{ "best_global_step": null, "best_metric": null, "best_model_checkpoint": null, "epoch": 1.0, "eval_steps": 500, "global_step": 80000, "is_hyper_param_search": false, "is_local_process_zero": true, "is_world_process_zero": true, "log_history": [ { "grad_norm": 0.5431153774261475, "learning_rate": 2.25e-07, "loss": 1.2846, "step": 10 }, { "grad_norm": 0.704882800579071, "learning_rate": 4.75e-07, "loss": 1.274, "step": 20 }, { "grad_norm": 0.8483231067657471, "learning_rate": 7.25e-07, "loss": 1.3093, "step": 30 }, { "grad_norm": 0.7457141876220703, "learning_rate": 9.75e-07, "loss": 1.2806, "step": 40 }, { "grad_norm": 0.763711154460907, "learning_rate": 1.2250000000000001e-06, "loss": 1.2756, "step": 50 }, { "grad_norm": 1.2313371896743774, "learning_rate": 1.475e-06, "loss": 1.267, "step": 60 }, { "grad_norm": 1.1704816818237305, "learning_rate": 1.7250000000000002e-06, "loss": 1.2196, "step": 70 }, { "grad_norm": 1.1192280054092407, "learning_rate": 1.975e-06, "loss": 1.199, "step": 80 }, { "grad_norm": 1.1373167037963867, "learning_rate": 2.225e-06, "loss": 1.206, "step": 90 }, { "grad_norm": 1.3954030275344849, "learning_rate": 2.4750000000000004e-06, "loss": 1.2092, "step": 100 }, { "grad_norm": 1.2224562168121338, "learning_rate": 2.725e-06, "loss": 1.1821, "step": 110 }, { "grad_norm": 1.4046024084091187, "learning_rate": 2.975e-06, "loss": 1.1745, "step": 120 }, { "grad_norm": 3.856203556060791, "learning_rate": 3.225e-06, "loss": 1.1559, "step": 130 }, { "grad_norm": 1.549608588218689, "learning_rate": 3.4750000000000006e-06, "loss": 1.141, "step": 140 }, { "grad_norm": 1.141764760017395, "learning_rate": 3.725e-06, "loss": 1.1088, "step": 150 }, { "grad_norm": 4.809648036956787, "learning_rate": 3.975e-06, "loss": 1.1195, "step": 160 }, { "grad_norm": 1.5821609497070312, "learning_rate": 4.225e-06, "loss": 1.0991, "step": 170 }, { "grad_norm": 1.0878759622573853, "learning_rate": 4.475e-06, "loss": 1.0987, "step": 180 }, { "grad_norm": 1.4275349378585815, "learning_rate": 4.7250000000000005e-06, "loss": 1.1194, "step": 190 }, { "grad_norm": 1.5334622859954834, "learning_rate": 4.975000000000001e-06, "loss": 1.1132, "step": 200 }, { "grad_norm": 1.1342462301254272, "learning_rate": 5.225e-06, "loss": 1.0931, "step": 210 }, { "grad_norm": 0.8981611728668213, "learning_rate": 5.475e-06, "loss": 1.094, "step": 220 }, { "grad_norm": 0.9375035166740417, "learning_rate": 5.725e-06, "loss": 1.0962, "step": 230 }, { "grad_norm": 1.0894956588745117, "learning_rate": 5.975e-06, "loss": 1.1104, "step": 240 }, { "grad_norm": 1.120235800743103, "learning_rate": 6.2250000000000005e-06, "loss": 1.0972, "step": 250 }, { "grad_norm": 0.9950898885726929, "learning_rate": 6.475000000000001e-06, "loss": 1.0758, "step": 260 }, { "grad_norm": 1.678780436515808, "learning_rate": 6.725000000000001e-06, "loss": 1.0885, "step": 270 }, { "grad_norm": 1.1942331790924072, "learning_rate": 6.975000000000001e-06, "loss": 1.1087, "step": 280 }, { "grad_norm": 1.4090739488601685, "learning_rate": 7.2249999999999994e-06, "loss": 1.0957, "step": 290 }, { "grad_norm": 1.3901479244232178, "learning_rate": 7.4750000000000004e-06, "loss": 1.0924, "step": 300 }, { "grad_norm": 1.2477613687515259, "learning_rate": 7.725e-06, "loss": 1.1259, "step": 310 }, { "grad_norm": 1.1780877113342285, "learning_rate": 7.975e-06, "loss": 1.0782, "step": 320 }, { "grad_norm": 0.8715211153030396, "learning_rate": 8.225e-06, "loss": 1.098, "step": 330 }, { "grad_norm": 1.097379207611084, "learning_rate": 8.475000000000001e-06, "loss": 1.0972, "step": 340 }, { "grad_norm": 1.343982219696045, "learning_rate": 8.725e-06, "loss": 1.0865, "step": 350 }, { "grad_norm": 0.9045400023460388, "learning_rate": 8.975e-06, "loss": 1.1127, "step": 360 }, { "grad_norm": 0.966303825378418, "learning_rate": 9.225e-06, "loss": 1.094, "step": 370 }, { "grad_norm": 1.1865202188491821, "learning_rate": 9.475e-06, "loss": 1.0818, "step": 380 }, { "grad_norm": 0.9908709526062012, "learning_rate": 9.725000000000001e-06, "loss": 1.0733, "step": 390 }, { "grad_norm": 1.1961721181869507, "learning_rate": 9.975e-06, "loss": 1.0668, "step": 400 }, { "grad_norm": 1.6272366046905518, "learning_rate": 1.0225e-05, "loss": 1.0567, "step": 410 }, { "grad_norm": 1.993451714515686, "learning_rate": 1.0475e-05, "loss": 1.0232, "step": 420 }, { "grad_norm": 2.56054949760437, "learning_rate": 1.0725e-05, "loss": 1.0248, "step": 430 }, { "grad_norm": 1.792136549949646, "learning_rate": 1.0975e-05, "loss": 0.9915, "step": 440 }, { "grad_norm": 3.412102222442627, "learning_rate": 1.1225e-05, "loss": 0.9405, "step": 450 }, { "grad_norm": 2.477004051208496, "learning_rate": 1.1475000000000001e-05, "loss": 0.9518, "step": 460 }, { "grad_norm": 2.128061056137085, "learning_rate": 1.1725e-05, "loss": 0.9072, "step": 470 }, { "grad_norm": 2.2953710556030273, "learning_rate": 1.1975e-05, "loss": 0.8957, "step": 480 }, { "grad_norm": 1.9724732637405396, "learning_rate": 1.2225e-05, "loss": 0.8574, "step": 490 }, { "grad_norm": 2.6025655269622803, "learning_rate": 1.2475e-05, "loss": 0.8078, "step": 500 }, { "grad_norm": 2.316500425338745, "learning_rate": 1.2725000000000001e-05, "loss": 0.776, "step": 510 }, { "grad_norm": 2.3263494968414307, "learning_rate": 1.2975e-05, "loss": 0.7557, "step": 520 }, { "grad_norm": 2.684110641479492, "learning_rate": 1.3225000000000001e-05, "loss": 0.7335, "step": 530 }, { "grad_norm": 2.1422369480133057, "learning_rate": 1.3475000000000002e-05, "loss": 0.665, "step": 540 }, { "grad_norm": 2.449409008026123, "learning_rate": 1.3725000000000002e-05, "loss": 0.6903, "step": 550 }, { "grad_norm": 2.6822845935821533, "learning_rate": 1.3975000000000003e-05, "loss": 0.6858, "step": 560 }, { "grad_norm": 3.0314767360687256, "learning_rate": 1.4225e-05, "loss": 0.6349, "step": 570 }, { "grad_norm": 2.008347988128662, "learning_rate": 1.4475e-05, "loss": 0.5799, "step": 580 }, { "grad_norm": 2.879486083984375, "learning_rate": 1.4725e-05, "loss": 0.5929, "step": 590 }, { "grad_norm": 2.340683937072754, "learning_rate": 1.4975e-05, "loss": 0.5634, "step": 600 }, { "grad_norm": 3.1592657566070557, "learning_rate": 1.5225e-05, "loss": 0.532, "step": 610 }, { "grad_norm": 2.8037824630737305, "learning_rate": 1.5475e-05, "loss": 0.5043, "step": 620 }, { "grad_norm": 3.2210662364959717, "learning_rate": 1.5725e-05, "loss": 0.4842, "step": 630 }, { "grad_norm": 2.9923315048217773, "learning_rate": 1.5975000000000002e-05, "loss": 0.4561, "step": 640 }, { "grad_norm": 3.3648886680603027, "learning_rate": 1.6225e-05, "loss": 0.4256, "step": 650 }, { "grad_norm": 3.65533185005188, "learning_rate": 1.6475e-05, "loss": 0.4299, "step": 660 }, { "grad_norm": 2.3232812881469727, "learning_rate": 1.6725000000000003e-05, "loss": 0.3792, "step": 670 }, { "grad_norm": 2.873311996459961, "learning_rate": 1.6975000000000003e-05, "loss": 0.3906, "step": 680 }, { "grad_norm": 3.3385281562805176, "learning_rate": 1.7225e-05, "loss": 0.3596, "step": 690 }, { "grad_norm": 2.924952507019043, "learning_rate": 1.7475e-05, "loss": 0.3706, "step": 700 }, { "grad_norm": 2.6991488933563232, "learning_rate": 1.7725e-05, "loss": 0.3187, "step": 710 }, { "grad_norm": 2.805422067642212, "learning_rate": 1.7975e-05, "loss": 0.3544, "step": 720 }, { "grad_norm": 2.990666627883911, "learning_rate": 1.8225e-05, "loss": 0.3118, "step": 730 }, { "grad_norm": 2.3594367504119873, "learning_rate": 1.8475000000000002e-05, "loss": 0.2647, "step": 740 }, { "grad_norm": 3.345263719558716, "learning_rate": 1.8725e-05, "loss": 0.2914, "step": 750 }, { "grad_norm": 1.8635302782058716, "learning_rate": 1.8975e-05, "loss": 0.2631, "step": 760 }, { "grad_norm": 2.944317102432251, "learning_rate": 1.9225e-05, "loss": 0.261, "step": 770 }, { "grad_norm": 2.555814027786255, "learning_rate": 1.9475000000000002e-05, "loss": 0.2894, "step": 780 }, { "grad_norm": 3.308335304260254, "learning_rate": 1.9725000000000002e-05, "loss": 0.2213, "step": 790 }, { "grad_norm": 2.7522993087768555, "learning_rate": 1.9975e-05, "loss": 0.2071, "step": 800 }, { "grad_norm": 2.6762218475341797, "learning_rate": 2.0225000000000004e-05, "loss": 0.2035, "step": 810 }, { "grad_norm": 2.7830991744995117, "learning_rate": 2.0475e-05, "loss": 0.2229, "step": 820 }, { "grad_norm": 3.4428389072418213, "learning_rate": 2.0725e-05, "loss": 0.2313, "step": 830 }, { "grad_norm": 2.4723029136657715, "learning_rate": 2.0975e-05, "loss": 0.2139, "step": 840 }, { "grad_norm": 2.835217237472534, "learning_rate": 2.1225e-05, "loss": 0.2125, "step": 850 }, { "grad_norm": 2.485422372817993, "learning_rate": 2.1475e-05, "loss": 0.1999, "step": 860 }, { "grad_norm": 2.946359157562256, "learning_rate": 2.1725e-05, "loss": 0.2294, "step": 870 }, { "grad_norm": 2.8859522342681885, "learning_rate": 2.1975000000000002e-05, "loss": 0.1872, "step": 880 }, { "grad_norm": 2.394087791442871, "learning_rate": 2.2225e-05, "loss": 0.2083, "step": 890 }, { "grad_norm": 2.312164783477783, "learning_rate": 2.2475e-05, "loss": 0.2075, "step": 900 }, { "grad_norm": 1.7804137468338013, "learning_rate": 2.2725000000000003e-05, "loss": 0.1958, "step": 910 }, { "grad_norm": 1.964760184288025, "learning_rate": 2.2975000000000003e-05, "loss": 0.1794, "step": 920 }, { "grad_norm": 2.4484925270080566, "learning_rate": 2.3225000000000002e-05, "loss": 0.1895, "step": 930 }, { "grad_norm": 2.800699472427368, "learning_rate": 2.3475e-05, "loss": 0.1994, "step": 940 }, { "grad_norm": 2.0316426753997803, "learning_rate": 2.3725e-05, "loss": 0.1667, "step": 950 }, { "grad_norm": 2.7410786151885986, "learning_rate": 2.3975e-05, "loss": 0.1846, "step": 960 }, { "grad_norm": 2.430562734603882, "learning_rate": 2.4225e-05, "loss": 0.2153, "step": 970 }, { "grad_norm": 2.692340850830078, "learning_rate": 2.4475000000000002e-05, "loss": 0.1591, "step": 980 }, { "grad_norm": 2.4501290321350098, "learning_rate": 2.4725e-05, "loss": 0.1582, "step": 990 }, { "grad_norm": 2.434159994125366, "learning_rate": 2.4975e-05, "loss": 0.1573, "step": 1000 }, { "grad_norm": 1.999289631843567, "learning_rate": 2.5225e-05, "loss": 0.16, "step": 1010 }, { "grad_norm": 2.5721986293792725, "learning_rate": 2.5475e-05, "loss": 0.1866, "step": 1020 }, { "grad_norm": 1.8383456468582153, "learning_rate": 2.5725e-05, "loss": 0.1542, "step": 1030 }, { "grad_norm": 2.3683457374572754, "learning_rate": 2.5974999999999998e-05, "loss": 0.1566, "step": 1040 }, { "grad_norm": 3.009570360183716, "learning_rate": 2.6225e-05, "loss": 0.19, "step": 1050 }, { "grad_norm": 2.0688016414642334, "learning_rate": 2.6475e-05, "loss": 0.1792, "step": 1060 }, { "grad_norm": 1.9425674676895142, "learning_rate": 2.6725e-05, "loss": 0.1936, "step": 1070 }, { "grad_norm": 2.5640499591827393, "learning_rate": 2.6975000000000002e-05, "loss": 0.1985, "step": 1080 }, { "grad_norm": 1.9155151844024658, "learning_rate": 2.7225e-05, "loss": 0.1756, "step": 1090 }, { "grad_norm": 2.2359235286712646, "learning_rate": 2.7475e-05, "loss": 0.1515, "step": 1100 }, { "grad_norm": 2.7123162746429443, "learning_rate": 2.7725e-05, "loss": 0.2232, "step": 1110 }, { "grad_norm": 2.0920968055725098, "learning_rate": 2.7975000000000002e-05, "loss": 0.156, "step": 1120 }, { "grad_norm": 1.9746192693710327, "learning_rate": 2.8225e-05, "loss": 0.1707, "step": 1130 }, { "grad_norm": 2.3495852947235107, "learning_rate": 2.8475e-05, "loss": 0.2076, "step": 1140 }, { "grad_norm": 2.398505210876465, "learning_rate": 2.8725e-05, "loss": 0.2024, "step": 1150 }, { "grad_norm": 2.1318910121917725, "learning_rate": 2.8975000000000003e-05, "loss": 0.1816, "step": 1160 }, { "grad_norm": 1.6881985664367676, "learning_rate": 2.9225000000000002e-05, "loss": 0.1823, "step": 1170 }, { "grad_norm": 2.080716609954834, "learning_rate": 2.9475e-05, "loss": 0.14, "step": 1180 }, { "grad_norm": 1.8937031030654907, "learning_rate": 2.9725000000000004e-05, "loss": 0.1719, "step": 1190 }, { "grad_norm": 2.991523504257202, "learning_rate": 2.9975000000000004e-05, "loss": 0.1806, "step": 1200 }, { "grad_norm": 2.852099895477295, "learning_rate": 3.0225000000000003e-05, "loss": 0.1879, "step": 1210 }, { "grad_norm": 2.382253885269165, "learning_rate": 3.0475000000000002e-05, "loss": 0.1725, "step": 1220 }, { "grad_norm": 1.809557318687439, "learning_rate": 3.0725e-05, "loss": 0.1691, "step": 1230 }, { "grad_norm": 1.6536983251571655, "learning_rate": 3.0975e-05, "loss": 0.1721, "step": 1240 }, { "grad_norm": 1.7845572233200073, "learning_rate": 3.122500000000001e-05, "loss": 0.1463, "step": 1250 }, { "grad_norm": 2.0346412658691406, "learning_rate": 3.1475e-05, "loss": 0.1717, "step": 1260 }, { "grad_norm": 1.952839970588684, "learning_rate": 3.1725e-05, "loss": 0.1573, "step": 1270 }, { "grad_norm": 1.8936820030212402, "learning_rate": 3.1975e-05, "loss": 0.1583, "step": 1280 }, { "grad_norm": 1.1636011600494385, "learning_rate": 3.2225e-05, "loss": 0.157, "step": 1290 }, { "grad_norm": 1.6308826208114624, "learning_rate": 3.2474999999999997e-05, "loss": 0.1583, "step": 1300 }, { "grad_norm": 1.4786405563354492, "learning_rate": 3.2725e-05, "loss": 0.1433, "step": 1310 }, { "grad_norm": 2.0649516582489014, "learning_rate": 3.2975e-05, "loss": 0.1698, "step": 1320 }, { "grad_norm": 2.0265157222747803, "learning_rate": 3.3225e-05, "loss": 0.1408, "step": 1330 }, { "grad_norm": 1.2851612567901611, "learning_rate": 3.3475e-05, "loss": 0.135, "step": 1340 }, { "grad_norm": 1.5551990270614624, "learning_rate": 3.3725e-05, "loss": 0.1635, "step": 1350 }, { "grad_norm": 2.3194570541381836, "learning_rate": 3.3975e-05, "loss": 0.153, "step": 1360 }, { "grad_norm": 1.4797348976135254, "learning_rate": 3.4225e-05, "loss": 0.1272, "step": 1370 }, { "grad_norm": 1.5963482856750488, "learning_rate": 3.4475000000000005e-05, "loss": 0.1347, "step": 1380 }, { "grad_norm": 1.4928956031799316, "learning_rate": 3.4725000000000004e-05, "loss": 0.156, "step": 1390 }, { "grad_norm": 1.3715198040008545, "learning_rate": 3.4975e-05, "loss": 0.1358, "step": 1400 }, { "grad_norm": 1.4973995685577393, "learning_rate": 3.5225e-05, "loss": 0.1438, "step": 1410 }, { "grad_norm": 1.7231431007385254, "learning_rate": 3.5475e-05, "loss": 0.1517, "step": 1420 }, { "grad_norm": 1.5333906412124634, "learning_rate": 3.5725e-05, "loss": 0.1326, "step": 1430 }, { "grad_norm": 1.525904655456543, "learning_rate": 3.5975e-05, "loss": 0.146, "step": 1440 }, { "grad_norm": 1.599575161933899, "learning_rate": 3.6225000000000006e-05, "loss": 0.1646, "step": 1450 }, { "grad_norm": 1.7930974960327148, "learning_rate": 3.6475000000000006e-05, "loss": 0.1491, "step": 1460 }, { "grad_norm": 1.2491945028305054, "learning_rate": 3.6725000000000005e-05, "loss": 0.1611, "step": 1470 }, { "grad_norm": 1.6500773429870605, "learning_rate": 3.6975000000000004e-05, "loss": 0.1311, "step": 1480 }, { "grad_norm": 1.325514793395996, "learning_rate": 3.7225000000000004e-05, "loss": 0.1492, "step": 1490 }, { "grad_norm": 1.1966854333877563, "learning_rate": 3.7475e-05, "loss": 0.1606, "step": 1500 }, { "grad_norm": 1.477803349494934, "learning_rate": 3.7725e-05, "loss": 0.135, "step": 1510 }, { "grad_norm": 1.8440725803375244, "learning_rate": 3.7975e-05, "loss": 0.165, "step": 1520 }, { "grad_norm": 1.1892518997192383, "learning_rate": 3.8225e-05, "loss": 0.1494, "step": 1530 }, { "grad_norm": 1.9365123510360718, "learning_rate": 3.8475e-05, "loss": 0.1525, "step": 1540 }, { "grad_norm": 1.780626654624939, "learning_rate": 3.8725e-05, "loss": 0.1324, "step": 1550 }, { "grad_norm": 1.376997947692871, "learning_rate": 3.8975e-05, "loss": 0.1543, "step": 1560 }, { "grad_norm": 1.5786669254302979, "learning_rate": 3.9225e-05, "loss": 0.1277, "step": 1570 }, { "grad_norm": 1.6603058576583862, "learning_rate": 3.9475000000000004e-05, "loss": 0.1554, "step": 1580 }, { "grad_norm": 2.022357940673828, "learning_rate": 3.9725e-05, "loss": 0.1582, "step": 1590 }, { "grad_norm": 1.2785974740982056, "learning_rate": 3.9975e-05, "loss": 0.1387, "step": 1600 }, { "grad_norm": 1.1644009351730347, "learning_rate": 4.0225e-05, "loss": 0.1285, "step": 1610 }, { "grad_norm": 1.737196683883667, "learning_rate": 4.0475e-05, "loss": 0.139, "step": 1620 }, { "grad_norm": 1.27861487865448, "learning_rate": 4.0725e-05, "loss": 0.1954, "step": 1630 }, { "grad_norm": 1.598375678062439, "learning_rate": 4.0975e-05, "loss": 0.1507, "step": 1640 }, { "grad_norm": 1.2664414644241333, "learning_rate": 4.1225e-05, "loss": 0.1492, "step": 1650 }, { "grad_norm": 1.518292784690857, "learning_rate": 4.1475000000000005e-05, "loss": 0.16, "step": 1660 }, { "grad_norm": 1.7412313222885132, "learning_rate": 4.1725000000000005e-05, "loss": 0.1329, "step": 1670 }, { "grad_norm": 1.3879241943359375, "learning_rate": 4.1975000000000004e-05, "loss": 0.1434, "step": 1680 }, { "grad_norm": 1.1349778175354004, "learning_rate": 4.2225e-05, "loss": 0.1407, "step": 1690 }, { "grad_norm": 1.428374171257019, "learning_rate": 4.2475e-05, "loss": 0.1373, "step": 1700 }, { "grad_norm": 1.679372787475586, "learning_rate": 4.2725e-05, "loss": 0.1665, "step": 1710 }, { "grad_norm": 1.321111798286438, "learning_rate": 4.2975e-05, "loss": 0.1593, "step": 1720 }, { "grad_norm": 1.2339612245559692, "learning_rate": 4.322500000000001e-05, "loss": 0.1306, "step": 1730 }, { "grad_norm": 1.346128225326538, "learning_rate": 4.3475000000000006e-05, "loss": 0.1318, "step": 1740 }, { "grad_norm": 1.4406754970550537, "learning_rate": 4.3725000000000006e-05, "loss": 0.1439, "step": 1750 }, { "grad_norm": 1.5731847286224365, "learning_rate": 4.3975e-05, "loss": 0.1419, "step": 1760 }, { "grad_norm": 0.9850104451179504, "learning_rate": 4.4225e-05, "loss": 0.1374, "step": 1770 }, { "grad_norm": 1.2982205152511597, "learning_rate": 4.4475e-05, "loss": 0.1618, "step": 1780 }, { "grad_norm": 1.6067787408828735, "learning_rate": 4.4725e-05, "loss": 0.1293, "step": 1790 }, { "grad_norm": 1.1752893924713135, "learning_rate": 4.4975e-05, "loss": 0.142, "step": 1800 }, { "grad_norm": 1.1445808410644531, "learning_rate": 4.5225e-05, "loss": 0.139, "step": 1810 }, { "grad_norm": 1.9155339002609253, "learning_rate": 4.5475e-05, "loss": 0.1484, "step": 1820 }, { "grad_norm": 2.004693031311035, "learning_rate": 4.5725e-05, "loss": 0.1431, "step": 1830 }, { "grad_norm": 1.1763824224472046, "learning_rate": 4.5975e-05, "loss": 0.1408, "step": 1840 }, { "grad_norm": 1.2791197299957275, "learning_rate": 4.6225e-05, "loss": 0.1698, "step": 1850 }, { "grad_norm": 1.1874176263809204, "learning_rate": 4.6475000000000005e-05, "loss": 0.1601, "step": 1860 }, { "grad_norm": 1.1339354515075684, "learning_rate": 4.6725000000000004e-05, "loss": 0.1446, "step": 1870 }, { "grad_norm": 1.2572120428085327, "learning_rate": 4.6975000000000003e-05, "loss": 0.1091, "step": 1880 }, { "grad_norm": 1.2498846054077148, "learning_rate": 4.7225e-05, "loss": 0.1481, "step": 1890 }, { "grad_norm": 1.1473273038864136, "learning_rate": 4.7475e-05, "loss": 0.131, "step": 1900 }, { "grad_norm": 2.287499189376831, "learning_rate": 4.7725e-05, "loss": 0.1757, "step": 1910 }, { "grad_norm": 1.2751128673553467, "learning_rate": 4.7975e-05, "loss": 0.1335, "step": 1920 }, { "grad_norm": 0.9073971509933472, "learning_rate": 4.822500000000001e-05, "loss": 0.1242, "step": 1930 }, { "grad_norm": 1.448928713798523, "learning_rate": 4.8475000000000006e-05, "loss": 0.1337, "step": 1940 }, { "grad_norm": 1.1824709177017212, "learning_rate": 4.8725000000000005e-05, "loss": 0.1319, "step": 1950 }, { "grad_norm": 1.306515097618103, "learning_rate": 4.8975000000000005e-05, "loss": 0.1781, "step": 1960 }, { "grad_norm": 1.0391674041748047, "learning_rate": 4.9225000000000004e-05, "loss": 0.1189, "step": 1970 }, { "grad_norm": 1.3401936292648315, "learning_rate": 4.9475e-05, "loss": 0.1544, "step": 1980 }, { "grad_norm": 1.1862478256225586, "learning_rate": 4.9725e-05, "loss": 0.1134, "step": 1990 }, { "grad_norm": 1.1880078315734863, "learning_rate": 4.9975e-05, "loss": 0.1418, "step": 2000 }, { "grad_norm": 1.1471537351608276, "learning_rate": 5.0225e-05, "loss": 0.1625, "step": 2010 }, { "grad_norm": 1.2696614265441895, "learning_rate": 5.047500000000001e-05, "loss": 0.1476, "step": 2020 }, { "grad_norm": 1.412797451019287, "learning_rate": 5.0725e-05, "loss": 0.1266, "step": 2030 }, { "grad_norm": 0.877478301525116, "learning_rate": 5.0975000000000006e-05, "loss": 0.1364, "step": 2040 }, { "grad_norm": 1.4326785802841187, "learning_rate": 5.1225e-05, "loss": 0.1412, "step": 2050 }, { "grad_norm": 1.4089504480361938, "learning_rate": 5.1475000000000004e-05, "loss": 0.1496, "step": 2060 }, { "grad_norm": 1.4427454471588135, "learning_rate": 5.1725000000000004e-05, "loss": 0.1286, "step": 2070 }, { "grad_norm": 1.0984492301940918, "learning_rate": 5.197500000000001e-05, "loss": 0.1378, "step": 2080 }, { "grad_norm": 1.204262137413025, "learning_rate": 5.2225e-05, "loss": 0.1393, "step": 2090 }, { "grad_norm": 1.1630460023880005, "learning_rate": 5.247500000000001e-05, "loss": 0.1377, "step": 2100 }, { "grad_norm": 1.238194227218628, "learning_rate": 5.2725e-05, "loss": 0.1823, "step": 2110 }, { "grad_norm": 1.1950664520263672, "learning_rate": 5.297500000000001e-05, "loss": 0.1341, "step": 2120 }, { "grad_norm": 0.8588013052940369, "learning_rate": 5.3225e-05, "loss": 0.1372, "step": 2130 }, { "grad_norm": 1.1154146194458008, "learning_rate": 5.3475e-05, "loss": 0.102, "step": 2140 }, { "grad_norm": 1.105975866317749, "learning_rate": 5.3725000000000005e-05, "loss": 0.1645, "step": 2150 }, { "grad_norm": 1.0765146017074585, "learning_rate": 5.3975e-05, "loss": 0.137, "step": 2160 }, { "grad_norm": 1.1347179412841797, "learning_rate": 5.4225000000000003e-05, "loss": 0.1299, "step": 2170 }, { "grad_norm": 0.9793242812156677, "learning_rate": 5.4474999999999996e-05, "loss": 0.1472, "step": 2180 }, { "grad_norm": 0.8273653984069824, "learning_rate": 5.4725e-05, "loss": 0.1241, "step": 2190 }, { "grad_norm": 1.0826293230056763, "learning_rate": 5.4975e-05, "loss": 0.1109, "step": 2200 }, { "grad_norm": 0.8819549679756165, "learning_rate": 5.522500000000001e-05, "loss": 0.122, "step": 2210 }, { "grad_norm": 1.080704689025879, "learning_rate": 5.5475e-05, "loss": 0.1187, "step": 2220 }, { "grad_norm": 1.071285367012024, "learning_rate": 5.5725000000000006e-05, "loss": 0.122, "step": 2230 }, { "grad_norm": 1.105644941329956, "learning_rate": 5.5975e-05, "loss": 0.1191, "step": 2240 }, { "grad_norm": 0.661395788192749, "learning_rate": 5.6225000000000005e-05, "loss": 0.1263, "step": 2250 }, { "grad_norm": 0.8865479230880737, "learning_rate": 5.6475e-05, "loss": 0.114, "step": 2260 }, { "grad_norm": 0.8789820075035095, "learning_rate": 5.6725e-05, "loss": 0.1122, "step": 2270 }, { "grad_norm": 1.0888891220092773, "learning_rate": 5.6975e-05, "loss": 0.1311, "step": 2280 }, { "grad_norm": 1.2141317129135132, "learning_rate": 5.722500000000001e-05, "loss": 0.1481, "step": 2290 }, { "grad_norm": 0.9565046429634094, "learning_rate": 5.7475e-05, "loss": 0.13, "step": 2300 }, { "grad_norm": 1.1403088569641113, "learning_rate": 5.772500000000001e-05, "loss": 0.1267, "step": 2310 }, { "grad_norm": 1.090891718864441, "learning_rate": 5.7975e-05, "loss": 0.1056, "step": 2320 }, { "grad_norm": 1.1995412111282349, "learning_rate": 5.8225000000000006e-05, "loss": 0.1313, "step": 2330 }, { "grad_norm": 0.8749694228172302, "learning_rate": 5.8475000000000005e-05, "loss": 0.1265, "step": 2340 }, { "grad_norm": 0.8057639598846436, "learning_rate": 5.8725000000000004e-05, "loss": 0.1269, "step": 2350 }, { "grad_norm": 0.8757561445236206, "learning_rate": 5.8975000000000004e-05, "loss": 0.1226, "step": 2360 }, { "grad_norm": 1.0051666498184204, "learning_rate": 5.922500000000001e-05, "loss": 0.1258, "step": 2370 }, { "grad_norm": 0.6594567894935608, "learning_rate": 5.9475e-05, "loss": 0.1245, "step": 2380 }, { "grad_norm": 1.0201544761657715, "learning_rate": 5.9724999999999995e-05, "loss": 0.1529, "step": 2390 }, { "grad_norm": 0.940999448299408, "learning_rate": 5.9975e-05, "loss": 0.1154, "step": 2400 }, { "grad_norm": 0.7241795063018799, "learning_rate": 6.0225e-05, "loss": 0.1197, "step": 2410 }, { "grad_norm": 0.9842861294746399, "learning_rate": 6.0475000000000006e-05, "loss": 0.1614, "step": 2420 }, { "grad_norm": 1.370078682899475, "learning_rate": 6.0725e-05, "loss": 0.1352, "step": 2430 }, { "grad_norm": 1.0890527963638306, "learning_rate": 6.0975000000000005e-05, "loss": 0.138, "step": 2440 }, { "grad_norm": 1.3318946361541748, "learning_rate": 6.1225e-05, "loss": 0.113, "step": 2450 }, { "grad_norm": 1.1467704772949219, "learning_rate": 6.1475e-05, "loss": 0.1207, "step": 2460 }, { "grad_norm": 1.0825399160385132, "learning_rate": 6.1725e-05, "loss": 0.1214, "step": 2470 }, { "grad_norm": 1.0997929573059082, "learning_rate": 6.1975e-05, "loss": 0.141, "step": 2480 }, { "grad_norm": 1.1544204950332642, "learning_rate": 6.2225e-05, "loss": 0.1288, "step": 2490 }, { "grad_norm": 0.7886078357696533, "learning_rate": 6.2475e-05, "loss": 0.1276, "step": 2500 }, { "grad_norm": 0.8849567770957947, "learning_rate": 6.2725e-05, "loss": 0.1064, "step": 2510 }, { "grad_norm": 1.058979868888855, "learning_rate": 6.297500000000001e-05, "loss": 0.1263, "step": 2520 }, { "grad_norm": 0.8957341313362122, "learning_rate": 6.3225e-05, "loss": 0.1151, "step": 2530 }, { "grad_norm": 1.037109375, "learning_rate": 6.347500000000001e-05, "loss": 0.1179, "step": 2540 }, { "grad_norm": 1.0356630086898804, "learning_rate": 6.3725e-05, "loss": 0.1365, "step": 2550 }, { "grad_norm": 0.9281898140907288, "learning_rate": 6.397500000000001e-05, "loss": 0.1277, "step": 2560 }, { "grad_norm": 0.9895304441452026, "learning_rate": 6.4225e-05, "loss": 0.1171, "step": 2570 }, { "grad_norm": 1.0366719961166382, "learning_rate": 6.447500000000001e-05, "loss": 0.1216, "step": 2580 }, { "grad_norm": 0.7852692604064941, "learning_rate": 6.4725e-05, "loss": 0.1102, "step": 2590 }, { "grad_norm": 1.2970002889633179, "learning_rate": 6.497500000000001e-05, "loss": 0.1268, "step": 2600 }, { "grad_norm": 0.9774898290634155, "learning_rate": 6.5225e-05, "loss": 0.1296, "step": 2610 }, { "grad_norm": 0.850584089756012, "learning_rate": 6.5475e-05, "loss": 0.1088, "step": 2620 }, { "grad_norm": 1.057862639427185, "learning_rate": 6.5725e-05, "loss": 0.1176, "step": 2630 }, { "grad_norm": 0.8347893953323364, "learning_rate": 6.5975e-05, "loss": 0.1379, "step": 2640 }, { "grad_norm": 0.9165388345718384, "learning_rate": 6.6225e-05, "loss": 0.1415, "step": 2650 }, { "grad_norm": 0.9493004679679871, "learning_rate": 6.6475e-05, "loss": 0.1258, "step": 2660 }, { "grad_norm": 0.8313294649124146, "learning_rate": 6.672500000000001e-05, "loss": 0.1259, "step": 2670 }, { "grad_norm": 0.9413134455680847, "learning_rate": 6.6975e-05, "loss": 0.1529, "step": 2680 }, { "grad_norm": 0.8114683628082275, "learning_rate": 6.722500000000001e-05, "loss": 0.1328, "step": 2690 }, { "grad_norm": 0.9074441194534302, "learning_rate": 6.7475e-05, "loss": 0.1123, "step": 2700 }, { "grad_norm": 0.9215481877326965, "learning_rate": 6.7725e-05, "loss": 0.1138, "step": 2710 }, { "grad_norm": 0.8135960698127747, "learning_rate": 6.7975e-05, "loss": 0.1401, "step": 2720 }, { "grad_norm": 1.2024155855178833, "learning_rate": 6.8225e-05, "loss": 0.1193, "step": 2730 }, { "grad_norm": 0.6929173469543457, "learning_rate": 6.8475e-05, "loss": 0.1137, "step": 2740 }, { "grad_norm": 0.7699432373046875, "learning_rate": 6.8725e-05, "loss": 0.1311, "step": 2750 }, { "grad_norm": 0.9267508387565613, "learning_rate": 6.8975e-05, "loss": 0.1127, "step": 2760 }, { "grad_norm": 0.9777426719665527, "learning_rate": 6.9225e-05, "loss": 0.1149, "step": 2770 }, { "grad_norm": 0.8240177035331726, "learning_rate": 6.9475e-05, "loss": 0.1314, "step": 2780 }, { "grad_norm": 0.7133148312568665, "learning_rate": 6.9725e-05, "loss": 0.1234, "step": 2790 }, { "grad_norm": 0.8226393461227417, "learning_rate": 6.997500000000001e-05, "loss": 0.1164, "step": 2800 }, { "grad_norm": 1.1519668102264404, "learning_rate": 7.022500000000001e-05, "loss": 0.1431, "step": 2810 }, { "grad_norm": 1.0757337808609009, "learning_rate": 7.0475e-05, "loss": 0.1373, "step": 2820 }, { "grad_norm": 0.8559907078742981, "learning_rate": 7.072500000000001e-05, "loss": 0.1071, "step": 2830 }, { "grad_norm": 0.7973321080207825, "learning_rate": 7.0975e-05, "loss": 0.1412, "step": 2840 }, { "grad_norm": 0.7857217788696289, "learning_rate": 7.122500000000001e-05, "loss": 0.1319, "step": 2850 }, { "grad_norm": 0.8988924026489258, "learning_rate": 7.1475e-05, "loss": 0.1261, "step": 2860 }, { "grad_norm": 0.6902256608009338, "learning_rate": 7.172500000000001e-05, "loss": 0.1077, "step": 2870 }, { "grad_norm": 0.726452648639679, "learning_rate": 7.1975e-05, "loss": 0.0964, "step": 2880 }, { "grad_norm": 0.9123789072036743, "learning_rate": 7.2225e-05, "loss": 0.1167, "step": 2890 }, { "grad_norm": 1.0389201641082764, "learning_rate": 7.2475e-05, "loss": 0.1266, "step": 2900 }, { "grad_norm": 1.056753396987915, "learning_rate": 7.272499999999999e-05, "loss": 0.1089, "step": 2910 }, { "grad_norm": 0.7325246334075928, "learning_rate": 7.2975e-05, "loss": 0.138, "step": 2920 }, { "grad_norm": 0.7351200580596924, "learning_rate": 7.3225e-05, "loss": 0.098, "step": 2930 }, { "grad_norm": 0.7859359383583069, "learning_rate": 7.347500000000001e-05, "loss": 0.1021, "step": 2940 }, { "grad_norm": 0.6897473335266113, "learning_rate": 7.3725e-05, "loss": 0.1075, "step": 2950 }, { "grad_norm": 0.8587526679039001, "learning_rate": 7.397500000000001e-05, "loss": 0.12, "step": 2960 }, { "grad_norm": 0.5999446511268616, "learning_rate": 7.4225e-05, "loss": 0.1011, "step": 2970 }, { "grad_norm": 0.8309486508369446, "learning_rate": 7.447500000000001e-05, "loss": 0.1126, "step": 2980 }, { "grad_norm": 0.9084790945053101, "learning_rate": 7.4725e-05, "loss": 0.1241, "step": 2990 }, { "grad_norm": 0.8626994490623474, "learning_rate": 7.4975e-05, "loss": 0.1211, "step": 3000 }, { "grad_norm": 0.6196193099021912, "learning_rate": 7.5225e-05, "loss": 0.0988, "step": 3010 }, { "grad_norm": 0.8170282244682312, "learning_rate": 7.5475e-05, "loss": 0.1129, "step": 3020 }, { "grad_norm": 0.6310524940490723, "learning_rate": 7.5725e-05, "loss": 0.1107, "step": 3030 }, { "grad_norm": 1.0861133337020874, "learning_rate": 7.5975e-05, "loss": 0.116, "step": 3040 }, { "grad_norm": 1.0807808637619019, "learning_rate": 7.6225e-05, "loss": 0.13, "step": 3050 }, { "grad_norm": 1.0874049663543701, "learning_rate": 7.6475e-05, "loss": 0.1346, "step": 3060 }, { "grad_norm": 0.6930229067802429, "learning_rate": 7.672500000000001e-05, "loss": 0.1138, "step": 3070 }, { "grad_norm": 0.6965377926826477, "learning_rate": 7.697500000000001e-05, "loss": 0.1019, "step": 3080 }, { "grad_norm": 0.9851542115211487, "learning_rate": 7.722500000000001e-05, "loss": 0.1333, "step": 3090 }, { "grad_norm": 0.9566983580589294, "learning_rate": 7.747500000000001e-05, "loss": 0.1088, "step": 3100 }, { "grad_norm": 1.046026349067688, "learning_rate": 7.7725e-05, "loss": 0.1003, "step": 3110 }, { "grad_norm": 1.0635675191879272, "learning_rate": 7.797500000000001e-05, "loss": 0.1118, "step": 3120 }, { "grad_norm": 0.8655579686164856, "learning_rate": 7.8225e-05, "loss": 0.1075, "step": 3130 }, { "grad_norm": 0.8679106831550598, "learning_rate": 7.8475e-05, "loss": 0.1224, "step": 3140 }, { "grad_norm": 0.6144790649414062, "learning_rate": 7.8725e-05, "loss": 0.1047, "step": 3150 }, { "grad_norm": 0.7917160391807556, "learning_rate": 7.8975e-05, "loss": 0.1091, "step": 3160 }, { "grad_norm": 0.6577690243721008, "learning_rate": 7.9225e-05, "loss": 0.1042, "step": 3170 }, { "grad_norm": 0.7563655376434326, "learning_rate": 7.9475e-05, "loss": 0.1187, "step": 3180 }, { "grad_norm": 0.7004166841506958, "learning_rate": 7.9725e-05, "loss": 0.116, "step": 3190 }, { "grad_norm": 0.8748149871826172, "learning_rate": 7.9975e-05, "loss": 0.117, "step": 3200 }, { "grad_norm": 0.5892912745475769, "learning_rate": 8.022500000000001e-05, "loss": 0.0974, "step": 3210 }, { "grad_norm": 0.9360044598579407, "learning_rate": 8.0475e-05, "loss": 0.1336, "step": 3220 }, { "grad_norm": 0.8930726051330566, "learning_rate": 8.072500000000001e-05, "loss": 0.1239, "step": 3230 }, { "grad_norm": 0.8324586153030396, "learning_rate": 8.0975e-05, "loss": 0.1087, "step": 3240 }, { "grad_norm": 0.8108780384063721, "learning_rate": 8.122500000000001e-05, "loss": 0.123, "step": 3250 }, { "grad_norm": 0.9115781784057617, "learning_rate": 8.1475e-05, "loss": 0.1172, "step": 3260 }, { "grad_norm": 0.9836762547492981, "learning_rate": 8.172500000000001e-05, "loss": 0.1263, "step": 3270 }, { "grad_norm": 0.802152693271637, "learning_rate": 8.1975e-05, "loss": 0.1095, "step": 3280 }, { "grad_norm": 0.8255699872970581, "learning_rate": 8.2225e-05, "loss": 0.1122, "step": 3290 }, { "grad_norm": 0.8026905059814453, "learning_rate": 8.2475e-05, "loss": 0.1135, "step": 3300 }, { "grad_norm": 0.7863026857376099, "learning_rate": 8.2725e-05, "loss": 0.1246, "step": 3310 }, { "grad_norm": 0.7901190519332886, "learning_rate": 8.2975e-05, "loss": 0.1413, "step": 3320 }, { "grad_norm": 0.7402108907699585, "learning_rate": 8.3225e-05, "loss": 0.1124, "step": 3330 }, { "grad_norm": 0.8288959264755249, "learning_rate": 8.347500000000001e-05, "loss": 0.1129, "step": 3340 }, { "grad_norm": 0.6901365518569946, "learning_rate": 8.3725e-05, "loss": 0.1166, "step": 3350 }, { "grad_norm": 0.7695791721343994, "learning_rate": 8.397500000000001e-05, "loss": 0.1069, "step": 3360 }, { "grad_norm": 0.6382176280021667, "learning_rate": 8.422500000000001e-05, "loss": 0.1126, "step": 3370 }, { "grad_norm": 0.6868104338645935, "learning_rate": 8.447500000000001e-05, "loss": 0.1375, "step": 3380 }, { "grad_norm": 0.7530983686447144, "learning_rate": 8.4725e-05, "loss": 0.1076, "step": 3390 }, { "grad_norm": 0.6837260723114014, "learning_rate": 8.4975e-05, "loss": 0.1122, "step": 3400 }, { "grad_norm": 0.6573678255081177, "learning_rate": 8.5225e-05, "loss": 0.1078, "step": 3410 }, { "grad_norm": 0.6465964317321777, "learning_rate": 8.5475e-05, "loss": 0.1028, "step": 3420 }, { "grad_norm": 0.9325146079063416, "learning_rate": 8.5725e-05, "loss": 0.117, "step": 3430 }, { "grad_norm": 0.6540305018424988, "learning_rate": 8.5975e-05, "loss": 0.1126, "step": 3440 }, { "grad_norm": 0.6768518090248108, "learning_rate": 8.6225e-05, "loss": 0.1035, "step": 3450 }, { "grad_norm": 0.6523855328559875, "learning_rate": 8.6475e-05, "loss": 0.0995, "step": 3460 }, { "grad_norm": 0.8906623721122742, "learning_rate": 8.672500000000001e-05, "loss": 0.1215, "step": 3470 }, { "grad_norm": 0.7852122187614441, "learning_rate": 8.6975e-05, "loss": 0.1046, "step": 3480 }, { "grad_norm": 0.7141943573951721, "learning_rate": 8.7225e-05, "loss": 0.1078, "step": 3490 }, { "grad_norm": 0.6753293871879578, "learning_rate": 8.747500000000001e-05, "loss": 0.0935, "step": 3500 }, { "grad_norm": 0.8204021453857422, "learning_rate": 8.7725e-05, "loss": 0.1153, "step": 3510 }, { "grad_norm": 0.7482382655143738, "learning_rate": 8.797500000000001e-05, "loss": 0.1214, "step": 3520 }, { "grad_norm": 0.6249631643295288, "learning_rate": 8.8225e-05, "loss": 0.1156, "step": 3530 }, { "grad_norm": 0.7522429823875427, "learning_rate": 8.847500000000001e-05, "loss": 0.124, "step": 3540 }, { "grad_norm": 0.7697792053222656, "learning_rate": 8.8725e-05, "loss": 0.117, "step": 3550 }, { "grad_norm": 0.622020959854126, "learning_rate": 8.897500000000001e-05, "loss": 0.09, "step": 3560 }, { "grad_norm": 0.709520161151886, "learning_rate": 8.9225e-05, "loss": 0.1129, "step": 3570 }, { "grad_norm": 0.8700830936431885, "learning_rate": 8.9475e-05, "loss": 0.1259, "step": 3580 }, { "grad_norm": 0.9189584255218506, "learning_rate": 8.9725e-05, "loss": 0.1265, "step": 3590 }, { "grad_norm": 0.7378640174865723, "learning_rate": 8.9975e-05, "loss": 0.1193, "step": 3600 }, { "grad_norm": 0.950532078742981, "learning_rate": 9.0225e-05, "loss": 0.1257, "step": 3610 }, { "grad_norm": 0.7328602075576782, "learning_rate": 9.0475e-05, "loss": 0.118, "step": 3620 }, { "grad_norm": 0.6293395161628723, "learning_rate": 9.072500000000001e-05, "loss": 0.1043, "step": 3630 }, { "grad_norm": 0.6343711614608765, "learning_rate": 9.0975e-05, "loss": 0.11, "step": 3640 }, { "grad_norm": 0.6469329595565796, "learning_rate": 9.122500000000001e-05, "loss": 0.1209, "step": 3650 }, { "grad_norm": 0.895849883556366, "learning_rate": 9.1475e-05, "loss": 0.1049, "step": 3660 }, { "grad_norm": 0.9123350977897644, "learning_rate": 9.172500000000001e-05, "loss": 0.1192, "step": 3670 }, { "grad_norm": 0.9163255095481873, "learning_rate": 9.1975e-05, "loss": 0.1075, "step": 3680 }, { "grad_norm": 0.7404049634933472, "learning_rate": 9.2225e-05, "loss": 0.0925, "step": 3690 }, { "grad_norm": 0.8366824388504028, "learning_rate": 9.2475e-05, "loss": 0.1015, "step": 3700 }, { "grad_norm": 0.6964174509048462, "learning_rate": 9.2725e-05, "loss": 0.116, "step": 3710 }, { "grad_norm": 0.7819623351097107, "learning_rate": 9.2975e-05, "loss": 0.1038, "step": 3720 }, { "grad_norm": 0.8979519009590149, "learning_rate": 9.3225e-05, "loss": 0.1053, "step": 3730 }, { "grad_norm": 0.8042652010917664, "learning_rate": 9.3475e-05, "loss": 0.1016, "step": 3740 }, { "grad_norm": 0.6529334187507629, "learning_rate": 9.3725e-05, "loss": 0.1186, "step": 3750 }, { "grad_norm": 0.5947971940040588, "learning_rate": 9.397500000000001e-05, "loss": 0.1044, "step": 3760 }, { "grad_norm": 0.7033646106719971, "learning_rate": 9.422500000000001e-05, "loss": 0.1161, "step": 3770 }, { "grad_norm": 0.8456103205680847, "learning_rate": 9.4475e-05, "loss": 0.1229, "step": 3780 }, { "grad_norm": 0.7358851432800293, "learning_rate": 9.472500000000001e-05, "loss": 0.0975, "step": 3790 }, { "grad_norm": 0.6940839290618896, "learning_rate": 9.4975e-05, "loss": 0.1346, "step": 3800 }, { "grad_norm": 0.7400931119918823, "learning_rate": 9.522500000000001e-05, "loss": 0.096, "step": 3810 }, { "grad_norm": 0.8584976196289062, "learning_rate": 9.5475e-05, "loss": 0.108, "step": 3820 }, { "grad_norm": 0.6099553108215332, "learning_rate": 9.572500000000001e-05, "loss": 0.0977, "step": 3830 }, { "grad_norm": 0.73540860414505, "learning_rate": 9.5975e-05, "loss": 0.1007, "step": 3840 }, { "grad_norm": 0.7388766407966614, "learning_rate": 9.622500000000001e-05, "loss": 0.1059, "step": 3850 }, { "grad_norm": 0.8991512656211853, "learning_rate": 9.6475e-05, "loss": 0.1033, "step": 3860 }, { "grad_norm": 0.9238452911376953, "learning_rate": 9.6725e-05, "loss": 0.0936, "step": 3870 }, { "grad_norm": 1.029613971710205, "learning_rate": 9.6975e-05, "loss": 0.0958, "step": 3880 }, { "grad_norm": 0.6654434204101562, "learning_rate": 9.7225e-05, "loss": 0.1112, "step": 3890 }, { "grad_norm": 0.671696126461029, "learning_rate": 9.747500000000001e-05, "loss": 0.1108, "step": 3900 }, { "grad_norm": 0.7442033886909485, "learning_rate": 9.7725e-05, "loss": 0.1208, "step": 3910 }, { "grad_norm": 0.9302428960800171, "learning_rate": 9.797500000000001e-05, "loss": 0.1211, "step": 3920 }, { "grad_norm": 0.8755017518997192, "learning_rate": 9.8225e-05, "loss": 0.1199, "step": 3930 }, { "grad_norm": 0.8395757675170898, "learning_rate": 9.847500000000001e-05, "loss": 0.1311, "step": 3940 }, { "grad_norm": 0.775713324546814, "learning_rate": 9.8725e-05, "loss": 0.1188, "step": 3950 }, { "grad_norm": 0.697736382484436, "learning_rate": 9.897500000000001e-05, "loss": 0.1293, "step": 3960 }, { "grad_norm": 0.7447744011878967, "learning_rate": 9.9225e-05, "loss": 0.0957, "step": 3970 }, { "grad_norm": 0.8380301594734192, "learning_rate": 9.9475e-05, "loss": 0.128, "step": 3980 }, { "grad_norm": 0.9440038204193115, "learning_rate": 9.9725e-05, "loss": 0.1306, "step": 3990 }, { "grad_norm": 0.956730306148529, "learning_rate": 9.9975e-05, "loss": 0.1259, "step": 4000 }, { "grad_norm": 0.654318630695343, "learning_rate": 9.999999653982884e-05, "loss": 0.1076, "step": 4010 }, { "grad_norm": 0.6905200481414795, "learning_rate": 9.999998457874392e-05, "loss": 0.105, "step": 4020 }, { "grad_norm": 0.8338927626609802, "learning_rate": 9.999996407402913e-05, "loss": 0.1196, "step": 4030 }, { "grad_norm": 0.5301967263221741, "learning_rate": 9.999993502568801e-05, "loss": 0.0926, "step": 4040 }, { "grad_norm": 0.776717483997345, "learning_rate": 9.999989743372548e-05, "loss": 0.1105, "step": 4050 }, { "grad_norm": 0.909007728099823, "learning_rate": 9.999985129814798e-05, "loss": 0.1161, "step": 4060 }, { "grad_norm": 0.7183696627616882, "learning_rate": 9.99997966189634e-05, "loss": 0.0972, "step": 4070 }, { "grad_norm": 0.7874764204025269, "learning_rate": 9.999973339618107e-05, "loss": 0.1191, "step": 4080 }, { "grad_norm": 0.8736000657081604, "learning_rate": 9.999966162981179e-05, "loss": 0.1015, "step": 4090 }, { "grad_norm": 0.6426836252212524, "learning_rate": 9.999958131986784e-05, "loss": 0.0965, "step": 4100 }, { "grad_norm": 0.8513848781585693, "learning_rate": 9.999949246636293e-05, "loss": 0.1037, "step": 4110 }, { "grad_norm": 0.6499559283256531, "learning_rate": 9.999939506931224e-05, "loss": 0.1159, "step": 4120 }, { "grad_norm": 0.603424608707428, "learning_rate": 9.999928912873243e-05, "loss": 0.1019, "step": 4130 }, { "grad_norm": 0.6842483282089233, "learning_rate": 9.999917464464159e-05, "loss": 0.1036, "step": 4140 }, { "grad_norm": 0.724621593952179, "learning_rate": 9.999905161705929e-05, "loss": 0.1218, "step": 4150 }, { "grad_norm": 0.8030492067337036, "learning_rate": 9.999892004600653e-05, "loss": 0.0944, "step": 4160 }, { "grad_norm": 0.6887372136116028, "learning_rate": 9.999877993150581e-05, "loss": 0.0958, "step": 4170 }, { "grad_norm": 0.8107737898826599, "learning_rate": 9.999863127358108e-05, "loss": 0.1036, "step": 4180 }, { "grad_norm": 0.7842081785202026, "learning_rate": 9.999847407225773e-05, "loss": 0.1173, "step": 4190 }, { "grad_norm": 0.8330972790718079, "learning_rate": 9.999830832756262e-05, "loss": 0.1188, "step": 4200 }, { "grad_norm": 0.7091305255889893, "learning_rate": 9.999813403952407e-05, "loss": 0.1049, "step": 4210 }, { "grad_norm": 0.6522459387779236, "learning_rate": 9.999795120817187e-05, "loss": 0.1114, "step": 4220 }, { "grad_norm": 0.7836375832557678, "learning_rate": 9.999775983353725e-05, "loss": 0.1252, "step": 4230 }, { "grad_norm": 0.5713881254196167, "learning_rate": 9.999755991565292e-05, "loss": 0.1159, "step": 4240 }, { "grad_norm": 0.6647621393203735, "learning_rate": 9.999735145455303e-05, "loss": 0.1175, "step": 4250 }, { "grad_norm": 0.802585780620575, "learning_rate": 9.99971344502732e-05, "loss": 0.1148, "step": 4260 }, { "grad_norm": 0.640131413936615, "learning_rate": 9.999690890285053e-05, "loss": 0.1044, "step": 4270 }, { "grad_norm": 0.7577351927757263, "learning_rate": 9.999667481232356e-05, "loss": 0.1014, "step": 4280 }, { "grad_norm": 0.8202180862426758, "learning_rate": 9.999643217873225e-05, "loss": 0.0989, "step": 4290 }, { "grad_norm": 0.7163066864013672, "learning_rate": 9.999618100211809e-05, "loss": 0.0944, "step": 4300 }, { "grad_norm": 0.7341286540031433, "learning_rate": 9.999592128252402e-05, "loss": 0.0936, "step": 4310 }, { "grad_norm": 0.8150373697280884, "learning_rate": 9.999565301999437e-05, "loss": 0.1223, "step": 4320 }, { "grad_norm": 0.6235664486885071, "learning_rate": 9.999537621457502e-05, "loss": 0.1031, "step": 4330 }, { "grad_norm": 0.5075170993804932, "learning_rate": 9.999509086631323e-05, "loss": 0.0865, "step": 4340 }, { "grad_norm": 0.6988273859024048, "learning_rate": 9.99947969752578e-05, "loss": 0.0926, "step": 4350 }, { "grad_norm": 0.6000894904136658, "learning_rate": 9.999449454145891e-05, "loss": 0.0975, "step": 4360 }, { "grad_norm": 0.7778390049934387, "learning_rate": 9.999418356496827e-05, "loss": 0.0927, "step": 4370 }, { "grad_norm": 0.6269701719284058, "learning_rate": 9.999386404583899e-05, "loss": 0.1076, "step": 4380 }, { "grad_norm": 0.6773362755775452, "learning_rate": 9.999353598412568e-05, "loss": 0.0925, "step": 4390 }, { "grad_norm": 0.5383150577545166, "learning_rate": 9.999319937988442e-05, "loss": 0.1082, "step": 4400 }, { "grad_norm": 0.8349189162254333, "learning_rate": 9.999285423317268e-05, "loss": 0.1015, "step": 4410 }, { "grad_norm": 0.6108231544494629, "learning_rate": 9.999250054404947e-05, "loss": 0.1093, "step": 4420 }, { "grad_norm": 0.6307428479194641, "learning_rate": 9.99921383125752e-05, "loss": 0.102, "step": 4430 }, { "grad_norm": 0.7155115604400635, "learning_rate": 9.99917675388118e-05, "loss": 0.1044, "step": 4440 }, { "grad_norm": 0.6564807891845703, "learning_rate": 9.99913882228226e-05, "loss": 0.1187, "step": 4450 }, { "grad_norm": 0.5921421647071838, "learning_rate": 9.999100036467242e-05, "loss": 0.1032, "step": 4460 }, { "grad_norm": 0.5411585569381714, "learning_rate": 9.999060396442753e-05, "loss": 0.103, "step": 4470 }, { "grad_norm": 0.6303417682647705, "learning_rate": 9.999019902215566e-05, "loss": 0.1134, "step": 4480 }, { "grad_norm": 0.6828523874282837, "learning_rate": 9.998978553792602e-05, "loss": 0.111, "step": 4490 }, { "grad_norm": 0.6598325371742249, "learning_rate": 9.998936351180926e-05, "loss": 0.1026, "step": 4500 }, { "grad_norm": 0.5982596278190613, "learning_rate": 9.998893294387747e-05, "loss": 0.1249, "step": 4510 }, { "grad_norm": 0.5990646481513977, "learning_rate": 9.998849383420426e-05, "loss": 0.1006, "step": 4520 }, { "grad_norm": 0.5051448345184326, "learning_rate": 9.998804618286465e-05, "loss": 0.0961, "step": 4530 }, { "grad_norm": 0.7178200483322144, "learning_rate": 9.99875899899351e-05, "loss": 0.1168, "step": 4540 }, { "grad_norm": 0.6541121602058411, "learning_rate": 9.99871252554936e-05, "loss": 0.1107, "step": 4550 }, { "grad_norm": 0.5533650517463684, "learning_rate": 9.998665197961955e-05, "loss": 0.0944, "step": 4560 }, { "grad_norm": 0.6944413185119629, "learning_rate": 9.998617016239379e-05, "loss": 0.1087, "step": 4570 }, { "grad_norm": 0.8968552350997925, "learning_rate": 9.998567980389869e-05, "loss": 0.1064, "step": 4580 }, { "grad_norm": 0.6860425472259521, "learning_rate": 9.998518090421802e-05, "loss": 0.0992, "step": 4590 }, { "grad_norm": 0.6173944473266602, "learning_rate": 9.998467346343703e-05, "loss": 0.1058, "step": 4600 }, { "grad_norm": 0.7420806288719177, "learning_rate": 9.998415748164243e-05, "loss": 0.0963, "step": 4610 }, { "grad_norm": 0.7079902291297913, "learning_rate": 9.998363295892238e-05, "loss": 0.0901, "step": 4620 }, { "grad_norm": 0.7289673089981079, "learning_rate": 9.998309989536652e-05, "loss": 0.111, "step": 4630 }, { "grad_norm": 0.6787611842155457, "learning_rate": 9.998255829106593e-05, "loss": 0.1012, "step": 4640 }, { "grad_norm": 0.6256413459777832, "learning_rate": 9.998200814611316e-05, "loss": 0.0826, "step": 4650 }, { "grad_norm": 0.7258480191230774, "learning_rate": 9.998144946060219e-05, "loss": 0.104, "step": 4660 }, { "grad_norm": 0.8481015563011169, "learning_rate": 9.998088223462852e-05, "loss": 0.0962, "step": 4670 }, { "grad_norm": 0.5217181444168091, "learning_rate": 9.998030646828905e-05, "loss": 0.0859, "step": 4680 }, { "grad_norm": 0.5354678630828857, "learning_rate": 9.997972216168217e-05, "loss": 0.1019, "step": 4690 }, { "grad_norm": 0.8314922451972961, "learning_rate": 9.997912931490771e-05, "loss": 0.0923, "step": 4700 }, { "grad_norm": 0.649173378944397, "learning_rate": 9.9978527928067e-05, "loss": 0.0913, "step": 4710 }, { "grad_norm": 0.6006035804748535, "learning_rate": 9.997791800126277e-05, "loss": 0.0909, "step": 4720 }, { "grad_norm": 0.8197608590126038, "learning_rate": 9.997729953459927e-05, "loss": 0.1189, "step": 4730 }, { "grad_norm": 0.6117826700210571, "learning_rate": 9.997667252818214e-05, "loss": 0.1016, "step": 4740 }, { "grad_norm": 0.5773656964302063, "learning_rate": 9.997603698211855e-05, "loss": 0.0794, "step": 4750 }, { "grad_norm": 1.214544653892517, "learning_rate": 9.99753928965171e-05, "loss": 0.0952, "step": 4760 }, { "grad_norm": 0.8136115670204163, "learning_rate": 9.997474027148781e-05, "loss": 0.1011, "step": 4770 }, { "grad_norm": 0.9374295473098755, "learning_rate": 9.997407910714223e-05, "loss": 0.1056, "step": 4780 }, { "grad_norm": 0.6015865802764893, "learning_rate": 9.997340940359332e-05, "loss": 0.0926, "step": 4790 }, { "grad_norm": 0.7256439328193665, "learning_rate": 9.997273116095552e-05, "loss": 0.0879, "step": 4800 }, { "grad_norm": 0.775257408618927, "learning_rate": 9.997204437934473e-05, "loss": 0.1026, "step": 4810 }, { "grad_norm": 0.6275255084037781, "learning_rate": 9.997134905887829e-05, "loss": 0.1028, "step": 4820 }, { "grad_norm": 0.9068632125854492, "learning_rate": 9.997064519967501e-05, "loss": 0.0991, "step": 4830 }, { "grad_norm": 0.9999529719352722, "learning_rate": 9.996993280185517e-05, "loss": 0.1115, "step": 4840 }, { "grad_norm": 0.8524535298347473, "learning_rate": 9.99692118655405e-05, "loss": 0.1037, "step": 4850 }, { "grad_norm": 0.5493072867393494, "learning_rate": 9.996848239085417e-05, "loss": 0.0872, "step": 4860 }, { "grad_norm": 0.6031056642532349, "learning_rate": 9.996774437792085e-05, "loss": 0.1012, "step": 4870 }, { "grad_norm": 0.6444302201271057, "learning_rate": 9.996699782686664e-05, "loss": 0.0891, "step": 4880 }, { "grad_norm": 0.764979898929596, "learning_rate": 9.996624273781909e-05, "loss": 0.1188, "step": 4890 }, { "grad_norm": 0.7246356010437012, "learning_rate": 9.996547911090725e-05, "loss": 0.0968, "step": 4900 }, { "grad_norm": 0.5261302590370178, "learning_rate": 9.996470694626157e-05, "loss": 0.1069, "step": 4910 }, { "grad_norm": 0.5655387043952942, "learning_rate": 9.996392624401403e-05, "loss": 0.1123, "step": 4920 }, { "grad_norm": 1.0294181108474731, "learning_rate": 9.996313700429801e-05, "loss": 0.097, "step": 4930 }, { "grad_norm": 0.5938118696212769, "learning_rate": 9.996233922724836e-05, "loss": 0.0882, "step": 4940 }, { "grad_norm": 0.6419538855552673, "learning_rate": 9.996153291300141e-05, "loss": 0.0935, "step": 4950 }, { "grad_norm": 0.6654691100120544, "learning_rate": 9.996071806169494e-05, "loss": 0.0981, "step": 4960 }, { "grad_norm": 0.6785321235656738, "learning_rate": 9.995989467346817e-05, "loss": 0.1042, "step": 4970 }, { "grad_norm": 0.4608953893184662, "learning_rate": 9.995906274846183e-05, "loss": 0.1034, "step": 4980 }, { "grad_norm": 0.8294247984886169, "learning_rate": 9.995822228681803e-05, "loss": 0.0971, "step": 4990 }, { "grad_norm": 0.7586800456047058, "learning_rate": 9.99573732886804e-05, "loss": 0.1041, "step": 5000 }, { "grad_norm": 0.5873317122459412, "learning_rate": 9.995651575419402e-05, "loss": 0.0994, "step": 5010 }, { "grad_norm": 0.743690550327301, "learning_rate": 9.995564968350541e-05, "loss": 0.0985, "step": 5020 }, { "grad_norm": 0.6709595918655396, "learning_rate": 9.995477507676256e-05, "loss": 0.0867, "step": 5030 }, { "grad_norm": 0.5809095501899719, "learning_rate": 9.995389193411493e-05, "loss": 0.0905, "step": 5040 }, { "grad_norm": 0.8104100227355957, "learning_rate": 9.995300025571339e-05, "loss": 0.102, "step": 5050 }, { "grad_norm": 0.6175168752670288, "learning_rate": 9.995210004171034e-05, "loss": 0.073, "step": 5060 }, { "grad_norm": 0.5577322840690613, "learning_rate": 9.995119129225956e-05, "loss": 0.1166, "step": 5070 }, { "grad_norm": 0.7184491753578186, "learning_rate": 9.995027400751637e-05, "loss": 0.1019, "step": 5080 }, { "grad_norm": 0.573292076587677, "learning_rate": 9.994934818763751e-05, "loss": 0.0854, "step": 5090 }, { "grad_norm": 0.6082231402397156, "learning_rate": 9.994841383278115e-05, "loss": 0.0878, "step": 5100 }, { "grad_norm": 0.5862478613853455, "learning_rate": 9.994747094310695e-05, "loss": 0.0988, "step": 5110 }, { "grad_norm": 0.6139740347862244, "learning_rate": 9.994651951877604e-05, "loss": 0.1108, "step": 5120 }, { "grad_norm": 0.6358111500740051, "learning_rate": 9.994555955995099e-05, "loss": 0.106, "step": 5130 }, { "grad_norm": 0.6261622309684753, "learning_rate": 9.994459106679581e-05, "loss": 0.0972, "step": 5140 }, { "grad_norm": 0.6327342987060547, "learning_rate": 9.994361403947603e-05, "loss": 0.0844, "step": 5150 }, { "grad_norm": 0.4969833195209503, "learning_rate": 9.994262847815854e-05, "loss": 0.1253, "step": 5160 }, { "grad_norm": 0.5726041793823242, "learning_rate": 9.99416343830118e-05, "loss": 0.0936, "step": 5170 }, { "grad_norm": 1.092392086982727, "learning_rate": 9.994063175420565e-05, "loss": 0.0966, "step": 5180 }, { "grad_norm": 0.8097949028015137, "learning_rate": 9.99396205919114e-05, "loss": 0.1283, "step": 5190 }, { "grad_norm": 0.5576263070106506, "learning_rate": 9.993860089630185e-05, "loss": 0.0876, "step": 5200 }, { "grad_norm": 0.7048825621604919, "learning_rate": 9.993757266755123e-05, "loss": 0.1035, "step": 5210 }, { "grad_norm": 0.7628759145736694, "learning_rate": 9.993653590583522e-05, "loss": 0.1135, "step": 5220 }, { "grad_norm": 0.8258781433105469, "learning_rate": 9.993549061133102e-05, "loss": 0.0867, "step": 5230 }, { "grad_norm": 0.6559590101242065, "learning_rate": 9.993443678421719e-05, "loss": 0.0918, "step": 5240 }, { "grad_norm": 0.6401426196098328, "learning_rate": 9.993337442467384e-05, "loss": 0.1158, "step": 5250 }, { "grad_norm": 0.6546245217323303, "learning_rate": 9.993230353288248e-05, "loss": 0.0874, "step": 5260 }, { "grad_norm": 0.701276421546936, "learning_rate": 9.993122410902608e-05, "loss": 0.0936, "step": 5270 }, { "grad_norm": 0.738377571105957, "learning_rate": 9.993013615328912e-05, "loss": 0.1147, "step": 5280 }, { "grad_norm": 0.7019073367118835, "learning_rate": 9.992903966585747e-05, "loss": 0.1107, "step": 5290 }, { "grad_norm": 0.5210531949996948, "learning_rate": 9.992793464691852e-05, "loss": 0.0897, "step": 5300 }, { "grad_norm": 0.5706728100776672, "learning_rate": 9.992682109666105e-05, "loss": 0.093, "step": 5310 }, { "grad_norm": 0.5324624180793762, "learning_rate": 9.992569901527538e-05, "loss": 0.0921, "step": 5320 }, { "grad_norm": 0.5395417809486389, "learning_rate": 9.99245684029532e-05, "loss": 0.1137, "step": 5330 }, { "grad_norm": 0.6429753303527832, "learning_rate": 9.992342925988774e-05, "loss": 0.1048, "step": 5340 }, { "grad_norm": 0.7836499810218811, "learning_rate": 9.992228158627361e-05, "loss": 0.0922, "step": 5350 }, { "grad_norm": 0.5512376427650452, "learning_rate": 9.992112538230693e-05, "loss": 0.0944, "step": 5360 }, { "grad_norm": 0.463318794965744, "learning_rate": 9.991996064818527e-05, "loss": 0.0939, "step": 5370 }, { "grad_norm": 0.7411442399024963, "learning_rate": 9.991878738410768e-05, "loss": 0.0902, "step": 5380 }, { "grad_norm": 0.5112854838371277, "learning_rate": 9.991760559027457e-05, "loss": 0.0904, "step": 5390 }, { "grad_norm": 0.7790564894676208, "learning_rate": 9.991641526688793e-05, "loss": 0.1014, "step": 5400 }, { "grad_norm": 0.521869957447052, "learning_rate": 9.991521641415113e-05, "loss": 0.0868, "step": 5410 }, { "grad_norm": 1.0100352764129639, "learning_rate": 9.991400903226904e-05, "loss": 0.0996, "step": 5420 }, { "grad_norm": 1.000142216682434, "learning_rate": 9.991279312144794e-05, "loss": 0.0838, "step": 5430 }, { "grad_norm": 0.6239786148071289, "learning_rate": 9.991156868189564e-05, "loss": 0.1044, "step": 5440 }, { "grad_norm": 0.5538811683654785, "learning_rate": 9.991033571382131e-05, "loss": 0.0843, "step": 5450 }, { "grad_norm": 0.7959867119789124, "learning_rate": 9.990909421743569e-05, "loss": 0.089, "step": 5460 }, { "grad_norm": 0.7101542353630066, "learning_rate": 9.990784419295085e-05, "loss": 0.0805, "step": 5470 }, { "grad_norm": 0.607963502407074, "learning_rate": 9.990658564058044e-05, "loss": 0.1012, "step": 5480 }, { "grad_norm": 0.6570059657096863, "learning_rate": 9.990531856053948e-05, "loss": 0.0992, "step": 5490 }, { "grad_norm": 0.7807677984237671, "learning_rate": 9.99040429530445e-05, "loss": 0.0948, "step": 5500 }, { "grad_norm": 0.7878811955451965, "learning_rate": 9.990275881831346e-05, "loss": 0.0896, "step": 5510 }, { "grad_norm": 0.5835066437721252, "learning_rate": 9.990146615656577e-05, "loss": 0.0824, "step": 5520 }, { "grad_norm": 0.6026182174682617, "learning_rate": 9.990016496802233e-05, "loss": 0.0943, "step": 5530 }, { "grad_norm": 0.686103343963623, "learning_rate": 9.989885525290548e-05, "loss": 0.0783, "step": 5540 }, { "grad_norm": 0.8524817228317261, "learning_rate": 9.989753701143897e-05, "loss": 0.082, "step": 5550 }, { "grad_norm": 0.5029655694961548, "learning_rate": 9.989621024384812e-05, "loss": 0.086, "step": 5560 }, { "grad_norm": 0.5475033521652222, "learning_rate": 9.989487495035959e-05, "loss": 0.0919, "step": 5570 }, { "grad_norm": 0.7506188154220581, "learning_rate": 9.989353113120156e-05, "loss": 0.1077, "step": 5580 }, { "grad_norm": 0.6919757127761841, "learning_rate": 9.989217878660366e-05, "loss": 0.0963, "step": 5590 }, { "grad_norm": 0.5959286093711853, "learning_rate": 9.989081791679695e-05, "loss": 0.0885, "step": 5600 }, { "grad_norm": 0.6099330186843872, "learning_rate": 9.988944852201397e-05, "loss": 0.0886, "step": 5610 }, { "grad_norm": 0.47548604011535645, "learning_rate": 9.988807060248873e-05, "loss": 0.0852, "step": 5620 }, { "grad_norm": 0.5842523574829102, "learning_rate": 9.988668415845665e-05, "loss": 0.0703, "step": 5630 }, { "grad_norm": 0.688912570476532, "learning_rate": 9.988528919015466e-05, "loss": 0.0833, "step": 5640 }, { "grad_norm": 0.5260763168334961, "learning_rate": 9.988388569782112e-05, "loss": 0.0989, "step": 5650 }, { "grad_norm": 0.5700973272323608, "learning_rate": 9.988247368169583e-05, "loss": 0.0844, "step": 5660 }, { "grad_norm": 0.7321475148200989, "learning_rate": 9.988105314202007e-05, "loss": 0.0832, "step": 5670 }, { "grad_norm": 0.5577324032783508, "learning_rate": 9.987962407903659e-05, "loss": 0.105, "step": 5680 }, { "grad_norm": 0.6606760621070862, "learning_rate": 9.987818649298957e-05, "loss": 0.087, "step": 5690 }, { "grad_norm": 0.474109411239624, "learning_rate": 9.987674038412465e-05, "loss": 0.0783, "step": 5700 }, { "grad_norm": 0.6337378621101379, "learning_rate": 9.987528575268891e-05, "loss": 0.0957, "step": 5710 }, { "grad_norm": 0.5255566835403442, "learning_rate": 9.987382259893095e-05, "loss": 0.0882, "step": 5720 }, { "grad_norm": 0.5392391681671143, "learning_rate": 9.987235092310074e-05, "loss": 0.0868, "step": 5730 }, { "grad_norm": 0.5863959789276123, "learning_rate": 9.987087072544978e-05, "loss": 0.108, "step": 5740 }, { "grad_norm": 0.7096166610717773, "learning_rate": 9.9869382006231e-05, "loss": 0.0821, "step": 5750 }, { "grad_norm": 0.7396399974822998, "learning_rate": 9.986788476569875e-05, "loss": 0.0954, "step": 5760 }, { "grad_norm": 1.0520908832550049, "learning_rate": 9.986637900410887e-05, "loss": 0.0952, "step": 5770 }, { "grad_norm": 0.5319347977638245, "learning_rate": 9.986486472171869e-05, "loss": 0.1063, "step": 5780 }, { "grad_norm": 0.7535521984100342, "learning_rate": 9.986334191878692e-05, "loss": 0.0946, "step": 5790 }, { "grad_norm": 0.5892050862312317, "learning_rate": 9.986181059557378e-05, "loss": 0.1016, "step": 5800 }, { "grad_norm": 0.6074603199958801, "learning_rate": 9.986027075234094e-05, "loss": 0.0883, "step": 5810 }, { "grad_norm": 0.6603528261184692, "learning_rate": 9.985872238935152e-05, "loss": 0.0868, "step": 5820 }, { "grad_norm": 0.8723049163818359, "learning_rate": 9.985716550687008e-05, "loss": 0.0974, "step": 5830 }, { "grad_norm": 0.5515657067298889, "learning_rate": 9.985560010516264e-05, "loss": 0.0844, "step": 5840 }, { "grad_norm": 0.7783278822898865, "learning_rate": 9.985402618449668e-05, "loss": 0.105, "step": 5850 }, { "grad_norm": 0.6559131741523743, "learning_rate": 9.985244374514118e-05, "loss": 0.118, "step": 5860 }, { "grad_norm": 0.4525586664676666, "learning_rate": 9.985085278736651e-05, "loss": 0.1097, "step": 5870 }, { "grad_norm": 0.6490414142608643, "learning_rate": 9.984925331144452e-05, "loss": 0.1098, "step": 5880 }, { "grad_norm": 0.5167040824890137, "learning_rate": 9.984764531764851e-05, "loss": 0.093, "step": 5890 }, { "grad_norm": 0.6130520105361938, "learning_rate": 9.984602880625326e-05, "loss": 0.1022, "step": 5900 }, { "grad_norm": 0.42590460181236267, "learning_rate": 9.9844403777535e-05, "loss": 0.0895, "step": 5910 }, { "grad_norm": 0.6400953531265259, "learning_rate": 9.984277023177135e-05, "loss": 0.0824, "step": 5920 }, { "grad_norm": 0.5437423586845398, "learning_rate": 9.984112816924148e-05, "loss": 0.0716, "step": 5930 }, { "grad_norm": 0.6007696986198425, "learning_rate": 9.983947759022596e-05, "loss": 0.0873, "step": 5940 }, { "grad_norm": 0.5442800521850586, "learning_rate": 9.983781849500682e-05, "loss": 0.0777, "step": 5950 }, { "grad_norm": 0.4387134611606598, "learning_rate": 9.98361508838676e-05, "loss": 0.0795, "step": 5960 }, { "grad_norm": 0.7410618662834167, "learning_rate": 9.98344747570932e-05, "loss": 0.088, "step": 5970 }, { "grad_norm": 0.4964122772216797, "learning_rate": 9.983279011497004e-05, "loss": 0.0946, "step": 5980 }, { "grad_norm": 0.7784399390220642, "learning_rate": 9.983109695778596e-05, "loss": 0.1053, "step": 5990 }, { "grad_norm": 0.565169632434845, "learning_rate": 9.982939528583032e-05, "loss": 0.1079, "step": 6000 }, { "grad_norm": 0.5990253686904907, "learning_rate": 9.982768509939385e-05, "loss": 0.0983, "step": 6010 }, { "grad_norm": 0.5134806632995605, "learning_rate": 9.982596639876879e-05, "loss": 0.0978, "step": 6020 }, { "grad_norm": 0.6633427739143372, "learning_rate": 9.982423918424881e-05, "loss": 0.1013, "step": 6030 }, { "grad_norm": 0.6665347218513489, "learning_rate": 9.982250345612908e-05, "loss": 0.0918, "step": 6040 }, { "grad_norm": 0.6216313242912292, "learning_rate": 9.982075921470611e-05, "loss": 0.1063, "step": 6050 }, { "grad_norm": 0.6217118501663208, "learning_rate": 9.981900646027802e-05, "loss": 0.0901, "step": 6060 }, { "grad_norm": 0.6999354362487793, "learning_rate": 9.981724519314425e-05, "loss": 0.0951, "step": 6070 }, { "grad_norm": 0.6256528496742249, "learning_rate": 9.981547541360581e-05, "loss": 0.1007, "step": 6080 }, { "grad_norm": 0.5362479090690613, "learning_rate": 9.981369712196508e-05, "loss": 0.1009, "step": 6090 }, { "grad_norm": 0.6865504384040833, "learning_rate": 9.981191031852592e-05, "loss": 0.0969, "step": 6100 }, { "grad_norm": 0.623001754283905, "learning_rate": 9.981011500359362e-05, "loss": 0.0861, "step": 6110 }, { "grad_norm": 0.6705141067504883, "learning_rate": 9.9808311177475e-05, "loss": 0.0865, "step": 6120 }, { "grad_norm": 0.7764915227890015, "learning_rate": 9.980649884047826e-05, "loss": 0.0936, "step": 6130 }, { "grad_norm": 0.6749064922332764, "learning_rate": 9.980467799291307e-05, "loss": 0.1038, "step": 6140 }, { "grad_norm": 1.03509521484375, "learning_rate": 9.980284863509058e-05, "loss": 0.1084, "step": 6150 }, { "grad_norm": 0.7284554243087769, "learning_rate": 9.980101076732334e-05, "loss": 0.1384, "step": 6160 }, { "grad_norm": 0.8236603140830994, "learning_rate": 9.979916438992544e-05, "loss": 0.1071, "step": 6170 }, { "grad_norm": 0.5409734845161438, "learning_rate": 9.979730950321237e-05, "loss": 0.0931, "step": 6180 }, { "grad_norm": 1.0628165006637573, "learning_rate": 9.979544610750104e-05, "loss": 0.1038, "step": 6190 }, { "grad_norm": 0.6557047963142395, "learning_rate": 9.97935742031099e-05, "loss": 0.0931, "step": 6200 }, { "grad_norm": 0.8035244941711426, "learning_rate": 9.979169379035878e-05, "loss": 0.0985, "step": 6210 }, { "grad_norm": 0.729661226272583, "learning_rate": 9.978980486956899e-05, "loss": 0.0875, "step": 6220 }, { "grad_norm": 0.5448691248893738, "learning_rate": 9.978790744106332e-05, "loss": 0.0862, "step": 6230 }, { "grad_norm": 0.5394471287727356, "learning_rate": 9.978600150516594e-05, "loss": 0.098, "step": 6240 }, { "grad_norm": 0.5322755575180054, "learning_rate": 9.978408706220259e-05, "loss": 0.1073, "step": 6250 }, { "grad_norm": 0.7940658926963806, "learning_rate": 9.978216411250032e-05, "loss": 0.0862, "step": 6260 }, { "grad_norm": 0.7208337187767029, "learning_rate": 9.978023265638778e-05, "loss": 0.0957, "step": 6270 }, { "grad_norm": 0.6360674500465393, "learning_rate": 9.977829269419495e-05, "loss": 0.0838, "step": 6280 }, { "grad_norm": 0.5619463324546814, "learning_rate": 9.977634422625335e-05, "loss": 0.0855, "step": 6290 }, { "grad_norm": 0.7428796887397766, "learning_rate": 9.97743872528959e-05, "loss": 0.099, "step": 6300 }, { "grad_norm": 0.6301143169403076, "learning_rate": 9.9772421774457e-05, "loss": 0.0899, "step": 6310 }, { "grad_norm": 0.6627640724182129, "learning_rate": 9.977044779127252e-05, "loss": 0.0938, "step": 6320 }, { "grad_norm": 0.5637080073356628, "learning_rate": 9.976846530367971e-05, "loss": 0.0914, "step": 6330 }, { "grad_norm": 0.6021554470062256, "learning_rate": 9.976647431201735e-05, "loss": 0.0951, "step": 6340 }, { "grad_norm": 0.5805454850196838, "learning_rate": 9.976447481662568e-05, "loss": 0.0831, "step": 6350 }, { "grad_norm": 0.5076398253440857, "learning_rate": 9.976246681784629e-05, "loss": 0.0798, "step": 6360 }, { "grad_norm": 0.485005646944046, "learning_rate": 9.976045031602234e-05, "loss": 0.0847, "step": 6370 }, { "grad_norm": 0.653906524181366, "learning_rate": 9.975842531149837e-05, "loss": 0.0855, "step": 6380 }, { "grad_norm": 0.6833205223083496, "learning_rate": 9.975639180462043e-05, "loss": 0.1023, "step": 6390 }, { "grad_norm": 0.45179957151412964, "learning_rate": 9.975434979573596e-05, "loss": 0.0865, "step": 6400 }, { "grad_norm": 0.5426590442657471, "learning_rate": 9.97522992851939e-05, "loss": 0.0749, "step": 6410 }, { "grad_norm": 0.46826907992362976, "learning_rate": 9.975024027334461e-05, "loss": 0.0805, "step": 6420 }, { "grad_norm": 0.5780354738235474, "learning_rate": 9.974817276053993e-05, "loss": 0.0891, "step": 6430 }, { "grad_norm": 0.6477095484733582, "learning_rate": 9.974609674713315e-05, "loss": 0.0866, "step": 6440 }, { "grad_norm": 0.7683902978897095, "learning_rate": 9.9744012233479e-05, "loss": 0.0975, "step": 6450 }, { "grad_norm": 0.5443863272666931, "learning_rate": 9.974191921993366e-05, "loss": 0.0933, "step": 6460 }, { "grad_norm": 0.5735291838645935, "learning_rate": 9.973981770685474e-05, "loss": 0.0952, "step": 6470 }, { "grad_norm": 0.6258144974708557, "learning_rate": 9.97377076946014e-05, "loss": 0.0825, "step": 6480 }, { "grad_norm": 0.7116514444351196, "learning_rate": 9.973558918353412e-05, "loss": 0.1195, "step": 6490 }, { "grad_norm": 0.7916361093521118, "learning_rate": 9.973346217401494e-05, "loss": 0.0811, "step": 6500 }, { "grad_norm": 0.4577432870864868, "learning_rate": 9.973132666640726e-05, "loss": 0.0779, "step": 6510 }, { "grad_norm": 0.40155041217803955, "learning_rate": 9.972918266107602e-05, "loss": 0.0803, "step": 6520 }, { "grad_norm": 0.6592113971710205, "learning_rate": 9.972703015838756e-05, "loss": 0.0917, "step": 6530 }, { "grad_norm": 0.7155427932739258, "learning_rate": 9.97248691587097e-05, "loss": 0.1034, "step": 6540 }, { "grad_norm": 0.598646879196167, "learning_rate": 9.972269966241166e-05, "loss": 0.0882, "step": 6550 }, { "grad_norm": 0.632796585559845, "learning_rate": 9.972052166986417e-05, "loss": 0.092, "step": 6560 }, { "grad_norm": 0.6573196053504944, "learning_rate": 9.971833518143938e-05, "loss": 0.0984, "step": 6570 }, { "grad_norm": 0.556714653968811, "learning_rate": 9.971614019751093e-05, "loss": 0.0776, "step": 6580 }, { "grad_norm": 0.6271039247512817, "learning_rate": 9.971393671845383e-05, "loss": 0.1021, "step": 6590 }, { "grad_norm": 0.7306895852088928, "learning_rate": 9.971172474464464e-05, "loss": 0.0967, "step": 6600 }, { "grad_norm": 0.5530976057052612, "learning_rate": 9.97095042764613e-05, "loss": 0.0743, "step": 6610 }, { "grad_norm": 0.5080495476722717, "learning_rate": 9.970727531428324e-05, "loss": 0.078, "step": 6620 }, { "grad_norm": 0.6833293437957764, "learning_rate": 9.970503785849132e-05, "loss": 0.0839, "step": 6630 }, { "grad_norm": 0.7806811928749084, "learning_rate": 9.970279190946788e-05, "loss": 0.0859, "step": 6640 }, { "grad_norm": 0.5671971440315247, "learning_rate": 9.970053746759667e-05, "loss": 0.1045, "step": 6650 }, { "grad_norm": 0.5018680691719055, "learning_rate": 9.969827453326292e-05, "loss": 0.0771, "step": 6660 }, { "grad_norm": 0.5193003416061401, "learning_rate": 9.969600310685332e-05, "loss": 0.0808, "step": 6670 }, { "grad_norm": 0.5510513186454773, "learning_rate": 9.969372318875596e-05, "loss": 0.096, "step": 6680 }, { "grad_norm": 0.5272420644760132, "learning_rate": 9.969143477936043e-05, "loss": 0.0966, "step": 6690 }, { "grad_norm": 0.6562424898147583, "learning_rate": 9.968913787905775e-05, "loss": 0.0934, "step": 6700 }, { "grad_norm": 0.8278453350067139, "learning_rate": 9.968683248824045e-05, "loss": 0.0892, "step": 6710 }, { "grad_norm": 0.5408686399459839, "learning_rate": 9.968451860730238e-05, "loss": 0.0902, "step": 6720 }, { "grad_norm": 0.7595577239990234, "learning_rate": 9.968219623663896e-05, "loss": 0.0841, "step": 6730 }, { "grad_norm": 0.8418247103691101, "learning_rate": 9.967986537664702e-05, "loss": 0.1234, "step": 6740 }, { "grad_norm": 0.5785623788833618, "learning_rate": 9.967752602772483e-05, "loss": 0.085, "step": 6750 }, { "grad_norm": 0.543594241142273, "learning_rate": 9.967517819027212e-05, "loss": 0.0828, "step": 6760 }, { "grad_norm": 0.6409075856208801, "learning_rate": 9.967282186469009e-05, "loss": 0.0775, "step": 6770 }, { "grad_norm": 0.5709226131439209, "learning_rate": 9.967045705138135e-05, "loss": 0.103, "step": 6780 }, { "grad_norm": 0.7213925719261169, "learning_rate": 9.966808375074998e-05, "loss": 0.0836, "step": 6790 }, { "grad_norm": 0.5187956690788269, "learning_rate": 9.966570196320154e-05, "loss": 0.102, "step": 6800 }, { "grad_norm": 0.7211189270019531, "learning_rate": 9.966331168914299e-05, "loss": 0.0862, "step": 6810 }, { "grad_norm": 0.5788689255714417, "learning_rate": 9.966091292898277e-05, "loss": 0.0925, "step": 6820 }, { "grad_norm": 0.6842591166496277, "learning_rate": 9.965850568313076e-05, "loss": 0.1025, "step": 6830 }, { "grad_norm": 0.6734321117401123, "learning_rate": 9.965608995199827e-05, "loss": 0.1012, "step": 6840 }, { "grad_norm": 0.5130741596221924, "learning_rate": 9.965366573599812e-05, "loss": 0.0812, "step": 6850 }, { "grad_norm": 0.5973415374755859, "learning_rate": 9.965123303554453e-05, "loss": 0.0791, "step": 6860 }, { "grad_norm": 0.7598349452018738, "learning_rate": 9.964879185105317e-05, "loss": 0.1054, "step": 6870 }, { "grad_norm": 0.8454861044883728, "learning_rate": 9.964634218294119e-05, "loss": 0.0982, "step": 6880 }, { "grad_norm": 0.5143091678619385, "learning_rate": 9.964388403162714e-05, "loss": 0.0843, "step": 6890 }, { "grad_norm": 0.6418402194976807, "learning_rate": 9.96414173975311e-05, "loss": 0.1057, "step": 6900 }, { "grad_norm": 0.7038930654525757, "learning_rate": 9.963894228107451e-05, "loss": 0.0875, "step": 6910 }, { "grad_norm": 0.5370308756828308, "learning_rate": 9.963645868268032e-05, "loss": 0.0924, "step": 6920 }, { "grad_norm": 0.5515524744987488, "learning_rate": 9.963396660277289e-05, "loss": 0.0958, "step": 6930 }, { "grad_norm": 0.5721067786216736, "learning_rate": 9.963146604177807e-05, "loss": 0.0831, "step": 6940 }, { "grad_norm": 0.39649906754493713, "learning_rate": 9.962895700012311e-05, "loss": 0.0766, "step": 6950 }, { "grad_norm": 0.504510223865509, "learning_rate": 9.962643947823677e-05, "loss": 0.0759, "step": 6960 }, { "grad_norm": 0.7469523549079895, "learning_rate": 9.962391347654921e-05, "loss": 0.096, "step": 6970 }, { "grad_norm": 0.5617644190788269, "learning_rate": 9.962137899549204e-05, "loss": 0.0872, "step": 6980 }, { "grad_norm": 0.6538245677947998, "learning_rate": 9.961883603549835e-05, "loss": 0.0733, "step": 6990 }, { "grad_norm": 0.5682275891304016, "learning_rate": 9.961628459700267e-05, "loss": 0.0917, "step": 7000 }, { "grad_norm": 0.49367156624794006, "learning_rate": 9.961372468044095e-05, "loss": 0.0914, "step": 7010 }, { "grad_norm": 0.614867091178894, "learning_rate": 9.961115628625062e-05, "loss": 0.0932, "step": 7020 }, { "grad_norm": 0.7495344281196594, "learning_rate": 9.960857941487056e-05, "loss": 0.1035, "step": 7030 }, { "grad_norm": 0.7037575244903564, "learning_rate": 9.960599406674106e-05, "loss": 0.0849, "step": 7040 }, { "grad_norm": 0.551581621170044, "learning_rate": 9.960340024230393e-05, "loss": 0.078, "step": 7050 }, { "grad_norm": 0.6938198804855347, "learning_rate": 9.960079794200232e-05, "loss": 0.0807, "step": 7060 }, { "grad_norm": 0.5711187720298767, "learning_rate": 9.959818716628096e-05, "loss": 0.0772, "step": 7070 }, { "grad_norm": 0.437593549489975, "learning_rate": 9.95955679155859e-05, "loss": 0.0968, "step": 7080 }, { "grad_norm": 0.7966926693916321, "learning_rate": 9.959294019036472e-05, "loss": 0.1058, "step": 7090 }, { "grad_norm": 0.6304031014442444, "learning_rate": 9.959030399106646e-05, "loss": 0.0869, "step": 7100 }, { "grad_norm": 0.47642722725868225, "learning_rate": 9.958765931814153e-05, "loss": 0.0868, "step": 7110 }, { "grad_norm": 0.49568718671798706, "learning_rate": 9.958500617204184e-05, "loss": 0.0689, "step": 7120 }, { "grad_norm": 0.7405747175216675, "learning_rate": 9.958234455322075e-05, "loss": 0.0761, "step": 7130 }, { "grad_norm": 0.5593428015708923, "learning_rate": 9.957967446213308e-05, "loss": 0.1155, "step": 7140 }, { "grad_norm": 0.5469794869422913, "learning_rate": 9.957699589923501e-05, "loss": 0.102, "step": 7150 }, { "grad_norm": 1.0593591928482056, "learning_rate": 9.957430886498431e-05, "loss": 0.1152, "step": 7160 }, { "grad_norm": 0.698313295841217, "learning_rate": 9.957161335984008e-05, "loss": 0.1021, "step": 7170 }, { "grad_norm": 0.6335567235946655, "learning_rate": 9.956890938426291e-05, "loss": 0.0839, "step": 7180 }, { "grad_norm": 0.5671165585517883, "learning_rate": 9.956619693871482e-05, "loss": 0.0919, "step": 7190 }, { "grad_norm": 0.5227290987968445, "learning_rate": 9.956347602365934e-05, "loss": 0.0854, "step": 7200 }, { "grad_norm": 0.671671450138092, "learning_rate": 9.956074663956135e-05, "loss": 0.0889, "step": 7210 }, { "grad_norm": 0.659842312335968, "learning_rate": 9.955800878688726e-05, "loss": 0.0963, "step": 7220 }, { "grad_norm": 0.5562698841094971, "learning_rate": 9.955526246610489e-05, "loss": 0.081, "step": 7230 }, { "grad_norm": 0.7411001324653625, "learning_rate": 9.955250767768349e-05, "loss": 0.0931, "step": 7240 }, { "grad_norm": 0.6721859574317932, "learning_rate": 9.95497444220938e-05, "loss": 0.0923, "step": 7250 }, { "grad_norm": 0.49203431606292725, "learning_rate": 9.954697269980797e-05, "loss": 0.0702, "step": 7260 }, { "grad_norm": 0.38848796486854553, "learning_rate": 9.954419251129962e-05, "loss": 0.0729, "step": 7270 }, { "grad_norm": 0.6487082242965698, "learning_rate": 9.95414038570438e-05, "loss": 0.0842, "step": 7280 }, { "grad_norm": 0.5126098394393921, "learning_rate": 9.953860673751703e-05, "loss": 0.0837, "step": 7290 }, { "grad_norm": 0.41901373863220215, "learning_rate": 9.953580115319725e-05, "loss": 0.0792, "step": 7300 }, { "grad_norm": 0.7278231382369995, "learning_rate": 9.953298710456387e-05, "loss": 0.0852, "step": 7310 }, { "grad_norm": 0.49166011810302734, "learning_rate": 9.953016459209771e-05, "loss": 0.0799, "step": 7320 }, { "grad_norm": 0.5362323522567749, "learning_rate": 9.952733361628108e-05, "loss": 0.0861, "step": 7330 }, { "grad_norm": 0.4875744581222534, "learning_rate": 9.952449417759772e-05, "loss": 0.0815, "step": 7340 }, { "grad_norm": 0.687927782535553, "learning_rate": 9.952164627653279e-05, "loss": 0.0822, "step": 7350 }, { "grad_norm": 0.6564334630966187, "learning_rate": 9.951878991357292e-05, "loss": 0.0735, "step": 7360 }, { "grad_norm": 0.6462326645851135, "learning_rate": 9.951592508920622e-05, "loss": 0.0969, "step": 7370 }, { "grad_norm": 0.6326659321784973, "learning_rate": 9.951305180392219e-05, "loss": 0.0772, "step": 7380 }, { "grad_norm": 0.6186691522598267, "learning_rate": 9.951017005821178e-05, "loss": 0.0839, "step": 7390 }, { "grad_norm": 0.4811253547668457, "learning_rate": 9.95072798525674e-05, "loss": 0.0816, "step": 7400 }, { "grad_norm": 0.6063529849052429, "learning_rate": 9.950438118748293e-05, "loss": 0.0927, "step": 7410 }, { "grad_norm": 0.6024909615516663, "learning_rate": 9.950147406345366e-05, "loss": 0.1004, "step": 7420 }, { "grad_norm": 0.494920015335083, "learning_rate": 9.949855848097635e-05, "loss": 0.093, "step": 7430 }, { "grad_norm": 0.47352421283721924, "learning_rate": 9.949563444054916e-05, "loss": 0.0843, "step": 7440 }, { "grad_norm": 0.5411509275436401, "learning_rate": 9.949270194267178e-05, "loss": 0.0708, "step": 7450 }, { "grad_norm": 0.7431241273880005, "learning_rate": 9.948976098784526e-05, "loss": 0.0912, "step": 7460 }, { "grad_norm": 0.6773921251296997, "learning_rate": 9.948681157657213e-05, "loss": 0.0951, "step": 7470 }, { "grad_norm": 0.6057146191596985, "learning_rate": 9.948385370935638e-05, "loss": 0.0844, "step": 7480 }, { "grad_norm": 0.4796041250228882, "learning_rate": 9.94808873867034e-05, "loss": 0.0867, "step": 7490 }, { "grad_norm": 0.7298924922943115, "learning_rate": 9.947791260912009e-05, "loss": 0.0755, "step": 7500 }, { "grad_norm": 0.6191756129264832, "learning_rate": 9.947492937711474e-05, "loss": 0.0726, "step": 7510 }, { "grad_norm": 0.576587975025177, "learning_rate": 9.947193769119707e-05, "loss": 0.0833, "step": 7520 }, { "grad_norm": 0.5880537629127502, "learning_rate": 9.946893755187834e-05, "loss": 0.0688, "step": 7530 }, { "grad_norm": 0.5277725458145142, "learning_rate": 9.946592895967115e-05, "loss": 0.0864, "step": 7540 }, { "grad_norm": 0.8402522206306458, "learning_rate": 9.94629119150896e-05, "loss": 0.0852, "step": 7550 }, { "grad_norm": 0.8065812587738037, "learning_rate": 9.94598864186492e-05, "loss": 0.1042, "step": 7560 }, { "grad_norm": 0.5571584701538086, "learning_rate": 9.945685247086696e-05, "loss": 0.0969, "step": 7570 }, { "grad_norm": 0.4804748594760895, "learning_rate": 9.945381007226129e-05, "loss": 0.0992, "step": 7580 }, { "grad_norm": 0.5788036584854126, "learning_rate": 9.945075922335203e-05, "loss": 0.089, "step": 7590 }, { "grad_norm": 0.4226996600627899, "learning_rate": 9.944769992466049e-05, "loss": 0.0875, "step": 7600 }, { "grad_norm": 0.40173542499542236, "learning_rate": 9.944463217670945e-05, "loss": 0.0826, "step": 7610 }, { "grad_norm": 1.0276834964752197, "learning_rate": 9.944155598002307e-05, "loss": 0.0813, "step": 7620 }, { "grad_norm": 0.5282548666000366, "learning_rate": 9.943847133512701e-05, "loss": 0.0862, "step": 7630 }, { "grad_norm": 0.7966116070747375, "learning_rate": 9.943537824254834e-05, "loss": 0.0797, "step": 7640 }, { "grad_norm": 0.45759645104408264, "learning_rate": 9.943227670281559e-05, "loss": 0.0797, "step": 7650 }, { "grad_norm": 0.5531575679779053, "learning_rate": 9.942916671645873e-05, "loss": 0.082, "step": 7660 }, { "grad_norm": 0.6301462054252625, "learning_rate": 9.942604828400916e-05, "loss": 0.0836, "step": 7670 }, { "grad_norm": 0.6010730862617493, "learning_rate": 9.942292140599975e-05, "loss": 0.0953, "step": 7680 }, { "grad_norm": 0.5641936659812927, "learning_rate": 9.94197860829648e-05, "loss": 0.1001, "step": 7690 }, { "grad_norm": 0.4487614035606384, "learning_rate": 9.941664231544004e-05, "loss": 0.0817, "step": 7700 }, { "grad_norm": 0.7090601921081543, "learning_rate": 9.941349010396264e-05, "loss": 0.0988, "step": 7710 }, { "grad_norm": 0.5512450337409973, "learning_rate": 9.941032944907125e-05, "loss": 0.0914, "step": 7720 }, { "grad_norm": 0.5818842649459839, "learning_rate": 9.940716035130596e-05, "loss": 0.0852, "step": 7730 }, { "grad_norm": 0.5948607921600342, "learning_rate": 9.940398281120821e-05, "loss": 0.0938, "step": 7740 }, { "grad_norm": 0.5594779253005981, "learning_rate": 9.940079682932102e-05, "loss": 0.0682, "step": 7750 }, { "grad_norm": 0.5103879570960999, "learning_rate": 9.939760240618877e-05, "loss": 0.0813, "step": 7760 }, { "grad_norm": 0.5685486197471619, "learning_rate": 9.939439954235729e-05, "loss": 0.0806, "step": 7770 }, { "grad_norm": 0.5749351978302002, "learning_rate": 9.939118823837387e-05, "loss": 0.0931, "step": 7780 }, { "grad_norm": 0.6563927531242371, "learning_rate": 9.938796849478725e-05, "loss": 0.0922, "step": 7790 }, { "grad_norm": 0.5092989802360535, "learning_rate": 9.938474031214755e-05, "loss": 0.083, "step": 7800 }, { "grad_norm": 0.560825765132904, "learning_rate": 9.938150369100643e-05, "loss": 0.0802, "step": 7810 }, { "grad_norm": 0.5561633706092834, "learning_rate": 9.93782586319169e-05, "loss": 0.0745, "step": 7820 }, { "grad_norm": 0.5513436794281006, "learning_rate": 9.937500513543348e-05, "loss": 0.0787, "step": 7830 }, { "grad_norm": 0.5611615180969238, "learning_rate": 9.937174320211207e-05, "loss": 0.0785, "step": 7840 }, { "grad_norm": 0.6810034513473511, "learning_rate": 9.936847283251009e-05, "loss": 0.0829, "step": 7850 }, { "grad_norm": 0.4416916072368622, "learning_rate": 9.936519402718632e-05, "loss": 0.0885, "step": 7860 }, { "grad_norm": 0.5582975149154663, "learning_rate": 9.936190678670102e-05, "loss": 0.0947, "step": 7870 }, { "grad_norm": 0.6671444177627563, "learning_rate": 9.935861111161593e-05, "loss": 0.0906, "step": 7880 }, { "grad_norm": 0.6968821287155151, "learning_rate": 9.935530700249416e-05, "loss": 0.0775, "step": 7890 }, { "grad_norm": 0.55149245262146, "learning_rate": 9.935199445990028e-05, "loss": 0.0956, "step": 7900 }, { "grad_norm": 0.6634159684181213, "learning_rate": 9.934867348440033e-05, "loss": 0.0887, "step": 7910 }, { "grad_norm": 0.5686479210853577, "learning_rate": 9.934534407656176e-05, "loss": 0.1217, "step": 7920 }, { "grad_norm": 0.5998662710189819, "learning_rate": 9.93420062369535e-05, "loss": 0.0851, "step": 7930 }, { "grad_norm": 0.5445342063903809, "learning_rate": 9.933865996614589e-05, "loss": 0.0861, "step": 7940 }, { "grad_norm": 0.5819941163063049, "learning_rate": 9.933530526471068e-05, "loss": 0.0816, "step": 7950 }, { "grad_norm": 0.5841440558433533, "learning_rate": 9.933194213322114e-05, "loss": 0.0814, "step": 7960 }, { "grad_norm": 0.6547318696975708, "learning_rate": 9.932857057225192e-05, "loss": 0.0891, "step": 7970 }, { "grad_norm": 0.6541323661804199, "learning_rate": 9.932519058237912e-05, "loss": 0.095, "step": 7980 }, { "grad_norm": 0.5961278676986694, "learning_rate": 9.932180216418032e-05, "loss": 0.0874, "step": 7990 }, { "grad_norm": 0.6337912082672119, "learning_rate": 9.931840531823446e-05, "loss": 0.0718, "step": 8000 }, { "grad_norm": 0.4454721510410309, "learning_rate": 9.9315000045122e-05, "loss": 0.074, "step": 8010 }, { "grad_norm": 0.5786619186401367, "learning_rate": 9.931158634542481e-05, "loss": 0.0933, "step": 8020 }, { "grad_norm": 0.42715102434158325, "learning_rate": 9.930816421972617e-05, "loss": 0.0685, "step": 8030 }, { "grad_norm": 0.6416335701942444, "learning_rate": 9.930473366861086e-05, "loss": 0.0772, "step": 8040 }, { "grad_norm": 0.4708614945411682, "learning_rate": 9.930129469266505e-05, "loss": 0.0866, "step": 8050 }, { "grad_norm": 0.5635308027267456, "learning_rate": 9.929784729247638e-05, "loss": 0.0856, "step": 8060 }, { "grad_norm": 0.6112752556800842, "learning_rate": 9.929439146863389e-05, "loss": 0.0809, "step": 8070 }, { "grad_norm": 0.4825780689716339, "learning_rate": 9.92909272217281e-05, "loss": 0.094, "step": 8080 }, { "grad_norm": 0.5949814915657043, "learning_rate": 9.928745455235097e-05, "loss": 0.0835, "step": 8090 }, { "grad_norm": 0.6704389452934265, "learning_rate": 9.928397346109588e-05, "loss": 0.1014, "step": 8100 }, { "grad_norm": 0.7012978196144104, "learning_rate": 9.928048394855762e-05, "loss": 0.0977, "step": 8110 }, { "grad_norm": 0.6753972768783569, "learning_rate": 9.92769860153325e-05, "loss": 0.0856, "step": 8120 }, { "grad_norm": 0.5678683519363403, "learning_rate": 9.927347966201819e-05, "loss": 0.1124, "step": 8130 }, { "grad_norm": 0.5248151421546936, "learning_rate": 9.926996488921383e-05, "loss": 0.0998, "step": 8140 }, { "grad_norm": 0.8490375876426697, "learning_rate": 9.926644169752001e-05, "loss": 0.0912, "step": 8150 }, { "grad_norm": 0.5701540112495422, "learning_rate": 9.926291008753875e-05, "loss": 0.0967, "step": 8160 }, { "grad_norm": 0.40385234355926514, "learning_rate": 9.92593700598735e-05, "loss": 0.0825, "step": 8170 }, { "grad_norm": 0.5475435853004456, "learning_rate": 9.925582161512915e-05, "loss": 0.0841, "step": 8180 }, { "grad_norm": 0.5462392568588257, "learning_rate": 9.925226475391205e-05, "loss": 0.0951, "step": 8190 }, { "grad_norm": 0.4721739590167999, "learning_rate": 9.924869947682993e-05, "loss": 0.076, "step": 8200 }, { "grad_norm": 0.44139793515205383, "learning_rate": 9.924512578449204e-05, "loss": 0.0676, "step": 8210 }, { "grad_norm": 0.5673798322677612, "learning_rate": 9.924154367750901e-05, "loss": 0.0734, "step": 8220 }, { "grad_norm": 0.5094850063323975, "learning_rate": 9.923795315649293e-05, "loss": 0.0821, "step": 8230 }, { "grad_norm": 0.5850591063499451, "learning_rate": 9.92343542220573e-05, "loss": 0.0884, "step": 8240 }, { "grad_norm": 0.7769544720649719, "learning_rate": 9.92307468748171e-05, "loss": 0.0804, "step": 8250 }, { "grad_norm": 0.441561222076416, "learning_rate": 9.922713111538873e-05, "loss": 0.0595, "step": 8260 }, { "grad_norm": 0.5389688611030579, "learning_rate": 9.922350694439003e-05, "loss": 0.0818, "step": 8270 }, { "grad_norm": 0.5877339839935303, "learning_rate": 9.921987436244024e-05, "loss": 0.0944, "step": 8280 }, { "grad_norm": 0.5641019940376282, "learning_rate": 9.921623337016008e-05, "loss": 0.0662, "step": 8290 }, { "grad_norm": 0.5579237937927246, "learning_rate": 9.921258396817172e-05, "loss": 0.0953, "step": 8300 }, { "grad_norm": 0.6572108268737793, "learning_rate": 9.920892615709874e-05, "loss": 0.0773, "step": 8310 }, { "grad_norm": 0.4020753800868988, "learning_rate": 9.920525993756612e-05, "loss": 0.0651, "step": 8320 }, { "grad_norm": 0.4223337769508362, "learning_rate": 9.920158531020036e-05, "loss": 0.0752, "step": 8330 }, { "grad_norm": 0.5387020111083984, "learning_rate": 9.919790227562933e-05, "loss": 0.0714, "step": 8340 }, { "grad_norm": 0.7773652672767639, "learning_rate": 9.919421083448237e-05, "loss": 0.0865, "step": 8350 }, { "grad_norm": 0.4664509892463684, "learning_rate": 9.919051098739022e-05, "loss": 0.0729, "step": 8360 }, { "grad_norm": 0.5562905073165894, "learning_rate": 9.918680273498514e-05, "loss": 0.068, "step": 8370 }, { "grad_norm": 0.5379325747489929, "learning_rate": 9.918308607790072e-05, "loss": 0.0765, "step": 8380 }, { "grad_norm": 0.5127636194229126, "learning_rate": 9.917936101677205e-05, "loss": 0.0617, "step": 8390 }, { "grad_norm": 0.5443041324615479, "learning_rate": 9.917562755223564e-05, "loss": 0.0814, "step": 8400 }, { "grad_norm": 0.5135800838470459, "learning_rate": 9.917188568492944e-05, "loss": 0.0668, "step": 8410 }, { "grad_norm": 0.6806768178939819, "learning_rate": 9.916813541549283e-05, "loss": 0.0733, "step": 8420 }, { "grad_norm": 0.7174573540687561, "learning_rate": 9.916437674456663e-05, "loss": 0.0843, "step": 8430 }, { "grad_norm": 0.493966668844223, "learning_rate": 9.916060967279308e-05, "loss": 0.0741, "step": 8440 }, { "grad_norm": 0.44398045539855957, "learning_rate": 9.91568342008159e-05, "loss": 0.0703, "step": 8450 }, { "grad_norm": 0.5635052919387817, "learning_rate": 9.915305032928019e-05, "loss": 0.0773, "step": 8460 }, { "grad_norm": 0.5520474314689636, "learning_rate": 9.914925805883253e-05, "loss": 0.0732, "step": 8470 }, { "grad_norm": 0.5739441514015198, "learning_rate": 9.914545739012088e-05, "loss": 0.0755, "step": 8480 }, { "grad_norm": 0.6403616070747375, "learning_rate": 9.91416483237947e-05, "loss": 0.0781, "step": 8490 }, { "grad_norm": 0.5310763716697693, "learning_rate": 9.913783086050485e-05, "loss": 0.0892, "step": 8500 }, { "grad_norm": 0.41553226113319397, "learning_rate": 9.913400500090364e-05, "loss": 0.0765, "step": 8510 }, { "grad_norm": 0.4380147457122803, "learning_rate": 9.913017074564479e-05, "loss": 0.0682, "step": 8520 }, { "grad_norm": 0.6840301752090454, "learning_rate": 9.912632809538348e-05, "loss": 0.0818, "step": 8530 }, { "grad_norm": 0.7919495701789856, "learning_rate": 9.912247705077629e-05, "loss": 0.0771, "step": 8540 }, { "grad_norm": 0.8158754706382751, "learning_rate": 9.911861761248127e-05, "loss": 0.0907, "step": 8550 }, { "grad_norm": 0.5989265441894531, "learning_rate": 9.91147497811579e-05, "loss": 0.0817, "step": 8560 }, { "grad_norm": 0.65377277135849, "learning_rate": 9.911087355746709e-05, "loss": 0.1221, "step": 8570 }, { "grad_norm": 0.558734655380249, "learning_rate": 9.910698894207117e-05, "loss": 0.1017, "step": 8580 }, { "grad_norm": 0.5796892046928406, "learning_rate": 9.910309593563392e-05, "loss": 0.0982, "step": 8590 }, { "grad_norm": 0.45696279406547546, "learning_rate": 9.909919453882057e-05, "loss": 0.0703, "step": 8600 }, { "grad_norm": 0.6035982370376587, "learning_rate": 9.90952847522977e-05, "loss": 0.0789, "step": 8610 }, { "grad_norm": 0.6262484192848206, "learning_rate": 9.909136657673346e-05, "loss": 0.0937, "step": 8620 }, { "grad_norm": 0.4356153905391693, "learning_rate": 9.908744001279731e-05, "loss": 0.0763, "step": 8630 }, { "grad_norm": 0.5606836676597595, "learning_rate": 9.90835050611602e-05, "loss": 0.0858, "step": 8640 }, { "grad_norm": 0.6173003911972046, "learning_rate": 9.90795617224945e-05, "loss": 0.0833, "step": 8650 }, { "grad_norm": 0.8469977378845215, "learning_rate": 9.907560999747405e-05, "loss": 0.081, "step": 8660 }, { "grad_norm": 0.631027340888977, "learning_rate": 9.907164988677408e-05, "loss": 0.0798, "step": 8670 }, { "grad_norm": 0.5408482551574707, "learning_rate": 9.906768139107124e-05, "loss": 0.0644, "step": 8680 }, { "grad_norm": 0.548490047454834, "learning_rate": 9.906370451104367e-05, "loss": 0.069, "step": 8690 }, { "grad_norm": 0.6540614366531372, "learning_rate": 9.905971924737088e-05, "loss": 0.0766, "step": 8700 }, { "grad_norm": 0.577683687210083, "learning_rate": 9.905572560073387e-05, "loss": 0.0752, "step": 8710 }, { "grad_norm": 0.7421601414680481, "learning_rate": 9.905172357181501e-05, "loss": 0.0821, "step": 8720 }, { "grad_norm": 0.4418056607246399, "learning_rate": 9.904771316129817e-05, "loss": 0.0834, "step": 8730 }, { "grad_norm": 0.5459570288658142, "learning_rate": 9.904369436986862e-05, "loss": 0.0787, "step": 8740 }, { "grad_norm": 0.48984333872795105, "learning_rate": 9.903966719821303e-05, "loss": 0.0906, "step": 8750 }, { "grad_norm": 0.4765840172767639, "learning_rate": 9.903563164701956e-05, "loss": 0.0774, "step": 8760 }, { "grad_norm": 0.8591476678848267, "learning_rate": 9.903158771697778e-05, "loss": 0.0904, "step": 8770 }, { "grad_norm": 0.8388614058494568, "learning_rate": 9.902753540877867e-05, "loss": 0.1098, "step": 8780 }, { "grad_norm": 0.7499213814735413, "learning_rate": 9.902347472311466e-05, "loss": 0.0839, "step": 8790 }, { "grad_norm": 0.4168194830417633, "learning_rate": 9.901940566067962e-05, "loss": 0.0696, "step": 8800 }, { "grad_norm": 0.6299332976341248, "learning_rate": 9.901532822216883e-05, "loss": 0.0709, "step": 8810 }, { "grad_norm": 0.7953328490257263, "learning_rate": 9.901124240827904e-05, "loss": 0.0727, "step": 8820 }, { "grad_norm": 0.6237785816192627, "learning_rate": 9.900714821970835e-05, "loss": 0.0805, "step": 8830 }, { "grad_norm": 0.6237820982933044, "learning_rate": 9.900304565715641e-05, "loss": 0.071, "step": 8840 }, { "grad_norm": 0.7253096103668213, "learning_rate": 9.899893472132419e-05, "loss": 0.0779, "step": 8850 }, { "grad_norm": 0.5572514533996582, "learning_rate": 9.899481541291415e-05, "loss": 0.0767, "step": 8860 }, { "grad_norm": 0.8421313762664795, "learning_rate": 9.899068773263016e-05, "loss": 0.0717, "step": 8870 }, { "grad_norm": 0.48259004950523376, "learning_rate": 9.898655168117754e-05, "loss": 0.072, "step": 8880 }, { "grad_norm": 0.5095600485801697, "learning_rate": 9.898240725926302e-05, "loss": 0.0868, "step": 8890 }, { "grad_norm": 0.4642258584499359, "learning_rate": 9.897825446759478e-05, "loss": 0.078, "step": 8900 }, { "grad_norm": 0.494861900806427, "learning_rate": 9.897409330688241e-05, "loss": 0.0864, "step": 8910 }, { "grad_norm": 0.6894025802612305, "learning_rate": 9.896992377783692e-05, "loss": 0.1041, "step": 8920 }, { "grad_norm": 0.4457980990409851, "learning_rate": 9.89657458811708e-05, "loss": 0.0763, "step": 8930 }, { "grad_norm": 0.4960051476955414, "learning_rate": 9.896155961759792e-05, "loss": 0.074, "step": 8940 }, { "grad_norm": 0.5872945785522461, "learning_rate": 9.895736498783361e-05, "loss": 0.0864, "step": 8950 }, { "grad_norm": 0.4728935956954956, "learning_rate": 9.895316199259462e-05, "loss": 0.0706, "step": 8960 }, { "grad_norm": 0.4500627815723419, "learning_rate": 9.894895063259909e-05, "loss": 0.0735, "step": 8970 }, { "grad_norm": 0.6555816531181335, "learning_rate": 9.894473090856667e-05, "loss": 0.087, "step": 8980 }, { "grad_norm": 0.7175841331481934, "learning_rate": 9.894050282121839e-05, "loss": 0.1086, "step": 8990 }, { "grad_norm": 0.45418959856033325, "learning_rate": 9.893626637127668e-05, "loss": 0.0774, "step": 9000 }, { "grad_norm": 0.6752191185951233, "learning_rate": 9.893202155946546e-05, "loss": 0.0705, "step": 9010 }, { "grad_norm": 0.7221190929412842, "learning_rate": 9.892776838651006e-05, "loss": 0.0798, "step": 9020 }, { "grad_norm": 0.6149332523345947, "learning_rate": 9.892350685313722e-05, "loss": 0.1031, "step": 9030 }, { "grad_norm": 0.5339863300323486, "learning_rate": 9.891923696007513e-05, "loss": 0.0991, "step": 9040 }, { "grad_norm": 0.5263338685035706, "learning_rate": 9.891495870805336e-05, "loss": 0.0809, "step": 9050 }, { "grad_norm": 0.631532609462738, "learning_rate": 9.891067209780298e-05, "loss": 0.0906, "step": 9060 }, { "grad_norm": 0.48455771803855896, "learning_rate": 9.890637713005646e-05, "loss": 0.0661, "step": 9070 }, { "grad_norm": 0.4865359961986542, "learning_rate": 9.890207380554767e-05, "loss": 0.0716, "step": 9080 }, { "grad_norm": 0.5044549107551575, "learning_rate": 9.889776212501196e-05, "loss": 0.0825, "step": 9090 }, { "grad_norm": 0.6519334316253662, "learning_rate": 9.889344208918605e-05, "loss": 0.0904, "step": 9100 }, { "grad_norm": 0.6852169036865234, "learning_rate": 9.888911369880812e-05, "loss": 0.0918, "step": 9110 }, { "grad_norm": 0.484353244304657, "learning_rate": 9.888477695461777e-05, "loss": 0.0805, "step": 9120 }, { "grad_norm": 0.6313307285308838, "learning_rate": 9.888043185735607e-05, "loss": 0.0977, "step": 9130 }, { "grad_norm": 0.5020358562469482, "learning_rate": 9.887607840776542e-05, "loss": 0.076, "step": 9140 }, { "grad_norm": 0.5087912082672119, "learning_rate": 9.887171660658975e-05, "loss": 0.0945, "step": 9150 }, { "grad_norm": 0.5803675055503845, "learning_rate": 9.886734645457435e-05, "loss": 0.0672, "step": 9160 }, { "grad_norm": 0.5957508683204651, "learning_rate": 9.886296795246597e-05, "loss": 0.0656, "step": 9170 }, { "grad_norm": 0.4468899071216583, "learning_rate": 9.885858110101276e-05, "loss": 0.0697, "step": 9180 }, { "grad_norm": 0.5651359558105469, "learning_rate": 9.885418590096434e-05, "loss": 0.0709, "step": 9190 }, { "grad_norm": 0.4469802677631378, "learning_rate": 9.88497823530717e-05, "loss": 0.0741, "step": 9200 }, { "grad_norm": 0.4023672342300415, "learning_rate": 9.884537045808732e-05, "loss": 0.0981, "step": 9210 }, { "grad_norm": 0.5055243372917175, "learning_rate": 9.884095021676502e-05, "loss": 0.0803, "step": 9220 }, { "grad_norm": 0.6483360528945923, "learning_rate": 9.883652162986017e-05, "loss": 0.0799, "step": 9230 }, { "grad_norm": 0.4846229553222656, "learning_rate": 9.883208469812943e-05, "loss": 0.0804, "step": 9240 }, { "grad_norm": 0.44350937008857727, "learning_rate": 9.882763942233098e-05, "loss": 0.0621, "step": 9250 }, { "grad_norm": 0.4939858317375183, "learning_rate": 9.882318580322441e-05, "loss": 0.075, "step": 9260 }, { "grad_norm": 0.4806790351867676, "learning_rate": 9.881872384157067e-05, "loss": 0.0764, "step": 9270 }, { "grad_norm": 0.46790021657943726, "learning_rate": 9.881425353813225e-05, "loss": 0.0753, "step": 9280 }, { "grad_norm": 0.5596435070037842, "learning_rate": 9.880977489367296e-05, "loss": 0.0788, "step": 9290 }, { "grad_norm": 0.5057677626609802, "learning_rate": 9.88052879089581e-05, "loss": 0.0953, "step": 9300 }, { "grad_norm": 0.5280722379684448, "learning_rate": 9.880079258475434e-05, "loss": 0.0864, "step": 9310 }, { "grad_norm": 0.42667946219444275, "learning_rate": 9.879628892182985e-05, "loss": 0.0884, "step": 9320 }, { "grad_norm": 0.6079956889152527, "learning_rate": 9.879177692095416e-05, "loss": 0.0731, "step": 9330 }, { "grad_norm": 0.513408362865448, "learning_rate": 9.878725658289825e-05, "loss": 0.0905, "step": 9340 }, { "grad_norm": 0.7937655448913574, "learning_rate": 9.878272790843454e-05, "loss": 0.0736, "step": 9350 }, { "grad_norm": 0.4611656963825226, "learning_rate": 9.877819089833682e-05, "loss": 0.0629, "step": 9360 }, { "grad_norm": 0.5025468468666077, "learning_rate": 9.877364555338038e-05, "loss": 0.086, "step": 9370 }, { "grad_norm": 0.5773407220840454, "learning_rate": 9.876909187434186e-05, "loss": 0.0828, "step": 9380 }, { "grad_norm": 0.5046210289001465, "learning_rate": 9.876452986199939e-05, "loss": 0.0718, "step": 9390 }, { "grad_norm": 0.5541767477989197, "learning_rate": 9.875995951713248e-05, "loss": 0.093, "step": 9400 }, { "grad_norm": 0.7334752082824707, "learning_rate": 9.875538084052207e-05, "loss": 0.0865, "step": 9410 }, { "grad_norm": 0.5480882525444031, "learning_rate": 9.875079383295053e-05, "loss": 0.0754, "step": 9420 }, { "grad_norm": 0.4588562846183777, "learning_rate": 9.874619849520167e-05, "loss": 0.0692, "step": 9430 }, { "grad_norm": 0.5179523825645447, "learning_rate": 9.874159482806069e-05, "loss": 0.0703, "step": 9440 }, { "grad_norm": 0.572407066822052, "learning_rate": 9.873698283231426e-05, "loss": 0.0662, "step": 9450 }, { "grad_norm": 0.7307089567184448, "learning_rate": 9.87323625087504e-05, "loss": 0.0671, "step": 9460 }, { "grad_norm": 0.6381977200508118, "learning_rate": 9.872773385815863e-05, "loss": 0.1, "step": 9470 }, { "grad_norm": 0.4237634837627411, "learning_rate": 9.872309688132986e-05, "loss": 0.0821, "step": 9480 }, { "grad_norm": 0.5285094380378723, "learning_rate": 9.871845157905639e-05, "loss": 0.0682, "step": 9490 }, { "grad_norm": 0.4415023922920227, "learning_rate": 9.871379795213201e-05, "loss": 0.067, "step": 9500 }, { "grad_norm": 0.44636255502700806, "learning_rate": 9.87091360013519e-05, "loss": 0.0706, "step": 9510 }, { "grad_norm": 0.7471835017204285, "learning_rate": 9.870446572751262e-05, "loss": 0.0763, "step": 9520 }, { "grad_norm": 0.4056589603424072, "learning_rate": 9.869978713141224e-05, "loss": 0.0814, "step": 9530 }, { "grad_norm": 0.7441840767860413, "learning_rate": 9.869510021385016e-05, "loss": 0.0842, "step": 9540 }, { "grad_norm": 0.39263275265693665, "learning_rate": 9.869040497562727e-05, "loss": 0.076, "step": 9550 }, { "grad_norm": 0.789309024810791, "learning_rate": 9.868570141754587e-05, "loss": 0.0829, "step": 9560 }, { "grad_norm": 0.5315515995025635, "learning_rate": 9.868098954040965e-05, "loss": 0.0911, "step": 9570 }, { "grad_norm": 0.6639784574508667, "learning_rate": 9.867626934502374e-05, "loss": 0.0878, "step": 9580 }, { "grad_norm": 0.47132399678230286, "learning_rate": 9.86715408321947e-05, "loss": 0.071, "step": 9590 }, { "grad_norm": 0.46468502283096313, "learning_rate": 9.86668040027305e-05, "loss": 0.0666, "step": 9600 }, { "grad_norm": 0.7183296084403992, "learning_rate": 9.866205885744053e-05, "loss": 0.0777, "step": 9610 }, { "grad_norm": 0.5324628353118896, "learning_rate": 9.865730539713563e-05, "loss": 0.0781, "step": 9620 }, { "grad_norm": 0.7203158736228943, "learning_rate": 9.8652543622628e-05, "loss": 0.0732, "step": 9630 }, { "grad_norm": 0.7367917895317078, "learning_rate": 9.864777353473132e-05, "loss": 0.0798, "step": 9640 }, { "grad_norm": 0.8221983313560486, "learning_rate": 9.864299513426068e-05, "loss": 0.0886, "step": 9650 }, { "grad_norm": 0.5419175028800964, "learning_rate": 9.863820842203254e-05, "loss": 0.0848, "step": 9660 }, { "grad_norm": 0.4883783459663391, "learning_rate": 9.863341339886483e-05, "loss": 0.082, "step": 9670 }, { "grad_norm": 0.48580217361450195, "learning_rate": 9.86286100655769e-05, "loss": 0.0704, "step": 9680 }, { "grad_norm": 0.8725981116294861, "learning_rate": 9.862379842298953e-05, "loss": 0.0759, "step": 9690 }, { "grad_norm": 0.6853016018867493, "learning_rate": 9.861897847192485e-05, "loss": 0.076, "step": 9700 }, { "grad_norm": 0.5301743149757385, "learning_rate": 9.86141502132065e-05, "loss": 0.0736, "step": 9710 }, { "grad_norm": 0.5661678314208984, "learning_rate": 9.860931364765946e-05, "loss": 0.0776, "step": 9720 }, { "grad_norm": 0.6006742715835571, "learning_rate": 9.860446877611021e-05, "loss": 0.0818, "step": 9730 }, { "grad_norm": 0.48271262645721436, "learning_rate": 9.859961559938655e-05, "loss": 0.063, "step": 9740 }, { "grad_norm": 0.6493464708328247, "learning_rate": 9.85947541183178e-05, "loss": 0.0712, "step": 9750 }, { "grad_norm": 0.8193683624267578, "learning_rate": 9.858988433373463e-05, "loss": 0.0845, "step": 9760 }, { "grad_norm": 0.44048619270324707, "learning_rate": 9.858500624646918e-05, "loss": 0.0704, "step": 9770 }, { "grad_norm": 0.5307182669639587, "learning_rate": 9.858011985735497e-05, "loss": 0.078, "step": 9780 }, { "grad_norm": 0.5999795198440552, "learning_rate": 9.857522516722693e-05, "loss": 0.0718, "step": 9790 }, { "grad_norm": 0.5383414626121521, "learning_rate": 9.857032217692145e-05, "loss": 0.0644, "step": 9800 }, { "grad_norm": 0.6474829316139221, "learning_rate": 9.856541088727631e-05, "loss": 0.0854, "step": 9810 }, { "grad_norm": 0.47787991166114807, "learning_rate": 9.856049129913072e-05, "loss": 0.0879, "step": 9820 }, { "grad_norm": 0.8143448829650879, "learning_rate": 9.85555634133253e-05, "loss": 0.0895, "step": 9830 }, { "grad_norm": 0.5933436751365662, "learning_rate": 9.855062723070208e-05, "loss": 0.0877, "step": 9840 }, { "grad_norm": 0.5155642628669739, "learning_rate": 9.854568275210454e-05, "loss": 0.0786, "step": 9850 }, { "grad_norm": 0.5698766708374023, "learning_rate": 9.854072997837754e-05, "loss": 0.0613, "step": 9860 }, { "grad_norm": 0.36295807361602783, "learning_rate": 9.853576891036737e-05, "loss": 0.0706, "step": 9870 }, { "grad_norm": 0.46446701884269714, "learning_rate": 9.853079954892177e-05, "loss": 0.0672, "step": 9880 }, { "grad_norm": 0.5392798185348511, "learning_rate": 9.852582189488983e-05, "loss": 0.0588, "step": 9890 }, { "grad_norm": 0.4794783294200897, "learning_rate": 9.852083594912212e-05, "loss": 0.0736, "step": 9900 }, { "grad_norm": 0.4501195549964905, "learning_rate": 9.851584171247058e-05, "loss": 0.0802, "step": 9910 }, { "grad_norm": 0.6189380288124084, "learning_rate": 9.851083918578863e-05, "loss": 0.0771, "step": 9920 }, { "grad_norm": 0.4191966950893402, "learning_rate": 9.850582836993103e-05, "loss": 0.063, "step": 9930 }, { "grad_norm": 0.5380081534385681, "learning_rate": 9.850080926575397e-05, "loss": 0.0779, "step": 9940 }, { "grad_norm": 0.4958268404006958, "learning_rate": 9.849578187411515e-05, "loss": 0.0715, "step": 9950 }, { "grad_norm": 0.7205268740653992, "learning_rate": 9.849074619587354e-05, "loss": 0.0746, "step": 9960 }, { "grad_norm": 0.5391332507133484, "learning_rate": 9.848570223188964e-05, "loss": 0.0743, "step": 9970 }, { "grad_norm": 0.4845001697540283, "learning_rate": 9.848064998302531e-05, "loss": 0.0753, "step": 9980 }, { "grad_norm": 0.4726444184780121, "learning_rate": 9.847558945014386e-05, "loss": 0.0844, "step": 9990 }, { "grad_norm": 0.6388817429542542, "learning_rate": 9.847052063410996e-05, "loss": 0.0828, "step": 10000 }, { "grad_norm": 2.4552597999572754, "learning_rate": 9.846544353578977e-05, "loss": 0.0828, "step": 10010 }, { "grad_norm": 0.4760039448738098, "learning_rate": 9.846035815605081e-05, "loss": 0.092, "step": 10020 }, { "grad_norm": 0.46088382601737976, "learning_rate": 9.845526449576204e-05, "loss": 0.075, "step": 10030 }, { "grad_norm": 0.642429769039154, "learning_rate": 9.845016255579383e-05, "loss": 0.0879, "step": 10040 }, { "grad_norm": 0.49918094277381897, "learning_rate": 9.844505233701794e-05, "loss": 0.0917, "step": 10050 }, { "grad_norm": 0.6783306002616882, "learning_rate": 9.843993384030757e-05, "loss": 0.0807, "step": 10060 }, { "grad_norm": 0.5536092519760132, "learning_rate": 9.843480706653737e-05, "loss": 0.0704, "step": 10070 }, { "grad_norm": 0.6317524909973145, "learning_rate": 9.84296720165833e-05, "loss": 0.08, "step": 10080 }, { "grad_norm": 0.550253689289093, "learning_rate": 9.842452869132286e-05, "loss": 0.079, "step": 10090 }, { "grad_norm": 0.47995689511299133, "learning_rate": 9.841937709163489e-05, "loss": 0.072, "step": 10100 }, { "grad_norm": 0.6544037461280823, "learning_rate": 9.841421721839962e-05, "loss": 0.0586, "step": 10110 }, { "grad_norm": 0.7880813479423523, "learning_rate": 9.840904907249879e-05, "loss": 0.0704, "step": 10120 }, { "grad_norm": 0.8539280891418457, "learning_rate": 9.840387265481545e-05, "loss": 0.0788, "step": 10130 }, { "grad_norm": 0.506828784942627, "learning_rate": 9.839868796623411e-05, "loss": 0.0723, "step": 10140 }, { "grad_norm": 0.6085572838783264, "learning_rate": 9.839349500764072e-05, "loss": 0.0943, "step": 10150 }, { "grad_norm": 0.514970064163208, "learning_rate": 9.83882937799226e-05, "loss": 0.077, "step": 10160 }, { "grad_norm": 0.6322582364082336, "learning_rate": 9.838308428396849e-05, "loss": 0.0709, "step": 10170 }, { "grad_norm": 0.5589413046836853, "learning_rate": 9.837786652066854e-05, "loss": 0.0926, "step": 10180 }, { "grad_norm": 0.7088934183120728, "learning_rate": 9.837264049091437e-05, "loss": 0.095, "step": 10190 }, { "grad_norm": 0.5695868730545044, "learning_rate": 9.836740619559893e-05, "loss": 0.0888, "step": 10200 }, { "grad_norm": 0.5326513051986694, "learning_rate": 9.836216363561659e-05, "loss": 0.0968, "step": 10210 }, { "grad_norm": 0.5745028257369995, "learning_rate": 9.835691281186322e-05, "loss": 0.0738, "step": 10220 }, { "grad_norm": 0.5815936923027039, "learning_rate": 9.8351653725236e-05, "loss": 0.0818, "step": 10230 }, { "grad_norm": 0.5928611755371094, "learning_rate": 9.83463863766336e-05, "loss": 0.0762, "step": 10240 }, { "grad_norm": 0.5496425628662109, "learning_rate": 9.834111076695602e-05, "loss": 0.0746, "step": 10250 }, { "grad_norm": 0.3881656229496002, "learning_rate": 9.833582689710477e-05, "loss": 0.0682, "step": 10260 }, { "grad_norm": 0.4938295781612396, "learning_rate": 9.833053476798268e-05, "loss": 0.0677, "step": 10270 }, { "grad_norm": 0.5383549332618713, "learning_rate": 9.832523438049404e-05, "loss": 0.0819, "step": 10280 }, { "grad_norm": 0.6472225785255432, "learning_rate": 9.831992573554454e-05, "loss": 0.0818, "step": 10290 }, { "grad_norm": 0.6583753824234009, "learning_rate": 9.831460883404128e-05, "loss": 0.073, "step": 10300 }, { "grad_norm": 0.6256301403045654, "learning_rate": 9.830928367689278e-05, "loss": 0.0716, "step": 10310 }, { "grad_norm": 0.4281242787837982, "learning_rate": 9.830395026500896e-05, "loss": 0.0796, "step": 10320 }, { "grad_norm": 0.6280300617218018, "learning_rate": 9.829860859930115e-05, "loss": 0.0743, "step": 10330 }, { "grad_norm": 0.6697205305099487, "learning_rate": 9.829325868068212e-05, "loss": 0.0801, "step": 10340 }, { "grad_norm": 0.4977855682373047, "learning_rate": 9.8287900510066e-05, "loss": 0.0791, "step": 10350 }, { "grad_norm": 0.4701065421104431, "learning_rate": 9.828253408836834e-05, "loss": 0.0717, "step": 10360 }, { "grad_norm": 0.6695235967636108, "learning_rate": 9.827715941650615e-05, "loss": 0.0777, "step": 10370 }, { "grad_norm": 0.5720806121826172, "learning_rate": 9.82717764953978e-05, "loss": 0.0755, "step": 10380 }, { "grad_norm": 0.877659797668457, "learning_rate": 9.826638532596308e-05, "loss": 0.0691, "step": 10390 }, { "grad_norm": 0.4745825529098511, "learning_rate": 9.82609859091232e-05, "loss": 0.0686, "step": 10400 }, { "grad_norm": 0.5783305168151855, "learning_rate": 9.825557824580076e-05, "loss": 0.0609, "step": 10410 }, { "grad_norm": 0.37768271565437317, "learning_rate": 9.82501623369198e-05, "loss": 0.0755, "step": 10420 }, { "grad_norm": 0.474982351064682, "learning_rate": 9.824473818340574e-05, "loss": 0.0615, "step": 10430 }, { "grad_norm": 0.5281403064727783, "learning_rate": 9.823930578618541e-05, "loss": 0.0746, "step": 10440 }, { "grad_norm": 0.5049290657043457, "learning_rate": 9.823386514618709e-05, "loss": 0.0932, "step": 10450 }, { "grad_norm": 0.7119255661964417, "learning_rate": 9.82284162643404e-05, "loss": 0.0764, "step": 10460 }, { "grad_norm": 0.5969694256782532, "learning_rate": 9.822295914157642e-05, "loss": 0.0774, "step": 10470 }, { "grad_norm": 0.7026344537734985, "learning_rate": 9.821749377882763e-05, "loss": 0.0719, "step": 10480 }, { "grad_norm": 0.585719645023346, "learning_rate": 9.821202017702791e-05, "loss": 0.0757, "step": 10490 }, { "grad_norm": 0.5562686324119568, "learning_rate": 9.820653833711253e-05, "loss": 0.0701, "step": 10500 }, { "grad_norm": 0.4495646357536316, "learning_rate": 9.820104826001822e-05, "loss": 0.0798, "step": 10510 }, { "grad_norm": 0.46479901671409607, "learning_rate": 9.819554994668305e-05, "loss": 0.059, "step": 10520 }, { "grad_norm": 0.5246067643165588, "learning_rate": 9.819004339804654e-05, "loss": 0.0692, "step": 10530 }, { "grad_norm": 0.44657185673713684, "learning_rate": 9.818452861504961e-05, "loss": 0.0688, "step": 10540 }, { "grad_norm": 0.49847355484962463, "learning_rate": 9.81790055986346e-05, "loss": 0.0632, "step": 10550 }, { "grad_norm": 0.39549100399017334, "learning_rate": 9.817347434974523e-05, "loss": 0.0633, "step": 10560 }, { "grad_norm": 0.42788177728652954, "learning_rate": 9.816793486932664e-05, "loss": 0.0616, "step": 10570 }, { "grad_norm": 0.5145291090011597, "learning_rate": 9.816238715832538e-05, "loss": 0.0722, "step": 10580 }, { "grad_norm": 0.5220904350280762, "learning_rate": 9.815683121768939e-05, "loss": 0.07, "step": 10590 }, { "grad_norm": 0.4785248041152954, "learning_rate": 9.815126704836804e-05, "loss": 0.0751, "step": 10600 }, { "grad_norm": 0.44612637162208557, "learning_rate": 9.81456946513121e-05, "loss": 0.0937, "step": 10610 }, { "grad_norm": 0.5280222296714783, "learning_rate": 9.814011402747373e-05, "loss": 0.0802, "step": 10620 }, { "grad_norm": 0.41991740465164185, "learning_rate": 9.813452517780651e-05, "loss": 0.0575, "step": 10630 }, { "grad_norm": 0.35960856080055237, "learning_rate": 9.81289281032654e-05, "loss": 0.0764, "step": 10640 }, { "grad_norm": 0.5195173025131226, "learning_rate": 9.812332280480683e-05, "loss": 0.0861, "step": 10650 }, { "grad_norm": 0.5635406374931335, "learning_rate": 9.811770928338854e-05, "loss": 0.0784, "step": 10660 }, { "grad_norm": 0.5805332064628601, "learning_rate": 9.811208753996979e-05, "loss": 0.0757, "step": 10670 }, { "grad_norm": 0.5491703152656555, "learning_rate": 9.810645757551113e-05, "loss": 0.0693, "step": 10680 }, { "grad_norm": 0.5627798438072205, "learning_rate": 9.810081939097459e-05, "loss": 0.0679, "step": 10690 }, { "grad_norm": 0.5661154985427856, "learning_rate": 9.809517298732356e-05, "loss": 0.0987, "step": 10700 }, { "grad_norm": 0.521872878074646, "learning_rate": 9.80895183655229e-05, "loss": 0.0794, "step": 10710 }, { "grad_norm": 0.521902322769165, "learning_rate": 9.808385552653877e-05, "loss": 0.0717, "step": 10720 }, { "grad_norm": 0.4973168671131134, "learning_rate": 9.807818447133886e-05, "loss": 0.0682, "step": 10730 }, { "grad_norm": 0.9777014851570129, "learning_rate": 9.807250520089215e-05, "loss": 0.0658, "step": 10740 }, { "grad_norm": 0.5799218416213989, "learning_rate": 9.806681771616908e-05, "loss": 0.0715, "step": 10750 }, { "grad_norm": 0.7228969931602478, "learning_rate": 9.80611220181415e-05, "loss": 0.0954, "step": 10760 }, { "grad_norm": 0.5619479417800903, "learning_rate": 9.805541810778264e-05, "loss": 0.087, "step": 10770 }, { "grad_norm": 0.5693924427032471, "learning_rate": 9.804970598606716e-05, "loss": 0.0824, "step": 10780 }, { "grad_norm": 0.6426829695701599, "learning_rate": 9.804398565397106e-05, "loss": 0.0741, "step": 10790 }, { "grad_norm": 0.6864905953407288, "learning_rate": 9.803825711247183e-05, "loss": 0.0838, "step": 10800 }, { "grad_norm": 0.47228601574897766, "learning_rate": 9.803252036254831e-05, "loss": 0.0643, "step": 10810 }, { "grad_norm": 0.44908326864242554, "learning_rate": 9.802677540518076e-05, "loss": 0.0844, "step": 10820 }, { "grad_norm": 0.44731730222702026, "learning_rate": 9.802102224135081e-05, "loss": 0.0862, "step": 10830 }, { "grad_norm": 0.647774338722229, "learning_rate": 9.801526087204155e-05, "loss": 0.0941, "step": 10840 }, { "grad_norm": 0.6506381034851074, "learning_rate": 9.800949129823743e-05, "loss": 0.0796, "step": 10850 }, { "grad_norm": 0.5751331448554993, "learning_rate": 9.80037135209243e-05, "loss": 0.0647, "step": 10860 }, { "grad_norm": 0.7092398405075073, "learning_rate": 9.799792754108946e-05, "loss": 0.077, "step": 10870 }, { "grad_norm": 0.49575045704841614, "learning_rate": 9.799213335972152e-05, "loss": 0.0774, "step": 10880 }, { "grad_norm": 0.8239098191261292, "learning_rate": 9.798633097781058e-05, "loss": 0.0776, "step": 10890 }, { "grad_norm": 0.7348482608795166, "learning_rate": 9.79805203963481e-05, "loss": 0.0925, "step": 10900 }, { "grad_norm": 0.4982756972312927, "learning_rate": 9.797470161632697e-05, "loss": 0.0774, "step": 10910 }, { "grad_norm": 0.5482870936393738, "learning_rate": 9.796887463874145e-05, "loss": 0.0762, "step": 10920 }, { "grad_norm": 0.5510295629501343, "learning_rate": 9.796303946458718e-05, "loss": 0.078, "step": 10930 }, { "grad_norm": 0.6686082482337952, "learning_rate": 9.795719609486127e-05, "loss": 0.0784, "step": 10940 }, { "grad_norm": 0.7605575919151306, "learning_rate": 9.795134453056219e-05, "loss": 0.0741, "step": 10950 }, { "grad_norm": 0.5193737149238586, "learning_rate": 9.794548477268979e-05, "loss": 0.0646, "step": 10960 }, { "grad_norm": 0.5099349021911621, "learning_rate": 9.793961682224537e-05, "loss": 0.073, "step": 10970 }, { "grad_norm": 0.6000353097915649, "learning_rate": 9.793374068023156e-05, "loss": 0.0856, "step": 10980 }, { "grad_norm": 0.47294554114341736, "learning_rate": 9.792785634765247e-05, "loss": 0.0734, "step": 10990 }, { "grad_norm": 0.5185801386833191, "learning_rate": 9.792196382551357e-05, "loss": 0.0811, "step": 11000 }, { "grad_norm": 0.5670230984687805, "learning_rate": 9.791606311482171e-05, "loss": 0.0734, "step": 11010 }, { "grad_norm": 0.5426215529441833, "learning_rate": 9.791015421658518e-05, "loss": 0.0745, "step": 11020 }, { "grad_norm": 0.5694929361343384, "learning_rate": 9.790423713181362e-05, "loss": 0.0812, "step": 11030 }, { "grad_norm": 0.5422756671905518, "learning_rate": 9.789831186151814e-05, "loss": 0.0698, "step": 11040 }, { "grad_norm": 0.5562644600868225, "learning_rate": 9.789237840671118e-05, "loss": 0.0701, "step": 11050 }, { "grad_norm": 0.5035049319267273, "learning_rate": 9.78864367684066e-05, "loss": 0.0681, "step": 11060 }, { "grad_norm": 0.536565899848938, "learning_rate": 9.788048694761968e-05, "loss": 0.0752, "step": 11070 }, { "grad_norm": 0.5079371333122253, "learning_rate": 9.787452894536709e-05, "loss": 0.0769, "step": 11080 }, { "grad_norm": 0.3986671566963196, "learning_rate": 9.786856276266685e-05, "loss": 0.0609, "step": 11090 }, { "grad_norm": 0.5291951298713684, "learning_rate": 9.786258840053845e-05, "loss": 0.0729, "step": 11100 }, { "grad_norm": 0.579871654510498, "learning_rate": 9.785660586000273e-05, "loss": 0.0735, "step": 11110 }, { "grad_norm": 0.4674587547779083, "learning_rate": 9.785061514208196e-05, "loss": 0.0631, "step": 11120 }, { "grad_norm": 0.6435877680778503, "learning_rate": 9.784461624779977e-05, "loss": 0.0744, "step": 11130 }, { "grad_norm": 0.5933992266654968, "learning_rate": 9.783860917818123e-05, "loss": 0.0671, "step": 11140 }, { "grad_norm": 0.44079241156578064, "learning_rate": 9.783259393425277e-05, "loss": 0.077, "step": 11150 }, { "grad_norm": 0.5367472171783447, "learning_rate": 9.782657051704221e-05, "loss": 0.0635, "step": 11160 }, { "grad_norm": 0.4848480820655823, "learning_rate": 9.782053892757883e-05, "loss": 0.0718, "step": 11170 }, { "grad_norm": 0.3771425187587738, "learning_rate": 9.781449916689324e-05, "loss": 0.0702, "step": 11180 }, { "grad_norm": 0.3765457570552826, "learning_rate": 9.780845123601746e-05, "loss": 0.0779, "step": 11190 }, { "grad_norm": 0.4909181594848633, "learning_rate": 9.780239513598492e-05, "loss": 0.0768, "step": 11200 }, { "grad_norm": 0.6328291893005371, "learning_rate": 9.779633086783047e-05, "loss": 0.072, "step": 11210 }, { "grad_norm": 0.6227254867553711, "learning_rate": 9.779025843259031e-05, "loss": 0.0683, "step": 11220 }, { "grad_norm": 0.4983925521373749, "learning_rate": 9.778417783130204e-05, "loss": 0.0858, "step": 11230 }, { "grad_norm": 0.49418795108795166, "learning_rate": 9.777808906500468e-05, "loss": 0.0622, "step": 11240 }, { "grad_norm": 0.5137078166007996, "learning_rate": 9.777199213473862e-05, "loss": 0.0582, "step": 11250 }, { "grad_norm": 0.411769300699234, "learning_rate": 9.77658870415457e-05, "loss": 0.0595, "step": 11260 }, { "grad_norm": 0.4208316206932068, "learning_rate": 9.775977378646906e-05, "loss": 0.0773, "step": 11270 }, { "grad_norm": 0.6169866919517517, "learning_rate": 9.775365237055331e-05, "loss": 0.0834, "step": 11280 }, { "grad_norm": 0.586180567741394, "learning_rate": 9.774752279484445e-05, "loss": 0.0687, "step": 11290 }, { "grad_norm": 0.5854020714759827, "learning_rate": 9.774138506038984e-05, "loss": 0.0683, "step": 11300 }, { "grad_norm": 0.7412139773368835, "learning_rate": 9.773523916823826e-05, "loss": 0.0637, "step": 11310 }, { "grad_norm": 0.42965027689933777, "learning_rate": 9.772908511943986e-05, "loss": 0.0635, "step": 11320 }, { "grad_norm": 0.658414900302887, "learning_rate": 9.77229229150462e-05, "loss": 0.0774, "step": 11330 }, { "grad_norm": 0.5130285620689392, "learning_rate": 9.771675255611024e-05, "loss": 0.0693, "step": 11340 }, { "grad_norm": 0.45580926537513733, "learning_rate": 9.771057404368632e-05, "loss": 0.0888, "step": 11350 }, { "grad_norm": 0.4311416745185852, "learning_rate": 9.770438737883018e-05, "loss": 0.0722, "step": 11360 }, { "grad_norm": 0.44391363859176636, "learning_rate": 9.769819256259898e-05, "loss": 0.0571, "step": 11370 }, { "grad_norm": 0.6147053241729736, "learning_rate": 9.769198959605119e-05, "loss": 0.064, "step": 11380 }, { "grad_norm": 0.35504254698753357, "learning_rate": 9.768577848024678e-05, "loss": 0.0805, "step": 11390 }, { "grad_norm": 0.4842676818370819, "learning_rate": 9.767955921624702e-05, "loss": 0.0735, "step": 11400 }, { "grad_norm": 0.653027355670929, "learning_rate": 9.767333180511465e-05, "loss": 0.0643, "step": 11410 }, { "grad_norm": 0.47115838527679443, "learning_rate": 9.766709624791373e-05, "loss": 0.0974, "step": 11420 }, { "grad_norm": 0.49486812949180603, "learning_rate": 9.766085254570975e-05, "loss": 0.0758, "step": 11430 }, { "grad_norm": 0.6159383654594421, "learning_rate": 9.76546006995696e-05, "loss": 0.0786, "step": 11440 }, { "grad_norm": 0.535308837890625, "learning_rate": 9.764834071056155e-05, "loss": 0.0697, "step": 11450 }, { "grad_norm": 0.44005048274993896, "learning_rate": 9.764207257975526e-05, "loss": 0.0687, "step": 11460 }, { "grad_norm": 0.5864790678024292, "learning_rate": 9.763579630822179e-05, "loss": 0.053, "step": 11470 }, { "grad_norm": 0.430507093667984, "learning_rate": 9.762951189703356e-05, "loss": 0.0644, "step": 11480 }, { "grad_norm": 0.5664432644844055, "learning_rate": 9.762321934726442e-05, "loss": 0.0707, "step": 11490 }, { "grad_norm": 0.49748295545578003, "learning_rate": 9.761691865998959e-05, "loss": 0.0582, "step": 11500 }, { "grad_norm": 0.4422622323036194, "learning_rate": 9.76106098362857e-05, "loss": 0.0636, "step": 11510 }, { "grad_norm": 0.40253937244415283, "learning_rate": 9.760429287723072e-05, "loss": 0.059, "step": 11520 }, { "grad_norm": 0.5248969793319702, "learning_rate": 9.759796778390406e-05, "loss": 0.067, "step": 11530 }, { "grad_norm": 0.5458329916000366, "learning_rate": 9.759163455738653e-05, "loss": 0.0636, "step": 11540 }, { "grad_norm": 0.5857506990432739, "learning_rate": 9.75852931987603e-05, "loss": 0.0737, "step": 11550 }, { "grad_norm": 0.4872324764728546, "learning_rate": 9.757894370910891e-05, "loss": 0.0644, "step": 11560 }, { "grad_norm": 0.4470299184322357, "learning_rate": 9.757258608951733e-05, "loss": 0.0664, "step": 11570 }, { "grad_norm": 0.7716564536094666, "learning_rate": 9.75662203410719e-05, "loss": 0.0837, "step": 11580 }, { "grad_norm": 0.5879382491111755, "learning_rate": 9.755984646486034e-05, "loss": 0.0715, "step": 11590 }, { "grad_norm": 0.5153780579566956, "learning_rate": 9.75534644619718e-05, "loss": 0.0665, "step": 11600 }, { "grad_norm": 0.4723442494869232, "learning_rate": 9.754707433349676e-05, "loss": 0.0895, "step": 11610 }, { "grad_norm": 0.6807955503463745, "learning_rate": 9.754067608052715e-05, "loss": 0.0875, "step": 11620 }, { "grad_norm": 0.490003377199173, "learning_rate": 9.753426970415622e-05, "loss": 0.0924, "step": 11630 }, { "grad_norm": 0.560819149017334, "learning_rate": 9.752785520547868e-05, "loss": 0.0609, "step": 11640 }, { "grad_norm": 0.6355839371681213, "learning_rate": 9.752143258559056e-05, "loss": 0.0738, "step": 11650 }, { "grad_norm": 0.4745175242424011, "learning_rate": 9.751500184558933e-05, "loss": 0.0639, "step": 11660 }, { "grad_norm": 0.4021434783935547, "learning_rate": 9.750856298657383e-05, "loss": 0.0548, "step": 11670 }, { "grad_norm": 0.44132041931152344, "learning_rate": 9.750211600964428e-05, "loss": 0.069, "step": 11680 }, { "grad_norm": 0.49351242184638977, "learning_rate": 9.749566091590226e-05, "loss": 0.064, "step": 11690 }, { "grad_norm": 0.7643229365348816, "learning_rate": 9.748919770645083e-05, "loss": 0.0628, "step": 11700 }, { "grad_norm": 0.6214596629142761, "learning_rate": 9.748272638239432e-05, "loss": 0.0755, "step": 11710 }, { "grad_norm": 0.5105453729629517, "learning_rate": 9.747624694483855e-05, "loss": 0.0772, "step": 11720 }, { "grad_norm": 0.4723168611526489, "learning_rate": 9.746975939489065e-05, "loss": 0.0687, "step": 11730 }, { "grad_norm": 0.5814576745033264, "learning_rate": 9.746326373365918e-05, "loss": 0.0722, "step": 11740 }, { "grad_norm": 0.4904615879058838, "learning_rate": 9.745675996225403e-05, "loss": 0.0776, "step": 11750 }, { "grad_norm": 0.45396631956100464, "learning_rate": 9.745024808178657e-05, "loss": 0.0691, "step": 11760 }, { "grad_norm": 0.47870612144470215, "learning_rate": 9.744372809336947e-05, "loss": 0.0644, "step": 11770 }, { "grad_norm": 0.5829415321350098, "learning_rate": 9.743719999811682e-05, "loss": 0.0636, "step": 11780 }, { "grad_norm": 0.5215325355529785, "learning_rate": 9.743066379714412e-05, "loss": 0.0669, "step": 11790 }, { "grad_norm": 0.45286449790000916, "learning_rate": 9.74241194915682e-05, "loss": 0.0628, "step": 11800 }, { "grad_norm": 0.4775339663028717, "learning_rate": 9.741756708250731e-05, "loss": 0.08, "step": 11810 }, { "grad_norm": 0.4450395107269287, "learning_rate": 9.741100657108109e-05, "loss": 0.0732, "step": 11820 }, { "grad_norm": 0.5050836205482483, "learning_rate": 9.740443795841054e-05, "loss": 0.0808, "step": 11830 }, { "grad_norm": 0.5468694567680359, "learning_rate": 9.739786124561805e-05, "loss": 0.0764, "step": 11840 }, { "grad_norm": 0.42162999510765076, "learning_rate": 9.73912764338274e-05, "loss": 0.0655, "step": 11850 }, { "grad_norm": 0.48114511370658875, "learning_rate": 9.738468352416377e-05, "loss": 0.0783, "step": 11860 }, { "grad_norm": 0.5048590898513794, "learning_rate": 9.737808251775369e-05, "loss": 0.0666, "step": 11870 }, { "grad_norm": 0.8306834697723389, "learning_rate": 9.737147341572512e-05, "loss": 0.0878, "step": 11880 }, { "grad_norm": 0.5565890073776245, "learning_rate": 9.736485621920735e-05, "loss": 0.0648, "step": 11890 }, { "grad_norm": 0.4144880175590515, "learning_rate": 9.735823092933108e-05, "loss": 0.0649, "step": 11900 }, { "grad_norm": 0.4737222492694855, "learning_rate": 9.735159754722838e-05, "loss": 0.0744, "step": 11910 }, { "grad_norm": 0.4629839062690735, "learning_rate": 9.734495607403275e-05, "loss": 0.053, "step": 11920 }, { "grad_norm": 0.4531034827232361, "learning_rate": 9.733830651087901e-05, "loss": 0.0676, "step": 11930 }, { "grad_norm": 0.4148541986942291, "learning_rate": 9.733164885890338e-05, "loss": 0.0626, "step": 11940 }, { "grad_norm": 0.32764798402786255, "learning_rate": 9.732498311924349e-05, "loss": 0.0512, "step": 11950 }, { "grad_norm": 0.4466022849082947, "learning_rate": 9.731830929303833e-05, "loss": 0.0534, "step": 11960 }, { "grad_norm": 0.4698883891105652, "learning_rate": 9.731162738142827e-05, "loss": 0.0551, "step": 11970 }, { "grad_norm": 0.5261984467506409, "learning_rate": 9.730493738555506e-05, "loss": 0.0723, "step": 11980 }, { "grad_norm": 0.4046032726764679, "learning_rate": 9.729823930656186e-05, "loss": 0.0724, "step": 11990 }, { "grad_norm": 0.588295042514801, "learning_rate": 9.729153314559316e-05, "loss": 0.0791, "step": 12000 }, { "grad_norm": 0.5625817775726318, "learning_rate": 9.728481890379486e-05, "loss": 0.0693, "step": 12010 }, { "grad_norm": 0.6622326970100403, "learning_rate": 9.727809658231428e-05, "loss": 0.0653, "step": 12020 }, { "grad_norm": 0.5666773319244385, "learning_rate": 9.727136618230003e-05, "loss": 0.0651, "step": 12030 }, { "grad_norm": 0.6852526664733887, "learning_rate": 9.726462770490219e-05, "loss": 0.0681, "step": 12040 }, { "grad_norm": 0.49590060114860535, "learning_rate": 9.725788115127214e-05, "loss": 0.0664, "step": 12050 }, { "grad_norm": 0.7259113788604736, "learning_rate": 9.725112652256274e-05, "loss": 0.0748, "step": 12060 }, { "grad_norm": 0.5394326448440552, "learning_rate": 9.724436381992812e-05, "loss": 0.0916, "step": 12070 }, { "grad_norm": 0.4830283522605896, "learning_rate": 9.723759304452387e-05, "loss": 0.0664, "step": 12080 }, { "grad_norm": 0.5555049180984497, "learning_rate": 9.72308141975069e-05, "loss": 0.057, "step": 12090 }, { "grad_norm": 0.43221741914749146, "learning_rate": 9.722402728003557e-05, "loss": 0.0638, "step": 12100 }, { "grad_norm": 0.6105766892433167, "learning_rate": 9.721723229326953e-05, "loss": 0.064, "step": 12110 }, { "grad_norm": 0.39734670519828796, "learning_rate": 9.721042923836992e-05, "loss": 0.0604, "step": 12120 }, { "grad_norm": 0.3830042779445648, "learning_rate": 9.720361811649914e-05, "loss": 0.0605, "step": 12130 }, { "grad_norm": 0.8738731145858765, "learning_rate": 9.719679892882106e-05, "loss": 0.0707, "step": 12140 }, { "grad_norm": 0.5566120743751526, "learning_rate": 9.718997167650085e-05, "loss": 0.0757, "step": 12150 }, { "grad_norm": 0.6000269651412964, "learning_rate": 9.718313636070515e-05, "loss": 0.0772, "step": 12160 }, { "grad_norm": 0.4536004662513733, "learning_rate": 9.717629298260192e-05, "loss": 0.0694, "step": 12170 }, { "grad_norm": 0.562367856502533, "learning_rate": 9.716944154336047e-05, "loss": 0.0822, "step": 12180 }, { "grad_norm": 0.6537054777145386, "learning_rate": 9.716258204415157e-05, "loss": 0.0645, "step": 12190 }, { "grad_norm": 0.5291623473167419, "learning_rate": 9.715571448614728e-05, "loss": 0.0774, "step": 12200 }, { "grad_norm": 0.7994457483291626, "learning_rate": 9.71488388705211e-05, "loss": 0.0852, "step": 12210 }, { "grad_norm": 0.5933327674865723, "learning_rate": 9.714195519844788e-05, "loss": 0.0711, "step": 12220 }, { "grad_norm": 0.5203611850738525, "learning_rate": 9.713506347110386e-05, "loss": 0.0678, "step": 12230 }, { "grad_norm": 0.45026087760925293, "learning_rate": 9.712816368966662e-05, "loss": 0.0718, "step": 12240 }, { "grad_norm": 0.4451008439064026, "learning_rate": 9.712125585531517e-05, "loss": 0.0817, "step": 12250 }, { "grad_norm": 0.6177807450294495, "learning_rate": 9.711433996922988e-05, "loss": 0.0791, "step": 12260 }, { "grad_norm": 0.4276295304298401, "learning_rate": 9.710741603259245e-05, "loss": 0.0713, "step": 12270 }, { "grad_norm": 0.5471391081809998, "learning_rate": 9.710048404658603e-05, "loss": 0.0669, "step": 12280 }, { "grad_norm": 0.5482802391052246, "learning_rate": 9.709354401239508e-05, "loss": 0.0801, "step": 12290 }, { "grad_norm": 0.5220881104469299, "learning_rate": 9.708659593120546e-05, "loss": 0.0648, "step": 12300 }, { "grad_norm": 0.45205268263816833, "learning_rate": 9.707963980420443e-05, "loss": 0.0627, "step": 12310 }, { "grad_norm": 0.596621572971344, "learning_rate": 9.707267563258058e-05, "loss": 0.065, "step": 12320 }, { "grad_norm": 0.42880433797836304, "learning_rate": 9.70657034175239e-05, "loss": 0.0616, "step": 12330 }, { "grad_norm": 0.5157198905944824, "learning_rate": 9.705872316022577e-05, "loss": 0.0719, "step": 12340 }, { "grad_norm": 0.5503883957862854, "learning_rate": 9.705173486187891e-05, "loss": 0.0636, "step": 12350 }, { "grad_norm": 0.4626397490501404, "learning_rate": 9.704473852367741e-05, "loss": 0.0818, "step": 12360 }, { "grad_norm": 0.6194549202919006, "learning_rate": 9.70377341468168e-05, "loss": 0.0588, "step": 12370 }, { "grad_norm": 0.6419069170951843, "learning_rate": 9.703072173249389e-05, "loss": 0.0689, "step": 12380 }, { "grad_norm": 0.6978343725204468, "learning_rate": 9.702370128190693e-05, "loss": 0.0884, "step": 12390 }, { "grad_norm": 0.5474632382392883, "learning_rate": 9.701667279625552e-05, "loss": 0.0724, "step": 12400 }, { "grad_norm": 0.5142436027526855, "learning_rate": 9.700963627674065e-05, "loss": 0.0743, "step": 12410 }, { "grad_norm": 0.6927314400672913, "learning_rate": 9.700259172456466e-05, "loss": 0.0624, "step": 12420 }, { "grad_norm": 0.6145099997520447, "learning_rate": 9.699553914093124e-05, "loss": 0.0868, "step": 12430 }, { "grad_norm": 0.571753203868866, "learning_rate": 9.698847852704553e-05, "loss": 0.0698, "step": 12440 }, { "grad_norm": 0.47922632098197937, "learning_rate": 9.6981409884114e-05, "loss": 0.0656, "step": 12450 }, { "grad_norm": 0.4435444176197052, "learning_rate": 9.697433321334443e-05, "loss": 0.0748, "step": 12460 }, { "grad_norm": 0.5951995253562927, "learning_rate": 9.696724851594607e-05, "loss": 0.0653, "step": 12470 }, { "grad_norm": 0.4378683865070343, "learning_rate": 9.696015579312952e-05, "loss": 0.0641, "step": 12480 }, { "grad_norm": 0.5260929465293884, "learning_rate": 9.695305504610668e-05, "loss": 0.0677, "step": 12490 }, { "grad_norm": 0.45083630084991455, "learning_rate": 9.694594627609092e-05, "loss": 0.0693, "step": 12500 }, { "grad_norm": 0.5430079102516174, "learning_rate": 9.693882948429691e-05, "loss": 0.0809, "step": 12510 }, { "grad_norm": 0.7170400023460388, "learning_rate": 9.693170467194071e-05, "loss": 0.0715, "step": 12520 }, { "grad_norm": 0.605512797832489, "learning_rate": 9.692457184023977e-05, "loss": 0.0696, "step": 12530 }, { "grad_norm": 0.476185142993927, "learning_rate": 9.691743099041291e-05, "loss": 0.0731, "step": 12540 }, { "grad_norm": 0.5325929522514343, "learning_rate": 9.691028212368027e-05, "loss": 0.0727, "step": 12550 }, { "grad_norm": 0.5928123593330383, "learning_rate": 9.690312524126342e-05, "loss": 0.0734, "step": 12560 }, { "grad_norm": 0.49097779393196106, "learning_rate": 9.689596034438527e-05, "loss": 0.079, "step": 12570 }, { "grad_norm": 0.39969661831855774, "learning_rate": 9.688878743427012e-05, "loss": 0.071, "step": 12580 }, { "grad_norm": 0.3578379452228546, "learning_rate": 9.688160651214359e-05, "loss": 0.0754, "step": 12590 }, { "grad_norm": 0.5474270582199097, "learning_rate": 9.687441757923273e-05, "loss": 0.0652, "step": 12600 }, { "grad_norm": 0.47615525126457214, "learning_rate": 9.68672206367659e-05, "loss": 0.0638, "step": 12610 }, { "grad_norm": 0.5454956293106079, "learning_rate": 9.686001568597291e-05, "loss": 0.0637, "step": 12620 }, { "grad_norm": 0.7532147169113159, "learning_rate": 9.685280272808486e-05, "loss": 0.0621, "step": 12630 }, { "grad_norm": 0.6357566118240356, "learning_rate": 9.684558176433424e-05, "loss": 0.0737, "step": 12640 }, { "grad_norm": 0.6361411809921265, "learning_rate": 9.683835279595495e-05, "loss": 0.0673, "step": 12650 }, { "grad_norm": 0.6801282167434692, "learning_rate": 9.683111582418216e-05, "loss": 0.082, "step": 12660 }, { "grad_norm": 0.5117168426513672, "learning_rate": 9.682387085025254e-05, "loss": 0.0769, "step": 12670 }, { "grad_norm": 0.48192283511161804, "learning_rate": 9.681661787540401e-05, "loss": 0.0818, "step": 12680 }, { "grad_norm": 0.5621492862701416, "learning_rate": 9.680935690087593e-05, "loss": 0.0718, "step": 12690 }, { "grad_norm": 0.47694218158721924, "learning_rate": 9.680208792790901e-05, "loss": 0.0569, "step": 12700 }, { "grad_norm": 0.4911132752895355, "learning_rate": 9.679481095774529e-05, "loss": 0.0702, "step": 12710 }, { "grad_norm": 0.6871297955513, "learning_rate": 9.678752599162822e-05, "loss": 0.0793, "step": 12720 }, { "grad_norm": 0.47312065958976746, "learning_rate": 9.678023303080259e-05, "loss": 0.0616, "step": 12730 }, { "grad_norm": 0.45213884115219116, "learning_rate": 9.677293207651459e-05, "loss": 0.0649, "step": 12740 }, { "grad_norm": 0.7146787047386169, "learning_rate": 9.676562313001173e-05, "loss": 0.0638, "step": 12750 }, { "grad_norm": 0.6128819584846497, "learning_rate": 9.675830619254293e-05, "loss": 0.0787, "step": 12760 }, { "grad_norm": 0.5871028900146484, "learning_rate": 9.675098126535843e-05, "loss": 0.0845, "step": 12770 }, { "grad_norm": 0.7511138319969177, "learning_rate": 9.674364834970988e-05, "loss": 0.081, "step": 12780 }, { "grad_norm": 0.4995322525501251, "learning_rate": 9.673630744685028e-05, "loss": 0.083, "step": 12790 }, { "grad_norm": 0.5993275046348572, "learning_rate": 9.672895855803397e-05, "loss": 0.0709, "step": 12800 }, { "grad_norm": 0.36088094115257263, "learning_rate": 9.672160168451667e-05, "loss": 0.0737, "step": 12810 }, { "grad_norm": 1.260447382926941, "learning_rate": 9.671423682755549e-05, "loss": 0.0761, "step": 12820 }, { "grad_norm": 0.7268471717834473, "learning_rate": 9.670686398840888e-05, "loss": 0.0791, "step": 12830 }, { "grad_norm": 0.49616819620132446, "learning_rate": 9.669948316833664e-05, "loss": 0.0621, "step": 12840 }, { "grad_norm": 0.5024594068527222, "learning_rate": 9.669209436859997e-05, "loss": 0.0609, "step": 12850 }, { "grad_norm": 0.46229270100593567, "learning_rate": 9.66846975904614e-05, "loss": 0.0656, "step": 12860 }, { "grad_norm": 0.7428377866744995, "learning_rate": 9.667729283518483e-05, "loss": 0.0578, "step": 12870 }, { "grad_norm": 0.41377025842666626, "learning_rate": 9.666988010403557e-05, "loss": 0.0653, "step": 12880 }, { "grad_norm": 0.40705031156539917, "learning_rate": 9.66624593982802e-05, "loss": 0.0613, "step": 12890 }, { "grad_norm": 0.4588109850883484, "learning_rate": 9.665503071918675e-05, "loss": 0.0786, "step": 12900 }, { "grad_norm": 0.6304764747619629, "learning_rate": 9.664759406802456e-05, "loss": 0.0665, "step": 12910 }, { "grad_norm": 0.5646094083786011, "learning_rate": 9.664014944606437e-05, "loss": 0.0616, "step": 12920 }, { "grad_norm": 0.4714445173740387, "learning_rate": 9.663269685457822e-05, "loss": 0.0731, "step": 12930 }, { "grad_norm": 0.551657497882843, "learning_rate": 9.662523629483962e-05, "loss": 0.0671, "step": 12940 }, { "grad_norm": 0.5848498344421387, "learning_rate": 9.661776776812333e-05, "loss": 0.0622, "step": 12950 }, { "grad_norm": 0.6234254240989685, "learning_rate": 9.661029127570553e-05, "loss": 0.0572, "step": 12960 }, { "grad_norm": 0.4790451228618622, "learning_rate": 9.660280681886373e-05, "loss": 0.0696, "step": 12970 }, { "grad_norm": 0.4485260248184204, "learning_rate": 9.659531439887685e-05, "loss": 0.0679, "step": 12980 }, { "grad_norm": 0.44148120284080505, "learning_rate": 9.658781401702511e-05, "loss": 0.0636, "step": 12990 }, { "grad_norm": 0.4859919250011444, "learning_rate": 9.658030567459015e-05, "loss": 0.0643, "step": 13000 }, { "grad_norm": 0.5333834886550903, "learning_rate": 9.65727893728549e-05, "loss": 0.0816, "step": 13010 }, { "grad_norm": 0.7074456214904785, "learning_rate": 9.656526511310375e-05, "loss": 0.0927, "step": 13020 }, { "grad_norm": 0.7111431956291199, "learning_rate": 9.655773289662233e-05, "loss": 0.0849, "step": 13030 }, { "grad_norm": 0.5401679873466492, "learning_rate": 9.655019272469772e-05, "loss": 0.0633, "step": 13040 }, { "grad_norm": 0.4485662579536438, "learning_rate": 9.654264459861832e-05, "loss": 0.0563, "step": 13050 }, { "grad_norm": 0.5540158152580261, "learning_rate": 9.653508851967391e-05, "loss": 0.0649, "step": 13060 }, { "grad_norm": 0.5042420029640198, "learning_rate": 9.65275244891556e-05, "loss": 0.0855, "step": 13070 }, { "grad_norm": 0.6029703617095947, "learning_rate": 9.651995250835591e-05, "loss": 0.0654, "step": 13080 }, { "grad_norm": 0.4641766846179962, "learning_rate": 9.651237257856862e-05, "loss": 0.0642, "step": 13090 }, { "grad_norm": 0.4543335437774658, "learning_rate": 9.6504784701089e-05, "loss": 0.0599, "step": 13100 }, { "grad_norm": 0.3718896210193634, "learning_rate": 9.649718887721357e-05, "loss": 0.0571, "step": 13110 }, { "grad_norm": 0.4866885542869568, "learning_rate": 9.648958510824028e-05, "loss": 0.0617, "step": 13120 }, { "grad_norm": 0.4857250452041626, "learning_rate": 9.648197339546837e-05, "loss": 0.0775, "step": 13130 }, { "grad_norm": 0.5358465313911438, "learning_rate": 9.647435374019851e-05, "loss": 0.066, "step": 13140 }, { "grad_norm": 0.521448016166687, "learning_rate": 9.646672614373266e-05, "loss": 0.0686, "step": 13150 }, { "grad_norm": 0.4888678193092346, "learning_rate": 9.645909060737418e-05, "loss": 0.0629, "step": 13160 }, { "grad_norm": 0.3755582273006439, "learning_rate": 9.645144713242778e-05, "loss": 0.06, "step": 13170 }, { "grad_norm": 0.5989090204238892, "learning_rate": 9.64437957201995e-05, "loss": 0.0622, "step": 13180 }, { "grad_norm": 0.771467387676239, "learning_rate": 9.643613637199678e-05, "loss": 0.0643, "step": 13190 }, { "grad_norm": 0.5663743019104004, "learning_rate": 9.642846908912839e-05, "loss": 0.0796, "step": 13200 }, { "grad_norm": 0.3084070384502411, "learning_rate": 9.642079387290444e-05, "loss": 0.0671, "step": 13210 }, { "grad_norm": 0.7612932920455933, "learning_rate": 9.641311072463644e-05, "loss": 0.0634, "step": 13220 }, { "grad_norm": 0.4363883137702942, "learning_rate": 9.640541964563722e-05, "loss": 0.0581, "step": 13230 }, { "grad_norm": 0.5292156338691711, "learning_rate": 9.639772063722096e-05, "loss": 0.0658, "step": 13240 }, { "grad_norm": 0.4274076223373413, "learning_rate": 9.639001370070324e-05, "loss": 0.0611, "step": 13250 }, { "grad_norm": 0.72112637758255, "learning_rate": 9.638229883740095e-05, "loss": 0.0613, "step": 13260 }, { "grad_norm": 0.4449668824672699, "learning_rate": 9.637457604863233e-05, "loss": 0.0578, "step": 13270 }, { "grad_norm": 0.6300351619720459, "learning_rate": 9.636684533571703e-05, "loss": 0.0683, "step": 13280 }, { "grad_norm": 0.3811010718345642, "learning_rate": 9.635910669997599e-05, "loss": 0.0645, "step": 13290 }, { "grad_norm": 0.560806393623352, "learning_rate": 9.635136014273154e-05, "loss": 0.0522, "step": 13300 }, { "grad_norm": 0.7318065166473389, "learning_rate": 9.634360566530735e-05, "loss": 0.069, "step": 13310 }, { "grad_norm": 0.553662121295929, "learning_rate": 9.633584326902845e-05, "loss": 0.0784, "step": 13320 }, { "grad_norm": 0.6367096900939941, "learning_rate": 9.632807295522124e-05, "loss": 0.0717, "step": 13330 }, { "grad_norm": 0.6002030372619629, "learning_rate": 9.632029472521342e-05, "loss": 0.0715, "step": 13340 }, { "grad_norm": 0.44375118613243103, "learning_rate": 9.631250858033409e-05, "loss": 0.0605, "step": 13350 }, { "grad_norm": 0.3751848042011261, "learning_rate": 9.630471452191371e-05, "loss": 0.0584, "step": 13360 }, { "grad_norm": 0.3667372167110443, "learning_rate": 9.629691255128405e-05, "loss": 0.0704, "step": 13370 }, { "grad_norm": 0.6345521211624146, "learning_rate": 9.628910266977825e-05, "loss": 0.0699, "step": 13380 }, { "grad_norm": 0.478718638420105, "learning_rate": 9.628128487873083e-05, "loss": 0.0692, "step": 13390 }, { "grad_norm": 0.4660668969154358, "learning_rate": 9.627345917947761e-05, "loss": 0.0734, "step": 13400 }, { "grad_norm": 0.5311440825462341, "learning_rate": 9.626562557335579e-05, "loss": 0.0682, "step": 13410 }, { "grad_norm": 0.6861147880554199, "learning_rate": 9.625778406170393e-05, "loss": 0.0724, "step": 13420 }, { "grad_norm": 0.6200295090675354, "learning_rate": 9.624993464586193e-05, "loss": 0.0782, "step": 13430 }, { "grad_norm": 0.6263048648834229, "learning_rate": 9.624207732717105e-05, "loss": 0.0804, "step": 13440 }, { "grad_norm": 0.7687500715255737, "learning_rate": 9.623421210697386e-05, "loss": 0.1032, "step": 13450 }, { "grad_norm": 0.49720898270606995, "learning_rate": 9.622633898661434e-05, "loss": 0.0798, "step": 13460 }, { "grad_norm": 0.4158327877521515, "learning_rate": 9.621845796743778e-05, "loss": 0.0815, "step": 13470 }, { "grad_norm": 0.4036085307598114, "learning_rate": 9.621056905079082e-05, "loss": 0.0686, "step": 13480 }, { "grad_norm": 0.5098157525062561, "learning_rate": 9.620267223802149e-05, "loss": 0.073, "step": 13490 }, { "grad_norm": 0.6662095785140991, "learning_rate": 9.619476753047911e-05, "loss": 0.0598, "step": 13500 }, { "grad_norm": 0.660835325717926, "learning_rate": 9.618685492951438e-05, "loss": 0.0788, "step": 13510 }, { "grad_norm": 0.38061070442199707, "learning_rate": 9.617893443647938e-05, "loss": 0.0751, "step": 13520 }, { "grad_norm": 0.5206005573272705, "learning_rate": 9.617100605272746e-05, "loss": 0.0775, "step": 13530 }, { "grad_norm": 0.42081236839294434, "learning_rate": 9.616306977961338e-05, "loss": 0.0726, "step": 13540 }, { "grad_norm": 0.49378225207328796, "learning_rate": 9.615512561849326e-05, "loss": 0.0597, "step": 13550 }, { "grad_norm": 0.550769567489624, "learning_rate": 9.61471735707245e-05, "loss": 0.0653, "step": 13560 }, { "grad_norm": 0.575873076915741, "learning_rate": 9.613921363766592e-05, "loss": 0.0673, "step": 13570 }, { "grad_norm": 0.5380557775497437, "learning_rate": 9.613124582067763e-05, "loss": 0.0648, "step": 13580 }, { "grad_norm": 0.5639264583587646, "learning_rate": 9.612327012112112e-05, "loss": 0.0841, "step": 13590 }, { "grad_norm": 0.45866987109184265, "learning_rate": 9.611528654035921e-05, "loss": 0.0892, "step": 13600 }, { "grad_norm": 0.4084102511405945, "learning_rate": 9.610729507975611e-05, "loss": 0.0644, "step": 13610 }, { "grad_norm": 0.6051228642463684, "learning_rate": 9.609929574067731e-05, "loss": 0.0806, "step": 13620 }, { "grad_norm": 0.38226059079170227, "learning_rate": 9.609128852448967e-05, "loss": 0.0665, "step": 13630 }, { "grad_norm": 1.3009254932403564, "learning_rate": 9.608327343256143e-05, "loss": 0.0644, "step": 13640 }, { "grad_norm": 0.5358849167823792, "learning_rate": 9.607525046626216e-05, "loss": 0.0736, "step": 13650 }, { "grad_norm": 0.6781689524650574, "learning_rate": 9.606721962696272e-05, "loss": 0.0763, "step": 13660 }, { "grad_norm": 0.6004512906074524, "learning_rate": 9.60591809160354e-05, "loss": 0.0667, "step": 13670 }, { "grad_norm": 0.5541675090789795, "learning_rate": 9.605113433485378e-05, "loss": 0.0704, "step": 13680 }, { "grad_norm": 0.5082460045814514, "learning_rate": 9.604307988479279e-05, "loss": 0.0597, "step": 13690 }, { "grad_norm": 0.43947669863700867, "learning_rate": 9.603501756722876e-05, "loss": 0.0598, "step": 13700 }, { "grad_norm": 0.3832433521747589, "learning_rate": 9.602694738353927e-05, "loss": 0.0588, "step": 13710 }, { "grad_norm": 0.47610709071159363, "learning_rate": 9.601886933510331e-05, "loss": 0.056, "step": 13720 }, { "grad_norm": 0.411920428276062, "learning_rate": 9.60107834233012e-05, "loss": 0.0546, "step": 13730 }, { "grad_norm": 0.5076626539230347, "learning_rate": 9.60026896495146e-05, "loss": 0.0482, "step": 13740 }, { "grad_norm": 0.45453810691833496, "learning_rate": 9.599458801512652e-05, "loss": 0.0571, "step": 13750 }, { "grad_norm": 0.5052747130393982, "learning_rate": 9.598647852152129e-05, "loss": 0.079, "step": 13760 }, { "grad_norm": 0.4577225148677826, "learning_rate": 9.597836117008462e-05, "loss": 0.057, "step": 13770 }, { "grad_norm": 0.4830833077430725, "learning_rate": 9.597023596220356e-05, "loss": 0.0601, "step": 13780 }, { "grad_norm": 0.5375869274139404, "learning_rate": 9.596210289926643e-05, "loss": 0.0664, "step": 13790 }, { "grad_norm": 0.49669861793518066, "learning_rate": 9.5953961982663e-05, "loss": 0.0681, "step": 13800 }, { "grad_norm": 0.5762737393379211, "learning_rate": 9.594581321378431e-05, "loss": 0.0669, "step": 13810 }, { "grad_norm": 0.4413909316062927, "learning_rate": 9.593765659402276e-05, "loss": 0.0678, "step": 13820 }, { "grad_norm": 0.7746898531913757, "learning_rate": 9.59294921247721e-05, "loss": 0.0689, "step": 13830 }, { "grad_norm": 0.46685364842414856, "learning_rate": 9.59213198074274e-05, "loss": 0.0815, "step": 13840 }, { "grad_norm": 0.8082584738731384, "learning_rate": 9.59131396433851e-05, "loss": 0.0683, "step": 13850 }, { "grad_norm": 0.6056733727455139, "learning_rate": 9.590495163404297e-05, "loss": 0.0626, "step": 13860 }, { "grad_norm": 0.4908236265182495, "learning_rate": 9.589675578080009e-05, "loss": 0.0713, "step": 13870 }, { "grad_norm": 0.6593511700630188, "learning_rate": 9.588855208505694e-05, "loss": 0.0607, "step": 13880 }, { "grad_norm": 0.7911823987960815, "learning_rate": 9.588034054821529e-05, "loss": 0.0782, "step": 13890 }, { "grad_norm": 0.48710983991622925, "learning_rate": 9.587212117167826e-05, "loss": 0.0662, "step": 13900 }, { "grad_norm": 0.6186926364898682, "learning_rate": 9.586389395685033e-05, "loss": 0.067, "step": 13910 }, { "grad_norm": 0.48065024614334106, "learning_rate": 9.585565890513733e-05, "loss": 0.0712, "step": 13920 }, { "grad_norm": 0.47172605991363525, "learning_rate": 9.584741601794636e-05, "loss": 0.0692, "step": 13930 }, { "grad_norm": 0.440668523311615, "learning_rate": 9.58391652966859e-05, "loss": 0.0512, "step": 13940 }, { "grad_norm": 0.468856543302536, "learning_rate": 9.583090674276583e-05, "loss": 0.0606, "step": 13950 }, { "grad_norm": 0.43069806694984436, "learning_rate": 9.582264035759726e-05, "loss": 0.0622, "step": 13960 }, { "grad_norm": 0.3783811926841736, "learning_rate": 9.58143661425927e-05, "loss": 0.0913, "step": 13970 }, { "grad_norm": 0.4874393343925476, "learning_rate": 9.580608409916601e-05, "loss": 0.0728, "step": 13980 }, { "grad_norm": 0.47942718863487244, "learning_rate": 9.579779422873233e-05, "loss": 0.065, "step": 13990 }, { "grad_norm": 0.5443252325057983, "learning_rate": 9.578949653270819e-05, "loss": 0.0731, "step": 14000 }, { "grad_norm": 0.45882648229599, "learning_rate": 9.578119101251144e-05, "loss": 0.0638, "step": 14010 }, { "grad_norm": 0.4568208158016205, "learning_rate": 9.577287766956127e-05, "loss": 0.0828, "step": 14020 }, { "grad_norm": 0.49484068155288696, "learning_rate": 9.57645565052782e-05, "loss": 0.0687, "step": 14030 }, { "grad_norm": 0.5494089126586914, "learning_rate": 9.575622752108407e-05, "loss": 0.0599, "step": 14040 }, { "grad_norm": 0.5492560267448425, "learning_rate": 9.57478907184021e-05, "loss": 0.0584, "step": 14050 }, { "grad_norm": 0.503540575504303, "learning_rate": 9.573954609865681e-05, "loss": 0.0748, "step": 14060 }, { "grad_norm": 0.6501151919364929, "learning_rate": 9.573119366327408e-05, "loss": 0.0594, "step": 14070 }, { "grad_norm": 0.38431984186172485, "learning_rate": 9.57228334136811e-05, "loss": 0.0613, "step": 14080 }, { "grad_norm": 0.5564183592796326, "learning_rate": 9.571446535130641e-05, "loss": 0.0558, "step": 14090 }, { "grad_norm": 0.32156920433044434, "learning_rate": 9.570608947757988e-05, "loss": 0.0701, "step": 14100 }, { "grad_norm": 0.38862282037734985, "learning_rate": 9.569770579393274e-05, "loss": 0.0667, "step": 14110 }, { "grad_norm": 0.7312856316566467, "learning_rate": 9.56893143017975e-05, "loss": 0.0638, "step": 14120 }, { "grad_norm": 0.4936133623123169, "learning_rate": 9.568091500260806e-05, "loss": 0.0678, "step": 14130 }, { "grad_norm": 0.5237100720405579, "learning_rate": 9.567250789779961e-05, "loss": 0.069, "step": 14140 }, { "grad_norm": 0.40217509865760803, "learning_rate": 9.566409298880872e-05, "loss": 0.0566, "step": 14150 }, { "grad_norm": 0.8234068155288696, "learning_rate": 9.565567027707326e-05, "loss": 0.0811, "step": 14160 }, { "grad_norm": 0.506203830242157, "learning_rate": 9.56472397640324e-05, "loss": 0.0571, "step": 14170 }, { "grad_norm": 0.4414336383342743, "learning_rate": 9.563880145112675e-05, "loss": 0.0704, "step": 14180 }, { "grad_norm": 0.5548738241195679, "learning_rate": 9.563035533979814e-05, "loss": 0.0726, "step": 14190 }, { "grad_norm": 0.5183058977127075, "learning_rate": 9.562190143148981e-05, "loss": 0.0709, "step": 14200 }, { "grad_norm": 0.4647201895713806, "learning_rate": 9.561343972764627e-05, "loss": 0.0704, "step": 14210 }, { "grad_norm": 0.44527921080589294, "learning_rate": 9.560497022971343e-05, "loss": 0.0689, "step": 14220 }, { "grad_norm": 0.4870621860027313, "learning_rate": 9.559649293913847e-05, "loss": 0.0646, "step": 14230 }, { "grad_norm": 0.5243958830833435, "learning_rate": 9.558800785736993e-05, "loss": 0.0652, "step": 14240 }, { "grad_norm": 0.534281313419342, "learning_rate": 9.557951498585767e-05, "loss": 0.067, "step": 14250 }, { "grad_norm": 0.4554073214530945, "learning_rate": 9.557101432605293e-05, "loss": 0.0644, "step": 14260 }, { "grad_norm": 0.4713858962059021, "learning_rate": 9.556250587940818e-05, "loss": 0.0734, "step": 14270 }, { "grad_norm": 0.8845525979995728, "learning_rate": 9.555398964737734e-05, "loss": 0.0626, "step": 14280 }, { "grad_norm": 0.505431592464447, "learning_rate": 9.554546563141555e-05, "loss": 0.0618, "step": 14290 }, { "grad_norm": 0.4310060143470764, "learning_rate": 9.553693383297937e-05, "loss": 0.058, "step": 14300 }, { "grad_norm": 0.6045358180999756, "learning_rate": 9.552839425352663e-05, "loss": 0.0566, "step": 14310 }, { "grad_norm": 0.45693957805633545, "learning_rate": 9.551984689451652e-05, "loss": 0.08, "step": 14320 }, { "grad_norm": 0.8559432625770569, "learning_rate": 9.551129175740953e-05, "loss": 0.066, "step": 14330 }, { "grad_norm": 0.36988648772239685, "learning_rate": 9.550272884366754e-05, "loss": 0.057, "step": 14340 }, { "grad_norm": 0.4478127956390381, "learning_rate": 9.549415815475369e-05, "loss": 0.0592, "step": 14350 }, { "grad_norm": 0.6516452431678772, "learning_rate": 9.548557969213247e-05, "loss": 0.0628, "step": 14360 }, { "grad_norm": 0.37117478251457214, "learning_rate": 9.547699345726972e-05, "loss": 0.0599, "step": 14370 }, { "grad_norm": 0.5034456253051758, "learning_rate": 9.546839945163257e-05, "loss": 0.0586, "step": 14380 }, { "grad_norm": 0.43702781200408936, "learning_rate": 9.545979767668953e-05, "loss": 0.0701, "step": 14390 }, { "grad_norm": 0.4708316922187805, "learning_rate": 9.54511881339104e-05, "loss": 0.0621, "step": 14400 }, { "grad_norm": 0.5233155488967896, "learning_rate": 9.54425708247663e-05, "loss": 0.0636, "step": 14410 }, { "grad_norm": 0.6377018094062805, "learning_rate": 9.543394575072972e-05, "loss": 0.0614, "step": 14420 }, { "grad_norm": 0.4290514588356018, "learning_rate": 9.542531291327441e-05, "loss": 0.0638, "step": 14430 }, { "grad_norm": 0.6209355592727661, "learning_rate": 9.541667231387552e-05, "loss": 0.0653, "step": 14440 }, { "grad_norm": 0.5398985743522644, "learning_rate": 9.540802395400949e-05, "loss": 0.0648, "step": 14450 }, { "grad_norm": 0.534273624420166, "learning_rate": 9.539936783515406e-05, "loss": 0.0635, "step": 14460 }, { "grad_norm": 0.4945164620876312, "learning_rate": 9.539070395878835e-05, "loss": 0.0591, "step": 14470 }, { "grad_norm": 0.6549351811408997, "learning_rate": 9.538203232639277e-05, "loss": 0.0673, "step": 14480 }, { "grad_norm": 0.6325492262840271, "learning_rate": 9.537335293944907e-05, "loss": 0.0648, "step": 14490 }, { "grad_norm": 0.500056266784668, "learning_rate": 9.536466579944032e-05, "loss": 0.0802, "step": 14500 }, { "grad_norm": 0.36255553364753723, "learning_rate": 9.535597090785091e-05, "loss": 0.0608, "step": 14510 }, { "grad_norm": 0.4792567491531372, "learning_rate": 9.534726826616656e-05, "loss": 0.0581, "step": 14520 }, { "grad_norm": 0.5876753926277161, "learning_rate": 9.53385578758743e-05, "loss": 0.0956, "step": 14530 }, { "grad_norm": 0.5563230514526367, "learning_rate": 9.532983973846252e-05, "loss": 0.0678, "step": 14540 }, { "grad_norm": 0.41720277070999146, "learning_rate": 9.53211138554209e-05, "loss": 0.059, "step": 14550 }, { "grad_norm": 0.5496581792831421, "learning_rate": 9.531238022824047e-05, "loss": 0.0628, "step": 14560 }, { "grad_norm": 0.41156327724456787, "learning_rate": 9.530363885841355e-05, "loss": 0.0478, "step": 14570 }, { "grad_norm": 0.5913909673690796, "learning_rate": 9.52948897474338e-05, "loss": 0.0747, "step": 14580 }, { "grad_norm": 0.4990389347076416, "learning_rate": 9.528613289679622e-05, "loss": 0.0679, "step": 14590 }, { "grad_norm": 0.38223302364349365, "learning_rate": 9.52773683079971e-05, "loss": 0.0637, "step": 14600 }, { "grad_norm": 0.35630741715431213, "learning_rate": 9.526859598253407e-05, "loss": 0.055, "step": 14610 }, { "grad_norm": 0.38886168599128723, "learning_rate": 9.525981592190609e-05, "loss": 0.0535, "step": 14620 }, { "grad_norm": 0.4956182539463043, "learning_rate": 9.525102812761342e-05, "loss": 0.0618, "step": 14630 }, { "grad_norm": 0.4368220865726471, "learning_rate": 9.524223260115768e-05, "loss": 0.0459, "step": 14640 }, { "grad_norm": 0.4585568606853485, "learning_rate": 9.523342934404175e-05, "loss": 0.0588, "step": 14650 }, { "grad_norm": 0.6686272025108337, "learning_rate": 9.522461835776989e-05, "loss": 0.0611, "step": 14660 }, { "grad_norm": 0.47571390867233276, "learning_rate": 9.521579964384764e-05, "loss": 0.0673, "step": 14670 }, { "grad_norm": 0.42145758867263794, "learning_rate": 9.52069732037819e-05, "loss": 0.0587, "step": 14680 }, { "grad_norm": 0.45893266797065735, "learning_rate": 9.519813903908083e-05, "loss": 0.0571, "step": 14690 }, { "grad_norm": 0.5235088467597961, "learning_rate": 9.5189297151254e-05, "loss": 0.0637, "step": 14700 }, { "grad_norm": 0.6791353821754456, "learning_rate": 9.518044754181218e-05, "loss": 0.0733, "step": 14710 }, { "grad_norm": 0.5365264415740967, "learning_rate": 9.51715902122676e-05, "loss": 0.0645, "step": 14720 }, { "grad_norm": 0.9561566114425659, "learning_rate": 9.516272516413368e-05, "loss": 0.0644, "step": 14730 }, { "grad_norm": 0.535889744758606, "learning_rate": 9.515385239892525e-05, "loss": 0.0683, "step": 14740 }, { "grad_norm": 0.4816092550754547, "learning_rate": 9.514497191815839e-05, "loss": 0.0775, "step": 14750 }, { "grad_norm": 1.232033371925354, "learning_rate": 9.513608372335055e-05, "loss": 0.0626, "step": 14760 }, { "grad_norm": 1.3213037252426147, "learning_rate": 9.512718781602045e-05, "loss": 0.0663, "step": 14770 }, { "grad_norm": 0.5130377411842346, "learning_rate": 9.511828419768823e-05, "loss": 0.064, "step": 14780 }, { "grad_norm": 0.42981627583503723, "learning_rate": 9.510937286987521e-05, "loss": 0.0605, "step": 14790 }, { "grad_norm": 0.4298648238182068, "learning_rate": 9.510045383410408e-05, "loss": 0.0584, "step": 14800 }, { "grad_norm": 0.9271475076675415, "learning_rate": 9.509152709189892e-05, "loss": 0.0653, "step": 14810 }, { "grad_norm": 0.5721539258956909, "learning_rate": 9.508259264478504e-05, "loss": 0.0668, "step": 14820 }, { "grad_norm": 0.578514814376831, "learning_rate": 9.507365049428909e-05, "loss": 0.0751, "step": 14830 }, { "grad_norm": 0.7645310163497925, "learning_rate": 9.506470064193902e-05, "loss": 0.091, "step": 14840 }, { "grad_norm": 0.4669584631919861, "learning_rate": 9.505574308926414e-05, "loss": 0.091, "step": 14850 }, { "grad_norm": 0.7486972808837891, "learning_rate": 9.504677783779505e-05, "loss": 0.0604, "step": 14860 }, { "grad_norm": 0.44874832034111023, "learning_rate": 9.503780488906365e-05, "loss": 0.0551, "step": 14870 }, { "grad_norm": 0.5353527069091797, "learning_rate": 9.502882424460319e-05, "loss": 0.0742, "step": 14880 }, { "grad_norm": 0.5548742413520813, "learning_rate": 9.501983590594821e-05, "loss": 0.0698, "step": 14890 }, { "grad_norm": 0.5447582006454468, "learning_rate": 9.501083987463455e-05, "loss": 0.0676, "step": 14900 }, { "grad_norm": 0.5114139318466187, "learning_rate": 9.500183615219942e-05, "loss": 0.0563, "step": 14910 }, { "grad_norm": 0.44137200713157654, "learning_rate": 9.49928247401813e-05, "loss": 0.0512, "step": 14920 }, { "grad_norm": 0.6095941662788391, "learning_rate": 9.498380564011997e-05, "loss": 0.0522, "step": 14930 }, { "grad_norm": 0.43905380368232727, "learning_rate": 9.497477885355656e-05, "loss": 0.0689, "step": 14940 }, { "grad_norm": 0.45565712451934814, "learning_rate": 9.496574438203353e-05, "loss": 0.0663, "step": 14950 }, { "grad_norm": 0.35161691904067993, "learning_rate": 9.495670222709459e-05, "loss": 0.0562, "step": 14960 }, { "grad_norm": 0.6681436896324158, "learning_rate": 9.494765239028483e-05, "loss": 0.0653, "step": 14970 }, { "grad_norm": 0.5271684527397156, "learning_rate": 9.493859487315057e-05, "loss": 0.0618, "step": 14980 }, { "grad_norm": 0.45285797119140625, "learning_rate": 9.492952967723953e-05, "loss": 0.0676, "step": 14990 }, { "grad_norm": 0.40828266739845276, "learning_rate": 9.492045680410068e-05, "loss": 0.0678, "step": 15000 }, { "grad_norm": 0.3978458642959595, "learning_rate": 9.491137625528436e-05, "loss": 0.0513, "step": 15010 }, { "grad_norm": 0.4016515016555786, "learning_rate": 9.490228803234215e-05, "loss": 0.054, "step": 15020 }, { "grad_norm": 0.5418629050254822, "learning_rate": 9.489319213682701e-05, "loss": 0.0564, "step": 15030 }, { "grad_norm": 0.44161468744277954, "learning_rate": 9.488408857029316e-05, "loss": 0.0668, "step": 15040 }, { "grad_norm": 0.42472729086875916, "learning_rate": 9.487497733429616e-05, "loss": 0.0763, "step": 15050 }, { "grad_norm": 0.6045131683349609, "learning_rate": 9.486585843039286e-05, "loss": 0.0712, "step": 15060 }, { "grad_norm": 0.36179399490356445, "learning_rate": 9.485673186014143e-05, "loss": 0.0579, "step": 15070 }, { "grad_norm": 0.4947940707206726, "learning_rate": 9.484759762510137e-05, "loss": 0.063, "step": 15080 }, { "grad_norm": 0.6672056913375854, "learning_rate": 9.483845572683346e-05, "loss": 0.0674, "step": 15090 }, { "grad_norm": 0.5287799835205078, "learning_rate": 9.48293061668998e-05, "loss": 0.0548, "step": 15100 }, { "grad_norm": 0.3850753903388977, "learning_rate": 9.48201489468638e-05, "loss": 0.0623, "step": 15110 }, { "grad_norm": 0.47106844186782837, "learning_rate": 9.481098406829016e-05, "loss": 0.0728, "step": 15120 }, { "grad_norm": 0.28838640451431274, "learning_rate": 9.480181153274495e-05, "loss": 0.0537, "step": 15130 }, { "grad_norm": 0.44383352994918823, "learning_rate": 9.479263134179548e-05, "loss": 0.0534, "step": 15140 }, { "grad_norm": 0.6556599736213684, "learning_rate": 9.478344349701039e-05, "loss": 0.071, "step": 15150 }, { "grad_norm": 0.43919771909713745, "learning_rate": 9.477424799995964e-05, "loss": 0.0699, "step": 15160 }, { "grad_norm": 0.3205508291721344, "learning_rate": 9.476504485221448e-05, "loss": 0.068, "step": 15170 }, { "grad_norm": 0.371995210647583, "learning_rate": 9.475583405534748e-05, "loss": 0.0575, "step": 15180 }, { "grad_norm": 0.49074995517730713, "learning_rate": 9.474661561093251e-05, "loss": 0.0621, "step": 15190 }, { "grad_norm": 0.5679596066474915, "learning_rate": 9.473738952054478e-05, "loss": 0.0697, "step": 15200 }, { "grad_norm": 0.48404252529144287, "learning_rate": 9.472815578576073e-05, "loss": 0.0577, "step": 15210 }, { "grad_norm": 0.6324400305747986, "learning_rate": 9.471891440815817e-05, "loss": 0.0632, "step": 15220 }, { "grad_norm": 0.4221877157688141, "learning_rate": 9.470966538931621e-05, "loss": 0.0534, "step": 15230 }, { "grad_norm": 0.6102300882339478, "learning_rate": 9.470040873081525e-05, "loss": 0.0605, "step": 15240 }, { "grad_norm": 0.5331792235374451, "learning_rate": 9.469114443423698e-05, "loss": 0.0829, "step": 15250 }, { "grad_norm": 0.5903159976005554, "learning_rate": 9.468187250116445e-05, "loss": 0.0607, "step": 15260 }, { "grad_norm": 0.35377737879753113, "learning_rate": 9.467259293318197e-05, "loss": 0.0566, "step": 15270 }, { "grad_norm": 0.4180314838886261, "learning_rate": 9.466330573187514e-05, "loss": 0.055, "step": 15280 }, { "grad_norm": 0.411441832780838, "learning_rate": 9.46540108988309e-05, "loss": 0.0669, "step": 15290 }, { "grad_norm": 0.5405778288841248, "learning_rate": 9.46447084356375e-05, "loss": 0.0666, "step": 15300 }, { "grad_norm": 0.3736920654773712, "learning_rate": 9.463539834388447e-05, "loss": 0.0667, "step": 15310 }, { "grad_norm": 0.7611107230186462, "learning_rate": 9.462608062516263e-05, "loss": 0.0558, "step": 15320 }, { "grad_norm": 0.4966685175895691, "learning_rate": 9.461675528106413e-05, "loss": 0.0649, "step": 15330 }, { "grad_norm": 0.5045683979988098, "learning_rate": 9.460742231318244e-05, "loss": 0.0556, "step": 15340 }, { "grad_norm": 0.43598026037216187, "learning_rate": 9.459808172311229e-05, "loss": 0.0656, "step": 15350 }, { "grad_norm": 0.5227437615394592, "learning_rate": 9.458873351244972e-05, "loss": 0.0569, "step": 15360 }, { "grad_norm": 0.45107153058052063, "learning_rate": 9.457937768279211e-05, "loss": 0.0599, "step": 15370 }, { "grad_norm": 0.4774399697780609, "learning_rate": 9.45700142357381e-05, "loss": 0.0624, "step": 15380 }, { "grad_norm": 0.5188800096511841, "learning_rate": 9.456064317288765e-05, "loss": 0.0557, "step": 15390 }, { "grad_norm": 0.5486508011817932, "learning_rate": 9.455126449584201e-05, "loss": 0.0612, "step": 15400 }, { "grad_norm": 0.4933741092681885, "learning_rate": 9.454187820620375e-05, "loss": 0.0534, "step": 15410 }, { "grad_norm": 0.41401153802871704, "learning_rate": 9.453248430557673e-05, "loss": 0.0619, "step": 15420 }, { "grad_norm": 0.490640252828598, "learning_rate": 9.452308279556611e-05, "loss": 0.0828, "step": 15430 }, { "grad_norm": 0.29529353976249695, "learning_rate": 9.451367367777835e-05, "loss": 0.0529, "step": 15440 }, { "grad_norm": 0.36909914016723633, "learning_rate": 9.450425695382122e-05, "loss": 0.0633, "step": 15450 }, { "grad_norm": 0.5456652045249939, "learning_rate": 9.449483262530375e-05, "loss": 0.0579, "step": 15460 }, { "grad_norm": 0.5447001457214355, "learning_rate": 9.448540069383633e-05, "loss": 0.0511, "step": 15470 }, { "grad_norm": 0.4658367335796356, "learning_rate": 9.447596116103061e-05, "loss": 0.0581, "step": 15480 }, { "grad_norm": 0.48959746956825256, "learning_rate": 9.446651402849955e-05, "loss": 0.0682, "step": 15490 }, { "grad_norm": 0.42748168110847473, "learning_rate": 9.44570592978574e-05, "loss": 0.0593, "step": 15500 }, { "grad_norm": 0.6854232549667358, "learning_rate": 9.444759697071972e-05, "loss": 0.0675, "step": 15510 }, { "grad_norm": 0.6434798240661621, "learning_rate": 9.443812704870336e-05, "loss": 0.0618, "step": 15520 }, { "grad_norm": 0.4066343605518341, "learning_rate": 9.442864953342649e-05, "loss": 0.063, "step": 15530 }, { "grad_norm": 0.6183857917785645, "learning_rate": 9.441916442650852e-05, "loss": 0.0518, "step": 15540 }, { "grad_norm": 0.5842982530593872, "learning_rate": 9.440967172957023e-05, "loss": 0.0574, "step": 15550 }, { "grad_norm": 0.47456443309783936, "learning_rate": 9.440017144423364e-05, "loss": 0.0642, "step": 15560 }, { "grad_norm": 0.5337789058685303, "learning_rate": 9.439066357212209e-05, "loss": 0.0584, "step": 15570 }, { "grad_norm": 0.4537689685821533, "learning_rate": 9.438114811486022e-05, "loss": 0.0613, "step": 15580 }, { "grad_norm": 0.5226091146469116, "learning_rate": 9.4371625074074e-05, "loss": 0.055, "step": 15590 }, { "grad_norm": 0.502898097038269, "learning_rate": 9.436209445139059e-05, "loss": 0.0508, "step": 15600 }, { "grad_norm": 0.45622915029525757, "learning_rate": 9.435255624843855e-05, "loss": 0.0579, "step": 15610 }, { "grad_norm": 0.5169779658317566, "learning_rate": 9.43430104668477e-05, "loss": 0.0466, "step": 15620 }, { "grad_norm": 0.43501096963882446, "learning_rate": 9.433345710824914e-05, "loss": 0.0527, "step": 15630 }, { "grad_norm": 0.37742623686790466, "learning_rate": 9.432389617427529e-05, "loss": 0.0494, "step": 15640 }, { "grad_norm": 0.39843830466270447, "learning_rate": 9.431432766655984e-05, "loss": 0.0606, "step": 15650 }, { "grad_norm": 0.5127043128013611, "learning_rate": 9.430475158673778e-05, "loss": 0.0561, "step": 15660 }, { "grad_norm": 0.6274949908256531, "learning_rate": 9.429516793644542e-05, "loss": 0.0582, "step": 15670 }, { "grad_norm": 0.5978347659111023, "learning_rate": 9.428557671732034e-05, "loss": 0.0591, "step": 15680 }, { "grad_norm": 0.5344213247299194, "learning_rate": 9.42759779310014e-05, "loss": 0.0618, "step": 15690 }, { "grad_norm": 0.4780386686325073, "learning_rate": 9.426637157912879e-05, "loss": 0.0537, "step": 15700 }, { "grad_norm": 0.536502480506897, "learning_rate": 9.425675766334397e-05, "loss": 0.0685, "step": 15710 }, { "grad_norm": 0.8064213395118713, "learning_rate": 9.424713618528968e-05, "loss": 0.0553, "step": 15720 }, { "grad_norm": 0.7883898615837097, "learning_rate": 9.423750714661e-05, "loss": 0.0669, "step": 15730 }, { "grad_norm": 0.4410579204559326, "learning_rate": 9.422787054895022e-05, "loss": 0.0668, "step": 15740 }, { "grad_norm": 0.6532207727432251, "learning_rate": 9.4218226393957e-05, "loss": 0.0701, "step": 15750 }, { "grad_norm": 0.4038883149623871, "learning_rate": 9.420857468327828e-05, "loss": 0.0661, "step": 15760 }, { "grad_norm": 0.49777930974960327, "learning_rate": 9.419891541856323e-05, "loss": 0.0534, "step": 15770 }, { "grad_norm": 0.4413801431655884, "learning_rate": 9.41892486014624e-05, "loss": 0.0607, "step": 15780 }, { "grad_norm": 0.44116634130477905, "learning_rate": 9.417957423362756e-05, "loss": 0.0728, "step": 15790 }, { "grad_norm": 0.3364131450653076, "learning_rate": 9.416989231671178e-05, "loss": 0.0585, "step": 15800 }, { "grad_norm": 0.34564119577407837, "learning_rate": 9.416020285236946e-05, "loss": 0.0569, "step": 15810 }, { "grad_norm": 0.45870187878608704, "learning_rate": 9.415050584225626e-05, "loss": 0.0579, "step": 15820 }, { "grad_norm": 0.5503543019294739, "learning_rate": 9.414080128802914e-05, "loss": 0.0662, "step": 15830 }, { "grad_norm": 0.6241634488105774, "learning_rate": 9.413108919134632e-05, "loss": 0.0564, "step": 15840 }, { "grad_norm": 0.4841318130493164, "learning_rate": 9.412136955386734e-05, "loss": 0.0632, "step": 15850 }, { "grad_norm": 0.6539974212646484, "learning_rate": 9.411164237725303e-05, "loss": 0.0574, "step": 15860 }, { "grad_norm": 0.5356277823448181, "learning_rate": 9.41019076631655e-05, "loss": 0.0756, "step": 15870 }, { "grad_norm": 0.7462782859802246, "learning_rate": 9.409216541326815e-05, "loss": 0.0652, "step": 15880 }, { "grad_norm": 0.5635814070701599, "learning_rate": 9.408241562922564e-05, "loss": 0.0627, "step": 15890 }, { "grad_norm": 0.5289958715438843, "learning_rate": 9.407265831270395e-05, "loss": 0.0588, "step": 15900 }, { "grad_norm": 0.6142227053642273, "learning_rate": 9.406289346537035e-05, "loss": 0.0668, "step": 15910 }, { "grad_norm": 0.41264021396636963, "learning_rate": 9.405312108889339e-05, "loss": 0.0577, "step": 15920 }, { "grad_norm": 0.5812535881996155, "learning_rate": 9.404334118494288e-05, "loss": 0.061, "step": 15930 }, { "grad_norm": 0.3973679542541504, "learning_rate": 9.403355375518995e-05, "loss": 0.0617, "step": 15940 }, { "grad_norm": 0.3993660509586334, "learning_rate": 9.4023758801307e-05, "loss": 0.0512, "step": 15950 }, { "grad_norm": 0.4613216817378998, "learning_rate": 9.401395632496774e-05, "loss": 0.0503, "step": 15960 }, { "grad_norm": 0.627045214176178, "learning_rate": 9.400414632784711e-05, "loss": 0.0701, "step": 15970 }, { "grad_norm": 0.580726146697998, "learning_rate": 9.39943288116214e-05, "loss": 0.0625, "step": 15980 }, { "grad_norm": 0.4937728941440582, "learning_rate": 9.398450377796815e-05, "loss": 0.064, "step": 15990 }, { "grad_norm": 0.8642576932907104, "learning_rate": 9.397467122856616e-05, "loss": 0.065, "step": 16000 }, { "grad_norm": 0.6824963092803955, "learning_rate": 9.396483116509558e-05, "loss": 0.0662, "step": 16010 }, { "grad_norm": 0.4906345009803772, "learning_rate": 9.39549835892378e-05, "loss": 0.0684, "step": 16020 }, { "grad_norm": 0.4228990375995636, "learning_rate": 9.39451285026755e-05, "loss": 0.0658, "step": 16030 }, { "grad_norm": 0.3947159945964813, "learning_rate": 9.393526590709262e-05, "loss": 0.0561, "step": 16040 }, { "grad_norm": 0.46829524636268616, "learning_rate": 9.392539580417444e-05, "loss": 0.054, "step": 16050 }, { "grad_norm": 0.4135490655899048, "learning_rate": 9.391551819560747e-05, "loss": 0.0627, "step": 16060 }, { "grad_norm": 0.6555302739143372, "learning_rate": 9.390563308307955e-05, "loss": 0.0645, "step": 16070 }, { "grad_norm": 0.39074793457984924, "learning_rate": 9.389574046827974e-05, "loss": 0.0573, "step": 16080 }, { "grad_norm": 0.568091630935669, "learning_rate": 9.388584035289845e-05, "loss": 0.0555, "step": 16090 }, { "grad_norm": 0.4803045392036438, "learning_rate": 9.387593273862732e-05, "loss": 0.0703, "step": 16100 }, { "grad_norm": 0.5392587184906006, "learning_rate": 9.386601762715929e-05, "loss": 0.0651, "step": 16110 }, { "grad_norm": 0.5229243040084839, "learning_rate": 9.38560950201886e-05, "loss": 0.065, "step": 16120 }, { "grad_norm": 0.7583493590354919, "learning_rate": 9.384616491941071e-05, "loss": 0.0722, "step": 16130 }, { "grad_norm": 0.43107208609580994, "learning_rate": 9.383622732652245e-05, "loss": 0.0647, "step": 16140 }, { "grad_norm": 0.5449812412261963, "learning_rate": 9.382628224322187e-05, "loss": 0.0716, "step": 16150 }, { "grad_norm": 0.6697376370429993, "learning_rate": 9.381632967120829e-05, "loss": 0.0634, "step": 16160 }, { "grad_norm": 0.3961394131183624, "learning_rate": 9.380636961218235e-05, "loss": 0.0612, "step": 16170 }, { "grad_norm": 0.5658188462257385, "learning_rate": 9.379640206784597e-05, "loss": 0.0457, "step": 16180 }, { "grad_norm": 0.47548040747642517, "learning_rate": 9.378642703990229e-05, "loss": 0.0565, "step": 16190 }, { "grad_norm": 0.6361516714096069, "learning_rate": 9.37764445300558e-05, "loss": 0.0636, "step": 16200 }, { "grad_norm": 0.3843062222003937, "learning_rate": 9.376645454001222e-05, "loss": 0.0673, "step": 16210 }, { "grad_norm": 0.7096383571624756, "learning_rate": 9.375645707147858e-05, "loss": 0.0591, "step": 16220 }, { "grad_norm": 0.5172668695449829, "learning_rate": 9.374645212616316e-05, "loss": 0.0717, "step": 16230 }, { "grad_norm": 0.49928539991378784, "learning_rate": 9.373643970577555e-05, "loss": 0.0681, "step": 16240 }, { "grad_norm": 0.5777023434638977, "learning_rate": 9.372641981202659e-05, "loss": 0.0576, "step": 16250 }, { "grad_norm": 0.5659461617469788, "learning_rate": 9.37163924466284e-05, "loss": 0.0619, "step": 16260 }, { "grad_norm": 0.5906530022621155, "learning_rate": 9.370635761129438e-05, "loss": 0.0562, "step": 16270 }, { "grad_norm": 0.3643147051334381, "learning_rate": 9.36963153077392e-05, "loss": 0.0563, "step": 16280 }, { "grad_norm": 0.38357117772102356, "learning_rate": 9.368626553767888e-05, "loss": 0.0505, "step": 16290 }, { "grad_norm": 0.515474259853363, "learning_rate": 9.367620830283057e-05, "loss": 0.0521, "step": 16300 }, { "grad_norm": 0.5349491834640503, "learning_rate": 9.366614360491281e-05, "loss": 0.0631, "step": 16310 }, { "grad_norm": 0.4280451834201813, "learning_rate": 9.365607144564539e-05, "loss": 0.0612, "step": 16320 }, { "grad_norm": 0.42014607787132263, "learning_rate": 9.364599182674934e-05, "loss": 0.0639, "step": 16330 }, { "grad_norm": 0.42594870924949646, "learning_rate": 9.3635904749947e-05, "loss": 0.0518, "step": 16340 }, { "grad_norm": 0.3297095000743866, "learning_rate": 9.362581021696202e-05, "loss": 0.0437, "step": 16350 }, { "grad_norm": 0.3405816853046417, "learning_rate": 9.361570822951921e-05, "loss": 0.0602, "step": 16360 }, { "grad_norm": 0.6311708688735962, "learning_rate": 9.360559878934476e-05, "loss": 0.061, "step": 16370 }, { "grad_norm": 0.4021540582180023, "learning_rate": 9.359548189816611e-05, "loss": 0.061, "step": 16380 }, { "grad_norm": 0.5378772020339966, "learning_rate": 9.358535755771193e-05, "loss": 0.0654, "step": 16390 }, { "grad_norm": 0.5398531556129456, "learning_rate": 9.357522576971221e-05, "loss": 0.0562, "step": 16400 }, { "grad_norm": 0.47402670979499817, "learning_rate": 9.356508653589819e-05, "loss": 0.0621, "step": 16410 }, { "grad_norm": 0.41486623883247375, "learning_rate": 9.355493985800237e-05, "loss": 0.0543, "step": 16420 }, { "grad_norm": 0.4595799148082733, "learning_rate": 9.354478573775857e-05, "loss": 0.0646, "step": 16430 }, { "grad_norm": 0.6349772810935974, "learning_rate": 9.353462417690186e-05, "loss": 0.0584, "step": 16440 }, { "grad_norm": 0.4583083689212799, "learning_rate": 9.352445517716853e-05, "loss": 0.0577, "step": 16450 }, { "grad_norm": 0.3866024613380432, "learning_rate": 9.351427874029621e-05, "loss": 0.0643, "step": 16460 }, { "grad_norm": 0.30885612964630127, "learning_rate": 9.350409486802379e-05, "loss": 0.0529, "step": 16470 }, { "grad_norm": 0.5719230771064758, "learning_rate": 9.349390356209138e-05, "loss": 0.0564, "step": 16480 }, { "grad_norm": 0.6188654899597168, "learning_rate": 9.348370482424042e-05, "loss": 0.062, "step": 16490 }, { "grad_norm": 0.6611849665641785, "learning_rate": 9.347349865621357e-05, "loss": 0.057, "step": 16500 }, { "grad_norm": 0.42797955870628357, "learning_rate": 9.346328505975481e-05, "loss": 0.0618, "step": 16510 }, { "grad_norm": 0.3336006999015808, "learning_rate": 9.345306403660936e-05, "loss": 0.0604, "step": 16520 }, { "grad_norm": 0.5069711804389954, "learning_rate": 9.344283558852371e-05, "loss": 0.061, "step": 16530 }, { "grad_norm": 0.49051910638809204, "learning_rate": 9.343259971724563e-05, "loss": 0.0515, "step": 16540 }, { "grad_norm": 0.47399359941482544, "learning_rate": 9.342235642452413e-05, "loss": 0.0643, "step": 16550 }, { "grad_norm": 0.6156262755393982, "learning_rate": 9.341210571210954e-05, "loss": 0.0652, "step": 16560 }, { "grad_norm": 0.452304482460022, "learning_rate": 9.340184758175338e-05, "loss": 0.0682, "step": 16570 }, { "grad_norm": 0.6064088940620422, "learning_rate": 9.339158203520854e-05, "loss": 0.0707, "step": 16580 }, { "grad_norm": 0.3047676682472229, "learning_rate": 9.338130907422908e-05, "loss": 0.0511, "step": 16590 }, { "grad_norm": 0.4632011950016022, "learning_rate": 9.337102870057037e-05, "loss": 0.0562, "step": 16600 }, { "grad_norm": 0.4036937355995178, "learning_rate": 9.336074091598907e-05, "loss": 0.0523, "step": 16610 }, { "grad_norm": 0.629181444644928, "learning_rate": 9.335044572224306e-05, "loss": 0.062, "step": 16620 }, { "grad_norm": 0.46297571063041687, "learning_rate": 9.334014312109151e-05, "loss": 0.0501, "step": 16630 }, { "grad_norm": 0.5387881398200989, "learning_rate": 9.332983311429486e-05, "loss": 0.0594, "step": 16640 }, { "grad_norm": 0.5094613432884216, "learning_rate": 9.33195157036148e-05, "loss": 0.0558, "step": 16650 }, { "grad_norm": 0.4473305642604828, "learning_rate": 9.330919089081432e-05, "loss": 0.0538, "step": 16660 }, { "grad_norm": 0.512860119342804, "learning_rate": 9.32988586776576e-05, "loss": 0.0743, "step": 16670 }, { "grad_norm": 0.7052309513092041, "learning_rate": 9.328851906591016e-05, "loss": 0.0553, "step": 16680 }, { "grad_norm": 0.5793055295944214, "learning_rate": 9.327817205733875e-05, "loss": 0.0662, "step": 16690 }, { "grad_norm": 0.3619213104248047, "learning_rate": 9.326781765371142e-05, "loss": 0.0598, "step": 16700 }, { "grad_norm": 0.6340780854225159, "learning_rate": 9.325745585679741e-05, "loss": 0.078, "step": 16710 }, { "grad_norm": 0.4656042754650116, "learning_rate": 9.32470866683673e-05, "loss": 0.0711, "step": 16720 }, { "grad_norm": 0.3797457814216614, "learning_rate": 9.323671009019288e-05, "loss": 0.071, "step": 16730 }, { "grad_norm": 0.339988648891449, "learning_rate": 9.322632612404725e-05, "loss": 0.0507, "step": 16740 }, { "grad_norm": 0.42239001393318176, "learning_rate": 9.321593477170471e-05, "loss": 0.0719, "step": 16750 }, { "grad_norm": 0.5138610601425171, "learning_rate": 9.320553603494088e-05, "loss": 0.0493, "step": 16760 }, { "grad_norm": 0.4188835024833679, "learning_rate": 9.319512991553261e-05, "loss": 0.0517, "step": 16770 }, { "grad_norm": 0.462832510471344, "learning_rate": 9.318471641525803e-05, "loss": 0.0549, "step": 16780 }, { "grad_norm": 0.33343085646629333, "learning_rate": 9.317429553589652e-05, "loss": 0.0468, "step": 16790 }, { "grad_norm": 0.46078795194625854, "learning_rate": 9.316386727922873e-05, "loss": 0.0583, "step": 16800 }, { "grad_norm": 0.5144149661064148, "learning_rate": 9.315343164703656e-05, "loss": 0.0582, "step": 16810 }, { "grad_norm": 0.41310662031173706, "learning_rate": 9.314298864110316e-05, "loss": 0.0574, "step": 16820 }, { "grad_norm": 0.34159746766090393, "learning_rate": 9.313253826321295e-05, "loss": 0.0466, "step": 16830 }, { "grad_norm": 0.6216355562210083, "learning_rate": 9.312208051515165e-05, "loss": 0.0556, "step": 16840 }, { "grad_norm": 0.3784632384777069, "learning_rate": 9.311161539870618e-05, "loss": 0.0543, "step": 16850 }, { "grad_norm": 0.5544754862785339, "learning_rate": 9.310114291566474e-05, "loss": 0.0688, "step": 16860 }, { "grad_norm": 0.41540518403053284, "learning_rate": 9.309066306781679e-05, "loss": 0.051, "step": 16870 }, { "grad_norm": 0.6579543352127075, "learning_rate": 9.308017585695306e-05, "loss": 0.0467, "step": 16880 }, { "grad_norm": 0.40091004967689514, "learning_rate": 9.306968128486552e-05, "loss": 0.0468, "step": 16890 }, { "grad_norm": 0.4718818664550781, "learning_rate": 9.30591793533474e-05, "loss": 0.0493, "step": 16900 }, { "grad_norm": 0.42586636543273926, "learning_rate": 9.304867006419321e-05, "loss": 0.0481, "step": 16910 }, { "grad_norm": 0.5187777280807495, "learning_rate": 9.303815341919868e-05, "loss": 0.0565, "step": 16920 }, { "grad_norm": 0.4328530728816986, "learning_rate": 9.302762942016084e-05, "loss": 0.0529, "step": 16930 }, { "grad_norm": 0.4306727349758148, "learning_rate": 9.301709806887792e-05, "loss": 0.046, "step": 16940 }, { "grad_norm": 0.3721405267715454, "learning_rate": 9.300655936714948e-05, "loss": 0.07, "step": 16950 }, { "grad_norm": 0.50131756067276, "learning_rate": 9.299601331677627e-05, "loss": 0.0486, "step": 16960 }, { "grad_norm": 0.5278552770614624, "learning_rate": 9.298545991956033e-05, "loss": 0.0622, "step": 16970 }, { "grad_norm": 0.3383072316646576, "learning_rate": 9.297489917730493e-05, "loss": 0.0477, "step": 16980 }, { "grad_norm": 0.34937581419944763, "learning_rate": 9.296433109181464e-05, "loss": 0.052, "step": 16990 }, { "grad_norm": 0.5364123582839966, "learning_rate": 9.295375566489523e-05, "loss": 0.054, "step": 17000 }, { "grad_norm": 0.5287961959838867, "learning_rate": 9.294317289835379e-05, "loss": 0.0659, "step": 17010 }, { "grad_norm": 0.43123912811279297, "learning_rate": 9.293258279399859e-05, "loss": 0.0617, "step": 17020 }, { "grad_norm": 0.5507157444953918, "learning_rate": 9.292198535363919e-05, "loss": 0.0523, "step": 17030 }, { "grad_norm": 0.49493956565856934, "learning_rate": 9.291138057908641e-05, "loss": 0.0515, "step": 17040 }, { "grad_norm": 0.5638753175735474, "learning_rate": 9.290076847215234e-05, "loss": 0.0567, "step": 17050 }, { "grad_norm": 0.5748199820518494, "learning_rate": 9.289014903465025e-05, "loss": 0.0754, "step": 17060 }, { "grad_norm": 0.4144059121608734, "learning_rate": 9.287952226839475e-05, "loss": 0.0569, "step": 17070 }, { "grad_norm": 0.43893948197364807, "learning_rate": 9.286888817520164e-05, "loss": 0.0536, "step": 17080 }, { "grad_norm": 0.6842093467712402, "learning_rate": 9.285824675688803e-05, "loss": 0.0537, "step": 17090 }, { "grad_norm": 0.3208901882171631, "learning_rate": 9.28475980152722e-05, "loss": 0.0529, "step": 17100 }, { "grad_norm": 0.44980865716934204, "learning_rate": 9.283694195217379e-05, "loss": 0.0672, "step": 17110 }, { "grad_norm": 0.569999098777771, "learning_rate": 9.282627856941356e-05, "loss": 0.0607, "step": 17120 }, { "grad_norm": 0.6757642030715942, "learning_rate": 9.281560786881363e-05, "loss": 0.057, "step": 17130 }, { "grad_norm": 0.639371395111084, "learning_rate": 9.280492985219733e-05, "loss": 0.0455, "step": 17140 }, { "grad_norm": 0.4570300877094269, "learning_rate": 9.279424452138924e-05, "loss": 0.0567, "step": 17150 }, { "grad_norm": 0.31996583938598633, "learning_rate": 9.278355187821517e-05, "loss": 0.0508, "step": 17160 }, { "grad_norm": 0.7199746966362, "learning_rate": 9.277285192450224e-05, "loss": 0.0569, "step": 17170 }, { "grad_norm": 0.7367189526557922, "learning_rate": 9.276214466207875e-05, "loss": 0.0607, "step": 17180 }, { "grad_norm": 0.514271080493927, "learning_rate": 9.275143009277427e-05, "loss": 0.0532, "step": 17190 }, { "grad_norm": 0.5348409414291382, "learning_rate": 9.274070821841964e-05, "loss": 0.0618, "step": 17200 }, { "grad_norm": 0.5853487253189087, "learning_rate": 9.272997904084696e-05, "loss": 0.0498, "step": 17210 }, { "grad_norm": 0.48864951729774475, "learning_rate": 9.271924256188951e-05, "loss": 0.0582, "step": 17220 }, { "grad_norm": 0.6162366271018982, "learning_rate": 9.270849878338189e-05, "loss": 0.057, "step": 17230 }, { "grad_norm": 0.4534902274608612, "learning_rate": 9.269774770715991e-05, "loss": 0.0549, "step": 17240 }, { "grad_norm": 0.3176596462726593, "learning_rate": 9.268698933506061e-05, "loss": 0.0612, "step": 17250 }, { "grad_norm": 0.4476422667503357, "learning_rate": 9.267622366892235e-05, "loss": 0.0526, "step": 17260 }, { "grad_norm": 0.5011503100395203, "learning_rate": 9.266545071058465e-05, "loss": 0.0583, "step": 17270 }, { "grad_norm": 0.5395255088806152, "learning_rate": 9.265467046188833e-05, "loss": 0.0717, "step": 17280 }, { "grad_norm": 0.3486761152744293, "learning_rate": 9.264388292467543e-05, "loss": 0.0586, "step": 17290 }, { "grad_norm": 0.40753135085105896, "learning_rate": 9.263308810078926e-05, "loss": 0.0544, "step": 17300 }, { "grad_norm": 0.3303479552268982, "learning_rate": 9.262228599207434e-05, "loss": 0.055, "step": 17310 }, { "grad_norm": 0.3683425784111023, "learning_rate": 9.261147660037647e-05, "loss": 0.0447, "step": 17320 }, { "grad_norm": 0.3934553861618042, "learning_rate": 9.26006599275427e-05, "loss": 0.0469, "step": 17330 }, { "grad_norm": 0.4824930727481842, "learning_rate": 9.258983597542124e-05, "loss": 0.0549, "step": 17340 }, { "grad_norm": 0.48395732045173645, "learning_rate": 9.257900474586167e-05, "loss": 0.0549, "step": 17350 }, { "grad_norm": 0.572215735912323, "learning_rate": 9.256816624071471e-05, "loss": 0.0628, "step": 17360 }, { "grad_norm": 0.5703082084655762, "learning_rate": 9.25573204618324e-05, "loss": 0.0566, "step": 17370 }, { "grad_norm": 0.6342107057571411, "learning_rate": 9.254646741106796e-05, "loss": 0.0553, "step": 17380 }, { "grad_norm": 0.5369377732276917, "learning_rate": 9.253560709027589e-05, "loss": 0.0573, "step": 17390 }, { "grad_norm": 0.4355736970901489, "learning_rate": 9.252473950131192e-05, "loss": 0.0588, "step": 17400 }, { "grad_norm": 0.42585551738739014, "learning_rate": 9.251386464603302e-05, "loss": 0.0483, "step": 17410 }, { "grad_norm": 0.43737295269966125, "learning_rate": 9.250298252629741e-05, "loss": 0.0521, "step": 17420 }, { "grad_norm": 0.37121301889419556, "learning_rate": 9.249209314396454e-05, "loss": 0.0458, "step": 17430 }, { "grad_norm": 0.5380534529685974, "learning_rate": 9.248119650089513e-05, "loss": 0.0595, "step": 17440 }, { "grad_norm": 0.5611270666122437, "learning_rate": 9.247029259895108e-05, "loss": 0.0594, "step": 17450 }, { "grad_norm": 0.34576866030693054, "learning_rate": 9.24593814399956e-05, "loss": 0.0603, "step": 17460 }, { "grad_norm": 0.5429890751838684, "learning_rate": 9.244846302589309e-05, "loss": 0.044, "step": 17470 }, { "grad_norm": 0.5134668946266174, "learning_rate": 9.243753735850923e-05, "loss": 0.057, "step": 17480 }, { "grad_norm": 0.33993253111839294, "learning_rate": 9.24266044397109e-05, "loss": 0.0497, "step": 17490 }, { "grad_norm": 1.0154528617858887, "learning_rate": 9.241566427136624e-05, "loss": 0.0456, "step": 17500 }, { "grad_norm": 0.45038822293281555, "learning_rate": 9.240471685534463e-05, "loss": 0.0546, "step": 17510 }, { "grad_norm": 0.3606150448322296, "learning_rate": 9.239376219351667e-05, "loss": 0.0534, "step": 17520 }, { "grad_norm": 0.702984094619751, "learning_rate": 9.238280028775425e-05, "loss": 0.0658, "step": 17530 }, { "grad_norm": 0.7813571691513062, "learning_rate": 9.237183113993041e-05, "loss": 0.0688, "step": 17540 }, { "grad_norm": 0.46909618377685547, "learning_rate": 9.236085475191952e-05, "loss": 0.0525, "step": 17550 }, { "grad_norm": 0.6490064263343811, "learning_rate": 9.234987112559709e-05, "loss": 0.0705, "step": 17560 }, { "grad_norm": 0.41877856850624084, "learning_rate": 9.233888026283999e-05, "loss": 0.0633, "step": 17570 }, { "grad_norm": 0.42870137095451355, "learning_rate": 9.232788216552619e-05, "loss": 0.0509, "step": 17580 }, { "grad_norm": 0.4248207211494446, "learning_rate": 9.231687683553502e-05, "loss": 0.0581, "step": 17590 }, { "grad_norm": 0.6037894487380981, "learning_rate": 9.230586427474698e-05, "loss": 0.0621, "step": 17600 }, { "grad_norm": 0.45263081789016724, "learning_rate": 9.229484448504379e-05, "loss": 0.0518, "step": 17610 }, { "grad_norm": 0.6122071743011475, "learning_rate": 9.228381746830843e-05, "loss": 0.0624, "step": 17620 }, { "grad_norm": 0.4819363057613373, "learning_rate": 9.227278322642514e-05, "loss": 0.0525, "step": 17630 }, { "grad_norm": 0.8440017700195312, "learning_rate": 9.226174176127937e-05, "loss": 0.0618, "step": 17640 }, { "grad_norm": 0.5981501340866089, "learning_rate": 9.22506930747578e-05, "loss": 0.063, "step": 17650 }, { "grad_norm": 0.6502760648727417, "learning_rate": 9.223963716874831e-05, "loss": 0.0668, "step": 17660 }, { "grad_norm": 0.5947176218032837, "learning_rate": 9.222857404514012e-05, "loss": 0.0665, "step": 17670 }, { "grad_norm": 0.5424709320068359, "learning_rate": 9.221750370582355e-05, "loss": 0.0581, "step": 17680 }, { "grad_norm": 0.3807879388332367, "learning_rate": 9.220642615269028e-05, "loss": 0.05, "step": 17690 }, { "grad_norm": 0.6957297325134277, "learning_rate": 9.219534138763311e-05, "loss": 0.0565, "step": 17700 }, { "grad_norm": 0.6390944719314575, "learning_rate": 9.218424941254613e-05, "loss": 0.0504, "step": 17710 }, { "grad_norm": 0.5313125252723694, "learning_rate": 9.217315022932468e-05, "loss": 0.0552, "step": 17720 }, { "grad_norm": 0.3813105821609497, "learning_rate": 9.216204383986528e-05, "loss": 0.0567, "step": 17730 }, { "grad_norm": 0.5153471231460571, "learning_rate": 9.215093024606574e-05, "loss": 0.0659, "step": 17740 }, { "grad_norm": 0.8781358003616333, "learning_rate": 9.213980944982506e-05, "loss": 0.073, "step": 17750 }, { "grad_norm": 0.39447733759880066, "learning_rate": 9.212868145304346e-05, "loss": 0.0694, "step": 17760 }, { "grad_norm": 0.5191782712936401, "learning_rate": 9.211754625762241e-05, "loss": 0.0523, "step": 17770 }, { "grad_norm": 0.5338747501373291, "learning_rate": 9.210640386546463e-05, "loss": 0.0704, "step": 17780 }, { "grad_norm": 0.6848440170288086, "learning_rate": 9.209525427847405e-05, "loss": 0.0544, "step": 17790 }, { "grad_norm": 1.2469470500946045, "learning_rate": 9.208409749855583e-05, "loss": 0.0523, "step": 17800 }, { "grad_norm": 0.32472777366638184, "learning_rate": 9.207293352761633e-05, "loss": 0.067, "step": 17810 }, { "grad_norm": 0.7426021099090576, "learning_rate": 9.206176236756319e-05, "loss": 0.0564, "step": 17820 }, { "grad_norm": 0.6659724116325378, "learning_rate": 9.205058402030525e-05, "loss": 0.0562, "step": 17830 }, { "grad_norm": 0.6826874613761902, "learning_rate": 9.203939848775259e-05, "loss": 0.0574, "step": 17840 }, { "grad_norm": 0.502433180809021, "learning_rate": 9.202820577181652e-05, "loss": 0.0552, "step": 17850 }, { "grad_norm": 0.440198689699173, "learning_rate": 9.201700587440953e-05, "loss": 0.0565, "step": 17860 }, { "grad_norm": 0.5933297276496887, "learning_rate": 9.200579879744544e-05, "loss": 0.063, "step": 17870 }, { "grad_norm": 0.5643934011459351, "learning_rate": 9.199458454283918e-05, "loss": 0.0513, "step": 17880 }, { "grad_norm": 0.566289484500885, "learning_rate": 9.198336311250697e-05, "loss": 0.0526, "step": 17890 }, { "grad_norm": 0.5867999196052551, "learning_rate": 9.197213450836626e-05, "loss": 0.0474, "step": 17900 }, { "grad_norm": 0.7316045761108398, "learning_rate": 9.19608987323357e-05, "loss": 0.0587, "step": 17910 }, { "grad_norm": 0.4739600718021393, "learning_rate": 9.194965578633517e-05, "loss": 0.0446, "step": 17920 }, { "grad_norm": 0.4226260781288147, "learning_rate": 9.193840567228582e-05, "loss": 0.0585, "step": 17930 }, { "grad_norm": 0.6452944874763489, "learning_rate": 9.192714839210994e-05, "loss": 0.0561, "step": 17940 }, { "grad_norm": 0.430397093296051, "learning_rate": 9.19158839477311e-05, "loss": 0.0646, "step": 17950 }, { "grad_norm": 0.49348223209381104, "learning_rate": 9.190461234107411e-05, "loss": 0.0593, "step": 17960 }, { "grad_norm": 0.45267853140830994, "learning_rate": 9.189333357406496e-05, "loss": 0.0487, "step": 17970 }, { "grad_norm": 0.5228747725486755, "learning_rate": 9.188204764863089e-05, "loss": 0.0641, "step": 17980 }, { "grad_norm": 0.5348548889160156, "learning_rate": 9.187075456670033e-05, "loss": 0.0582, "step": 17990 }, { "grad_norm": 0.4367138743400574, "learning_rate": 9.1859454330203e-05, "loss": 0.0589, "step": 18000 }, { "grad_norm": 0.35842180252075195, "learning_rate": 9.18481469410698e-05, "loss": 0.0663, "step": 18010 }, { "grad_norm": 0.28809088468551636, "learning_rate": 9.183683240123281e-05, "loss": 0.0571, "step": 18020 }, { "grad_norm": 0.4677867293357849, "learning_rate": 9.182551071262541e-05, "loss": 0.0591, "step": 18030 }, { "grad_norm": 0.46515366435050964, "learning_rate": 9.181418187718218e-05, "loss": 0.07, "step": 18040 }, { "grad_norm": 0.4791201949119568, "learning_rate": 9.180284589683888e-05, "loss": 0.066, "step": 18050 }, { "grad_norm": 0.4180237352848053, "learning_rate": 9.17915027735325e-05, "loss": 0.0563, "step": 18060 }, { "grad_norm": 0.4621897339820862, "learning_rate": 9.178015250920133e-05, "loss": 0.0522, "step": 18070 }, { "grad_norm": 0.36921262741088867, "learning_rate": 9.176879510578477e-05, "loss": 0.0469, "step": 18080 }, { "grad_norm": 0.8857194781303406, "learning_rate": 9.17574305652235e-05, "loss": 0.0545, "step": 18090 }, { "grad_norm": 0.34082886576652527, "learning_rate": 9.174605888945942e-05, "loss": 0.0575, "step": 18100 }, { "grad_norm": 0.4252270460128784, "learning_rate": 9.173468008043564e-05, "loss": 0.0546, "step": 18110 }, { "grad_norm": 0.418486624956131, "learning_rate": 9.172329414009648e-05, "loss": 0.0539, "step": 18120 }, { "grad_norm": 0.6735098958015442, "learning_rate": 9.171190107038747e-05, "loss": 0.06, "step": 18130 }, { "grad_norm": 0.5621034502983093, "learning_rate": 9.170050087325541e-05, "loss": 0.0527, "step": 18140 }, { "grad_norm": 0.4466221332550049, "learning_rate": 9.168909355064824e-05, "loss": 0.0513, "step": 18150 }, { "grad_norm": 0.5104590058326721, "learning_rate": 9.167767910451519e-05, "loss": 0.0571, "step": 18160 }, { "grad_norm": 0.5175824761390686, "learning_rate": 9.166625753680669e-05, "loss": 0.057, "step": 18170 }, { "grad_norm": 0.34727397561073303, "learning_rate": 9.165482884947431e-05, "loss": 0.045, "step": 18180 }, { "grad_norm": 0.43622884154319763, "learning_rate": 9.164339304447098e-05, "loss": 0.0519, "step": 18190 }, { "grad_norm": 0.568386435508728, "learning_rate": 9.163195012375072e-05, "loss": 0.0461, "step": 18200 }, { "grad_norm": 0.6848515272140503, "learning_rate": 9.16205000892688e-05, "loss": 0.0535, "step": 18210 }, { "grad_norm": 0.6346338987350464, "learning_rate": 9.160904294298175e-05, "loss": 0.0475, "step": 18220 }, { "grad_norm": 0.35594791173934937, "learning_rate": 9.159757868684727e-05, "loss": 0.0493, "step": 18230 }, { "grad_norm": 1.2062833309173584, "learning_rate": 9.15861073228243e-05, "loss": 0.0535, "step": 18240 }, { "grad_norm": 0.5927925705909729, "learning_rate": 9.157462885287296e-05, "loss": 0.0637, "step": 18250 }, { "grad_norm": 0.5197908878326416, "learning_rate": 9.156314327895461e-05, "loss": 0.0605, "step": 18260 }, { "grad_norm": 0.344714492559433, "learning_rate": 9.155165060303185e-05, "loss": 0.0552, "step": 18270 }, { "grad_norm": 0.39346957206726074, "learning_rate": 9.154015082706841e-05, "loss": 0.0506, "step": 18280 }, { "grad_norm": 0.5673379302024841, "learning_rate": 9.152864395302936e-05, "loss": 0.0668, "step": 18290 }, { "grad_norm": 0.44096657633781433, "learning_rate": 9.151712998288085e-05, "loss": 0.0591, "step": 18300 }, { "grad_norm": 0.5428701639175415, "learning_rate": 9.150560891859031e-05, "loss": 0.0559, "step": 18310 }, { "grad_norm": 0.49211493134498596, "learning_rate": 9.14940807621264e-05, "loss": 0.0566, "step": 18320 }, { "grad_norm": 0.6043431758880615, "learning_rate": 9.148254551545894e-05, "loss": 0.0526, "step": 18330 }, { "grad_norm": 0.5078730583190918, "learning_rate": 9.147100318055901e-05, "loss": 0.0552, "step": 18340 }, { "grad_norm": 0.4033052623271942, "learning_rate": 9.145945375939888e-05, "loss": 0.0486, "step": 18350 }, { "grad_norm": 0.5125427842140198, "learning_rate": 9.144789725395203e-05, "loss": 0.0482, "step": 18360 }, { "grad_norm": 0.47791072726249695, "learning_rate": 9.14363336661931e-05, "loss": 0.0507, "step": 18370 }, { "grad_norm": 0.5455305576324463, "learning_rate": 9.142476299809806e-05, "loss": 0.0567, "step": 18380 }, { "grad_norm": 0.6047282218933105, "learning_rate": 9.1413185251644e-05, "loss": 0.0604, "step": 18390 }, { "grad_norm": 0.43486323952674866, "learning_rate": 9.140160042880923e-05, "loss": 0.0575, "step": 18400 }, { "grad_norm": 0.4536930322647095, "learning_rate": 9.139000853157327e-05, "loss": 0.0595, "step": 18410 }, { "grad_norm": 0.46186938881874084, "learning_rate": 9.137840956191688e-05, "loss": 0.0627, "step": 18420 }, { "grad_norm": 0.6296013593673706, "learning_rate": 9.136680352182199e-05, "loss": 0.053, "step": 18430 }, { "grad_norm": 0.3923041820526123, "learning_rate": 9.135519041327177e-05, "loss": 0.0399, "step": 18440 }, { "grad_norm": 0.3453211784362793, "learning_rate": 9.134357023825058e-05, "loss": 0.0545, "step": 18450 }, { "grad_norm": 0.4217585623264313, "learning_rate": 9.133194299874398e-05, "loss": 0.0596, "step": 18460 }, { "grad_norm": 0.585760772228241, "learning_rate": 9.132030869673876e-05, "loss": 0.0584, "step": 18470 }, { "grad_norm": 0.476273775100708, "learning_rate": 9.130866733422288e-05, "loss": 0.0487, "step": 18480 }, { "grad_norm": 0.47802767157554626, "learning_rate": 9.129701891318556e-05, "loss": 0.0531, "step": 18490 }, { "grad_norm": 0.24811576306819916, "learning_rate": 9.128536343561718e-05, "loss": 0.0631, "step": 18500 }, { "grad_norm": 0.2776155471801758, "learning_rate": 9.127370090350934e-05, "loss": 0.0552, "step": 18510 }, { "grad_norm": 0.36996957659721375, "learning_rate": 9.126203131885487e-05, "loss": 0.0553, "step": 18520 }, { "grad_norm": 0.47491776943206787, "learning_rate": 9.125035468364775e-05, "loss": 0.0578, "step": 18530 }, { "grad_norm": 0.34746846556663513, "learning_rate": 9.123867099988322e-05, "loss": 0.0468, "step": 18540 }, { "grad_norm": 0.5232239961624146, "learning_rate": 9.122698026955769e-05, "loss": 0.0557, "step": 18550 }, { "grad_norm": 0.6021501421928406, "learning_rate": 9.12152824946688e-05, "loss": 0.0719, "step": 18560 }, { "grad_norm": 0.5720585584640503, "learning_rate": 9.120357767721538e-05, "loss": 0.0581, "step": 18570 }, { "grad_norm": 0.563089907169342, "learning_rate": 9.119186581919745e-05, "loss": 0.0503, "step": 18580 }, { "grad_norm": 0.5659278631210327, "learning_rate": 9.118014692261624e-05, "loss": 0.0507, "step": 18590 }, { "grad_norm": 0.5211268663406372, "learning_rate": 9.116842098947422e-05, "loss": 0.0694, "step": 18600 }, { "grad_norm": 0.5209834575653076, "learning_rate": 9.115668802177499e-05, "loss": 0.0564, "step": 18610 }, { "grad_norm": 0.4575141370296478, "learning_rate": 9.114494802152342e-05, "loss": 0.0653, "step": 18620 }, { "grad_norm": 0.612768292427063, "learning_rate": 9.113320099072555e-05, "loss": 0.0825, "step": 18630 }, { "grad_norm": 0.5723153352737427, "learning_rate": 9.112144693138864e-05, "loss": 0.0544, "step": 18640 }, { "grad_norm": 0.5220708250999451, "learning_rate": 9.110968584552111e-05, "loss": 0.0545, "step": 18650 }, { "grad_norm": 0.5281069278717041, "learning_rate": 9.109791773513264e-05, "loss": 0.0571, "step": 18660 }, { "grad_norm": 0.4785918593406677, "learning_rate": 9.108614260223403e-05, "loss": 0.0612, "step": 18670 }, { "grad_norm": 0.46798500418663025, "learning_rate": 9.107436044883738e-05, "loss": 0.0528, "step": 18680 }, { "grad_norm": 0.473031610250473, "learning_rate": 9.10625712769559e-05, "loss": 0.0516, "step": 18690 }, { "grad_norm": 0.37887904047966003, "learning_rate": 9.105077508860406e-05, "loss": 0.0441, "step": 18700 }, { "grad_norm": 1.8328883647918701, "learning_rate": 9.103897188579751e-05, "loss": 0.049, "step": 18710 }, { "grad_norm": 0.42846986651420593, "learning_rate": 9.102716167055308e-05, "loss": 0.054, "step": 18720 }, { "grad_norm": 0.48055559396743774, "learning_rate": 9.10153444448888e-05, "loss": 0.0858, "step": 18730 }, { "grad_norm": 0.5391286015510559, "learning_rate": 9.100352021082393e-05, "loss": 0.0522, "step": 18740 }, { "grad_norm": 0.6786708831787109, "learning_rate": 9.099168897037891e-05, "loss": 0.0592, "step": 18750 }, { "grad_norm": 0.5380852222442627, "learning_rate": 9.097985072557538e-05, "loss": 0.0545, "step": 18760 }, { "grad_norm": 0.41390788555145264, "learning_rate": 9.096800547843615e-05, "loss": 0.0784, "step": 18770 }, { "grad_norm": 0.5158698558807373, "learning_rate": 9.095615323098526e-05, "loss": 0.0723, "step": 18780 }, { "grad_norm": 0.5444900393486023, "learning_rate": 9.094429398524795e-05, "loss": 0.0522, "step": 18790 }, { "grad_norm": 0.5089141726493835, "learning_rate": 9.093242774325061e-05, "loss": 0.0496, "step": 18800 }, { "grad_norm": 0.4268593192100525, "learning_rate": 9.092055450702088e-05, "loss": 0.0557, "step": 18810 }, { "grad_norm": 0.51356041431427, "learning_rate": 9.090867427858756e-05, "loss": 0.047, "step": 18820 }, { "grad_norm": 0.4635311961174011, "learning_rate": 9.089678705998066e-05, "loss": 0.0547, "step": 18830 }, { "grad_norm": 0.5659905672073364, "learning_rate": 9.088489285323139e-05, "loss": 0.0615, "step": 18840 }, { "grad_norm": 0.6245047450065613, "learning_rate": 9.087299166037212e-05, "loss": 0.0628, "step": 18850 }, { "grad_norm": 0.42203930020332336, "learning_rate": 9.086108348343647e-05, "loss": 0.0586, "step": 18860 }, { "grad_norm": 0.3582535982131958, "learning_rate": 9.08491683244592e-05, "loss": 0.0598, "step": 18870 }, { "grad_norm": 0.5825019478797913, "learning_rate": 9.08372461854763e-05, "loss": 0.0523, "step": 18880 }, { "grad_norm": 0.5062922239303589, "learning_rate": 9.082531706852492e-05, "loss": 0.0521, "step": 18890 }, { "grad_norm": 0.41434434056282043, "learning_rate": 9.081338097564342e-05, "loss": 0.0621, "step": 18900 }, { "grad_norm": 0.48886916041374207, "learning_rate": 9.080143790887137e-05, "loss": 0.0465, "step": 18910 }, { "grad_norm": 0.5123961567878723, "learning_rate": 9.07894878702495e-05, "loss": 0.0497, "step": 18920 }, { "grad_norm": 0.6192600131034851, "learning_rate": 9.077753086181974e-05, "loss": 0.0553, "step": 18930 }, { "grad_norm": 0.49170562624931335, "learning_rate": 9.076556688562524e-05, "loss": 0.043, "step": 18940 }, { "grad_norm": 0.9748948812484741, "learning_rate": 9.075359594371029e-05, "loss": 0.0459, "step": 18950 }, { "grad_norm": 0.4376855194568634, "learning_rate": 9.07416180381204e-05, "loss": 0.0555, "step": 18960 }, { "grad_norm": 0.39298319816589355, "learning_rate": 9.072963317090228e-05, "loss": 0.0527, "step": 18970 }, { "grad_norm": 0.35177990794181824, "learning_rate": 9.071764134410382e-05, "loss": 0.0548, "step": 18980 }, { "grad_norm": 0.5174527168273926, "learning_rate": 9.070564255977407e-05, "loss": 0.0435, "step": 18990 }, { "grad_norm": 0.41346481442451477, "learning_rate": 9.06936368199633e-05, "loss": 0.0585, "step": 19000 }, { "grad_norm": 1.0452598333358765, "learning_rate": 9.0681624126723e-05, "loss": 0.0521, "step": 19010 }, { "grad_norm": 0.4192759692668915, "learning_rate": 9.066960448210576e-05, "loss": 0.0557, "step": 19020 }, { "grad_norm": 0.5163227915763855, "learning_rate": 9.065757788816543e-05, "loss": 0.057, "step": 19030 }, { "grad_norm": 0.6592527627944946, "learning_rate": 9.064554434695705e-05, "loss": 0.0649, "step": 19040 }, { "grad_norm": 0.4117596447467804, "learning_rate": 9.063350386053677e-05, "loss": 0.0606, "step": 19050 }, { "grad_norm": 0.4540465176105499, "learning_rate": 9.062145643096202e-05, "loss": 0.0524, "step": 19060 }, { "grad_norm": 0.5492464303970337, "learning_rate": 9.060940206029136e-05, "loss": 0.0573, "step": 19070 }, { "grad_norm": 0.4367077052593231, "learning_rate": 9.059734075058457e-05, "loss": 0.0535, "step": 19080 }, { "grad_norm": 0.39113089442253113, "learning_rate": 9.058527250390257e-05, "loss": 0.061, "step": 19090 }, { "grad_norm": 0.5092663168907166, "learning_rate": 9.057319732230752e-05, "loss": 0.0632, "step": 19100 }, { "grad_norm": 0.5032614469528198, "learning_rate": 9.056111520786273e-05, "loss": 0.0519, "step": 19110 }, { "grad_norm": 0.5167850852012634, "learning_rate": 9.054902616263268e-05, "loss": 0.0466, "step": 19120 }, { "grad_norm": 0.39899858832359314, "learning_rate": 9.05369301886831e-05, "loss": 0.0673, "step": 19130 }, { "grad_norm": 0.5332270264625549, "learning_rate": 9.052482728808083e-05, "loss": 0.0506, "step": 19140 }, { "grad_norm": 0.5019622445106506, "learning_rate": 9.051271746289391e-05, "loss": 0.039, "step": 19150 }, { "grad_norm": 0.3657451272010803, "learning_rate": 9.050060071519162e-05, "loss": 0.0402, "step": 19160 }, { "grad_norm": 0.5596876740455627, "learning_rate": 9.048847704704437e-05, "loss": 0.0447, "step": 19170 }, { "grad_norm": 0.6122698187828064, "learning_rate": 9.047634646052376e-05, "loss": 0.0488, "step": 19180 }, { "grad_norm": 0.605193555355072, "learning_rate": 9.046420895770256e-05, "loss": 0.0644, "step": 19190 }, { "grad_norm": 0.4316907525062561, "learning_rate": 9.045206454065473e-05, "loss": 0.0504, "step": 19200 }, { "grad_norm": 0.5150677561759949, "learning_rate": 9.043991321145546e-05, "loss": 0.0492, "step": 19210 }, { "grad_norm": 0.41491588950157166, "learning_rate": 9.042775497218105e-05, "loss": 0.0449, "step": 19220 }, { "grad_norm": 0.527043879032135, "learning_rate": 9.041558982490901e-05, "loss": 0.0584, "step": 19230 }, { "grad_norm": 0.5579095482826233, "learning_rate": 9.040341777171805e-05, "loss": 0.0509, "step": 19240 }, { "grad_norm": 0.49734383821487427, "learning_rate": 9.039123881468802e-05, "loss": 0.0488, "step": 19250 }, { "grad_norm": 0.4209725558757782, "learning_rate": 9.037905295589998e-05, "loss": 0.0474, "step": 19260 }, { "grad_norm": 0.3415996730327606, "learning_rate": 9.036686019743617e-05, "loss": 0.0436, "step": 19270 }, { "grad_norm": 0.6045629382133484, "learning_rate": 9.035466054137997e-05, "loss": 0.0463, "step": 19280 }, { "grad_norm": 0.4799043536186218, "learning_rate": 9.0342453989816e-05, "loss": 0.0583, "step": 19290 }, { "grad_norm": 0.4580201506614685, "learning_rate": 9.033024054483e-05, "loss": 0.0589, "step": 19300 }, { "grad_norm": 0.3555520474910736, "learning_rate": 9.031802020850894e-05, "loss": 0.0563, "step": 19310 }, { "grad_norm": 0.45860302448272705, "learning_rate": 9.030579298294092e-05, "loss": 0.0656, "step": 19320 }, { "grad_norm": 0.5507845282554626, "learning_rate": 9.029355887021524e-05, "loss": 0.0713, "step": 19330 }, { "grad_norm": 0.702667236328125, "learning_rate": 9.028131787242238e-05, "loss": 0.0754, "step": 19340 }, { "grad_norm": 0.4012373685836792, "learning_rate": 9.026906999165399e-05, "loss": 0.0625, "step": 19350 }, { "grad_norm": 0.5077152252197266, "learning_rate": 9.025681523000291e-05, "loss": 0.0552, "step": 19360 }, { "grad_norm": 0.3292287588119507, "learning_rate": 9.024455358956315e-05, "loss": 0.0483, "step": 19370 }, { "grad_norm": 0.426308274269104, "learning_rate": 9.023228507242984e-05, "loss": 0.0595, "step": 19380 }, { "grad_norm": 0.41500523686408997, "learning_rate": 9.022000968069937e-05, "loss": 0.0629, "step": 19390 }, { "grad_norm": 0.447209894657135, "learning_rate": 9.020772741646928e-05, "loss": 0.0486, "step": 19400 }, { "grad_norm": 0.47407636046409607, "learning_rate": 9.019543828183826e-05, "loss": 0.0502, "step": 19410 }, { "grad_norm": 0.6151295900344849, "learning_rate": 9.018314227890616e-05, "loss": 0.0558, "step": 19420 }, { "grad_norm": 0.5557923316955566, "learning_rate": 9.017083940977408e-05, "loss": 0.0682, "step": 19430 }, { "grad_norm": 0.3932513892650604, "learning_rate": 9.015852967654422e-05, "loss": 0.0399, "step": 19440 }, { "grad_norm": 0.39470213651657104, "learning_rate": 9.014621308131996e-05, "loss": 0.0476, "step": 19450 }, { "grad_norm": 0.4331963062286377, "learning_rate": 9.01338896262059e-05, "loss": 0.0441, "step": 19460 }, { "grad_norm": 0.8103521466255188, "learning_rate": 9.012155931330777e-05, "loss": 0.0648, "step": 19470 }, { "grad_norm": 0.4734151363372803, "learning_rate": 9.010922214473246e-05, "loss": 0.0477, "step": 19480 }, { "grad_norm": 0.5908887982368469, "learning_rate": 9.009687812258808e-05, "loss": 0.067, "step": 19490 }, { "grad_norm": 0.4744894504547119, "learning_rate": 9.00845272489839e-05, "loss": 0.057, "step": 19500 }, { "grad_norm": 0.4696533679962158, "learning_rate": 9.007216952603031e-05, "loss": 0.0777, "step": 19510 }, { "grad_norm": 0.3417789936065674, "learning_rate": 9.005980495583894e-05, "loss": 0.0508, "step": 19520 }, { "grad_norm": 0.44381633400917053, "learning_rate": 9.004743354052252e-05, "loss": 0.0431, "step": 19530 }, { "grad_norm": 0.30574676394462585, "learning_rate": 9.003505528219503e-05, "loss": 0.0439, "step": 19540 }, { "grad_norm": 0.39785802364349365, "learning_rate": 9.002267018297154e-05, "loss": 0.049, "step": 19550 }, { "grad_norm": 0.6203267574310303, "learning_rate": 9.001027824496834e-05, "loss": 0.0694, "step": 19560 }, { "grad_norm": 0.4606442153453827, "learning_rate": 8.999787947030287e-05, "loss": 0.0502, "step": 19570 }, { "grad_norm": 0.42593804001808167, "learning_rate": 8.998547386109376e-05, "loss": 0.059, "step": 19580 }, { "grad_norm": 0.4923399090766907, "learning_rate": 8.997306141946073e-05, "loss": 0.0555, "step": 19590 }, { "grad_norm": 0.556938648223877, "learning_rate": 8.996064214752481e-05, "loss": 0.0703, "step": 19600 }, { "grad_norm": 0.5652723908424377, "learning_rate": 8.994821604740806e-05, "loss": 0.0628, "step": 19610 }, { "grad_norm": 0.4318135678768158, "learning_rate": 8.993578312123377e-05, "loss": 0.0607, "step": 19620 }, { "grad_norm": 0.6437371969223022, "learning_rate": 8.992334337112639e-05, "loss": 0.0715, "step": 19630 }, { "grad_norm": 0.46937045454978943, "learning_rate": 8.991089679921154e-05, "loss": 0.0505, "step": 19640 }, { "grad_norm": 0.5169073939323425, "learning_rate": 8.989844340761599e-05, "loss": 0.0527, "step": 19650 }, { "grad_norm": 0.7428557276725769, "learning_rate": 8.988598319846768e-05, "loss": 0.0564, "step": 19660 }, { "grad_norm": 0.4333978593349457, "learning_rate": 8.987351617389574e-05, "loss": 0.0503, "step": 19670 }, { "grad_norm": 0.339313268661499, "learning_rate": 8.98610423360304e-05, "loss": 0.053, "step": 19680 }, { "grad_norm": 0.6500771045684814, "learning_rate": 8.984856168700317e-05, "loss": 0.0721, "step": 19690 }, { "grad_norm": 0.6960025429725647, "learning_rate": 8.983607422894658e-05, "loss": 0.0574, "step": 19700 }, { "grad_norm": 0.40725642442703247, "learning_rate": 8.982357996399442e-05, "loss": 0.0589, "step": 19710 }, { "grad_norm": 0.4422830641269684, "learning_rate": 8.981107889428164e-05, "loss": 0.0538, "step": 19720 }, { "grad_norm": 0.45275628566741943, "learning_rate": 8.979857102194428e-05, "loss": 0.0483, "step": 19730 }, { "grad_norm": 0.36615854501724243, "learning_rate": 8.978605634911968e-05, "loss": 0.0456, "step": 19740 }, { "grad_norm": 0.4949999153614044, "learning_rate": 8.977353487794616e-05, "loss": 0.0476, "step": 19750 }, { "grad_norm": 0.50687575340271, "learning_rate": 8.976100661056334e-05, "loss": 0.0555, "step": 19760 }, { "grad_norm": 0.33857694268226624, "learning_rate": 8.974847154911197e-05, "loss": 0.0516, "step": 19770 }, { "grad_norm": 0.3219503164291382, "learning_rate": 8.973592969573393e-05, "loss": 0.0535, "step": 19780 }, { "grad_norm": 0.45552217960357666, "learning_rate": 8.972338105257228e-05, "loss": 0.0491, "step": 19790 }, { "grad_norm": 0.29528236389160156, "learning_rate": 8.971082562177125e-05, "loss": 0.0433, "step": 19800 }, { "grad_norm": 0.5871306657791138, "learning_rate": 8.96982634054762e-05, "loss": 0.0522, "step": 19810 }, { "grad_norm": 0.541684091091156, "learning_rate": 8.96856944058337e-05, "loss": 0.0505, "step": 19820 }, { "grad_norm": 0.6238228678703308, "learning_rate": 8.967311862499144e-05, "loss": 0.0418, "step": 19830 }, { "grad_norm": 0.44916650652885437, "learning_rate": 8.966053606509825e-05, "loss": 0.0411, "step": 19840 }, { "grad_norm": 0.5548610687255859, "learning_rate": 8.964794672830417e-05, "loss": 0.0644, "step": 19850 }, { "grad_norm": 0.4762863218784332, "learning_rate": 8.963535061676038e-05, "loss": 0.0582, "step": 19860 }, { "grad_norm": 0.4626893699169159, "learning_rate": 8.962274773261918e-05, "loss": 0.0485, "step": 19870 }, { "grad_norm": 0.3822260797023773, "learning_rate": 8.961013807803409e-05, "loss": 0.0535, "step": 19880 }, { "grad_norm": 0.4080527126789093, "learning_rate": 8.959752165515973e-05, "loss": 0.0543, "step": 19890 }, { "grad_norm": 0.6147271394729614, "learning_rate": 8.958489846615193e-05, "loss": 0.0707, "step": 19900 }, { "grad_norm": 0.29148969054222107, "learning_rate": 8.957226851316762e-05, "loss": 0.0557, "step": 19910 }, { "grad_norm": 1.3847965002059937, "learning_rate": 8.955963179836493e-05, "loss": 0.0507, "step": 19920 }, { "grad_norm": 0.39873361587524414, "learning_rate": 8.954698832390312e-05, "loss": 0.0496, "step": 19930 }, { "grad_norm": 0.5329576134681702, "learning_rate": 8.953433809194263e-05, "loss": 0.0477, "step": 19940 }, { "grad_norm": 0.43259793519973755, "learning_rate": 8.9521681104645e-05, "loss": 0.046, "step": 19950 }, { "grad_norm": 0.37194862961769104, "learning_rate": 8.9509017364173e-05, "loss": 0.0453, "step": 19960 }, { "grad_norm": 0.5157869458198547, "learning_rate": 8.949634687269052e-05, "loss": 0.0503, "step": 19970 }, { "grad_norm": 0.5186125040054321, "learning_rate": 8.948366963236259e-05, "loss": 0.0634, "step": 19980 }, { "grad_norm": 0.43977609276771545, "learning_rate": 8.947098564535538e-05, "loss": 0.0546, "step": 19990 }, { "grad_norm": 0.5391355752944946, "learning_rate": 8.945829491383627e-05, "loss": 0.0578, "step": 20000 }, { "grad_norm": 0.5975368618965149, "learning_rate": 8.944559743997374e-05, "loss": 0.0681, "step": 20010 }, { "grad_norm": 0.656181275844574, "learning_rate": 8.943289322593746e-05, "loss": 0.0509, "step": 20020 }, { "grad_norm": 0.39197060465812683, "learning_rate": 8.942018227389821e-05, "loss": 0.0461, "step": 20030 }, { "grad_norm": 0.4731140732765198, "learning_rate": 8.940746458602795e-05, "loss": 0.0613, "step": 20040 }, { "grad_norm": 1.0002617835998535, "learning_rate": 8.939474016449979e-05, "loss": 0.0585, "step": 20050 }, { "grad_norm": 0.7170873880386353, "learning_rate": 8.938200901148799e-05, "loss": 0.0479, "step": 20060 }, { "grad_norm": 0.6718692779541016, "learning_rate": 8.936927112916795e-05, "loss": 0.0672, "step": 20070 }, { "grad_norm": 0.3381393253803253, "learning_rate": 8.935652651971622e-05, "loss": 0.0465, "step": 20080 }, { "grad_norm": 0.5828379392623901, "learning_rate": 8.934377518531052e-05, "loss": 0.0489, "step": 20090 }, { "grad_norm": 0.6261377930641174, "learning_rate": 8.933101712812967e-05, "loss": 0.056, "step": 20100 }, { "grad_norm": 0.5831194519996643, "learning_rate": 8.931825235035374e-05, "loss": 0.0606, "step": 20110 }, { "grad_norm": 0.6322402358055115, "learning_rate": 8.930548085416382e-05, "loss": 0.0535, "step": 20120 }, { "grad_norm": 0.5600718855857849, "learning_rate": 8.92927026417422e-05, "loss": 0.0593, "step": 20130 }, { "grad_norm": 0.4701082110404968, "learning_rate": 8.92799177152724e-05, "loss": 0.0429, "step": 20140 }, { "grad_norm": 0.31485870480537415, "learning_rate": 8.926712607693895e-05, "loss": 0.0466, "step": 20150 }, { "grad_norm": 0.5820761322975159, "learning_rate": 8.925432772892762e-05, "loss": 0.0548, "step": 20160 }, { "grad_norm": 0.5069075226783752, "learning_rate": 8.924152267342529e-05, "loss": 0.0544, "step": 20170 }, { "grad_norm": 0.4341265857219696, "learning_rate": 8.922871091261998e-05, "loss": 0.0566, "step": 20180 }, { "grad_norm": 0.3587571680545807, "learning_rate": 8.92158924487009e-05, "loss": 0.0476, "step": 20190 }, { "grad_norm": 0.5838915109634399, "learning_rate": 8.920306728385834e-05, "loss": 0.0476, "step": 20200 }, { "grad_norm": 0.4975401759147644, "learning_rate": 8.919023542028379e-05, "loss": 0.0535, "step": 20210 }, { "grad_norm": 0.5200842022895813, "learning_rate": 8.917739686016988e-05, "loss": 0.0603, "step": 20220 }, { "grad_norm": 0.7264800667762756, "learning_rate": 8.916455160571033e-05, "loss": 0.0712, "step": 20230 }, { "grad_norm": 0.4606236219406128, "learning_rate": 8.915169965910008e-05, "loss": 0.054, "step": 20240 }, { "grad_norm": 0.4487776756286621, "learning_rate": 8.913884102253514e-05, "loss": 0.0491, "step": 20250 }, { "grad_norm": 0.6731328964233398, "learning_rate": 8.912597569821273e-05, "loss": 0.0432, "step": 20260 }, { "grad_norm": 0.573594868183136, "learning_rate": 8.911310368833118e-05, "loss": 0.0577, "step": 20270 }, { "grad_norm": 0.6521437764167786, "learning_rate": 8.910022499508994e-05, "loss": 0.0556, "step": 20280 }, { "grad_norm": 1.0748718976974487, "learning_rate": 8.908733962068965e-05, "loss": 0.0638, "step": 20290 }, { "grad_norm": 0.4773772060871124, "learning_rate": 8.907444756733207e-05, "loss": 0.0501, "step": 20300 }, { "grad_norm": 0.37512174248695374, "learning_rate": 8.906154883722006e-05, "loss": 0.0541, "step": 20310 }, { "grad_norm": 0.9992350935935974, "learning_rate": 8.904864343255773e-05, "loss": 0.0572, "step": 20320 }, { "grad_norm": 0.3877463638782501, "learning_rate": 8.90357313555502e-05, "loss": 0.0521, "step": 20330 }, { "grad_norm": 0.39984026551246643, "learning_rate": 8.90228126084038e-05, "loss": 0.0611, "step": 20340 }, { "grad_norm": 0.4398178160190582, "learning_rate": 8.900988719332601e-05, "loss": 0.0598, "step": 20350 }, { "grad_norm": 0.3502620458602905, "learning_rate": 8.899695511252542e-05, "loss": 0.0552, "step": 20360 }, { "grad_norm": 0.42104148864746094, "learning_rate": 8.898401636821176e-05, "loss": 0.0589, "step": 20370 }, { "grad_norm": 0.522936999797821, "learning_rate": 8.897107096259593e-05, "loss": 0.0668, "step": 20380 }, { "grad_norm": 0.4384613633155823, "learning_rate": 8.895811889788994e-05, "loss": 0.0667, "step": 20390 }, { "grad_norm": 0.31856220960617065, "learning_rate": 8.894516017630692e-05, "loss": 0.0445, "step": 20400 }, { "grad_norm": 0.44200900197029114, "learning_rate": 8.893219480006118e-05, "loss": 0.0518, "step": 20410 }, { "grad_norm": 0.6891664266586304, "learning_rate": 8.891922277136817e-05, "loss": 0.0449, "step": 20420 }, { "grad_norm": 0.4879913330078125, "learning_rate": 8.890624409244441e-05, "loss": 0.0483, "step": 20430 }, { "grad_norm": 0.4757779836654663, "learning_rate": 8.889325876550763e-05, "loss": 0.0527, "step": 20440 }, { "grad_norm": 0.3897044062614441, "learning_rate": 8.888026679277666e-05, "loss": 0.0546, "step": 20450 }, { "grad_norm": 0.5411668419837952, "learning_rate": 8.886726817647147e-05, "loss": 0.0735, "step": 20460 }, { "grad_norm": 0.5978240966796875, "learning_rate": 8.885426291881319e-05, "loss": 0.0661, "step": 20470 }, { "grad_norm": 0.4703297019004822, "learning_rate": 8.884125102202401e-05, "loss": 0.051, "step": 20480 }, { "grad_norm": 0.4034339189529419, "learning_rate": 8.882823248832736e-05, "loss": 0.0417, "step": 20490 }, { "grad_norm": 0.42968878149986267, "learning_rate": 8.881520731994772e-05, "loss": 0.0551, "step": 20500 }, { "grad_norm": 0.4703235626220703, "learning_rate": 8.880217551911077e-05, "loss": 0.0609, "step": 20510 }, { "grad_norm": 0.3481132686138153, "learning_rate": 8.878913708804323e-05, "loss": 0.0529, "step": 20520 }, { "grad_norm": 0.41943320631980896, "learning_rate": 8.877609202897308e-05, "loss": 0.0581, "step": 20530 }, { "grad_norm": 0.4356236159801483, "learning_rate": 8.876304034412933e-05, "loss": 0.0456, "step": 20540 }, { "grad_norm": 0.5073036551475525, "learning_rate": 8.874998203574214e-05, "loss": 0.0524, "step": 20550 }, { "grad_norm": 0.6247925758361816, "learning_rate": 8.873691710604284e-05, "loss": 0.0447, "step": 20560 }, { "grad_norm": 0.4690244495868683, "learning_rate": 8.872384555726387e-05, "loss": 0.0485, "step": 20570 }, { "grad_norm": 0.5302172899246216, "learning_rate": 8.871076739163878e-05, "loss": 0.0487, "step": 20580 }, { "grad_norm": 0.25114142894744873, "learning_rate": 8.86976826114023e-05, "loss": 0.0486, "step": 20590 }, { "grad_norm": 0.42916831374168396, "learning_rate": 8.868459121879023e-05, "loss": 0.0578, "step": 20600 }, { "grad_norm": 0.587990403175354, "learning_rate": 8.867149321603956e-05, "loss": 0.0576, "step": 20610 }, { "grad_norm": 0.449551522731781, "learning_rate": 8.865838860538835e-05, "loss": 0.0544, "step": 20620 }, { "grad_norm": 0.5311264395713806, "learning_rate": 8.864527738907585e-05, "loss": 0.0517, "step": 20630 }, { "grad_norm": 0.6178710460662842, "learning_rate": 8.863215956934239e-05, "loss": 0.049, "step": 20640 }, { "grad_norm": 1.4615232944488525, "learning_rate": 8.861903514842947e-05, "loss": 0.0407, "step": 20650 }, { "grad_norm": 0.4196693003177643, "learning_rate": 8.860590412857966e-05, "loss": 0.0461, "step": 20660 }, { "grad_norm": 0.3098660111427307, "learning_rate": 8.85927665120367e-05, "loss": 0.0399, "step": 20670 }, { "grad_norm": 0.4703763723373413, "learning_rate": 8.857962230104546e-05, "loss": 0.0447, "step": 20680 }, { "grad_norm": 0.46919623017311096, "learning_rate": 8.856647149785193e-05, "loss": 0.0502, "step": 20690 }, { "grad_norm": 0.9228067994117737, "learning_rate": 8.855331410470322e-05, "loss": 0.0559, "step": 20700 }, { "grad_norm": 0.3114263117313385, "learning_rate": 8.854015012384756e-05, "loss": 0.0515, "step": 20710 }, { "grad_norm": 0.3644132614135742, "learning_rate": 8.852697955753433e-05, "loss": 0.054, "step": 20720 }, { "grad_norm": 0.33606767654418945, "learning_rate": 8.851380240801399e-05, "loss": 0.0522, "step": 20730 }, { "grad_norm": 0.3952462673187256, "learning_rate": 8.850061867753818e-05, "loss": 0.0356, "step": 20740 }, { "grad_norm": 0.6062120795249939, "learning_rate": 8.848742836835963e-05, "loss": 0.0567, "step": 20750 }, { "grad_norm": 0.37977346777915955, "learning_rate": 8.847423148273221e-05, "loss": 0.0508, "step": 20760 }, { "grad_norm": 0.495470255613327, "learning_rate": 8.846102802291092e-05, "loss": 0.0579, "step": 20770 }, { "grad_norm": 0.46408501267433167, "learning_rate": 8.844781799115183e-05, "loss": 0.0552, "step": 20780 }, { "grad_norm": 0.4493902921676636, "learning_rate": 8.84346013897122e-05, "loss": 0.0422, "step": 20790 }, { "grad_norm": 0.5891146063804626, "learning_rate": 8.842137822085038e-05, "loss": 0.0563, "step": 20800 }, { "grad_norm": 0.6688717603683472, "learning_rate": 8.840814848682585e-05, "loss": 0.0454, "step": 20810 }, { "grad_norm": 0.4136405289173126, "learning_rate": 8.83949121898992e-05, "loss": 0.0685, "step": 20820 }, { "grad_norm": 0.33442187309265137, "learning_rate": 8.838166933233217e-05, "loss": 0.0453, "step": 20830 }, { "grad_norm": 0.4235641360282898, "learning_rate": 8.83684199163876e-05, "loss": 0.0537, "step": 20840 }, { "grad_norm": 0.34721729159355164, "learning_rate": 8.835516394432943e-05, "loss": 0.0578, "step": 20850 }, { "grad_norm": 0.609378457069397, "learning_rate": 8.834190141842276e-05, "loss": 0.0546, "step": 20860 }, { "grad_norm": 0.4686500132083893, "learning_rate": 8.83286323409338e-05, "loss": 0.0524, "step": 20870 }, { "grad_norm": 0.39550095796585083, "learning_rate": 8.831535671412986e-05, "loss": 0.0466, "step": 20880 }, { "grad_norm": 0.34392300248146057, "learning_rate": 8.830207454027938e-05, "loss": 0.0501, "step": 20890 }, { "grad_norm": 0.576627254486084, "learning_rate": 8.828878582165192e-05, "loss": 0.0518, "step": 20900 }, { "grad_norm": 0.6828538179397583, "learning_rate": 8.827549056051818e-05, "loss": 0.0424, "step": 20910 }, { "grad_norm": 0.5803424715995789, "learning_rate": 8.826218875914993e-05, "loss": 0.0515, "step": 20920 }, { "grad_norm": 0.41193777322769165, "learning_rate": 8.82488804198201e-05, "loss": 0.0516, "step": 20930 }, { "grad_norm": 0.3089330792427063, "learning_rate": 8.82355655448027e-05, "loss": 0.0581, "step": 20940 }, { "grad_norm": 0.6272918581962585, "learning_rate": 8.822224413637293e-05, "loss": 0.0594, "step": 20950 }, { "grad_norm": 0.34205392003059387, "learning_rate": 8.820891619680697e-05, "loss": 0.0487, "step": 20960 }, { "grad_norm": 0.44036340713500977, "learning_rate": 8.819558172838227e-05, "loss": 0.0452, "step": 20970 }, { "grad_norm": 0.2957168221473694, "learning_rate": 8.818224073337731e-05, "loss": 0.0414, "step": 20980 }, { "grad_norm": 0.7096449732780457, "learning_rate": 8.816889321407169e-05, "loss": 0.0527, "step": 20990 }, { "grad_norm": 0.3360976278781891, "learning_rate": 8.815553917274615e-05, "loss": 0.0485, "step": 21000 }, { "grad_norm": 0.4420130252838135, "learning_rate": 8.81421786116825e-05, "loss": 0.0557, "step": 21010 }, { "grad_norm": 0.9724693298339844, "learning_rate": 8.812881153316373e-05, "loss": 0.0493, "step": 21020 }, { "grad_norm": 0.40369105339050293, "learning_rate": 8.81154379394739e-05, "loss": 0.0563, "step": 21030 }, { "grad_norm": 0.5153401494026184, "learning_rate": 8.810205783289818e-05, "loss": 0.0494, "step": 21040 }, { "grad_norm": 0.46104446053504944, "learning_rate": 8.808867121572286e-05, "loss": 0.049, "step": 21050 }, { "grad_norm": 0.505578875541687, "learning_rate": 8.807527809023537e-05, "loss": 0.0644, "step": 21060 }, { "grad_norm": 0.44097718596458435, "learning_rate": 8.80618784587242e-05, "loss": 0.0557, "step": 21070 }, { "grad_norm": 0.48608338832855225, "learning_rate": 8.804847232347902e-05, "loss": 0.0467, "step": 21080 }, { "grad_norm": 0.42436710000038147, "learning_rate": 8.803505968679054e-05, "loss": 0.0598, "step": 21090 }, { "grad_norm": 0.43708574771881104, "learning_rate": 8.802164055095061e-05, "loss": 0.0538, "step": 21100 }, { "grad_norm": 0.7824268341064453, "learning_rate": 8.80082149182522e-05, "loss": 0.0539, "step": 21110 }, { "grad_norm": 0.39038917422294617, "learning_rate": 8.79947827909894e-05, "loss": 0.0552, "step": 21120 }, { "grad_norm": 0.45566901564598083, "learning_rate": 8.798134417145738e-05, "loss": 0.0551, "step": 21130 }, { "grad_norm": 0.4827585816383362, "learning_rate": 8.796789906195243e-05, "loss": 0.0475, "step": 21140 }, { "grad_norm": 0.8066015243530273, "learning_rate": 8.795444746477195e-05, "loss": 0.0562, "step": 21150 }, { "grad_norm": 0.48756200075149536, "learning_rate": 8.794098938221446e-05, "loss": 0.0547, "step": 21160 }, { "grad_norm": 0.46640244126319885, "learning_rate": 8.792752481657957e-05, "loss": 0.0541, "step": 21170 }, { "grad_norm": 0.8475204706192017, "learning_rate": 8.791405377016802e-05, "loss": 0.0528, "step": 21180 }, { "grad_norm": 0.4001831114292145, "learning_rate": 8.790057624528163e-05, "loss": 0.0526, "step": 21190 }, { "grad_norm": 0.45697733759880066, "learning_rate": 8.788709224422333e-05, "loss": 0.0518, "step": 21200 }, { "grad_norm": 0.5968452095985413, "learning_rate": 8.787360176929717e-05, "loss": 0.0515, "step": 21210 }, { "grad_norm": 0.4528087377548218, "learning_rate": 8.786010482280834e-05, "loss": 0.0488, "step": 21220 }, { "grad_norm": 0.5552605986595154, "learning_rate": 8.784660140706306e-05, "loss": 0.0595, "step": 21230 }, { "grad_norm": 0.5084413886070251, "learning_rate": 8.783309152436872e-05, "loss": 0.0463, "step": 21240 }, { "grad_norm": 0.37269818782806396, "learning_rate": 8.781957517703375e-05, "loss": 0.0492, "step": 21250 }, { "grad_norm": 0.40402162075042725, "learning_rate": 8.780605236736776e-05, "loss": 0.0506, "step": 21260 }, { "grad_norm": 0.42880210280418396, "learning_rate": 8.779252309768142e-05, "loss": 0.0636, "step": 21270 }, { "grad_norm": 0.32690009474754333, "learning_rate": 8.777898737028652e-05, "loss": 0.047, "step": 21280 }, { "grad_norm": 0.45669665932655334, "learning_rate": 8.776544518749591e-05, "loss": 0.0444, "step": 21290 }, { "grad_norm": 0.39031901955604553, "learning_rate": 8.775189655162364e-05, "loss": 0.0475, "step": 21300 }, { "grad_norm": 0.4325757622718811, "learning_rate": 8.773834146498474e-05, "loss": 0.0512, "step": 21310 }, { "grad_norm": 0.4208393692970276, "learning_rate": 8.772477992989545e-05, "loss": 0.041, "step": 21320 }, { "grad_norm": 0.4304693937301636, "learning_rate": 8.771121194867304e-05, "loss": 0.0514, "step": 21330 }, { "grad_norm": 0.5489794611930847, "learning_rate": 8.769763752363589e-05, "loss": 0.0431, "step": 21340 }, { "grad_norm": 0.4035125970840454, "learning_rate": 8.768405665710352e-05, "loss": 0.0518, "step": 21350 }, { "grad_norm": 0.49710771441459656, "learning_rate": 8.767046935139655e-05, "loss": 0.0548, "step": 21360 }, { "grad_norm": 0.5747507810592651, "learning_rate": 8.765687560883666e-05, "loss": 0.044, "step": 21370 }, { "grad_norm": 0.727399468421936, "learning_rate": 8.764327543174664e-05, "loss": 0.0399, "step": 21380 }, { "grad_norm": 0.4035472571849823, "learning_rate": 8.762966882245038e-05, "loss": 0.0471, "step": 21390 }, { "grad_norm": 0.41025087237358093, "learning_rate": 8.761605578327291e-05, "loss": 0.0531, "step": 21400 }, { "grad_norm": 0.3815823197364807, "learning_rate": 8.76024363165403e-05, "loss": 0.0656, "step": 21410 }, { "grad_norm": 0.4736960232257843, "learning_rate": 8.758881042457976e-05, "loss": 0.0535, "step": 21420 }, { "grad_norm": 0.6863701939582825, "learning_rate": 8.757517810971957e-05, "loss": 0.0492, "step": 21430 }, { "grad_norm": 0.4388402998447418, "learning_rate": 8.756153937428913e-05, "loss": 0.0521, "step": 21440 }, { "grad_norm": 0.5166118144989014, "learning_rate": 8.754789422061889e-05, "loss": 0.0608, "step": 21450 }, { "grad_norm": 0.6599544882774353, "learning_rate": 8.753424265104052e-05, "loss": 0.049, "step": 21460 }, { "grad_norm": 0.440029501914978, "learning_rate": 8.752058466788659e-05, "loss": 0.0575, "step": 21470 }, { "grad_norm": 0.38217321038246155, "learning_rate": 8.750692027349097e-05, "loss": 0.0473, "step": 21480 }, { "grad_norm": 0.5267980694770813, "learning_rate": 8.749324947018847e-05, "loss": 0.0478, "step": 21490 }, { "grad_norm": 0.48250746726989746, "learning_rate": 8.747957226031507e-05, "loss": 0.0529, "step": 21500 }, { "grad_norm": 0.4046100378036499, "learning_rate": 8.746588864620787e-05, "loss": 0.051, "step": 21510 }, { "grad_norm": 0.5571536421775818, "learning_rate": 8.745219863020498e-05, "loss": 0.0483, "step": 21520 }, { "grad_norm": 0.4849887788295746, "learning_rate": 8.743850221464564e-05, "loss": 0.0472, "step": 21530 }, { "grad_norm": 0.5377779603004456, "learning_rate": 8.742479940187026e-05, "loss": 0.0554, "step": 21540 }, { "grad_norm": 0.655647873878479, "learning_rate": 8.74110901942202e-05, "loss": 0.0528, "step": 21550 }, { "grad_norm": 0.44080987572669983, "learning_rate": 8.739737459403803e-05, "loss": 0.0483, "step": 21560 }, { "grad_norm": 0.5368458032608032, "learning_rate": 8.738365260366737e-05, "loss": 0.0538, "step": 21570 }, { "grad_norm": 0.3867794871330261, "learning_rate": 8.736992422545292e-05, "loss": 0.0438, "step": 21580 }, { "grad_norm": 0.3680649995803833, "learning_rate": 8.73561894617405e-05, "loss": 0.0534, "step": 21590 }, { "grad_norm": 1.4462703466415405, "learning_rate": 8.734244831487697e-05, "loss": 0.0504, "step": 21600 }, { "grad_norm": 0.6135678887367249, "learning_rate": 8.732870078721035e-05, "loss": 0.0657, "step": 21610 }, { "grad_norm": 0.4100572466850281, "learning_rate": 8.731494688108972e-05, "loss": 0.0465, "step": 21620 }, { "grad_norm": 0.781844973564148, "learning_rate": 8.730118659886523e-05, "loss": 0.0501, "step": 21630 }, { "grad_norm": 0.38538965582847595, "learning_rate": 8.728741994288814e-05, "loss": 0.0534, "step": 21640 }, { "grad_norm": 0.32723620533943176, "learning_rate": 8.727364691551079e-05, "loss": 0.045, "step": 21650 }, { "grad_norm": 0.4933920204639435, "learning_rate": 8.725986751908661e-05, "loss": 0.0495, "step": 21660 }, { "grad_norm": 0.48398375511169434, "learning_rate": 8.724608175597016e-05, "loss": 0.0406, "step": 21670 }, { "grad_norm": 0.27908268570899963, "learning_rate": 8.723228962851699e-05, "loss": 0.0386, "step": 21680 }, { "grad_norm": 0.5246523022651672, "learning_rate": 8.721849113908385e-05, "loss": 0.0555, "step": 21690 }, { "grad_norm": 0.4629148244857788, "learning_rate": 8.720468629002848e-05, "loss": 0.0468, "step": 21700 }, { "grad_norm": 0.6231836676597595, "learning_rate": 8.719087508370978e-05, "loss": 0.0465, "step": 21710 }, { "grad_norm": 0.3936467468738556, "learning_rate": 8.717705752248772e-05, "loss": 0.0507, "step": 21720 }, { "grad_norm": 0.5460842847824097, "learning_rate": 8.71632336087233e-05, "loss": 0.0413, "step": 21730 }, { "grad_norm": 0.4482916593551636, "learning_rate": 8.71494033447787e-05, "loss": 0.0444, "step": 21740 }, { "grad_norm": 0.44144877791404724, "learning_rate": 8.713556673301708e-05, "loss": 0.041, "step": 21750 }, { "grad_norm": 0.7946435809135437, "learning_rate": 8.712172377580278e-05, "loss": 0.0665, "step": 21760 }, { "grad_norm": 0.5316463112831116, "learning_rate": 8.710787447550114e-05, "loss": 0.0567, "step": 21770 }, { "grad_norm": 0.47441989183425903, "learning_rate": 8.70940188344787e-05, "loss": 0.0588, "step": 21780 }, { "grad_norm": 0.4539816677570343, "learning_rate": 8.708015685510293e-05, "loss": 0.042, "step": 21790 }, { "grad_norm": 0.4659745991230011, "learning_rate": 8.706628853974252e-05, "loss": 0.0514, "step": 21800 }, { "grad_norm": 0.5068458914756775, "learning_rate": 8.705241389076715e-05, "loss": 0.0431, "step": 21810 }, { "grad_norm": 0.323630154132843, "learning_rate": 8.703853291054764e-05, "loss": 0.041, "step": 21820 }, { "grad_norm": 0.4481382966041565, "learning_rate": 8.702464560145587e-05, "loss": 0.0468, "step": 21830 }, { "grad_norm": 0.3622024357318878, "learning_rate": 8.701075196586476e-05, "loss": 0.0381, "step": 21840 }, { "grad_norm": 0.45113059878349304, "learning_rate": 8.699685200614842e-05, "loss": 0.0458, "step": 21850 }, { "grad_norm": 0.5513678193092346, "learning_rate": 8.698294572468193e-05, "loss": 0.0564, "step": 21860 }, { "grad_norm": 1.6041419506072998, "learning_rate": 8.696903312384148e-05, "loss": 0.0594, "step": 21870 }, { "grad_norm": 0.5674536228179932, "learning_rate": 8.695511420600439e-05, "loss": 0.046, "step": 21880 }, { "grad_norm": 0.3852362632751465, "learning_rate": 8.694118897354901e-05, "loss": 0.0414, "step": 21890 }, { "grad_norm": 0.4247594475746155, "learning_rate": 8.692725742885478e-05, "loss": 0.0505, "step": 21900 }, { "grad_norm": 0.5782657861709595, "learning_rate": 8.691331957430221e-05, "loss": 0.054, "step": 21910 }, { "grad_norm": 0.34600046277046204, "learning_rate": 8.68993754122729e-05, "loss": 0.0528, "step": 21920 }, { "grad_norm": 0.4136679172515869, "learning_rate": 8.688542494514955e-05, "loss": 0.0451, "step": 21930 }, { "grad_norm": 0.5763400197029114, "learning_rate": 8.68714681753159e-05, "loss": 0.0591, "step": 21940 }, { "grad_norm": 0.7235249280929565, "learning_rate": 8.685750510515676e-05, "loss": 0.0507, "step": 21950 }, { "grad_norm": 0.4332274794578552, "learning_rate": 8.684353573705805e-05, "loss": 0.0483, "step": 21960 }, { "grad_norm": 0.455151230096817, "learning_rate": 8.682956007340677e-05, "loss": 0.0465, "step": 21970 }, { "grad_norm": 0.52569180727005, "learning_rate": 8.681557811659095e-05, "loss": 0.0444, "step": 21980 }, { "grad_norm": 0.6050779819488525, "learning_rate": 8.680158986899974e-05, "loss": 0.0596, "step": 21990 }, { "grad_norm": 0.4510556161403656, "learning_rate": 8.678759533302335e-05, "loss": 0.0523, "step": 22000 }, { "grad_norm": 0.5193684101104736, "learning_rate": 8.677359451105308e-05, "loss": 0.05, "step": 22010 }, { "grad_norm": 0.3324309289455414, "learning_rate": 8.675958740548123e-05, "loss": 0.0552, "step": 22020 }, { "grad_norm": 1.1835675239562988, "learning_rate": 8.674557401870129e-05, "loss": 0.0526, "step": 22030 }, { "grad_norm": 0.49577251076698303, "learning_rate": 8.673155435310775e-05, "loss": 0.0636, "step": 22040 }, { "grad_norm": 0.44059672951698303, "learning_rate": 8.671752841109617e-05, "loss": 0.0542, "step": 22050 }, { "grad_norm": 0.5644385814666748, "learning_rate": 8.670349619506321e-05, "loss": 0.0476, "step": 22060 }, { "grad_norm": 0.32566341757774353, "learning_rate": 8.66894577074066e-05, "loss": 0.0518, "step": 22070 }, { "grad_norm": 0.5089062452316284, "learning_rate": 8.667541295052513e-05, "loss": 0.0474, "step": 22080 }, { "grad_norm": 1.1478972434997559, "learning_rate": 8.666136192681865e-05, "loss": 0.0668, "step": 22090 }, { "grad_norm": 0.4600183963775635, "learning_rate": 8.664730463868811e-05, "loss": 0.0563, "step": 22100 }, { "grad_norm": 0.626093864440918, "learning_rate": 8.663324108853552e-05, "loss": 0.0599, "step": 22110 }, { "grad_norm": 0.46917909383773804, "learning_rate": 8.661917127876395e-05, "loss": 0.0659, "step": 22120 }, { "grad_norm": 0.4462677836418152, "learning_rate": 8.660509521177754e-05, "loss": 0.0545, "step": 22130 }, { "grad_norm": 0.46261805295944214, "learning_rate": 8.65910128899815e-05, "loss": 0.0452, "step": 22140 }, { "grad_norm": 0.4657350182533264, "learning_rate": 8.657692431578214e-05, "loss": 0.0476, "step": 22150 }, { "grad_norm": 0.32349860668182373, "learning_rate": 8.656282949158679e-05, "loss": 0.0409, "step": 22160 }, { "grad_norm": 0.5807945728302002, "learning_rate": 8.654872841980388e-05, "loss": 0.0507, "step": 22170 }, { "grad_norm": 0.4550827741622925, "learning_rate": 8.653462110284289e-05, "loss": 0.0472, "step": 22180 }, { "grad_norm": 0.3761017918586731, "learning_rate": 8.652050754311437e-05, "loss": 0.0515, "step": 22190 }, { "grad_norm": 0.4445497691631317, "learning_rate": 8.650638774302995e-05, "loss": 0.0414, "step": 22200 }, { "grad_norm": 0.4266735911369324, "learning_rate": 8.649226170500233e-05, "loss": 0.0531, "step": 22210 }, { "grad_norm": 0.33392226696014404, "learning_rate": 8.647812943144524e-05, "loss": 0.0363, "step": 22220 }, { "grad_norm": 0.5116717219352722, "learning_rate": 8.646399092477351e-05, "loss": 0.0419, "step": 22230 }, { "grad_norm": 0.5009613633155823, "learning_rate": 8.644984618740301e-05, "loss": 0.0534, "step": 22240 }, { "grad_norm": 0.33296284079551697, "learning_rate": 8.643569522175073e-05, "loss": 0.0393, "step": 22250 }, { "grad_norm": 0.3865097165107727, "learning_rate": 8.642153803023463e-05, "loss": 0.0588, "step": 22260 }, { "grad_norm": 0.49673032760620117, "learning_rate": 8.640737461527383e-05, "loss": 0.0419, "step": 22270 }, { "grad_norm": 0.40006616711616516, "learning_rate": 8.639320497928845e-05, "loss": 0.045, "step": 22280 }, { "grad_norm": 0.45205917954444885, "learning_rate": 8.637902912469969e-05, "loss": 0.0419, "step": 22290 }, { "grad_norm": 0.3169107437133789, "learning_rate": 8.636484705392982e-05, "loss": 0.0426, "step": 22300 }, { "grad_norm": 0.5717531442642212, "learning_rate": 8.635065876940216e-05, "loss": 0.0433, "step": 22310 }, { "grad_norm": 0.40435484051704407, "learning_rate": 8.633646427354112e-05, "loss": 0.0482, "step": 22320 }, { "grad_norm": 0.3982279896736145, "learning_rate": 8.632226356877213e-05, "loss": 0.0462, "step": 22330 }, { "grad_norm": 0.4505939781665802, "learning_rate": 8.630805665752173e-05, "loss": 0.0515, "step": 22340 }, { "grad_norm": 0.4078729450702667, "learning_rate": 8.629384354221748e-05, "loss": 0.0503, "step": 22350 }, { "grad_norm": 0.339046448469162, "learning_rate": 8.627962422528797e-05, "loss": 0.0455, "step": 22360 }, { "grad_norm": 0.35904303193092346, "learning_rate": 8.626539870916296e-05, "loss": 0.0683, "step": 22370 }, { "grad_norm": 0.34505361318588257, "learning_rate": 8.625116699627317e-05, "loss": 0.046, "step": 22380 }, { "grad_norm": 0.4868096113204956, "learning_rate": 8.623692908905041e-05, "loss": 0.0432, "step": 22390 }, { "grad_norm": 0.43026596307754517, "learning_rate": 8.622268498992755e-05, "loss": 0.0545, "step": 22400 }, { "grad_norm": 0.3381105661392212, "learning_rate": 8.620843470133851e-05, "loss": 0.0668, "step": 22410 }, { "grad_norm": 0.9826821684837341, "learning_rate": 8.619417822571829e-05, "loss": 0.0512, "step": 22420 }, { "grad_norm": 0.5898223519325256, "learning_rate": 8.617991556550292e-05, "loss": 0.0529, "step": 22430 }, { "grad_norm": 0.5701519846916199, "learning_rate": 8.616564672312952e-05, "loss": 0.054, "step": 22440 }, { "grad_norm": 0.467218816280365, "learning_rate": 8.61513717010362e-05, "loss": 0.0562, "step": 22450 }, { "grad_norm": 0.45456957817077637, "learning_rate": 8.613709050166221e-05, "loss": 0.0498, "step": 22460 }, { "grad_norm": 0.7020622491836548, "learning_rate": 8.61228031274478e-05, "loss": 0.0535, "step": 22470 }, { "grad_norm": 0.6305111646652222, "learning_rate": 8.610850958083431e-05, "loss": 0.0485, "step": 22480 }, { "grad_norm": 0.569119930267334, "learning_rate": 8.609420986426409e-05, "loss": 0.0457, "step": 22490 }, { "grad_norm": 0.5033904314041138, "learning_rate": 8.60799039801806e-05, "loss": 0.048, "step": 22500 }, { "grad_norm": 0.644302487373352, "learning_rate": 8.606559193102828e-05, "loss": 0.0438, "step": 22510 }, { "grad_norm": 0.37093767523765564, "learning_rate": 8.605127371925273e-05, "loss": 0.0618, "step": 22520 }, { "grad_norm": 0.6836467385292053, "learning_rate": 8.603694934730047e-05, "loss": 0.0533, "step": 22530 }, { "grad_norm": 0.43828657269477844, "learning_rate": 8.602261881761919e-05, "loss": 0.0481, "step": 22540 }, { "grad_norm": 0.5695018172264099, "learning_rate": 8.600828213265759e-05, "loss": 0.0582, "step": 22550 }, { "grad_norm": 0.5362601280212402, "learning_rate": 8.599393929486539e-05, "loss": 0.0635, "step": 22560 }, { "grad_norm": 0.4012851417064667, "learning_rate": 8.59795903066934e-05, "loss": 0.0467, "step": 22570 }, { "grad_norm": 0.2966401278972626, "learning_rate": 8.596523517059347e-05, "loss": 0.0407, "step": 22580 }, { "grad_norm": 0.4145204424858093, "learning_rate": 8.59508738890185e-05, "loss": 0.0366, "step": 22590 }, { "grad_norm": 0.44798970222473145, "learning_rate": 8.593650646442246e-05, "loss": 0.0426, "step": 22600 }, { "grad_norm": 0.46058839559555054, "learning_rate": 8.59221328992603e-05, "loss": 0.0459, "step": 22610 }, { "grad_norm": 0.6483567357063293, "learning_rate": 8.590775319598813e-05, "loss": 0.0476, "step": 22620 }, { "grad_norm": 0.4939172565937042, "learning_rate": 8.589336735706301e-05, "loss": 0.0473, "step": 22630 }, { "grad_norm": 0.4298166036605835, "learning_rate": 8.587897538494307e-05, "loss": 0.0574, "step": 22640 }, { "grad_norm": 0.4038276672363281, "learning_rate": 8.586457728208756e-05, "loss": 0.0525, "step": 22650 }, { "grad_norm": 0.3814943730831146, "learning_rate": 8.585017305095667e-05, "loss": 0.0428, "step": 22660 }, { "grad_norm": 0.40636390447616577, "learning_rate": 8.583576269401173e-05, "loss": 0.0482, "step": 22670 }, { "grad_norm": 0.46137475967407227, "learning_rate": 8.582134621371504e-05, "loss": 0.0618, "step": 22680 }, { "grad_norm": 0.3661853075027466, "learning_rate": 8.580692361253e-05, "loss": 0.0447, "step": 22690 }, { "grad_norm": 0.4649789035320282, "learning_rate": 8.579249489292104e-05, "loss": 0.0519, "step": 22700 }, { "grad_norm": 0.47907841205596924, "learning_rate": 8.577806005735363e-05, "loss": 0.0631, "step": 22710 }, { "grad_norm": 0.4633365571498871, "learning_rate": 8.576361910829429e-05, "loss": 0.0449, "step": 22720 }, { "grad_norm": 0.47501808404922485, "learning_rate": 8.574917204821057e-05, "loss": 0.0463, "step": 22730 }, { "grad_norm": 0.6010297536849976, "learning_rate": 8.57347188795711e-05, "loss": 0.0489, "step": 22740 }, { "grad_norm": 0.43322286009788513, "learning_rate": 8.572025960484551e-05, "loss": 0.0611, "step": 22750 }, { "grad_norm": 0.3798511326313019, "learning_rate": 8.57057942265045e-05, "loss": 0.0558, "step": 22760 }, { "grad_norm": 0.5338422060012817, "learning_rate": 8.569132274701984e-05, "loss": 0.0529, "step": 22770 }, { "grad_norm": 0.3469119668006897, "learning_rate": 8.567684516886427e-05, "loss": 0.0386, "step": 22780 }, { "grad_norm": 0.4271344840526581, "learning_rate": 8.56623614945116e-05, "loss": 0.0408, "step": 22790 }, { "grad_norm": 0.34897324442863464, "learning_rate": 8.564787172643675e-05, "loss": 0.0422, "step": 22800 }, { "grad_norm": 0.45960405468940735, "learning_rate": 8.563337586711559e-05, "loss": 0.0437, "step": 22810 }, { "grad_norm": 0.4458675980567932, "learning_rate": 8.561887391902506e-05, "loss": 0.0539, "step": 22820 }, { "grad_norm": 0.6416196823120117, "learning_rate": 8.560436588464316e-05, "loss": 0.0406, "step": 22830 }, { "grad_norm": 0.35313230752944946, "learning_rate": 8.55898517664489e-05, "loss": 0.0537, "step": 22840 }, { "grad_norm": 0.526464581489563, "learning_rate": 8.557533156692236e-05, "loss": 0.0674, "step": 22850 }, { "grad_norm": 0.37832334637641907, "learning_rate": 8.556080528854467e-05, "loss": 0.042, "step": 22860 }, { "grad_norm": 0.5253424644470215, "learning_rate": 8.554627293379791e-05, "loss": 0.0462, "step": 22870 }, { "grad_norm": 0.3895505368709564, "learning_rate": 8.553173450516531e-05, "loss": 0.0513, "step": 22880 }, { "grad_norm": 0.5881906151771545, "learning_rate": 8.551719000513108e-05, "loss": 0.0508, "step": 22890 }, { "grad_norm": 0.39079299569129944, "learning_rate": 8.550263943618049e-05, "loss": 0.0615, "step": 22900 }, { "grad_norm": 0.37506985664367676, "learning_rate": 8.54880828007998e-05, "loss": 0.0453, "step": 22910 }, { "grad_norm": 0.7433834671974182, "learning_rate": 8.547352010147637e-05, "loss": 0.0522, "step": 22920 }, { "grad_norm": 0.41977423429489136, "learning_rate": 8.545895134069855e-05, "loss": 0.0437, "step": 22930 }, { "grad_norm": 0.3777415156364441, "learning_rate": 8.544437652095576e-05, "loss": 0.0432, "step": 22940 }, { "grad_norm": 0.44752857089042664, "learning_rate": 8.542979564473843e-05, "loss": 0.0534, "step": 22950 }, { "grad_norm": 0.5896140336990356, "learning_rate": 8.541520871453802e-05, "loss": 0.0439, "step": 22960 }, { "grad_norm": 0.42283400893211365, "learning_rate": 8.540061573284705e-05, "loss": 0.0374, "step": 22970 }, { "grad_norm": 0.5735599398612976, "learning_rate": 8.538601670215906e-05, "loss": 0.0417, "step": 22980 }, { "grad_norm": 0.5148605704307556, "learning_rate": 8.537141162496864e-05, "loss": 0.0553, "step": 22990 }, { "grad_norm": 0.5806919932365417, "learning_rate": 8.535680050377137e-05, "loss": 0.043, "step": 23000 }, { "grad_norm": 0.4538303017616272, "learning_rate": 8.534218334106391e-05, "loss": 0.0444, "step": 23010 }, { "grad_norm": 0.30027449131011963, "learning_rate": 8.532756013934393e-05, "loss": 0.0465, "step": 23020 }, { "grad_norm": 1.057640552520752, "learning_rate": 8.531293090111012e-05, "loss": 0.0519, "step": 23030 }, { "grad_norm": 0.4252682328224182, "learning_rate": 8.529829562886225e-05, "loss": 0.0501, "step": 23040 }, { "grad_norm": 0.3924197554588318, "learning_rate": 8.528365432510105e-05, "loss": 0.0624, "step": 23050 }, { "grad_norm": 0.554050862789154, "learning_rate": 8.526900699232833e-05, "loss": 0.0535, "step": 23060 }, { "grad_norm": 0.6304863691329956, "learning_rate": 8.525435363304695e-05, "loss": 0.0493, "step": 23070 }, { "grad_norm": 0.4499381482601166, "learning_rate": 8.523969424976072e-05, "loss": 0.0508, "step": 23080 }, { "grad_norm": 0.6736518144607544, "learning_rate": 8.522502884497457e-05, "loss": 0.0481, "step": 23090 }, { "grad_norm": 0.5181863903999329, "learning_rate": 8.521035742119437e-05, "loss": 0.0485, "step": 23100 }, { "grad_norm": 0.4286479353904724, "learning_rate": 8.519567998092712e-05, "loss": 0.0538, "step": 23110 }, { "grad_norm": 0.3291845917701721, "learning_rate": 8.518099652668075e-05, "loss": 0.0486, "step": 23120 }, { "grad_norm": 0.33987951278686523, "learning_rate": 8.516630706096429e-05, "loss": 0.048, "step": 23130 }, { "grad_norm": 0.4460534155368805, "learning_rate": 8.515161158628773e-05, "loss": 0.042, "step": 23140 }, { "grad_norm": 0.46620070934295654, "learning_rate": 8.513691010516216e-05, "loss": 0.0693, "step": 23150 }, { "grad_norm": 0.8356783986091614, "learning_rate": 8.512220262009966e-05, "loss": 0.043, "step": 23160 }, { "grad_norm": 0.4094713628292084, "learning_rate": 8.510748913361332e-05, "loss": 0.0489, "step": 23170 }, { "grad_norm": 0.39939382672309875, "learning_rate": 8.509276964821726e-05, "loss": 0.0423, "step": 23180 }, { "grad_norm": 1.9265556335449219, "learning_rate": 8.507804416642669e-05, "loss": 0.0559, "step": 23190 }, { "grad_norm": 0.5177720189094543, "learning_rate": 8.506331269075774e-05, "loss": 0.0529, "step": 23200 }, { "grad_norm": 0.42422738671302795, "learning_rate": 8.504857522372765e-05, "loss": 0.0532, "step": 23210 }, { "grad_norm": 0.43318209052085876, "learning_rate": 8.503383176785461e-05, "loss": 0.0468, "step": 23220 }, { "grad_norm": 0.7278298139572144, "learning_rate": 8.501908232565792e-05, "loss": 0.054, "step": 23230 }, { "grad_norm": 0.4312978982925415, "learning_rate": 8.50043268996578e-05, "loss": 0.0551, "step": 23240 }, { "grad_norm": 0.47116827964782715, "learning_rate": 8.498956549237562e-05, "loss": 0.0482, "step": 23250 }, { "grad_norm": 0.6514401435852051, "learning_rate": 8.497479810633366e-05, "loss": 0.0534, "step": 23260 }, { "grad_norm": 0.6363171935081482, "learning_rate": 8.496002474405525e-05, "loss": 0.0512, "step": 23270 }, { "grad_norm": 0.4447367191314697, "learning_rate": 8.494524540806478e-05, "loss": 0.0525, "step": 23280 }, { "grad_norm": 0.4721403121948242, "learning_rate": 8.493046010088761e-05, "loss": 0.0519, "step": 23290 }, { "grad_norm": 0.5177345275878906, "learning_rate": 8.491566882505018e-05, "loss": 0.0547, "step": 23300 }, { "grad_norm": 0.3690294027328491, "learning_rate": 8.490087158307988e-05, "loss": 0.0512, "step": 23310 }, { "grad_norm": 0.5837888121604919, "learning_rate": 8.488606837750518e-05, "loss": 0.0451, "step": 23320 }, { "grad_norm": 0.6416813135147095, "learning_rate": 8.487125921085552e-05, "loss": 0.0486, "step": 23330 }, { "grad_norm": 0.42129969596862793, "learning_rate": 8.485644408566141e-05, "loss": 0.0463, "step": 23340 }, { "grad_norm": 0.7641489505767822, "learning_rate": 8.484162300445431e-05, "loss": 0.0522, "step": 23350 }, { "grad_norm": 0.5225024819374084, "learning_rate": 8.482679596976676e-05, "loss": 0.0606, "step": 23360 }, { "grad_norm": 0.5288516879081726, "learning_rate": 8.48119629841323e-05, "loss": 0.0583, "step": 23370 }, { "grad_norm": 0.4893631935119629, "learning_rate": 8.479712405008547e-05, "loss": 0.0586, "step": 23380 }, { "grad_norm": 0.7369239330291748, "learning_rate": 8.478227917016184e-05, "loss": 0.0552, "step": 23390 }, { "grad_norm": 0.4158259928226471, "learning_rate": 8.476742834689801e-05, "loss": 0.0604, "step": 23400 }, { "grad_norm": 0.536973774433136, "learning_rate": 8.475257158283157e-05, "loss": 0.0568, "step": 23410 }, { "grad_norm": 0.4626109302043915, "learning_rate": 8.473770888050112e-05, "loss": 0.0475, "step": 23420 }, { "grad_norm": 0.431330144405365, "learning_rate": 8.47228402424463e-05, "loss": 0.0569, "step": 23430 }, { "grad_norm": 0.645388662815094, "learning_rate": 8.470796567120775e-05, "loss": 0.0493, "step": 23440 }, { "grad_norm": 0.706895112991333, "learning_rate": 8.469308516932714e-05, "loss": 0.0595, "step": 23450 }, { "grad_norm": 0.41696658730506897, "learning_rate": 8.467819873934714e-05, "loss": 0.0434, "step": 23460 }, { "grad_norm": 0.4647001326084137, "learning_rate": 8.466330638381143e-05, "loss": 0.0443, "step": 23470 }, { "grad_norm": 0.44866126775741577, "learning_rate": 8.464840810526469e-05, "loss": 0.0528, "step": 23480 }, { "grad_norm": 0.3664382994174957, "learning_rate": 8.463350390625264e-05, "loss": 0.0668, "step": 23490 }, { "grad_norm": 0.8514575362205505, "learning_rate": 8.4618593789322e-05, "loss": 0.0656, "step": 23500 }, { "grad_norm": 0.4397822618484497, "learning_rate": 8.46036777570205e-05, "loss": 0.0468, "step": 23510 }, { "grad_norm": 0.43234145641326904, "learning_rate": 8.458875581189688e-05, "loss": 0.0797, "step": 23520 }, { "grad_norm": 0.45033010840415955, "learning_rate": 8.457382795650092e-05, "loss": 0.0458, "step": 23530 }, { "grad_norm": 0.4181717336177826, "learning_rate": 8.455889419338335e-05, "loss": 0.0552, "step": 23540 }, { "grad_norm": 0.5533812046051025, "learning_rate": 8.454395452509593e-05, "loss": 0.0437, "step": 23550 }, { "grad_norm": 0.4660018980503082, "learning_rate": 8.452900895419146e-05, "loss": 0.0474, "step": 23560 }, { "grad_norm": 0.5697498917579651, "learning_rate": 8.451405748322376e-05, "loss": 0.0473, "step": 23570 }, { "grad_norm": 0.3547360301017761, "learning_rate": 8.449910011474759e-05, "loss": 0.044, "step": 23580 }, { "grad_norm": 0.3154691755771637, "learning_rate": 8.448413685131876e-05, "loss": 0.0536, "step": 23590 }, { "grad_norm": 0.4522012770175934, "learning_rate": 8.446916769549407e-05, "loss": 0.0477, "step": 23600 }, { "grad_norm": 0.3311137557029724, "learning_rate": 8.445419264983136e-05, "loss": 0.0483, "step": 23610 }, { "grad_norm": 0.3619922697544098, "learning_rate": 8.443921171688947e-05, "loss": 0.0607, "step": 23620 }, { "grad_norm": 0.35141047835350037, "learning_rate": 8.442422489922819e-05, "loss": 0.0454, "step": 23630 }, { "grad_norm": 0.5090125799179077, "learning_rate": 8.440923219940838e-05, "loss": 0.061, "step": 23640 }, { "grad_norm": 0.35505279898643494, "learning_rate": 8.439423361999189e-05, "loss": 0.0478, "step": 23650 }, { "grad_norm": 0.6353744864463806, "learning_rate": 8.437922916354155e-05, "loss": 0.0519, "step": 23660 }, { "grad_norm": 0.38189953565597534, "learning_rate": 8.436421883262123e-05, "loss": 0.0519, "step": 23670 }, { "grad_norm": 0.3947927951812744, "learning_rate": 8.434920262979577e-05, "loss": 0.0618, "step": 23680 }, { "grad_norm": 1.0776675939559937, "learning_rate": 8.433418055763104e-05, "loss": 0.0432, "step": 23690 }, { "grad_norm": 0.5008506774902344, "learning_rate": 8.431915261869389e-05, "loss": 0.062, "step": 23700 }, { "grad_norm": 0.4638838768005371, "learning_rate": 8.43041188155522e-05, "loss": 0.048, "step": 23710 }, { "grad_norm": 0.4050893783569336, "learning_rate": 8.428907915077481e-05, "loss": 0.058, "step": 23720 }, { "grad_norm": 0.5323772430419922, "learning_rate": 8.42740336269316e-05, "loss": 0.0603, "step": 23730 }, { "grad_norm": 0.5796324610710144, "learning_rate": 8.425898224659345e-05, "loss": 0.0398, "step": 23740 }, { "grad_norm": 0.35752803087234497, "learning_rate": 8.42439250123322e-05, "loss": 0.0412, "step": 23750 }, { "grad_norm": 0.5539013147354126, "learning_rate": 8.422886192672076e-05, "loss": 0.0492, "step": 23760 }, { "grad_norm": 0.592237651348114, "learning_rate": 8.421379299233297e-05, "loss": 0.0466, "step": 23770 }, { "grad_norm": 0.45751675963401794, "learning_rate": 8.419871821174371e-05, "loss": 0.0501, "step": 23780 }, { "grad_norm": 0.38749057054519653, "learning_rate": 8.418363758752884e-05, "loss": 0.0465, "step": 23790 }, { "grad_norm": 0.39561083912849426, "learning_rate": 8.416855112226523e-05, "loss": 0.0477, "step": 23800 }, { "grad_norm": 0.4529789686203003, "learning_rate": 8.415345881853075e-05, "loss": 0.05, "step": 23810 }, { "grad_norm": 0.7881451845169067, "learning_rate": 8.413836067890426e-05, "loss": 0.0462, "step": 23820 }, { "grad_norm": 0.315940260887146, "learning_rate": 8.41232567059656e-05, "loss": 0.0469, "step": 23830 }, { "grad_norm": 0.5024480223655701, "learning_rate": 8.410814690229565e-05, "loss": 0.0509, "step": 23840 }, { "grad_norm": 0.31455573439598083, "learning_rate": 8.409303127047626e-05, "loss": 0.0375, "step": 23850 }, { "grad_norm": 0.2995401620864868, "learning_rate": 8.407790981309028e-05, "loss": 0.0369, "step": 23860 }, { "grad_norm": 0.4537748098373413, "learning_rate": 8.406278253272153e-05, "loss": 0.0531, "step": 23870 }, { "grad_norm": 0.44536569714546204, "learning_rate": 8.404764943195487e-05, "loss": 0.0386, "step": 23880 }, { "grad_norm": 0.41171425580978394, "learning_rate": 8.403251051337613e-05, "loss": 0.0492, "step": 23890 }, { "grad_norm": 0.3841175436973572, "learning_rate": 8.401736577957214e-05, "loss": 0.0351, "step": 23900 }, { "grad_norm": 0.5211271643638611, "learning_rate": 8.40022152331307e-05, "loss": 0.0512, "step": 23910 }, { "grad_norm": 0.8623242974281311, "learning_rate": 8.398705887664064e-05, "loss": 0.0506, "step": 23920 }, { "grad_norm": 0.3999151885509491, "learning_rate": 8.397189671269177e-05, "loss": 0.0426, "step": 23930 }, { "grad_norm": 0.36264798045158386, "learning_rate": 8.395672874387488e-05, "loss": 0.0367, "step": 23940 }, { "grad_norm": 0.35512202978134155, "learning_rate": 8.394155497278177e-05, "loss": 0.036, "step": 23950 }, { "grad_norm": 0.5221197009086609, "learning_rate": 8.392637540200523e-05, "loss": 0.0673, "step": 23960 }, { "grad_norm": 0.36490893363952637, "learning_rate": 8.391119003413902e-05, "loss": 0.0411, "step": 23970 }, { "grad_norm": 0.3201865255832672, "learning_rate": 8.38959988717779e-05, "loss": 0.0448, "step": 23980 }, { "grad_norm": 0.39716270565986633, "learning_rate": 8.388080191751764e-05, "loss": 0.0459, "step": 23990 }, { "grad_norm": 0.4316622018814087, "learning_rate": 8.386559917395496e-05, "loss": 0.0339, "step": 24000 }, { "grad_norm": 0.2938394844532013, "learning_rate": 8.385039064368761e-05, "loss": 0.0372, "step": 24010 }, { "grad_norm": 0.3897663950920105, "learning_rate": 8.383517632931431e-05, "loss": 0.0436, "step": 24020 }, { "grad_norm": 0.615912139415741, "learning_rate": 8.381995623343477e-05, "loss": 0.0455, "step": 24030 }, { "grad_norm": 0.3697468042373657, "learning_rate": 8.380473035864968e-05, "loss": 0.0458, "step": 24040 }, { "grad_norm": 0.6855111122131348, "learning_rate": 8.378949870756076e-05, "loss": 0.055, "step": 24050 }, { "grad_norm": 0.5697231292724609, "learning_rate": 8.377426128277063e-05, "loss": 0.0547, "step": 24060 }, { "grad_norm": 0.35627758502960205, "learning_rate": 8.375901808688298e-05, "loss": 0.0349, "step": 24070 }, { "grad_norm": 0.5553974509239197, "learning_rate": 8.374376912250246e-05, "loss": 0.0502, "step": 24080 }, { "grad_norm": 0.30487751960754395, "learning_rate": 8.372851439223468e-05, "loss": 0.0407, "step": 24090 }, { "grad_norm": 0.3773460388183594, "learning_rate": 8.371325389868627e-05, "loss": 0.0481, "step": 24100 }, { "grad_norm": 0.39281317591667175, "learning_rate": 8.369798764446482e-05, "loss": 0.0435, "step": 24110 }, { "grad_norm": 0.39260515570640564, "learning_rate": 8.368271563217893e-05, "loss": 0.0418, "step": 24120 }, { "grad_norm": 0.48057761788368225, "learning_rate": 8.366743786443817e-05, "loss": 0.0603, "step": 24130 }, { "grad_norm": 0.37214186787605286, "learning_rate": 8.365215434385309e-05, "loss": 0.0557, "step": 24140 }, { "grad_norm": 0.4361361563205719, "learning_rate": 8.36368650730352e-05, "loss": 0.0491, "step": 24150 }, { "grad_norm": 0.7700260877609253, "learning_rate": 8.362157005459705e-05, "loss": 0.0529, "step": 24160 }, { "grad_norm": 0.40999311208724976, "learning_rate": 8.360626929115213e-05, "loss": 0.0446, "step": 24170 }, { "grad_norm": 0.37307584285736084, "learning_rate": 8.359096278531492e-05, "loss": 0.0399, "step": 24180 }, { "grad_norm": 0.7051408886909485, "learning_rate": 8.357565053970088e-05, "loss": 0.0471, "step": 24190 }, { "grad_norm": 0.570364773273468, "learning_rate": 8.356033255692647e-05, "loss": 0.0373, "step": 24200 }, { "grad_norm": 0.5112131237983704, "learning_rate": 8.354500883960911e-05, "loss": 0.0419, "step": 24210 }, { "grad_norm": 0.5756241679191589, "learning_rate": 8.352967939036717e-05, "loss": 0.0569, "step": 24220 }, { "grad_norm": 0.4048575460910797, "learning_rate": 8.35143442118201e-05, "loss": 0.0523, "step": 24230 }, { "grad_norm": 0.42769283056259155, "learning_rate": 8.349900330658819e-05, "loss": 0.0523, "step": 24240 }, { "grad_norm": 0.45384982228279114, "learning_rate": 8.348365667729284e-05, "loss": 0.0525, "step": 24250 }, { "grad_norm": 0.3923954367637634, "learning_rate": 8.346830432655633e-05, "loss": 0.0478, "step": 24260 }, { "grad_norm": 0.39565902948379517, "learning_rate": 8.345294625700195e-05, "loss": 0.0464, "step": 24270 }, { "grad_norm": 0.975391685962677, "learning_rate": 8.343758247125402e-05, "loss": 0.0516, "step": 24280 }, { "grad_norm": 0.4679974913597107, "learning_rate": 8.342221297193776e-05, "loss": 0.0569, "step": 24290 }, { "grad_norm": 0.8500255346298218, "learning_rate": 8.34068377616794e-05, "loss": 0.0557, "step": 24300 }, { "grad_norm": 0.6053717732429504, "learning_rate": 8.339145684310615e-05, "loss": 0.0506, "step": 24310 }, { "grad_norm": 0.4373026192188263, "learning_rate": 8.337607021884618e-05, "loss": 0.0434, "step": 24320 }, { "grad_norm": 0.32709068059921265, "learning_rate": 8.336067789152867e-05, "loss": 0.0378, "step": 24330 }, { "grad_norm": 0.33279746770858765, "learning_rate": 8.334527986378369e-05, "loss": 0.0473, "step": 24340 }, { "grad_norm": 0.33600351214408875, "learning_rate": 8.332987613824239e-05, "loss": 0.0499, "step": 24350 }, { "grad_norm": 0.3366025388240814, "learning_rate": 8.331446671753685e-05, "loss": 0.0426, "step": 24360 }, { "grad_norm": 0.3878895044326782, "learning_rate": 8.329905160430007e-05, "loss": 0.0409, "step": 24370 }, { "grad_norm": 0.4381489157676697, "learning_rate": 8.328363080116611e-05, "loss": 0.0402, "step": 24380 }, { "grad_norm": 0.6928077340126038, "learning_rate": 8.326820431076997e-05, "loss": 0.048, "step": 24390 }, { "grad_norm": 0.5937249064445496, "learning_rate": 8.325277213574759e-05, "loss": 0.0502, "step": 24400 }, { "grad_norm": 0.40691307187080383, "learning_rate": 8.32373342787359e-05, "loss": 0.0459, "step": 24410 }, { "grad_norm": 0.6026660203933716, "learning_rate": 8.322189074237285e-05, "loss": 0.0432, "step": 24420 }, { "grad_norm": 0.7363903522491455, "learning_rate": 8.32064415292973e-05, "loss": 0.0475, "step": 24430 }, { "grad_norm": 0.41051116585731506, "learning_rate": 8.319098664214907e-05, "loss": 0.0363, "step": 24440 }, { "grad_norm": 0.5713452696800232, "learning_rate": 8.3175526083569e-05, "loss": 0.047, "step": 24450 }, { "grad_norm": 0.5116363763809204, "learning_rate": 8.316005985619889e-05, "loss": 0.0457, "step": 24460 }, { "grad_norm": 0.456361323595047, "learning_rate": 8.314458796268147e-05, "loss": 0.0446, "step": 24470 }, { "grad_norm": 0.6291637420654297, "learning_rate": 8.312911040566047e-05, "loss": 0.0482, "step": 24480 }, { "grad_norm": 0.342752069234848, "learning_rate": 8.31136271877806e-05, "loss": 0.0444, "step": 24490 }, { "grad_norm": 0.4596729576587677, "learning_rate": 8.309813831168748e-05, "loss": 0.0508, "step": 24500 }, { "grad_norm": 0.5649973750114441, "learning_rate": 8.308264378002777e-05, "loss": 0.038, "step": 24510 }, { "grad_norm": 0.3899560272693634, "learning_rate": 8.306714359544906e-05, "loss": 0.0562, "step": 24520 }, { "grad_norm": 0.4080280065536499, "learning_rate": 8.30516377605999e-05, "loss": 0.0421, "step": 24530 }, { "grad_norm": 0.38401737809181213, "learning_rate": 8.30361262781298e-05, "loss": 0.0478, "step": 24540 }, { "grad_norm": 0.39131537079811096, "learning_rate": 8.302060915068924e-05, "loss": 0.0481, "step": 24550 }, { "grad_norm": 0.37318313121795654, "learning_rate": 8.300508638092972e-05, "loss": 0.0356, "step": 24560 }, { "grad_norm": 0.31710413098335266, "learning_rate": 8.298955797150361e-05, "loss": 0.0412, "step": 24570 }, { "grad_norm": 0.4319509267807007, "learning_rate": 8.297402392506433e-05, "loss": 0.0396, "step": 24580 }, { "grad_norm": 0.3614477515220642, "learning_rate": 8.295848424426617e-05, "loss": 0.0383, "step": 24590 }, { "grad_norm": 0.38693657517433167, "learning_rate": 8.29429389317645e-05, "loss": 0.0463, "step": 24600 }, { "grad_norm": 0.3215349614620209, "learning_rate": 8.292738799021556e-05, "loss": 0.0519, "step": 24610 }, { "grad_norm": 0.41306373476982117, "learning_rate": 8.291183142227656e-05, "loss": 0.0434, "step": 24620 }, { "grad_norm": 0.40409448742866516, "learning_rate": 8.289626923060572e-05, "loss": 0.0519, "step": 24630 }, { "grad_norm": 0.4950672686100006, "learning_rate": 8.288070141786218e-05, "loss": 0.0431, "step": 24640 }, { "grad_norm": 0.5155262351036072, "learning_rate": 8.286512798670605e-05, "loss": 0.0515, "step": 24650 }, { "grad_norm": 0.4078425467014313, "learning_rate": 8.284954893979842e-05, "loss": 0.0485, "step": 24660 }, { "grad_norm": 0.3313245177268982, "learning_rate": 8.283396427980131e-05, "loss": 0.0468, "step": 24670 }, { "grad_norm": 0.5020387768745422, "learning_rate": 8.281837400937771e-05, "loss": 0.0479, "step": 24680 }, { "grad_norm": 0.4554130733013153, "learning_rate": 8.28027781311916e-05, "loss": 0.0469, "step": 24690 }, { "grad_norm": 0.46463924646377563, "learning_rate": 8.278717664790785e-05, "loss": 0.0573, "step": 24700 }, { "grad_norm": 0.5845319032669067, "learning_rate": 8.277156956219234e-05, "loss": 0.0494, "step": 24710 }, { "grad_norm": 0.4387023448944092, "learning_rate": 8.275595687671189e-05, "loss": 0.0484, "step": 24720 }, { "grad_norm": 0.3859962821006775, "learning_rate": 8.27403385941343e-05, "loss": 0.0498, "step": 24730 }, { "grad_norm": 0.47817111015319824, "learning_rate": 8.272471471712828e-05, "loss": 0.0505, "step": 24740 }, { "grad_norm": 0.49405592679977417, "learning_rate": 8.270908524836355e-05, "loss": 0.0583, "step": 24750 }, { "grad_norm": 0.4820529818534851, "learning_rate": 8.269345019051074e-05, "loss": 0.0602, "step": 24760 }, { "grad_norm": 0.47249746322631836, "learning_rate": 8.267780954624147e-05, "loss": 0.0512, "step": 24770 }, { "grad_norm": 0.5145324468612671, "learning_rate": 8.266216331822827e-05, "loss": 0.0592, "step": 24780 }, { "grad_norm": 0.46088096499443054, "learning_rate": 8.264651150914469e-05, "loss": 0.0469, "step": 24790 }, { "grad_norm": 0.5396301746368408, "learning_rate": 8.263085412166517e-05, "loss": 0.0422, "step": 24800 }, { "grad_norm": 0.6089586615562439, "learning_rate": 8.261519115846514e-05, "loss": 0.05, "step": 24810 }, { "grad_norm": 0.4870002567768097, "learning_rate": 8.259952262222096e-05, "loss": 0.0538, "step": 24820 }, { "grad_norm": 0.27783849835395813, "learning_rate": 8.258384851560997e-05, "loss": 0.0479, "step": 24830 }, { "grad_norm": 0.5770269632339478, "learning_rate": 8.256816884131044e-05, "loss": 0.0522, "step": 24840 }, { "grad_norm": 0.514711856842041, "learning_rate": 8.255248360200159e-05, "loss": 0.0394, "step": 24850 }, { "grad_norm": 0.43654701113700867, "learning_rate": 8.253679280036359e-05, "loss": 0.0522, "step": 24860 }, { "grad_norm": 0.46336910128593445, "learning_rate": 8.252109643907762e-05, "loss": 0.0644, "step": 24870 }, { "grad_norm": 0.4372580647468567, "learning_rate": 8.250539452082569e-05, "loss": 0.0492, "step": 24880 }, { "grad_norm": 0.4227525293827057, "learning_rate": 8.248968704829087e-05, "loss": 0.0552, "step": 24890 }, { "grad_norm": 0.5116661190986633, "learning_rate": 8.247397402415714e-05, "loss": 0.0577, "step": 24900 }, { "grad_norm": 0.5076478719711304, "learning_rate": 8.24582554511094e-05, "loss": 0.0476, "step": 24910 }, { "grad_norm": 0.4275442361831665, "learning_rate": 8.244253133183355e-05, "loss": 0.0435, "step": 24920 }, { "grad_norm": 0.37857940793037415, "learning_rate": 8.24268016690164e-05, "loss": 0.0429, "step": 24930 }, { "grad_norm": 0.47314634919166565, "learning_rate": 8.241106646534571e-05, "loss": 0.0459, "step": 24940 }, { "grad_norm": 0.47632747888565063, "learning_rate": 8.23953257235102e-05, "loss": 0.0485, "step": 24950 }, { "grad_norm": 0.4337039291858673, "learning_rate": 8.237957944619956e-05, "loss": 0.042, "step": 24960 }, { "grad_norm": 0.4622589647769928, "learning_rate": 8.236382763610437e-05, "loss": 0.0554, "step": 24970 }, { "grad_norm": 0.3888557553291321, "learning_rate": 8.234807029591619e-05, "loss": 0.0455, "step": 24980 }, { "grad_norm": 0.676067590713501, "learning_rate": 8.233230742832752e-05, "loss": 0.0583, "step": 24990 }, { "grad_norm": 0.4326360821723938, "learning_rate": 8.231653903603178e-05, "loss": 0.0462, "step": 25000 }, { "grad_norm": 0.32549548149108887, "learning_rate": 8.23007651217234e-05, "loss": 0.044, "step": 25010 }, { "grad_norm": 0.5323083400726318, "learning_rate": 8.228498568809769e-05, "loss": 0.0449, "step": 25020 }, { "grad_norm": 0.3857411742210388, "learning_rate": 8.22692007378509e-05, "loss": 0.0543, "step": 25030 }, { "grad_norm": 0.48854362964630127, "learning_rate": 8.225341027368028e-05, "loss": 0.0466, "step": 25040 }, { "grad_norm": 0.3389459252357483, "learning_rate": 8.223761429828399e-05, "loss": 0.0425, "step": 25050 }, { "grad_norm": 0.4904734194278717, "learning_rate": 8.22218128143611e-05, "loss": 0.05, "step": 25060 }, { "grad_norm": 0.665497362613678, "learning_rate": 8.220600582461166e-05, "loss": 0.0495, "step": 25070 }, { "grad_norm": 0.92596834897995, "learning_rate": 8.219019333173668e-05, "loss": 0.0505, "step": 25080 }, { "grad_norm": 0.6235218644142151, "learning_rate": 8.217437533843805e-05, "loss": 0.0572, "step": 25090 }, { "grad_norm": 0.47228461503982544, "learning_rate": 8.215855184741867e-05, "loss": 0.0572, "step": 25100 }, { "grad_norm": 0.44748783111572266, "learning_rate": 8.21427228613823e-05, "loss": 0.0427, "step": 25110 }, { "grad_norm": 0.41701531410217285, "learning_rate": 8.21268883830337e-05, "loss": 0.0515, "step": 25120 }, { "grad_norm": 0.3418148159980774, "learning_rate": 8.211104841507855e-05, "loss": 0.0621, "step": 25130 }, { "grad_norm": 0.40020647644996643, "learning_rate": 8.209520296022346e-05, "loss": 0.0523, "step": 25140 }, { "grad_norm": 1.7655150890350342, "learning_rate": 8.207935202117599e-05, "loss": 0.0517, "step": 25150 }, { "grad_norm": 0.5505461692810059, "learning_rate": 8.206349560064463e-05, "loss": 0.0481, "step": 25160 }, { "grad_norm": 0.4272196888923645, "learning_rate": 8.204763370133881e-05, "loss": 0.0469, "step": 25170 }, { "grad_norm": 0.8607397079467773, "learning_rate": 8.203176632596892e-05, "loss": 0.053, "step": 25180 }, { "grad_norm": 0.5395998954772949, "learning_rate": 8.20158934772462e-05, "loss": 0.0661, "step": 25190 }, { "grad_norm": 0.5213334560394287, "learning_rate": 8.200001515788294e-05, "loss": 0.0434, "step": 25200 }, { "grad_norm": 0.5054654479026794, "learning_rate": 8.198413137059228e-05, "loss": 0.048, "step": 25210 }, { "grad_norm": 0.4432279169559479, "learning_rate": 8.196824211808835e-05, "loss": 0.0408, "step": 25220 }, { "grad_norm": 0.38057461380958557, "learning_rate": 8.195234740308617e-05, "loss": 0.0466, "step": 25230 }, { "grad_norm": 0.3281056880950928, "learning_rate": 8.193644722830171e-05, "loss": 0.0412, "step": 25240 }, { "grad_norm": 0.425879567861557, "learning_rate": 8.19205415964519e-05, "loss": 0.0381, "step": 25250 }, { "grad_norm": 1.1379469633102417, "learning_rate": 8.190463051025456e-05, "loss": 0.0423, "step": 25260 }, { "grad_norm": 0.4718812108039856, "learning_rate": 8.188871397242843e-05, "loss": 0.0462, "step": 25270 }, { "grad_norm": 0.48418739438056946, "learning_rate": 8.187279198569326e-05, "loss": 0.0487, "step": 25280 }, { "grad_norm": 0.4155771732330322, "learning_rate": 8.185686455276966e-05, "loss": 0.0418, "step": 25290 }, { "grad_norm": 0.6064079999923706, "learning_rate": 8.184093167637921e-05, "loss": 0.0554, "step": 25300 }, { "grad_norm": 0.3905450403690338, "learning_rate": 8.182499335924437e-05, "loss": 0.0513, "step": 25310 }, { "grad_norm": 0.3716517686843872, "learning_rate": 8.18090496040886e-05, "loss": 0.0477, "step": 25320 }, { "grad_norm": 0.4569169282913208, "learning_rate": 8.179310041363621e-05, "loss": 0.0563, "step": 25330 }, { "grad_norm": 0.3360488712787628, "learning_rate": 8.17771457906125e-05, "loss": 0.041, "step": 25340 }, { "grad_norm": 0.5650113821029663, "learning_rate": 8.176118573774371e-05, "loss": 0.0434, "step": 25350 }, { "grad_norm": 0.5247783660888672, "learning_rate": 8.174522025775692e-05, "loss": 0.0481, "step": 25360 }, { "grad_norm": 0.40433812141418457, "learning_rate": 8.172924935338022e-05, "loss": 0.0479, "step": 25370 }, { "grad_norm": 0.6926612257957458, "learning_rate": 8.171327302734262e-05, "loss": 0.0375, "step": 25380 }, { "grad_norm": 0.6030600666999817, "learning_rate": 8.169729128237401e-05, "loss": 0.0414, "step": 25390 }, { "grad_norm": 0.374055951833725, "learning_rate": 8.168130412120525e-05, "loss": 0.0473, "step": 25400 }, { "grad_norm": 0.5026649832725525, "learning_rate": 8.16653115465681e-05, "loss": 0.0496, "step": 25410 }, { "grad_norm": 0.4384991526603699, "learning_rate": 8.164931356119526e-05, "loss": 0.0388, "step": 25420 }, { "grad_norm": 0.44884493947029114, "learning_rate": 8.163331016782032e-05, "loss": 0.0539, "step": 25430 }, { "grad_norm": 0.5555216670036316, "learning_rate": 8.161730136917785e-05, "loss": 0.0431, "step": 25440 }, { "grad_norm": 0.3187556564807892, "learning_rate": 8.160128716800333e-05, "loss": 0.0497, "step": 25450 }, { "grad_norm": 0.38292232155799866, "learning_rate": 8.158526756703313e-05, "loss": 0.038, "step": 25460 }, { "grad_norm": 0.5180206298828125, "learning_rate": 8.156924256900455e-05, "loss": 0.0427, "step": 25470 }, { "grad_norm": 0.4045247435569763, "learning_rate": 8.155321217665584e-05, "loss": 0.0433, "step": 25480 }, { "grad_norm": 0.31767386198043823, "learning_rate": 8.153717639272614e-05, "loss": 0.0449, "step": 25490 }, { "grad_norm": 0.45408618450164795, "learning_rate": 8.152113521995555e-05, "loss": 0.0521, "step": 25500 }, { "grad_norm": 0.41009965538978577, "learning_rate": 8.150508866108505e-05, "loss": 0.0446, "step": 25510 }, { "grad_norm": 0.44009265303611755, "learning_rate": 8.148903671885657e-05, "loss": 0.0396, "step": 25520 }, { "grad_norm": 0.4380711615085602, "learning_rate": 8.147297939601292e-05, "loss": 0.0474, "step": 25530 }, { "grad_norm": 0.48341935873031616, "learning_rate": 8.145691669529792e-05, "loss": 0.0469, "step": 25540 }, { "grad_norm": 0.3339202404022217, "learning_rate": 8.144084861945618e-05, "loss": 0.0506, "step": 25550 }, { "grad_norm": 0.3671453893184662, "learning_rate": 8.142477517123333e-05, "loss": 0.0471, "step": 25560 }, { "grad_norm": 0.4805566966533661, "learning_rate": 8.140869635337586e-05, "loss": 0.0495, "step": 25570 }, { "grad_norm": 0.37929534912109375, "learning_rate": 8.139261216863123e-05, "loss": 0.0405, "step": 25580 }, { "grad_norm": 0.3142184913158417, "learning_rate": 8.137652261974776e-05, "loss": 0.04, "step": 25590 }, { "grad_norm": 0.32005128264427185, "learning_rate": 8.136042770947472e-05, "loss": 0.0382, "step": 25600 }, { "grad_norm": 0.48460373282432556, "learning_rate": 8.134432744056228e-05, "loss": 0.0494, "step": 25610 }, { "grad_norm": 0.4563581347465515, "learning_rate": 8.132822181576158e-05, "loss": 0.0547, "step": 25620 }, { "grad_norm": 0.4679286777973175, "learning_rate": 8.131211083782459e-05, "loss": 0.0449, "step": 25630 }, { "grad_norm": 0.3793441355228424, "learning_rate": 8.129599450950424e-05, "loss": 0.0422, "step": 25640 }, { "grad_norm": 0.4079468846321106, "learning_rate": 8.127987283355438e-05, "loss": 0.0456, "step": 25650 }, { "grad_norm": 0.4168282747268677, "learning_rate": 8.126374581272976e-05, "loss": 0.0384, "step": 25660 }, { "grad_norm": 0.6537474393844604, "learning_rate": 8.124761344978605e-05, "loss": 0.065, "step": 25670 }, { "grad_norm": 0.4426495134830475, "learning_rate": 8.12314757474798e-05, "loss": 0.0528, "step": 25680 }, { "grad_norm": 0.35584163665771484, "learning_rate": 8.121533270856856e-05, "loss": 0.0499, "step": 25690 }, { "grad_norm": 0.4728916585445404, "learning_rate": 8.119918433581069e-05, "loss": 0.0423, "step": 25700 }, { "grad_norm": 0.4121393859386444, "learning_rate": 8.118303063196551e-05, "loss": 0.0482, "step": 25710 }, { "grad_norm": 0.7714487314224243, "learning_rate": 8.116687159979326e-05, "loss": 0.0432, "step": 25720 }, { "grad_norm": 0.5856634378433228, "learning_rate": 8.115070724205508e-05, "loss": 0.0607, "step": 25730 }, { "grad_norm": 0.4330069124698639, "learning_rate": 8.113453756151296e-05, "loss": 0.0433, "step": 25740 }, { "grad_norm": 0.37222158908843994, "learning_rate": 8.111836256092995e-05, "loss": 0.0403, "step": 25750 }, { "grad_norm": 0.30976903438568115, "learning_rate": 8.110218224306985e-05, "loss": 0.0373, "step": 25760 }, { "grad_norm": 0.39571622014045715, "learning_rate": 8.108599661069745e-05, "loss": 0.0467, "step": 25770 }, { "grad_norm": 0.40515050292015076, "learning_rate": 8.106980566657845e-05, "loss": 0.0356, "step": 25780 }, { "grad_norm": 0.4189046621322632, "learning_rate": 8.10536094134794e-05, "loss": 0.0422, "step": 25790 }, { "grad_norm": 0.5766568779945374, "learning_rate": 8.103740785416783e-05, "loss": 0.0477, "step": 25800 }, { "grad_norm": 1.2208805084228516, "learning_rate": 8.102120099141212e-05, "loss": 0.0515, "step": 25810 }, { "grad_norm": 0.4120160937309265, "learning_rate": 8.100498882798163e-05, "loss": 0.0399, "step": 25820 }, { "grad_norm": 0.587142825126648, "learning_rate": 8.09887713666465e-05, "loss": 0.043, "step": 25830 }, { "grad_norm": 0.3326663374900818, "learning_rate": 8.09725486101779e-05, "loss": 0.0556, "step": 25840 }, { "grad_norm": 0.5131575465202332, "learning_rate": 8.095632056134784e-05, "loss": 0.0438, "step": 25850 }, { "grad_norm": 0.3042997419834137, "learning_rate": 8.094008722292925e-05, "loss": 0.0423, "step": 25860 }, { "grad_norm": 0.37382057309150696, "learning_rate": 8.092384859769598e-05, "loss": 0.0432, "step": 25870 }, { "grad_norm": 0.3641442656517029, "learning_rate": 8.090760468842275e-05, "loss": 0.067, "step": 25880 }, { "grad_norm": 0.4460107982158661, "learning_rate": 8.089135549788521e-05, "loss": 0.045, "step": 25890 }, { "grad_norm": 0.6151760816574097, "learning_rate": 8.087510102885987e-05, "loss": 0.0369, "step": 25900 }, { "grad_norm": 0.36466702818870544, "learning_rate": 8.085884128412422e-05, "loss": 0.0566, "step": 25910 }, { "grad_norm": 0.6481053829193115, "learning_rate": 8.084257626645659e-05, "loss": 0.0521, "step": 25920 }, { "grad_norm": 0.5588774085044861, "learning_rate": 8.08263059786362e-05, "loss": 0.0416, "step": 25930 }, { "grad_norm": 0.3444460928440094, "learning_rate": 8.081003042344325e-05, "loss": 0.0415, "step": 25940 }, { "grad_norm": 0.43107205629348755, "learning_rate": 8.079374960365872e-05, "loss": 0.0521, "step": 25950 }, { "grad_norm": 0.3379327058792114, "learning_rate": 8.077746352206463e-05, "loss": 0.053, "step": 25960 }, { "grad_norm": 0.46659064292907715, "learning_rate": 8.076117218144377e-05, "loss": 0.0503, "step": 25970 }, { "grad_norm": 0.5229030847549438, "learning_rate": 8.074487558457991e-05, "loss": 0.0424, "step": 25980 }, { "grad_norm": 0.36138665676116943, "learning_rate": 8.072857373425768e-05, "loss": 0.0482, "step": 25990 }, { "grad_norm": 0.3602260649204254, "learning_rate": 8.071226663326264e-05, "loss": 0.0444, "step": 26000 }, { "grad_norm": 0.4869520962238312, "learning_rate": 8.069595428438121e-05, "loss": 0.0538, "step": 26010 }, { "grad_norm": 0.46533915400505066, "learning_rate": 8.067963669040072e-05, "loss": 0.0525, "step": 26020 }, { "grad_norm": 0.46324267983436584, "learning_rate": 8.066331385410942e-05, "loss": 0.0558, "step": 26030 }, { "grad_norm": 0.6942493319511414, "learning_rate": 8.064698577829641e-05, "loss": 0.0559, "step": 26040 }, { "grad_norm": 0.41417011618614197, "learning_rate": 8.063065246575175e-05, "loss": 0.045, "step": 26050 }, { "grad_norm": 0.345502644777298, "learning_rate": 8.061431391926631e-05, "loss": 0.0438, "step": 26060 }, { "grad_norm": 0.5515331029891968, "learning_rate": 8.059797014163195e-05, "loss": 0.0584, "step": 26070 }, { "grad_norm": 0.43378570675849915, "learning_rate": 8.058162113564133e-05, "loss": 0.0706, "step": 26080 }, { "grad_norm": 0.510964572429657, "learning_rate": 8.056526690408806e-05, "loss": 0.0579, "step": 26090 }, { "grad_norm": 0.7315894365310669, "learning_rate": 8.054890744976666e-05, "loss": 0.0555, "step": 26100 }, { "grad_norm": 0.48023244738578796, "learning_rate": 8.053254277547248e-05, "loss": 0.0482, "step": 26110 }, { "grad_norm": 0.3519739508628845, "learning_rate": 8.051617288400182e-05, "loss": 0.0435, "step": 26120 }, { "grad_norm": 0.611386775970459, "learning_rate": 8.049979777815182e-05, "loss": 0.0452, "step": 26130 }, { "grad_norm": 0.31452664732933044, "learning_rate": 8.048341746072054e-05, "loss": 0.0412, "step": 26140 }, { "grad_norm": 0.3690938651561737, "learning_rate": 8.046703193450696e-05, "loss": 0.048, "step": 26150 }, { "grad_norm": 0.33920466899871826, "learning_rate": 8.04506412023109e-05, "loss": 0.0382, "step": 26160 }, { "grad_norm": 0.4423552453517914, "learning_rate": 8.043424526693306e-05, "loss": 0.0406, "step": 26170 }, { "grad_norm": 0.3873532712459564, "learning_rate": 8.04178441311751e-05, "loss": 0.0506, "step": 26180 }, { "grad_norm": 0.5265120267868042, "learning_rate": 8.04014377978395e-05, "loss": 0.0482, "step": 26190 }, { "grad_norm": 0.39398208260536194, "learning_rate": 8.038502626972967e-05, "loss": 0.0495, "step": 26200 }, { "grad_norm": 0.5478465557098389, "learning_rate": 8.036860954964989e-05, "loss": 0.0452, "step": 26210 }, { "grad_norm": 0.3056565225124359, "learning_rate": 8.035218764040531e-05, "loss": 0.0448, "step": 26220 }, { "grad_norm": 0.41484588384628296, "learning_rate": 8.033576054480199e-05, "loss": 0.0442, "step": 26230 }, { "grad_norm": 0.5144966244697571, "learning_rate": 8.031932826564688e-05, "loss": 0.0452, "step": 26240 }, { "grad_norm": 1.073105812072754, "learning_rate": 8.030289080574782e-05, "loss": 0.0452, "step": 26250 }, { "grad_norm": 0.4298485219478607, "learning_rate": 8.028644816791349e-05, "loss": 0.0566, "step": 26260 }, { "grad_norm": 0.3374505937099457, "learning_rate": 8.027000035495351e-05, "loss": 0.0508, "step": 26270 }, { "grad_norm": 0.563184380531311, "learning_rate": 8.025354736967836e-05, "loss": 0.0594, "step": 26280 }, { "grad_norm": 0.35160669684410095, "learning_rate": 8.023708921489941e-05, "loss": 0.0612, "step": 26290 }, { "grad_norm": 0.45490536093711853, "learning_rate": 8.022062589342887e-05, "loss": 0.0481, "step": 26300 }, { "grad_norm": 0.3709096908569336, "learning_rate": 8.020415740807993e-05, "loss": 0.0555, "step": 26310 }, { "grad_norm": 0.5505017638206482, "learning_rate": 8.018768376166656e-05, "loss": 0.05, "step": 26320 }, { "grad_norm": 0.41147229075431824, "learning_rate": 8.017120495700368e-05, "loss": 0.0459, "step": 26330 }, { "grad_norm": 0.4230954647064209, "learning_rate": 8.015472099690704e-05, "loss": 0.0455, "step": 26340 }, { "grad_norm": 0.3425554931163788, "learning_rate": 8.013823188419332e-05, "loss": 0.04, "step": 26350 }, { "grad_norm": 0.39862504601478577, "learning_rate": 8.012173762168006e-05, "loss": 0.0455, "step": 26360 }, { "grad_norm": 0.49000003933906555, "learning_rate": 8.010523821218567e-05, "loss": 0.0426, "step": 26370 }, { "grad_norm": 0.4183892011642456, "learning_rate": 8.008873365852945e-05, "loss": 0.0413, "step": 26380 }, { "grad_norm": 0.4461885988712311, "learning_rate": 8.007222396353157e-05, "loss": 0.045, "step": 26390 }, { "grad_norm": 0.5751295685768127, "learning_rate": 8.00557091300131e-05, "loss": 0.0417, "step": 26400 }, { "grad_norm": 0.38095250725746155, "learning_rate": 8.003918916079597e-05, "loss": 0.0468, "step": 26410 }, { "grad_norm": 0.3590182662010193, "learning_rate": 8.002266405870298e-05, "loss": 0.0421, "step": 26420 }, { "grad_norm": 0.4047601819038391, "learning_rate": 8.000613382655782e-05, "loss": 0.0471, "step": 26430 }, { "grad_norm": 0.38752326369285583, "learning_rate": 7.998959846718505e-05, "loss": 0.0393, "step": 26440 }, { "grad_norm": 0.4820268154144287, "learning_rate": 7.997305798341012e-05, "loss": 0.0455, "step": 26450 }, { "grad_norm": 0.3464396893978119, "learning_rate": 7.995651237805937e-05, "loss": 0.0381, "step": 26460 }, { "grad_norm": 0.4056612551212311, "learning_rate": 7.993996165395996e-05, "loss": 0.0377, "step": 26470 }, { "grad_norm": 0.4407738745212555, "learning_rate": 7.992340581393996e-05, "loss": 0.0345, "step": 26480 }, { "grad_norm": 0.3394278287887573, "learning_rate": 7.990684486082831e-05, "loss": 0.0383, "step": 26490 }, { "grad_norm": 0.47234153747558594, "learning_rate": 7.989027879745482e-05, "loss": 0.0434, "step": 26500 }, { "grad_norm": 0.43655645847320557, "learning_rate": 7.98737076266502e-05, "loss": 0.0498, "step": 26510 }, { "grad_norm": 0.3588062524795532, "learning_rate": 7.985713135124598e-05, "loss": 0.0369, "step": 26520 }, { "grad_norm": 0.33891916275024414, "learning_rate": 7.98405499740746e-05, "loss": 0.0473, "step": 26530 }, { "grad_norm": 0.39001190662384033, "learning_rate": 7.98239634979694e-05, "loss": 0.0466, "step": 26540 }, { "grad_norm": 0.4586679935455322, "learning_rate": 7.98073719257645e-05, "loss": 0.0418, "step": 26550 }, { "grad_norm": 0.40764373540878296, "learning_rate": 7.979077526029499e-05, "loss": 0.0445, "step": 26560 }, { "grad_norm": 0.3412081301212311, "learning_rate": 7.977417350439675e-05, "loss": 0.0388, "step": 26570 }, { "grad_norm": 0.5091781616210938, "learning_rate": 7.97575666609066e-05, "loss": 0.0374, "step": 26580 }, { "grad_norm": 0.40852952003479004, "learning_rate": 7.974095473266216e-05, "loss": 0.0525, "step": 26590 }, { "grad_norm": 0.3742083013057709, "learning_rate": 7.972433772250198e-05, "loss": 0.0445, "step": 26600 }, { "grad_norm": 0.4989628493785858, "learning_rate": 7.970771563326544e-05, "loss": 0.0376, "step": 26610 }, { "grad_norm": 0.448097825050354, "learning_rate": 7.96910884677928e-05, "loss": 0.0443, "step": 26620 }, { "grad_norm": 0.7390444874763489, "learning_rate": 7.967445622892523e-05, "loss": 0.0468, "step": 26630 }, { "grad_norm": 0.5268713235855103, "learning_rate": 7.965781891950465e-05, "loss": 0.0428, "step": 26640 }, { "grad_norm": 0.39993661642074585, "learning_rate": 7.964117654237397e-05, "loss": 0.0403, "step": 26650 }, { "grad_norm": 0.5548281073570251, "learning_rate": 7.962452910037692e-05, "loss": 0.0578, "step": 26660 }, { "grad_norm": 0.44386935234069824, "learning_rate": 7.96078765963581e-05, "loss": 0.0377, "step": 26670 }, { "grad_norm": 0.469063401222229, "learning_rate": 7.95912190331629e-05, "loss": 0.0443, "step": 26680 }, { "grad_norm": 0.5333355665206909, "learning_rate": 7.957455641363772e-05, "loss": 0.0454, "step": 26690 }, { "grad_norm": 0.50777268409729, "learning_rate": 7.955788874062968e-05, "loss": 0.0536, "step": 26700 }, { "grad_norm": 0.615077018737793, "learning_rate": 7.95412160169869e-05, "loss": 0.0446, "step": 26710 }, { "grad_norm": 0.31619149446487427, "learning_rate": 7.952453824555824e-05, "loss": 0.0378, "step": 26720 }, { "grad_norm": 0.39039096236228943, "learning_rate": 7.95078554291935e-05, "loss": 0.0396, "step": 26730 }, { "grad_norm": 0.355471134185791, "learning_rate": 7.94911675707433e-05, "loss": 0.0369, "step": 26740 }, { "grad_norm": 0.3508443236351013, "learning_rate": 7.947447467305915e-05, "loss": 0.0528, "step": 26750 }, { "grad_norm": 1.0481091737747192, "learning_rate": 7.94577767389934e-05, "loss": 0.0493, "step": 26760 }, { "grad_norm": 0.35508590936660767, "learning_rate": 7.944107377139928e-05, "loss": 0.044, "step": 26770 }, { "grad_norm": 0.559756338596344, "learning_rate": 7.942436577313088e-05, "loss": 0.0472, "step": 26780 }, { "grad_norm": 0.6241933703422546, "learning_rate": 7.940765274704312e-05, "loss": 0.0423, "step": 26790 }, { "grad_norm": 0.3751288950443268, "learning_rate": 7.939093469599181e-05, "loss": 0.0484, "step": 26800 }, { "grad_norm": 0.4364127218723297, "learning_rate": 7.93742116228336e-05, "loss": 0.0467, "step": 26810 }, { "grad_norm": 0.5124316215515137, "learning_rate": 7.935748353042602e-05, "loss": 0.0541, "step": 26820 }, { "grad_norm": 0.3443808853626251, "learning_rate": 7.934075042162744e-05, "loss": 0.0425, "step": 26830 }, { "grad_norm": 0.5029804706573486, "learning_rate": 7.932401229929705e-05, "loss": 0.0496, "step": 26840 }, { "grad_norm": 0.5575343370437622, "learning_rate": 7.9307269166295e-05, "loss": 0.0452, "step": 26850 }, { "grad_norm": 0.4306146502494812, "learning_rate": 7.92905210254822e-05, "loss": 0.0545, "step": 26860 }, { "grad_norm": 0.3759489357471466, "learning_rate": 7.927376787972045e-05, "loss": 0.0483, "step": 26870 }, { "grad_norm": 0.5096621513366699, "learning_rate": 7.92570097318724e-05, "loss": 0.0472, "step": 26880 }, { "grad_norm": 2.3466899394989014, "learning_rate": 7.924024658480158e-05, "loss": 0.0489, "step": 26890 }, { "grad_norm": 0.4441997706890106, "learning_rate": 7.922347844137233e-05, "loss": 0.0431, "step": 26900 }, { "grad_norm": 0.43792107701301575, "learning_rate": 7.92067053044499e-05, "loss": 0.0517, "step": 26910 }, { "grad_norm": 0.4773937463760376, "learning_rate": 7.918992717690031e-05, "loss": 0.0538, "step": 26920 }, { "grad_norm": 0.9145843386650085, "learning_rate": 7.917314406159053e-05, "loss": 0.0443, "step": 26930 }, { "grad_norm": 0.5817670822143555, "learning_rate": 7.915635596138832e-05, "loss": 0.0666, "step": 26940 }, { "grad_norm": 0.3823172152042389, "learning_rate": 7.913956287916228e-05, "loss": 0.0402, "step": 26950 }, { "grad_norm": 0.43064844608306885, "learning_rate": 7.912276481778193e-05, "loss": 0.0481, "step": 26960 }, { "grad_norm": 0.3245788812637329, "learning_rate": 7.910596178011759e-05, "loss": 0.0453, "step": 26970 }, { "grad_norm": 0.44301292300224304, "learning_rate": 7.908915376904043e-05, "loss": 0.0466, "step": 26980 }, { "grad_norm": 0.3992122411727905, "learning_rate": 7.907234078742247e-05, "loss": 0.0516, "step": 26990 }, { "grad_norm": 0.40390798449516296, "learning_rate": 7.90555228381366e-05, "loss": 0.0343, "step": 27000 }, { "grad_norm": 0.44071224331855774, "learning_rate": 7.903869992405656e-05, "loss": 0.0337, "step": 27010 }, { "grad_norm": 0.8795284032821655, "learning_rate": 7.902187204805691e-05, "loss": 0.0587, "step": 27020 }, { "grad_norm": 0.30824124813079834, "learning_rate": 7.900503921301308e-05, "loss": 0.0422, "step": 27030 }, { "grad_norm": 0.5500854253768921, "learning_rate": 7.898820142180133e-05, "loss": 0.0579, "step": 27040 }, { "grad_norm": 0.44693413376808167, "learning_rate": 7.897135867729879e-05, "loss": 0.0438, "step": 27050 }, { "grad_norm": 0.3194526433944702, "learning_rate": 7.89545109823834e-05, "loss": 0.0426, "step": 27060 }, { "grad_norm": 0.33939921855926514, "learning_rate": 7.8937658339934e-05, "loss": 0.0394, "step": 27070 }, { "grad_norm": 0.4418858587741852, "learning_rate": 7.892080075283026e-05, "loss": 0.0408, "step": 27080 }, { "grad_norm": 0.34113970398902893, "learning_rate": 7.890393822395263e-05, "loss": 0.041, "step": 27090 }, { "grad_norm": 0.3619801700115204, "learning_rate": 7.88870707561825e-05, "loss": 0.0396, "step": 27100 }, { "grad_norm": 0.5419780015945435, "learning_rate": 7.887019835240203e-05, "loss": 0.0438, "step": 27110 }, { "grad_norm": 0.45275142788887024, "learning_rate": 7.885332101549427e-05, "loss": 0.0451, "step": 27120 }, { "grad_norm": 0.3527544140815735, "learning_rate": 7.883643874834308e-05, "loss": 0.0403, "step": 27130 }, { "grad_norm": 0.49407392740249634, "learning_rate": 7.881955155383321e-05, "loss": 0.0579, "step": 27140 }, { "grad_norm": 0.4950307607650757, "learning_rate": 7.880265943485017e-05, "loss": 0.0429, "step": 27150 }, { "grad_norm": 0.3910384774208069, "learning_rate": 7.878576239428038e-05, "loss": 0.039, "step": 27160 }, { "grad_norm": 0.34402549266815186, "learning_rate": 7.87688604350111e-05, "loss": 0.0419, "step": 27170 }, { "grad_norm": 0.5787535905838013, "learning_rate": 7.875195355993042e-05, "loss": 0.0446, "step": 27180 }, { "grad_norm": 0.4692564010620117, "learning_rate": 7.873504177192724e-05, "loss": 0.0474, "step": 27190 }, { "grad_norm": 0.3643520176410675, "learning_rate": 7.87181250738913e-05, "loss": 0.0506, "step": 27200 }, { "grad_norm": 0.7236224412918091, "learning_rate": 7.870120346871324e-05, "loss": 0.0459, "step": 27210 }, { "grad_norm": 0.4184383153915405, "learning_rate": 7.86842769592845e-05, "loss": 0.0399, "step": 27220 }, { "grad_norm": 0.6425241827964783, "learning_rate": 7.866734554849732e-05, "loss": 0.0406, "step": 27230 }, { "grad_norm": 0.31296634674072266, "learning_rate": 7.865040923924486e-05, "loss": 0.0363, "step": 27240 }, { "grad_norm": 0.4814597964286804, "learning_rate": 7.863346803442104e-05, "loss": 0.0448, "step": 27250 }, { "grad_norm": 0.44358503818511963, "learning_rate": 7.861652193692067e-05, "loss": 0.0343, "step": 27260 }, { "grad_norm": 0.31328341364860535, "learning_rate": 7.859957094963937e-05, "loss": 0.033, "step": 27270 }, { "grad_norm": 0.36140963435173035, "learning_rate": 7.858261507547357e-05, "loss": 0.0362, "step": 27280 }, { "grad_norm": 0.5415652990341187, "learning_rate": 7.856565431732061e-05, "loss": 0.0359, "step": 27290 }, { "grad_norm": 0.39309611916542053, "learning_rate": 7.854868867807859e-05, "loss": 0.0315, "step": 27300 }, { "grad_norm": 0.40749210119247437, "learning_rate": 7.85317181606465e-05, "loss": 0.0397, "step": 27310 }, { "grad_norm": 0.4970971345901489, "learning_rate": 7.85147427679241e-05, "loss": 0.037, "step": 27320 }, { "grad_norm": 0.30558037757873535, "learning_rate": 7.849776250281205e-05, "loss": 0.0312, "step": 27330 }, { "grad_norm": 0.35177692770957947, "learning_rate": 7.84807773682118e-05, "loss": 0.0443, "step": 27340 }, { "grad_norm": 0.34509900212287903, "learning_rate": 7.846378736702565e-05, "loss": 0.0451, "step": 27350 }, { "grad_norm": 0.35478612780570984, "learning_rate": 7.844679250215671e-05, "loss": 0.0503, "step": 27360 }, { "grad_norm": 0.3550502359867096, "learning_rate": 7.842979277650898e-05, "loss": 0.0367, "step": 27370 }, { "grad_norm": 0.3490087389945984, "learning_rate": 7.84127881929872e-05, "loss": 0.0385, "step": 27380 }, { "grad_norm": 0.3872968852519989, "learning_rate": 7.839577875449704e-05, "loss": 0.0345, "step": 27390 }, { "grad_norm": 0.4212762713432312, "learning_rate": 7.837876446394489e-05, "loss": 0.0361, "step": 27400 }, { "grad_norm": 0.32495683431625366, "learning_rate": 7.836174532423805e-05, "loss": 0.04, "step": 27410 }, { "grad_norm": 0.3391246795654297, "learning_rate": 7.834472133828466e-05, "loss": 0.0389, "step": 27420 }, { "grad_norm": 0.45814934372901917, "learning_rate": 7.832769250899359e-05, "loss": 0.0481, "step": 27430 }, { "grad_norm": 0.5430766940116882, "learning_rate": 7.831065883927464e-05, "loss": 0.048, "step": 27440 }, { "grad_norm": 0.32162216305732727, "learning_rate": 7.829362033203841e-05, "loss": 0.0423, "step": 27450 }, { "grad_norm": 0.5996665954589844, "learning_rate": 7.827657699019628e-05, "loss": 0.0414, "step": 27460 }, { "grad_norm": 0.46366631984710693, "learning_rate": 7.825952881666052e-05, "loss": 0.0626, "step": 27470 }, { "grad_norm": 0.8686415553092957, "learning_rate": 7.824247581434418e-05, "loss": 0.0479, "step": 27480 }, { "grad_norm": 0.5003138184547424, "learning_rate": 7.822541798616116e-05, "loss": 0.0549, "step": 27490 }, { "grad_norm": 0.8360114097595215, "learning_rate": 7.820835533502617e-05, "loss": 0.0649, "step": 27500 }, { "grad_norm": 0.434688001871109, "learning_rate": 7.819128786385475e-05, "loss": 0.0531, "step": 27510 }, { "grad_norm": 0.4445961117744446, "learning_rate": 7.817421557556329e-05, "loss": 0.051, "step": 27520 }, { "grad_norm": 0.43111029267311096, "learning_rate": 7.815713847306893e-05, "loss": 0.0556, "step": 27530 }, { "grad_norm": 0.83547443151474, "learning_rate": 7.81400565592897e-05, "loss": 0.0433, "step": 27540 }, { "grad_norm": 0.45232608914375305, "learning_rate": 7.812296983714444e-05, "loss": 0.0471, "step": 27550 }, { "grad_norm": 1.056609869003296, "learning_rate": 7.810587830955281e-05, "loss": 0.0415, "step": 27560 }, { "grad_norm": 0.37326616048812866, "learning_rate": 7.808878197943528e-05, "loss": 0.0471, "step": 27570 }, { "grad_norm": 0.3966928720474243, "learning_rate": 7.807168084971312e-05, "loss": 0.0478, "step": 27580 }, { "grad_norm": 0.3621335029602051, "learning_rate": 7.805457492330849e-05, "loss": 0.0441, "step": 27590 }, { "grad_norm": 0.5205698609352112, "learning_rate": 7.803746420314428e-05, "loss": 0.0422, "step": 27600 }, { "grad_norm": 0.32028213143348694, "learning_rate": 7.802034869214428e-05, "loss": 0.0453, "step": 27610 }, { "grad_norm": 0.3806569278240204, "learning_rate": 7.800322839323303e-05, "loss": 0.0398, "step": 27620 }, { "grad_norm": 0.31493377685546875, "learning_rate": 7.798610330933593e-05, "loss": 0.0359, "step": 27630 }, { "grad_norm": 0.47631216049194336, "learning_rate": 7.796897344337922e-05, "loss": 0.0453, "step": 27640 }, { "grad_norm": 0.5400141477584839, "learning_rate": 7.795183879828989e-05, "loss": 0.0512, "step": 27650 }, { "grad_norm": 0.32523879408836365, "learning_rate": 7.793469937699579e-05, "loss": 0.0541, "step": 27660 }, { "grad_norm": 0.4167729318141937, "learning_rate": 7.791755518242558e-05, "loss": 0.0502, "step": 27670 }, { "grad_norm": 0.42641162872314453, "learning_rate": 7.790040621750876e-05, "loss": 0.0443, "step": 27680 }, { "grad_norm": 0.6656023263931274, "learning_rate": 7.788325248517558e-05, "loss": 0.0397, "step": 27690 }, { "grad_norm": 0.5133889317512512, "learning_rate": 7.786609398835715e-05, "loss": 0.043, "step": 27700 }, { "grad_norm": 0.34340816736221313, "learning_rate": 7.784893072998541e-05, "loss": 0.0401, "step": 27710 }, { "grad_norm": 0.3286517560482025, "learning_rate": 7.783176271299306e-05, "loss": 0.0374, "step": 27720 }, { "grad_norm": 0.661565899848938, "learning_rate": 7.781458994031368e-05, "loss": 0.0444, "step": 27730 }, { "grad_norm": 1.119673490524292, "learning_rate": 7.779741241488161e-05, "loss": 0.0432, "step": 27740 }, { "grad_norm": 0.44047415256500244, "learning_rate": 7.7780230139632e-05, "loss": 0.0446, "step": 27750 }, { "grad_norm": 0.4375230371952057, "learning_rate": 7.776304311750087e-05, "loss": 0.0413, "step": 27760 }, { "grad_norm": 0.36224931478500366, "learning_rate": 7.7745851351425e-05, "loss": 0.0536, "step": 27770 }, { "grad_norm": 0.28120094537734985, "learning_rate": 7.772865484434197e-05, "loss": 0.0469, "step": 27780 }, { "grad_norm": 0.4876856803894043, "learning_rate": 7.77114535991902e-05, "loss": 0.0459, "step": 27790 }, { "grad_norm": 0.34179630875587463, "learning_rate": 7.769424761890893e-05, "loss": 0.0422, "step": 27800 }, { "grad_norm": 0.45319220423698425, "learning_rate": 7.767703690643817e-05, "loss": 0.0473, "step": 27810 }, { "grad_norm": 0.4071083962917328, "learning_rate": 7.76598214647188e-05, "loss": 0.0394, "step": 27820 }, { "grad_norm": 0.38636377453804016, "learning_rate": 7.764260129669241e-05, "loss": 0.051, "step": 27830 }, { "grad_norm": 0.6684812903404236, "learning_rate": 7.76253764053015e-05, "loss": 0.0509, "step": 27840 }, { "grad_norm": 0.3860377371311188, "learning_rate": 7.760814679348932e-05, "loss": 0.038, "step": 27850 }, { "grad_norm": 0.45233646035194397, "learning_rate": 7.759091246419992e-05, "loss": 0.0331, "step": 27860 }, { "grad_norm": 0.5108296275138855, "learning_rate": 7.75736734203782e-05, "loss": 0.0318, "step": 27870 }, { "grad_norm": 0.634654700756073, "learning_rate": 7.755642966496985e-05, "loss": 0.0567, "step": 27880 }, { "grad_norm": 0.8536920547485352, "learning_rate": 7.753918120092132e-05, "loss": 0.0454, "step": 27890 }, { "grad_norm": 3.0330066680908203, "learning_rate": 7.752192803117993e-05, "loss": 0.0466, "step": 27900 }, { "grad_norm": 0.4342512786388397, "learning_rate": 7.750467015869377e-05, "loss": 0.0419, "step": 27910 }, { "grad_norm": 0.4256817698478699, "learning_rate": 7.748740758641174e-05, "loss": 0.047, "step": 27920 }, { "grad_norm": 1.4723470211029053, "learning_rate": 7.74701403172835e-05, "loss": 0.0528, "step": 27930 }, { "grad_norm": 0.4513952434062958, "learning_rate": 7.745286835425962e-05, "loss": 0.0472, "step": 27940 }, { "grad_norm": 0.4008081555366516, "learning_rate": 7.743559170029138e-05, "loss": 0.0413, "step": 27950 }, { "grad_norm": 0.38559406995773315, "learning_rate": 7.741831035833087e-05, "loss": 0.0508, "step": 27960 }, { "grad_norm": 0.3974566161632538, "learning_rate": 7.740102433133102e-05, "loss": 0.0409, "step": 27970 }, { "grad_norm": 0.3019939363002777, "learning_rate": 7.738373362224553e-05, "loss": 0.0316, "step": 27980 }, { "grad_norm": 0.44754308462142944, "learning_rate": 7.73664382340289e-05, "loss": 0.0483, "step": 27990 }, { "grad_norm": 0.468394935131073, "learning_rate": 7.734913816963647e-05, "loss": 0.0474, "step": 28000 }, { "grad_norm": 0.5312840938568115, "learning_rate": 7.73318334320243e-05, "loss": 0.0416, "step": 28010 }, { "grad_norm": 0.4729278087615967, "learning_rate": 7.731452402414934e-05, "loss": 0.0449, "step": 28020 }, { "grad_norm": 0.4596042335033417, "learning_rate": 7.729720994896928e-05, "loss": 0.0449, "step": 28030 }, { "grad_norm": 0.4479197859764099, "learning_rate": 7.727989120944262e-05, "loss": 0.0424, "step": 28040 }, { "grad_norm": 0.48597452044487, "learning_rate": 7.726256780852865e-05, "loss": 0.0371, "step": 28050 }, { "grad_norm": 0.31188762187957764, "learning_rate": 7.724523974918749e-05, "loss": 0.0345, "step": 28060 }, { "grad_norm": 0.37941646575927734, "learning_rate": 7.722790703438002e-05, "loss": 0.0347, "step": 28070 }, { "grad_norm": 0.3422989845275879, "learning_rate": 7.72105696670679e-05, "loss": 0.0373, "step": 28080 }, { "grad_norm": 0.31995314359664917, "learning_rate": 7.719322765021364e-05, "loss": 0.0419, "step": 28090 }, { "grad_norm": 0.575401782989502, "learning_rate": 7.717588098678051e-05, "loss": 0.0451, "step": 28100 }, { "grad_norm": 0.39463794231414795, "learning_rate": 7.715852967973258e-05, "loss": 0.0394, "step": 28110 }, { "grad_norm": 0.5147845149040222, "learning_rate": 7.714117373203474e-05, "loss": 0.0453, "step": 28120 }, { "grad_norm": 0.5557206869125366, "learning_rate": 7.712381314665259e-05, "loss": 0.036, "step": 28130 }, { "grad_norm": 0.44052329659461975, "learning_rate": 7.710644792655261e-05, "loss": 0.0374, "step": 28140 }, { "grad_norm": 0.6399142146110535, "learning_rate": 7.708907807470207e-05, "loss": 0.0528, "step": 28150 }, { "grad_norm": 0.39042380452156067, "learning_rate": 7.707170359406896e-05, "loss": 0.0432, "step": 28160 }, { "grad_norm": 0.2787654399871826, "learning_rate": 7.705432448762213e-05, "loss": 0.0371, "step": 28170 }, { "grad_norm": 0.4404609203338623, "learning_rate": 7.703694075833117e-05, "loss": 0.0468, "step": 28180 }, { "grad_norm": 0.43203428387641907, "learning_rate": 7.70195524091665e-05, "loss": 0.0588, "step": 28190 }, { "grad_norm": 0.5888636112213135, "learning_rate": 7.70021594430993e-05, "loss": 0.0529, "step": 28200 }, { "grad_norm": 0.5918638706207275, "learning_rate": 7.698476186310157e-05, "loss": 0.0438, "step": 28210 }, { "grad_norm": 0.4637017846107483, "learning_rate": 7.696735967214608e-05, "loss": 0.0383, "step": 28220 }, { "grad_norm": 0.43962785601615906, "learning_rate": 7.694995287320636e-05, "loss": 0.0485, "step": 28230 }, { "grad_norm": 0.32835057377815247, "learning_rate": 7.693254146925679e-05, "loss": 0.0373, "step": 28240 }, { "grad_norm": 0.5383363962173462, "learning_rate": 7.691512546327251e-05, "loss": 0.0512, "step": 28250 }, { "grad_norm": 0.4012707471847534, "learning_rate": 7.689770485822939e-05, "loss": 0.0405, "step": 28260 }, { "grad_norm": 0.32544493675231934, "learning_rate": 7.688027965710416e-05, "loss": 0.04, "step": 28270 }, { "grad_norm": 0.34809550642967224, "learning_rate": 7.686284986287433e-05, "loss": 0.0397, "step": 28280 }, { "grad_norm": 0.38380274176597595, "learning_rate": 7.684541547851817e-05, "loss": 0.0428, "step": 28290 }, { "grad_norm": 0.388911634683609, "learning_rate": 7.68279765070147e-05, "loss": 0.0436, "step": 28300 }, { "grad_norm": 0.39079052209854126, "learning_rate": 7.68105329513438e-05, "loss": 0.0368, "step": 28310 }, { "grad_norm": 0.5118090510368347, "learning_rate": 7.67930848144861e-05, "loss": 0.0451, "step": 28320 }, { "grad_norm": 0.47073814272880554, "learning_rate": 7.6775632099423e-05, "loss": 0.045, "step": 28330 }, { "grad_norm": 0.33638209104537964, "learning_rate": 7.675817480913667e-05, "loss": 0.0395, "step": 28340 }, { "grad_norm": 0.33841466903686523, "learning_rate": 7.674071294661011e-05, "loss": 0.0346, "step": 28350 }, { "grad_norm": 0.3258403539657593, "learning_rate": 7.672324651482707e-05, "loss": 0.0423, "step": 28360 }, { "grad_norm": 1.1296528577804565, "learning_rate": 7.670577551677209e-05, "loss": 0.0393, "step": 28370 }, { "grad_norm": 0.3092230260372162, "learning_rate": 7.668829995543047e-05, "loss": 0.0368, "step": 28380 }, { "grad_norm": 0.6180593967437744, "learning_rate": 7.667081983378832e-05, "loss": 0.0517, "step": 28390 }, { "grad_norm": 0.33979329466819763, "learning_rate": 7.66533351548325e-05, "loss": 0.0405, "step": 28400 }, { "grad_norm": 0.3849603533744812, "learning_rate": 7.663584592155069e-05, "loss": 0.0348, "step": 28410 }, { "grad_norm": 0.49483391642570496, "learning_rate": 7.661835213693129e-05, "loss": 0.0515, "step": 28420 }, { "grad_norm": 0.511709988117218, "learning_rate": 7.660085380396353e-05, "loss": 0.0417, "step": 28430 }, { "grad_norm": 0.41351959109306335, "learning_rate": 7.658335092563738e-05, "loss": 0.0417, "step": 28440 }, { "grad_norm": 0.9343783259391785, "learning_rate": 7.656584350494362e-05, "loss": 0.042, "step": 28450 }, { "grad_norm": 0.4653586149215698, "learning_rate": 7.654833154487378e-05, "loss": 0.0527, "step": 28460 }, { "grad_norm": 0.29300329089164734, "learning_rate": 7.653081504842017e-05, "loss": 0.0374, "step": 28470 }, { "grad_norm": 0.57813960313797, "learning_rate": 7.65132940185759e-05, "loss": 0.0491, "step": 28480 }, { "grad_norm": 0.5731943845748901, "learning_rate": 7.649576845833481e-05, "loss": 0.0486, "step": 28490 }, { "grad_norm": 0.273667573928833, "learning_rate": 7.647823837069156e-05, "loss": 0.0437, "step": 28500 }, { "grad_norm": 0.34047552943229675, "learning_rate": 7.646070375864156e-05, "loss": 0.0402, "step": 28510 }, { "grad_norm": 0.40121549367904663, "learning_rate": 7.644316462518099e-05, "loss": 0.0436, "step": 28520 }, { "grad_norm": 0.7553543448448181, "learning_rate": 7.642562097330679e-05, "loss": 0.0427, "step": 28530 }, { "grad_norm": 0.36695894598960876, "learning_rate": 7.640807280601671e-05, "loss": 0.0361, "step": 28540 }, { "grad_norm": 0.35468801856040955, "learning_rate": 7.639052012630927e-05, "loss": 0.05, "step": 28550 }, { "grad_norm": 0.4326346218585968, "learning_rate": 7.63729629371837e-05, "loss": 0.0459, "step": 28560 }, { "grad_norm": 0.3793697655200958, "learning_rate": 7.635540124164009e-05, "loss": 0.053, "step": 28570 }, { "grad_norm": 0.43618708848953247, "learning_rate": 7.633783504267922e-05, "loss": 0.048, "step": 28580 }, { "grad_norm": 0.6312034726142883, "learning_rate": 7.632026434330269e-05, "loss": 0.0482, "step": 28590 }, { "grad_norm": 0.42592287063598633, "learning_rate": 7.630268914651282e-05, "loss": 0.0404, "step": 28600 }, { "grad_norm": 0.4283890724182129, "learning_rate": 7.628510945531278e-05, "loss": 0.0542, "step": 28610 }, { "grad_norm": 0.322933554649353, "learning_rate": 7.626752527270641e-05, "loss": 0.0425, "step": 28620 }, { "grad_norm": 0.24184681475162506, "learning_rate": 7.62499366016984e-05, "loss": 0.0355, "step": 28630 }, { "grad_norm": 0.23700454831123352, "learning_rate": 7.623234344529416e-05, "loss": 0.0431, "step": 28640 }, { "grad_norm": 0.39329618215560913, "learning_rate": 7.62147458064999e-05, "loss": 0.0427, "step": 28650 }, { "grad_norm": 0.5549142956733704, "learning_rate": 7.619714368832254e-05, "loss": 0.0523, "step": 28660 }, { "grad_norm": 0.4652426540851593, "learning_rate": 7.61795370937698e-05, "loss": 0.0576, "step": 28670 }, { "grad_norm": 0.3609826862812042, "learning_rate": 7.61619260258502e-05, "loss": 0.0366, "step": 28680 }, { "grad_norm": 0.42676496505737305, "learning_rate": 7.614431048757298e-05, "loss": 0.0464, "step": 28690 }, { "grad_norm": 0.34491342306137085, "learning_rate": 7.612669048194814e-05, "loss": 0.0449, "step": 28700 }, { "grad_norm": 0.5529112815856934, "learning_rate": 7.610906601198646e-05, "loss": 0.0371, "step": 28710 }, { "grad_norm": 0.30481475591659546, "learning_rate": 7.60914370806995e-05, "loss": 0.043, "step": 28720 }, { "grad_norm": 0.35724371671676636, "learning_rate": 7.607380369109953e-05, "loss": 0.0661, "step": 28730 }, { "grad_norm": 0.4062212407588959, "learning_rate": 7.605616584619961e-05, "loss": 0.0365, "step": 28740 }, { "grad_norm": 0.4676084518432617, "learning_rate": 7.603852354901362e-05, "loss": 0.0449, "step": 28750 }, { "grad_norm": 0.5501010417938232, "learning_rate": 7.602087680255609e-05, "loss": 0.0572, "step": 28760 }, { "grad_norm": 0.34024038910865784, "learning_rate": 7.600322560984238e-05, "loss": 0.0357, "step": 28770 }, { "grad_norm": 0.5600265860557556, "learning_rate": 7.598556997388863e-05, "loss": 0.0448, "step": 28780 }, { "grad_norm": 0.9377803802490234, "learning_rate": 7.596790989771166e-05, "loss": 0.0493, "step": 28790 }, { "grad_norm": 0.4603292942047119, "learning_rate": 7.595024538432914e-05, "loss": 0.0437, "step": 28800 }, { "grad_norm": 0.5049324631690979, "learning_rate": 7.59325764367594e-05, "loss": 0.0393, "step": 28810 }, { "grad_norm": 0.41373035311698914, "learning_rate": 7.59149030580216e-05, "loss": 0.0485, "step": 28820 }, { "grad_norm": 0.41663140058517456, "learning_rate": 7.589722525113562e-05, "loss": 0.0345, "step": 28830 }, { "grad_norm": 0.4788146913051605, "learning_rate": 7.587954301912216e-05, "loss": 0.0348, "step": 28840 }, { "grad_norm": 0.5512632727622986, "learning_rate": 7.586185636500263e-05, "loss": 0.0514, "step": 28850 }, { "grad_norm": 0.5269279479980469, "learning_rate": 7.584416529179914e-05, "loss": 0.044, "step": 28860 }, { "grad_norm": 0.4318508505821228, "learning_rate": 7.582646980253465e-05, "loss": 0.0433, "step": 28870 }, { "grad_norm": 0.37519657611846924, "learning_rate": 7.580876990023282e-05, "loss": 0.0529, "step": 28880 }, { "grad_norm": 0.47138863801956177, "learning_rate": 7.579106558791809e-05, "loss": 0.0407, "step": 28890 }, { "grad_norm": 0.8753939270973206, "learning_rate": 7.577335686861565e-05, "loss": 0.0402, "step": 28900 }, { "grad_norm": 0.36981841921806335, "learning_rate": 7.575564374535141e-05, "loss": 0.0293, "step": 28910 }, { "grad_norm": 0.4108924865722656, "learning_rate": 7.573792622115207e-05, "loss": 0.0358, "step": 28920 }, { "grad_norm": 0.5526973009109497, "learning_rate": 7.572020429904507e-05, "loss": 0.0393, "step": 28930 }, { "grad_norm": 0.574558675289154, "learning_rate": 7.570247798205861e-05, "loss": 0.0488, "step": 28940 }, { "grad_norm": 0.5911026000976562, "learning_rate": 7.568474727322164e-05, "loss": 0.0403, "step": 28950 }, { "grad_norm": 0.33168962597846985, "learning_rate": 7.566701217556384e-05, "loss": 0.0381, "step": 28960 }, { "grad_norm": 0.48931005597114563, "learning_rate": 7.564927269211564e-05, "loss": 0.0516, "step": 28970 }, { "grad_norm": 0.6005535125732422, "learning_rate": 7.563152882590824e-05, "loss": 0.0391, "step": 28980 }, { "grad_norm": 0.4236428439617157, "learning_rate": 7.56137805799736e-05, "loss": 0.0367, "step": 28990 }, { "grad_norm": 0.399344801902771, "learning_rate": 7.559602795734439e-05, "loss": 0.0508, "step": 29000 }, { "grad_norm": 0.4291248321533203, "learning_rate": 7.557827096105408e-05, "loss": 0.0375, "step": 29010 }, { "grad_norm": 0.5149611830711365, "learning_rate": 7.55605095941368e-05, "loss": 0.0435, "step": 29020 }, { "grad_norm": 0.3313141167163849, "learning_rate": 7.554274385962752e-05, "loss": 0.0432, "step": 29030 }, { "grad_norm": 0.43135449290275574, "learning_rate": 7.552497376056191e-05, "loss": 0.0555, "step": 29040 }, { "grad_norm": 0.4264967441558838, "learning_rate": 7.550719929997639e-05, "loss": 0.045, "step": 29050 }, { "grad_norm": 0.4706692099571228, "learning_rate": 7.548942048090813e-05, "loss": 0.0443, "step": 29060 }, { "grad_norm": 0.48833152651786804, "learning_rate": 7.547163730639506e-05, "loss": 0.0455, "step": 29070 }, { "grad_norm": 0.4407796263694763, "learning_rate": 7.545384977947583e-05, "loss": 0.0489, "step": 29080 }, { "grad_norm": 0.36375778913497925, "learning_rate": 7.543605790318981e-05, "loss": 0.0435, "step": 29090 }, { "grad_norm": 0.8288133144378662, "learning_rate": 7.54182616805772e-05, "loss": 0.0448, "step": 29100 }, { "grad_norm": 0.6844086050987244, "learning_rate": 7.540046111467885e-05, "loss": 0.0394, "step": 29110 }, { "grad_norm": 0.35854193568229675, "learning_rate": 7.53826562085364e-05, "loss": 0.039, "step": 29120 }, { "grad_norm": 0.446887731552124, "learning_rate": 7.536484696519221e-05, "loss": 0.0358, "step": 29130 }, { "grad_norm": 0.4782381057739258, "learning_rate": 7.534703338768942e-05, "loss": 0.0448, "step": 29140 }, { "grad_norm": 0.4950411319732666, "learning_rate": 7.532921547907185e-05, "loss": 0.0379, "step": 29150 }, { "grad_norm": 1.1183042526245117, "learning_rate": 7.531139324238412e-05, "loss": 0.0431, "step": 29160 }, { "grad_norm": 0.5762420892715454, "learning_rate": 7.529356668067157e-05, "loss": 0.0527, "step": 29170 }, { "grad_norm": 0.5926799178123474, "learning_rate": 7.527573579698023e-05, "loss": 0.0485, "step": 29180 }, { "grad_norm": 0.45829683542251587, "learning_rate": 7.525790059435693e-05, "loss": 0.0405, "step": 29190 }, { "grad_norm": 0.45044150948524475, "learning_rate": 7.524006107584926e-05, "loss": 0.0477, "step": 29200 }, { "grad_norm": 0.3346621096134186, "learning_rate": 7.522221724450544e-05, "loss": 0.0372, "step": 29210 }, { "grad_norm": 0.5325780510902405, "learning_rate": 7.520436910337451e-05, "loss": 0.0513, "step": 29220 }, { "grad_norm": 1.3204642534255981, "learning_rate": 7.518651665550627e-05, "loss": 0.0362, "step": 29230 }, { "grad_norm": 0.2888697683811188, "learning_rate": 7.516865990395117e-05, "loss": 0.0474, "step": 29240 }, { "grad_norm": 0.4387681186199188, "learning_rate": 7.515079885176047e-05, "loss": 0.0433, "step": 29250 }, { "grad_norm": 0.35308095812797546, "learning_rate": 7.513293350198612e-05, "loss": 0.037, "step": 29260 }, { "grad_norm": 0.5300085544586182, "learning_rate": 7.511506385768081e-05, "loss": 0.0384, "step": 29270 }, { "grad_norm": 0.6056085824966431, "learning_rate": 7.509718992189801e-05, "loss": 0.047, "step": 29280 }, { "grad_norm": 0.44989487528800964, "learning_rate": 7.507931169769182e-05, "loss": 0.0423, "step": 29290 }, { "grad_norm": 0.4032541811466217, "learning_rate": 7.506142918811722e-05, "loss": 0.0386, "step": 29300 }, { "grad_norm": 0.3070141077041626, "learning_rate": 7.504354239622978e-05, "loss": 0.0451, "step": 29310 }, { "grad_norm": 0.5133172273635864, "learning_rate": 7.50256513250859e-05, "loss": 0.0455, "step": 29320 }, { "grad_norm": 0.4008387327194214, "learning_rate": 7.500775597774265e-05, "loss": 0.0362, "step": 29330 }, { "grad_norm": 0.3815937638282776, "learning_rate": 7.498985635725788e-05, "loss": 0.0357, "step": 29340 }, { "grad_norm": 0.657171368598938, "learning_rate": 7.497195246669012e-05, "loss": 0.0425, "step": 29350 }, { "grad_norm": 0.5241934657096863, "learning_rate": 7.495404430909868e-05, "loss": 0.0447, "step": 29360 }, { "grad_norm": 0.3037532866001129, "learning_rate": 7.493613188754356e-05, "loss": 0.0423, "step": 29370 }, { "grad_norm": 0.3719695806503296, "learning_rate": 7.49182152050855e-05, "loss": 0.041, "step": 29380 }, { "grad_norm": 0.5045694708824158, "learning_rate": 7.490029426478598e-05, "loss": 0.0358, "step": 29390 }, { "grad_norm": 0.37821078300476074, "learning_rate": 7.488236906970719e-05, "loss": 0.0385, "step": 29400 }, { "grad_norm": 0.4568997919559479, "learning_rate": 7.486443962291207e-05, "loss": 0.0442, "step": 29410 }, { "grad_norm": 0.366560697555542, "learning_rate": 7.484650592746424e-05, "loss": 0.0346, "step": 29420 }, { "grad_norm": 0.37287476658821106, "learning_rate": 7.482856798642811e-05, "loss": 0.0366, "step": 29430 }, { "grad_norm": 0.4778156280517578, "learning_rate": 7.481062580286878e-05, "loss": 0.0494, "step": 29440 }, { "grad_norm": 0.38114064931869507, "learning_rate": 7.479267937985208e-05, "loss": 0.0422, "step": 29450 }, { "grad_norm": 0.4700707197189331, "learning_rate": 7.477472872044456e-05, "loss": 0.0385, "step": 29460 }, { "grad_norm": 0.9130417108535767, "learning_rate": 7.475677382771347e-05, "loss": 0.0443, "step": 29470 }, { "grad_norm": 0.3531750738620758, "learning_rate": 7.473881470472683e-05, "loss": 0.0454, "step": 29480 }, { "grad_norm": 0.4949168264865875, "learning_rate": 7.472085135455337e-05, "loss": 0.0439, "step": 29490 }, { "grad_norm": 0.5421390533447266, "learning_rate": 7.470288378026256e-05, "loss": 0.0531, "step": 29500 }, { "grad_norm": 0.7016215324401855, "learning_rate": 7.468491198492451e-05, "loss": 0.0348, "step": 29510 }, { "grad_norm": 0.38418295979499817, "learning_rate": 7.466693597161013e-05, "loss": 0.0328, "step": 29520 }, { "grad_norm": 0.3955647349357605, "learning_rate": 7.464895574339104e-05, "loss": 0.0442, "step": 29530 }, { "grad_norm": 0.3359333574771881, "learning_rate": 7.463097130333958e-05, "loss": 0.0477, "step": 29540 }, { "grad_norm": 0.44831475615501404, "learning_rate": 7.461298265452876e-05, "loss": 0.0444, "step": 29550 }, { "grad_norm": 0.3177972435951233, "learning_rate": 7.459498980003239e-05, "loss": 0.0349, "step": 29560 }, { "grad_norm": 0.5263708829879761, "learning_rate": 7.457699274292493e-05, "loss": 0.0438, "step": 29570 }, { "grad_norm": 0.31350308656692505, "learning_rate": 7.455899148628159e-05, "loss": 0.0511, "step": 29580 }, { "grad_norm": 0.40585342049598694, "learning_rate": 7.45409860331783e-05, "loss": 0.0381, "step": 29590 }, { "grad_norm": 0.4144717752933502, "learning_rate": 7.452297638669169e-05, "loss": 0.0314, "step": 29600 }, { "grad_norm": 0.35012415051460266, "learning_rate": 7.450496254989911e-05, "loss": 0.0417, "step": 29610 }, { "grad_norm": 0.3319641053676605, "learning_rate": 7.448694452587866e-05, "loss": 0.0406, "step": 29620 }, { "grad_norm": 0.3573160469532013, "learning_rate": 7.44689223177091e-05, "loss": 0.0544, "step": 29630 }, { "grad_norm": 0.44116032123565674, "learning_rate": 7.445089592846994e-05, "loss": 0.036, "step": 29640 }, { "grad_norm": 0.5432223677635193, "learning_rate": 7.443286536124141e-05, "loss": 0.0425, "step": 29650 }, { "grad_norm": 0.43264201283454895, "learning_rate": 7.441483061910443e-05, "loss": 0.0412, "step": 29660 }, { "grad_norm": 0.7458211183547974, "learning_rate": 7.439679170514064e-05, "loss": 0.0507, "step": 29670 }, { "grad_norm": 0.38854897022247314, "learning_rate": 7.43787486224324e-05, "loss": 0.0382, "step": 29680 }, { "grad_norm": 0.4509827196598053, "learning_rate": 7.436070137406276e-05, "loss": 0.042, "step": 29690 }, { "grad_norm": 0.35232409834861755, "learning_rate": 7.434264996311556e-05, "loss": 0.035, "step": 29700 }, { "grad_norm": 0.5035144686698914, "learning_rate": 7.432459439267525e-05, "loss": 0.0396, "step": 29710 }, { "grad_norm": 0.41554123163223267, "learning_rate": 7.430653466582701e-05, "loss": 0.0384, "step": 29720 }, { "grad_norm": 0.3483515977859497, "learning_rate": 7.42884707856568e-05, "loss": 0.0475, "step": 29730 }, { "grad_norm": 0.3867361843585968, "learning_rate": 7.427040275525122e-05, "loss": 0.0373, "step": 29740 }, { "grad_norm": 0.4209800958633423, "learning_rate": 7.42523305776976e-05, "loss": 0.0385, "step": 29750 }, { "grad_norm": 0.508438229560852, "learning_rate": 7.4234254256084e-05, "loss": 0.0449, "step": 29760 }, { "grad_norm": 0.6921877264976501, "learning_rate": 7.421617379349915e-05, "loss": 0.0546, "step": 29770 }, { "grad_norm": 0.3913876414299011, "learning_rate": 7.41980891930325e-05, "loss": 0.0415, "step": 29780 }, { "grad_norm": 0.4745946228504181, "learning_rate": 7.418000045777425e-05, "loss": 0.0332, "step": 29790 }, { "grad_norm": 0.5422821044921875, "learning_rate": 7.416190759081523e-05, "loss": 0.0416, "step": 29800 }, { "grad_norm": 0.5463343262672424, "learning_rate": 7.414381059524704e-05, "loss": 0.0391, "step": 29810 }, { "grad_norm": 0.39869266748428345, "learning_rate": 7.412570947416195e-05, "loss": 0.0325, "step": 29820 }, { "grad_norm": 0.5234299898147583, "learning_rate": 7.410760423065295e-05, "loss": 0.0526, "step": 29830 }, { "grad_norm": 1.3636740446090698, "learning_rate": 7.408949486781372e-05, "loss": 0.0546, "step": 29840 }, { "grad_norm": 0.44051963090896606, "learning_rate": 7.407138138873868e-05, "loss": 0.053, "step": 29850 }, { "grad_norm": 0.44113701581954956, "learning_rate": 7.405326379652292e-05, "loss": 0.0529, "step": 29860 }, { "grad_norm": 0.35936421155929565, "learning_rate": 7.403514209426222e-05, "loss": 0.0472, "step": 29870 }, { "grad_norm": 0.4761769473552704, "learning_rate": 7.40170162850531e-05, "loss": 0.0475, "step": 29880 }, { "grad_norm": 0.48325812816619873, "learning_rate": 7.399888637199278e-05, "loss": 0.0363, "step": 29890 }, { "grad_norm": 0.5653668642044067, "learning_rate": 7.398075235817914e-05, "loss": 0.0546, "step": 29900 }, { "grad_norm": 0.8358132839202881, "learning_rate": 7.39626142467108e-05, "loss": 0.0434, "step": 29910 }, { "grad_norm": 1.2715836763381958, "learning_rate": 7.394447204068706e-05, "loss": 0.0396, "step": 29920 }, { "grad_norm": 0.5110361576080322, "learning_rate": 7.392632574320793e-05, "loss": 0.0437, "step": 29930 }, { "grad_norm": 0.968330442905426, "learning_rate": 7.390817535737411e-05, "loss": 0.0385, "step": 29940 }, { "grad_norm": 0.4901718497276306, "learning_rate": 7.389002088628703e-05, "loss": 0.0387, "step": 29950 }, { "grad_norm": 0.45402199029922485, "learning_rate": 7.387186233304877e-05, "loss": 0.0463, "step": 29960 }, { "grad_norm": 0.43527644872665405, "learning_rate": 7.385369970076212e-05, "loss": 0.0344, "step": 29970 }, { "grad_norm": 0.5969868302345276, "learning_rate": 7.38355329925306e-05, "loss": 0.0497, "step": 29980 }, { "grad_norm": 0.3814965486526489, "learning_rate": 7.381736221145838e-05, "loss": 0.0423, "step": 29990 }, { "grad_norm": 0.4214581847190857, "learning_rate": 7.37991873606504e-05, "loss": 0.0338, "step": 30000 }, { "grad_norm": 0.36761534214019775, "learning_rate": 7.378100844321218e-05, "loss": 0.0483, "step": 30010 }, { "grad_norm": 0.5649234652519226, "learning_rate": 7.376282546225004e-05, "loss": 0.0332, "step": 30020 }, { "grad_norm": 0.6794222593307495, "learning_rate": 7.374463842087094e-05, "loss": 0.0396, "step": 30030 }, { "grad_norm": 0.5289500951766968, "learning_rate": 7.372644732218254e-05, "loss": 0.0389, "step": 30040 }, { "grad_norm": 0.35739850997924805, "learning_rate": 7.370825216929322e-05, "loss": 0.0477, "step": 30050 }, { "grad_norm": 0.3290834128856659, "learning_rate": 7.369005296531205e-05, "loss": 0.0494, "step": 30060 }, { "grad_norm": 0.44718438386917114, "learning_rate": 7.367184971334873e-05, "loss": 0.0422, "step": 30070 }, { "grad_norm": 0.39329981803894043, "learning_rate": 7.365364241651371e-05, "loss": 0.0424, "step": 30080 }, { "grad_norm": 0.43077439069747925, "learning_rate": 7.363543107791815e-05, "loss": 0.0387, "step": 30090 }, { "grad_norm": 0.5380855202674866, "learning_rate": 7.361721570067384e-05, "loss": 0.0429, "step": 30100 }, { "grad_norm": 0.42499926686286926, "learning_rate": 7.359899628789331e-05, "loss": 0.0574, "step": 30110 }, { "grad_norm": 0.4440183937549591, "learning_rate": 7.358077284268974e-05, "loss": 0.045, "step": 30120 }, { "grad_norm": 0.6876449584960938, "learning_rate": 7.356254536817702e-05, "loss": 0.0391, "step": 30130 }, { "grad_norm": 0.4105169475078583, "learning_rate": 7.354431386746973e-05, "loss": 0.0401, "step": 30140 }, { "grad_norm": 0.4995388984680176, "learning_rate": 7.352607834368316e-05, "loss": 0.0348, "step": 30150 }, { "grad_norm": 0.4147016108036041, "learning_rate": 7.350783879993324e-05, "loss": 0.0375, "step": 30160 }, { "grad_norm": 0.31791749596595764, "learning_rate": 7.348959523933658e-05, "loss": 0.0394, "step": 30170 }, { "grad_norm": 0.3424052894115448, "learning_rate": 7.347134766501057e-05, "loss": 0.0331, "step": 30180 }, { "grad_norm": 0.46926772594451904, "learning_rate": 7.345309608007315e-05, "loss": 0.0366, "step": 30190 }, { "grad_norm": 0.5544354915618896, "learning_rate": 7.343484048764308e-05, "loss": 0.0451, "step": 30200 }, { "grad_norm": 0.46366873383522034, "learning_rate": 7.341658089083972e-05, "loss": 0.0431, "step": 30210 }, { "grad_norm": 0.34469282627105713, "learning_rate": 7.339831729278313e-05, "loss": 0.0413, "step": 30220 }, { "grad_norm": 0.42572757601737976, "learning_rate": 7.338004969659404e-05, "loss": 0.0341, "step": 30230 }, { "grad_norm": 0.4194553792476654, "learning_rate": 7.336177810539391e-05, "loss": 0.0497, "step": 30240 }, { "grad_norm": 0.40200120210647583, "learning_rate": 7.334350252230485e-05, "loss": 0.0326, "step": 30250 }, { "grad_norm": 0.2707642912864685, "learning_rate": 7.332522295044965e-05, "loss": 0.0379, "step": 30260 }, { "grad_norm": 0.44489139318466187, "learning_rate": 7.33069393929518e-05, "loss": 0.0424, "step": 30270 }, { "grad_norm": 0.4298955202102661, "learning_rate": 7.328865185293545e-05, "loss": 0.0513, "step": 30280 }, { "grad_norm": 0.47334039211273193, "learning_rate": 7.327036033352546e-05, "loss": 0.0479, "step": 30290 }, { "grad_norm": 0.3840359151363373, "learning_rate": 7.325206483784733e-05, "loss": 0.0358, "step": 30300 }, { "grad_norm": 0.5249068140983582, "learning_rate": 7.323376536902724e-05, "loss": 0.0559, "step": 30310 }, { "grad_norm": 0.35897311568260193, "learning_rate": 7.321546193019213e-05, "loss": 0.0461, "step": 30320 }, { "grad_norm": 0.38594749569892883, "learning_rate": 7.31971545244695e-05, "loss": 0.0442, "step": 30330 }, { "grad_norm": 0.48897963762283325, "learning_rate": 7.31788431549876e-05, "loss": 0.0435, "step": 30340 }, { "grad_norm": 0.46491825580596924, "learning_rate": 7.316052782487534e-05, "loss": 0.037, "step": 30350 }, { "grad_norm": 0.5368831157684326, "learning_rate": 7.314220853726234e-05, "loss": 0.0435, "step": 30360 }, { "grad_norm": 0.6082685589790344, "learning_rate": 7.312388529527884e-05, "loss": 0.0441, "step": 30370 }, { "grad_norm": 0.4168534278869629, "learning_rate": 7.310555810205577e-05, "loss": 0.0414, "step": 30380 }, { "grad_norm": 0.26309308409690857, "learning_rate": 7.308722696072476e-05, "loss": 0.0355, "step": 30390 }, { "grad_norm": 0.5898239612579346, "learning_rate": 7.306889187441811e-05, "loss": 0.0411, "step": 30400 }, { "grad_norm": 0.3420807719230652, "learning_rate": 7.305055284626876e-05, "loss": 0.0431, "step": 30410 }, { "grad_norm": 0.3831409215927124, "learning_rate": 7.303220987941037e-05, "loss": 0.0359, "step": 30420 }, { "grad_norm": 0.5237017869949341, "learning_rate": 7.301386297697726e-05, "loss": 0.038, "step": 30430 }, { "grad_norm": 0.43355390429496765, "learning_rate": 7.299551214210438e-05, "loss": 0.0445, "step": 30440 }, { "grad_norm": 0.28595736622810364, "learning_rate": 7.297715737792738e-05, "loss": 0.0335, "step": 30450 }, { "grad_norm": 0.5246178507804871, "learning_rate": 7.295879868758265e-05, "loss": 0.036, "step": 30460 }, { "grad_norm": 0.31051936745643616, "learning_rate": 7.294043607420713e-05, "loss": 0.0362, "step": 30470 }, { "grad_norm": 0.3223739564418793, "learning_rate": 7.292206954093852e-05, "loss": 0.0431, "step": 30480 }, { "grad_norm": 0.4234927296638489, "learning_rate": 7.290369909091515e-05, "loss": 0.0454, "step": 30490 }, { "grad_norm": 0.5030019283294678, "learning_rate": 7.2885324727276e-05, "loss": 0.0472, "step": 30500 }, { "grad_norm": 0.43132075667381287, "learning_rate": 7.286694645316076e-05, "loss": 0.0469, "step": 30510 }, { "grad_norm": 0.33098939061164856, "learning_rate": 7.284856427170982e-05, "loss": 0.039, "step": 30520 }, { "grad_norm": 0.4352603852748871, "learning_rate": 7.283017818606414e-05, "loss": 0.0503, "step": 30530 }, { "grad_norm": 0.5927151441574097, "learning_rate": 7.28117881993654e-05, "loss": 0.0489, "step": 30540 }, { "grad_norm": 0.43300357460975647, "learning_rate": 7.279339431475598e-05, "loss": 0.0379, "step": 30550 }, { "grad_norm": 0.41575151681900024, "learning_rate": 7.277499653537887e-05, "loss": 0.0396, "step": 30560 }, { "grad_norm": 0.5558313131332397, "learning_rate": 7.275659486437776e-05, "loss": 0.0464, "step": 30570 }, { "grad_norm": 0.4904414713382721, "learning_rate": 7.273818930489695e-05, "loss": 0.0444, "step": 30580 }, { "grad_norm": 0.3758291006088257, "learning_rate": 7.271977986008151e-05, "loss": 0.0429, "step": 30590 }, { "grad_norm": 0.4440297484397888, "learning_rate": 7.270136653307705e-05, "loss": 0.0375, "step": 30600 }, { "grad_norm": 0.40464743971824646, "learning_rate": 7.268294932702994e-05, "loss": 0.0353, "step": 30610 }, { "grad_norm": 0.4101018011569977, "learning_rate": 7.266452824508719e-05, "loss": 0.0443, "step": 30620 }, { "grad_norm": 0.4306139647960663, "learning_rate": 7.264610329039643e-05, "loss": 0.0385, "step": 30630 }, { "grad_norm": 0.38227778673171997, "learning_rate": 7.262767446610599e-05, "loss": 0.0391, "step": 30640 }, { "grad_norm": 0.6007211208343506, "learning_rate": 7.260924177536485e-05, "loss": 0.0507, "step": 30650 }, { "grad_norm": 0.5061405301094055, "learning_rate": 7.259080522132265e-05, "loss": 0.0435, "step": 30660 }, { "grad_norm": 0.47244346141815186, "learning_rate": 7.257236480712972e-05, "loss": 0.0369, "step": 30670 }, { "grad_norm": 0.42679083347320557, "learning_rate": 7.255392053593697e-05, "loss": 0.0314, "step": 30680 }, { "grad_norm": 0.3856189250946045, "learning_rate": 7.253547241089607e-05, "loss": 0.0364, "step": 30690 }, { "grad_norm": 0.40810832381248474, "learning_rate": 7.251702043515927e-05, "loss": 0.0357, "step": 30700 }, { "grad_norm": 1.070789098739624, "learning_rate": 7.249856461187952e-05, "loss": 0.0361, "step": 30710 }, { "grad_norm": 0.35791563987731934, "learning_rate": 7.248010494421042e-05, "loss": 0.0338, "step": 30720 }, { "grad_norm": 0.3753274977207184, "learning_rate": 7.246164143530622e-05, "loss": 0.0473, "step": 30730 }, { "grad_norm": 0.34911128878593445, "learning_rate": 7.244317408832181e-05, "loss": 0.0409, "step": 30740 }, { "grad_norm": 0.35241663455963135, "learning_rate": 7.242470290641279e-05, "loss": 0.0364, "step": 30750 }, { "grad_norm": 0.49244287610054016, "learning_rate": 7.240622789273536e-05, "loss": 0.0371, "step": 30760 }, { "grad_norm": 0.6551334857940674, "learning_rate": 7.238774905044638e-05, "loss": 0.0471, "step": 30770 }, { "grad_norm": 0.4152946174144745, "learning_rate": 7.236926638270341e-05, "loss": 0.0395, "step": 30780 }, { "grad_norm": 0.47355031967163086, "learning_rate": 7.23507798926646e-05, "loss": 0.0437, "step": 30790 }, { "grad_norm": 0.4338340163230896, "learning_rate": 7.23322895834888e-05, "loss": 0.042, "step": 30800 }, { "grad_norm": 0.34941527247428894, "learning_rate": 7.231379545833552e-05, "loss": 0.0384, "step": 30810 }, { "grad_norm": 0.44594889879226685, "learning_rate": 7.229529752036487e-05, "loss": 0.0384, "step": 30820 }, { "grad_norm": 0.369986355304718, "learning_rate": 7.227679577273765e-05, "loss": 0.0456, "step": 30830 }, { "grad_norm": 0.3997175097465515, "learning_rate": 7.225829021861529e-05, "loss": 0.038, "step": 30840 }, { "grad_norm": 0.2568216919898987, "learning_rate": 7.223978086115992e-05, "loss": 0.0333, "step": 30850 }, { "grad_norm": 0.6364225149154663, "learning_rate": 7.222126770353425e-05, "loss": 0.0463, "step": 30860 }, { "grad_norm": 0.3979567587375641, "learning_rate": 7.22027507489017e-05, "loss": 0.0391, "step": 30870 }, { "grad_norm": 0.36461007595062256, "learning_rate": 7.218423000042627e-05, "loss": 0.0414, "step": 30880 }, { "grad_norm": 0.43419149518013, "learning_rate": 7.216570546127268e-05, "loss": 0.0394, "step": 30890 }, { "grad_norm": 0.3098463714122772, "learning_rate": 7.214717713460626e-05, "loss": 0.0446, "step": 30900 }, { "grad_norm": 0.39106476306915283, "learning_rate": 7.2128645023593e-05, "loss": 0.0415, "step": 30910 }, { "grad_norm": 0.2931218147277832, "learning_rate": 7.211010913139951e-05, "loss": 0.0408, "step": 30920 }, { "grad_norm": 0.5792466402053833, "learning_rate": 7.209156946119308e-05, "loss": 0.0309, "step": 30930 }, { "grad_norm": 0.40748411417007446, "learning_rate": 7.207302601614166e-05, "loss": 0.0364, "step": 30940 }, { "grad_norm": 0.2551114857196808, "learning_rate": 7.205447879941378e-05, "loss": 0.0348, "step": 30950 }, { "grad_norm": 0.312710702419281, "learning_rate": 7.203592781417866e-05, "loss": 0.0399, "step": 30960 }, { "grad_norm": 0.342037558555603, "learning_rate": 7.201737306360617e-05, "loss": 0.0328, "step": 30970 }, { "grad_norm": 0.5761469602584839, "learning_rate": 7.19988145508668e-05, "loss": 0.0366, "step": 30980 }, { "grad_norm": 0.39653337001800537, "learning_rate": 7.198025227913168e-05, "loss": 0.0412, "step": 30990 }, { "grad_norm": 0.5849342346191406, "learning_rate": 7.196168625157261e-05, "loss": 0.0415, "step": 31000 }, { "grad_norm": 0.3698936104774475, "learning_rate": 7.194311647136201e-05, "loss": 0.0356, "step": 31010 }, { "grad_norm": 0.5304430723190308, "learning_rate": 7.192454294167297e-05, "loss": 0.0422, "step": 31020 }, { "grad_norm": 0.4456908106803894, "learning_rate": 7.190596566567917e-05, "loss": 0.0367, "step": 31030 }, { "grad_norm": 0.4256921410560608, "learning_rate": 7.188738464655496e-05, "loss": 0.0393, "step": 31040 }, { "grad_norm": 0.540919840335846, "learning_rate": 7.186879988747533e-05, "loss": 0.0337, "step": 31050 }, { "grad_norm": 0.49367040395736694, "learning_rate": 7.185021139161592e-05, "loss": 0.0367, "step": 31060 }, { "grad_norm": 0.4246712327003479, "learning_rate": 7.1831619162153e-05, "loss": 0.0433, "step": 31070 }, { "grad_norm": 0.49982818961143494, "learning_rate": 7.181302320226345e-05, "loss": 0.0377, "step": 31080 }, { "grad_norm": 0.3229430913925171, "learning_rate": 7.179442351512482e-05, "loss": 0.0421, "step": 31090 }, { "grad_norm": 0.4236776828765869, "learning_rate": 7.177582010391528e-05, "loss": 0.0397, "step": 31100 }, { "grad_norm": 0.4760687053203583, "learning_rate": 7.175721297181366e-05, "loss": 0.0435, "step": 31110 }, { "grad_norm": 0.41726914048194885, "learning_rate": 7.173860212199942e-05, "loss": 0.0424, "step": 31120 }, { "grad_norm": 0.6967774033546448, "learning_rate": 7.171998755765263e-05, "loss": 0.0376, "step": 31130 }, { "grad_norm": 0.36748528480529785, "learning_rate": 7.170136928195398e-05, "loss": 0.039, "step": 31140 }, { "grad_norm": 0.3180055320262909, "learning_rate": 7.168274729808489e-05, "loss": 0.0301, "step": 31150 }, { "grad_norm": 0.4758792519569397, "learning_rate": 7.166412160922728e-05, "loss": 0.0302, "step": 31160 }, { "grad_norm": 0.44292566180229187, "learning_rate": 7.164549221856382e-05, "loss": 0.0444, "step": 31170 }, { "grad_norm": 0.3160606026649475, "learning_rate": 7.162685912927775e-05, "loss": 0.0491, "step": 31180 }, { "grad_norm": 0.4991343021392822, "learning_rate": 7.160822234455294e-05, "loss": 0.0452, "step": 31190 }, { "grad_norm": 0.45010337233543396, "learning_rate": 7.158958186757391e-05, "loss": 0.0365, "step": 31200 }, { "grad_norm": 0.3804376423358917, "learning_rate": 7.157093770152582e-05, "loss": 0.0403, "step": 31210 }, { "grad_norm": 0.365174263715744, "learning_rate": 7.155228984959446e-05, "loss": 0.0362, "step": 31220 }, { "grad_norm": 0.27160823345184326, "learning_rate": 7.153363831496621e-05, "loss": 0.034, "step": 31230 }, { "grad_norm": 0.5220775604248047, "learning_rate": 7.151498310082811e-05, "loss": 0.0413, "step": 31240 }, { "grad_norm": 0.6383504271507263, "learning_rate": 7.149632421036784e-05, "loss": 0.0415, "step": 31250 }, { "grad_norm": 0.5306386947631836, "learning_rate": 7.147766164677369e-05, "loss": 0.0607, "step": 31260 }, { "grad_norm": 0.5671952366828918, "learning_rate": 7.145899541323459e-05, "loss": 0.0373, "step": 31270 }, { "grad_norm": 0.48504573106765747, "learning_rate": 7.144032551294007e-05, "loss": 0.055, "step": 31280 }, { "grad_norm": 0.5943192839622498, "learning_rate": 7.14216519490803e-05, "loss": 0.0447, "step": 31290 }, { "grad_norm": 0.6646921038627625, "learning_rate": 7.140297472484609e-05, "loss": 0.0506, "step": 31300 }, { "grad_norm": 0.49241673946380615, "learning_rate": 7.138429384342891e-05, "loss": 0.0569, "step": 31310 }, { "grad_norm": 0.39144617319107056, "learning_rate": 7.136560930802074e-05, "loss": 0.0354, "step": 31320 }, { "grad_norm": 0.4303186535835266, "learning_rate": 7.134692112181431e-05, "loss": 0.0381, "step": 31330 }, { "grad_norm": 0.2815077602863312, "learning_rate": 7.13282292880029e-05, "loss": 0.0378, "step": 31340 }, { "grad_norm": 0.37768489122390747, "learning_rate": 7.130953380978043e-05, "loss": 0.0423, "step": 31350 }, { "grad_norm": 0.2695557177066803, "learning_rate": 7.129083469034144e-05, "loss": 0.0556, "step": 31360 }, { "grad_norm": 0.33619222044944763, "learning_rate": 7.127213193288112e-05, "loss": 0.0461, "step": 31370 }, { "grad_norm": 0.3256155550479889, "learning_rate": 7.125342554059522e-05, "loss": 0.0298, "step": 31380 }, { "grad_norm": 0.40536269545555115, "learning_rate": 7.12347155166802e-05, "loss": 0.0383, "step": 31390 }, { "grad_norm": 0.3608044683933258, "learning_rate": 7.121600186433306e-05, "loss": 0.037, "step": 31400 }, { "grad_norm": 0.41590508818626404, "learning_rate": 7.119728458675148e-05, "loss": 0.0389, "step": 31410 }, { "grad_norm": 0.36924755573272705, "learning_rate": 7.117856368713369e-05, "loss": 0.0421, "step": 31420 }, { "grad_norm": 0.3674830198287964, "learning_rate": 7.115983916867861e-05, "loss": 0.0358, "step": 31430 }, { "grad_norm": 0.45157966017723083, "learning_rate": 7.114111103458574e-05, "loss": 0.0419, "step": 31440 }, { "grad_norm": 0.295202374458313, "learning_rate": 7.11223792880552e-05, "loss": 0.0407, "step": 31450 }, { "grad_norm": 0.43585005402565, "learning_rate": 7.110364393228773e-05, "loss": 0.0364, "step": 31460 }, { "grad_norm": 0.41606462001800537, "learning_rate": 7.108490497048471e-05, "loss": 0.0283, "step": 31470 }, { "grad_norm": 0.38511720299720764, "learning_rate": 7.10661624058481e-05, "loss": 0.045, "step": 31480 }, { "grad_norm": 0.3909663259983063, "learning_rate": 7.10474162415805e-05, "loss": 0.0411, "step": 31490 }, { "grad_norm": 0.385699063539505, "learning_rate": 7.102866648088511e-05, "loss": 0.0379, "step": 31500 }, { "grad_norm": 0.40740537643432617, "learning_rate": 7.100991312696576e-05, "loss": 0.0348, "step": 31510 }, { "grad_norm": 0.3910471498966217, "learning_rate": 7.099115618302686e-05, "loss": 0.0497, "step": 31520 }, { "grad_norm": 0.5101051926612854, "learning_rate": 7.097239565227349e-05, "loss": 0.0482, "step": 31530 }, { "grad_norm": 0.2943921387195587, "learning_rate": 7.09536315379113e-05, "loss": 0.0363, "step": 31540 }, { "grad_norm": 0.4791998863220215, "learning_rate": 7.093486384314656e-05, "loss": 0.0372, "step": 31550 }, { "grad_norm": 0.5479989051818848, "learning_rate": 7.091609257118616e-05, "loss": 0.0388, "step": 31560 }, { "grad_norm": 0.38533827662467957, "learning_rate": 7.08973177252376e-05, "loss": 0.0389, "step": 31570 }, { "grad_norm": 0.38884398341178894, "learning_rate": 7.087853930850898e-05, "loss": 0.0355, "step": 31580 }, { "grad_norm": 0.5305756330490112, "learning_rate": 7.085975732420903e-05, "loss": 0.0335, "step": 31590 }, { "grad_norm": 0.4160500466823578, "learning_rate": 7.084097177554706e-05, "loss": 0.0305, "step": 31600 }, { "grad_norm": 0.5383226871490479, "learning_rate": 7.082218266573301e-05, "loss": 0.0476, "step": 31610 }, { "grad_norm": 0.419597864151001, "learning_rate": 7.080338999797743e-05, "loss": 0.04, "step": 31620 }, { "grad_norm": 0.3744939863681793, "learning_rate": 7.07845937754915e-05, "loss": 0.0369, "step": 31630 }, { "grad_norm": 0.4896147549152374, "learning_rate": 7.076579400148693e-05, "loss": 0.0421, "step": 31640 }, { "grad_norm": 0.303606241941452, "learning_rate": 7.074699067917611e-05, "loss": 0.0427, "step": 31650 }, { "grad_norm": 0.48894965648651123, "learning_rate": 7.072818381177201e-05, "loss": 0.0367, "step": 31660 }, { "grad_norm": 0.42805927991867065, "learning_rate": 7.070937340248823e-05, "loss": 0.0439, "step": 31670 }, { "grad_norm": 0.3839535713195801, "learning_rate": 7.069055945453893e-05, "loss": 0.0392, "step": 31680 }, { "grad_norm": 0.40828511118888855, "learning_rate": 7.067174197113892e-05, "loss": 0.0373, "step": 31690 }, { "grad_norm": 0.34787434339523315, "learning_rate": 7.065292095550355e-05, "loss": 0.0412, "step": 31700 }, { "grad_norm": 0.5299034118652344, "learning_rate": 7.063409641084887e-05, "loss": 0.039, "step": 31710 }, { "grad_norm": 0.512820303440094, "learning_rate": 7.061526834039145e-05, "loss": 0.0499, "step": 31720 }, { "grad_norm": 0.4006543755531311, "learning_rate": 7.05964367473485e-05, "loss": 0.0284, "step": 31730 }, { "grad_norm": 0.32955482602119446, "learning_rate": 7.057760163493783e-05, "loss": 0.0424, "step": 31740 }, { "grad_norm": 0.41120392084121704, "learning_rate": 7.055876300637783e-05, "loss": 0.0346, "step": 31750 }, { "grad_norm": 0.5105316042900085, "learning_rate": 7.053992086488753e-05, "loss": 0.0438, "step": 31760 }, { "grad_norm": 0.42267096042633057, "learning_rate": 7.052107521368651e-05, "loss": 0.0393, "step": 31770 }, { "grad_norm": 0.37626898288726807, "learning_rate": 7.0502226055995e-05, "loss": 0.047, "step": 31780 }, { "grad_norm": 0.4927668273448944, "learning_rate": 7.048337339503379e-05, "loss": 0.0557, "step": 31790 }, { "grad_norm": 0.44810134172439575, "learning_rate": 7.046451723402427e-05, "loss": 0.0376, "step": 31800 }, { "grad_norm": 0.33802345395088196, "learning_rate": 7.044565757618848e-05, "loss": 0.0531, "step": 31810 }, { "grad_norm": 0.33764538168907166, "learning_rate": 7.042679442474899e-05, "loss": 0.0431, "step": 31820 }, { "grad_norm": 0.8570422530174255, "learning_rate": 7.040792778292902e-05, "loss": 0.0487, "step": 31830 }, { "grad_norm": 0.3196007311344147, "learning_rate": 7.038905765395234e-05, "loss": 0.0526, "step": 31840 }, { "grad_norm": 0.39822882413864136, "learning_rate": 7.037018404104334e-05, "loss": 0.0392, "step": 31850 }, { "grad_norm": 1.087408185005188, "learning_rate": 7.035130694742702e-05, "loss": 0.0391, "step": 31860 }, { "grad_norm": 0.4920200705528259, "learning_rate": 7.033242637632897e-05, "loss": 0.0466, "step": 31870 }, { "grad_norm": 0.5066297650337219, "learning_rate": 7.031354233097534e-05, "loss": 0.0493, "step": 31880 }, { "grad_norm": 0.43958958983421326, "learning_rate": 7.029465481459289e-05, "loss": 0.0526, "step": 31890 }, { "grad_norm": 0.38685342669487, "learning_rate": 7.027576383040898e-05, "loss": 0.0447, "step": 31900 }, { "grad_norm": 0.46558520197868347, "learning_rate": 7.025686938165159e-05, "loss": 0.0498, "step": 31910 }, { "grad_norm": 0.5332244038581848, "learning_rate": 7.023797147154924e-05, "loss": 0.0556, "step": 31920 }, { "grad_norm": 0.34334251284599304, "learning_rate": 7.021907010333111e-05, "loss": 0.04, "step": 31930 }, { "grad_norm": 0.3453975319862366, "learning_rate": 7.020016528022685e-05, "loss": 0.0406, "step": 31940 }, { "grad_norm": 0.4399991035461426, "learning_rate": 7.018125700546683e-05, "loss": 0.0446, "step": 31950 }, { "grad_norm": 0.4323357343673706, "learning_rate": 7.016234528228196e-05, "loss": 0.0456, "step": 31960 }, { "grad_norm": 0.3650605082511902, "learning_rate": 7.014343011390372e-05, "loss": 0.0471, "step": 31970 }, { "grad_norm": 0.32134178280830383, "learning_rate": 7.01245115035642e-05, "loss": 0.0373, "step": 31980 }, { "grad_norm": 0.40730923414230347, "learning_rate": 7.010558945449606e-05, "loss": 0.0471, "step": 31990 }, { "grad_norm": 0.5742233991622925, "learning_rate": 7.008666396993258e-05, "loss": 0.033, "step": 32000 }, { "grad_norm": 0.48975491523742676, "learning_rate": 7.006773505310759e-05, "loss": 0.0453, "step": 32010 }, { "grad_norm": 0.41289496421813965, "learning_rate": 7.004880270725553e-05, "loss": 0.0472, "step": 32020 }, { "grad_norm": 0.6868689060211182, "learning_rate": 7.002986693561144e-05, "loss": 0.044, "step": 32030 }, { "grad_norm": 0.4650551974773407, "learning_rate": 7.001092774141089e-05, "loss": 0.0405, "step": 32040 }, { "grad_norm": 0.3120664954185486, "learning_rate": 6.999198512789009e-05, "loss": 0.0415, "step": 32050 }, { "grad_norm": 0.5135970711708069, "learning_rate": 6.997303909828584e-05, "loss": 0.0516, "step": 32060 }, { "grad_norm": 0.5344609022140503, "learning_rate": 6.995408965583544e-05, "loss": 0.0459, "step": 32070 }, { "grad_norm": 0.39458611607551575, "learning_rate": 6.993513680377688e-05, "loss": 0.0405, "step": 32080 }, { "grad_norm": 0.3777574896812439, "learning_rate": 6.991618054534868e-05, "loss": 0.0403, "step": 32090 }, { "grad_norm": 0.41899070143699646, "learning_rate": 6.989722088378991e-05, "loss": 0.0328, "step": 32100 }, { "grad_norm": 0.4240906834602356, "learning_rate": 6.987825782234027e-05, "loss": 0.0362, "step": 32110 }, { "grad_norm": 0.47570791840553284, "learning_rate": 6.985929136424006e-05, "loss": 0.0344, "step": 32120 }, { "grad_norm": 0.41651588678359985, "learning_rate": 6.984032151273012e-05, "loss": 0.0405, "step": 32130 }, { "grad_norm": 0.5314726829528809, "learning_rate": 6.982134827105186e-05, "loss": 0.0368, "step": 32140 }, { "grad_norm": 0.4106559157371521, "learning_rate": 6.980237164244729e-05, "loss": 0.0329, "step": 32150 }, { "grad_norm": 0.31445983052253723, "learning_rate": 6.9783391630159e-05, "loss": 0.0317, "step": 32160 }, { "grad_norm": 0.49414172768592834, "learning_rate": 6.976440823743015e-05, "loss": 0.0397, "step": 32170 }, { "grad_norm": 0.31175073981285095, "learning_rate": 6.974542146750451e-05, "loss": 0.0299, "step": 32180 }, { "grad_norm": 0.5635617971420288, "learning_rate": 6.972643132362637e-05, "loss": 0.033, "step": 32190 }, { "grad_norm": 0.47250863909721375, "learning_rate": 6.970743780904064e-05, "loss": 0.0399, "step": 32200 }, { "grad_norm": 0.43412911891937256, "learning_rate": 6.968844092699277e-05, "loss": 0.0368, "step": 32210 }, { "grad_norm": 0.33632519841194153, "learning_rate": 6.966944068072883e-05, "loss": 0.0365, "step": 32220 }, { "grad_norm": 0.3785035014152527, "learning_rate": 6.965043707349545e-05, "loss": 0.035, "step": 32230 }, { "grad_norm": 0.34708160161972046, "learning_rate": 6.963143010853982e-05, "loss": 0.0376, "step": 32240 }, { "grad_norm": 0.47436559200286865, "learning_rate": 6.961241978910971e-05, "loss": 0.0399, "step": 32250 }, { "grad_norm": 0.4367481768131256, "learning_rate": 6.959340611845344e-05, "loss": 0.036, "step": 32260 }, { "grad_norm": 0.5079392790794373, "learning_rate": 6.957438909981995e-05, "loss": 0.0305, "step": 32270 }, { "grad_norm": 0.33368733525276184, "learning_rate": 6.955536873645872e-05, "loss": 0.034, "step": 32280 }, { "grad_norm": 0.4729241132736206, "learning_rate": 6.953634503161982e-05, "loss": 0.0401, "step": 32290 }, { "grad_norm": 0.46052345633506775, "learning_rate": 6.951731798855387e-05, "loss": 0.0397, "step": 32300 }, { "grad_norm": 0.516650915145874, "learning_rate": 6.949828761051208e-05, "loss": 0.0482, "step": 32310 }, { "grad_norm": 0.5076043009757996, "learning_rate": 6.947925390074622e-05, "loss": 0.0383, "step": 32320 }, { "grad_norm": 0.30654284358024597, "learning_rate": 6.946021686250863e-05, "loss": 0.0374, "step": 32330 }, { "grad_norm": 0.38454121351242065, "learning_rate": 6.94411764990522e-05, "loss": 0.0383, "step": 32340 }, { "grad_norm": 0.4003232717514038, "learning_rate": 6.942213281363044e-05, "loss": 0.0374, "step": 32350 }, { "grad_norm": 0.34751978516578674, "learning_rate": 6.940308580949737e-05, "loss": 0.0366, "step": 32360 }, { "grad_norm": 0.5030838251113892, "learning_rate": 6.93840354899076e-05, "loss": 0.0351, "step": 32370 }, { "grad_norm": 0.41907674074172974, "learning_rate": 6.936498185811633e-05, "loss": 0.0414, "step": 32380 }, { "grad_norm": 0.3090747594833374, "learning_rate": 6.93459249173793e-05, "loss": 0.0298, "step": 32390 }, { "grad_norm": 0.4195398986339569, "learning_rate": 6.932686467095279e-05, "loss": 0.0388, "step": 32400 }, { "grad_norm": 0.44499534368515015, "learning_rate": 6.930780112209373e-05, "loss": 0.0347, "step": 32410 }, { "grad_norm": 0.40056338906288147, "learning_rate": 6.92887342740595e-05, "loss": 0.037, "step": 32420 }, { "grad_norm": 0.373050332069397, "learning_rate": 6.926966413010816e-05, "loss": 0.0501, "step": 32430 }, { "grad_norm": 0.42897215485572815, "learning_rate": 6.925059069349824e-05, "loss": 0.0332, "step": 32440 }, { "grad_norm": 0.3951212465763092, "learning_rate": 6.923151396748886e-05, "loss": 0.0338, "step": 32450 }, { "grad_norm": 0.33320844173431396, "learning_rate": 6.921243395533974e-05, "loss": 0.0381, "step": 32460 }, { "grad_norm": 0.529500424861908, "learning_rate": 6.919335066031109e-05, "loss": 0.0429, "step": 32470 }, { "grad_norm": 0.29018014669418335, "learning_rate": 6.917426408566379e-05, "loss": 0.0409, "step": 32480 }, { "grad_norm": 0.460147500038147, "learning_rate": 6.915517423465916e-05, "loss": 0.0318, "step": 32490 }, { "grad_norm": 0.598904550075531, "learning_rate": 6.913608111055914e-05, "loss": 0.0417, "step": 32500 }, { "grad_norm": 0.47845757007598877, "learning_rate": 6.911698471662623e-05, "loss": 0.0355, "step": 32510 }, { "grad_norm": 0.24244560301303864, "learning_rate": 6.90978850561235e-05, "loss": 0.0412, "step": 32520 }, { "grad_norm": 0.30380818247795105, "learning_rate": 6.907878213231454e-05, "loss": 0.0337, "step": 32530 }, { "grad_norm": 0.43911629915237427, "learning_rate": 6.90596759484635e-05, "loss": 0.0384, "step": 32540 }, { "grad_norm": 0.3794873356819153, "learning_rate": 6.904056650783514e-05, "loss": 0.0394, "step": 32550 }, { "grad_norm": 0.36790934205055237, "learning_rate": 6.902145381369471e-05, "loss": 0.0405, "step": 32560 }, { "grad_norm": 0.35929304361343384, "learning_rate": 6.900233786930808e-05, "loss": 0.0441, "step": 32570 }, { "grad_norm": 0.5408797264099121, "learning_rate": 6.898321867794161e-05, "loss": 0.0468, "step": 32580 }, { "grad_norm": 0.8114152550697327, "learning_rate": 6.896409624286226e-05, "loss": 0.0378, "step": 32590 }, { "grad_norm": 0.3174239993095398, "learning_rate": 6.894497056733754e-05, "loss": 0.0398, "step": 32600 }, { "grad_norm": 0.4839909076690674, "learning_rate": 6.89258416546355e-05, "loss": 0.0382, "step": 32610 }, { "grad_norm": 0.39984130859375, "learning_rate": 6.890670950802474e-05, "loss": 0.0452, "step": 32620 }, { "grad_norm": 0.26213786005973816, "learning_rate": 6.88875741307744e-05, "loss": 0.0345, "step": 32630 }, { "grad_norm": 0.28550657629966736, "learning_rate": 6.886843552615425e-05, "loss": 0.0442, "step": 32640 }, { "grad_norm": 0.41738685965538025, "learning_rate": 6.884929369743451e-05, "loss": 0.035, "step": 32650 }, { "grad_norm": 0.7934902906417847, "learning_rate": 6.8830148647886e-05, "loss": 0.0382, "step": 32660 }, { "grad_norm": 0.3250032961368561, "learning_rate": 6.88110003807801e-05, "loss": 0.0409, "step": 32670 }, { "grad_norm": 0.5655002593994141, "learning_rate": 6.87918488993887e-05, "loss": 0.0373, "step": 32680 }, { "grad_norm": 0.5087494254112244, "learning_rate": 6.877269420698431e-05, "loss": 0.0416, "step": 32690 }, { "grad_norm": 0.45026740431785583, "learning_rate": 6.875353630683989e-05, "loss": 0.0397, "step": 32700 }, { "grad_norm": 0.6164796948432922, "learning_rate": 6.873437520222905e-05, "loss": 0.0456, "step": 32710 }, { "grad_norm": 0.4483093023300171, "learning_rate": 6.871521089642585e-05, "loss": 0.0348, "step": 32720 }, { "grad_norm": 0.32275164127349854, "learning_rate": 6.869604339270498e-05, "loss": 0.0329, "step": 32730 }, { "grad_norm": 0.34987130761146545, "learning_rate": 6.867687269434164e-05, "loss": 0.0448, "step": 32740 }, { "grad_norm": 0.44213810563087463, "learning_rate": 6.865769880461156e-05, "loss": 0.0423, "step": 32750 }, { "grad_norm": 0.4531612992286682, "learning_rate": 6.863852172679104e-05, "loss": 0.0373, "step": 32760 }, { "grad_norm": 0.39664819836616516, "learning_rate": 6.861934146415693e-05, "loss": 0.0367, "step": 32770 }, { "grad_norm": 0.35879752039909363, "learning_rate": 6.86001580199866e-05, "loss": 0.0436, "step": 32780 }, { "grad_norm": 0.40368887782096863, "learning_rate": 6.858097139755798e-05, "loss": 0.0326, "step": 32790 }, { "grad_norm": 0.3954155147075653, "learning_rate": 6.856178160014955e-05, "loss": 0.0354, "step": 32800 }, { "grad_norm": 0.3626139163970947, "learning_rate": 6.85425886310403e-05, "loss": 0.0348, "step": 32810 }, { "grad_norm": 0.45381537079811096, "learning_rate": 6.852339249350979e-05, "loss": 0.0389, "step": 32820 }, { "grad_norm": 0.5921798944473267, "learning_rate": 6.850419319083812e-05, "loss": 0.0498, "step": 32830 }, { "grad_norm": 0.3427823781967163, "learning_rate": 6.848499072630592e-05, "loss": 0.0296, "step": 32840 }, { "grad_norm": 0.4311017692089081, "learning_rate": 6.846578510319439e-05, "loss": 0.0498, "step": 32850 }, { "grad_norm": 0.3233408033847809, "learning_rate": 6.844657632478519e-05, "loss": 0.0408, "step": 32860 }, { "grad_norm": 0.4812699556350708, "learning_rate": 6.842736439436063e-05, "loss": 0.0393, "step": 32870 }, { "grad_norm": 0.2943544387817383, "learning_rate": 6.84081493152035e-05, "loss": 0.0409, "step": 32880 }, { "grad_norm": 0.6836777329444885, "learning_rate": 6.83889310905971e-05, "loss": 0.0326, "step": 32890 }, { "grad_norm": 0.4534972310066223, "learning_rate": 6.836970972382533e-05, "loss": 0.0389, "step": 32900 }, { "grad_norm": 0.4065239131450653, "learning_rate": 6.835048521817257e-05, "loss": 0.0417, "step": 32910 }, { "grad_norm": 0.4213314354419708, "learning_rate": 6.833125757692379e-05, "loss": 0.0381, "step": 32920 }, { "grad_norm": 0.3556375205516815, "learning_rate": 6.831202680336441e-05, "loss": 0.033, "step": 32930 }, { "grad_norm": 0.30379998683929443, "learning_rate": 6.829279290078052e-05, "loss": 0.0336, "step": 32940 }, { "grad_norm": 0.39199987053871155, "learning_rate": 6.827355587245863e-05, "loss": 0.036, "step": 32950 }, { "grad_norm": 0.5134928226470947, "learning_rate": 6.82543157216858e-05, "loss": 0.0328, "step": 32960 }, { "grad_norm": 0.44198882579803467, "learning_rate": 6.823507245174969e-05, "loss": 0.0373, "step": 32970 }, { "grad_norm": 0.3816802203655243, "learning_rate": 6.821582606593841e-05, "loss": 0.0354, "step": 32980 }, { "grad_norm": 0.38628053665161133, "learning_rate": 6.81965765675407e-05, "loss": 0.0377, "step": 32990 }, { "grad_norm": 0.989875853061676, "learning_rate": 6.81773239598457e-05, "loss": 0.041, "step": 33000 }, { "grad_norm": 0.41845059394836426, "learning_rate": 6.815806824614319e-05, "loss": 0.0452, "step": 33010 }, { "grad_norm": 0.5012903213500977, "learning_rate": 6.813880942972343e-05, "loss": 0.0333, "step": 33020 }, { "grad_norm": 0.427259624004364, "learning_rate": 6.811954751387726e-05, "loss": 0.0357, "step": 33030 }, { "grad_norm": 0.3235084116458893, "learning_rate": 6.810028250189598e-05, "loss": 0.0367, "step": 33040 }, { "grad_norm": 0.4563050866127014, "learning_rate": 6.808101439707147e-05, "loss": 0.041, "step": 33050 }, { "grad_norm": 0.3619121015071869, "learning_rate": 6.806174320269609e-05, "loss": 0.0428, "step": 33060 }, { "grad_norm": 0.39747029542922974, "learning_rate": 6.804246892206281e-05, "loss": 0.0414, "step": 33070 }, { "grad_norm": 0.466653048992157, "learning_rate": 6.802319155846506e-05, "loss": 0.0578, "step": 33080 }, { "grad_norm": 0.4885951578617096, "learning_rate": 6.800391111519679e-05, "loss": 0.0358, "step": 33090 }, { "grad_norm": 0.3956272602081299, "learning_rate": 6.798462759555253e-05, "loss": 0.0405, "step": 33100 }, { "grad_norm": 0.3266656696796417, "learning_rate": 6.79653410028273e-05, "loss": 0.0374, "step": 33110 }, { "grad_norm": 0.2745707631111145, "learning_rate": 6.794605134031663e-05, "loss": 0.0381, "step": 33120 }, { "grad_norm": 0.37776637077331543, "learning_rate": 6.792675861131661e-05, "loss": 0.0332, "step": 33130 }, { "grad_norm": 0.5055871605873108, "learning_rate": 6.790746281912386e-05, "loss": 0.0454, "step": 33140 }, { "grad_norm": 0.28628453612327576, "learning_rate": 6.788816396703546e-05, "loss": 0.0302, "step": 33150 }, { "grad_norm": 0.33233827352523804, "learning_rate": 6.78688620583491e-05, "loss": 0.0353, "step": 33160 }, { "grad_norm": 0.516899049282074, "learning_rate": 6.784955709636292e-05, "loss": 0.0389, "step": 33170 }, { "grad_norm": 0.4437871277332306, "learning_rate": 6.783024908437564e-05, "loss": 0.0384, "step": 33180 }, { "grad_norm": 0.3457331359386444, "learning_rate": 6.781093802568641e-05, "loss": 0.0291, "step": 33190 }, { "grad_norm": 0.4279782772064209, "learning_rate": 6.779162392359504e-05, "loss": 0.0319, "step": 33200 }, { "grad_norm": 0.3427669107913971, "learning_rate": 6.777230678140172e-05, "loss": 0.0354, "step": 33210 }, { "grad_norm": 0.5310565829277039, "learning_rate": 6.775298660240726e-05, "loss": 0.0412, "step": 33220 }, { "grad_norm": 0.34154367446899414, "learning_rate": 6.773366338991292e-05, "loss": 0.0489, "step": 33230 }, { "grad_norm": 0.33546915650367737, "learning_rate": 6.771433714722052e-05, "loss": 0.0305, "step": 33240 }, { "grad_norm": 0.43907493352890015, "learning_rate": 6.769500787763239e-05, "loss": 0.0438, "step": 33250 }, { "grad_norm": 0.6977565288543701, "learning_rate": 6.76756755844514e-05, "loss": 0.0383, "step": 33260 }, { "grad_norm": 0.3987012207508087, "learning_rate": 6.765634027098087e-05, "loss": 0.0354, "step": 33270 }, { "grad_norm": 0.40252605080604553, "learning_rate": 6.763700194052468e-05, "loss": 0.0353, "step": 33280 }, { "grad_norm": 0.441959410905838, "learning_rate": 6.761766059638723e-05, "loss": 0.0382, "step": 33290 }, { "grad_norm": 0.4556809961795807, "learning_rate": 6.759831624187345e-05, "loss": 0.0402, "step": 33300 }, { "grad_norm": 0.5701584815979004, "learning_rate": 6.757896888028871e-05, "loss": 0.0463, "step": 33310 }, { "grad_norm": 0.32164183259010315, "learning_rate": 6.7559618514939e-05, "loss": 0.0305, "step": 33320 }, { "grad_norm": 0.4370386004447937, "learning_rate": 6.754026514913073e-05, "loss": 0.0382, "step": 33330 }, { "grad_norm": 0.46354517340660095, "learning_rate": 6.752090878617087e-05, "loss": 0.0441, "step": 33340 }, { "grad_norm": 0.43585214018821716, "learning_rate": 6.75015494293669e-05, "loss": 0.0361, "step": 33350 }, { "grad_norm": 0.4620259404182434, "learning_rate": 6.74821870820268e-05, "loss": 0.0363, "step": 33360 }, { "grad_norm": 0.367848664522171, "learning_rate": 6.746282174745907e-05, "loss": 0.0366, "step": 33370 }, { "grad_norm": 0.454277902841568, "learning_rate": 6.744345342897271e-05, "loss": 0.0381, "step": 33380 }, { "grad_norm": 0.42098483443260193, "learning_rate": 6.742408212987724e-05, "loss": 0.0453, "step": 33390 }, { "grad_norm": 0.4210885167121887, "learning_rate": 6.740470785348269e-05, "loss": 0.0305, "step": 33400 }, { "grad_norm": 0.3134020268917084, "learning_rate": 6.738533060309958e-05, "loss": 0.0372, "step": 33410 }, { "grad_norm": 0.4971160590648651, "learning_rate": 6.736595038203894e-05, "loss": 0.0358, "step": 33420 }, { "grad_norm": 0.5596634149551392, "learning_rate": 6.734656719361236e-05, "loss": 0.0465, "step": 33430 }, { "grad_norm": 0.38362234830856323, "learning_rate": 6.732718104113189e-05, "loss": 0.0418, "step": 33440 }, { "grad_norm": 0.4988180696964264, "learning_rate": 6.730779192791006e-05, "loss": 0.0387, "step": 33450 }, { "grad_norm": 0.4056105613708496, "learning_rate": 6.728839985725997e-05, "loss": 0.0359, "step": 33460 }, { "grad_norm": 0.4231628179550171, "learning_rate": 6.726900483249517e-05, "loss": 0.0341, "step": 33470 }, { "grad_norm": 0.579829216003418, "learning_rate": 6.724960685692976e-05, "loss": 0.0357, "step": 33480 }, { "grad_norm": 0.5120840072631836, "learning_rate": 6.723020593387833e-05, "loss": 0.0373, "step": 33490 }, { "grad_norm": 0.484369695186615, "learning_rate": 6.721080206665593e-05, "loss": 0.0441, "step": 33500 }, { "grad_norm": 0.35641664266586304, "learning_rate": 6.719139525857819e-05, "loss": 0.0497, "step": 33510 }, { "grad_norm": 0.4658120572566986, "learning_rate": 6.717198551296117e-05, "loss": 0.0357, "step": 33520 }, { "grad_norm": 0.5124161839485168, "learning_rate": 6.715257283312148e-05, "loss": 0.0417, "step": 33530 }, { "grad_norm": 0.538946270942688, "learning_rate": 6.713315722237623e-05, "loss": 0.0432, "step": 33540 }, { "grad_norm": 0.30936262011528015, "learning_rate": 6.7113738684043e-05, "loss": 0.0406, "step": 33550 }, { "grad_norm": 2.5029890537261963, "learning_rate": 6.709431722143989e-05, "loss": 0.0433, "step": 33560 }, { "grad_norm": 0.49041512608528137, "learning_rate": 6.70748928378855e-05, "loss": 0.0434, "step": 33570 }, { "grad_norm": 0.49639779329299927, "learning_rate": 6.705546553669891e-05, "loss": 0.0363, "step": 33580 }, { "grad_norm": 0.4639999270439148, "learning_rate": 6.703603532119974e-05, "loss": 0.0358, "step": 33590 }, { "grad_norm": 0.24131976068019867, "learning_rate": 6.701660219470808e-05, "loss": 0.0319, "step": 33600 }, { "grad_norm": 0.47930994629859924, "learning_rate": 6.69971661605445e-05, "loss": 0.0325, "step": 33610 }, { "grad_norm": 0.6776825785636902, "learning_rate": 6.697772722203008e-05, "loss": 0.0452, "step": 33620 }, { "grad_norm": 0.47835075855255127, "learning_rate": 6.695828538248643e-05, "loss": 0.0355, "step": 33630 }, { "grad_norm": 0.29880255460739136, "learning_rate": 6.693884064523563e-05, "loss": 0.0387, "step": 33640 }, { "grad_norm": 0.4035781919956207, "learning_rate": 6.691939301360023e-05, "loss": 0.0413, "step": 33650 }, { "grad_norm": 0.2901895344257355, "learning_rate": 6.689994249090333e-05, "loss": 0.0323, "step": 33660 }, { "grad_norm": 0.4102861285209656, "learning_rate": 6.688048908046845e-05, "loss": 0.0391, "step": 33670 }, { "grad_norm": 0.5444868803024292, "learning_rate": 6.686103278561969e-05, "loss": 0.0312, "step": 33680 }, { "grad_norm": 0.5356267690658569, "learning_rate": 6.684157360968156e-05, "loss": 0.0353, "step": 33690 }, { "grad_norm": 0.43437543511390686, "learning_rate": 6.682211155597911e-05, "loss": 0.0294, "step": 33700 }, { "grad_norm": 0.4615095555782318, "learning_rate": 6.680264662783789e-05, "loss": 0.0377, "step": 33710 }, { "grad_norm": 0.5954592823982239, "learning_rate": 6.678317882858391e-05, "loss": 0.0416, "step": 33720 }, { "grad_norm": 0.34845227003097534, "learning_rate": 6.67637081615437e-05, "loss": 0.0361, "step": 33730 }, { "grad_norm": 0.3614341914653778, "learning_rate": 6.674423463004427e-05, "loss": 0.0357, "step": 33740 }, { "grad_norm": 0.5487259030342102, "learning_rate": 6.672475823741308e-05, "loss": 0.0353, "step": 33750 }, { "grad_norm": 0.5917842388153076, "learning_rate": 6.670527898697811e-05, "loss": 0.0377, "step": 33760 }, { "grad_norm": 0.41405075788497925, "learning_rate": 6.668579688206788e-05, "loss": 0.038, "step": 33770 }, { "grad_norm": 0.3514423370361328, "learning_rate": 6.666631192601131e-05, "loss": 0.0346, "step": 33780 }, { "grad_norm": 0.30736467242240906, "learning_rate": 6.664682412213785e-05, "loss": 0.0339, "step": 33790 }, { "grad_norm": 0.37586233019828796, "learning_rate": 6.662733347377745e-05, "loss": 0.0393, "step": 33800 }, { "grad_norm": 0.39934247732162476, "learning_rate": 6.660783998426051e-05, "loss": 0.0337, "step": 33810 }, { "grad_norm": 0.2619069814682007, "learning_rate": 6.658834365691794e-05, "loss": 0.0338, "step": 33820 }, { "grad_norm": 0.30611956119537354, "learning_rate": 6.656884449508115e-05, "loss": 0.0301, "step": 33830 }, { "grad_norm": 1.1052825450897217, "learning_rate": 6.654934250208198e-05, "loss": 0.0373, "step": 33840 }, { "grad_norm": 0.39502957463264465, "learning_rate": 6.65298376812528e-05, "loss": 0.0323, "step": 33850 }, { "grad_norm": 0.5050266981124878, "learning_rate": 6.651033003592646e-05, "loss": 0.0354, "step": 33860 }, { "grad_norm": 0.4566768407821655, "learning_rate": 6.649081956943626e-05, "loss": 0.0346, "step": 33870 }, { "grad_norm": 0.6386871933937073, "learning_rate": 6.647130628511604e-05, "loss": 0.0399, "step": 33880 }, { "grad_norm": 0.36975082755088806, "learning_rate": 6.645179018630005e-05, "loss": 0.0336, "step": 33890 }, { "grad_norm": 0.6003490686416626, "learning_rate": 6.643227127632309e-05, "loss": 0.039, "step": 33900 }, { "grad_norm": 0.5516213178634644, "learning_rate": 6.641274955852038e-05, "loss": 0.0447, "step": 33910 }, { "grad_norm": 0.4592325985431671, "learning_rate": 6.639322503622768e-05, "loss": 0.0512, "step": 33920 }, { "grad_norm": 0.41513362526893616, "learning_rate": 6.637369771278116e-05, "loss": 0.0361, "step": 33930 }, { "grad_norm": 0.3203142583370209, "learning_rate": 6.635416759151751e-05, "loss": 0.0323, "step": 33940 }, { "grad_norm": 0.7024002075195312, "learning_rate": 6.633463467577394e-05, "loss": 0.0592, "step": 33950 }, { "grad_norm": 0.34478089213371277, "learning_rate": 6.631509896888803e-05, "loss": 0.0358, "step": 33960 }, { "grad_norm": 0.3088957965373993, "learning_rate": 6.629556047419794e-05, "loss": 0.0337, "step": 33970 }, { "grad_norm": 0.5506462454795837, "learning_rate": 6.627601919504223e-05, "loss": 0.0464, "step": 33980 }, { "grad_norm": 0.33197322487831116, "learning_rate": 6.625647513476001e-05, "loss": 0.0346, "step": 33990 }, { "grad_norm": 0.6444472074508667, "learning_rate": 6.62369282966908e-05, "loss": 0.0462, "step": 34000 }, { "grad_norm": 0.22257620096206665, "learning_rate": 6.621737868417464e-05, "loss": 0.0344, "step": 34010 }, { "grad_norm": 0.3064539134502411, "learning_rate": 6.619782630055198e-05, "loss": 0.046, "step": 34020 }, { "grad_norm": 0.4169784188270569, "learning_rate": 6.617827114916382e-05, "loss": 0.0412, "step": 34030 }, { "grad_norm": 0.5475894212722778, "learning_rate": 6.615871323335161e-05, "loss": 0.0386, "step": 34040 }, { "grad_norm": 0.4663859009742737, "learning_rate": 6.613915255645725e-05, "loss": 0.0411, "step": 34050 }, { "grad_norm": 0.32885828614234924, "learning_rate": 6.611958912182312e-05, "loss": 0.0382, "step": 34060 }, { "grad_norm": 0.3831658363342285, "learning_rate": 6.610002293279207e-05, "loss": 0.0464, "step": 34070 }, { "grad_norm": 0.3802947700023651, "learning_rate": 6.608045399270746e-05, "loss": 0.0386, "step": 34080 }, { "grad_norm": 0.4627615809440613, "learning_rate": 6.606088230491304e-05, "loss": 0.0399, "step": 34090 }, { "grad_norm": 0.5392831563949585, "learning_rate": 6.604130787275312e-05, "loss": 0.0429, "step": 34100 }, { "grad_norm": 0.26142117381095886, "learning_rate": 6.602173069957242e-05, "loss": 0.0272, "step": 34110 }, { "grad_norm": 0.41583648324012756, "learning_rate": 6.600215078871612e-05, "loss": 0.0347, "step": 34120 }, { "grad_norm": 0.3953542709350586, "learning_rate": 6.598256814352992e-05, "loss": 0.0419, "step": 34130 }, { "grad_norm": 0.44517093896865845, "learning_rate": 6.596298276735995e-05, "loss": 0.0383, "step": 34140 }, { "grad_norm": 0.5829938054084778, "learning_rate": 6.594339466355282e-05, "loss": 0.0379, "step": 34150 }, { "grad_norm": 0.4664449393749237, "learning_rate": 6.592380383545558e-05, "loss": 0.0394, "step": 34160 }, { "grad_norm": 0.6131563186645508, "learning_rate": 6.590421028641577e-05, "loss": 0.0429, "step": 34170 }, { "grad_norm": 0.5027982592582703, "learning_rate": 6.588461401978143e-05, "loss": 0.0444, "step": 34180 }, { "grad_norm": 0.31923699378967285, "learning_rate": 6.586501503890099e-05, "loss": 0.0347, "step": 34190 }, { "grad_norm": 0.31188637018203735, "learning_rate": 6.584541334712338e-05, "loss": 0.0357, "step": 34200 }, { "grad_norm": 0.291858047246933, "learning_rate": 6.582580894779802e-05, "loss": 0.0349, "step": 34210 }, { "grad_norm": 0.5854943990707397, "learning_rate": 6.580620184427473e-05, "loss": 0.0379, "step": 34220 }, { "grad_norm": 0.437336266040802, "learning_rate": 6.578659203990385e-05, "loss": 0.031, "step": 34230 }, { "grad_norm": 0.34352561831474304, "learning_rate": 6.576697953803615e-05, "loss": 0.0316, "step": 34240 }, { "grad_norm": 0.29420390725135803, "learning_rate": 6.57473643420229e-05, "loss": 0.0379, "step": 34250 }, { "grad_norm": 0.4297981560230255, "learning_rate": 6.572774645521574e-05, "loss": 0.0399, "step": 34260 }, { "grad_norm": 0.4563969373703003, "learning_rate": 6.570812588096688e-05, "loss": 0.0431, "step": 34270 }, { "grad_norm": 0.26020348072052, "learning_rate": 6.568850262262893e-05, "loss": 0.043, "step": 34280 }, { "grad_norm": 0.5341201424598694, "learning_rate": 6.566887668355497e-05, "loss": 0.0374, "step": 34290 }, { "grad_norm": 0.27353063225746155, "learning_rate": 6.564924806709851e-05, "loss": 0.034, "step": 34300 }, { "grad_norm": 0.5215060114860535, "learning_rate": 6.562961677661359e-05, "loss": 0.039, "step": 34310 }, { "grad_norm": 0.5188571214675903, "learning_rate": 6.56099828154546e-05, "loss": 0.031, "step": 34320 }, { "grad_norm": 0.48110225796699524, "learning_rate": 6.55903461869765e-05, "loss": 0.0338, "step": 34330 }, { "grad_norm": 0.2964994013309479, "learning_rate": 6.557070689453465e-05, "loss": 0.0304, "step": 34340 }, { "grad_norm": 0.3065626621246338, "learning_rate": 6.555106494148482e-05, "loss": 0.0357, "step": 34350 }, { "grad_norm": 0.3528689444065094, "learning_rate": 6.553142033118333e-05, "loss": 0.0368, "step": 34360 }, { "grad_norm": 0.42826053500175476, "learning_rate": 6.551177306698688e-05, "loss": 0.0483, "step": 34370 }, { "grad_norm": 0.6560186743736267, "learning_rate": 6.549212315225267e-05, "loss": 0.0443, "step": 34380 }, { "grad_norm": 0.28736552596092224, "learning_rate": 6.547247059033833e-05, "loss": 0.0395, "step": 34390 }, { "grad_norm": 0.49749505519866943, "learning_rate": 6.545281538460193e-05, "loss": 0.0401, "step": 34400 }, { "grad_norm": 0.3286372125148773, "learning_rate": 6.543315753840202e-05, "loss": 0.0321, "step": 34410 }, { "grad_norm": 0.4492506682872772, "learning_rate": 6.541349705509758e-05, "loss": 0.0417, "step": 34420 }, { "grad_norm": 0.2098393440246582, "learning_rate": 6.539383393804805e-05, "loss": 0.038, "step": 34430 }, { "grad_norm": 0.2667076587677002, "learning_rate": 6.537416819061333e-05, "loss": 0.0438, "step": 34440 }, { "grad_norm": 0.4390351474285126, "learning_rate": 6.535449981615375e-05, "loss": 0.0406, "step": 34450 }, { "grad_norm": 0.3239001929759979, "learning_rate": 6.53348288180301e-05, "loss": 0.0334, "step": 34460 }, { "grad_norm": 0.4522438943386078, "learning_rate": 6.531515519960361e-05, "loss": 0.0499, "step": 34470 }, { "grad_norm": 0.37588658928871155, "learning_rate": 6.529547896423597e-05, "loss": 0.0357, "step": 34480 }, { "grad_norm": 0.3860267102718353, "learning_rate": 6.52758001152893e-05, "loss": 0.0371, "step": 34490 }, { "grad_norm": 0.23203477263450623, "learning_rate": 6.525611865612618e-05, "loss": 0.0349, "step": 34500 }, { "grad_norm": 0.4306982159614563, "learning_rate": 6.523643459010966e-05, "loss": 0.0455, "step": 34510 }, { "grad_norm": 0.45966029167175293, "learning_rate": 6.521674792060317e-05, "loss": 0.0427, "step": 34520 }, { "grad_norm": 0.5151593089103699, "learning_rate": 6.519705865097063e-05, "loss": 0.0442, "step": 34530 }, { "grad_norm": 0.46521079540252686, "learning_rate": 6.517736678457641e-05, "loss": 0.0318, "step": 34540 }, { "grad_norm": 0.36196020245552063, "learning_rate": 6.515767232478534e-05, "loss": 0.0326, "step": 34550 }, { "grad_norm": 0.5865984559059143, "learning_rate": 6.51379752749626e-05, "loss": 0.0406, "step": 34560 }, { "grad_norm": 0.40978002548217773, "learning_rate": 6.511827563847393e-05, "loss": 0.0305, "step": 34570 }, { "grad_norm": 1.0176770687103271, "learning_rate": 6.509857341868542e-05, "loss": 0.0437, "step": 34580 }, { "grad_norm": 0.5213173627853394, "learning_rate": 6.507886861896367e-05, "loss": 0.0316, "step": 34590 }, { "grad_norm": 0.4422640800476074, "learning_rate": 6.505916124267567e-05, "loss": 0.0462, "step": 34600 }, { "grad_norm": 0.6748618483543396, "learning_rate": 6.503945129318891e-05, "loss": 0.046, "step": 34610 }, { "grad_norm": 0.33593207597732544, "learning_rate": 6.501973877387122e-05, "loss": 0.0372, "step": 34620 }, { "grad_norm": 0.5353114604949951, "learning_rate": 6.500002368809098e-05, "loss": 0.0343, "step": 34630 }, { "grad_norm": 0.519262969493866, "learning_rate": 6.498030603921694e-05, "loss": 0.0411, "step": 34640 }, { "grad_norm": 0.33774247765541077, "learning_rate": 6.496058583061832e-05, "loss": 0.0337, "step": 34650 }, { "grad_norm": 0.41669961810112, "learning_rate": 6.494086306566475e-05, "loss": 0.0251, "step": 34660 }, { "grad_norm": 0.38710519671440125, "learning_rate": 6.492113774772632e-05, "loss": 0.0331, "step": 34670 }, { "grad_norm": 0.42033326625823975, "learning_rate": 6.490140988017354e-05, "loss": 0.0403, "step": 34680 }, { "grad_norm": 0.43923866748809814, "learning_rate": 6.488167946637736e-05, "loss": 0.0392, "step": 34690 }, { "grad_norm": 0.46003642678260803, "learning_rate": 6.486194650970915e-05, "loss": 0.0416, "step": 34700 }, { "grad_norm": 0.3540642261505127, "learning_rate": 6.48422110135408e-05, "loss": 0.0389, "step": 34710 }, { "grad_norm": 0.3988944888114929, "learning_rate": 6.482247298124451e-05, "loss": 0.0422, "step": 34720 }, { "grad_norm": 0.5820536017417908, "learning_rate": 6.480273241619297e-05, "loss": 0.0358, "step": 34730 }, { "grad_norm": 0.4582962095737457, "learning_rate": 6.478298932175933e-05, "loss": 0.0406, "step": 34740 }, { "grad_norm": 0.3098306357860565, "learning_rate": 6.476324370131712e-05, "loss": 0.0393, "step": 34750 }, { "grad_norm": 0.4925588369369507, "learning_rate": 6.474349555824036e-05, "loss": 0.0478, "step": 34760 }, { "grad_norm": 0.2787818908691406, "learning_rate": 6.472374489590342e-05, "loss": 0.0251, "step": 34770 }, { "grad_norm": 0.36726707220077515, "learning_rate": 6.470399171768118e-05, "loss": 0.0427, "step": 34780 }, { "grad_norm": 0.43888765573501587, "learning_rate": 6.468423602694891e-05, "loss": 0.0344, "step": 34790 }, { "grad_norm": 0.5336930751800537, "learning_rate": 6.466447782708232e-05, "loss": 0.0389, "step": 34800 }, { "grad_norm": 0.8078598976135254, "learning_rate": 6.464471712145754e-05, "loss": 0.0377, "step": 34810 }, { "grad_norm": 0.5844798684120178, "learning_rate": 6.462495391345114e-05, "loss": 0.0398, "step": 34820 }, { "grad_norm": 0.46768054366111755, "learning_rate": 6.46051882064401e-05, "loss": 0.0372, "step": 34830 }, { "grad_norm": 0.4844091832637787, "learning_rate": 6.458542000380186e-05, "loss": 0.0344, "step": 34840 }, { "grad_norm": 0.4253314733505249, "learning_rate": 6.456564930891424e-05, "loss": 0.0373, "step": 34850 }, { "grad_norm": 0.4129495322704315, "learning_rate": 6.454587612515555e-05, "loss": 0.0454, "step": 34860 }, { "grad_norm": 0.38108736276626587, "learning_rate": 6.452610045590444e-05, "loss": 0.0496, "step": 34870 }, { "grad_norm": 0.42690104246139526, "learning_rate": 6.450632230454005e-05, "loss": 0.0349, "step": 34880 }, { "grad_norm": 0.838356614112854, "learning_rate": 6.448654167444195e-05, "loss": 0.0475, "step": 34890 }, { "grad_norm": 0.3550873398780823, "learning_rate": 6.446675856899005e-05, "loss": 0.0341, "step": 34900 }, { "grad_norm": 0.49694934487342834, "learning_rate": 6.444697299156481e-05, "loss": 0.0357, "step": 34910 }, { "grad_norm": 0.3583577871322632, "learning_rate": 6.442718494554701e-05, "loss": 0.0372, "step": 34920 }, { "grad_norm": 0.2613445818424225, "learning_rate": 6.440739443431787e-05, "loss": 0.0293, "step": 34930 }, { "grad_norm": 0.45370036363601685, "learning_rate": 6.438760146125906e-05, "loss": 0.0329, "step": 34940 }, { "grad_norm": 0.32397738099098206, "learning_rate": 6.436780602975267e-05, "loss": 0.0414, "step": 34950 }, { "grad_norm": 0.4564352333545685, "learning_rate": 6.43480081431812e-05, "loss": 0.039, "step": 34960 }, { "grad_norm": 0.48585474491119385, "learning_rate": 6.432820780492756e-05, "loss": 0.0421, "step": 34970 }, { "grad_norm": 0.38566839694976807, "learning_rate": 6.430840501837506e-05, "loss": 0.0484, "step": 34980 }, { "grad_norm": 0.2796672284603119, "learning_rate": 6.428859978690748e-05, "loss": 0.0337, "step": 34990 }, { "grad_norm": 0.3524704575538635, "learning_rate": 6.426879211390901e-05, "loss": 0.0348, "step": 35000 }, { "grad_norm": 0.5965819358825684, "learning_rate": 6.424898200276422e-05, "loss": 0.0393, "step": 35010 }, { "grad_norm": 0.5375567078590393, "learning_rate": 6.42291694568581e-05, "loss": 0.0389, "step": 35020 }, { "grad_norm": 0.408247172832489, "learning_rate": 6.42093544795761e-05, "loss": 0.0424, "step": 35030 }, { "grad_norm": 0.43670937418937683, "learning_rate": 6.418953707430403e-05, "loss": 0.0374, "step": 35040 }, { "grad_norm": 0.3524174988269806, "learning_rate": 6.416971724442819e-05, "loss": 0.0296, "step": 35050 }, { "grad_norm": 0.3430725336074829, "learning_rate": 6.414989499333519e-05, "loss": 0.0345, "step": 35060 }, { "grad_norm": 0.5195139646530151, "learning_rate": 6.413007032441214e-05, "loss": 0.0408, "step": 35070 }, { "grad_norm": 0.35975638031959534, "learning_rate": 6.411024324104653e-05, "loss": 0.036, "step": 35080 }, { "grad_norm": 0.3984256684780121, "learning_rate": 6.409041374662628e-05, "loss": 0.0329, "step": 35090 }, { "grad_norm": 0.34636378288269043, "learning_rate": 6.407058184453967e-05, "loss": 0.0522, "step": 35100 }, { "grad_norm": 0.3417045772075653, "learning_rate": 6.405074753817548e-05, "loss": 0.0391, "step": 35110 }, { "grad_norm": 0.36982470750808716, "learning_rate": 6.40309108309228e-05, "loss": 0.0317, "step": 35120 }, { "grad_norm": 0.46107980608940125, "learning_rate": 6.401107172617122e-05, "loss": 0.0358, "step": 35130 }, { "grad_norm": 0.3393160104751587, "learning_rate": 6.399123022731068e-05, "loss": 0.0487, "step": 35140 }, { "grad_norm": 0.414142906665802, "learning_rate": 6.397138633773157e-05, "loss": 0.0328, "step": 35150 }, { "grad_norm": 0.39658263325691223, "learning_rate": 6.395154006082463e-05, "loss": 0.0343, "step": 35160 }, { "grad_norm": 0.434637188911438, "learning_rate": 6.393169139998109e-05, "loss": 0.0385, "step": 35170 }, { "grad_norm": 0.4116481840610504, "learning_rate": 6.39118403585925e-05, "loss": 0.0351, "step": 35180 }, { "grad_norm": 0.6741134524345398, "learning_rate": 6.38919869400509e-05, "loss": 0.0393, "step": 35190 }, { "grad_norm": 0.39971014857292175, "learning_rate": 6.387213114774865e-05, "loss": 0.0388, "step": 35200 }, { "grad_norm": 0.29184526205062866, "learning_rate": 6.385227298507863e-05, "loss": 0.0519, "step": 35210 }, { "grad_norm": 0.38630571961402893, "learning_rate": 6.3832412455434e-05, "loss": 0.0307, "step": 35220 }, { "grad_norm": 0.2966330945491791, "learning_rate": 6.381254956220841e-05, "loss": 0.0287, "step": 35230 }, { "grad_norm": 0.29700878262519836, "learning_rate": 6.379268430879586e-05, "loss": 0.0294, "step": 35240 }, { "grad_norm": 0.3066340982913971, "learning_rate": 6.37728166985908e-05, "loss": 0.0326, "step": 35250 }, { "grad_norm": 0.36591199040412903, "learning_rate": 6.375294673498804e-05, "loss": 0.0411, "step": 35260 }, { "grad_norm": 0.5109749436378479, "learning_rate": 6.373307442138284e-05, "loss": 0.0396, "step": 35270 }, { "grad_norm": 0.49697253108024597, "learning_rate": 6.371319976117081e-05, "loss": 0.0455, "step": 35280 }, { "grad_norm": 0.36649614572525024, "learning_rate": 6.3693322757748e-05, "loss": 0.033, "step": 35290 }, { "grad_norm": 0.3418906629085541, "learning_rate": 6.367344341451086e-05, "loss": 0.033, "step": 35300 }, { "grad_norm": 0.3071225881576538, "learning_rate": 6.36535617348562e-05, "loss": 0.0383, "step": 35310 }, { "grad_norm": 0.5274714827537537, "learning_rate": 6.363367772218128e-05, "loss": 0.0441, "step": 35320 }, { "grad_norm": 0.4846079349517822, "learning_rate": 6.36137913798837e-05, "loss": 0.0322, "step": 35330 }, { "grad_norm": 0.3951490521430969, "learning_rate": 6.359390271136151e-05, "loss": 0.0316, "step": 35340 }, { "grad_norm": 0.5253846049308777, "learning_rate": 6.357401172001314e-05, "loss": 0.0404, "step": 35350 }, { "grad_norm": 0.22470910847187042, "learning_rate": 6.355411840923742e-05, "loss": 0.0312, "step": 35360 }, { "grad_norm": 0.47057536244392395, "learning_rate": 6.353422278243358e-05, "loss": 0.0427, "step": 35370 }, { "grad_norm": 0.35990530252456665, "learning_rate": 6.351432484300121e-05, "loss": 0.0409, "step": 35380 }, { "grad_norm": 0.5167772173881531, "learning_rate": 6.349442459434036e-05, "loss": 0.0394, "step": 35390 }, { "grad_norm": 0.4372369050979614, "learning_rate": 6.34745220398514e-05, "loss": 0.042, "step": 35400 }, { "grad_norm": 0.4516526758670807, "learning_rate": 6.345461718293518e-05, "loss": 0.0307, "step": 35410 }, { "grad_norm": 0.4153570830821991, "learning_rate": 6.343471002699286e-05, "loss": 0.0414, "step": 35420 }, { "grad_norm": 0.4106340706348419, "learning_rate": 6.341480057542602e-05, "loss": 0.0344, "step": 35430 }, { "grad_norm": 0.4243904948234558, "learning_rate": 6.339488883163667e-05, "loss": 0.0408, "step": 35440 }, { "grad_norm": 0.37498942017555237, "learning_rate": 6.337497479902716e-05, "loss": 0.0317, "step": 35450 }, { "grad_norm": 0.3689463138580322, "learning_rate": 6.335505848100027e-05, "loss": 0.0413, "step": 35460 }, { "grad_norm": 0.7822089791297913, "learning_rate": 6.333513988095915e-05, "loss": 0.0339, "step": 35470 }, { "grad_norm": 0.28801071643829346, "learning_rate": 6.331521900230735e-05, "loss": 0.0306, "step": 35480 }, { "grad_norm": 0.395772248506546, "learning_rate": 6.329529584844878e-05, "loss": 0.0329, "step": 35490 }, { "grad_norm": 0.4565408527851105, "learning_rate": 6.327537042278777e-05, "loss": 0.0344, "step": 35500 }, { "grad_norm": 0.5200827717781067, "learning_rate": 6.325544272872905e-05, "loss": 0.0367, "step": 35510 }, { "grad_norm": 0.31957924365997314, "learning_rate": 6.323551276967771e-05, "loss": 0.0353, "step": 35520 }, { "grad_norm": 0.3739723861217499, "learning_rate": 6.321558054903922e-05, "loss": 0.0328, "step": 35530 }, { "grad_norm": 0.3201850652694702, "learning_rate": 6.319564607021947e-05, "loss": 0.0358, "step": 35540 }, { "grad_norm": 0.35669994354248047, "learning_rate": 6.31757093366247e-05, "loss": 0.0327, "step": 35550 }, { "grad_norm": 0.38408657908439636, "learning_rate": 6.315577035166154e-05, "loss": 0.0407, "step": 35560 }, { "grad_norm": 0.3819826543331146, "learning_rate": 6.313582911873708e-05, "loss": 0.0332, "step": 35570 }, { "grad_norm": 0.5000622868537903, "learning_rate": 6.311588564125865e-05, "loss": 0.0329, "step": 35580 }, { "grad_norm": 0.5142940282821655, "learning_rate": 6.30959399226341e-05, "loss": 0.0428, "step": 35590 }, { "grad_norm": 0.3460054397583008, "learning_rate": 6.30759919662716e-05, "loss": 0.0423, "step": 35600 }, { "grad_norm": 0.2961997389793396, "learning_rate": 6.30560417755797e-05, "loss": 0.0299, "step": 35610 }, { "grad_norm": 0.294121652841568, "learning_rate": 6.303608935396735e-05, "loss": 0.032, "step": 35620 }, { "grad_norm": 0.4863431751728058, "learning_rate": 6.301613470484386e-05, "loss": 0.036, "step": 35630 }, { "grad_norm": 0.3676512837409973, "learning_rate": 6.299617783161893e-05, "loss": 0.0364, "step": 35640 }, { "grad_norm": 0.6573317050933838, "learning_rate": 6.297621873770266e-05, "loss": 0.038, "step": 35650 }, { "grad_norm": 0.3612161874771118, "learning_rate": 6.29562574265055e-05, "loss": 0.034, "step": 35660 }, { "grad_norm": 0.6666446328163147, "learning_rate": 6.293629390143834e-05, "loss": 0.0381, "step": 35670 }, { "grad_norm": 0.3959156572818756, "learning_rate": 6.291632816591232e-05, "loss": 0.0372, "step": 35680 }, { "grad_norm": 0.4494524896144867, "learning_rate": 6.28963602233391e-05, "loss": 0.0313, "step": 35690 }, { "grad_norm": 0.4201417565345764, "learning_rate": 6.287639007713062e-05, "loss": 0.0417, "step": 35700 }, { "grad_norm": 0.2838888466358185, "learning_rate": 6.285641773069926e-05, "loss": 0.0334, "step": 35710 }, { "grad_norm": 0.19686947762966156, "learning_rate": 6.283644318745773e-05, "loss": 0.0278, "step": 35720 }, { "grad_norm": 0.5223777890205383, "learning_rate": 6.281646645081912e-05, "loss": 0.0375, "step": 35730 }, { "grad_norm": 0.3100718557834625, "learning_rate": 6.279648752419693e-05, "loss": 0.0315, "step": 35740 }, { "grad_norm": 0.5591443181037903, "learning_rate": 6.2776506411005e-05, "loss": 0.0393, "step": 35750 }, { "grad_norm": 0.4648686647415161, "learning_rate": 6.275652311465758e-05, "loss": 0.0362, "step": 35760 }, { "grad_norm": 0.3190547227859497, "learning_rate": 6.273653763856926e-05, "loss": 0.0362, "step": 35770 }, { "grad_norm": 0.3191397488117218, "learning_rate": 6.271654998615501e-05, "loss": 0.0343, "step": 35780 }, { "grad_norm": 0.3042318522930145, "learning_rate": 6.269656016083013e-05, "loss": 0.0308, "step": 35790 }, { "grad_norm": 0.3631298243999481, "learning_rate": 6.267656816601038e-05, "loss": 0.043, "step": 35800 }, { "grad_norm": 0.3870183825492859, "learning_rate": 6.265657400511185e-05, "loss": 0.0409, "step": 35810 }, { "grad_norm": 0.39566048979759216, "learning_rate": 6.263657768155098e-05, "loss": 0.0406, "step": 35820 }, { "grad_norm": 0.3121087849140167, "learning_rate": 6.261657919874457e-05, "loss": 0.0373, "step": 35830 }, { "grad_norm": 0.41038957238197327, "learning_rate": 6.259657856010986e-05, "loss": 0.037, "step": 35840 }, { "grad_norm": 0.4396035969257355, "learning_rate": 6.257657576906439e-05, "loss": 0.0324, "step": 35850 }, { "grad_norm": 0.512620747089386, "learning_rate": 6.255657082902609e-05, "loss": 0.031, "step": 35860 }, { "grad_norm": 0.5060451626777649, "learning_rate": 6.253656374341325e-05, "loss": 0.0313, "step": 35870 }, { "grad_norm": 0.888392448425293, "learning_rate": 6.251655451564457e-05, "loss": 0.0411, "step": 35880 }, { "grad_norm": 0.3782109320163727, "learning_rate": 6.249654314913902e-05, "loss": 0.0307, "step": 35890 }, { "grad_norm": 0.5788094401359558, "learning_rate": 6.247652964731604e-05, "loss": 0.0347, "step": 35900 }, { "grad_norm": 0.42065030336380005, "learning_rate": 6.245651401359537e-05, "loss": 0.0527, "step": 35910 }, { "grad_norm": 0.33984285593032837, "learning_rate": 6.243649625139715e-05, "loss": 0.0411, "step": 35920 }, { "grad_norm": 0.40204572677612305, "learning_rate": 6.241647636414185e-05, "loss": 0.0337, "step": 35930 }, { "grad_norm": 0.5083188414573669, "learning_rate": 6.239645435525034e-05, "loss": 0.0415, "step": 35940 }, { "grad_norm": 0.31260982155799866, "learning_rate": 6.237643022814381e-05, "loss": 0.0375, "step": 35950 }, { "grad_norm": 0.3114618957042694, "learning_rate": 6.235640398624386e-05, "loss": 0.0363, "step": 35960 }, { "grad_norm": 0.4760865867137909, "learning_rate": 6.233637563297243e-05, "loss": 0.0448, "step": 35970 }, { "grad_norm": 0.4590393900871277, "learning_rate": 6.23163451717518e-05, "loss": 0.0422, "step": 35980 }, { "grad_norm": 0.3017921447753906, "learning_rate": 6.229631260600463e-05, "loss": 0.0352, "step": 35990 }, { "grad_norm": 0.326445996761322, "learning_rate": 6.227627793915392e-05, "loss": 0.0426, "step": 36000 }, { "grad_norm": 0.2855369448661804, "learning_rate": 6.225624117462309e-05, "loss": 0.0279, "step": 36010 }, { "grad_norm": 0.3130032420158386, "learning_rate": 6.223620231583586e-05, "loss": 0.0418, "step": 36020 }, { "grad_norm": 0.17084632813930511, "learning_rate": 6.221616136621629e-05, "loss": 0.0305, "step": 36030 }, { "grad_norm": 0.31790098547935486, "learning_rate": 6.219611832918887e-05, "loss": 0.0388, "step": 36040 }, { "grad_norm": 0.3316575884819031, "learning_rate": 6.217607320817838e-05, "loss": 0.0306, "step": 36050 }, { "grad_norm": 0.3114827573299408, "learning_rate": 6.215602600661001e-05, "loss": 0.0309, "step": 36060 }, { "grad_norm": 0.4454807639122009, "learning_rate": 6.213597672790925e-05, "loss": 0.0364, "step": 36070 }, { "grad_norm": 0.6055686473846436, "learning_rate": 6.2115925375502e-05, "loss": 0.033, "step": 36080 }, { "grad_norm": 0.49643373489379883, "learning_rate": 6.209587195281447e-05, "loss": 0.0355, "step": 36090 }, { "grad_norm": 0.5804153680801392, "learning_rate": 6.207581646327324e-05, "loss": 0.0438, "step": 36100 }, { "grad_norm": 0.37742307782173157, "learning_rate": 6.205575891030526e-05, "loss": 0.0279, "step": 36110 }, { "grad_norm": 0.3316954970359802, "learning_rate": 6.203569929733781e-05, "loss": 0.0274, "step": 36120 }, { "grad_norm": 0.5189766883850098, "learning_rate": 6.201563762779852e-05, "loss": 0.0329, "step": 36130 }, { "grad_norm": 0.924802839756012, "learning_rate": 6.199557390511538e-05, "loss": 0.0293, "step": 36140 }, { "grad_norm": 3.1396431922912598, "learning_rate": 6.197550813271675e-05, "loss": 0.0385, "step": 36150 }, { "grad_norm": 0.3727574050426483, "learning_rate": 6.195544031403131e-05, "loss": 0.0405, "step": 36160 }, { "grad_norm": 0.469662606716156, "learning_rate": 6.19353704524881e-05, "loss": 0.0318, "step": 36170 }, { "grad_norm": 0.3676559031009674, "learning_rate": 6.191529855151652e-05, "loss": 0.0368, "step": 36180 }, { "grad_norm": 0.3284766376018524, "learning_rate": 6.189522461454629e-05, "loss": 0.0429, "step": 36190 }, { "grad_norm": 0.38528013229370117, "learning_rate": 6.187514864500752e-05, "loss": 0.0364, "step": 36200 }, { "grad_norm": 0.37397846579551697, "learning_rate": 6.185507064633062e-05, "loss": 0.031, "step": 36210 }, { "grad_norm": 0.4613967537879944, "learning_rate": 6.18349906219464e-05, "loss": 0.0452, "step": 36220 }, { "grad_norm": 0.5768287181854248, "learning_rate": 6.181490857528596e-05, "loss": 0.0382, "step": 36230 }, { "grad_norm": 0.4315197467803955, "learning_rate": 6.179482450978077e-05, "loss": 0.0355, "step": 36240 }, { "grad_norm": 0.3082074522972107, "learning_rate": 6.177473842886269e-05, "loss": 0.0275, "step": 36250 }, { "grad_norm": 0.3606102466583252, "learning_rate": 6.175465033596382e-05, "loss": 0.0338, "step": 36260 }, { "grad_norm": 0.28486305475234985, "learning_rate": 6.173456023451671e-05, "loss": 0.0395, "step": 36270 }, { "grad_norm": 0.3646036386489868, "learning_rate": 6.171446812795422e-05, "loss": 0.044, "step": 36280 }, { "grad_norm": 0.328829288482666, "learning_rate": 6.169437401970949e-05, "loss": 0.0417, "step": 36290 }, { "grad_norm": 0.48395901918411255, "learning_rate": 6.16742779132161e-05, "loss": 0.032, "step": 36300 }, { "grad_norm": 0.4711693227291107, "learning_rate": 6.165417981190789e-05, "loss": 0.0303, "step": 36310 }, { "grad_norm": 0.5748862028121948, "learning_rate": 6.16340797192191e-05, "loss": 0.0315, "step": 36320 }, { "grad_norm": 0.40557584166526794, "learning_rate": 6.161397763858427e-05, "loss": 0.0327, "step": 36330 }, { "grad_norm": 0.457709401845932, "learning_rate": 6.159387357343834e-05, "loss": 0.0292, "step": 36340 }, { "grad_norm": 0.3385459780693054, "learning_rate": 6.157376752721648e-05, "loss": 0.0367, "step": 36350 }, { "grad_norm": 0.6675351858139038, "learning_rate": 6.155365950335428e-05, "loss": 0.0321, "step": 36360 }, { "grad_norm": 0.41132238507270813, "learning_rate": 6.153354950528768e-05, "loss": 0.0327, "step": 36370 }, { "grad_norm": 0.2945319712162018, "learning_rate": 6.151343753645293e-05, "loss": 0.0342, "step": 36380 }, { "grad_norm": 0.31468790769577026, "learning_rate": 6.149332360028657e-05, "loss": 0.0399, "step": 36390 }, { "grad_norm": 0.30167046189308167, "learning_rate": 6.147320770022555e-05, "loss": 0.0378, "step": 36400 }, { "grad_norm": 0.44982606172561646, "learning_rate": 6.145308983970715e-05, "loss": 0.0366, "step": 36410 }, { "grad_norm": 0.30012303590774536, "learning_rate": 6.143297002216892e-05, "loss": 0.0342, "step": 36420 }, { "grad_norm": 0.3326670527458191, "learning_rate": 6.141284825104882e-05, "loss": 0.0376, "step": 36430 }, { "grad_norm": 0.3207145631313324, "learning_rate": 6.13927245297851e-05, "loss": 0.0441, "step": 36440 }, { "grad_norm": 0.27478304505348206, "learning_rate": 6.137259886181633e-05, "loss": 0.0361, "step": 36450 }, { "grad_norm": 0.35363060235977173, "learning_rate": 6.135247125058145e-05, "loss": 0.0377, "step": 36460 }, { "grad_norm": 0.36172986030578613, "learning_rate": 6.133234169951974e-05, "loss": 0.0286, "step": 36470 }, { "grad_norm": 0.4389502704143524, "learning_rate": 6.131221021207078e-05, "loss": 0.0392, "step": 36480 }, { "grad_norm": 0.3172625005245209, "learning_rate": 6.129207679167448e-05, "loss": 0.0337, "step": 36490 }, { "grad_norm": 0.46636924147605896, "learning_rate": 6.127194144177109e-05, "loss": 0.0289, "step": 36500 }, { "grad_norm": 0.3980743885040283, "learning_rate": 6.125180416580118e-05, "loss": 0.0276, "step": 36510 }, { "grad_norm": 0.3924206495285034, "learning_rate": 6.123166496720571e-05, "loss": 0.0292, "step": 36520 }, { "grad_norm": 0.3869207501411438, "learning_rate": 6.121152384942588e-05, "loss": 0.0374, "step": 36530 }, { "grad_norm": 1.6428042650222778, "learning_rate": 6.119138081590324e-05, "loss": 0.0354, "step": 36540 }, { "grad_norm": 0.3990453779697418, "learning_rate": 6.117123587007971e-05, "loss": 0.039, "step": 36550 }, { "grad_norm": 0.2603200674057007, "learning_rate": 6.11510890153975e-05, "loss": 0.0354, "step": 36560 }, { "grad_norm": 0.36431071162223816, "learning_rate": 6.113094025529916e-05, "loss": 0.034, "step": 36570 }, { "grad_norm": 0.4347541034221649, "learning_rate": 6.111078959322757e-05, "loss": 0.0521, "step": 36580 }, { "grad_norm": 0.40351247787475586, "learning_rate": 6.109063703262592e-05, "loss": 0.0384, "step": 36590 }, { "grad_norm": 0.3749068081378937, "learning_rate": 6.107048257693772e-05, "loss": 0.0407, "step": 36600 }, { "grad_norm": 0.5360490083694458, "learning_rate": 6.105032622960683e-05, "loss": 0.0371, "step": 36610 }, { "grad_norm": 0.3728407621383667, "learning_rate": 6.103016799407743e-05, "loss": 0.035, "step": 36620 }, { "grad_norm": 0.3850688338279724, "learning_rate": 6.1010007873793984e-05, "loss": 0.0359, "step": 36630 }, { "grad_norm": 0.3531491160392761, "learning_rate": 6.098984587220131e-05, "loss": 0.033, "step": 36640 }, { "grad_norm": 0.33859914541244507, "learning_rate": 6.096968199274456e-05, "loss": 0.0372, "step": 36650 }, { "grad_norm": 0.3679613173007965, "learning_rate": 6.0949516238869166e-05, "loss": 0.0319, "step": 36660 }, { "grad_norm": 0.5219504833221436, "learning_rate": 6.092934861402092e-05, "loss": 0.0511, "step": 36670 }, { "grad_norm": 0.43306249380111694, "learning_rate": 6.0909179121645924e-05, "loss": 0.031, "step": 36680 }, { "grad_norm": 0.3661085069179535, "learning_rate": 6.0889007765190576e-05, "loss": 0.0378, "step": 36690 }, { "grad_norm": 0.4799378514289856, "learning_rate": 6.086883454810162e-05, "loss": 0.0355, "step": 36700 }, { "grad_norm": 0.32222333550453186, "learning_rate": 6.0848659473826084e-05, "loss": 0.0334, "step": 36710 }, { "grad_norm": 0.4805249571800232, "learning_rate": 6.082848254581138e-05, "loss": 0.0418, "step": 36720 }, { "grad_norm": 0.3694900870323181, "learning_rate": 6.080830376750517e-05, "loss": 0.0318, "step": 36730 }, { "grad_norm": 0.33113718032836914, "learning_rate": 6.0788123142355445e-05, "loss": 0.0316, "step": 36740 }, { "grad_norm": 0.4719506800174713, "learning_rate": 6.076794067381052e-05, "loss": 0.0332, "step": 36750 }, { "grad_norm": 0.46908360719680786, "learning_rate": 6.074775636531905e-05, "loss": 0.0357, "step": 36760 }, { "grad_norm": 0.514593243598938, "learning_rate": 6.072757022032997e-05, "loss": 0.0377, "step": 36770 }, { "grad_norm": 0.48608124256134033, "learning_rate": 6.070738224229253e-05, "loss": 0.0394, "step": 36780 }, { "grad_norm": 0.39841148257255554, "learning_rate": 6.0687192434656314e-05, "loss": 0.0329, "step": 36790 }, { "grad_norm": 0.7512547969818115, "learning_rate": 6.066700080087121e-05, "loss": 0.0342, "step": 36800 }, { "grad_norm": 0.2909466028213501, "learning_rate": 6.0646807344387424e-05, "loss": 0.038, "step": 36810 }, { "grad_norm": 0.39604321122169495, "learning_rate": 6.062661206865543e-05, "loss": 0.0348, "step": 36820 }, { "grad_norm": 0.6090066432952881, "learning_rate": 6.06064149771261e-05, "loss": 0.0302, "step": 36830 }, { "grad_norm": 0.3915582299232483, "learning_rate": 6.058621607325051e-05, "loss": 0.0295, "step": 36840 }, { "grad_norm": 0.3744930326938629, "learning_rate": 6.056601536048014e-05, "loss": 0.0311, "step": 36850 }, { "grad_norm": 0.41452285647392273, "learning_rate": 6.0545812842266725e-05, "loss": 0.0326, "step": 36860 }, { "grad_norm": 0.2803637683391571, "learning_rate": 6.052560852206232e-05, "loss": 0.032, "step": 36870 }, { "grad_norm": 0.3979117274284363, "learning_rate": 6.05054024033193e-05, "loss": 0.0377, "step": 36880 }, { "grad_norm": 0.3944333791732788, "learning_rate": 6.048519448949032e-05, "loss": 0.0317, "step": 36890 }, { "grad_norm": 0.3558533489704132, "learning_rate": 6.046498478402839e-05, "loss": 0.0404, "step": 36900 }, { "grad_norm": 0.44729045033454895, "learning_rate": 6.044477329038677e-05, "loss": 0.0336, "step": 36910 }, { "grad_norm": 0.474080353975296, "learning_rate": 6.042456001201906e-05, "loss": 0.0414, "step": 36920 }, { "grad_norm": 0.32378891110420227, "learning_rate": 6.040434495237917e-05, "loss": 0.0413, "step": 36930 }, { "grad_norm": 0.8664717674255371, "learning_rate": 6.0384128114921256e-05, "loss": 0.0489, "step": 36940 }, { "grad_norm": 0.28672870993614197, "learning_rate": 6.036390950309987e-05, "loss": 0.0293, "step": 36950 }, { "grad_norm": 0.37439829111099243, "learning_rate": 6.0343689120369805e-05, "loss": 0.0363, "step": 36960 }, { "grad_norm": 0.39479655027389526, "learning_rate": 6.032346697018616e-05, "loss": 0.044, "step": 36970 }, { "grad_norm": 0.3762202560901642, "learning_rate": 6.0303243056004375e-05, "loss": 0.0557, "step": 36980 }, { "grad_norm": 0.6437246203422546, "learning_rate": 6.0283017381280136e-05, "loss": 0.0303, "step": 36990 }, { "grad_norm": 0.3933965265750885, "learning_rate": 6.026278994946948e-05, "loss": 0.0336, "step": 37000 }, { "grad_norm": 0.31233105063438416, "learning_rate": 6.02425607640287e-05, "loss": 0.0376, "step": 37010 }, { "grad_norm": 0.6460931897163391, "learning_rate": 6.022232982841441e-05, "loss": 0.0355, "step": 37020 }, { "grad_norm": 0.40908727049827576, "learning_rate": 6.020209714608355e-05, "loss": 0.0256, "step": 37030 }, { "grad_norm": 0.37106427550315857, "learning_rate": 6.018186272049331e-05, "loss": 0.0258, "step": 37040 }, { "grad_norm": 0.3222174644470215, "learning_rate": 6.01616265551012e-05, "loss": 0.0344, "step": 37050 }, { "grad_norm": 0.35377076268196106, "learning_rate": 6.014138865336503e-05, "loss": 0.0292, "step": 37060 }, { "grad_norm": 0.31514424085617065, "learning_rate": 6.0121149018742905e-05, "loss": 0.0256, "step": 37070 }, { "grad_norm": 0.20663972198963165, "learning_rate": 6.010090765469325e-05, "loss": 0.0276, "step": 37080 }, { "grad_norm": 0.5339118242263794, "learning_rate": 6.008066456467473e-05, "loss": 0.0314, "step": 37090 }, { "grad_norm": 0.2183542549610138, "learning_rate": 6.0060419752146335e-05, "loss": 0.0314, "step": 37100 }, { "grad_norm": 0.2792440354824066, "learning_rate": 6.0040173220567353e-05, "loss": 0.0388, "step": 37110 }, { "grad_norm": 0.263942688703537, "learning_rate": 6.001992497339737e-05, "loss": 0.0326, "step": 37120 }, { "grad_norm": 0.25505778193473816, "learning_rate": 5.999967501409626e-05, "loss": 0.0324, "step": 37130 }, { "grad_norm": 0.5605981945991516, "learning_rate": 5.997942334612418e-05, "loss": 0.0328, "step": 37140 }, { "grad_norm": 0.23892396688461304, "learning_rate": 5.995916997294158e-05, "loss": 0.0239, "step": 37150 }, { "grad_norm": 0.30632781982421875, "learning_rate": 5.9938914898009214e-05, "loss": 0.0332, "step": 37160 }, { "grad_norm": 0.3244042694568634, "learning_rate": 5.991865812478813e-05, "loss": 0.0348, "step": 37170 }, { "grad_norm": 0.4490131139755249, "learning_rate": 5.989839965673964e-05, "loss": 0.0505, "step": 37180 }, { "grad_norm": 0.27536633610725403, "learning_rate": 5.987813949732539e-05, "loss": 0.0312, "step": 37190 }, { "grad_norm": 0.3805648982524872, "learning_rate": 5.9857877650007255e-05, "loss": 0.0439, "step": 37200 }, { "grad_norm": 0.26746976375579834, "learning_rate": 5.983761411824744e-05, "loss": 0.0285, "step": 37210 }, { "grad_norm": 0.3476649224758148, "learning_rate": 5.981734890550844e-05, "loss": 0.04, "step": 37220 }, { "grad_norm": 0.27663934230804443, "learning_rate": 5.979708201525301e-05, "loss": 0.0343, "step": 37230 }, { "grad_norm": 0.5430079698562622, "learning_rate": 5.977681345094422e-05, "loss": 0.0401, "step": 37240 }, { "grad_norm": 0.3948136866092682, "learning_rate": 5.97565432160454e-05, "loss": 0.0368, "step": 37250 }, { "grad_norm": 0.33991748094558716, "learning_rate": 5.9736271314020186e-05, "loss": 0.0319, "step": 37260 }, { "grad_norm": 0.6009336113929749, "learning_rate": 5.971599774833251e-05, "loss": 0.0387, "step": 37270 }, { "grad_norm": 0.30774441361427307, "learning_rate": 5.9695722522446525e-05, "loss": 0.0372, "step": 37280 }, { "grad_norm": 0.2881905734539032, "learning_rate": 5.9675445639826765e-05, "loss": 0.0296, "step": 37290 }, { "grad_norm": 0.27415281534194946, "learning_rate": 5.965516710393796e-05, "loss": 0.0337, "step": 37300 }, { "grad_norm": 0.35584384202957153, "learning_rate": 5.963488691824516e-05, "loss": 0.0405, "step": 37310 }, { "grad_norm": 0.40991783142089844, "learning_rate": 5.96146050862137e-05, "loss": 0.0359, "step": 37320 }, { "grad_norm": 0.401361346244812, "learning_rate": 5.959432161130919e-05, "loss": 0.0343, "step": 37330 }, { "grad_norm": 0.37554770708084106, "learning_rate": 5.9574036496997545e-05, "loss": 0.0465, "step": 37340 }, { "grad_norm": 0.31808269023895264, "learning_rate": 5.955374974674488e-05, "loss": 0.0283, "step": 37350 }, { "grad_norm": 0.3348691463470459, "learning_rate": 5.9533461364017696e-05, "loss": 0.0311, "step": 37360 }, { "grad_norm": 0.33435380458831787, "learning_rate": 5.9513171352282716e-05, "loss": 0.0406, "step": 37370 }, { "grad_norm": 0.2983368933200836, "learning_rate": 5.949287971500692e-05, "loss": 0.0353, "step": 37380 }, { "grad_norm": 0.4840184450149536, "learning_rate": 5.947258645565762e-05, "loss": 0.0346, "step": 37390 }, { "grad_norm": 0.31778305768966675, "learning_rate": 5.945229157770237e-05, "loss": 0.031, "step": 37400 }, { "grad_norm": 0.22953613102436066, "learning_rate": 5.9431995084609006e-05, "loss": 0.0387, "step": 37410 }, { "grad_norm": 0.43991973996162415, "learning_rate": 5.941169697984564e-05, "loss": 0.0334, "step": 37420 }, { "grad_norm": 0.36853882670402527, "learning_rate": 5.9391397266880675e-05, "loss": 0.0288, "step": 37430 }, { "grad_norm": 0.33926764130592346, "learning_rate": 5.937109594918279e-05, "loss": 0.0474, "step": 37440 }, { "grad_norm": 0.31048211455345154, "learning_rate": 5.9350793030220884e-05, "loss": 0.0409, "step": 37450 }, { "grad_norm": 0.34590837359428406, "learning_rate": 5.933048851346421e-05, "loss": 0.029, "step": 37460 }, { "grad_norm": 0.3318617641925812, "learning_rate": 5.931018240238224e-05, "loss": 0.0501, "step": 37470 }, { "grad_norm": 0.38863617181777954, "learning_rate": 5.928987470044471e-05, "loss": 0.0354, "step": 37480 }, { "grad_norm": 0.546741783618927, "learning_rate": 5.9269565411121695e-05, "loss": 0.0344, "step": 37490 }, { "grad_norm": 0.38581582903862, "learning_rate": 5.924925453788347e-05, "loss": 0.0366, "step": 37500 }, { "grad_norm": 0.22406256198883057, "learning_rate": 5.92289420842006e-05, "loss": 0.0308, "step": 37510 }, { "grad_norm": 0.4773154854774475, "learning_rate": 5.9208628053543945e-05, "loss": 0.0331, "step": 37520 }, { "grad_norm": 0.4157751500606537, "learning_rate": 5.918831244938462e-05, "loss": 0.0309, "step": 37530 }, { "grad_norm": 0.45927315950393677, "learning_rate": 5.9167995275194e-05, "loss": 0.0384, "step": 37540 }, { "grad_norm": 0.449656218290329, "learning_rate": 5.914767653444373e-05, "loss": 0.0389, "step": 37550 }, { "grad_norm": 0.5435166954994202, "learning_rate": 5.912735623060572e-05, "loss": 0.0414, "step": 37560 }, { "grad_norm": 0.7803773283958435, "learning_rate": 5.910703436715217e-05, "loss": 0.0347, "step": 37570 }, { "grad_norm": 0.2897496521472931, "learning_rate": 5.908671094755552e-05, "loss": 0.0318, "step": 37580 }, { "grad_norm": 0.42819589376449585, "learning_rate": 5.906638597528851e-05, "loss": 0.0353, "step": 37590 }, { "grad_norm": 0.36155298352241516, "learning_rate": 5.9046059453824076e-05, "loss": 0.0344, "step": 37600 }, { "grad_norm": 0.40109729766845703, "learning_rate": 5.9025731386635505e-05, "loss": 0.0328, "step": 37610 }, { "grad_norm": 0.44358715415000916, "learning_rate": 5.900540177719629e-05, "loss": 0.0412, "step": 37620 }, { "grad_norm": 0.4882548153400421, "learning_rate": 5.898507062898021e-05, "loss": 0.0403, "step": 37630 }, { "grad_norm": 0.42155373096466064, "learning_rate": 5.8964737945461316e-05, "loss": 0.0293, "step": 37640 }, { "grad_norm": 0.28422319889068604, "learning_rate": 5.8944403730113885e-05, "loss": 0.0323, "step": 37650 }, { "grad_norm": 0.4772372841835022, "learning_rate": 5.892406798641248e-05, "loss": 0.0445, "step": 37660 }, { "grad_norm": 0.39963796734809875, "learning_rate": 5.890373071783193e-05, "loss": 0.0342, "step": 37670 }, { "grad_norm": 0.414958119392395, "learning_rate": 5.888339192784732e-05, "loss": 0.0359, "step": 37680 }, { "grad_norm": 0.23587368428707123, "learning_rate": 5.8863051619934003e-05, "loss": 0.036, "step": 37690 }, { "grad_norm": 0.8476446866989136, "learning_rate": 5.8842709797567554e-05, "loss": 0.0429, "step": 37700 }, { "grad_norm": 0.3890610933303833, "learning_rate": 5.8822366464223855e-05, "loss": 0.0256, "step": 37710 }, { "grad_norm": 0.3148840069770813, "learning_rate": 5.880202162337901e-05, "loss": 0.0429, "step": 37720 }, { "grad_norm": 0.26593583822250366, "learning_rate": 5.8781675278509405e-05, "loss": 0.0326, "step": 37730 }, { "grad_norm": 0.43636277318000793, "learning_rate": 5.8761327433091696e-05, "loss": 0.0364, "step": 37740 }, { "grad_norm": 0.2846131920814514, "learning_rate": 5.874097809060275e-05, "loss": 0.0393, "step": 37750 }, { "grad_norm": 0.4259558916091919, "learning_rate": 5.87206272545197e-05, "loss": 0.0517, "step": 37760 }, { "grad_norm": 0.47241735458374023, "learning_rate": 5.8700274928319955e-05, "loss": 0.0418, "step": 37770 }, { "grad_norm": 1.1044408082962036, "learning_rate": 5.867992111548118e-05, "loss": 0.0414, "step": 37780 }, { "grad_norm": 0.3345806896686554, "learning_rate": 5.865956581948131e-05, "loss": 0.0344, "step": 37790 }, { "grad_norm": 0.6645082235336304, "learning_rate": 5.863920904379845e-05, "loss": 0.0352, "step": 37800 }, { "grad_norm": 0.23265479505062103, "learning_rate": 5.8618850791911064e-05, "loss": 0.0339, "step": 37810 }, { "grad_norm": 0.47325271368026733, "learning_rate": 5.859849106729779e-05, "loss": 0.0321, "step": 37820 }, { "grad_norm": 0.7124536037445068, "learning_rate": 5.857812987343758e-05, "loss": 0.0305, "step": 37830 }, { "grad_norm": 0.37048426270484924, "learning_rate": 5.855776721380957e-05, "loss": 0.0378, "step": 37840 }, { "grad_norm": 0.2984601855278015, "learning_rate": 5.8537403091893217e-05, "loss": 0.032, "step": 37850 }, { "grad_norm": 0.48470574617385864, "learning_rate": 5.851703751116816e-05, "loss": 0.0404, "step": 37860 }, { "grad_norm": 0.3837432563304901, "learning_rate": 5.8496670475114336e-05, "loss": 0.0347, "step": 37870 }, { "grad_norm": 0.384686142206192, "learning_rate": 5.84763019872119e-05, "loss": 0.0382, "step": 37880 }, { "grad_norm": 0.36610114574432373, "learning_rate": 5.845593205094131e-05, "loss": 0.0394, "step": 37890 }, { "grad_norm": 0.3084951639175415, "learning_rate": 5.843556066978318e-05, "loss": 0.0341, "step": 37900 }, { "grad_norm": 0.3967215120792389, "learning_rate": 5.8415187847218455e-05, "loss": 0.0363, "step": 37910 }, { "grad_norm": 0.5282396078109741, "learning_rate": 5.839481358672827e-05, "loss": 0.037, "step": 37920 }, { "grad_norm": 0.2849845588207245, "learning_rate": 5.837443789179407e-05, "loss": 0.0335, "step": 37930 }, { "grad_norm": 0.44151535630226135, "learning_rate": 5.8354060765897445e-05, "loss": 0.0445, "step": 37940 }, { "grad_norm": 0.45449450612068176, "learning_rate": 5.8333682212520334e-05, "loss": 0.05, "step": 37950 }, { "grad_norm": 0.35233739018440247, "learning_rate": 5.831330223514486e-05, "loss": 0.0378, "step": 37960 }, { "grad_norm": 0.3280646800994873, "learning_rate": 5.8292920837253396e-05, "loss": 0.0344, "step": 37970 }, { "grad_norm": 0.3389405310153961, "learning_rate": 5.827253802232857e-05, "loss": 0.0288, "step": 37980 }, { "grad_norm": 0.4061473309993744, "learning_rate": 5.825215379385327e-05, "loss": 0.027, "step": 37990 }, { "grad_norm": 0.3513490557670593, "learning_rate": 5.823176815531057e-05, "loss": 0.0468, "step": 38000 }, { "grad_norm": 0.3845105469226837, "learning_rate": 5.8211381110183826e-05, "loss": 0.0401, "step": 38010 }, { "grad_norm": 0.4166778326034546, "learning_rate": 5.8190992661956645e-05, "loss": 0.0329, "step": 38020 }, { "grad_norm": 0.3231939971446991, "learning_rate": 5.817060281411284e-05, "loss": 0.0339, "step": 38030 }, { "grad_norm": 0.41903597116470337, "learning_rate": 5.815021157013647e-05, "loss": 0.0328, "step": 38040 }, { "grad_norm": 0.5539596676826477, "learning_rate": 5.8129818933511856e-05, "loss": 0.0451, "step": 38050 }, { "grad_norm": 0.5645294785499573, "learning_rate": 5.8109424907723544e-05, "loss": 0.0445, "step": 38060 }, { "grad_norm": 0.2851985692977905, "learning_rate": 5.80890294962563e-05, "loss": 0.0304, "step": 38070 }, { "grad_norm": 0.3729383647441864, "learning_rate": 5.806863270259515e-05, "loss": 0.037, "step": 38080 }, { "grad_norm": 0.4651769697666168, "learning_rate": 5.804823453022536e-05, "loss": 0.0374, "step": 38090 }, { "grad_norm": 0.40702229738235474, "learning_rate": 5.80278349826324e-05, "loss": 0.0352, "step": 38100 }, { "grad_norm": 0.5006536841392517, "learning_rate": 5.8007434063302014e-05, "loss": 0.0326, "step": 38110 }, { "grad_norm": 0.4218916893005371, "learning_rate": 5.7987031775720136e-05, "loss": 0.0334, "step": 38120 }, { "grad_norm": 0.5158666372299194, "learning_rate": 5.7966628123372976e-05, "loss": 0.0385, "step": 38130 }, { "grad_norm": 0.42646172642707825, "learning_rate": 5.7946223109746956e-05, "loss": 0.0424, "step": 38140 }, { "grad_norm": 0.48802274465560913, "learning_rate": 5.7925816738328754e-05, "loss": 0.0297, "step": 38150 }, { "grad_norm": 0.46540290117263794, "learning_rate": 5.790540901260521e-05, "loss": 0.0353, "step": 38160 }, { "grad_norm": 0.6701451539993286, "learning_rate": 5.788499993606351e-05, "loss": 0.0296, "step": 38170 }, { "grad_norm": 0.2394055724143982, "learning_rate": 5.786458951219096e-05, "loss": 0.0344, "step": 38180 }, { "grad_norm": 0.39814284443855286, "learning_rate": 5.784417774447517e-05, "loss": 0.0399, "step": 38190 }, { "grad_norm": 0.3030100464820862, "learning_rate": 5.782376463640393e-05, "loss": 0.0357, "step": 38200 }, { "grad_norm": 0.504069983959198, "learning_rate": 5.780335019146531e-05, "loss": 0.0321, "step": 38210 }, { "grad_norm": 0.3284755349159241, "learning_rate": 5.778293441314755e-05, "loss": 0.0338, "step": 38220 }, { "grad_norm": 0.45450371503829956, "learning_rate": 5.776251730493917e-05, "loss": 0.0327, "step": 38230 }, { "grad_norm": 0.36157482862472534, "learning_rate": 5.774209887032887e-05, "loss": 0.0326, "step": 38240 }, { "grad_norm": 0.2335880994796753, "learning_rate": 5.772167911280565e-05, "loss": 0.0343, "step": 38250 }, { "grad_norm": 0.2762213945388794, "learning_rate": 5.770125803585864e-05, "loss": 0.0316, "step": 38260 }, { "grad_norm": 0.6661598086357117, "learning_rate": 5.768083564297726e-05, "loss": 0.0477, "step": 38270 }, { "grad_norm": 0.5947515368461609, "learning_rate": 5.766041193765114e-05, "loss": 0.0378, "step": 38280 }, { "grad_norm": 0.2809041142463684, "learning_rate": 5.763998692337015e-05, "loss": 0.0313, "step": 38290 }, { "grad_norm": 0.43069565296173096, "learning_rate": 5.761956060362433e-05, "loss": 0.0434, "step": 38300 }, { "grad_norm": 0.30057263374328613, "learning_rate": 5.7599132981904e-05, "loss": 0.0287, "step": 38310 }, { "grad_norm": 0.43378186225891113, "learning_rate": 5.75787040616997e-05, "loss": 0.0364, "step": 38320 }, { "grad_norm": 0.34386152029037476, "learning_rate": 5.755827384650212e-05, "loss": 0.031, "step": 38330 }, { "grad_norm": 0.38890713453292847, "learning_rate": 5.753784233980228e-05, "loss": 0.0375, "step": 38340 }, { "grad_norm": 0.2591014504432678, "learning_rate": 5.751740954509135e-05, "loss": 0.0268, "step": 38350 }, { "grad_norm": 0.5919404029846191, "learning_rate": 5.7496975465860715e-05, "loss": 0.0253, "step": 38360 }, { "grad_norm": 0.24692022800445557, "learning_rate": 5.747654010560202e-05, "loss": 0.0264, "step": 38370 }, { "grad_norm": 0.2335079312324524, "learning_rate": 5.7456103467807097e-05, "loss": 0.0326, "step": 38380 }, { "grad_norm": 3.286588668823242, "learning_rate": 5.7435665555968046e-05, "loss": 0.0326, "step": 38390 }, { "grad_norm": 0.38991278409957886, "learning_rate": 5.74152263735771e-05, "loss": 0.0387, "step": 38400 }, { "grad_norm": 0.40886515378952026, "learning_rate": 5.739478592412677e-05, "loss": 0.0344, "step": 38410 }, { "grad_norm": 0.38937637209892273, "learning_rate": 5.7374344211109766e-05, "loss": 0.0378, "step": 38420 }, { "grad_norm": 0.2961386740207672, "learning_rate": 5.735390123801904e-05, "loss": 0.0363, "step": 38430 }, { "grad_norm": 0.6203613877296448, "learning_rate": 5.7333457008347704e-05, "loss": 0.0401, "step": 38440 }, { "grad_norm": 0.4349805414676666, "learning_rate": 5.7313011525589156e-05, "loss": 0.0341, "step": 38450 }, { "grad_norm": 0.46111825108528137, "learning_rate": 5.729256479323694e-05, "loss": 0.0343, "step": 38460 }, { "grad_norm": 0.49339428544044495, "learning_rate": 5.727211681478485e-05, "loss": 0.0357, "step": 38470 }, { "grad_norm": 0.4214938282966614, "learning_rate": 5.7251667593726886e-05, "loss": 0.0393, "step": 38480 }, { "grad_norm": 0.49185192584991455, "learning_rate": 5.723121713355728e-05, "loss": 0.0498, "step": 38490 }, { "grad_norm": 0.3426356911659241, "learning_rate": 5.721076543777044e-05, "loss": 0.0336, "step": 38500 }, { "grad_norm": 0.23221756517887115, "learning_rate": 5.7190312509860986e-05, "loss": 0.0248, "step": 38510 }, { "grad_norm": 0.28966817259788513, "learning_rate": 5.716985835332379e-05, "loss": 0.0405, "step": 38520 }, { "grad_norm": 0.2468900978565216, "learning_rate": 5.714940297165389e-05, "loss": 0.0294, "step": 38530 }, { "grad_norm": 0.4096447825431824, "learning_rate": 5.712894636834656e-05, "loss": 0.0333, "step": 38540 }, { "grad_norm": 0.41967901587486267, "learning_rate": 5.7108488546897276e-05, "loss": 0.0249, "step": 38550 }, { "grad_norm": 0.42139339447021484, "learning_rate": 5.708802951080172e-05, "loss": 0.028, "step": 38560 }, { "grad_norm": 0.3335167467594147, "learning_rate": 5.706756926355576e-05, "loss": 0.0341, "step": 38570 }, { "grad_norm": 0.5769187808036804, "learning_rate": 5.704710780865554e-05, "loss": 0.0367, "step": 38580 }, { "grad_norm": 0.3278234302997589, "learning_rate": 5.7026645149597325e-05, "loss": 0.0316, "step": 38590 }, { "grad_norm": 0.41784942150115967, "learning_rate": 5.700618128987764e-05, "loss": 0.0379, "step": 38600 }, { "grad_norm": 0.22360269725322723, "learning_rate": 5.698571623299317e-05, "loss": 0.029, "step": 38610 }, { "grad_norm": 0.2593012750148773, "learning_rate": 5.696524998244086e-05, "loss": 0.0239, "step": 38620 }, { "grad_norm": 0.33716243505477905, "learning_rate": 5.6944782541717836e-05, "loss": 0.0398, "step": 38630 }, { "grad_norm": 0.28842201828956604, "learning_rate": 5.69243139143214e-05, "loss": 0.0275, "step": 38640 }, { "grad_norm": 0.32358187437057495, "learning_rate": 5.6903844103749125e-05, "loss": 0.034, "step": 38650 }, { "grad_norm": 0.49169257283210754, "learning_rate": 5.688337311349869e-05, "loss": 0.0327, "step": 38660 }, { "grad_norm": 0.4543604850769043, "learning_rate": 5.6862900947068074e-05, "loss": 0.0385, "step": 38670 }, { "grad_norm": 0.33968573808670044, "learning_rate": 5.6842427607955374e-05, "loss": 0.0386, "step": 38680 }, { "grad_norm": 0.41270408034324646, "learning_rate": 5.682195309965893e-05, "loss": 0.0339, "step": 38690 }, { "grad_norm": 0.4661698341369629, "learning_rate": 5.6801477425677294e-05, "loss": 0.0321, "step": 38700 }, { "grad_norm": 0.3980875611305237, "learning_rate": 5.678100058950917e-05, "loss": 0.0388, "step": 38710 }, { "grad_norm": 0.414610892534256, "learning_rate": 5.676052259465352e-05, "loss": 0.0363, "step": 38720 }, { "grad_norm": 0.38930198550224304, "learning_rate": 5.674004344460945e-05, "loss": 0.0389, "step": 38730 }, { "grad_norm": 0.3614974021911621, "learning_rate": 5.6719563142876295e-05, "loss": 0.0329, "step": 38740 }, { "grad_norm": 0.3584858477115631, "learning_rate": 5.669908169295359e-05, "loss": 0.0318, "step": 38750 }, { "grad_norm": 0.32728737592697144, "learning_rate": 5.667859909834105e-05, "loss": 0.0273, "step": 38760 }, { "grad_norm": 0.4780517518520355, "learning_rate": 5.6658115362538565e-05, "loss": 0.0395, "step": 38770 }, { "grad_norm": 0.456415057182312, "learning_rate": 5.663763048904628e-05, "loss": 0.0345, "step": 38780 }, { "grad_norm": 0.5928784012794495, "learning_rate": 5.661714448136447e-05, "loss": 0.0355, "step": 38790 }, { "grad_norm": 0.48373040556907654, "learning_rate": 5.659665734299366e-05, "loss": 0.0354, "step": 38800 }, { "grad_norm": 0.34741899371147156, "learning_rate": 5.6576169077434516e-05, "loss": 0.0446, "step": 38810 }, { "grad_norm": 0.4237832725048065, "learning_rate": 5.6555679688187944e-05, "loss": 0.0354, "step": 38820 }, { "grad_norm": 0.391265332698822, "learning_rate": 5.6535189178755e-05, "loss": 0.0359, "step": 38830 }, { "grad_norm": 0.44667986035346985, "learning_rate": 5.651469755263698e-05, "loss": 0.0334, "step": 38840 }, { "grad_norm": 0.31070950627326965, "learning_rate": 5.6494204813335316e-05, "loss": 0.0433, "step": 38850 }, { "grad_norm": 0.35506322979927063, "learning_rate": 5.647371096435168e-05, "loss": 0.0279, "step": 38860 }, { "grad_norm": 0.7883604168891907, "learning_rate": 5.645321600918788e-05, "loss": 0.0325, "step": 38870 }, { "grad_norm": 0.3736245930194855, "learning_rate": 5.643271995134597e-05, "loss": 0.0393, "step": 38880 }, { "grad_norm": 0.34473949670791626, "learning_rate": 5.641222279432814e-05, "loss": 0.0326, "step": 38890 }, { "grad_norm": 0.5168132781982422, "learning_rate": 5.6391724541636834e-05, "loss": 0.0367, "step": 38900 }, { "grad_norm": 0.39528989791870117, "learning_rate": 5.6371225196774605e-05, "loss": 0.0381, "step": 38910 }, { "grad_norm": 0.3259337842464447, "learning_rate": 5.635072476324423e-05, "loss": 0.0411, "step": 38920 }, { "grad_norm": 0.4524807035923004, "learning_rate": 5.63302232445487e-05, "loss": 0.031, "step": 38930 }, { "grad_norm": 0.376474529504776, "learning_rate": 5.6309720644191144e-05, "loss": 0.035, "step": 38940 }, { "grad_norm": 0.5964246392250061, "learning_rate": 5.628921696567491e-05, "loss": 0.0374, "step": 38950 }, { "grad_norm": 0.4180201590061188, "learning_rate": 5.62687122125035e-05, "loss": 0.0369, "step": 38960 }, { "grad_norm": 0.5173056125640869, "learning_rate": 5.624820638818062e-05, "loss": 0.0307, "step": 38970 }, { "grad_norm": 0.48244404792785645, "learning_rate": 5.6227699496210164e-05, "loss": 0.0402, "step": 38980 }, { "grad_norm": 0.5102797746658325, "learning_rate": 5.6207191540096195e-05, "loss": 0.0461, "step": 38990 }, { "grad_norm": 0.6308307647705078, "learning_rate": 5.618668252334296e-05, "loss": 0.0388, "step": 39000 }, { "grad_norm": 0.3724716603755951, "learning_rate": 5.616617244945488e-05, "loss": 0.0305, "step": 39010 }, { "grad_norm": 0.5342318415641785, "learning_rate": 5.614566132193656e-05, "loss": 0.0355, "step": 39020 }, { "grad_norm": 0.36301690340042114, "learning_rate": 5.612514914429282e-05, "loss": 0.0377, "step": 39030 }, { "grad_norm": 0.30816441774368286, "learning_rate": 5.610463592002863e-05, "loss": 0.0257, "step": 39040 }, { "grad_norm": 0.4786248505115509, "learning_rate": 5.608412165264909e-05, "loss": 0.0303, "step": 39050 }, { "grad_norm": 0.30411824584007263, "learning_rate": 5.606360634565959e-05, "loss": 0.0309, "step": 39060 }, { "grad_norm": 0.5524266958236694, "learning_rate": 5.604309000256559e-05, "loss": 0.0319, "step": 39070 }, { "grad_norm": 0.3351418673992157, "learning_rate": 5.6022572626872785e-05, "loss": 0.0292, "step": 39080 }, { "grad_norm": 0.5620212554931641, "learning_rate": 5.600205422208704e-05, "loss": 0.0341, "step": 39090 }, { "grad_norm": 0.3852422833442688, "learning_rate": 5.59815347917144e-05, "loss": 0.0395, "step": 39100 }, { "grad_norm": 0.49863117933273315, "learning_rate": 5.596101433926103e-05, "loss": 0.0325, "step": 39110 }, { "grad_norm": 0.3592854142189026, "learning_rate": 5.5940492868233364e-05, "loss": 0.0288, "step": 39120 }, { "grad_norm": 0.21709676086902618, "learning_rate": 5.591997038213793e-05, "loss": 0.0292, "step": 39130 }, { "grad_norm": 0.42118367552757263, "learning_rate": 5.5899446884481475e-05, "loss": 0.0302, "step": 39140 }, { "grad_norm": 0.3770845830440521, "learning_rate": 5.5878922378770906e-05, "loss": 0.0484, "step": 39150 }, { "grad_norm": 0.6877844929695129, "learning_rate": 5.5858396868513285e-05, "loss": 0.0362, "step": 39160 }, { "grad_norm": 0.3754054605960846, "learning_rate": 5.583787035721586e-05, "loss": 0.0351, "step": 39170 }, { "grad_norm": 0.37297454476356506, "learning_rate": 5.581734284838606e-05, "loss": 0.0347, "step": 39180 }, { "grad_norm": 0.4281522333621979, "learning_rate": 5.579681434553147e-05, "loss": 0.0331, "step": 39190 }, { "grad_norm": 0.38704532384872437, "learning_rate": 5.5776284852159854e-05, "loss": 0.0341, "step": 39200 }, { "grad_norm": 0.5624945163726807, "learning_rate": 5.575575437177913e-05, "loss": 0.0404, "step": 39210 }, { "grad_norm": 0.3827647268772125, "learning_rate": 5.573522290789742e-05, "loss": 0.0295, "step": 39220 }, { "grad_norm": 0.3853335678577423, "learning_rate": 5.571469046402298e-05, "loss": 0.0379, "step": 39230 }, { "grad_norm": 0.360592246055603, "learning_rate": 5.5694157043664205e-05, "loss": 0.0357, "step": 39240 }, { "grad_norm": 0.3743932843208313, "learning_rate": 5.567362265032975e-05, "loss": 0.0271, "step": 39250 }, { "grad_norm": 0.36391034722328186, "learning_rate": 5.565308728752836e-05, "loss": 0.0357, "step": 39260 }, { "grad_norm": 0.33129218220710754, "learning_rate": 5.5632550958768956e-05, "loss": 0.0299, "step": 39270 }, { "grad_norm": 0.5227932333946228, "learning_rate": 5.5612013667560636e-05, "loss": 0.0348, "step": 39280 }, { "grad_norm": 0.29879164695739746, "learning_rate": 5.5591475417412676e-05, "loss": 0.0466, "step": 39290 }, { "grad_norm": 0.37623369693756104, "learning_rate": 5.557093621183451e-05, "loss": 0.03, "step": 39300 }, { "grad_norm": 0.3146575093269348, "learning_rate": 5.55503960543357e-05, "loss": 0.0371, "step": 39310 }, { "grad_norm": 0.4004872739315033, "learning_rate": 5.552985494842601e-05, "loss": 0.0336, "step": 39320 }, { "grad_norm": 0.26192957162857056, "learning_rate": 5.550931289761534e-05, "loss": 0.0346, "step": 39330 }, { "grad_norm": 0.7452594637870789, "learning_rate": 5.548876990541378e-05, "loss": 0.0367, "step": 39340 }, { "grad_norm": 0.4678291380405426, "learning_rate": 5.5468225975331564e-05, "loss": 0.0456, "step": 39350 }, { "grad_norm": 0.4434296786785126, "learning_rate": 5.544768111087909e-05, "loss": 0.0315, "step": 39360 }, { "grad_norm": 0.5112227201461792, "learning_rate": 5.54271353155669e-05, "loss": 0.0402, "step": 39370 }, { "grad_norm": 0.4128543436527252, "learning_rate": 5.5406588592905715e-05, "loss": 0.0332, "step": 39380 }, { "grad_norm": 0.2849378287792206, "learning_rate": 5.5386040946406416e-05, "loss": 0.0327, "step": 39390 }, { "grad_norm": 0.4260163903236389, "learning_rate": 5.536549237958004e-05, "loss": 0.0435, "step": 39400 }, { "grad_norm": 0.34289515018463135, "learning_rate": 5.5344942895937744e-05, "loss": 0.0439, "step": 39410 }, { "grad_norm": 0.5332036018371582, "learning_rate": 5.5324392498990904e-05, "loss": 0.0325, "step": 39420 }, { "grad_norm": 0.3931787610054016, "learning_rate": 5.5303841192251e-05, "loss": 0.0361, "step": 39430 }, { "grad_norm": 0.4942307472229004, "learning_rate": 5.52832889792297e-05, "loss": 0.037, "step": 39440 }, { "grad_norm": 0.32241231203079224, "learning_rate": 5.526273586343881e-05, "loss": 0.0312, "step": 39450 }, { "grad_norm": 0.38408851623535156, "learning_rate": 5.5242181848390306e-05, "loss": 0.0293, "step": 39460 }, { "grad_norm": 0.47192516922950745, "learning_rate": 5.5221626937596285e-05, "loss": 0.0287, "step": 39470 }, { "grad_norm": 0.3091696798801422, "learning_rate": 5.520107113456903e-05, "loss": 0.0382, "step": 39480 }, { "grad_norm": 0.4388064742088318, "learning_rate": 5.5180514442820974e-05, "loss": 0.0467, "step": 39490 }, { "grad_norm": 0.4736238718032837, "learning_rate": 5.515995686586469e-05, "loss": 0.033, "step": 39500 }, { "grad_norm": 0.47810086607933044, "learning_rate": 5.5139398407212916e-05, "loss": 0.0275, "step": 39510 }, { "grad_norm": 0.5889946818351746, "learning_rate": 5.511883907037849e-05, "loss": 0.0335, "step": 39520 }, { "grad_norm": 0.4185038208961487, "learning_rate": 5.509827885887449e-05, "loss": 0.0329, "step": 39530 }, { "grad_norm": 0.4040605425834656, "learning_rate": 5.507771777621406e-05, "loss": 0.0318, "step": 39540 }, { "grad_norm": 0.33120250701904297, "learning_rate": 5.505715582591052e-05, "loss": 0.0446, "step": 39550 }, { "grad_norm": 0.37256920337677, "learning_rate": 5.50365930114774e-05, "loss": 0.0373, "step": 39560 }, { "grad_norm": 0.3539451062679291, "learning_rate": 5.5016029336428255e-05, "loss": 0.0298, "step": 39570 }, { "grad_norm": 0.29592615365982056, "learning_rate": 5.49954648042769e-05, "loss": 0.0382, "step": 39580 }, { "grad_norm": 0.39471113681793213, "learning_rate": 5.4974899418537226e-05, "loss": 0.0375, "step": 39590 }, { "grad_norm": 0.3486326038837433, "learning_rate": 5.4954333182723316e-05, "loss": 0.0291, "step": 39600 }, { "grad_norm": 0.4401375949382782, "learning_rate": 5.493376610034937e-05, "loss": 0.0341, "step": 39610 }, { "grad_norm": 0.5786091685295105, "learning_rate": 5.4913198174929735e-05, "loss": 0.0348, "step": 39620 }, { "grad_norm": 0.24518483877182007, "learning_rate": 5.489262940997889e-05, "loss": 0.0363, "step": 39630 }, { "grad_norm": 0.24121937155723572, "learning_rate": 5.487205980901151e-05, "loss": 0.0376, "step": 39640 }, { "grad_norm": 0.45670920610427856, "learning_rate": 5.485148937554234e-05, "loss": 0.05, "step": 39650 }, { "grad_norm": 0.4665985107421875, "learning_rate": 5.483091811308635e-05, "loss": 0.0316, "step": 39660 }, { "grad_norm": 0.4161461591720581, "learning_rate": 5.4810346025158564e-05, "loss": 0.035, "step": 39670 }, { "grad_norm": 0.3265329599380493, "learning_rate": 5.478977311527421e-05, "loss": 0.0323, "step": 39680 }, { "grad_norm": 0.296413779258728, "learning_rate": 5.476919938694863e-05, "loss": 0.0306, "step": 39690 }, { "grad_norm": 0.8752837777137756, "learning_rate": 5.474862484369733e-05, "loss": 0.0314, "step": 39700 }, { "grad_norm": 0.33689460158348083, "learning_rate": 5.472804948903589e-05, "loss": 0.0332, "step": 39710 }, { "grad_norm": 0.33111125230789185, "learning_rate": 5.470747332648013e-05, "loss": 0.0289, "step": 39720 }, { "grad_norm": 0.47101420164108276, "learning_rate": 5.468689635954591e-05, "loss": 0.044, "step": 39730 }, { "grad_norm": 0.4175773859024048, "learning_rate": 5.46663185917493e-05, "loss": 0.0364, "step": 39740 }, { "grad_norm": 0.4698968827724457, "learning_rate": 5.464574002660645e-05, "loss": 0.0414, "step": 39750 }, { "grad_norm": 0.4889221489429474, "learning_rate": 5.4625160667633724e-05, "loss": 0.0364, "step": 39760 }, { "grad_norm": 0.3363637626171112, "learning_rate": 5.4604580518347505e-05, "loss": 0.0358, "step": 39770 }, { "grad_norm": 0.24544000625610352, "learning_rate": 5.458399958226442e-05, "loss": 0.0359, "step": 39780 }, { "grad_norm": 0.2891780436038971, "learning_rate": 5.456341786290119e-05, "loss": 0.0359, "step": 39790 }, { "grad_norm": 0.31166979670524597, "learning_rate": 5.454283536377465e-05, "loss": 0.0437, "step": 39800 }, { "grad_norm": 0.44798851013183594, "learning_rate": 5.452225208840179e-05, "loss": 0.0372, "step": 39810 }, { "grad_norm": 0.2836386263370514, "learning_rate": 5.450166804029975e-05, "loss": 0.0454, "step": 39820 }, { "grad_norm": 0.37065237760543823, "learning_rate": 5.448108322298574e-05, "loss": 0.0332, "step": 39830 }, { "grad_norm": 0.3849884569644928, "learning_rate": 5.446049763997717e-05, "loss": 0.0332, "step": 39840 }, { "grad_norm": 0.42376160621643066, "learning_rate": 5.4439911294791546e-05, "loss": 0.0279, "step": 39850 }, { "grad_norm": 0.327374130487442, "learning_rate": 5.441932419094652e-05, "loss": 0.0295, "step": 39860 }, { "grad_norm": 0.3939153254032135, "learning_rate": 5.439873633195985e-05, "loss": 0.0299, "step": 39870 }, { "grad_norm": 0.3743765354156494, "learning_rate": 5.437814772134947e-05, "loss": 0.0342, "step": 39880 }, { "grad_norm": 0.7067796587944031, "learning_rate": 5.4357558362633366e-05, "loss": 0.0347, "step": 39890 }, { "grad_norm": 0.3761989176273346, "learning_rate": 5.4336968259329726e-05, "loss": 0.0255, "step": 39900 }, { "grad_norm": 0.3553122580051422, "learning_rate": 5.431637741495681e-05, "loss": 0.0336, "step": 39910 }, { "grad_norm": 0.41687023639678955, "learning_rate": 5.429578583303307e-05, "loss": 0.0364, "step": 39920 }, { "grad_norm": 0.33976054191589355, "learning_rate": 5.427519351707701e-05, "loss": 0.031, "step": 39930 }, { "grad_norm": 0.2959953546524048, "learning_rate": 5.42546004706073e-05, "loss": 0.0317, "step": 39940 }, { "grad_norm": 1.3740901947021484, "learning_rate": 5.4234006697142735e-05, "loss": 0.0371, "step": 39950 }, { "grad_norm": 0.5141887664794922, "learning_rate": 5.421341220020224e-05, "loss": 0.0298, "step": 39960 }, { "grad_norm": 0.37134864926338196, "learning_rate": 5.419281698330482e-05, "loss": 0.0376, "step": 39970 }, { "grad_norm": 0.2945578694343567, "learning_rate": 5.4172221049969665e-05, "loss": 0.0334, "step": 39980 }, { "grad_norm": 0.41557204723358154, "learning_rate": 5.415162440371604e-05, "loss": 0.0369, "step": 39990 }, { "grad_norm": 0.6268703937530518, "learning_rate": 5.413102704806334e-05, "loss": 0.0519, "step": 40000 }, { "grad_norm": 0.40730351209640503, "learning_rate": 5.41104289865311e-05, "loss": 0.0352, "step": 40010 }, { "grad_norm": 0.5615933537483215, "learning_rate": 5.408983022263898e-05, "loss": 0.0364, "step": 40020 }, { "grad_norm": 0.4961620271205902, "learning_rate": 5.406923075990673e-05, "loss": 0.0431, "step": 40030 }, { "grad_norm": 0.474825918674469, "learning_rate": 5.404863060185423e-05, "loss": 0.039, "step": 40040 }, { "grad_norm": 0.3479293882846832, "learning_rate": 5.40280297520015e-05, "loss": 0.0354, "step": 40050 }, { "grad_norm": 0.4284614622592926, "learning_rate": 5.400742821386865e-05, "loss": 0.0329, "step": 40060 }, { "grad_norm": 0.34819692373275757, "learning_rate": 5.398682599097592e-05, "loss": 0.0349, "step": 40070 }, { "grad_norm": 0.39057761430740356, "learning_rate": 5.396622308684367e-05, "loss": 0.0308, "step": 40080 }, { "grad_norm": 0.4842693507671356, "learning_rate": 5.394561950499236e-05, "loss": 0.0376, "step": 40090 }, { "grad_norm": 0.23464076220989227, "learning_rate": 5.39250152489426e-05, "loss": 0.0313, "step": 40100 }, { "grad_norm": 0.34582802653312683, "learning_rate": 5.390441032221507e-05, "loss": 0.0328, "step": 40110 }, { "grad_norm": 0.43984872102737427, "learning_rate": 5.388380472833062e-05, "loss": 0.0374, "step": 40120 }, { "grad_norm": 0.3557141125202179, "learning_rate": 5.386319847081016e-05, "loss": 0.0334, "step": 40130 }, { "grad_norm": 0.42878082394599915, "learning_rate": 5.384259155317473e-05, "loss": 0.0319, "step": 40140 }, { "grad_norm": 0.4507593810558319, "learning_rate": 5.38219839789455e-05, "loss": 0.0334, "step": 40150 }, { "grad_norm": 0.4818006157875061, "learning_rate": 5.380137575164376e-05, "loss": 0.0335, "step": 40160 }, { "grad_norm": 0.31589362025260925, "learning_rate": 5.378076687479085e-05, "loss": 0.0305, "step": 40170 }, { "grad_norm": 0.27855363488197327, "learning_rate": 5.3760157351908305e-05, "loss": 0.0304, "step": 40180 }, { "grad_norm": 0.37675079703330994, "learning_rate": 5.373954718651768e-05, "loss": 0.0299, "step": 40190 }, { "grad_norm": 0.25908976793289185, "learning_rate": 5.371893638214074e-05, "loss": 0.0257, "step": 40200 }, { "grad_norm": 0.6317040920257568, "learning_rate": 5.369832494229927e-05, "loss": 0.0336, "step": 40210 }, { "grad_norm": 0.4485970735549927, "learning_rate": 5.367771287051524e-05, "loss": 0.0341, "step": 40220 }, { "grad_norm": 1.6034787893295288, "learning_rate": 5.365710017031065e-05, "loss": 0.0289, "step": 40230 }, { "grad_norm": 0.390784353017807, "learning_rate": 5.3636486845207654e-05, "loss": 0.0371, "step": 40240 }, { "grad_norm": 0.46748724579811096, "learning_rate": 5.3615872898728526e-05, "loss": 0.0319, "step": 40250 }, { "grad_norm": 0.8081892728805542, "learning_rate": 5.3595258334395614e-05, "loss": 0.0329, "step": 40260 }, { "grad_norm": 0.35841861367225647, "learning_rate": 5.3574643155731396e-05, "loss": 0.0369, "step": 40270 }, { "grad_norm": 0.3137405514717102, "learning_rate": 5.3554027366258405e-05, "loss": 0.0247, "step": 40280 }, { "grad_norm": 0.431375116109848, "learning_rate": 5.3533410969499355e-05, "loss": 0.031, "step": 40290 }, { "grad_norm": 0.34294742345809937, "learning_rate": 5.3512793968977e-05, "loss": 0.0332, "step": 40300 }, { "grad_norm": 0.47509610652923584, "learning_rate": 5.349217636821423e-05, "loss": 0.0381, "step": 40310 }, { "grad_norm": 0.2924949824810028, "learning_rate": 5.347155817073405e-05, "loss": 0.023, "step": 40320 }, { "grad_norm": 0.39490780234336853, "learning_rate": 5.34509393800595e-05, "loss": 0.0331, "step": 40330 }, { "grad_norm": 0.34828031063079834, "learning_rate": 5.343031999971381e-05, "loss": 0.0353, "step": 40340 }, { "grad_norm": 0.3251493573188782, "learning_rate": 5.340970003322026e-05, "loss": 0.0316, "step": 40350 }, { "grad_norm": 0.4339792728424072, "learning_rate": 5.33890794841022e-05, "loss": 0.0304, "step": 40360 }, { "grad_norm": 0.4599890112876892, "learning_rate": 5.336845835588318e-05, "loss": 0.023, "step": 40370 }, { "grad_norm": 0.44158127903938293, "learning_rate": 5.334783665208674e-05, "loss": 0.0346, "step": 40380 }, { "grad_norm": 0.32858407497406006, "learning_rate": 5.332721437623657e-05, "loss": 0.0278, "step": 40390 }, { "grad_norm": 0.3307833969593048, "learning_rate": 5.3306591531856464e-05, "loss": 0.0363, "step": 40400 }, { "grad_norm": 0.35797908902168274, "learning_rate": 5.3285968122470295e-05, "loss": 0.0284, "step": 40410 }, { "grad_norm": 0.31225165724754333, "learning_rate": 5.326534415160207e-05, "loss": 0.0352, "step": 40420 }, { "grad_norm": 0.39181947708129883, "learning_rate": 5.324471962277582e-05, "loss": 0.0356, "step": 40430 }, { "grad_norm": 0.3695782423019409, "learning_rate": 5.3224094539515746e-05, "loss": 0.03, "step": 40440 }, { "grad_norm": 0.8755583763122559, "learning_rate": 5.3203468905346076e-05, "loss": 0.0318, "step": 40450 }, { "grad_norm": 0.42584386467933655, "learning_rate": 5.3182842723791195e-05, "loss": 0.0285, "step": 40460 }, { "grad_norm": 0.3970796465873718, "learning_rate": 5.316221599837554e-05, "loss": 0.0257, "step": 40470 }, { "grad_norm": 0.34541386365890503, "learning_rate": 5.314158873262366e-05, "loss": 0.033, "step": 40480 }, { "grad_norm": 0.4867195785045624, "learning_rate": 5.312096093006018e-05, "loss": 0.0294, "step": 40490 }, { "grad_norm": 0.5824280381202698, "learning_rate": 5.3100332594209846e-05, "loss": 0.0325, "step": 40500 }, { "grad_norm": 0.486189603805542, "learning_rate": 5.3079703728597454e-05, "loss": 0.0427, "step": 40510 }, { "grad_norm": 0.33403509855270386, "learning_rate": 5.305907433674794e-05, "loss": 0.0318, "step": 40520 }, { "grad_norm": 0.34151387214660645, "learning_rate": 5.3038444422186284e-05, "loss": 0.027, "step": 40530 }, { "grad_norm": 0.749370813369751, "learning_rate": 5.301781398843757e-05, "loss": 0.0282, "step": 40540 }, { "grad_norm": 0.4829536974430084, "learning_rate": 5.299718303902699e-05, "loss": 0.0434, "step": 40550 }, { "grad_norm": 0.34980419278144836, "learning_rate": 5.297655157747979e-05, "loss": 0.0358, "step": 40560 }, { "grad_norm": 0.375501811504364, "learning_rate": 5.295591960732136e-05, "loss": 0.0309, "step": 40570 }, { "grad_norm": 0.24628420174121857, "learning_rate": 5.293528713207708e-05, "loss": 0.0326, "step": 40580 }, { "grad_norm": 0.27137497067451477, "learning_rate": 5.291465415527253e-05, "loss": 0.0352, "step": 40590 }, { "grad_norm": 0.36204615235328674, "learning_rate": 5.2894020680433296e-05, "loss": 0.0362, "step": 40600 }, { "grad_norm": 0.36922991275787354, "learning_rate": 5.287338671108507e-05, "loss": 0.0316, "step": 40610 }, { "grad_norm": 0.4074614346027374, "learning_rate": 5.285275225075367e-05, "loss": 0.0313, "step": 40620 }, { "grad_norm": 0.3314659595489502, "learning_rate": 5.283211730296492e-05, "loss": 0.0306, "step": 40630 }, { "grad_norm": 0.4554198086261749, "learning_rate": 5.281148187124477e-05, "loss": 0.0407, "step": 40640 }, { "grad_norm": 0.4565872550010681, "learning_rate": 5.279084595911927e-05, "loss": 0.0325, "step": 40650 }, { "grad_norm": 0.39677858352661133, "learning_rate": 5.277020957011453e-05, "loss": 0.0446, "step": 40660 }, { "grad_norm": 0.31252408027648926, "learning_rate": 5.274957270775673e-05, "loss": 0.0348, "step": 40670 }, { "grad_norm": 0.26025113463401794, "learning_rate": 5.2728935375572164e-05, "loss": 0.0258, "step": 40680 }, { "grad_norm": 0.3229038417339325, "learning_rate": 5.2708297577087176e-05, "loss": 0.0297, "step": 40690 }, { "grad_norm": 0.4572701156139374, "learning_rate": 5.2687659315828196e-05, "loss": 0.0255, "step": 40700 }, { "grad_norm": 0.4798358082771301, "learning_rate": 5.266702059532175e-05, "loss": 0.0252, "step": 40710 }, { "grad_norm": 0.38921889662742615, "learning_rate": 5.264638141909444e-05, "loss": 0.0358, "step": 40720 }, { "grad_norm": 0.329077810049057, "learning_rate": 5.2625741790672925e-05, "loss": 0.0323, "step": 40730 }, { "grad_norm": 1.0030826330184937, "learning_rate": 5.2605101713583935e-05, "loss": 0.0323, "step": 40740 }, { "grad_norm": 0.38927513360977173, "learning_rate": 5.2584461191354315e-05, "loss": 0.037, "step": 40750 }, { "grad_norm": 0.4115799069404602, "learning_rate": 5.2563820227510964e-05, "loss": 0.0325, "step": 40760 }, { "grad_norm": 0.48373275995254517, "learning_rate": 5.2543178825580864e-05, "loss": 0.0314, "step": 40770 }, { "grad_norm": 0.4612232446670532, "learning_rate": 5.2522536989091054e-05, "loss": 0.0384, "step": 40780 }, { "grad_norm": 0.5534194111824036, "learning_rate": 5.2501894721568655e-05, "loss": 0.0335, "step": 40790 }, { "grad_norm": 0.4431593418121338, "learning_rate": 5.248125202654089e-05, "loss": 0.0303, "step": 40800 }, { "grad_norm": 0.3593579828739166, "learning_rate": 5.246060890753501e-05, "loss": 0.0403, "step": 40810 }, { "grad_norm": 0.3266165852546692, "learning_rate": 5.243996536807837e-05, "loss": 0.0283, "step": 40820 }, { "grad_norm": 0.4298870861530304, "learning_rate": 5.2419321411698384e-05, "loss": 0.0349, "step": 40830 }, { "grad_norm": 0.6967145800590515, "learning_rate": 5.239867704192253e-05, "loss": 0.036, "step": 40840 }, { "grad_norm": 0.3031332492828369, "learning_rate": 5.237803226227838e-05, "loss": 0.0312, "step": 40850 }, { "grad_norm": 0.21711193025112152, "learning_rate": 5.235738707629354e-05, "loss": 0.0272, "step": 40860 }, { "grad_norm": 0.32387974858283997, "learning_rate": 5.233674148749575e-05, "loss": 0.0366, "step": 40870 }, { "grad_norm": 0.3908642530441284, "learning_rate": 5.231609549941272e-05, "loss": 0.0374, "step": 40880 }, { "grad_norm": 0.3593493402004242, "learning_rate": 5.229544911557233e-05, "loss": 0.0375, "step": 40890 }, { "grad_norm": 0.3251184821128845, "learning_rate": 5.227480233950246e-05, "loss": 0.0334, "step": 40900 }, { "grad_norm": 0.7394295930862427, "learning_rate": 5.22541551747311e-05, "loss": 0.0405, "step": 40910 }, { "grad_norm": 0.4152580797672272, "learning_rate": 5.223350762478626e-05, "loss": 0.0276, "step": 40920 }, { "grad_norm": 0.35727283358573914, "learning_rate": 5.221285969319606e-05, "loss": 0.0292, "step": 40930 }, { "grad_norm": 0.6202654838562012, "learning_rate": 5.219221138348865e-05, "loss": 0.0297, "step": 40940 }, { "grad_norm": 0.6585808396339417, "learning_rate": 5.217156269919228e-05, "loss": 0.0417, "step": 40950 }, { "grad_norm": 0.2751542627811432, "learning_rate": 5.215091364383523e-05, "loss": 0.0387, "step": 40960 }, { "grad_norm": 0.46106240153312683, "learning_rate": 5.213026422094588e-05, "loss": 0.0369, "step": 40970 }, { "grad_norm": 0.31920289993286133, "learning_rate": 5.210961443405262e-05, "loss": 0.0274, "step": 40980 }, { "grad_norm": 0.5042022466659546, "learning_rate": 5.208896428668396e-05, "loss": 0.0336, "step": 40990 }, { "grad_norm": 0.3646062910556793, "learning_rate": 5.206831378236845e-05, "loss": 0.0347, "step": 41000 }, { "grad_norm": 0.3012208342552185, "learning_rate": 5.2047662924634666e-05, "loss": 0.0349, "step": 41010 }, { "grad_norm": 0.26833075284957886, "learning_rate": 5.20270117170113e-05, "loss": 0.0267, "step": 41020 }, { "grad_norm": 0.5305628776550293, "learning_rate": 5.200636016302707e-05, "loss": 0.049, "step": 41030 }, { "grad_norm": 0.2934819757938385, "learning_rate": 5.1985708266210754e-05, "loss": 0.0294, "step": 41040 }, { "grad_norm": 0.33853545784950256, "learning_rate": 5.1965056030091206e-05, "loss": 0.0379, "step": 41050 }, { "grad_norm": 0.355922669172287, "learning_rate": 5.194440345819732e-05, "loss": 0.0262, "step": 41060 }, { "grad_norm": 0.3420458137989044, "learning_rate": 5.1923750554058084e-05, "loss": 0.0433, "step": 41070 }, { "grad_norm": 0.3673223555088043, "learning_rate": 5.1903097321202476e-05, "loss": 0.0258, "step": 41080 }, { "grad_norm": 0.47498196363449097, "learning_rate": 5.18824437631596e-05, "loss": 0.0492, "step": 41090 }, { "grad_norm": 0.33436301350593567, "learning_rate": 5.186178988345856e-05, "loss": 0.0364, "step": 41100 }, { "grad_norm": 0.40364283323287964, "learning_rate": 5.184113568562855e-05, "loss": 0.0364, "step": 41110 }, { "grad_norm": 0.46328720450401306, "learning_rate": 5.18204811731988e-05, "loss": 0.039, "step": 41120 }, { "grad_norm": 0.3857249319553375, "learning_rate": 5.179982634969862e-05, "loss": 0.0326, "step": 41130 }, { "grad_norm": 0.530927300453186, "learning_rate": 5.177917121865734e-05, "loss": 0.0401, "step": 41140 }, { "grad_norm": 0.3831010162830353, "learning_rate": 5.1758515783604346e-05, "loss": 0.0294, "step": 41150 }, { "grad_norm": 0.3343365788459778, "learning_rate": 5.173786004806911e-05, "loss": 0.0432, "step": 41160 }, { "grad_norm": 0.6574912071228027, "learning_rate": 5.1717204015581135e-05, "loss": 0.0399, "step": 41170 }, { "grad_norm": 0.318745881319046, "learning_rate": 5.169654768966994e-05, "loss": 0.0295, "step": 41180 }, { "grad_norm": 0.40453943610191345, "learning_rate": 5.1675891073865156e-05, "loss": 0.0383, "step": 41190 }, { "grad_norm": 0.3645862340927124, "learning_rate": 5.1655234171696424e-05, "loss": 0.0263, "step": 41200 }, { "grad_norm": 0.2854591906070709, "learning_rate": 5.163457698669343e-05, "loss": 0.0283, "step": 41210 }, { "grad_norm": 0.48246586322784424, "learning_rate": 5.1613919522385936e-05, "loss": 0.0317, "step": 41220 }, { "grad_norm": 0.4315643906593323, "learning_rate": 5.1593261782303746e-05, "loss": 0.0307, "step": 41230 }, { "grad_norm": 0.27411696314811707, "learning_rate": 5.157260376997669e-05, "loss": 0.0264, "step": 41240 }, { "grad_norm": 0.5576100945472717, "learning_rate": 5.155194548893466e-05, "loss": 0.0344, "step": 41250 }, { "grad_norm": 0.25629910826683044, "learning_rate": 5.153128694270759e-05, "loss": 0.0373, "step": 41260 }, { "grad_norm": 0.4990192949771881, "learning_rate": 5.151062813482548e-05, "loss": 0.0369, "step": 41270 }, { "grad_norm": 0.2692052125930786, "learning_rate": 5.148996906881832e-05, "loss": 0.0243, "step": 41280 }, { "grad_norm": 0.29606151580810547, "learning_rate": 5.1469309748216196e-05, "loss": 0.03, "step": 41290 }, { "grad_norm": 0.2621966004371643, "learning_rate": 5.144865017654923e-05, "loss": 0.0246, "step": 41300 }, { "grad_norm": 0.3691510856151581, "learning_rate": 5.1427990357347556e-05, "loss": 0.0415, "step": 41310 }, { "grad_norm": 0.333368182182312, "learning_rate": 5.140733029414139e-05, "loss": 0.027, "step": 41320 }, { "grad_norm": 0.35585159063339233, "learning_rate": 5.138666999046098e-05, "loss": 0.0282, "step": 41330 }, { "grad_norm": 0.44528821110725403, "learning_rate": 5.136600944983658e-05, "loss": 0.0402, "step": 41340 }, { "grad_norm": 0.382576584815979, "learning_rate": 5.134534867579853e-05, "loss": 0.0323, "step": 41350 }, { "grad_norm": 0.3805457055568695, "learning_rate": 5.1324687671877184e-05, "loss": 0.0253, "step": 41360 }, { "grad_norm": 0.3493150472640991, "learning_rate": 5.130402644160296e-05, "loss": 0.0342, "step": 41370 }, { "grad_norm": 0.30834102630615234, "learning_rate": 5.128336498850628e-05, "loss": 0.0286, "step": 41380 }, { "grad_norm": 0.3276524841785431, "learning_rate": 5.126270331611761e-05, "loss": 0.0273, "step": 41390 }, { "grad_norm": 0.5277291536331177, "learning_rate": 5.124204142796748e-05, "loss": 0.03, "step": 41400 }, { "grad_norm": 0.26708272099494934, "learning_rate": 5.122137932758644e-05, "loss": 0.0292, "step": 41410 }, { "grad_norm": 0.38680365681648254, "learning_rate": 5.1200717018505074e-05, "loss": 0.0308, "step": 41420 }, { "grad_norm": 0.20384512841701508, "learning_rate": 5.118005450425403e-05, "loss": 0.0261, "step": 41430 }, { "grad_norm": 0.2846038043498993, "learning_rate": 5.115939178836392e-05, "loss": 0.0261, "step": 41440 }, { "grad_norm": 0.2704465091228485, "learning_rate": 5.113872887436547e-05, "loss": 0.0305, "step": 41450 }, { "grad_norm": 0.46934500336647034, "learning_rate": 5.11180657657894e-05, "loss": 0.0321, "step": 41460 }, { "grad_norm": 0.30290648341178894, "learning_rate": 5.109740246616649e-05, "loss": 0.0242, "step": 41470 }, { "grad_norm": 0.5845567584037781, "learning_rate": 5.10767389790275e-05, "loss": 0.0459, "step": 41480 }, { "grad_norm": 0.33738335967063904, "learning_rate": 5.1056075307903265e-05, "loss": 0.0353, "step": 41490 }, { "grad_norm": 0.300060510635376, "learning_rate": 5.1035411456324634e-05, "loss": 0.0256, "step": 41500 }, { "grad_norm": 0.2785648703575134, "learning_rate": 5.1014747427822504e-05, "loss": 0.0343, "step": 41510 }, { "grad_norm": 0.5445599555969238, "learning_rate": 5.0994083225927804e-05, "loss": 0.0322, "step": 41520 }, { "grad_norm": 0.8540360927581787, "learning_rate": 5.0973418854171475e-05, "loss": 0.0289, "step": 41530 }, { "grad_norm": 0.37034156918525696, "learning_rate": 5.095275431608447e-05, "loss": 0.0307, "step": 41540 }, { "grad_norm": 0.29469215869903564, "learning_rate": 5.093208961519782e-05, "loss": 0.0314, "step": 41550 }, { "grad_norm": 0.3586345314979553, "learning_rate": 5.091142475504255e-05, "loss": 0.0277, "step": 41560 }, { "grad_norm": 0.471769779920578, "learning_rate": 5.089075973914971e-05, "loss": 0.0341, "step": 41570 }, { "grad_norm": 0.4530966877937317, "learning_rate": 5.08700945710504e-05, "loss": 0.0413, "step": 41580 }, { "grad_norm": 0.31916114687919617, "learning_rate": 5.0849429254275714e-05, "loss": 0.0279, "step": 41590 }, { "grad_norm": 0.3194199800491333, "learning_rate": 5.0828763792356804e-05, "loss": 0.0291, "step": 41600 }, { "grad_norm": 0.3143458664417267, "learning_rate": 5.080809818882483e-05, "loss": 0.0318, "step": 41610 }, { "grad_norm": 0.43017709255218506, "learning_rate": 5.078743244721097e-05, "loss": 0.0259, "step": 41620 }, { "grad_norm": 0.3000352084636688, "learning_rate": 5.0766766571046455e-05, "loss": 0.0259, "step": 41630 }, { "grad_norm": 0.29107511043548584, "learning_rate": 5.07461005638625e-05, "loss": 0.0241, "step": 41640 }, { "grad_norm": 0.7141829133033752, "learning_rate": 5.072543442919037e-05, "loss": 0.0296, "step": 41650 }, { "grad_norm": 0.26688119769096375, "learning_rate": 5.070476817056132e-05, "loss": 0.0345, "step": 41660 }, { "grad_norm": 0.3929726183414459, "learning_rate": 5.068410179150668e-05, "loss": 0.0261, "step": 41670 }, { "grad_norm": 0.25349920988082886, "learning_rate": 5.066343529555775e-05, "loss": 0.0318, "step": 41680 }, { "grad_norm": 0.524384617805481, "learning_rate": 5.064276868624587e-05, "loss": 0.0297, "step": 41690 }, { "grad_norm": 0.26093506813049316, "learning_rate": 5.0622101967102396e-05, "loss": 0.0322, "step": 41700 }, { "grad_norm": 0.41737374663352966, "learning_rate": 5.0601435141658714e-05, "loss": 0.0305, "step": 41710 }, { "grad_norm": 0.35351917147636414, "learning_rate": 5.058076821344622e-05, "loss": 0.035, "step": 41720 }, { "grad_norm": 0.3936912417411804, "learning_rate": 5.056010118599632e-05, "loss": 0.0399, "step": 41730 }, { "grad_norm": 0.6662709712982178, "learning_rate": 5.053943406284044e-05, "loss": 0.0371, "step": 41740 }, { "grad_norm": 0.5306934118270874, "learning_rate": 5.051876684751006e-05, "loss": 0.0307, "step": 41750 }, { "grad_norm": 0.3567822575569153, "learning_rate": 5.0498099543536584e-05, "loss": 0.0289, "step": 41760 }, { "grad_norm": 0.31774160265922546, "learning_rate": 5.047743215445152e-05, "loss": 0.0321, "step": 41770 }, { "grad_norm": 0.3796725571155548, "learning_rate": 5.045676468378637e-05, "loss": 0.0342, "step": 41780 }, { "grad_norm": 0.3501778244972229, "learning_rate": 5.0436097135072626e-05, "loss": 0.0361, "step": 41790 }, { "grad_norm": 0.28642183542251587, "learning_rate": 5.041542951184181e-05, "loss": 0.0308, "step": 41800 }, { "grad_norm": 0.3117964267730713, "learning_rate": 5.039476181762545e-05, "loss": 0.0224, "step": 41810 }, { "grad_norm": 0.27007558941841125, "learning_rate": 5.037409405595508e-05, "loss": 0.0253, "step": 41820 }, { "grad_norm": 0.43623030185699463, "learning_rate": 5.035342623036229e-05, "loss": 0.027, "step": 41830 }, { "grad_norm": 0.4378749430179596, "learning_rate": 5.033275834437862e-05, "loss": 0.0328, "step": 41840 }, { "grad_norm": 0.5095057487487793, "learning_rate": 5.031209040153564e-05, "loss": 0.0261, "step": 41850 }, { "grad_norm": 0.464733749628067, "learning_rate": 5.0291422405364955e-05, "loss": 0.0414, "step": 41860 }, { "grad_norm": 0.5523807406425476, "learning_rate": 5.0270754359398133e-05, "loss": 0.031, "step": 41870 }, { "grad_norm": 0.5732932686805725, "learning_rate": 5.025008626716682e-05, "loss": 0.0241, "step": 41880 }, { "grad_norm": 0.4840470552444458, "learning_rate": 5.0229418132202585e-05, "loss": 0.0288, "step": 41890 }, { "grad_norm": 0.3238082826137543, "learning_rate": 5.020874995803707e-05, "loss": 0.0333, "step": 41900 }, { "grad_norm": 0.3847835063934326, "learning_rate": 5.01880817482019e-05, "loss": 0.032, "step": 41910 }, { "grad_norm": 0.4614737927913666, "learning_rate": 5.01674135062287e-05, "loss": 0.0304, "step": 41920 }, { "grad_norm": 0.3758070468902588, "learning_rate": 5.0146745235649115e-05, "loss": 0.0359, "step": 41930 }, { "grad_norm": 0.3970884084701538, "learning_rate": 5.0126076939994795e-05, "loss": 0.0292, "step": 41940 }, { "grad_norm": 0.5471968650817871, "learning_rate": 5.010540862279736e-05, "loss": 0.0338, "step": 41950 }, { "grad_norm": 0.32728222012519836, "learning_rate": 5.008474028758846e-05, "loss": 0.0295, "step": 41960 }, { "grad_norm": 0.3684370219707489, "learning_rate": 5.0064071937899784e-05, "loss": 0.0273, "step": 41970 }, { "grad_norm": 0.4040882885456085, "learning_rate": 5.004340357726296e-05, "loss": 0.0282, "step": 41980 }, { "grad_norm": 0.3381153345108032, "learning_rate": 5.002273520920965e-05, "loss": 0.0333, "step": 41990 }, { "grad_norm": 0.34581831097602844, "learning_rate": 5.000206683727151e-05, "loss": 0.0295, "step": 42000 }, { "grad_norm": 0.45437923073768616, "learning_rate": 4.998139846498021e-05, "loss": 0.0308, "step": 42010 }, { "grad_norm": 0.5957505106925964, "learning_rate": 4.9960730095867405e-05, "loss": 0.031, "step": 42020 }, { "grad_norm": 0.32587388157844543, "learning_rate": 4.9940061733464755e-05, "loss": 0.0365, "step": 42030 }, { "grad_norm": 0.5404424667358398, "learning_rate": 4.991939338130392e-05, "loss": 0.041, "step": 42040 }, { "grad_norm": 0.40745559334754944, "learning_rate": 4.989872504291653e-05, "loss": 0.0393, "step": 42050 }, { "grad_norm": 0.3939675986766815, "learning_rate": 4.9878056721834273e-05, "loss": 0.0289, "step": 42060 }, { "grad_norm": 0.3937116861343384, "learning_rate": 4.98573884215888e-05, "loss": 0.0294, "step": 42070 }, { "grad_norm": 0.458609014749527, "learning_rate": 4.9836720145711715e-05, "loss": 0.032, "step": 42080 }, { "grad_norm": 0.3674200475215912, "learning_rate": 4.9816051897734725e-05, "loss": 0.0349, "step": 42090 }, { "grad_norm": 0.4145602583885193, "learning_rate": 4.979538368118941e-05, "loss": 0.0314, "step": 42100 }, { "grad_norm": 0.2952553927898407, "learning_rate": 4.9774715499607446e-05, "loss": 0.0343, "step": 42110 }, { "grad_norm": 0.4234563410282135, "learning_rate": 4.9754047356520416e-05, "loss": 0.036, "step": 42120 }, { "grad_norm": 0.3127861022949219, "learning_rate": 4.973337925545999e-05, "loss": 0.0269, "step": 42130 }, { "grad_norm": 0.27206674218177795, "learning_rate": 4.9712711199957746e-05, "loss": 0.0302, "step": 42140 }, { "grad_norm": 0.3254161775112152, "learning_rate": 4.969204319354531e-05, "loss": 0.0302, "step": 42150 }, { "grad_norm": 0.45194029808044434, "learning_rate": 4.9671375239754267e-05, "loss": 0.0283, "step": 42160 }, { "grad_norm": 0.2915351390838623, "learning_rate": 4.9650707342116195e-05, "loss": 0.0303, "step": 42170 }, { "grad_norm": 0.3978866636753082, "learning_rate": 4.96300395041627e-05, "loss": 0.0294, "step": 42180 }, { "grad_norm": 0.5183194875717163, "learning_rate": 4.960937172942532e-05, "loss": 0.0292, "step": 42190 }, { "grad_norm": 0.28836023807525635, "learning_rate": 4.958870402143565e-05, "loss": 0.0257, "step": 42200 }, { "grad_norm": 0.24203498661518097, "learning_rate": 4.9568036383725186e-05, "loss": 0.0355, "step": 42210 }, { "grad_norm": 0.3619263768196106, "learning_rate": 4.9547368819825516e-05, "loss": 0.0289, "step": 42220 }, { "grad_norm": 0.5059977173805237, "learning_rate": 4.952670133326812e-05, "loss": 0.0307, "step": 42230 }, { "grad_norm": 0.5222558379173279, "learning_rate": 4.950603392758453e-05, "loss": 0.0398, "step": 42240 }, { "grad_norm": 0.36607468128204346, "learning_rate": 4.948536660630621e-05, "loss": 0.0297, "step": 42250 }, { "grad_norm": 0.3024350106716156, "learning_rate": 4.9464699372964685e-05, "loss": 0.0417, "step": 42260 }, { "grad_norm": 0.31672874093055725, "learning_rate": 4.9444032231091395e-05, "loss": 0.0302, "step": 42270 }, { "grad_norm": 0.3793778121471405, "learning_rate": 4.9423365184217776e-05, "loss": 0.0297, "step": 42280 }, { "grad_norm": 0.36770710349082947, "learning_rate": 4.940269823587529e-05, "loss": 0.0301, "step": 42290 }, { "grad_norm": 0.4187788963317871, "learning_rate": 4.938203138959533e-05, "loss": 0.0262, "step": 42300 }, { "grad_norm": 0.4708707630634308, "learning_rate": 4.9361364648909325e-05, "loss": 0.0303, "step": 42310 }, { "grad_norm": 0.4184524118900299, "learning_rate": 4.9340698017348605e-05, "loss": 0.0241, "step": 42320 }, { "grad_norm": 0.4093882739543915, "learning_rate": 4.932003149844458e-05, "loss": 0.0265, "step": 42330 }, { "grad_norm": 0.2837868928909302, "learning_rate": 4.929936509572857e-05, "loss": 0.0314, "step": 42340 }, { "grad_norm": 0.4254237711429596, "learning_rate": 4.927869881273191e-05, "loss": 0.032, "step": 42350 }, { "grad_norm": 0.34733808040618896, "learning_rate": 4.9258032652985894e-05, "loss": 0.0248, "step": 42360 }, { "grad_norm": 0.41634419560432434, "learning_rate": 4.9237366620021786e-05, "loss": 0.0292, "step": 42370 }, { "grad_norm": 0.22392192482948303, "learning_rate": 4.921670071737089e-05, "loss": 0.0262, "step": 42380 }, { "grad_norm": 0.3875906467437744, "learning_rate": 4.91960349485644e-05, "loss": 0.0287, "step": 42390 }, { "grad_norm": 0.41154393553733826, "learning_rate": 4.9175369317133566e-05, "loss": 0.0269, "step": 42400 }, { "grad_norm": 0.225599005818367, "learning_rate": 4.915470382660953e-05, "loss": 0.0296, "step": 42410 }, { "grad_norm": 0.46651241183280945, "learning_rate": 4.9134038480523524e-05, "loss": 0.0215, "step": 42420 }, { "grad_norm": 0.3684956133365631, "learning_rate": 4.911337328240664e-05, "loss": 0.0288, "step": 42430 }, { "grad_norm": 0.3021056652069092, "learning_rate": 4.909270823579003e-05, "loss": 0.0259, "step": 42440 }, { "grad_norm": 0.40927305817604065, "learning_rate": 4.907204334420476e-05, "loss": 0.0253, "step": 42450 }, { "grad_norm": 0.3275335431098938, "learning_rate": 4.9051378611181884e-05, "loss": 0.0291, "step": 42460 }, { "grad_norm": 0.8620375990867615, "learning_rate": 4.9030714040252486e-05, "loss": 0.0334, "step": 42470 }, { "grad_norm": 0.2971115708351135, "learning_rate": 4.901004963494752e-05, "loss": 0.0262, "step": 42480 }, { "grad_norm": 0.36846551299095154, "learning_rate": 4.898938539879802e-05, "loss": 0.0348, "step": 42490 }, { "grad_norm": 0.3347975015640259, "learning_rate": 4.8968721335334904e-05, "loss": 0.0315, "step": 42500 }, { "grad_norm": 0.33382663130760193, "learning_rate": 4.894805744808912e-05, "loss": 0.0329, "step": 42510 }, { "grad_norm": 0.36885902285575867, "learning_rate": 4.8927393740591524e-05, "loss": 0.0298, "step": 42520 }, { "grad_norm": 0.40259942412376404, "learning_rate": 4.890673021637302e-05, "loss": 0.0301, "step": 42530 }, { "grad_norm": 0.29339730739593506, "learning_rate": 4.888606687896442e-05, "loss": 0.0245, "step": 42540 }, { "grad_norm": 0.31524989008903503, "learning_rate": 4.886540373189652e-05, "loss": 0.0286, "step": 42550 }, { "grad_norm": 0.31028908491134644, "learning_rate": 4.88447407787001e-05, "loss": 0.0308, "step": 42560 }, { "grad_norm": 0.44316455721855164, "learning_rate": 4.8824078022905864e-05, "loss": 0.0227, "step": 42570 }, { "grad_norm": 0.3046759068965912, "learning_rate": 4.880341546804455e-05, "loss": 0.0261, "step": 42580 }, { "grad_norm": 0.36051228642463684, "learning_rate": 4.878275311764679e-05, "loss": 0.0318, "step": 42590 }, { "grad_norm": 0.32870182394981384, "learning_rate": 4.8762090975243254e-05, "loss": 0.0297, "step": 42600 }, { "grad_norm": 0.2538427412509918, "learning_rate": 4.874142904436448e-05, "loss": 0.0315, "step": 42610 }, { "grad_norm": 0.24767035245895386, "learning_rate": 4.8720767328541094e-05, "loss": 0.0412, "step": 42620 }, { "grad_norm": 0.3360055983066559, "learning_rate": 4.870010583130358e-05, "loss": 0.0302, "step": 42630 }, { "grad_norm": 0.41277387738227844, "learning_rate": 4.867944455618241e-05, "loss": 0.0262, "step": 42640 }, { "grad_norm": 0.3483937382698059, "learning_rate": 4.865878350670807e-05, "loss": 0.0321, "step": 42650 }, { "grad_norm": 0.37387073040008545, "learning_rate": 4.8638122686410914e-05, "loss": 0.0302, "step": 42660 }, { "grad_norm": 0.40535491704940796, "learning_rate": 4.861746209882137e-05, "loss": 0.0219, "step": 42670 }, { "grad_norm": 0.4148902893066406, "learning_rate": 4.859680174746972e-05, "loss": 0.0285, "step": 42680 }, { "grad_norm": 0.4884273409843445, "learning_rate": 4.857614163588629e-05, "loss": 0.0389, "step": 42690 }, { "grad_norm": 0.24276241660118103, "learning_rate": 4.855548176760131e-05, "loss": 0.0299, "step": 42700 }, { "grad_norm": 0.24643608927726746, "learning_rate": 4.8534822146144986e-05, "loss": 0.0309, "step": 42710 }, { "grad_norm": 0.4723329544067383, "learning_rate": 4.851416277504749e-05, "loss": 0.0366, "step": 42720 }, { "grad_norm": 0.5606285333633423, "learning_rate": 4.8493503657838923e-05, "loss": 0.0254, "step": 42730 }, { "grad_norm": 0.4919135868549347, "learning_rate": 4.84728447980494e-05, "loss": 0.0319, "step": 42740 }, { "grad_norm": 0.4006243944168091, "learning_rate": 4.8452186199208914e-05, "loss": 0.0335, "step": 42750 }, { "grad_norm": 0.39273929595947266, "learning_rate": 4.843152786484749e-05, "loss": 0.0304, "step": 42760 }, { "grad_norm": 0.2819238305091858, "learning_rate": 4.8410869798495035e-05, "loss": 0.0275, "step": 42770 }, { "grad_norm": 0.5931539535522461, "learning_rate": 4.8390212003681486e-05, "loss": 0.033, "step": 42780 }, { "grad_norm": 0.37709254026412964, "learning_rate": 4.836955448393667e-05, "loss": 0.0344, "step": 42790 }, { "grad_norm": 0.3990976810455322, "learning_rate": 4.8348897242790406e-05, "loss": 0.0317, "step": 42800 }, { "grad_norm": 0.31498152017593384, "learning_rate": 4.832824028377243e-05, "loss": 0.0256, "step": 42810 }, { "grad_norm": 0.34512564539909363, "learning_rate": 4.830758361041249e-05, "loss": 0.0269, "step": 42820 }, { "grad_norm": 1.059679627418518, "learning_rate": 4.828692722624022e-05, "loss": 0.0375, "step": 42830 }, { "grad_norm": 0.8889297246932983, "learning_rate": 4.826627113478522e-05, "loss": 0.0444, "step": 42840 }, { "grad_norm": 0.37251681089401245, "learning_rate": 4.824561533957708e-05, "loss": 0.0283, "step": 42850 }, { "grad_norm": 0.595277726650238, "learning_rate": 4.822495984414527e-05, "loss": 0.0268, "step": 42860 }, { "grad_norm": 0.48169848322868347, "learning_rate": 4.8204304652019304e-05, "loss": 0.0286, "step": 42870 }, { "grad_norm": 0.411215215921402, "learning_rate": 4.8183649766728537e-05, "loss": 0.0254, "step": 42880 }, { "grad_norm": 0.4480641484260559, "learning_rate": 4.816299519180237e-05, "loss": 0.031, "step": 42890 }, { "grad_norm": 0.29636961221694946, "learning_rate": 4.8142340930770066e-05, "loss": 0.0282, "step": 42900 }, { "grad_norm": 0.7522193193435669, "learning_rate": 4.81216869871609e-05, "loss": 0.0245, "step": 42910 }, { "grad_norm": 0.31628143787384033, "learning_rate": 4.8101033364504064e-05, "loss": 0.0286, "step": 42920 }, { "grad_norm": 0.28428351879119873, "learning_rate": 4.808038006632866e-05, "loss": 0.0325, "step": 42930 }, { "grad_norm": 0.3552131950855255, "learning_rate": 4.805972709616383e-05, "loss": 0.0367, "step": 42940 }, { "grad_norm": 1.2539016008377075, "learning_rate": 4.8039074457538556e-05, "loss": 0.0307, "step": 42950 }, { "grad_norm": 0.2853073477745056, "learning_rate": 4.801842215398184e-05, "loss": 0.0315, "step": 42960 }, { "grad_norm": 0.3586648404598236, "learning_rate": 4.799777018902256e-05, "loss": 0.0339, "step": 42970 }, { "grad_norm": 0.5469627976417542, "learning_rate": 4.797711856618961e-05, "loss": 0.0337, "step": 42980 }, { "grad_norm": 0.31834596395492554, "learning_rate": 4.7956467289011765e-05, "loss": 0.028, "step": 42990 }, { "grad_norm": 0.4940647482872009, "learning_rate": 4.793581636101778e-05, "loss": 0.0302, "step": 43000 }, { "grad_norm": 0.405996173620224, "learning_rate": 4.7915165785736326e-05, "loss": 0.029, "step": 43010 }, { "grad_norm": 0.2954234480857849, "learning_rate": 4.789451556669599e-05, "loss": 0.0269, "step": 43020 }, { "grad_norm": 0.3434615135192871, "learning_rate": 4.7873865707425384e-05, "loss": 0.0327, "step": 43030 }, { "grad_norm": 0.549415111541748, "learning_rate": 4.7853216211452975e-05, "loss": 0.03, "step": 43040 }, { "grad_norm": 0.36316099762916565, "learning_rate": 4.78325670823072e-05, "loss": 0.026, "step": 43050 }, { "grad_norm": 0.3870411813259125, "learning_rate": 4.781191832351641e-05, "loss": 0.0324, "step": 43060 }, { "grad_norm": 0.3153805136680603, "learning_rate": 4.7791269938608965e-05, "loss": 0.0313, "step": 43070 }, { "grad_norm": 0.2644408047199249, "learning_rate": 4.777062193111305e-05, "loss": 0.0309, "step": 43080 }, { "grad_norm": 0.29886093735694885, "learning_rate": 4.77499743045569e-05, "loss": 0.0346, "step": 43090 }, { "grad_norm": 0.3334413766860962, "learning_rate": 4.7729327062468596e-05, "loss": 0.028, "step": 43100 }, { "grad_norm": 0.36285409331321716, "learning_rate": 4.770868020837619e-05, "loss": 0.0355, "step": 43110 }, { "grad_norm": 0.28638505935668945, "learning_rate": 4.768803374580768e-05, "loss": 0.0221, "step": 43120 }, { "grad_norm": 0.44857627153396606, "learning_rate": 4.766738767829094e-05, "loss": 0.032, "step": 43130 }, { "grad_norm": 0.363700807094574, "learning_rate": 4.764674200935388e-05, "loss": 0.0275, "step": 43140 }, { "grad_norm": 0.403158962726593, "learning_rate": 4.762609674252424e-05, "loss": 0.0272, "step": 43150 }, { "grad_norm": 0.378235399723053, "learning_rate": 4.760545188132974e-05, "loss": 0.0408, "step": 43160 }, { "grad_norm": 0.2513982951641083, "learning_rate": 4.758480742929801e-05, "loss": 0.0271, "step": 43170 }, { "grad_norm": 0.36193495988845825, "learning_rate": 4.756416338995664e-05, "loss": 0.0369, "step": 43180 }, { "grad_norm": 0.5957789421081543, "learning_rate": 4.7543519766833145e-05, "loss": 0.0319, "step": 43190 }, { "grad_norm": 0.3601643443107605, "learning_rate": 4.752287656345492e-05, "loss": 0.0315, "step": 43200 }, { "grad_norm": 0.43203258514404297, "learning_rate": 4.7502233783349355e-05, "loss": 0.0332, "step": 43210 }, { "grad_norm": 0.4529266953468323, "learning_rate": 4.7481591430043694e-05, "loss": 0.0247, "step": 43220 }, { "grad_norm": 0.2553369402885437, "learning_rate": 4.746094950706521e-05, "loss": 0.0286, "step": 43230 }, { "grad_norm": 0.3193790316581726, "learning_rate": 4.744030801794099e-05, "loss": 0.031, "step": 43240 }, { "grad_norm": 0.44174742698669434, "learning_rate": 4.741966696619813e-05, "loss": 0.0345, "step": 43250 }, { "grad_norm": 0.2654026746749878, "learning_rate": 4.739902635536361e-05, "loss": 0.0299, "step": 43260 }, { "grad_norm": 0.39372172951698303, "learning_rate": 4.737838618896436e-05, "loss": 0.0244, "step": 43270 }, { "grad_norm": 0.5490168333053589, "learning_rate": 4.7357746470527203e-05, "loss": 0.0311, "step": 43280 }, { "grad_norm": 0.41679954528808594, "learning_rate": 4.73371072035789e-05, "loss": 0.0325, "step": 43290 }, { "grad_norm": 0.41244205832481384, "learning_rate": 4.731646839164616e-05, "loss": 0.0313, "step": 43300 }, { "grad_norm": 0.4470407962799072, "learning_rate": 4.729583003825556e-05, "loss": 0.03, "step": 43310 }, { "grad_norm": 0.3136054575443268, "learning_rate": 4.727519214693367e-05, "loss": 0.0323, "step": 43320 }, { "grad_norm": 0.3921659588813782, "learning_rate": 4.725455472120689e-05, "loss": 0.0355, "step": 43330 }, { "grad_norm": 0.19752629101276398, "learning_rate": 4.723391776460164e-05, "loss": 0.0317, "step": 43340 }, { "grad_norm": 0.40870386362075806, "learning_rate": 4.7213281280644186e-05, "loss": 0.0275, "step": 43350 }, { "grad_norm": 0.3904080092906952, "learning_rate": 4.719264527286075e-05, "loss": 0.0316, "step": 43360 }, { "grad_norm": 0.4083063006401062, "learning_rate": 4.717200974477744e-05, "loss": 0.0328, "step": 43370 }, { "grad_norm": 0.3320886194705963, "learning_rate": 4.715137469992034e-05, "loss": 0.0312, "step": 43380 }, { "grad_norm": 0.42670318484306335, "learning_rate": 4.713074014181539e-05, "loss": 0.0251, "step": 43390 }, { "grad_norm": 0.5396502017974854, "learning_rate": 4.711010607398846e-05, "loss": 0.0241, "step": 43400 }, { "grad_norm": 0.4450935125350952, "learning_rate": 4.708947249996538e-05, "loss": 0.0245, "step": 43410 }, { "grad_norm": 0.3655507266521454, "learning_rate": 4.706883942327183e-05, "loss": 0.0242, "step": 43420 }, { "grad_norm": 0.31558480858802795, "learning_rate": 4.704820684743347e-05, "loss": 0.0331, "step": 43430 }, { "grad_norm": 0.2861931324005127, "learning_rate": 4.702757477597581e-05, "loss": 0.0283, "step": 43440 }, { "grad_norm": 0.30346232652664185, "learning_rate": 4.700694321242435e-05, "loss": 0.0301, "step": 43450 }, { "grad_norm": 0.7922858595848083, "learning_rate": 4.698631216030441e-05, "loss": 0.0346, "step": 43460 }, { "grad_norm": 0.2656305432319641, "learning_rate": 4.6965681623141314e-05, "loss": 0.0241, "step": 43470 }, { "grad_norm": 0.2945968210697174, "learning_rate": 4.694505160446024e-05, "loss": 0.0204, "step": 43480 }, { "grad_norm": 0.2862091362476349, "learning_rate": 4.692442210778628e-05, "loss": 0.0286, "step": 43490 }, { "grad_norm": 0.4942086637020111, "learning_rate": 4.690379313664448e-05, "loss": 0.0291, "step": 43500 }, { "grad_norm": 0.3173930048942566, "learning_rate": 4.688316469455973e-05, "loss": 0.0271, "step": 43510 }, { "grad_norm": 0.2041737139225006, "learning_rate": 4.6862536785056914e-05, "loss": 0.0299, "step": 43520 }, { "grad_norm": 0.31770947575569153, "learning_rate": 4.684190941166073e-05, "loss": 0.0327, "step": 43530 }, { "grad_norm": 0.33671626448631287, "learning_rate": 4.682128257789587e-05, "loss": 0.0311, "step": 43540 }, { "grad_norm": 0.40721753239631653, "learning_rate": 4.680065628728687e-05, "loss": 0.0303, "step": 43550 }, { "grad_norm": 0.35210946202278137, "learning_rate": 4.678003054335822e-05, "loss": 0.0336, "step": 43560 }, { "grad_norm": 0.35470709204673767, "learning_rate": 4.675940534963428e-05, "loss": 0.0297, "step": 43570 }, { "grad_norm": 0.38444632291793823, "learning_rate": 4.673878070963931e-05, "loss": 0.0357, "step": 43580 }, { "grad_norm": 0.3260277509689331, "learning_rate": 4.671815662689756e-05, "loss": 0.0258, "step": 43590 }, { "grad_norm": 0.30281221866607666, "learning_rate": 4.669753310493306e-05, "loss": 0.0325, "step": 43600 }, { "grad_norm": 0.40637242794036865, "learning_rate": 4.6676910147269845e-05, "loss": 0.0333, "step": 43610 }, { "grad_norm": 0.3197762668132782, "learning_rate": 4.665628775743177e-05, "loss": 0.0276, "step": 43620 }, { "grad_norm": 0.4293917417526245, "learning_rate": 4.66356659389427e-05, "loss": 0.0288, "step": 43630 }, { "grad_norm": 0.37935200333595276, "learning_rate": 4.661504469532629e-05, "loss": 0.0274, "step": 43640 }, { "grad_norm": 0.3242097795009613, "learning_rate": 4.659442403010617e-05, "loss": 0.026, "step": 43650 }, { "grad_norm": 0.2806794345378876, "learning_rate": 4.6573803946805845e-05, "loss": 0.0215, "step": 43660 }, { "grad_norm": 0.2990965247154236, "learning_rate": 4.6553184448948695e-05, "loss": 0.0325, "step": 43670 }, { "grad_norm": 0.26399362087249756, "learning_rate": 4.653256554005807e-05, "loss": 0.0299, "step": 43680 }, { "grad_norm": 0.25540366768836975, "learning_rate": 4.6511947223657145e-05, "loss": 0.0335, "step": 43690 }, { "grad_norm": 0.5110959410667419, "learning_rate": 4.649132950326906e-05, "loss": 0.0317, "step": 43700 }, { "grad_norm": 0.36846983432769775, "learning_rate": 4.647071238241679e-05, "loss": 0.0322, "step": 43710 }, { "grad_norm": 0.3077659606933594, "learning_rate": 4.6450095864623264e-05, "loss": 0.0309, "step": 43720 }, { "grad_norm": 0.43964794278144836, "learning_rate": 4.642947995341123e-05, "loss": 0.0324, "step": 43730 }, { "grad_norm": 0.40640705823898315, "learning_rate": 4.640886465230345e-05, "loss": 0.0308, "step": 43740 }, { "grad_norm": 0.3840571343898773, "learning_rate": 4.6388249964822485e-05, "loss": 0.0348, "step": 43750 }, { "grad_norm": 0.537630558013916, "learning_rate": 4.63676358944908e-05, "loss": 0.0282, "step": 43760 }, { "grad_norm": 0.5493066906929016, "learning_rate": 4.634702244483081e-05, "loss": 0.0354, "step": 43770 }, { "grad_norm": 0.3941568434238434, "learning_rate": 4.632640961936475e-05, "loss": 0.028, "step": 43780 }, { "grad_norm": 0.2959149181842804, "learning_rate": 4.6305797421614835e-05, "loss": 0.0246, "step": 43790 }, { "grad_norm": 0.31355559825897217, "learning_rate": 4.6285185855103105e-05, "loss": 0.0319, "step": 43800 }, { "grad_norm": 0.4858769178390503, "learning_rate": 4.626457492335151e-05, "loss": 0.0295, "step": 43810 }, { "grad_norm": 0.544800877571106, "learning_rate": 4.624396462988188e-05, "loss": 0.0243, "step": 43820 }, { "grad_norm": 0.3383084833621979, "learning_rate": 4.6223354978215985e-05, "loss": 0.0292, "step": 43830 }, { "grad_norm": 0.36651843786239624, "learning_rate": 4.620274597187544e-05, "loss": 0.0248, "step": 43840 }, { "grad_norm": 0.4373173415660858, "learning_rate": 4.6182137614381736e-05, "loss": 0.0301, "step": 43850 }, { "grad_norm": 0.4407889246940613, "learning_rate": 4.616152990925631e-05, "loss": 0.0295, "step": 43860 }, { "grad_norm": 0.2547391653060913, "learning_rate": 4.614092286002041e-05, "loss": 0.0245, "step": 43870 }, { "grad_norm": 0.358704149723053, "learning_rate": 4.612031647019528e-05, "loss": 0.0276, "step": 43880 }, { "grad_norm": 0.3815101087093353, "learning_rate": 4.609971074330193e-05, "loss": 0.0284, "step": 43890 }, { "grad_norm": 0.36779168248176575, "learning_rate": 4.607910568286136e-05, "loss": 0.0255, "step": 43900 }, { "grad_norm": 0.27985456585884094, "learning_rate": 4.605850129239438e-05, "loss": 0.0325, "step": 43910 }, { "grad_norm": 0.44417765736579895, "learning_rate": 4.603789757542174e-05, "loss": 0.0269, "step": 43920 }, { "grad_norm": 0.44479063153266907, "learning_rate": 4.601729453546404e-05, "loss": 0.0342, "step": 43930 }, { "grad_norm": 0.37252452969551086, "learning_rate": 4.599669217604177e-05, "loss": 0.0267, "step": 43940 }, { "grad_norm": 0.41478654742240906, "learning_rate": 4.597609050067532e-05, "loss": 0.0335, "step": 43950 }, { "grad_norm": 0.35836532711982727, "learning_rate": 4.595548951288495e-05, "loss": 0.0347, "step": 43960 }, { "grad_norm": 0.22113581001758575, "learning_rate": 4.593488921619081e-05, "loss": 0.036, "step": 43970 }, { "grad_norm": 0.29622405767440796, "learning_rate": 4.591428961411289e-05, "loss": 0.0341, "step": 43980 }, { "grad_norm": 0.3636987805366516, "learning_rate": 4.589369071017117e-05, "loss": 0.0375, "step": 43990 }, { "grad_norm": 0.27779102325439453, "learning_rate": 4.587309250788538e-05, "loss": 0.0294, "step": 44000 }, { "grad_norm": 0.436574250459671, "learning_rate": 4.585249501077522e-05, "loss": 0.0287, "step": 44010 }, { "grad_norm": 0.26839160919189453, "learning_rate": 4.583189822236021e-05, "loss": 0.0359, "step": 44020 }, { "grad_norm": 0.32636263966560364, "learning_rate": 4.5811302146159816e-05, "loss": 0.0278, "step": 44030 }, { "grad_norm": 0.5550341010093689, "learning_rate": 4.579070678569332e-05, "loss": 0.0263, "step": 44040 }, { "grad_norm": 0.665432333946228, "learning_rate": 4.5770112144479904e-05, "loss": 0.0347, "step": 44050 }, { "grad_norm": 0.4689013957977295, "learning_rate": 4.5749518226038645e-05, "loss": 0.0305, "step": 44060 }, { "grad_norm": 0.3404510021209717, "learning_rate": 4.572892503388845e-05, "loss": 0.034, "step": 44070 }, { "grad_norm": 0.2870306968688965, "learning_rate": 4.570833257154817e-05, "loss": 0.0296, "step": 44080 }, { "grad_norm": 0.3801044821739197, "learning_rate": 4.568774084253646e-05, "loss": 0.0375, "step": 44090 }, { "grad_norm": 0.3343062698841095, "learning_rate": 4.566714985037191e-05, "loss": 0.0356, "step": 44100 }, { "grad_norm": 0.2803919017314911, "learning_rate": 4.564655959857295e-05, "loss": 0.0314, "step": 44110 }, { "grad_norm": 0.366940975189209, "learning_rate": 4.5625970090657893e-05, "loss": 0.0287, "step": 44120 }, { "grad_norm": 0.4859829246997833, "learning_rate": 4.560538133014491e-05, "loss": 0.0285, "step": 44130 }, { "grad_norm": 0.2953607141971588, "learning_rate": 4.5584793320552055e-05, "loss": 0.026, "step": 44140 }, { "grad_norm": 0.307709276676178, "learning_rate": 4.556420606539728e-05, "loss": 0.0303, "step": 44150 }, { "grad_norm": 0.36406463384628296, "learning_rate": 4.554361956819836e-05, "loss": 0.0302, "step": 44160 }, { "grad_norm": 1.0907838344573975, "learning_rate": 4.552303383247299e-05, "loss": 0.0295, "step": 44170 }, { "grad_norm": 0.28709110617637634, "learning_rate": 4.5502448861738664e-05, "loss": 0.0315, "step": 44180 }, { "grad_norm": 0.3886435329914093, "learning_rate": 4.548186465951284e-05, "loss": 0.0387, "step": 44190 }, { "grad_norm": 0.24583867192268372, "learning_rate": 4.546128122931277e-05, "loss": 0.0328, "step": 44200 }, { "grad_norm": 0.30203545093536377, "learning_rate": 4.544069857465561e-05, "loss": 0.0291, "step": 44210 }, { "grad_norm": 0.35192015767097473, "learning_rate": 4.542011669905837e-05, "loss": 0.0267, "step": 44220 }, { "grad_norm": 0.3826117515563965, "learning_rate": 4.53995356060379e-05, "loss": 0.0304, "step": 44230 }, { "grad_norm": 0.21996967494487762, "learning_rate": 4.5378955299110994e-05, "loss": 0.0279, "step": 44240 }, { "grad_norm": 0.390426903963089, "learning_rate": 4.5358375781794233e-05, "loss": 0.0337, "step": 44250 }, { "grad_norm": 0.31065046787261963, "learning_rate": 4.533779705760411e-05, "loss": 0.031, "step": 44260 }, { "grad_norm": 0.40645912289619446, "learning_rate": 4.5317219130056934e-05, "loss": 0.0305, "step": 44270 }, { "grad_norm": 0.33019962906837463, "learning_rate": 4.5296642002668964e-05, "loss": 0.0301, "step": 44280 }, { "grad_norm": 0.31131449341773987, "learning_rate": 4.5276065678956216e-05, "loss": 0.0315, "step": 44290 }, { "grad_norm": 0.2751663029193878, "learning_rate": 4.525549016243466e-05, "loss": 0.0341, "step": 44300 }, { "grad_norm": 0.4457300007343292, "learning_rate": 4.523491545662008e-05, "loss": 0.0303, "step": 44310 }, { "grad_norm": 0.3680853843688965, "learning_rate": 4.5214341565028104e-05, "loss": 0.0349, "step": 44320 }, { "grad_norm": 0.38750019669532776, "learning_rate": 4.519376849117428e-05, "loss": 0.028, "step": 44330 }, { "grad_norm": 0.3010810315608978, "learning_rate": 4.517319623857395e-05, "loss": 0.0341, "step": 44340 }, { "grad_norm": 0.3398100733757019, "learning_rate": 4.5152624810742394e-05, "loss": 0.0363, "step": 44350 }, { "grad_norm": 0.29048973321914673, "learning_rate": 4.5132054211194664e-05, "loss": 0.0262, "step": 44360 }, { "grad_norm": 0.30913734436035156, "learning_rate": 4.511148444344574e-05, "loss": 0.0252, "step": 44370 }, { "grad_norm": 0.36923450231552124, "learning_rate": 4.509091551101041e-05, "loss": 0.0359, "step": 44380 }, { "grad_norm": 0.4730159640312195, "learning_rate": 4.507034741740338e-05, "loss": 0.0261, "step": 44390 }, { "grad_norm": 0.30237406492233276, "learning_rate": 4.5049780166139145e-05, "loss": 0.033, "step": 44400 }, { "grad_norm": 0.3261999189853668, "learning_rate": 4.5029213760732075e-05, "loss": 0.0453, "step": 44410 }, { "grad_norm": 0.38519835472106934, "learning_rate": 4.5008648204696434e-05, "loss": 0.0318, "step": 44420 }, { "grad_norm": 0.3958258032798767, "learning_rate": 4.4988083501546284e-05, "loss": 0.03, "step": 44430 }, { "grad_norm": 0.4432253837585449, "learning_rate": 4.4967519654795606e-05, "loss": 0.0264, "step": 44440 }, { "grad_norm": 0.4792671799659729, "learning_rate": 4.494695666795816e-05, "loss": 0.0282, "step": 44450 }, { "grad_norm": 0.3866070806980133, "learning_rate": 4.4926394544547644e-05, "loss": 0.0317, "step": 44460 }, { "grad_norm": 0.33116450905799866, "learning_rate": 4.490583328807752e-05, "loss": 0.0295, "step": 44470 }, { "grad_norm": 0.2414981722831726, "learning_rate": 4.488527290206117e-05, "loss": 0.0253, "step": 44480 }, { "grad_norm": 0.2353808730840683, "learning_rate": 4.48647133900118e-05, "loss": 0.0277, "step": 44490 }, { "grad_norm": 0.5244398713111877, "learning_rate": 4.484415475544244e-05, "loss": 0.0305, "step": 44500 }, { "grad_norm": 0.4866044819355011, "learning_rate": 4.4823597001866035e-05, "loss": 0.028, "step": 44510 }, { "grad_norm": 0.1755732297897339, "learning_rate": 4.480304013279532e-05, "loss": 0.024, "step": 44520 }, { "grad_norm": 0.2711450159549713, "learning_rate": 4.478248415174292e-05, "loss": 0.0271, "step": 44530 }, { "grad_norm": 0.23191577196121216, "learning_rate": 4.476192906222126e-05, "loss": 0.0277, "step": 44540 }, { "grad_norm": 0.6926229596138, "learning_rate": 4.474137486774268e-05, "loss": 0.036, "step": 44550 }, { "grad_norm": 0.4831656515598297, "learning_rate": 4.4720821571819296e-05, "loss": 0.0281, "step": 44560 }, { "grad_norm": 0.40847235918045044, "learning_rate": 4.470026917796314e-05, "loss": 0.0359, "step": 44570 }, { "grad_norm": 0.47482797503471375, "learning_rate": 4.4679717689686005e-05, "loss": 0.0434, "step": 44580 }, { "grad_norm": 0.30743497610092163, "learning_rate": 4.465916711049963e-05, "loss": 0.025, "step": 44590 }, { "grad_norm": 0.5444673299789429, "learning_rate": 4.4638617443915524e-05, "loss": 0.036, "step": 44600 }, { "grad_norm": 0.46932217478752136, "learning_rate": 4.4618068693445055e-05, "loss": 0.0293, "step": 44610 }, { "grad_norm": 0.30420684814453125, "learning_rate": 4.459752086259946e-05, "loss": 0.0296, "step": 44620 }, { "grad_norm": 0.5698896646499634, "learning_rate": 4.457697395488977e-05, "loss": 0.032, "step": 44630 }, { "grad_norm": 0.29585200548171997, "learning_rate": 4.455642797382693e-05, "loss": 0.0317, "step": 44640 }, { "grad_norm": 0.4124166667461395, "learning_rate": 4.4535882922921643e-05, "loss": 0.029, "step": 44650 }, { "grad_norm": 0.39324280619621277, "learning_rate": 4.451533880568455e-05, "loss": 0.0294, "step": 44660 }, { "grad_norm": 0.3245912790298462, "learning_rate": 4.449479562562603e-05, "loss": 0.0394, "step": 44670 }, { "grad_norm": 0.37973982095718384, "learning_rate": 4.447425338625639e-05, "loss": 0.038, "step": 44680 }, { "grad_norm": 0.3331994414329529, "learning_rate": 4.4453712091085705e-05, "loss": 0.0328, "step": 44690 }, { "grad_norm": 0.3664778172969818, "learning_rate": 4.443317174362392e-05, "loss": 0.0342, "step": 44700 }, { "grad_norm": 0.2191201150417328, "learning_rate": 4.441263234738085e-05, "loss": 0.0283, "step": 44710 }, { "grad_norm": 0.431781530380249, "learning_rate": 4.4392093905866086e-05, "loss": 0.0246, "step": 44720 }, { "grad_norm": 0.43834054470062256, "learning_rate": 4.437155642258911e-05, "loss": 0.0286, "step": 44730 }, { "grad_norm": 0.2621007263660431, "learning_rate": 4.4351019901059177e-05, "loss": 0.0326, "step": 44740 }, { "grad_norm": 0.44905757904052734, "learning_rate": 4.433048434478547e-05, "loss": 0.0395, "step": 44750 }, { "grad_norm": 0.4637875258922577, "learning_rate": 4.4309949757276906e-05, "loss": 0.027, "step": 44760 }, { "grad_norm": 0.3907768130302429, "learning_rate": 4.428941614204233e-05, "loss": 0.0381, "step": 44770 }, { "grad_norm": 0.5123491883277893, "learning_rate": 4.426888350259034e-05, "loss": 0.0299, "step": 44780 }, { "grad_norm": 0.3398131728172302, "learning_rate": 4.424835184242939e-05, "loss": 0.037, "step": 44790 }, { "grad_norm": 0.3397325873374939, "learning_rate": 4.422782116506784e-05, "loss": 0.0315, "step": 44800 }, { "grad_norm": 0.39639517664909363, "learning_rate": 4.4207291474013754e-05, "loss": 0.0297, "step": 44810 }, { "grad_norm": 0.3740972876548767, "learning_rate": 4.418676277277514e-05, "loss": 0.0447, "step": 44820 }, { "grad_norm": 0.3014534115791321, "learning_rate": 4.416623506485976e-05, "loss": 0.0371, "step": 44830 }, { "grad_norm": 0.4616142511367798, "learning_rate": 4.414570835377527e-05, "loss": 0.0272, "step": 44840 }, { "grad_norm": 0.2838352620601654, "learning_rate": 4.412518264302909e-05, "loss": 0.0293, "step": 44850 }, { "grad_norm": 0.42361828684806824, "learning_rate": 4.4104657936128535e-05, "loss": 0.0271, "step": 44860 }, { "grad_norm": 0.45434805750846863, "learning_rate": 4.408413423658071e-05, "loss": 0.0322, "step": 44870 }, { "grad_norm": 0.3587099313735962, "learning_rate": 4.406361154789253e-05, "loss": 0.0314, "step": 44880 }, { "grad_norm": 0.47670575976371765, "learning_rate": 4.4043089873570776e-05, "loss": 0.0318, "step": 44890 }, { "grad_norm": 0.38102179765701294, "learning_rate": 4.4022569217122025e-05, "loss": 0.027, "step": 44900 }, { "grad_norm": 0.4742155969142914, "learning_rate": 4.400204958205274e-05, "loss": 0.027, "step": 44910 }, { "grad_norm": 0.3500191271305084, "learning_rate": 4.3981530971869125e-05, "loss": 0.0422, "step": 44920 }, { "grad_norm": 0.251671701669693, "learning_rate": 4.396101339007727e-05, "loss": 0.0285, "step": 44930 }, { "grad_norm": 0.38403332233428955, "learning_rate": 4.394049684018304e-05, "loss": 0.0327, "step": 44940 }, { "grad_norm": 0.29276689887046814, "learning_rate": 4.3919981325692186e-05, "loss": 0.031, "step": 44950 }, { "grad_norm": 0.5578926801681519, "learning_rate": 4.389946685011024e-05, "loss": 0.0274, "step": 44960 }, { "grad_norm": 0.40608566999435425, "learning_rate": 4.387895341694255e-05, "loss": 0.0376, "step": 44970 }, { "grad_norm": 0.3463403880596161, "learning_rate": 4.3858441029694324e-05, "loss": 0.0395, "step": 44980 }, { "grad_norm": 0.5666441917419434, "learning_rate": 4.3837929691870527e-05, "loss": 0.0293, "step": 44990 }, { "grad_norm": 0.29715725779533386, "learning_rate": 4.381741940697604e-05, "loss": 0.0292, "step": 45000 }, { "grad_norm": 0.5815876126289368, "learning_rate": 4.379691017851547e-05, "loss": 0.0249, "step": 45010 }, { "grad_norm": 0.2901636064052582, "learning_rate": 4.3776402009993304e-05, "loss": 0.0316, "step": 45020 }, { "grad_norm": 0.717158854007721, "learning_rate": 4.3755894904913794e-05, "loss": 0.0257, "step": 45030 }, { "grad_norm": 0.28937724232673645, "learning_rate": 4.373538886678109e-05, "loss": 0.0247, "step": 45040 }, { "grad_norm": 0.219411700963974, "learning_rate": 4.371488389909909e-05, "loss": 0.026, "step": 45050 }, { "grad_norm": 0.3752163052558899, "learning_rate": 4.3694380005371515e-05, "loss": 0.0287, "step": 45060 }, { "grad_norm": 0.247737318277359, "learning_rate": 4.367387718910196e-05, "loss": 0.026, "step": 45070 }, { "grad_norm": 0.4095258116722107, "learning_rate": 4.3653375453793764e-05, "loss": 0.0231, "step": 45080 }, { "grad_norm": 0.25692352652549744, "learning_rate": 4.3632874802950136e-05, "loss": 0.0217, "step": 45090 }, { "grad_norm": 0.33323681354522705, "learning_rate": 4.3612375240074034e-05, "loss": 0.0295, "step": 45100 }, { "grad_norm": 0.4118829071521759, "learning_rate": 4.3591876768668325e-05, "loss": 0.0289, "step": 45110 }, { "grad_norm": 0.3839343190193176, "learning_rate": 4.3571379392235605e-05, "loss": 0.0266, "step": 45120 }, { "grad_norm": 0.333766907453537, "learning_rate": 4.3550883114278335e-05, "loss": 0.0267, "step": 45130 }, { "grad_norm": 0.2916804552078247, "learning_rate": 4.353038793829876e-05, "loss": 0.0301, "step": 45140 }, { "grad_norm": 0.3829896152019501, "learning_rate": 4.350989386779891e-05, "loss": 0.0235, "step": 45150 }, { "grad_norm": 1.38401460647583, "learning_rate": 4.3489400906280724e-05, "loss": 0.0388, "step": 45160 }, { "grad_norm": 0.4084412753582001, "learning_rate": 4.3468909057245845e-05, "loss": 0.0304, "step": 45170 }, { "grad_norm": 0.4525783658027649, "learning_rate": 4.3448418324195794e-05, "loss": 0.0393, "step": 45180 }, { "grad_norm": 0.36686405539512634, "learning_rate": 4.342792871063184e-05, "loss": 0.0339, "step": 45190 }, { "grad_norm": 0.5394913554191589, "learning_rate": 4.3407440220055145e-05, "loss": 0.029, "step": 45200 }, { "grad_norm": 0.38633039593696594, "learning_rate": 4.33869528559666e-05, "loss": 0.0264, "step": 45210 }, { "grad_norm": 0.2734375, "learning_rate": 4.336646662186696e-05, "loss": 0.024, "step": 45220 }, { "grad_norm": 0.48882779479026794, "learning_rate": 4.334598152125672e-05, "loss": 0.031, "step": 45230 }, { "grad_norm": 0.29443565011024475, "learning_rate": 4.3325497557636276e-05, "loss": 0.0272, "step": 45240 }, { "grad_norm": 0.38394659757614136, "learning_rate": 4.330501473450574e-05, "loss": 0.0325, "step": 45250 }, { "grad_norm": 0.4530005156993866, "learning_rate": 4.328453305536507e-05, "loss": 0.0237, "step": 45260 }, { "grad_norm": 0.32095015048980713, "learning_rate": 4.326405252371404e-05, "loss": 0.0261, "step": 45270 }, { "grad_norm": 0.2656947076320648, "learning_rate": 4.324357314305221e-05, "loss": 0.0225, "step": 45280 }, { "grad_norm": 0.22459563612937927, "learning_rate": 4.3223094916878945e-05, "loss": 0.0257, "step": 45290 }, { "grad_norm": 0.30733025074005127, "learning_rate": 4.320261784869338e-05, "loss": 0.0397, "step": 45300 }, { "grad_norm": 0.37934058904647827, "learning_rate": 4.318214194199455e-05, "loss": 0.0337, "step": 45310 }, { "grad_norm": 0.27651727199554443, "learning_rate": 4.316166720028118e-05, "loss": 0.0264, "step": 45320 }, { "grad_norm": 0.4251936376094818, "learning_rate": 4.3141193627051864e-05, "loss": 0.027, "step": 45330 }, { "grad_norm": 0.5723696351051331, "learning_rate": 4.312072122580496e-05, "loss": 0.0335, "step": 45340 }, { "grad_norm": 0.21855920553207397, "learning_rate": 4.3100250000038646e-05, "loss": 0.039, "step": 45350 }, { "grad_norm": 0.49749425053596497, "learning_rate": 4.307977995325091e-05, "loss": 0.0388, "step": 45360 }, { "grad_norm": 0.3227722644805908, "learning_rate": 4.30593110889395e-05, "loss": 0.0289, "step": 45370 }, { "grad_norm": 0.29171550273895264, "learning_rate": 4.3038843410602016e-05, "loss": 0.0311, "step": 45380 }, { "grad_norm": 0.23042945563793182, "learning_rate": 4.3018376921735774e-05, "loss": 0.0242, "step": 45390 }, { "grad_norm": 0.29239729046821594, "learning_rate": 4.299791162583799e-05, "loss": 0.028, "step": 45400 }, { "grad_norm": 0.3863494098186493, "learning_rate": 4.29774475264056e-05, "loss": 0.0327, "step": 45410 }, { "grad_norm": 0.3505631983280182, "learning_rate": 4.2956984626935365e-05, "loss": 0.0252, "step": 45420 }, { "grad_norm": 0.3084408640861511, "learning_rate": 4.293652293092383e-05, "loss": 0.027, "step": 45430 }, { "grad_norm": 0.5547205805778503, "learning_rate": 4.2916062441867324e-05, "loss": 0.0308, "step": 45440 }, { "grad_norm": 0.2733219861984253, "learning_rate": 4.289560316326201e-05, "loss": 0.035, "step": 45450 }, { "grad_norm": 0.6365563273429871, "learning_rate": 4.28751450986038e-05, "loss": 0.0356, "step": 45460 }, { "grad_norm": 0.35966014862060547, "learning_rate": 4.2854688251388444e-05, "loss": 0.0271, "step": 45470 }, { "grad_norm": 0.37967702746391296, "learning_rate": 4.2834232625111425e-05, "loss": 0.0328, "step": 45480 }, { "grad_norm": 0.4507845938205719, "learning_rate": 4.2813778223268086e-05, "loss": 0.026, "step": 45490 }, { "grad_norm": 0.5096461772918701, "learning_rate": 4.2793325049353477e-05, "loss": 0.0384, "step": 45500 }, { "grad_norm": 0.2952439486980438, "learning_rate": 4.2772873106862535e-05, "loss": 0.034, "step": 45510 }, { "grad_norm": 0.5313515067100525, "learning_rate": 4.275242239928993e-05, "loss": 0.0293, "step": 45520 }, { "grad_norm": 0.44070926308631897, "learning_rate": 4.273197293013009e-05, "loss": 0.0241, "step": 45530 }, { "grad_norm": 0.3246995508670807, "learning_rate": 4.271152470287731e-05, "loss": 0.0209, "step": 45540 }, { "grad_norm": 0.3710494339466095, "learning_rate": 4.2691077721025594e-05, "loss": 0.023, "step": 45550 }, { "grad_norm": 0.296095073223114, "learning_rate": 4.267063198806883e-05, "loss": 0.0263, "step": 45560 }, { "grad_norm": 0.29983288049697876, "learning_rate": 4.2650187507500574e-05, "loss": 0.0222, "step": 45570 }, { "grad_norm": 0.41951432824134827, "learning_rate": 4.2629744282814275e-05, "loss": 0.03, "step": 45580 }, { "grad_norm": 0.5740137100219727, "learning_rate": 4.2609302317503074e-05, "loss": 0.0231, "step": 45590 }, { "grad_norm": 0.536246657371521, "learning_rate": 4.258886161505999e-05, "loss": 0.0277, "step": 45600 }, { "grad_norm": 0.3691101372241974, "learning_rate": 4.2568422178977775e-05, "loss": 0.0234, "step": 45610 }, { "grad_norm": 0.3742140531539917, "learning_rate": 4.254798401274894e-05, "loss": 0.0337, "step": 45620 }, { "grad_norm": 0.38153529167175293, "learning_rate": 4.252754711986583e-05, "loss": 0.0417, "step": 45630 }, { "grad_norm": 0.35612165927886963, "learning_rate": 4.250711150382052e-05, "loss": 0.0303, "step": 45640 }, { "grad_norm": 0.42189836502075195, "learning_rate": 4.248667716810495e-05, "loss": 0.0337, "step": 45650 }, { "grad_norm": 0.4420050382614136, "learning_rate": 4.246624411621074e-05, "loss": 0.03, "step": 45660 }, { "grad_norm": 0.28994259238243103, "learning_rate": 4.244581235162938e-05, "loss": 0.0251, "step": 45670 }, { "grad_norm": 0.4401211142539978, "learning_rate": 4.2425381877852075e-05, "loss": 0.0343, "step": 45680 }, { "grad_norm": 0.37431010603904724, "learning_rate": 4.2404952698369856e-05, "loss": 0.0244, "step": 45690 }, { "grad_norm": 0.27068960666656494, "learning_rate": 4.238452481667349e-05, "loss": 0.0268, "step": 45700 }, { "grad_norm": 0.4906728267669678, "learning_rate": 4.2364098236253526e-05, "loss": 0.0265, "step": 45710 }, { "grad_norm": 0.32211706042289734, "learning_rate": 4.2343672960600356e-05, "loss": 0.0275, "step": 45720 }, { "grad_norm": 0.4094953238964081, "learning_rate": 4.232324899320406e-05, "loss": 0.0302, "step": 45730 }, { "grad_norm": 0.2514864504337311, "learning_rate": 4.230282633755457e-05, "loss": 0.0278, "step": 45740 }, { "grad_norm": 0.3963785171508789, "learning_rate": 4.2282404997141515e-05, "loss": 0.0322, "step": 45750 }, { "grad_norm": 0.2297336757183075, "learning_rate": 4.2261984975454397e-05, "loss": 0.0336, "step": 45760 }, { "grad_norm": 0.4620351195335388, "learning_rate": 4.224156627598239e-05, "loss": 0.024, "step": 45770 }, { "grad_norm": 0.2612062990665436, "learning_rate": 4.222114890221453e-05, "loss": 0.0252, "step": 45780 }, { "grad_norm": 0.3487444221973419, "learning_rate": 4.2200732857639546e-05, "loss": 0.0315, "step": 45790 }, { "grad_norm": 0.2851693630218506, "learning_rate": 4.2180318145746035e-05, "loss": 0.0297, "step": 45800 }, { "grad_norm": 0.572207510471344, "learning_rate": 4.215990477002227e-05, "loss": 0.034, "step": 45810 }, { "grad_norm": 0.2876013517379761, "learning_rate": 4.2139492733956356e-05, "loss": 0.0255, "step": 45820 }, { "grad_norm": 0.39098432660102844, "learning_rate": 4.211908204103615e-05, "loss": 0.0381, "step": 45830 }, { "grad_norm": 0.42042940855026245, "learning_rate": 4.2098672694749265e-05, "loss": 0.0344, "step": 45840 }, { "grad_norm": 0.33195698261260986, "learning_rate": 4.2078264698583133e-05, "loss": 0.033, "step": 45850 }, { "grad_norm": 0.2774858772754669, "learning_rate": 4.205785805602488e-05, "loss": 0.0272, "step": 45860 }, { "grad_norm": 0.45802024006843567, "learning_rate": 4.203745277056149e-05, "loss": 0.0314, "step": 45870 }, { "grad_norm": 0.4160047769546509, "learning_rate": 4.201704884567964e-05, "loss": 0.0326, "step": 45880 }, { "grad_norm": 0.31519681215286255, "learning_rate": 4.1996646284865816e-05, "loss": 0.035, "step": 45890 }, { "grad_norm": 0.30360347032546997, "learning_rate": 4.197624509160625e-05, "loss": 0.024, "step": 45900 }, { "grad_norm": 0.2896144688129425, "learning_rate": 4.195584526938692e-05, "loss": 0.0213, "step": 45910 }, { "grad_norm": 0.228115051984787, "learning_rate": 4.193544682169365e-05, "loss": 0.0251, "step": 45920 }, { "grad_norm": 0.5256645679473877, "learning_rate": 4.1915049752011946e-05, "loss": 0.0283, "step": 45930 }, { "grad_norm": 0.3693898916244507, "learning_rate": 4.189465406382712e-05, "loss": 0.0222, "step": 45940 }, { "grad_norm": 0.271107941865921, "learning_rate": 4.187425976062422e-05, "loss": 0.0276, "step": 45950 }, { "grad_norm": 0.3907988369464874, "learning_rate": 4.18538668458881e-05, "loss": 0.029, "step": 45960 }, { "grad_norm": 0.251796156167984, "learning_rate": 4.183347532310333e-05, "loss": 0.0244, "step": 45970 }, { "grad_norm": 0.3304281532764435, "learning_rate": 4.181308519575429e-05, "loss": 0.0265, "step": 45980 }, { "grad_norm": 0.2886459231376648, "learning_rate": 4.179269646732507e-05, "loss": 0.0346, "step": 45990 }, { "grad_norm": 0.24280814826488495, "learning_rate": 4.177230914129954e-05, "loss": 0.0256, "step": 46000 }, { "grad_norm": 0.32522717118263245, "learning_rate": 4.175192322116136e-05, "loss": 0.025, "step": 46010 }, { "grad_norm": 0.3204248547554016, "learning_rate": 4.173153871039391e-05, "loss": 0.0255, "step": 46020 }, { "grad_norm": 0.4122067391872406, "learning_rate": 4.171115561248036e-05, "loss": 0.0304, "step": 46030 }, { "grad_norm": 0.5215545892715454, "learning_rate": 4.16907739309036e-05, "loss": 0.0306, "step": 46040 }, { "grad_norm": 0.28089994192123413, "learning_rate": 4.167039366914633e-05, "loss": 0.0248, "step": 46050 }, { "grad_norm": 0.3673288822174072, "learning_rate": 4.165001483069096e-05, "loss": 0.0272, "step": 46060 }, { "grad_norm": 0.30377984046936035, "learning_rate": 4.16296374190197e-05, "loss": 0.0405, "step": 46070 }, { "grad_norm": 0.29111993312835693, "learning_rate": 4.1609261437614464e-05, "loss": 0.0282, "step": 46080 }, { "grad_norm": 0.47687602043151855, "learning_rate": 4.158888688995696e-05, "loss": 0.0351, "step": 46090 }, { "grad_norm": 0.435210645198822, "learning_rate": 4.1568513779528645e-05, "loss": 0.026, "step": 46100 }, { "grad_norm": 0.3962881565093994, "learning_rate": 4.1548142109810704e-05, "loss": 0.0334, "step": 46110 }, { "grad_norm": 0.23333773016929626, "learning_rate": 4.152777188428414e-05, "loss": 0.027, "step": 46120 }, { "grad_norm": 0.395453542470932, "learning_rate": 4.1507403106429646e-05, "loss": 0.0291, "step": 46130 }, { "grad_norm": 0.3215789198875427, "learning_rate": 4.148703577972768e-05, "loss": 0.027, "step": 46140 }, { "grad_norm": 0.30423954129219055, "learning_rate": 4.146666990765846e-05, "loss": 0.0271, "step": 46150 }, { "grad_norm": 0.28471291065216064, "learning_rate": 4.1446305493701986e-05, "loss": 0.0273, "step": 46160 }, { "grad_norm": 0.2803080379962921, "learning_rate": 4.142594254133796e-05, "loss": 0.0241, "step": 46170 }, { "grad_norm": 0.3668181002140045, "learning_rate": 4.1405581054045835e-05, "loss": 0.0288, "step": 46180 }, { "grad_norm": 0.3189215362071991, "learning_rate": 4.1385221035304864e-05, "loss": 0.0307, "step": 46190 }, { "grad_norm": 0.4059112071990967, "learning_rate": 4.136486248859398e-05, "loss": 0.0271, "step": 46200 }, { "grad_norm": 0.27989649772644043, "learning_rate": 4.1344505417391955e-05, "loss": 0.0355, "step": 46210 }, { "grad_norm": 0.5443186163902283, "learning_rate": 4.132414982517721e-05, "loss": 0.0278, "step": 46220 }, { "grad_norm": 0.49802711606025696, "learning_rate": 4.130379571542798e-05, "loss": 0.0263, "step": 46230 }, { "grad_norm": 0.4946482479572296, "learning_rate": 4.12834430916222e-05, "loss": 0.0259, "step": 46240 }, { "grad_norm": 0.4426209628582001, "learning_rate": 4.126309195723763e-05, "loss": 0.0248, "step": 46250 }, { "grad_norm": 0.3451743423938751, "learning_rate": 4.124274231575168e-05, "loss": 0.0336, "step": 46260 }, { "grad_norm": 0.20933087170124054, "learning_rate": 4.122239417064154e-05, "loss": 0.024, "step": 46270 }, { "grad_norm": 0.362160861492157, "learning_rate": 4.1202047525384184e-05, "loss": 0.0256, "step": 46280 }, { "grad_norm": 0.3270241618156433, "learning_rate": 4.118170238345627e-05, "loss": 0.0266, "step": 46290 }, { "grad_norm": 0.33950844407081604, "learning_rate": 4.1161358748334256e-05, "loss": 0.032, "step": 46300 }, { "grad_norm": 0.3967711925506592, "learning_rate": 4.1141016623494266e-05, "loss": 0.0272, "step": 46310 }, { "grad_norm": 0.3389616012573242, "learning_rate": 4.112067601241227e-05, "loss": 0.0237, "step": 46320 }, { "grad_norm": 0.25026100873947144, "learning_rate": 4.110033691856387e-05, "loss": 0.0242, "step": 46330 }, { "grad_norm": 0.9240832924842834, "learning_rate": 4.107999934542451e-05, "loss": 0.0383, "step": 46340 }, { "grad_norm": 0.44692325592041016, "learning_rate": 4.105966329646928e-05, "loss": 0.0355, "step": 46350 }, { "grad_norm": 0.21149064600467682, "learning_rate": 4.103932877517308e-05, "loss": 0.0285, "step": 46360 }, { "grad_norm": 0.4871106743812561, "learning_rate": 4.101899578501052e-05, "loss": 0.0238, "step": 46370 }, { "grad_norm": 0.25401192903518677, "learning_rate": 4.099866432945595e-05, "loss": 0.038, "step": 46380 }, { "grad_norm": 0.4386689066886902, "learning_rate": 4.097833441198346e-05, "loss": 0.0362, "step": 46390 }, { "grad_norm": 0.36244383454322815, "learning_rate": 4.095800603606685e-05, "loss": 0.0277, "step": 46400 }, { "grad_norm": 0.34089913964271545, "learning_rate": 4.093767920517975e-05, "loss": 0.0319, "step": 46410 }, { "grad_norm": 0.3261166214942932, "learning_rate": 4.091735392279539e-05, "loss": 0.0318, "step": 46420 }, { "grad_norm": 0.3836362361907959, "learning_rate": 4.089703019238685e-05, "loss": 0.0283, "step": 46430 }, { "grad_norm": 0.3718344271183014, "learning_rate": 4.0876708017426866e-05, "loss": 0.027, "step": 46440 }, { "grad_norm": 0.6085076928138733, "learning_rate": 4.085638740138798e-05, "loss": 0.0329, "step": 46450 }, { "grad_norm": 0.38026970624923706, "learning_rate": 4.08360683477424e-05, "loss": 0.0273, "step": 46460 }, { "grad_norm": 0.3840717375278473, "learning_rate": 4.0815750859962085e-05, "loss": 0.0313, "step": 46470 }, { "grad_norm": 0.3095989227294922, "learning_rate": 4.079543494151879e-05, "loss": 0.0334, "step": 46480 }, { "grad_norm": 0.41756677627563477, "learning_rate": 4.07751205958839e-05, "loss": 0.0307, "step": 46490 }, { "grad_norm": 0.2574009299278259, "learning_rate": 4.0754807826528615e-05, "loss": 0.0282, "step": 46500 }, { "grad_norm": 0.34180405735969543, "learning_rate": 4.073449663692379e-05, "loss": 0.0306, "step": 46510 }, { "grad_norm": 0.41595280170440674, "learning_rate": 4.07141870305401e-05, "loss": 0.0236, "step": 46520 }, { "grad_norm": 0.4557713270187378, "learning_rate": 4.0693879010847866e-05, "loss": 0.0488, "step": 46530 }, { "grad_norm": 0.42907601594924927, "learning_rate": 4.06735725813172e-05, "loss": 0.0299, "step": 46540 }, { "grad_norm": 0.2625732421875, "learning_rate": 4.0653267745417903e-05, "loss": 0.0245, "step": 46550 }, { "grad_norm": 0.2854016125202179, "learning_rate": 4.063296450661949e-05, "loss": 0.0196, "step": 46560 }, { "grad_norm": 0.25760290026664734, "learning_rate": 4.061266286839128e-05, "loss": 0.0288, "step": 46570 }, { "grad_norm": 0.23878134787082672, "learning_rate": 4.0592362834202225e-05, "loss": 0.0236, "step": 46580 }, { "grad_norm": 0.5189093947410583, "learning_rate": 4.057206440752107e-05, "loss": 0.0222, "step": 46590 }, { "grad_norm": 0.45703741908073425, "learning_rate": 4.0551767591816245e-05, "loss": 0.0297, "step": 46600 }, { "grad_norm": 0.5214962959289551, "learning_rate": 4.0531472390555935e-05, "loss": 0.0311, "step": 46610 }, { "grad_norm": 0.33927789330482483, "learning_rate": 4.051117880720802e-05, "loss": 0.026, "step": 46620 }, { "grad_norm": 0.2689524292945862, "learning_rate": 4.049088684524015e-05, "loss": 0.0214, "step": 46630 }, { "grad_norm": 0.348456472158432, "learning_rate": 4.0470596508119636e-05, "loss": 0.0227, "step": 46640 }, { "grad_norm": 0.5776980519294739, "learning_rate": 4.0450307799313524e-05, "loss": 0.0298, "step": 46650 }, { "grad_norm": 0.3413599133491516, "learning_rate": 4.0430020722288656e-05, "loss": 0.029, "step": 46660 }, { "grad_norm": 0.3228619396686554, "learning_rate": 4.040973528051148e-05, "loss": 0.0284, "step": 46670 }, { "grad_norm": 0.4637688100337982, "learning_rate": 4.038945147744827e-05, "loss": 0.0317, "step": 46680 }, { "grad_norm": 0.27457407116889954, "learning_rate": 4.0369169316564945e-05, "loss": 0.0257, "step": 46690 }, { "grad_norm": 0.2516029477119446, "learning_rate": 4.034888880132718e-05, "loss": 0.0245, "step": 46700 }, { "grad_norm": 0.48402833938598633, "learning_rate": 4.032860993520035e-05, "loss": 0.0302, "step": 46710 }, { "grad_norm": 0.2669902443885803, "learning_rate": 4.030833272164959e-05, "loss": 0.0271, "step": 46720 }, { "grad_norm": 0.35898175835609436, "learning_rate": 4.0288057164139705e-05, "loss": 0.0259, "step": 46730 }, { "grad_norm": 0.3851933181285858, "learning_rate": 4.0267783266135205e-05, "loss": 0.0221, "step": 46740 }, { "grad_norm": 0.2505688965320587, "learning_rate": 4.024751103110039e-05, "loss": 0.0286, "step": 46750 }, { "grad_norm": 0.6914085149765015, "learning_rate": 4.0227240462499176e-05, "loss": 0.0303, "step": 46760 }, { "grad_norm": 0.30479535460472107, "learning_rate": 4.020697156379531e-05, "loss": 0.0315, "step": 46770 }, { "grad_norm": 0.3835068345069885, "learning_rate": 4.018670433845215e-05, "loss": 0.0311, "step": 46780 }, { "grad_norm": 0.2660562992095947, "learning_rate": 4.016643878993284e-05, "loss": 0.0223, "step": 46790 }, { "grad_norm": 0.3469025492668152, "learning_rate": 4.014617492170017e-05, "loss": 0.0298, "step": 46800 }, { "grad_norm": 0.2786354124546051, "learning_rate": 4.0125912737216726e-05, "loss": 0.0356, "step": 46810 }, { "grad_norm": 0.242559552192688, "learning_rate": 4.0105652239944735e-05, "loss": 0.0315, "step": 46820 }, { "grad_norm": 0.3408702313899994, "learning_rate": 4.0085393433346144e-05, "loss": 0.0294, "step": 46830 }, { "grad_norm": 0.3601163923740387, "learning_rate": 4.006513632088268e-05, "loss": 0.0228, "step": 46840 }, { "grad_norm": 0.2922094166278839, "learning_rate": 4.004488090601567e-05, "loss": 0.0233, "step": 46850 }, { "grad_norm": 0.3486013114452362, "learning_rate": 4.002462719220626e-05, "loss": 0.0235, "step": 46860 }, { "grad_norm": 0.3489064574241638, "learning_rate": 4.000437518291522e-05, "loss": 0.023, "step": 46870 }, { "grad_norm": 0.3948470652103424, "learning_rate": 3.9984124881603094e-05, "loss": 0.0294, "step": 46880 }, { "grad_norm": 0.2582068145275116, "learning_rate": 3.9963876291730086e-05, "loss": 0.0299, "step": 46890 }, { "grad_norm": 0.5309450626373291, "learning_rate": 3.994362941675614e-05, "loss": 0.0258, "step": 46900 }, { "grad_norm": 0.2434004545211792, "learning_rate": 3.992338426014088e-05, "loss": 0.029, "step": 46910 }, { "grad_norm": 0.5189502239227295, "learning_rate": 3.9903140825343636e-05, "loss": 0.0349, "step": 46920 }, { "grad_norm": 0.3908790349960327, "learning_rate": 3.98828991158235e-05, "loss": 0.0246, "step": 46930 }, { "grad_norm": 0.3023780286312103, "learning_rate": 3.9862659135039185e-05, "loss": 0.0316, "step": 46940 }, { "grad_norm": 0.2920268177986145, "learning_rate": 3.984242088644918e-05, "loss": 0.0254, "step": 46950 }, { "grad_norm": 0.2597799003124237, "learning_rate": 3.9822184373511615e-05, "loss": 0.023, "step": 46960 }, { "grad_norm": 0.4014386534690857, "learning_rate": 3.98019495996844e-05, "loss": 0.0288, "step": 46970 }, { "grad_norm": 0.4097512662410736, "learning_rate": 3.978171656842507e-05, "loss": 0.0297, "step": 46980 }, { "grad_norm": 0.5199683904647827, "learning_rate": 3.976148528319091e-05, "loss": 0.0292, "step": 46990 }, { "grad_norm": 0.32131293416023254, "learning_rate": 3.974125574743888e-05, "loss": 0.0264, "step": 47000 }, { "grad_norm": 0.35083961486816406, "learning_rate": 3.9721027964625686e-05, "loss": 0.0308, "step": 47010 }, { "grad_norm": 0.3335585296154022, "learning_rate": 3.9700801938207676e-05, "loss": 0.0239, "step": 47020 }, { "grad_norm": 0.29825615882873535, "learning_rate": 3.9680577671640916e-05, "loss": 0.0198, "step": 47030 }, { "grad_norm": 0.3399379849433899, "learning_rate": 3.966035516838121e-05, "loss": 0.032, "step": 47040 }, { "grad_norm": 0.2933800220489502, "learning_rate": 3.9640134431884014e-05, "loss": 0.0221, "step": 47050 }, { "grad_norm": 0.6344345211982727, "learning_rate": 3.961991546560451e-05, "loss": 0.0236, "step": 47060 }, { "grad_norm": 0.2896655797958374, "learning_rate": 3.959969827299753e-05, "loss": 0.0263, "step": 47070 }, { "grad_norm": 0.3920303285121918, "learning_rate": 3.9579482857517684e-05, "loss": 0.0241, "step": 47080 }, { "grad_norm": 0.3691975474357605, "learning_rate": 3.955926922261921e-05, "loss": 0.0205, "step": 47090 }, { "grad_norm": 0.32431885600090027, "learning_rate": 3.9539057371756084e-05, "loss": 0.0249, "step": 47100 }, { "grad_norm": 0.25970348715782166, "learning_rate": 3.951884730838195e-05, "loss": 0.0323, "step": 47110 }, { "grad_norm": 0.3559805154800415, "learning_rate": 3.949863903595012e-05, "loss": 0.024, "step": 47120 }, { "grad_norm": 0.286592036485672, "learning_rate": 3.947843255791369e-05, "loss": 0.0268, "step": 47130 }, { "grad_norm": 0.3328596353530884, "learning_rate": 3.9458227877725364e-05, "loss": 0.0275, "step": 47140 }, { "grad_norm": 0.3622424304485321, "learning_rate": 3.943802499883758e-05, "loss": 0.0257, "step": 47150 }, { "grad_norm": 0.4017845392227173, "learning_rate": 3.9417823924702437e-05, "loss": 0.0293, "step": 47160 }, { "grad_norm": 0.23468483984470367, "learning_rate": 3.939762465877178e-05, "loss": 0.0269, "step": 47170 }, { "grad_norm": 0.3573455214500427, "learning_rate": 3.937742720449708e-05, "loss": 0.0289, "step": 47180 }, { "grad_norm": 0.29611238837242126, "learning_rate": 3.9357231565329563e-05, "loss": 0.0266, "step": 47190 }, { "grad_norm": 0.34352123737335205, "learning_rate": 3.933703774472008e-05, "loss": 0.0226, "step": 47200 }, { "grad_norm": 0.4211899936199188, "learning_rate": 3.93168457461192e-05, "loss": 0.029, "step": 47210 }, { "grad_norm": 0.39668694138526917, "learning_rate": 3.9296655572977216e-05, "loss": 0.0309, "step": 47220 }, { "grad_norm": 0.44479915499687195, "learning_rate": 3.927646722874404e-05, "loss": 0.0295, "step": 47230 }, { "grad_norm": 0.28193917870521545, "learning_rate": 3.925628071686934e-05, "loss": 0.0261, "step": 47240 }, { "grad_norm": 0.4467827379703522, "learning_rate": 3.9236096040802415e-05, "loss": 0.0267, "step": 47250 }, { "grad_norm": 0.5167788863182068, "learning_rate": 3.9215913203992294e-05, "loss": 0.0391, "step": 47260 }, { "grad_norm": 0.4889582395553589, "learning_rate": 3.9195732209887645e-05, "loss": 0.0261, "step": 47270 }, { "grad_norm": 0.31086602807044983, "learning_rate": 3.9175553061936875e-05, "loss": 0.0283, "step": 47280 }, { "grad_norm": 0.3312920331954956, "learning_rate": 3.9155375763588045e-05, "loss": 0.034, "step": 47290 }, { "grad_norm": 0.3609091341495514, "learning_rate": 3.913520031828889e-05, "loss": 0.0279, "step": 47300 }, { "grad_norm": 0.5697543621063232, "learning_rate": 3.911502672948685e-05, "loss": 0.0258, "step": 47310 }, { "grad_norm": 0.3637995421886444, "learning_rate": 3.9094855000629014e-05, "loss": 0.0283, "step": 47320 }, { "grad_norm": 0.34535494446754456, "learning_rate": 3.907468513516223e-05, "loss": 0.0335, "step": 47330 }, { "grad_norm": 0.4909859001636505, "learning_rate": 3.905451713653294e-05, "loss": 0.0267, "step": 47340 }, { "grad_norm": 0.3611668348312378, "learning_rate": 3.903435100818731e-05, "loss": 0.0234, "step": 47350 }, { "grad_norm": 0.44373780488967896, "learning_rate": 3.901418675357117e-05, "loss": 0.029, "step": 47360 }, { "grad_norm": 0.2729963958263397, "learning_rate": 3.8994024376130075e-05, "loss": 0.0324, "step": 47370 }, { "grad_norm": 0.3913370370864868, "learning_rate": 3.8973863879309194e-05, "loss": 0.031, "step": 47380 }, { "grad_norm": 0.2391713410615921, "learning_rate": 3.8953705266553394e-05, "loss": 0.0212, "step": 47390 }, { "grad_norm": 0.7024469375610352, "learning_rate": 3.893354854130727e-05, "loss": 0.0302, "step": 47400 }, { "grad_norm": 0.36085519194602966, "learning_rate": 3.8913393707015006e-05, "loss": 0.0299, "step": 47410 }, { "grad_norm": 0.3672705292701721, "learning_rate": 3.889324076712056e-05, "loss": 0.0305, "step": 47420 }, { "grad_norm": 0.24380022287368774, "learning_rate": 3.8873089725067476e-05, "loss": 0.0278, "step": 47430 }, { "grad_norm": 0.3898701071739197, "learning_rate": 3.885294058429905e-05, "loss": 0.0406, "step": 47440 }, { "grad_norm": 0.3355731666088104, "learning_rate": 3.8832793348258206e-05, "loss": 0.0262, "step": 47450 }, { "grad_norm": 0.2874908745288849, "learning_rate": 3.881264802038756e-05, "loss": 0.0338, "step": 47460 }, { "grad_norm": 0.4536120295524597, "learning_rate": 3.879250460412939e-05, "loss": 0.0316, "step": 47470 }, { "grad_norm": 0.31112515926361084, "learning_rate": 3.8772363102925644e-05, "loss": 0.0276, "step": 47480 }, { "grad_norm": 0.7480918765068054, "learning_rate": 3.875222352021798e-05, "loss": 0.0287, "step": 47490 }, { "grad_norm": 0.3235584795475006, "learning_rate": 3.8732085859447686e-05, "loss": 0.0278, "step": 47500 }, { "grad_norm": 0.25934135913848877, "learning_rate": 3.871195012405575e-05, "loss": 0.027, "step": 47510 }, { "grad_norm": 0.3211771547794342, "learning_rate": 3.869181631748278e-05, "loss": 0.0235, "step": 47520 }, { "grad_norm": 0.3034675717353821, "learning_rate": 3.867168444316915e-05, "loss": 0.0193, "step": 47530 }, { "grad_norm": 0.3756037950515747, "learning_rate": 3.865155450455481e-05, "loss": 0.0253, "step": 47540 }, { "grad_norm": 0.27652835845947266, "learning_rate": 3.8631426505079426e-05, "loss": 0.0336, "step": 47550 }, { "grad_norm": 0.5254383683204651, "learning_rate": 3.8611300448182304e-05, "loss": 0.0358, "step": 47560 }, { "grad_norm": 0.37290874123573303, "learning_rate": 3.859117633730248e-05, "loss": 0.0285, "step": 47570 }, { "grad_norm": 0.40781715512275696, "learning_rate": 3.857105417587858e-05, "loss": 0.0332, "step": 47580 }, { "grad_norm": 0.36573877930641174, "learning_rate": 3.855093396734894e-05, "loss": 0.0255, "step": 47590 }, { "grad_norm": 0.24446357786655426, "learning_rate": 3.8530815715151545e-05, "loss": 0.0413, "step": 47600 }, { "grad_norm": 0.3851059377193451, "learning_rate": 3.851069942272405e-05, "loss": 0.0283, "step": 47610 }, { "grad_norm": 0.3065102696418762, "learning_rate": 3.849058509350382e-05, "loss": 0.0299, "step": 47620 }, { "grad_norm": 0.4039205312728882, "learning_rate": 3.8470472730927783e-05, "loss": 0.0295, "step": 47630 }, { "grad_norm": 0.20435252785682678, "learning_rate": 3.845036233843264e-05, "loss": 0.021, "step": 47640 }, { "grad_norm": 0.28829750418663025, "learning_rate": 3.843025391945469e-05, "loss": 0.0261, "step": 47650 }, { "grad_norm": 0.31757161021232605, "learning_rate": 3.841014747742992e-05, "loss": 0.0248, "step": 47660 }, { "grad_norm": 0.46680665016174316, "learning_rate": 3.839004301579397e-05, "loss": 0.0248, "step": 47670 }, { "grad_norm": 0.20426419377326965, "learning_rate": 3.8369940537982097e-05, "loss": 0.0254, "step": 47680 }, { "grad_norm": 0.31669217348098755, "learning_rate": 3.834984004742933e-05, "loss": 0.0254, "step": 47690 }, { "grad_norm": 0.7091426849365234, "learning_rate": 3.832974154757026e-05, "loss": 0.0248, "step": 47700 }, { "grad_norm": 0.28932005167007446, "learning_rate": 3.830964504183919e-05, "loss": 0.0274, "step": 47710 }, { "grad_norm": 1.9123497009277344, "learning_rate": 3.828955053367003e-05, "loss": 0.0311, "step": 47720 }, { "grad_norm": 0.3397364616394043, "learning_rate": 3.826945802649642e-05, "loss": 0.0257, "step": 47730 }, { "grad_norm": 0.1938943713903427, "learning_rate": 3.824936752375159e-05, "loss": 0.042, "step": 47740 }, { "grad_norm": 0.2725509703159332, "learning_rate": 3.822927902886848e-05, "loss": 0.0213, "step": 47750 }, { "grad_norm": 0.31447651982307434, "learning_rate": 3.8209192545279653e-05, "loss": 0.0283, "step": 47760 }, { "grad_norm": 0.49088090658187866, "learning_rate": 3.8189108076417326e-05, "loss": 0.0297, "step": 47770 }, { "grad_norm": 0.32572686672210693, "learning_rate": 3.816902562571342e-05, "loss": 0.0344, "step": 47780 }, { "grad_norm": 0.4544518291950226, "learning_rate": 3.814894519659944e-05, "loss": 0.0256, "step": 47790 }, { "grad_norm": 0.2589573562145233, "learning_rate": 3.812886679250661e-05, "loss": 0.0219, "step": 47800 }, { "grad_norm": 0.17776429653167725, "learning_rate": 3.810879041686575e-05, "loss": 0.0216, "step": 47810 }, { "grad_norm": 0.7110867500305176, "learning_rate": 3.808871607310741e-05, "loss": 0.0329, "step": 47820 }, { "grad_norm": 0.39076635241508484, "learning_rate": 3.806864376466169e-05, "loss": 0.0246, "step": 47830 }, { "grad_norm": 0.319812148809433, "learning_rate": 3.804857349495845e-05, "loss": 0.0209, "step": 47840 }, { "grad_norm": 0.33491483330726624, "learning_rate": 3.802850526742713e-05, "loss": 0.0428, "step": 47850 }, { "grad_norm": 0.3436764180660248, "learning_rate": 3.800843908549683e-05, "loss": 0.0197, "step": 47860 }, { "grad_norm": 0.32537704706192017, "learning_rate": 3.7988374952596325e-05, "loss": 0.0229, "step": 47870 }, { "grad_norm": 0.2738727629184723, "learning_rate": 3.7968312872154e-05, "loss": 0.0228, "step": 47880 }, { "grad_norm": 1.0170177221298218, "learning_rate": 3.7948252847597965e-05, "loss": 0.028, "step": 47890 }, { "grad_norm": 0.28797706961631775, "learning_rate": 3.7928194882355885e-05, "loss": 0.0254, "step": 47900 }, { "grad_norm": 0.3258446156978607, "learning_rate": 3.790813897985515e-05, "loss": 0.0249, "step": 47910 }, { "grad_norm": 0.4229603111743927, "learning_rate": 3.7888085143522726e-05, "loss": 0.0291, "step": 47920 }, { "grad_norm": 0.3183020055294037, "learning_rate": 3.7868033376785314e-05, "loss": 0.0228, "step": 47930 }, { "grad_norm": 0.4271391034126282, "learning_rate": 3.784798368306919e-05, "loss": 0.0243, "step": 47940 }, { "grad_norm": 0.36716851592063904, "learning_rate": 3.782793606580029e-05, "loss": 0.0252, "step": 47950 }, { "grad_norm": 0.3925037086009979, "learning_rate": 3.7807890528404205e-05, "loss": 0.0281, "step": 47960 }, { "grad_norm": 0.2592800557613373, "learning_rate": 3.778784707430616e-05, "loss": 0.0204, "step": 47970 }, { "grad_norm": 0.2793722450733185, "learning_rate": 3.776780570693107e-05, "loss": 0.0261, "step": 47980 }, { "grad_norm": 0.23501291871070862, "learning_rate": 3.774776642970342e-05, "loss": 0.0346, "step": 47990 }, { "grad_norm": 0.36374586820602417, "learning_rate": 3.77277292460474e-05, "loss": 0.028, "step": 48000 }, { "grad_norm": 0.31383636593818665, "learning_rate": 3.770769415938678e-05, "loss": 0.026, "step": 48010 }, { "grad_norm": 0.24621310830116272, "learning_rate": 3.768766117314506e-05, "loss": 0.0332, "step": 48020 }, { "grad_norm": 0.781632125377655, "learning_rate": 3.76676302907453e-05, "loss": 0.0265, "step": 48030 }, { "grad_norm": 0.32140442728996277, "learning_rate": 3.764760151561021e-05, "loss": 0.03, "step": 48040 }, { "grad_norm": 0.3469124138355255, "learning_rate": 3.76275748511622e-05, "loss": 0.0262, "step": 48050 }, { "grad_norm": 0.3264645040035248, "learning_rate": 3.7607550300823255e-05, "loss": 0.0306, "step": 48060 }, { "grad_norm": 0.44460055232048035, "learning_rate": 3.7587527868015044e-05, "loss": 0.0419, "step": 48070 }, { "grad_norm": 0.380765825510025, "learning_rate": 3.756750755615881e-05, "loss": 0.0274, "step": 48080 }, { "grad_norm": 0.22444158792495728, "learning_rate": 3.754748936867553e-05, "loss": 0.0267, "step": 48090 }, { "grad_norm": 0.41594821214675903, "learning_rate": 3.752747330898573e-05, "loss": 0.0271, "step": 48100 }, { "grad_norm": 0.30344676971435547, "learning_rate": 3.750745938050962e-05, "loss": 0.0227, "step": 48110 }, { "grad_norm": 0.44502684473991394, "learning_rate": 3.7487447586667025e-05, "loss": 0.0259, "step": 48120 }, { "grad_norm": 0.6365350484848022, "learning_rate": 3.7467437930877426e-05, "loss": 0.0258, "step": 48130 }, { "grad_norm": 0.42253899574279785, "learning_rate": 3.744743041655992e-05, "loss": 0.0243, "step": 48140 }, { "grad_norm": 0.312550812959671, "learning_rate": 3.742742504713324e-05, "loss": 0.025, "step": 48150 }, { "grad_norm": 0.32185640931129456, "learning_rate": 3.740742182601576e-05, "loss": 0.0287, "step": 48160 }, { "grad_norm": 0.29933539032936096, "learning_rate": 3.7387420756625464e-05, "loss": 0.0306, "step": 48170 }, { "grad_norm": 0.33758220076560974, "learning_rate": 3.736742184238002e-05, "loss": 0.0309, "step": 48180 }, { "grad_norm": 0.3187558948993683, "learning_rate": 3.7347425086696684e-05, "loss": 0.0293, "step": 48190 }, { "grad_norm": 0.3891465663909912, "learning_rate": 3.732743049299235e-05, "loss": 0.0288, "step": 48200 }, { "grad_norm": 0.3571043312549591, "learning_rate": 3.730743806468354e-05, "loss": 0.0221, "step": 48210 }, { "grad_norm": 0.32769903540611267, "learning_rate": 3.7287447805186436e-05, "loss": 0.0202, "step": 48220 }, { "grad_norm": 0.3454093635082245, "learning_rate": 3.726745971791682e-05, "loss": 0.0325, "step": 48230 }, { "grad_norm": 0.40121832489967346, "learning_rate": 3.724747380629008e-05, "loss": 0.035, "step": 48240 }, { "grad_norm": 0.5277709364891052, "learning_rate": 3.72274900737213e-05, "loss": 0.0302, "step": 48250 }, { "grad_norm": 0.328146368265152, "learning_rate": 3.7207508523625123e-05, "loss": 0.0245, "step": 48260 }, { "grad_norm": 0.3921009302139282, "learning_rate": 3.718752915941588e-05, "loss": 0.0188, "step": 48270 }, { "grad_norm": 0.34332290291786194, "learning_rate": 3.7167551984507464e-05, "loss": 0.0239, "step": 48280 }, { "grad_norm": 0.3443804681301117, "learning_rate": 3.714757700231346e-05, "loss": 0.0293, "step": 48290 }, { "grad_norm": 0.19313181936740875, "learning_rate": 3.712760421624703e-05, "loss": 0.0261, "step": 48300 }, { "grad_norm": 0.3862783908843994, "learning_rate": 3.710763362972099e-05, "loss": 0.0276, "step": 48310 }, { "grad_norm": 0.466998815536499, "learning_rate": 3.708766524614774e-05, "loss": 0.0307, "step": 48320 }, { "grad_norm": 0.2784213721752167, "learning_rate": 3.7067699068939335e-05, "loss": 0.0279, "step": 48330 }, { "grad_norm": 0.3272513747215271, "learning_rate": 3.704773510150748e-05, "loss": 0.0245, "step": 48340 }, { "grad_norm": 0.3770833909511566, "learning_rate": 3.702777334726344e-05, "loss": 0.0274, "step": 48350 }, { "grad_norm": 0.23436778783798218, "learning_rate": 3.7007813809618164e-05, "loss": 0.026, "step": 48360 }, { "grad_norm": 0.3542858362197876, "learning_rate": 3.6987856491982145e-05, "loss": 0.031, "step": 48370 }, { "grad_norm": 0.25033140182495117, "learning_rate": 3.696790139776558e-05, "loss": 0.0273, "step": 48380 }, { "grad_norm": 0.21330483257770538, "learning_rate": 3.6947948530378235e-05, "loss": 0.026, "step": 48390 }, { "grad_norm": 0.2570861577987671, "learning_rate": 3.6927997893229516e-05, "loss": 0.023, "step": 48400 }, { "grad_norm": 0.22576278448104858, "learning_rate": 3.690804948972845e-05, "loss": 0.0277, "step": 48410 }, { "grad_norm": 0.2470293641090393, "learning_rate": 3.688810332328363e-05, "loss": 0.0257, "step": 48420 }, { "grad_norm": 0.37249431014060974, "learning_rate": 3.686815939730336e-05, "loss": 0.0302, "step": 48430 }, { "grad_norm": 0.2688099145889282, "learning_rate": 3.684821771519548e-05, "loss": 0.0202, "step": 48440 }, { "grad_norm": 0.3080603778362274, "learning_rate": 3.68282782803675e-05, "loss": 0.0285, "step": 48450 }, { "grad_norm": 0.24637803435325623, "learning_rate": 3.6808341096226504e-05, "loss": 0.0263, "step": 48460 }, { "grad_norm": 0.5407806038856506, "learning_rate": 3.678840616617924e-05, "loss": 0.0306, "step": 48470 }, { "grad_norm": 0.3974955677986145, "learning_rate": 3.6768473493632e-05, "loss": 0.0315, "step": 48480 }, { "grad_norm": 0.2625234127044678, "learning_rate": 3.6748543081990783e-05, "loss": 0.0236, "step": 48490 }, { "grad_norm": 0.41634199023246765, "learning_rate": 3.672861493466112e-05, "loss": 0.0265, "step": 48500 }, { "grad_norm": 0.39388659596443176, "learning_rate": 3.670868905504818e-05, "loss": 0.029, "step": 48510 }, { "grad_norm": 0.43440091609954834, "learning_rate": 3.6688765446556784e-05, "loss": 0.0276, "step": 48520 }, { "grad_norm": 0.411937415599823, "learning_rate": 3.6668844112591276e-05, "loss": 0.0232, "step": 48530 }, { "grad_norm": 0.45549142360687256, "learning_rate": 3.664892505655573e-05, "loss": 0.0253, "step": 48540 }, { "grad_norm": 0.29931336641311646, "learning_rate": 3.662900828185373e-05, "loss": 0.0256, "step": 48550 }, { "grad_norm": 0.2241119146347046, "learning_rate": 3.6609093791888516e-05, "loss": 0.0217, "step": 48560 }, { "grad_norm": 0.3898005187511444, "learning_rate": 3.658918159006292e-05, "loss": 0.028, "step": 48570 }, { "grad_norm": 0.4724913537502289, "learning_rate": 3.656927167977942e-05, "loss": 0.0314, "step": 48580 }, { "grad_norm": 0.3680015504360199, "learning_rate": 3.654936406444006e-05, "loss": 0.0256, "step": 48590 }, { "grad_norm": 0.3434707522392273, "learning_rate": 3.65294587474465e-05, "loss": 0.0204, "step": 48600 }, { "grad_norm": 0.5056021213531494, "learning_rate": 3.650955573220002e-05, "loss": 0.059, "step": 48610 }, { "grad_norm": 0.5191631317138672, "learning_rate": 3.648965502210149e-05, "loss": 0.0295, "step": 48620 }, { "grad_norm": 0.3323349952697754, "learning_rate": 3.646975662055142e-05, "loss": 0.0272, "step": 48630 }, { "grad_norm": 0.22947277128696442, "learning_rate": 3.644986053094987e-05, "loss": 0.0254, "step": 48640 }, { "grad_norm": 0.3910246789455414, "learning_rate": 3.642996675669659e-05, "loss": 0.0266, "step": 48650 }, { "grad_norm": 0.27747970819473267, "learning_rate": 3.641007530119083e-05, "loss": 0.0209, "step": 48660 }, { "grad_norm": 0.4329083263874054, "learning_rate": 3.639018616783153e-05, "loss": 0.0284, "step": 48670 }, { "grad_norm": 0.28730785846710205, "learning_rate": 3.637029936001719e-05, "loss": 0.0222, "step": 48680 }, { "grad_norm": 0.3481724262237549, "learning_rate": 3.6350414881145886e-05, "loss": 0.0309, "step": 48690 }, { "grad_norm": 0.3548574447631836, "learning_rate": 3.6330532734615386e-05, "loss": 0.0229, "step": 48700 }, { "grad_norm": 0.3660385012626648, "learning_rate": 3.6310652923822975e-05, "loss": 0.0269, "step": 48710 }, { "grad_norm": 0.26479196548461914, "learning_rate": 3.629077545216558e-05, "loss": 0.0217, "step": 48720 }, { "grad_norm": 0.29677751660346985, "learning_rate": 3.627090032303969e-05, "loss": 0.0353, "step": 48730 }, { "grad_norm": 0.3197411000728607, "learning_rate": 3.625102753984146e-05, "loss": 0.0342, "step": 48740 }, { "grad_norm": 0.2700355350971222, "learning_rate": 3.623115710596659e-05, "loss": 0.0213, "step": 48750 }, { "grad_norm": 0.4553071856498718, "learning_rate": 3.6211289024810395e-05, "loss": 0.0323, "step": 48760 }, { "grad_norm": 0.23904798924922943, "learning_rate": 3.619142329976777e-05, "loss": 0.019, "step": 48770 }, { "grad_norm": 0.35719308257102966, "learning_rate": 3.6171559934233247e-05, "loss": 0.0193, "step": 48780 }, { "grad_norm": 1.3386240005493164, "learning_rate": 3.615169893160093e-05, "loss": 0.0267, "step": 48790 }, { "grad_norm": 0.29588568210601807, "learning_rate": 3.61318402952645e-05, "loss": 0.0258, "step": 48800 }, { "grad_norm": 0.6075724959373474, "learning_rate": 3.6111984028617285e-05, "loss": 0.0326, "step": 48810 }, { "grad_norm": 0.8548670411109924, "learning_rate": 3.6092130135052134e-05, "loss": 0.0261, "step": 48820 }, { "grad_norm": 0.2540898323059082, "learning_rate": 3.6072278617961584e-05, "loss": 0.0292, "step": 48830 }, { "grad_norm": 0.3872198164463043, "learning_rate": 3.605242948073767e-05, "loss": 0.0305, "step": 48840 }, { "grad_norm": 0.4798900783061981, "learning_rate": 3.603258272677212e-05, "loss": 0.0248, "step": 48850 }, { "grad_norm": 0.423401415348053, "learning_rate": 3.601273835945616e-05, "loss": 0.0235, "step": 48860 }, { "grad_norm": 0.1925499588251114, "learning_rate": 3.5992896382180664e-05, "loss": 0.0262, "step": 48870 }, { "grad_norm": 0.3152678608894348, "learning_rate": 3.597305679833609e-05, "loss": 0.0289, "step": 48880 }, { "grad_norm": 0.5648614764213562, "learning_rate": 3.595321961131245e-05, "loss": 0.0254, "step": 48890 }, { "grad_norm": 0.4495076537132263, "learning_rate": 3.593338482449942e-05, "loss": 0.0273, "step": 48900 }, { "grad_norm": 0.5825023651123047, "learning_rate": 3.591355244128618e-05, "loss": 0.0372, "step": 48910 }, { "grad_norm": 0.28599128127098083, "learning_rate": 3.589372246506158e-05, "loss": 0.0299, "step": 48920 }, { "grad_norm": 0.30928924679756165, "learning_rate": 3.5873894899213984e-05, "loss": 0.0244, "step": 48930 }, { "grad_norm": 0.2770242989063263, "learning_rate": 3.5854069747131416e-05, "loss": 0.0231, "step": 48940 }, { "grad_norm": 0.2975689172744751, "learning_rate": 3.583424701220143e-05, "loss": 0.0279, "step": 48950 }, { "grad_norm": 0.30331236124038696, "learning_rate": 3.581442669781121e-05, "loss": 0.0259, "step": 48960 }, { "grad_norm": 0.31008487939834595, "learning_rate": 3.579460880734749e-05, "loss": 0.0241, "step": 48970 }, { "grad_norm": 0.23901306092739105, "learning_rate": 3.577479334419657e-05, "loss": 0.0307, "step": 48980 }, { "grad_norm": 0.30537673830986023, "learning_rate": 3.575498031174444e-05, "loss": 0.0267, "step": 48990 }, { "grad_norm": 0.3957051932811737, "learning_rate": 3.573516971337657e-05, "loss": 0.0336, "step": 49000 }, { "grad_norm": 0.433799147605896, "learning_rate": 3.5715361552478046e-05, "loss": 0.0298, "step": 49010 }, { "grad_norm": 0.34661537408828735, "learning_rate": 3.5695555832433536e-05, "loss": 0.0249, "step": 49020 }, { "grad_norm": 0.3203428089618683, "learning_rate": 3.5675752556627325e-05, "loss": 0.0277, "step": 49030 }, { "grad_norm": 0.367867112159729, "learning_rate": 3.565595172844322e-05, "loss": 0.0238, "step": 49040 }, { "grad_norm": 0.28614094853401184, "learning_rate": 3.5636153351264666e-05, "loss": 0.0295, "step": 49050 }, { "grad_norm": 0.5546087026596069, "learning_rate": 3.5616357428474655e-05, "loss": 0.0292, "step": 49060 }, { "grad_norm": 0.2627180218696594, "learning_rate": 3.559656396345575e-05, "loss": 0.0316, "step": 49070 }, { "grad_norm": 0.22931721806526184, "learning_rate": 3.5576772959590146e-05, "loss": 0.0275, "step": 49080 }, { "grad_norm": 0.2629697620868683, "learning_rate": 3.5556984420259545e-05, "loss": 0.0269, "step": 49090 }, { "grad_norm": 0.3852669298648834, "learning_rate": 3.5537198348845305e-05, "loss": 0.0243, "step": 49100 }, { "grad_norm": 0.2593989670276642, "learning_rate": 3.551741474872831e-05, "loss": 0.0266, "step": 49110 }, { "grad_norm": 0.25562313199043274, "learning_rate": 3.549763362328903e-05, "loss": 0.0291, "step": 49120 }, { "grad_norm": 0.2353864163160324, "learning_rate": 3.5477854975907515e-05, "loss": 0.0231, "step": 49130 }, { "grad_norm": 0.21132050454616547, "learning_rate": 3.5458078809963416e-05, "loss": 0.0239, "step": 49140 }, { "grad_norm": 0.3711107075214386, "learning_rate": 3.543830512883594e-05, "loss": 0.0216, "step": 49150 }, { "grad_norm": 0.32004785537719727, "learning_rate": 3.5418533935903824e-05, "loss": 0.0281, "step": 49160 }, { "grad_norm": 0.383210688829422, "learning_rate": 3.539876523454547e-05, "loss": 0.0225, "step": 49170 }, { "grad_norm": 0.2637437880039215, "learning_rate": 3.537899902813878e-05, "loss": 0.02, "step": 49180 }, { "grad_norm": 0.3627282977104187, "learning_rate": 3.5359235320061293e-05, "loss": 0.026, "step": 49190 }, { "grad_norm": 0.3210900127887726, "learning_rate": 3.533947411369003e-05, "loss": 0.0261, "step": 49200 }, { "grad_norm": 0.31992241740226746, "learning_rate": 3.53197154124017e-05, "loss": 0.0301, "step": 49210 }, { "grad_norm": 0.37128016352653503, "learning_rate": 3.52999592195725e-05, "loss": 0.0282, "step": 49220 }, { "grad_norm": 0.257862389087677, "learning_rate": 3.5280205538578224e-05, "loss": 0.0232, "step": 49230 }, { "grad_norm": 0.36426636576652527, "learning_rate": 3.5260454372794236e-05, "loss": 0.0268, "step": 49240 }, { "grad_norm": 0.4387573301792145, "learning_rate": 3.524070572559545e-05, "loss": 0.0295, "step": 49250 }, { "grad_norm": 0.41026127338409424, "learning_rate": 3.5220959600356395e-05, "loss": 0.0265, "step": 49260 }, { "grad_norm": 0.35683077573776245, "learning_rate": 3.5201216000451145e-05, "loss": 0.0227, "step": 49270 }, { "grad_norm": 0.2578093409538269, "learning_rate": 3.5181474929253335e-05, "loss": 0.0235, "step": 49280 }, { "grad_norm": 0.6190739274024963, "learning_rate": 3.516173639013615e-05, "loss": 0.0291, "step": 49290 }, { "grad_norm": 0.2906874418258667, "learning_rate": 3.5142000386472406e-05, "loss": 0.0237, "step": 49300 }, { "grad_norm": 0.3581008017063141, "learning_rate": 3.5122266921634427e-05, "loss": 0.0274, "step": 49310 }, { "grad_norm": 0.3151693642139435, "learning_rate": 3.510253599899413e-05, "loss": 0.0235, "step": 49320 }, { "grad_norm": 0.3759486973285675, "learning_rate": 3.5082807621922965e-05, "loss": 0.0246, "step": 49330 }, { "grad_norm": 0.3050794303417206, "learning_rate": 3.506308179379201e-05, "loss": 0.0301, "step": 49340 }, { "grad_norm": 0.3241746127605438, "learning_rate": 3.5043358517971844e-05, "loss": 0.0276, "step": 49350 }, { "grad_norm": 0.2389039695262909, "learning_rate": 3.502363779783264e-05, "loss": 0.0286, "step": 49360 }, { "grad_norm": 0.31590035557746887, "learning_rate": 3.500391963674415e-05, "loss": 0.022, "step": 49370 }, { "grad_norm": 0.34411489963531494, "learning_rate": 3.4984204038075615e-05, "loss": 0.0307, "step": 49380 }, { "grad_norm": 0.7083561420440674, "learning_rate": 3.496449100519595e-05, "loss": 0.0338, "step": 49390 }, { "grad_norm": 0.30331021547317505, "learning_rate": 3.494478054147354e-05, "loss": 0.0221, "step": 49400 }, { "grad_norm": 0.24330061674118042, "learning_rate": 3.4925072650276395e-05, "loss": 0.0208, "step": 49410 }, { "grad_norm": 0.31662166118621826, "learning_rate": 3.4905367334972016e-05, "loss": 0.0261, "step": 49420 }, { "grad_norm": 0.4462803900241852, "learning_rate": 3.488566459892752e-05, "loss": 0.0268, "step": 49430 }, { "grad_norm": 0.3454715609550476, "learning_rate": 3.4865964445509585e-05, "loss": 0.0257, "step": 49440 }, { "grad_norm": 0.36348626017570496, "learning_rate": 3.484626687808438e-05, "loss": 0.0271, "step": 49450 }, { "grad_norm": 0.4273124039173126, "learning_rate": 3.4826571900017735e-05, "loss": 0.0219, "step": 49460 }, { "grad_norm": 0.33739450573921204, "learning_rate": 3.480687951467495e-05, "loss": 0.0279, "step": 49470 }, { "grad_norm": 0.44902151823043823, "learning_rate": 3.4787189725420925e-05, "loss": 0.0295, "step": 49480 }, { "grad_norm": 0.30012625455856323, "learning_rate": 3.4767502535620086e-05, "loss": 0.024, "step": 49490 }, { "grad_norm": 0.3090725243091583, "learning_rate": 3.474781794863648e-05, "loss": 0.0332, "step": 49500 }, { "grad_norm": 0.2907199561595917, "learning_rate": 3.472813596783363e-05, "loss": 0.0305, "step": 49510 }, { "grad_norm": 0.47671452164649963, "learning_rate": 3.470845659657466e-05, "loss": 0.0294, "step": 49520 }, { "grad_norm": 0.3125942349433899, "learning_rate": 3.468877983822223e-05, "loss": 0.0238, "step": 49530 }, { "grad_norm": 0.2831939160823822, "learning_rate": 3.466910569613855e-05, "loss": 0.0191, "step": 49540 }, { "grad_norm": 0.3315834701061249, "learning_rate": 3.464943417368542e-05, "loss": 0.028, "step": 49550 }, { "grad_norm": 0.49113553762435913, "learning_rate": 3.462976527422415e-05, "loss": 0.0268, "step": 49560 }, { "grad_norm": 0.3386616110801697, "learning_rate": 3.461009900111562e-05, "loss": 0.0298, "step": 49570 }, { "grad_norm": 0.3270740211009979, "learning_rate": 3.459043535772023e-05, "loss": 0.0285, "step": 49580 }, { "grad_norm": 0.28190526366233826, "learning_rate": 3.4570774347398014e-05, "loss": 0.0274, "step": 49590 }, { "grad_norm": 0.3504902422428131, "learning_rate": 3.4551115973508454e-05, "loss": 0.034, "step": 49600 }, { "grad_norm": 0.28762245178222656, "learning_rate": 3.453146023941066e-05, "loss": 0.0339, "step": 49610 }, { "grad_norm": 0.2558461129665375, "learning_rate": 3.451180714846325e-05, "loss": 0.0339, "step": 49620 }, { "grad_norm": 0.3109598457813263, "learning_rate": 3.449215670402438e-05, "loss": 0.0242, "step": 49630 }, { "grad_norm": 0.44015225768089294, "learning_rate": 3.447250890945181e-05, "loss": 0.0257, "step": 49640 }, { "grad_norm": 0.24193337559700012, "learning_rate": 3.4452863768102754e-05, "loss": 0.0255, "step": 49650 }, { "grad_norm": 0.5564740300178528, "learning_rate": 3.443322128333409e-05, "loss": 0.03, "step": 49660 }, { "grad_norm": 0.4093388617038727, "learning_rate": 3.441358145850215e-05, "loss": 0.0267, "step": 49670 }, { "grad_norm": 0.3020649552345276, "learning_rate": 3.439394429696286e-05, "loss": 0.0225, "step": 49680 }, { "grad_norm": 0.27132758498191833, "learning_rate": 3.4374309802071644e-05, "loss": 0.0251, "step": 49690 }, { "grad_norm": 0.28606992959976196, "learning_rate": 3.435467797718353e-05, "loss": 0.0294, "step": 49700 }, { "grad_norm": 0.24797673523426056, "learning_rate": 3.433504882565306e-05, "loss": 0.0221, "step": 49710 }, { "grad_norm": 0.34005415439605713, "learning_rate": 3.43154223508343e-05, "loss": 0.0368, "step": 49720 }, { "grad_norm": 0.41124287247657776, "learning_rate": 3.429579855608089e-05, "loss": 0.0256, "step": 49730 }, { "grad_norm": 0.2497229278087616, "learning_rate": 3.427617744474597e-05, "loss": 0.0254, "step": 49740 }, { "grad_norm": 0.581526517868042, "learning_rate": 3.425655902018231e-05, "loss": 0.0312, "step": 49750 }, { "grad_norm": 0.35596963763237, "learning_rate": 3.423694328574211e-05, "loss": 0.0284, "step": 49760 }, { "grad_norm": 0.4397595524787903, "learning_rate": 3.42173302447772e-05, "loss": 0.0279, "step": 49770 }, { "grad_norm": 0.21425937116146088, "learning_rate": 3.419771990063886e-05, "loss": 0.0237, "step": 49780 }, { "grad_norm": 0.31105223298072815, "learning_rate": 3.417811225667803e-05, "loss": 0.0261, "step": 49790 }, { "grad_norm": 0.5525304079055786, "learning_rate": 3.415850731624508e-05, "loss": 0.0237, "step": 49800 }, { "grad_norm": 0.2759499251842499, "learning_rate": 3.4138905082689945e-05, "loss": 0.0344, "step": 49810 }, { "grad_norm": 0.4599136710166931, "learning_rate": 3.4119305559362145e-05, "loss": 0.0297, "step": 49820 }, { "grad_norm": 0.36746808886528015, "learning_rate": 3.4099708749610684e-05, "loss": 0.0281, "step": 49830 }, { "grad_norm": 0.3385954201221466, "learning_rate": 3.408011465678413e-05, "loss": 0.0289, "step": 49840 }, { "grad_norm": 0.2567513883113861, "learning_rate": 3.406052328423055e-05, "loss": 0.0295, "step": 49850 }, { "grad_norm": 0.33354440331459045, "learning_rate": 3.4040934635297615e-05, "loss": 0.0237, "step": 49860 }, { "grad_norm": 0.21331152319908142, "learning_rate": 3.4021348713332466e-05, "loss": 0.0239, "step": 49870 }, { "grad_norm": 0.3709389269351959, "learning_rate": 3.4001765521681807e-05, "loss": 0.0229, "step": 49880 }, { "grad_norm": 0.3347066342830658, "learning_rate": 3.398218506369188e-05, "loss": 0.0306, "step": 49890 }, { "grad_norm": 0.19907228648662567, "learning_rate": 3.3962607342708404e-05, "loss": 0.025, "step": 49900 }, { "grad_norm": 0.4613457918167114, "learning_rate": 3.394303236207673e-05, "loss": 0.0276, "step": 49910 }, { "grad_norm": 0.48255136609077454, "learning_rate": 3.392346012514166e-05, "loss": 0.0264, "step": 49920 }, { "grad_norm": 0.3205840587615967, "learning_rate": 3.390389063524757e-05, "loss": 0.0256, "step": 49930 }, { "grad_norm": 0.5951923131942749, "learning_rate": 3.3884323895738324e-05, "loss": 0.0237, "step": 49940 }, { "grad_norm": 0.32397276163101196, "learning_rate": 3.386475990995738e-05, "loss": 0.0313, "step": 49950 }, { "grad_norm": 0.3935748040676117, "learning_rate": 3.384519868124765e-05, "loss": 0.0262, "step": 49960 }, { "grad_norm": 0.38199496269226074, "learning_rate": 3.3825640212951645e-05, "loss": 0.0277, "step": 49970 }, { "grad_norm": 0.26885879039764404, "learning_rate": 3.380608450841134e-05, "loss": 0.0245, "step": 49980 }, { "grad_norm": 0.4621042311191559, "learning_rate": 3.3786531570968305e-05, "loss": 0.0278, "step": 49990 }, { "grad_norm": 0.2740171551704407, "learning_rate": 3.3766981403963584e-05, "loss": 0.0316, "step": 50000 }, { "grad_norm": 0.35581251978874207, "learning_rate": 3.374743401073775e-05, "loss": 0.031, "step": 50010 }, { "grad_norm": 0.36164483428001404, "learning_rate": 3.372788939463095e-05, "loss": 0.0264, "step": 50020 }, { "grad_norm": 0.49712300300598145, "learning_rate": 3.370834755898281e-05, "loss": 0.021, "step": 50030 }, { "grad_norm": 0.34622418880462646, "learning_rate": 3.3688808507132493e-05, "loss": 0.0243, "step": 50040 }, { "grad_norm": 0.19465981423854828, "learning_rate": 3.3669272242418685e-05, "loss": 0.026, "step": 50050 }, { "grad_norm": 0.32359105348587036, "learning_rate": 3.364973876817961e-05, "loss": 0.0243, "step": 50060 }, { "grad_norm": 0.2505585253238678, "learning_rate": 3.363020808775299e-05, "loss": 0.0228, "step": 50070 }, { "grad_norm": 0.23079724609851837, "learning_rate": 3.361068020447611e-05, "loss": 0.0264, "step": 50080 }, { "grad_norm": 0.27367275953292847, "learning_rate": 3.3591155121685724e-05, "loss": 0.0292, "step": 50090 }, { "grad_norm": 0.32904180884361267, "learning_rate": 3.357163284271814e-05, "loss": 0.0239, "step": 50100 }, { "grad_norm": 0.5446228981018066, "learning_rate": 3.355211337090919e-05, "loss": 0.0285, "step": 50110 }, { "grad_norm": 0.35481300950050354, "learning_rate": 3.353259670959421e-05, "loss": 0.0271, "step": 50120 }, { "grad_norm": 0.21763625741004944, "learning_rate": 3.351308286210808e-05, "loss": 0.0253, "step": 50130 }, { "grad_norm": 0.3212783932685852, "learning_rate": 3.3493571831785156e-05, "loss": 0.0199, "step": 50140 }, { "grad_norm": 0.26202496886253357, "learning_rate": 3.347406362195936e-05, "loss": 0.0186, "step": 50150 }, { "grad_norm": 0.29577672481536865, "learning_rate": 3.345455823596411e-05, "loss": 0.0266, "step": 50160 }, { "grad_norm": 0.9018177390098572, "learning_rate": 3.3435055677132346e-05, "loss": 0.0286, "step": 50170 }, { "grad_norm": 0.33236950635910034, "learning_rate": 3.3415555948796505e-05, "loss": 0.0255, "step": 50180 }, { "grad_norm": 0.28134429454803467, "learning_rate": 3.3396059054288556e-05, "loss": 0.0252, "step": 50190 }, { "grad_norm": 0.2430768758058548, "learning_rate": 3.3376564996940015e-05, "loss": 0.0303, "step": 50200 }, { "grad_norm": 0.5435144305229187, "learning_rate": 3.3357073780081836e-05, "loss": 0.026, "step": 50210 }, { "grad_norm": 0.38501185178756714, "learning_rate": 3.333758540704459e-05, "loss": 0.0287, "step": 50220 }, { "grad_norm": 0.27276450395584106, "learning_rate": 3.331809988115827e-05, "loss": 0.0251, "step": 50230 }, { "grad_norm": 0.3153790831565857, "learning_rate": 3.329861720575244e-05, "loss": 0.0279, "step": 50240 }, { "grad_norm": 0.3528486490249634, "learning_rate": 3.3279137384156126e-05, "loss": 0.0209, "step": 50250 }, { "grad_norm": 0.4187476933002472, "learning_rate": 3.3259660419697934e-05, "loss": 0.0271, "step": 50260 }, { "grad_norm": 0.6562569737434387, "learning_rate": 3.3240186315705926e-05, "loss": 0.0307, "step": 50270 }, { "grad_norm": 0.584367573261261, "learning_rate": 3.322071507550769e-05, "loss": 0.0359, "step": 50280 }, { "grad_norm": 0.6840082406997681, "learning_rate": 3.320124670243033e-05, "loss": 0.0265, "step": 50290 }, { "grad_norm": 0.5562723875045776, "learning_rate": 3.318178119980045e-05, "loss": 0.0271, "step": 50300 }, { "grad_norm": 0.4010748863220215, "learning_rate": 3.316231857094421e-05, "loss": 0.0262, "step": 50310 }, { "grad_norm": 0.44598495960235596, "learning_rate": 3.31428588191872e-05, "loss": 0.0285, "step": 50320 }, { "grad_norm": 0.2984946370124817, "learning_rate": 3.312340194785458e-05, "loss": 0.0271, "step": 50330 }, { "grad_norm": 0.43300527334213257, "learning_rate": 3.3103947960270975e-05, "loss": 0.0272, "step": 50340 }, { "grad_norm": 0.2538352608680725, "learning_rate": 3.308449685976058e-05, "loss": 0.0204, "step": 50350 }, { "grad_norm": 0.39431869983673096, "learning_rate": 3.3065048649647024e-05, "loss": 0.0256, "step": 50360 }, { "grad_norm": 0.35128992795944214, "learning_rate": 3.304560333325348e-05, "loss": 0.0264, "step": 50370 }, { "grad_norm": 0.334809273481369, "learning_rate": 3.3026160913902624e-05, "loss": 0.0255, "step": 50380 }, { "grad_norm": 0.43559691309928894, "learning_rate": 3.300672139491662e-05, "loss": 0.0305, "step": 50390 }, { "grad_norm": 0.35354337096214294, "learning_rate": 3.298728477961717e-05, "loss": 0.0257, "step": 50400 }, { "grad_norm": 0.22749213874340057, "learning_rate": 3.296785107132544e-05, "loss": 0.0275, "step": 50410 }, { "grad_norm": 0.3502025604248047, "learning_rate": 3.2948420273362156e-05, "loss": 0.0255, "step": 50420 }, { "grad_norm": 0.2847127318382263, "learning_rate": 3.292899238904747e-05, "loss": 0.0241, "step": 50430 }, { "grad_norm": 0.29145970940589905, "learning_rate": 3.29095674217011e-05, "loss": 0.023, "step": 50440 }, { "grad_norm": 0.4883892238140106, "learning_rate": 3.289014537464224e-05, "loss": 0.0303, "step": 50450 }, { "grad_norm": 0.24209527671337128, "learning_rate": 3.287072625118955e-05, "loss": 0.0213, "step": 50460 }, { "grad_norm": 0.23447036743164062, "learning_rate": 3.285131005466129e-05, "loss": 0.0225, "step": 50470 }, { "grad_norm": 0.34430912137031555, "learning_rate": 3.2831896788375105e-05, "loss": 0.0223, "step": 50480 }, { "grad_norm": 0.3153681457042694, "learning_rate": 3.281248645564822e-05, "loss": 0.0229, "step": 50490 }, { "grad_norm": 0.2991546392440796, "learning_rate": 3.2793079059797306e-05, "loss": 0.0257, "step": 50500 }, { "grad_norm": 0.45029929280281067, "learning_rate": 3.277367460413859e-05, "loss": 0.0247, "step": 50510 }, { "grad_norm": 0.6652340292930603, "learning_rate": 3.275427309198773e-05, "loss": 0.0237, "step": 50520 }, { "grad_norm": 0.3103904128074646, "learning_rate": 3.273487452665993e-05, "loss": 0.0233, "step": 50530 }, { "grad_norm": 0.26065006852149963, "learning_rate": 3.271547891146986e-05, "loss": 0.0259, "step": 50540 }, { "grad_norm": 0.23260708153247833, "learning_rate": 3.269608624973173e-05, "loss": 0.0204, "step": 50550 }, { "grad_norm": 0.3163350224494934, "learning_rate": 3.26766965447592e-05, "loss": 0.0224, "step": 50560 }, { "grad_norm": 0.27781370282173157, "learning_rate": 3.2657309799865424e-05, "loss": 0.026, "step": 50570 }, { "grad_norm": 0.43800535798072815, "learning_rate": 3.2637926018363084e-05, "loss": 0.0211, "step": 50580 }, { "grad_norm": 0.7757429480552673, "learning_rate": 3.261854520356432e-05, "loss": 0.0232, "step": 50590 }, { "grad_norm": 0.2735312581062317, "learning_rate": 3.2599167358780816e-05, "loss": 0.0263, "step": 50600 }, { "grad_norm": 0.287484347820282, "learning_rate": 3.2579792487323676e-05, "loss": 0.0223, "step": 50610 }, { "grad_norm": 0.40007734298706055, "learning_rate": 3.256042059250358e-05, "loss": 0.0281, "step": 50620 }, { "grad_norm": 0.4224660098552704, "learning_rate": 3.254105167763062e-05, "loss": 0.0308, "step": 50630 }, { "grad_norm": 0.2996913194656372, "learning_rate": 3.252168574601443e-05, "loss": 0.0277, "step": 50640 }, { "grad_norm": 0.45793765783309937, "learning_rate": 3.2502322800964115e-05, "loss": 0.0264, "step": 50650 }, { "grad_norm": 0.2931588888168335, "learning_rate": 3.248296284578825e-05, "loss": 0.0249, "step": 50660 }, { "grad_norm": 0.3913111090660095, "learning_rate": 3.246360588379497e-05, "loss": 0.0303, "step": 50670 }, { "grad_norm": 0.32507777214050293, "learning_rate": 3.24442519182918e-05, "loss": 0.0351, "step": 50680 }, { "grad_norm": 0.3240801692008972, "learning_rate": 3.2424900952585845e-05, "loss": 0.0251, "step": 50690 }, { "grad_norm": 0.29725924134254456, "learning_rate": 3.240555298998362e-05, "loss": 0.0269, "step": 50700 }, { "grad_norm": 0.41008785367012024, "learning_rate": 3.2386208033791194e-05, "loss": 0.0235, "step": 50710 }, { "grad_norm": 0.34435930848121643, "learning_rate": 3.2366866087314064e-05, "loss": 0.0281, "step": 50720 }, { "grad_norm": 0.36195147037506104, "learning_rate": 3.234752715385727e-05, "loss": 0.0272, "step": 50730 }, { "grad_norm": 0.32637709379196167, "learning_rate": 3.232819123672529e-05, "loss": 0.0301, "step": 50740 }, { "grad_norm": 0.5690141320228577, "learning_rate": 3.2308858339222076e-05, "loss": 0.0286, "step": 50750 }, { "grad_norm": 0.2878626585006714, "learning_rate": 3.2289528464651144e-05, "loss": 0.0285, "step": 50760 }, { "grad_norm": 0.7958336472511292, "learning_rate": 3.22702016163154e-05, "loss": 0.0354, "step": 50770 }, { "grad_norm": 0.28041937947273254, "learning_rate": 3.2250877797517295e-05, "loss": 0.0268, "step": 50780 }, { "grad_norm": 0.26113495230674744, "learning_rate": 3.223155701155872e-05, "loss": 0.0282, "step": 50790 }, { "grad_norm": 0.3449813425540924, "learning_rate": 3.2212239261741105e-05, "loss": 0.0324, "step": 50800 }, { "grad_norm": 0.44901934266090393, "learning_rate": 3.219292455136528e-05, "loss": 0.0239, "step": 50810 }, { "grad_norm": 0.4565117657184601, "learning_rate": 3.217361288373165e-05, "loss": 0.0198, "step": 50820 }, { "grad_norm": 0.26268213987350464, "learning_rate": 3.215430426214002e-05, "loss": 0.0218, "step": 50830 }, { "grad_norm": 0.35488635301589966, "learning_rate": 3.2134998689889705e-05, "loss": 0.0232, "step": 50840 }, { "grad_norm": 0.2376396656036377, "learning_rate": 3.211569617027952e-05, "loss": 0.0266, "step": 50850 }, { "grad_norm": 0.40458983182907104, "learning_rate": 3.209639670660769e-05, "loss": 0.0276, "step": 50860 }, { "grad_norm": 0.5868089199066162, "learning_rate": 3.207710030217202e-05, "loss": 0.0282, "step": 50870 }, { "grad_norm": 0.33224785327911377, "learning_rate": 3.205780696026972e-05, "loss": 0.0253, "step": 50880 }, { "grad_norm": 0.2644689679145813, "learning_rate": 3.203851668419749e-05, "loss": 0.0196, "step": 50890 }, { "grad_norm": 0.32016924023628235, "learning_rate": 3.201922947725149e-05, "loss": 0.0229, "step": 50900 }, { "grad_norm": 0.33162030577659607, "learning_rate": 3.199994534272742e-05, "loss": 0.0317, "step": 50910 }, { "grad_norm": 0.32075366377830505, "learning_rate": 3.1980664283920394e-05, "loss": 0.0269, "step": 50920 }, { "grad_norm": 0.39734023809432983, "learning_rate": 3.196138630412499e-05, "loss": 0.0234, "step": 50930 }, { "grad_norm": 0.22650687396526337, "learning_rate": 3.194211140663532e-05, "loss": 0.0219, "step": 50940 }, { "grad_norm": 0.21300294995307922, "learning_rate": 3.1922839594744914e-05, "loss": 0.0255, "step": 50950 }, { "grad_norm": 0.3463926911354065, "learning_rate": 3.190357087174683e-05, "loss": 0.0267, "step": 50960 }, { "grad_norm": 0.22772467136383057, "learning_rate": 3.188430524093353e-05, "loss": 0.0211, "step": 50970 }, { "grad_norm": 0.3247641921043396, "learning_rate": 3.1865042705597014e-05, "loss": 0.0276, "step": 50980 }, { "grad_norm": 0.4047309458255768, "learning_rate": 3.184578326902868e-05, "loss": 0.0233, "step": 50990 }, { "grad_norm": 0.3632564842700958, "learning_rate": 3.182652693451949e-05, "loss": 0.0262, "step": 51000 }, { "grad_norm": 0.4687947928905487, "learning_rate": 3.180727370535981e-05, "loss": 0.0289, "step": 51010 }, { "grad_norm": 0.4110884964466095, "learning_rate": 3.178802358483945e-05, "loss": 0.0267, "step": 51020 }, { "grad_norm": 0.23185043036937714, "learning_rate": 3.17687765762478e-05, "loss": 0.0223, "step": 51030 }, { "grad_norm": 0.4821174144744873, "learning_rate": 3.174953268287357e-05, "loss": 0.0277, "step": 51040 }, { "grad_norm": 0.4299059808254242, "learning_rate": 3.173029190800508e-05, "loss": 0.0229, "step": 51050 }, { "grad_norm": 0.3438865840435028, "learning_rate": 3.171105425493e-05, "loss": 0.03, "step": 51060 }, { "grad_norm": 0.4706786572933197, "learning_rate": 3.169181972693557e-05, "loss": 0.0263, "step": 51070 }, { "grad_norm": 0.47706276178359985, "learning_rate": 3.1672588327308404e-05, "loss": 0.0197, "step": 51080 }, { "grad_norm": 0.3576062321662903, "learning_rate": 3.1653360059334646e-05, "loss": 0.0266, "step": 51090 }, { "grad_norm": 0.5761720538139343, "learning_rate": 3.163413492629985e-05, "loss": 0.0398, "step": 51100 }, { "grad_norm": 0.7255602478981018, "learning_rate": 3.161491293148912e-05, "loss": 0.0283, "step": 51110 }, { "grad_norm": 0.5970762372016907, "learning_rate": 3.1595694078186925e-05, "loss": 0.026, "step": 51120 }, { "grad_norm": 0.29811590909957886, "learning_rate": 3.157647836967724e-05, "loss": 0.024, "step": 51130 }, { "grad_norm": 0.427082896232605, "learning_rate": 3.155726580924354e-05, "loss": 0.0218, "step": 51140 }, { "grad_norm": 0.4091486930847168, "learning_rate": 3.1538056400168676e-05, "loss": 0.0373, "step": 51150 }, { "grad_norm": 0.3396218419075012, "learning_rate": 3.151885014573506e-05, "loss": 0.0291, "step": 51160 }, { "grad_norm": 0.43809524178504944, "learning_rate": 3.149964704922448e-05, "loss": 0.0365, "step": 51170 }, { "grad_norm": 0.25344550609588623, "learning_rate": 3.148044711391825e-05, "loss": 0.0272, "step": 51180 }, { "grad_norm": 0.3225420117378235, "learning_rate": 3.146125034309707e-05, "loss": 0.0308, "step": 51190 }, { "grad_norm": 0.4088955819606781, "learning_rate": 3.1442056740041195e-05, "loss": 0.0308, "step": 51200 }, { "grad_norm": 0.38558927178382874, "learning_rate": 3.142286630803026e-05, "loss": 0.0371, "step": 51210 }, { "grad_norm": 0.3566494584083557, "learning_rate": 3.140367905034337e-05, "loss": 0.0283, "step": 51220 }, { "grad_norm": 0.16907833516597748, "learning_rate": 3.138449497025914e-05, "loss": 0.0285, "step": 51230 }, { "grad_norm": 0.4683540165424347, "learning_rate": 3.136531407105557e-05, "loss": 0.0243, "step": 51240 }, { "grad_norm": 0.385128378868103, "learning_rate": 3.1346136356010184e-05, "loss": 0.032, "step": 51250 }, { "grad_norm": 0.342888206243515, "learning_rate": 3.132696182839988e-05, "loss": 0.0291, "step": 51260 }, { "grad_norm": 0.31726565957069397, "learning_rate": 3.130779049150111e-05, "loss": 0.0323, "step": 51270 }, { "grad_norm": 0.3785193860530853, "learning_rate": 3.128862234858971e-05, "loss": 0.0272, "step": 51280 }, { "grad_norm": 0.3038731515407562, "learning_rate": 3.1269457402941005e-05, "loss": 0.0251, "step": 51290 }, { "grad_norm": 0.43932363390922546, "learning_rate": 3.125029565782974e-05, "loss": 0.0268, "step": 51300 }, { "grad_norm": 0.6709438562393188, "learning_rate": 3.123113711653013e-05, "loss": 0.0272, "step": 51310 }, { "grad_norm": 0.2767361104488373, "learning_rate": 3.121198178231587e-05, "loss": 0.0353, "step": 51320 }, { "grad_norm": 0.19863949716091156, "learning_rate": 3.119282965846006e-05, "loss": 0.0264, "step": 51330 }, { "grad_norm": 0.41505154967308044, "learning_rate": 3.1173680748235303e-05, "loss": 0.0307, "step": 51340 }, { "grad_norm": 0.31174609065055847, "learning_rate": 3.115453505491358e-05, "loss": 0.0237, "step": 51350 }, { "grad_norm": 0.22890883684158325, "learning_rate": 3.1135392581766404e-05, "loss": 0.0273, "step": 51360 }, { "grad_norm": 0.25937139987945557, "learning_rate": 3.111625333206467e-05, "loss": 0.0258, "step": 51370 }, { "grad_norm": 0.28119686245918274, "learning_rate": 3.109711730907881e-05, "loss": 0.0156, "step": 51380 }, { "grad_norm": 0.2425898313522339, "learning_rate": 3.107798451607859e-05, "loss": 0.0203, "step": 51390 }, { "grad_norm": 0.425863116979599, "learning_rate": 3.105885495633328e-05, "loss": 0.028, "step": 51400 }, { "grad_norm": 0.27155911922454834, "learning_rate": 3.103972863311163e-05, "loss": 0.018, "step": 51410 }, { "grad_norm": 0.371181845664978, "learning_rate": 3.102060554968178e-05, "loss": 0.0238, "step": 51420 }, { "grad_norm": 0.4384758770465851, "learning_rate": 3.100148570931137e-05, "loss": 0.0324, "step": 51430 }, { "grad_norm": 0.43679535388946533, "learning_rate": 3.098236911526744e-05, "loss": 0.0294, "step": 51440 }, { "grad_norm": 0.229755699634552, "learning_rate": 3.0963255770816495e-05, "loss": 0.0167, "step": 51450 }, { "grad_norm": 0.4770936667919159, "learning_rate": 3.094414567922447e-05, "loss": 0.0254, "step": 51460 }, { "grad_norm": 0.38194605708122253, "learning_rate": 3.092503884375678e-05, "loss": 0.0249, "step": 51470 }, { "grad_norm": 0.3866026699542999, "learning_rate": 3.0905935267678255e-05, "loss": 0.0303, "step": 51480 }, { "grad_norm": 0.2607730031013489, "learning_rate": 3.0886834954253166e-05, "loss": 0.0222, "step": 51490 }, { "grad_norm": 0.4940936267375946, "learning_rate": 3.086773790674524e-05, "loss": 0.0292, "step": 51500 }, { "grad_norm": 0.2845010459423065, "learning_rate": 3.0848644128417617e-05, "loss": 0.0279, "step": 51510 }, { "grad_norm": 0.340378999710083, "learning_rate": 3.082955362253294e-05, "loss": 0.0272, "step": 51520 }, { "grad_norm": 0.28732559084892273, "learning_rate": 3.081046639235323e-05, "loss": 0.024, "step": 51530 }, { "grad_norm": 0.18576478958129883, "learning_rate": 3.0791382441139986e-05, "loss": 0.0191, "step": 51540 }, { "grad_norm": 0.5097030997276306, "learning_rate": 3.077230177215411e-05, "loss": 0.0318, "step": 51550 }, { "grad_norm": 0.17253145575523376, "learning_rate": 3.0753224388656e-05, "loss": 0.0241, "step": 51560 }, { "grad_norm": 0.2874200642108917, "learning_rate": 3.073415029390544e-05, "loss": 0.025, "step": 51570 }, { "grad_norm": 0.3271150290966034, "learning_rate": 3.071507949116166e-05, "loss": 0.0236, "step": 51580 }, { "grad_norm": 0.5568315982818604, "learning_rate": 3.069601198368337e-05, "loss": 0.0242, "step": 51590 }, { "grad_norm": 0.2488960474729538, "learning_rate": 3.067694777472864e-05, "loss": 0.0254, "step": 51600 }, { "grad_norm": 0.4309334456920624, "learning_rate": 3.065788686755508e-05, "loss": 0.0211, "step": 51610 }, { "grad_norm": 0.2950570285320282, "learning_rate": 3.063882926541961e-05, "loss": 0.0263, "step": 51620 }, { "grad_norm": 0.3857293128967285, "learning_rate": 3.061977497157872e-05, "loss": 0.0242, "step": 51630 }, { "grad_norm": 0.2907068729400635, "learning_rate": 3.0600723989288224e-05, "loss": 0.0238, "step": 51640 }, { "grad_norm": 0.4416877329349518, "learning_rate": 3.058167632180343e-05, "loss": 0.0242, "step": 51650 }, { "grad_norm": 0.3709798753261566, "learning_rate": 3.0562631972379074e-05, "loss": 0.0249, "step": 51660 }, { "grad_norm": 0.1934712827205658, "learning_rate": 3.0543590944269276e-05, "loss": 0.0237, "step": 51670 }, { "grad_norm": 0.4309910833835602, "learning_rate": 3.052455324072766e-05, "loss": 0.0347, "step": 51680 }, { "grad_norm": 0.2749313414096832, "learning_rate": 3.0505518865007244e-05, "loss": 0.0255, "step": 51690 }, { "grad_norm": 0.25791484117507935, "learning_rate": 3.048648782036048e-05, "loss": 0.021, "step": 51700 }, { "grad_norm": 0.26280349493026733, "learning_rate": 3.0467460110039236e-05, "loss": 0.0197, "step": 51710 }, { "grad_norm": 0.31372103095054626, "learning_rate": 3.0448435737294866e-05, "loss": 0.0229, "step": 51720 }, { "grad_norm": 0.32743725180625916, "learning_rate": 3.042941470537808e-05, "loss": 0.0196, "step": 51730 }, { "grad_norm": 0.24150462448596954, "learning_rate": 3.0410397017539072e-05, "loss": 0.0199, "step": 51740 }, { "grad_norm": 0.27546051144599915, "learning_rate": 3.039138267702742e-05, "loss": 0.0224, "step": 51750 }, { "grad_norm": 0.26411378383636475, "learning_rate": 3.0372371687092193e-05, "loss": 0.0227, "step": 51760 }, { "grad_norm": 0.39474695920944214, "learning_rate": 3.0353364050981823e-05, "loss": 0.0281, "step": 51770 }, { "grad_norm": 0.20354188978672028, "learning_rate": 3.033435977194418e-05, "loss": 0.0257, "step": 51780 }, { "grad_norm": 0.35566791892051697, "learning_rate": 3.0315358853226618e-05, "loss": 0.022, "step": 51790 }, { "grad_norm": 0.2944329082965851, "learning_rate": 3.0296361298075844e-05, "loss": 0.027, "step": 51800 }, { "grad_norm": 0.18847370147705078, "learning_rate": 3.027736710973803e-05, "loss": 0.0234, "step": 51810 }, { "grad_norm": 0.26523441076278687, "learning_rate": 3.025837629145875e-05, "loss": 0.025, "step": 51820 }, { "grad_norm": 0.31221169233322144, "learning_rate": 3.0239388846483048e-05, "loss": 0.0265, "step": 51830 }, { "grad_norm": 0.48757752776145935, "learning_rate": 3.022040477805532e-05, "loss": 0.0287, "step": 51840 }, { "grad_norm": 0.4268462359905243, "learning_rate": 3.020142408941946e-05, "loss": 0.0275, "step": 51850 }, { "grad_norm": 0.27540016174316406, "learning_rate": 3.018244678381873e-05, "loss": 0.0279, "step": 51860 }, { "grad_norm": 0.4279480278491974, "learning_rate": 3.0163472864495812e-05, "loss": 0.0342, "step": 51870 }, { "grad_norm": 0.7299830913543701, "learning_rate": 3.014450233469287e-05, "loss": 0.0283, "step": 51880 }, { "grad_norm": 0.3864697217941284, "learning_rate": 3.0125535197651422e-05, "loss": 0.0273, "step": 51890 }, { "grad_norm": 0.2788996696472168, "learning_rate": 3.0106571456612448e-05, "loss": 0.0207, "step": 51900 }, { "grad_norm": 0.34061339497566223, "learning_rate": 3.0087611114816305e-05, "loss": 0.0278, "step": 51910 }, { "grad_norm": 0.4021199643611908, "learning_rate": 3.006865417550284e-05, "loss": 0.0227, "step": 51920 }, { "grad_norm": 0.39839956164360046, "learning_rate": 3.0049700641911242e-05, "loss": 0.0288, "step": 51930 }, { "grad_norm": 0.3691042363643646, "learning_rate": 3.0030750517280175e-05, "loss": 0.0253, "step": 51940 }, { "grad_norm": 0.2966301739215851, "learning_rate": 3.0011803804847682e-05, "loss": 0.0324, "step": 51950 }, { "grad_norm": 0.4207884669303894, "learning_rate": 2.9992860507851224e-05, "loss": 0.0288, "step": 51960 }, { "grad_norm": 0.2941702902317047, "learning_rate": 2.9973920629527723e-05, "loss": 0.0267, "step": 51970 }, { "grad_norm": 0.32868218421936035, "learning_rate": 2.9954984173113453e-05, "loss": 0.0237, "step": 51980 }, { "grad_norm": 0.2903427183628082, "learning_rate": 2.993605114184418e-05, "loss": 0.0276, "step": 51990 }, { "grad_norm": 0.39162030816078186, "learning_rate": 2.9917121538955005e-05, "loss": 0.0324, "step": 52000 }, { "grad_norm": 0.4024960398674011, "learning_rate": 2.98981953676805e-05, "loss": 0.0368, "step": 52010 }, { "grad_norm": 0.2250290960073471, "learning_rate": 2.9879272631254594e-05, "loss": 0.0202, "step": 52020 }, { "grad_norm": 0.19203706085681915, "learning_rate": 2.986035333291073e-05, "loss": 0.0305, "step": 52030 }, { "grad_norm": 0.21398334205150604, "learning_rate": 2.9841437475881646e-05, "loss": 0.026, "step": 52040 }, { "grad_norm": 0.3544965386390686, "learning_rate": 2.9822525063399555e-05, "loss": 0.0255, "step": 52050 }, { "grad_norm": 0.21885772049427032, "learning_rate": 2.9803616098696087e-05, "loss": 0.0281, "step": 52060 }, { "grad_norm": 0.29012632369995117, "learning_rate": 2.9784710585002228e-05, "loss": 0.0278, "step": 52070 }, { "grad_norm": 0.33571383357048035, "learning_rate": 2.9765808525548467e-05, "loss": 0.0226, "step": 52080 }, { "grad_norm": 0.7457991242408752, "learning_rate": 2.974690992356461e-05, "loss": 0.0249, "step": 52090 }, { "grad_norm": 0.25656402111053467, "learning_rate": 2.9728014782279934e-05, "loss": 0.0295, "step": 52100 }, { "grad_norm": 0.4344411790370941, "learning_rate": 2.970912310492307e-05, "loss": 0.0247, "step": 52110 }, { "grad_norm": 0.25955191254615784, "learning_rate": 2.9690234894722134e-05, "loss": 0.0221, "step": 52120 }, { "grad_norm": 0.28088414669036865, "learning_rate": 2.9671350154904577e-05, "loss": 0.0298, "step": 52130 }, { "grad_norm": 0.4150089919567108, "learning_rate": 2.9652468888697282e-05, "loss": 0.0266, "step": 52140 }, { "grad_norm": 0.24728591740131378, "learning_rate": 2.9633591099326562e-05, "loss": 0.0185, "step": 52150 }, { "grad_norm": 1.0209333896636963, "learning_rate": 2.9614716790018072e-05, "loss": 0.0217, "step": 52160 }, { "grad_norm": 0.31535908579826355, "learning_rate": 2.959584596399697e-05, "loss": 0.02, "step": 52170 }, { "grad_norm": 0.3661200702190399, "learning_rate": 2.9576978624487717e-05, "loss": 0.0336, "step": 52180 }, { "grad_norm": 0.23708726465702057, "learning_rate": 2.9558114774714263e-05, "loss": 0.029, "step": 52190 }, { "grad_norm": 0.22193555533885956, "learning_rate": 2.9539254417899897e-05, "loss": 0.0242, "step": 52200 }, { "grad_norm": 0.3552039563655853, "learning_rate": 2.9520397557267365e-05, "loss": 0.0231, "step": 52210 }, { "grad_norm": 0.2586785852909088, "learning_rate": 2.9501544196038765e-05, "loss": 0.0244, "step": 52220 }, { "grad_norm": 0.2263137400150299, "learning_rate": 2.9482694337435614e-05, "loss": 0.0328, "step": 52230 }, { "grad_norm": 0.347311794757843, "learning_rate": 2.946384798467887e-05, "loss": 0.0226, "step": 52240 }, { "grad_norm": 0.26510101556777954, "learning_rate": 2.9445005140988824e-05, "loss": 0.0207, "step": 52250 }, { "grad_norm": 0.26813793182373047, "learning_rate": 2.942616580958524e-05, "loss": 0.0209, "step": 52260 }, { "grad_norm": 0.422091543674469, "learning_rate": 2.9407329993687193e-05, "loss": 0.0257, "step": 52270 }, { "grad_norm": 0.3512860834598541, "learning_rate": 2.938849769651326e-05, "loss": 0.0227, "step": 52280 }, { "grad_norm": 0.29163122177124023, "learning_rate": 2.9369668921281345e-05, "loss": 0.0236, "step": 52290 }, { "grad_norm": 0.3911783993244171, "learning_rate": 2.9350843671208773e-05, "loss": 0.0191, "step": 52300 }, { "grad_norm": 0.33625662326812744, "learning_rate": 2.9332021949512245e-05, "loss": 0.0254, "step": 52310 }, { "grad_norm": 0.5015355944633484, "learning_rate": 2.9313203759407908e-05, "loss": 0.0258, "step": 52320 }, { "grad_norm": 0.20330286026000977, "learning_rate": 2.929438910411127e-05, "loss": 0.027, "step": 52330 }, { "grad_norm": 0.3113930821418762, "learning_rate": 2.927557798683722e-05, "loss": 0.0292, "step": 52340 }, { "grad_norm": 0.3123323619365692, "learning_rate": 2.9256770410800095e-05, "loss": 0.0162, "step": 52350 }, { "grad_norm": 0.2547025978565216, "learning_rate": 2.9237966379213554e-05, "loss": 0.032, "step": 52360 }, { "grad_norm": 0.24115681648254395, "learning_rate": 2.9219165895290736e-05, "loss": 0.0215, "step": 52370 }, { "grad_norm": 0.8514545559883118, "learning_rate": 2.9200368962244083e-05, "loss": 0.0273, "step": 52380 }, { "grad_norm": 0.2864883542060852, "learning_rate": 2.9181575583285536e-05, "loss": 0.0224, "step": 52390 }, { "grad_norm": 0.32468318939208984, "learning_rate": 2.916278576162632e-05, "loss": 0.0246, "step": 52400 }, { "grad_norm": 0.3123946487903595, "learning_rate": 2.9143999500477116e-05, "loss": 0.0222, "step": 52410 }, { "grad_norm": 0.32255104184150696, "learning_rate": 2.9125216803048004e-05, "loss": 0.0277, "step": 52420 }, { "grad_norm": 0.41371747851371765, "learning_rate": 2.9106437672548403e-05, "loss": 0.0277, "step": 52430 }, { "grad_norm": 0.31288963556289673, "learning_rate": 2.9087662112187154e-05, "loss": 0.0229, "step": 52440 }, { "grad_norm": 0.3268146812915802, "learning_rate": 2.9068890125172498e-05, "loss": 0.0179, "step": 52450 }, { "grad_norm": 0.1319279670715332, "learning_rate": 2.9050121714712058e-05, "loss": 0.0302, "step": 52460 }, { "grad_norm": 0.46983420848846436, "learning_rate": 2.9031356884012835e-05, "loss": 0.0215, "step": 52470 }, { "grad_norm": 0.3496924936771393, "learning_rate": 2.9012595636281247e-05, "loss": 0.0309, "step": 52480 }, { "grad_norm": 0.26695743203163147, "learning_rate": 2.8993837974723016e-05, "loss": 0.0411, "step": 52490 }, { "grad_norm": 0.2577723562717438, "learning_rate": 2.8975083902543394e-05, "loss": 0.0263, "step": 52500 }, { "grad_norm": 0.18990419805049896, "learning_rate": 2.895633342294688e-05, "loss": 0.0193, "step": 52510 }, { "grad_norm": 0.3137800097465515, "learning_rate": 2.893758653913744e-05, "loss": 0.0249, "step": 52520 }, { "grad_norm": 0.22145695984363556, "learning_rate": 2.8918843254318423e-05, "loss": 0.021, "step": 52530 }, { "grad_norm": 0.40532100200653076, "learning_rate": 2.8900103571692483e-05, "loss": 0.0246, "step": 52540 }, { "grad_norm": 0.5477864146232605, "learning_rate": 2.8881367494461796e-05, "loss": 0.0283, "step": 52550 }, { "grad_norm": 0.252284437417984, "learning_rate": 2.886263502582779e-05, "loss": 0.0218, "step": 52560 }, { "grad_norm": 0.2530975639820099, "learning_rate": 2.8843906168991353e-05, "loss": 0.0251, "step": 52570 }, { "grad_norm": 0.26437893509864807, "learning_rate": 2.8825180927152728e-05, "loss": 0.0234, "step": 52580 }, { "grad_norm": 0.24629728496074677, "learning_rate": 2.8806459303511547e-05, "loss": 0.0281, "step": 52590 }, { "grad_norm": 0.5561893582344055, "learning_rate": 2.8787741301266852e-05, "loss": 0.0245, "step": 52600 }, { "grad_norm": 0.2857604920864105, "learning_rate": 2.876902692361697e-05, "loss": 0.0206, "step": 52610 }, { "grad_norm": 0.3096562325954437, "learning_rate": 2.875031617375976e-05, "loss": 0.0192, "step": 52620 }, { "grad_norm": 0.48691707849502563, "learning_rate": 2.8731609054892318e-05, "loss": 0.024, "step": 52630 }, { "grad_norm": 0.34368249773979187, "learning_rate": 2.871290557021119e-05, "loss": 0.0222, "step": 52640 }, { "grad_norm": 0.2918304204940796, "learning_rate": 2.8694205722912298e-05, "loss": 0.0232, "step": 52650 }, { "grad_norm": 0.23462024331092834, "learning_rate": 2.8675509516190936e-05, "loss": 0.0282, "step": 52660 }, { "grad_norm": 0.37867307662963867, "learning_rate": 2.8656816953241772e-05, "loss": 0.03, "step": 52670 }, { "grad_norm": 0.343351811170578, "learning_rate": 2.863812803725887e-05, "loss": 0.0214, "step": 52680 }, { "grad_norm": 0.32986271381378174, "learning_rate": 2.8619442771435623e-05, "loss": 0.0164, "step": 52690 }, { "grad_norm": 0.36257874965667725, "learning_rate": 2.860076115896484e-05, "loss": 0.0335, "step": 52700 }, { "grad_norm": 0.29965662956237793, "learning_rate": 2.8582083203038712e-05, "loss": 0.0231, "step": 52710 }, { "grad_norm": 0.33402183651924133, "learning_rate": 2.8563408906848778e-05, "loss": 0.0199, "step": 52720 }, { "grad_norm": 0.48032107949256897, "learning_rate": 2.8544738273585993e-05, "loss": 0.0239, "step": 52730 }, { "grad_norm": 0.17850732803344727, "learning_rate": 2.8526071306440595e-05, "loss": 0.0181, "step": 52740 }, { "grad_norm": 0.27683088183403015, "learning_rate": 2.850740800860233e-05, "loss": 0.0221, "step": 52750 }, { "grad_norm": 0.41034820675849915, "learning_rate": 2.8488748383260178e-05, "loss": 0.0209, "step": 52760 }, { "grad_norm": 0.5881392359733582, "learning_rate": 2.8470092433602635e-05, "loss": 0.0303, "step": 52770 }, { "grad_norm": 0.32464492321014404, "learning_rate": 2.845144016281745e-05, "loss": 0.0223, "step": 52780 }, { "grad_norm": 0.4869917333126068, "learning_rate": 2.8432791574091744e-05, "loss": 0.0355, "step": 52790 }, { "grad_norm": 0.3141070604324341, "learning_rate": 2.8414146670612134e-05, "loss": 0.0245, "step": 52800 }, { "grad_norm": 0.2626979947090149, "learning_rate": 2.8395505455564446e-05, "loss": 0.026, "step": 52810 }, { "grad_norm": 0.24297726154327393, "learning_rate": 2.837686793213403e-05, "loss": 0.0208, "step": 52820 }, { "grad_norm": 0.2689044773578644, "learning_rate": 2.8358234103505477e-05, "loss": 0.0233, "step": 52830 }, { "grad_norm": 0.36349448561668396, "learning_rate": 2.8339603972862806e-05, "loss": 0.0269, "step": 52840 }, { "grad_norm": 0.39160069823265076, "learning_rate": 2.83209775433894e-05, "loss": 0.027, "step": 52850 }, { "grad_norm": 0.25027403235435486, "learning_rate": 2.8302354818268013e-05, "loss": 0.0218, "step": 52860 }, { "grad_norm": 0.3038659989833832, "learning_rate": 2.8283735800680754e-05, "loss": 0.0232, "step": 52870 }, { "grad_norm": 0.27142953872680664, "learning_rate": 2.826512049380913e-05, "loss": 0.023, "step": 52880 }, { "grad_norm": 0.26212015748023987, "learning_rate": 2.8246508900833935e-05, "loss": 0.0252, "step": 52890 }, { "grad_norm": 0.23306401073932648, "learning_rate": 2.822790102493541e-05, "loss": 0.0215, "step": 52900 }, { "grad_norm": 0.308756947517395, "learning_rate": 2.820929686929314e-05, "loss": 0.027, "step": 52910 }, { "grad_norm": 0.29793113470077515, "learning_rate": 2.819069643708605e-05, "loss": 0.0251, "step": 52920 }, { "grad_norm": 0.28588414192199707, "learning_rate": 2.8172099731492484e-05, "loss": 0.0266, "step": 52930 }, { "grad_norm": 0.3501029312610626, "learning_rate": 2.815350675569003e-05, "loss": 0.0335, "step": 52940 }, { "grad_norm": 0.28792116045951843, "learning_rate": 2.8134917512855807e-05, "loss": 0.0263, "step": 52950 }, { "grad_norm": 0.30106714367866516, "learning_rate": 2.8116332006166134e-05, "loss": 0.0194, "step": 52960 }, { "grad_norm": 0.3835007846355438, "learning_rate": 2.809775023879685e-05, "loss": 0.0302, "step": 52970 }, { "grad_norm": 0.2300378680229187, "learning_rate": 2.807917221392299e-05, "loss": 0.0225, "step": 52980 }, { "grad_norm": 0.3397236168384552, "learning_rate": 2.8060597934719067e-05, "loss": 0.0183, "step": 52990 }, { "grad_norm": 0.35890474915504456, "learning_rate": 2.8042027404358927e-05, "loss": 0.0272, "step": 53000 }, { "grad_norm": 0.32673418521881104, "learning_rate": 2.802346062601572e-05, "loss": 0.0201, "step": 53010 }, { "grad_norm": 0.4134257435798645, "learning_rate": 2.800489760286207e-05, "loss": 0.0324, "step": 53020 }, { "grad_norm": 0.3478892743587494, "learning_rate": 2.7986338338069834e-05, "loss": 0.0216, "step": 53030 }, { "grad_norm": 0.4550318419933319, "learning_rate": 2.7967782834810297e-05, "loss": 0.0296, "step": 53040 }, { "grad_norm": 0.2333212047815323, "learning_rate": 2.794923109625409e-05, "loss": 0.0243, "step": 53050 }, { "grad_norm": 0.3413878381252289, "learning_rate": 2.7930683125571205e-05, "loss": 0.0209, "step": 53060 }, { "grad_norm": 0.30096885561943054, "learning_rate": 2.7912138925930997e-05, "loss": 0.0263, "step": 53070 }, { "grad_norm": 0.2898387014865875, "learning_rate": 2.7893598500502117e-05, "loss": 0.0416, "step": 53080 }, { "grad_norm": 0.313090980052948, "learning_rate": 2.7875061852452644e-05, "loss": 0.0241, "step": 53090 }, { "grad_norm": 0.3549706041812897, "learning_rate": 2.7856528984949982e-05, "loss": 0.0255, "step": 53100 }, { "grad_norm": 0.5908457040786743, "learning_rate": 2.7837999901160888e-05, "loss": 0.0186, "step": 53110 }, { "grad_norm": 0.22948363423347473, "learning_rate": 2.7819474604251484e-05, "loss": 0.0206, "step": 53120 }, { "grad_norm": 0.36833760142326355, "learning_rate": 2.780095309738725e-05, "loss": 0.0236, "step": 53130 }, { "grad_norm": 0.2761562168598175, "learning_rate": 2.778243538373294e-05, "loss": 0.0213, "step": 53140 }, { "grad_norm": 0.28073760867118835, "learning_rate": 2.7763921466452826e-05, "loss": 0.0261, "step": 53150 }, { "grad_norm": 0.35958626866340637, "learning_rate": 2.7745411348710336e-05, "loss": 0.0338, "step": 53160 }, { "grad_norm": 0.35451239347457886, "learning_rate": 2.7726905033668395e-05, "loss": 0.0246, "step": 53170 }, { "grad_norm": 0.217522993683815, "learning_rate": 2.7708402524489214e-05, "loss": 0.029, "step": 53180 }, { "grad_norm": 0.3009614646434784, "learning_rate": 2.7689903824334364e-05, "loss": 0.0284, "step": 53190 }, { "grad_norm": 0.47296005487442017, "learning_rate": 2.7671408936364785e-05, "loss": 0.0285, "step": 53200 }, { "grad_norm": 0.20973116159439087, "learning_rate": 2.765291786374069e-05, "loss": 0.0222, "step": 53210 }, { "grad_norm": 0.2819311022758484, "learning_rate": 2.7634430609621775e-05, "loss": 0.0204, "step": 53220 }, { "grad_norm": 0.2941216230392456, "learning_rate": 2.7615947177166956e-05, "loss": 0.0207, "step": 53230 }, { "grad_norm": 0.29199814796447754, "learning_rate": 2.7597467569534553e-05, "loss": 0.0227, "step": 53240 }, { "grad_norm": 0.2583925426006317, "learning_rate": 2.757899178988226e-05, "loss": 0.0214, "step": 53250 }, { "grad_norm": 0.23254084587097168, "learning_rate": 2.7560519841367005e-05, "loss": 0.0232, "step": 53260 }, { "grad_norm": 0.35353991389274597, "learning_rate": 2.7542051727145237e-05, "loss": 0.0212, "step": 53270 }, { "grad_norm": 0.34561529755592346, "learning_rate": 2.7523587450372578e-05, "loss": 0.0278, "step": 53280 }, { "grad_norm": 0.3014874756336212, "learning_rate": 2.750512701420409e-05, "loss": 0.0244, "step": 53290 }, { "grad_norm": 0.4347054958343506, "learning_rate": 2.7486670421794158e-05, "loss": 0.0275, "step": 53300 }, { "grad_norm": 0.2978466749191284, "learning_rate": 2.7468217676296515e-05, "loss": 0.0323, "step": 53310 }, { "grad_norm": 0.4307342767715454, "learning_rate": 2.7449768780864226e-05, "loss": 0.026, "step": 53320 }, { "grad_norm": 0.3559168577194214, "learning_rate": 2.7431323738649724e-05, "loss": 0.0226, "step": 53330 }, { "grad_norm": 0.3975730836391449, "learning_rate": 2.7412882552804713e-05, "loss": 0.0305, "step": 53340 }, { "grad_norm": 0.4700160026550293, "learning_rate": 2.739444522648032e-05, "loss": 0.0243, "step": 53350 }, { "grad_norm": 0.26806148886680603, "learning_rate": 2.737601176282697e-05, "loss": 0.0199, "step": 53360 }, { "grad_norm": 0.47213494777679443, "learning_rate": 2.735758216499445e-05, "loss": 0.0258, "step": 53370 }, { "grad_norm": 0.21984612941741943, "learning_rate": 2.7339156436131864e-05, "loss": 0.0239, "step": 53380 }, { "grad_norm": 0.3172493577003479, "learning_rate": 2.7320734579387663e-05, "loss": 0.0198, "step": 53390 }, { "grad_norm": 0.5472540855407715, "learning_rate": 2.730231659790966e-05, "loss": 0.0217, "step": 53400 }, { "grad_norm": 0.3184978663921356, "learning_rate": 2.728390249484494e-05, "loss": 0.0303, "step": 53410 }, { "grad_norm": 0.365701287984848, "learning_rate": 2.7265492273340032e-05, "loss": 0.021, "step": 53420 }, { "grad_norm": 0.3205854296684265, "learning_rate": 2.7247085936540678e-05, "loss": 0.0245, "step": 53430 }, { "grad_norm": 0.2913914620876312, "learning_rate": 2.722868348759205e-05, "loss": 0.0233, "step": 53440 }, { "grad_norm": 0.2681647539138794, "learning_rate": 2.7210284929638635e-05, "loss": 0.0266, "step": 53450 }, { "grad_norm": 0.4593690037727356, "learning_rate": 2.7191890265824183e-05, "loss": 0.0244, "step": 53460 }, { "grad_norm": 0.2611878216266632, "learning_rate": 2.7173499499291926e-05, "loss": 0.0188, "step": 53470 }, { "grad_norm": 0.45138075947761536, "learning_rate": 2.7155112633184277e-05, "loss": 0.0223, "step": 53480 }, { "grad_norm": 0.31182172894477844, "learning_rate": 2.713672967064307e-05, "loss": 0.024, "step": 53490 }, { "grad_norm": 0.19765028357505798, "learning_rate": 2.711835061480945e-05, "loss": 0.0178, "step": 53500 }, { "grad_norm": 0.26115450263023376, "learning_rate": 2.7099975468823896e-05, "loss": 0.0262, "step": 53510 }, { "grad_norm": 0.2354046106338501, "learning_rate": 2.708160423582622e-05, "loss": 0.0212, "step": 53520 }, { "grad_norm": 0.49142149090766907, "learning_rate": 2.706323691895557e-05, "loss": 0.0316, "step": 53530 }, { "grad_norm": 0.36933591961860657, "learning_rate": 2.70448735213504e-05, "loss": 0.0233, "step": 53540 }, { "grad_norm": 0.30231669545173645, "learning_rate": 2.702651404614852e-05, "loss": 0.0275, "step": 53550 }, { "grad_norm": 0.23891273140907288, "learning_rate": 2.7008158496487056e-05, "loss": 0.0239, "step": 53560 }, { "grad_norm": 1.0905215740203857, "learning_rate": 2.6989806875502487e-05, "loss": 0.0264, "step": 53570 }, { "grad_norm": 0.43665269017219543, "learning_rate": 2.6971459186330584e-05, "loss": 0.0228, "step": 53580 }, { "grad_norm": 0.24281349778175354, "learning_rate": 2.695311543210648e-05, "loss": 0.021, "step": 53590 }, { "grad_norm": 0.34900420904159546, "learning_rate": 2.693477561596463e-05, "loss": 0.025, "step": 53600 }, { "grad_norm": 0.4381483495235443, "learning_rate": 2.6916439741038756e-05, "loss": 0.0246, "step": 53610 }, { "grad_norm": 0.3127874433994293, "learning_rate": 2.689810781046203e-05, "loss": 0.0222, "step": 53620 }, { "grad_norm": 0.4008442759513855, "learning_rate": 2.6879779827366823e-05, "loss": 0.0243, "step": 53630 }, { "grad_norm": 0.31690576672554016, "learning_rate": 2.6861455794884904e-05, "loss": 0.0224, "step": 53640 }, { "grad_norm": 0.3264029920101166, "learning_rate": 2.6843135716147373e-05, "loss": 0.0252, "step": 53650 }, { "grad_norm": 0.4691690504550934, "learning_rate": 2.6824819594284556e-05, "loss": 0.0273, "step": 53660 }, { "grad_norm": 0.3116459548473358, "learning_rate": 2.6806507432426275e-05, "loss": 0.0218, "step": 53670 }, { "grad_norm": 0.35208624601364136, "learning_rate": 2.6788199233701512e-05, "loss": 0.0249, "step": 53680 }, { "grad_norm": 0.5122194886207581, "learning_rate": 2.6769895001238652e-05, "loss": 0.0344, "step": 53690 }, { "grad_norm": 0.33954620361328125, "learning_rate": 2.67515947381654e-05, "loss": 0.0277, "step": 53700 }, { "grad_norm": 0.3342248499393463, "learning_rate": 2.6733298447608745e-05, "loss": 0.0221, "step": 53710 }, { "grad_norm": 0.21985986828804016, "learning_rate": 2.6715006132695074e-05, "loss": 0.0276, "step": 53720 }, { "grad_norm": 0.303244411945343, "learning_rate": 2.6696717796549985e-05, "loss": 0.0296, "step": 53730 }, { "grad_norm": 0.3532012403011322, "learning_rate": 2.667843344229848e-05, "loss": 0.025, "step": 53740 }, { "grad_norm": 0.2944257855415344, "learning_rate": 2.6660153073064846e-05, "loss": 0.0214, "step": 53750 }, { "grad_norm": 0.19028200209140778, "learning_rate": 2.664187669197271e-05, "loss": 0.0351, "step": 53760 }, { "grad_norm": 0.28280365467071533, "learning_rate": 2.6623604302145005e-05, "loss": 0.0263, "step": 53770 }, { "grad_norm": 0.47209638357162476, "learning_rate": 2.6605335906703975e-05, "loss": 0.0225, "step": 53780 }, { "grad_norm": 0.5611914396286011, "learning_rate": 2.65870715087712e-05, "loss": 0.0298, "step": 53790 }, { "grad_norm": 0.3906201422214508, "learning_rate": 2.6568811111467573e-05, "loss": 0.0252, "step": 53800 }, { "grad_norm": 0.385365754365921, "learning_rate": 2.6550554717913258e-05, "loss": 0.0227, "step": 53810 }, { "grad_norm": 0.41750892996788025, "learning_rate": 2.6532302331227805e-05, "loss": 0.028, "step": 53820 }, { "grad_norm": 0.34655267000198364, "learning_rate": 2.651405395453004e-05, "loss": 0.0232, "step": 53830 }, { "grad_norm": 0.5356264710426331, "learning_rate": 2.6495809590938115e-05, "loss": 0.0309, "step": 53840 }, { "grad_norm": 0.36812475323677063, "learning_rate": 2.64775692435695e-05, "loss": 0.0279, "step": 53850 }, { "grad_norm": 0.25945666432380676, "learning_rate": 2.6459332915540928e-05, "loss": 0.02, "step": 53860 }, { "grad_norm": 0.33498650789260864, "learning_rate": 2.644110060996856e-05, "loss": 0.0296, "step": 53870 }, { "grad_norm": 0.27790871262550354, "learning_rate": 2.642287232996774e-05, "loss": 0.0212, "step": 53880 }, { "grad_norm": 0.3686608374118805, "learning_rate": 2.6404648078653205e-05, "loss": 0.0236, "step": 53890 }, { "grad_norm": 0.2629036009311676, "learning_rate": 2.6386427859139002e-05, "loss": 0.0162, "step": 53900 }, { "grad_norm": 0.47093722224235535, "learning_rate": 2.63682116745384e-05, "loss": 0.0305, "step": 53910 }, { "grad_norm": 0.3102227747440338, "learning_rate": 2.6349999527964138e-05, "loss": 0.0213, "step": 53920 }, { "grad_norm": 0.41830480098724365, "learning_rate": 2.63317914225281e-05, "loss": 0.0285, "step": 53930 }, { "grad_norm": 0.33896228671073914, "learning_rate": 2.6313587361341585e-05, "loss": 0.0285, "step": 53940 }, { "grad_norm": 0.527157187461853, "learning_rate": 2.6295387347515165e-05, "loss": 0.0245, "step": 53950 }, { "grad_norm": 0.3343249559402466, "learning_rate": 2.6277191384158727e-05, "loss": 0.03, "step": 53960 }, { "grad_norm": 0.30131128430366516, "learning_rate": 2.6258999474381457e-05, "loss": 0.0262, "step": 53970 }, { "grad_norm": 0.3163471817970276, "learning_rate": 2.624081162129186e-05, "loss": 0.0235, "step": 53980 }, { "grad_norm": 0.39184027910232544, "learning_rate": 2.6222627827997765e-05, "loss": 0.035, "step": 53990 }, { "grad_norm": 0.23887479305267334, "learning_rate": 2.6204448097606236e-05, "loss": 0.024, "step": 54000 }, { "grad_norm": 0.3530261516571045, "learning_rate": 2.6186272433223726e-05, "loss": 0.0221, "step": 54010 }, { "grad_norm": 0.3207967281341553, "learning_rate": 2.6168100837955943e-05, "loss": 0.0241, "step": 54020 }, { "grad_norm": 0.34953397512435913, "learning_rate": 2.6149933314907926e-05, "loss": 0.0302, "step": 54030 }, { "grad_norm": 0.47101160883903503, "learning_rate": 2.6131769867184e-05, "loss": 0.0322, "step": 54040 }, { "grad_norm": 0.22849567234516144, "learning_rate": 2.611361049788783e-05, "loss": 0.0262, "step": 54050 }, { "grad_norm": 0.2779828906059265, "learning_rate": 2.6095455210122292e-05, "loss": 0.0227, "step": 54060 }, { "grad_norm": 0.3148435950279236, "learning_rate": 2.6077304006989712e-05, "loss": 0.0237, "step": 54070 }, { "grad_norm": 0.3612370491027832, "learning_rate": 2.6059156891591562e-05, "loss": 0.0215, "step": 54080 }, { "grad_norm": 0.44828349351882935, "learning_rate": 2.6041013867028718e-05, "loss": 0.0249, "step": 54090 }, { "grad_norm": 0.3183101713657379, "learning_rate": 2.6022874936401347e-05, "loss": 0.0203, "step": 54100 }, { "grad_norm": 0.6980341672897339, "learning_rate": 2.6004740102808832e-05, "loss": 0.0308, "step": 54110 }, { "grad_norm": 0.36120423674583435, "learning_rate": 2.598660936935e-05, "loss": 0.0195, "step": 54120 }, { "grad_norm": 0.36310043931007385, "learning_rate": 2.5968482739122845e-05, "loss": 0.0342, "step": 54130 }, { "grad_norm": 0.3293682932853699, "learning_rate": 2.595036021522472e-05, "loss": 0.0206, "step": 54140 }, { "grad_norm": 0.20293158292770386, "learning_rate": 2.5932241800752278e-05, "loss": 0.0248, "step": 54150 }, { "grad_norm": 0.246717631816864, "learning_rate": 2.5914127498801453e-05, "loss": 0.0333, "step": 54160 }, { "grad_norm": 0.5541892051696777, "learning_rate": 2.5896017312467497e-05, "loss": 0.0217, "step": 54170 }, { "grad_norm": 0.4501093029975891, "learning_rate": 2.587791124484493e-05, "loss": 0.0203, "step": 54180 }, { "grad_norm": 0.20825275778770447, "learning_rate": 2.5859809299027615e-05, "loss": 0.0264, "step": 54190 }, { "grad_norm": 0.47383660078048706, "learning_rate": 2.5841711478108632e-05, "loss": 0.0244, "step": 54200 }, { "grad_norm": 0.3699464797973633, "learning_rate": 2.582361778518043e-05, "loss": 0.0194, "step": 54210 }, { "grad_norm": 0.3138551414012909, "learning_rate": 2.580552822333472e-05, "loss": 0.0222, "step": 54220 }, { "grad_norm": 0.3072049915790558, "learning_rate": 2.578744279566252e-05, "loss": 0.0248, "step": 54230 }, { "grad_norm": 0.36530423164367676, "learning_rate": 2.576936150525413e-05, "loss": 0.0315, "step": 54240 }, { "grad_norm": 0.26718851923942566, "learning_rate": 2.5751284355199168e-05, "loss": 0.0172, "step": 54250 }, { "grad_norm": 0.3198896646499634, "learning_rate": 2.573321134858646e-05, "loss": 0.0197, "step": 54260 }, { "grad_norm": 0.26339712738990784, "learning_rate": 2.5715142488504286e-05, "loss": 0.0222, "step": 54270 }, { "grad_norm": 0.32836368680000305, "learning_rate": 2.5697077778040042e-05, "loss": 0.0242, "step": 54280 }, { "grad_norm": 0.35737454891204834, "learning_rate": 2.5679017220280522e-05, "loss": 0.0385, "step": 54290 }, { "grad_norm": 0.3690197467803955, "learning_rate": 2.5660960818311796e-05, "loss": 0.0268, "step": 54300 }, { "grad_norm": 0.5014678835868835, "learning_rate": 2.564290857521915e-05, "loss": 0.027, "step": 54310 }, { "grad_norm": 0.3147900700569153, "learning_rate": 2.5624860494087298e-05, "loss": 0.0276, "step": 54320 }, { "grad_norm": 0.28714680671691895, "learning_rate": 2.5606816578000115e-05, "loss": 0.0335, "step": 54330 }, { "grad_norm": 0.39265134930610657, "learning_rate": 2.558877683004082e-05, "loss": 0.0247, "step": 54340 }, { "grad_norm": 0.6659910678863525, "learning_rate": 2.557074125329192e-05, "loss": 0.0269, "step": 54350 }, { "grad_norm": 0.34640541672706604, "learning_rate": 2.5552709850835195e-05, "loss": 0.0252, "step": 54360 }, { "grad_norm": 0.18473342061042786, "learning_rate": 2.5534682625751738e-05, "loss": 0.0235, "step": 54370 }, { "grad_norm": 0.25072890520095825, "learning_rate": 2.551665958112186e-05, "loss": 0.0313, "step": 54380 }, { "grad_norm": 0.25507891178131104, "learning_rate": 2.549864072002527e-05, "loss": 0.0221, "step": 54390 }, { "grad_norm": 0.20090562105178833, "learning_rate": 2.5480626045540858e-05, "loss": 0.0285, "step": 54400 }, { "grad_norm": 0.22171975672245026, "learning_rate": 2.546261556074684e-05, "loss": 0.022, "step": 54410 }, { "grad_norm": 0.23410923779010773, "learning_rate": 2.5444609268720726e-05, "loss": 0.0224, "step": 54420 }, { "grad_norm": 0.2007909119129181, "learning_rate": 2.5426607172539297e-05, "loss": 0.0207, "step": 54430 }, { "grad_norm": 0.3865693211555481, "learning_rate": 2.5408609275278617e-05, "loss": 0.0338, "step": 54440 }, { "grad_norm": 0.22273732721805573, "learning_rate": 2.5390615580014055e-05, "loss": 0.0276, "step": 54450 }, { "grad_norm": 0.20350109040737152, "learning_rate": 2.5372626089820207e-05, "loss": 0.0227, "step": 54460 }, { "grad_norm": 0.3107542395591736, "learning_rate": 2.5354640807770997e-05, "loss": 0.0181, "step": 54470 }, { "grad_norm": 0.32089173793792725, "learning_rate": 2.5336659736939622e-05, "loss": 0.0262, "step": 54480 }, { "grad_norm": 0.284802109003067, "learning_rate": 2.5318682880398554e-05, "loss": 0.0256, "step": 54490 }, { "grad_norm": 0.2697809636592865, "learning_rate": 2.530071024121956e-05, "loss": 0.0246, "step": 54500 }, { "grad_norm": 0.28268322348594666, "learning_rate": 2.5282741822473627e-05, "loss": 0.0203, "step": 54510 }, { "grad_norm": 0.41388484835624695, "learning_rate": 2.526477762723114e-05, "loss": 0.03, "step": 54520 }, { "grad_norm": 0.312845915555954, "learning_rate": 2.5246817658561618e-05, "loss": 0.0207, "step": 54530 }, { "grad_norm": 0.21240732073783875, "learning_rate": 2.5228861919533965e-05, "loss": 0.0237, "step": 54540 }, { "grad_norm": 0.20923136174678802, "learning_rate": 2.5210910413216326e-05, "loss": 0.0379, "step": 54550 }, { "grad_norm": 0.3908064663410187, "learning_rate": 2.5192963142676086e-05, "loss": 0.0159, "step": 54560 }, { "grad_norm": 0.2385266274213791, "learning_rate": 2.517502011098001e-05, "loss": 0.0192, "step": 54570 }, { "grad_norm": 0.27904683351516724, "learning_rate": 2.5157081321193987e-05, "loss": 0.0204, "step": 54580 }, { "grad_norm": 0.25594621896743774, "learning_rate": 2.5139146776383356e-05, "loss": 0.0211, "step": 54590 }, { "grad_norm": 0.25488945841789246, "learning_rate": 2.5121216479612575e-05, "loss": 0.0222, "step": 54600 }, { "grad_norm": 0.25207653641700745, "learning_rate": 2.510329043394546e-05, "loss": 0.0343, "step": 54610 }, { "grad_norm": 0.31482675671577454, "learning_rate": 2.508536864244508e-05, "loss": 0.0259, "step": 54620 }, { "grad_norm": 0.35931524634361267, "learning_rate": 2.5067451108173778e-05, "loss": 0.0233, "step": 54630 }, { "grad_norm": 0.22810009121894836, "learning_rate": 2.5049537834193204e-05, "loss": 0.0209, "step": 54640 }, { "grad_norm": 0.3979566991329193, "learning_rate": 2.5031628823564194e-05, "loss": 0.0228, "step": 54650 }, { "grad_norm": 0.31228676438331604, "learning_rate": 2.5013724079346933e-05, "loss": 0.026, "step": 54660 }, { "grad_norm": 0.3903358578681946, "learning_rate": 2.4995823604600854e-05, "loss": 0.0209, "step": 54670 }, { "grad_norm": 0.22207339107990265, "learning_rate": 2.497792740238465e-05, "loss": 0.0184, "step": 54680 }, { "grad_norm": 0.30149948596954346, "learning_rate": 2.49600354757563e-05, "loss": 0.031, "step": 54690 }, { "grad_norm": 0.28528180718421936, "learning_rate": 2.494214782777306e-05, "loss": 0.0228, "step": 54700 }, { "grad_norm": 0.4115441143512726, "learning_rate": 2.4924264461491386e-05, "loss": 0.0208, "step": 54710 }, { "grad_norm": 0.33095023036003113, "learning_rate": 2.4906385379967133e-05, "loss": 0.026, "step": 54720 }, { "grad_norm": 0.3645992875099182, "learning_rate": 2.4888510586255285e-05, "loss": 0.0219, "step": 54730 }, { "grad_norm": 0.1707889586687088, "learning_rate": 2.487064008341018e-05, "loss": 0.0226, "step": 54740 }, { "grad_norm": 0.2794345021247864, "learning_rate": 2.4852773874485407e-05, "loss": 0.0224, "step": 54750 }, { "grad_norm": 0.3604079484939575, "learning_rate": 2.483491196253377e-05, "loss": 0.0255, "step": 54760 }, { "grad_norm": 0.34592384099960327, "learning_rate": 2.4817054350607443e-05, "loss": 0.0275, "step": 54770 }, { "grad_norm": 0.22927455604076385, "learning_rate": 2.4799201041757742e-05, "loss": 0.0183, "step": 54780 }, { "grad_norm": 0.6189382076263428, "learning_rate": 2.4781352039035373e-05, "loss": 0.0221, "step": 54790 }, { "grad_norm": 0.27856558561325073, "learning_rate": 2.4763507345490194e-05, "loss": 0.022, "step": 54800 }, { "grad_norm": 0.32907289266586304, "learning_rate": 2.4745666964171386e-05, "loss": 0.0255, "step": 54810 }, { "grad_norm": 0.38002344965934753, "learning_rate": 2.4727830898127387e-05, "loss": 0.0211, "step": 54820 }, { "grad_norm": 0.325495183467865, "learning_rate": 2.4709999150405895e-05, "loss": 0.0292, "step": 54830 }, { "grad_norm": 0.38688090443611145, "learning_rate": 2.469217172405388e-05, "loss": 0.0245, "step": 54840 }, { "grad_norm": 0.20987765491008759, "learning_rate": 2.4674348622117527e-05, "loss": 0.016, "step": 54850 }, { "grad_norm": 0.20756663382053375, "learning_rate": 2.465652984764234e-05, "loss": 0.0171, "step": 54860 }, { "grad_norm": 0.23408399522304535, "learning_rate": 2.4638715403673056e-05, "loss": 0.0189, "step": 54870 }, { "grad_norm": 0.4920370578765869, "learning_rate": 2.4620905293253676e-05, "loss": 0.0297, "step": 54880 }, { "grad_norm": 0.3180072009563446, "learning_rate": 2.4603099519427463e-05, "loss": 0.0264, "step": 54890 }, { "grad_norm": 0.32644736766815186, "learning_rate": 2.458529808523695e-05, "loss": 0.0207, "step": 54900 }, { "grad_norm": 0.2248375415802002, "learning_rate": 2.456750099372387e-05, "loss": 0.022, "step": 54910 }, { "grad_norm": 0.21537241339683533, "learning_rate": 2.4549708247929327e-05, "loss": 0.0287, "step": 54920 }, { "grad_norm": 0.2312149852514267, "learning_rate": 2.4531919850893554e-05, "loss": 0.0204, "step": 54930 }, { "grad_norm": 0.7572106719017029, "learning_rate": 2.4514135805656125e-05, "loss": 0.0227, "step": 54940 }, { "grad_norm": 0.3398599624633789, "learning_rate": 2.449635611525587e-05, "loss": 0.0263, "step": 54950 }, { "grad_norm": 0.3678204119205475, "learning_rate": 2.447858078273079e-05, "loss": 0.0278, "step": 54960 }, { "grad_norm": 0.4180872440338135, "learning_rate": 2.4460809811118275e-05, "loss": 0.0305, "step": 54970 }, { "grad_norm": 0.2538881301879883, "learning_rate": 2.444304320345483e-05, "loss": 0.0182, "step": 54980 }, { "grad_norm": 0.2797315716743469, "learning_rate": 2.4425280962776354e-05, "loss": 0.0216, "step": 54990 }, { "grad_norm": 0.24673841893672943, "learning_rate": 2.4407523092117875e-05, "loss": 0.026, "step": 55000 }, { "grad_norm": 0.28001272678375244, "learning_rate": 2.438976959451374e-05, "loss": 0.0201, "step": 55010 }, { "grad_norm": 0.27802249789237976, "learning_rate": 2.4372020472997565e-05, "loss": 0.0224, "step": 55020 }, { "grad_norm": 0.28574082255363464, "learning_rate": 2.4354275730602122e-05, "loss": 0.0201, "step": 55030 }, { "grad_norm": 0.2927297353744507, "learning_rate": 2.433653537035958e-05, "loss": 0.0239, "step": 55040 }, { "grad_norm": 0.2246994972229004, "learning_rate": 2.431879939530123e-05, "loss": 0.0206, "step": 55050 }, { "grad_norm": 0.303305447101593, "learning_rate": 2.4301067808457684e-05, "loss": 0.0218, "step": 55060 }, { "grad_norm": 0.24089279770851135, "learning_rate": 2.428334061285878e-05, "loss": 0.0243, "step": 55070 }, { "grad_norm": 0.3419538736343384, "learning_rate": 2.426561781153361e-05, "loss": 0.0217, "step": 55080 }, { "grad_norm": 0.30764704942703247, "learning_rate": 2.424789940751052e-05, "loss": 0.02, "step": 55090 }, { "grad_norm": 0.2058582603931427, "learning_rate": 2.423018540381712e-05, "loss": 0.0237, "step": 55100 }, { "grad_norm": 0.2074912190437317, "learning_rate": 2.421247580348021e-05, "loss": 0.0183, "step": 55110 }, { "grad_norm": 0.40000826120376587, "learning_rate": 2.4194770609525897e-05, "loss": 0.023, "step": 55120 }, { "grad_norm": 0.34834080934524536, "learning_rate": 2.417706982497951e-05, "loss": 0.025, "step": 55130 }, { "grad_norm": 0.7200471758842468, "learning_rate": 2.415937345286563e-05, "loss": 0.0289, "step": 55140 }, { "grad_norm": 0.6455214023590088, "learning_rate": 2.4141681496208087e-05, "loss": 0.0238, "step": 55150 }, { "grad_norm": 0.3874194324016571, "learning_rate": 2.4123993958029946e-05, "loss": 0.0243, "step": 55160 }, { "grad_norm": 0.29375407099723816, "learning_rate": 2.4106310841353548e-05, "loss": 0.0186, "step": 55170 }, { "grad_norm": 0.28627175092697144, "learning_rate": 2.4088632149200398e-05, "loss": 0.0288, "step": 55180 }, { "grad_norm": 0.2054237276315689, "learning_rate": 2.4070957884591367e-05, "loss": 0.0166, "step": 55190 }, { "grad_norm": 0.4184304177761078, "learning_rate": 2.4053288050546464e-05, "loss": 0.0237, "step": 55200 }, { "grad_norm": 0.1941237449645996, "learning_rate": 2.403562265008498e-05, "loss": 0.0225, "step": 55210 }, { "grad_norm": 0.35685697197914124, "learning_rate": 2.4017961686225483e-05, "loss": 0.0296, "step": 55220 }, { "grad_norm": 0.2015124261379242, "learning_rate": 2.400030516198568e-05, "loss": 0.0368, "step": 55230 }, { "grad_norm": 0.2617192268371582, "learning_rate": 2.3982653080382673e-05, "loss": 0.0236, "step": 55240 }, { "grad_norm": 0.4204629361629486, "learning_rate": 2.396500544443266e-05, "loss": 0.0223, "step": 55250 }, { "grad_norm": 0.43985703587532043, "learning_rate": 2.3947362257151156e-05, "loss": 0.0247, "step": 55260 }, { "grad_norm": 0.29712578654289246, "learning_rate": 2.39297235215529e-05, "loss": 0.0304, "step": 55270 }, { "grad_norm": 0.2981297969818115, "learning_rate": 2.3912089240651873e-05, "loss": 0.0227, "step": 55280 }, { "grad_norm": 0.215323343873024, "learning_rate": 2.389445941746129e-05, "loss": 0.0231, "step": 55290 }, { "grad_norm": 0.39545738697052, "learning_rate": 2.3876834054993625e-05, "loss": 0.0284, "step": 55300 }, { "grad_norm": 0.2973591387271881, "learning_rate": 2.3859213156260522e-05, "loss": 0.0239, "step": 55310 }, { "grad_norm": 0.3268504738807678, "learning_rate": 2.3841596724272948e-05, "loss": 0.027, "step": 55320 }, { "grad_norm": 0.27858731150627136, "learning_rate": 2.382398476204106e-05, "loss": 0.0234, "step": 55330 }, { "grad_norm": 0.46965107321739197, "learning_rate": 2.3806377272574254e-05, "loss": 0.023, "step": 55340 }, { "grad_norm": 0.3615536093711853, "learning_rate": 2.3788774258881174e-05, "loss": 0.0214, "step": 55350 }, { "grad_norm": 0.30876824259757996, "learning_rate": 2.37711757239697e-05, "loss": 0.0226, "step": 55360 }, { "grad_norm": 0.3528195321559906, "learning_rate": 2.3753581670846954e-05, "loss": 0.0268, "step": 55370 }, { "grad_norm": 0.22962477803230286, "learning_rate": 2.3735992102519216e-05, "loss": 0.0296, "step": 55380 }, { "grad_norm": 0.3595069646835327, "learning_rate": 2.371840702199215e-05, "loss": 0.0244, "step": 55390 }, { "grad_norm": 0.31991592049598694, "learning_rate": 2.3700826432270494e-05, "loss": 0.0248, "step": 55400 }, { "grad_norm": 0.3192961513996124, "learning_rate": 2.3683250336358326e-05, "loss": 0.0298, "step": 55410 }, { "grad_norm": 0.23291444778442383, "learning_rate": 2.3665678737258923e-05, "loss": 0.0269, "step": 55420 }, { "grad_norm": 0.34162667393684387, "learning_rate": 2.3648111637974745e-05, "loss": 0.0209, "step": 55430 }, { "grad_norm": 0.3979685604572296, "learning_rate": 2.36305490415076e-05, "loss": 0.0304, "step": 55440 }, { "grad_norm": 0.43497520685195923, "learning_rate": 2.3612990950858392e-05, "loss": 0.0199, "step": 55450 }, { "grad_norm": 0.3763565421104431, "learning_rate": 2.359543736902735e-05, "loss": 0.0224, "step": 55460 }, { "grad_norm": 0.4695974886417389, "learning_rate": 2.3577888299013896e-05, "loss": 0.0227, "step": 55470 }, { "grad_norm": 0.44887667894363403, "learning_rate": 2.3560343743816683e-05, "loss": 0.0228, "step": 55480 }, { "grad_norm": 0.19048452377319336, "learning_rate": 2.354280370643362e-05, "loss": 0.0262, "step": 55490 }, { "grad_norm": 0.313694030046463, "learning_rate": 2.3525268189861777e-05, "loss": 0.0206, "step": 55500 }, { "grad_norm": 0.3628353476524353, "learning_rate": 2.3507737197097513e-05, "loss": 0.0237, "step": 55510 }, { "grad_norm": 0.3688868582248688, "learning_rate": 2.34902107311364e-05, "loss": 0.0204, "step": 55520 }, { "grad_norm": 0.2963472604751587, "learning_rate": 2.3472688794973225e-05, "loss": 0.0258, "step": 55530 }, { "grad_norm": 0.32507526874542236, "learning_rate": 2.3455171391602016e-05, "loss": 0.0224, "step": 55540 }, { "grad_norm": 0.29643478989601135, "learning_rate": 2.343765852401601e-05, "loss": 0.0233, "step": 55550 }, { "grad_norm": 0.23703625798225403, "learning_rate": 2.342015019520768e-05, "loss": 0.0257, "step": 55560 }, { "grad_norm": 0.5857123136520386, "learning_rate": 2.3402646408168742e-05, "loss": 0.0201, "step": 55570 }, { "grad_norm": 0.4258562922477722, "learning_rate": 2.3385147165890074e-05, "loss": 0.0274, "step": 55580 }, { "grad_norm": 0.27191418409347534, "learning_rate": 2.336765247136184e-05, "loss": 0.0242, "step": 55590 }, { "grad_norm": 0.47380366921424866, "learning_rate": 2.33501623275734e-05, "loss": 0.0283, "step": 55600 }, { "grad_norm": 0.4509795010089874, "learning_rate": 2.333267673751334e-05, "loss": 0.0223, "step": 55610 }, { "grad_norm": 0.37458571791648865, "learning_rate": 2.33151957041695e-05, "loss": 0.0211, "step": 55620 }, { "grad_norm": 0.3578108549118042, "learning_rate": 2.329771923052884e-05, "loss": 0.022, "step": 55630 }, { "grad_norm": 0.4020440876483917, "learning_rate": 2.3280247319577697e-05, "loss": 0.022, "step": 55640 }, { "grad_norm": 0.24249736964702606, "learning_rate": 2.3262779974301473e-05, "loss": 0.0221, "step": 55650 }, { "grad_norm": 0.45715516805648804, "learning_rate": 2.3245317197684895e-05, "loss": 0.0244, "step": 55660 }, { "grad_norm": 0.2849610149860382, "learning_rate": 2.322785899271188e-05, "loss": 0.0197, "step": 55670 }, { "grad_norm": 0.7676052451133728, "learning_rate": 2.3210405362365507e-05, "loss": 0.027, "step": 55680 }, { "grad_norm": 0.3680586516857147, "learning_rate": 2.31929563096282e-05, "loss": 0.0287, "step": 55690 }, { "grad_norm": 0.29755476117134094, "learning_rate": 2.317551183748146e-05, "loss": 0.023, "step": 55700 }, { "grad_norm": 0.26960229873657227, "learning_rate": 2.3158071948906103e-05, "loss": 0.0282, "step": 55710 }, { "grad_norm": 0.32142266631126404, "learning_rate": 2.314063664688212e-05, "loss": 0.0257, "step": 55720 }, { "grad_norm": 0.365494042634964, "learning_rate": 2.3123205934388725e-05, "loss": 0.0232, "step": 55730 }, { "grad_norm": 0.31306907534599304, "learning_rate": 2.310577981440436e-05, "loss": 0.0327, "step": 55740 }, { "grad_norm": 0.33509811758995056, "learning_rate": 2.3088358289906665e-05, "loss": 0.0238, "step": 55750 }, { "grad_norm": 0.2381923347711563, "learning_rate": 2.307094136387252e-05, "loss": 0.0199, "step": 55760 }, { "grad_norm": 0.38027486205101013, "learning_rate": 2.305352903927796e-05, "loss": 0.0255, "step": 55770 }, { "grad_norm": 0.20876435935497284, "learning_rate": 2.303612131909831e-05, "loss": 0.0195, "step": 55780 }, { "grad_norm": 0.32101303339004517, "learning_rate": 2.3018718206308054e-05, "loss": 0.0251, "step": 55790 }, { "grad_norm": 0.4282403290271759, "learning_rate": 2.3001319703880925e-05, "loss": 0.0247, "step": 55800 }, { "grad_norm": 0.3105543553829193, "learning_rate": 2.2983925814789835e-05, "loss": 0.0223, "step": 55810 }, { "grad_norm": 0.2625584304332733, "learning_rate": 2.2966536542006957e-05, "loss": 0.026, "step": 55820 }, { "grad_norm": 0.24512144923210144, "learning_rate": 2.294915188850358e-05, "loss": 0.0274, "step": 55830 }, { "grad_norm": 0.2568272352218628, "learning_rate": 2.2931771857250333e-05, "loss": 0.0232, "step": 55840 }, { "grad_norm": 0.6634745597839355, "learning_rate": 2.2914396451216946e-05, "loss": 0.0207, "step": 55850 }, { "grad_norm": 0.4464957118034363, "learning_rate": 2.2897025673372412e-05, "loss": 0.0213, "step": 55860 }, { "grad_norm": 0.20061799883842468, "learning_rate": 2.287965952668494e-05, "loss": 0.0207, "step": 55870 }, { "grad_norm": 0.6774054765701294, "learning_rate": 2.2862298014121873e-05, "loss": 0.0264, "step": 55880 }, { "grad_norm": 0.4773190915584564, "learning_rate": 2.28449411386499e-05, "loss": 0.0273, "step": 55890 }, { "grad_norm": 0.26889917254447937, "learning_rate": 2.282758890323477e-05, "loss": 0.0226, "step": 55900 }, { "grad_norm": 0.17429211735725403, "learning_rate": 2.2810241310841528e-05, "loss": 0.0197, "step": 55910 }, { "grad_norm": 0.5073997974395752, "learning_rate": 2.2792898364434412e-05, "loss": 0.0251, "step": 55920 }, { "grad_norm": 0.375622421503067, "learning_rate": 2.2775560066976843e-05, "loss": 0.0227, "step": 55930 }, { "grad_norm": 0.3851422965526581, "learning_rate": 2.275822642143147e-05, "loss": 0.0263, "step": 55940 }, { "grad_norm": 0.27874094247817993, "learning_rate": 2.274089743076014e-05, "loss": 0.0247, "step": 55950 }, { "grad_norm": 0.425025999546051, "learning_rate": 2.2723573097923917e-05, "loss": 0.0215, "step": 55960 }, { "grad_norm": 0.28357771039009094, "learning_rate": 2.2706253425883024e-05, "loss": 0.0202, "step": 55970 }, { "grad_norm": 0.3941583037376404, "learning_rate": 2.2688938417596933e-05, "loss": 0.0229, "step": 55980 }, { "grad_norm": 0.3445571959018707, "learning_rate": 2.2671628076024305e-05, "loss": 0.0219, "step": 55990 }, { "grad_norm": 0.4507107734680176, "learning_rate": 2.2654322404123008e-05, "loss": 0.0202, "step": 56000 }, { "grad_norm": 0.252693235874176, "learning_rate": 2.2637021404850105e-05, "loss": 0.017, "step": 56010 }, { "grad_norm": 0.3547894060611725, "learning_rate": 2.2619725081161885e-05, "loss": 0.0233, "step": 56020 }, { "grad_norm": 0.22761858999729156, "learning_rate": 2.260243343601376e-05, "loss": 0.0225, "step": 56030 }, { "grad_norm": 0.28412824869155884, "learning_rate": 2.2585146472360473e-05, "loss": 0.0262, "step": 56040 }, { "grad_norm": 0.2599223554134369, "learning_rate": 2.2567864193155834e-05, "loss": 0.0221, "step": 56050 }, { "grad_norm": 0.33374401926994324, "learning_rate": 2.255058660135294e-05, "loss": 0.0307, "step": 56060 }, { "grad_norm": 0.45988160371780396, "learning_rate": 2.253331369990407e-05, "loss": 0.0265, "step": 56070 }, { "grad_norm": 0.14654682576656342, "learning_rate": 2.251604549176063e-05, "loss": 0.0292, "step": 56080 }, { "grad_norm": 0.24172960221767426, "learning_rate": 2.249878197987337e-05, "loss": 0.0224, "step": 56090 }, { "grad_norm": 0.22041042149066925, "learning_rate": 2.2481523167192087e-05, "loss": 0.0183, "step": 56100 }, { "grad_norm": 0.2862248122692108, "learning_rate": 2.2464269056665867e-05, "loss": 0.0183, "step": 56110 }, { "grad_norm": 0.23440192639827728, "learning_rate": 2.2447019651242958e-05, "loss": 0.0256, "step": 56120 }, { "grad_norm": 0.5867108106613159, "learning_rate": 2.2429774953870818e-05, "loss": 0.0246, "step": 56130 }, { "grad_norm": 0.3357658088207245, "learning_rate": 2.241253496749611e-05, "loss": 0.0226, "step": 56140 }, { "grad_norm": 0.2917572259902954, "learning_rate": 2.2395299695064614e-05, "loss": 0.0193, "step": 56150 }, { "grad_norm": 0.253776878118515, "learning_rate": 2.237806913952145e-05, "loss": 0.0216, "step": 56160 }, { "grad_norm": 0.26024720072746277, "learning_rate": 2.2360843303810798e-05, "loss": 0.0173, "step": 56170 }, { "grad_norm": 0.3335039019584656, "learning_rate": 2.2343622190876084e-05, "loss": 0.0192, "step": 56180 }, { "grad_norm": 0.357187956571579, "learning_rate": 2.2326405803659935e-05, "loss": 0.0237, "step": 56190 }, { "grad_norm": 0.2714084982872009, "learning_rate": 2.230919414510416e-05, "loss": 0.0255, "step": 56200 }, { "grad_norm": 0.38491174578666687, "learning_rate": 2.229198721814976e-05, "loss": 0.02, "step": 56210 }, { "grad_norm": 0.365596204996109, "learning_rate": 2.2274785025736948e-05, "loss": 0.0197, "step": 56220 }, { "grad_norm": 0.40989574790000916, "learning_rate": 2.225758757080507e-05, "loss": 0.0232, "step": 56230 }, { "grad_norm": 0.38211894035339355, "learning_rate": 2.2240394856292723e-05, "loss": 0.0228, "step": 56240 }, { "grad_norm": 0.3361327052116394, "learning_rate": 2.2223206885137664e-05, "loss": 0.0179, "step": 56250 }, { "grad_norm": 0.44732969999313354, "learning_rate": 2.2206023660276853e-05, "loss": 0.0263, "step": 56260 }, { "grad_norm": 0.2500164806842804, "learning_rate": 2.218884518464645e-05, "loss": 0.0218, "step": 56270 }, { "grad_norm": 0.3390917479991913, "learning_rate": 2.2171671461181732e-05, "loss": 0.0222, "step": 56280 }, { "grad_norm": 0.36214423179626465, "learning_rate": 2.2154502492817292e-05, "loss": 0.0248, "step": 56290 }, { "grad_norm": 0.25032591819763184, "learning_rate": 2.2137338282486782e-05, "loss": 0.0201, "step": 56300 }, { "grad_norm": 0.24417197704315186, "learning_rate": 2.2120178833123113e-05, "loss": 0.0207, "step": 56310 }, { "grad_norm": 0.3866821825504303, "learning_rate": 2.210302414765838e-05, "loss": 0.0214, "step": 56320 }, { "grad_norm": 0.25240325927734375, "learning_rate": 2.2085874229023808e-05, "loss": 0.0171, "step": 56330 }, { "grad_norm": 0.3492435812950134, "learning_rate": 2.2068729080149907e-05, "loss": 0.023, "step": 56340 }, { "grad_norm": 0.3725776970386505, "learning_rate": 2.205158870396625e-05, "loss": 0.0227, "step": 56350 }, { "grad_norm": 0.2696378529071808, "learning_rate": 2.2034453103401732e-05, "loss": 0.0261, "step": 56360 }, { "grad_norm": 0.6063113212585449, "learning_rate": 2.20173222813843e-05, "loss": 0.0185, "step": 56370 }, { "grad_norm": 0.28493353724479675, "learning_rate": 2.200019624084116e-05, "loss": 0.0248, "step": 56380 }, { "grad_norm": 0.35599425435066223, "learning_rate": 2.1983074984698687e-05, "loss": 0.0223, "step": 56390 }, { "grad_norm": 0.41411882638931274, "learning_rate": 2.1965958515882433e-05, "loss": 0.0302, "step": 56400 }, { "grad_norm": 0.30281949043273926, "learning_rate": 2.1948846837317162e-05, "loss": 0.0248, "step": 56410 }, { "grad_norm": 0.36371856927871704, "learning_rate": 2.1931739951926738e-05, "loss": 0.0187, "step": 56420 }, { "grad_norm": 0.3397684693336487, "learning_rate": 2.191463786263429e-05, "loss": 0.0165, "step": 56430 }, { "grad_norm": 0.3794100284576416, "learning_rate": 2.1897540572362095e-05, "loss": 0.0288, "step": 56440 }, { "grad_norm": 0.2670448124408722, "learning_rate": 2.1880448084031614e-05, "loss": 0.0253, "step": 56450 }, { "grad_norm": 0.1789930760860443, "learning_rate": 2.1863360400563482e-05, "loss": 0.0218, "step": 56460 }, { "grad_norm": 0.2724074125289917, "learning_rate": 2.184627752487754e-05, "loss": 0.0228, "step": 56470 }, { "grad_norm": 0.3632802367210388, "learning_rate": 2.1829199459892725e-05, "loss": 0.0213, "step": 56480 }, { "grad_norm": 0.2753942012786865, "learning_rate": 2.1812126208527282e-05, "loss": 0.0201, "step": 56490 }, { "grad_norm": 0.3315288722515106, "learning_rate": 2.179505777369852e-05, "loss": 0.0253, "step": 56500 }, { "grad_norm": 0.29652121663093567, "learning_rate": 2.1777994158322974e-05, "loss": 0.0212, "step": 56510 }, { "grad_norm": 0.2934688627719879, "learning_rate": 2.176093536531637e-05, "loss": 0.0272, "step": 56520 }, { "grad_norm": 0.7383435368537903, "learning_rate": 2.1743881397593537e-05, "loss": 0.023, "step": 56530 }, { "grad_norm": 0.39747607707977295, "learning_rate": 2.1726832258068595e-05, "loss": 0.0354, "step": 56540 }, { "grad_norm": 0.34760549664497375, "learning_rate": 2.170978794965472e-05, "loss": 0.023, "step": 56550 }, { "grad_norm": 0.3343811333179474, "learning_rate": 2.169274847526438e-05, "loss": 0.027, "step": 56560 }, { "grad_norm": 0.22711896896362305, "learning_rate": 2.1675713837809103e-05, "loss": 0.0227, "step": 56570 }, { "grad_norm": 0.3479638397693634, "learning_rate": 2.1658684040199655e-05, "loss": 0.0258, "step": 56580 }, { "grad_norm": 0.28819435834884644, "learning_rate": 2.1641659085345974e-05, "loss": 0.0177, "step": 56590 }, { "grad_norm": 0.44404760003089905, "learning_rate": 2.1624638976157154e-05, "loss": 0.0245, "step": 56600 }, { "grad_norm": 0.38712674379348755, "learning_rate": 2.1607623715541476e-05, "loss": 0.0288, "step": 56610 }, { "grad_norm": 0.49248263239860535, "learning_rate": 2.159061330640636e-05, "loss": 0.02, "step": 56620 }, { "grad_norm": 0.2691385746002197, "learning_rate": 2.1573607751658425e-05, "loss": 0.0174, "step": 56630 }, { "grad_norm": 0.2793791592121124, "learning_rate": 2.155660705420347e-05, "loss": 0.024, "step": 56640 }, { "grad_norm": 0.2803950309753418, "learning_rate": 2.153961121694644e-05, "loss": 0.0232, "step": 56650 }, { "grad_norm": 0.2685133218765259, "learning_rate": 2.152262024279145e-05, "loss": 0.0324, "step": 56660 }, { "grad_norm": 0.49916741251945496, "learning_rate": 2.150563413464183e-05, "loss": 0.0231, "step": 56670 }, { "grad_norm": 0.20232190191745758, "learning_rate": 2.1488652895399974e-05, "loss": 0.0189, "step": 56680 }, { "grad_norm": 0.3444764316082001, "learning_rate": 2.1471676527967587e-05, "loss": 0.0181, "step": 56690 }, { "grad_norm": 0.327396422624588, "learning_rate": 2.1454705035245414e-05, "loss": 0.0281, "step": 56700 }, { "grad_norm": 0.2784736156463623, "learning_rate": 2.143773842013343e-05, "loss": 0.0191, "step": 56710 }, { "grad_norm": 0.2582986056804657, "learning_rate": 2.1420776685530796e-05, "loss": 0.0194, "step": 56720 }, { "grad_norm": 0.30593451857566833, "learning_rate": 2.1403819834335742e-05, "loss": 0.0222, "step": 56730 }, { "grad_norm": 0.3418177664279938, "learning_rate": 2.1386867869445808e-05, "loss": 0.0212, "step": 56740 }, { "grad_norm": 0.3133464753627777, "learning_rate": 2.1369920793757548e-05, "loss": 0.0266, "step": 56750 }, { "grad_norm": 0.24980303645133972, "learning_rate": 2.1352978610166828e-05, "loss": 0.0165, "step": 56760 }, { "grad_norm": 0.3545105457305908, "learning_rate": 2.1336041321568546e-05, "loss": 0.0258, "step": 56770 }, { "grad_norm": 0.3165968954563141, "learning_rate": 2.131910893085684e-05, "loss": 0.0219, "step": 56780 }, { "grad_norm": 0.2728061079978943, "learning_rate": 2.1302181440925012e-05, "loss": 0.0209, "step": 56790 }, { "grad_norm": 0.26449939608573914, "learning_rate": 2.128525885466546e-05, "loss": 0.0209, "step": 56800 }, { "grad_norm": 0.310777485370636, "learning_rate": 2.1268341174969847e-05, "loss": 0.0307, "step": 56810 }, { "grad_norm": 0.21498043835163116, "learning_rate": 2.1251428404728894e-05, "loss": 0.0192, "step": 56820 }, { "grad_norm": 0.36163508892059326, "learning_rate": 2.123452054683256e-05, "loss": 0.0208, "step": 56830 }, { "grad_norm": 0.299816757440567, "learning_rate": 2.1217617604169927e-05, "loss": 0.0259, "step": 56840 }, { "grad_norm": 0.3918727934360504, "learning_rate": 2.120071957962924e-05, "loss": 0.0246, "step": 56850 }, { "grad_norm": 0.5053849816322327, "learning_rate": 2.1183826476097917e-05, "loss": 0.0295, "step": 56860 }, { "grad_norm": 0.19777289032936096, "learning_rate": 2.116693829646254e-05, "loss": 0.0209, "step": 56870 }, { "grad_norm": 0.3001774847507477, "learning_rate": 2.1150055043608806e-05, "loss": 0.0207, "step": 56880 }, { "grad_norm": 0.35063791275024414, "learning_rate": 2.113317672042162e-05, "loss": 0.0274, "step": 56890 }, { "grad_norm": 0.3358888030052185, "learning_rate": 2.1116303329785025e-05, "loss": 0.0207, "step": 56900 }, { "grad_norm": 0.2893292009830475, "learning_rate": 2.109943487458222e-05, "loss": 0.0273, "step": 56910 }, { "grad_norm": 0.41789042949676514, "learning_rate": 2.1082571357695574e-05, "loss": 0.0234, "step": 56920 }, { "grad_norm": 0.25268635153770447, "learning_rate": 2.1065712782006557e-05, "loss": 0.0232, "step": 56930 }, { "grad_norm": 0.21538333594799042, "learning_rate": 2.104885915039591e-05, "loss": 0.0218, "step": 56940 }, { "grad_norm": 0.21503040194511414, "learning_rate": 2.103201046574338e-05, "loss": 0.0182, "step": 56950 }, { "grad_norm": 0.5804761648178101, "learning_rate": 2.1015166730928022e-05, "loss": 0.0253, "step": 56960 }, { "grad_norm": 0.3001454174518585, "learning_rate": 2.0998327948827912e-05, "loss": 0.0244, "step": 56970 }, { "grad_norm": 0.29245033860206604, "learning_rate": 2.0981494122320363e-05, "loss": 0.0247, "step": 56980 }, { "grad_norm": 0.5368181467056274, "learning_rate": 2.0964665254281822e-05, "loss": 0.0252, "step": 56990 }, { "grad_norm": 0.2766213119029999, "learning_rate": 2.094784134758784e-05, "loss": 0.0273, "step": 57000 }, { "grad_norm": 0.29460465908050537, "learning_rate": 2.0931022405113226e-05, "loss": 0.0306, "step": 57010 }, { "grad_norm": 0.4553791880607605, "learning_rate": 2.091420842973183e-05, "loss": 0.032, "step": 57020 }, { "grad_norm": 0.25098371505737305, "learning_rate": 2.0897399424316715e-05, "loss": 0.0192, "step": 57030 }, { "grad_norm": 0.3128313720226288, "learning_rate": 2.0880595391740078e-05, "loss": 0.0244, "step": 57040 }, { "grad_norm": 0.2582782506942749, "learning_rate": 2.0863796334873277e-05, "loss": 0.0243, "step": 57050 }, { "grad_norm": 0.21847465634346008, "learning_rate": 2.08470022565868e-05, "loss": 0.0224, "step": 57060 }, { "grad_norm": 0.1948004961013794, "learning_rate": 2.0830213159750317e-05, "loss": 0.0161, "step": 57070 }, { "grad_norm": 0.3398032784461975, "learning_rate": 2.0813429047232596e-05, "loss": 0.018, "step": 57080 }, { "grad_norm": 0.27956005930900574, "learning_rate": 2.0796649921901594e-05, "loss": 0.0284, "step": 57090 }, { "grad_norm": 0.27718818187713623, "learning_rate": 2.077987578662441e-05, "loss": 0.0172, "step": 57100 }, { "grad_norm": 0.4486047327518463, "learning_rate": 2.0763106644267277e-05, "loss": 0.0242, "step": 57110 }, { "grad_norm": 0.193595752120018, "learning_rate": 2.0746342497695607e-05, "loss": 0.0149, "step": 57120 }, { "grad_norm": 0.28804028034210205, "learning_rate": 2.0729583349773886e-05, "loss": 0.0168, "step": 57130 }, { "grad_norm": 0.3621664047241211, "learning_rate": 2.0712829203365853e-05, "loss": 0.0212, "step": 57140 }, { "grad_norm": 0.26613783836364746, "learning_rate": 2.0696080061334267e-05, "loss": 0.0225, "step": 57150 }, { "grad_norm": 0.2448064535856247, "learning_rate": 2.067933592654117e-05, "loss": 0.0195, "step": 57160 }, { "grad_norm": 0.33873337507247925, "learning_rate": 2.066259680184763e-05, "loss": 0.0212, "step": 57170 }, { "grad_norm": 0.23642559349536896, "learning_rate": 2.0645862690113908e-05, "loss": 0.0169, "step": 57180 }, { "grad_norm": 0.3572443425655365, "learning_rate": 2.0629133594199436e-05, "loss": 0.0275, "step": 57190 }, { "grad_norm": 0.38754957914352417, "learning_rate": 2.0612409516962704e-05, "loss": 0.0196, "step": 57200 }, { "grad_norm": 0.26850426197052, "learning_rate": 2.0595690461261467e-05, "loss": 0.0233, "step": 57210 }, { "grad_norm": 0.3740499019622803, "learning_rate": 2.0578976429952503e-05, "loss": 0.025, "step": 57220 }, { "grad_norm": 0.25622475147247314, "learning_rate": 2.0562267425891802e-05, "loss": 0.024, "step": 57230 }, { "grad_norm": 0.42410939931869507, "learning_rate": 2.0545563451934467e-05, "loss": 0.0233, "step": 57240 }, { "grad_norm": 0.21454991400241852, "learning_rate": 2.0528864510934764e-05, "loss": 0.0255, "step": 57250 }, { "grad_norm": 0.24397024512290955, "learning_rate": 2.0512170605746096e-05, "loss": 0.0297, "step": 57260 }, { "grad_norm": 0.22454902529716492, "learning_rate": 2.0495481739220963e-05, "loss": 0.0211, "step": 57270 }, { "grad_norm": 0.3502456247806549, "learning_rate": 2.0478797914211045e-05, "loss": 0.0286, "step": 57280 }, { "grad_norm": 0.3367049992084503, "learning_rate": 2.046211913356716e-05, "loss": 0.0239, "step": 57290 }, { "grad_norm": 0.39878278970718384, "learning_rate": 2.0445445400139247e-05, "loss": 0.0268, "step": 57300 }, { "grad_norm": 0.6440215706825256, "learning_rate": 2.0428776716776405e-05, "loss": 0.0248, "step": 57310 }, { "grad_norm": 0.3746025562286377, "learning_rate": 2.0412113086326856e-05, "loss": 0.0207, "step": 57320 }, { "grad_norm": 0.23526543378829956, "learning_rate": 2.0395454511637918e-05, "loss": 0.0184, "step": 57330 }, { "grad_norm": 0.4131799042224884, "learning_rate": 2.037880099555616e-05, "loss": 0.0212, "step": 57340 }, { "grad_norm": 0.2816566228866577, "learning_rate": 2.0362152540927144e-05, "loss": 0.019, "step": 57350 }, { "grad_norm": 0.31564992666244507, "learning_rate": 2.0345509150595666e-05, "loss": 0.0352, "step": 57360 }, { "grad_norm": 0.2920887768268585, "learning_rate": 2.0328870827405617e-05, "loss": 0.0284, "step": 57370 }, { "grad_norm": 0.3052370548248291, "learning_rate": 2.0312237574200043e-05, "loss": 0.0222, "step": 57380 }, { "grad_norm": 0.5288344621658325, "learning_rate": 2.029560939382112e-05, "loss": 0.0209, "step": 57390 }, { "grad_norm": 0.31911563873291016, "learning_rate": 2.0278986289110097e-05, "loss": 0.0213, "step": 57400 }, { "grad_norm": 0.344774067401886, "learning_rate": 2.0262368262907484e-05, "loss": 0.0261, "step": 57410 }, { "grad_norm": 0.48791930079460144, "learning_rate": 2.024575531805279e-05, "loss": 0.0198, "step": 57420 }, { "grad_norm": 0.3294578790664673, "learning_rate": 2.0229147457384735e-05, "loss": 0.0157, "step": 57430 }, { "grad_norm": 0.28009936213493347, "learning_rate": 2.0212544683741157e-05, "loss": 0.0321, "step": 57440 }, { "grad_norm": 0.3589041531085968, "learning_rate": 2.0195946999958976e-05, "loss": 0.0302, "step": 57450 }, { "grad_norm": 0.24008411169052124, "learning_rate": 2.017935440887434e-05, "loss": 0.0249, "step": 57460 }, { "grad_norm": 0.42426517605781555, "learning_rate": 2.0162766913322423e-05, "loss": 0.0198, "step": 57470 }, { "grad_norm": 0.3528551757335663, "learning_rate": 2.0146184516137588e-05, "loss": 0.0252, "step": 57480 }, { "grad_norm": 0.32528141140937805, "learning_rate": 2.012960722015332e-05, "loss": 0.0291, "step": 57490 }, { "grad_norm": 0.26879942417144775, "learning_rate": 2.0113035028202214e-05, "loss": 0.0218, "step": 57500 }, { "grad_norm": 0.3031480610370636, "learning_rate": 2.009646794311602e-05, "loss": 0.0241, "step": 57510 }, { "grad_norm": 0.5018938779830933, "learning_rate": 2.007990596772559e-05, "loss": 0.0261, "step": 57520 }, { "grad_norm": 0.3486347496509552, "learning_rate": 2.0063349104860923e-05, "loss": 0.0197, "step": 57530 }, { "grad_norm": 0.3001502752304077, "learning_rate": 2.0046797357351116e-05, "loss": 0.0248, "step": 57540 }, { "grad_norm": 0.25080034136772156, "learning_rate": 2.0030250728024412e-05, "loss": 0.0235, "step": 57550 }, { "grad_norm": 0.7580573558807373, "learning_rate": 2.001370921970819e-05, "loss": 0.0208, "step": 57560 }, { "grad_norm": 0.3849342465400696, "learning_rate": 1.9997172835228932e-05, "loss": 0.0217, "step": 57570 }, { "grad_norm": 0.4434317648410797, "learning_rate": 1.9980641577412262e-05, "loss": 0.0244, "step": 57580 }, { "grad_norm": 0.25415322184562683, "learning_rate": 1.9964115449082925e-05, "loss": 0.0209, "step": 57590 }, { "grad_norm": 0.5379010438919067, "learning_rate": 1.9947594453064742e-05, "loss": 0.0327, "step": 57600 }, { "grad_norm": 0.3699694275856018, "learning_rate": 1.9931078592180774e-05, "loss": 0.0195, "step": 57610 }, { "grad_norm": 0.35265928506851196, "learning_rate": 1.9914567869253065e-05, "loss": 0.02, "step": 57620 }, { "grad_norm": 0.18746814131736755, "learning_rate": 1.989806228710287e-05, "loss": 0.0298, "step": 57630 }, { "grad_norm": 0.39912399649620056, "learning_rate": 1.9881561848550555e-05, "loss": 0.0201, "step": 57640 }, { "grad_norm": 0.24598583579063416, "learning_rate": 1.9865066556415544e-05, "loss": 0.0253, "step": 57650 }, { "grad_norm": 0.17866398394107819, "learning_rate": 1.98485764135165e-05, "loss": 0.0223, "step": 57660 }, { "grad_norm": 0.30739665031433105, "learning_rate": 1.983209142267109e-05, "loss": 0.0242, "step": 57670 }, { "grad_norm": 0.3881002366542816, "learning_rate": 1.9815611586696165e-05, "loss": 0.0216, "step": 57680 }, { "grad_norm": 0.2878772020339966, "learning_rate": 1.9799136908407667e-05, "loss": 0.0198, "step": 57690 }, { "grad_norm": 0.21050380170345306, "learning_rate": 1.9782667390620678e-05, "loss": 0.0252, "step": 57700 }, { "grad_norm": 0.3053979277610779, "learning_rate": 1.976620303614939e-05, "loss": 0.0166, "step": 57710 }, { "grad_norm": 0.3429754674434662, "learning_rate": 1.9749743847807108e-05, "loss": 0.017, "step": 57720 }, { "grad_norm": 0.39496949315071106, "learning_rate": 1.9733289828406272e-05, "loss": 0.0148, "step": 57730 }, { "grad_norm": 0.32566672563552856, "learning_rate": 1.9716840980758382e-05, "loss": 0.0275, "step": 57740 }, { "grad_norm": 0.30183714628219604, "learning_rate": 1.9700397307674134e-05, "loss": 0.0216, "step": 57750 }, { "grad_norm": 0.4021514654159546, "learning_rate": 1.968395881196328e-05, "loss": 0.0252, "step": 57760 }, { "grad_norm": 0.456143319606781, "learning_rate": 1.966752549643473e-05, "loss": 0.0214, "step": 57770 }, { "grad_norm": 0.5634891986846924, "learning_rate": 1.965109736389647e-05, "loss": 0.0187, "step": 57780 }, { "grad_norm": 0.3404306471347809, "learning_rate": 1.9634674417155645e-05, "loss": 0.019, "step": 57790 }, { "grad_norm": 0.3209132254123688, "learning_rate": 1.9618256659018434e-05, "loss": 0.036, "step": 57800 }, { "grad_norm": 0.5362692475318909, "learning_rate": 1.9601844092290257e-05, "loss": 0.0222, "step": 57810 }, { "grad_norm": 0.3806287348270416, "learning_rate": 1.9585436719775512e-05, "loss": 0.0216, "step": 57820 }, { "grad_norm": 0.3039809465408325, "learning_rate": 1.9569034544277793e-05, "loss": 0.024, "step": 57830 }, { "grad_norm": 0.2951682507991791, "learning_rate": 1.9552637568599798e-05, "loss": 0.0198, "step": 57840 }, { "grad_norm": 0.2721361815929413, "learning_rate": 1.953624579554327e-05, "loss": 0.0312, "step": 57850 }, { "grad_norm": 0.5256319046020508, "learning_rate": 1.951985922790918e-05, "loss": 0.0231, "step": 57860 }, { "grad_norm": 0.23669660091400146, "learning_rate": 1.9503477868497505e-05, "loss": 0.017, "step": 57870 }, { "grad_norm": 0.3885689675807953, "learning_rate": 1.9487101720107375e-05, "loss": 0.0224, "step": 57880 }, { "grad_norm": 0.3519076108932495, "learning_rate": 1.9470730785537032e-05, "loss": 0.023, "step": 57890 }, { "grad_norm": 0.17034506797790527, "learning_rate": 1.9454365067583823e-05, "loss": 0.0234, "step": 57900 }, { "grad_norm": 0.34967437386512756, "learning_rate": 1.9438004569044215e-05, "loss": 0.0215, "step": 57910 }, { "grad_norm": 0.24866820871829987, "learning_rate": 1.9421649292713724e-05, "loss": 0.0376, "step": 57920 }, { "grad_norm": 0.33942365646362305, "learning_rate": 1.9405299241387076e-05, "loss": 0.0238, "step": 57930 }, { "grad_norm": 0.22099505364894867, "learning_rate": 1.9388954417858007e-05, "loss": 0.016, "step": 57940 }, { "grad_norm": 0.4129413366317749, "learning_rate": 1.9372614824919417e-05, "loss": 0.0235, "step": 57950 }, { "grad_norm": 0.29573872685432434, "learning_rate": 1.9356280465363284e-05, "loss": 0.0287, "step": 57960 }, { "grad_norm": 0.36093801259994507, "learning_rate": 1.9339951341980723e-05, "loss": 0.0191, "step": 57970 }, { "grad_norm": 0.27984029054641724, "learning_rate": 1.9323627457561916e-05, "loss": 0.0214, "step": 57980 }, { "grad_norm": 0.236110657453537, "learning_rate": 1.9307308814896198e-05, "loss": 0.014, "step": 57990 }, { "grad_norm": 0.2440984845161438, "learning_rate": 1.9290995416771935e-05, "loss": 0.0174, "step": 58000 }, { "grad_norm": 0.3439417779445648, "learning_rate": 1.9274687265976665e-05, "loss": 0.0194, "step": 58010 }, { "grad_norm": 0.24384792149066925, "learning_rate": 1.9258384365297e-05, "loss": 0.0239, "step": 58020 }, { "grad_norm": 0.21644000709056854, "learning_rate": 1.924208671751866e-05, "loss": 0.0208, "step": 58030 }, { "grad_norm": 0.20729689300060272, "learning_rate": 1.9225794325426492e-05, "loss": 0.0174, "step": 58040 }, { "grad_norm": 0.3061050772666931, "learning_rate": 1.920950719180436e-05, "loss": 0.0226, "step": 58050 }, { "grad_norm": 0.18447090685367584, "learning_rate": 1.919322531943536e-05, "loss": 0.023, "step": 58060 }, { "grad_norm": 0.20199546217918396, "learning_rate": 1.917694871110157e-05, "loss": 0.0244, "step": 58070 }, { "grad_norm": 0.2834928035736084, "learning_rate": 1.9160677369584234e-05, "loss": 0.0218, "step": 58080 }, { "grad_norm": 0.2567550837993622, "learning_rate": 1.9144411297663694e-05, "loss": 0.0192, "step": 58090 }, { "grad_norm": 0.22533664107322693, "learning_rate": 1.9128150498119328e-05, "loss": 0.0215, "step": 58100 }, { "grad_norm": 0.20545272529125214, "learning_rate": 1.9111894973729726e-05, "loss": 0.0174, "step": 58110 }, { "grad_norm": 0.5474928021430969, "learning_rate": 1.9095644727272454e-05, "loss": 0.0238, "step": 58120 }, { "grad_norm": 0.5133388638496399, "learning_rate": 1.907939976152429e-05, "loss": 0.0255, "step": 58130 }, { "grad_norm": 0.4242402911186218, "learning_rate": 1.906316007926101e-05, "loss": 0.0222, "step": 58140 }, { "grad_norm": 0.17889098823070526, "learning_rate": 1.904692568325755e-05, "loss": 0.0198, "step": 58150 }, { "grad_norm": 0.3246012032032013, "learning_rate": 1.9030696576287925e-05, "loss": 0.0199, "step": 58160 }, { "grad_norm": 0.27605903148651123, "learning_rate": 1.9014472761125242e-05, "loss": 0.0198, "step": 58170 }, { "grad_norm": 0.1425829380750656, "learning_rate": 1.899825424054172e-05, "loss": 0.019, "step": 58180 }, { "grad_norm": 0.33461955189704895, "learning_rate": 1.898204101730863e-05, "loss": 0.0228, "step": 58190 }, { "grad_norm": 0.3570164442062378, "learning_rate": 1.8965833094196394e-05, "loss": 0.031, "step": 58200 }, { "grad_norm": 0.4904595911502838, "learning_rate": 1.8949630473974495e-05, "loss": 0.0231, "step": 58210 }, { "grad_norm": 0.2321893572807312, "learning_rate": 1.8933433159411517e-05, "loss": 0.0254, "step": 58220 }, { "grad_norm": 0.2387728989124298, "learning_rate": 1.891724115327514e-05, "loss": 0.0207, "step": 58230 }, { "grad_norm": 0.42605993151664734, "learning_rate": 1.8901054458332152e-05, "loss": 0.0271, "step": 58240 }, { "grad_norm": 0.33038467168807983, "learning_rate": 1.8884873077348364e-05, "loss": 0.0257, "step": 58250 }, { "grad_norm": 0.3987918496131897, "learning_rate": 1.88686970130888e-05, "loss": 0.0239, "step": 58260 }, { "grad_norm": 0.29436975717544556, "learning_rate": 1.8852526268317456e-05, "loss": 0.0261, "step": 58270 }, { "grad_norm": 0.27727609872817993, "learning_rate": 1.883636084579749e-05, "loss": 0.02, "step": 58280 }, { "grad_norm": 0.27312353253364563, "learning_rate": 1.8820200748291145e-05, "loss": 0.0249, "step": 58290 }, { "grad_norm": 0.2088889628648758, "learning_rate": 1.8804045978559686e-05, "loss": 0.0228, "step": 58300 }, { "grad_norm": 0.33664700388908386, "learning_rate": 1.8787896539363594e-05, "loss": 0.0206, "step": 58310 }, { "grad_norm": 0.23583534359931946, "learning_rate": 1.8771752433462298e-05, "loss": 0.0186, "step": 58320 }, { "grad_norm": 0.4020259380340576, "learning_rate": 1.8755613663614447e-05, "loss": 0.025, "step": 58330 }, { "grad_norm": 0.33288848400115967, "learning_rate": 1.873948023257767e-05, "loss": 0.021, "step": 58340 }, { "grad_norm": 0.238463893532753, "learning_rate": 1.8723352143108747e-05, "loss": 0.0193, "step": 58350 }, { "grad_norm": 0.29417940974235535, "learning_rate": 1.870722939796352e-05, "loss": 0.0211, "step": 58360 }, { "grad_norm": 0.2893313467502594, "learning_rate": 1.869111199989693e-05, "loss": 0.0234, "step": 58370 }, { "grad_norm": 0.3322242796421051, "learning_rate": 1.8674999951663018e-05, "loss": 0.0259, "step": 58380 }, { "grad_norm": 0.19706156849861145, "learning_rate": 1.8658893256014857e-05, "loss": 0.0234, "step": 58390 }, { "grad_norm": 0.4245797097682953, "learning_rate": 1.8642791915704655e-05, "loss": 0.0229, "step": 58400 }, { "grad_norm": 0.31386497616767883, "learning_rate": 1.8626695933483697e-05, "loss": 0.0168, "step": 58410 }, { "grad_norm": 0.2336893528699875, "learning_rate": 1.8610605312102335e-05, "loss": 0.0233, "step": 58420 }, { "grad_norm": 0.4000946283340454, "learning_rate": 1.8594520054310032e-05, "loss": 0.0269, "step": 58430 }, { "grad_norm": 0.4976254403591156, "learning_rate": 1.8578440162855326e-05, "loss": 0.0213, "step": 58440 }, { "grad_norm": 0.2819148302078247, "learning_rate": 1.8562365640485774e-05, "loss": 0.0235, "step": 58450 }, { "grad_norm": 0.3383188247680664, "learning_rate": 1.854629648994815e-05, "loss": 0.0213, "step": 58460 }, { "grad_norm": 0.14622840285301208, "learning_rate": 1.8530232713988183e-05, "loss": 0.0224, "step": 58470 }, { "grad_norm": 0.3963339924812317, "learning_rate": 1.851417431535074e-05, "loss": 0.0202, "step": 58480 }, { "grad_norm": 0.23057474195957184, "learning_rate": 1.8498121296779785e-05, "loss": 0.0322, "step": 58490 }, { "grad_norm": 0.28852134943008423, "learning_rate": 1.848207366101829e-05, "loss": 0.0261, "step": 58500 }, { "grad_norm": 0.4475004971027374, "learning_rate": 1.8466031410808422e-05, "loss": 0.0201, "step": 58510 }, { "grad_norm": 0.37896886467933655, "learning_rate": 1.844999454889129e-05, "loss": 0.0226, "step": 58520 }, { "grad_norm": 0.38659465312957764, "learning_rate": 1.8433963078007237e-05, "loss": 0.0294, "step": 58530 }, { "grad_norm": 0.18624629080295563, "learning_rate": 1.841793700089554e-05, "loss": 0.0234, "step": 58540 }, { "grad_norm": 0.37590479850769043, "learning_rate": 1.8401916320294644e-05, "loss": 0.021, "step": 58550 }, { "grad_norm": 0.3309749662876129, "learning_rate": 1.838590103894205e-05, "loss": 0.018, "step": 58560 }, { "grad_norm": 0.2676120698451996, "learning_rate": 1.8369891159574292e-05, "loss": 0.023, "step": 58570 }, { "grad_norm": 0.3466245234012604, "learning_rate": 1.835388668492708e-05, "loss": 0.0232, "step": 58580 }, { "grad_norm": 0.3319447338581085, "learning_rate": 1.8337887617735095e-05, "loss": 0.0225, "step": 58590 }, { "grad_norm": 0.19264665246009827, "learning_rate": 1.8321893960732157e-05, "loss": 0.0225, "step": 58600 }, { "grad_norm": 0.3924033045768738, "learning_rate": 1.8305905716651138e-05, "loss": 0.0206, "step": 58610 }, { "grad_norm": 0.2948078513145447, "learning_rate": 1.8289922888224e-05, "loss": 0.0242, "step": 58620 }, { "grad_norm": 0.3688456416130066, "learning_rate": 1.8273945478181765e-05, "loss": 0.0212, "step": 58630 }, { "grad_norm": 0.4890916049480438, "learning_rate": 1.8257973489254558e-05, "loss": 0.0188, "step": 58640 }, { "grad_norm": 0.2658434212207794, "learning_rate": 1.824200692417152e-05, "loss": 0.0192, "step": 58650 }, { "grad_norm": 0.24038492143154144, "learning_rate": 1.8226045785660912e-05, "loss": 0.0208, "step": 58660 }, { "grad_norm": 0.3849576413631439, "learning_rate": 1.821009007645006e-05, "loss": 0.0234, "step": 58670 }, { "grad_norm": 0.3876170217990875, "learning_rate": 1.8194139799265357e-05, "loss": 0.0247, "step": 58680 }, { "grad_norm": 0.2956200838088989, "learning_rate": 1.8178194956832295e-05, "loss": 0.0256, "step": 58690 }, { "grad_norm": 0.21328027546405792, "learning_rate": 1.8162255551875346e-05, "loss": 0.019, "step": 58700 }, { "grad_norm": 0.43501266837120056, "learning_rate": 1.8146321587118194e-05, "loss": 0.0278, "step": 58710 }, { "grad_norm": 0.338826447725296, "learning_rate": 1.8130393065283448e-05, "loss": 0.0197, "step": 58720 }, { "grad_norm": 0.29932549595832825, "learning_rate": 1.8114469989092925e-05, "loss": 0.0236, "step": 58730 }, { "grad_norm": 0.338694304227829, "learning_rate": 1.8098552361267397e-05, "loss": 0.0181, "step": 58740 }, { "grad_norm": 0.3291743993759155, "learning_rate": 1.8082640184526763e-05, "loss": 0.0185, "step": 58750 }, { "grad_norm": 0.27018222212791443, "learning_rate": 1.806673346158999e-05, "loss": 0.0195, "step": 58760 }, { "grad_norm": 0.29661989212036133, "learning_rate": 1.8050832195175067e-05, "loss": 0.0202, "step": 58770 }, { "grad_norm": 0.23704619705677032, "learning_rate": 1.8034936387999136e-05, "loss": 0.0212, "step": 58780 }, { "grad_norm": 0.26359063386917114, "learning_rate": 1.8019046042778315e-05, "loss": 0.0193, "step": 58790 }, { "grad_norm": 0.2676191031932831, "learning_rate": 1.800316116222785e-05, "loss": 0.024, "step": 58800 }, { "grad_norm": 0.3901877701282501, "learning_rate": 1.7987281749062018e-05, "loss": 0.0435, "step": 58810 }, { "grad_norm": 0.20742174983024597, "learning_rate": 1.7971407805994195e-05, "loss": 0.0255, "step": 58820 }, { "grad_norm": 0.8109086751937866, "learning_rate": 1.7955539335736787e-05, "loss": 0.0269, "step": 58830 }, { "grad_norm": 0.3249921202659607, "learning_rate": 1.7939676341001304e-05, "loss": 0.021, "step": 58840 }, { "grad_norm": 0.5790721774101257, "learning_rate": 1.7923818824498275e-05, "loss": 0.0203, "step": 58850 }, { "grad_norm": 0.37481677532196045, "learning_rate": 1.7907966788937315e-05, "loss": 0.026, "step": 58860 }, { "grad_norm": 0.22651952505111694, "learning_rate": 1.7892120237027116e-05, "loss": 0.0258, "step": 58870 }, { "grad_norm": 0.48490509390830994, "learning_rate": 1.7876279171475413e-05, "loss": 0.0222, "step": 58880 }, { "grad_norm": 0.25361117720603943, "learning_rate": 1.7860443594989028e-05, "loss": 0.0188, "step": 58890 }, { "grad_norm": 0.6043986678123474, "learning_rate": 1.784461351027379e-05, "loss": 0.0177, "step": 58900 }, { "grad_norm": 0.4574798047542572, "learning_rate": 1.7828788920034677e-05, "loss": 0.0291, "step": 58910 }, { "grad_norm": 0.5673665404319763, "learning_rate": 1.7812969826975623e-05, "loss": 0.0266, "step": 58920 }, { "grad_norm": 0.24458985030651093, "learning_rate": 1.7797156233799738e-05, "loss": 0.0235, "step": 58930 }, { "grad_norm": 0.41531628370285034, "learning_rate": 1.7781348143209094e-05, "loss": 0.0254, "step": 58940 }, { "grad_norm": 0.2895691394805908, "learning_rate": 1.7765545557904862e-05, "loss": 0.0298, "step": 58950 }, { "grad_norm": 0.44729843735694885, "learning_rate": 1.7749748480587302e-05, "loss": 0.0227, "step": 58960 }, { "grad_norm": 0.26066240668296814, "learning_rate": 1.773395691395564e-05, "loss": 0.0205, "step": 58970 }, { "grad_norm": 0.35718685388565063, "learning_rate": 1.7718170860708305e-05, "loss": 0.0216, "step": 58980 }, { "grad_norm": 0.513559877872467, "learning_rate": 1.770239032354264e-05, "loss": 0.0191, "step": 58990 }, { "grad_norm": 0.23912334442138672, "learning_rate": 1.7686615305155124e-05, "loss": 0.026, "step": 59000 }, { "grad_norm": 0.33205851912498474, "learning_rate": 1.767084580824128e-05, "loss": 0.0199, "step": 59010 }, { "grad_norm": 0.7430047392845154, "learning_rate": 1.765508183549569e-05, "loss": 0.0254, "step": 59020 }, { "grad_norm": 0.4046354591846466, "learning_rate": 1.763932338961199e-05, "loss": 0.0213, "step": 59030 }, { "grad_norm": 0.253825306892395, "learning_rate": 1.762357047328284e-05, "loss": 0.0244, "step": 59040 }, { "grad_norm": 0.2584270238876343, "learning_rate": 1.7607823089199997e-05, "loss": 0.0231, "step": 59050 }, { "grad_norm": 0.47649598121643066, "learning_rate": 1.7592081240054265e-05, "loss": 0.0193, "step": 59060 }, { "grad_norm": 0.29844507575035095, "learning_rate": 1.7576344928535488e-05, "loss": 0.0201, "step": 59070 }, { "grad_norm": 0.4067957103252411, "learning_rate": 1.7560614157332573e-05, "loss": 0.0231, "step": 59080 }, { "grad_norm": 0.31508925557136536, "learning_rate": 1.7544888929133495e-05, "loss": 0.0239, "step": 59090 }, { "grad_norm": 0.30979517102241516, "learning_rate": 1.752916924662522e-05, "loss": 0.0205, "step": 59100 }, { "grad_norm": 0.1803845465183258, "learning_rate": 1.751345511249387e-05, "loss": 0.0178, "step": 59110 }, { "grad_norm": 0.2544323205947876, "learning_rate": 1.749774652942452e-05, "loss": 0.02, "step": 59120 }, { "grad_norm": 0.2719653248786926, "learning_rate": 1.748204350010135e-05, "loss": 0.0232, "step": 59130 }, { "grad_norm": 0.3136404752731323, "learning_rate": 1.7466346027207574e-05, "loss": 0.0207, "step": 59140 }, { "grad_norm": 0.4261303246021271, "learning_rate": 1.745065411342547e-05, "loss": 0.0235, "step": 59150 }, { "grad_norm": 0.3792402148246765, "learning_rate": 1.7434967761436366e-05, "loss": 0.0227, "step": 59160 }, { "grad_norm": 0.4293364882469177, "learning_rate": 1.741928697392058e-05, "loss": 0.0232, "step": 59170 }, { "grad_norm": 0.371436208486557, "learning_rate": 1.7403611753557597e-05, "loss": 0.0214, "step": 59180 }, { "grad_norm": 0.5399484038352966, "learning_rate": 1.738794210302584e-05, "loss": 0.017, "step": 59190 }, { "grad_norm": 0.3814188539981842, "learning_rate": 1.7372278025002837e-05, "loss": 0.0186, "step": 59200 }, { "grad_norm": 0.3801649510860443, "learning_rate": 1.7356619522165164e-05, "loss": 0.0174, "step": 59210 }, { "grad_norm": 0.3586256504058838, "learning_rate": 1.7340966597188378e-05, "loss": 0.0209, "step": 59220 }, { "grad_norm": 0.14123013615608215, "learning_rate": 1.732531925274722e-05, "loss": 0.0187, "step": 59230 }, { "grad_norm": 0.28159299492836, "learning_rate": 1.7309677491515318e-05, "loss": 0.0179, "step": 59240 }, { "grad_norm": 0.35211509466171265, "learning_rate": 1.7294041316165455e-05, "loss": 0.0191, "step": 59250 }, { "grad_norm": 0.24859629571437836, "learning_rate": 1.727841072936942e-05, "loss": 0.0215, "step": 59260 }, { "grad_norm": 0.3527863919734955, "learning_rate": 1.7262785733798058e-05, "loss": 0.0202, "step": 59270 }, { "grad_norm": 0.30379557609558105, "learning_rate": 1.7247166332121246e-05, "loss": 0.0222, "step": 59280 }, { "grad_norm": 0.3050723075866699, "learning_rate": 1.7231552527007933e-05, "loss": 0.0225, "step": 59290 }, { "grad_norm": 0.2940429449081421, "learning_rate": 1.721594432112606e-05, "loss": 0.017, "step": 59300 }, { "grad_norm": 0.43545588850975037, "learning_rate": 1.7200341717142655e-05, "loss": 0.0241, "step": 59310 }, { "grad_norm": 0.38658958673477173, "learning_rate": 1.718474471772377e-05, "loss": 0.0221, "step": 59320 }, { "grad_norm": 0.3123478889465332, "learning_rate": 1.7169153325534528e-05, "loss": 0.0205, "step": 59330 }, { "grad_norm": 0.35571908950805664, "learning_rate": 1.7153567543239045e-05, "loss": 0.0221, "step": 59340 }, { "grad_norm": 0.4911543130874634, "learning_rate": 1.7137987373500525e-05, "loss": 0.022, "step": 59350 }, { "grad_norm": 0.2792973518371582, "learning_rate": 1.7122412818981198e-05, "loss": 0.0233, "step": 59360 }, { "grad_norm": 0.36194393038749695, "learning_rate": 1.7106843882342283e-05, "loss": 0.0199, "step": 59370 }, { "grad_norm": 0.25088462233543396, "learning_rate": 1.709128056624415e-05, "loss": 0.02, "step": 59380 }, { "grad_norm": 0.30857545137405396, "learning_rate": 1.70757228733461e-05, "loss": 0.0194, "step": 59390 }, { "grad_norm": 0.26210445165634155, "learning_rate": 1.706017080630653e-05, "loss": 0.0187, "step": 59400 }, { "grad_norm": 0.22790879011154175, "learning_rate": 1.7044624367782873e-05, "loss": 0.02, "step": 59410 }, { "grad_norm": 0.21230681240558624, "learning_rate": 1.7029083560431553e-05, "loss": 0.0175, "step": 59420 }, { "grad_norm": 0.38558465242385864, "learning_rate": 1.701354838690813e-05, "loss": 0.0203, "step": 59430 }, { "grad_norm": 0.2616972327232361, "learning_rate": 1.6998018849867087e-05, "loss": 0.019, "step": 59440 }, { "grad_norm": 0.3092726767063141, "learning_rate": 1.698249495196202e-05, "loss": 0.0204, "step": 59450 }, { "grad_norm": 0.43728020787239075, "learning_rate": 1.6966976695845528e-05, "loss": 0.0244, "step": 59460 }, { "grad_norm": 0.34592771530151367, "learning_rate": 1.6951464084169268e-05, "loss": 0.017, "step": 59470 }, { "grad_norm": 0.3967914879322052, "learning_rate": 1.6935957119583916e-05, "loss": 0.0256, "step": 59480 }, { "grad_norm": 0.29297125339508057, "learning_rate": 1.6920455804739205e-05, "loss": 0.0244, "step": 59490 }, { "grad_norm": 0.30795910954475403, "learning_rate": 1.6904960142283855e-05, "loss": 0.0221, "step": 59500 }, { "grad_norm": 0.3541252911090851, "learning_rate": 1.6889470134865666e-05, "loss": 0.0222, "step": 59510 }, { "grad_norm": 0.35644614696502686, "learning_rate": 1.687398578513145e-05, "loss": 0.0201, "step": 59520 }, { "grad_norm": 0.34316983819007874, "learning_rate": 1.6858507095727066e-05, "loss": 0.0167, "step": 59530 }, { "grad_norm": 0.436199426651001, "learning_rate": 1.6843034069297403e-05, "loss": 0.0245, "step": 59540 }, { "grad_norm": 0.3039008378982544, "learning_rate": 1.682756670848637e-05, "loss": 0.0286, "step": 59550 }, { "grad_norm": 0.2932259440422058, "learning_rate": 1.6812105015936937e-05, "loss": 0.019, "step": 59560 }, { "grad_norm": 0.29730015993118286, "learning_rate": 1.6796648994291027e-05, "loss": 0.027, "step": 59570 }, { "grad_norm": 0.3866438865661621, "learning_rate": 1.678119864618973e-05, "loss": 0.0172, "step": 59580 }, { "grad_norm": 0.32987499237060547, "learning_rate": 1.6765753974273023e-05, "loss": 0.0154, "step": 59590 }, { "grad_norm": 0.3681165874004364, "learning_rate": 1.675031498118001e-05, "loss": 0.0234, "step": 59600 }, { "grad_norm": 0.2692250609397888, "learning_rate": 1.6734881669548804e-05, "loss": 0.0225, "step": 59610 }, { "grad_norm": 0.27740776538848877, "learning_rate": 1.671945404201648e-05, "loss": 0.0295, "step": 59620 }, { "grad_norm": 0.27201899886131287, "learning_rate": 1.670403210121927e-05, "loss": 0.0189, "step": 59630 }, { "grad_norm": 0.3634907603263855, "learning_rate": 1.6688615849792312e-05, "loss": 0.0161, "step": 59640 }, { "grad_norm": 0.39575257897377014, "learning_rate": 1.6673205290369832e-05, "loss": 0.0227, "step": 59650 }, { "grad_norm": 0.31910985708236694, "learning_rate": 1.6657800425585086e-05, "loss": 0.0222, "step": 59660 }, { "grad_norm": 0.6039173007011414, "learning_rate": 1.6642401258070327e-05, "loss": 0.0297, "step": 59670 }, { "grad_norm": 0.305672287940979, "learning_rate": 1.6627007790456884e-05, "loss": 0.0269, "step": 59680 }, { "grad_norm": 0.13185471296310425, "learning_rate": 1.6611620025375037e-05, "loss": 0.0187, "step": 59690 }, { "grad_norm": 0.4010993540287018, "learning_rate": 1.6596237965454154e-05, "loss": 0.028, "step": 59700 }, { "grad_norm": 0.4046582281589508, "learning_rate": 1.6580861613322607e-05, "loss": 0.027, "step": 59710 }, { "grad_norm": 0.38850563764572144, "learning_rate": 1.6565490971607796e-05, "loss": 0.022, "step": 59720 }, { "grad_norm": 0.42017701268196106, "learning_rate": 1.655012604293614e-05, "loss": 0.025, "step": 59730 }, { "grad_norm": 0.3538277745246887, "learning_rate": 1.653476682993309e-05, "loss": 0.0239, "step": 59740 }, { "grad_norm": 0.25600481033325195, "learning_rate": 1.651941333522311e-05, "loss": 0.021, "step": 59750 }, { "grad_norm": 0.39820677042007446, "learning_rate": 1.6504065561429715e-05, "loss": 0.0242, "step": 59760 }, { "grad_norm": 0.3153803050518036, "learning_rate": 1.6488723511175385e-05, "loss": 0.0225, "step": 59770 }, { "grad_norm": 0.3101905584335327, "learning_rate": 1.6473387187081668e-05, "loss": 0.0209, "step": 59780 }, { "grad_norm": 0.36441847681999207, "learning_rate": 1.6458056591769123e-05, "loss": 0.0178, "step": 59790 }, { "grad_norm": 0.6439487338066101, "learning_rate": 1.6442731727857335e-05, "loss": 0.0209, "step": 59800 }, { "grad_norm": 0.26267924904823303, "learning_rate": 1.642741259796492e-05, "loss": 0.0175, "step": 59810 }, { "grad_norm": 0.2607787251472473, "learning_rate": 1.641209920470944e-05, "loss": 0.0197, "step": 59820 }, { "grad_norm": 0.44455277919769287, "learning_rate": 1.639679155070762e-05, "loss": 0.0224, "step": 59830 }, { "grad_norm": 0.3867541253566742, "learning_rate": 1.6381489638575048e-05, "loss": 0.0175, "step": 59840 }, { "grad_norm": 0.17133183777332306, "learning_rate": 1.636619347092643e-05, "loss": 0.0249, "step": 59850 }, { "grad_norm": 0.28183645009994507, "learning_rate": 1.6350903050375476e-05, "loss": 0.0192, "step": 59860 }, { "grad_norm": 0.3387426435947418, "learning_rate": 1.6335618379534856e-05, "loss": 0.0191, "step": 59870 }, { "grad_norm": 0.26473763585090637, "learning_rate": 1.6320339461016364e-05, "loss": 0.0197, "step": 59880 }, { "grad_norm": 0.3522094488143921, "learning_rate": 1.6305066297430687e-05, "loss": 0.0195, "step": 59890 }, { "grad_norm": 0.21884316205978394, "learning_rate": 1.6289798891387654e-05, "loss": 0.0216, "step": 59900 }, { "grad_norm": 0.34202730655670166, "learning_rate": 1.6274537245495995e-05, "loss": 0.0198, "step": 59910 }, { "grad_norm": 0.2040940225124359, "learning_rate": 1.6259281362363527e-05, "loss": 0.018, "step": 59920 }, { "grad_norm": 0.49175581336021423, "learning_rate": 1.6244031244597068e-05, "loss": 0.0281, "step": 59930 }, { "grad_norm": 0.23260968923568726, "learning_rate": 1.622878689480244e-05, "loss": 0.0226, "step": 59940 }, { "grad_norm": 0.24955470860004425, "learning_rate": 1.6213548315584498e-05, "loss": 0.0237, "step": 59950 }, { "grad_norm": 0.3548307716846466, "learning_rate": 1.6198315509547074e-05, "loss": 0.0242, "step": 59960 }, { "grad_norm": 0.3142443597316742, "learning_rate": 1.6183088479293056e-05, "loss": 0.0169, "step": 59970 }, { "grad_norm": 0.22430114448070526, "learning_rate": 1.6167867227424316e-05, "loss": 0.0211, "step": 59980 }, { "grad_norm": 0.42803841829299927, "learning_rate": 1.615265175654176e-05, "loss": 0.0208, "step": 59990 }, { "grad_norm": 0.4602993428707123, "learning_rate": 1.613744206924529e-05, "loss": 0.0278, "step": 60000 }, { "grad_norm": 0.3201342523097992, "learning_rate": 1.6122238168133845e-05, "loss": 0.0222, "step": 60010 }, { "grad_norm": 0.19819287955760956, "learning_rate": 1.6107040055805305e-05, "loss": 0.0225, "step": 60020 }, { "grad_norm": 0.3838360905647278, "learning_rate": 1.6091847734856673e-05, "loss": 0.0178, "step": 60030 }, { "grad_norm": 0.3388914167881012, "learning_rate": 1.6076661207883865e-05, "loss": 0.0298, "step": 60040 }, { "grad_norm": 0.2536170780658722, "learning_rate": 1.6061480477481848e-05, "loss": 0.0212, "step": 60050 }, { "grad_norm": 0.4295891523361206, "learning_rate": 1.604630554624461e-05, "loss": 0.0216, "step": 60060 }, { "grad_norm": 0.32583746314048767, "learning_rate": 1.603113641676509e-05, "loss": 0.0202, "step": 60070 }, { "grad_norm": 0.2688530385494232, "learning_rate": 1.6015973091635338e-05, "loss": 0.0183, "step": 60080 }, { "grad_norm": 0.13989777863025665, "learning_rate": 1.6000815573446288e-05, "loss": 0.0195, "step": 60090 }, { "grad_norm": 0.3427346646785736, "learning_rate": 1.5985663864788002e-05, "loss": 0.0288, "step": 60100 }, { "grad_norm": 0.3565884828567505, "learning_rate": 1.597051796824946e-05, "loss": 0.0253, "step": 60110 }, { "grad_norm": 0.2440202236175537, "learning_rate": 1.5955377886418682e-05, "loss": 0.023, "step": 60120 }, { "grad_norm": 0.3156742453575134, "learning_rate": 1.5940243621882704e-05, "loss": 0.0215, "step": 60130 }, { "grad_norm": 0.180843323469162, "learning_rate": 1.5925115177227555e-05, "loss": 0.0198, "step": 60140 }, { "grad_norm": 0.24577268958091736, "learning_rate": 1.5909992555038288e-05, "loss": 0.0221, "step": 60150 }, { "grad_norm": 0.8730204105377197, "learning_rate": 1.5894875757898912e-05, "loss": 0.0253, "step": 60160 }, { "grad_norm": 0.2726135849952698, "learning_rate": 1.58797647883925e-05, "loss": 0.024, "step": 60170 }, { "grad_norm": 0.29414626955986023, "learning_rate": 1.586465964910109e-05, "loss": 0.0248, "step": 60180 }, { "grad_norm": 0.30859851837158203, "learning_rate": 1.5849560342605734e-05, "loss": 0.0215, "step": 60190 }, { "grad_norm": 0.29941490292549133, "learning_rate": 1.583446687148651e-05, "loss": 0.0179, "step": 60200 }, { "grad_norm": 0.5375967025756836, "learning_rate": 1.581937923832248e-05, "loss": 0.0241, "step": 60210 }, { "grad_norm": 0.20960266888141632, "learning_rate": 1.5804297445691663e-05, "loss": 0.0208, "step": 60220 }, { "grad_norm": 0.3957342505455017, "learning_rate": 1.578922149617119e-05, "loss": 0.0254, "step": 60230 }, { "grad_norm": 0.35682010650634766, "learning_rate": 1.5774151392337084e-05, "loss": 0.0178, "step": 60240 }, { "grad_norm": 0.3287383019924164, "learning_rate": 1.575908713676442e-05, "loss": 0.0199, "step": 60250 }, { "grad_norm": 0.334352970123291, "learning_rate": 1.5744028732027293e-05, "loss": 0.0212, "step": 60260 }, { "grad_norm": 0.40991613268852234, "learning_rate": 1.5728976180698723e-05, "loss": 0.021, "step": 60270 }, { "grad_norm": 0.2669774889945984, "learning_rate": 1.5713929485350836e-05, "loss": 0.0233, "step": 60280 }, { "grad_norm": 0.16626904904842377, "learning_rate": 1.569888864855465e-05, "loss": 0.0229, "step": 60290 }, { "grad_norm": 0.24946001172065735, "learning_rate": 1.568385367288028e-05, "loss": 0.0217, "step": 60300 }, { "grad_norm": 0.31615498661994934, "learning_rate": 1.5668824560896755e-05, "loss": 0.0202, "step": 60310 }, { "grad_norm": 0.28303590416908264, "learning_rate": 1.5653801315172156e-05, "loss": 0.0178, "step": 60320 }, { "grad_norm": 0.23728522658348083, "learning_rate": 1.5638783938273554e-05, "loss": 0.0181, "step": 60330 }, { "grad_norm": 0.45550069212913513, "learning_rate": 1.5623772432766966e-05, "loss": 0.0325, "step": 60340 }, { "grad_norm": 0.18975865840911865, "learning_rate": 1.5608766801217507e-05, "loss": 0.0212, "step": 60350 }, { "grad_norm": 0.2303621619939804, "learning_rate": 1.5593767046189184e-05, "loss": 0.0203, "step": 60360 }, { "grad_norm": 0.25129467248916626, "learning_rate": 1.5578773170245064e-05, "loss": 0.0157, "step": 60370 }, { "grad_norm": 0.45248815417289734, "learning_rate": 1.5563785175947182e-05, "loss": 0.0246, "step": 60380 }, { "grad_norm": 0.17330119013786316, "learning_rate": 1.5548803065856582e-05, "loss": 0.0174, "step": 60390 }, { "grad_norm": 0.32482486963272095, "learning_rate": 1.55338268425333e-05, "loss": 0.0167, "step": 60400 }, { "grad_norm": 0.3289438784122467, "learning_rate": 1.5518856508536373e-05, "loss": 0.0172, "step": 60410 }, { "grad_norm": 0.4678795337677002, "learning_rate": 1.5503892066423786e-05, "loss": 0.0238, "step": 60420 }, { "grad_norm": 0.4004950523376465, "learning_rate": 1.548893351875258e-05, "loss": 0.0323, "step": 60430 }, { "grad_norm": 0.321065217256546, "learning_rate": 1.547398086807875e-05, "loss": 0.0225, "step": 60440 }, { "grad_norm": 0.46690285205841064, "learning_rate": 1.5459034116957305e-05, "loss": 0.0276, "step": 60450 }, { "grad_norm": 0.22190292179584503, "learning_rate": 1.544409326794225e-05, "loss": 0.0262, "step": 60460 }, { "grad_norm": 0.27461084723472595, "learning_rate": 1.542915832358651e-05, "loss": 0.0163, "step": 60470 }, { "grad_norm": 0.4243040680885315, "learning_rate": 1.541422928644214e-05, "loss": 0.0223, "step": 60480 }, { "grad_norm": 0.42906588315963745, "learning_rate": 1.5399306159060024e-05, "loss": 0.0303, "step": 60490 }, { "grad_norm": 0.13592040538787842, "learning_rate": 1.5384388943990185e-05, "loss": 0.0175, "step": 60500 }, { "grad_norm": 0.4118170440196991, "learning_rate": 1.5369477643781526e-05, "loss": 0.022, "step": 60510 }, { "grad_norm": 0.4479491114616394, "learning_rate": 1.5354572260981985e-05, "loss": 0.0253, "step": 60520 }, { "grad_norm": 0.48671072721481323, "learning_rate": 1.533967279813851e-05, "loss": 0.0207, "step": 60530 }, { "grad_norm": 0.2952896058559418, "learning_rate": 1.5324779257796956e-05, "loss": 0.0178, "step": 60540 }, { "grad_norm": 0.23504315316677094, "learning_rate": 1.5309891642502293e-05, "loss": 0.0195, "step": 60550 }, { "grad_norm": 0.1866922527551651, "learning_rate": 1.5295009954798357e-05, "loss": 0.015, "step": 60560 }, { "grad_norm": 0.3207535147666931, "learning_rate": 1.5280134197228036e-05, "loss": 0.0222, "step": 60570 }, { "grad_norm": 0.3848746716976166, "learning_rate": 1.526526437233319e-05, "loss": 0.0205, "step": 60580 }, { "grad_norm": 0.19920994341373444, "learning_rate": 1.525040048265466e-05, "loss": 0.0199, "step": 60590 }, { "grad_norm": 0.24214553833007812, "learning_rate": 1.5235542530732288e-05, "loss": 0.0164, "step": 60600 }, { "grad_norm": 0.2600077986717224, "learning_rate": 1.5220690519104901e-05, "loss": 0.0277, "step": 60610 }, { "grad_norm": 0.5355350971221924, "learning_rate": 1.5205844450310275e-05, "loss": 0.023, "step": 60620 }, { "grad_norm": 0.5180989503860474, "learning_rate": 1.519100432688521e-05, "loss": 0.0187, "step": 60630 }, { "grad_norm": 0.35215118527412415, "learning_rate": 1.5176170151365476e-05, "loss": 0.0244, "step": 60640 }, { "grad_norm": 0.3012922406196594, "learning_rate": 1.5161341926285833e-05, "loss": 0.0172, "step": 60650 }, { "grad_norm": 0.2126961648464203, "learning_rate": 1.5146519654180025e-05, "loss": 0.0218, "step": 60660 }, { "grad_norm": 0.2737146019935608, "learning_rate": 1.5131703337580739e-05, "loss": 0.0235, "step": 60670 }, { "grad_norm": 0.22605623304843903, "learning_rate": 1.5116892979019731e-05, "loss": 0.0219, "step": 60680 }, { "grad_norm": 0.20962318778038025, "learning_rate": 1.5102088581027623e-05, "loss": 0.017, "step": 60690 }, { "grad_norm": 0.3640688359737396, "learning_rate": 1.5087290146134154e-05, "loss": 0.0206, "step": 60700 }, { "grad_norm": 0.561966061592102, "learning_rate": 1.5072497676867915e-05, "loss": 0.0179, "step": 60710 }, { "grad_norm": 0.2869585454463959, "learning_rate": 1.505771117575655e-05, "loss": 0.0183, "step": 60720 }, { "grad_norm": 0.2725394368171692, "learning_rate": 1.5042930645326691e-05, "loss": 0.0163, "step": 60730 }, { "grad_norm": 0.5607691407203674, "learning_rate": 1.5028156088103879e-05, "loss": 0.0288, "step": 60740 }, { "grad_norm": 0.5025760531425476, "learning_rate": 1.5013387506612735e-05, "loss": 0.0259, "step": 60750 }, { "grad_norm": 0.31459787487983704, "learning_rate": 1.4998624903376767e-05, "loss": 0.0244, "step": 60760 }, { "grad_norm": 0.4789130687713623, "learning_rate": 1.4983868280918517e-05, "loss": 0.0194, "step": 60770 }, { "grad_norm": 0.3748418688774109, "learning_rate": 1.4969117641759478e-05, "loss": 0.025, "step": 60780 }, { "grad_norm": 0.34579890966415405, "learning_rate": 1.495437298842014e-05, "loss": 0.0226, "step": 60790 }, { "grad_norm": 0.20369777083396912, "learning_rate": 1.4939634323419976e-05, "loss": 0.0216, "step": 60800 }, { "grad_norm": 0.28656479716300964, "learning_rate": 1.4924901649277384e-05, "loss": 0.0195, "step": 60810 }, { "grad_norm": 0.3080810308456421, "learning_rate": 1.4910174968509793e-05, "loss": 0.0207, "step": 60820 }, { "grad_norm": 0.3236381709575653, "learning_rate": 1.4895454283633598e-05, "loss": 0.0214, "step": 60830 }, { "grad_norm": 0.3320329189300537, "learning_rate": 1.488073959716415e-05, "loss": 0.0174, "step": 60840 }, { "grad_norm": 0.324425607919693, "learning_rate": 1.4866030911615791e-05, "loss": 0.0264, "step": 60850 }, { "grad_norm": 0.28726086020469666, "learning_rate": 1.4851328229501849e-05, "loss": 0.0219, "step": 60860 }, { "grad_norm": 0.30060186982154846, "learning_rate": 1.4836631553334562e-05, "loss": 0.0225, "step": 60870 }, { "grad_norm": 0.35492631793022156, "learning_rate": 1.4821940885625251e-05, "loss": 0.0194, "step": 60880 }, { "grad_norm": 0.2812405228614807, "learning_rate": 1.4807256228884109e-05, "loss": 0.0195, "step": 60890 }, { "grad_norm": 0.3037707209587097, "learning_rate": 1.4792577585620353e-05, "loss": 0.0191, "step": 60900 }, { "grad_norm": 0.28334060311317444, "learning_rate": 1.4777904958342164e-05, "loss": 0.0241, "step": 60910 }, { "grad_norm": 0.21822912991046906, "learning_rate": 1.4763238349556691e-05, "loss": 0.0211, "step": 60920 }, { "grad_norm": 0.29316481947898865, "learning_rate": 1.4748577761770072e-05, "loss": 0.0268, "step": 60930 }, { "grad_norm": 0.259811133146286, "learning_rate": 1.4733923197487354e-05, "loss": 0.0215, "step": 60940 }, { "grad_norm": 0.34644564986228943, "learning_rate": 1.4719274659212662e-05, "loss": 0.0299, "step": 60950 }, { "grad_norm": 0.3110404908657074, "learning_rate": 1.4704632149448983e-05, "loss": 0.0188, "step": 60960 }, { "grad_norm": 0.3427475690841675, "learning_rate": 1.4689995670698343e-05, "loss": 0.0308, "step": 60970 }, { "grad_norm": 0.2005738466978073, "learning_rate": 1.4675365225461728e-05, "loss": 0.0256, "step": 60980 }, { "grad_norm": 0.4709676504135132, "learning_rate": 1.4660740816239032e-05, "loss": 0.0167, "step": 60990 }, { "grad_norm": 0.3541196584701538, "learning_rate": 1.464612244552923e-05, "loss": 0.0192, "step": 61000 }, { "grad_norm": 0.48939695954322815, "learning_rate": 1.4631510115830161e-05, "loss": 0.0175, "step": 61010 }, { "grad_norm": 0.24461695551872253, "learning_rate": 1.4616903829638679e-05, "loss": 0.013, "step": 61020 }, { "grad_norm": 0.20404866337776184, "learning_rate": 1.46023035894506e-05, "loss": 0.0203, "step": 61030 }, { "grad_norm": 0.38340768218040466, "learning_rate": 1.4587709397760713e-05, "loss": 0.0231, "step": 61040 }, { "grad_norm": 0.46903282403945923, "learning_rate": 1.4573121257062755e-05, "loss": 0.0229, "step": 61050 }, { "grad_norm": 0.4312417507171631, "learning_rate": 1.4558539169849472e-05, "loss": 0.0278, "step": 61060 }, { "grad_norm": 0.2835380434989929, "learning_rate": 1.4543963138612499e-05, "loss": 0.0203, "step": 61070 }, { "grad_norm": 0.2967749834060669, "learning_rate": 1.4529393165842498e-05, "loss": 0.0184, "step": 61080 }, { "grad_norm": 0.3859702944755554, "learning_rate": 1.4514829254029084e-05, "loss": 0.0241, "step": 61090 }, { "grad_norm": 0.4357488453388214, "learning_rate": 1.450027140566083e-05, "loss": 0.0198, "step": 61100 }, { "grad_norm": 0.250780314207077, "learning_rate": 1.4485719623225268e-05, "loss": 0.0179, "step": 61110 }, { "grad_norm": 0.2563796043395996, "learning_rate": 1.4471173909208912e-05, "loss": 0.0204, "step": 61120 }, { "grad_norm": 0.25614893436431885, "learning_rate": 1.4456634266097236e-05, "loss": 0.0189, "step": 61130 }, { "grad_norm": 0.19187040627002716, "learning_rate": 1.444210069637461e-05, "loss": 0.029, "step": 61140 }, { "grad_norm": 0.15848927199840546, "learning_rate": 1.4427573202524502e-05, "loss": 0.0182, "step": 61150 }, { "grad_norm": 0.21598131954669952, "learning_rate": 1.4413051787029208e-05, "loss": 0.0211, "step": 61160 }, { "grad_norm": 0.28914907574653625, "learning_rate": 1.439853645237006e-05, "loss": 0.0232, "step": 61170 }, { "grad_norm": 0.38103386759757996, "learning_rate": 1.4384027201027344e-05, "loss": 0.0195, "step": 61180 }, { "grad_norm": 0.2966947853565216, "learning_rate": 1.4369524035480253e-05, "loss": 0.0198, "step": 61190 }, { "grad_norm": 0.25222206115722656, "learning_rate": 1.4355026958207035e-05, "loss": 0.0194, "step": 61200 }, { "grad_norm": 0.5703363418579102, "learning_rate": 1.434053597168481e-05, "loss": 0.0199, "step": 61210 }, { "grad_norm": 0.3066975474357605, "learning_rate": 1.432605107838969e-05, "loss": 0.0204, "step": 61220 }, { "grad_norm": 0.39557746052742004, "learning_rate": 1.4311572280796759e-05, "loss": 0.0209, "step": 61230 }, { "grad_norm": 0.3561338782310486, "learning_rate": 1.4297099581380047e-05, "loss": 0.027, "step": 61240 }, { "grad_norm": 0.31030458211898804, "learning_rate": 1.4282632982612538e-05, "loss": 0.0235, "step": 61250 }, { "grad_norm": 0.29001569747924805, "learning_rate": 1.4268172486966197e-05, "loss": 0.0256, "step": 61260 }, { "grad_norm": 0.3004131019115448, "learning_rate": 1.4253718096911894e-05, "loss": 0.0207, "step": 61270 }, { "grad_norm": 0.3881623148918152, "learning_rate": 1.4239269814919504e-05, "loss": 0.0255, "step": 61280 }, { "grad_norm": 0.3117944896221161, "learning_rate": 1.4224827643457843e-05, "loss": 0.0256, "step": 61290 }, { "grad_norm": 0.278139591217041, "learning_rate": 1.4210391584994686e-05, "loss": 0.0175, "step": 61300 }, { "grad_norm": 0.2626122832298279, "learning_rate": 1.4195961641996763e-05, "loss": 0.0207, "step": 61310 }, { "grad_norm": 0.2816338539123535, "learning_rate": 1.4181537816929751e-05, "loss": 0.0223, "step": 61320 }, { "grad_norm": 0.23705831170082092, "learning_rate": 1.4167120112258303e-05, "loss": 0.0226, "step": 61330 }, { "grad_norm": 0.30557915568351746, "learning_rate": 1.4152708530445963e-05, "loss": 0.0177, "step": 61340 }, { "grad_norm": 0.344985693693161, "learning_rate": 1.4138303073955344e-05, "loss": 0.0161, "step": 61350 }, { "grad_norm": 0.31205424666404724, "learning_rate": 1.4123903745247897e-05, "loss": 0.0203, "step": 61360 }, { "grad_norm": 0.29332414269447327, "learning_rate": 1.4109510546784082e-05, "loss": 0.0155, "step": 61370 }, { "grad_norm": 0.20365118980407715, "learning_rate": 1.4095123481023325e-05, "loss": 0.0234, "step": 61380 }, { "grad_norm": 0.24298259615898132, "learning_rate": 1.4080742550423936e-05, "loss": 0.0218, "step": 61390 }, { "grad_norm": 0.3207651674747467, "learning_rate": 1.4066367757443278e-05, "loss": 0.0171, "step": 61400 }, { "grad_norm": 0.37264880537986755, "learning_rate": 1.4051999104537572e-05, "loss": 0.0239, "step": 61410 }, { "grad_norm": 0.4420812129974365, "learning_rate": 1.4037636594162034e-05, "loss": 0.0231, "step": 61420 }, { "grad_norm": 0.7784394025802612, "learning_rate": 1.402328022877083e-05, "loss": 0.0196, "step": 61430 }, { "grad_norm": 0.27606919407844543, "learning_rate": 1.4008930010817073e-05, "loss": 0.0211, "step": 61440 }, { "grad_norm": 0.2587326467037201, "learning_rate": 1.3994585942752831e-05, "loss": 0.0181, "step": 61450 }, { "grad_norm": 0.390654057264328, "learning_rate": 1.398024802702909e-05, "loss": 0.0183, "step": 61460 }, { "grad_norm": 0.33471935987472534, "learning_rate": 1.3965916266095819e-05, "loss": 0.022, "step": 61470 }, { "grad_norm": 0.39076656103134155, "learning_rate": 1.395159066240192e-05, "loss": 0.0221, "step": 61480 }, { "grad_norm": 0.3492795526981354, "learning_rate": 1.3937271218395248e-05, "loss": 0.0176, "step": 61490 }, { "grad_norm": 0.4074450731277466, "learning_rate": 1.3922957936522613e-05, "loss": 0.0196, "step": 61500 }, { "grad_norm": 0.39279648661613464, "learning_rate": 1.3908650819229758e-05, "loss": 0.0229, "step": 61510 }, { "grad_norm": 0.4218583405017853, "learning_rate": 1.389434986896137e-05, "loss": 0.0178, "step": 61520 }, { "grad_norm": 0.2907707393169403, "learning_rate": 1.388005508816112e-05, "loss": 0.0164, "step": 61530 }, { "grad_norm": 0.3338853418827057, "learning_rate": 1.3865766479271557e-05, "loss": 0.0252, "step": 61540 }, { "grad_norm": 0.47367605566978455, "learning_rate": 1.3851484044734225e-05, "loss": 0.0295, "step": 61550 }, { "grad_norm": 0.2721937596797943, "learning_rate": 1.3837207786989608e-05, "loss": 0.0286, "step": 61560 }, { "grad_norm": 0.2845126688480377, "learning_rate": 1.382293770847713e-05, "loss": 0.0215, "step": 61570 }, { "grad_norm": 0.47224634885787964, "learning_rate": 1.3808673811635159e-05, "loss": 0.0206, "step": 61580 }, { "grad_norm": 0.32727307081222534, "learning_rate": 1.3794416098900975e-05, "loss": 0.028, "step": 61590 }, { "grad_norm": 0.30848973989486694, "learning_rate": 1.3780164572710886e-05, "loss": 0.0228, "step": 61600 }, { "grad_norm": 0.2542398273944855, "learning_rate": 1.3765919235500035e-05, "loss": 0.0237, "step": 61610 }, { "grad_norm": 0.33701080083847046, "learning_rate": 1.375168008970259e-05, "loss": 0.0199, "step": 61620 }, { "grad_norm": 0.16114453971385956, "learning_rate": 1.3737447137751635e-05, "loss": 0.0131, "step": 61630 }, { "grad_norm": 0.31824883818626404, "learning_rate": 1.372322038207915e-05, "loss": 0.0159, "step": 61640 }, { "grad_norm": 0.38873931765556335, "learning_rate": 1.3708999825116159e-05, "loss": 0.0298, "step": 61650 }, { "grad_norm": 0.28242743015289307, "learning_rate": 1.3694785469292526e-05, "loss": 0.0262, "step": 61660 }, { "grad_norm": 0.41420143842697144, "learning_rate": 1.3680577317037101e-05, "loss": 0.0175, "step": 61670 }, { "grad_norm": 0.28402817249298096, "learning_rate": 1.3666375370777678e-05, "loss": 0.0263, "step": 61680 }, { "grad_norm": 0.253918319940567, "learning_rate": 1.3652179632940981e-05, "loss": 0.0176, "step": 61690 }, { "grad_norm": 0.2295607626438141, "learning_rate": 1.3637990105952664e-05, "loss": 0.0209, "step": 61700 }, { "grad_norm": 0.22144678235054016, "learning_rate": 1.3623806792237337e-05, "loss": 0.0152, "step": 61710 }, { "grad_norm": 0.30803054571151733, "learning_rate": 1.3609629694218551e-05, "loss": 0.0178, "step": 61720 }, { "grad_norm": 0.33024024963378906, "learning_rate": 1.3595458814318762e-05, "loss": 0.0185, "step": 61730 }, { "grad_norm": 0.3230046331882477, "learning_rate": 1.3581294154959389e-05, "loss": 0.0206, "step": 61740 }, { "grad_norm": 0.4094984233379364, "learning_rate": 1.3567135718560792e-05, "loss": 0.0177, "step": 61750 }, { "grad_norm": 0.1840570867061615, "learning_rate": 1.3552983507542261e-05, "loss": 0.0215, "step": 61760 }, { "grad_norm": 0.24664826691150665, "learning_rate": 1.3538837524322023e-05, "loss": 0.0266, "step": 61770 }, { "grad_norm": 0.09886405616998672, "learning_rate": 1.3524697771317251e-05, "loss": 0.0297, "step": 61780 }, { "grad_norm": 0.24955430626869202, "learning_rate": 1.3510564250943997e-05, "loss": 0.0191, "step": 61790 }, { "grad_norm": 0.42379486560821533, "learning_rate": 1.3496436965617353e-05, "loss": 0.0211, "step": 61800 }, { "grad_norm": 0.2620012164115906, "learning_rate": 1.3482315917751243e-05, "loss": 0.0247, "step": 61810 }, { "grad_norm": 0.16229292750358582, "learning_rate": 1.3468201109758582e-05, "loss": 0.0155, "step": 61820 }, { "grad_norm": 0.5717675685882568, "learning_rate": 1.3454092544051222e-05, "loss": 0.0227, "step": 61830 }, { "grad_norm": 0.21867823600769043, "learning_rate": 1.3439990223039878e-05, "loss": 0.0209, "step": 61840 }, { "grad_norm": 0.44398200511932373, "learning_rate": 1.3425894149134321e-05, "loss": 0.0177, "step": 61850 }, { "grad_norm": 0.25186464190483093, "learning_rate": 1.3411804324743132e-05, "loss": 0.0199, "step": 61860 }, { "grad_norm": 0.27138614654541016, "learning_rate": 1.3397720752273896e-05, "loss": 0.0198, "step": 61870 }, { "grad_norm": 0.19949646294116974, "learning_rate": 1.3383643434133108e-05, "loss": 0.0214, "step": 61880 }, { "grad_norm": 0.27205023169517517, "learning_rate": 1.3369572372726196e-05, "loss": 0.0216, "step": 61890 }, { "grad_norm": 0.17269369959831238, "learning_rate": 1.3355507570457525e-05, "loss": 0.0173, "step": 61900 }, { "grad_norm": 0.42440950870513916, "learning_rate": 1.334144902973038e-05, "loss": 0.0158, "step": 61910 }, { "grad_norm": 0.2684394419193268, "learning_rate": 1.3327396752946997e-05, "loss": 0.0219, "step": 61920 }, { "grad_norm": 0.2894410789012909, "learning_rate": 1.3313350742508496e-05, "loss": 0.0172, "step": 61930 }, { "grad_norm": 0.4904462397098541, "learning_rate": 1.3299311000814973e-05, "loss": 0.029, "step": 61940 }, { "grad_norm": 0.18122565746307373, "learning_rate": 1.3285277530265432e-05, "loss": 0.0199, "step": 61950 }, { "grad_norm": 0.3617701232433319, "learning_rate": 1.3271250333257813e-05, "loss": 0.0237, "step": 61960 }, { "grad_norm": 0.18278314173221588, "learning_rate": 1.3257229412188977e-05, "loss": 0.0197, "step": 61970 }, { "grad_norm": 0.29037556052207947, "learning_rate": 1.3243214769454726e-05, "loss": 0.0226, "step": 61980 }, { "grad_norm": 0.43639686703681946, "learning_rate": 1.3229206407449751e-05, "loss": 0.0203, "step": 61990 }, { "grad_norm": 0.5424256324768066, "learning_rate": 1.3215204328567742e-05, "loss": 0.0233, "step": 62000 }, { "grad_norm": 0.22858652472496033, "learning_rate": 1.3201208535201232e-05, "loss": 0.0229, "step": 62010 }, { "grad_norm": 0.22821146249771118, "learning_rate": 1.3187219029741737e-05, "loss": 0.018, "step": 62020 }, { "grad_norm": 0.34852349758148193, "learning_rate": 1.3173235814579693e-05, "loss": 0.0169, "step": 62030 }, { "grad_norm": 0.4431566298007965, "learning_rate": 1.3159258892104398e-05, "loss": 0.0173, "step": 62040 }, { "grad_norm": 0.1956612467765808, "learning_rate": 1.3145288264704198e-05, "loss": 0.0217, "step": 62050 }, { "grad_norm": 0.2509126663208008, "learning_rate": 1.3131323934766237e-05, "loss": 0.0201, "step": 62060 }, { "grad_norm": 0.2932169437408447, "learning_rate": 1.311736590467666e-05, "loss": 0.0163, "step": 62070 }, { "grad_norm": 0.19990111887454987, "learning_rate": 1.3103414176820505e-05, "loss": 0.0181, "step": 62080 }, { "grad_norm": 0.17870381474494934, "learning_rate": 1.308946875358174e-05, "loss": 0.027, "step": 62090 }, { "grad_norm": 0.30624809861183167, "learning_rate": 1.3075529637343276e-05, "loss": 0.0347, "step": 62100 }, { "grad_norm": 0.8523943424224854, "learning_rate": 1.3061596830486883e-05, "loss": 0.0207, "step": 62110 }, { "grad_norm": 0.359595388174057, "learning_rate": 1.3047670335393353e-05, "loss": 0.017, "step": 62120 }, { "grad_norm": 0.3737773299217224, "learning_rate": 1.3033750154442298e-05, "loss": 0.0238, "step": 62130 }, { "grad_norm": 0.3695244789123535, "learning_rate": 1.3019836290012316e-05, "loss": 0.0185, "step": 62140 }, { "grad_norm": 0.3171972930431366, "learning_rate": 1.30059287444809e-05, "loss": 0.0227, "step": 62150 }, { "grad_norm": 0.28175729513168335, "learning_rate": 1.299202752022447e-05, "loss": 0.0294, "step": 62160 }, { "grad_norm": 0.1634306162595749, "learning_rate": 1.2978132619618371e-05, "loss": 0.0136, "step": 62170 }, { "grad_norm": 0.2645154893398285, "learning_rate": 1.2964244045036866e-05, "loss": 0.0278, "step": 62180 }, { "grad_norm": 0.2710907757282257, "learning_rate": 1.295036179885311e-05, "loss": 0.0176, "step": 62190 }, { "grad_norm": 0.3241714537143707, "learning_rate": 1.293648588343922e-05, "loss": 0.0248, "step": 62200 }, { "grad_norm": 0.3988826870918274, "learning_rate": 1.2922616301166196e-05, "loss": 0.0183, "step": 62210 }, { "grad_norm": 0.3927291929721832, "learning_rate": 1.2908753054403976e-05, "loss": 0.0197, "step": 62220 }, { "grad_norm": 0.27237004041671753, "learning_rate": 1.2894896145521429e-05, "loss": 0.0187, "step": 62230 }, { "grad_norm": 0.3367251455783844, "learning_rate": 1.2881045576886275e-05, "loss": 0.0172, "step": 62240 }, { "grad_norm": 0.8261880278587341, "learning_rate": 1.2867201350865254e-05, "loss": 0.0224, "step": 62250 }, { "grad_norm": 0.27672624588012695, "learning_rate": 1.2853363469823914e-05, "loss": 0.018, "step": 62260 }, { "grad_norm": 0.3483906388282776, "learning_rate": 1.283953193612682e-05, "loss": 0.0165, "step": 62270 }, { "grad_norm": 0.36189910769462585, "learning_rate": 1.2825706752137372e-05, "loss": 0.0196, "step": 62280 }, { "grad_norm": 0.24570344388484955, "learning_rate": 1.2811887920217896e-05, "loss": 0.0156, "step": 62290 }, { "grad_norm": 0.24826982617378235, "learning_rate": 1.2798075442729707e-05, "loss": 0.0175, "step": 62300 }, { "grad_norm": 0.21799305081367493, "learning_rate": 1.2784269322032922e-05, "loss": 0.0163, "step": 62310 }, { "grad_norm": 0.2852824032306671, "learning_rate": 1.2770469560486686e-05, "loss": 0.0213, "step": 62320 }, { "grad_norm": 0.26501768827438354, "learning_rate": 1.2756676160448956e-05, "loss": 0.0213, "step": 62330 }, { "grad_norm": 0.3392171561717987, "learning_rate": 1.2742889124276663e-05, "loss": 0.0212, "step": 62340 }, { "grad_norm": 0.44596636295318604, "learning_rate": 1.2729108454325639e-05, "loss": 0.0246, "step": 62350 }, { "grad_norm": 0.2363341599702835, "learning_rate": 1.2715334152950614e-05, "loss": 0.0188, "step": 62360 }, { "grad_norm": 0.36160463094711304, "learning_rate": 1.2701566222505246e-05, "loss": 0.0245, "step": 62370 }, { "grad_norm": 0.32078754901885986, "learning_rate": 1.2687804665342107e-05, "loss": 0.0191, "step": 62380 }, { "grad_norm": 0.28955593705177307, "learning_rate": 1.267404948381265e-05, "loss": 0.0183, "step": 62390 }, { "grad_norm": 0.26106807589530945, "learning_rate": 1.2660300680267267e-05, "loss": 0.0114, "step": 62400 }, { "grad_norm": 0.32503899931907654, "learning_rate": 1.2646558257055257e-05, "loss": 0.0199, "step": 62410 }, { "grad_norm": 0.16260790824890137, "learning_rate": 1.263282221652482e-05, "loss": 0.0161, "step": 62420 }, { "grad_norm": 0.29441729187965393, "learning_rate": 1.2619092561023088e-05, "loss": 0.0183, "step": 62430 }, { "grad_norm": 0.25123006105422974, "learning_rate": 1.2605369292896036e-05, "loss": 0.0222, "step": 62440 }, { "grad_norm": 0.35026827454566956, "learning_rate": 1.2591652414488658e-05, "loss": 0.0225, "step": 62450 }, { "grad_norm": 0.20973972976207733, "learning_rate": 1.2577941928144732e-05, "loss": 0.0186, "step": 62460 }, { "grad_norm": 0.24224717915058136, "learning_rate": 1.256423783620706e-05, "loss": 0.022, "step": 62470 }, { "grad_norm": 0.2880557179450989, "learning_rate": 1.2550540141017264e-05, "loss": 0.0192, "step": 62480 }, { "grad_norm": 0.1702958643436432, "learning_rate": 1.253684884491591e-05, "loss": 0.0208, "step": 62490 }, { "grad_norm": 0.5155977606773376, "learning_rate": 1.2523163950242483e-05, "loss": 0.0181, "step": 62500 }, { "grad_norm": 0.3119107186794281, "learning_rate": 1.2509485459335313e-05, "loss": 0.0153, "step": 62510 }, { "grad_norm": 0.23211808502674103, "learning_rate": 1.2495813374531739e-05, "loss": 0.0137, "step": 62520 }, { "grad_norm": 0.2968406677246094, "learning_rate": 1.2482147698167907e-05, "loss": 0.0163, "step": 62530 }, { "grad_norm": 0.2622933089733124, "learning_rate": 1.2468488432578911e-05, "loss": 0.0209, "step": 62540 }, { "grad_norm": 0.25837016105651855, "learning_rate": 1.2454835580098761e-05, "loss": 0.0203, "step": 62550 }, { "grad_norm": 0.3245193064212799, "learning_rate": 1.2441189143060338e-05, "loss": 0.0209, "step": 62560 }, { "grad_norm": 0.3005630075931549, "learning_rate": 1.242754912379548e-05, "loss": 0.0223, "step": 62570 }, { "grad_norm": 0.27502015233039856, "learning_rate": 1.2413915524634844e-05, "loss": 0.0366, "step": 62580 }, { "grad_norm": 0.269302636384964, "learning_rate": 1.2400288347908073e-05, "loss": 0.0198, "step": 62590 }, { "grad_norm": 0.19599656760692596, "learning_rate": 1.2386667595943663e-05, "loss": 0.0201, "step": 62600 }, { "grad_norm": 0.2694498896598816, "learning_rate": 1.2373053271069035e-05, "loss": 0.0179, "step": 62610 }, { "grad_norm": 0.34249478578567505, "learning_rate": 1.2359445375610501e-05, "loss": 0.0309, "step": 62620 }, { "grad_norm": 0.3583613634109497, "learning_rate": 1.2345843911893301e-05, "loss": 0.0165, "step": 62630 }, { "grad_norm": 0.28463098406791687, "learning_rate": 1.2332248882241498e-05, "loss": 0.0321, "step": 62640 }, { "grad_norm": 0.30334317684173584, "learning_rate": 1.2318660288978178e-05, "loss": 0.0194, "step": 62650 }, { "grad_norm": 0.34736168384552, "learning_rate": 1.2305078134425213e-05, "loss": 0.0196, "step": 62660 }, { "grad_norm": 0.2878805994987488, "learning_rate": 1.2291502420903434e-05, "loss": 0.0178, "step": 62670 }, { "grad_norm": 0.17317257821559906, "learning_rate": 1.2277933150732567e-05, "loss": 0.0201, "step": 62680 }, { "grad_norm": 0.21378308534622192, "learning_rate": 1.2264370326231216e-05, "loss": 0.0165, "step": 62690 }, { "grad_norm": 0.2676932215690613, "learning_rate": 1.2250813949716927e-05, "loss": 0.0165, "step": 62700 }, { "grad_norm": 0.2656797766685486, "learning_rate": 1.2237264023506063e-05, "loss": 0.0201, "step": 62710 }, { "grad_norm": 0.3050667643547058, "learning_rate": 1.2223720549913987e-05, "loss": 0.02, "step": 62720 }, { "grad_norm": 0.2650316059589386, "learning_rate": 1.221018353125487e-05, "loss": 0.0176, "step": 62730 }, { "grad_norm": 0.24031883478164673, "learning_rate": 1.2196652969841837e-05, "loss": 0.0191, "step": 62740 }, { "grad_norm": 0.25702205300331116, "learning_rate": 1.2183128867986904e-05, "loss": 0.0251, "step": 62750 }, { "grad_norm": 0.3515136241912842, "learning_rate": 1.2169611228000927e-05, "loss": 0.0193, "step": 62760 }, { "grad_norm": 0.7085626721382141, "learning_rate": 1.2156100052193752e-05, "loss": 0.0238, "step": 62770 }, { "grad_norm": 0.25049424171447754, "learning_rate": 1.214259534287403e-05, "loss": 0.0241, "step": 62780 }, { "grad_norm": 0.1689920574426651, "learning_rate": 1.2129097102349363e-05, "loss": 0.0172, "step": 62790 }, { "grad_norm": 0.4198821783065796, "learning_rate": 1.2115605332926227e-05, "loss": 0.0204, "step": 62800 }, { "grad_norm": 0.26884666085243225, "learning_rate": 1.2102120036910003e-05, "loss": 0.0305, "step": 62810 }, { "grad_norm": 0.6267996430397034, "learning_rate": 1.2088641216604956e-05, "loss": 0.0208, "step": 62820 }, { "grad_norm": 0.2963850200176239, "learning_rate": 1.2075168874314264e-05, "loss": 0.0152, "step": 62830 }, { "grad_norm": 0.16059452295303345, "learning_rate": 1.2061703012339942e-05, "loss": 0.0181, "step": 62840 }, { "grad_norm": 0.24602095782756805, "learning_rate": 1.204824363298297e-05, "loss": 0.0203, "step": 62850 }, { "grad_norm": 0.5242286324501038, "learning_rate": 1.2034790738543173e-05, "loss": 0.027, "step": 62860 }, { "grad_norm": 0.30467987060546875, "learning_rate": 1.2021344331319284e-05, "loss": 0.0267, "step": 62870 }, { "grad_norm": 0.19260284304618835, "learning_rate": 1.200790441360894e-05, "loss": 0.0217, "step": 62880 }, { "grad_norm": 0.20347939431667328, "learning_rate": 1.199447098770864e-05, "loss": 0.027, "step": 62890 }, { "grad_norm": 0.4647511839866638, "learning_rate": 1.198104405591381e-05, "loss": 0.0214, "step": 62900 }, { "grad_norm": 0.25447332859039307, "learning_rate": 1.1967623620518697e-05, "loss": 0.0167, "step": 62910 }, { "grad_norm": 0.2678420841693878, "learning_rate": 1.1954209683816554e-05, "loss": 0.0186, "step": 62920 }, { "grad_norm": 0.2506680488586426, "learning_rate": 1.1940802248099402e-05, "loss": 0.0181, "step": 62930 }, { "grad_norm": 0.4215489625930786, "learning_rate": 1.1927401315658232e-05, "loss": 0.0282, "step": 62940 }, { "grad_norm": 0.31444600224494934, "learning_rate": 1.1914006888782898e-05, "loss": 0.0185, "step": 62950 }, { "grad_norm": 0.35940611362457275, "learning_rate": 1.1900618969762106e-05, "loss": 0.0208, "step": 62960 }, { "grad_norm": 0.41023769974708557, "learning_rate": 1.188723756088354e-05, "loss": 0.0179, "step": 62970 }, { "grad_norm": 0.28209033608436584, "learning_rate": 1.1873862664433672e-05, "loss": 0.0207, "step": 62980 }, { "grad_norm": 0.22579868137836456, "learning_rate": 1.1860494282697927e-05, "loss": 0.0193, "step": 62990 }, { "grad_norm": 0.2877483069896698, "learning_rate": 1.1847132417960588e-05, "loss": 0.0214, "step": 63000 }, { "grad_norm": 0.38714075088500977, "learning_rate": 1.1833777072504832e-05, "loss": 0.0187, "step": 63010 }, { "grad_norm": 0.24427247047424316, "learning_rate": 1.1820428248612731e-05, "loss": 0.0182, "step": 63020 }, { "grad_norm": 0.5485484600067139, "learning_rate": 1.1807085948565245e-05, "loss": 0.0219, "step": 63030 }, { "grad_norm": 0.5102887749671936, "learning_rate": 1.1793750174642172e-05, "loss": 0.0222, "step": 63040 }, { "grad_norm": 0.8082175850868225, "learning_rate": 1.1780420929122254e-05, "loss": 0.0245, "step": 63050 }, { "grad_norm": 0.46820926666259766, "learning_rate": 1.176709821428309e-05, "loss": 0.0208, "step": 63060 }, { "grad_norm": 0.4171127676963806, "learning_rate": 1.1753782032401173e-05, "loss": 0.0225, "step": 63070 }, { "grad_norm": 0.31407588720321655, "learning_rate": 1.1740472385751866e-05, "loss": 0.0189, "step": 63080 }, { "grad_norm": 0.18015800416469574, "learning_rate": 1.172716927660943e-05, "loss": 0.0177, "step": 63090 }, { "grad_norm": 0.3548154830932617, "learning_rate": 1.1713872707247015e-05, "loss": 0.0226, "step": 63100 }, { "grad_norm": 0.20201534032821655, "learning_rate": 1.1700582679936595e-05, "loss": 0.0163, "step": 63110 }, { "grad_norm": 0.3308849334716797, "learning_rate": 1.1687299196949136e-05, "loss": 0.0266, "step": 63120 }, { "grad_norm": 0.35670578479766846, "learning_rate": 1.1674022260554374e-05, "loss": 0.0181, "step": 63130 }, { "grad_norm": 0.2557908296585083, "learning_rate": 1.1660751873020987e-05, "loss": 0.0218, "step": 63140 }, { "grad_norm": 0.358290433883667, "learning_rate": 1.1647488036616538e-05, "loss": 0.0213, "step": 63150 }, { "grad_norm": 0.3088708817958832, "learning_rate": 1.1634230753607417e-05, "loss": 0.0248, "step": 63160 }, { "grad_norm": 0.20260512828826904, "learning_rate": 1.1620980026258982e-05, "loss": 0.0214, "step": 63170 }, { "grad_norm": 0.24621883034706116, "learning_rate": 1.1607735856835373e-05, "loss": 0.0212, "step": 63180 }, { "grad_norm": 0.30150285363197327, "learning_rate": 1.1594498247599677e-05, "loss": 0.0182, "step": 63190 }, { "grad_norm": 0.2827376127243042, "learning_rate": 1.158126720081384e-05, "loss": 0.0237, "step": 63200 }, { "grad_norm": 0.49077415466308594, "learning_rate": 1.156804271873868e-05, "loss": 0.0175, "step": 63210 }, { "grad_norm": 0.4387534558773041, "learning_rate": 1.1554824803633924e-05, "loss": 0.0234, "step": 63220 }, { "grad_norm": 0.23649829626083374, "learning_rate": 1.154161345775811e-05, "loss": 0.02, "step": 63230 }, { "grad_norm": 0.1515868455171585, "learning_rate": 1.1528408683368724e-05, "loss": 0.0208, "step": 63240 }, { "grad_norm": 0.39566299319267273, "learning_rate": 1.1515210482722088e-05, "loss": 0.0264, "step": 63250 }, { "grad_norm": 0.28147080540657043, "learning_rate": 1.150201885807342e-05, "loss": 0.0284, "step": 63260 }, { "grad_norm": 0.4326026737689972, "learning_rate": 1.1488833811676807e-05, "loss": 0.0172, "step": 63270 }, { "grad_norm": 0.1997324377298355, "learning_rate": 1.1475655345785213e-05, "loss": 0.0215, "step": 63280 }, { "grad_norm": 0.2975814938545227, "learning_rate": 1.1462483462650481e-05, "loss": 0.0174, "step": 63290 }, { "grad_norm": 0.37367865443229675, "learning_rate": 1.1449318164523331e-05, "loss": 0.0181, "step": 63300 }, { "grad_norm": 0.21002843976020813, "learning_rate": 1.1436159453653334e-05, "loss": 0.0187, "step": 63310 }, { "grad_norm": 0.34143301844596863, "learning_rate": 1.1423007332288955e-05, "loss": 0.0256, "step": 63320 }, { "grad_norm": 0.2991519570350647, "learning_rate": 1.1409861802677546e-05, "loss": 0.0202, "step": 63330 }, { "grad_norm": 0.3520830273628235, "learning_rate": 1.1396722867065313e-05, "loss": 0.0186, "step": 63340 }, { "grad_norm": 0.34987419843673706, "learning_rate": 1.1383590527697352e-05, "loss": 0.0204, "step": 63350 }, { "grad_norm": 1.3075283765792847, "learning_rate": 1.137046478681758e-05, "loss": 0.0228, "step": 63360 }, { "grad_norm": 0.2670069932937622, "learning_rate": 1.1357345646668888e-05, "loss": 0.0325, "step": 63370 }, { "grad_norm": 0.22797490656375885, "learning_rate": 1.1344233109492924e-05, "loss": 0.0195, "step": 63380 }, { "grad_norm": 0.23839148879051208, "learning_rate": 1.1331127177530292e-05, "loss": 0.0172, "step": 63390 }, { "grad_norm": 0.2885686457157135, "learning_rate": 1.1318027853020441e-05, "loss": 0.0193, "step": 63400 }, { "grad_norm": 0.3865234851837158, "learning_rate": 1.1304935138201645e-05, "loss": 0.0188, "step": 63410 }, { "grad_norm": 0.4355016350746155, "learning_rate": 1.1291849035311153e-05, "loss": 0.0206, "step": 63420 }, { "grad_norm": 0.31243839859962463, "learning_rate": 1.1278769546584972e-05, "loss": 0.0171, "step": 63430 }, { "grad_norm": 0.2829006016254425, "learning_rate": 1.1265696674258052e-05, "loss": 0.0174, "step": 63440 }, { "grad_norm": 0.17033971846103668, "learning_rate": 1.1252630420564186e-05, "loss": 0.0187, "step": 63450 }, { "grad_norm": 0.22889572381973267, "learning_rate": 1.1239570787736036e-05, "loss": 0.0216, "step": 63460 }, { "grad_norm": 0.3436417579650879, "learning_rate": 1.1226517778005135e-05, "loss": 0.0215, "step": 63470 }, { "grad_norm": 0.22835403680801392, "learning_rate": 1.1213471393601893e-05, "loss": 0.0212, "step": 63480 }, { "grad_norm": 0.1797051578760147, "learning_rate": 1.1200431636755587e-05, "loss": 0.0206, "step": 63490 }, { "grad_norm": 0.8920679688453674, "learning_rate": 1.1187398509694336e-05, "loss": 0.0265, "step": 63500 }, { "grad_norm": 0.20494651794433594, "learning_rate": 1.1174372014645146e-05, "loss": 0.0171, "step": 63510 }, { "grad_norm": 0.3721790909767151, "learning_rate": 1.1161352153833899e-05, "loss": 0.0279, "step": 63520 }, { "grad_norm": 0.29657217860221863, "learning_rate": 1.1148338929485325e-05, "loss": 0.0206, "step": 63530 }, { "grad_norm": 0.31196528673171997, "learning_rate": 1.113533234382304e-05, "loss": 0.0159, "step": 63540 }, { "grad_norm": 0.18827572464942932, "learning_rate": 1.1122332399069513e-05, "loss": 0.018, "step": 63550 }, { "grad_norm": 0.7518406510353088, "learning_rate": 1.1109339097446047e-05, "loss": 0.0262, "step": 63560 }, { "grad_norm": 0.4043770134449005, "learning_rate": 1.1096352441172897e-05, "loss": 0.0271, "step": 63570 }, { "grad_norm": 0.3977196216583252, "learning_rate": 1.1083372432469086e-05, "loss": 0.017, "step": 63580 }, { "grad_norm": 0.38600656390190125, "learning_rate": 1.107039907355255e-05, "loss": 0.0262, "step": 63590 }, { "grad_norm": 0.31759145855903625, "learning_rate": 1.1057432366640103e-05, "loss": 0.0181, "step": 63600 }, { "grad_norm": 0.339386522769928, "learning_rate": 1.1044472313947352e-05, "loss": 0.0186, "step": 63610 }, { "grad_norm": 0.3242480158805847, "learning_rate": 1.1031518917688877e-05, "loss": 0.0249, "step": 63620 }, { "grad_norm": 0.24190294742584229, "learning_rate": 1.1018572180078007e-05, "loss": 0.0278, "step": 63630 }, { "grad_norm": 0.26687514781951904, "learning_rate": 1.1005632103327018e-05, "loss": 0.0204, "step": 63640 }, { "grad_norm": 0.3545421063899994, "learning_rate": 1.0992698689646996e-05, "loss": 0.0204, "step": 63650 }, { "grad_norm": 0.4407674968242645, "learning_rate": 1.0979771941247919e-05, "loss": 0.0237, "step": 63660 }, { "grad_norm": 0.3575204312801361, "learning_rate": 1.0966851860338611e-05, "loss": 0.0194, "step": 63670 }, { "grad_norm": 0.14137107133865356, "learning_rate": 1.0953938449126766e-05, "loss": 0.0165, "step": 63680 }, { "grad_norm": 0.42703813314437866, "learning_rate": 1.0941031709818933e-05, "loss": 0.0342, "step": 63690 }, { "grad_norm": 0.290261834859848, "learning_rate": 1.0928131644620509e-05, "loss": 0.0254, "step": 63700 }, { "grad_norm": 0.47607991099357605, "learning_rate": 1.0915238255735766e-05, "loss": 0.0272, "step": 63710 }, { "grad_norm": 0.2881160378456116, "learning_rate": 1.0902351545367833e-05, "loss": 0.0194, "step": 63720 }, { "grad_norm": 0.3188246786594391, "learning_rate": 1.0889471515718702e-05, "loss": 0.0197, "step": 63730 }, { "grad_norm": 0.23062466084957123, "learning_rate": 1.087659816898921e-05, "loss": 0.0148, "step": 63740 }, { "grad_norm": 0.45026615262031555, "learning_rate": 1.0863731507379082e-05, "loss": 0.0212, "step": 63750 }, { "grad_norm": 0.34938472509384155, "learning_rate": 1.0850871533086827e-05, "loss": 0.0201, "step": 63760 }, { "grad_norm": 0.48647981882095337, "learning_rate": 1.0838018248309927e-05, "loss": 0.0216, "step": 63770 }, { "grad_norm": 0.3565562963485718, "learning_rate": 1.0825171655244615e-05, "loss": 0.0212, "step": 63780 }, { "grad_norm": 0.260093092918396, "learning_rate": 1.0812331756086025e-05, "loss": 0.018, "step": 63790 }, { "grad_norm": 0.4049442708492279, "learning_rate": 1.079949855302817e-05, "loss": 0.0199, "step": 63800 }, { "grad_norm": 0.3099919259548187, "learning_rate": 1.0786672048263852e-05, "loss": 0.0242, "step": 63810 }, { "grad_norm": 0.24573266506195068, "learning_rate": 1.0773852243984817e-05, "loss": 0.0216, "step": 63820 }, { "grad_norm": 0.34424638748168945, "learning_rate": 1.0761039142381586e-05, "loss": 0.0272, "step": 63830 }, { "grad_norm": 0.4579274356365204, "learning_rate": 1.0748232745643577e-05, "loss": 0.019, "step": 63840 }, { "grad_norm": 0.312000036239624, "learning_rate": 1.0735433055959055e-05, "loss": 0.0234, "step": 63850 }, { "grad_norm": 0.655272364616394, "learning_rate": 1.0722640075515133e-05, "loss": 0.0224, "step": 63860 }, { "grad_norm": 0.20968936383724213, "learning_rate": 1.0709853806497795e-05, "loss": 0.0147, "step": 63870 }, { "grad_norm": 0.3397890031337738, "learning_rate": 1.0697074251091831e-05, "loss": 0.0236, "step": 63880 }, { "grad_norm": 0.3019152581691742, "learning_rate": 1.0684301411480962e-05, "loss": 0.0203, "step": 63890 }, { "grad_norm": 0.3023720383644104, "learning_rate": 1.067153528984769e-05, "loss": 0.0174, "step": 63900 }, { "grad_norm": 0.408481240272522, "learning_rate": 1.0658775888373395e-05, "loss": 0.0195, "step": 63910 }, { "grad_norm": 0.3138797879219055, "learning_rate": 1.0646023209238314e-05, "loss": 0.0271, "step": 63920 }, { "grad_norm": 0.21073558926582336, "learning_rate": 1.0633277254621537e-05, "loss": 0.0216, "step": 63930 }, { "grad_norm": 0.17807495594024658, "learning_rate": 1.0620538026700994e-05, "loss": 0.0188, "step": 63940 }, { "grad_norm": 0.2655394971370697, "learning_rate": 1.0607805527653486e-05, "loss": 0.0196, "step": 63950 }, { "grad_norm": 0.24770359694957733, "learning_rate": 1.059507975965462e-05, "loss": 0.0209, "step": 63960 }, { "grad_norm": 0.16552776098251343, "learning_rate": 1.0582360724878898e-05, "loss": 0.0149, "step": 63970 }, { "grad_norm": 0.26760146021842957, "learning_rate": 1.0569648425499651e-05, "loss": 0.0219, "step": 63980 }, { "grad_norm": 0.22348597645759583, "learning_rate": 1.0556942863689063e-05, "loss": 0.0206, "step": 63990 }, { "grad_norm": 0.3134969174861908, "learning_rate": 1.054424404161819e-05, "loss": 0.0233, "step": 64000 }, { "grad_norm": 0.30711761116981506, "learning_rate": 1.053155196145686e-05, "loss": 0.0157, "step": 64010 }, { "grad_norm": 0.2716337740421295, "learning_rate": 1.0518866625373863e-05, "loss": 0.0181, "step": 64020 }, { "grad_norm": 0.20579591393470764, "learning_rate": 1.0506188035536735e-05, "loss": 0.0202, "step": 64030 }, { "grad_norm": 0.2955949604511261, "learning_rate": 1.0493516194111919e-05, "loss": 0.02, "step": 64040 }, { "grad_norm": 0.21179521083831787, "learning_rate": 1.0480851103264688e-05, "loss": 0.0174, "step": 64050 }, { "grad_norm": 0.4021323621273041, "learning_rate": 1.046819276515913e-05, "loss": 0.0125, "step": 64060 }, { "grad_norm": 0.33543121814727783, "learning_rate": 1.0455541181958256e-05, "loss": 0.0172, "step": 64070 }, { "grad_norm": 0.25627273321151733, "learning_rate": 1.0442896355823822e-05, "loss": 0.0208, "step": 64080 }, { "grad_norm": 0.3579299747943878, "learning_rate": 1.0430258288916539e-05, "loss": 0.0168, "step": 64090 }, { "grad_norm": 0.29103395342826843, "learning_rate": 1.0417626983395868e-05, "loss": 0.0271, "step": 64100 }, { "grad_norm": 0.29461178183555603, "learning_rate": 1.0405002441420165e-05, "loss": 0.0169, "step": 64110 }, { "grad_norm": 0.4044371247291565, "learning_rate": 1.039238466514662e-05, "loss": 0.0221, "step": 64120 }, { "grad_norm": 0.17794901132583618, "learning_rate": 1.0379773656731262e-05, "loss": 0.0212, "step": 64130 }, { "grad_norm": 0.256635844707489, "learning_rate": 1.0367169418328986e-05, "loss": 0.0177, "step": 64140 }, { "grad_norm": 0.30595436692237854, "learning_rate": 1.0354571952093484e-05, "loss": 0.0219, "step": 64150 }, { "grad_norm": 0.24923831224441528, "learning_rate": 1.034198126017733e-05, "loss": 0.0163, "step": 64160 }, { "grad_norm": 0.3386511206626892, "learning_rate": 1.0329397344731928e-05, "loss": 0.0182, "step": 64170 }, { "grad_norm": 0.3398599922657013, "learning_rate": 1.0316820207907524e-05, "loss": 0.0233, "step": 64180 }, { "grad_norm": 0.4410220980644226, "learning_rate": 1.0304249851853215e-05, "loss": 0.022, "step": 64190 }, { "grad_norm": 0.4269769787788391, "learning_rate": 1.0291686278716933e-05, "loss": 0.0185, "step": 64200 }, { "grad_norm": 0.3240775167942047, "learning_rate": 1.0279129490645418e-05, "loss": 0.0207, "step": 64210 }, { "grad_norm": 0.2262033224105835, "learning_rate": 1.0266579489784328e-05, "loss": 0.0154, "step": 64220 }, { "grad_norm": 0.2834761142730713, "learning_rate": 1.0254036278278084e-05, "loss": 0.0151, "step": 64230 }, { "grad_norm": 0.2953384518623352, "learning_rate": 1.0241499858269982e-05, "loss": 0.0158, "step": 64240 }, { "grad_norm": 0.4043244421482086, "learning_rate": 1.0228970231902169e-05, "loss": 0.021, "step": 64250 }, { "grad_norm": 0.3440195918083191, "learning_rate": 1.0216447401315582e-05, "loss": 0.0203, "step": 64260 }, { "grad_norm": 0.5473034381866455, "learning_rate": 1.020393136865007e-05, "loss": 0.0183, "step": 64270 }, { "grad_norm": 0.23771043121814728, "learning_rate": 1.0191422136044242e-05, "loss": 0.0246, "step": 64280 }, { "grad_norm": 0.20014294981956482, "learning_rate": 1.017891970563563e-05, "loss": 0.0195, "step": 64290 }, { "grad_norm": 0.20975439250469208, "learning_rate": 1.0166424079560516e-05, "loss": 0.0269, "step": 64300 }, { "grad_norm": 0.334189236164093, "learning_rate": 1.0153935259954078e-05, "loss": 0.0191, "step": 64310 }, { "grad_norm": 0.20483817160129547, "learning_rate": 1.0141453248950311e-05, "loss": 0.0226, "step": 64320 }, { "grad_norm": 0.6102038621902466, "learning_rate": 1.0128978048682054e-05, "loss": 0.0261, "step": 64330 }, { "grad_norm": 0.3972408175468445, "learning_rate": 1.0116509661280982e-05, "loss": 0.022, "step": 64340 }, { "grad_norm": 0.20253153145313263, "learning_rate": 1.0104048088877576e-05, "loss": 0.0147, "step": 64350 }, { "grad_norm": 0.22560365498065948, "learning_rate": 1.0091593333601201e-05, "loss": 0.0161, "step": 64360 }, { "grad_norm": 0.3449494540691376, "learning_rate": 1.0079145397580031e-05, "loss": 0.0211, "step": 64370 }, { "grad_norm": 0.2914640009403229, "learning_rate": 1.006670428294107e-05, "loss": 0.0216, "step": 64380 }, { "grad_norm": 0.30150559544563293, "learning_rate": 1.0054269991810166e-05, "loss": 0.0171, "step": 64390 }, { "grad_norm": 0.2491379976272583, "learning_rate": 1.0041842526312024e-05, "loss": 0.0151, "step": 64400 }, { "grad_norm": 0.3178468942642212, "learning_rate": 1.0029421888570101e-05, "loss": 0.0158, "step": 64410 }, { "grad_norm": 0.13682988286018372, "learning_rate": 1.0017008080706813e-05, "loss": 0.0271, "step": 64420 }, { "grad_norm": 0.24968321621418, "learning_rate": 1.0004601104843287e-05, "loss": 0.0195, "step": 64430 }, { "grad_norm": 0.26427987217903137, "learning_rate": 9.992200963099562e-06, "loss": 0.0178, "step": 64440 }, { "grad_norm": 0.1255844086408615, "learning_rate": 9.979807657594486e-06, "loss": 0.0149, "step": 64450 }, { "grad_norm": 0.30821868777275085, "learning_rate": 9.967421190445703e-06, "loss": 0.018, "step": 64460 }, { "grad_norm": 0.24288183450698853, "learning_rate": 9.955041563769769e-06, "loss": 0.019, "step": 64470 }, { "grad_norm": 0.24610143899917603, "learning_rate": 9.942668779681974e-06, "loss": 0.0218, "step": 64480 }, { "grad_norm": 0.47715240716934204, "learning_rate": 9.930302840296541e-06, "loss": 0.0198, "step": 64490 }, { "grad_norm": 0.13943471014499664, "learning_rate": 9.917943747726426e-06, "loss": 0.0161, "step": 64500 }, { "grad_norm": 0.1933203637599945, "learning_rate": 9.905591504083484e-06, "loss": 0.0184, "step": 64510 }, { "grad_norm": 0.26201218366622925, "learning_rate": 9.893246111478382e-06, "loss": 0.0191, "step": 64520 }, { "grad_norm": 0.47579506039619446, "learning_rate": 9.88090757202057e-06, "loss": 0.0188, "step": 64530 }, { "grad_norm": 0.3317889869213104, "learning_rate": 9.868575887818421e-06, "loss": 0.0228, "step": 64540 }, { "grad_norm": 0.19522713124752045, "learning_rate": 9.856251060979044e-06, "loss": 0.0176, "step": 64550 }, { "grad_norm": 0.25338372588157654, "learning_rate": 9.843933093608426e-06, "loss": 0.0213, "step": 64560 }, { "grad_norm": 0.24539296329021454, "learning_rate": 9.831621987811368e-06, "loss": 0.0271, "step": 64570 }, { "grad_norm": 0.23900581896305084, "learning_rate": 9.819317745691509e-06, "loss": 0.0176, "step": 64580 }, { "grad_norm": 0.3199784457683563, "learning_rate": 9.8070203693513e-06, "loss": 0.0172, "step": 64590 }, { "grad_norm": 0.4272356629371643, "learning_rate": 9.794729860892048e-06, "loss": 0.0188, "step": 64600 }, { "grad_norm": 0.32373449206352234, "learning_rate": 9.782446222413827e-06, "loss": 0.0201, "step": 64610 }, { "grad_norm": 0.15309977531433105, "learning_rate": 9.770169456015598e-06, "loss": 0.025, "step": 64620 }, { "grad_norm": 0.32489436864852905, "learning_rate": 9.75789956379512e-06, "loss": 0.0195, "step": 64630 }, { "grad_norm": 0.35492944717407227, "learning_rate": 9.74563654784898e-06, "loss": 0.0166, "step": 64640 }, { "grad_norm": 0.11181777715682983, "learning_rate": 9.733380410272596e-06, "loss": 0.0163, "step": 64650 }, { "grad_norm": 0.37463489174842834, "learning_rate": 9.721131153160207e-06, "loss": 0.0164, "step": 64660 }, { "grad_norm": 0.4577048122882843, "learning_rate": 9.708888778604886e-06, "loss": 0.0204, "step": 64670 }, { "grad_norm": 0.28917863965034485, "learning_rate": 9.696653288698488e-06, "loss": 0.0214, "step": 64680 }, { "grad_norm": 0.25744158029556274, "learning_rate": 9.684424685531763e-06, "loss": 0.0249, "step": 64690 }, { "grad_norm": 1.9001506567001343, "learning_rate": 9.672202971194216e-06, "loss": 0.0267, "step": 64700 }, { "grad_norm": 0.28431031107902527, "learning_rate": 9.659988147774213e-06, "loss": 0.0189, "step": 64710 }, { "grad_norm": 0.459370493888855, "learning_rate": 9.647780217358942e-06, "loss": 0.0216, "step": 64720 }, { "grad_norm": 0.29454857110977173, "learning_rate": 9.635579182034376e-06, "loss": 0.0211, "step": 64730 }, { "grad_norm": 0.529758095741272, "learning_rate": 9.623385043885386e-06, "loss": 0.0296, "step": 64740 }, { "grad_norm": 0.2173670381307602, "learning_rate": 9.61119780499557e-06, "loss": 0.0169, "step": 64750 }, { "grad_norm": 0.3321610689163208, "learning_rate": 9.599017467447418e-06, "loss": 0.0195, "step": 64760 }, { "grad_norm": 0.29340311884880066, "learning_rate": 9.586844033322206e-06, "loss": 0.0155, "step": 64770 }, { "grad_norm": 0.3862443268299103, "learning_rate": 9.574677504700052e-06, "loss": 0.0211, "step": 64780 }, { "grad_norm": 0.39834076166152954, "learning_rate": 9.562517883659877e-06, "loss": 0.0274, "step": 64790 }, { "grad_norm": 0.265482634305954, "learning_rate": 9.55036517227944e-06, "loss": 0.0159, "step": 64800 }, { "grad_norm": 0.4929830729961395, "learning_rate": 9.538219372635282e-06, "loss": 0.0164, "step": 64810 }, { "grad_norm": 0.20684286952018738, "learning_rate": 9.526080486802802e-06, "loss": 0.0151, "step": 64820 }, { "grad_norm": 0.46015772223472595, "learning_rate": 9.513948516856203e-06, "loss": 0.0198, "step": 64830 }, { "grad_norm": 0.4078460931777954, "learning_rate": 9.501823464868504e-06, "loss": 0.0179, "step": 64840 }, { "grad_norm": 0.3011852204799652, "learning_rate": 9.489705332911547e-06, "loss": 0.0194, "step": 64850 }, { "grad_norm": 0.23103711009025574, "learning_rate": 9.477594123055994e-06, "loss": 0.0201, "step": 64860 }, { "grad_norm": 0.2757290303707123, "learning_rate": 9.465489837371321e-06, "loss": 0.021, "step": 64870 }, { "grad_norm": 0.28134313225746155, "learning_rate": 9.453392477925794e-06, "loss": 0.0182, "step": 64880 }, { "grad_norm": 0.23684191703796387, "learning_rate": 9.441302046786566e-06, "loss": 0.0189, "step": 64890 }, { "grad_norm": 0.33635252714157104, "learning_rate": 9.429218546019519e-06, "loss": 0.0217, "step": 64900 }, { "grad_norm": 0.31013786792755127, "learning_rate": 9.41714197768941e-06, "loss": 0.0231, "step": 64910 }, { "grad_norm": 0.408484548330307, "learning_rate": 9.405072343859805e-06, "loss": 0.019, "step": 64920 }, { "grad_norm": 0.3949126899242401, "learning_rate": 9.393009646593043e-06, "loss": 0.0175, "step": 64930 }, { "grad_norm": 0.30621349811553955, "learning_rate": 9.38095388795035e-06, "loss": 0.0203, "step": 64940 }, { "grad_norm": 0.28518328070640564, "learning_rate": 9.36890506999169e-06, "loss": 0.0202, "step": 64950 }, { "grad_norm": 0.1828630119562149, "learning_rate": 9.356863194775894e-06, "loss": 0.0204, "step": 64960 }, { "grad_norm": 0.33256205916404724, "learning_rate": 9.344828264360583e-06, "loss": 0.0215, "step": 64970 }, { "grad_norm": 0.2399013489484787, "learning_rate": 9.332800280802201e-06, "loss": 0.0172, "step": 64980 }, { "grad_norm": 0.1776336431503296, "learning_rate": 9.32077924615602e-06, "loss": 0.0151, "step": 64990 }, { "grad_norm": 0.2605087161064148, "learning_rate": 9.308765162476063e-06, "loss": 0.0156, "step": 65000 }, { "grad_norm": 0.20113036036491394, "learning_rate": 9.296758031815239e-06, "loss": 0.0202, "step": 65010 }, { "grad_norm": 0.37200927734375, "learning_rate": 9.284757856225229e-06, "loss": 0.0162, "step": 65020 }, { "grad_norm": 0.20731885731220245, "learning_rate": 9.272764637756538e-06, "loss": 0.0197, "step": 65030 }, { "grad_norm": 0.31486672163009644, "learning_rate": 9.260778378458479e-06, "loss": 0.0194, "step": 65040 }, { "grad_norm": 0.13171546161174774, "learning_rate": 9.248799080379172e-06, "loss": 0.0207, "step": 65050 }, { "grad_norm": 0.24685145914554596, "learning_rate": 9.236826745565558e-06, "loss": 0.0198, "step": 65060 }, { "grad_norm": 0.2424004077911377, "learning_rate": 9.224861376063388e-06, "loss": 0.018, "step": 65070 }, { "grad_norm": 0.20992013812065125, "learning_rate": 9.212902973917192e-06, "loss": 0.013, "step": 65080 }, { "grad_norm": 0.30833882093429565, "learning_rate": 9.20095154117035e-06, "loss": 0.0192, "step": 65090 }, { "grad_norm": 0.2582911550998688, "learning_rate": 9.189007079865036e-06, "loss": 0.0146, "step": 65100 }, { "grad_norm": 0.28294697403907776, "learning_rate": 9.177069592042226e-06, "loss": 0.0154, "step": 65110 }, { "grad_norm": 0.2891315221786499, "learning_rate": 9.165139079741724e-06, "loss": 0.0215, "step": 65120 }, { "grad_norm": 0.3441847860813141, "learning_rate": 9.153215545002098e-06, "loss": 0.0232, "step": 65130 }, { "grad_norm": 0.30261480808258057, "learning_rate": 9.141298989860798e-06, "loss": 0.0268, "step": 65140 }, { "grad_norm": 0.2866029143333435, "learning_rate": 9.129389416353994e-06, "loss": 0.0175, "step": 65150 }, { "grad_norm": 0.32107308506965637, "learning_rate": 9.11748682651673e-06, "loss": 0.0144, "step": 65160 }, { "grad_norm": 0.3599679470062256, "learning_rate": 9.105591222382837e-06, "loss": 0.0169, "step": 65170 }, { "grad_norm": 0.28323987126350403, "learning_rate": 9.093702605984915e-06, "loss": 0.0172, "step": 65180 }, { "grad_norm": 0.35194990038871765, "learning_rate": 9.081820979354455e-06, "loss": 0.0178, "step": 65190 }, { "grad_norm": 0.2602311372756958, "learning_rate": 9.069946344521663e-06, "loss": 0.0179, "step": 65200 }, { "grad_norm": 0.2909355163574219, "learning_rate": 9.058078703515598e-06, "loss": 0.0193, "step": 65210 }, { "grad_norm": 0.5554068684577942, "learning_rate": 9.046218058364125e-06, "loss": 0.0282, "step": 65220 }, { "grad_norm": 0.36786478757858276, "learning_rate": 9.034364411093893e-06, "loss": 0.0201, "step": 65230 }, { "grad_norm": 0.40686383843421936, "learning_rate": 9.022517763730371e-06, "loss": 0.0239, "step": 65240 }, { "grad_norm": 0.2591566741466522, "learning_rate": 9.010678118297827e-06, "loss": 0.0197, "step": 65250 }, { "grad_norm": 0.3512421250343323, "learning_rate": 8.998845476819345e-06, "loss": 0.0212, "step": 65260 }, { "grad_norm": 0.5009858012199402, "learning_rate": 8.987019841316773e-06, "loss": 0.0189, "step": 65270 }, { "grad_norm": 0.21699805557727814, "learning_rate": 8.975201213810802e-06, "loss": 0.0149, "step": 65280 }, { "grad_norm": 0.2531862258911133, "learning_rate": 8.963389596320915e-06, "loss": 0.0192, "step": 65290 }, { "grad_norm": 0.33991163969039917, "learning_rate": 8.951584990865391e-06, "loss": 0.0199, "step": 65300 }, { "grad_norm": 0.26293304562568665, "learning_rate": 8.939787399461319e-06, "loss": 0.0241, "step": 65310 }, { "grad_norm": 0.43710649013519287, "learning_rate": 8.927996824124591e-06, "loss": 0.0203, "step": 65320 }, { "grad_norm": 0.3233884274959564, "learning_rate": 8.916213266869854e-06, "loss": 0.0233, "step": 65330 }, { "grad_norm": 0.3069345951080322, "learning_rate": 8.904436729710658e-06, "loss": 0.0182, "step": 65340 }, { "grad_norm": 0.20231282711029053, "learning_rate": 8.892667214659245e-06, "loss": 0.0151, "step": 65350 }, { "grad_norm": 0.3747479319572449, "learning_rate": 8.880904723726713e-06, "loss": 0.018, "step": 65360 }, { "grad_norm": 0.3679766058921814, "learning_rate": 8.869149258922971e-06, "loss": 0.0149, "step": 65370 }, { "grad_norm": 0.3127557337284088, "learning_rate": 8.857400822256662e-06, "loss": 0.0152, "step": 65380 }, { "grad_norm": 0.2475188821554184, "learning_rate": 8.845659415735324e-06, "loss": 0.0246, "step": 65390 }, { "grad_norm": 0.387192964553833, "learning_rate": 8.83392504136521e-06, "loss": 0.017, "step": 65400 }, { "grad_norm": 0.4551902413368225, "learning_rate": 8.822197701151407e-06, "loss": 0.0224, "step": 65410 }, { "grad_norm": 0.3049565553665161, "learning_rate": 8.810477397097804e-06, "loss": 0.0187, "step": 65420 }, { "grad_norm": 0.19412226974964142, "learning_rate": 8.798764131207077e-06, "loss": 0.0183, "step": 65430 }, { "grad_norm": 0.2675231397151947, "learning_rate": 8.787057905480706e-06, "loss": 0.0194, "step": 65440 }, { "grad_norm": 0.273533433675766, "learning_rate": 8.775358721918958e-06, "loss": 0.0197, "step": 65450 }, { "grad_norm": 0.2371797263622284, "learning_rate": 8.763666582520923e-06, "loss": 0.0211, "step": 65460 }, { "grad_norm": 0.2147098332643509, "learning_rate": 8.751981489284445e-06, "loss": 0.0179, "step": 65470 }, { "grad_norm": 0.3326804041862488, "learning_rate": 8.740303444206188e-06, "loss": 0.0165, "step": 65480 }, { "grad_norm": 0.19308675825595856, "learning_rate": 8.72863244928162e-06, "loss": 0.0143, "step": 65490 }, { "grad_norm": 0.1897769570350647, "learning_rate": 8.716968506504991e-06, "loss": 0.0267, "step": 65500 }, { "grad_norm": 0.4374825358390808, "learning_rate": 8.70531161786935e-06, "loss": 0.0159, "step": 65510 }, { "grad_norm": 0.3449057638645172, "learning_rate": 8.693661785366558e-06, "loss": 0.0273, "step": 65520 }, { "grad_norm": 0.2544824779033661, "learning_rate": 8.682019010987208e-06, "loss": 0.023, "step": 65530 }, { "grad_norm": 0.3878398835659027, "learning_rate": 8.670383296720786e-06, "loss": 0.0166, "step": 65540 }, { "grad_norm": 0.4424940049648285, "learning_rate": 8.658754644555478e-06, "loss": 0.0181, "step": 65550 }, { "grad_norm": 0.4257112145423889, "learning_rate": 8.647133056478313e-06, "loss": 0.0217, "step": 65560 }, { "grad_norm": 0.2926003932952881, "learning_rate": 8.635518534475123e-06, "loss": 0.0221, "step": 65570 }, { "grad_norm": 0.18089479207992554, "learning_rate": 8.623911080530467e-06, "loss": 0.0173, "step": 65580 }, { "grad_norm": 0.5217428207397461, "learning_rate": 8.61231069662779e-06, "loss": 0.0228, "step": 65590 }, { "grad_norm": 0.2572307288646698, "learning_rate": 8.600717384749252e-06, "loss": 0.0177, "step": 65600 }, { "grad_norm": 0.7716846466064453, "learning_rate": 8.58913114687584e-06, "loss": 0.0185, "step": 65610 }, { "grad_norm": 0.4416961669921875, "learning_rate": 8.57755198498732e-06, "loss": 0.0236, "step": 65620 }, { "grad_norm": 0.22971127927303314, "learning_rate": 8.565979901062265e-06, "loss": 0.0217, "step": 65630 }, { "grad_norm": 0.20373718440532684, "learning_rate": 8.554414897078033e-06, "loss": 0.015, "step": 65640 }, { "grad_norm": 0.1540980190038681, "learning_rate": 8.542856975010727e-06, "loss": 0.0147, "step": 65650 }, { "grad_norm": 0.19647297263145447, "learning_rate": 8.531306136835337e-06, "loss": 0.014, "step": 65660 }, { "grad_norm": 0.3253435492515564, "learning_rate": 8.519762384525548e-06, "loss": 0.0179, "step": 65670 }, { "grad_norm": 0.48056960105895996, "learning_rate": 8.508225720053875e-06, "loss": 0.0188, "step": 65680 }, { "grad_norm": 0.282784640789032, "learning_rate": 8.496696145391625e-06, "loss": 0.0202, "step": 65690 }, { "grad_norm": 0.1967531144618988, "learning_rate": 8.485173662508889e-06, "loss": 0.0182, "step": 65700 }, { "grad_norm": 0.49811699986457825, "learning_rate": 8.473658273374536e-06, "loss": 0.0207, "step": 65710 }, { "grad_norm": 0.19853734970092773, "learning_rate": 8.462149979956253e-06, "loss": 0.0202, "step": 65720 }, { "grad_norm": 0.2557740807533264, "learning_rate": 8.450648784220461e-06, "loss": 0.0213, "step": 65730 }, { "grad_norm": 0.3751753866672516, "learning_rate": 8.439154688132417e-06, "loss": 0.0188, "step": 65740 }, { "grad_norm": 0.28698232769966125, "learning_rate": 8.427667693656143e-06, "loss": 0.02, "step": 65750 }, { "grad_norm": 0.31169021129608154, "learning_rate": 8.416187802754454e-06, "loss": 0.0149, "step": 65760 }, { "grad_norm": 0.14311373233795166, "learning_rate": 8.404715017388965e-06, "loss": 0.0174, "step": 65770 }, { "grad_norm": 0.37151843309402466, "learning_rate": 8.393249339520015e-06, "loss": 0.0265, "step": 65780 }, { "grad_norm": 0.35792216658592224, "learning_rate": 8.381790771106834e-06, "loss": 0.015, "step": 65790 }, { "grad_norm": 0.3646656572818756, "learning_rate": 8.370339314107339e-06, "loss": 0.0178, "step": 65800 }, { "grad_norm": 0.3703261613845825, "learning_rate": 8.358894970478281e-06, "loss": 0.0188, "step": 65810 }, { "grad_norm": 0.26977068185806274, "learning_rate": 8.347457742175196e-06, "loss": 0.0173, "step": 65820 }, { "grad_norm": 0.2759196162223816, "learning_rate": 8.33602763115236e-06, "loss": 0.019, "step": 65830 }, { "grad_norm": 0.3661595284938812, "learning_rate": 8.324604639362916e-06, "loss": 0.0289, "step": 65840 }, { "grad_norm": 0.303081214427948, "learning_rate": 8.31318876875869e-06, "loss": 0.0171, "step": 65850 }, { "grad_norm": 0.3960222899913788, "learning_rate": 8.30178002129039e-06, "loss": 0.0169, "step": 65860 }, { "grad_norm": 0.23053893446922302, "learning_rate": 8.290378398907423e-06, "loss": 0.0193, "step": 65870 }, { "grad_norm": 0.7088867425918579, "learning_rate": 8.278983903558029e-06, "loss": 0.025, "step": 65880 }, { "grad_norm": 0.41661229729652405, "learning_rate": 8.26759653718921e-06, "loss": 0.0169, "step": 65890 }, { "grad_norm": 0.34260037541389465, "learning_rate": 8.25621630174676e-06, "loss": 0.0233, "step": 65900 }, { "grad_norm": 0.18596766889095306, "learning_rate": 8.244843199175261e-06, "loss": 0.0252, "step": 65910 }, { "grad_norm": 0.44592729210853577, "learning_rate": 8.233477231418046e-06, "loss": 0.0159, "step": 65920 }, { "grad_norm": 0.47701942920684814, "learning_rate": 8.22211840041725e-06, "loss": 0.0191, "step": 65930 }, { "grad_norm": 0.39913249015808105, "learning_rate": 8.210766708113792e-06, "loss": 0.0236, "step": 65940 }, { "grad_norm": 0.6691967248916626, "learning_rate": 8.199422156447367e-06, "loss": 0.0196, "step": 65950 }, { "grad_norm": 0.23963092267513275, "learning_rate": 8.188084747356451e-06, "loss": 0.0178, "step": 65960 }, { "grad_norm": 0.9752946496009827, "learning_rate": 8.176754482778299e-06, "loss": 0.0225, "step": 65970 }, { "grad_norm": 0.36547982692718506, "learning_rate": 8.165431364648918e-06, "loss": 0.0198, "step": 65980 }, { "grad_norm": 0.3799917995929718, "learning_rate": 8.154115394903162e-06, "loss": 0.017, "step": 65990 }, { "grad_norm": 0.2533198893070221, "learning_rate": 8.142806575474582e-06, "loss": 0.0144, "step": 66000 }, { "grad_norm": 0.1591930091381073, "learning_rate": 8.131504908295562e-06, "loss": 0.0158, "step": 66010 }, { "grad_norm": 0.31316637992858887, "learning_rate": 8.120210395297262e-06, "loss": 0.0168, "step": 66020 }, { "grad_norm": 0.23004652559757233, "learning_rate": 8.108923038409565e-06, "loss": 0.0219, "step": 66030 }, { "grad_norm": 0.4256870746612549, "learning_rate": 8.097642839561226e-06, "loss": 0.0209, "step": 66040 }, { "grad_norm": 0.3409790098667145, "learning_rate": 8.08636980067966e-06, "loss": 0.0172, "step": 66050 }, { "grad_norm": 0.1770329475402832, "learning_rate": 8.075103923691186e-06, "loss": 0.0195, "step": 66060 }, { "grad_norm": 0.2301761656999588, "learning_rate": 8.063845210520793e-06, "loss": 0.0172, "step": 66070 }, { "grad_norm": 0.4973220229148865, "learning_rate": 8.052593663092295e-06, "loss": 0.0211, "step": 66080 }, { "grad_norm": 0.43315190076828003, "learning_rate": 8.04134928332827e-06, "loss": 0.0181, "step": 66090 }, { "grad_norm": 0.2879354953765869, "learning_rate": 8.030112073150086e-06, "loss": 0.0192, "step": 66100 }, { "grad_norm": 0.25995633006095886, "learning_rate": 8.018882034477881e-06, "loss": 0.0146, "step": 66110 }, { "grad_norm": 0.35852131247520447, "learning_rate": 8.007659169230541e-06, "loss": 0.0178, "step": 66120 }, { "grad_norm": 0.2988186776638031, "learning_rate": 7.996443479325755e-06, "loss": 0.016, "step": 66130 }, { "grad_norm": 0.18888016045093536, "learning_rate": 7.985234966679977e-06, "loss": 0.02, "step": 66140 }, { "grad_norm": 0.31215786933898926, "learning_rate": 7.974033633208438e-06, "loss": 0.017, "step": 66150 }, { "grad_norm": 0.19572028517723083, "learning_rate": 7.962839480825135e-06, "loss": 0.0184, "step": 66160 }, { "grad_norm": 0.3617725968360901, "learning_rate": 7.951652511442858e-06, "loss": 0.0243, "step": 66170 }, { "grad_norm": 0.3388327360153198, "learning_rate": 7.940472726973125e-06, "loss": 0.0201, "step": 66180 }, { "grad_norm": 0.21541982889175415, "learning_rate": 7.929300129326289e-06, "loss": 0.018, "step": 66190 }, { "grad_norm": 0.348739892244339, "learning_rate": 7.91813472041142e-06, "loss": 0.0222, "step": 66200 }, { "grad_norm": 0.18993370234966278, "learning_rate": 7.906976502136376e-06, "loss": 0.0178, "step": 66210 }, { "grad_norm": 0.6008769869804382, "learning_rate": 7.89582547640782e-06, "loss": 0.0181, "step": 66220 }, { "grad_norm": 0.33730465173721313, "learning_rate": 7.884681645131115e-06, "loss": 0.0198, "step": 66230 }, { "grad_norm": 0.49886423349380493, "learning_rate": 7.87354501021048e-06, "loss": 0.0167, "step": 66240 }, { "grad_norm": 0.29383453726768494, "learning_rate": 7.86241557354882e-06, "loss": 0.0188, "step": 66250 }, { "grad_norm": 0.3243809640407562, "learning_rate": 7.8512933370479e-06, "loss": 0.0214, "step": 66260 }, { "grad_norm": 0.31786981225013733, "learning_rate": 7.840178302608158e-06, "loss": 0.0241, "step": 66270 }, { "grad_norm": 0.19824890792369843, "learning_rate": 7.82907047212888e-06, "loss": 0.0157, "step": 66280 }, { "grad_norm": 0.30523866415023804, "learning_rate": 7.81796984750809e-06, "loss": 0.0206, "step": 66290 }, { "grad_norm": 0.20706504583358765, "learning_rate": 7.806876430642546e-06, "loss": 0.0185, "step": 66300 }, { "grad_norm": 0.24947446584701538, "learning_rate": 7.795790223427862e-06, "loss": 0.0178, "step": 66310 }, { "grad_norm": 0.325444757938385, "learning_rate": 7.784711227758324e-06, "loss": 0.0212, "step": 66320 }, { "grad_norm": 1.172594666481018, "learning_rate": 7.773639445527053e-06, "loss": 0.0215, "step": 66330 }, { "grad_norm": 0.2342672199010849, "learning_rate": 7.762574878625905e-06, "loss": 0.0192, "step": 66340 }, { "grad_norm": 0.34176111221313477, "learning_rate": 7.751517528945513e-06, "loss": 0.0137, "step": 66350 }, { "grad_norm": 0.36194610595703125, "learning_rate": 7.740467398375278e-06, "loss": 0.0165, "step": 66360 }, { "grad_norm": 0.3858156204223633, "learning_rate": 7.729424488803378e-06, "loss": 0.0153, "step": 66370 }, { "grad_norm": 0.2657317519187927, "learning_rate": 7.71838880211671e-06, "loss": 0.0172, "step": 66380 }, { "grad_norm": 0.2886527478694916, "learning_rate": 7.707360340200997e-06, "loss": 0.018, "step": 66390 }, { "grad_norm": 0.2828902304172516, "learning_rate": 7.696339104940697e-06, "loss": 0.0181, "step": 66400 }, { "grad_norm": 0.24410198628902435, "learning_rate": 7.685325098219038e-06, "loss": 0.0166, "step": 66410 }, { "grad_norm": 0.22983071208000183, "learning_rate": 7.674318321918017e-06, "loss": 0.0196, "step": 66420 }, { "grad_norm": 0.1776379495859146, "learning_rate": 7.663318777918366e-06, "loss": 0.0176, "step": 66430 }, { "grad_norm": 0.25334689021110535, "learning_rate": 7.652326468099647e-06, "loss": 0.0239, "step": 66440 }, { "grad_norm": 0.6069362163543701, "learning_rate": 7.641341394340096e-06, "loss": 0.0176, "step": 66450 }, { "grad_norm": 0.33677637577056885, "learning_rate": 7.630363558516818e-06, "loss": 0.0217, "step": 66460 }, { "grad_norm": 0.3907201290130615, "learning_rate": 7.619392962505578e-06, "loss": 0.0261, "step": 66470 }, { "grad_norm": 0.549440860748291, "learning_rate": 7.6084296081809725e-06, "loss": 0.022, "step": 66480 }, { "grad_norm": 0.4069176912307739, "learning_rate": 7.597473497416347e-06, "loss": 0.0218, "step": 66490 }, { "grad_norm": 0.19235080480575562, "learning_rate": 7.586524632083764e-06, "loss": 0.0148, "step": 66500 }, { "grad_norm": 0.3221823275089264, "learning_rate": 7.5755830140541326e-06, "loss": 0.0172, "step": 66510 }, { "grad_norm": 0.16213765740394592, "learning_rate": 7.5646486451970475e-06, "loss": 0.0188, "step": 66520 }, { "grad_norm": 0.31089064478874207, "learning_rate": 7.553721527380897e-06, "loss": 0.0279, "step": 66530 }, { "grad_norm": 0.3071037232875824, "learning_rate": 7.542801662472826e-06, "loss": 0.0159, "step": 66540 }, { "grad_norm": 0.2395133227109909, "learning_rate": 7.5318890523387475e-06, "loss": 0.0181, "step": 66550 }, { "grad_norm": 0.3333837687969208, "learning_rate": 7.52098369884332e-06, "loss": 0.0215, "step": 66560 }, { "grad_norm": 0.15695655345916748, "learning_rate": 7.510085603849992e-06, "loss": 0.0168, "step": 66570 }, { "grad_norm": 0.21146750450134277, "learning_rate": 7.499194769220918e-06, "loss": 0.0186, "step": 66580 }, { "grad_norm": 0.37190723419189453, "learning_rate": 7.48831119681706e-06, "loss": 0.0244, "step": 66590 }, { "grad_norm": 0.2939205467700958, "learning_rate": 7.477434888498119e-06, "loss": 0.0168, "step": 66600 }, { "grad_norm": 0.3669973313808441, "learning_rate": 7.466565846122564e-06, "loss": 0.0247, "step": 66610 }, { "grad_norm": 0.565683126449585, "learning_rate": 7.455704071547626e-06, "loss": 0.0199, "step": 66620 }, { "grad_norm": 0.3082376718521118, "learning_rate": 7.444849566629247e-06, "loss": 0.0183, "step": 66630 }, { "grad_norm": 0.35137811303138733, "learning_rate": 7.434002333222212e-06, "loss": 0.0145, "step": 66640 }, { "grad_norm": 0.36013683676719666, "learning_rate": 7.423162373179976e-06, "loss": 0.0239, "step": 66650 }, { "grad_norm": 0.32158133387565613, "learning_rate": 7.412329688354835e-06, "loss": 0.0192, "step": 66660 }, { "grad_norm": 0.18024764955043793, "learning_rate": 7.40150428059776e-06, "loss": 0.0156, "step": 66670 }, { "grad_norm": 0.4829274117946625, "learning_rate": 7.3906861517585354e-06, "loss": 0.0192, "step": 66680 }, { "grad_norm": 0.2752749025821686, "learning_rate": 7.37987530368569e-06, "loss": 0.0182, "step": 66690 }, { "grad_norm": 0.372783362865448, "learning_rate": 7.369071738226474e-06, "loss": 0.0167, "step": 66700 }, { "grad_norm": 0.28934723138809204, "learning_rate": 7.358275457226954e-06, "loss": 0.0214, "step": 66710 }, { "grad_norm": 0.2850050926208496, "learning_rate": 7.347486462531899e-06, "loss": 0.0168, "step": 66720 }, { "grad_norm": 0.26504066586494446, "learning_rate": 7.336704755984858e-06, "loss": 0.0194, "step": 66730 }, { "grad_norm": 0.20119665563106537, "learning_rate": 7.325930339428133e-06, "loss": 0.0211, "step": 66740 }, { "grad_norm": 0.2446516752243042, "learning_rate": 7.315163214702769e-06, "loss": 0.0202, "step": 66750 }, { "grad_norm": 0.31222033500671387, "learning_rate": 7.304403383648595e-06, "loss": 0.0206, "step": 66760 }, { "grad_norm": 0.2264283001422882, "learning_rate": 7.293650848104139e-06, "loss": 0.0192, "step": 66770 }, { "grad_norm": 0.32226356863975525, "learning_rate": 7.2829056099067374e-06, "loss": 0.0197, "step": 66780 }, { "grad_norm": 0.3085905909538269, "learning_rate": 7.2721676708924494e-06, "loss": 0.016, "step": 66790 }, { "grad_norm": 0.43674635887145996, "learning_rate": 7.261437032896096e-06, "loss": 0.0293, "step": 66800 }, { "grad_norm": 0.27160748839378357, "learning_rate": 7.250713697751255e-06, "loss": 0.0214, "step": 66810 }, { "grad_norm": 0.17144744098186493, "learning_rate": 7.239997667290255e-06, "loss": 0.015, "step": 66820 }, { "grad_norm": 0.1939915269613266, "learning_rate": 7.2292889433441425e-06, "loss": 0.0177, "step": 66830 }, { "grad_norm": 0.3092845380306244, "learning_rate": 7.218587527742793e-06, "loss": 0.0133, "step": 66840 }, { "grad_norm": 0.533862292766571, "learning_rate": 7.207893422314749e-06, "loss": 0.021, "step": 66850 }, { "grad_norm": 0.1840783655643463, "learning_rate": 7.1972066288873545e-06, "loss": 0.0152, "step": 66860 }, { "grad_norm": 0.21258544921875, "learning_rate": 7.186527149286687e-06, "loss": 0.0324, "step": 66870 }, { "grad_norm": 0.22902488708496094, "learning_rate": 7.175854985337576e-06, "loss": 0.0177, "step": 66880 }, { "grad_norm": 0.2694411277770996, "learning_rate": 7.16519013886362e-06, "loss": 0.021, "step": 66890 }, { "grad_norm": 0.436219722032547, "learning_rate": 7.154532611687109e-06, "loss": 0.0165, "step": 66900 }, { "grad_norm": 0.32008859515190125, "learning_rate": 7.143882405629177e-06, "loss": 0.0147, "step": 66910 }, { "grad_norm": 0.4646762013435364, "learning_rate": 7.133239522509605e-06, "loss": 0.0137, "step": 66920 }, { "grad_norm": 0.35959482192993164, "learning_rate": 7.1226039641469956e-06, "loss": 0.0179, "step": 66930 }, { "grad_norm": 0.23518085479736328, "learning_rate": 7.111975732358678e-06, "loss": 0.018, "step": 66940 }, { "grad_norm": 1.3811488151550293, "learning_rate": 7.101354828960693e-06, "loss": 0.0183, "step": 66950 }, { "grad_norm": 0.3020351827144623, "learning_rate": 7.090741255767918e-06, "loss": 0.0153, "step": 66960 }, { "grad_norm": 0.31796520948410034, "learning_rate": 7.080135014593875e-06, "loss": 0.0213, "step": 66970 }, { "grad_norm": 0.3809508979320526, "learning_rate": 7.069536107250896e-06, "loss": 0.019, "step": 66980 }, { "grad_norm": 0.5001930594444275, "learning_rate": 7.058944535550049e-06, "loss": 0.0161, "step": 66990 }, { "grad_norm": 0.23081405460834503, "learning_rate": 7.048360301301138e-06, "loss": 0.0169, "step": 67000 }, { "grad_norm": 0.1912052184343338, "learning_rate": 7.03778340631272e-06, "loss": 0.0189, "step": 67010 }, { "grad_norm": 0.23745346069335938, "learning_rate": 7.0272138523921e-06, "loss": 0.0207, "step": 67020 }, { "grad_norm": 0.2900422513484955, "learning_rate": 7.016651641345334e-06, "loss": 0.0218, "step": 67030 }, { "grad_norm": 0.3420645296573639, "learning_rate": 7.0060967749772e-06, "loss": 0.0174, "step": 67040 }, { "grad_norm": 0.5138292908668518, "learning_rate": 6.995549255091238e-06, "loss": 0.0328, "step": 67050 }, { "grad_norm": 0.2446611076593399, "learning_rate": 6.98500908348973e-06, "loss": 0.0208, "step": 67060 }, { "grad_norm": 0.26528704166412354, "learning_rate": 6.974476261973711e-06, "loss": 0.0196, "step": 67070 }, { "grad_norm": 0.1974492371082306, "learning_rate": 6.963950792342949e-06, "loss": 0.0144, "step": 67080 }, { "grad_norm": 0.2387256920337677, "learning_rate": 6.9534326763959715e-06, "loss": 0.0147, "step": 67090 }, { "grad_norm": 0.8444380164146423, "learning_rate": 6.942921915929995e-06, "loss": 0.0247, "step": 67100 }, { "grad_norm": 0.2259713113307953, "learning_rate": 6.9324185127410735e-06, "loss": 0.0166, "step": 67110 }, { "grad_norm": 0.3227326571941376, "learning_rate": 6.921922468623921e-06, "loss": 0.0225, "step": 67120 }, { "grad_norm": 0.2981090843677521, "learning_rate": 6.911433785372023e-06, "loss": 0.0154, "step": 67130 }, { "grad_norm": 0.20537452399730682, "learning_rate": 6.900952464777632e-06, "loss": 0.0169, "step": 67140 }, { "grad_norm": 0.32750582695007324, "learning_rate": 6.8904785086316815e-06, "loss": 0.0235, "step": 67150 }, { "grad_norm": 0.4997003674507141, "learning_rate": 6.880011918723927e-06, "loss": 0.0164, "step": 67160 }, { "grad_norm": 0.2843892276287079, "learning_rate": 6.86955269684279e-06, "loss": 0.0174, "step": 67170 }, { "grad_norm": 0.47154226899147034, "learning_rate": 6.859100844775473e-06, "loss": 0.0287, "step": 67180 }, { "grad_norm": 0.17539525032043457, "learning_rate": 6.84865636430792e-06, "loss": 0.0213, "step": 67190 }, { "grad_norm": 0.2352449744939804, "learning_rate": 6.838219257224804e-06, "loss": 0.0154, "step": 67200 }, { "grad_norm": 0.15129730105400085, "learning_rate": 6.827789525309536e-06, "loss": 0.019, "step": 67210 }, { "grad_norm": 0.41744205355644226, "learning_rate": 6.81736717034428e-06, "loss": 0.0192, "step": 67220 }, { "grad_norm": 0.4263651669025421, "learning_rate": 6.806952194109933e-06, "loss": 0.0312, "step": 67230 }, { "grad_norm": 0.3308121860027313, "learning_rate": 6.796544598386112e-06, "loss": 0.0169, "step": 67240 }, { "grad_norm": 0.32540664076805115, "learning_rate": 6.786144384951204e-06, "loss": 0.0206, "step": 67250 }, { "grad_norm": 0.2926005423069, "learning_rate": 6.775751555582322e-06, "loss": 0.0152, "step": 67260 }, { "grad_norm": 0.28122061491012573, "learning_rate": 6.76536611205531e-06, "loss": 0.0223, "step": 67270 }, { "grad_norm": 0.18876977264881134, "learning_rate": 6.754988056144762e-06, "loss": 0.0173, "step": 67280 }, { "grad_norm": 0.27339422702789307, "learning_rate": 6.744617389624014e-06, "loss": 0.0199, "step": 67290 }, { "grad_norm": 0.31715133786201477, "learning_rate": 6.734254114265087e-06, "loss": 0.0216, "step": 67300 }, { "grad_norm": 0.41858628392219543, "learning_rate": 6.723898231838843e-06, "loss": 0.0197, "step": 67310 }, { "grad_norm": 0.16699428856372833, "learning_rate": 6.713549744114766e-06, "loss": 0.0181, "step": 67320 }, { "grad_norm": 0.22062788903713226, "learning_rate": 6.703208652861159e-06, "loss": 0.0177, "step": 67330 }, { "grad_norm": 0.4123510718345642, "learning_rate": 6.69287495984503e-06, "loss": 0.0232, "step": 67340 }, { "grad_norm": 0.3742798864841461, "learning_rate": 6.6825486668321e-06, "loss": 0.0157, "step": 67350 }, { "grad_norm": 0.33552291989326477, "learning_rate": 6.672229775586886e-06, "loss": 0.0216, "step": 67360 }, { "grad_norm": 0.26258301734924316, "learning_rate": 6.661918287872576e-06, "loss": 0.0182, "step": 67370 }, { "grad_norm": 0.14890456199645996, "learning_rate": 6.6516142054511346e-06, "loss": 0.0212, "step": 67380 }, { "grad_norm": 0.3454698622226715, "learning_rate": 6.641317530083241e-06, "loss": 0.0156, "step": 67390 }, { "grad_norm": 0.23894968628883362, "learning_rate": 6.631028263528322e-06, "loss": 0.0169, "step": 67400 }, { "grad_norm": 0.28235700726509094, "learning_rate": 6.620746407544537e-06, "loss": 0.0194, "step": 67410 }, { "grad_norm": 0.29370731115341187, "learning_rate": 6.610471963888742e-06, "loss": 0.0209, "step": 67420 }, { "grad_norm": 0.2972545325756073, "learning_rate": 6.600204934316606e-06, "loss": 0.0144, "step": 67430 }, { "grad_norm": 0.20444580912590027, "learning_rate": 6.589945320582452e-06, "loss": 0.0141, "step": 67440 }, { "grad_norm": 0.2874758243560791, "learning_rate": 6.579693124439374e-06, "loss": 0.0211, "step": 67450 }, { "grad_norm": 0.32096943259239197, "learning_rate": 6.56944834763919e-06, "loss": 0.0153, "step": 67460 }, { "grad_norm": 0.3309219479560852, "learning_rate": 6.5592109919324575e-06, "loss": 0.0173, "step": 67470 }, { "grad_norm": 0.2645427882671356, "learning_rate": 6.54898105906846e-06, "loss": 0.0239, "step": 67480 }, { "grad_norm": 0.2277437299489975, "learning_rate": 6.5387585507952155e-06, "loss": 0.021, "step": 67490 }, { "grad_norm": 0.2520219087600708, "learning_rate": 6.528543468859461e-06, "loss": 0.0196, "step": 67500 }, { "grad_norm": 0.44068557024002075, "learning_rate": 6.518335815006682e-06, "loss": 0.0245, "step": 67510 }, { "grad_norm": 0.33215391635894775, "learning_rate": 6.5081355909810845e-06, "loss": 0.0143, "step": 67520 }, { "grad_norm": 0.38466349244117737, "learning_rate": 6.497942798525608e-06, "loss": 0.0214, "step": 67530 }, { "grad_norm": 0.9226799011230469, "learning_rate": 6.487757439381936e-06, "loss": 0.0233, "step": 67540 }, { "grad_norm": 0.28714022040367126, "learning_rate": 6.477579515290433e-06, "loss": 0.0178, "step": 67550 }, { "grad_norm": 0.18555869162082672, "learning_rate": 6.467409027990273e-06, "loss": 0.023, "step": 67560 }, { "grad_norm": 0.35980138182640076, "learning_rate": 6.457245979219279e-06, "loss": 0.0207, "step": 67570 }, { "grad_norm": 0.3149774670600891, "learning_rate": 6.447090370714054e-06, "loss": 0.0286, "step": 67580 }, { "grad_norm": 0.4662572145462036, "learning_rate": 6.436942204209917e-06, "loss": 0.0213, "step": 67590 }, { "grad_norm": 0.21310937404632568, "learning_rate": 6.4268014814408804e-06, "loss": 0.0258, "step": 67600 }, { "grad_norm": 0.5339140295982361, "learning_rate": 6.41666820413977e-06, "loss": 0.0211, "step": 67610 }, { "grad_norm": 0.2570440471172333, "learning_rate": 6.406542374038033e-06, "loss": 0.0179, "step": 67620 }, { "grad_norm": 0.2491789162158966, "learning_rate": 6.396423992865935e-06, "loss": 0.0164, "step": 67630 }, { "grad_norm": 0.20520102977752686, "learning_rate": 6.386313062352412e-06, "loss": 0.0193, "step": 67640 }, { "grad_norm": 0.32411473989486694, "learning_rate": 6.376209584225152e-06, "loss": 0.0134, "step": 67650 }, { "grad_norm": 0.2517563998699188, "learning_rate": 6.366113560210557e-06, "loss": 0.0195, "step": 67660 }, { "grad_norm": 0.37828052043914795, "learning_rate": 6.356024992033766e-06, "loss": 0.0234, "step": 67670 }, { "grad_norm": 0.30911239981651306, "learning_rate": 6.345943881418648e-06, "loss": 0.0164, "step": 67680 }, { "grad_norm": 0.5282211899757385, "learning_rate": 6.33587023008777e-06, "loss": 0.0153, "step": 67690 }, { "grad_norm": 0.24555575847625732, "learning_rate": 6.32580403976245e-06, "loss": 0.016, "step": 67700 }, { "grad_norm": 0.3077729642391205, "learning_rate": 6.31574531216273e-06, "loss": 0.0165, "step": 67710 }, { "grad_norm": 0.3162842392921448, "learning_rate": 6.305694049007371e-06, "loss": 0.0178, "step": 67720 }, { "grad_norm": 0.6004505157470703, "learning_rate": 6.2956502520138575e-06, "loss": 0.0164, "step": 67730 }, { "grad_norm": 0.36442312598228455, "learning_rate": 6.285613922898409e-06, "loss": 0.0133, "step": 67740 }, { "grad_norm": 0.3347938358783722, "learning_rate": 6.275585063375927e-06, "loss": 0.0221, "step": 67750 }, { "grad_norm": 0.34187665581703186, "learning_rate": 6.265563675160113e-06, "loss": 0.0184, "step": 67760 }, { "grad_norm": 0.3648822009563446, "learning_rate": 6.25554975996332e-06, "loss": 0.0178, "step": 67770 }, { "grad_norm": 0.3649149537086487, "learning_rate": 6.245543319496661e-06, "loss": 0.0166, "step": 67780 }, { "grad_norm": 0.33095356822013855, "learning_rate": 6.2355443554699685e-06, "loss": 0.019, "step": 67790 }, { "grad_norm": 0.2793204188346863, "learning_rate": 6.225552869591766e-06, "loss": 0.0177, "step": 67800 }, { "grad_norm": 0.317452996969223, "learning_rate": 6.215568863569365e-06, "loss": 0.0185, "step": 67810 }, { "grad_norm": 0.44910308718681335, "learning_rate": 6.205592339108712e-06, "loss": 0.0243, "step": 67820 }, { "grad_norm": 0.2543931305408478, "learning_rate": 6.195623297914577e-06, "loss": 0.0172, "step": 67830 }, { "grad_norm": 0.3058653771877289, "learning_rate": 6.185661741690357e-06, "loss": 0.0171, "step": 67840 }, { "grad_norm": 0.1875969022512436, "learning_rate": 6.1757076721382145e-06, "loss": 0.0213, "step": 67850 }, { "grad_norm": 0.2471097707748413, "learning_rate": 6.165761090959038e-06, "loss": 0.0217, "step": 67860 }, { "grad_norm": 0.2664788067340851, "learning_rate": 6.155821999852424e-06, "loss": 0.0167, "step": 67870 }, { "grad_norm": 0.12183800339698792, "learning_rate": 6.145890400516696e-06, "loss": 0.0172, "step": 67880 }, { "grad_norm": 0.4039529263973236, "learning_rate": 6.1359662946488816e-06, "loss": 0.0165, "step": 67890 }, { "grad_norm": 0.2755805552005768, "learning_rate": 6.12604968394474e-06, "loss": 0.0174, "step": 67900 }, { "grad_norm": 0.3511403203010559, "learning_rate": 6.11614057009875e-06, "loss": 0.0235, "step": 67910 }, { "grad_norm": 0.3073151111602783, "learning_rate": 6.106238954804111e-06, "loss": 0.0171, "step": 67920 }, { "grad_norm": 0.3331543803215027, "learning_rate": 6.096344839752738e-06, "loss": 0.0186, "step": 67930 }, { "grad_norm": 0.19955749809741974, "learning_rate": 6.086458226635278e-06, "loss": 0.0194, "step": 67940 }, { "grad_norm": 0.3281504511833191, "learning_rate": 6.076579117141046e-06, "loss": 0.0258, "step": 67950 }, { "grad_norm": 0.31324461102485657, "learning_rate": 6.066707512958153e-06, "loss": 0.016, "step": 67960 }, { "grad_norm": 0.22753237187862396, "learning_rate": 6.056843415773361e-06, "loss": 0.0169, "step": 67970 }, { "grad_norm": 0.2745470404624939, "learning_rate": 6.0469868272721776e-06, "loss": 0.0172, "step": 67980 }, { "grad_norm": 0.7026659846305847, "learning_rate": 6.037137749138844e-06, "loss": 0.0254, "step": 67990 }, { "grad_norm": 0.3216991424560547, "learning_rate": 6.027296183056252e-06, "loss": 0.019, "step": 68000 }, { "grad_norm": 0.3124057650566101, "learning_rate": 6.017462130706114e-06, "loss": 0.0168, "step": 68010 }, { "grad_norm": 0.32561904191970825, "learning_rate": 6.0076355937687516e-06, "loss": 0.0202, "step": 68020 }, { "grad_norm": 0.21575215458869934, "learning_rate": 5.9978165739232925e-06, "loss": 0.0174, "step": 68030 }, { "grad_norm": 0.19731032848358154, "learning_rate": 5.988005072847508e-06, "loss": 0.0169, "step": 68040 }, { "grad_norm": 0.46245041489601135, "learning_rate": 5.978201092217928e-06, "loss": 0.0259, "step": 68050 }, { "grad_norm": 0.20754484832286835, "learning_rate": 5.9684046337097895e-06, "loss": 0.0136, "step": 68060 }, { "grad_norm": 0.1910625547170639, "learning_rate": 5.958615698997016e-06, "loss": 0.018, "step": 68070 }, { "grad_norm": 0.3473137617111206, "learning_rate": 5.948834289752303e-06, "loss": 0.0181, "step": 68080 }, { "grad_norm": 0.26936909556388855, "learning_rate": 5.939060407646996e-06, "loss": 0.019, "step": 68090 }, { "grad_norm": 0.27100029587745667, "learning_rate": 5.929294054351198e-06, "loss": 0.0179, "step": 68100 }, { "grad_norm": 0.31162720918655396, "learning_rate": 5.919535231533707e-06, "loss": 0.0181, "step": 68110 }, { "grad_norm": 0.26744717359542847, "learning_rate": 5.909783940862046e-06, "loss": 0.0158, "step": 68120 }, { "grad_norm": 0.24330613017082214, "learning_rate": 5.900040184002436e-06, "loss": 0.0158, "step": 68130 }, { "grad_norm": 0.308474063873291, "learning_rate": 5.890303962619831e-06, "loss": 0.0165, "step": 68140 }, { "grad_norm": 0.3899141550064087, "learning_rate": 5.88057527837787e-06, "loss": 0.0234, "step": 68150 }, { "grad_norm": 0.3552807867527008, "learning_rate": 5.870854132938924e-06, "loss": 0.0236, "step": 68160 }, { "grad_norm": 0.3542662262916565, "learning_rate": 5.861140527964071e-06, "loss": 0.0183, "step": 68170 }, { "grad_norm": 0.22133703529834747, "learning_rate": 5.851434465113098e-06, "loss": 0.0133, "step": 68180 }, { "grad_norm": 0.4072325527667999, "learning_rate": 5.841735946044524e-06, "loss": 0.0167, "step": 68190 }, { "grad_norm": 0.18336930871009827, "learning_rate": 5.832044972415523e-06, "loss": 0.0182, "step": 68200 }, { "grad_norm": 0.18788841366767883, "learning_rate": 5.8223615458820655e-06, "loss": 0.0207, "step": 68210 }, { "grad_norm": 0.27659016847610474, "learning_rate": 5.812685668098733e-06, "loss": 0.0177, "step": 68220 }, { "grad_norm": 0.168935164809227, "learning_rate": 5.803017340718913e-06, "loss": 0.02, "step": 68230 }, { "grad_norm": 0.2563614249229431, "learning_rate": 5.793356565394636e-06, "loss": 0.0151, "step": 68240 }, { "grad_norm": 0.23134054243564606, "learning_rate": 5.783703343776658e-06, "loss": 0.0176, "step": 68250 }, { "grad_norm": 0.2005244493484497, "learning_rate": 5.774057677514477e-06, "loss": 0.0144, "step": 68260 }, { "grad_norm": 0.2661445736885071, "learning_rate": 5.764419568256235e-06, "loss": 0.0163, "step": 68270 }, { "grad_norm": 0.4453794062137604, "learning_rate": 5.754789017648865e-06, "loss": 0.0353, "step": 68280 }, { "grad_norm": 0.1968599408864975, "learning_rate": 5.745166027337934e-06, "loss": 0.0221, "step": 68290 }, { "grad_norm": 0.18979589641094208, "learning_rate": 5.735550598967754e-06, "loss": 0.0181, "step": 68300 }, { "grad_norm": 0.40674078464508057, "learning_rate": 5.725942734181339e-06, "loss": 0.0207, "step": 68310 }, { "grad_norm": 0.32555195689201355, "learning_rate": 5.7163424346204146e-06, "loss": 0.0186, "step": 68320 }, { "grad_norm": 0.3341864049434662, "learning_rate": 5.706749701925407e-06, "loss": 0.017, "step": 68330 }, { "grad_norm": 0.24177932739257812, "learning_rate": 5.697164537735461e-06, "loss": 0.0155, "step": 68340 }, { "grad_norm": 0.17927499115467072, "learning_rate": 5.687586943688406e-06, "loss": 0.0175, "step": 68350 }, { "grad_norm": 0.2556787133216858, "learning_rate": 5.678016921420792e-06, "loss": 0.0164, "step": 68360 }, { "grad_norm": 0.20512790977954865, "learning_rate": 5.668454472567875e-06, "loss": 0.024, "step": 68370 }, { "grad_norm": 0.40703198313713074, "learning_rate": 5.658899598763617e-06, "loss": 0.0193, "step": 68380 }, { "grad_norm": 0.3116375505924225, "learning_rate": 5.6493523016407035e-06, "loss": 0.0143, "step": 68390 }, { "grad_norm": 0.14705587923526764, "learning_rate": 5.6398125828304615e-06, "loss": 0.0188, "step": 68400 }, { "grad_norm": 0.3495938777923584, "learning_rate": 5.630280443963015e-06, "loss": 0.0213, "step": 68410 }, { "grad_norm": 0.2908940017223358, "learning_rate": 5.620755886667112e-06, "loss": 0.0192, "step": 68420 }, { "grad_norm": 0.2615600526332855, "learning_rate": 5.611238912570271e-06, "loss": 0.0155, "step": 68430 }, { "grad_norm": 0.47452181577682495, "learning_rate": 5.601729523298649e-06, "loss": 0.027, "step": 68440 }, { "grad_norm": 0.2661920189857483, "learning_rate": 5.592227720477161e-06, "loss": 0.0207, "step": 68450 }, { "grad_norm": 0.34612590074539185, "learning_rate": 5.582733505729415e-06, "loss": 0.0205, "step": 68460 }, { "grad_norm": 0.3034812808036804, "learning_rate": 5.573246880677668e-06, "loss": 0.0195, "step": 68470 }, { "grad_norm": 0.23768995702266693, "learning_rate": 5.56376784694298e-06, "loss": 0.0207, "step": 68480 }, { "grad_norm": 0.4173407554626465, "learning_rate": 5.554296406145027e-06, "loss": 0.0215, "step": 68490 }, { "grad_norm": 0.29612594842910767, "learning_rate": 5.544832559902219e-06, "loss": 0.0179, "step": 68500 }, { "grad_norm": 0.22610995173454285, "learning_rate": 5.53537630983168e-06, "loss": 0.0232, "step": 68510 }, { "grad_norm": 0.215338334441185, "learning_rate": 5.5259276575492125e-06, "loss": 0.0199, "step": 68520 }, { "grad_norm": 0.32006368041038513, "learning_rate": 5.516486604669357e-06, "loss": 0.0224, "step": 68530 }, { "grad_norm": 0.5618679523468018, "learning_rate": 5.507053152805303e-06, "loss": 0.0246, "step": 68540 }, { "grad_norm": 0.24691620469093323, "learning_rate": 5.497627303568975e-06, "loss": 0.0221, "step": 68550 }, { "grad_norm": 0.15000319480895996, "learning_rate": 5.488209058571003e-06, "loss": 0.0136, "step": 68560 }, { "grad_norm": 0.23671887814998627, "learning_rate": 5.4787984194207054e-06, "loss": 0.0189, "step": 68570 }, { "grad_norm": 0.26619380712509155, "learning_rate": 5.469395387726095e-06, "loss": 0.0197, "step": 68580 }, { "grad_norm": 0.376139372587204, "learning_rate": 5.45999996509391e-06, "loss": 0.0156, "step": 68590 }, { "grad_norm": 0.31980133056640625, "learning_rate": 5.450612153129536e-06, "loss": 0.0212, "step": 68600 }, { "grad_norm": 0.36666950583457947, "learning_rate": 5.4412319534371426e-06, "loss": 0.024, "step": 68610 }, { "grad_norm": 0.3217290937900543, "learning_rate": 5.431859367619513e-06, "loss": 0.0191, "step": 68620 }, { "grad_norm": 0.3056755065917969, "learning_rate": 5.422494397278172e-06, "loss": 0.0183, "step": 68630 }, { "grad_norm": 0.3357579708099365, "learning_rate": 5.413137044013344e-06, "loss": 0.0186, "step": 68640 }, { "grad_norm": 0.2559845447540283, "learning_rate": 5.403787309423941e-06, "loss": 0.0213, "step": 68650 }, { "grad_norm": 0.3101915717124939, "learning_rate": 5.3944451951075835e-06, "loss": 0.0178, "step": 68660 }, { "grad_norm": 0.244329571723938, "learning_rate": 5.385110702660562e-06, "loss": 0.0214, "step": 68670 }, { "grad_norm": 0.10802048444747925, "learning_rate": 5.375783833677922e-06, "loss": 0.0228, "step": 68680 }, { "grad_norm": 0.3697652518749237, "learning_rate": 5.36646458975334e-06, "loss": 0.0208, "step": 68690 }, { "grad_norm": 0.2610010504722595, "learning_rate": 5.3571529724792294e-06, "loss": 0.0155, "step": 68700 }, { "grad_norm": 0.38130512833595276, "learning_rate": 5.347848983446702e-06, "loss": 0.0188, "step": 68710 }, { "grad_norm": 0.20599304139614105, "learning_rate": 5.3385526242455185e-06, "loss": 0.0208, "step": 68720 }, { "grad_norm": 0.40685319900512695, "learning_rate": 5.329263896464226e-06, "loss": 0.017, "step": 68730 }, { "grad_norm": 0.23560751974582672, "learning_rate": 5.3199828016899715e-06, "loss": 0.0157, "step": 68740 }, { "grad_norm": 0.26161903142929077, "learning_rate": 5.310709341508657e-06, "loss": 0.0186, "step": 68750 }, { "grad_norm": 0.22755885124206543, "learning_rate": 5.301443517504861e-06, "loss": 0.0177, "step": 68760 }, { "grad_norm": 0.2839200794696808, "learning_rate": 5.292185331261862e-06, "loss": 0.0141, "step": 68770 }, { "grad_norm": 0.3586856722831726, "learning_rate": 5.28293478436162e-06, "loss": 0.018, "step": 68780 }, { "grad_norm": 0.29923462867736816, "learning_rate": 5.273691878384829e-06, "loss": 0.02, "step": 68790 }, { "grad_norm": 0.3193727135658264, "learning_rate": 5.264456614910812e-06, "loss": 0.0199, "step": 68800 }, { "grad_norm": 0.27016618847846985, "learning_rate": 5.255228995517647e-06, "loss": 0.0255, "step": 68810 }, { "grad_norm": 0.2446564882993698, "learning_rate": 5.246009021782067e-06, "loss": 0.0183, "step": 68820 }, { "grad_norm": 0.22624577581882477, "learning_rate": 5.2367966952795225e-06, "loss": 0.025, "step": 68830 }, { "grad_norm": 0.30512773990631104, "learning_rate": 5.2275920175841485e-06, "loss": 0.0154, "step": 68840 }, { "grad_norm": 0.1732935607433319, "learning_rate": 5.218394990268771e-06, "loss": 0.0222, "step": 68850 }, { "grad_norm": 0.2826295495033264, "learning_rate": 5.209205614904916e-06, "loss": 0.015, "step": 68860 }, { "grad_norm": 0.29671910405158997, "learning_rate": 5.200023893062777e-06, "loss": 0.0192, "step": 68870 }, { "grad_norm": 0.4071011245250702, "learning_rate": 5.190849826311289e-06, "loss": 0.0175, "step": 68880 }, { "grad_norm": 0.31889915466308594, "learning_rate": 5.181683416218025e-06, "loss": 0.0193, "step": 68890 }, { "grad_norm": 0.36378830671310425, "learning_rate": 5.172524664349276e-06, "loss": 0.0173, "step": 68900 }, { "grad_norm": 0.24938784539699554, "learning_rate": 5.1633735722700416e-06, "loss": 0.0202, "step": 68910 }, { "grad_norm": 0.5071117281913757, "learning_rate": 5.154230141543958e-06, "loss": 0.019, "step": 68920 }, { "grad_norm": 0.5213088989257812, "learning_rate": 5.145094373733433e-06, "loss": 0.017, "step": 68930 }, { "grad_norm": 0.3210793733596802, "learning_rate": 5.135966270399478e-06, "loss": 0.0145, "step": 68940 }, { "grad_norm": 0.5911820530891418, "learning_rate": 5.126845833101857e-06, "loss": 0.0228, "step": 68950 }, { "grad_norm": 0.1585512012243271, "learning_rate": 5.1177330633989994e-06, "loss": 0.0155, "step": 68960 }, { "grad_norm": 0.5679561495780945, "learning_rate": 5.108627962848033e-06, "loss": 0.035, "step": 68970 }, { "grad_norm": 0.4473973512649536, "learning_rate": 5.099530533004759e-06, "loss": 0.0234, "step": 68980 }, { "grad_norm": 0.3072747588157654, "learning_rate": 5.090440775423699e-06, "loss": 0.0151, "step": 68990 }, { "grad_norm": 0.26929837465286255, "learning_rate": 5.081358691658022e-06, "loss": 0.0158, "step": 69000 }, { "grad_norm": 0.30676761269569397, "learning_rate": 5.072284283259621e-06, "loss": 0.0226, "step": 69010 }, { "grad_norm": 0.1596948206424713, "learning_rate": 5.063217551779053e-06, "loss": 0.0148, "step": 69020 }, { "grad_norm": 0.35326120257377625, "learning_rate": 5.054158498765588e-06, "loss": 0.0149, "step": 69030 }, { "grad_norm": 0.3376810848712921, "learning_rate": 5.0451071257671625e-06, "loss": 0.0173, "step": 69040 }, { "grad_norm": 0.4680936932563782, "learning_rate": 5.036063434330407e-06, "loss": 0.02, "step": 69050 }, { "grad_norm": 0.29719826579093933, "learning_rate": 5.027027426000652e-06, "loss": 0.0182, "step": 69060 }, { "grad_norm": 0.5415589809417725, "learning_rate": 5.017999102321886e-06, "loss": 0.017, "step": 69070 }, { "grad_norm": 0.29276323318481445, "learning_rate": 5.0089784648368224e-06, "loss": 0.0179, "step": 69080 }, { "grad_norm": 0.2957035005092621, "learning_rate": 4.999965515086829e-06, "loss": 0.0169, "step": 69090 }, { "grad_norm": 0.294704407453537, "learning_rate": 4.990960254611976e-06, "loss": 0.0169, "step": 69100 }, { "grad_norm": 0.40771257877349854, "learning_rate": 4.9819626849510194e-06, "loss": 0.0214, "step": 69110 }, { "grad_norm": 0.20960482954978943, "learning_rate": 4.97297280764138e-06, "loss": 0.0219, "step": 69120 }, { "grad_norm": 0.3260379135608673, "learning_rate": 4.963990624219211e-06, "loss": 0.0183, "step": 69130 }, { "grad_norm": 0.1409696787595749, "learning_rate": 4.955016136219298e-06, "loss": 0.0174, "step": 69140 }, { "grad_norm": 0.26194679737091064, "learning_rate": 4.946049345175146e-06, "loss": 0.015, "step": 69150 }, { "grad_norm": 0.41980502009391785, "learning_rate": 4.937090252618937e-06, "loss": 0.0163, "step": 69160 }, { "grad_norm": 0.17712657153606415, "learning_rate": 4.928138860081521e-06, "loss": 0.0206, "step": 69170 }, { "grad_norm": 0.24538396298885345, "learning_rate": 4.919195169092472e-06, "loss": 0.019, "step": 69180 }, { "grad_norm": 0.28564825654029846, "learning_rate": 4.910259181179994e-06, "loss": 0.0193, "step": 69190 }, { "grad_norm": 0.20157304406166077, "learning_rate": 4.901330897871015e-06, "loss": 0.0228, "step": 69200 }, { "grad_norm": 0.27414464950561523, "learning_rate": 4.8924103206911375e-06, "loss": 0.0195, "step": 69210 }, { "grad_norm": 0.31837570667266846, "learning_rate": 4.883497451164637e-06, "loss": 0.0241, "step": 69220 }, { "grad_norm": 0.40281808376312256, "learning_rate": 4.874592290814484e-06, "loss": 0.0183, "step": 69230 }, { "grad_norm": 0.4630926251411438, "learning_rate": 4.865694841162327e-06, "loss": 0.0245, "step": 69240 }, { "grad_norm": 0.146310493350029, "learning_rate": 4.856805103728496e-06, "loss": 0.0168, "step": 69250 }, { "grad_norm": 0.5673907995223999, "learning_rate": 4.847923080032007e-06, "loss": 0.0174, "step": 69260 }, { "grad_norm": 0.4140741527080536, "learning_rate": 4.8390487715905486e-06, "loss": 0.0138, "step": 69270 }, { "grad_norm": 0.2498014271259308, "learning_rate": 4.8301821799205e-06, "loss": 0.0196, "step": 69280 }, { "grad_norm": 0.22511152923107147, "learning_rate": 4.821323306536918e-06, "loss": 0.0146, "step": 69290 }, { "grad_norm": 0.3724518418312073, "learning_rate": 4.8124721529535455e-06, "loss": 0.0153, "step": 69300 }, { "grad_norm": 0.5411677956581116, "learning_rate": 4.803628720682807e-06, "loss": 0.0239, "step": 69310 }, { "grad_norm": 0.4078406095504761, "learning_rate": 4.794793011235776e-06, "loss": 0.0182, "step": 69320 }, { "grad_norm": 0.26919034123420715, "learning_rate": 4.78596502612228e-06, "loss": 0.0191, "step": 69330 }, { "grad_norm": 0.2562040686607361, "learning_rate": 4.777144766850738e-06, "loss": 0.0228, "step": 69340 }, { "grad_norm": 0.42828619480133057, "learning_rate": 4.768332234928313e-06, "loss": 0.0189, "step": 69350 }, { "grad_norm": 0.2680705487728119, "learning_rate": 4.759527431860828e-06, "loss": 0.0152, "step": 69360 }, { "grad_norm": 0.4195023775100708, "learning_rate": 4.750730359152755e-06, "loss": 0.021, "step": 69370 }, { "grad_norm": 0.1574394553899765, "learning_rate": 4.741941018307311e-06, "loss": 0.0184, "step": 69380 }, { "grad_norm": 0.3793236315250397, "learning_rate": 4.733159410826321e-06, "loss": 0.0201, "step": 69390 }, { "grad_norm": 0.28777459263801575, "learning_rate": 4.724385538210357e-06, "loss": 0.0191, "step": 69400 }, { "grad_norm": 0.22746558487415314, "learning_rate": 4.715619401958599e-06, "loss": 0.0168, "step": 69410 }, { "grad_norm": 0.31203189492225647, "learning_rate": 4.706861003568958e-06, "loss": 0.0233, "step": 69420 }, { "grad_norm": 0.18005569279193878, "learning_rate": 4.698110344538003e-06, "loss": 0.0215, "step": 69430 }, { "grad_norm": 0.4058433473110199, "learning_rate": 4.689367426360975e-06, "loss": 0.0153, "step": 69440 }, { "grad_norm": 0.2599612772464752, "learning_rate": 4.680632250531819e-06, "loss": 0.0197, "step": 69450 }, { "grad_norm": 0.2690353989601135, "learning_rate": 4.671904818543115e-06, "loss": 0.0146, "step": 69460 }, { "grad_norm": 0.1738823503255844, "learning_rate": 4.66318513188615e-06, "loss": 0.0131, "step": 69470 }, { "grad_norm": 0.3700665831565857, "learning_rate": 4.654473192050884e-06, "loss": 0.0184, "step": 69480 }, { "grad_norm": 0.2781636118888855, "learning_rate": 4.64576900052594e-06, "loss": 0.0188, "step": 69490 }, { "grad_norm": 0.21502165496349335, "learning_rate": 4.637072558798638e-06, "loss": 0.0144, "step": 69500 }, { "grad_norm": 0.2580063045024872, "learning_rate": 4.628383868354969e-06, "loss": 0.0173, "step": 69510 }, { "grad_norm": 0.4121415615081787, "learning_rate": 4.6197029306795595e-06, "loss": 0.0185, "step": 69520 }, { "grad_norm": 0.24204427003860474, "learning_rate": 4.611029747255779e-06, "loss": 0.0158, "step": 69530 }, { "grad_norm": 0.2607172131538391, "learning_rate": 4.6023643195656164e-06, "loss": 0.0191, "step": 69540 }, { "grad_norm": 0.22257792949676514, "learning_rate": 4.593706649089768e-06, "loss": 0.0144, "step": 69550 }, { "grad_norm": 0.2452651858329773, "learning_rate": 4.585056737307597e-06, "loss": 0.0167, "step": 69560 }, { "grad_norm": 0.2683905363082886, "learning_rate": 4.576414585697103e-06, "loss": 0.0125, "step": 69570 }, { "grad_norm": 0.34547892212867737, "learning_rate": 4.567780195735044e-06, "loss": 0.0198, "step": 69580 }, { "grad_norm": 0.2084176391363144, "learning_rate": 4.559153568896757e-06, "loss": 0.0126, "step": 69590 }, { "grad_norm": 0.5462530851364136, "learning_rate": 4.550534706656329e-06, "loss": 0.0155, "step": 69600 }, { "grad_norm": 0.18328437209129333, "learning_rate": 4.541923610486465e-06, "loss": 0.017, "step": 69610 }, { "grad_norm": 0.49798262119293213, "learning_rate": 4.533320281858578e-06, "loss": 0.0154, "step": 69620 }, { "grad_norm": 0.1757507175207138, "learning_rate": 4.52472472224274e-06, "loss": 0.0181, "step": 69630 }, { "grad_norm": 0.32015204429626465, "learning_rate": 4.51613693310769e-06, "loss": 0.0216, "step": 69640 }, { "grad_norm": 0.19288969039916992, "learning_rate": 4.507556915920868e-06, "loss": 0.0199, "step": 69650 }, { "grad_norm": 0.3476569950580597, "learning_rate": 4.498984672148332e-06, "loss": 0.0166, "step": 69660 }, { "grad_norm": 0.606049120426178, "learning_rate": 4.490420203254864e-06, "loss": 0.0216, "step": 69670 }, { "grad_norm": 0.38665562868118286, "learning_rate": 4.481863510703893e-06, "loss": 0.0188, "step": 69680 }, { "grad_norm": 0.09253662824630737, "learning_rate": 4.473314595957523e-06, "loss": 0.0191, "step": 69690 }, { "grad_norm": 0.31554022431373596, "learning_rate": 4.464773460476535e-06, "loss": 0.0212, "step": 69700 }, { "grad_norm": 0.26839032769203186, "learning_rate": 4.456240105720372e-06, "loss": 0.0143, "step": 69710 }, { "grad_norm": 0.1944490671157837, "learning_rate": 4.4477145331471405e-06, "loss": 0.0298, "step": 69720 }, { "grad_norm": 0.3469655215740204, "learning_rate": 4.439196744213653e-06, "loss": 0.0174, "step": 69730 }, { "grad_norm": 0.3114969730377197, "learning_rate": 4.430686740375339e-06, "loss": 0.018, "step": 69740 }, { "grad_norm": 0.2732439637184143, "learning_rate": 4.422184523086342e-06, "loss": 0.0187, "step": 69750 }, { "grad_norm": 0.18831747770309448, "learning_rate": 4.41369009379946e-06, "loss": 0.0164, "step": 69760 }, { "grad_norm": 0.21926943957805634, "learning_rate": 4.405203453966139e-06, "loss": 0.0136, "step": 69770 }, { "grad_norm": 0.35912007093429565, "learning_rate": 4.396724605036539e-06, "loss": 0.0237, "step": 69780 }, { "grad_norm": 0.2527310848236084, "learning_rate": 4.388253548459437e-06, "loss": 0.0191, "step": 69790 }, { "grad_norm": 0.3224947452545166, "learning_rate": 4.3797902856823395e-06, "loss": 0.0161, "step": 69800 }, { "grad_norm": 0.2890653610229492, "learning_rate": 4.371334818151357e-06, "loss": 0.0146, "step": 69810 }, { "grad_norm": 0.27546852827072144, "learning_rate": 4.362887147311306e-06, "loss": 0.0175, "step": 69820 }, { "grad_norm": 0.1415005624294281, "learning_rate": 4.354447274605672e-06, "loss": 0.0158, "step": 69830 }, { "grad_norm": 0.3794343173503876, "learning_rate": 4.346015201476572e-06, "loss": 0.0232, "step": 69840 }, { "grad_norm": 0.2393304705619812, "learning_rate": 4.337590929364855e-06, "loss": 0.0147, "step": 69850 }, { "grad_norm": 0.2801820635795593, "learning_rate": 4.329174459709973e-06, "loss": 0.0151, "step": 69860 }, { "grad_norm": 0.2177029550075531, "learning_rate": 4.320765793950071e-06, "loss": 0.0182, "step": 69870 }, { "grad_norm": 0.24351920187473297, "learning_rate": 4.312364933521962e-06, "loss": 0.0216, "step": 69880 }, { "grad_norm": 0.21796418726444244, "learning_rate": 4.303971879861129e-06, "loss": 0.0136, "step": 69890 }, { "grad_norm": 0.3177243769168854, "learning_rate": 4.295586634401716e-06, "loss": 0.02, "step": 69900 }, { "grad_norm": 0.28527018427848816, "learning_rate": 4.287209198576536e-06, "loss": 0.017, "step": 69910 }, { "grad_norm": 0.20185717940330505, "learning_rate": 4.278839573817045e-06, "loss": 0.0188, "step": 69920 }, { "grad_norm": 0.4075617790222168, "learning_rate": 4.270477761553399e-06, "loss": 0.0209, "step": 69930 }, { "grad_norm": 0.2063494622707367, "learning_rate": 4.262123763214393e-06, "loss": 0.0147, "step": 69940 }, { "grad_norm": 0.30918607115745544, "learning_rate": 4.253777580227508e-06, "loss": 0.0225, "step": 69950 }, { "grad_norm": 0.2519476115703583, "learning_rate": 4.2454392140188755e-06, "loss": 0.0161, "step": 69960 }, { "grad_norm": 0.38837456703186035, "learning_rate": 4.2371086660132785e-06, "loss": 0.0179, "step": 69970 }, { "grad_norm": 0.20067401230335236, "learning_rate": 4.228785937634205e-06, "loss": 0.0144, "step": 69980 }, { "grad_norm": 0.5283110737800598, "learning_rate": 4.220471030303758e-06, "loss": 0.0178, "step": 69990 }, { "grad_norm": 0.23642519116401672, "learning_rate": 4.212163945442754e-06, "loss": 0.0179, "step": 70000 }, { "grad_norm": 0.25175657868385315, "learning_rate": 4.203864684470621e-06, "loss": 0.0158, "step": 70010 }, { "grad_norm": 0.27940335869789124, "learning_rate": 4.19557324880549e-06, "loss": 0.0198, "step": 70020 }, { "grad_norm": 0.34952741861343384, "learning_rate": 4.187289639864145e-06, "loss": 0.015, "step": 70030 }, { "grad_norm": 0.32914477586746216, "learning_rate": 4.1790138590619974e-06, "loss": 0.0145, "step": 70040 }, { "grad_norm": 0.49891406297683716, "learning_rate": 4.170745907813195e-06, "loss": 0.0218, "step": 70050 }, { "grad_norm": 0.14628635346889496, "learning_rate": 4.162485787530479e-06, "loss": 0.0182, "step": 70060 }, { "grad_norm": 0.26428791880607605, "learning_rate": 4.154233499625282e-06, "loss": 0.0167, "step": 70070 }, { "grad_norm": 0.325910747051239, "learning_rate": 4.145989045507692e-06, "loss": 0.0169, "step": 70080 }, { "grad_norm": 0.2716531455516815, "learning_rate": 4.1377524265864666e-06, "loss": 0.0212, "step": 70090 }, { "grad_norm": 0.19775506854057312, "learning_rate": 4.1295236442690175e-06, "loss": 0.0194, "step": 70100 }, { "grad_norm": 0.37723997235298157, "learning_rate": 4.121302699961421e-06, "loss": 0.0173, "step": 70110 }, { "grad_norm": 0.30713993310928345, "learning_rate": 4.113089595068403e-06, "loss": 0.021, "step": 70120 }, { "grad_norm": 0.3999093174934387, "learning_rate": 4.104884330993364e-06, "loss": 0.0183, "step": 70130 }, { "grad_norm": 0.25994741916656494, "learning_rate": 4.0966869091383585e-06, "loss": 0.0243, "step": 70140 }, { "grad_norm": 0.19339382648468018, "learning_rate": 4.088497330904101e-06, "loss": 0.0164, "step": 70150 }, { "grad_norm": 0.389165997505188, "learning_rate": 4.080315597689976e-06, "loss": 0.0224, "step": 70160 }, { "grad_norm": 0.2932758629322052, "learning_rate": 4.072141710893995e-06, "loss": 0.0274, "step": 70170 }, { "grad_norm": 0.2233782410621643, "learning_rate": 4.063975671912879e-06, "loss": 0.0265, "step": 70180 }, { "grad_norm": 0.17244938015937805, "learning_rate": 4.055817482141949e-06, "loss": 0.0146, "step": 70190 }, { "grad_norm": 0.30085447430610657, "learning_rate": 4.047667142975259e-06, "loss": 0.0161, "step": 70200 }, { "grad_norm": 1.4490885734558105, "learning_rate": 4.039524655805443e-06, "loss": 0.03, "step": 70210 }, { "grad_norm": 0.25423333048820496, "learning_rate": 4.03139002202384e-06, "loss": 0.0154, "step": 70220 }, { "grad_norm": 0.3458104431629181, "learning_rate": 4.023263243020447e-06, "loss": 0.0202, "step": 70230 }, { "grad_norm": 0.36049535870552063, "learning_rate": 4.015144320183884e-06, "loss": 0.0191, "step": 70240 }, { "grad_norm": 0.2753370702266693, "learning_rate": 4.007033254901482e-06, "loss": 0.0174, "step": 70250 }, { "grad_norm": 0.31010255217552185, "learning_rate": 3.9989300485591795e-06, "loss": 0.0224, "step": 70260 }, { "grad_norm": 0.16202959418296814, "learning_rate": 3.990834702541601e-06, "loss": 0.0217, "step": 70270 }, { "grad_norm": 0.23325416445732117, "learning_rate": 3.982747218232019e-06, "loss": 0.0152, "step": 70280 }, { "grad_norm": 0.2621232867240906, "learning_rate": 3.97466759701236e-06, "loss": 0.0224, "step": 70290 }, { "grad_norm": 0.5336433053016663, "learning_rate": 3.966595840263226e-06, "loss": 0.0195, "step": 70300 }, { "grad_norm": 0.2757708728313446, "learning_rate": 3.958531949363836e-06, "loss": 0.019, "step": 70310 }, { "grad_norm": 0.32553830742836, "learning_rate": 3.950475925692098e-06, "loss": 0.0215, "step": 70320 }, { "grad_norm": 0.39129748940467834, "learning_rate": 3.9424277706245685e-06, "loss": 0.018, "step": 70330 }, { "grad_norm": 0.2252846211194992, "learning_rate": 3.934387485536451e-06, "loss": 0.0147, "step": 70340 }, { "grad_norm": 0.8483794927597046, "learning_rate": 3.926355071801619e-06, "loss": 0.0146, "step": 70350 }, { "grad_norm": 0.21217618882656097, "learning_rate": 3.9183305307925965e-06, "loss": 0.019, "step": 70360 }, { "grad_norm": 0.2907619774341583, "learning_rate": 3.910313863880533e-06, "loss": 0.018, "step": 70370 }, { "grad_norm": 0.4674713611602783, "learning_rate": 3.902305072435292e-06, "loss": 0.0159, "step": 70380 }, { "grad_norm": 0.35784193873405457, "learning_rate": 3.894304157825329e-06, "loss": 0.0158, "step": 70390 }, { "grad_norm": 0.25898614525794983, "learning_rate": 3.886311121417791e-06, "loss": 0.0133, "step": 70400 }, { "grad_norm": 0.32385188341140747, "learning_rate": 3.878325964578472e-06, "loss": 0.0189, "step": 70410 }, { "grad_norm": 0.1941865086555481, "learning_rate": 3.870348688671815e-06, "loss": 0.0134, "step": 70420 }, { "grad_norm": 0.23979246616363525, "learning_rate": 3.862379295060931e-06, "loss": 0.0132, "step": 70430 }, { "grad_norm": 0.2181229144334793, "learning_rate": 3.854417785107534e-06, "loss": 0.0271, "step": 70440 }, { "grad_norm": 0.15357954800128937, "learning_rate": 3.8464641601720755e-06, "loss": 0.0144, "step": 70450 }, { "grad_norm": 0.23955711722373962, "learning_rate": 3.838518421613579e-06, "loss": 0.0176, "step": 70460 }, { "grad_norm": 0.27680036425590515, "learning_rate": 3.830580570789766e-06, "loss": 0.0181, "step": 70470 }, { "grad_norm": 0.4347785413265228, "learning_rate": 3.822650609057005e-06, "loss": 0.018, "step": 70480 }, { "grad_norm": 0.17677366733551025, "learning_rate": 3.814728537770285e-06, "loss": 0.0158, "step": 70490 }, { "grad_norm": 0.248679518699646, "learning_rate": 3.806814358283306e-06, "loss": 0.0189, "step": 70500 }, { "grad_norm": 0.35350456833839417, "learning_rate": 3.7989080719483596e-06, "loss": 0.014, "step": 70510 }, { "grad_norm": 0.2938254177570343, "learning_rate": 3.7910096801164143e-06, "loss": 0.0186, "step": 70520 }, { "grad_norm": 0.2180214673280716, "learning_rate": 3.7831191841371016e-06, "loss": 0.0181, "step": 70530 }, { "grad_norm": 0.36667197942733765, "learning_rate": 3.775236585358688e-06, "loss": 0.0179, "step": 70540 }, { "grad_norm": 1.0503839254379272, "learning_rate": 3.7673618851280843e-06, "loss": 0.0185, "step": 70550 }, { "grad_norm": 0.3852272033691406, "learning_rate": 3.759495084790887e-06, "loss": 0.0221, "step": 70560 }, { "grad_norm": 0.24002407491207123, "learning_rate": 3.751636185691282e-06, "loss": 0.0195, "step": 70570 }, { "grad_norm": 0.2669445276260376, "learning_rate": 3.7437851891721607e-06, "loss": 0.0199, "step": 70580 }, { "grad_norm": 0.3034772276878357, "learning_rate": 3.7359420965750404e-06, "loss": 0.017, "step": 70590 }, { "grad_norm": 0.46232160925865173, "learning_rate": 3.7281069092400922e-06, "loss": 0.0184, "step": 70600 }, { "grad_norm": 0.23040713369846344, "learning_rate": 3.7202796285061348e-06, "loss": 0.0197, "step": 70610 }, { "grad_norm": 0.3328237235546112, "learning_rate": 3.712460255710637e-06, "loss": 0.019, "step": 70620 }, { "grad_norm": 0.1900823712348938, "learning_rate": 3.704648792189719e-06, "loss": 0.0152, "step": 70630 }, { "grad_norm": 0.3453202247619629, "learning_rate": 3.696845239278124e-06, "loss": 0.0152, "step": 70640 }, { "grad_norm": 0.3400867283344269, "learning_rate": 3.689049598309302e-06, "loss": 0.0223, "step": 70650 }, { "grad_norm": 0.38546809554100037, "learning_rate": 3.681261870615288e-06, "loss": 0.0261, "step": 70660 }, { "grad_norm": 0.18476682901382446, "learning_rate": 3.6734820575268004e-06, "loss": 0.0214, "step": 70670 }, { "grad_norm": 0.523615300655365, "learning_rate": 3.665710160373209e-06, "loss": 0.0144, "step": 70680 }, { "grad_norm": 0.5172936916351318, "learning_rate": 3.65794618048248e-06, "loss": 0.0157, "step": 70690 }, { "grad_norm": 0.5923133492469788, "learning_rate": 3.6501901191813125e-06, "loss": 0.0164, "step": 70700 }, { "grad_norm": 0.1988176852464676, "learning_rate": 3.642441977794975e-06, "loss": 0.0186, "step": 70710 }, { "grad_norm": 0.2280898094177246, "learning_rate": 3.634701757647424e-06, "loss": 0.0191, "step": 70720 }, { "grad_norm": 0.33686354756355286, "learning_rate": 3.6269694600612468e-06, "loss": 0.0165, "step": 70730 }, { "grad_norm": 0.2284626066684723, "learning_rate": 3.619245086357681e-06, "loss": 0.0147, "step": 70740 }, { "grad_norm": 0.292639821767807, "learning_rate": 3.611528637856615e-06, "loss": 0.0183, "step": 70750 }, { "grad_norm": 0.4982788562774658, "learning_rate": 3.6038201158765884e-06, "loss": 0.0152, "step": 70760 }, { "grad_norm": 0.5388558506965637, "learning_rate": 3.5961195217347534e-06, "loss": 0.0271, "step": 70770 }, { "grad_norm": 0.2629197835922241, "learning_rate": 3.588426856746946e-06, "loss": 0.019, "step": 70780 }, { "grad_norm": 0.17539547383785248, "learning_rate": 3.5807421222276316e-06, "loss": 0.0302, "step": 70790 }, { "grad_norm": 0.23553705215454102, "learning_rate": 3.5730653194899156e-06, "loss": 0.0178, "step": 70800 }, { "grad_norm": 0.22468553483486176, "learning_rate": 3.565396449845554e-06, "loss": 0.0166, "step": 70810 }, { "grad_norm": 0.3694254159927368, "learning_rate": 3.557735514604954e-06, "loss": 0.0191, "step": 70820 }, { "grad_norm": 0.2971087098121643, "learning_rate": 3.5500825150771633e-06, "loss": 0.0154, "step": 70830 }, { "grad_norm": 0.23639346659183502, "learning_rate": 3.5424374525698466e-06, "loss": 0.0185, "step": 70840 }, { "grad_norm": 0.364690363407135, "learning_rate": 3.5348003283893704e-06, "loss": 0.0132, "step": 70850 }, { "grad_norm": 0.344058096408844, "learning_rate": 3.527171143840685e-06, "loss": 0.0163, "step": 70860 }, { "grad_norm": 0.3499659299850464, "learning_rate": 3.519549900227409e-06, "loss": 0.0241, "step": 70870 }, { "grad_norm": 0.2639172077178955, "learning_rate": 3.511936598851828e-06, "loss": 0.0185, "step": 70880 }, { "grad_norm": 0.23812724649906158, "learning_rate": 3.504331241014813e-06, "loss": 0.0158, "step": 70890 }, { "grad_norm": 0.3484855890274048, "learning_rate": 3.4967338280159414e-06, "loss": 0.0148, "step": 70900 }, { "grad_norm": 0.37203821539878845, "learning_rate": 3.4891443611533846e-06, "loss": 0.0274, "step": 70910 }, { "grad_norm": 0.2724023163318634, "learning_rate": 3.481562841723984e-06, "loss": 0.0225, "step": 70920 }, { "grad_norm": 0.7068766355514526, "learning_rate": 3.473989271023215e-06, "loss": 0.0193, "step": 70930 }, { "grad_norm": 0.3115542531013489, "learning_rate": 3.4664236503451854e-06, "loss": 0.0168, "step": 70940 }, { "grad_norm": 0.26830223202705383, "learning_rate": 3.4588659809826674e-06, "loss": 0.0152, "step": 70950 }, { "grad_norm": 0.37722107768058777, "learning_rate": 3.451316264227039e-06, "loss": 0.0273, "step": 70960 }, { "grad_norm": 0.17151667177677155, "learning_rate": 3.4437745013683574e-06, "loss": 0.0171, "step": 70970 }, { "grad_norm": 0.3114445209503174, "learning_rate": 3.4362406936952916e-06, "loss": 0.0157, "step": 70980 }, { "grad_norm": 0.39574646949768066, "learning_rate": 3.4287148424951786e-06, "loss": 0.0186, "step": 70990 }, { "grad_norm": 0.29325586557388306, "learning_rate": 3.4211969490539617e-06, "loss": 0.0171, "step": 71000 }, { "grad_norm": 0.15879051387310028, "learning_rate": 3.4136870146562584e-06, "loss": 0.0171, "step": 71010 }, { "grad_norm": 0.30131494998931885, "learning_rate": 3.406185040585308e-06, "loss": 0.0206, "step": 71020 }, { "grad_norm": 0.3540199100971222, "learning_rate": 3.3986910281229966e-06, "loss": 0.0193, "step": 71030 }, { "grad_norm": 0.5747552514076233, "learning_rate": 3.391204978549828e-06, "loss": 0.017, "step": 71040 }, { "grad_norm": 0.531495213508606, "learning_rate": 3.3837268931449785e-06, "loss": 0.0236, "step": 71050 }, { "grad_norm": 0.27412179112434387, "learning_rate": 3.376256773186248e-06, "loss": 0.0242, "step": 71060 }, { "grad_norm": 0.29300713539123535, "learning_rate": 3.3687946199500664e-06, "loss": 0.0221, "step": 71070 }, { "grad_norm": 0.21012310683727264, "learning_rate": 3.3613404347115295e-06, "loss": 0.0217, "step": 71080 }, { "grad_norm": 0.22029289603233337, "learning_rate": 3.35389421874433e-06, "loss": 0.0141, "step": 71090 }, { "grad_norm": 0.34951189160346985, "learning_rate": 3.3464559733208446e-06, "loss": 0.0202, "step": 71100 }, { "grad_norm": 0.30308398604393005, "learning_rate": 3.3390256997120505e-06, "loss": 0.022, "step": 71110 }, { "grad_norm": 0.4203411936759949, "learning_rate": 3.331603399187583e-06, "loss": 0.0239, "step": 71120 }, { "grad_norm": 0.227501779794693, "learning_rate": 3.324189073015721e-06, "loss": 0.0273, "step": 71130 }, { "grad_norm": 0.16008026897907257, "learning_rate": 3.3167827224633465e-06, "loss": 0.0189, "step": 71140 }, { "grad_norm": 0.3083643317222595, "learning_rate": 3.309384348796024e-06, "loss": 0.0156, "step": 71150 }, { "grad_norm": 0.30340656638145447, "learning_rate": 3.301993953277921e-06, "loss": 0.017, "step": 71160 }, { "grad_norm": 0.29379916191101074, "learning_rate": 3.29461153717186e-06, "loss": 0.0158, "step": 71170 }, { "grad_norm": 0.2851024866104126, "learning_rate": 3.287237101739293e-06, "loss": 0.0283, "step": 71180 }, { "grad_norm": 0.361986368894577, "learning_rate": 3.2798706482403075e-06, "loss": 0.0264, "step": 71190 }, { "grad_norm": 0.20852914452552795, "learning_rate": 3.2725121779336285e-06, "loss": 0.0173, "step": 71200 }, { "grad_norm": 0.5344870686531067, "learning_rate": 3.265161692076618e-06, "loss": 0.0206, "step": 71210 }, { "grad_norm": 0.2244049608707428, "learning_rate": 3.257819191925282e-06, "loss": 0.0153, "step": 71220 }, { "grad_norm": 0.33589470386505127, "learning_rate": 3.2504846787342392e-06, "loss": 0.0168, "step": 71230 }, { "grad_norm": 0.1785147339105606, "learning_rate": 3.243158153756759e-06, "loss": 0.0159, "step": 71240 }, { "grad_norm": 0.37166446447372437, "learning_rate": 3.2358396182447514e-06, "loss": 0.014, "step": 71250 }, { "grad_norm": 0.1845325082540512, "learning_rate": 3.2285290734487496e-06, "loss": 0.0166, "step": 71260 }, { "grad_norm": 0.2052312046289444, "learning_rate": 3.221226520617926e-06, "loss": 0.0125, "step": 71270 }, { "grad_norm": 0.25545790791511536, "learning_rate": 3.213931961000094e-06, "loss": 0.0195, "step": 71280 }, { "grad_norm": 0.3237418830394745, "learning_rate": 3.2066453958416733e-06, "loss": 0.0165, "step": 71290 }, { "grad_norm": 0.29909422993659973, "learning_rate": 3.199366826387773e-06, "loss": 0.0172, "step": 71300 }, { "grad_norm": 0.2513222396373749, "learning_rate": 3.192096253882071e-06, "loss": 0.0185, "step": 71310 }, { "grad_norm": 0.22857490181922913, "learning_rate": 3.1848336795669177e-06, "loss": 0.0145, "step": 71320 }, { "grad_norm": 0.3083916902542114, "learning_rate": 3.1775791046833035e-06, "loss": 0.022, "step": 71330 }, { "grad_norm": 0.10573253780603409, "learning_rate": 3.170332530470804e-06, "loss": 0.0163, "step": 71340 }, { "grad_norm": 0.26536932587623596, "learning_rate": 3.1630939581677012e-06, "loss": 0.0214, "step": 71350 }, { "grad_norm": 0.27955201268196106, "learning_rate": 3.155863389010838e-06, "loss": 0.0188, "step": 71360 }, { "grad_norm": 0.16605053842067719, "learning_rate": 3.1486408242357323e-06, "loss": 0.0173, "step": 71370 }, { "grad_norm": 0.40470752120018005, "learning_rate": 3.1414262650765248e-06, "loss": 0.0345, "step": 71380 }, { "grad_norm": 0.3439341187477112, "learning_rate": 3.134219712765979e-06, "loss": 0.0263, "step": 71390 }, { "grad_norm": 0.2197609543800354, "learning_rate": 3.1270211685355044e-06, "loss": 0.0215, "step": 71400 }, { "grad_norm": 0.33093804121017456, "learning_rate": 3.1198306336151338e-06, "loss": 0.0232, "step": 71410 }, { "grad_norm": 0.18339873850345612, "learning_rate": 3.11264810923354e-06, "loss": 0.0179, "step": 71420 }, { "grad_norm": 0.290660560131073, "learning_rate": 3.105473596618008e-06, "loss": 0.0181, "step": 71430 }, { "grad_norm": 0.24029308557510376, "learning_rate": 3.0983070969944683e-06, "loss": 0.016, "step": 71440 }, { "grad_norm": 0.30505627393722534, "learning_rate": 3.09114861158748e-06, "loss": 0.0141, "step": 71450 }, { "grad_norm": 0.479500412940979, "learning_rate": 3.0839981416202314e-06, "loss": 0.0244, "step": 71460 }, { "grad_norm": 0.37531766295433044, "learning_rate": 3.0768556883145507e-06, "loss": 0.0155, "step": 71470 }, { "grad_norm": 0.2506403923034668, "learning_rate": 3.0697212528908834e-06, "loss": 0.0158, "step": 71480 }, { "grad_norm": 0.3239920139312744, "learning_rate": 3.0625948365683e-06, "loss": 0.0204, "step": 71490 }, { "grad_norm": 0.2511579692363739, "learning_rate": 3.055476440564531e-06, "loss": 0.0167, "step": 71500 }, { "grad_norm": 0.1903119534254074, "learning_rate": 3.0483660660958924e-06, "loss": 0.0174, "step": 71510 }, { "grad_norm": 0.41426411271095276, "learning_rate": 3.0412637143773624e-06, "loss": 0.0239, "step": 71520 }, { "grad_norm": 0.20178399980068207, "learning_rate": 3.034169386622554e-06, "loss": 0.0186, "step": 71530 }, { "grad_norm": 0.1971258521080017, "learning_rate": 3.027083084043658e-06, "loss": 0.0164, "step": 71540 }, { "grad_norm": 0.290121853351593, "learning_rate": 3.0200048078515676e-06, "loss": 0.02, "step": 71550 }, { "grad_norm": 0.24185939133167267, "learning_rate": 3.012934559255742e-06, "loss": 0.0145, "step": 71560 }, { "grad_norm": 0.3447025418281555, "learning_rate": 3.0058723394642994e-06, "loss": 0.0202, "step": 71570 }, { "grad_norm": 0.24567782878875732, "learning_rate": 2.998818149683985e-06, "loss": 0.0235, "step": 71580 }, { "grad_norm": 0.3597593605518341, "learning_rate": 2.9917719911201627e-06, "loss": 0.0176, "step": 71590 }, { "grad_norm": 0.20089732110500336, "learning_rate": 2.9847338649768352e-06, "loss": 0.0138, "step": 71600 }, { "grad_norm": 0.494780957698822, "learning_rate": 2.977703772456608e-06, "loss": 0.0145, "step": 71610 }, { "grad_norm": 0.2812142074108124, "learning_rate": 2.9706817147607535e-06, "loss": 0.0159, "step": 71620 }, { "grad_norm": 0.16183353960514069, "learning_rate": 2.9636676930891393e-06, "loss": 0.0191, "step": 71630 }, { "grad_norm": 0.28579843044281006, "learning_rate": 2.9566617086402625e-06, "loss": 0.0175, "step": 71640 }, { "grad_norm": 0.3879407048225403, "learning_rate": 2.9496637626112655e-06, "loss": 0.0192, "step": 71650 }, { "grad_norm": 0.35577592253685, "learning_rate": 2.9426738561979027e-06, "loss": 0.0183, "step": 71660 }, { "grad_norm": 0.3150431215763092, "learning_rate": 2.9356919905945524e-06, "loss": 0.0318, "step": 71670 }, { "grad_norm": 0.2538457214832306, "learning_rate": 2.928718166994243e-06, "loss": 0.0196, "step": 71680 }, { "grad_norm": 0.22479033470153809, "learning_rate": 2.9217523865885833e-06, "loss": 0.0228, "step": 71690 }, { "grad_norm": 0.2807861566543579, "learning_rate": 2.9147946505678537e-06, "loss": 0.0197, "step": 71700 }, { "grad_norm": 0.3907826840877533, "learning_rate": 2.9078449601209313e-06, "loss": 0.0159, "step": 71710 }, { "grad_norm": 0.4238656461238861, "learning_rate": 2.900903316435333e-06, "loss": 0.0206, "step": 71720 }, { "grad_norm": 0.38937294483184814, "learning_rate": 2.8939697206971983e-06, "loss": 0.0212, "step": 71730 }, { "grad_norm": 0.3799721598625183, "learning_rate": 2.8870441740912746e-06, "loss": 0.0146, "step": 71740 }, { "grad_norm": 0.7079835534095764, "learning_rate": 2.8801266778009762e-06, "loss": 0.0221, "step": 71750 }, { "grad_norm": 0.23891255259513855, "learning_rate": 2.8732172330082798e-06, "loss": 0.0233, "step": 71760 }, { "grad_norm": 0.21573148667812347, "learning_rate": 2.8663158408938517e-06, "loss": 0.0159, "step": 71770 }, { "grad_norm": 0.2938653826713562, "learning_rate": 2.859422502636938e-06, "loss": 0.0131, "step": 71780 }, { "grad_norm": 0.21692928671836853, "learning_rate": 2.8525372194154076e-06, "loss": 0.0194, "step": 71790 }, { "grad_norm": 0.21968822181224823, "learning_rate": 2.8456599924057913e-06, "loss": 0.0146, "step": 71800 }, { "grad_norm": 0.2443179190158844, "learning_rate": 2.8387908227831995e-06, "loss": 0.0186, "step": 71810 }, { "grad_norm": 0.17058654129505157, "learning_rate": 2.831929711721404e-06, "loss": 0.0187, "step": 71820 }, { "grad_norm": 0.29636961221694946, "learning_rate": 2.825076660392767e-06, "loss": 0.0167, "step": 71830 }, { "grad_norm": 0.24839530885219574, "learning_rate": 2.8182316699682908e-06, "loss": 0.0186, "step": 71840 }, { "grad_norm": 0.2481747269630432, "learning_rate": 2.811394741617601e-06, "loss": 0.0202, "step": 71850 }, { "grad_norm": 0.4615330696105957, "learning_rate": 2.8045658765089356e-06, "loss": 0.0167, "step": 71860 }, { "grad_norm": 0.34951168298721313, "learning_rate": 2.7977450758091605e-06, "loss": 0.0267, "step": 71870 }, { "grad_norm": 0.21978320181369781, "learning_rate": 2.790932340683783e-06, "loss": 0.022, "step": 71880 }, { "grad_norm": 0.21651090681552887, "learning_rate": 2.7841276722968823e-06, "loss": 0.0159, "step": 71890 }, { "grad_norm": 0.4010801315307617, "learning_rate": 2.7773310718112067e-06, "loss": 0.0165, "step": 71900 }, { "grad_norm": 0.31357529759407043, "learning_rate": 2.7705425403881102e-06, "loss": 0.0219, "step": 71910 }, { "grad_norm": 0.16674496233463287, "learning_rate": 2.763762079187565e-06, "loss": 0.0206, "step": 71920 }, { "grad_norm": 0.3624163866043091, "learning_rate": 2.756989689368178e-06, "loss": 0.0167, "step": 71930 }, { "grad_norm": 0.21000176668167114, "learning_rate": 2.750225372087134e-06, "loss": 0.0144, "step": 71940 }, { "grad_norm": 0.30298206210136414, "learning_rate": 2.7434691285003033e-06, "loss": 0.0154, "step": 71950 }, { "grad_norm": 0.3217720687389374, "learning_rate": 2.7367209597621178e-06, "loss": 0.0176, "step": 71960 }, { "grad_norm": 0.3061532974243164, "learning_rate": 2.729980867025683e-06, "loss": 0.0199, "step": 71970 }, { "grad_norm": 0.2599084973335266, "learning_rate": 2.7232488514426724e-06, "loss": 0.0142, "step": 71980 }, { "grad_norm": 0.2381954938173294, "learning_rate": 2.7165249141634097e-06, "loss": 0.0155, "step": 71990 }, { "grad_norm": 0.31977662444114685, "learning_rate": 2.709809056336837e-06, "loss": 0.0258, "step": 72000 }, { "grad_norm": 0.21166247129440308, "learning_rate": 2.703101279110498e-06, "loss": 0.0172, "step": 72010 }, { "grad_norm": 0.32509511709213257, "learning_rate": 2.696401583630587e-06, "loss": 0.023, "step": 72020 }, { "grad_norm": 0.33925336599349976, "learning_rate": 2.689709971041887e-06, "loss": 0.0149, "step": 72030 }, { "grad_norm": 0.3272038400173187, "learning_rate": 2.6830264424878126e-06, "loss": 0.0187, "step": 72040 }, { "grad_norm": 0.3011888861656189, "learning_rate": 2.676350999110394e-06, "loss": 0.0209, "step": 72050 }, { "grad_norm": 0.23308934271335602, "learning_rate": 2.6696836420502856e-06, "loss": 0.0223, "step": 72060 }, { "grad_norm": 0.29455092549324036, "learning_rate": 2.663024372446765e-06, "loss": 0.0169, "step": 72070 }, { "grad_norm": 0.3362521231174469, "learning_rate": 2.6563731914377055e-06, "loss": 0.0191, "step": 72080 }, { "grad_norm": 0.7021117210388184, "learning_rate": 2.6497301001596087e-06, "loss": 0.0251, "step": 72090 }, { "grad_norm": 0.4036547839641571, "learning_rate": 2.643095099747611e-06, "loss": 0.0206, "step": 72100 }, { "grad_norm": 0.3117155134677887, "learning_rate": 2.636468191335445e-06, "loss": 0.0213, "step": 72110 }, { "grad_norm": 0.2670019567012787, "learning_rate": 2.629849376055471e-06, "loss": 0.0174, "step": 72120 }, { "grad_norm": 0.2755616307258606, "learning_rate": 2.6232386550386678e-06, "loss": 0.0182, "step": 72130 }, { "grad_norm": 0.21571364998817444, "learning_rate": 2.6166360294146097e-06, "loss": 0.0141, "step": 72140 }, { "grad_norm": 0.25074782967567444, "learning_rate": 2.6100415003115275e-06, "loss": 0.0211, "step": 72150 }, { "grad_norm": 0.6730160117149353, "learning_rate": 2.603455068856225e-06, "loss": 0.0241, "step": 72160 }, { "grad_norm": 0.18329133093357086, "learning_rate": 2.5968767361741584e-06, "loss": 0.017, "step": 72170 }, { "grad_norm": 0.19519945979118347, "learning_rate": 2.590306503389378e-06, "loss": 0.0181, "step": 72180 }, { "grad_norm": 0.22082075476646423, "learning_rate": 2.583744371624558e-06, "loss": 0.0234, "step": 72190 }, { "grad_norm": 0.26076775789260864, "learning_rate": 2.57719034200099e-06, "loss": 0.0124, "step": 72200 }, { "grad_norm": 0.3148651123046875, "learning_rate": 2.570644415638568e-06, "loss": 0.0177, "step": 72210 }, { "grad_norm": 0.19268742203712463, "learning_rate": 2.5641065936558295e-06, "loss": 0.0196, "step": 72220 }, { "grad_norm": 0.303486704826355, "learning_rate": 2.557576877169898e-06, "loss": 0.0249, "step": 72230 }, { "grad_norm": 0.3632378578186035, "learning_rate": 2.5510552672965205e-06, "loss": 0.0155, "step": 72240 }, { "grad_norm": 0.45825329422950745, "learning_rate": 2.544541765150077e-06, "loss": 0.016, "step": 72250 }, { "grad_norm": 0.2590601444244385, "learning_rate": 2.5380363718435165e-06, "loss": 0.0226, "step": 72260 }, { "grad_norm": 0.2542249262332916, "learning_rate": 2.5315390884884714e-06, "loss": 0.0175, "step": 72270 }, { "grad_norm": 0.3327382504940033, "learning_rate": 2.5250499161951214e-06, "loss": 0.0145, "step": 72280 }, { "grad_norm": 0.28205227851867676, "learning_rate": 2.518568856072301e-06, "loss": 0.0159, "step": 72290 }, { "grad_norm": 0.2532337009906769, "learning_rate": 2.512095909227441e-06, "loss": 0.0221, "step": 72300 }, { "grad_norm": 0.33208486437797546, "learning_rate": 2.50563107676659e-06, "loss": 0.0219, "step": 72310 }, { "grad_norm": 0.41357919573783875, "learning_rate": 2.4991743597944152e-06, "loss": 0.0147, "step": 72320 }, { "grad_norm": 0.23643334209918976, "learning_rate": 2.492725759414205e-06, "loss": 0.0182, "step": 72330 }, { "grad_norm": 0.5467807054519653, "learning_rate": 2.4862852767278234e-06, "loss": 0.0175, "step": 72340 }, { "grad_norm": 0.8293163180351257, "learning_rate": 2.479852912835784e-06, "loss": 0.014, "step": 72350 }, { "grad_norm": 0.30771610140800476, "learning_rate": 2.4734286688372075e-06, "loss": 0.0178, "step": 72360 }, { "grad_norm": 0.25874534249305725, "learning_rate": 2.467012545829811e-06, "loss": 0.0239, "step": 72370 }, { "grad_norm": 0.23755042254924774, "learning_rate": 2.460604544909945e-06, "loss": 0.014, "step": 72380 }, { "grad_norm": 0.2936660051345825, "learning_rate": 2.454204667172555e-06, "loss": 0.016, "step": 72390 }, { "grad_norm": 0.24975433945655823, "learning_rate": 2.447812913711217e-06, "loss": 0.0176, "step": 72400 }, { "grad_norm": 0.2598729729652405, "learning_rate": 2.44142928561808e-06, "loss": 0.0243, "step": 72410 }, { "grad_norm": 0.33179280161857605, "learning_rate": 2.435053783983965e-06, "loss": 0.0139, "step": 72420 }, { "grad_norm": 0.4675399959087372, "learning_rate": 2.4286864098982453e-06, "loss": 0.0196, "step": 72430 }, { "grad_norm": 0.17841018736362457, "learning_rate": 2.4223271644489397e-06, "loss": 0.0179, "step": 72440 }, { "grad_norm": 0.1953360140323639, "learning_rate": 2.415976048722679e-06, "loss": 0.0171, "step": 72450 }, { "grad_norm": 0.2570886015892029, "learning_rate": 2.4096330638046673e-06, "loss": 0.0167, "step": 72460 }, { "grad_norm": 0.35739055275917053, "learning_rate": 2.4032982107787816e-06, "loss": 0.0178, "step": 72470 }, { "grad_norm": 0.2321765422821045, "learning_rate": 2.396971490727451e-06, "loss": 0.0222, "step": 72480 }, { "grad_norm": 0.2968592047691345, "learning_rate": 2.3906529047317493e-06, "loss": 0.0173, "step": 72490 }, { "grad_norm": 0.22180598974227905, "learning_rate": 2.3843424538713465e-06, "loss": 0.0165, "step": 72500 }, { "grad_norm": 0.37608128786087036, "learning_rate": 2.3780401392245298e-06, "loss": 0.0207, "step": 72510 }, { "grad_norm": 0.17550037801265717, "learning_rate": 2.3717459618681883e-06, "loss": 0.015, "step": 72520 }, { "grad_norm": 0.20586949586868286, "learning_rate": 2.3654599228778274e-06, "loss": 0.0135, "step": 72530 }, { "grad_norm": 0.22431276738643646, "learning_rate": 2.3591820233275607e-06, "loss": 0.0186, "step": 72540 }, { "grad_norm": 0.3413768708705902, "learning_rate": 2.3529122642901024e-06, "loss": 0.0179, "step": 72550 }, { "grad_norm": 0.2892898917198181, "learning_rate": 2.3466506468367845e-06, "loss": 0.0282, "step": 72560 }, { "grad_norm": 0.23132996261119843, "learning_rate": 2.3403971720375506e-06, "loss": 0.0194, "step": 72570 }, { "grad_norm": 0.30771970748901367, "learning_rate": 2.334151840960952e-06, "loss": 0.0159, "step": 72580 }, { "grad_norm": 0.28972944617271423, "learning_rate": 2.3279146546741347e-06, "loss": 0.0143, "step": 72590 }, { "grad_norm": 0.19008170068264008, "learning_rate": 2.32168561424288e-06, "loss": 0.0146, "step": 72600 }, { "grad_norm": 0.1582489162683487, "learning_rate": 2.3154647207315307e-06, "loss": 0.0214, "step": 72610 }, { "grad_norm": 0.21831674873828888, "learning_rate": 2.309251975203103e-06, "loss": 0.0173, "step": 72620 }, { "grad_norm": 0.49589625000953674, "learning_rate": 2.303047378719159e-06, "loss": 0.0194, "step": 72630 }, { "grad_norm": 0.2763683795928955, "learning_rate": 2.2968509323399e-06, "loss": 0.0214, "step": 72640 }, { "grad_norm": 0.1694081872701645, "learning_rate": 2.290662637124147e-06, "loss": 0.0159, "step": 72650 }, { "grad_norm": 0.1715206354856491, "learning_rate": 2.2844824941292807e-06, "loss": 0.0134, "step": 72660 }, { "grad_norm": 0.29701370000839233, "learning_rate": 2.2783105044113406e-06, "loss": 0.0208, "step": 72670 }, { "grad_norm": 0.26348942518234253, "learning_rate": 2.272146669024944e-06, "loss": 0.0155, "step": 72680 }, { "grad_norm": 0.2098609209060669, "learning_rate": 2.26599098902332e-06, "loss": 0.015, "step": 72690 }, { "grad_norm": 0.27174460887908936, "learning_rate": 2.2598434654583113e-06, "loss": 0.0151, "step": 72700 }, { "grad_norm": 0.41223618388175964, "learning_rate": 2.253704099380355e-06, "loss": 0.0258, "step": 72710 }, { "grad_norm": 0.16821227967739105, "learning_rate": 2.247572891838512e-06, "loss": 0.0235, "step": 72720 }, { "grad_norm": 0.46860796213150024, "learning_rate": 2.241449843880422e-06, "loss": 0.0195, "step": 72730 }, { "grad_norm": 0.17692777514457703, "learning_rate": 2.235334956552354e-06, "loss": 0.0189, "step": 72740 }, { "grad_norm": 0.18363964557647705, "learning_rate": 2.2292282308991775e-06, "loss": 0.0156, "step": 72750 }, { "grad_norm": 0.4204758107662201, "learning_rate": 2.223129667964363e-06, "loss": 0.0202, "step": 72760 }, { "grad_norm": 0.362976610660553, "learning_rate": 2.2170392687899834e-06, "loss": 0.0126, "step": 72770 }, { "grad_norm": 0.2494065761566162, "learning_rate": 2.210957034416733e-06, "loss": 0.0146, "step": 72780 }, { "grad_norm": 0.2327132374048233, "learning_rate": 2.2048829658838867e-06, "loss": 0.0168, "step": 72790 }, { "grad_norm": 0.5923736095428467, "learning_rate": 2.1988170642293525e-06, "loss": 0.0243, "step": 72800 }, { "grad_norm": 0.14923648536205292, "learning_rate": 2.1927593304896075e-06, "loss": 0.0138, "step": 72810 }, { "grad_norm": 0.12923015654087067, "learning_rate": 2.1867097656997626e-06, "loss": 0.012, "step": 72820 }, { "grad_norm": 0.3659295439720154, "learning_rate": 2.180668370893524e-06, "loss": 0.023, "step": 72830 }, { "grad_norm": 0.3752562999725342, "learning_rate": 2.174635147103199e-06, "loss": 0.0188, "step": 72840 }, { "grad_norm": 0.2852883040904999, "learning_rate": 2.1686100953597075e-06, "loss": 0.0177, "step": 72850 }, { "grad_norm": 0.248831644654274, "learning_rate": 2.1625932166925433e-06, "loss": 0.0168, "step": 72860 }, { "grad_norm": 0.26074397563934326, "learning_rate": 2.1565845121298556e-06, "loss": 0.0164, "step": 72870 }, { "grad_norm": 0.2567916810512543, "learning_rate": 2.150583982698351e-06, "loss": 0.0225, "step": 72880 }, { "grad_norm": 0.17347609996795654, "learning_rate": 2.144591629423359e-06, "loss": 0.0156, "step": 72890 }, { "grad_norm": 0.2426069974899292, "learning_rate": 2.138607453328817e-06, "loss": 0.0187, "step": 72900 }, { "grad_norm": 0.39720216393470764, "learning_rate": 2.132631455437234e-06, "loss": 0.0271, "step": 72910 }, { "grad_norm": 0.26967519521713257, "learning_rate": 2.1266636367697714e-06, "loss": 0.0244, "step": 72920 }, { "grad_norm": 0.559450089931488, "learning_rate": 2.120703998346152e-06, "loss": 0.0151, "step": 72930 }, { "grad_norm": 0.3205781579017639, "learning_rate": 2.1147525411847114e-06, "loss": 0.0165, "step": 72940 }, { "grad_norm": 0.187524676322937, "learning_rate": 2.108809266302403e-06, "loss": 0.016, "step": 72950 }, { "grad_norm": 0.32727667689323425, "learning_rate": 2.102874174714764e-06, "loss": 0.02, "step": 72960 }, { "grad_norm": 0.44559943675994873, "learning_rate": 2.0969472674359393e-06, "loss": 0.0161, "step": 72970 }, { "grad_norm": 0.20816859602928162, "learning_rate": 2.091028545478668e-06, "loss": 0.0197, "step": 72980 }, { "grad_norm": 0.18406732380390167, "learning_rate": 2.08511800985432e-06, "loss": 0.0137, "step": 72990 }, { "grad_norm": 0.5573335886001587, "learning_rate": 2.07921566157282e-06, "loss": 0.0224, "step": 73000 }, { "grad_norm": 0.24640998244285583, "learning_rate": 2.073321501642728e-06, "loss": 0.0185, "step": 73010 }, { "grad_norm": 0.2532716691493988, "learning_rate": 2.0674355310711937e-06, "loss": 0.0205, "step": 73020 }, { "grad_norm": 0.22772353887557983, "learning_rate": 2.0615577508639682e-06, "loss": 0.017, "step": 73030 }, { "grad_norm": 0.18522915244102478, "learning_rate": 2.055688162025399e-06, "loss": 0.0171, "step": 73040 }, { "grad_norm": 0.25870490074157715, "learning_rate": 2.0498267655584546e-06, "loss": 0.0245, "step": 73050 }, { "grad_norm": 0.2771621644496918, "learning_rate": 2.0439735624646626e-06, "loss": 0.0181, "step": 73060 }, { "grad_norm": 0.2660213112831116, "learning_rate": 2.0381285537442008e-06, "loss": 0.0175, "step": 73070 }, { "grad_norm": 0.20191779732704163, "learning_rate": 2.032291740395803e-06, "loss": 0.0178, "step": 73080 }, { "grad_norm": 0.21871328353881836, "learning_rate": 2.0264631234168276e-06, "loss": 0.0153, "step": 73090 }, { "grad_norm": 0.5664663910865784, "learning_rate": 2.0206427038032326e-06, "loss": 0.0212, "step": 73100 }, { "grad_norm": 0.28824201226234436, "learning_rate": 2.0148304825495457e-06, "loss": 0.0133, "step": 73110 }, { "grad_norm": 0.6165202260017395, "learning_rate": 2.0090264606489496e-06, "loss": 0.0184, "step": 73120 }, { "grad_norm": 0.2430839240550995, "learning_rate": 2.003230639093162e-06, "loss": 0.0211, "step": 73130 }, { "grad_norm": 0.7815138697624207, "learning_rate": 1.9974430188725525e-06, "loss": 0.0191, "step": 73140 }, { "grad_norm": 0.6643864512443542, "learning_rate": 1.9916636009760513e-06, "loss": 0.0228, "step": 73150 }, { "grad_norm": 0.29894641041755676, "learning_rate": 1.985892386391208e-06, "loss": 0.0209, "step": 73160 }, { "grad_norm": 0.37955591082572937, "learning_rate": 1.980129376104173e-06, "loss": 0.0213, "step": 73170 }, { "grad_norm": 0.3506140112876892, "learning_rate": 1.9743745710996796e-06, "loss": 0.0198, "step": 73180 }, { "grad_norm": 0.26650094985961914, "learning_rate": 1.9686279723610757e-06, "loss": 0.0144, "step": 73190 }, { "grad_norm": 0.24918289482593536, "learning_rate": 1.962889580870281e-06, "loss": 0.0188, "step": 73200 }, { "grad_norm": 0.32304370403289795, "learning_rate": 1.9571593976078385e-06, "loss": 0.0133, "step": 73210 }, { "grad_norm": 0.68206387758255, "learning_rate": 1.9514374235528876e-06, "loss": 0.0181, "step": 73220 }, { "grad_norm": 0.4560382068157196, "learning_rate": 1.9457236596831408e-06, "loss": 0.0141, "step": 73230 }, { "grad_norm": 0.21346977353096008, "learning_rate": 1.940018106974939e-06, "loss": 0.0169, "step": 73240 }, { "grad_norm": 0.34824255108833313, "learning_rate": 1.9343207664032025e-06, "loss": 0.0211, "step": 73250 }, { "grad_norm": 0.22797204554080963, "learning_rate": 1.928631638941436e-06, "loss": 0.0142, "step": 73260 }, { "grad_norm": 0.3717302680015564, "learning_rate": 1.922950725561784e-06, "loss": 0.0175, "step": 73270 }, { "grad_norm": 0.5225013494491577, "learning_rate": 1.9172780272349312e-06, "loss": 0.0163, "step": 73280 }, { "grad_norm": 0.3128058612346649, "learning_rate": 1.911613544930202e-06, "loss": 0.017, "step": 73290 }, { "grad_norm": 0.213596910238266, "learning_rate": 1.9059572796155e-06, "loss": 0.0168, "step": 73300 }, { "grad_norm": 0.32242175936698914, "learning_rate": 1.9003092322573192e-06, "loss": 0.0135, "step": 73310 }, { "grad_norm": 0.2263810634613037, "learning_rate": 1.8946694038207647e-06, "loss": 0.0215, "step": 73320 }, { "grad_norm": 0.29251670837402344, "learning_rate": 1.8890377952695215e-06, "loss": 0.015, "step": 73330 }, { "grad_norm": 0.22123201191425323, "learning_rate": 1.8834144075658865e-06, "loss": 0.0199, "step": 73340 }, { "grad_norm": 0.3148459494113922, "learning_rate": 1.8777992416707357e-06, "loss": 0.0235, "step": 73350 }, { "grad_norm": 0.4028085172176361, "learning_rate": 1.8721922985435458e-06, "loss": 0.0165, "step": 73360 }, { "grad_norm": 0.18780431151390076, "learning_rate": 1.8665935791424006e-06, "loss": 0.0146, "step": 73370 }, { "grad_norm": 0.210432767868042, "learning_rate": 1.8610030844239512e-06, "loss": 0.0172, "step": 73380 }, { "grad_norm": 0.21095962822437286, "learning_rate": 1.8554208153434838e-06, "loss": 0.0127, "step": 73390 }, { "grad_norm": 0.33998796343803406, "learning_rate": 1.8498467728548296e-06, "loss": 0.0139, "step": 73400 }, { "grad_norm": 0.1846293807029724, "learning_rate": 1.8442809579104547e-06, "loss": 0.0161, "step": 73410 }, { "grad_norm": 0.3868832588195801, "learning_rate": 1.8387233714614039e-06, "loss": 0.0178, "step": 73420 }, { "grad_norm": 0.2969917356967926, "learning_rate": 1.8331740144573174e-06, "loss": 0.0211, "step": 73430 }, { "grad_norm": 0.26563578844070435, "learning_rate": 1.8276328878464255e-06, "loss": 0.0164, "step": 73440 }, { "grad_norm": 0.19879616796970367, "learning_rate": 1.82209999257556e-06, "loss": 0.0168, "step": 73450 }, { "grad_norm": 0.2511128783226013, "learning_rate": 1.8165753295901311e-06, "loss": 0.0155, "step": 73460 }, { "grad_norm": 0.15816764533519745, "learning_rate": 1.8110588998341616e-06, "loss": 0.0147, "step": 73470 }, { "grad_norm": 0.28578871488571167, "learning_rate": 1.805550704250253e-06, "loss": 0.0342, "step": 73480 }, { "grad_norm": 0.32192277908325195, "learning_rate": 1.8000507437796077e-06, "loss": 0.0132, "step": 73490 }, { "grad_norm": 0.2918054163455963, "learning_rate": 1.7945590193620242e-06, "loss": 0.0219, "step": 73500 }, { "grad_norm": 0.30604463815689087, "learning_rate": 1.7890755319358742e-06, "loss": 0.017, "step": 73510 }, { "grad_norm": 0.3014923632144928, "learning_rate": 1.7836002824381525e-06, "loss": 0.0222, "step": 73520 }, { "grad_norm": 0.21567530930042267, "learning_rate": 1.7781332718044165e-06, "loss": 0.0112, "step": 73530 }, { "grad_norm": 0.26864245533943176, "learning_rate": 1.772674500968835e-06, "loss": 0.0199, "step": 73540 }, { "grad_norm": 0.33579862117767334, "learning_rate": 1.7672239708641624e-06, "loss": 0.018, "step": 73550 }, { "grad_norm": 0.13357439637184143, "learning_rate": 1.7617816824217371e-06, "loss": 0.012, "step": 73560 }, { "grad_norm": 0.3370354473590851, "learning_rate": 1.7563476365715148e-06, "loss": 0.0235, "step": 73570 }, { "grad_norm": 0.25975656509399414, "learning_rate": 1.750921834242003e-06, "loss": 0.0205, "step": 73580 }, { "grad_norm": 0.27050307393074036, "learning_rate": 1.7455042763603436e-06, "loss": 0.018, "step": 73590 }, { "grad_norm": 0.22727762162685394, "learning_rate": 1.7400949638522345e-06, "loss": 0.0129, "step": 73600 }, { "grad_norm": 0.27448704838752747, "learning_rate": 1.734693897641987e-06, "loss": 0.0271, "step": 73610 }, { "grad_norm": 0.30261555314064026, "learning_rate": 1.7293010786524955e-06, "loss": 0.0132, "step": 73620 }, { "grad_norm": 0.29964327812194824, "learning_rate": 1.72391650780524e-06, "loss": 0.0201, "step": 73630 }, { "grad_norm": 0.3085564076900482, "learning_rate": 1.718540186020301e-06, "loss": 0.017, "step": 73640 }, { "grad_norm": 0.24590720236301422, "learning_rate": 1.7131721142163437e-06, "loss": 0.0158, "step": 73650 }, { "grad_norm": 0.3244333565235138, "learning_rate": 1.7078122933106233e-06, "loss": 0.022, "step": 73660 }, { "grad_norm": 0.22298483550548553, "learning_rate": 1.7024607242189905e-06, "loss": 0.0159, "step": 73670 }, { "grad_norm": 0.20756232738494873, "learning_rate": 1.6971174078558749e-06, "loss": 0.015, "step": 73680 }, { "grad_norm": 0.2726976275444031, "learning_rate": 1.6917823451343073e-06, "loss": 0.0225, "step": 73690 }, { "grad_norm": 0.3215000629425049, "learning_rate": 1.6864555369659142e-06, "loss": 0.0201, "step": 73700 }, { "grad_norm": 0.3707546889781952, "learning_rate": 1.681136984260878e-06, "loss": 0.0203, "step": 73710 }, { "grad_norm": 0.4582623243331909, "learning_rate": 1.6758266879280172e-06, "loss": 0.0191, "step": 73720 }, { "grad_norm": 0.3238498270511627, "learning_rate": 1.6705246488746996e-06, "loss": 0.0145, "step": 73730 }, { "grad_norm": 0.36891159415245056, "learning_rate": 1.6652308680069062e-06, "loss": 0.019, "step": 73740 }, { "grad_norm": 0.3806281089782715, "learning_rate": 1.6599453462292081e-06, "loss": 0.0196, "step": 73750 }, { "grad_norm": 0.4715626835823059, "learning_rate": 1.6546680844447326e-06, "loss": 0.0185, "step": 73760 }, { "grad_norm": 0.157191663980484, "learning_rate": 1.6493990835552475e-06, "loss": 0.016, "step": 73770 }, { "grad_norm": 0.26822951436042786, "learning_rate": 1.6441383444610547e-06, "loss": 0.0112, "step": 73780 }, { "grad_norm": 0.1844625174999237, "learning_rate": 1.6388858680610908e-06, "loss": 0.0196, "step": 73790 }, { "grad_norm": 0.35236650705337524, "learning_rate": 1.6336416552528543e-06, "loss": 0.0148, "step": 73800 }, { "grad_norm": 0.17457181215286255, "learning_rate": 1.6284057069324288e-06, "loss": 0.022, "step": 73810 }, { "grad_norm": 0.19533635675907135, "learning_rate": 1.623178023994504e-06, "loss": 0.0217, "step": 73820 }, { "grad_norm": 0.27473723888397217, "learning_rate": 1.6179586073323483e-06, "loss": 0.019, "step": 73830 }, { "grad_norm": 0.46922025084495544, "learning_rate": 1.612747457837821e-06, "loss": 0.0163, "step": 73840 }, { "grad_norm": 0.30994150042533875, "learning_rate": 1.607544576401354e-06, "loss": 0.0167, "step": 73850 }, { "grad_norm": 0.2619658410549164, "learning_rate": 1.6023499639119754e-06, "loss": 0.0184, "step": 73860 }, { "grad_norm": 0.3586304783821106, "learning_rate": 1.5971636212573138e-06, "loss": 0.0141, "step": 73870 }, { "grad_norm": 0.24133165180683136, "learning_rate": 1.5919855493235714e-06, "loss": 0.0174, "step": 73880 }, { "grad_norm": 0.1926807016134262, "learning_rate": 1.586815748995535e-06, "loss": 0.0137, "step": 73890 }, { "grad_norm": 0.26452288031578064, "learning_rate": 1.5816542211565866e-06, "loss": 0.0151, "step": 73900 }, { "grad_norm": 0.321848064661026, "learning_rate": 1.5765009666886765e-06, "loss": 0.0162, "step": 73910 }, { "grad_norm": 0.2683040499687195, "learning_rate": 1.5713559864723781e-06, "loss": 0.0264, "step": 73920 }, { "grad_norm": 0.27345654368400574, "learning_rate": 1.5662192813868104e-06, "loss": 0.0186, "step": 73930 }, { "grad_norm": 0.407610148191452, "learning_rate": 1.561090852309699e-06, "loss": 0.0236, "step": 73940 }, { "grad_norm": 0.20657634735107422, "learning_rate": 1.5559707001173651e-06, "loss": 0.0173, "step": 73950 }, { "grad_norm": 0.5511232614517212, "learning_rate": 1.5508588256846757e-06, "loss": 0.0177, "step": 73960 }, { "grad_norm": 0.4947553873062134, "learning_rate": 1.5457552298851319e-06, "loss": 0.0165, "step": 73970 }, { "grad_norm": 0.387644499540329, "learning_rate": 1.5406599135907918e-06, "loss": 0.0198, "step": 73980 }, { "grad_norm": 0.19120699167251587, "learning_rate": 1.5355728776723088e-06, "loss": 0.0184, "step": 73990 }, { "grad_norm": 0.18235354125499725, "learning_rate": 1.5304941229989155e-06, "loss": 0.0136, "step": 74000 }, { "grad_norm": 0.4435024559497833, "learning_rate": 1.5254236504384345e-06, "loss": 0.0151, "step": 74010 }, { "grad_norm": 0.33442115783691406, "learning_rate": 1.5203614608572726e-06, "loss": 0.0173, "step": 74020 }, { "grad_norm": 0.16946490108966827, "learning_rate": 1.5153075551204044e-06, "loss": 0.0134, "step": 74030 }, { "grad_norm": 0.3311481773853302, "learning_rate": 1.5102619340914225e-06, "loss": 0.0311, "step": 74040 }, { "grad_norm": 0.2542192041873932, "learning_rate": 1.505224598632482e-06, "loss": 0.0134, "step": 74050 }, { "grad_norm": 0.28130868077278137, "learning_rate": 1.5001955496043164e-06, "loss": 0.0298, "step": 74060 }, { "grad_norm": 0.3868196904659271, "learning_rate": 1.4951747878662604e-06, "loss": 0.0138, "step": 74070 }, { "grad_norm": 0.47797176241874695, "learning_rate": 1.4901623142762221e-06, "loss": 0.017, "step": 74080 }, { "grad_norm": 0.252322256565094, "learning_rate": 1.4851581296907001e-06, "loss": 0.018, "step": 74090 }, { "grad_norm": 0.26098552346229553, "learning_rate": 1.4801622349647714e-06, "loss": 0.0118, "step": 74100 }, { "grad_norm": 0.2503635883331299, "learning_rate": 1.4751746309520976e-06, "loss": 0.0178, "step": 74110 }, { "grad_norm": 0.2191632091999054, "learning_rate": 1.4701953185049132e-06, "loss": 0.0172, "step": 74120 }, { "grad_norm": 0.3078483045101166, "learning_rate": 1.4652242984740661e-06, "loss": 0.0199, "step": 74130 }, { "grad_norm": 0.25000736117362976, "learning_rate": 1.4602615717089484e-06, "loss": 0.028, "step": 74140 }, { "grad_norm": 0.28134316205978394, "learning_rate": 1.4553071390575657e-06, "loss": 0.0218, "step": 74150 }, { "grad_norm": 0.1871599406003952, "learning_rate": 1.450361001366496e-06, "loss": 0.0222, "step": 74160 }, { "grad_norm": 0.2510392367839813, "learning_rate": 1.4454231594809021e-06, "loss": 0.0191, "step": 74170 }, { "grad_norm": 0.16684484481811523, "learning_rate": 1.4404936142445036e-06, "loss": 0.0201, "step": 74180 }, { "grad_norm": 0.2003968358039856, "learning_rate": 1.4355723664996546e-06, "loss": 0.0145, "step": 74190 }, { "grad_norm": 0.32234007120132446, "learning_rate": 1.4306594170872433e-06, "loss": 0.0146, "step": 74200 }, { "grad_norm": 0.22748789191246033, "learning_rate": 1.4257547668467598e-06, "loss": 0.0198, "step": 74210 }, { "grad_norm": 0.3359679579734802, "learning_rate": 1.4208584166162886e-06, "loss": 0.0228, "step": 74220 }, { "grad_norm": 0.22991140186786652, "learning_rate": 1.4159703672324553e-06, "loss": 0.0157, "step": 74230 }, { "grad_norm": 0.2341320961713791, "learning_rate": 1.4110906195305252e-06, "loss": 0.0198, "step": 74240 }, { "grad_norm": 0.23348623514175415, "learning_rate": 1.406219174344292e-06, "loss": 0.017, "step": 74250 }, { "grad_norm": 0.2654930055141449, "learning_rate": 1.401356032506157e-06, "loss": 0.0135, "step": 74260 }, { "grad_norm": 0.5045037865638733, "learning_rate": 1.3965011948471051e-06, "loss": 0.0225, "step": 74270 }, { "grad_norm": 0.18763570487499237, "learning_rate": 1.3916546621966842e-06, "loss": 0.0183, "step": 74280 }, { "grad_norm": 0.25224587321281433, "learning_rate": 1.3868164353830482e-06, "loss": 0.0125, "step": 74290 }, { "grad_norm": 0.24996650218963623, "learning_rate": 1.3819865152329082e-06, "loss": 0.013, "step": 74300 }, { "grad_norm": 0.20930784940719604, "learning_rate": 1.377164902571565e-06, "loss": 0.0228, "step": 74310 }, { "grad_norm": 0.3128790557384491, "learning_rate": 1.372351598222904e-06, "loss": 0.0197, "step": 74320 }, { "grad_norm": 1.2874425649642944, "learning_rate": 1.3675466030093843e-06, "loss": 0.0202, "step": 74330 }, { "grad_norm": 0.30254268646240234, "learning_rate": 1.3627499177520486e-06, "loss": 0.0205, "step": 74340 }, { "grad_norm": 0.45113176107406616, "learning_rate": 1.3579615432705196e-06, "loss": 0.0183, "step": 74350 }, { "grad_norm": 0.48579326272010803, "learning_rate": 1.3531814803829978e-06, "loss": 0.0189, "step": 74360 }, { "grad_norm": 0.30397874116897583, "learning_rate": 1.3484097299062803e-06, "loss": 0.0185, "step": 74370 }, { "grad_norm": 0.2600756287574768, "learning_rate": 1.343646292655698e-06, "loss": 0.0151, "step": 74380 }, { "grad_norm": 0.23676709830760956, "learning_rate": 1.3388911694452277e-06, "loss": 0.0199, "step": 74390 }, { "grad_norm": 0.36489614844322205, "learning_rate": 1.3341443610873638e-06, "loss": 0.0206, "step": 74400 }, { "grad_norm": 0.5254427194595337, "learning_rate": 1.329405868393213e-06, "loss": 0.0184, "step": 74410 }, { "grad_norm": 0.27158206701278687, "learning_rate": 1.3246756921724613e-06, "loss": 0.0204, "step": 74420 }, { "grad_norm": 0.23819278180599213, "learning_rate": 1.3199538332333506e-06, "loss": 0.0146, "step": 74430 }, { "grad_norm": 0.20083773136138916, "learning_rate": 1.3152402923827411e-06, "loss": 0.018, "step": 74440 }, { "grad_norm": 0.35767045617103577, "learning_rate": 1.3105350704260277e-06, "loss": 0.0193, "step": 74450 }, { "grad_norm": 0.515473484992981, "learning_rate": 1.305838168167206e-06, "loss": 0.015, "step": 74460 }, { "grad_norm": 0.35078901052474976, "learning_rate": 1.301149586408862e-06, "loss": 0.0148, "step": 74470 }, { "grad_norm": 0.44887399673461914, "learning_rate": 1.2964693259521322e-06, "loss": 0.0194, "step": 74480 }, { "grad_norm": 0.22823014855384827, "learning_rate": 1.2917973875967548e-06, "loss": 0.0133, "step": 74490 }, { "grad_norm": 0.5495733022689819, "learning_rate": 1.2871337721410249e-06, "loss": 0.0178, "step": 74500 }, { "grad_norm": 0.27678996324539185, "learning_rate": 1.2824784803818379e-06, "loss": 0.0183, "step": 74510 }, { "grad_norm": 0.28531280159950256, "learning_rate": 1.2778315131146467e-06, "loss": 0.0154, "step": 74520 }, { "grad_norm": 0.2893119156360626, "learning_rate": 1.2731928711334994e-06, "loss": 0.0167, "step": 74530 }, { "grad_norm": 0.23719972372055054, "learning_rate": 1.2685625552310122e-06, "loss": 0.0159, "step": 74540 }, { "grad_norm": 0.21789301931858063, "learning_rate": 1.263940566198374e-06, "loss": 0.0169, "step": 74550 }, { "grad_norm": 0.20655310153961182, "learning_rate": 1.259326904825353e-06, "loss": 0.0158, "step": 74560 }, { "grad_norm": 0.30008864402770996, "learning_rate": 1.2547215719003137e-06, "loss": 0.0182, "step": 74570 }, { "grad_norm": 0.1909877508878708, "learning_rate": 1.2501245682101703e-06, "loss": 0.0139, "step": 74580 }, { "grad_norm": 0.2220126986503601, "learning_rate": 1.2455358945404171e-06, "loss": 0.0164, "step": 74590 }, { "grad_norm": 0.2525823712348938, "learning_rate": 1.2409555516751493e-06, "loss": 0.0182, "step": 74600 }, { "grad_norm": 0.23236115276813507, "learning_rate": 1.2363835403970125e-06, "loss": 0.0176, "step": 74610 }, { "grad_norm": 0.2787877917289734, "learning_rate": 1.2318198614872489e-06, "loss": 0.0181, "step": 74620 }, { "grad_norm": 0.25303715467453003, "learning_rate": 1.2272645157256457e-06, "loss": 0.0163, "step": 74630 }, { "grad_norm": 0.3055722117424011, "learning_rate": 1.222717503890608e-06, "loss": 0.0161, "step": 74640 }, { "grad_norm": 0.35444745421409607, "learning_rate": 1.218178826759081e-06, "loss": 0.0222, "step": 74650 }, { "grad_norm": 0.2867055833339691, "learning_rate": 1.213648485106611e-06, "loss": 0.0184, "step": 74660 }, { "grad_norm": 0.4092257618904114, "learning_rate": 1.2091264797073066e-06, "loss": 0.0274, "step": 74670 }, { "grad_norm": 0.2685275375843048, "learning_rate": 1.2046128113338494e-06, "loss": 0.0296, "step": 74680 }, { "grad_norm": 0.25788968801498413, "learning_rate": 1.200107480757512e-06, "loss": 0.0166, "step": 74690 }, { "grad_norm": 0.27231961488723755, "learning_rate": 1.1956104887481168e-06, "loss": 0.0275, "step": 74700 }, { "grad_norm": 0.271919846534729, "learning_rate": 1.191121836074094e-06, "loss": 0.0168, "step": 74710 }, { "grad_norm": 0.3229617774486542, "learning_rate": 1.1866415235024186e-06, "loss": 0.0154, "step": 74720 }, { "grad_norm": 0.2957933843135834, "learning_rate": 1.182169551798662e-06, "loss": 0.0209, "step": 74730 }, { "grad_norm": 0.18812738358974457, "learning_rate": 1.177705921726957e-06, "loss": 0.0171, "step": 74740 }, { "grad_norm": 0.4230494499206543, "learning_rate": 1.173250634050016e-06, "loss": 0.0215, "step": 74750 }, { "grad_norm": 0.2245691567659378, "learning_rate": 1.1688036895291354e-06, "loss": 0.0186, "step": 74760 }, { "grad_norm": 0.2756723165512085, "learning_rate": 1.1643650889241574e-06, "loss": 0.0185, "step": 74770 }, { "grad_norm": 0.23487135767936707, "learning_rate": 1.1599348329935311e-06, "loss": 0.0159, "step": 74780 }, { "grad_norm": 0.3619285821914673, "learning_rate": 1.1555129224942673e-06, "loss": 0.0275, "step": 74790 }, { "grad_norm": 0.29219532012939453, "learning_rate": 1.1510993581819396e-06, "loss": 0.0169, "step": 74800 }, { "grad_norm": 0.2234981507062912, "learning_rate": 1.1466941408107112e-06, "loss": 0.0243, "step": 74810 }, { "grad_norm": 0.3995361030101776, "learning_rate": 1.1422972711333247e-06, "loss": 0.0182, "step": 74820 }, { "grad_norm": 0.3004937171936035, "learning_rate": 1.1379087499010565e-06, "loss": 0.0151, "step": 74830 }, { "grad_norm": 0.46455952525138855, "learning_rate": 1.1335285778638127e-06, "loss": 0.013, "step": 74840 }, { "grad_norm": 0.2567795515060425, "learning_rate": 1.1291567557700333e-06, "loss": 0.016, "step": 74850 }, { "grad_norm": 0.5532861948013306, "learning_rate": 1.124793284366743e-06, "loss": 0.0206, "step": 74860 }, { "grad_norm": 0.22479502856731415, "learning_rate": 1.1204381643995455e-06, "loss": 0.0184, "step": 74870 }, { "grad_norm": 0.4848722219467163, "learning_rate": 1.116091396612595e-06, "loss": 0.0185, "step": 74880 }, { "grad_norm": 0.28985732793807983, "learning_rate": 1.1117529817486538e-06, "loss": 0.0177, "step": 74890 }, { "grad_norm": 0.47069254517555237, "learning_rate": 1.107422920549034e-06, "loss": 0.015, "step": 74900 }, { "grad_norm": 0.3860110640525818, "learning_rate": 1.1031012137536157e-06, "loss": 0.0223, "step": 74910 }, { "grad_norm": 0.16540466248989105, "learning_rate": 1.0987878621008695e-06, "loss": 0.0155, "step": 74920 }, { "grad_norm": 0.3405489921569824, "learning_rate": 1.0944828663278284e-06, "loss": 0.0162, "step": 74930 }, { "grad_norm": 0.16415275633335114, "learning_rate": 1.0901862271700925e-06, "loss": 0.0176, "step": 74940 }, { "grad_norm": 0.237229123711586, "learning_rate": 1.0858979453618467e-06, "loss": 0.0151, "step": 74950 }, { "grad_norm": 0.3790428936481476, "learning_rate": 1.0816180216358441e-06, "loss": 0.018, "step": 74960 }, { "grad_norm": 0.32599255442619324, "learning_rate": 1.0773464567233937e-06, "loss": 0.0139, "step": 74970 }, { "grad_norm": 0.17815794050693512, "learning_rate": 1.0730832513543953e-06, "loss": 0.0173, "step": 74980 }, { "grad_norm": 0.26637259125709534, "learning_rate": 1.0688284062573218e-06, "loss": 0.0175, "step": 74990 }, { "grad_norm": 0.22014518082141876, "learning_rate": 1.0645819221591969e-06, "loss": 0.0188, "step": 75000 }, { "grad_norm": 0.2560734748840332, "learning_rate": 1.0603437997856347e-06, "loss": 0.0167, "step": 75010 }, { "grad_norm": 0.2999805510044098, "learning_rate": 1.0561140398608228e-06, "loss": 0.0145, "step": 75020 }, { "grad_norm": 0.2606717050075531, "learning_rate": 1.0518926431074937e-06, "loss": 0.014, "step": 75030 }, { "grad_norm": 0.24038763344287872, "learning_rate": 1.0476796102469877e-06, "loss": 0.0191, "step": 75040 }, { "grad_norm": 0.32663044333457947, "learning_rate": 1.0434749419991786e-06, "loss": 0.0176, "step": 75050 }, { "grad_norm": 0.36422550678253174, "learning_rate": 1.0392786390825415e-06, "loss": 0.0216, "step": 75060 }, { "grad_norm": 0.3115825951099396, "learning_rate": 1.0350907022141087e-06, "loss": 0.0226, "step": 75070 }, { "grad_norm": 0.24780133366584778, "learning_rate": 1.0309111321094744e-06, "loss": 0.0142, "step": 75080 }, { "grad_norm": 0.19502317905426025, "learning_rate": 1.026739929482834e-06, "loss": 0.0118, "step": 75090 }, { "grad_norm": 0.2951733469963074, "learning_rate": 1.022577095046906e-06, "loss": 0.0182, "step": 75100 }, { "grad_norm": 0.16438797116279602, "learning_rate": 1.0184226295130217e-06, "loss": 0.0166, "step": 75110 }, { "grad_norm": 0.35638758540153503, "learning_rate": 1.0142765335910575e-06, "loss": 0.0184, "step": 75120 }, { "grad_norm": 0.3627910315990448, "learning_rate": 1.0101388079894746e-06, "loss": 0.0168, "step": 75130 }, { "grad_norm": 0.23100122809410095, "learning_rate": 1.0060094534152908e-06, "loss": 0.0213, "step": 75140 }, { "grad_norm": 0.30234748125076294, "learning_rate": 1.0018884705741028e-06, "loss": 0.0136, "step": 75150 }, { "grad_norm": 0.3652385473251343, "learning_rate": 9.977758601700805e-07, "loss": 0.0213, "step": 75160 }, { "grad_norm": 0.31156063079833984, "learning_rate": 9.93671622905945e-07, "loss": 0.0225, "step": 75170 }, { "grad_norm": 0.19608227908611298, "learning_rate": 9.895757594830024e-07, "loss": 0.0153, "step": 75180 }, { "grad_norm": 0.41429030895233154, "learning_rate": 9.854882706011204e-07, "loss": 0.0155, "step": 75190 }, { "grad_norm": 0.28829652070999146, "learning_rate": 9.814091569587513e-07, "loss": 0.0189, "step": 75200 }, { "grad_norm": 0.39064034819602966, "learning_rate": 9.77338419252888e-07, "loss": 0.0222, "step": 75210 }, { "grad_norm": 0.23559802770614624, "learning_rate": 9.732760581791289e-07, "loss": 0.0153, "step": 75220 }, { "grad_norm": 0.17873400449752808, "learning_rate": 9.692220744315972e-07, "loss": 0.0202, "step": 75230 }, { "grad_norm": 0.1561957746744156, "learning_rate": 9.651764687030162e-07, "loss": 0.0176, "step": 75240 }, { "grad_norm": 0.18473170697689056, "learning_rate": 9.611392416846776e-07, "loss": 0.0148, "step": 75250 }, { "grad_norm": 0.44551345705986023, "learning_rate": 9.571103940664238e-07, "loss": 0.021, "step": 75260 }, { "grad_norm": 0.31753915548324585, "learning_rate": 9.530899265366822e-07, "loss": 0.0192, "step": 75270 }, { "grad_norm": 0.31716495752334595, "learning_rate": 9.490778397824307e-07, "loss": 0.0194, "step": 75280 }, { "grad_norm": 0.3290148377418518, "learning_rate": 9.450741344892322e-07, "loss": 0.0197, "step": 75290 }, { "grad_norm": 0.3271871507167816, "learning_rate": 9.410788113412116e-07, "loss": 0.0129, "step": 75300 }, { "grad_norm": 0.24134796857833862, "learning_rate": 9.370918710210563e-07, "loss": 0.0161, "step": 75310 }, { "grad_norm": 0.2807842195034027, "learning_rate": 9.331133142100268e-07, "loss": 0.0147, "step": 75320 }, { "grad_norm": 0.19295907020568848, "learning_rate": 9.291431415879459e-07, "loss": 0.0141, "step": 75330 }, { "grad_norm": 0.30990761518478394, "learning_rate": 9.251813538332155e-07, "loss": 0.0242, "step": 75340 }, { "grad_norm": 0.11268159747123718, "learning_rate": 9.212279516227884e-07, "loss": 0.0145, "step": 75350 }, { "grad_norm": 0.2276173233985901, "learning_rate": 9.172829356322021e-07, "loss": 0.0124, "step": 75360 }, { "grad_norm": 0.1609441041946411, "learning_rate": 9.133463065355452e-07, "loss": 0.0144, "step": 75370 }, { "grad_norm": 0.2610210180282593, "learning_rate": 9.094180650054796e-07, "loss": 0.0251, "step": 75380 }, { "grad_norm": 0.2984462380409241, "learning_rate": 9.054982117132404e-07, "loss": 0.0157, "step": 75390 }, { "grad_norm": 0.6409175992012024, "learning_rate": 9.015867473286143e-07, "loss": 0.0175, "step": 75400 }, { "grad_norm": 0.30903810262680054, "learning_rate": 8.976836725199778e-07, "loss": 0.0171, "step": 75410 }, { "grad_norm": 0.3175199329853058, "learning_rate": 8.937889879542416e-07, "loss": 0.0175, "step": 75420 }, { "grad_norm": 0.17996950447559357, "learning_rate": 8.899026942969068e-07, "loss": 0.017, "step": 75430 }, { "grad_norm": 0.268954336643219, "learning_rate": 8.860247922120424e-07, "loss": 0.0125, "step": 75440 }, { "grad_norm": 0.33019837737083435, "learning_rate": 8.821552823622737e-07, "loss": 0.0167, "step": 75450 }, { "grad_norm": 0.17973771691322327, "learning_rate": 8.782941654087884e-07, "loss": 0.0182, "step": 75460 }, { "grad_norm": 0.4309979975223541, "learning_rate": 8.744414420113478e-07, "loss": 0.0254, "step": 75470 }, { "grad_norm": 0.3153364658355713, "learning_rate": 8.705971128282753e-07, "loss": 0.0158, "step": 75480 }, { "grad_norm": 0.34599021077156067, "learning_rate": 8.66761178516473e-07, "loss": 0.0154, "step": 75490 }, { "grad_norm": 0.3555338680744171, "learning_rate": 8.629336397313781e-07, "loss": 0.0164, "step": 75500 }, { "grad_norm": 0.2517484724521637, "learning_rate": 8.591144971270282e-07, "loss": 0.0163, "step": 75510 }, { "grad_norm": 0.3118005692958832, "learning_rate": 8.553037513560069e-07, "loss": 0.0198, "step": 75520 }, { "grad_norm": 0.2986644506454468, "learning_rate": 8.515014030694546e-07, "loss": 0.0147, "step": 75530 }, { "grad_norm": 0.20138004422187805, "learning_rate": 8.47707452917107e-07, "loss": 0.0144, "step": 75540 }, { "grad_norm": 0.37556779384613037, "learning_rate": 8.439219015472343e-07, "loss": 0.0232, "step": 75550 }, { "grad_norm": 0.20489519834518433, "learning_rate": 8.401447496066861e-07, "loss": 0.0138, "step": 75560 }, { "grad_norm": 0.2729356586933136, "learning_rate": 8.363759977408792e-07, "loss": 0.0158, "step": 75570 }, { "grad_norm": 0.17493784427642822, "learning_rate": 8.326156465937817e-07, "loss": 0.0148, "step": 75580 }, { "grad_norm": 0.6081076264381409, "learning_rate": 8.28863696807941e-07, "loss": 0.023, "step": 75590 }, { "grad_norm": 0.3495067059993744, "learning_rate": 8.251201490244609e-07, "loss": 0.0202, "step": 75600 }, { "grad_norm": 0.33035099506378174, "learning_rate": 8.213850038830129e-07, "loss": 0.0152, "step": 75610 }, { "grad_norm": 0.34714993834495544, "learning_rate": 8.176582620218309e-07, "loss": 0.0146, "step": 75620 }, { "grad_norm": 0.2053098976612091, "learning_rate": 8.139399240777057e-07, "loss": 0.0161, "step": 75630 }, { "grad_norm": 0.3235895335674286, "learning_rate": 8.102299906860122e-07, "loss": 0.0187, "step": 75640 }, { "grad_norm": 0.40786272287368774, "learning_rate": 8.065284624806657e-07, "loss": 0.0208, "step": 75650 }, { "grad_norm": 0.22486470639705658, "learning_rate": 8.028353400941601e-07, "loss": 0.0157, "step": 75660 }, { "grad_norm": 0.26377832889556885, "learning_rate": 7.991506241575575e-07, "loss": 0.0194, "step": 75670 }, { "grad_norm": 0.3266942799091339, "learning_rate": 7.954743153004541e-07, "loss": 0.0195, "step": 75680 }, { "grad_norm": 0.21951036155223846, "learning_rate": 7.918064141510528e-07, "loss": 0.0166, "step": 75690 }, { "grad_norm": 0.16948726773262024, "learning_rate": 7.881469213360859e-07, "loss": 0.0189, "step": 75700 }, { "grad_norm": 0.28655552864074707, "learning_rate": 7.844958374808642e-07, "loss": 0.0158, "step": 75710 }, { "grad_norm": 0.19431819021701813, "learning_rate": 7.808531632092608e-07, "loss": 0.0168, "step": 75720 }, { "grad_norm": 0.23253478109836578, "learning_rate": 7.772188991436946e-07, "loss": 0.0179, "step": 75730 }, { "grad_norm": 0.2405296415090561, "learning_rate": 7.735930459051799e-07, "loss": 0.014, "step": 75740 }, { "grad_norm": 0.27950242161750793, "learning_rate": 7.699756041132655e-07, "loss": 0.0168, "step": 75750 }, { "grad_norm": 0.38027769327163696, "learning_rate": 7.663665743860793e-07, "loss": 0.0174, "step": 75760 }, { "grad_norm": 0.28000757098197937, "learning_rate": 7.627659573403001e-07, "loss": 0.0176, "step": 75770 }, { "grad_norm": 0.38906967639923096, "learning_rate": 7.591737535911802e-07, "loss": 0.0186, "step": 75780 }, { "grad_norm": 0.318957656621933, "learning_rate": 7.555899637525288e-07, "loss": 0.0177, "step": 75790 }, { "grad_norm": 0.5096712112426758, "learning_rate": 7.520145884367058e-07, "loss": 0.0151, "step": 75800 }, { "grad_norm": 0.4289866089820862, "learning_rate": 7.484476282546615e-07, "loss": 0.0168, "step": 75810 }, { "grad_norm": 0.4955591559410095, "learning_rate": 7.448890838158806e-07, "loss": 0.0244, "step": 75820 }, { "grad_norm": 0.41796061396598816, "learning_rate": 7.413389557284267e-07, "loss": 0.0155, "step": 75830 }, { "grad_norm": 0.19531464576721191, "learning_rate": 7.377972445989201e-07, "loss": 0.0171, "step": 75840 }, { "grad_norm": 0.31010180711746216, "learning_rate": 7.342639510325377e-07, "loss": 0.0197, "step": 75850 }, { "grad_norm": 0.6234784126281738, "learning_rate": 7.307390756330246e-07, "loss": 0.0202, "step": 75860 }, { "grad_norm": 0.39045560359954834, "learning_rate": 7.272226190026876e-07, "loss": 0.0147, "step": 75870 }, { "grad_norm": 0.18465253710746765, "learning_rate": 7.237145817423907e-07, "loss": 0.0141, "step": 75880 }, { "grad_norm": 0.1471930742263794, "learning_rate": 7.202149644515654e-07, "loss": 0.0173, "step": 75890 }, { "grad_norm": 0.31139805912971497, "learning_rate": 7.167237677281946e-07, "loss": 0.0164, "step": 75900 }, { "grad_norm": 0.41597384214401245, "learning_rate": 7.132409921688288e-07, "loss": 0.0169, "step": 75910 }, { "grad_norm": 0.2780321538448334, "learning_rate": 7.09766638368592e-07, "loss": 0.0167, "step": 75920 }, { "grad_norm": 0.4312070906162262, "learning_rate": 7.063007069211313e-07, "loss": 0.0157, "step": 75930 }, { "grad_norm": 0.28161120414733887, "learning_rate": 7.028431984187067e-07, "loss": 0.0163, "step": 75940 }, { "grad_norm": 0.36802181601524353, "learning_rate": 6.993941134520898e-07, "loss": 0.0142, "step": 75950 }, { "grad_norm": 0.27587178349494934, "learning_rate": 6.959534526106537e-07, "loss": 0.0159, "step": 75960 }, { "grad_norm": 0.16244806349277496, "learning_rate": 6.925212164822948e-07, "loss": 0.0151, "step": 75970 }, { "grad_norm": 0.3333382308483124, "learning_rate": 6.890974056535049e-07, "loss": 0.0243, "step": 75980 }, { "grad_norm": 0.4940394163131714, "learning_rate": 6.856820207093106e-07, "loss": 0.0164, "step": 75990 }, { "grad_norm": 0.16399918496608734, "learning_rate": 6.822750622333063e-07, "loss": 0.0114, "step": 76000 }, { "grad_norm": 0.32072341442108154, "learning_rate": 6.788765308076539e-07, "loss": 0.0234, "step": 76010 }, { "grad_norm": 0.22303062677383423, "learning_rate": 6.754864270130668e-07, "loss": 0.0132, "step": 76020 }, { "grad_norm": 0.10182664543390274, "learning_rate": 6.721047514288203e-07, "loss": 0.0144, "step": 76030 }, { "grad_norm": 0.2970001697540283, "learning_rate": 6.687315046327469e-07, "loss": 0.0181, "step": 76040 }, { "grad_norm": 0.4094281494617462, "learning_rate": 6.65366687201252e-07, "loss": 0.0155, "step": 76050 }, { "grad_norm": 0.22923925518989563, "learning_rate": 6.620102997092814e-07, "loss": 0.0211, "step": 76060 }, { "grad_norm": 0.17600010335445404, "learning_rate": 6.586623427303596e-07, "loss": 0.015, "step": 76070 }, { "grad_norm": 0.2287420779466629, "learning_rate": 6.553228168365455e-07, "loss": 0.0183, "step": 76080 }, { "grad_norm": 0.38284701108932495, "learning_rate": 6.519917225984884e-07, "loss": 0.0221, "step": 76090 }, { "grad_norm": 0.38630256056785583, "learning_rate": 6.486690605853718e-07, "loss": 0.0181, "step": 76100 }, { "grad_norm": 0.31488001346588135, "learning_rate": 6.453548313649527e-07, "loss": 0.0183, "step": 76110 }, { "grad_norm": 0.4017728269100189, "learning_rate": 6.420490355035446e-07, "loss": 0.0169, "step": 76120 }, { "grad_norm": 0.1567453145980835, "learning_rate": 6.387516735660071e-07, "loss": 0.0154, "step": 76130 }, { "grad_norm": 0.21865886449813843, "learning_rate": 6.35462746115778e-07, "loss": 0.0163, "step": 76140 }, { "grad_norm": 0.24346201121807098, "learning_rate": 6.321822537148358e-07, "loss": 0.0207, "step": 76150 }, { "grad_norm": 0.2963792383670807, "learning_rate": 6.289101969237432e-07, "loss": 0.0215, "step": 76160 }, { "grad_norm": 0.41291239857673645, "learning_rate": 6.256465763015918e-07, "loss": 0.0151, "step": 76170 }, { "grad_norm": 0.16345953941345215, "learning_rate": 6.22391392406052e-07, "loss": 0.0145, "step": 76180 }, { "grad_norm": 0.4421111047267914, "learning_rate": 6.191446457933403e-07, "loss": 0.0157, "step": 76190 }, { "grad_norm": 0.24295081198215485, "learning_rate": 6.159063370182406e-07, "loss": 0.0196, "step": 76200 }, { "grad_norm": 0.5268405079841614, "learning_rate": 6.126764666340879e-07, "loss": 0.0153, "step": 76210 }, { "grad_norm": 0.38117265701293945, "learning_rate": 6.094550351927852e-07, "loss": 0.0182, "step": 76220 }, { "grad_norm": 0.22186188399791718, "learning_rate": 6.062420432447757e-07, "loss": 0.0163, "step": 76230 }, { "grad_norm": 0.2761910557746887, "learning_rate": 6.030374913390813e-07, "loss": 0.0163, "step": 76240 }, { "grad_norm": 0.31321150064468384, "learning_rate": 5.998413800232694e-07, "loss": 0.0148, "step": 76250 }, { "grad_norm": 0.5382955074310303, "learning_rate": 5.966537098434755e-07, "loss": 0.0182, "step": 76260 }, { "grad_norm": 0.2946998178958893, "learning_rate": 5.934744813443694e-07, "loss": 0.0159, "step": 76270 }, { "grad_norm": 0.2625904083251953, "learning_rate": 5.903036950692054e-07, "loss": 0.0171, "step": 76280 }, { "grad_norm": 0.20153750479221344, "learning_rate": 5.871413515597835e-07, "loss": 0.0184, "step": 76290 }, { "grad_norm": 0.2070002406835556, "learning_rate": 5.839874513564547e-07, "loss": 0.0127, "step": 76300 }, { "grad_norm": 0.38509446382522583, "learning_rate": 5.808419949981436e-07, "loss": 0.0176, "step": 76310 }, { "grad_norm": 1.0441288948059082, "learning_rate": 5.777049830223257e-07, "loss": 0.0192, "step": 76320 }, { "grad_norm": 0.1676703542470932, "learning_rate": 5.745764159650114e-07, "loss": 0.0154, "step": 76330 }, { "grad_norm": 0.4432491958141327, "learning_rate": 5.71456294360806e-07, "loss": 0.0169, "step": 76340 }, { "grad_norm": 0.26458045840263367, "learning_rate": 5.683446187428443e-07, "loss": 0.016, "step": 76350 }, { "grad_norm": 0.4007011950016022, "learning_rate": 5.652413896428288e-07, "loss": 0.0155, "step": 76360 }, { "grad_norm": 0.23979495465755463, "learning_rate": 5.621466075910131e-07, "loss": 0.0166, "step": 76370 }, { "grad_norm": 0.3819877803325653, "learning_rate": 5.590602731162187e-07, "loss": 0.0201, "step": 76380 }, { "grad_norm": 0.5597291588783264, "learning_rate": 5.559823867458125e-07, "loss": 0.0174, "step": 76390 }, { "grad_norm": 0.19536080956459045, "learning_rate": 5.529129490057128e-07, "loss": 0.0221, "step": 76400 }, { "grad_norm": 0.29926976561546326, "learning_rate": 5.498519604204167e-07, "loss": 0.015, "step": 76410 }, { "grad_norm": 0.21614739298820496, "learning_rate": 5.467994215129501e-07, "loss": 0.0137, "step": 76420 }, { "grad_norm": 0.3937314450740814, "learning_rate": 5.437553328049183e-07, "loss": 0.0181, "step": 76430 }, { "grad_norm": 0.18430382013320923, "learning_rate": 5.407196948164661e-07, "loss": 0.0136, "step": 76440 }, { "grad_norm": 0.16925576329231262, "learning_rate": 5.376925080663009e-07, "loss": 0.0125, "step": 76450 }, { "grad_norm": 0.2634519934654236, "learning_rate": 5.346737730716977e-07, "loss": 0.0207, "step": 76460 }, { "grad_norm": 0.23841789364814758, "learning_rate": 5.316634903484607e-07, "loss": 0.0155, "step": 76470 }, { "grad_norm": 0.22748686373233795, "learning_rate": 5.286616604109728e-07, "loss": 0.0154, "step": 76480 }, { "grad_norm": 0.1781555563211441, "learning_rate": 5.256682837721627e-07, "loss": 0.018, "step": 76490 }, { "grad_norm": 0.21845892071723938, "learning_rate": 5.226833609435156e-07, "loss": 0.0143, "step": 76500 }, { "grad_norm": 0.2632956802845001, "learning_rate": 5.19706892435079e-07, "loss": 0.0165, "step": 76510 }, { "grad_norm": 0.28336793184280396, "learning_rate": 5.167388787554406e-07, "loss": 0.0187, "step": 76520 }, { "grad_norm": 0.4514523446559906, "learning_rate": 5.137793204117614e-07, "loss": 0.0175, "step": 76530 }, { "grad_norm": 0.43192633986473083, "learning_rate": 5.108282179097479e-07, "loss": 0.0219, "step": 76540 }, { "grad_norm": 0.30229756236076355, "learning_rate": 5.078855717536523e-07, "loss": 0.0148, "step": 76550 }, { "grad_norm": 0.18351858854293823, "learning_rate": 5.049513824463059e-07, "loss": 0.0144, "step": 76560 }, { "grad_norm": 0.17239969968795776, "learning_rate": 5.020256504890742e-07, "loss": 0.023, "step": 76570 }, { "grad_norm": 0.20492763817310333, "learning_rate": 4.991083763818849e-07, "loss": 0.017, "step": 76580 }, { "grad_norm": 0.7494048476219177, "learning_rate": 4.961995606232228e-07, "loss": 0.0273, "step": 76590 }, { "grad_norm": 0.25990286469459534, "learning_rate": 4.932992037101236e-07, "loss": 0.0158, "step": 76600 }, { "grad_norm": 0.40400657057762146, "learning_rate": 4.904073061381798e-07, "loss": 0.0213, "step": 76610 }, { "grad_norm": 0.28520113229751587, "learning_rate": 4.875238684015349e-07, "loss": 0.0192, "step": 76620 }, { "grad_norm": 0.2739093005657196, "learning_rate": 4.846488909928948e-07, "loss": 0.0144, "step": 76630 }, { "grad_norm": 0.3062308728694916, "learning_rate": 4.81782374403511e-07, "loss": 0.0233, "step": 76640 }, { "grad_norm": 0.3231646716594696, "learning_rate": 4.789243191231918e-07, "loss": 0.0184, "step": 76650 }, { "grad_norm": 0.5128914713859558, "learning_rate": 4.76074725640302e-07, "loss": 0.017, "step": 76660 }, { "grad_norm": 0.2916349470615387, "learning_rate": 4.732335944417632e-07, "loss": 0.0192, "step": 76670 }, { "grad_norm": 0.4786636233329773, "learning_rate": 4.704009260130371e-07, "loss": 0.0174, "step": 76680 }, { "grad_norm": 0.47180119156837463, "learning_rate": 4.675767208381587e-07, "loss": 0.0236, "step": 76690 }, { "grad_norm": 0.3288190960884094, "learning_rate": 4.647609793997032e-07, "loss": 0.0199, "step": 76700 }, { "grad_norm": 0.3275269567966461, "learning_rate": 4.619537021788023e-07, "loss": 0.0153, "step": 76710 }, { "grad_norm": 0.269583135843277, "learning_rate": 4.5915488965515005e-07, "loss": 0.0162, "step": 76720 }, { "grad_norm": 0.31802481412887573, "learning_rate": 4.563645423069807e-07, "loss": 0.0136, "step": 76730 }, { "grad_norm": 0.4037818908691406, "learning_rate": 4.535826606110849e-07, "loss": 0.0157, "step": 76740 }, { "grad_norm": 0.47493407130241394, "learning_rate": 4.508092450428214e-07, "loss": 0.0169, "step": 76750 }, { "grad_norm": 0.21279394626617432, "learning_rate": 4.480442960760778e-07, "loss": 0.0211, "step": 76760 }, { "grad_norm": 0.21241231262683868, "learning_rate": 4.4528781418332053e-07, "loss": 0.0158, "step": 76770 }, { "grad_norm": 0.3187108337879181, "learning_rate": 4.4253979983555073e-07, "loss": 0.0169, "step": 76780 }, { "grad_norm": 0.1981847733259201, "learning_rate": 4.3980025350233154e-07, "loss": 0.0136, "step": 76790 }, { "grad_norm": 0.23125310242176056, "learning_rate": 4.370691756517664e-07, "loss": 0.0198, "step": 76800 }, { "grad_norm": 0.26463207602500916, "learning_rate": 4.3434656675053753e-07, "loss": 0.0154, "step": 76810 }, { "grad_norm": 0.24359571933746338, "learning_rate": 4.3163242726385613e-07, "loss": 0.016, "step": 76820 }, { "grad_norm": 0.48860225081443787, "learning_rate": 4.289267576554956e-07, "loss": 0.0171, "step": 76830 }, { "grad_norm": 0.34589099884033203, "learning_rate": 4.262295583877807e-07, "loss": 0.0146, "step": 76840 }, { "grad_norm": 0.2041931450366974, "learning_rate": 4.235408299215815e-07, "loss": 0.0196, "step": 76850 }, { "grad_norm": 0.3456515669822693, "learning_rate": 4.2086057271634173e-07, "loss": 0.0114, "step": 76860 }, { "grad_norm": 0.30292317271232605, "learning_rate": 4.181887872300394e-07, "loss": 0.0189, "step": 76870 }, { "grad_norm": 0.32266703248023987, "learning_rate": 4.155254739192038e-07, "loss": 0.0159, "step": 76880 }, { "grad_norm": 0.35369759798049927, "learning_rate": 4.128706332389265e-07, "loss": 0.0196, "step": 76890 }, { "grad_norm": 0.3724673390388489, "learning_rate": 4.1022426564284453e-07, "loss": 0.0135, "step": 76900 }, { "grad_norm": 0.3656960129737854, "learning_rate": 4.075863715831574e-07, "loss": 0.0166, "step": 76910 }, { "grad_norm": 0.27588507533073425, "learning_rate": 4.0495695151059887e-07, "loss": 0.0202, "step": 76920 }, { "grad_norm": 0.27089613676071167, "learning_rate": 4.023360058744763e-07, "loss": 0.0139, "step": 76930 }, { "grad_norm": 0.3243729770183563, "learning_rate": 3.997235351226314e-07, "loss": 0.0145, "step": 76940 }, { "grad_norm": 0.28094998002052307, "learning_rate": 3.971195397014571e-07, "loss": 0.0175, "step": 76950 }, { "grad_norm": 0.3447355031967163, "learning_rate": 3.9452402005591417e-07, "loss": 0.0213, "step": 76960 }, { "grad_norm": 0.24685510993003845, "learning_rate": 3.9193697662950334e-07, "loss": 0.0152, "step": 76970 }, { "grad_norm": 0.386227548122406, "learning_rate": 3.893584098642822e-07, "loss": 0.0204, "step": 76980 }, { "grad_norm": 0.20572903752326965, "learning_rate": 3.867883202008538e-07, "loss": 0.0156, "step": 76990 }, { "grad_norm": 0.36403337121009827, "learning_rate": 3.842267080783779e-07, "loss": 0.0178, "step": 77000 }, { "grad_norm": 0.09395406395196915, "learning_rate": 3.816735739345656e-07, "loss": 0.0196, "step": 77010 }, { "grad_norm": 0.22758188843727112, "learning_rate": 3.791289182056679e-07, "loss": 0.0197, "step": 77020 }, { "grad_norm": 0.15918952226638794, "learning_rate": 3.7659274132650913e-07, "loss": 0.0178, "step": 77030 }, { "grad_norm": 0.21730826795101166, "learning_rate": 3.7406504373044826e-07, "loss": 0.0148, "step": 77040 }, { "grad_norm": 0.27801549434661865, "learning_rate": 3.7154582584939533e-07, "loss": 0.0173, "step": 77050 }, { "grad_norm": 0.38915592432022095, "learning_rate": 3.690350881138227e-07, "loss": 0.0193, "step": 77060 }, { "grad_norm": 0.2709554433822632, "learning_rate": 3.665328309527427e-07, "loss": 0.0148, "step": 77070 }, { "grad_norm": 0.15579015016555786, "learning_rate": 3.640390547937245e-07, "loss": 0.0176, "step": 77080 }, { "grad_norm": 0.6372251510620117, "learning_rate": 3.615537600628882e-07, "loss": 0.019, "step": 77090 }, { "grad_norm": 0.2942940294742584, "learning_rate": 3.5907694718489427e-07, "loss": 0.0162, "step": 77100 }, { "grad_norm": 0.21921664476394653, "learning_rate": 3.566086165829707e-07, "loss": 0.0192, "step": 77110 }, { "grad_norm": 0.5002244710922241, "learning_rate": 3.5414876867888024e-07, "loss": 0.0203, "step": 77120 }, { "grad_norm": 0.317415326833725, "learning_rate": 3.5169740389295326e-07, "loss": 0.0205, "step": 77130 }, { "grad_norm": 0.48980817198753357, "learning_rate": 3.4925452264405466e-07, "loss": 0.018, "step": 77140 }, { "grad_norm": 0.3436902165412903, "learning_rate": 3.468201253496062e-07, "loss": 0.0141, "step": 77150 }, { "grad_norm": 0.3541587293148041, "learning_rate": 3.443942124255861e-07, "loss": 0.0187, "step": 77160 }, { "grad_norm": 0.1922972947359085, "learning_rate": 3.4197678428650183e-07, "loss": 0.0187, "step": 77170 }, { "grad_norm": 0.18573665618896484, "learning_rate": 3.3956784134544504e-07, "loss": 0.0195, "step": 77180 }, { "grad_norm": 0.2165464609861374, "learning_rate": 3.3716738401402547e-07, "loss": 0.0198, "step": 77190 }, { "grad_norm": 0.23176443576812744, "learning_rate": 3.3477541270241495e-07, "loss": 0.021, "step": 77200 }, { "grad_norm": 0.20066185295581818, "learning_rate": 3.3239192781934215e-07, "loss": 0.0188, "step": 77210 }, { "grad_norm": 0.2777591049671173, "learning_rate": 3.3001692977207563e-07, "loss": 0.0185, "step": 77220 }, { "grad_norm": 0.3103960454463959, "learning_rate": 3.276504189664409e-07, "loss": 0.0182, "step": 77230 }, { "grad_norm": 0.22291702032089233, "learning_rate": 3.252923958068088e-07, "loss": 0.0147, "step": 77240 }, { "grad_norm": 0.27172550559043884, "learning_rate": 3.2294286069609046e-07, "loss": 0.0191, "step": 77250 }, { "grad_norm": 0.3615378439426422, "learning_rate": 3.2060181403577583e-07, "loss": 0.0216, "step": 77260 }, { "grad_norm": 0.24167953431606293, "learning_rate": 3.1826925622587823e-07, "loss": 0.0154, "step": 77270 }, { "grad_norm": 0.331723690032959, "learning_rate": 3.159451876649622e-07, "loss": 0.0185, "step": 77280 }, { "grad_norm": 0.3963790237903595, "learning_rate": 3.1362960875015444e-07, "loss": 0.0129, "step": 77290 }, { "grad_norm": 0.47270363569259644, "learning_rate": 3.1132251987711637e-07, "loss": 0.0253, "step": 77300 }, { "grad_norm": 0.22568227350711823, "learning_rate": 3.0902392144007695e-07, "loss": 0.0174, "step": 77310 }, { "grad_norm": 0.23398658633232117, "learning_rate": 3.0673381383179436e-07, "loss": 0.0157, "step": 77320 }, { "grad_norm": 0.2570967972278595, "learning_rate": 3.0445219744358875e-07, "loss": 0.0197, "step": 77330 }, { "grad_norm": 0.2673448324203491, "learning_rate": 3.021790726653262e-07, "loss": 0.0233, "step": 77340 }, { "grad_norm": 0.2741694748401642, "learning_rate": 2.9991443988542366e-07, "loss": 0.0142, "step": 77350 }, { "grad_norm": 0.26177361607551575, "learning_rate": 2.976582994908439e-07, "loss": 0.0182, "step": 77360 }, { "grad_norm": 0.3424692153930664, "learning_rate": 2.954106518671007e-07, "loss": 0.0148, "step": 77370 }, { "grad_norm": 0.3235324025154114, "learning_rate": 2.9317149739825356e-07, "loss": 0.0171, "step": 77380 }, { "grad_norm": 0.35095342993736267, "learning_rate": 2.909408364669075e-07, "loss": 0.0186, "step": 77390 }, { "grad_norm": 0.4330078661441803, "learning_rate": 2.887186694542299e-07, "loss": 0.0194, "step": 77400 }, { "grad_norm": 0.2257402241230011, "learning_rate": 2.865049967399225e-07, "loss": 0.0188, "step": 77410 }, { "grad_norm": 0.4422999322414398, "learning_rate": 2.842998187022439e-07, "loss": 0.0243, "step": 77420 }, { "grad_norm": 0.2785041034221649, "learning_rate": 2.8210313571800374e-07, "loss": 0.0161, "step": 77430 }, { "grad_norm": 0.4347185492515564, "learning_rate": 2.7991494816255184e-07, "loss": 0.0147, "step": 77440 }, { "grad_norm": 0.23832806944847107, "learning_rate": 2.777352564097779e-07, "loss": 0.0146, "step": 77450 }, { "grad_norm": 0.28081873059272766, "learning_rate": 2.755640608321508e-07, "loss": 0.0165, "step": 77460 }, { "grad_norm": 0.25835302472114563, "learning_rate": 2.73401361800657e-07, "loss": 0.0163, "step": 77470 }, { "grad_norm": 0.3457399904727936, "learning_rate": 2.7124715968484537e-07, "loss": 0.0163, "step": 77480 }, { "grad_norm": 0.39151760935783386, "learning_rate": 2.691014548528104e-07, "loss": 0.0184, "step": 77490 }, { "grad_norm": 0.3371528685092926, "learning_rate": 2.669642476711864e-07, "loss": 0.018, "step": 77500 }, { "grad_norm": 0.423445463180542, "learning_rate": 2.648355385051815e-07, "loss": 0.0193, "step": 77510 }, { "grad_norm": 0.6351311206817627, "learning_rate": 2.627153277185157e-07, "loss": 0.0173, "step": 77520 }, { "grad_norm": 0.2903412878513336, "learning_rate": 2.606036156734881e-07, "loss": 0.0158, "step": 77530 }, { "grad_norm": 0.2878100872039795, "learning_rate": 2.5850040273092127e-07, "loss": 0.0184, "step": 77540 }, { "grad_norm": 0.3526661694049835, "learning_rate": 2.5640568925020536e-07, "loss": 0.023, "step": 77550 }, { "grad_norm": 0.5444995164871216, "learning_rate": 2.5431947558927083e-07, "loss": 0.0261, "step": 77560 }, { "grad_norm": 0.33422088623046875, "learning_rate": 2.522417621045825e-07, "loss": 0.0157, "step": 77570 }, { "grad_norm": 0.296424001455307, "learning_rate": 2.501725491511786e-07, "loss": 0.019, "step": 77580 }, { "grad_norm": 0.4131651222705841, "learning_rate": 2.481118370826263e-07, "loss": 0.0156, "step": 77590 }, { "grad_norm": 0.22808687388896942, "learning_rate": 2.4605962625104393e-07, "loss": 0.0151, "step": 77600 }, { "grad_norm": 0.3900807201862335, "learning_rate": 2.440159170070955e-07, "loss": 0.0179, "step": 77610 }, { "grad_norm": 0.26851364970207214, "learning_rate": 2.41980709699996e-07, "loss": 0.0186, "step": 77620 }, { "grad_norm": 0.42227742075920105, "learning_rate": 2.3995400467751174e-07, "loss": 0.0167, "step": 77630 }, { "grad_norm": 0.6485266089439392, "learning_rate": 2.3793580228594902e-07, "loss": 0.0261, "step": 77640 }, { "grad_norm": 0.25894203782081604, "learning_rate": 2.3592610287015982e-07, "loss": 0.0186, "step": 77650 }, { "grad_norm": 0.27077624201774597, "learning_rate": 2.3392490677354718e-07, "loss": 0.019, "step": 77660 }, { "grad_norm": 0.33226925134658813, "learning_rate": 2.3193221433806533e-07, "loss": 0.0162, "step": 77670 }, { "grad_norm": 0.24275964498519897, "learning_rate": 2.299480259042086e-07, "loss": 0.0143, "step": 77680 }, { "grad_norm": 0.36243003606796265, "learning_rate": 2.2797234181102244e-07, "loss": 0.0179, "step": 77690 }, { "grad_norm": 0.25507816672325134, "learning_rate": 2.2600516239609238e-07, "loss": 0.017, "step": 77700 }, { "grad_norm": 0.24989238381385803, "learning_rate": 2.240464879955606e-07, "loss": 0.0246, "step": 77710 }, { "grad_norm": 0.26704925298690796, "learning_rate": 2.2209631894410387e-07, "loss": 0.0143, "step": 77720 }, { "grad_norm": 0.21287372708320618, "learning_rate": 2.2015465557496117e-07, "loss": 0.0173, "step": 77730 }, { "grad_norm": 0.3419766426086426, "learning_rate": 2.1822149821990602e-07, "loss": 0.0175, "step": 77740 }, { "grad_norm": 0.3108263313770294, "learning_rate": 2.1629684720926303e-07, "loss": 0.02, "step": 77750 }, { "grad_norm": 0.3878709077835083, "learning_rate": 2.1438070287189693e-07, "loss": 0.0156, "step": 77760 }, { "grad_norm": 0.27002131938934326, "learning_rate": 2.1247306553523472e-07, "loss": 0.0156, "step": 77770 }, { "grad_norm": 0.28615424036979675, "learning_rate": 2.105739355252323e-07, "loss": 0.0186, "step": 77780 }, { "grad_norm": 0.15610149502754211, "learning_rate": 2.0868331316639678e-07, "loss": 0.0134, "step": 77790 }, { "grad_norm": 0.4593779146671295, "learning_rate": 2.0680119878179193e-07, "loss": 0.0144, "step": 77800 }, { "grad_norm": 0.5870087146759033, "learning_rate": 2.0492759269301054e-07, "loss": 0.018, "step": 77810 }, { "grad_norm": 0.2061602771282196, "learning_rate": 2.0306249522021315e-07, "loss": 0.017, "step": 77820 }, { "grad_norm": 0.26071402430534363, "learning_rate": 2.0120590668207816e-07, "loss": 0.0149, "step": 77830 }, { "grad_norm": 0.2643389403820038, "learning_rate": 1.993578273958574e-07, "loss": 0.0167, "step": 77840 }, { "grad_norm": 0.3501549959182739, "learning_rate": 1.975182576773371e-07, "loss": 0.0183, "step": 77850 }, { "grad_norm": 0.5106648802757263, "learning_rate": 1.9568719784083812e-07, "loss": 0.0228, "step": 77860 }, { "grad_norm": 0.642571747303009, "learning_rate": 1.9386464819924898e-07, "loss": 0.0166, "step": 77870 }, { "grad_norm": 0.5085776448249817, "learning_rate": 1.9205060906399285e-07, "loss": 0.0201, "step": 77880 }, { "grad_norm": 0.13747194409370422, "learning_rate": 1.9024508074503845e-07, "loss": 0.0174, "step": 77890 }, { "grad_norm": 0.2840857207775116, "learning_rate": 1.8844806355089452e-07, "loss": 0.0248, "step": 77900 }, { "grad_norm": 0.477591872215271, "learning_rate": 1.8665955778863208e-07, "loss": 0.028, "step": 77910 }, { "grad_norm": 0.4068700969219208, "learning_rate": 1.848795637638512e-07, "loss": 0.0167, "step": 77920 }, { "grad_norm": 0.3458728492259979, "learning_rate": 1.831080817807085e-07, "loss": 0.0218, "step": 77930 }, { "grad_norm": 0.3742872178554535, "learning_rate": 1.813451121418952e-07, "loss": 0.0261, "step": 77940 }, { "grad_norm": 0.47312042117118835, "learning_rate": 1.795906551486648e-07, "loss": 0.0158, "step": 77950 }, { "grad_norm": 0.23312094807624817, "learning_rate": 1.7784471110079414e-07, "loss": 0.0195, "step": 77960 }, { "grad_norm": 0.2569481134414673, "learning_rate": 1.7610728029662793e-07, "loss": 0.0165, "step": 77970 }, { "grad_norm": 0.1783342957496643, "learning_rate": 1.7437836303303979e-07, "loss": 0.0194, "step": 77980 }, { "grad_norm": 0.2797814607620239, "learning_rate": 1.726579596054545e-07, "loss": 0.0139, "step": 77990 }, { "grad_norm": 0.31289809942245483, "learning_rate": 1.7094607030784803e-07, "loss": 0.0178, "step": 78000 }, { "grad_norm": 0.1839347630739212, "learning_rate": 1.6924269543272532e-07, "loss": 0.02, "step": 78010 }, { "grad_norm": 0.25211092829704285, "learning_rate": 1.6754783527115348e-07, "loss": 0.0143, "step": 78020 }, { "grad_norm": 0.2272985726594925, "learning_rate": 1.6586149011273422e-07, "loss": 0.0179, "step": 78030 }, { "grad_norm": 0.16357852518558502, "learning_rate": 1.6418366024562038e-07, "loss": 0.0203, "step": 78040 }, { "grad_norm": 0.28580358624458313, "learning_rate": 1.6251434595651037e-07, "loss": 0.0164, "step": 78050 }, { "grad_norm": 0.46933773159980774, "learning_rate": 1.608535475306372e-07, "loss": 0.0196, "step": 78060 }, { "grad_norm": 0.25783172249794006, "learning_rate": 1.5920126525179048e-07, "loss": 0.0195, "step": 78070 }, { "grad_norm": 0.24937671422958374, "learning_rate": 1.5755749940229435e-07, "loss": 0.0132, "step": 78080 }, { "grad_norm": 0.2691832184791565, "learning_rate": 1.559222502630353e-07, "loss": 0.0165, "step": 78090 }, { "grad_norm": 0.2333197444677353, "learning_rate": 1.5429551811341757e-07, "loss": 0.0159, "step": 78100 }, { "grad_norm": 0.3193853199481964, "learning_rate": 1.5267730323141882e-07, "loss": 0.0154, "step": 78110 }, { "grad_norm": 0.27957814931869507, "learning_rate": 1.5106760589353454e-07, "loss": 0.0163, "step": 78120 }, { "grad_norm": 0.21797682344913483, "learning_rate": 1.4946642637483355e-07, "loss": 0.0167, "step": 78130 }, { "grad_norm": 0.6110532283782959, "learning_rate": 1.4787376494889703e-07, "loss": 0.0197, "step": 78140 }, { "grad_norm": 0.24455638229846954, "learning_rate": 1.4628962188787955e-07, "loss": 0.0148, "step": 78150 }, { "grad_norm": 0.3547496497631073, "learning_rate": 1.4471399746247006e-07, "loss": 0.0204, "step": 78160 }, { "grad_norm": 0.22291892766952515, "learning_rate": 1.4314689194188103e-07, "loss": 0.0158, "step": 78170 }, { "grad_norm": 0.35225552320480347, "learning_rate": 1.4158830559390933e-07, "loss": 0.0179, "step": 78180 }, { "grad_norm": 0.24701310694217682, "learning_rate": 1.4003823868486422e-07, "loss": 0.0295, "step": 78190 }, { "grad_norm": 0.38688623905181885, "learning_rate": 1.3849669147960598e-07, "loss": 0.0181, "step": 78200 }, { "grad_norm": 0.3641364276409149, "learning_rate": 1.3696366424155726e-07, "loss": 0.0158, "step": 78210 }, { "grad_norm": 0.5076432228088379, "learning_rate": 1.35439157232653e-07, "loss": 0.0266, "step": 78220 }, { "grad_norm": 0.2753865122795105, "learning_rate": 1.3392317071340144e-07, "loss": 0.0161, "step": 78230 }, { "grad_norm": 0.5182250738143921, "learning_rate": 1.3241570494283984e-07, "loss": 0.0137, "step": 78240 }, { "grad_norm": 0.2688538432121277, "learning_rate": 1.3091676017855103e-07, "loss": 0.0198, "step": 78250 }, { "grad_norm": 0.17320598661899567, "learning_rate": 1.2942633667666904e-07, "loss": 0.0181, "step": 78260 }, { "grad_norm": 0.22825506329536438, "learning_rate": 1.2794443469185679e-07, "loss": 0.0243, "step": 78270 }, { "grad_norm": 0.34467825293540955, "learning_rate": 1.2647105447734508e-07, "loss": 0.0179, "step": 78280 }, { "grad_norm": 0.2256319522857666, "learning_rate": 1.2500619628488254e-07, "loss": 0.0178, "step": 78290 }, { "grad_norm": 0.33918070793151855, "learning_rate": 1.2354986036477446e-07, "loss": 0.0261, "step": 78300 }, { "grad_norm": 0.2877885103225708, "learning_rate": 1.221020469658718e-07, "loss": 0.0184, "step": 78310 }, { "grad_norm": 0.23323272168636322, "learning_rate": 1.2066275633556556e-07, "loss": 0.0163, "step": 78320 }, { "grad_norm": 0.2859955430030823, "learning_rate": 1.192319887197979e-07, "loss": 0.0133, "step": 78330 }, { "grad_norm": 0.1324760615825653, "learning_rate": 1.1780974436303438e-07, "loss": 0.0117, "step": 78340 }, { "grad_norm": 0.18675902485847473, "learning_rate": 1.1639602350830836e-07, "loss": 0.0155, "step": 78350 }, { "grad_norm": 0.21337145566940308, "learning_rate": 1.1499082639718217e-07, "loss": 0.0184, "step": 78360 }, { "grad_norm": 0.345670223236084, "learning_rate": 1.1359415326976374e-07, "loss": 0.0146, "step": 78370 }, { "grad_norm": 0.33184316754341125, "learning_rate": 1.1220600436470663e-07, "loss": 0.0169, "step": 78380 }, { "grad_norm": 0.2465442717075348, "learning_rate": 1.1082637991920996e-07, "loss": 0.0205, "step": 78390 }, { "grad_norm": 0.30271053314208984, "learning_rate": 1.0945528016901851e-07, "loss": 0.0154, "step": 78400 }, { "grad_norm": 0.29321280121803284, "learning_rate": 1.0809270534840599e-07, "loss": 0.0156, "step": 78410 }, { "grad_norm": 0.33118730783462524, "learning_rate": 1.067386556902028e-07, "loss": 0.0162, "step": 78420 }, { "grad_norm": 0.4603951573371887, "learning_rate": 1.05393131425785e-07, "loss": 0.0102, "step": 78430 }, { "grad_norm": 0.29465681314468384, "learning_rate": 1.0405613278505199e-07, "loss": 0.0152, "step": 78440 }, { "grad_norm": 0.26037630438804626, "learning_rate": 1.027276599964766e-07, "loss": 0.016, "step": 78450 }, { "grad_norm": 0.3061026632785797, "learning_rate": 1.0140771328704391e-07, "loss": 0.015, "step": 78460 }, { "grad_norm": 0.1714918613433838, "learning_rate": 1.0009629288231237e-07, "loss": 0.016, "step": 78470 }, { "grad_norm": 0.198738232254982, "learning_rate": 9.879339900635276e-08, "loss": 0.0176, "step": 78480 }, { "grad_norm": 0.28818339109420776, "learning_rate": 9.74990318817981e-08, "loss": 0.0201, "step": 78490 }, { "grad_norm": 0.3196941614151001, "learning_rate": 9.621319172982701e-08, "loss": 0.0164, "step": 78500 }, { "grad_norm": 0.323229044675827, "learning_rate": 9.493587877015264e-08, "loss": 0.0153, "step": 78510 }, { "grad_norm": 0.35718750953674316, "learning_rate": 9.366709322102263e-08, "loss": 0.0164, "step": 78520 }, { "grad_norm": 0.28100305795669556, "learning_rate": 9.240683529924688e-08, "loss": 0.0158, "step": 78530 }, { "grad_norm": 0.337070494890213, "learning_rate": 9.115510522016979e-08, "loss": 0.0163, "step": 78540 }, { "grad_norm": 0.32622256875038147, "learning_rate": 8.991190319767583e-08, "loss": 0.0138, "step": 78550 }, { "grad_norm": 0.1301460713148117, "learning_rate": 8.867722944419509e-08, "loss": 0.0125, "step": 78560 }, { "grad_norm": 0.20666956901550293, "learning_rate": 8.745108417069214e-08, "loss": 0.0128, "step": 78570 }, { "grad_norm": 0.343831330537796, "learning_rate": 8.623346758669381e-08, "loss": 0.0125, "step": 78580 }, { "grad_norm": 0.16701145470142365, "learning_rate": 8.502437990025037e-08, "loss": 0.0167, "step": 78590 }, { "grad_norm": 0.2244405597448349, "learning_rate": 8.382382131795764e-08, "loss": 0.0179, "step": 78600 }, { "grad_norm": 0.22487731277942657, "learning_rate": 8.263179204496818e-08, "loss": 0.0187, "step": 78610 }, { "grad_norm": 0.17760278284549713, "learning_rate": 8.144829228496354e-08, "loss": 0.0154, "step": 78620 }, { "grad_norm": 0.3880169093608856, "learning_rate": 8.027332224016526e-08, "loss": 0.0181, "step": 78630 }, { "grad_norm": 0.18196611106395721, "learning_rate": 7.910688211135164e-08, "loss": 0.0207, "step": 78640 }, { "grad_norm": 0.4164658486843109, "learning_rate": 7.794897209783546e-08, "loss": 0.0211, "step": 78650 }, { "grad_norm": 0.2320578247308731, "learning_rate": 7.679959239746403e-08, "loss": 0.0193, "step": 78660 }, { "grad_norm": 0.2394530475139618, "learning_rate": 7.565874320664135e-08, "loss": 0.0206, "step": 78670 }, { "grad_norm": 0.31162381172180176, "learning_rate": 7.45264247203059e-08, "loss": 0.0182, "step": 78680 }, { "grad_norm": 0.3754615783691406, "learning_rate": 7.340263713194184e-08, "loss": 0.0176, "step": 78690 }, { "grad_norm": 0.30657562613487244, "learning_rate": 7.228738063356777e-08, "loss": 0.013, "step": 78700 }, { "grad_norm": 0.29346248507499695, "learning_rate": 7.118065541575903e-08, "loss": 0.0207, "step": 78710 }, { "grad_norm": 0.2665378749370575, "learning_rate": 7.008246166761435e-08, "loss": 0.0111, "step": 78720 }, { "grad_norm": 0.19123773276805878, "learning_rate": 6.899279957679472e-08, "loss": 0.0171, "step": 78730 }, { "grad_norm": 0.3338286876678467, "learning_rate": 6.791166932949011e-08, "loss": 0.0192, "step": 78740 }, { "grad_norm": 0.3072771430015564, "learning_rate": 6.683907111043608e-08, "loss": 0.0195, "step": 78750 }, { "grad_norm": 0.15273192524909973, "learning_rate": 6.577500510290824e-08, "loss": 0.0168, "step": 78760 }, { "grad_norm": 0.19269077479839325, "learning_rate": 6.47194714887278e-08, "loss": 0.0309, "step": 78770 }, { "grad_norm": 0.2339017689228058, "learning_rate": 6.367247044825608e-08, "loss": 0.014, "step": 78780 }, { "grad_norm": 0.24542339146137238, "learning_rate": 6.263400216039994e-08, "loss": 0.0157, "step": 78790 }, { "grad_norm": 0.21860867738723755, "learning_rate": 6.160406680260078e-08, "loss": 0.025, "step": 78800 }, { "grad_norm": 0.49997037649154663, "learning_rate": 6.058266455084561e-08, "loss": 0.0151, "step": 78810 }, { "grad_norm": 0.24904514849185944, "learning_rate": 5.956979557967257e-08, "loss": 0.015, "step": 78820 }, { "grad_norm": 0.39930203557014465, "learning_rate": 5.856546006214325e-08, "loss": 0.0186, "step": 78830 }, { "grad_norm": 0.1470295786857605, "learning_rate": 5.7569658169881466e-08, "loss": 0.0138, "step": 78840 }, { "grad_norm": 0.15627369284629822, "learning_rate": 5.658239007303445e-08, "loss": 0.0199, "step": 78850 }, { "grad_norm": 0.3375135064125061, "learning_rate": 5.560365594030059e-08, "loss": 0.0225, "step": 78860 }, { "grad_norm": 1.114258885383606, "learning_rate": 5.463345593891833e-08, "loss": 0.0173, "step": 78870 }, { "grad_norm": 0.22833997011184692, "learning_rate": 5.367179023467173e-08, "loss": 0.0216, "step": 78880 }, { "grad_norm": 0.3505624532699585, "learning_rate": 5.271865899187378e-08, "loss": 0.0157, "step": 78890 }, { "grad_norm": 0.185491144657135, "learning_rate": 5.177406237340532e-08, "loss": 0.0254, "step": 78900 }, { "grad_norm": 0.19733121991157532, "learning_rate": 5.083800054065946e-08, "loss": 0.0153, "step": 78910 }, { "grad_norm": 0.41122400760650635, "learning_rate": 4.991047365358603e-08, "loss": 0.0164, "step": 78920 }, { "grad_norm": 0.49141478538513184, "learning_rate": 4.899148187067493e-08, "loss": 0.0242, "step": 78930 }, { "grad_norm": 0.49405837059020996, "learning_rate": 4.808102534895609e-08, "loss": 0.0159, "step": 78940 }, { "grad_norm": 0.26599934697151184, "learning_rate": 4.717910424400507e-08, "loss": 0.0174, "step": 78950 }, { "grad_norm": 0.41289055347442627, "learning_rate": 4.628571870993193e-08, "loss": 0.0179, "step": 78960 }, { "grad_norm": 0.28762853145599365, "learning_rate": 4.5400868899392324e-08, "loss": 0.0219, "step": 78970 }, { "grad_norm": 0.36505889892578125, "learning_rate": 4.4524554963576436e-08, "loss": 0.0188, "step": 78980 }, { "grad_norm": 0.41225749254226685, "learning_rate": 4.36567770522367e-08, "loss": 0.0185, "step": 78990 }, { "grad_norm": 0.34305310249328613, "learning_rate": 4.279753531364339e-08, "loss": 0.0145, "step": 79000 }, { "grad_norm": 0.41907742619514465, "learning_rate": 4.1946829894617955e-08, "loss": 0.0199, "step": 79010 }, { "grad_norm": 0.2102549970149994, "learning_rate": 4.11046609405219e-08, "loss": 0.0199, "step": 79020 }, { "grad_norm": 0.1997666358947754, "learning_rate": 4.027102859526233e-08, "loss": 0.0114, "step": 79030 }, { "grad_norm": 0.33445143699645996, "learning_rate": 3.9445933001280856e-08, "loss": 0.0192, "step": 79040 }, { "grad_norm": 0.2159121185541153, "learning_rate": 3.8629374299564704e-08, "loss": 0.02, "step": 79050 }, { "grad_norm": 0.2416372150182724, "learning_rate": 3.78213526296467e-08, "loss": 0.0178, "step": 79060 }, { "grad_norm": 0.2273857146501541, "learning_rate": 3.702186812958308e-08, "loss": 0.0161, "step": 79070 }, { "grad_norm": 0.8339418172836304, "learning_rate": 3.623092093599789e-08, "loss": 0.0215, "step": 79080 }, { "grad_norm": 0.3488524258136749, "learning_rate": 3.544851118403303e-08, "loss": 0.0195, "step": 79090 }, { "grad_norm": 0.5360185503959656, "learning_rate": 3.46746390073871e-08, "loss": 0.0219, "step": 79100 }, { "grad_norm": 0.31734320521354675, "learning_rate": 3.390930453829322e-08, "loss": 0.0244, "step": 79110 }, { "grad_norm": 1.1314830780029297, "learning_rate": 3.3152507907519005e-08, "loss": 0.0221, "step": 79120 }, { "grad_norm": 0.20165257155895233, "learning_rate": 3.240424924438323e-08, "loss": 0.0152, "step": 79130 }, { "grad_norm": 0.3518361449241638, "learning_rate": 3.1664528676750295e-08, "loss": 0.0206, "step": 79140 }, { "grad_norm": 0.29954105615615845, "learning_rate": 3.093334633100797e-08, "loss": 0.027, "step": 79150 }, { "grad_norm": 0.30506473779678345, "learning_rate": 3.021070233210077e-08, "loss": 0.0173, "step": 79160 }, { "grad_norm": 0.26628780364990234, "learning_rate": 2.9496596803507692e-08, "loss": 0.0189, "step": 79170 }, { "grad_norm": 0.27402159571647644, "learning_rate": 2.879102986725335e-08, "loss": 0.0197, "step": 79180 }, { "grad_norm": 0.2715528607368469, "learning_rate": 2.809400164389131e-08, "loss": 0.0155, "step": 79190 }, { "grad_norm": 0.2256147861480713, "learning_rate": 2.740551225253185e-08, "loss": 0.0186, "step": 79200 }, { "grad_norm": 0.23146165907382965, "learning_rate": 2.6725561810819754e-08, "loss": 0.023, "step": 79210 }, { "grad_norm": 0.3574049770832062, "learning_rate": 2.605415043493431e-08, "loss": 0.0145, "step": 79220 }, { "grad_norm": 0.28940311074256897, "learning_rate": 2.5391278239605965e-08, "loss": 0.0176, "step": 79230 }, { "grad_norm": 0.250046044588089, "learning_rate": 2.4736945338094118e-08, "loss": 0.0186, "step": 79240 }, { "grad_norm": 0.23148947954177856, "learning_rate": 2.4091151842214887e-08, "loss": 0.0167, "step": 79250 }, { "grad_norm": 0.2176656424999237, "learning_rate": 2.3453897862318884e-08, "loss": 0.0228, "step": 79260 }, { "grad_norm": 0.17532485723495483, "learning_rate": 2.2825183507285686e-08, "loss": 0.0195, "step": 79270 }, { "grad_norm": 0.6493067145347595, "learning_rate": 2.220500888455157e-08, "loss": 0.0178, "step": 79280 }, { "grad_norm": 0.28143978118896484, "learning_rate": 2.1593374100081777e-08, "loss": 0.0251, "step": 79290 }, { "grad_norm": 0.2682535946369171, "learning_rate": 2.09902792583927e-08, "loss": 0.024, "step": 79300 }, { "grad_norm": 0.2934476137161255, "learning_rate": 2.0395724462540788e-08, "loss": 0.0131, "step": 79310 }, { "grad_norm": 0.2568696141242981, "learning_rate": 1.9809709814111453e-08, "loss": 0.0143, "step": 79320 }, { "grad_norm": 0.25582432746887207, "learning_rate": 1.923223541324126e-08, "loss": 0.0201, "step": 79330 }, { "grad_norm": 0.2172917276620865, "learning_rate": 1.8663301358606833e-08, "loss": 0.0206, "step": 79340 }, { "grad_norm": 0.23036596179008484, "learning_rate": 1.8102907747419295e-08, "loss": 0.0221, "step": 79350 }, { "grad_norm": 0.30460065603256226, "learning_rate": 1.7551054675435385e-08, "loss": 0.0136, "step": 79360 }, { "grad_norm": 0.17887179553508759, "learning_rate": 1.7007742236957447e-08, "loss": 0.0202, "step": 79370 }, { "grad_norm": 0.34318268299102783, "learning_rate": 1.647297052481678e-08, "loss": 0.0194, "step": 79380 }, { "grad_norm": 0.2935844659805298, "learning_rate": 1.5946739630390285e-08, "loss": 0.0198, "step": 79390 }, { "grad_norm": 0.29499122500419617, "learning_rate": 1.5429049643606032e-08, "loss": 0.0213, "step": 79400 }, { "grad_norm": 0.23150794208049774, "learning_rate": 1.4919900652909935e-08, "loss": 0.0236, "step": 79410 }, { "grad_norm": 0.18963037431240082, "learning_rate": 1.4419292745310175e-08, "loss": 0.019, "step": 79420 }, { "grad_norm": 0.16272872686386108, "learning_rate": 1.3927226006343885e-08, "loss": 0.0132, "step": 79430 }, { "grad_norm": 0.36660251021385193, "learning_rate": 1.3443700520093805e-08, "loss": 0.0166, "step": 79440 }, { "grad_norm": 0.25052449107170105, "learning_rate": 1.296871636917718e-08, "loss": 0.0197, "step": 79450 }, { "grad_norm": 0.4758073091506958, "learning_rate": 1.2502273634762419e-08, "loss": 0.0159, "step": 79460 }, { "grad_norm": 0.3890957534313202, "learning_rate": 1.2044372396546876e-08, "loss": 0.0264, "step": 79470 }, { "grad_norm": 0.45131462812423706, "learning_rate": 1.1595012732773524e-08, "loss": 0.0187, "step": 79480 }, { "grad_norm": 0.4980248510837555, "learning_rate": 1.1154194720225386e-08, "loss": 0.019, "step": 79490 }, { "grad_norm": 0.294886976480484, "learning_rate": 1.0721918434231093e-08, "loss": 0.0157, "step": 79500 }, { "grad_norm": 0.30291295051574707, "learning_rate": 1.0298183948648232e-08, "loss": 0.0173, "step": 79510 }, { "grad_norm": 0.21014736592769623, "learning_rate": 9.88299133588555e-09, "loss": 0.0145, "step": 79520 }, { "grad_norm": 0.1672791987657547, "learning_rate": 9.476340666891847e-09, "loss": 0.0182, "step": 79530 }, { "grad_norm": 0.2494708001613617, "learning_rate": 9.078232011139332e-09, "loss": 0.0142, "step": 79540 }, { "grad_norm": 0.346238911151886, "learning_rate": 8.68866543666802e-09, "loss": 0.0177, "step": 79550 }, { "grad_norm": 0.2838231921195984, "learning_rate": 8.307641010035782e-09, "loss": 0.0149, "step": 79560 }, { "grad_norm": 0.20937983691692352, "learning_rate": 7.93515879635165e-09, "loss": 0.0156, "step": 79570 }, { "grad_norm": 0.21172820031642914, "learning_rate": 7.571218859264706e-09, "loss": 0.0196, "step": 79580 }, { "grad_norm": 0.16413016617298126, "learning_rate": 7.215821260958544e-09, "loss": 0.0171, "step": 79590 }, { "grad_norm": 0.21569330990314484, "learning_rate": 6.868966062162363e-09, "loss": 0.0197, "step": 79600 }, { "grad_norm": 0.13447652757167816, "learning_rate": 6.530653322145419e-09, "loss": 0.0152, "step": 79610 }, { "grad_norm": 0.1514381766319275, "learning_rate": 6.200883098717025e-09, "loss": 0.0115, "step": 79620 }, { "grad_norm": 0.35559171438217163, "learning_rate": 5.879655448226551e-09, "loss": 0.0224, "step": 79630 }, { "grad_norm": 0.19280554354190826, "learning_rate": 5.566970425557872e-09, "loss": 0.0156, "step": 79640 }, { "grad_norm": 0.3321743607521057, "learning_rate": 5.26282808414047e-09, "loss": 0.0156, "step": 79650 }, { "grad_norm": 0.27486053109169006, "learning_rate": 4.967228475949437e-09, "loss": 0.0196, "step": 79660 }, { "grad_norm": 0.26461657881736755, "learning_rate": 4.680171651494369e-09, "loss": 0.014, "step": 79670 }, { "grad_norm": 0.35353943705558777, "learning_rate": 4.40165765981937e-09, "loss": 0.0193, "step": 79680 }, { "grad_norm": 0.7870636582374573, "learning_rate": 4.131686548519698e-09, "loss": 0.0169, "step": 79690 }, { "grad_norm": 0.26792827248573303, "learning_rate": 3.8702583637251214e-09, "loss": 0.016, "step": 79700 }, { "grad_norm": 0.41061848402023315, "learning_rate": 3.617373150105463e-09, "loss": 0.0161, "step": 79710 }, { "grad_norm": 0.3165629506111145, "learning_rate": 3.3730309508706036e-09, "loss": 0.0168, "step": 79720 }, { "grad_norm": 0.3126426637172699, "learning_rate": 3.137231807781582e-09, "loss": 0.0143, "step": 79730 }, { "grad_norm": 0.24700015783309937, "learning_rate": 2.9099757611172894e-09, "loss": 0.0179, "step": 79740 }, { "grad_norm": 0.34407809376716614, "learning_rate": 2.6912628497133275e-09, "loss": 0.0183, "step": 79750 }, { "grad_norm": 0.23032665252685547, "learning_rate": 2.481093110945354e-09, "loss": 0.0249, "step": 79760 }, { "grad_norm": 0.3223065733909607, "learning_rate": 2.2794665807235327e-09, "loss": 0.0224, "step": 79770 }, { "grad_norm": 0.2625441551208496, "learning_rate": 2.0863832934980843e-09, "loss": 0.0183, "step": 79780 }, { "grad_norm": 0.23718659579753876, "learning_rate": 1.901843282264837e-09, "loss": 0.0183, "step": 79790 }, { "grad_norm": 0.24113619327545166, "learning_rate": 1.7258465785541245e-09, "loss": 0.0173, "step": 79800 }, { "grad_norm": 0.44358110427856445, "learning_rate": 1.5583932124418887e-09, "loss": 0.0163, "step": 79810 }, { "grad_norm": 0.27363264560699463, "learning_rate": 1.3994832125441282e-09, "loss": 0.0158, "step": 79820 }, { "grad_norm": 0.3336828649044037, "learning_rate": 1.2491166060057958e-09, "loss": 0.0165, "step": 79830 }, { "grad_norm": 0.23772159218788147, "learning_rate": 1.1072934185230032e-09, "loss": 0.0195, "step": 79840 }, { "grad_norm": 0.30676257610321045, "learning_rate": 9.740136743319195e-10, "loss": 0.0185, "step": 79850 }, { "grad_norm": 0.1828165501356125, "learning_rate": 8.492773962087696e-10, "loss": 0.0137, "step": 79860 }, { "grad_norm": 0.28355488181114197, "learning_rate": 7.330846054587337e-10, "loss": 0.0205, "step": 79870 }, { "grad_norm": 0.5700218081474304, "learning_rate": 6.254353219492526e-10, "loss": 0.0185, "step": 79880 }, { "grad_norm": 0.366496205329895, "learning_rate": 5.263295640600685e-10, "loss": 0.0151, "step": 79890 }, { "grad_norm": 0.5739659070968628, "learning_rate": 4.357673487387359e-10, "loss": 0.0161, "step": 79900 }, { "grad_norm": 0.38331523537635803, "learning_rate": 3.537486914506616e-10, "loss": 0.0172, "step": 79910 }, { "grad_norm": 0.31476444005966187, "learning_rate": 2.8027360621241117e-10, "loss": 0.0181, "step": 79920 }, { "grad_norm": 0.4061132073402405, "learning_rate": 2.153421055806071e-10, "loss": 0.0176, "step": 79930 }, { "grad_norm": 0.32733047008514404, "learning_rate": 1.5895420064637733e-10, "loss": 0.02, "step": 79940 }, { "grad_norm": 0.21202188730239868, "learning_rate": 1.1110990105200891e-10, "loss": 0.0145, "step": 79950 }, { "grad_norm": 0.15971660614013672, "learning_rate": 7.180921496874326e-11, "loss": 0.0187, "step": 79960 }, { "grad_norm": 0.33633652329444885, "learning_rate": 4.10521491134297e-11, "loss": 0.0143, "step": 79970 }, { "grad_norm": 0.2236708551645279, "learning_rate": 1.883870874297422e-11, "loss": 0.0176, "step": 79980 }, { "grad_norm": 0.29792940616607666, "learning_rate": 5.1688976432373584e-12, "loss": 0.0188, "step": 79990 }, { "grad_norm": 0.31837213039398193, "learning_rate": 4.271816234080461e-14, "loss": 0.0176, "step": 80000 } ], "logging_steps": 10, "max_steps": 80000, "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 }