning_rate": 1.7091191936803308e-07, "loss": 0.0216, "step": 97500 }, { "grad_norm": 1.5924842357635498, "learning_rate": 1.6954867453281563e-07, "loss": 0.0264, "step": 97510 }, { "grad_norm": 0.22240573167800903, "learning_rate": 1.6819087908084285e-07, "loss": 0.0189, "step": 97520 }, { "grad_norm": 0.36721178889274597, "learning_rate": 1.6683853316058485e-07, "loss": 0.025, "step": 97530 }, { "grad_norm": 0.32807931303977966, "learning_rate": 1.6549163691995106e-07, "loss": 0.0188, "step": 97540 }, { "grad_norm": 0.3858393430709839, "learning_rate": 1.641501905062237e-07, "loss": 0.031, "step": 97550 }, { "grad_norm": 0.41084009408950806, "learning_rate": 1.628141940661021e-07, "loss": 0.0251, "step": 97560 }, { "grad_norm": 0.3487796187400818, "learning_rate": 1.6148364774569158e-07, "loss": 0.0216, "step": 97570 }, { "grad_norm": 0.1797354817390442, "learning_rate": 1.6015855169050355e-07, "loss": 0.023, "step": 97580 }, { "grad_norm": 0.3676944375038147, "learning_rate": 1.5883890604543873e-07, "loss": 0.0158, "step": 97590 }, { "grad_norm": 0.21235042810440063, "learning_rate": 1.5752471095482057e-07, "loss": 0.0273, "step": 97600 }, { "grad_norm": 0.45287150144577026, "learning_rate": 1.5621596656235638e-07, "loss": 0.0207, "step": 97610 }, { "grad_norm": 0.26386821269989014, "learning_rate": 1.5491267301117608e-07, "loss": 0.0171, "step": 97620 }, { "grad_norm": 0.3877030313014984, "learning_rate": 1.536148304438101e-07, "loss": 0.0166, "step": 97630 }, { "grad_norm": 0.3781571388244629, "learning_rate": 1.5232243900217823e-07, "loss": 0.0237, "step": 97640 }, { "grad_norm": 0.38393619656562805, "learning_rate": 1.51035498827623e-07, "loss": 0.0194, "step": 97650 }, { "grad_norm": 0.29555749893188477, "learning_rate": 1.497540100608763e-07, "loss": 0.0289, "step": 97660 }, { "grad_norm": 0.3064590394496918, "learning_rate": 1.4847797284208153e-07, "loss": 0.0217, "step": 97670 }, { "grad_norm": 0.2990761697292328, "learning_rate": 1.472073873107882e-07, "loss": 0.0291, "step": 97680 }, { "grad_norm": 0.4269690215587616, "learning_rate": 1.459422536059407e-07, "loss": 0.0274, "step": 97690 }, { "grad_norm": 0.5563403964042664, "learning_rate": 1.4468257186589508e-07, "loss": 0.0206, "step": 97700 }, { "grad_norm": 0.33666160702705383, "learning_rate": 1.4342834222840217e-07, "loss": 0.0165, "step": 97710 }, { "grad_norm": 0.31280413269996643, "learning_rate": 1.4217956483063566e-07, "loss": 0.0213, "step": 97720 }, { "grad_norm": 0.9657325148582458, "learning_rate": 1.409362398091474e-07, "loss": 0.0237, "step": 97730 }, { "grad_norm": 0.2966781556606293, "learning_rate": 1.3969836729990638e-07, "loss": 0.0173, "step": 97740 }, { "grad_norm": 0.25557929277420044, "learning_rate": 1.384659474382932e-07, "loss": 0.0164, "step": 97750 }, { "grad_norm": 0.2819545269012451, "learning_rate": 1.372389803590779e-07, "loss": 0.0197, "step": 97760 }, { "grad_norm": 0.3441816568374634, "learning_rate": 1.360174661964364e-07, "loss": 0.0182, "step": 97770 }, { "grad_norm": 0.9416369795799255, "learning_rate": 1.3480140508396188e-07, "loss": 0.0243, "step": 97780 }, { "grad_norm": 0.25630712509155273, "learning_rate": 1.335907971546313e-07, "loss": 0.0264, "step": 97790 }, { "grad_norm": 0.3550782799720764, "learning_rate": 1.3238564254083875e-07, "loss": 0.0211, "step": 97800 }, { "grad_norm": 0.28311964869499207, "learning_rate": 1.3118594137437323e-07, "loss": 0.0209, "step": 97810 }, { "grad_norm": 0.3056652545928955, "learning_rate": 1.2999169378644094e-07, "loss": 0.0193, "step": 97820 }, { "grad_norm": 0.2667657434940338, "learning_rate": 1.2880289990763738e-07, "loss": 0.023, "step": 97830 }, { "grad_norm": 0.4563572406768799, "learning_rate": 1.2761955986796968e-07, "loss": 0.0244, "step": 97840 }, { "grad_norm": 0.20279507339000702, "learning_rate": 1.264416737968399e-07, "loss": 0.0308, "step": 97850 }, { "grad_norm": 0.256756991147995, "learning_rate": 1.2526924182307275e-07, "loss": 0.0177, "step": 97860 }, { "grad_norm": 0.24948877096176147, "learning_rate": 1.2410226407487124e-07, "loss": 0.0167, "step": 97870 }, { "grad_norm": 0.20963676273822784, "learning_rate": 1.229407406798555e-07, "loss": 0.0228, "step": 97880 }, { "grad_norm": 0.17941133677959442, "learning_rate": 1.2178467176505726e-07, "loss": 0.0186, "step": 97890 }, { "grad_norm": 0.34534400701522827, "learning_rate": 1.2063405745689205e-07, "loss": 0.0241, "step": 97900 }, { "grad_norm": 0.3604156970977783, "learning_rate": 1.1948889788119812e-07, "loss": 0.0185, "step": 97910 }, { "grad_norm": 0.3643667995929718, "learning_rate": 1.1834919316320303e-07, "loss": 0.0198, "step": 97920 }, { "grad_norm": 0.3643830418586731, "learning_rate": 1.1721494342754048e-07, "loss": 0.0209, "step": 97930 }, { "grad_norm": 0.26585322618484497, "learning_rate": 1.1608614879825563e-07, "loss": 0.0179, "step": 97940 }, { "grad_norm": 0.43050673604011536, "learning_rate": 1.1496280939879422e-07, "loss": 0.0207, "step": 97950 }, { "grad_norm": 0.4419335722923279, "learning_rate": 1.1384492535199687e-07, "loss": 0.0175, "step": 97960 }, { "grad_norm": 0.21128565073013306, "learning_rate": 1.1273249678011578e-07, "loss": 0.0196, "step": 97970 }, { "grad_norm": 0.4630313515663147, "learning_rate": 1.1162552380480362e-07, "loss": 0.0238, "step": 97980 }, { "grad_norm": 0.36156776547431946, "learning_rate": 1.1052400654711359e-07, "loss": 0.0203, "step": 97990 }, { "grad_norm": 0.24554584920406342, "learning_rate": 1.0942794512751597e-07, "loss": 0.0178, "step": 98000 }, { "grad_norm": 0.2735036611557007, "learning_rate": 1.0833733966587045e-07, "loss": 0.0234, "step": 98010 }, { "grad_norm": 0.3261817991733551, "learning_rate": 1.0725219028143718e-07, "loss": 0.0226, "step": 98020 }, { "grad_norm": 0.2880729138851166, "learning_rate": 1.0617249709289345e-07, "loss": 0.0155, "step": 98030 }, { "grad_norm": 0.3605578541755676, "learning_rate": 1.0509826021831703e-07, "loss": 0.0184, "step": 98040 }, { "grad_norm": 0.3166407346725464, "learning_rate": 1.0402947977517508e-07, "loss": 0.0188, "step": 98050 }, { "grad_norm": 0.2611561417579651, "learning_rate": 1.0296615588035185e-07, "loss": 0.0207, "step": 98060 }, { "grad_norm": 0.5486974716186523, "learning_rate": 1.0190828865012659e-07, "loss": 0.019, "step": 98070 }, { "grad_norm": 0.2193998396396637, "learning_rate": 1.008558782001956e-07, "loss": 0.0154, "step": 98080 }, { "grad_norm": 0.186213880777359, "learning_rate": 9.980892464563906e-08, "loss": 0.0185, "step": 98090 }, { "grad_norm": 0.1742624193429947, "learning_rate": 9.876742810095985e-08, "loss": 0.018, "step": 98100 }, { "grad_norm": 0.3354796767234802, "learning_rate": 9.77313886800446e-08, "loss": 0.0197, "step": 98110 }, { "grad_norm": 0.5167422890663147, "learning_rate": 9.670080649619717e-08, "loss": 0.0299, "step": 98120 }, { "grad_norm": 0.4031149744987488, "learning_rate": 9.567568166212737e-08, "loss": 0.0228, "step": 98130 }, { "grad_norm": 0.20701013505458832, "learning_rate": 9.465601428992888e-08, "loss": 0.0174, "step": 98140 }, { "grad_norm": 0.39065030217170715, "learning_rate": 9.364180449111803e-08, "loss": 0.0217, "step": 98150 }, { "grad_norm": 0.20628201961517334, "learning_rate": 9.263305237661169e-08, "loss": 0.02, "step": 98160 }, { "grad_norm": 0.479608952999115, "learning_rate": 9.162975805671603e-08, "loss": 0.0189, "step": 98170 }, { "grad_norm": 0.3581943213939667, "learning_rate": 9.063192164115442e-08, "loss": 0.0212, "step": 98180 }, { "grad_norm": 0.3636305034160614, "learning_rate": 8.963954323904511e-08, "loss": 0.0169, "step": 98190 }, { "grad_norm": 0.19576239585876465, "learning_rate": 8.865262295891796e-08, "loss": 0.0185, "step": 98200 }, { "grad_norm": 0.7431902289390564, "learning_rate": 8.767116090870331e-08, "loss": 0.0223, "step": 98210 }, { "grad_norm": 0.19138996303081512, "learning_rate": 8.669515719572086e-08, "loss": 0.0159, "step": 98220 }, { "grad_norm": 0.6178662180900574, "learning_rate": 8.572461192671855e-08, "loss": 0.0247, "step": 98230 }, { "grad_norm": 0.2482353001832962, "learning_rate": 8.475952520782815e-08, "loss": 0.0138, "step": 98240 }, { "grad_norm": 0.3957439959049225, "learning_rate": 8.379989714458747e-08, "loss": 0.0182, "step": 98250 }, { "grad_norm": 0.2969415783882141, "learning_rate": 8.284572784194034e-08, "loss": 0.0315, "step": 98260 }, { "grad_norm": 0.3646571934223175, "learning_rate": 8.189701740424216e-08, "loss": 0.027, "step": 98270 }, { "grad_norm": 0.5510919094085693, "learning_rate": 8.095376593522663e-08, "loss": 0.0217, "step": 98280 }, { "grad_norm": 0.25709760189056396, "learning_rate": 8.001597353806123e-08, "loss": 0.0323, "step": 98290 }, { "grad_norm": 0.3262171447277069, "learning_rate": 7.90836403152917e-08, "loss": 0.0239, "step": 98300 }, { "grad_norm": 1.0054187774658203, "learning_rate": 7.815676636888092e-08, "loss": 0.0161, "step": 98310 }, { "grad_norm": 0.4280869662761688, "learning_rate": 7.723535180019226e-08, "loss": 0.0201, "step": 98320 }, { "grad_norm": 0.31794169545173645, "learning_rate": 7.6319396709984e-08, "loss": 0.0287, "step": 98330 }, { "grad_norm": 0.7173303961753845, "learning_rate": 7.540890119843158e-08, "loss": 0.0265, "step": 98340 }, { "grad_norm": 0.28293097019195557, "learning_rate": 7.450386536509979e-08, "loss": 0.0225, "step": 98350 }, { "grad_norm": 0.20698148012161255, "learning_rate": 7.360428930895947e-08, "loss": 0.0286, "step": 98360 }, { "grad_norm": 0.1856945902109146, "learning_rate": 7.271017312839302e-08, "loss": 0.0307, "step": 98370 }, { "grad_norm": 0.3834667205810547, "learning_rate": 7.182151692118333e-08, "loss": 0.0268, "step": 98380 }, { "grad_norm": 0.4976453483104706, "learning_rate": 7.093832078449713e-08, "loss": 0.0265, "step": 98390 }, { "grad_norm": 0.7859475016593933, "learning_rate": 7.006058481493493e-08, "loss": 0.0336, "step": 98400 }, { "grad_norm": 0.17308616638183594, "learning_rate": 6.918830910847552e-08, "loss": 0.0149, "step": 98410 }, { "grad_norm": 0.39043498039245605, "learning_rate": 6.832149376052033e-08, "loss": 0.0205, "step": 98420 }, { "grad_norm": 0.5200351476669312, "learning_rate": 6.746013886584912e-08, "loss": 0.0301, "step": 98430 }, { "grad_norm": 0.36702755093574524, "learning_rate": 6.660424451866431e-08, "loss": 0.0173, "step": 98440 }, { "grad_norm": 0.31699514389038086, "learning_rate": 6.575381081256327e-08, "loss": 0.0264, "step": 98450 }, { "grad_norm": 0.43168482184410095, "learning_rate": 6.49088378405549e-08, "loss": 0.0255, "step": 98460 }, { "grad_norm": 0.6442358493804932, "learning_rate": 6.406932569503755e-08, "loss": 0.0275, "step": 98470 }, { "grad_norm": 0.4305477440357208, "learning_rate": 6.323527446782662e-08, "loss": 0.0252, "step": 98480 }, { "grad_norm": 0.272936075925827, "learning_rate": 6.240668425012142e-08, "loss": 0.027, "step": 98490 }, { "grad_norm": 0.2823338508605957, "learning_rate": 6.158355513254388e-08, "loss": 0.0193, "step": 98500 }, { "grad_norm": 0.24482949078083038, "learning_rate": 6.076588720510534e-08, "loss": 0.019, "step": 98510 }, { "grad_norm": 0.6520809531211853, "learning_rate": 5.995368055722316e-08, "loss": 0.0247, "step": 98520 }, { "grad_norm": 0.5771741271018982, "learning_rate": 5.914693527773185e-08, "loss": 0.0164, "step": 98530 }, { "grad_norm": 0.29433709383010864, "learning_rate": 5.834565145483861e-08, "loss": 0.0166, "step": 98540 }, { "grad_norm": 0.37669408321380615, "learning_rate": 5.754982917618446e-08, "loss": 0.0185, "step": 98550 }, { "grad_norm": 0.14093267917633057, "learning_rate": 5.675946852878866e-08, "loss": 0.014, "step": 98560 }, { "grad_norm": 0.27852562069892883, "learning_rate": 5.59745695990932e-08, "loss": 0.0207, "step": 98570 }, { "grad_norm": 0.2591770887374878, "learning_rate": 5.51951324729294e-08, "loss": 0.0162, "step": 98580 }, { "grad_norm": 0.38872915506362915, "learning_rate": 5.4421157235529093e-08, "loss": 0.0148, "step": 98590 }, { "grad_norm": 0.7015690803527832, "learning_rate": 5.365264397154679e-08, "loss": 0.022, "step": 98600 }, { "grad_norm": 0.19356770813465118, "learning_rate": 5.288959276501526e-08, "loss": 0.0188, "step": 98610 }, { "grad_norm": 0.5321481823921204, "learning_rate": 5.2132003699378876e-08, "loss": 0.0202, "step": 98620 }, { "grad_norm": 0.29009994864463806, "learning_rate": 5.137987685749912e-08, "loss": 0.0165, "step": 98630 }, { "grad_norm": 0.33441105484962463, "learning_rate": 5.0633212321610227e-08, "loss": 0.0225, "step": 98640 }, { "grad_norm": 0.3054264783859253, "learning_rate": 4.989201017338019e-08, "loss": 0.0207, "step": 98650 }, { "grad_norm": 0.2187442034482956, "learning_rate": 4.9156270493860846e-08, "loss": 0.0199, "step": 98660 }, { "grad_norm": 0.44712966680526733, "learning_rate": 4.842599336351561e-08, "loss": 0.0222, "step": 98670 }, { "grad_norm": 0.3816584646701813, "learning_rate": 4.7701178862197274e-08, "loss": 0.023, "step": 98680 }, { "grad_norm": 0.24922266602516174, "learning_rate": 4.698182706918131e-08, "loss": 0.0183, "step": 98690 }, { "grad_norm": 0.15025189518928528, "learning_rate": 4.6267938063121466e-08, "loss": 0.018, "step": 98700 }, { "grad_norm": 0.5934537649154663, "learning_rate": 4.555951192209973e-08, "loss": 0.0265, "step": 98710 }, { "grad_norm": 1.1194287538528442, "learning_rate": 4.485654872358747e-08, "loss": 0.0179, "step": 98720 }, { "grad_norm": 0.1848548799753189, "learning_rate": 4.415904854445097e-08, "loss": 0.017, "step": 98730 }, { "grad_norm": 0.3604137599468231, "learning_rate": 4.34670114609792e-08, "loss": 0.0221, "step": 98740 }, { "grad_norm": 0.28491124510765076, "learning_rate": 4.278043754884498e-08, "loss": 0.019, "step": 98750 }, { "grad_norm": 0.7382993698120117, "learning_rate": 4.209932688312712e-08, "loss": 0.0219, "step": 98760 }, { "grad_norm": 0.2775775194168091, "learning_rate": 4.14236795383216e-08, "loss": 0.0236, "step": 98770 }, { "grad_norm": 0.28383374214172363, "learning_rate": 4.075349558830821e-08, "loss": 0.0201, "step": 98780 }, { "grad_norm": 0.21727146208286285, "learning_rate": 4.008877510638387e-08, "loss": 0.0181, "step": 98790 }, { "grad_norm": 0.23092976212501526, "learning_rate": 3.942951816523488e-08, "loss": 0.0173, "step": 98800 }, { "grad_norm": 0.3070470094680786, "learning_rate": 3.8775724836959125e-08, "loss": 0.0214, "step": 98810 }, { "grad_norm": 0.28993189334869385, "learning_rate": 3.812739519305497e-08, "loss": 0.024, "step": 98820 }, { "grad_norm": 0.2772357761859894, "learning_rate": 3.748452930442126e-08, "loss": 0.0242, "step": 98830 }, { "grad_norm": 0.6119892001152039, "learning_rate": 3.6847127241362855e-08, "loss": 0.0305, "step": 98840 }, { "grad_norm": 0.34297487139701843, "learning_rate": 3.621518907358512e-08, "loss": 0.0285, "step": 98850 }, { "grad_norm": 0.5484989881515503, "learning_rate": 3.558871487019388e-08, "loss": 0.0317, "step": 98860 }, { "grad_norm": 0.383974552154541, "learning_rate": 3.4967704699701007e-08, "loss": 0.0168, "step": 98870 }, { "grad_norm": 0.34046053886413574, "learning_rate": 3.435215863001884e-08, "loss": 0.0187, "step": 98880 }, { "grad_norm": 0.19521988928318024, "learning_rate": 3.374207672846019e-08, "loss": 0.0228, "step": 98890 }, { "grad_norm": 0.3533931374549866, "learning_rate": 3.313745906174392e-08, "loss": 0.0235, "step": 98900 }, { "grad_norm": 0.34202900528907776, "learning_rate": 3.253830569599492e-08, "loss": 0.0251, "step": 98910 }, { "grad_norm": 0.17636097967624664, "learning_rate": 3.1944616696727436e-08, "loss": 0.0213, "step": 98920 }, { "grad_norm": 0.34969568252563477, "learning_rate": 3.135639212887287e-08, "loss": 0.0292, "step": 98930 }, { "grad_norm": 0.24822072684764862, "learning_rate": 3.077363205675754e-08, "loss": 0.0201, "step": 98940 }, { "grad_norm": 0.3084951341152191, "learning_rate": 3.0196336544113804e-08, "loss": 0.0152, "step": 98950 }, { "grad_norm": 0.3685940206050873, "learning_rate": 2.9624505654063383e-08, "loss": 0.0273, "step": 98960 }, { "grad_norm": 0.2539595067501068, "learning_rate": 2.90581394491507e-08, "loss": 0.0236, "step": 98970 }, { "grad_norm": 0.23909993469715118, "learning_rate": 2.8497237991309545e-08, "loss": 0.0229, "step": 98980 }, { "grad_norm": 0.39416974782943726, "learning_rate": 2.7941801341879735e-08, "loss": 0.02, "step": 98990 }, { "grad_norm": 0.29284295439720154, "learning_rate": 2.7391829561601578e-08, "loss": 0.0198, "step": 99000 }, { "grad_norm": 0.3654163181781769, "learning_rate": 2.6847322710621403e-08, "loss": 0.0166, "step": 99010 }, { "grad_norm": 0.34735456109046936, "learning_rate": 2.6308280848486022e-08, "loss": 0.0223, "step": 99020 }, { "grad_norm": 0.29368799924850464, "learning_rate": 2.5774704034137177e-08, "loss": 0.0206, "step": 99030 }, { "grad_norm": 0.32380205392837524, "learning_rate": 2.524659232593929e-08, "loss": 0.0164, "step": 99040 }, { "grad_norm": 0.4916025996208191, "learning_rate": 2.472394578163506e-08, "loss": 0.0264, "step": 99050 }, { "grad_norm": 0.23198753595352173, "learning_rate": 2.4206764458378772e-08, "loss": 0.0187, "step": 99060 }, { "grad_norm": 1.2426258325576782, "learning_rate": 2.3695048412736285e-08, "loss": 0.0195, "step": 99070 }, { "grad_norm": 0.7020143866539001, "learning_rate": 2.3188797700668395e-08, "loss": 0.0222, "step": 99080 }, { "grad_norm": 0.4161253273487091, "learning_rate": 2.268801237753082e-08, "loss": 0.0195, "step": 99090 }, { "grad_norm": 0.3861084580421448, "learning_rate": 2.2192692498090862e-08, "loss": 0.0178, "step": 99100 }, { "grad_norm": 0.48290541768074036, "learning_rate": 2.170283811652185e-08, "loss": 0.0295, "step": 99110 }, { "grad_norm": 0.37414035201072693, "learning_rate": 2.12184492863865e-08, "loss": 0.022, "step": 99120 }, { "grad_norm": 0.5565318465232849, "learning_rate": 2.073952606066465e-08, "loss": 0.0204, "step": 99130 }, { "grad_norm": 0.5865601301193237, "learning_rate": 2.026606849171997e-08, "loss": 0.0218, "step": 99140 }, { "grad_norm": 0.3050520718097687, "learning_rate": 1.9798076631333262e-08, "loss": 0.0308, "step": 99150 }, { "grad_norm": 0.22276999056339264, "learning_rate": 1.933555053069136e-08, "loss": 0.0191, "step": 99160 }, { "grad_norm": 0.31465059518814087, "learning_rate": 1.887849024036492e-08, "loss": 0.0284, "step": 99170 }, { "grad_norm": 0.2040049433708191, "learning_rate": 1.8426895810341736e-08, "loss": 0.0287, "step": 99180 }, { "grad_norm": 0.2890198528766632, "learning_rate": 1.798076729000453e-08, "loss": 0.0228, "step": 99190 }, { "grad_norm": 0.3629124164581299, "learning_rate": 1.75401047281476e-08, "loss": 0.021, "step": 99200 }, { "grad_norm": 0.3274589478969574, "learning_rate": 1.7104908172954626e-08, "loss": 0.0226, "step": 99210 }, { "grad_norm": 0.3091385066509247, "learning_rate": 1.6675177672020868e-08, "loss": 0.0271, "step": 99220 }, { "grad_norm": 0.15837500989437103, "learning_rate": 1.6250913272342073e-08, "loss": 0.017, "step": 99230 }, { "grad_norm": 0.42592549324035645, "learning_rate": 1.5832115020314453e-08, "loss": 0.0153, "step": 99240 }, { "grad_norm": 0.6323513388633728, "learning_rate": 1.5418782961734712e-08, "loss": 0.0166, "step": 99250 }, { "grad_norm": 0.4839460849761963, "learning_rate": 1.501091714181113e-08, "loss": 0.0207, "step": 99260 }, { "grad_norm": 0.4265737235546112, "learning_rate": 1.460851760513582e-08, "loss": 0.0189, "step": 99270 }, { "grad_norm": 0.1874968260526657, "learning_rate": 1.4211584395723566e-08, "loss": 0.0212, "step": 99280 }, { "grad_norm": 0.34583595395088196, "learning_rate": 1.382011755697854e-08, "loss": 0.0241, "step": 99290 }, { "grad_norm": 0.42215338349342346, "learning_rate": 1.3434117131716494e-08, "loss": 0.0201, "step": 99300 }, { "grad_norm": 0.3662542402744293, "learning_rate": 1.3053583162142557e-08, "loss": 0.0221, "step": 99310 }, { "grad_norm": 0.27559149265289307, "learning_rate": 1.2678515689873438e-08, "loss": 0.0181, "step": 99320 }, { "grad_norm": 0.49794986844062805, "learning_rate": 1.2308914755931877e-08, "loss": 0.0193, "step": 99330 }, { "grad_norm": 0.3199872374534607, "learning_rate": 1.1944780400729993e-08, "loss": 0.028, "step": 99340 }, { "grad_norm": 0.4681510627269745, "learning_rate": 1.1586112664085935e-08, "loss": 0.0294, "step": 99350 }, { "grad_norm": 0.1464800387620926, "learning_rate": 1.1232911585234984e-08, "loss": 0.02, "step": 99360 }, { "grad_norm": 0.27985095977783203, "learning_rate": 1.0885177202790697e-08, "loss": 0.0303, "step": 99370 }, { "grad_norm": 0.23119361698627472, "learning_rate": 1.0542909554789315e-08, "loss": 0.0261, "step": 99380 }, { "grad_norm": 0.36770322918891907, "learning_rate": 1.0206108678656456e-08, "loss": 0.0182, "step": 99390 }, { "grad_norm": 0.39852356910705566, "learning_rate": 9.874774611223769e-09, "loss": 0.0198, "step": 99400 }, { "grad_norm": 0.2749834358692169, "learning_rate": 9.548907388728933e-09, "loss": 0.0178, "step": 99410 }, { "grad_norm": 0.4910098910331726, "learning_rate": 9.228507046804557e-09, "loss": 0.0202, "step": 99420 }, { "grad_norm": 0.2655166685581207, "learning_rate": 8.91357362048928e-09, "loss": 0.0251, "step": 99430 }, { "grad_norm": 0.3050340414047241, "learning_rate": 8.604107144227769e-09, "loss": 0.0168, "step": 99440 }, { "grad_norm": 0.3115820586681366, "learning_rate": 8.300107651859623e-09, "loss": 0.019, "step": 99450 }, { "grad_norm": 0.3464454710483551, "learning_rate": 8.001575176630472e-09, "loss": 0.019, "step": 99460 }, { "grad_norm": 0.4443289339542389, "learning_rate": 7.708509751186422e-09, "loss": 0.0204, "step": 99470 }, { "grad_norm": 0.19366928935050964, "learning_rate": 7.420911407579611e-09, "loss": 0.0208, "step": 99480 }, { "grad_norm": 0.7202664613723755, "learning_rate": 7.13878017725711e-09, "loss": 0.0247, "step": 99490 }, { "grad_norm": 0.36436280608177185, "learning_rate": 6.8621160910720125e-09, "loss": 0.0225, "step": 99500 }, { "grad_norm": 0.40986716747283936, "learning_rate": 6.59091917928345e-09, "loss": 0.0182, "step": 99510 }, { "grad_norm": 0.34941428899765015, "learning_rate": 6.32518947155103e-09, "loss": 0.0183, "step": 99520 }, { "grad_norm": 0.5638836622238159, "learning_rate": 6.064926996929288e-09, "loss": 0.0256, "step": 99530 }, { "grad_norm": 0.26227033138275146, "learning_rate": 5.810131783884343e-09, "loss": 0.0187, "step": 99540 }, { "grad_norm": 0.20217004418373108, "learning_rate": 5.5608038602772415e-09, "loss": 0.0235, "step": 99550 }, { "grad_norm": 0.3405892550945282, "learning_rate": 5.316943253375062e-09, "loss": 0.0174, "step": 99560 }, { "grad_norm": 0.4141310751438141, "learning_rate": 5.07854998984536e-09, "loss": 0.0244, "step": 99570 }, { "grad_norm": 0.2855057716369629, "learning_rate": 4.845624095756174e-09, "loss": 0.0171, "step": 99580 }, { "grad_norm": 0.3844063878059387, "learning_rate": 4.618165596587121e-09, "loss": 0.0245, "step": 99590 }, { "grad_norm": 0.3236703872680664, "learning_rate": 4.396174517207197e-09, "loss": 0.0207, "step": 99600 }, { "grad_norm": 0.1822250634431839, "learning_rate": 4.179650881896979e-09, "loss": 0.0228, "step": 99610 }, { "grad_norm": 0.27744171023368835, "learning_rate": 3.968594714331975e-09, "loss": 0.0207, "step": 99620 }, { "grad_norm": 0.6555664539337158, "learning_rate": 3.763006037593719e-09, "loss": 0.0248, "step": 99630 }, { "grad_norm": 0.2123641073703766, "learning_rate": 3.5628848741586786e-09, "loss": 0.0194, "step": 99640 }, { "grad_norm": 0.26259106397628784, "learning_rate": 3.368231245926001e-09, "loss": 0.0191, "step": 99650 }, { "grad_norm": 0.4950551688671112, "learning_rate": 3.179045174167561e-09, "loss": 0.0208, "step": 99660 }, { "grad_norm": 0.3246896266937256, "learning_rate": 2.9953266795834657e-09, "loss": 0.02, "step": 99670 }, { "grad_norm": 0.3764232099056244, "learning_rate": 2.817075782263201e-09, "loss": 0.0201, "step": 99680 }, { "grad_norm": 0.39212438464164734, "learning_rate": 2.6442925016967325e-09, "loss": 0.0209, "step": 99690 }, { "grad_norm": 0.598941445350647, "learning_rate": 2.4769768567800555e-09, "loss": 0.0174, "step": 99700 }, { "grad_norm": 0.3805243968963623, "learning_rate": 2.315128865809646e-09, "loss": 0.0282, "step": 99710 }, { "grad_norm": 0.3259055018424988, "learning_rate": 2.1587485464880097e-09, "loss": 0.0242, "step": 99720 }, { "grad_norm": 0.6211592555046082, "learning_rate": 2.0078359159125813e-09, "loss": 0.0297, "step": 99730 }, { "grad_norm": 0.40211421251296997, "learning_rate": 1.8623909905923776e-09, "loss": 0.0221, "step": 99740 }, { "grad_norm": 0.2589254677295685, "learning_rate": 1.7224137864257916e-09, "loss": 0.0251, "step": 99750 }, { "grad_norm": 0.2687188982963562, "learning_rate": 1.5879043187283505e-09, "loss": 0.0201, "step": 99760 }, { "grad_norm": 0.223525732755661, "learning_rate": 1.458862602204958e-09, "loss": 0.0177, "step": 99770 }, { "grad_norm": 0.3214530050754547, "learning_rate": 1.3352886509720996e-09, "loss": 0.0219, "step": 99780 }, { "grad_norm": 0.19934391975402832, "learning_rate": 1.2171824785356389e-09, "loss": 0.0228, "step": 99790 }, { "grad_norm": 0.2799447476863861, "learning_rate": 1.1045440978185718e-09, "loss": 0.0166, "step": 99800 }, { "grad_norm": 0.30285513401031494, "learning_rate": 9.973735211332714e-10, "loss": 0.0211, "step": 99810 }, { "grad_norm": 0.3935718834400177, "learning_rate": 8.956707602036929e-10, "loss": 0.0278, "step": 99820 }, { "grad_norm": 0.23277506232261658, "learning_rate": 7.994358261542712e-10, "loss": 0.0197, "step": 99830 }, { "grad_norm": 0.3563784658908844, "learning_rate": 7.086687295043692e-10, "loss": 0.0188, "step": 99840 }, { "grad_norm": 0.26015058159828186, "learning_rate": 6.23369480179381e-10, "loss": 0.0224, "step": 99850 }, { "grad_norm": 0.2617815434932709, "learning_rate": 5.43538087510731e-10, "loss": 0.0304, "step": 99860 }, { "grad_norm": 0.6615999341011047, "learning_rate": 4.691745602303233e-10, "loss": 0.0281, "step": 99870 }, { "grad_norm": 0.21442003548145294, "learning_rate": 4.0027890646499034e-10, "loss": 0.0268, "step": 99880 }, { "grad_norm": 0.2424648553133011, "learning_rate": 3.368511337531466e-10, "loss": 0.0212, "step": 99890 }, { "grad_norm": 0.37822744250297546, "learning_rate": 2.788912490281348e-10, "loss": 0.0172, "step": 99900 }, { "grad_norm": 0.2875105142593384, "learning_rate": 2.2639925863487954e-10, "loss": 0.0179, "step": 99910 }, { "grad_norm": 0.22411873936653137, "learning_rate": 1.7937516830213163e-10, "loss": 0.0247, "step": 99920 }, { "grad_norm": 0.2807525396347046, "learning_rate": 1.3781898318687702e-10, "loss": 0.0245, "step": 99930 }, { "grad_norm": 0.37593597173690796, "learning_rate": 1.0173070781882566e-10, "loss": 0.0192, "step": 99940 }, { "grad_norm": 0.3545752763748169, "learning_rate": 7.111034615592261e-11, "loss": 0.0202, "step": 99950 }, { "grad_norm": 0.24164706468582153, "learning_rate": 4.595790153993918e-11, "loss": 0.0209, "step": 99960 }, { "grad_norm": 0.4709622859954834, "learning_rate": 2.6273376729779585e-11, "loss": 0.0253, "step": 99970 }, { "grad_norm": 0.18756555020809174, "learning_rate": 1.2056773868174276e-11, "loss": 0.0189, "step": 99980 }, { "grad_norm": 0.22345435619354248, "learning_rate": 3.3080945149865926e-12, "loss": 0.0175, "step": 99990 }, { "grad_norm": 0.3014679551124573, "learning_rate": 2.7339625008337977e-14, "loss": 0.0227, "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 }