{ "best_global_step": null, "best_metric": null, "best_model_checkpoint": null, "epoch": 1.0, "eval_steps": 500, "global_step": 100000, "is_hyper_param_search": false, "is_local_process_zero": true, "is_world_process_zero": true, "log_history": [ { "grad_norm": 0.7135694026947021, "learning_rate": 1.8e-07, "loss": 1.4126, "step": 10 }, { "grad_norm": 0.7490555047988892, "learning_rate": 3.8e-07, "loss": 1.4187, "step": 20 }, { "grad_norm": 1.2258727550506592, "learning_rate": 5.8e-07, "loss": 1.4095, "step": 30 }, { "grad_norm": 1.1767369508743286, "learning_rate": 7.8e-07, "loss": 1.3965, "step": 40 }, { "grad_norm": 1.4625039100646973, "learning_rate": 9.8e-07, "loss": 1.3783, "step": 50 }, { "grad_norm": 1.3737322092056274, "learning_rate": 1.18e-06, "loss": 1.313, "step": 60 }, { "grad_norm": 1.1101115942001343, "learning_rate": 1.3800000000000001e-06, "loss": 1.2839, "step": 70 }, { "grad_norm": 0.8307448029518127, "learning_rate": 1.5800000000000003e-06, "loss": 1.2522, "step": 80 }, { "grad_norm": 1.2965688705444336, "learning_rate": 1.7800000000000001e-06, "loss": 1.2205, "step": 90 }, { "grad_norm": 0.9033528566360474, "learning_rate": 1.98e-06, "loss": 1.2522, "step": 100 }, { "grad_norm": 1.0183218717575073, "learning_rate": 2.1800000000000003e-06, "loss": 1.256, "step": 110 }, { "grad_norm": 1.173508644104004, "learning_rate": 2.38e-06, "loss": 1.2322, "step": 120 }, { "grad_norm": 1.2035298347473145, "learning_rate": 2.5800000000000003e-06, "loss": 1.2155, "step": 130 }, { "grad_norm": 1.4349178075790405, "learning_rate": 2.78e-06, "loss": 1.2115, "step": 140 }, { "grad_norm": 1.8562787771224976, "learning_rate": 2.9800000000000003e-06, "loss": 1.1933, "step": 150 }, { "grad_norm": 2.615771770477295, "learning_rate": 3.1800000000000005e-06, "loss": 1.18, "step": 160 }, { "grad_norm": 1.6185033321380615, "learning_rate": 3.38e-06, "loss": 1.1331, "step": 170 }, { "grad_norm": 0.9788058400154114, "learning_rate": 3.58e-06, "loss": 1.1425, "step": 180 }, { "grad_norm": 4.847921848297119, "learning_rate": 3.7800000000000002e-06, "loss": 1.1192, "step": 190 }, { "grad_norm": 4.644264221191406, "learning_rate": 3.98e-06, "loss": 1.1043, "step": 200 }, { "grad_norm": 3.956115245819092, "learning_rate": 4.18e-06, "loss": 1.0925, "step": 210 }, { "grad_norm": 1.621024250984192, "learning_rate": 4.38e-06, "loss": 1.1069, "step": 220 }, { "grad_norm": 1.86665940284729, "learning_rate": 4.58e-06, "loss": 1.1135, "step": 230 }, { "grad_norm": 2.3148744106292725, "learning_rate": 4.780000000000001e-06, "loss": 1.0975, "step": 240 }, { "grad_norm": 1.2103192806243896, "learning_rate": 4.98e-06, "loss": 1.0863, "step": 250 }, { "grad_norm": 0.8899864554405212, "learning_rate": 5.18e-06, "loss": 1.0792, "step": 260 }, { "grad_norm": 8.153668403625488, "learning_rate": 5.38e-06, "loss": 1.0965, "step": 270 }, { "grad_norm": 1.5999597311019897, "learning_rate": 5.580000000000001e-06, "loss": 1.0748, "step": 280 }, { "grad_norm": 1.4676125049591064, "learning_rate": 5.78e-06, "loss": 1.0703, "step": 290 }, { "grad_norm": 1.125103235244751, "learning_rate": 5.98e-06, "loss": 1.068, "step": 300 }, { "grad_norm": 14.250452995300293, "learning_rate": 6.18e-06, "loss": 1.094, "step": 310 }, { "grad_norm": 2.7759547233581543, "learning_rate": 6.38e-06, "loss": 1.0883, "step": 320 }, { "grad_norm": 1.2951332330703735, "learning_rate": 6.58e-06, "loss": 1.0587, "step": 330 }, { "grad_norm": 2.054816484451294, "learning_rate": 6.78e-06, "loss": 1.0924, "step": 340 }, { "grad_norm": 1.689043641090393, "learning_rate": 6.98e-06, "loss": 1.0891, "step": 350 }, { "grad_norm": 2.6435325145721436, "learning_rate": 7.180000000000001e-06, "loss": 1.0817, "step": 360 }, { "grad_norm": 2.1370887756347656, "learning_rate": 7.3800000000000005e-06, "loss": 1.0718, "step": 370 }, { "grad_norm": 1.6449432373046875, "learning_rate": 7.580000000000001e-06, "loss": 1.0621, "step": 380 }, { "grad_norm": 2.0735859870910645, "learning_rate": 7.78e-06, "loss": 1.0321, "step": 390 }, { "grad_norm": 1.7150187492370605, "learning_rate": 7.98e-06, "loss": 1.022, "step": 400 }, { "grad_norm": 1.5348047018051147, "learning_rate": 8.18e-06, "loss": 1.006, "step": 410 }, { "grad_norm": 1.4560785293579102, "learning_rate": 8.380000000000001e-06, "loss": 0.9684, "step": 420 }, { "grad_norm": 2.1532111167907715, "learning_rate": 8.580000000000001e-06, "loss": 0.9642, "step": 430 }, { "grad_norm": 1.6117520332336426, "learning_rate": 8.78e-06, "loss": 0.9543, "step": 440 }, { "grad_norm": 2.0199620723724365, "learning_rate": 8.98e-06, "loss": 0.8902, "step": 450 }, { "grad_norm": 2.9788053035736084, "learning_rate": 9.180000000000002e-06, "loss": 0.8878, "step": 460 }, { "grad_norm": 1.9907711744308472, "learning_rate": 9.38e-06, "loss": 0.8469, "step": 470 }, { "grad_norm": 2.6271486282348633, "learning_rate": 9.58e-06, "loss": 0.8251, "step": 480 }, { "grad_norm": 2.9042158126831055, "learning_rate": 9.78e-06, "loss": 0.7842, "step": 490 }, { "grad_norm": 2.2392687797546387, "learning_rate": 9.980000000000001e-06, "loss": 0.728, "step": 500 }, { "grad_norm": 2.5498151779174805, "learning_rate": 1.018e-05, "loss": 0.7142, "step": 510 }, { "grad_norm": 2.216747522354126, "learning_rate": 1.038e-05, "loss": 0.6721, "step": 520 }, { "grad_norm": 4.161910057067871, "learning_rate": 1.058e-05, "loss": 0.6669, "step": 530 }, { "grad_norm": 2.008030414581299, "learning_rate": 1.0780000000000002e-05, "loss": 0.5958, "step": 540 }, { "grad_norm": 2.465325355529785, "learning_rate": 1.098e-05, "loss": 0.6191, "step": 550 }, { "grad_norm": 2.379769802093506, "learning_rate": 1.118e-05, "loss": 0.5782, "step": 560 }, { "grad_norm": 2.4178168773651123, "learning_rate": 1.1380000000000001e-05, "loss": 0.4883, "step": 570 }, { "grad_norm": 4.055697917938232, "learning_rate": 1.1580000000000001e-05, "loss": 0.505, "step": 580 }, { "grad_norm": 2.449341058731079, "learning_rate": 1.178e-05, "loss": 0.4783, "step": 590 }, { "grad_norm": 3.108570098876953, "learning_rate": 1.198e-05, "loss": 0.4488, "step": 600 }, { "grad_norm": 3.1312294006347656, "learning_rate": 1.2180000000000002e-05, "loss": 0.4205, "step": 610 }, { "grad_norm": 2.551774263381958, "learning_rate": 1.238e-05, "loss": 0.4133, "step": 620 }, { "grad_norm": 3.0258262157440186, "learning_rate": 1.258e-05, "loss": 0.3698, "step": 630 }, { "grad_norm": 3.0825018882751465, "learning_rate": 1.278e-05, "loss": 0.3482, "step": 640 }, { "grad_norm": 3.1992905139923096, "learning_rate": 1.2980000000000001e-05, "loss": 0.3152, "step": 650 }, { "grad_norm": 3.268580436706543, "learning_rate": 1.3180000000000001e-05, "loss": 0.3598, "step": 660 }, { "grad_norm": 2.254060983657837, "learning_rate": 1.338e-05, "loss": 0.2833, "step": 670 }, { "grad_norm": 4.519505500793457, "learning_rate": 1.358e-05, "loss": 0.3135, "step": 680 }, { "grad_norm": 2.5062620639801025, "learning_rate": 1.3780000000000002e-05, "loss": 0.28, "step": 690 }, { "grad_norm": 3.1997976303100586, "learning_rate": 1.3980000000000002e-05, "loss": 0.2993, "step": 700 }, { "grad_norm": 2.350235939025879, "learning_rate": 1.4180000000000001e-05, "loss": 0.2295, "step": 710 }, { "grad_norm": 3.538104295730591, "learning_rate": 1.4380000000000001e-05, "loss": 0.2513, "step": 720 }, { "grad_norm": 2.5094470977783203, "learning_rate": 1.4580000000000003e-05, "loss": 0.2474, "step": 730 }, { "grad_norm": 3.8917133808135986, "learning_rate": 1.4779999999999999e-05, "loss": 0.1935, "step": 740 }, { "grad_norm": 2.8398382663726807, "learning_rate": 1.4979999999999999e-05, "loss": 0.2434, "step": 750 }, { "grad_norm": 1.713679313659668, "learning_rate": 1.518e-05, "loss": 0.2232, "step": 760 }, { "grad_norm": 2.483713150024414, "learning_rate": 1.538e-05, "loss": 0.2153, "step": 770 }, { "grad_norm": 2.295012950897217, "learning_rate": 1.558e-05, "loss": 0.2598, "step": 780 }, { "grad_norm": 3.302203416824341, "learning_rate": 1.578e-05, "loss": 0.1976, "step": 790 }, { "grad_norm": 3.127641439437866, "learning_rate": 1.598e-05, "loss": 0.2043, "step": 800 }, { "grad_norm": 1.9548513889312744, "learning_rate": 1.618e-05, "loss": 0.1993, "step": 810 }, { "grad_norm": 3.1830902099609375, "learning_rate": 1.6380000000000002e-05, "loss": 0.2102, "step": 820 }, { "grad_norm": 2.212949514389038, "learning_rate": 1.658e-05, "loss": 0.2208, "step": 830 }, { "grad_norm": 1.9943091869354248, "learning_rate": 1.6780000000000002e-05, "loss": 0.2101, "step": 840 }, { "grad_norm": 2.3580546379089355, "learning_rate": 1.698e-05, "loss": 0.1872, "step": 850 }, { "grad_norm": 2.412992000579834, "learning_rate": 1.718e-05, "loss": 0.1839, "step": 860 }, { "grad_norm": 2.2870121002197266, "learning_rate": 1.7380000000000003e-05, "loss": 0.2102, "step": 870 }, { "grad_norm": 2.4377708435058594, "learning_rate": 1.758e-05, "loss": 0.1787, "step": 880 }, { "grad_norm": 2.5455098152160645, "learning_rate": 1.7780000000000003e-05, "loss": 0.1954, "step": 890 }, { "grad_norm": 2.7864370346069336, "learning_rate": 1.798e-05, "loss": 0.2091, "step": 900 }, { "grad_norm": 2.2925477027893066, "learning_rate": 1.818e-05, "loss": 0.171, "step": 910 }, { "grad_norm": 1.7930569648742676, "learning_rate": 1.838e-05, "loss": 0.1557, "step": 920 }, { "grad_norm": 1.8293399810791016, "learning_rate": 1.858e-05, "loss": 0.1687, "step": 930 }, { "grad_norm": 1.899845838546753, "learning_rate": 1.878e-05, "loss": 0.1878, "step": 940 }, { "grad_norm": 1.7196276187896729, "learning_rate": 1.898e-05, "loss": 0.1705, "step": 950 }, { "grad_norm": 2.2073793411254883, "learning_rate": 1.918e-05, "loss": 0.1654, "step": 960 }, { "grad_norm": 2.288602590560913, "learning_rate": 1.938e-05, "loss": 0.2008, "step": 970 }, { "grad_norm": 1.6840791702270508, "learning_rate": 1.9580000000000002e-05, "loss": 0.1663, "step": 980 }, { "grad_norm": 1.6462293863296509, "learning_rate": 1.978e-05, "loss": 0.1446, "step": 990 }, { "grad_norm": 1.9386152029037476, "learning_rate": 1.9980000000000002e-05, "loss": 0.151, "step": 1000 }, { "grad_norm": 1.3349230289459229, "learning_rate": 2.0180000000000003e-05, "loss": 0.1456, "step": 1010 }, { "grad_norm": 1.9599257707595825, "learning_rate": 2.038e-05, "loss": 0.1854, "step": 1020 }, { "grad_norm": 1.6491239070892334, "learning_rate": 2.0580000000000003e-05, "loss": 0.1319, "step": 1030 }, { "grad_norm": 1.44796884059906, "learning_rate": 2.078e-05, "loss": 0.1352, "step": 1040 }, { "grad_norm": 2.0833323001861572, "learning_rate": 2.098e-05, "loss": 0.1701, "step": 1050 }, { "grad_norm": 2.375728130340576, "learning_rate": 2.118e-05, "loss": 0.1622, "step": 1060 }, { "grad_norm": 2.6484904289245605, "learning_rate": 2.138e-05, "loss": 0.1696, "step": 1070 }, { "grad_norm": 1.7132600545883179, "learning_rate": 2.158e-05, "loss": 0.1826, "step": 1080 }, { "grad_norm": 1.6837999820709229, "learning_rate": 2.178e-05, "loss": 0.1504, "step": 1090 }, { "grad_norm": 1.5824546813964844, "learning_rate": 2.198e-05, "loss": 0.1411, "step": 1100 }, { "grad_norm": 1.8623790740966797, "learning_rate": 2.218e-05, "loss": 0.1807, "step": 1110 }, { "grad_norm": 1.6623926162719727, "learning_rate": 2.2380000000000003e-05, "loss": 0.1273, "step": 1120 }, { "grad_norm": 2.3538784980773926, "learning_rate": 2.258e-05, "loss": 0.1556, "step": 1130 }, { "grad_norm": 2.268850088119507, "learning_rate": 2.2780000000000002e-05, "loss": 0.1858, "step": 1140 }, { "grad_norm": 2.3277015686035156, "learning_rate": 2.298e-05, "loss": 0.1787, "step": 1150 }, { "grad_norm": 1.9763201475143433, "learning_rate": 2.318e-05, "loss": 0.1521, "step": 1160 }, { "grad_norm": 1.4203376770019531, "learning_rate": 2.3380000000000003e-05, "loss": 0.1371, "step": 1170 }, { "grad_norm": 1.5801386833190918, "learning_rate": 2.358e-05, "loss": 0.1281, "step": 1180 }, { "grad_norm": 1.4023513793945312, "learning_rate": 2.3780000000000003e-05, "loss": 0.1348, "step": 1190 }, { "grad_norm": 2.7823331356048584, "learning_rate": 2.398e-05, "loss": 0.1459, "step": 1200 }, { "grad_norm": 2.4884612560272217, "learning_rate": 2.418e-05, "loss": 0.1733, "step": 1210 }, { "grad_norm": 1.5038458108901978, "learning_rate": 2.438e-05, "loss": 0.1449, "step": 1220 }, { "grad_norm": 2.4974076747894287, "learning_rate": 2.4580000000000002e-05, "loss": 0.1539, "step": 1230 }, { "grad_norm": 1.8610228300094604, "learning_rate": 2.478e-05, "loss": 0.1523, "step": 1240 }, { "grad_norm": 2.1034810543060303, "learning_rate": 2.498e-05, "loss": 0.1351, "step": 1250 }, { "grad_norm": 2.318661689758301, "learning_rate": 2.5180000000000003e-05, "loss": 0.1449, "step": 1260 }, { "grad_norm": 1.534422755241394, "learning_rate": 2.5380000000000004e-05, "loss": 0.1283, "step": 1270 }, { "grad_norm": 1.4347800016403198, "learning_rate": 2.5580000000000002e-05, "loss": 0.1421, "step": 1280 }, { "grad_norm": 1.8132617473602295, "learning_rate": 2.5779999999999997e-05, "loss": 0.1485, "step": 1290 }, { "grad_norm": 1.9587937593460083, "learning_rate": 2.598e-05, "loss": 0.1373, "step": 1300 }, { "grad_norm": 1.1723653078079224, "learning_rate": 2.618e-05, "loss": 0.1263, "step": 1310 }, { "grad_norm": 1.2906330823898315, "learning_rate": 2.6379999999999998e-05, "loss": 0.1273, "step": 1320 }, { "grad_norm": 1.540407657623291, "learning_rate": 2.658e-05, "loss": 0.1094, "step": 1330 }, { "grad_norm": 1.2812793254852295, "learning_rate": 2.678e-05, "loss": 0.1275, "step": 1340 }, { "grad_norm": 1.4135925769805908, "learning_rate": 2.698e-05, "loss": 0.1592, "step": 1350 }, { "grad_norm": 2.454845666885376, "learning_rate": 2.718e-05, "loss": 0.1483, "step": 1360 }, { "grad_norm": 1.3855727910995483, "learning_rate": 2.738e-05, "loss": 0.1256, "step": 1370 }, { "grad_norm": 1.392602562904358, "learning_rate": 2.758e-05, "loss": 0.1211, "step": 1380 }, { "grad_norm": 1.492092251777649, "learning_rate": 2.778e-05, "loss": 0.1271, "step": 1390 }, { "grad_norm": 4.455747127532959, "learning_rate": 2.798e-05, "loss": 0.1196, "step": 1400 }, { "grad_norm": 1.8190170526504517, "learning_rate": 2.818e-05, "loss": 0.131, "step": 1410 }, { "grad_norm": 2.3321728706359863, "learning_rate": 2.8380000000000003e-05, "loss": 0.1376, "step": 1420 }, { "grad_norm": 1.6330598592758179, "learning_rate": 2.858e-05, "loss": 0.111, "step": 1430 }, { "grad_norm": 1.4097862243652344, "learning_rate": 2.8780000000000002e-05, "loss": 0.1313, "step": 1440 }, { "grad_norm": 1.1436898708343506, "learning_rate": 2.898e-05, "loss": 0.1424, "step": 1450 }, { "grad_norm": 1.3706566095352173, "learning_rate": 2.9180000000000002e-05, "loss": 0.1371, "step": 1460 }, { "grad_norm": 1.2787177562713623, "learning_rate": 2.9380000000000003e-05, "loss": 0.1442, "step": 1470 }, { "grad_norm": 1.596131682395935, "learning_rate": 2.958e-05, "loss": 0.1356, "step": 1480 }, { "grad_norm": 1.7032685279846191, "learning_rate": 2.9780000000000003e-05, "loss": 0.1299, "step": 1490 }, { "grad_norm": 1.7498518228530884, "learning_rate": 2.998e-05, "loss": 0.1512, "step": 1500 }, { "grad_norm": 1.028283715248108, "learning_rate": 3.0180000000000002e-05, "loss": 0.1231, "step": 1510 }, { "grad_norm": 1.5455892086029053, "learning_rate": 3.0380000000000004e-05, "loss": 0.1374, "step": 1520 }, { "grad_norm": 1.171690821647644, "learning_rate": 3.058e-05, "loss": 0.1521, "step": 1530 }, { "grad_norm": 1.364059329032898, "learning_rate": 3.078e-05, "loss": 0.1313, "step": 1540 }, { "grad_norm": 1.6391243934631348, "learning_rate": 3.0980000000000005e-05, "loss": 0.1243, "step": 1550 }, { "grad_norm": 1.0896271467208862, "learning_rate": 3.118e-05, "loss": 0.1481, "step": 1560 }, { "grad_norm": 1.2053844928741455, "learning_rate": 3.138e-05, "loss": 0.1142, "step": 1570 }, { "grad_norm": 1.428505301475525, "learning_rate": 3.1580000000000006e-05, "loss": 0.1172, "step": 1580 }, { "grad_norm": 1.7531874179840088, "learning_rate": 3.1780000000000004e-05, "loss": 0.1358, "step": 1590 }, { "grad_norm": 1.6201013326644897, "learning_rate": 3.198e-05, "loss": 0.1248, "step": 1600 }, { "grad_norm": 1.2440476417541504, "learning_rate": 3.218e-05, "loss": 0.1013, "step": 1610 }, { "grad_norm": 1.3853747844696045, "learning_rate": 3.238e-05, "loss": 0.1297, "step": 1620 }, { "grad_norm": 1.1818870306015015, "learning_rate": 3.2579999999999996e-05, "loss": 0.1495, "step": 1630 }, { "grad_norm": 1.2643859386444092, "learning_rate": 3.278e-05, "loss": 0.1349, "step": 1640 }, { "grad_norm": 1.4103283882141113, "learning_rate": 3.298e-05, "loss": 0.1315, "step": 1650 }, { "grad_norm": 0.873501718044281, "learning_rate": 3.318e-05, "loss": 0.1228, "step": 1660 }, { "grad_norm": 1.1958553791046143, "learning_rate": 3.338e-05, "loss": 0.1025, "step": 1670 }, { "grad_norm": 1.428358793258667, "learning_rate": 3.358e-05, "loss": 0.1286, "step": 1680 }, { "grad_norm": 1.0675383806228638, "learning_rate": 3.378e-05, "loss": 0.1256, "step": 1690 }, { "grad_norm": 0.9069299697875977, "learning_rate": 3.398e-05, "loss": 0.1281, "step": 1700 }, { "grad_norm": 1.3612985610961914, "learning_rate": 3.418e-05, "loss": 0.1539, "step": 1710 }, { "grad_norm": 1.237984538078308, "learning_rate": 3.438e-05, "loss": 0.132, "step": 1720 }, { "grad_norm": 1.4954118728637695, "learning_rate": 3.4580000000000004e-05, "loss": 0.1282, "step": 1730 }, { "grad_norm": 1.1870088577270508, "learning_rate": 3.478e-05, "loss": 0.1202, "step": 1740 }, { "grad_norm": 1.065847396850586, "learning_rate": 3.498e-05, "loss": 0.1307, "step": 1750 }, { "grad_norm": 1.2188146114349365, "learning_rate": 3.518e-05, "loss": 0.1164, "step": 1760 }, { "grad_norm": 1.1863892078399658, "learning_rate": 3.5380000000000003e-05, "loss": 0.1159, "step": 1770 }, { "grad_norm": 1.245065689086914, "learning_rate": 3.558e-05, "loss": 0.1387, "step": 1780 }, { "grad_norm": 1.4569973945617676, "learning_rate": 3.578e-05, "loss": 0.1209, "step": 1790 }, { "grad_norm": 0.8841944932937622, "learning_rate": 3.5980000000000004e-05, "loss": 0.1193, "step": 1800 }, { "grad_norm": 1.2130645513534546, "learning_rate": 3.618e-05, "loss": 0.1078, "step": 1810 }, { "grad_norm": 1.3286411762237549, "learning_rate": 3.638e-05, "loss": 0.1214, "step": 1820 }, { "grad_norm": 1.8060963153839111, "learning_rate": 3.6580000000000006e-05, "loss": 0.1214, "step": 1830 }, { "grad_norm": 1.3449182510375977, "learning_rate": 3.6780000000000004e-05, "loss": 0.1355, "step": 1840 }, { "grad_norm": 1.1445468664169312, "learning_rate": 3.698e-05, "loss": 0.1546, "step": 1850 }, { "grad_norm": 1.3300421237945557, "learning_rate": 3.7180000000000007e-05, "loss": 0.1435, "step": 1860 }, { "grad_norm": 1.186694860458374, "learning_rate": 3.7380000000000005e-05, "loss": 0.1361, "step": 1870 }, { "grad_norm": 1.050958514213562, "learning_rate": 3.758e-05, "loss": 0.1088, "step": 1880 }, { "grad_norm": 1.1559392213821411, "learning_rate": 3.778000000000001e-05, "loss": 0.1079, "step": 1890 }, { "grad_norm": 0.8240066170692444, "learning_rate": 3.7980000000000006e-05, "loss": 0.113, "step": 1900 }, { "grad_norm": 1.4721447229385376, "learning_rate": 3.818e-05, "loss": 0.1358, "step": 1910 }, { "grad_norm": 0.7367483973503113, "learning_rate": 3.838e-05, "loss": 0.11, "step": 1920 }, { "grad_norm": 1.1847565174102783, "learning_rate": 3.858e-05, "loss": 0.0947, "step": 1930 }, { "grad_norm": 1.200070858001709, "learning_rate": 3.878e-05, "loss": 0.1242, "step": 1940 }, { "grad_norm": 0.9848204255104065, "learning_rate": 3.898e-05, "loss": 0.117, "step": 1950 }, { "grad_norm": 1.6384382247924805, "learning_rate": 3.918e-05, "loss": 0.1415, "step": 1960 }, { "grad_norm": 1.1129554510116577, "learning_rate": 3.938e-05, "loss": 0.1224, "step": 1970 }, { "grad_norm": 1.2044916152954102, "learning_rate": 3.958e-05, "loss": 0.1216, "step": 1980 }, { "grad_norm": 1.077362060546875, "learning_rate": 3.978e-05, "loss": 0.1012, "step": 1990 }, { "grad_norm": 1.057317852973938, "learning_rate": 3.998e-05, "loss": 0.1225, "step": 2000 }, { "grad_norm": 1.0900568962097168, "learning_rate": 4.018e-05, "loss": 0.1529, "step": 2010 }, { "grad_norm": 1.3373651504516602, "learning_rate": 4.038e-05, "loss": 0.1239, "step": 2020 }, { "grad_norm": 1.1174768209457397, "learning_rate": 4.058e-05, "loss": 0.1067, "step": 2030 }, { "grad_norm": 0.9141257405281067, "learning_rate": 4.078e-05, "loss": 0.1269, "step": 2040 }, { "grad_norm": 0.9103925228118896, "learning_rate": 4.0980000000000004e-05, "loss": 0.1224, "step": 2050 }, { "grad_norm": 1.6959460973739624, "learning_rate": 4.118e-05, "loss": 0.1175, "step": 2060 }, { "grad_norm": 1.0457115173339844, "learning_rate": 4.138e-05, "loss": 0.1162, "step": 2070 }, { "grad_norm": 0.9177454113960266, "learning_rate": 4.1580000000000005e-05, "loss": 0.1135, "step": 2080 }, { "grad_norm": 0.8668966889381409, "learning_rate": 4.178e-05, "loss": 0.1118, "step": 2090 }, { "grad_norm": 0.9004907608032227, "learning_rate": 4.198e-05, "loss": 0.1031, "step": 2100 }, { "grad_norm": 1.116608738899231, "learning_rate": 4.2180000000000006e-05, "loss": 0.1306, "step": 2110 }, { "grad_norm": 1.3128107786178589, "learning_rate": 4.2380000000000004e-05, "loss": 0.123, "step": 2120 }, { "grad_norm": 1.182716727256775, "learning_rate": 4.258e-05, "loss": 0.1099, "step": 2130 }, { "grad_norm": 1.0242983102798462, "learning_rate": 4.278e-05, "loss": 0.0898, "step": 2140 }, { "grad_norm": 1.1042135953903198, "learning_rate": 4.2980000000000005e-05, "loss": 0.1162, "step": 2150 }, { "grad_norm": 1.1468818187713623, "learning_rate": 4.318e-05, "loss": 0.1196, "step": 2160 }, { "grad_norm": 0.7834706902503967, "learning_rate": 4.338e-05, "loss": 0.1084, "step": 2170 }, { "grad_norm": 0.9882423281669617, "learning_rate": 4.3580000000000006e-05, "loss": 0.1331, "step": 2180 }, { "grad_norm": 1.1989305019378662, "learning_rate": 4.3780000000000004e-05, "loss": 0.1117, "step": 2190 }, { "grad_norm": 1.002230167388916, "learning_rate": 4.398e-05, "loss": 0.1177, "step": 2200 }, { "grad_norm": 0.7538912296295166, "learning_rate": 4.418000000000001e-05, "loss": 0.1162, "step": 2210 }, { "grad_norm": 0.9868258237838745, "learning_rate": 4.438e-05, "loss": 0.1089, "step": 2220 }, { "grad_norm": 0.9078338146209717, "learning_rate": 4.458e-05, "loss": 0.1015, "step": 2230 }, { "grad_norm": 0.8649982213973999, "learning_rate": 4.478e-05, "loss": 0.1128, "step": 2240 }, { "grad_norm": 0.8306825757026672, "learning_rate": 4.498e-05, "loss": 0.1182, "step": 2250 }, { "grad_norm": 1.1022247076034546, "learning_rate": 4.518e-05, "loss": 0.1044, "step": 2260 }, { "grad_norm": 0.6999926567077637, "learning_rate": 4.538e-05, "loss": 0.1116, "step": 2270 }, { "grad_norm": 0.8462041020393372, "learning_rate": 4.558e-05, "loss": 0.1034, "step": 2280 }, { "grad_norm": 1.3252503871917725, "learning_rate": 4.578e-05, "loss": 0.1287, "step": 2290 }, { "grad_norm": 1.6194137334823608, "learning_rate": 4.5980000000000004e-05, "loss": 0.1018, "step": 2300 }, { "grad_norm": 1.051092505455017, "learning_rate": 4.618e-05, "loss": 0.1082, "step": 2310 }, { "grad_norm": 1.1975361108779907, "learning_rate": 4.638e-05, "loss": 0.0948, "step": 2320 }, { "grad_norm": 0.8491308689117432, "learning_rate": 4.6580000000000005e-05, "loss": 0.1011, "step": 2330 }, { "grad_norm": 0.8856850266456604, "learning_rate": 4.678e-05, "loss": 0.1178, "step": 2340 }, { "grad_norm": 0.7360876798629761, "learning_rate": 4.698e-05, "loss": 0.1094, "step": 2350 }, { "grad_norm": 0.9410296678543091, "learning_rate": 4.718e-05, "loss": 0.11, "step": 2360 }, { "grad_norm": 0.8211159110069275, "learning_rate": 4.7380000000000004e-05, "loss": 0.0966, "step": 2370 }, { "grad_norm": 0.9037021398544312, "learning_rate": 4.758e-05, "loss": 0.1171, "step": 2380 }, { "grad_norm": 0.746242880821228, "learning_rate": 4.778e-05, "loss": 0.1152, "step": 2390 }, { "grad_norm": 0.8676548004150391, "learning_rate": 4.7980000000000005e-05, "loss": 0.1105, "step": 2400 }, { "grad_norm": 0.8953896760940552, "learning_rate": 4.818e-05, "loss": 0.109, "step": 2410 }, { "grad_norm": 0.8831972479820251, "learning_rate": 4.838e-05, "loss": 0.1494, "step": 2420 }, { "grad_norm": 1.0545687675476074, "learning_rate": 4.8580000000000006e-05, "loss": 0.1069, "step": 2430 }, { "grad_norm": 1.3093780279159546, "learning_rate": 4.8780000000000004e-05, "loss": 0.1379, "step": 2440 }, { "grad_norm": 0.950481653213501, "learning_rate": 4.898e-05, "loss": 0.099, "step": 2450 }, { "grad_norm": 1.0052409172058105, "learning_rate": 4.918000000000001e-05, "loss": 0.1093, "step": 2460 }, { "grad_norm": 0.8567463755607605, "learning_rate": 4.9380000000000005e-05, "loss": 0.103, "step": 2470 }, { "grad_norm": 0.8036132454872131, "learning_rate": 4.958e-05, "loss": 0.1297, "step": 2480 }, { "grad_norm": 0.7557700872421265, "learning_rate": 4.978e-05, "loss": 0.0915, "step": 2490 }, { "grad_norm": 0.7311651706695557, "learning_rate": 4.9980000000000006e-05, "loss": 0.0979, "step": 2500 }, { "grad_norm": 0.7151845097541809, "learning_rate": 5.0180000000000004e-05, "loss": 0.0903, "step": 2510 }, { "grad_norm": 0.6662042737007141, "learning_rate": 5.038e-05, "loss": 0.1125, "step": 2520 }, { "grad_norm": 1.067001461982727, "learning_rate": 5.058000000000001e-05, "loss": 0.0915, "step": 2530 }, { "grad_norm": 0.7130573391914368, "learning_rate": 5.0780000000000005e-05, "loss": 0.1021, "step": 2540 }, { "grad_norm": 0.7981851100921631, "learning_rate": 5.098e-05, "loss": 0.1086, "step": 2550 }, { "grad_norm": 0.7445557117462158, "learning_rate": 5.118000000000001e-05, "loss": 0.1006, "step": 2560 }, { "grad_norm": 0.6987658739089966, "learning_rate": 5.1380000000000006e-05, "loss": 0.1107, "step": 2570 }, { "grad_norm": 0.9254143238067627, "learning_rate": 5.1580000000000004e-05, "loss": 0.1109, "step": 2580 }, { "grad_norm": 0.6869136691093445, "learning_rate": 5.178000000000001e-05, "loss": 0.1007, "step": 2590 }, { "grad_norm": 0.696145236492157, "learning_rate": 5.198000000000001e-05, "loss": 0.1261, "step": 2600 }, { "grad_norm": 0.8219096660614014, "learning_rate": 5.2180000000000005e-05, "loss": 0.1156, "step": 2610 }, { "grad_norm": 0.8296982645988464, "learning_rate": 5.238000000000001e-05, "loss": 0.0945, "step": 2620 }, { "grad_norm": 0.7050186395645142, "learning_rate": 5.258000000000001e-05, "loss": 0.1085, "step": 2630 }, { "grad_norm": 0.6537378430366516, "learning_rate": 5.2780000000000006e-05, "loss": 0.1104, "step": 2640 }, { "grad_norm": 0.7944967746734619, "learning_rate": 5.2980000000000004e-05, "loss": 0.1233, "step": 2650 }, { "grad_norm": 1.1838456392288208, "learning_rate": 5.318000000000001e-05, "loss": 0.1208, "step": 2660 }, { "grad_norm": 0.9060463309288025, "learning_rate": 5.338000000000001e-05, "loss": 0.1052, "step": 2670 }, { "grad_norm": 1.1154558658599854, "learning_rate": 5.3580000000000005e-05, "loss": 0.1398, "step": 2680 }, { "grad_norm": 0.6107066869735718, "learning_rate": 5.378e-05, "loss": 0.1076, "step": 2690 }, { "grad_norm": 0.9505775570869446, "learning_rate": 5.3979999999999995e-05, "loss": 0.1076, "step": 2700 }, { "grad_norm": 0.6918302178382874, "learning_rate": 5.418e-05, "loss": 0.1138, "step": 2710 }, { "grad_norm": 0.7684422731399536, "learning_rate": 5.438e-05, "loss": 0.1248, "step": 2720 }, { "grad_norm": 1.107714056968689, "learning_rate": 5.4579999999999996e-05, "loss": 0.1109, "step": 2730 }, { "grad_norm": 0.7014469504356384, "learning_rate": 5.478e-05, "loss": 0.0881, "step": 2740 }, { "grad_norm": 0.7320536971092224, "learning_rate": 5.498e-05, "loss": 0.1069, "step": 2750 }, { "grad_norm": 0.9067598581314087, "learning_rate": 5.518e-05, "loss": 0.0843, "step": 2760 }, { "grad_norm": 1.0791234970092773, "learning_rate": 5.538e-05, "loss": 0.0912, "step": 2770 }, { "grad_norm": 0.662813663482666, "learning_rate": 5.558e-05, "loss": 0.1061, "step": 2780 }, { "grad_norm": 0.6703081727027893, "learning_rate": 5.578e-05, "loss": 0.1086, "step": 2790 }, { "grad_norm": 0.792402446269989, "learning_rate": 5.5979999999999996e-05, "loss": 0.0958, "step": 2800 }, { "grad_norm": 0.9399141073226929, "learning_rate": 5.618e-05, "loss": 0.1265, "step": 2810 }, { "grad_norm": 0.8354750871658325, "learning_rate": 5.638e-05, "loss": 0.1157, "step": 2820 }, { "grad_norm": 1.2728431224822998, "learning_rate": 5.658e-05, "loss": 0.0966, "step": 2830 }, { "grad_norm": 0.9039880633354187, "learning_rate": 5.678e-05, "loss": 0.1199, "step": 2840 }, { "grad_norm": 1.1136295795440674, "learning_rate": 5.698e-05, "loss": 0.1107, "step": 2850 }, { "grad_norm": 0.7336447834968567, "learning_rate": 5.718e-05, "loss": 0.0912, "step": 2860 }, { "grad_norm": 0.5776678323745728, "learning_rate": 5.738e-05, "loss": 0.0982, "step": 2870 }, { "grad_norm": 0.6034001708030701, "learning_rate": 5.758e-05, "loss": 0.0859, "step": 2880 }, { "grad_norm": 0.8625809550285339, "learning_rate": 5.778e-05, "loss": 0.1048, "step": 2890 }, { "grad_norm": 0.5045586228370667, "learning_rate": 5.7980000000000004e-05, "loss": 0.1019, "step": 2900 }, { "grad_norm": 0.8899149894714355, "learning_rate": 5.818e-05, "loss": 0.0997, "step": 2910 }, { "grad_norm": 1.0156158208847046, "learning_rate": 5.838e-05, "loss": 0.1131, "step": 2920 }, { "grad_norm": 0.7592231631278992, "learning_rate": 5.858e-05, "loss": 0.0986, "step": 2930 }, { "grad_norm": 0.8557561635971069, "learning_rate": 5.878e-05, "loss": 0.1059, "step": 2940 }, { "grad_norm": 1.0484845638275146, "learning_rate": 5.898e-05, "loss": 0.099, "step": 2950 }, { "grad_norm": 0.7089977860450745, "learning_rate": 5.918e-05, "loss": 0.1014, "step": 2960 }, { "grad_norm": 0.7150424122810364, "learning_rate": 5.9380000000000004e-05, "loss": 0.0806, "step": 2970 }, { "grad_norm": 0.794983446598053, "learning_rate": 5.958e-05, "loss": 0.099, "step": 2980 }, { "grad_norm": 0.7351647615432739, "learning_rate": 5.978e-05, "loss": 0.1121, "step": 2990 }, { "grad_norm": 1.1034300327301025, "learning_rate": 5.9980000000000005e-05, "loss": 0.1145, "step": 3000 }, { "grad_norm": 0.6453109979629517, "learning_rate": 6.018e-05, "loss": 0.0927, "step": 3010 }, { "grad_norm": 0.6729623079299927, "learning_rate": 6.038e-05, "loss": 0.0875, "step": 3020 }, { "grad_norm": 0.5860438346862793, "learning_rate": 6.0580000000000006e-05, "loss": 0.1015, "step": 3030 }, { "grad_norm": 0.7965766787528992, "learning_rate": 6.0780000000000004e-05, "loss": 0.0964, "step": 3040 }, { "grad_norm": 0.6567987203598022, "learning_rate": 6.098e-05, "loss": 0.1032, "step": 3050 }, { "grad_norm": 0.8836715221405029, "learning_rate": 6.118000000000001e-05, "loss": 0.1107, "step": 3060 }, { "grad_norm": 0.7244722843170166, "learning_rate": 6.138e-05, "loss": 0.093, "step": 3070 }, { "grad_norm": 0.7205469012260437, "learning_rate": 6.158e-05, "loss": 0.0943, "step": 3080 }, { "grad_norm": 1.0581443309783936, "learning_rate": 6.178000000000001e-05, "loss": 0.1267, "step": 3090 }, { "grad_norm": 0.7386137843132019, "learning_rate": 6.198e-05, "loss": 0.0849, "step": 3100 }, { "grad_norm": 0.8669150471687317, "learning_rate": 6.218e-05, "loss": 0.0881, "step": 3110 }, { "grad_norm": 0.8035977482795715, "learning_rate": 6.238000000000001e-05, "loss": 0.085, "step": 3120 }, { "grad_norm": 0.7959375381469727, "learning_rate": 6.258e-05, "loss": 0.0822, "step": 3130 }, { "grad_norm": 0.6513484716415405, "learning_rate": 6.278e-05, "loss": 0.0971, "step": 3140 }, { "grad_norm": 1.0763953924179077, "learning_rate": 6.298000000000001e-05, "loss": 0.1101, "step": 3150 }, { "grad_norm": 0.6855977177619934, "learning_rate": 6.318e-05, "loss": 0.1029, "step": 3160 }, { "grad_norm": 0.47879964113235474, "learning_rate": 6.338e-05, "loss": 0.0856, "step": 3170 }, { "grad_norm": 0.8671007752418518, "learning_rate": 6.358000000000001e-05, "loss": 0.1113, "step": 3180 }, { "grad_norm": 0.581312894821167, "learning_rate": 6.378e-05, "loss": 0.1074, "step": 3190 }, { "grad_norm": 0.684444785118103, "learning_rate": 6.398000000000001e-05, "loss": 0.1021, "step": 3200 }, { "grad_norm": 0.7713956236839294, "learning_rate": 6.418000000000001e-05, "loss": 0.091, "step": 3210 }, { "grad_norm": 1.1187372207641602, "learning_rate": 6.438e-05, "loss": 0.1202, "step": 3220 }, { "grad_norm": 0.6324926018714905, "learning_rate": 6.458000000000001e-05, "loss": 0.1022, "step": 3230 }, { "grad_norm": 0.9575175046920776, "learning_rate": 6.478000000000001e-05, "loss": 0.1018, "step": 3240 }, { "grad_norm": 0.586545467376709, "learning_rate": 6.498e-05, "loss": 0.0925, "step": 3250 }, { "grad_norm": 0.6788380742073059, "learning_rate": 6.518000000000001e-05, "loss": 0.0905, "step": 3260 }, { "grad_norm": 0.7310558557510376, "learning_rate": 6.538000000000001e-05, "loss": 0.0866, "step": 3270 }, { "grad_norm": 0.6013817191123962, "learning_rate": 6.558e-05, "loss": 0.0742, "step": 3280 }, { "grad_norm": 0.6581671237945557, "learning_rate": 6.578000000000001e-05, "loss": 0.0906, "step": 3290 }, { "grad_norm": 0.8179090023040771, "learning_rate": 6.598e-05, "loss": 0.0976, "step": 3300 }, { "grad_norm": 1.1396957635879517, "learning_rate": 6.618e-05, "loss": 0.1077, "step": 3310 }, { "grad_norm": 0.8733755946159363, "learning_rate": 6.638e-05, "loss": 0.1152, "step": 3320 }, { "grad_norm": 0.5678859353065491, "learning_rate": 6.658e-05, "loss": 0.0987, "step": 3330 }, { "grad_norm": 0.7155807018280029, "learning_rate": 6.678e-05, "loss": 0.0967, "step": 3340 }, { "grad_norm": 0.6751067042350769, "learning_rate": 6.698e-05, "loss": 0.0849, "step": 3350 }, { "grad_norm": 0.6738308668136597, "learning_rate": 6.718e-05, "loss": 0.1099, "step": 3360 }, { "grad_norm": 0.8349173069000244, "learning_rate": 6.738e-05, "loss": 0.0983, "step": 3370 }, { "grad_norm": 0.5953673720359802, "learning_rate": 6.758e-05, "loss": 0.1392, "step": 3380 }, { "grad_norm": 0.6354956030845642, "learning_rate": 6.778e-05, "loss": 0.0911, "step": 3390 }, { "grad_norm": 0.5907008647918701, "learning_rate": 6.798e-05, "loss": 0.0965, "step": 3400 }, { "grad_norm": 0.7270945310592651, "learning_rate": 6.818e-05, "loss": 0.0962, "step": 3410 }, { "grad_norm": 0.5979515910148621, "learning_rate": 6.838e-05, "loss": 0.0884, "step": 3420 }, { "grad_norm": 0.6168406009674072, "learning_rate": 6.858e-05, "loss": 0.0818, "step": 3430 }, { "grad_norm": 0.6034042239189148, "learning_rate": 6.878e-05, "loss": 0.0915, "step": 3440 }, { "grad_norm": 0.5479382276535034, "learning_rate": 6.898e-05, "loss": 0.0887, "step": 3450 }, { "grad_norm": 0.7155365347862244, "learning_rate": 6.918e-05, "loss": 0.0816, "step": 3460 }, { "grad_norm": 0.6591423749923706, "learning_rate": 6.938e-05, "loss": 0.0944, "step": 3470 }, { "grad_norm": 0.6967965364456177, "learning_rate": 6.958e-05, "loss": 0.1005, "step": 3480 }, { "grad_norm": 0.6126736402511597, "learning_rate": 6.978e-05, "loss": 0.0926, "step": 3490 }, { "grad_norm": 0.6977865099906921, "learning_rate": 6.998e-05, "loss": 0.0823, "step": 3500 }, { "grad_norm": 1.1090445518493652, "learning_rate": 7.018e-05, "loss": 0.1056, "step": 3510 }, { "grad_norm": 0.658547043800354, "learning_rate": 7.038e-05, "loss": 0.1026, "step": 3520 }, { "grad_norm": 0.5667654871940613, "learning_rate": 7.058e-05, "loss": 0.0921, "step": 3530 }, { "grad_norm": 0.6308541297912598, "learning_rate": 7.078e-05, "loss": 0.0903, "step": 3540 }, { "grad_norm": 0.5504917502403259, "learning_rate": 7.098e-05, "loss": 0.0777, "step": 3550 }, { "grad_norm": 0.9623975157737732, "learning_rate": 7.118e-05, "loss": 0.0834, "step": 3560 }, { "grad_norm": 0.6113068461418152, "learning_rate": 7.138e-05, "loss": 0.1065, "step": 3570 }, { "grad_norm": 0.6908184289932251, "learning_rate": 7.158e-05, "loss": 0.1039, "step": 3580 }, { "grad_norm": 0.8039638996124268, "learning_rate": 7.178000000000001e-05, "loss": 0.1154, "step": 3590 }, { "grad_norm": 0.7086897492408752, "learning_rate": 7.198e-05, "loss": 0.0979, "step": 3600 }, { "grad_norm": 0.5818172097206116, "learning_rate": 7.218e-05, "loss": 0.096, "step": 3610 }, { "grad_norm": 0.5694104433059692, "learning_rate": 7.238000000000001e-05, "loss": 0.09, "step": 3620 }, { "grad_norm": 0.5629916787147522, "learning_rate": 7.258e-05, "loss": 0.0861, "step": 3630 }, { "grad_norm": 0.5781622529029846, "learning_rate": 7.278e-05, "loss": 0.0831, "step": 3640 }, { "grad_norm": 0.6772428154945374, "learning_rate": 7.298000000000001e-05, "loss": 0.0996, "step": 3650 }, { "grad_norm": 0.589460551738739, "learning_rate": 7.318e-05, "loss": 0.0868, "step": 3660 }, { "grad_norm": 0.6162086725234985, "learning_rate": 7.338e-05, "loss": 0.1071, "step": 3670 }, { "grad_norm": 0.5484318137168884, "learning_rate": 7.358000000000001e-05, "loss": 0.087, "step": 3680 }, { "grad_norm": 0.6355458498001099, "learning_rate": 7.378e-05, "loss": 0.0744, "step": 3690 }, { "grad_norm": 0.5999330878257751, "learning_rate": 7.398e-05, "loss": 0.0878, "step": 3700 }, { "grad_norm": 0.6614571809768677, "learning_rate": 7.418000000000001e-05, "loss": 0.0864, "step": 3710 }, { "grad_norm": 0.6056673526763916, "learning_rate": 7.438e-05, "loss": 0.0867, "step": 3720 }, { "grad_norm": 0.6694501042366028, "learning_rate": 7.458000000000001e-05, "loss": 0.0782, "step": 3730 }, { "grad_norm": 0.8218725919723511, "learning_rate": 7.478e-05, "loss": 0.0887, "step": 3740 }, { "grad_norm": 0.7275111675262451, "learning_rate": 7.498e-05, "loss": 0.0952, "step": 3750 }, { "grad_norm": 0.5827742218971252, "learning_rate": 7.518000000000001e-05, "loss": 0.0822, "step": 3760 }, { "grad_norm": 0.6306923031806946, "learning_rate": 7.538e-05, "loss": 0.1015, "step": 3770 }, { "grad_norm": 1.0139321088790894, "learning_rate": 7.558e-05, "loss": 0.1207, "step": 3780 }, { "grad_norm": 0.6089711785316467, "learning_rate": 7.578000000000001e-05, "loss": 0.0961, "step": 3790 }, { "grad_norm": 0.4698846936225891, "learning_rate": 7.598e-05, "loss": 0.117, "step": 3800 }, { "grad_norm": 0.5377482175827026, "learning_rate": 7.618e-05, "loss": 0.0822, "step": 3810 }, { "grad_norm": 0.623175323009491, "learning_rate": 7.638000000000001e-05, "loss": 0.0896, "step": 3820 }, { "grad_norm": 1.1342462301254272, "learning_rate": 7.658e-05, "loss": 0.1027, "step": 3830 }, { "grad_norm": 0.661013662815094, "learning_rate": 7.678000000000001e-05, "loss": 0.1011, "step": 3840 }, { "grad_norm": 0.6606666445732117, "learning_rate": 7.698000000000001e-05, "loss": 0.1041, "step": 3850 }, { "grad_norm": 0.894855797290802, "learning_rate": 7.718e-05, "loss": 0.0993, "step": 3860 }, { "grad_norm": 0.6326399445533752, "learning_rate": 7.738000000000001e-05, "loss": 0.0915, "step": 3870 }, { "grad_norm": 0.7396665811538696, "learning_rate": 7.758000000000001e-05, "loss": 0.0945, "step": 3880 }, { "grad_norm": 0.6839178800582886, "learning_rate": 7.778e-05, "loss": 0.1008, "step": 3890 }, { "grad_norm": 1.1570183038711548, "learning_rate": 7.798000000000001e-05, "loss": 0.096, "step": 3900 }, { "grad_norm": 0.8008842468261719, "learning_rate": 7.818000000000001e-05, "loss": 0.1044, "step": 3910 }, { "grad_norm": 0.9642969369888306, "learning_rate": 7.838e-05, "loss": 0.106, "step": 3920 }, { "grad_norm": 0.4200981557369232, "learning_rate": 7.858000000000001e-05, "loss": 0.0877, "step": 3930 }, { "grad_norm": 0.8494421243667603, "learning_rate": 7.878e-05, "loss": 0.1168, "step": 3940 }, { "grad_norm": 0.6376171708106995, "learning_rate": 7.897999999999999e-05, "loss": 0.1034, "step": 3950 }, { "grad_norm": 0.6363007426261902, "learning_rate": 7.918e-05, "loss": 0.1067, "step": 3960 }, { "grad_norm": 0.5338305234909058, "learning_rate": 7.938e-05, "loss": 0.0758, "step": 3970 }, { "grad_norm": 0.43976831436157227, "learning_rate": 7.958e-05, "loss": 0.1019, "step": 3980 }, { "grad_norm": 0.7008061408996582, "learning_rate": 7.978e-05, "loss": 0.1068, "step": 3990 }, { "grad_norm": 0.9667803645133972, "learning_rate": 7.998e-05, "loss": 0.1146, "step": 4000 }, { "grad_norm": 0.8154069781303406, "learning_rate": 8.018e-05, "loss": 0.1011, "step": 4010 }, { "grad_norm": 0.6706463694572449, "learning_rate": 8.038e-05, "loss": 0.1009, "step": 4020 }, { "grad_norm": 0.6775062084197998, "learning_rate": 8.058e-05, "loss": 0.1168, "step": 4030 }, { "grad_norm": 0.7755032181739807, "learning_rate": 8.078e-05, "loss": 0.0929, "step": 4040 }, { "grad_norm": 0.5313028693199158, "learning_rate": 8.098e-05, "loss": 0.0959, "step": 4050 }, { "grad_norm": 0.7578502893447876, "learning_rate": 8.118e-05, "loss": 0.0899, "step": 4060 }, { "grad_norm": 0.6821162104606628, "learning_rate": 8.138e-05, "loss": 0.095, "step": 4070 }, { "grad_norm": 0.629173994064331, "learning_rate": 8.158e-05, "loss": 0.0981, "step": 4080 }, { "grad_norm": 0.5211255550384521, "learning_rate": 8.178e-05, "loss": 0.1088, "step": 4090 }, { "grad_norm": 0.5910434722900391, "learning_rate": 8.198e-05, "loss": 0.0935, "step": 4100 }, { "grad_norm": 0.5177809596061707, "learning_rate": 8.218e-05, "loss": 0.0963, "step": 4110 }, { "grad_norm": 0.9107664227485657, "learning_rate": 8.238000000000001e-05, "loss": 0.1197, "step": 4120 }, { "grad_norm": 0.41474249958992004, "learning_rate": 8.258e-05, "loss": 0.0865, "step": 4130 }, { "grad_norm": 0.5993481278419495, "learning_rate": 8.278e-05, "loss": 0.1066, "step": 4140 }, { "grad_norm": 0.6775844097137451, "learning_rate": 8.298000000000001e-05, "loss": 0.1068, "step": 4150 }, { "grad_norm": 0.8169371485710144, "learning_rate": 8.318e-05, "loss": 0.0939, "step": 4160 }, { "grad_norm": 0.6345911622047424, "learning_rate": 8.338e-05, "loss": 0.0946, "step": 4170 }, { "grad_norm": 0.453382283449173, "learning_rate": 8.358e-05, "loss": 0.0785, "step": 4180 }, { "grad_norm": 0.7507769465446472, "learning_rate": 8.378e-05, "loss": 0.0904, "step": 4190 }, { "grad_norm": 0.7657016515731812, "learning_rate": 8.398e-05, "loss": 0.1108, "step": 4200 }, { "grad_norm": 0.776472806930542, "learning_rate": 8.418e-05, "loss": 0.1097, "step": 4210 }, { "grad_norm": 0.7323020100593567, "learning_rate": 8.438e-05, "loss": 0.1116, "step": 4220 }, { "grad_norm": 0.6162874102592468, "learning_rate": 8.458e-05, "loss": 0.1045, "step": 4230 }, { "grad_norm": 0.554460883140564, "learning_rate": 8.478e-05, "loss": 0.1078, "step": 4240 }, { "grad_norm": 0.5947408080101013, "learning_rate": 8.498e-05, "loss": 0.1171, "step": 4250 }, { "grad_norm": 0.6880836486816406, "learning_rate": 8.518000000000001e-05, "loss": 0.1024, "step": 4260 }, { "grad_norm": 0.574665367603302, "learning_rate": 8.538e-05, "loss": 0.1001, "step": 4270 }, { "grad_norm": 0.5929573178291321, "learning_rate": 8.558e-05, "loss": 0.096, "step": 4280 }, { "grad_norm": 0.582463800907135, "learning_rate": 8.578000000000001e-05, "loss": 0.0877, "step": 4290 }, { "grad_norm": 0.48772290349006653, "learning_rate": 8.598e-05, "loss": 0.0855, "step": 4300 }, { "grad_norm": 0.8245601654052734, "learning_rate": 8.618e-05, "loss": 0.0958, "step": 4310 }, { "grad_norm": 0.5410375595092773, "learning_rate": 8.638000000000001e-05, "loss": 0.1279, "step": 4320 }, { "grad_norm": 0.8249337673187256, "learning_rate": 8.658e-05, "loss": 0.0888, "step": 4330 }, { "grad_norm": 1.0432721376419067, "learning_rate": 8.678e-05, "loss": 0.0933, "step": 4340 }, { "grad_norm": 0.8913310170173645, "learning_rate": 8.698000000000001e-05, "loss": 0.088, "step": 4350 }, { "grad_norm": 0.7037708759307861, "learning_rate": 8.718e-05, "loss": 0.0932, "step": 4360 }, { "grad_norm": 0.5712260007858276, "learning_rate": 8.738000000000001e-05, "loss": 0.0942, "step": 4370 }, { "grad_norm": 0.6363220810890198, "learning_rate": 8.758000000000001e-05, "loss": 0.0909, "step": 4380 }, { "grad_norm": 0.42182570695877075, "learning_rate": 8.778e-05, "loss": 0.0834, "step": 4390 }, { "grad_norm": 0.5564857125282288, "learning_rate": 8.798000000000001e-05, "loss": 0.1076, "step": 4400 }, { "grad_norm": 0.5961993932723999, "learning_rate": 8.818000000000001e-05, "loss": 0.101, "step": 4410 }, { "grad_norm": 0.6750171184539795, "learning_rate": 8.838e-05, "loss": 0.1021, "step": 4420 }, { "grad_norm": 0.6220190525054932, "learning_rate": 8.858000000000001e-05, "loss": 0.101, "step": 4430 }, { "grad_norm": 0.5936095714569092, "learning_rate": 8.878000000000001e-05, "loss": 0.1058, "step": 4440 }, { "grad_norm": 0.6778795123100281, "learning_rate": 8.898e-05, "loss": 0.1235, "step": 4450 }, { "grad_norm": 0.8313892483711243, "learning_rate": 8.918000000000001e-05, "loss": 0.0919, "step": 4460 }, { "grad_norm": 0.5395525693893433, "learning_rate": 8.938e-05, "loss": 0.1064, "step": 4470 }, { "grad_norm": 1.054982304573059, "learning_rate": 8.958e-05, "loss": 0.1024, "step": 4480 }, { "grad_norm": 0.5933322906494141, "learning_rate": 8.978000000000001e-05, "loss": 0.0944, "step": 4490 }, { "grad_norm": 0.47189733386039734, "learning_rate": 8.998e-05, "loss": 0.1018, "step": 4500 }, { "grad_norm": 0.4988139569759369, "learning_rate": 9.018000000000001e-05, "loss": 0.1099, "step": 4510 }, { "grad_norm": 0.5896015763282776, "learning_rate": 9.038000000000001e-05, "loss": 0.1076, "step": 4520 }, { "grad_norm": 0.607282817363739, "learning_rate": 9.058e-05, "loss": 0.094, "step": 4530 }, { "grad_norm": 0.6803016066551208, "learning_rate": 9.078000000000001e-05, "loss": 0.1046, "step": 4540 }, { "grad_norm": 0.624024510383606, "learning_rate": 9.098000000000001e-05, "loss": 0.1072, "step": 4550 }, { "grad_norm": 0.6654129028320312, "learning_rate": 9.118e-05, "loss": 0.1188, "step": 4560 }, { "grad_norm": 0.5989634394645691, "learning_rate": 9.138e-05, "loss": 0.1131, "step": 4570 }, { "grad_norm": 0.9547699093818665, "learning_rate": 9.158e-05, "loss": 0.1051, "step": 4580 }, { "grad_norm": 0.6293616890907288, "learning_rate": 9.178e-05, "loss": 0.1041, "step": 4590 }, { "grad_norm": 1.2888609170913696, "learning_rate": 9.198e-05, "loss": 0.0942, "step": 4600 }, { "grad_norm": 0.5381483435630798, "learning_rate": 9.218e-05, "loss": 0.0976, "step": 4610 }, { "grad_norm": 0.743687629699707, "learning_rate": 9.238e-05, "loss": 0.0814, "step": 4620 }, { "grad_norm": 0.8595154881477356, "learning_rate": 9.258e-05, "loss": 0.1084, "step": 4630 }, { "grad_norm": 0.5442202687263489, "learning_rate": 9.278e-05, "loss": 0.0997, "step": 4640 }, { "grad_norm": 0.5188539028167725, "learning_rate": 9.298e-05, "loss": 0.0871, "step": 4650 }, { "grad_norm": 0.5793619155883789, "learning_rate": 9.318e-05, "loss": 0.0948, "step": 4660 }, { "grad_norm": 0.7123681306838989, "learning_rate": 9.338e-05, "loss": 0.0885, "step": 4670 }, { "grad_norm": 0.5400840640068054, "learning_rate": 9.358e-05, "loss": 0.0859, "step": 4680 }, { "grad_norm": 0.5139044523239136, "learning_rate": 9.378e-05, "loss": 0.0909, "step": 4690 }, { "grad_norm": 0.516180157661438, "learning_rate": 9.398e-05, "loss": 0.0952, "step": 4700 }, { "grad_norm": 0.8767381310462952, "learning_rate": 9.418e-05, "loss": 0.1021, "step": 4710 }, { "grad_norm": 0.579221248626709, "learning_rate": 9.438e-05, "loss": 0.0911, "step": 4720 }, { "grad_norm": 0.5078046917915344, "learning_rate": 9.458e-05, "loss": 0.1079, "step": 4730 }, { "grad_norm": 0.4611551761627197, "learning_rate": 9.478e-05, "loss": 0.0931, "step": 4740 }, { "grad_norm": 0.6035267114639282, "learning_rate": 9.498e-05, "loss": 0.0873, "step": 4750 }, { "grad_norm": 0.6513412594795227, "learning_rate": 9.518000000000001e-05, "loss": 0.0976, "step": 4760 }, { "grad_norm": 0.564170777797699, "learning_rate": 9.538e-05, "loss": 0.1027, "step": 4770 }, { "grad_norm": 0.6456611156463623, "learning_rate": 9.558e-05, "loss": 0.1046, "step": 4780 }, { "grad_norm": 0.6209012269973755, "learning_rate": 9.578000000000001e-05, "loss": 0.0817, "step": 4790 }, { "grad_norm": 0.5091963410377502, "learning_rate": 9.598e-05, "loss": 0.1122, "step": 4800 }, { "grad_norm": 0.5510321855545044, "learning_rate": 9.618e-05, "loss": 0.0968, "step": 4810 }, { "grad_norm": 0.6248212456703186, "learning_rate": 9.638000000000001e-05, "loss": 0.1, "step": 4820 }, { "grad_norm": 0.8517590165138245, "learning_rate": 9.658e-05, "loss": 0.1015, "step": 4830 }, { "grad_norm": 0.4873780906200409, "learning_rate": 9.678e-05, "loss": 0.0892, "step": 4840 }, { "grad_norm": 0.7027065753936768, "learning_rate": 9.698000000000001e-05, "loss": 0.0895, "step": 4850 }, { "grad_norm": 1.2185274362564087, "learning_rate": 9.718e-05, "loss": 0.0841, "step": 4860 }, { "grad_norm": 0.5047493577003479, "learning_rate": 9.738e-05, "loss": 0.0922, "step": 4870 }, { "grad_norm": 0.48274222016334534, "learning_rate": 9.758000000000001e-05, "loss": 0.0883, "step": 4880 }, { "grad_norm": 0.8473528027534485, "learning_rate": 9.778e-05, "loss": 0.0941, "step": 4890 }, { "grad_norm": 0.6275802254676819, "learning_rate": 9.798000000000001e-05, "loss": 0.1034, "step": 4900 }, { "grad_norm": 0.5995041131973267, "learning_rate": 9.818000000000001e-05, "loss": 0.0877, "step": 4910 }, { "grad_norm": 0.8629735112190247, "learning_rate": 9.838e-05, "loss": 0.1151, "step": 4920 }, { "grad_norm": 0.4841732680797577, "learning_rate": 9.858000000000001e-05, "loss": 0.102, "step": 4930 }, { "grad_norm": 0.5262433290481567, "learning_rate": 9.878e-05, "loss": 0.0864, "step": 4940 }, { "grad_norm": 0.5250185132026672, "learning_rate": 9.898e-05, "loss": 0.091, "step": 4950 }, { "grad_norm": 0.5781921148300171, "learning_rate": 9.918000000000001e-05, "loss": 0.0982, "step": 4960 }, { "grad_norm": 0.602575957775116, "learning_rate": 9.938e-05, "loss": 0.1084, "step": 4970 }, { "grad_norm": 0.6278376579284668, "learning_rate": 9.958e-05, "loss": 0.1041, "step": 4980 }, { "grad_norm": 0.5196207761764526, "learning_rate": 9.978000000000001e-05, "loss": 0.0796, "step": 4990 }, { "grad_norm": 0.6767387390136719, "learning_rate": 9.998e-05, "loss": 0.0991, "step": 5000 }, { "grad_norm": 0.6830164194107056, "learning_rate": 9.999999778549045e-05, "loss": 0.0868, "step": 5010 }, { "grad_norm": 0.6296982765197754, "learning_rate": 9.999999013039593e-05, "loss": 0.0857, "step": 5020 }, { "grad_norm": 0.6790051460266113, "learning_rate": 9.999997700737766e-05, "loss": 0.0837, "step": 5030 }, { "grad_norm": 0.6964753270149231, "learning_rate": 9.999995841643709e-05, "loss": 0.0914, "step": 5040 }, { "grad_norm": 0.7476134300231934, "learning_rate": 9.999993435757623e-05, "loss": 0.1183, "step": 5050 }, { "grad_norm": 0.48251548409461975, "learning_rate": 9.999990483079773e-05, "loss": 0.0957, "step": 5060 }, { "grad_norm": 0.7169930338859558, "learning_rate": 9.999986983610481e-05, "loss": 0.1167, "step": 5070 }, { "grad_norm": 0.939412534236908, "learning_rate": 9.99998293735013e-05, "loss": 0.1093, "step": 5080 }, { "grad_norm": 0.4950328767299652, "learning_rate": 9.999978344299161e-05, "loss": 0.0889, "step": 5090 }, { "grad_norm": 0.5439654588699341, "learning_rate": 9.99997320445808e-05, "loss": 0.0967, "step": 5100 }, { "grad_norm": 0.4972032904624939, "learning_rate": 9.999967517827444e-05, "loss": 0.0931, "step": 5110 }, { "grad_norm": 0.5825090408325195, "learning_rate": 9.999961284407879e-05, "loss": 0.113, "step": 5120 }, { "grad_norm": 0.6754448413848877, "learning_rate": 9.999954504200067e-05, "loss": 0.113, "step": 5130 }, { "grad_norm": 0.6978862285614014, "learning_rate": 9.999947177204744e-05, "loss": 0.1066, "step": 5140 }, { "grad_norm": 0.5665680766105652, "learning_rate": 9.999939303422718e-05, "loss": 0.0915, "step": 5150 }, { "grad_norm": 0.7546841502189636, "learning_rate": 9.999930882854847e-05, "loss": 0.1127, "step": 5160 }, { "grad_norm": 1.3653168678283691, "learning_rate": 9.999921915502051e-05, "loss": 0.1013, "step": 5170 }, { "grad_norm": 0.6230541467666626, "learning_rate": 9.99991240136531e-05, "loss": 0.0978, "step": 5180 }, { "grad_norm": 0.7530800104141235, "learning_rate": 9.999902340445668e-05, "loss": 0.119, "step": 5190 }, { "grad_norm": 0.5857192873954773, "learning_rate": 9.999891732744224e-05, "loss": 0.1083, "step": 5200 }, { "grad_norm": 0.6995623707771301, "learning_rate": 9.999880578262135e-05, "loss": 0.0935, "step": 5210 }, { "grad_norm": 0.5489845871925354, "learning_rate": 9.999868877000624e-05, "loss": 0.0952, "step": 5220 }, { "grad_norm": 0.5932523012161255, "learning_rate": 9.99985662896097e-05, "loss": 0.0918, "step": 5230 }, { "grad_norm": 0.5885301232337952, "learning_rate": 9.999843834144513e-05, "loss": 0.0934, "step": 5240 }, { "grad_norm": 0.5515293478965759, "learning_rate": 9.99983049255265e-05, "loss": 0.1091, "step": 5250 }, { "grad_norm": 0.7974238991737366, "learning_rate": 9.999816604186843e-05, "loss": 0.0952, "step": 5260 }, { "grad_norm": 0.5812280774116516, "learning_rate": 9.999802169048609e-05, "loss": 0.0979, "step": 5270 }, { "grad_norm": 0.6429814696311951, "learning_rate": 9.999787187139527e-05, "loss": 0.1194, "step": 5280 }, { "grad_norm": 0.673903226852417, "learning_rate": 9.999771658461234e-05, "loss": 0.1046, "step": 5290 }, { "grad_norm": 0.7114124894142151, "learning_rate": 9.999755583015431e-05, "loss": 0.0878, "step": 5300 }, { "grad_norm": 1.251287579536438, "learning_rate": 9.999738960803874e-05, "loss": 0.093, "step": 5310 }, { "grad_norm": 0.5236539840698242, "learning_rate": 9.99972179182838e-05, "loss": 0.0895, "step": 5320 }, { "grad_norm": 0.6034770607948303, "learning_rate": 9.99970407609083e-05, "loss": 0.1, "step": 5330 }, { "grad_norm": 0.6025089621543884, "learning_rate": 9.999685813593159e-05, "loss": 0.1012, "step": 5340 }, { "grad_norm": 0.6621538996696472, "learning_rate": 9.999667004337362e-05, "loss": 0.0819, "step": 5350 }, { "grad_norm": 0.8579690456390381, "learning_rate": 9.9996476483255e-05, "loss": 0.102, "step": 5360 }, { "grad_norm": 0.5721607804298401, "learning_rate": 9.999627745559688e-05, "loss": 0.0888, "step": 5370 }, { "grad_norm": 0.6390917301177979, "learning_rate": 9.999607296042101e-05, "loss": 0.0942, "step": 5380 }, { "grad_norm": 0.673575758934021, "learning_rate": 9.99958629977498e-05, "loss": 0.1064, "step": 5390 }, { "grad_norm": 0.7151557803153992, "learning_rate": 9.999564756760615e-05, "loss": 0.1012, "step": 5400 }, { "grad_norm": 0.5364000201225281, "learning_rate": 9.999542667001366e-05, "loss": 0.0979, "step": 5410 }, { "grad_norm": 0.6029332876205444, "learning_rate": 9.999520030499647e-05, "loss": 0.0982, "step": 5420 }, { "grad_norm": 0.7869920134544373, "learning_rate": 9.999496847257936e-05, "loss": 0.097, "step": 5430 }, { "grad_norm": 0.7192692756652832, "learning_rate": 9.999473117278764e-05, "loss": 0.0972, "step": 5440 }, { "grad_norm": 0.5260676145553589, "learning_rate": 9.999448840564731e-05, "loss": 0.0835, "step": 5450 }, { "grad_norm": 0.6510304808616638, "learning_rate": 9.999424017118488e-05, "loss": 0.0854, "step": 5460 }, { "grad_norm": 0.43092575669288635, "learning_rate": 9.999398646942751e-05, "loss": 0.0736, "step": 5470 }, { "grad_norm": 0.4877787232398987, "learning_rate": 9.999372730040296e-05, "loss": 0.083, "step": 5480 }, { "grad_norm": 0.5552425384521484, "learning_rate": 9.999346266413953e-05, "loss": 0.0787, "step": 5490 }, { "grad_norm": 0.5004018545150757, "learning_rate": 9.99931925606662e-05, "loss": 0.0805, "step": 5500 }, { "grad_norm": 0.7599260807037354, "learning_rate": 9.99929169900125e-05, "loss": 0.0859, "step": 5510 }, { "grad_norm": 0.6935972571372986, "learning_rate": 9.999263595220855e-05, "loss": 0.075, "step": 5520 }, { "grad_norm": 0.4129263460636139, "learning_rate": 9.99923494472851e-05, "loss": 0.0803, "step": 5530 }, { "grad_norm": 0.6044372916221619, "learning_rate": 9.999205747527348e-05, "loss": 0.0744, "step": 5540 }, { "grad_norm": 0.6383456587791443, "learning_rate": 9.999176003620561e-05, "loss": 0.0749, "step": 5550 }, { "grad_norm": 0.4339534640312195, "learning_rate": 9.999145713011405e-05, "loss": 0.0707, "step": 5560 }, { "grad_norm": 0.5574052333831787, "learning_rate": 9.999114875703186e-05, "loss": 0.083, "step": 5570 }, { "grad_norm": 0.5758464336395264, "learning_rate": 9.999083491699281e-05, "loss": 0.0976, "step": 5580 }, { "grad_norm": 0.483314573764801, "learning_rate": 9.999051561003123e-05, "loss": 0.082, "step": 5590 }, { "grad_norm": 0.6498728394508362, "learning_rate": 9.999019083618202e-05, "loss": 0.1037, "step": 5600 }, { "grad_norm": 0.37652620673179626, "learning_rate": 9.99898605954807e-05, "loss": 0.0805, "step": 5610 }, { "grad_norm": 0.39435920119285583, "learning_rate": 9.998952488796338e-05, "loss": 0.0773, "step": 5620 }, { "grad_norm": 0.8756068348884583, "learning_rate": 9.998918371366676e-05, "loss": 0.0785, "step": 5630 }, { "grad_norm": 0.6454023122787476, "learning_rate": 9.99888370726282e-05, "loss": 0.0767, "step": 5640 }, { "grad_norm": 0.7314372062683105, "learning_rate": 9.998848496488556e-05, "loss": 0.0906, "step": 5650 }, { "grad_norm": 0.5638194680213928, "learning_rate": 9.998812739047736e-05, "loss": 0.0832, "step": 5660 }, { "grad_norm": 0.5082799792289734, "learning_rate": 9.99877643494427e-05, "loss": 0.0838, "step": 5670 }, { "grad_norm": 0.4346458315849304, "learning_rate": 9.998739584182128e-05, "loss": 0.0761, "step": 5680 }, { "grad_norm": 0.7183325886726379, "learning_rate": 9.998702186765342e-05, "loss": 0.0952, "step": 5690 }, { "grad_norm": 0.7628253698348999, "learning_rate": 9.998664242698e-05, "loss": 0.0865, "step": 5700 }, { "grad_norm": 0.732089102268219, "learning_rate": 9.998625751984251e-05, "loss": 0.0957, "step": 5710 }, { "grad_norm": 0.7235324382781982, "learning_rate": 9.998586714628307e-05, "loss": 0.0839, "step": 5720 }, { "grad_norm": 0.6585373878479004, "learning_rate": 9.998547130634432e-05, "loss": 0.09, "step": 5730 }, { "grad_norm": 0.508874773979187, "learning_rate": 9.99850700000696e-05, "loss": 0.0956, "step": 5740 }, { "grad_norm": 0.3894718587398529, "learning_rate": 9.998466322750278e-05, "loss": 0.0858, "step": 5750 }, { "grad_norm": 0.7035272121429443, "learning_rate": 9.998425098868834e-05, "loss": 0.0877, "step": 5760 }, { "grad_norm": 0.5206483006477356, "learning_rate": 9.998383328367136e-05, "loss": 0.0736, "step": 5770 }, { "grad_norm": 0.5150696039199829, "learning_rate": 9.99834101124975e-05, "loss": 0.0883, "step": 5780 }, { "grad_norm": 0.532234251499176, "learning_rate": 9.998298147521309e-05, "loss": 0.0761, "step": 5790 }, { "grad_norm": 0.6084682941436768, "learning_rate": 9.998254737186496e-05, "loss": 0.0874, "step": 5800 }, { "grad_norm": 0.5619395971298218, "learning_rate": 9.99821078025006e-05, "loss": 0.0866, "step": 5810 }, { "grad_norm": 0.7414094805717468, "learning_rate": 9.998166276716807e-05, "loss": 0.0923, "step": 5820 }, { "grad_norm": 0.5682315230369568, "learning_rate": 9.998121226591606e-05, "loss": 0.0934, "step": 5830 }, { "grad_norm": 0.6098215579986572, "learning_rate": 9.998075629879382e-05, "loss": 0.085, "step": 5840 }, { "grad_norm": 0.4930187165737152, "learning_rate": 9.99802948658512e-05, "loss": 0.0726, "step": 5850 }, { "grad_norm": 0.5084232091903687, "learning_rate": 9.99798279671387e-05, "loss": 0.0805, "step": 5860 }, { "grad_norm": 0.5231969952583313, "learning_rate": 9.997935560270734e-05, "loss": 0.0895, "step": 5870 }, { "grad_norm": 0.46494293212890625, "learning_rate": 9.997887777260879e-05, "loss": 0.0926, "step": 5880 }, { "grad_norm": 0.4581216275691986, "learning_rate": 9.997839447689532e-05, "loss": 0.0791, "step": 5890 }, { "grad_norm": 0.6309420466423035, "learning_rate": 9.997790571561978e-05, "loss": 0.11, "step": 5900 }, { "grad_norm": 1.0411049127578735, "learning_rate": 9.99774114888356e-05, "loss": 0.0952, "step": 5910 }, { "grad_norm": 0.8423668742179871, "learning_rate": 9.997691179659684e-05, "loss": 0.0845, "step": 5920 }, { "grad_norm": 0.5351660847663879, "learning_rate": 9.997640663895815e-05, "loss": 0.0827, "step": 5930 }, { "grad_norm": 0.5629948973655701, "learning_rate": 9.997589601597477e-05, "loss": 0.0716, "step": 5940 }, { "grad_norm": 0.5758786797523499, "learning_rate": 9.997537992770252e-05, "loss": 0.0871, "step": 5950 }, { "grad_norm": 0.8300349712371826, "learning_rate": 9.997485837419788e-05, "loss": 0.08, "step": 5960 }, { "grad_norm": 0.8274376392364502, "learning_rate": 9.997433135551786e-05, "loss": 0.0905, "step": 5970 }, { "grad_norm": 0.4599776864051819, "learning_rate": 9.997379887172009e-05, "loss": 0.0822, "step": 5980 }, { "grad_norm": 0.4812140464782715, "learning_rate": 9.997326092286281e-05, "loss": 0.0764, "step": 5990 }, { "grad_norm": 0.7825086116790771, "learning_rate": 9.997271750900486e-05, "loss": 0.1156, "step": 6000 }, { "grad_norm": 0.5299860239028931, "learning_rate": 9.997216863020565e-05, "loss": 0.0866, "step": 6010 }, { "grad_norm": 0.5712087154388428, "learning_rate": 9.99716142865252e-05, "loss": 0.0831, "step": 6020 }, { "grad_norm": 0.663847029209137, "learning_rate": 9.997105447802415e-05, "loss": 0.0783, "step": 6030 }, { "grad_norm": 0.6217572689056396, "learning_rate": 9.997048920476373e-05, "loss": 0.0765, "step": 6040 }, { "grad_norm": 0.694840669631958, "learning_rate": 9.996991846680572e-05, "loss": 0.0902, "step": 6050 }, { "grad_norm": 0.5093729496002197, "learning_rate": 9.996934226421257e-05, "loss": 0.0722, "step": 6060 }, { "grad_norm": 0.5856476426124573, "learning_rate": 9.996876059704726e-05, "loss": 0.0858, "step": 6070 }, { "grad_norm": 0.5382934212684631, "learning_rate": 9.996817346537343e-05, "loss": 0.0939, "step": 6080 }, { "grad_norm": 0.4749177396297455, "learning_rate": 9.996758086925526e-05, "loss": 0.0819, "step": 6090 }, { "grad_norm": 0.43044090270996094, "learning_rate": 9.996698280875759e-05, "loss": 0.0952, "step": 6100 }, { "grad_norm": 0.5690100193023682, "learning_rate": 9.99663792839458e-05, "loss": 0.0673, "step": 6110 }, { "grad_norm": 0.5105999708175659, "learning_rate": 9.99657702948859e-05, "loss": 0.082, "step": 6120 }, { "grad_norm": 0.5516719818115234, "learning_rate": 9.996515584164448e-05, "loss": 0.0661, "step": 6130 }, { "grad_norm": 0.5671621561050415, "learning_rate": 9.996453592428873e-05, "loss": 0.0818, "step": 6140 }, { "grad_norm": 0.5945515632629395, "learning_rate": 9.996391054288646e-05, "loss": 0.1036, "step": 6150 }, { "grad_norm": 0.7571638226509094, "learning_rate": 9.996327969750605e-05, "loss": 0.1122, "step": 6160 }, { "grad_norm": 0.5393619537353516, "learning_rate": 9.996264338821649e-05, "loss": 0.0992, "step": 6170 }, { "grad_norm": 0.6404428482055664, "learning_rate": 9.996200161508735e-05, "loss": 0.0814, "step": 6180 }, { "grad_norm": 0.5335026383399963, "learning_rate": 9.996135437818885e-05, "loss": 0.1009, "step": 6190 }, { "grad_norm": 0.4876450300216675, "learning_rate": 9.996070167759175e-05, "loss": 0.0859, "step": 6200 }, { "grad_norm": 0.6842040419578552, "learning_rate": 9.996004351336743e-05, "loss": 0.0848, "step": 6210 }, { "grad_norm": 0.4995551109313965, "learning_rate": 9.995937988558785e-05, "loss": 0.0807, "step": 6220 }, { "grad_norm": 0.49707695841789246, "learning_rate": 9.995871079432561e-05, "loss": 0.0792, "step": 6230 }, { "grad_norm": 0.47044187784194946, "learning_rate": 9.995803623965389e-05, "loss": 0.0823, "step": 6240 }, { "grad_norm": 0.85899817943573, "learning_rate": 9.995735622164641e-05, "loss": 0.0793, "step": 6250 }, { "grad_norm": 0.7818295955657959, "learning_rate": 9.995667074037758e-05, "loss": 0.0713, "step": 6260 }, { "grad_norm": 0.41526445746421814, "learning_rate": 9.995597979592232e-05, "loss": 0.0786, "step": 6270 }, { "grad_norm": 0.500050961971283, "learning_rate": 9.995528338835625e-05, "loss": 0.0825, "step": 6280 }, { "grad_norm": 0.5479579567909241, "learning_rate": 9.995458151775547e-05, "loss": 0.0772, "step": 6290 }, { "grad_norm": 0.5520066022872925, "learning_rate": 9.995387418419677e-05, "loss": 0.0998, "step": 6300 }, { "grad_norm": 0.5575284361839294, "learning_rate": 9.99531613877575e-05, "loss": 0.0982, "step": 6310 }, { "grad_norm": 0.43886950612068176, "learning_rate": 9.995244312851559e-05, "loss": 0.0858, "step": 6320 }, { "grad_norm": 0.5890887975692749, "learning_rate": 9.995171940654961e-05, "loss": 0.0845, "step": 6330 }, { "grad_norm": 0.5624409317970276, "learning_rate": 9.995099022193871e-05, "loss": 0.0906, "step": 6340 }, { "grad_norm": 0.5009239912033081, "learning_rate": 9.995025557476261e-05, "loss": 0.0861, "step": 6350 }, { "grad_norm": 0.646872878074646, "learning_rate": 9.994951546510165e-05, "loss": 0.0832, "step": 6360 }, { "grad_norm": 0.5669960975646973, "learning_rate": 9.994876989303679e-05, "loss": 0.0867, "step": 6370 }, { "grad_norm": 0.5040633082389832, "learning_rate": 9.994801885864955e-05, "loss": 0.0857, "step": 6380 }, { "grad_norm": 0.6021477580070496, "learning_rate": 9.994726236202205e-05, "loss": 0.0792, "step": 6390 }, { "grad_norm": 0.43056634068489075, "learning_rate": 9.994650040323704e-05, "loss": 0.0841, "step": 6400 }, { "grad_norm": 0.7628158330917358, "learning_rate": 9.994573298237784e-05, "loss": 0.077, "step": 6410 }, { "grad_norm": 0.5627318620681763, "learning_rate": 9.994496009952837e-05, "loss": 0.0847, "step": 6420 }, { "grad_norm": 0.5331496596336365, "learning_rate": 9.994418175477316e-05, "loss": 0.0754, "step": 6430 }, { "grad_norm": 0.5359110236167908, "learning_rate": 9.994339794819733e-05, "loss": 0.0795, "step": 6440 }, { "grad_norm": 0.48072776198387146, "learning_rate": 9.994260867988658e-05, "loss": 0.0686, "step": 6450 }, { "grad_norm": 0.44873857498168945, "learning_rate": 9.994181394992723e-05, "loss": 0.072, "step": 6460 }, { "grad_norm": 0.4802972376346588, "learning_rate": 9.994101375840618e-05, "loss": 0.0767, "step": 6470 }, { "grad_norm": 0.5417256951332092, "learning_rate": 9.994020810541098e-05, "loss": 0.0666, "step": 6480 }, { "grad_norm": 0.5975290536880493, "learning_rate": 9.99393969910297e-05, "loss": 0.0864, "step": 6490 }, { "grad_norm": 0.5163251161575317, "learning_rate": 9.993858041535104e-05, "loss": 0.0838, "step": 6500 }, { "grad_norm": 0.4529055058956146, "learning_rate": 9.99377583784643e-05, "loss": 0.072, "step": 6510 }, { "grad_norm": 0.4419057369232178, "learning_rate": 9.993693088045939e-05, "loss": 0.0816, "step": 6520 }, { "grad_norm": 0.5368260145187378, "learning_rate": 9.99360979214268e-05, "loss": 0.0952, "step": 6530 }, { "grad_norm": 0.44346290826797485, "learning_rate": 9.99352595014576e-05, "loss": 0.0944, "step": 6540 }, { "grad_norm": 0.6158344745635986, "learning_rate": 9.993441562064354e-05, "loss": 0.0866, "step": 6550 }, { "grad_norm": 0.47308337688446045, "learning_rate": 9.993356627907685e-05, "loss": 0.0801, "step": 6560 }, { "grad_norm": 0.608809769153595, "learning_rate": 9.99327114768504e-05, "loss": 0.0932, "step": 6570 }, { "grad_norm": 0.6439595222473145, "learning_rate": 9.99318512140577e-05, "loss": 0.0862, "step": 6580 }, { "grad_norm": 0.8498368859291077, "learning_rate": 9.993098549079284e-05, "loss": 0.1025, "step": 6590 }, { "grad_norm": 0.4808712601661682, "learning_rate": 9.993011430715047e-05, "loss": 0.1009, "step": 6600 }, { "grad_norm": 0.741065263748169, "learning_rate": 9.992923766322586e-05, "loss": 0.0785, "step": 6610 }, { "grad_norm": 0.5041429996490479, "learning_rate": 9.99283555591149e-05, "loss": 0.0797, "step": 6620 }, { "grad_norm": 1.0963410139083862, "learning_rate": 9.992746799491404e-05, "loss": 0.0798, "step": 6630 }, { "grad_norm": 0.902653694152832, "learning_rate": 9.992657497072033e-05, "loss": 0.0863, "step": 6640 }, { "grad_norm": 0.44988641142845154, "learning_rate": 9.992567648663147e-05, "loss": 0.103, "step": 6650 }, { "grad_norm": 0.4422292113304138, "learning_rate": 9.992477254274568e-05, "loss": 0.0728, "step": 6660 }, { "grad_norm": 0.5714475512504578, "learning_rate": 9.992386313916183e-05, "loss": 0.0809, "step": 6670 }, { "grad_norm": 1.746603012084961, "learning_rate": 9.992294827597934e-05, "loss": 0.117, "step": 6680 }, { "grad_norm": 0.6609501242637634, "learning_rate": 9.992202795329831e-05, "loss": 0.0925, "step": 6690 }, { "grad_norm": 0.5740496516227722, "learning_rate": 9.992110217121936e-05, "loss": 0.0858, "step": 6700 }, { "grad_norm": 0.4753771424293518, "learning_rate": 9.992017092984372e-05, "loss": 0.086, "step": 6710 }, { "grad_norm": 0.5633362531661987, "learning_rate": 9.991923422927326e-05, "loss": 0.084, "step": 6720 }, { "grad_norm": 0.629802942276001, "learning_rate": 9.991829206961037e-05, "loss": 0.0944, "step": 6730 }, { "grad_norm": 0.39654576778411865, "learning_rate": 9.991734445095813e-05, "loss": 0.1072, "step": 6740 }, { "grad_norm": 0.6198650598526001, "learning_rate": 9.991639137342015e-05, "loss": 0.0892, "step": 6750 }, { "grad_norm": 0.5939840078353882, "learning_rate": 9.991543283710064e-05, "loss": 0.0826, "step": 6760 }, { "grad_norm": 0.6905688643455505, "learning_rate": 9.991446884210445e-05, "loss": 0.0777, "step": 6770 }, { "grad_norm": 0.5937542915344238, "learning_rate": 9.9913499388537e-05, "loss": 0.0855, "step": 6780 }, { "grad_norm": 0.7624496817588806, "learning_rate": 9.99125244765043e-05, "loss": 0.0894, "step": 6790 }, { "grad_norm": 0.6255714297294617, "learning_rate": 9.991154410611296e-05, "loss": 0.1037, "step": 6800 }, { "grad_norm": 0.4790450930595398, "learning_rate": 9.99105582774702e-05, "loss": 0.0816, "step": 6810 }, { "grad_norm": 0.45049870014190674, "learning_rate": 9.990956699068384e-05, "loss": 0.0948, "step": 6820 }, { "grad_norm": 0.44296127557754517, "learning_rate": 9.990857024586224e-05, "loss": 0.0882, "step": 6830 }, { "grad_norm": 0.4829314649105072, "learning_rate": 9.990756804311446e-05, "loss": 0.1008, "step": 6840 }, { "grad_norm": 0.5568238496780396, "learning_rate": 9.990656038255006e-05, "loss": 0.0819, "step": 6850 }, { "grad_norm": 0.6187124848365784, "learning_rate": 9.990554726427926e-05, "loss": 0.0721, "step": 6860 }, { "grad_norm": 0.8295088410377502, "learning_rate": 9.990452868841284e-05, "loss": 0.1044, "step": 6870 }, { "grad_norm": 0.46162131428718567, "learning_rate": 9.99035046550622e-05, "loss": 0.0794, "step": 6880 }, { "grad_norm": 0.509081244468689, "learning_rate": 9.99024751643393e-05, "loss": 0.0714, "step": 6890 }, { "grad_norm": 0.8333366513252258, "learning_rate": 9.990144021635677e-05, "loss": 0.0892, "step": 6900 }, { "grad_norm": 0.6823648810386658, "learning_rate": 9.990039981122775e-05, "loss": 0.0775, "step": 6910 }, { "grad_norm": 0.6949301958084106, "learning_rate": 9.989935394906602e-05, "loss": 0.0893, "step": 6920 }, { "grad_norm": 0.46465855836868286, "learning_rate": 9.989830262998598e-05, "loss": 0.0839, "step": 6930 }, { "grad_norm": 0.5795624256134033, "learning_rate": 9.989724585410259e-05, "loss": 0.0782, "step": 6940 }, { "grad_norm": 0.5859248638153076, "learning_rate": 9.989618362153139e-05, "loss": 0.0704, "step": 6950 }, { "grad_norm": 0.48996439576148987, "learning_rate": 9.989511593238859e-05, "loss": 0.0709, "step": 6960 }, { "grad_norm": 0.6675187945365906, "learning_rate": 9.98940427867909e-05, "loss": 0.0725, "step": 6970 }, { "grad_norm": 0.5399051308631897, "learning_rate": 9.989296418485573e-05, "loss": 0.0828, "step": 6980 }, { "grad_norm": 0.6558834910392761, "learning_rate": 9.989188012670101e-05, "loss": 0.0769, "step": 6990 }, { "grad_norm": 0.48665305972099304, "learning_rate": 9.989079061244528e-05, "loss": 0.0752, "step": 7000 }, { "grad_norm": 0.4502890706062317, "learning_rate": 9.988969564220769e-05, "loss": 0.0867, "step": 7010 }, { "grad_norm": 0.5547441244125366, "learning_rate": 9.988859521610801e-05, "loss": 0.0825, "step": 7020 }, { "grad_norm": 0.5063604116439819, "learning_rate": 9.988748933426656e-05, "loss": 0.0948, "step": 7030 }, { "grad_norm": 0.49813100695610046, "learning_rate": 9.988637799680428e-05, "loss": 0.0781, "step": 7040 }, { "grad_norm": 0.41337278485298157, "learning_rate": 9.98852612038427e-05, "loss": 0.0791, "step": 7050 }, { "grad_norm": 0.48750439286231995, "learning_rate": 9.988413895550397e-05, "loss": 0.08, "step": 7060 }, { "grad_norm": 0.48774588108062744, "learning_rate": 9.98830112519108e-05, "loss": 0.0747, "step": 7070 }, { "grad_norm": 0.591492772102356, "learning_rate": 9.98818780931865e-05, "loss": 0.0897, "step": 7080 }, { "grad_norm": 0.5640372037887573, "learning_rate": 9.988073947945502e-05, "loss": 0.087, "step": 7090 }, { "grad_norm": 0.7391645312309265, "learning_rate": 9.987959541084087e-05, "loss": 0.0838, "step": 7100 }, { "grad_norm": 0.3877103924751282, "learning_rate": 9.987844588746915e-05, "loss": 0.0737, "step": 7110 }, { "grad_norm": 0.3597821593284607, "learning_rate": 9.987729090946558e-05, "loss": 0.0589, "step": 7120 }, { "grad_norm": 0.3474735915660858, "learning_rate": 9.987613047695647e-05, "loss": 0.0565, "step": 7130 }, { "grad_norm": 0.49103844165802, "learning_rate": 9.987496459006871e-05, "loss": 0.0835, "step": 7140 }, { "grad_norm": 0.3797304034233093, "learning_rate": 9.987379324892982e-05, "loss": 0.069, "step": 7150 }, { "grad_norm": 0.777701735496521, "learning_rate": 9.987261645366788e-05, "loss": 0.083, "step": 7160 }, { "grad_norm": 0.9701107740402222, "learning_rate": 9.987143420441158e-05, "loss": 0.0742, "step": 7170 }, { "grad_norm": 0.5071896910667419, "learning_rate": 9.987024650129022e-05, "loss": 0.0792, "step": 7180 }, { "grad_norm": 0.9283631443977356, "learning_rate": 9.986905334443368e-05, "loss": 0.0883, "step": 7190 }, { "grad_norm": 0.5094850063323975, "learning_rate": 9.986785473397245e-05, "loss": 0.072, "step": 7200 }, { "grad_norm": 0.4589231014251709, "learning_rate": 9.98666506700376e-05, "loss": 0.085, "step": 7210 }, { "grad_norm": 0.5603319406509399, "learning_rate": 9.986544115276081e-05, "loss": 0.0891, "step": 7220 }, { "grad_norm": 0.8664730191230774, "learning_rate": 9.986422618227433e-05, "loss": 0.0889, "step": 7230 }, { "grad_norm": 0.583276629447937, "learning_rate": 9.986300575871106e-05, "loss": 0.0948, "step": 7240 }, { "grad_norm": 0.44265997409820557, "learning_rate": 9.986177988220444e-05, "loss": 0.0982, "step": 7250 }, { "grad_norm": 0.544095516204834, "learning_rate": 9.986054855288856e-05, "loss": 0.0827, "step": 7260 }, { "grad_norm": 0.845276951789856, "learning_rate": 9.985931177089802e-05, "loss": 0.0974, "step": 7270 }, { "grad_norm": 0.6964960098266602, "learning_rate": 9.985806953636814e-05, "loss": 0.0772, "step": 7280 }, { "grad_norm": 0.4152548611164093, "learning_rate": 9.985682184943471e-05, "loss": 0.0791, "step": 7290 }, { "grad_norm": 0.5581075549125671, "learning_rate": 9.98555687102342e-05, "loss": 0.0754, "step": 7300 }, { "grad_norm": 0.6170472502708435, "learning_rate": 9.985431011890367e-05, "loss": 0.07, "step": 7310 }, { "grad_norm": 0.42365366220474243, "learning_rate": 9.985304607558075e-05, "loss": 0.0619, "step": 7320 }, { "grad_norm": 0.4947020411491394, "learning_rate": 9.985177658040364e-05, "loss": 0.0799, "step": 7330 }, { "grad_norm": 0.4351385831832886, "learning_rate": 9.985050163351119e-05, "loss": 0.0891, "step": 7340 }, { "grad_norm": 0.42952096462249756, "learning_rate": 9.984922123504286e-05, "loss": 0.0795, "step": 7350 }, { "grad_norm": 0.6613431572914124, "learning_rate": 9.984793538513862e-05, "loss": 0.0899, "step": 7360 }, { "grad_norm": 0.4196169674396515, "learning_rate": 9.984664408393912e-05, "loss": 0.0765, "step": 7370 }, { "grad_norm": 0.46992480754852295, "learning_rate": 9.984534733158556e-05, "loss": 0.0789, "step": 7380 }, { "grad_norm": 0.3992558717727661, "learning_rate": 9.984404512821977e-05, "loss": 0.0741, "step": 7390 }, { "grad_norm": 0.46972882747650146, "learning_rate": 9.984273747398411e-05, "loss": 0.0769, "step": 7400 }, { "grad_norm": 0.7504937052726746, "learning_rate": 9.984142436902165e-05, "loss": 0.09, "step": 7410 }, { "grad_norm": 0.6396647691726685, "learning_rate": 9.984010581347596e-05, "loss": 0.0953, "step": 7420 }, { "grad_norm": 0.7074894309043884, "learning_rate": 9.983878180749121e-05, "loss": 0.0988, "step": 7430 }, { "grad_norm": 0.7239507436752319, "learning_rate": 9.983745235121222e-05, "loss": 0.0778, "step": 7440 }, { "grad_norm": 0.43905407190322876, "learning_rate": 9.983611744478438e-05, "loss": 0.0724, "step": 7450 }, { "grad_norm": 0.5643949508666992, "learning_rate": 9.983477708835365e-05, "loss": 0.0895, "step": 7460 }, { "grad_norm": 0.5512492060661316, "learning_rate": 9.983343128206664e-05, "loss": 0.0928, "step": 7470 }, { "grad_norm": 0.5796912908554077, "learning_rate": 9.983208002607049e-05, "loss": 0.0986, "step": 7480 }, { "grad_norm": 0.5290583372116089, "learning_rate": 9.9830723320513e-05, "loss": 0.088, "step": 7490 }, { "grad_norm": 0.37465766072273254, "learning_rate": 9.982936116554254e-05, "loss": 0.0933, "step": 7500 }, { "grad_norm": 0.5846315622329712, "learning_rate": 9.982799356130803e-05, "loss": 0.0805, "step": 7510 }, { "grad_norm": 0.5571986436843872, "learning_rate": 9.982662050795908e-05, "loss": 0.0712, "step": 7520 }, { "grad_norm": 0.4341917335987091, "learning_rate": 9.982524200564583e-05, "loss": 0.0745, "step": 7530 }, { "grad_norm": 0.3366028666496277, "learning_rate": 9.982385805451901e-05, "loss": 0.073, "step": 7540 }, { "grad_norm": 0.45623520016670227, "learning_rate": 9.982246865472998e-05, "loss": 0.0704, "step": 7550 }, { "grad_norm": 0.6206967234611511, "learning_rate": 9.982107380643069e-05, "loss": 0.0744, "step": 7560 }, { "grad_norm": 0.7144664525985718, "learning_rate": 9.981967350977368e-05, "loss": 0.0873, "step": 7570 }, { "grad_norm": 0.556751549243927, "learning_rate": 9.981826776491208e-05, "loss": 0.0823, "step": 7580 }, { "grad_norm": 0.5886870622634888, "learning_rate": 9.98168565719996e-05, "loss": 0.0839, "step": 7590 }, { "grad_norm": 1.0006585121154785, "learning_rate": 9.98154399311906e-05, "loss": 0.0912, "step": 7600 }, { "grad_norm": 0.42369574308395386, "learning_rate": 9.981401784263997e-05, "loss": 0.0737, "step": 7610 }, { "grad_norm": 0.685171365737915, "learning_rate": 9.981259030650326e-05, "loss": 0.0742, "step": 7620 }, { "grad_norm": 0.5332068204879761, "learning_rate": 9.981115732293655e-05, "loss": 0.0851, "step": 7630 }, { "grad_norm": 0.46598154306411743, "learning_rate": 9.980971889209659e-05, "loss": 0.0705, "step": 7640 }, { "grad_norm": 0.4073695242404938, "learning_rate": 9.980827501414064e-05, "loss": 0.0792, "step": 7650 }, { "grad_norm": 0.586333155632019, "learning_rate": 9.980682568922663e-05, "loss": 0.069, "step": 7660 }, { "grad_norm": 0.5215618014335632, "learning_rate": 9.980537091751304e-05, "loss": 0.0753, "step": 7670 }, { "grad_norm": 0.5233636498451233, "learning_rate": 9.980391069915897e-05, "loss": 0.0715, "step": 7680 }, { "grad_norm": 0.45928439497947693, "learning_rate": 9.98024450343241e-05, "loss": 0.0723, "step": 7690 }, { "grad_norm": 0.4666677415370941, "learning_rate": 9.980097392316872e-05, "loss": 0.0696, "step": 7700 }, { "grad_norm": 0.5777807235717773, "learning_rate": 9.97994973658537e-05, "loss": 0.0826, "step": 7710 }, { "grad_norm": 0.3659957945346832, "learning_rate": 9.979801536254054e-05, "loss": 0.0929, "step": 7720 }, { "grad_norm": 0.42091137170791626, "learning_rate": 9.979652791339127e-05, "loss": 0.0675, "step": 7730 }, { "grad_norm": 0.4152044951915741, "learning_rate": 9.97950350185686e-05, "loss": 0.068, "step": 7740 }, { "grad_norm": 0.43301594257354736, "learning_rate": 9.979353667823574e-05, "loss": 0.0705, "step": 7750 }, { "grad_norm": 0.3979814350605011, "learning_rate": 9.979203289255658e-05, "loss": 0.0708, "step": 7760 }, { "grad_norm": 0.43790382146835327, "learning_rate": 9.979052366169557e-05, "loss": 0.0741, "step": 7770 }, { "grad_norm": 0.44089463353157043, "learning_rate": 9.978900898581775e-05, "loss": 0.0746, "step": 7780 }, { "grad_norm": 0.5593128800392151, "learning_rate": 9.978748886508875e-05, "loss": 0.082, "step": 7790 }, { "grad_norm": 0.646177351474762, "learning_rate": 9.978596329967484e-05, "loss": 0.1004, "step": 7800 }, { "grad_norm": 0.408924400806427, "learning_rate": 9.978443228974284e-05, "loss": 0.0682, "step": 7810 }, { "grad_norm": 0.5677200555801392, "learning_rate": 9.978289583546015e-05, "loss": 0.0694, "step": 7820 }, { "grad_norm": 0.4090668261051178, "learning_rate": 9.978135393699484e-05, "loss": 0.0643, "step": 7830 }, { "grad_norm": 0.5954995155334473, "learning_rate": 9.977980659451548e-05, "loss": 0.0736, "step": 7840 }, { "grad_norm": 1.6989519596099854, "learning_rate": 9.977825380819135e-05, "loss": 0.0884, "step": 7850 }, { "grad_norm": 0.5133607983589172, "learning_rate": 9.97766955781922e-05, "loss": 0.0922, "step": 7860 }, { "grad_norm": 0.6693598628044128, "learning_rate": 9.977513190468848e-05, "loss": 0.0829, "step": 7870 }, { "grad_norm": 0.4122161269187927, "learning_rate": 9.977356278785116e-05, "loss": 0.0781, "step": 7880 }, { "grad_norm": 0.6194725036621094, "learning_rate": 9.977198822785184e-05, "loss": 0.0718, "step": 7890 }, { "grad_norm": 0.34420865774154663, "learning_rate": 9.977040822486273e-05, "loss": 0.106, "step": 7900 }, { "grad_norm": 0.4658917188644409, "learning_rate": 9.97688227790566e-05, "loss": 0.0749, "step": 7910 }, { "grad_norm": 0.6367108821868896, "learning_rate": 9.976723189060684e-05, "loss": 0.0979, "step": 7920 }, { "grad_norm": 0.4981233477592468, "learning_rate": 9.976563555968742e-05, "loss": 0.0669, "step": 7930 }, { "grad_norm": 0.3253311216831207, "learning_rate": 9.976403378647292e-05, "loss": 0.0734, "step": 7940 }, { "grad_norm": 0.6184011101722717, "learning_rate": 9.97624265711385e-05, "loss": 0.0837, "step": 7950 }, { "grad_norm": 0.37201324105262756, "learning_rate": 9.976081391385993e-05, "loss": 0.0684, "step": 7960 }, { "grad_norm": 0.7601473331451416, "learning_rate": 9.975919581481356e-05, "loss": 0.0664, "step": 7970 }, { "grad_norm": 0.5009045004844666, "learning_rate": 9.975757227417634e-05, "loss": 0.078, "step": 7980 }, { "grad_norm": 0.6164166927337646, "learning_rate": 9.975594329212586e-05, "loss": 0.0867, "step": 7990 }, { "grad_norm": 0.5105886459350586, "learning_rate": 9.97543088688402e-05, "loss": 0.07, "step": 8000 }, { "grad_norm": 0.6547042727470398, "learning_rate": 9.975266900449814e-05, "loss": 0.0754, "step": 8010 }, { "grad_norm": 0.4216683804988861, "learning_rate": 9.975102369927898e-05, "loss": 0.0716, "step": 8020 }, { "grad_norm": 0.567327082157135, "learning_rate": 9.974937295336269e-05, "loss": 0.0695, "step": 8030 }, { "grad_norm": 0.43913790583610535, "learning_rate": 9.974771676692975e-05, "loss": 0.0784, "step": 8040 }, { "grad_norm": 0.47907042503356934, "learning_rate": 9.974605514016131e-05, "loss": 0.0804, "step": 8050 }, { "grad_norm": 0.5230476260185242, "learning_rate": 9.974438807323907e-05, "loss": 0.0686, "step": 8060 }, { "grad_norm": 0.48555171489715576, "learning_rate": 9.974271556634535e-05, "loss": 0.0693, "step": 8070 }, { "grad_norm": 0.5562572479248047, "learning_rate": 9.974103761966302e-05, "loss": 0.0875, "step": 8080 }, { "grad_norm": 0.5963612198829651, "learning_rate": 9.973935423337563e-05, "loss": 0.0634, "step": 8090 }, { "grad_norm": 0.4265187084674835, "learning_rate": 9.973766540766722e-05, "loss": 0.0747, "step": 8100 }, { "grad_norm": 0.577989935874939, "learning_rate": 9.97359711427225e-05, "loss": 0.0759, "step": 8110 }, { "grad_norm": 0.699626624584198, "learning_rate": 9.973427143872677e-05, "loss": 0.1024, "step": 8120 }, { "grad_norm": 0.6553930044174194, "learning_rate": 9.973256629586589e-05, "loss": 0.092, "step": 8130 }, { "grad_norm": 0.4825073182582855, "learning_rate": 9.973085571432632e-05, "loss": 0.0712, "step": 8140 }, { "grad_norm": 0.45634737610816956, "learning_rate": 9.972913969429513e-05, "loss": 0.0677, "step": 8150 }, { "grad_norm": 0.36222174763679504, "learning_rate": 9.972741823596e-05, "loss": 0.0763, "step": 8160 }, { "grad_norm": 0.5648095607757568, "learning_rate": 9.972569133950917e-05, "loss": 0.0775, "step": 8170 }, { "grad_norm": 0.40402257442474365, "learning_rate": 9.972395900513151e-05, "loss": 0.073, "step": 8180 }, { "grad_norm": 0.4686986207962036, "learning_rate": 9.972222123301645e-05, "loss": 0.0769, "step": 8190 }, { "grad_norm": 0.3656535744667053, "learning_rate": 9.972047802335403e-05, "loss": 0.074, "step": 8200 }, { "grad_norm": 0.6746503710746765, "learning_rate": 9.971872937633488e-05, "loss": 0.0924, "step": 8210 }, { "grad_norm": 0.5060656070709229, "learning_rate": 9.971697529215024e-05, "loss": 0.0811, "step": 8220 }, { "grad_norm": 0.4881823658943176, "learning_rate": 9.971521577099192e-05, "loss": 0.0846, "step": 8230 }, { "grad_norm": 0.4540814161300659, "learning_rate": 9.971345081305236e-05, "loss": 0.0908, "step": 8240 }, { "grad_norm": 0.5079106688499451, "learning_rate": 9.971168041852456e-05, "loss": 0.07, "step": 8250 }, { "grad_norm": 0.4135422110557556, "learning_rate": 9.970990458760215e-05, "loss": 0.067, "step": 8260 }, { "grad_norm": 0.36278820037841797, "learning_rate": 9.970812332047929e-05, "loss": 0.0685, "step": 8270 }, { "grad_norm": 0.47569769620895386, "learning_rate": 9.97063366173508e-05, "loss": 0.0823, "step": 8280 }, { "grad_norm": 0.3993738889694214, "learning_rate": 9.970454447841207e-05, "loss": 0.0698, "step": 8290 }, { "grad_norm": 0.3519040644168854, "learning_rate": 9.970274690385909e-05, "loss": 0.0689, "step": 8300 }, { "grad_norm": 0.4837682843208313, "learning_rate": 9.970094389388844e-05, "loss": 0.0772, "step": 8310 }, { "grad_norm": 0.521850049495697, "learning_rate": 9.969913544869728e-05, "loss": 0.058, "step": 8320 }, { "grad_norm": 0.49797219038009644, "learning_rate": 9.96973215684834e-05, "loss": 0.069, "step": 8330 }, { "grad_norm": 0.42630958557128906, "learning_rate": 9.969550225344513e-05, "loss": 0.0749, "step": 8340 }, { "grad_norm": 0.35259100794792175, "learning_rate": 9.969367750378147e-05, "loss": 0.0757, "step": 8350 }, { "grad_norm": 0.4522119164466858, "learning_rate": 9.969184731969194e-05, "loss": 0.063, "step": 8360 }, { "grad_norm": 0.43675553798675537, "learning_rate": 9.96900117013767e-05, "loss": 0.0568, "step": 8370 }, { "grad_norm": 0.4986533522605896, "learning_rate": 9.96881706490365e-05, "loss": 0.0714, "step": 8380 }, { "grad_norm": 0.7677559852600098, "learning_rate": 9.968632416287265e-05, "loss": 0.0601, "step": 8390 }, { "grad_norm": 0.5310751795768738, "learning_rate": 9.96844722430871e-05, "loss": 0.0699, "step": 8400 }, { "grad_norm": 0.5001355409622192, "learning_rate": 9.968261488988235e-05, "loss": 0.0702, "step": 8410 }, { "grad_norm": 0.4655325412750244, "learning_rate": 9.968075210346155e-05, "loss": 0.065, "step": 8420 }, { "grad_norm": 0.3915587067604065, "learning_rate": 9.967888388402839e-05, "loss": 0.0648, "step": 8430 }, { "grad_norm": 0.3708435297012329, "learning_rate": 9.967701023178717e-05, "loss": 0.063, "step": 8440 }, { "grad_norm": 0.7394161224365234, "learning_rate": 9.967513114694282e-05, "loss": 0.0718, "step": 8450 }, { "grad_norm": 0.4484173059463501, "learning_rate": 9.967324662970079e-05, "loss": 0.0652, "step": 8460 }, { "grad_norm": 0.5111478567123413, "learning_rate": 9.96713566802672e-05, "loss": 0.0696, "step": 8470 }, { "grad_norm": 0.588520348072052, "learning_rate": 9.966946129884873e-05, "loss": 0.0649, "step": 8480 }, { "grad_norm": 0.5216318964958191, "learning_rate": 9.966756048565265e-05, "loss": 0.0656, "step": 8490 }, { "grad_norm": 0.4770970046520233, "learning_rate": 9.966565424088681e-05, "loss": 0.0786, "step": 8500 }, { "grad_norm": 0.511369526386261, "learning_rate": 9.96637425647597e-05, "loss": 0.0711, "step": 8510 }, { "grad_norm": 0.30698511004447937, "learning_rate": 9.966182545748038e-05, "loss": 0.0574, "step": 8520 }, { "grad_norm": 0.6824235916137695, "learning_rate": 9.96599029192585e-05, "loss": 0.0662, "step": 8530 }, { "grad_norm": 0.4006899297237396, "learning_rate": 9.965797495030428e-05, "loss": 0.0605, "step": 8540 }, { "grad_norm": 0.4027666747570038, "learning_rate": 9.96560415508286e-05, "loss": 0.0673, "step": 8550 }, { "grad_norm": 0.3970046639442444, "learning_rate": 9.965410272104286e-05, "loss": 0.0664, "step": 8560 }, { "grad_norm": 0.8023781180381775, "learning_rate": 9.96521584611591e-05, "loss": 0.0922, "step": 8570 }, { "grad_norm": 0.47723010182380676, "learning_rate": 9.965020877138994e-05, "loss": 0.0721, "step": 8580 }, { "grad_norm": 0.46166321635246277, "learning_rate": 9.964825365194861e-05, "loss": 0.0666, "step": 8590 }, { "grad_norm": 0.4630085229873657, "learning_rate": 9.96462931030489e-05, "loss": 0.0671, "step": 8600 }, { "grad_norm": 0.43791690468788147, "learning_rate": 9.96443271249052e-05, "loss": 0.0641, "step": 8610 }, { "grad_norm": 0.4001566767692566, "learning_rate": 9.964235571773255e-05, "loss": 0.0692, "step": 8620 }, { "grad_norm": 0.3812680244445801, "learning_rate": 9.96403788817465e-05, "loss": 0.0696, "step": 8630 }, { "grad_norm": 0.5536967515945435, "learning_rate": 9.963839661716325e-05, "loss": 0.0703, "step": 8640 }, { "grad_norm": 0.3633366525173187, "learning_rate": 9.963640892419958e-05, "loss": 0.0739, "step": 8650 }, { "grad_norm": 0.502629280090332, "learning_rate": 9.963441580307286e-05, "loss": 0.07, "step": 8660 }, { "grad_norm": 0.7408014535903931, "learning_rate": 9.963241725400104e-05, "loss": 0.0757, "step": 8670 }, { "grad_norm": 0.4477762281894684, "learning_rate": 9.963041327720271e-05, "loss": 0.0707, "step": 8680 }, { "grad_norm": 0.9249809980392456, "learning_rate": 9.962840387289697e-05, "loss": 0.0765, "step": 8690 }, { "grad_norm": 0.675334095954895, "learning_rate": 9.962638904130363e-05, "loss": 0.0833, "step": 8700 }, { "grad_norm": 0.7110618352890015, "learning_rate": 9.962436878264298e-05, "loss": 0.0952, "step": 8710 }, { "grad_norm": 0.6178496479988098, "learning_rate": 9.962234309713598e-05, "loss": 0.0949, "step": 8720 }, { "grad_norm": 0.45574116706848145, "learning_rate": 9.962031198500414e-05, "loss": 0.062, "step": 8730 }, { "grad_norm": 0.46809402108192444, "learning_rate": 9.961827544646958e-05, "loss": 0.0728, "step": 8740 }, { "grad_norm": 0.7206579446792603, "learning_rate": 9.961623348175501e-05, "loss": 0.0668, "step": 8750 }, { "grad_norm": 0.5185737013816833, "learning_rate": 9.961418609108377e-05, "loss": 0.0714, "step": 8760 }, { "grad_norm": 0.687811017036438, "learning_rate": 9.961213327467971e-05, "loss": 0.0723, "step": 8770 }, { "grad_norm": 0.4285285174846649, "learning_rate": 9.961007503276736e-05, "loss": 0.0819, "step": 8780 }, { "grad_norm": 0.4620485305786133, "learning_rate": 9.960801136557179e-05, "loss": 0.0724, "step": 8790 }, { "grad_norm": 0.3784882426261902, "learning_rate": 9.960594227331866e-05, "loss": 0.0554, "step": 8800 }, { "grad_norm": 0.450941801071167, "learning_rate": 9.960386775623429e-05, "loss": 0.0686, "step": 8810 }, { "grad_norm": 0.8933266997337341, "learning_rate": 9.96017878145455e-05, "loss": 0.0649, "step": 8820 }, { "grad_norm": 0.4630638360977173, "learning_rate": 9.959970244847977e-05, "loss": 0.0696, "step": 8830 }, { "grad_norm": 0.39966827630996704, "learning_rate": 9.959761165826518e-05, "loss": 0.0586, "step": 8840 }, { "grad_norm": 0.6963910460472107, "learning_rate": 9.959551544413033e-05, "loss": 0.0728, "step": 8850 }, { "grad_norm": 0.4622602164745331, "learning_rate": 9.959341380630448e-05, "loss": 0.0774, "step": 8860 }, { "grad_norm": 0.47267431020736694, "learning_rate": 9.959130674501746e-05, "loss": 0.0795, "step": 8870 }, { "grad_norm": 0.5692779421806335, "learning_rate": 9.958919426049968e-05, "loss": 0.0833, "step": 8880 }, { "grad_norm": 0.6794158220291138, "learning_rate": 9.958707635298219e-05, "loss": 0.0838, "step": 8890 }, { "grad_norm": 0.5501949787139893, "learning_rate": 9.958495302269657e-05, "loss": 0.0781, "step": 8900 }, { "grad_norm": 0.348989337682724, "learning_rate": 9.958282426987503e-05, "loss": 0.0675, "step": 8910 }, { "grad_norm": 0.5469598174095154, "learning_rate": 9.95806900947504e-05, "loss": 0.0779, "step": 8920 }, { "grad_norm": 0.5588148236274719, "learning_rate": 9.957855049755604e-05, "loss": 0.0672, "step": 8930 }, { "grad_norm": 0.7323089241981506, "learning_rate": 9.957640547852593e-05, "loss": 0.0661, "step": 8940 }, { "grad_norm": 0.5045587420463562, "learning_rate": 9.957425503789466e-05, "loss": 0.0713, "step": 8950 }, { "grad_norm": 0.6254780888557434, "learning_rate": 9.957209917589738e-05, "loss": 0.0622, "step": 8960 }, { "grad_norm": 0.7001920938491821, "learning_rate": 9.956993789276987e-05, "loss": 0.0562, "step": 8970 }, { "grad_norm": 1.423056721687317, "learning_rate": 9.956777118874847e-05, "loss": 0.0716, "step": 8980 }, { "grad_norm": 0.8736857175827026, "learning_rate": 9.956559906407016e-05, "loss": 0.1102, "step": 8990 }, { "grad_norm": 0.5760020613670349, "learning_rate": 9.956342151897245e-05, "loss": 0.0923, "step": 9000 }, { "grad_norm": 0.5487590432167053, "learning_rate": 9.956123855369346e-05, "loss": 0.0704, "step": 9010 }, { "grad_norm": 0.5305182933807373, "learning_rate": 9.955905016847196e-05, "loss": 0.0677, "step": 9020 }, { "grad_norm": 0.44396328926086426, "learning_rate": 9.955685636354723e-05, "loss": 0.0853, "step": 9030 }, { "grad_norm": 0.5523348450660706, "learning_rate": 9.95546571391592e-05, "loss": 0.0735, "step": 9040 }, { "grad_norm": 0.5560973882675171, "learning_rate": 9.955245249554837e-05, "loss": 0.0663, "step": 9050 }, { "grad_norm": 0.6033662557601929, "learning_rate": 9.955024243295582e-05, "loss": 0.0785, "step": 9060 }, { "grad_norm": 0.4458974599838257, "learning_rate": 9.954802695162328e-05, "loss": 0.0698, "step": 9070 }, { "grad_norm": 0.40443262457847595, "learning_rate": 9.954580605179302e-05, "loss": 0.0583, "step": 9080 }, { "grad_norm": 0.4138258993625641, "learning_rate": 9.954357973370788e-05, "loss": 0.0859, "step": 9090 }, { "grad_norm": 0.4300827085971832, "learning_rate": 9.954134799761135e-05, "loss": 0.0882, "step": 9100 }, { "grad_norm": 0.5121161937713623, "learning_rate": 9.953911084374748e-05, "loss": 0.0691, "step": 9110 }, { "grad_norm": 0.4564792513847351, "learning_rate": 9.953686827236093e-05, "loss": 0.0579, "step": 9120 }, { "grad_norm": 0.4903402626514435, "learning_rate": 9.953462028369695e-05, "loss": 0.0814, "step": 9130 }, { "grad_norm": 0.4909287691116333, "learning_rate": 9.953236687800136e-05, "loss": 0.0778, "step": 9140 }, { "grad_norm": 0.485933780670166, "learning_rate": 9.95301080555206e-05, "loss": 0.0753, "step": 9150 }, { "grad_norm": 0.3931553065776825, "learning_rate": 9.952784381650171e-05, "loss": 0.0642, "step": 9160 }, { "grad_norm": 0.657849907875061, "learning_rate": 9.952557416119226e-05, "loss": 0.0794, "step": 9170 }, { "grad_norm": 0.3607873022556305, "learning_rate": 9.95232990898405e-05, "loss": 0.0723, "step": 9180 }, { "grad_norm": 0.6470350027084351, "learning_rate": 9.95210186026952e-05, "loss": 0.069, "step": 9190 }, { "grad_norm": 0.7934017181396484, "learning_rate": 9.951873270000576e-05, "loss": 0.0727, "step": 9200 }, { "grad_norm": 0.38532888889312744, "learning_rate": 9.951644138202216e-05, "loss": 0.0734, "step": 9210 }, { "grad_norm": 1.0992201566696167, "learning_rate": 9.951414464899498e-05, "loss": 0.071, "step": 9220 }, { "grad_norm": 0.5222402215003967, "learning_rate": 9.951184250117538e-05, "loss": 0.0701, "step": 9230 }, { "grad_norm": 0.47030699253082275, "learning_rate": 9.950953493881513e-05, "loss": 0.0899, "step": 9240 }, { "grad_norm": 0.3994715213775635, "learning_rate": 9.950722196216658e-05, "loss": 0.0708, "step": 9250 }, { "grad_norm": 0.44299978017807007, "learning_rate": 9.950490357148265e-05, "loss": 0.0694, "step": 9260 }, { "grad_norm": 0.46805158257484436, "learning_rate": 9.950257976701692e-05, "loss": 0.0704, "step": 9270 }, { "grad_norm": 0.3634122312068939, "learning_rate": 9.950025054902348e-05, "loss": 0.0683, "step": 9280 }, { "grad_norm": 0.4818827211856842, "learning_rate": 9.949791591775706e-05, "loss": 0.0822, "step": 9290 }, { "grad_norm": 0.4589516818523407, "learning_rate": 9.949557587347298e-05, "loss": 0.0821, "step": 9300 }, { "grad_norm": 0.5120392441749573, "learning_rate": 9.949323041642713e-05, "loss": 0.0611, "step": 9310 }, { "grad_norm": 0.3402548134326935, "learning_rate": 9.949087954687602e-05, "loss": 0.0849, "step": 9320 }, { "grad_norm": 0.35966572165489197, "learning_rate": 9.948852326507672e-05, "loss": 0.0761, "step": 9330 }, { "grad_norm": 0.5637399554252625, "learning_rate": 9.948616157128694e-05, "loss": 0.0645, "step": 9340 }, { "grad_norm": 0.3191740810871124, "learning_rate": 9.948379446576493e-05, "loss": 0.06, "step": 9350 }, { "grad_norm": 1.0081220865249634, "learning_rate": 9.948142194876952e-05, "loss": 0.0592, "step": 9360 }, { "grad_norm": 0.5673500895500183, "learning_rate": 9.947904402056024e-05, "loss": 0.0751, "step": 9370 }, { "grad_norm": 0.46685826778411865, "learning_rate": 9.947666068139708e-05, "loss": 0.0764, "step": 9380 }, { "grad_norm": 0.44081270694732666, "learning_rate": 9.947427193154071e-05, "loss": 0.0734, "step": 9390 }, { "grad_norm": 0.47667670249938965, "learning_rate": 9.947187777125233e-05, "loss": 0.0763, "step": 9400 }, { "grad_norm": 0.506192684173584, "learning_rate": 9.946947820079377e-05, "loss": 0.0729, "step": 9410 }, { "grad_norm": 0.7120671272277832, "learning_rate": 9.946707322042747e-05, "loss": 0.0855, "step": 9420 }, { "grad_norm": 0.38095125555992126, "learning_rate": 9.94646628304164e-05, "loss": 0.0643, "step": 9430 }, { "grad_norm": 0.4832693636417389, "learning_rate": 9.946224703102418e-05, "loss": 0.0655, "step": 9440 }, { "grad_norm": 0.3982566297054291, "learning_rate": 9.945982582251498e-05, "loss": 0.0603, "step": 9450 }, { "grad_norm": 0.3832383155822754, "learning_rate": 9.94573992051536e-05, "loss": 0.0638, "step": 9460 }, { "grad_norm": 0.460651159286499, "learning_rate": 9.94549671792054e-05, "loss": 0.0871, "step": 9470 }, { "grad_norm": 0.5287819504737854, "learning_rate": 9.945252974493635e-05, "loss": 0.0869, "step": 9480 }, { "grad_norm": 1.585832953453064, "learning_rate": 9.9450086902613e-05, "loss": 0.0681, "step": 9490 }, { "grad_norm": 0.47945845127105713, "learning_rate": 9.944763865250248e-05, "loss": 0.0716, "step": 9500 }, { "grad_norm": 0.8088494539260864, "learning_rate": 9.944518499487254e-05, "loss": 0.085, "step": 9510 }, { "grad_norm": 0.41041985154151917, "learning_rate": 9.944272592999151e-05, "loss": 0.0713, "step": 9520 }, { "grad_norm": 0.32957783341407776, "learning_rate": 9.94402614581283e-05, "loss": 0.0704, "step": 9530 }, { "grad_norm": 0.48526403307914734, "learning_rate": 9.943779157955244e-05, "loss": 0.0776, "step": 9540 }, { "grad_norm": 0.4335469901561737, "learning_rate": 9.943531629453403e-05, "loss": 0.0658, "step": 9550 }, { "grad_norm": 0.5319211483001709, "learning_rate": 9.943283560334375e-05, "loss": 0.0743, "step": 9560 }, { "grad_norm": 0.6040117144584656, "learning_rate": 9.943034950625288e-05, "loss": 0.087, "step": 9570 }, { "grad_norm": 0.4957962930202484, "learning_rate": 9.942785800353332e-05, "loss": 0.0792, "step": 9580 }, { "grad_norm": 0.4804360270500183, "learning_rate": 9.942536109545751e-05, "loss": 0.0652, "step": 9590 }, { "grad_norm": 0.47088947892189026, "learning_rate": 9.942285878229853e-05, "loss": 0.0801, "step": 9600 }, { "grad_norm": 0.3975810706615448, "learning_rate": 9.942035106433001e-05, "loss": 0.0789, "step": 9610 }, { "grad_norm": 0.6893861293792725, "learning_rate": 9.94178379418262e-05, "loss": 0.0779, "step": 9620 }, { "grad_norm": 0.33982741832733154, "learning_rate": 9.941531941506194e-05, "loss": 0.0724, "step": 9630 }, { "grad_norm": 0.34846365451812744, "learning_rate": 9.941279548431263e-05, "loss": 0.0682, "step": 9640 }, { "grad_norm": 0.4755953252315521, "learning_rate": 9.941026614985431e-05, "loss": 0.0796, "step": 9650 }, { "grad_norm": 0.5015513896942139, "learning_rate": 9.940773141196357e-05, "loss": 0.0718, "step": 9660 }, { "grad_norm": 0.630063533782959, "learning_rate": 9.94051912709176e-05, "loss": 0.0743, "step": 9670 }, { "grad_norm": 0.574956476688385, "learning_rate": 9.940264572699421e-05, "loss": 0.0716, "step": 9680 }, { "grad_norm": 0.3626740276813507, "learning_rate": 9.940009478047174e-05, "loss": 0.0618, "step": 9690 }, { "grad_norm": 0.42495834827423096, "learning_rate": 9.939753843162918e-05, "loss": 0.0561, "step": 9700 }, { "grad_norm": 0.6246792674064636, "learning_rate": 9.939497668074609e-05, "loss": 0.0633, "step": 9710 }, { "grad_norm": 0.524569571018219, "learning_rate": 9.93924095281026e-05, "loss": 0.0641, "step": 9720 }, { "grad_norm": 0.4071705639362335, "learning_rate": 9.938983697397948e-05, "loss": 0.0793, "step": 9730 }, { "grad_norm": 0.44662967324256897, "learning_rate": 9.938725901865805e-05, "loss": 0.0686, "step": 9740 }, { "grad_norm": 0.391162633895874, "learning_rate": 9.93846756624202e-05, "loss": 0.072, "step": 9750 }, { "grad_norm": 0.4783325493335724, "learning_rate": 9.938208690554849e-05, "loss": 0.0687, "step": 9760 }, { "grad_norm": 0.44876158237457275, "learning_rate": 9.9379492748326e-05, "loss": 0.0573, "step": 9770 }, { "grad_norm": 0.36829620599746704, "learning_rate": 9.937689319103641e-05, "loss": 0.0691, "step": 9780 }, { "grad_norm": 0.7129588723182678, "learning_rate": 9.937428823396404e-05, "loss": 0.0685, "step": 9790 }, { "grad_norm": 0.5918910503387451, "learning_rate": 9.937167787739372e-05, "loss": 0.0704, "step": 9800 }, { "grad_norm": 0.45085304975509644, "learning_rate": 9.936906212161095e-05, "loss": 0.0732, "step": 9810 }, { "grad_norm": 0.5915669202804565, "learning_rate": 9.936644096690176e-05, "loss": 0.0758, "step": 9820 }, { "grad_norm": 0.4759649932384491, "learning_rate": 9.936381441355282e-05, "loss": 0.0701, "step": 9830 }, { "grad_norm": 0.3213818669319153, "learning_rate": 9.936118246185136e-05, "loss": 0.0824, "step": 9840 }, { "grad_norm": 0.3625492453575134, "learning_rate": 9.935854511208518e-05, "loss": 0.0691, "step": 9850 }, { "grad_norm": 0.39841610193252563, "learning_rate": 9.935590236454272e-05, "loss": 0.05, "step": 9860 }, { "grad_norm": 0.39654457569122314, "learning_rate": 9.935325421951298e-05, "loss": 0.0741, "step": 9870 }, { "grad_norm": 0.8242678642272949, "learning_rate": 9.935060067728557e-05, "loss": 0.0582, "step": 9880 }, { "grad_norm": 0.4466853737831116, "learning_rate": 9.934794173815067e-05, "loss": 0.0621, "step": 9890 }, { "grad_norm": 0.46167439222335815, "learning_rate": 9.934527740239906e-05, "loss": 0.0613, "step": 9900 }, { "grad_norm": 0.319099098443985, "learning_rate": 9.934260767032209e-05, "loss": 0.0524, "step": 9910 }, { "grad_norm": 0.4477906823158264, "learning_rate": 9.933993254221172e-05, "loss": 0.0598, "step": 9920 }, { "grad_norm": 0.36945632100105286, "learning_rate": 9.933725201836053e-05, "loss": 0.0623, "step": 9930 }, { "grad_norm": 0.43428388237953186, "learning_rate": 9.933456609906162e-05, "loss": 0.0517, "step": 9940 }, { "grad_norm": 0.4370114207267761, "learning_rate": 9.933187478460875e-05, "loss": 0.052, "step": 9950 }, { "grad_norm": 0.30784133076667786, "learning_rate": 9.93291780752962e-05, "loss": 0.0552, "step": 9960 }, { "grad_norm": 0.4063231945037842, "learning_rate": 9.932647597141893e-05, "loss": 0.0685, "step": 9970 }, { "grad_norm": 0.34557944536209106, "learning_rate": 9.932376847327239e-05, "loss": 0.0704, "step": 9980 }, { "grad_norm": 0.8614240884780884, "learning_rate": 9.932105558115268e-05, "loss": 0.0732, "step": 9990 }, { "grad_norm": 0.5945726633071899, "learning_rate": 9.931833729535651e-05, "loss": 0.0747, "step": 10000 }, { "grad_norm": 0.4038606584072113, "learning_rate": 9.931561361618111e-05, "loss": 0.0704, "step": 10010 }, { "grad_norm": 0.4365256726741791, "learning_rate": 9.931288454392435e-05, "loss": 0.0641, "step": 10020 }, { "grad_norm": 0.31108465790748596, "learning_rate": 9.931015007888467e-05, "loss": 0.0526, "step": 10030 }, { "grad_norm": 0.4429807662963867, "learning_rate": 9.930741022136112e-05, "loss": 0.0807, "step": 10040 }, { "grad_norm": 0.38073021173477173, "learning_rate": 9.930466497165333e-05, "loss": 0.0642, "step": 10050 }, { "grad_norm": 0.4381420612335205, "learning_rate": 9.93019143300615e-05, "loss": 0.0609, "step": 10060 }, { "grad_norm": 0.36821651458740234, "learning_rate": 9.929915829688644e-05, "loss": 0.0637, "step": 10070 }, { "grad_norm": 0.4994083046913147, "learning_rate": 9.929639687242955e-05, "loss": 0.0627, "step": 10080 }, { "grad_norm": 0.5242270827293396, "learning_rate": 9.929363005699281e-05, "loss": 0.0622, "step": 10090 }, { "grad_norm": 0.5033246278762817, "learning_rate": 9.92908578508788e-05, "loss": 0.0727, "step": 10100 }, { "grad_norm": 0.6131588220596313, "learning_rate": 9.928808025439069e-05, "loss": 0.0584, "step": 10110 }, { "grad_norm": 0.722909152507782, "learning_rate": 9.928529726783223e-05, "loss": 0.0721, "step": 10120 }, { "grad_norm": 0.4713866710662842, "learning_rate": 9.928250889150774e-05, "loss": 0.0682, "step": 10130 }, { "grad_norm": 0.32492348551750183, "learning_rate": 9.92797151257222e-05, "loss": 0.0679, "step": 10140 }, { "grad_norm": 0.8099474906921387, "learning_rate": 9.927691597078108e-05, "loss": 0.0778, "step": 10150 }, { "grad_norm": 0.4488596022129059, "learning_rate": 9.927411142699053e-05, "loss": 0.0765, "step": 10160 }, { "grad_norm": 0.42184528708457947, "learning_rate": 9.927130149465725e-05, "loss": 0.0596, "step": 10170 }, { "grad_norm": 0.34589332342147827, "learning_rate": 9.92684861740885e-05, "loss": 0.0693, "step": 10180 }, { "grad_norm": 0.7128167748451233, "learning_rate": 9.926566546559217e-05, "loss": 0.0824, "step": 10190 }, { "grad_norm": 0.44404107332229614, "learning_rate": 9.926283936947673e-05, "loss": 0.0659, "step": 10200 }, { "grad_norm": 0.3349900543689728, "learning_rate": 9.926000788605126e-05, "loss": 0.0713, "step": 10210 }, { "grad_norm": 0.5594521164894104, "learning_rate": 9.92571710156254e-05, "loss": 0.0711, "step": 10220 }, { "grad_norm": 0.5016708970069885, "learning_rate": 9.925432875850936e-05, "loss": 0.0734, "step": 10230 }, { "grad_norm": 0.8161987066268921, "learning_rate": 9.925148111501396e-05, "loss": 0.0691, "step": 10240 }, { "grad_norm": 0.715812087059021, "learning_rate": 9.924862808545066e-05, "loss": 0.0636, "step": 10250 }, { "grad_norm": 0.4963158965110779, "learning_rate": 9.924576967013141e-05, "loss": 0.0657, "step": 10260 }, { "grad_norm": 0.4769311249256134, "learning_rate": 9.924290586936887e-05, "loss": 0.0648, "step": 10270 }, { "grad_norm": 0.333125501871109, "learning_rate": 9.924003668347614e-05, "loss": 0.0584, "step": 10280 }, { "grad_norm": 0.39715445041656494, "learning_rate": 9.923716211276704e-05, "loss": 0.0758, "step": 10290 }, { "grad_norm": 0.3806394934654236, "learning_rate": 9.923428215755594e-05, "loss": 0.0602, "step": 10300 }, { "grad_norm": 0.5005134344100952, "learning_rate": 9.923139681815775e-05, "loss": 0.057, "step": 10310 }, { "grad_norm": 0.4791213274002075, "learning_rate": 9.922850609488801e-05, "loss": 0.0642, "step": 10320 }, { "grad_norm": 0.5762342810630798, "learning_rate": 9.922560998806287e-05, "loss": 0.0672, "step": 10330 }, { "grad_norm": 0.6119033694267273, "learning_rate": 9.922270849799905e-05, "loss": 0.0762, "step": 10340 }, { "grad_norm": 0.5471627116203308, "learning_rate": 9.92198016250138e-05, "loss": 0.0675, "step": 10350 }, { "grad_norm": 0.455251544713974, "learning_rate": 9.921688936942506e-05, "loss": 0.0823, "step": 10360 }, { "grad_norm": 0.4056386649608612, "learning_rate": 9.921397173155129e-05, "loss": 0.056, "step": 10370 }, { "grad_norm": 0.45116308331489563, "learning_rate": 9.921104871171157e-05, "loss": 0.0766, "step": 10380 }, { "grad_norm": 0.46664947271347046, "learning_rate": 9.920812031022554e-05, "loss": 0.0612, "step": 10390 }, { "grad_norm": 0.5969527363777161, "learning_rate": 9.920518652741348e-05, "loss": 0.0688, "step": 10400 }, { "grad_norm": 0.3688623607158661, "learning_rate": 9.920224736359618e-05, "loss": 0.0682, "step": 10410 }, { "grad_norm": 0.3904699385166168, "learning_rate": 9.91993028190951e-05, "loss": 0.0582, "step": 10420 }, { "grad_norm": 0.45626261830329895, "learning_rate": 9.919635289423222e-05, "loss": 0.0577, "step": 10430 }, { "grad_norm": 0.38524290919303894, "learning_rate": 9.919339758933015e-05, "loss": 0.0592, "step": 10440 }, { "grad_norm": 0.34790611267089844, "learning_rate": 9.919043690471209e-05, "loss": 0.069, "step": 10450 }, { "grad_norm": 0.32862940430641174, "learning_rate": 9.91874708407018e-05, "loss": 0.0655, "step": 10460 }, { "grad_norm": 0.3448886275291443, "learning_rate": 9.918449939762367e-05, "loss": 0.0642, "step": 10470 }, { "grad_norm": 0.45658835768699646, "learning_rate": 9.91815225758026e-05, "loss": 0.0579, "step": 10480 }, { "grad_norm": 0.5057335495948792, "learning_rate": 9.917854037556419e-05, "loss": 0.0523, "step": 10490 }, { "grad_norm": 0.33902889490127563, "learning_rate": 9.917555279723454e-05, "loss": 0.0562, "step": 10500 }, { "grad_norm": 0.4347708225250244, "learning_rate": 9.917255984114036e-05, "loss": 0.0649, "step": 10510 }, { "grad_norm": 0.33828049898147583, "learning_rate": 9.916956150760896e-05, "loss": 0.0615, "step": 10520 }, { "grad_norm": 0.2921109199523926, "learning_rate": 9.916655779696826e-05, "loss": 0.0593, "step": 10530 }, { "grad_norm": 0.4169688820838928, "learning_rate": 9.916354870954671e-05, "loss": 0.0727, "step": 10540 }, { "grad_norm": 0.35158899426460266, "learning_rate": 9.91605342456734e-05, "loss": 0.0521, "step": 10550 }, { "grad_norm": 0.5452243089675903, "learning_rate": 9.915751440567795e-05, "loss": 0.0569, "step": 10560 }, { "grad_norm": 0.3754562735557556, "learning_rate": 9.915448918989066e-05, "loss": 0.0654, "step": 10570 }, { "grad_norm": 0.5528964400291443, "learning_rate": 9.915145859864232e-05, "loss": 0.0831, "step": 10580 }, { "grad_norm": 0.6203136444091797, "learning_rate": 9.914842263226437e-05, "loss": 0.0644, "step": 10590 }, { "grad_norm": 0.38478249311447144, "learning_rate": 9.914538129108882e-05, "loss": 0.071, "step": 10600 }, { "grad_norm": 0.4223262071609497, "learning_rate": 9.914233457544825e-05, "loss": 0.0804, "step": 10610 }, { "grad_norm": 0.5472383499145508, "learning_rate": 9.913928248567586e-05, "loss": 0.0764, "step": 10620 }, { "grad_norm": 0.44723832607269287, "learning_rate": 9.913622502210542e-05, "loss": 0.0583, "step": 10630 }, { "grad_norm": 0.3424271047115326, "learning_rate": 9.913316218507128e-05, "loss": 0.0756, "step": 10640 }, { "grad_norm": 0.7385749220848083, "learning_rate": 9.91300939749084e-05, "loss": 0.066, "step": 10650 }, { "grad_norm": 0.4604965150356293, "learning_rate": 9.91270203919523e-05, "loss": 0.0642, "step": 10660 }, { "grad_norm": 0.5124915838241577, "learning_rate": 9.912394143653912e-05, "loss": 0.0793, "step": 10670 }, { "grad_norm": 0.4226537346839905, "learning_rate": 9.912085710900555e-05, "loss": 0.056, "step": 10680 }, { "grad_norm": 0.5210351347923279, "learning_rate": 9.911776740968892e-05, "loss": 0.0576, "step": 10690 }, { "grad_norm": 0.47203755378723145, "learning_rate": 9.911467233892709e-05, "loss": 0.0767, "step": 10700 }, { "grad_norm": 0.3983309268951416, "learning_rate": 9.911157189705853e-05, "loss": 0.0626, "step": 10710 }, { "grad_norm": 0.5816956758499146, "learning_rate": 9.910846608442229e-05, "loss": 0.0591, "step": 10720 }, { "grad_norm": 0.4134751856327057, "learning_rate": 9.910535490135805e-05, "loss": 0.0558, "step": 10730 }, { "grad_norm": 0.35905808210372925, "learning_rate": 9.910223834820603e-05, "loss": 0.0548, "step": 10740 }, { "grad_norm": 0.3068734407424927, "learning_rate": 9.909911642530703e-05, "loss": 0.0697, "step": 10750 }, { "grad_norm": 0.3676179051399231, "learning_rate": 9.909598913300249e-05, "loss": 0.0791, "step": 10760 }, { "grad_norm": 0.7117173671722412, "learning_rate": 9.909285647163438e-05, "loss": 0.0785, "step": 10770 }, { "grad_norm": 0.45324358344078064, "learning_rate": 9.908971844154531e-05, "loss": 0.0586, "step": 10780 }, { "grad_norm": 0.45409834384918213, "learning_rate": 9.908657504307843e-05, "loss": 0.0639, "step": 10790 }, { "grad_norm": 0.5484837889671326, "learning_rate": 9.908342627657751e-05, "loss": 0.0655, "step": 10800 }, { "grad_norm": 0.3462132215499878, "learning_rate": 9.908027214238689e-05, "loss": 0.0512, "step": 10810 }, { "grad_norm": 0.5369640588760376, "learning_rate": 9.90771126408515e-05, "loss": 0.0656, "step": 10820 }, { "grad_norm": 0.3557003140449524, "learning_rate": 9.907394777231685e-05, "loss": 0.0565, "step": 10830 }, { "grad_norm": 0.39115607738494873, "learning_rate": 9.907077753712905e-05, "loss": 0.0759, "step": 10840 }, { "grad_norm": 0.482492059469223, "learning_rate": 9.906760193563482e-05, "loss": 0.0597, "step": 10850 }, { "grad_norm": 0.4114564061164856, "learning_rate": 9.906442096818139e-05, "loss": 0.0502, "step": 10860 }, { "grad_norm": 0.49010705947875977, "learning_rate": 9.906123463511665e-05, "loss": 0.0674, "step": 10870 }, { "grad_norm": 0.43242231011390686, "learning_rate": 9.905804293678907e-05, "loss": 0.0651, "step": 10880 }, { "grad_norm": 0.5369090437889099, "learning_rate": 9.905484587354766e-05, "loss": 0.0651, "step": 10890 }, { "grad_norm": 0.45517340302467346, "learning_rate": 9.905164344574205e-05, "loss": 0.073, "step": 10900 }, { "grad_norm": 0.47868549823760986, "learning_rate": 9.904843565372248e-05, "loss": 0.067, "step": 10910 }, { "grad_norm": 0.665456235408783, "learning_rate": 9.904522249783972e-05, "loss": 0.0779, "step": 10920 }, { "grad_norm": 0.3766871392726898, "learning_rate": 9.904200397844517e-05, "loss": 0.0794, "step": 10930 }, { "grad_norm": 0.39065834879875183, "learning_rate": 9.903878009589078e-05, "loss": 0.0762, "step": 10940 }, { "grad_norm": 0.7962613701820374, "learning_rate": 9.903555085052915e-05, "loss": 0.0538, "step": 10950 }, { "grad_norm": 0.299896240234375, "learning_rate": 9.903231624271338e-05, "loss": 0.0639, "step": 10960 }, { "grad_norm": 0.49555981159210205, "learning_rate": 9.902907627279724e-05, "loss": 0.0644, "step": 10970 }, { "grad_norm": 0.2890872359275818, "learning_rate": 9.902583094113504e-05, "loss": 0.0497, "step": 10980 }, { "grad_norm": 0.5033740401268005, "learning_rate": 9.902258024808168e-05, "loss": 0.0616, "step": 10990 }, { "grad_norm": 0.3028638958930969, "learning_rate": 9.901932419399264e-05, "loss": 0.0656, "step": 11000 }, { "grad_norm": 0.4422679841518402, "learning_rate": 9.9016062779224e-05, "loss": 0.0586, "step": 11010 }, { "grad_norm": 0.4598238170146942, "learning_rate": 9.901279600413242e-05, "loss": 0.0549, "step": 11020 }, { "grad_norm": 0.45095035433769226, "learning_rate": 9.900952386907518e-05, "loss": 0.06, "step": 11030 }, { "grad_norm": 0.4989805519580841, "learning_rate": 9.90062463744101e-05, "loss": 0.0677, "step": 11040 }, { "grad_norm": 0.42285090684890747, "learning_rate": 9.900296352049558e-05, "loss": 0.057, "step": 11050 }, { "grad_norm": 0.4211635887622833, "learning_rate": 9.899967530769065e-05, "loss": 0.0647, "step": 11060 }, { "grad_norm": 0.5155559182167053, "learning_rate": 9.899638173635489e-05, "loss": 0.052, "step": 11070 }, { "grad_norm": 0.37950199842453003, "learning_rate": 9.899308280684849e-05, "loss": 0.0669, "step": 11080 }, { "grad_norm": 0.5209943056106567, "learning_rate": 9.898977851953222e-05, "loss": 0.0558, "step": 11090 }, { "grad_norm": 0.28757229447364807, "learning_rate": 9.898646887476741e-05, "loss": 0.0698, "step": 11100 }, { "grad_norm": 0.4942101538181305, "learning_rate": 9.898315387291603e-05, "loss": 0.0608, "step": 11110 }, { "grad_norm": 0.4656786620616913, "learning_rate": 9.89798335143406e-05, "loss": 0.0673, "step": 11120 }, { "grad_norm": 0.40266069769859314, "learning_rate": 9.897650779940419e-05, "loss": 0.0728, "step": 11130 }, { "grad_norm": 0.3714785575866699, "learning_rate": 9.897317672847054e-05, "loss": 0.0625, "step": 11140 }, { "grad_norm": 0.5611922144889832, "learning_rate": 9.89698403019039e-05, "loss": 0.0751, "step": 11150 }, { "grad_norm": 0.595370888710022, "learning_rate": 9.896649852006917e-05, "loss": 0.0662, "step": 11160 }, { "grad_norm": 0.38752856850624084, "learning_rate": 9.896315138333177e-05, "loss": 0.06, "step": 11170 }, { "grad_norm": 0.4064400792121887, "learning_rate": 9.895979889205774e-05, "loss": 0.0577, "step": 11180 }, { "grad_norm": 0.3791373372077942, "learning_rate": 9.895644104661372e-05, "loss": 0.0491, "step": 11190 }, { "grad_norm": 0.3986258804798126, "learning_rate": 9.895307784736691e-05, "loss": 0.0621, "step": 11200 }, { "grad_norm": 0.4087057411670685, "learning_rate": 9.894970929468512e-05, "loss": 0.0563, "step": 11210 }, { "grad_norm": 0.4806055724620819, "learning_rate": 9.89463353889367e-05, "loss": 0.0539, "step": 11220 }, { "grad_norm": 0.34167325496673584, "learning_rate": 9.894295613049065e-05, "loss": 0.0765, "step": 11230 }, { "grad_norm": 0.33999764919281006, "learning_rate": 9.893957151971649e-05, "loss": 0.0621, "step": 11240 }, { "grad_norm": 0.351114422082901, "learning_rate": 9.893618155698436e-05, "loss": 0.0665, "step": 11250 }, { "grad_norm": 0.3808758556842804, "learning_rate": 9.8932786242665e-05, "loss": 0.0573, "step": 11260 }, { "grad_norm": 0.3177148699760437, "learning_rate": 9.89293855771297e-05, "loss": 0.065, "step": 11270 }, { "grad_norm": 0.40708109736442566, "learning_rate": 9.892597956075036e-05, "loss": 0.0681, "step": 11280 }, { "grad_norm": 0.38609614968299866, "learning_rate": 9.892256819389947e-05, "loss": 0.0536, "step": 11290 }, { "grad_norm": 0.5961116552352905, "learning_rate": 9.891915147695006e-05, "loss": 0.0723, "step": 11300 }, { "grad_norm": 0.43083643913269043, "learning_rate": 9.891572941027577e-05, "loss": 0.0563, "step": 11310 }, { "grad_norm": 0.4457092881202698, "learning_rate": 9.89123019942509e-05, "loss": 0.0568, "step": 11320 }, { "grad_norm": 0.31288689374923706, "learning_rate": 9.89088692292502e-05, "loss": 0.0473, "step": 11330 }, { "grad_norm": 0.38478922843933105, "learning_rate": 9.89054311156491e-05, "loss": 0.0624, "step": 11340 }, { "grad_norm": 0.3952314555644989, "learning_rate": 9.890198765382357e-05, "loss": 0.078, "step": 11350 }, { "grad_norm": 0.6460043787956238, "learning_rate": 9.889853884415021e-05, "loss": 0.0568, "step": 11360 }, { "grad_norm": 0.25010183453559875, "learning_rate": 9.889508468700614e-05, "loss": 0.0502, "step": 11370 }, { "grad_norm": 0.383219450712204, "learning_rate": 9.889162518276915e-05, "loss": 0.0504, "step": 11380 }, { "grad_norm": 0.6245751976966858, "learning_rate": 9.888816033181752e-05, "loss": 0.0654, "step": 11390 }, { "grad_norm": 0.39747485518455505, "learning_rate": 9.888469013453018e-05, "loss": 0.071, "step": 11400 }, { "grad_norm": 0.473581999540329, "learning_rate": 9.888121459128663e-05, "loss": 0.0545, "step": 11410 }, { "grad_norm": 0.633489727973938, "learning_rate": 9.887773370246693e-05, "loss": 0.0675, "step": 11420 }, { "grad_norm": 0.5172019600868225, "learning_rate": 9.887424746845177e-05, "loss": 0.0691, "step": 11430 }, { "grad_norm": 0.5112085342407227, "learning_rate": 9.887075588962239e-05, "loss": 0.0537, "step": 11440 }, { "grad_norm": 0.648453414440155, "learning_rate": 9.88672589663606e-05, "loss": 0.0555, "step": 11450 }, { "grad_norm": 0.36550915241241455, "learning_rate": 9.886375669904886e-05, "loss": 0.063, "step": 11460 }, { "grad_norm": 1.0860317945480347, "learning_rate": 9.886024908807014e-05, "loss": 0.0612, "step": 11470 }, { "grad_norm": 0.5288704633712769, "learning_rate": 9.885673613380806e-05, "loss": 0.0598, "step": 11480 }, { "grad_norm": 0.43293851613998413, "learning_rate": 9.885321783664676e-05, "loss": 0.0752, "step": 11490 }, { "grad_norm": 0.621856153011322, "learning_rate": 9.884969419697101e-05, "loss": 0.0748, "step": 11500 }, { "grad_norm": 0.5570642948150635, "learning_rate": 9.884616521516614e-05, "loss": 0.0682, "step": 11510 }, { "grad_norm": 0.42367565631866455, "learning_rate": 9.88426308916181e-05, "loss": 0.0622, "step": 11520 }, { "grad_norm": 0.42665088176727295, "learning_rate": 9.883909122671335e-05, "loss": 0.0588, "step": 11530 }, { "grad_norm": 0.4431483745574951, "learning_rate": 9.883554622083904e-05, "loss": 0.0608, "step": 11540 }, { "grad_norm": 0.4484812319278717, "learning_rate": 9.88319958743828e-05, "loss": 0.0597, "step": 11550 }, { "grad_norm": 0.29409241676330566, "learning_rate": 9.882844018773291e-05, "loss": 0.0519, "step": 11560 }, { "grad_norm": 0.41244372725486755, "learning_rate": 9.882487916127823e-05, "loss": 0.0531, "step": 11570 }, { "grad_norm": 0.4007129371166229, "learning_rate": 9.882131279540815e-05, "loss": 0.0595, "step": 11580 }, { "grad_norm": 0.6317788362503052, "learning_rate": 9.881774109051271e-05, "loss": 0.0705, "step": 11590 }, { "grad_norm": 0.3290461003780365, "learning_rate": 9.881416404698252e-05, "loss": 0.0534, "step": 11600 }, { "grad_norm": 0.5181831121444702, "learning_rate": 9.881058166520873e-05, "loss": 0.0682, "step": 11610 }, { "grad_norm": 0.4630926549434662, "learning_rate": 9.880699394558311e-05, "loss": 0.0629, "step": 11620 }, { "grad_norm": 0.4456664025783539, "learning_rate": 9.880340088849801e-05, "loss": 0.0678, "step": 11630 }, { "grad_norm": 0.5213413834571838, "learning_rate": 9.879980249434637e-05, "loss": 0.0539, "step": 11640 }, { "grad_norm": 0.676836371421814, "learning_rate": 9.879619876352168e-05, "loss": 0.0648, "step": 11650 }, { "grad_norm": 0.4267866611480713, "learning_rate": 9.879258969641809e-05, "loss": 0.0509, "step": 11660 }, { "grad_norm": 0.34238824248313904, "learning_rate": 9.878897529343023e-05, "loss": 0.0488, "step": 11670 }, { "grad_norm": 0.4186929166316986, "learning_rate": 9.878535555495338e-05, "loss": 0.0552, "step": 11680 }, { "grad_norm": 0.6691641807556152, "learning_rate": 9.87817304813834e-05, "loss": 0.0632, "step": 11690 }, { "grad_norm": 0.5224377512931824, "learning_rate": 9.877810007311671e-05, "loss": 0.0513, "step": 11700 }, { "grad_norm": 0.5512650012969971, "learning_rate": 9.877446433055035e-05, "loss": 0.065, "step": 11710 }, { "grad_norm": 0.6531388759613037, "learning_rate": 9.877082325408191e-05, "loss": 0.0587, "step": 11720 }, { "grad_norm": 0.5379595160484314, "learning_rate": 9.876717684410954e-05, "loss": 0.0589, "step": 11730 }, { "grad_norm": 0.38580965995788574, "learning_rate": 9.876352510103204e-05, "loss": 0.0526, "step": 11740 }, { "grad_norm": 0.4069638252258301, "learning_rate": 9.875986802524875e-05, "loss": 0.0634, "step": 11750 }, { "grad_norm": 0.4204708933830261, "learning_rate": 9.87562056171596e-05, "loss": 0.0565, "step": 11760 }, { "grad_norm": 1.003463625907898, "learning_rate": 9.875253787716511e-05, "loss": 0.0585, "step": 11770 }, { "grad_norm": 0.4304141104221344, "learning_rate": 9.874886480566637e-05, "loss": 0.0446, "step": 11780 }, { "grad_norm": 0.4068470895290375, "learning_rate": 9.874518640306507e-05, "loss": 0.0478, "step": 11790 }, { "grad_norm": 0.3990546464920044, "learning_rate": 9.874150266976347e-05, "loss": 0.0551, "step": 11800 }, { "grad_norm": 0.6879760026931763, "learning_rate": 9.873781360616443e-05, "loss": 0.0775, "step": 11810 }, { "grad_norm": 0.41714897751808167, "learning_rate": 9.873411921267137e-05, "loss": 0.0574, "step": 11820 }, { "grad_norm": 0.38168463110923767, "learning_rate": 9.873041948968829e-05, "loss": 0.0694, "step": 11830 }, { "grad_norm": 0.51066654920578, "learning_rate": 9.872671443761981e-05, "loss": 0.0579, "step": 11840 }, { "grad_norm": 0.48447972536087036, "learning_rate": 9.872300405687109e-05, "loss": 0.0586, "step": 11850 }, { "grad_norm": 0.3992230296134949, "learning_rate": 9.871928834784792e-05, "loss": 0.0565, "step": 11860 }, { "grad_norm": 0.31013625860214233, "learning_rate": 9.871556731095661e-05, "loss": 0.0462, "step": 11870 }, { "grad_norm": 0.39031240344047546, "learning_rate": 9.871184094660411e-05, "loss": 0.0648, "step": 11880 }, { "grad_norm": 0.3525523245334625, "learning_rate": 9.870810925519791e-05, "loss": 0.0523, "step": 11890 }, { "grad_norm": 0.4043876826763153, "learning_rate": 9.870437223714612e-05, "loss": 0.0532, "step": 11900 }, { "grad_norm": 0.8848881125450134, "learning_rate": 9.87006298928574e-05, "loss": 0.073, "step": 11910 }, { "grad_norm": 0.3092653155326843, "learning_rate": 9.869688222274103e-05, "loss": 0.058, "step": 11920 }, { "grad_norm": 0.351779043674469, "learning_rate": 9.869312922720681e-05, "loss": 0.0575, "step": 11930 }, { "grad_norm": 0.4076462984085083, "learning_rate": 9.868937090666521e-05, "loss": 0.0536, "step": 11940 }, { "grad_norm": 0.41229647397994995, "learning_rate": 9.86856072615272e-05, "loss": 0.0546, "step": 11950 }, { "grad_norm": 0.708653450012207, "learning_rate": 9.868183829220438e-05, "loss": 0.0512, "step": 11960 }, { "grad_norm": 0.3415394425392151, "learning_rate": 9.867806399910893e-05, "loss": 0.0501, "step": 11970 }, { "grad_norm": 0.3886515498161316, "learning_rate": 9.867428438265356e-05, "loss": 0.0546, "step": 11980 }, { "grad_norm": 0.3332265317440033, "learning_rate": 9.867049944325165e-05, "loss": 0.051, "step": 11990 }, { "grad_norm": 0.5547340512275696, "learning_rate": 9.86667091813171e-05, "loss": 0.0712, "step": 12000 }, { "grad_norm": 0.48410674929618835, "learning_rate": 9.866291359726438e-05, "loss": 0.0504, "step": 12010 }, { "grad_norm": 0.3844154179096222, "learning_rate": 9.865911269150861e-05, "loss": 0.0525, "step": 12020 }, { "grad_norm": 0.720989465713501, "learning_rate": 9.865530646446544e-05, "loss": 0.0914, "step": 12030 }, { "grad_norm": 0.5078722834587097, "learning_rate": 9.86514949165511e-05, "loss": 0.0761, "step": 12040 }, { "grad_norm": 0.27972081303596497, "learning_rate": 9.864767804818243e-05, "loss": 0.0601, "step": 12050 }, { "grad_norm": 0.32762596011161804, "learning_rate": 9.86438558597768e-05, "loss": 0.0624, "step": 12060 }, { "grad_norm": 0.5963965654373169, "learning_rate": 9.864002835175225e-05, "loss": 0.0729, "step": 12070 }, { "grad_norm": 0.4171138107776642, "learning_rate": 9.863619552452734e-05, "loss": 0.0535, "step": 12080 }, { "grad_norm": 0.6731846332550049, "learning_rate": 9.863235737852119e-05, "loss": 0.0572, "step": 12090 }, { "grad_norm": 0.5996744632720947, "learning_rate": 9.862851391415356e-05, "loss": 0.0698, "step": 12100 }, { "grad_norm": 0.3939007818698883, "learning_rate": 9.862466513184477e-05, "loss": 0.0647, "step": 12110 }, { "grad_norm": 0.4641454517841339, "learning_rate": 9.86208110320157e-05, "loss": 0.0545, "step": 12120 }, { "grad_norm": 0.5475053787231445, "learning_rate": 9.861695161508784e-05, "loss": 0.0711, "step": 12130 }, { "grad_norm": 0.47144800424575806, "learning_rate": 9.861308688148324e-05, "loss": 0.0644, "step": 12140 }, { "grad_norm": 0.6529697775840759, "learning_rate": 9.860921683162455e-05, "loss": 0.0621, "step": 12150 }, { "grad_norm": 0.4674578011035919, "learning_rate": 9.860534146593499e-05, "loss": 0.0628, "step": 12160 }, { "grad_norm": 0.344698041677475, "learning_rate": 9.860146078483836e-05, "loss": 0.0617, "step": 12170 }, { "grad_norm": 0.3810987174510956, "learning_rate": 9.859757478875905e-05, "loss": 0.0654, "step": 12180 }, { "grad_norm": 0.4822886884212494, "learning_rate": 9.859368347812204e-05, "loss": 0.0593, "step": 12190 }, { "grad_norm": 0.4001966714859009, "learning_rate": 9.858978685335285e-05, "loss": 0.0627, "step": 12200 }, { "grad_norm": 0.49793732166290283, "learning_rate": 9.858588491487763e-05, "loss": 0.0727, "step": 12210 }, { "grad_norm": 0.3719596862792969, "learning_rate": 9.858197766312308e-05, "loss": 0.0476, "step": 12220 }, { "grad_norm": 0.34818387031555176, "learning_rate": 9.857806509851649e-05, "loss": 0.0602, "step": 12230 }, { "grad_norm": 0.3701023459434509, "learning_rate": 9.857414722148574e-05, "loss": 0.0452, "step": 12240 }, { "grad_norm": 0.321217805147171, "learning_rate": 9.857022403245928e-05, "loss": 0.0569, "step": 12250 }, { "grad_norm": 0.5410432815551758, "learning_rate": 9.856629553186615e-05, "loss": 0.0578, "step": 12260 }, { "grad_norm": 0.4707779288291931, "learning_rate": 9.856236172013595e-05, "loss": 0.0626, "step": 12270 }, { "grad_norm": 0.3954041004180908, "learning_rate": 9.85584225976989e-05, "loss": 0.0583, "step": 12280 }, { "grad_norm": 0.428585022687912, "learning_rate": 9.855447816498575e-05, "loss": 0.063, "step": 12290 }, { "grad_norm": 0.3533538281917572, "learning_rate": 9.855052842242787e-05, "loss": 0.0556, "step": 12300 }, { "grad_norm": 0.491729736328125, "learning_rate": 9.85465733704572e-05, "loss": 0.0577, "step": 12310 }, { "grad_norm": 0.4175650477409363, "learning_rate": 9.854261300950624e-05, "loss": 0.0671, "step": 12320 }, { "grad_norm": 0.5202690362930298, "learning_rate": 9.853864734000813e-05, "loss": 0.0618, "step": 12330 }, { "grad_norm": 0.8392542600631714, "learning_rate": 9.85346763623965e-05, "loss": 0.0698, "step": 12340 }, { "grad_norm": 0.4029737114906311, "learning_rate": 9.853070007710564e-05, "loss": 0.0522, "step": 12350 }, { "grad_norm": 0.39825570583343506, "learning_rate": 9.85267184845704e-05, "loss": 0.0685, "step": 12360 }, { "grad_norm": 0.29875022172927856, "learning_rate": 9.852273158522616e-05, "loss": 0.0538, "step": 12370 }, { "grad_norm": 0.4861697554588318, "learning_rate": 9.851873937950896e-05, "loss": 0.0583, "step": 12380 }, { "grad_norm": 0.4331324100494385, "learning_rate": 9.851474186785537e-05, "loss": 0.0623, "step": 12390 }, { "grad_norm": 0.4963541030883789, "learning_rate": 9.851073905070254e-05, "loss": 0.057, "step": 12400 }, { "grad_norm": 0.6152244806289673, "learning_rate": 9.850673092848824e-05, "loss": 0.0576, "step": 12410 }, { "grad_norm": 0.3201034367084503, "learning_rate": 9.850271750165077e-05, "loss": 0.0472, "step": 12420 }, { "grad_norm": 0.5722818970680237, "learning_rate": 9.849869877062902e-05, "loss": 0.0615, "step": 12430 }, { "grad_norm": 0.4399576485157013, "learning_rate": 9.849467473586252e-05, "loss": 0.0566, "step": 12440 }, { "grad_norm": 0.4337202310562134, "learning_rate": 9.849064539779127e-05, "loss": 0.0621, "step": 12450 }, { "grad_norm": 0.6891626715660095, "learning_rate": 9.848661075685594e-05, "loss": 0.0674, "step": 12460 }, { "grad_norm": 0.36807674169540405, "learning_rate": 9.848257081349778e-05, "loss": 0.0686, "step": 12470 }, { "grad_norm": 0.7388619184494019, "learning_rate": 9.847852556815856e-05, "loss": 0.0736, "step": 12480 }, { "grad_norm": 0.3392692506313324, "learning_rate": 9.847447502128067e-05, "loss": 0.0603, "step": 12490 }, { "grad_norm": 0.6777485609054565, "learning_rate": 9.847041917330708e-05, "loss": 0.074, "step": 12500 }, { "grad_norm": 0.41447141766548157, "learning_rate": 9.846635802468132e-05, "loss": 0.0739, "step": 12510 }, { "grad_norm": 0.5509321689605713, "learning_rate": 9.84622915758475e-05, "loss": 0.0656, "step": 12520 }, { "grad_norm": 0.6372785568237305, "learning_rate": 9.845821982725034e-05, "loss": 0.0632, "step": 12530 }, { "grad_norm": 0.42557835578918457, "learning_rate": 9.845414277933514e-05, "loss": 0.0635, "step": 12540 }, { "grad_norm": 0.44779273867607117, "learning_rate": 9.845006043254771e-05, "loss": 0.0684, "step": 12550 }, { "grad_norm": 0.4238632023334503, "learning_rate": 9.844597278733451e-05, "loss": 0.053, "step": 12560 }, { "grad_norm": 0.3197530508041382, "learning_rate": 9.844187984414259e-05, "loss": 0.0583, "step": 12570 }, { "grad_norm": 0.26081761717796326, "learning_rate": 9.84377816034195e-05, "loss": 0.0654, "step": 12580 }, { "grad_norm": 0.4194890260696411, "learning_rate": 9.843367806561345e-05, "loss": 0.0655, "step": 12590 }, { "grad_norm": 0.46171340346336365, "learning_rate": 9.842956923117317e-05, "loss": 0.0582, "step": 12600 }, { "grad_norm": 0.30778706073760986, "learning_rate": 9.842545510054802e-05, "loss": 0.051, "step": 12610 }, { "grad_norm": 0.42330697178840637, "learning_rate": 9.842133567418792e-05, "loss": 0.0639, "step": 12620 }, { "grad_norm": 0.5142356753349304, "learning_rate": 9.841721095254333e-05, "loss": 0.0572, "step": 12630 }, { "grad_norm": 0.34035515785217285, "learning_rate": 9.841308093606537e-05, "loss": 0.0569, "step": 12640 }, { "grad_norm": 0.33369511365890503, "learning_rate": 9.840894562520565e-05, "loss": 0.052, "step": 12650 }, { "grad_norm": 0.3968375623226166, "learning_rate": 9.840480502041642e-05, "loss": 0.0504, "step": 12660 }, { "grad_norm": 0.5389741063117981, "learning_rate": 9.840065912215049e-05, "loss": 0.0597, "step": 12670 }, { "grad_norm": 0.7345328330993652, "learning_rate": 9.839650793086124e-05, "loss": 0.0698, "step": 12680 }, { "grad_norm": 0.586225688457489, "learning_rate": 9.839235144700265e-05, "loss": 0.064, "step": 12690 }, { "grad_norm": 0.5249252915382385, "learning_rate": 9.838818967102926e-05, "loss": 0.0628, "step": 12700 }, { "grad_norm": 0.4391450583934784, "learning_rate": 9.83840226033962e-05, "loss": 0.0517, "step": 12710 }, { "grad_norm": 0.4656866192817688, "learning_rate": 9.837985024455918e-05, "loss": 0.0575, "step": 12720 }, { "grad_norm": 0.3309883177280426, "learning_rate": 9.837567259497447e-05, "loss": 0.05, "step": 12730 }, { "grad_norm": 0.4233601689338684, "learning_rate": 9.837148965509894e-05, "loss": 0.063, "step": 12740 }, { "grad_norm": 0.5886220932006836, "learning_rate": 9.836730142539001e-05, "loss": 0.0537, "step": 12750 }, { "grad_norm": 0.4852317273616791, "learning_rate": 9.836310790630574e-05, "loss": 0.0689, "step": 12760 }, { "grad_norm": 0.43510228395462036, "learning_rate": 9.83589090983047e-05, "loss": 0.0652, "step": 12770 }, { "grad_norm": 0.578069269657135, "learning_rate": 9.835470500184605e-05, "loss": 0.0634, "step": 12780 }, { "grad_norm": 0.37945041060447693, "learning_rate": 9.835049561738957e-05, "loss": 0.0575, "step": 12790 }, { "grad_norm": 0.458603173494339, "learning_rate": 9.834628094539558e-05, "loss": 0.0511, "step": 12800 }, { "grad_norm": 0.3832399249076843, "learning_rate": 9.834206098632499e-05, "loss": 0.0544, "step": 12810 }, { "grad_norm": 0.358277291059494, "learning_rate": 9.833783574063931e-05, "loss": 0.0559, "step": 12820 }, { "grad_norm": 0.4879895746707916, "learning_rate": 9.833360520880058e-05, "loss": 0.0579, "step": 12830 }, { "grad_norm": 0.47430217266082764, "learning_rate": 9.832936939127144e-05, "loss": 0.0441, "step": 12840 }, { "grad_norm": 0.3175238370895386, "learning_rate": 9.832512828851515e-05, "loss": 0.0486, "step": 12850 }, { "grad_norm": 0.3962570130825043, "learning_rate": 9.832088190099546e-05, "loss": 0.0496, "step": 12860 }, { "grad_norm": 0.4295032322406769, "learning_rate": 9.831663022917679e-05, "loss": 0.0487, "step": 12870 }, { "grad_norm": 0.4427366852760315, "learning_rate": 9.831237327352407e-05, "loss": 0.0529, "step": 12880 }, { "grad_norm": 0.35465651750564575, "learning_rate": 9.830811103450286e-05, "loss": 0.0652, "step": 12890 }, { "grad_norm": 0.724989652633667, "learning_rate": 9.830384351257924e-05, "loss": 0.0829, "step": 12900 }, { "grad_norm": 0.6183024048805237, "learning_rate": 9.829957070821993e-05, "loss": 0.0637, "step": 12910 }, { "grad_norm": 0.3575976490974426, "learning_rate": 9.829529262189218e-05, "loss": 0.0495, "step": 12920 }, { "grad_norm": 0.3817242980003357, "learning_rate": 9.829100925406385e-05, "loss": 0.067, "step": 12930 }, { "grad_norm": 0.5861304402351379, "learning_rate": 9.828672060520333e-05, "loss": 0.0622, "step": 12940 }, { "grad_norm": 0.4683319330215454, "learning_rate": 9.828242667577966e-05, "loss": 0.0607, "step": 12950 }, { "grad_norm": 0.46359339356422424, "learning_rate": 9.82781274662624e-05, "loss": 0.0608, "step": 12960 }, { "grad_norm": 0.3442693054676056, "learning_rate": 9.82738229771217e-05, "loss": 0.0659, "step": 12970 }, { "grad_norm": 0.39198675751686096, "learning_rate": 9.826951320882829e-05, "loss": 0.0557, "step": 12980 }, { "grad_norm": 0.31898683309555054, "learning_rate": 9.826519816185351e-05, "loss": 0.0529, "step": 12990 }, { "grad_norm": 0.3848154544830322, "learning_rate": 9.826087783666921e-05, "loss": 0.0515, "step": 13000 }, { "grad_norm": 0.3552359342575073, "learning_rate": 9.825655223374787e-05, "loss": 0.0613, "step": 13010 }, { "grad_norm": 0.6147911548614502, "learning_rate": 9.825222135356253e-05, "loss": 0.0654, "step": 13020 }, { "grad_norm": 0.41712018847465515, "learning_rate": 9.82478851965868e-05, "loss": 0.062, "step": 13030 }, { "grad_norm": 0.3899552822113037, "learning_rate": 9.82435437632949e-05, "loss": 0.0496, "step": 13040 }, { "grad_norm": 0.24746856093406677, "learning_rate": 9.823919705416158e-05, "loss": 0.0523, "step": 13050 }, { "grad_norm": 0.6390419006347656, "learning_rate": 9.82348450696622e-05, "loss": 0.0674, "step": 13060 }, { "grad_norm": 0.35174426436424255, "learning_rate": 9.823048781027268e-05, "loss": 0.0769, "step": 13070 }, { "grad_norm": 0.613960862159729, "learning_rate": 9.822612527646953e-05, "loss": 0.055, "step": 13080 }, { "grad_norm": 0.3712557256221771, "learning_rate": 9.822175746872984e-05, "loss": 0.0489, "step": 13090 }, { "grad_norm": 0.3710648715496063, "learning_rate": 9.821738438753123e-05, "loss": 0.0597, "step": 13100 }, { "grad_norm": 0.5593953132629395, "learning_rate": 9.821300603335196e-05, "loss": 0.0598, "step": 13110 }, { "grad_norm": 0.34384047985076904, "learning_rate": 9.820862240667085e-05, "loss": 0.0549, "step": 13120 }, { "grad_norm": 0.3052879571914673, "learning_rate": 9.820423350796726e-05, "loss": 0.0543, "step": 13130 }, { "grad_norm": 0.555351197719574, "learning_rate": 9.819983933772118e-05, "loss": 0.0487, "step": 13140 }, { "grad_norm": 0.5073156952857971, "learning_rate": 9.819543989641314e-05, "loss": 0.0614, "step": 13150 }, { "grad_norm": 0.3871437907218933, "learning_rate": 9.819103518452423e-05, "loss": 0.0509, "step": 13160 }, { "grad_norm": 0.40674829483032227, "learning_rate": 9.818662520253618e-05, "loss": 0.0498, "step": 13170 }, { "grad_norm": 0.519288957118988, "learning_rate": 9.818220995093126e-05, "loss": 0.0605, "step": 13180 }, { "grad_norm": 0.3518896698951721, "learning_rate": 9.817778943019228e-05, "loss": 0.0572, "step": 13190 }, { "grad_norm": 0.40595391392707825, "learning_rate": 9.81733636408027e-05, "loss": 0.0525, "step": 13200 }, { "grad_norm": 0.6806679964065552, "learning_rate": 9.816893258324649e-05, "loss": 0.0532, "step": 13210 }, { "grad_norm": 0.3946031630039215, "learning_rate": 9.816449625800823e-05, "loss": 0.0514, "step": 13220 }, { "grad_norm": 0.3513343632221222, "learning_rate": 9.816005466557308e-05, "loss": 0.0496, "step": 13230 }, { "grad_norm": 0.4223090410232544, "learning_rate": 9.815560780642674e-05, "loss": 0.0602, "step": 13240 }, { "grad_norm": 0.4065419137477875, "learning_rate": 9.815115568105555e-05, "loss": 0.0471, "step": 13250 }, { "grad_norm": 0.45909085869789124, "learning_rate": 9.814669828994638e-05, "loss": 0.0587, "step": 13260 }, { "grad_norm": 0.44435545802116394, "learning_rate": 9.814223563358665e-05, "loss": 0.0481, "step": 13270 }, { "grad_norm": 0.33437108993530273, "learning_rate": 9.813776771246443e-05, "loss": 0.0559, "step": 13280 }, { "grad_norm": 0.49813312292099, "learning_rate": 9.813329452706829e-05, "loss": 0.0514, "step": 13290 }, { "grad_norm": 0.39809492230415344, "learning_rate": 9.812881607788744e-05, "loss": 0.0477, "step": 13300 }, { "grad_norm": 0.4358217418193817, "learning_rate": 9.812433236541163e-05, "loss": 0.0537, "step": 13310 }, { "grad_norm": 0.4440642297267914, "learning_rate": 9.811984339013116e-05, "loss": 0.0624, "step": 13320 }, { "grad_norm": 0.36938905715942383, "learning_rate": 9.811534915253698e-05, "loss": 0.051, "step": 13330 }, { "grad_norm": 0.3361254036426544, "learning_rate": 9.811084965312056e-05, "loss": 0.053, "step": 13340 }, { "grad_norm": 0.271318256855011, "learning_rate": 9.810634489237396e-05, "loss": 0.0411, "step": 13350 }, { "grad_norm": 0.5202323198318481, "learning_rate": 9.81018348707898e-05, "loss": 0.0561, "step": 13360 }, { "grad_norm": 0.410245418548584, "learning_rate": 9.809731958886131e-05, "loss": 0.0706, "step": 13370 }, { "grad_norm": 0.598802924156189, "learning_rate": 9.809279904708224e-05, "loss": 0.06, "step": 13380 }, { "grad_norm": 0.867315948009491, "learning_rate": 9.808827324594699e-05, "loss": 0.0721, "step": 13390 }, { "grad_norm": 0.42777055501937866, "learning_rate": 9.808374218595046e-05, "loss": 0.0715, "step": 13400 }, { "grad_norm": 0.5358802676200867, "learning_rate": 9.80792058675882e-05, "loss": 0.0538, "step": 13410 }, { "grad_norm": 0.6207070350646973, "learning_rate": 9.807466429135627e-05, "loss": 0.0651, "step": 13420 }, { "grad_norm": 0.2645619511604309, "learning_rate": 9.807011745775132e-05, "loss": 0.0474, "step": 13430 }, { "grad_norm": 0.3504679203033447, "learning_rate": 9.806556536727061e-05, "loss": 0.066, "step": 13440 }, { "grad_norm": 0.49302536249160767, "learning_rate": 9.806100802041193e-05, "loss": 0.0665, "step": 13450 }, { "grad_norm": 0.7572987675666809, "learning_rate": 9.805644541767368e-05, "loss": 0.0695, "step": 13460 }, { "grad_norm": 0.6660945415496826, "learning_rate": 9.805187755955478e-05, "loss": 0.0655, "step": 13470 }, { "grad_norm": 0.5798060894012451, "learning_rate": 9.804730444655483e-05, "loss": 0.0749, "step": 13480 }, { "grad_norm": 0.9509345293045044, "learning_rate": 9.804272607917388e-05, "loss": 0.0632, "step": 13490 }, { "grad_norm": 0.5590643286705017, "learning_rate": 9.803814245791265e-05, "loss": 0.0612, "step": 13500 }, { "grad_norm": 0.5799415707588196, "learning_rate": 9.803355358327239e-05, "loss": 0.0681, "step": 13510 }, { "grad_norm": 0.2966032922267914, "learning_rate": 9.802895945575492e-05, "loss": 0.0677, "step": 13520 }, { "grad_norm": 0.6816831231117249, "learning_rate": 9.802436007586266e-05, "loss": 0.0757, "step": 13530 }, { "grad_norm": 0.4274328052997589, "learning_rate": 9.801975544409858e-05, "loss": 0.0588, "step": 13540 }, { "grad_norm": 0.5154897570610046, "learning_rate": 9.801514556096625e-05, "loss": 0.0628, "step": 13550 }, { "grad_norm": 0.372050940990448, "learning_rate": 9.801053042696977e-05, "loss": 0.0615, "step": 13560 }, { "grad_norm": 0.3366428017616272, "learning_rate": 9.800591004261388e-05, "loss": 0.0578, "step": 13570 }, { "grad_norm": 0.3371473550796509, "learning_rate": 9.800128440840385e-05, "loss": 0.0697, "step": 13580 }, { "grad_norm": 0.445989727973938, "learning_rate": 9.799665352484552e-05, "loss": 0.0775, "step": 13590 }, { "grad_norm": 0.4128187894821167, "learning_rate": 9.799201739244532e-05, "loss": 0.078, "step": 13600 }, { "grad_norm": 0.41151684522628784, "learning_rate": 9.798737601171025e-05, "loss": 0.0645, "step": 13610 }, { "grad_norm": 0.3949223458766937, "learning_rate": 9.79827293831479e-05, "loss": 0.0748, "step": 13620 }, { "grad_norm": 1.2170027494430542, "learning_rate": 9.797807750726638e-05, "loss": 0.0554, "step": 13630 }, { "grad_norm": 0.31556642055511475, "learning_rate": 9.797342038457446e-05, "loss": 0.0602, "step": 13640 }, { "grad_norm": 0.4486563801765442, "learning_rate": 9.796875801558141e-05, "loss": 0.0748, "step": 13650 }, { "grad_norm": 0.533737301826477, "learning_rate": 9.79640904007971e-05, "loss": 0.0662, "step": 13660 }, { "grad_norm": 0.4724898636341095, "learning_rate": 9.795941754073199e-05, "loss": 0.0673, "step": 13670 }, { "grad_norm": 1.2554608583450317, "learning_rate": 9.795473943589705e-05, "loss": 0.0684, "step": 13680 }, { "grad_norm": 0.40953415632247925, "learning_rate": 9.795005608680394e-05, "loss": 0.0571, "step": 13690 }, { "grad_norm": 0.5488870143890381, "learning_rate": 9.794536749396477e-05, "loss": 0.0592, "step": 13700 }, { "grad_norm": 0.35869964957237244, "learning_rate": 9.79406736578923e-05, "loss": 0.0591, "step": 13710 }, { "grad_norm": 0.34553062915802, "learning_rate": 9.793597457909984e-05, "loss": 0.0574, "step": 13720 }, { "grad_norm": 0.6885492205619812, "learning_rate": 9.793127025810127e-05, "loss": 0.0621, "step": 13730 }, { "grad_norm": 0.7285597920417786, "learning_rate": 9.792656069541104e-05, "loss": 0.0694, "step": 13740 }, { "grad_norm": 0.3110388219356537, "learning_rate": 9.79218458915442e-05, "loss": 0.0581, "step": 13750 }, { "grad_norm": 1.2052943706512451, "learning_rate": 9.791712584701634e-05, "loss": 0.0737, "step": 13760 }, { "grad_norm": 0.6509580016136169, "learning_rate": 9.791240056234364e-05, "loss": 0.0616, "step": 13770 }, { "grad_norm": 0.4334964454174042, "learning_rate": 9.790767003804283e-05, "loss": 0.0602, "step": 13780 }, { "grad_norm": 0.4064197242259979, "learning_rate": 9.790293427463126e-05, "loss": 0.054, "step": 13790 }, { "grad_norm": 0.33906465768814087, "learning_rate": 9.789819327262684e-05, "loss": 0.0681, "step": 13800 }, { "grad_norm": 0.5630972385406494, "learning_rate": 9.7893447032548e-05, "loss": 0.0704, "step": 13810 }, { "grad_norm": 0.36823031306266785, "learning_rate": 9.78886955549138e-05, "loss": 0.0647, "step": 13820 }, { "grad_norm": 0.45277518033981323, "learning_rate": 9.788393884024387e-05, "loss": 0.0653, "step": 13830 }, { "grad_norm": 0.3909164071083069, "learning_rate": 9.787917688905836e-05, "loss": 0.0744, "step": 13840 }, { "grad_norm": 0.3859253525733948, "learning_rate": 9.787440970187807e-05, "loss": 0.0552, "step": 13850 }, { "grad_norm": 0.3958010971546173, "learning_rate": 9.786963727922429e-05, "loss": 0.0562, "step": 13860 }, { "grad_norm": 0.5804046988487244, "learning_rate": 9.786485962161897e-05, "loss": 0.0572, "step": 13870 }, { "grad_norm": 0.37955525517463684, "learning_rate": 9.786007672958455e-05, "loss": 0.0481, "step": 13880 }, { "grad_norm": 0.4925856292247772, "learning_rate": 9.78552886036441e-05, "loss": 0.0694, "step": 13890 }, { "grad_norm": 0.7158657908439636, "learning_rate": 9.785049524432124e-05, "loss": 0.0523, "step": 13900 }, { "grad_norm": 0.5711793899536133, "learning_rate": 9.784569665214016e-05, "loss": 0.0505, "step": 13910 }, { "grad_norm": 0.48664167523384094, "learning_rate": 9.784089282762563e-05, "loss": 0.06, "step": 13920 }, { "grad_norm": 0.34221068024635315, "learning_rate": 9.7836083771303e-05, "loss": 0.0565, "step": 13930 }, { "grad_norm": 0.5320858359336853, "learning_rate": 9.783126948369817e-05, "loss": 0.0569, "step": 13940 }, { "grad_norm": 0.30022287368774414, "learning_rate": 9.78264499653376e-05, "loss": 0.0575, "step": 13950 }, { "grad_norm": 0.6571076512336731, "learning_rate": 9.782162521674838e-05, "loss": 0.0698, "step": 13960 }, { "grad_norm": 0.531173586845398, "learning_rate": 9.781679523845812e-05, "loss": 0.084, "step": 13970 }, { "grad_norm": 0.4086587429046631, "learning_rate": 9.781196003099502e-05, "loss": 0.0709, "step": 13980 }, { "grad_norm": 0.3788236081600189, "learning_rate": 9.780711959488786e-05, "loss": 0.0504, "step": 13990 }, { "grad_norm": 0.3622128665447235, "learning_rate": 9.780227393066599e-05, "loss": 0.0575, "step": 14000 }, { "grad_norm": 0.21499602496623993, "learning_rate": 9.77974230388593e-05, "loss": 0.0614, "step": 14010 }, { "grad_norm": 0.6176992654800415, "learning_rate": 9.779256691999829e-05, "loss": 0.0664, "step": 14020 }, { "grad_norm": 0.3431169390678406, "learning_rate": 9.778770557461403e-05, "loss": 0.058, "step": 14030 }, { "grad_norm": 0.5650629997253418, "learning_rate": 9.778283900323812e-05, "loss": 0.0618, "step": 14040 }, { "grad_norm": 0.4199686348438263, "learning_rate": 9.777796720640277e-05, "loss": 0.0557, "step": 14050 }, { "grad_norm": 0.7393726706504822, "learning_rate": 9.777309018464078e-05, "loss": 0.06, "step": 14060 }, { "grad_norm": 0.42794686555862427, "learning_rate": 9.776820793848547e-05, "loss": 0.0617, "step": 14070 }, { "grad_norm": 0.492011159658432, "learning_rate": 9.776332046847075e-05, "loss": 0.0598, "step": 14080 }, { "grad_norm": 0.42129290103912354, "learning_rate": 9.775842777513111e-05, "loss": 0.0655, "step": 14090 }, { "grad_norm": 0.47617390751838684, "learning_rate": 9.775352985900163e-05, "loss": 0.0572, "step": 14100 }, { "grad_norm": 0.4174290597438812, "learning_rate": 9.774862672061791e-05, "loss": 0.0709, "step": 14110 }, { "grad_norm": 0.3111671507358551, "learning_rate": 9.774371836051616e-05, "loss": 0.0506, "step": 14120 }, { "grad_norm": 0.38148316740989685, "learning_rate": 9.773880477923315e-05, "loss": 0.0677, "step": 14130 }, { "grad_norm": 0.38196635246276855, "learning_rate": 9.773388597730623e-05, "loss": 0.0631, "step": 14140 }, { "grad_norm": 0.37262797355651855, "learning_rate": 9.77289619552733e-05, "loss": 0.0511, "step": 14150 }, { "grad_norm": 0.41217854619026184, "learning_rate": 9.772403271367285e-05, "loss": 0.0621, "step": 14160 }, { "grad_norm": 0.39545688033103943, "learning_rate": 9.771909825304396e-05, "loss": 0.0551, "step": 14170 }, { "grad_norm": 0.367429256439209, "learning_rate": 9.771415857392619e-05, "loss": 0.053, "step": 14180 }, { "grad_norm": 0.565189778804779, "learning_rate": 9.770921367685978e-05, "loss": 0.0736, "step": 14190 }, { "grad_norm": 0.4261168837547302, "learning_rate": 9.770426356238551e-05, "loss": 0.0644, "step": 14200 }, { "grad_norm": 0.3848499357700348, "learning_rate": 9.769930823104469e-05, "loss": 0.0493, "step": 14210 }, { "grad_norm": 0.407436728477478, "learning_rate": 9.769434768337926e-05, "loss": 0.0658, "step": 14220 }, { "grad_norm": 0.7880037426948547, "learning_rate": 9.768938191993164e-05, "loss": 0.0532, "step": 14230 }, { "grad_norm": 0.39034757018089294, "learning_rate": 9.768441094124494e-05, "loss": 0.0563, "step": 14240 }, { "grad_norm": 0.39484938979148865, "learning_rate": 9.767943474786275e-05, "loss": 0.0541, "step": 14250 }, { "grad_norm": 0.4687204658985138, "learning_rate": 9.767445334032923e-05, "loss": 0.0513, "step": 14260 }, { "grad_norm": 0.3323075473308563, "learning_rate": 9.766946671918919e-05, "loss": 0.054, "step": 14270 }, { "grad_norm": 0.5150803327560425, "learning_rate": 9.766447488498796e-05, "loss": 0.0622, "step": 14280 }, { "grad_norm": 0.38034433126449585, "learning_rate": 9.765947783827139e-05, "loss": 0.0527, "step": 14290 }, { "grad_norm": 0.37018540501594543, "learning_rate": 9.765447557958599e-05, "loss": 0.0578, "step": 14300 }, { "grad_norm": 0.464346319437027, "learning_rate": 9.764946810947879e-05, "loss": 0.0469, "step": 14310 }, { "grad_norm": 0.3358563780784607, "learning_rate": 9.764445542849738e-05, "loss": 0.0654, "step": 14320 }, { "grad_norm": 0.5257922410964966, "learning_rate": 9.763943753718998e-05, "loss": 0.06, "step": 14330 }, { "grad_norm": 0.35386142134666443, "learning_rate": 9.76344144361053e-05, "loss": 0.0559, "step": 14340 }, { "grad_norm": 0.6280863881111145, "learning_rate": 9.762938612579269e-05, "loss": 0.0683, "step": 14350 }, { "grad_norm": 0.6066476106643677, "learning_rate": 9.762435260680202e-05, "loss": 0.069, "step": 14360 }, { "grad_norm": 0.4476766884326935, "learning_rate": 9.761931387968373e-05, "loss": 0.0665, "step": 14370 }, { "grad_norm": 0.3101426959037781, "learning_rate": 9.76142699449889e-05, "loss": 0.0553, "step": 14380 }, { "grad_norm": 0.41718819737434387, "learning_rate": 9.760922080326908e-05, "loss": 0.0695, "step": 14390 }, { "grad_norm": 0.4768635034561157, "learning_rate": 9.760416645507644e-05, "loss": 0.0531, "step": 14400 }, { "grad_norm": 0.32419484853744507, "learning_rate": 9.759910690096375e-05, "loss": 0.0488, "step": 14410 }, { "grad_norm": 0.25879567861557007, "learning_rate": 9.759404214148429e-05, "loss": 0.0449, "step": 14420 }, { "grad_norm": 0.5278331637382507, "learning_rate": 9.758897217719191e-05, "loss": 0.0503, "step": 14430 }, { "grad_norm": 0.3456684350967407, "learning_rate": 9.758389700864113e-05, "loss": 0.0496, "step": 14440 }, { "grad_norm": 0.426742821931839, "learning_rate": 9.757881663638688e-05, "loss": 0.0501, "step": 14450 }, { "grad_norm": 0.3203500211238861, "learning_rate": 9.757373106098478e-05, "loss": 0.047, "step": 14460 }, { "grad_norm": 0.6042081713676453, "learning_rate": 9.756864028299097e-05, "loss": 0.0506, "step": 14470 }, { "grad_norm": 0.4872748851776123, "learning_rate": 9.75635443029622e-05, "loss": 0.0536, "step": 14480 }, { "grad_norm": 0.5457574129104614, "learning_rate": 9.755844312145572e-05, "loss": 0.0422, "step": 14490 }, { "grad_norm": 0.9479235410690308, "learning_rate": 9.755333673902941e-05, "loss": 0.0728, "step": 14500 }, { "grad_norm": 0.4016554355621338, "learning_rate": 9.75482251562417e-05, "loss": 0.0616, "step": 14510 }, { "grad_norm": 0.2863680124282837, "learning_rate": 9.754310837365155e-05, "loss": 0.0565, "step": 14520 }, { "grad_norm": 0.5006407499313354, "learning_rate": 9.753798639181856e-05, "loss": 0.072, "step": 14530 }, { "grad_norm": 0.25224265456199646, "learning_rate": 9.753285921130286e-05, "loss": 0.0466, "step": 14540 }, { "grad_norm": 0.4758531153202057, "learning_rate": 9.752772683266512e-05, "loss": 0.0471, "step": 14550 }, { "grad_norm": 0.4560045003890991, "learning_rate": 9.752258925646665e-05, "loss": 0.0457, "step": 14560 }, { "grad_norm": 0.3095613121986389, "learning_rate": 9.751744648326926e-05, "loss": 0.0465, "step": 14570 }, { "grad_norm": 0.6247802376747131, "learning_rate": 9.751229851363536e-05, "loss": 0.0507, "step": 14580 }, { "grad_norm": 0.4268445074558258, "learning_rate": 9.750714534812793e-05, "loss": 0.055, "step": 14590 }, { "grad_norm": 0.4068394601345062, "learning_rate": 9.750198698731053e-05, "loss": 0.0582, "step": 14600 }, { "grad_norm": 0.3406752347946167, "learning_rate": 9.749682343174722e-05, "loss": 0.0521, "step": 14610 }, { "grad_norm": 0.42334744334220886, "learning_rate": 9.749165468200272e-05, "loss": 0.0591, "step": 14620 }, { "grad_norm": 0.55129474401474, "learning_rate": 9.748648073864229e-05, "loss": 0.0544, "step": 14630 }, { "grad_norm": 0.4489625096321106, "learning_rate": 9.748130160223168e-05, "loss": 0.0535, "step": 14640 }, { "grad_norm": 0.69437575340271, "learning_rate": 9.747611727333734e-05, "loss": 0.0567, "step": 14650 }, { "grad_norm": 0.7493348121643066, "learning_rate": 9.74709277525262e-05, "loss": 0.0556, "step": 14660 }, { "grad_norm": 0.34174683690071106, "learning_rate": 9.746573304036576e-05, "loss": 0.0639, "step": 14670 }, { "grad_norm": 0.3935542106628418, "learning_rate": 9.746053313742412e-05, "loss": 0.0519, "step": 14680 }, { "grad_norm": 0.4589988887310028, "learning_rate": 9.745532804426994e-05, "loss": 0.0524, "step": 14690 }, { "grad_norm": 0.5297170877456665, "learning_rate": 9.745011776147242e-05, "loss": 0.0621, "step": 14700 }, { "grad_norm": 0.5365943908691406, "learning_rate": 9.744490228960138e-05, "loss": 0.0653, "step": 14710 }, { "grad_norm": 0.377311646938324, "learning_rate": 9.743968162922713e-05, "loss": 0.0511, "step": 14720 }, { "grad_norm": 0.49534955620765686, "learning_rate": 9.743445578092064e-05, "loss": 0.0561, "step": 14730 }, { "grad_norm": 0.4688047468662262, "learning_rate": 9.742922474525338e-05, "loss": 0.0463, "step": 14740 }, { "grad_norm": 0.2601907551288605, "learning_rate": 9.742398852279741e-05, "loss": 0.0508, "step": 14750 }, { "grad_norm": 0.30921539664268494, "learning_rate": 9.741874711412535e-05, "loss": 0.0529, "step": 14760 }, { "grad_norm": 0.30015474557876587, "learning_rate": 9.741350051981042e-05, "loss": 0.0529, "step": 14770 }, { "grad_norm": 0.42279794812202454, "learning_rate": 9.740824874042633e-05, "loss": 0.0523, "step": 14780 }, { "grad_norm": 0.3250093162059784, "learning_rate": 9.740299177654746e-05, "loss": 0.0455, "step": 14790 }, { "grad_norm": 0.5336729884147644, "learning_rate": 9.739772962874867e-05, "loss": 0.0542, "step": 14800 }, { "grad_norm": 0.6025897860527039, "learning_rate": 9.739246229760541e-05, "loss": 0.0426, "step": 14810 }, { "grad_norm": 0.3154866695404053, "learning_rate": 9.738718978369376e-05, "loss": 0.0519, "step": 14820 }, { "grad_norm": 0.4617200493812561, "learning_rate": 9.738191208759025e-05, "loss": 0.0697, "step": 14830 }, { "grad_norm": 0.5564742088317871, "learning_rate": 9.73766292098721e-05, "loss": 0.0699, "step": 14840 }, { "grad_norm": 0.7249313592910767, "learning_rate": 9.737134115111699e-05, "loss": 0.0862, "step": 14850 }, { "grad_norm": 0.7737976908683777, "learning_rate": 9.736604791190323e-05, "loss": 0.054, "step": 14860 }, { "grad_norm": 0.3247060775756836, "learning_rate": 9.73607494928097e-05, "loss": 0.0557, "step": 14870 }, { "grad_norm": 0.3920322358608246, "learning_rate": 9.735544589441581e-05, "loss": 0.0577, "step": 14880 }, { "grad_norm": 0.4903883934020996, "learning_rate": 9.735013711730154e-05, "loss": 0.049, "step": 14890 }, { "grad_norm": 0.8208074569702148, "learning_rate": 9.734482316204747e-05, "loss": 0.0484, "step": 14900 }, { "grad_norm": 0.34416329860687256, "learning_rate": 9.733950402923473e-05, "loss": 0.0553, "step": 14910 }, { "grad_norm": 0.3248659074306488, "learning_rate": 9.7334179719445e-05, "loss": 0.0378, "step": 14920 }, { "grad_norm": 0.3458632528781891, "learning_rate": 9.732885023326053e-05, "loss": 0.0423, "step": 14930 }, { "grad_norm": 0.39694029092788696, "learning_rate": 9.732351557126418e-05, "loss": 0.0591, "step": 14940 }, { "grad_norm": 0.4270515441894531, "learning_rate": 9.731817573403929e-05, "loss": 0.075, "step": 14950 }, { "grad_norm": 0.38398507237434387, "learning_rate": 9.731283072216985e-05, "loss": 0.0569, "step": 14960 }, { "grad_norm": 0.29121702909469604, "learning_rate": 9.730748053624039e-05, "loss": 0.0525, "step": 14970 }, { "grad_norm": 0.4761861562728882, "learning_rate": 9.730212517683598e-05, "loss": 0.0664, "step": 14980 }, { "grad_norm": 0.462291955947876, "learning_rate": 9.729676464454228e-05, "loss": 0.0581, "step": 14990 }, { "grad_norm": 0.36227571964263916, "learning_rate": 9.72913989399455e-05, "loss": 0.0531, "step": 15000 }, { "grad_norm": 0.31313556432724, "learning_rate": 9.728602806363242e-05, "loss": 0.0526, "step": 15010 }, { "grad_norm": 0.39073827862739563, "learning_rate": 9.728065201619043e-05, "loss": 0.0587, "step": 15020 }, { "grad_norm": 0.44850969314575195, "learning_rate": 9.727527079820742e-05, "loss": 0.0513, "step": 15030 }, { "grad_norm": 0.3612494170665741, "learning_rate": 9.726988441027186e-05, "loss": 0.0601, "step": 15040 }, { "grad_norm": 0.3842093348503113, "learning_rate": 9.726449285297281e-05, "loss": 0.0575, "step": 15050 }, { "grad_norm": 0.626576840877533, "learning_rate": 9.72590961268999e-05, "loss": 0.072, "step": 15060 }, { "grad_norm": 0.3550819456577301, "learning_rate": 9.725369423264328e-05, "loss": 0.0594, "step": 15070 }, { "grad_norm": 0.4429287910461426, "learning_rate": 9.72482871707937e-05, "loss": 0.0641, "step": 15080 }, { "grad_norm": 0.5703012943267822, "learning_rate": 9.724287494194247e-05, "loss": 0.0612, "step": 15090 }, { "grad_norm": 0.4888669550418854, "learning_rate": 9.723745754668147e-05, "loss": 0.063, "step": 15100 }, { "grad_norm": 0.3833734095096588, "learning_rate": 9.723203498560313e-05, "loss": 0.0661, "step": 15110 }, { "grad_norm": 0.3953171670436859, "learning_rate": 9.722660725930046e-05, "loss": 0.0556, "step": 15120 }, { "grad_norm": 0.4743669927120209, "learning_rate": 9.722117436836702e-05, "loss": 0.0585, "step": 15130 }, { "grad_norm": 0.39109694957733154, "learning_rate": 9.721573631339696e-05, "loss": 0.0505, "step": 15140 }, { "grad_norm": 0.48026221990585327, "learning_rate": 9.721029309498494e-05, "loss": 0.0538, "step": 15150 }, { "grad_norm": 0.33927613496780396, "learning_rate": 9.720484471372627e-05, "loss": 0.0553, "step": 15160 }, { "grad_norm": 0.3612862229347229, "learning_rate": 9.719939117021673e-05, "loss": 0.0704, "step": 15170 }, { "grad_norm": 0.5024193525314331, "learning_rate": 9.719393246505275e-05, "loss": 0.0531, "step": 15180 }, { "grad_norm": 0.44788575172424316, "learning_rate": 9.718846859883128e-05, "loss": 0.0408, "step": 15190 }, { "grad_norm": 0.38610392808914185, "learning_rate": 9.718299957214982e-05, "loss": 0.0587, "step": 15200 }, { "grad_norm": 0.7122905850410461, "learning_rate": 9.717752538560646e-05, "loss": 0.0553, "step": 15210 }, { "grad_norm": 0.6204729676246643, "learning_rate": 9.717204603979986e-05, "loss": 0.0728, "step": 15220 }, { "grad_norm": 0.4297749996185303, "learning_rate": 9.716656153532922e-05, "loss": 0.0516, "step": 15230 }, { "grad_norm": 0.3297552466392517, "learning_rate": 9.716107187279434e-05, "loss": 0.0613, "step": 15240 }, { "grad_norm": 0.6627450585365295, "learning_rate": 9.715557705279555e-05, "loss": 0.0791, "step": 15250 }, { "grad_norm": 0.5807854533195496, "learning_rate": 9.715007707593372e-05, "loss": 0.0592, "step": 15260 }, { "grad_norm": 0.3110291063785553, "learning_rate": 9.714457194281036e-05, "loss": 0.0559, "step": 15270 }, { "grad_norm": 0.462498277425766, "learning_rate": 9.713906165402751e-05, "loss": 0.0546, "step": 15280 }, { "grad_norm": 0.37270495295524597, "learning_rate": 9.713354621018774e-05, "loss": 0.0602, "step": 15290 }, { "grad_norm": 0.3078727722167969, "learning_rate": 9.712802561189422e-05, "loss": 0.0466, "step": 15300 }, { "grad_norm": 0.34054651856422424, "learning_rate": 9.712249985975069e-05, "loss": 0.0625, "step": 15310 }, { "grad_norm": 0.3972044885158539, "learning_rate": 9.71169689543614e-05, "loss": 0.0657, "step": 15320 }, { "grad_norm": 0.7011252045631409, "learning_rate": 9.711143289633123e-05, "loss": 0.079, "step": 15330 }, { "grad_norm": 0.41270485520362854, "learning_rate": 9.710589168626561e-05, "loss": 0.0575, "step": 15340 }, { "grad_norm": 0.5083094835281372, "learning_rate": 9.710034532477048e-05, "loss": 0.0676, "step": 15350 }, { "grad_norm": 0.8654505610466003, "learning_rate": 9.709479381245239e-05, "loss": 0.0591, "step": 15360 }, { "grad_norm": 0.5102522373199463, "learning_rate": 9.708923714991847e-05, "loss": 0.0679, "step": 15370 }, { "grad_norm": 0.29061785340309143, "learning_rate": 9.708367533777638e-05, "loss": 0.0566, "step": 15380 }, { "grad_norm": 0.3806493580341339, "learning_rate": 9.707810837663431e-05, "loss": 0.0638, "step": 15390 }, { "grad_norm": 0.27951157093048096, "learning_rate": 9.707253626710113e-05, "loss": 0.053, "step": 15400 }, { "grad_norm": 0.6040575504302979, "learning_rate": 9.706695900978613e-05, "loss": 0.052, "step": 15410 }, { "grad_norm": 0.3882785439491272, "learning_rate": 9.706137660529926e-05, "loss": 0.0581, "step": 15420 }, { "grad_norm": 0.38112694025039673, "learning_rate": 9.705578905425101e-05, "loss": 0.0652, "step": 15430 }, { "grad_norm": 0.30178189277648926, "learning_rate": 9.705019635725241e-05, "loss": 0.0536, "step": 15440 }, { "grad_norm": 0.32297080755233765, "learning_rate": 9.704459851491508e-05, "loss": 0.0595, "step": 15450 }, { "grad_norm": 0.43479427695274353, "learning_rate": 9.703899552785118e-05, "loss": 0.0565, "step": 15460 }, { "grad_norm": 0.479657381772995, "learning_rate": 9.703338739667346e-05, "loss": 0.0493, "step": 15470 }, { "grad_norm": 0.40584370493888855, "learning_rate": 9.70277741219952e-05, "loss": 0.0523, "step": 15480 }, { "grad_norm": 0.49656009674072266, "learning_rate": 9.702215570443027e-05, "loss": 0.0495, "step": 15490 }, { "grad_norm": 0.43176400661468506, "learning_rate": 9.701653214459309e-05, "loss": 0.0554, "step": 15500 }, { "grad_norm": 0.40229564905166626, "learning_rate": 9.701090344309865e-05, "loss": 0.0588, "step": 15510 }, { "grad_norm": 0.44451987743377686, "learning_rate": 9.700526960056247e-05, "loss": 0.051, "step": 15520 }, { "grad_norm": 0.5007333755493164, "learning_rate": 9.699963061760068e-05, "loss": 0.0526, "step": 15530 }, { "grad_norm": 0.29776838421821594, "learning_rate": 9.699398649482997e-05, "loss": 0.0516, "step": 15540 }, { "grad_norm": 0.3341871500015259, "learning_rate": 9.698833723286753e-05, "loss": 0.0564, "step": 15550 }, { "grad_norm": 0.4672217071056366, "learning_rate": 9.698268283233118e-05, "loss": 0.0505, "step": 15560 }, { "grad_norm": 0.3950842320919037, "learning_rate": 9.697702329383929e-05, "loss": 0.0611, "step": 15570 }, { "grad_norm": 0.4717801511287689, "learning_rate": 9.697135861801074e-05, "loss": 0.0591, "step": 15580 }, { "grad_norm": 0.3841693103313446, "learning_rate": 9.696568880546505e-05, "loss": 0.048, "step": 15590 }, { "grad_norm": 0.28790950775146484, "learning_rate": 9.696001385682223e-05, "loss": 0.0494, "step": 15600 }, { "grad_norm": 0.5162932872772217, "learning_rate": 9.695433377270291e-05, "loss": 0.0654, "step": 15610 }, { "grad_norm": 0.3564940392971039, "learning_rate": 9.694864855372824e-05, "loss": 0.0428, "step": 15620 }, { "grad_norm": 0.49320119619369507, "learning_rate": 9.694295820051995e-05, "loss": 0.0559, "step": 15630 }, { "grad_norm": 0.39851778745651245, "learning_rate": 9.693726271370032e-05, "loss": 0.0619, "step": 15640 }, { "grad_norm": 0.5240485072135925, "learning_rate": 9.693156209389221e-05, "loss": 0.0581, "step": 15650 }, { "grad_norm": 0.4818536043167114, "learning_rate": 9.692585634171905e-05, "loss": 0.0571, "step": 15660 }, { "grad_norm": 0.3198942244052887, "learning_rate": 9.692014545780476e-05, "loss": 0.0528, "step": 15670 }, { "grad_norm": 0.33907318115234375, "learning_rate": 9.691442944277393e-05, "loss": 0.0458, "step": 15680 }, { "grad_norm": 0.3362446427345276, "learning_rate": 9.690870829725162e-05, "loss": 0.0505, "step": 15690 }, { "grad_norm": 0.2567596137523651, "learning_rate": 9.69029820218635e-05, "loss": 0.0486, "step": 15700 }, { "grad_norm": 0.2513679563999176, "learning_rate": 9.689725061723579e-05, "loss": 0.054, "step": 15710 }, { "grad_norm": 0.44766634702682495, "learning_rate": 9.689151408399527e-05, "loss": 0.0523, "step": 15720 }, { "grad_norm": 0.3799903392791748, "learning_rate": 9.688577242276924e-05, "loss": 0.0646, "step": 15730 }, { "grad_norm": 0.34789007902145386, "learning_rate": 9.688002563418566e-05, "loss": 0.054, "step": 15740 }, { "grad_norm": 0.6416984796524048, "learning_rate": 9.687427371887293e-05, "loss": 0.0591, "step": 15750 }, { "grad_norm": 0.42457157373428345, "learning_rate": 9.686851667746012e-05, "loss": 0.0559, "step": 15760 }, { "grad_norm": 0.374592661857605, "learning_rate": 9.686275451057677e-05, "loss": 0.0501, "step": 15770 }, { "grad_norm": 0.2569778859615326, "learning_rate": 9.685698721885308e-05, "loss": 0.0496, "step": 15780 }, { "grad_norm": 0.33343908190727234, "learning_rate": 9.68512148029197e-05, "loss": 0.0656, "step": 15790 }, { "grad_norm": 0.29783543944358826, "learning_rate": 9.684543726340791e-05, "loss": 0.0684, "step": 15800 }, { "grad_norm": 0.635382890701294, "learning_rate": 9.683965460094952e-05, "loss": 0.0525, "step": 15810 }, { "grad_norm": 0.34624701738357544, "learning_rate": 9.683386681617694e-05, "loss": 0.0588, "step": 15820 }, { "grad_norm": 0.3855621814727783, "learning_rate": 9.68280739097231e-05, "loss": 0.0559, "step": 15830 }, { "grad_norm": 0.4856884777545929, "learning_rate": 9.682227588222148e-05, "loss": 0.058, "step": 15840 }, { "grad_norm": 0.2975888252258301, "learning_rate": 9.681647273430618e-05, "loss": 0.0521, "step": 15850 }, { "grad_norm": 0.5958489775657654, "learning_rate": 9.681066446661182e-05, "loss": 0.0515, "step": 15860 }, { "grad_norm": 0.3680373430252075, "learning_rate": 9.680485107977357e-05, "loss": 0.0609, "step": 15870 }, { "grad_norm": 0.3702380657196045, "learning_rate": 9.679903257442716e-05, "loss": 0.0534, "step": 15880 }, { "grad_norm": 0.3500135838985443, "learning_rate": 9.679320895120891e-05, "loss": 0.046, "step": 15890 }, { "grad_norm": 0.45465829968452454, "learning_rate": 9.67873802107557e-05, "loss": 0.0404, "step": 15900 }, { "grad_norm": 0.4690186083316803, "learning_rate": 9.67815463537049e-05, "loss": 0.0565, "step": 15910 }, { "grad_norm": 0.3021708130836487, "learning_rate": 9.677570738069457e-05, "loss": 0.0411, "step": 15920 }, { "grad_norm": 0.3498341739177704, "learning_rate": 9.676986329236318e-05, "loss": 0.0596, "step": 15930 }, { "grad_norm": 0.3828352987766266, "learning_rate": 9.676401408934987e-05, "loss": 0.0509, "step": 15940 }, { "grad_norm": 0.29967960715293884, "learning_rate": 9.675815977229428e-05, "loss": 0.0375, "step": 15950 }, { "grad_norm": 0.3119756281375885, "learning_rate": 9.675230034183664e-05, "loss": 0.0392, "step": 15960 }, { "grad_norm": 0.46908339858055115, "learning_rate": 9.674643579861773e-05, "loss": 0.05, "step": 15970 }, { "grad_norm": 0.7418333888053894, "learning_rate": 9.674056614327886e-05, "loss": 0.057, "step": 15980 }, { "grad_norm": 0.3203902542591095, "learning_rate": 9.673469137646198e-05, "loss": 0.0676, "step": 15990 }, { "grad_norm": 0.4318184554576874, "learning_rate": 9.67288114988095e-05, "loss": 0.0484, "step": 16000 }, { "grad_norm": 0.424654096364975, "learning_rate": 9.672292651096447e-05, "loss": 0.055, "step": 16010 }, { "grad_norm": 0.3142888844013214, "learning_rate": 9.671703641357042e-05, "loss": 0.0552, "step": 16020 }, { "grad_norm": 0.24213379621505737, "learning_rate": 9.67111412072715e-05, "loss": 0.0575, "step": 16030 }, { "grad_norm": 0.6151400804519653, "learning_rate": 9.670524089271242e-05, "loss": 0.04, "step": 16040 }, { "grad_norm": 0.4031842350959778, "learning_rate": 9.669933547053842e-05, "loss": 0.0491, "step": 16050 }, { "grad_norm": 0.3863207995891571, "learning_rate": 9.669342494139531e-05, "loss": 0.0506, "step": 16060 }, { "grad_norm": 0.40976643562316895, "learning_rate": 9.668750930592943e-05, "loss": 0.0464, "step": 16070 }, { "grad_norm": 0.35776329040527344, "learning_rate": 9.668158856478775e-05, "loss": 0.0445, "step": 16080 }, { "grad_norm": 0.4266743063926697, "learning_rate": 9.66756627186177e-05, "loss": 0.045, "step": 16090 }, { "grad_norm": 0.4224739372730255, "learning_rate": 9.666973176806737e-05, "loss": 0.0628, "step": 16100 }, { "grad_norm": 0.39421457052230835, "learning_rate": 9.666379571378534e-05, "loss": 0.0557, "step": 16110 }, { "grad_norm": 0.3998565673828125, "learning_rate": 9.665785455642076e-05, "loss": 0.0563, "step": 16120 }, { "grad_norm": 0.3176048994064331, "learning_rate": 9.665190829662337e-05, "loss": 0.0481, "step": 16130 }, { "grad_norm": 0.36022475361824036, "learning_rate": 9.664595693504342e-05, "loss": 0.0541, "step": 16140 }, { "grad_norm": 0.3021596074104309, "learning_rate": 9.664000047233175e-05, "loss": 0.0619, "step": 16150 }, { "grad_norm": 0.29334184527397156, "learning_rate": 9.663403890913976e-05, "loss": 0.0456, "step": 16160 }, { "grad_norm": 0.3843955099582672, "learning_rate": 9.662807224611938e-05, "loss": 0.0552, "step": 16170 }, { "grad_norm": 0.29100653529167175, "learning_rate": 9.662210048392311e-05, "loss": 0.0477, "step": 16180 }, { "grad_norm": 0.42950570583343506, "learning_rate": 9.661612362320405e-05, "loss": 0.0452, "step": 16190 }, { "grad_norm": 0.9888516664505005, "learning_rate": 9.661014166461579e-05, "loss": 0.0483, "step": 16200 }, { "grad_norm": 0.4112706184387207, "learning_rate": 9.66041546088125e-05, "loss": 0.0604, "step": 16210 }, { "grad_norm": 0.36765071749687195, "learning_rate": 9.659816245644895e-05, "loss": 0.0556, "step": 16220 }, { "grad_norm": 0.758520781993866, "learning_rate": 9.65921652081804e-05, "loss": 0.0678, "step": 16230 }, { "grad_norm": 0.3408832252025604, "learning_rate": 9.658616286466271e-05, "loss": 0.061, "step": 16240 }, { "grad_norm": 0.4286729693412781, "learning_rate": 9.65801554265523e-05, "loss": 0.0571, "step": 16250 }, { "grad_norm": 0.42791855335235596, "learning_rate": 9.657414289450612e-05, "loss": 0.0509, "step": 16260 }, { "grad_norm": 0.36640557646751404, "learning_rate": 9.656812526918171e-05, "loss": 0.0488, "step": 16270 }, { "grad_norm": 0.2894496023654938, "learning_rate": 9.656210255123712e-05, "loss": 0.0557, "step": 16280 }, { "grad_norm": 0.2617413103580475, "learning_rate": 9.6556074741331e-05, "loss": 0.0464, "step": 16290 }, { "grad_norm": 0.3819311559200287, "learning_rate": 9.655004184012256e-05, "loss": 0.0434, "step": 16300 }, { "grad_norm": 0.53171706199646, "learning_rate": 9.654400384827152e-05, "loss": 0.0571, "step": 16310 }, { "grad_norm": 0.28005996346473694, "learning_rate": 9.653796076643818e-05, "loss": 0.0424, "step": 16320 }, { "grad_norm": 0.2798978388309479, "learning_rate": 9.653191259528344e-05, "loss": 0.0444, "step": 16330 }, { "grad_norm": 0.6521334052085876, "learning_rate": 9.65258593354687e-05, "loss": 0.0488, "step": 16340 }, { "grad_norm": 0.6496895551681519, "learning_rate": 9.651980098765591e-05, "loss": 0.0446, "step": 16350 }, { "grad_norm": 0.5053685307502747, "learning_rate": 9.651373755250765e-05, "loss": 0.0532, "step": 16360 }, { "grad_norm": 0.3487279415130615, "learning_rate": 9.650766903068697e-05, "loss": 0.047, "step": 16370 }, { "grad_norm": 0.37764155864715576, "learning_rate": 9.650159542285753e-05, "loss": 0.0464, "step": 16380 }, { "grad_norm": 0.32617801427841187, "learning_rate": 9.649551672968353e-05, "loss": 0.0542, "step": 16390 }, { "grad_norm": 0.4310944080352783, "learning_rate": 9.648943295182973e-05, "loss": 0.0556, "step": 16400 }, { "grad_norm": 0.43843695521354675, "learning_rate": 9.648334408996144e-05, "loss": 0.0665, "step": 16410 }, { "grad_norm": 0.35721343755722046, "learning_rate": 9.647725014474452e-05, "loss": 0.0436, "step": 16420 }, { "grad_norm": 0.44956594705581665, "learning_rate": 9.64711511168454e-05, "loss": 0.0477, "step": 16430 }, { "grad_norm": 0.3560231328010559, "learning_rate": 9.646504700693108e-05, "loss": 0.0588, "step": 16440 }, { "grad_norm": 0.3329145610332489, "learning_rate": 9.645893781566907e-05, "loss": 0.0465, "step": 16450 }, { "grad_norm": 0.3191578686237335, "learning_rate": 9.645282354372744e-05, "loss": 0.0455, "step": 16460 }, { "grad_norm": 0.3551603853702545, "learning_rate": 9.644670419177491e-05, "loss": 0.0409, "step": 16470 }, { "grad_norm": 0.32908353209495544, "learning_rate": 9.644057976048062e-05, "loss": 0.0466, "step": 16480 }, { "grad_norm": 0.4730241000652313, "learning_rate": 9.643445025051435e-05, "loss": 0.0457, "step": 16490 }, { "grad_norm": 0.8207025527954102, "learning_rate": 9.642831566254641e-05, "loss": 0.0399, "step": 16500 }, { "grad_norm": 0.35653507709503174, "learning_rate": 9.642217599724769e-05, "loss": 0.0484, "step": 16510 }, { "grad_norm": 0.7796677350997925, "learning_rate": 9.64160312552896e-05, "loss": 0.0771, "step": 16520 }, { "grad_norm": 0.5514439940452576, "learning_rate": 9.64098814373441e-05, "loss": 0.0745, "step": 16530 }, { "grad_norm": 0.3731875419616699, "learning_rate": 9.640372654408374e-05, "loss": 0.0562, "step": 16540 }, { "grad_norm": 0.3928159773349762, "learning_rate": 9.639756657618162e-05, "loss": 0.0639, "step": 16550 }, { "grad_norm": 0.569991946220398, "learning_rate": 9.639140153431138e-05, "loss": 0.0646, "step": 16560 }, { "grad_norm": 0.2556438744068146, "learning_rate": 9.638523141914721e-05, "loss": 0.0509, "step": 16570 }, { "grad_norm": 0.43360769748687744, "learning_rate": 9.637905623136388e-05, "loss": 0.053, "step": 16580 }, { "grad_norm": 0.35681700706481934, "learning_rate": 9.637287597163669e-05, "loss": 0.0517, "step": 16590 }, { "grad_norm": 0.3517395555973053, "learning_rate": 9.63666906406415e-05, "loss": 0.0452, "step": 16600 }, { "grad_norm": 0.4957995116710663, "learning_rate": 9.636050023905473e-05, "loss": 0.0588, "step": 16610 }, { "grad_norm": 0.29428133368492126, "learning_rate": 9.635430476755336e-05, "loss": 0.0437, "step": 16620 }, { "grad_norm": 0.4365566372871399, "learning_rate": 9.63481042268149e-05, "loss": 0.0638, "step": 16630 }, { "grad_norm": 0.47170305252075195, "learning_rate": 9.634189861751745e-05, "loss": 0.0579, "step": 16640 }, { "grad_norm": 0.6314541101455688, "learning_rate": 9.633568794033967e-05, "loss": 0.0539, "step": 16650 }, { "grad_norm": 0.31712806224823, "learning_rate": 9.63294721959607e-05, "loss": 0.0474, "step": 16660 }, { "grad_norm": 0.32610151171684265, "learning_rate": 9.63232513850603e-05, "loss": 0.0626, "step": 16670 }, { "grad_norm": 0.4101214110851288, "learning_rate": 9.631702550831878e-05, "loss": 0.0552, "step": 16680 }, { "grad_norm": 0.40992313623428345, "learning_rate": 9.631079456641698e-05, "loss": 0.0545, "step": 16690 }, { "grad_norm": 0.40536147356033325, "learning_rate": 9.630455856003632e-05, "loss": 0.047, "step": 16700 }, { "grad_norm": 0.6735473275184631, "learning_rate": 9.629831748985876e-05, "loss": 0.0519, "step": 16710 }, { "grad_norm": 0.3646441400051117, "learning_rate": 9.629207135656679e-05, "loss": 0.0658, "step": 16720 }, { "grad_norm": 0.34431353211402893, "learning_rate": 9.628582016084353e-05, "loss": 0.0549, "step": 16730 }, { "grad_norm": 0.7391538619995117, "learning_rate": 9.627956390337254e-05, "loss": 0.0527, "step": 16740 }, { "grad_norm": 0.5152788162231445, "learning_rate": 9.627330258483802e-05, "loss": 0.0675, "step": 16750 }, { "grad_norm": 0.5641005039215088, "learning_rate": 9.62670362059247e-05, "loss": 0.0528, "step": 16760 }, { "grad_norm": 0.31752967834472656, "learning_rate": 9.626076476731786e-05, "loss": 0.0432, "step": 16770 }, { "grad_norm": 0.3088204562664032, "learning_rate": 9.625448826970336e-05, "loss": 0.0474, "step": 16780 }, { "grad_norm": 0.295441597700119, "learning_rate": 9.624820671376755e-05, "loss": 0.0412, "step": 16790 }, { "grad_norm": 0.3616272807121277, "learning_rate": 9.62419201001974e-05, "loss": 0.0511, "step": 16800 }, { "grad_norm": 0.43666407465934753, "learning_rate": 9.623562842968037e-05, "loss": 0.0511, "step": 16810 }, { "grad_norm": 0.8258205056190491, "learning_rate": 9.622933170290454e-05, "loss": 0.0475, "step": 16820 }, { "grad_norm": 0.5419214963912964, "learning_rate": 9.622302992055849e-05, "loss": 0.0446, "step": 16830 }, { "grad_norm": 0.46948859095573425, "learning_rate": 9.62167230833314e-05, "loss": 0.0475, "step": 16840 }, { "grad_norm": 0.3967577815055847, "learning_rate": 9.621041119191295e-05, "loss": 0.0522, "step": 16850 }, { "grad_norm": 0.390941858291626, "learning_rate": 9.620409424699342e-05, "loss": 0.0552, "step": 16860 }, { "grad_norm": 0.4814775884151459, "learning_rate": 9.619777224926359e-05, "loss": 0.051, "step": 16870 }, { "grad_norm": 0.6001628637313843, "learning_rate": 9.619144519941485e-05, "loss": 0.0455, "step": 16880 }, { "grad_norm": 0.3553895354270935, "learning_rate": 9.618511309813912e-05, "loss": 0.0511, "step": 16890 }, { "grad_norm": 0.6852762699127197, "learning_rate": 9.617877594612886e-05, "loss": 0.0441, "step": 16900 }, { "grad_norm": 0.40376341342926025, "learning_rate": 9.617243374407707e-05, "loss": 0.0476, "step": 16910 }, { "grad_norm": 0.34544867277145386, "learning_rate": 9.616608649267736e-05, "loss": 0.0501, "step": 16920 }, { "grad_norm": 0.46361398696899414, "learning_rate": 9.615973419262385e-05, "loss": 0.0581, "step": 16930 }, { "grad_norm": 0.5161553621292114, "learning_rate": 9.615337684461119e-05, "loss": 0.0459, "step": 16940 }, { "grad_norm": 0.35937702655792236, "learning_rate": 9.614701444933465e-05, "loss": 0.0569, "step": 16950 }, { "grad_norm": 0.4807637631893158, "learning_rate": 9.614064700748997e-05, "loss": 0.056, "step": 16960 }, { "grad_norm": 0.45481154322624207, "learning_rate": 9.613427451977352e-05, "loss": 0.0511, "step": 16970 }, { "grad_norm": 0.38836658000946045, "learning_rate": 9.612789698688216e-05, "loss": 0.0527, "step": 16980 }, { "grad_norm": 0.3274594247341156, "learning_rate": 9.612151440951334e-05, "loss": 0.0474, "step": 16990 }, { "grad_norm": 0.45832908153533936, "learning_rate": 9.611512678836506e-05, "loss": 0.0415, "step": 17000 }, { "grad_norm": 0.40540847182273865, "learning_rate": 9.610873412413584e-05, "loss": 0.0701, "step": 17010 }, { "grad_norm": 0.256888747215271, "learning_rate": 9.610233641752476e-05, "loss": 0.0579, "step": 17020 }, { "grad_norm": 0.3611418604850769, "learning_rate": 9.609593366923151e-05, "loss": 0.0421, "step": 17030 }, { "grad_norm": 0.38272592425346375, "learning_rate": 9.608952587995625e-05, "loss": 0.0445, "step": 17040 }, { "grad_norm": 0.4315459430217743, "learning_rate": 9.608311305039972e-05, "loss": 0.0482, "step": 17050 }, { "grad_norm": 0.5420951247215271, "learning_rate": 9.607669518126326e-05, "loss": 0.0713, "step": 17060 }, { "grad_norm": 0.975211501121521, "learning_rate": 9.607027227324866e-05, "loss": 0.0651, "step": 17070 }, { "grad_norm": 0.41023072600364685, "learning_rate": 9.606384432705837e-05, "loss": 0.0546, "step": 17080 }, { "grad_norm": 0.34753426909446716, "learning_rate": 9.60574113433953e-05, "loss": 0.0556, "step": 17090 }, { "grad_norm": 0.37309709191322327, "learning_rate": 9.6050973322963e-05, "loss": 0.0564, "step": 17100 }, { "grad_norm": 0.36285972595214844, "learning_rate": 9.604453026646547e-05, "loss": 0.0541, "step": 17110 }, { "grad_norm": 0.3692886233329773, "learning_rate": 9.603808217460735e-05, "loss": 0.0388, "step": 17120 }, { "grad_norm": 0.43231621384620667, "learning_rate": 9.603162904809377e-05, "loss": 0.0452, "step": 17130 }, { "grad_norm": 0.4513810873031616, "learning_rate": 9.602517088763045e-05, "loss": 0.056, "step": 17140 }, { "grad_norm": 0.3911888003349304, "learning_rate": 9.601870769392365e-05, "loss": 0.0561, "step": 17150 }, { "grad_norm": 0.42832398414611816, "learning_rate": 9.601223946768017e-05, "loss": 0.0538, "step": 17160 }, { "grad_norm": 0.3775603771209717, "learning_rate": 9.600576620960734e-05, "loss": 0.0556, "step": 17170 }, { "grad_norm": 0.23884965479373932, "learning_rate": 9.599928792041308e-05, "loss": 0.0468, "step": 17180 }, { "grad_norm": 0.46606484055519104, "learning_rate": 9.599280460080587e-05, "loss": 0.0505, "step": 17190 }, { "grad_norm": 0.32383862137794495, "learning_rate": 9.59863162514947e-05, "loss": 0.0458, "step": 17200 }, { "grad_norm": 0.26145994663238525, "learning_rate": 9.597982287318911e-05, "loss": 0.0394, "step": 17210 }, { "grad_norm": 0.326538622379303, "learning_rate": 9.597332446659923e-05, "loss": 0.0496, "step": 17220 }, { "grad_norm": 0.46922770142555237, "learning_rate": 9.59668210324357e-05, "loss": 0.0521, "step": 17230 }, { "grad_norm": 0.2689809203147888, "learning_rate": 9.596031257140974e-05, "loss": 0.0499, "step": 17240 }, { "grad_norm": 0.36954963207244873, "learning_rate": 9.59537990842331e-05, "loss": 0.0681, "step": 17250 }, { "grad_norm": 0.4601404070854187, "learning_rate": 9.594728057161806e-05, "loss": 0.0627, "step": 17260 }, { "grad_norm": 0.2352907359600067, "learning_rate": 9.594075703427752e-05, "loss": 0.0451, "step": 17270 }, { "grad_norm": 0.386987566947937, "learning_rate": 9.593422847292486e-05, "loss": 0.0484, "step": 17280 }, { "grad_norm": 0.3299630582332611, "learning_rate": 9.592769488827402e-05, "loss": 0.0485, "step": 17290 }, { "grad_norm": 0.44331252574920654, "learning_rate": 9.592115628103952e-05, "loss": 0.0458, "step": 17300 }, { "grad_norm": 0.4122704565525055, "learning_rate": 9.591461265193643e-05, "loss": 0.0423, "step": 17310 }, { "grad_norm": 0.4558379054069519, "learning_rate": 9.590806400168032e-05, "loss": 0.047, "step": 17320 }, { "grad_norm": 0.3380287289619446, "learning_rate": 9.590151033098735e-05, "loss": 0.0436, "step": 17330 }, { "grad_norm": 0.5800278186798096, "learning_rate": 9.589495164057423e-05, "loss": 0.0487, "step": 17340 }, { "grad_norm": 0.6095606088638306, "learning_rate": 9.58883879311582e-05, "loss": 0.0424, "step": 17350 }, { "grad_norm": 0.34421998262405396, "learning_rate": 9.588181920345705e-05, "loss": 0.0593, "step": 17360 }, { "grad_norm": 0.7315945625305176, "learning_rate": 9.587524545818913e-05, "loss": 0.0557, "step": 17370 }, { "grad_norm": 0.4154987037181854, "learning_rate": 9.586866669607335e-05, "loss": 0.0613, "step": 17380 }, { "grad_norm": 0.4017108380794525, "learning_rate": 9.586208291782915e-05, "loss": 0.0468, "step": 17390 }, { "grad_norm": 0.36889898777008057, "learning_rate": 9.58554941241765e-05, "loss": 0.0466, "step": 17400 }, { "grad_norm": 0.33743351697921753, "learning_rate": 9.584890031583596e-05, "loss": 0.043, "step": 17410 }, { "grad_norm": 0.41017431020736694, "learning_rate": 9.584230149352861e-05, "loss": 0.0543, "step": 17420 }, { "grad_norm": 1.171126127243042, "learning_rate": 9.58356976579761e-05, "loss": 0.0467, "step": 17430 }, { "grad_norm": 0.5176495909690857, "learning_rate": 9.58290888099006e-05, "loss": 0.0616, "step": 17440 }, { "grad_norm": 0.4168725609779358, "learning_rate": 9.582247495002486e-05, "loss": 0.0517, "step": 17450 }, { "grad_norm": 0.38180428743362427, "learning_rate": 9.581585607907214e-05, "loss": 0.0531, "step": 17460 }, { "grad_norm": 0.4119875729084015, "learning_rate": 9.580923219776628e-05, "loss": 0.0605, "step": 17470 }, { "grad_norm": 0.2858964502811432, "learning_rate": 9.580260330683167e-05, "loss": 0.0464, "step": 17480 }, { "grad_norm": 0.37646231055259705, "learning_rate": 9.579596940699322e-05, "loss": 0.0522, "step": 17490 }, { "grad_norm": 0.3593918979167938, "learning_rate": 9.578933049897643e-05, "loss": 0.0432, "step": 17500 }, { "grad_norm": 0.4476747512817383, "learning_rate": 9.578268658350728e-05, "loss": 0.0509, "step": 17510 }, { "grad_norm": 0.3588750958442688, "learning_rate": 9.577603766131235e-05, "loss": 0.0518, "step": 17520 }, { "grad_norm": 0.4475458562374115, "learning_rate": 9.576938373311878e-05, "loss": 0.0393, "step": 17530 }, { "grad_norm": 0.2956116497516632, "learning_rate": 9.576272479965421e-05, "loss": 0.0527, "step": 17540 }, { "grad_norm": 0.5831186175346375, "learning_rate": 9.575606086164687e-05, "loss": 0.0466, "step": 17550 }, { "grad_norm": 0.45015764236450195, "learning_rate": 9.57493919198255e-05, "loss": 0.0585, "step": 17560 }, { "grad_norm": 0.37367767095565796, "learning_rate": 9.57427179749194e-05, "loss": 0.0511, "step": 17570 }, { "grad_norm": 0.41224217414855957, "learning_rate": 9.573603902765846e-05, "loss": 0.0444, "step": 17580 }, { "grad_norm": 0.4955519735813141, "learning_rate": 9.572935507877304e-05, "loss": 0.0479, "step": 17590 }, { "grad_norm": 0.47350096702575684, "learning_rate": 9.57226661289941e-05, "loss": 0.0626, "step": 17600 }, { "grad_norm": 0.29468750953674316, "learning_rate": 9.571597217905315e-05, "loss": 0.0558, "step": 17610 }, { "grad_norm": 0.3941427767276764, "learning_rate": 9.57092732296822e-05, "loss": 0.0511, "step": 17620 }, { "grad_norm": 0.5886451601982117, "learning_rate": 9.570256928161385e-05, "loss": 0.0358, "step": 17630 }, { "grad_norm": 0.33459797501564026, "learning_rate": 9.569586033558126e-05, "loss": 0.0505, "step": 17640 }, { "grad_norm": 0.5290727019309998, "learning_rate": 9.568914639231807e-05, "loss": 0.0533, "step": 17650 }, { "grad_norm": 0.3643973767757416, "learning_rate": 9.568242745255852e-05, "loss": 0.0584, "step": 17660 }, { "grad_norm": 0.4162476658821106, "learning_rate": 9.567570351703739e-05, "loss": 0.0521, "step": 17670 }, { "grad_norm": 0.42913737893104553, "learning_rate": 9.566897458649001e-05, "loss": 0.0591, "step": 17680 }, { "grad_norm": 0.3390141725540161, "learning_rate": 9.566224066165221e-05, "loss": 0.052, "step": 17690 }, { "grad_norm": 0.34860649704933167, "learning_rate": 9.565550174326043e-05, "loss": 0.057, "step": 17700 }, { "grad_norm": 0.34457987546920776, "learning_rate": 9.564875783205162e-05, "loss": 0.0542, "step": 17710 }, { "grad_norm": 0.3426571786403656, "learning_rate": 9.564200892876328e-05, "loss": 0.055, "step": 17720 }, { "grad_norm": 0.36226823925971985, "learning_rate": 9.563525503413348e-05, "loss": 0.043, "step": 17730 }, { "grad_norm": 1.048751950263977, "learning_rate": 9.562849614890079e-05, "loss": 0.0496, "step": 17740 }, { "grad_norm": 0.41554123163223267, "learning_rate": 9.562173227380436e-05, "loss": 0.0635, "step": 17750 }, { "grad_norm": 0.3463945686817169, "learning_rate": 9.561496340958389e-05, "loss": 0.0541, "step": 17760 }, { "grad_norm": 0.475399374961853, "learning_rate": 9.560818955697959e-05, "loss": 0.0427, "step": 17770 }, { "grad_norm": 0.5395157337188721, "learning_rate": 9.560141071673228e-05, "loss": 0.0505, "step": 17780 }, { "grad_norm": 0.5803005695343018, "learning_rate": 9.559462688958323e-05, "loss": 0.0464, "step": 17790 }, { "grad_norm": 0.46902263164520264, "learning_rate": 9.558783807627434e-05, "loss": 0.041, "step": 17800 }, { "grad_norm": 0.43067461252212524, "learning_rate": 9.558104427754801e-05, "loss": 0.0607, "step": 17810 }, { "grad_norm": 0.4198478162288666, "learning_rate": 9.557424549414722e-05, "loss": 0.0543, "step": 17820 }, { "grad_norm": 0.5677474737167358, "learning_rate": 9.556744172681546e-05, "loss": 0.0529, "step": 17830 }, { "grad_norm": 0.8331289291381836, "learning_rate": 9.556063297629677e-05, "loss": 0.0526, "step": 17840 }, { "grad_norm": 0.785799503326416, "learning_rate": 9.555381924333578e-05, "loss": 0.0616, "step": 17850 }, { "grad_norm": 0.55129075050354, "learning_rate": 9.554700052867758e-05, "loss": 0.0558, "step": 17860 }, { "grad_norm": 0.3629668056964874, "learning_rate": 9.554017683306789e-05, "loss": 0.0514, "step": 17870 }, { "grad_norm": 0.4574887156486511, "learning_rate": 9.553334815725294e-05, "loss": 0.0513, "step": 17880 }, { "grad_norm": 0.6332955360412598, "learning_rate": 9.552651450197949e-05, "loss": 0.0551, "step": 17890 }, { "grad_norm": 0.38552233576774597, "learning_rate": 9.551967586799486e-05, "loss": 0.0509, "step": 17900 }, { "grad_norm": 0.3347938358783722, "learning_rate": 9.551283225604692e-05, "loss": 0.0459, "step": 17910 }, { "grad_norm": 0.28984513878822327, "learning_rate": 9.550598366688406e-05, "loss": 0.0516, "step": 17920 }, { "grad_norm": 0.5603975653648376, "learning_rate": 9.549913010125526e-05, "loss": 0.0483, "step": 17930 }, { "grad_norm": 0.33496299386024475, "learning_rate": 9.549227155990999e-05, "loss": 0.0441, "step": 17940 }, { "grad_norm": 0.4408473074436188, "learning_rate": 9.548540804359828e-05, "loss": 0.0596, "step": 17950 }, { "grad_norm": 0.375902384519577, "learning_rate": 9.547853955307077e-05, "loss": 0.0615, "step": 17960 }, { "grad_norm": 0.2753642499446869, "learning_rate": 9.547166608907853e-05, "loss": 0.0445, "step": 17970 }, { "grad_norm": 0.4279259443283081, "learning_rate": 9.546478765237326e-05, "loss": 0.0614, "step": 17980 }, { "grad_norm": 0.2830798625946045, "learning_rate": 9.545790424370715e-05, "loss": 0.0363, "step": 17990 }, { "grad_norm": 0.3132559657096863, "learning_rate": 9.5451015863833e-05, "loss": 0.0526, "step": 18000 }, { "grad_norm": 0.5817726850509644, "learning_rate": 9.544412251350408e-05, "loss": 0.059, "step": 18010 }, { "grad_norm": 0.41289710998535156, "learning_rate": 9.543722419347422e-05, "loss": 0.0405, "step": 18020 }, { "grad_norm": 0.29880982637405396, "learning_rate": 9.543032090449788e-05, "loss": 0.0373, "step": 18030 }, { "grad_norm": 0.4432965815067291, "learning_rate": 9.542341264732992e-05, "loss": 0.0485, "step": 18040 }, { "grad_norm": 0.3987765312194824, "learning_rate": 9.541649942272585e-05, "loss": 0.0468, "step": 18050 }, { "grad_norm": 0.3354209065437317, "learning_rate": 9.54095812314417e-05, "loss": 0.0488, "step": 18060 }, { "grad_norm": 0.31024354696273804, "learning_rate": 9.540265807423401e-05, "loss": 0.0513, "step": 18070 }, { "grad_norm": 0.2742711901664734, "learning_rate": 9.53957299518599e-05, "loss": 0.0452, "step": 18080 }, { "grad_norm": 0.38444021344184875, "learning_rate": 9.5388796865077e-05, "loss": 0.0426, "step": 18090 }, { "grad_norm": 0.4238685965538025, "learning_rate": 9.538185881464353e-05, "loss": 0.0496, "step": 18100 }, { "grad_norm": 0.4686732292175293, "learning_rate": 9.537491580131821e-05, "loss": 0.0386, "step": 18110 }, { "grad_norm": 0.4560915529727936, "learning_rate": 9.53679678258603e-05, "loss": 0.0561, "step": 18120 }, { "grad_norm": 0.4846409857273102, "learning_rate": 9.536101488902966e-05, "loss": 0.0578, "step": 18130 }, { "grad_norm": 0.2904343903064728, "learning_rate": 9.535405699158663e-05, "loss": 0.0484, "step": 18140 }, { "grad_norm": 0.48174721002578735, "learning_rate": 9.53470941342921e-05, "loss": 0.0574, "step": 18150 }, { "grad_norm": 0.7402368187904358, "learning_rate": 9.534012631790756e-05, "loss": 0.0603, "step": 18160 }, { "grad_norm": 0.4752486050128937, "learning_rate": 9.533315354319494e-05, "loss": 0.0635, "step": 18170 }, { "grad_norm": 0.4585202932357788, "learning_rate": 9.532617581091682e-05, "loss": 0.0521, "step": 18180 }, { "grad_norm": 0.5293006896972656, "learning_rate": 9.531919312183629e-05, "loss": 0.0645, "step": 18190 }, { "grad_norm": 0.38053563237190247, "learning_rate": 9.531220547671688e-05, "loss": 0.0422, "step": 18200 }, { "grad_norm": 0.3373056948184967, "learning_rate": 9.530521287632285e-05, "loss": 0.0521, "step": 18210 }, { "grad_norm": 0.30727043747901917, "learning_rate": 9.529821532141884e-05, "loss": 0.0362, "step": 18220 }, { "grad_norm": 0.3533334732055664, "learning_rate": 9.52912128127701e-05, "loss": 0.0451, "step": 18230 }, { "grad_norm": 0.34244289994239807, "learning_rate": 9.528420535114244e-05, "loss": 0.0448, "step": 18240 }, { "grad_norm": 0.32819777727127075, "learning_rate": 9.527719293730215e-05, "loss": 0.0533, "step": 18250 }, { "grad_norm": 0.4440157115459442, "learning_rate": 9.527017557201611e-05, "loss": 0.046, "step": 18260 }, { "grad_norm": 0.4053414762020111, "learning_rate": 9.526315325605176e-05, "loss": 0.0485, "step": 18270 }, { "grad_norm": 0.3444061279296875, "learning_rate": 9.525612599017699e-05, "loss": 0.061, "step": 18280 }, { "grad_norm": 0.48015204071998596, "learning_rate": 9.524909377516033e-05, "loss": 0.0509, "step": 18290 }, { "grad_norm": 0.35273319482803345, "learning_rate": 9.524205661177081e-05, "loss": 0.0491, "step": 18300 }, { "grad_norm": 0.4757229685783386, "learning_rate": 9.523501450077801e-05, "loss": 0.0619, "step": 18310 }, { "grad_norm": 0.34673231840133667, "learning_rate": 9.522796744295202e-05, "loss": 0.0621, "step": 18320 }, { "grad_norm": 0.35561129450798035, "learning_rate": 9.522091543906352e-05, "loss": 0.0441, "step": 18330 }, { "grad_norm": 0.31632521748542786, "learning_rate": 9.521385848988369e-05, "loss": 0.0562, "step": 18340 }, { "grad_norm": 0.9191197752952576, "learning_rate": 9.520679659618428e-05, "loss": 0.0523, "step": 18350 }, { "grad_norm": 0.3110382854938507, "learning_rate": 9.519972975873754e-05, "loss": 0.0448, "step": 18360 }, { "grad_norm": 0.4878231883049011, "learning_rate": 9.519265797831633e-05, "loss": 0.0681, "step": 18370 }, { "grad_norm": 0.43517377972602844, "learning_rate": 9.518558125569399e-05, "loss": 0.0464, "step": 18380 }, { "grad_norm": 0.39319998025894165, "learning_rate": 9.517849959164442e-05, "loss": 0.0499, "step": 18390 }, { "grad_norm": 0.42350298166275024, "learning_rate": 9.517141298694205e-05, "loss": 0.0588, "step": 18400 }, { "grad_norm": 0.4721560776233673, "learning_rate": 9.516432144236188e-05, "loss": 0.0638, "step": 18410 }, { "grad_norm": 0.4111000597476959, "learning_rate": 9.515722495867941e-05, "loss": 0.0577, "step": 18420 }, { "grad_norm": 0.41876092553138733, "learning_rate": 9.515012353667072e-05, "loss": 0.053, "step": 18430 }, { "grad_norm": 0.39202290773391724, "learning_rate": 9.51430171771124e-05, "loss": 0.0434, "step": 18440 }, { "grad_norm": 0.4742405414581299, "learning_rate": 9.513590588078159e-05, "loss": 0.0512, "step": 18450 }, { "grad_norm": 0.35828492045402527, "learning_rate": 9.512878964845597e-05, "loss": 0.0397, "step": 18460 }, { "grad_norm": 0.3806658089160919, "learning_rate": 9.512166848091377e-05, "loss": 0.0493, "step": 18470 }, { "grad_norm": 0.5009909868240356, "learning_rate": 9.511454237893376e-05, "loss": 0.0451, "step": 18480 }, { "grad_norm": 0.35007598996162415, "learning_rate": 9.51074113432952e-05, "loss": 0.0368, "step": 18490 }, { "grad_norm": 0.356178879737854, "learning_rate": 9.510027537477797e-05, "loss": 0.0554, "step": 18500 }, { "grad_norm": 1.0488581657409668, "learning_rate": 9.509313447416242e-05, "loss": 0.0483, "step": 18510 }, { "grad_norm": 0.3293653428554535, "learning_rate": 9.508598864222949e-05, "loss": 0.0473, "step": 18520 }, { "grad_norm": 0.373587429523468, "learning_rate": 9.507883787976062e-05, "loss": 0.0498, "step": 18530 }, { "grad_norm": 0.9993255734443665, "learning_rate": 9.507168218753781e-05, "loss": 0.0481, "step": 18540 }, { "grad_norm": 0.44593343138694763, "learning_rate": 9.506452156634362e-05, "loss": 0.0403, "step": 18550 }, { "grad_norm": 0.4315359890460968, "learning_rate": 9.505735601696109e-05, "loss": 0.0582, "step": 18560 }, { "grad_norm": 0.4333573579788208, "learning_rate": 9.505018554017385e-05, "loss": 0.0486, "step": 18570 }, { "grad_norm": 0.42972803115844727, "learning_rate": 9.504301013676604e-05, "loss": 0.0478, "step": 18580 }, { "grad_norm": 0.40136072039604187, "learning_rate": 9.503582980752238e-05, "loss": 0.0459, "step": 18590 }, { "grad_norm": 0.4639167785644531, "learning_rate": 9.502864455322809e-05, "loss": 0.0683, "step": 18600 }, { "grad_norm": 0.40254536271095276, "learning_rate": 9.502145437466891e-05, "loss": 0.0497, "step": 18610 }, { "grad_norm": 0.2715066075325012, "learning_rate": 9.501425927263116e-05, "loss": 0.0522, "step": 18620 }, { "grad_norm": 0.45143306255340576, "learning_rate": 9.500705924790172e-05, "loss": 0.0715, "step": 18630 }, { "grad_norm": 0.4498295783996582, "learning_rate": 9.499985430126794e-05, "loss": 0.0479, "step": 18640 }, { "grad_norm": 0.32062190771102905, "learning_rate": 9.499264443351775e-05, "loss": 0.0451, "step": 18650 }, { "grad_norm": 0.35003331303596497, "learning_rate": 9.498542964543961e-05, "loss": 0.056, "step": 18660 }, { "grad_norm": 0.36053717136383057, "learning_rate": 9.497820993782252e-05, "loss": 0.0492, "step": 18670 }, { "grad_norm": 0.890367329120636, "learning_rate": 9.497098531145601e-05, "loss": 0.0523, "step": 18680 }, { "grad_norm": 0.4672947824001312, "learning_rate": 9.496375576713017e-05, "loss": 0.0564, "step": 18690 }, { "grad_norm": 0.311801552772522, "learning_rate": 9.49565213056356e-05, "loss": 0.0533, "step": 18700 }, { "grad_norm": 0.877900242805481, "learning_rate": 9.494928192776342e-05, "loss": 0.0522, "step": 18710 }, { "grad_norm": 0.6007913947105408, "learning_rate": 9.494203763430538e-05, "loss": 0.0487, "step": 18720 }, { "grad_norm": 0.43543559312820435, "learning_rate": 9.493478842605366e-05, "loss": 0.0838, "step": 18730 }, { "grad_norm": 0.4431493282318115, "learning_rate": 9.492753430380105e-05, "loss": 0.0559, "step": 18740 }, { "grad_norm": 0.26989373564720154, "learning_rate": 9.492027526834083e-05, "loss": 0.0428, "step": 18750 }, { "grad_norm": 0.3742123246192932, "learning_rate": 9.491301132046684e-05, "loss": 0.0469, "step": 18760 }, { "grad_norm": 0.4045634865760803, "learning_rate": 9.490574246097345e-05, "loss": 0.0686, "step": 18770 }, { "grad_norm": 0.4216637909412384, "learning_rate": 9.48984686906556e-05, "loss": 0.0628, "step": 18780 }, { "grad_norm": 0.3306225836277008, "learning_rate": 9.489119001030871e-05, "loss": 0.0529, "step": 18790 }, { "grad_norm": 0.32351434230804443, "learning_rate": 9.488390642072878e-05, "loss": 0.062, "step": 18800 }, { "grad_norm": 0.34227219223976135, "learning_rate": 9.48766179227123e-05, "loss": 0.0538, "step": 18810 }, { "grad_norm": 0.3015289306640625, "learning_rate": 9.486932451705636e-05, "loss": 0.0472, "step": 18820 }, { "grad_norm": 0.3010396361351013, "learning_rate": 9.486202620455857e-05, "loss": 0.0434, "step": 18830 }, { "grad_norm": 0.4941382110118866, "learning_rate": 9.485472298601704e-05, "loss": 0.051, "step": 18840 }, { "grad_norm": 0.33930516242980957, "learning_rate": 9.484741486223043e-05, "loss": 0.0457, "step": 18850 }, { "grad_norm": 0.3197035491466522, "learning_rate": 9.484010183399797e-05, "loss": 0.05, "step": 18860 }, { "grad_norm": 0.41231173276901245, "learning_rate": 9.483278390211938e-05, "loss": 0.0547, "step": 18870 }, { "grad_norm": 0.5420534014701843, "learning_rate": 9.482546106739496e-05, "loss": 0.0456, "step": 18880 }, { "grad_norm": 0.5061304569244385, "learning_rate": 9.48181333306255e-05, "loss": 0.0485, "step": 18890 }, { "grad_norm": 0.4860251545906067, "learning_rate": 9.481080069261237e-05, "loss": 0.0573, "step": 18900 }, { "grad_norm": 0.34725743532180786, "learning_rate": 9.480346315415745e-05, "loss": 0.0499, "step": 18910 }, { "grad_norm": 0.3921125531196594, "learning_rate": 9.479612071606314e-05, "loss": 0.0413, "step": 18920 }, { "grad_norm": 0.29144805669784546, "learning_rate": 9.478877337913244e-05, "loss": 0.0489, "step": 18930 }, { "grad_norm": 0.39092278480529785, "learning_rate": 9.478142114416881e-05, "loss": 0.0405, "step": 18940 }, { "grad_norm": 0.2569447457790375, "learning_rate": 9.47740640119763e-05, "loss": 0.0471, "step": 18950 }, { "grad_norm": 0.3712750971317291, "learning_rate": 9.476670198335947e-05, "loss": 0.0429, "step": 18960 }, { "grad_norm": 0.31387874484062195, "learning_rate": 9.47593350591234e-05, "loss": 0.0456, "step": 18970 }, { "grad_norm": 0.7257832884788513, "learning_rate": 9.475196324007376e-05, "loss": 0.0557, "step": 18980 }, { "grad_norm": 0.24615240097045898, "learning_rate": 9.474458652701669e-05, "loss": 0.0511, "step": 18990 }, { "grad_norm": 0.38260918855667114, "learning_rate": 9.473720492075892e-05, "loss": 0.045, "step": 19000 }, { "grad_norm": 0.294063001871109, "learning_rate": 9.472981842210768e-05, "loss": 0.0529, "step": 19010 }, { "grad_norm": 0.3158189058303833, "learning_rate": 9.472242703187074e-05, "loss": 0.0492, "step": 19020 }, { "grad_norm": 0.47640910744667053, "learning_rate": 9.471503075085643e-05, "loss": 0.0412, "step": 19030 }, { "grad_norm": 0.2747668921947479, "learning_rate": 9.470762957987359e-05, "loss": 0.0555, "step": 19040 }, { "grad_norm": 0.34680482745170593, "learning_rate": 9.470022351973158e-05, "loss": 0.0494, "step": 19050 }, { "grad_norm": 0.35867518186569214, "learning_rate": 9.469281257124034e-05, "loss": 0.0625, "step": 19060 }, { "grad_norm": 0.3582550287246704, "learning_rate": 9.46853967352103e-05, "loss": 0.0561, "step": 19070 }, { "grad_norm": 0.3213408887386322, "learning_rate": 9.467797601245246e-05, "loss": 0.0385, "step": 19080 }, { "grad_norm": 0.36506861448287964, "learning_rate": 9.467055040377834e-05, "loss": 0.0475, "step": 19090 }, { "grad_norm": 0.31339535117149353, "learning_rate": 9.466311990999999e-05, "loss": 0.0678, "step": 19100 }, { "grad_norm": 0.44869619607925415, "learning_rate": 9.465568453193e-05, "loss": 0.0436, "step": 19110 }, { "grad_norm": 0.5369935035705566, "learning_rate": 9.464824427038148e-05, "loss": 0.0458, "step": 19120 }, { "grad_norm": 0.6381813287734985, "learning_rate": 9.46407991261681e-05, "loss": 0.0611, "step": 19130 }, { "grad_norm": 0.3558575212955475, "learning_rate": 9.463334910010404e-05, "loss": 0.0417, "step": 19140 }, { "grad_norm": 0.3716224431991577, "learning_rate": 9.462589419300403e-05, "loss": 0.0474, "step": 19150 }, { "grad_norm": 0.3291948437690735, "learning_rate": 9.461843440568333e-05, "loss": 0.0517, "step": 19160 }, { "grad_norm": 0.499275267124176, "learning_rate": 9.461096973895773e-05, "loss": 0.0415, "step": 19170 }, { "grad_norm": 0.28457677364349365, "learning_rate": 9.460350019364355e-05, "loss": 0.0483, "step": 19180 }, { "grad_norm": 0.49993979930877686, "learning_rate": 9.459602577055764e-05, "loss": 0.0617, "step": 19190 }, { "grad_norm": 0.3739563822746277, "learning_rate": 9.45885464705174e-05, "loss": 0.0437, "step": 19200 }, { "grad_norm": 0.5581665635108948, "learning_rate": 9.458106229434076e-05, "loss": 0.0518, "step": 19210 }, { "grad_norm": 0.4587426483631134, "learning_rate": 9.457357324284617e-05, "loss": 0.045, "step": 19220 }, { "grad_norm": 0.45765364170074463, "learning_rate": 9.456607931685262e-05, "loss": 0.047, "step": 19230 }, { "grad_norm": 0.555525004863739, "learning_rate": 9.455858051717965e-05, "loss": 0.0495, "step": 19240 }, { "grad_norm": 0.4424363076686859, "learning_rate": 9.45510768446473e-05, "loss": 0.0492, "step": 19250 }, { "grad_norm": 0.2755178213119507, "learning_rate": 9.454356830007618e-05, "loss": 0.0464, "step": 19260 }, { "grad_norm": 0.2819059193134308, "learning_rate": 9.45360548842874e-05, "loss": 0.0441, "step": 19270 }, { "grad_norm": 0.39241456985473633, "learning_rate": 9.452853659810261e-05, "loss": 0.0557, "step": 19280 }, { "grad_norm": 0.2768126130104065, "learning_rate": 9.452101344234401e-05, "loss": 0.0468, "step": 19290 }, { "grad_norm": 0.38355281949043274, "learning_rate": 9.451348541783431e-05, "loss": 0.046, "step": 19300 }, { "grad_norm": 0.48116424679756165, "learning_rate": 9.450595252539678e-05, "loss": 0.049, "step": 19310 }, { "grad_norm": 0.40849044919013977, "learning_rate": 9.449841476585518e-05, "loss": 0.0482, "step": 19320 }, { "grad_norm": 0.35548242926597595, "learning_rate": 9.449087214003384e-05, "loss": 0.0645, "step": 19330 }, { "grad_norm": 0.3952885568141937, "learning_rate": 9.448332464875765e-05, "loss": 0.059, "step": 19340 }, { "grad_norm": 0.5682221055030823, "learning_rate": 9.447577229285192e-05, "loss": 0.0503, "step": 19350 }, { "grad_norm": 0.38206708431243896, "learning_rate": 9.446821507314261e-05, "loss": 0.0426, "step": 19360 }, { "grad_norm": 0.37633809447288513, "learning_rate": 9.446065299045617e-05, "loss": 0.0463, "step": 19370 }, { "grad_norm": 0.6649489402770996, "learning_rate": 9.445308604561955e-05, "loss": 0.0513, "step": 19380 }, { "grad_norm": 0.4434382915496826, "learning_rate": 9.444551423946028e-05, "loss": 0.0471, "step": 19390 }, { "grad_norm": 0.5871487855911255, "learning_rate": 9.443793757280638e-05, "loss": 0.0619, "step": 19400 }, { "grad_norm": 0.34151068329811096, "learning_rate": 9.443035604648646e-05, "loss": 0.0563, "step": 19410 }, { "grad_norm": 0.34732672572135925, "learning_rate": 9.44227696613296e-05, "loss": 0.0445, "step": 19420 }, { "grad_norm": 0.3736587166786194, "learning_rate": 9.441517841816542e-05, "loss": 0.0616, "step": 19430 }, { "grad_norm": 0.3739520311355591, "learning_rate": 9.440758231782413e-05, "loss": 0.0391, "step": 19440 }, { "grad_norm": 0.3035103380680084, "learning_rate": 9.439998136113639e-05, "loss": 0.0464, "step": 19450 }, { "grad_norm": 0.4911500811576843, "learning_rate": 9.439237554893344e-05, "loss": 0.0483, "step": 19460 }, { "grad_norm": 0.4586029648780823, "learning_rate": 9.438476488204705e-05, "loss": 0.0568, "step": 19470 }, { "grad_norm": 0.43136292695999146, "learning_rate": 9.43771493613095e-05, "loss": 0.0527, "step": 19480 }, { "grad_norm": 0.7372437119483948, "learning_rate": 9.436952898755362e-05, "loss": 0.0681, "step": 19490 }, { "grad_norm": 0.363872766494751, "learning_rate": 9.436190376161276e-05, "loss": 0.0434, "step": 19500 }, { "grad_norm": 0.39661577343940735, "learning_rate": 9.43542736843208e-05, "loss": 0.0676, "step": 19510 }, { "grad_norm": 0.2939746379852295, "learning_rate": 9.434663875651216e-05, "loss": 0.0375, "step": 19520 }, { "grad_norm": 0.7551473379135132, "learning_rate": 9.433899897902177e-05, "loss": 0.0513, "step": 19530 }, { "grad_norm": 0.3993140757083893, "learning_rate": 9.433135435268511e-05, "loss": 0.0412, "step": 19540 }, { "grad_norm": 0.5931510329246521, "learning_rate": 9.432370487833819e-05, "loss": 0.0441, "step": 19550 }, { "grad_norm": 0.3900865614414215, "learning_rate": 9.431605055681756e-05, "loss": 0.0597, "step": 19560 }, { "grad_norm": 0.46406272053718567, "learning_rate": 9.430839138896026e-05, "loss": 0.0515, "step": 19570 }, { "grad_norm": 0.4012851119041443, "learning_rate": 9.43007273756039e-05, "loss": 0.0534, "step": 19580 }, { "grad_norm": 0.34557995200157166, "learning_rate": 9.429305851758658e-05, "loss": 0.0533, "step": 19590 }, { "grad_norm": 0.40072351694107056, "learning_rate": 9.428538481574699e-05, "loss": 0.0596, "step": 19600 }, { "grad_norm": 0.5272044539451599, "learning_rate": 9.42777062709243e-05, "loss": 0.0529, "step": 19610 }, { "grad_norm": 0.765654444694519, "learning_rate": 9.427002288395821e-05, "loss": 0.0667, "step": 19620 }, { "grad_norm": 0.353414386510849, "learning_rate": 9.426233465568898e-05, "loss": 0.0588, "step": 19630 }, { "grad_norm": 0.4006589949131012, "learning_rate": 9.42546415869574e-05, "loss": 0.0531, "step": 19640 }, { "grad_norm": 0.5825399160385132, "learning_rate": 9.424694367860473e-05, "loss": 0.0535, "step": 19650 }, { "grad_norm": 0.5804132223129272, "learning_rate": 9.423924093147284e-05, "loss": 0.0469, "step": 19660 }, { "grad_norm": 0.4017391502857208, "learning_rate": 9.423153334640407e-05, "loss": 0.0442, "step": 19670 }, { "grad_norm": 0.6113275289535522, "learning_rate": 9.42238209242413e-05, "loss": 0.0521, "step": 19680 }, { "grad_norm": 0.551359236240387, "learning_rate": 9.421610366582798e-05, "loss": 0.0542, "step": 19690 }, { "grad_norm": 0.433358758687973, "learning_rate": 9.420838157200803e-05, "loss": 0.0457, "step": 19700 }, { "grad_norm": 0.31345972418785095, "learning_rate": 9.420065464362594e-05, "loss": 0.0602, "step": 19710 }, { "grad_norm": 0.3015410602092743, "learning_rate": 9.419292288152673e-05, "loss": 0.0507, "step": 19720 }, { "grad_norm": 0.331787109375, "learning_rate": 9.418518628655588e-05, "loss": 0.0513, "step": 19730 }, { "grad_norm": 0.5570451617240906, "learning_rate": 9.417744485955951e-05, "loss": 0.061, "step": 19740 }, { "grad_norm": 0.5460230112075806, "learning_rate": 9.41696986013842e-05, "loss": 0.0423, "step": 19750 }, { "grad_norm": 0.34029269218444824, "learning_rate": 9.416194751287705e-05, "loss": 0.0588, "step": 19760 }, { "grad_norm": 0.2616148889064789, "learning_rate": 9.415419159488572e-05, "loss": 0.0532, "step": 19770 }, { "grad_norm": 0.4195351004600525, "learning_rate": 9.414643084825837e-05, "loss": 0.0507, "step": 19780 }, { "grad_norm": 0.3175259232521057, "learning_rate": 9.413866527384372e-05, "loss": 0.0582, "step": 19790 }, { "grad_norm": 0.29761242866516113, "learning_rate": 9.4130894872491e-05, "loss": 0.0461, "step": 19800 }, { "grad_norm": 0.3246433734893799, "learning_rate": 9.412311964504998e-05, "loss": 0.0461, "step": 19810 }, { "grad_norm": 0.9969501495361328, "learning_rate": 9.411533959237091e-05, "loss": 0.0409, "step": 19820 }, { "grad_norm": 0.32712334394454956, "learning_rate": 9.410755471530464e-05, "loss": 0.0429, "step": 19830 }, { "grad_norm": 0.7248774766921997, "learning_rate": 9.40997650147025e-05, "loss": 0.0569, "step": 19840 }, { "grad_norm": 0.3938688039779663, "learning_rate": 9.409197049141637e-05, "loss": 0.0603, "step": 19850 }, { "grad_norm": 0.3652009665966034, "learning_rate": 9.408417114629863e-05, "loss": 0.06, "step": 19860 }, { "grad_norm": 0.4224506914615631, "learning_rate": 9.40763669802022e-05, "loss": 0.047, "step": 19870 }, { "grad_norm": 0.4135993421077728, "learning_rate": 9.406855799398056e-05, "loss": 0.0576, "step": 19880 }, { "grad_norm": 0.31260058283805847, "learning_rate": 9.406074418848767e-05, "loss": 0.049, "step": 19890 }, { "grad_norm": 0.44943639636039734, "learning_rate": 9.405292556457805e-05, "loss": 0.0598, "step": 19900 }, { "grad_norm": 0.3157280683517456, "learning_rate": 9.404510212310671e-05, "loss": 0.0432, "step": 19910 }, { "grad_norm": 0.2899739444255829, "learning_rate": 9.403727386492924e-05, "loss": 0.0458, "step": 19920 }, { "grad_norm": 0.3137354850769043, "learning_rate": 9.40294407909017e-05, "loss": 0.0428, "step": 19930 }, { "grad_norm": 0.34274163842201233, "learning_rate": 9.40216029018807e-05, "loss": 0.0436, "step": 19940 }, { "grad_norm": 0.3419102430343628, "learning_rate": 9.401376019872338e-05, "loss": 0.0427, "step": 19950 }, { "grad_norm": 0.29604625701904297, "learning_rate": 9.400591268228746e-05, "loss": 0.0476, "step": 19960 }, { "grad_norm": 0.36637386679649353, "learning_rate": 9.399806035343106e-05, "loss": 0.0382, "step": 19970 }, { "grad_norm": 0.36121150851249695, "learning_rate": 9.399020321301294e-05, "loss": 0.0494, "step": 19980 }, { "grad_norm": 0.4723649322986603, "learning_rate": 9.398234126189234e-05, "loss": 0.0438, "step": 19990 }, { "grad_norm": 0.432375967502594, "learning_rate": 9.397447450092902e-05, "loss": 0.0418, "step": 20000 }, { "grad_norm": 0.20960815250873566, "learning_rate": 9.39666029309833e-05, "loss": 0.0374, "step": 20010 }, { "grad_norm": 0.8967154622077942, "learning_rate": 9.395872655291596e-05, "loss": 0.0485, "step": 20020 }, { "grad_norm": 0.4369119703769684, "learning_rate": 9.395084536758838e-05, "loss": 0.0427, "step": 20030 }, { "grad_norm": 0.3296007513999939, "learning_rate": 9.394295937586243e-05, "loss": 0.0567, "step": 20040 }, { "grad_norm": 0.26325786113739014, "learning_rate": 9.393506857860052e-05, "loss": 0.0536, "step": 20050 }, { "grad_norm": 0.40706321597099304, "learning_rate": 9.392717297666555e-05, "loss": 0.0539, "step": 20060 }, { "grad_norm": 0.5592760443687439, "learning_rate": 9.391927257092101e-05, "loss": 0.0565, "step": 20070 }, { "grad_norm": 0.4219704568386078, "learning_rate": 9.391136736223085e-05, "loss": 0.0554, "step": 20080 }, { "grad_norm": 0.3313888609409332, "learning_rate": 9.390345735145956e-05, "loss": 0.049, "step": 20090 }, { "grad_norm": 0.4299201965332031, "learning_rate": 9.389554253947219e-05, "loss": 0.0477, "step": 20100 }, { "grad_norm": 0.32062849402427673, "learning_rate": 9.388762292713428e-05, "loss": 0.0518, "step": 20110 }, { "grad_norm": 0.5287749767303467, "learning_rate": 9.38796985153119e-05, "loss": 0.0641, "step": 20120 }, { "grad_norm": 0.30636757612228394, "learning_rate": 9.387176930487169e-05, "loss": 0.0494, "step": 20130 }, { "grad_norm": 0.4014318883419037, "learning_rate": 9.386383529668072e-05, "loss": 0.0396, "step": 20140 }, { "grad_norm": 0.702074408531189, "learning_rate": 9.385589649160669e-05, "loss": 0.0479, "step": 20150 }, { "grad_norm": 0.435173362493515, "learning_rate": 9.384795289051775e-05, "loss": 0.0526, "step": 20160 }, { "grad_norm": 0.29967695474624634, "learning_rate": 9.384000449428261e-05, "loss": 0.0529, "step": 20170 }, { "grad_norm": 0.2860206663608551, "learning_rate": 9.383205130377048e-05, "loss": 0.0436, "step": 20180 }, { "grad_norm": 0.35953786969184875, "learning_rate": 9.382409331985114e-05, "loss": 0.0362, "step": 20190 }, { "grad_norm": 0.6653789281845093, "learning_rate": 9.381613054339482e-05, "loss": 0.0404, "step": 20200 }, { "grad_norm": 0.5504899621009827, "learning_rate": 9.380816297527235e-05, "loss": 0.0475, "step": 20210 }, { "grad_norm": 0.47626793384552, "learning_rate": 9.380019061635506e-05, "loss": 0.049, "step": 20220 }, { "grad_norm": 0.5695226788520813, "learning_rate": 9.379221346751474e-05, "loss": 0.0574, "step": 20230 }, { "grad_norm": 0.4177599847316742, "learning_rate": 9.378423152962382e-05, "loss": 0.0478, "step": 20240 }, { "grad_norm": 0.4712597727775574, "learning_rate": 9.377624480355517e-05, "loss": 0.0393, "step": 20250 }, { "grad_norm": 0.347096711397171, "learning_rate": 9.376825329018219e-05, "loss": 0.0417, "step": 20260 }, { "grad_norm": 0.38564684987068176, "learning_rate": 9.376025699037884e-05, "loss": 0.0431, "step": 20270 }, { "grad_norm": 0.27520984411239624, "learning_rate": 9.37522559050196e-05, "loss": 0.0448, "step": 20280 }, { "grad_norm": 0.3090883195400238, "learning_rate": 9.37442500349794e-05, "loss": 0.0443, "step": 20290 }, { "grad_norm": 0.30060717463493347, "learning_rate": 9.373623938113381e-05, "loss": 0.0453, "step": 20300 }, { "grad_norm": 0.6169137954711914, "learning_rate": 9.372822394435883e-05, "loss": 0.0454, "step": 20310 }, { "grad_norm": 0.3272407352924347, "learning_rate": 9.372020372553102e-05, "loss": 0.0588, "step": 20320 }, { "grad_norm": 0.42567527294158936, "learning_rate": 9.371217872552746e-05, "loss": 0.0517, "step": 20330 }, { "grad_norm": 0.3884611427783966, "learning_rate": 9.370414894522576e-05, "loss": 0.0655, "step": 20340 }, { "grad_norm": 0.3831130862236023, "learning_rate": 9.369611438550406e-05, "loss": 0.0652, "step": 20350 }, { "grad_norm": 0.6565023064613342, "learning_rate": 9.368807504724095e-05, "loss": 0.059, "step": 20360 }, { "grad_norm": 0.44059157371520996, "learning_rate": 9.368003093131565e-05, "loss": 0.0701, "step": 20370 }, { "grad_norm": 0.518364667892456, "learning_rate": 9.367198203860785e-05, "loss": 0.0677, "step": 20380 }, { "grad_norm": 0.3286553621292114, "learning_rate": 9.366392836999774e-05, "loss": 0.0575, "step": 20390 }, { "grad_norm": 0.28934815526008606, "learning_rate": 9.365586992636607e-05, "loss": 0.0447, "step": 20400 }, { "grad_norm": 0.5930657982826233, "learning_rate": 9.364780670859412e-05, "loss": 0.0586, "step": 20410 }, { "grad_norm": 0.6763738393783569, "learning_rate": 9.363973871756364e-05, "loss": 0.0709, "step": 20420 }, { "grad_norm": 0.37154802680015564, "learning_rate": 9.363166595415696e-05, "loss": 0.0487, "step": 20430 }, { "grad_norm": 0.3683359622955322, "learning_rate": 9.362358841925686e-05, "loss": 0.0533, "step": 20440 }, { "grad_norm": 0.47761639952659607, "learning_rate": 9.361550611374674e-05, "loss": 0.0532, "step": 20450 }, { "grad_norm": 0.5040227174758911, "learning_rate": 9.360741903851043e-05, "loss": 0.0689, "step": 20460 }, { "grad_norm": 0.3925781846046448, "learning_rate": 9.359932719443236e-05, "loss": 0.069, "step": 20470 }, { "grad_norm": 0.441303014755249, "learning_rate": 9.35912305823974e-05, "loss": 0.0483, "step": 20480 }, { "grad_norm": 0.5815380811691284, "learning_rate": 9.358312920329101e-05, "loss": 0.0514, "step": 20490 }, { "grad_norm": 0.35224390029907227, "learning_rate": 9.357502305799914e-05, "loss": 0.0633, "step": 20500 }, { "grad_norm": 0.528305172920227, "learning_rate": 9.356691214740824e-05, "loss": 0.0603, "step": 20510 }, { "grad_norm": 0.5373373031616211, "learning_rate": 9.355879647240535e-05, "loss": 0.0524, "step": 20520 }, { "grad_norm": 0.3113265037536621, "learning_rate": 9.355067603387798e-05, "loss": 0.0618, "step": 20530 }, { "grad_norm": 0.4113186001777649, "learning_rate": 9.354255083271412e-05, "loss": 0.0539, "step": 20540 }, { "grad_norm": 0.34988781809806824, "learning_rate": 9.353442086980239e-05, "loss": 0.0587, "step": 20550 }, { "grad_norm": 0.34268006682395935, "learning_rate": 9.352628614603185e-05, "loss": 0.0519, "step": 20560 }, { "grad_norm": 0.35116472840309143, "learning_rate": 9.351814666229209e-05, "loss": 0.0491, "step": 20570 }, { "grad_norm": 0.7176857590675354, "learning_rate": 9.351000241947324e-05, "loss": 0.0482, "step": 20580 }, { "grad_norm": 0.4364546239376068, "learning_rate": 9.350185341846594e-05, "loss": 0.0545, "step": 20590 }, { "grad_norm": 0.35958918929100037, "learning_rate": 9.349369966016134e-05, "loss": 0.0495, "step": 20600 }, { "grad_norm": 0.577360987663269, "learning_rate": 9.348554114545117e-05, "loss": 0.0587, "step": 20610 }, { "grad_norm": 0.37167298793792725, "learning_rate": 9.347737787522758e-05, "loss": 0.0504, "step": 20620 }, { "grad_norm": 0.34435513615608215, "learning_rate": 9.346920985038332e-05, "loss": 0.0426, "step": 20630 }, { "grad_norm": 0.48078101873397827, "learning_rate": 9.346103707181162e-05, "loss": 0.0481, "step": 20640 }, { "grad_norm": 0.43883565068244934, "learning_rate": 9.345285954040626e-05, "loss": 0.043, "step": 20650 }, { "grad_norm": 0.3280922472476959, "learning_rate": 9.34446772570615e-05, "loss": 0.0442, "step": 20660 }, { "grad_norm": 0.3576987683773041, "learning_rate": 9.343649022267214e-05, "loss": 0.0426, "step": 20670 }, { "grad_norm": 0.28892749547958374, "learning_rate": 9.342829843813353e-05, "loss": 0.051, "step": 20680 }, { "grad_norm": 0.558322548866272, "learning_rate": 9.342010190434149e-05, "loss": 0.0483, "step": 20690 }, { "grad_norm": 1.0533440113067627, "learning_rate": 9.34119006221924e-05, "loss": 0.0504, "step": 20700 }, { "grad_norm": 0.31582701206207275, "learning_rate": 9.340369459258313e-05, "loss": 0.0425, "step": 20710 }, { "grad_norm": 0.30020564794540405, "learning_rate": 9.339548381641106e-05, "loss": 0.0464, "step": 20720 }, { "grad_norm": 0.38882219791412354, "learning_rate": 9.338726829457413e-05, "loss": 0.0542, "step": 20730 }, { "grad_norm": 0.4085119962692261, "learning_rate": 9.337904802797078e-05, "loss": 0.0461, "step": 20740 }, { "grad_norm": 1.3359788656234741, "learning_rate": 9.337082301749993e-05, "loss": 0.0586, "step": 20750 }, { "grad_norm": 0.5152600407600403, "learning_rate": 9.336259326406109e-05, "loss": 0.0446, "step": 20760 }, { "grad_norm": 2.1994454860687256, "learning_rate": 9.335435876855427e-05, "loss": 0.0476, "step": 20770 }, { "grad_norm": 1.5543146133422852, "learning_rate": 9.334611953187994e-05, "loss": 0.0787, "step": 20780 }, { "grad_norm": 0.9057533144950867, "learning_rate": 9.333787555493914e-05, "loss": 0.0534, "step": 20790 }, { "grad_norm": 0.41645461320877075, "learning_rate": 9.332962683863345e-05, "loss": 0.0654, "step": 20800 }, { "grad_norm": 0.4113345444202423, "learning_rate": 9.332137338386489e-05, "loss": 0.0516, "step": 20810 }, { "grad_norm": 0.3113968074321747, "learning_rate": 9.33131151915361e-05, "loss": 0.0517, "step": 20820 }, { "grad_norm": 0.6784585118293762, "learning_rate": 9.330485226255012e-05, "loss": 0.045, "step": 20830 }, { "grad_norm": 1.0525574684143066, "learning_rate": 9.329658459781061e-05, "loss": 0.0512, "step": 20840 }, { "grad_norm": 0.49150392413139343, "learning_rate": 9.328831219822172e-05, "loss": 0.0527, "step": 20850 }, { "grad_norm": 0.5401943325996399, "learning_rate": 9.328003506468808e-05, "loss": 0.0698, "step": 20860 }, { "grad_norm": 0.5037384033203125, "learning_rate": 9.327175319811488e-05, "loss": 0.0647, "step": 20870 }, { "grad_norm": 0.34953391551971436, "learning_rate": 9.326346659940781e-05, "loss": 0.0532, "step": 20880 }, { "grad_norm": 0.34448903799057007, "learning_rate": 9.325517526947308e-05, "loss": 0.0495, "step": 20890 }, { "grad_norm": 0.37334713339805603, "learning_rate": 9.32468792092174e-05, "loss": 0.0491, "step": 20900 }, { "grad_norm": 0.3738628029823303, "learning_rate": 9.323857841954803e-05, "loss": 0.0615, "step": 20910 }, { "grad_norm": 0.42365580797195435, "learning_rate": 9.323027290137276e-05, "loss": 0.0492, "step": 20920 }, { "grad_norm": 0.5110760927200317, "learning_rate": 9.322196265559981e-05, "loss": 0.0489, "step": 20930 }, { "grad_norm": 0.5586082935333252, "learning_rate": 9.321364768313803e-05, "loss": 0.0481, "step": 20940 }, { "grad_norm": 0.33081233501434326, "learning_rate": 9.32053279848967e-05, "loss": 0.0477, "step": 20950 }, { "grad_norm": 0.3926832973957062, "learning_rate": 9.319700356178567e-05, "loss": 0.048, "step": 20960 }, { "grad_norm": 0.5477834343910217, "learning_rate": 9.318867441471527e-05, "loss": 0.0429, "step": 20970 }, { "grad_norm": 0.27831098437309265, "learning_rate": 9.318034054459637e-05, "loss": 0.0382, "step": 20980 }, { "grad_norm": 0.36777636408805847, "learning_rate": 9.317200195234034e-05, "loss": 0.0472, "step": 20990 }, { "grad_norm": 0.32997366786003113, "learning_rate": 9.316365863885909e-05, "loss": 0.0534, "step": 21000 }, { "grad_norm": 0.3740238547325134, "learning_rate": 9.315531060506502e-05, "loss": 0.0563, "step": 21010 }, { "grad_norm": 0.32456958293914795, "learning_rate": 9.314695785187108e-05, "loss": 0.0477, "step": 21020 }, { "grad_norm": 0.5241549611091614, "learning_rate": 9.313860038019069e-05, "loss": 0.0513, "step": 21030 }, { "grad_norm": 0.24279941618442535, "learning_rate": 9.313023819093782e-05, "loss": 0.0467, "step": 21040 }, { "grad_norm": 0.40677037835121155, "learning_rate": 9.312187128502695e-05, "loss": 0.0421, "step": 21050 }, { "grad_norm": 0.26305028796195984, "learning_rate": 9.311349966337307e-05, "loss": 0.0485, "step": 21060 }, { "grad_norm": 0.3007529377937317, "learning_rate": 9.310512332689169e-05, "loss": 0.042, "step": 21070 }, { "grad_norm": 0.28252318501472473, "learning_rate": 9.309674227649883e-05, "loss": 0.0366, "step": 21080 }, { "grad_norm": 0.35782310366630554, "learning_rate": 9.308835651311103e-05, "loss": 0.0481, "step": 21090 }, { "grad_norm": 0.2796698212623596, "learning_rate": 9.307996603764533e-05, "loss": 0.0445, "step": 21100 }, { "grad_norm": 0.3169427216053009, "learning_rate": 9.307157085101932e-05, "loss": 0.0418, "step": 21110 }, { "grad_norm": 0.24706095457077026, "learning_rate": 9.306317095415109e-05, "loss": 0.0354, "step": 21120 }, { "grad_norm": 0.5990288853645325, "learning_rate": 9.305476634795922e-05, "loss": 0.0475, "step": 21130 }, { "grad_norm": 0.31138351559638977, "learning_rate": 9.304635703336284e-05, "loss": 0.0554, "step": 21140 }, { "grad_norm": 0.36266031861305237, "learning_rate": 9.303794301128157e-05, "loss": 0.0424, "step": 21150 }, { "grad_norm": 0.29598483443260193, "learning_rate": 9.302952428263555e-05, "loss": 0.0376, "step": 21160 }, { "grad_norm": 0.2809413969516754, "learning_rate": 9.302110084834545e-05, "loss": 0.0422, "step": 21170 }, { "grad_norm": 0.27889150381088257, "learning_rate": 9.301267270933245e-05, "loss": 0.0422, "step": 21180 }, { "grad_norm": 0.36756566166877747, "learning_rate": 9.300423986651823e-05, "loss": 0.0412, "step": 21190 }, { "grad_norm": 0.5038533210754395, "learning_rate": 9.299580232082501e-05, "loss": 0.0502, "step": 21200 }, { "grad_norm": 0.25615084171295166, "learning_rate": 9.298736007317547e-05, "loss": 0.0348, "step": 21210 }, { "grad_norm": 0.33857783675193787, "learning_rate": 9.297891312449288e-05, "loss": 0.0439, "step": 21220 }, { "grad_norm": 0.41510942578315735, "learning_rate": 9.297046147570094e-05, "loss": 0.0536, "step": 21230 }, { "grad_norm": 0.38124823570251465, "learning_rate": 9.296200512772396e-05, "loss": 0.0377, "step": 21240 }, { "grad_norm": 0.49103784561157227, "learning_rate": 9.295354408148668e-05, "loss": 0.0476, "step": 21250 }, { "grad_norm": 0.4336002767086029, "learning_rate": 9.294507833791441e-05, "loss": 0.0474, "step": 21260 }, { "grad_norm": 0.3248753845691681, "learning_rate": 9.293660789793295e-05, "loss": 0.0511, "step": 21270 }, { "grad_norm": 0.3050256073474884, "learning_rate": 9.292813276246858e-05, "loss": 0.0454, "step": 21280 }, { "grad_norm": 0.373341828584671, "learning_rate": 9.291965293244816e-05, "loss": 0.0425, "step": 21290 }, { "grad_norm": 0.36905527114868164, "learning_rate": 9.291116840879904e-05, "loss": 0.0432, "step": 21300 }, { "grad_norm": 0.4314521849155426, "learning_rate": 9.290267919244904e-05, "loss": 0.051, "step": 21310 }, { "grad_norm": 0.3645634055137634, "learning_rate": 9.289418528432655e-05, "loss": 0.0509, "step": 21320 }, { "grad_norm": 0.4634406268596649, "learning_rate": 9.288568668536045e-05, "loss": 0.0448, "step": 21330 }, { "grad_norm": 0.38767293095588684, "learning_rate": 9.287718339648013e-05, "loss": 0.043, "step": 21340 }, { "grad_norm": 0.351693332195282, "learning_rate": 9.28686754186155e-05, "loss": 0.0432, "step": 21350 }, { "grad_norm": 0.46667957305908203, "learning_rate": 9.286016275269698e-05, "loss": 0.0558, "step": 21360 }, { "grad_norm": 0.3546038269996643, "learning_rate": 9.285164539965551e-05, "loss": 0.0388, "step": 21370 }, { "grad_norm": 0.46996405720710754, "learning_rate": 9.284312336042251e-05, "loss": 0.0473, "step": 21380 }, { "grad_norm": 0.3960878252983093, "learning_rate": 9.283459663592996e-05, "loss": 0.045, "step": 21390 }, { "grad_norm": 0.2769704759120941, "learning_rate": 9.282606522711033e-05, "loss": 0.0431, "step": 21400 }, { "grad_norm": 0.5887336134910583, "learning_rate": 9.281752913489657e-05, "loss": 0.049, "step": 21410 }, { "grad_norm": 0.2982948124408722, "learning_rate": 9.280898836022222e-05, "loss": 0.0382, "step": 21420 }, { "grad_norm": 0.2682998776435852, "learning_rate": 9.280044290402126e-05, "loss": 0.0395, "step": 21430 }, { "grad_norm": 0.5288773775100708, "learning_rate": 9.279189276722821e-05, "loss": 0.0484, "step": 21440 }, { "grad_norm": 0.40475907921791077, "learning_rate": 9.278333795077812e-05, "loss": 0.0488, "step": 21450 }, { "grad_norm": 0.3428836762905121, "learning_rate": 9.27747784556065e-05, "loss": 0.0449, "step": 21460 }, { "grad_norm": 0.37765222787857056, "learning_rate": 9.276621428264942e-05, "loss": 0.0511, "step": 21470 }, { "grad_norm": 0.5984172821044922, "learning_rate": 9.275764543284345e-05, "loss": 0.0441, "step": 21480 }, { "grad_norm": 0.5873965620994568, "learning_rate": 9.274907190712566e-05, "loss": 0.0445, "step": 21490 }, { "grad_norm": 0.5250704884529114, "learning_rate": 9.274049370643363e-05, "loss": 0.0399, "step": 21500 }, { "grad_norm": 0.541553258895874, "learning_rate": 9.273191083170547e-05, "loss": 0.0533, "step": 21510 }, { "grad_norm": 0.41873687505722046, "learning_rate": 9.27233232838798e-05, "loss": 0.0496, "step": 21520 }, { "grad_norm": 0.33482715487480164, "learning_rate": 9.27147310638957e-05, "loss": 0.0452, "step": 21530 }, { "grad_norm": 0.553834855556488, "learning_rate": 9.270613417269286e-05, "loss": 0.0559, "step": 21540 }, { "grad_norm": 0.4091596305370331, "learning_rate": 9.269753261121138e-05, "loss": 0.0531, "step": 21550 }, { "grad_norm": 0.38686761260032654, "learning_rate": 9.268892638039194e-05, "loss": 0.0571, "step": 21560 }, { "grad_norm": 0.3967103660106659, "learning_rate": 9.268031548117569e-05, "loss": 0.0577, "step": 21570 }, { "grad_norm": 0.4758848547935486, "learning_rate": 9.26716999145043e-05, "loss": 0.0521, "step": 21580 }, { "grad_norm": 0.37020769715309143, "learning_rate": 9.266307968131998e-05, "loss": 0.0559, "step": 21590 }, { "grad_norm": 0.36396268010139465, "learning_rate": 9.26544547825654e-05, "loss": 0.0467, "step": 21600 }, { "grad_norm": 0.38585418462753296, "learning_rate": 9.264582521918376e-05, "loss": 0.0436, "step": 21610 }, { "grad_norm": 0.4061623811721802, "learning_rate": 9.263719099211881e-05, "loss": 0.0421, "step": 21620 }, { "grad_norm": 0.4503259062767029, "learning_rate": 9.262855210231476e-05, "loss": 0.047, "step": 21630 }, { "grad_norm": 0.45771998167037964, "learning_rate": 9.261990855071633e-05, "loss": 0.0414, "step": 21640 }, { "grad_norm": 0.3514363169670105, "learning_rate": 9.261126033826878e-05, "loss": 0.0389, "step": 21650 }, { "grad_norm": 0.47780051827430725, "learning_rate": 9.260260746591786e-05, "loss": 0.061, "step": 21660 }, { "grad_norm": 0.26287418603897095, "learning_rate": 9.259394993460985e-05, "loss": 0.0385, "step": 21670 }, { "grad_norm": 0.2883005738258362, "learning_rate": 9.258528774529151e-05, "loss": 0.0419, "step": 21680 }, { "grad_norm": 0.2842267155647278, "learning_rate": 9.257662089891013e-05, "loss": 0.0392, "step": 21690 }, { "grad_norm": 0.36705276370048523, "learning_rate": 9.25679493964135e-05, "loss": 0.0418, "step": 21700 }, { "grad_norm": 0.4373585283756256, "learning_rate": 9.255927323874994e-05, "loss": 0.0363, "step": 21710 }, { "grad_norm": 0.5585453510284424, "learning_rate": 9.255059242686822e-05, "loss": 0.0492, "step": 21720 }, { "grad_norm": 0.38658690452575684, "learning_rate": 9.254190696171769e-05, "loss": 0.0394, "step": 21730 }, { "grad_norm": 0.475606769323349, "learning_rate": 9.25332168442482e-05, "loss": 0.0474, "step": 21740 }, { "grad_norm": 0.3410329222679138, "learning_rate": 9.252452207541004e-05, "loss": 0.038, "step": 21750 }, { "grad_norm": 0.37830987572669983, "learning_rate": 9.251582265615409e-05, "loss": 0.0503, "step": 21760 }, { "grad_norm": 0.44284337759017944, "learning_rate": 9.250711858743169e-05, "loss": 0.0513, "step": 21770 }, { "grad_norm": 0.2748567461967468, "learning_rate": 9.24984098701947e-05, "loss": 0.056, "step": 21780 }, { "grad_norm": 0.6121186017990112, "learning_rate": 9.248969650539552e-05, "loss": 0.0456, "step": 21790 }, { "grad_norm": 0.4202228784561157, "learning_rate": 9.2480978493987e-05, "loss": 0.0514, "step": 21800 }, { "grad_norm": 0.38255900144577026, "learning_rate": 9.247225583692256e-05, "loss": 0.0398, "step": 21810 }, { "grad_norm": 0.3380793035030365, "learning_rate": 9.246352853515607e-05, "loss": 0.0346, "step": 21820 }, { "grad_norm": 0.37183475494384766, "learning_rate": 9.245479658964194e-05, "loss": 0.0365, "step": 21830 }, { "grad_norm": 0.831921398639679, "learning_rate": 9.244606000133507e-05, "loss": 0.0465, "step": 21840 }, { "grad_norm": 0.29586294293403625, "learning_rate": 9.24373187711909e-05, "loss": 0.0333, "step": 21850 }, { "grad_norm": 0.3435073792934418, "learning_rate": 9.242857290016537e-05, "loss": 0.0466, "step": 21860 }, { "grad_norm": 0.25784358382225037, "learning_rate": 9.241982238921488e-05, "loss": 0.0457, "step": 21870 }, { "grad_norm": 0.5200949907302856, "learning_rate": 9.24110672392964e-05, "loss": 0.0431, "step": 21880 }, { "grad_norm": 0.32795146107673645, "learning_rate": 9.240230745136737e-05, "loss": 0.0485, "step": 21890 }, { "grad_norm": 0.305253803730011, "learning_rate": 9.239354302638575e-05, "loss": 0.037, "step": 21900 }, { "grad_norm": 0.34032249450683594, "learning_rate": 9.238477396531e-05, "loss": 0.0434, "step": 21910 }, { "grad_norm": 0.279870867729187, "learning_rate": 9.23760002690991e-05, "loss": 0.038, "step": 21920 }, { "grad_norm": 0.388823002576828, "learning_rate": 9.236722193871252e-05, "loss": 0.0389, "step": 21930 }, { "grad_norm": 0.390023410320282, "learning_rate": 9.235843897511023e-05, "loss": 0.0412, "step": 21940 }, { "grad_norm": 0.3521849811077118, "learning_rate": 9.234965137925276e-05, "loss": 0.0442, "step": 21950 }, { "grad_norm": 0.3153517544269562, "learning_rate": 9.234085915210108e-05, "loss": 0.0498, "step": 21960 }, { "grad_norm": 0.46997594833374023, "learning_rate": 9.23320622946167e-05, "loss": 0.0397, "step": 21970 }, { "grad_norm": 0.45446157455444336, "learning_rate": 9.232326080776163e-05, "loss": 0.039, "step": 21980 }, { "grad_norm": 0.3275923728942871, "learning_rate": 9.23144546924984e-05, "loss": 0.0435, "step": 21990 }, { "grad_norm": 0.34069472551345825, "learning_rate": 9.230564394979e-05, "loss": 0.0401, "step": 22000 }, { "grad_norm": 0.32211390137672424, "learning_rate": 9.22968285806e-05, "loss": 0.0363, "step": 22010 }, { "grad_norm": 0.3191034495830536, "learning_rate": 9.228800858589242e-05, "loss": 0.0494, "step": 22020 }, { "grad_norm": 0.4701422452926636, "learning_rate": 9.227918396663179e-05, "loss": 0.0516, "step": 22030 }, { "grad_norm": 0.4717274308204651, "learning_rate": 9.227035472378319e-05, "loss": 0.0549, "step": 22040 }, { "grad_norm": 0.776802659034729, "learning_rate": 9.226152085831213e-05, "loss": 0.0383, "step": 22050 }, { "grad_norm": 0.4190305173397064, "learning_rate": 9.22526823711847e-05, "loss": 0.0479, "step": 22060 }, { "grad_norm": 0.268539696931839, "learning_rate": 9.224383926336745e-05, "loss": 0.0473, "step": 22070 }, { "grad_norm": 0.29213255643844604, "learning_rate": 9.223499153582744e-05, "loss": 0.0483, "step": 22080 }, { "grad_norm": 0.38629278540611267, "learning_rate": 9.222613918953226e-05, "loss": 0.0594, "step": 22090 }, { "grad_norm": 0.36161354184150696, "learning_rate": 9.221728222544999e-05, "loss": 0.0491, "step": 22100 }, { "grad_norm": 0.39670538902282715, "learning_rate": 9.22084206445492e-05, "loss": 0.05, "step": 22110 }, { "grad_norm": 0.57256019115448, "learning_rate": 9.2199554447799e-05, "loss": 0.0551, "step": 22120 }, { "grad_norm": 0.3217293918132782, "learning_rate": 9.219068363616897e-05, "loss": 0.0453, "step": 22130 }, { "grad_norm": 0.38431718945503235, "learning_rate": 9.218180821062919e-05, "loss": 0.0483, "step": 22140 }, { "grad_norm": 0.49927273392677307, "learning_rate": 9.21729281721503e-05, "loss": 0.0403, "step": 22150 }, { "grad_norm": 0.3135511577129364, "learning_rate": 9.216404352170339e-05, "loss": 0.0444, "step": 22160 }, { "grad_norm": 0.4645102024078369, "learning_rate": 9.215515426026007e-05, "loss": 0.0446, "step": 22170 }, { "grad_norm": 0.33814042806625366, "learning_rate": 9.214626038879246e-05, "loss": 0.039, "step": 22180 }, { "grad_norm": 0.2994438409805298, "learning_rate": 9.21373619082732e-05, "loss": 0.0384, "step": 22190 }, { "grad_norm": 0.3264652490615845, "learning_rate": 9.212845881967535e-05, "loss": 0.0578, "step": 22200 }, { "grad_norm": 0.5831136703491211, "learning_rate": 9.211955112397262e-05, "loss": 0.0506, "step": 22210 }, { "grad_norm": 0.39688318967819214, "learning_rate": 9.211063882213909e-05, "loss": 0.0481, "step": 22220 }, { "grad_norm": 0.3384591042995453, "learning_rate": 9.210172191514942e-05, "loss": 0.042, "step": 22230 }, { "grad_norm": 0.8842236399650574, "learning_rate": 9.209280040397874e-05, "loss": 0.0522, "step": 22240 }, { "grad_norm": 0.3948633670806885, "learning_rate": 9.208387428960268e-05, "loss": 0.0421, "step": 22250 }, { "grad_norm": 0.45048296451568604, "learning_rate": 9.20749435729974e-05, "loss": 0.0498, "step": 22260 }, { "grad_norm": 0.8068967461585999, "learning_rate": 9.206600825513957e-05, "loss": 0.0402, "step": 22270 }, { "grad_norm": 0.2881636619567871, "learning_rate": 9.20570683370063e-05, "loss": 0.0498, "step": 22280 }, { "grad_norm": 0.39145758748054504, "learning_rate": 9.204812381957528e-05, "loss": 0.0462, "step": 22290 }, { "grad_norm": 0.36338338255882263, "learning_rate": 9.203917470382465e-05, "loss": 0.0495, "step": 22300 }, { "grad_norm": 0.7811047434806824, "learning_rate": 9.203022099073309e-05, "loss": 0.0489, "step": 22310 }, { "grad_norm": 0.40912774205207825, "learning_rate": 9.202126268127976e-05, "loss": 0.0616, "step": 22320 }, { "grad_norm": 0.3625476062297821, "learning_rate": 9.20122997764443e-05, "loss": 0.0519, "step": 22330 }, { "grad_norm": 0.3779207468032837, "learning_rate": 9.200333227720692e-05, "loss": 0.052, "step": 22340 }, { "grad_norm": 0.3402426838874817, "learning_rate": 9.199436018454826e-05, "loss": 0.0541, "step": 22350 }, { "grad_norm": 0.3478364050388336, "learning_rate": 9.198538349944952e-05, "loss": 0.0482, "step": 22360 }, { "grad_norm": 0.4578298330307007, "learning_rate": 9.197640222289234e-05, "loss": 0.0715, "step": 22370 }, { "grad_norm": 0.37344351410865784, "learning_rate": 9.196741635585895e-05, "loss": 0.047, "step": 22380 }, { "grad_norm": 0.4339611530303955, "learning_rate": 9.195842589933199e-05, "loss": 0.0407, "step": 22390 }, { "grad_norm": 0.41184625029563904, "learning_rate": 9.194943085429466e-05, "loss": 0.0407, "step": 22400 }, { "grad_norm": 0.43040764331817627, "learning_rate": 9.194043122173065e-05, "loss": 0.0592, "step": 22410 }, { "grad_norm": 0.4861229360103607, "learning_rate": 9.193142700262413e-05, "loss": 0.0441, "step": 22420 }, { "grad_norm": 0.2701532244682312, "learning_rate": 9.192241819795979e-05, "loss": 0.0419, "step": 22430 }, { "grad_norm": 0.42506441473960876, "learning_rate": 9.191340480872284e-05, "loss": 0.0443, "step": 22440 }, { "grad_norm": 0.34150442481040955, "learning_rate": 9.190438683589895e-05, "loss": 0.0491, "step": 22450 }, { "grad_norm": 0.30955690145492554, "learning_rate": 9.189536428047432e-05, "loss": 0.0389, "step": 22460 }, { "grad_norm": 0.3263419270515442, "learning_rate": 9.188633714343564e-05, "loss": 0.0504, "step": 22470 }, { "grad_norm": 0.43417730927467346, "learning_rate": 9.18773054257701e-05, "loss": 0.0452, "step": 22480 }, { "grad_norm": 0.3749244809150696, "learning_rate": 9.18682691284654e-05, "loss": 0.0428, "step": 22490 }, { "grad_norm": 0.40600571036338806, "learning_rate": 9.185922825250974e-05, "loss": 0.0373, "step": 22500 }, { "grad_norm": 0.34079718589782715, "learning_rate": 9.185018279889181e-05, "loss": 0.0325, "step": 22510 }, { "grad_norm": 0.37577375769615173, "learning_rate": 9.184113276860082e-05, "loss": 0.0469, "step": 22520 }, { "grad_norm": 0.37121686339378357, "learning_rate": 9.183207816262645e-05, "loss": 0.0481, "step": 22530 }, { "grad_norm": 0.28871282935142517, "learning_rate": 9.182301898195891e-05, "loss": 0.0407, "step": 22540 }, { "grad_norm": 0.28778156638145447, "learning_rate": 9.181395522758889e-05, "loss": 0.0456, "step": 22550 }, { "grad_norm": 0.623581051826477, "learning_rate": 9.180488690050759e-05, "loss": 0.0564, "step": 22560 }, { "grad_norm": 0.36255964636802673, "learning_rate": 9.179581400170671e-05, "loss": 0.0377, "step": 22570 }, { "grad_norm": 0.2599315643310547, "learning_rate": 9.178673653217845e-05, "loss": 0.0313, "step": 22580 }, { "grad_norm": 0.3777864873409271, "learning_rate": 9.177765449291551e-05, "loss": 0.0359, "step": 22590 }, { "grad_norm": 0.37761273980140686, "learning_rate": 9.176856788491109e-05, "loss": 0.0377, "step": 22600 }, { "grad_norm": 0.4360474646091461, "learning_rate": 9.175947670915887e-05, "loss": 0.0498, "step": 22610 }, { "grad_norm": 0.31637251377105713, "learning_rate": 9.175038096665309e-05, "loss": 0.0506, "step": 22620 }, { "grad_norm": 0.34464362263679504, "learning_rate": 9.17412806583884e-05, "loss": 0.0453, "step": 22630 }, { "grad_norm": 0.421803742647171, "learning_rate": 9.173217578536002e-05, "loss": 0.0501, "step": 22640 }, { "grad_norm": 0.42035382986068726, "learning_rate": 9.172306634856362e-05, "loss": 0.0435, "step": 22650 }, { "grad_norm": 0.2772374153137207, "learning_rate": 9.171395234899545e-05, "loss": 0.0528, "step": 22660 }, { "grad_norm": 0.30995163321495056, "learning_rate": 9.170483378765214e-05, "loss": 0.0396, "step": 22670 }, { "grad_norm": 0.500014066696167, "learning_rate": 9.169571066553091e-05, "loss": 0.0534, "step": 22680 }, { "grad_norm": 0.3225460648536682, "learning_rate": 9.168658298362946e-05, "loss": 0.0403, "step": 22690 }, { "grad_norm": 0.339408814907074, "learning_rate": 9.167745074294598e-05, "loss": 0.0448, "step": 22700 }, { "grad_norm": 0.38191986083984375, "learning_rate": 9.166831394447913e-05, "loss": 0.051, "step": 22710 }, { "grad_norm": 0.3885268270969391, "learning_rate": 9.165917258922812e-05, "loss": 0.0442, "step": 22720 }, { "grad_norm": 0.4295201301574707, "learning_rate": 9.165002667819262e-05, "loss": 0.0451, "step": 22730 }, { "grad_norm": 0.4713437855243683, "learning_rate": 9.164087621237282e-05, "loss": 0.0477, "step": 22740 }, { "grad_norm": 0.4393973648548126, "learning_rate": 9.163172119276942e-05, "loss": 0.054, "step": 22750 }, { "grad_norm": 0.5054776668548584, "learning_rate": 9.162256162038358e-05, "loss": 0.0582, "step": 22760 }, { "grad_norm": 0.4000076949596405, "learning_rate": 9.161339749621698e-05, "loss": 0.0395, "step": 22770 }, { "grad_norm": 0.2192409783601761, "learning_rate": 9.160422882127177e-05, "loss": 0.0432, "step": 22780 }, { "grad_norm": 0.29986199736595154, "learning_rate": 9.159505559655069e-05, "loss": 0.0376, "step": 22790 }, { "grad_norm": 0.24354229867458344, "learning_rate": 9.158587782305684e-05, "loss": 0.0373, "step": 22800 }, { "grad_norm": 0.33520933985710144, "learning_rate": 9.157669550179391e-05, "loss": 0.0461, "step": 22810 }, { "grad_norm": 0.2907232940196991, "learning_rate": 9.156750863376609e-05, "loss": 0.0466, "step": 22820 }, { "grad_norm": 0.5093023180961609, "learning_rate": 9.155831721997801e-05, "loss": 0.0364, "step": 22830 }, { "grad_norm": 0.2357989251613617, "learning_rate": 9.154912126143484e-05, "loss": 0.0412, "step": 22840 }, { "grad_norm": 0.466225266456604, "learning_rate": 9.153992075914224e-05, "loss": 0.055, "step": 22850 }, { "grad_norm": 0.40541890263557434, "learning_rate": 9.153071571410635e-05, "loss": 0.0492, "step": 22860 }, { "grad_norm": 0.3751359283924103, "learning_rate": 9.152150612733384e-05, "loss": 0.0503, "step": 22870 }, { "grad_norm": 0.3359023928642273, "learning_rate": 9.151229199983184e-05, "loss": 0.0518, "step": 22880 }, { "grad_norm": 0.5083742141723633, "learning_rate": 9.150307333260802e-05, "loss": 0.0448, "step": 22890 }, { "grad_norm": 0.4091525971889496, "learning_rate": 9.149385012667048e-05, "loss": 0.0543, "step": 22900 }, { "grad_norm": 0.27521997690200806, "learning_rate": 9.148462238302788e-05, "loss": 0.0445, "step": 22910 }, { "grad_norm": 0.49705538153648376, "learning_rate": 9.147539010268936e-05, "loss": 0.0498, "step": 22920 }, { "grad_norm": 0.41694188117980957, "learning_rate": 9.14661532866645e-05, "loss": 0.0504, "step": 22930 }, { "grad_norm": 0.46447649598121643, "learning_rate": 9.145691193596348e-05, "loss": 0.0425, "step": 22940 }, { "grad_norm": 0.5778586864471436, "learning_rate": 9.144766605159691e-05, "loss": 0.0493, "step": 22950 }, { "grad_norm": 0.44760769605636597, "learning_rate": 9.14384156345759e-05, "loss": 0.0417, "step": 22960 }, { "grad_norm": 0.33883950114250183, "learning_rate": 9.142916068591204e-05, "loss": 0.0407, "step": 22970 }, { "grad_norm": 0.35182344913482666, "learning_rate": 9.141990120661746e-05, "loss": 0.0474, "step": 22980 }, { "grad_norm": 0.3551713824272156, "learning_rate": 9.141063719770475e-05, "loss": 0.0454, "step": 22990 }, { "grad_norm": 0.40190306305885315, "learning_rate": 9.140136866018704e-05, "loss": 0.0402, "step": 23000 }, { "grad_norm": 0.4523436725139618, "learning_rate": 9.139209559507788e-05, "loss": 0.0417, "step": 23010 }, { "grad_norm": 0.3651514947414398, "learning_rate": 9.13828180033914e-05, "loss": 0.0429, "step": 23020 }, { "grad_norm": 0.4032096862792969, "learning_rate": 9.137353588614212e-05, "loss": 0.0429, "step": 23030 }, { "grad_norm": 0.3348528742790222, "learning_rate": 9.136424924434519e-05, "loss": 0.0401, "step": 23040 }, { "grad_norm": 0.37325337529182434, "learning_rate": 9.135495807901615e-05, "loss": 0.0487, "step": 23050 }, { "grad_norm": 0.3424908518791199, "learning_rate": 9.134566239117108e-05, "loss": 0.0449, "step": 23060 }, { "grad_norm": 0.49584826827049255, "learning_rate": 9.13363621818265e-05, "loss": 0.0426, "step": 23070 }, { "grad_norm": 0.43845969438552856, "learning_rate": 9.132705745199953e-05, "loss": 0.0544, "step": 23080 }, { "grad_norm": 0.4736239016056061, "learning_rate": 9.131774820270768e-05, "loss": 0.0507, "step": 23090 }, { "grad_norm": 0.4083918035030365, "learning_rate": 9.130843443496901e-05, "loss": 0.0644, "step": 23100 }, { "grad_norm": 0.39891937375068665, "learning_rate": 9.129911614980206e-05, "loss": 0.0693, "step": 23110 }, { "grad_norm": 0.7665024995803833, "learning_rate": 9.128979334822584e-05, "loss": 0.0474, "step": 23120 }, { "grad_norm": 0.3271966576576233, "learning_rate": 9.128046603125992e-05, "loss": 0.0503, "step": 23130 }, { "grad_norm": 0.5108307003974915, "learning_rate": 9.12711341999243e-05, "loss": 0.0494, "step": 23140 }, { "grad_norm": 0.33011260628700256, "learning_rate": 9.12617978552395e-05, "loss": 0.0601, "step": 23150 }, { "grad_norm": 0.542803168296814, "learning_rate": 9.12524569982265e-05, "loss": 0.0459, "step": 23160 }, { "grad_norm": 0.36252039670944214, "learning_rate": 9.124311162990684e-05, "loss": 0.0354, "step": 23170 }, { "grad_norm": 0.5100041627883911, "learning_rate": 9.12337617513025e-05, "loss": 0.0458, "step": 23180 }, { "grad_norm": 0.355673223733902, "learning_rate": 9.122440736343596e-05, "loss": 0.0441, "step": 23190 }, { "grad_norm": 0.3776749074459076, "learning_rate": 9.12150484673302e-05, "loss": 0.0375, "step": 23200 }, { "grad_norm": 0.5952038764953613, "learning_rate": 9.120568506400873e-05, "loss": 0.0452, "step": 23210 }, { "grad_norm": 0.31120017170906067, "learning_rate": 9.119631715449548e-05, "loss": 0.0454, "step": 23220 }, { "grad_norm": 0.3946845531463623, "learning_rate": 9.118694473981493e-05, "loss": 0.0413, "step": 23230 }, { "grad_norm": 0.3305130898952484, "learning_rate": 9.117756782099203e-05, "loss": 0.0411, "step": 23240 }, { "grad_norm": 0.38019803166389465, "learning_rate": 9.11681863990522e-05, "loss": 0.0495, "step": 23250 }, { "grad_norm": 0.4083479344844818, "learning_rate": 9.115880047502142e-05, "loss": 0.036, "step": 23260 }, { "grad_norm": 0.34348732233047485, "learning_rate": 9.114941004992609e-05, "loss": 0.0479, "step": 23270 }, { "grad_norm": 0.44593825936317444, "learning_rate": 9.114001512479317e-05, "loss": 0.0459, "step": 23280 }, { "grad_norm": 0.3741427958011627, "learning_rate": 9.113061570065003e-05, "loss": 0.0465, "step": 23290 }, { "grad_norm": 0.6181777119636536, "learning_rate": 9.112121177852459e-05, "loss": 0.0513, "step": 23300 }, { "grad_norm": 0.42911210656166077, "learning_rate": 9.111180335944527e-05, "loss": 0.0525, "step": 23310 }, { "grad_norm": 0.4820404052734375, "learning_rate": 9.110239044444093e-05, "loss": 0.0427, "step": 23320 }, { "grad_norm": 0.4098185896873474, "learning_rate": 9.109297303454099e-05, "loss": 0.0385, "step": 23330 }, { "grad_norm": 0.3318181335926056, "learning_rate": 9.108355113077526e-05, "loss": 0.0469, "step": 23340 }, { "grad_norm": 0.40949898958206177, "learning_rate": 9.107412473417419e-05, "loss": 0.055, "step": 23350 }, { "grad_norm": 0.31722646951675415, "learning_rate": 9.106469384576858e-05, "loss": 0.0453, "step": 23360 }, { "grad_norm": 0.3225176930427551, "learning_rate": 9.105525846658978e-05, "loss": 0.0488, "step": 23370 }, { "grad_norm": 0.32901546359062195, "learning_rate": 9.104581859766965e-05, "loss": 0.0518, "step": 23380 }, { "grad_norm": 0.4925488233566284, "learning_rate": 9.10363742400405e-05, "loss": 0.0472, "step": 23390 }, { "grad_norm": 0.704037070274353, "learning_rate": 9.102692539473518e-05, "loss": 0.0542, "step": 23400 }, { "grad_norm": 0.29287832975387573, "learning_rate": 9.101747206278697e-05, "loss": 0.0404, "step": 23410 }, { "grad_norm": 0.3467637896537781, "learning_rate": 9.100801424522968e-05, "loss": 0.0394, "step": 23420 }, { "grad_norm": 0.494817316532135, "learning_rate": 9.099855194309762e-05, "loss": 0.0419, "step": 23430 }, { "grad_norm": 0.44231829047203064, "learning_rate": 9.098908515742554e-05, "loss": 0.0389, "step": 23440 }, { "grad_norm": 0.4893515706062317, "learning_rate": 9.097961388924873e-05, "loss": 0.0435, "step": 23450 }, { "grad_norm": 0.6065194010734558, "learning_rate": 9.097013813960298e-05, "loss": 0.0383, "step": 23460 }, { "grad_norm": 0.2567960023880005, "learning_rate": 9.09606579095245e-05, "loss": 0.0391, "step": 23470 }, { "grad_norm": 0.262548565864563, "learning_rate": 9.095117320005008e-05, "loss": 0.0373, "step": 23480 }, { "grad_norm": 0.2540869414806366, "learning_rate": 9.094168401221691e-05, "loss": 0.0423, "step": 23490 }, { "grad_norm": 0.31834712624549866, "learning_rate": 9.093219034706273e-05, "loss": 0.0601, "step": 23500 }, { "grad_norm": 0.31238770484924316, "learning_rate": 9.092269220562577e-05, "loss": 0.0393, "step": 23510 }, { "grad_norm": 0.3275942802429199, "learning_rate": 9.09131895889447e-05, "loss": 0.0589, "step": 23520 }, { "grad_norm": 0.43610021471977234, "learning_rate": 9.090368249805873e-05, "loss": 0.0339, "step": 23530 }, { "grad_norm": 0.2771140933036804, "learning_rate": 9.089417093400754e-05, "loss": 0.0457, "step": 23540 }, { "grad_norm": 0.3186388313770294, "learning_rate": 9.088465489783131e-05, "loss": 0.0382, "step": 23550 }, { "grad_norm": 0.30498170852661133, "learning_rate": 9.087513439057068e-05, "loss": 0.0416, "step": 23560 }, { "grad_norm": 0.3127652704715729, "learning_rate": 9.08656094132668e-05, "loss": 0.0413, "step": 23570 }, { "grad_norm": 0.29689592123031616, "learning_rate": 9.085607996696134e-05, "loss": 0.0326, "step": 23580 }, { "grad_norm": 0.25656744837760925, "learning_rate": 9.084654605269639e-05, "loss": 0.0546, "step": 23590 }, { "grad_norm": 0.38114801049232483, "learning_rate": 9.083700767151457e-05, "loss": 0.0434, "step": 23600 }, { "grad_norm": 0.39599862694740295, "learning_rate": 9.082746482445898e-05, "loss": 0.0453, "step": 23610 }, { "grad_norm": 0.25989216566085815, "learning_rate": 9.081791751257325e-05, "loss": 0.0444, "step": 23620 }, { "grad_norm": 0.7751835584640503, "learning_rate": 9.080836573690142e-05, "loss": 0.0404, "step": 23630 }, { "grad_norm": 0.41701969504356384, "learning_rate": 9.079880949848805e-05, "loss": 0.0478, "step": 23640 }, { "grad_norm": 0.4654902219772339, "learning_rate": 9.078924879837822e-05, "loss": 0.048, "step": 23650 }, { "grad_norm": 0.8013936877250671, "learning_rate": 9.077968363761747e-05, "loss": 0.0413, "step": 23660 }, { "grad_norm": 0.22661685943603516, "learning_rate": 9.077011401725182e-05, "loss": 0.0453, "step": 23670 }, { "grad_norm": 0.3895590603351593, "learning_rate": 9.07605399383278e-05, "loss": 0.0474, "step": 23680 }, { "grad_norm": 0.3956400454044342, "learning_rate": 9.075096140189243e-05, "loss": 0.0474, "step": 23690 }, { "grad_norm": 0.2961104214191437, "learning_rate": 9.074137840899318e-05, "loss": 0.0391, "step": 23700 }, { "grad_norm": 0.5530878901481628, "learning_rate": 9.073179096067804e-05, "loss": 0.0514, "step": 23710 }, { "grad_norm": 0.36155006289482117, "learning_rate": 9.072219905799549e-05, "loss": 0.0443, "step": 23720 }, { "grad_norm": 0.4774215519428253, "learning_rate": 9.071260270199447e-05, "loss": 0.0558, "step": 23730 }, { "grad_norm": 0.28716760873794556, "learning_rate": 9.070300189372441e-05, "loss": 0.0449, "step": 23740 }, { "grad_norm": 0.3478676676750183, "learning_rate": 9.069339663423528e-05, "loss": 0.0429, "step": 23750 }, { "grad_norm": 0.4605434536933899, "learning_rate": 9.068378692457747e-05, "loss": 0.047, "step": 23760 }, { "grad_norm": 0.3536325693130493, "learning_rate": 9.067417276580189e-05, "loss": 0.0467, "step": 23770 }, { "grad_norm": 0.376907616853714, "learning_rate": 9.066455415895993e-05, "loss": 0.0459, "step": 23780 }, { "grad_norm": 0.3756680190563202, "learning_rate": 9.065493110510346e-05, "loss": 0.0456, "step": 23790 }, { "grad_norm": 0.5610321164131165, "learning_rate": 9.064530360528484e-05, "loss": 0.0528, "step": 23800 }, { "grad_norm": 0.24362915754318237, "learning_rate": 9.063567166055695e-05, "loss": 0.0437, "step": 23810 }, { "grad_norm": 0.5121342539787292, "learning_rate": 9.062603527197308e-05, "loss": 0.0469, "step": 23820 }, { "grad_norm": 0.2972973883152008, "learning_rate": 9.06163944405871e-05, "loss": 0.0426, "step": 23830 }, { "grad_norm": 0.4284428358078003, "learning_rate": 9.060674916745327e-05, "loss": 0.0438, "step": 23840 }, { "grad_norm": 0.30242979526519775, "learning_rate": 9.05970994536264e-05, "loss": 0.043, "step": 23850 }, { "grad_norm": 0.23079560697078705, "learning_rate": 9.05874453001618e-05, "loss": 0.043, "step": 23860 }, { "grad_norm": 0.37380462884902954, "learning_rate": 9.057778670811517e-05, "loss": 0.049, "step": 23870 }, { "grad_norm": 0.7185073494911194, "learning_rate": 9.056812367854281e-05, "loss": 0.0432, "step": 23880 }, { "grad_norm": 0.4156869351863861, "learning_rate": 9.055845621250143e-05, "loss": 0.0476, "step": 23890 }, { "grad_norm": 0.30423006415367126, "learning_rate": 9.054878431104825e-05, "loss": 0.0398, "step": 23900 }, { "grad_norm": 0.4218680262565613, "learning_rate": 9.0539107975241e-05, "loss": 0.035, "step": 23910 }, { "grad_norm": 0.2621612846851349, "learning_rate": 9.052942720613784e-05, "loss": 0.0447, "step": 23920 }, { "grad_norm": 0.7017450928688049, "learning_rate": 9.051974200479745e-05, "loss": 0.0433, "step": 23930 }, { "grad_norm": 0.4740833640098572, "learning_rate": 9.051005237227901e-05, "loss": 0.0512, "step": 23940 }, { "grad_norm": 0.35418838262557983, "learning_rate": 9.050035830964215e-05, "loss": 0.0439, "step": 23950 }, { "grad_norm": 0.49421852827072144, "learning_rate": 9.049065981794698e-05, "loss": 0.0461, "step": 23960 }, { "grad_norm": 0.39810436964035034, "learning_rate": 9.048095689825414e-05, "loss": 0.0486, "step": 23970 }, { "grad_norm": 0.6286148428916931, "learning_rate": 9.047124955162472e-05, "loss": 0.0413, "step": 23980 }, { "grad_norm": 0.28966110944747925, "learning_rate": 9.046153777912028e-05, "loss": 0.0396, "step": 23990 }, { "grad_norm": 0.33371052145957947, "learning_rate": 9.045182158180292e-05, "loss": 0.0488, "step": 24000 }, { "grad_norm": 0.3656078577041626, "learning_rate": 9.044210096073516e-05, "loss": 0.039, "step": 24010 }, { "grad_norm": 0.2697644531726837, "learning_rate": 9.043237591698004e-05, "loss": 0.0435, "step": 24020 }, { "grad_norm": 0.3337746858596802, "learning_rate": 9.04226464516011e-05, "loss": 0.0417, "step": 24030 }, { "grad_norm": 0.356317937374115, "learning_rate": 9.041291256566229e-05, "loss": 0.0407, "step": 24040 }, { "grad_norm": 0.5898444056510925, "learning_rate": 9.040317426022814e-05, "loss": 0.0409, "step": 24050 }, { "grad_norm": 0.3529130816459656, "learning_rate": 9.03934315363636e-05, "loss": 0.0451, "step": 24060 }, { "grad_norm": 0.2792249619960785, "learning_rate": 9.038368439513409e-05, "loss": 0.0399, "step": 24070 }, { "grad_norm": 0.3143593668937683, "learning_rate": 9.03739328376056e-05, "loss": 0.0369, "step": 24080 }, { "grad_norm": 0.24899950623512268, "learning_rate": 9.036417686484451e-05, "loss": 0.0334, "step": 24090 }, { "grad_norm": 0.565961480140686, "learning_rate": 9.035441647791773e-05, "loss": 0.0552, "step": 24100 }, { "grad_norm": 0.420458048582077, "learning_rate": 9.034465167789263e-05, "loss": 0.0372, "step": 24110 }, { "grad_norm": 0.30911505222320557, "learning_rate": 9.033488246583706e-05, "loss": 0.0381, "step": 24120 }, { "grad_norm": 0.3967691957950592, "learning_rate": 9.032510884281941e-05, "loss": 0.0554, "step": 24130 }, { "grad_norm": 0.3061392307281494, "learning_rate": 9.031533080990848e-05, "loss": 0.0454, "step": 24140 }, { "grad_norm": 0.4179416000843048, "learning_rate": 9.030554836817358e-05, "loss": 0.0373, "step": 24150 }, { "grad_norm": 0.2425709217786789, "learning_rate": 9.029576151868451e-05, "loss": 0.0325, "step": 24160 }, { "grad_norm": 0.296432763338089, "learning_rate": 9.028597026251155e-05, "loss": 0.0365, "step": 24170 }, { "grad_norm": 0.36042502522468567, "learning_rate": 9.027617460072547e-05, "loss": 0.0465, "step": 24180 }, { "grad_norm": 0.3364031910896301, "learning_rate": 9.026637453439745e-05, "loss": 0.0402, "step": 24190 }, { "grad_norm": 0.4437991678714752, "learning_rate": 9.025657006459927e-05, "loss": 0.0314, "step": 24200 }, { "grad_norm": 0.3269263803958893, "learning_rate": 9.024676119240311e-05, "loss": 0.0379, "step": 24210 }, { "grad_norm": 0.26461806893348694, "learning_rate": 9.023694791888166e-05, "loss": 0.0399, "step": 24220 }, { "grad_norm": 0.4379410743713379, "learning_rate": 9.022713024510808e-05, "loss": 0.0527, "step": 24230 }, { "grad_norm": 0.4031730890274048, "learning_rate": 9.021730817215601e-05, "loss": 0.0395, "step": 24240 }, { "grad_norm": 0.3200777471065521, "learning_rate": 9.02074817010996e-05, "loss": 0.0388, "step": 24250 }, { "grad_norm": 0.3858821988105774, "learning_rate": 9.019765083301342e-05, "loss": 0.0462, "step": 24260 }, { "grad_norm": 0.3327317535877228, "learning_rate": 9.01878155689726e-05, "loss": 0.0461, "step": 24270 }, { "grad_norm": 0.32918548583984375, "learning_rate": 9.017797591005268e-05, "loss": 0.0363, "step": 24280 }, { "grad_norm": 0.30896350741386414, "learning_rate": 9.016813185732972e-05, "loss": 0.0467, "step": 24290 }, { "grad_norm": 0.4566667377948761, "learning_rate": 9.015828341188027e-05, "loss": 0.0492, "step": 24300 }, { "grad_norm": 0.4408932030200958, "learning_rate": 9.01484305747813e-05, "loss": 0.0429, "step": 24310 }, { "grad_norm": 0.3368363082408905, "learning_rate": 9.013857334711033e-05, "loss": 0.0378, "step": 24320 }, { "grad_norm": 0.4570847749710083, "learning_rate": 9.012871172994534e-05, "loss": 0.0359, "step": 24330 }, { "grad_norm": 0.4127565920352936, "learning_rate": 9.011884572436476e-05, "loss": 0.0549, "step": 24340 }, { "grad_norm": 0.408372163772583, "learning_rate": 9.010897533144754e-05, "loss": 0.0404, "step": 24350 }, { "grad_norm": 0.546145498752594, "learning_rate": 9.009910055227306e-05, "loss": 0.0466, "step": 24360 }, { "grad_norm": 0.36087995767593384, "learning_rate": 9.008922138792124e-05, "loss": 0.0469, "step": 24370 }, { "grad_norm": 0.30959707498550415, "learning_rate": 9.007933783947244e-05, "loss": 0.0442, "step": 24380 }, { "grad_norm": 0.33778929710388184, "learning_rate": 9.006944990800752e-05, "loss": 0.0395, "step": 24390 }, { "grad_norm": 0.49404728412628174, "learning_rate": 9.005955759460779e-05, "loss": 0.0443, "step": 24400 }, { "grad_norm": 0.42712244391441345, "learning_rate": 9.004966090035508e-05, "loss": 0.0389, "step": 24410 }, { "grad_norm": 0.2732807993888855, "learning_rate": 9.003975982633166e-05, "loss": 0.0407, "step": 24420 }, { "grad_norm": 0.3893149197101593, "learning_rate": 9.00298543736203e-05, "loss": 0.0344, "step": 24430 }, { "grad_norm": 0.22261959314346313, "learning_rate": 9.001994454330427e-05, "loss": 0.035, "step": 24440 }, { "grad_norm": 0.31643664836883545, "learning_rate": 9.001003033646727e-05, "loss": 0.0413, "step": 24450 }, { "grad_norm": 0.3032538890838623, "learning_rate": 9.00001117541935e-05, "loss": 0.0373, "step": 24460 }, { "grad_norm": 0.5629661679267883, "learning_rate": 8.999018879756764e-05, "loss": 0.0357, "step": 24470 }, { "grad_norm": 0.26851046085357666, "learning_rate": 8.998026146767487e-05, "loss": 0.0479, "step": 24480 }, { "grad_norm": 0.7253178358078003, "learning_rate": 8.99703297656008e-05, "loss": 0.0521, "step": 24490 }, { "grad_norm": 0.28851446509361267, "learning_rate": 8.996039369243156e-05, "loss": 0.0414, "step": 24500 }, { "grad_norm": 0.3041139245033264, "learning_rate": 8.995045324925378e-05, "loss": 0.041, "step": 24510 }, { "grad_norm": 0.7052353620529175, "learning_rate": 8.994050843715448e-05, "loss": 0.0561, "step": 24520 }, { "grad_norm": 0.5176982879638672, "learning_rate": 8.993055925722121e-05, "loss": 0.0489, "step": 24530 }, { "grad_norm": 0.422315388917923, "learning_rate": 8.992060571054202e-05, "loss": 0.0443, "step": 24540 }, { "grad_norm": 0.3351377546787262, "learning_rate": 8.991064779820542e-05, "loss": 0.0432, "step": 24550 }, { "grad_norm": 0.3372708261013031, "learning_rate": 8.990068552130036e-05, "loss": 0.0343, "step": 24560 }, { "grad_norm": 0.4317016899585724, "learning_rate": 8.989071888091634e-05, "loss": 0.043, "step": 24570 }, { "grad_norm": 0.338726669549942, "learning_rate": 8.988074787814329e-05, "loss": 0.0442, "step": 24580 }, { "grad_norm": 0.41069498658180237, "learning_rate": 8.987077251407158e-05, "loss": 0.0411, "step": 24590 }, { "grad_norm": 1.093336582183838, "learning_rate": 8.986079278979216e-05, "loss": 0.0463, "step": 24600 }, { "grad_norm": 0.9371621012687683, "learning_rate": 8.985080870639635e-05, "loss": 0.0519, "step": 24610 }, { "grad_norm": 0.5696249604225159, "learning_rate": 8.984082026497603e-05, "loss": 0.044, "step": 24620 }, { "grad_norm": 0.368210107088089, "learning_rate": 8.98308274666235e-05, "loss": 0.0702, "step": 24630 }, { "grad_norm": 0.4440128803253174, "learning_rate": 8.982083031243155e-05, "loss": 0.0535, "step": 24640 }, { "grad_norm": 0.5552011132240295, "learning_rate": 8.98108288034935e-05, "loss": 0.0571, "step": 24650 }, { "grad_norm": 0.38759225606918335, "learning_rate": 8.980082294090305e-05, "loss": 0.0458, "step": 24660 }, { "grad_norm": 0.35051214694976807, "learning_rate": 8.979081272575443e-05, "loss": 0.0508, "step": 24670 }, { "grad_norm": 0.4614960253238678, "learning_rate": 8.978079815914236e-05, "loss": 0.0502, "step": 24680 }, { "grad_norm": 0.39844438433647156, "learning_rate": 8.977077924216202e-05, "loss": 0.049, "step": 24690 }, { "grad_norm": 0.3821462094783783, "learning_rate": 8.976075597590905e-05, "loss": 0.0519, "step": 24700 }, { "grad_norm": 0.3955009877681732, "learning_rate": 8.975072836147958e-05, "loss": 0.0575, "step": 24710 }, { "grad_norm": 0.6051958799362183, "learning_rate": 8.974069639997025e-05, "loss": 0.0475, "step": 24720 }, { "grad_norm": 0.4526841640472412, "learning_rate": 8.973066009247808e-05, "loss": 0.0469, "step": 24730 }, { "grad_norm": 0.35406017303466797, "learning_rate": 8.972061944010066e-05, "loss": 0.0573, "step": 24740 }, { "grad_norm": 0.3151538670063019, "learning_rate": 8.971057444393603e-05, "loss": 0.0471, "step": 24750 }, { "grad_norm": 0.3783821165561676, "learning_rate": 8.970052510508268e-05, "loss": 0.047, "step": 24760 }, { "grad_norm": 0.36871010065078735, "learning_rate": 8.969047142463959e-05, "loss": 0.0535, "step": 24770 }, { "grad_norm": 0.528439998626709, "learning_rate": 8.968041340370621e-05, "loss": 0.0499, "step": 24780 }, { "grad_norm": 0.4055100679397583, "learning_rate": 8.96703510433825e-05, "loss": 0.0442, "step": 24790 }, { "grad_norm": 0.3724214434623718, "learning_rate": 8.966028434476883e-05, "loss": 0.0428, "step": 24800 }, { "grad_norm": 0.3133236765861511, "learning_rate": 8.96502133089661e-05, "loss": 0.0433, "step": 24810 }, { "grad_norm": 0.26295915246009827, "learning_rate": 8.964013793707564e-05, "loss": 0.0397, "step": 24820 }, { "grad_norm": 0.3902495205402374, "learning_rate": 8.963005823019932e-05, "loss": 0.0454, "step": 24830 }, { "grad_norm": 0.3477022051811218, "learning_rate": 8.961997418943939e-05, "loss": 0.045, "step": 24840 }, { "grad_norm": 0.283596009016037, "learning_rate": 8.960988581589865e-05, "loss": 0.0395, "step": 24850 }, { "grad_norm": 0.3743489980697632, "learning_rate": 8.959979311068037e-05, "loss": 0.0336, "step": 24860 }, { "grad_norm": 0.6521921753883362, "learning_rate": 8.958969607488823e-05, "loss": 0.04, "step": 24870 }, { "grad_norm": 0.273865669965744, "learning_rate": 8.957959470962647e-05, "loss": 0.039, "step": 24880 }, { "grad_norm": 0.26695430278778076, "learning_rate": 8.956948901599971e-05, "loss": 0.0419, "step": 24890 }, { "grad_norm": 0.2873345911502838, "learning_rate": 8.955937899511315e-05, "loss": 0.0413, "step": 24900 }, { "grad_norm": 0.2889424264431, "learning_rate": 8.954926464807238e-05, "loss": 0.0381, "step": 24910 }, { "grad_norm": 0.3493858277797699, "learning_rate": 8.953914597598347e-05, "loss": 0.0407, "step": 24920 }, { "grad_norm": 0.44471240043640137, "learning_rate": 8.952902297995303e-05, "loss": 0.0436, "step": 24930 }, { "grad_norm": 0.4331432580947876, "learning_rate": 8.951889566108804e-05, "loss": 0.0475, "step": 24940 }, { "grad_norm": 0.4039077162742615, "learning_rate": 8.950876402049606e-05, "loss": 0.0458, "step": 24950 }, { "grad_norm": 0.3365856409072876, "learning_rate": 8.949862805928504e-05, "loss": 0.0413, "step": 24960 }, { "grad_norm": 1.2405613660812378, "learning_rate": 8.948848777856343e-05, "loss": 0.0408, "step": 24970 }, { "grad_norm": 0.39274266362190247, "learning_rate": 8.947834317944017e-05, "loss": 0.0441, "step": 24980 }, { "grad_norm": 0.46231597661972046, "learning_rate": 8.946819426302466e-05, "loss": 0.0578, "step": 24990 }, { "grad_norm": 0.26082971692085266, "learning_rate": 8.945804103042676e-05, "loss": 0.0384, "step": 25000 }, { "grad_norm": 0.44920462369918823, "learning_rate": 8.944788348275681e-05, "loss": 0.0412, "step": 25010 }, { "grad_norm": 0.3903338313102722, "learning_rate": 8.943772162112565e-05, "loss": 0.0384, "step": 25020 }, { "grad_norm": 0.9360179901123047, "learning_rate": 8.942755544664454e-05, "loss": 0.0499, "step": 25030 }, { "grad_norm": 0.5452401041984558, "learning_rate": 8.941738496042525e-05, "loss": 0.0461, "step": 25040 }, { "grad_norm": 0.35956963896751404, "learning_rate": 8.940721016357999e-05, "loss": 0.0364, "step": 25050 }, { "grad_norm": 0.548710823059082, "learning_rate": 8.939703105722148e-05, "loss": 0.046, "step": 25060 }, { "grad_norm": 0.38518673181533813, "learning_rate": 8.93868476424629e-05, "loss": 0.0409, "step": 25070 }, { "grad_norm": 0.338151752948761, "learning_rate": 8.937665992041786e-05, "loss": 0.0444, "step": 25080 }, { "grad_norm": 0.5088019371032715, "learning_rate": 8.93664678922005e-05, "loss": 0.0478, "step": 25090 }, { "grad_norm": 0.32541945576667786, "learning_rate": 8.93562715589254e-05, "loss": 0.0452, "step": 25100 }, { "grad_norm": 0.3039392828941345, "learning_rate": 8.934607092170762e-05, "loss": 0.033, "step": 25110 }, { "grad_norm": 0.7164227366447449, "learning_rate": 8.933586598166266e-05, "loss": 0.046, "step": 25120 }, { "grad_norm": 0.31901103258132935, "learning_rate": 8.932565673990655e-05, "loss": 0.0588, "step": 25130 }, { "grad_norm": 0.6143819689750671, "learning_rate": 8.931544319755574e-05, "loss": 0.0529, "step": 25140 }, { "grad_norm": 0.2890176773071289, "learning_rate": 8.930522535572718e-05, "loss": 0.0388, "step": 25150 }, { "grad_norm": 0.39962849020957947, "learning_rate": 8.929500321553826e-05, "loss": 0.0466, "step": 25160 }, { "grad_norm": 0.45402055978775024, "learning_rate": 8.928477677810686e-05, "loss": 0.0462, "step": 25170 }, { "grad_norm": 0.40892282128334045, "learning_rate": 8.927454604455137e-05, "loss": 0.0383, "step": 25180 }, { "grad_norm": 0.40020957589149475, "learning_rate": 8.926431101599053e-05, "loss": 0.0476, "step": 25190 }, { "grad_norm": 0.3962438702583313, "learning_rate": 8.925407169354369e-05, "loss": 0.0393, "step": 25200 }, { "grad_norm": 0.38302338123321533, "learning_rate": 8.92438280783306e-05, "loss": 0.0445, "step": 25210 }, { "grad_norm": 1.4254690408706665, "learning_rate": 8.923358017147146e-05, "loss": 0.0493, "step": 25220 }, { "grad_norm": 0.6091961860656738, "learning_rate": 8.922332797408697e-05, "loss": 0.0444, "step": 25230 }, { "grad_norm": 0.28479015827178955, "learning_rate": 8.921307148729831e-05, "loss": 0.0446, "step": 25240 }, { "grad_norm": 0.36173591017723083, "learning_rate": 8.920281071222712e-05, "loss": 0.0328, "step": 25250 }, { "grad_norm": 0.3134620189666748, "learning_rate": 8.919254564999548e-05, "loss": 0.0432, "step": 25260 }, { "grad_norm": 0.3640928268432617, "learning_rate": 8.918227630172598e-05, "loss": 0.0374, "step": 25270 }, { "grad_norm": 0.25004562735557556, "learning_rate": 8.917200266854165e-05, "loss": 0.0417, "step": 25280 }, { "grad_norm": 0.25460323691368103, "learning_rate": 8.9161724751566e-05, "loss": 0.0413, "step": 25290 }, { "grad_norm": 0.6006272435188293, "learning_rate": 8.915144255192302e-05, "loss": 0.0483, "step": 25300 }, { "grad_norm": 0.27635201811790466, "learning_rate": 8.914115607073714e-05, "loss": 0.0458, "step": 25310 }, { "grad_norm": 0.4336325228214264, "learning_rate": 8.913086530913327e-05, "loss": 0.034, "step": 25320 }, { "grad_norm": 0.37953197956085205, "learning_rate": 8.912057026823681e-05, "loss": 0.0436, "step": 25330 }, { "grad_norm": 0.33730947971343994, "learning_rate": 8.91102709491736e-05, "loss": 0.0409, "step": 25340 }, { "grad_norm": 0.338186651468277, "learning_rate": 8.909996735306996e-05, "loss": 0.0358, "step": 25350 }, { "grad_norm": 0.25019240379333496, "learning_rate": 8.908965948105268e-05, "loss": 0.0365, "step": 25360 }, { "grad_norm": 0.28457367420196533, "learning_rate": 8.907934733424901e-05, "loss": 0.0447, "step": 25370 }, { "grad_norm": 0.37669941782951355, "learning_rate": 8.906903091378666e-05, "loss": 0.0432, "step": 25380 }, { "grad_norm": 0.30941981077194214, "learning_rate": 8.905871022079384e-05, "loss": 0.0373, "step": 25390 }, { "grad_norm": 0.31130626797676086, "learning_rate": 8.90483852563992e-05, "loss": 0.0325, "step": 25400 }, { "grad_norm": 0.26513752341270447, "learning_rate": 8.903805602173185e-05, "loss": 0.0383, "step": 25410 }, { "grad_norm": 0.6487002968788147, "learning_rate": 8.902772251792137e-05, "loss": 0.036, "step": 25420 }, { "grad_norm": 0.4634423851966858, "learning_rate": 8.901738474609786e-05, "loss": 0.0451, "step": 25430 }, { "grad_norm": 0.21798627078533173, "learning_rate": 8.900704270739179e-05, "loss": 0.0374, "step": 25440 }, { "grad_norm": 0.39668387174606323, "learning_rate": 8.89966964029342e-05, "loss": 0.0551, "step": 25450 }, { "grad_norm": 0.4230329096317291, "learning_rate": 8.898634583385652e-05, "loss": 0.0441, "step": 25460 }, { "grad_norm": 0.32162028551101685, "learning_rate": 8.897599100129065e-05, "loss": 0.0406, "step": 25470 }, { "grad_norm": 0.3356838822364807, "learning_rate": 8.896563190636903e-05, "loss": 0.033, "step": 25480 }, { "grad_norm": 0.39769473671913147, "learning_rate": 8.895526855022448e-05, "loss": 0.0407, "step": 25490 }, { "grad_norm": 0.2926677465438843, "learning_rate": 8.894490093399033e-05, "loss": 0.0432, "step": 25500 }, { "grad_norm": 0.3117738962173462, "learning_rate": 8.893452905880035e-05, "loss": 0.0367, "step": 25510 }, { "grad_norm": 0.3504948318004608, "learning_rate": 8.892415292578883e-05, "loss": 0.0407, "step": 25520 }, { "grad_norm": 0.3441716730594635, "learning_rate": 8.891377253609046e-05, "loss": 0.053, "step": 25530 }, { "grad_norm": 0.3295609652996063, "learning_rate": 8.890338789084043e-05, "loss": 0.045, "step": 25540 }, { "grad_norm": 0.47248077392578125, "learning_rate": 8.88929989911744e-05, "loss": 0.0421, "step": 25550 }, { "grad_norm": 0.35595420002937317, "learning_rate": 8.888260583822847e-05, "loss": 0.0543, "step": 25560 }, { "grad_norm": 0.2758628726005554, "learning_rate": 8.887220843313921e-05, "loss": 0.044, "step": 25570 }, { "grad_norm": 0.2710055112838745, "learning_rate": 8.88618067770437e-05, "loss": 0.0458, "step": 25580 }, { "grad_norm": 0.38678890466690063, "learning_rate": 8.885140087107942e-05, "loss": 0.0478, "step": 25590 }, { "grad_norm": 0.31707388162612915, "learning_rate": 8.884099071638436e-05, "loss": 0.0445, "step": 25600 }, { "grad_norm": 0.5179728269577026, "learning_rate": 8.883057631409695e-05, "loss": 0.0464, "step": 25610 }, { "grad_norm": 0.2776808738708496, "learning_rate": 8.882015766535608e-05, "loss": 0.0392, "step": 25620 }, { "grad_norm": 0.33650273084640503, "learning_rate": 8.880973477130115e-05, "loss": 0.0379, "step": 25630 }, { "grad_norm": 0.4065229892730713, "learning_rate": 8.879930763307197e-05, "loss": 0.0435, "step": 25640 }, { "grad_norm": 0.4449862837791443, "learning_rate": 8.878887625180884e-05, "loss": 0.0444, "step": 25650 }, { "grad_norm": 0.31391143798828125, "learning_rate": 8.877844062865253e-05, "loss": 0.0375, "step": 25660 }, { "grad_norm": 0.2680973708629608, "learning_rate": 8.876800076474424e-05, "loss": 0.0539, "step": 25670 }, { "grad_norm": 0.28434962034225464, "learning_rate": 8.875755666122568e-05, "loss": 0.041, "step": 25680 }, { "grad_norm": 0.4654112756252289, "learning_rate": 8.8747108319239e-05, "loss": 0.0317, "step": 25690 }, { "grad_norm": 0.3717789947986603, "learning_rate": 8.87366557399268e-05, "loss": 0.0383, "step": 25700 }, { "grad_norm": 0.27595341205596924, "learning_rate": 8.872619892443217e-05, "loss": 0.0354, "step": 25710 }, { "grad_norm": 0.22119243443012238, "learning_rate": 8.871573787389865e-05, "loss": 0.0339, "step": 25720 }, { "grad_norm": 0.32708823680877686, "learning_rate": 8.870527258947024e-05, "loss": 0.0486, "step": 25730 }, { "grad_norm": 0.34193238615989685, "learning_rate": 8.869480307229143e-05, "loss": 0.0495, "step": 25740 }, { "grad_norm": 0.34654465317726135, "learning_rate": 8.868432932350712e-05, "loss": 0.0363, "step": 25750 }, { "grad_norm": 0.3746630549430847, "learning_rate": 8.867385134426272e-05, "loss": 0.0464, "step": 25760 }, { "grad_norm": 0.24313673377037048, "learning_rate": 8.866336913570407e-05, "loss": 0.0382, "step": 25770 }, { "grad_norm": 0.49939125776290894, "learning_rate": 8.865288269897751e-05, "loss": 0.033, "step": 25780 }, { "grad_norm": 0.26965659856796265, "learning_rate": 8.864239203522981e-05, "loss": 0.0346, "step": 25790 }, { "grad_norm": 0.40156289935112, "learning_rate": 8.863189714560822e-05, "loss": 0.0343, "step": 25800 }, { "grad_norm": 0.4394378364086151, "learning_rate": 8.862139803126043e-05, "loss": 0.0395, "step": 25810 }, { "grad_norm": 0.2505299746990204, "learning_rate": 8.861089469333463e-05, "loss": 0.0322, "step": 25820 }, { "grad_norm": 0.23819255828857422, "learning_rate": 8.860038713297944e-05, "loss": 0.0363, "step": 25830 }, { "grad_norm": 0.43868520855903625, "learning_rate": 8.858987535134394e-05, "loss": 0.0459, "step": 25840 }, { "grad_norm": 0.30713003873825073, "learning_rate": 8.857935934957769e-05, "loss": 0.0443, "step": 25850 }, { "grad_norm": 0.4636926054954529, "learning_rate": 8.856883912883071e-05, "loss": 0.0371, "step": 25860 }, { "grad_norm": 0.47885772585868835, "learning_rate": 8.855831469025346e-05, "loss": 0.04, "step": 25870 }, { "grad_norm": 0.4559437930583954, "learning_rate": 8.854778603499689e-05, "loss": 0.0463, "step": 25880 }, { "grad_norm": 0.593222975730896, "learning_rate": 8.85372531642124e-05, "loss": 0.0365, "step": 25890 }, { "grad_norm": 0.38318881392478943, "learning_rate": 8.852671607905185e-05, "loss": 0.0349, "step": 25900 }, { "grad_norm": 0.5305522084236145, "learning_rate": 8.851617478066754e-05, "loss": 0.0415, "step": 25910 }, { "grad_norm": 0.5284160375595093, "learning_rate": 8.850562927021227e-05, "loss": 0.0366, "step": 25920 }, { "grad_norm": 0.5247029662132263, "learning_rate": 8.849507954883928e-05, "loss": 0.0363, "step": 25930 }, { "grad_norm": 0.3227526843547821, "learning_rate": 8.848452561770226e-05, "loss": 0.0433, "step": 25940 }, { "grad_norm": 0.3611487150192261, "learning_rate": 8.847396747795538e-05, "loss": 0.0416, "step": 25950 }, { "grad_norm": 0.8237918615341187, "learning_rate": 8.846340513075327e-05, "loss": 0.0385, "step": 25960 }, { "grad_norm": 0.49349677562713623, "learning_rate": 8.845283857725099e-05, "loss": 0.0396, "step": 25970 }, { "grad_norm": 0.25266769528388977, "learning_rate": 8.844226781860409e-05, "loss": 0.037, "step": 25980 }, { "grad_norm": 0.3646477460861206, "learning_rate": 8.84316928559686e-05, "loss": 0.0478, "step": 25990 }, { "grad_norm": 0.34904223680496216, "learning_rate": 8.842111369050094e-05, "loss": 0.0464, "step": 26000 }, { "grad_norm": 0.30375102162361145, "learning_rate": 8.841053032335808e-05, "loss": 0.0474, "step": 26010 }, { "grad_norm": 0.3123350441455841, "learning_rate": 8.839994275569735e-05, "loss": 0.0435, "step": 26020 }, { "grad_norm": 0.42613038420677185, "learning_rate": 8.838935098867662e-05, "loss": 0.0477, "step": 26030 }, { "grad_norm": 0.3388694226741791, "learning_rate": 8.837875502345418e-05, "loss": 0.0531, "step": 26040 }, { "grad_norm": 0.3483731150627136, "learning_rate": 8.83681548611888e-05, "loss": 0.0465, "step": 26050 }, { "grad_norm": 0.7310535907745361, "learning_rate": 8.835755050303969e-05, "loss": 0.0382, "step": 26060 }, { "grad_norm": 0.4002419710159302, "learning_rate": 8.834694195016653e-05, "loss": 0.0511, "step": 26070 }, { "grad_norm": 0.5181463360786438, "learning_rate": 8.833632920372942e-05, "loss": 0.0616, "step": 26080 }, { "grad_norm": 0.8811393976211548, "learning_rate": 8.832571226488903e-05, "loss": 0.0601, "step": 26090 }, { "grad_norm": 0.5230762958526611, "learning_rate": 8.831509113480634e-05, "loss": 0.0649, "step": 26100 }, { "grad_norm": 0.32356762886047363, "learning_rate": 8.83044658146429e-05, "loss": 0.0372, "step": 26110 }, { "grad_norm": 0.3452855050563812, "learning_rate": 8.829383630556067e-05, "loss": 0.0452, "step": 26120 }, { "grad_norm": 0.2379506677389145, "learning_rate": 8.828320260872207e-05, "loss": 0.043, "step": 26130 }, { "grad_norm": 0.5188124179840088, "learning_rate": 8.827256472529e-05, "loss": 0.0472, "step": 26140 }, { "grad_norm": 0.30282267928123474, "learning_rate": 8.826192265642778e-05, "loss": 0.0545, "step": 26150 }, { "grad_norm": 0.38485684990882874, "learning_rate": 8.825127640329923e-05, "loss": 0.0396, "step": 26160 }, { "grad_norm": 0.3005872964859009, "learning_rate": 8.824062596706861e-05, "loss": 0.0483, "step": 26170 }, { "grad_norm": 0.2806885540485382, "learning_rate": 8.822997134890062e-05, "loss": 0.0402, "step": 26180 }, { "grad_norm": 0.38325631618499756, "learning_rate": 8.821931254996044e-05, "loss": 0.0422, "step": 26190 }, { "grad_norm": 0.27494749426841736, "learning_rate": 8.82086495714137e-05, "loss": 0.0515, "step": 26200 }, { "grad_norm": 0.6173917651176453, "learning_rate": 8.81979824144265e-05, "loss": 0.0459, "step": 26210 }, { "grad_norm": 0.4139973819255829, "learning_rate": 8.818731108016536e-05, "loss": 0.0391, "step": 26220 }, { "grad_norm": 0.4730604887008667, "learning_rate": 8.81766355697973e-05, "loss": 0.0422, "step": 26230 }, { "grad_norm": 0.3996771574020386, "learning_rate": 8.816595588448977e-05, "loss": 0.0482, "step": 26240 }, { "grad_norm": 0.3031848967075348, "learning_rate": 8.81552720254107e-05, "loss": 0.0388, "step": 26250 }, { "grad_norm": 0.4577155113220215, "learning_rate": 8.814458399372842e-05, "loss": 0.0443, "step": 26260 }, { "grad_norm": 0.232248455286026, "learning_rate": 8.813389179061181e-05, "loss": 0.0376, "step": 26270 }, { "grad_norm": 0.5156363844871521, "learning_rate": 8.812319541723012e-05, "loss": 0.0471, "step": 26280 }, { "grad_norm": 0.2925766110420227, "learning_rate": 8.811249487475309e-05, "loss": 0.0515, "step": 26290 }, { "grad_norm": 0.27503371238708496, "learning_rate": 8.810179016435092e-05, "loss": 0.0421, "step": 26300 }, { "grad_norm": 0.24983428418636322, "learning_rate": 8.809108128719428e-05, "loss": 0.0496, "step": 26310 }, { "grad_norm": 0.4591754972934723, "learning_rate": 8.808036824445424e-05, "loss": 0.0432, "step": 26320 }, { "grad_norm": 0.5513346791267395, "learning_rate": 8.806965103730238e-05, "loss": 0.0472, "step": 26330 }, { "grad_norm": 0.4407426714897156, "learning_rate": 8.805892966691074e-05, "loss": 0.037, "step": 26340 }, { "grad_norm": 0.3832164406776428, "learning_rate": 8.804820413445175e-05, "loss": 0.0377, "step": 26350 }, { "grad_norm": 0.4522498548030853, "learning_rate": 8.803747444109837e-05, "loss": 0.0425, "step": 26360 }, { "grad_norm": 0.25652897357940674, "learning_rate": 8.802674058802399e-05, "loss": 0.0391, "step": 26370 }, { "grad_norm": 0.36020877957344055, "learning_rate": 8.801600257640241e-05, "loss": 0.0412, "step": 26380 }, { "grad_norm": 0.5350365042686462, "learning_rate": 8.800526040740795e-05, "loss": 0.0374, "step": 26390 }, { "grad_norm": 0.4546343684196472, "learning_rate": 8.799451408221535e-05, "loss": 0.0344, "step": 26400 }, { "grad_norm": 0.23790761828422546, "learning_rate": 8.798376360199982e-05, "loss": 0.0365, "step": 26410 }, { "grad_norm": 0.328619122505188, "learning_rate": 8.797300896793701e-05, "loss": 0.0325, "step": 26420 }, { "grad_norm": 0.27436885237693787, "learning_rate": 8.796225018120302e-05, "loss": 0.04, "step": 26430 }, { "grad_norm": 0.3473621904850006, "learning_rate": 8.795148724297444e-05, "loss": 0.0448, "step": 26440 }, { "grad_norm": 0.31978073716163635, "learning_rate": 8.794072015442825e-05, "loss": 0.0426, "step": 26450 }, { "grad_norm": 0.4416113793849945, "learning_rate": 8.792994891674198e-05, "loss": 0.0456, "step": 26460 }, { "grad_norm": 0.40445154905319214, "learning_rate": 8.79191735310935e-05, "loss": 0.0385, "step": 26470 }, { "grad_norm": 0.2891925871372223, "learning_rate": 8.790839399866122e-05, "loss": 0.0352, "step": 26480 }, { "grad_norm": 0.4511677920818329, "learning_rate": 8.789761032062397e-05, "loss": 0.038, "step": 26490 }, { "grad_norm": 0.4227549731731415, "learning_rate": 8.788682249816103e-05, "loss": 0.0459, "step": 26500 }, { "grad_norm": 0.2968071699142456, "learning_rate": 8.787603053245215e-05, "loss": 0.0477, "step": 26510 }, { "grad_norm": 0.24025948345661163, "learning_rate": 8.78652344246775e-05, "loss": 0.0395, "step": 26520 }, { "grad_norm": 0.3313574194908142, "learning_rate": 8.785443417601776e-05, "loss": 0.0398, "step": 26530 }, { "grad_norm": 0.2845149338245392, "learning_rate": 8.784362978765401e-05, "loss": 0.0402, "step": 26540 }, { "grad_norm": 0.5283274054527283, "learning_rate": 8.783282126076779e-05, "loss": 0.0445, "step": 26550 }, { "grad_norm": 0.468206524848938, "learning_rate": 8.782200859654112e-05, "loss": 0.0392, "step": 26560 }, { "grad_norm": 0.32265013456344604, "learning_rate": 8.781119179615646e-05, "loss": 0.0396, "step": 26570 }, { "grad_norm": 0.3929537534713745, "learning_rate": 8.780037086079674e-05, "loss": 0.0466, "step": 26580 }, { "grad_norm": 0.33842968940734863, "learning_rate": 8.778954579164527e-05, "loss": 0.0378, "step": 26590 }, { "grad_norm": 0.28163251280784607, "learning_rate": 8.777871658988588e-05, "loss": 0.0398, "step": 26600 }, { "grad_norm": 0.2990071475505829, "learning_rate": 8.776788325670285e-05, "loss": 0.0325, "step": 26610 }, { "grad_norm": 0.32574185729026794, "learning_rate": 8.775704579328089e-05, "loss": 0.0429, "step": 26620 }, { "grad_norm": 0.9909359812736511, "learning_rate": 8.774620420080517e-05, "loss": 0.0431, "step": 26630 }, { "grad_norm": 0.39446815848350525, "learning_rate": 8.773535848046131e-05, "loss": 0.0421, "step": 26640 }, { "grad_norm": 0.36362165212631226, "learning_rate": 8.772450863343538e-05, "loss": 0.0417, "step": 26650 }, { "grad_norm": 0.3769621253013611, "learning_rate": 8.77136546609139e-05, "loss": 0.0492, "step": 26660 }, { "grad_norm": 0.3443608283996582, "learning_rate": 8.770279656408385e-05, "loss": 0.0432, "step": 26670 }, { "grad_norm": 0.3942749798297882, "learning_rate": 8.769193434413265e-05, "loss": 0.0416, "step": 26680 }, { "grad_norm": 0.33254191279411316, "learning_rate": 8.76810680022482e-05, "loss": 0.0383, "step": 26690 }, { "grad_norm": 0.4055637717247009, "learning_rate": 8.767019753961878e-05, "loss": 0.0501, "step": 26700 }, { "grad_norm": 0.4413089454174042, "learning_rate": 8.765932295743321e-05, "loss": 0.0394, "step": 26710 }, { "grad_norm": 0.8591364622116089, "learning_rate": 8.764844425688068e-05, "loss": 0.0445, "step": 26720 }, { "grad_norm": 0.313022255897522, "learning_rate": 8.763756143915092e-05, "loss": 0.0384, "step": 26730 }, { "grad_norm": 0.3981771767139435, "learning_rate": 8.7626674505434e-05, "loss": 0.0426, "step": 26740 }, { "grad_norm": 0.3376806378364563, "learning_rate": 8.761578345692053e-05, "loss": 0.0571, "step": 26750 }, { "grad_norm": 0.38092702627182007, "learning_rate": 8.760488829480156e-05, "loss": 0.0473, "step": 26760 }, { "grad_norm": 0.36036357283592224, "learning_rate": 8.759398902026854e-05, "loss": 0.0321, "step": 26770 }, { "grad_norm": 0.26876410841941833, "learning_rate": 8.758308563451339e-05, "loss": 0.0367, "step": 26780 }, { "grad_norm": 0.301434189081192, "learning_rate": 8.75721781387285e-05, "loss": 0.0437, "step": 26790 }, { "grad_norm": 0.30764180421829224, "learning_rate": 8.75612665341067e-05, "loss": 0.0388, "step": 26800 }, { "grad_norm": 0.26075270771980286, "learning_rate": 8.755035082184126e-05, "loss": 0.0337, "step": 26810 }, { "grad_norm": 0.31689298152923584, "learning_rate": 8.753943100312592e-05, "loss": 0.0332, "step": 26820 }, { "grad_norm": 0.43287166953086853, "learning_rate": 8.752850707915484e-05, "loss": 0.039, "step": 26830 }, { "grad_norm": 0.33745795488357544, "learning_rate": 8.751757905112264e-05, "loss": 0.0452, "step": 26840 }, { "grad_norm": 0.5965906977653503, "learning_rate": 8.75066469202244e-05, "loss": 0.0412, "step": 26850 }, { "grad_norm": 0.39100614190101624, "learning_rate": 8.749571068765567e-05, "loss": 0.0475, "step": 26860 }, { "grad_norm": 0.3579765558242798, "learning_rate": 8.748477035461238e-05, "loss": 0.0425, "step": 26870 }, { "grad_norm": 0.46788308024406433, "learning_rate": 8.747382592229095e-05, "loss": 0.0442, "step": 26880 }, { "grad_norm": 0.30218061804771423, "learning_rate": 8.746287739188828e-05, "loss": 0.0371, "step": 26890 }, { "grad_norm": 0.40579211711883545, "learning_rate": 8.745192476460165e-05, "loss": 0.0406, "step": 26900 }, { "grad_norm": 0.24939149618148804, "learning_rate": 8.744096804162882e-05, "loss": 0.0447, "step": 26910 }, { "grad_norm": 0.41942232847213745, "learning_rate": 8.743000722416804e-05, "loss": 0.0434, "step": 26920 }, { "grad_norm": 0.33743640780448914, "learning_rate": 8.741904231341793e-05, "loss": 0.04, "step": 26930 }, { "grad_norm": 0.46519914269447327, "learning_rate": 8.740807331057762e-05, "loss": 0.0569, "step": 26940 }, { "grad_norm": 0.3518822193145752, "learning_rate": 8.739710021684667e-05, "loss": 0.0462, "step": 26950 }, { "grad_norm": 0.5243620276451111, "learning_rate": 8.738612303342503e-05, "loss": 0.0386, "step": 26960 }, { "grad_norm": 0.22087489068508148, "learning_rate": 8.73751417615132e-05, "loss": 0.0503, "step": 26970 }, { "grad_norm": 0.36140570044517517, "learning_rate": 8.736415640231208e-05, "loss": 0.0426, "step": 26980 }, { "grad_norm": 0.21771861612796783, "learning_rate": 8.735316695702297e-05, "loss": 0.0406, "step": 26990 }, { "grad_norm": 0.2906704246997833, "learning_rate": 8.734217342684769e-05, "loss": 0.0389, "step": 27000 }, { "grad_norm": 0.35073569416999817, "learning_rate": 8.733117581298847e-05, "loss": 0.0357, "step": 27010 }, { "grad_norm": 0.21863287687301636, "learning_rate": 8.732017411664796e-05, "loss": 0.0393, "step": 27020 }, { "grad_norm": 0.5857628583908081, "learning_rate": 8.730916833902936e-05, "loss": 0.0406, "step": 27030 }, { "grad_norm": 0.34702712297439575, "learning_rate": 8.729815848133618e-05, "loss": 0.0451, "step": 27040 }, { "grad_norm": 0.2729613184928894, "learning_rate": 8.728714454477247e-05, "loss": 0.0399, "step": 27050 }, { "grad_norm": 0.36221277713775635, "learning_rate": 8.727612653054269e-05, "loss": 0.0459, "step": 27060 }, { "grad_norm": 0.24080924689769745, "learning_rate": 8.726510443985176e-05, "loss": 0.0391, "step": 27070 }, { "grad_norm": 0.23702049255371094, "learning_rate": 8.725407827390503e-05, "loss": 0.0424, "step": 27080 }, { "grad_norm": 0.26858633756637573, "learning_rate": 8.724304803390833e-05, "loss": 0.0409, "step": 27090 }, { "grad_norm": 0.39326372742652893, "learning_rate": 8.723201372106788e-05, "loss": 0.0389, "step": 27100 }, { "grad_norm": 0.4148292541503906, "learning_rate": 8.722097533659038e-05, "loss": 0.0413, "step": 27110 }, { "grad_norm": 0.3574638366699219, "learning_rate": 8.720993288168299e-05, "loss": 0.0524, "step": 27120 }, { "grad_norm": 0.35399118065834045, "learning_rate": 8.719888635755327e-05, "loss": 0.037, "step": 27130 }, { "grad_norm": 0.5004224181175232, "learning_rate": 8.718783576540928e-05, "loss": 0.0475, "step": 27140 }, { "grad_norm": 0.29325637221336365, "learning_rate": 8.717678110645948e-05, "loss": 0.0327, "step": 27150 }, { "grad_norm": 0.300462543964386, "learning_rate": 8.716572238191279e-05, "loss": 0.0377, "step": 27160 }, { "grad_norm": 0.48965662717819214, "learning_rate": 8.715465959297857e-05, "loss": 0.042, "step": 27170 }, { "grad_norm": 0.3825879991054535, "learning_rate": 8.714359274086665e-05, "loss": 0.0375, "step": 27180 }, { "grad_norm": 0.29604291915893555, "learning_rate": 8.713252182678726e-05, "loss": 0.0449, "step": 27190 }, { "grad_norm": 0.2814534306526184, "learning_rate": 8.712144685195112e-05, "loss": 0.0519, "step": 27200 }, { "grad_norm": 0.34949761629104614, "learning_rate": 8.711036781756936e-05, "loss": 0.0342, "step": 27210 }, { "grad_norm": 0.32311680912971497, "learning_rate": 8.709928472485357e-05, "loss": 0.0401, "step": 27220 }, { "grad_norm": 0.2982695698738098, "learning_rate": 8.708819757501579e-05, "loss": 0.0416, "step": 27230 }, { "grad_norm": 0.4406491816043854, "learning_rate": 8.707710636926846e-05, "loss": 0.041, "step": 27240 }, { "grad_norm": 0.33284083008766174, "learning_rate": 8.706601110882455e-05, "loss": 0.0589, "step": 27250 }, { "grad_norm": 0.4798690974712372, "learning_rate": 8.705491179489738e-05, "loss": 0.0403, "step": 27260 }, { "grad_norm": 0.3624672591686249, "learning_rate": 8.704380842870077e-05, "loss": 0.0318, "step": 27270 }, { "grad_norm": 0.7830280065536499, "learning_rate": 8.703270101144895e-05, "loss": 0.0487, "step": 27280 }, { "grad_norm": 0.25996318459510803, "learning_rate": 8.702158954435664e-05, "loss": 0.0312, "step": 27290 }, { "grad_norm": 0.9941256046295166, "learning_rate": 8.701047402863896e-05, "loss": 0.0374, "step": 27300 }, { "grad_norm": 0.8899745941162109, "learning_rate": 8.699935446551148e-05, "loss": 0.0434, "step": 27310 }, { "grad_norm": 0.47278469800949097, "learning_rate": 8.698823085619022e-05, "loss": 0.0462, "step": 27320 }, { "grad_norm": 0.33603182435035706, "learning_rate": 8.697710320189166e-05, "loss": 0.0364, "step": 27330 }, { "grad_norm": 0.28328946232795715, "learning_rate": 8.696597150383268e-05, "loss": 0.0365, "step": 27340 }, { "grad_norm": 0.2544209063053131, "learning_rate": 8.695483576323063e-05, "loss": 0.0391, "step": 27350 }, { "grad_norm": 0.5323811173439026, "learning_rate": 8.69436959813033e-05, "loss": 0.0503, "step": 27360 }, { "grad_norm": 0.2633238732814789, "learning_rate": 8.693255215926892e-05, "loss": 0.0411, "step": 27370 }, { "grad_norm": 0.468659371137619, "learning_rate": 8.692140429834617e-05, "loss": 0.0431, "step": 27380 }, { "grad_norm": 0.24431008100509644, "learning_rate": 8.691025239975415e-05, "loss": 0.0331, "step": 27390 }, { "grad_norm": 0.2428765892982483, "learning_rate": 8.689909646471243e-05, "loss": 0.0322, "step": 27400 }, { "grad_norm": 0.20977704226970673, "learning_rate": 8.688793649444099e-05, "loss": 0.0399, "step": 27410 }, { "grad_norm": 0.41398584842681885, "learning_rate": 8.687677249016029e-05, "loss": 0.0314, "step": 27420 }, { "grad_norm": 0.483484148979187, "learning_rate": 8.686560445309118e-05, "loss": 0.0407, "step": 27430 }, { "grad_norm": 0.39162978529930115, "learning_rate": 8.685443238445499e-05, "loss": 0.0405, "step": 27440 }, { "grad_norm": 0.24794816970825195, "learning_rate": 8.68432562854735e-05, "loss": 0.0351, "step": 27450 }, { "grad_norm": 0.2513883113861084, "learning_rate": 8.683207615736887e-05, "loss": 0.0468, "step": 27460 }, { "grad_norm": 0.35421663522720337, "learning_rate": 8.682089200136379e-05, "loss": 0.0481, "step": 27470 }, { "grad_norm": 0.294297456741333, "learning_rate": 8.680970381868132e-05, "loss": 0.0389, "step": 27480 }, { "grad_norm": 0.33903375267982483, "learning_rate": 8.679851161054498e-05, "loss": 0.0376, "step": 27490 }, { "grad_norm": 0.3866511881351471, "learning_rate": 8.678731537817873e-05, "loss": 0.0486, "step": 27500 }, { "grad_norm": 0.3274959921836853, "learning_rate": 8.677611512280697e-05, "loss": 0.0331, "step": 27510 }, { "grad_norm": 0.2554576098918915, "learning_rate": 8.676491084565457e-05, "loss": 0.037, "step": 27520 }, { "grad_norm": 0.26722317934036255, "learning_rate": 8.675370254794678e-05, "loss": 0.0333, "step": 27530 }, { "grad_norm": 0.3589686453342438, "learning_rate": 8.674249023090935e-05, "loss": 0.035, "step": 27540 }, { "grad_norm": 0.3474130630493164, "learning_rate": 8.673127389576843e-05, "loss": 0.0343, "step": 27550 }, { "grad_norm": 0.3947076201438904, "learning_rate": 8.67200535437506e-05, "loss": 0.0303, "step": 27560 }, { "grad_norm": 0.2989254891872406, "learning_rate": 8.670882917608296e-05, "loss": 0.0319, "step": 27570 }, { "grad_norm": 0.20864346623420715, "learning_rate": 8.669760079399292e-05, "loss": 0.0316, "step": 27580 }, { "grad_norm": 0.32424861192703247, "learning_rate": 8.668636839870845e-05, "loss": 0.0403, "step": 27590 }, { "grad_norm": 2.1041882038116455, "learning_rate": 8.667513199145789e-05, "loss": 0.0477, "step": 27600 }, { "grad_norm": 0.40220117568969727, "learning_rate": 8.666389157347002e-05, "loss": 0.0444, "step": 27610 }, { "grad_norm": 0.3711780607700348, "learning_rate": 8.66526471459741e-05, "loss": 0.0459, "step": 27620 }, { "grad_norm": 0.3465292751789093, "learning_rate": 8.66413987101998e-05, "loss": 0.0436, "step": 27630 }, { "grad_norm": 0.31180647015571594, "learning_rate": 8.663014626737723e-05, "loss": 0.0457, "step": 27640 }, { "grad_norm": 0.44068583846092224, "learning_rate": 8.661888981873691e-05, "loss": 0.0569, "step": 27650 }, { "grad_norm": 0.36771079897880554, "learning_rate": 8.660762936550988e-05, "loss": 0.0596, "step": 27660 }, { "grad_norm": 0.7423506379127502, "learning_rate": 8.659636490892753e-05, "loss": 0.0598, "step": 27670 }, { "grad_norm": 0.46343013644218445, "learning_rate": 8.658509645022174e-05, "loss": 0.0544, "step": 27680 }, { "grad_norm": 0.4414062201976776, "learning_rate": 8.657382399062481e-05, "loss": 0.0385, "step": 27690 }, { "grad_norm": 0.673509418964386, "learning_rate": 8.656254753136946e-05, "loss": 0.0429, "step": 27700 }, { "grad_norm": 0.3379499912261963, "learning_rate": 8.655126707368891e-05, "loss": 0.0324, "step": 27710 }, { "grad_norm": 0.3005549907684326, "learning_rate": 8.653998261881672e-05, "loss": 0.0315, "step": 27720 }, { "grad_norm": 0.44532889127731323, "learning_rate": 8.652869416798699e-05, "loss": 0.0453, "step": 27730 }, { "grad_norm": 0.6483521461486816, "learning_rate": 8.651740172243417e-05, "loss": 0.0445, "step": 27740 }, { "grad_norm": 0.28195831179618835, "learning_rate": 8.65061052833932e-05, "loss": 0.035, "step": 27750 }, { "grad_norm": 0.36005139350891113, "learning_rate": 8.649480485209945e-05, "loss": 0.0346, "step": 27760 }, { "grad_norm": 0.29433420300483704, "learning_rate": 8.64835004297887e-05, "loss": 0.0409, "step": 27770 }, { "grad_norm": 0.2934008538722992, "learning_rate": 8.64721920176972e-05, "loss": 0.0399, "step": 27780 }, { "grad_norm": 0.2860315442085266, "learning_rate": 8.646087961706164e-05, "loss": 0.037, "step": 27790 }, { "grad_norm": 0.41959038376808167, "learning_rate": 8.644956322911908e-05, "loss": 0.0442, "step": 27800 }, { "grad_norm": 0.3914596736431122, "learning_rate": 8.643824285510709e-05, "loss": 0.0453, "step": 27810 }, { "grad_norm": 0.3736063241958618, "learning_rate": 8.642691849626364e-05, "loss": 0.0368, "step": 27820 }, { "grad_norm": 0.2690572142601013, "learning_rate": 8.641559015382717e-05, "loss": 0.0371, "step": 27830 }, { "grad_norm": 0.29341477155685425, "learning_rate": 8.640425782903649e-05, "loss": 0.0403, "step": 27840 }, { "grad_norm": 0.2624547481536865, "learning_rate": 8.639292152313091e-05, "loss": 0.0353, "step": 27850 }, { "grad_norm": 0.21532003581523895, "learning_rate": 8.638158123735015e-05, "loss": 0.0365, "step": 27860 }, { "grad_norm": 0.30544930696487427, "learning_rate": 8.637023697293436e-05, "loss": 0.0299, "step": 27870 }, { "grad_norm": 0.4543147385120392, "learning_rate": 8.635888873112414e-05, "loss": 0.0557, "step": 27880 }, { "grad_norm": 0.35529825091362, "learning_rate": 8.634753651316052e-05, "loss": 0.038, "step": 27890 }, { "grad_norm": 0.30388975143432617, "learning_rate": 8.633618032028496e-05, "loss": 0.0382, "step": 27900 }, { "grad_norm": 0.4562527537345886, "learning_rate": 8.632482015373934e-05, "loss": 0.0419, "step": 27910 }, { "grad_norm": 0.3267274498939514, "learning_rate": 8.6313456014766e-05, "loss": 0.0309, "step": 27920 }, { "grad_norm": 0.4618401825428009, "learning_rate": 8.630208790460771e-05, "loss": 0.0424, "step": 27930 }, { "grad_norm": 0.2862476110458374, "learning_rate": 8.629071582450768e-05, "loss": 0.0368, "step": 27940 }, { "grad_norm": 0.5342169404029846, "learning_rate": 8.62793397757095e-05, "loss": 0.0403, "step": 27950 }, { "grad_norm": 0.354917049407959, "learning_rate": 8.626795975945729e-05, "loss": 0.0512, "step": 27960 }, { "grad_norm": 0.3987850546836853, "learning_rate": 8.625657577699551e-05, "loss": 0.0428, "step": 27970 }, { "grad_norm": 0.23664847016334534, "learning_rate": 8.624518782956914e-05, "loss": 0.0293, "step": 27980 }, { "grad_norm": 0.46080151200294495, "learning_rate": 8.62337959184235e-05, "loss": 0.0447, "step": 27990 }, { "grad_norm": 0.2581998407840729, "learning_rate": 8.622240004480441e-05, "loss": 0.0324, "step": 28000 }, { "grad_norm": 0.235935777425766, "learning_rate": 8.621100020995814e-05, "loss": 0.0262, "step": 28010 }, { "grad_norm": 0.3322802186012268, "learning_rate": 8.619959641513132e-05, "loss": 0.0487, "step": 28020 }, { "grad_norm": 0.3420617878437042, "learning_rate": 8.618818866157105e-05, "loss": 0.0473, "step": 28030 }, { "grad_norm": 0.3723759949207306, "learning_rate": 8.617677695052487e-05, "loss": 0.0332, "step": 28040 }, { "grad_norm": 0.3149554133415222, "learning_rate": 8.616536128324078e-05, "loss": 0.0347, "step": 28050 }, { "grad_norm": 0.38664957880973816, "learning_rate": 8.615394166096712e-05, "loss": 0.0342, "step": 28060 }, { "grad_norm": 0.23660999536514282, "learning_rate": 8.614251808495279e-05, "loss": 0.0342, "step": 28070 }, { "grad_norm": 0.405758798122406, "learning_rate": 8.6131090556447e-05, "loss": 0.0496, "step": 28080 }, { "grad_norm": 0.30757784843444824, "learning_rate": 8.611965907669947e-05, "loss": 0.0429, "step": 28090 }, { "grad_norm": 0.53425133228302, "learning_rate": 8.610822364696034e-05, "loss": 0.0382, "step": 28100 }, { "grad_norm": 0.23924300074577332, "learning_rate": 8.609678426848015e-05, "loss": 0.0365, "step": 28110 }, { "grad_norm": 0.41907986998558044, "learning_rate": 8.60853409425099e-05, "loss": 0.0384, "step": 28120 }, { "grad_norm": 0.28823554515838623, "learning_rate": 8.607389367030104e-05, "loss": 0.0385, "step": 28130 }, { "grad_norm": 0.39081695675849915, "learning_rate": 8.606244245310538e-05, "loss": 0.0315, "step": 28140 }, { "grad_norm": 0.3822290301322937, "learning_rate": 8.605098729217525e-05, "loss": 0.0439, "step": 28150 }, { "grad_norm": 0.32391357421875, "learning_rate": 8.603952818876335e-05, "loss": 0.0422, "step": 28160 }, { "grad_norm": 0.29950791597366333, "learning_rate": 8.602806514412281e-05, "loss": 0.0341, "step": 28170 }, { "grad_norm": 0.24773827195167542, "learning_rate": 8.601659815950726e-05, "loss": 0.0448, "step": 28180 }, { "grad_norm": 0.418301522731781, "learning_rate": 8.600512723617067e-05, "loss": 0.0722, "step": 28190 }, { "grad_norm": 0.31599828600883484, "learning_rate": 8.59936523753675e-05, "loss": 0.0405, "step": 28200 }, { "grad_norm": 0.24977631866931915, "learning_rate": 8.598217357835264e-05, "loss": 0.0455, "step": 28210 }, { "grad_norm": 0.28189754486083984, "learning_rate": 8.597069084638135e-05, "loss": 0.0442, "step": 28220 }, { "grad_norm": 0.383186012506485, "learning_rate": 8.595920418070939e-05, "loss": 0.0436, "step": 28230 }, { "grad_norm": 0.24869033694267273, "learning_rate": 8.594771358259295e-05, "loss": 0.032, "step": 28240 }, { "grad_norm": 0.45868170261383057, "learning_rate": 8.593621905328858e-05, "loss": 0.0468, "step": 28250 }, { "grad_norm": 0.3282680809497833, "learning_rate": 8.592472059405333e-05, "loss": 0.0431, "step": 28260 }, { "grad_norm": 0.30914562940597534, "learning_rate": 8.591321820614464e-05, "loss": 0.0399, "step": 28270 }, { "grad_norm": 0.3691658675670624, "learning_rate": 8.590171189082041e-05, "loss": 0.0343, "step": 28280 }, { "grad_norm": 0.3018491864204407, "learning_rate": 8.589020164933894e-05, "loss": 0.0267, "step": 28290 }, { "grad_norm": 0.6217901110649109, "learning_rate": 8.587868748295898e-05, "loss": 0.0462, "step": 28300 }, { "grad_norm": 0.24063174426555634, "learning_rate": 8.586716939293971e-05, "loss": 0.0367, "step": 28310 }, { "grad_norm": 0.3236628472805023, "learning_rate": 8.58556473805407e-05, "loss": 0.0363, "step": 28320 }, { "grad_norm": 0.44079431891441345, "learning_rate": 8.584412144702202e-05, "loss": 0.0403, "step": 28330 }, { "grad_norm": 1.5147485733032227, "learning_rate": 8.58325915936441e-05, "loss": 0.0355, "step": 28340 }, { "grad_norm": 0.35303202271461487, "learning_rate": 8.582105782166783e-05, "loss": 0.0313, "step": 28350 }, { "grad_norm": 0.4196009039878845, "learning_rate": 8.580952013235455e-05, "loss": 0.0387, "step": 28360 }, { "grad_norm": 0.41005292534828186, "learning_rate": 8.579797852696596e-05, "loss": 0.0373, "step": 28370 }, { "grad_norm": 0.43140313029289246, "learning_rate": 8.578643300676428e-05, "loss": 0.0539, "step": 28380 }, { "grad_norm": 0.9366371035575867, "learning_rate": 8.577488357301209e-05, "loss": 0.0395, "step": 28390 }, { "grad_norm": 0.3594188392162323, "learning_rate": 8.576333022697242e-05, "loss": 0.0386, "step": 28400 }, { "grad_norm": 0.5028604865074158, "learning_rate": 8.575177296990873e-05, "loss": 0.0522, "step": 28410 }, { "grad_norm": 0.24622824788093567, "learning_rate": 8.574021180308489e-05, "loss": 0.0484, "step": 28420 }, { "grad_norm": 0.36355170607566833, "learning_rate": 8.572864672776523e-05, "loss": 0.0433, "step": 28430 }, { "grad_norm": 0.41445887088775635, "learning_rate": 8.571707774521447e-05, "loss": 0.0407, "step": 28440 }, { "grad_norm": 0.3635046184062958, "learning_rate": 8.57055048566978e-05, "loss": 0.035, "step": 28450 }, { "grad_norm": 0.49786481261253357, "learning_rate": 8.569392806348078e-05, "loss": 0.0539, "step": 28460 }, { "grad_norm": 0.4331667125225067, "learning_rate": 8.568234736682947e-05, "loss": 0.0435, "step": 28470 }, { "grad_norm": 0.3379708230495453, "learning_rate": 8.567076276801029e-05, "loss": 0.043, "step": 28480 }, { "grad_norm": 0.35119861364364624, "learning_rate": 8.565917426829013e-05, "loss": 0.0365, "step": 28490 }, { "grad_norm": 1.6814374923706055, "learning_rate": 8.564758186893628e-05, "loss": 0.0431, "step": 28500 }, { "grad_norm": 0.43483301997184753, "learning_rate": 8.563598557121649e-05, "loss": 0.0331, "step": 28510 }, { "grad_norm": 0.3582538068294525, "learning_rate": 8.562438537639888e-05, "loss": 0.0411, "step": 28520 }, { "grad_norm": 0.488988995552063, "learning_rate": 8.561278128575206e-05, "loss": 0.0396, "step": 28530 }, { "grad_norm": 0.2875147759914398, "learning_rate": 8.5601173300545e-05, "loss": 0.0403, "step": 28540 }, { "grad_norm": 0.4771023094654083, "learning_rate": 8.558956142204717e-05, "loss": 0.0394, "step": 28550 }, { "grad_norm": 0.35268962383270264, "learning_rate": 8.55779456515284e-05, "loss": 0.0446, "step": 28560 }, { "grad_norm": 0.3939008116722107, "learning_rate": 8.556632599025898e-05, "loss": 0.0488, "step": 28570 }, { "grad_norm": 0.5516713857650757, "learning_rate": 8.555470243950964e-05, "loss": 0.0435, "step": 28580 }, { "grad_norm": 1.4690625667572021, "learning_rate": 8.554307500055148e-05, "loss": 0.0474, "step": 28590 }, { "grad_norm": 0.5896278023719788, "learning_rate": 8.553144367465609e-05, "loss": 0.0417, "step": 28600 }, { "grad_norm": 0.4416000545024872, "learning_rate": 8.551980846309544e-05, "loss": 0.0535, "step": 28610 }, { "grad_norm": 0.400572270154953, "learning_rate": 8.550816936714193e-05, "loss": 0.0397, "step": 28620 }, { "grad_norm": 0.37809473276138306, "learning_rate": 8.549652638806841e-05, "loss": 0.0399, "step": 28630 }, { "grad_norm": 0.5859066247940063, "learning_rate": 8.548487952714812e-05, "loss": 0.0449, "step": 28640 }, { "grad_norm": 0.3938581645488739, "learning_rate": 8.547322878565478e-05, "loss": 0.0381, "step": 28650 }, { "grad_norm": 0.381316214799881, "learning_rate": 8.546157416486245e-05, "loss": 0.0421, "step": 28660 }, { "grad_norm": 0.32036274671554565, "learning_rate": 8.54499156660457e-05, "loss": 0.0353, "step": 28670 }, { "grad_norm": 0.48846808075904846, "learning_rate": 8.543825329047947e-05, "loss": 0.0439, "step": 28680 }, { "grad_norm": 0.3264347314834595, "learning_rate": 8.542658703943913e-05, "loss": 0.0413, "step": 28690 }, { "grad_norm": 0.5467832684516907, "learning_rate": 8.541491691420051e-05, "loss": 0.0471, "step": 28700 }, { "grad_norm": 0.2753008306026459, "learning_rate": 8.54032429160398e-05, "loss": 0.0384, "step": 28710 }, { "grad_norm": 0.44494420289993286, "learning_rate": 8.539156504623369e-05, "loss": 0.0418, "step": 28720 }, { "grad_norm": 0.3459649980068207, "learning_rate": 8.537988330605923e-05, "loss": 0.0525, "step": 28730 }, { "grad_norm": 0.36191505193710327, "learning_rate": 8.536819769679393e-05, "loss": 0.0444, "step": 28740 }, { "grad_norm": 0.37000179290771484, "learning_rate": 8.53565082197157e-05, "loss": 0.0356, "step": 28750 }, { "grad_norm": 0.32565224170684814, "learning_rate": 8.534481487610289e-05, "loss": 0.0532, "step": 28760 }, { "grad_norm": 0.5846612453460693, "learning_rate": 8.533311766723428e-05, "loss": 0.0367, "step": 28770 }, { "grad_norm": 0.3006758689880371, "learning_rate": 8.532141659438901e-05, "loss": 0.0355, "step": 28780 }, { "grad_norm": 0.4487796723842621, "learning_rate": 8.530971165884675e-05, "loss": 0.0477, "step": 28790 }, { "grad_norm": 0.40857014060020447, "learning_rate": 8.529800286188752e-05, "loss": 0.04, "step": 28800 }, { "grad_norm": 0.5108959078788757, "learning_rate": 8.528629020479175e-05, "loss": 0.0449, "step": 28810 }, { "grad_norm": 0.3685985505580902, "learning_rate": 8.527457368884033e-05, "loss": 0.0399, "step": 28820 }, { "grad_norm": 0.382183700799942, "learning_rate": 8.526285331531458e-05, "loss": 0.0299, "step": 28830 }, { "grad_norm": 0.5279740691184998, "learning_rate": 8.525112908549621e-05, "loss": 0.0484, "step": 28840 }, { "grad_norm": 0.4141520857810974, "learning_rate": 8.523940100066735e-05, "loss": 0.0428, "step": 28850 }, { "grad_norm": 0.3513302803039551, "learning_rate": 8.52276690621106e-05, "loss": 0.0422, "step": 28860 }, { "grad_norm": 0.3290916383266449, "learning_rate": 8.521593327110889e-05, "loss": 0.0368, "step": 28870 }, { "grad_norm": 0.2158178836107254, "learning_rate": 8.520419362894569e-05, "loss": 0.0391, "step": 28880 }, { "grad_norm": 0.3072427213191986, "learning_rate": 8.51924501369048e-05, "loss": 0.0328, "step": 28890 }, { "grad_norm": 0.31005898118019104, "learning_rate": 8.518070279627047e-05, "loss": 0.0348, "step": 28900 }, { "grad_norm": 0.33588936924934387, "learning_rate": 8.516895160832737e-05, "loss": 0.0357, "step": 28910 }, { "grad_norm": 0.44085925817489624, "learning_rate": 8.515719657436061e-05, "loss": 0.0371, "step": 28920 }, { "grad_norm": 0.22538742423057556, "learning_rate": 8.514543769565568e-05, "loss": 0.0322, "step": 28930 }, { "grad_norm": 0.34005025029182434, "learning_rate": 8.513367497349853e-05, "loss": 0.0393, "step": 28940 }, { "grad_norm": 0.34142330288887024, "learning_rate": 8.51219084091755e-05, "loss": 0.0428, "step": 28950 }, { "grad_norm": 0.5151235461235046, "learning_rate": 8.511013800397338e-05, "loss": 0.0431, "step": 28960 }, { "grad_norm": 0.35084307193756104, "learning_rate": 8.509836375917937e-05, "loss": 0.0462, "step": 28970 }, { "grad_norm": 0.5302925705909729, "learning_rate": 8.508658567608104e-05, "loss": 0.0406, "step": 28980 }, { "grad_norm": 0.3389400243759155, "learning_rate": 8.507480375596647e-05, "loss": 0.0338, "step": 28990 }, { "grad_norm": 0.4558046758174896, "learning_rate": 8.506301800012408e-05, "loss": 0.0417, "step": 29000 }, { "grad_norm": 0.5431724786758423, "learning_rate": 8.505122840984278e-05, "loss": 0.0374, "step": 29010 }, { "grad_norm": 0.31433722376823425, "learning_rate": 8.503943498641182e-05, "loss": 0.0376, "step": 29020 }, { "grad_norm": 0.34702059626579285, "learning_rate": 8.502763773112095e-05, "loss": 0.0291, "step": 29030 }, { "grad_norm": 0.3790374994277954, "learning_rate": 8.501583664526026e-05, "loss": 0.0378, "step": 29040 }, { "grad_norm": 0.278913289308548, "learning_rate": 8.500403173012032e-05, "loss": 0.0366, "step": 29050 }, { "grad_norm": 0.3223758041858673, "learning_rate": 8.499222298699211e-05, "loss": 0.0319, "step": 29060 }, { "grad_norm": 0.29842743277549744, "learning_rate": 8.498041041716701e-05, "loss": 0.0367, "step": 29070 }, { "grad_norm": 0.2690417766571045, "learning_rate": 8.496859402193681e-05, "loss": 0.0356, "step": 29080 }, { "grad_norm": 0.4029424786567688, "learning_rate": 8.495677380259374e-05, "loss": 0.0323, "step": 29090 }, { "grad_norm": 0.2825126647949219, "learning_rate": 8.494494976043045e-05, "loss": 0.0377, "step": 29100 }, { "grad_norm": 0.19026421010494232, "learning_rate": 8.493312189673998e-05, "loss": 0.0318, "step": 29110 }, { "grad_norm": 0.48030176758766174, "learning_rate": 8.492129021281584e-05, "loss": 0.0401, "step": 29120 }, { "grad_norm": 0.30337679386138916, "learning_rate": 8.490945470995188e-05, "loss": 0.0313, "step": 29130 }, { "grad_norm": 0.6045801639556885, "learning_rate": 8.489761538944247e-05, "loss": 0.0358, "step": 29140 }, { "grad_norm": 0.2513841986656189, "learning_rate": 8.48857722525823e-05, "loss": 0.0335, "step": 29150 }, { "grad_norm": 0.31824344396591187, "learning_rate": 8.487392530066652e-05, "loss": 0.0394, "step": 29160 }, { "grad_norm": 0.34424951672554016, "learning_rate": 8.486207453499069e-05, "loss": 0.0429, "step": 29170 }, { "grad_norm": 0.44412845373153687, "learning_rate": 8.485021995685082e-05, "loss": 0.0492, "step": 29180 }, { "grad_norm": 0.48349735140800476, "learning_rate": 8.483836156754328e-05, "loss": 0.0516, "step": 29190 }, { "grad_norm": 0.26754331588745117, "learning_rate": 8.482649936836491e-05, "loss": 0.0396, "step": 29200 }, { "grad_norm": 0.6397508382797241, "learning_rate": 8.481463336061293e-05, "loss": 0.036, "step": 29210 }, { "grad_norm": 0.7179964780807495, "learning_rate": 8.480276354558496e-05, "loss": 0.0701, "step": 29220 }, { "grad_norm": 0.43704327940940857, "learning_rate": 8.479088992457913e-05, "loss": 0.0416, "step": 29230 }, { "grad_norm": 0.5426940321922302, "learning_rate": 8.477901249889387e-05, "loss": 0.0604, "step": 29240 }, { "grad_norm": 0.31209662556648254, "learning_rate": 8.47671312698281e-05, "loss": 0.0304, "step": 29250 }, { "grad_norm": 0.3452214002609253, "learning_rate": 8.475524623868112e-05, "loss": 0.0443, "step": 29260 }, { "grad_norm": 0.5426904559135437, "learning_rate": 8.474335740675266e-05, "loss": 0.0347, "step": 29270 }, { "grad_norm": 0.5965123772621155, "learning_rate": 8.473146477534289e-05, "loss": 0.0406, "step": 29280 }, { "grad_norm": 0.34061598777770996, "learning_rate": 8.471956834575232e-05, "loss": 0.041, "step": 29290 }, { "grad_norm": 0.4836970567703247, "learning_rate": 8.470766811928197e-05, "loss": 0.0369, "step": 29300 }, { "grad_norm": 0.34268608689308167, "learning_rate": 8.469576409723323e-05, "loss": 0.0468, "step": 29310 }, { "grad_norm": 0.36537501215934753, "learning_rate": 8.468385628090788e-05, "loss": 0.0385, "step": 29320 }, { "grad_norm": 0.37774574756622314, "learning_rate": 8.467194467160815e-05, "loss": 0.0413, "step": 29330 }, { "grad_norm": 0.37029772996902466, "learning_rate": 8.466002927063667e-05, "loss": 0.035, "step": 29340 }, { "grad_norm": 0.41992300748825073, "learning_rate": 8.464811007929651e-05, "loss": 0.0478, "step": 29350 }, { "grad_norm": 0.40433213114738464, "learning_rate": 8.463618709889114e-05, "loss": 0.0375, "step": 29360 }, { "grad_norm": 0.40438809990882874, "learning_rate": 8.462426033072442e-05, "loss": 0.0392, "step": 29370 }, { "grad_norm": 0.2827891409397125, "learning_rate": 8.461232977610061e-05, "loss": 0.0319, "step": 29380 }, { "grad_norm": 0.6789244413375854, "learning_rate": 8.46003954363245e-05, "loss": 0.0358, "step": 29390 }, { "grad_norm": 0.5329113602638245, "learning_rate": 8.458845731270115e-05, "loss": 0.0423, "step": 29400 }, { "grad_norm": 0.5180900692939758, "learning_rate": 8.45765154065361e-05, "loss": 0.0467, "step": 29410 }, { "grad_norm": 0.22045516967773438, "learning_rate": 8.456456971913532e-05, "loss": 0.0282, "step": 29420 }, { "grad_norm": 0.2396380603313446, "learning_rate": 8.455262025180517e-05, "loss": 0.0345, "step": 29430 }, { "grad_norm": 0.4941868484020233, "learning_rate": 8.454066700585242e-05, "loss": 0.0383, "step": 29440 }, { "grad_norm": 0.2692984938621521, "learning_rate": 8.452870998258423e-05, "loss": 0.0357, "step": 29450 }, { "grad_norm": 0.26712071895599365, "learning_rate": 8.451674918330825e-05, "loss": 0.0367, "step": 29460 }, { "grad_norm": 0.5171303153038025, "learning_rate": 8.450478460933246e-05, "loss": 0.0382, "step": 29470 }, { "grad_norm": 0.34462058544158936, "learning_rate": 8.449281626196532e-05, "loss": 0.0469, "step": 29480 }, { "grad_norm": 0.40842264890670776, "learning_rate": 8.448084414251564e-05, "loss": 0.0493, "step": 29490 }, { "grad_norm": 4.754133701324463, "learning_rate": 8.446886825229271e-05, "loss": 0.0523, "step": 29500 }, { "grad_norm": 0.4241630733013153, "learning_rate": 8.445688859260615e-05, "loss": 0.0411, "step": 29510 }, { "grad_norm": 0.4415193498134613, "learning_rate": 8.444490516476606e-05, "loss": 0.0364, "step": 29520 }, { "grad_norm": 0.26845860481262207, "learning_rate": 8.443291797008293e-05, "loss": 0.0371, "step": 29530 }, { "grad_norm": 0.19826163351535797, "learning_rate": 8.442092700986765e-05, "loss": 0.0502, "step": 29540 }, { "grad_norm": 0.2810412049293518, "learning_rate": 8.440893228543156e-05, "loss": 0.0393, "step": 29550 }, { "grad_norm": 0.4552595019340515, "learning_rate": 8.439693379808638e-05, "loss": 0.0371, "step": 29560 }, { "grad_norm": 0.36266419291496277, "learning_rate": 8.43849315491442e-05, "loss": 0.0377, "step": 29570 }, { "grad_norm": 0.3705275058746338, "learning_rate": 8.437292553991763e-05, "loss": 0.0325, "step": 29580 }, { "grad_norm": 0.3820542097091675, "learning_rate": 8.436091577171959e-05, "loss": 0.0402, "step": 29590 }, { "grad_norm": 0.35705214738845825, "learning_rate": 8.434890224586347e-05, "loss": 0.0391, "step": 29600 }, { "grad_norm": 0.5969372391700745, "learning_rate": 8.433688496366303e-05, "loss": 0.0303, "step": 29610 }, { "grad_norm": 0.42417600750923157, "learning_rate": 8.432486392643248e-05, "loss": 0.0387, "step": 29620 }, { "grad_norm": 0.35446834564208984, "learning_rate": 8.431283913548643e-05, "loss": 0.0349, "step": 29630 }, { "grad_norm": 0.2572995126247406, "learning_rate": 8.430081059213985e-05, "loss": 0.0349, "step": 29640 }, { "grad_norm": 0.333010733127594, "learning_rate": 8.428877829770823e-05, "loss": 0.0373, "step": 29650 }, { "grad_norm": 0.2597094178199768, "learning_rate": 8.427674225350735e-05, "loss": 0.0456, "step": 29660 }, { "grad_norm": 0.2787008583545685, "learning_rate": 8.426470246085347e-05, "loss": 0.0335, "step": 29670 }, { "grad_norm": 0.38245120644569397, "learning_rate": 8.425265892106324e-05, "loss": 0.0389, "step": 29680 }, { "grad_norm": 0.27612045407295227, "learning_rate": 8.424061163545374e-05, "loss": 0.0273, "step": 29690 }, { "grad_norm": 0.4081169664859772, "learning_rate": 8.422856060534243e-05, "loss": 0.0405, "step": 29700 }, { "grad_norm": 0.4055878818035126, "learning_rate": 8.421650583204718e-05, "loss": 0.034, "step": 29710 }, { "grad_norm": 0.30993396043777466, "learning_rate": 8.420444731688633e-05, "loss": 0.037, "step": 29720 }, { "grad_norm": 0.26753169298171997, "learning_rate": 8.419238506117852e-05, "loss": 0.0488, "step": 29730 }, { "grad_norm": 0.3400731384754181, "learning_rate": 8.418031906624289e-05, "loss": 0.0349, "step": 29740 }, { "grad_norm": 0.24060693383216858, "learning_rate": 8.416824933339898e-05, "loss": 0.0386, "step": 29750 }, { "grad_norm": 0.26689252257347107, "learning_rate": 8.415617586396667e-05, "loss": 0.0349, "step": 29760 }, { "grad_norm": 0.2895442843437195, "learning_rate": 8.414409865926632e-05, "loss": 0.0378, "step": 29770 }, { "grad_norm": 0.290506511926651, "learning_rate": 8.413201772061867e-05, "loss": 0.0402, "step": 29780 }, { "grad_norm": 0.35152673721313477, "learning_rate": 8.411993304934488e-05, "loss": 0.0367, "step": 29790 }, { "grad_norm": 0.2420547753572464, "learning_rate": 8.410784464676654e-05, "loss": 0.0317, "step": 29800 }, { "grad_norm": 0.2900025248527527, "learning_rate": 8.409575251420556e-05, "loss": 0.0268, "step": 29810 }, { "grad_norm": 0.42928892374038696, "learning_rate": 8.408365665298435e-05, "loss": 0.0292, "step": 29820 }, { "grad_norm": 0.45695364475250244, "learning_rate": 8.40715570644257e-05, "loss": 0.0411, "step": 29830 }, { "grad_norm": 0.8482176661491394, "learning_rate": 8.40594537498528e-05, "loss": 0.0473, "step": 29840 }, { "grad_norm": 0.46355950832366943, "learning_rate": 8.404734671058924e-05, "loss": 0.0414, "step": 29850 }, { "grad_norm": 0.3605497479438782, "learning_rate": 8.403523594795902e-05, "loss": 0.0499, "step": 29860 }, { "grad_norm": 0.42415258288383484, "learning_rate": 8.402312146328659e-05, "loss": 0.0324, "step": 29870 }, { "grad_norm": 0.4984363615512848, "learning_rate": 8.401100325789675e-05, "loss": 0.0467, "step": 29880 }, { "grad_norm": 0.2874836027622223, "learning_rate": 8.399888133311472e-05, "loss": 0.0388, "step": 29890 }, { "grad_norm": 0.26505246758461, "learning_rate": 8.398675569026613e-05, "loss": 0.0385, "step": 29900 }, { "grad_norm": 0.3883230686187744, "learning_rate": 8.397462633067705e-05, "loss": 0.031, "step": 29910 }, { "grad_norm": 0.35569173097610474, "learning_rate": 8.396249325567392e-05, "loss": 0.0405, "step": 29920 }, { "grad_norm": 0.350348562002182, "learning_rate": 8.395035646658357e-05, "loss": 0.0416, "step": 29930 }, { "grad_norm": 0.27044105529785156, "learning_rate": 8.39382159647333e-05, "loss": 0.0389, "step": 29940 }, { "grad_norm": 0.2804214656352997, "learning_rate": 8.392607175145075e-05, "loss": 0.0377, "step": 29950 }, { "grad_norm": 0.24791425466537476, "learning_rate": 8.3913923828064e-05, "loss": 0.0367, "step": 29960 }, { "grad_norm": 0.27886486053466797, "learning_rate": 8.390177219590152e-05, "loss": 0.0356, "step": 29970 }, { "grad_norm": 0.2542877495288849, "learning_rate": 8.388961685629222e-05, "loss": 0.0348, "step": 29980 }, { "grad_norm": 0.4578014314174652, "learning_rate": 8.387745781056536e-05, "loss": 0.0378, "step": 29990 }, { "grad_norm": 0.4479272961616516, "learning_rate": 8.386529506005065e-05, "loss": 0.0348, "step": 30000 }, { "grad_norm": 0.3420987129211426, "learning_rate": 8.38531286060782e-05, "loss": 0.037, "step": 30010 }, { "grad_norm": 0.35531091690063477, "learning_rate": 8.384095844997849e-05, "loss": 0.0341, "step": 30020 }, { "grad_norm": 0.30186963081359863, "learning_rate": 8.382878459308245e-05, "loss": 0.0399, "step": 30030 }, { "grad_norm": 0.25760215520858765, "learning_rate": 8.381660703672138e-05, "loss": 0.0377, "step": 30040 }, { "grad_norm": 0.3126625716686249, "learning_rate": 8.380442578222702e-05, "loss": 0.0504, "step": 30050 }, { "grad_norm": 0.712372362613678, "learning_rate": 8.379224083093146e-05, "loss": 0.0358, "step": 30060 }, { "grad_norm": 0.8077777028083801, "learning_rate": 8.378005218416727e-05, "loss": 0.048, "step": 30070 }, { "grad_norm": 0.6892715692520142, "learning_rate": 8.376785984326735e-05, "loss": 0.0472, "step": 30080 }, { "grad_norm": 0.36294904351234436, "learning_rate": 8.375566380956506e-05, "loss": 0.0351, "step": 30090 }, { "grad_norm": 0.38872218132019043, "learning_rate": 8.374346408439411e-05, "loss": 0.0327, "step": 30100 }, { "grad_norm": 0.6870675683021545, "learning_rate": 8.373126066908868e-05, "loss": 0.0433, "step": 30110 }, { "grad_norm": 0.2930174469947815, "learning_rate": 8.371905356498326e-05, "loss": 0.0393, "step": 30120 }, { "grad_norm": 0.3609110414981842, "learning_rate": 8.370684277341288e-05, "loss": 0.0383, "step": 30130 }, { "grad_norm": 0.24745605885982513, "learning_rate": 8.369462829571282e-05, "loss": 0.0356, "step": 30140 }, { "grad_norm": 0.4871610701084137, "learning_rate": 8.36824101332189e-05, "loss": 0.0375, "step": 30150 }, { "grad_norm": 0.2120341807603836, "learning_rate": 8.367018828726721e-05, "loss": 0.038, "step": 30160 }, { "grad_norm": 0.6375918388366699, "learning_rate": 8.365796275919438e-05, "loss": 0.0388, "step": 30170 }, { "grad_norm": 0.3763818144798279, "learning_rate": 8.364573355033734e-05, "loss": 0.0329, "step": 30180 }, { "grad_norm": 0.2577761113643646, "learning_rate": 8.363350066203346e-05, "loss": 0.0318, "step": 30190 }, { "grad_norm": 0.27061545848846436, "learning_rate": 8.362126409562053e-05, "loss": 0.0412, "step": 30200 }, { "grad_norm": 0.41368651390075684, "learning_rate": 8.360902385243667e-05, "loss": 0.0489, "step": 30210 }, { "grad_norm": 0.31074783205986023, "learning_rate": 8.359677993382052e-05, "loss": 0.0376, "step": 30220 }, { "grad_norm": 0.7394621968269348, "learning_rate": 8.358453234111103e-05, "loss": 0.0381, "step": 30230 }, { "grad_norm": 0.3115876615047455, "learning_rate": 8.357228107564756e-05, "loss": 0.0479, "step": 30240 }, { "grad_norm": 0.5252290964126587, "learning_rate": 8.356002613876993e-05, "loss": 0.0386, "step": 30250 }, { "grad_norm": 0.33084359765052795, "learning_rate": 8.35477675318183e-05, "loss": 0.0364, "step": 30260 }, { "grad_norm": 0.3783540427684784, "learning_rate": 8.353550525613323e-05, "loss": 0.0425, "step": 30270 }, { "grad_norm": 0.3047431707382202, "learning_rate": 8.352323931305572e-05, "loss": 0.0557, "step": 30280 }, { "grad_norm": 0.270060658454895, "learning_rate": 8.351096970392717e-05, "loss": 0.041, "step": 30290 }, { "grad_norm": 0.2720925509929657, "learning_rate": 8.349869643008937e-05, "loss": 0.0326, "step": 30300 }, { "grad_norm": 0.33041346073150635, "learning_rate": 8.348641949288449e-05, "loss": 0.0468, "step": 30310 }, { "grad_norm": 0.3762701153755188, "learning_rate": 8.34741388936551e-05, "loss": 0.0435, "step": 30320 }, { "grad_norm": 0.36318284273147583, "learning_rate": 8.346185463374423e-05, "loss": 0.047, "step": 30330 }, { "grad_norm": 0.4295693039894104, "learning_rate": 8.344956671449524e-05, "loss": 0.0437, "step": 30340 }, { "grad_norm": 0.3844248354434967, "learning_rate": 8.343727513725192e-05, "loss": 0.0398, "step": 30350 }, { "grad_norm": 0.3554269075393677, "learning_rate": 8.342497990335847e-05, "loss": 0.0461, "step": 30360 }, { "grad_norm": 0.38089606165885925, "learning_rate": 8.341268101415946e-05, "loss": 0.0468, "step": 30370 }, { "grad_norm": 0.28243932127952576, "learning_rate": 8.34003784709999e-05, "loss": 0.0405, "step": 30380 }, { "grad_norm": 0.7278969287872314, "learning_rate": 8.338807227522516e-05, "loss": 0.0376, "step": 30390 }, { "grad_norm": 0.25419482588768005, "learning_rate": 8.337576242818103e-05, "loss": 0.0387, "step": 30400 }, { "grad_norm": 0.3935230076313019, "learning_rate": 8.336344893121372e-05, "loss": 0.0333, "step": 30410 }, { "grad_norm": 0.35841497778892517, "learning_rate": 8.335113178566977e-05, "loss": 0.0446, "step": 30420 }, { "grad_norm": 0.5848546028137207, "learning_rate": 8.33388109928962e-05, "loss": 0.0388, "step": 30430 }, { "grad_norm": 0.3379873037338257, "learning_rate": 8.33264865542404e-05, "loss": 0.0368, "step": 30440 }, { "grad_norm": 0.25081977248191833, "learning_rate": 8.331415847105013e-05, "loss": 0.043, "step": 30450 }, { "grad_norm": 0.3069287836551666, "learning_rate": 8.330182674467357e-05, "loss": 0.034, "step": 30460 }, { "grad_norm": 0.3472833037376404, "learning_rate": 8.32894913764593e-05, "loss": 0.0326, "step": 30470 }, { "grad_norm": 0.30876249074935913, "learning_rate": 8.327715236775633e-05, "loss": 0.0349, "step": 30480 }, { "grad_norm": 0.2947734296321869, "learning_rate": 8.326480971991398e-05, "loss": 0.0357, "step": 30490 }, { "grad_norm": 0.6008239984512329, "learning_rate": 8.325246343428206e-05, "loss": 0.0472, "step": 30500 }, { "grad_norm": 0.24488648772239685, "learning_rate": 8.324011351221072e-05, "loss": 0.0337, "step": 30510 }, { "grad_norm": 0.34229427576065063, "learning_rate": 8.322775995505057e-05, "loss": 0.0311, "step": 30520 }, { "grad_norm": 0.33231285214424133, "learning_rate": 8.321540276415254e-05, "loss": 0.0399, "step": 30530 }, { "grad_norm": 0.4345304071903229, "learning_rate": 8.320304194086798e-05, "loss": 0.0387, "step": 30540 }, { "grad_norm": 0.3918163776397705, "learning_rate": 8.31906774865487e-05, "loss": 0.0394, "step": 30550 }, { "grad_norm": 0.4756021499633789, "learning_rate": 8.317830940254682e-05, "loss": 0.0325, "step": 30560 }, { "grad_norm": 0.2892128527164459, "learning_rate": 8.316593769021491e-05, "loss": 0.0385, "step": 30570 }, { "grad_norm": 0.737396240234375, "learning_rate": 8.315356235090592e-05, "loss": 0.0486, "step": 30580 }, { "grad_norm": 0.3069036602973938, "learning_rate": 8.314118338597319e-05, "loss": 0.0439, "step": 30590 }, { "grad_norm": 0.5599225163459778, "learning_rate": 8.312880079677048e-05, "loss": 0.0394, "step": 30600 }, { "grad_norm": 0.32681044936180115, "learning_rate": 8.311641458465191e-05, "loss": 0.0371, "step": 30610 }, { "grad_norm": 0.4474082589149475, "learning_rate": 8.310402475097205e-05, "loss": 0.0322, "step": 30620 }, { "grad_norm": 0.36262547969818115, "learning_rate": 8.309163129708581e-05, "loss": 0.0342, "step": 30630 }, { "grad_norm": 0.6887705326080322, "learning_rate": 8.307923422434852e-05, "loss": 0.0411, "step": 30640 }, { "grad_norm": 0.49646636843681335, "learning_rate": 8.30668335341159e-05, "loss": 0.0471, "step": 30650 }, { "grad_norm": 0.3375691771507263, "learning_rate": 8.305442922774408e-05, "loss": 0.0353, "step": 30660 }, { "grad_norm": 0.40806978940963745, "learning_rate": 8.304202130658959e-05, "loss": 0.0326, "step": 30670 }, { "grad_norm": 0.32898205518722534, "learning_rate": 8.302960977200931e-05, "loss": 0.0333, "step": 30680 }, { "grad_norm": 0.5665314793586731, "learning_rate": 8.301719462536058e-05, "loss": 0.0326, "step": 30690 }, { "grad_norm": 0.38572561740875244, "learning_rate": 8.300477586800108e-05, "loss": 0.0463, "step": 30700 }, { "grad_norm": 0.963496208190918, "learning_rate": 8.299235350128892e-05, "loss": 0.0501, "step": 30710 }, { "grad_norm": 0.33761486411094666, "learning_rate": 8.297992752658258e-05, "loss": 0.0388, "step": 30720 }, { "grad_norm": 0.25707337260246277, "learning_rate": 8.296749794524094e-05, "loss": 0.0517, "step": 30730 }, { "grad_norm": 0.30654340982437134, "learning_rate": 8.295506475862332e-05, "loss": 0.0489, "step": 30740 }, { "grad_norm": 0.2983115315437317, "learning_rate": 8.294262796808933e-05, "loss": 0.0344, "step": 30750 }, { "grad_norm": 0.48841366171836853, "learning_rate": 8.293018757499909e-05, "loss": 0.0431, "step": 30760 }, { "grad_norm": 0.5026473999023438, "learning_rate": 8.291774358071305e-05, "loss": 0.0445, "step": 30770 }, { "grad_norm": 0.3654652535915375, "learning_rate": 8.290529598659205e-05, "loss": 0.0389, "step": 30780 }, { "grad_norm": 2.3937408924102783, "learning_rate": 8.289284479399738e-05, "loss": 0.0351, "step": 30790 }, { "grad_norm": 0.4679868817329407, "learning_rate": 8.288039000429064e-05, "loss": 0.0465, "step": 30800 }, { "grad_norm": 0.4117625653743744, "learning_rate": 8.286793161883388e-05, "loss": 0.0407, "step": 30810 }, { "grad_norm": 1.1519848108291626, "learning_rate": 8.285546963898954e-05, "loss": 0.0453, "step": 30820 }, { "grad_norm": 0.3092482388019562, "learning_rate": 8.284300406612044e-05, "loss": 0.0398, "step": 30830 }, { "grad_norm": 0.3136131167411804, "learning_rate": 8.283053490158978e-05, "loss": 0.0496, "step": 30840 }, { "grad_norm": 0.534511148929596, "learning_rate": 8.28180621467612e-05, "loss": 0.0378, "step": 30850 }, { "grad_norm": 0.31133580207824707, "learning_rate": 8.280558580299867e-05, "loss": 0.0417, "step": 30860 }, { "grad_norm": 0.4878869652748108, "learning_rate": 8.279310587166661e-05, "loss": 0.0366, "step": 30870 }, { "grad_norm": 0.4031529724597931, "learning_rate": 8.278062235412978e-05, "loss": 0.0401, "step": 30880 }, { "grad_norm": 0.23657900094985962, "learning_rate": 8.276813525175339e-05, "loss": 0.038, "step": 30890 }, { "grad_norm": 0.22000667452812195, "learning_rate": 8.2755644565903e-05, "loss": 0.0455, "step": 30900 }, { "grad_norm": 0.47588050365448, "learning_rate": 8.274315029794454e-05, "loss": 0.0435, "step": 30910 }, { "grad_norm": 0.2930246889591217, "learning_rate": 8.273065244924443e-05, "loss": 0.0341, "step": 30920 }, { "grad_norm": 0.3045109212398529, "learning_rate": 8.271815102116936e-05, "loss": 0.0454, "step": 30930 }, { "grad_norm": 0.3986615836620331, "learning_rate": 8.270564601508648e-05, "loss": 0.0423, "step": 30940 }, { "grad_norm": 0.37651848793029785, "learning_rate": 8.269313743236333e-05, "loss": 0.0425, "step": 30950 }, { "grad_norm": 0.3639039397239685, "learning_rate": 8.268062527436783e-05, "loss": 0.0445, "step": 30960 }, { "grad_norm": 0.3596018850803375, "learning_rate": 8.266810954246828e-05, "loss": 0.0395, "step": 30970 }, { "grad_norm": 0.3167949318885803, "learning_rate": 8.265559023803338e-05, "loss": 0.0405, "step": 30980 }, { "grad_norm": 0.24507442116737366, "learning_rate": 8.264306736243223e-05, "loss": 0.0411, "step": 30990 }, { "grad_norm": 0.527935266494751, "learning_rate": 8.263054091703432e-05, "loss": 0.0479, "step": 31000 }, { "grad_norm": 0.32273489236831665, "learning_rate": 8.26180109032095e-05, "loss": 0.0352, "step": 31010 }, { "grad_norm": 0.3102593421936035, "learning_rate": 8.260547732232807e-05, "loss": 0.0371, "step": 31020 }, { "grad_norm": 0.4521332383155823, "learning_rate": 8.259294017576063e-05, "loss": 0.0445, "step": 31030 }, { "grad_norm": 0.5572641491889954, "learning_rate": 8.258039946487828e-05, "loss": 0.037, "step": 31040 }, { "grad_norm": 0.3119651973247528, "learning_rate": 8.256785519105241e-05, "loss": 0.0378, "step": 31050 }, { "grad_norm": 0.34645143151283264, "learning_rate": 8.255530735565488e-05, "loss": 0.0301, "step": 31060 }, { "grad_norm": 0.42501580715179443, "learning_rate": 8.254275596005787e-05, "loss": 0.0461, "step": 31070 }, { "grad_norm": 0.3316183090209961, "learning_rate": 8.2530201005634e-05, "loss": 0.0328, "step": 31080 }, { "grad_norm": 0.6705784201622009, "learning_rate": 8.251764249375626e-05, "loss": 0.0418, "step": 31090 }, { "grad_norm": 0.3305554986000061, "learning_rate": 8.250508042579803e-05, "loss": 0.0409, "step": 31100 }, { "grad_norm": 0.3318779170513153, "learning_rate": 8.249251480313307e-05, "loss": 0.0443, "step": 31110 }, { "grad_norm": 0.32211193442344666, "learning_rate": 8.247994562713555e-05, "loss": 0.0351, "step": 31120 }, { "grad_norm": 0.2773047089576721, "learning_rate": 8.246737289917998e-05, "loss": 0.0377, "step": 31130 }, { "grad_norm": 0.22993303835391998, "learning_rate": 8.245479662064135e-05, "loss": 0.0357, "step": 31140 }, { "grad_norm": 0.36457353830337524, "learning_rate": 8.244221679289496e-05, "loss": 0.0316, "step": 31150 }, { "grad_norm": 0.4686535596847534, "learning_rate": 8.242963341731652e-05, "loss": 0.0415, "step": 31160 }, { "grad_norm": 0.40393489599227905, "learning_rate": 8.24170464952821e-05, "loss": 0.0395, "step": 31170 }, { "grad_norm": 0.32617369294166565, "learning_rate": 8.240445602816824e-05, "loss": 0.0411, "step": 31180 }, { "grad_norm": 0.29764673113822937, "learning_rate": 8.239186201735179e-05, "loss": 0.0339, "step": 31190 }, { "grad_norm": 0.5868545770645142, "learning_rate": 8.237926446420998e-05, "loss": 0.0362, "step": 31200 }, { "grad_norm": 0.6279767751693726, "learning_rate": 8.236666337012052e-05, "loss": 0.0443, "step": 31210 }, { "grad_norm": 0.3342876136302948, "learning_rate": 8.23540587364614e-05, "loss": 0.0388, "step": 31220 }, { "grad_norm": 0.25293490290641785, "learning_rate": 8.234145056461107e-05, "loss": 0.0462, "step": 31230 }, { "grad_norm": 0.3959311246871948, "learning_rate": 8.232883885594832e-05, "loss": 0.0431, "step": 31240 }, { "grad_norm": 0.39235901832580566, "learning_rate": 8.231622361185236e-05, "loss": 0.0408, "step": 31250 }, { "grad_norm": 0.4485073983669281, "learning_rate": 8.230360483370278e-05, "loss": 0.0566, "step": 31260 }, { "grad_norm": 0.3131929934024811, "learning_rate": 8.229098252287953e-05, "loss": 0.0369, "step": 31270 }, { "grad_norm": 0.2796340584754944, "learning_rate": 8.2278356680763e-05, "loss": 0.0487, "step": 31280 }, { "grad_norm": 0.2868461608886719, "learning_rate": 8.22657273087339e-05, "loss": 0.0402, "step": 31290 }, { "grad_norm": 0.3882789611816406, "learning_rate": 8.225309440817336e-05, "loss": 0.0452, "step": 31300 }, { "grad_norm": 0.29773399233818054, "learning_rate": 8.224045798046293e-05, "loss": 0.0502, "step": 31310 }, { "grad_norm": 0.2736762464046478, "learning_rate": 8.22278180269845e-05, "loss": 0.036, "step": 31320 }, { "grad_norm": 0.37682080268859863, "learning_rate": 8.221517454912031e-05, "loss": 0.0399, "step": 31330 }, { "grad_norm": 0.3400808870792389, "learning_rate": 8.220252754825308e-05, "loss": 0.03, "step": 31340 }, { "grad_norm": 0.3657175898551941, "learning_rate": 8.218987702576586e-05, "loss": 0.0379, "step": 31350 }, { "grad_norm": 0.30296072363853455, "learning_rate": 8.217722298304207e-05, "loss": 0.0368, "step": 31360 }, { "grad_norm": 0.40287208557128906, "learning_rate": 8.216456542146557e-05, "loss": 0.0449, "step": 31370 }, { "grad_norm": 0.2914368808269501, "learning_rate": 8.215190434242055e-05, "loss": 0.0344, "step": 31380 }, { "grad_norm": 0.3533870279788971, "learning_rate": 8.213923974729161e-05, "loss": 0.0478, "step": 31390 }, { "grad_norm": 0.22309459745883942, "learning_rate": 8.212657163746373e-05, "loss": 0.0351, "step": 31400 }, { "grad_norm": 0.2897528409957886, "learning_rate": 8.211390001432227e-05, "loss": 0.0391, "step": 31410 }, { "grad_norm": 0.456397145986557, "learning_rate": 8.210122487925297e-05, "loss": 0.0467, "step": 31420 }, { "grad_norm": 0.4098323881626129, "learning_rate": 8.208854623364202e-05, "loss": 0.0435, "step": 31430 }, { "grad_norm": 0.8522638082504272, "learning_rate": 8.207586407887589e-05, "loss": 0.0462, "step": 31440 }, { "grad_norm": 0.3460673689842224, "learning_rate": 8.206317841634148e-05, "loss": 0.0404, "step": 31450 }, { "grad_norm": 0.3768135905265808, "learning_rate": 8.205048924742609e-05, "loss": 0.0408, "step": 31460 }, { "grad_norm": 0.33779260516166687, "learning_rate": 8.203779657351738e-05, "loss": 0.0368, "step": 31470 }, { "grad_norm": 0.3062061369419098, "learning_rate": 8.20251003960034e-05, "loss": 0.0487, "step": 31480 }, { "grad_norm": 0.3571013808250427, "learning_rate": 8.201240071627258e-05, "loss": 0.033, "step": 31490 }, { "grad_norm": 0.2894783914089203, "learning_rate": 8.199969753571377e-05, "loss": 0.0404, "step": 31500 }, { "grad_norm": 0.7481649518013, "learning_rate": 8.198699085571615e-05, "loss": 0.0296, "step": 31510 }, { "grad_norm": 0.47279661893844604, "learning_rate": 8.197428067766928e-05, "loss": 0.0427, "step": 31520 }, { "grad_norm": 0.38398823142051697, "learning_rate": 8.196156700296316e-05, "loss": 0.0477, "step": 31530 }, { "grad_norm": 0.29384341835975647, "learning_rate": 8.194884983298814e-05, "loss": 0.0339, "step": 31540 }, { "grad_norm": 0.34587159752845764, "learning_rate": 8.193612916913491e-05, "loss": 0.0345, "step": 31550 }, { "grad_norm": 0.29138118028640747, "learning_rate": 8.192340501279463e-05, "loss": 0.0399, "step": 31560 }, { "grad_norm": 0.3421734869480133, "learning_rate": 8.191067736535876e-05, "loss": 0.0351, "step": 31570 }, { "grad_norm": 0.4385833740234375, "learning_rate": 8.18979462282192e-05, "loss": 0.0307, "step": 31580 }, { "grad_norm": 0.2708422541618347, "learning_rate": 8.188521160276819e-05, "loss": 0.0346, "step": 31590 }, { "grad_norm": 0.3404846489429474, "learning_rate": 8.187247349039837e-05, "loss": 0.0342, "step": 31600 }, { "grad_norm": 0.40431246161460876, "learning_rate": 8.185973189250278e-05, "loss": 0.0361, "step": 31610 }, { "grad_norm": 0.42619800567626953, "learning_rate": 8.184698681047482e-05, "loss": 0.0365, "step": 31620 }, { "grad_norm": 0.31234949827194214, "learning_rate": 8.183423824570827e-05, "loss": 0.0323, "step": 31630 }, { "grad_norm": 0.4437505900859833, "learning_rate": 8.182148619959725e-05, "loss": 0.034, "step": 31640 }, { "grad_norm": 0.20046648383140564, "learning_rate": 8.180873067353636e-05, "loss": 0.0343, "step": 31650 }, { "grad_norm": 0.3163817226886749, "learning_rate": 8.179597166892052e-05, "loss": 0.0365, "step": 31660 }, { "grad_norm": 0.25451335310935974, "learning_rate": 8.178320918714501e-05, "loss": 0.0277, "step": 31670 }, { "grad_norm": 0.36819881200790405, "learning_rate": 8.177044322960554e-05, "loss": 0.033, "step": 31680 }, { "grad_norm": 0.27520060539245605, "learning_rate": 8.175767379769816e-05, "loss": 0.0288, "step": 31690 }, { "grad_norm": 0.33315786719322205, "learning_rate": 8.174490089281932e-05, "loss": 0.0527, "step": 31700 }, { "grad_norm": 0.372254878282547, "learning_rate": 8.173212451636584e-05, "loss": 0.0364, "step": 31710 }, { "grad_norm": 0.32811760902404785, "learning_rate": 8.171934466973493e-05, "loss": 0.0467, "step": 31720 }, { "grad_norm": 0.2382611632347107, "learning_rate": 8.170656135432418e-05, "loss": 0.0282, "step": 31730 }, { "grad_norm": 0.23958970606327057, "learning_rate": 8.169377457153155e-05, "loss": 0.0329, "step": 31740 }, { "grad_norm": 0.23447099328041077, "learning_rate": 8.168098432275539e-05, "loss": 0.0312, "step": 31750 }, { "grad_norm": 0.27884024381637573, "learning_rate": 8.166819060939442e-05, "loss": 0.0282, "step": 31760 }, { "grad_norm": 0.5192039608955383, "learning_rate": 8.165539343284772e-05, "loss": 0.0375, "step": 31770 }, { "grad_norm": 0.47869575023651123, "learning_rate": 8.16425927945148e-05, "loss": 0.0405, "step": 31780 }, { "grad_norm": 0.26004084944725037, "learning_rate": 8.162978869579551e-05, "loss": 0.0396, "step": 31790 }, { "grad_norm": 0.21522575616836548, "learning_rate": 8.161698113809007e-05, "loss": 0.0306, "step": 31800 }, { "grad_norm": 0.3811260759830475, "learning_rate": 8.160417012279911e-05, "loss": 0.05, "step": 31810 }, { "grad_norm": 0.26710668206214905, "learning_rate": 8.159135565132363e-05, "loss": 0.0401, "step": 31820 }, { "grad_norm": 0.23619891703128815, "learning_rate": 8.157853772506498e-05, "loss": 0.0353, "step": 31830 }, { "grad_norm": 0.2963066101074219, "learning_rate": 8.156571634542494e-05, "loss": 0.0473, "step": 31840 }, { "grad_norm": 0.31622856855392456, "learning_rate": 8.15528915138056e-05, "loss": 0.0374, "step": 31850 }, { "grad_norm": 0.3150166869163513, "learning_rate": 8.154006323160949e-05, "loss": 0.0347, "step": 31860 }, { "grad_norm": 0.21522901952266693, "learning_rate": 8.152723150023949e-05, "loss": 0.0358, "step": 31870 }, { "grad_norm": 0.34887808561325073, "learning_rate": 8.151439632109886e-05, "loss": 0.0336, "step": 31880 }, { "grad_norm": 0.26626506447792053, "learning_rate": 8.150155769559122e-05, "loss": 0.0441, "step": 31890 }, { "grad_norm": 0.34794607758522034, "learning_rate": 8.148871562512058e-05, "loss": 0.0396, "step": 31900 }, { "grad_norm": 0.24880564212799072, "learning_rate": 8.147587011109136e-05, "loss": 0.0433, "step": 31910 }, { "grad_norm": 0.2344679832458496, "learning_rate": 8.14630211549083e-05, "loss": 0.0445, "step": 31920 }, { "grad_norm": 0.32255101203918457, "learning_rate": 8.145016875797655e-05, "loss": 0.0365, "step": 31930 }, { "grad_norm": 0.40809449553489685, "learning_rate": 8.143731292170164e-05, "loss": 0.0379, "step": 31940 }, { "grad_norm": 0.23327326774597168, "learning_rate": 8.142445364748944e-05, "loss": 0.0421, "step": 31950 }, { "grad_norm": 0.396898090839386, "learning_rate": 8.141159093674624e-05, "loss": 0.0397, "step": 31960 }, { "grad_norm": 0.3325514495372772, "learning_rate": 8.139872479087869e-05, "loss": 0.0484, "step": 31970 }, { "grad_norm": 0.28536784648895264, "learning_rate": 8.13858552112938e-05, "loss": 0.0411, "step": 31980 }, { "grad_norm": 0.5553591251373291, "learning_rate": 8.137298219939895e-05, "loss": 0.0514, "step": 31990 }, { "grad_norm": 0.30874234437942505, "learning_rate": 8.136010575660196e-05, "loss": 0.0381, "step": 32000 }, { "grad_norm": 0.36869731545448303, "learning_rate": 8.134722588431095e-05, "loss": 0.0417, "step": 32010 }, { "grad_norm": 0.30882006883621216, "learning_rate": 8.133434258393444e-05, "loss": 0.0483, "step": 32020 }, { "grad_norm": 0.41970768570899963, "learning_rate": 8.132145585688134e-05, "loss": 0.0397, "step": 32030 }, { "grad_norm": 0.49422094225883484, "learning_rate": 8.130856570456093e-05, "loss": 0.051, "step": 32040 }, { "grad_norm": 0.5949768424034119, "learning_rate": 8.129567212838283e-05, "loss": 0.0417, "step": 32050 }, { "grad_norm": 0.42598485946655273, "learning_rate": 8.128277512975708e-05, "loss": 0.0389, "step": 32060 }, { "grad_norm": 0.4250110685825348, "learning_rate": 8.126987471009408e-05, "loss": 0.0357, "step": 32070 }, { "grad_norm": 0.30819278955459595, "learning_rate": 8.125697087080459e-05, "loss": 0.0357, "step": 32080 }, { "grad_norm": 0.2409232258796692, "learning_rate": 8.124406361329976e-05, "loss": 0.0381, "step": 32090 }, { "grad_norm": 0.3560454249382019, "learning_rate": 8.123115293899111e-05, "loss": 0.0307, "step": 32100 }, { "grad_norm": 0.28160154819488525, "learning_rate": 8.121823884929055e-05, "loss": 0.0346, "step": 32110 }, { "grad_norm": 0.3485710918903351, "learning_rate": 8.12053213456103e-05, "loss": 0.0301, "step": 32120 }, { "grad_norm": 0.3244517147541046, "learning_rate": 8.119240042936303e-05, "loss": 0.0347, "step": 32130 }, { "grad_norm": 0.19950033724308014, "learning_rate": 8.117947610196175e-05, "loss": 0.0338, "step": 32140 }, { "grad_norm": 0.33833208680152893, "learning_rate": 8.116654836481982e-05, "loss": 0.0331, "step": 32150 }, { "grad_norm": 0.345819354057312, "learning_rate": 8.115361721935106e-05, "loss": 0.0304, "step": 32160 }, { "grad_norm": 0.25292178988456726, "learning_rate": 8.114068266696953e-05, "loss": 0.027, "step": 32170 }, { "grad_norm": 0.3658265471458435, "learning_rate": 8.112774470908978e-05, "loss": 0.0307, "step": 32180 }, { "grad_norm": 0.49515265226364136, "learning_rate": 8.111480334712665e-05, "loss": 0.0351, "step": 32190 }, { "grad_norm": 0.4311598837375641, "learning_rate": 8.110185858249542e-05, "loss": 0.0352, "step": 32200 }, { "grad_norm": 0.3288404047489166, "learning_rate": 8.108891041661168e-05, "loss": 0.0443, "step": 32210 }, { "grad_norm": 0.39026153087615967, "learning_rate": 8.107595885089146e-05, "loss": 0.0342, "step": 32220 }, { "grad_norm": 0.25448980927467346, "learning_rate": 8.106300388675108e-05, "loss": 0.0364, "step": 32230 }, { "grad_norm": 0.3177433907985687, "learning_rate": 8.105004552560731e-05, "loss": 0.0367, "step": 32240 }, { "grad_norm": 1.182356357574463, "learning_rate": 8.103708376887724e-05, "loss": 0.043, "step": 32250 }, { "grad_norm": 0.3230714499950409, "learning_rate": 8.102411861797836e-05, "loss": 0.0346, "step": 32260 }, { "grad_norm": 0.2540574371814728, "learning_rate": 8.101115007432851e-05, "loss": 0.0327, "step": 32270 }, { "grad_norm": 0.35098403692245483, "learning_rate": 8.09981781393459e-05, "loss": 0.0358, "step": 32280 }, { "grad_norm": 0.29250913858413696, "learning_rate": 8.098520281444915e-05, "loss": 0.0423, "step": 32290 }, { "grad_norm": 0.2996760308742523, "learning_rate": 8.09722241010572e-05, "loss": 0.0413, "step": 32300 }, { "grad_norm": 0.24071311950683594, "learning_rate": 8.095924200058939e-05, "loss": 0.0412, "step": 32310 }, { "grad_norm": 0.7109031081199646, "learning_rate": 8.094625651446541e-05, "loss": 0.0317, "step": 32320 }, { "grad_norm": 0.4114578068256378, "learning_rate": 8.093326764410536e-05, "loss": 0.0369, "step": 32330 }, { "grad_norm": 0.4673689007759094, "learning_rate": 8.092027539092966e-05, "loss": 0.0339, "step": 32340 }, { "grad_norm": 0.2567424476146698, "learning_rate": 8.090727975635913e-05, "loss": 0.0337, "step": 32350 }, { "grad_norm": 0.27965307235717773, "learning_rate": 8.089428074181497e-05, "loss": 0.0344, "step": 32360 }, { "grad_norm": 0.3569695055484772, "learning_rate": 8.088127834871871e-05, "loss": 0.0272, "step": 32370 }, { "grad_norm": 0.36501753330230713, "learning_rate": 8.086827257849226e-05, "loss": 0.0326, "step": 32380 }, { "grad_norm": 0.2606300413608551, "learning_rate": 8.085526343255795e-05, "loss": 0.0354, "step": 32390 }, { "grad_norm": 0.41563957929611206, "learning_rate": 8.084225091233842e-05, "loss": 0.0447, "step": 32400 }, { "grad_norm": 0.47525808215141296, "learning_rate": 8.082923501925668e-05, "loss": 0.0373, "step": 32410 }, { "grad_norm": 0.3203459680080414, "learning_rate": 8.081621575473617e-05, "loss": 0.0421, "step": 32420 }, { "grad_norm": 0.25998878479003906, "learning_rate": 8.080319312020064e-05, "loss": 0.0426, "step": 32430 }, { "grad_norm": 0.3102365732192993, "learning_rate": 8.079016711707421e-05, "loss": 0.0352, "step": 32440 }, { "grad_norm": 0.379508912563324, "learning_rate": 8.077713774678139e-05, "loss": 0.0354, "step": 32450 }, { "grad_norm": 0.2381795048713684, "learning_rate": 8.076410501074707e-05, "loss": 0.0357, "step": 32460 }, { "grad_norm": 0.3483901023864746, "learning_rate": 8.075106891039647e-05, "loss": 0.0382, "step": 32470 }, { "grad_norm": 0.4130554497241974, "learning_rate": 8.073802944715523e-05, "loss": 0.0482, "step": 32480 }, { "grad_norm": 0.41073471307754517, "learning_rate": 8.072498662244929e-05, "loss": 0.0346, "step": 32490 }, { "grad_norm": 0.3409956693649292, "learning_rate": 8.0711940437705e-05, "loss": 0.0324, "step": 32500 }, { "grad_norm": 0.32292842864990234, "learning_rate": 8.06988908943491e-05, "loss": 0.0343, "step": 32510 }, { "grad_norm": 0.2901027798652649, "learning_rate": 8.068583799380863e-05, "loss": 0.028, "step": 32520 }, { "grad_norm": 0.2710975110530853, "learning_rate": 8.067278173751104e-05, "loss": 0.0298, "step": 32530 }, { "grad_norm": 0.35253235697746277, "learning_rate": 8.065972212688417e-05, "loss": 0.0298, "step": 32540 }, { "grad_norm": 0.3779687285423279, "learning_rate": 8.064665916335618e-05, "loss": 0.0305, "step": 32550 }, { "grad_norm": 0.4436417520046234, "learning_rate": 8.063359284835564e-05, "loss": 0.043, "step": 32560 }, { "grad_norm": 0.3406980633735657, "learning_rate": 8.062052318331142e-05, "loss": 0.0414, "step": 32570 }, { "grad_norm": 0.2686534523963928, "learning_rate": 8.060745016965283e-05, "loss": 0.0313, "step": 32580 }, { "grad_norm": 0.2983876168727875, "learning_rate": 8.059437380880952e-05, "loss": 0.0323, "step": 32590 }, { "grad_norm": 0.4603743255138397, "learning_rate": 8.058129410221146e-05, "loss": 0.0492, "step": 32600 }, { "grad_norm": 0.32766827940940857, "learning_rate": 8.056821105128908e-05, "loss": 0.0362, "step": 32610 }, { "grad_norm": 0.24710853397846222, "learning_rate": 8.05551246574731e-05, "loss": 0.0366, "step": 32620 }, { "grad_norm": 0.32185497879981995, "learning_rate": 8.05420349221946e-05, "loss": 0.0436, "step": 32630 }, { "grad_norm": 0.21226656436920166, "learning_rate": 8.05289418468851e-05, "loss": 0.0298, "step": 32640 }, { "grad_norm": 0.6669334173202515, "learning_rate": 8.051584543297642e-05, "loss": 0.0373, "step": 32650 }, { "grad_norm": 0.45230981707572937, "learning_rate": 8.050274568190074e-05, "loss": 0.0366, "step": 32660 }, { "grad_norm": 0.2863248288631439, "learning_rate": 8.048964259509067e-05, "loss": 0.0362, "step": 32670 }, { "grad_norm": 0.41687825322151184, "learning_rate": 8.047653617397914e-05, "loss": 0.0343, "step": 32680 }, { "grad_norm": 0.5162659287452698, "learning_rate": 8.046342641999941e-05, "loss": 0.0449, "step": 32690 }, { "grad_norm": 0.36708036065101624, "learning_rate": 8.045031333458517e-05, "loss": 0.0453, "step": 32700 }, { "grad_norm": 0.47378912568092346, "learning_rate": 8.043719691917047e-05, "loss": 0.0493, "step": 32710 }, { "grad_norm": 0.32515421509742737, "learning_rate": 8.042407717518966e-05, "loss": 0.0298, "step": 32720 }, { "grad_norm": 0.5750540494918823, "learning_rate": 8.041095410407751e-05, "loss": 0.0427, "step": 32730 }, { "grad_norm": 0.2972069978713989, "learning_rate": 8.039782770726913e-05, "loss": 0.034, "step": 32740 }, { "grad_norm": 0.23216882348060608, "learning_rate": 8.038469798620004e-05, "loss": 0.0358, "step": 32750 }, { "grad_norm": 0.43491989374160767, "learning_rate": 8.037156494230604e-05, "loss": 0.0381, "step": 32760 }, { "grad_norm": 0.3685729205608368, "learning_rate": 8.035842857702338e-05, "loss": 0.0375, "step": 32770 }, { "grad_norm": 0.37248796224594116, "learning_rate": 8.03452888917886e-05, "loss": 0.0419, "step": 32780 }, { "grad_norm": 0.49144071340560913, "learning_rate": 8.033214588803866e-05, "loss": 0.0342, "step": 32790 }, { "grad_norm": 0.3152526617050171, "learning_rate": 8.031899956721083e-05, "loss": 0.0337, "step": 32800 }, { "grad_norm": 0.52218097448349, "learning_rate": 8.030584993074282e-05, "loss": 0.0399, "step": 32810 }, { "grad_norm": 0.3205784857273102, "learning_rate": 8.02926969800726e-05, "loss": 0.0404, "step": 32820 }, { "grad_norm": 0.2998005747795105, "learning_rate": 8.027954071663859e-05, "loss": 0.0487, "step": 32830 }, { "grad_norm": 0.4313710629940033, "learning_rate": 8.026638114187954e-05, "loss": 0.0391, "step": 32840 }, { "grad_norm": 0.4737600088119507, "learning_rate": 8.025321825723456e-05, "loss": 0.0493, "step": 32850 }, { "grad_norm": 0.5135376453399658, "learning_rate": 8.02400520641431e-05, "loss": 0.0432, "step": 32860 }, { "grad_norm": 0.26417696475982666, "learning_rate": 8.022688256404501e-05, "loss": 0.0341, "step": 32870 }, { "grad_norm": 0.3499424457550049, "learning_rate": 8.02137097583805e-05, "loss": 0.0352, "step": 32880 }, { "grad_norm": 0.2893335521221161, "learning_rate": 8.02005336485901e-05, "loss": 0.0331, "step": 32890 }, { "grad_norm": 0.3938600718975067, "learning_rate": 8.018735423611476e-05, "loss": 0.0371, "step": 32900 }, { "grad_norm": 1.3370486497879028, "learning_rate": 8.017417152239574e-05, "loss": 0.0425, "step": 32910 }, { "grad_norm": 0.30429473519325256, "learning_rate": 8.01609855088747e-05, "loss": 0.0371, "step": 32920 }, { "grad_norm": 0.3003927171230316, "learning_rate": 8.014779619699362e-05, "loss": 0.0327, "step": 32930 }, { "grad_norm": 0.3698314130306244, "learning_rate": 8.013460358819489e-05, "loss": 0.0421, "step": 32940 }, { "grad_norm": 0.2212209850549698, "learning_rate": 8.01214076839212e-05, "loss": 0.0337, "step": 32950 }, { "grad_norm": 0.3396742641925812, "learning_rate": 8.010820848561565e-05, "loss": 0.0371, "step": 32960 }, { "grad_norm": 0.7367702126502991, "learning_rate": 8.009500599472171e-05, "loss": 0.0326, "step": 32970 }, { "grad_norm": 0.5739076733589172, "learning_rate": 8.008180021268314e-05, "loss": 0.0357, "step": 32980 }, { "grad_norm": 0.40435078740119934, "learning_rate": 8.006859114094414e-05, "loss": 0.0384, "step": 32990 }, { "grad_norm": 0.3246689736843109, "learning_rate": 8.005537878094921e-05, "loss": 0.0377, "step": 33000 }, { "grad_norm": 0.21225330233573914, "learning_rate": 8.004216313414323e-05, "loss": 0.0283, "step": 33010 }, { "grad_norm": 0.45536309480667114, "learning_rate": 8.002894420197149e-05, "loss": 0.0386, "step": 33020 }, { "grad_norm": 0.44215691089630127, "learning_rate": 8.001572198587954e-05, "loss": 0.0379, "step": 33030 }, { "grad_norm": 0.3897514045238495, "learning_rate": 8.000249648731338e-05, "loss": 0.0313, "step": 33040 }, { "grad_norm": 0.40553975105285645, "learning_rate": 7.998926770771928e-05, "loss": 0.0451, "step": 33050 }, { "grad_norm": 0.38902920484542847, "learning_rate": 7.997603564854397e-05, "loss": 0.0453, "step": 33060 }, { "grad_norm": 0.35277581214904785, "learning_rate": 7.996280031123448e-05, "loss": 0.0412, "step": 33070 }, { "grad_norm": 0.3935230076313019, "learning_rate": 7.994956169723818e-05, "loss": 0.0447, "step": 33080 }, { "grad_norm": 0.3277231454849243, "learning_rate": 7.993631980800285e-05, "loss": 0.0332, "step": 33090 }, { "grad_norm": 0.280313640832901, "learning_rate": 7.992307464497659e-05, "loss": 0.0409, "step": 33100 }, { "grad_norm": 0.2675308883190155, "learning_rate": 7.990982620960787e-05, "loss": 0.0306, "step": 33110 }, { "grad_norm": 0.28881537914276123, "learning_rate": 7.989657450334554e-05, "loss": 0.0331, "step": 33120 }, { "grad_norm": 0.333290696144104, "learning_rate": 7.988331952763877e-05, "loss": 0.0336, "step": 33130 }, { "grad_norm": 0.3290376365184784, "learning_rate": 7.987006128393709e-05, "loss": 0.0326, "step": 33140 }, { "grad_norm": 0.26761573553085327, "learning_rate": 7.985679977369043e-05, "loss": 0.0266, "step": 33150 }, { "grad_norm": 0.3592274487018585, "learning_rate": 7.984353499834902e-05, "loss": 0.0408, "step": 33160 }, { "grad_norm": 0.33392050862312317, "learning_rate": 7.983026695936351e-05, "loss": 0.0281, "step": 33170 }, { "grad_norm": 0.2961103618144989, "learning_rate": 7.981699565818486e-05, "loss": 0.0355, "step": 33180 }, { "grad_norm": 0.353800892829895, "learning_rate": 7.980372109626437e-05, "loss": 0.0307, "step": 33190 }, { "grad_norm": 0.47084134817123413, "learning_rate": 7.979044327505375e-05, "loss": 0.0323, "step": 33200 }, { "grad_norm": 0.433240681886673, "learning_rate": 7.977716219600506e-05, "loss": 0.0352, "step": 33210 }, { "grad_norm": 0.46241459250450134, "learning_rate": 7.97638778605707e-05, "loss": 0.0404, "step": 33220 }, { "grad_norm": 0.5074854493141174, "learning_rate": 7.975059027020338e-05, "loss": 0.0453, "step": 33230 }, { "grad_norm": 0.28482094407081604, "learning_rate": 7.973729942635623e-05, "loss": 0.0354, "step": 33240 }, { "grad_norm": 0.7074388861656189, "learning_rate": 7.972400533048273e-05, "loss": 0.0419, "step": 33250 }, { "grad_norm": 0.9643957018852234, "learning_rate": 7.97107079840367e-05, "loss": 0.045, "step": 33260 }, { "grad_norm": 0.3123049736022949, "learning_rate": 7.969740738847231e-05, "loss": 0.0383, "step": 33270 }, { "grad_norm": 0.31651008129119873, "learning_rate": 7.968410354524411e-05, "loss": 0.0399, "step": 33280 }, { "grad_norm": 0.2900896668434143, "learning_rate": 7.967079645580697e-05, "loss": 0.0335, "step": 33290 }, { "grad_norm": 0.30763962864875793, "learning_rate": 7.965748612161612e-05, "loss": 0.0357, "step": 33300 }, { "grad_norm": 0.3808804154396057, "learning_rate": 7.96441725441272e-05, "loss": 0.0333, "step": 33310 }, { "grad_norm": 0.22477903962135315, "learning_rate": 7.963085572479614e-05, "loss": 0.0293, "step": 33320 }, { "grad_norm": 0.28624749183654785, "learning_rate": 7.961753566507924e-05, "loss": 0.0392, "step": 33330 }, { "grad_norm": 0.29024648666381836, "learning_rate": 7.960421236643316e-05, "loss": 0.0388, "step": 33340 }, { "grad_norm": 0.34615740180015564, "learning_rate": 7.959088583031496e-05, "loss": 0.0386, "step": 33350 }, { "grad_norm": 0.22984157502651215, "learning_rate": 7.957755605818194e-05, "loss": 0.0353, "step": 33360 }, { "grad_norm": 0.2567773759365082, "learning_rate": 7.956422305149185e-05, "loss": 0.0413, "step": 33370 }, { "grad_norm": 0.35761579871177673, "learning_rate": 7.95508868117028e-05, "loss": 0.0385, "step": 33380 }, { "grad_norm": 0.45959433913230896, "learning_rate": 7.953754734027318e-05, "loss": 0.0521, "step": 33390 }, { "grad_norm": 0.36325839161872864, "learning_rate": 7.952420463866182e-05, "loss": 0.0381, "step": 33400 }, { "grad_norm": 0.31746289134025574, "learning_rate": 7.951085870832782e-05, "loss": 0.0332, "step": 33410 }, { "grad_norm": 0.27474433183670044, "learning_rate": 7.949750955073067e-05, "loss": 0.0315, "step": 33420 }, { "grad_norm": 0.33382555842399597, "learning_rate": 7.948415716733022e-05, "loss": 0.049, "step": 33430 }, { "grad_norm": 0.4355982840061188, "learning_rate": 7.947080155958669e-05, "loss": 0.0386, "step": 33440 }, { "grad_norm": 0.2259020358324051, "learning_rate": 7.94574427289606e-05, "loss": 0.0301, "step": 33450 }, { "grad_norm": 0.3297049105167389, "learning_rate": 7.944408067691284e-05, "loss": 0.0406, "step": 33460 }, { "grad_norm": 0.44514986872673035, "learning_rate": 7.943071540490473e-05, "loss": 0.031, "step": 33470 }, { "grad_norm": 0.3641186058521271, "learning_rate": 7.94173469143978e-05, "loss": 0.0283, "step": 33480 }, { "grad_norm": 0.294531911611557, "learning_rate": 7.940397520685406e-05, "loss": 0.0372, "step": 33490 }, { "grad_norm": 0.370016872882843, "learning_rate": 7.939060028373577e-05, "loss": 0.0383, "step": 33500 }, { "grad_norm": 0.3359000086784363, "learning_rate": 7.937722214650565e-05, "loss": 0.0459, "step": 33510 }, { "grad_norm": 0.4196845293045044, "learning_rate": 7.936384079662666e-05, "loss": 0.0359, "step": 33520 }, { "grad_norm": 0.3376953899860382, "learning_rate": 7.93504562355622e-05, "loss": 0.0393, "step": 33530 }, { "grad_norm": 0.5064026117324829, "learning_rate": 7.933706846477599e-05, "loss": 0.0426, "step": 33540 }, { "grad_norm": 0.36117494106292725, "learning_rate": 7.932367748573206e-05, "loss": 0.0413, "step": 33550 }, { "grad_norm": 0.30517786741256714, "learning_rate": 7.931028329989485e-05, "loss": 0.0393, "step": 33560 }, { "grad_norm": 0.27602633833885193, "learning_rate": 7.929688590872913e-05, "loss": 0.0435, "step": 33570 }, { "grad_norm": 0.49918878078460693, "learning_rate": 7.928348531370003e-05, "loss": 0.0375, "step": 33580 }, { "grad_norm": 0.7362155318260193, "learning_rate": 7.927008151627297e-05, "loss": 0.0334, "step": 33590 }, { "grad_norm": 0.2847193479537964, "learning_rate": 7.925667451791383e-05, "loss": 0.0351, "step": 33600 }, { "grad_norm": 0.2874413728713989, "learning_rate": 7.924326432008874e-05, "loss": 0.0377, "step": 33610 }, { "grad_norm": 0.4454072415828705, "learning_rate": 7.922985092426422e-05, "loss": 0.0406, "step": 33620 }, { "grad_norm": 0.3069995045661926, "learning_rate": 7.921643433190717e-05, "loss": 0.0349, "step": 33630 }, { "grad_norm": 0.24402062594890594, "learning_rate": 7.920301454448478e-05, "loss": 0.0321, "step": 33640 }, { "grad_norm": 0.36244839429855347, "learning_rate": 7.918959156346461e-05, "loss": 0.0387, "step": 33650 }, { "grad_norm": 0.3033449649810791, "learning_rate": 7.91761653903146e-05, "loss": 0.041, "step": 33660 }, { "grad_norm": 0.37705492973327637, "learning_rate": 7.916273602650302e-05, "loss": 0.0377, "step": 33670 }, { "grad_norm": 0.24605241417884827, "learning_rate": 7.914930347349847e-05, "loss": 0.03, "step": 33680 }, { "grad_norm": 0.3914673626422882, "learning_rate": 7.913586773276992e-05, "loss": 0.0375, "step": 33690 }, { "grad_norm": 0.33948513865470886, "learning_rate": 7.912242880578667e-05, "loss": 0.0375, "step": 33700 }, { "grad_norm": 0.4278040826320648, "learning_rate": 7.910898669401839e-05, "loss": 0.0434, "step": 33710 }, { "grad_norm": 0.32731375098228455, "learning_rate": 7.909554139893511e-05, "loss": 0.033, "step": 33720 }, { "grad_norm": 0.2887725234031677, "learning_rate": 7.908209292200715e-05, "loss": 0.0398, "step": 33730 }, { "grad_norm": 0.24887414276599884, "learning_rate": 7.906864126470523e-05, "loss": 0.0342, "step": 33740 }, { "grad_norm": 0.33423227071762085, "learning_rate": 7.905518642850041e-05, "loss": 0.0369, "step": 33750 }, { "grad_norm": 0.3365819454193115, "learning_rate": 7.904172841486409e-05, "loss": 0.0353, "step": 33760 }, { "grad_norm": 0.4688853323459625, "learning_rate": 7.902826722526801e-05, "loss": 0.035, "step": 33770 }, { "grad_norm": 0.28802546858787537, "learning_rate": 7.901480286118427e-05, "loss": 0.0328, "step": 33780 }, { "grad_norm": 0.2407655119895935, "learning_rate": 7.900133532408531e-05, "loss": 0.0358, "step": 33790 }, { "grad_norm": 0.3568786084651947, "learning_rate": 7.898786461544395e-05, "loss": 0.0379, "step": 33800 }, { "grad_norm": 0.2828184962272644, "learning_rate": 7.897439073673325e-05, "loss": 0.0327, "step": 33810 }, { "grad_norm": 0.30302131175994873, "learning_rate": 7.896091368942677e-05, "loss": 0.0282, "step": 33820 }, { "grad_norm": 0.24021165072917938, "learning_rate": 7.894743347499832e-05, "loss": 0.0369, "step": 33830 }, { "grad_norm": 0.30461952090263367, "learning_rate": 7.893395009492203e-05, "loss": 0.0344, "step": 33840 }, { "grad_norm": 0.41990751028060913, "learning_rate": 7.892046355067248e-05, "loss": 0.0313, "step": 33850 }, { "grad_norm": 0.3131566643714905, "learning_rate": 7.890697384372451e-05, "loss": 0.0348, "step": 33860 }, { "grad_norm": 0.23579804599285126, "learning_rate": 7.889348097555336e-05, "loss": 0.0317, "step": 33870 }, { "grad_norm": 0.6174931526184082, "learning_rate": 7.887998494763455e-05, "loss": 0.0349, "step": 33880 }, { "grad_norm": 0.3526611030101776, "learning_rate": 7.886648576144404e-05, "loss": 0.0387, "step": 33890 }, { "grad_norm": 0.2863526940345764, "learning_rate": 7.885298341845802e-05, "loss": 0.0357, "step": 33900 }, { "grad_norm": 0.38137397170066833, "learning_rate": 7.883947792015311e-05, "loss": 0.0338, "step": 33910 }, { "grad_norm": 0.4397212266921997, "learning_rate": 7.882596926800628e-05, "loss": 0.0582, "step": 33920 }, { "grad_norm": 0.42462438344955444, "learning_rate": 7.881245746349477e-05, "loss": 0.0345, "step": 33930 }, { "grad_norm": 0.433310329914093, "learning_rate": 7.879894250809623e-05, "loss": 0.0337, "step": 33940 }, { "grad_norm": 0.3690340220928192, "learning_rate": 7.878542440328865e-05, "loss": 0.0546, "step": 33950 }, { "grad_norm": 0.3298906087875366, "learning_rate": 7.877190315055031e-05, "loss": 0.0361, "step": 33960 }, { "grad_norm": 0.347944051027298, "learning_rate": 7.875837875135991e-05, "loss": 0.0359, "step": 33970 }, { "grad_norm": 0.25469204783439636, "learning_rate": 7.874485120719646e-05, "loss": 0.0306, "step": 33980 }, { "grad_norm": 0.3231489062309265, "learning_rate": 7.873132051953928e-05, "loss": 0.0281, "step": 33990 }, { "grad_norm": 0.21503911912441254, "learning_rate": 7.87177866898681e-05, "loss": 0.0313, "step": 34000 }, { "grad_norm": 0.25782859325408936, "learning_rate": 7.870424971966294e-05, "loss": 0.0331, "step": 34010 }, { "grad_norm": 0.3009248971939087, "learning_rate": 7.869070961040419e-05, "loss": 0.0317, "step": 34020 }, { "grad_norm": 0.3734458386898041, "learning_rate": 7.867716636357257e-05, "loss": 0.034, "step": 34030 }, { "grad_norm": 0.2766494154930115, "learning_rate": 7.866361998064915e-05, "loss": 0.0354, "step": 34040 }, { "grad_norm": 0.3247620761394501, "learning_rate": 7.865007046311534e-05, "loss": 0.0311, "step": 34050 }, { "grad_norm": 0.3458101749420166, "learning_rate": 7.86365178124529e-05, "loss": 0.0255, "step": 34060 }, { "grad_norm": 0.2341909110546112, "learning_rate": 7.862296203014394e-05, "loss": 0.0313, "step": 34070 }, { "grad_norm": 0.45910438895225525, "learning_rate": 7.860940311767088e-05, "loss": 0.0337, "step": 34080 }, { "grad_norm": 0.36315497756004333, "learning_rate": 7.85958410765165e-05, "loss": 0.0306, "step": 34090 }, { "grad_norm": 0.34128591418266296, "learning_rate": 7.858227590816394e-05, "loss": 0.0295, "step": 34100 }, { "grad_norm": 0.30696046352386475, "learning_rate": 7.856870761409664e-05, "loss": 0.0362, "step": 34110 }, { "grad_norm": 0.2900685667991638, "learning_rate": 7.855513619579846e-05, "loss": 0.0314, "step": 34120 }, { "grad_norm": 0.4001160264015198, "learning_rate": 7.85415616547535e-05, "loss": 0.0322, "step": 34130 }, { "grad_norm": 0.42180776596069336, "learning_rate": 7.852798399244627e-05, "loss": 0.0285, "step": 34140 }, { "grad_norm": 0.3293079137802124, "learning_rate": 7.851440321036161e-05, "loss": 0.0243, "step": 34150 }, { "grad_norm": 0.2263602465391159, "learning_rate": 7.850081930998468e-05, "loss": 0.0352, "step": 34160 }, { "grad_norm": 0.41925448179244995, "learning_rate": 7.8487232292801e-05, "loss": 0.0379, "step": 34170 }, { "grad_norm": 0.3554747402667999, "learning_rate": 7.847364216029642e-05, "loss": 0.0347, "step": 34180 }, { "grad_norm": 0.5431614518165588, "learning_rate": 7.846004891395716e-05, "loss": 0.0345, "step": 34190 }, { "grad_norm": 0.2710590958595276, "learning_rate": 7.844645255526972e-05, "loss": 0.0321, "step": 34200 }, { "grad_norm": 0.35557568073272705, "learning_rate": 7.843285308572101e-05, "loss": 0.0267, "step": 34210 }, { "grad_norm": 0.3280973732471466, "learning_rate": 7.841925050679823e-05, "loss": 0.0335, "step": 34220 }, { "grad_norm": 0.46259114146232605, "learning_rate": 7.840564481998895e-05, "loss": 0.0428, "step": 34230 }, { "grad_norm": 0.30673328042030334, "learning_rate": 7.839203602678105e-05, "loss": 0.0275, "step": 34240 }, { "grad_norm": 0.32056674361228943, "learning_rate": 7.837842412866279e-05, "loss": 0.0348, "step": 34250 }, { "grad_norm": 0.26009780168533325, "learning_rate": 7.836480912712272e-05, "loss": 0.0314, "step": 34260 }, { "grad_norm": 0.4117041528224945, "learning_rate": 7.835119102364976e-05, "loss": 0.039, "step": 34270 }, { "grad_norm": 0.32385098934173584, "learning_rate": 7.83375698197332e-05, "loss": 0.0339, "step": 34280 }, { "grad_norm": 0.36194953322410583, "learning_rate": 7.83239455168626e-05, "loss": 0.0349, "step": 34290 }, { "grad_norm": 0.24092033505439758, "learning_rate": 7.83103181165279e-05, "loss": 0.0251, "step": 34300 }, { "grad_norm": 0.26485636830329895, "learning_rate": 7.829668762021937e-05, "loss": 0.0367, "step": 34310 }, { "grad_norm": 0.9526070952415466, "learning_rate": 7.828305402942764e-05, "loss": 0.0306, "step": 34320 }, { "grad_norm": 0.3207401931285858, "learning_rate": 7.826941734564363e-05, "loss": 0.0283, "step": 34330 }, { "grad_norm": 0.22192750871181488, "learning_rate": 7.825577757035865e-05, "loss": 0.0293, "step": 34340 }, { "grad_norm": 0.20166225731372833, "learning_rate": 7.824213470506431e-05, "loss": 0.0367, "step": 34350 }, { "grad_norm": 0.5143682360649109, "learning_rate": 7.822848875125257e-05, "loss": 0.041, "step": 34360 }, { "grad_norm": 0.3140559494495392, "learning_rate": 7.821483971041576e-05, "loss": 0.0329, "step": 34370 }, { "grad_norm": 0.2996448874473572, "learning_rate": 7.820118758404649e-05, "loss": 0.0386, "step": 34380 }, { "grad_norm": 0.2769199013710022, "learning_rate": 7.818753237363776e-05, "loss": 0.0383, "step": 34390 }, { "grad_norm": 0.40897223353385925, "learning_rate": 7.817387408068286e-05, "loss": 0.0316, "step": 34400 }, { "grad_norm": 1.0305417776107788, "learning_rate": 7.816021270667544e-05, "loss": 0.0443, "step": 34410 }, { "grad_norm": 0.3414754569530487, "learning_rate": 7.81465482531095e-05, "loss": 0.0299, "step": 34420 }, { "grad_norm": 0.5667785406112671, "learning_rate": 7.813288072147938e-05, "loss": 0.0435, "step": 34430 }, { "grad_norm": 0.5389432311058044, "learning_rate": 7.811921011327972e-05, "loss": 0.0413, "step": 34440 }, { "grad_norm": 0.20337873697280884, "learning_rate": 7.810553643000549e-05, "loss": 0.0441, "step": 34450 }, { "grad_norm": 0.3879426121711731, "learning_rate": 7.809185967315206e-05, "loss": 0.0387, "step": 34460 }, { "grad_norm": 0.34993451833724976, "learning_rate": 7.80781798442151e-05, "loss": 0.0391, "step": 34470 }, { "grad_norm": 0.3949001133441925, "learning_rate": 7.806449694469061e-05, "loss": 0.0276, "step": 34480 }, { "grad_norm": 0.25646740198135376, "learning_rate": 7.805081097607492e-05, "loss": 0.0416, "step": 34490 }, { "grad_norm": 0.43993818759918213, "learning_rate": 7.803712193986474e-05, "loss": 0.0306, "step": 34500 }, { "grad_norm": 0.3073083460330963, "learning_rate": 7.802342983755702e-05, "loss": 0.0334, "step": 34510 }, { "grad_norm": 0.529073178768158, "learning_rate": 7.800973467064918e-05, "loss": 0.042, "step": 34520 }, { "grad_norm": 0.4724183678627014, "learning_rate": 7.799603644063884e-05, "loss": 0.0434, "step": 34530 }, { "grad_norm": 0.3997385799884796, "learning_rate": 7.798233514902405e-05, "loss": 0.035, "step": 34540 }, { "grad_norm": 0.2651580274105072, "learning_rate": 7.796863079730318e-05, "loss": 0.0356, "step": 34550 }, { "grad_norm": 0.30540579557418823, "learning_rate": 7.795492338697488e-05, "loss": 0.0308, "step": 34560 }, { "grad_norm": 0.3729653060436249, "learning_rate": 7.794121291953819e-05, "loss": 0.0396, "step": 34570 }, { "grad_norm": 0.3412588834762573, "learning_rate": 7.792749939649246e-05, "loss": 0.0388, "step": 34580 }, { "grad_norm": 0.306649774312973, "learning_rate": 7.79137828193374e-05, "loss": 0.027, "step": 34590 }, { "grad_norm": 0.38774609565734863, "learning_rate": 7.790006318957301e-05, "loss": 0.0421, "step": 34600 }, { "grad_norm": 0.31285279989242554, "learning_rate": 7.788634050869965e-05, "loss": 0.0345, "step": 34610 }, { "grad_norm": 0.35755595564842224, "learning_rate": 7.787261477821803e-05, "loss": 0.0355, "step": 34620 }, { "grad_norm": 0.39406225085258484, "learning_rate": 7.785888599962916e-05, "loss": 0.0256, "step": 34630 }, { "grad_norm": 0.4164443612098694, "learning_rate": 7.784515417443439e-05, "loss": 0.0358, "step": 34640 }, { "grad_norm": 0.34750354290008545, "learning_rate": 7.783141930413545e-05, "loss": 0.043, "step": 34650 }, { "grad_norm": 0.2603650391101837, "learning_rate": 7.78176813902343e-05, "loss": 0.0272, "step": 34660 }, { "grad_norm": 0.25292399525642395, "learning_rate": 7.780394043423336e-05, "loss": 0.0319, "step": 34670 }, { "grad_norm": 0.3490259647369385, "learning_rate": 7.77901964376353e-05, "loss": 0.0344, "step": 34680 }, { "grad_norm": 0.2738434672355652, "learning_rate": 7.777644940194316e-05, "loss": 0.04, "step": 34690 }, { "grad_norm": 0.3942566215991974, "learning_rate": 7.776269932866023e-05, "loss": 0.0372, "step": 34700 }, { "grad_norm": 0.6964448690414429, "learning_rate": 7.774894621929026e-05, "loss": 0.0353, "step": 34710 }, { "grad_norm": 0.3314773738384247, "learning_rate": 7.773519007533725e-05, "loss": 0.0521, "step": 34720 }, { "grad_norm": 0.2784637212753296, "learning_rate": 7.772143089830556e-05, "loss": 0.0345, "step": 34730 }, { "grad_norm": 0.41192498803138733, "learning_rate": 7.770766868969985e-05, "loss": 0.034, "step": 34740 }, { "grad_norm": 0.4449058473110199, "learning_rate": 7.769390345102518e-05, "loss": 0.0348, "step": 34750 }, { "grad_norm": 0.33842310309410095, "learning_rate": 7.768013518378683e-05, "loss": 0.0435, "step": 34760 }, { "grad_norm": 0.42665043473243713, "learning_rate": 7.766636388949053e-05, "loss": 0.0278, "step": 34770 }, { "grad_norm": 0.4121442437171936, "learning_rate": 7.765258956964229e-05, "loss": 0.055, "step": 34780 }, { "grad_norm": 0.30590322613716125, "learning_rate": 7.76388122257484e-05, "loss": 0.0324, "step": 34790 }, { "grad_norm": 0.5803946852684021, "learning_rate": 7.762503185931558e-05, "loss": 0.0283, "step": 34800 }, { "grad_norm": 0.27764663100242615, "learning_rate": 7.76112484718508e-05, "loss": 0.0386, "step": 34810 }, { "grad_norm": 0.6028018593788147, "learning_rate": 7.75974620648614e-05, "loss": 0.0483, "step": 34820 }, { "grad_norm": 0.35191088914871216, "learning_rate": 7.758367263985503e-05, "loss": 0.0438, "step": 34830 }, { "grad_norm": 0.5010019540786743, "learning_rate": 7.75698801983397e-05, "loss": 0.0296, "step": 34840 }, { "grad_norm": 0.47284621000289917, "learning_rate": 7.755608474182372e-05, "loss": 0.0532, "step": 34850 }, { "grad_norm": 0.7928063869476318, "learning_rate": 7.754228627181574e-05, "loss": 0.0448, "step": 34860 }, { "grad_norm": 0.38481900095939636, "learning_rate": 7.752848478982476e-05, "loss": 0.0431, "step": 34870 }, { "grad_norm": 0.49346065521240234, "learning_rate": 7.751468029736006e-05, "loss": 0.0368, "step": 34880 }, { "grad_norm": 1.1035584211349487, "learning_rate": 7.750087279593129e-05, "loss": 0.0466, "step": 34890 }, { "grad_norm": 0.3174028992652893, "learning_rate": 7.748706228704843e-05, "loss": 0.0333, "step": 34900 }, { "grad_norm": 0.2579417824745178, "learning_rate": 7.747324877222176e-05, "loss": 0.0414, "step": 34910 }, { "grad_norm": 0.49320513010025024, "learning_rate": 7.745943225296188e-05, "loss": 0.0388, "step": 34920 }, { "grad_norm": 0.30834922194480896, "learning_rate": 7.744561273077981e-05, "loss": 0.0335, "step": 34930 }, { "grad_norm": 0.5074669718742371, "learning_rate": 7.743179020718678e-05, "loss": 0.0407, "step": 34940 }, { "grad_norm": 0.4530579149723053, "learning_rate": 7.741796468369443e-05, "loss": 0.0493, "step": 34950 }, { "grad_norm": 0.45311787724494934, "learning_rate": 7.740413616181466e-05, "loss": 0.0469, "step": 34960 }, { "grad_norm": 0.2590677738189697, "learning_rate": 7.739030464305978e-05, "loss": 0.0398, "step": 34970 }, { "grad_norm": 0.23222245275974274, "learning_rate": 7.737647012894235e-05, "loss": 0.041, "step": 34980 }, { "grad_norm": 0.31209707260131836, "learning_rate": 7.736263262097532e-05, "loss": 0.037, "step": 34990 }, { "grad_norm": 0.35328614711761475, "learning_rate": 7.734879212067192e-05, "loss": 0.0387, "step": 35000 }, { "grad_norm": 0.30510565638542175, "learning_rate": 7.733494862954573e-05, "loss": 0.0396, "step": 35010 }, { "grad_norm": 0.33856821060180664, "learning_rate": 7.732110214911066e-05, "loss": 0.0411, "step": 35020 }, { "grad_norm": 0.4045318365097046, "learning_rate": 7.730725268088094e-05, "loss": 0.0437, "step": 35030 }, { "grad_norm": 0.3693859875202179, "learning_rate": 7.729340022637112e-05, "loss": 0.0389, "step": 35040 }, { "grad_norm": 0.23784981667995453, "learning_rate": 7.727954478709607e-05, "loss": 0.0323, "step": 35050 }, { "grad_norm": 0.5615198612213135, "learning_rate": 7.726568636457103e-05, "loss": 0.0339, "step": 35060 }, { "grad_norm": 0.3228866755962372, "learning_rate": 7.725182496031153e-05, "loss": 0.0393, "step": 35070 }, { "grad_norm": 0.2795256972312927, "learning_rate": 7.723796057583341e-05, "loss": 0.0349, "step": 35080 }, { "grad_norm": 0.32289978861808777, "learning_rate": 7.722409321265287e-05, "loss": 0.0375, "step": 35090 }, { "grad_norm": 0.5062304139137268, "learning_rate": 7.721022287228645e-05, "loss": 0.0461, "step": 35100 }, { "grad_norm": 0.6848072409629822, "learning_rate": 7.719634955625094e-05, "loss": 0.0274, "step": 35110 }, { "grad_norm": 0.31371384859085083, "learning_rate": 7.718247326606355e-05, "loss": 0.0281, "step": 35120 }, { "grad_norm": 0.3235814869403839, "learning_rate": 7.716859400324175e-05, "loss": 0.0308, "step": 35130 }, { "grad_norm": 0.34667426347732544, "learning_rate": 7.715471176930335e-05, "loss": 0.0501, "step": 35140 }, { "grad_norm": 0.25272077322006226, "learning_rate": 7.714082656576651e-05, "loss": 0.0281, "step": 35150 }, { "grad_norm": 0.40144649147987366, "learning_rate": 7.712693839414968e-05, "loss": 0.0376, "step": 35160 }, { "grad_norm": 0.37000206112861633, "learning_rate": 7.711304725597164e-05, "loss": 0.0323, "step": 35170 }, { "grad_norm": 0.3598545789718628, "learning_rate": 7.709915315275151e-05, "loss": 0.0306, "step": 35180 }, { "grad_norm": 0.6548917889595032, "learning_rate": 7.708525608600876e-05, "loss": 0.0353, "step": 35190 }, { "grad_norm": 0.273237943649292, "learning_rate": 7.707135605726311e-05, "loss": 0.0346, "step": 35200 }, { "grad_norm": 0.24954457581043243, "learning_rate": 7.705745306803466e-05, "loss": 0.0501, "step": 35210 }, { "grad_norm": 0.3205318748950958, "learning_rate": 7.704354711984383e-05, "loss": 0.0374, "step": 35220 }, { "grad_norm": 0.30175307393074036, "learning_rate": 7.702963821421133e-05, "loss": 0.0381, "step": 35230 }, { "grad_norm": 0.4543519616127014, "learning_rate": 7.701572635265826e-05, "loss": 0.0323, "step": 35240 }, { "grad_norm": 0.31601187586784363, "learning_rate": 7.700181153670596e-05, "loss": 0.0315, "step": 35250 }, { "grad_norm": 0.29144200682640076, "learning_rate": 7.698789376787616e-05, "loss": 0.0334, "step": 35260 }, { "grad_norm": 0.2742979824542999, "learning_rate": 7.697397304769085e-05, "loss": 0.0375, "step": 35270 }, { "grad_norm": 0.34495025873184204, "learning_rate": 7.696004937767241e-05, "loss": 0.0418, "step": 35280 }, { "grad_norm": 0.3065648376941681, "learning_rate": 7.694612275934352e-05, "loss": 0.0289, "step": 35290 }, { "grad_norm": 0.2535524070262909, "learning_rate": 7.693219319422714e-05, "loss": 0.0278, "step": 35300 }, { "grad_norm": 2.523146867752075, "learning_rate": 7.69182606838466e-05, "loss": 0.033, "step": 35310 }, { "grad_norm": 0.28924182057380676, "learning_rate": 7.690432522972558e-05, "loss": 0.0407, "step": 35320 }, { "grad_norm": 0.8669989705085754, "learning_rate": 7.689038683338796e-05, "loss": 0.0415, "step": 35330 }, { "grad_norm": 0.3167360723018646, "learning_rate": 7.687644549635808e-05, "loss": 0.033, "step": 35340 }, { "grad_norm": 0.8080126643180847, "learning_rate": 7.686250122016053e-05, "loss": 0.0451, "step": 35350 }, { "grad_norm": 0.22601565718650818, "learning_rate": 7.684855400632022e-05, "loss": 0.0271, "step": 35360 }, { "grad_norm": 0.3423301875591278, "learning_rate": 7.683460385636243e-05, "loss": 0.03, "step": 35370 }, { "grad_norm": 0.3234097957611084, "learning_rate": 7.68206507718127e-05, "loss": 0.0317, "step": 35380 }, { "grad_norm": 0.6065872311592102, "learning_rate": 7.680669475419693e-05, "loss": 0.0389, "step": 35390 }, { "grad_norm": 0.2692949175834656, "learning_rate": 7.679273580504132e-05, "loss": 0.0385, "step": 35400 }, { "grad_norm": 0.25153201818466187, "learning_rate": 7.67787739258724e-05, "loss": 0.031, "step": 35410 }, { "grad_norm": 0.3806419372558594, "learning_rate": 7.676480911821705e-05, "loss": 0.0398, "step": 35420 }, { "grad_norm": 0.30760878324508667, "learning_rate": 7.675084138360238e-05, "loss": 0.0331, "step": 35430 }, { "grad_norm": 0.31768864393234253, "learning_rate": 7.673687072355592e-05, "loss": 0.0364, "step": 35440 }, { "grad_norm": 0.2878972291946411, "learning_rate": 7.67228971396055e-05, "loss": 0.0363, "step": 35450 }, { "grad_norm": 0.37697944045066833, "learning_rate": 7.670892063327922e-05, "loss": 0.0346, "step": 35460 }, { "grad_norm": 0.3680899143218994, "learning_rate": 7.669494120610552e-05, "loss": 0.0427, "step": 35470 }, { "grad_norm": 0.5920754671096802, "learning_rate": 7.66809588596132e-05, "loss": 0.0421, "step": 35480 }, { "grad_norm": 0.35448822379112244, "learning_rate": 7.666697359533132e-05, "loss": 0.0347, "step": 35490 }, { "grad_norm": 0.2703777849674225, "learning_rate": 7.665298541478932e-05, "loss": 0.0319, "step": 35500 }, { "grad_norm": 0.3392300605773926, "learning_rate": 7.66389943195169e-05, "loss": 0.0438, "step": 35510 }, { "grad_norm": 0.31760793924331665, "learning_rate": 7.66250003110441e-05, "loss": 0.0392, "step": 35520 }, { "grad_norm": 0.37545377016067505, "learning_rate": 7.66110033909013e-05, "loss": 0.0395, "step": 35530 }, { "grad_norm": 0.35463836789131165, "learning_rate": 7.659700356061918e-05, "loss": 0.0385, "step": 35540 }, { "grad_norm": 0.23895540833473206, "learning_rate": 7.658300082172875e-05, "loss": 0.0339, "step": 35550 }, { "grad_norm": 0.5512027740478516, "learning_rate": 7.65689951757613e-05, "loss": 0.0354, "step": 35560 }, { "grad_norm": 0.3099910318851471, "learning_rate": 7.655498662424849e-05, "loss": 0.0282, "step": 35570 }, { "grad_norm": 0.2504037618637085, "learning_rate": 7.654097516872227e-05, "loss": 0.0316, "step": 35580 }, { "grad_norm": 0.29670286178588867, "learning_rate": 7.65269608107149e-05, "loss": 0.0301, "step": 35590 }, { "grad_norm": 0.36565008759498596, "learning_rate": 7.6512943551759e-05, "loss": 0.0423, "step": 35600 }, { "grad_norm": 0.3511713147163391, "learning_rate": 7.649892339338743e-05, "loss": 0.0303, "step": 35610 }, { "grad_norm": 0.23814526200294495, "learning_rate": 7.648490033713344e-05, "loss": 0.0282, "step": 35620 }, { "grad_norm": 0.21418607234954834, "learning_rate": 7.647087438453058e-05, "loss": 0.0346, "step": 35630 }, { "grad_norm": 0.38168880343437195, "learning_rate": 7.64568455371127e-05, "loss": 0.0414, "step": 35640 }, { "grad_norm": 0.33895060420036316, "learning_rate": 7.644281379641396e-05, "loss": 0.0414, "step": 35650 }, { "grad_norm": 0.33267685770988464, "learning_rate": 7.642877916396887e-05, "loss": 0.0335, "step": 35660 }, { "grad_norm": 0.36615750193595886, "learning_rate": 7.641474164131221e-05, "loss": 0.0333, "step": 35670 }, { "grad_norm": 0.39413920044898987, "learning_rate": 7.640070122997913e-05, "loss": 0.0357, "step": 35680 }, { "grad_norm": 0.29731932282447815, "learning_rate": 7.638665793150505e-05, "loss": 0.0274, "step": 35690 }, { "grad_norm": 0.3628135323524475, "learning_rate": 7.637261174742574e-05, "loss": 0.0496, "step": 35700 }, { "grad_norm": 0.23888052999973297, "learning_rate": 7.635856267927725e-05, "loss": 0.0286, "step": 35710 }, { "grad_norm": 0.7204527258872986, "learning_rate": 7.634451072859597e-05, "loss": 0.0366, "step": 35720 }, { "grad_norm": 0.3708757162094116, "learning_rate": 7.633045589691863e-05, "loss": 0.0391, "step": 35730 }, { "grad_norm": 0.2989940047264099, "learning_rate": 7.63163981857822e-05, "loss": 0.0287, "step": 35740 }, { "grad_norm": 0.46999913454055786, "learning_rate": 7.630233759672403e-05, "loss": 0.0382, "step": 35750 }, { "grad_norm": 0.3502756953239441, "learning_rate": 7.628827413128177e-05, "loss": 0.042, "step": 35760 }, { "grad_norm": 0.30278512835502625, "learning_rate": 7.627420779099338e-05, "loss": 0.0356, "step": 35770 }, { "grad_norm": 0.2751677334308624, "learning_rate": 7.626013857739711e-05, "loss": 0.0411, "step": 35780 }, { "grad_norm": 0.4692207872867584, "learning_rate": 7.624606649203158e-05, "loss": 0.0296, "step": 35790 }, { "grad_norm": 0.2041904330253601, "learning_rate": 7.623199153643569e-05, "loss": 0.0481, "step": 35800 }, { "grad_norm": 0.29231491684913635, "learning_rate": 7.621791371214863e-05, "loss": 0.0449, "step": 35810 }, { "grad_norm": 0.41260120272636414, "learning_rate": 7.620383302070995e-05, "loss": 0.0443, "step": 35820 }, { "grad_norm": 0.2824031412601471, "learning_rate": 7.61897494636595e-05, "loss": 0.0385, "step": 35830 }, { "grad_norm": 0.35115382075309753, "learning_rate": 7.617566304253739e-05, "loss": 0.0439, "step": 35840 }, { "grad_norm": 0.3383215367794037, "learning_rate": 7.616157375888416e-05, "loss": 0.0382, "step": 35850 }, { "grad_norm": 0.32998543977737427, "learning_rate": 7.614748161424053e-05, "loss": 0.0357, "step": 35860 }, { "grad_norm": 0.46293774247169495, "learning_rate": 7.613338661014763e-05, "loss": 0.0408, "step": 35870 }, { "grad_norm": 0.3527509868144989, "learning_rate": 7.611928874814686e-05, "loss": 0.0354, "step": 35880 }, { "grad_norm": 0.25345349311828613, "learning_rate": 7.610518802977996e-05, "loss": 0.0355, "step": 35890 }, { "grad_norm": 0.3325011134147644, "learning_rate": 7.609108445658893e-05, "loss": 0.0361, "step": 35900 }, { "grad_norm": 0.3161720335483551, "learning_rate": 7.607697803011612e-05, "loss": 0.0427, "step": 35910 }, { "grad_norm": 0.36303776502609253, "learning_rate": 7.606286875190421e-05, "loss": 0.0392, "step": 35920 }, { "grad_norm": 0.3267717957496643, "learning_rate": 7.604875662349617e-05, "loss": 0.03, "step": 35930 }, { "grad_norm": 0.3667326271533966, "learning_rate": 7.603464164643525e-05, "loss": 0.0374, "step": 35940 }, { "grad_norm": 0.21771767735481262, "learning_rate": 7.602052382226507e-05, "loss": 0.0333, "step": 35950 }, { "grad_norm": 0.2926994264125824, "learning_rate": 7.600640315252954e-05, "loss": 0.0292, "step": 35960 }, { "grad_norm": 0.362684428691864, "learning_rate": 7.599227963877284e-05, "loss": 0.0383, "step": 35970 }, { "grad_norm": 0.3337101638317108, "learning_rate": 7.597815328253954e-05, "loss": 0.0313, "step": 35980 }, { "grad_norm": 0.23705528676509857, "learning_rate": 7.596402408537443e-05, "loss": 0.0284, "step": 35990 }, { "grad_norm": 0.3510713279247284, "learning_rate": 7.59498920488227e-05, "loss": 0.0439, "step": 36000 }, { "grad_norm": 0.2815088927745819, "learning_rate": 7.593575717442979e-05, "loss": 0.0325, "step": 36010 }, { "grad_norm": 0.34400737285614014, "learning_rate": 7.592161946374147e-05, "loss": 0.0333, "step": 36020 }, { "grad_norm": 0.23668351769447327, "learning_rate": 7.590747891830381e-05, "loss": 0.0326, "step": 36030 }, { "grad_norm": 0.375296950340271, "learning_rate": 7.58933355396632e-05, "loss": 0.0342, "step": 36040 }, { "grad_norm": 0.2517029941082001, "learning_rate": 7.587918932936636e-05, "loss": 0.0279, "step": 36050 }, { "grad_norm": 0.3432581424713135, "learning_rate": 7.586504028896028e-05, "loss": 0.0288, "step": 36060 }, { "grad_norm": 0.28860801458358765, "learning_rate": 7.585088841999228e-05, "loss": 0.028, "step": 36070 }, { "grad_norm": 0.2666780650615692, "learning_rate": 7.583673372400999e-05, "loss": 0.0251, "step": 36080 }, { "grad_norm": 0.19802120327949524, "learning_rate": 7.582257620256134e-05, "loss": 0.0268, "step": 36090 }, { "grad_norm": 0.3234136998653412, "learning_rate": 7.580841585719458e-05, "loss": 0.0293, "step": 36100 }, { "grad_norm": 0.26657816767692566, "learning_rate": 7.579425268945825e-05, "loss": 0.0283, "step": 36110 }, { "grad_norm": 0.3401118814945221, "learning_rate": 7.578008670090127e-05, "loss": 0.0347, "step": 36120 }, { "grad_norm": 0.37066832184791565, "learning_rate": 7.576591789307272e-05, "loss": 0.0355, "step": 36130 }, { "grad_norm": 0.2494097650051117, "learning_rate": 7.575174626752216e-05, "loss": 0.0317, "step": 36140 }, { "grad_norm": 0.40151655673980713, "learning_rate": 7.573757182579934e-05, "loss": 0.0368, "step": 36150 }, { "grad_norm": 0.5601347088813782, "learning_rate": 7.572339456945435e-05, "loss": 0.0321, "step": 36160 }, { "grad_norm": 0.38737058639526367, "learning_rate": 7.570921450003762e-05, "loss": 0.0392, "step": 36170 }, { "grad_norm": 0.4302552342414856, "learning_rate": 7.569503161909983e-05, "loss": 0.0339, "step": 36180 }, { "grad_norm": 0.2851134240627289, "learning_rate": 7.568084592819202e-05, "loss": 0.0376, "step": 36190 }, { "grad_norm": 0.4524170458316803, "learning_rate": 7.566665742886551e-05, "loss": 0.0301, "step": 36200 }, { "grad_norm": 0.6998880505561829, "learning_rate": 7.565246612267194e-05, "loss": 0.0552, "step": 36210 }, { "grad_norm": 0.3580162525177002, "learning_rate": 7.563827201116325e-05, "loss": 0.0329, "step": 36220 }, { "grad_norm": 0.43848398327827454, "learning_rate": 7.562407509589164e-05, "loss": 0.0362, "step": 36230 }, { "grad_norm": 0.42695239186286926, "learning_rate": 7.560987537840973e-05, "loss": 0.0314, "step": 36240 }, { "grad_norm": 0.3215920925140381, "learning_rate": 7.559567286027036e-05, "loss": 0.0333, "step": 36250 }, { "grad_norm": 0.27847927808761597, "learning_rate": 7.558146754302668e-05, "loss": 0.0334, "step": 36260 }, { "grad_norm": 0.351672887802124, "learning_rate": 7.556725942823217e-05, "loss": 0.0283, "step": 36270 }, { "grad_norm": 0.3139965534210205, "learning_rate": 7.555304851744061e-05, "loss": 0.0439, "step": 36280 }, { "grad_norm": 0.30281519889831543, "learning_rate": 7.553883481220608e-05, "loss": 0.0429, "step": 36290 }, { "grad_norm": 0.29692885279655457, "learning_rate": 7.552461831408298e-05, "loss": 0.0262, "step": 36300 }, { "grad_norm": 0.3637649714946747, "learning_rate": 7.551039902462599e-05, "loss": 0.0364, "step": 36310 }, { "grad_norm": 0.419820636510849, "learning_rate": 7.549617694539011e-05, "loss": 0.0299, "step": 36320 }, { "grad_norm": 0.3148355185985565, "learning_rate": 7.548195207793066e-05, "loss": 0.0337, "step": 36330 }, { "grad_norm": 0.35584619641304016, "learning_rate": 7.546772442380323e-05, "loss": 0.0288, "step": 36340 }, { "grad_norm": 0.26932039856910706, "learning_rate": 7.545349398456376e-05, "loss": 0.0361, "step": 36350 }, { "grad_norm": 1.3854659795761108, "learning_rate": 7.543926076176845e-05, "loss": 0.0299, "step": 36360 }, { "grad_norm": 0.2149466872215271, "learning_rate": 7.542502475697385e-05, "loss": 0.0283, "step": 36370 }, { "grad_norm": 0.2709490954875946, "learning_rate": 7.541078597173675e-05, "loss": 0.0345, "step": 36380 }, { "grad_norm": 0.8735479116439819, "learning_rate": 7.539654440761431e-05, "loss": 0.0454, "step": 36390 }, { "grad_norm": 0.22253476083278656, "learning_rate": 7.538230006616395e-05, "loss": 0.0356, "step": 36400 }, { "grad_norm": 0.31320762634277344, "learning_rate": 7.536805294894342e-05, "loss": 0.0524, "step": 36410 }, { "grad_norm": 0.33595260977745056, "learning_rate": 7.535380305751076e-05, "loss": 0.0404, "step": 36420 }, { "grad_norm": 0.27438193559646606, "learning_rate": 7.533955039342431e-05, "loss": 0.0382, "step": 36430 }, { "grad_norm": 0.4214576780796051, "learning_rate": 7.532529495824274e-05, "loss": 0.0409, "step": 36440 }, { "grad_norm": 0.25710245966911316, "learning_rate": 7.531103675352497e-05, "loss": 0.0348, "step": 36450 }, { "grad_norm": 0.29325297474861145, "learning_rate": 7.529677578083029e-05, "loss": 0.0329, "step": 36460 }, { "grad_norm": 0.35293787717819214, "learning_rate": 7.528251204171823e-05, "loss": 0.03, "step": 36470 }, { "grad_norm": 0.25030267238616943, "learning_rate": 7.526824553774867e-05, "loss": 0.0379, "step": 36480 }, { "grad_norm": 0.7599924802780151, "learning_rate": 7.525397627048178e-05, "loss": 0.0384, "step": 36490 }, { "grad_norm": 0.3161139488220215, "learning_rate": 7.523970424147802e-05, "loss": 0.0314, "step": 36500 }, { "grad_norm": 0.34930285811424255, "learning_rate": 7.522542945229813e-05, "loss": 0.0398, "step": 36510 }, { "grad_norm": 0.37119239568710327, "learning_rate": 7.521115190450321e-05, "loss": 0.0271, "step": 36520 }, { "grad_norm": 0.29290688037872314, "learning_rate": 7.519687159965462e-05, "loss": 0.0421, "step": 36530 }, { "grad_norm": 0.25537458062171936, "learning_rate": 7.518258853931403e-05, "loss": 0.0318, "step": 36540 }, { "grad_norm": 0.28547707200050354, "learning_rate": 7.516830272504342e-05, "loss": 0.0295, "step": 36550 }, { "grad_norm": 0.3027907907962799, "learning_rate": 7.515401415840509e-05, "loss": 0.0419, "step": 36560 }, { "grad_norm": 0.4963127374649048, "learning_rate": 7.513972284096155e-05, "loss": 0.0381, "step": 36570 }, { "grad_norm": 0.40148496627807617, "learning_rate": 7.512542877427574e-05, "loss": 0.0519, "step": 36580 }, { "grad_norm": 0.4333936870098114, "learning_rate": 7.51111319599108e-05, "loss": 0.0401, "step": 36590 }, { "grad_norm": 0.370593786239624, "learning_rate": 7.509683239943024e-05, "loss": 0.0406, "step": 36600 }, { "grad_norm": 0.2833469808101654, "learning_rate": 7.50825300943978e-05, "loss": 0.0296, "step": 36610 }, { "grad_norm": 0.34274670481681824, "learning_rate": 7.50682250463776e-05, "loss": 0.0364, "step": 36620 }, { "grad_norm": 0.3733518719673157, "learning_rate": 7.5053917256934e-05, "loss": 0.0431, "step": 36630 }, { "grad_norm": 0.20040149986743927, "learning_rate": 7.503960672763166e-05, "loss": 0.0352, "step": 36640 }, { "grad_norm": 0.3231887221336365, "learning_rate": 7.502529346003559e-05, "loss": 0.0382, "step": 36650 }, { "grad_norm": 0.27775758504867554, "learning_rate": 7.501097745571106e-05, "loss": 0.0286, "step": 36660 }, { "grad_norm": 0.44628000259399414, "learning_rate": 7.499665871622361e-05, "loss": 0.033, "step": 36670 }, { "grad_norm": 0.3053397238254547, "learning_rate": 7.498233724313917e-05, "loss": 0.0349, "step": 36680 }, { "grad_norm": 0.23249131441116333, "learning_rate": 7.496801303802387e-05, "loss": 0.036, "step": 36690 }, { "grad_norm": 0.2490009069442749, "learning_rate": 7.495368610244423e-05, "loss": 0.0281, "step": 36700 }, { "grad_norm": 0.3690761923789978, "learning_rate": 7.493935643796697e-05, "loss": 0.0377, "step": 36710 }, { "grad_norm": 0.28236308693885803, "learning_rate": 7.49250240461592e-05, "loss": 0.0327, "step": 36720 }, { "grad_norm": 0.3888886868953705, "learning_rate": 7.491068892858826e-05, "loss": 0.032, "step": 36730 }, { "grad_norm": 0.38043078780174255, "learning_rate": 7.489635108682184e-05, "loss": 0.0361, "step": 36740 }, { "grad_norm": 0.2982163727283478, "learning_rate": 7.48820105224279e-05, "loss": 0.0302, "step": 36750 }, { "grad_norm": 0.2860713005065918, "learning_rate": 7.48676672369747e-05, "loss": 0.0292, "step": 36760 }, { "grad_norm": 0.3916427195072174, "learning_rate": 7.485332123203079e-05, "loss": 0.0287, "step": 36770 }, { "grad_norm": 0.3065102696418762, "learning_rate": 7.483897250916504e-05, "loss": 0.0436, "step": 36780 }, { "grad_norm": 0.29429948329925537, "learning_rate": 7.48246210699466e-05, "loss": 0.0345, "step": 36790 }, { "grad_norm": 0.3396293818950653, "learning_rate": 7.481026691594492e-05, "loss": 0.0357, "step": 36800 }, { "grad_norm": 0.2124784290790558, "learning_rate": 7.479591004872975e-05, "loss": 0.0253, "step": 36810 }, { "grad_norm": 0.30280691385269165, "learning_rate": 7.478155046987114e-05, "loss": 0.0383, "step": 36820 }, { "grad_norm": 0.5393773317337036, "learning_rate": 7.476718818093944e-05, "loss": 0.0337, "step": 36830 }, { "grad_norm": 0.2565869987010956, "learning_rate": 7.475282318350527e-05, "loss": 0.0339, "step": 36840 }, { "grad_norm": 0.2792752683162689, "learning_rate": 7.473845547913957e-05, "loss": 0.0349, "step": 36850 }, { "grad_norm": 0.25125426054000854, "learning_rate": 7.472408506941357e-05, "loss": 0.0442, "step": 36860 }, { "grad_norm": 0.3675500154495239, "learning_rate": 7.470971195589881e-05, "loss": 0.0367, "step": 36870 }, { "grad_norm": 0.31314972043037415, "learning_rate": 7.46953361401671e-05, "loss": 0.0336, "step": 36880 }, { "grad_norm": 0.2377094328403473, "learning_rate": 7.468095762379055e-05, "loss": 0.0292, "step": 36890 }, { "grad_norm": 0.3778097331523895, "learning_rate": 7.466657640834158e-05, "loss": 0.0311, "step": 36900 }, { "grad_norm": 0.2615266740322113, "learning_rate": 7.46521924953929e-05, "loss": 0.0323, "step": 36910 }, { "grad_norm": 0.36869463324546814, "learning_rate": 7.463780588651752e-05, "loss": 0.0291, "step": 36920 }, { "grad_norm": 0.30129802227020264, "learning_rate": 7.462341658328873e-05, "loss": 0.0328, "step": 36930 }, { "grad_norm": 0.3371216058731079, "learning_rate": 7.460902458728012e-05, "loss": 0.0323, "step": 36940 }, { "grad_norm": 0.43958044052124023, "learning_rate": 7.459462990006558e-05, "loss": 0.0293, "step": 36950 }, { "grad_norm": 0.345219224691391, "learning_rate": 7.458023252321929e-05, "loss": 0.0341, "step": 36960 }, { "grad_norm": 0.4615834653377533, "learning_rate": 7.456583245831573e-05, "loss": 0.0317, "step": 36970 }, { "grad_norm": 0.3514408767223358, "learning_rate": 7.455142970692968e-05, "loss": 0.0462, "step": 36980 }, { "grad_norm": 0.1717945784330368, "learning_rate": 7.453702427063615e-05, "loss": 0.0241, "step": 36990 }, { "grad_norm": 0.6083803176879883, "learning_rate": 7.452261615101057e-05, "loss": 0.0344, "step": 37000 }, { "grad_norm": 0.23601104319095612, "learning_rate": 7.450820534962856e-05, "loss": 0.0319, "step": 37010 }, { "grad_norm": 0.359661728143692, "learning_rate": 7.449379186806604e-05, "loss": 0.0331, "step": 37020 }, { "grad_norm": 0.16223649680614471, "learning_rate": 7.44793757078993e-05, "loss": 0.0274, "step": 37030 }, { "grad_norm": 0.2693748474121094, "learning_rate": 7.44649568707048e-05, "loss": 0.0367, "step": 37040 }, { "grad_norm": 0.2437906116247177, "learning_rate": 7.445053535805942e-05, "loss": 0.0321, "step": 37050 }, { "grad_norm": 0.5912423729896545, "learning_rate": 7.443611117154024e-05, "loss": 0.0341, "step": 37060 }, { "grad_norm": 0.20788517594337463, "learning_rate": 7.442168431272469e-05, "loss": 0.0297, "step": 37070 }, { "grad_norm": 0.8469058871269226, "learning_rate": 7.440725478319044e-05, "loss": 0.0348, "step": 37080 }, { "grad_norm": 0.5814629793167114, "learning_rate": 7.439282258451553e-05, "loss": 0.0307, "step": 37090 }, { "grad_norm": 0.24459479749202728, "learning_rate": 7.43783877182782e-05, "loss": 0.0336, "step": 37100 }, { "grad_norm": 0.2885412573814392, "learning_rate": 7.436395018605707e-05, "loss": 0.0254, "step": 37110 }, { "grad_norm": 0.1718156635761261, "learning_rate": 7.434950998943094e-05, "loss": 0.0309, "step": 37120 }, { "grad_norm": 0.18017342686653137, "learning_rate": 7.433506712997902e-05, "loss": 0.0266, "step": 37130 }, { "grad_norm": 0.3566274642944336, "learning_rate": 7.432062160928077e-05, "loss": 0.0294, "step": 37140 }, { "grad_norm": 0.2279699295759201, "learning_rate": 7.430617342891588e-05, "loss": 0.0319, "step": 37150 }, { "grad_norm": 0.3346247673034668, "learning_rate": 7.429172259046442e-05, "loss": 0.0382, "step": 37160 }, { "grad_norm": 0.3437800705432892, "learning_rate": 7.42772690955067e-05, "loss": 0.0319, "step": 37170 }, { "grad_norm": 0.32960012555122375, "learning_rate": 7.426281294562332e-05, "loss": 0.0483, "step": 37180 }, { "grad_norm": 0.3327553868293762, "learning_rate": 7.424835414239519e-05, "loss": 0.0301, "step": 37190 }, { "grad_norm": 0.33964046835899353, "learning_rate": 7.423389268740352e-05, "loss": 0.0332, "step": 37200 }, { "grad_norm": 0.3402246832847595, "learning_rate": 7.42194285822298e-05, "loss": 0.0411, "step": 37210 }, { "grad_norm": 0.5558066368103027, "learning_rate": 7.420496182845576e-05, "loss": 0.0399, "step": 37220 }, { "grad_norm": 0.286186158657074, "learning_rate": 7.419049242766348e-05, "loss": 0.0322, "step": 37230 }, { "grad_norm": 0.35198357701301575, "learning_rate": 7.417602038143533e-05, "loss": 0.036, "step": 37240 }, { "grad_norm": 0.37486720085144043, "learning_rate": 7.416154569135393e-05, "loss": 0.0375, "step": 37250 }, { "grad_norm": 0.2739946246147156, "learning_rate": 7.414706835900222e-05, "loss": 0.0247, "step": 37260 }, { "grad_norm": 0.29983603954315186, "learning_rate": 7.413258838596345e-05, "loss": 0.0271, "step": 37270 }, { "grad_norm": 0.2892038822174072, "learning_rate": 7.411810577382106e-05, "loss": 0.0324, "step": 37280 }, { "grad_norm": 0.26835113763809204, "learning_rate": 7.410362052415889e-05, "loss": 0.0288, "step": 37290 }, { "grad_norm": 0.250920832157135, "learning_rate": 7.408913263856102e-05, "loss": 0.0311, "step": 37300 }, { "grad_norm": 0.4273272752761841, "learning_rate": 7.407464211861183e-05, "loss": 0.0362, "step": 37310 }, { "grad_norm": 0.2220062017440796, "learning_rate": 7.406014896589597e-05, "loss": 0.0353, "step": 37320 }, { "grad_norm": 0.3913092017173767, "learning_rate": 7.404565318199838e-05, "loss": 0.0352, "step": 37330 }, { "grad_norm": 0.3640958070755005, "learning_rate": 7.403115476850432e-05, "loss": 0.0487, "step": 37340 }, { "grad_norm": 0.29891762137413025, "learning_rate": 7.401665372699932e-05, "loss": 0.0281, "step": 37350 }, { "grad_norm": 0.29725587368011475, "learning_rate": 7.400215005906916e-05, "loss": 0.0385, "step": 37360 }, { "grad_norm": 0.25962135195732117, "learning_rate": 7.398764376629996e-05, "loss": 0.0323, "step": 37370 }, { "grad_norm": 0.21414627134799957, "learning_rate": 7.39731348502781e-05, "loss": 0.0244, "step": 37380 }, { "grad_norm": 0.31117069721221924, "learning_rate": 7.395862331259028e-05, "loss": 0.0342, "step": 37390 }, { "grad_norm": 0.5197910070419312, "learning_rate": 7.394410915482341e-05, "loss": 0.0378, "step": 37400 }, { "grad_norm": 0.5233649611473083, "learning_rate": 7.392959237856476e-05, "loss": 0.0445, "step": 37410 }, { "grad_norm": 0.3334294557571411, "learning_rate": 7.391507298540187e-05, "loss": 0.0297, "step": 37420 }, { "grad_norm": 0.3255740702152252, "learning_rate": 7.390055097692257e-05, "loss": 0.0313, "step": 37430 }, { "grad_norm": 0.2825770080089569, "learning_rate": 7.388602635471494e-05, "loss": 0.0386, "step": 37440 }, { "grad_norm": 0.4098237156867981, "learning_rate": 7.387149912036738e-05, "loss": 0.0363, "step": 37450 }, { "grad_norm": 0.3125038146972656, "learning_rate": 7.385696927546854e-05, "loss": 0.0271, "step": 37460 }, { "grad_norm": 0.43238335847854614, "learning_rate": 7.384243682160743e-05, "loss": 0.0436, "step": 37470 }, { "grad_norm": 0.2785450220108032, "learning_rate": 7.382790176037329e-05, "loss": 0.0281, "step": 37480 }, { "grad_norm": 0.35308778285980225, "learning_rate": 7.381336409335562e-05, "loss": 0.0265, "step": 37490 }, { "grad_norm": 0.3122829794883728, "learning_rate": 7.379882382214426e-05, "loss": 0.0255, "step": 37500 }, { "grad_norm": 0.22983287274837494, "learning_rate": 7.378428094832931e-05, "loss": 0.0355, "step": 37510 }, { "grad_norm": 0.27339816093444824, "learning_rate": 7.376973547350114e-05, "loss": 0.0378, "step": 37520 }, { "grad_norm": 0.3872798681259155, "learning_rate": 7.375518739925046e-05, "loss": 0.0341, "step": 37530 }, { "grad_norm": 0.3144654333591461, "learning_rate": 7.374063672716818e-05, "loss": 0.0292, "step": 37540 }, { "grad_norm": 0.5233893990516663, "learning_rate": 7.372608345884558e-05, "loss": 0.032, "step": 37550 }, { "grad_norm": 0.48472484946250916, "learning_rate": 7.371152759587415e-05, "loss": 0.0341, "step": 37560 }, { "grad_norm": 0.36018455028533936, "learning_rate": 7.369696913984572e-05, "loss": 0.0289, "step": 37570 }, { "grad_norm": 0.44558578729629517, "learning_rate": 7.368240809235237e-05, "loss": 0.0326, "step": 37580 }, { "grad_norm": 0.2552087903022766, "learning_rate": 7.366784445498649e-05, "loss": 0.0301, "step": 37590 }, { "grad_norm": 0.24505078792572021, "learning_rate": 7.36532782293407e-05, "loss": 0.0285, "step": 37600 }, { "grad_norm": 0.24368713796138763, "learning_rate": 7.363870941700797e-05, "loss": 0.0286, "step": 37610 }, { "grad_norm": 0.22264111042022705, "learning_rate": 7.362413801958152e-05, "loss": 0.0329, "step": 37620 }, { "grad_norm": 0.38175395131111145, "learning_rate": 7.360956403865486e-05, "loss": 0.0378, "step": 37630 }, { "grad_norm": 0.3468181788921356, "learning_rate": 7.359498747582177e-05, "loss": 0.0331, "step": 37640 }, { "grad_norm": 0.5599763989448547, "learning_rate": 7.358040833267634e-05, "loss": 0.0306, "step": 37650 }, { "grad_norm": 0.24572809040546417, "learning_rate": 7.356582661081288e-05, "loss": 0.0361, "step": 37660 }, { "grad_norm": 0.2593163847923279, "learning_rate": 7.355124231182607e-05, "loss": 0.0289, "step": 37670 }, { "grad_norm": 0.3839353024959564, "learning_rate": 7.353665543731083e-05, "loss": 0.0333, "step": 37680 }, { "grad_norm": 0.2527346611022949, "learning_rate": 7.352206598886229e-05, "loss": 0.0302, "step": 37690 }, { "grad_norm": 0.1708616018295288, "learning_rate": 7.350747396807601e-05, "loss": 0.0287, "step": 37700 }, { "grad_norm": 0.3885330557823181, "learning_rate": 7.349287937654772e-05, "loss": 0.0256, "step": 37710 }, { "grad_norm": 0.3325341045856476, "learning_rate": 7.347828221587345e-05, "loss": 0.0359, "step": 37720 }, { "grad_norm": 0.23183369636535645, "learning_rate": 7.346368248764955e-05, "loss": 0.0277, "step": 37730 }, { "grad_norm": 0.2263101190328598, "learning_rate": 7.34490801934726e-05, "loss": 0.03, "step": 37740 }, { "grad_norm": 0.3497181236743927, "learning_rate": 7.343447533493947e-05, "loss": 0.0366, "step": 37750 }, { "grad_norm": 0.32005029916763306, "learning_rate": 7.341986791364739e-05, "loss": 0.0446, "step": 37760 }, { "grad_norm": 0.29926392436027527, "learning_rate": 7.340525793119374e-05, "loss": 0.037, "step": 37770 }, { "grad_norm": 0.30859267711639404, "learning_rate": 7.33906453891763e-05, "loss": 0.0333, "step": 37780 }, { "grad_norm": 0.36344531178474426, "learning_rate": 7.337603028919303e-05, "loss": 0.0382, "step": 37790 }, { "grad_norm": 0.2526860535144806, "learning_rate": 7.336141263284225e-05, "loss": 0.0366, "step": 37800 }, { "grad_norm": 0.30347108840942383, "learning_rate": 7.334679242172249e-05, "loss": 0.0305, "step": 37810 }, { "grad_norm": 0.26947158575057983, "learning_rate": 7.33321696574326e-05, "loss": 0.0331, "step": 37820 }, { "grad_norm": 0.2669346034526825, "learning_rate": 7.331754434157174e-05, "loss": 0.0342, "step": 37830 }, { "grad_norm": 0.32648998498916626, "learning_rate": 7.330291647573928e-05, "loss": 0.036, "step": 37840 }, { "grad_norm": 0.2784041166305542, "learning_rate": 7.32882860615349e-05, "loss": 0.0312, "step": 37850 }, { "grad_norm": 0.30619195103645325, "learning_rate": 7.327365310055858e-05, "loss": 0.0312, "step": 37860 }, { "grad_norm": 0.4299352765083313, "learning_rate": 7.325901759441055e-05, "loss": 0.0383, "step": 37870 }, { "grad_norm": 0.48050209879875183, "learning_rate": 7.324437954469132e-05, "loss": 0.0414, "step": 37880 }, { "grad_norm": 0.3001105785369873, "learning_rate": 7.32297389530017e-05, "loss": 0.0393, "step": 37890 }, { "grad_norm": 0.3211295008659363, "learning_rate": 7.321509582094274e-05, "loss": 0.116, "step": 37900 }, { "grad_norm": 0.48902010917663574, "learning_rate": 7.320045015011582e-05, "loss": 0.0349, "step": 37910 }, { "grad_norm": 0.271726131439209, "learning_rate": 7.318580194212253e-05, "loss": 0.0257, "step": 37920 }, { "grad_norm": 0.40607988834381104, "learning_rate": 7.317115119856483e-05, "loss": 0.0297, "step": 37930 }, { "grad_norm": 0.3192613422870636, "learning_rate": 7.315649792104485e-05, "loss": 0.0302, "step": 37940 }, { "grad_norm": 0.3731788396835327, "learning_rate": 7.314184211116507e-05, "loss": 0.0397, "step": 37950 }, { "grad_norm": 0.32114365696907043, "learning_rate": 7.312718377052823e-05, "loss": 0.0365, "step": 37960 }, { "grad_norm": 0.3999779522418976, "learning_rate": 7.311252290073737e-05, "loss": 0.0393, "step": 37970 }, { "grad_norm": 0.257226824760437, "learning_rate": 7.309785950339572e-05, "loss": 0.0377, "step": 37980 }, { "grad_norm": 0.2982597351074219, "learning_rate": 7.308319358010692e-05, "loss": 0.0328, "step": 37990 }, { "grad_norm": 0.34383028745651245, "learning_rate": 7.306852513247474e-05, "loss": 0.0545, "step": 38000 }, { "grad_norm": 0.3576603829860687, "learning_rate": 7.305385416210336e-05, "loss": 0.0451, "step": 38010 }, { "grad_norm": 0.47030994296073914, "learning_rate": 7.303918067059711e-05, "loss": 0.0359, "step": 38020 }, { "grad_norm": 0.33557191491127014, "learning_rate": 7.302450465956075e-05, "loss": 0.0323, "step": 38030 }, { "grad_norm": 0.8883193731307983, "learning_rate": 7.300982613059914e-05, "loss": 0.0276, "step": 38040 }, { "grad_norm": 0.3383873701095581, "learning_rate": 7.299514508531757e-05, "loss": 0.0428, "step": 38050 }, { "grad_norm": 0.6020694375038147, "learning_rate": 7.298046152532148e-05, "loss": 0.0372, "step": 38060 }, { "grad_norm": 0.45070093870162964, "learning_rate": 7.296577545221668e-05, "loss": 0.0266, "step": 38070 }, { "grad_norm": 0.3593887984752655, "learning_rate": 7.29510868676092e-05, "loss": 0.0352, "step": 38080 }, { "grad_norm": 0.18035906553268433, "learning_rate": 7.293639577310538e-05, "loss": 0.0278, "step": 38090 }, { "grad_norm": 0.33389467000961304, "learning_rate": 7.292170217031179e-05, "loss": 0.0368, "step": 38100 }, { "grad_norm": 0.3981846868991852, "learning_rate": 7.290700606083532e-05, "loss": 0.0317, "step": 38110 }, { "grad_norm": 0.3923346698284149, "learning_rate": 7.28923074462831e-05, "loss": 0.0355, "step": 38120 }, { "grad_norm": 0.2711300849914551, "learning_rate": 7.287760632826255e-05, "loss": 0.0404, "step": 38130 }, { "grad_norm": 0.40883669257164, "learning_rate": 7.286290270838139e-05, "loss": 0.0423, "step": 38140 }, { "grad_norm": 0.3650742173194885, "learning_rate": 7.284819658824756e-05, "loss": 0.0303, "step": 38150 }, { "grad_norm": 0.27675724029541016, "learning_rate": 7.283348796946929e-05, "loss": 0.032, "step": 38160 }, { "grad_norm": 0.24120782315731049, "learning_rate": 7.28187768536551e-05, "loss": 0.0343, "step": 38170 }, { "grad_norm": 0.47385647892951965, "learning_rate": 7.280406324241379e-05, "loss": 0.0356, "step": 38180 }, { "grad_norm": 0.25939255952835083, "learning_rate": 7.278934713735441e-05, "loss": 0.0396, "step": 38190 }, { "grad_norm": 0.38372552394866943, "learning_rate": 7.277462854008629e-05, "loss": 0.0306, "step": 38200 }, { "grad_norm": 0.34065160155296326, "learning_rate": 7.275990745221904e-05, "loss": 0.0302, "step": 38210 }, { "grad_norm": 0.3373106122016907, "learning_rate": 7.274518387536252e-05, "loss": 0.0317, "step": 38220 }, { "grad_norm": 0.3906705975532532, "learning_rate": 7.273045781112689e-05, "loss": 0.0369, "step": 38230 }, { "grad_norm": 0.3213169276714325, "learning_rate": 7.271572926112257e-05, "loss": 0.0348, "step": 38240 }, { "grad_norm": 0.24858006834983826, "learning_rate": 7.270099822696024e-05, "loss": 0.0304, "step": 38250 }, { "grad_norm": 0.19296512007713318, "learning_rate": 7.26862647102509e-05, "loss": 0.0259, "step": 38260 }, { "grad_norm": 0.21904119849205017, "learning_rate": 7.267152871260573e-05, "loss": 0.0397, "step": 38270 }, { "grad_norm": 0.3768961727619171, "learning_rate": 7.265679023563629e-05, "loss": 0.0305, "step": 38280 }, { "grad_norm": 0.3894908130168915, "learning_rate": 7.264204928095433e-05, "loss": 0.0344, "step": 38290 }, { "grad_norm": 0.33240699768066406, "learning_rate": 7.262730585017188e-05, "loss": 0.0376, "step": 38300 }, { "grad_norm": 0.3561054766178131, "learning_rate": 7.261255994490131e-05, "loss": 0.0322, "step": 38310 }, { "grad_norm": 0.27908995747566223, "learning_rate": 7.259781156675517e-05, "loss": 0.0467, "step": 38320 }, { "grad_norm": 0.3600788414478302, "learning_rate": 7.258306071734633e-05, "loss": 0.0293, "step": 38330 }, { "grad_norm": 0.3148813843727112, "learning_rate": 7.256830739828792e-05, "loss": 0.0345, "step": 38340 }, { "grad_norm": 0.23786500096321106, "learning_rate": 7.255355161119336e-05, "loss": 0.0349, "step": 38350 }, { "grad_norm": 0.302850604057312, "learning_rate": 7.253879335767628e-05, "loss": 0.0319, "step": 38360 }, { "grad_norm": 0.9044826030731201, "learning_rate": 7.252403263935067e-05, "loss": 0.0368, "step": 38370 }, { "grad_norm": 0.4381861686706543, "learning_rate": 7.25092694578307e-05, "loss": 0.037, "step": 38380 }, { "grad_norm": 0.8137790560722351, "learning_rate": 7.249450381473087e-05, "loss": 0.0323, "step": 38390 }, { "grad_norm": 0.32157060503959656, "learning_rate": 7.247973571166593e-05, "loss": 0.0395, "step": 38400 }, { "grad_norm": 0.3502677083015442, "learning_rate": 7.246496515025089e-05, "loss": 0.0288, "step": 38410 }, { "grad_norm": 1.3341619968414307, "learning_rate": 7.245019213210101e-05, "loss": 0.0389, "step": 38420 }, { "grad_norm": 0.23606181144714355, "learning_rate": 7.24354166588319e-05, "loss": 0.036, "step": 38430 }, { "grad_norm": 0.30899903178215027, "learning_rate": 7.242063873205935e-05, "loss": 0.0337, "step": 38440 }, { "grad_norm": 0.3543858826160431, "learning_rate": 7.240585835339946e-05, "loss": 0.0393, "step": 38450 }, { "grad_norm": 0.3360940217971802, "learning_rate": 7.239107552446857e-05, "loss": 0.0353, "step": 38460 }, { "grad_norm": 0.4012167155742645, "learning_rate": 7.237629024688334e-05, "loss": 0.0345, "step": 38470 }, { "grad_norm": 0.2693166732788086, "learning_rate": 7.236150252226064e-05, "loss": 0.0354, "step": 38480 }, { "grad_norm": 0.2837105691432953, "learning_rate": 7.234671235221765e-05, "loss": 0.04, "step": 38490 }, { "grad_norm": 0.29271209239959717, "learning_rate": 7.233191973837179e-05, "loss": 0.0285, "step": 38500 }, { "grad_norm": 0.32874158024787903, "learning_rate": 7.231712468234078e-05, "loss": 0.0375, "step": 38510 }, { "grad_norm": 0.3081772029399872, "learning_rate": 7.230232718574254e-05, "loss": 0.0282, "step": 38520 }, { "grad_norm": 0.2182351052761078, "learning_rate": 7.228752725019535e-05, "loss": 0.0287, "step": 38530 }, { "grad_norm": 0.3238835632801056, "learning_rate": 7.227272487731769e-05, "loss": 0.0266, "step": 38540 }, { "grad_norm": 0.31678569316864014, "learning_rate": 7.225792006872831e-05, "loss": 0.0263, "step": 38550 }, { "grad_norm": 0.285537451505661, "learning_rate": 7.224311282604628e-05, "loss": 0.0293, "step": 38560 }, { "grad_norm": 0.28657737374305725, "learning_rate": 7.222830315089085e-05, "loss": 0.03, "step": 38570 }, { "grad_norm": 0.29443541169166565, "learning_rate": 7.22134910448816e-05, "loss": 0.0337, "step": 38580 }, { "grad_norm": 0.349058598279953, "learning_rate": 7.219867650963839e-05, "loss": 0.0425, "step": 38590 }, { "grad_norm": 0.6483020186424255, "learning_rate": 7.21838595467813e-05, "loss": 0.0427, "step": 38600 }, { "grad_norm": 0.3558037281036377, "learning_rate": 7.216904015793067e-05, "loss": 0.0374, "step": 38610 }, { "grad_norm": 0.34921252727508545, "learning_rate": 7.215421834470713e-05, "loss": 0.0339, "step": 38620 }, { "grad_norm": 0.3968047797679901, "learning_rate": 7.213939410873158e-05, "loss": 0.0385, "step": 38630 }, { "grad_norm": 0.27033400535583496, "learning_rate": 7.212456745162519e-05, "loss": 0.0291, "step": 38640 }, { "grad_norm": 0.29454129934310913, "learning_rate": 7.210973837500937e-05, "loss": 0.0289, "step": 38650 }, { "grad_norm": 0.2885720133781433, "learning_rate": 7.209490688050578e-05, "loss": 0.0369, "step": 38660 }, { "grad_norm": 0.2672286629676819, "learning_rate": 7.208007296973641e-05, "loss": 0.0284, "step": 38670 }, { "grad_norm": 0.2790466547012329, "learning_rate": 7.206523664432345e-05, "loss": 0.0296, "step": 38680 }, { "grad_norm": 0.201217919588089, "learning_rate": 7.205039790588939e-05, "loss": 0.0228, "step": 38690 }, { "grad_norm": 0.24458937346935272, "learning_rate": 7.203555675605697e-05, "loss": 0.0278, "step": 38700 }, { "grad_norm": 0.23970167338848114, "learning_rate": 7.202071319644917e-05, "loss": 0.0402, "step": 38710 }, { "grad_norm": 0.40681686997413635, "learning_rate": 7.200586722868932e-05, "loss": 0.0299, "step": 38720 }, { "grad_norm": 0.2726300358772278, "learning_rate": 7.199101885440088e-05, "loss": 0.0297, "step": 38730 }, { "grad_norm": 0.2436281144618988, "learning_rate": 7.197616807520771e-05, "loss": 0.0272, "step": 38740 }, { "grad_norm": 0.33033618330955505, "learning_rate": 7.196131489273381e-05, "loss": 0.0327, "step": 38750 }, { "grad_norm": 0.31726619601249695, "learning_rate": 7.194645930860355e-05, "loss": 0.0362, "step": 38760 }, { "grad_norm": 0.3184623420238495, "learning_rate": 7.19316013244415e-05, "loss": 0.0341, "step": 38770 }, { "grad_norm": 0.33263838291168213, "learning_rate": 7.191674094187248e-05, "loss": 0.0303, "step": 38780 }, { "grad_norm": 0.39237937331199646, "learning_rate": 7.190187816252165e-05, "loss": 0.0297, "step": 38790 }, { "grad_norm": 0.3356042206287384, "learning_rate": 7.188701298801435e-05, "loss": 0.029, "step": 38800 }, { "grad_norm": 0.33271661400794983, "learning_rate": 7.18721454199762e-05, "loss": 0.0351, "step": 38810 }, { "grad_norm": 0.2029736191034317, "learning_rate": 7.185727546003312e-05, "loss": 0.0349, "step": 38820 }, { "grad_norm": 0.4016329348087311, "learning_rate": 7.184240310981126e-05, "loss": 0.0335, "step": 38830 }, { "grad_norm": 0.4452700912952423, "learning_rate": 7.182752837093704e-05, "loss": 0.0292, "step": 38840 }, { "grad_norm": 0.23850195109844208, "learning_rate": 7.181265124503711e-05, "loss": 0.0356, "step": 38850 }, { "grad_norm": 0.26191022992134094, "learning_rate": 7.179777173373847e-05, "loss": 0.0236, "step": 38860 }, { "grad_norm": 0.382879376411438, "learning_rate": 7.178288983866826e-05, "loss": 0.0268, "step": 38870 }, { "grad_norm": 0.348708838224411, "learning_rate": 7.176800556145397e-05, "loss": 0.0327, "step": 38880 }, { "grad_norm": 0.2299070656299591, "learning_rate": 7.175311890372334e-05, "loss": 0.0387, "step": 38890 }, { "grad_norm": 0.372045636177063, "learning_rate": 7.17382298671043e-05, "loss": 0.0338, "step": 38900 }, { "grad_norm": 0.6876275539398193, "learning_rate": 7.172333845322515e-05, "loss": 0.0304, "step": 38910 }, { "grad_norm": 0.30030229687690735, "learning_rate": 7.170844466371436e-05, "loss": 0.0355, "step": 38920 }, { "grad_norm": 0.30502188205718994, "learning_rate": 7.16935485002007e-05, "loss": 0.0323, "step": 38930 }, { "grad_norm": 0.24155665934085846, "learning_rate": 7.167864996431319e-05, "loss": 0.028, "step": 38940 }, { "grad_norm": 0.29415714740753174, "learning_rate": 7.166374905768111e-05, "loss": 0.0286, "step": 38950 }, { "grad_norm": 0.3220629096031189, "learning_rate": 7.164884578193401e-05, "loss": 0.0301, "step": 38960 }, { "grad_norm": 0.31733018159866333, "learning_rate": 7.163394013870168e-05, "loss": 0.0309, "step": 38970 }, { "grad_norm": 0.3810935616493225, "learning_rate": 7.161903212961419e-05, "loss": 0.0381, "step": 38980 }, { "grad_norm": 0.23698365688323975, "learning_rate": 7.160412175630183e-05, "loss": 0.0348, "step": 38990 }, { "grad_norm": 0.2811804711818695, "learning_rate": 7.158920902039521e-05, "loss": 0.0342, "step": 39000 }, { "grad_norm": 0.5300208926200867, "learning_rate": 7.157429392352514e-05, "loss": 0.0339, "step": 39010 }, { "grad_norm": 0.3919732868671417, "learning_rate": 7.155937646732275e-05, "loss": 0.0329, "step": 39020 }, { "grad_norm": 0.8333520889282227, "learning_rate": 7.154445665341933e-05, "loss": 0.041, "step": 39030 }, { "grad_norm": 0.2655348777770996, "learning_rate": 7.152953448344654e-05, "loss": 0.033, "step": 39040 }, { "grad_norm": 0.4234931170940399, "learning_rate": 7.151460995903624e-05, "loss": 0.0387, "step": 39050 }, { "grad_norm": 0.2876938283443451, "learning_rate": 7.149968308182052e-05, "loss": 0.0316, "step": 39060 }, { "grad_norm": 0.37322521209716797, "learning_rate": 7.14847538534318e-05, "loss": 0.0383, "step": 39070 }, { "grad_norm": 0.496588796377182, "learning_rate": 7.14698222755027e-05, "loss": 0.0304, "step": 39080 }, { "grad_norm": 0.2348201870918274, "learning_rate": 7.14548883496661e-05, "loss": 0.0271, "step": 39090 }, { "grad_norm": 0.15421020984649658, "learning_rate": 7.143995207755517e-05, "loss": 0.0313, "step": 39100 }, { "grad_norm": 0.22948122024536133, "learning_rate": 7.142501346080333e-05, "loss": 0.037, "step": 39110 }, { "grad_norm": 0.5859676003456116, "learning_rate": 7.141007250104421e-05, "loss": 0.0318, "step": 39120 }, { "grad_norm": 0.28741690516471863, "learning_rate": 7.139512919991176e-05, "loss": 0.0299, "step": 39130 }, { "grad_norm": 0.37074118852615356, "learning_rate": 7.138018355904014e-05, "loss": 0.0311, "step": 39140 }, { "grad_norm": 0.6041899919509888, "learning_rate": 7.13652355800638e-05, "loss": 0.0537, "step": 39150 }, { "grad_norm": 0.2356332391500473, "learning_rate": 7.135028526461743e-05, "loss": 0.0349, "step": 39160 }, { "grad_norm": 0.6095538139343262, "learning_rate": 7.133533261433594e-05, "loss": 0.0318, "step": 39170 }, { "grad_norm": 0.22881799936294556, "learning_rate": 7.132037763085457e-05, "loss": 0.0307, "step": 39180 }, { "grad_norm": 0.35879674553871155, "learning_rate": 7.130542031580875e-05, "loss": 0.0336, "step": 39190 }, { "grad_norm": 0.5044896006584167, "learning_rate": 7.12904606708342e-05, "loss": 0.0296, "step": 39200 }, { "grad_norm": 0.3571154773235321, "learning_rate": 7.127549869756687e-05, "loss": 0.031, "step": 39210 }, { "grad_norm": 0.5169345736503601, "learning_rate": 7.126053439764299e-05, "loss": 0.0279, "step": 39220 }, { "grad_norm": 0.4022354483604431, "learning_rate": 7.124556777269904e-05, "loss": 0.0266, "step": 39230 }, { "grad_norm": 0.30338940024375916, "learning_rate": 7.123059882437174e-05, "loss": 0.0303, "step": 39240 }, { "grad_norm": 0.18050220608711243, "learning_rate": 7.121562755429807e-05, "loss": 0.0237, "step": 39250 }, { "grad_norm": 0.19085216522216797, "learning_rate": 7.120065396411528e-05, "loss": 0.0242, "step": 39260 }, { "grad_norm": 0.22431010007858276, "learning_rate": 7.118567805546084e-05, "loss": 0.0254, "step": 39270 }, { "grad_norm": 0.31779050827026367, "learning_rate": 7.117069982997248e-05, "loss": 0.0268, "step": 39280 }, { "grad_norm": 0.33165243268013, "learning_rate": 7.115571928928825e-05, "loss": 0.0472, "step": 39290 }, { "grad_norm": 1.998639464378357, "learning_rate": 7.114073643504635e-05, "loss": 0.0353, "step": 39300 }, { "grad_norm": 0.27389848232269287, "learning_rate": 7.11257512688853e-05, "loss": 0.033, "step": 39310 }, { "grad_norm": 0.18824177980422974, "learning_rate": 7.111076379244384e-05, "loss": 0.0303, "step": 39320 }, { "grad_norm": 1.379625678062439, "learning_rate": 7.109577400736101e-05, "loss": 0.0415, "step": 39330 }, { "grad_norm": 0.40851542353630066, "learning_rate": 7.108078191527605e-05, "loss": 0.0285, "step": 39340 }, { "grad_norm": 0.3039572834968567, "learning_rate": 7.106578751782847e-05, "loss": 0.0417, "step": 39350 }, { "grad_norm": 0.18900637328624725, "learning_rate": 7.105079081665803e-05, "loss": 0.0371, "step": 39360 }, { "grad_norm": 0.4494205415248871, "learning_rate": 7.103579181340476e-05, "loss": 0.0334, "step": 39370 }, { "grad_norm": 0.3780823349952698, "learning_rate": 7.102079050970893e-05, "loss": 0.0276, "step": 39380 }, { "grad_norm": 0.25272539258003235, "learning_rate": 7.100578690721105e-05, "loss": 0.0266, "step": 39390 }, { "grad_norm": 0.5155367851257324, "learning_rate": 7.09907810075519e-05, "loss": 0.0346, "step": 39400 }, { "grad_norm": 0.37329232692718506, "learning_rate": 7.097577281237249e-05, "loss": 0.0424, "step": 39410 }, { "grad_norm": 0.22724854946136475, "learning_rate": 7.09607623233141e-05, "loss": 0.0318, "step": 39420 }, { "grad_norm": 0.3694271743297577, "learning_rate": 7.094574954201827e-05, "loss": 0.0332, "step": 39430 }, { "grad_norm": 0.5603849291801453, "learning_rate": 7.093073447012675e-05, "loss": 0.0352, "step": 39440 }, { "grad_norm": 0.2812104821205139, "learning_rate": 7.09157171092816e-05, "loss": 0.0296, "step": 39450 }, { "grad_norm": 0.3305143415927887, "learning_rate": 7.090069746112504e-05, "loss": 0.0382, "step": 39460 }, { "grad_norm": 0.395779013633728, "learning_rate": 7.088567552729965e-05, "loss": 0.0312, "step": 39470 }, { "grad_norm": 0.40161919593811035, "learning_rate": 7.087065130944818e-05, "loss": 0.0379, "step": 39480 }, { "grad_norm": 0.36393335461616516, "learning_rate": 7.085562480921366e-05, "loss": 0.0435, "step": 39490 }, { "grad_norm": 0.25961148738861084, "learning_rate": 7.084059602823937e-05, "loss": 0.0398, "step": 39500 }, { "grad_norm": 0.41312506794929504, "learning_rate": 7.082556496816882e-05, "loss": 0.0289, "step": 39510 }, { "grad_norm": 0.28908947110176086, "learning_rate": 7.081053163064582e-05, "loss": 0.0349, "step": 39520 }, { "grad_norm": 0.2873140573501587, "learning_rate": 7.079549601731434e-05, "loss": 0.0397, "step": 39530 }, { "grad_norm": 0.43741974234580994, "learning_rate": 7.07804581298187e-05, "loss": 0.0397, "step": 39540 }, { "grad_norm": 0.5069283246994019, "learning_rate": 7.07654179698034e-05, "loss": 0.0461, "step": 39550 }, { "grad_norm": 0.33881381154060364, "learning_rate": 7.075037553891321e-05, "loss": 0.0384, "step": 39560 }, { "grad_norm": 0.4038686156272888, "learning_rate": 7.073533083879315e-05, "loss": 0.0326, "step": 39570 }, { "grad_norm": 0.2927762567996979, "learning_rate": 7.072028387108849e-05, "loss": 0.0338, "step": 39580 }, { "grad_norm": 0.36372485756874084, "learning_rate": 7.070523463744472e-05, "loss": 0.0445, "step": 39590 }, { "grad_norm": 0.3057897984981537, "learning_rate": 7.069018313950763e-05, "loss": 0.0285, "step": 39600 }, { "grad_norm": 0.43117642402648926, "learning_rate": 7.067512937892322e-05, "loss": 0.0341, "step": 39610 }, { "grad_norm": 0.7609483599662781, "learning_rate": 7.066007335733775e-05, "loss": 0.0414, "step": 39620 }, { "grad_norm": 0.3665832281112671, "learning_rate": 7.064501507639772e-05, "loss": 0.0429, "step": 39630 }, { "grad_norm": 0.27810895442962646, "learning_rate": 7.062995453774987e-05, "loss": 0.0479, "step": 39640 }, { "grad_norm": 0.405435711145401, "learning_rate": 7.061489174304121e-05, "loss": 0.0468, "step": 39650 }, { "grad_norm": 0.20412087440490723, "learning_rate": 7.0599826693919e-05, "loss": 0.0336, "step": 39660 }, { "grad_norm": 0.2719198167324066, "learning_rate": 7.05847593920307e-05, "loss": 0.0284, "step": 39670 }, { "grad_norm": 0.30138877034187317, "learning_rate": 7.056968983902406e-05, "loss": 0.0256, "step": 39680 }, { "grad_norm": 0.30809250473976135, "learning_rate": 7.055461803654706e-05, "loss": 0.0286, "step": 39690 }, { "grad_norm": 0.2182672768831253, "learning_rate": 7.053954398624794e-05, "loss": 0.0255, "step": 39700 }, { "grad_norm": 0.3481593728065491, "learning_rate": 7.052446768977518e-05, "loss": 0.0392, "step": 39710 }, { "grad_norm": 0.29304179549217224, "learning_rate": 7.050938914877748e-05, "loss": 0.0301, "step": 39720 }, { "grad_norm": 0.29983410239219666, "learning_rate": 7.049430836490382e-05, "loss": 0.0418, "step": 39730 }, { "grad_norm": 0.42851194739341736, "learning_rate": 7.047922533980342e-05, "loss": 0.0262, "step": 39740 }, { "grad_norm": 0.28103432059288025, "learning_rate": 7.046414007512571e-05, "loss": 0.0358, "step": 39750 }, { "grad_norm": 0.27171507477760315, "learning_rate": 7.044905257252042e-05, "loss": 0.0346, "step": 39760 }, { "grad_norm": 0.43332475423812866, "learning_rate": 7.04339628336375e-05, "loss": 0.0409, "step": 39770 }, { "grad_norm": 0.46180397272109985, "learning_rate": 7.041887086012711e-05, "loss": 0.0384, "step": 39780 }, { "grad_norm": 0.2243545949459076, "learning_rate": 7.040377665363969e-05, "loss": 0.0303, "step": 39790 }, { "grad_norm": 0.23464606702327728, "learning_rate": 7.038868021582594e-05, "loss": 0.0332, "step": 39800 }, { "grad_norm": 0.33700087666511536, "learning_rate": 7.037358154833679e-05, "loss": 0.0351, "step": 39810 }, { "grad_norm": 0.2055254429578781, "learning_rate": 7.035848065282339e-05, "loss": 0.0347, "step": 39820 }, { "grad_norm": 0.9253800511360168, "learning_rate": 7.034337753093714e-05, "loss": 0.0273, "step": 39830 }, { "grad_norm": 0.2089449018239975, "learning_rate": 7.032827218432972e-05, "loss": 0.0251, "step": 39840 }, { "grad_norm": 1.4763379096984863, "learning_rate": 7.031316461465302e-05, "loss": 0.034, "step": 39850 }, { "grad_norm": 0.3059585392475128, "learning_rate": 7.029805482355915e-05, "loss": 0.031, "step": 39860 }, { "grad_norm": 0.3023515045642853, "learning_rate": 7.028294281270055e-05, "loss": 0.0339, "step": 39870 }, { "grad_norm": 0.18029561638832092, "learning_rate": 7.026782858372982e-05, "loss": 0.0327, "step": 39880 }, { "grad_norm": 0.3316384553909302, "learning_rate": 7.025271213829982e-05, "loss": 0.0316, "step": 39890 }, { "grad_norm": 0.3535819351673126, "learning_rate": 7.023759347806366e-05, "loss": 0.0248, "step": 39900 }, { "grad_norm": 0.3926122784614563, "learning_rate": 7.022247260467472e-05, "loss": 0.0362, "step": 39910 }, { "grad_norm": 0.3440747559070587, "learning_rate": 7.020734951978658e-05, "loss": 0.0337, "step": 39920 }, { "grad_norm": 0.3310454785823822, "learning_rate": 7.019222422505307e-05, "loss": 0.0386, "step": 39930 }, { "grad_norm": 0.19499541819095612, "learning_rate": 7.017709672212827e-05, "loss": 0.0266, "step": 39940 }, { "grad_norm": 0.3886280655860901, "learning_rate": 7.016196701266652e-05, "loss": 0.0343, "step": 39950 }, { "grad_norm": 0.3228731155395508, "learning_rate": 7.014683509832238e-05, "loss": 0.0326, "step": 39960 }, { "grad_norm": 0.29276782274246216, "learning_rate": 7.013170098075063e-05, "loss": 0.0343, "step": 39970 }, { "grad_norm": 0.236771360039711, "learning_rate": 7.011656466160632e-05, "loss": 0.0347, "step": 39980 }, { "grad_norm": 0.26765361428260803, "learning_rate": 7.010142614254475e-05, "loss": 0.0349, "step": 39990 }, { "grad_norm": 0.3717787265777588, "learning_rate": 7.008628542522147e-05, "loss": 0.0418, "step": 40000 }, { "grad_norm": 0.38221484422683716, "learning_rate": 7.007114251129218e-05, "loss": 0.0348, "step": 40010 }, { "grad_norm": 0.3102669417858124, "learning_rate": 7.005599740241293e-05, "loss": 0.0302, "step": 40020 }, { "grad_norm": 0.3992285132408142, "learning_rate": 7.004085010023996e-05, "loss": 0.0349, "step": 40030 }, { "grad_norm": 0.4054008424282074, "learning_rate": 7.002570060642976e-05, "loss": 0.0317, "step": 40040 }, { "grad_norm": 0.23846079409122467, "learning_rate": 7.001054892263903e-05, "loss": 0.031, "step": 40050 }, { "grad_norm": 0.2873106598854065, "learning_rate": 6.999539505052477e-05, "loss": 0.0329, "step": 40060 }, { "grad_norm": 0.3414248824119568, "learning_rate": 6.998023899174418e-05, "loss": 0.0348, "step": 40070 }, { "grad_norm": 0.45708388090133667, "learning_rate": 6.996508074795467e-05, "loss": 0.0386, "step": 40080 }, { "grad_norm": 0.3985428214073181, "learning_rate": 6.994992032081396e-05, "loss": 0.0384, "step": 40090 }, { "grad_norm": 0.28883183002471924, "learning_rate": 6.993475771197995e-05, "loss": 0.0382, "step": 40100 }, { "grad_norm": 0.2655448019504547, "learning_rate": 6.991959292311082e-05, "loss": 0.0321, "step": 40110 }, { "grad_norm": 0.298223078250885, "learning_rate": 6.990442595586495e-05, "loss": 0.0352, "step": 40120 }, { "grad_norm": 0.292116641998291, "learning_rate": 6.988925681190098e-05, "loss": 0.0327, "step": 40130 }, { "grad_norm": 0.4091491997241974, "learning_rate": 6.987408549287778e-05, "loss": 0.0394, "step": 40140 }, { "grad_norm": 0.47011297941207886, "learning_rate": 6.985891200045449e-05, "loss": 0.0406, "step": 40150 }, { "grad_norm": 0.5575920939445496, "learning_rate": 6.984373633629045e-05, "loss": 0.0381, "step": 40160 }, { "grad_norm": 0.42947089672088623, "learning_rate": 6.98285585020452e-05, "loss": 0.0289, "step": 40170 }, { "grad_norm": 0.2552793622016907, "learning_rate": 6.981337849937864e-05, "loss": 0.03, "step": 40180 }, { "grad_norm": 0.42493030428886414, "learning_rate": 6.979819632995078e-05, "loss": 0.0354, "step": 40190 }, { "grad_norm": 0.26263388991355896, "learning_rate": 6.978301199542193e-05, "loss": 0.0281, "step": 40200 }, { "grad_norm": 0.26229479908943176, "learning_rate": 6.976782549745263e-05, "loss": 0.03, "step": 40210 }, { "grad_norm": 0.393291711807251, "learning_rate": 6.975263683770367e-05, "loss": 0.035, "step": 40220 }, { "grad_norm": 0.295198917388916, "learning_rate": 6.973744601783604e-05, "loss": 0.0289, "step": 40230 }, { "grad_norm": 0.6814598441123962, "learning_rate": 6.972225303951097e-05, "loss": 0.041, "step": 40240 }, { "grad_norm": 0.3213421702384949, "learning_rate": 6.970705790438998e-05, "loss": 0.0342, "step": 40250 }, { "grad_norm": 0.44632747769355774, "learning_rate": 6.969186061413477e-05, "loss": 0.038, "step": 40260 }, { "grad_norm": 0.28699934482574463, "learning_rate": 6.967666117040727e-05, "loss": 0.0352, "step": 40270 }, { "grad_norm": 0.30192723870277405, "learning_rate": 6.966145957486972e-05, "loss": 0.026, "step": 40280 }, { "grad_norm": 0.3367542624473572, "learning_rate": 6.964625582918449e-05, "loss": 0.0309, "step": 40290 }, { "grad_norm": 0.3419952690601349, "learning_rate": 6.963104993501425e-05, "loss": 0.0342, "step": 40300 }, { "grad_norm": 0.5039871335029602, "learning_rate": 6.961584189402192e-05, "loss": 0.0284, "step": 40310 }, { "grad_norm": 0.37630176544189453, "learning_rate": 6.96006317078706e-05, "loss": 0.0262, "step": 40320 }, { "grad_norm": 0.2999221384525299, "learning_rate": 6.958541937822367e-05, "loss": 0.037, "step": 40330 }, { "grad_norm": 0.2550029754638672, "learning_rate": 6.957020490674473e-05, "loss": 0.0303, "step": 40340 }, { "grad_norm": 0.2708478569984436, "learning_rate": 6.95549882950976e-05, "loss": 0.0295, "step": 40350 }, { "grad_norm": 0.307951956987381, "learning_rate": 6.953976954494635e-05, "loss": 0.03, "step": 40360 }, { "grad_norm": 0.29391369223594666, "learning_rate": 6.952454865795528e-05, "loss": 0.0288, "step": 40370 }, { "grad_norm": 0.3164633512496948, "learning_rate": 6.950932563578892e-05, "loss": 0.032, "step": 40380 }, { "grad_norm": 0.2636929154396057, "learning_rate": 6.949410048011206e-05, "loss": 0.0239, "step": 40390 }, { "grad_norm": 0.4818274974822998, "learning_rate": 6.947887319258966e-05, "loss": 0.0427, "step": 40400 }, { "grad_norm": 0.23233211040496826, "learning_rate": 6.946364377488696e-05, "loss": 0.0304, "step": 40410 }, { "grad_norm": 0.32652655243873596, "learning_rate": 6.944841222866947e-05, "loss": 0.0307, "step": 40420 }, { "grad_norm": 0.4609036445617676, "learning_rate": 6.943317855560284e-05, "loss": 0.0446, "step": 40430 }, { "grad_norm": 0.2850201427936554, "learning_rate": 6.941794275735302e-05, "loss": 0.0364, "step": 40440 }, { "grad_norm": 0.3512972295284271, "learning_rate": 6.94027048355862e-05, "loss": 0.0303, "step": 40450 }, { "grad_norm": 0.34391483664512634, "learning_rate": 6.938746479196871e-05, "loss": 0.0296, "step": 40460 }, { "grad_norm": 0.21594014763832092, "learning_rate": 6.937222262816724e-05, "loss": 0.0256, "step": 40470 }, { "grad_norm": 0.2308375984430313, "learning_rate": 6.935697834584865e-05, "loss": 0.0269, "step": 40480 }, { "grad_norm": 0.2410380244255066, "learning_rate": 6.934173194667998e-05, "loss": 0.0288, "step": 40490 }, { "grad_norm": 0.3751291036605835, "learning_rate": 6.93264834323286e-05, "loss": 0.0286, "step": 40500 }, { "grad_norm": 0.22295387089252472, "learning_rate": 6.931123280446206e-05, "loss": 0.0268, "step": 40510 }, { "grad_norm": 0.30297350883483887, "learning_rate": 6.929598006474811e-05, "loss": 0.0297, "step": 40520 }, { "grad_norm": 0.32655850052833557, "learning_rate": 6.92807252148548e-05, "loss": 0.0276, "step": 40530 }, { "grad_norm": 0.26383572816848755, "learning_rate": 6.926546825645037e-05, "loss": 0.0271, "step": 40540 }, { "grad_norm": 0.34905797243118286, "learning_rate": 6.92502091912033e-05, "loss": 0.0313, "step": 40550 }, { "grad_norm": 0.32924726605415344, "learning_rate": 6.923494802078229e-05, "loss": 0.0208, "step": 40560 }, { "grad_norm": 0.3533030152320862, "learning_rate": 6.921968474685628e-05, "loss": 0.0366, "step": 40570 }, { "grad_norm": 0.3070703148841858, "learning_rate": 6.920441937109445e-05, "loss": 0.0295, "step": 40580 }, { "grad_norm": 0.39892080426216125, "learning_rate": 6.91891518951662e-05, "loss": 0.0275, "step": 40590 }, { "grad_norm": 0.18816770613193512, "learning_rate": 6.917388232074114e-05, "loss": 0.0313, "step": 40600 }, { "grad_norm": 0.34587743878364563, "learning_rate": 6.915861064948914e-05, "loss": 0.0323, "step": 40610 }, { "grad_norm": 0.4638550579547882, "learning_rate": 6.914333688308025e-05, "loss": 0.0294, "step": 40620 }, { "grad_norm": 0.3569824993610382, "learning_rate": 6.912806102318487e-05, "loss": 0.0344, "step": 40630 }, { "grad_norm": 0.23232817649841309, "learning_rate": 6.911278307147347e-05, "loss": 0.0339, "step": 40640 }, { "grad_norm": 0.34957587718963623, "learning_rate": 6.909750302961684e-05, "loss": 0.0292, "step": 40650 }, { "grad_norm": 0.25166311860084534, "learning_rate": 6.9082220899286e-05, "loss": 0.0322, "step": 40660 }, { "grad_norm": 0.424384742975235, "learning_rate": 6.906693668215219e-05, "loss": 0.0375, "step": 40670 }, { "grad_norm": 0.284853994846344, "learning_rate": 6.905165037988683e-05, "loss": 0.0276, "step": 40680 }, { "grad_norm": 0.24521824717521667, "learning_rate": 6.903636199416164e-05, "loss": 0.0331, "step": 40690 }, { "grad_norm": 0.2816130816936493, "learning_rate": 6.902107152664851e-05, "loss": 0.0231, "step": 40700 }, { "grad_norm": 0.32047516107559204, "learning_rate": 6.90057789790196e-05, "loss": 0.0303, "step": 40710 }, { "grad_norm": 0.3933655023574829, "learning_rate": 6.899048435294728e-05, "loss": 0.0308, "step": 40720 }, { "grad_norm": 0.33906176686286926, "learning_rate": 6.897518765010415e-05, "loss": 0.0355, "step": 40730 }, { "grad_norm": 0.31725823879241943, "learning_rate": 6.895988887216302e-05, "loss": 0.0314, "step": 40740 }, { "grad_norm": 0.3032546937465668, "learning_rate": 6.894458802079694e-05, "loss": 0.0405, "step": 40750 }, { "grad_norm": 0.24431703984737396, "learning_rate": 6.892928509767922e-05, "loss": 0.0295, "step": 40760 }, { "grad_norm": 0.23852889239788055, "learning_rate": 6.891398010448333e-05, "loss": 0.0338, "step": 40770 }, { "grad_norm": 0.3513862192630768, "learning_rate": 6.8898673042883e-05, "loss": 0.0333, "step": 40780 }, { "grad_norm": 0.2665936052799225, "learning_rate": 6.888336391455222e-05, "loss": 0.0346, "step": 40790 }, { "grad_norm": 0.22643914818763733, "learning_rate": 6.886805272116513e-05, "loss": 0.027, "step": 40800 }, { "grad_norm": 0.3095985949039459, "learning_rate": 6.885273946439617e-05, "loss": 0.0469, "step": 40810 }, { "grad_norm": 0.365843802690506, "learning_rate": 6.883742414591998e-05, "loss": 0.0298, "step": 40820 }, { "grad_norm": 0.3320804238319397, "learning_rate": 6.88221067674114e-05, "loss": 0.0317, "step": 40830 }, { "grad_norm": 0.2560215890407562, "learning_rate": 6.88067873305455e-05, "loss": 0.0379, "step": 40840 }, { "grad_norm": 0.39614903926849365, "learning_rate": 6.879146583699765e-05, "loss": 0.0241, "step": 40850 }, { "grad_norm": 0.4103647768497467, "learning_rate": 6.877614228844334e-05, "loss": 0.0305, "step": 40860 }, { "grad_norm": 0.3045719563961029, "learning_rate": 6.876081668655832e-05, "loss": 0.0316, "step": 40870 }, { "grad_norm": 0.24714785814285278, "learning_rate": 6.874548903301861e-05, "loss": 0.0309, "step": 40880 }, { "grad_norm": 0.20194417238235474, "learning_rate": 6.873015932950039e-05, "loss": 0.0254, "step": 40890 }, { "grad_norm": 0.27663806080818176, "learning_rate": 6.871482757768012e-05, "loss": 0.0322, "step": 40900 }, { "grad_norm": 0.3124968409538269, "learning_rate": 6.869949377923442e-05, "loss": 0.0316, "step": 40910 }, { "grad_norm": 0.6628016233444214, "learning_rate": 6.868415793584022e-05, "loss": 0.0249, "step": 40920 }, { "grad_norm": 0.42629802227020264, "learning_rate": 6.86688200491746e-05, "loss": 0.0315, "step": 40930 }, { "grad_norm": 0.23101958632469177, "learning_rate": 6.865348012091488e-05, "loss": 0.0315, "step": 40940 }, { "grad_norm": 0.2511853873729706, "learning_rate": 6.86381381527386e-05, "loss": 0.0404, "step": 40950 }, { "grad_norm": 0.551016092300415, "learning_rate": 6.862279414632359e-05, "loss": 0.0337, "step": 40960 }, { "grad_norm": 0.32724839448928833, "learning_rate": 6.860744810334778e-05, "loss": 0.0318, "step": 40970 }, { "grad_norm": 0.5565049052238464, "learning_rate": 6.859210002548943e-05, "loss": 0.0295, "step": 40980 }, { "grad_norm": 0.7485968470573425, "learning_rate": 6.8576749914427e-05, "loss": 0.0274, "step": 40990 }, { "grad_norm": 0.32863953709602356, "learning_rate": 6.85613977718391e-05, "loss": 0.0325, "step": 41000 }, { "grad_norm": 0.27512478828430176, "learning_rate": 6.854604359940464e-05, "loss": 0.0318, "step": 41010 }, { "grad_norm": 0.25537875294685364, "learning_rate": 6.853068739880276e-05, "loss": 0.0275, "step": 41020 }, { "grad_norm": 0.5501304864883423, "learning_rate": 6.851532917171276e-05, "loss": 0.0469, "step": 41030 }, { "grad_norm": 0.3159637749195099, "learning_rate": 6.849996891981418e-05, "loss": 0.0365, "step": 41040 }, { "grad_norm": 0.31024178862571716, "learning_rate": 6.848460664478684e-05, "loss": 0.0398, "step": 41050 }, { "grad_norm": 0.3815573453903198, "learning_rate": 6.846924234831069e-05, "loss": 0.0323, "step": 41060 }, { "grad_norm": 0.4422680139541626, "learning_rate": 6.845387603206597e-05, "loss": 0.0336, "step": 41070 }, { "grad_norm": 0.3532586097717285, "learning_rate": 6.843850769773312e-05, "loss": 0.031, "step": 41080 }, { "grad_norm": 0.2839930057525635, "learning_rate": 6.842313734699278e-05, "loss": 0.0364, "step": 41090 }, { "grad_norm": 0.3294929563999176, "learning_rate": 6.840776498152584e-05, "loss": 0.0293, "step": 41100 }, { "grad_norm": 0.2022247016429901, "learning_rate": 6.83923906030134e-05, "loss": 0.0238, "step": 41110 }, { "grad_norm": 0.3940066397190094, "learning_rate": 6.837701421313677e-05, "loss": 0.036, "step": 41120 }, { "grad_norm": 0.24226868152618408, "learning_rate": 6.83616358135775e-05, "loss": 0.0284, "step": 41130 }, { "grad_norm": 0.43272900581359863, "learning_rate": 6.834625540601734e-05, "loss": 0.0374, "step": 41140 }, { "grad_norm": 0.26667648553848267, "learning_rate": 6.833087299213829e-05, "loss": 0.027, "step": 41150 }, { "grad_norm": 0.21283303201198578, "learning_rate": 6.83154885736225e-05, "loss": 0.0373, "step": 41160 }, { "grad_norm": 0.36069217324256897, "learning_rate": 6.830010215215244e-05, "loss": 0.031, "step": 41170 }, { "grad_norm": 0.2620772421360016, "learning_rate": 6.828471372941072e-05, "loss": 0.0242, "step": 41180 }, { "grad_norm": 0.3349637985229492, "learning_rate": 6.826932330708018e-05, "loss": 0.0366, "step": 41190 }, { "grad_norm": 0.32090628147125244, "learning_rate": 6.825393088684393e-05, "loss": 0.0351, "step": 41200 }, { "grad_norm": 0.24660368263721466, "learning_rate": 6.823853647038524e-05, "loss": 0.0277, "step": 41210 }, { "grad_norm": 0.30662357807159424, "learning_rate": 6.822314005938762e-05, "loss": 0.032, "step": 41220 }, { "grad_norm": 0.2445564717054367, "learning_rate": 6.820774165553479e-05, "loss": 0.0386, "step": 41230 }, { "grad_norm": 0.2353477030992508, "learning_rate": 6.819234126051072e-05, "loss": 0.026, "step": 41240 }, { "grad_norm": 0.2124718874692917, "learning_rate": 6.817693887599956e-05, "loss": 0.0328, "step": 41250 }, { "grad_norm": 0.2911224663257599, "learning_rate": 6.816153450368568e-05, "loss": 0.0391, "step": 41260 }, { "grad_norm": 0.3103621304035187, "learning_rate": 6.814612814525372e-05, "loss": 0.0381, "step": 41270 }, { "grad_norm": 0.2925177812576294, "learning_rate": 6.813071980238846e-05, "loss": 0.0295, "step": 41280 }, { "grad_norm": 0.32026365399360657, "learning_rate": 6.811530947677492e-05, "loss": 0.0263, "step": 41290 }, { "grad_norm": 0.1959509551525116, "learning_rate": 6.809989717009839e-05, "loss": 0.0279, "step": 41300 }, { "grad_norm": 0.31721749901771545, "learning_rate": 6.808448288404431e-05, "loss": 0.0307, "step": 41310 }, { "grad_norm": 0.23071660101413727, "learning_rate": 6.806906662029838e-05, "loss": 0.0268, "step": 41320 }, { "grad_norm": 0.3637261390686035, "learning_rate": 6.805364838054647e-05, "loss": 0.0276, "step": 41330 }, { "grad_norm": 0.25703004002571106, "learning_rate": 6.803822816647474e-05, "loss": 0.0272, "step": 41340 }, { "grad_norm": 0.3863103985786438, "learning_rate": 6.802280597976949e-05, "loss": 0.0313, "step": 41350 }, { "grad_norm": 0.45901161432266235, "learning_rate": 6.800738182211727e-05, "loss": 0.0256, "step": 41360 }, { "grad_norm": 0.3158973157405853, "learning_rate": 6.799195569520485e-05, "loss": 0.031, "step": 41370 }, { "grad_norm": 0.5530687570571899, "learning_rate": 6.79765276007192e-05, "loss": 0.028, "step": 41380 }, { "grad_norm": 0.205453559756279, "learning_rate": 6.796109754034753e-05, "loss": 0.0244, "step": 41390 }, { "grad_norm": 0.31932777166366577, "learning_rate": 6.794566551577724e-05, "loss": 0.0234, "step": 41400 }, { "grad_norm": 0.34365811944007874, "learning_rate": 6.793023152869594e-05, "loss": 0.0272, "step": 41410 }, { "grad_norm": 0.22808220982551575, "learning_rate": 6.791479558079147e-05, "loss": 0.0254, "step": 41420 }, { "grad_norm": 0.46241408586502075, "learning_rate": 6.789935767375191e-05, "loss": 0.0308, "step": 41430 }, { "grad_norm": 0.2856612205505371, "learning_rate": 6.78839178092655e-05, "loss": 0.0208, "step": 41440 }, { "grad_norm": 0.26893171668052673, "learning_rate": 6.786847598902072e-05, "loss": 0.0262, "step": 41450 }, { "grad_norm": 0.7945329546928406, "learning_rate": 6.785303221470627e-05, "loss": 0.0322, "step": 41460 }, { "grad_norm": 0.3135852515697479, "learning_rate": 6.783758648801108e-05, "loss": 0.0321, "step": 41470 }, { "grad_norm": 0.5032624006271362, "learning_rate": 6.782213881062422e-05, "loss": 0.0311, "step": 41480 }, { "grad_norm": 0.31699520349502563, "learning_rate": 6.780668918423508e-05, "loss": 0.0317, "step": 41490 }, { "grad_norm": 0.40283238887786865, "learning_rate": 6.779123761053317e-05, "loss": 0.0265, "step": 41500 }, { "grad_norm": 0.3194638788700104, "learning_rate": 6.777578409120826e-05, "loss": 0.0302, "step": 41510 }, { "grad_norm": 0.36215338110923767, "learning_rate": 6.776032862795033e-05, "loss": 0.0379, "step": 41520 }, { "grad_norm": 0.42067649960517883, "learning_rate": 6.774487122244956e-05, "loss": 0.0334, "step": 41530 }, { "grad_norm": 0.6044965386390686, "learning_rate": 6.772941187639637e-05, "loss": 0.023, "step": 41540 }, { "grad_norm": 0.26941055059432983, "learning_rate": 6.771395059148134e-05, "loss": 0.0338, "step": 41550 }, { "grad_norm": 0.4874454140663147, "learning_rate": 6.76984873693953e-05, "loss": 0.0329, "step": 41560 }, { "grad_norm": 0.3889409899711609, "learning_rate": 6.768302221182931e-05, "loss": 0.0486, "step": 41570 }, { "grad_norm": 0.28968289494514465, "learning_rate": 6.766755512047457e-05, "loss": 0.0377, "step": 41580 }, { "grad_norm": 0.22032202780246735, "learning_rate": 6.765208609702259e-05, "loss": 0.027, "step": 41590 }, { "grad_norm": 0.3507164716720581, "learning_rate": 6.763661514316499e-05, "loss": 0.0296, "step": 41600 }, { "grad_norm": 0.27055779099464417, "learning_rate": 6.76211422605937e-05, "loss": 0.0304, "step": 41610 }, { "grad_norm": 0.3028360605239868, "learning_rate": 6.760566745100076e-05, "loss": 0.0269, "step": 41620 }, { "grad_norm": 0.3258419334888458, "learning_rate": 6.75901907160785e-05, "loss": 0.0423, "step": 41630 }, { "grad_norm": 0.7390488982200623, "learning_rate": 6.757471205751943e-05, "loss": 0.0283, "step": 41640 }, { "grad_norm": 0.22447915375232697, "learning_rate": 6.75592314770163e-05, "loss": 0.0282, "step": 41650 }, { "grad_norm": 0.4077018201351166, "learning_rate": 6.7543748976262e-05, "loss": 0.0317, "step": 41660 }, { "grad_norm": 0.23081718385219574, "learning_rate": 6.752826455694968e-05, "loss": 0.0253, "step": 41670 }, { "grad_norm": 0.346951961517334, "learning_rate": 6.751277822077271e-05, "loss": 0.0325, "step": 41680 }, { "grad_norm": 0.32768094539642334, "learning_rate": 6.749728996942465e-05, "loss": 0.0359, "step": 41690 }, { "grad_norm": 0.3448559045791626, "learning_rate": 6.748179980459924e-05, "loss": 0.033, "step": 41700 }, { "grad_norm": 0.3063351511955261, "learning_rate": 6.746630772799052e-05, "loss": 0.0347, "step": 41710 }, { "grad_norm": 0.33917123079299927, "learning_rate": 6.745081374129262e-05, "loss": 0.0332, "step": 41720 }, { "grad_norm": 0.35750898718833923, "learning_rate": 6.743531784619998e-05, "loss": 0.0372, "step": 41730 }, { "grad_norm": 0.301902174949646, "learning_rate": 6.741982004440719e-05, "loss": 0.0273, "step": 41740 }, { "grad_norm": 0.26963943243026733, "learning_rate": 6.740432033760907e-05, "loss": 0.0282, "step": 41750 }, { "grad_norm": 0.23680830001831055, "learning_rate": 6.738881872750066e-05, "loss": 0.0311, "step": 41760 }, { "grad_norm": 0.3131257891654968, "learning_rate": 6.737331521577715e-05, "loss": 0.0304, "step": 41770 }, { "grad_norm": 0.3352122902870178, "learning_rate": 6.735780980413403e-05, "loss": 0.0263, "step": 41780 }, { "grad_norm": 0.458911657333374, "learning_rate": 6.734230249426692e-05, "loss": 0.0327, "step": 41790 }, { "grad_norm": 0.2996344566345215, "learning_rate": 6.732679328787168e-05, "loss": 0.0302, "step": 41800 }, { "grad_norm": 0.28395283222198486, "learning_rate": 6.731128218664438e-05, "loss": 0.0235, "step": 41810 }, { "grad_norm": 0.3291434943675995, "learning_rate": 6.729576919228129e-05, "loss": 0.031, "step": 41820 }, { "grad_norm": 0.2258094698190689, "learning_rate": 6.728025430647888e-05, "loss": 0.032, "step": 41830 }, { "grad_norm": 0.23978643119335175, "learning_rate": 6.726473753093383e-05, "loss": 0.0426, "step": 41840 }, { "grad_norm": 0.3504679203033447, "learning_rate": 6.724921886734305e-05, "loss": 0.03, "step": 41850 }, { "grad_norm": 0.25128960609436035, "learning_rate": 6.723369831740363e-05, "loss": 0.0377, "step": 41860 }, { "grad_norm": 0.23063498735427856, "learning_rate": 6.721817588281288e-05, "loss": 0.0324, "step": 41870 }, { "grad_norm": 0.30046549439430237, "learning_rate": 6.720265156526828e-05, "loss": 0.0338, "step": 41880 }, { "grad_norm": 0.23481440544128418, "learning_rate": 6.718712536646758e-05, "loss": 0.0273, "step": 41890 }, { "grad_norm": 0.3243497908115387, "learning_rate": 6.71715972881087e-05, "loss": 0.0318, "step": 41900 }, { "grad_norm": 0.29837682843208313, "learning_rate": 6.715606733188974e-05, "loss": 0.0284, "step": 41910 }, { "grad_norm": 0.22574374079704285, "learning_rate": 6.714053549950909e-05, "loss": 0.0288, "step": 41920 }, { "grad_norm": 0.2900630831718445, "learning_rate": 6.71250017926652e-05, "loss": 0.0289, "step": 41930 }, { "grad_norm": 0.3209679126739502, "learning_rate": 6.71094662130569e-05, "loss": 0.0296, "step": 41940 }, { "grad_norm": 0.41886255145072937, "learning_rate": 6.709392876238307e-05, "loss": 0.0372, "step": 41950 }, { "grad_norm": 0.3035832941532135, "learning_rate": 6.707838944234289e-05, "loss": 0.0277, "step": 41960 }, { "grad_norm": 0.3548857271671295, "learning_rate": 6.706284825463574e-05, "loss": 0.0286, "step": 41970 }, { "grad_norm": 0.33012160658836365, "learning_rate": 6.704730520096114e-05, "loss": 0.0307, "step": 41980 }, { "grad_norm": 0.23633305728435516, "learning_rate": 6.703176028301888e-05, "loss": 0.0329, "step": 41990 }, { "grad_norm": 0.38183730840682983, "learning_rate": 6.701621350250892e-05, "loss": 0.0284, "step": 42000 }, { "grad_norm": 0.7976521253585815, "learning_rate": 6.700066486113144e-05, "loss": 0.0316, "step": 42010 }, { "grad_norm": 0.43562746047973633, "learning_rate": 6.69851143605868e-05, "loss": 0.0315, "step": 42020 }, { "grad_norm": 0.33119383454322815, "learning_rate": 6.696956200257561e-05, "loss": 0.0353, "step": 42030 }, { "grad_norm": 0.8667761087417603, "learning_rate": 6.695400778879863e-05, "loss": 0.0403, "step": 42040 }, { "grad_norm": 0.44409695267677307, "learning_rate": 6.693845172095683e-05, "loss": 0.0429, "step": 42050 }, { "grad_norm": 0.2936406433582306, "learning_rate": 6.692289380075142e-05, "loss": 0.0288, "step": 42060 }, { "grad_norm": 0.3816591799259186, "learning_rate": 6.690733402988379e-05, "loss": 0.0313, "step": 42070 }, { "grad_norm": 0.3411063253879547, "learning_rate": 6.689177241005553e-05, "loss": 0.0381, "step": 42080 }, { "grad_norm": 0.25560057163238525, "learning_rate": 6.687620894296844e-05, "loss": 0.0353, "step": 42090 }, { "grad_norm": 0.3374910354614258, "learning_rate": 6.686064363032451e-05, "loss": 0.0344, "step": 42100 }, { "grad_norm": 0.4119403064250946, "learning_rate": 6.684507647382596e-05, "loss": 0.0282, "step": 42110 }, { "grad_norm": 1.0447179079055786, "learning_rate": 6.682950747517513e-05, "loss": 0.041, "step": 42120 }, { "grad_norm": 0.44743627309799194, "learning_rate": 6.68139366360747e-05, "loss": 0.0316, "step": 42130 }, { "grad_norm": 0.3038797974586487, "learning_rate": 6.679836395822744e-05, "loss": 0.0352, "step": 42140 }, { "grad_norm": 0.3487132787704468, "learning_rate": 6.678278944333633e-05, "loss": 0.0333, "step": 42150 }, { "grad_norm": 0.3268249034881592, "learning_rate": 6.676721309310462e-05, "loss": 0.027, "step": 42160 }, { "grad_norm": 0.5195966958999634, "learning_rate": 6.67516349092357e-05, "loss": 0.0313, "step": 42170 }, { "grad_norm": 0.3373323082923889, "learning_rate": 6.673605489343315e-05, "loss": 0.0282, "step": 42180 }, { "grad_norm": 0.32034748792648315, "learning_rate": 6.672047304740081e-05, "loss": 0.036, "step": 42190 }, { "grad_norm": 0.27295881509780884, "learning_rate": 6.670488937284268e-05, "loss": 0.0308, "step": 42200 }, { "grad_norm": 0.2967739999294281, "learning_rate": 6.668930387146294e-05, "loss": 0.0339, "step": 42210 }, { "grad_norm": 0.2782098352909088, "learning_rate": 6.667371654496605e-05, "loss": 0.0302, "step": 42220 }, { "grad_norm": 0.4139430522918701, "learning_rate": 6.665812739505659e-05, "loss": 0.034, "step": 42230 }, { "grad_norm": 0.24632826447486877, "learning_rate": 6.664253642343935e-05, "loss": 0.0349, "step": 42240 }, { "grad_norm": 0.30098140239715576, "learning_rate": 6.662694363181935e-05, "loss": 0.0325, "step": 42250 }, { "grad_norm": 0.35373035073280334, "learning_rate": 6.66113490219018e-05, "loss": 0.0365, "step": 42260 }, { "grad_norm": 0.44110530614852905, "learning_rate": 6.659575259539207e-05, "loss": 0.0366, "step": 42270 }, { "grad_norm": 0.33527693152427673, "learning_rate": 6.65801543539958e-05, "loss": 0.0319, "step": 42280 }, { "grad_norm": 0.2740277647972107, "learning_rate": 6.65645542994188e-05, "loss": 0.0334, "step": 42290 }, { "grad_norm": 0.4830504059791565, "learning_rate": 6.654895243336702e-05, "loss": 0.0288, "step": 42300 }, { "grad_norm": 0.3083871901035309, "learning_rate": 6.65333487575467e-05, "loss": 0.0307, "step": 42310 }, { "grad_norm": 0.30856087803840637, "learning_rate": 6.651774327366422e-05, "loss": 0.0251, "step": 42320 }, { "grad_norm": 0.2918223738670349, "learning_rate": 6.650213598342613e-05, "loss": 0.0248, "step": 42330 }, { "grad_norm": 17.959022521972656, "learning_rate": 6.648652688853931e-05, "loss": 0.0423, "step": 42340 }, { "grad_norm": 0.26454243063926697, "learning_rate": 6.647091599071066e-05, "loss": 0.0451, "step": 42350 }, { "grad_norm": 0.33378610014915466, "learning_rate": 6.645530329164743e-05, "loss": 0.033, "step": 42360 }, { "grad_norm": 0.30567625164985657, "learning_rate": 6.643968879305694e-05, "loss": 0.0418, "step": 42370 }, { "grad_norm": 0.4951317310333252, "learning_rate": 6.642407249664683e-05, "loss": 0.0272, "step": 42380 }, { "grad_norm": 0.22761520743370056, "learning_rate": 6.640845440412483e-05, "loss": 0.0335, "step": 42390 }, { "grad_norm": 0.27549371123313904, "learning_rate": 6.639283451719893e-05, "loss": 0.0336, "step": 42400 }, { "grad_norm": 0.4374396800994873, "learning_rate": 6.637721283757729e-05, "loss": 0.0414, "step": 42410 }, { "grad_norm": 0.3775833547115326, "learning_rate": 6.636158936696828e-05, "loss": 0.0298, "step": 42420 }, { "grad_norm": 0.30935361981391907, "learning_rate": 6.634596410708047e-05, "loss": 0.0349, "step": 42430 }, { "grad_norm": 0.2601487338542938, "learning_rate": 6.633033705962257e-05, "loss": 0.0354, "step": 42440 }, { "grad_norm": 0.39032626152038574, "learning_rate": 6.631470822630359e-05, "loss": 0.0318, "step": 42450 }, { "grad_norm": 0.27846723794937134, "learning_rate": 6.629907760883263e-05, "loss": 0.0319, "step": 42460 }, { "grad_norm": 0.24010668694972992, "learning_rate": 6.628344520891905e-05, "loss": 0.0273, "step": 42470 }, { "grad_norm": 0.48813721537590027, "learning_rate": 6.62678110282724e-05, "loss": 0.0333, "step": 42480 }, { "grad_norm": 0.2796073257923126, "learning_rate": 6.625217506860237e-05, "loss": 0.0365, "step": 42490 }, { "grad_norm": 0.3288590610027313, "learning_rate": 6.623653733161892e-05, "loss": 0.0359, "step": 42500 }, { "grad_norm": 0.36560776829719543, "learning_rate": 6.622089781903217e-05, "loss": 0.0328, "step": 42510 }, { "grad_norm": 0.2314416617155075, "learning_rate": 6.620525653255241e-05, "loss": 0.0305, "step": 42520 }, { "grad_norm": 0.3336690664291382, "learning_rate": 6.618961347389015e-05, "loss": 0.0282, "step": 42530 }, { "grad_norm": 0.37016454339027405, "learning_rate": 6.617396864475613e-05, "loss": 0.0314, "step": 42540 }, { "grad_norm": 0.2499016970396042, "learning_rate": 6.61583220468612e-05, "loss": 0.0242, "step": 42550 }, { "grad_norm": 0.25254595279693604, "learning_rate": 6.614267368191645e-05, "loss": 0.0287, "step": 42560 }, { "grad_norm": 0.5403438210487366, "learning_rate": 6.61270235516332e-05, "loss": 0.0341, "step": 42570 }, { "grad_norm": 0.21767182648181915, "learning_rate": 6.611137165772287e-05, "loss": 0.0309, "step": 42580 }, { "grad_norm": 0.3771848678588867, "learning_rate": 6.60957180018972e-05, "loss": 0.0444, "step": 42590 }, { "grad_norm": 0.3389802575111389, "learning_rate": 6.608006258586797e-05, "loss": 0.0354, "step": 42600 }, { "grad_norm": 0.22345182299613953, "learning_rate": 6.606440541134731e-05, "loss": 0.033, "step": 42610 }, { "grad_norm": 0.29171112179756165, "learning_rate": 6.60487464800474e-05, "loss": 0.0364, "step": 42620 }, { "grad_norm": 0.30335575342178345, "learning_rate": 6.603308579368071e-05, "loss": 0.0329, "step": 42630 }, { "grad_norm": 0.34137842059135437, "learning_rate": 6.601742335395987e-05, "loss": 0.0361, "step": 42640 }, { "grad_norm": 0.2730488181114197, "learning_rate": 6.600175916259769e-05, "loss": 0.0366, "step": 42650 }, { "grad_norm": 0.3546374440193176, "learning_rate": 6.598609322130718e-05, "loss": 0.0496, "step": 42660 }, { "grad_norm": 1.2631996870040894, "learning_rate": 6.597042553180155e-05, "loss": 0.041, "step": 42670 }, { "grad_norm": 0.3510931432247162, "learning_rate": 6.59547560957942e-05, "loss": 0.0338, "step": 42680 }, { "grad_norm": 0.29388827085494995, "learning_rate": 6.59390849149987e-05, "loss": 0.042, "step": 42690 }, { "grad_norm": 0.23245497047901154, "learning_rate": 6.592341199112886e-05, "loss": 0.0297, "step": 42700 }, { "grad_norm": 0.25293803215026855, "learning_rate": 6.590773732589861e-05, "loss": 0.0364, "step": 42710 }, { "grad_norm": 0.4422770142555237, "learning_rate": 6.589206092102213e-05, "loss": 0.0452, "step": 42720 }, { "grad_norm": 0.16363249719142914, "learning_rate": 6.587638277821375e-05, "loss": 0.03, "step": 42730 }, { "grad_norm": 0.30899617075920105, "learning_rate": 6.586070289918803e-05, "loss": 0.027, "step": 42740 }, { "grad_norm": 0.36364877223968506, "learning_rate": 6.584502128565968e-05, "loss": 0.0346, "step": 42750 }, { "grad_norm": 0.47395631670951843, "learning_rate": 6.582933793934363e-05, "loss": 0.0287, "step": 42760 }, { "grad_norm": 0.28143346309661865, "learning_rate": 6.581365286195498e-05, "loss": 0.0239, "step": 42770 }, { "grad_norm": 0.33879488706588745, "learning_rate": 6.579796605520903e-05, "loss": 0.036, "step": 42780 }, { "grad_norm": 0.30838465690612793, "learning_rate": 6.578227752082127e-05, "loss": 0.0372, "step": 42790 }, { "grad_norm": 0.4198318421840668, "learning_rate": 6.576658726050735e-05, "loss": 0.0301, "step": 42800 }, { "grad_norm": 0.22841644287109375, "learning_rate": 6.575089527598316e-05, "loss": 0.0281, "step": 42810 }, { "grad_norm": 0.6909507513046265, "learning_rate": 6.573520156896476e-05, "loss": 0.0303, "step": 42820 }, { "grad_norm": 0.32455673813819885, "learning_rate": 6.571950614116835e-05, "loss": 0.0276, "step": 42830 }, { "grad_norm": 0.20610985159873962, "learning_rate": 6.570380899431039e-05, "loss": 0.0435, "step": 42840 }, { "grad_norm": 0.23158441483974457, "learning_rate": 6.568811013010749e-05, "loss": 0.0266, "step": 42850 }, { "grad_norm": 0.34986215829849243, "learning_rate": 6.567240955027645e-05, "loss": 0.0288, "step": 42860 }, { "grad_norm": 0.3120068609714508, "learning_rate": 6.565670725653427e-05, "loss": 0.0267, "step": 42870 }, { "grad_norm": 0.189449742436409, "learning_rate": 6.564100325059812e-05, "loss": 0.0212, "step": 42880 }, { "grad_norm": 0.3255798816680908, "learning_rate": 6.562529753418538e-05, "loss": 0.0287, "step": 42890 }, { "grad_norm": 0.41670501232147217, "learning_rate": 6.56095901090136e-05, "loss": 0.0297, "step": 42900 }, { "grad_norm": 0.379690557718277, "learning_rate": 6.55938809768005e-05, "loss": 0.0272, "step": 42910 }, { "grad_norm": 0.2625478208065033, "learning_rate": 6.557817013926402e-05, "loss": 0.0215, "step": 42920 }, { "grad_norm": 0.34912389516830444, "learning_rate": 6.556245759812227e-05, "loss": 0.0305, "step": 42930 }, { "grad_norm": 0.4118184745311737, "learning_rate": 6.554674335509357e-05, "loss": 0.0249, "step": 42940 }, { "grad_norm": 0.22581374645233154, "learning_rate": 6.553102741189638e-05, "loss": 0.0219, "step": 42950 }, { "grad_norm": 0.27568042278289795, "learning_rate": 6.55153097702494e-05, "loss": 0.0293, "step": 42960 }, { "grad_norm": 0.33865663409233093, "learning_rate": 6.549959043187146e-05, "loss": 0.0322, "step": 42970 }, { "grad_norm": 0.4168829321861267, "learning_rate": 6.54838693984816e-05, "loss": 0.0339, "step": 42980 }, { "grad_norm": 0.23662105202674866, "learning_rate": 6.546814667179909e-05, "loss": 0.0324, "step": 42990 }, { "grad_norm": 0.23926177620887756, "learning_rate": 6.545242225354328e-05, "loss": 0.0281, "step": 43000 }, { "grad_norm": 0.23808135092258453, "learning_rate": 6.543669614543383e-05, "loss": 0.0266, "step": 43010 }, { "grad_norm": 0.3564523160457611, "learning_rate": 6.542096834919049e-05, "loss": 0.0398, "step": 43020 }, { "grad_norm": 0.23212416470050812, "learning_rate": 6.540523886653324e-05, "loss": 0.0329, "step": 43030 }, { "grad_norm": 0.2928062975406647, "learning_rate": 6.538950769918222e-05, "loss": 0.0329, "step": 43040 }, { "grad_norm": 0.5086935758590698, "learning_rate": 6.537377484885779e-05, "loss": 0.0245, "step": 43050 }, { "grad_norm": 0.29600584506988525, "learning_rate": 6.535804031728045e-05, "loss": 0.0464, "step": 43060 }, { "grad_norm": 0.41021835803985596, "learning_rate": 6.534230410617092e-05, "loss": 0.0372, "step": 43070 }, { "grad_norm": 0.278137743473053, "learning_rate": 6.532656621725007e-05, "loss": 0.0295, "step": 43080 }, { "grad_norm": 1.0619350671768188, "learning_rate": 6.531082665223898e-05, "loss": 0.039, "step": 43090 }, { "grad_norm": 0.36484581232070923, "learning_rate": 6.529508541285889e-05, "loss": 0.0314, "step": 43100 }, { "grad_norm": 0.33359774947166443, "learning_rate": 6.527934250083128e-05, "loss": 0.0378, "step": 43110 }, { "grad_norm": 0.39749637246131897, "learning_rate": 6.526359791787775e-05, "loss": 0.0289, "step": 43120 }, { "grad_norm": 0.3150077164173126, "learning_rate": 6.524785166572009e-05, "loss": 0.0303, "step": 43130 }, { "grad_norm": 0.14511893689632416, "learning_rate": 6.523210374608029e-05, "loss": 0.0325, "step": 43140 }, { "grad_norm": 0.4558276832103729, "learning_rate": 6.521635416068054e-05, "loss": 0.0322, "step": 43150 }, { "grad_norm": 0.3606725037097931, "learning_rate": 6.520060291124317e-05, "loss": 0.037, "step": 43160 }, { "grad_norm": 0.3655315339565277, "learning_rate": 6.518484999949072e-05, "loss": 0.026, "step": 43170 }, { "grad_norm": 0.6036680340766907, "learning_rate": 6.516909542714591e-05, "loss": 0.0329, "step": 43180 }, { "grad_norm": 0.27638471126556396, "learning_rate": 6.515333919593161e-05, "loss": 0.0288, "step": 43190 }, { "grad_norm": 0.18777887523174286, "learning_rate": 6.513758130757094e-05, "loss": 0.0308, "step": 43200 }, { "grad_norm": 0.24530792236328125, "learning_rate": 6.512182176378713e-05, "loss": 0.0238, "step": 43210 }, { "grad_norm": 0.19650126993656158, "learning_rate": 6.510606056630362e-05, "loss": 0.0227, "step": 43220 }, { "grad_norm": 0.20743228495121002, "learning_rate": 6.509029771684405e-05, "loss": 0.0278, "step": 43230 }, { "grad_norm": 0.3390566408634186, "learning_rate": 6.50745332171322e-05, "loss": 0.0278, "step": 43240 }, { "grad_norm": 0.331387996673584, "learning_rate": 6.505876706889207e-05, "loss": 0.0275, "step": 43250 }, { "grad_norm": 0.30412954092025757, "learning_rate": 6.504299927384781e-05, "loss": 0.0321, "step": 43260 }, { "grad_norm": 0.4196111857891083, "learning_rate": 6.502722983372378e-05, "loss": 0.024, "step": 43270 }, { "grad_norm": 0.2790205776691437, "learning_rate": 6.501145875024446e-05, "loss": 0.0253, "step": 43280 }, { "grad_norm": 0.31009653210639954, "learning_rate": 6.499568602513459e-05, "loss": 0.0251, "step": 43290 }, { "grad_norm": 0.34036785364151, "learning_rate": 6.497991166011903e-05, "loss": 0.0299, "step": 43300 }, { "grad_norm": 0.29398679733276367, "learning_rate": 6.49641356569229e-05, "loss": 0.0357, "step": 43310 }, { "grad_norm": 0.35221877694129944, "learning_rate": 6.494835801727135e-05, "loss": 0.0339, "step": 43320 }, { "grad_norm": 0.24943006038665771, "learning_rate": 6.493257874288986e-05, "loss": 0.0309, "step": 43330 }, { "grad_norm": 0.21903808414936066, "learning_rate": 6.491679783550399e-05, "loss": 0.027, "step": 43340 }, { "grad_norm": 0.34331366419792175, "learning_rate": 6.490101529683957e-05, "loss": 0.0303, "step": 43350 }, { "grad_norm": 0.396438866853714, "learning_rate": 6.488523112862248e-05, "loss": 0.0324, "step": 43360 }, { "grad_norm": 0.2742571532726288, "learning_rate": 6.486944533257891e-05, "loss": 0.0343, "step": 43370 }, { "grad_norm": 0.313803493976593, "learning_rate": 6.485365791043517e-05, "loss": 0.039, "step": 43380 }, { "grad_norm": 0.30123719573020935, "learning_rate": 6.483786886391772e-05, "loss": 0.0415, "step": 43390 }, { "grad_norm": 0.2304268181324005, "learning_rate": 6.482207819475323e-05, "loss": 0.022, "step": 43400 }, { "grad_norm": 0.5004900693893433, "learning_rate": 6.480628590466857e-05, "loss": 0.0331, "step": 43410 }, { "grad_norm": 0.2226569503545761, "learning_rate": 6.479049199539074e-05, "loss": 0.0288, "step": 43420 }, { "grad_norm": 0.3098169267177582, "learning_rate": 6.477469646864693e-05, "loss": 0.0285, "step": 43430 }, { "grad_norm": 0.3267117738723755, "learning_rate": 6.475889932616454e-05, "loss": 0.0379, "step": 43440 }, { "grad_norm": 0.3277619481086731, "learning_rate": 6.474310056967111e-05, "loss": 0.0303, "step": 43450 }, { "grad_norm": 0.36737680435180664, "learning_rate": 6.472730020089437e-05, "loss": 0.0291, "step": 43460 }, { "grad_norm": 0.4142972230911255, "learning_rate": 6.471149822156222e-05, "loss": 0.0245, "step": 43470 }, { "grad_norm": 0.2845761179924011, "learning_rate": 6.469569463340275e-05, "loss": 0.0247, "step": 43480 }, { "grad_norm": 0.33743220567703247, "learning_rate": 6.467988943814418e-05, "loss": 0.0317, "step": 43490 }, { "grad_norm": 0.5431020855903625, "learning_rate": 6.4664082637515e-05, "loss": 0.0265, "step": 43500 }, { "grad_norm": 0.3032192587852478, "learning_rate": 6.464827423324377e-05, "loss": 0.032, "step": 43510 }, { "grad_norm": 0.2903501093387604, "learning_rate": 6.463246422705931e-05, "loss": 0.0221, "step": 43520 }, { "grad_norm": 0.39283955097198486, "learning_rate": 6.461665262069055e-05, "loss": 0.0311, "step": 43530 }, { "grad_norm": 0.331792950630188, "learning_rate": 6.460083941586665e-05, "loss": 0.0308, "step": 43540 }, { "grad_norm": 0.342274010181427, "learning_rate": 6.45850246143169e-05, "loss": 0.0272, "step": 43550 }, { "grad_norm": 0.2588483691215515, "learning_rate": 6.456920821777077e-05, "loss": 0.0227, "step": 43560 }, { "grad_norm": 0.37722739577293396, "learning_rate": 6.455339022795796e-05, "loss": 0.0323, "step": 43570 }, { "grad_norm": 0.32448241114616394, "learning_rate": 6.453757064660825e-05, "loss": 0.025, "step": 43580 }, { "grad_norm": 0.5071043372154236, "learning_rate": 6.452174947545169e-05, "loss": 0.0248, "step": 43590 }, { "grad_norm": 0.2719459533691406, "learning_rate": 6.450592671621842e-05, "loss": 0.0264, "step": 43600 }, { "grad_norm": 0.3960312008857727, "learning_rate": 6.449010237063882e-05, "loss": 0.0229, "step": 43610 }, { "grad_norm": 0.2725895941257477, "learning_rate": 6.447427644044343e-05, "loss": 0.0241, "step": 43620 }, { "grad_norm": 0.30976560711860657, "learning_rate": 6.44584489273629e-05, "loss": 0.0316, "step": 43630 }, { "grad_norm": 0.2800668179988861, "learning_rate": 6.444261983312817e-05, "loss": 0.0261, "step": 43640 }, { "grad_norm": 0.2623043358325958, "learning_rate": 6.442678915947023e-05, "loss": 0.0305, "step": 43650 }, { "grad_norm": 0.24077734351158142, "learning_rate": 6.441095690812032e-05, "loss": 0.0246, "step": 43660 }, { "grad_norm": 0.33686333894729614, "learning_rate": 6.439512308080982e-05, "loss": 0.0336, "step": 43670 }, { "grad_norm": 0.7968010306358337, "learning_rate": 6.43792876792703e-05, "loss": 0.03, "step": 43680 }, { "grad_norm": 0.2526887357234955, "learning_rate": 6.43634507052335e-05, "loss": 0.0303, "step": 43690 }, { "grad_norm": 0.3248482644557953, "learning_rate": 6.434761216043135e-05, "loss": 0.0363, "step": 43700 }, { "grad_norm": 0.39977946877479553, "learning_rate": 6.433177204659589e-05, "loss": 0.0385, "step": 43710 }, { "grad_norm": 0.33532828092575073, "learning_rate": 6.431593036545938e-05, "loss": 0.0283, "step": 43720 }, { "grad_norm": 0.4008045792579651, "learning_rate": 6.430008711875426e-05, "loss": 0.0322, "step": 43730 }, { "grad_norm": 0.3628676235675812, "learning_rate": 6.42842423082131e-05, "loss": 0.033, "step": 43740 }, { "grad_norm": 0.236603245139122, "learning_rate": 6.426839593556868e-05, "loss": 0.025, "step": 43750 }, { "grad_norm": 0.25121501088142395, "learning_rate": 6.425254800255392e-05, "loss": 0.0213, "step": 43760 }, { "grad_norm": 0.41124871373176575, "learning_rate": 6.423669851090196e-05, "loss": 0.0254, "step": 43770 }, { "grad_norm": 0.1985686719417572, "learning_rate": 6.422084746234604e-05, "loss": 0.0259, "step": 43780 }, { "grad_norm": 0.3076229393482208, "learning_rate": 6.420499485861965e-05, "loss": 0.0263, "step": 43790 }, { "grad_norm": 0.25977641344070435, "learning_rate": 6.418914070145636e-05, "loss": 0.0257, "step": 43800 }, { "grad_norm": 0.2607707381248474, "learning_rate": 6.417328499258996e-05, "loss": 0.0283, "step": 43810 }, { "grad_norm": 0.30295825004577637, "learning_rate": 6.415742773375446e-05, "loss": 0.0268, "step": 43820 }, { "grad_norm": 0.2173633873462677, "learning_rate": 6.414156892668393e-05, "loss": 0.0365, "step": 43830 }, { "grad_norm": 0.33486858010292053, "learning_rate": 6.412570857311267e-05, "loss": 0.0323, "step": 43840 }, { "grad_norm": 0.4233172535896301, "learning_rate": 6.410984667477518e-05, "loss": 0.0317, "step": 43850 }, { "grad_norm": 0.371088445186615, "learning_rate": 6.409398323340607e-05, "loss": 0.0296, "step": 43860 }, { "grad_norm": 0.5642900466918945, "learning_rate": 6.407811825074012e-05, "loss": 0.0397, "step": 43870 }, { "grad_norm": 0.29655125737190247, "learning_rate": 6.406225172851234e-05, "loss": 0.0292, "step": 43880 }, { "grad_norm": 0.2868921458721161, "learning_rate": 6.404638366845786e-05, "loss": 0.0279, "step": 43890 }, { "grad_norm": 0.3377489447593689, "learning_rate": 6.403051407231196e-05, "loss": 0.0347, "step": 43900 }, { "grad_norm": 0.3866300880908966, "learning_rate": 6.401464294181016e-05, "loss": 0.0294, "step": 43910 }, { "grad_norm": 0.5324925184249878, "learning_rate": 6.399877027868808e-05, "loss": 0.0317, "step": 43920 }, { "grad_norm": 0.33498919010162354, "learning_rate": 6.39828960846815e-05, "loss": 0.03, "step": 43930 }, { "grad_norm": 0.4321739971637726, "learning_rate": 6.396702036152644e-05, "loss": 0.0305, "step": 43940 }, { "grad_norm": 0.4764220714569092, "learning_rate": 6.395114311095905e-05, "loss": 0.0377, "step": 43950 }, { "grad_norm": 1.1615350246429443, "learning_rate": 6.393526433471559e-05, "loss": 0.0487, "step": 43960 }, { "grad_norm": 0.5936533212661743, "learning_rate": 6.391938403453262e-05, "loss": 0.0451, "step": 43970 }, { "grad_norm": 0.3655567765235901, "learning_rate": 6.390350221214671e-05, "loss": 0.038, "step": 43980 }, { "grad_norm": 0.33603355288505554, "learning_rate": 6.38876188692947e-05, "loss": 0.0387, "step": 43990 }, { "grad_norm": 0.3533877432346344, "learning_rate": 6.387173400771358e-05, "loss": 0.03, "step": 44000 }, { "grad_norm": 0.24586287140846252, "learning_rate": 6.385584762914047e-05, "loss": 0.0277, "step": 44010 }, { "grad_norm": 0.34120413661003113, "learning_rate": 6.38399597353127e-05, "loss": 0.0316, "step": 44020 }, { "grad_norm": 0.38570186495780945, "learning_rate": 6.382407032796775e-05, "loss": 0.0363, "step": 44030 }, { "grad_norm": 0.18948045372962952, "learning_rate": 6.380817940884325e-05, "loss": 0.0253, "step": 44040 }, { "grad_norm": 0.32831308245658875, "learning_rate": 6.379228697967702e-05, "loss": 0.0257, "step": 44050 }, { "grad_norm": 0.3166705369949341, "learning_rate": 6.3776393042207e-05, "loss": 0.0279, "step": 44060 }, { "grad_norm": 0.571225643157959, "learning_rate": 6.376049759817137e-05, "loss": 0.0247, "step": 44070 }, { "grad_norm": 0.32004767656326294, "learning_rate": 6.37446006493084e-05, "loss": 0.0258, "step": 44080 }, { "grad_norm": 0.2982292175292969, "learning_rate": 6.372870219735656e-05, "loss": 0.0371, "step": 44090 }, { "grad_norm": 0.35094234347343445, "learning_rate": 6.371280224405452e-05, "loss": 0.0328, "step": 44100 }, { "grad_norm": 0.24212633073329926, "learning_rate": 6.369690079114101e-05, "loss": 0.0227, "step": 44110 }, { "grad_norm": 0.28237614035606384, "learning_rate": 6.368099784035504e-05, "loss": 0.0275, "step": 44120 }, { "grad_norm": 0.34134379029273987, "learning_rate": 6.366509339343572e-05, "loss": 0.0331, "step": 44130 }, { "grad_norm": 0.26965874433517456, "learning_rate": 6.364918745212232e-05, "loss": 0.0305, "step": 44140 }, { "grad_norm": 0.4687215983867645, "learning_rate": 6.363328001815431e-05, "loss": 0.0319, "step": 44150 }, { "grad_norm": 0.26346859335899353, "learning_rate": 6.361737109327128e-05, "loss": 0.031, "step": 44160 }, { "grad_norm": 0.34358951449394226, "learning_rate": 6.360146067921301e-05, "loss": 0.032, "step": 44170 }, { "grad_norm": 0.42018723487854004, "learning_rate": 6.35855487777195e-05, "loss": 0.0341, "step": 44180 }, { "grad_norm": 0.3904542326927185, "learning_rate": 6.356963539053076e-05, "loss": 0.038, "step": 44190 }, { "grad_norm": 0.2583659589290619, "learning_rate": 6.355372051938711e-05, "loss": 0.03, "step": 44200 }, { "grad_norm": 0.42449310421943665, "learning_rate": 6.353780416602894e-05, "loss": 0.0296, "step": 44210 }, { "grad_norm": 0.24084864556789398, "learning_rate": 6.352188633219689e-05, "loss": 0.028, "step": 44220 }, { "grad_norm": 0.30270230770111084, "learning_rate": 6.350596701963166e-05, "loss": 0.0298, "step": 44230 }, { "grad_norm": 0.29315510392189026, "learning_rate": 6.349004623007419e-05, "loss": 0.034, "step": 44240 }, { "grad_norm": 0.3122449219226837, "learning_rate": 6.347412396526555e-05, "loss": 0.031, "step": 44250 }, { "grad_norm": 0.279488742351532, "learning_rate": 6.345820022694696e-05, "loss": 0.029, "step": 44260 }, { "grad_norm": 0.24690523743629456, "learning_rate": 6.344227501685984e-05, "loss": 0.0254, "step": 44270 }, { "grad_norm": 0.3196890950202942, "learning_rate": 6.342634833674572e-05, "loss": 0.0295, "step": 44280 }, { "grad_norm": 0.3847827911376953, "learning_rate": 6.341042018834635e-05, "loss": 0.0314, "step": 44290 }, { "grad_norm": 0.2533314824104309, "learning_rate": 6.339449057340359e-05, "loss": 0.028, "step": 44300 }, { "grad_norm": 0.7598565220832825, "learning_rate": 6.337855949365945e-05, "loss": 0.0278, "step": 44310 }, { "grad_norm": 0.32994964718818665, "learning_rate": 6.336262695085619e-05, "loss": 0.0305, "step": 44320 }, { "grad_norm": 0.34476304054260254, "learning_rate": 6.334669294673612e-05, "loss": 0.0269, "step": 44330 }, { "grad_norm": 0.3298689126968384, "learning_rate": 6.333075748304179e-05, "loss": 0.031, "step": 44340 }, { "grad_norm": 0.3393583595752716, "learning_rate": 6.331482056151585e-05, "loss": 0.0322, "step": 44350 }, { "grad_norm": 0.22402966022491455, "learning_rate": 6.329888218390117e-05, "loss": 0.0258, "step": 44360 }, { "grad_norm": 0.2538594901561737, "learning_rate": 6.328294235194072e-05, "loss": 0.0313, "step": 44370 }, { "grad_norm": 0.37604546546936035, "learning_rate": 6.326700106737765e-05, "loss": 0.0292, "step": 44380 }, { "grad_norm": 0.258883535861969, "learning_rate": 6.325105833195531e-05, "loss": 0.0302, "step": 44390 }, { "grad_norm": 0.38823267817497253, "learning_rate": 6.323511414741715e-05, "loss": 0.0348, "step": 44400 }, { "grad_norm": 1.4319077730178833, "learning_rate": 6.321916851550678e-05, "loss": 0.0471, "step": 44410 }, { "grad_norm": 0.2471563071012497, "learning_rate": 6.320322143796806e-05, "loss": 0.051, "step": 44420 }, { "grad_norm": 0.501146674156189, "learning_rate": 6.318727291654488e-05, "loss": 0.0374, "step": 44430 }, { "grad_norm": 0.5680443048477173, "learning_rate": 6.317132295298134e-05, "loss": 0.0313, "step": 44440 }, { "grad_norm": 0.22109313309192657, "learning_rate": 6.315537154902173e-05, "loss": 0.0334, "step": 44450 }, { "grad_norm": 0.3104986250400543, "learning_rate": 6.313941870641048e-05, "loss": 0.0357, "step": 44460 }, { "grad_norm": 0.29223376512527466, "learning_rate": 6.312346442689214e-05, "loss": 0.0286, "step": 44470 }, { "grad_norm": 0.363901823759079, "learning_rate": 6.310750871221147e-05, "loss": 0.0324, "step": 44480 }, { "grad_norm": 0.3291229009628296, "learning_rate": 6.309155156411335e-05, "loss": 0.0323, "step": 44490 }, { "grad_norm": 0.3656606376171112, "learning_rate": 6.307559298434284e-05, "loss": 0.0278, "step": 44500 }, { "grad_norm": 0.26367977261543274, "learning_rate": 6.305963297464512e-05, "loss": 0.0306, "step": 44510 }, { "grad_norm": 0.4031590223312378, "learning_rate": 6.304367153676561e-05, "loss": 0.0347, "step": 44520 }, { "grad_norm": 0.2891556918621063, "learning_rate": 6.302770867244978e-05, "loss": 0.0322, "step": 44530 }, { "grad_norm": 0.30471229553222656, "learning_rate": 6.301174438344328e-05, "loss": 0.0372, "step": 44540 }, { "grad_norm": 0.29854723811149597, "learning_rate": 6.299577867149202e-05, "loss": 0.0398, "step": 44550 }, { "grad_norm": 0.30570995807647705, "learning_rate": 6.297981153834192e-05, "loss": 0.0328, "step": 44560 }, { "grad_norm": 0.21071527898311615, "learning_rate": 6.296384298573916e-05, "loss": 0.0387, "step": 44570 }, { "grad_norm": 0.45362141728401184, "learning_rate": 6.294787301543001e-05, "loss": 0.051, "step": 44580 }, { "grad_norm": 1.0316716432571411, "learning_rate": 6.293190162916095e-05, "loss": 0.0285, "step": 44590 }, { "grad_norm": 0.2967340350151062, "learning_rate": 6.291592882867855e-05, "loss": 0.024, "step": 44600 }, { "grad_norm": 0.37861859798431396, "learning_rate": 6.28999546157296e-05, "loss": 0.0517, "step": 44610 }, { "grad_norm": 0.3639981746673584, "learning_rate": 6.288397899206102e-05, "loss": 0.0361, "step": 44620 }, { "grad_norm": 0.26823586225509644, "learning_rate": 6.286800195941984e-05, "loss": 0.0324, "step": 44630 }, { "grad_norm": 0.31746962666511536, "learning_rate": 6.285202351955334e-05, "loss": 0.0378, "step": 44640 }, { "grad_norm": 0.26109838485717773, "learning_rate": 6.283604367420887e-05, "loss": 0.031, "step": 44650 }, { "grad_norm": 0.34037283062934875, "learning_rate": 6.282006242513394e-05, "loss": 0.0299, "step": 44660 }, { "grad_norm": 0.2868157625198364, "learning_rate": 6.280407977407628e-05, "loss": 0.0265, "step": 44670 }, { "grad_norm": 0.34376680850982666, "learning_rate": 6.27880957227837e-05, "loss": 0.0339, "step": 44680 }, { "grad_norm": 0.25283119082450867, "learning_rate": 6.27721102730042e-05, "loss": 0.0273, "step": 44690 }, { "grad_norm": 0.2752116024494171, "learning_rate": 6.275612342648592e-05, "loss": 0.0394, "step": 44700 }, { "grad_norm": 0.32489728927612305, "learning_rate": 6.274013518497716e-05, "loss": 0.0287, "step": 44710 }, { "grad_norm": 0.23208075761795044, "learning_rate": 6.272414555022636e-05, "loss": 0.0291, "step": 44720 }, { "grad_norm": 0.3703249990940094, "learning_rate": 6.270815452398215e-05, "loss": 0.0297, "step": 44730 }, { "grad_norm": 0.30374857783317566, "learning_rate": 6.269216210799326e-05, "loss": 0.0373, "step": 44740 }, { "grad_norm": 0.2277410924434662, "learning_rate": 6.26761683040086e-05, "loss": 0.0402, "step": 44750 }, { "grad_norm": 0.287204772233963, "learning_rate": 6.266017311377723e-05, "loss": 0.0316, "step": 44760 }, { "grad_norm": 0.18392322957515717, "learning_rate": 6.264417653904839e-05, "loss": 0.0296, "step": 44770 }, { "grad_norm": 0.4091900587081909, "learning_rate": 6.262817858157139e-05, "loss": 0.0392, "step": 44780 }, { "grad_norm": 0.22437481582164764, "learning_rate": 6.261217924309576e-05, "loss": 0.0369, "step": 44790 }, { "grad_norm": 0.2618394196033478, "learning_rate": 6.259617852537118e-05, "loss": 0.0311, "step": 44800 }, { "grad_norm": 0.4207141399383545, "learning_rate": 6.258017643014747e-05, "loss": 0.0385, "step": 44810 }, { "grad_norm": 0.27490442991256714, "learning_rate": 6.256417295917456e-05, "loss": 0.0343, "step": 44820 }, { "grad_norm": 0.22105838358402252, "learning_rate": 6.254816811420258e-05, "loss": 0.0433, "step": 44830 }, { "grad_norm": 0.3961787819862366, "learning_rate": 6.253216189698183e-05, "loss": 0.0304, "step": 44840 }, { "grad_norm": 0.24693994224071503, "learning_rate": 6.251615430926267e-05, "loss": 0.0368, "step": 44850 }, { "grad_norm": 0.47032782435417175, "learning_rate": 6.25001453527957e-05, "loss": 0.0371, "step": 44860 }, { "grad_norm": 0.342488169670105, "learning_rate": 6.248413502933164e-05, "loss": 0.0317, "step": 44870 }, { "grad_norm": 0.36378470063209534, "learning_rate": 6.246812334062133e-05, "loss": 0.029, "step": 44880 }, { "grad_norm": 0.2977908253669739, "learning_rate": 6.245211028841579e-05, "loss": 0.0347, "step": 44890 }, { "grad_norm": 0.30538174510002136, "learning_rate": 6.24360958744662e-05, "loss": 0.0313, "step": 44900 }, { "grad_norm": 0.36647552251815796, "learning_rate": 6.242008010052387e-05, "loss": 0.0257, "step": 44910 }, { "grad_norm": 0.3114210069179535, "learning_rate": 6.240406296834024e-05, "loss": 0.0327, "step": 44920 }, { "grad_norm": 0.3034570515155792, "learning_rate": 6.238804447966694e-05, "loss": 0.0272, "step": 44930 }, { "grad_norm": 0.9168070554733276, "learning_rate": 6.237202463625573e-05, "loss": 0.0267, "step": 44940 }, { "grad_norm": 0.38046708703041077, "learning_rate": 6.235600343985848e-05, "loss": 0.0359, "step": 44950 }, { "grad_norm": 0.33893346786499023, "learning_rate": 6.233998089222729e-05, "loss": 0.0262, "step": 44960 }, { "grad_norm": 0.2200274020433426, "learning_rate": 6.232395699511433e-05, "loss": 0.039, "step": 44970 }, { "grad_norm": 0.22867630422115326, "learning_rate": 6.230793175027199e-05, "loss": 0.0324, "step": 44980 }, { "grad_norm": 0.2622843384742737, "learning_rate": 6.22919051594527e-05, "loss": 0.0302, "step": 44990 }, { "grad_norm": 0.29190585017204285, "learning_rate": 6.227587722440917e-05, "loss": 0.0302, "step": 45000 }, { "grad_norm": 0.25319620966911316, "learning_rate": 6.225984794689414e-05, "loss": 0.0314, "step": 45010 }, { "grad_norm": 0.22875431180000305, "learning_rate": 6.22438173286606e-05, "loss": 0.032, "step": 45020 }, { "grad_norm": 0.40662193298339844, "learning_rate": 6.222778537146159e-05, "loss": 0.0222, "step": 45030 }, { "grad_norm": 0.34244221448898315, "learning_rate": 6.221175207705037e-05, "loss": 0.0284, "step": 45040 }, { "grad_norm": 0.3570960462093353, "learning_rate": 6.219571744718027e-05, "loss": 0.03, "step": 45050 }, { "grad_norm": 0.29933783411979675, "learning_rate": 6.217968148360487e-05, "loss": 0.0312, "step": 45060 }, { "grad_norm": 0.25496286153793335, "learning_rate": 6.216364418807782e-05, "loss": 0.0261, "step": 45070 }, { "grad_norm": 0.22414368391036987, "learning_rate": 6.214760556235292e-05, "loss": 0.0193, "step": 45080 }, { "grad_norm": 0.4435468018054962, "learning_rate": 6.213156560818414e-05, "loss": 0.0292, "step": 45090 }, { "grad_norm": 0.23472581803798676, "learning_rate": 6.211552432732559e-05, "loss": 0.0304, "step": 45100 }, { "grad_norm": 0.4142882525920868, "learning_rate": 6.20994817215315e-05, "loss": 0.0333, "step": 45110 }, { "grad_norm": 0.28860047459602356, "learning_rate": 6.208343779255628e-05, "loss": 0.0255, "step": 45120 }, { "grad_norm": 0.2739396393299103, "learning_rate": 6.206739254215449e-05, "loss": 0.0263, "step": 45130 }, { "grad_norm": 0.35333988070487976, "learning_rate": 6.205134597208077e-05, "loss": 0.0253, "step": 45140 }, { "grad_norm": 0.2115221470594406, "learning_rate": 6.203529808409e-05, "loss": 0.0229, "step": 45150 }, { "grad_norm": 0.3566688895225525, "learning_rate": 6.201924887993712e-05, "loss": 0.0333, "step": 45160 }, { "grad_norm": 0.2921079695224762, "learning_rate": 6.200319836137725e-05, "loss": 0.0238, "step": 45170 }, { "grad_norm": 0.32726138830184937, "learning_rate": 6.198714653016565e-05, "loss": 0.0289, "step": 45180 }, { "grad_norm": 0.406731516122818, "learning_rate": 6.197109338805774e-05, "loss": 0.0334, "step": 45190 }, { "grad_norm": 0.2080984264612198, "learning_rate": 6.195503893680903e-05, "loss": 0.0259, "step": 45200 }, { "grad_norm": 0.30318695306777954, "learning_rate": 6.193898317817524e-05, "loss": 0.0257, "step": 45210 }, { "grad_norm": 0.2142820954322815, "learning_rate": 6.192292611391221e-05, "loss": 0.0221, "step": 45220 }, { "grad_norm": 0.3813074827194214, "learning_rate": 6.190686774577591e-05, "loss": 0.0328, "step": 45230 }, { "grad_norm": 0.6278638243675232, "learning_rate": 6.189080807552245e-05, "loss": 0.0351, "step": 45240 }, { "grad_norm": 0.2667016386985779, "learning_rate": 6.187474710490809e-05, "loss": 0.0313, "step": 45250 }, { "grad_norm": 0.24286699295043945, "learning_rate": 6.185868483568926e-05, "loss": 0.0271, "step": 45260 }, { "grad_norm": 0.2104528844356537, "learning_rate": 6.184262126962245e-05, "loss": 0.0347, "step": 45270 }, { "grad_norm": 0.39670658111572266, "learning_rate": 6.182655640846442e-05, "loss": 0.0326, "step": 45280 }, { "grad_norm": 0.28516241908073425, "learning_rate": 6.181049025397196e-05, "loss": 0.0249, "step": 45290 }, { "grad_norm": 0.37932732701301575, "learning_rate": 6.179442280790202e-05, "loss": 0.035, "step": 45300 }, { "grad_norm": 0.34947413206100464, "learning_rate": 6.177835407201174e-05, "loss": 0.0313, "step": 45310 }, { "grad_norm": 0.29788196086883545, "learning_rate": 6.176228404805839e-05, "loss": 0.0284, "step": 45320 }, { "grad_norm": 0.33560433983802795, "learning_rate": 6.174621273779932e-05, "loss": 0.0327, "step": 45330 }, { "grad_norm": 0.3079582154750824, "learning_rate": 6.17301401429921e-05, "loss": 0.0255, "step": 45340 }, { "grad_norm": 0.24640555679798126, "learning_rate": 6.171406626539436e-05, "loss": 0.0333, "step": 45350 }, { "grad_norm": 0.7162890434265137, "learning_rate": 6.169799110676398e-05, "loss": 0.0375, "step": 45360 }, { "grad_norm": 0.2562437951564789, "learning_rate": 6.168191466885885e-05, "loss": 0.0272, "step": 45370 }, { "grad_norm": 0.28846246004104614, "learning_rate": 6.166583695343711e-05, "loss": 0.0258, "step": 45380 }, { "grad_norm": 0.2487739771604538, "learning_rate": 6.164975796225698e-05, "loss": 0.0367, "step": 45390 }, { "grad_norm": 0.309039831161499, "learning_rate": 6.163367769707683e-05, "loss": 0.0286, "step": 45400 }, { "grad_norm": 0.5057035088539124, "learning_rate": 6.161759615965519e-05, "loss": 0.0337, "step": 45410 }, { "grad_norm": 0.23677091300487518, "learning_rate": 6.16015133517507e-05, "loss": 0.0255, "step": 45420 }, { "grad_norm": 0.24006302654743195, "learning_rate": 6.158542927512214e-05, "loss": 0.0275, "step": 45430 }, { "grad_norm": 1.5701255798339844, "learning_rate": 6.156934393152846e-05, "loss": 0.0299, "step": 45440 }, { "grad_norm": 0.2652587592601776, "learning_rate": 6.15532573227287e-05, "loss": 0.0293, "step": 45450 }, { "grad_norm": 0.3127739131450653, "learning_rate": 6.153716945048212e-05, "loss": 0.0417, "step": 45460 }, { "grad_norm": 0.3781064450740814, "learning_rate": 6.152108031654802e-05, "loss": 0.0284, "step": 45470 }, { "grad_norm": 0.26605212688446045, "learning_rate": 6.15049899226859e-05, "loss": 0.025, "step": 45480 }, { "grad_norm": 0.2427665740251541, "learning_rate": 6.148889827065537e-05, "loss": 0.0258, "step": 45490 }, { "grad_norm": 0.26884812116622925, "learning_rate": 6.147280536221622e-05, "loss": 0.0255, "step": 45500 }, { "grad_norm": 0.254989892244339, "learning_rate": 6.145671119912832e-05, "loss": 0.0248, "step": 45510 }, { "grad_norm": 0.2618529796600342, "learning_rate": 6.144061578315169e-05, "loss": 0.0361, "step": 45520 }, { "grad_norm": 1.1221576929092407, "learning_rate": 6.142451911604654e-05, "loss": 0.0299, "step": 45530 }, { "grad_norm": 0.34542807936668396, "learning_rate": 6.140842119957315e-05, "loss": 0.0259, "step": 45540 }, { "grad_norm": 0.22252106666564941, "learning_rate": 6.139232203549197e-05, "loss": 0.0261, "step": 45550 }, { "grad_norm": 0.13669544458389282, "learning_rate": 6.137622162556357e-05, "loss": 0.0253, "step": 45560 }, { "grad_norm": 0.21626508235931396, "learning_rate": 6.13601199715487e-05, "loss": 0.0219, "step": 45570 }, { "grad_norm": 0.2996539771556854, "learning_rate": 6.134401707520816e-05, "loss": 0.0282, "step": 45580 }, { "grad_norm": 0.5727747678756714, "learning_rate": 6.132791293830298e-05, "loss": 0.0303, "step": 45590 }, { "grad_norm": 0.31689831614494324, "learning_rate": 6.131180756259428e-05, "loss": 0.0282, "step": 45600 }, { "grad_norm": 0.30836576223373413, "learning_rate": 6.129570094984331e-05, "loss": 0.0265, "step": 45610 }, { "grad_norm": 0.37032631039619446, "learning_rate": 6.127959310181145e-05, "loss": 0.0329, "step": 45620 }, { "grad_norm": 0.37775784730911255, "learning_rate": 6.126348402026026e-05, "loss": 0.0297, "step": 45630 }, { "grad_norm": 0.306924968957901, "learning_rate": 6.12473737069514e-05, "loss": 0.028, "step": 45640 }, { "grad_norm": 0.30480650067329407, "learning_rate": 6.123126216364665e-05, "loss": 0.0252, "step": 45650 }, { "grad_norm": 0.3230557143688202, "learning_rate": 6.121514939210797e-05, "loss": 0.0274, "step": 45660 }, { "grad_norm": 0.4539893865585327, "learning_rate": 6.119903539409741e-05, "loss": 0.034, "step": 45670 }, { "grad_norm": 0.37623903155326843, "learning_rate": 6.118292017137716e-05, "loss": 0.039, "step": 45680 }, { "grad_norm": 0.37235820293426514, "learning_rate": 6.116680372570959e-05, "loss": 0.0365, "step": 45690 }, { "grad_norm": 0.25001904368400574, "learning_rate": 6.115068605885713e-05, "loss": 0.0303, "step": 45700 }, { "grad_norm": 0.2632010579109192, "learning_rate": 6.113456717258243e-05, "loss": 0.0348, "step": 45710 }, { "grad_norm": 0.32190248370170593, "learning_rate": 6.11184470686482e-05, "loss": 0.0324, "step": 45720 }, { "grad_norm": 0.22242134809494019, "learning_rate": 6.110232574881731e-05, "loss": 0.0271, "step": 45730 }, { "grad_norm": 0.6161972284317017, "learning_rate": 6.108620321485277e-05, "loss": 0.0322, "step": 45740 }, { "grad_norm": 0.4087355136871338, "learning_rate": 6.107007946851773e-05, "loss": 0.0333, "step": 45750 }, { "grad_norm": 0.4157026410102844, "learning_rate": 6.105395451157542e-05, "loss": 0.0262, "step": 45760 }, { "grad_norm": 0.27341940999031067, "learning_rate": 6.103782834578928e-05, "loss": 0.0387, "step": 45770 }, { "grad_norm": 0.2384682148694992, "learning_rate": 6.102170097292281e-05, "loss": 0.0296, "step": 45780 }, { "grad_norm": 0.3209526538848877, "learning_rate": 6.1005572394739716e-05, "loss": 0.0335, "step": 45790 }, { "grad_norm": 0.3996790051460266, "learning_rate": 6.0989442613003765e-05, "loss": 0.0326, "step": 45800 }, { "grad_norm": 0.5528028011322021, "learning_rate": 6.097331162947888e-05, "loss": 0.0338, "step": 45810 }, { "grad_norm": 0.37646451592445374, "learning_rate": 6.095717944592914e-05, "loss": 0.0357, "step": 45820 }, { "grad_norm": 0.49369698762893677, "learning_rate": 6.094104606411873e-05, "loss": 0.0402, "step": 45830 }, { "grad_norm": 0.3309535086154938, "learning_rate": 6.0924911485811966e-05, "loss": 0.0402, "step": 45840 }, { "grad_norm": 0.4344029426574707, "learning_rate": 6.090877571277331e-05, "loss": 0.0359, "step": 45850 }, { "grad_norm": 0.2562806308269501, "learning_rate": 6.089263874676736e-05, "loss": 0.034, "step": 45860 }, { "grad_norm": 0.2635153830051422, "learning_rate": 6.0876500589558796e-05, "loss": 0.0268, "step": 45870 }, { "grad_norm": 0.5385434031486511, "learning_rate": 6.086036124291248e-05, "loss": 0.0298, "step": 45880 }, { "grad_norm": 0.5649303197860718, "learning_rate": 6.084422070859339e-05, "loss": 0.0342, "step": 45890 }, { "grad_norm": 0.42304685711860657, "learning_rate": 6.082807898836663e-05, "loss": 0.0319, "step": 45900 }, { "grad_norm": 0.4210191071033478, "learning_rate": 6.081193608399742e-05, "loss": 0.0308, "step": 45910 }, { "grad_norm": 0.2696475088596344, "learning_rate": 6.0795791997251164e-05, "loss": 0.0266, "step": 45920 }, { "grad_norm": 0.7179797887802124, "learning_rate": 6.0779646729893294e-05, "loss": 0.0348, "step": 45930 }, { "grad_norm": 0.35844725370407104, "learning_rate": 6.0763500283689476e-05, "loss": 0.0295, "step": 45940 }, { "grad_norm": 0.24625544250011444, "learning_rate": 6.0747352660405455e-05, "loss": 0.0278, "step": 45950 }, { "grad_norm": 0.21434065699577332, "learning_rate": 6.073120386180709e-05, "loss": 0.0288, "step": 45960 }, { "grad_norm": 0.3699074387550354, "learning_rate": 6.0715053889660425e-05, "loss": 0.0257, "step": 45970 }, { "grad_norm": 0.2663491666316986, "learning_rate": 6.069890274573157e-05, "loss": 0.0299, "step": 45980 }, { "grad_norm": 0.252487450838089, "learning_rate": 6.068275043178679e-05, "loss": 0.0338, "step": 45990 }, { "grad_norm": 0.2772486209869385, "learning_rate": 6.066659694959248e-05, "loss": 0.0348, "step": 46000 }, { "grad_norm": 0.3101842403411865, "learning_rate": 6.0650442300915176e-05, "loss": 0.0252, "step": 46010 }, { "grad_norm": 0.24878236651420593, "learning_rate": 6.063428648752152e-05, "loss": 0.0247, "step": 46020 }, { "grad_norm": 0.3440761864185333, "learning_rate": 6.0618129511178266e-05, "loss": 0.0326, "step": 46030 }, { "grad_norm": 0.2881053686141968, "learning_rate": 6.060197137365234e-05, "loss": 0.0356, "step": 46040 }, { "grad_norm": 0.2734110653400421, "learning_rate": 6.058581207671077e-05, "loss": 0.0289, "step": 46050 }, { "grad_norm": 0.4067889153957367, "learning_rate": 6.056965162212072e-05, "loss": 0.0296, "step": 46060 }, { "grad_norm": 0.2777262330055237, "learning_rate": 6.055349001164943e-05, "loss": 0.041, "step": 46070 }, { "grad_norm": 0.4308554232120514, "learning_rate": 6.0537327247064347e-05, "loss": 0.031, "step": 46080 }, { "grad_norm": 0.4189145565032959, "learning_rate": 6.0521163330133014e-05, "loss": 0.0387, "step": 46090 }, { "grad_norm": 0.39690423011779785, "learning_rate": 6.050499826262306e-05, "loss": 0.0301, "step": 46100 }, { "grad_norm": 0.292800635099411, "learning_rate": 6.0488832046302294e-05, "loss": 0.0296, "step": 46110 }, { "grad_norm": 1.9340214729309082, "learning_rate": 6.0472664682938626e-05, "loss": 0.0289, "step": 46120 }, { "grad_norm": 0.33014053106307983, "learning_rate": 6.045649617430009e-05, "loss": 0.0344, "step": 46130 }, { "grad_norm": 0.38721364736557007, "learning_rate": 6.0440326522154866e-05, "loss": 0.0268, "step": 46140 }, { "grad_norm": 0.22459614276885986, "learning_rate": 6.0424155728271224e-05, "loss": 0.0272, "step": 46150 }, { "grad_norm": 0.2255968451499939, "learning_rate": 6.040798379441758e-05, "loss": 0.0323, "step": 46160 }, { "grad_norm": 0.341133177280426, "learning_rate": 6.0391810722362485e-05, "loss": 0.0301, "step": 46170 }, { "grad_norm": 0.25098416209220886, "learning_rate": 6.037563651387458e-05, "loss": 0.0254, "step": 46180 }, { "grad_norm": 0.22949941456317902, "learning_rate": 6.0359461170722666e-05, "loss": 0.0281, "step": 46190 }, { "grad_norm": 0.2025047391653061, "learning_rate": 6.034328469467566e-05, "loss": 0.0295, "step": 46200 }, { "grad_norm": 0.2127958983182907, "learning_rate": 6.0327107087502596e-05, "loss": 0.0327, "step": 46210 }, { "grad_norm": 0.31169357895851135, "learning_rate": 6.031092835097262e-05, "loss": 0.0261, "step": 46220 }, { "grad_norm": 0.33318445086479187, "learning_rate": 6.0294748486855024e-05, "loss": 0.0205, "step": 46230 }, { "grad_norm": 0.3068305552005768, "learning_rate": 6.0278567496919216e-05, "loss": 0.0265, "step": 46240 }, { "grad_norm": 0.6789649724960327, "learning_rate": 6.026238538293472e-05, "loss": 0.0262, "step": 46250 }, { "grad_norm": 0.20337079465389252, "learning_rate": 6.024620214667118e-05, "loss": 0.031, "step": 46260 }, { "grad_norm": 0.30157458782196045, "learning_rate": 6.0230017789898384e-05, "loss": 0.0239, "step": 46270 }, { "grad_norm": 0.17144224047660828, "learning_rate": 6.021383231438622e-05, "loss": 0.0225, "step": 46280 }, { "grad_norm": 0.2959647476673126, "learning_rate": 6.0197645721904704e-05, "loss": 0.0235, "step": 46290 }, { "grad_norm": 0.3112697899341583, "learning_rate": 6.0181458014224e-05, "loss": 0.0212, "step": 46300 }, { "grad_norm": 0.25221318006515503, "learning_rate": 6.016526919311435e-05, "loss": 0.0235, "step": 46310 }, { "grad_norm": 0.36061322689056396, "learning_rate": 6.014907926034613e-05, "loss": 0.0308, "step": 46320 }, { "grad_norm": 0.2336743324995041, "learning_rate": 6.0132888217689875e-05, "loss": 0.0303, "step": 46330 }, { "grad_norm": 0.5902330279350281, "learning_rate": 6.01166960669162e-05, "loss": 0.0333, "step": 46340 }, { "grad_norm": 0.2248220145702362, "learning_rate": 6.0100502809795845e-05, "loss": 0.023, "step": 46350 }, { "grad_norm": 0.3986828029155731, "learning_rate": 6.00843084480997e-05, "loss": 0.0365, "step": 46360 }, { "grad_norm": 0.32235556840896606, "learning_rate": 6.0068112983598736e-05, "loss": 0.0289, "step": 46370 }, { "grad_norm": 0.376355916261673, "learning_rate": 6.005191641806407e-05, "loss": 0.0305, "step": 46380 }, { "grad_norm": 0.2823944687843323, "learning_rate": 6.003571875326694e-05, "loss": 0.029, "step": 46390 }, { "grad_norm": 0.3401317894458771, "learning_rate": 6.00195199909787e-05, "loss": 0.0322, "step": 46400 }, { "grad_norm": 0.40019333362579346, "learning_rate": 6.000332013297082e-05, "loss": 0.0246, "step": 46410 }, { "grad_norm": 0.2938203811645508, "learning_rate": 5.998711918101487e-05, "loss": 0.0319, "step": 46420 }, { "grad_norm": 0.3095267713069916, "learning_rate": 5.997091713688261e-05, "loss": 0.0282, "step": 46430 }, { "grad_norm": 0.1938820481300354, "learning_rate": 5.9954714002345836e-05, "loss": 0.0318, "step": 46440 }, { "grad_norm": 1.0575886964797974, "learning_rate": 5.993850977917649e-05, "loss": 0.0421, "step": 46450 }, { "grad_norm": 0.19000910222530365, "learning_rate": 5.992230446914667e-05, "loss": 0.0295, "step": 46460 }, { "grad_norm": 0.30436578392982483, "learning_rate": 5.990609807402855e-05, "loss": 0.0339, "step": 46470 }, { "grad_norm": 0.2247982919216156, "learning_rate": 5.988989059559443e-05, "loss": 0.0253, "step": 46480 }, { "grad_norm": 0.34339264035224915, "learning_rate": 5.987368203561675e-05, "loss": 0.0271, "step": 46490 }, { "grad_norm": 0.16451503336429596, "learning_rate": 5.9857472395868055e-05, "loss": 0.0217, "step": 46500 }, { "grad_norm": 0.3365669548511505, "learning_rate": 5.9841261678120983e-05, "loss": 0.0307, "step": 46510 }, { "grad_norm": 0.3053838312625885, "learning_rate": 5.982504988414834e-05, "loss": 0.0227, "step": 46520 }, { "grad_norm": 0.343412309885025, "learning_rate": 5.9808837015723015e-05, "loss": 0.0473, "step": 46530 }, { "grad_norm": 0.3097294270992279, "learning_rate": 5.9792623074618016e-05, "loss": 0.0258, "step": 46540 }, { "grad_norm": 0.26769551634788513, "learning_rate": 5.977640806260648e-05, "loss": 0.0253, "step": 46550 }, { "grad_norm": 0.30466482043266296, "learning_rate": 5.976019198146166e-05, "loss": 0.0203, "step": 46560 }, { "grad_norm": 0.2997635006904602, "learning_rate": 5.974397483295692e-05, "loss": 0.0234, "step": 46570 }, { "grad_norm": 0.27229082584381104, "learning_rate": 5.972775661886575e-05, "loss": 0.0262, "step": 46580 }, { "grad_norm": 0.3114837408065796, "learning_rate": 5.971153734096173e-05, "loss": 0.0335, "step": 46590 }, { "grad_norm": 0.2535277307033539, "learning_rate": 5.969531700101859e-05, "loss": 0.0295, "step": 46600 }, { "grad_norm": 0.2011626660823822, "learning_rate": 5.9679095600810155e-05, "loss": 0.0258, "step": 46610 }, { "grad_norm": 0.31327781081199646, "learning_rate": 5.9662873142110384e-05, "loss": 0.0233, "step": 46620 }, { "grad_norm": 0.3379671573638916, "learning_rate": 5.964664962669333e-05, "loss": 0.0295, "step": 46630 }, { "grad_norm": 0.2924256920814514, "learning_rate": 5.9630425056333186e-05, "loss": 0.0264, "step": 46640 }, { "grad_norm": 0.24404236674308777, "learning_rate": 5.961419943280422e-05, "loss": 0.0316, "step": 46650 }, { "grad_norm": 0.30723249912261963, "learning_rate": 5.959797275788087e-05, "loss": 0.0252, "step": 46660 }, { "grad_norm": 0.3185877203941345, "learning_rate": 5.958174503333765e-05, "loss": 0.0272, "step": 46670 }, { "grad_norm": 0.23873494565486908, "learning_rate": 5.9565516260949195e-05, "loss": 0.0345, "step": 46680 }, { "grad_norm": 0.27129289507865906, "learning_rate": 5.9549286442490273e-05, "loss": 0.0269, "step": 46690 }, { "grad_norm": 0.26207780838012695, "learning_rate": 5.953305557973572e-05, "loss": 0.0237, "step": 46700 }, { "grad_norm": 0.5010031461715698, "learning_rate": 5.9516823674460564e-05, "loss": 0.0301, "step": 46710 }, { "grad_norm": 0.311979204416275, "learning_rate": 5.9500590728439875e-05, "loss": 0.025, "step": 46720 }, { "grad_norm": 0.33969810605049133, "learning_rate": 5.948435674344886e-05, "loss": 0.0353, "step": 46730 }, { "grad_norm": 0.38034456968307495, "learning_rate": 5.946812172126285e-05, "loss": 0.0273, "step": 46740 }, { "grad_norm": 0.2927910089492798, "learning_rate": 5.945188566365729e-05, "loss": 0.0287, "step": 46750 }, { "grad_norm": 0.2285035103559494, "learning_rate": 5.943564857240773e-05, "loss": 0.0311, "step": 46760 }, { "grad_norm": 0.4352577030658722, "learning_rate": 5.941941044928981e-05, "loss": 0.0301, "step": 46770 }, { "grad_norm": 0.2659836411476135, "learning_rate": 5.940317129607935e-05, "loss": 0.0237, "step": 46780 }, { "grad_norm": 0.47768113017082214, "learning_rate": 5.9386931114552204e-05, "loss": 0.0311, "step": 46790 }, { "grad_norm": 0.3786112368106842, "learning_rate": 5.937068990648438e-05, "loss": 0.0316, "step": 46800 }, { "grad_norm": 0.25105637311935425, "learning_rate": 5.935444767365199e-05, "loss": 0.0421, "step": 46810 }, { "grad_norm": 0.450412780046463, "learning_rate": 5.933820441783129e-05, "loss": 0.0295, "step": 46820 }, { "grad_norm": 0.29655882716178894, "learning_rate": 5.932196014079857e-05, "loss": 0.0237, "step": 46830 }, { "grad_norm": 0.2978893518447876, "learning_rate": 5.930571484433032e-05, "loss": 0.0211, "step": 46840 }, { "grad_norm": 0.41006818413734436, "learning_rate": 5.928946853020309e-05, "loss": 0.0286, "step": 46850 }, { "grad_norm": 0.46103179454803467, "learning_rate": 5.927322120019352e-05, "loss": 0.0311, "step": 46860 }, { "grad_norm": 0.26592767238616943, "learning_rate": 5.925697285607844e-05, "loss": 0.0281, "step": 46870 }, { "grad_norm": 0.19994474947452545, "learning_rate": 5.924072349963472e-05, "loss": 0.0246, "step": 46880 }, { "grad_norm": 0.21610091626644135, "learning_rate": 5.922447313263938e-05, "loss": 0.0206, "step": 46890 }, { "grad_norm": 0.3340885639190674, "learning_rate": 5.920822175686952e-05, "loss": 0.0307, "step": 46900 }, { "grad_norm": 0.1880345195531845, "learning_rate": 5.919196937410239e-05, "loss": 0.0266, "step": 46910 }, { "grad_norm": 0.22249199450016022, "learning_rate": 5.917571598611529e-05, "loss": 0.0246, "step": 46920 }, { "grad_norm": 0.28426679968833923, "learning_rate": 5.9159461594685686e-05, "loss": 0.029, "step": 46930 }, { "grad_norm": 0.4202539324760437, "learning_rate": 5.914320620159114e-05, "loss": 0.0293, "step": 46940 }, { "grad_norm": 0.24471266567707062, "learning_rate": 5.912694980860931e-05, "loss": 0.0286, "step": 46950 }, { "grad_norm": 0.6174537539482117, "learning_rate": 5.9110692417517964e-05, "loss": 0.0257, "step": 46960 }, { "grad_norm": 0.3173363208770752, "learning_rate": 5.9094434030095e-05, "loss": 0.0246, "step": 46970 }, { "grad_norm": 0.2859749495983124, "learning_rate": 5.907817464811841e-05, "loss": 0.0273, "step": 46980 }, { "grad_norm": 0.3793732821941376, "learning_rate": 5.906191427336627e-05, "loss": 0.0251, "step": 46990 }, { "grad_norm": 0.23776179552078247, "learning_rate": 5.9045652907616835e-05, "loss": 0.028, "step": 47000 }, { "grad_norm": 0.25800883769989014, "learning_rate": 5.902939055264838e-05, "loss": 0.0251, "step": 47010 }, { "grad_norm": 0.23643861711025238, "learning_rate": 5.901312721023935e-05, "loss": 0.0257, "step": 47020 }, { "grad_norm": 0.3076287508010864, "learning_rate": 5.8996862882168294e-05, "loss": 0.0225, "step": 47030 }, { "grad_norm": 0.18661105632781982, "learning_rate": 5.8980597570213826e-05, "loss": 0.0263, "step": 47040 }, { "grad_norm": 0.36677783727645874, "learning_rate": 5.896433127615471e-05, "loss": 0.0256, "step": 47050 }, { "grad_norm": 0.22976885735988617, "learning_rate": 5.894806400176981e-05, "loss": 0.0236, "step": 47060 }, { "grad_norm": 0.532321572303772, "learning_rate": 5.893179574883808e-05, "loss": 0.0261, "step": 47070 }, { "grad_norm": 0.22348164021968842, "learning_rate": 5.8915526519138585e-05, "loss": 0.0243, "step": 47080 }, { "grad_norm": 0.3413746654987335, "learning_rate": 5.889925631445053e-05, "loss": 0.0245, "step": 47090 }, { "grad_norm": 0.34818172454833984, "learning_rate": 5.888298513655318e-05, "loss": 0.0303, "step": 47100 }, { "grad_norm": 0.2784135639667511, "learning_rate": 5.886671298722592e-05, "loss": 0.0254, "step": 47110 }, { "grad_norm": 0.2776047885417938, "learning_rate": 5.885043986824828e-05, "loss": 0.0226, "step": 47120 }, { "grad_norm": 0.32297155261039734, "learning_rate": 5.883416578139982e-05, "loss": 0.033, "step": 47130 }, { "grad_norm": 0.28183239698410034, "learning_rate": 5.881789072846029e-05, "loss": 0.0313, "step": 47140 }, { "grad_norm": 0.44613033533096313, "learning_rate": 5.8801614711209474e-05, "loss": 0.0334, "step": 47150 }, { "grad_norm": 0.27170127630233765, "learning_rate": 5.878533773142732e-05, "loss": 0.0297, "step": 47160 }, { "grad_norm": 0.4500122368335724, "learning_rate": 5.8769059790893834e-05, "loss": 0.0286, "step": 47170 }, { "grad_norm": 0.2799986004829407, "learning_rate": 5.8752780891389145e-05, "loss": 0.0267, "step": 47180 }, { "grad_norm": 0.2813739776611328, "learning_rate": 5.873650103469351e-05, "loss": 0.0234, "step": 47190 }, { "grad_norm": 0.2371394783258438, "learning_rate": 5.872022022258725e-05, "loss": 0.0178, "step": 47200 }, { "grad_norm": 0.23360562324523926, "learning_rate": 5.8703938456850814e-05, "loss": 0.0252, "step": 47210 }, { "grad_norm": 0.18038715422153473, "learning_rate": 5.868765573926477e-05, "loss": 0.0287, "step": 47220 }, { "grad_norm": 0.3184298574924469, "learning_rate": 5.867137207160974e-05, "loss": 0.0273, "step": 47230 }, { "grad_norm": 0.20185419917106628, "learning_rate": 5.86550874556665e-05, "loss": 0.026, "step": 47240 }, { "grad_norm": 0.3018316924571991, "learning_rate": 5.863880189321592e-05, "loss": 0.0326, "step": 47250 }, { "grad_norm": 0.3900303840637207, "learning_rate": 5.862251538603894e-05, "loss": 0.029, "step": 47260 }, { "grad_norm": 0.24686260521411896, "learning_rate": 5.8606227935916656e-05, "loss": 0.0259, "step": 47270 }, { "grad_norm": 0.1728874146938324, "learning_rate": 5.858993954463021e-05, "loss": 0.0208, "step": 47280 }, { "grad_norm": 0.2934775948524475, "learning_rate": 5.8573650213960896e-05, "loss": 0.0263, "step": 47290 }, { "grad_norm": 0.39944931864738464, "learning_rate": 5.8557359945690084e-05, "loss": 0.0189, "step": 47300 }, { "grad_norm": 0.19819986820220947, "learning_rate": 5.8541068741599246e-05, "loss": 0.0281, "step": 47310 }, { "grad_norm": 0.2154642790555954, "learning_rate": 5.8524776603469985e-05, "loss": 0.0264, "step": 47320 }, { "grad_norm": 0.29609280824661255, "learning_rate": 5.850848353308397e-05, "loss": 0.0289, "step": 47330 }, { "grad_norm": 0.31308484077453613, "learning_rate": 5.849218953222297e-05, "loss": 0.0258, "step": 47340 }, { "grad_norm": 0.2812456786632538, "learning_rate": 5.847589460266891e-05, "loss": 0.0311, "step": 47350 }, { "grad_norm": 0.252332478761673, "learning_rate": 5.845959874620375e-05, "loss": 0.0256, "step": 47360 }, { "grad_norm": 0.23308034241199493, "learning_rate": 5.84433019646096e-05, "loss": 0.0282, "step": 47370 }, { "grad_norm": 0.2694456875324249, "learning_rate": 5.842700425966863e-05, "loss": 0.0323, "step": 47380 }, { "grad_norm": 0.4027070701122284, "learning_rate": 5.841070563316315e-05, "loss": 0.0244, "step": 47390 }, { "grad_norm": 0.2496279925107956, "learning_rate": 5.839440608687554e-05, "loss": 0.0378, "step": 47400 }, { "grad_norm": 0.3459221124649048, "learning_rate": 5.837810562258831e-05, "loss": 0.0277, "step": 47410 }, { "grad_norm": 0.23478886485099792, "learning_rate": 5.836180424208405e-05, "loss": 0.0306, "step": 47420 }, { "grad_norm": 0.2296585738658905, "learning_rate": 5.834550194714543e-05, "loss": 0.0318, "step": 47430 }, { "grad_norm": 0.3207264542579651, "learning_rate": 5.832919873955528e-05, "loss": 0.0305, "step": 47440 }, { "grad_norm": 0.19974224269390106, "learning_rate": 5.831289462109648e-05, "loss": 0.0249, "step": 47450 }, { "grad_norm": 0.17788077890872955, "learning_rate": 5.829658959355202e-05, "loss": 0.0407, "step": 47460 }, { "grad_norm": 0.8457727432250977, "learning_rate": 5.8280283658704995e-05, "loss": 0.0354, "step": 47470 }, { "grad_norm": 0.24280625581741333, "learning_rate": 5.826397681833859e-05, "loss": 0.0373, "step": 47480 }, { "grad_norm": 0.34820833802223206, "learning_rate": 5.824766907423612e-05, "loss": 0.0328, "step": 47490 }, { "grad_norm": 0.2223072648048401, "learning_rate": 5.8231360428180945e-05, "loss": 0.0251, "step": 47500 }, { "grad_norm": 0.2563765347003937, "learning_rate": 5.821505088195658e-05, "loss": 0.027, "step": 47510 }, { "grad_norm": 0.27930566668510437, "learning_rate": 5.819874043734661e-05, "loss": 0.0293, "step": 47520 }, { "grad_norm": 0.30189234018325806, "learning_rate": 5.8182429096134695e-05, "loss": 0.024, "step": 47530 }, { "grad_norm": 0.28135088086128235, "learning_rate": 5.816611686010465e-05, "loss": 0.0217, "step": 47540 }, { "grad_norm": 0.3297702670097351, "learning_rate": 5.814980373104033e-05, "loss": 0.0312, "step": 47550 }, { "grad_norm": 0.4380609691143036, "learning_rate": 5.813348971072572e-05, "loss": 0.0317, "step": 47560 }, { "grad_norm": 0.266976535320282, "learning_rate": 5.811717480094492e-05, "loss": 0.0223, "step": 47570 }, { "grad_norm": 0.5221756100654602, "learning_rate": 5.810085900348209e-05, "loss": 0.028, "step": 47580 }, { "grad_norm": 0.3021552264690399, "learning_rate": 5.8084542320121483e-05, "loss": 0.0236, "step": 47590 }, { "grad_norm": 0.1626214236021042, "learning_rate": 5.8068224752647497e-05, "loss": 0.0402, "step": 47600 }, { "grad_norm": 0.190687894821167, "learning_rate": 5.805190630284457e-05, "loss": 0.0245, "step": 47610 }, { "grad_norm": 0.29866158962249756, "learning_rate": 5.803558697249729e-05, "loss": 0.027, "step": 47620 }, { "grad_norm": 0.6914454102516174, "learning_rate": 5.8019266763390276e-05, "loss": 0.0373, "step": 47630 }, { "grad_norm": 0.44970834255218506, "learning_rate": 5.800294567730833e-05, "loss": 0.0302, "step": 47640 }, { "grad_norm": 0.466385155916214, "learning_rate": 5.798662371603628e-05, "loss": 0.0262, "step": 47650 }, { "grad_norm": 0.4978683590888977, "learning_rate": 5.797030088135904e-05, "loss": 0.0278, "step": 47660 }, { "grad_norm": 0.2797005772590637, "learning_rate": 5.7953977175061705e-05, "loss": 0.0314, "step": 47670 }, { "grad_norm": 0.3054952621459961, "learning_rate": 5.793765259892938e-05, "loss": 0.0324, "step": 47680 }, { "grad_norm": 0.21681435406208038, "learning_rate": 5.792132715474729e-05, "loss": 0.0276, "step": 47690 }, { "grad_norm": 0.3007465600967407, "learning_rate": 5.790500084430078e-05, "loss": 0.0283, "step": 47700 }, { "grad_norm": 0.3840237855911255, "learning_rate": 5.7888673669375274e-05, "loss": 0.0284, "step": 47710 }, { "grad_norm": 0.28933998942375183, "learning_rate": 5.787234563175625e-05, "loss": 0.026, "step": 47720 }, { "grad_norm": 0.33880335092544556, "learning_rate": 5.7856016733229355e-05, "loss": 0.03, "step": 47730 }, { "grad_norm": 0.25275489687919617, "learning_rate": 5.7839686975580297e-05, "loss": 0.0374, "step": 47740 }, { "grad_norm": 0.7632362246513367, "learning_rate": 5.782335636059484e-05, "loss": 0.0345, "step": 47750 }, { "grad_norm": 0.30654287338256836, "learning_rate": 5.780702489005889e-05, "loss": 0.0274, "step": 47760 }, { "grad_norm": 0.5201886296272278, "learning_rate": 5.779069256575845e-05, "loss": 0.0326, "step": 47770 }, { "grad_norm": 0.2859387695789337, "learning_rate": 5.7774359389479574e-05, "loss": 0.0316, "step": 47780 }, { "grad_norm": 0.39183372259140015, "learning_rate": 5.775802536300845e-05, "loss": 0.0291, "step": 47790 }, { "grad_norm": 0.24981528520584106, "learning_rate": 5.774169048813134e-05, "loss": 0.0231, "step": 47800 }, { "grad_norm": 0.23808377981185913, "learning_rate": 5.77253547666346e-05, "loss": 0.0243, "step": 47810 }, { "grad_norm": 0.4302658438682556, "learning_rate": 5.770901820030465e-05, "loss": 0.0227, "step": 47820 }, { "grad_norm": 0.23778343200683594, "learning_rate": 5.769268079092809e-05, "loss": 0.0338, "step": 47830 }, { "grad_norm": 0.25156697630882263, "learning_rate": 5.767634254029151e-05, "loss": 0.0292, "step": 47840 }, { "grad_norm": 0.3352768123149872, "learning_rate": 5.7660003450181655e-05, "loss": 0.0401, "step": 47850 }, { "grad_norm": 0.18903011083602905, "learning_rate": 5.764366352238534e-05, "loss": 0.0227, "step": 47860 }, { "grad_norm": 0.3017178475856781, "learning_rate": 5.7627322758689474e-05, "loss": 0.025, "step": 47870 }, { "grad_norm": 0.20227012038230896, "learning_rate": 5.761098116088105e-05, "loss": 0.0271, "step": 47880 }, { "grad_norm": 0.670952320098877, "learning_rate": 5.759463873074717e-05, "loss": 0.0278, "step": 47890 }, { "grad_norm": 0.20098280906677246, "learning_rate": 5.757829547007504e-05, "loss": 0.029, "step": 47900 }, { "grad_norm": 0.21951542794704437, "learning_rate": 5.756195138065189e-05, "loss": 0.0222, "step": 47910 }, { "grad_norm": 0.3403526544570923, "learning_rate": 5.754560646426511e-05, "loss": 0.0242, "step": 47920 }, { "grad_norm": 0.27837029099464417, "learning_rate": 5.752926072270216e-05, "loss": 0.0234, "step": 47930 }, { "grad_norm": 0.3021240532398224, "learning_rate": 5.7512914157750563e-05, "loss": 0.023, "step": 47940 }, { "grad_norm": 0.22505232691764832, "learning_rate": 5.749656677119798e-05, "loss": 0.0314, "step": 47950 }, { "grad_norm": 0.1882697194814682, "learning_rate": 5.7480218564832125e-05, "loss": 0.0202, "step": 47960 }, { "grad_norm": 0.1973930299282074, "learning_rate": 5.746386954044082e-05, "loss": 0.0271, "step": 47970 }, { "grad_norm": 0.3815626800060272, "learning_rate": 5.744751969981195e-05, "loss": 0.0226, "step": 47980 }, { "grad_norm": 0.26714086532592773, "learning_rate": 5.7431169044733526e-05, "loss": 0.0278, "step": 47990 }, { "grad_norm": 1.8312898874282837, "learning_rate": 5.741481757699364e-05, "loss": 0.0254, "step": 48000 }, { "grad_norm": 0.20981353521347046, "learning_rate": 5.7398465298380434e-05, "loss": 0.025, "step": 48010 }, { "grad_norm": 0.4000118672847748, "learning_rate": 5.7382112210682193e-05, "loss": 0.0331, "step": 48020 }, { "grad_norm": 0.41903191804885864, "learning_rate": 5.7365758315687266e-05, "loss": 0.028, "step": 48030 }, { "grad_norm": 0.310173898935318, "learning_rate": 5.734940361518407e-05, "loss": 0.0311, "step": 48040 }, { "grad_norm": 0.28752750158309937, "learning_rate": 5.733304811096116e-05, "loss": 0.0246, "step": 48050 }, { "grad_norm": 0.22352449595928192, "learning_rate": 5.731669180480713e-05, "loss": 0.0263, "step": 48060 }, { "grad_norm": 0.3179892301559448, "learning_rate": 5.730033469851067e-05, "loss": 0.0419, "step": 48070 }, { "grad_norm": 0.2547716796398163, "learning_rate": 5.7283976793860607e-05, "loss": 0.0293, "step": 48080 }, { "grad_norm": 0.15820379555225372, "learning_rate": 5.726761809264577e-05, "loss": 0.0219, "step": 48090 }, { "grad_norm": 0.21534231305122375, "learning_rate": 5.7251258596655155e-05, "loss": 0.0259, "step": 48100 }, { "grad_norm": 0.5432296395301819, "learning_rate": 5.72348983076778e-05, "loss": 0.0319, "step": 48110 }, { "grad_norm": 0.22355353832244873, "learning_rate": 5.7218537227502854e-05, "loss": 0.026, "step": 48120 }, { "grad_norm": 0.16917012631893158, "learning_rate": 5.720217535791951e-05, "loss": 0.0259, "step": 48130 }, { "grad_norm": 0.2312084436416626, "learning_rate": 5.718581270071711e-05, "loss": 0.0273, "step": 48140 }, { "grad_norm": 0.24235554039478302, "learning_rate": 5.716944925768505e-05, "loss": 0.0299, "step": 48150 }, { "grad_norm": 0.3068114221096039, "learning_rate": 5.7153085030612786e-05, "loss": 0.0326, "step": 48160 }, { "grad_norm": 0.39334630966186523, "learning_rate": 5.713672002128989e-05, "loss": 0.0333, "step": 48170 }, { "grad_norm": 0.2500554025173187, "learning_rate": 5.712035423150602e-05, "loss": 0.0266, "step": 48180 }, { "grad_norm": 0.32230252027511597, "learning_rate": 5.710398766305094e-05, "loss": 0.0261, "step": 48190 }, { "grad_norm": 0.3444017469882965, "learning_rate": 5.708762031771442e-05, "loss": 0.0301, "step": 48200 }, { "grad_norm": 0.27432680130004883, "learning_rate": 5.707125219728642e-05, "loss": 0.0251, "step": 48210 }, { "grad_norm": 0.40256884694099426, "learning_rate": 5.7054883303556905e-05, "loss": 0.0284, "step": 48220 }, { "grad_norm": 0.33888742327690125, "learning_rate": 5.703851363831595e-05, "loss": 0.0305, "step": 48230 }, { "grad_norm": 0.30402496457099915, "learning_rate": 5.702214320335374e-05, "loss": 0.0401, "step": 48240 }, { "grad_norm": 0.23036275804042816, "learning_rate": 5.7005772000460514e-05, "loss": 0.0283, "step": 48250 }, { "grad_norm": 0.4401937425136566, "learning_rate": 5.698940003142656e-05, "loss": 0.0265, "step": 48260 }, { "grad_norm": 0.27522122859954834, "learning_rate": 5.697302729804236e-05, "loss": 0.0294, "step": 48270 }, { "grad_norm": 0.5330785512924194, "learning_rate": 5.695665380209837e-05, "loss": 0.0245, "step": 48280 }, { "grad_norm": 0.29949134588241577, "learning_rate": 5.6940279545385165e-05, "loss": 0.026, "step": 48290 }, { "grad_norm": 0.22555701434612274, "learning_rate": 5.692390452969344e-05, "loss": 0.0214, "step": 48300 }, { "grad_norm": 0.3411048650741577, "learning_rate": 5.690752875681392e-05, "loss": 0.0323, "step": 48310 }, { "grad_norm": 0.30183351039886475, "learning_rate": 5.6891152228537435e-05, "loss": 0.0303, "step": 48320 }, { "grad_norm": 0.28320273756980896, "learning_rate": 5.687477494665492e-05, "loss": 0.0232, "step": 48330 }, { "grad_norm": 0.3231810927391052, "learning_rate": 5.685839691295733e-05, "loss": 0.0262, "step": 48340 }, { "grad_norm": 0.7930566668510437, "learning_rate": 5.6842018129235786e-05, "loss": 0.0269, "step": 48350 }, { "grad_norm": 0.21467337012290955, "learning_rate": 5.6825638597281404e-05, "loss": 0.0218, "step": 48360 }, { "grad_norm": 0.2725805938243866, "learning_rate": 5.680925831888546e-05, "loss": 0.0294, "step": 48370 }, { "grad_norm": 0.32780125737190247, "learning_rate": 5.6792877295839274e-05, "loss": 0.0283, "step": 48380 }, { "grad_norm": 0.7185071110725403, "learning_rate": 5.6776495529934224e-05, "loss": 0.0358, "step": 48390 }, { "grad_norm": 0.370918869972229, "learning_rate": 5.6760113022961824e-05, "loss": 0.0269, "step": 48400 }, { "grad_norm": 0.5076048374176025, "learning_rate": 5.6743729776713617e-05, "loss": 0.0311, "step": 48410 }, { "grad_norm": 0.29735469818115234, "learning_rate": 5.672734579298126e-05, "loss": 0.03, "step": 48420 }, { "grad_norm": 0.2014411836862564, "learning_rate": 5.671096107355649e-05, "loss": 0.0253, "step": 48430 }, { "grad_norm": 0.251495897769928, "learning_rate": 5.66945756202311e-05, "loss": 0.0286, "step": 48440 }, { "grad_norm": 0.3542611300945282, "learning_rate": 5.667818943479699e-05, "loss": 0.0244, "step": 48450 }, { "grad_norm": 0.304790198802948, "learning_rate": 5.666180251904612e-05, "loss": 0.0261, "step": 48460 }, { "grad_norm": 0.34050247073173523, "learning_rate": 5.6645414874770555e-05, "loss": 0.0247, "step": 48470 }, { "grad_norm": 0.22786343097686768, "learning_rate": 5.66290265037624e-05, "loss": 0.0235, "step": 48480 }, { "grad_norm": 0.26349693536758423, "learning_rate": 5.661263740781386e-05, "loss": 0.0229, "step": 48490 }, { "grad_norm": 0.22254440188407898, "learning_rate": 5.6596247588717254e-05, "loss": 0.0273, "step": 48500 }, { "grad_norm": 0.3633420169353485, "learning_rate": 5.6579857048264926e-05, "loss": 0.0273, "step": 48510 }, { "grad_norm": 0.45558223128318787, "learning_rate": 5.6563465788249314e-05, "loss": 0.0324, "step": 48520 }, { "grad_norm": 0.33305779099464417, "learning_rate": 5.6547073810462956e-05, "loss": 0.0285, "step": 48530 }, { "grad_norm": 0.19577762484550476, "learning_rate": 5.653068111669846e-05, "loss": 0.0265, "step": 48540 }, { "grad_norm": 0.4075428247451782, "learning_rate": 5.651428770874848e-05, "loss": 0.0264, "step": 48550 }, { "grad_norm": 0.27546054124832153, "learning_rate": 5.64978935884058e-05, "loss": 0.0211, "step": 48560 }, { "grad_norm": 1.8246992826461792, "learning_rate": 5.6481498757463244e-05, "loss": 0.0293, "step": 48570 }, { "grad_norm": 0.2334388643503189, "learning_rate": 5.646510321771373e-05, "loss": 0.0299, "step": 48580 }, { "grad_norm": 0.1390998810529709, "learning_rate": 5.644870697095024e-05, "loss": 0.0255, "step": 48590 }, { "grad_norm": 0.21044421195983887, "learning_rate": 5.643231001896586e-05, "loss": 0.0283, "step": 48600 }, { "grad_norm": 0.24898898601531982, "learning_rate": 5.6415912363553726e-05, "loss": 0.0608, "step": 48610 }, { "grad_norm": 0.22232413291931152, "learning_rate": 5.639951400650706e-05, "loss": 0.0207, "step": 48620 }, { "grad_norm": 0.34774279594421387, "learning_rate": 5.6383114949619165e-05, "loss": 0.0268, "step": 48630 }, { "grad_norm": 0.3499963581562042, "learning_rate": 5.636671519468342e-05, "loss": 0.0273, "step": 48640 }, { "grad_norm": 0.3142949044704437, "learning_rate": 5.635031474349327e-05, "loss": 0.0252, "step": 48650 }, { "grad_norm": 0.26036882400512695, "learning_rate": 5.6333913597842246e-05, "loss": 0.0229, "step": 48660 }, { "grad_norm": 0.28299111127853394, "learning_rate": 5.6317511759523955e-05, "loss": 0.0316, "step": 48670 }, { "grad_norm": 0.29537561535835266, "learning_rate": 5.630110923033207e-05, "loss": 0.0239, "step": 48680 }, { "grad_norm": 0.17832575738430023, "learning_rate": 5.628470601206036e-05, "loss": 0.0321, "step": 48690 }, { "grad_norm": 0.17293183505535126, "learning_rate": 5.626830210650266e-05, "loss": 0.0272, "step": 48700 }, { "grad_norm": 0.1642238050699234, "learning_rate": 5.625189751545285e-05, "loss": 0.037, "step": 48710 }, { "grad_norm": 0.5049245953559875, "learning_rate": 5.6235492240704936e-05, "loss": 0.0321, "step": 48720 }, { "grad_norm": 0.15578287839889526, "learning_rate": 5.621908628405296e-05, "loss": 0.0291, "step": 48730 }, { "grad_norm": 0.23699124157428741, "learning_rate": 5.620267964729106e-05, "loss": 0.0348, "step": 48740 }, { "grad_norm": 0.4278936982154846, "learning_rate": 5.618627233221344e-05, "loss": 0.0195, "step": 48750 }, { "grad_norm": 0.332440048456192, "learning_rate": 5.616986434061438e-05, "loss": 0.0299, "step": 48760 }, { "grad_norm": 0.28421342372894287, "learning_rate": 5.615345567428822e-05, "loss": 0.021, "step": 48770 }, { "grad_norm": 0.46755823493003845, "learning_rate": 5.613704633502941e-05, "loss": 0.0278, "step": 48780 }, { "grad_norm": 0.281739741563797, "learning_rate": 5.612063632463245e-05, "loss": 0.0289, "step": 48790 }, { "grad_norm": 0.49279123544692993, "learning_rate": 5.610422564489188e-05, "loss": 0.0276, "step": 48800 }, { "grad_norm": 0.2821773886680603, "learning_rate": 5.608781429760239e-05, "loss": 0.0274, "step": 48810 }, { "grad_norm": 0.3432987630367279, "learning_rate": 5.607140228455866e-05, "loss": 0.0286, "step": 48820 }, { "grad_norm": 0.3745759129524231, "learning_rate": 5.605498960755553e-05, "loss": 0.0345, "step": 48830 }, { "grad_norm": 0.34053513407707214, "learning_rate": 5.603857626838782e-05, "loss": 0.0334, "step": 48840 }, { "grad_norm": 0.26106685400009155, "learning_rate": 5.60221622688505e-05, "loss": 0.0359, "step": 48850 }, { "grad_norm": 0.2311643809080124, "learning_rate": 5.6005747610738565e-05, "loss": 0.0214, "step": 48860 }, { "grad_norm": 0.2641846835613251, "learning_rate": 5.59893322958471e-05, "loss": 0.0271, "step": 48870 }, { "grad_norm": 0.1967451125383377, "learning_rate": 5.5972916325971256e-05, "loss": 0.0264, "step": 48880 }, { "grad_norm": 0.22268417477607727, "learning_rate": 5.595649970290628e-05, "loss": 0.0189, "step": 48890 }, { "grad_norm": 0.2059318870306015, "learning_rate": 5.5940082428447426e-05, "loss": 0.0287, "step": 48900 }, { "grad_norm": 0.6930224299430847, "learning_rate": 5.592366450439012e-05, "loss": 0.0298, "step": 48910 }, { "grad_norm": 0.304275244474411, "learning_rate": 5.590724593252975e-05, "loss": 0.0294, "step": 48920 }, { "grad_norm": 0.2409793585538864, "learning_rate": 5.589082671466184e-05, "loss": 0.0286, "step": 48930 }, { "grad_norm": 0.3007207214832306, "learning_rate": 5.587440685258199e-05, "loss": 0.0328, "step": 48940 }, { "grad_norm": 0.5964112877845764, "learning_rate": 5.585798634808583e-05, "loss": 0.0345, "step": 48950 }, { "grad_norm": 0.3958895802497864, "learning_rate": 5.584156520296909e-05, "loss": 0.0345, "step": 48960 }, { "grad_norm": 0.3189370632171631, "learning_rate": 5.582514341902757e-05, "loss": 0.0291, "step": 48970 }, { "grad_norm": 0.2699219584465027, "learning_rate": 5.580872099805713e-05, "loss": 0.0327, "step": 48980 }, { "grad_norm": 0.4809788167476654, "learning_rate": 5.5792297941853674e-05, "loss": 0.0276, "step": 48990 }, { "grad_norm": 0.3509714901447296, "learning_rate": 5.5775874252213247e-05, "loss": 0.0299, "step": 49000 }, { "grad_norm": 0.3454830050468445, "learning_rate": 5.575944993093189e-05, "loss": 0.0286, "step": 49010 }, { "grad_norm": 0.3100604712963104, "learning_rate": 5.574302497980574e-05, "loss": 0.0259, "step": 49020 }, { "grad_norm": 0.2885773181915283, "learning_rate": 5.5726599400631005e-05, "loss": 0.0299, "step": 49030 }, { "grad_norm": 0.2884446084499359, "learning_rate": 5.571017319520399e-05, "loss": 0.0325, "step": 49040 }, { "grad_norm": 0.18441075086593628, "learning_rate": 5.5693746365320984e-05, "loss": 0.0252, "step": 49050 }, { "grad_norm": 0.37663519382476807, "learning_rate": 5.567731891277848e-05, "loss": 0.0254, "step": 49060 }, { "grad_norm": 0.38013795018196106, "learning_rate": 5.566089083937288e-05, "loss": 0.0332, "step": 49070 }, { "grad_norm": 0.31584230065345764, "learning_rate": 5.564446214690079e-05, "loss": 0.0296, "step": 49080 }, { "grad_norm": 0.22005827724933624, "learning_rate": 5.56280328371588e-05, "loss": 0.0248, "step": 49090 }, { "grad_norm": 0.3306129574775696, "learning_rate": 5.5611602911943596e-05, "loss": 0.0253, "step": 49100 }, { "grad_norm": 0.27487534284591675, "learning_rate": 5.559517237305194e-05, "loss": 0.0324, "step": 49110 }, { "grad_norm": 0.5727698802947998, "learning_rate": 5.557874122228064e-05, "loss": 0.0354, "step": 49120 }, { "grad_norm": 0.5240165591239929, "learning_rate": 5.55623094614266e-05, "loss": 0.0252, "step": 49130 }, { "grad_norm": 0.31602922081947327, "learning_rate": 5.5545877092286743e-05, "loss": 0.0287, "step": 49140 }, { "grad_norm": 0.28677695989608765, "learning_rate": 5.552944411665811e-05, "loss": 0.0233, "step": 49150 }, { "grad_norm": 0.44941115379333496, "learning_rate": 5.5513010536337795e-05, "loss": 0.0277, "step": 49160 }, { "grad_norm": 0.25813716650009155, "learning_rate": 5.5496576353122934e-05, "loss": 0.0253, "step": 49170 }, { "grad_norm": 0.38557133078575134, "learning_rate": 5.548014156881074e-05, "loss": 0.0265, "step": 49180 }, { "grad_norm": 0.24353201687335968, "learning_rate": 5.5463706185198494e-05, "loss": 0.0322, "step": 49190 }, { "grad_norm": 0.4116317629814148, "learning_rate": 5.544727020408358e-05, "loss": 0.0319, "step": 49200 }, { "grad_norm": 0.8955342769622803, "learning_rate": 5.5430833627263366e-05, "loss": 0.0343, "step": 49210 }, { "grad_norm": 0.24954482913017273, "learning_rate": 5.5414396456535354e-05, "loss": 0.0415, "step": 49220 }, { "grad_norm": 0.20910362899303436, "learning_rate": 5.539795869369709e-05, "loss": 0.0275, "step": 49230 }, { "grad_norm": 0.39186057448387146, "learning_rate": 5.538152034054618e-05, "loss": 0.0331, "step": 49240 }, { "grad_norm": 0.23286844789981842, "learning_rate": 5.536508139888028e-05, "loss": 0.026, "step": 49250 }, { "grad_norm": 0.3663317561149597, "learning_rate": 5.534864187049716e-05, "loss": 0.0346, "step": 49260 }, { "grad_norm": 0.2874506115913391, "learning_rate": 5.5332201757194615e-05, "loss": 0.0244, "step": 49270 }, { "grad_norm": 0.29248684644699097, "learning_rate": 5.531576106077049e-05, "loss": 0.0238, "step": 49280 }, { "grad_norm": 0.2062457948923111, "learning_rate": 5.529931978302272e-05, "loss": 0.0261, "step": 49290 }, { "grad_norm": 0.26771897077560425, "learning_rate": 5.528287792574932e-05, "loss": 0.0213, "step": 49300 }, { "grad_norm": 0.2663329243659973, "learning_rate": 5.5266435490748324e-05, "loss": 0.0268, "step": 49310 }, { "grad_norm": 0.23628319799900055, "learning_rate": 5.524999247981787e-05, "loss": 0.0261, "step": 49320 }, { "grad_norm": 0.29475751519203186, "learning_rate": 5.523354889475613e-05, "loss": 0.0275, "step": 49330 }, { "grad_norm": 0.3647278845310211, "learning_rate": 5.521710473736134e-05, "loss": 0.0314, "step": 49340 }, { "grad_norm": 0.22717024385929108, "learning_rate": 5.5200660009431826e-05, "loss": 0.0232, "step": 49350 }, { "grad_norm": 0.19936271011829376, "learning_rate": 5.518421471276596e-05, "loss": 0.0258, "step": 49360 }, { "grad_norm": 0.19760213792324066, "learning_rate": 5.5167768849162174e-05, "loss": 0.022, "step": 49370 }, { "grad_norm": 0.26695844531059265, "learning_rate": 5.515132242041893e-05, "loss": 0.0319, "step": 49380 }, { "grad_norm": 0.3221837282180786, "learning_rate": 5.513487542833483e-05, "loss": 0.0327, "step": 49390 }, { "grad_norm": 0.28923264145851135, "learning_rate": 5.5118427874708466e-05, "loss": 0.0245, "step": 49400 }, { "grad_norm": 0.3157689571380615, "learning_rate": 5.510197976133853e-05, "loss": 0.0245, "step": 49410 }, { "grad_norm": 0.24223476648330688, "learning_rate": 5.508553109002376e-05, "loss": 0.0201, "step": 49420 }, { "grad_norm": 0.29941287636756897, "learning_rate": 5.5069081862562957e-05, "loss": 0.0227, "step": 49430 }, { "grad_norm": 0.33492469787597656, "learning_rate": 5.5052632080754965e-05, "loss": 0.0259, "step": 49440 }, { "grad_norm": 0.22823724150657654, "learning_rate": 5.503618174639874e-05, "loss": 0.0211, "step": 49450 }, { "grad_norm": 0.4778278172016144, "learning_rate": 5.501973086129325e-05, "loss": 0.026, "step": 49460 }, { "grad_norm": 0.1775216907262802, "learning_rate": 5.500327942723753e-05, "loss": 0.0234, "step": 49470 }, { "grad_norm": 0.20218035578727722, "learning_rate": 5.498682744603071e-05, "loss": 0.0216, "step": 49480 }, { "grad_norm": 0.24840296804904938, "learning_rate": 5.497037491947194e-05, "loss": 0.024, "step": 49490 }, { "grad_norm": 0.2633209824562073, "learning_rate": 5.4953921849360424e-05, "loss": 0.0305, "step": 49500 }, { "grad_norm": 0.2200046181678772, "learning_rate": 5.493746823749547e-05, "loss": 0.0332, "step": 49510 }, { "grad_norm": 0.32143059372901917, "learning_rate": 5.49210140856764e-05, "loss": 0.0245, "step": 49520 }, { "grad_norm": 0.9113864898681641, "learning_rate": 5.4904559395702647e-05, "loss": 0.0319, "step": 49530 }, { "grad_norm": 0.3628729581832886, "learning_rate": 5.488810416937364e-05, "loss": 0.0267, "step": 49540 }, { "grad_norm": 0.4147409498691559, "learning_rate": 5.4871648408488926e-05, "loss": 0.0308, "step": 49550 }, { "grad_norm": 0.2032414972782135, "learning_rate": 5.485519211484807e-05, "loss": 0.0224, "step": 49560 }, { "grad_norm": 0.3580559194087982, "learning_rate": 5.4838735290250676e-05, "loss": 0.0297, "step": 49570 }, { "grad_norm": 0.2845054268836975, "learning_rate": 5.482227793649649e-05, "loss": 0.0329, "step": 49580 }, { "grad_norm": 0.25555965304374695, "learning_rate": 5.480582005538524e-05, "loss": 0.0255, "step": 49590 }, { "grad_norm": 0.27863165736198425, "learning_rate": 5.478936164871671e-05, "loss": 0.0323, "step": 49600 }, { "grad_norm": 0.260721355676651, "learning_rate": 5.477290271829081e-05, "loss": 0.0296, "step": 49610 }, { "grad_norm": 0.37872621417045593, "learning_rate": 5.475644326590744e-05, "loss": 0.0253, "step": 49620 }, { "grad_norm": 0.32822558283805847, "learning_rate": 5.473998329336658e-05, "loss": 0.0297, "step": 49630 }, { "grad_norm": 0.22661717236042023, "learning_rate": 5.4723522802468286e-05, "loss": 0.0225, "step": 49640 }, { "grad_norm": 0.235635444521904, "learning_rate": 5.470706179501264e-05, "loss": 0.0281, "step": 49650 }, { "grad_norm": 0.3089011609554291, "learning_rate": 5.4690600272799776e-05, "loss": 0.0283, "step": 49660 }, { "grad_norm": 0.38522401452064514, "learning_rate": 5.467413823762993e-05, "loss": 0.0315, "step": 49670 }, { "grad_norm": 0.2672646939754486, "learning_rate": 5.465767569130335e-05, "loss": 0.0292, "step": 49680 }, { "grad_norm": 0.2566545605659485, "learning_rate": 5.464121263562036e-05, "loss": 0.0259, "step": 49690 }, { "grad_norm": 0.16248586773872375, "learning_rate": 5.4624749072381343e-05, "loss": 0.0266, "step": 49700 }, { "grad_norm": 0.3187214136123657, "learning_rate": 5.460828500338672e-05, "loss": 0.0269, "step": 49710 }, { "grad_norm": 0.381744384765625, "learning_rate": 5.459182043043698e-05, "loss": 0.0355, "step": 49720 }, { "grad_norm": 0.26491644978523254, "learning_rate": 5.457535535533265e-05, "loss": 0.0217, "step": 49730 }, { "grad_norm": 0.3440778851509094, "learning_rate": 5.4558889779874334e-05, "loss": 0.0283, "step": 49740 }, { "grad_norm": 0.27769145369529724, "learning_rate": 5.454242370586269e-05, "loss": 0.0276, "step": 49750 }, { "grad_norm": 0.1907481551170349, "learning_rate": 5.452595713509843e-05, "loss": 0.0271, "step": 49760 }, { "grad_norm": 0.2516319751739502, "learning_rate": 5.4509490069382274e-05, "loss": 0.0288, "step": 49770 }, { "grad_norm": 0.2774985730648041, "learning_rate": 5.449302251051509e-05, "loss": 0.0266, "step": 49780 }, { "grad_norm": 0.22437752783298492, "learning_rate": 5.44765544602977e-05, "loss": 0.0244, "step": 49790 }, { "grad_norm": 0.4282563328742981, "learning_rate": 5.446008592053107e-05, "loss": 0.0296, "step": 49800 }, { "grad_norm": 0.47077059745788574, "learning_rate": 5.4443616893016135e-05, "loss": 0.0317, "step": 49810 }, { "grad_norm": 0.21487542986869812, "learning_rate": 5.4427147379553934e-05, "loss": 0.0289, "step": 49820 }, { "grad_norm": 0.5819941759109497, "learning_rate": 5.441067738194556e-05, "loss": 0.027, "step": 49830 }, { "grad_norm": 0.3409827649593353, "learning_rate": 5.439420690199214e-05, "loss": 0.0376, "step": 49840 }, { "grad_norm": 0.6915009021759033, "learning_rate": 5.437773594149484e-05, "loss": 0.03, "step": 49850 }, { "grad_norm": 0.2555111050605774, "learning_rate": 5.436126450225495e-05, "loss": 0.024, "step": 49860 }, { "grad_norm": 0.25007110834121704, "learning_rate": 5.434479258607371e-05, "loss": 0.0336, "step": 49870 }, { "grad_norm": 0.3708314597606659, "learning_rate": 5.43283201947525e-05, "loss": 0.0272, "step": 49880 }, { "grad_norm": 0.30597683787345886, "learning_rate": 5.431184733009268e-05, "loss": 0.0349, "step": 49890 }, { "grad_norm": 0.3223402500152588, "learning_rate": 5.4295373993895736e-05, "loss": 0.0255, "step": 49900 }, { "grad_norm": 0.6689013838768005, "learning_rate": 5.4278900187963157e-05, "loss": 0.0331, "step": 49910 }, { "grad_norm": 0.2204393744468689, "learning_rate": 5.426242591409646e-05, "loss": 0.0256, "step": 49920 }, { "grad_norm": 0.22498171031475067, "learning_rate": 5.42459511740973e-05, "loss": 0.0313, "step": 49930 }, { "grad_norm": 0.23328357934951782, "learning_rate": 5.422947596976729e-05, "loss": 0.0274, "step": 49940 }, { "grad_norm": 0.2997707426548004, "learning_rate": 5.4213000302908134e-05, "loss": 0.0349, "step": 49950 }, { "grad_norm": 0.24776500463485718, "learning_rate": 5.419652417532162e-05, "loss": 0.0223, "step": 49960 }, { "grad_norm": 0.31263867020606995, "learning_rate": 5.4180047588809534e-05, "loss": 0.0244, "step": 49970 }, { "grad_norm": 0.40008828043937683, "learning_rate": 5.4163570545173704e-05, "loss": 0.0262, "step": 49980 }, { "grad_norm": 0.24155019223690033, "learning_rate": 5.414709304621608e-05, "loss": 0.022, "step": 49990 }, { "grad_norm": 0.29884248971939087, "learning_rate": 5.413061509373858e-05, "loss": 0.0255, "step": 50000 }, { "grad_norm": 0.40311601758003235, "learning_rate": 5.411413668954324e-05, "loss": 0.0239, "step": 50010 }, { "grad_norm": 0.25736552476882935, "learning_rate": 5.4097657835432083e-05, "loss": 0.0219, "step": 50020 }, { "grad_norm": 0.2984789311885834, "learning_rate": 5.408117853320723e-05, "loss": 0.0286, "step": 50030 }, { "grad_norm": 0.21253132820129395, "learning_rate": 5.406469878467084e-05, "loss": 0.02, "step": 50040 }, { "grad_norm": 0.14830820262432098, "learning_rate": 5.404821859162509e-05, "loss": 0.0231, "step": 50050 }, { "grad_norm": 0.18057100474834442, "learning_rate": 5.403173795587225e-05, "loss": 0.0291, "step": 50060 }, { "grad_norm": 0.2777455151081085, "learning_rate": 5.4015256879214606e-05, "loss": 0.026, "step": 50070 }, { "grad_norm": 0.21533820033073425, "learning_rate": 5.39987753634545e-05, "loss": 0.0258, "step": 50080 }, { "grad_norm": 0.21152439713478088, "learning_rate": 5.3982293410394335e-05, "loss": 0.0235, "step": 50090 }, { "grad_norm": 0.4302290380001068, "learning_rate": 5.396581102183655e-05, "loss": 0.0239, "step": 50100 }, { "grad_norm": 0.2811957001686096, "learning_rate": 5.3949328199583634e-05, "loss": 0.0177, "step": 50110 }, { "grad_norm": 0.3516838550567627, "learning_rate": 5.393284494543813e-05, "loss": 0.0269, "step": 50120 }, { "grad_norm": 0.35248133540153503, "learning_rate": 5.391636126120262e-05, "loss": 0.0275, "step": 50130 }, { "grad_norm": 0.21624436974525452, "learning_rate": 5.389987714867971e-05, "loss": 0.0253, "step": 50140 }, { "grad_norm": 0.2626928985118866, "learning_rate": 5.3883392609672114e-05, "loss": 0.0256, "step": 50150 }, { "grad_norm": 0.26447537541389465, "learning_rate": 5.3866907645982545e-05, "loss": 0.0265, "step": 50160 }, { "grad_norm": 0.21833200752735138, "learning_rate": 5.385042225941375e-05, "loss": 0.0253, "step": 50170 }, { "grad_norm": 0.2982526123523712, "learning_rate": 5.3833936451768585e-05, "loss": 0.0212, "step": 50180 }, { "grad_norm": 0.32206180691719055, "learning_rate": 5.3817450224849884e-05, "loss": 0.0264, "step": 50190 }, { "grad_norm": 0.2473161816596985, "learning_rate": 5.380096358046056e-05, "loss": 0.0316, "step": 50200 }, { "grad_norm": 0.23711781203746796, "learning_rate": 5.378447652040359e-05, "loss": 0.0253, "step": 50210 }, { "grad_norm": 0.27053770422935486, "learning_rate": 5.376798904648195e-05, "loss": 0.0221, "step": 50220 }, { "grad_norm": 0.2342582494020462, "learning_rate": 5.375150116049869e-05, "loss": 0.0254, "step": 50230 }, { "grad_norm": 0.33751732110977173, "learning_rate": 5.373501286425691e-05, "loss": 0.0265, "step": 50240 }, { "grad_norm": 0.3236194849014282, "learning_rate": 5.3718524159559726e-05, "loss": 0.0248, "step": 50250 }, { "grad_norm": 0.24993294477462769, "learning_rate": 5.370203504821034e-05, "loss": 0.0217, "step": 50260 }, { "grad_norm": 0.17724229395389557, "learning_rate": 5.368554553201196e-05, "loss": 0.0197, "step": 50270 }, { "grad_norm": 0.3535228371620178, "learning_rate": 5.3669055612767874e-05, "loss": 0.0314, "step": 50280 }, { "grad_norm": 0.2525198459625244, "learning_rate": 5.3652565292281376e-05, "loss": 0.0261, "step": 50290 }, { "grad_norm": 0.3008333742618561, "learning_rate": 5.3636074572355824e-05, "loss": 0.0276, "step": 50300 }, { "grad_norm": 0.2701754868030548, "learning_rate": 5.361958345479463e-05, "loss": 0.0303, "step": 50310 }, { "grad_norm": 0.2962202727794647, "learning_rate": 5.360309194140123e-05, "loss": 0.0299, "step": 50320 }, { "grad_norm": 0.23660948872566223, "learning_rate": 5.3586600033979106e-05, "loss": 0.0217, "step": 50330 }, { "grad_norm": 0.222639799118042, "learning_rate": 5.357010773433181e-05, "loss": 0.0201, "step": 50340 }, { "grad_norm": 0.3246467113494873, "learning_rate": 5.3553615044262895e-05, "loss": 0.0254, "step": 50350 }, { "grad_norm": 0.27441030740737915, "learning_rate": 5.3537121965575964e-05, "loss": 0.022, "step": 50360 }, { "grad_norm": 0.3317601978778839, "learning_rate": 5.352062850007471e-05, "loss": 0.0255, "step": 50370 }, { "grad_norm": 0.3415776193141937, "learning_rate": 5.350413464956282e-05, "loss": 0.031, "step": 50380 }, { "grad_norm": 0.40511152148246765, "learning_rate": 5.348764041584403e-05, "loss": 0.0288, "step": 50390 }, { "grad_norm": 0.8295385837554932, "learning_rate": 5.347114580072212e-05, "loss": 0.0303, "step": 50400 }, { "grad_norm": 0.40393736958503723, "learning_rate": 5.345465080600093e-05, "loss": 0.0315, "step": 50410 }, { "grad_norm": 0.23477374017238617, "learning_rate": 5.3438155433484326e-05, "loss": 0.0233, "step": 50420 }, { "grad_norm": 0.2690977454185486, "learning_rate": 5.3421659684976197e-05, "loss": 0.0297, "step": 50430 }, { "grad_norm": 0.2747360169887543, "learning_rate": 5.340516356228052e-05, "loss": 0.0278, "step": 50440 }, { "grad_norm": 0.34779876470565796, "learning_rate": 5.338866706720128e-05, "loss": 0.0297, "step": 50450 }, { "grad_norm": 0.16686514019966125, "learning_rate": 5.337217020154249e-05, "loss": 0.0251, "step": 50460 }, { "grad_norm": 0.2607506513595581, "learning_rate": 5.335567296710825e-05, "loss": 0.0248, "step": 50470 }, { "grad_norm": 0.3078882694244385, "learning_rate": 5.333917536570265e-05, "loss": 0.0228, "step": 50480 }, { "grad_norm": 0.25830042362213135, "learning_rate": 5.332267739912986e-05, "loss": 0.0257, "step": 50490 }, { "grad_norm": 0.2440018504858017, "learning_rate": 5.330617906919405e-05, "loss": 0.0282, "step": 50500 }, { "grad_norm": 0.18328222632408142, "learning_rate": 5.328968037769949e-05, "loss": 0.0241, "step": 50510 }, { "grad_norm": 0.35947519540786743, "learning_rate": 5.327318132645042e-05, "loss": 0.0245, "step": 50520 }, { "grad_norm": 0.4055701494216919, "learning_rate": 5.325668191725116e-05, "loss": 0.0194, "step": 50530 }, { "grad_norm": 0.29294660687446594, "learning_rate": 5.324018215190606e-05, "loss": 0.0282, "step": 50540 }, { "grad_norm": 0.32971182465553284, "learning_rate": 5.3223682032219515e-05, "loss": 0.0247, "step": 50550 }, { "grad_norm": 0.2549819052219391, "learning_rate": 5.320718155999595e-05, "loss": 0.0222, "step": 50560 }, { "grad_norm": 0.31038698554039, "learning_rate": 5.3190680737039835e-05, "loss": 0.0266, "step": 50570 }, { "grad_norm": 0.29273346066474915, "learning_rate": 5.317417956515567e-05, "loss": 0.0281, "step": 50580 }, { "grad_norm": 0.29676181077957153, "learning_rate": 5.315767804614798e-05, "loss": 0.0238, "step": 50590 }, { "grad_norm": 0.3157370388507843, "learning_rate": 5.3141176181821395e-05, "loss": 0.0228, "step": 50600 }, { "grad_norm": 0.23917634785175323, "learning_rate": 5.31246739739805e-05, "loss": 0.0247, "step": 50610 }, { "grad_norm": 0.27227285504341125, "learning_rate": 5.310817142442994e-05, "loss": 0.0278, "step": 50620 }, { "grad_norm": 0.6600841879844666, "learning_rate": 5.309166853497445e-05, "loss": 0.0334, "step": 50630 }, { "grad_norm": 0.2665392756462097, "learning_rate": 5.307516530741873e-05, "loss": 0.0218, "step": 50640 }, { "grad_norm": 0.5048534870147705, "learning_rate": 5.305866174356754e-05, "loss": 0.0377, "step": 50650 }, { "grad_norm": 0.2640458047389984, "learning_rate": 5.304215784522571e-05, "loss": 0.0276, "step": 50660 }, { "grad_norm": 0.2728336453437805, "learning_rate": 5.302565361419808e-05, "loss": 0.0357, "step": 50670 }, { "grad_norm": 0.2641170024871826, "learning_rate": 5.3009149052289507e-05, "loss": 0.0333, "step": 50680 }, { "grad_norm": 0.2791822850704193, "learning_rate": 5.299264416130493e-05, "loss": 0.0312, "step": 50690 }, { "grad_norm": 0.31321027874946594, "learning_rate": 5.297613894304928e-05, "loss": 0.0295, "step": 50700 }, { "grad_norm": 0.3707423806190491, "learning_rate": 5.2959633399327534e-05, "loss": 0.0248, "step": 50710 }, { "grad_norm": 0.2722293436527252, "learning_rate": 5.294312753194476e-05, "loss": 0.0233, "step": 50720 }, { "grad_norm": 0.20587104558944702, "learning_rate": 5.292662134270596e-05, "loss": 0.028, "step": 50730 }, { "grad_norm": 1.4208495616912842, "learning_rate": 5.291011483341626e-05, "loss": 0.0392, "step": 50740 }, { "grad_norm": 0.33999964594841003, "learning_rate": 5.2893608005880767e-05, "loss": 0.0308, "step": 50750 }, { "grad_norm": 0.3345325291156769, "learning_rate": 5.287710086190467e-05, "loss": 0.0267, "step": 50760 }, { "grad_norm": 0.44105422496795654, "learning_rate": 5.286059340329314e-05, "loss": 0.0252, "step": 50770 }, { "grad_norm": 0.4140715003013611, "learning_rate": 5.284408563185141e-05, "loss": 0.0262, "step": 50780 }, { "grad_norm": 0.3068844974040985, "learning_rate": 5.282757754938476e-05, "loss": 0.0275, "step": 50790 }, { "grad_norm": 0.30582350492477417, "learning_rate": 5.281106915769849e-05, "loss": 0.0333, "step": 50800 }, { "grad_norm": 0.14268194139003754, "learning_rate": 5.2794560458597897e-05, "loss": 0.0219, "step": 50810 }, { "grad_norm": 0.5168461799621582, "learning_rate": 5.27780514538884e-05, "loss": 0.025, "step": 50820 }, { "grad_norm": 0.31104037165641785, "learning_rate": 5.2761542145375365e-05, "loss": 0.0228, "step": 50830 }, { "grad_norm": 0.23166917264461517, "learning_rate": 5.274503253486421e-05, "loss": 0.0227, "step": 50840 }, { "grad_norm": 0.21052861213684082, "learning_rate": 5.272852262416046e-05, "loss": 0.021, "step": 50850 }, { "grad_norm": 0.5767438411712646, "learning_rate": 5.2712012415069555e-05, "loss": 0.0282, "step": 50860 }, { "grad_norm": 0.438877671957016, "learning_rate": 5.269550190939705e-05, "loss": 0.0284, "step": 50870 }, { "grad_norm": 0.30891966819763184, "learning_rate": 5.267899110894852e-05, "loss": 0.028, "step": 50880 }, { "grad_norm": 0.16421057283878326, "learning_rate": 5.266248001552955e-05, "loss": 0.023, "step": 50890 }, { "grad_norm": 0.3104360103607178, "learning_rate": 5.264596863094575e-05, "loss": 0.0262, "step": 50900 }, { "grad_norm": 0.22431808710098267, "learning_rate": 5.262945695700282e-05, "loss": 0.0307, "step": 50910 }, { "grad_norm": 0.3988405466079712, "learning_rate": 5.261294499550643e-05, "loss": 0.033, "step": 50920 }, { "grad_norm": 0.25868940353393555, "learning_rate": 5.25964327482623e-05, "loss": 0.026, "step": 50930 }, { "grad_norm": 0.2421487271785736, "learning_rate": 5.257992021707617e-05, "loss": 0.0233, "step": 50940 }, { "grad_norm": 0.24868932366371155, "learning_rate": 5.256340740375387e-05, "loss": 0.0291, "step": 50950 }, { "grad_norm": 0.2630682587623596, "learning_rate": 5.254689431010117e-05, "loss": 0.0213, "step": 50960 }, { "grad_norm": 0.4453825056552887, "learning_rate": 5.253038093792395e-05, "loss": 0.0257, "step": 50970 }, { "grad_norm": 0.18174000084400177, "learning_rate": 5.251386728902806e-05, "loss": 0.0255, "step": 50980 }, { "grad_norm": 0.3165661692619324, "learning_rate": 5.2497353365219446e-05, "loss": 0.023, "step": 50990 }, { "grad_norm": 0.24549712240695953, "learning_rate": 5.2480839168304e-05, "loss": 0.0272, "step": 51000 }, { "grad_norm": 0.25053149461746216, "learning_rate": 5.246432470008773e-05, "loss": 0.028, "step": 51010 }, { "grad_norm": 0.20129269361495972, "learning_rate": 5.2447809962376616e-05, "loss": 0.0279, "step": 51020 }, { "grad_norm": 0.2581225335597992, "learning_rate": 5.243129495697666e-05, "loss": 0.0214, "step": 51030 }, { "grad_norm": 0.4738737940788269, "learning_rate": 5.241477968569397e-05, "loss": 0.0291, "step": 51040 }, { "grad_norm": 0.31553682684898376, "learning_rate": 5.23982641503346e-05, "loss": 0.0258, "step": 51050 }, { "grad_norm": 0.273123174905777, "learning_rate": 5.238174835270465e-05, "loss": 0.0294, "step": 51060 }, { "grad_norm": 0.3616613745689392, "learning_rate": 5.23652322946103e-05, "loss": 0.0254, "step": 51070 }, { "grad_norm": 0.20243823528289795, "learning_rate": 5.234871597785771e-05, "loss": 0.025, "step": 51080 }, { "grad_norm": 0.3492916524410248, "learning_rate": 5.233219940425306e-05, "loss": 0.0324, "step": 51090 }, { "grad_norm": 0.6423605680465698, "learning_rate": 5.2315682575602576e-05, "loss": 0.0438, "step": 51100 }, { "grad_norm": 0.36863309144973755, "learning_rate": 5.229916549371254e-05, "loss": 0.0253, "step": 51110 }, { "grad_norm": 0.2514468729496002, "learning_rate": 5.228264816038924e-05, "loss": 0.0232, "step": 51120 }, { "grad_norm": 0.5023626089096069, "learning_rate": 5.2266130577438935e-05, "loss": 0.0197, "step": 51130 }, { "grad_norm": 0.3454095721244812, "learning_rate": 5.224961274666802e-05, "loss": 0.028, "step": 51140 }, { "grad_norm": 0.38163766264915466, "learning_rate": 5.223309466988283e-05, "loss": 0.0347, "step": 51150 }, { "grad_norm": 0.6053348183631897, "learning_rate": 5.221657634888976e-05, "loss": 0.0342, "step": 51160 }, { "grad_norm": 0.24640730023384094, "learning_rate": 5.2200057785495235e-05, "loss": 0.0292, "step": 51170 }, { "grad_norm": 0.5239002108573914, "learning_rate": 5.2183538981505706e-05, "loss": 0.0293, "step": 51180 }, { "grad_norm": 0.2788885235786438, "learning_rate": 5.216701993872762e-05, "loss": 0.0216, "step": 51190 }, { "grad_norm": 0.2173335701227188, "learning_rate": 5.215050065896752e-05, "loss": 0.0313, "step": 51200 }, { "grad_norm": 0.33720535039901733, "learning_rate": 5.2133981144031875e-05, "loss": 0.0432, "step": 51210 }, { "grad_norm": 0.20058345794677734, "learning_rate": 5.211746139572727e-05, "loss": 0.0226, "step": 51220 }, { "grad_norm": 0.15709288418293, "learning_rate": 5.210094141586026e-05, "loss": 0.031, "step": 51230 }, { "grad_norm": 0.22429881989955902, "learning_rate": 5.2084421206237476e-05, "loss": 0.0287, "step": 51240 }, { "grad_norm": 0.23874635994434357, "learning_rate": 5.20679007686655e-05, "loss": 0.0269, "step": 51250 }, { "grad_norm": 0.24679818749427795, "learning_rate": 5.205138010495102e-05, "loss": 0.0253, "step": 51260 }, { "grad_norm": 0.336124449968338, "learning_rate": 5.20348592169007e-05, "loss": 0.0254, "step": 51270 }, { "grad_norm": 0.21767571568489075, "learning_rate": 5.201833810632123e-05, "loss": 0.0264, "step": 51280 }, { "grad_norm": 0.29924464225769043, "learning_rate": 5.200181677501933e-05, "loss": 0.0238, "step": 51290 }, { "grad_norm": 0.2856455147266388, "learning_rate": 5.1985295224801777e-05, "loss": 0.029, "step": 51300 }, { "grad_norm": 0.33877259492874146, "learning_rate": 5.196877345747531e-05, "loss": 0.0232, "step": 51310 }, { "grad_norm": 0.2161029726266861, "learning_rate": 5.195225147484674e-05, "loss": 0.0215, "step": 51320 }, { "grad_norm": 0.26668378710746765, "learning_rate": 5.193572927872288e-05, "loss": 0.0293, "step": 51330 }, { "grad_norm": 0.31689420342445374, "learning_rate": 5.191920687091059e-05, "loss": 0.0324, "step": 51340 }, { "grad_norm": 0.7513831257820129, "learning_rate": 5.1902684253216695e-05, "loss": 0.0317, "step": 51350 }, { "grad_norm": 0.37185487151145935, "learning_rate": 5.1886161427448134e-05, "loss": 0.0272, "step": 51360 }, { "grad_norm": 0.36622411012649536, "learning_rate": 5.186963839541179e-05, "loss": 0.0259, "step": 51370 }, { "grad_norm": 0.26742079854011536, "learning_rate": 5.185311515891459e-05, "loss": 0.0281, "step": 51380 }, { "grad_norm": 0.20841963589191437, "learning_rate": 5.1836591719763516e-05, "loss": 0.025, "step": 51390 }, { "grad_norm": 0.35367926955223083, "learning_rate": 5.182006807976554e-05, "loss": 0.024, "step": 51400 }, { "grad_norm": 0.4370170533657074, "learning_rate": 5.180354424072763e-05, "loss": 0.0306, "step": 51410 }, { "grad_norm": 0.2554035782814026, "learning_rate": 5.178702020445686e-05, "loss": 0.0267, "step": 51420 }, { "grad_norm": 0.26915520429611206, "learning_rate": 5.177049597276024e-05, "loss": 0.0251, "step": 51430 }, { "grad_norm": 0.23932023346424103, "learning_rate": 5.175397154744482e-05, "loss": 0.03, "step": 51440 }, { "grad_norm": 0.37065866589546204, "learning_rate": 5.1737446930317736e-05, "loss": 0.0261, "step": 51450 }, { "grad_norm": 0.29685020446777344, "learning_rate": 5.172092212318604e-05, "loss": 0.0302, "step": 51460 }, { "grad_norm": 0.3642536401748657, "learning_rate": 5.170439712785692e-05, "loss": 0.0265, "step": 51470 }, { "grad_norm": 0.3122626841068268, "learning_rate": 5.1687871946137467e-05, "loss": 0.0267, "step": 51480 }, { "grad_norm": 1.4966884851455688, "learning_rate": 5.16713465798349e-05, "loss": 0.0284, "step": 51490 }, { "grad_norm": 0.6183701157569885, "learning_rate": 5.165482103075637e-05, "loss": 0.0323, "step": 51500 }, { "grad_norm": 0.32001373171806335, "learning_rate": 5.163829530070909e-05, "loss": 0.0345, "step": 51510 }, { "grad_norm": 0.42161625623703003, "learning_rate": 5.162176939150032e-05, "loss": 0.0315, "step": 51520 }, { "grad_norm": 0.2563427984714508, "learning_rate": 5.160524330493728e-05, "loss": 0.0215, "step": 51530 }, { "grad_norm": 0.18225915729999542, "learning_rate": 5.158871704282724e-05, "loss": 0.0193, "step": 51540 }, { "grad_norm": 0.2048635482788086, "learning_rate": 5.15721906069775e-05, "loss": 0.0217, "step": 51550 }, { "grad_norm": 0.20155714452266693, "learning_rate": 5.155566399919536e-05, "loss": 0.0257, "step": 51560 }, { "grad_norm": 0.2763381004333496, "learning_rate": 5.153913722128813e-05, "loss": 0.03, "step": 51570 }, { "grad_norm": 0.2042502760887146, "learning_rate": 5.1522610275063176e-05, "loss": 0.0229, "step": 51580 }, { "grad_norm": 0.2734319269657135, "learning_rate": 5.150608316232786e-05, "loss": 0.0218, "step": 51590 }, { "grad_norm": 0.29525983333587646, "learning_rate": 5.148955588488955e-05, "loss": 0.0181, "step": 51600 }, { "grad_norm": 0.4664643108844757, "learning_rate": 5.147302844455564e-05, "loss": 0.0222, "step": 51610 }, { "grad_norm": 0.34340783953666687, "learning_rate": 5.145650084313356e-05, "loss": 0.024, "step": 51620 }, { "grad_norm": 0.2569911479949951, "learning_rate": 5.1439973082430734e-05, "loss": 0.0182, "step": 51630 }, { "grad_norm": 0.24826838076114655, "learning_rate": 5.142344516425461e-05, "loss": 0.0227, "step": 51640 }, { "grad_norm": 0.8077351450920105, "learning_rate": 5.1406917090412673e-05, "loss": 0.0303, "step": 51650 }, { "grad_norm": 0.35927510261535645, "learning_rate": 5.1390388862712404e-05, "loss": 0.0294, "step": 51660 }, { "grad_norm": 0.30745741724967957, "learning_rate": 5.137386048296129e-05, "loss": 0.0252, "step": 51670 }, { "grad_norm": 0.36826807260513306, "learning_rate": 5.135733195296686e-05, "loss": 0.0246, "step": 51680 }, { "grad_norm": 0.3915417492389679, "learning_rate": 5.134080327453664e-05, "loss": 0.0249, "step": 51690 }, { "grad_norm": 0.28116029500961304, "learning_rate": 5.1324274449478204e-05, "loss": 0.0233, "step": 51700 }, { "grad_norm": 0.262207567691803, "learning_rate": 5.130774547959909e-05, "loss": 0.0282, "step": 51710 }, { "grad_norm": 0.3071041703224182, "learning_rate": 5.129121636670692e-05, "loss": 0.0257, "step": 51720 }, { "grad_norm": 0.21551388502120972, "learning_rate": 5.1274687112609246e-05, "loss": 0.0213, "step": 51730 }, { "grad_norm": 0.39214634895324707, "learning_rate": 5.125815771911372e-05, "loss": 0.026, "step": 51740 }, { "grad_norm": 0.30922433733940125, "learning_rate": 5.1241628188027956e-05, "loss": 0.0227, "step": 51750 }, { "grad_norm": 0.2487039864063263, "learning_rate": 5.1225098521159585e-05, "loss": 0.0205, "step": 51760 }, { "grad_norm": 0.27225813269615173, "learning_rate": 5.12085687203163e-05, "loss": 0.0231, "step": 51770 }, { "grad_norm": 0.18112677335739136, "learning_rate": 5.119203878730575e-05, "loss": 0.0189, "step": 51780 }, { "grad_norm": 0.20982828736305237, "learning_rate": 5.117550872393564e-05, "loss": 0.0242, "step": 51790 }, { "grad_norm": 0.20964014530181885, "learning_rate": 5.1158978532013645e-05, "loss": 0.0225, "step": 51800 }, { "grad_norm": 0.27664855122566223, "learning_rate": 5.114244821334751e-05, "loss": 0.0255, "step": 51810 }, { "grad_norm": 0.3155674636363983, "learning_rate": 5.112591776974497e-05, "loss": 0.0249, "step": 51820 }, { "grad_norm": 0.24683307111263275, "learning_rate": 5.110938720301374e-05, "loss": 0.0349, "step": 51830 }, { "grad_norm": 0.17957811057567596, "learning_rate": 5.1092856514961595e-05, "loss": 0.0284, "step": 51840 }, { "grad_norm": 0.41836073994636536, "learning_rate": 5.107632570739631e-05, "loss": 0.0308, "step": 51850 }, { "grad_norm": 0.23356260359287262, "learning_rate": 5.1059794782125645e-05, "loss": 0.0274, "step": 51860 }, { "grad_norm": 0.4437057673931122, "learning_rate": 5.104326374095744e-05, "loss": 0.0347, "step": 51870 }, { "grad_norm": 0.29333633184432983, "learning_rate": 5.102673258569947e-05, "loss": 0.0254, "step": 51880 }, { "grad_norm": 0.2974103093147278, "learning_rate": 5.101020131815956e-05, "loss": 0.0291, "step": 51890 }, { "grad_norm": 0.22997614741325378, "learning_rate": 5.099366994014556e-05, "loss": 0.022, "step": 51900 }, { "grad_norm": 0.5382049679756165, "learning_rate": 5.097713845346532e-05, "loss": 0.0395, "step": 51910 }, { "grad_norm": 0.2669340968132019, "learning_rate": 5.096060685992668e-05, "loss": 0.0231, "step": 51920 }, { "grad_norm": 0.29063016176223755, "learning_rate": 5.0944075161337514e-05, "loss": 0.0231, "step": 51930 }, { "grad_norm": 0.20189112424850464, "learning_rate": 5.09275433595057e-05, "loss": 0.022, "step": 51940 }, { "grad_norm": 0.342283695936203, "learning_rate": 5.0911011456239157e-05, "loss": 0.0278, "step": 51950 }, { "grad_norm": 0.4393146336078644, "learning_rate": 5.089447945334574e-05, "loss": 0.0255, "step": 51960 }, { "grad_norm": 0.2684921324253082, "learning_rate": 5.087794735263343e-05, "loss": 0.0234, "step": 51970 }, { "grad_norm": 0.26958051323890686, "learning_rate": 5.086141515591011e-05, "loss": 0.0252, "step": 51980 }, { "grad_norm": 0.348280668258667, "learning_rate": 5.084488286498371e-05, "loss": 0.0275, "step": 51990 }, { "grad_norm": 0.29861441254615784, "learning_rate": 5.08283504816622e-05, "loss": 0.0286, "step": 52000 }, { "grad_norm": 0.296968013048172, "learning_rate": 5.081181800775353e-05, "loss": 0.0313, "step": 52010 }, { "grad_norm": 0.2427283525466919, "learning_rate": 5.079528544506565e-05, "loss": 0.0231, "step": 52020 }, { "grad_norm": 0.3098490536212921, "learning_rate": 5.077875279540657e-05, "loss": 0.0267, "step": 52030 }, { "grad_norm": 0.26371094584465027, "learning_rate": 5.076222006058424e-05, "loss": 0.03, "step": 52040 }, { "grad_norm": 0.32854995131492615, "learning_rate": 5.074568724240667e-05, "loss": 0.0272, "step": 52050 }, { "grad_norm": 0.21736885607242584, "learning_rate": 5.072915434268187e-05, "loss": 0.0257, "step": 52060 }, { "grad_norm": 0.2755344808101654, "learning_rate": 5.071262136321785e-05, "loss": 0.0286, "step": 52070 }, { "grad_norm": 0.1784265637397766, "learning_rate": 5.069608830582262e-05, "loss": 0.0234, "step": 52080 }, { "grad_norm": 0.4503851532936096, "learning_rate": 5.067955517230423e-05, "loss": 0.0275, "step": 52090 }, { "grad_norm": 0.3552623689174652, "learning_rate": 5.06630219644707e-05, "loss": 0.0336, "step": 52100 }, { "grad_norm": 0.30553802847862244, "learning_rate": 5.064648868413008e-05, "loss": 0.0233, "step": 52110 }, { "grad_norm": 0.17389683425426483, "learning_rate": 5.062995533309044e-05, "loss": 0.0223, "step": 52120 }, { "grad_norm": 0.3722037971019745, "learning_rate": 5.061342191315983e-05, "loss": 0.0376, "step": 52130 }, { "grad_norm": 0.23915787041187286, "learning_rate": 5.0596888426146325e-05, "loss": 0.0295, "step": 52140 }, { "grad_norm": 0.30386844277381897, "learning_rate": 5.058035487385798e-05, "loss": 0.0211, "step": 52150 }, { "grad_norm": 0.24100126326084137, "learning_rate": 5.0563821258102915e-05, "loss": 0.0226, "step": 52160 }, { "grad_norm": 0.3348242938518524, "learning_rate": 5.05472875806892e-05, "loss": 0.0195, "step": 52170 }, { "grad_norm": 0.27862703800201416, "learning_rate": 5.053075384342494e-05, "loss": 0.0266, "step": 52180 }, { "grad_norm": 0.21805526316165924, "learning_rate": 5.0514220048118234e-05, "loss": 0.0257, "step": 52190 }, { "grad_norm": 0.27543824911117554, "learning_rate": 5.049768619657721e-05, "loss": 0.0307, "step": 52200 }, { "grad_norm": 0.15938971936702728, "learning_rate": 5.048115229060996e-05, "loss": 0.0172, "step": 52210 }, { "grad_norm": 0.29854610562324524, "learning_rate": 5.046461833202462e-05, "loss": 0.0307, "step": 52220 }, { "grad_norm": 0.2238595187664032, "learning_rate": 5.044808432262932e-05, "loss": 0.019, "step": 52230 }, { "grad_norm": 0.2575317323207855, "learning_rate": 5.043155026423219e-05, "loss": 0.0251, "step": 52240 }, { "grad_norm": 0.18872816860675812, "learning_rate": 5.041501615864137e-05, "loss": 0.0211, "step": 52250 }, { "grad_norm": 0.2814441919326782, "learning_rate": 5.039848200766502e-05, "loss": 0.0249, "step": 52260 }, { "grad_norm": 0.31145021319389343, "learning_rate": 5.038194781311125e-05, "loss": 0.0261, "step": 52270 }, { "grad_norm": 0.7442765235900879, "learning_rate": 5.0365413576788264e-05, "loss": 0.0215, "step": 52280 }, { "grad_norm": 0.16737580299377441, "learning_rate": 5.0348879300504194e-05, "loss": 0.0194, "step": 52290 }, { "grad_norm": 0.3277972638607025, "learning_rate": 5.033234498606721e-05, "loss": 0.0241, "step": 52300 }, { "grad_norm": 0.30325427651405334, "learning_rate": 5.0315810635285476e-05, "loss": 0.0235, "step": 52310 }, { "grad_norm": 0.314862996339798, "learning_rate": 5.0299276249967164e-05, "loss": 0.0223, "step": 52320 }, { "grad_norm": 0.288274884223938, "learning_rate": 5.0282741831920454e-05, "loss": 0.0294, "step": 52330 }, { "grad_norm": 0.3860117197036743, "learning_rate": 5.026620738295351e-05, "loss": 0.0294, "step": 52340 }, { "grad_norm": 0.3177035450935364, "learning_rate": 5.024967290487455e-05, "loss": 0.0174, "step": 52350 }, { "grad_norm": 0.281823068857193, "learning_rate": 5.023313839949172e-05, "loss": 0.0228, "step": 52360 }, { "grad_norm": 0.28406789898872375, "learning_rate": 5.0216603868613224e-05, "loss": 0.0261, "step": 52370 }, { "grad_norm": 0.3259580433368683, "learning_rate": 5.020006931404726e-05, "loss": 0.0321, "step": 52380 }, { "grad_norm": 0.2741295099258423, "learning_rate": 5.0183534737602014e-05, "loss": 0.0242, "step": 52390 }, { "grad_norm": 0.33317652344703674, "learning_rate": 5.0167000141085686e-05, "loss": 0.0248, "step": 52400 }, { "grad_norm": 0.3238459527492523, "learning_rate": 5.015046552630647e-05, "loss": 0.0229, "step": 52410 }, { "grad_norm": 0.2969243824481964, "learning_rate": 5.013393089507258e-05, "loss": 0.0262, "step": 52420 }, { "grad_norm": 0.16193313896656036, "learning_rate": 5.011739624919219e-05, "loss": 0.0228, "step": 52430 }, { "grad_norm": 0.2360880970954895, "learning_rate": 5.0100861590473545e-05, "loss": 0.0233, "step": 52440 }, { "grad_norm": 0.3276991546154022, "learning_rate": 5.008432692072482e-05, "loss": 0.024, "step": 52450 }, { "grad_norm": 0.22849120199680328, "learning_rate": 5.006779224175423e-05, "loss": 0.0269, "step": 52460 }, { "grad_norm": 0.2264799028635025, "learning_rate": 5.005125755536998e-05, "loss": 0.022, "step": 52470 }, { "grad_norm": 0.4888931214809418, "learning_rate": 5.003472286338029e-05, "loss": 0.0246, "step": 52480 }, { "grad_norm": 0.36011144518852234, "learning_rate": 5.001818816759335e-05, "loss": 0.0433, "step": 52490 }, { "grad_norm": 0.30961164832115173, "learning_rate": 5.000165346981738e-05, "loss": 0.0259, "step": 52500 }, { "grad_norm": 0.24496908485889435, "learning_rate": 4.9985118771860595e-05, "loss": 0.0214, "step": 52510 }, { "grad_norm": 0.2728715240955353, "learning_rate": 4.996858407553119e-05, "loss": 0.0247, "step": 52520 }, { "grad_norm": 0.3105597198009491, "learning_rate": 4.99520493826374e-05, "loss": 0.0266, "step": 52530 }, { "grad_norm": 0.2755359709262848, "learning_rate": 4.9935514694987384e-05, "loss": 0.0251, "step": 52540 }, { "grad_norm": 0.4514990448951721, "learning_rate": 4.991898001438941e-05, "loss": 0.0272, "step": 52550 }, { "grad_norm": 0.22610312700271606, "learning_rate": 4.990244534265166e-05, "loss": 0.0243, "step": 52560 }, { "grad_norm": 0.23226991295814514, "learning_rate": 4.988591068158234e-05, "loss": 0.0257, "step": 52570 }, { "grad_norm": 0.2847628891468048, "learning_rate": 4.986937603298965e-05, "loss": 0.0226, "step": 52580 }, { "grad_norm": 0.2486243098974228, "learning_rate": 4.985284139868181e-05, "loss": 0.0269, "step": 52590 }, { "grad_norm": 0.21462248265743256, "learning_rate": 4.9836306780467e-05, "loss": 0.0196, "step": 52600 }, { "grad_norm": 0.2513958215713501, "learning_rate": 4.981977218015343e-05, "loss": 0.0248, "step": 52610 }, { "grad_norm": 0.24045006930828094, "learning_rate": 4.980323759954932e-05, "loss": 0.0221, "step": 52620 }, { "grad_norm": 0.17556704580783844, "learning_rate": 4.978670304046284e-05, "loss": 0.0247, "step": 52630 }, { "grad_norm": 0.3078168034553528, "learning_rate": 4.9770168504702195e-05, "loss": 0.0318, "step": 52640 }, { "grad_norm": 0.2899329662322998, "learning_rate": 4.975363399407559e-05, "loss": 0.0342, "step": 52650 }, { "grad_norm": 0.1992940455675125, "learning_rate": 4.9737099510391194e-05, "loss": 0.0268, "step": 52660 }, { "grad_norm": 0.22187106311321259, "learning_rate": 4.972056505545719e-05, "loss": 0.0271, "step": 52670 }, { "grad_norm": 0.7040793895721436, "learning_rate": 4.9704030631081775e-05, "loss": 0.0229, "step": 52680 }, { "grad_norm": 0.33084967732429504, "learning_rate": 4.968749623907313e-05, "loss": 0.0238, "step": 52690 }, { "grad_norm": 0.38070428371429443, "learning_rate": 4.967096188123943e-05, "loss": 0.0279, "step": 52700 }, { "grad_norm": 0.6042217016220093, "learning_rate": 4.965442755938884e-05, "loss": 0.0237, "step": 52710 }, { "grad_norm": 0.37559157609939575, "learning_rate": 4.963789327532954e-05, "loss": 0.0224, "step": 52720 }, { "grad_norm": 0.37279459834098816, "learning_rate": 4.962135903086968e-05, "loss": 0.0215, "step": 52730 }, { "grad_norm": 0.48405173420906067, "learning_rate": 4.960482482781741e-05, "loss": 0.0192, "step": 52740 }, { "grad_norm": 0.19084681570529938, "learning_rate": 4.958829066798092e-05, "loss": 0.0245, "step": 52750 }, { "grad_norm": 0.21164724230766296, "learning_rate": 4.957175655316835e-05, "loss": 0.021, "step": 52760 }, { "grad_norm": 0.3029460310935974, "learning_rate": 4.9555222485187834e-05, "loss": 0.0298, "step": 52770 }, { "grad_norm": 0.17160214483737946, "learning_rate": 4.953868846584752e-05, "loss": 0.0201, "step": 52780 }, { "grad_norm": 0.29399892687797546, "learning_rate": 4.9522154496955544e-05, "loss": 0.0315, "step": 52790 }, { "grad_norm": 0.6448541283607483, "learning_rate": 4.950562058032001e-05, "loss": 0.0268, "step": 52800 }, { "grad_norm": 0.324796199798584, "learning_rate": 4.948908671774909e-05, "loss": 0.0249, "step": 52810 }, { "grad_norm": 0.1852339208126068, "learning_rate": 4.947255291105087e-05, "loss": 0.0244, "step": 52820 }, { "grad_norm": 0.3304572105407715, "learning_rate": 4.945601916203348e-05, "loss": 0.0284, "step": 52830 }, { "grad_norm": 0.3584371507167816, "learning_rate": 4.9439485472505016e-05, "loss": 0.027, "step": 52840 }, { "grad_norm": 0.23919397592544556, "learning_rate": 4.942295184427356e-05, "loss": 0.0258, "step": 52850 }, { "grad_norm": 0.41427749395370483, "learning_rate": 4.9406418279147234e-05, "loss": 0.0281, "step": 52860 }, { "grad_norm": 0.3165811002254486, "learning_rate": 4.9389884778934116e-05, "loss": 0.0331, "step": 52870 }, { "grad_norm": 0.36671310663223267, "learning_rate": 4.937335134544228e-05, "loss": 0.0321, "step": 52880 }, { "grad_norm": 0.27985629439353943, "learning_rate": 4.935681798047979e-05, "loss": 0.0271, "step": 52890 }, { "grad_norm": 0.3208126127719879, "learning_rate": 4.934028468585473e-05, "loss": 0.0284, "step": 52900 }, { "grad_norm": 0.21817803382873535, "learning_rate": 4.932375146337514e-05, "loss": 0.0262, "step": 52910 }, { "grad_norm": 0.30963775515556335, "learning_rate": 4.930721831484908e-05, "loss": 0.0258, "step": 52920 }, { "grad_norm": 0.3196706175804138, "learning_rate": 4.929068524208456e-05, "loss": 0.0249, "step": 52930 }, { "grad_norm": 0.23615337908267975, "learning_rate": 4.927415224688965e-05, "loss": 0.0329, "step": 52940 }, { "grad_norm": 0.16353575885295868, "learning_rate": 4.9257619331072375e-05, "loss": 0.0299, "step": 52950 }, { "grad_norm": 0.27850690484046936, "learning_rate": 4.924108649644072e-05, "loss": 0.0187, "step": 52960 }, { "grad_norm": 0.33470073342323303, "learning_rate": 4.922455374480271e-05, "loss": 0.0248, "step": 52970 }, { "grad_norm": 0.2359745353460312, "learning_rate": 4.920802107796634e-05, "loss": 0.0275, "step": 52980 }, { "grad_norm": 0.17928937077522278, "learning_rate": 4.9191488497739564e-05, "loss": 0.0243, "step": 52990 }, { "grad_norm": 0.3168119788169861, "learning_rate": 4.917495600593042e-05, "loss": 0.0245, "step": 53000 }, { "grad_norm": 0.28855687379837036, "learning_rate": 4.915842360434684e-05, "loss": 0.0247, "step": 53010 }, { "grad_norm": 0.40103450417518616, "learning_rate": 4.914189129479679e-05, "loss": 0.0337, "step": 53020 }, { "grad_norm": 0.3361178934574127, "learning_rate": 4.912535907908822e-05, "loss": 0.0226, "step": 53030 }, { "grad_norm": 0.2827274799346924, "learning_rate": 4.910882695902907e-05, "loss": 0.027, "step": 53040 }, { "grad_norm": 0.2079702764749527, "learning_rate": 4.9092294936427234e-05, "loss": 0.0313, "step": 53050 }, { "grad_norm": 0.2695930600166321, "learning_rate": 4.9075763013090685e-05, "loss": 0.0258, "step": 53060 }, { "grad_norm": 0.3897058069705963, "learning_rate": 4.9059231190827305e-05, "loss": 0.0321, "step": 53070 }, { "grad_norm": 0.24908268451690674, "learning_rate": 4.904269947144499e-05, "loss": 0.0393, "step": 53080 }, { "grad_norm": 0.2827737629413605, "learning_rate": 4.902616785675161e-05, "loss": 0.0279, "step": 53090 }, { "grad_norm": 0.31054097414016724, "learning_rate": 4.900963634855505e-05, "loss": 0.034, "step": 53100 }, { "grad_norm": 0.316638320684433, "learning_rate": 4.8993104948663173e-05, "loss": 0.022, "step": 53110 }, { "grad_norm": 0.8539215326309204, "learning_rate": 4.8976573658883823e-05, "loss": 0.0275, "step": 53120 }, { "grad_norm": 0.44279369711875916, "learning_rate": 4.896004248102484e-05, "loss": 0.0259, "step": 53130 }, { "grad_norm": 0.5839421153068542, "learning_rate": 4.8943511416894045e-05, "loss": 0.0286, "step": 53140 }, { "grad_norm": 0.3568513095378876, "learning_rate": 4.892698046829926e-05, "loss": 0.0239, "step": 53150 }, { "grad_norm": 0.27945777773857117, "learning_rate": 4.8910449637048275e-05, "loss": 0.0371, "step": 53160 }, { "grad_norm": 0.4055232107639313, "learning_rate": 4.8893918924948884e-05, "loss": 0.0311, "step": 53170 }, { "grad_norm": 0.521965742111206, "learning_rate": 4.8877388333808835e-05, "loss": 0.0296, "step": 53180 }, { "grad_norm": 0.31768813729286194, "learning_rate": 4.8860857865435925e-05, "loss": 0.0223, "step": 53190 }, { "grad_norm": 0.27683544158935547, "learning_rate": 4.884432752163789e-05, "loss": 0.0255, "step": 53200 }, { "grad_norm": 0.4571533501148224, "learning_rate": 4.8827797304222466e-05, "loss": 0.0229, "step": 53210 }, { "grad_norm": 0.41824063658714294, "learning_rate": 4.8811267214997366e-05, "loss": 0.0229, "step": 53220 }, { "grad_norm": 0.31377077102661133, "learning_rate": 4.879473725577029e-05, "loss": 0.0251, "step": 53230 }, { "grad_norm": 0.22520406544208527, "learning_rate": 4.877820742834891e-05, "loss": 0.0213, "step": 53240 }, { "grad_norm": 0.4884343445301056, "learning_rate": 4.8761677734540956e-05, "loss": 0.0255, "step": 53250 }, { "grad_norm": 0.3503408432006836, "learning_rate": 4.874514817615407e-05, "loss": 0.0239, "step": 53260 }, { "grad_norm": 0.22326244413852692, "learning_rate": 4.872861875499588e-05, "loss": 0.0183, "step": 53270 }, { "grad_norm": 0.26405298709869385, "learning_rate": 4.871208947287404e-05, "loss": 0.0248, "step": 53280 }, { "grad_norm": 0.6303079128265381, "learning_rate": 4.869556033159615e-05, "loss": 0.0316, "step": 53290 }, { "grad_norm": 0.34564605355262756, "learning_rate": 4.8679031332969835e-05, "loss": 0.0241, "step": 53300 }, { "grad_norm": 0.18819595873355865, "learning_rate": 4.8662502478802635e-05, "loss": 0.0311, "step": 53310 }, { "grad_norm": 0.33358991146087646, "learning_rate": 4.8645973770902176e-05, "loss": 0.0205, "step": 53320 }, { "grad_norm": 0.21016483008861542, "learning_rate": 4.862944521107599e-05, "loss": 0.0204, "step": 53330 }, { "grad_norm": 0.38829129934310913, "learning_rate": 4.861291680113162e-05, "loss": 0.0244, "step": 53340 }, { "grad_norm": 0.2568191885948181, "learning_rate": 4.8596388542876563e-05, "loss": 0.0209, "step": 53350 }, { "grad_norm": 0.3545586168766022, "learning_rate": 4.857986043811836e-05, "loss": 0.0202, "step": 53360 }, { "grad_norm": 0.20985235273838043, "learning_rate": 4.856333248866448e-05, "loss": 0.0206, "step": 53370 }, { "grad_norm": 0.1526196002960205, "learning_rate": 4.85468046963224e-05, "loss": 0.022, "step": 53380 }, { "grad_norm": 0.33461254835128784, "learning_rate": 4.853027706289957e-05, "loss": 0.0297, "step": 53390 }, { "grad_norm": 0.3923158049583435, "learning_rate": 4.851374959020344e-05, "loss": 0.0214, "step": 53400 }, { "grad_norm": 0.23035837709903717, "learning_rate": 4.849722228004141e-05, "loss": 0.0291, "step": 53410 }, { "grad_norm": 0.3457166254520416, "learning_rate": 4.848069513422091e-05, "loss": 0.0229, "step": 53420 }, { "grad_norm": 0.23872816562652588, "learning_rate": 4.8464168154549294e-05, "loss": 0.0237, "step": 53430 }, { "grad_norm": 0.3290480673313141, "learning_rate": 4.844764134283392e-05, "loss": 0.0201, "step": 53440 }, { "grad_norm": 0.22713342308998108, "learning_rate": 4.843111470088218e-05, "loss": 0.0246, "step": 53450 }, { "grad_norm": 0.2287740260362625, "learning_rate": 4.841458823050137e-05, "loss": 0.0224, "step": 53460 }, { "grad_norm": 0.406841516494751, "learning_rate": 4.8398061933498816e-05, "loss": 0.0229, "step": 53470 }, { "grad_norm": 0.300326406955719, "learning_rate": 4.83815358116818e-05, "loss": 0.023, "step": 53480 }, { "grad_norm": 0.2758144438266754, "learning_rate": 4.8365009866857584e-05, "loss": 0.0342, "step": 53490 }, { "grad_norm": 0.2626221179962158, "learning_rate": 4.834848410083342e-05, "loss": 0.0254, "step": 53500 }, { "grad_norm": 0.3063908815383911, "learning_rate": 4.833195851541657e-05, "loss": 0.0212, "step": 53510 }, { "grad_norm": 0.257406622171402, "learning_rate": 4.831543311241423e-05, "loss": 0.0225, "step": 53520 }, { "grad_norm": 0.5196455121040344, "learning_rate": 4.829890789363359e-05, "loss": 0.0355, "step": 53530 }, { "grad_norm": 0.3528517484664917, "learning_rate": 4.828238286088182e-05, "loss": 0.0241, "step": 53540 }, { "grad_norm": 0.23399561643600464, "learning_rate": 4.826585801596609e-05, "loss": 0.0254, "step": 53550 }, { "grad_norm": 0.2339276373386383, "learning_rate": 4.8249333360693486e-05, "loss": 0.025, "step": 53560 }, { "grad_norm": 0.26177576184272766, "learning_rate": 4.8232808896871185e-05, "loss": 0.022, "step": 53570 }, { "grad_norm": 0.19857698678970337, "learning_rate": 4.821628462630624e-05, "loss": 0.0204, "step": 53580 }, { "grad_norm": 0.21633681654930115, "learning_rate": 4.8199760550805735e-05, "loss": 0.024, "step": 53590 }, { "grad_norm": 0.31147852540016174, "learning_rate": 4.818323667217669e-05, "loss": 0.0253, "step": 53600 }, { "grad_norm": 0.34209778904914856, "learning_rate": 4.8166712992226165e-05, "loss": 0.024, "step": 53610 }, { "grad_norm": 0.2951171398162842, "learning_rate": 4.8150189512761154e-05, "loss": 0.0178, "step": 53620 }, { "grad_norm": 0.29586607217788696, "learning_rate": 4.813366623558862e-05, "loss": 0.0282, "step": 53630 }, { "grad_norm": 0.2685345709323883, "learning_rate": 4.811714316251554e-05, "loss": 0.0268, "step": 53640 }, { "grad_norm": 0.3267233073711395, "learning_rate": 4.810062029534886e-05, "loss": 0.021, "step": 53650 }, { "grad_norm": 0.35424500703811646, "learning_rate": 4.8084097635895505e-05, "loss": 0.0268, "step": 53660 }, { "grad_norm": 0.2872857451438904, "learning_rate": 4.8067575185962335e-05, "loss": 0.0248, "step": 53670 }, { "grad_norm": 0.241486057639122, "learning_rate": 4.805105294735625e-05, "loss": 0.0236, "step": 53680 }, { "grad_norm": 0.272533118724823, "learning_rate": 4.803453092188406e-05, "loss": 0.0234, "step": 53690 }, { "grad_norm": 0.3130205571651459, "learning_rate": 4.801800911135263e-05, "loss": 0.0223, "step": 53700 }, { "grad_norm": 0.24858838319778442, "learning_rate": 4.8001487517568744e-05, "loss": 0.0237, "step": 53710 }, { "grad_norm": 0.1627940982580185, "learning_rate": 4.798496614233918e-05, "loss": 0.0287, "step": 53720 }, { "grad_norm": 0.29443359375, "learning_rate": 4.796844498747069e-05, "loss": 0.0303, "step": 53730 }, { "grad_norm": 0.26445716619491577, "learning_rate": 4.795192405477e-05, "loss": 0.0213, "step": 53740 }, { "grad_norm": 0.4115550220012665, "learning_rate": 4.79354033460438e-05, "loss": 0.0297, "step": 53750 }, { "grad_norm": 0.36528512835502625, "learning_rate": 4.79188828630988e-05, "loss": 0.0325, "step": 53760 }, { "grad_norm": 0.33487799763679504, "learning_rate": 4.790236260774165e-05, "loss": 0.0315, "step": 53770 }, { "grad_norm": 0.22834007441997528, "learning_rate": 4.788584258177896e-05, "loss": 0.0295, "step": 53780 }, { "grad_norm": 0.3943234384059906, "learning_rate": 4.7869322787017356e-05, "loss": 0.0282, "step": 53790 }, { "grad_norm": 0.45453009009361267, "learning_rate": 4.78528032252634e-05, "loss": 0.0298, "step": 53800 }, { "grad_norm": 0.40162283182144165, "learning_rate": 4.783628389832366e-05, "loss": 0.0278, "step": 53810 }, { "grad_norm": 0.31731224060058594, "learning_rate": 4.7819764808004644e-05, "loss": 0.0322, "step": 53820 }, { "grad_norm": 0.20724034309387207, "learning_rate": 4.780324595611289e-05, "loss": 0.0225, "step": 53830 }, { "grad_norm": 0.35380467772483826, "learning_rate": 4.7786727344454854e-05, "loss": 0.0297, "step": 53840 }, { "grad_norm": 0.6504601240158081, "learning_rate": 4.777020897483697e-05, "loss": 0.03, "step": 53850 }, { "grad_norm": 0.205575093626976, "learning_rate": 4.775369084906569e-05, "loss": 0.0264, "step": 53860 }, { "grad_norm": 0.2921169102191925, "learning_rate": 4.773717296894742e-05, "loss": 0.0314, "step": 53870 }, { "grad_norm": 0.3363368809223175, "learning_rate": 4.772065533628848e-05, "loss": 0.0335, "step": 53880 }, { "grad_norm": 0.13199858367443085, "learning_rate": 4.7704137952895265e-05, "loss": 0.0185, "step": 53890 }, { "grad_norm": 0.2747299075126648, "learning_rate": 4.768762082057407e-05, "loss": 0.0232, "step": 53900 }, { "grad_norm": 0.3668123483657837, "learning_rate": 4.7671103941131186e-05, "loss": 0.034, "step": 53910 }, { "grad_norm": 0.2342677116394043, "learning_rate": 4.765458731637287e-05, "loss": 0.0253, "step": 53920 }, { "grad_norm": 0.3541928827762604, "learning_rate": 4.763807094810537e-05, "loss": 0.0337, "step": 53930 }, { "grad_norm": 0.2817330062389374, "learning_rate": 4.762155483813487e-05, "loss": 0.0315, "step": 53940 }, { "grad_norm": 0.373009592294693, "learning_rate": 4.760503898826754e-05, "loss": 0.0243, "step": 53950 }, { "grad_norm": 0.47354209423065186, "learning_rate": 4.758852340030957e-05, "loss": 0.0271, "step": 53960 }, { "grad_norm": 0.21657922863960266, "learning_rate": 4.757200807606705e-05, "loss": 0.0311, "step": 53970 }, { "grad_norm": 0.3425664007663727, "learning_rate": 4.755549301734607e-05, "loss": 0.0384, "step": 53980 }, { "grad_norm": 0.18476799130439758, "learning_rate": 4.753897822595271e-05, "loss": 0.0379, "step": 53990 }, { "grad_norm": 0.30110666155815125, "learning_rate": 4.7522463703692994e-05, "loss": 0.0229, "step": 54000 }, { "grad_norm": 0.22758443653583527, "learning_rate": 4.75059494523729e-05, "loss": 0.0252, "step": 54010 }, { "grad_norm": 0.3674283027648926, "learning_rate": 4.748943547379843e-05, "loss": 0.027, "step": 54020 }, { "grad_norm": 0.6302533149719238, "learning_rate": 4.7472921769775535e-05, "loss": 0.034, "step": 54030 }, { "grad_norm": 0.29785022139549255, "learning_rate": 4.745640834211011e-05, "loss": 0.0248, "step": 54040 }, { "grad_norm": 0.2145659178495407, "learning_rate": 4.7439895192608054e-05, "loss": 0.0214, "step": 54050 }, { "grad_norm": 0.3100215494632721, "learning_rate": 4.7423382323075186e-05, "loss": 0.0241, "step": 54060 }, { "grad_norm": 0.2085709571838379, "learning_rate": 4.740686973531737e-05, "loss": 0.0252, "step": 54070 }, { "grad_norm": 0.32211819291114807, "learning_rate": 4.7390357431140375e-05, "loss": 0.0241, "step": 54080 }, { "grad_norm": 0.2981770932674408, "learning_rate": 4.737384541234998e-05, "loss": 0.023, "step": 54090 }, { "grad_norm": 0.3174159824848175, "learning_rate": 4.735733368075189e-05, "loss": 0.0189, "step": 54100 }, { "grad_norm": 0.357891321182251, "learning_rate": 4.734082223815182e-05, "loss": 0.0255, "step": 54110 }, { "grad_norm": 0.2892208993434906, "learning_rate": 4.7324311086355436e-05, "loss": 0.0301, "step": 54120 }, { "grad_norm": 0.2707813084125519, "learning_rate": 4.730780022716836e-05, "loss": 0.0307, "step": 54130 }, { "grad_norm": 0.2742173671722412, "learning_rate": 4.729128966239619e-05, "loss": 0.0175, "step": 54140 }, { "grad_norm": 0.3330574333667755, "learning_rate": 4.7274779393844536e-05, "loss": 0.0215, "step": 54150 }, { "grad_norm": 0.36094313859939575, "learning_rate": 4.725826942331891e-05, "loss": 0.0356, "step": 54160 }, { "grad_norm": 0.3638065457344055, "learning_rate": 4.724175975262481e-05, "loss": 0.0211, "step": 54170 }, { "grad_norm": 0.3507100045681, "learning_rate": 4.722525038356772e-05, "loss": 0.0195, "step": 54180 }, { "grad_norm": 0.1359744518995285, "learning_rate": 4.7208741317953075e-05, "loss": 0.0272, "step": 54190 }, { "grad_norm": 0.3602071702480316, "learning_rate": 4.7192232557586264e-05, "loss": 0.0309, "step": 54200 }, { "grad_norm": 0.2548564076423645, "learning_rate": 4.717572410427269e-05, "loss": 0.0259, "step": 54210 }, { "grad_norm": 0.3091421127319336, "learning_rate": 4.71592159598177e-05, "loss": 0.0267, "step": 54220 }, { "grad_norm": 0.3209463655948639, "learning_rate": 4.714270812602657e-05, "loss": 0.0265, "step": 54230 }, { "grad_norm": 0.30529528856277466, "learning_rate": 4.712620060470458e-05, "loss": 0.0349, "step": 54240 }, { "grad_norm": 0.41798335313796997, "learning_rate": 4.7109693397656986e-05, "loss": 0.0186, "step": 54250 }, { "grad_norm": 0.22762560844421387, "learning_rate": 4.709318650668894e-05, "loss": 0.019, "step": 54260 }, { "grad_norm": 0.2916783094406128, "learning_rate": 4.7076679933605676e-05, "loss": 0.0231, "step": 54270 }, { "grad_norm": 0.28072449564933777, "learning_rate": 4.70601736802123e-05, "loss": 0.0287, "step": 54280 }, { "grad_norm": 0.4090518355369568, "learning_rate": 4.704366774831391e-05, "loss": 0.034, "step": 54290 }, { "grad_norm": 0.24042731523513794, "learning_rate": 4.702716213971557e-05, "loss": 0.0306, "step": 54300 }, { "grad_norm": 0.3016507923603058, "learning_rate": 4.70106568562223e-05, "loss": 0.0285, "step": 54310 }, { "grad_norm": 0.8328573107719421, "learning_rate": 4.6994151899639105e-05, "loss": 0.0261, "step": 54320 }, { "grad_norm": 0.23215487599372864, "learning_rate": 4.697764727177093e-05, "loss": 0.0246, "step": 54330 }, { "grad_norm": 0.3942018151283264, "learning_rate": 4.696114297442272e-05, "loss": 0.0253, "step": 54340 }, { "grad_norm": 0.17653287947177887, "learning_rate": 4.694463900939933e-05, "loss": 0.02, "step": 54350 }, { "grad_norm": 0.2603466808795929, "learning_rate": 4.692813537850564e-05, "loss": 0.0269, "step": 54360 }, { "grad_norm": 0.2116892784833908, "learning_rate": 4.6911632083546445e-05, "loss": 0.0246, "step": 54370 }, { "grad_norm": 0.3893607556819916, "learning_rate": 4.6895129126326525e-05, "loss": 0.0316, "step": 54380 }, { "grad_norm": 0.30452781915664673, "learning_rate": 4.687862650865059e-05, "loss": 0.0263, "step": 54390 }, { "grad_norm": 0.2031451314687729, "learning_rate": 4.686212423232339e-05, "loss": 0.0227, "step": 54400 }, { "grad_norm": 0.25323566794395447, "learning_rate": 4.6845622299149565e-05, "loss": 0.0214, "step": 54410 }, { "grad_norm": 0.25866347551345825, "learning_rate": 4.682912071093374e-05, "loss": 0.0209, "step": 54420 }, { "grad_norm": 0.1685071587562561, "learning_rate": 4.681261946948052e-05, "loss": 0.0252, "step": 54430 }, { "grad_norm": 0.18094681203365326, "learning_rate": 4.679611857659443e-05, "loss": 0.0388, "step": 54440 }, { "grad_norm": 0.2843966484069824, "learning_rate": 4.6779618034079975e-05, "loss": 0.027, "step": 54450 }, { "grad_norm": 0.2747495174407959, "learning_rate": 4.6763117843741675e-05, "loss": 0.029, "step": 54460 }, { "grad_norm": 0.23153488337993622, "learning_rate": 4.674661800738394e-05, "loss": 0.0219, "step": 54470 }, { "grad_norm": 0.3389367461204529, "learning_rate": 4.673011852681117e-05, "loss": 0.0241, "step": 54480 }, { "grad_norm": 0.3592827022075653, "learning_rate": 4.671361940382772e-05, "loss": 0.0265, "step": 54490 }, { "grad_norm": 0.29254651069641113, "learning_rate": 4.669712064023791e-05, "loss": 0.0278, "step": 54500 }, { "grad_norm": 0.21530401706695557, "learning_rate": 4.668062223784602e-05, "loss": 0.0211, "step": 54510 }, { "grad_norm": 0.18078964948654175, "learning_rate": 4.6664124198456274e-05, "loss": 0.0192, "step": 54520 }, { "grad_norm": 0.3144531846046448, "learning_rate": 4.6647626523872905e-05, "loss": 0.0194, "step": 54530 }, { "grad_norm": 0.19988523423671722, "learning_rate": 4.6631129215900066e-05, "loss": 0.0219, "step": 54540 }, { "grad_norm": 0.812175989151001, "learning_rate": 4.661463227634187e-05, "loss": 0.0349, "step": 54550 }, { "grad_norm": 0.2927475571632385, "learning_rate": 4.659813570700238e-05, "loss": 0.0154, "step": 54560 }, { "grad_norm": 0.23765374720096588, "learning_rate": 4.658163950968568e-05, "loss": 0.0208, "step": 54570 }, { "grad_norm": 0.17559993267059326, "learning_rate": 4.6565143686195726e-05, "loss": 0.027, "step": 54580 }, { "grad_norm": 0.29400888085365295, "learning_rate": 4.654864823833649e-05, "loss": 0.0281, "step": 54590 }, { "grad_norm": 0.25782671570777893, "learning_rate": 4.65321531679119e-05, "loss": 0.0254, "step": 54600 }, { "grad_norm": 0.21809720993041992, "learning_rate": 4.6515658476725834e-05, "loss": 0.0442, "step": 54610 }, { "grad_norm": 0.20985069870948792, "learning_rate": 4.649916416658212e-05, "loss": 0.032, "step": 54620 }, { "grad_norm": 0.23256511986255646, "learning_rate": 4.648267023928455e-05, "loss": 0.0266, "step": 54630 }, { "grad_norm": 0.21644319593906403, "learning_rate": 4.646617669663687e-05, "loss": 0.0233, "step": 54640 }, { "grad_norm": 0.3903070390224457, "learning_rate": 4.644968354044277e-05, "loss": 0.0237, "step": 54650 }, { "grad_norm": 0.33430346846580505, "learning_rate": 4.643319077250596e-05, "loss": 0.0218, "step": 54660 }, { "grad_norm": 0.6318581104278564, "learning_rate": 4.6416698394630055e-05, "loss": 0.0255, "step": 54670 }, { "grad_norm": 0.462166965007782, "learning_rate": 4.640020640861862e-05, "loss": 0.0261, "step": 54680 }, { "grad_norm": 0.4490732252597809, "learning_rate": 4.6383714816275195e-05, "loss": 0.029, "step": 54690 }, { "grad_norm": 0.5391114354133606, "learning_rate": 4.636722361940329e-05, "loss": 0.0216, "step": 54700 }, { "grad_norm": 0.3020777702331543, "learning_rate": 4.635073281980632e-05, "loss": 0.0227, "step": 54710 }, { "grad_norm": 0.3332771062850952, "learning_rate": 4.633424241928775e-05, "loss": 0.0245, "step": 54720 }, { "grad_norm": 0.2527998387813568, "learning_rate": 4.6317752419650914e-05, "loss": 0.0253, "step": 54730 }, { "grad_norm": 0.23625820875167847, "learning_rate": 4.630126282269914e-05, "loss": 0.0251, "step": 54740 }, { "grad_norm": 0.27371153235435486, "learning_rate": 4.628477363023569e-05, "loss": 0.0324, "step": 54750 }, { "grad_norm": 0.27113279700279236, "learning_rate": 4.626828484406383e-05, "loss": 0.0222, "step": 54760 }, { "grad_norm": 0.3408263921737671, "learning_rate": 4.625179646598669e-05, "loss": 0.0243, "step": 54770 }, { "grad_norm": 0.23368960618972778, "learning_rate": 4.6235308497807487e-05, "loss": 0.0237, "step": 54780 }, { "grad_norm": 0.2390938550233841, "learning_rate": 4.6218820941329276e-05, "loss": 0.027, "step": 54790 }, { "grad_norm": 0.1941971778869629, "learning_rate": 4.620233379835513e-05, "loss": 0.0248, "step": 54800 }, { "grad_norm": 0.17957210540771484, "learning_rate": 4.618584707068804e-05, "loss": 0.02, "step": 54810 }, { "grad_norm": 0.26553821563720703, "learning_rate": 4.616936076013099e-05, "loss": 0.0273, "step": 54820 }, { "grad_norm": 0.6135122179985046, "learning_rate": 4.615287486848688e-05, "loss": 0.0397, "step": 54830 }, { "grad_norm": 0.33852967619895935, "learning_rate": 4.6136389397558584e-05, "loss": 0.0261, "step": 54840 }, { "grad_norm": 0.34357500076293945, "learning_rate": 4.6119904349148935e-05, "loss": 0.0264, "step": 54850 }, { "grad_norm": 0.6735550761222839, "learning_rate": 4.6103419725060726e-05, "loss": 0.027, "step": 54860 }, { "grad_norm": 0.25746580958366394, "learning_rate": 4.608693552709667e-05, "loss": 0.0199, "step": 54870 }, { "grad_norm": 0.31211015582084656, "learning_rate": 4.607045175705947e-05, "loss": 0.0246, "step": 54880 }, { "grad_norm": 0.24186351895332336, "learning_rate": 4.605396841675176e-05, "loss": 0.0325, "step": 54890 }, { "grad_norm": 0.4519156813621521, "learning_rate": 4.6037485507976094e-05, "loss": 0.0291, "step": 54900 }, { "grad_norm": 0.3340100049972534, "learning_rate": 4.602100303253509e-05, "loss": 0.0307, "step": 54910 }, { "grad_norm": 0.32134678959846497, "learning_rate": 4.600452099223122e-05, "loss": 0.0348, "step": 54920 }, { "grad_norm": 0.42484018206596375, "learning_rate": 4.598803938886691e-05, "loss": 0.0219, "step": 54930 }, { "grad_norm": 0.25921326875686646, "learning_rate": 4.597155822424459e-05, "loss": 0.0234, "step": 54940 }, { "grad_norm": 0.4493367671966553, "learning_rate": 4.59550775001666e-05, "loss": 0.0266, "step": 54950 }, { "grad_norm": 0.21259063482284546, "learning_rate": 4.5938597218435234e-05, "loss": 0.0233, "step": 54960 }, { "grad_norm": 0.45195692777633667, "learning_rate": 4.592211738085279e-05, "loss": 0.0273, "step": 54970 }, { "grad_norm": 0.14036297798156738, "learning_rate": 4.590563798922146e-05, "loss": 0.0185, "step": 54980 }, { "grad_norm": 0.27385765314102173, "learning_rate": 4.5889159045343404e-05, "loss": 0.0321, "step": 54990 }, { "grad_norm": 0.4850521385669708, "learning_rate": 4.5872680551020734e-05, "loss": 0.0257, "step": 55000 }, { "grad_norm": 0.25589311122894287, "learning_rate": 4.585620250805552e-05, "loss": 0.0286, "step": 55010 }, { "grad_norm": 0.17523276805877686, "learning_rate": 4.583972491824975e-05, "loss": 0.0257, "step": 55020 }, { "grad_norm": 0.26900482177734375, "learning_rate": 4.582324778340541e-05, "loss": 0.0242, "step": 55030 }, { "grad_norm": 0.3726732134819031, "learning_rate": 4.5806771105324424e-05, "loss": 0.0255, "step": 55040 }, { "grad_norm": 0.22277683019638062, "learning_rate": 4.579029488580865e-05, "loss": 0.0203, "step": 55050 }, { "grad_norm": 0.24836552143096924, "learning_rate": 4.5773819126659876e-05, "loss": 0.0205, "step": 55060 }, { "grad_norm": 0.3483913540840149, "learning_rate": 4.575734382967991e-05, "loss": 0.022, "step": 55070 }, { "grad_norm": 0.3189204931259155, "learning_rate": 4.574086899667044e-05, "loss": 0.024, "step": 55080 }, { "grad_norm": 0.2687486410140991, "learning_rate": 4.5724394629433106e-05, "loss": 0.0227, "step": 55090 }, { "grad_norm": 0.324923038482666, "learning_rate": 4.570792072976958e-05, "loss": 0.0282, "step": 55100 }, { "grad_norm": 0.3833504319190979, "learning_rate": 4.569144729948138e-05, "loss": 0.0295, "step": 55110 }, { "grad_norm": 0.31570062041282654, "learning_rate": 4.5674974340370033e-05, "loss": 0.024, "step": 55120 }, { "grad_norm": 0.29522794485092163, "learning_rate": 4.565850185423699e-05, "loss": 0.0242, "step": 55130 }, { "grad_norm": 0.24298585951328278, "learning_rate": 4.564202984288365e-05, "loss": 0.0277, "step": 55140 }, { "grad_norm": 0.2190198451280594, "learning_rate": 4.5625558308111356e-05, "loss": 0.0225, "step": 55150 }, { "grad_norm": 0.2422376126050949, "learning_rate": 4.560908725172145e-05, "loss": 0.0204, "step": 55160 }, { "grad_norm": 0.3002125024795532, "learning_rate": 4.559261667551516e-05, "loss": 0.0226, "step": 55170 }, { "grad_norm": 0.34656986594200134, "learning_rate": 4.5576146581293685e-05, "loss": 0.0254, "step": 55180 }, { "grad_norm": 0.30708956718444824, "learning_rate": 4.5559676970858164e-05, "loss": 0.029, "step": 55190 }, { "grad_norm": 0.4978211224079132, "learning_rate": 4.5543207846009705e-05, "loss": 0.0333, "step": 55200 }, { "grad_norm": 0.3331950306892395, "learning_rate": 4.552673920854933e-05, "loss": 0.0288, "step": 55210 }, { "grad_norm": 0.31184032559394836, "learning_rate": 4.5510271060278006e-05, "loss": 0.0268, "step": 55220 }, { "grad_norm": 0.3198888897895813, "learning_rate": 4.5493803402996704e-05, "loss": 0.0304, "step": 55230 }, { "grad_norm": 0.4849141240119934, "learning_rate": 4.5477336238506295e-05, "loss": 0.0211, "step": 55240 }, { "grad_norm": 0.19717569649219513, "learning_rate": 4.546086956860759e-05, "loss": 0.0195, "step": 55250 }, { "grad_norm": 0.14830276370048523, "learning_rate": 4.5444403395101374e-05, "loss": 0.0229, "step": 55260 }, { "grad_norm": 0.4016326665878296, "learning_rate": 4.542793771978834e-05, "loss": 0.0297, "step": 55270 }, { "grad_norm": 0.3062208890914917, "learning_rate": 4.541147254446916e-05, "loss": 0.0231, "step": 55280 }, { "grad_norm": 0.20016302168369293, "learning_rate": 4.539500787094445e-05, "loss": 0.028, "step": 55290 }, { "grad_norm": 0.24552418291568756, "learning_rate": 4.537854370101475e-05, "loss": 0.0189, "step": 55300 }, { "grad_norm": 0.24898776412010193, "learning_rate": 4.5362080036480566e-05, "loss": 0.0225, "step": 55310 }, { "grad_norm": 0.2690068483352661, "learning_rate": 4.534561687914235e-05, "loss": 0.0294, "step": 55320 }, { "grad_norm": 0.26889461278915405, "learning_rate": 4.5329154230800466e-05, "loss": 0.0254, "step": 55330 }, { "grad_norm": 0.2477731704711914, "learning_rate": 4.531269209325525e-05, "loss": 0.0215, "step": 55340 }, { "grad_norm": 0.2577611207962036, "learning_rate": 4.529623046830696e-05, "loss": 0.0276, "step": 55350 }, { "grad_norm": 0.26894742250442505, "learning_rate": 4.5279769357755866e-05, "loss": 0.0194, "step": 55360 }, { "grad_norm": 0.21192176640033722, "learning_rate": 4.5263308763402084e-05, "loss": 0.0276, "step": 55370 }, { "grad_norm": 0.32743191719055176, "learning_rate": 4.524684868704574e-05, "loss": 0.0258, "step": 55380 }, { "grad_norm": 0.24262818694114685, "learning_rate": 4.523038913048688e-05, "loss": 0.0235, "step": 55390 }, { "grad_norm": 0.21691808104515076, "learning_rate": 4.521393009552549e-05, "loss": 0.0264, "step": 55400 }, { "grad_norm": 0.7677584290504456, "learning_rate": 4.5197471583961484e-05, "loss": 0.0297, "step": 55410 }, { "grad_norm": 0.29357218742370605, "learning_rate": 4.51810135975948e-05, "loss": 0.0354, "step": 55420 }, { "grad_norm": 0.1954122632741928, "learning_rate": 4.516455613822521e-05, "loss": 0.0241, "step": 55430 }, { "grad_norm": 0.24686700105667114, "learning_rate": 4.514809920765249e-05, "loss": 0.0341, "step": 55440 }, { "grad_norm": 0.30923259258270264, "learning_rate": 4.5131642807676356e-05, "loss": 0.018, "step": 55450 }, { "grad_norm": 0.38498467206954956, "learning_rate": 4.511518694009644e-05, "loss": 0.0216, "step": 55460 }, { "grad_norm": 0.2944170832633972, "learning_rate": 4.509873160671231e-05, "loss": 0.0263, "step": 55470 }, { "grad_norm": 0.3421267867088318, "learning_rate": 4.508227680932356e-05, "loss": 0.0265, "step": 55480 }, { "grad_norm": 0.32912495732307434, "learning_rate": 4.5065822549729605e-05, "loss": 0.0247, "step": 55490 }, { "grad_norm": 0.30610567331314087, "learning_rate": 4.504936882972989e-05, "loss": 0.0284, "step": 55500 }, { "grad_norm": 0.24398352205753326, "learning_rate": 4.503291565112374e-05, "loss": 0.0233, "step": 55510 }, { "grad_norm": 0.4564068615436554, "learning_rate": 4.501646301571048e-05, "loss": 0.04, "step": 55520 }, { "grad_norm": 0.23665115237236023, "learning_rate": 4.500001092528933e-05, "loss": 0.0281, "step": 55530 }, { "grad_norm": 0.24985043704509735, "learning_rate": 4.4983559381659455e-05, "loss": 0.0256, "step": 55540 }, { "grad_norm": 0.39459583163261414, "learning_rate": 4.496710838661999e-05, "loss": 0.026, "step": 55550 }, { "grad_norm": 0.44846463203430176, "learning_rate": 4.495065794196999e-05, "loss": 0.0249, "step": 55560 }, { "grad_norm": 0.2983628511428833, "learning_rate": 4.4934208049508455e-05, "loss": 0.0254, "step": 55570 }, { "grad_norm": 0.38650059700012207, "learning_rate": 4.4917758711034303e-05, "loss": 0.0298, "step": 55580 }, { "grad_norm": 0.46043962240219116, "learning_rate": 4.490130992834642e-05, "loss": 0.0231, "step": 55590 }, { "grad_norm": 0.3878248333930969, "learning_rate": 4.488486170324359e-05, "loss": 0.0382, "step": 55600 }, { "grad_norm": 0.29570427536964417, "learning_rate": 4.486841403752463e-05, "loss": 0.0249, "step": 55610 }, { "grad_norm": 0.13852614164352417, "learning_rate": 4.4851966932988174e-05, "loss": 0.0196, "step": 55620 }, { "grad_norm": 0.6325373649597168, "learning_rate": 4.4835520391432894e-05, "loss": 0.021, "step": 55630 }, { "grad_norm": 0.25133639574050903, "learning_rate": 4.481907441465733e-05, "loss": 0.0226, "step": 55640 }, { "grad_norm": 0.2057870328426361, "learning_rate": 4.480262900446e-05, "loss": 0.0274, "step": 55650 }, { "grad_norm": 0.23267598450183868, "learning_rate": 4.478618416263933e-05, "loss": 0.0237, "step": 55660 }, { "grad_norm": 0.19259437918663025, "learning_rate": 4.476973989099373e-05, "loss": 0.0214, "step": 55670 }, { "grad_norm": 0.20893831551074982, "learning_rate": 4.475329619132153e-05, "loss": 0.0278, "step": 55680 }, { "grad_norm": 0.17919720709323883, "learning_rate": 4.4736853065420956e-05, "loss": 0.0186, "step": 55690 }, { "grad_norm": 0.2572145462036133, "learning_rate": 4.472041051509023e-05, "loss": 0.0224, "step": 55700 }, { "grad_norm": 0.25187182426452637, "learning_rate": 4.470396854212746e-05, "loss": 0.02, "step": 55710 }, { "grad_norm": 0.17308691143989563, "learning_rate": 4.468752714833072e-05, "loss": 0.024, "step": 55720 }, { "grad_norm": 0.2707671523094177, "learning_rate": 4.467108633549803e-05, "loss": 0.024, "step": 55730 }, { "grad_norm": 0.20343488454818726, "learning_rate": 4.4654646105427335e-05, "loss": 0.0272, "step": 55740 }, { "grad_norm": 0.8567943572998047, "learning_rate": 4.463820645991651e-05, "loss": 0.0272, "step": 55750 }, { "grad_norm": 0.38086697459220886, "learning_rate": 4.462176740076334e-05, "loss": 0.0215, "step": 55760 }, { "grad_norm": 0.2956780791282654, "learning_rate": 4.4605328929765624e-05, "loss": 0.0232, "step": 55770 }, { "grad_norm": 0.23738259077072144, "learning_rate": 4.458889104872102e-05, "loss": 0.0167, "step": 55780 }, { "grad_norm": 0.25028741359710693, "learning_rate": 4.457245375942715e-05, "loss": 0.0197, "step": 55790 }, { "grad_norm": 0.24368916451931, "learning_rate": 4.455601706368157e-05, "loss": 0.0252, "step": 55800 }, { "grad_norm": 0.32069727778434753, "learning_rate": 4.453958096328179e-05, "loss": 0.0247, "step": 55810 }, { "grad_norm": 0.3328680396080017, "learning_rate": 4.452314546002523e-05, "loss": 0.0252, "step": 55820 }, { "grad_norm": 0.4028741121292114, "learning_rate": 4.4506710555709246e-05, "loss": 0.0241, "step": 55830 }, { "grad_norm": 0.4337410628795624, "learning_rate": 4.449027625213114e-05, "loss": 0.0297, "step": 55840 }, { "grad_norm": 0.22704710066318512, "learning_rate": 4.4473842551088135e-05, "loss": 0.025, "step": 55850 }, { "grad_norm": 0.22117480635643005, "learning_rate": 4.445740945437738e-05, "loss": 0.0211, "step": 55860 }, { "grad_norm": 0.3284018337726593, "learning_rate": 4.444097696379602e-05, "loss": 0.0323, "step": 55870 }, { "grad_norm": 0.3735906183719635, "learning_rate": 4.442454508114106e-05, "loss": 0.0249, "step": 55880 }, { "grad_norm": 0.27675294876098633, "learning_rate": 4.4408113808209465e-05, "loss": 0.0256, "step": 55890 }, { "grad_norm": 0.20964857935905457, "learning_rate": 4.439168314679813e-05, "loss": 0.0255, "step": 55900 }, { "grad_norm": 0.5341759324073792, "learning_rate": 4.43752530987039e-05, "loss": 0.0269, "step": 55910 }, { "grad_norm": 0.2803433835506439, "learning_rate": 4.4358823665723515e-05, "loss": 0.0193, "step": 55920 }, { "grad_norm": 0.2209654450416565, "learning_rate": 4.434239484965371e-05, "loss": 0.0192, "step": 55930 }, { "grad_norm": 0.26338616013526917, "learning_rate": 4.4325966652291103e-05, "loss": 0.0234, "step": 55940 }, { "grad_norm": 0.29621952772140503, "learning_rate": 4.430953907543225e-05, "loss": 0.0218, "step": 55950 }, { "grad_norm": 0.17943742871284485, "learning_rate": 4.4293112120873645e-05, "loss": 0.0209, "step": 55960 }, { "grad_norm": 0.3182453215122223, "learning_rate": 4.427668579041172e-05, "loss": 0.017, "step": 55970 }, { "grad_norm": 0.2719675600528717, "learning_rate": 4.4260260085842816e-05, "loss": 0.0256, "step": 55980 }, { "grad_norm": 0.2631016969680786, "learning_rate": 4.4243835008963256e-05, "loss": 0.0218, "step": 55990 }, { "grad_norm": 0.2668587267398834, "learning_rate": 4.422741056156925e-05, "loss": 0.0218, "step": 56000 }, { "grad_norm": 0.4705972969532013, "learning_rate": 4.421098674545693e-05, "loss": 0.0181, "step": 56010 }, { "grad_norm": 0.31589004397392273, "learning_rate": 4.419456356242241e-05, "loss": 0.0271, "step": 56020 }, { "grad_norm": 0.1900094896554947, "learning_rate": 4.4178141014261685e-05, "loss": 0.0257, "step": 56030 }, { "grad_norm": 0.3310341536998749, "learning_rate": 4.416171910277071e-05, "loss": 0.0227, "step": 56040 }, { "grad_norm": 0.2616174519062042, "learning_rate": 4.414529782974533e-05, "loss": 0.0249, "step": 56050 }, { "grad_norm": 0.23302778601646423, "learning_rate": 4.41288771969814e-05, "loss": 0.027, "step": 56060 }, { "grad_norm": 0.3158029317855835, "learning_rate": 4.411245720627464e-05, "loss": 0.0264, "step": 56070 }, { "grad_norm": 0.20333948731422424, "learning_rate": 4.40960378594207e-05, "loss": 0.0296, "step": 56080 }, { "grad_norm": 0.16394023597240448, "learning_rate": 4.407961915821519e-05, "loss": 0.0265, "step": 56090 }, { "grad_norm": 0.2819938361644745, "learning_rate": 4.406320110445362e-05, "loss": 0.026, "step": 56100 }, { "grad_norm": 0.2809957265853882, "learning_rate": 4.404678369993144e-05, "loss": 0.022, "step": 56110 }, { "grad_norm": 0.19602741301059723, "learning_rate": 4.403036694644406e-05, "loss": 0.0208, "step": 56120 }, { "grad_norm": 0.304555207490921, "learning_rate": 4.4013950845786764e-05, "loss": 0.0168, "step": 56130 }, { "grad_norm": 0.39426037669181824, "learning_rate": 4.399753539975482e-05, "loss": 0.0266, "step": 56140 }, { "grad_norm": 0.16421082615852356, "learning_rate": 4.398112061014337e-05, "loss": 0.0176, "step": 56150 }, { "grad_norm": 0.20121583342552185, "learning_rate": 4.396470647874753e-05, "loss": 0.0281, "step": 56160 }, { "grad_norm": 0.3683145046234131, "learning_rate": 4.394829300736229e-05, "loss": 0.0225, "step": 56170 }, { "grad_norm": 0.28950753808021545, "learning_rate": 4.393188019778265e-05, "loss": 0.019, "step": 56180 }, { "grad_norm": 0.23057855665683746, "learning_rate": 4.391546805180347e-05, "loss": 0.0211, "step": 56190 }, { "grad_norm": 0.20407208800315857, "learning_rate": 4.389905657121955e-05, "loss": 0.0236, "step": 56200 }, { "grad_norm": 0.2894473671913147, "learning_rate": 4.3882645757825635e-05, "loss": 0.0229, "step": 56210 }, { "grad_norm": 0.30142852663993835, "learning_rate": 4.386623561341637e-05, "loss": 0.0275, "step": 56220 }, { "grad_norm": 0.5657771825790405, "learning_rate": 4.384982613978637e-05, "loss": 0.0257, "step": 56230 }, { "grad_norm": 0.3594295084476471, "learning_rate": 4.383341733873012e-05, "loss": 0.0227, "step": 56240 }, { "grad_norm": 0.18607191741466522, "learning_rate": 4.381700921204209e-05, "loss": 0.0235, "step": 56250 }, { "grad_norm": 0.33397072553634644, "learning_rate": 4.380060176151663e-05, "loss": 0.0346, "step": 56260 }, { "grad_norm": 0.17731861770153046, "learning_rate": 4.378419498894805e-05, "loss": 0.0203, "step": 56270 }, { "grad_norm": 0.33652263879776, "learning_rate": 4.376778889613056e-05, "loss": 0.0248, "step": 56280 }, { "grad_norm": 0.2656186521053314, "learning_rate": 4.375138348485831e-05, "loss": 0.0249, "step": 56290 }, { "grad_norm": 0.2927137017250061, "learning_rate": 4.3734978756925346e-05, "loss": 0.0202, "step": 56300 }, { "grad_norm": 0.3089520037174225, "learning_rate": 4.3718574714125706e-05, "loss": 0.0191, "step": 56310 }, { "grad_norm": 0.20640182495117188, "learning_rate": 4.370217135825329e-05, "loss": 0.0211, "step": 56320 }, { "grad_norm": 0.19534236192703247, "learning_rate": 4.368576869110194e-05, "loss": 0.0175, "step": 56330 }, { "grad_norm": 0.1977289915084839, "learning_rate": 4.366936671446544e-05, "loss": 0.0192, "step": 56340 }, { "grad_norm": 0.33915242552757263, "learning_rate": 4.3652965430137474e-05, "loss": 0.0259, "step": 56350 }, { "grad_norm": 0.274975448846817, "learning_rate": 4.3636564839911646e-05, "loss": 0.0238, "step": 56360 }, { "grad_norm": 0.22394931316375732, "learning_rate": 4.3620164945581545e-05, "loss": 0.0186, "step": 56370 }, { "grad_norm": 1.0241700410842896, "learning_rate": 4.360376574894061e-05, "loss": 0.0248, "step": 56380 }, { "grad_norm": 0.1918228715658188, "learning_rate": 4.358736725178224e-05, "loss": 0.0186, "step": 56390 }, { "grad_norm": 0.24375200271606445, "learning_rate": 4.357096945589974e-05, "loss": 0.0253, "step": 56400 }, { "grad_norm": 0.20443840324878693, "learning_rate": 4.355457236308636e-05, "loss": 0.0192, "step": 56410 }, { "grad_norm": 0.2094947248697281, "learning_rate": 4.353817597513526e-05, "loss": 0.0207, "step": 56420 }, { "grad_norm": 0.4438033103942871, "learning_rate": 4.352178029383948e-05, "loss": 0.0194, "step": 56430 }, { "grad_norm": 0.28196191787719727, "learning_rate": 4.3505385320992105e-05, "loss": 0.0226, "step": 56440 }, { "grad_norm": 0.32889053225517273, "learning_rate": 4.348899105838602e-05, "loss": 0.0259, "step": 56450 }, { "grad_norm": 0.395480751991272, "learning_rate": 4.3472597507814087e-05, "loss": 0.0195, "step": 56460 }, { "grad_norm": 0.359815776348114, "learning_rate": 4.3456204671069066e-05, "loss": 0.0213, "step": 56470 }, { "grad_norm": 0.32553258538246155, "learning_rate": 4.343981254994367e-05, "loss": 0.0258, "step": 56480 }, { "grad_norm": 0.28423571586608887, "learning_rate": 4.34234211462305e-05, "loss": 0.0241, "step": 56490 }, { "grad_norm": 0.2729843854904175, "learning_rate": 4.340703046172213e-05, "loss": 0.0285, "step": 56500 }, { "grad_norm": 0.2961578667163849, "learning_rate": 4.339064049821097e-05, "loss": 0.024, "step": 56510 }, { "grad_norm": 0.3259829580783844, "learning_rate": 4.3374251257489446e-05, "loss": 0.0297, "step": 56520 }, { "grad_norm": 0.198843851685524, "learning_rate": 4.335786274134984e-05, "loss": 0.017, "step": 56530 }, { "grad_norm": 0.4173210859298706, "learning_rate": 4.334147495158438e-05, "loss": 0.0262, "step": 56540 }, { "grad_norm": 0.29501867294311523, "learning_rate": 4.33250878899852e-05, "loss": 0.0222, "step": 56550 }, { "grad_norm": 0.24694067239761353, "learning_rate": 4.330870155834437e-05, "loss": 0.0216, "step": 56560 }, { "grad_norm": 0.33496713638305664, "learning_rate": 4.3292315958453884e-05, "loss": 0.0183, "step": 56570 }, { "grad_norm": 0.3172733783721924, "learning_rate": 4.3275931092105646e-05, "loss": 0.028, "step": 56580 }, { "grad_norm": 0.20883965492248535, "learning_rate": 4.325954696109147e-05, "loss": 0.0192, "step": 56590 }, { "grad_norm": 0.2592051029205322, "learning_rate": 4.32431635672031e-05, "loss": 0.0257, "step": 56600 }, { "grad_norm": 0.31130197644233704, "learning_rate": 4.3226780912232214e-05, "loss": 0.026, "step": 56610 }, { "grad_norm": 0.5578295588493347, "learning_rate": 4.321039899797036e-05, "loss": 0.0199, "step": 56620 }, { "grad_norm": 0.4082328975200653, "learning_rate": 4.319401782620908e-05, "loss": 0.0208, "step": 56630 }, { "grad_norm": 0.24880541861057281, "learning_rate": 4.317763739873978e-05, "loss": 0.0321, "step": 56640 }, { "grad_norm": 0.46210435032844543, "learning_rate": 4.31612577173538e-05, "loss": 0.0197, "step": 56650 }, { "grad_norm": 0.1669134944677353, "learning_rate": 4.314487878384239e-05, "loss": 0.0244, "step": 56660 }, { "grad_norm": 0.4503347873687744, "learning_rate": 4.3128500599996736e-05, "loss": 0.0314, "step": 56670 }, { "grad_norm": 0.3531939685344696, "learning_rate": 4.31121231676079e-05, "loss": 0.0221, "step": 56680 }, { "grad_norm": 0.2239570915699005, "learning_rate": 4.309574648846694e-05, "loss": 0.0162, "step": 56690 }, { "grad_norm": 0.22198878228664398, "learning_rate": 4.3079370564364755e-05, "loss": 0.0225, "step": 56700 }, { "grad_norm": 0.225915789604187, "learning_rate": 4.306299539709221e-05, "loss": 0.0216, "step": 56710 }, { "grad_norm": 0.31744441390037537, "learning_rate": 4.3046620988440046e-05, "loss": 0.0185, "step": 56720 }, { "grad_norm": 0.35472720861434937, "learning_rate": 4.303024734019897e-05, "loss": 0.0181, "step": 56730 }, { "grad_norm": 0.21762320399284363, "learning_rate": 4.3013874454159565e-05, "loss": 0.0219, "step": 56740 }, { "grad_norm": 0.5758216381072998, "learning_rate": 4.299750233211233e-05, "loss": 0.0257, "step": 56750 }, { "grad_norm": 0.19770033657550812, "learning_rate": 4.2981130975847715e-05, "loss": 0.0202, "step": 56760 }, { "grad_norm": 0.2486880123615265, "learning_rate": 4.2964760387156076e-05, "loss": 0.0257, "step": 56770 }, { "grad_norm": 0.38830989599227905, "learning_rate": 4.294839056782767e-05, "loss": 0.028, "step": 56780 }, { "grad_norm": 0.20379382371902466, "learning_rate": 4.293202151965266e-05, "loss": 0.0215, "step": 56790 }, { "grad_norm": 0.19495202600955963, "learning_rate": 4.291565324442115e-05, "loss": 0.0226, "step": 56800 }, { "grad_norm": 0.35646066069602966, "learning_rate": 4.2899285743923135e-05, "loss": 0.0212, "step": 56810 }, { "grad_norm": 0.25275856256484985, "learning_rate": 4.288291901994857e-05, "loss": 0.021, "step": 56820 }, { "grad_norm": 0.224105104804039, "learning_rate": 4.286655307428729e-05, "loss": 0.0244, "step": 56830 }, { "grad_norm": 0.31284961104393005, "learning_rate": 4.285018790872903e-05, "loss": 0.0233, "step": 56840 }, { "grad_norm": 0.3378584384918213, "learning_rate": 4.283382352506347e-05, "loss": 0.0235, "step": 56850 }, { "grad_norm": 0.30886971950531006, "learning_rate": 4.2817459925080205e-05, "loss": 0.0318, "step": 56860 }, { "grad_norm": 0.30253851413726807, "learning_rate": 4.28010971105687e-05, "loss": 0.0233, "step": 56870 }, { "grad_norm": 0.3812732994556427, "learning_rate": 4.278473508331841e-05, "loss": 0.0208, "step": 56880 }, { "grad_norm": 0.34280407428741455, "learning_rate": 4.276837384511864e-05, "loss": 0.0243, "step": 56890 }, { "grad_norm": 0.3345496952533722, "learning_rate": 4.275201339775864e-05, "loss": 0.0175, "step": 56900 }, { "grad_norm": 0.7793554067611694, "learning_rate": 4.273565374302756e-05, "loss": 0.0299, "step": 56910 }, { "grad_norm": 0.2931651473045349, "learning_rate": 4.2719294882714454e-05, "loss": 0.0238, "step": 56920 }, { "grad_norm": 0.17636528611183167, "learning_rate": 4.270293681860831e-05, "loss": 0.0189, "step": 56930 }, { "grad_norm": 0.2256781905889511, "learning_rate": 4.268657955249802e-05, "loss": 0.0183, "step": 56940 }, { "grad_norm": 0.22051815688610077, "learning_rate": 4.267022308617241e-05, "loss": 0.0237, "step": 56950 }, { "grad_norm": 0.26854807138442993, "learning_rate": 4.265386742142018e-05, "loss": 0.0304, "step": 56960 }, { "grad_norm": 0.3937433063983917, "learning_rate": 4.263751256002996e-05, "loss": 0.0291, "step": 56970 }, { "grad_norm": 0.2870316803455353, "learning_rate": 4.26211585037903e-05, "loss": 0.0264, "step": 56980 }, { "grad_norm": 0.18596011400222778, "learning_rate": 4.2604805254489656e-05, "loss": 0.0264, "step": 56990 }, { "grad_norm": 0.46213045716285706, "learning_rate": 4.258845281391638e-05, "loss": 0.0277, "step": 57000 }, { "grad_norm": 0.19605529308319092, "learning_rate": 4.257210118385876e-05, "loss": 0.0311, "step": 57010 }, { "grad_norm": 0.34527647495269775, "learning_rate": 4.2555750366104994e-05, "loss": 0.0225, "step": 57020 }, { "grad_norm": 0.27244412899017334, "learning_rate": 4.2539400362443184e-05, "loss": 0.0205, "step": 57030 }, { "grad_norm": 0.22273357212543488, "learning_rate": 4.252305117466133e-05, "loss": 0.0209, "step": 57040 }, { "grad_norm": 0.16272851824760437, "learning_rate": 4.2506702804547356e-05, "loss": 0.0256, "step": 57050 }, { "grad_norm": 0.16495773196220398, "learning_rate": 4.249035525388907e-05, "loss": 0.0213, "step": 57060 }, { "grad_norm": 0.3157229721546173, "learning_rate": 4.247400852447427e-05, "loss": 0.0243, "step": 57070 }, { "grad_norm": 0.2239651083946228, "learning_rate": 4.245766261809059e-05, "loss": 0.0203, "step": 57080 }, { "grad_norm": 0.3098433017730713, "learning_rate": 4.244131753652558e-05, "loss": 0.0243, "step": 57090 }, { "grad_norm": 0.2649187743663788, "learning_rate": 4.242497328156672e-05, "loss": 0.0196, "step": 57100 }, { "grad_norm": 0.3710934519767761, "learning_rate": 4.240862985500139e-05, "loss": 0.0313, "step": 57110 }, { "grad_norm": 0.2565908133983612, "learning_rate": 4.2392287258616894e-05, "loss": 0.0253, "step": 57120 }, { "grad_norm": 0.15390998125076294, "learning_rate": 4.2375945494200396e-05, "loss": 0.0193, "step": 57130 }, { "grad_norm": 0.293140709400177, "learning_rate": 4.2359604563539054e-05, "loss": 0.0214, "step": 57140 }, { "grad_norm": 0.1910610944032669, "learning_rate": 4.2343264468419885e-05, "loss": 0.0262, "step": 57150 }, { "grad_norm": 0.2232954055070877, "learning_rate": 4.232692521062979e-05, "loss": 0.0209, "step": 57160 }, { "grad_norm": 0.24792736768722534, "learning_rate": 4.231058679195562e-05, "loss": 0.019, "step": 57170 }, { "grad_norm": 0.24560990929603577, "learning_rate": 4.229424921418411e-05, "loss": 0.0211, "step": 57180 }, { "grad_norm": 0.2396773248910904, "learning_rate": 4.2277912479101916e-05, "loss": 0.0252, "step": 57190 }, { "grad_norm": 0.22844500839710236, "learning_rate": 4.226157658849562e-05, "loss": 0.0237, "step": 57200 }, { "grad_norm": 0.2577117681503296, "learning_rate": 4.2245241544151664e-05, "loss": 0.0263, "step": 57210 }, { "grad_norm": 0.17653167247772217, "learning_rate": 4.2228907347856425e-05, "loss": 0.0199, "step": 57220 }, { "grad_norm": 0.21254615485668182, "learning_rate": 4.2212574001396214e-05, "loss": 0.0261, "step": 57230 }, { "grad_norm": 0.25200971961021423, "learning_rate": 4.21962415065572e-05, "loss": 0.0195, "step": 57240 }, { "grad_norm": 0.22155267000198364, "learning_rate": 4.217990986512548e-05, "loss": 0.0288, "step": 57250 }, { "grad_norm": 0.768805742263794, "learning_rate": 4.216357907888704e-05, "loss": 0.0331, "step": 57260 }, { "grad_norm": 0.2812502384185791, "learning_rate": 4.2147249149627824e-05, "loss": 0.0216, "step": 57270 }, { "grad_norm": 0.27459654211997986, "learning_rate": 4.213092007913364e-05, "loss": 0.0261, "step": 57280 }, { "grad_norm": 0.579148530960083, "learning_rate": 4.2114591869190205e-05, "loss": 0.0297, "step": 57290 }, { "grad_norm": 0.1579338014125824, "learning_rate": 4.209826452158315e-05, "loss": 0.0313, "step": 57300 }, { "grad_norm": 0.36227548122406006, "learning_rate": 4.2081938038098e-05, "loss": 0.0231, "step": 57310 }, { "grad_norm": 0.3174285292625427, "learning_rate": 4.2065612420520185e-05, "loss": 0.0305, "step": 57320 }, { "grad_norm": 0.2651410698890686, "learning_rate": 4.204928767063509e-05, "loss": 0.0219, "step": 57330 }, { "grad_norm": 0.24766859412193298, "learning_rate": 4.2032963790227944e-05, "loss": 0.0246, "step": 57340 }, { "grad_norm": 0.1785520315170288, "learning_rate": 4.2016640781083887e-05, "loss": 0.02, "step": 57350 }, { "grad_norm": 0.22850602865219116, "learning_rate": 4.2000318644988004e-05, "loss": 0.0312, "step": 57360 }, { "grad_norm": 0.3025439381599426, "learning_rate": 4.198399738372524e-05, "loss": 0.0264, "step": 57370 }, { "grad_norm": 0.28625571727752686, "learning_rate": 4.196767699908045e-05, "loss": 0.019, "step": 57380 }, { "grad_norm": 0.25294065475463867, "learning_rate": 4.195135749283845e-05, "loss": 0.0218, "step": 57390 }, { "grad_norm": 0.40347009897232056, "learning_rate": 4.1935038866783894e-05, "loss": 0.0216, "step": 57400 }, { "grad_norm": 0.22067692875862122, "learning_rate": 4.1918721122701355e-05, "loss": 0.0206, "step": 57410 }, { "grad_norm": 0.3793669044971466, "learning_rate": 4.190240426237533e-05, "loss": 0.0258, "step": 57420 }, { "grad_norm": 0.5513628125190735, "learning_rate": 4.188608828759019e-05, "loss": 0.025, "step": 57430 }, { "grad_norm": 0.22377541661262512, "learning_rate": 4.1869773200130246e-05, "loss": 0.0306, "step": 57440 }, { "grad_norm": 0.3107903301715851, "learning_rate": 4.185345900177967e-05, "loss": 0.0295, "step": 57450 }, { "grad_norm": 0.4868868291378021, "learning_rate": 4.183714569432258e-05, "loss": 0.0269, "step": 57460 }, { "grad_norm": 0.18771100044250488, "learning_rate": 4.182083327954295e-05, "loss": 0.0191, "step": 57470 }, { "grad_norm": 0.33721521496772766, "learning_rate": 4.1804521759224714e-05, "loss": 0.0267, "step": 57480 }, { "grad_norm": 0.2502797245979309, "learning_rate": 4.178821113515165e-05, "loss": 0.026, "step": 57490 }, { "grad_norm": 0.29557928442955017, "learning_rate": 4.1771901409107476e-05, "loss": 0.0266, "step": 57500 }, { "grad_norm": 0.21145999431610107, "learning_rate": 4.175559258287577e-05, "loss": 0.0267, "step": 57510 }, { "grad_norm": 0.18878251314163208, "learning_rate": 4.173928465824008e-05, "loss": 0.0248, "step": 57520 }, { "grad_norm": 0.19502602517604828, "learning_rate": 4.17229776369838e-05, "loss": 0.0204, "step": 57530 }, { "grad_norm": 0.24937883019447327, "learning_rate": 4.170667152089025e-05, "loss": 0.0266, "step": 57540 }, { "grad_norm": 0.3872062563896179, "learning_rate": 4.1690366311742626e-05, "loss": 0.0208, "step": 57550 }, { "grad_norm": 0.18252670764923096, "learning_rate": 4.1674062011324055e-05, "loss": 0.0231, "step": 57560 }, { "grad_norm": 0.25299277901649475, "learning_rate": 4.165775862141752e-05, "loss": 0.0215, "step": 57570 }, { "grad_norm": 0.24314063787460327, "learning_rate": 4.164145614380598e-05, "loss": 0.0182, "step": 57580 }, { "grad_norm": 0.254408061504364, "learning_rate": 4.162515458027223e-05, "loss": 0.0211, "step": 57590 }, { "grad_norm": 0.2998355031013489, "learning_rate": 4.160885393259899e-05, "loss": 0.0264, "step": 57600 }, { "grad_norm": 0.25871342420578003, "learning_rate": 4.159255420256886e-05, "loss": 0.0205, "step": 57610 }, { "grad_norm": 0.3404320478439331, "learning_rate": 4.157625539196436e-05, "loss": 0.0229, "step": 57620 }, { "grad_norm": 0.19632619619369507, "learning_rate": 4.155995750256791e-05, "loss": 0.0274, "step": 57630 }, { "grad_norm": 0.476927787065506, "learning_rate": 4.15436605361618e-05, "loss": 0.0205, "step": 57640 }, { "grad_norm": 0.39280468225479126, "learning_rate": 4.152736449452827e-05, "loss": 0.0318, "step": 57650 }, { "grad_norm": 0.2909305989742279, "learning_rate": 4.151106937944943e-05, "loss": 0.023, "step": 57660 }, { "grad_norm": 0.28325483202934265, "learning_rate": 4.149477519270727e-05, "loss": 0.0259, "step": 57670 }, { "grad_norm": 0.32097360491752625, "learning_rate": 4.147848193608371e-05, "loss": 0.0211, "step": 57680 }, { "grad_norm": 0.2740035057067871, "learning_rate": 4.146218961136056e-05, "loss": 0.0239, "step": 57690 }, { "grad_norm": 0.3305578827857971, "learning_rate": 4.14458982203195e-05, "loss": 0.025, "step": 57700 }, { "grad_norm": 0.21529392898082733, "learning_rate": 4.1429607764742165e-05, "loss": 0.0182, "step": 57710 }, { "grad_norm": 0.3754701316356659, "learning_rate": 4.141331824641003e-05, "loss": 0.032, "step": 57720 }, { "grad_norm": 0.2679893970489502, "learning_rate": 4.139702966710452e-05, "loss": 0.0223, "step": 57730 }, { "grad_norm": 0.31342390179634094, "learning_rate": 4.13807420286069e-05, "loss": 0.024, "step": 57740 }, { "grad_norm": 0.2320673018693924, "learning_rate": 4.136445533269837e-05, "loss": 0.0207, "step": 57750 }, { "grad_norm": 0.2291593849658966, "learning_rate": 4.134816958116003e-05, "loss": 0.0255, "step": 57760 }, { "grad_norm": 0.47806355357170105, "learning_rate": 4.133188477577283e-05, "loss": 0.024, "step": 57770 }, { "grad_norm": 0.2944481670856476, "learning_rate": 4.131560091831771e-05, "loss": 0.0191, "step": 57780 }, { "grad_norm": 0.24164724349975586, "learning_rate": 4.1299318010575405e-05, "loss": 0.0217, "step": 57790 }, { "grad_norm": 0.3206495940685272, "learning_rate": 4.128303605432661e-05, "loss": 0.0298, "step": 57800 }, { "grad_norm": 0.2848871350288391, "learning_rate": 4.126675505135188e-05, "loss": 0.0222, "step": 57810 }, { "grad_norm": 0.29970693588256836, "learning_rate": 4.125047500343169e-05, "loss": 0.0257, "step": 57820 }, { "grad_norm": 0.20849855244159698, "learning_rate": 4.1234195912346376e-05, "loss": 0.0225, "step": 57830 }, { "grad_norm": 0.238628089427948, "learning_rate": 4.1217917779876235e-05, "loss": 0.025, "step": 57840 }, { "grad_norm": 0.4900509715080261, "learning_rate": 4.120164060780141e-05, "loss": 0.0336, "step": 57850 }, { "grad_norm": 0.29872602224349976, "learning_rate": 4.1185364397901935e-05, "loss": 0.0277, "step": 57860 }, { "grad_norm": 0.2556122839450836, "learning_rate": 4.116908915195776e-05, "loss": 0.0188, "step": 57870 }, { "grad_norm": 0.22234852612018585, "learning_rate": 4.115281487174872e-05, "loss": 0.0224, "step": 57880 }, { "grad_norm": 0.3162917196750641, "learning_rate": 4.113654155905452e-05, "loss": 0.0213, "step": 57890 }, { "grad_norm": 0.23997969925403595, "learning_rate": 4.112026921565483e-05, "loss": 0.025, "step": 57900 }, { "grad_norm": 0.18091394007205963, "learning_rate": 4.1103997843329147e-05, "loss": 0.0192, "step": 57910 }, { "grad_norm": 0.2566729187965393, "learning_rate": 4.108772744385689e-05, "loss": 0.045, "step": 57920 }, { "grad_norm": 0.17875367403030396, "learning_rate": 4.1071458019017346e-05, "loss": 0.0254, "step": 57930 }, { "grad_norm": 0.2898063063621521, "learning_rate": 4.1055189570589734e-05, "loss": 0.0241, "step": 57940 }, { "grad_norm": 0.22261521220207214, "learning_rate": 4.103892210035315e-05, "loss": 0.0194, "step": 57950 }, { "grad_norm": 0.3776737153530121, "learning_rate": 4.102265561008656e-05, "loss": 0.0271, "step": 57960 }, { "grad_norm": 0.3113832175731659, "learning_rate": 4.100639010156885e-05, "loss": 0.0208, "step": 57970 }, { "grad_norm": 0.22032633423805237, "learning_rate": 4.099012557657882e-05, "loss": 0.0236, "step": 57980 }, { "grad_norm": 0.2303338646888733, "learning_rate": 4.09738620368951e-05, "loss": 0.019, "step": 57990 }, { "grad_norm": 0.28434044122695923, "learning_rate": 4.095759948429626e-05, "loss": 0.0266, "step": 58000 }, { "grad_norm": 0.2063591182231903, "learning_rate": 4.094133792056075e-05, "loss": 0.0171, "step": 58010 }, { "grad_norm": 0.5671514272689819, "learning_rate": 4.092507734746688e-05, "loss": 0.0282, "step": 58020 }, { "grad_norm": 0.26574331521987915, "learning_rate": 4.090881776679293e-05, "loss": 0.0219, "step": 58030 }, { "grad_norm": 0.28700271248817444, "learning_rate": 4.0892559180316994e-05, "loss": 0.0223, "step": 58040 }, { "grad_norm": 0.1720285564661026, "learning_rate": 4.08763015898171e-05, "loss": 0.0216, "step": 58050 }, { "grad_norm": 0.27510198950767517, "learning_rate": 4.0860044997071145e-05, "loss": 0.0209, "step": 58060 }, { "grad_norm": 0.1875491589307785, "learning_rate": 4.084378940385693e-05, "loss": 0.0254, "step": 58070 }, { "grad_norm": 0.30186527967453003, "learning_rate": 4.082753481195211e-05, "loss": 0.0235, "step": 58080 }, { "grad_norm": 0.32556620240211487, "learning_rate": 4.0811281223134324e-05, "loss": 0.0228, "step": 58090 }, { "grad_norm": 0.3637300729751587, "learning_rate": 4.0795028639181e-05, "loss": 0.0314, "step": 58100 }, { "grad_norm": 0.5531733632087708, "learning_rate": 4.077877706186951e-05, "loss": 0.0236, "step": 58110 }, { "grad_norm": 0.3751230537891388, "learning_rate": 4.076252649297709e-05, "loss": 0.0271, "step": 58120 }, { "grad_norm": 0.23893912136554718, "learning_rate": 4.07462769342809e-05, "loss": 0.0185, "step": 58130 }, { "grad_norm": 0.301880806684494, "learning_rate": 4.073002838755793e-05, "loss": 0.0236, "step": 58140 }, { "grad_norm": 0.20387929677963257, "learning_rate": 4.0713780854585115e-05, "loss": 0.0184, "step": 58150 }, { "grad_norm": 0.29691484570503235, "learning_rate": 4.069753433713927e-05, "loss": 0.0217, "step": 58160 }, { "grad_norm": 0.27500665187835693, "learning_rate": 4.068128883699709e-05, "loss": 0.0185, "step": 58170 }, { "grad_norm": 0.2024846374988556, "learning_rate": 4.0665044355935155e-05, "loss": 0.0215, "step": 58180 }, { "grad_norm": 0.2045551985502243, "learning_rate": 4.064880089572993e-05, "loss": 0.022, "step": 58190 }, { "grad_norm": 0.32109585404396057, "learning_rate": 4.063255845815779e-05, "loss": 0.0308, "step": 58200 }, { "grad_norm": 0.17363615334033966, "learning_rate": 4.0616317044994945e-05, "loss": 0.0179, "step": 58210 }, { "grad_norm": 0.2929372191429138, "learning_rate": 4.0600076658017585e-05, "loss": 0.025, "step": 58220 }, { "grad_norm": 0.4153757095336914, "learning_rate": 4.0583837299001706e-05, "loss": 0.025, "step": 58230 }, { "grad_norm": 0.2951405942440033, "learning_rate": 4.056759896972322e-05, "loss": 0.0221, "step": 58240 }, { "grad_norm": 0.20316214859485626, "learning_rate": 4.055136167195793e-05, "loss": 0.0249, "step": 58250 }, { "grad_norm": 0.20381902158260345, "learning_rate": 4.053512540748152e-05, "loss": 0.0241, "step": 58260 }, { "grad_norm": 0.37069571018218994, "learning_rate": 4.051889017806955e-05, "loss": 0.0272, "step": 58270 }, { "grad_norm": 0.293846070766449, "learning_rate": 4.0502655985497516e-05, "loss": 0.0173, "step": 58280 }, { "grad_norm": 0.22784660756587982, "learning_rate": 4.048642283154074e-05, "loss": 0.0256, "step": 58290 }, { "grad_norm": 0.4747799336910248, "learning_rate": 4.0470190717974464e-05, "loss": 0.0228, "step": 58300 }, { "grad_norm": 0.9362044334411621, "learning_rate": 4.045395964657381e-05, "loss": 0.0259, "step": 58310 }, { "grad_norm": 0.6252413392066956, "learning_rate": 4.043772961911376e-05, "loss": 0.0263, "step": 58320 }, { "grad_norm": 0.24666531383991241, "learning_rate": 4.042150063736924e-05, "loss": 0.0265, "step": 58330 }, { "grad_norm": 0.31454575061798096, "learning_rate": 4.0405272703114976e-05, "loss": 0.0228, "step": 58340 }, { "grad_norm": 0.3348636031150818, "learning_rate": 4.038904581812569e-05, "loss": 0.0215, "step": 58350 }, { "grad_norm": 0.17610110342502594, "learning_rate": 4.03728199841759e-05, "loss": 0.0229, "step": 58360 }, { "grad_norm": 0.25218772888183594, "learning_rate": 4.0356595203040044e-05, "loss": 0.019, "step": 58370 }, { "grad_norm": 0.3436114192008972, "learning_rate": 4.0340371476492426e-05, "loss": 0.0172, "step": 58380 }, { "grad_norm": 0.20424707233905792, "learning_rate": 4.032414880630727e-05, "loss": 0.0265, "step": 58390 }, { "grad_norm": 0.3195939362049103, "learning_rate": 4.0307927194258646e-05, "loss": 0.0199, "step": 58400 }, { "grad_norm": 0.15080669522285461, "learning_rate": 4.029170664212054e-05, "loss": 0.0207, "step": 58410 }, { "grad_norm": 0.1497398018836975, "learning_rate": 4.0275487151666794e-05, "loss": 0.022, "step": 58420 }, { "grad_norm": 0.30858591198921204, "learning_rate": 4.025926872467116e-05, "loss": 0.0309, "step": 58430 }, { "grad_norm": 0.3447875678539276, "learning_rate": 4.024305136290725e-05, "loss": 0.0233, "step": 58440 }, { "grad_norm": 0.22821931540966034, "learning_rate": 4.022683506814858e-05, "loss": 0.0213, "step": 58450 }, { "grad_norm": 0.24186427891254425, "learning_rate": 4.021061984216853e-05, "loss": 0.0181, "step": 58460 }, { "grad_norm": 0.2587270736694336, "learning_rate": 4.019440568674035e-05, "loss": 0.0228, "step": 58470 }, { "grad_norm": 0.2255181521177292, "learning_rate": 4.017819260363724e-05, "loss": 0.0234, "step": 58480 }, { "grad_norm": 0.24280555546283722, "learning_rate": 4.016198059463223e-05, "loss": 0.0272, "step": 58490 }, { "grad_norm": 0.2558560073375702, "learning_rate": 4.0145769661498233e-05, "loss": 0.0212, "step": 58500 }, { "grad_norm": 0.2332518994808197, "learning_rate": 4.012955980600804e-05, "loss": 0.019, "step": 58510 }, { "grad_norm": 0.27769193053245544, "learning_rate": 4.0113351029934355e-05, "loss": 0.0248, "step": 58520 }, { "grad_norm": 0.32136914134025574, "learning_rate": 4.0097143335049714e-05, "loss": 0.0275, "step": 58530 }, { "grad_norm": 0.24517062306404114, "learning_rate": 4.0080936723126606e-05, "loss": 0.023, "step": 58540 }, { "grad_norm": 0.2780115604400635, "learning_rate": 4.006473119593735e-05, "loss": 0.0254, "step": 58550 }, { "grad_norm": 0.2872839868068695, "learning_rate": 4.004852675525415e-05, "loss": 0.0225, "step": 58560 }, { "grad_norm": 0.20604637265205383, "learning_rate": 4.00323234028491e-05, "loss": 0.0215, "step": 58570 }, { "grad_norm": 0.17962110042572021, "learning_rate": 4.0016121140494186e-05, "loss": 0.0257, "step": 58580 }, { "grad_norm": 0.28319793939590454, "learning_rate": 3.999991996996122e-05, "loss": 0.0222, "step": 58590 }, { "grad_norm": 0.19561457633972168, "learning_rate": 3.9983719893022e-05, "loss": 0.0227, "step": 58600 }, { "grad_norm": 0.1607777178287506, "learning_rate": 3.996752091144812e-05, "loss": 0.0233, "step": 58610 }, { "grad_norm": 0.6686266660690308, "learning_rate": 3.9951323027011055e-05, "loss": 0.0261, "step": 58620 }, { "grad_norm": 0.22429095208644867, "learning_rate": 3.993512624148219e-05, "loss": 0.022, "step": 58630 }, { "grad_norm": 0.7766305804252625, "learning_rate": 3.99189305566328e-05, "loss": 0.0184, "step": 58640 }, { "grad_norm": 0.18123024702072144, "learning_rate": 3.990273597423401e-05, "loss": 0.0208, "step": 58650 }, { "grad_norm": 0.18381622433662415, "learning_rate": 3.988654249605681e-05, "loss": 0.0233, "step": 58660 }, { "grad_norm": 0.18882551789283752, "learning_rate": 3.987035012387213e-05, "loss": 0.0159, "step": 58670 }, { "grad_norm": 0.22498387098312378, "learning_rate": 3.9854158859450735e-05, "loss": 0.022, "step": 58680 }, { "grad_norm": 0.2715042233467102, "learning_rate": 3.983796870456328e-05, "loss": 0.03, "step": 58690 }, { "grad_norm": 0.24634547531604767, "learning_rate": 3.982177966098028e-05, "loss": 0.0189, "step": 58700 }, { "grad_norm": 0.1857643723487854, "learning_rate": 3.980559173047216e-05, "loss": 0.0202, "step": 58710 }, { "grad_norm": 0.28473663330078125, "learning_rate": 3.9789404914809176e-05, "loss": 0.024, "step": 58720 }, { "grad_norm": 0.19626179337501526, "learning_rate": 3.9773219215761544e-05, "loss": 0.019, "step": 58730 }, { "grad_norm": 0.22233286499977112, "learning_rate": 3.975703463509928e-05, "loss": 0.0223, "step": 58740 }, { "grad_norm": 0.20718804001808167, "learning_rate": 3.974085117459231e-05, "loss": 0.0239, "step": 58750 }, { "grad_norm": 0.3353211283683777, "learning_rate": 3.972466883601043e-05, "loss": 0.0208, "step": 58760 }, { "grad_norm": 0.16877669095993042, "learning_rate": 3.9708487621123316e-05, "loss": 0.0185, "step": 58770 }, { "grad_norm": 0.24763235449790955, "learning_rate": 3.969230753170051e-05, "loss": 0.024, "step": 58780 }, { "grad_norm": 0.2392110824584961, "learning_rate": 3.967612856951146e-05, "loss": 0.0231, "step": 58790 }, { "grad_norm": 0.3890983462333679, "learning_rate": 3.9659950736325476e-05, "loss": 0.0273, "step": 58800 }, { "grad_norm": 0.2594270408153534, "learning_rate": 3.964377403391174e-05, "loss": 0.0441, "step": 58810 }, { "grad_norm": 0.18984127044677734, "learning_rate": 3.9627598464039303e-05, "loss": 0.0325, "step": 58820 }, { "grad_norm": 0.2888786196708679, "learning_rate": 3.9611424028477096e-05, "loss": 0.0267, "step": 58830 }, { "grad_norm": 0.23883841931819916, "learning_rate": 3.959525072899394e-05, "loss": 0.0323, "step": 58840 }, { "grad_norm": 0.28020626306533813, "learning_rate": 3.9579078567358506e-05, "loss": 0.0208, "step": 58850 }, { "grad_norm": 0.3590371608734131, "learning_rate": 3.956290754533939e-05, "loss": 0.0201, "step": 58860 }, { "grad_norm": 0.22391708195209503, "learning_rate": 3.954673766470501e-05, "loss": 0.0269, "step": 58870 }, { "grad_norm": 0.1951218843460083, "learning_rate": 3.9530568927223674e-05, "loss": 0.022, "step": 58880 }, { "grad_norm": 0.20221403241157532, "learning_rate": 3.95144013346636e-05, "loss": 0.0213, "step": 58890 }, { "grad_norm": 0.5719707012176514, "learning_rate": 3.949823488879282e-05, "loss": 0.02, "step": 58900 }, { "grad_norm": 0.356263667345047, "learning_rate": 3.9482069591379276e-05, "loss": 0.0315, "step": 58910 }, { "grad_norm": 0.22587035596370697, "learning_rate": 3.9465905444190784e-05, "loss": 0.0189, "step": 58920 }, { "grad_norm": 0.2659224271774292, "learning_rate": 3.944974244899505e-05, "loss": 0.0268, "step": 58930 }, { "grad_norm": 0.521331787109375, "learning_rate": 3.9433580607559616e-05, "loss": 0.0248, "step": 58940 }, { "grad_norm": 0.30377084016799927, "learning_rate": 3.941741992165193e-05, "loss": 0.0379, "step": 58950 }, { "grad_norm": 0.335101455450058, "learning_rate": 3.940126039303928e-05, "loss": 0.0226, "step": 58960 }, { "grad_norm": 0.24264217913150787, "learning_rate": 3.938510202348886e-05, "loss": 0.0202, "step": 58970 }, { "grad_norm": 0.41086193919181824, "learning_rate": 3.93689448147677e-05, "loss": 0.0285, "step": 58980 }, { "grad_norm": 0.27207890152931213, "learning_rate": 3.935278876864278e-05, "loss": 0.0268, "step": 58990 }, { "grad_norm": 0.24910055100917816, "learning_rate": 3.933663388688087e-05, "loss": 0.0289, "step": 59000 }, { "grad_norm": 0.2089831680059433, "learning_rate": 3.9320480171248645e-05, "loss": 0.0213, "step": 59010 }, { "grad_norm": 0.4105614721775055, "learning_rate": 3.930432762351265e-05, "loss": 0.0243, "step": 59020 }, { "grad_norm": 0.16144511103630066, "learning_rate": 3.928817624543931e-05, "loss": 0.0221, "step": 59030 }, { "grad_norm": 0.21101851761341095, "learning_rate": 3.9272026038794885e-05, "loss": 0.0202, "step": 59040 }, { "grad_norm": 0.3260725736618042, "learning_rate": 3.9255877005345586e-05, "loss": 0.0266, "step": 59050 }, { "grad_norm": 0.49694088101387024, "learning_rate": 3.923972914685743e-05, "loss": 0.0223, "step": 59060 }, { "grad_norm": 0.13982626795768738, "learning_rate": 3.922358246509631e-05, "loss": 0.0235, "step": 59070 }, { "grad_norm": 0.6781744360923767, "learning_rate": 3.920743696182801e-05, "loss": 0.0293, "step": 59080 }, { "grad_norm": 0.20947980880737305, "learning_rate": 3.9191292638818164e-05, "loss": 0.0262, "step": 59090 }, { "grad_norm": 0.27363696694374084, "learning_rate": 3.9175149497832316e-05, "loss": 0.0208, "step": 59100 }, { "grad_norm": 0.1989753693342209, "learning_rate": 3.915900754063584e-05, "loss": 0.0231, "step": 59110 }, { "grad_norm": 0.5780044794082642, "learning_rate": 3.9142866768994014e-05, "loss": 0.0186, "step": 59120 }, { "grad_norm": 0.25901344418525696, "learning_rate": 3.912672718467193e-05, "loss": 0.0344, "step": 59130 }, { "grad_norm": 0.4095786511898041, "learning_rate": 3.911058878943463e-05, "loss": 0.0219, "step": 59140 }, { "grad_norm": 0.31213703751564026, "learning_rate": 3.9094451585046974e-05, "loss": 0.0244, "step": 59150 }, { "grad_norm": 0.2506805956363678, "learning_rate": 3.90783155732737e-05, "loss": 0.0262, "step": 59160 }, { "grad_norm": 0.294456422328949, "learning_rate": 3.9062180755879373e-05, "loss": 0.0189, "step": 59170 }, { "grad_norm": 0.3073605000972748, "learning_rate": 3.904604713462855e-05, "loss": 0.0211, "step": 59180 }, { "grad_norm": 0.2719675898551941, "learning_rate": 3.902991471128554e-05, "loss": 0.0209, "step": 59190 }, { "grad_norm": 0.18093976378440857, "learning_rate": 3.901378348761456e-05, "loss": 0.0192, "step": 59200 }, { "grad_norm": 0.19052141904830933, "learning_rate": 3.899765346537971e-05, "loss": 0.0214, "step": 59210 }, { "grad_norm": 0.2209140956401825, "learning_rate": 3.898152464634493e-05, "loss": 0.0211, "step": 59220 }, { "grad_norm": 0.16403113305568695, "learning_rate": 3.896539703227402e-05, "loss": 0.0204, "step": 59230 }, { "grad_norm": 0.37183675169944763, "learning_rate": 3.894927062493073e-05, "loss": 0.0199, "step": 59240 }, { "grad_norm": 0.3030518889427185, "learning_rate": 3.893314542607859e-05, "loss": 0.0172, "step": 59250 }, { "grad_norm": 0.236222505569458, "learning_rate": 3.891702143748103e-05, "loss": 0.0233, "step": 59260 }, { "grad_norm": 0.4181675910949707, "learning_rate": 3.890089866090133e-05, "loss": 0.0232, "step": 59270 }, { "grad_norm": 0.381332129240036, "learning_rate": 3.888477709810267e-05, "loss": 0.0286, "step": 59280 }, { "grad_norm": 0.24986135959625244, "learning_rate": 3.886865675084806e-05, "loss": 0.0224, "step": 59290 }, { "grad_norm": 0.23840713500976562, "learning_rate": 3.885253762090043e-05, "loss": 0.0203, "step": 59300 }, { "grad_norm": 0.34736043214797974, "learning_rate": 3.8836419710022534e-05, "loss": 0.0282, "step": 59310 }, { "grad_norm": 0.2906090319156647, "learning_rate": 3.882030301997698e-05, "loss": 0.0221, "step": 59320 }, { "grad_norm": 0.35817450284957886, "learning_rate": 3.880418755252629e-05, "loss": 0.0242, "step": 59330 }, { "grad_norm": 0.5216262936592102, "learning_rate": 3.87880733094328e-05, "loss": 0.0213, "step": 59340 }, { "grad_norm": 0.21884092688560486, "learning_rate": 3.877196029245877e-05, "loss": 0.0175, "step": 59350 }, { "grad_norm": 0.24930354952812195, "learning_rate": 3.875584850336627e-05, "loss": 0.0214, "step": 59360 }, { "grad_norm": 0.36653611063957214, "learning_rate": 3.873973794391728e-05, "loss": 0.0227, "step": 59370 }, { "grad_norm": 0.1174183189868927, "learning_rate": 3.872362861587361e-05, "loss": 0.0207, "step": 59380 }, { "grad_norm": 0.2620244026184082, "learning_rate": 3.870752052099698e-05, "loss": 0.0225, "step": 59390 }, { "grad_norm": 0.4255024194717407, "learning_rate": 3.8691413661048916e-05, "loss": 0.0196, "step": 59400 }, { "grad_norm": 0.370492160320282, "learning_rate": 3.867530803779085e-05, "loss": 0.0226, "step": 59410 }, { "grad_norm": 0.23582859337329865, "learning_rate": 3.8659203652984054e-05, "loss": 0.021, "step": 59420 }, { "grad_norm": 0.33655110001564026, "learning_rate": 3.864310050838972e-05, "loss": 0.0208, "step": 59430 }, { "grad_norm": 0.2557816207408905, "learning_rate": 3.8626998605768835e-05, "loss": 0.0172, "step": 59440 }, { "grad_norm": 0.2509121298789978, "learning_rate": 3.8610897946882294e-05, "loss": 0.0213, "step": 59450 }, { "grad_norm": 0.27723875641822815, "learning_rate": 3.859479853349082e-05, "loss": 0.0219, "step": 59460 }, { "grad_norm": 0.23829703032970428, "learning_rate": 3.857870036735503e-05, "loss": 0.0212, "step": 59470 }, { "grad_norm": 0.2416037768125534, "learning_rate": 3.856260345023538e-05, "loss": 0.0194, "step": 59480 }, { "grad_norm": 0.27371951937675476, "learning_rate": 3.854650778389224e-05, "loss": 0.0209, "step": 59490 }, { "grad_norm": 0.33083653450012207, "learning_rate": 3.8530413370085786e-05, "loss": 0.0244, "step": 59500 }, { "grad_norm": 0.16908542811870575, "learning_rate": 3.851432021057608e-05, "loss": 0.0154, "step": 59510 }, { "grad_norm": 0.1945725977420807, "learning_rate": 3.849822830712306e-05, "loss": 0.0252, "step": 59520 }, { "grad_norm": 0.29206037521362305, "learning_rate": 3.848213766148649e-05, "loss": 0.0209, "step": 59530 }, { "grad_norm": 0.4011744260787964, "learning_rate": 3.846604827542602e-05, "loss": 0.0205, "step": 59540 }, { "grad_norm": 0.7240447998046875, "learning_rate": 3.844996015070117e-05, "loss": 0.0324, "step": 59550 }, { "grad_norm": 0.23250794410705566, "learning_rate": 3.843387328907132e-05, "loss": 0.02, "step": 59560 }, { "grad_norm": 0.23537428677082062, "learning_rate": 3.8417787692295696e-05, "loss": 0.0251, "step": 59570 }, { "grad_norm": 0.24640683829784393, "learning_rate": 3.84017033621334e-05, "loss": 0.0209, "step": 59580 }, { "grad_norm": 0.14551299810409546, "learning_rate": 3.838562030034338e-05, "loss": 0.0162, "step": 59590 }, { "grad_norm": 0.31684452295303345, "learning_rate": 3.836953850868447e-05, "loss": 0.0297, "step": 59600 }, { "grad_norm": 1.1545335054397583, "learning_rate": 3.835345798891533e-05, "loss": 0.0204, "step": 59610 }, { "grad_norm": 0.2671886086463928, "learning_rate": 3.833737874279452e-05, "loss": 0.0275, "step": 59620 }, { "grad_norm": 0.1941949874162674, "learning_rate": 3.8321300772080433e-05, "loss": 0.0158, "step": 59630 }, { "grad_norm": 0.27880242466926575, "learning_rate": 3.8305224078531345e-05, "loss": 0.023, "step": 59640 }, { "grad_norm": 0.35301706194877625, "learning_rate": 3.828914866390537e-05, "loss": 0.0259, "step": 59650 }, { "grad_norm": 0.2626921236515045, "learning_rate": 3.827307452996048e-05, "loss": 0.0257, "step": 59660 }, { "grad_norm": 0.21049490571022034, "learning_rate": 3.825700167845454e-05, "loss": 0.0197, "step": 59670 }, { "grad_norm": 0.22213208675384521, "learning_rate": 3.8240930111145214e-05, "loss": 0.0264, "step": 59680 }, { "grad_norm": 0.20606301724910736, "learning_rate": 3.8224859829790116e-05, "loss": 0.0263, "step": 59690 }, { "grad_norm": 0.278466135263443, "learning_rate": 3.820879083614664e-05, "loss": 0.0321, "step": 59700 }, { "grad_norm": 0.2621876299381256, "learning_rate": 3.819272313197208e-05, "loss": 0.0225, "step": 59710 }, { "grad_norm": 0.3819223940372467, "learning_rate": 3.817665671902355e-05, "loss": 0.0189, "step": 59720 }, { "grad_norm": 0.18733426928520203, "learning_rate": 3.816059159905808e-05, "loss": 0.0248, "step": 59730 }, { "grad_norm": 0.19793684780597687, "learning_rate": 3.8144527773832476e-05, "loss": 0.0298, "step": 59740 }, { "grad_norm": 0.1874447613954544, "learning_rate": 3.812846524510352e-05, "loss": 0.0214, "step": 59750 }, { "grad_norm": 0.32387039065361023, "learning_rate": 3.811240401462775e-05, "loss": 0.0296, "step": 59760 }, { "grad_norm": 0.2702152729034424, "learning_rate": 3.809634408416162e-05, "loss": 0.0303, "step": 59770 }, { "grad_norm": 0.30210986733436584, "learning_rate": 3.8080285455461394e-05, "loss": 0.0335, "step": 59780 }, { "grad_norm": 0.6625841856002808, "learning_rate": 3.806422813028323e-05, "loss": 0.0332, "step": 59790 }, { "grad_norm": 0.25986215472221375, "learning_rate": 3.8048172110383114e-05, "loss": 0.0207, "step": 59800 }, { "grad_norm": 0.2788229286670685, "learning_rate": 3.8032117397516944e-05, "loss": 0.0184, "step": 59810 }, { "grad_norm": 0.21401774883270264, "learning_rate": 3.801606399344042e-05, "loss": 0.0224, "step": 59820 }, { "grad_norm": 0.2309282273054123, "learning_rate": 3.800001189990913e-05, "loss": 0.0223, "step": 59830 }, { "grad_norm": 0.38109952211380005, "learning_rate": 3.798396111867847e-05, "loss": 0.0242, "step": 59840 }, { "grad_norm": 0.213038831949234, "learning_rate": 3.796791165150378e-05, "loss": 0.0222, "step": 59850 }, { "grad_norm": 0.2979833781719208, "learning_rate": 3.795186350014018e-05, "loss": 0.025, "step": 59860 }, { "grad_norm": 0.3080933392047882, "learning_rate": 3.793581666634266e-05, "loss": 0.0254, "step": 59870 }, { "grad_norm": 0.34149470925331116, "learning_rate": 3.791977115186609e-05, "loss": 0.0197, "step": 59880 }, { "grad_norm": 0.21634440124034882, "learning_rate": 3.79037269584652e-05, "loss": 0.0228, "step": 59890 }, { "grad_norm": 0.20967893302440643, "learning_rate": 3.7887684087894545e-05, "loss": 0.0227, "step": 59900 }, { "grad_norm": 0.2678648829460144, "learning_rate": 3.787164254190854e-05, "loss": 0.0213, "step": 59910 }, { "grad_norm": 0.2696346044540405, "learning_rate": 3.7855602322261473e-05, "loss": 0.023, "step": 59920 }, { "grad_norm": 0.2595871686935425, "learning_rate": 3.783956343070746e-05, "loss": 0.0307, "step": 59930 }, { "grad_norm": 0.5511003732681274, "learning_rate": 3.7823525869000526e-05, "loss": 0.0212, "step": 59940 }, { "grad_norm": 0.22741933166980743, "learning_rate": 3.78074896388945e-05, "loss": 0.0237, "step": 59950 }, { "grad_norm": 0.6871237754821777, "learning_rate": 3.7791454742143075e-05, "loss": 0.0288, "step": 59960 }, { "grad_norm": 0.16240829229354858, "learning_rate": 3.7775421180499795e-05, "loss": 0.0189, "step": 59970 }, { "grad_norm": 0.3884752690792084, "learning_rate": 3.7759388955718086e-05, "loss": 0.0213, "step": 59980 }, { "grad_norm": 0.2588235139846802, "learning_rate": 3.774335806955117e-05, "loss": 0.0232, "step": 59990 }, { "grad_norm": 0.3128032386302948, "learning_rate": 3.7727328523752215e-05, "loss": 0.0226, "step": 60000 }, { "grad_norm": 0.21080341935157776, "learning_rate": 3.771130032007415e-05, "loss": 0.0184, "step": 60010 }, { "grad_norm": 0.2837831974029541, "learning_rate": 3.769527346026982e-05, "loss": 0.025, "step": 60020 }, { "grad_norm": 0.20599417388439178, "learning_rate": 3.767924794609188e-05, "loss": 0.0182, "step": 60030 }, { "grad_norm": 0.54438316822052, "learning_rate": 3.7663223779292854e-05, "loss": 0.0211, "step": 60040 }, { "grad_norm": 0.34357762336730957, "learning_rate": 3.764720096162512e-05, "loss": 0.0307, "step": 60050 }, { "grad_norm": 0.2842610776424408, "learning_rate": 3.763117949484092e-05, "loss": 0.0216, "step": 60060 }, { "grad_norm": 0.24101522564888, "learning_rate": 3.7615159380692344e-05, "loss": 0.0221, "step": 60070 }, { "grad_norm": 0.2558269798755646, "learning_rate": 3.759914062093132e-05, "loss": 0.0218, "step": 60080 }, { "grad_norm": 0.3082420527935028, "learning_rate": 3.7583123217309615e-05, "loss": 0.0279, "step": 60090 }, { "grad_norm": 0.45869261026382446, "learning_rate": 3.7567107171578904e-05, "loss": 0.0293, "step": 60100 }, { "grad_norm": 0.1971054971218109, "learning_rate": 3.755109248549066e-05, "loss": 0.023, "step": 60110 }, { "grad_norm": 0.15420712530612946, "learning_rate": 3.75350791607962e-05, "loss": 0.0169, "step": 60120 }, { "grad_norm": 0.25083649158477783, "learning_rate": 3.751906719924676e-05, "loss": 0.0246, "step": 60130 }, { "grad_norm": 0.17646077275276184, "learning_rate": 3.750305660259337e-05, "loss": 0.0296, "step": 60140 }, { "grad_norm": 0.2595354914665222, "learning_rate": 3.748704737258693e-05, "loss": 0.0183, "step": 60150 }, { "grad_norm": 0.6221306920051575, "learning_rate": 3.747103951097816e-05, "loss": 0.024, "step": 60160 }, { "grad_norm": 0.34490257501602173, "learning_rate": 3.745503301951768e-05, "loss": 0.0182, "step": 60170 }, { "grad_norm": 0.25288593769073486, "learning_rate": 3.743902789995592e-05, "loss": 0.0252, "step": 60180 }, { "grad_norm": 0.2679027318954468, "learning_rate": 3.7423024154043183e-05, "loss": 0.0212, "step": 60190 }, { "grad_norm": 0.2554752230644226, "learning_rate": 3.740702178352963e-05, "loss": 0.0211, "step": 60200 }, { "grad_norm": 0.24452504515647888, "learning_rate": 3.739102079016523e-05, "loss": 0.0204, "step": 60210 }, { "grad_norm": 0.17523854970932007, "learning_rate": 3.7375021175699846e-05, "loss": 0.0192, "step": 60220 }, { "grad_norm": 0.22272032499313354, "learning_rate": 3.735902294188317e-05, "loss": 0.0222, "step": 60230 }, { "grad_norm": 0.21871677041053772, "learning_rate": 3.734302609046473e-05, "loss": 0.0154, "step": 60240 }, { "grad_norm": 0.22938643395900726, "learning_rate": 3.732703062319391e-05, "loss": 0.0199, "step": 60250 }, { "grad_norm": 0.2454812228679657, "learning_rate": 3.731103654181999e-05, "loss": 0.0244, "step": 60260 }, { "grad_norm": 0.2239534556865692, "learning_rate": 3.729504384809203e-05, "loss": 0.0261, "step": 60270 }, { "grad_norm": 0.14653511345386505, "learning_rate": 3.727905254375898e-05, "loss": 0.0234, "step": 60280 }, { "grad_norm": 0.3689629137516022, "learning_rate": 3.7263062630569614e-05, "loss": 0.0186, "step": 60290 }, { "grad_norm": 0.2210279256105423, "learning_rate": 3.724707411027256e-05, "loss": 0.019, "step": 60300 }, { "grad_norm": 0.27913451194763184, "learning_rate": 3.723108698461631e-05, "loss": 0.0218, "step": 60310 }, { "grad_norm": 0.19260306656360626, "learning_rate": 3.72151012553492e-05, "loss": 0.0261, "step": 60320 }, { "grad_norm": 0.22236429154872894, "learning_rate": 3.719911692421939e-05, "loss": 0.0188, "step": 60330 }, { "grad_norm": 0.23268422484397888, "learning_rate": 3.71831339929749e-05, "loss": 0.0166, "step": 60340 }, { "grad_norm": 0.2770915627479553, "learning_rate": 3.716715246336362e-05, "loss": 0.0212, "step": 60350 }, { "grad_norm": 0.271955668926239, "learning_rate": 3.715117233713324e-05, "loss": 0.0177, "step": 60360 }, { "grad_norm": 0.2591732144355774, "learning_rate": 3.713519361603135e-05, "loss": 0.0157, "step": 60370 }, { "grad_norm": 0.2794182002544403, "learning_rate": 3.711921630180532e-05, "loss": 0.0192, "step": 60380 }, { "grad_norm": 0.34697869420051575, "learning_rate": 3.710324039620245e-05, "loss": 0.0205, "step": 60390 }, { "grad_norm": 0.25939422845840454, "learning_rate": 3.708726590096982e-05, "loss": 0.0184, "step": 60400 }, { "grad_norm": 0.2962356507778168, "learning_rate": 3.707129281785437e-05, "loss": 0.0223, "step": 60410 }, { "grad_norm": 0.24757710099220276, "learning_rate": 3.705532114860291e-05, "loss": 0.0254, "step": 60420 }, { "grad_norm": 0.2994142770767212, "learning_rate": 3.703935089496207e-05, "loss": 0.0279, "step": 60430 }, { "grad_norm": 0.27904951572418213, "learning_rate": 3.702338205867829e-05, "loss": 0.0222, "step": 60440 }, { "grad_norm": 0.3750552237033844, "learning_rate": 3.7007414641497976e-05, "loss": 0.0235, "step": 60450 }, { "grad_norm": 0.17913095653057098, "learning_rate": 3.699144864516726e-05, "loss": 0.027, "step": 60460 }, { "grad_norm": 0.25604888796806335, "learning_rate": 3.6975484071432153e-05, "loss": 0.022, "step": 60470 }, { "grad_norm": 0.17420506477355957, "learning_rate": 3.695952092203854e-05, "loss": 0.0181, "step": 60480 }, { "grad_norm": 0.3298814594745636, "learning_rate": 3.69435591987321e-05, "loss": 0.0288, "step": 60490 }, { "grad_norm": 0.21821938455104828, "learning_rate": 3.6927598903258374e-05, "loss": 0.0201, "step": 60500 }, { "grad_norm": 0.30330631136894226, "learning_rate": 3.691164003736281e-05, "loss": 0.0222, "step": 60510 }, { "grad_norm": 0.1650894284248352, "learning_rate": 3.689568260279061e-05, "loss": 0.0242, "step": 60520 }, { "grad_norm": 0.48340660333633423, "learning_rate": 3.687972660128686e-05, "loss": 0.0232, "step": 60530 }, { "grad_norm": 0.38295209407806396, "learning_rate": 3.686377203459648e-05, "loss": 0.0205, "step": 60540 }, { "grad_norm": 0.2971491813659668, "learning_rate": 3.6847818904464226e-05, "loss": 0.0208, "step": 60550 }, { "grad_norm": 0.20761258900165558, "learning_rate": 3.683186721263474e-05, "loss": 0.015, "step": 60560 }, { "grad_norm": 0.24951490759849548, "learning_rate": 3.6815916960852435e-05, "loss": 0.0222, "step": 60570 }, { "grad_norm": 0.23240840435028076, "learning_rate": 3.679996815086165e-05, "loss": 0.0164, "step": 60580 }, { "grad_norm": 0.3339521288871765, "learning_rate": 3.678402078440649e-05, "loss": 0.0165, "step": 60590 }, { "grad_norm": 0.16858552396297455, "learning_rate": 3.676807486323096e-05, "loss": 0.0175, "step": 60600 }, { "grad_norm": 0.28007569909095764, "learning_rate": 3.6752130389078864e-05, "loss": 0.0221, "step": 60610 }, { "grad_norm": 0.2636098861694336, "learning_rate": 3.673618736369388e-05, "loss": 0.0154, "step": 60620 }, { "grad_norm": 0.23918645083904266, "learning_rate": 3.6720245788819474e-05, "loss": 0.0162, "step": 60630 }, { "grad_norm": 0.24788442254066467, "learning_rate": 3.6704305666199044e-05, "loss": 0.0174, "step": 60640 }, { "grad_norm": 0.153248131275177, "learning_rate": 3.668836699757576e-05, "loss": 0.0201, "step": 60650 }, { "grad_norm": 0.23869730532169342, "learning_rate": 3.667242978469265e-05, "loss": 0.0204, "step": 60660 }, { "grad_norm": 0.30848804116249084, "learning_rate": 3.6656494029292575e-05, "loss": 0.0223, "step": 60670 }, { "grad_norm": 0.2050754278898239, "learning_rate": 3.664055973311825e-05, "loss": 0.022, "step": 60680 }, { "grad_norm": 0.4053982198238373, "learning_rate": 3.662462689791221e-05, "loss": 0.0208, "step": 60690 }, { "grad_norm": 1.1527155637741089, "learning_rate": 3.660869552541689e-05, "loss": 0.0244, "step": 60700 }, { "grad_norm": 0.2100982517004013, "learning_rate": 3.659276561737448e-05, "loss": 0.0214, "step": 60710 }, { "grad_norm": 0.2700226306915283, "learning_rate": 3.657683717552708e-05, "loss": 0.0214, "step": 60720 }, { "grad_norm": 0.22839215397834778, "learning_rate": 3.6560910201616574e-05, "loss": 0.0211, "step": 60730 }, { "grad_norm": 0.30822160840034485, "learning_rate": 3.654498469738472e-05, "loss": 0.0215, "step": 60740 }, { "grad_norm": 0.35871240496635437, "learning_rate": 3.6529060664573126e-05, "loss": 0.0224, "step": 60750 }, { "grad_norm": 0.29390135407447815, "learning_rate": 3.6513138104923176e-05, "loss": 0.021, "step": 60760 }, { "grad_norm": 0.285086989402771, "learning_rate": 3.6497217020176176e-05, "loss": 0.0169, "step": 60770 }, { "grad_norm": 0.2535487115383148, "learning_rate": 3.648129741207323e-05, "loss": 0.0237, "step": 60780 }, { "grad_norm": 0.21293552219867706, "learning_rate": 3.646537928235527e-05, "loss": 0.0212, "step": 60790 }, { "grad_norm": 0.23903053998947144, "learning_rate": 3.6449462632763075e-05, "loss": 0.0298, "step": 60800 }, { "grad_norm": 0.5087917447090149, "learning_rate": 3.643354746503729e-05, "loss": 0.0263, "step": 60810 }, { "grad_norm": 0.3563205599784851, "learning_rate": 3.6417633780918345e-05, "loss": 0.0176, "step": 60820 }, { "grad_norm": 0.2930293381214142, "learning_rate": 3.640172158214655e-05, "loss": 0.0219, "step": 60830 }, { "grad_norm": 0.18457593023777008, "learning_rate": 3.6385810870462035e-05, "loss": 0.0172, "step": 60840 }, { "grad_norm": 0.18577840924263, "learning_rate": 3.636990164760479e-05, "loss": 0.0206, "step": 60850 }, { "grad_norm": 0.17654216289520264, "learning_rate": 3.6353993915314596e-05, "loss": 0.0214, "step": 60860 }, { "grad_norm": 0.2724170386791229, "learning_rate": 3.6338087675331124e-05, "loss": 0.0195, "step": 60870 }, { "grad_norm": 0.21400322020053864, "learning_rate": 3.632218292939383e-05, "loss": 0.0182, "step": 60880 }, { "grad_norm": 0.4374273717403412, "learning_rate": 3.6306279679242033e-05, "loss": 0.0189, "step": 60890 }, { "grad_norm": 0.6447880268096924, "learning_rate": 3.629037792661492e-05, "loss": 0.0215, "step": 60900 }, { "grad_norm": 0.3289303183555603, "learning_rate": 3.627447767325147e-05, "loss": 0.0281, "step": 60910 }, { "grad_norm": 0.32493069767951965, "learning_rate": 3.62585789208905e-05, "loss": 0.0285, "step": 60920 }, { "grad_norm": 0.27584224939346313, "learning_rate": 3.624268167127068e-05, "loss": 0.0287, "step": 60930 }, { "grad_norm": 0.22580872476100922, "learning_rate": 3.622678592613051e-05, "loss": 0.0233, "step": 60940 }, { "grad_norm": 0.6463112831115723, "learning_rate": 3.6210891687208304e-05, "loss": 0.0286, "step": 60950 }, { "grad_norm": 0.24096670746803284, "learning_rate": 3.619499895624228e-05, "loss": 0.0193, "step": 60960 }, { "grad_norm": 0.18500211834907532, "learning_rate": 3.61791077349704e-05, "loss": 0.0236, "step": 60970 }, { "grad_norm": 0.2006470263004303, "learning_rate": 3.616321802513053e-05, "loss": 0.0274, "step": 60980 }, { "grad_norm": 0.2072136551141739, "learning_rate": 3.6147329828460325e-05, "loss": 0.0221, "step": 60990 }, { "grad_norm": 0.25264254212379456, "learning_rate": 3.6131443146697306e-05, "loss": 0.0261, "step": 61000 }, { "grad_norm": 0.30687251687049866, "learning_rate": 3.611555798157879e-05, "loss": 0.0207, "step": 61010 }, { "grad_norm": 0.2403390109539032, "learning_rate": 3.6099674334842e-05, "loss": 0.0197, "step": 61020 }, { "grad_norm": 0.2361106425523758, "learning_rate": 3.6083792208223934e-05, "loss": 0.0187, "step": 61030 }, { "grad_norm": 0.28960806131362915, "learning_rate": 3.606791160346142e-05, "loss": 0.0352, "step": 61040 }, { "grad_norm": 0.20851509273052216, "learning_rate": 3.6052032522291134e-05, "loss": 0.0223, "step": 61050 }, { "grad_norm": 0.2697763442993164, "learning_rate": 3.603615496644962e-05, "loss": 0.0204, "step": 61060 }, { "grad_norm": 0.3511408269405365, "learning_rate": 3.60202789376732e-05, "loss": 0.0238, "step": 61070 }, { "grad_norm": 0.35613057017326355, "learning_rate": 3.6004404437698036e-05, "loss": 0.0244, "step": 61080 }, { "grad_norm": 0.33241403102874756, "learning_rate": 3.5988531468260184e-05, "loss": 0.0213, "step": 61090 }, { "grad_norm": 0.2969881296157837, "learning_rate": 3.597266003109546e-05, "loss": 0.0254, "step": 61100 }, { "grad_norm": 0.25440728664398193, "learning_rate": 3.5956790127939557e-05, "loss": 0.0175, "step": 61110 }, { "grad_norm": 0.236962229013443, "learning_rate": 3.594092176052796e-05, "loss": 0.0178, "step": 61120 }, { "grad_norm": 0.24725307524204254, "learning_rate": 3.592505493059603e-05, "loss": 0.0216, "step": 61130 }, { "grad_norm": 0.16433066129684448, "learning_rate": 3.5909189639878905e-05, "loss": 0.0226, "step": 61140 }, { "grad_norm": 0.17173360288143158, "learning_rate": 3.5893325890111626e-05, "loss": 0.0173, "step": 61150 }, { "grad_norm": 0.15520767867565155, "learning_rate": 3.5877463683029034e-05, "loss": 0.0247, "step": 61160 }, { "grad_norm": 0.20113100111484528, "learning_rate": 3.586160302036578e-05, "loss": 0.0198, "step": 61170 }, { "grad_norm": 0.19777944684028625, "learning_rate": 3.584574390385636e-05, "loss": 0.02, "step": 61180 }, { "grad_norm": 0.3643399178981781, "learning_rate": 3.582988633523511e-05, "loss": 0.0238, "step": 61190 }, { "grad_norm": 0.23811045289039612, "learning_rate": 3.581403031623616e-05, "loss": 0.0248, "step": 61200 }, { "grad_norm": 0.2553798258304596, "learning_rate": 3.579817584859354e-05, "loss": 0.0203, "step": 61210 }, { "grad_norm": 0.3267751932144165, "learning_rate": 3.578232293404107e-05, "loss": 0.023, "step": 61220 }, { "grad_norm": 0.3005784749984741, "learning_rate": 3.5766471574312385e-05, "loss": 0.0224, "step": 61230 }, { "grad_norm": 0.2000287026166916, "learning_rate": 3.5750621771140964e-05, "loss": 0.0279, "step": 61240 }, { "grad_norm": 0.30080893635749817, "learning_rate": 3.5734773526260115e-05, "loss": 0.0211, "step": 61250 }, { "grad_norm": 0.27195319533348083, "learning_rate": 3.5718926841402996e-05, "loss": 0.036, "step": 61260 }, { "grad_norm": 0.2862326502799988, "learning_rate": 3.5703081718302555e-05, "loss": 0.0268, "step": 61270 }, { "grad_norm": 0.31313765048980713, "learning_rate": 3.568723815869162e-05, "loss": 0.0246, "step": 61280 }, { "grad_norm": 0.1715613603591919, "learning_rate": 3.5671396164302777e-05, "loss": 0.0279, "step": 61290 }, { "grad_norm": 0.22870798408985138, "learning_rate": 3.565555573686853e-05, "loss": 0.0221, "step": 61300 }, { "grad_norm": 0.5087738037109375, "learning_rate": 3.563971687812113e-05, "loss": 0.0203, "step": 61310 }, { "grad_norm": 0.39503946900367737, "learning_rate": 3.5623879589792706e-05, "loss": 0.019, "step": 61320 }, { "grad_norm": 0.1632579267024994, "learning_rate": 3.5608043873615185e-05, "loss": 0.0244, "step": 61330 }, { "grad_norm": 0.17033261060714722, "learning_rate": 3.559220973132035e-05, "loss": 0.0175, "step": 61340 }, { "grad_norm": 0.24798186123371124, "learning_rate": 3.557637716463981e-05, "loss": 0.0255, "step": 61350 }, { "grad_norm": 0.2614022493362427, "learning_rate": 3.556054617530497e-05, "loss": 0.0248, "step": 61360 }, { "grad_norm": 0.22135865688323975, "learning_rate": 3.5544716765047106e-05, "loss": 0.0208, "step": 61370 }, { "grad_norm": 0.3007397949695587, "learning_rate": 3.5528888935597274e-05, "loss": 0.0208, "step": 61380 }, { "grad_norm": 0.25153648853302, "learning_rate": 3.551306268868636e-05, "loss": 0.0176, "step": 61390 }, { "grad_norm": 0.21583276987075806, "learning_rate": 3.5497238026045166e-05, "loss": 0.022, "step": 61400 }, { "grad_norm": 0.20551174879074097, "learning_rate": 3.548141494940421e-05, "loss": 0.0173, "step": 61410 }, { "grad_norm": 0.2093268781900406, "learning_rate": 3.546559346049388e-05, "loss": 0.022, "step": 61420 }, { "grad_norm": 0.20364241302013397, "learning_rate": 3.544977356104441e-05, "loss": 0.021, "step": 61430 }, { "grad_norm": 0.18633556365966797, "learning_rate": 3.543395525278582e-05, "loss": 0.021, "step": 61440 }, { "grad_norm": 0.1719716638326645, "learning_rate": 3.5418138537447985e-05, "loss": 0.0224, "step": 61450 }, { "grad_norm": 0.18762405216693878, "learning_rate": 3.5402323416760574e-05, "loss": 0.0227, "step": 61460 }, { "grad_norm": 0.23669837415218353, "learning_rate": 3.538650989245316e-05, "loss": 0.0201, "step": 61470 }, { "grad_norm": 0.25077736377716064, "learning_rate": 3.537069796625504e-05, "loss": 0.0201, "step": 61480 }, { "grad_norm": 0.12251246720552444, "learning_rate": 3.53548876398954e-05, "loss": 0.0174, "step": 61490 }, { "grad_norm": 0.420314222574234, "learning_rate": 3.5339078915103214e-05, "loss": 0.0245, "step": 61500 }, { "grad_norm": 0.32409340143203735, "learning_rate": 3.532327179360733e-05, "loss": 0.0229, "step": 61510 }, { "grad_norm": 0.21865031123161316, "learning_rate": 3.530746627713636e-05, "loss": 0.0177, "step": 61520 }, { "grad_norm": 0.2008407711982727, "learning_rate": 3.5291662367418805e-05, "loss": 0.0165, "step": 61530 }, { "grad_norm": 0.32689762115478516, "learning_rate": 3.527586006618293e-05, "loss": 0.0218, "step": 61540 }, { "grad_norm": 0.259467750787735, "learning_rate": 3.526005937515687e-05, "loss": 0.0235, "step": 61550 }, { "grad_norm": 0.4111478626728058, "learning_rate": 3.524426029606856e-05, "loss": 0.0276, "step": 61560 }, { "grad_norm": 0.31372854113578796, "learning_rate": 3.5228462830645746e-05, "loss": 0.0168, "step": 61570 }, { "grad_norm": 0.2530093789100647, "learning_rate": 3.5212666980616036e-05, "loss": 0.0189, "step": 61580 }, { "grad_norm": 0.275235116481781, "learning_rate": 3.519687274770681e-05, "loss": 0.0188, "step": 61590 }, { "grad_norm": 0.12274007499217987, "learning_rate": 3.518108013364535e-05, "loss": 0.0267, "step": 61600 }, { "grad_norm": 0.1505005806684494, "learning_rate": 3.516528914015868e-05, "loss": 0.0193, "step": 61610 }, { "grad_norm": 0.1797659546136856, "learning_rate": 3.5149499768973694e-05, "loss": 0.0202, "step": 61620 }, { "grad_norm": 0.25998321175575256, "learning_rate": 3.513371202181708e-05, "loss": 0.0229, "step": 61630 }, { "grad_norm": 0.17873801290988922, "learning_rate": 3.511792590041537e-05, "loss": 0.0193, "step": 61640 }, { "grad_norm": 0.2961612343788147, "learning_rate": 3.510214140649488e-05, "loss": 0.0295, "step": 61650 }, { "grad_norm": 0.21432387828826904, "learning_rate": 3.508635854178184e-05, "loss": 0.022, "step": 61660 }, { "grad_norm": 0.19123199582099915, "learning_rate": 3.50705773080022e-05, "loss": 0.0232, "step": 61670 }, { "grad_norm": 0.32615941762924194, "learning_rate": 3.505479770688178e-05, "loss": 0.0206, "step": 61680 }, { "grad_norm": 0.2323056310415268, "learning_rate": 3.503901974014622e-05, "loss": 0.0168, "step": 61690 }, { "grad_norm": 0.13252444565296173, "learning_rate": 3.502324340952095e-05, "loss": 0.02, "step": 61700 }, { "grad_norm": 0.27107498049736023, "learning_rate": 3.500746871673126e-05, "loss": 0.0184, "step": 61710 }, { "grad_norm": 0.16438785195350647, "learning_rate": 3.499169566350227e-05, "loss": 0.0262, "step": 61720 }, { "grad_norm": 0.37536540627479553, "learning_rate": 3.4975924251558876e-05, "loss": 0.02, "step": 61730 }, { "grad_norm": 0.2752428948879242, "learning_rate": 3.496015448262582e-05, "loss": 0.0225, "step": 61740 }, { "grad_norm": 0.23643355071544647, "learning_rate": 3.494438635842764e-05, "loss": 0.0176, "step": 61750 }, { "grad_norm": 0.29441606998443604, "learning_rate": 3.492861988068874e-05, "loss": 0.0222, "step": 61760 }, { "grad_norm": 0.2825900614261627, "learning_rate": 3.491285505113332e-05, "loss": 0.0212, "step": 61770 }, { "grad_norm": 0.24877946078777313, "learning_rate": 3.4897091871485376e-05, "loss": 0.0256, "step": 61780 }, { "grad_norm": 0.1721239984035492, "learning_rate": 3.488133034346875e-05, "loss": 0.022, "step": 61790 }, { "grad_norm": 0.4390536844730377, "learning_rate": 3.486557046880713e-05, "loss": 0.022, "step": 61800 }, { "grad_norm": 0.16166779398918152, "learning_rate": 3.484981224922395e-05, "loss": 0.0244, "step": 61810 }, { "grad_norm": 0.20410867035388947, "learning_rate": 3.4834055686442544e-05, "loss": 0.0189, "step": 61820 }, { "grad_norm": 0.27437829971313477, "learning_rate": 3.4818300782186e-05, "loss": 0.0229, "step": 61830 }, { "grad_norm": 0.5615166425704956, "learning_rate": 3.4802547538177236e-05, "loss": 0.0226, "step": 61840 }, { "grad_norm": 0.15784116089344025, "learning_rate": 3.478679595613905e-05, "loss": 0.0183, "step": 61850 }, { "grad_norm": 0.289277583360672, "learning_rate": 3.4771046037793975e-05, "loss": 0.0231, "step": 61860 }, { "grad_norm": 0.25229236483573914, "learning_rate": 3.4755297784864416e-05, "loss": 0.0247, "step": 61870 }, { "grad_norm": 0.19201786816120148, "learning_rate": 3.473955119907257e-05, "loss": 0.0202, "step": 61880 }, { "grad_norm": 0.22501009702682495, "learning_rate": 3.472380628214046e-05, "loss": 0.0188, "step": 61890 }, { "grad_norm": 0.3019930422306061, "learning_rate": 3.4708063035789906e-05, "loss": 0.0216, "step": 61900 }, { "grad_norm": 0.3912009596824646, "learning_rate": 3.4692321461742614e-05, "loss": 0.021, "step": 61910 }, { "grad_norm": 0.5315520167350769, "learning_rate": 3.4676581561720023e-05, "loss": 0.0283, "step": 61920 }, { "grad_norm": 0.21874211728572845, "learning_rate": 3.466084333744344e-05, "loss": 0.0215, "step": 61930 }, { "grad_norm": 0.2869795858860016, "learning_rate": 3.4645106790633964e-05, "loss": 0.0254, "step": 61940 }, { "grad_norm": 0.23645584285259247, "learning_rate": 3.4629371923012526e-05, "loss": 0.0255, "step": 61950 }, { "grad_norm": 0.26357460021972656, "learning_rate": 3.4613638736299856e-05, "loss": 0.0268, "step": 61960 }, { "grad_norm": 0.3178781569004059, "learning_rate": 3.459790723221652e-05, "loss": 0.0244, "step": 61970 }, { "grad_norm": 0.23349077999591827, "learning_rate": 3.45821774124829e-05, "loss": 0.0191, "step": 61980 }, { "grad_norm": 0.19702614843845367, "learning_rate": 3.456644927881918e-05, "loss": 0.0195, "step": 61990 }, { "grad_norm": 0.23134192824363708, "learning_rate": 3.4550722832945354e-05, "loss": 0.0214, "step": 62000 }, { "grad_norm": 0.1563768982887268, "learning_rate": 3.453499807658126e-05, "loss": 0.0239, "step": 62010 }, { "grad_norm": 0.13959015905857086, "learning_rate": 3.451927501144653e-05, "loss": 0.0252, "step": 62020 }, { "grad_norm": 0.31255295872688293, "learning_rate": 3.450355363926059e-05, "loss": 0.0214, "step": 62030 }, { "grad_norm": 0.2337816059589386, "learning_rate": 3.448783396174273e-05, "loss": 0.0261, "step": 62040 }, { "grad_norm": 0.31795811653137207, "learning_rate": 3.4472115980612034e-05, "loss": 0.0199, "step": 62050 }, { "grad_norm": 0.27022987604141235, "learning_rate": 3.4456399697587396e-05, "loss": 0.0204, "step": 62060 }, { "grad_norm": 0.22771187126636505, "learning_rate": 3.4440685114387506e-05, "loss": 0.0156, "step": 62070 }, { "grad_norm": 0.1965923011302948, "learning_rate": 3.442497223273091e-05, "loss": 0.0224, "step": 62080 }, { "grad_norm": 0.5200836658477783, "learning_rate": 3.4409261054335926e-05, "loss": 0.028, "step": 62090 }, { "grad_norm": 0.22943975031375885, "learning_rate": 3.439355158092069e-05, "loss": 0.0323, "step": 62100 }, { "grad_norm": 0.2686789631843567, "learning_rate": 3.437784381420322e-05, "loss": 0.0242, "step": 62110 }, { "grad_norm": 0.2242952585220337, "learning_rate": 3.436213775590126e-05, "loss": 0.0197, "step": 62120 }, { "grad_norm": 0.19029484689235687, "learning_rate": 3.4346433407732405e-05, "loss": 0.0219, "step": 62130 }, { "grad_norm": 0.20794694125652313, "learning_rate": 3.433073077141405e-05, "loss": 0.0202, "step": 62140 }, { "grad_norm": 0.3084031641483307, "learning_rate": 3.431502984866343e-05, "loss": 0.0274, "step": 62150 }, { "grad_norm": 0.2637105882167816, "learning_rate": 3.429933064119753e-05, "loss": 0.0327, "step": 62160 }, { "grad_norm": 0.18530000746250153, "learning_rate": 3.4283633150733255e-05, "loss": 0.0284, "step": 62170 }, { "grad_norm": 0.18452292680740356, "learning_rate": 3.4267937378987225e-05, "loss": 0.0239, "step": 62180 }, { "grad_norm": 0.26414376497268677, "learning_rate": 3.425224332767592e-05, "loss": 0.0169, "step": 62190 }, { "grad_norm": 0.26004454493522644, "learning_rate": 3.4236550998515595e-05, "loss": 0.0254, "step": 62200 }, { "grad_norm": 0.34851503372192383, "learning_rate": 3.422086039322235e-05, "loss": 0.0221, "step": 62210 }, { "grad_norm": 0.3311505615711212, "learning_rate": 3.420517151351209e-05, "loss": 0.0257, "step": 62220 }, { "grad_norm": 0.2829195261001587, "learning_rate": 3.418948436110053e-05, "loss": 0.0201, "step": 62230 }, { "grad_norm": 0.19095659255981445, "learning_rate": 3.417379893770319e-05, "loss": 0.026, "step": 62240 }, { "grad_norm": 0.1743898242712021, "learning_rate": 3.41581152450354e-05, "loss": 0.0214, "step": 62250 }, { "grad_norm": 0.17624929547309875, "learning_rate": 3.414243328481231e-05, "loss": 0.0206, "step": 62260 }, { "grad_norm": 0.28371334075927734, "learning_rate": 3.412675305874888e-05, "loss": 0.0212, "step": 62270 }, { "grad_norm": 0.2736736238002777, "learning_rate": 3.4111074568559876e-05, "loss": 0.0171, "step": 62280 }, { "grad_norm": 0.46243730187416077, "learning_rate": 3.4095397815959846e-05, "loss": 0.021, "step": 62290 }, { "grad_norm": 0.4582955837249756, "learning_rate": 3.407972280266322e-05, "loss": 0.0253, "step": 62300 }, { "grad_norm": 0.2769286334514618, "learning_rate": 3.4064049530384166e-05, "loss": 0.0178, "step": 62310 }, { "grad_norm": 0.2775408625602722, "learning_rate": 3.40483780008367e-05, "loss": 0.0211, "step": 62320 }, { "grad_norm": 0.3028006851673126, "learning_rate": 3.4032708215734635e-05, "loss": 0.0192, "step": 62330 }, { "grad_norm": 0.2553558349609375, "learning_rate": 3.4017040176791584e-05, "loss": 0.0182, "step": 62340 }, { "grad_norm": 0.17310908436775208, "learning_rate": 3.4001373885720976e-05, "loss": 0.0255, "step": 62350 }, { "grad_norm": 0.2076588273048401, "learning_rate": 3.398570934423608e-05, "loss": 0.0209, "step": 62360 }, { "grad_norm": 0.21867890655994415, "learning_rate": 3.397004655404994e-05, "loss": 0.02, "step": 62370 }, { "grad_norm": 0.24075983464717865, "learning_rate": 3.395438551687541e-05, "loss": 0.0195, "step": 62380 }, { "grad_norm": 0.2509669363498688, "learning_rate": 3.393872623442515e-05, "loss": 0.0175, "step": 62390 }, { "grad_norm": 0.19432711601257324, "learning_rate": 3.392306870841164e-05, "loss": 0.0193, "step": 62400 }, { "grad_norm": 0.2413565069437027, "learning_rate": 3.390741294054715e-05, "loss": 0.0262, "step": 62410 }, { "grad_norm": 0.24267123639583588, "learning_rate": 3.38917589325438e-05, "loss": 0.014, "step": 62420 }, { "grad_norm": 0.2098780870437622, "learning_rate": 3.387610668611349e-05, "loss": 0.017, "step": 62430 }, { "grad_norm": 0.4271436631679535, "learning_rate": 3.38604562029679e-05, "loss": 0.0224, "step": 62440 }, { "grad_norm": 0.18553821742534637, "learning_rate": 3.384480748481855e-05, "loss": 0.021, "step": 62450 }, { "grad_norm": 0.34458687901496887, "learning_rate": 3.382916053337676e-05, "loss": 0.0201, "step": 62460 }, { "grad_norm": 0.32599911093711853, "learning_rate": 3.381351535035366e-05, "loss": 0.0229, "step": 62470 }, { "grad_norm": 0.2458563596010208, "learning_rate": 3.379787193746019e-05, "loss": 0.017, "step": 62480 }, { "grad_norm": 0.18185661733150482, "learning_rate": 3.378223029640709e-05, "loss": 0.0176, "step": 62490 }, { "grad_norm": 0.3138333857059479, "learning_rate": 3.376659042890489e-05, "loss": 0.0162, "step": 62500 }, { "grad_norm": 0.4283859431743622, "learning_rate": 3.375095233666397e-05, "loss": 0.0245, "step": 62510 }, { "grad_norm": 0.23753993213176727, "learning_rate": 3.373531602139447e-05, "loss": 0.0162, "step": 62520 }, { "grad_norm": 0.22748735547065735, "learning_rate": 3.371968148480635e-05, "loss": 0.0201, "step": 62530 }, { "grad_norm": 0.3267623484134674, "learning_rate": 3.370404872860937e-05, "loss": 0.0256, "step": 62540 }, { "grad_norm": 0.6419336795806885, "learning_rate": 3.3688417754513145e-05, "loss": 0.0162, "step": 62550 }, { "grad_norm": 0.19320857524871826, "learning_rate": 3.367278856422703e-05, "loss": 0.0217, "step": 62560 }, { "grad_norm": 0.235734760761261, "learning_rate": 3.365716115946021e-05, "loss": 0.0186, "step": 62570 }, { "grad_norm": 0.26670756936073303, "learning_rate": 3.3641535541921666e-05, "loss": 0.0336, "step": 62580 }, { "grad_norm": 0.34884244203567505, "learning_rate": 3.36259117133202e-05, "loss": 0.0226, "step": 62590 }, { "grad_norm": 0.21950282156467438, "learning_rate": 3.36102896753644e-05, "loss": 0.023, "step": 62600 }, { "grad_norm": 0.348344624042511, "learning_rate": 3.35946694297627e-05, "loss": 0.0227, "step": 62610 }, { "grad_norm": 0.1372450739145279, "learning_rate": 3.357905097822328e-05, "loss": 0.0272, "step": 62620 }, { "grad_norm": 0.3262473940849304, "learning_rate": 3.356343432245416e-05, "loss": 0.0201, "step": 62630 }, { "grad_norm": 0.2526337504386902, "learning_rate": 3.354781946416315e-05, "loss": 0.0283, "step": 62640 }, { "grad_norm": 0.2063649594783783, "learning_rate": 3.3532206405057864e-05, "loss": 0.0165, "step": 62650 }, { "grad_norm": 0.36257362365722656, "learning_rate": 3.351659514684573e-05, "loss": 0.0204, "step": 62660 }, { "grad_norm": 0.298259973526001, "learning_rate": 3.350098569123395e-05, "loss": 0.0196, "step": 62670 }, { "grad_norm": 0.20578791201114655, "learning_rate": 3.348537803992958e-05, "loss": 0.023, "step": 62680 }, { "grad_norm": 0.18472066521644592, "learning_rate": 3.346977219463945e-05, "loss": 0.0225, "step": 62690 }, { "grad_norm": 0.46001383662223816, "learning_rate": 3.345416815707017e-05, "loss": 0.0185, "step": 62700 }, { "grad_norm": 0.1967371702194214, "learning_rate": 3.3438565928928185e-05, "loss": 0.0166, "step": 62710 }, { "grad_norm": 0.21823890507221222, "learning_rate": 3.3422965511919736e-05, "loss": 0.0202, "step": 62720 }, { "grad_norm": 0.8450927734375, "learning_rate": 3.3407366907750846e-05, "loss": 0.0226, "step": 62730 }, { "grad_norm": 0.15270884335041046, "learning_rate": 3.339177011812737e-05, "loss": 0.017, "step": 62740 }, { "grad_norm": 0.2435881346464157, "learning_rate": 3.3376175144754935e-05, "loss": 0.0189, "step": 62750 }, { "grad_norm": 0.21517165005207062, "learning_rate": 3.3360581989339004e-05, "loss": 0.0197, "step": 62760 }, { "grad_norm": 0.17033939063549042, "learning_rate": 3.3344990653584804e-05, "loss": 0.0178, "step": 62770 }, { "grad_norm": 0.22874663770198822, "learning_rate": 3.332940113919739e-05, "loss": 0.0194, "step": 62780 }, { "grad_norm": 0.34959882497787476, "learning_rate": 3.331381344788161e-05, "loss": 0.0189, "step": 62790 }, { "grad_norm": 0.15161117911338806, "learning_rate": 3.3298227581342076e-05, "loss": 0.0193, "step": 62800 }, { "grad_norm": 0.2248753011226654, "learning_rate": 3.328264354128329e-05, "loss": 0.0251, "step": 62810 }, { "grad_norm": 0.3241165578365326, "learning_rate": 3.3267061329409465e-05, "loss": 0.0248, "step": 62820 }, { "grad_norm": 0.15949112176895142, "learning_rate": 3.325148094742466e-05, "loss": 0.0202, "step": 62830 }, { "grad_norm": 0.21207140386104584, "learning_rate": 3.323590239703271e-05, "loss": 0.0156, "step": 62840 }, { "grad_norm": 0.29408031702041626, "learning_rate": 3.322032567993728e-05, "loss": 0.0222, "step": 62850 }, { "grad_norm": 0.20883868634700775, "learning_rate": 3.3204750797841774e-05, "loss": 0.0247, "step": 62860 }, { "grad_norm": 0.19191347062587738, "learning_rate": 3.31891777524495e-05, "loss": 0.0179, "step": 62870 }, { "grad_norm": 0.24028460681438446, "learning_rate": 3.317360654546347e-05, "loss": 0.0253, "step": 62880 }, { "grad_norm": 0.24330754578113556, "learning_rate": 3.3158037178586534e-05, "loss": 0.0194, "step": 62890 }, { "grad_norm": 0.3166406750679016, "learning_rate": 3.3142469653521316e-05, "loss": 0.023, "step": 62900 }, { "grad_norm": 0.14460746943950653, "learning_rate": 3.3126903971970284e-05, "loss": 0.021, "step": 62910 }, { "grad_norm": 0.3779143691062927, "learning_rate": 3.311134013563564e-05, "loss": 0.0186, "step": 62920 }, { "grad_norm": 0.3739180862903595, "learning_rate": 3.309577814621947e-05, "loss": 0.0212, "step": 62930 }, { "grad_norm": 0.2909143567085266, "learning_rate": 3.308021800542359e-05, "loss": 0.0252, "step": 62940 }, { "grad_norm": 0.2558484971523285, "learning_rate": 3.306465971494963e-05, "loss": 0.0214, "step": 62950 }, { "grad_norm": 0.3421063721179962, "learning_rate": 3.304910327649901e-05, "loss": 0.0236, "step": 62960 }, { "grad_norm": 0.5705118179321289, "learning_rate": 3.303354869177297e-05, "loss": 0.0228, "step": 62970 }, { "grad_norm": 0.3693068325519562, "learning_rate": 3.301799596247255e-05, "loss": 0.02, "step": 62980 }, { "grad_norm": 0.24077723920345306, "learning_rate": 3.300244509029854e-05, "loss": 0.0209, "step": 62990 }, { "grad_norm": 0.2713775634765625, "learning_rate": 3.2986896076951585e-05, "loss": 0.0236, "step": 63000 }, { "grad_norm": 0.26852577924728394, "learning_rate": 3.297134892413211e-05, "loss": 0.024, "step": 63010 }, { "grad_norm": 0.39581823348999023, "learning_rate": 3.295580363354031e-05, "loss": 0.0202, "step": 63020 }, { "grad_norm": 0.2902878224849701, "learning_rate": 3.2940260206876194e-05, "loss": 0.0212, "step": 63030 }, { "grad_norm": 0.2958255112171173, "learning_rate": 3.292471864583958e-05, "loss": 0.0228, "step": 63040 }, { "grad_norm": 0.18183784186840057, "learning_rate": 3.290917895213004e-05, "loss": 0.0197, "step": 63050 }, { "grad_norm": 0.2357066571712494, "learning_rate": 3.289364112744701e-05, "loss": 0.0248, "step": 63060 }, { "grad_norm": 0.22741840779781342, "learning_rate": 3.287810517348967e-05, "loss": 0.0225, "step": 63070 }, { "grad_norm": 0.445689857006073, "learning_rate": 3.2862571091957014e-05, "loss": 0.0205, "step": 63080 }, { "grad_norm": 0.21198222041130066, "learning_rate": 3.284703888454781e-05, "loss": 0.0198, "step": 63090 }, { "grad_norm": 0.2616937458515167, "learning_rate": 3.2831508552960646e-05, "loss": 0.0259, "step": 63100 }, { "grad_norm": 0.3141031563282013, "learning_rate": 3.281598009889387e-05, "loss": 0.0192, "step": 63110 }, { "grad_norm": 0.30036625266075134, "learning_rate": 3.28004535240457e-05, "loss": 0.0216, "step": 63120 }, { "grad_norm": 0.3017023801803589, "learning_rate": 3.278492883011407e-05, "loss": 0.0247, "step": 63130 }, { "grad_norm": 0.30838504433631897, "learning_rate": 3.276940601879675e-05, "loss": 0.022, "step": 63140 }, { "grad_norm": 0.29123637080192566, "learning_rate": 3.275388509179127e-05, "loss": 0.0218, "step": 63150 }, { "grad_norm": 0.22274769842624664, "learning_rate": 3.2738366050794985e-05, "loss": 0.0238, "step": 63160 }, { "grad_norm": 0.251615434885025, "learning_rate": 3.2722848897505034e-05, "loss": 0.0225, "step": 63170 }, { "grad_norm": 0.24465425312519073, "learning_rate": 3.2707333633618356e-05, "loss": 0.0186, "step": 63180 }, { "grad_norm": 0.309541255235672, "learning_rate": 3.269182026083167e-05, "loss": 0.0192, "step": 63190 }, { "grad_norm": 0.2765350043773651, "learning_rate": 3.2676308780841516e-05, "loss": 0.0243, "step": 63200 }, { "grad_norm": 0.189924418926239, "learning_rate": 3.266079919534416e-05, "loss": 0.0194, "step": 63210 }, { "grad_norm": 0.20447883009910583, "learning_rate": 3.264529150603575e-05, "loss": 0.0247, "step": 63220 }, { "grad_norm": 0.2906535267829895, "learning_rate": 3.262978571461217e-05, "loss": 0.0221, "step": 63230 }, { "grad_norm": 0.18792788684368134, "learning_rate": 3.26142818227691e-05, "loss": 0.0197, "step": 63240 }, { "grad_norm": 0.3628159165382385, "learning_rate": 3.259877983220203e-05, "loss": 0.0259, "step": 63250 }, { "grad_norm": 0.34462130069732666, "learning_rate": 3.2583279744606246e-05, "loss": 0.029, "step": 63260 }, { "grad_norm": 0.31349098682403564, "learning_rate": 3.25677815616768e-05, "loss": 0.0161, "step": 63270 }, { "grad_norm": 0.22178952395915985, "learning_rate": 3.255228528510856e-05, "loss": 0.0218, "step": 63280 }, { "grad_norm": 0.25759047269821167, "learning_rate": 3.253679091659617e-05, "loss": 0.0201, "step": 63290 }, { "grad_norm": 0.2632317841053009, "learning_rate": 3.252129845783404e-05, "loss": 0.0295, "step": 63300 }, { "grad_norm": 0.3713562488555908, "learning_rate": 3.2505807910516456e-05, "loss": 0.0176, "step": 63310 }, { "grad_norm": 0.17400553822517395, "learning_rate": 3.249031927633742e-05, "loss": 0.022, "step": 63320 }, { "grad_norm": 0.24854905903339386, "learning_rate": 3.2474832556990737e-05, "loss": 0.0173, "step": 63330 }, { "grad_norm": 0.19271276891231537, "learning_rate": 3.245934775417002e-05, "loss": 0.0163, "step": 63340 }, { "grad_norm": 0.3030480742454529, "learning_rate": 3.244386486956866e-05, "loss": 0.025, "step": 63350 }, { "grad_norm": 0.20349304378032684, "learning_rate": 3.2428383904879846e-05, "loss": 0.0165, "step": 63360 }, { "grad_norm": 0.2895183563232422, "learning_rate": 3.241290486179652e-05, "loss": 0.0301, "step": 63370 }, { "grad_norm": 0.10124674439430237, "learning_rate": 3.239742774201151e-05, "loss": 0.0184, "step": 63380 }, { "grad_norm": 0.17838439345359802, "learning_rate": 3.238195254721733e-05, "loss": 0.0214, "step": 63390 }, { "grad_norm": 0.33253875374794006, "learning_rate": 3.236647927910633e-05, "loss": 0.0177, "step": 63400 }, { "grad_norm": 0.2103324681520462, "learning_rate": 3.2351007939370646e-05, "loss": 0.0203, "step": 63410 }, { "grad_norm": 0.24293093383312225, "learning_rate": 3.2335538529702184e-05, "loss": 0.023, "step": 63420 }, { "grad_norm": 0.28962990641593933, "learning_rate": 3.232007105179268e-05, "loss": 0.0209, "step": 63430 }, { "grad_norm": 0.29940274357795715, "learning_rate": 3.230460550733363e-05, "loss": 0.0197, "step": 63440 }, { "grad_norm": 0.5681014060974121, "learning_rate": 3.2289141898016315e-05, "loss": 0.0197, "step": 63450 }, { "grad_norm": 0.2216930240392685, "learning_rate": 3.227368022553181e-05, "loss": 0.0205, "step": 63460 }, { "grad_norm": 0.2140827625989914, "learning_rate": 3.225822049157099e-05, "loss": 0.0199, "step": 63470 }, { "grad_norm": 0.2805328667163849, "learning_rate": 3.2242762697824494e-05, "loss": 0.0184, "step": 63480 }, { "grad_norm": 0.11505934596061707, "learning_rate": 3.222730684598279e-05, "loss": 0.0277, "step": 63490 }, { "grad_norm": 0.389531672000885, "learning_rate": 3.221185293773605e-05, "loss": 0.022, "step": 63500 }, { "grad_norm": 0.20934343338012695, "learning_rate": 3.2196400974774354e-05, "loss": 0.0183, "step": 63510 }, { "grad_norm": 0.16109339892864227, "learning_rate": 3.2180950958787485e-05, "loss": 0.0285, "step": 63520 }, { "grad_norm": 0.20335666835308075, "learning_rate": 3.2165502891465024e-05, "loss": 0.0212, "step": 63530 }, { "grad_norm": 0.17525207996368408, "learning_rate": 3.215005677449636e-05, "loss": 0.0193, "step": 63540 }, { "grad_norm": 0.1626300811767578, "learning_rate": 3.2134612609570644e-05, "loss": 0.0223, "step": 63550 }, { "grad_norm": 0.22111763060092926, "learning_rate": 3.211917039837682e-05, "loss": 0.0203, "step": 63560 }, { "grad_norm": 0.23870569467544556, "learning_rate": 3.210373014260365e-05, "loss": 0.0256, "step": 63570 }, { "grad_norm": 0.25071433186531067, "learning_rate": 3.208829184393965e-05, "loss": 0.0161, "step": 63580 }, { "grad_norm": 0.27817386388778687, "learning_rate": 3.207285550407312e-05, "loss": 0.0243, "step": 63590 }, { "grad_norm": 0.26263248920440674, "learning_rate": 3.205742112469217e-05, "loss": 0.0262, "step": 63600 }, { "grad_norm": 0.2641906142234802, "learning_rate": 3.204198870748465e-05, "loss": 0.022, "step": 63610 }, { "grad_norm": 0.2898683547973633, "learning_rate": 3.202655825413824e-05, "loss": 0.0157, "step": 63620 }, { "grad_norm": 0.25087738037109375, "learning_rate": 3.2011129766340416e-05, "loss": 0.0254, "step": 63630 }, { "grad_norm": 0.1413842886686325, "learning_rate": 3.1995703245778384e-05, "loss": 0.0168, "step": 63640 }, { "grad_norm": 0.22483797371387482, "learning_rate": 3.1980278694139186e-05, "loss": 0.0194, "step": 63650 }, { "grad_norm": 0.5657219290733337, "learning_rate": 3.196485611310962e-05, "loss": 0.0205, "step": 63660 }, { "grad_norm": 0.28859803080558777, "learning_rate": 3.1949435504376257e-05, "loss": 0.0185, "step": 63670 }, { "grad_norm": 0.6582997441291809, "learning_rate": 3.1934016869625505e-05, "loss": 0.0226, "step": 63680 }, { "grad_norm": 0.5284136533737183, "learning_rate": 3.191860021054349e-05, "loss": 0.0376, "step": 63690 }, { "grad_norm": 0.25122109055519104, "learning_rate": 3.190318552881618e-05, "loss": 0.0276, "step": 63700 }, { "grad_norm": 0.24374361336231232, "learning_rate": 3.1887772826129285e-05, "loss": 0.0195, "step": 63710 }, { "grad_norm": 0.28906524181365967, "learning_rate": 3.1872362104168335e-05, "loss": 0.0204, "step": 63720 }, { "grad_norm": 0.3444157838821411, "learning_rate": 3.185695336461861e-05, "loss": 0.0138, "step": 63730 }, { "grad_norm": 0.22123578190803528, "learning_rate": 3.184154660916519e-05, "loss": 0.016, "step": 63740 }, { "grad_norm": 0.19470879435539246, "learning_rate": 3.1826141839492906e-05, "loss": 0.0205, "step": 63750 }, { "grad_norm": 0.47964268922805786, "learning_rate": 3.181073905728644e-05, "loss": 0.0249, "step": 63760 }, { "grad_norm": 0.22952131927013397, "learning_rate": 3.1795338264230214e-05, "loss": 0.0219, "step": 63770 }, { "grad_norm": 0.24603407084941864, "learning_rate": 3.1779939462008426e-05, "loss": 0.0206, "step": 63780 }, { "grad_norm": 0.26032379269599915, "learning_rate": 3.1764542652305066e-05, "loss": 0.0181, "step": 63790 }, { "grad_norm": 0.17804691195487976, "learning_rate": 3.17491478368039e-05, "loss": 0.0177, "step": 63800 }, { "grad_norm": 0.23892353475093842, "learning_rate": 3.173375501718847e-05, "loss": 0.0214, "step": 63810 }, { "grad_norm": 0.27503836154937744, "learning_rate": 3.171836419514215e-05, "loss": 0.0204, "step": 63820 }, { "grad_norm": 0.26171499490737915, "learning_rate": 3.170297537234803e-05, "loss": 0.0223, "step": 63830 }, { "grad_norm": 0.22660303115844727, "learning_rate": 3.168758855048901e-05, "loss": 0.0227, "step": 63840 }, { "grad_norm": 0.16375179588794708, "learning_rate": 3.1672203731247785e-05, "loss": 0.0253, "step": 63850 }, { "grad_norm": 0.24704031646251678, "learning_rate": 3.16568209163068e-05, "loss": 0.0199, "step": 63860 }, { "grad_norm": 0.4398811459541321, "learning_rate": 3.1641440107348295e-05, "loss": 0.02, "step": 63870 }, { "grad_norm": 0.21476203203201294, "learning_rate": 3.162606130605429e-05, "loss": 0.0192, "step": 63880 }, { "grad_norm": 0.1685485690832138, "learning_rate": 3.16106845141066e-05, "loss": 0.0175, "step": 63890 }, { "grad_norm": 0.2519683241844177, "learning_rate": 3.159530973318682e-05, "loss": 0.0168, "step": 63900 }, { "grad_norm": 0.48248860239982605, "learning_rate": 3.1579936964976286e-05, "loss": 0.0186, "step": 63910 }, { "grad_norm": 0.987955629825592, "learning_rate": 3.156456621115615e-05, "loss": 0.0225, "step": 63920 }, { "grad_norm": 0.22340059280395508, "learning_rate": 3.154919747340734e-05, "loss": 0.0269, "step": 63930 }, { "grad_norm": 0.3449763357639313, "learning_rate": 3.153383075341054e-05, "loss": 0.0181, "step": 63940 }, { "grad_norm": 0.568334698677063, "learning_rate": 3.1518466052846264e-05, "loss": 0.0267, "step": 63950 }, { "grad_norm": 0.2147631198167801, "learning_rate": 3.150310337339474e-05, "loss": 0.0215, "step": 63960 }, { "grad_norm": 0.2407759577035904, "learning_rate": 3.148774271673603e-05, "loss": 0.0253, "step": 63970 }, { "grad_norm": 0.20547538995742798, "learning_rate": 3.147238408454995e-05, "loss": 0.0244, "step": 63980 }, { "grad_norm": 0.172278493642807, "learning_rate": 3.1457027478516086e-05, "loss": 0.0166, "step": 63990 }, { "grad_norm": 0.201475128531456, "learning_rate": 3.144167290031382e-05, "loss": 0.0165, "step": 64000 }, { "grad_norm": 0.29969075322151184, "learning_rate": 3.142632035162227e-05, "loss": 0.022, "step": 64010 }, { "grad_norm": 0.42784383893013, "learning_rate": 3.141096983412044e-05, "loss": 0.0191, "step": 64020 }, { "grad_norm": 0.2170373648405075, "learning_rate": 3.139562134948699e-05, "loss": 0.0239, "step": 64030 }, { "grad_norm": 0.5983903408050537, "learning_rate": 3.1380274899400424e-05, "loss": 0.0206, "step": 64040 }, { "grad_norm": 0.23991821706295013, "learning_rate": 3.1364930485539e-05, "loss": 0.0225, "step": 64050 }, { "grad_norm": 0.17348915338516235, "learning_rate": 3.134958810958076e-05, "loss": 0.0179, "step": 64060 }, { "grad_norm": 0.3429158627986908, "learning_rate": 3.13342477732035e-05, "loss": 0.0178, "step": 64070 }, { "grad_norm": 0.46939826011657715, "learning_rate": 3.131890947808487e-05, "loss": 0.0215, "step": 64080 }, { "grad_norm": 0.23954077064990997, "learning_rate": 3.13035732259022e-05, "loss": 0.0182, "step": 64090 }, { "grad_norm": 0.41892513632774353, "learning_rate": 3.1288239018332656e-05, "loss": 0.0184, "step": 64100 }, { "grad_norm": 0.19932639598846436, "learning_rate": 3.1272906857053164e-05, "loss": 0.0145, "step": 64110 }, { "grad_norm": 0.25581446290016174, "learning_rate": 3.125757674374042e-05, "loss": 0.0163, "step": 64120 }, { "grad_norm": 0.1531868278980255, "learning_rate": 3.124224868007088e-05, "loss": 0.0216, "step": 64130 }, { "grad_norm": 0.2039199024438858, "learning_rate": 3.1226922667720854e-05, "loss": 0.0171, "step": 64140 }, { "grad_norm": 0.2422664910554886, "learning_rate": 3.121159870836633e-05, "loss": 0.0221, "step": 64150 }, { "grad_norm": 0.26225632429122925, "learning_rate": 3.119627680368312e-05, "loss": 0.0196, "step": 64160 }, { "grad_norm": 0.32486632466316223, "learning_rate": 3.1180956955346816e-05, "loss": 0.0248, "step": 64170 }, { "grad_norm": 0.2459881603717804, "learning_rate": 3.116563916503276e-05, "loss": 0.0211, "step": 64180 }, { "grad_norm": 0.16167400777339935, "learning_rate": 3.1150323434416095e-05, "loss": 0.0174, "step": 64190 }, { "grad_norm": 0.35249432921409607, "learning_rate": 3.1135009765171696e-05, "loss": 0.0209, "step": 64200 }, { "grad_norm": 0.24905288219451904, "learning_rate": 3.111969815897429e-05, "loss": 0.0193, "step": 64210 }, { "grad_norm": 0.1896214336156845, "learning_rate": 3.1104388617498295e-05, "loss": 0.0184, "step": 64220 }, { "grad_norm": 0.16270340979099274, "learning_rate": 3.1089081142417966e-05, "loss": 0.0176, "step": 64230 }, { "grad_norm": 0.22921498119831085, "learning_rate": 3.1073775735407285e-05, "loss": 0.0168, "step": 64240 }, { "grad_norm": 0.2931804358959198, "learning_rate": 3.1058472398140035e-05, "loss": 0.0219, "step": 64250 }, { "grad_norm": 0.20924749970436096, "learning_rate": 3.1043171132289736e-05, "loss": 0.0237, "step": 64260 }, { "grad_norm": 0.2506581246852875, "learning_rate": 3.1027871939529774e-05, "loss": 0.0199, "step": 64270 }, { "grad_norm": 0.21551159024238586, "learning_rate": 3.1012574821533214e-05, "loss": 0.0267, "step": 64280 }, { "grad_norm": 0.24466544389724731, "learning_rate": 3.099727977997291e-05, "loss": 0.0189, "step": 64290 }, { "grad_norm": 0.6675326824188232, "learning_rate": 3.098198681652154e-05, "loss": 0.0254, "step": 64300 }, { "grad_norm": 0.19730839133262634, "learning_rate": 3.0966695932851475e-05, "loss": 0.0205, "step": 64310 }, { "grad_norm": 0.2737921178340912, "learning_rate": 3.0951407130634926e-05, "loss": 0.0213, "step": 64320 }, { "grad_norm": 0.26416632533073425, "learning_rate": 3.0936120411543864e-05, "loss": 0.0166, "step": 64330 }, { "grad_norm": 0.25508734583854675, "learning_rate": 3.0920835777250015e-05, "loss": 0.0187, "step": 64340 }, { "grad_norm": 0.5462625622749329, "learning_rate": 3.090555322942488e-05, "loss": 0.0183, "step": 64350 }, { "grad_norm": 1.216229796409607, "learning_rate": 3.089027276973973e-05, "loss": 0.0289, "step": 64360 }, { "grad_norm": 0.2010754495859146, "learning_rate": 3.0874994399865615e-05, "loss": 0.0211, "step": 64370 }, { "grad_norm": 0.37934550642967224, "learning_rate": 3.085971812147337e-05, "loss": 0.0265, "step": 64380 }, { "grad_norm": 0.38640904426574707, "learning_rate": 3.0844443936233557e-05, "loss": 0.0174, "step": 64390 }, { "grad_norm": 0.49954545497894287, "learning_rate": 3.082917184581657e-05, "loss": 0.0298, "step": 64400 }, { "grad_norm": 0.2099291980266571, "learning_rate": 3.0813901851892513e-05, "loss": 0.0185, "step": 64410 }, { "grad_norm": 0.2782834768295288, "learning_rate": 3.0798633956131316e-05, "loss": 0.0274, "step": 64420 }, { "grad_norm": 0.3850732445716858, "learning_rate": 3.078336816020264e-05, "loss": 0.0205, "step": 64430 }, { "grad_norm": 0.24028272926807404, "learning_rate": 3.0768104465775935e-05, "loss": 0.0247, "step": 64440 }, { "grad_norm": 0.24455218017101288, "learning_rate": 3.0752842874520396e-05, "loss": 0.0182, "step": 64450 }, { "grad_norm": 0.24721626937389374, "learning_rate": 3.073758338810503e-05, "loss": 0.0179, "step": 64460 }, { "grad_norm": 0.35192441940307617, "learning_rate": 3.072232600819859e-05, "loss": 0.0216, "step": 64470 }, { "grad_norm": 0.23853187263011932, "learning_rate": 3.070707073646959e-05, "loss": 0.0254, "step": 64480 }, { "grad_norm": 0.25643298029899597, "learning_rate": 3.069181757458633e-05, "loss": 0.0206, "step": 64490 }, { "grad_norm": 0.31733474135398865, "learning_rate": 3.067656652421687e-05, "loss": 0.0167, "step": 64500 }, { "grad_norm": 0.23920421302318573, "learning_rate": 3.0661317587029025e-05, "loss": 0.0174, "step": 64510 }, { "grad_norm": 0.18143777549266815, "learning_rate": 3.064607076469043e-05, "loss": 0.0158, "step": 64520 }, { "grad_norm": 0.2802187204360962, "learning_rate": 3.063082605886845e-05, "loss": 0.02, "step": 64530 }, { "grad_norm": 0.13305813074111938, "learning_rate": 3.061558347123019e-05, "loss": 0.0196, "step": 64540 }, { "grad_norm": 0.2513536810874939, "learning_rate": 3.060034300344259e-05, "loss": 0.0199, "step": 64550 }, { "grad_norm": 0.24467529356479645, "learning_rate": 3.058510465717231e-05, "loss": 0.0254, "step": 64560 }, { "grad_norm": 0.23369437456130981, "learning_rate": 3.0569868434085804e-05, "loss": 0.0229, "step": 64570 }, { "grad_norm": 0.1943805068731308, "learning_rate": 3.055463433584924e-05, "loss": 0.021, "step": 64580 }, { "grad_norm": 0.2405952513217926, "learning_rate": 3.0539402364128656e-05, "loss": 0.0197, "step": 64590 }, { "grad_norm": 0.3414735198020935, "learning_rate": 3.052417252058977e-05, "loss": 0.0184, "step": 64600 }, { "grad_norm": 0.3627944588661194, "learning_rate": 3.0508944806898092e-05, "loss": 0.021, "step": 64610 }, { "grad_norm": 1.5134696960449219, "learning_rate": 3.04937192247189e-05, "loss": 0.0265, "step": 64620 }, { "grad_norm": 0.2000976949930191, "learning_rate": 3.047849577571725e-05, "loss": 0.0192, "step": 64630 }, { "grad_norm": 0.17524269223213196, "learning_rate": 3.046327446155795e-05, "loss": 0.0148, "step": 64640 }, { "grad_norm": 0.17382103204727173, "learning_rate": 3.0448055283905587e-05, "loss": 0.0157, "step": 64650 }, { "grad_norm": 0.22676150500774384, "learning_rate": 3.0432838244424494e-05, "loss": 0.0223, "step": 64660 }, { "grad_norm": 0.256503701210022, "learning_rate": 3.04176233447788e-05, "loss": 0.0178, "step": 64670 }, { "grad_norm": 0.15218748152256012, "learning_rate": 3.040241058663238e-05, "loss": 0.0217, "step": 64680 }, { "grad_norm": 0.1950240433216095, "learning_rate": 3.038719997164887e-05, "loss": 0.027, "step": 64690 }, { "grad_norm": 0.22710251808166504, "learning_rate": 3.0371991501491683e-05, "loss": 0.0232, "step": 64700 }, { "grad_norm": 0.27290669083595276, "learning_rate": 3.0356785177823975e-05, "loss": 0.0214, "step": 64710 }, { "grad_norm": 0.2531993091106415, "learning_rate": 3.0341581002308726e-05, "loss": 0.0154, "step": 64720 }, { "grad_norm": 0.13128793239593506, "learning_rate": 3.0326378976608627e-05, "loss": 0.0176, "step": 64730 }, { "grad_norm": 0.2227804809808731, "learning_rate": 3.0311179102386146e-05, "loss": 0.021, "step": 64740 }, { "grad_norm": 0.1473211795091629, "learning_rate": 3.0295981381303513e-05, "loss": 0.0187, "step": 64750 }, { "grad_norm": 0.23010960221290588, "learning_rate": 3.0280785815022727e-05, "loss": 0.0191, "step": 64760 }, { "grad_norm": 0.1922808736562729, "learning_rate": 3.0265592405205544e-05, "loss": 0.0186, "step": 64770 }, { "grad_norm": 0.3874109089374542, "learning_rate": 3.0250401153513517e-05, "loss": 0.0205, "step": 64780 }, { "grad_norm": 0.2804718315601349, "learning_rate": 3.0235212061607936e-05, "loss": 0.0249, "step": 64790 }, { "grad_norm": 0.27159884572029114, "learning_rate": 3.0220025131149842e-05, "loss": 0.0197, "step": 64800 }, { "grad_norm": 0.18141710758209229, "learning_rate": 3.020484036380007e-05, "loss": 0.0192, "step": 64810 }, { "grad_norm": 0.19674056768417358, "learning_rate": 3.0189657761219188e-05, "loss": 0.024, "step": 64820 }, { "grad_norm": 0.16781415045261383, "learning_rate": 3.017447732506753e-05, "loss": 0.0182, "step": 64830 }, { "grad_norm": 0.3616121709346771, "learning_rate": 3.015929905700524e-05, "loss": 0.0182, "step": 64840 }, { "grad_norm": 0.7468639612197876, "learning_rate": 3.014412295869218e-05, "loss": 0.0171, "step": 64850 }, { "grad_norm": 0.22822313010692596, "learning_rate": 3.0128949031787977e-05, "loss": 0.019, "step": 64860 }, { "grad_norm": 0.17959032952785492, "learning_rate": 3.0113777277952022e-05, "loss": 0.0195, "step": 64870 }, { "grad_norm": 0.20420946180820465, "learning_rate": 3.0098607698843496e-05, "loss": 0.0176, "step": 64880 }, { "grad_norm": 0.1865638941526413, "learning_rate": 3.0083440296121312e-05, "loss": 0.0165, "step": 64890 }, { "grad_norm": 0.35673773288726807, "learning_rate": 3.0068275071444142e-05, "loss": 0.0192, "step": 64900 }, { "grad_norm": 0.3189878761768341, "learning_rate": 3.0053112026470442e-05, "loss": 0.0253, "step": 64910 }, { "grad_norm": 0.2635301351547241, "learning_rate": 3.0037951162858434e-05, "loss": 0.0189, "step": 64920 }, { "grad_norm": 0.21851833164691925, "learning_rate": 3.002279248226607e-05, "loss": 0.0189, "step": 64930 }, { "grad_norm": 0.1565244197845459, "learning_rate": 3.000763598635109e-05, "loss": 0.0143, "step": 64940 }, { "grad_norm": 0.2990487813949585, "learning_rate": 2.999248167677098e-05, "loss": 0.0204, "step": 64950 }, { "grad_norm": 0.28666290640830994, "learning_rate": 2.9977329555182976e-05, "loss": 0.0262, "step": 64960 }, { "grad_norm": 0.37473753094673157, "learning_rate": 2.9962179623244124e-05, "loss": 0.023, "step": 64970 }, { "grad_norm": 0.19892120361328125, "learning_rate": 2.9947031882611175e-05, "loss": 0.0195, "step": 64980 }, { "grad_norm": 0.16437356173992157, "learning_rate": 2.9931886334940685e-05, "loss": 0.0167, "step": 64990 }, { "grad_norm": 0.3211299180984497, "learning_rate": 2.9916742981888923e-05, "loss": 0.0165, "step": 65000 }, { "grad_norm": 0.3712040185928345, "learning_rate": 2.990160182511196e-05, "loss": 0.0194, "step": 65010 }, { "grad_norm": 0.30321139097213745, "learning_rate": 2.9886462866265574e-05, "loss": 0.0249, "step": 65020 }, { "grad_norm": 0.2486986368894577, "learning_rate": 2.9871326107005392e-05, "loss": 0.0186, "step": 65030 }, { "grad_norm": 0.19291293621063232, "learning_rate": 2.9856191548986735e-05, "loss": 0.0185, "step": 65040 }, { "grad_norm": 0.1346118301153183, "learning_rate": 2.9841059193864672e-05, "loss": 0.019, "step": 65050 }, { "grad_norm": 0.1874442845582962, "learning_rate": 2.982592904329407e-05, "loss": 0.0228, "step": 65060 }, { "grad_norm": 0.3692358732223511, "learning_rate": 2.981080109892954e-05, "loss": 0.0189, "step": 65070 }, { "grad_norm": 0.22483929991722107, "learning_rate": 2.9795675362425434e-05, "loss": 0.021, "step": 65080 }, { "grad_norm": 0.245339497923851, "learning_rate": 2.97805518354359e-05, "loss": 0.0167, "step": 65090 }, { "grad_norm": 0.19755029678344727, "learning_rate": 2.9765430519614825e-05, "loss": 0.0161, "step": 65100 }, { "grad_norm": 0.14050187170505524, "learning_rate": 2.975031141661584e-05, "loss": 0.0136, "step": 65110 }, { "grad_norm": 0.17807386815547943, "learning_rate": 2.9735194528092347e-05, "loss": 0.024, "step": 65120 }, { "grad_norm": 0.24550040066242218, "learning_rate": 2.9720079855697525e-05, "loss": 0.0277, "step": 65130 }, { "grad_norm": 0.26027947664260864, "learning_rate": 2.970496740108427e-05, "loss": 0.0298, "step": 65140 }, { "grad_norm": 0.3834074139595032, "learning_rate": 2.968985716590527e-05, "loss": 0.0226, "step": 65150 }, { "grad_norm": 0.23262903094291687, "learning_rate": 2.9674749151812953e-05, "loss": 0.0151, "step": 65160 }, { "grad_norm": 0.2734106779098511, "learning_rate": 2.965964336045951e-05, "loss": 0.0181, "step": 65170 }, { "grad_norm": 0.23209963738918304, "learning_rate": 2.9644539793496894e-05, "loss": 0.0196, "step": 65180 }, { "grad_norm": 0.20205387473106384, "learning_rate": 2.9629438452576807e-05, "loss": 0.0179, "step": 65190 }, { "grad_norm": 0.18749593198299408, "learning_rate": 2.9614339339350698e-05, "loss": 0.0179, "step": 65200 }, { "grad_norm": 0.23094487190246582, "learning_rate": 2.959924245546979e-05, "loss": 0.0207, "step": 65210 }, { "grad_norm": 0.3780864477157593, "learning_rate": 2.9584147802585038e-05, "loss": 0.0199, "step": 65220 }, { "grad_norm": 0.2516965866088867, "learning_rate": 2.956905538234721e-05, "loss": 0.0245, "step": 65230 }, { "grad_norm": 0.28931725025177, "learning_rate": 2.9553965196406764e-05, "loss": 0.0229, "step": 65240 }, { "grad_norm": 0.18661868572235107, "learning_rate": 2.9538877246413943e-05, "loss": 0.0175, "step": 65250 }, { "grad_norm": 0.2641508877277374, "learning_rate": 2.952379153401874e-05, "loss": 0.0255, "step": 65260 }, { "grad_norm": 0.3274953067302704, "learning_rate": 2.9508708060870914e-05, "loss": 0.0234, "step": 65270 }, { "grad_norm": 0.5385094881057739, "learning_rate": 2.9493626828619948e-05, "loss": 0.0196, "step": 65280 }, { "grad_norm": 0.28391918540000916, "learning_rate": 2.9478547838915137e-05, "loss": 0.0177, "step": 65290 }, { "grad_norm": 0.18225319683551788, "learning_rate": 2.946347109340548e-05, "loss": 0.0206, "step": 65300 }, { "grad_norm": 0.2618941068649292, "learning_rate": 2.944839659373976e-05, "loss": 0.0233, "step": 65310 }, { "grad_norm": 0.3191359341144562, "learning_rate": 2.943332434156647e-05, "loss": 0.0261, "step": 65320 }, { "grad_norm": 0.26307588815689087, "learning_rate": 2.9418254338533912e-05, "loss": 0.0191, "step": 65330 }, { "grad_norm": 0.1842251718044281, "learning_rate": 2.940318658629011e-05, "loss": 0.0181, "step": 65340 }, { "grad_norm": 0.21306654810905457, "learning_rate": 2.938812108648287e-05, "loss": 0.0181, "step": 65350 }, { "grad_norm": 0.24691838026046753, "learning_rate": 2.937305784075972e-05, "loss": 0.0181, "step": 65360 }, { "grad_norm": 0.15099184215068817, "learning_rate": 2.9357996850767937e-05, "loss": 0.014, "step": 65370 }, { "grad_norm": 0.29726147651672363, "learning_rate": 2.9342938118154595e-05, "loss": 0.0195, "step": 65380 }, { "grad_norm": 0.37107136845588684, "learning_rate": 2.9327881644566485e-05, "loss": 0.0307, "step": 65390 }, { "grad_norm": 0.19703450798988342, "learning_rate": 2.9312827431650157e-05, "loss": 0.0154, "step": 65400 }, { "grad_norm": 0.4055021405220032, "learning_rate": 2.92977754810519e-05, "loss": 0.0278, "step": 65410 }, { "grad_norm": 0.15352539718151093, "learning_rate": 2.928272579441781e-05, "loss": 0.024, "step": 65420 }, { "grad_norm": 0.310269832611084, "learning_rate": 2.9267678373393676e-05, "loss": 0.0228, "step": 65430 }, { "grad_norm": 0.24061904847621918, "learning_rate": 2.925263321962507e-05, "loss": 0.0233, "step": 65440 }, { "grad_norm": 0.33825936913490295, "learning_rate": 2.9237590334757292e-05, "loss": 0.0189, "step": 65450 }, { "grad_norm": 0.16041019558906555, "learning_rate": 2.922254972043543e-05, "loss": 0.021, "step": 65460 }, { "grad_norm": 0.2848949134349823, "learning_rate": 2.9207511378304265e-05, "loss": 0.0191, "step": 65470 }, { "grad_norm": 0.24340328574180603, "learning_rate": 2.9192475310008428e-05, "loss": 0.0211, "step": 65480 }, { "grad_norm": 0.32469844818115234, "learning_rate": 2.917744151719219e-05, "loss": 0.0202, "step": 65490 }, { "grad_norm": 0.33972904086112976, "learning_rate": 2.916241000149964e-05, "loss": 0.0262, "step": 65500 }, { "grad_norm": 0.32259055972099304, "learning_rate": 2.9147380764574622e-05, "loss": 0.0216, "step": 65510 }, { "grad_norm": 0.23103366792201996, "learning_rate": 2.9132353808060674e-05, "loss": 0.0274, "step": 65520 }, { "grad_norm": 0.1683027595281601, "learning_rate": 2.9117329133601134e-05, "loss": 0.0295, "step": 65530 }, { "grad_norm": 0.21935264766216278, "learning_rate": 2.9102306742839093e-05, "loss": 0.0157, "step": 65540 }, { "grad_norm": 0.21826006472110748, "learning_rate": 2.9087286637417375e-05, "loss": 0.0238, "step": 65550 }, { "grad_norm": 0.17370417714118958, "learning_rate": 2.9072268818978537e-05, "loss": 0.0208, "step": 65560 }, { "grad_norm": 0.36945199966430664, "learning_rate": 2.905725328916493e-05, "loss": 0.0176, "step": 65570 }, { "grad_norm": 0.3059404492378235, "learning_rate": 2.9042240049618608e-05, "loss": 0.0246, "step": 65580 }, { "grad_norm": 0.27738726139068604, "learning_rate": 2.9027229101981413e-05, "loss": 0.0201, "step": 65590 }, { "grad_norm": 0.2307407557964325, "learning_rate": 2.901222044789488e-05, "loss": 0.0148, "step": 65600 }, { "grad_norm": 0.24524997174739838, "learning_rate": 2.89972140890004e-05, "loss": 0.0223, "step": 65610 }, { "grad_norm": 0.3424474000930786, "learning_rate": 2.8982210026938988e-05, "loss": 0.0254, "step": 65620 }, { "grad_norm": 0.22543315589427948, "learning_rate": 2.896720826335151e-05, "loss": 0.0216, "step": 65630 }, { "grad_norm": 0.22134587168693542, "learning_rate": 2.8952208799878493e-05, "loss": 0.0187, "step": 65640 }, { "grad_norm": 0.27951717376708984, "learning_rate": 2.89372116381603e-05, "loss": 0.015, "step": 65650 }, { "grad_norm": 0.2473505139350891, "learning_rate": 2.8922216779836932e-05, "loss": 0.0183, "step": 65660 }, { "grad_norm": 0.3070608973503113, "learning_rate": 2.8907224226548285e-05, "loss": 0.0198, "step": 65670 }, { "grad_norm": 0.31947407126426697, "learning_rate": 2.8892233979933868e-05, "loss": 0.0172, "step": 65680 }, { "grad_norm": 0.19894467294216156, "learning_rate": 2.8877246041633026e-05, "loss": 0.0203, "step": 65690 }, { "grad_norm": 0.20478998124599457, "learning_rate": 2.886226041328478e-05, "loss": 0.0155, "step": 65700 }, { "grad_norm": 0.20103459060192108, "learning_rate": 2.8847277096527957e-05, "loss": 0.0187, "step": 65710 }, { "grad_norm": 0.2255065143108368, "learning_rate": 2.8832296093001105e-05, "loss": 0.0192, "step": 65720 }, { "grad_norm": 0.1852157562971115, "learning_rate": 2.8817317404342547e-05, "loss": 0.016, "step": 65730 }, { "grad_norm": 0.4391062557697296, "learning_rate": 2.880234103219029e-05, "loss": 0.0221, "step": 65740 }, { "grad_norm": 0.1985049992799759, "learning_rate": 2.878736697818215e-05, "loss": 0.0167, "step": 65750 }, { "grad_norm": 0.3495914041996002, "learning_rate": 2.877239524395568e-05, "loss": 0.0222, "step": 65760 }, { "grad_norm": 0.48824551701545715, "learning_rate": 2.8757425831148134e-05, "loss": 0.0188, "step": 65770 }, { "grad_norm": 0.35849031805992126, "learning_rate": 2.874245874139658e-05, "loss": 0.022, "step": 65780 }, { "grad_norm": 1.0477814674377441, "learning_rate": 2.8727493976337737e-05, "loss": 0.0235, "step": 65790 }, { "grad_norm": 0.44701454043388367, "learning_rate": 2.8712531537608218e-05, "loss": 0.0177, "step": 65800 }, { "grad_norm": 0.35271093249320984, "learning_rate": 2.8697571426844217e-05, "loss": 0.0172, "step": 65810 }, { "grad_norm": 0.23707976937294006, "learning_rate": 2.86826136456818e-05, "loss": 0.0203, "step": 65820 }, { "grad_norm": 0.25336122512817383, "learning_rate": 2.8667658195756685e-05, "loss": 0.0174, "step": 65830 }, { "grad_norm": 0.19361107051372528, "learning_rate": 2.8652705078704412e-05, "loss": 0.0238, "step": 65840 }, { "grad_norm": 0.24578362703323364, "learning_rate": 2.863775429616018e-05, "loss": 0.0207, "step": 65850 }, { "grad_norm": 0.25134149193763733, "learning_rate": 2.862280584975906e-05, "loss": 0.0214, "step": 65860 }, { "grad_norm": 0.28972312808036804, "learning_rate": 2.8607859741135735e-05, "loss": 0.0222, "step": 65870 }, { "grad_norm": 0.25337934494018555, "learning_rate": 2.8592915971924723e-05, "loss": 0.021, "step": 65880 }, { "grad_norm": 0.3185546398162842, "learning_rate": 2.8577974543760212e-05, "loss": 0.0162, "step": 65890 }, { "grad_norm": 0.23059578239917755, "learning_rate": 2.8563035458276222e-05, "loss": 0.0244, "step": 65900 }, { "grad_norm": 0.22967587411403656, "learning_rate": 2.8548098717106424e-05, "loss": 0.02, "step": 65910 }, { "grad_norm": 0.3581782877445221, "learning_rate": 2.85331643218843e-05, "loss": 0.0177, "step": 65920 }, { "grad_norm": 0.1689385622739792, "learning_rate": 2.8518232274243046e-05, "loss": 0.0196, "step": 65930 }, { "grad_norm": 0.3689733147621155, "learning_rate": 2.8503302575815638e-05, "loss": 0.0177, "step": 65940 }, { "grad_norm": 0.2718534469604492, "learning_rate": 2.8488375228234716e-05, "loss": 0.025, "step": 65950 }, { "grad_norm": 0.28036704659461975, "learning_rate": 2.8473450233132738e-05, "loss": 0.0221, "step": 65960 }, { "grad_norm": 0.152693510055542, "learning_rate": 2.8458527592141898e-05, "loss": 0.0164, "step": 65970 }, { "grad_norm": 0.24254952371120453, "learning_rate": 2.8443607306894073e-05, "loss": 0.0209, "step": 65980 }, { "grad_norm": 0.18212899565696716, "learning_rate": 2.8428689379020946e-05, "loss": 0.021, "step": 65990 }, { "grad_norm": 0.22967886924743652, "learning_rate": 2.841377381015391e-05, "loss": 0.0184, "step": 66000 }, { "grad_norm": 0.5405160784721375, "learning_rate": 2.8398860601924144e-05, "loss": 0.0178, "step": 66010 }, { "grad_norm": 0.381975919008255, "learning_rate": 2.838394975596248e-05, "loss": 0.0177, "step": 66020 }, { "grad_norm": 0.3137006461620331, "learning_rate": 2.83690412738996e-05, "loss": 0.0204, "step": 66030 }, { "grad_norm": 0.712196946144104, "learning_rate": 2.835413515736579e-05, "loss": 0.0204, "step": 66040 }, { "grad_norm": 0.21533489227294922, "learning_rate": 2.8339231407991267e-05, "loss": 0.0226, "step": 66050 }, { "grad_norm": 0.1406676024198532, "learning_rate": 2.832433002740581e-05, "loss": 0.0113, "step": 66060 }, { "grad_norm": 0.288526713848114, "learning_rate": 2.8309431017239047e-05, "loss": 0.0154, "step": 66070 }, { "grad_norm": 0.20349077880382538, "learning_rate": 2.829453437912028e-05, "loss": 0.0155, "step": 66080 }, { "grad_norm": 0.21743173897266388, "learning_rate": 2.8279640114678617e-05, "loss": 0.0164, "step": 66090 }, { "grad_norm": 0.21612077951431274, "learning_rate": 2.8264748225542836e-05, "loss": 0.0224, "step": 66100 }, { "grad_norm": 0.253248393535614, "learning_rate": 2.8249858713341516e-05, "loss": 0.0196, "step": 66110 }, { "grad_norm": 0.19283224642276764, "learning_rate": 2.823497157970294e-05, "loss": 0.0178, "step": 66120 }, { "grad_norm": 0.16423602402210236, "learning_rate": 2.8220086826255164e-05, "loss": 0.014, "step": 66130 }, { "grad_norm": 0.17633652687072754, "learning_rate": 2.8205204454625934e-05, "loss": 0.0199, "step": 66140 }, { "grad_norm": 0.14193251729011536, "learning_rate": 2.8190324466442798e-05, "loss": 0.0147, "step": 66150 }, { "grad_norm": 0.1893300861120224, "learning_rate": 2.8175446863332965e-05, "loss": 0.0117, "step": 66160 }, { "grad_norm": 0.2524695098400116, "learning_rate": 2.816057164692345e-05, "loss": 0.0209, "step": 66170 }, { "grad_norm": 0.22459737956523895, "learning_rate": 2.8145698818840992e-05, "loss": 0.0199, "step": 66180 }, { "grad_norm": 0.16028355062007904, "learning_rate": 2.8130828380712065e-05, "loss": 0.0232, "step": 66190 }, { "grad_norm": 0.20678414404392242, "learning_rate": 2.811596033416285e-05, "loss": 0.0213, "step": 66200 }, { "grad_norm": 0.33314502239227295, "learning_rate": 2.8101094680819307e-05, "loss": 0.0175, "step": 66210 }, { "grad_norm": 0.18379878997802734, "learning_rate": 2.808623142230715e-05, "loss": 0.0193, "step": 66220 }, { "grad_norm": 0.20805884897708893, "learning_rate": 2.8071370560251757e-05, "loss": 0.015, "step": 66230 }, { "grad_norm": 0.40242093801498413, "learning_rate": 2.805651209627831e-05, "loss": 0.02, "step": 66240 }, { "grad_norm": 0.7666864991188049, "learning_rate": 2.804165603201171e-05, "loss": 0.0213, "step": 66250 }, { "grad_norm": 0.2956392467021942, "learning_rate": 2.802680236907661e-05, "loss": 0.0178, "step": 66260 }, { "grad_norm": 0.215886652469635, "learning_rate": 2.8011951109097344e-05, "loss": 0.0167, "step": 66270 }, { "grad_norm": 0.47403740882873535, "learning_rate": 2.7997102253698072e-05, "loss": 0.0179, "step": 66280 }, { "grad_norm": 0.2544442117214203, "learning_rate": 2.7982255804502588e-05, "loss": 0.0156, "step": 66290 }, { "grad_norm": 0.3050386309623718, "learning_rate": 2.7967411763134516e-05, "loss": 0.0168, "step": 66300 }, { "grad_norm": 0.3762936592102051, "learning_rate": 2.7952570131217166e-05, "loss": 0.0198, "step": 66310 }, { "grad_norm": 0.2889019548892975, "learning_rate": 2.7937730910373618e-05, "loss": 0.0229, "step": 66320 }, { "grad_norm": 0.22425797581672668, "learning_rate": 2.7922894102226628e-05, "loss": 0.0196, "step": 66330 }, { "grad_norm": 0.15130172669887543, "learning_rate": 2.790805970839876e-05, "loss": 0.0165, "step": 66340 }, { "grad_norm": 0.21691925823688507, "learning_rate": 2.789322773051225e-05, "loss": 0.014, "step": 66350 }, { "grad_norm": 0.27132201194763184, "learning_rate": 2.7878398170189124e-05, "loss": 0.0186, "step": 66360 }, { "grad_norm": 0.17222648859024048, "learning_rate": 2.7863571029051112e-05, "loss": 0.0178, "step": 66370 }, { "grad_norm": 0.2788502871990204, "learning_rate": 2.784874630871971e-05, "loss": 0.0229, "step": 66380 }, { "grad_norm": 0.2770897150039673, "learning_rate": 2.7833924010816086e-05, "loss": 0.0197, "step": 66390 }, { "grad_norm": 0.4005102217197418, "learning_rate": 2.7819104136961226e-05, "loss": 0.0197, "step": 66400 }, { "grad_norm": 0.2594197392463684, "learning_rate": 2.780428668877577e-05, "loss": 0.0201, "step": 66410 }, { "grad_norm": 0.546837568283081, "learning_rate": 2.7789471667880152e-05, "loss": 0.0237, "step": 66420 }, { "grad_norm": 0.16501714289188385, "learning_rate": 2.7774659075894523e-05, "loss": 0.0219, "step": 66430 }, { "grad_norm": 0.2218000441789627, "learning_rate": 2.7759848914438778e-05, "loss": 0.0189, "step": 66440 }, { "grad_norm": 0.3712881803512573, "learning_rate": 2.7745041185132496e-05, "loss": 0.0197, "step": 66450 }, { "grad_norm": 0.27380046248435974, "learning_rate": 2.773023588959506e-05, "loss": 0.0204, "step": 66460 }, { "grad_norm": 0.2695342004299164, "learning_rate": 2.7715433029445555e-05, "loss": 0.0212, "step": 66470 }, { "grad_norm": 0.2641507387161255, "learning_rate": 2.7700632606302766e-05, "loss": 0.0167, "step": 66480 }, { "grad_norm": 0.2577228844165802, "learning_rate": 2.768583462178528e-05, "loss": 0.0219, "step": 66490 }, { "grad_norm": 0.35981255769729614, "learning_rate": 2.767103907751136e-05, "loss": 0.0199, "step": 66500 }, { "grad_norm": 0.3789379596710205, "learning_rate": 2.7656245975099065e-05, "loss": 0.0192, "step": 66510 }, { "grad_norm": 0.4568842053413391, "learning_rate": 2.7641455316166093e-05, "loss": 0.026, "step": 66520 }, { "grad_norm": 0.232621967792511, "learning_rate": 2.7626667102329973e-05, "loss": 0.0166, "step": 66530 }, { "grad_norm": 0.33123546838760376, "learning_rate": 2.761188133520788e-05, "loss": 0.0184, "step": 66540 }, { "grad_norm": 0.6437942385673523, "learning_rate": 2.7597098016416784e-05, "loss": 0.0192, "step": 66550 }, { "grad_norm": 0.18258348107337952, "learning_rate": 2.7582317147573368e-05, "loss": 0.0203, "step": 66560 }, { "grad_norm": 0.31762614846229553, "learning_rate": 2.7567538730294064e-05, "loss": 0.0213, "step": 66570 }, { "grad_norm": 0.3876390755176544, "learning_rate": 2.7552762766194973e-05, "loss": 0.0212, "step": 66580 }, { "grad_norm": 0.20847192406654358, "learning_rate": 2.753798925689202e-05, "loss": 0.0257, "step": 66590 }, { "grad_norm": 0.24030892550945282, "learning_rate": 2.7523218204000774e-05, "loss": 0.0216, "step": 66600 }, { "grad_norm": 0.21036690473556519, "learning_rate": 2.7508449609136612e-05, "loss": 0.022, "step": 66610 }, { "grad_norm": 0.26773303747177124, "learning_rate": 2.749368347391454e-05, "loss": 0.0172, "step": 66620 }, { "grad_norm": 0.17324045300483704, "learning_rate": 2.7478919799949443e-05, "loss": 0.0209, "step": 66630 }, { "grad_norm": 0.35110974311828613, "learning_rate": 2.7464158588855803e-05, "loss": 0.0167, "step": 66640 }, { "grad_norm": 0.3051111400127411, "learning_rate": 2.7449399842247918e-05, "loss": 0.0204, "step": 66650 }, { "grad_norm": 0.4434513747692108, "learning_rate": 2.7434643561739736e-05, "loss": 0.0271, "step": 66660 }, { "grad_norm": 0.36442825198173523, "learning_rate": 2.7419889748945016e-05, "loss": 0.0186, "step": 66670 }, { "grad_norm": 0.21142391860485077, "learning_rate": 2.7405138405477203e-05, "loss": 0.0206, "step": 66680 }, { "grad_norm": 0.6185985803604126, "learning_rate": 2.73903895329495e-05, "loss": 0.0238, "step": 66690 }, { "grad_norm": 0.2543701231479645, "learning_rate": 2.737564313297478e-05, "loss": 0.018, "step": 66700 }, { "grad_norm": 0.27394407987594604, "learning_rate": 2.736089920716572e-05, "loss": 0.0187, "step": 66710 }, { "grad_norm": 0.23641204833984375, "learning_rate": 2.7346157757134706e-05, "loss": 0.0175, "step": 66720 }, { "grad_norm": 0.2889525592327118, "learning_rate": 2.733141878449379e-05, "loss": 0.0172, "step": 66730 }, { "grad_norm": 0.1946924775838852, "learning_rate": 2.7316682290854844e-05, "loss": 0.0164, "step": 66740 }, { "grad_norm": 0.34158530831336975, "learning_rate": 2.7301948277829414e-05, "loss": 0.0194, "step": 66750 }, { "grad_norm": 0.2934848964214325, "learning_rate": 2.7287216747028814e-05, "loss": 0.0188, "step": 66760 }, { "grad_norm": 0.29187721014022827, "learning_rate": 2.7272487700064025e-05, "loss": 0.0203, "step": 66770 }, { "grad_norm": 0.14374732971191406, "learning_rate": 2.725776113854583e-05, "loss": 0.0195, "step": 66780 }, { "grad_norm": 0.4374508559703827, "learning_rate": 2.7243037064084664e-05, "loss": 0.0176, "step": 66790 }, { "grad_norm": 0.45864227414131165, "learning_rate": 2.7228315478290766e-05, "loss": 0.0283, "step": 66800 }, { "grad_norm": 0.28063535690307617, "learning_rate": 2.721359638277402e-05, "loss": 0.0182, "step": 66810 }, { "grad_norm": 0.22159388661384583, "learning_rate": 2.7198879779144148e-05, "loss": 0.0183, "step": 66820 }, { "grad_norm": 0.3270178735256195, "learning_rate": 2.7184165669010485e-05, "loss": 0.018, "step": 66830 }, { "grad_norm": 0.5773646235466003, "learning_rate": 2.716945405398218e-05, "loss": 0.0215, "step": 66840 }, { "grad_norm": 0.2112795114517212, "learning_rate": 2.715474493566803e-05, "loss": 0.014, "step": 66850 }, { "grad_norm": 0.24655164778232574, "learning_rate": 2.7140038315676652e-05, "loss": 0.0173, "step": 66860 }, { "grad_norm": 0.33360999822616577, "learning_rate": 2.7125334195616275e-05, "loss": 0.0312, "step": 66870 }, { "grad_norm": 0.21684211492538452, "learning_rate": 2.7110632577094997e-05, "loss": 0.0201, "step": 66880 }, { "grad_norm": 0.2412978708744049, "learning_rate": 2.709593346172051e-05, "loss": 0.0215, "step": 66890 }, { "grad_norm": 0.1831061691045761, "learning_rate": 2.708123685110032e-05, "loss": 0.02, "step": 66900 }, { "grad_norm": 0.3328489661216736, "learning_rate": 2.7066542746841588e-05, "loss": 0.0142, "step": 66910 }, { "grad_norm": 0.2820464074611664, "learning_rate": 2.7051851150551267e-05, "loss": 0.0175, "step": 66920 }, { "grad_norm": 0.17699308693408966, "learning_rate": 2.7037162063835996e-05, "loss": 0.0162, "step": 66930 }, { "grad_norm": 0.19653263688087463, "learning_rate": 2.702247548830218e-05, "loss": 0.016, "step": 66940 }, { "grad_norm": 0.22555437684059143, "learning_rate": 2.7007791425555883e-05, "loss": 0.0149, "step": 66950 }, { "grad_norm": 0.2685612440109253, "learning_rate": 2.6993109877202945e-05, "loss": 0.0179, "step": 66960 }, { "grad_norm": 0.3179655969142914, "learning_rate": 2.6978430844848935e-05, "loss": 0.025, "step": 66970 }, { "grad_norm": 0.24759003520011902, "learning_rate": 2.69637543300991e-05, "loss": 0.0145, "step": 66980 }, { "grad_norm": 0.3669315278530121, "learning_rate": 2.694908033455848e-05, "loss": 0.0148, "step": 66990 }, { "grad_norm": 0.3060433268547058, "learning_rate": 2.6934408859831733e-05, "loss": 0.0188, "step": 67000 }, { "grad_norm": 0.21183907985687256, "learning_rate": 2.69197399075234e-05, "loss": 0.0186, "step": 67010 }, { "grad_norm": 0.18227317929267883, "learning_rate": 2.6905073479237584e-05, "loss": 0.0203, "step": 67020 }, { "grad_norm": 0.20302556455135345, "learning_rate": 2.6890409576578235e-05, "loss": 0.0196, "step": 67030 }, { "grad_norm": 0.20704159140586853, "learning_rate": 2.687574820114892e-05, "loss": 0.0161, "step": 67040 }, { "grad_norm": 0.35022228956222534, "learning_rate": 2.686108935455305e-05, "loss": 0.0225, "step": 67050 }, { "grad_norm": 0.2965242266654968, "learning_rate": 2.6846433038393603e-05, "loss": 0.0211, "step": 67060 }, { "grad_norm": 0.2620284855365753, "learning_rate": 2.6831779254273482e-05, "loss": 0.0182, "step": 67070 }, { "grad_norm": 0.40837839245796204, "learning_rate": 2.6817128003795122e-05, "loss": 0.023, "step": 67080 }, { "grad_norm": 0.2400922328233719, "learning_rate": 2.680247928856081e-05, "loss": 0.0195, "step": 67090 }, { "grad_norm": 0.22105836868286133, "learning_rate": 2.6787833110172466e-05, "loss": 0.0158, "step": 67100 }, { "grad_norm": 0.15889538824558258, "learning_rate": 2.6773189470231807e-05, "loss": 0.0173, "step": 67110 }, { "grad_norm": 0.23757287859916687, "learning_rate": 2.675854837034022e-05, "loss": 0.0209, "step": 67120 }, { "grad_norm": 0.5015692114830017, "learning_rate": 2.6743909812098833e-05, "loss": 0.0169, "step": 67130 }, { "grad_norm": 0.20572873950004578, "learning_rate": 2.67292737971085e-05, "loss": 0.0225, "step": 67140 }, { "grad_norm": 0.32618600130081177, "learning_rate": 2.671464032696982e-05, "loss": 0.0203, "step": 67150 }, { "grad_norm": 0.2576082646846771, "learning_rate": 2.6700009403283055e-05, "loss": 0.0162, "step": 67160 }, { "grad_norm": 0.33076968789100647, "learning_rate": 2.6685381027648214e-05, "loss": 0.0167, "step": 67170 }, { "grad_norm": 0.3073323965072632, "learning_rate": 2.6670755201665086e-05, "loss": 0.0304, "step": 67180 }, { "grad_norm": 0.17047840356826782, "learning_rate": 2.6656131926933066e-05, "loss": 0.0145, "step": 67190 }, { "grad_norm": 0.24891822040081024, "learning_rate": 2.664151120505136e-05, "loss": 0.0237, "step": 67200 }, { "grad_norm": 0.3042602837085724, "learning_rate": 2.6626893037618873e-05, "loss": 0.0183, "step": 67210 }, { "grad_norm": 0.26031097769737244, "learning_rate": 2.6612277426234243e-05, "loss": 0.0228, "step": 67220 }, { "grad_norm": 0.21969535946846008, "learning_rate": 2.659766437249577e-05, "loss": 0.0209, "step": 67230 }, { "grad_norm": 0.223881334066391, "learning_rate": 2.6583053878001557e-05, "loss": 0.0169, "step": 67240 }, { "grad_norm": 0.2581663131713867, "learning_rate": 2.6568445944349323e-05, "loss": 0.0186, "step": 67250 }, { "grad_norm": 0.2013711929321289, "learning_rate": 2.6553840573136657e-05, "loss": 0.0155, "step": 67260 }, { "grad_norm": 0.2563018500804901, "learning_rate": 2.6539237765960706e-05, "loss": 0.0202, "step": 67270 }, { "grad_norm": 0.5251278877258301, "learning_rate": 2.6524637524418465e-05, "loss": 0.0209, "step": 67280 }, { "grad_norm": 0.19993478059768677, "learning_rate": 2.651003985010655e-05, "loss": 0.016, "step": 67290 }, { "grad_norm": 0.20907263457775116, "learning_rate": 2.649544474462138e-05, "loss": 0.0162, "step": 67300 }, { "grad_norm": 0.43077918887138367, "learning_rate": 2.6480852209559015e-05, "loss": 0.0173, "step": 67310 }, { "grad_norm": 0.27529260516166687, "learning_rate": 2.6466262246515296e-05, "loss": 0.0185, "step": 67320 }, { "grad_norm": 0.29421064257621765, "learning_rate": 2.6451674857085746e-05, "loss": 0.017, "step": 67330 }, { "grad_norm": 0.30188533663749695, "learning_rate": 2.6437090042865655e-05, "loss": 0.0169, "step": 67340 }, { "grad_norm": 0.1839655637741089, "learning_rate": 2.6422507805449947e-05, "loss": 0.0156, "step": 67350 }, { "grad_norm": 0.19315505027770996, "learning_rate": 2.6407928146433357e-05, "loss": 0.0131, "step": 67360 }, { "grad_norm": 0.36567696928977966, "learning_rate": 2.639335106741026e-05, "loss": 0.0169, "step": 67370 }, { "grad_norm": 0.18697747588157654, "learning_rate": 2.6378776569974796e-05, "loss": 0.0221, "step": 67380 }, { "grad_norm": 0.2568453252315521, "learning_rate": 2.6364204655720814e-05, "loss": 0.0147, "step": 67390 }, { "grad_norm": 0.19462096691131592, "learning_rate": 2.6349635326241894e-05, "loss": 0.0182, "step": 67400 }, { "grad_norm": 0.44898271560668945, "learning_rate": 2.633506858313127e-05, "loss": 0.0181, "step": 67410 }, { "grad_norm": 0.2547313868999481, "learning_rate": 2.6320504427981975e-05, "loss": 0.0201, "step": 67420 }, { "grad_norm": 0.1998068392276764, "learning_rate": 2.6305942862386734e-05, "loss": 0.0176, "step": 67430 }, { "grad_norm": 0.2591383457183838, "learning_rate": 2.629138388793793e-05, "loss": 0.0219, "step": 67440 }, { "grad_norm": 0.2691422998905182, "learning_rate": 2.6276827506227743e-05, "loss": 0.021, "step": 67450 }, { "grad_norm": 0.2656872570514679, "learning_rate": 2.6262273718848037e-05, "loss": 0.0153, "step": 67460 }, { "grad_norm": 0.23948021233081818, "learning_rate": 2.6247722527390407e-05, "loss": 0.0165, "step": 67470 }, { "grad_norm": 0.1873885840177536, "learning_rate": 2.6233173933446104e-05, "loss": 0.0167, "step": 67480 }, { "grad_norm": 0.22586530447006226, "learning_rate": 2.6218627938606188e-05, "loss": 0.0251, "step": 67490 }, { "grad_norm": 0.20691083371639252, "learning_rate": 2.620408454446135e-05, "loss": 0.0176, "step": 67500 }, { "grad_norm": 0.14157819747924805, "learning_rate": 2.6189543752602046e-05, "loss": 0.0207, "step": 67510 }, { "grad_norm": 0.23185423016548157, "learning_rate": 2.6175005564618444e-05, "loss": 0.0226, "step": 67520 }, { "grad_norm": 0.28272679448127747, "learning_rate": 2.6160469982100428e-05, "loss": 0.0279, "step": 67530 }, { "grad_norm": 0.3649405837059021, "learning_rate": 2.6145937006637555e-05, "loss": 0.0181, "step": 67540 }, { "grad_norm": 0.26888808608055115, "learning_rate": 2.6131406639819166e-05, "loss": 0.0202, "step": 67550 }, { "grad_norm": 0.11805086582899094, "learning_rate": 2.611687888323424e-05, "loss": 0.0219, "step": 67560 }, { "grad_norm": 0.4703974723815918, "learning_rate": 2.610235373847153e-05, "loss": 0.0207, "step": 67570 }, { "grad_norm": 0.31094059348106384, "learning_rate": 2.6087831207119494e-05, "loss": 0.0214, "step": 67580 }, { "grad_norm": 0.3091999292373657, "learning_rate": 2.60733112907663e-05, "loss": 0.0233, "step": 67590 }, { "grad_norm": 0.2530962824821472, "learning_rate": 2.6058793990999786e-05, "loss": 0.0203, "step": 67600 }, { "grad_norm": 0.43526676297187805, "learning_rate": 2.6044279309407593e-05, "loss": 0.0204, "step": 67610 }, { "grad_norm": 0.349735826253891, "learning_rate": 2.6029767247576976e-05, "loss": 0.0202, "step": 67620 }, { "grad_norm": 0.34146052598953247, "learning_rate": 2.601525780709497e-05, "loss": 0.0204, "step": 67630 }, { "grad_norm": 0.14406687021255493, "learning_rate": 2.6000750989548318e-05, "loss": 0.0154, "step": 67640 }, { "grad_norm": 0.1705549657344818, "learning_rate": 2.5986246796523462e-05, "loss": 0.0167, "step": 67650 }, { "grad_norm": 0.25681206583976746, "learning_rate": 2.5971745229606546e-05, "loss": 0.017, "step": 67660 }, { "grad_norm": 0.39475885033607483, "learning_rate": 2.5957246290383442e-05, "loss": 0.02, "step": 67670 }, { "grad_norm": 0.3381889760494232, "learning_rate": 2.594274998043975e-05, "loss": 0.0193, "step": 67680 }, { "grad_norm": 0.2722413241863251, "learning_rate": 2.5928256301360733e-05, "loss": 0.0195, "step": 67690 }, { "grad_norm": 0.22836390137672424, "learning_rate": 2.591376525473142e-05, "loss": 0.0204, "step": 67700 }, { "grad_norm": 0.21867965161800385, "learning_rate": 2.5899276842136523e-05, "loss": 0.0159, "step": 67710 }, { "grad_norm": 0.17305901646614075, "learning_rate": 2.5884791065160495e-05, "loss": 0.023, "step": 67720 }, { "grad_norm": 0.26573777198791504, "learning_rate": 2.5870307925387436e-05, "loss": 0.0177, "step": 67730 }, { "grad_norm": 0.44095224142074585, "learning_rate": 2.585582742440125e-05, "loss": 0.0201, "step": 67740 }, { "grad_norm": 0.23503051698207855, "learning_rate": 2.5841349563785455e-05, "loss": 0.0204, "step": 67750 }, { "grad_norm": 0.3074599802494049, "learning_rate": 2.5826874345123355e-05, "loss": 0.0176, "step": 67760 }, { "grad_norm": 0.2098325490951538, "learning_rate": 2.581240176999792e-05, "loss": 0.0145, "step": 67770 }, { "grad_norm": 0.32785090804100037, "learning_rate": 2.579793183999189e-05, "loss": 0.0144, "step": 67780 }, { "grad_norm": 0.20424090325832367, "learning_rate": 2.5783464556687627e-05, "loss": 0.0183, "step": 67790 }, { "grad_norm": 0.5876897573471069, "learning_rate": 2.576899992166729e-05, "loss": 0.0205, "step": 67800 }, { "grad_norm": 0.20710279047489166, "learning_rate": 2.5754537936512668e-05, "loss": 0.0194, "step": 67810 }, { "grad_norm": 0.5143638253211975, "learning_rate": 2.574007860280535e-05, "loss": 0.0257, "step": 67820 }, { "grad_norm": 0.22481904923915863, "learning_rate": 2.5725621922126518e-05, "loss": 0.0182, "step": 67830 }, { "grad_norm": 0.2654724717140198, "learning_rate": 2.571116789605722e-05, "loss": 0.0147, "step": 67840 }, { "grad_norm": 0.15891033411026, "learning_rate": 2.5696716526178054e-05, "loss": 0.0224, "step": 67850 }, { "grad_norm": 0.13389091193675995, "learning_rate": 2.568226781406946e-05, "loss": 0.0252, "step": 67860 }, { "grad_norm": 0.2551439702510834, "learning_rate": 2.5667821761311474e-05, "loss": 0.0162, "step": 67870 }, { "grad_norm": 0.17649972438812256, "learning_rate": 2.5653378369483917e-05, "loss": 0.0208, "step": 67880 }, { "grad_norm": 0.1590019315481186, "learning_rate": 2.5638937640166292e-05, "loss": 0.0135, "step": 67890 }, { "grad_norm": 0.1801488846540451, "learning_rate": 2.5624499574937842e-05, "loss": 0.0181, "step": 67900 }, { "grad_norm": 0.19771403074264526, "learning_rate": 2.5610064175377456e-05, "loss": 0.0196, "step": 67910 }, { "grad_norm": 0.16078898310661316, "learning_rate": 2.5595631443063782e-05, "loss": 0.0203, "step": 67920 }, { "grad_norm": 0.29305994510650635, "learning_rate": 2.5581201379575177e-05, "loss": 0.0183, "step": 67930 }, { "grad_norm": 0.3196980655193329, "learning_rate": 2.5566773986489655e-05, "loss": 0.0196, "step": 67940 }, { "grad_norm": 0.20100706815719604, "learning_rate": 2.5552349265385e-05, "loss": 0.034, "step": 67950 }, { "grad_norm": 0.28330567479133606, "learning_rate": 2.5537927217838675e-05, "loss": 0.0175, "step": 67960 }, { "grad_norm": 0.22807981073856354, "learning_rate": 2.5523507845427864e-05, "loss": 0.0167, "step": 67970 }, { "grad_norm": 0.32830318808555603, "learning_rate": 2.5509091149729413e-05, "loss": 0.0196, "step": 67980 }, { "grad_norm": 0.2330883890390396, "learning_rate": 2.5494677132319962e-05, "loss": 0.0193, "step": 67990 }, { "grad_norm": 0.24496901035308838, "learning_rate": 2.548026579477575e-05, "loss": 0.0303, "step": 68000 }, { "grad_norm": 0.20725040137767792, "learning_rate": 2.5465857138672834e-05, "loss": 0.0203, "step": 68010 }, { "grad_norm": 0.21538145840168, "learning_rate": 2.5451451165586848e-05, "loss": 0.0188, "step": 68020 }, { "grad_norm": 0.15950201451778412, "learning_rate": 2.543704787709329e-05, "loss": 0.0203, "step": 68030 }, { "grad_norm": 0.176493838429451, "learning_rate": 2.542264727476723e-05, "loss": 0.0185, "step": 68040 }, { "grad_norm": 0.18787768483161926, "learning_rate": 2.5408249360183535e-05, "loss": 0.0311, "step": 68050 }, { "grad_norm": 0.1993742734193802, "learning_rate": 2.5393854134916694e-05, "loss": 0.0188, "step": 68060 }, { "grad_norm": 0.10774111747741699, "learning_rate": 2.5379461600540993e-05, "loss": 0.0151, "step": 68070 }, { "grad_norm": 0.3513701260089874, "learning_rate": 2.5365071758630317e-05, "loss": 0.0176, "step": 68080 }, { "grad_norm": 0.23044688999652863, "learning_rate": 2.535068461075839e-05, "loss": 0.0155, "step": 68090 }, { "grad_norm": 0.4765866994857788, "learning_rate": 2.5336300158498516e-05, "loss": 0.0209, "step": 68100 }, { "grad_norm": 0.2300935685634613, "learning_rate": 2.5321918403423793e-05, "loss": 0.0146, "step": 68110 }, { "grad_norm": 0.2899009883403778, "learning_rate": 2.5307539347106957e-05, "loss": 0.0202, "step": 68120 }, { "grad_norm": 0.23643876612186432, "learning_rate": 2.529316299112049e-05, "loss": 0.0164, "step": 68130 }, { "grad_norm": 0.34221479296684265, "learning_rate": 2.5278789337036567e-05, "loss": 0.0149, "step": 68140 }, { "grad_norm": 0.30389079451560974, "learning_rate": 2.52644183864271e-05, "loss": 0.0176, "step": 68150 }, { "grad_norm": 0.4125787019729614, "learning_rate": 2.525005014086363e-05, "loss": 0.02, "step": 68160 }, { "grad_norm": 0.28479647636413574, "learning_rate": 2.5235684601917465e-05, "loss": 0.018, "step": 68170 }, { "grad_norm": 0.21479706466197968, "learning_rate": 2.5221321771159618e-05, "loss": 0.0147, "step": 68180 }, { "grad_norm": 0.20164625346660614, "learning_rate": 2.5206961650160743e-05, "loss": 0.0162, "step": 68190 }, { "grad_norm": 0.24400903284549713, "learning_rate": 2.519260424049129e-05, "loss": 0.0213, "step": 68200 }, { "grad_norm": 0.27233508229255676, "learning_rate": 2.5178249543721304e-05, "loss": 0.0236, "step": 68210 }, { "grad_norm": 0.24449345469474792, "learning_rate": 2.5163897561420668e-05, "loss": 0.0193, "step": 68220 }, { "grad_norm": 0.19610823690891266, "learning_rate": 2.514954829515883e-05, "loss": 0.0206, "step": 68230 }, { "grad_norm": 0.4322252869606018, "learning_rate": 2.513520174650505e-05, "loss": 0.0236, "step": 68240 }, { "grad_norm": 0.5746244788169861, "learning_rate": 2.5120857917028207e-05, "loss": 0.0185, "step": 68250 }, { "grad_norm": 0.4260275959968567, "learning_rate": 2.5106516808296953e-05, "loss": 0.0178, "step": 68260 }, { "grad_norm": 0.2449936419725418, "learning_rate": 2.5092178421879552e-05, "loss": 0.0207, "step": 68270 }, { "grad_norm": 0.26089709997177124, "learning_rate": 2.5077842759344116e-05, "loss": 0.0239, "step": 68280 }, { "grad_norm": 0.2723820209503174, "learning_rate": 2.5063509822258304e-05, "loss": 0.0211, "step": 68290 }, { "grad_norm": 0.1900734007358551, "learning_rate": 2.5049179612189576e-05, "loss": 0.0172, "step": 68300 }, { "grad_norm": 0.25409135222435, "learning_rate": 2.5034852130705043e-05, "loss": 0.0215, "step": 68310 }, { "grad_norm": 0.178033709526062, "learning_rate": 2.5020527379371572e-05, "loss": 0.0161, "step": 68320 }, { "grad_norm": 0.3384227454662323, "learning_rate": 2.500620535975564e-05, "loss": 0.022, "step": 68330 }, { "grad_norm": 0.17536917328834534, "learning_rate": 2.4991886073423515e-05, "loss": 0.0167, "step": 68340 }, { "grad_norm": 0.28878921270370483, "learning_rate": 2.4977569521941134e-05, "loss": 0.0168, "step": 68350 }, { "grad_norm": 0.28755810856819153, "learning_rate": 2.4963255706874143e-05, "loss": 0.0241, "step": 68360 }, { "grad_norm": 0.24624642729759216, "learning_rate": 2.4948944629787853e-05, "loss": 0.0172, "step": 68370 }, { "grad_norm": 0.2910829186439514, "learning_rate": 2.493463629224731e-05, "loss": 0.0192, "step": 68380 }, { "grad_norm": 0.3095255196094513, "learning_rate": 2.492033069581728e-05, "loss": 0.0212, "step": 68390 }, { "grad_norm": 0.3201926350593567, "learning_rate": 2.4906027842062137e-05, "loss": 0.0221, "step": 68400 }, { "grad_norm": 0.2143789827823639, "learning_rate": 2.48917277325461e-05, "loss": 0.0221, "step": 68410 }, { "grad_norm": 0.17148256301879883, "learning_rate": 2.487743036883295e-05, "loss": 0.0202, "step": 68420 }, { "grad_norm": 0.25618985295295715, "learning_rate": 2.486313575248626e-05, "loss": 0.0183, "step": 68430 }, { "grad_norm": 0.2448936104774475, "learning_rate": 2.4848843885069235e-05, "loss": 0.022, "step": 68440 }, { "grad_norm": 0.2508942782878876, "learning_rate": 2.483455476814484e-05, "loss": 0.0173, "step": 68450 }, { "grad_norm": 0.20487014949321747, "learning_rate": 2.4820268403275672e-05, "loss": 0.0204, "step": 68460 }, { "grad_norm": 0.20671692490577698, "learning_rate": 2.480598479202413e-05, "loss": 0.016, "step": 68470 }, { "grad_norm": 0.246091827750206, "learning_rate": 2.4791703935952193e-05, "loss": 0.0238, "step": 68480 }, { "grad_norm": 0.3845326602458954, "learning_rate": 2.477742583662163e-05, "loss": 0.0182, "step": 68490 }, { "grad_norm": 0.3459606468677521, "learning_rate": 2.476315049559384e-05, "loss": 0.0205, "step": 68500 }, { "grad_norm": 0.1755530834197998, "learning_rate": 2.4748877914429987e-05, "loss": 0.0211, "step": 68510 }, { "grad_norm": 0.40408411622047424, "learning_rate": 2.4734608094690857e-05, "loss": 0.0193, "step": 68520 }, { "grad_norm": 0.17686644196510315, "learning_rate": 2.4720341037937006e-05, "loss": 0.0246, "step": 68530 }, { "grad_norm": 0.34791216254234314, "learning_rate": 2.4706076745728645e-05, "loss": 0.0267, "step": 68540 }, { "grad_norm": 0.18867053091526031, "learning_rate": 2.469181521962573e-05, "loss": 0.0318, "step": 68550 }, { "grad_norm": 0.20503200590610504, "learning_rate": 2.467755646118783e-05, "loss": 0.0127, "step": 68560 }, { "grad_norm": 0.18127009272575378, "learning_rate": 2.46633004719743e-05, "loss": 0.0223, "step": 68570 }, { "grad_norm": 0.21452027559280396, "learning_rate": 2.4649047253544123e-05, "loss": 0.0161, "step": 68580 }, { "grad_norm": 0.28603395819664, "learning_rate": 2.4634796807456016e-05, "loss": 0.0201, "step": 68590 }, { "grad_norm": 0.35176944732666016, "learning_rate": 2.46205491352684e-05, "loss": 0.0197, "step": 68600 }, { "grad_norm": 0.15482720732688904, "learning_rate": 2.4606304238539375e-05, "loss": 0.0161, "step": 68610 }, { "grad_norm": 0.189406156539917, "learning_rate": 2.4592062118826753e-05, "loss": 0.015, "step": 68620 }, { "grad_norm": 0.16921567916870117, "learning_rate": 2.4577822777688004e-05, "loss": 0.0185, "step": 68630 }, { "grad_norm": 0.3464577794075012, "learning_rate": 2.4563586216680347e-05, "loss": 0.0225, "step": 68640 }, { "grad_norm": 0.21041041612625122, "learning_rate": 2.4549352437360622e-05, "loss": 0.0177, "step": 68650 }, { "grad_norm": 0.255193293094635, "learning_rate": 2.4535121441285493e-05, "loss": 0.0157, "step": 68660 }, { "grad_norm": 0.2033071368932724, "learning_rate": 2.4520893230011172e-05, "loss": 0.0144, "step": 68670 }, { "grad_norm": 0.21833747625350952, "learning_rate": 2.450666780509368e-05, "loss": 0.0218, "step": 68680 }, { "grad_norm": 0.20188646018505096, "learning_rate": 2.4492445168088656e-05, "loss": 0.0181, "step": 68690 }, { "grad_norm": 0.2804190218448639, "learning_rate": 2.447822532055149e-05, "loss": 0.0156, "step": 68700 }, { "grad_norm": 0.27191638946533203, "learning_rate": 2.4464008264037212e-05, "loss": 0.016, "step": 68710 }, { "grad_norm": 0.22941000759601593, "learning_rate": 2.4449794000100605e-05, "loss": 0.0163, "step": 68720 }, { "grad_norm": 0.2556372284889221, "learning_rate": 2.4435582530296108e-05, "loss": 0.0174, "step": 68730 }, { "grad_norm": 0.19731581211090088, "learning_rate": 2.4421373856177887e-05, "loss": 0.0171, "step": 68740 }, { "grad_norm": 0.1464844048023224, "learning_rate": 2.440716797929975e-05, "loss": 0.0146, "step": 68750 }, { "grad_norm": 0.18952372670173645, "learning_rate": 2.439296490121526e-05, "loss": 0.0187, "step": 68760 }, { "grad_norm": 0.20842567086219788, "learning_rate": 2.4378764623477608e-05, "loss": 0.0155, "step": 68770 }, { "grad_norm": 0.4110994040966034, "learning_rate": 2.436456714763974e-05, "loss": 0.0184, "step": 68780 }, { "grad_norm": 0.24442899227142334, "learning_rate": 2.4350372475254262e-05, "loss": 0.0189, "step": 68790 }, { "grad_norm": 0.14763736724853516, "learning_rate": 2.433618060787351e-05, "loss": 0.0174, "step": 68800 }, { "grad_norm": 0.23159532248973846, "learning_rate": 2.432199154704944e-05, "loss": 0.0259, "step": 68810 }, { "grad_norm": 0.28464818000793457, "learning_rate": 2.430780529433377e-05, "loss": 0.0152, "step": 68820 }, { "grad_norm": 0.1882958859205246, "learning_rate": 2.42936218512779e-05, "loss": 0.0261, "step": 68830 }, { "grad_norm": 0.3066869080066681, "learning_rate": 2.427944121943288e-05, "loss": 0.0141, "step": 68840 }, { "grad_norm": 0.23929400742053986, "learning_rate": 2.4265263400349497e-05, "loss": 0.0187, "step": 68850 }, { "grad_norm": 0.369282603263855, "learning_rate": 2.4251088395578213e-05, "loss": 0.0212, "step": 68860 }, { "grad_norm": 0.2839522659778595, "learning_rate": 2.4236916206669214e-05, "loss": 0.0159, "step": 68870 }, { "grad_norm": 0.3457190692424774, "learning_rate": 2.4222746835172304e-05, "loss": 0.0189, "step": 68880 }, { "grad_norm": 0.2503361403942108, "learning_rate": 2.4208580282637065e-05, "loss": 0.0193, "step": 68890 }, { "grad_norm": 0.5659804344177246, "learning_rate": 2.4194416550612687e-05, "loss": 0.0168, "step": 68900 }, { "grad_norm": 0.40680748224258423, "learning_rate": 2.418025564064812e-05, "loss": 0.0146, "step": 68910 }, { "grad_norm": 0.3809712529182434, "learning_rate": 2.4166097554291978e-05, "loss": 0.0193, "step": 68920 }, { "grad_norm": 0.3937820494174957, "learning_rate": 2.4151942293092583e-05, "loss": 0.0144, "step": 68930 }, { "grad_norm": 0.2717595398426056, "learning_rate": 2.41377898585979e-05, "loss": 0.0273, "step": 68940 }, { "grad_norm": 0.3521191477775574, "learning_rate": 2.4123640252355652e-05, "loss": 0.0217, "step": 68950 }, { "grad_norm": 0.17777501046657562, "learning_rate": 2.410949347591318e-05, "loss": 0.0196, "step": 68960 }, { "grad_norm": 0.23831912875175476, "learning_rate": 2.4095349530817578e-05, "loss": 0.0311, "step": 68970 }, { "grad_norm": 0.24857042729854584, "learning_rate": 2.4081208418615603e-05, "loss": 0.0179, "step": 68980 }, { "grad_norm": 0.3759708106517792, "learning_rate": 2.4067070140853725e-05, "loss": 0.02, "step": 68990 }, { "grad_norm": 0.2514500916004181, "learning_rate": 2.405293469907805e-05, "loss": 0.0174, "step": 69000 }, { "grad_norm": 0.18532010912895203, "learning_rate": 2.4038802094834446e-05, "loss": 0.0188, "step": 69010 }, { "grad_norm": 0.23150837421417236, "learning_rate": 2.4024672329668395e-05, "loss": 0.0144, "step": 69020 }, { "grad_norm": 0.27261143922805786, "learning_rate": 2.4010545405125125e-05, "loss": 0.0211, "step": 69030 }, { "grad_norm": 0.2772425711154938, "learning_rate": 2.3996421322749528e-05, "loss": 0.0163, "step": 69040 }, { "grad_norm": 0.28062328696250916, "learning_rate": 2.3982300084086224e-05, "loss": 0.0221, "step": 69050 }, { "grad_norm": 0.4020952880382538, "learning_rate": 2.396818169067944e-05, "loss": 0.0182, "step": 69060 }, { "grad_norm": 0.18303778767585754, "learning_rate": 2.3954066144073178e-05, "loss": 0.0169, "step": 69070 }, { "grad_norm": 0.2853838801383972, "learning_rate": 2.39399534458111e-05, "loss": 0.0198, "step": 69080 }, { "grad_norm": 0.22025248408317566, "learning_rate": 2.392584359743651e-05, "loss": 0.0193, "step": 69090 }, { "grad_norm": 0.26949647068977356, "learning_rate": 2.3911736600492463e-05, "loss": 0.0135, "step": 69100 }, { "grad_norm": 0.3528699576854706, "learning_rate": 2.3897632456521672e-05, "loss": 0.0203, "step": 69110 }, { "grad_norm": 0.1340494155883789, "learning_rate": 2.388353116706658e-05, "loss": 0.0222, "step": 69120 }, { "grad_norm": 0.19131629168987274, "learning_rate": 2.3869432733669232e-05, "loss": 0.0147, "step": 69130 }, { "grad_norm": 0.40674903988838196, "learning_rate": 2.385533715787145e-05, "loss": 0.018, "step": 69140 }, { "grad_norm": 0.2329266518354416, "learning_rate": 2.3841244441214673e-05, "loss": 0.0207, "step": 69150 }, { "grad_norm": 0.2721647024154663, "learning_rate": 2.3827154585240073e-05, "loss": 0.0174, "step": 69160 }, { "grad_norm": 0.18096618354320526, "learning_rate": 2.3813067591488498e-05, "loss": 0.0199, "step": 69170 }, { "grad_norm": 0.20788678526878357, "learning_rate": 2.3798983461500494e-05, "loss": 0.0187, "step": 69180 }, { "grad_norm": 0.21923261880874634, "learning_rate": 2.3784902196816256e-05, "loss": 0.0179, "step": 69190 }, { "grad_norm": 0.2743329405784607, "learning_rate": 2.3770823798975716e-05, "loss": 0.0174, "step": 69200 }, { "grad_norm": 0.1868203580379486, "learning_rate": 2.3756748269518424e-05, "loss": 0.016, "step": 69210 }, { "grad_norm": 0.23652434349060059, "learning_rate": 2.374267560998371e-05, "loss": 0.0248, "step": 69220 }, { "grad_norm": 0.23688043653964996, "learning_rate": 2.3728605821910477e-05, "loss": 0.0138, "step": 69230 }, { "grad_norm": 0.38375726342201233, "learning_rate": 2.3714538906837453e-05, "loss": 0.0187, "step": 69240 }, { "grad_norm": 0.11895240843296051, "learning_rate": 2.3700474866302913e-05, "loss": 0.0157, "step": 69250 }, { "grad_norm": 0.22916825115680695, "learning_rate": 2.3686413701844917e-05, "loss": 0.019, "step": 69260 }, { "grad_norm": 0.28669750690460205, "learning_rate": 2.367235541500114e-05, "loss": 0.0157, "step": 69270 }, { "grad_norm": 0.29254603385925293, "learning_rate": 2.3658300007308993e-05, "loss": 0.0196, "step": 69280 }, { "grad_norm": 0.44751647114753723, "learning_rate": 2.364424748030555e-05, "loss": 0.0232, "step": 69290 }, { "grad_norm": 0.19176647067070007, "learning_rate": 2.3630197835527595e-05, "loss": 0.0146, "step": 69300 }, { "grad_norm": 0.26089560985565186, "learning_rate": 2.3616151074511546e-05, "loss": 0.0183, "step": 69310 }, { "grad_norm": 0.2382216602563858, "learning_rate": 2.3602107198793545e-05, "loss": 0.0183, "step": 69320 }, { "grad_norm": 0.20216822624206543, "learning_rate": 2.3588066209909427e-05, "loss": 0.0194, "step": 69330 }, { "grad_norm": 0.22298137843608856, "learning_rate": 2.3574028109394662e-05, "loss": 0.0206, "step": 69340 }, { "grad_norm": 0.20615537464618683, "learning_rate": 2.3559992898784454e-05, "loss": 0.0189, "step": 69350 }, { "grad_norm": 0.2413080781698227, "learning_rate": 2.354596057961367e-05, "loss": 0.0185, "step": 69360 }, { "grad_norm": 0.2669556438922882, "learning_rate": 2.353193115341688e-05, "loss": 0.0262, "step": 69370 }, { "grad_norm": 0.23033049702644348, "learning_rate": 2.3517904621728294e-05, "loss": 0.0215, "step": 69380 }, { "grad_norm": 0.16204231977462769, "learning_rate": 2.3503880986081855e-05, "loss": 0.0199, "step": 69390 }, { "grad_norm": 0.2229299247264862, "learning_rate": 2.3489860248011136e-05, "loss": 0.0187, "step": 69400 }, { "grad_norm": 0.20980475842952728, "learning_rate": 2.3475842409049464e-05, "loss": 0.0176, "step": 69410 }, { "grad_norm": 0.19694723188877106, "learning_rate": 2.346182747072975e-05, "loss": 0.0258, "step": 69420 }, { "grad_norm": 0.21564620733261108, "learning_rate": 2.3447815434584718e-05, "loss": 0.0182, "step": 69430 }, { "grad_norm": 0.3567235767841339, "learning_rate": 2.3433806302146656e-05, "loss": 0.0159, "step": 69440 }, { "grad_norm": 0.3633061349391937, "learning_rate": 2.3419800074947612e-05, "loss": 0.018, "step": 69450 }, { "grad_norm": 0.42341551184654236, "learning_rate": 2.3405796754519248e-05, "loss": 0.023, "step": 69460 }, { "grad_norm": 0.42066556215286255, "learning_rate": 2.3391796342392992e-05, "loss": 0.0161, "step": 69470 }, { "grad_norm": 0.21795228123664856, "learning_rate": 2.337779884009984e-05, "loss": 0.0142, "step": 69480 }, { "grad_norm": 0.26838743686676025, "learning_rate": 2.3363804249170624e-05, "loss": 0.0167, "step": 69490 }, { "grad_norm": 0.29986464977264404, "learning_rate": 2.3349812571135703e-05, "loss": 0.0186, "step": 69500 }, { "grad_norm": 0.16754311323165894, "learning_rate": 2.333582380752523e-05, "loss": 0.0215, "step": 69510 }, { "grad_norm": 0.3315275013446808, "learning_rate": 2.3321837959868963e-05, "loss": 0.0168, "step": 69520 }, { "grad_norm": 0.22594420611858368, "learning_rate": 2.3307855029696383e-05, "loss": 0.0137, "step": 69530 }, { "grad_norm": 0.17924006283283234, "learning_rate": 2.3293875018536648e-05, "loss": 0.0149, "step": 69540 }, { "grad_norm": 0.14466962218284607, "learning_rate": 2.327989792791861e-05, "loss": 0.0147, "step": 69550 }, { "grad_norm": 0.2462613880634308, "learning_rate": 2.326592375937074e-05, "loss": 0.0187, "step": 69560 }, { "grad_norm": 0.13159361481666565, "learning_rate": 2.3251952514421248e-05, "loss": 0.0163, "step": 69570 }, { "grad_norm": 0.3466153144836426, "learning_rate": 2.323798419459804e-05, "loss": 0.0179, "step": 69580 }, { "grad_norm": 0.6004816889762878, "learning_rate": 2.3224018801428625e-05, "loss": 0.0203, "step": 69590 }, { "grad_norm": 0.2819785177707672, "learning_rate": 2.3210056336440273e-05, "loss": 0.0176, "step": 69600 }, { "grad_norm": 0.22141700983047485, "learning_rate": 2.3196096801159844e-05, "loss": 0.013, "step": 69610 }, { "grad_norm": 1.083906650543213, "learning_rate": 2.3182140197114012e-05, "loss": 0.0202, "step": 69620 }, { "grad_norm": 0.13668963313102722, "learning_rate": 2.3168186525828985e-05, "loss": 0.021, "step": 69630 }, { "grad_norm": 0.1903681606054306, "learning_rate": 2.3154235788830757e-05, "loss": 0.0212, "step": 69640 }, { "grad_norm": 0.28707608580589294, "learning_rate": 2.3140287987644927e-05, "loss": 0.0197, "step": 69650 }, { "grad_norm": 0.41378381848335266, "learning_rate": 2.3126343123796833e-05, "loss": 0.0185, "step": 69660 }, { "grad_norm": 0.21462886035442352, "learning_rate": 2.3112401198811418e-05, "loss": 0.0168, "step": 69670 }, { "grad_norm": 0.3156585693359375, "learning_rate": 2.3098462214213418e-05, "loss": 0.0182, "step": 69680 }, { "grad_norm": 0.17970722913742065, "learning_rate": 2.308452617152712e-05, "loss": 0.0207, "step": 69690 }, { "grad_norm": 0.5914601683616638, "learning_rate": 2.3070593072276603e-05, "loss": 0.0191, "step": 69700 }, { "grad_norm": 0.22893640398979187, "learning_rate": 2.3056662917985518e-05, "loss": 0.0146, "step": 69710 }, { "grad_norm": 0.5603847503662109, "learning_rate": 2.3042735710177283e-05, "loss": 0.0281, "step": 69720 }, { "grad_norm": 0.3232627809047699, "learning_rate": 2.302881145037492e-05, "loss": 0.0199, "step": 69730 }, { "grad_norm": 0.30098703503608704, "learning_rate": 2.3014890140101192e-05, "loss": 0.0192, "step": 69740 }, { "grad_norm": 0.3138960301876068, "learning_rate": 2.30009717808785e-05, "loss": 0.0221, "step": 69750 }, { "grad_norm": 0.23317237198352814, "learning_rate": 2.2987056374228967e-05, "loss": 0.0169, "step": 69760 }, { "grad_norm": 0.2575100064277649, "learning_rate": 2.297314392167431e-05, "loss": 0.0173, "step": 69770 }, { "grad_norm": 0.23983950912952423, "learning_rate": 2.295923442473601e-05, "loss": 0.0173, "step": 69780 }, { "grad_norm": 0.2126821130514145, "learning_rate": 2.294532788493519e-05, "loss": 0.0149, "step": 69790 }, { "grad_norm": 0.18048547208309174, "learning_rate": 2.293142430379262e-05, "loss": 0.0189, "step": 69800 }, { "grad_norm": 0.44592535495758057, "learning_rate": 2.291752368282879e-05, "loss": 0.0196, "step": 69810 }, { "grad_norm": 2.2472026348114014, "learning_rate": 2.290362602356385e-05, "loss": 0.0208, "step": 69820 }, { "grad_norm": 0.18628054857254028, "learning_rate": 2.288973132751765e-05, "loss": 0.0146, "step": 69830 }, { "grad_norm": 0.46472474932670593, "learning_rate": 2.287583959620965e-05, "loss": 0.0246, "step": 69840 }, { "grad_norm": 0.3123161196708679, "learning_rate": 2.2861950831159074e-05, "loss": 0.0165, "step": 69850 }, { "grad_norm": 0.23134368658065796, "learning_rate": 2.284806503388472e-05, "loss": 0.0219, "step": 69860 }, { "grad_norm": 0.22577661275863647, "learning_rate": 2.283418220590518e-05, "loss": 0.0154, "step": 69870 }, { "grad_norm": 0.30645403265953064, "learning_rate": 2.282030234873862e-05, "loss": 0.023, "step": 69880 }, { "grad_norm": 0.17185264825820923, "learning_rate": 2.2806425463902948e-05, "loss": 0.0181, "step": 69890 }, { "grad_norm": 0.32807403802871704, "learning_rate": 2.279255155291568e-05, "loss": 0.0219, "step": 69900 }, { "grad_norm": 0.14655718207359314, "learning_rate": 2.277868061729409e-05, "loss": 0.0184, "step": 69910 }, { "grad_norm": 0.21635526418685913, "learning_rate": 2.276481265855504e-05, "loss": 0.0217, "step": 69920 }, { "grad_norm": 0.3158171772956848, "learning_rate": 2.2750947678215128e-05, "loss": 0.0221, "step": 69930 }, { "grad_norm": 0.19861392676830292, "learning_rate": 2.2737085677790616e-05, "loss": 0.0159, "step": 69940 }, { "grad_norm": 0.23629721999168396, "learning_rate": 2.2723226658797443e-05, "loss": 0.0163, "step": 69950 }, { "grad_norm": 0.46652403473854065, "learning_rate": 2.270937062275117e-05, "loss": 0.0192, "step": 69960 }, { "grad_norm": 0.19416971504688263, "learning_rate": 2.2695517571167113e-05, "loss": 0.017, "step": 69970 }, { "grad_norm": 0.1401597112417221, "learning_rate": 2.268166750556019e-05, "loss": 0.0127, "step": 69980 }, { "grad_norm": 0.279329389333725, "learning_rate": 2.266782042744503e-05, "loss": 0.0171, "step": 69990 }, { "grad_norm": 0.4339011013507843, "learning_rate": 2.2653976338335935e-05, "loss": 0.0186, "step": 70000 }, { "grad_norm": 0.6974875330924988, "learning_rate": 2.2640135239746895e-05, "loss": 0.0221, "step": 70010 }, { "grad_norm": 0.23138515651226044, "learning_rate": 2.2626297133191515e-05, "loss": 0.018, "step": 70020 }, { "grad_norm": 0.18936395645141602, "learning_rate": 2.261246202018312e-05, "loss": 0.015, "step": 70030 }, { "grad_norm": 0.1818755865097046, "learning_rate": 2.2598629902234718e-05, "loss": 0.0214, "step": 70040 }, { "grad_norm": 0.5618298053741455, "learning_rate": 2.258480078085894e-05, "loss": 0.0181, "step": 70050 }, { "grad_norm": 0.20304203033447266, "learning_rate": 2.2570974657568126e-05, "loss": 0.0173, "step": 70060 }, { "grad_norm": 0.253761351108551, "learning_rate": 2.2557151533874286e-05, "loss": 0.0161, "step": 70070 }, { "grad_norm": 0.1899767965078354, "learning_rate": 2.2543331411289116e-05, "loss": 0.0209, "step": 70080 }, { "grad_norm": 0.26864510774612427, "learning_rate": 2.252951429132392e-05, "loss": 0.0249, "step": 70090 }, { "grad_norm": 0.2550142705440521, "learning_rate": 2.2515700175489762e-05, "loss": 0.0202, "step": 70100 }, { "grad_norm": 0.265658438205719, "learning_rate": 2.2501889065297288e-05, "loss": 0.0177, "step": 70110 }, { "grad_norm": 0.24171708524227142, "learning_rate": 2.2488080962256887e-05, "loss": 0.0188, "step": 70120 }, { "grad_norm": 0.24980151653289795, "learning_rate": 2.2474275867878585e-05, "loss": 0.0242, "step": 70130 }, { "grad_norm": 0.24209798872470856, "learning_rate": 2.246047378367211e-05, "loss": 0.0288, "step": 70140 }, { "grad_norm": 0.2909429669380188, "learning_rate": 2.2446674711146798e-05, "loss": 0.02, "step": 70150 }, { "grad_norm": 0.4655148684978485, "learning_rate": 2.2432878651811734e-05, "loss": 0.0207, "step": 70160 }, { "grad_norm": 1.4992413520812988, "learning_rate": 2.2419085607175594e-05, "loss": 0.0254, "step": 70170 }, { "grad_norm": 0.19679073989391327, "learning_rate": 2.240529557874678e-05, "loss": 0.0257, "step": 70180 }, { "grad_norm": 0.3132782280445099, "learning_rate": 2.2391508568033358e-05, "loss": 0.0167, "step": 70190 }, { "grad_norm": 0.45885682106018066, "learning_rate": 2.237772457654307e-05, "loss": 0.0183, "step": 70200 }, { "grad_norm": 0.1302216351032257, "learning_rate": 2.236394360578327e-05, "loss": 0.0162, "step": 70210 }, { "grad_norm": 0.19722193479537964, "learning_rate": 2.2350165657261073e-05, "loss": 0.0195, "step": 70220 }, { "grad_norm": 0.2823536992073059, "learning_rate": 2.2336390732483164e-05, "loss": 0.0202, "step": 70230 }, { "grad_norm": 0.49584195017814636, "learning_rate": 2.2322618832955976e-05, "loss": 0.0183, "step": 70240 }, { "grad_norm": 0.21084439754486084, "learning_rate": 2.2308849960185578e-05, "loss": 0.0169, "step": 70250 }, { "grad_norm": 0.32252874970436096, "learning_rate": 2.229508411567774e-05, "loss": 0.0194, "step": 70260 }, { "grad_norm": 0.4217907786369324, "learning_rate": 2.2281321300937823e-05, "loss": 0.0207, "step": 70270 }, { "grad_norm": 0.3420788645744324, "learning_rate": 2.2267561517470937e-05, "loss": 0.0168, "step": 70280 }, { "grad_norm": 0.24151892960071564, "learning_rate": 2.2253804766781845e-05, "loss": 0.0209, "step": 70290 }, { "grad_norm": 0.25275230407714844, "learning_rate": 2.2240051050374933e-05, "loss": 0.0161, "step": 70300 }, { "grad_norm": 0.4787880480289459, "learning_rate": 2.22263003697543e-05, "loss": 0.0202, "step": 70310 }, { "grad_norm": 0.651287317276001, "learning_rate": 2.2212552726423702e-05, "loss": 0.0268, "step": 70320 }, { "grad_norm": 0.22738853096961975, "learning_rate": 2.2198808121886578e-05, "loss": 0.0144, "step": 70330 }, { "grad_norm": 0.3122963607311249, "learning_rate": 2.218506655764598e-05, "loss": 0.0179, "step": 70340 }, { "grad_norm": 0.19923678040504456, "learning_rate": 2.2171328035204704e-05, "loss": 0.0154, "step": 70350 }, { "grad_norm": 0.23758041858673096, "learning_rate": 2.2157592556065133e-05, "loss": 0.0168, "step": 70360 }, { "grad_norm": 0.16201163828372955, "learning_rate": 2.2143860121729382e-05, "loss": 0.0164, "step": 70370 }, { "grad_norm": 0.31110304594039917, "learning_rate": 2.2130130733699206e-05, "loss": 0.0157, "step": 70380 }, { "grad_norm": 0.5744664669036865, "learning_rate": 2.2116404393476052e-05, "loss": 0.0165, "step": 70390 }, { "grad_norm": 0.3294171392917633, "learning_rate": 2.2102681102560967e-05, "loss": 0.015, "step": 70400 }, { "grad_norm": 0.2711206376552582, "learning_rate": 2.208896086245476e-05, "loss": 0.0186, "step": 70410 }, { "grad_norm": 0.4701823592185974, "learning_rate": 2.2075243674657808e-05, "loss": 0.0182, "step": 70420 }, { "grad_norm": 0.19722704589366913, "learning_rate": 2.206152954067024e-05, "loss": 0.0142, "step": 70430 }, { "grad_norm": 0.22759591042995453, "learning_rate": 2.2047818461991758e-05, "loss": 0.0231, "step": 70440 }, { "grad_norm": 0.1576961874961853, "learning_rate": 2.2034110440121864e-05, "loss": 0.0176, "step": 70450 }, { "grad_norm": 0.22813287377357483, "learning_rate": 2.2020405476559585e-05, "loss": 0.021, "step": 70460 }, { "grad_norm": 0.26119163632392883, "learning_rate": 2.2006703572803722e-05, "loss": 0.0195, "step": 70470 }, { "grad_norm": 0.21752923727035522, "learning_rate": 2.1993004730352646e-05, "loss": 0.0194, "step": 70480 }, { "grad_norm": 0.19220805168151855, "learning_rate": 2.1979308950704464e-05, "loss": 0.016, "step": 70490 }, { "grad_norm": 0.2594376504421234, "learning_rate": 2.1965616235356923e-05, "loss": 0.0216, "step": 70500 }, { "grad_norm": 0.2365005612373352, "learning_rate": 2.1951926585807465e-05, "loss": 0.0152, "step": 70510 }, { "grad_norm": 0.3254340589046478, "learning_rate": 2.193824000355313e-05, "loss": 0.0171, "step": 70520 }, { "grad_norm": 0.2019214779138565, "learning_rate": 2.192455649009067e-05, "loss": 0.0143, "step": 70530 }, { "grad_norm": 0.22569626569747925, "learning_rate": 2.1910876046916523e-05, "loss": 0.0193, "step": 70540 }, { "grad_norm": 0.17999953031539917, "learning_rate": 2.1897198675526725e-05, "loss": 0.0167, "step": 70550 }, { "grad_norm": 0.2495022416114807, "learning_rate": 2.1883524377417024e-05, "loss": 0.018, "step": 70560 }, { "grad_norm": 0.2328701913356781, "learning_rate": 2.1869853154082825e-05, "loss": 0.0144, "step": 70570 }, { "grad_norm": 0.13788816332817078, "learning_rate": 2.1856185007019215e-05, "loss": 0.0168, "step": 70580 }, { "grad_norm": 0.19364632666110992, "learning_rate": 2.1842519937720874e-05, "loss": 0.0185, "step": 70590 }, { "grad_norm": 0.17374272644519806, "learning_rate": 2.182885794768224e-05, "loss": 0.0172, "step": 70600 }, { "grad_norm": 0.3336679935455322, "learning_rate": 2.1815199038397326e-05, "loss": 0.0166, "step": 70610 }, { "grad_norm": 0.26130908727645874, "learning_rate": 2.1801543211359893e-05, "loss": 0.0171, "step": 70620 }, { "grad_norm": 0.13041317462921143, "learning_rate": 2.178789046806326e-05, "loss": 0.0167, "step": 70630 }, { "grad_norm": 0.21123264729976654, "learning_rate": 2.1774240810000547e-05, "loss": 0.0263, "step": 70640 }, { "grad_norm": 0.23943021893501282, "learning_rate": 2.17605942386644e-05, "loss": 0.0162, "step": 70650 }, { "grad_norm": 0.18175242841243744, "learning_rate": 2.174695075554723e-05, "loss": 0.0213, "step": 70660 }, { "grad_norm": 0.3295077681541443, "learning_rate": 2.173331036214103e-05, "loss": 0.0269, "step": 70670 }, { "grad_norm": 0.2437981367111206, "learning_rate": 2.171967305993753e-05, "loss": 0.0128, "step": 70680 }, { "grad_norm": 0.1378680020570755, "learning_rate": 2.1706038850428024e-05, "loss": 0.0155, "step": 70690 }, { "grad_norm": 0.1864544153213501, "learning_rate": 2.169240773510361e-05, "loss": 0.0147, "step": 70700 }, { "grad_norm": 0.29318588972091675, "learning_rate": 2.1678779715454906e-05, "loss": 0.0194, "step": 70710 }, { "grad_norm": 0.27701956033706665, "learning_rate": 2.166515479297229e-05, "loss": 0.0213, "step": 70720 }, { "grad_norm": 0.19331514835357666, "learning_rate": 2.1651532969145732e-05, "loss": 0.0141, "step": 70730 }, { "grad_norm": 0.1800042688846588, "learning_rate": 2.1637914245464903e-05, "loss": 0.0178, "step": 70740 }, { "grad_norm": 0.19346120953559875, "learning_rate": 2.1624298623419133e-05, "loss": 0.0153, "step": 70750 }, { "grad_norm": 0.19596248865127563, "learning_rate": 2.1610686104497413e-05, "loss": 0.0159, "step": 70760 }, { "grad_norm": 0.35985323786735535, "learning_rate": 2.1597076690188366e-05, "loss": 0.0256, "step": 70770 }, { "grad_norm": 0.2560305893421173, "learning_rate": 2.1583470381980315e-05, "loss": 0.0137, "step": 70780 }, { "grad_norm": 0.3592453896999359, "learning_rate": 2.156986718136123e-05, "loss": 0.0277, "step": 70790 }, { "grad_norm": 0.3393177092075348, "learning_rate": 2.155626708981871e-05, "loss": 0.017, "step": 70800 }, { "grad_norm": 0.2787415683269501, "learning_rate": 2.1542670108840075e-05, "loss": 0.0181, "step": 70810 }, { "grad_norm": 0.3056449592113495, "learning_rate": 2.152907623991221e-05, "loss": 0.0177, "step": 70820 }, { "grad_norm": 0.2089085578918457, "learning_rate": 2.1515485484521802e-05, "loss": 0.0151, "step": 70830 }, { "grad_norm": 0.23481805622577667, "learning_rate": 2.150189784415506e-05, "loss": 0.0214, "step": 70840 }, { "grad_norm": 0.17195887863636017, "learning_rate": 2.148831332029794e-05, "loss": 0.0168, "step": 70850 }, { "grad_norm": 0.2432098388671875, "learning_rate": 2.147473191443599e-05, "loss": 0.0185, "step": 70860 }, { "grad_norm": 0.2063840627670288, "learning_rate": 2.1461153628054492e-05, "loss": 0.0203, "step": 70870 }, { "grad_norm": 0.25171491503715515, "learning_rate": 2.1447578462638284e-05, "loss": 0.0156, "step": 70880 }, { "grad_norm": 0.29730716347694397, "learning_rate": 2.143400641967201e-05, "loss": 0.0154, "step": 70890 }, { "grad_norm": 0.43304938077926636, "learning_rate": 2.1420437500639823e-05, "loss": 0.0155, "step": 70900 }, { "grad_norm": 0.25442320108413696, "learning_rate": 2.1406871707025646e-05, "loss": 0.0195, "step": 70910 }, { "grad_norm": 0.22274057567119598, "learning_rate": 2.1393309040312963e-05, "loss": 0.0199, "step": 70920 }, { "grad_norm": 0.3107645511627197, "learning_rate": 2.137974950198502e-05, "loss": 0.0184, "step": 70930 }, { "grad_norm": 0.3936622440814972, "learning_rate": 2.1366193093524617e-05, "loss": 0.0153, "step": 70940 }, { "grad_norm": 0.17309191823005676, "learning_rate": 2.1352639816414288e-05, "loss": 0.0216, "step": 70950 }, { "grad_norm": 0.25377053022384644, "learning_rate": 2.1339089672136202e-05, "loss": 0.0172, "step": 70960 }, { "grad_norm": 0.21088053286075592, "learning_rate": 2.1325542662172193e-05, "loss": 0.0249, "step": 70970 }, { "grad_norm": 0.2117006629705429, "learning_rate": 2.1311998788003712e-05, "loss": 0.0217, "step": 70980 }, { "grad_norm": 0.3016246557235718, "learning_rate": 2.1298458051111907e-05, "loss": 0.0239, "step": 70990 }, { "grad_norm": 0.143482968211174, "learning_rate": 2.1284920452977602e-05, "loss": 0.0177, "step": 71000 }, { "grad_norm": 0.14656397700309753, "learning_rate": 2.12713859950812e-05, "loss": 0.0106, "step": 71010 }, { "grad_norm": 0.20328164100646973, "learning_rate": 2.1257854678902832e-05, "loss": 0.0183, "step": 71020 }, { "grad_norm": 0.2753831446170807, "learning_rate": 2.124432650592227e-05, "loss": 0.0167, "step": 71030 }, { "grad_norm": 0.29180437326431274, "learning_rate": 2.1230801477618937e-05, "loss": 0.0218, "step": 71040 }, { "grad_norm": 0.34569528698921204, "learning_rate": 2.121727959547189e-05, "loss": 0.0175, "step": 71050 }, { "grad_norm": 0.3412224352359772, "learning_rate": 2.1203760860959886e-05, "loss": 0.0194, "step": 71060 }, { "grad_norm": 0.23454003036022186, "learning_rate": 2.119024527556127e-05, "loss": 0.0181, "step": 71070 }, { "grad_norm": 0.19778208434581757, "learning_rate": 2.117673284075415e-05, "loss": 0.0172, "step": 71080 }, { "grad_norm": 0.14695709943771362, "learning_rate": 2.1163223558016165e-05, "loss": 0.0131, "step": 71090 }, { "grad_norm": 0.2146514356136322, "learning_rate": 2.1149717428824722e-05, "loss": 0.0185, "step": 71100 }, { "grad_norm": 0.6208985447883606, "learning_rate": 2.1136214454656784e-05, "loss": 0.0255, "step": 71110 }, { "grad_norm": 0.31833139061927795, "learning_rate": 2.1122714636989056e-05, "loss": 0.0292, "step": 71120 }, { "grad_norm": 0.18516333401203156, "learning_rate": 2.1109217977297823e-05, "loss": 0.0211, "step": 71130 }, { "grad_norm": 0.20923300087451935, "learning_rate": 2.1095724477059077e-05, "loss": 0.014, "step": 71140 }, { "grad_norm": 0.3649798035621643, "learning_rate": 2.1082234137748434e-05, "loss": 0.0157, "step": 71150 }, { "grad_norm": 0.17092132568359375, "learning_rate": 2.106874696084122e-05, "loss": 0.0177, "step": 71160 }, { "grad_norm": 0.1991136223077774, "learning_rate": 2.1055262947812315e-05, "loss": 0.0198, "step": 71170 }, { "grad_norm": 0.41199198365211487, "learning_rate": 2.104178210013636e-05, "loss": 0.0282, "step": 71180 }, { "grad_norm": 0.5393324494361877, "learning_rate": 2.1028304419287554e-05, "loss": 0.0247, "step": 71190 }, { "grad_norm": 0.2013266533613205, "learning_rate": 2.101482990673982e-05, "loss": 0.018, "step": 71200 }, { "grad_norm": 0.2672332525253296, "learning_rate": 2.1001358563966707e-05, "loss": 0.0196, "step": 71210 }, { "grad_norm": 0.1126801073551178, "learning_rate": 2.098789039244144e-05, "loss": 0.0167, "step": 71220 }, { "grad_norm": 0.30806106328964233, "learning_rate": 2.097442539363685e-05, "loss": 0.0186, "step": 71230 }, { "grad_norm": 0.21332237124443054, "learning_rate": 2.0960963569025448e-05, "loss": 0.0179, "step": 71240 }, { "grad_norm": 0.2629149854183197, "learning_rate": 2.0947504920079435e-05, "loss": 0.0199, "step": 71250 }, { "grad_norm": 0.14758223295211792, "learning_rate": 2.093404944827059e-05, "loss": 0.0155, "step": 71260 }, { "grad_norm": 0.3969325125217438, "learning_rate": 2.0920597155070394e-05, "loss": 0.0161, "step": 71270 }, { "grad_norm": 0.22564609348773956, "learning_rate": 2.090714804194997e-05, "loss": 0.0221, "step": 71280 }, { "grad_norm": 0.22426243126392365, "learning_rate": 2.0893702110380114e-05, "loss": 0.0108, "step": 71290 }, { "grad_norm": 0.18253745138645172, "learning_rate": 2.0880259361831216e-05, "loss": 0.0148, "step": 71300 }, { "grad_norm": 0.18302352726459503, "learning_rate": 2.0866819797773397e-05, "loss": 0.0205, "step": 71310 }, { "grad_norm": 0.20379318296909332, "learning_rate": 2.085338341967634e-05, "loss": 0.0191, "step": 71320 }, { "grad_norm": 0.11333419382572174, "learning_rate": 2.083995022900946e-05, "loss": 0.019, "step": 71330 }, { "grad_norm": 0.4990660846233368, "learning_rate": 2.0826520227241768e-05, "loss": 0.0212, "step": 71340 }, { "grad_norm": 0.25106486678123474, "learning_rate": 2.081309341584199e-05, "loss": 0.0163, "step": 71350 }, { "grad_norm": 0.28996121883392334, "learning_rate": 2.0799669796278413e-05, "loss": 0.0188, "step": 71360 }, { "grad_norm": 0.23155182600021362, "learning_rate": 2.078624937001906e-05, "loss": 0.0175, "step": 71370 }, { "grad_norm": 0.3911094069480896, "learning_rate": 2.0772832138531544e-05, "loss": 0.0307, "step": 71380 }, { "grad_norm": 0.23028025031089783, "learning_rate": 2.0759418103283158e-05, "loss": 0.0179, "step": 71390 }, { "grad_norm": 0.18006587028503418, "learning_rate": 2.0746007265740846e-05, "loss": 0.0177, "step": 71400 }, { "grad_norm": 0.22488681972026825, "learning_rate": 2.0732599627371215e-05, "loss": 0.0184, "step": 71410 }, { "grad_norm": 0.38212332129478455, "learning_rate": 2.0719195189640468e-05, "loss": 0.0165, "step": 71420 }, { "grad_norm": 0.1659853458404541, "learning_rate": 2.0705793954014528e-05, "loss": 0.0159, "step": 71430 }, { "grad_norm": 0.24423089623451233, "learning_rate": 2.06923959219589e-05, "loss": 0.0172, "step": 71440 }, { "grad_norm": 0.6025957465171814, "learning_rate": 2.0679001094938793e-05, "loss": 0.0159, "step": 71450 }, { "grad_norm": 0.21194471418857574, "learning_rate": 2.066560947441904e-05, "loss": 0.0175, "step": 71460 }, { "grad_norm": 0.23476983606815338, "learning_rate": 2.0652221061864152e-05, "loss": 0.0161, "step": 71470 }, { "grad_norm": 0.23079833388328552, "learning_rate": 2.0638835858738226e-05, "loss": 0.0157, "step": 71480 }, { "grad_norm": 0.2903236150741577, "learning_rate": 2.0625453866505068e-05, "loss": 0.0151, "step": 71490 }, { "grad_norm": 0.12544097006320953, "learning_rate": 2.0612075086628123e-05, "loss": 0.0159, "step": 71500 }, { "grad_norm": 0.316936731338501, "learning_rate": 2.059869952057045e-05, "loss": 0.0163, "step": 71510 }, { "grad_norm": 0.28982216119766235, "learning_rate": 2.0585327169794793e-05, "loss": 0.0204, "step": 71520 }, { "grad_norm": 0.23202082514762878, "learning_rate": 2.0571958035763534e-05, "loss": 0.0172, "step": 71530 }, { "grad_norm": 0.17984573543071747, "learning_rate": 2.0558592119938715e-05, "loss": 0.0144, "step": 71540 }, { "grad_norm": 0.2108495831489563, "learning_rate": 2.054522942378198e-05, "loss": 0.0168, "step": 71550 }, { "grad_norm": 0.1750689148902893, "learning_rate": 2.053186994875469e-05, "loss": 0.0146, "step": 71560 }, { "grad_norm": 0.20016808807849884, "learning_rate": 2.0518513696317775e-05, "loss": 0.0168, "step": 71570 }, { "grad_norm": 0.24894538521766663, "learning_rate": 2.050516066793188e-05, "loss": 0.0208, "step": 71580 }, { "grad_norm": 0.23675848543643951, "learning_rate": 2.0491810865057276e-05, "loss": 0.018, "step": 71590 }, { "grad_norm": 0.16093172132968903, "learning_rate": 2.0478464289153882e-05, "loss": 0.0164, "step": 71600 }, { "grad_norm": 0.16690973937511444, "learning_rate": 2.0465120941681232e-05, "loss": 0.0139, "step": 71610 }, { "grad_norm": 0.3854069709777832, "learning_rate": 2.0451780824098566e-05, "loss": 0.0189, "step": 71620 }, { "grad_norm": 0.20790943503379822, "learning_rate": 2.043844393786471e-05, "loss": 0.015, "step": 71630 }, { "grad_norm": 0.28330540657043457, "learning_rate": 2.0425110284438193e-05, "loss": 0.019, "step": 71640 }, { "grad_norm": 0.2110934853553772, "learning_rate": 2.041177986527712e-05, "loss": 0.0163, "step": 71650 }, { "grad_norm": 0.30604350566864014, "learning_rate": 2.0398452681839347e-05, "loss": 0.0124, "step": 71660 }, { "grad_norm": 0.2436242252588272, "learning_rate": 2.0385128735582272e-05, "loss": 0.0289, "step": 71670 }, { "grad_norm": 0.22784975171089172, "learning_rate": 2.0371808027963003e-05, "loss": 0.0168, "step": 71680 }, { "grad_norm": 0.3401505947113037, "learning_rate": 2.035849056043825e-05, "loss": 0.0245, "step": 71690 }, { "grad_norm": 0.3236316740512848, "learning_rate": 2.03451763344644e-05, "loss": 0.0184, "step": 71700 }, { "grad_norm": 0.25700652599334717, "learning_rate": 2.0331865351497482e-05, "loss": 0.0195, "step": 71710 }, { "grad_norm": 0.23568931221961975, "learning_rate": 2.0318557612993185e-05, "loss": 0.0256, "step": 71720 }, { "grad_norm": 0.2726360857486725, "learning_rate": 2.0305253120406782e-05, "loss": 0.0182, "step": 71730 }, { "grad_norm": 0.22291994094848633, "learning_rate": 2.029195187519325e-05, "loss": 0.0148, "step": 71740 }, { "grad_norm": 0.2189021110534668, "learning_rate": 2.0278653878807217e-05, "loss": 0.0192, "step": 71750 }, { "grad_norm": 0.2206203192472458, "learning_rate": 2.0265359132702892e-05, "loss": 0.0201, "step": 71760 }, { "grad_norm": 0.17416027188301086, "learning_rate": 2.0252067638334182e-05, "loss": 0.0184, "step": 71770 }, { "grad_norm": 0.21273837983608246, "learning_rate": 2.023877939715464e-05, "loss": 0.0163, "step": 71780 }, { "grad_norm": 0.4478318393230438, "learning_rate": 2.0225494410617448e-05, "loss": 0.0199, "step": 71790 }, { "grad_norm": 0.19846481084823608, "learning_rate": 2.02122126801754e-05, "loss": 0.0131, "step": 71800 }, { "grad_norm": 0.27652111649513245, "learning_rate": 2.0198934207281018e-05, "loss": 0.017, "step": 71810 }, { "grad_norm": 0.24670623242855072, "learning_rate": 2.0185658993386357e-05, "loss": 0.0161, "step": 71820 }, { "grad_norm": 0.17607258260250092, "learning_rate": 2.0172387039943226e-05, "loss": 0.0166, "step": 71830 }, { "grad_norm": 0.13599929213523865, "learning_rate": 2.0159118348402968e-05, "loss": 0.0205, "step": 71840 }, { "grad_norm": 0.27875351905822754, "learning_rate": 2.0145852920216697e-05, "loss": 0.0165, "step": 71850 }, { "grad_norm": 0.39211371541023254, "learning_rate": 2.0132590756835045e-05, "loss": 0.0216, "step": 71860 }, { "grad_norm": 0.24077537655830383, "learning_rate": 2.0119331859708386e-05, "loss": 0.0264, "step": 71870 }, { "grad_norm": 0.26548534631729126, "learning_rate": 2.0106076230286654e-05, "loss": 0.0256, "step": 71880 }, { "grad_norm": 0.2488197535276413, "learning_rate": 2.0092823870019488e-05, "loss": 0.0183, "step": 71890 }, { "grad_norm": 0.15540237724781036, "learning_rate": 2.0079574780356115e-05, "loss": 0.0164, "step": 71900 }, { "grad_norm": 0.28837528824806213, "learning_rate": 2.0066328962745486e-05, "loss": 0.0198, "step": 71910 }, { "grad_norm": 0.12733900547027588, "learning_rate": 2.00530864186361e-05, "loss": 0.0181, "step": 71920 }, { "grad_norm": 0.5616328716278076, "learning_rate": 2.003984714947618e-05, "loss": 0.017, "step": 71930 }, { "grad_norm": 0.4575677216053009, "learning_rate": 2.0026611156713515e-05, "loss": 0.0264, "step": 71940 }, { "grad_norm": 0.21534216403961182, "learning_rate": 2.0013378441795584e-05, "loss": 0.0154, "step": 71950 }, { "grad_norm": 0.18064798414707184, "learning_rate": 2.0000149006169506e-05, "loss": 0.0176, "step": 71960 }, { "grad_norm": 0.21667852997779846, "learning_rate": 1.9986922851282048e-05, "loss": 0.0192, "step": 71970 }, { "grad_norm": 0.2112332582473755, "learning_rate": 1.997369997857956e-05, "loss": 0.0139, "step": 71980 }, { "grad_norm": 0.20440064370632172, "learning_rate": 1.9960480389508103e-05, "loss": 0.0136, "step": 71990 }, { "grad_norm": 0.21419154107570648, "learning_rate": 1.9947264085513363e-05, "loss": 0.0169, "step": 72000 }, { "grad_norm": 0.15571200847625732, "learning_rate": 1.993405106804062e-05, "loss": 0.0197, "step": 72010 }, { "grad_norm": 0.33851587772369385, "learning_rate": 1.9920841338534867e-05, "loss": 0.0232, "step": 72020 }, { "grad_norm": 0.256368488073349, "learning_rate": 1.9907634898440646e-05, "loss": 0.0145, "step": 72030 }, { "grad_norm": 0.1938144415616989, "learning_rate": 1.989443174920227e-05, "loss": 0.0155, "step": 72040 }, { "grad_norm": 0.23134100437164307, "learning_rate": 1.988123189226355e-05, "loss": 0.0252, "step": 72050 }, { "grad_norm": 0.43572670221328735, "learning_rate": 1.986803532906805e-05, "loss": 0.0214, "step": 72060 }, { "grad_norm": 0.19025352597236633, "learning_rate": 1.985484206105889e-05, "loss": 0.011, "step": 72070 }, { "grad_norm": 0.16507619619369507, "learning_rate": 1.9841652089678887e-05, "loss": 0.0195, "step": 72080 }, { "grad_norm": 0.19698990881443024, "learning_rate": 1.9828465416370435e-05, "loss": 0.0223, "step": 72090 }, { "grad_norm": 0.44150158762931824, "learning_rate": 1.9815282042575685e-05, "loss": 0.0271, "step": 72100 }, { "grad_norm": 0.17012560367584229, "learning_rate": 1.9802101969736285e-05, "loss": 0.0154, "step": 72110 }, { "grad_norm": 0.19418774545192719, "learning_rate": 1.9788925199293633e-05, "loss": 0.0189, "step": 72120 }, { "grad_norm": 0.2741372287273407, "learning_rate": 1.977575173268868e-05, "loss": 0.0136, "step": 72130 }, { "grad_norm": 0.1495293378829956, "learning_rate": 1.9762581571362094e-05, "loss": 0.0175, "step": 72140 }, { "grad_norm": 0.34404781460762024, "learning_rate": 1.9749414716754112e-05, "loss": 0.0248, "step": 72150 }, { "grad_norm": 0.22906385362148285, "learning_rate": 1.973625117030465e-05, "loss": 0.0167, "step": 72160 }, { "grad_norm": 0.2130032628774643, "learning_rate": 1.972309093345326e-05, "loss": 0.0189, "step": 72170 }, { "grad_norm": 0.2592487633228302, "learning_rate": 1.970993400763915e-05, "loss": 0.0214, "step": 72180 }, { "grad_norm": 0.22636844217777252, "learning_rate": 1.969678039430109e-05, "loss": 0.0231, "step": 72190 }, { "grad_norm": 0.47452637553215027, "learning_rate": 1.9683630094877564e-05, "loss": 0.0131, "step": 72200 }, { "grad_norm": 0.1446428894996643, "learning_rate": 1.9670483110806696e-05, "loss": 0.0133, "step": 72210 }, { "grad_norm": 0.21497993171215057, "learning_rate": 1.9657339443526172e-05, "loss": 0.0251, "step": 72220 }, { "grad_norm": 0.4329822361469269, "learning_rate": 1.964419909447339e-05, "loss": 0.0203, "step": 72230 }, { "grad_norm": 0.2710953950881958, "learning_rate": 1.963106206508536e-05, "loss": 0.0163, "step": 72240 }, { "grad_norm": 0.16321927309036255, "learning_rate": 1.9617928356798738e-05, "loss": 0.0145, "step": 72250 }, { "grad_norm": 0.29986438155174255, "learning_rate": 1.960479797104977e-05, "loss": 0.0211, "step": 72260 }, { "grad_norm": 0.2523258626461029, "learning_rate": 1.9591670909274418e-05, "loss": 0.0182, "step": 72270 }, { "grad_norm": 0.17092202603816986, "learning_rate": 1.9578547172908184e-05, "loss": 0.0176, "step": 72280 }, { "grad_norm": 0.27870652079582214, "learning_rate": 1.9565426763386325e-05, "loss": 0.0186, "step": 72290 }, { "grad_norm": 0.22895829379558563, "learning_rate": 1.9552309682143622e-05, "loss": 0.0155, "step": 72300 }, { "grad_norm": 0.26891759037971497, "learning_rate": 1.9539195930614574e-05, "loss": 0.0219, "step": 72310 }, { "grad_norm": 0.8331323266029358, "learning_rate": 1.952608551023324e-05, "loss": 0.0157, "step": 72320 }, { "grad_norm": 0.3056681454181671, "learning_rate": 1.9512978422433397e-05, "loss": 0.0166, "step": 72330 }, { "grad_norm": 0.1083008274435997, "learning_rate": 1.9499874668648376e-05, "loss": 0.0154, "step": 72340 }, { "grad_norm": 0.2115100473165512, "learning_rate": 1.9486774250311213e-05, "loss": 0.015, "step": 72350 }, { "grad_norm": 0.2266327440738678, "learning_rate": 1.9473677168854537e-05, "loss": 0.0197, "step": 72360 }, { "grad_norm": 0.15053533017635345, "learning_rate": 1.946058342571065e-05, "loss": 0.0232, "step": 72370 }, { "grad_norm": 0.22222289443016052, "learning_rate": 1.9447493022311424e-05, "loss": 0.0138, "step": 72380 }, { "grad_norm": 0.1551225334405899, "learning_rate": 1.9434405960088442e-05, "loss": 0.0146, "step": 72390 }, { "grad_norm": 0.3551100194454193, "learning_rate": 1.9421322240472857e-05, "loss": 0.0176, "step": 72400 }, { "grad_norm": 0.1942453235387802, "learning_rate": 1.9408241864895494e-05, "loss": 0.0165, "step": 72410 }, { "grad_norm": 0.19040542840957642, "learning_rate": 1.939516483478681e-05, "loss": 0.0132, "step": 72420 }, { "grad_norm": 0.2857000529766083, "learning_rate": 1.9382091151576902e-05, "loss": 0.019, "step": 72430 }, { "grad_norm": 0.2561831474304199, "learning_rate": 1.9369020816695454e-05, "loss": 0.017, "step": 72440 }, { "grad_norm": 0.24765650928020477, "learning_rate": 1.935595383157184e-05, "loss": 0.0127, "step": 72450 }, { "grad_norm": 0.4806866943836212, "learning_rate": 1.9342890197635057e-05, "loss": 0.0165, "step": 72460 }, { "grad_norm": 0.28046149015426636, "learning_rate": 1.9329829916313685e-05, "loss": 0.0173, "step": 72470 }, { "grad_norm": 0.12169212102890015, "learning_rate": 1.9316772989036013e-05, "loss": 0.0158, "step": 72480 }, { "grad_norm": 0.29867029190063477, "learning_rate": 1.930371941722991e-05, "loss": 0.0149, "step": 72490 }, { "grad_norm": 0.26607218384742737, "learning_rate": 1.929066920232292e-05, "loss": 0.0155, "step": 72500 }, { "grad_norm": 0.1781906634569168, "learning_rate": 1.9277622345742152e-05, "loss": 0.0154, "step": 72510 }, { "grad_norm": 0.16640037298202515, "learning_rate": 1.926457884891444e-05, "loss": 0.0144, "step": 72520 }, { "grad_norm": 0.20127154886722565, "learning_rate": 1.9251538713266147e-05, "loss": 0.0168, "step": 72530 }, { "grad_norm": 0.2538530230522156, "learning_rate": 1.9238501940223358e-05, "loss": 0.0198, "step": 72540 }, { "grad_norm": 0.21595297753810883, "learning_rate": 1.922546853121175e-05, "loss": 0.0139, "step": 72550 }, { "grad_norm": 0.13412758708000183, "learning_rate": 1.9212438487656647e-05, "loss": 0.0157, "step": 72560 }, { "grad_norm": 0.13348659873008728, "learning_rate": 1.9199411810982965e-05, "loss": 0.0149, "step": 72570 }, { "grad_norm": 0.1920870542526245, "learning_rate": 1.918638850261532e-05, "loss": 0.019, "step": 72580 }, { "grad_norm": 0.45330795645713806, "learning_rate": 1.9173368563977885e-05, "loss": 0.0192, "step": 72590 }, { "grad_norm": 0.14467689394950867, "learning_rate": 1.9160351996494513e-05, "loss": 0.0189, "step": 72600 }, { "grad_norm": 0.2006203830242157, "learning_rate": 1.9147338801588677e-05, "loss": 0.0148, "step": 72610 }, { "grad_norm": 0.22734937071800232, "learning_rate": 1.913432898068351e-05, "loss": 0.0173, "step": 72620 }, { "grad_norm": 0.3309996724128723, "learning_rate": 1.912132253520169e-05, "loss": 0.0193, "step": 72630 }, { "grad_norm": 0.2642941176891327, "learning_rate": 1.910831946656564e-05, "loss": 0.0215, "step": 72640 }, { "grad_norm": 0.29489561915397644, "learning_rate": 1.909531977619731e-05, "loss": 0.0171, "step": 72650 }, { "grad_norm": 0.31695258617401123, "learning_rate": 1.908232346551834e-05, "loss": 0.0118, "step": 72660 }, { "grad_norm": 0.3681302070617676, "learning_rate": 1.9069330535949993e-05, "loss": 0.0246, "step": 72670 }, { "grad_norm": 0.8608875274658203, "learning_rate": 1.9056340988913167e-05, "loss": 0.0179, "step": 72680 }, { "grad_norm": 0.15621593594551086, "learning_rate": 1.9043354825828357e-05, "loss": 0.0148, "step": 72690 }, { "grad_norm": 0.3232012093067169, "learning_rate": 1.9030372048115724e-05, "loss": 0.017, "step": 72700 }, { "grad_norm": 0.2820318639278412, "learning_rate": 1.9017392657195056e-05, "loss": 0.0316, "step": 72710 }, { "grad_norm": 0.18852469325065613, "learning_rate": 1.9004416654485724e-05, "loss": 0.0171, "step": 72720 }, { "grad_norm": 0.24813970923423767, "learning_rate": 1.8991444041406785e-05, "loss": 0.0226, "step": 72730 }, { "grad_norm": 0.39042210578918457, "learning_rate": 1.8978474819376912e-05, "loss": 0.0195, "step": 72740 }, { "grad_norm": 0.2593066096305847, "learning_rate": 1.8965508989814406e-05, "loss": 0.0174, "step": 72750 }, { "grad_norm": 0.22185319662094116, "learning_rate": 1.8952546554137158e-05, "loss": 0.0129, "step": 72760 }, { "grad_norm": 0.2137143313884735, "learning_rate": 1.8939587513762757e-05, "loss": 0.0176, "step": 72770 }, { "grad_norm": 0.11204379051923752, "learning_rate": 1.892663187010835e-05, "loss": 0.0155, "step": 72780 }, { "grad_norm": 0.16913153231143951, "learning_rate": 1.891367962459077e-05, "loss": 0.0163, "step": 72790 }, { "grad_norm": 0.2816745638847351, "learning_rate": 1.890073077862644e-05, "loss": 0.0156, "step": 72800 }, { "grad_norm": 0.2912369668483734, "learning_rate": 1.888778533363146e-05, "loss": 0.0149, "step": 72810 }, { "grad_norm": 0.2740756571292877, "learning_rate": 1.8874843291021478e-05, "loss": 0.0121, "step": 72820 }, { "grad_norm": 0.2927146255970001, "learning_rate": 1.8861904652211853e-05, "loss": 0.0227, "step": 72830 }, { "grad_norm": 0.16711436212062836, "learning_rate": 1.8848969418617507e-05, "loss": 0.0161, "step": 72840 }, { "grad_norm": 0.2532496750354767, "learning_rate": 1.8836037591653044e-05, "loss": 0.0146, "step": 72850 }, { "grad_norm": 0.18716856837272644, "learning_rate": 1.882310917273262e-05, "loss": 0.0136, "step": 72860 }, { "grad_norm": 0.1495150327682495, "learning_rate": 1.881018416327014e-05, "loss": 0.0129, "step": 72870 }, { "grad_norm": 0.3032839596271515, "learning_rate": 1.8797262564679014e-05, "loss": 0.0135, "step": 72880 }, { "grad_norm": 0.22631338238716125, "learning_rate": 1.8784344378372355e-05, "loss": 0.0214, "step": 72890 }, { "grad_norm": 0.46218013763427734, "learning_rate": 1.8771429605762843e-05, "loss": 0.0163, "step": 72900 }, { "grad_norm": 0.21890169382095337, "learning_rate": 1.8758518248262835e-05, "loss": 0.0187, "step": 72910 }, { "grad_norm": 0.2644137442111969, "learning_rate": 1.8745610307284305e-05, "loss": 0.0281, "step": 72920 }, { "grad_norm": 0.3963540196418762, "learning_rate": 1.8732705784238862e-05, "loss": 0.0178, "step": 72930 }, { "grad_norm": 0.11983920633792877, "learning_rate": 1.8719804680537682e-05, "loss": 0.0232, "step": 72940 }, { "grad_norm": 0.21712234616279602, "learning_rate": 1.870690699759164e-05, "loss": 0.0153, "step": 72950 }, { "grad_norm": 0.22473065555095673, "learning_rate": 1.8694012736811213e-05, "loss": 0.0145, "step": 72960 }, { "grad_norm": 0.24030090868473053, "learning_rate": 1.8681121899606463e-05, "loss": 0.0177, "step": 72970 }, { "grad_norm": 0.25513380765914917, "learning_rate": 1.8668234487387144e-05, "loss": 0.0159, "step": 72980 }, { "grad_norm": 0.2061319649219513, "learning_rate": 1.8655350501562592e-05, "loss": 0.0149, "step": 72990 }, { "grad_norm": 0.1466873586177826, "learning_rate": 1.8642469943541796e-05, "loss": 0.0223, "step": 73000 }, { "grad_norm": 0.44471344351768494, "learning_rate": 1.8629592814733328e-05, "loss": 0.0147, "step": 73010 }, { "grad_norm": 0.21367420256137848, "learning_rate": 1.8616719116545438e-05, "loss": 0.0198, "step": 73020 }, { "grad_norm": 0.24852974712848663, "learning_rate": 1.8603848850385952e-05, "loss": 0.0136, "step": 73030 }, { "grad_norm": 0.2605249881744385, "learning_rate": 1.859098201766236e-05, "loss": 0.0193, "step": 73040 }, { "grad_norm": 0.16782955825328827, "learning_rate": 1.8578118619781726e-05, "loss": 0.0196, "step": 73050 }, { "grad_norm": 0.2086249440908432, "learning_rate": 1.856525865815083e-05, "loss": 0.0165, "step": 73060 }, { "grad_norm": 0.29949110746383667, "learning_rate": 1.8552402134175966e-05, "loss": 0.0167, "step": 73070 }, { "grad_norm": 0.2082853764295578, "learning_rate": 1.8539549049263143e-05, "loss": 0.015, "step": 73080 }, { "grad_norm": 0.19538474082946777, "learning_rate": 1.8526699404817917e-05, "loss": 0.0142, "step": 73090 }, { "grad_norm": 0.19721032679080963, "learning_rate": 1.8513853202245546e-05, "loss": 0.0151, "step": 73100 }, { "grad_norm": 0.15909449756145477, "learning_rate": 1.850101044295081e-05, "loss": 0.0093, "step": 73110 }, { "grad_norm": 0.21173880994319916, "learning_rate": 1.848817112833825e-05, "loss": 0.0136, "step": 73120 }, { "grad_norm": 0.26346439123153687, "learning_rate": 1.8475335259811905e-05, "loss": 0.0147, "step": 73130 }, { "grad_norm": 0.18407078087329865, "learning_rate": 1.8462502838775508e-05, "loss": 0.0212, "step": 73140 }, { "grad_norm": 0.2828422784805298, "learning_rate": 1.8449673866632378e-05, "loss": 0.0181, "step": 73150 }, { "grad_norm": 0.3426313102245331, "learning_rate": 1.8436848344785473e-05, "loss": 0.0234, "step": 73160 }, { "grad_norm": 0.33776959776878357, "learning_rate": 1.8424026274637378e-05, "loss": 0.0182, "step": 73170 }, { "grad_norm": 0.3343336880207062, "learning_rate": 1.8411207657590312e-05, "loss": 0.0215, "step": 73180 }, { "grad_norm": 0.273562490940094, "learning_rate": 1.839839249504607e-05, "loss": 0.0168, "step": 73190 }, { "grad_norm": 0.2155861109495163, "learning_rate": 1.8385580788406108e-05, "loss": 0.0222, "step": 73200 }, { "grad_norm": 0.1836349219083786, "learning_rate": 1.837277253907152e-05, "loss": 0.0144, "step": 73210 }, { "grad_norm": 0.22341911494731903, "learning_rate": 1.8359967748442957e-05, "loss": 0.0184, "step": 73220 }, { "grad_norm": 0.6460384130477905, "learning_rate": 1.8347166417920774e-05, "loss": 0.0168, "step": 73230 }, { "grad_norm": 0.2865510582923889, "learning_rate": 1.833436854890484e-05, "loss": 0.0178, "step": 73240 }, { "grad_norm": 1.197016954421997, "learning_rate": 1.8321574142794796e-05, "loss": 0.0127, "step": 73250 }, { "grad_norm": 0.1385461688041687, "learning_rate": 1.8308783200989748e-05, "loss": 0.0158, "step": 73260 }, { "grad_norm": 0.5244852900505066, "learning_rate": 1.829599572488855e-05, "loss": 0.0193, "step": 73270 }, { "grad_norm": 0.25721922516822815, "learning_rate": 1.828321171588957e-05, "loss": 0.0174, "step": 73280 }, { "grad_norm": 0.19844594597816467, "learning_rate": 1.827043117539089e-05, "loss": 0.0186, "step": 73290 }, { "grad_norm": 0.14916814863681793, "learning_rate": 1.8257654104790116e-05, "loss": 0.0143, "step": 73300 }, { "grad_norm": 0.16390974819660187, "learning_rate": 1.82448805054846e-05, "loss": 0.0133, "step": 73310 }, { "grad_norm": 0.26387742161750793, "learning_rate": 1.823211037887119e-05, "loss": 0.016, "step": 73320 }, { "grad_norm": 0.36923056840896606, "learning_rate": 1.821934372634645e-05, "loss": 0.014, "step": 73330 }, { "grad_norm": 0.12125350534915924, "learning_rate": 1.8206580549306472e-05, "loss": 0.0125, "step": 73340 }, { "grad_norm": 0.1340058296918869, "learning_rate": 1.819382084914707e-05, "loss": 0.0198, "step": 73350 }, { "grad_norm": 0.4991625249385834, "learning_rate": 1.8181064627263577e-05, "loss": 0.0123, "step": 73360 }, { "grad_norm": 0.29400917887687683, "learning_rate": 1.8168311885051022e-05, "loss": 0.0198, "step": 73370 }, { "grad_norm": 0.2919062376022339, "learning_rate": 1.815556262390402e-05, "loss": 0.0153, "step": 73380 }, { "grad_norm": 0.2939445376396179, "learning_rate": 1.8142816845216826e-05, "loss": 0.0158, "step": 73390 }, { "grad_norm": 0.5652218461036682, "learning_rate": 1.8130074550383276e-05, "loss": 0.0181, "step": 73400 }, { "grad_norm": 0.2217129021883011, "learning_rate": 1.8117335740796848e-05, "loss": 0.0189, "step": 73410 }, { "grad_norm": 0.5847073793411255, "learning_rate": 1.810460041785067e-05, "loss": 0.0168, "step": 73420 }, { "grad_norm": 0.16375890374183655, "learning_rate": 1.809186858293742e-05, "loss": 0.0264, "step": 73430 }, { "grad_norm": 0.1879919320344925, "learning_rate": 1.8079140237449448e-05, "loss": 0.0149, "step": 73440 }, { "grad_norm": 0.21576504409313202, "learning_rate": 1.8066415382778707e-05, "loss": 0.0206, "step": 73450 }, { "grad_norm": 0.1569223552942276, "learning_rate": 1.8053694020316786e-05, "loss": 0.0154, "step": 73460 }, { "grad_norm": 0.2718014121055603, "learning_rate": 1.8040976151454837e-05, "loss": 0.016, "step": 73470 }, { "grad_norm": 0.20997722446918488, "learning_rate": 1.8028261777583704e-05, "loss": 0.0271, "step": 73480 }, { "grad_norm": 0.19354656338691711, "learning_rate": 1.8015550900093758e-05, "loss": 0.0154, "step": 73490 }, { "grad_norm": 0.19539839029312134, "learning_rate": 1.800284352037512e-05, "loss": 0.0164, "step": 73500 }, { "grad_norm": 0.1485058069229126, "learning_rate": 1.799013963981738e-05, "loss": 0.0178, "step": 73510 }, { "grad_norm": 0.21871137619018555, "learning_rate": 1.797743925980987e-05, "loss": 0.0152, "step": 73520 }, { "grad_norm": 0.19464170932769775, "learning_rate": 1.7964742381741435e-05, "loss": 0.0172, "step": 73530 }, { "grad_norm": 0.2813572287559509, "learning_rate": 1.7952049007000627e-05, "loss": 0.0134, "step": 73540 }, { "grad_norm": 0.430112361907959, "learning_rate": 1.7939359136975535e-05, "loss": 0.0174, "step": 73550 }, { "grad_norm": 0.5593467950820923, "learning_rate": 1.7926672773053932e-05, "loss": 0.0154, "step": 73560 }, { "grad_norm": 0.2466837614774704, "learning_rate": 1.791398991662317e-05, "loss": 0.0241, "step": 73570 }, { "grad_norm": 0.24865280091762543, "learning_rate": 1.7901310569070244e-05, "loss": 0.0245, "step": 73580 }, { "grad_norm": 0.11613701283931732, "learning_rate": 1.7888634731781724e-05, "loss": 0.0147, "step": 73590 }, { "grad_norm": 0.2910471260547638, "learning_rate": 1.7875962406143844e-05, "loss": 0.0164, "step": 73600 }, { "grad_norm": 0.1510864645242691, "learning_rate": 1.7863293593542403e-05, "loss": 0.0293, "step": 73610 }, { "grad_norm": 0.11697960644960403, "learning_rate": 1.7850628295362853e-05, "loss": 0.0141, "step": 73620 }, { "grad_norm": 0.15373042225837708, "learning_rate": 1.7837966512990263e-05, "loss": 0.0169, "step": 73630 }, { "grad_norm": 0.41343051195144653, "learning_rate": 1.782530824780931e-05, "loss": 0.0187, "step": 73640 }, { "grad_norm": 0.2534441649913788, "learning_rate": 1.7812653501204263e-05, "loss": 0.0152, "step": 73650 }, { "grad_norm": 0.2022218108177185, "learning_rate": 1.780000227455903e-05, "loss": 0.0172, "step": 73660 }, { "grad_norm": 0.25859636068344116, "learning_rate": 1.778735456925716e-05, "loss": 0.0176, "step": 73670 }, { "grad_norm": 0.15961244702339172, "learning_rate": 1.7774710386681737e-05, "loss": 0.0137, "step": 73680 }, { "grad_norm": 0.22717563807964325, "learning_rate": 1.7762069728215545e-05, "loss": 0.0235, "step": 73690 }, { "grad_norm": 0.21327614784240723, "learning_rate": 1.7749432595240932e-05, "loss": 0.0151, "step": 73700 }, { "grad_norm": 0.3076542317867279, "learning_rate": 1.773679898913991e-05, "loss": 0.0211, "step": 73710 }, { "grad_norm": 0.2729393541812897, "learning_rate": 1.7724168911294015e-05, "loss": 0.0169, "step": 73720 }, { "grad_norm": 0.27227315306663513, "learning_rate": 1.7711542363084504e-05, "loss": 0.0162, "step": 73730 }, { "grad_norm": 0.17760948836803436, "learning_rate": 1.7698919345892158e-05, "loss": 0.0164, "step": 73740 }, { "grad_norm": 0.21756061911582947, "learning_rate": 1.768629986109743e-05, "loss": 0.0172, "step": 73750 }, { "grad_norm": 0.2173600047826767, "learning_rate": 1.7673683910080357e-05, "loss": 0.0141, "step": 73760 }, { "grad_norm": 0.243570938706398, "learning_rate": 1.7661071494220638e-05, "loss": 0.0165, "step": 73770 }, { "grad_norm": 0.21621334552764893, "learning_rate": 1.7648462614897498e-05, "loss": 0.0139, "step": 73780 }, { "grad_norm": 0.3205423653125763, "learning_rate": 1.7635857273489863e-05, "loss": 0.016, "step": 73790 }, { "grad_norm": 0.1739601194858551, "learning_rate": 1.7623255471376198e-05, "loss": 0.0154, "step": 73800 }, { "grad_norm": 0.27771279215812683, "learning_rate": 1.7610657209934646e-05, "loss": 0.0221, "step": 73810 }, { "grad_norm": 0.3425881862640381, "learning_rate": 1.7598062490542915e-05, "loss": 0.0176, "step": 73820 }, { "grad_norm": 0.19290311634540558, "learning_rate": 1.7585471314578377e-05, "loss": 0.0153, "step": 73830 }, { "grad_norm": 0.2045312076807022, "learning_rate": 1.7572883683417945e-05, "loss": 0.0168, "step": 73840 }, { "grad_norm": 0.26895827054977417, "learning_rate": 1.7560299598438212e-05, "loss": 0.0212, "step": 73850 }, { "grad_norm": 0.24135005474090576, "learning_rate": 1.7547719061015334e-05, "loss": 0.015, "step": 73860 }, { "grad_norm": 0.17340445518493652, "learning_rate": 1.75351420725251e-05, "loss": 0.0135, "step": 73870 }, { "grad_norm": 0.24248477816581726, "learning_rate": 1.7522568634342923e-05, "loss": 0.0131, "step": 73880 }, { "grad_norm": 0.16991852223873138, "learning_rate": 1.7509998747843825e-05, "loss": 0.0107, "step": 73890 }, { "grad_norm": 0.2752077281475067, "learning_rate": 1.7497432414402403e-05, "loss": 0.016, "step": 73900 }, { "grad_norm": 0.16117580235004425, "learning_rate": 1.7484869635392903e-05, "loss": 0.0145, "step": 73910 }, { "grad_norm": 0.18917705118656158, "learning_rate": 1.747231041218919e-05, "loss": 0.0159, "step": 73920 }, { "grad_norm": 0.19144149124622345, "learning_rate": 1.7459754746164692e-05, "loss": 0.0189, "step": 73930 }, { "grad_norm": 0.35754910111427307, "learning_rate": 1.744720263869249e-05, "loss": 0.0198, "step": 73940 }, { "grad_norm": 0.25125017762184143, "learning_rate": 1.7434654091145263e-05, "loss": 0.0152, "step": 73950 }, { "grad_norm": 0.984916627407074, "learning_rate": 1.742210910489532e-05, "loss": 0.0202, "step": 73960 }, { "grad_norm": 0.27129727602005005, "learning_rate": 1.7409567681314532e-05, "loss": 0.0157, "step": 73970 }, { "grad_norm": 0.2650489807128906, "learning_rate": 1.7397029821774434e-05, "loss": 0.0179, "step": 73980 }, { "grad_norm": 0.10867644101381302, "learning_rate": 1.7384495527646126e-05, "loss": 0.0212, "step": 73990 }, { "grad_norm": 0.18665996193885803, "learning_rate": 1.737196480030035e-05, "loss": 0.0126, "step": 74000 }, { "grad_norm": 0.3790092170238495, "learning_rate": 1.7359437641107447e-05, "loss": 0.0185, "step": 74010 }, { "grad_norm": 0.17119085788726807, "learning_rate": 1.7346914051437387e-05, "loss": 0.0171, "step": 74020 }, { "grad_norm": 0.3012326955795288, "learning_rate": 1.733439403265969e-05, "loss": 0.0155, "step": 74030 }, { "grad_norm": 0.2106815129518509, "learning_rate": 1.732187758614357e-05, "loss": 0.0227, "step": 74040 }, { "grad_norm": 0.20193608105182648, "learning_rate": 1.730936471325777e-05, "loss": 0.018, "step": 74050 }, { "grad_norm": 0.19711077213287354, "learning_rate": 1.7296855415370718e-05, "loss": 0.0196, "step": 74060 }, { "grad_norm": 0.23660290241241455, "learning_rate": 1.728434969385035e-05, "loss": 0.018, "step": 74070 }, { "grad_norm": 0.19369147717952728, "learning_rate": 1.7271847550064346e-05, "loss": 0.0197, "step": 74080 }, { "grad_norm": 0.24446462094783783, "learning_rate": 1.725934898537987e-05, "loss": 0.0185, "step": 74090 }, { "grad_norm": 0.1655256301164627, "learning_rate": 1.7246854001163783e-05, "loss": 0.0163, "step": 74100 }, { "grad_norm": 0.3250425457954407, "learning_rate": 1.7234362598782476e-05, "loss": 0.0167, "step": 74110 }, { "grad_norm": 0.24645033478736877, "learning_rate": 1.722187477960202e-05, "loss": 0.0197, "step": 74120 }, { "grad_norm": 1.1357879638671875, "learning_rate": 1.7209390544988057e-05, "loss": 0.0233, "step": 74130 }, { "grad_norm": 0.17542032897472382, "learning_rate": 1.7196909896305856e-05, "loss": 0.0209, "step": 74140 }, { "grad_norm": 0.26324382424354553, "learning_rate": 1.718443283492026e-05, "loss": 0.0179, "step": 74150 }, { "grad_norm": 0.18512454628944397, "learning_rate": 1.7171959362195746e-05, "loss": 0.0266, "step": 74160 }, { "grad_norm": 0.13779284060001373, "learning_rate": 1.715948947949642e-05, "loss": 0.0191, "step": 74170 }, { "grad_norm": 0.15579944849014282, "learning_rate": 1.714702318818593e-05, "loss": 0.0164, "step": 74180 }, { "grad_norm": 0.16491450369358063, "learning_rate": 1.7134560489627592e-05, "loss": 0.0171, "step": 74190 }, { "grad_norm": 0.29683223366737366, "learning_rate": 1.712210138518431e-05, "loss": 0.0163, "step": 74200 }, { "grad_norm": 0.2540862262248993, "learning_rate": 1.71096458762186e-05, "loss": 0.0324, "step": 74210 }, { "grad_norm": 0.3580063283443451, "learning_rate": 1.7097193964092552e-05, "loss": 0.0207, "step": 74220 }, { "grad_norm": 0.22525060176849365, "learning_rate": 1.708474565016792e-05, "loss": 0.0189, "step": 74230 }, { "grad_norm": 0.25442829728126526, "learning_rate": 1.7072300935806e-05, "loss": 0.0176, "step": 74240 }, { "grad_norm": 0.28293803334236145, "learning_rate": 1.705985982236776e-05, "loss": 0.0163, "step": 74250 }, { "grad_norm": 0.23053902387619019, "learning_rate": 1.7047422311213697e-05, "loss": 0.0141, "step": 74260 }, { "grad_norm": 0.11634553223848343, "learning_rate": 1.703498840370401e-05, "loss": 0.0155, "step": 74270 }, { "grad_norm": 0.2969934344291687, "learning_rate": 1.702255810119842e-05, "loss": 0.0171, "step": 74280 }, { "grad_norm": 0.15885911881923676, "learning_rate": 1.7010131405056305e-05, "loss": 0.0165, "step": 74290 }, { "grad_norm": 0.15763670206069946, "learning_rate": 1.6997708316636613e-05, "loss": 0.0158, "step": 74300 }, { "grad_norm": 0.1555415540933609, "learning_rate": 1.6985288837297932e-05, "loss": 0.019, "step": 74310 }, { "grad_norm": 0.43692126870155334, "learning_rate": 1.6972872968398394e-05, "loss": 0.0183, "step": 74320 }, { "grad_norm": 0.29088839888572693, "learning_rate": 1.6960460711295845e-05, "loss": 0.0178, "step": 74330 }, { "grad_norm": 0.25766122341156006, "learning_rate": 1.6948052067347623e-05, "loss": 0.0194, "step": 74340 }, { "grad_norm": 0.1535627245903015, "learning_rate": 1.693564703791075e-05, "loss": 0.0168, "step": 74350 }, { "grad_norm": 0.17631791532039642, "learning_rate": 1.6923245624341793e-05, "loss": 0.02, "step": 74360 }, { "grad_norm": 0.29222196340560913, "learning_rate": 1.6910847827996962e-05, "loss": 0.0162, "step": 74370 }, { "grad_norm": 0.24061515927314758, "learning_rate": 1.689845365023206e-05, "loss": 0.0156, "step": 74380 }, { "grad_norm": 0.21528761088848114, "learning_rate": 1.6886063092402505e-05, "loss": 0.0174, "step": 74390 }, { "grad_norm": 0.443432092666626, "learning_rate": 1.6873676155863317e-05, "loss": 0.0251, "step": 74400 }, { "grad_norm": 0.16357764601707458, "learning_rate": 1.6861292841969085e-05, "loss": 0.0144, "step": 74410 }, { "grad_norm": 0.15417087078094482, "learning_rate": 1.684891315207406e-05, "loss": 0.0176, "step": 74420 }, { "grad_norm": 0.21459141373634338, "learning_rate": 1.6836537087532034e-05, "loss": 0.0164, "step": 74430 }, { "grad_norm": 0.15660926699638367, "learning_rate": 1.6824164649696465e-05, "loss": 0.0162, "step": 74440 }, { "grad_norm": 0.1742984801530838, "learning_rate": 1.681179583992034e-05, "loss": 0.0154, "step": 74450 }, { "grad_norm": 0.1253269910812378, "learning_rate": 1.679943065955636e-05, "loss": 0.0133, "step": 74460 }, { "grad_norm": 0.5373473167419434, "learning_rate": 1.678706910995671e-05, "loss": 0.0171, "step": 74470 }, { "grad_norm": 0.2721037268638611, "learning_rate": 1.6774711192473262e-05, "loss": 0.0204, "step": 74480 }, { "grad_norm": 0.21079878509044647, "learning_rate": 1.676235690845743e-05, "loss": 0.0165, "step": 74490 }, { "grad_norm": 0.26839956641197205, "learning_rate": 1.6750006259260285e-05, "loss": 0.0167, "step": 74500 }, { "grad_norm": 0.14141146838665009, "learning_rate": 1.673765924623244e-05, "loss": 0.0146, "step": 74510 }, { "grad_norm": 0.29445314407348633, "learning_rate": 1.67253158707242e-05, "loss": 0.0186, "step": 74520 }, { "grad_norm": 0.3007930517196655, "learning_rate": 1.671297613408537e-05, "loss": 0.0155, "step": 74530 }, { "grad_norm": 0.40509459376335144, "learning_rate": 1.6700640037665444e-05, "loss": 0.0164, "step": 74540 }, { "grad_norm": 0.2590121328830719, "learning_rate": 1.668830758281344e-05, "loss": 0.0169, "step": 74550 }, { "grad_norm": 0.612755537033081, "learning_rate": 1.6675978770878052e-05, "loss": 0.0147, "step": 74560 }, { "grad_norm": 0.19867746531963348, "learning_rate": 1.6663653603207506e-05, "loss": 0.0177, "step": 74570 }, { "grad_norm": 0.2369145154953003, "learning_rate": 1.6651332081149685e-05, "loss": 0.0163, "step": 74580 }, { "grad_norm": 0.2829420566558838, "learning_rate": 1.6639014206052046e-05, "loss": 0.0168, "step": 74590 }, { "grad_norm": 0.16342641413211823, "learning_rate": 1.6626699979261657e-05, "loss": 0.0133, "step": 74600 }, { "grad_norm": 0.29067936539649963, "learning_rate": 1.6614389402125197e-05, "loss": 0.0171, "step": 74610 }, { "grad_norm": 0.1876353770494461, "learning_rate": 1.66020824759889e-05, "loss": 0.0143, "step": 74620 }, { "grad_norm": 0.18228726089000702, "learning_rate": 1.6589779202198675e-05, "loss": 0.0134, "step": 74630 }, { "grad_norm": 0.2471468299627304, "learning_rate": 1.6577479582099924e-05, "loss": 0.0168, "step": 74640 }, { "grad_norm": 0.21973393857479095, "learning_rate": 1.6565183617037788e-05, "loss": 0.0147, "step": 74650 }, { "grad_norm": 0.30041417479515076, "learning_rate": 1.6552891308356894e-05, "loss": 0.0168, "step": 74660 }, { "grad_norm": 0.28612735867500305, "learning_rate": 1.6540602657401526e-05, "loss": 0.0219, "step": 74670 }, { "grad_norm": 0.16511747241020203, "learning_rate": 1.6528317665515537e-05, "loss": 0.0261, "step": 74680 }, { "grad_norm": 0.25478222966194153, "learning_rate": 1.6516036334042416e-05, "loss": 0.0221, "step": 74690 }, { "grad_norm": 0.3088504374027252, "learning_rate": 1.650375866432519e-05, "loss": 0.0224, "step": 74700 }, { "grad_norm": 0.2972053587436676, "learning_rate": 1.649148465770659e-05, "loss": 0.0172, "step": 74710 }, { "grad_norm": 0.20284821093082428, "learning_rate": 1.647921431552883e-05, "loss": 0.0194, "step": 74720 }, { "grad_norm": 0.356928288936615, "learning_rate": 1.646694763913381e-05, "loss": 0.0176, "step": 74730 }, { "grad_norm": 0.24561525881290436, "learning_rate": 1.6454684629862966e-05, "loss": 0.0138, "step": 74740 }, { "grad_norm": 0.21014665067195892, "learning_rate": 1.6442425289057388e-05, "loss": 0.015, "step": 74750 }, { "grad_norm": 0.24173016846179962, "learning_rate": 1.6430169618057718e-05, "loss": 0.0135, "step": 74760 }, { "grad_norm": 0.200398787856102, "learning_rate": 1.641791761820422e-05, "loss": 0.0175, "step": 74770 }, { "grad_norm": 0.2740817368030548, "learning_rate": 1.640566929083676e-05, "loss": 0.0266, "step": 74780 }, { "grad_norm": 0.1661539226770401, "learning_rate": 1.6393424637294823e-05, "loss": 0.0286, "step": 74790 }, { "grad_norm": 0.16484396159648895, "learning_rate": 1.638118365891742e-05, "loss": 0.0191, "step": 74800 }, { "grad_norm": 0.189507395029068, "learning_rate": 1.636894635704323e-05, "loss": 0.0217, "step": 74810 }, { "grad_norm": 0.17860695719718933, "learning_rate": 1.6356712733010517e-05, "loss": 0.0166, "step": 74820 }, { "grad_norm": 0.1689353734254837, "learning_rate": 1.6344482788157096e-05, "loss": 0.0152, "step": 74830 }, { "grad_norm": 0.3037956655025482, "learning_rate": 1.6332256523820444e-05, "loss": 0.0144, "step": 74840 }, { "grad_norm": 0.17554569244384766, "learning_rate": 1.6320033941337597e-05, "loss": 0.0146, "step": 74850 }, { "grad_norm": 0.32082730531692505, "learning_rate": 1.6307815042045216e-05, "loss": 0.0179, "step": 74860 }, { "grad_norm": 0.1536220759153366, "learning_rate": 1.6295599827279513e-05, "loss": 0.0147, "step": 74870 }, { "grad_norm": 0.31765326857566833, "learning_rate": 1.6283388298376356e-05, "loss": 0.0207, "step": 74880 }, { "grad_norm": 0.27716949582099915, "learning_rate": 1.6271180456671125e-05, "loss": 0.0183, "step": 74890 }, { "grad_norm": 0.1265663057565689, "learning_rate": 1.625897630349893e-05, "loss": 0.0143, "step": 74900 }, { "grad_norm": 0.2497253119945526, "learning_rate": 1.6246775840194335e-05, "loss": 0.016, "step": 74910 }, { "grad_norm": 0.13625690340995789, "learning_rate": 1.6234579068091604e-05, "loss": 0.0155, "step": 74920 }, { "grad_norm": 0.1678418219089508, "learning_rate": 1.6222385988524524e-05, "loss": 0.0125, "step": 74930 }, { "grad_norm": 0.19935953617095947, "learning_rate": 1.6210196602826544e-05, "loss": 0.0225, "step": 74940 }, { "grad_norm": 0.12955045700073242, "learning_rate": 1.6198010912330646e-05, "loss": 0.0123, "step": 74950 }, { "grad_norm": 0.3408057689666748, "learning_rate": 1.6185828918369456e-05, "loss": 0.0174, "step": 74960 }, { "grad_norm": 0.15230682492256165, "learning_rate": 1.617365062227517e-05, "loss": 0.0132, "step": 74970 }, { "grad_norm": 0.2427407056093216, "learning_rate": 1.616147602537962e-05, "loss": 0.0229, "step": 74980 }, { "grad_norm": 0.2682199776172638, "learning_rate": 1.614930512901415e-05, "loss": 0.0193, "step": 74990 }, { "grad_norm": 0.30360373854637146, "learning_rate": 1.6137137934509804e-05, "loss": 0.0176, "step": 75000 }, { "grad_norm": 0.246768057346344, "learning_rate": 1.6124974443197123e-05, "loss": 0.0169, "step": 75010 }, { "grad_norm": 0.19545169174671173, "learning_rate": 1.6112814656406306e-05, "loss": 0.0162, "step": 75020 }, { "grad_norm": 0.3514261245727539, "learning_rate": 1.6100658575467125e-05, "loss": 0.0198, "step": 75030 }, { "grad_norm": 0.17484313249588013, "learning_rate": 1.6088506201708974e-05, "loss": 0.0167, "step": 75040 }, { "grad_norm": 0.21834823489189148, "learning_rate": 1.6076357536460785e-05, "loss": 0.0169, "step": 75050 }, { "grad_norm": 0.14388340711593628, "learning_rate": 1.606421258105113e-05, "loss": 0.0159, "step": 75060 }, { "grad_norm": 0.21298529207706451, "learning_rate": 1.6052071336808183e-05, "loss": 0.0145, "step": 75070 }, { "grad_norm": 0.25840553641319275, "learning_rate": 1.6039933805059658e-05, "loss": 0.0142, "step": 75080 }, { "grad_norm": 0.15036027133464813, "learning_rate": 1.6027799987132915e-05, "loss": 0.0139, "step": 75090 }, { "grad_norm": 0.2725134491920471, "learning_rate": 1.6015669884354888e-05, "loss": 0.0184, "step": 75100 }, { "grad_norm": 0.2205890715122223, "learning_rate": 1.6003543498052125e-05, "loss": 0.0136, "step": 75110 }, { "grad_norm": 0.3034122884273529, "learning_rate": 1.5991420829550715e-05, "loss": 0.0172, "step": 75120 }, { "grad_norm": 0.3114423155784607, "learning_rate": 1.5979301880176405e-05, "loss": 0.0158, "step": 75130 }, { "grad_norm": 0.24385903775691986, "learning_rate": 1.5967186651254485e-05, "loss": 0.0168, "step": 75140 }, { "grad_norm": 0.22682078182697296, "learning_rate": 1.5955075144109853e-05, "loss": 0.0135, "step": 75150 }, { "grad_norm": 0.24700692296028137, "learning_rate": 1.594296736006703e-05, "loss": 0.0182, "step": 75160 }, { "grad_norm": 0.1570092886686325, "learning_rate": 1.59308633004501e-05, "loss": 0.0256, "step": 75170 }, { "grad_norm": 0.28880831599235535, "learning_rate": 1.5918762966582724e-05, "loss": 0.0155, "step": 75180 }, { "grad_norm": 0.137921541929245, "learning_rate": 1.5906666359788203e-05, "loss": 0.015, "step": 75190 }, { "grad_norm": 0.3760054409503937, "learning_rate": 1.589457348138938e-05, "loss": 0.0166, "step": 75200 }, { "grad_norm": 0.27794769406318665, "learning_rate": 1.588248433270872e-05, "loss": 0.0218, "step": 75210 }, { "grad_norm": 0.24421614408493042, "learning_rate": 1.587039891506828e-05, "loss": 0.0148, "step": 75220 }, { "grad_norm": 0.19588376581668854, "learning_rate": 1.5858317229789715e-05, "loss": 0.0156, "step": 75230 }, { "grad_norm": 0.24379436671733856, "learning_rate": 1.584623927819423e-05, "loss": 0.0141, "step": 75240 }, { "grad_norm": 0.23578591644763947, "learning_rate": 1.583416506160269e-05, "loss": 0.0117, "step": 75250 }, { "grad_norm": 0.21889282763004303, "learning_rate": 1.5822094581335472e-05, "loss": 0.0173, "step": 75260 }, { "grad_norm": 0.18744871020317078, "learning_rate": 1.5810027838712604e-05, "loss": 0.0173, "step": 75270 }, { "grad_norm": 0.1569328010082245, "learning_rate": 1.5797964835053687e-05, "loss": 0.0166, "step": 75280 }, { "grad_norm": 0.19040776789188385, "learning_rate": 1.5785905571677933e-05, "loss": 0.0165, "step": 75290 }, { "grad_norm": 0.20831555128097534, "learning_rate": 1.5773850049904086e-05, "loss": 0.0206, "step": 75300 }, { "grad_norm": 0.3315785527229309, "learning_rate": 1.5761798271050543e-05, "loss": 0.0196, "step": 75310 }, { "grad_norm": 0.33994367718696594, "learning_rate": 1.5749750236435277e-05, "loss": 0.0174, "step": 75320 }, { "grad_norm": 0.24576054513454437, "learning_rate": 1.5737705947375815e-05, "loss": 0.0164, "step": 75330 }, { "grad_norm": 0.2657853662967682, "learning_rate": 1.5725665405189322e-05, "loss": 0.0162, "step": 75340 }, { "grad_norm": 0.3010832965373993, "learning_rate": 1.571362861119253e-05, "loss": 0.0144, "step": 75350 }, { "grad_norm": 0.19270363450050354, "learning_rate": 1.570159556670177e-05, "loss": 0.0144, "step": 75360 }, { "grad_norm": 0.1938496083021164, "learning_rate": 1.5689566273032946e-05, "loss": 0.0166, "step": 75370 }, { "grad_norm": 0.18802110850811005, "learning_rate": 1.5677540731501583e-05, "loss": 0.0178, "step": 75380 }, { "grad_norm": 0.23270165920257568, "learning_rate": 1.5665518943422746e-05, "loss": 0.0138, "step": 75390 }, { "grad_norm": 0.3097327947616577, "learning_rate": 1.5653500910111134e-05, "loss": 0.0164, "step": 75400 }, { "grad_norm": 0.25700244307518005, "learning_rate": 1.5641486632881023e-05, "loss": 0.0179, "step": 75410 }, { "grad_norm": 0.20913589000701904, "learning_rate": 1.56294761130463e-05, "loss": 0.0217, "step": 75420 }, { "grad_norm": 0.27073803544044495, "learning_rate": 1.5617469351920373e-05, "loss": 0.0163, "step": 75430 }, { "grad_norm": 0.2728828489780426, "learning_rate": 1.5605466350816315e-05, "loss": 0.0173, "step": 75440 }, { "grad_norm": 0.3156985342502594, "learning_rate": 1.559346711104674e-05, "loss": 0.0152, "step": 75450 }, { "grad_norm": 1.086182951927185, "learning_rate": 1.5581471633923883e-05, "loss": 0.015, "step": 75460 }, { "grad_norm": 0.22000877559185028, "learning_rate": 1.5569479920759505e-05, "loss": 0.0189, "step": 75470 }, { "grad_norm": 0.35486334562301636, "learning_rate": 1.5557491972865072e-05, "loss": 0.0158, "step": 75480 }, { "grad_norm": 0.19303038716316223, "learning_rate": 1.554550779155152e-05, "loss": 0.0135, "step": 75490 }, { "grad_norm": 0.5739419460296631, "learning_rate": 1.5533527378129453e-05, "loss": 0.0169, "step": 75500 }, { "grad_norm": 0.15147021412849426, "learning_rate": 1.552155073390899e-05, "loss": 0.0166, "step": 75510 }, { "grad_norm": 0.28658527135849, "learning_rate": 1.5509577860199903e-05, "loss": 0.0168, "step": 75520 }, { "grad_norm": 0.189740389585495, "learning_rate": 1.5497608758311525e-05, "loss": 0.0154, "step": 75530 }, { "grad_norm": 0.1699860543012619, "learning_rate": 1.5485643429552797e-05, "loss": 0.0168, "step": 75540 }, { "grad_norm": 0.3782729208469391, "learning_rate": 1.5473681875232197e-05, "loss": 0.0217, "step": 75550 }, { "grad_norm": 0.3044566512107849, "learning_rate": 1.5461724096657837e-05, "loss": 0.0151, "step": 75560 }, { "grad_norm": 0.17312946915626526, "learning_rate": 1.5449770095137416e-05, "loss": 0.0133, "step": 75570 }, { "grad_norm": 0.17152419686317444, "learning_rate": 1.5437819871978178e-05, "loss": 0.0132, "step": 75580 }, { "grad_norm": 0.39065489172935486, "learning_rate": 1.5425873428486992e-05, "loss": 0.0258, "step": 75590 }, { "grad_norm": 0.2072681188583374, "learning_rate": 1.5413930765970312e-05, "loss": 0.0168, "step": 75600 }, { "grad_norm": 0.24129201471805573, "learning_rate": 1.5401991885734175e-05, "loss": 0.0158, "step": 75610 }, { "grad_norm": 0.12611551582813263, "learning_rate": 1.5390056789084174e-05, "loss": 0.0157, "step": 75620 }, { "grad_norm": 0.21692757308483124, "learning_rate": 1.5378125477325545e-05, "loss": 0.0156, "step": 75630 }, { "grad_norm": 0.12741589546203613, "learning_rate": 1.5366197951763046e-05, "loss": 0.0178, "step": 75640 }, { "grad_norm": 0.38624510169029236, "learning_rate": 1.5354274213701075e-05, "loss": 0.0222, "step": 75650 }, { "grad_norm": 0.14256404340267181, "learning_rate": 1.5342354264443566e-05, "loss": 0.0202, "step": 75660 }, { "grad_norm": 0.5938220620155334, "learning_rate": 1.533043810529411e-05, "loss": 0.0215, "step": 75670 }, { "grad_norm": 0.3184404969215393, "learning_rate": 1.5318525737555807e-05, "loss": 0.0223, "step": 75680 }, { "grad_norm": 0.09652987122535706, "learning_rate": 1.53066171625314e-05, "loss": 0.0138, "step": 75690 }, { "grad_norm": 0.14034971594810486, "learning_rate": 1.529471238152317e-05, "loss": 0.0196, "step": 75700 }, { "grad_norm": 0.24082478880882263, "learning_rate": 1.5282811395833023e-05, "loss": 0.014, "step": 75710 }, { "grad_norm": 0.15008871257305145, "learning_rate": 1.5270914206762397e-05, "loss": 0.0189, "step": 75720 }, { "grad_norm": 0.680070161819458, "learning_rate": 1.525902081561241e-05, "loss": 0.0212, "step": 75730 }, { "grad_norm": 0.2579296827316284, "learning_rate": 1.5247131223683664e-05, "loss": 0.0196, "step": 75740 }, { "grad_norm": 0.2449246346950531, "learning_rate": 1.5235245432276408e-05, "loss": 0.0177, "step": 75750 }, { "grad_norm": 0.18604670464992523, "learning_rate": 1.5223363442690424e-05, "loss": 0.0164, "step": 75760 }, { "grad_norm": 0.20744508504867554, "learning_rate": 1.5211485256225133e-05, "loss": 0.02, "step": 75770 }, { "grad_norm": 0.21733815968036652, "learning_rate": 1.519961087417951e-05, "loss": 0.0186, "step": 75780 }, { "grad_norm": 0.1471085399389267, "learning_rate": 1.5187740297852137e-05, "loss": 0.0186, "step": 75790 }, { "grad_norm": 0.3118530511856079, "learning_rate": 1.517587352854113e-05, "loss": 0.0155, "step": 75800 }, { "grad_norm": 0.29778164625167847, "learning_rate": 1.516401056754423e-05, "loss": 0.0184, "step": 75810 }, { "grad_norm": 0.299968957901001, "learning_rate": 1.5152151416158783e-05, "loss": 0.0177, "step": 75820 }, { "grad_norm": 0.16415053606033325, "learning_rate": 1.5140296075681642e-05, "loss": 0.0185, "step": 75830 }, { "grad_norm": 0.17541882395744324, "learning_rate": 1.5128444547409331e-05, "loss": 0.0134, "step": 75840 }, { "grad_norm": 0.25455036759376526, "learning_rate": 1.5116596832637864e-05, "loss": 0.0173, "step": 75850 }, { "grad_norm": 0.19407902657985687, "learning_rate": 1.5104752932662952e-05, "loss": 0.0108, "step": 75860 }, { "grad_norm": 0.5101727247238159, "learning_rate": 1.5092912848779778e-05, "loss": 0.0192, "step": 75870 }, { "grad_norm": 0.23112614452838898, "learning_rate": 1.5081076582283194e-05, "loss": 0.0144, "step": 75880 }, { "grad_norm": 1.3607077598571777, "learning_rate": 1.5069244134467553e-05, "loss": 0.0197, "step": 75890 }, { "grad_norm": 0.15011227130889893, "learning_rate": 1.5057415506626876e-05, "loss": 0.0136, "step": 75900 }, { "grad_norm": 0.26184386014938354, "learning_rate": 1.5045590700054668e-05, "loss": 0.0176, "step": 75910 }, { "grad_norm": 0.6484853625297546, "learning_rate": 1.5033769716044144e-05, "loss": 0.0228, "step": 75920 }, { "grad_norm": 0.20847512781620026, "learning_rate": 1.5021952555887975e-05, "loss": 0.0212, "step": 75930 }, { "grad_norm": 0.25037071108818054, "learning_rate": 1.50101392208785e-05, "loss": 0.0141, "step": 75940 }, { "grad_norm": 0.13778340816497803, "learning_rate": 1.4998329712307579e-05, "loss": 0.0162, "step": 75950 }, { "grad_norm": 0.17094898223876953, "learning_rate": 1.4986524031466708e-05, "loss": 0.0141, "step": 75960 }, { "grad_norm": 0.2045142650604248, "learning_rate": 1.4974722179646916e-05, "loss": 0.0175, "step": 75970 }, { "grad_norm": 0.239545077085495, "learning_rate": 1.4962924158138847e-05, "loss": 0.0144, "step": 75980 }, { "grad_norm": 0.10475389659404755, "learning_rate": 1.4951129968232714e-05, "loss": 0.0108, "step": 75990 }, { "grad_norm": 0.18274544179439545, "learning_rate": 1.4939339611218329e-05, "loss": 0.012, "step": 76000 }, { "grad_norm": 0.2666427195072174, "learning_rate": 1.4927553088385032e-05, "loss": 0.0217, "step": 76010 }, { "grad_norm": 0.17745783925056458, "learning_rate": 1.4915770401021801e-05, "loss": 0.0128, "step": 76020 }, { "grad_norm": 0.15077608823776245, "learning_rate": 1.4903991550417185e-05, "loss": 0.0126, "step": 76030 }, { "grad_norm": 0.22931358218193054, "learning_rate": 1.4892216537859277e-05, "loss": 0.0133, "step": 76040 }, { "grad_norm": 0.2762255370616913, "learning_rate": 1.4880445364635775e-05, "loss": 0.0161, "step": 76050 }, { "grad_norm": 0.21556183695793152, "learning_rate": 1.4868678032033973e-05, "loss": 0.0157, "step": 76060 }, { "grad_norm": 0.15110842883586884, "learning_rate": 1.4856914541340739e-05, "loss": 0.0132, "step": 76070 }, { "grad_norm": 0.2433263212442398, "learning_rate": 1.4845154893842472e-05, "loss": 0.0193, "step": 76080 }, { "grad_norm": 0.2219780683517456, "learning_rate": 1.4833399090825234e-05, "loss": 0.017, "step": 76090 }, { "grad_norm": 0.20007899403572083, "learning_rate": 1.4821647133574562e-05, "loss": 0.0147, "step": 76100 }, { "grad_norm": 0.20015154778957367, "learning_rate": 1.4809899023375706e-05, "loss": 0.0155, "step": 76110 }, { "grad_norm": 0.41871634125709534, "learning_rate": 1.4798154761513366e-05, "loss": 0.0138, "step": 76120 }, { "grad_norm": 0.2041018009185791, "learning_rate": 1.4786414349271916e-05, "loss": 0.0153, "step": 76130 }, { "grad_norm": 0.1722695529460907, "learning_rate": 1.4774677787935226e-05, "loss": 0.0118, "step": 76140 }, { "grad_norm": 0.2852577865123749, "learning_rate": 1.4762945078786838e-05, "loss": 0.0153, "step": 76150 }, { "grad_norm": 0.13767434656620026, "learning_rate": 1.4751216223109777e-05, "loss": 0.0201, "step": 76160 }, { "grad_norm": 0.2901448905467987, "learning_rate": 1.4739491222186714e-05, "loss": 0.0193, "step": 76170 }, { "grad_norm": 0.18319210410118103, "learning_rate": 1.4727770077299879e-05, "loss": 0.014, "step": 76180 }, { "grad_norm": 0.14819003641605377, "learning_rate": 1.4716052789731094e-05, "loss": 0.0115, "step": 76190 }, { "grad_norm": 0.17891696095466614, "learning_rate": 1.470433936076171e-05, "loss": 0.019, "step": 76200 }, { "grad_norm": 0.16732552647590637, "learning_rate": 1.4692629791672719e-05, "loss": 0.0132, "step": 76210 }, { "grad_norm": 0.1263428032398224, "learning_rate": 1.468092408374464e-05, "loss": 0.0147, "step": 76220 }, { "grad_norm": 0.4782952666282654, "learning_rate": 1.4669222238257602e-05, "loss": 0.0183, "step": 76230 }, { "grad_norm": 0.19349652528762817, "learning_rate": 1.4657524256491301e-05, "loss": 0.0188, "step": 76240 }, { "grad_norm": 0.335015207529068, "learning_rate": 1.464583013972502e-05, "loss": 0.018, "step": 76250 }, { "grad_norm": 0.17610634863376617, "learning_rate": 1.4634139889237591e-05, "loss": 0.0167, "step": 76260 }, { "grad_norm": 0.14965982735157013, "learning_rate": 1.4622453506307448e-05, "loss": 0.0155, "step": 76270 }, { "grad_norm": 0.24375218152999878, "learning_rate": 1.461077099221262e-05, "loss": 0.019, "step": 76280 }, { "grad_norm": 0.17339584231376648, "learning_rate": 1.4599092348230652e-05, "loss": 0.0131, "step": 76290 }, { "grad_norm": 0.28319352865219116, "learning_rate": 1.4587417575638724e-05, "loss": 0.0178, "step": 76300 }, { "grad_norm": 0.1840561330318451, "learning_rate": 1.4575746675713563e-05, "loss": 0.015, "step": 76310 }, { "grad_norm": 0.17657674849033356, "learning_rate": 1.4564079649731504e-05, "loss": 0.0147, "step": 76320 }, { "grad_norm": 0.16464577615261078, "learning_rate": 1.4552416498968402e-05, "loss": 0.01, "step": 76330 }, { "grad_norm": 0.2436913400888443, "learning_rate": 1.4540757224699747e-05, "loss": 0.0172, "step": 76340 }, { "grad_norm": 0.24841633439064026, "learning_rate": 1.452910182820056e-05, "loss": 0.0201, "step": 76350 }, { "grad_norm": 0.2117973417043686, "learning_rate": 1.4517450310745468e-05, "loss": 0.0151, "step": 76360 }, { "grad_norm": 0.26116320490837097, "learning_rate": 1.450580267360866e-05, "loss": 0.0176, "step": 76370 }, { "grad_norm": 0.21575729548931122, "learning_rate": 1.4494158918063922e-05, "loss": 0.014, "step": 76380 }, { "grad_norm": 0.1523890346288681, "learning_rate": 1.4482519045384568e-05, "loss": 0.0145, "step": 76390 }, { "grad_norm": 0.1750343292951584, "learning_rate": 1.4470883056843548e-05, "loss": 0.0207, "step": 76400 }, { "grad_norm": 0.3182447850704193, "learning_rate": 1.4459250953713322e-05, "loss": 0.0185, "step": 76410 }, { "grad_norm": 0.1933072954416275, "learning_rate": 1.444762273726598e-05, "loss": 0.0162, "step": 76420 }, { "grad_norm": 0.17104780673980713, "learning_rate": 1.443599840877316e-05, "loss": 0.0134, "step": 76430 }, { "grad_norm": 0.23355281352996826, "learning_rate": 1.4424377969506103e-05, "loss": 0.0148, "step": 76440 }, { "grad_norm": 0.12261898070573807, "learning_rate": 1.4412761420735565e-05, "loss": 0.0137, "step": 76450 }, { "grad_norm": 0.23964224755764008, "learning_rate": 1.4401148763731953e-05, "loss": 0.0173, "step": 76460 }, { "grad_norm": 0.3101727366447449, "learning_rate": 1.4389539999765173e-05, "loss": 0.0175, "step": 76470 }, { "grad_norm": 0.14359936118125916, "learning_rate": 1.4377935130104758e-05, "loss": 0.0136, "step": 76480 }, { "grad_norm": 0.37998977303504944, "learning_rate": 1.4366334156019801e-05, "loss": 0.0223, "step": 76490 }, { "grad_norm": 0.2742574214935303, "learning_rate": 1.435473707877898e-05, "loss": 0.0153, "step": 76500 }, { "grad_norm": 0.22661292552947998, "learning_rate": 1.4343143899650508e-05, "loss": 0.0163, "step": 76510 }, { "grad_norm": 0.18802955746650696, "learning_rate": 1.4331554619902204e-05, "loss": 0.0154, "step": 76520 }, { "grad_norm": 0.23606127500534058, "learning_rate": 1.4319969240801484e-05, "loss": 0.022, "step": 76530 }, { "grad_norm": 0.2756166458129883, "learning_rate": 1.430838776361526e-05, "loss": 0.0156, "step": 76540 }, { "grad_norm": 0.2658686637878418, "learning_rate": 1.4296810189610093e-05, "loss": 0.0176, "step": 76550 }, { "grad_norm": 0.22464509308338165, "learning_rate": 1.4285236520052087e-05, "loss": 0.0149, "step": 76560 }, { "grad_norm": 0.17711211740970612, "learning_rate": 1.4273666756206933e-05, "loss": 0.0153, "step": 76570 }, { "grad_norm": 0.9267430901527405, "learning_rate": 1.4262100899339848e-05, "loss": 0.0174, "step": 76580 }, { "grad_norm": 0.24753791093826294, "learning_rate": 1.4250538950715696e-05, "loss": 0.0146, "step": 76590 }, { "grad_norm": 0.20784235000610352, "learning_rate": 1.4238980911598842e-05, "loss": 0.0157, "step": 76600 }, { "grad_norm": 0.30829477310180664, "learning_rate": 1.4227426783253267e-05, "loss": 0.0171, "step": 76610 }, { "grad_norm": 0.2211325466632843, "learning_rate": 1.4215876566942521e-05, "loss": 0.0161, "step": 76620 }, { "grad_norm": 0.2241465002298355, "learning_rate": 1.4204330263929721e-05, "loss": 0.0175, "step": 76630 }, { "grad_norm": 0.5639873743057251, "learning_rate": 1.4192787875477537e-05, "loss": 0.0199, "step": 76640 }, { "grad_norm": 0.11911359429359436, "learning_rate": 1.4181249402848246e-05, "loss": 0.0127, "step": 76650 }, { "grad_norm": 0.22042839229106903, "learning_rate": 1.4169714847303656e-05, "loss": 0.0205, "step": 76660 }, { "grad_norm": 0.17047782242298126, "learning_rate": 1.4158184210105197e-05, "loss": 0.0206, "step": 76670 }, { "grad_norm": 0.279764324426651, "learning_rate": 1.4146657492513787e-05, "loss": 0.0198, "step": 76680 }, { "grad_norm": 0.6185944676399231, "learning_rate": 1.4135134695790047e-05, "loss": 0.0129, "step": 76690 }, { "grad_norm": 0.3162998855113983, "learning_rate": 1.4123615821194037e-05, "loss": 0.0185, "step": 76700 }, { "grad_norm": 0.4040246307849884, "learning_rate": 1.4112100869985473e-05, "loss": 0.0188, "step": 76710 }, { "grad_norm": 0.22877097129821777, "learning_rate": 1.4100589843423589e-05, "loss": 0.0139, "step": 76720 }, { "grad_norm": 0.2500668466091156, "learning_rate": 1.408908274276723e-05, "loss": 0.0175, "step": 76730 }, { "grad_norm": 0.20851828157901764, "learning_rate": 1.4077579569274786e-05, "loss": 0.0179, "step": 76740 }, { "grad_norm": 0.20023395121097565, "learning_rate": 1.4066080324204245e-05, "loss": 0.0291, "step": 76750 }, { "grad_norm": 0.3482038080692291, "learning_rate": 1.405458500881312e-05, "loss": 0.0183, "step": 76760 }, { "grad_norm": 0.5908530950546265, "learning_rate": 1.4043093624358538e-05, "loss": 0.0154, "step": 76770 }, { "grad_norm": 0.16047848761081696, "learning_rate": 1.4031606172097189e-05, "loss": 0.0165, "step": 76780 }, { "grad_norm": 0.20422545075416565, "learning_rate": 1.40201226532853e-05, "loss": 0.0188, "step": 76790 }, { "grad_norm": 0.28089600801467896, "learning_rate": 1.4008643069178701e-05, "loss": 0.0171, "step": 76800 }, { "grad_norm": 0.13981357216835022, "learning_rate": 1.399716742103278e-05, "loss": 0.0133, "step": 76810 }, { "grad_norm": 0.2458159327507019, "learning_rate": 1.398569571010252e-05, "loss": 0.0152, "step": 76820 }, { "grad_norm": 0.2578922510147095, "learning_rate": 1.3974227937642414e-05, "loss": 0.021, "step": 76830 }, { "grad_norm": 0.1398237943649292, "learning_rate": 1.3962764104906595e-05, "loss": 0.0138, "step": 76840 }, { "grad_norm": 0.2700616419315338, "learning_rate": 1.3951304213148696e-05, "loss": 0.0178, "step": 76850 }, { "grad_norm": 0.2152814120054245, "learning_rate": 1.3939848263621979e-05, "loss": 0.0152, "step": 76860 }, { "grad_norm": 0.33674585819244385, "learning_rate": 1.3928396257579219e-05, "loss": 0.015, "step": 76870 }, { "grad_norm": 0.3591702878475189, "learning_rate": 1.3916948196272833e-05, "loss": 0.0219, "step": 76880 }, { "grad_norm": 0.2304280698299408, "learning_rate": 1.3905504080954734e-05, "loss": 0.0217, "step": 76890 }, { "grad_norm": 0.19207978248596191, "learning_rate": 1.3894063912876454e-05, "loss": 0.0151, "step": 76900 }, { "grad_norm": 0.23024839162826538, "learning_rate": 1.3882627693289046e-05, "loss": 0.0167, "step": 76910 }, { "grad_norm": 0.17265702784061432, "learning_rate": 1.3871195423443183e-05, "loss": 0.0184, "step": 76920 }, { "grad_norm": 0.2538512051105499, "learning_rate": 1.3859767104589044e-05, "loss": 0.019, "step": 76930 }, { "grad_norm": 0.20243358612060547, "learning_rate": 1.3848342737976467e-05, "loss": 0.0139, "step": 76940 }, { "grad_norm": 0.3741148114204407, "learning_rate": 1.3836922324854757e-05, "loss": 0.0159, "step": 76950 }, { "grad_norm": 0.37382054328918457, "learning_rate": 1.382550586647287e-05, "loss": 0.0161, "step": 76960 }, { "grad_norm": 0.1402835100889206, "learning_rate": 1.3814093364079257e-05, "loss": 0.0172, "step": 76970 }, { "grad_norm": 0.2766039967536926, "learning_rate": 1.3802684818921991e-05, "loss": 0.0176, "step": 76980 }, { "grad_norm": 0.33993130922317505, "learning_rate": 1.3791280232248687e-05, "loss": 0.0154, "step": 76990 }, { "grad_norm": 0.22540000081062317, "learning_rate": 1.3779879605306555e-05, "loss": 0.0169, "step": 77000 }, { "grad_norm": 0.2039317786693573, "learning_rate": 1.376848293934232e-05, "loss": 0.0142, "step": 77010 }, { "grad_norm": 0.20595847070217133, "learning_rate": 1.3757090235602316e-05, "loss": 0.0173, "step": 77020 }, { "grad_norm": 0.21852803230285645, "learning_rate": 1.3745701495332447e-05, "loss": 0.0158, "step": 77030 }, { "grad_norm": 0.14324745535850525, "learning_rate": 1.3734316719778145e-05, "loss": 0.0153, "step": 77040 }, { "grad_norm": 0.2565140724182129, "learning_rate": 1.3722935910184454e-05, "loss": 0.0206, "step": 77050 }, { "grad_norm": 0.25810620188713074, "learning_rate": 1.3711559067795921e-05, "loss": 0.0173, "step": 77060 }, { "grad_norm": 0.17772147059440613, "learning_rate": 1.3700186193856756e-05, "loss": 0.0112, "step": 77070 }, { "grad_norm": 0.2612544894218445, "learning_rate": 1.368881728961064e-05, "loss": 0.0156, "step": 77080 }, { "grad_norm": 0.21321505308151245, "learning_rate": 1.3677452356300885e-05, "loss": 0.0145, "step": 77090 }, { "grad_norm": 0.32633936405181885, "learning_rate": 1.3666091395170321e-05, "loss": 0.0183, "step": 77100 }, { "grad_norm": 0.2577025890350342, "learning_rate": 1.3654734407461384e-05, "loss": 0.0167, "step": 77110 }, { "grad_norm": 0.3846037983894348, "learning_rate": 1.3643381394416022e-05, "loss": 0.0168, "step": 77120 }, { "grad_norm": 0.2795707583427429, "learning_rate": 1.3632032357275838e-05, "loss": 0.02, "step": 77130 }, { "grad_norm": 0.20912706851959229, "learning_rate": 1.3620687297281898e-05, "loss": 0.0153, "step": 77140 }, { "grad_norm": 0.19980770349502563, "learning_rate": 1.3609346215674917e-05, "loss": 0.0128, "step": 77150 }, { "grad_norm": 0.2470492273569107, "learning_rate": 1.3598009113695109e-05, "loss": 0.0185, "step": 77160 }, { "grad_norm": 0.14377455413341522, "learning_rate": 1.358667599258231e-05, "loss": 0.0113, "step": 77170 }, { "grad_norm": 0.19750818610191345, "learning_rate": 1.357534685357586e-05, "loss": 0.0141, "step": 77180 }, { "grad_norm": 0.17876410484313965, "learning_rate": 1.3564021697914715e-05, "loss": 0.0127, "step": 77190 }, { "grad_norm": 0.22932091355323792, "learning_rate": 1.3552700526837381e-05, "loss": 0.0155, "step": 77200 }, { "grad_norm": 0.3642025887966156, "learning_rate": 1.3541383341581925e-05, "loss": 0.0156, "step": 77210 }, { "grad_norm": 0.2507172226905823, "learning_rate": 1.3530070143385965e-05, "loss": 0.0164, "step": 77220 }, { "grad_norm": 0.24900434911251068, "learning_rate": 1.3518760933486702e-05, "loss": 0.0156, "step": 77230 }, { "grad_norm": 0.14507481455802917, "learning_rate": 1.3507455713120908e-05, "loss": 0.0168, "step": 77240 }, { "grad_norm": 0.17248652875423431, "learning_rate": 1.3496154483524876e-05, "loss": 0.0159, "step": 77250 }, { "grad_norm": 0.23557467758655548, "learning_rate": 1.3484857245934506e-05, "loss": 0.0131, "step": 77260 }, { "grad_norm": 0.19915911555290222, "learning_rate": 1.3473564001585248e-05, "loss": 0.017, "step": 77270 }, { "grad_norm": 0.19917048513889313, "learning_rate": 1.3462274751712129e-05, "loss": 0.0154, "step": 77280 }, { "grad_norm": 0.21364009380340576, "learning_rate": 1.3450989497549698e-05, "loss": 0.0159, "step": 77290 }, { "grad_norm": 0.29161885380744934, "learning_rate": 1.3439708240332115e-05, "loss": 0.0175, "step": 77300 }, { "grad_norm": 0.20251218974590302, "learning_rate": 1.3428430981293043e-05, "loss": 0.013, "step": 77310 }, { "grad_norm": 0.32551029324531555, "learning_rate": 1.3417157721665808e-05, "loss": 0.0151, "step": 77320 }, { "grad_norm": 0.2543564438819885, "learning_rate": 1.3405888462683186e-05, "loss": 0.012, "step": 77330 }, { "grad_norm": 0.4673227071762085, "learning_rate": 1.3394623205577595e-05, "loss": 0.0212, "step": 77340 }, { "grad_norm": 0.28782373666763306, "learning_rate": 1.3383361951580963e-05, "loss": 0.0223, "step": 77350 }, { "grad_norm": 0.27711114287376404, "learning_rate": 1.337210470192483e-05, "loss": 0.0141, "step": 77360 }, { "grad_norm": 0.1735595315694809, "learning_rate": 1.3360851457840245e-05, "loss": 0.0137, "step": 77370 }, { "grad_norm": 0.1771293729543686, "learning_rate": 1.3349602220557856e-05, "loss": 0.0146, "step": 77380 }, { "grad_norm": 0.3319433629512787, "learning_rate": 1.3338356991307865e-05, "loss": 0.018, "step": 77390 }, { "grad_norm": 0.3650157153606415, "learning_rate": 1.3327115771320048e-05, "loss": 0.0164, "step": 77400 }, { "grad_norm": 0.3118187487125397, "learning_rate": 1.3315878561823697e-05, "loss": 0.0149, "step": 77410 }, { "grad_norm": 0.278306245803833, "learning_rate": 1.330464536404773e-05, "loss": 0.0148, "step": 77420 }, { "grad_norm": 0.2597675919532776, "learning_rate": 1.3293416179220563e-05, "loss": 0.0204, "step": 77430 }, { "grad_norm": 0.2030450701713562, "learning_rate": 1.328219100857021e-05, "loss": 0.0125, "step": 77440 }, { "grad_norm": 0.19448235630989075, "learning_rate": 1.327096985332425e-05, "loss": 0.0181, "step": 77450 }, { "grad_norm": 0.2532019317150116, "learning_rate": 1.3259752714709816e-05, "loss": 0.0146, "step": 77460 }, { "grad_norm": 0.21288080513477325, "learning_rate": 1.3248539593953569e-05, "loss": 0.0136, "step": 77470 }, { "grad_norm": 0.2219868004322052, "learning_rate": 1.323733049228178e-05, "loss": 0.0127, "step": 77480 }, { "grad_norm": 0.1675061732530594, "learning_rate": 1.3226125410920276e-05, "loss": 0.0139, "step": 77490 }, { "grad_norm": 0.14519616961479187, "learning_rate": 1.3214924351094388e-05, "loss": 0.0168, "step": 77500 }, { "grad_norm": 0.25768059492111206, "learning_rate": 1.3203727314029069e-05, "loss": 0.0143, "step": 77510 }, { "grad_norm": 0.2155086100101471, "learning_rate": 1.3192534300948811e-05, "loss": 0.0154, "step": 77520 }, { "grad_norm": 0.1346374750137329, "learning_rate": 1.3181345313077675e-05, "loss": 0.0159, "step": 77530 }, { "grad_norm": 0.12990540266036987, "learning_rate": 1.3170160351639243e-05, "loss": 0.015, "step": 77540 }, { "grad_norm": 0.22355349361896515, "learning_rate": 1.3158979417856721e-05, "loss": 0.0209, "step": 77550 }, { "grad_norm": 0.13705690205097198, "learning_rate": 1.31478025129528e-05, "loss": 0.0142, "step": 77560 }, { "grad_norm": 0.2154603898525238, "learning_rate": 1.3136629638149794e-05, "loss": 0.0135, "step": 77570 }, { "grad_norm": 0.21426910161972046, "learning_rate": 1.3125460794669547e-05, "loss": 0.0175, "step": 77580 }, { "grad_norm": 0.17118355631828308, "learning_rate": 1.3114295983733483e-05, "loss": 0.017, "step": 77590 }, { "grad_norm": 0.19904860854148865, "learning_rate": 1.3103135206562534e-05, "loss": 0.0144, "step": 77600 }, { "grad_norm": 0.209694966673851, "learning_rate": 1.3091978464377264e-05, "loss": 0.0146, "step": 77610 }, { "grad_norm": 0.2918602228164673, "learning_rate": 1.3080825758397724e-05, "loss": 0.0116, "step": 77620 }, { "grad_norm": 0.4211743474006653, "learning_rate": 1.3069677089843568e-05, "loss": 0.0129, "step": 77630 }, { "grad_norm": 0.34682634472846985, "learning_rate": 1.305853245993401e-05, "loss": 0.0198, "step": 77640 }, { "grad_norm": 0.1959245800971985, "learning_rate": 1.3047391869887809e-05, "loss": 0.0158, "step": 77650 }, { "grad_norm": 0.1417206972837448, "learning_rate": 1.3036255320923263e-05, "loss": 0.0186, "step": 77660 }, { "grad_norm": 0.2777811884880066, "learning_rate": 1.3025122814258273e-05, "loss": 0.0108, "step": 77670 }, { "grad_norm": 0.31111183762550354, "learning_rate": 1.3013994351110249e-05, "loss": 0.0208, "step": 77680 }, { "grad_norm": 0.5659412145614624, "learning_rate": 1.3002869932696194e-05, "loss": 0.0175, "step": 77690 }, { "grad_norm": 0.15135198831558228, "learning_rate": 1.2991749560232664e-05, "loss": 0.0122, "step": 77700 }, { "grad_norm": 0.38487303256988525, "learning_rate": 1.2980633234935768e-05, "loss": 0.0163, "step": 77710 }, { "grad_norm": 0.1675254851579666, "learning_rate": 1.2969520958021153e-05, "loss": 0.0162, "step": 77720 }, { "grad_norm": 0.33238154649734497, "learning_rate": 1.295841273070405e-05, "loss": 0.015, "step": 77730 }, { "grad_norm": 0.15080536901950836, "learning_rate": 1.2947308554199255e-05, "loss": 0.0205, "step": 77740 }, { "grad_norm": 0.34967362880706787, "learning_rate": 1.2936208429721075e-05, "loss": 0.0246, "step": 77750 }, { "grad_norm": 0.3433946669101715, "learning_rate": 1.2925112358483416e-05, "loss": 0.0136, "step": 77760 }, { "grad_norm": 0.248549222946167, "learning_rate": 1.2914020341699728e-05, "loss": 0.0184, "step": 77770 }, { "grad_norm": 0.1638142615556717, "learning_rate": 1.2902932380583032e-05, "loss": 0.0146, "step": 77780 }, { "grad_norm": 0.1598559468984604, "learning_rate": 1.2891848476345863e-05, "loss": 0.0134, "step": 77790 }, { "grad_norm": 0.2208585888147354, "learning_rate": 1.2880768630200368e-05, "loss": 0.018, "step": 77800 }, { "grad_norm": 0.2628882825374603, "learning_rate": 1.286969284335819e-05, "loss": 0.0135, "step": 77810 }, { "grad_norm": 0.21187527477741241, "learning_rate": 1.2858621117030578e-05, "loss": 0.0174, "step": 77820 }, { "grad_norm": 0.23772242665290833, "learning_rate": 1.2847553452428324e-05, "loss": 0.0115, "step": 77830 }, { "grad_norm": 0.15317393839359283, "learning_rate": 1.283648985076178e-05, "loss": 0.016, "step": 77840 }, { "grad_norm": 0.1645827293395996, "learning_rate": 1.2825430313240817e-05, "loss": 0.0188, "step": 77850 }, { "grad_norm": 0.3699178099632263, "learning_rate": 1.281437484107491e-05, "loss": 0.0208, "step": 77860 }, { "grad_norm": 0.15091508626937866, "learning_rate": 1.2803323435473058e-05, "loss": 0.0155, "step": 77870 }, { "grad_norm": 0.23875749111175537, "learning_rate": 1.2792276097643841e-05, "loss": 0.015, "step": 77880 }, { "grad_norm": 0.2423119843006134, "learning_rate": 1.2781232828795342e-05, "loss": 0.0147, "step": 77890 }, { "grad_norm": 0.22813042998313904, "learning_rate": 1.2770193630135292e-05, "loss": 0.0202, "step": 77900 }, { "grad_norm": 0.25964951515197754, "learning_rate": 1.2759158502870877e-05, "loss": 0.0268, "step": 77910 }, { "grad_norm": 0.1748582273721695, "learning_rate": 1.2748127448208913e-05, "loss": 0.0125, "step": 77920 }, { "grad_norm": 0.21074603497982025, "learning_rate": 1.2737100467355706e-05, "loss": 0.0175, "step": 77930 }, { "grad_norm": 0.24356867372989655, "learning_rate": 1.2726077561517169e-05, "loss": 0.0176, "step": 77940 }, { "grad_norm": 0.17818637192249298, "learning_rate": 1.2715058731898755e-05, "loss": 0.0114, "step": 77950 }, { "grad_norm": 0.24639996886253357, "learning_rate": 1.2704043979705471e-05, "loss": 0.0129, "step": 77960 }, { "grad_norm": 0.3651498258113861, "learning_rate": 1.269303330614185e-05, "loss": 0.0149, "step": 77970 }, { "grad_norm": 0.18725743889808655, "learning_rate": 1.2682026712412015e-05, "loss": 0.0192, "step": 77980 }, { "grad_norm": 0.14273157715797424, "learning_rate": 1.2671024199719644e-05, "loss": 0.011, "step": 77990 }, { "grad_norm": 0.29093918204307556, "learning_rate": 1.2660025769267936e-05, "loss": 0.0207, "step": 78000 }, { "grad_norm": 0.3712780475616455, "learning_rate": 1.2649031422259667e-05, "loss": 0.0184, "step": 78010 }, { "grad_norm": 0.2639889419078827, "learning_rate": 1.2638041159897163e-05, "loss": 0.0179, "step": 78020 }, { "grad_norm": 0.15766392648220062, "learning_rate": 1.262705498338232e-05, "loss": 0.0151, "step": 78030 }, { "grad_norm": 0.23328109085559845, "learning_rate": 1.2616072893916542e-05, "loss": 0.0179, "step": 78040 }, { "grad_norm": 0.4245861768722534, "learning_rate": 1.2605094892700836e-05, "loss": 0.0157, "step": 78050 }, { "grad_norm": 0.092173732817173, "learning_rate": 1.259412098093572e-05, "loss": 0.0145, "step": 78060 }, { "grad_norm": 0.21745672821998596, "learning_rate": 1.258315115982131e-05, "loss": 0.013, "step": 78070 }, { "grad_norm": 0.18036122620105743, "learning_rate": 1.2572185430557203e-05, "loss": 0.0141, "step": 78080 }, { "grad_norm": 0.18073594570159912, "learning_rate": 1.2561223794342659e-05, "loss": 0.0153, "step": 78090 }, { "grad_norm": 0.2721760869026184, "learning_rate": 1.255026625237637e-05, "loss": 0.0245, "step": 78100 }, { "grad_norm": 0.3518787920475006, "learning_rate": 1.2539312805856678e-05, "loss": 0.0175, "step": 78110 }, { "grad_norm": 0.41115206480026245, "learning_rate": 1.25283634559814e-05, "loss": 0.0187, "step": 78120 }, { "grad_norm": 0.14028021693229675, "learning_rate": 1.2517418203947972e-05, "loss": 0.017, "step": 78130 }, { "grad_norm": 0.20585690438747406, "learning_rate": 1.2506477050953308e-05, "loss": 0.0168, "step": 78140 }, { "grad_norm": 0.19669188559055328, "learning_rate": 1.2495539998193972e-05, "loss": 0.0191, "step": 78150 }, { "grad_norm": 0.20962512493133545, "learning_rate": 1.2484607046865975e-05, "loss": 0.0242, "step": 78160 }, { "grad_norm": 0.2611388564109802, "learning_rate": 1.2473678198164968e-05, "loss": 0.0154, "step": 78170 }, { "grad_norm": 0.2795366942882538, "learning_rate": 1.2462753453286075e-05, "loss": 0.0158, "step": 78180 }, { "grad_norm": 0.21800681948661804, "learning_rate": 1.2451832813424031e-05, "loss": 0.0238, "step": 78190 }, { "grad_norm": 0.3400343358516693, "learning_rate": 1.2440916279773095e-05, "loss": 0.0199, "step": 78200 }, { "grad_norm": 0.1709802895784378, "learning_rate": 1.2430003853527101e-05, "loss": 0.0181, "step": 78210 }, { "grad_norm": 0.566843569278717, "learning_rate": 1.2419095535879383e-05, "loss": 0.0239, "step": 78220 }, { "grad_norm": 0.08456382900476456, "learning_rate": 1.2408191328022878e-05, "loss": 0.0148, "step": 78230 }, { "grad_norm": 0.2739526331424713, "learning_rate": 1.2397291231150066e-05, "loss": 0.0147, "step": 78240 }, { "grad_norm": 0.15994767844676971, "learning_rate": 1.2386395246452937e-05, "loss": 0.0182, "step": 78250 }, { "grad_norm": 0.22540314495563507, "learning_rate": 1.2375503375123083e-05, "loss": 0.0142, "step": 78260 }, { "grad_norm": 0.5517428517341614, "learning_rate": 1.2364615618351583e-05, "loss": 0.0143, "step": 78270 }, { "grad_norm": 0.20273196697235107, "learning_rate": 1.2353731977329169e-05, "loss": 0.0136, "step": 78280 }, { "grad_norm": 0.2421232908964157, "learning_rate": 1.2342852453246007e-05, "loss": 0.0104, "step": 78290 }, { "grad_norm": 0.34634828567504883, "learning_rate": 1.2331977047291898e-05, "loss": 0.0238, "step": 78300 }, { "grad_norm": 0.2427489161491394, "learning_rate": 1.2321105760656137e-05, "loss": 0.0166, "step": 78310 }, { "grad_norm": 0.19614067673683167, "learning_rate": 1.2310238594527613e-05, "loss": 0.0153, "step": 78320 }, { "grad_norm": 0.3606569468975067, "learning_rate": 1.2299375550094693e-05, "loss": 0.0124, "step": 78330 }, { "grad_norm": 0.22169098258018494, "learning_rate": 1.2288516628545421e-05, "loss": 0.0133, "step": 78340 }, { "grad_norm": 0.4733874499797821, "learning_rate": 1.2277661831067255e-05, "loss": 0.0208, "step": 78350 }, { "grad_norm": 0.19029463827610016, "learning_rate": 1.2266811158847286e-05, "loss": 0.016, "step": 78360 }, { "grad_norm": 0.20691251754760742, "learning_rate": 1.2255964613072107e-05, "loss": 0.014, "step": 78370 }, { "grad_norm": 0.3224846124649048, "learning_rate": 1.2245122194927905e-05, "loss": 0.0134, "step": 78380 }, { "grad_norm": 0.23965975642204285, "learning_rate": 1.2234283905600364e-05, "loss": 0.0156, "step": 78390 }, { "grad_norm": 0.1750614047050476, "learning_rate": 1.222344974627475e-05, "loss": 0.0157, "step": 78400 }, { "grad_norm": 0.1332535743713379, "learning_rate": 1.2212619718135875e-05, "loss": 0.0139, "step": 78410 }, { "grad_norm": 0.2534009516239166, "learning_rate": 1.2201793822368118e-05, "loss": 0.015, "step": 78420 }, { "grad_norm": 0.2961881160736084, "learning_rate": 1.2190972060155337e-05, "loss": 0.0115, "step": 78430 }, { "grad_norm": 0.15551774203777313, "learning_rate": 1.2180154432681013e-05, "loss": 0.0127, "step": 78440 }, { "grad_norm": 0.2741302251815796, "learning_rate": 1.2169340941128148e-05, "loss": 0.0174, "step": 78450 }, { "grad_norm": 0.18654771149158478, "learning_rate": 1.2158531586679267e-05, "loss": 0.0144, "step": 78460 }, { "grad_norm": 0.3032861053943634, "learning_rate": 1.2147726370516477e-05, "loss": 0.0132, "step": 78470 }, { "grad_norm": 0.16926978528499603, "learning_rate": 1.2136925293821422e-05, "loss": 0.016, "step": 78480 }, { "grad_norm": 0.3459644615650177, "learning_rate": 1.2126128357775306e-05, "loss": 0.0184, "step": 78490 }, { "grad_norm": 0.44602134823799133, "learning_rate": 1.2115335563558838e-05, "loss": 0.0168, "step": 78500 }, { "grad_norm": 0.14590542018413544, "learning_rate": 1.2104546912352327e-05, "loss": 0.0137, "step": 78510 }, { "grad_norm": 0.23876668512821198, "learning_rate": 1.2093762405335557e-05, "loss": 0.0134, "step": 78520 }, { "grad_norm": 0.2792809009552002, "learning_rate": 1.2082982043687974e-05, "loss": 0.0142, "step": 78530 }, { "grad_norm": 0.1547694057226181, "learning_rate": 1.207220582858845e-05, "loss": 0.0128, "step": 78540 }, { "grad_norm": 0.15939365327358246, "learning_rate": 1.206143376121549e-05, "loss": 0.0142, "step": 78550 }, { "grad_norm": 0.1515609174966812, "learning_rate": 1.2050665842747078e-05, "loss": 0.0126, "step": 78560 }, { "grad_norm": 0.1844908744096756, "learning_rate": 1.2039902074360804e-05, "loss": 0.0098, "step": 78570 }, { "grad_norm": 0.17757265269756317, "learning_rate": 1.2029142457233755e-05, "loss": 0.0187, "step": 78580 }, { "grad_norm": 0.2568035423755646, "learning_rate": 1.2018386992542601e-05, "loss": 0.0154, "step": 78590 }, { "grad_norm": 0.24477848410606384, "learning_rate": 1.2007635681463542e-05, "loss": 0.0147, "step": 78600 }, { "grad_norm": 0.2507912516593933, "learning_rate": 1.1996888525172334e-05, "loss": 0.0183, "step": 78610 }, { "grad_norm": 0.274374395608902, "learning_rate": 1.198614552484425e-05, "loss": 0.0167, "step": 78620 }, { "grad_norm": 0.2661822438240051, "learning_rate": 1.197540668165415e-05, "loss": 0.0144, "step": 78630 }, { "grad_norm": 0.3525776267051697, "learning_rate": 1.1964671996776395e-05, "loss": 0.0189, "step": 78640 }, { "grad_norm": 0.22444410622119904, "learning_rate": 1.1953941471384922e-05, "loss": 0.0169, "step": 78650 }, { "grad_norm": 0.17045411467552185, "learning_rate": 1.1943215106653205e-05, "loss": 0.016, "step": 78660 }, { "grad_norm": 0.42505770921707153, "learning_rate": 1.193249290375429e-05, "loss": 0.0168, "step": 78670 }, { "grad_norm": 0.22465072572231293, "learning_rate": 1.1921774863860696e-05, "loss": 0.013, "step": 78680 }, { "grad_norm": 0.4006812870502472, "learning_rate": 1.1911060988144556e-05, "loss": 0.0165, "step": 78690 }, { "grad_norm": 0.14104534685611725, "learning_rate": 1.1900351277777533e-05, "loss": 0.0117, "step": 78700 }, { "grad_norm": 0.22997158765792847, "learning_rate": 1.1889645733930798e-05, "loss": 0.0156, "step": 78710 }, { "grad_norm": 0.2270863950252533, "learning_rate": 1.1878944357775107e-05, "loss": 0.0123, "step": 78720 }, { "grad_norm": 0.1619345247745514, "learning_rate": 1.1868247150480749e-05, "loss": 0.0167, "step": 78730 }, { "grad_norm": 0.2077779918909073, "learning_rate": 1.1857554113217567e-05, "loss": 0.0135, "step": 78740 }, { "grad_norm": 0.19039179384708405, "learning_rate": 1.1846865247154903e-05, "loss": 0.0185, "step": 78750 }, { "grad_norm": 0.1948879063129425, "learning_rate": 1.1836180553461711e-05, "loss": 0.0202, "step": 78760 }, { "grad_norm": 0.20185783505439758, "learning_rate": 1.1825500033306419e-05, "loss": 0.0233, "step": 78770 }, { "grad_norm": 0.14803485572338104, "learning_rate": 1.1814823687857052e-05, "loss": 0.0102, "step": 78780 }, { "grad_norm": 0.3273865878582001, "learning_rate": 1.1804151518281153e-05, "loss": 0.0131, "step": 78790 }, { "grad_norm": 0.19399511814117432, "learning_rate": 1.179348352574583e-05, "loss": 0.0281, "step": 78800 }, { "grad_norm": 0.2065078616142273, "learning_rate": 1.1782819711417697e-05, "loss": 0.0148, "step": 78810 }, { "grad_norm": 0.21107961237430573, "learning_rate": 1.1772160076462951e-05, "loss": 0.0142, "step": 78820 }, { "grad_norm": 0.2371932864189148, "learning_rate": 1.1761504622047293e-05, "loss": 0.0167, "step": 78830 }, { "grad_norm": 0.29937753081321716, "learning_rate": 1.1750853349335999e-05, "loss": 0.0156, "step": 78840 }, { "grad_norm": 0.19951525330543518, "learning_rate": 1.1740206259493874e-05, "loss": 0.0164, "step": 78850 }, { "grad_norm": 0.1975063979625702, "learning_rate": 1.1729563353685286e-05, "loss": 0.0154, "step": 78860 }, { "grad_norm": 0.2542542517185211, "learning_rate": 1.1718924633074091e-05, "loss": 0.0172, "step": 78870 }, { "grad_norm": 0.2088891714811325, "learning_rate": 1.1708290098823765e-05, "loss": 0.0159, "step": 78880 }, { "grad_norm": 0.1774512231349945, "learning_rate": 1.1697659752097245e-05, "loss": 0.0089, "step": 78890 }, { "grad_norm": 0.23558300733566284, "learning_rate": 1.1687033594057068e-05, "loss": 0.0182, "step": 78900 }, { "grad_norm": 0.17278707027435303, "learning_rate": 1.1676411625865297e-05, "loss": 0.0166, "step": 78910 }, { "grad_norm": 0.3501850366592407, "learning_rate": 1.1665793848683542e-05, "loss": 0.0202, "step": 78920 }, { "grad_norm": 0.3775269687175751, "learning_rate": 1.1655180263672927e-05, "loss": 0.0243, "step": 78930 }, { "grad_norm": 0.2341199368238449, "learning_rate": 1.1644570871994142e-05, "loss": 0.015, "step": 78940 }, { "grad_norm": 0.1557840257883072, "learning_rate": 1.163396567480744e-05, "loss": 0.0123, "step": 78950 }, { "grad_norm": 0.18209059536457062, "learning_rate": 1.1623364673272552e-05, "loss": 0.0146, "step": 78960 }, { "grad_norm": 0.17503298819065094, "learning_rate": 1.1612767868548807e-05, "loss": 0.0129, "step": 78970 }, { "grad_norm": 0.21050812304019928, "learning_rate": 1.1602175261795057e-05, "loss": 0.0174, "step": 78980 }, { "grad_norm": 0.2681971788406372, "learning_rate": 1.1591586854169696e-05, "loss": 0.0169, "step": 78990 }, { "grad_norm": 0.18161696195602417, "learning_rate": 1.158100264683064e-05, "loss": 0.0139, "step": 79000 }, { "grad_norm": 0.28365665674209595, "learning_rate": 1.1570422640935386e-05, "loss": 0.0163, "step": 79010 }, { "grad_norm": 0.2714693546295166, "learning_rate": 1.155984683764092e-05, "loss": 0.0206, "step": 79020 }, { "grad_norm": 0.34193697571754456, "learning_rate": 1.1549275238103812e-05, "loss": 0.0144, "step": 79030 }, { "grad_norm": 0.17226822674274445, "learning_rate": 1.1538707843480152e-05, "loss": 0.0214, "step": 79040 }, { "grad_norm": 0.14351755380630493, "learning_rate": 1.1528144654925593e-05, "loss": 0.0203, "step": 79050 }, { "grad_norm": 0.31946513056755066, "learning_rate": 1.1517585673595277e-05, "loss": 0.0159, "step": 79060 }, { "grad_norm": 0.20153675973415375, "learning_rate": 1.150703090064395e-05, "loss": 0.0162, "step": 79070 }, { "grad_norm": 0.2617126405239105, "learning_rate": 1.1496480337225835e-05, "loss": 0.0195, "step": 79080 }, { "grad_norm": 0.2704489529132843, "learning_rate": 1.1485933984494751e-05, "loss": 0.014, "step": 79090 }, { "grad_norm": 0.391303151845932, "learning_rate": 1.1475391843604e-05, "loss": 0.0152, "step": 79100 }, { "grad_norm": 0.19434545934200287, "learning_rate": 1.1464853915706503e-05, "loss": 0.0243, "step": 79110 }, { "grad_norm": 0.22070327401161194, "learning_rate": 1.1454320201954626e-05, "loss": 0.0133, "step": 79120 }, { "grad_norm": 0.21240027248859406, "learning_rate": 1.1443790703500356e-05, "loss": 0.0148, "step": 79130 }, { "grad_norm": 0.22470639646053314, "learning_rate": 1.1433265421495154e-05, "loss": 0.0163, "step": 79140 }, { "grad_norm": 0.2039603888988495, "learning_rate": 1.1422744357090065e-05, "loss": 0.0234, "step": 79150 }, { "grad_norm": 0.3154541254043579, "learning_rate": 1.1412227511435647e-05, "loss": 0.0163, "step": 79160 }, { "grad_norm": 0.3208432197570801, "learning_rate": 1.1401714885682025e-05, "loss": 0.0113, "step": 79170 }, { "grad_norm": 0.17680855095386505, "learning_rate": 1.1391206480978823e-05, "loss": 0.0129, "step": 79180 }, { "grad_norm": 0.14564718306064606, "learning_rate": 1.1380702298475227e-05, "loss": 0.0158, "step": 79190 }, { "grad_norm": 0.2044553905725479, "learning_rate": 1.1370202339319985e-05, "loss": 0.0151, "step": 79200 }, { "grad_norm": 0.2036547213792801, "learning_rate": 1.1359706604661319e-05, "loss": 0.0161, "step": 79210 }, { "grad_norm": 0.2583516836166382, "learning_rate": 1.1349215095647042e-05, "loss": 0.0194, "step": 79220 }, { "grad_norm": 0.29973673820495605, "learning_rate": 1.1338727813424493e-05, "loss": 0.0175, "step": 79230 }, { "grad_norm": 0.1495901644229889, "learning_rate": 1.132824475914056e-05, "loss": 0.0156, "step": 79240 }, { "grad_norm": 0.19693586230278015, "learning_rate": 1.131776593394162e-05, "loss": 0.017, "step": 79250 }, { "grad_norm": 0.21253125369548798, "learning_rate": 1.1307291338973657e-05, "loss": 0.0189, "step": 79260 }, { "grad_norm": 0.18177033960819244, "learning_rate": 1.1296820975382121e-05, "loss": 0.0156, "step": 79270 }, { "grad_norm": 0.23539595305919647, "learning_rate": 1.1286354844312074e-05, "loss": 0.021, "step": 79280 }, { "grad_norm": 0.25275811553001404, "learning_rate": 1.1275892946908023e-05, "loss": 0.0193, "step": 79290 }, { "grad_norm": 0.4008253514766693, "learning_rate": 1.126543528431413e-05, "loss": 0.0169, "step": 79300 }, { "grad_norm": 0.4332684576511383, "learning_rate": 1.125498185767398e-05, "loss": 0.0181, "step": 79310 }, { "grad_norm": 0.17005161941051483, "learning_rate": 1.1244532668130781e-05, "loss": 0.0137, "step": 79320 }, { "grad_norm": 0.19330179691314697, "learning_rate": 1.1234087716827203e-05, "loss": 0.0191, "step": 79330 }, { "grad_norm": 0.20669089257717133, "learning_rate": 1.1223647004905524e-05, "loss": 0.0154, "step": 79340 }, { "grad_norm": 0.19339491426944733, "learning_rate": 1.1213210533507485e-05, "loss": 0.0187, "step": 79350 }, { "grad_norm": 0.8264492750167847, "learning_rate": 1.1202778303774447e-05, "loss": 0.0154, "step": 79360 }, { "grad_norm": 0.1519668996334076, "learning_rate": 1.1192350316847234e-05, "loss": 0.0151, "step": 79370 }, { "grad_norm": 0.4037747085094452, "learning_rate": 1.1181926573866258e-05, "loss": 0.0171, "step": 79380 }, { "grad_norm": 0.13531476259231567, "learning_rate": 1.1171507075971416e-05, "loss": 0.0163, "step": 79390 }, { "grad_norm": 0.1762983202934265, "learning_rate": 1.1161091824302177e-05, "loss": 0.0215, "step": 79400 }, { "grad_norm": 0.12951672077178955, "learning_rate": 1.115068081999755e-05, "loss": 0.0182, "step": 79410 }, { "grad_norm": 0.38547661900520325, "learning_rate": 1.1140274064196071e-05, "loss": 0.0176, "step": 79420 }, { "grad_norm": 0.24455180764198303, "learning_rate": 1.112987155803578e-05, "loss": 0.0236, "step": 79430 }, { "grad_norm": 0.2043790966272354, "learning_rate": 1.1119473302654298e-05, "loss": 0.0149, "step": 79440 }, { "grad_norm": 0.3647637963294983, "learning_rate": 1.1109079299188769e-05, "loss": 0.0141, "step": 79450 }, { "grad_norm": 0.19300571084022522, "learning_rate": 1.1098689548775847e-05, "loss": 0.0146, "step": 79460 }, { "grad_norm": 0.34477943181991577, "learning_rate": 1.1088304052551762e-05, "loss": 0.0187, "step": 79470 }, { "grad_norm": 0.1773141622543335, "learning_rate": 1.1077922811652215e-05, "loss": 0.0182, "step": 79480 }, { "grad_norm": 0.19669365882873535, "learning_rate": 1.1067545827212534e-05, "loss": 0.0129, "step": 79490 }, { "grad_norm": 0.22268109023571014, "learning_rate": 1.1057173100367496e-05, "loss": 0.0109, "step": 79500 }, { "grad_norm": 0.15913699567317963, "learning_rate": 1.1046804632251474e-05, "loss": 0.016, "step": 79510 }, { "grad_norm": 0.32133573293685913, "learning_rate": 1.1036440423998318e-05, "loss": 0.0131, "step": 79520 }, { "grad_norm": 0.23334205150604248, "learning_rate": 1.1026080476741469e-05, "loss": 0.0165, "step": 79530 }, { "grad_norm": 0.21543298661708832, "learning_rate": 1.1015724791613836e-05, "loss": 0.0134, "step": 79540 }, { "grad_norm": 0.3100019097328186, "learning_rate": 1.1005373369747951e-05, "loss": 0.0172, "step": 79550 }, { "grad_norm": 0.20424090325832367, "learning_rate": 1.09950262122758e-05, "loss": 0.0156, "step": 79560 }, { "grad_norm": 0.14062795042991638, "learning_rate": 1.0984683320328948e-05, "loss": 0.0147, "step": 79570 }, { "grad_norm": 0.17136505246162415, "learning_rate": 1.0974344695038458e-05, "loss": 0.0145, "step": 79580 }, { "grad_norm": 0.37506288290023804, "learning_rate": 1.0964010337534968e-05, "loss": 0.0145, "step": 79590 }, { "grad_norm": 0.2559433579444885, "learning_rate": 1.0953680248948611e-05, "loss": 0.0171, "step": 79600 }, { "grad_norm": 0.7084786295890808, "learning_rate": 1.0943354430409069e-05, "loss": 0.023, "step": 79610 }, { "grad_norm": 0.14196202158927917, "learning_rate": 1.0933032883045574e-05, "loss": 0.0125, "step": 79620 }, { "grad_norm": 0.2344001829624176, "learning_rate": 1.0922715607986872e-05, "loss": 0.014, "step": 79630 }, { "grad_norm": 0.21366724371910095, "learning_rate": 1.0912402606361234e-05, "loss": 0.0117, "step": 79640 }, { "grad_norm": 0.2653209865093231, "learning_rate": 1.090209387929647e-05, "loss": 0.0132, "step": 79650 }, { "grad_norm": 0.8451084494590759, "learning_rate": 1.0891789427919957e-05, "loss": 0.0201, "step": 79660 }, { "grad_norm": 0.26941531896591187, "learning_rate": 1.088148925335854e-05, "loss": 0.0143, "step": 79670 }, { "grad_norm": 0.1935897171497345, "learning_rate": 1.087119335673864e-05, "loss": 0.0145, "step": 79680 }, { "grad_norm": 0.17853839695453644, "learning_rate": 1.0860901739186208e-05, "loss": 0.0162, "step": 79690 }, { "grad_norm": 0.16421762108802795, "learning_rate": 1.085061440182673e-05, "loss": 0.0166, "step": 79700 }, { "grad_norm": 0.27492448687553406, "learning_rate": 1.0840331345785193e-05, "loss": 0.0164, "step": 79710 }, { "grad_norm": 0.1797025352716446, "learning_rate": 1.083005257218615e-05, "loss": 0.0145, "step": 79720 }, { "grad_norm": 0.18181918561458588, "learning_rate": 1.0819778082153642e-05, "loss": 0.0153, "step": 79730 }, { "grad_norm": 0.229512020945549, "learning_rate": 1.0809507876811325e-05, "loss": 0.016, "step": 79740 }, { "grad_norm": 0.2925126850605011, "learning_rate": 1.0799241957282292e-05, "loss": 0.0154, "step": 79750 }, { "grad_norm": 0.20637674629688263, "learning_rate": 1.078898032468923e-05, "loss": 0.0169, "step": 79760 }, { "grad_norm": 0.1749749332666397, "learning_rate": 1.0778722980154315e-05, "loss": 0.0238, "step": 79770 }, { "grad_norm": 0.36347731947898865, "learning_rate": 1.0768469924799302e-05, "loss": 0.0159, "step": 79780 }, { "grad_norm": 0.20265811681747437, "learning_rate": 1.0758221159745419e-05, "loss": 0.0155, "step": 79790 }, { "grad_norm": 0.13566862046718597, "learning_rate": 1.074797668611347e-05, "loss": 0.0193, "step": 79800 }, { "grad_norm": 0.2674289047718048, "learning_rate": 1.0737736505023777e-05, "loss": 0.011, "step": 79810 }, { "grad_norm": 0.19145400822162628, "learning_rate": 1.0727500617596203e-05, "loss": 0.0181, "step": 79820 }, { "grad_norm": 0.33244022727012634, "learning_rate": 1.0717269024950104e-05, "loss": 0.0176, "step": 79830 }, { "grad_norm": 0.25697848200798035, "learning_rate": 1.070704172820442e-05, "loss": 0.013, "step": 79840 }, { "grad_norm": 0.2541239559650421, "learning_rate": 1.0696818728477559e-05, "loss": 0.0163, "step": 79850 }, { "grad_norm": 0.16847553849220276, "learning_rate": 1.0686600026887511e-05, "loss": 0.0144, "step": 79860 }, { "grad_norm": 0.24276851117610931, "learning_rate": 1.0676385624551777e-05, "loss": 0.0188, "step": 79870 }, { "grad_norm": 0.3713575601577759, "learning_rate": 1.0666175522587402e-05, "loss": 0.0122, "step": 79880 }, { "grad_norm": 0.16951203346252441, "learning_rate": 1.0655969722110925e-05, "loss": 0.0122, "step": 79890 }, { "grad_norm": 0.25732725858688354, "learning_rate": 1.0645768224238439e-05, "loss": 0.0144, "step": 79900 }, { "grad_norm": 0.23957136273384094, "learning_rate": 1.063557103008559e-05, "loss": 0.0135, "step": 79910 }, { "grad_norm": 0.1916193962097168, "learning_rate": 1.0625378140767494e-05, "loss": 0.0122, "step": 79920 }, { "grad_norm": 0.21612775325775146, "learning_rate": 1.0615189557398846e-05, "loss": 0.0129, "step": 79930 }, { "grad_norm": 0.44660836458206177, "learning_rate": 1.060500528109386e-05, "loss": 0.0186, "step": 79940 }, { "grad_norm": 0.28067243099212646, "learning_rate": 1.0594825312966272e-05, "loss": 0.0143, "step": 79950 }, { "grad_norm": 0.2607419192790985, "learning_rate": 1.0584649654129336e-05, "loss": 0.0206, "step": 79960 }, { "grad_norm": 0.2796930968761444, "learning_rate": 1.0574478305695863e-05, "loss": 0.0158, "step": 79970 }, { "grad_norm": 0.19984747469425201, "learning_rate": 1.0564311268778154e-05, "loss": 0.016, "step": 79980 }, { "grad_norm": 0.1881057173013687, "learning_rate": 1.0554148544488073e-05, "loss": 0.0141, "step": 79990 }, { "grad_norm": 0.29417842626571655, "learning_rate": 1.0543990133937004e-05, "loss": 0.0203, "step": 80000 }, { "grad_norm": 0.19252045452594757, "learning_rate": 1.0533836038235868e-05, "loss": 0.0123, "step": 80010 }, { "grad_norm": 0.23804615437984467, "learning_rate": 1.052368625849507e-05, "loss": 0.0133, "step": 80020 }, { "grad_norm": 0.23274478316307068, "learning_rate": 1.05135407958246e-05, "loss": 0.0148, "step": 80030 }, { "grad_norm": 0.16221995651721954, "learning_rate": 1.0503399651333934e-05, "loss": 0.016, "step": 80040 }, { "grad_norm": 0.2551744282245636, "learning_rate": 1.04932628261321e-05, "loss": 0.0158, "step": 80050 }, { "grad_norm": 0.2819530963897705, "learning_rate": 1.0483130321327644e-05, "loss": 0.012, "step": 80060 }, { "grad_norm": 0.22470532357692719, "learning_rate": 1.0473002138028654e-05, "loss": 0.0134, "step": 80070 }, { "grad_norm": 0.4891889691352844, "learning_rate": 1.0462878277342713e-05, "loss": 0.0164, "step": 80080 }, { "grad_norm": 0.229198157787323, "learning_rate": 1.0452758740376972e-05, "loss": 0.0164, "step": 80090 }, { "grad_norm": 0.15928766131401062, "learning_rate": 1.0442643528238067e-05, "loss": 0.0103, "step": 80100 }, { "grad_norm": 0.31309688091278076, "learning_rate": 1.0432532642032188e-05, "loss": 0.0161, "step": 80110 }, { "grad_norm": 0.45358267426490784, "learning_rate": 1.042242608286505e-05, "loss": 0.0168, "step": 80120 }, { "grad_norm": 0.2833254635334015, "learning_rate": 1.041232385184191e-05, "loss": 0.0162, "step": 80130 }, { "grad_norm": 0.1732039898633957, "learning_rate": 1.0402225950067495e-05, "loss": 0.013, "step": 80140 }, { "grad_norm": 0.46468329429626465, "learning_rate": 1.0392132378646119e-05, "loss": 0.0167, "step": 80150 }, { "grad_norm": 0.21542738378047943, "learning_rate": 1.0382043138681607e-05, "loss": 0.0193, "step": 80160 }, { "grad_norm": 0.20721516013145447, "learning_rate": 1.0371958231277284e-05, "loss": 0.0126, "step": 80170 }, { "grad_norm": 0.2769557535648346, "learning_rate": 1.0361877657536024e-05, "loss": 0.0164, "step": 80180 }, { "grad_norm": 0.23692280054092407, "learning_rate": 1.035180141856023e-05, "loss": 0.0171, "step": 80190 }, { "grad_norm": 0.33674660325050354, "learning_rate": 1.0341729515451836e-05, "loss": 0.0189, "step": 80200 }, { "grad_norm": 0.14410780370235443, "learning_rate": 1.0331661949312265e-05, "loss": 0.0151, "step": 80210 }, { "grad_norm": 0.181797057390213, "learning_rate": 1.0321598721242514e-05, "loss": 0.0168, "step": 80220 }, { "grad_norm": 0.2295663058757782, "learning_rate": 1.0311539832343064e-05, "loss": 0.0136, "step": 80230 }, { "grad_norm": 0.17258179187774658, "learning_rate": 1.0301485283713942e-05, "loss": 0.0148, "step": 80240 }, { "grad_norm": 0.21577827632427216, "learning_rate": 1.0291435076454703e-05, "loss": 0.0192, "step": 80250 }, { "grad_norm": 0.19453920423984528, "learning_rate": 1.028138921166444e-05, "loss": 0.0152, "step": 80260 }, { "grad_norm": 0.22771669924259186, "learning_rate": 1.0271347690441719e-05, "loss": 0.0157, "step": 80270 }, { "grad_norm": 0.3883882164955139, "learning_rate": 1.0261310513884698e-05, "loss": 0.0159, "step": 80280 }, { "grad_norm": 0.2768265902996063, "learning_rate": 1.0251277683090998e-05, "loss": 0.0152, "step": 80290 }, { "grad_norm": 0.2492198348045349, "learning_rate": 1.0241249199157826e-05, "loss": 0.0163, "step": 80300 }, { "grad_norm": 0.1309279352426529, "learning_rate": 1.0231225063181837e-05, "loss": 0.0122, "step": 80310 }, { "grad_norm": 0.23467528820037842, "learning_rate": 1.0221205276259316e-05, "loss": 0.0133, "step": 80320 }, { "grad_norm": 0.1829078197479248, "learning_rate": 1.0211189839485958e-05, "loss": 0.0159, "step": 80330 }, { "grad_norm": 0.2160598337650299, "learning_rate": 1.0201178753957075e-05, "loss": 0.0103, "step": 80340 }, { "grad_norm": 0.17197245359420776, "learning_rate": 1.0191172020767432e-05, "loss": 0.0175, "step": 80350 }, { "grad_norm": 1.13823664188385, "learning_rate": 1.0181169641011362e-05, "loss": 0.0183, "step": 80360 }, { "grad_norm": 0.21198125183582306, "learning_rate": 1.0171171615782721e-05, "loss": 0.013, "step": 80370 }, { "grad_norm": 0.29139387607574463, "learning_rate": 1.0161177946174866e-05, "loss": 0.0179, "step": 80380 }, { "grad_norm": 0.2415381371974945, "learning_rate": 1.0151188633280706e-05, "loss": 0.0129, "step": 80390 }, { "grad_norm": 0.27474480867385864, "learning_rate": 1.0141203678192635e-05, "loss": 0.0146, "step": 80400 }, { "grad_norm": 0.19820205867290497, "learning_rate": 1.0131223082002612e-05, "loss": 0.0245, "step": 80410 }, { "grad_norm": 0.475588321685791, "learning_rate": 1.012124684580208e-05, "loss": 0.0117, "step": 80420 }, { "grad_norm": 0.11636443436145782, "learning_rate": 1.0111274970682039e-05, "loss": 0.0144, "step": 80430 }, { "grad_norm": 0.13347788155078888, "learning_rate": 1.010130745773299e-05, "loss": 0.0123, "step": 80440 }, { "grad_norm": 0.20754291117191315, "learning_rate": 1.0091344308044986e-05, "loss": 0.0135, "step": 80450 }, { "grad_norm": 0.17518992722034454, "learning_rate": 1.0081385522707554e-05, "loss": 0.0163, "step": 80460 }, { "grad_norm": 0.20826995372772217, "learning_rate": 1.0071431102809803e-05, "loss": 0.0133, "step": 80470 }, { "grad_norm": 0.1595558524131775, "learning_rate": 1.0061481049440297e-05, "loss": 0.0113, "step": 80480 }, { "grad_norm": 0.20984385907649994, "learning_rate": 1.0051535363687187e-05, "loss": 0.0126, "step": 80490 }, { "grad_norm": 0.168077290058136, "learning_rate": 1.0041594046638087e-05, "loss": 0.014, "step": 80500 }, { "grad_norm": 0.16251979768276215, "learning_rate": 1.0031657099380204e-05, "loss": 0.0131, "step": 80510 }, { "grad_norm": 0.1967688500881195, "learning_rate": 1.0021724523000204e-05, "loss": 0.0118, "step": 80520 }, { "grad_norm": 0.21916718780994415, "learning_rate": 1.001179631858431e-05, "loss": 0.0143, "step": 80530 }, { "grad_norm": 0.18154314160346985, "learning_rate": 1.0001872487218245e-05, "loss": 0.0159, "step": 80540 }, { "grad_norm": 0.214527428150177, "learning_rate": 9.991953029987272e-06, "loss": 0.0182, "step": 80550 }, { "grad_norm": 0.34344905614852905, "learning_rate": 9.982037947976147e-06, "loss": 0.019, "step": 80560 }, { "grad_norm": 0.20619072020053864, "learning_rate": 9.972127242269208e-06, "loss": 0.013, "step": 80570 }, { "grad_norm": 0.16592764854431152, "learning_rate": 9.962220913950238e-06, "loss": 0.0149, "step": 80580 }, { "grad_norm": 0.21480761468410492, "learning_rate": 9.952318964102591e-06, "loss": 0.02, "step": 80590 }, { "grad_norm": 0.1305551528930664, "learning_rate": 9.942421393809148e-06, "loss": 0.0145, "step": 80600 }, { "grad_norm": 0.2176366001367569, "learning_rate": 9.932528204152264e-06, "loss": 0.0124, "step": 80610 }, { "grad_norm": 0.2752223610877991, "learning_rate": 9.922639396213856e-06, "loss": 0.0248, "step": 80620 }, { "grad_norm": 0.2543978989124298, "learning_rate": 9.912754971075344e-06, "loss": 0.0137, "step": 80630 }, { "grad_norm": 0.2027515321969986, "learning_rate": 9.902874929817696e-06, "loss": 0.0147, "step": 80640 }, { "grad_norm": 0.11708950996398926, "learning_rate": 9.892999273521342e-06, "loss": 0.0159, "step": 80650 }, { "grad_norm": 0.19900794327259064, "learning_rate": 9.883128003266302e-06, "loss": 0.0135, "step": 80660 }, { "grad_norm": 0.20933270454406738, "learning_rate": 9.87326112013206e-06, "loss": 0.0166, "step": 80670 }, { "grad_norm": 0.22517190873622894, "learning_rate": 9.863398625197662e-06, "loss": 0.015, "step": 80680 }, { "grad_norm": 0.21172870695590973, "learning_rate": 9.853540519541615e-06, "loss": 0.0116, "step": 80690 }, { "grad_norm": 0.2307683527469635, "learning_rate": 9.843686804242053e-06, "loss": 0.0178, "step": 80700 }, { "grad_norm": 0.14165592193603516, "learning_rate": 9.833837480376506e-06, "loss": 0.0193, "step": 80710 }, { "grad_norm": 0.2017793506383896, "learning_rate": 9.82399254902212e-06, "loss": 0.0167, "step": 80720 }, { "grad_norm": 0.16440850496292114, "learning_rate": 9.814152011255494e-06, "loss": 0.0146, "step": 80730 }, { "grad_norm": 0.1765068769454956, "learning_rate": 9.804315868152797e-06, "loss": 0.0147, "step": 80740 }, { "grad_norm": 0.21293890476226807, "learning_rate": 9.79448412078966e-06, "loss": 0.0128, "step": 80750 }, { "grad_norm": 0.22588321566581726, "learning_rate": 9.784656770241312e-06, "loss": 0.0175, "step": 80760 }, { "grad_norm": 0.1501474529504776, "learning_rate": 9.77483381758243e-06, "loss": 0.0106, "step": 80770 }, { "grad_norm": 0.177979975938797, "learning_rate": 9.76501526388725e-06, "loss": 0.0114, "step": 80780 }, { "grad_norm": 0.31611520051956177, "learning_rate": 9.7552011102295e-06, "loss": 0.0292, "step": 80790 }, { "grad_norm": 0.1882239580154419, "learning_rate": 9.745391357682449e-06, "loss": 0.0157, "step": 80800 }, { "grad_norm": 0.19938796758651733, "learning_rate": 9.735586007318875e-06, "loss": 0.0206, "step": 80810 }, { "grad_norm": 0.32704323530197144, "learning_rate": 9.725785060211096e-06, "loss": 0.0136, "step": 80820 }, { "grad_norm": 0.31167933344841003, "learning_rate": 9.715988517430896e-06, "loss": 0.0154, "step": 80830 }, { "grad_norm": 0.22406600415706635, "learning_rate": 9.706196380049625e-06, "loss": 0.0124, "step": 80840 }, { "grad_norm": 0.15538053214550018, "learning_rate": 9.69640864913815e-06, "loss": 0.0116, "step": 80850 }, { "grad_norm": 0.189891517162323, "learning_rate": 9.686625325766818e-06, "loss": 0.0133, "step": 80860 }, { "grad_norm": 0.16740871965885162, "learning_rate": 9.67684641100554e-06, "loss": 0.0166, "step": 80870 }, { "grad_norm": 0.3635464012622833, "learning_rate": 9.667071905923692e-06, "loss": 0.0163, "step": 80880 }, { "grad_norm": 0.2206576019525528, "learning_rate": 9.657301811590242e-06, "loss": 0.0132, "step": 80890 }, { "grad_norm": 0.2543797492980957, "learning_rate": 9.647536129073603e-06, "loss": 0.0143, "step": 80900 }, { "grad_norm": 0.403879314661026, "learning_rate": 9.637774859441751e-06, "loss": 0.0145, "step": 80910 }, { "grad_norm": 0.1469523012638092, "learning_rate": 9.628018003762152e-06, "loss": 0.0156, "step": 80920 }, { "grad_norm": 0.19460520148277283, "learning_rate": 9.618265563101813e-06, "loss": 0.0155, "step": 80930 }, { "grad_norm": 0.2453661859035492, "learning_rate": 9.608517538527211e-06, "loss": 0.0153, "step": 80940 }, { "grad_norm": 0.1376766711473465, "learning_rate": 9.598773931104431e-06, "loss": 0.0152, "step": 80950 }, { "grad_norm": 0.26628923416137695, "learning_rate": 9.589034741898983e-06, "loss": 0.0161, "step": 80960 }, { "grad_norm": 0.22490549087524414, "learning_rate": 9.579299971975947e-06, "loss": 0.0186, "step": 80970 }, { "grad_norm": 0.19221587479114532, "learning_rate": 9.569569622399882e-06, "loss": 0.0128, "step": 80980 }, { "grad_norm": 0.29239335656166077, "learning_rate": 9.559843694234916e-06, "loss": 0.0121, "step": 80990 }, { "grad_norm": 0.2018846571445465, "learning_rate": 9.550122188544625e-06, "loss": 0.0154, "step": 81000 }, { "grad_norm": 0.23571781814098358, "learning_rate": 9.540405106392152e-06, "loss": 0.0146, "step": 81010 }, { "grad_norm": 0.16651953756809235, "learning_rate": 9.53069244884015e-06, "loss": 0.0145, "step": 81020 }, { "grad_norm": 0.2328580617904663, "learning_rate": 9.52098421695079e-06, "loss": 0.0147, "step": 81030 }, { "grad_norm": 0.1638888269662857, "learning_rate": 9.511280411785728e-06, "loss": 0.0144, "step": 81040 }, { "grad_norm": 0.1801232099533081, "learning_rate": 9.50158103440616e-06, "loss": 0.012, "step": 81050 }, { "grad_norm": 0.11461833119392395, "learning_rate": 9.491886085872815e-06, "loss": 0.0143, "step": 81060 }, { "grad_norm": 0.212516188621521, "learning_rate": 9.482195567245895e-06, "loss": 0.0153, "step": 81070 }, { "grad_norm": 0.2518816590309143, "learning_rate": 9.472509479585151e-06, "loss": 0.012, "step": 81080 }, { "grad_norm": 0.16581735014915466, "learning_rate": 9.462827823949833e-06, "loss": 0.0142, "step": 81090 }, { "grad_norm": 0.22731512784957886, "learning_rate": 9.453150601398735e-06, "loss": 0.0135, "step": 81100 }, { "grad_norm": 0.3019365966320038, "learning_rate": 9.44347781299011e-06, "loss": 0.0201, "step": 81110 }, { "grad_norm": 0.12602555751800537, "learning_rate": 9.43380945978179e-06, "loss": 0.0121, "step": 81120 }, { "grad_norm": 0.2555358111858368, "learning_rate": 9.424145542831047e-06, "loss": 0.014, "step": 81130 }, { "grad_norm": 0.2153303176164627, "learning_rate": 9.414486063194778e-06, "loss": 0.0173, "step": 81140 }, { "grad_norm": 0.5415668487548828, "learning_rate": 9.404831021929273e-06, "loss": 0.0167, "step": 81150 }, { "grad_norm": 0.1928206980228424, "learning_rate": 9.395180420090427e-06, "loss": 0.0149, "step": 81160 }, { "grad_norm": 0.2601640522480011, "learning_rate": 9.385534258733591e-06, "loss": 0.0168, "step": 81170 }, { "grad_norm": 0.2186536341905594, "learning_rate": 9.375892538913677e-06, "loss": 0.0157, "step": 81180 }, { "grad_norm": 0.1812550127506256, "learning_rate": 9.36625526168507e-06, "loss": 0.0128, "step": 81190 }, { "grad_norm": 0.17603030800819397, "learning_rate": 9.356622428101686e-06, "loss": 0.012, "step": 81200 }, { "grad_norm": 0.3724331557750702, "learning_rate": 9.346994039216972e-06, "loss": 0.0162, "step": 81210 }, { "grad_norm": 0.11335683614015579, "learning_rate": 9.337370096083886e-06, "loss": 0.0113, "step": 81220 }, { "grad_norm": 0.21039383113384247, "learning_rate": 9.327750599754854e-06, "loss": 0.0132, "step": 81230 }, { "grad_norm": 0.1511090248823166, "learning_rate": 9.318135551281875e-06, "loss": 0.014, "step": 81240 }, { "grad_norm": 0.22172677516937256, "learning_rate": 9.308524951716424e-06, "loss": 0.0121, "step": 81250 }, { "grad_norm": 0.11060021072626114, "learning_rate": 9.298918802109497e-06, "loss": 0.0129, "step": 81260 }, { "grad_norm": 0.18535393476486206, "learning_rate": 9.289317103511625e-06, "loss": 0.0139, "step": 81270 }, { "grad_norm": 0.15954141318798065, "learning_rate": 9.279719856972835e-06, "loss": 0.0133, "step": 81280 }, { "grad_norm": 0.21858195960521698, "learning_rate": 9.270127063542645e-06, "loss": 0.0156, "step": 81290 }, { "grad_norm": 0.12887415289878845, "learning_rate": 9.260538724270124e-06, "loss": 0.014, "step": 81300 }, { "grad_norm": 0.11697458475828171, "learning_rate": 9.250954840203852e-06, "loss": 0.0144, "step": 81310 }, { "grad_norm": 0.19310536980628967, "learning_rate": 9.241375412391878e-06, "loss": 0.0127, "step": 81320 }, { "grad_norm": 0.1386314481496811, "learning_rate": 9.23180044188181e-06, "loss": 0.0235, "step": 81330 }, { "grad_norm": 0.3538129925727844, "learning_rate": 9.222229929720755e-06, "loss": 0.0205, "step": 81340 }, { "grad_norm": 0.483428031206131, "learning_rate": 9.212663876955336e-06, "loss": 0.0182, "step": 81350 }, { "grad_norm": 0.3643285632133484, "learning_rate": 9.20310228463166e-06, "loss": 0.0221, "step": 81360 }, { "grad_norm": 0.19463041424751282, "learning_rate": 9.19354515379539e-06, "loss": 0.0181, "step": 81370 }, { "grad_norm": 0.3495132029056549, "learning_rate": 9.183992485491666e-06, "loss": 0.014, "step": 81380 }, { "grad_norm": 0.15476690232753754, "learning_rate": 9.174444280765154e-06, "loss": 0.0142, "step": 81390 }, { "grad_norm": 0.31628191471099854, "learning_rate": 9.164900540660032e-06, "loss": 0.0143, "step": 81400 }, { "grad_norm": 0.2173372358083725, "learning_rate": 9.155361266220008e-06, "loss": 0.0147, "step": 81410 }, { "grad_norm": 0.1943153738975525, "learning_rate": 9.145826458488254e-06, "loss": 0.0198, "step": 81420 }, { "grad_norm": 0.15664862096309662, "learning_rate": 9.136296118507509e-06, "loss": 0.0104, "step": 81430 }, { "grad_norm": 0.31966495513916016, "learning_rate": 9.126770247319966e-06, "loss": 0.0182, "step": 81440 }, { "grad_norm": 0.28208908438682556, "learning_rate": 9.11724884596738e-06, "loss": 0.0139, "step": 81450 }, { "grad_norm": 0.14206954836845398, "learning_rate": 9.107731915490996e-06, "loss": 0.0147, "step": 81460 }, { "grad_norm": 0.1739642173051834, "learning_rate": 9.098219456931579e-06, "loss": 0.0123, "step": 81470 }, { "grad_norm": 0.3924446403980255, "learning_rate": 9.088711471329376e-06, "loss": 0.0146, "step": 81480 }, { "grad_norm": 0.16399052739143372, "learning_rate": 9.079207959724184e-06, "loss": 0.0129, "step": 81490 }, { "grad_norm": 0.13637933135032654, "learning_rate": 9.069708923155279e-06, "loss": 0.0112, "step": 81500 }, { "grad_norm": 0.5080175995826721, "learning_rate": 9.060214362661468e-06, "loss": 0.0135, "step": 81510 }, { "grad_norm": 0.2698426842689514, "learning_rate": 9.050724279281064e-06, "loss": 0.0148, "step": 81520 }, { "grad_norm": 0.4377714991569519, "learning_rate": 9.041238674051894e-06, "loss": 0.0143, "step": 81530 }, { "grad_norm": 0.2641901671886444, "learning_rate": 9.031757548011277e-06, "loss": 0.0158, "step": 81540 }, { "grad_norm": 0.18816256523132324, "learning_rate": 9.022280902196051e-06, "loss": 0.0109, "step": 81550 }, { "grad_norm": 0.25717777013778687, "learning_rate": 9.012808737642592e-06, "loss": 0.0159, "step": 81560 }, { "grad_norm": 0.13909918069839478, "learning_rate": 9.003341055386739e-06, "loss": 0.0117, "step": 81570 }, { "grad_norm": 0.295341819524765, "learning_rate": 8.993877856463862e-06, "loss": 0.0131, "step": 81580 }, { "grad_norm": 0.14015349745750427, "learning_rate": 8.984419141908856e-06, "loss": 0.0129, "step": 81590 }, { "grad_norm": 0.23537981510162354, "learning_rate": 8.974964912756117e-06, "loss": 0.0173, "step": 81600 }, { "grad_norm": 0.19501203298568726, "learning_rate": 8.965515170039523e-06, "loss": 0.0131, "step": 81610 }, { "grad_norm": 0.14713473618030548, "learning_rate": 8.956069914792514e-06, "loss": 0.0132, "step": 81620 }, { "grad_norm": 0.08934289962053299, "learning_rate": 8.946629148047969e-06, "loss": 0.0141, "step": 81630 }, { "grad_norm": 0.30523210763931274, "learning_rate": 8.93719287083834e-06, "loss": 0.0224, "step": 81640 }, { "grad_norm": 0.2765932083129883, "learning_rate": 8.927761084195569e-06, "loss": 0.0125, "step": 81650 }, { "grad_norm": 0.21263813972473145, "learning_rate": 8.918333789151096e-06, "loss": 0.0147, "step": 81660 }, { "grad_norm": 0.2996288239955902, "learning_rate": 8.90891098673587e-06, "loss": 0.0139, "step": 81670 }, { "grad_norm": 0.1468958705663681, "learning_rate": 8.89949267798037e-06, "loss": 0.0151, "step": 81680 }, { "grad_norm": 0.1555652916431427, "learning_rate": 8.890078863914542e-06, "loss": 0.0149, "step": 81690 }, { "grad_norm": 0.12427296489477158, "learning_rate": 8.880669545567894e-06, "loss": 0.0127, "step": 81700 }, { "grad_norm": 0.17105872929096222, "learning_rate": 8.871264723969381e-06, "loss": 0.0116, "step": 81710 }, { "grad_norm": 0.41803520917892456, "learning_rate": 8.86186440014754e-06, "loss": 0.0186, "step": 81720 }, { "grad_norm": 0.20978249609470367, "learning_rate": 8.852468575130351e-06, "loss": 0.019, "step": 81730 }, { "grad_norm": 0.1804816722869873, "learning_rate": 8.843077249945347e-06, "loss": 0.0134, "step": 81740 }, { "grad_norm": 0.13689285516738892, "learning_rate": 8.833690425619523e-06, "loss": 0.0163, "step": 81750 }, { "grad_norm": 0.1716332733631134, "learning_rate": 8.82430810317942e-06, "loss": 0.0179, "step": 81760 }, { "grad_norm": 0.34029608964920044, "learning_rate": 8.814930283651074e-06, "loss": 0.0128, "step": 81770 }, { "grad_norm": 0.2599310576915741, "learning_rate": 8.805556968060048e-06, "loss": 0.013, "step": 81780 }, { "grad_norm": 0.1231059804558754, "learning_rate": 8.796188157431362e-06, "loss": 0.0215, "step": 81790 }, { "grad_norm": 0.18087729811668396, "learning_rate": 8.786823852789595e-06, "loss": 0.0129, "step": 81800 }, { "grad_norm": 0.1790269911289215, "learning_rate": 8.777464055158813e-06, "loss": 0.0128, "step": 81810 }, { "grad_norm": 0.1980964094400406, "learning_rate": 8.768108765562578e-06, "loss": 0.0197, "step": 81820 }, { "grad_norm": 0.32127052545547485, "learning_rate": 8.758757985023975e-06, "loss": 0.0139, "step": 81830 }, { "grad_norm": 0.4863080680370331, "learning_rate": 8.749411714565598e-06, "loss": 0.0134, "step": 81840 }, { "grad_norm": 0.15936671197414398, "learning_rate": 8.74006995520954e-06, "loss": 0.019, "step": 81850 }, { "grad_norm": 0.1840267926454544, "learning_rate": 8.730732707977397e-06, "loss": 0.0138, "step": 81860 }, { "grad_norm": 0.15510766208171844, "learning_rate": 8.721399973890281e-06, "loss": 0.0138, "step": 81870 }, { "grad_norm": 0.13001541793346405, "learning_rate": 8.712071753968792e-06, "loss": 0.0119, "step": 81880 }, { "grad_norm": 0.27112871408462524, "learning_rate": 8.702748049233072e-06, "loss": 0.0118, "step": 81890 }, { "grad_norm": 0.12547504901885986, "learning_rate": 8.693428860702718e-06, "loss": 0.0158, "step": 81900 }, { "grad_norm": 0.30841362476348877, "learning_rate": 8.684114189396898e-06, "loss": 0.0137, "step": 81910 }, { "grad_norm": 0.12633317708969116, "learning_rate": 8.674804036334222e-06, "loss": 0.0119, "step": 81920 }, { "grad_norm": 0.32701992988586426, "learning_rate": 8.665498402532857e-06, "loss": 0.0173, "step": 81930 }, { "grad_norm": 0.3089441955089569, "learning_rate": 8.656197289010432e-06, "loss": 0.0152, "step": 81940 }, { "grad_norm": 0.27299773693084717, "learning_rate": 8.646900696784122e-06, "loss": 0.0121, "step": 81950 }, { "grad_norm": 0.18641570210456848, "learning_rate": 8.637608626870553e-06, "loss": 0.018, "step": 81960 }, { "grad_norm": 0.21720652282238007, "learning_rate": 8.628321080285944e-06, "loss": 0.0147, "step": 81970 }, { "grad_norm": 0.366921991109848, "learning_rate": 8.619038058045925e-06, "loss": 0.0119, "step": 81980 }, { "grad_norm": 0.6321474313735962, "learning_rate": 8.609759561165704e-06, "loss": 0.0135, "step": 81990 }, { "grad_norm": 0.2767791748046875, "learning_rate": 8.600485590659934e-06, "loss": 0.0144, "step": 82000 }, { "grad_norm": 0.2339823991060257, "learning_rate": 8.591216147542813e-06, "loss": 0.0147, "step": 82010 }, { "grad_norm": 0.241554394364357, "learning_rate": 8.581951232828045e-06, "loss": 0.0223, "step": 82020 }, { "grad_norm": 0.19543397426605225, "learning_rate": 8.572690847528825e-06, "loss": 0.0122, "step": 82030 }, { "grad_norm": 0.22272413969039917, "learning_rate": 8.563434992657838e-06, "loss": 0.0163, "step": 82040 }, { "grad_norm": 0.24149252474308014, "learning_rate": 8.554183669227307e-06, "loss": 0.0125, "step": 82050 }, { "grad_norm": 0.19152288138866425, "learning_rate": 8.544936878248937e-06, "loss": 0.0125, "step": 82060 }, { "grad_norm": 0.2045208066701889, "learning_rate": 8.535694620733942e-06, "loss": 0.019, "step": 82070 }, { "grad_norm": 0.15633994340896606, "learning_rate": 8.526456897693053e-06, "loss": 0.0121, "step": 82080 }, { "grad_norm": 0.23290136456489563, "learning_rate": 8.517223710136456e-06, "loss": 0.0232, "step": 82090 }, { "grad_norm": 0.2175177037715912, "learning_rate": 8.507995059073931e-06, "loss": 0.0177, "step": 82100 }, { "grad_norm": 0.22189423441886902, "learning_rate": 8.498770945514673e-06, "loss": 0.0167, "step": 82110 }, { "grad_norm": 0.4837396740913391, "learning_rate": 8.489551370467436e-06, "loss": 0.0148, "step": 82120 }, { "grad_norm": 0.15704947710037231, "learning_rate": 8.480336334940442e-06, "loss": 0.0173, "step": 82130 }, { "grad_norm": 0.28678983449935913, "learning_rate": 8.471125839941458e-06, "loss": 0.016, "step": 82140 }, { "grad_norm": 0.21632301807403564, "learning_rate": 8.461919886477682e-06, "loss": 0.0148, "step": 82150 }, { "grad_norm": 0.21378354728221893, "learning_rate": 8.452718475555927e-06, "loss": 0.0153, "step": 82160 }, { "grad_norm": 0.3037031292915344, "learning_rate": 8.443521608182398e-06, "loss": 0.0179, "step": 82170 }, { "grad_norm": 0.35857319831848145, "learning_rate": 8.43432928536288e-06, "loss": 0.0131, "step": 82180 }, { "grad_norm": 0.2487080693244934, "learning_rate": 8.425141508102607e-06, "loss": 0.0195, "step": 82190 }, { "grad_norm": 0.18185938894748688, "learning_rate": 8.415958277406365e-06, "loss": 0.0144, "step": 82200 }, { "grad_norm": 0.15291441977024078, "learning_rate": 8.406779594278392e-06, "loss": 0.0171, "step": 82210 }, { "grad_norm": 0.22893503308296204, "learning_rate": 8.397605459722468e-06, "loss": 0.0173, "step": 82220 }, { "grad_norm": 0.26635053753852844, "learning_rate": 8.388435874741863e-06, "loss": 0.017, "step": 82230 }, { "grad_norm": 0.21268218755722046, "learning_rate": 8.37927084033936e-06, "loss": 0.0152, "step": 82240 }, { "grad_norm": 0.17000167071819305, "learning_rate": 8.370110357517208e-06, "loss": 0.0142, "step": 82250 }, { "grad_norm": 0.08010399341583252, "learning_rate": 8.360954427277201e-06, "loss": 0.0133, "step": 82260 }, { "grad_norm": 0.14009328186511993, "learning_rate": 8.351803050620627e-06, "loss": 0.0146, "step": 82270 }, { "grad_norm": 0.2714771032333374, "learning_rate": 8.34265622854824e-06, "loss": 0.0182, "step": 82280 }, { "grad_norm": 0.18805231153964996, "learning_rate": 8.333513962060335e-06, "loss": 0.0207, "step": 82290 }, { "grad_norm": 0.18616920709609985, "learning_rate": 8.324376252156701e-06, "loss": 0.0107, "step": 82300 }, { "grad_norm": 0.22552268207073212, "learning_rate": 8.315243099836633e-06, "loss": 0.0166, "step": 82310 }, { "grad_norm": 0.25023505091667175, "learning_rate": 8.3061145060989e-06, "loss": 0.0197, "step": 82320 }, { "grad_norm": 0.17996352910995483, "learning_rate": 8.296990471941806e-06, "loss": 0.0125, "step": 82330 }, { "grad_norm": 0.15450842678546906, "learning_rate": 8.287870998363112e-06, "loss": 0.0148, "step": 82340 }, { "grad_norm": 0.14315320551395416, "learning_rate": 8.278756086360156e-06, "loss": 0.0116, "step": 82350 }, { "grad_norm": 0.2713419198989868, "learning_rate": 8.269645736929693e-06, "loss": 0.0113, "step": 82360 }, { "grad_norm": 0.2134319245815277, "learning_rate": 8.260539951068048e-06, "loss": 0.0111, "step": 82370 }, { "grad_norm": 0.2492504119873047, "learning_rate": 8.251438729770983e-06, "loss": 0.0174, "step": 82380 }, { "grad_norm": 0.13796183466911316, "learning_rate": 8.242342074033826e-06, "loss": 0.0144, "step": 82390 }, { "grad_norm": 0.1800893098115921, "learning_rate": 8.233249984851344e-06, "loss": 0.0184, "step": 82400 }, { "grad_norm": 0.2722328305244446, "learning_rate": 8.224162463217849e-06, "loss": 0.02, "step": 82410 }, { "grad_norm": 0.18855667114257812, "learning_rate": 8.215079510127139e-06, "loss": 0.014, "step": 82420 }, { "grad_norm": 0.15341290831565857, "learning_rate": 8.206001126572521e-06, "loss": 0.0147, "step": 82430 }, { "grad_norm": 0.18650616705417633, "learning_rate": 8.196927313546776e-06, "loss": 0.0182, "step": 82440 }, { "grad_norm": 0.2150147408246994, "learning_rate": 8.18785807204222e-06, "loss": 0.0134, "step": 82450 }, { "grad_norm": 0.5558782815933228, "learning_rate": 8.178793403050627e-06, "loss": 0.0168, "step": 82460 }, { "grad_norm": 0.27783524990081787, "learning_rate": 8.16973330756331e-06, "loss": 0.0155, "step": 82470 }, { "grad_norm": 0.255144327878952, "learning_rate": 8.16067778657107e-06, "loss": 0.0136, "step": 82480 }, { "grad_norm": 0.19103394448757172, "learning_rate": 8.151626841064214e-06, "loss": 0.0143, "step": 82490 }, { "grad_norm": 0.13465657830238342, "learning_rate": 8.142580472032518e-06, "loss": 0.0136, "step": 82500 }, { "grad_norm": 0.4338621497154236, "learning_rate": 8.133538680465285e-06, "loss": 0.016, "step": 82510 }, { "grad_norm": 0.7817443609237671, "learning_rate": 8.124501467351336e-06, "loss": 0.0171, "step": 82520 }, { "grad_norm": 0.18133680522441864, "learning_rate": 8.11546883367893e-06, "loss": 0.0148, "step": 82530 }, { "grad_norm": 0.2558709681034088, "learning_rate": 8.106440780435882e-06, "loss": 0.0163, "step": 82540 }, { "grad_norm": 0.18295960128307343, "learning_rate": 8.097417308609485e-06, "loss": 0.0132, "step": 82550 }, { "grad_norm": 0.2589297890663147, "learning_rate": 8.088398419186544e-06, "loss": 0.0127, "step": 82560 }, { "grad_norm": 0.4194488823413849, "learning_rate": 8.079384113153321e-06, "loss": 0.0154, "step": 82570 }, { "grad_norm": 0.2779392898082733, "learning_rate": 8.070374391495644e-06, "loss": 0.0144, "step": 82580 }, { "grad_norm": 0.1579972356557846, "learning_rate": 8.061369255198764e-06, "loss": 0.0157, "step": 82590 }, { "grad_norm": 0.2226559966802597, "learning_rate": 8.052368705247493e-06, "loss": 0.0177, "step": 82600 }, { "grad_norm": 0.09806650131940842, "learning_rate": 8.043372742626114e-06, "loss": 0.0223, "step": 82610 }, { "grad_norm": 0.1435810923576355, "learning_rate": 8.034381368318422e-06, "loss": 0.0139, "step": 82620 }, { "grad_norm": 0.21101266145706177, "learning_rate": 8.025394583307677e-06, "loss": 0.0173, "step": 82630 }, { "grad_norm": 0.22331979870796204, "learning_rate": 8.016412388576683e-06, "loss": 0.0212, "step": 82640 }, { "grad_norm": 0.11440322548151016, "learning_rate": 8.007434785107699e-06, "loss": 0.0138, "step": 82650 }, { "grad_norm": 0.18942111730575562, "learning_rate": 7.998461773882509e-06, "loss": 0.0149, "step": 82660 }, { "grad_norm": 0.1890857070684433, "learning_rate": 7.989493355882393e-06, "loss": 0.0203, "step": 82670 }, { "grad_norm": 0.2605401873588562, "learning_rate": 7.980529532088137e-06, "loss": 0.0143, "step": 82680 }, { "grad_norm": 0.21179868280887604, "learning_rate": 7.97157030347998e-06, "loss": 0.0141, "step": 82690 }, { "grad_norm": 0.20966003835201263, "learning_rate": 7.96261567103772e-06, "loss": 0.0131, "step": 82700 }, { "grad_norm": 0.25792211294174194, "learning_rate": 7.953665635740593e-06, "loss": 0.0135, "step": 82710 }, { "grad_norm": 0.2822815477848053, "learning_rate": 7.94472019856738e-06, "loss": 0.0178, "step": 82720 }, { "grad_norm": 0.2585230767726898, "learning_rate": 7.935779360496337e-06, "loss": 0.0124, "step": 82730 }, { "grad_norm": 0.24082256853580475, "learning_rate": 7.92684312250523e-06, "loss": 0.0173, "step": 82740 }, { "grad_norm": 0.3313295245170593, "learning_rate": 7.917911485571289e-06, "loss": 0.0142, "step": 82750 }, { "grad_norm": 0.6432223320007324, "learning_rate": 7.90898445067128e-06, "loss": 0.0174, "step": 82760 }, { "grad_norm": 0.1705676019191742, "learning_rate": 7.900062018781462e-06, "loss": 0.0121, "step": 82770 }, { "grad_norm": 0.14053800702095032, "learning_rate": 7.891144190877553e-06, "loss": 0.0141, "step": 82780 }, { "grad_norm": 0.16877387464046478, "learning_rate": 7.882230967934806e-06, "loss": 0.0151, "step": 82790 }, { "grad_norm": 0.1530982404947281, "learning_rate": 7.873322350927959e-06, "loss": 0.0185, "step": 82800 }, { "grad_norm": 0.3477779030799866, "learning_rate": 7.864418340831253e-06, "loss": 0.016, "step": 82810 }, { "grad_norm": 0.2927045524120331, "learning_rate": 7.855518938618395e-06, "loss": 0.014, "step": 82820 }, { "grad_norm": 0.2318563610315323, "learning_rate": 7.846624145262633e-06, "loss": 0.0116, "step": 82830 }, { "grad_norm": 0.6979709267616272, "learning_rate": 7.837733961736672e-06, "loss": 0.0167, "step": 82840 }, { "grad_norm": 0.2631949782371521, "learning_rate": 7.828848389012727e-06, "loss": 0.014, "step": 82850 }, { "grad_norm": 0.21238289773464203, "learning_rate": 7.819967428062524e-06, "loss": 0.0166, "step": 82860 }, { "grad_norm": 0.30587247014045715, "learning_rate": 7.811091079857274e-06, "loss": 0.0177, "step": 82870 }, { "grad_norm": 0.1707238107919693, "learning_rate": 7.802219345367667e-06, "loss": 0.0142, "step": 82880 }, { "grad_norm": 0.12861810624599457, "learning_rate": 7.793352225563916e-06, "loss": 0.0147, "step": 82890 }, { "grad_norm": 0.17705997824668884, "learning_rate": 7.7844897214157e-06, "loss": 0.0135, "step": 82900 }, { "grad_norm": 0.20631687343120575, "learning_rate": 7.775631833892228e-06, "loss": 0.0198, "step": 82910 }, { "grad_norm": 0.21834085881710052, "learning_rate": 7.76677856396215e-06, "loss": 0.0169, "step": 82920 }, { "grad_norm": 0.28054100275039673, "learning_rate": 7.757929912593697e-06, "loss": 0.0157, "step": 82930 }, { "grad_norm": 0.11118628829717636, "learning_rate": 7.749085880754509e-06, "loss": 0.0109, "step": 82940 }, { "grad_norm": 0.21851247549057007, "learning_rate": 7.740246469411777e-06, "loss": 0.0108, "step": 82950 }, { "grad_norm": 0.2857752740383148, "learning_rate": 7.731411679532142e-06, "loss": 0.0171, "step": 82960 }, { "grad_norm": 0.15025168657302856, "learning_rate": 7.722581512081779e-06, "loss": 0.0126, "step": 82970 }, { "grad_norm": 0.47980621457099915, "learning_rate": 7.713755968026331e-06, "loss": 0.0141, "step": 82980 }, { "grad_norm": 0.1970069855451584, "learning_rate": 7.704935048330969e-06, "loss": 0.0129, "step": 82990 }, { "grad_norm": 0.41592010855674744, "learning_rate": 7.696118753960313e-06, "loss": 0.0154, "step": 83000 }, { "grad_norm": 0.23701311647891998, "learning_rate": 7.687307085878503e-06, "loss": 0.0131, "step": 83010 }, { "grad_norm": 0.12068338692188263, "learning_rate": 7.678500045049186e-06, "loss": 0.0126, "step": 83020 }, { "grad_norm": 0.14815308153629303, "learning_rate": 7.669697632435463e-06, "loss": 0.0155, "step": 83030 }, { "grad_norm": 0.22238944470882416, "learning_rate": 7.660899848999963e-06, "loss": 0.0124, "step": 83040 }, { "grad_norm": 0.1425531506538391, "learning_rate": 7.652106695704803e-06, "loss": 0.0156, "step": 83050 }, { "grad_norm": 0.12350988388061523, "learning_rate": 7.643318173511598e-06, "loss": 0.0132, "step": 83060 }, { "grad_norm": 0.31983643770217896, "learning_rate": 7.634534283381423e-06, "loss": 0.0124, "step": 83070 }, { "grad_norm": 0.1361398696899414, "learning_rate": 7.625755026274889e-06, "loss": 0.0191, "step": 83080 }, { "grad_norm": 0.28797799348831177, "learning_rate": 7.616980403152069e-06, "loss": 0.0132, "step": 83090 }, { "grad_norm": 0.20240329205989838, "learning_rate": 7.608210414972561e-06, "loss": 0.0168, "step": 83100 }, { "grad_norm": 0.15195311605930328, "learning_rate": 7.599445062695404e-06, "loss": 0.0122, "step": 83110 }, { "grad_norm": 0.1875651329755783, "learning_rate": 7.590684347279214e-06, "loss": 0.0192, "step": 83120 }, { "grad_norm": 0.18272843956947327, "learning_rate": 7.581928269682004e-06, "loss": 0.0135, "step": 83130 }, { "grad_norm": 0.1986387073993683, "learning_rate": 7.573176830861361e-06, "loss": 0.0129, "step": 83140 }, { "grad_norm": 0.24428869783878326, "learning_rate": 7.5644300317743e-06, "loss": 0.0167, "step": 83150 }, { "grad_norm": 0.1750422716140747, "learning_rate": 7.555687873377376e-06, "loss": 0.0111, "step": 83160 }, { "grad_norm": 0.11107617616653442, "learning_rate": 7.546950356626592e-06, "loss": 0.0174, "step": 83170 }, { "grad_norm": 0.19261473417282104, "learning_rate": 7.538217482477516e-06, "loss": 0.0138, "step": 83180 }, { "grad_norm": 0.16113153100013733, "learning_rate": 7.529489251885119e-06, "loss": 0.0192, "step": 83190 }, { "grad_norm": 0.13484260439872742, "learning_rate": 7.520765665803941e-06, "loss": 0.0134, "step": 83200 }, { "grad_norm": 0.21573060750961304, "learning_rate": 7.512046725187949e-06, "loss": 0.0106, "step": 83210 }, { "grad_norm": 0.27466145157814026, "learning_rate": 7.503332430990645e-06, "loss": 0.0125, "step": 83220 }, { "grad_norm": 0.2402973175048828, "learning_rate": 7.494622784165017e-06, "loss": 0.0191, "step": 83230 }, { "grad_norm": 0.27753597497940063, "learning_rate": 7.485917785663543e-06, "loss": 0.0128, "step": 83240 }, { "grad_norm": 0.30421462655067444, "learning_rate": 7.477217436438172e-06, "loss": 0.0152, "step": 83250 }, { "grad_norm": 0.2287275791168213, "learning_rate": 7.46852173744037e-06, "loss": 0.0119, "step": 83260 }, { "grad_norm": 0.2673291563987732, "learning_rate": 7.4598306896211004e-06, "loss": 0.0161, "step": 83270 }, { "grad_norm": 0.2268487960100174, "learning_rate": 7.451144293930773e-06, "loss": 0.0119, "step": 83280 }, { "grad_norm": 0.24844583868980408, "learning_rate": 7.4424625513193455e-06, "loss": 0.0125, "step": 83290 }, { "grad_norm": 0.18458527326583862, "learning_rate": 7.433785462736209e-06, "loss": 0.0164, "step": 83300 }, { "grad_norm": 0.6443570256233215, "learning_rate": 7.425113029130315e-06, "loss": 0.0163, "step": 83310 }, { "grad_norm": 0.2369544506072998, "learning_rate": 7.416445251450044e-06, "loss": 0.0183, "step": 83320 }, { "grad_norm": 0.20652058720588684, "learning_rate": 7.4077821306433035e-06, "loss": 0.0119, "step": 83330 }, { "grad_norm": 0.3008899688720703, "learning_rate": 7.399123667657465e-06, "loss": 0.0164, "step": 83340 }, { "grad_norm": 0.1439266949892044, "learning_rate": 7.390469863439426e-06, "loss": 0.0157, "step": 83350 }, { "grad_norm": 0.2653723955154419, "learning_rate": 7.381820718935512e-06, "loss": 0.0147, "step": 83360 }, { "grad_norm": 0.2632422149181366, "learning_rate": 7.373176235091633e-06, "loss": 0.0174, "step": 83370 }, { "grad_norm": 0.1729651689529419, "learning_rate": 7.364536412853101e-06, "loss": 0.0141, "step": 83380 }, { "grad_norm": 0.18506678938865662, "learning_rate": 7.355901253164776e-06, "loss": 0.0155, "step": 83390 }, { "grad_norm": 0.19666725397109985, "learning_rate": 7.34727075697097e-06, "loss": 0.0114, "step": 83400 }, { "grad_norm": 0.254526823759079, "learning_rate": 7.3386449252155095e-06, "loss": 0.0164, "step": 83410 }, { "grad_norm": 0.2532329261302948, "learning_rate": 7.330023758841692e-06, "loss": 0.0147, "step": 83420 }, { "grad_norm": 0.14271056652069092, "learning_rate": 7.321407258792323e-06, "loss": 0.011, "step": 83430 }, { "grad_norm": 0.1707235723733902, "learning_rate": 7.312795426009694e-06, "loss": 0.0167, "step": 83440 }, { "grad_norm": 0.19289664924144745, "learning_rate": 7.304188261435585e-06, "loss": 0.0179, "step": 83450 }, { "grad_norm": 0.22792793810367584, "learning_rate": 7.295585766011242e-06, "loss": 0.0154, "step": 83460 }, { "grad_norm": 0.1685154139995575, "learning_rate": 7.286987940677436e-06, "loss": 0.0124, "step": 83470 }, { "grad_norm": 0.4362557530403137, "learning_rate": 7.278394786374426e-06, "loss": 0.016, "step": 83480 }, { "grad_norm": 0.245247021317482, "learning_rate": 7.269806304041915e-06, "loss": 0.0135, "step": 83490 }, { "grad_norm": 0.15946315228939056, "learning_rate": 7.261222494619147e-06, "loss": 0.0113, "step": 83500 }, { "grad_norm": 0.17194058001041412, "learning_rate": 7.252643359044825e-06, "loss": 0.0166, "step": 83510 }, { "grad_norm": 0.1873312145471573, "learning_rate": 7.244068898257172e-06, "loss": 0.0119, "step": 83520 }, { "grad_norm": 0.19751371443271637, "learning_rate": 7.2354991131938495e-06, "loss": 0.0152, "step": 83530 }, { "grad_norm": 0.15345068275928497, "learning_rate": 7.2269340047920574e-06, "loss": 0.0103, "step": 83540 }, { "grad_norm": 0.1938934326171875, "learning_rate": 7.218373573988435e-06, "loss": 0.0139, "step": 83550 }, { "grad_norm": 0.3269789516925812, "learning_rate": 7.209817821719178e-06, "loss": 0.0163, "step": 83560 }, { "grad_norm": 0.305232435464859, "learning_rate": 7.201266748919894e-06, "loss": 0.0199, "step": 83570 }, { "grad_norm": 0.19521036744117737, "learning_rate": 7.19272035652575e-06, "loss": 0.0139, "step": 83580 }, { "grad_norm": 0.19153186678886414, "learning_rate": 7.184178645471334e-06, "loss": 0.0147, "step": 83590 }, { "grad_norm": 0.20836900174617767, "learning_rate": 7.175641616690775e-06, "loss": 0.0123, "step": 83600 }, { "grad_norm": 0.2599732279777527, "learning_rate": 7.167109271117661e-06, "loss": 0.0149, "step": 83610 }, { "grad_norm": 0.3020564913749695, "learning_rate": 7.158581609685072e-06, "loss": 0.017, "step": 83620 }, { "grad_norm": 0.17265747487545013, "learning_rate": 7.150058633325591e-06, "loss": 0.0141, "step": 83630 }, { "grad_norm": 0.24333786964416504, "learning_rate": 7.141540342971287e-06, "loss": 0.013, "step": 83640 }, { "grad_norm": 0.3061951696872711, "learning_rate": 7.133026739553677e-06, "loss": 0.0124, "step": 83650 }, { "grad_norm": 0.22664323449134827, "learning_rate": 7.124517824003829e-06, "loss": 0.0145, "step": 83660 }, { "grad_norm": 0.21680045127868652, "learning_rate": 7.116013597252236e-06, "loss": 0.0185, "step": 83670 }, { "grad_norm": 0.13652993738651276, "learning_rate": 7.107514060228921e-06, "loss": 0.0155, "step": 83680 }, { "grad_norm": 0.21347787976264954, "learning_rate": 7.099019213863378e-06, "loss": 0.0186, "step": 83690 }, { "grad_norm": 0.4695698022842407, "learning_rate": 7.090529059084611e-06, "loss": 0.0165, "step": 83700 }, { "grad_norm": 0.2026580274105072, "learning_rate": 7.082043596821058e-06, "loss": 0.0099, "step": 83710 }, { "grad_norm": 0.41289153695106506, "learning_rate": 7.07356282800069e-06, "loss": 0.0143, "step": 83720 }, { "grad_norm": 0.3096455931663513, "learning_rate": 7.0650867535509645e-06, "loss": 0.0108, "step": 83730 }, { "grad_norm": 0.19694183766841888, "learning_rate": 7.0566153743987895e-06, "loss": 0.0148, "step": 83740 }, { "grad_norm": 0.21614420413970947, "learning_rate": 7.04814869147059e-06, "loss": 0.011, "step": 83750 }, { "grad_norm": 0.15866528451442719, "learning_rate": 7.0396867056922725e-06, "loss": 0.0109, "step": 83760 }, { "grad_norm": 0.1279449611902237, "learning_rate": 7.031229417989244e-06, "loss": 0.0166, "step": 83770 }, { "grad_norm": 0.15317736566066742, "learning_rate": 7.022776829286348e-06, "loss": 0.0175, "step": 83780 }, { "grad_norm": 0.20457562804222107, "learning_rate": 7.014328940507975e-06, "loss": 0.0147, "step": 83790 }, { "grad_norm": 0.1768283247947693, "learning_rate": 7.005885752577945e-06, "loss": 0.0186, "step": 83800 }, { "grad_norm": 0.6750417351722717, "learning_rate": 6.997447266419615e-06, "loss": 0.0125, "step": 83810 }, { "grad_norm": 0.21271838247776031, "learning_rate": 6.989013482955797e-06, "loss": 0.0118, "step": 83820 }, { "grad_norm": 0.25029316544532776, "learning_rate": 6.980584403108808e-06, "loss": 0.0141, "step": 83830 }, { "grad_norm": 0.18845762312412262, "learning_rate": 6.972160027800417e-06, "loss": 0.0135, "step": 83840 }, { "grad_norm": 0.23545275628566742, "learning_rate": 6.963740357951931e-06, "loss": 0.0131, "step": 83850 }, { "grad_norm": 0.15215077996253967, "learning_rate": 6.95532539448408e-06, "loss": 0.0158, "step": 83860 }, { "grad_norm": 0.2593776285648346, "learning_rate": 6.946915138317128e-06, "loss": 0.0194, "step": 83870 }, { "grad_norm": 0.3268672823905945, "learning_rate": 6.9385095903708065e-06, "loss": 0.0185, "step": 83880 }, { "grad_norm": 0.21740089356899261, "learning_rate": 6.930108751564346e-06, "loss": 0.016, "step": 83890 }, { "grad_norm": 0.26937761902809143, "learning_rate": 6.921712622816428e-06, "loss": 0.0174, "step": 83900 }, { "grad_norm": 0.14404526352882385, "learning_rate": 6.913321205045259e-06, "loss": 0.0114, "step": 83910 }, { "grad_norm": 0.33608710765838623, "learning_rate": 6.904934499168491e-06, "loss": 0.0133, "step": 83920 }, { "grad_norm": 0.20894771814346313, "learning_rate": 6.896552506103299e-06, "loss": 0.0142, "step": 83930 }, { "grad_norm": 0.22720079123973846, "learning_rate": 6.888175226766313e-06, "loss": 0.0136, "step": 83940 }, { "grad_norm": 0.09681076556444168, "learning_rate": 6.879802662073686e-06, "loss": 0.0105, "step": 83950 }, { "grad_norm": 0.30298686027526855, "learning_rate": 6.871434812940991e-06, "loss": 0.021, "step": 83960 }, { "grad_norm": 0.40126529335975647, "learning_rate": 6.863071680283345e-06, "loss": 0.0136, "step": 83970 }, { "grad_norm": 0.2241198867559433, "learning_rate": 6.854713265015339e-06, "loss": 0.021, "step": 83980 }, { "grad_norm": 0.19226889312267303, "learning_rate": 6.8463595680510025e-06, "loss": 0.0128, "step": 83990 }, { "grad_norm": 0.2065150886774063, "learning_rate": 6.838010590303911e-06, "loss": 0.0138, "step": 84000 }, { "grad_norm": 0.32027366757392883, "learning_rate": 6.829666332687085e-06, "loss": 0.0108, "step": 84010 }, { "grad_norm": 0.21652357280254364, "learning_rate": 6.821326796113059e-06, "loss": 0.0122, "step": 84020 }, { "grad_norm": 0.14155539870262146, "learning_rate": 6.812991981493799e-06, "loss": 0.0139, "step": 84030 }, { "grad_norm": 0.24796532094478607, "learning_rate": 6.804661889740821e-06, "loss": 0.0155, "step": 84040 }, { "grad_norm": 0.25787654519081116, "learning_rate": 6.7963365217650625e-06, "loss": 0.0163, "step": 84050 }, { "grad_norm": 0.2206285446882248, "learning_rate": 6.788015878476983e-06, "loss": 0.0123, "step": 84060 }, { "grad_norm": 0.2021360844373703, "learning_rate": 6.779699960786523e-06, "loss": 0.0126, "step": 84070 }, { "grad_norm": 0.2590068280696869, "learning_rate": 6.771388769603104e-06, "loss": 0.0132, "step": 84080 }, { "grad_norm": 0.162130206823349, "learning_rate": 6.763082305835605e-06, "loss": 0.0196, "step": 84090 }, { "grad_norm": 0.2159026861190796, "learning_rate": 6.754780570392433e-06, "loss": 0.0152, "step": 84100 }, { "grad_norm": 0.18089058995246887, "learning_rate": 6.7464835641814215e-06, "loss": 0.013, "step": 84110 }, { "grad_norm": 0.2312842160463333, "learning_rate": 6.7381912881099515e-06, "loss": 0.0113, "step": 84120 }, { "grad_norm": 0.4378229081630707, "learning_rate": 6.72990374308482e-06, "loss": 0.0165, "step": 84130 }, { "grad_norm": 0.17073701322078705, "learning_rate": 6.721620930012379e-06, "loss": 0.0137, "step": 84140 }, { "grad_norm": 0.28482624888420105, "learning_rate": 6.713342849798393e-06, "loss": 0.0128, "step": 84150 }, { "grad_norm": 0.18638959527015686, "learning_rate": 6.705069503348166e-06, "loss": 0.0127, "step": 84160 }, { "grad_norm": 0.17288196086883545, "learning_rate": 6.696800891566435e-06, "loss": 0.0144, "step": 84170 }, { "grad_norm": 0.16628873348236084, "learning_rate": 6.688537015357449e-06, "loss": 0.0124, "step": 84180 }, { "grad_norm": 0.22638168931007385, "learning_rate": 6.680277875624941e-06, "loss": 0.0122, "step": 84190 }, { "grad_norm": 0.19167685508728027, "learning_rate": 6.672023473272121e-06, "loss": 0.0142, "step": 84200 }, { "grad_norm": 0.2480163723230362, "learning_rate": 6.663773809201667e-06, "loss": 0.0125, "step": 84210 }, { "grad_norm": 0.19202479720115662, "learning_rate": 6.655528884315748e-06, "loss": 0.014, "step": 84220 }, { "grad_norm": 0.24618999660015106, "learning_rate": 6.647288699516036e-06, "loss": 0.0209, "step": 84230 }, { "grad_norm": 0.25253352522850037, "learning_rate": 6.639053255703642e-06, "loss": 0.0187, "step": 84240 }, { "grad_norm": 0.29432642459869385, "learning_rate": 6.6308225537791925e-06, "loss": 0.0141, "step": 84250 }, { "grad_norm": 0.188676118850708, "learning_rate": 6.622596594642777e-06, "loss": 0.015, "step": 84260 }, { "grad_norm": 0.17767855525016785, "learning_rate": 6.614375379193999e-06, "loss": 0.0126, "step": 84270 }, { "grad_norm": 0.2421407252550125, "learning_rate": 6.606158908331889e-06, "loss": 0.0168, "step": 84280 }, { "grad_norm": 0.1694543957710266, "learning_rate": 6.597947182955005e-06, "loss": 0.016, "step": 84290 }, { "grad_norm": 0.1922045350074768, "learning_rate": 6.589740203961359e-06, "loss": 0.0237, "step": 84300 }, { "grad_norm": 0.3652366101741791, "learning_rate": 6.581537972248464e-06, "loss": 0.0245, "step": 84310 }, { "grad_norm": 0.11243237555027008, "learning_rate": 6.573340488713275e-06, "loss": 0.0131, "step": 84320 }, { "grad_norm": 0.2057713121175766, "learning_rate": 6.565147754252304e-06, "loss": 0.0164, "step": 84330 }, { "grad_norm": 0.22156353294849396, "learning_rate": 6.556959769761462e-06, "loss": 0.0112, "step": 84340 }, { "grad_norm": 0.3304648995399475, "learning_rate": 6.548776536136192e-06, "loss": 0.0152, "step": 84350 }, { "grad_norm": 0.22377242147922516, "learning_rate": 6.54059805427138e-06, "loss": 0.0141, "step": 84360 }, { "grad_norm": 0.2152179777622223, "learning_rate": 6.532424325061437e-06, "loss": 0.0168, "step": 84370 }, { "grad_norm": 0.13840298354625702, "learning_rate": 6.5242553494001935e-06, "loss": 0.0129, "step": 84380 }, { "grad_norm": 0.15052208304405212, "learning_rate": 6.516091128181046e-06, "loss": 0.0114, "step": 84390 }, { "grad_norm": 0.1694003790616989, "learning_rate": 6.507931662296785e-06, "loss": 0.0196, "step": 84400 }, { "grad_norm": 0.20292118191719055, "learning_rate": 6.4997769526397354e-06, "loss": 0.0144, "step": 84410 }, { "grad_norm": 0.30022507905960083, "learning_rate": 6.491627000101663e-06, "loss": 0.0141, "step": 84420 }, { "grad_norm": 0.13682544231414795, "learning_rate": 6.4834818055738545e-06, "loss": 0.0146, "step": 84430 }, { "grad_norm": 0.13351009786128998, "learning_rate": 6.4753413699470465e-06, "loss": 0.012, "step": 84440 }, { "grad_norm": 0.265605628490448, "learning_rate": 6.467205694111478e-06, "loss": 0.0141, "step": 84450 }, { "grad_norm": 0.19725190103054047, "learning_rate": 6.459074778956836e-06, "loss": 0.0143, "step": 84460 }, { "grad_norm": 0.2814267575740814, "learning_rate": 6.450948625372311e-06, "loss": 0.0117, "step": 84470 }, { "grad_norm": 0.28666022419929504, "learning_rate": 6.442827234246585e-06, "loss": 0.0134, "step": 84480 }, { "grad_norm": 0.22306717932224274, "learning_rate": 6.43471060646777e-06, "loss": 0.0154, "step": 84490 }, { "grad_norm": 0.13058443367481232, "learning_rate": 6.426598742923512e-06, "loss": 0.0114, "step": 84500 }, { "grad_norm": 0.4488551616668701, "learning_rate": 6.418491644500885e-06, "loss": 0.0181, "step": 84510 }, { "grad_norm": 0.16609476506710052, "learning_rate": 6.410389312086512e-06, "loss": 0.0206, "step": 84520 }, { "grad_norm": 0.15847748517990112, "learning_rate": 6.4022917465664135e-06, "loss": 0.0167, "step": 84530 }, { "grad_norm": 0.16881999373435974, "learning_rate": 6.3941989488261575e-06, "loss": 0.0149, "step": 84540 }, { "grad_norm": 0.2520597279071808, "learning_rate": 6.386110919750726e-06, "loss": 0.0131, "step": 84550 }, { "grad_norm": 0.1625383049249649, "learning_rate": 6.378027660224651e-06, "loss": 0.0134, "step": 84560 }, { "grad_norm": 0.16349928081035614, "learning_rate": 6.369949171131856e-06, "loss": 0.0168, "step": 84570 }, { "grad_norm": 0.23365476727485657, "learning_rate": 6.361875453355848e-06, "loss": 0.0131, "step": 84580 }, { "grad_norm": 0.19134904444217682, "learning_rate": 6.353806507779525e-06, "loss": 0.0117, "step": 84590 }, { "grad_norm": 0.19973570108413696, "learning_rate": 6.3457423352853044e-06, "loss": 0.0112, "step": 84600 }, { "grad_norm": 0.2167983502149582, "learning_rate": 6.337682936755062e-06, "loss": 0.0148, "step": 84610 }, { "grad_norm": 0.17291197180747986, "learning_rate": 6.329628313070179e-06, "loss": 0.0123, "step": 84620 }, { "grad_norm": 0.2092655897140503, "learning_rate": 6.321578465111477e-06, "loss": 0.0154, "step": 84630 }, { "grad_norm": 0.22041943669319153, "learning_rate": 6.31353339375928e-06, "loss": 0.0135, "step": 84640 }, { "grad_norm": 0.2584579288959503, "learning_rate": 6.305493099893395e-06, "loss": 0.0149, "step": 84650 }, { "grad_norm": 0.18211187422275543, "learning_rate": 6.297457584393096e-06, "loss": 0.0109, "step": 84660 }, { "grad_norm": 0.21584735810756683, "learning_rate": 6.289426848137126e-06, "loss": 0.0128, "step": 84670 }, { "grad_norm": 0.29246774315834045, "learning_rate": 6.281400892003713e-06, "loss": 0.0107, "step": 84680 }, { "grad_norm": 0.23426270484924316, "learning_rate": 6.273379716870581e-06, "loss": 0.0119, "step": 84690 }, { "grad_norm": 0.16245068609714508, "learning_rate": 6.265363323614887e-06, "loss": 0.0154, "step": 84700 }, { "grad_norm": 0.11741741746664047, "learning_rate": 6.257351713113308e-06, "loss": 0.0132, "step": 84710 }, { "grad_norm": 0.2003614604473114, "learning_rate": 6.249344886241981e-06, "loss": 0.0167, "step": 84720 }, { "grad_norm": 0.3000571131706238, "learning_rate": 6.2413428438765295e-06, "loss": 0.0105, "step": 84730 }, { "grad_norm": 0.218595951795578, "learning_rate": 6.233345586892026e-06, "loss": 0.0094, "step": 84740 }, { "grad_norm": 0.24808178842067719, "learning_rate": 6.225353116163052e-06, "loss": 0.0135, "step": 84750 }, { "grad_norm": 0.16933573782444, "learning_rate": 6.217365432563627e-06, "loss": 0.0164, "step": 84760 }, { "grad_norm": 0.6434735059738159, "learning_rate": 6.209382536967318e-06, "loss": 0.0171, "step": 84770 }, { "grad_norm": 0.37072041630744934, "learning_rate": 6.201404430247082e-06, "loss": 0.015, "step": 84780 }, { "grad_norm": 0.20614014565944672, "learning_rate": 6.1934311132754165e-06, "loss": 0.0147, "step": 84790 }, { "grad_norm": 0.2313961535692215, "learning_rate": 6.185462586924257e-06, "loss": 0.0161, "step": 84800 }, { "grad_norm": 0.14887820184230804, "learning_rate": 6.177498852065039e-06, "loss": 0.0123, "step": 84810 }, { "grad_norm": 0.11817354708909988, "learning_rate": 6.169539909568655e-06, "loss": 0.0131, "step": 84820 }, { "grad_norm": 0.13424566388130188, "learning_rate": 6.161585760305488e-06, "loss": 0.0128, "step": 84830 }, { "grad_norm": 0.2172904759645462, "learning_rate": 6.153636405145391e-06, "loss": 0.0167, "step": 84840 }, { "grad_norm": 0.152107834815979, "learning_rate": 6.145691844957702e-06, "loss": 0.0121, "step": 84850 }, { "grad_norm": 0.22978238761425018, "learning_rate": 6.1377520806112145e-06, "loss": 0.0151, "step": 84860 }, { "grad_norm": 0.1149401068687439, "learning_rate": 6.129817112974218e-06, "loss": 0.0158, "step": 84870 }, { "grad_norm": 0.2113540768623352, "learning_rate": 6.121886942914457e-06, "loss": 0.0146, "step": 84880 }, { "grad_norm": 0.20172275602817535, "learning_rate": 6.11396157129917e-06, "loss": 0.0116, "step": 84890 }, { "grad_norm": 0.2700105905532837, "learning_rate": 6.106040998995066e-06, "loss": 0.0139, "step": 84900 }, { "grad_norm": 0.300245463848114, "learning_rate": 6.098125226868329e-06, "loss": 0.0165, "step": 84910 }, { "grad_norm": 0.2000177800655365, "learning_rate": 6.090214255784604e-06, "loss": 0.0187, "step": 84920 }, { "grad_norm": 0.43117696046829224, "learning_rate": 6.0823080866090346e-06, "loss": 0.022, "step": 84930 }, { "grad_norm": 0.18542669713497162, "learning_rate": 6.074406720206233e-06, "loss": 0.0112, "step": 84940 }, { "grad_norm": 0.1899644434452057, "learning_rate": 6.066510157440259e-06, "loss": 0.0126, "step": 84950 }, { "grad_norm": 0.2757413685321808, "learning_rate": 6.058618399174676e-06, "loss": 0.0136, "step": 84960 }, { "grad_norm": 0.42907318472862244, "learning_rate": 6.050731446272528e-06, "loss": 0.0129, "step": 84970 }, { "grad_norm": 0.49119997024536133, "learning_rate": 6.042849299596321e-06, "loss": 0.0158, "step": 84980 }, { "grad_norm": 0.14981384575366974, "learning_rate": 6.034971960008013e-06, "loss": 0.0112, "step": 84990 }, { "grad_norm": 0.08016029000282288, "learning_rate": 6.027099428369082e-06, "loss": 0.0147, "step": 85000 }, { "grad_norm": 0.1752379834651947, "learning_rate": 6.019231705540435e-06, "loss": 0.0098, "step": 85010 }, { "grad_norm": 0.25730517506599426, "learning_rate": 6.011368792382488e-06, "loss": 0.0126, "step": 85020 }, { "grad_norm": 0.8245256543159485, "learning_rate": 6.00351068975511e-06, "loss": 0.021, "step": 85030 }, { "grad_norm": 0.1795894056558609, "learning_rate": 5.9956573985176674e-06, "loss": 0.0095, "step": 85040 }, { "grad_norm": 0.15198498964309692, "learning_rate": 5.987808919528959e-06, "loss": 0.012, "step": 85050 }, { "grad_norm": 0.1890851855278015, "learning_rate": 5.979965253647307e-06, "loss": 0.0184, "step": 85060 }, { "grad_norm": 0.2768590748310089, "learning_rate": 5.9721264017304655e-06, "loss": 0.0199, "step": 85070 }, { "grad_norm": 0.20747743546962738, "learning_rate": 5.964292364635682e-06, "loss": 0.02, "step": 85080 }, { "grad_norm": 0.08129797130823135, "learning_rate": 5.956463143219682e-06, "loss": 0.0156, "step": 85090 }, { "grad_norm": 0.28189611434936523, "learning_rate": 5.948638738338663e-06, "loss": 0.018, "step": 85100 }, { "grad_norm": 0.2660778760910034, "learning_rate": 5.940819150848271e-06, "loss": 0.0141, "step": 85110 }, { "grad_norm": 0.19001421332359314, "learning_rate": 5.933004381603663e-06, "loss": 0.0135, "step": 85120 }, { "grad_norm": 0.2004666030406952, "learning_rate": 5.92519443145943e-06, "loss": 0.0127, "step": 85130 }, { "grad_norm": 0.27572041749954224, "learning_rate": 5.917389301269671e-06, "loss": 0.0154, "step": 85140 }, { "grad_norm": 0.4827643036842346, "learning_rate": 5.9095889918879435e-06, "loss": 0.0171, "step": 85150 }, { "grad_norm": 0.19486214220523834, "learning_rate": 5.901793504167286e-06, "loss": 0.0156, "step": 85160 }, { "grad_norm": 0.28821131587028503, "learning_rate": 5.894002838960172e-06, "loss": 0.0198, "step": 85170 }, { "grad_norm": 0.11769832670688629, "learning_rate": 5.886216997118604e-06, "loss": 0.0165, "step": 85180 }, { "grad_norm": 0.2615814507007599, "learning_rate": 5.878435979494023e-06, "loss": 0.0118, "step": 85190 }, { "grad_norm": 0.12185057252645493, "learning_rate": 5.870659786937344e-06, "loss": 0.012, "step": 85200 }, { "grad_norm": 0.18917974829673767, "learning_rate": 5.862888420298962e-06, "loss": 0.0106, "step": 85210 }, { "grad_norm": 0.15357688069343567, "learning_rate": 5.855121880428743e-06, "loss": 0.0103, "step": 85220 }, { "grad_norm": 0.1826280653476715, "learning_rate": 5.84736016817603e-06, "loss": 0.0113, "step": 85230 }, { "grad_norm": 0.19308748841285706, "learning_rate": 5.839603284389622e-06, "loss": 0.0123, "step": 85240 }, { "grad_norm": 0.1652265340089798, "learning_rate": 5.831851229917812e-06, "loss": 0.0131, "step": 85250 }, { "grad_norm": 0.23399455845355988, "learning_rate": 5.824104005608338e-06, "loss": 0.0151, "step": 85260 }, { "grad_norm": 0.5220572352409363, "learning_rate": 5.81636161230843e-06, "loss": 0.0171, "step": 85270 }, { "grad_norm": 0.23732861876487732, "learning_rate": 5.808624050864791e-06, "loss": 0.0181, "step": 85280 }, { "grad_norm": 0.2472408264875412, "learning_rate": 5.800891322123592e-06, "loss": 0.0159, "step": 85290 }, { "grad_norm": 0.09238860756158829, "learning_rate": 5.793163426930454e-06, "loss": 0.0144, "step": 85300 }, { "grad_norm": 0.3192233443260193, "learning_rate": 5.785440366130512e-06, "loss": 0.0131, "step": 85310 }, { "grad_norm": 0.20771728456020355, "learning_rate": 5.777722140568326e-06, "loss": 0.0146, "step": 85320 }, { "grad_norm": 0.22259998321533203, "learning_rate": 5.7700087510879685e-06, "loss": 0.0141, "step": 85330 }, { "grad_norm": 0.22139872610569, "learning_rate": 5.762300198532933e-06, "loss": 0.011, "step": 85340 }, { "grad_norm": 0.1797044575214386, "learning_rate": 5.7545964837462606e-06, "loss": 0.0129, "step": 85350 }, { "grad_norm": 0.21431761980056763, "learning_rate": 5.746897607570384e-06, "loss": 0.0133, "step": 85360 }, { "grad_norm": 0.24323557317256927, "learning_rate": 5.7392035708472626e-06, "loss": 0.0156, "step": 85370 }, { "grad_norm": 0.195164754986763, "learning_rate": 5.731514374418278e-06, "loss": 0.0124, "step": 85380 }, { "grad_norm": 0.1111498698592186, "learning_rate": 5.7238300191243325e-06, "loss": 0.0132, "step": 85390 }, { "grad_norm": 0.6138683557510376, "learning_rate": 5.716150505805762e-06, "loss": 0.0153, "step": 85400 }, { "grad_norm": 0.3355933725833893, "learning_rate": 5.708475835302407e-06, "loss": 0.014, "step": 85410 }, { "grad_norm": 0.16414999961853027, "learning_rate": 5.70080600845353e-06, "loss": 0.0175, "step": 85420 }, { "grad_norm": 0.21550020575523376, "learning_rate": 5.693141026097909e-06, "loss": 0.0182, "step": 85430 }, { "grad_norm": 0.1981479972600937, "learning_rate": 5.6854808890737745e-06, "loss": 0.0136, "step": 85440 }, { "grad_norm": 0.1672178953886032, "learning_rate": 5.6778255982188225e-06, "loss": 0.0182, "step": 85450 }, { "grad_norm": 0.265632688999176, "learning_rate": 5.670175154370217e-06, "loss": 0.0203, "step": 85460 }, { "grad_norm": 0.3093161880970001, "learning_rate": 5.662529558364616e-06, "loss": 0.0166, "step": 85470 }, { "grad_norm": 0.1405952274799347, "learning_rate": 5.654888811038128e-06, "loss": 0.0194, "step": 85480 }, { "grad_norm": 0.23979812860488892, "learning_rate": 5.647252913226314e-06, "loss": 0.0136, "step": 85490 }, { "grad_norm": 0.2087005376815796, "learning_rate": 5.639621865764255e-06, "loss": 0.0143, "step": 85500 }, { "grad_norm": 0.7146527171134949, "learning_rate": 5.63199566948644e-06, "loss": 0.0149, "step": 85510 }, { "grad_norm": 0.12374221533536911, "learning_rate": 5.6243743252268835e-06, "loss": 0.0125, "step": 85520 }, { "grad_norm": 0.1423562914133072, "learning_rate": 5.616757833819009e-06, "loss": 0.0167, "step": 85530 }, { "grad_norm": 0.13259513676166534, "learning_rate": 5.609146196095794e-06, "loss": 0.0112, "step": 85540 }, { "grad_norm": 0.3181099593639374, "learning_rate": 5.6015394128896005e-06, "loss": 0.0117, "step": 85550 }, { "grad_norm": 0.4021434187889099, "learning_rate": 5.593937485032308e-06, "loss": 0.0161, "step": 85560 }, { "grad_norm": 0.2221345752477646, "learning_rate": 5.586340413355246e-06, "loss": 0.0138, "step": 85570 }, { "grad_norm": 0.5722909569740295, "learning_rate": 5.578748198689227e-06, "loss": 0.015, "step": 85580 }, { "grad_norm": 0.094510018825531, "learning_rate": 5.571160841864504e-06, "loss": 0.0119, "step": 85590 }, { "grad_norm": 0.12534089386463165, "learning_rate": 5.563578343710846e-06, "loss": 0.0147, "step": 85600 }, { "grad_norm": 0.14848285913467407, "learning_rate": 5.556000705057446e-06, "loss": 0.0094, "step": 85610 }, { "grad_norm": 0.2018442004919052, "learning_rate": 5.548427926732996e-06, "loss": 0.0159, "step": 85620 }, { "grad_norm": 0.16819937527179718, "learning_rate": 5.540860009565629e-06, "loss": 0.0118, "step": 85630 }, { "grad_norm": 0.17301660776138306, "learning_rate": 5.533296954382966e-06, "loss": 0.0118, "step": 85640 }, { "grad_norm": 0.305695503950119, "learning_rate": 5.525738762012095e-06, "loss": 0.0155, "step": 85650 }, { "grad_norm": 0.4014759659767151, "learning_rate": 5.518185433279577e-06, "loss": 0.0185, "step": 85660 }, { "grad_norm": 0.1565442830324173, "learning_rate": 5.510636969011407e-06, "loss": 0.0115, "step": 85670 }, { "grad_norm": 0.15980283915996552, "learning_rate": 5.503093370033096e-06, "loss": 0.0104, "step": 85680 }, { "grad_norm": 0.23169121146202087, "learning_rate": 5.495554637169603e-06, "loss": 0.0112, "step": 85690 }, { "grad_norm": 0.21960338950157166, "learning_rate": 5.488020771245328e-06, "loss": 0.0156, "step": 85700 }, { "grad_norm": 0.231561541557312, "learning_rate": 5.480491773084195e-06, "loss": 0.0126, "step": 85710 }, { "grad_norm": 0.23425394296646118, "learning_rate": 5.472967643509525e-06, "loss": 0.0131, "step": 85720 }, { "grad_norm": 0.8391238451004028, "learning_rate": 5.46544838334419e-06, "loss": 0.0155, "step": 85730 }, { "grad_norm": 0.14064042270183563, "learning_rate": 5.4579339934104525e-06, "loss": 0.013, "step": 85740 }, { "grad_norm": 0.19173209369182587, "learning_rate": 5.450424474530091e-06, "loss": 0.0155, "step": 85750 }, { "grad_norm": 0.30339354276657104, "learning_rate": 5.442919827524323e-06, "loss": 0.0109, "step": 85760 }, { "grad_norm": 0.16453757882118225, "learning_rate": 5.435420053213863e-06, "loss": 0.016, "step": 85770 }, { "grad_norm": 0.2605343759059906, "learning_rate": 5.427925152418845e-06, "loss": 0.0271, "step": 85780 }, { "grad_norm": 0.24772465229034424, "learning_rate": 5.420435125958934e-06, "loss": 0.0147, "step": 85790 }, { "grad_norm": 0.20334531366825104, "learning_rate": 5.412949974653209e-06, "loss": 0.0135, "step": 85800 }, { "grad_norm": 0.17526821792125702, "learning_rate": 5.405469699320248e-06, "loss": 0.0108, "step": 85810 }, { "grad_norm": 0.12622395157814026, "learning_rate": 5.397994300778064e-06, "loss": 0.016, "step": 85820 }, { "grad_norm": 0.2783313989639282, "learning_rate": 5.39052377984417e-06, "loss": 0.014, "step": 85830 }, { "grad_norm": 0.4313603639602661, "learning_rate": 5.383058137335523e-06, "loss": 0.0127, "step": 85840 }, { "grad_norm": 0.18544408679008484, "learning_rate": 5.375597374068553e-06, "loss": 0.0171, "step": 85850 }, { "grad_norm": 0.15806570649147034, "learning_rate": 5.368141490859163e-06, "loss": 0.0139, "step": 85860 }, { "grad_norm": 0.3587166368961334, "learning_rate": 5.360690488522729e-06, "loss": 0.0132, "step": 85870 }, { "grad_norm": 0.25062835216522217, "learning_rate": 5.353244367874055e-06, "loss": 0.0126, "step": 85880 }, { "grad_norm": 0.23938380181789398, "learning_rate": 5.345803129727456e-06, "loss": 0.013, "step": 85890 }, { "grad_norm": 0.21336452662944794, "learning_rate": 5.338366774896697e-06, "loss": 0.0107, "step": 85900 }, { "grad_norm": 0.34395670890808105, "learning_rate": 5.330935304194995e-06, "loss": 0.0161, "step": 85910 }, { "grad_norm": 0.2352748066186905, "learning_rate": 5.323508718435044e-06, "loss": 0.012, "step": 85920 }, { "grad_norm": 0.19223810732364655, "learning_rate": 5.316087018429011e-06, "loss": 0.0127, "step": 85930 }, { "grad_norm": 0.19327282905578613, "learning_rate": 5.30867020498853e-06, "loss": 0.0124, "step": 85940 }, { "grad_norm": 0.13636815547943115, "learning_rate": 5.30125827892467e-06, "loss": 0.0123, "step": 85950 }, { "grad_norm": 0.21302521228790283, "learning_rate": 5.293851241048015e-06, "loss": 0.0168, "step": 85960 }, { "grad_norm": 0.19895967841148376, "learning_rate": 5.286449092168555e-06, "loss": 0.0142, "step": 85970 }, { "grad_norm": 0.2969738245010376, "learning_rate": 5.2790518330958105e-06, "loss": 0.0184, "step": 85980 }, { "grad_norm": 0.1446482092142105, "learning_rate": 5.271659464638717e-06, "loss": 0.0098, "step": 85990 }, { "grad_norm": 0.3432171642780304, "learning_rate": 5.264271987605701e-06, "loss": 0.0141, "step": 86000 }, { "grad_norm": 0.13442328572273254, "learning_rate": 5.256889402804632e-06, "loss": 0.0106, "step": 86010 }, { "grad_norm": 0.2199031561613083, "learning_rate": 5.249511711042881e-06, "loss": 0.0187, "step": 86020 }, { "grad_norm": 0.31423330307006836, "learning_rate": 5.242138913127237e-06, "loss": 0.0143, "step": 86030 }, { "grad_norm": 0.22868557274341583, "learning_rate": 5.234771009863987e-06, "loss": 0.0117, "step": 86040 }, { "grad_norm": 0.20719163119792938, "learning_rate": 5.227408002058881e-06, "loss": 0.0153, "step": 86050 }, { "grad_norm": 0.250512957572937, "learning_rate": 5.220049890517126e-06, "loss": 0.0117, "step": 86060 }, { "grad_norm": 0.25263339281082153, "learning_rate": 5.212696676043377e-06, "loss": 0.0162, "step": 86070 }, { "grad_norm": 0.09284847974777222, "learning_rate": 5.205348359441798e-06, "loss": 0.0128, "step": 86080 }, { "grad_norm": 0.185678631067276, "learning_rate": 5.198004941515966e-06, "loss": 0.0111, "step": 86090 }, { "grad_norm": 0.1807076632976532, "learning_rate": 5.190666423068952e-06, "loss": 0.0145, "step": 86100 }, { "grad_norm": 0.30706340074539185, "learning_rate": 5.183332804903284e-06, "loss": 0.0191, "step": 86110 }, { "grad_norm": 0.22089534997940063, "learning_rate": 5.176004087820979e-06, "loss": 0.0128, "step": 86120 }, { "grad_norm": 0.11780764907598495, "learning_rate": 5.168680272623455e-06, "loss": 0.0101, "step": 86130 }, { "grad_norm": 0.2558825612068176, "learning_rate": 5.161361360111661e-06, "loss": 0.018, "step": 86140 }, { "grad_norm": 0.20006680488586426, "learning_rate": 5.154047351085984e-06, "loss": 0.0129, "step": 86150 }, { "grad_norm": 0.29266414046287537, "learning_rate": 5.1467382463462385e-06, "loss": 0.0116, "step": 86160 }, { "grad_norm": 0.13208770751953125, "learning_rate": 5.139434046691788e-06, "loss": 0.0139, "step": 86170 }, { "grad_norm": 0.13824708759784698, "learning_rate": 5.13213475292137e-06, "loss": 0.0119, "step": 86180 }, { "grad_norm": 0.12249508500099182, "learning_rate": 5.12484036583325e-06, "loss": 0.0147, "step": 86190 }, { "grad_norm": 0.18772253394126892, "learning_rate": 5.117550886225109e-06, "loss": 0.0138, "step": 86200 }, { "grad_norm": 0.29210761189460754, "learning_rate": 5.11026631489413e-06, "loss": 0.0131, "step": 86210 }, { "grad_norm": 0.1748363971710205, "learning_rate": 5.102986652636926e-06, "loss": 0.0111, "step": 86220 }, { "grad_norm": 0.21676304936408997, "learning_rate": 5.095711900249606e-06, "loss": 0.0146, "step": 86230 }, { "grad_norm": 0.29227977991104126, "learning_rate": 5.088442058527715e-06, "loss": 0.0147, "step": 86240 }, { "grad_norm": 0.18029364943504333, "learning_rate": 5.08117712826629e-06, "loss": 0.0137, "step": 86250 }, { "grad_norm": 0.23221273720264435, "learning_rate": 5.073917110259785e-06, "loss": 0.0148, "step": 86260 }, { "grad_norm": 0.12016543745994568, "learning_rate": 5.066662005302175e-06, "loss": 0.0138, "step": 86270 }, { "grad_norm": 0.07767312228679657, "learning_rate": 5.059411814186838e-06, "loss": 0.014, "step": 86280 }, { "grad_norm": 0.19814257323741913, "learning_rate": 5.05216653770666e-06, "loss": 0.0142, "step": 86290 }, { "grad_norm": 0.16915355622768402, "learning_rate": 5.044926176653969e-06, "loss": 0.0103, "step": 86300 }, { "grad_norm": 0.15473677217960358, "learning_rate": 5.037690731820571e-06, "loss": 0.0125, "step": 86310 }, { "grad_norm": 0.14804551005363464, "learning_rate": 5.030460203997711e-06, "loss": 0.0164, "step": 86320 }, { "grad_norm": 0.2049851268529892, "learning_rate": 5.0232345939761174e-06, "loss": 0.0135, "step": 86330 }, { "grad_norm": 0.19463534653186798, "learning_rate": 5.016013902545957e-06, "loss": 0.0176, "step": 86340 }, { "grad_norm": 0.13811582326889038, "learning_rate": 5.008798130496889e-06, "loss": 0.0136, "step": 86350 }, { "grad_norm": 0.2142954021692276, "learning_rate": 5.001587278618008e-06, "loss": 0.0154, "step": 86360 }, { "grad_norm": 0.22929179668426514, "learning_rate": 4.994381347697891e-06, "loss": 0.0155, "step": 86370 }, { "grad_norm": 0.1779121309518814, "learning_rate": 4.987180338524572e-06, "loss": 0.0137, "step": 86380 }, { "grad_norm": 0.16819767653942108, "learning_rate": 4.9799842518855275e-06, "loss": 0.0123, "step": 86390 }, { "grad_norm": 0.2403058558702469, "learning_rate": 4.9727930885677285e-06, "loss": 0.0123, "step": 86400 }, { "grad_norm": 0.12766605615615845, "learning_rate": 4.965606849357573e-06, "loss": 0.0172, "step": 86410 }, { "grad_norm": 0.26852869987487793, "learning_rate": 4.9584255350409445e-06, "loss": 0.0094, "step": 86420 }, { "grad_norm": 0.18349944055080414, "learning_rate": 4.951249146403175e-06, "loss": 0.0138, "step": 86430 }, { "grad_norm": 0.4269252121448517, "learning_rate": 4.9440776842290815e-06, "loss": 0.0133, "step": 86440 }, { "grad_norm": 0.1259334534406662, "learning_rate": 4.936911149302903e-06, "loss": 0.011, "step": 86450 }, { "grad_norm": 0.4805227816104889, "learning_rate": 4.929749542408379e-06, "loss": 0.0123, "step": 86460 }, { "grad_norm": 0.28702518343925476, "learning_rate": 4.922592864328673e-06, "loss": 0.012, "step": 86470 }, { "grad_norm": 0.1597287952899933, "learning_rate": 4.915441115846436e-06, "loss": 0.0168, "step": 86480 }, { "grad_norm": 0.24626290798187256, "learning_rate": 4.908294297743782e-06, "loss": 0.0233, "step": 86490 }, { "grad_norm": 0.1651109904050827, "learning_rate": 4.901152410802273e-06, "loss": 0.013, "step": 86500 }, { "grad_norm": 0.3030761182308197, "learning_rate": 4.894015455802919e-06, "loss": 0.0124, "step": 86510 }, { "grad_norm": 0.15857695043087006, "learning_rate": 4.8868834335262346e-06, "loss": 0.0105, "step": 86520 }, { "grad_norm": 0.46777623891830444, "learning_rate": 4.87975634475214e-06, "loss": 0.0137, "step": 86530 }, { "grad_norm": 0.18059182167053223, "learning_rate": 4.8726341902600614e-06, "loss": 0.0121, "step": 86540 }, { "grad_norm": 0.10724338889122009, "learning_rate": 4.865516970828843e-06, "loss": 0.0142, "step": 86550 }, { "grad_norm": 0.2520970106124878, "learning_rate": 4.85840468723685e-06, "loss": 0.0126, "step": 86560 }, { "grad_norm": 0.4206370711326599, "learning_rate": 4.85129734026184e-06, "loss": 0.0139, "step": 86570 }, { "grad_norm": 0.20382560789585114, "learning_rate": 4.844194930681073e-06, "loss": 0.0102, "step": 86580 }, { "grad_norm": 0.1902748942375183, "learning_rate": 4.837097459271267e-06, "loss": 0.0136, "step": 86590 }, { "grad_norm": 0.11752405017614365, "learning_rate": 4.830004926808573e-06, "loss": 0.0132, "step": 86600 }, { "grad_norm": 0.2887931764125824, "learning_rate": 4.822917334068627e-06, "loss": 0.0151, "step": 86610 }, { "grad_norm": 0.1683429330587387, "learning_rate": 4.81583468182652e-06, "loss": 0.0083, "step": 86620 }, { "grad_norm": 0.13308516144752502, "learning_rate": 4.808756970856809e-06, "loss": 0.012, "step": 86630 }, { "grad_norm": 0.20183660089969635, "learning_rate": 4.801684201933476e-06, "loss": 0.0136, "step": 86640 }, { "grad_norm": 0.11758401989936829, "learning_rate": 4.794616375830019e-06, "loss": 0.0094, "step": 86650 }, { "grad_norm": 0.14715030789375305, "learning_rate": 4.787553493319341e-06, "loss": 0.0122, "step": 86660 }, { "grad_norm": 0.11323495209217072, "learning_rate": 4.780495555173836e-06, "loss": 0.0135, "step": 86670 }, { "grad_norm": 0.25685611367225647, "learning_rate": 4.7734425621653475e-06, "loss": 0.0146, "step": 86680 }, { "grad_norm": 0.18478284776210785, "learning_rate": 4.766394515065198e-06, "loss": 0.0125, "step": 86690 }, { "grad_norm": 0.28593358397483826, "learning_rate": 4.75935141464412e-06, "loss": 0.0158, "step": 86700 }, { "grad_norm": 0.2960527837276459, "learning_rate": 4.75231326167237e-06, "loss": 0.0118, "step": 86710 }, { "grad_norm": 0.4063073992729187, "learning_rate": 4.745280056919599e-06, "loss": 0.0159, "step": 86720 }, { "grad_norm": 0.20181210339069366, "learning_rate": 4.738251801154975e-06, "loss": 0.0152, "step": 86730 }, { "grad_norm": 0.3159126937389374, "learning_rate": 4.731228495147061e-06, "loss": 0.0124, "step": 86740 }, { "grad_norm": 0.2863233685493469, "learning_rate": 4.7242101396639595e-06, "loss": 0.0117, "step": 86750 }, { "grad_norm": 0.09199266135692596, "learning_rate": 4.717196735473151e-06, "loss": 0.0086, "step": 86760 }, { "grad_norm": 0.11877618730068207, "learning_rate": 4.710188283341643e-06, "loss": 0.0136, "step": 86770 }, { "grad_norm": 0.22144059836864471, "learning_rate": 4.703184784035835e-06, "loss": 0.0126, "step": 86780 }, { "grad_norm": 0.5445333123207092, "learning_rate": 4.696186238321642e-06, "loss": 0.0124, "step": 86790 }, { "grad_norm": 0.2421153038740158, "learning_rate": 4.689192646964408e-06, "loss": 0.0212, "step": 86800 }, { "grad_norm": 0.22305013239383698, "learning_rate": 4.682204010728952e-06, "loss": 0.0153, "step": 86810 }, { "grad_norm": 0.45539751648902893, "learning_rate": 4.67522033037952e-06, "loss": 0.0123, "step": 86820 }, { "grad_norm": 0.3734687864780426, "learning_rate": 4.668241606679846e-06, "loss": 0.0107, "step": 86830 }, { "grad_norm": 0.21420243382453918, "learning_rate": 4.66126784039313e-06, "loss": 0.0167, "step": 86840 }, { "grad_norm": 0.5899531841278076, "learning_rate": 4.6542990322819836e-06, "loss": 0.0155, "step": 86850 }, { "grad_norm": 0.15073181688785553, "learning_rate": 4.64733518310852e-06, "loss": 0.0175, "step": 86860 }, { "grad_norm": 0.23555587232112885, "learning_rate": 4.640376293634291e-06, "loss": 0.0179, "step": 86870 }, { "grad_norm": 0.3008974492549896, "learning_rate": 4.633422364620327e-06, "loss": 0.0124, "step": 86880 }, { "grad_norm": 0.2968873679637909, "learning_rate": 4.6264733968270745e-06, "loss": 0.0127, "step": 86890 }, { "grad_norm": 0.44076529145240784, "learning_rate": 4.619529391014482e-06, "loss": 0.0228, "step": 86900 }, { "grad_norm": 0.19255506992340088, "learning_rate": 4.61259034794192e-06, "loss": 0.0115, "step": 86910 }, { "grad_norm": 0.32998353242874146, "learning_rate": 4.605656268368247e-06, "loss": 0.0213, "step": 86920 }, { "grad_norm": 0.26272356510162354, "learning_rate": 4.598727153051735e-06, "loss": 0.0121, "step": 86930 }, { "grad_norm": 0.09968622028827667, "learning_rate": 4.591803002750178e-06, "loss": 0.0132, "step": 86940 }, { "grad_norm": 0.17195139825344086, "learning_rate": 4.58488381822077e-06, "loss": 0.011, "step": 86950 }, { "grad_norm": 0.15897172689437866, "learning_rate": 4.5779696002201946e-06, "loss": 0.0119, "step": 86960 }, { "grad_norm": 0.2611728310585022, "learning_rate": 4.571060349504563e-06, "loss": 0.0108, "step": 86970 }, { "grad_norm": 1.2425992488861084, "learning_rate": 4.564156066829478e-06, "loss": 0.0197, "step": 86980 }, { "grad_norm": 0.42339232563972473, "learning_rate": 4.557256752949957e-06, "loss": 0.0191, "step": 86990 }, { "grad_norm": 0.27921855449676514, "learning_rate": 4.550362408620534e-06, "loss": 0.0134, "step": 87000 }, { "grad_norm": 0.28281447291374207, "learning_rate": 4.543473034595136e-06, "loss": 0.0205, "step": 87010 }, { "grad_norm": 0.24177441000938416, "learning_rate": 4.536588631627192e-06, "loss": 0.0158, "step": 87020 }, { "grad_norm": 0.14012400805950165, "learning_rate": 4.5297092004695565e-06, "loss": 0.01, "step": 87030 }, { "grad_norm": 0.22023126482963562, "learning_rate": 4.522834741874555e-06, "loss": 0.0143, "step": 87040 }, { "grad_norm": 0.1633271425962448, "learning_rate": 4.515965256593974e-06, "loss": 0.0103, "step": 87050 }, { "grad_norm": 0.22748243808746338, "learning_rate": 4.509100745379058e-06, "loss": 0.0152, "step": 87060 }, { "grad_norm": 0.16425155103206635, "learning_rate": 4.502241208980479e-06, "loss": 0.01, "step": 87070 }, { "grad_norm": 0.19903263449668884, "learning_rate": 4.495386648148398e-06, "loss": 0.012, "step": 87080 }, { "grad_norm": 0.19911620020866394, "learning_rate": 4.488537063632431e-06, "loss": 0.0119, "step": 87090 }, { "grad_norm": 0.17791913449764252, "learning_rate": 4.481692456181608e-06, "loss": 0.0138, "step": 87100 }, { "grad_norm": 0.17979170382022858, "learning_rate": 4.474852826544479e-06, "loss": 0.0107, "step": 87110 }, { "grad_norm": 0.2824729084968567, "learning_rate": 4.468018175468974e-06, "loss": 0.0168, "step": 87120 }, { "grad_norm": 0.3279064893722534, "learning_rate": 4.461188503702568e-06, "loss": 0.017, "step": 87130 }, { "grad_norm": 0.1989942193031311, "learning_rate": 4.454363811992107e-06, "loss": 0.0154, "step": 87140 }, { "grad_norm": 0.1891358345746994, "learning_rate": 4.447544101083956e-06, "loss": 0.013, "step": 87150 }, { "grad_norm": 0.22662243247032166, "learning_rate": 4.440729371723884e-06, "loss": 0.0163, "step": 87160 }, { "grad_norm": 0.35259339213371277, "learning_rate": 4.433919624657162e-06, "loss": 0.0098, "step": 87170 }, { "grad_norm": 0.42792364954948425, "learning_rate": 4.427114860628462e-06, "loss": 0.0138, "step": 87180 }, { "grad_norm": 0.21472014486789703, "learning_rate": 4.420315080381982e-06, "loss": 0.0133, "step": 87190 }, { "grad_norm": 0.18572503328323364, "learning_rate": 4.4135202846613045e-06, "loss": 0.0149, "step": 87200 }, { "grad_norm": 0.10867653042078018, "learning_rate": 4.406730474209525e-06, "loss": 0.0126, "step": 87210 }, { "grad_norm": 0.28452154994010925, "learning_rate": 4.399945649769138e-06, "loss": 0.0124, "step": 87220 }, { "grad_norm": 0.2443116009235382, "learning_rate": 4.393165812082151e-06, "loss": 0.0134, "step": 87230 }, { "grad_norm": 0.23277944326400757, "learning_rate": 4.386390961889969e-06, "loss": 0.0245, "step": 87240 }, { "grad_norm": 0.1343553215265274, "learning_rate": 4.3796210999334895e-06, "loss": 0.0127, "step": 87250 }, { "grad_norm": 0.14379867911338806, "learning_rate": 4.372856226953065e-06, "loss": 0.0126, "step": 87260 }, { "grad_norm": 0.17828771471977234, "learning_rate": 4.366096343688486e-06, "loss": 0.0148, "step": 87270 }, { "grad_norm": 0.19063180685043335, "learning_rate": 4.359341450878995e-06, "loss": 0.0144, "step": 87280 }, { "grad_norm": 0.33301180601119995, "learning_rate": 4.352591549263302e-06, "loss": 0.0149, "step": 87290 }, { "grad_norm": 0.27183249592781067, "learning_rate": 4.345846639579582e-06, "loss": 0.0125, "step": 87300 }, { "grad_norm": 0.18415294587612152, "learning_rate": 4.339106722565417e-06, "loss": 0.0149, "step": 87310 }, { "grad_norm": 0.22285524010658264, "learning_rate": 4.332371798957896e-06, "loss": 0.0164, "step": 87320 }, { "grad_norm": 0.3073655068874359, "learning_rate": 4.3256418694935295e-06, "loss": 0.0106, "step": 87330 }, { "grad_norm": 0.20378363132476807, "learning_rate": 4.3189169349083125e-06, "loss": 0.0124, "step": 87340 }, { "grad_norm": 0.35521048307418823, "learning_rate": 4.312196995937645e-06, "loss": 0.0115, "step": 87350 }, { "grad_norm": 0.17887921631336212, "learning_rate": 4.305482053316434e-06, "loss": 0.0115, "step": 87360 }, { "grad_norm": 0.2904885709285736, "learning_rate": 4.2987721077789925e-06, "loss": 0.013, "step": 87370 }, { "grad_norm": 0.5189143419265747, "learning_rate": 4.2920671600591325e-06, "loss": 0.0193, "step": 87380 }, { "grad_norm": 0.9315716624259949, "learning_rate": 4.285367210890084e-06, "loss": 0.0144, "step": 87390 }, { "grad_norm": 0.21406912803649902, "learning_rate": 4.278672261004551e-06, "loss": 0.0202, "step": 87400 }, { "grad_norm": 0.0785181075334549, "learning_rate": 4.2719823111346704e-06, "loss": 0.0103, "step": 87410 }, { "grad_norm": 0.24886392056941986, "learning_rate": 4.265297362012061e-06, "loss": 0.0148, "step": 87420 }, { "grad_norm": 0.2683662176132202, "learning_rate": 4.258617414367766e-06, "loss": 0.0147, "step": 87430 }, { "grad_norm": 0.24737638235092163, "learning_rate": 4.251942468932296e-06, "loss": 0.0148, "step": 87440 }, { "grad_norm": 0.1945480853319168, "learning_rate": 4.2452725264356175e-06, "loss": 0.0153, "step": 87450 }, { "grad_norm": 0.23180435597896576, "learning_rate": 4.2386075876071575e-06, "loss": 0.0109, "step": 87460 }, { "grad_norm": 0.28432783484458923, "learning_rate": 4.231947653175755e-06, "loss": 0.0101, "step": 87470 }, { "grad_norm": 0.19914574921131134, "learning_rate": 4.225292723869761e-06, "loss": 0.0147, "step": 87480 }, { "grad_norm": 0.28253450989723206, "learning_rate": 4.218642800416916e-06, "loss": 0.0116, "step": 87490 }, { "grad_norm": 0.21655862033367157, "learning_rate": 4.211997883544472e-06, "loss": 0.0149, "step": 87500 }, { "grad_norm": 0.15007445216178894, "learning_rate": 4.205357973979096e-06, "loss": 0.0118, "step": 87510 }, { "grad_norm": 0.17674867808818817, "learning_rate": 4.198723072446925e-06, "loss": 0.0125, "step": 87520 }, { "grad_norm": 0.12226659804582596, "learning_rate": 4.192093179673534e-06, "loss": 0.0135, "step": 87530 }, { "grad_norm": 0.1587405651807785, "learning_rate": 4.185468296383959e-06, "loss": 0.013, "step": 87540 }, { "grad_norm": 0.14837957918643951, "learning_rate": 4.1788484233027005e-06, "loss": 0.0119, "step": 87550 }, { "grad_norm": 0.21146829426288605, "learning_rate": 4.172233561153677e-06, "loss": 0.0168, "step": 87560 }, { "grad_norm": 0.17636114358901978, "learning_rate": 4.165623710660288e-06, "loss": 0.0105, "step": 87570 }, { "grad_norm": 0.16761371493339539, "learning_rate": 4.1590188725453835e-06, "loss": 0.0109, "step": 87580 }, { "grad_norm": 0.42323175072669983, "learning_rate": 4.152419047531259e-06, "loss": 0.0169, "step": 87590 }, { "grad_norm": 0.303256630897522, "learning_rate": 4.145824236339651e-06, "loss": 0.0144, "step": 87600 }, { "grad_norm": 0.16474173963069916, "learning_rate": 4.1392344396917695e-06, "loss": 0.0142, "step": 87610 }, { "grad_norm": 0.21323226392269135, "learning_rate": 4.132649658308252e-06, "loss": 0.0165, "step": 87620 }, { "grad_norm": 0.2516479194164276, "learning_rate": 4.126069892909213e-06, "loss": 0.0159, "step": 87630 }, { "grad_norm": 0.1494232714176178, "learning_rate": 4.119495144214191e-06, "loss": 0.0146, "step": 87640 }, { "grad_norm": 0.232207790017128, "learning_rate": 4.112925412942214e-06, "loss": 0.0133, "step": 87650 }, { "grad_norm": 0.1090439036488533, "learning_rate": 4.106360699811712e-06, "loss": 0.01, "step": 87660 }, { "grad_norm": 0.14218440651893616, "learning_rate": 4.099801005540616e-06, "loss": 0.0145, "step": 87670 }, { "grad_norm": 0.23214782774448395, "learning_rate": 4.093246330846256e-06, "loss": 0.0144, "step": 87680 }, { "grad_norm": 0.1462831199169159, "learning_rate": 4.086696676445467e-06, "loss": 0.0138, "step": 87690 }, { "grad_norm": 0.37517932057380676, "learning_rate": 4.0801520430544914e-06, "loss": 0.0174, "step": 87700 }, { "grad_norm": 0.11784428358078003, "learning_rate": 4.073612431389062e-06, "loss": 0.0139, "step": 87710 }, { "grad_norm": 0.24070370197296143, "learning_rate": 4.067077842164319e-06, "loss": 0.0185, "step": 87720 }, { "grad_norm": 0.2680981755256653, "learning_rate": 4.0605482760948945e-06, "loss": 0.0159, "step": 87730 }, { "grad_norm": 0.39782947301864624, "learning_rate": 4.054023733894829e-06, "loss": 0.0162, "step": 87740 }, { "grad_norm": 0.1507306694984436, "learning_rate": 4.047504216277653e-06, "loss": 0.0136, "step": 87750 }, { "grad_norm": 0.31767547130584717, "learning_rate": 4.040989723956323e-06, "loss": 0.0184, "step": 87760 }, { "grad_norm": 0.15987379848957062, "learning_rate": 4.03448025764327e-06, "loss": 0.0138, "step": 87770 }, { "grad_norm": 0.3833370506763458, "learning_rate": 4.027975818050339e-06, "loss": 0.0154, "step": 87780 }, { "grad_norm": 0.14940916001796722, "learning_rate": 4.021476405888858e-06, "loss": 0.0119, "step": 87790 }, { "grad_norm": 0.3399578928947449, "learning_rate": 4.01498202186959e-06, "loss": 0.0124, "step": 87800 }, { "grad_norm": 0.22391454875469208, "learning_rate": 4.008492666702751e-06, "loss": 0.0099, "step": 87810 }, { "grad_norm": 0.2178000658750534, "learning_rate": 4.002008341098001e-06, "loss": 0.0155, "step": 87820 }, { "grad_norm": 0.1775415539741516, "learning_rate": 3.995529045764462e-06, "loss": 0.0116, "step": 87830 }, { "grad_norm": 0.3153533637523651, "learning_rate": 3.989054781410712e-06, "loss": 0.013, "step": 87840 }, { "grad_norm": 0.24399444460868835, "learning_rate": 3.982585548744744e-06, "loss": 0.0136, "step": 87850 }, { "grad_norm": 0.12828965485095978, "learning_rate": 3.9761213484740435e-06, "loss": 0.0095, "step": 87860 }, { "grad_norm": 0.18758173286914825, "learning_rate": 3.9696621813055055e-06, "loss": 0.0106, "step": 87870 }, { "grad_norm": 0.27725327014923096, "learning_rate": 3.9632080479455044e-06, "loss": 0.0118, "step": 87880 }, { "grad_norm": 0.3563957214355469, "learning_rate": 3.956758949099859e-06, "loss": 0.0141, "step": 87890 }, { "grad_norm": 0.2406684011220932, "learning_rate": 3.9503148854738385e-06, "loss": 0.0133, "step": 87900 }, { "grad_norm": 0.24746978282928467, "learning_rate": 3.943875857772139e-06, "loss": 0.0257, "step": 87910 }, { "grad_norm": 0.16550272703170776, "learning_rate": 3.9374418666989375e-06, "loss": 0.0121, "step": 87920 }, { "grad_norm": 0.1926887482404709, "learning_rate": 3.931012912957838e-06, "loss": 0.012, "step": 87930 }, { "grad_norm": 0.21518594026565552, "learning_rate": 3.9245889972519045e-06, "loss": 0.0184, "step": 87940 }, { "grad_norm": 0.43348428606987, "learning_rate": 3.9181701202836265e-06, "loss": 0.0106, "step": 87950 }, { "grad_norm": 0.2563555836677551, "learning_rate": 3.911756282755003e-06, "loss": 0.0134, "step": 87960 }, { "grad_norm": 0.5095511078834534, "learning_rate": 3.905347485367411e-06, "loss": 0.0141, "step": 87970 }, { "grad_norm": 0.29294902086257935, "learning_rate": 3.898943728821725e-06, "loss": 0.0131, "step": 87980 }, { "grad_norm": 0.2658323049545288, "learning_rate": 3.892545013818227e-06, "loss": 0.0146, "step": 87990 }, { "grad_norm": 0.13861589133739471, "learning_rate": 3.886151341056687e-06, "loss": 0.0172, "step": 88000 }, { "grad_norm": 0.16552644968032837, "learning_rate": 3.8797627112363066e-06, "loss": 0.0151, "step": 88010 }, { "grad_norm": 0.1205802634358406, "learning_rate": 3.873379125055743e-06, "loss": 0.0123, "step": 88020 }, { "grad_norm": 0.1951959878206253, "learning_rate": 3.867000583213082e-06, "loss": 0.0141, "step": 88030 }, { "grad_norm": 0.15236876904964447, "learning_rate": 3.860627086405871e-06, "loss": 0.0118, "step": 88040 }, { "grad_norm": 0.23091822862625122, "learning_rate": 3.854258635331126e-06, "loss": 0.013, "step": 88050 }, { "grad_norm": 0.24343229830265045, "learning_rate": 3.8478952306852665e-06, "loss": 0.0154, "step": 88060 }, { "grad_norm": 0.23799392580986023, "learning_rate": 3.841536873164198e-06, "loss": 0.0173, "step": 88070 }, { "grad_norm": 0.2853836715221405, "learning_rate": 3.8351835634632626e-06, "loss": 0.0148, "step": 88080 }, { "grad_norm": 0.21233035624027252, "learning_rate": 3.828835302277245e-06, "loss": 0.0139, "step": 88090 }, { "grad_norm": 0.15118080377578735, "learning_rate": 3.822492090300378e-06, "loss": 0.0124, "step": 88100 }, { "grad_norm": 0.16863925755023956, "learning_rate": 3.8161539282263566e-06, "loss": 0.011, "step": 88110 }, { "grad_norm": 0.2524709105491638, "learning_rate": 3.8098208167482995e-06, "loss": 0.0203, "step": 88120 }, { "grad_norm": 0.6431238055229187, "learning_rate": 3.803492756558796e-06, "loss": 0.0109, "step": 88130 }, { "grad_norm": 0.26530444622039795, "learning_rate": 3.797169748349855e-06, "loss": 0.0107, "step": 88140 }, { "grad_norm": 0.2832159101963043, "learning_rate": 3.790851792812977e-06, "loss": 0.0113, "step": 88150 }, { "grad_norm": 0.3825232982635498, "learning_rate": 3.784538890639072e-06, "loss": 0.0178, "step": 88160 }, { "grad_norm": 0.24076396226882935, "learning_rate": 3.7782310425185153e-06, "loss": 0.0149, "step": 88170 }, { "grad_norm": 0.09640849381685257, "learning_rate": 3.77192824914111e-06, "loss": 0.0141, "step": 88180 }, { "grad_norm": 0.11343994736671448, "learning_rate": 3.7656305111961333e-06, "loss": 0.0113, "step": 88190 }, { "grad_norm": 0.23821236193180084, "learning_rate": 3.7593378293722782e-06, "loss": 0.0123, "step": 88200 }, { "grad_norm": 0.11360049247741699, "learning_rate": 3.7530502043577333e-06, "loss": 0.0119, "step": 88210 }, { "grad_norm": 0.1260053515434265, "learning_rate": 3.746767636840076e-06, "loss": 0.0134, "step": 88220 }, { "grad_norm": 0.14395517110824585, "learning_rate": 3.74049012750638e-06, "loss": 0.0134, "step": 88230 }, { "grad_norm": 0.2924899160861969, "learning_rate": 3.734217677043128e-06, "loss": 0.0103, "step": 88240 }, { "grad_norm": 0.14849478006362915, "learning_rate": 3.727950286136267e-06, "loss": 0.0115, "step": 88250 }, { "grad_norm": 0.15667246282100677, "learning_rate": 3.7216879554711925e-06, "loss": 0.013, "step": 88260 }, { "grad_norm": 0.2536674737930298, "learning_rate": 3.715430685732757e-06, "loss": 0.0172, "step": 88270 }, { "grad_norm": 0.18049050867557526, "learning_rate": 3.7091784776052243e-06, "loss": 0.0127, "step": 88280 }, { "grad_norm": 0.43703335523605347, "learning_rate": 3.7029313317723356e-06, "loss": 0.0129, "step": 88290 }, { "grad_norm": 0.2613210380077362, "learning_rate": 3.6966892489172787e-06, "loss": 0.0185, "step": 88300 }, { "grad_norm": 0.1741558462381363, "learning_rate": 3.6904522297226575e-06, "loss": 0.0112, "step": 88310 }, { "grad_norm": 0.2605130076408386, "learning_rate": 3.6842202748705647e-06, "loss": 0.0121, "step": 88320 }, { "grad_norm": 0.24064017832279205, "learning_rate": 3.677993385042494e-06, "loss": 0.012, "step": 88330 }, { "grad_norm": 0.13851167261600494, "learning_rate": 3.67177156091944e-06, "loss": 0.0148, "step": 88340 }, { "grad_norm": 0.22231361269950867, "learning_rate": 3.6655548031817866e-06, "loss": 0.0155, "step": 88350 }, { "grad_norm": 0.19685465097427368, "learning_rate": 3.6593431125094e-06, "loss": 0.012, "step": 88360 }, { "grad_norm": 0.3708641827106476, "learning_rate": 3.6531364895815766e-06, "loss": 0.0121, "step": 88370 }, { "grad_norm": 0.12507744133472443, "learning_rate": 3.6469349350770677e-06, "loss": 0.0116, "step": 88380 }, { "grad_norm": 0.2632950246334076, "learning_rate": 3.6407384496740525e-06, "loss": 0.0217, "step": 88390 }, { "grad_norm": 0.19593048095703125, "learning_rate": 3.6345470340501954e-06, "loss": 0.0141, "step": 88400 }, { "grad_norm": 0.40511345863342285, "learning_rate": 3.6283606888825537e-06, "loss": 0.0147, "step": 88410 }, { "grad_norm": 0.965112030506134, "learning_rate": 3.622179414847676e-06, "loss": 0.0201, "step": 88420 }, { "grad_norm": 0.371987521648407, "learning_rate": 3.616003212621527e-06, "loss": 0.0218, "step": 88430 }, { "grad_norm": 0.5971371531486511, "learning_rate": 3.609832082879533e-06, "loss": 0.0157, "step": 88440 }, { "grad_norm": 0.14477863907814026, "learning_rate": 3.6036660262965485e-06, "loss": 0.0124, "step": 88450 }, { "grad_norm": 0.3998546600341797, "learning_rate": 3.597505043546895e-06, "loss": 0.0121, "step": 88460 }, { "grad_norm": 0.2414163500070572, "learning_rate": 3.5913491353043227e-06, "loss": 0.0106, "step": 88470 }, { "grad_norm": 0.20516741275787354, "learning_rate": 3.5851983022420375e-06, "loss": 0.0125, "step": 88480 }, { "grad_norm": 0.25739234685897827, "learning_rate": 3.5790525450326797e-06, "loss": 0.0135, "step": 88490 }, { "grad_norm": 0.16883093118667603, "learning_rate": 3.572911864348344e-06, "loss": 0.0141, "step": 88500 }, { "grad_norm": 0.5743488073348999, "learning_rate": 3.5667762608605717e-06, "loss": 0.0132, "step": 88510 }, { "grad_norm": 0.08108299225568771, "learning_rate": 3.5606457352403255e-06, "loss": 0.0114, "step": 88520 }, { "grad_norm": 0.13415715098381042, "learning_rate": 3.554520288158042e-06, "loss": 0.0102, "step": 88530 }, { "grad_norm": 0.21375487744808197, "learning_rate": 3.5483999202835903e-06, "loss": 0.0145, "step": 88540 }, { "grad_norm": 0.15928196907043457, "learning_rate": 3.5422846322862913e-06, "loss": 0.0163, "step": 88550 }, { "grad_norm": 0.1900145560503006, "learning_rate": 3.5361744248348925e-06, "loss": 0.0126, "step": 88560 }, { "grad_norm": 0.25781089067459106, "learning_rate": 3.5300692985976047e-06, "loss": 0.013, "step": 88570 }, { "grad_norm": 0.6247997283935547, "learning_rate": 3.52396925424206e-06, "loss": 0.0121, "step": 88580 }, { "grad_norm": 0.21805612742900848, "learning_rate": 3.5178742924353747e-06, "loss": 0.0142, "step": 88590 }, { "grad_norm": 0.1977132260799408, "learning_rate": 3.511784413844066e-06, "loss": 0.0122, "step": 88600 }, { "grad_norm": 0.7088650465011597, "learning_rate": 3.5056996191341287e-06, "loss": 0.0168, "step": 88610 }, { "grad_norm": 0.2035706788301468, "learning_rate": 3.4996199089709692e-06, "loss": 0.0124, "step": 88620 }, { "grad_norm": 0.24792999029159546, "learning_rate": 3.4935452840194725e-06, "loss": 0.0148, "step": 88630 }, { "grad_norm": 0.18088607490062714, "learning_rate": 3.48747574494393e-06, "loss": 0.0123, "step": 88640 }, { "grad_norm": 0.30827590823173523, "learning_rate": 3.4814112924081154e-06, "loss": 0.0127, "step": 88650 }, { "grad_norm": 0.2729593813419342, "learning_rate": 3.475351927075221e-06, "loss": 0.0137, "step": 88660 }, { "grad_norm": 0.3509972393512726, "learning_rate": 3.469297649607894e-06, "loss": 0.0165, "step": 88670 }, { "grad_norm": 0.22025234997272491, "learning_rate": 3.463248460668217e-06, "loss": 0.0124, "step": 88680 }, { "grad_norm": 0.20197561383247375, "learning_rate": 3.457204360917726e-06, "loss": 0.0164, "step": 88690 }, { "grad_norm": 0.2220529466867447, "learning_rate": 3.4511653510173827e-06, "loss": 0.0088, "step": 88700 }, { "grad_norm": 0.17579962313175201, "learning_rate": 3.4451314316276074e-06, "loss": 0.0168, "step": 88710 }, { "grad_norm": 0.2705823481082916, "learning_rate": 3.4391026034082674e-06, "loss": 0.0137, "step": 88720 }, { "grad_norm": 0.2556025981903076, "learning_rate": 3.433078867018674e-06, "loss": 0.0156, "step": 88730 }, { "grad_norm": 0.2638096511363983, "learning_rate": 3.427060223117551e-06, "loss": 0.012, "step": 88740 }, { "grad_norm": 0.17749446630477905, "learning_rate": 3.4210466723631042e-06, "loss": 0.0102, "step": 88750 }, { "grad_norm": 0.17175932228565216, "learning_rate": 3.415038215412969e-06, "loss": 0.0113, "step": 88760 }, { "grad_norm": 0.24938538670539856, "learning_rate": 3.4090348529242088e-06, "loss": 0.012, "step": 88770 }, { "grad_norm": 0.11232635378837585, "learning_rate": 3.4030365855533487e-06, "loss": 0.0167, "step": 88780 }, { "grad_norm": 0.2974066734313965, "learning_rate": 3.397043413956347e-06, "loss": 0.0146, "step": 88790 }, { "grad_norm": 0.12087040394544601, "learning_rate": 3.3910553387886237e-06, "loss": 0.0119, "step": 88800 }, { "grad_norm": 0.21336445212364197, "learning_rate": 3.3850723607049996e-06, "loss": 0.0133, "step": 88810 }, { "grad_norm": 0.2577914297580719, "learning_rate": 3.3790944803597902e-06, "loss": 0.0122, "step": 88820 }, { "grad_norm": 0.3438476026058197, "learning_rate": 3.373121698406706e-06, "loss": 0.0182, "step": 88830 }, { "grad_norm": 0.20597557723522186, "learning_rate": 3.3671540154989302e-06, "loss": 0.0137, "step": 88840 }, { "grad_norm": 0.19972214102745056, "learning_rate": 3.361191432289079e-06, "loss": 0.0146, "step": 88850 }, { "grad_norm": 0.1836048662662506, "learning_rate": 3.3552339494292207e-06, "loss": 0.0131, "step": 88860 }, { "grad_norm": 0.1608652025461197, "learning_rate": 3.349281567570839e-06, "loss": 0.0146, "step": 88870 }, { "grad_norm": 0.22178292274475098, "learning_rate": 3.3433342873648965e-06, "loss": 0.0146, "step": 88880 }, { "grad_norm": 0.28086134791374207, "learning_rate": 3.337392109461762e-06, "loss": 0.0123, "step": 88890 }, { "grad_norm": 0.14874829351902008, "learning_rate": 3.331455034511266e-06, "loss": 0.0178, "step": 88900 }, { "grad_norm": 0.09554997831583023, "learning_rate": 3.325523063162689e-06, "loss": 0.0111, "step": 88910 }, { "grad_norm": 0.3369692862033844, "learning_rate": 3.3195961960647393e-06, "loss": 0.0147, "step": 88920 }, { "grad_norm": 0.17616496980190277, "learning_rate": 3.31367443386556e-06, "loss": 0.0142, "step": 88930 }, { "grad_norm": 0.1919231414794922, "learning_rate": 3.3077577772127556e-06, "loss": 0.009, "step": 88940 }, { "grad_norm": 0.15891645848751068, "learning_rate": 3.301846226753358e-06, "loss": 0.0147, "step": 88950 }, { "grad_norm": 0.2598559260368347, "learning_rate": 3.2959397831338444e-06, "loss": 0.0158, "step": 88960 }, { "grad_norm": 0.2168455272912979, "learning_rate": 3.2900384470001432e-06, "loss": 0.0168, "step": 88970 }, { "grad_norm": 0.6505937576293945, "learning_rate": 3.28414221899761e-06, "loss": 0.0142, "step": 88980 }, { "grad_norm": 0.2351512312889099, "learning_rate": 3.278251099771046e-06, "loss": 0.0096, "step": 88990 }, { "grad_norm": 0.5954622626304626, "learning_rate": 3.2723650899646906e-06, "loss": 0.0249, "step": 89000 }, { "grad_norm": 0.30571597814559937, "learning_rate": 3.2664841902222466e-06, "loss": 0.0148, "step": 89010 }, { "grad_norm": 0.22143451869487762, "learning_rate": 3.2606084011868208e-06, "loss": 0.0182, "step": 89020 }, { "grad_norm": 0.2723664939403534, "learning_rate": 3.254737723500989e-06, "loss": 0.0204, "step": 89030 }, { "grad_norm": 0.21964916586875916, "learning_rate": 3.2488721578067595e-06, "loss": 0.0125, "step": 89040 }, { "grad_norm": 0.1239556148648262, "learning_rate": 3.243011704745591e-06, "loss": 0.0117, "step": 89050 }, { "grad_norm": 0.24286651611328125, "learning_rate": 3.237156364958355e-06, "loss": 0.0129, "step": 89060 }, { "grad_norm": 0.22073781490325928, "learning_rate": 3.231306139085394e-06, "loss": 0.0133, "step": 89070 }, { "grad_norm": 0.17754940688610077, "learning_rate": 3.2254610277664797e-06, "loss": 0.0128, "step": 89080 }, { "grad_norm": 0.1333211064338684, "learning_rate": 3.2196210316408183e-06, "loss": 0.0137, "step": 89090 }, { "grad_norm": 0.19712640345096588, "learning_rate": 3.2137861513470644e-06, "loss": 0.0132, "step": 89100 }, { "grad_norm": 0.22420506179332733, "learning_rate": 3.2079563875233252e-06, "loss": 0.0167, "step": 89110 }, { "grad_norm": 0.17246617376804352, "learning_rate": 3.2021317408071184e-06, "loss": 0.0122, "step": 89120 }, { "grad_norm": 0.18137329816818237, "learning_rate": 3.196312211835434e-06, "loss": 0.0089, "step": 89130 }, { "grad_norm": 0.17438629269599915, "learning_rate": 3.1904978012446686e-06, "loss": 0.0146, "step": 89140 }, { "grad_norm": 0.2518532872200012, "learning_rate": 3.184688509670691e-06, "loss": 0.0146, "step": 89150 }, { "grad_norm": 0.32676753401756287, "learning_rate": 3.178884337748783e-06, "loss": 0.0124, "step": 89160 }, { "grad_norm": 0.23623286187648773, "learning_rate": 3.173085286113697e-06, "loss": 0.0139, "step": 89170 }, { "grad_norm": 0.22891636192798615, "learning_rate": 3.1672913553995986e-06, "loss": 0.0109, "step": 89180 }, { "grad_norm": 0.11802387237548828, "learning_rate": 3.1615025462401138e-06, "loss": 0.0109, "step": 89190 }, { "grad_norm": 0.126401886343956, "learning_rate": 3.1557188592682808e-06, "loss": 0.0138, "step": 89200 }, { "grad_norm": 0.18780794739723206, "learning_rate": 3.1499402951166045e-06, "loss": 0.0123, "step": 89210 }, { "grad_norm": 0.15088383853435516, "learning_rate": 3.144166854417019e-06, "loss": 0.0136, "step": 89220 }, { "grad_norm": 0.21525812149047852, "learning_rate": 3.1383985378009073e-06, "loss": 0.0129, "step": 89230 }, { "grad_norm": 0.2765349745750427, "learning_rate": 3.1326353458990654e-06, "loss": 0.0137, "step": 89240 }, { "grad_norm": 0.1994163990020752, "learning_rate": 3.126877279341761e-06, "loss": 0.0117, "step": 89250 }, { "grad_norm": 0.3223077356815338, "learning_rate": 3.1211243387586964e-06, "loss": 0.0152, "step": 89260 }, { "grad_norm": 0.6033549308776855, "learning_rate": 3.115376524778979e-06, "loss": 0.0166, "step": 89270 }, { "grad_norm": 0.12743903696537018, "learning_rate": 3.109633838031195e-06, "loss": 0.0177, "step": 89280 }, { "grad_norm": 0.19735755026340485, "learning_rate": 3.103896279143359e-06, "loss": 0.0126, "step": 89290 }, { "grad_norm": 0.2749103605747223, "learning_rate": 3.098163848742924e-06, "loss": 0.0134, "step": 89300 }, { "grad_norm": 0.16008655726909637, "learning_rate": 3.0924365474567675e-06, "loss": 0.0125, "step": 89310 }, { "grad_norm": 0.4002965986728668, "learning_rate": 3.086714375911237e-06, "loss": 0.0113, "step": 89320 }, { "grad_norm": 0.13215015828609467, "learning_rate": 3.0809973347320775e-06, "loss": 0.0129, "step": 89330 }, { "grad_norm": 0.19562621414661407, "learning_rate": 3.075285424544516e-06, "loss": 0.0112, "step": 89340 }, { "grad_norm": 0.22874845564365387, "learning_rate": 3.069578645973176e-06, "loss": 0.0147, "step": 89350 }, { "grad_norm": 0.23373566567897797, "learning_rate": 3.0638769996421735e-06, "loss": 0.0132, "step": 89360 }, { "grad_norm": 0.3112083971500397, "learning_rate": 3.058180486175005e-06, "loss": 0.0164, "step": 89370 }, { "grad_norm": 0.2677735686302185, "learning_rate": 3.052489106194645e-06, "loss": 0.0128, "step": 89380 }, { "grad_norm": 0.11804037541151047, "learning_rate": 3.0468028603234946e-06, "loss": 0.0117, "step": 89390 }, { "grad_norm": 0.3012777268886566, "learning_rate": 3.04112174918339e-06, "loss": 0.0097, "step": 89400 }, { "grad_norm": 0.27138540148735046, "learning_rate": 3.0354457733955954e-06, "loss": 0.0121, "step": 89410 }, { "grad_norm": 0.23368749022483826, "learning_rate": 3.029774933580859e-06, "loss": 0.0099, "step": 89420 }, { "grad_norm": 0.09079050272703171, "learning_rate": 3.0241092303593065e-06, "loss": 0.0121, "step": 89430 }, { "grad_norm": 0.19085709750652313, "learning_rate": 3.018448664350554e-06, "loss": 0.0126, "step": 89440 }, { "grad_norm": 0.2621099054813385, "learning_rate": 3.0127932361736055e-06, "loss": 0.0143, "step": 89450 }, { "grad_norm": 0.34033721685409546, "learning_rate": 3.00714294644695e-06, "loss": 0.0127, "step": 89460 }, { "grad_norm": 0.21743890643119812, "learning_rate": 3.0014977957884828e-06, "loss": 0.0166, "step": 89470 }, { "grad_norm": 0.14878517389297485, "learning_rate": 2.99585778481557e-06, "loss": 0.0132, "step": 89480 }, { "grad_norm": 0.16172218322753906, "learning_rate": 2.990222914144969e-06, "loss": 0.0169, "step": 89490 }, { "grad_norm": 0.11045045405626297, "learning_rate": 2.984593184392914e-06, "loss": 0.0129, "step": 89500 }, { "grad_norm": 0.3645727336406708, "learning_rate": 2.978968596175069e-06, "loss": 0.0128, "step": 89510 }, { "grad_norm": 0.13922090828418732, "learning_rate": 2.9733491501065137e-06, "loss": 0.0107, "step": 89520 }, { "grad_norm": 0.15475380420684814, "learning_rate": 2.9677348468018005e-06, "loss": 0.0132, "step": 89530 }, { "grad_norm": 0.15645629167556763, "learning_rate": 2.9621256868748726e-06, "loss": 0.0093, "step": 89540 }, { "grad_norm": 0.18177197873592377, "learning_rate": 2.956521670939183e-06, "loss": 0.0127, "step": 89550 }, { "grad_norm": 0.39656659960746765, "learning_rate": 2.950922799607542e-06, "loss": 0.0215, "step": 89560 }, { "grad_norm": 0.1268470734357834, "learning_rate": 2.9453290734922535e-06, "loss": 0.011, "step": 89570 }, { "grad_norm": 0.4868472218513489, "learning_rate": 2.939740493205029e-06, "loss": 0.0172, "step": 89580 }, { "grad_norm": 0.10548987239599228, "learning_rate": 2.9341570593570345e-06, "loss": 0.014, "step": 89590 }, { "grad_norm": 0.4095280170440674, "learning_rate": 2.928578772558843e-06, "loss": 0.0149, "step": 89600 }, { "grad_norm": 0.3219032287597656, "learning_rate": 2.923005633420528e-06, "loss": 0.0123, "step": 89610 }, { "grad_norm": 0.2896403670310974, "learning_rate": 2.9174376425515237e-06, "loss": 0.0178, "step": 89620 }, { "grad_norm": 0.3386164605617523, "learning_rate": 2.91187480056076e-06, "loss": 0.0195, "step": 89630 }, { "grad_norm": 0.41678693890571594, "learning_rate": 2.9063171080565664e-06, "loss": 0.0162, "step": 89640 }, { "grad_norm": 0.1874707192182541, "learning_rate": 2.900764565646741e-06, "loss": 0.01, "step": 89650 }, { "grad_norm": 0.27857735753059387, "learning_rate": 2.895217173938475e-06, "loss": 0.0137, "step": 89660 }, { "grad_norm": 0.11485981941223145, "learning_rate": 2.8896749335384445e-06, "loss": 0.0141, "step": 89670 }, { "grad_norm": 0.19191782176494598, "learning_rate": 2.884137845052731e-06, "loss": 0.011, "step": 89680 }, { "grad_norm": 0.21249330043792725, "learning_rate": 2.8786059090868734e-06, "loss": 0.0121, "step": 89690 }, { "grad_norm": 0.12990830838680267, "learning_rate": 2.87307912624582e-06, "loss": 0.0127, "step": 89700 }, { "grad_norm": 0.2567714750766754, "learning_rate": 2.867557497133977e-06, "loss": 0.011, "step": 89710 }, { "grad_norm": 0.12127077579498291, "learning_rate": 2.862041022355194e-06, "loss": 0.0118, "step": 89720 }, { "grad_norm": 0.2480572760105133, "learning_rate": 2.8565297025127226e-06, "loss": 0.0107, "step": 89730 }, { "grad_norm": 0.2357439547777176, "learning_rate": 2.8510235382092853e-06, "loss": 0.0087, "step": 89740 }, { "grad_norm": 0.417365163564682, "learning_rate": 2.845522530047029e-06, "loss": 0.011, "step": 89750 }, { "grad_norm": 0.26623013615608215, "learning_rate": 2.8400266786275387e-06, "loss": 0.0138, "step": 89760 }, { "grad_norm": 0.2045985460281372, "learning_rate": 2.834535984551817e-06, "loss": 0.0144, "step": 89770 }, { "grad_norm": 0.2269923835992813, "learning_rate": 2.8290504484203397e-06, "loss": 0.0242, "step": 89780 }, { "grad_norm": 0.22014069557189941, "learning_rate": 2.8235700708329705e-06, "loss": 0.0134, "step": 89790 }, { "grad_norm": 0.17444944381713867, "learning_rate": 2.818094852389058e-06, "loss": 0.0168, "step": 89800 }, { "grad_norm": 0.4532504975795746, "learning_rate": 2.8126247936873516e-06, "loss": 0.0144, "step": 89810 }, { "grad_norm": 0.2058350294828415, "learning_rate": 2.8071598953260614e-06, "loss": 0.012, "step": 89820 }, { "grad_norm": 0.20739424228668213, "learning_rate": 2.8017001579028035e-06, "loss": 0.0135, "step": 89830 }, { "grad_norm": 0.2219105362892151, "learning_rate": 2.7962455820146617e-06, "loss": 0.0142, "step": 89840 }, { "grad_norm": 0.17688989639282227, "learning_rate": 2.7907961682581253e-06, "loss": 0.0124, "step": 89850 }, { "grad_norm": 0.19484122097492218, "learning_rate": 2.7853519172291453e-06, "loss": 0.0172, "step": 89860 }, { "grad_norm": 0.2087729573249817, "learning_rate": 2.7799128295230904e-06, "loss": 0.0101, "step": 89870 }, { "grad_norm": 0.6714837551116943, "learning_rate": 2.774478905734784e-06, "loss": 0.0122, "step": 89880 }, { "grad_norm": 0.23837214708328247, "learning_rate": 2.7690501464584516e-06, "loss": 0.0148, "step": 89890 }, { "grad_norm": 0.16798155009746552, "learning_rate": 2.763626552287796e-06, "loss": 0.0145, "step": 89900 }, { "grad_norm": 0.1562202125787735, "learning_rate": 2.7582081238159083e-06, "loss": 0.0111, "step": 89910 }, { "grad_norm": 0.19889624416828156, "learning_rate": 2.7527948616353548e-06, "loss": 0.0117, "step": 89920 }, { "grad_norm": 0.3007037341594696, "learning_rate": 2.7473867663381226e-06, "loss": 0.0144, "step": 89930 }, { "grad_norm": 0.20919747650623322, "learning_rate": 2.7419838385156383e-06, "loss": 0.0127, "step": 89940 }, { "grad_norm": 0.31122490763664246, "learning_rate": 2.7365860787587407e-06, "loss": 0.0169, "step": 89950 }, { "grad_norm": 0.2891768217086792, "learning_rate": 2.7311934876577295e-06, "loss": 0.0182, "step": 89960 }, { "grad_norm": 0.5807521343231201, "learning_rate": 2.725806065802339e-06, "loss": 0.0128, "step": 89970 }, { "grad_norm": 0.24028721451759338, "learning_rate": 2.720423813781714e-06, "loss": 0.0121, "step": 89980 }, { "grad_norm": 0.3352678418159485, "learning_rate": 2.715046732184462e-06, "loss": 0.0123, "step": 89990 }, { "grad_norm": 0.2678270637989044, "learning_rate": 2.7096748215986013e-06, "loss": 0.014, "step": 90000 }, { "grad_norm": 0.2174602597951889, "learning_rate": 2.704308082611612e-06, "loss": 0.012, "step": 90010 }, { "grad_norm": 0.2883131504058838, "learning_rate": 2.698946515810374e-06, "loss": 0.0106, "step": 90020 }, { "grad_norm": 0.21330757439136505, "learning_rate": 2.6935901217812367e-06, "loss": 0.0124, "step": 90030 }, { "grad_norm": 0.11512524634599686, "learning_rate": 2.688238901109952e-06, "loss": 0.0103, "step": 90040 }, { "grad_norm": 0.2702820301055908, "learning_rate": 2.68289285438173e-06, "loss": 0.016, "step": 90050 }, { "grad_norm": 0.41051462292671204, "learning_rate": 2.6775519821812033e-06, "loss": 0.0117, "step": 90060 }, { "grad_norm": 0.21754072606563568, "learning_rate": 2.672216285092449e-06, "loss": 0.0129, "step": 90070 }, { "grad_norm": 0.42432427406311035, "learning_rate": 2.666885763698962e-06, "loss": 0.0144, "step": 90080 }, { "grad_norm": 0.4462567865848541, "learning_rate": 2.661560418583692e-06, "loss": 0.017, "step": 90090 }, { "grad_norm": 0.6143259406089783, "learning_rate": 2.656240250328995e-06, "loss": 0.0137, "step": 90100 }, { "grad_norm": 0.2652808725833893, "learning_rate": 2.650925259516679e-06, "loss": 0.014, "step": 90110 }, { "grad_norm": 0.21832512319087982, "learning_rate": 2.645615446727995e-06, "loss": 0.0157, "step": 90120 }, { "grad_norm": 0.19395072758197784, "learning_rate": 2.6403108125436115e-06, "loss": 0.0122, "step": 90130 }, { "grad_norm": 0.15757130086421967, "learning_rate": 2.6350113575436264e-06, "loss": 0.0133, "step": 90140 }, { "grad_norm": 0.16571113467216492, "learning_rate": 2.629717082307598e-06, "loss": 0.0146, "step": 90150 }, { "grad_norm": 0.18639568984508514, "learning_rate": 2.6244279874144793e-06, "loss": 0.0137, "step": 90160 }, { "grad_norm": 0.13292711973190308, "learning_rate": 2.6191440734426862e-06, "loss": 0.0124, "step": 90170 }, { "grad_norm": 0.17694754898548126, "learning_rate": 2.613865340970062e-06, "loss": 0.0134, "step": 90180 }, { "grad_norm": 0.23551493883132935, "learning_rate": 2.608591790573889e-06, "loss": 0.0118, "step": 90190 }, { "grad_norm": 0.17489375174045563, "learning_rate": 2.603323422830861e-06, "loss": 0.0122, "step": 90200 }, { "grad_norm": 0.1744665801525116, "learning_rate": 2.5980602383171183e-06, "loss": 0.0137, "step": 90210 }, { "grad_norm": 0.21826784312725067, "learning_rate": 2.5928022376082496e-06, "loss": 0.0133, "step": 90220 }, { "grad_norm": 0.1389678716659546, "learning_rate": 2.5875494212792505e-06, "loss": 0.0087, "step": 90230 }, { "grad_norm": 0.18620474636554718, "learning_rate": 2.5823017899045565e-06, "loss": 0.0138, "step": 90240 }, { "grad_norm": 0.3458799421787262, "learning_rate": 2.577059344058047e-06, "loss": 0.013, "step": 90250 }, { "grad_norm": 0.2489006072282791, "learning_rate": 2.5718220843130415e-06, "loss": 0.0124, "step": 90260 }, { "grad_norm": 0.2627333998680115, "learning_rate": 2.5665900112422535e-06, "loss": 0.0119, "step": 90270 }, { "grad_norm": 0.20175334811210632, "learning_rate": 2.561363125417876e-06, "loss": 0.0153, "step": 90280 }, { "grad_norm": 0.4267016053199768, "learning_rate": 2.556141427411496e-06, "loss": 0.0191, "step": 90290 }, { "grad_norm": 0.24843700230121613, "learning_rate": 2.550924917794156e-06, "loss": 0.0141, "step": 90300 }, { "grad_norm": 0.2799307107925415, "learning_rate": 2.545713597136329e-06, "loss": 0.0107, "step": 90310 }, { "grad_norm": 0.17313015460968018, "learning_rate": 2.5405074660079298e-06, "loss": 0.0106, "step": 90320 }, { "grad_norm": 0.23477692902088165, "learning_rate": 2.5353065249782647e-06, "loss": 0.0112, "step": 90330 }, { "grad_norm": 0.18925032019615173, "learning_rate": 2.530110774616129e-06, "loss": 0.0118, "step": 90340 }, { "grad_norm": 0.1503610908985138, "learning_rate": 2.524920215489701e-06, "loss": 0.0096, "step": 90350 }, { "grad_norm": 0.12440913915634155, "learning_rate": 2.519734848166627e-06, "loss": 0.0125, "step": 90360 }, { "grad_norm": 0.15257546305656433, "learning_rate": 2.514554673213948e-06, "loss": 0.01, "step": 90370 }, { "grad_norm": 0.25052663683891296, "learning_rate": 2.5093796911981936e-06, "loss": 0.0109, "step": 90380 }, { "grad_norm": 0.3506583869457245, "learning_rate": 2.5042099026852672e-06, "loss": 0.0167, "step": 90390 }, { "grad_norm": 0.1491374969482422, "learning_rate": 2.499045308240544e-06, "loss": 0.0112, "step": 90400 }, { "grad_norm": 0.1635715812444687, "learning_rate": 2.4938859084288003e-06, "loss": 0.0123, "step": 90410 }, { "grad_norm": 0.16203996539115906, "learning_rate": 2.4887317038142734e-06, "loss": 0.0134, "step": 90420 }, { "grad_norm": 0.20593395829200745, "learning_rate": 2.4835826949606124e-06, "loss": 0.0114, "step": 90430 }, { "grad_norm": 0.2342529594898224, "learning_rate": 2.4784388824309167e-06, "loss": 0.0185, "step": 90440 }, { "grad_norm": 0.25065651535987854, "learning_rate": 2.473300266787687e-06, "loss": 0.0134, "step": 90450 }, { "grad_norm": 0.2108641415834427, "learning_rate": 2.46816684859289e-06, "loss": 0.0136, "step": 90460 }, { "grad_norm": 0.20929625630378723, "learning_rate": 2.4630386284079043e-06, "loss": 0.0116, "step": 90470 }, { "grad_norm": 0.20662468671798706, "learning_rate": 2.4579156067935427e-06, "loss": 0.0138, "step": 90480 }, { "grad_norm": 0.19181816279888153, "learning_rate": 2.452797784310046e-06, "loss": 0.012, "step": 90490 }, { "grad_norm": 0.19296111166477203, "learning_rate": 2.447685161517105e-06, "loss": 0.0126, "step": 90500 }, { "grad_norm": 0.20223475992679596, "learning_rate": 2.4425777389738224e-06, "loss": 0.0106, "step": 90510 }, { "grad_norm": 0.2325645387172699, "learning_rate": 2.43747551723873e-06, "loss": 0.0125, "step": 90520 }, { "grad_norm": 0.13800761103630066, "learning_rate": 2.4323784968698136e-06, "loss": 0.0149, "step": 90530 }, { "grad_norm": 0.17287051677703857, "learning_rate": 2.4272866784244565e-06, "loss": 0.0154, "step": 90540 }, { "grad_norm": 0.15836720168590546, "learning_rate": 2.422200062459518e-06, "loss": 0.0138, "step": 90550 }, { "grad_norm": 0.16290904581546783, "learning_rate": 2.41711864953123e-06, "loss": 0.0106, "step": 90560 }, { "grad_norm": 0.28067973256111145, "learning_rate": 2.412042440195322e-06, "loss": 0.0095, "step": 90570 }, { "grad_norm": 0.28190818428993225, "learning_rate": 2.4069714350068993e-06, "loss": 0.013, "step": 90580 }, { "grad_norm": 0.137663796544075, "learning_rate": 2.4019056345205294e-06, "loss": 0.0126, "step": 90590 }, { "grad_norm": 0.26905253529548645, "learning_rate": 2.396845039290191e-06, "loss": 0.0146, "step": 90600 }, { "grad_norm": 0.2261231392621994, "learning_rate": 2.3917896498693194e-06, "loss": 0.0156, "step": 90610 }, { "grad_norm": 0.12540827691555023, "learning_rate": 2.3867394668107335e-06, "loss": 0.0098, "step": 90620 }, { "grad_norm": 0.12305574119091034, "learning_rate": 2.3816944906667527e-06, "loss": 0.015, "step": 90630 }, { "grad_norm": 0.3421143889427185, "learning_rate": 2.3766547219890634e-06, "loss": 0.0138, "step": 90640 }, { "grad_norm": 0.2000044286251068, "learning_rate": 2.371620161328819e-06, "loss": 0.0142, "step": 90650 }, { "grad_norm": 0.27578043937683105, "learning_rate": 2.366590809236574e-06, "loss": 0.0145, "step": 90660 }, { "grad_norm": 0.09891275316476822, "learning_rate": 2.3615666662623494e-06, "loss": 0.0134, "step": 90670 }, { "grad_norm": 0.3449992537498474, "learning_rate": 2.3565477329555664e-06, "loss": 0.0119, "step": 90680 }, { "grad_norm": 0.17593775689601898, "learning_rate": 2.3515340098650975e-06, "loss": 0.0106, "step": 90690 }, { "grad_norm": 0.39684122800827026, "learning_rate": 2.3465254975392258e-06, "loss": 0.0214, "step": 90700 }, { "grad_norm": 0.3069799840450287, "learning_rate": 2.3415221965256805e-06, "loss": 0.0217, "step": 90710 }, { "grad_norm": 0.20342211425304413, "learning_rate": 2.3365241073716237e-06, "loss": 0.0125, "step": 90720 }, { "grad_norm": 0.2143711894750595, "learning_rate": 2.3315312306236183e-06, "loss": 0.0116, "step": 90730 }, { "grad_norm": 0.1345893293619156, "learning_rate": 2.326543566827699e-06, "loss": 0.0115, "step": 90740 }, { "grad_norm": 0.1979658603668213, "learning_rate": 2.3215611165292862e-06, "loss": 0.0104, "step": 90750 }, { "grad_norm": 0.21593846380710602, "learning_rate": 2.3165838802732765e-06, "loss": 0.0108, "step": 90760 }, { "grad_norm": 0.34901130199432373, "learning_rate": 2.311611858603957e-06, "loss": 0.0139, "step": 90770 }, { "grad_norm": 0.31237366795539856, "learning_rate": 2.3066450520650763e-06, "loss": 0.0143, "step": 90780 }, { "grad_norm": 0.24123547971248627, "learning_rate": 2.3016834611997773e-06, "loss": 0.018, "step": 90790 }, { "grad_norm": 0.17446057498455048, "learning_rate": 2.296727086550665e-06, "loss": 0.013, "step": 90800 }, { "grad_norm": 0.7374404668807983, "learning_rate": 2.2917759286597507e-06, "loss": 0.0163, "step": 90810 }, { "grad_norm": 0.4314829111099243, "learning_rate": 2.2868299880685006e-06, "loss": 0.0171, "step": 90820 }, { "grad_norm": 0.16823740303516388, "learning_rate": 2.2818892653177827e-06, "loss": 0.0112, "step": 90830 }, { "grad_norm": 0.15578696131706238, "learning_rate": 2.2769537609479196e-06, "loss": 0.0116, "step": 90840 }, { "grad_norm": 0.1748734414577484, "learning_rate": 2.2720234754986357e-06, "loss": 0.0186, "step": 90850 }, { "grad_norm": 0.2256627380847931, "learning_rate": 2.2670984095091108e-06, "loss": 0.0169, "step": 90860 }, { "grad_norm": 0.16891032457351685, "learning_rate": 2.262178563517936e-06, "loss": 0.0179, "step": 90870 }, { "grad_norm": 0.19918927550315857, "learning_rate": 2.257263938063137e-06, "loss": 0.0104, "step": 90880 }, { "grad_norm": 0.21315035223960876, "learning_rate": 2.2523545336821715e-06, "loss": 0.0183, "step": 90890 }, { "grad_norm": 0.3098277151584625, "learning_rate": 2.247450350911939e-06, "loss": 0.0149, "step": 90900 }, { "grad_norm": 0.39276981353759766, "learning_rate": 2.2425513902887328e-06, "loss": 0.0126, "step": 90910 }, { "grad_norm": 0.23109132051467896, "learning_rate": 2.2376576523483015e-06, "loss": 0.0138, "step": 90920 }, { "grad_norm": 0.16050833463668823, "learning_rate": 2.232769137625829e-06, "loss": 0.0162, "step": 90930 }, { "grad_norm": 0.14886942505836487, "learning_rate": 2.227885846655903e-06, "loss": 0.0205, "step": 90940 }, { "grad_norm": 0.2200794368982315, "learning_rate": 2.2230077799725523e-06, "loss": 0.0127, "step": 90950 }, { "grad_norm": 0.11807847023010254, "learning_rate": 2.2181349381092396e-06, "loss": 0.0112, "step": 90960 }, { "grad_norm": 0.2521762549877167, "learning_rate": 2.2132673215988552e-06, "loss": 0.0122, "step": 90970 }, { "grad_norm": 0.26621392369270325, "learning_rate": 2.2084049309737065e-06, "loss": 0.0123, "step": 90980 }, { "grad_norm": 0.17928943037986755, "learning_rate": 2.2035477667655513e-06, "loss": 0.0167, "step": 90990 }, { "grad_norm": 0.5725240111351013, "learning_rate": 2.1986958295055315e-06, "loss": 0.0119, "step": 91000 }, { "grad_norm": 0.34291720390319824, "learning_rate": 2.193849119724284e-06, "loss": 0.0136, "step": 91010 }, { "grad_norm": 0.3123305141925812, "learning_rate": 2.1890076379518175e-06, "loss": 0.01, "step": 91020 }, { "grad_norm": 0.4238510727882385, "learning_rate": 2.1841713847175925e-06, "loss": 0.0148, "step": 91030 }, { "grad_norm": 0.24342310428619385, "learning_rate": 2.1793403605504968e-06, "loss": 0.0174, "step": 91040 }, { "grad_norm": 0.1637515127658844, "learning_rate": 2.1745145659788413e-06, "loss": 0.0131, "step": 91050 }, { "grad_norm": 0.2569599747657776, "learning_rate": 2.1696940015303645e-06, "loss": 0.0128, "step": 91060 }, { "grad_norm": 0.2461307793855667, "learning_rate": 2.1648786677322397e-06, "loss": 0.0164, "step": 91070 }, { "grad_norm": 0.31561604142189026, "learning_rate": 2.1600685651110673e-06, "loss": 0.0114, "step": 91080 }, { "grad_norm": 0.1277211755514145, "learning_rate": 2.1552636941928717e-06, "loss": 0.0106, "step": 91090 }, { "grad_norm": 0.21372169256210327, "learning_rate": 2.1504640555031043e-06, "loss": 0.0163, "step": 91100 }, { "grad_norm": 0.19805297255516052, "learning_rate": 2.1456696495666506e-06, "loss": 0.0128, "step": 91110 }, { "grad_norm": 0.08428476005792618, "learning_rate": 2.1408804769078084e-06, "loss": 0.0143, "step": 91120 }, { "grad_norm": 0.16909660398960114, "learning_rate": 2.136096538050325e-06, "loss": 0.0175, "step": 91130 }, { "grad_norm": 0.21991069614887238, "learning_rate": 2.13131783351736e-06, "loss": 0.0198, "step": 91140 }, { "grad_norm": 0.6512226462364197, "learning_rate": 2.126544363831512e-06, "loss": 0.0144, "step": 91150 }, { "grad_norm": 0.349933385848999, "learning_rate": 2.121776129514791e-06, "loss": 0.0118, "step": 91160 }, { "grad_norm": 0.2369774430990219, "learning_rate": 2.1170131310886523e-06, "loss": 0.0122, "step": 91170 }, { "grad_norm": 0.202315554022789, "learning_rate": 2.112255369073968e-06, "loss": 0.0136, "step": 91180 }, { "grad_norm": 0.12053020298480988, "learning_rate": 2.1075028439910328e-06, "loss": 0.0116, "step": 91190 }, { "grad_norm": 0.33267757296562195, "learning_rate": 2.1027555563595813e-06, "loss": 0.0149, "step": 91200 }, { "grad_norm": 0.35701385140419006, "learning_rate": 2.0980135066987706e-06, "loss": 0.0129, "step": 91210 }, { "grad_norm": 0.15380339324474335, "learning_rate": 2.093276695527191e-06, "loss": 0.0151, "step": 91220 }, { "grad_norm": 0.2102363407611847, "learning_rate": 2.0885451233628396e-06, "loss": 0.015, "step": 91230 }, { "grad_norm": 0.20277568697929382, "learning_rate": 2.0838187907231633e-06, "loss": 0.012, "step": 91240 }, { "grad_norm": 0.24870049953460693, "learning_rate": 2.0790976981250153e-06, "loss": 0.0123, "step": 91250 }, { "grad_norm": 0.2865908443927765, "learning_rate": 2.0743818460847e-06, "loss": 0.013, "step": 91260 }, { "grad_norm": 0.15145432949066162, "learning_rate": 2.069671235117926e-06, "loss": 0.0126, "step": 91270 }, { "grad_norm": 0.23181810975074768, "learning_rate": 2.0649658657398484e-06, "loss": 0.0105, "step": 91280 }, { "grad_norm": 0.10914339125156403, "learning_rate": 2.060265738465034e-06, "loss": 0.0141, "step": 91290 }, { "grad_norm": 0.2030908167362213, "learning_rate": 2.055570853807487e-06, "loss": 0.0116, "step": 91300 }, { "grad_norm": 0.26541706919670105, "learning_rate": 2.0508812122806196e-06, "loss": 0.0132, "step": 91310 }, { "grad_norm": 0.2384525090456009, "learning_rate": 2.046196814397294e-06, "loss": 0.0136, "step": 91320 }, { "grad_norm": 0.33367812633514404, "learning_rate": 2.0415176606697896e-06, "loss": 0.0155, "step": 91330 }, { "grad_norm": 0.17291170358657837, "learning_rate": 2.0368437516098126e-06, "loss": 0.0139, "step": 91340 }, { "grad_norm": 0.16865161061286926, "learning_rate": 2.032175087728483e-06, "loss": 0.0137, "step": 91350 }, { "grad_norm": 0.13047447800636292, "learning_rate": 2.027511669536375e-06, "loss": 0.0125, "step": 91360 }, { "grad_norm": 0.16104789078235626, "learning_rate": 2.0228534975434643e-06, "loss": 0.009, "step": 91370 }, { "grad_norm": 0.4655359983444214, "learning_rate": 2.0182005722591548e-06, "loss": 0.0135, "step": 91380 }, { "grad_norm": 0.22256717085838318, "learning_rate": 2.0135528941922944e-06, "loss": 0.0137, "step": 91390 }, { "grad_norm": 0.4048621654510498, "learning_rate": 2.008910463851149e-06, "loss": 0.0114, "step": 91400 }, { "grad_norm": 0.17009171843528748, "learning_rate": 2.0042732817433963e-06, "loss": 0.0141, "step": 91410 }, { "grad_norm": 0.27117782831192017, "learning_rate": 1.999641348376158e-06, "loss": 0.0103, "step": 91420 }, { "grad_norm": 0.14960598945617676, "learning_rate": 1.9950146642559787e-06, "loss": 0.0141, "step": 91430 }, { "grad_norm": 0.28145307302474976, "learning_rate": 1.990393229888815e-06, "loss": 0.0166, "step": 91440 }, { "grad_norm": 0.4719223380088806, "learning_rate": 1.9857770457800684e-06, "loss": 0.0173, "step": 91450 }, { "grad_norm": 0.26997750997543335, "learning_rate": 1.9811661124345517e-06, "loss": 0.0149, "step": 91460 }, { "grad_norm": 0.19840992987155914, "learning_rate": 1.976560430356522e-06, "loss": 0.0143, "step": 91470 }, { "grad_norm": 0.15667878091335297, "learning_rate": 1.9719600000496383e-06, "loss": 0.0123, "step": 91480 }, { "grad_norm": 0.21431021392345428, "learning_rate": 1.967364822017004e-06, "loss": 0.0155, "step": 91490 }, { "grad_norm": 0.1637612134218216, "learning_rate": 1.962774896761127e-06, "loss": 0.0145, "step": 91500 }, { "grad_norm": 0.12171169370412827, "learning_rate": 1.958190224783968e-06, "loss": 0.0135, "step": 91510 }, { "grad_norm": 0.2384658008813858, "learning_rate": 1.9536108065868975e-06, "loss": 0.0136, "step": 91520 }, { "grad_norm": 0.17543470859527588, "learning_rate": 1.949036642670715e-06, "loss": 0.0112, "step": 91530 }, { "grad_norm": 0.2838841378688812, "learning_rate": 1.9444677335356377e-06, "loss": 0.0143, "step": 91540 }, { "grad_norm": 0.08429121971130371, "learning_rate": 1.939904079681326e-06, "loss": 0.0138, "step": 91550 }, { "grad_norm": 0.24954786896705627, "learning_rate": 1.935345681606837e-06, "loss": 0.0123, "step": 91560 }, { "grad_norm": 0.3135845363140106, "learning_rate": 1.930792539810694e-06, "loss": 0.0106, "step": 91570 }, { "grad_norm": 0.19384193420410156, "learning_rate": 1.9262446547907877e-06, "loss": 0.0113, "step": 91580 }, { "grad_norm": 0.1556629091501236, "learning_rate": 1.921702027044503e-06, "loss": 0.012, "step": 91590 }, { "grad_norm": 0.24953092634677887, "learning_rate": 1.917164657068593e-06, "loss": 0.0128, "step": 91600 }, { "grad_norm": 0.19890880584716797, "learning_rate": 1.912632545359272e-06, "loss": 0.0128, "step": 91610 }, { "grad_norm": 0.20975467562675476, "learning_rate": 1.9081056924121488e-06, "loss": 0.0173, "step": 91620 }, { "grad_norm": 0.23444533348083496, "learning_rate": 1.9035840987222887e-06, "loss": 0.0119, "step": 91630 }, { "grad_norm": 0.09956136345863342, "learning_rate": 1.8990677647841515e-06, "loss": 0.0131, "step": 91640 }, { "grad_norm": 0.24982252717018127, "learning_rate": 1.8945566910916534e-06, "loss": 0.0145, "step": 91650 }, { "grad_norm": 0.13969308137893677, "learning_rate": 1.8900508781381054e-06, "loss": 0.0129, "step": 91660 }, { "grad_norm": 0.1384059488773346, "learning_rate": 1.8855503264162633e-06, "loss": 0.0144, "step": 91670 }, { "grad_norm": 0.22225657105445862, "learning_rate": 1.8810550364182998e-06, "loss": 0.0185, "step": 91680 }, { "grad_norm": 0.20933201909065247, "learning_rate": 1.8765650086358112e-06, "loss": 0.0127, "step": 91690 }, { "grad_norm": 0.24482771754264832, "learning_rate": 1.8720802435598151e-06, "loss": 0.0132, "step": 91700 }, { "grad_norm": 0.19625505805015564, "learning_rate": 1.8676007416807694e-06, "loss": 0.0151, "step": 91710 }, { "grad_norm": 0.28017592430114746, "learning_rate": 1.8631265034885437e-06, "loss": 0.0167, "step": 91720 }, { "grad_norm": 0.3056952953338623, "learning_rate": 1.858657529472424e-06, "loss": 0.0134, "step": 91730 }, { "grad_norm": 0.14596645534038544, "learning_rate": 1.8541938201211473e-06, "loss": 0.0097, "step": 91740 }, { "grad_norm": 0.3130747675895691, "learning_rate": 1.8497353759228398e-06, "loss": 0.0158, "step": 91750 }, { "grad_norm": 0.22690710425376892, "learning_rate": 1.8452821973650835e-06, "loss": 0.0102, "step": 91760 }, { "grad_norm": 0.4551289975643158, "learning_rate": 1.840834284934856e-06, "loss": 0.0177, "step": 91770 }, { "grad_norm": 0.17798586189746857, "learning_rate": 1.8363916391185954e-06, "loss": 0.012, "step": 91780 }, { "grad_norm": 0.18280132114887238, "learning_rate": 1.8319542604021245e-06, "loss": 0.0131, "step": 91790 }, { "grad_norm": 0.4631052613258362, "learning_rate": 1.8275221492707273e-06, "loss": 0.0133, "step": 91800 }, { "grad_norm": 0.19276344776153564, "learning_rate": 1.8230953062090717e-06, "loss": 0.0108, "step": 91810 }, { "grad_norm": 0.18040558695793152, "learning_rate": 1.8186737317012926e-06, "loss": 0.0132, "step": 91820 }, { "grad_norm": 0.1576230227947235, "learning_rate": 1.8142574262308977e-06, "loss": 0.0135, "step": 91830 }, { "grad_norm": 0.18402931094169617, "learning_rate": 1.8098463902808727e-06, "loss": 0.0156, "step": 91840 }, { "grad_norm": 0.14631816744804382, "learning_rate": 1.805440624333593e-06, "loss": 0.0099, "step": 91850 }, { "grad_norm": 0.2331552356481552, "learning_rate": 1.8010401288708734e-06, "loss": 0.0144, "step": 91860 }, { "grad_norm": 0.16571813821792603, "learning_rate": 1.7966449043739343e-06, "loss": 0.0101, "step": 91870 }, { "grad_norm": 0.32801929116249084, "learning_rate": 1.79225495132343e-06, "loss": 0.0134, "step": 91880 }, { "grad_norm": 0.27032896876335144, "learning_rate": 1.7878702701994487e-06, "loss": 0.0163, "step": 91890 }, { "grad_norm": 0.23537132143974304, "learning_rate": 1.783490861481496e-06, "loss": 0.0128, "step": 91900 }, { "grad_norm": 0.29420313239097595, "learning_rate": 1.7791167256484832e-06, "loss": 0.0119, "step": 91910 }, { "grad_norm": 0.3232286274433136, "learning_rate": 1.774747863178766e-06, "loss": 0.0194, "step": 91920 }, { "grad_norm": 0.18270736932754517, "learning_rate": 1.7703842745501186e-06, "loss": 0.0122, "step": 91930 }, { "grad_norm": 0.11356700211763382, "learning_rate": 1.7660259602397366e-06, "loss": 0.0119, "step": 91940 }, { "grad_norm": 0.17652490735054016, "learning_rate": 1.7616729207242388e-06, "loss": 0.014, "step": 91950 }, { "grad_norm": 0.1406644731760025, "learning_rate": 1.7573251564796556e-06, "loss": 0.0107, "step": 91960 }, { "grad_norm": 0.40629473328590393, "learning_rate": 1.7529826679814676e-06, "loss": 0.014, "step": 91970 }, { "grad_norm": 0.31221845746040344, "learning_rate": 1.7486454557045562e-06, "loss": 0.0127, "step": 91980 }, { "grad_norm": 0.17924070358276367, "learning_rate": 1.7443135201232364e-06, "loss": 0.0115, "step": 91990 }, { "grad_norm": 0.23212884366512299, "learning_rate": 1.7399868617112347e-06, "loss": 0.0151, "step": 92000 }, { "grad_norm": 0.2758144736289978, "learning_rate": 1.7356654809417171e-06, "loss": 0.0122, "step": 92010 }, { "grad_norm": 0.22542421519756317, "learning_rate": 1.7313493782872448e-06, "loss": 0.0135, "step": 92020 }, { "grad_norm": 0.3625730872154236, "learning_rate": 1.7270385542198453e-06, "loss": 0.013, "step": 92030 }, { "grad_norm": 0.13809163868427277, "learning_rate": 1.7227330092109307e-06, "loss": 0.0115, "step": 92040 }, { "grad_norm": 0.19614847004413605, "learning_rate": 1.718432743731352e-06, "loss": 0.0126, "step": 92050 }, { "grad_norm": 0.1729516088962555, "learning_rate": 1.7141377582513718e-06, "loss": 0.0108, "step": 92060 }, { "grad_norm": 0.19904941320419312, "learning_rate": 1.709848053240698e-06, "loss": 0.0137, "step": 92070 }, { "grad_norm": 0.20637337863445282, "learning_rate": 1.7055636291684328e-06, "loss": 0.0137, "step": 92080 }, { "grad_norm": 0.30680420994758606, "learning_rate": 1.7012844865031185e-06, "loss": 0.0134, "step": 92090 }, { "grad_norm": 0.1044277474284172, "learning_rate": 1.6970106257127138e-06, "loss": 0.0124, "step": 92100 }, { "grad_norm": 0.18344910442829132, "learning_rate": 1.6927420472646117e-06, "loss": 0.0116, "step": 92110 }, { "grad_norm": 0.11882515996694565, "learning_rate": 1.6884787516256106e-06, "loss": 0.016, "step": 92120 }, { "grad_norm": 0.1967632919549942, "learning_rate": 1.6842207392619325e-06, "loss": 0.011, "step": 92130 }, { "grad_norm": 0.23830784857273102, "learning_rate": 1.679968010639238e-06, "loss": 0.0142, "step": 92140 }, { "grad_norm": 0.1346563994884491, "learning_rate": 1.6757205662225829e-06, "loss": 0.0147, "step": 92150 }, { "grad_norm": 0.3821476697921753, "learning_rate": 1.6714784064764844e-06, "loss": 0.0139, "step": 92160 }, { "grad_norm": 0.19949355721473694, "learning_rate": 1.6672415318648383e-06, "loss": 0.0116, "step": 92170 }, { "grad_norm": 0.37975433468818665, "learning_rate": 1.6630099428510016e-06, "loss": 0.0102, "step": 92180 }, { "grad_norm": 0.1877620965242386, "learning_rate": 1.6587836398977153e-06, "loss": 0.0201, "step": 92190 }, { "grad_norm": 0.11713218688964844, "learning_rate": 1.6545626234671762e-06, "loss": 0.0156, "step": 92200 }, { "grad_norm": 0.1258004903793335, "learning_rate": 1.650346894020971e-06, "loss": 0.0127, "step": 92210 }, { "grad_norm": 0.2681800425052643, "learning_rate": 1.6461364520201472e-06, "loss": 0.0119, "step": 92220 }, { "grad_norm": 0.2982521653175354, "learning_rate": 1.6419312979251366e-06, "loss": 0.0184, "step": 92230 }, { "grad_norm": 0.18607528507709503, "learning_rate": 1.6377314321958214e-06, "loss": 0.0137, "step": 92240 }, { "grad_norm": 0.10452558845281601, "learning_rate": 1.6335368552914787e-06, "loss": 0.0132, "step": 92250 }, { "grad_norm": 0.2095278650522232, "learning_rate": 1.6293475676708358e-06, "loss": 0.0134, "step": 92260 }, { "grad_norm": 0.1743469387292862, "learning_rate": 1.6251635697920097e-06, "loss": 0.0174, "step": 92270 }, { "grad_norm": 0.38218843936920166, "learning_rate": 1.6209848621125624e-06, "loss": 0.0149, "step": 92280 }, { "grad_norm": 0.1800544410943985, "learning_rate": 1.6168114450894778e-06, "loss": 0.0145, "step": 92290 }, { "grad_norm": 0.12693443894386292, "learning_rate": 1.612643319179158e-06, "loss": 0.0138, "step": 92300 }, { "grad_norm": 0.2500194311141968, "learning_rate": 1.6084804848374046e-06, "loss": 0.0209, "step": 92310 }, { "grad_norm": 0.1539222002029419, "learning_rate": 1.6043229425194817e-06, "loss": 0.0126, "step": 92320 }, { "grad_norm": 0.20712341368198395, "learning_rate": 1.6001706926800308e-06, "loss": 0.0143, "step": 92330 }, { "grad_norm": 0.26888930797576904, "learning_rate": 1.5960237357731444e-06, "loss": 0.0124, "step": 92340 }, { "grad_norm": 0.17208240926265717, "learning_rate": 1.5918820722523264e-06, "loss": 0.0157, "step": 92350 }, { "grad_norm": 0.19225317239761353, "learning_rate": 1.587745702570509e-06, "loss": 0.0123, "step": 92360 }, { "grad_norm": 0.17839884757995605, "learning_rate": 1.5836146271800356e-06, "loss": 0.0159, "step": 92370 }, { "grad_norm": 0.26003214716911316, "learning_rate": 1.5794888465326728e-06, "loss": 0.0125, "step": 92380 }, { "grad_norm": 0.32710424065589905, "learning_rate": 1.5753683610796155e-06, "loss": 0.0172, "step": 92390 }, { "grad_norm": 0.13071231544017792, "learning_rate": 1.5712531712714528e-06, "loss": 0.0154, "step": 92400 }, { "grad_norm": 0.19840046763420105, "learning_rate": 1.5671432775582474e-06, "loss": 0.0115, "step": 92410 }, { "grad_norm": 0.14343778789043427, "learning_rate": 1.563038680389428e-06, "loss": 0.0131, "step": 92420 }, { "grad_norm": 0.16427409648895264, "learning_rate": 1.5589393802138808e-06, "loss": 0.0128, "step": 92430 }, { "grad_norm": 0.20261448621749878, "learning_rate": 1.554845377479891e-06, "loss": 0.0132, "step": 92440 }, { "grad_norm": 0.1659628003835678, "learning_rate": 1.550756672635173e-06, "loss": 0.0139, "step": 92450 }, { "grad_norm": 0.16777020692825317, "learning_rate": 1.5466732661268635e-06, "loss": 0.0124, "step": 92460 }, { "grad_norm": 0.18010838329792023, "learning_rate": 1.5425951584015163e-06, "loss": 0.011, "step": 92470 }, { "grad_norm": 0.16145454347133636, "learning_rate": 1.5385223499051026e-06, "loss": 0.0191, "step": 92480 }, { "grad_norm": 0.1600257009267807, "learning_rate": 1.5344548410830328e-06, "loss": 0.0188, "step": 92490 }, { "grad_norm": 0.16040760278701782, "learning_rate": 1.5303926323801121e-06, "loss": 0.0113, "step": 92500 }, { "grad_norm": 0.2955939471721649, "learning_rate": 1.5263357242405795e-06, "loss": 0.0114, "step": 92510 }, { "grad_norm": 0.12152265757322311, "learning_rate": 1.5222841171080915e-06, "loss": 0.0145, "step": 92520 }, { "grad_norm": 0.3168542981147766, "learning_rate": 1.5182378114257268e-06, "loss": 0.0118, "step": 92530 }, { "grad_norm": 0.266353964805603, "learning_rate": 1.5141968076359869e-06, "loss": 0.0153, "step": 92540 }, { "grad_norm": 0.2338286191225052, "learning_rate": 1.5101611061807908e-06, "loss": 0.0212, "step": 92550 }, { "grad_norm": 0.40504685044288635, "learning_rate": 1.506130707501463e-06, "loss": 0.0117, "step": 92560 }, { "grad_norm": 0.20927341282367706, "learning_rate": 1.5021056120387734e-06, "loss": 0.0139, "step": 92570 }, { "grad_norm": 0.1259615272283554, "learning_rate": 1.4980858202329084e-06, "loss": 0.0139, "step": 92580 }, { "grad_norm": 0.48185646533966064, "learning_rate": 1.4940713325234445e-06, "loss": 0.0162, "step": 92590 }, { "grad_norm": 0.13719750940799713, "learning_rate": 1.4900621493494137e-06, "loss": 0.0108, "step": 92600 }, { "grad_norm": 0.19049683213233948, "learning_rate": 1.4860582711492544e-06, "loss": 0.02, "step": 92610 }, { "grad_norm": 0.3446223735809326, "learning_rate": 1.4820596983608325e-06, "loss": 0.0142, "step": 92620 }, { "grad_norm": 0.18309064209461212, "learning_rate": 1.4780664314214044e-06, "loss": 0.0134, "step": 92630 }, { "grad_norm": 0.26759207248687744, "learning_rate": 1.474078470767687e-06, "loss": 0.0168, "step": 92640 }, { "grad_norm": 0.2721782326698303, "learning_rate": 1.470095816835787e-06, "loss": 0.0155, "step": 92650 }, { "grad_norm": 0.4361276924610138, "learning_rate": 1.4661184700612395e-06, "loss": 0.0189, "step": 92660 }, { "grad_norm": 0.3399799168109894, "learning_rate": 1.4621464308790133e-06, "loss": 0.0148, "step": 92670 }, { "grad_norm": 0.26166000962257385, "learning_rate": 1.4581796997234775e-06, "loss": 0.0179, "step": 92680 }, { "grad_norm": 0.15680399537086487, "learning_rate": 1.4542182770284297e-06, "loss": 0.0128, "step": 92690 }, { "grad_norm": 0.1283225268125534, "learning_rate": 1.450262163227084e-06, "loss": 0.0137, "step": 92700 }, { "grad_norm": 0.17653505504131317, "learning_rate": 1.4463113587520671e-06, "loss": 0.0212, "step": 92710 }, { "grad_norm": 0.17770788073539734, "learning_rate": 1.4423658640354443e-06, "loss": 0.015, "step": 92720 }, { "grad_norm": 0.3198164999485016, "learning_rate": 1.438425679508687e-06, "loss": 0.0207, "step": 92730 }, { "grad_norm": 0.34788838028907776, "learning_rate": 1.4344908056026895e-06, "loss": 0.0114, "step": 92740 }, { "grad_norm": 0.5341761708259583, "learning_rate": 1.4305612427477521e-06, "loss": 0.0127, "step": 92750 }, { "grad_norm": 0.2444353848695755, "learning_rate": 1.42663699137362e-06, "loss": 0.0134, "step": 92760 }, { "grad_norm": 0.14104589819908142, "learning_rate": 1.4227180519094386e-06, "loss": 0.0149, "step": 92770 }, { "grad_norm": 0.24457260966300964, "learning_rate": 1.4188044247837706e-06, "loss": 0.0128, "step": 92780 }, { "grad_norm": 0.18456481397151947, "learning_rate": 1.4148961104246128e-06, "loss": 0.0085, "step": 92790 }, { "grad_norm": 0.3553259074687958, "learning_rate": 1.4109931092593731e-06, "loss": 0.0151, "step": 92800 }, { "grad_norm": 0.2618369460105896, "learning_rate": 1.4070954217148658e-06, "loss": 0.0141, "step": 92810 }, { "grad_norm": 0.20400232076644897, "learning_rate": 1.4032030482173497e-06, "loss": 0.0151, "step": 92820 }, { "grad_norm": 0.17932221293449402, "learning_rate": 1.3993159891924901e-06, "loss": 0.0131, "step": 92830 }, { "grad_norm": 0.43633776903152466, "learning_rate": 1.3954342450653524e-06, "loss": 0.0197, "step": 92840 }, { "grad_norm": 0.4593716561794281, "learning_rate": 1.3915578162604582e-06, "loss": 0.0118, "step": 92850 }, { "grad_norm": 0.19592022895812988, "learning_rate": 1.3876867032017126e-06, "loss": 0.0132, "step": 92860 }, { "grad_norm": 0.34513378143310547, "learning_rate": 1.383820906312472e-06, "loss": 0.0144, "step": 92870 }, { "grad_norm": 0.27130088210105896, "learning_rate": 1.3799604260154752e-06, "loss": 0.0168, "step": 92880 }, { "grad_norm": 0.2694464325904846, "learning_rate": 1.376105262732913e-06, "loss": 0.0231, "step": 92890 }, { "grad_norm": 0.2890242338180542, "learning_rate": 1.37225541688637e-06, "loss": 0.0176, "step": 92900 }, { "grad_norm": 0.13614438474178314, "learning_rate": 1.3684108888968594e-06, "loss": 0.0173, "step": 92910 }, { "grad_norm": 0.4841700494289398, "learning_rate": 1.364571679184823e-06, "loss": 0.0161, "step": 92920 }, { "grad_norm": 0.26624345779418945, "learning_rate": 1.3607377881701077e-06, "loss": 0.0211, "step": 92930 }, { "grad_norm": 0.16938331723213196, "learning_rate": 1.3569092162719787e-06, "loss": 0.0099, "step": 92940 }, { "grad_norm": 0.1678124964237213, "learning_rate": 1.353085963909123e-06, "loss": 0.0138, "step": 92950 }, { "grad_norm": 0.2173452377319336, "learning_rate": 1.3492680314996454e-06, "loss": 0.0157, "step": 92960 }, { "grad_norm": 0.18250715732574463, "learning_rate": 1.345455419461078e-06, "loss": 0.0149, "step": 92970 }, { "grad_norm": 0.40999895334243774, "learning_rate": 1.3416481282103431e-06, "loss": 0.0127, "step": 92980 }, { "grad_norm": 0.20050348341464996, "learning_rate": 1.337846158163819e-06, "loss": 0.013, "step": 92990 }, { "grad_norm": 0.14613361656665802, "learning_rate": 1.334049509737273e-06, "loss": 0.0166, "step": 93000 }, { "grad_norm": 0.4826492965221405, "learning_rate": 1.3302581833459116e-06, "loss": 0.0111, "step": 93010 }, { "grad_norm": 0.1775023490190506, "learning_rate": 1.3264721794043366e-06, "loss": 0.0128, "step": 93020 }, { "grad_norm": 0.18736104667186737, "learning_rate": 1.3226914983265836e-06, "loss": 0.0208, "step": 93030 }, { "grad_norm": 0.3460903763771057, "learning_rate": 1.3189161405260997e-06, "loss": 0.0243, "step": 93040 }, { "grad_norm": 0.1489582657814026, "learning_rate": 1.3151461064157656e-06, "loss": 0.0133, "step": 93050 }, { "grad_norm": 0.11921030282974243, "learning_rate": 1.3113813964078459e-06, "loss": 0.0142, "step": 93060 }, { "grad_norm": 0.2072642594575882, "learning_rate": 1.3076220109140614e-06, "loss": 0.0121, "step": 93070 }, { "grad_norm": 0.1356992870569229, "learning_rate": 1.3038679503455275e-06, "loss": 0.0115, "step": 93080 }, { "grad_norm": 2.320141553878784, "learning_rate": 1.3001192151127773e-06, "loss": 0.017, "step": 93090 }, { "grad_norm": 0.5253239870071411, "learning_rate": 1.2963758056257657e-06, "loss": 0.0174, "step": 93100 }, { "grad_norm": 0.19645348191261292, "learning_rate": 1.2926377222938768e-06, "loss": 0.011, "step": 93110 }, { "grad_norm": 0.2583538591861725, "learning_rate": 1.2889049655259e-06, "loss": 0.0177, "step": 93120 }, { "grad_norm": 0.1341986507177353, "learning_rate": 1.2851775357300312e-06, "loss": 0.0167, "step": 93130 }, { "grad_norm": 0.22604891657829285, "learning_rate": 1.281455433313916e-06, "loss": 0.0144, "step": 93140 }, { "grad_norm": 0.14598815143108368, "learning_rate": 1.2777386586845796e-06, "loss": 0.0118, "step": 93150 }, { "grad_norm": 0.1418221890926361, "learning_rate": 1.2740272122484964e-06, "loss": 0.0155, "step": 93160 }, { "grad_norm": 0.24343271553516388, "learning_rate": 1.2703210944115308e-06, "loss": 0.0115, "step": 93170 }, { "grad_norm": 0.2088214010000229, "learning_rate": 1.2666203055789916e-06, "loss": 0.0135, "step": 93180 }, { "grad_norm": 0.1651209592819214, "learning_rate": 1.2629248461555832e-06, "loss": 0.0126, "step": 93190 }, { "grad_norm": 0.16181021928787231, "learning_rate": 1.2592347165454431e-06, "loss": 0.0104, "step": 93200 }, { "grad_norm": 0.5600239634513855, "learning_rate": 1.2555499171521102e-06, "loss": 0.0151, "step": 93210 }, { "grad_norm": 0.1729440540075302, "learning_rate": 1.251870448378556e-06, "loss": 0.0128, "step": 93220 }, { "grad_norm": 0.1159467101097107, "learning_rate": 1.2481963106271478e-06, "loss": 0.0143, "step": 93230 }, { "grad_norm": 0.13189387321472168, "learning_rate": 1.2445275042997084e-06, "loss": 0.0168, "step": 93240 }, { "grad_norm": 0.1217760220170021, "learning_rate": 1.2408640297974283e-06, "loss": 0.0123, "step": 93250 }, { "grad_norm": 0.2135845124721527, "learning_rate": 1.2372058875209535e-06, "loss": 0.0116, "step": 93260 }, { "grad_norm": 0.32990384101867676, "learning_rate": 1.2335530778703252e-06, "loss": 0.012, "step": 93270 }, { "grad_norm": 0.12627308070659637, "learning_rate": 1.2299056012450182e-06, "loss": 0.013, "step": 93280 }, { "grad_norm": 0.351121187210083, "learning_rate": 1.2262634580439026e-06, "loss": 0.0142, "step": 93290 }, { "grad_norm": 0.19148467481136322, "learning_rate": 1.2226266486652926e-06, "loss": 0.0129, "step": 93300 }, { "grad_norm": 0.20685075223445892, "learning_rate": 1.2189951735068927e-06, "loss": 0.0117, "step": 93310 }, { "grad_norm": 0.4264853298664093, "learning_rate": 1.2153690329658407e-06, "loss": 0.014, "step": 93320 }, { "grad_norm": 0.2707774043083191, "learning_rate": 1.2117482274386859e-06, "loss": 0.0115, "step": 93330 }, { "grad_norm": 0.2061261385679245, "learning_rate": 1.2081327573213953e-06, "loss": 0.0156, "step": 93340 }, { "grad_norm": 0.23556199669837952, "learning_rate": 1.2045226230093465e-06, "loss": 0.015, "step": 93350 }, { "grad_norm": 0.2495098114013672, "learning_rate": 1.2009178248973296e-06, "loss": 0.0123, "step": 93360 }, { "grad_norm": 0.20753103494644165, "learning_rate": 1.1973183633795849e-06, "loss": 0.0092, "step": 93370 }, { "grad_norm": 0.17323553562164307, "learning_rate": 1.193724238849725e-06, "loss": 0.0105, "step": 93380 }, { "grad_norm": 0.22010394930839539, "learning_rate": 1.190135451700808e-06, "loss": 0.012, "step": 93390 }, { "grad_norm": 0.3365742564201355, "learning_rate": 1.186552002325292e-06, "loss": 0.0148, "step": 93400 }, { "grad_norm": 0.2013663947582245, "learning_rate": 1.1829738911150578e-06, "loss": 0.0139, "step": 93410 }, { "grad_norm": 0.1497442126274109, "learning_rate": 1.1794011184613985e-06, "loss": 0.01, "step": 93420 }, { "grad_norm": 0.17914621531963348, "learning_rate": 1.17583368475504e-06, "loss": 0.0094, "step": 93430 }, { "grad_norm": 0.22891587018966675, "learning_rate": 1.1722715903860982e-06, "loss": 0.017, "step": 93440 }, { "grad_norm": 0.2652893662452698, "learning_rate": 1.168714835744128e-06, "loss": 0.0087, "step": 93450 }, { "grad_norm": 0.2839089334011078, "learning_rate": 1.165163421218085e-06, "loss": 0.0162, "step": 93460 }, { "grad_norm": 0.1572142392396927, "learning_rate": 1.1616173471963477e-06, "loss": 0.0128, "step": 93470 }, { "grad_norm": 0.2696397602558136, "learning_rate": 1.1580766140667109e-06, "loss": 0.0128, "step": 93480 }, { "grad_norm": 0.4689636826515198, "learning_rate": 1.1545412222163764e-06, "loss": 0.0111, "step": 93490 }, { "grad_norm": 0.1982717663049698, "learning_rate": 1.1510111720319795e-06, "loss": 0.0145, "step": 93500 }, { "grad_norm": 0.16452696919441223, "learning_rate": 1.1474864638995663e-06, "loss": 0.0106, "step": 93510 }, { "grad_norm": 0.1527385413646698, "learning_rate": 1.1439670982045735e-06, "loss": 0.012, "step": 93520 }, { "grad_norm": 0.18112874031066895, "learning_rate": 1.1404530753318877e-06, "loss": 0.0146, "step": 93530 }, { "grad_norm": 0.7228615283966064, "learning_rate": 1.136944395665801e-06, "loss": 0.0121, "step": 93540 }, { "grad_norm": 0.24506829679012299, "learning_rate": 1.1334410595900014e-06, "loss": 0.0148, "step": 93550 }, { "grad_norm": 0.1531211882829666, "learning_rate": 1.129943067487621e-06, "loss": 0.0153, "step": 93560 }, { "grad_norm": 0.29630744457244873, "learning_rate": 1.126450419741193e-06, "loss": 0.02, "step": 93570 }, { "grad_norm": 0.1436561942100525, "learning_rate": 1.1229631167326671e-06, "loss": 0.0105, "step": 93580 }, { "grad_norm": 0.20689523220062256, "learning_rate": 1.1194811588434052e-06, "loss": 0.0118, "step": 93590 }, { "grad_norm": 0.9109403491020203, "learning_rate": 1.1160045464542023e-06, "loss": 0.0118, "step": 93600 }, { "grad_norm": 0.24590498208999634, "learning_rate": 1.1125332799452325e-06, "loss": 0.0184, "step": 93610 }, { "grad_norm": 0.16958506405353546, "learning_rate": 1.109067359696131e-06, "loss": 0.0166, "step": 93620 }, { "grad_norm": 0.4595428705215454, "learning_rate": 1.1056067860859166e-06, "loss": 0.0127, "step": 93630 }, { "grad_norm": 0.0615183562040329, "learning_rate": 1.102151559493031e-06, "loss": 0.0121, "step": 93640 }, { "grad_norm": 0.19609741866588593, "learning_rate": 1.0987016802953332e-06, "loss": 0.0121, "step": 93650 }, { "grad_norm": 0.16580182313919067, "learning_rate": 1.0952571488701048e-06, "loss": 0.0138, "step": 93660 }, { "grad_norm": 0.08004845678806305, "learning_rate": 1.0918179655940163e-06, "loss": 0.0127, "step": 93670 }, { "grad_norm": 0.20343148708343506, "learning_rate": 1.088384130843184e-06, "loss": 0.0103, "step": 93680 }, { "grad_norm": 0.18405388295650482, "learning_rate": 1.0849556449931296e-06, "loss": 0.015, "step": 93690 }, { "grad_norm": 0.16456186771392822, "learning_rate": 1.0815325084187866e-06, "loss": 0.0134, "step": 93700 }, { "grad_norm": 0.09839529544115067, "learning_rate": 1.0781147214944943e-06, "loss": 0.0106, "step": 93710 }, { "grad_norm": 0.24216486513614655, "learning_rate": 1.0747022845940313e-06, "loss": 0.0115, "step": 93720 }, { "grad_norm": 0.20117872953414917, "learning_rate": 1.0712951980905605e-06, "loss": 0.0099, "step": 93730 }, { "grad_norm": 0.35223153233528137, "learning_rate": 1.0678934623566839e-06, "loss": 0.0114, "step": 93740 }, { "grad_norm": 0.1428382396697998, "learning_rate": 1.0644970777644091e-06, "loss": 0.0105, "step": 93750 }, { "grad_norm": 0.20398633182048798, "learning_rate": 1.0611060446851673e-06, "loss": 0.0118, "step": 93760 }, { "grad_norm": 0.1383151412010193, "learning_rate": 1.0577203634897836e-06, "loss": 0.0167, "step": 93770 }, { "grad_norm": 0.13275815546512604, "learning_rate": 1.0543400345485178e-06, "loss": 0.0134, "step": 93780 }, { "grad_norm": 0.3643468916416168, "learning_rate": 1.0509650582310404e-06, "loss": 0.0136, "step": 93790 }, { "grad_norm": 0.3388165831565857, "learning_rate": 1.0475954349064287e-06, "loss": 0.0153, "step": 93800 }, { "grad_norm": 0.23822703957557678, "learning_rate": 1.0442311649431825e-06, "loss": 0.0129, "step": 93810 }, { "grad_norm": 0.1597556620836258, "learning_rate": 1.0408722487092071e-06, "loss": 0.0154, "step": 93820 }, { "grad_norm": 0.15308168530464172, "learning_rate": 1.0375186865718424e-06, "loss": 0.0101, "step": 93830 }, { "grad_norm": 0.11664465069770813, "learning_rate": 1.0341704788978113e-06, "loss": 0.0128, "step": 93840 }, { "grad_norm": 0.29455092549324036, "learning_rate": 1.0308276260532879e-06, "loss": 0.014, "step": 93850 }, { "grad_norm": 0.1731710582971573, "learning_rate": 1.0274901284038241e-06, "loss": 0.0097, "step": 93860 }, { "grad_norm": 0.19344960153102875, "learning_rate": 1.0241579863144114e-06, "loss": 0.0162, "step": 93870 }, { "grad_norm": 0.13281606137752533, "learning_rate": 1.0208312001494413e-06, "loss": 0.0156, "step": 93880 }, { "grad_norm": 0.14758606255054474, "learning_rate": 1.0175097702727453e-06, "loss": 0.0108, "step": 93890 }, { "grad_norm": 0.2877001464366913, "learning_rate": 1.0141936970475275e-06, "loss": 0.012, "step": 93900 }, { "grad_norm": 0.2385270744562149, "learning_rate": 1.0108829808364417e-06, "loss": 0.0153, "step": 93910 }, { "grad_norm": 0.2231370210647583, "learning_rate": 1.0075776220015376e-06, "loss": 0.0107, "step": 93920 }, { "grad_norm": 0.1845402717590332, "learning_rate": 1.0042776209042816e-06, "loss": 0.0121, "step": 93930 }, { "grad_norm": 0.19222024083137512, "learning_rate": 1.000982977905568e-06, "loss": 0.0147, "step": 93940 }, { "grad_norm": 0.29043540358543396, "learning_rate": 9.976936933656867e-07, "loss": 0.0125, "step": 93950 }, { "grad_norm": 0.12080293148756027, "learning_rate": 9.944097676443442e-07, "loss": 0.0134, "step": 93960 }, { "grad_norm": 0.15218883752822876, "learning_rate": 9.911312011006757e-07, "loss": 0.0101, "step": 93970 }, { "grad_norm": 0.27735498547554016, "learning_rate": 9.878579940932109e-07, "loss": 0.0152, "step": 93980 }, { "grad_norm": 0.15850621461868286, "learning_rate": 9.84590146979908e-07, "loss": 0.0137, "step": 93990 }, { "grad_norm": 0.22262555360794067, "learning_rate": 9.813276601181365e-07, "loss": 0.0161, "step": 94000 }, { "grad_norm": 0.510784387588501, "learning_rate": 9.780705338646724e-07, "loss": 0.0104, "step": 94010 }, { "grad_norm": 0.5072484612464905, "learning_rate": 9.748187685757083e-07, "loss": 0.0171, "step": 94020 }, { "grad_norm": 0.20221872627735138, "learning_rate": 9.715723646068543e-07, "loss": 0.01, "step": 94030 }, { "grad_norm": 0.15757380425930023, "learning_rate": 9.68331322313143e-07, "loss": 0.0111, "step": 94040 }, { "grad_norm": 0.19421425461769104, "learning_rate": 9.650956420489909e-07, "loss": 0.0131, "step": 94050 }, { "grad_norm": 0.19778728485107422, "learning_rate": 9.618653241682595e-07, "loss": 0.0186, "step": 94060 }, { "grad_norm": 0.22205597162246704, "learning_rate": 9.586403690242053e-07, "loss": 0.0133, "step": 94070 }, { "grad_norm": 0.19575271010398865, "learning_rate": 9.554207769695122e-07, "loss": 0.0153, "step": 94080 }, { "grad_norm": 0.23841221630573273, "learning_rate": 9.522065483562659e-07, "loss": 0.0148, "step": 94090 }, { "grad_norm": 0.3556775450706482, "learning_rate": 9.489976835359681e-07, "loss": 0.0139, "step": 94100 }, { "grad_norm": 0.171748548746109, "learning_rate": 9.457941828595329e-07, "loss": 0.0084, "step": 94110 }, { "grad_norm": 0.2841361463069916, "learning_rate": 9.425960466772965e-07, "loss": 0.0117, "step": 94120 }, { "grad_norm": 0.2065436691045761, "learning_rate": 9.394032753390014e-07, "loss": 0.0122, "step": 94130 }, { "grad_norm": 0.40767934918403625, "learning_rate": 9.362158691938017e-07, "loss": 0.0128, "step": 94140 }, { "grad_norm": 0.18107973039150238, "learning_rate": 9.330338285902685e-07, "loss": 0.0116, "step": 94150 }, { "grad_norm": 0.221259206533432, "learning_rate": 9.298571538763845e-07, "loss": 0.0112, "step": 94160 }, { "grad_norm": 0.11795185506343842, "learning_rate": 9.266858453995441e-07, "loss": 0.0145, "step": 94170 }, { "grad_norm": 0.12144122272729874, "learning_rate": 9.235199035065645e-07, "loss": 0.0136, "step": 94180 }, { "grad_norm": 0.15592031180858612, "learning_rate": 9.203593285436574e-07, "loss": 0.0122, "step": 94190 }, { "grad_norm": 0.33206281065940857, "learning_rate": 9.172041208564686e-07, "loss": 0.0134, "step": 94200 }, { "grad_norm": 0.14604032039642334, "learning_rate": 9.140542807900332e-07, "loss": 0.0109, "step": 94210 }, { "grad_norm": 0.1603134572505951, "learning_rate": 9.109098086888313e-07, "loss": 0.013, "step": 94220 }, { "grad_norm": 0.13584725558757782, "learning_rate": 9.077707048967265e-07, "loss": 0.0122, "step": 94230 }, { "grad_norm": 0.20652666687965393, "learning_rate": 9.046369697570057e-07, "loss": 0.0112, "step": 94240 }, { "grad_norm": 0.13910631835460663, "learning_rate": 9.015086036123721e-07, "loss": 0.0158, "step": 94250 }, { "grad_norm": 0.1098000705242157, "learning_rate": 8.983856068049413e-07, "loss": 0.0154, "step": 94260 }, { "grad_norm": 0.22297275066375732, "learning_rate": 8.952679796762398e-07, "loss": 0.0116, "step": 94270 }, { "grad_norm": 0.44127556681632996, "learning_rate": 8.921557225671951e-07, "loss": 0.016, "step": 94280 }, { "grad_norm": 0.18074525892734528, "learning_rate": 8.890488358181792e-07, "loss": 0.0169, "step": 94290 }, { "grad_norm": 0.12556321918964386, "learning_rate": 8.859473197689372e-07, "loss": 0.0113, "step": 94300 }, { "grad_norm": 0.32746678590774536, "learning_rate": 8.828511747586588e-07, "loss": 0.0161, "step": 94310 }, { "grad_norm": 0.15173941850662231, "learning_rate": 8.797604011259287e-07, "loss": 0.0116, "step": 94320 }, { "grad_norm": 0.41794389486312866, "learning_rate": 8.766749992087487e-07, "loss": 0.0144, "step": 94330 }, { "grad_norm": 0.28225645422935486, "learning_rate": 8.735949693445378e-07, "loss": 0.0117, "step": 94340 }, { "grad_norm": 0.24526885151863098, "learning_rate": 8.705203118701211e-07, "loss": 0.0142, "step": 94350 }, { "grad_norm": 0.2352294772863388, "learning_rate": 8.674510271217295e-07, "loss": 0.0198, "step": 94360 }, { "grad_norm": 0.2201186567544937, "learning_rate": 8.643871154350336e-07, "loss": 0.0158, "step": 94370 }, { "grad_norm": 0.23946727812290192, "learning_rate": 8.613285771450818e-07, "loss": 0.0154, "step": 94380 }, { "grad_norm": 0.21336498856544495, "learning_rate": 8.582754125863623e-07, "loss": 0.0152, "step": 94390 }, { "grad_norm": 0.4793837070465088, "learning_rate": 8.55227622092758e-07, "loss": 0.0154, "step": 94400 }, { "grad_norm": 0.20603077113628387, "learning_rate": 8.521852059975743e-07, "loss": 0.0132, "step": 94410 }, { "grad_norm": 0.37529948353767395, "learning_rate": 8.491481646335231e-07, "loss": 0.0157, "step": 94420 }, { "grad_norm": 0.3274744153022766, "learning_rate": 8.46116498332733e-07, "loss": 0.0128, "step": 94430 }, { "grad_norm": 0.16765595972537994, "learning_rate": 8.430902074267333e-07, "loss": 0.0116, "step": 94440 }, { "grad_norm": 0.14422695338726044, "learning_rate": 8.400692922464925e-07, "loss": 0.0102, "step": 94450 }, { "grad_norm": 0.13549263775348663, "learning_rate": 8.370537531223577e-07, "loss": 0.0114, "step": 94460 }, { "grad_norm": 0.13569432497024536, "learning_rate": 8.340435903841092e-07, "loss": 0.0121, "step": 94470 }, { "grad_norm": 0.3168726861476898, "learning_rate": 8.310388043609341e-07, "loss": 0.0147, "step": 94480 }, { "grad_norm": 0.536801815032959, "learning_rate": 8.280393953814303e-07, "loss": 0.0118, "step": 94490 }, { "grad_norm": 0.21259747445583344, "learning_rate": 8.250453637736077e-07, "loss": 0.0128, "step": 94500 }, { "grad_norm": 0.18969014286994934, "learning_rate": 8.22056709864899e-07, "loss": 0.0144, "step": 94510 }, { "grad_norm": 0.2418944239616394, "learning_rate": 8.190734339821204e-07, "loss": 0.0109, "step": 94520 }, { "grad_norm": 0.19963055849075317, "learning_rate": 8.160955364515333e-07, "loss": 0.0108, "step": 94530 }, { "grad_norm": 0.16665197908878326, "learning_rate": 8.131230175987936e-07, "loss": 0.0097, "step": 94540 }, { "grad_norm": 0.13876089453697205, "learning_rate": 8.101558777489693e-07, "loss": 0.0191, "step": 94550 }, { "grad_norm": 0.13504743576049805, "learning_rate": 8.071941172265396e-07, "loss": 0.014, "step": 94560 }, { "grad_norm": 0.4782986342906952, "learning_rate": 8.042377363553955e-07, "loss": 0.0148, "step": 94570 }, { "grad_norm": 0.18267115950584412, "learning_rate": 8.01286735458856e-07, "loss": 0.012, "step": 94580 }, { "grad_norm": 0.1465669870376587, "learning_rate": 7.983411148596298e-07, "loss": 0.0126, "step": 94590 }, { "grad_norm": 0.1705484241247177, "learning_rate": 7.954008748798425e-07, "loss": 0.0149, "step": 94600 }, { "grad_norm": 0.2534191906452179, "learning_rate": 7.924660158410369e-07, "loss": 0.0115, "step": 94610 }, { "grad_norm": 0.17614483833312988, "learning_rate": 7.895365380641728e-07, "loss": 0.0234, "step": 94620 }, { "grad_norm": 0.1561119556427002, "learning_rate": 7.866124418695941e-07, "loss": 0.0139, "step": 94630 }, { "grad_norm": 0.18780387938022614, "learning_rate": 7.83693727577095e-07, "loss": 0.0125, "step": 94640 }, { "grad_norm": 0.2597835958003998, "learning_rate": 7.807803955058535e-07, "loss": 0.0128, "step": 94650 }, { "grad_norm": 0.3028505742549896, "learning_rate": 7.778724459744702e-07, "loss": 0.015, "step": 94660 }, { "grad_norm": 0.47006064653396606, "learning_rate": 7.749698793009464e-07, "loss": 0.0167, "step": 94670 }, { "grad_norm": 0.16655388474464417, "learning_rate": 7.720726958027169e-07, "loss": 0.0132, "step": 94680 }, { "grad_norm": 0.8903468251228333, "learning_rate": 7.691808957965951e-07, "loss": 0.0147, "step": 94690 }, { "grad_norm": 0.23579999804496765, "learning_rate": 7.662944795988336e-07, "loss": 0.013, "step": 94700 }, { "grad_norm": 0.1801070272922516, "learning_rate": 7.634134475250854e-07, "loss": 0.013, "step": 94710 }, { "grad_norm": 0.2618879973888397, "learning_rate": 7.605377998904262e-07, "loss": 0.0154, "step": 94720 }, { "grad_norm": 0.23132242262363434, "learning_rate": 7.5766753700931e-07, "loss": 0.0141, "step": 94730 }, { "grad_norm": 0.3416522443294525, "learning_rate": 7.54802659195647e-07, "loss": 0.0115, "step": 94740 }, { "grad_norm": 0.24996908009052277, "learning_rate": 7.519431667627253e-07, "loss": 0.0118, "step": 94750 }, { "grad_norm": 0.2609139680862427, "learning_rate": 7.490890600232503e-07, "loss": 0.012, "step": 94760 }, { "grad_norm": 0.1583196371793747, "learning_rate": 7.462403392893502e-07, "loss": 0.0163, "step": 94770 }, { "grad_norm": 0.24831463396549225, "learning_rate": 7.433970048725536e-07, "loss": 0.0139, "step": 94780 }, { "grad_norm": 0.1237117126584053, "learning_rate": 7.405590570838062e-07, "loss": 0.0098, "step": 94790 }, { "grad_norm": 0.10970523208379745, "learning_rate": 7.377264962334596e-07, "loss": 0.0106, "step": 94800 }, { "grad_norm": 0.2595217823982239, "learning_rate": 7.348993226312773e-07, "loss": 0.0116, "step": 94810 }, { "grad_norm": 0.3730257451534271, "learning_rate": 7.32077536586434e-07, "loss": 0.013, "step": 94820 }, { "grad_norm": 0.1625930219888687, "learning_rate": 7.292611384075221e-07, "loss": 0.0137, "step": 94830 }, { "grad_norm": 0.16827334463596344, "learning_rate": 7.264501284025338e-07, "loss": 0.0141, "step": 94840 }, { "grad_norm": 0.6472328901290894, "learning_rate": 7.236445068788789e-07, "loss": 0.0147, "step": 94850 }, { "grad_norm": 0.26627087593078613, "learning_rate": 7.20844274143373e-07, "loss": 0.0123, "step": 94860 }, { "grad_norm": 0.17285600304603577, "learning_rate": 7.180494305022489e-07, "loss": 0.0158, "step": 94870 }, { "grad_norm": 0.14299030601978302, "learning_rate": 7.152599762611456e-07, "loss": 0.0133, "step": 94880 }, { "grad_norm": 0.14602026343345642, "learning_rate": 7.124759117251079e-07, "loss": 0.031, "step": 94890 }, { "grad_norm": 0.1377396583557129, "learning_rate": 7.096972371986033e-07, "loss": 0.0102, "step": 94900 }, { "grad_norm": 0.13975732028484344, "learning_rate": 7.069239529855054e-07, "loss": 0.016, "step": 94910 }, { "grad_norm": 0.2220412790775299, "learning_rate": 7.041560593890939e-07, "loss": 0.0115, "step": 94920 }, { "grad_norm": 0.2683263123035431, "learning_rate": 7.013935567120599e-07, "loss": 0.0115, "step": 94930 }, { "grad_norm": 0.14439399540424347, "learning_rate": 6.986364452565009e-07, "loss": 0.0096, "step": 94940 }, { "grad_norm": 0.17929160594940186, "learning_rate": 6.95884725323942e-07, "loss": 0.0108, "step": 94950 }, { "grad_norm": 0.2528790533542633, "learning_rate": 6.931383972152982e-07, "loss": 0.0145, "step": 94960 }, { "grad_norm": 0.17856401205062866, "learning_rate": 6.903974612309127e-07, "loss": 0.0166, "step": 94970 }, { "grad_norm": 0.23260129988193512, "learning_rate": 6.876619176705179e-07, "loss": 0.0107, "step": 94980 }, { "grad_norm": 0.25917503237724304, "learning_rate": 6.8493176683328e-07, "loss": 0.0158, "step": 94990 }, { "grad_norm": 0.3133563995361328, "learning_rate": 6.822070090177657e-07, "loss": 0.0217, "step": 95000 }, { "grad_norm": 0.3090555965900421, "learning_rate": 6.794876445219367e-07, "loss": 0.0169, "step": 95010 }, { "grad_norm": 0.35873425006866455, "learning_rate": 6.767736736431885e-07, "loss": 0.0132, "step": 95020 }, { "grad_norm": 0.29737555980682373, "learning_rate": 6.740650966783113e-07, "loss": 0.0148, "step": 95030 }, { "grad_norm": 0.37435659766197205, "learning_rate": 6.713619139235239e-07, "loss": 0.0107, "step": 95040 }, { "grad_norm": 0.0897759422659874, "learning_rate": 6.686641256744286e-07, "loss": 0.0144, "step": 95050 }, { "grad_norm": 0.1366322785615921, "learning_rate": 6.659717322260562e-07, "loss": 0.0176, "step": 95060 }, { "grad_norm": 0.16637109220027924, "learning_rate": 6.632847338728377e-07, "loss": 0.0132, "step": 95070 }, { "grad_norm": 0.11555840820074081, "learning_rate": 6.60603130908627e-07, "loss": 0.0098, "step": 95080 }, { "grad_norm": 0.16741353273391724, "learning_rate": 6.579269236266783e-07, "loss": 0.0123, "step": 95090 }, { "grad_norm": 0.2870389521121979, "learning_rate": 6.552561123196632e-07, "loss": 0.01, "step": 95100 }, { "grad_norm": 0.0856483057141304, "learning_rate": 6.525906972796425e-07, "loss": 0.0122, "step": 95110 }, { "grad_norm": 0.18447893857955933, "learning_rate": 6.499306787981108e-07, "loss": 0.0136, "step": 95120 }, { "grad_norm": 0.20047074556350708, "learning_rate": 6.472760571659631e-07, "loss": 0.0133, "step": 95130 }, { "grad_norm": 0.34600454568862915, "learning_rate": 6.446268326735061e-07, "loss": 0.0199, "step": 95140 }, { "grad_norm": 0.18756000697612762, "learning_rate": 6.41983005610458e-07, "loss": 0.0108, "step": 95150 }, { "grad_norm": 0.10974819958209991, "learning_rate": 6.393445762659378e-07, "loss": 0.0152, "step": 95160 }, { "grad_norm": 0.17971642315387726, "learning_rate": 6.36711544928481e-07, "loss": 0.0132, "step": 95170 }, { "grad_norm": 0.3588612973690033, "learning_rate": 6.340839118860298e-07, "loss": 0.017, "step": 95180 }, { "grad_norm": 0.2208077311515808, "learning_rate": 6.314616774259485e-07, "loss": 0.0126, "step": 95190 }, { "grad_norm": 0.1492266058921814, "learning_rate": 6.288448418349858e-07, "loss": 0.0129, "step": 95200 }, { "grad_norm": 0.27824893593788147, "learning_rate": 6.262334053993235e-07, "loss": 0.0142, "step": 95210 }, { "grad_norm": 0.14958344399929047, "learning_rate": 6.236273684045501e-07, "loss": 0.0106, "step": 95220 }, { "grad_norm": 0.17769981920719147, "learning_rate": 6.210267311356488e-07, "loss": 0.0136, "step": 95230 }, { "grad_norm": 0.2738686203956604, "learning_rate": 6.184314938770197e-07, "loss": 0.0151, "step": 95240 }, { "grad_norm": 0.22810739278793335, "learning_rate": 6.158416569124858e-07, "loss": 0.0119, "step": 95250 }, { "grad_norm": 0.19246937334537506, "learning_rate": 6.132572205252541e-07, "loss": 0.0137, "step": 95260 }, { "grad_norm": 0.11392150074243546, "learning_rate": 6.106781849979648e-07, "loss": 0.0101, "step": 95270 }, { "grad_norm": 0.2549493610858917, "learning_rate": 6.081045506126482e-07, "loss": 0.0123, "step": 95280 }, { "grad_norm": 0.21531203389167786, "learning_rate": 6.055363176507678e-07, "loss": 0.0137, "step": 95290 }, { "grad_norm": 0.13859303295612335, "learning_rate": 6.029734863931602e-07, "loss": 0.01, "step": 95300 }, { "grad_norm": 0.2495073825120926, "learning_rate": 6.004160571201179e-07, "loss": 0.013, "step": 95310 }, { "grad_norm": 0.25384581089019775, "learning_rate": 5.978640301112947e-07, "loss": 0.0115, "step": 95320 }, { "grad_norm": 0.25768622756004333, "learning_rate": 5.9531740564579e-07, "loss": 0.0158, "step": 95330 }, { "grad_norm": 0.0751492902636528, "learning_rate": 5.927761840020918e-07, "loss": 0.0094, "step": 95340 }, { "grad_norm": 0.14118751883506775, "learning_rate": 5.902403654581113e-07, "loss": 0.0092, "step": 95350 }, { "grad_norm": 0.37597689032554626, "learning_rate": 5.877099502911542e-07, "loss": 0.013, "step": 95360 }, { "grad_norm": 0.3786240220069885, "learning_rate": 5.851849387779496e-07, "loss": 0.0158, "step": 95370 }, { "grad_norm": 0.19956547021865845, "learning_rate": 5.826653311946262e-07, "loss": 0.0188, "step": 95380 }, { "grad_norm": 0.3212580680847168, "learning_rate": 5.801511278167249e-07, "loss": 0.0136, "step": 95390 }, { "grad_norm": 0.14738546311855316, "learning_rate": 5.776423289191868e-07, "loss": 0.0146, "step": 95400 }, { "grad_norm": 0.24327072501182556, "learning_rate": 5.751389347763869e-07, "loss": 0.0142, "step": 95410 }, { "grad_norm": 0.14677263796329498, "learning_rate": 5.726409456620841e-07, "loss": 0.0115, "step": 95420 }, { "grad_norm": 0.500808835029602, "learning_rate": 5.701483618494541e-07, "loss": 0.0105, "step": 95430 }, { "grad_norm": 0.18181024491786957, "learning_rate": 5.67661183611079e-07, "loss": 0.015, "step": 95440 }, { "grad_norm": 0.1760280430316925, "learning_rate": 5.651794112189579e-07, "loss": 0.0089, "step": 95450 }, { "grad_norm": 0.11567524075508118, "learning_rate": 5.627030449444903e-07, "loss": 0.0151, "step": 95460 }, { "grad_norm": 0.3056802451610565, "learning_rate": 5.60232085058493e-07, "loss": 0.0152, "step": 95470 }, { "grad_norm": 0.20562858879566193, "learning_rate": 5.577665318311776e-07, "loss": 0.0152, "step": 95480 }, { "grad_norm": 0.18139758706092834, "learning_rate": 5.553063855321838e-07, "loss": 0.0163, "step": 95490 }, { "grad_norm": 0.25956669449806213, "learning_rate": 5.528516464305467e-07, "loss": 0.0156, "step": 95500 }, { "grad_norm": 0.07584423571825027, "learning_rate": 5.504023147947068e-07, "loss": 0.0096, "step": 95510 }, { "grad_norm": 0.29492494463920593, "learning_rate": 5.479583908925279e-07, "loss": 0.0209, "step": 95520 }, { "grad_norm": 0.1931421458721161, "learning_rate": 5.455198749912626e-07, "loss": 0.0131, "step": 95530 }, { "grad_norm": 0.2823066711425781, "learning_rate": 5.430867673575979e-07, "loss": 0.0174, "step": 95540 }, { "grad_norm": 0.2509342133998871, "learning_rate": 5.406590682575985e-07, "loss": 0.0086, "step": 95550 }, { "grad_norm": 0.26274433732032776, "learning_rate": 5.382367779567687e-07, "loss": 0.0105, "step": 95560 }, { "grad_norm": 0.21194662153720856, "learning_rate": 5.358198967200023e-07, "loss": 0.01, "step": 95570 }, { "grad_norm": 0.1565653532743454, "learning_rate": 5.334084248115989e-07, "loss": 0.0153, "step": 95580 }, { "grad_norm": 0.24902263283729553, "learning_rate": 5.310023624952754e-07, "loss": 0.0119, "step": 95590 }, { "grad_norm": 0.11379153281450272, "learning_rate": 5.286017100341656e-07, "loss": 0.0133, "step": 95600 }, { "grad_norm": 0.1832612007856369, "learning_rate": 5.262064676907874e-07, "loss": 0.0147, "step": 95610 }, { "grad_norm": 0.20426234602928162, "learning_rate": 5.238166357270924e-07, "loss": 0.0115, "step": 95620 }, { "grad_norm": 0.3394119143486023, "learning_rate": 5.214322144044215e-07, "loss": 0.0136, "step": 95630 }, { "grad_norm": 0.21702633798122406, "learning_rate": 5.190532039835327e-07, "loss": 0.0167, "step": 95640 }, { "grad_norm": 0.32136303186416626, "learning_rate": 5.166796047245903e-07, "loss": 0.0119, "step": 95650 }, { "grad_norm": 0.25626903772354126, "learning_rate": 5.143114168871699e-07, "loss": 0.0145, "step": 95660 }, { "grad_norm": 0.4030296504497528, "learning_rate": 5.119486407302532e-07, "loss": 0.0182, "step": 95670 }, { "grad_norm": 0.0822155699133873, "learning_rate": 5.095912765122335e-07, "loss": 0.0194, "step": 95680 }, { "grad_norm": 0.19116386771202087, "learning_rate": 5.07239324490899e-07, "loss": 0.0114, "step": 95690 }, { "grad_norm": 0.19811637699604034, "learning_rate": 5.048927849234608e-07, "loss": 0.0183, "step": 95700 }, { "grad_norm": 0.11444554477930069, "learning_rate": 5.025516580665358e-07, "loss": 0.0131, "step": 95710 }, { "grad_norm": 0.3109772205352783, "learning_rate": 5.002159441761412e-07, "loss": 0.014, "step": 95720 }, { "grad_norm": 0.2146032601594925, "learning_rate": 4.978856435077062e-07, "loss": 0.0164, "step": 95730 }, { "grad_norm": 0.15839123725891113, "learning_rate": 4.955607563160769e-07, "loss": 0.0113, "step": 95740 }, { "grad_norm": 0.3429237902164459, "learning_rate": 4.932412828554944e-07, "loss": 0.014, "step": 95750 }, { "grad_norm": 0.21639148890972137, "learning_rate": 4.909272233796114e-07, "loss": 0.0122, "step": 95760 }, { "grad_norm": 0.1437605768442154, "learning_rate": 4.88618578141492e-07, "loss": 0.0141, "step": 95770 }, { "grad_norm": 0.11983528733253479, "learning_rate": 4.863153473936011e-07, "loss": 0.013, "step": 95780 }, { "grad_norm": 0.2158631682395935, "learning_rate": 4.840175313878259e-07, "loss": 0.0107, "step": 95790 }, { "grad_norm": 0.5141080617904663, "learning_rate": 4.817251303754433e-07, "loss": 0.0135, "step": 95800 }, { "grad_norm": 0.2087244838476181, "learning_rate": 4.794381446071583e-07, "loss": 0.0105, "step": 95810 }, { "grad_norm": 0.16864041984081268, "learning_rate": 4.771565743330597e-07, "loss": 0.0148, "step": 95820 }, { "grad_norm": 0.29026931524276733, "learning_rate": 4.7488041980266463e-07, "loss": 0.0133, "step": 95830 }, { "grad_norm": 0.3384160101413727, "learning_rate": 4.726096812648795e-07, "loss": 0.0151, "step": 95840 }, { "grad_norm": 1.548931360244751, "learning_rate": 4.703443589680445e-07, "loss": 0.0123, "step": 95850 }, { "grad_norm": 0.23292966187000275, "learning_rate": 4.6808445315987826e-07, "loss": 0.0143, "step": 95860 }, { "grad_norm": 0.24829594790935516, "learning_rate": 4.658299640875274e-07, "loss": 0.0123, "step": 95870 }, { "grad_norm": 0.20540311932563782, "learning_rate": 4.635808919975393e-07, "loss": 0.0112, "step": 95880 }, { "grad_norm": 0.23906554281711578, "learning_rate": 4.6133723713586706e-07, "loss": 0.013, "step": 95890 }, { "grad_norm": 0.17003439366817474, "learning_rate": 4.590989997478756e-07, "loss": 0.0109, "step": 95900 }, { "grad_norm": 0.15627539157867432, "learning_rate": 4.5686618007833027e-07, "loss": 0.0114, "step": 95910 }, { "grad_norm": 0.16641907393932343, "learning_rate": 4.54638778371419e-07, "loss": 0.0148, "step": 95920 }, { "grad_norm": 0.5964504480361938, "learning_rate": 4.524167948707192e-07, "loss": 0.0156, "step": 95930 }, { "grad_norm": 0.14030013978481293, "learning_rate": 4.5020022981921986e-07, "loss": 0.011, "step": 95940 }, { "grad_norm": 0.1917537897825241, "learning_rate": 4.47989083459327e-07, "loss": 0.0182, "step": 95950 }, { "grad_norm": 0.1767444908618927, "learning_rate": 4.457833560328528e-07, "loss": 0.0145, "step": 95960 }, { "grad_norm": 0.2265121340751648, "learning_rate": 4.4358304778100434e-07, "loss": 0.0142, "step": 95970 }, { "grad_norm": 0.1355040818452835, "learning_rate": 4.413881589444058e-07, "loss": 0.0122, "step": 95980 }, { "grad_norm": 0.15761980414390564, "learning_rate": 4.391986897630873e-07, "loss": 0.0119, "step": 95990 }, { "grad_norm": 0.35994356870651245, "learning_rate": 4.370146404764908e-07, "loss": 0.0116, "step": 96000 }, { "grad_norm": 0.22591441869735718, "learning_rate": 4.3483601132345285e-07, "loss": 0.015, "step": 96010 }, { "grad_norm": 0.16166992485523224, "learning_rate": 4.32662802542233e-07, "loss": 0.0129, "step": 96020 }, { "grad_norm": 0.2669063210487366, "learning_rate": 4.3049501437047444e-07, "loss": 0.0139, "step": 96030 }, { "grad_norm": 0.1748625636100769, "learning_rate": 4.283326470452653e-07, "loss": 0.017, "step": 96040 }, { "grad_norm": 0.1952759176492691, "learning_rate": 4.2617570080306647e-07, "loss": 0.0112, "step": 96050 }, { "grad_norm": 0.1627230942249298, "learning_rate": 4.2402417587975585e-07, "loss": 0.0121, "step": 96060 }, { "grad_norm": 0.25329160690307617, "learning_rate": 4.218780725106286e-07, "loss": 0.0133, "step": 96070 }, { "grad_norm": 0.46959245204925537, "learning_rate": 4.197373909303748e-07, "loss": 0.0184, "step": 96080 }, { "grad_norm": 0.17133422195911407, "learning_rate": 4.17602131373096e-07, "loss": 0.0097, "step": 96090 }, { "grad_norm": 0.29146867990493774, "learning_rate": 4.154722940723055e-07, "loss": 0.0213, "step": 96100 }, { "grad_norm": 0.2086680829524994, "learning_rate": 4.1334787926091136e-07, "loss": 0.0136, "step": 96110 }, { "grad_norm": 0.30103498697280884, "learning_rate": 4.1122888717124444e-07, "loss": 0.0143, "step": 96120 }, { "grad_norm": 0.17156344652175903, "learning_rate": 4.091153180350249e-07, "loss": 0.0155, "step": 96130 }, { "grad_norm": 0.1700643002986908, "learning_rate": 4.0700717208340125e-07, "loss": 0.0167, "step": 96140 }, { "grad_norm": 0.13098940253257751, "learning_rate": 4.049044495469112e-07, "loss": 0.0097, "step": 96150 }, { "grad_norm": 0.18641653656959534, "learning_rate": 4.0280715065549866e-07, "loss": 0.0128, "step": 96160 }, { "grad_norm": 0.21014444530010223, "learning_rate": 4.0071527563853575e-07, "loss": 0.0152, "step": 96170 }, { "grad_norm": 0.3540858030319214, "learning_rate": 3.9862882472477823e-07, "loss": 0.0125, "step": 96180 }, { "grad_norm": 0.2763548791408539, "learning_rate": 3.9654779814239375e-07, "loss": 0.0113, "step": 96190 }, { "grad_norm": 0.12225966155529022, "learning_rate": 3.944721961189668e-07, "loss": 0.0093, "step": 96200 }, { "grad_norm": 0.12145361304283142, "learning_rate": 3.9240201888147697e-07, "loss": 0.0099, "step": 96210 }, { "grad_norm": 0.35854747891426086, "learning_rate": 3.9033726665632096e-07, "loss": 0.0158, "step": 96220 }, { "grad_norm": 0.12127738445997238, "learning_rate": 3.882779396692959e-07, "loss": 0.0101, "step": 96230 }, { "grad_norm": 0.13306793570518494, "learning_rate": 3.8622403814559947e-07, "loss": 0.0111, "step": 96240 }, { "grad_norm": 0.2083505094051361, "learning_rate": 3.841755623098575e-07, "loss": 0.0098, "step": 96250 }, { "grad_norm": 0.3428638279438019, "learning_rate": 3.821325123860742e-07, "loss": 0.0161, "step": 96260 }, { "grad_norm": 0.3387907147407532, "learning_rate": 3.800948885976818e-07, "loss": 0.0151, "step": 96270 }, { "grad_norm": 0.24279914796352386, "learning_rate": 3.780626911675078e-07, "loss": 0.0147, "step": 96280 }, { "grad_norm": 0.2295670360326767, "learning_rate": 3.76035920317791e-07, "loss": 0.0114, "step": 96290 }, { "grad_norm": 0.2843547761440277, "learning_rate": 3.740145762701819e-07, "loss": 0.0119, "step": 96300 }, { "grad_norm": 0.13019269704818726, "learning_rate": 3.7199865924572585e-07, "loss": 0.0126, "step": 96310 }, { "grad_norm": 0.5526909828186035, "learning_rate": 3.699881694648799e-07, "loss": 0.0111, "step": 96320 }, { "grad_norm": 0.15173350274562836, "learning_rate": 3.679831071475126e-07, "loss": 0.017, "step": 96330 }, { "grad_norm": 0.2260495275259018, "learning_rate": 3.65983472512893e-07, "loss": 0.0132, "step": 96340 }, { "grad_norm": 0.33053824305534363, "learning_rate": 3.6398926577969616e-07, "loss": 0.014, "step": 96350 }, { "grad_norm": 0.24862168729305267, "learning_rate": 3.620004871660032e-07, "loss": 0.0142, "step": 96360 }, { "grad_norm": 0.29348963499069214, "learning_rate": 3.600171368893124e-07, "loss": 0.0099, "step": 96370 }, { "grad_norm": 0.20670166611671448, "learning_rate": 3.580392151665113e-07, "loss": 0.0129, "step": 96380 }, { "grad_norm": 0.12562879920005798, "learning_rate": 3.5606672221391024e-07, "loss": 0.0133, "step": 96390 }, { "grad_norm": 0.16872750222682953, "learning_rate": 3.540996582472089e-07, "loss": 0.0106, "step": 96400 }, { "grad_norm": 0.1213701143860817, "learning_rate": 3.521380234815297e-07, "loss": 0.0143, "step": 96410 }, { "grad_norm": 0.21146884560585022, "learning_rate": 3.5018181813138985e-07, "loss": 0.014, "step": 96420 }, { "grad_norm": 0.24923627078533173, "learning_rate": 3.482310424107238e-07, "loss": 0.0169, "step": 96430 }, { "grad_norm": 0.28365954756736755, "learning_rate": 3.4628569653285537e-07, "loss": 0.012, "step": 96440 }, { "grad_norm": 0.1945478767156601, "learning_rate": 3.44345780710531e-07, "loss": 0.0152, "step": 96450 }, { "grad_norm": 0.3569580018520355, "learning_rate": 3.424112951558978e-07, "loss": 0.0137, "step": 96460 }, { "grad_norm": 0.21720950305461884, "learning_rate": 3.40482240080503e-07, "loss": 0.0145, "step": 96470 }, { "grad_norm": 0.2160143107175827, "learning_rate": 3.3855861569531133e-07, "loss": 0.0129, "step": 96480 }, { "grad_norm": 0.20444288849830627, "learning_rate": 3.3664042221068226e-07, "loss": 0.0141, "step": 96490 }, { "grad_norm": 0.19410209357738495, "learning_rate": 3.347276598363924e-07, "loss": 0.0103, "step": 96500 }, { "grad_norm": 0.23421034216880798, "learning_rate": 3.328203287816134e-07, "loss": 0.0138, "step": 96510 }, { "grad_norm": 0.3805862367153168, "learning_rate": 3.3091842925493386e-07, "loss": 0.0127, "step": 96520 }, { "grad_norm": 0.15537230670452118, "learning_rate": 3.2902196146433196e-07, "loss": 0.0172, "step": 96530 }, { "grad_norm": 0.24687738716602325, "learning_rate": 3.2713092561720836e-07, "loss": 0.0151, "step": 96540 }, { "grad_norm": 2.6752450466156006, "learning_rate": 3.2524532192036996e-07, "loss": 0.0158, "step": 96550 }, { "grad_norm": 0.17478236556053162, "learning_rate": 3.2336515058001836e-07, "loss": 0.011, "step": 96560 }, { "grad_norm": 0.25751280784606934, "learning_rate": 3.2149041180176143e-07, "loss": 0.0108, "step": 96570 }, { "grad_norm": 0.16273923218250275, "learning_rate": 3.1962110579062954e-07, "loss": 0.0135, "step": 96580 }, { "grad_norm": 0.2076369673013687, "learning_rate": 3.177572327510314e-07, "loss": 0.0115, "step": 96590 }, { "grad_norm": 0.1844930499792099, "learning_rate": 3.158987928868151e-07, "loss": 0.0173, "step": 96600 }, { "grad_norm": 0.19633325934410095, "learning_rate": 3.1404578640120144e-07, "loss": 0.0138, "step": 96610 }, { "grad_norm": 0.5872905850410461, "learning_rate": 3.1219821349683933e-07, "loss": 0.0186, "step": 96620 }, { "grad_norm": 0.12058953940868378, "learning_rate": 3.103560743757783e-07, "loss": 0.0134, "step": 96630 }, { "grad_norm": 0.16771143674850464, "learning_rate": 3.0851936923946834e-07, "loss": 0.013, "step": 96640 }, { "grad_norm": 0.2581571936607361, "learning_rate": 3.06688098288771e-07, "loss": 0.0133, "step": 96650 }, { "grad_norm": 0.29478931427001953, "learning_rate": 3.0486226172394825e-07, "loss": 0.0153, "step": 96660 }, { "grad_norm": 0.3348988890647888, "learning_rate": 3.030418597446738e-07, "loss": 0.013, "step": 96670 }, { "grad_norm": 0.18260763585567474, "learning_rate": 3.012268925500272e-07, "loss": 0.0138, "step": 96680 }, { "grad_norm": 0.21317270398139954, "learning_rate": 2.994173603384831e-07, "loss": 0.0134, "step": 96690 }, { "grad_norm": 0.23299333453178406, "learning_rate": 2.9761326330793317e-07, "loss": 0.0114, "step": 96700 }, { "grad_norm": 0.1626458317041397, "learning_rate": 2.958146016556751e-07, "loss": 0.0126, "step": 96710 }, { "grad_norm": 0.23449085652828217, "learning_rate": 2.94021375578396e-07, "loss": 0.0131, "step": 96720 }, { "grad_norm": 0.2523106336593628, "learning_rate": 2.9223358527221135e-07, "loss": 0.0158, "step": 96730 }, { "grad_norm": 0.10933845490217209, "learning_rate": 2.904512309326257e-07, "loss": 0.0126, "step": 96740 }, { "grad_norm": 0.1433555781841278, "learning_rate": 2.8867431275455545e-07, "loss": 0.0154, "step": 96750 }, { "grad_norm": 0.28686878085136414, "learning_rate": 2.86902830932323e-07, "loss": 0.015, "step": 96760 }, { "grad_norm": 0.31052225828170776, "learning_rate": 2.851367856596565e-07, "loss": 0.0085, "step": 96770 }, { "grad_norm": 0.3988397419452667, "learning_rate": 2.83376177129685e-07, "loss": 0.0158, "step": 96780 }, { "grad_norm": 0.14091812074184418, "learning_rate": 2.8162100553494884e-07, "loss": 0.0108, "step": 96790 }, { "grad_norm": 0.18470120429992676, "learning_rate": 2.798712710673834e-07, "loss": 0.0116, "step": 96800 }, { "grad_norm": 0.19284401834011078, "learning_rate": 2.7812697391834676e-07, "loss": 0.0129, "step": 96810 }, { "grad_norm": 0.2378794550895691, "learning_rate": 2.7638811427858627e-07, "loss": 0.0105, "step": 96820 }, { "grad_norm": 0.31615737080574036, "learning_rate": 2.746546923382665e-07, "loss": 0.014, "step": 96830 }, { "grad_norm": 0.31602784991264343, "learning_rate": 2.729267082869469e-07, "loss": 0.0161, "step": 96840 }, { "grad_norm": 0.15030449628829956, "learning_rate": 2.712041623135986e-07, "loss": 0.0144, "step": 96850 }, { "grad_norm": 0.1865754872560501, "learning_rate": 2.694870546065931e-07, "loss": 0.0162, "step": 96860 }, { "grad_norm": 0.2320328950881958, "learning_rate": 2.677753853537246e-07, "loss": 0.0121, "step": 96870 }, { "grad_norm": 0.30728328227996826, "learning_rate": 2.660691547421601e-07, "loss": 0.0138, "step": 96880 }, { "grad_norm": 0.1781897246837616, "learning_rate": 2.6436836295850033e-07, "loss": 0.0105, "step": 96890 }, { "grad_norm": 0.2348705232143402, "learning_rate": 2.6267301018874093e-07, "loss": 0.0119, "step": 96900 }, { "grad_norm": 0.23830270767211914, "learning_rate": 2.6098309661828355e-07, "loss": 0.0156, "step": 96910 }, { "grad_norm": 0.16680839657783508, "learning_rate": 2.592986224319305e-07, "loss": 0.0172, "step": 96920 }, { "grad_norm": 0.26033321022987366, "learning_rate": 2.576195878139009e-07, "loss": 0.012, "step": 96930 }, { "grad_norm": 0.18072839081287384, "learning_rate": 2.5594599294780364e-07, "loss": 0.0195, "step": 96940 }, { "grad_norm": 0.17576618492603302, "learning_rate": 2.542778380166644e-07, "loss": 0.0159, "step": 96950 }, { "grad_norm": 0.1791963428258896, "learning_rate": 2.526151232029095e-07, "loss": 0.0126, "step": 96960 }, { "grad_norm": 0.20292240381240845, "learning_rate": 2.5095784868836567e-07, "loss": 0.0118, "step": 96970 }, { "grad_norm": 0.14961886405944824, "learning_rate": 2.493060146542825e-07, "loss": 0.0114, "step": 96980 }, { "grad_norm": 0.29503482580184937, "learning_rate": 2.476596212812876e-07, "loss": 0.011, "step": 96990 }, { "grad_norm": 0.1380809098482132, "learning_rate": 2.4601866874943703e-07, "loss": 0.0103, "step": 97000 }, { "grad_norm": 0.16231270134449005, "learning_rate": 2.4438315723818163e-07, "loss": 0.016, "step": 97010 }, { "grad_norm": 0.16730235517024994, "learning_rate": 2.4275308692637833e-07, "loss": 0.011, "step": 97020 }, { "grad_norm": 0.3092177212238312, "learning_rate": 2.411284579922901e-07, "loss": 0.0112, "step": 97030 }, { "grad_norm": 0.2500259280204773, "learning_rate": 2.395092706135804e-07, "loss": 0.0204, "step": 97040 }, { "grad_norm": 0.16992658376693726, "learning_rate": 2.3789552496731872e-07, "loss": 0.0107, "step": 97050 }, { "grad_norm": 0.19225481152534485, "learning_rate": 2.3628722122999158e-07, "loss": 0.0124, "step": 97060 }, { "grad_norm": 0.1729513555765152, "learning_rate": 2.3468435957747503e-07, "loss": 0.0111, "step": 97070 }, { "grad_norm": 0.11957523971796036, "learning_rate": 2.330869401850566e-07, "loss": 0.0134, "step": 97080 }, { "grad_norm": 0.2901398837566376, "learning_rate": 2.3149496322742437e-07, "loss": 0.0138, "step": 97090 }, { "grad_norm": 0.22778815031051636, "learning_rate": 2.2990842887868346e-07, "loss": 0.0142, "step": 97100 }, { "grad_norm": 0.22366994619369507, "learning_rate": 2.2832733731232292e-07, "loss": 0.0149, "step": 97110 }, { "grad_norm": 0.15685586631298065, "learning_rate": 2.2675168870124886e-07, "loss": 0.0118, "step": 97120 }, { "grad_norm": 0.13569201529026031, "learning_rate": 2.2518148321778455e-07, "loss": 0.0113, "step": 97130 }, { "grad_norm": 0.22692035138607025, "learning_rate": 2.2361672103363706e-07, "loss": 0.0136, "step": 97140 }, { "grad_norm": 0.3760940134525299, "learning_rate": 2.220574023199251e-07, "loss": 0.0146, "step": 97150 }, { "grad_norm": 0.28097403049468994, "learning_rate": 2.2050352724717894e-07, "loss": 0.0126, "step": 97160 }, { "grad_norm": 0.10437711328268051, "learning_rate": 2.1895509598532372e-07, "loss": 0.008, "step": 97170 }, { "grad_norm": 0.2588857412338257, "learning_rate": 2.1741210870369066e-07, "loss": 0.0145, "step": 97180 }, { "grad_norm": 0.16489437222480774, "learning_rate": 2.1587456557102814e-07, "loss": 0.008, "step": 97190 }, { "grad_norm": 0.31659844517707825, "learning_rate": 2.143424667554683e-07, "loss": 0.0151, "step": 97200 }, { "grad_norm": 0.3658197224140167, "learning_rate": 2.128158124245716e-07, "loss": 0.0115, "step": 97210 }, { "grad_norm": 0.16368673741817474, "learning_rate": 2.1129460274527668e-07, "loss": 0.0108, "step": 97220 }, { "grad_norm": 0.17904794216156006, "learning_rate": 2.0977883788395046e-07, "loss": 0.0118, "step": 97230 }, { "grad_norm": 0.1738096922636032, "learning_rate": 2.0826851800634928e-07, "loss": 0.0104, "step": 97240 }, { "grad_norm": 0.12157715111970901, "learning_rate": 2.0676364327764653e-07, "loss": 0.0148, "step": 97250 }, { "grad_norm": 0.10501997917890549, "learning_rate": 2.0526421386240502e-07, "loss": 0.012, "step": 97260 }, { "grad_norm": 0.19310054183006287, "learning_rate": 2.0377022992461025e-07, "loss": 0.0185, "step": 97270 }, { "grad_norm": 0.1798851490020752, "learning_rate": 2.0228169162763156e-07, "loss": 0.012, "step": 97280 }, { "grad_norm": 0.2632352113723755, "learning_rate": 2.0079859913425536e-07, "loss": 0.0134, "step": 97290 }, { "grad_norm": 0.2752716541290283, "learning_rate": 1.9932095260667417e-07, "loss": 0.0115, "step": 97300 }, { "grad_norm": 0.17710307240486145, "learning_rate": 1.9784875220648092e-07, "loss": 0.0111, "step": 97310 }, { "grad_norm": 0.612799346446991, "learning_rate": 1.9638199809466907e-07, "loss": 0.0125, "step": 97320 }, { "grad_norm": 0.26443833112716675, "learning_rate": 1.9492069043164364e-07, "loss": 0.0148, "step": 97330 }, { "grad_norm": 0.25589612126350403, "learning_rate": 1.9346482937721567e-07, "loss": 0.0121, "step": 97340 }, { "grad_norm": 0.2604415714740753, "learning_rate": 1.920144150905856e-07, "loss": 0.0124, "step": 97350 }, { "grad_norm": 0.1906847208738327, "learning_rate": 1.9056944773037655e-07, "loss": 0.0141, "step": 97360 }, { "grad_norm": 0.18605327606201172, "learning_rate": 1.8912992745460102e-07, "loss": 0.0134, "step": 97370 }, { "grad_norm": 0.2023962289094925, "learning_rate": 1.876958544206886e-07, "loss": 0.0149, "step": 97380 }, { "grad_norm": 0.08186829090118408, "learning_rate": 1.8626722878546942e-07, "loss": 0.021, "step": 97390 }, { "grad_norm": 0.1656477451324463, "learning_rate": 1.8484405070516851e-07, "loss": 0.0094, "step": 97400 }, { "grad_norm": 0.22334156930446625, "learning_rate": 1.8342632033542805e-07, "loss": 0.0143, "step": 97410 }, { "grad_norm": 0.20915259420871735, "learning_rate": 1.8201403783129066e-07, "loss": 0.0126, "step": 97420 }, { "grad_norm": 0.12384209781885147, "learning_rate": 1.806072033471884e-07, "loss": 0.0099, "step": 97430 }, { "grad_norm": 0.3482813537120819, "learning_rate": 1.79205817036987e-07, "loss": 0.0149, "step": 97440 }, { "grad_norm": 0.25984010100364685, "learning_rate": 1.7780987905393065e-07, "loss": 0.0141, "step": 97450 }, { "grad_norm": 0.255751371383667, "learning_rate": 1.764193895506805e-07, "loss": 0.012, "step": 97460 }, { "grad_norm": 0.3117106258869171, "learning_rate": 1.7503434867929824e-07, "loss": 0.0104, "step": 97470 }, { "grad_norm": 0.4319151043891907, "learning_rate": 1.7365475659125165e-07, "loss": 0.0091, "step": 97480 }, { "grad_norm": 0.13417913019657135, "learning_rate": 1.7228061343740332e-07, "loss": 0.0165, "step": 97490 }, { "grad_norm": 0.2018049955368042, "learning_rate": 1.7091191936803308e-07, "loss": 0.0146, "step": 97500 }, { "grad_norm": 0.37681737542152405, "learning_rate": 1.6954867453281563e-07, "loss": 0.0128, "step": 97510 }, { "grad_norm": 0.16500185430049896, "learning_rate": 1.6819087908084285e-07, "loss": 0.011, "step": 97520 }, { "grad_norm": 0.2560889422893524, "learning_rate": 1.6683853316058485e-07, "loss": 0.0119, "step": 97530 }, { "grad_norm": 0.16143903136253357, "learning_rate": 1.6549163691995106e-07, "loss": 0.0144, "step": 97540 }, { "grad_norm": 0.21286876499652863, "learning_rate": 1.641501905062237e-07, "loss": 0.0215, "step": 97550 }, { "grad_norm": 0.20932893455028534, "learning_rate": 1.628141940661021e-07, "loss": 0.0146, "step": 97560 }, { "grad_norm": 0.13153143227100372, "learning_rate": 1.6148364774569158e-07, "loss": 0.0103, "step": 97570 }, { "grad_norm": 0.13993500173091888, "learning_rate": 1.6015855169050355e-07, "loss": 0.0161, "step": 97580 }, { "grad_norm": 0.16928844153881073, "learning_rate": 1.5883890604543873e-07, "loss": 0.011, "step": 97590 }, { "grad_norm": 0.1490863412618637, "learning_rate": 1.5752471095482057e-07, "loss": 0.0086, "step": 97600 }, { "grad_norm": 0.10545940697193146, "learning_rate": 1.5621596656235638e-07, "loss": 0.0126, "step": 97610 }, { "grad_norm": 0.3899686932563782, "learning_rate": 1.5491267301117608e-07, "loss": 0.0136, "step": 97620 }, { "grad_norm": 0.09928487241268158, "learning_rate": 1.536148304438101e-07, "loss": 0.0116, "step": 97630 }, { "grad_norm": 0.28248825669288635, "learning_rate": 1.5232243900217823e-07, "loss": 0.0119, "step": 97640 }, { "grad_norm": 0.15467685461044312, "learning_rate": 1.51035498827623e-07, "loss": 0.0132, "step": 97650 }, { "grad_norm": 0.2662312388420105, "learning_rate": 1.497540100608763e-07, "loss": 0.0166, "step": 97660 }, { "grad_norm": 0.18281206488609314, "learning_rate": 1.4847797284208153e-07, "loss": 0.014, "step": 97670 }, { "grad_norm": 0.29178908467292786, "learning_rate": 1.472073873107882e-07, "loss": 0.0175, "step": 97680 }, { "grad_norm": 0.487434983253479, "learning_rate": 1.459422536059407e-07, "loss": 0.0128, "step": 97690 }, { "grad_norm": 0.13023073971271515, "learning_rate": 1.4468257186589508e-07, "loss": 0.0087, "step": 97700 }, { "grad_norm": 0.26083752512931824, "learning_rate": 1.4342834222840217e-07, "loss": 0.0113, "step": 97710 }, { "grad_norm": 0.2708565592765808, "learning_rate": 1.4217956483063566e-07, "loss": 0.0115, "step": 97720 }, { "grad_norm": 0.19946958124637604, "learning_rate": 1.409362398091474e-07, "loss": 0.0139, "step": 97730 }, { "grad_norm": 0.224117711186409, "learning_rate": 1.3969836729990638e-07, "loss": 0.0139, "step": 97740 }, { "grad_norm": 0.21039070188999176, "learning_rate": 1.384659474382932e-07, "loss": 0.01, "step": 97750 }, { "grad_norm": 0.17539742588996887, "learning_rate": 1.372389803590779e-07, "loss": 0.0126, "step": 97760 }, { "grad_norm": 0.22252702713012695, "learning_rate": 1.360174661964364e-07, "loss": 0.0104, "step": 97770 }, { "grad_norm": 0.2663957476615906, "learning_rate": 1.3480140508396188e-07, "loss": 0.0101, "step": 97780 }, { "grad_norm": 0.15671803057193756, "learning_rate": 1.335907971546313e-07, "loss": 0.0122, "step": 97790 }, { "grad_norm": 0.3309383690357208, "learning_rate": 1.3238564254083875e-07, "loss": 0.0171, "step": 97800 }, { "grad_norm": 0.22772471606731415, "learning_rate": 1.3118594137437323e-07, "loss": 0.0124, "step": 97810 }, { "grad_norm": 0.34608522057533264, "learning_rate": 1.2999169378644094e-07, "loss": 0.011, "step": 97820 }, { "grad_norm": 0.41541221737861633, "learning_rate": 1.2880289990763738e-07, "loss": 0.0169, "step": 97830 }, { "grad_norm": 0.2134835124015808, "learning_rate": 1.2761955986796968e-07, "loss": 0.0105, "step": 97840 }, { "grad_norm": 0.42397022247314453, "learning_rate": 1.264416737968399e-07, "loss": 0.0164, "step": 97850 }, { "grad_norm": 0.25916925072669983, "learning_rate": 1.2526924182307275e-07, "loss": 0.0122, "step": 97860 }, { "grad_norm": 0.12811920046806335, "learning_rate": 1.2410226407487124e-07, "loss": 0.0133, "step": 97870 }, { "grad_norm": 0.15226119756698608, "learning_rate": 1.229407406798555e-07, "loss": 0.0148, "step": 97880 }, { "grad_norm": 0.23781634867191315, "learning_rate": 1.2178467176505726e-07, "loss": 0.012, "step": 97890 }, { "grad_norm": 0.1644911915063858, "learning_rate": 1.2063405745689205e-07, "loss": 0.0144, "step": 97900 }, { "grad_norm": 0.18829038739204407, "learning_rate": 1.1948889788119812e-07, "loss": 0.0132, "step": 97910 }, { "grad_norm": 0.16371062397956848, "learning_rate": 1.1834919316320303e-07, "loss": 0.0102, "step": 97920 }, { "grad_norm": 0.276430606842041, "learning_rate": 1.1721494342754048e-07, "loss": 0.0144, "step": 97930 }, { "grad_norm": 0.2838088274002075, "learning_rate": 1.1608614879825563e-07, "loss": 0.0139, "step": 97940 }, { "grad_norm": 0.21094003319740295, "learning_rate": 1.1496280939879422e-07, "loss": 0.0157, "step": 97950 }, { "grad_norm": 0.2291085124015808, "learning_rate": 1.1384492535199687e-07, "loss": 0.013, "step": 97960 }, { "grad_norm": 0.29898586869239807, "learning_rate": 1.1273249678011578e-07, "loss": 0.0128, "step": 97970 }, { "grad_norm": 0.21032683551311493, "learning_rate": 1.1162552380480362e-07, "loss": 0.0107, "step": 97980 }, { "grad_norm": 0.311733603477478, "learning_rate": 1.1052400654711359e-07, "loss": 0.0116, "step": 97990 }, { "grad_norm": 0.3188871145248413, "learning_rate": 1.0942794512751597e-07, "loss": 0.0138, "step": 98000 }, { "grad_norm": 0.24323321878910065, "learning_rate": 1.0833733966587045e-07, "loss": 0.0131, "step": 98010 }, { "grad_norm": 0.14201998710632324, "learning_rate": 1.0725219028143718e-07, "loss": 0.0111, "step": 98020 }, { "grad_norm": 0.11657512933015823, "learning_rate": 1.0617249709289345e-07, "loss": 0.0122, "step": 98030 }, { "grad_norm": 0.204069122672081, "learning_rate": 1.0509826021831703e-07, "loss": 0.0094, "step": 98040 }, { "grad_norm": 0.17875657975673676, "learning_rate": 1.0402947977517508e-07, "loss": 0.016, "step": 98050 }, { "grad_norm": 0.34016573429107666, "learning_rate": 1.0296615588035185e-07, "loss": 0.0122, "step": 98060 }, { "grad_norm": 0.20613104104995728, "learning_rate": 1.0190828865012659e-07, "loss": 0.0139, "step": 98070 }, { "grad_norm": 0.13048553466796875, "learning_rate": 1.008558782001956e-07, "loss": 0.0137, "step": 98080 }, { "grad_norm": 0.23011848330497742, "learning_rate": 9.980892464563906e-08, "loss": 0.0134, "step": 98090 }, { "grad_norm": 0.2906040549278259, "learning_rate": 9.876742810095985e-08, "loss": 0.0118, "step": 98100 }, { "grad_norm": 0.19907529652118683, "learning_rate": 9.77313886800446e-08, "loss": 0.0136, "step": 98110 }, { "grad_norm": 0.22944819927215576, "learning_rate": 9.670080649619717e-08, "loss": 0.0214, "step": 98120 }, { "grad_norm": 0.16128873825073242, "learning_rate": 9.567568166212737e-08, "loss": 0.0151, "step": 98130 }, { "grad_norm": 0.1632593721151352, "learning_rate": 9.465601428992888e-08, "loss": 0.0168, "step": 98140 }, { "grad_norm": 0.1930260807275772, "learning_rate": 9.364180449111803e-08, "loss": 0.0167, "step": 98150 }, { "grad_norm": 0.1513087898492813, "learning_rate": 9.263305237661169e-08, "loss": 0.0125, "step": 98160 }, { "grad_norm": 0.2888353168964386, "learning_rate": 9.162975805671603e-08, "loss": 0.0145, "step": 98170 }, { "grad_norm": 0.1621038019657135, "learning_rate": 9.063192164115442e-08, "loss": 0.0104, "step": 98180 }, { "grad_norm": 0.16810227930545807, "learning_rate": 8.963954323904511e-08, "loss": 0.0123, "step": 98190 }, { "grad_norm": 0.2306395024061203, "learning_rate": 8.865262295891796e-08, "loss": 0.0095, "step": 98200 }, { "grad_norm": 0.2888695299625397, "learning_rate": 8.767116090870331e-08, "loss": 0.0115, "step": 98210 }, { "grad_norm": 0.17975878715515137, "learning_rate": 8.669515719572086e-08, "loss": 0.0168, "step": 98220 }, { "grad_norm": 0.2712341547012329, "learning_rate": 8.572461192671855e-08, "loss": 0.01, "step": 98230 }, { "grad_norm": 0.24970637261867523, "learning_rate": 8.475952520782815e-08, "loss": 0.0128, "step": 98240 }, { "grad_norm": 0.16530585289001465, "learning_rate": 8.379989714458747e-08, "loss": 0.0094, "step": 98250 }, { "grad_norm": 0.20854249596595764, "learning_rate": 8.284572784194034e-08, "loss": 0.0105, "step": 98260 }, { "grad_norm": 0.44231677055358887, "learning_rate": 8.189701740424216e-08, "loss": 0.0092, "step": 98270 }, { "grad_norm": 0.2477256953716278, "learning_rate": 8.095376593522663e-08, "loss": 0.0141, "step": 98280 }, { "grad_norm": 0.1618356853723526, "learning_rate": 8.001597353806123e-08, "loss": 0.0118, "step": 98290 }, { "grad_norm": 0.15657676756381989, "learning_rate": 7.90836403152917e-08, "loss": 0.0114, "step": 98300 }, { "grad_norm": 0.1925557255744934, "learning_rate": 7.815676636888092e-08, "loss": 0.0105, "step": 98310 }, { "grad_norm": 0.13466982543468475, "learning_rate": 7.723535180019226e-08, "loss": 0.0092, "step": 98320 }, { "grad_norm": 0.19997820258140564, "learning_rate": 7.6319396709984e-08, "loss": 0.0142, "step": 98330 }, { "grad_norm": 0.13291506469249725, "learning_rate": 7.540890119843158e-08, "loss": 0.0133, "step": 98340 }, { "grad_norm": 0.20184436440467834, "learning_rate": 7.450386536509979e-08, "loss": 0.0126, "step": 98350 }, { "grad_norm": 0.16861505806446075, "learning_rate": 7.360428930895947e-08, "loss": 0.0141, "step": 98360 }, { "grad_norm": 0.16035161912441254, "learning_rate": 7.271017312839302e-08, "loss": 0.0125, "step": 98370 }, { "grad_norm": 0.15923184156417847, "learning_rate": 7.182151692118333e-08, "loss": 0.0132, "step": 98380 }, { "grad_norm": 0.23712821304798126, "learning_rate": 7.093832078449713e-08, "loss": 0.0141, "step": 98390 }, { "grad_norm": 0.23923362791538239, "learning_rate": 7.006058481493493e-08, "loss": 0.0139, "step": 98400 }, { "grad_norm": 0.28203651309013367, "learning_rate": 6.918830910847552e-08, "loss": 0.0099, "step": 98410 }, { "grad_norm": 0.3838714361190796, "learning_rate": 6.832149376052033e-08, "loss": 0.0088, "step": 98420 }, { "grad_norm": 0.2040804922580719, "learning_rate": 6.746013886584912e-08, "loss": 0.0159, "step": 98430 }, { "grad_norm": 0.1762520968914032, "learning_rate": 6.660424451866431e-08, "loss": 0.0085, "step": 98440 }, { "grad_norm": 0.16081389784812927, "learning_rate": 6.575381081256327e-08, "loss": 0.0118, "step": 98450 }, { "grad_norm": 0.2092452049255371, "learning_rate": 6.49088378405549e-08, "loss": 0.0148, "step": 98460 }, { "grad_norm": 0.196304589509964, "learning_rate": 6.406932569503755e-08, "loss": 0.0161, "step": 98470 }, { "grad_norm": 0.12891307473182678, "learning_rate": 6.323527446782662e-08, "loss": 0.0108, "step": 98480 }, { "grad_norm": 0.08510005474090576, "learning_rate": 6.240668425012142e-08, "loss": 0.0131, "step": 98490 }, { "grad_norm": 0.1374914050102234, "learning_rate": 6.158355513254388e-08, "loss": 0.0114, "step": 98500 }, { "grad_norm": 0.08586720377206802, "learning_rate": 6.076588720510534e-08, "loss": 0.0075, "step": 98510 }, { "grad_norm": 0.21044375002384186, "learning_rate": 5.995368055722316e-08, "loss": 0.0176, "step": 98520 }, { "grad_norm": 0.20480814576148987, "learning_rate": 5.914693527773185e-08, "loss": 0.011, "step": 98530 }, { "grad_norm": 0.24108028411865234, "learning_rate": 5.834565145483861e-08, "loss": 0.0154, "step": 98540 }, { "grad_norm": 0.20167377591133118, "learning_rate": 5.754982917618446e-08, "loss": 0.0154, "step": 98550 }, { "grad_norm": 0.19432175159454346, "learning_rate": 5.675946852878866e-08, "loss": 0.011, "step": 98560 }, { "grad_norm": 0.20691433548927307, "learning_rate": 5.59745695990932e-08, "loss": 0.0143, "step": 98570 }, { "grad_norm": 0.18311567604541779, "learning_rate": 5.51951324729294e-08, "loss": 0.0136, "step": 98580 }, { "grad_norm": 0.15613189339637756, "learning_rate": 5.4421157235529093e-08, "loss": 0.0105, "step": 98590 }, { "grad_norm": 0.3130335509777069, "learning_rate": 5.365264397154679e-08, "loss": 0.0145, "step": 98600 }, { "grad_norm": 0.6538580656051636, "learning_rate": 5.288959276501526e-08, "loss": 0.0148, "step": 98610 }, { "grad_norm": 0.23922830820083618, "learning_rate": 5.2132003699378876e-08, "loss": 0.0116, "step": 98620 }, { "grad_norm": 0.2028811275959015, "learning_rate": 5.137987685749912e-08, "loss": 0.0135, "step": 98630 }, { "grad_norm": 0.18816208839416504, "learning_rate": 5.0633212321610227e-08, "loss": 0.0102, "step": 98640 }, { "grad_norm": 0.18260492384433746, "learning_rate": 4.989201017338019e-08, "loss": 0.0152, "step": 98650 }, { "grad_norm": 0.16287560760974884, "learning_rate": 4.9156270493860846e-08, "loss": 0.0136, "step": 98660 }, { "grad_norm": 0.20475034415721893, "learning_rate": 4.842599336351561e-08, "loss": 0.016, "step": 98670 }, { "grad_norm": 0.2073577344417572, "learning_rate": 4.7701178862197274e-08, "loss": 0.0165, "step": 98680 }, { "grad_norm": 0.20954063534736633, "learning_rate": 4.698182706918131e-08, "loss": 0.0157, "step": 98690 }, { "grad_norm": 0.20712712407112122, "learning_rate": 4.6267938063121466e-08, "loss": 0.0108, "step": 98700 }, { "grad_norm": 0.24647724628448486, "learning_rate": 4.555951192209973e-08, "loss": 0.013, "step": 98710 }, { "grad_norm": 0.19368265569210052, "learning_rate": 4.485654872358747e-08, "loss": 0.0122, "step": 98720 }, { "grad_norm": 0.20704545080661774, "learning_rate": 4.415904854445097e-08, "loss": 0.0167, "step": 98730 }, { "grad_norm": 0.14402922987937927, "learning_rate": 4.34670114609792e-08, "loss": 0.0104, "step": 98740 }, { "grad_norm": 0.3199903070926666, "learning_rate": 4.278043754884498e-08, "loss": 0.0162, "step": 98750 }, { "grad_norm": 0.11669963598251343, "learning_rate": 4.209932688312712e-08, "loss": 0.0109, "step": 98760 }, { "grad_norm": 0.1963784247636795, "learning_rate": 4.14236795383216e-08, "loss": 0.0142, "step": 98770 }, { "grad_norm": 0.21424277126789093, "learning_rate": 4.075349558830821e-08, "loss": 0.0102, "step": 98780 }, { "grad_norm": 0.3381419777870178, "learning_rate": 4.008877510638387e-08, "loss": 0.0114, "step": 98790 }, { "grad_norm": 0.2597411572933197, "learning_rate": 3.942951816523488e-08, "loss": 0.0156, "step": 98800 }, { "grad_norm": 0.29832160472869873, "learning_rate": 3.8775724836959125e-08, "loss": 0.0133, "step": 98810 }, { "grad_norm": 0.24498742818832397, "learning_rate": 3.812739519305497e-08, "loss": 0.0102, "step": 98820 }, { "grad_norm": 0.22564701735973358, "learning_rate": 3.748452930442126e-08, "loss": 0.0123, "step": 98830 }, { "grad_norm": 0.22014188766479492, "learning_rate": 3.6847127241362855e-08, "loss": 0.013, "step": 98840 }, { "grad_norm": 0.271019846200943, "learning_rate": 3.621518907358512e-08, "loss": 0.0201, "step": 98850 }, { "grad_norm": 0.21872973442077637, "learning_rate": 3.558871487019388e-08, "loss": 0.0121, "step": 98860 }, { "grad_norm": 0.17557604610919952, "learning_rate": 3.4967704699701007e-08, "loss": 0.0147, "step": 98870 }, { "grad_norm": 0.3950962424278259, "learning_rate": 3.435215863001884e-08, "loss": 0.0109, "step": 98880 }, { "grad_norm": 0.2527128756046295, "learning_rate": 3.374207672846019e-08, "loss": 0.0162, "step": 98890 }, { "grad_norm": 0.17059868574142456, "learning_rate": 3.313745906174392e-08, "loss": 0.0161, "step": 98900 }, { "grad_norm": 0.17533008754253387, "learning_rate": 3.253830569599492e-08, "loss": 0.0133, "step": 98910 }, { "grad_norm": 0.21071678400039673, "learning_rate": 3.1944616696727436e-08, "loss": 0.0094, "step": 98920 }, { "grad_norm": 0.21797089278697968, "learning_rate": 3.135639212887287e-08, "loss": 0.0131, "step": 98930 }, { "grad_norm": 0.2993500232696533, "learning_rate": 3.077363205675754e-08, "loss": 0.0128, "step": 98940 }, { "grad_norm": 0.26321643590927124, "learning_rate": 3.0196336544113804e-08, "loss": 0.0108, "step": 98950 }, { "grad_norm": 0.23471210896968842, "learning_rate": 2.9624505654063383e-08, "loss": 0.0159, "step": 98960 }, { "grad_norm": 0.11123396456241608, "learning_rate": 2.90581394491507e-08, "loss": 0.0161, "step": 98970 }, { "grad_norm": 0.1028321385383606, "learning_rate": 2.8497237991309545e-08, "loss": 0.0133, "step": 98980 }, { "grad_norm": 0.29775792360305786, "learning_rate": 2.7941801341879735e-08, "loss": 0.0134, "step": 98990 }, { "grad_norm": 0.20827749371528625, "learning_rate": 2.7391829561601578e-08, "loss": 0.0112, "step": 99000 }, { "grad_norm": 0.22461152076721191, "learning_rate": 2.6847322710621403e-08, "loss": 0.0094, "step": 99010 }, { "grad_norm": 0.19002030789852142, "learning_rate": 2.6308280848486022e-08, "loss": 0.0126, "step": 99020 }, { "grad_norm": 0.3638128936290741, "learning_rate": 2.5774704034137177e-08, "loss": 0.0166, "step": 99030 }, { "grad_norm": 0.24015723168849945, "learning_rate": 2.524659232593929e-08, "loss": 0.01, "step": 99040 }, { "grad_norm": 0.1955549418926239, "learning_rate": 2.472394578163506e-08, "loss": 0.0139, "step": 99050 }, { "grad_norm": 0.15676701068878174, "learning_rate": 2.4206764458378772e-08, "loss": 0.0156, "step": 99060 }, { "grad_norm": 0.28220584988594055, "learning_rate": 2.3695048412736285e-08, "loss": 0.0147, "step": 99070 }, { "grad_norm": 0.1199350357055664, "learning_rate": 2.3188797700668395e-08, "loss": 0.0177, "step": 99080 }, { "grad_norm": 0.2674621045589447, "learning_rate": 2.268801237753082e-08, "loss": 0.0108, "step": 99090 }, { "grad_norm": 0.12742988765239716, "learning_rate": 2.2192692498090862e-08, "loss": 0.0126, "step": 99100 }, { "grad_norm": 0.19125786423683167, "learning_rate": 2.170283811652185e-08, "loss": 0.0193, "step": 99110 }, { "grad_norm": 0.22470875084400177, "learning_rate": 2.12184492863865e-08, "loss": 0.0161, "step": 99120 }, { "grad_norm": 0.25898677110671997, "learning_rate": 2.073952606066465e-08, "loss": 0.0132, "step": 99130 }, { "grad_norm": 0.2728320360183716, "learning_rate": 2.026606849171997e-08, "loss": 0.015, "step": 99140 }, { "grad_norm": 0.24174663424491882, "learning_rate": 1.9798076631333262e-08, "loss": 0.0161, "step": 99150 }, { "grad_norm": 0.15398716926574707, "learning_rate": 1.933555053069136e-08, "loss": 0.0098, "step": 99160 }, { "grad_norm": 0.2688850164413452, "learning_rate": 1.887849024036492e-08, "loss": 0.0114, "step": 99170 }, { "grad_norm": 0.2157583385705948, "learning_rate": 1.8426895810341736e-08, "loss": 0.0159, "step": 99180 }, { "grad_norm": 0.10491684079170227, "learning_rate": 1.798076729000453e-08, "loss": 0.011, "step": 99190 }, { "grad_norm": 0.15950481593608856, "learning_rate": 1.75401047281476e-08, "loss": 0.0095, "step": 99200 }, { "grad_norm": 0.4883289039134979, "learning_rate": 1.7104908172954626e-08, "loss": 0.0158, "step": 99210 }, { "grad_norm": 0.1754850149154663, "learning_rate": 1.6675177672020868e-08, "loss": 0.0146, "step": 99220 }, { "grad_norm": 0.26868680119514465, "learning_rate": 1.6250913272342073e-08, "loss": 0.0107, "step": 99230 }, { "grad_norm": 0.18491002917289734, "learning_rate": 1.5832115020314453e-08, "loss": 0.0087, "step": 99240 }, { "grad_norm": 0.2538490891456604, "learning_rate": 1.5418782961734712e-08, "loss": 0.014, "step": 99250 }, { "grad_norm": 0.1760701835155487, "learning_rate": 1.501091714181113e-08, "loss": 0.0113, "step": 99260 }, { "grad_norm": 0.15283139050006866, "learning_rate": 1.460851760513582e-08, "loss": 0.0121, "step": 99270 }, { "grad_norm": 0.1860455721616745, "learning_rate": 1.4211584395723566e-08, "loss": 0.0112, "step": 99280 }, { "grad_norm": 0.22024419903755188, "learning_rate": 1.382011755697854e-08, "loss": 0.0169, "step": 99290 }, { "grad_norm": 0.6440390348434448, "learning_rate": 1.3434117131716494e-08, "loss": 0.0114, "step": 99300 }, { "grad_norm": 0.40442100167274475, "learning_rate": 1.3053583162142557e-08, "loss": 0.0138, "step": 99310 }, { "grad_norm": 0.24535433948040009, "learning_rate": 1.2678515689873438e-08, "loss": 0.0127, "step": 99320 }, { "grad_norm": 0.6137586832046509, "learning_rate": 1.2308914755931877e-08, "loss": 0.0133, "step": 99330 }, { "grad_norm": 0.2046302855014801, "learning_rate": 1.1944780400729993e-08, "loss": 0.0125, "step": 99340 }, { "grad_norm": 0.26413029432296753, "learning_rate": 1.1586112664085935e-08, "loss": 0.0133, "step": 99350 }, { "grad_norm": 0.24220891296863556, "learning_rate": 1.1232911585234984e-08, "loss": 0.0132, "step": 99360 }, { "grad_norm": 0.2635551989078522, "learning_rate": 1.0885177202790697e-08, "loss": 0.0145, "step": 99370 }, { "grad_norm": 0.16299770772457123, "learning_rate": 1.0542909554789315e-08, "loss": 0.0159, "step": 99380 }, { "grad_norm": 0.32791867852211, "learning_rate": 1.0206108678656456e-08, "loss": 0.0185, "step": 99390 }, { "grad_norm": 0.09756524860858917, "learning_rate": 9.874774611223769e-09, "loss": 0.0113, "step": 99400 }, { "grad_norm": 0.2774074673652649, "learning_rate": 9.548907388728933e-09, "loss": 0.0112, "step": 99410 }, { "grad_norm": 0.25968873500823975, "learning_rate": 9.228507046804557e-09, "loss": 0.0132, "step": 99420 }, { "grad_norm": 0.07530657202005386, "learning_rate": 8.91357362048928e-09, "loss": 0.0167, "step": 99430 }, { "grad_norm": 0.29132750630378723, "learning_rate": 8.604107144227769e-09, "loss": 0.014, "step": 99440 }, { "grad_norm": 0.22101019322872162, "learning_rate": 8.300107651859623e-09, "loss": 0.0131, "step": 99450 }, { "grad_norm": 0.2657967507839203, "learning_rate": 8.001575176630472e-09, "loss": 0.0116, "step": 99460 }, { "grad_norm": 0.2793196141719818, "learning_rate": 7.708509751186422e-09, "loss": 0.0098, "step": 99470 }, { "grad_norm": 0.1539784073829651, "learning_rate": 7.420911407579611e-09, "loss": 0.0108, "step": 99480 }, { "grad_norm": 0.2731967270374298, "learning_rate": 7.13878017725711e-09, "loss": 0.011, "step": 99490 }, { "grad_norm": 0.25270843505859375, "learning_rate": 6.8621160910720125e-09, "loss": 0.0127, "step": 99500 }, { "grad_norm": 0.2096681445837021, "learning_rate": 6.59091917928345e-09, "loss": 0.0116, "step": 99510 }, { "grad_norm": 0.3166196644306183, "learning_rate": 6.32518947155103e-09, "loss": 0.0141, "step": 99520 }, { "grad_norm": 0.22792737185955048, "learning_rate": 6.064926996929288e-09, "loss": 0.0101, "step": 99530 }, { "grad_norm": 0.26600566506385803, "learning_rate": 5.810131783884343e-09, "loss": 0.0093, "step": 99540 }, { "grad_norm": 0.20781956613063812, "learning_rate": 5.5608038602772415e-09, "loss": 0.0132, "step": 99550 }, { "grad_norm": 0.18501563370227814, "learning_rate": 5.316943253375062e-09, "loss": 0.011, "step": 99560 }, { "grad_norm": 0.13622792065143585, "learning_rate": 5.07854998984536e-09, "loss": 0.0106, "step": 99570 }, { "grad_norm": 0.27533069252967834, "learning_rate": 4.845624095756174e-09, "loss": 0.0143, "step": 99580 }, { "grad_norm": 0.44599246978759766, "learning_rate": 4.618165596587121e-09, "loss": 0.018, "step": 99590 }, { "grad_norm": 0.2094675898551941, "learning_rate": 4.396174517207197e-09, "loss": 0.0133, "step": 99600 }, { "grad_norm": 0.18294568359851837, "learning_rate": 4.179650881896979e-09, "loss": 0.0146, "step": 99610 }, { "grad_norm": 0.17476525902748108, "learning_rate": 3.968594714331975e-09, "loss": 0.0082, "step": 99620 }, { "grad_norm": 0.25942495465278625, "learning_rate": 3.763006037593719e-09, "loss": 0.011, "step": 99630 }, { "grad_norm": 0.4400799572467804, "learning_rate": 3.5628848741586786e-09, "loss": 0.0103, "step": 99640 }, { "grad_norm": 0.14418189227581024, "learning_rate": 3.368231245926001e-09, "loss": 0.0141, "step": 99650 }, { "grad_norm": 0.14542321860790253, "learning_rate": 3.179045174167561e-09, "loss": 0.0111, "step": 99660 }, { "grad_norm": 0.1821506768465042, "learning_rate": 2.9953266795834657e-09, "loss": 0.0143, "step": 99670 }, { "grad_norm": 0.18111921846866608, "learning_rate": 2.817075782263201e-09, "loss": 0.014, "step": 99680 }, { "grad_norm": 0.22432641685009003, "learning_rate": 2.6442925016967325e-09, "loss": 0.0104, "step": 99690 }, { "grad_norm": 0.2277468740940094, "learning_rate": 2.4769768567800555e-09, "loss": 0.013, "step": 99700 }, { "grad_norm": 0.17133644223213196, "learning_rate": 2.315128865809646e-09, "loss": 0.0134, "step": 99710 }, { "grad_norm": 0.34669530391693115, "learning_rate": 2.1587485464880097e-09, "loss": 0.0129, "step": 99720 }, { "grad_norm": 0.20091603696346283, "learning_rate": 2.0078359159125813e-09, "loss": 0.021, "step": 99730 }, { "grad_norm": 0.3985915184020996, "learning_rate": 1.8623909905923776e-09, "loss": 0.0144, "step": 99740 }, { "grad_norm": 0.14602141082286835, "learning_rate": 1.7224137864257916e-09, "loss": 0.0142, "step": 99750 }, { "grad_norm": 0.08290017396211624, "learning_rate": 1.5879043187283505e-09, "loss": 0.0134, "step": 99760 }, { "grad_norm": 0.23217734694480896, "learning_rate": 1.458862602204958e-09, "loss": 0.0157, "step": 99770 }, { "grad_norm": 0.13855214416980743, "learning_rate": 1.3352886509720996e-09, "loss": 0.0099, "step": 99780 }, { "grad_norm": 0.13437117636203766, "learning_rate": 1.2171824785356389e-09, "loss": 0.0141, "step": 99790 }, { "grad_norm": 0.42971959710121155, "learning_rate": 1.1045440978185718e-09, "loss": 0.0139, "step": 99800 }, { "grad_norm": 0.12676936388015747, "learning_rate": 9.973735211332714e-10, "loss": 0.0123, "step": 99810 }, { "grad_norm": 0.28271985054016113, "learning_rate": 8.956707602036929e-10, "loss": 0.0142, "step": 99820 }, { "grad_norm": 0.13829632103443146, "learning_rate": 7.994358261542712e-10, "loss": 0.0094, "step": 99830 }, { "grad_norm": 0.19435535371303558, "learning_rate": 7.086687295043692e-10, "loss": 0.0137, "step": 99840 }, { "grad_norm": 0.13188523054122925, "learning_rate": 6.23369480179381e-10, "loss": 0.0141, "step": 99850 }, { "grad_norm": 0.18308037519454956, "learning_rate": 5.43538087510731e-10, "loss": 0.0131, "step": 99860 }, { "grad_norm": 0.22145138680934906, "learning_rate": 4.691745602303233e-10, "loss": 0.021, "step": 99870 }, { "grad_norm": 0.17531011998653412, "learning_rate": 4.0027890646499034e-10, "loss": 0.023, "step": 99880 }, { "grad_norm": 0.30033358931541443, "learning_rate": 3.368511337531466e-10, "loss": 0.0111, "step": 99890 }, { "grad_norm": 0.22735336422920227, "learning_rate": 2.788912490281348e-10, "loss": 0.0106, "step": 99900 }, { "grad_norm": 0.17188219726085663, "learning_rate": 2.2639925863487954e-10, "loss": 0.0136, "step": 99910 }, { "grad_norm": 0.2898416221141815, "learning_rate": 1.7937516830213163e-10, "loss": 0.0123, "step": 99920 }, { "grad_norm": 0.2519025206565857, "learning_rate": 1.3781898318687702e-10, "loss": 0.0115, "step": 99930 }, { "grad_norm": 0.25101733207702637, "learning_rate": 1.0173070781882566e-10, "loss": 0.0189, "step": 99940 }, { "grad_norm": 0.16791893541812897, "learning_rate": 7.111034615592261e-11, "loss": 0.0123, "step": 99950 }, { "grad_norm": 0.19208039343357086, "learning_rate": 4.595790153993918e-11, "loss": 0.0155, "step": 99960 }, { "grad_norm": 0.33522626757621765, "learning_rate": 2.6273376729779585e-11, "loss": 0.0151, "step": 99970 }, { "grad_norm": 0.22597068548202515, "learning_rate": 1.2056773868174276e-11, "loss": 0.0135, "step": 99980 }, { "grad_norm": 0.1701051890850067, "learning_rate": 3.3080945149865926e-12, "loss": 0.0115, "step": 99990 }, { "grad_norm": 0.17228925228118896, "learning_rate": 2.7339625008337977e-14, "loss": 0.0138, "step": 100000 } ], "logging_steps": 10, "max_steps": 100000, "num_input_tokens_seen": 0, "num_train_epochs": 9223372036854775807, "save_steps": 20000, "stateful_callbacks": { "TrainerControl": { "args": { "should_epoch_stop": false, "should_evaluate": false, "should_log": false, "should_save": true, "should_training_stop": true }, "attributes": {} } }, "total_flos": 0.0, "train_batch_size": 8, "trial_name": null, "trial_params": null }