{ "best_global_step": null, "best_metric": null, "best_model_checkpoint": null, "epoch": 0.75, "eval_steps": 500, "global_step": 60000, "is_hyper_param_search": false, "is_local_process_zero": true, "is_world_process_zero": true, "log_history": [ { "grad_norm": 1.037879228591919, "learning_rate": 2.25e-07, "loss": 1.5137, "step": 10 }, { "grad_norm": 1.1727687120437622, "learning_rate": 4.75e-07, "loss": 1.5522, "step": 20 }, { "grad_norm": 1.2154980897903442, "learning_rate": 7.25e-07, "loss": 1.5359, "step": 30 }, { "grad_norm": 1.0449926853179932, "learning_rate": 9.75e-07, "loss": 1.5195, "step": 40 }, { "grad_norm": 1.1831287145614624, "learning_rate": 1.2250000000000001e-06, "loss": 1.503, "step": 50 }, { "grad_norm": 2.169935941696167, "learning_rate": 1.475e-06, "loss": 1.3874, "step": 60 }, { "grad_norm": 3.9084229469299316, "learning_rate": 1.7250000000000002e-06, "loss": 1.2721, "step": 70 }, { "grad_norm": 3.6716809272766113, "learning_rate": 1.975e-06, "loss": 1.228, "step": 80 }, { "grad_norm": 1.327792763710022, "learning_rate": 2.225e-06, "loss": 1.208, "step": 90 }, { "grad_norm": 1.3557993173599243, "learning_rate": 2.4750000000000004e-06, "loss": 1.1943, "step": 100 }, { "grad_norm": 1.5463353395462036, "learning_rate": 2.725e-06, "loss": 1.1625, "step": 110 }, { "grad_norm": 1.9236607551574707, "learning_rate": 2.975e-06, "loss": 1.1505, "step": 120 }, { "grad_norm": 1.410109043121338, "learning_rate": 3.225e-06, "loss": 1.1584, "step": 130 }, { "grad_norm": 1.6414321660995483, "learning_rate": 3.4750000000000006e-06, "loss": 1.1321, "step": 140 }, { "grad_norm": 1.5918344259262085, "learning_rate": 3.725e-06, "loss": 1.1194, "step": 150 }, { "grad_norm": 1.7760788202285767, "learning_rate": 3.975e-06, "loss": 1.0897, "step": 160 }, { "grad_norm": 1.251808524131775, "learning_rate": 4.225e-06, "loss": 1.1414, "step": 170 }, { "grad_norm": 1.717016339302063, "learning_rate": 4.475e-06, "loss": 1.1366, "step": 180 }, { "grad_norm": 1.540271520614624, "learning_rate": 4.7250000000000005e-06, "loss": 1.1339, "step": 190 }, { "grad_norm": 1.5406213998794556, "learning_rate": 4.975000000000001e-06, "loss": 1.1203, "step": 200 }, { "grad_norm": 1.7959346771240234, "learning_rate": 5.225e-06, "loss": 1.1115, "step": 210 }, { "grad_norm": 1.244638204574585, "learning_rate": 5.475e-06, "loss": 1.1047, "step": 220 }, { "grad_norm": 1.5479416847229004, "learning_rate": 5.725e-06, "loss": 1.1187, "step": 230 }, { "grad_norm": 1.0193949937820435, "learning_rate": 5.975e-06, "loss": 1.0935, "step": 240 }, { "grad_norm": 1.423858404159546, "learning_rate": 6.2250000000000005e-06, "loss": 1.0606, "step": 250 }, { "grad_norm": 1.337644338607788, "learning_rate": 6.475000000000001e-06, "loss": 1.0661, "step": 260 }, { "grad_norm": 2.561272144317627, "learning_rate": 6.725000000000001e-06, "loss": 1.0936, "step": 270 }, { "grad_norm": 1.4948028326034546, "learning_rate": 6.975000000000001e-06, "loss": 1.0846, "step": 280 }, { "grad_norm": 2.2803125381469727, "learning_rate": 7.2249999999999994e-06, "loss": 1.081, "step": 290 }, { "grad_norm": 1.5281028747558594, "learning_rate": 7.4750000000000004e-06, "loss": 1.0777, "step": 300 }, { "grad_norm": 1.4752074480056763, "learning_rate": 7.725e-06, "loss": 1.0666, "step": 310 }, { "grad_norm": 1.7930450439453125, "learning_rate": 7.975e-06, "loss": 1.0773, "step": 320 }, { "grad_norm": 1.6157379150390625, "learning_rate": 8.225e-06, "loss": 1.0669, "step": 330 }, { "grad_norm": 1.6522464752197266, "learning_rate": 8.475000000000001e-06, "loss": 1.0598, "step": 340 }, { "grad_norm": 1.4196891784667969, "learning_rate": 8.725e-06, "loss": 1.0661, "step": 350 }, { "grad_norm": 1.3078888654708862, "learning_rate": 8.975e-06, "loss": 1.0713, "step": 360 }, { "grad_norm": 1.760909914970398, "learning_rate": 9.225e-06, "loss": 1.0681, "step": 370 }, { "grad_norm": 1.5030139684677124, "learning_rate": 9.475e-06, "loss": 1.0613, "step": 380 }, { "grad_norm": 1.1136548519134521, "learning_rate": 9.725000000000001e-06, "loss": 1.0462, "step": 390 }, { "grad_norm": 1.015451192855835, "learning_rate": 9.975e-06, "loss": 1.0637, "step": 400 }, { "grad_norm": 1.4158153533935547, "learning_rate": 1.0225e-05, "loss": 1.0787, "step": 410 }, { "grad_norm": 1.1128000020980835, "learning_rate": 1.0475e-05, "loss": 1.0664, "step": 420 }, { "grad_norm": 1.8625755310058594, "learning_rate": 1.0725e-05, "loss": 1.0593, "step": 430 }, { "grad_norm": 1.4809938669204712, "learning_rate": 1.0975e-05, "loss": 1.0563, "step": 440 }, { "grad_norm": 1.376676082611084, "learning_rate": 1.1225e-05, "loss": 1.0273, "step": 450 }, { "grad_norm": 1.7043354511260986, "learning_rate": 1.1475000000000001e-05, "loss": 1.0218, "step": 460 }, { "grad_norm": 1.9443575143814087, "learning_rate": 1.1725e-05, "loss": 1.0089, "step": 470 }, { "grad_norm": 1.8139256238937378, "learning_rate": 1.1975e-05, "loss": 0.9903, "step": 480 }, { "grad_norm": 2.3824384212493896, "learning_rate": 1.2225e-05, "loss": 0.941, "step": 490 }, { "grad_norm": 2.25228214263916, "learning_rate": 1.2475e-05, "loss": 0.9007, "step": 500 }, { "grad_norm": 2.0508220195770264, "learning_rate": 1.2725000000000001e-05, "loss": 0.8873, "step": 510 }, { "grad_norm": 2.220612049102783, "learning_rate": 1.2975e-05, "loss": 0.8453, "step": 520 }, { "grad_norm": 2.390648365020752, "learning_rate": 1.3225000000000001e-05, "loss": 0.8196, "step": 530 }, { "grad_norm": 2.274989604949951, "learning_rate": 1.3475000000000002e-05, "loss": 0.7402, "step": 540 }, { "grad_norm": 2.382871627807617, "learning_rate": 1.3725000000000002e-05, "loss": 0.7466, "step": 550 }, { "grad_norm": 2.260232448577881, "learning_rate": 1.3975000000000003e-05, "loss": 0.7241, "step": 560 }, { "grad_norm": 2.041764736175537, "learning_rate": 1.4225e-05, "loss": 0.6714, "step": 570 }, { "grad_norm": 2.6011979579925537, "learning_rate": 1.4475e-05, "loss": 0.6289, "step": 580 }, { "grad_norm": 2.656662702560425, "learning_rate": 1.4725e-05, "loss": 0.6286, "step": 590 }, { "grad_norm": 2.562751054763794, "learning_rate": 1.4975e-05, "loss": 0.5903, "step": 600 }, { "grad_norm": 3.201646566390991, "learning_rate": 1.5225e-05, "loss": 0.5855, "step": 610 }, { "grad_norm": 4.484045028686523, "learning_rate": 1.5475e-05, "loss": 0.572, "step": 620 }, { "grad_norm": 2.4270331859588623, "learning_rate": 1.5725e-05, "loss": 0.5334, "step": 630 }, { "grad_norm": 2.671125888824463, "learning_rate": 1.5975000000000002e-05, "loss": 0.5082, "step": 640 }, { "grad_norm": 2.353020191192627, "learning_rate": 1.6225e-05, "loss": 0.4752, "step": 650 }, { "grad_norm": 3.055009126663208, "learning_rate": 1.6475e-05, "loss": 0.477, "step": 660 }, { "grad_norm": 1.9944814443588257, "learning_rate": 1.6725000000000003e-05, "loss": 0.419, "step": 670 }, { "grad_norm": 2.7653472423553467, "learning_rate": 1.6975000000000003e-05, "loss": 0.4374, "step": 680 }, { "grad_norm": 2.4157450199127197, "learning_rate": 1.7225e-05, "loss": 0.4182, "step": 690 }, { "grad_norm": 3.2104804515838623, "learning_rate": 1.7475e-05, "loss": 0.4178, "step": 700 }, { "grad_norm": 2.760042667388916, "learning_rate": 1.7725e-05, "loss": 0.36, "step": 710 }, { "grad_norm": 2.7874627113342285, "learning_rate": 1.7975e-05, "loss": 0.3941, "step": 720 }, { "grad_norm": 2.407700538635254, "learning_rate": 1.8225e-05, "loss": 0.3453, "step": 730 }, { "grad_norm": 2.0070459842681885, "learning_rate": 1.8475000000000002e-05, "loss": 0.3084, "step": 740 }, { "grad_norm": 2.5090672969818115, "learning_rate": 1.8725e-05, "loss": 0.3178, "step": 750 }, { "grad_norm": 2.4462997913360596, "learning_rate": 1.8975e-05, "loss": 0.307, "step": 760 }, { "grad_norm": 2.1990556716918945, "learning_rate": 1.9225e-05, "loss": 0.2901, "step": 770 }, { "grad_norm": 2.957414388656616, "learning_rate": 1.9475000000000002e-05, "loss": 0.3012, "step": 780 }, { "grad_norm": 2.3204641342163086, "learning_rate": 1.9725000000000002e-05, "loss": 0.2476, "step": 790 }, { "grad_norm": 2.804839611053467, "learning_rate": 1.9975e-05, "loss": 0.2315, "step": 800 }, { "grad_norm": 2.6934008598327637, "learning_rate": 2.0225000000000004e-05, "loss": 0.2457, "step": 810 }, { "grad_norm": 2.2680625915527344, "learning_rate": 2.0475e-05, "loss": 0.2356, "step": 820 }, { "grad_norm": 2.1193125247955322, "learning_rate": 2.0725e-05, "loss": 0.2536, "step": 830 }, { "grad_norm": 2.6217591762542725, "learning_rate": 2.0975e-05, "loss": 0.2496, "step": 840 }, { "grad_norm": 2.087918281555176, "learning_rate": 2.1225e-05, "loss": 0.2281, "step": 850 }, { "grad_norm": 2.7572813034057617, "learning_rate": 2.1475e-05, "loss": 0.1994, "step": 860 }, { "grad_norm": 2.521813154220581, "learning_rate": 2.1725e-05, "loss": 0.2319, "step": 870 }, { "grad_norm": 2.790123701095581, "learning_rate": 2.1975000000000002e-05, "loss": 0.198, "step": 880 }, { "grad_norm": 2.363448143005371, "learning_rate": 2.2225e-05, "loss": 0.1966, "step": 890 }, { "grad_norm": 2.434621572494507, "learning_rate": 2.2475e-05, "loss": 0.2073, "step": 900 }, { "grad_norm": 2.4361016750335693, "learning_rate": 2.2725000000000003e-05, "loss": 0.2082, "step": 910 }, { "grad_norm": 1.7169686555862427, "learning_rate": 2.2975000000000003e-05, "loss": 0.1785, "step": 920 }, { "grad_norm": 2.3596432209014893, "learning_rate": 2.3225000000000002e-05, "loss": 0.1929, "step": 930 }, { "grad_norm": 2.5425853729248047, "learning_rate": 2.3475e-05, "loss": 0.1765, "step": 940 }, { "grad_norm": 1.7971521615982056, "learning_rate": 2.3725e-05, "loss": 0.1697, "step": 950 }, { "grad_norm": 2.6153271198272705, "learning_rate": 2.3975e-05, "loss": 0.1714, "step": 960 }, { "grad_norm": 2.3928918838500977, "learning_rate": 2.4225e-05, "loss": 0.2124, "step": 970 }, { "grad_norm": 1.8646035194396973, "learning_rate": 2.4475000000000002e-05, "loss": 0.1603, "step": 980 }, { "grad_norm": 2.3566696643829346, "learning_rate": 2.4725e-05, "loss": 0.145, "step": 990 }, { "grad_norm": 2.7308926582336426, "learning_rate": 2.4975e-05, "loss": 0.1606, "step": 1000 }, { "grad_norm": 2.1203367710113525, "learning_rate": 2.5225e-05, "loss": 0.1579, "step": 1010 }, { "grad_norm": 2.1489696502685547, "learning_rate": 2.5475e-05, "loss": 0.1873, "step": 1020 }, { "grad_norm": 2.061659097671509, "learning_rate": 2.5725e-05, "loss": 0.1486, "step": 1030 }, { "grad_norm": 3.1126954555511475, "learning_rate": 2.5974999999999998e-05, "loss": 0.1657, "step": 1040 }, { "grad_norm": 2.429140329360962, "learning_rate": 2.6225e-05, "loss": 0.1618, "step": 1050 }, { "grad_norm": 2.2163336277008057, "learning_rate": 2.6475e-05, "loss": 0.1791, "step": 1060 }, { "grad_norm": 2.0744268894195557, "learning_rate": 2.6725e-05, "loss": 0.1947, "step": 1070 }, { "grad_norm": 2.059906005859375, "learning_rate": 2.6975000000000002e-05, "loss": 0.1949, "step": 1080 }, { "grad_norm": 1.44491708278656, "learning_rate": 2.7225e-05, "loss": 0.1657, "step": 1090 }, { "grad_norm": 1.9222891330718994, "learning_rate": 2.7475e-05, "loss": 0.1286, "step": 1100 }, { "grad_norm": 2.1851351261138916, "learning_rate": 2.7725e-05, "loss": 0.2032, "step": 1110 }, { "grad_norm": 1.5166385173797607, "learning_rate": 2.7975000000000002e-05, "loss": 0.1329, "step": 1120 }, { "grad_norm": 1.8134747743606567, "learning_rate": 2.8225e-05, "loss": 0.1597, "step": 1130 }, { "grad_norm": 2.429729700088501, "learning_rate": 2.8475e-05, "loss": 0.2096, "step": 1140 }, { "grad_norm": 2.1573598384857178, "learning_rate": 2.8725e-05, "loss": 0.1883, "step": 1150 }, { "grad_norm": 1.7671619653701782, "learning_rate": 2.8975000000000003e-05, "loss": 0.1663, "step": 1160 }, { "grad_norm": 1.843886375427246, "learning_rate": 2.9225000000000002e-05, "loss": 0.1729, "step": 1170 }, { "grad_norm": 1.8944097757339478, "learning_rate": 2.9475e-05, "loss": 0.1409, "step": 1180 }, { "grad_norm": 1.9085065126419067, "learning_rate": 2.9725000000000004e-05, "loss": 0.1624, "step": 1190 }, { "grad_norm": 3.052446126937866, "learning_rate": 2.9975000000000004e-05, "loss": 0.1639, "step": 1200 }, { "grad_norm": 3.762261152267456, "learning_rate": 3.0225000000000003e-05, "loss": 0.185, "step": 1210 }, { "grad_norm": 1.3778365850448608, "learning_rate": 3.0475000000000002e-05, "loss": 0.16, "step": 1220 }, { "grad_norm": 2.101390838623047, "learning_rate": 3.0725e-05, "loss": 0.164, "step": 1230 }, { "grad_norm": 1.6146845817565918, "learning_rate": 3.0975e-05, "loss": 0.1649, "step": 1240 }, { "grad_norm": 2.053755521774292, "learning_rate": 3.122500000000001e-05, "loss": 0.1444, "step": 1250 }, { "grad_norm": 1.8005622625350952, "learning_rate": 3.1475e-05, "loss": 0.1537, "step": 1260 }, { "grad_norm": 1.6267493963241577, "learning_rate": 3.1725e-05, "loss": 0.1408, "step": 1270 }, { "grad_norm": 1.2232205867767334, "learning_rate": 3.1975e-05, "loss": 0.1564, "step": 1280 }, { "grad_norm": 1.5040483474731445, "learning_rate": 3.2225e-05, "loss": 0.1575, "step": 1290 }, { "grad_norm": 1.4879934787750244, "learning_rate": 3.2474999999999997e-05, "loss": 0.1403, "step": 1300 }, { "grad_norm": 1.90468430519104, "learning_rate": 3.2725e-05, "loss": 0.1412, "step": 1310 }, { "grad_norm": 1.8211917877197266, "learning_rate": 3.2975e-05, "loss": 0.1539, "step": 1320 }, { "grad_norm": 1.306514859199524, "learning_rate": 3.3225e-05, "loss": 0.126, "step": 1330 }, { "grad_norm": 2.218984842300415, "learning_rate": 3.3475e-05, "loss": 0.1359, "step": 1340 }, { "grad_norm": 1.9924246072769165, "learning_rate": 3.3725e-05, "loss": 0.1707, "step": 1350 }, { "grad_norm": 2.1078197956085205, "learning_rate": 3.3975e-05, "loss": 0.1468, "step": 1360 }, { "grad_norm": 2.2034804821014404, "learning_rate": 3.4225e-05, "loss": 0.1265, "step": 1370 }, { "grad_norm": 1.9018747806549072, "learning_rate": 3.4475000000000005e-05, "loss": 0.1268, "step": 1380 }, { "grad_norm": 1.9234524965286255, "learning_rate": 3.4725000000000004e-05, "loss": 0.1405, "step": 1390 }, { "grad_norm": 1.9374897480010986, "learning_rate": 3.4975e-05, "loss": 0.1293, "step": 1400 }, { "grad_norm": 1.3408483266830444, "learning_rate": 3.5225e-05, "loss": 0.1448, "step": 1410 }, { "grad_norm": 1.755257487297058, "learning_rate": 3.5475e-05, "loss": 0.1378, "step": 1420 }, { "grad_norm": 1.4595181941986084, "learning_rate": 3.5725e-05, "loss": 0.115, "step": 1430 }, { "grad_norm": 1.7297132015228271, "learning_rate": 3.5975e-05, "loss": 0.1487, "step": 1440 }, { "grad_norm": 1.3787968158721924, "learning_rate": 3.6225000000000006e-05, "loss": 0.1584, "step": 1450 }, { "grad_norm": 1.399754285812378, "learning_rate": 3.6475000000000006e-05, "loss": 0.1577, "step": 1460 }, { "grad_norm": 1.4554942846298218, "learning_rate": 3.6725000000000005e-05, "loss": 0.1408, "step": 1470 }, { "grad_norm": 1.744908332824707, "learning_rate": 3.6975000000000004e-05, "loss": 0.134, "step": 1480 }, { "grad_norm": 1.5110799074172974, "learning_rate": 3.7225000000000004e-05, "loss": 0.1309, "step": 1490 }, { "grad_norm": 1.0980877876281738, "learning_rate": 3.7475e-05, "loss": 0.1431, "step": 1500 }, { "grad_norm": 1.353761911392212, "learning_rate": 3.7725e-05, "loss": 0.1148, "step": 1510 }, { "grad_norm": 1.720003366470337, "learning_rate": 3.7975e-05, "loss": 0.1325, "step": 1520 }, { "grad_norm": 1.3130781650543213, "learning_rate": 3.8225e-05, "loss": 0.1404, "step": 1530 }, { "grad_norm": 1.489071249961853, "learning_rate": 3.8475e-05, "loss": 0.1263, "step": 1540 }, { "grad_norm": 1.3261351585388184, "learning_rate": 3.8725e-05, "loss": 0.1267, "step": 1550 }, { "grad_norm": 1.4746724367141724, "learning_rate": 3.8975e-05, "loss": 0.1505, "step": 1560 }, { "grad_norm": 1.6342463493347168, "learning_rate": 3.9225e-05, "loss": 0.1195, "step": 1570 }, { "grad_norm": 1.473227858543396, "learning_rate": 3.9475000000000004e-05, "loss": 0.1329, "step": 1580 }, { "grad_norm": 1.5189464092254639, "learning_rate": 3.9725e-05, "loss": 0.1401, "step": 1590 }, { "grad_norm": 1.2169787883758545, "learning_rate": 3.9975e-05, "loss": 0.1186, "step": 1600 }, { "grad_norm": 1.3533635139465332, "learning_rate": 4.0225e-05, "loss": 0.1138, "step": 1610 }, { "grad_norm": 1.69119131565094, "learning_rate": 4.0475e-05, "loss": 0.1397, "step": 1620 }, { "grad_norm": 1.5980162620544434, "learning_rate": 4.0725e-05, "loss": 0.1741, "step": 1630 }, { "grad_norm": 1.1757237911224365, "learning_rate": 4.0975e-05, "loss": 0.1316, "step": 1640 }, { "grad_norm": 1.588876724243164, "learning_rate": 4.1225e-05, "loss": 0.1505, "step": 1650 }, { "grad_norm": 1.3505851030349731, "learning_rate": 4.1475000000000005e-05, "loss": 0.1371, "step": 1660 }, { "grad_norm": 1.2319358587265015, "learning_rate": 4.1725000000000005e-05, "loss": 0.1117, "step": 1670 }, { "grad_norm": 1.5464437007904053, "learning_rate": 4.1975000000000004e-05, "loss": 0.1231, "step": 1680 }, { "grad_norm": 1.2073643207550049, "learning_rate": 4.2225e-05, "loss": 0.1238, "step": 1690 }, { "grad_norm": 1.4788508415222168, "learning_rate": 4.2475e-05, "loss": 0.1348, "step": 1700 }, { "grad_norm": 1.4668617248535156, "learning_rate": 4.2725e-05, "loss": 0.1589, "step": 1710 }, { "grad_norm": 1.2036510705947876, "learning_rate": 4.2975e-05, "loss": 0.1352, "step": 1720 }, { "grad_norm": 1.1560145616531372, "learning_rate": 4.322500000000001e-05, "loss": 0.1351, "step": 1730 }, { "grad_norm": 1.4351190328598022, "learning_rate": 4.3475000000000006e-05, "loss": 0.1279, "step": 1740 }, { "grad_norm": 1.527893304824829, "learning_rate": 4.3725000000000006e-05, "loss": 0.1398, "step": 1750 }, { "grad_norm": 1.4643199443817139, "learning_rate": 4.3975e-05, "loss": 0.137, "step": 1760 }, { "grad_norm": 1.0274198055267334, "learning_rate": 4.4225e-05, "loss": 0.1201, "step": 1770 }, { "grad_norm": 1.215057373046875, "learning_rate": 4.4475e-05, "loss": 0.1371, "step": 1780 }, { "grad_norm": 1.3651336431503296, "learning_rate": 4.4725e-05, "loss": 0.1153, "step": 1790 }, { "grad_norm": 1.2669894695281982, "learning_rate": 4.4975e-05, "loss": 0.1302, "step": 1800 }, { "grad_norm": 0.8995741009712219, "learning_rate": 4.5225e-05, "loss": 0.1238, "step": 1810 }, { "grad_norm": 1.659932017326355, "learning_rate": 4.5475e-05, "loss": 0.1317, "step": 1820 }, { "grad_norm": 1.855787754058838, "learning_rate": 4.5725e-05, "loss": 0.1233, "step": 1830 }, { "grad_norm": 1.2045735120773315, "learning_rate": 4.5975e-05, "loss": 0.1307, "step": 1840 }, { "grad_norm": 1.1679713726043701, "learning_rate": 4.6225e-05, "loss": 0.1538, "step": 1850 }, { "grad_norm": 1.1519006490707397, "learning_rate": 4.6475000000000005e-05, "loss": 0.1617, "step": 1860 }, { "grad_norm": 1.1891311407089233, "learning_rate": 4.6725000000000004e-05, "loss": 0.1213, "step": 1870 }, { "grad_norm": 1.20973801612854, "learning_rate": 4.6975000000000003e-05, "loss": 0.1135, "step": 1880 }, { "grad_norm": 1.3255629539489746, "learning_rate": 4.7225e-05, "loss": 0.1268, "step": 1890 }, { "grad_norm": 0.8107821941375732, "learning_rate": 4.7475e-05, "loss": 0.1172, "step": 1900 }, { "grad_norm": 1.5125285387039185, "learning_rate": 4.7725e-05, "loss": 0.1523, "step": 1910 }, { "grad_norm": 1.093984842300415, "learning_rate": 4.7975e-05, "loss": 0.1231, "step": 1920 }, { "grad_norm": 1.3651256561279297, "learning_rate": 4.822500000000001e-05, "loss": 0.1128, "step": 1930 }, { "grad_norm": 1.3601741790771484, "learning_rate": 4.8475000000000006e-05, "loss": 0.1296, "step": 1940 }, { "grad_norm": 1.1680980920791626, "learning_rate": 4.8725000000000005e-05, "loss": 0.1065, "step": 1950 }, { "grad_norm": 1.4747815132141113, "learning_rate": 4.8975000000000005e-05, "loss": 0.1468, "step": 1960 }, { "grad_norm": 1.170799732208252, "learning_rate": 4.9225000000000004e-05, "loss": 0.0951, "step": 1970 }, { "grad_norm": 1.2053275108337402, "learning_rate": 4.9475e-05, "loss": 0.1355, "step": 1980 }, { "grad_norm": 1.0177805423736572, "learning_rate": 4.9725e-05, "loss": 0.1017, "step": 1990 }, { "grad_norm": 1.2434666156768799, "learning_rate": 4.9975e-05, "loss": 0.1311, "step": 2000 }, { "grad_norm": 1.2162634134292603, "learning_rate": 5.0225e-05, "loss": 0.1573, "step": 2010 }, { "grad_norm": 1.2801063060760498, "learning_rate": 5.047500000000001e-05, "loss": 0.1325, "step": 2020 }, { "grad_norm": 1.1615440845489502, "learning_rate": 5.0725e-05, "loss": 0.1092, "step": 2030 }, { "grad_norm": 1.3040574789047241, "learning_rate": 5.0975000000000006e-05, "loss": 0.1235, "step": 2040 }, { "grad_norm": 1.2717483043670654, "learning_rate": 5.1225e-05, "loss": 0.1307, "step": 2050 }, { "grad_norm": 1.018047571182251, "learning_rate": 5.1475000000000004e-05, "loss": 0.1192, "step": 2060 }, { "grad_norm": 0.8629509806632996, "learning_rate": 5.1725000000000004e-05, "loss": 0.1081, "step": 2070 }, { "grad_norm": 1.1743873357772827, "learning_rate": 5.197500000000001e-05, "loss": 0.1199, "step": 2080 }, { "grad_norm": 1.253340721130371, "learning_rate": 5.2225e-05, "loss": 0.1298, "step": 2090 }, { "grad_norm": 1.0681616067886353, "learning_rate": 5.247500000000001e-05, "loss": 0.1196, "step": 2100 }, { "grad_norm": 1.1930660009384155, "learning_rate": 5.2725e-05, "loss": 0.1437, "step": 2110 }, { "grad_norm": 1.0572696924209595, "learning_rate": 5.297500000000001e-05, "loss": 0.1118, "step": 2120 }, { "grad_norm": 1.2202396392822266, "learning_rate": 5.3225e-05, "loss": 0.1111, "step": 2130 }, { "grad_norm": 0.943488597869873, "learning_rate": 5.3475e-05, "loss": 0.1053, "step": 2140 }, { "grad_norm": 0.9266957640647888, "learning_rate": 5.3725000000000005e-05, "loss": 0.1416, "step": 2150 }, { "grad_norm": 0.9358584880828857, "learning_rate": 5.3975e-05, "loss": 0.1197, "step": 2160 }, { "grad_norm": 1.0563747882843018, "learning_rate": 5.4225000000000003e-05, "loss": 0.1071, "step": 2170 }, { "grad_norm": 1.252631664276123, "learning_rate": 5.4474999999999996e-05, "loss": 0.1298, "step": 2180 }, { "grad_norm": 0.9878680109977722, "learning_rate": 5.4725e-05, "loss": 0.1208, "step": 2190 }, { "grad_norm": 1.3607057332992554, "learning_rate": 5.4975e-05, "loss": 0.1227, "step": 2200 }, { "grad_norm": 1.198927879333496, "learning_rate": 5.522500000000001e-05, "loss": 0.1227, "step": 2210 }, { "grad_norm": 1.2198066711425781, "learning_rate": 5.5475e-05, "loss": 0.1166, "step": 2220 }, { "grad_norm": 1.1140468120574951, "learning_rate": 5.5725000000000006e-05, "loss": 0.1068, "step": 2230 }, { "grad_norm": 1.0755599737167358, "learning_rate": 5.5975e-05, "loss": 0.1054, "step": 2240 }, { "grad_norm": 1.042974591255188, "learning_rate": 5.6225000000000005e-05, "loss": 0.1217, "step": 2250 }, { "grad_norm": 1.0967648029327393, "learning_rate": 5.6475e-05, "loss": 0.1137, "step": 2260 }, { "grad_norm": 1.0321788787841797, "learning_rate": 5.6725e-05, "loss": 0.1129, "step": 2270 }, { "grad_norm": 1.048147201538086, "learning_rate": 5.6975e-05, "loss": 0.1103, "step": 2280 }, { "grad_norm": 1.217652678489685, "learning_rate": 5.722500000000001e-05, "loss": 0.1406, "step": 2290 }, { "grad_norm": 1.146368384361267, "learning_rate": 5.7475e-05, "loss": 0.1042, "step": 2300 }, { "grad_norm": 1.2576117515563965, "learning_rate": 5.772500000000001e-05, "loss": 0.1128, "step": 2310 }, { "grad_norm": 0.9563098549842834, "learning_rate": 5.7975e-05, "loss": 0.098, "step": 2320 }, { "grad_norm": 1.1709794998168945, "learning_rate": 5.8225000000000006e-05, "loss": 0.1316, "step": 2330 }, { "grad_norm": 0.8115894198417664, "learning_rate": 5.8475000000000005e-05, "loss": 0.1113, "step": 2340 }, { "grad_norm": 1.0980546474456787, "learning_rate": 5.8725000000000004e-05, "loss": 0.1252, "step": 2350 }, { "grad_norm": 0.8628365397453308, "learning_rate": 5.8975000000000004e-05, "loss": 0.1245, "step": 2360 }, { "grad_norm": 0.9222819209098816, "learning_rate": 5.922500000000001e-05, "loss": 0.1122, "step": 2370 }, { "grad_norm": 0.8934987783432007, "learning_rate": 5.9475e-05, "loss": 0.1142, "step": 2380 }, { "grad_norm": 0.9733268618583679, "learning_rate": 5.9724999999999995e-05, "loss": 0.126, "step": 2390 }, { "grad_norm": 0.7214210629463196, "learning_rate": 5.9975e-05, "loss": 0.1255, "step": 2400 }, { "grad_norm": 1.0429095029830933, "learning_rate": 6.0225e-05, "loss": 0.1112, "step": 2410 }, { "grad_norm": 0.8871645927429199, "learning_rate": 6.0475000000000006e-05, "loss": 0.149, "step": 2420 }, { "grad_norm": 1.336459755897522, "learning_rate": 6.0725e-05, "loss": 0.1143, "step": 2430 }, { "grad_norm": 1.1515284776687622, "learning_rate": 6.0975000000000005e-05, "loss": 0.1277, "step": 2440 }, { "grad_norm": 0.8831538558006287, "learning_rate": 6.1225e-05, "loss": 0.0992, "step": 2450 }, { "grad_norm": 0.7443053126335144, "learning_rate": 6.1475e-05, "loss": 0.1008, "step": 2460 }, { "grad_norm": 1.1608682870864868, "learning_rate": 6.1725e-05, "loss": 0.1057, "step": 2470 }, { "grad_norm": 1.0224337577819824, "learning_rate": 6.1975e-05, "loss": 0.135, "step": 2480 }, { "grad_norm": 1.1087498664855957, "learning_rate": 6.2225e-05, "loss": 0.1081, "step": 2490 }, { "grad_norm": 0.7603142857551575, "learning_rate": 6.2475e-05, "loss": 0.1005, "step": 2500 }, { "grad_norm": 0.6725986003875732, "learning_rate": 6.2725e-05, "loss": 0.0995, "step": 2510 }, { "grad_norm": 0.9372966885566711, "learning_rate": 6.297500000000001e-05, "loss": 0.1129, "step": 2520 }, { "grad_norm": 1.4925291538238525, "learning_rate": 6.3225e-05, "loss": 0.1046, "step": 2530 }, { "grad_norm": 0.9479634761810303, "learning_rate": 6.347500000000001e-05, "loss": 0.1107, "step": 2540 }, { "grad_norm": 0.8416444659233093, "learning_rate": 6.3725e-05, "loss": 0.1134, "step": 2550 }, { "grad_norm": 1.35062575340271, "learning_rate": 6.397500000000001e-05, "loss": 0.1271, "step": 2560 }, { "grad_norm": 1.2156401872634888, "learning_rate": 6.4225e-05, "loss": 0.1127, "step": 2570 }, { "grad_norm": 1.1134980916976929, "learning_rate": 6.447500000000001e-05, "loss": 0.1198, "step": 2580 }, { "grad_norm": 0.7143936157226562, "learning_rate": 6.4725e-05, "loss": 0.1029, "step": 2590 }, { "grad_norm": 0.8293321132659912, "learning_rate": 6.497500000000001e-05, "loss": 0.1253, "step": 2600 }, { "grad_norm": 1.0394586324691772, "learning_rate": 6.5225e-05, "loss": 0.1094, "step": 2610 }, { "grad_norm": 0.8121738433837891, "learning_rate": 6.5475e-05, "loss": 0.1037, "step": 2620 }, { "grad_norm": 1.0385233163833618, "learning_rate": 6.5725e-05, "loss": 0.1004, "step": 2630 }, { "grad_norm": 1.015921950340271, "learning_rate": 6.5975e-05, "loss": 0.1222, "step": 2640 }, { "grad_norm": 0.8493093252182007, "learning_rate": 6.6225e-05, "loss": 0.1196, "step": 2650 }, { "grad_norm": 0.9565926790237427, "learning_rate": 6.6475e-05, "loss": 0.1145, "step": 2660 }, { "grad_norm": 0.9933330416679382, "learning_rate": 6.672500000000001e-05, "loss": 0.1242, "step": 2670 }, { "grad_norm": 1.0645827054977417, "learning_rate": 6.6975e-05, "loss": 0.1493, "step": 2680 }, { "grad_norm": 0.7827352285385132, "learning_rate": 6.722500000000001e-05, "loss": 0.1147, "step": 2690 }, { "grad_norm": 0.994665801525116, "learning_rate": 6.7475e-05, "loss": 0.1018, "step": 2700 }, { "grad_norm": 0.7434802651405334, "learning_rate": 6.7725e-05, "loss": 0.1058, "step": 2710 }, { "grad_norm": 0.7673996090888977, "learning_rate": 6.7975e-05, "loss": 0.1299, "step": 2720 }, { "grad_norm": 1.1512700319290161, "learning_rate": 6.8225e-05, "loss": 0.1073, "step": 2730 }, { "grad_norm": 0.67215895652771, "learning_rate": 6.8475e-05, "loss": 0.1054, "step": 2740 }, { "grad_norm": 1.0621843338012695, "learning_rate": 6.8725e-05, "loss": 0.1124, "step": 2750 }, { "grad_norm": 0.8716442584991455, "learning_rate": 6.8975e-05, "loss": 0.0996, "step": 2760 }, { "grad_norm": 0.9091720581054688, "learning_rate": 6.9225e-05, "loss": 0.1164, "step": 2770 }, { "grad_norm": 0.8813672661781311, "learning_rate": 6.9475e-05, "loss": 0.1209, "step": 2780 }, { "grad_norm": 0.9543070197105408, "learning_rate": 6.9725e-05, "loss": 0.1114, "step": 2790 }, { "grad_norm": 0.8144809603691101, "learning_rate": 6.997500000000001e-05, "loss": 0.096, "step": 2800 }, { "grad_norm": 1.086455225944519, "learning_rate": 7.022500000000001e-05, "loss": 0.1413, "step": 2810 }, { "grad_norm": 1.3268401622772217, "learning_rate": 7.0475e-05, "loss": 0.1112, "step": 2820 }, { "grad_norm": 0.6280853748321533, "learning_rate": 7.072500000000001e-05, "loss": 0.1008, "step": 2830 }, { "grad_norm": 0.7107454538345337, "learning_rate": 7.0975e-05, "loss": 0.1344, "step": 2840 }, { "grad_norm": 1.1063814163208008, "learning_rate": 7.122500000000001e-05, "loss": 0.1184, "step": 2850 }, { "grad_norm": 0.9749586582183838, "learning_rate": 7.1475e-05, "loss": 0.105, "step": 2860 }, { "grad_norm": 0.8268616199493408, "learning_rate": 7.172500000000001e-05, "loss": 0.091, "step": 2870 }, { "grad_norm": 0.9473592042922974, "learning_rate": 7.1975e-05, "loss": 0.099, "step": 2880 }, { "grad_norm": 0.7489947080612183, "learning_rate": 7.2225e-05, "loss": 0.1007, "step": 2890 }, { "grad_norm": 0.8029493689537048, "learning_rate": 7.2475e-05, "loss": 0.1073, "step": 2900 }, { "grad_norm": 1.0153167247772217, "learning_rate": 7.272499999999999e-05, "loss": 0.0993, "step": 2910 }, { "grad_norm": 0.6517780423164368, "learning_rate": 7.2975e-05, "loss": 0.1063, "step": 2920 }, { "grad_norm": 0.7716829180717468, "learning_rate": 7.3225e-05, "loss": 0.085, "step": 2930 }, { "grad_norm": 0.8208396434783936, "learning_rate": 7.347500000000001e-05, "loss": 0.0984, "step": 2940 }, { "grad_norm": 1.1097429990768433, "learning_rate": 7.3725e-05, "loss": 0.0923, "step": 2950 }, { "grad_norm": 0.7229036688804626, "learning_rate": 7.397500000000001e-05, "loss": 0.0973, "step": 2960 }, { "grad_norm": 0.732600748538971, "learning_rate": 7.4225e-05, "loss": 0.0851, "step": 2970 }, { "grad_norm": 0.9079328179359436, "learning_rate": 7.447500000000001e-05, "loss": 0.1073, "step": 2980 }, { "grad_norm": 0.9559067487716675, "learning_rate": 7.4725e-05, "loss": 0.1124, "step": 2990 }, { "grad_norm": 0.8981481194496155, "learning_rate": 7.4975e-05, "loss": 0.1102, "step": 3000 }, { "grad_norm": 0.6410384774208069, "learning_rate": 7.5225e-05, "loss": 0.0881, "step": 3010 }, { "grad_norm": 0.6838392019271851, "learning_rate": 7.5475e-05, "loss": 0.0906, "step": 3020 }, { "grad_norm": 0.9712271094322205, "learning_rate": 7.5725e-05, "loss": 0.0955, "step": 3030 }, { "grad_norm": 0.8994407653808594, "learning_rate": 7.5975e-05, "loss": 0.098, "step": 3040 }, { "grad_norm": 1.0481009483337402, "learning_rate": 7.6225e-05, "loss": 0.112, "step": 3050 }, { "grad_norm": 0.6689869165420532, "learning_rate": 7.6475e-05, "loss": 0.1023, "step": 3060 }, { "grad_norm": 0.6697691679000854, "learning_rate": 7.672500000000001e-05, "loss": 0.0842, "step": 3070 }, { "grad_norm": 0.6981144547462463, "learning_rate": 7.697500000000001e-05, "loss": 0.1014, "step": 3080 }, { "grad_norm": 1.151863694190979, "learning_rate": 7.722500000000001e-05, "loss": 0.123, "step": 3090 }, { "grad_norm": 1.0347213745117188, "learning_rate": 7.747500000000001e-05, "loss": 0.0947, "step": 3100 }, { "grad_norm": 0.7930265665054321, "learning_rate": 7.7725e-05, "loss": 0.0794, "step": 3110 }, { "grad_norm": 0.7757615447044373, "learning_rate": 7.797500000000001e-05, "loss": 0.0898, "step": 3120 }, { "grad_norm": 0.9546329379081726, "learning_rate": 7.8225e-05, "loss": 0.0867, "step": 3130 }, { "grad_norm": 0.7230708599090576, "learning_rate": 7.8475e-05, "loss": 0.1114, "step": 3140 }, { "grad_norm": 0.7930948138237, "learning_rate": 7.8725e-05, "loss": 0.099, "step": 3150 }, { "grad_norm": 0.5867165923118591, "learning_rate": 7.8975e-05, "loss": 0.105, "step": 3160 }, { "grad_norm": 0.6422441601753235, "learning_rate": 7.9225e-05, "loss": 0.0799, "step": 3170 }, { "grad_norm": 0.7431588768959045, "learning_rate": 7.9475e-05, "loss": 0.1109, "step": 3180 }, { "grad_norm": 0.6603161692619324, "learning_rate": 7.9725e-05, "loss": 0.1054, "step": 3190 }, { "grad_norm": 0.7784467339515686, "learning_rate": 7.9975e-05, "loss": 0.1023, "step": 3200 }, { "grad_norm": 0.7415934801101685, "learning_rate": 8.022500000000001e-05, "loss": 0.0832, "step": 3210 }, { "grad_norm": 1.0568358898162842, "learning_rate": 8.0475e-05, "loss": 0.1227, "step": 3220 }, { "grad_norm": 0.6925896406173706, "learning_rate": 8.072500000000001e-05, "loss": 0.1075, "step": 3230 }, { "grad_norm": 0.45894700288772583, "learning_rate": 8.0975e-05, "loss": 0.083, "step": 3240 }, { "grad_norm": 0.6919277906417847, "learning_rate": 8.122500000000001e-05, "loss": 0.1128, "step": 3250 }, { "grad_norm": 0.8741796016693115, "learning_rate": 8.1475e-05, "loss": 0.1099, "step": 3260 }, { "grad_norm": 0.786183774471283, "learning_rate": 8.172500000000001e-05, "loss": 0.1023, "step": 3270 }, { "grad_norm": 0.8755525946617126, "learning_rate": 8.1975e-05, "loss": 0.088, "step": 3280 }, { "grad_norm": 0.7256410121917725, "learning_rate": 8.2225e-05, "loss": 0.0947, "step": 3290 }, { "grad_norm": 0.7042533755302429, "learning_rate": 8.2475e-05, "loss": 0.0874, "step": 3300 }, { "grad_norm": 1.0808988809585571, "learning_rate": 8.2725e-05, "loss": 0.1065, "step": 3310 }, { "grad_norm": 0.6708264350891113, "learning_rate": 8.2975e-05, "loss": 0.1181, "step": 3320 }, { "grad_norm": 0.6069616675376892, "learning_rate": 8.3225e-05, "loss": 0.0997, "step": 3330 }, { "grad_norm": 1.1354366540908813, "learning_rate": 8.347500000000001e-05, "loss": 0.1042, "step": 3340 }, { "grad_norm": 0.6768868565559387, "learning_rate": 8.3725e-05, "loss": 0.0974, "step": 3350 }, { "grad_norm": 0.9865342378616333, "learning_rate": 8.397500000000001e-05, "loss": 0.0909, "step": 3360 }, { "grad_norm": 0.7558157444000244, "learning_rate": 8.422500000000001e-05, "loss": 0.1007, "step": 3370 }, { "grad_norm": 0.8284071087837219, "learning_rate": 8.447500000000001e-05, "loss": 0.1547, "step": 3380 }, { "grad_norm": 0.8602030277252197, "learning_rate": 8.4725e-05, "loss": 0.1011, "step": 3390 }, { "grad_norm": 0.8202283978462219, "learning_rate": 8.4975e-05, "loss": 0.1054, "step": 3400 }, { "grad_norm": 1.1545939445495605, "learning_rate": 8.5225e-05, "loss": 0.1105, "step": 3410 }, { "grad_norm": 0.806383490562439, "learning_rate": 8.5475e-05, "loss": 0.0861, "step": 3420 }, { "grad_norm": 0.7937048077583313, "learning_rate": 8.5725e-05, "loss": 0.1104, "step": 3430 }, { "grad_norm": 0.8773503303527832, "learning_rate": 8.5975e-05, "loss": 0.1183, "step": 3440 }, { "grad_norm": 1.269039511680603, "learning_rate": 8.6225e-05, "loss": 0.1076, "step": 3450 }, { "grad_norm": 0.8712294697761536, "learning_rate": 8.6475e-05, "loss": 0.0924, "step": 3460 }, { "grad_norm": 0.7702007293701172, "learning_rate": 8.672500000000001e-05, "loss": 0.123, "step": 3470 }, { "grad_norm": 0.7717063426971436, "learning_rate": 8.6975e-05, "loss": 0.1153, "step": 3480 }, { "grad_norm": 0.7439447641372681, "learning_rate": 8.7225e-05, "loss": 0.119, "step": 3490 }, { "grad_norm": 0.8528266549110413, "learning_rate": 8.747500000000001e-05, "loss": 0.0876, "step": 3500 }, { "grad_norm": 0.9483062028884888, "learning_rate": 8.7725e-05, "loss": 0.0979, "step": 3510 }, { "grad_norm": 0.7391209602355957, "learning_rate": 8.797500000000001e-05, "loss": 0.0939, "step": 3520 }, { "grad_norm": 0.8769018650054932, "learning_rate": 8.8225e-05, "loss": 0.097, "step": 3530 }, { "grad_norm": 0.8593559861183167, "learning_rate": 8.847500000000001e-05, "loss": 0.1025, "step": 3540 }, { "grad_norm": 0.7571817636489868, "learning_rate": 8.8725e-05, "loss": 0.1021, "step": 3550 }, { "grad_norm": 0.6437474489212036, "learning_rate": 8.897500000000001e-05, "loss": 0.0818, "step": 3560 }, { "grad_norm": 0.9197229146957397, "learning_rate": 8.9225e-05, "loss": 0.114, "step": 3570 }, { "grad_norm": 0.8331531882286072, "learning_rate": 8.9475e-05, "loss": 0.1161, "step": 3580 }, { "grad_norm": 0.6494369506835938, "learning_rate": 8.9725e-05, "loss": 0.1121, "step": 3590 }, { "grad_norm": 0.615815281867981, "learning_rate": 8.9975e-05, "loss": 0.0914, "step": 3600 }, { "grad_norm": 0.694008469581604, "learning_rate": 9.0225e-05, "loss": 0.0935, "step": 3610 }, { "grad_norm": 0.6946540474891663, "learning_rate": 9.0475e-05, "loss": 0.0961, "step": 3620 }, { "grad_norm": 0.5122930407524109, "learning_rate": 9.072500000000001e-05, "loss": 0.0826, "step": 3630 }, { "grad_norm": 0.7755635380744934, "learning_rate": 9.0975e-05, "loss": 0.0902, "step": 3640 }, { "grad_norm": 0.7140825986862183, "learning_rate": 9.122500000000001e-05, "loss": 0.0938, "step": 3650 }, { "grad_norm": 0.7029992341995239, "learning_rate": 9.1475e-05, "loss": 0.1004, "step": 3660 }, { "grad_norm": 0.9221705794334412, "learning_rate": 9.172500000000001e-05, "loss": 0.1239, "step": 3670 }, { "grad_norm": 0.6815193891525269, "learning_rate": 9.1975e-05, "loss": 0.0967, "step": 3680 }, { "grad_norm": 0.8873224258422852, "learning_rate": 9.2225e-05, "loss": 0.1112, "step": 3690 }, { "grad_norm": 0.6958267092704773, "learning_rate": 9.2475e-05, "loss": 0.1043, "step": 3700 }, { "grad_norm": 0.8579167127609253, "learning_rate": 9.2725e-05, "loss": 0.1056, "step": 3710 }, { "grad_norm": 0.6995607614517212, "learning_rate": 9.2975e-05, "loss": 0.1042, "step": 3720 }, { "grad_norm": 0.6878102421760559, "learning_rate": 9.3225e-05, "loss": 0.096, "step": 3730 }, { "grad_norm": 0.6405667066574097, "learning_rate": 9.3475e-05, "loss": 0.0974, "step": 3740 }, { "grad_norm": 0.6888028979301453, "learning_rate": 9.3725e-05, "loss": 0.1077, "step": 3750 }, { "grad_norm": 0.7258442044258118, "learning_rate": 9.397500000000001e-05, "loss": 0.0875, "step": 3760 }, { "grad_norm": 0.7302910089492798, "learning_rate": 9.422500000000001e-05, "loss": 0.101, "step": 3770 }, { "grad_norm": 0.7250649333000183, "learning_rate": 9.4475e-05, "loss": 0.097, "step": 3780 }, { "grad_norm": 0.7425843477249146, "learning_rate": 9.472500000000001e-05, "loss": 0.0988, "step": 3790 }, { "grad_norm": 0.6457201838493347, "learning_rate": 9.4975e-05, "loss": 0.1249, "step": 3800 }, { "grad_norm": 0.617525041103363, "learning_rate": 9.522500000000001e-05, "loss": 0.0848, "step": 3810 }, { "grad_norm": 0.7148045301437378, "learning_rate": 9.5475e-05, "loss": 0.1033, "step": 3820 }, { "grad_norm": 0.6809467673301697, "learning_rate": 9.572500000000001e-05, "loss": 0.0849, "step": 3830 }, { "grad_norm": 0.684630811214447, "learning_rate": 9.5975e-05, "loss": 0.0906, "step": 3840 }, { "grad_norm": 0.8629801869392395, "learning_rate": 9.622500000000001e-05, "loss": 0.1154, "step": 3850 }, { "grad_norm": 0.8409157395362854, "learning_rate": 9.6475e-05, "loss": 0.0863, "step": 3860 }, { "grad_norm": 1.1738810539245605, "learning_rate": 9.6725e-05, "loss": 0.0853, "step": 3870 }, { "grad_norm": 0.5484836101531982, "learning_rate": 9.6975e-05, "loss": 0.0928, "step": 3880 }, { "grad_norm": 0.8082861304283142, "learning_rate": 9.7225e-05, "loss": 0.0961, "step": 3890 }, { "grad_norm": 0.8948726654052734, "learning_rate": 9.747500000000001e-05, "loss": 0.0975, "step": 3900 }, { "grad_norm": 0.9523621201515198, "learning_rate": 9.7725e-05, "loss": 0.1037, "step": 3910 }, { "grad_norm": 0.912493884563446, "learning_rate": 9.797500000000001e-05, "loss": 0.112, "step": 3920 }, { "grad_norm": 0.5441131591796875, "learning_rate": 9.8225e-05, "loss": 0.0875, "step": 3930 }, { "grad_norm": 0.7441449761390686, "learning_rate": 9.847500000000001e-05, "loss": 0.1103, "step": 3940 }, { "grad_norm": 0.7273918390274048, "learning_rate": 9.8725e-05, "loss": 0.0952, "step": 3950 }, { "grad_norm": 0.6768333315849304, "learning_rate": 9.897500000000001e-05, "loss": 0.1071, "step": 3960 }, { "grad_norm": 0.7240888476371765, "learning_rate": 9.9225e-05, "loss": 0.0976, "step": 3970 }, { "grad_norm": 0.7176719307899475, "learning_rate": 9.9475e-05, "loss": 0.1056, "step": 3980 }, { "grad_norm": 0.9587885141372681, "learning_rate": 9.9725e-05, "loss": 0.1037, "step": 3990 }, { "grad_norm": 0.816663384437561, "learning_rate": 9.9975e-05, "loss": 0.0957, "step": 4000 }, { "grad_norm": 0.80731201171875, "learning_rate": 9.999999653982884e-05, "loss": 0.098, "step": 4010 }, { "grad_norm": 0.5805341005325317, "learning_rate": 9.999998457874392e-05, "loss": 0.1007, "step": 4020 }, { "grad_norm": 0.8491482734680176, "learning_rate": 9.999996407402913e-05, "loss": 0.1, "step": 4030 }, { "grad_norm": 0.7060011625289917, "learning_rate": 9.999993502568801e-05, "loss": 0.0723, "step": 4040 }, { "grad_norm": 0.6693379282951355, "learning_rate": 9.999989743372548e-05, "loss": 0.0949, "step": 4050 }, { "grad_norm": 0.6909465789794922, "learning_rate": 9.999985129814798e-05, "loss": 0.0897, "step": 4060 }, { "grad_norm": 0.7055634260177612, "learning_rate": 9.99997966189634e-05, "loss": 0.0863, "step": 4070 }, { "grad_norm": 0.649992823600769, "learning_rate": 9.999973339618107e-05, "loss": 0.1071, "step": 4080 }, { "grad_norm": 0.7644646763801575, "learning_rate": 9.999966162981179e-05, "loss": 0.1042, "step": 4090 }, { "grad_norm": 0.8050379157066345, "learning_rate": 9.999958131986784e-05, "loss": 0.0923, "step": 4100 }, { "grad_norm": 0.5830329656600952, "learning_rate": 9.999949246636293e-05, "loss": 0.0981, "step": 4110 }, { "grad_norm": 0.7326309680938721, "learning_rate": 9.999939506931224e-05, "loss": 0.1071, "step": 4120 }, { "grad_norm": 0.7302266359329224, "learning_rate": 9.999928912873243e-05, "loss": 0.0966, "step": 4130 }, { "grad_norm": 0.8372926712036133, "learning_rate": 9.999917464464159e-05, "loss": 0.1088, "step": 4140 }, { "grad_norm": 0.6294270753860474, "learning_rate": 9.999905161705929e-05, "loss": 0.1097, "step": 4150 }, { "grad_norm": 0.6173607110977173, "learning_rate": 9.999892004600653e-05, "loss": 0.0991, "step": 4160 }, { "grad_norm": 0.7258467674255371, "learning_rate": 9.999877993150581e-05, "loss": 0.0894, "step": 4170 }, { "grad_norm": 0.5222402215003967, "learning_rate": 9.999863127358108e-05, "loss": 0.0812, "step": 4180 }, { "grad_norm": 0.6972135901451111, "learning_rate": 9.999847407225773e-05, "loss": 0.0833, "step": 4190 }, { "grad_norm": 0.7253326773643494, "learning_rate": 9.999830832756262e-05, "loss": 0.09, "step": 4200 }, { "grad_norm": 0.7528601288795471, "learning_rate": 9.999813403952407e-05, "loss": 0.0941, "step": 4210 }, { "grad_norm": 0.6715784072875977, "learning_rate": 9.999795120817187e-05, "loss": 0.1046, "step": 4220 }, { "grad_norm": 0.8208842277526855, "learning_rate": 9.999775983353725e-05, "loss": 0.1254, "step": 4230 }, { "grad_norm": 0.7496628165245056, "learning_rate": 9.999755991565292e-05, "loss": 0.1127, "step": 4240 }, { "grad_norm": 0.6591829657554626, "learning_rate": 9.999735145455303e-05, "loss": 0.1038, "step": 4250 }, { "grad_norm": 0.7059779167175293, "learning_rate": 9.99971344502732e-05, "loss": 0.1111, "step": 4260 }, { "grad_norm": 0.5739170908927917, "learning_rate": 9.999690890285053e-05, "loss": 0.1012, "step": 4270 }, { "grad_norm": 0.7312772870063782, "learning_rate": 9.999667481232356e-05, "loss": 0.082, "step": 4280 }, { "grad_norm": 0.6249011158943176, "learning_rate": 9.999643217873225e-05, "loss": 0.0901, "step": 4290 }, { "grad_norm": 0.6558547019958496, "learning_rate": 9.999618100211809e-05, "loss": 0.0802, "step": 4300 }, { "grad_norm": 0.46935388445854187, "learning_rate": 9.999592128252402e-05, "loss": 0.0927, "step": 4310 }, { "grad_norm": 0.8966103792190552, "learning_rate": 9.999565301999437e-05, "loss": 0.1252, "step": 4320 }, { "grad_norm": 0.46997204422950745, "learning_rate": 9.999537621457502e-05, "loss": 0.0909, "step": 4330 }, { "grad_norm": 0.5478244423866272, "learning_rate": 9.999509086631323e-05, "loss": 0.0755, "step": 4340 }, { "grad_norm": 0.6294610500335693, "learning_rate": 9.99947969752578e-05, "loss": 0.0781, "step": 4350 }, { "grad_norm": 0.6800083518028259, "learning_rate": 9.999449454145891e-05, "loss": 0.0867, "step": 4360 }, { "grad_norm": 0.7026942372322083, "learning_rate": 9.999418356496827e-05, "loss": 0.0903, "step": 4370 }, { "grad_norm": 0.4981493353843689, "learning_rate": 9.999386404583899e-05, "loss": 0.0948, "step": 4380 }, { "grad_norm": 0.9832645058631897, "learning_rate": 9.999353598412568e-05, "loss": 0.0849, "step": 4390 }, { "grad_norm": 0.6194189786911011, "learning_rate": 9.999319937988442e-05, "loss": 0.111, "step": 4400 }, { "grad_norm": 0.5876731276512146, "learning_rate": 9.999285423317268e-05, "loss": 0.1077, "step": 4410 }, { "grad_norm": 0.5327100157737732, "learning_rate": 9.999250054404947e-05, "loss": 0.1265, "step": 4420 }, { "grad_norm": 0.6086254715919495, "learning_rate": 9.99921383125752e-05, "loss": 0.0995, "step": 4430 }, { "grad_norm": 0.6422713398933411, "learning_rate": 9.99917675388118e-05, "loss": 0.0969, "step": 4440 }, { "grad_norm": 0.7151811122894287, "learning_rate": 9.99913882228226e-05, "loss": 0.1047, "step": 4450 }, { "grad_norm": 0.601818323135376, "learning_rate": 9.999100036467242e-05, "loss": 0.0885, "step": 4460 }, { "grad_norm": 0.7426432967185974, "learning_rate": 9.999060396442753e-05, "loss": 0.0848, "step": 4470 }, { "grad_norm": 0.7140601277351379, "learning_rate": 9.999019902215566e-05, "loss": 0.1036, "step": 4480 }, { "grad_norm": 0.6685841679573059, "learning_rate": 9.998978553792602e-05, "loss": 0.0925, "step": 4490 }, { "grad_norm": 0.793701171875, "learning_rate": 9.998936351180926e-05, "loss": 0.0905, "step": 4500 }, { "grad_norm": 0.7201404571533203, "learning_rate": 9.998893294387747e-05, "loss": 0.1229, "step": 4510 }, { "grad_norm": 0.7527884244918823, "learning_rate": 9.998849383420426e-05, "loss": 0.1005, "step": 4520 }, { "grad_norm": 0.7701876759529114, "learning_rate": 9.998804618286465e-05, "loss": 0.1039, "step": 4530 }, { "grad_norm": 0.8960968852043152, "learning_rate": 9.99875899899351e-05, "loss": 0.1231, "step": 4540 }, { "grad_norm": 0.7024566531181335, "learning_rate": 9.99871252554936e-05, "loss": 0.1153, "step": 4550 }, { "grad_norm": 0.7817474603652954, "learning_rate": 9.998665197961955e-05, "loss": 0.094, "step": 4560 }, { "grad_norm": 0.6959978938102722, "learning_rate": 9.998617016239379e-05, "loss": 0.1195, "step": 4570 }, { "grad_norm": 0.974354088306427, "learning_rate": 9.998567980389869e-05, "loss": 0.1043, "step": 4580 }, { "grad_norm": 0.64463210105896, "learning_rate": 9.998518090421802e-05, "loss": 0.0908, "step": 4590 }, { "grad_norm": 0.7417001724243164, "learning_rate": 9.998467346343703e-05, "loss": 0.0917, "step": 4600 }, { "grad_norm": 0.6521580219268799, "learning_rate": 9.998415748164243e-05, "loss": 0.091, "step": 4610 }, { "grad_norm": 0.663733184337616, "learning_rate": 9.998363295892238e-05, "loss": 0.0873, "step": 4620 }, { "grad_norm": 0.7824060916900635, "learning_rate": 9.998309989536652e-05, "loss": 0.092, "step": 4630 }, { "grad_norm": 0.5672284364700317, "learning_rate": 9.998255829106593e-05, "loss": 0.1027, "step": 4640 }, { "grad_norm": 0.6259315013885498, "learning_rate": 9.998200814611316e-05, "loss": 0.0877, "step": 4650 }, { "grad_norm": 0.7068696022033691, "learning_rate": 9.998144946060219e-05, "loss": 0.0975, "step": 4660 }, { "grad_norm": 0.964932382106781, "learning_rate": 9.998088223462852e-05, "loss": 0.0907, "step": 4670 }, { "grad_norm": 0.945661187171936, "learning_rate": 9.998030646828905e-05, "loss": 0.0882, "step": 4680 }, { "grad_norm": 0.5922400951385498, "learning_rate": 9.997972216168217e-05, "loss": 0.0971, "step": 4690 }, { "grad_norm": 0.648963987827301, "learning_rate": 9.997912931490771e-05, "loss": 0.0914, "step": 4700 }, { "grad_norm": 0.7344776391983032, "learning_rate": 9.9978527928067e-05, "loss": 0.0929, "step": 4710 }, { "grad_norm": 0.6816878914833069, "learning_rate": 9.997791800126277e-05, "loss": 0.086, "step": 4720 }, { "grad_norm": 0.4595393240451813, "learning_rate": 9.997729953459927e-05, "loss": 0.1102, "step": 4730 }, { "grad_norm": 0.5680307149887085, "learning_rate": 9.997667252818214e-05, "loss": 0.0969, "step": 4740 }, { "grad_norm": 0.6610550284385681, "learning_rate": 9.997603698211855e-05, "loss": 0.083, "step": 4750 }, { "grad_norm": 0.8365136384963989, "learning_rate": 9.99753928965171e-05, "loss": 0.0976, "step": 4760 }, { "grad_norm": 0.8221967816352844, "learning_rate": 9.997474027148781e-05, "loss": 0.0951, "step": 4770 }, { "grad_norm": 0.7737025022506714, "learning_rate": 9.997407910714223e-05, "loss": 0.1061, "step": 4780 }, { "grad_norm": 0.5566369295120239, "learning_rate": 9.997340940359332e-05, "loss": 0.0784, "step": 4790 }, { "grad_norm": 0.8592110276222229, "learning_rate": 9.997273116095552e-05, "loss": 0.0855, "step": 4800 }, { "grad_norm": 0.6218822598457336, "learning_rate": 9.997204437934473e-05, "loss": 0.0865, "step": 4810 }, { "grad_norm": 0.7261716723442078, "learning_rate": 9.997134905887829e-05, "loss": 0.0914, "step": 4820 }, { "grad_norm": 0.5298239588737488, "learning_rate": 9.997064519967501e-05, "loss": 0.0916, "step": 4830 }, { "grad_norm": 0.5580573678016663, "learning_rate": 9.996993280185517e-05, "loss": 0.0911, "step": 4840 }, { "grad_norm": 0.799647331237793, "learning_rate": 9.99692118655405e-05, "loss": 0.0865, "step": 4850 }, { "grad_norm": 0.8499962091445923, "learning_rate": 9.996848239085417e-05, "loss": 0.0818, "step": 4860 }, { "grad_norm": 0.568547785282135, "learning_rate": 9.996774437792085e-05, "loss": 0.0762, "step": 4870 }, { "grad_norm": 0.631234884262085, "learning_rate": 9.996699782686664e-05, "loss": 0.0773, "step": 4880 }, { "grad_norm": 0.7187631130218506, "learning_rate": 9.996624273781909e-05, "loss": 0.1011, "step": 4890 }, { "grad_norm": 0.6149592995643616, "learning_rate": 9.996547911090725e-05, "loss": 0.0782, "step": 4900 }, { "grad_norm": 0.6341006755828857, "learning_rate": 9.996470694626157e-05, "loss": 0.0849, "step": 4910 }, { "grad_norm": 0.567354142665863, "learning_rate": 9.996392624401403e-05, "loss": 0.0844, "step": 4920 }, { "grad_norm": 0.7183308601379395, "learning_rate": 9.996313700429801e-05, "loss": 0.0802, "step": 4930 }, { "grad_norm": 0.5529053211212158, "learning_rate": 9.996233922724836e-05, "loss": 0.0746, "step": 4940 }, { "grad_norm": 0.5053432583808899, "learning_rate": 9.996153291300141e-05, "loss": 0.1029, "step": 4950 }, { "grad_norm": 0.4523994028568268, "learning_rate": 9.996071806169494e-05, "loss": 0.0756, "step": 4960 }, { "grad_norm": 0.5395421981811523, "learning_rate": 9.995989467346817e-05, "loss": 0.0939, "step": 4970 }, { "grad_norm": 0.6202993988990784, "learning_rate": 9.995906274846183e-05, "loss": 0.0862, "step": 4980 }, { "grad_norm": 0.9989957213401794, "learning_rate": 9.995822228681803e-05, "loss": 0.0923, "step": 4990 }, { "grad_norm": 0.57069331407547, "learning_rate": 9.99573732886804e-05, "loss": 0.0939, "step": 5000 }, { "grad_norm": 0.757864773273468, "learning_rate": 9.995651575419402e-05, "loss": 0.0926, "step": 5010 }, { "grad_norm": 0.9190270304679871, "learning_rate": 9.995564968350541e-05, "loss": 0.0918, "step": 5020 }, { "grad_norm": 0.6685031652450562, "learning_rate": 9.995477507676256e-05, "loss": 0.0766, "step": 5030 }, { "grad_norm": 0.5065507888793945, "learning_rate": 9.995389193411493e-05, "loss": 0.0831, "step": 5040 }, { "grad_norm": 0.591122567653656, "learning_rate": 9.995300025571339e-05, "loss": 0.0927, "step": 5050 }, { "grad_norm": 0.5673381686210632, "learning_rate": 9.995210004171034e-05, "loss": 0.0753, "step": 5060 }, { "grad_norm": 0.5630785822868347, "learning_rate": 9.995119129225956e-05, "loss": 0.1038, "step": 5070 }, { "grad_norm": 0.6515262722969055, "learning_rate": 9.995027400751637e-05, "loss": 0.0989, "step": 5080 }, { "grad_norm": 0.6229056715965271, "learning_rate": 9.994934818763751e-05, "loss": 0.0849, "step": 5090 }, { "grad_norm": 0.9056984186172485, "learning_rate": 9.994841383278115e-05, "loss": 0.0823, "step": 5100 }, { "grad_norm": 0.6804705858230591, "learning_rate": 9.994747094310695e-05, "loss": 0.0893, "step": 5110 }, { "grad_norm": 0.8070368766784668, "learning_rate": 9.994651951877604e-05, "loss": 0.1062, "step": 5120 }, { "grad_norm": 0.6486148238182068, "learning_rate": 9.994555955995099e-05, "loss": 0.103, "step": 5130 }, { "grad_norm": 0.5857247114181519, "learning_rate": 9.994459106679581e-05, "loss": 0.0933, "step": 5140 }, { "grad_norm": 0.5793846249580383, "learning_rate": 9.994361403947603e-05, "loss": 0.0779, "step": 5150 }, { "grad_norm": 0.6521289348602295, "learning_rate": 9.994262847815854e-05, "loss": 0.1147, "step": 5160 }, { "grad_norm": 0.7272570729255676, "learning_rate": 9.99416343830118e-05, "loss": 0.0966, "step": 5170 }, { "grad_norm": 0.6852249503135681, "learning_rate": 9.994063175420565e-05, "loss": 0.0838, "step": 5180 }, { "grad_norm": 0.7334885597229004, "learning_rate": 9.99396205919114e-05, "loss": 0.1206, "step": 5190 }, { "grad_norm": 0.6187967658042908, "learning_rate": 9.993860089630185e-05, "loss": 0.0795, "step": 5200 }, { "grad_norm": 0.5253296494483948, "learning_rate": 9.993757266755123e-05, "loss": 0.0855, "step": 5210 }, { "grad_norm": 0.6467638611793518, "learning_rate": 9.993653590583522e-05, "loss": 0.0849, "step": 5220 }, { "grad_norm": 0.49420928955078125, "learning_rate": 9.993549061133102e-05, "loss": 0.0764, "step": 5230 }, { "grad_norm": 0.6136224865913391, "learning_rate": 9.993443678421719e-05, "loss": 0.0813, "step": 5240 }, { "grad_norm": 0.6506721377372742, "learning_rate": 9.993337442467384e-05, "loss": 0.09, "step": 5250 }, { "grad_norm": 0.744143009185791, "learning_rate": 9.993230353288248e-05, "loss": 0.0747, "step": 5260 }, { "grad_norm": 0.6281589269638062, "learning_rate": 9.993122410902608e-05, "loss": 0.0793, "step": 5270 }, { "grad_norm": 0.5616821050643921, "learning_rate": 9.993013615328912e-05, "loss": 0.0917, "step": 5280 }, { "grad_norm": 0.6345070600509644, "learning_rate": 9.992903966585747e-05, "loss": 0.0932, "step": 5290 }, { "grad_norm": 0.54500412940979, "learning_rate": 9.992793464691852e-05, "loss": 0.0749, "step": 5300 }, { "grad_norm": 0.6992676854133606, "learning_rate": 9.992682109666105e-05, "loss": 0.0884, "step": 5310 }, { "grad_norm": 0.5907784104347229, "learning_rate": 9.992569901527538e-05, "loss": 0.0858, "step": 5320 }, { "grad_norm": 0.6044657826423645, "learning_rate": 9.99245684029532e-05, "loss": 0.095, "step": 5330 }, { "grad_norm": 0.7255027890205383, "learning_rate": 9.992342925988774e-05, "loss": 0.0861, "step": 5340 }, { "grad_norm": 0.6217452883720398, "learning_rate": 9.992228158627361e-05, "loss": 0.0787, "step": 5350 }, { "grad_norm": 0.6405369639396667, "learning_rate": 9.992112538230693e-05, "loss": 0.0809, "step": 5360 }, { "grad_norm": 0.5752140283584595, "learning_rate": 9.991996064818527e-05, "loss": 0.0793, "step": 5370 }, { "grad_norm": 0.6843039393424988, "learning_rate": 9.991878738410768e-05, "loss": 0.075, "step": 5380 }, { "grad_norm": 0.8028432726860046, "learning_rate": 9.991760559027457e-05, "loss": 0.0781, "step": 5390 }, { "grad_norm": 0.44660240411758423, "learning_rate": 9.991641526688793e-05, "loss": 0.1066, "step": 5400 }, { "grad_norm": 0.6244422197341919, "learning_rate": 9.991521641415113e-05, "loss": 0.0848, "step": 5410 }, { "grad_norm": 0.5593048334121704, "learning_rate": 9.991400903226904e-05, "loss": 0.0755, "step": 5420 }, { "grad_norm": 0.8304235339164734, "learning_rate": 9.991279312144794e-05, "loss": 0.0817, "step": 5430 }, { "grad_norm": 0.7673569321632385, "learning_rate": 9.991156868189564e-05, "loss": 0.0937, "step": 5440 }, { "grad_norm": 0.594325065612793, "learning_rate": 9.991033571382131e-05, "loss": 0.0636, "step": 5450 }, { "grad_norm": 0.6292912364006042, "learning_rate": 9.990909421743569e-05, "loss": 0.07, "step": 5460 }, { "grad_norm": 0.7371444702148438, "learning_rate": 9.990784419295085e-05, "loss": 0.0741, "step": 5470 }, { "grad_norm": 0.5848631262779236, "learning_rate": 9.990658564058044e-05, "loss": 0.0739, "step": 5480 }, { "grad_norm": 0.894788920879364, "learning_rate": 9.990531856053948e-05, "loss": 0.0834, "step": 5490 }, { "grad_norm": 0.5574281811714172, "learning_rate": 9.99040429530445e-05, "loss": 0.093, "step": 5500 }, { "grad_norm": 0.6771383881568909, "learning_rate": 9.990275881831346e-05, "loss": 0.0891, "step": 5510 }, { "grad_norm": 0.5071075558662415, "learning_rate": 9.990146615656577e-05, "loss": 0.0798, "step": 5520 }, { "grad_norm": 0.47177854180336, "learning_rate": 9.990016496802233e-05, "loss": 0.0782, "step": 5530 }, { "grad_norm": 0.5009283423423767, "learning_rate": 9.989885525290548e-05, "loss": 0.0665, "step": 5540 }, { "grad_norm": 0.6071903109550476, "learning_rate": 9.989753701143897e-05, "loss": 0.0729, "step": 5550 }, { "grad_norm": 0.5023123025894165, "learning_rate": 9.989621024384812e-05, "loss": 0.0734, "step": 5560 }, { "grad_norm": 0.7049488425254822, "learning_rate": 9.989487495035959e-05, "loss": 0.0808, "step": 5570 }, { "grad_norm": 0.8489465117454529, "learning_rate": 9.989353113120156e-05, "loss": 0.0966, "step": 5580 }, { "grad_norm": 0.7115804553031921, "learning_rate": 9.989217878660366e-05, "loss": 0.0781, "step": 5590 }, { "grad_norm": 0.671501874923706, "learning_rate": 9.989081791679695e-05, "loss": 0.0894, "step": 5600 }, { "grad_norm": 0.4873778223991394, "learning_rate": 9.988944852201397e-05, "loss": 0.0769, "step": 5610 }, { "grad_norm": 0.5432367324829102, "learning_rate": 9.988807060248873e-05, "loss": 0.0733, "step": 5620 }, { "grad_norm": 0.5530681610107422, "learning_rate": 9.988668415845665e-05, "loss": 0.0675, "step": 5630 }, { "grad_norm": 0.749701976776123, "learning_rate": 9.988528919015466e-05, "loss": 0.0747, "step": 5640 }, { "grad_norm": 0.5856157541275024, "learning_rate": 9.988388569782112e-05, "loss": 0.1041, "step": 5650 }, { "grad_norm": 0.6665031909942627, "learning_rate": 9.988247368169583e-05, "loss": 0.0775, "step": 5660 }, { "grad_norm": 0.631603479385376, "learning_rate": 9.988105314202007e-05, "loss": 0.075, "step": 5670 }, { "grad_norm": 0.6231602430343628, "learning_rate": 9.987962407903659e-05, "loss": 0.0887, "step": 5680 }, { "grad_norm": 0.7892677783966064, "learning_rate": 9.987818649298957e-05, "loss": 0.0801, "step": 5690 }, { "grad_norm": 0.4515496492385864, "learning_rate": 9.987674038412465e-05, "loss": 0.0798, "step": 5700 }, { "grad_norm": 0.5550165176391602, "learning_rate": 9.987528575268891e-05, "loss": 0.0789, "step": 5710 }, { "grad_norm": 0.5258063673973083, "learning_rate": 9.987382259893095e-05, "loss": 0.0722, "step": 5720 }, { "grad_norm": 0.5668755769729614, "learning_rate": 9.987235092310074e-05, "loss": 0.0719, "step": 5730 }, { "grad_norm": 0.5692359805107117, "learning_rate": 9.987087072544978e-05, "loss": 0.0877, "step": 5740 }, { "grad_norm": 0.6510366201400757, "learning_rate": 9.9869382006231e-05, "loss": 0.0776, "step": 5750 }, { "grad_norm": 0.5913699269294739, "learning_rate": 9.986788476569875e-05, "loss": 0.0864, "step": 5760 }, { "grad_norm": 0.5619310736656189, "learning_rate": 9.986637900410887e-05, "loss": 0.0786, "step": 5770 }, { "grad_norm": 0.5320875644683838, "learning_rate": 9.986486472171869e-05, "loss": 0.0958, "step": 5780 }, { "grad_norm": 0.6404861807823181, "learning_rate": 9.986334191878692e-05, "loss": 0.0747, "step": 5790 }, { "grad_norm": 0.630691647529602, "learning_rate": 9.986181059557378e-05, "loss": 0.0907, "step": 5800 }, { "grad_norm": 0.5591872930526733, "learning_rate": 9.986027075234094e-05, "loss": 0.0737, "step": 5810 }, { "grad_norm": 0.5618623495101929, "learning_rate": 9.985872238935152e-05, "loss": 0.0859, "step": 5820 }, { "grad_norm": 0.6224364042282104, "learning_rate": 9.985716550687008e-05, "loss": 0.0876, "step": 5830 }, { "grad_norm": 0.48722898960113525, "learning_rate": 9.985560010516264e-05, "loss": 0.0751, "step": 5840 }, { "grad_norm": 0.5383521318435669, "learning_rate": 9.985402618449668e-05, "loss": 0.0684, "step": 5850 }, { "grad_norm": 0.6486960649490356, "learning_rate": 9.985244374514118e-05, "loss": 0.0924, "step": 5860 }, { "grad_norm": 0.4553612470626831, "learning_rate": 9.985085278736651e-05, "loss": 0.0817, "step": 5870 }, { "grad_norm": 0.5198431611061096, "learning_rate": 9.984925331144452e-05, "loss": 0.0794, "step": 5880 }, { "grad_norm": 0.5975961089134216, "learning_rate": 9.984764531764851e-05, "loss": 0.0848, "step": 5890 }, { "grad_norm": 0.6732950210571289, "learning_rate": 9.984602880625326e-05, "loss": 0.0936, "step": 5900 }, { "grad_norm": 0.5922431349754333, "learning_rate": 9.9844403777535e-05, "loss": 0.0851, "step": 5910 }, { "grad_norm": 0.7270139455795288, "learning_rate": 9.984277023177135e-05, "loss": 0.0754, "step": 5920 }, { "grad_norm": 0.5880727767944336, "learning_rate": 9.984112816924148e-05, "loss": 0.0749, "step": 5930 }, { "grad_norm": 0.47867685556411743, "learning_rate": 9.983947759022596e-05, "loss": 0.0806, "step": 5940 }, { "grad_norm": 0.5544881224632263, "learning_rate": 9.983781849500682e-05, "loss": 0.0757, "step": 5950 }, { "grad_norm": 0.8749607801437378, "learning_rate": 9.98361508838676e-05, "loss": 0.0783, "step": 5960 }, { "grad_norm": 0.6329994797706604, "learning_rate": 9.98344747570932e-05, "loss": 0.0798, "step": 5970 }, { "grad_norm": 0.6850922107696533, "learning_rate": 9.983279011497004e-05, "loss": 0.0849, "step": 5980 }, { "grad_norm": 0.9556413888931274, "learning_rate": 9.983109695778596e-05, "loss": 0.1041, "step": 5990 }, { "grad_norm": 0.5013958215713501, "learning_rate": 9.982939528583032e-05, "loss": 0.1102, "step": 6000 }, { "grad_norm": 0.5844435691833496, "learning_rate": 9.982768509939385e-05, "loss": 0.0864, "step": 6010 }, { "grad_norm": 0.6694456338882446, "learning_rate": 9.982596639876879e-05, "loss": 0.0832, "step": 6020 }, { "grad_norm": 0.542803943157196, "learning_rate": 9.982423918424881e-05, "loss": 0.0808, "step": 6030 }, { "grad_norm": 0.6814818382263184, "learning_rate": 9.982250345612908e-05, "loss": 0.0762, "step": 6040 }, { "grad_norm": 0.6827667355537415, "learning_rate": 9.982075921470611e-05, "loss": 0.0833, "step": 6050 }, { "grad_norm": 0.6212171316146851, "learning_rate": 9.981900646027802e-05, "loss": 0.0763, "step": 6060 }, { "grad_norm": 0.614452600479126, "learning_rate": 9.981724519314425e-05, "loss": 0.0722, "step": 6070 }, { "grad_norm": 0.5147626996040344, "learning_rate": 9.981547541360581e-05, "loss": 0.0831, "step": 6080 }, { "grad_norm": 0.7736936807632446, "learning_rate": 9.981369712196508e-05, "loss": 0.0876, "step": 6090 }, { "grad_norm": 0.6910366415977478, "learning_rate": 9.981191031852592e-05, "loss": 0.0967, "step": 6100 }, { "grad_norm": 0.5440165400505066, "learning_rate": 9.981011500359362e-05, "loss": 0.0743, "step": 6110 }, { "grad_norm": 0.45481231808662415, "learning_rate": 9.9808311177475e-05, "loss": 0.0802, "step": 6120 }, { "grad_norm": 0.6233107447624207, "learning_rate": 9.980649884047826e-05, "loss": 0.0659, "step": 6130 }, { "grad_norm": 0.6389815211296082, "learning_rate": 9.980467799291307e-05, "loss": 0.0696, "step": 6140 }, { "grad_norm": 0.5599873661994934, "learning_rate": 9.980284863509058e-05, "loss": 0.1037, "step": 6150 }, { "grad_norm": 0.6137582659721375, "learning_rate": 9.980101076732334e-05, "loss": 0.1062, "step": 6160 }, { "grad_norm": 0.5943885445594788, "learning_rate": 9.979916438992544e-05, "loss": 0.0857, "step": 6170 }, { "grad_norm": 0.5195236206054688, "learning_rate": 9.979730950321237e-05, "loss": 0.0708, "step": 6180 }, { "grad_norm": 0.523967444896698, "learning_rate": 9.979544610750104e-05, "loss": 0.0847, "step": 6190 }, { "grad_norm": 0.43815332651138306, "learning_rate": 9.97935742031099e-05, "loss": 0.0735, "step": 6200 }, { "grad_norm": 0.6826993227005005, "learning_rate": 9.979169379035878e-05, "loss": 0.0779, "step": 6210 }, { "grad_norm": 0.6485916376113892, "learning_rate": 9.978980486956899e-05, "loss": 0.0712, "step": 6220 }, { "grad_norm": 0.389155775308609, "learning_rate": 9.978790744106332e-05, "loss": 0.0721, "step": 6230 }, { "grad_norm": 0.6246785521507263, "learning_rate": 9.978600150516594e-05, "loss": 0.0791, "step": 6240 }, { "grad_norm": 0.5323804616928101, "learning_rate": 9.978408706220259e-05, "loss": 0.0784, "step": 6250 }, { "grad_norm": 0.4743943512439728, "learning_rate": 9.978216411250032e-05, "loss": 0.0682, "step": 6260 }, { "grad_norm": 0.5960530042648315, "learning_rate": 9.978023265638778e-05, "loss": 0.0779, "step": 6270 }, { "grad_norm": 0.5744068622589111, "learning_rate": 9.977829269419495e-05, "loss": 0.0822, "step": 6280 }, { "grad_norm": 0.5739364624023438, "learning_rate": 9.977634422625335e-05, "loss": 0.0784, "step": 6290 }, { "grad_norm": 0.5968267917633057, "learning_rate": 9.97743872528959e-05, "loss": 0.0811, "step": 6300 }, { "grad_norm": 0.621635913848877, "learning_rate": 9.9772421774457e-05, "loss": 0.0773, "step": 6310 }, { "grad_norm": 0.5558649897575378, "learning_rate": 9.977044779127252e-05, "loss": 0.0728, "step": 6320 }, { "grad_norm": 0.6649143099784851, "learning_rate": 9.976846530367971e-05, "loss": 0.0726, "step": 6330 }, { "grad_norm": 0.8420178890228271, "learning_rate": 9.976647431201735e-05, "loss": 0.0897, "step": 6340 }, { "grad_norm": 0.547081470489502, "learning_rate": 9.976447481662568e-05, "loss": 0.0718, "step": 6350 }, { "grad_norm": 0.5103899836540222, "learning_rate": 9.976246681784629e-05, "loss": 0.0839, "step": 6360 }, { "grad_norm": 0.4733942151069641, "learning_rate": 9.976045031602234e-05, "loss": 0.0662, "step": 6370 }, { "grad_norm": 0.5225886702537537, "learning_rate": 9.975842531149837e-05, "loss": 0.0793, "step": 6380 }, { "grad_norm": 0.7768915891647339, "learning_rate": 9.975639180462043e-05, "loss": 0.0841, "step": 6390 }, { "grad_norm": 0.6118035316467285, "learning_rate": 9.975434979573596e-05, "loss": 0.078, "step": 6400 }, { "grad_norm": 0.4436512887477875, "learning_rate": 9.97522992851939e-05, "loss": 0.0587, "step": 6410 }, { "grad_norm": 0.4987613558769226, "learning_rate": 9.975024027334461e-05, "loss": 0.0676, "step": 6420 }, { "grad_norm": 0.3984963297843933, "learning_rate": 9.974817276053993e-05, "loss": 0.0674, "step": 6430 }, { "grad_norm": 0.891409158706665, "learning_rate": 9.974609674713315e-05, "loss": 0.0779, "step": 6440 }, { "grad_norm": 0.6503852009773254, "learning_rate": 9.9744012233479e-05, "loss": 0.072, "step": 6450 }, { "grad_norm": 0.5002050995826721, "learning_rate": 9.974191921993366e-05, "loss": 0.0729, "step": 6460 }, { "grad_norm": 0.43562671542167664, "learning_rate": 9.973981770685474e-05, "loss": 0.0695, "step": 6470 }, { "grad_norm": 0.6380531191825867, "learning_rate": 9.97377076946014e-05, "loss": 0.0649, "step": 6480 }, { "grad_norm": 0.5945307612419128, "learning_rate": 9.973558918353412e-05, "loss": 0.0804, "step": 6490 }, { "grad_norm": 0.426464319229126, "learning_rate": 9.973346217401494e-05, "loss": 0.0676, "step": 6500 }, { "grad_norm": 0.5707445740699768, "learning_rate": 9.973132666640726e-05, "loss": 0.0612, "step": 6510 }, { "grad_norm": 0.5123850107192993, "learning_rate": 9.972918266107602e-05, "loss": 0.0679, "step": 6520 }, { "grad_norm": 0.8884283900260925, "learning_rate": 9.972703015838756e-05, "loss": 0.0723, "step": 6530 }, { "grad_norm": 0.5870687365531921, "learning_rate": 9.97248691587097e-05, "loss": 0.0775, "step": 6540 }, { "grad_norm": 0.6840682625770569, "learning_rate": 9.972269966241166e-05, "loss": 0.0694, "step": 6550 }, { "grad_norm": 0.5448303818702698, "learning_rate": 9.972052166986417e-05, "loss": 0.0781, "step": 6560 }, { "grad_norm": 0.6677625179290771, "learning_rate": 9.971833518143938e-05, "loss": 0.0778, "step": 6570 }, { "grad_norm": 0.6106738448143005, "learning_rate": 9.971614019751093e-05, "loss": 0.066, "step": 6580 }, { "grad_norm": 0.6357826590538025, "learning_rate": 9.971393671845383e-05, "loss": 0.0785, "step": 6590 }, { "grad_norm": 0.4150644540786743, "learning_rate": 9.971172474464464e-05, "loss": 0.0882, "step": 6600 }, { "grad_norm": 0.5151309370994568, "learning_rate": 9.97095042764613e-05, "loss": 0.0674, "step": 6610 }, { "grad_norm": 0.5647066235542297, "learning_rate": 9.970727531428324e-05, "loss": 0.0635, "step": 6620 }, { "grad_norm": 0.6519758701324463, "learning_rate": 9.970503785849132e-05, "loss": 0.0667, "step": 6630 }, { "grad_norm": 0.4606937766075134, "learning_rate": 9.970279190946788e-05, "loss": 0.0606, "step": 6640 }, { "grad_norm": 0.4479175806045532, "learning_rate": 9.970053746759667e-05, "loss": 0.0736, "step": 6650 }, { "grad_norm": 0.4184899926185608, "learning_rate": 9.969827453326292e-05, "loss": 0.0638, "step": 6660 }, { "grad_norm": 0.41636836528778076, "learning_rate": 9.969600310685332e-05, "loss": 0.0615, "step": 6670 }, { "grad_norm": 0.4649190306663513, "learning_rate": 9.969372318875596e-05, "loss": 0.0754, "step": 6680 }, { "grad_norm": 0.5239422917366028, "learning_rate": 9.969143477936043e-05, "loss": 0.0804, "step": 6690 }, { "grad_norm": 0.536268413066864, "learning_rate": 9.968913787905775e-05, "loss": 0.0771, "step": 6700 }, { "grad_norm": 0.6392862200737, "learning_rate": 9.968683248824045e-05, "loss": 0.0784, "step": 6710 }, { "grad_norm": 0.5592299103736877, "learning_rate": 9.968451860730238e-05, "loss": 0.0885, "step": 6720 }, { "grad_norm": 0.5224854946136475, "learning_rate": 9.968219623663896e-05, "loss": 0.0749, "step": 6730 }, { "grad_norm": 0.6589832901954651, "learning_rate": 9.967986537664702e-05, "loss": 0.1029, "step": 6740 }, { "grad_norm": 0.44783496856689453, "learning_rate": 9.967752602772483e-05, "loss": 0.0714, "step": 6750 }, { "grad_norm": 0.6542580723762512, "learning_rate": 9.967517819027212e-05, "loss": 0.0815, "step": 6760 }, { "grad_norm": 0.560693085193634, "learning_rate": 9.967282186469009e-05, "loss": 0.0629, "step": 6770 }, { "grad_norm": 0.5876162052154541, "learning_rate": 9.967045705138135e-05, "loss": 0.0783, "step": 6780 }, { "grad_norm": 0.5983959436416626, "learning_rate": 9.966808375074998e-05, "loss": 0.0681, "step": 6790 }, { "grad_norm": 0.5592385530471802, "learning_rate": 9.966570196320154e-05, "loss": 0.0744, "step": 6800 }, { "grad_norm": 0.5389854311943054, "learning_rate": 9.966331168914299e-05, "loss": 0.0689, "step": 6810 }, { "grad_norm": 0.5204718708992004, "learning_rate": 9.966091292898277e-05, "loss": 0.0788, "step": 6820 }, { "grad_norm": 0.5998125076293945, "learning_rate": 9.965850568313076e-05, "loss": 0.0754, "step": 6830 }, { "grad_norm": 0.5560948252677917, "learning_rate": 9.965608995199827e-05, "loss": 0.0795, "step": 6840 }, { "grad_norm": 1.018811583518982, "learning_rate": 9.965366573599812e-05, "loss": 0.0802, "step": 6850 }, { "grad_norm": 0.47565412521362305, "learning_rate": 9.965123303554453e-05, "loss": 0.0751, "step": 6860 }, { "grad_norm": 0.65594083070755, "learning_rate": 9.964879185105317e-05, "loss": 0.0836, "step": 6870 }, { "grad_norm": 0.5321378111839294, "learning_rate": 9.964634218294119e-05, "loss": 0.0819, "step": 6880 }, { "grad_norm": 0.51924067735672, "learning_rate": 9.964388403162714e-05, "loss": 0.0794, "step": 6890 }, { "grad_norm": 0.8064002990722656, "learning_rate": 9.96414173975311e-05, "loss": 0.0813, "step": 6900 }, { "grad_norm": 0.5148372054100037, "learning_rate": 9.963894228107451e-05, "loss": 0.0697, "step": 6910 }, { "grad_norm": 0.5310919284820557, "learning_rate": 9.963645868268032e-05, "loss": 0.0781, "step": 6920 }, { "grad_norm": 0.5636833310127258, "learning_rate": 9.963396660277289e-05, "loss": 0.082, "step": 6930 }, { "grad_norm": 0.6057842373847961, "learning_rate": 9.963146604177807e-05, "loss": 0.0687, "step": 6940 }, { "grad_norm": 0.5356021523475647, "learning_rate": 9.962895700012311e-05, "loss": 0.0695, "step": 6950 }, { "grad_norm": 0.5541054010391235, "learning_rate": 9.962643947823677e-05, "loss": 0.0731, "step": 6960 }, { "grad_norm": 0.5091695189476013, "learning_rate": 9.962391347654921e-05, "loss": 0.0834, "step": 6970 }, { "grad_norm": 0.526309072971344, "learning_rate": 9.962137899549204e-05, "loss": 0.0823, "step": 6980 }, { "grad_norm": 0.5006219148635864, "learning_rate": 9.961883603549835e-05, "loss": 0.0647, "step": 6990 }, { "grad_norm": 0.5000775456428528, "learning_rate": 9.961628459700267e-05, "loss": 0.0696, "step": 7000 }, { "grad_norm": 0.5215116739273071, "learning_rate": 9.961372468044095e-05, "loss": 0.0747, "step": 7010 }, { "grad_norm": 0.5379998683929443, "learning_rate": 9.961115628625062e-05, "loss": 0.0633, "step": 7020 }, { "grad_norm": 0.7058515548706055, "learning_rate": 9.960857941487056e-05, "loss": 0.0841, "step": 7030 }, { "grad_norm": 0.5131456255912781, "learning_rate": 9.960599406674106e-05, "loss": 0.0799, "step": 7040 }, { "grad_norm": 0.5755428671836853, "learning_rate": 9.960340024230393e-05, "loss": 0.068, "step": 7050 }, { "grad_norm": 0.49272361397743225, "learning_rate": 9.960079794200232e-05, "loss": 0.0682, "step": 7060 }, { "grad_norm": 0.674399733543396, "learning_rate": 9.959818716628096e-05, "loss": 0.0747, "step": 7070 }, { "grad_norm": 0.5598304867744446, "learning_rate": 9.95955679155859e-05, "loss": 0.0922, "step": 7080 }, { "grad_norm": 0.7120679020881653, "learning_rate": 9.959294019036472e-05, "loss": 0.0872, "step": 7090 }, { "grad_norm": 0.7238549590110779, "learning_rate": 9.959030399106646e-05, "loss": 0.0717, "step": 7100 }, { "grad_norm": 0.5670225024223328, "learning_rate": 9.958765931814153e-05, "loss": 0.0825, "step": 7110 }, { "grad_norm": 0.4525257647037506, "learning_rate": 9.958500617204184e-05, "loss": 0.0656, "step": 7120 }, { "grad_norm": 0.654962956905365, "learning_rate": 9.958234455322075e-05, "loss": 0.068, "step": 7130 }, { "grad_norm": 0.5715701580047607, "learning_rate": 9.957967446213308e-05, "loss": 0.0899, "step": 7140 }, { "grad_norm": 0.5631451606750488, "learning_rate": 9.957699589923501e-05, "loss": 0.0868, "step": 7150 }, { "grad_norm": 0.8464711308479309, "learning_rate": 9.957430886498431e-05, "loss": 0.079, "step": 7160 }, { "grad_norm": 0.5429198145866394, "learning_rate": 9.957161335984008e-05, "loss": 0.0806, "step": 7170 }, { "grad_norm": 0.6436806321144104, "learning_rate": 9.956890938426291e-05, "loss": 0.0652, "step": 7180 }, { "grad_norm": 0.49071913957595825, "learning_rate": 9.956619693871482e-05, "loss": 0.0825, "step": 7190 }, { "grad_norm": 0.9494631290435791, "learning_rate": 9.956347602365934e-05, "loss": 0.0754, "step": 7200 }, { "grad_norm": 0.6399387121200562, "learning_rate": 9.956074663956135e-05, "loss": 0.0851, "step": 7210 }, { "grad_norm": 0.7039481401443481, "learning_rate": 9.955800878688726e-05, "loss": 0.0694, "step": 7220 }, { "grad_norm": 0.5969107747077942, "learning_rate": 9.955526246610489e-05, "loss": 0.0772, "step": 7230 }, { "grad_norm": 0.5533707737922668, "learning_rate": 9.955250767768349e-05, "loss": 0.0764, "step": 7240 }, { "grad_norm": 0.4438389241695404, "learning_rate": 9.95497444220938e-05, "loss": 0.064, "step": 7250 }, { "grad_norm": 0.682032585144043, "learning_rate": 9.954697269980797e-05, "loss": 0.075, "step": 7260 }, { "grad_norm": 0.3834315836429596, "learning_rate": 9.954419251129962e-05, "loss": 0.0657, "step": 7270 }, { "grad_norm": 0.4768194258213043, "learning_rate": 9.95414038570438e-05, "loss": 0.0629, "step": 7280 }, { "grad_norm": 0.3975251615047455, "learning_rate": 9.953860673751703e-05, "loss": 0.0711, "step": 7290 }, { "grad_norm": 0.5031237602233887, "learning_rate": 9.953580115319725e-05, "loss": 0.0738, "step": 7300 }, { "grad_norm": 0.41801023483276367, "learning_rate": 9.953298710456387e-05, "loss": 0.0763, "step": 7310 }, { "grad_norm": 0.5511636137962341, "learning_rate": 9.953016459209771e-05, "loss": 0.0695, "step": 7320 }, { "grad_norm": 0.509403645992279, "learning_rate": 9.952733361628108e-05, "loss": 0.0621, "step": 7330 }, { "grad_norm": 0.5313684940338135, "learning_rate": 9.952449417759772e-05, "loss": 0.0728, "step": 7340 }, { "grad_norm": 0.5414009690284729, "learning_rate": 9.952164627653279e-05, "loss": 0.0715, "step": 7350 }, { "grad_norm": 0.5127963423728943, "learning_rate": 9.951878991357292e-05, "loss": 0.0679, "step": 7360 }, { "grad_norm": 0.49422070384025574, "learning_rate": 9.951592508920622e-05, "loss": 0.0784, "step": 7370 }, { "grad_norm": 0.4906255900859833, "learning_rate": 9.951305180392219e-05, "loss": 0.0713, "step": 7380 }, { "grad_norm": 0.42722973227500916, "learning_rate": 9.951017005821178e-05, "loss": 0.0684, "step": 7390 }, { "grad_norm": 0.4979342818260193, "learning_rate": 9.95072798525674e-05, "loss": 0.0639, "step": 7400 }, { "grad_norm": 0.5871387720108032, "learning_rate": 9.950438118748293e-05, "loss": 0.0722, "step": 7410 }, { "grad_norm": 0.5148041844367981, "learning_rate": 9.950147406345366e-05, "loss": 0.0852, "step": 7420 }, { "grad_norm": 0.5125296711921692, "learning_rate": 9.949855848097635e-05, "loss": 0.0793, "step": 7430 }, { "grad_norm": 0.4633112847805023, "learning_rate": 9.949563444054916e-05, "loss": 0.073, "step": 7440 }, { "grad_norm": 0.5386579036712646, "learning_rate": 9.949270194267178e-05, "loss": 0.0647, "step": 7450 }, { "grad_norm": 0.6267468929290771, "learning_rate": 9.948976098784526e-05, "loss": 0.0678, "step": 7460 }, { "grad_norm": 0.46155881881713867, "learning_rate": 9.948681157657213e-05, "loss": 0.0772, "step": 7470 }, { "grad_norm": 0.5190286636352539, "learning_rate": 9.948385370935638e-05, "loss": 0.0804, "step": 7480 }, { "grad_norm": 0.5560651421546936, "learning_rate": 9.94808873867034e-05, "loss": 0.0759, "step": 7490 }, { "grad_norm": 0.5940940380096436, "learning_rate": 9.947791260912009e-05, "loss": 0.0728, "step": 7500 }, { "grad_norm": 0.5731062293052673, "learning_rate": 9.947492937711474e-05, "loss": 0.0634, "step": 7510 }, { "grad_norm": 0.44373127818107605, "learning_rate": 9.947193769119707e-05, "loss": 0.0753, "step": 7520 }, { "grad_norm": 0.4105991721153259, "learning_rate": 9.946893755187834e-05, "loss": 0.0685, "step": 7530 }, { "grad_norm": 0.4912089705467224, "learning_rate": 9.946592895967115e-05, "loss": 0.0736, "step": 7540 }, { "grad_norm": 0.46915537118911743, "learning_rate": 9.94629119150896e-05, "loss": 0.057, "step": 7550 }, { "grad_norm": 0.48484790325164795, "learning_rate": 9.94598864186492e-05, "loss": 0.0684, "step": 7560 }, { "grad_norm": 0.4978044629096985, "learning_rate": 9.945685247086696e-05, "loss": 0.0775, "step": 7570 }, { "grad_norm": 0.6035637855529785, "learning_rate": 9.945381007226129e-05, "loss": 0.0802, "step": 7580 }, { "grad_norm": 0.5079963803291321, "learning_rate": 9.945075922335203e-05, "loss": 0.0742, "step": 7590 }, { "grad_norm": 0.5211001634597778, "learning_rate": 9.944769992466049e-05, "loss": 0.0905, "step": 7600 }, { "grad_norm": 0.46609625220298767, "learning_rate": 9.944463217670945e-05, "loss": 0.0817, "step": 7610 }, { "grad_norm": 0.4944364130496979, "learning_rate": 9.944155598002307e-05, "loss": 0.0698, "step": 7620 }, { "grad_norm": 0.43898847699165344, "learning_rate": 9.943847133512701e-05, "loss": 0.068, "step": 7630 }, { "grad_norm": 0.6899057030677795, "learning_rate": 9.943537824254834e-05, "loss": 0.0689, "step": 7640 }, { "grad_norm": 0.5510210394859314, "learning_rate": 9.943227670281559e-05, "loss": 0.0625, "step": 7650 }, { "grad_norm": 0.42597001791000366, "learning_rate": 9.942916671645873e-05, "loss": 0.057, "step": 7660 }, { "grad_norm": 0.6361128091812134, "learning_rate": 9.942604828400916e-05, "loss": 0.0679, "step": 7670 }, { "grad_norm": 0.5280230641365051, "learning_rate": 9.942292140599975e-05, "loss": 0.0648, "step": 7680 }, { "grad_norm": 0.5942790508270264, "learning_rate": 9.94197860829648e-05, "loss": 0.0692, "step": 7690 }, { "grad_norm": 0.48870232701301575, "learning_rate": 9.941664231544004e-05, "loss": 0.0579, "step": 7700 }, { "grad_norm": 0.615508496761322, "learning_rate": 9.941349010396264e-05, "loss": 0.0784, "step": 7710 }, { "grad_norm": 0.5525040626525879, "learning_rate": 9.941032944907125e-05, "loss": 0.0726, "step": 7720 }, { "grad_norm": 0.47463011741638184, "learning_rate": 9.940716035130596e-05, "loss": 0.068, "step": 7730 }, { "grad_norm": 0.6757602095603943, "learning_rate": 9.940398281120821e-05, "loss": 0.0699, "step": 7740 }, { "grad_norm": 0.4643668830394745, "learning_rate": 9.940079682932102e-05, "loss": 0.0608, "step": 7750 }, { "grad_norm": 0.5512277483940125, "learning_rate": 9.939760240618877e-05, "loss": 0.0719, "step": 7760 }, { "grad_norm": 0.5704546570777893, "learning_rate": 9.939439954235729e-05, "loss": 0.0748, "step": 7770 }, { "grad_norm": 0.58283531665802, "learning_rate": 9.939118823837387e-05, "loss": 0.0746, "step": 7780 }, { "grad_norm": 0.4736770987510681, "learning_rate": 9.938796849478725e-05, "loss": 0.0736, "step": 7790 }, { "grad_norm": 0.6247066259384155, "learning_rate": 9.938474031214755e-05, "loss": 0.0742, "step": 7800 }, { "grad_norm": 0.5837437510490417, "learning_rate": 9.938150369100643e-05, "loss": 0.0728, "step": 7810 }, { "grad_norm": 0.5628241300582886, "learning_rate": 9.93782586319169e-05, "loss": 0.0578, "step": 7820 }, { "grad_norm": 0.5226519107818604, "learning_rate": 9.937500513543348e-05, "loss": 0.075, "step": 7830 }, { "grad_norm": 0.5261894464492798, "learning_rate": 9.937174320211207e-05, "loss": 0.0648, "step": 7840 }, { "grad_norm": 0.60637366771698, "learning_rate": 9.936847283251009e-05, "loss": 0.0741, "step": 7850 }, { "grad_norm": 0.5370784401893616, "learning_rate": 9.936519402718632e-05, "loss": 0.0683, "step": 7860 }, { "grad_norm": 0.6187614798545837, "learning_rate": 9.936190678670102e-05, "loss": 0.0853, "step": 7870 }, { "grad_norm": 0.5178565979003906, "learning_rate": 9.935861111161593e-05, "loss": 0.0828, "step": 7880 }, { "grad_norm": 0.4242347776889801, "learning_rate": 9.935530700249416e-05, "loss": 0.0716, "step": 7890 }, { "grad_norm": 0.4177737534046173, "learning_rate": 9.935199445990028e-05, "loss": 0.0834, "step": 7900 }, { "grad_norm": 0.5227515697479248, "learning_rate": 9.934867348440033e-05, "loss": 0.0809, "step": 7910 }, { "grad_norm": 0.43903395533561707, "learning_rate": 9.934534407656176e-05, "loss": 0.1009, "step": 7920 }, { "grad_norm": 0.5056912899017334, "learning_rate": 9.93420062369535e-05, "loss": 0.0684, "step": 7930 }, { "grad_norm": 0.82958984375, "learning_rate": 9.933865996614589e-05, "loss": 0.0714, "step": 7940 }, { "grad_norm": 0.6318562030792236, "learning_rate": 9.933530526471068e-05, "loss": 0.0634, "step": 7950 }, { "grad_norm": 0.4152906835079193, "learning_rate": 9.933194213322114e-05, "loss": 0.0665, "step": 7960 }, { "grad_norm": 0.5213850736618042, "learning_rate": 9.932857057225192e-05, "loss": 0.0723, "step": 7970 }, { "grad_norm": 0.43639031052589417, "learning_rate": 9.932519058237912e-05, "loss": 0.062, "step": 7980 }, { "grad_norm": 0.4171326160430908, "learning_rate": 9.932180216418032e-05, "loss": 0.0679, "step": 7990 }, { "grad_norm": 0.5952378511428833, "learning_rate": 9.931840531823446e-05, "loss": 0.0646, "step": 8000 }, { "grad_norm": 0.5556617379188538, "learning_rate": 9.9315000045122e-05, "loss": 0.0651, "step": 8010 }, { "grad_norm": 0.5303550362586975, "learning_rate": 9.931158634542481e-05, "loss": 0.059, "step": 8020 }, { "grad_norm": 0.4328838884830475, "learning_rate": 9.930816421972617e-05, "loss": 0.0609, "step": 8030 }, { "grad_norm": 0.6107727289199829, "learning_rate": 9.930473366861086e-05, "loss": 0.067, "step": 8040 }, { "grad_norm": 0.5726258754730225, "learning_rate": 9.930129469266505e-05, "loss": 0.0719, "step": 8050 }, { "grad_norm": 0.6803480982780457, "learning_rate": 9.929784729247638e-05, "loss": 0.063, "step": 8060 }, { "grad_norm": 0.5867189168930054, "learning_rate": 9.929439146863389e-05, "loss": 0.0681, "step": 8070 }, { "grad_norm": 0.8171291947364807, "learning_rate": 9.92909272217281e-05, "loss": 0.0926, "step": 8080 }, { "grad_norm": 0.6574304699897766, "learning_rate": 9.928745455235097e-05, "loss": 0.0808, "step": 8090 }, { "grad_norm": 0.546486496925354, "learning_rate": 9.928397346109588e-05, "loss": 0.0853, "step": 8100 }, { "grad_norm": 0.6328953504562378, "learning_rate": 9.928048394855762e-05, "loss": 0.0768, "step": 8110 }, { "grad_norm": 0.4288747012615204, "learning_rate": 9.92769860153325e-05, "loss": 0.0708, "step": 8120 }, { "grad_norm": 0.7061898708343506, "learning_rate": 9.927347966201819e-05, "loss": 0.0845, "step": 8130 }, { "grad_norm": 0.5425888895988464, "learning_rate": 9.926996488921383e-05, "loss": 0.0843, "step": 8140 }, { "grad_norm": 0.6308217644691467, "learning_rate": 9.926644169752001e-05, "loss": 0.0805, "step": 8150 }, { "grad_norm": 0.4570934772491455, "learning_rate": 9.926291008753875e-05, "loss": 0.0791, "step": 8160 }, { "grad_norm": 0.4758504331111908, "learning_rate": 9.92593700598735e-05, "loss": 0.0719, "step": 8170 }, { "grad_norm": 0.420696884393692, "learning_rate": 9.925582161512915e-05, "loss": 0.0732, "step": 8180 }, { "grad_norm": 0.44436559081077576, "learning_rate": 9.925226475391205e-05, "loss": 0.0697, "step": 8190 }, { "grad_norm": 0.48338890075683594, "learning_rate": 9.924869947682993e-05, "loss": 0.0653, "step": 8200 }, { "grad_norm": 0.3817466199398041, "learning_rate": 9.924512578449204e-05, "loss": 0.0516, "step": 8210 }, { "grad_norm": 0.4464687705039978, "learning_rate": 9.924154367750901e-05, "loss": 0.0626, "step": 8220 }, { "grad_norm": 0.4957428276538849, "learning_rate": 9.923795315649293e-05, "loss": 0.08, "step": 8230 }, { "grad_norm": 0.4266600012779236, "learning_rate": 9.92343542220573e-05, "loss": 0.0713, "step": 8240 }, { "grad_norm": 0.6675784587860107, "learning_rate": 9.92307468748171e-05, "loss": 0.0753, "step": 8250 }, { "grad_norm": 0.6567960381507874, "learning_rate": 9.922713111538873e-05, "loss": 0.068, "step": 8260 }, { "grad_norm": 0.4890688359737396, "learning_rate": 9.922350694439003e-05, "loss": 0.0615, "step": 8270 }, { "grad_norm": 0.5633100867271423, "learning_rate": 9.921987436244024e-05, "loss": 0.0805, "step": 8280 }, { "grad_norm": 0.47718381881713867, "learning_rate": 9.921623337016008e-05, "loss": 0.0618, "step": 8290 }, { "grad_norm": 0.3853529989719391, "learning_rate": 9.921258396817172e-05, "loss": 0.0711, "step": 8300 }, { "grad_norm": 0.5278508067131042, "learning_rate": 9.920892615709874e-05, "loss": 0.0687, "step": 8310 }, { "grad_norm": 0.4226672053337097, "learning_rate": 9.920525993756612e-05, "loss": 0.0595, "step": 8320 }, { "grad_norm": 0.6014373898506165, "learning_rate": 9.920158531020036e-05, "loss": 0.0747, "step": 8330 }, { "grad_norm": 0.5273570418357849, "learning_rate": 9.919790227562933e-05, "loss": 0.0659, "step": 8340 }, { "grad_norm": 0.49598225951194763, "learning_rate": 9.919421083448237e-05, "loss": 0.0837, "step": 8350 }, { "grad_norm": 0.6267748475074768, "learning_rate": 9.919051098739022e-05, "loss": 0.0647, "step": 8360 }, { "grad_norm": 0.46919581294059753, "learning_rate": 9.918680273498514e-05, "loss": 0.062, "step": 8370 }, { "grad_norm": 0.4897213876247406, "learning_rate": 9.918308607790072e-05, "loss": 0.0696, "step": 8380 }, { "grad_norm": 0.5509771108627319, "learning_rate": 9.917936101677205e-05, "loss": 0.0587, "step": 8390 }, { "grad_norm": 0.7149263024330139, "learning_rate": 9.917562755223564e-05, "loss": 0.0694, "step": 8400 }, { "grad_norm": 0.5478460192680359, "learning_rate": 9.917188568492944e-05, "loss": 0.1275, "step": 8410 }, { "grad_norm": 0.43333879113197327, "learning_rate": 9.916813541549283e-05, "loss": 0.0612, "step": 8420 }, { "grad_norm": 0.5720267295837402, "learning_rate": 9.916437674456663e-05, "loss": 0.0667, "step": 8430 }, { "grad_norm": 0.6261653900146484, "learning_rate": 9.916060967279308e-05, "loss": 0.0694, "step": 8440 }, { "grad_norm": 0.886306881904602, "learning_rate": 9.91568342008159e-05, "loss": 0.0714, "step": 8450 }, { "grad_norm": 0.46946990489959717, "learning_rate": 9.915305032928019e-05, "loss": 0.0668, "step": 8460 }, { "grad_norm": 0.49219009280204773, "learning_rate": 9.914925805883253e-05, "loss": 0.0662, "step": 8470 }, { "grad_norm": 0.4825650751590729, "learning_rate": 9.914545739012088e-05, "loss": 0.0605, "step": 8480 }, { "grad_norm": 0.5862884521484375, "learning_rate": 9.91416483237947e-05, "loss": 0.0645, "step": 8490 }, { "grad_norm": 0.6750649809837341, "learning_rate": 9.913783086050485e-05, "loss": 0.07, "step": 8500 }, { "grad_norm": 0.48771417140960693, "learning_rate": 9.913400500090364e-05, "loss": 0.0594, "step": 8510 }, { "grad_norm": 0.5051802396774292, "learning_rate": 9.913017074564479e-05, "loss": 0.0503, "step": 8520 }, { "grad_norm": 0.44861388206481934, "learning_rate": 9.912632809538348e-05, "loss": 0.0679, "step": 8530 }, { "grad_norm": 0.48398855328559875, "learning_rate": 9.912247705077629e-05, "loss": 0.0677, "step": 8540 }, { "grad_norm": 0.5332832336425781, "learning_rate": 9.911861761248127e-05, "loss": 0.0643, "step": 8550 }, { "grad_norm": 0.49102020263671875, "learning_rate": 9.91147497811579e-05, "loss": 0.0654, "step": 8560 }, { "grad_norm": 0.6545950770378113, "learning_rate": 9.911087355746709e-05, "loss": 0.0941, "step": 8570 }, { "grad_norm": 0.5182519555091858, "learning_rate": 9.910698894207117e-05, "loss": 0.084, "step": 8580 }, { "grad_norm": 0.4486962556838989, "learning_rate": 9.910309593563392e-05, "loss": 0.0772, "step": 8590 }, { "grad_norm": 0.3981485962867737, "learning_rate": 9.909919453882057e-05, "loss": 0.0613, "step": 8600 }, { "grad_norm": 0.6555084586143494, "learning_rate": 9.90952847522977e-05, "loss": 0.0567, "step": 8610 }, { "grad_norm": 0.6270935535430908, "learning_rate": 9.909136657673346e-05, "loss": 0.0853, "step": 8620 }, { "grad_norm": 0.4889476001262665, "learning_rate": 9.908744001279731e-05, "loss": 0.0672, "step": 8630 }, { "grad_norm": 0.7306236624717712, "learning_rate": 9.90835050611602e-05, "loss": 0.08, "step": 8640 }, { "grad_norm": 0.5370626449584961, "learning_rate": 9.90795617224945e-05, "loss": 0.0792, "step": 8650 }, { "grad_norm": 0.5990990400314331, "learning_rate": 9.907560999747405e-05, "loss": 0.0646, "step": 8660 }, { "grad_norm": 0.48506319522857666, "learning_rate": 9.907164988677408e-05, "loss": 0.0545, "step": 8670 }, { "grad_norm": 0.6293627023696899, "learning_rate": 9.906768139107124e-05, "loss": 0.0618, "step": 8680 }, { "grad_norm": 0.5406057238578796, "learning_rate": 9.906370451104367e-05, "loss": 0.0634, "step": 8690 }, { "grad_norm": 0.4694159924983978, "learning_rate": 9.905971924737088e-05, "loss": 0.0695, "step": 8700 }, { "grad_norm": 0.5048756003379822, "learning_rate": 9.905572560073387e-05, "loss": 0.0674, "step": 8710 }, { "grad_norm": 0.8242479562759399, "learning_rate": 9.905172357181501e-05, "loss": 0.0884, "step": 8720 }, { "grad_norm": 0.5479881167411804, "learning_rate": 9.904771316129817e-05, "loss": 0.0693, "step": 8730 }, { "grad_norm": 0.4449993371963501, "learning_rate": 9.904369436986862e-05, "loss": 0.0666, "step": 8740 }, { "grad_norm": 0.6059691905975342, "learning_rate": 9.903966719821303e-05, "loss": 0.0726, "step": 8750 }, { "grad_norm": 0.6519355177879333, "learning_rate": 9.903563164701956e-05, "loss": 0.0678, "step": 8760 }, { "grad_norm": 0.42957574129104614, "learning_rate": 9.903158771697778e-05, "loss": 0.0655, "step": 8770 }, { "grad_norm": 0.4703845679759979, "learning_rate": 9.902753540877867e-05, "loss": 0.0657, "step": 8780 }, { "grad_norm": 0.47848162055015564, "learning_rate": 9.902347472311466e-05, "loss": 0.0556, "step": 8790 }, { "grad_norm": 0.572902262210846, "learning_rate": 9.901940566067962e-05, "loss": 0.0512, "step": 8800 }, { "grad_norm": 0.3607097864151001, "learning_rate": 9.901532822216883e-05, "loss": 0.0589, "step": 8810 }, { "grad_norm": 0.5360502600669861, "learning_rate": 9.901124240827904e-05, "loss": 0.0715, "step": 8820 }, { "grad_norm": 0.4063662886619568, "learning_rate": 9.900714821970835e-05, "loss": 0.0567, "step": 8830 }, { "grad_norm": 0.7503601908683777, "learning_rate": 9.900304565715641e-05, "loss": 0.0675, "step": 8840 }, { "grad_norm": 0.47445541620254517, "learning_rate": 9.899893472132419e-05, "loss": 0.0635, "step": 8850 }, { "grad_norm": 0.48820123076438904, "learning_rate": 9.899481541291415e-05, "loss": 0.0609, "step": 8860 }, { "grad_norm": 0.7726028561592102, "learning_rate": 9.899068773263016e-05, "loss": 0.0683, "step": 8870 }, { "grad_norm": 0.5336052775382996, "learning_rate": 9.898655168117754e-05, "loss": 0.068, "step": 8880 }, { "grad_norm": 0.40871381759643555, "learning_rate": 9.898240725926302e-05, "loss": 0.0655, "step": 8890 }, { "grad_norm": 0.47433337569236755, "learning_rate": 9.897825446759478e-05, "loss": 0.065, "step": 8900 }, { "grad_norm": 0.4607186019420624, "learning_rate": 9.897409330688241e-05, "loss": 0.059, "step": 8910 }, { "grad_norm": 0.6440395712852478, "learning_rate": 9.896992377783692e-05, "loss": 0.0749, "step": 8920 }, { "grad_norm": 0.6035154461860657, "learning_rate": 9.89657458811708e-05, "loss": 0.0504, "step": 8930 }, { "grad_norm": 0.41375234723091125, "learning_rate": 9.896155961759792e-05, "loss": 0.0643, "step": 8940 }, { "grad_norm": 0.3570089042186737, "learning_rate": 9.895736498783361e-05, "loss": 0.0622, "step": 8950 }, { "grad_norm": 0.49096405506134033, "learning_rate": 9.895316199259462e-05, "loss": 0.0531, "step": 8960 }, { "grad_norm": 0.4829016625881195, "learning_rate": 9.894895063259909e-05, "loss": 0.0493, "step": 8970 }, { "grad_norm": 0.42404904961586, "learning_rate": 9.894473090856667e-05, "loss": 0.0588, "step": 8980 }, { "grad_norm": 0.6653497815132141, "learning_rate": 9.894050282121839e-05, "loss": 0.0859, "step": 8990 }, { "grad_norm": 0.6202142834663391, "learning_rate": 9.893626637127668e-05, "loss": 0.0617, "step": 9000 }, { "grad_norm": 0.5063773989677429, "learning_rate": 9.893202155946546e-05, "loss": 0.0498, "step": 9010 }, { "grad_norm": 0.6846022605895996, "learning_rate": 9.892776838651006e-05, "loss": 0.0615, "step": 9020 }, { "grad_norm": 0.5182366371154785, "learning_rate": 9.892350685313722e-05, "loss": 0.0823, "step": 9030 }, { "grad_norm": 0.643796980381012, "learning_rate": 9.891923696007513e-05, "loss": 0.0726, "step": 9040 }, { "grad_norm": 0.6249280571937561, "learning_rate": 9.891495870805336e-05, "loss": 0.0624, "step": 9050 }, { "grad_norm": 0.5307380557060242, "learning_rate": 9.891067209780298e-05, "loss": 0.0793, "step": 9060 }, { "grad_norm": 0.5465173125267029, "learning_rate": 9.890637713005646e-05, "loss": 0.0583, "step": 9070 }, { "grad_norm": 0.4754057228565216, "learning_rate": 9.890207380554767e-05, "loss": 0.0654, "step": 9080 }, { "grad_norm": 0.5983615517616272, "learning_rate": 9.889776212501196e-05, "loss": 0.0742, "step": 9090 }, { "grad_norm": 0.44310295581817627, "learning_rate": 9.889344208918605e-05, "loss": 0.0894, "step": 9100 }, { "grad_norm": 0.6278305053710938, "learning_rate": 9.888911369880812e-05, "loss": 0.0752, "step": 9110 }, { "grad_norm": 0.4518612027168274, "learning_rate": 9.888477695461777e-05, "loss": 0.0703, "step": 9120 }, { "grad_norm": 0.5691809058189392, "learning_rate": 9.888043185735607e-05, "loss": 0.0732, "step": 9130 }, { "grad_norm": 0.5324551463127136, "learning_rate": 9.887607840776542e-05, "loss": 0.0588, "step": 9140 }, { "grad_norm": 0.6847150921821594, "learning_rate": 9.887171660658975e-05, "loss": 0.0728, "step": 9150 }, { "grad_norm": 0.3897951543331146, "learning_rate": 9.886734645457435e-05, "loss": 0.0519, "step": 9160 }, { "grad_norm": 0.46344834566116333, "learning_rate": 9.886296795246597e-05, "loss": 0.0558, "step": 9170 }, { "grad_norm": 0.4110753834247589, "learning_rate": 9.885858110101276e-05, "loss": 0.055, "step": 9180 }, { "grad_norm": 0.3707296550273895, "learning_rate": 9.885418590096434e-05, "loss": 0.0519, "step": 9190 }, { "grad_norm": 0.4284105598926544, "learning_rate": 9.88497823530717e-05, "loss": 0.0583, "step": 9200 }, { "grad_norm": 0.7158948183059692, "learning_rate": 9.884537045808732e-05, "loss": 0.0827, "step": 9210 }, { "grad_norm": 0.782511830329895, "learning_rate": 9.884095021676502e-05, "loss": 0.0806, "step": 9220 }, { "grad_norm": 0.6144619584083557, "learning_rate": 9.883652162986017e-05, "loss": 0.067, "step": 9230 }, { "grad_norm": 0.4491519629955292, "learning_rate": 9.883208469812943e-05, "loss": 0.0733, "step": 9240 }, { "grad_norm": 0.42262938618659973, "learning_rate": 9.882763942233098e-05, "loss": 0.054, "step": 9250 }, { "grad_norm": 0.414000540971756, "learning_rate": 9.882318580322441e-05, "loss": 0.0558, "step": 9260 }, { "grad_norm": 0.4775155782699585, "learning_rate": 9.881872384157067e-05, "loss": 0.0587, "step": 9270 }, { "grad_norm": 0.4252065122127533, "learning_rate": 9.881425353813225e-05, "loss": 0.0639, "step": 9280 }, { "grad_norm": 0.6273467540740967, "learning_rate": 9.880977489367296e-05, "loss": 0.0742, "step": 9290 }, { "grad_norm": 0.45408689975738525, "learning_rate": 9.88052879089581e-05, "loss": 0.0698, "step": 9300 }, { "grad_norm": 0.47775042057037354, "learning_rate": 9.880079258475434e-05, "loss": 0.0484, "step": 9310 }, { "grad_norm": 0.41576021909713745, "learning_rate": 9.879628892182985e-05, "loss": 0.0726, "step": 9320 }, { "grad_norm": 0.40257054567337036, "learning_rate": 9.879177692095416e-05, "loss": 0.0586, "step": 9330 }, { "grad_norm": 0.5232487916946411, "learning_rate": 9.878725658289825e-05, "loss": 0.0748, "step": 9340 }, { "grad_norm": 0.41623589396476746, "learning_rate": 9.878272790843454e-05, "loss": 0.0556, "step": 9350 }, { "grad_norm": 0.509713351726532, "learning_rate": 9.877819089833682e-05, "loss": 0.0534, "step": 9360 }, { "grad_norm": 0.5645818114280701, "learning_rate": 9.877364555338038e-05, "loss": 0.0732, "step": 9370 }, { "grad_norm": 0.4391767382621765, "learning_rate": 9.876909187434186e-05, "loss": 0.0715, "step": 9380 }, { "grad_norm": 0.6010136604309082, "learning_rate": 9.876452986199939e-05, "loss": 0.0663, "step": 9390 }, { "grad_norm": 0.6206870079040527, "learning_rate": 9.875995951713248e-05, "loss": 0.0757, "step": 9400 }, { "grad_norm": 0.5567260980606079, "learning_rate": 9.875538084052207e-05, "loss": 0.0674, "step": 9410 }, { "grad_norm": 0.6030465960502625, "learning_rate": 9.875079383295053e-05, "loss": 0.0618, "step": 9420 }, { "grad_norm": 0.424704909324646, "learning_rate": 9.874619849520167e-05, "loss": 0.0523, "step": 9430 }, { "grad_norm": 0.47548139095306396, "learning_rate": 9.874159482806069e-05, "loss": 0.0729, "step": 9440 }, { "grad_norm": 0.5932033658027649, "learning_rate": 9.873698283231426e-05, "loss": 0.0651, "step": 9450 }, { "grad_norm": 0.6316573023796082, "learning_rate": 9.87323625087504e-05, "loss": 0.058, "step": 9460 }, { "grad_norm": 0.5786330699920654, "learning_rate": 9.872773385815863e-05, "loss": 0.0797, "step": 9470 }, { "grad_norm": 0.5010982751846313, "learning_rate": 9.872309688132986e-05, "loss": 0.0611, "step": 9480 }, { "grad_norm": 0.3931989073753357, "learning_rate": 9.871845157905639e-05, "loss": 0.0604, "step": 9490 }, { "grad_norm": 0.4166123569011688, "learning_rate": 9.871379795213201e-05, "loss": 0.0536, "step": 9500 }, { "grad_norm": 0.5203947424888611, "learning_rate": 9.87091360013519e-05, "loss": 0.0697, "step": 9510 }, { "grad_norm": 0.473317950963974, "learning_rate": 9.870446572751262e-05, "loss": 0.0641, "step": 9520 }, { "grad_norm": 0.46522995829582214, "learning_rate": 9.869978713141224e-05, "loss": 0.0601, "step": 9530 }, { "grad_norm": 0.6542088985443115, "learning_rate": 9.869510021385016e-05, "loss": 0.0698, "step": 9540 }, { "grad_norm": 0.4595860242843628, "learning_rate": 9.869040497562727e-05, "loss": 0.0586, "step": 9550 }, { "grad_norm": 0.5141375660896301, "learning_rate": 9.868570141754587e-05, "loss": 0.0717, "step": 9560 }, { "grad_norm": 0.5653437376022339, "learning_rate": 9.868098954040965e-05, "loss": 0.0813, "step": 9570 }, { "grad_norm": 0.6497948169708252, "learning_rate": 9.867626934502374e-05, "loss": 0.0735, "step": 9580 }, { "grad_norm": 0.5258520245552063, "learning_rate": 9.86715408321947e-05, "loss": 0.0658, "step": 9590 }, { "grad_norm": 0.5142753720283508, "learning_rate": 9.86668040027305e-05, "loss": 0.0605, "step": 9600 }, { "grad_norm": 0.42259785532951355, "learning_rate": 9.866205885744053e-05, "loss": 0.0628, "step": 9610 }, { "grad_norm": 0.45583319664001465, "learning_rate": 9.865730539713563e-05, "loss": 0.0781, "step": 9620 }, { "grad_norm": 0.42095789313316345, "learning_rate": 9.8652543622628e-05, "loss": 0.0638, "step": 9630 }, { "grad_norm": 0.46260473132133484, "learning_rate": 9.864777353473132e-05, "loss": 0.0599, "step": 9640 }, { "grad_norm": 0.39356809854507446, "learning_rate": 9.864299513426068e-05, "loss": 0.0599, "step": 9650 }, { "grad_norm": 0.42618849873542786, "learning_rate": 9.863820842203254e-05, "loss": 0.0585, "step": 9660 }, { "grad_norm": 0.48902028799057007, "learning_rate": 9.863341339886483e-05, "loss": 0.0649, "step": 9670 }, { "grad_norm": 0.4112946689128876, "learning_rate": 9.86286100655769e-05, "loss": 0.0653, "step": 9680 }, { "grad_norm": 1.0037940740585327, "learning_rate": 9.862379842298953e-05, "loss": 0.0763, "step": 9690 }, { "grad_norm": 0.6025242805480957, "learning_rate": 9.861897847192485e-05, "loss": 0.0619, "step": 9700 }, { "grad_norm": 0.47899678349494934, "learning_rate": 9.86141502132065e-05, "loss": 0.0779, "step": 9710 }, { "grad_norm": 0.5037445425987244, "learning_rate": 9.860931364765946e-05, "loss": 0.0603, "step": 9720 }, { "grad_norm": 1.1988749504089355, "learning_rate": 9.860446877611021e-05, "loss": 0.0768, "step": 9730 }, { "grad_norm": 0.5338507294654846, "learning_rate": 9.859961559938655e-05, "loss": 0.0759, "step": 9740 }, { "grad_norm": 0.6123878359794617, "learning_rate": 9.85947541183178e-05, "loss": 0.0676, "step": 9750 }, { "grad_norm": 0.6703047752380371, "learning_rate": 9.858988433373463e-05, "loss": 0.0722, "step": 9760 }, { "grad_norm": 0.45104101300239563, "learning_rate": 9.858500624646918e-05, "loss": 0.0654, "step": 9770 }, { "grad_norm": 0.6899626851081848, "learning_rate": 9.858011985735497e-05, "loss": 0.0628, "step": 9780 }, { "grad_norm": 0.8473296165466309, "learning_rate": 9.857522516722693e-05, "loss": 0.0695, "step": 9790 }, { "grad_norm": 0.5361933708190918, "learning_rate": 9.857032217692145e-05, "loss": 0.0526, "step": 9800 }, { "grad_norm": 0.4887063801288605, "learning_rate": 9.856541088727631e-05, "loss": 0.0653, "step": 9810 }, { "grad_norm": 0.482338547706604, "learning_rate": 9.856049129913072e-05, "loss": 0.0734, "step": 9820 }, { "grad_norm": 0.4609827697277069, "learning_rate": 9.85555634133253e-05, "loss": 0.072, "step": 9830 }, { "grad_norm": 0.5114256143569946, "learning_rate": 9.855062723070208e-05, "loss": 0.0643, "step": 9840 }, { "grad_norm": 0.40850764513015747, "learning_rate": 9.854568275210454e-05, "loss": 0.0699, "step": 9850 }, { "grad_norm": 0.44655904173851013, "learning_rate": 9.854072997837754e-05, "loss": 0.0596, "step": 9860 }, { "grad_norm": 0.4498429596424103, "learning_rate": 9.853576891036737e-05, "loss": 0.0645, "step": 9870 }, { "grad_norm": 0.46200746297836304, "learning_rate": 9.853079954892177e-05, "loss": 0.0634, "step": 9880 }, { "grad_norm": 0.36810481548309326, "learning_rate": 9.852582189488983e-05, "loss": 0.0529, "step": 9890 }, { "grad_norm": 0.5962795615196228, "learning_rate": 9.852083594912212e-05, "loss": 0.0658, "step": 9900 }, { "grad_norm": 0.45179131627082825, "learning_rate": 9.851584171247058e-05, "loss": 0.0624, "step": 9910 }, { "grad_norm": 0.5970026254653931, "learning_rate": 9.851083918578863e-05, "loss": 0.0682, "step": 9920 }, { "grad_norm": 0.572847306728363, "learning_rate": 9.850582836993103e-05, "loss": 0.0615, "step": 9930 }, { "grad_norm": 0.44001445174217224, "learning_rate": 9.850080926575397e-05, "loss": 0.0605, "step": 9940 }, { "grad_norm": 0.36874979734420776, "learning_rate": 9.849578187411515e-05, "loss": 0.054, "step": 9950 }, { "grad_norm": 0.3939172625541687, "learning_rate": 9.849074619587354e-05, "loss": 0.0456, "step": 9960 }, { "grad_norm": 0.48721393942832947, "learning_rate": 9.848570223188964e-05, "loss": 0.0634, "step": 9970 }, { "grad_norm": 0.4774163067340851, "learning_rate": 9.848064998302531e-05, "loss": 0.0562, "step": 9980 }, { "grad_norm": 0.6921958923339844, "learning_rate": 9.847558945014386e-05, "loss": 0.0654, "step": 9990 }, { "grad_norm": 0.6543124914169312, "learning_rate": 9.847052063410996e-05, "loss": 0.068, "step": 10000 }, { "grad_norm": 0.38625553250312805, "learning_rate": 9.846544353578977e-05, "loss": 0.0651, "step": 10010 }, { "grad_norm": 0.4190619885921478, "learning_rate": 9.846035815605081e-05, "loss": 0.068, "step": 10020 }, { "grad_norm": 0.40990784764289856, "learning_rate": 9.845526449576204e-05, "loss": 0.0544, "step": 10030 }, { "grad_norm": 0.5395299792289734, "learning_rate": 9.845016255579383e-05, "loss": 0.0646, "step": 10040 }, { "grad_norm": 0.6260796785354614, "learning_rate": 9.844505233701794e-05, "loss": 0.0635, "step": 10050 }, { "grad_norm": 0.44381195306777954, "learning_rate": 9.843993384030757e-05, "loss": 0.0631, "step": 10060 }, { "grad_norm": 0.4783567190170288, "learning_rate": 9.843480706653737e-05, "loss": 0.0464, "step": 10070 }, { "grad_norm": 0.5137284398078918, "learning_rate": 9.84296720165833e-05, "loss": 0.0599, "step": 10080 }, { "grad_norm": 0.5647042989730835, "learning_rate": 9.842452869132286e-05, "loss": 0.0508, "step": 10090 }, { "grad_norm": 0.5509084463119507, "learning_rate": 9.841937709163489e-05, "loss": 0.0595, "step": 10100 }, { "grad_norm": 0.40811359882354736, "learning_rate": 9.841421721839962e-05, "loss": 0.0525, "step": 10110 }, { "grad_norm": 0.4761819541454315, "learning_rate": 9.840904907249879e-05, "loss": 0.0466, "step": 10120 }, { "grad_norm": 0.4966767430305481, "learning_rate": 9.840387265481545e-05, "loss": 0.0655, "step": 10130 }, { "grad_norm": 0.5052791833877563, "learning_rate": 9.839868796623411e-05, "loss": 0.0516, "step": 10140 }, { "grad_norm": 0.5694230198860168, "learning_rate": 9.839349500764072e-05, "loss": 0.0682, "step": 10150 }, { "grad_norm": 0.5806097388267517, "learning_rate": 9.83882937799226e-05, "loss": 0.0642, "step": 10160 }, { "grad_norm": 0.4469643533229828, "learning_rate": 9.838308428396849e-05, "loss": 0.0549, "step": 10170 }, { "grad_norm": 0.4467284083366394, "learning_rate": 9.837786652066854e-05, "loss": 0.0709, "step": 10180 }, { "grad_norm": 0.5390962362289429, "learning_rate": 9.837264049091437e-05, "loss": 0.0776, "step": 10190 }, { "grad_norm": 0.5099854469299316, "learning_rate": 9.836740619559893e-05, "loss": 0.0614, "step": 10200 }, { "grad_norm": 0.4153328537940979, "learning_rate": 9.836216363561659e-05, "loss": 0.0577, "step": 10210 }, { "grad_norm": 0.46430709958076477, "learning_rate": 9.835691281186322e-05, "loss": 0.0618, "step": 10220 }, { "grad_norm": 0.45401135087013245, "learning_rate": 9.8351653725236e-05, "loss": 0.0606, "step": 10230 }, { "grad_norm": 0.39338743686676025, "learning_rate": 9.83463863766336e-05, "loss": 0.059, "step": 10240 }, { "grad_norm": 0.680844247341156, "learning_rate": 9.834111076695602e-05, "loss": 0.0526, "step": 10250 }, { "grad_norm": 0.42817193269729614, "learning_rate": 9.833582689710477e-05, "loss": 0.0412, "step": 10260 }, { "grad_norm": 0.3915446996688843, "learning_rate": 9.833053476798268e-05, "loss": 0.0489, "step": 10270 }, { "grad_norm": 0.4481029510498047, "learning_rate": 9.832523438049404e-05, "loss": 0.0582, "step": 10280 }, { "grad_norm": 0.5533279776573181, "learning_rate": 9.831992573554454e-05, "loss": 0.066, "step": 10290 }, { "grad_norm": 0.5626020431518555, "learning_rate": 9.831460883404128e-05, "loss": 0.0668, "step": 10300 }, { "grad_norm": 0.5027697682380676, "learning_rate": 9.830928367689278e-05, "loss": 0.052, "step": 10310 }, { "grad_norm": 0.4525052011013031, "learning_rate": 9.830395026500896e-05, "loss": 0.057, "step": 10320 }, { "grad_norm": 0.46236374974250793, "learning_rate": 9.829860859930115e-05, "loss": 0.0538, "step": 10330 }, { "grad_norm": 0.4886053502559662, "learning_rate": 9.829325868068212e-05, "loss": 0.0575, "step": 10340 }, { "grad_norm": 0.4220532178878784, "learning_rate": 9.8287900510066e-05, "loss": 0.0661, "step": 10350 }, { "grad_norm": 0.38570278882980347, "learning_rate": 9.828253408836834e-05, "loss": 0.0504, "step": 10360 }, { "grad_norm": 0.41280725598335266, "learning_rate": 9.827715941650615e-05, "loss": 0.0482, "step": 10370 }, { "grad_norm": 0.525398850440979, "learning_rate": 9.82717764953978e-05, "loss": 0.0547, "step": 10380 }, { "grad_norm": 0.4612923264503479, "learning_rate": 9.826638532596308e-05, "loss": 0.057, "step": 10390 }, { "grad_norm": 0.38469377160072327, "learning_rate": 9.82609859091232e-05, "loss": 0.0608, "step": 10400 }, { "grad_norm": 0.42871448397636414, "learning_rate": 9.825557824580076e-05, "loss": 0.0528, "step": 10410 }, { "grad_norm": 0.639885663986206, "learning_rate": 9.82501623369198e-05, "loss": 0.0548, "step": 10420 }, { "grad_norm": 0.3872023820877075, "learning_rate": 9.824473818340574e-05, "loss": 0.0534, "step": 10430 }, { "grad_norm": 0.4241308569908142, "learning_rate": 9.823930578618541e-05, "loss": 0.0471, "step": 10440 }, { "grad_norm": 0.4380960762500763, "learning_rate": 9.823386514618709e-05, "loss": 0.0589, "step": 10450 }, { "grad_norm": 0.4605562388896942, "learning_rate": 9.82284162643404e-05, "loss": 0.0545, "step": 10460 }, { "grad_norm": 0.3777429163455963, "learning_rate": 9.822295914157642e-05, "loss": 0.0539, "step": 10470 }, { "grad_norm": 0.5574368238449097, "learning_rate": 9.821749377882763e-05, "loss": 0.048, "step": 10480 }, { "grad_norm": 0.4114394783973694, "learning_rate": 9.821202017702791e-05, "loss": 0.0479, "step": 10490 }, { "grad_norm": 0.3520287871360779, "learning_rate": 9.820653833711253e-05, "loss": 0.0519, "step": 10500 }, { "grad_norm": 0.48815757036209106, "learning_rate": 9.820104826001822e-05, "loss": 0.0617, "step": 10510 }, { "grad_norm": 0.4426628351211548, "learning_rate": 9.819554994668305e-05, "loss": 0.0454, "step": 10520 }, { "grad_norm": 0.4862055778503418, "learning_rate": 9.819004339804654e-05, "loss": 0.0725, "step": 10530 }, { "grad_norm": 0.4396548271179199, "learning_rate": 9.818452861504961e-05, "loss": 0.0522, "step": 10540 }, { "grad_norm": 0.36903926730155945, "learning_rate": 9.81790055986346e-05, "loss": 0.0437, "step": 10550 }, { "grad_norm": 0.4102110266685486, "learning_rate": 9.817347434974523e-05, "loss": 0.0442, "step": 10560 }, { "grad_norm": 0.5046717524528503, "learning_rate": 9.816793486932664e-05, "loss": 0.0399, "step": 10570 }, { "grad_norm": 0.3397158980369568, "learning_rate": 9.816238715832538e-05, "loss": 0.0491, "step": 10580 }, { "grad_norm": 0.42330047488212585, "learning_rate": 9.815683121768939e-05, "loss": 0.0578, "step": 10590 }, { "grad_norm": 0.4829716384410858, "learning_rate": 9.815126704836804e-05, "loss": 0.063, "step": 10600 }, { "grad_norm": 0.39684638381004333, "learning_rate": 9.81456946513121e-05, "loss": 0.0814, "step": 10610 }, { "grad_norm": 0.6385467648506165, "learning_rate": 9.814011402747373e-05, "loss": 0.0757, "step": 10620 }, { "grad_norm": 0.4740777313709259, "learning_rate": 9.813452517780651e-05, "loss": 0.0537, "step": 10630 }, { "grad_norm": 0.3736051917076111, "learning_rate": 9.81289281032654e-05, "loss": 0.0606, "step": 10640 }, { "grad_norm": 0.3937983810901642, "learning_rate": 9.812332280480683e-05, "loss": 0.06, "step": 10650 }, { "grad_norm": 0.5022880434989929, "learning_rate": 9.811770928338854e-05, "loss": 0.0574, "step": 10660 }, { "grad_norm": 0.5895534157752991, "learning_rate": 9.811208753996979e-05, "loss": 0.0556, "step": 10670 }, { "grad_norm": 0.5459941625595093, "learning_rate": 9.810645757551113e-05, "loss": 0.0457, "step": 10680 }, { "grad_norm": 0.3774429261684418, "learning_rate": 9.810081939097459e-05, "loss": 0.0526, "step": 10690 }, { "grad_norm": 0.5120177268981934, "learning_rate": 9.809517298732356e-05, "loss": 0.074, "step": 10700 }, { "grad_norm": 0.4902516305446625, "learning_rate": 9.80895183655229e-05, "loss": 0.0605, "step": 10710 }, { "grad_norm": 0.47945401072502136, "learning_rate": 9.808385552653877e-05, "loss": 0.052, "step": 10720 }, { "grad_norm": 0.4249904751777649, "learning_rate": 9.807818447133886e-05, "loss": 0.0454, "step": 10730 }, { "grad_norm": 0.4933639466762543, "learning_rate": 9.807250520089215e-05, "loss": 0.0503, "step": 10740 }, { "grad_norm": 0.42845848202705383, "learning_rate": 9.806681771616908e-05, "loss": 0.0611, "step": 10750 }, { "grad_norm": 0.42908889055252075, "learning_rate": 9.80611220181415e-05, "loss": 0.0638, "step": 10760 }, { "grad_norm": 0.6179621815681458, "learning_rate": 9.805541810778264e-05, "loss": 0.0648, "step": 10770 }, { "grad_norm": 0.6002838611602783, "learning_rate": 9.804970598606716e-05, "loss": 0.061, "step": 10780 }, { "grad_norm": 0.39166513085365295, "learning_rate": 9.804398565397106e-05, "loss": 0.0586, "step": 10790 }, { "grad_norm": 0.5489034056663513, "learning_rate": 9.803825711247183e-05, "loss": 0.0612, "step": 10800 }, { "grad_norm": 0.5050615072250366, "learning_rate": 9.803252036254831e-05, "loss": 0.052, "step": 10810 }, { "grad_norm": 0.41446685791015625, "learning_rate": 9.802677540518076e-05, "loss": 0.0585, "step": 10820 }, { "grad_norm": 0.3428248167037964, "learning_rate": 9.802102224135081e-05, "loss": 0.048, "step": 10830 }, { "grad_norm": 0.43777260184288025, "learning_rate": 9.801526087204155e-05, "loss": 0.0645, "step": 10840 }, { "grad_norm": 0.532526969909668, "learning_rate": 9.800949129823743e-05, "loss": 0.0559, "step": 10850 }, { "grad_norm": 0.438697874546051, "learning_rate": 9.80037135209243e-05, "loss": 0.0463, "step": 10860 }, { "grad_norm": 0.5056961178779602, "learning_rate": 9.799792754108946e-05, "loss": 0.0507, "step": 10870 }, { "grad_norm": 0.48475128412246704, "learning_rate": 9.799213335972152e-05, "loss": 0.0532, "step": 10880 }, { "grad_norm": 0.46078479290008545, "learning_rate": 9.798633097781058e-05, "loss": 0.0522, "step": 10890 }, { "grad_norm": 0.4455903470516205, "learning_rate": 9.79805203963481e-05, "loss": 0.0563, "step": 10900 }, { "grad_norm": 0.5325824618339539, "learning_rate": 9.797470161632697e-05, "loss": 0.0506, "step": 10910 }, { "grad_norm": 0.45014166831970215, "learning_rate": 9.796887463874145e-05, "loss": 0.0564, "step": 10920 }, { "grad_norm": 0.38291266560554504, "learning_rate": 9.796303946458718e-05, "loss": 0.0645, "step": 10930 }, { "grad_norm": 0.43569216132164, "learning_rate": 9.795719609486127e-05, "loss": 0.0611, "step": 10940 }, { "grad_norm": 0.5435948371887207, "learning_rate": 9.795134453056219e-05, "loss": 0.0696, "step": 10950 }, { "grad_norm": 0.4018741250038147, "learning_rate": 9.794548477268979e-05, "loss": 0.0506, "step": 10960 }, { "grad_norm": 0.4625226557254791, "learning_rate": 9.793961682224537e-05, "loss": 0.06, "step": 10970 }, { "grad_norm": 0.46033182740211487, "learning_rate": 9.793374068023156e-05, "loss": 0.0583, "step": 10980 }, { "grad_norm": 0.5096797943115234, "learning_rate": 9.792785634765247e-05, "loss": 0.0516, "step": 10990 }, { "grad_norm": 0.4706266522407532, "learning_rate": 9.792196382551357e-05, "loss": 0.0491, "step": 11000 }, { "grad_norm": 0.5515569448471069, "learning_rate": 9.791606311482171e-05, "loss": 0.0522, "step": 11010 }, { "grad_norm": 0.45978865027427673, "learning_rate": 9.791015421658518e-05, "loss": 0.0492, "step": 11020 }, { "grad_norm": 0.5614347457885742, "learning_rate": 9.790423713181362e-05, "loss": 0.0576, "step": 11030 }, { "grad_norm": 0.4792032837867737, "learning_rate": 9.789831186151814e-05, "loss": 0.0524, "step": 11040 }, { "grad_norm": 0.426980197429657, "learning_rate": 9.789237840671118e-05, "loss": 0.0501, "step": 11050 }, { "grad_norm": 0.4140438437461853, "learning_rate": 9.78864367684066e-05, "loss": 0.0559, "step": 11060 }, { "grad_norm": 0.4440847635269165, "learning_rate": 9.788048694761968e-05, "loss": 0.0499, "step": 11070 }, { "grad_norm": 0.4861035645008087, "learning_rate": 9.787452894536709e-05, "loss": 0.0517, "step": 11080 }, { "grad_norm": 0.36333319544792175, "learning_rate": 9.786856276266685e-05, "loss": 0.0499, "step": 11090 }, { "grad_norm": 0.3490222990512848, "learning_rate": 9.786258840053845e-05, "loss": 0.061, "step": 11100 }, { "grad_norm": 0.5229003429412842, "learning_rate": 9.785660586000273e-05, "loss": 0.0614, "step": 11110 }, { "grad_norm": 0.4331909716129303, "learning_rate": 9.785061514208196e-05, "loss": 0.0544, "step": 11120 }, { "grad_norm": 0.44897523522377014, "learning_rate": 9.784461624779977e-05, "loss": 0.0547, "step": 11130 }, { "grad_norm": 0.42480218410491943, "learning_rate": 9.783860917818123e-05, "loss": 0.0491, "step": 11140 }, { "grad_norm": 0.4630144238471985, "learning_rate": 9.783259393425277e-05, "loss": 0.0408, "step": 11150 }, { "grad_norm": 0.5039328336715698, "learning_rate": 9.782657051704221e-05, "loss": 0.0497, "step": 11160 }, { "grad_norm": 0.42655256390571594, "learning_rate": 9.782053892757883e-05, "loss": 0.0536, "step": 11170 }, { "grad_norm": 0.45102640986442566, "learning_rate": 9.781449916689324e-05, "loss": 0.0469, "step": 11180 }, { "grad_norm": 0.4327525198459625, "learning_rate": 9.780845123601746e-05, "loss": 0.0546, "step": 11190 }, { "grad_norm": 0.4179236590862274, "learning_rate": 9.780239513598492e-05, "loss": 0.0493, "step": 11200 }, { "grad_norm": 0.4315072000026703, "learning_rate": 9.779633086783047e-05, "loss": 0.05, "step": 11210 }, { "grad_norm": 0.3781377673149109, "learning_rate": 9.779025843259031e-05, "loss": 0.0491, "step": 11220 }, { "grad_norm": 0.5366251468658447, "learning_rate": 9.778417783130204e-05, "loss": 0.0636, "step": 11230 }, { "grad_norm": 0.5226191282272339, "learning_rate": 9.777808906500468e-05, "loss": 0.0543, "step": 11240 }, { "grad_norm": 0.4193602204322815, "learning_rate": 9.777199213473862e-05, "loss": 0.0398, "step": 11250 }, { "grad_norm": 0.466550350189209, "learning_rate": 9.77658870415457e-05, "loss": 0.0456, "step": 11260 }, { "grad_norm": 0.4048946499824524, "learning_rate": 9.775977378646906e-05, "loss": 0.0478, "step": 11270 }, { "grad_norm": 0.5729438662528992, "learning_rate": 9.775365237055331e-05, "loss": 0.061, "step": 11280 }, { "grad_norm": 0.4649139642715454, "learning_rate": 9.774752279484445e-05, "loss": 0.0539, "step": 11290 }, { "grad_norm": 0.4380496144294739, "learning_rate": 9.774138506038984e-05, "loss": 0.051, "step": 11300 }, { "grad_norm": 0.5854820013046265, "learning_rate": 9.773523916823826e-05, "loss": 0.055, "step": 11310 }, { "grad_norm": 0.4098604917526245, "learning_rate": 9.772908511943986e-05, "loss": 0.0522, "step": 11320 }, { "grad_norm": 0.5033850073814392, "learning_rate": 9.77229229150462e-05, "loss": 0.0507, "step": 11330 }, { "grad_norm": 0.4306310713291168, "learning_rate": 9.771675255611024e-05, "loss": 0.0501, "step": 11340 }, { "grad_norm": 0.5236417055130005, "learning_rate": 9.771057404368632e-05, "loss": 0.0796, "step": 11350 }, { "grad_norm": 0.472854346036911, "learning_rate": 9.770438737883018e-05, "loss": 0.0561, "step": 11360 }, { "grad_norm": 0.44846779108047485, "learning_rate": 9.769819256259898e-05, "loss": 0.0454, "step": 11370 }, { "grad_norm": 0.4503997266292572, "learning_rate": 9.769198959605119e-05, "loss": 0.0507, "step": 11380 }, { "grad_norm": 0.4217078387737274, "learning_rate": 9.768577848024678e-05, "loss": 0.0533, "step": 11390 }, { "grad_norm": 0.3887816369533539, "learning_rate": 9.767955921624702e-05, "loss": 0.061, "step": 11400 }, { "grad_norm": 0.6489797830581665, "learning_rate": 9.767333180511465e-05, "loss": 0.0476, "step": 11410 }, { "grad_norm": 0.38814496994018555, "learning_rate": 9.766709624791373e-05, "loss": 0.0611, "step": 11420 }, { "grad_norm": 0.4384576380252838, "learning_rate": 9.766085254570975e-05, "loss": 0.0557, "step": 11430 }, { "grad_norm": 0.5481582283973694, "learning_rate": 9.76546006995696e-05, "loss": 0.049, "step": 11440 }, { "grad_norm": 0.6249732971191406, "learning_rate": 9.764834071056155e-05, "loss": 0.0566, "step": 11450 }, { "grad_norm": 0.3788849413394928, "learning_rate": 9.764207257975526e-05, "loss": 0.0605, "step": 11460 }, { "grad_norm": 0.6036906838417053, "learning_rate": 9.763579630822179e-05, "loss": 0.0572, "step": 11470 }, { "grad_norm": 0.6515277624130249, "learning_rate": 9.762951189703356e-05, "loss": 0.0617, "step": 11480 }, { "grad_norm": 0.41330552101135254, "learning_rate": 9.762321934726442e-05, "loss": 0.0673, "step": 11490 }, { "grad_norm": 0.3852963447570801, "learning_rate": 9.761691865998959e-05, "loss": 0.0559, "step": 11500 }, { "grad_norm": 0.4374261498451233, "learning_rate": 9.76106098362857e-05, "loss": 0.0552, "step": 11510 }, { "grad_norm": 0.47945332527160645, "learning_rate": 9.760429287723072e-05, "loss": 0.0594, "step": 11520 }, { "grad_norm": 0.46801629662513733, "learning_rate": 9.759796778390406e-05, "loss": 0.052, "step": 11530 }, { "grad_norm": 0.30105578899383545, "learning_rate": 9.759163455738653e-05, "loss": 0.0545, "step": 11540 }, { "grad_norm": 0.492861270904541, "learning_rate": 9.75852931987603e-05, "loss": 0.0534, "step": 11550 }, { "grad_norm": 0.4163826107978821, "learning_rate": 9.757894370910891e-05, "loss": 0.0488, "step": 11560 }, { "grad_norm": 0.37383052706718445, "learning_rate": 9.757258608951733e-05, "loss": 0.0471, "step": 11570 }, { "grad_norm": 0.5077378749847412, "learning_rate": 9.75662203410719e-05, "loss": 0.0587, "step": 11580 }, { "grad_norm": 0.39263200759887695, "learning_rate": 9.755984646486034e-05, "loss": 0.0546, "step": 11590 }, { "grad_norm": 0.4850635528564453, "learning_rate": 9.75534644619718e-05, "loss": 0.0513, "step": 11600 }, { "grad_norm": 0.3812474310398102, "learning_rate": 9.754707433349676e-05, "loss": 0.0639, "step": 11610 }, { "grad_norm": 0.5680201649665833, "learning_rate": 9.754067608052715e-05, "loss": 0.0633, "step": 11620 }, { "grad_norm": 0.5344335436820984, "learning_rate": 9.753426970415622e-05, "loss": 0.0733, "step": 11630 }, { "grad_norm": 0.5747514963150024, "learning_rate": 9.752785520547868e-05, "loss": 0.054, "step": 11640 }, { "grad_norm": 0.48369768261909485, "learning_rate": 9.752143258559056e-05, "loss": 0.0549, "step": 11650 }, { "grad_norm": 0.7218174338340759, "learning_rate": 9.751500184558933e-05, "loss": 0.0526, "step": 11660 }, { "grad_norm": 0.4198892116546631, "learning_rate": 9.750856298657383e-05, "loss": 0.0436, "step": 11670 }, { "grad_norm": 0.875187873840332, "learning_rate": 9.750211600964428e-05, "loss": 0.06, "step": 11680 }, { "grad_norm": 0.6030828356742859, "learning_rate": 9.749566091590226e-05, "loss": 0.0662, "step": 11690 }, { "grad_norm": 0.5193609595298767, "learning_rate": 9.748919770645083e-05, "loss": 0.0671, "step": 11700 }, { "grad_norm": 0.3545781672000885, "learning_rate": 9.748272638239432e-05, "loss": 0.0539, "step": 11710 }, { "grad_norm": 0.4433625042438507, "learning_rate": 9.747624694483855e-05, "loss": 0.0537, "step": 11720 }, { "grad_norm": 0.4214731752872467, "learning_rate": 9.746975939489065e-05, "loss": 0.0533, "step": 11730 }, { "grad_norm": 0.4746861457824707, "learning_rate": 9.746326373365918e-05, "loss": 0.0471, "step": 11740 }, { "grad_norm": 0.4226950407028198, "learning_rate": 9.745675996225403e-05, "loss": 0.0586, "step": 11750 }, { "grad_norm": 0.33744102716445923, "learning_rate": 9.745024808178657e-05, "loss": 0.0493, "step": 11760 }, { "grad_norm": 0.49541983008384705, "learning_rate": 9.744372809336947e-05, "loss": 0.0514, "step": 11770 }, { "grad_norm": 0.5797685980796814, "learning_rate": 9.743719999811682e-05, "loss": 0.048, "step": 11780 }, { "grad_norm": 0.2737135887145996, "learning_rate": 9.743066379714412e-05, "loss": 0.0453, "step": 11790 }, { "grad_norm": 0.31091490387916565, "learning_rate": 9.74241194915682e-05, "loss": 0.0473, "step": 11800 }, { "grad_norm": 0.42853572964668274, "learning_rate": 9.741756708250731e-05, "loss": 0.0635, "step": 11810 }, { "grad_norm": 0.3794018030166626, "learning_rate": 9.741100657108109e-05, "loss": 0.0493, "step": 11820 }, { "grad_norm": 0.47025156021118164, "learning_rate": 9.740443795841054e-05, "loss": 0.0671, "step": 11830 }, { "grad_norm": 0.4633864462375641, "learning_rate": 9.739786124561805e-05, "loss": 0.0459, "step": 11840 }, { "grad_norm": 0.3305177092552185, "learning_rate": 9.73912764338274e-05, "loss": 0.0484, "step": 11850 }, { "grad_norm": 0.46766233444213867, "learning_rate": 9.738468352416377e-05, "loss": 0.0589, "step": 11860 }, { "grad_norm": 0.4221186935901642, "learning_rate": 9.737808251775369e-05, "loss": 0.0528, "step": 11870 }, { "grad_norm": 0.42215847969055176, "learning_rate": 9.737147341572512e-05, "loss": 0.0607, "step": 11880 }, { "grad_norm": 0.5035384297370911, "learning_rate": 9.736485621920735e-05, "loss": 0.05, "step": 11890 }, { "grad_norm": 0.5563461184501648, "learning_rate": 9.735823092933108e-05, "loss": 0.0558, "step": 11900 }, { "grad_norm": 0.5849642753601074, "learning_rate": 9.735159754722838e-05, "loss": 0.0659, "step": 11910 }, { "grad_norm": 0.49081501364707947, "learning_rate": 9.734495607403275e-05, "loss": 0.0499, "step": 11920 }, { "grad_norm": 0.4273568391799927, "learning_rate": 9.733830651087901e-05, "loss": 0.0543, "step": 11930 }, { "grad_norm": 0.4155169427394867, "learning_rate": 9.733164885890338e-05, "loss": 0.0465, "step": 11940 }, { "grad_norm": 0.3141757547855377, "learning_rate": 9.732498311924349e-05, "loss": 0.0362, "step": 11950 }, { "grad_norm": 0.36699628829956055, "learning_rate": 9.731830929303833e-05, "loss": 0.0439, "step": 11960 }, { "grad_norm": 0.9209628701210022, "learning_rate": 9.731162738142827e-05, "loss": 0.0661, "step": 11970 }, { "grad_norm": 0.48834550380706787, "learning_rate": 9.730493738555506e-05, "loss": 0.0552, "step": 11980 }, { "grad_norm": 0.3321431577205658, "learning_rate": 9.729823930656186e-05, "loss": 0.055, "step": 11990 }, { "grad_norm": 0.46223577857017517, "learning_rate": 9.729153314559316e-05, "loss": 0.0616, "step": 12000 }, { "grad_norm": 0.44230690598487854, "learning_rate": 9.728481890379486e-05, "loss": 0.0544, "step": 12010 }, { "grad_norm": 0.4304066300392151, "learning_rate": 9.727809658231428e-05, "loss": 0.0516, "step": 12020 }, { "grad_norm": 0.3280954658985138, "learning_rate": 9.727136618230003e-05, "loss": 0.0477, "step": 12030 }, { "grad_norm": 0.5380178689956665, "learning_rate": 9.726462770490219e-05, "loss": 0.0592, "step": 12040 }, { "grad_norm": 0.421547532081604, "learning_rate": 9.725788115127214e-05, "loss": 0.0574, "step": 12050 }, { "grad_norm": 0.45991602540016174, "learning_rate": 9.725112652256274e-05, "loss": 0.0504, "step": 12060 }, { "grad_norm": 0.42021453380584717, "learning_rate": 9.724436381992812e-05, "loss": 0.069, "step": 12070 }, { "grad_norm": 0.5344069600105286, "learning_rate": 9.723759304452387e-05, "loss": 0.0472, "step": 12080 }, { "grad_norm": 0.49333029985427856, "learning_rate": 9.72308141975069e-05, "loss": 0.0461, "step": 12090 }, { "grad_norm": 0.4343907833099365, "learning_rate": 9.722402728003557e-05, "loss": 0.0649, "step": 12100 }, { "grad_norm": 0.48773109912872314, "learning_rate": 9.721723229326953e-05, "loss": 0.0524, "step": 12110 }, { "grad_norm": 0.5022515654563904, "learning_rate": 9.721042923836992e-05, "loss": 0.0439, "step": 12120 }, { "grad_norm": 0.3573954105377197, "learning_rate": 9.720361811649914e-05, "loss": 0.0477, "step": 12130 }, { "grad_norm": 0.4760834276676178, "learning_rate": 9.719679892882106e-05, "loss": 0.0462, "step": 12140 }, { "grad_norm": 0.5172815918922424, "learning_rate": 9.718997167650085e-05, "loss": 0.0486, "step": 12150 }, { "grad_norm": 0.29308757185935974, "learning_rate": 9.718313636070515e-05, "loss": 0.0413, "step": 12160 }, { "grad_norm": 0.5028605461120605, "learning_rate": 9.717629298260192e-05, "loss": 0.0489, "step": 12170 }, { "grad_norm": 0.5792133212089539, "learning_rate": 9.716944154336047e-05, "loss": 0.0616, "step": 12180 }, { "grad_norm": 0.367824912071228, "learning_rate": 9.716258204415157e-05, "loss": 0.0465, "step": 12190 }, { "grad_norm": 0.3860955834388733, "learning_rate": 9.715571448614728e-05, "loss": 0.0581, "step": 12200 }, { "grad_norm": 0.3648185133934021, "learning_rate": 9.71488388705211e-05, "loss": 0.0691, "step": 12210 }, { "grad_norm": 0.5638396143913269, "learning_rate": 9.714195519844788e-05, "loss": 0.0411, "step": 12220 }, { "grad_norm": 0.49782606959342957, "learning_rate": 9.713506347110386e-05, "loss": 0.0585, "step": 12230 }, { "grad_norm": 0.4258947968482971, "learning_rate": 9.712816368966662e-05, "loss": 0.049, "step": 12240 }, { "grad_norm": 0.3458767831325531, "learning_rate": 9.712125585531517e-05, "loss": 0.068, "step": 12250 }, { "grad_norm": 0.5135696530342102, "learning_rate": 9.711433996922988e-05, "loss": 0.0686, "step": 12260 }, { "grad_norm": 0.5062606334686279, "learning_rate": 9.710741603259245e-05, "loss": 0.0595, "step": 12270 }, { "grad_norm": 0.4220620095729828, "learning_rate": 9.710048404658603e-05, "loss": 0.0498, "step": 12280 }, { "grad_norm": 0.36396729946136475, "learning_rate": 9.709354401239508e-05, "loss": 0.0522, "step": 12290 }, { "grad_norm": 0.5068788528442383, "learning_rate": 9.708659593120546e-05, "loss": 0.0426, "step": 12300 }, { "grad_norm": 0.4490620493888855, "learning_rate": 9.707963980420443e-05, "loss": 0.0452, "step": 12310 }, { "grad_norm": 0.603279709815979, "learning_rate": 9.707267563258058e-05, "loss": 0.0583, "step": 12320 }, { "grad_norm": 0.3576047718524933, "learning_rate": 9.70657034175239e-05, "loss": 0.0514, "step": 12330 }, { "grad_norm": 0.6046767830848694, "learning_rate": 9.705872316022577e-05, "loss": 0.0598, "step": 12340 }, { "grad_norm": 0.5193299055099487, "learning_rate": 9.705173486187891e-05, "loss": 0.054, "step": 12350 }, { "grad_norm": 0.4384704530239105, "learning_rate": 9.704473852367741e-05, "loss": 0.0593, "step": 12360 }, { "grad_norm": 0.41912996768951416, "learning_rate": 9.70377341468168e-05, "loss": 0.0447, "step": 12370 }, { "grad_norm": 0.41253185272216797, "learning_rate": 9.703072173249389e-05, "loss": 0.0494, "step": 12380 }, { "grad_norm": 0.5555065274238586, "learning_rate": 9.702370128190693e-05, "loss": 0.0526, "step": 12390 }, { "grad_norm": 0.47067633271217346, "learning_rate": 9.701667279625552e-05, "loss": 0.0593, "step": 12400 }, { "grad_norm": 0.5501346588134766, "learning_rate": 9.700963627674065e-05, "loss": 0.0602, "step": 12410 }, { "grad_norm": 0.4453848600387573, "learning_rate": 9.700259172456466e-05, "loss": 0.0497, "step": 12420 }, { "grad_norm": 0.541316568851471, "learning_rate": 9.699553914093124e-05, "loss": 0.0605, "step": 12430 }, { "grad_norm": 0.5831606984138489, "learning_rate": 9.698847852704553e-05, "loss": 0.0615, "step": 12440 }, { "grad_norm": 0.5179435014724731, "learning_rate": 9.6981409884114e-05, "loss": 0.0498, "step": 12450 }, { "grad_norm": 0.4920286536216736, "learning_rate": 9.697433321334443e-05, "loss": 0.0563, "step": 12460 }, { "grad_norm": 0.34904879331588745, "learning_rate": 9.696724851594607e-05, "loss": 0.0579, "step": 12470 }, { "grad_norm": 0.439928263425827, "learning_rate": 9.696015579312952e-05, "loss": 0.0433, "step": 12480 }, { "grad_norm": 0.3364410400390625, "learning_rate": 9.695305504610668e-05, "loss": 0.0446, "step": 12490 }, { "grad_norm": 0.41461825370788574, "learning_rate": 9.694594627609092e-05, "loss": 0.0503, "step": 12500 }, { "grad_norm": 0.4841984510421753, "learning_rate": 9.693882948429691e-05, "loss": 0.0635, "step": 12510 }, { "grad_norm": 0.4082094132900238, "learning_rate": 9.693170467194071e-05, "loss": 0.0408, "step": 12520 }, { "grad_norm": 0.30651476979255676, "learning_rate": 9.692457184023977e-05, "loss": 0.0455, "step": 12530 }, { "grad_norm": 0.3186545670032501, "learning_rate": 9.691743099041291e-05, "loss": 0.0508, "step": 12540 }, { "grad_norm": 0.4322033226490021, "learning_rate": 9.691028212368027e-05, "loss": 0.0486, "step": 12550 }, { "grad_norm": 0.4615146219730377, "learning_rate": 9.690312524126342e-05, "loss": 0.0414, "step": 12560 }, { "grad_norm": 0.42286041378974915, "learning_rate": 9.689596034438527e-05, "loss": 0.0478, "step": 12570 }, { "grad_norm": 0.29536980390548706, "learning_rate": 9.688878743427012e-05, "loss": 0.0462, "step": 12580 }, { "grad_norm": 0.37473949790000916, "learning_rate": 9.688160651214359e-05, "loss": 0.0543, "step": 12590 }, { "grad_norm": 0.4055265486240387, "learning_rate": 9.687441757923273e-05, "loss": 0.0485, "step": 12600 }, { "grad_norm": 0.3816889524459839, "learning_rate": 9.68672206367659e-05, "loss": 0.0457, "step": 12610 }, { "grad_norm": 0.4035714864730835, "learning_rate": 9.686001568597291e-05, "loss": 0.0462, "step": 12620 }, { "grad_norm": 0.43110114336013794, "learning_rate": 9.685280272808486e-05, "loss": 0.0431, "step": 12630 }, { "grad_norm": 0.5270542502403259, "learning_rate": 9.684558176433424e-05, "loss": 0.0591, "step": 12640 }, { "grad_norm": 0.5458003282546997, "learning_rate": 9.683835279595495e-05, "loss": 0.0458, "step": 12650 }, { "grad_norm": 0.47334447503089905, "learning_rate": 9.683111582418216e-05, "loss": 0.055, "step": 12660 }, { "grad_norm": 0.4050378203392029, "learning_rate": 9.682387085025254e-05, "loss": 0.045, "step": 12670 }, { "grad_norm": 0.501141369342804, "learning_rate": 9.681661787540401e-05, "loss": 0.058, "step": 12680 }, { "grad_norm": 0.44272878766059875, "learning_rate": 9.680935690087593e-05, "loss": 0.0528, "step": 12690 }, { "grad_norm": 0.5325649380683899, "learning_rate": 9.680208792790901e-05, "loss": 0.0585, "step": 12700 }, { "grad_norm": 0.3738967180252075, "learning_rate": 9.679481095774529e-05, "loss": 0.0459, "step": 12710 }, { "grad_norm": 0.5468589067459106, "learning_rate": 9.678752599162822e-05, "loss": 0.0528, "step": 12720 }, { "grad_norm": 0.5624083280563354, "learning_rate": 9.678023303080259e-05, "loss": 0.047, "step": 12730 }, { "grad_norm": 0.323495090007782, "learning_rate": 9.677293207651459e-05, "loss": 0.051, "step": 12740 }, { "grad_norm": 0.39185425639152527, "learning_rate": 9.676562313001173e-05, "loss": 0.046, "step": 12750 }, { "grad_norm": 0.523337721824646, "learning_rate": 9.675830619254293e-05, "loss": 0.0656, "step": 12760 }, { "grad_norm": 0.5721856355667114, "learning_rate": 9.675098126535843e-05, "loss": 0.0611, "step": 12770 }, { "grad_norm": 0.532538115978241, "learning_rate": 9.674364834970988e-05, "loss": 0.0594, "step": 12780 }, { "grad_norm": 0.44536522030830383, "learning_rate": 9.673630744685028e-05, "loss": 0.0544, "step": 12790 }, { "grad_norm": 0.43662169575691223, "learning_rate": 9.672895855803397e-05, "loss": 0.0544, "step": 12800 }, { "grad_norm": 0.3055674135684967, "learning_rate": 9.672160168451667e-05, "loss": 0.0472, "step": 12810 }, { "grad_norm": 0.58818519115448, "learning_rate": 9.671423682755549e-05, "loss": 0.0611, "step": 12820 }, { "grad_norm": 0.6531535387039185, "learning_rate": 9.670686398840888e-05, "loss": 0.066, "step": 12830 }, { "grad_norm": 0.4382178485393524, "learning_rate": 9.669948316833664e-05, "loss": 0.0477, "step": 12840 }, { "grad_norm": 0.3592139482498169, "learning_rate": 9.669209436859997e-05, "loss": 0.0471, "step": 12850 }, { "grad_norm": 0.3909594714641571, "learning_rate": 9.66846975904614e-05, "loss": 0.0474, "step": 12860 }, { "grad_norm": 0.49002400040626526, "learning_rate": 9.667729283518483e-05, "loss": 0.0444, "step": 12870 }, { "grad_norm": 0.5567033886909485, "learning_rate": 9.666988010403557e-05, "loss": 0.0426, "step": 12880 }, { "grad_norm": 0.3334484398365021, "learning_rate": 9.66624593982802e-05, "loss": 0.0423, "step": 12890 }, { "grad_norm": 0.34237703680992126, "learning_rate": 9.665503071918675e-05, "loss": 0.0571, "step": 12900 }, { "grad_norm": 0.42947208881378174, "learning_rate": 9.664759406802456e-05, "loss": 0.0433, "step": 12910 }, { "grad_norm": 0.46343186497688293, "learning_rate": 9.664014944606437e-05, "loss": 0.038, "step": 12920 }, { "grad_norm": 0.36658141016960144, "learning_rate": 9.663269685457822e-05, "loss": 0.0528, "step": 12930 }, { "grad_norm": 0.4054850935935974, "learning_rate": 9.662523629483962e-05, "loss": 0.0493, "step": 12940 }, { "grad_norm": 0.3625718057155609, "learning_rate": 9.661776776812333e-05, "loss": 0.0398, "step": 12950 }, { "grad_norm": 0.48869410157203674, "learning_rate": 9.661029127570553e-05, "loss": 0.0455, "step": 12960 }, { "grad_norm": 0.4810537099838257, "learning_rate": 9.660280681886373e-05, "loss": 0.0491, "step": 12970 }, { "grad_norm": 0.41480812430381775, "learning_rate": 9.659531439887685e-05, "loss": 0.0374, "step": 12980 }, { "grad_norm": 0.5305384993553162, "learning_rate": 9.658781401702511e-05, "loss": 0.0561, "step": 12990 }, { "grad_norm": 0.514242947101593, "learning_rate": 9.658030567459015e-05, "loss": 0.045, "step": 13000 }, { "grad_norm": 0.4086378514766693, "learning_rate": 9.65727893728549e-05, "loss": 0.0664, "step": 13010 }, { "grad_norm": 0.5261985063552856, "learning_rate": 9.656526511310375e-05, "loss": 0.0609, "step": 13020 }, { "grad_norm": 0.3061433434486389, "learning_rate": 9.655773289662233e-05, "loss": 0.058, "step": 13030 }, { "grad_norm": 0.5772855281829834, "learning_rate": 9.655019272469772e-05, "loss": 0.0573, "step": 13040 }, { "grad_norm": 0.33871641755104065, "learning_rate": 9.654264459861832e-05, "loss": 0.0417, "step": 13050 }, { "grad_norm": 0.32237252593040466, "learning_rate": 9.653508851967391e-05, "loss": 0.0463, "step": 13060 }, { "grad_norm": 0.5205962061882019, "learning_rate": 9.65275244891556e-05, "loss": 0.0571, "step": 13070 }, { "grad_norm": 0.4214637577533722, "learning_rate": 9.651995250835591e-05, "loss": 0.0515, "step": 13080 }, { "grad_norm": 0.49197718501091003, "learning_rate": 9.651237257856862e-05, "loss": 0.0506, "step": 13090 }, { "grad_norm": 0.5714525580406189, "learning_rate": 9.6504784701089e-05, "loss": 0.0535, "step": 13100 }, { "grad_norm": 0.41109129786491394, "learning_rate": 9.649718887721357e-05, "loss": 0.0512, "step": 13110 }, { "grad_norm": 0.6182546615600586, "learning_rate": 9.648958510824028e-05, "loss": 0.0494, "step": 13120 }, { "grad_norm": 0.4380488991737366, "learning_rate": 9.648197339546837e-05, "loss": 0.0574, "step": 13130 }, { "grad_norm": 0.41569435596466064, "learning_rate": 9.647435374019851e-05, "loss": 0.0502, "step": 13140 }, { "grad_norm": 0.5048779845237732, "learning_rate": 9.646672614373266e-05, "loss": 0.0505, "step": 13150 }, { "grad_norm": 0.7061945796012878, "learning_rate": 9.645909060737418e-05, "loss": 0.0461, "step": 13160 }, { "grad_norm": 0.3518708348274231, "learning_rate": 9.645144713242778e-05, "loss": 0.0438, "step": 13170 }, { "grad_norm": 0.30560722947120667, "learning_rate": 9.64437957201995e-05, "loss": 0.0515, "step": 13180 }, { "grad_norm": 0.4734482765197754, "learning_rate": 9.643613637199678e-05, "loss": 0.0488, "step": 13190 }, { "grad_norm": 0.5642409324645996, "learning_rate": 9.642846908912839e-05, "loss": 0.0654, "step": 13200 }, { "grad_norm": 0.4294516146183014, "learning_rate": 9.642079387290444e-05, "loss": 0.0437, "step": 13210 }, { "grad_norm": 0.4072403609752655, "learning_rate": 9.641311072463644e-05, "loss": 0.047, "step": 13220 }, { "grad_norm": 0.5891801714897156, "learning_rate": 9.640541964563722e-05, "loss": 0.0557, "step": 13230 }, { "grad_norm": 0.521308183670044, "learning_rate": 9.639772063722096e-05, "loss": 0.0615, "step": 13240 }, { "grad_norm": 0.3392868638038635, "learning_rate": 9.639001370070324e-05, "loss": 0.0447, "step": 13250 }, { "grad_norm": 0.4565582573413849, "learning_rate": 9.638229883740095e-05, "loss": 0.0466, "step": 13260 }, { "grad_norm": 0.4437249004840851, "learning_rate": 9.637457604863233e-05, "loss": 0.0428, "step": 13270 }, { "grad_norm": 0.39833447337150574, "learning_rate": 9.636684533571703e-05, "loss": 0.0475, "step": 13280 }, { "grad_norm": 0.3731980621814728, "learning_rate": 9.635910669997599e-05, "loss": 0.0519, "step": 13290 }, { "grad_norm": 1.2648110389709473, "learning_rate": 9.635136014273154e-05, "loss": 0.0545, "step": 13300 }, { "grad_norm": 0.46464046835899353, "learning_rate": 9.634360566530735e-05, "loss": 0.0579, "step": 13310 }, { "grad_norm": 0.5143049359321594, "learning_rate": 9.633584326902845e-05, "loss": 0.0755, "step": 13320 }, { "grad_norm": 0.46410951018333435, "learning_rate": 9.632807295522124e-05, "loss": 0.0522, "step": 13330 }, { "grad_norm": 0.4262242615222931, "learning_rate": 9.632029472521342e-05, "loss": 0.0514, "step": 13340 }, { "grad_norm": 0.3724451959133148, "learning_rate": 9.631250858033409e-05, "loss": 0.0479, "step": 13350 }, { "grad_norm": 0.48451507091522217, "learning_rate": 9.630471452191371e-05, "loss": 0.0527, "step": 13360 }, { "grad_norm": 0.3035520911216736, "learning_rate": 9.629691255128405e-05, "loss": 0.0584, "step": 13370 }, { "grad_norm": 0.5587412118911743, "learning_rate": 9.628910266977825e-05, "loss": 0.0573, "step": 13380 }, { "grad_norm": 0.4211120903491974, "learning_rate": 9.628128487873083e-05, "loss": 0.0528, "step": 13390 }, { "grad_norm": 0.3905750811100006, "learning_rate": 9.627345917947761e-05, "loss": 0.0579, "step": 13400 }, { "grad_norm": 0.42304617166519165, "learning_rate": 9.626562557335579e-05, "loss": 0.0507, "step": 13410 }, { "grad_norm": 0.8138939738273621, "learning_rate": 9.625778406170393e-05, "loss": 0.0515, "step": 13420 }, { "grad_norm": 0.5270801186561584, "learning_rate": 9.624993464586193e-05, "loss": 0.0484, "step": 13430 }, { "grad_norm": 0.43407970666885376, "learning_rate": 9.624207732717105e-05, "loss": 0.0528, "step": 13440 }, { "grad_norm": 0.5655791759490967, "learning_rate": 9.623421210697386e-05, "loss": 0.0604, "step": 13450 }, { "grad_norm": 0.5369436740875244, "learning_rate": 9.622633898661434e-05, "loss": 0.0592, "step": 13460 }, { "grad_norm": 0.3910553753376007, "learning_rate": 9.621845796743778e-05, "loss": 0.0644, "step": 13470 }, { "grad_norm": 0.35257238149642944, "learning_rate": 9.621056905079082e-05, "loss": 0.0572, "step": 13480 }, { "grad_norm": 0.45205917954444885, "learning_rate": 9.620267223802149e-05, "loss": 0.0573, "step": 13490 }, { "grad_norm": 0.5291292071342468, "learning_rate": 9.619476753047911e-05, "loss": 0.0537, "step": 13500 }, { "grad_norm": 0.46953970193862915, "learning_rate": 9.618685492951438e-05, "loss": 0.0623, "step": 13510 }, { "grad_norm": 0.4353081285953522, "learning_rate": 9.617893443647938e-05, "loss": 0.0642, "step": 13520 }, { "grad_norm": 0.5669013857841492, "learning_rate": 9.617100605272746e-05, "loss": 0.0642, "step": 13530 }, { "grad_norm": 0.3810495436191559, "learning_rate": 9.616306977961338e-05, "loss": 0.0575, "step": 13540 }, { "grad_norm": 0.431448370218277, "learning_rate": 9.615512561849326e-05, "loss": 0.0479, "step": 13550 }, { "grad_norm": 0.4083310067653656, "learning_rate": 9.61471735707245e-05, "loss": 0.0538, "step": 13560 }, { "grad_norm": 0.4097582697868347, "learning_rate": 9.613921363766592e-05, "loss": 0.0438, "step": 13570 }, { "grad_norm": 0.4545203149318695, "learning_rate": 9.613124582067763e-05, "loss": 0.0522, "step": 13580 }, { "grad_norm": 0.5354859828948975, "learning_rate": 9.612327012112112e-05, "loss": 0.0606, "step": 13590 }, { "grad_norm": 0.4819281995296478, "learning_rate": 9.611528654035921e-05, "loss": 0.0719, "step": 13600 }, { "grad_norm": 0.356118381023407, "learning_rate": 9.610729507975611e-05, "loss": 0.047, "step": 13610 }, { "grad_norm": 0.464885950088501, "learning_rate": 9.609929574067731e-05, "loss": 0.0662, "step": 13620 }, { "grad_norm": 0.412220299243927, "learning_rate": 9.609128852448967e-05, "loss": 0.0442, "step": 13630 }, { "grad_norm": 0.30138301849365234, "learning_rate": 9.608327343256143e-05, "loss": 0.0488, "step": 13640 }, { "grad_norm": 0.45272794365882874, "learning_rate": 9.607525046626216e-05, "loss": 0.0474, "step": 13650 }, { "grad_norm": 0.43684935569763184, "learning_rate": 9.606721962696272e-05, "loss": 0.0575, "step": 13660 }, { "grad_norm": 0.467344731092453, "learning_rate": 9.60591809160354e-05, "loss": 0.0509, "step": 13670 }, { "grad_norm": 0.3950978219509125, "learning_rate": 9.605113433485378e-05, "loss": 0.0512, "step": 13680 }, { "grad_norm": 0.6296749114990234, "learning_rate": 9.604307988479279e-05, "loss": 0.048, "step": 13690 }, { "grad_norm": 0.44861531257629395, "learning_rate": 9.603501756722876e-05, "loss": 0.0539, "step": 13700 }, { "grad_norm": 0.4090619385242462, "learning_rate": 9.602694738353927e-05, "loss": 0.0402, "step": 13710 }, { "grad_norm": 0.3763015866279602, "learning_rate": 9.601886933510331e-05, "loss": 0.0412, "step": 13720 }, { "grad_norm": 0.4857856035232544, "learning_rate": 9.60107834233012e-05, "loss": 0.0409, "step": 13730 }, { "grad_norm": 0.4102412164211273, "learning_rate": 9.60026896495146e-05, "loss": 0.0434, "step": 13740 }, { "grad_norm": 0.41359561681747437, "learning_rate": 9.599458801512652e-05, "loss": 0.0477, "step": 13750 }, { "grad_norm": 0.515537440776825, "learning_rate": 9.598647852152129e-05, "loss": 0.0686, "step": 13760 }, { "grad_norm": 0.46628105640411377, "learning_rate": 9.597836117008462e-05, "loss": 0.0629, "step": 13770 }, { "grad_norm": 0.4019376337528229, "learning_rate": 9.597023596220356e-05, "loss": 0.0432, "step": 13780 }, { "grad_norm": 0.5490701198577881, "learning_rate": 9.596210289926643e-05, "loss": 0.0428, "step": 13790 }, { "grad_norm": 0.4447329044342041, "learning_rate": 9.5953961982663e-05, "loss": 0.0374, "step": 13800 }, { "grad_norm": 0.37694066762924194, "learning_rate": 9.594581321378431e-05, "loss": 0.0462, "step": 13810 }, { "grad_norm": 0.3436850309371948, "learning_rate": 9.593765659402276e-05, "loss": 0.0484, "step": 13820 }, { "grad_norm": 0.6247482299804688, "learning_rate": 9.59294921247721e-05, "loss": 0.0516, "step": 13830 }, { "grad_norm": 0.5178403854370117, "learning_rate": 9.59213198074274e-05, "loss": 0.0565, "step": 13840 }, { "grad_norm": 0.4851975440979004, "learning_rate": 9.59131396433851e-05, "loss": 0.0404, "step": 13850 }, { "grad_norm": 0.4609600305557251, "learning_rate": 9.590495163404297e-05, "loss": 0.0661, "step": 13860 }, { "grad_norm": 0.46096372604370117, "learning_rate": 9.589675578080009e-05, "loss": 0.0606, "step": 13870 }, { "grad_norm": 0.4564734101295471, "learning_rate": 9.588855208505694e-05, "loss": 0.0494, "step": 13880 }, { "grad_norm": 0.653256356716156, "learning_rate": 9.588034054821529e-05, "loss": 0.0658, "step": 13890 }, { "grad_norm": 0.4675126373767853, "learning_rate": 9.587212117167826e-05, "loss": 0.0416, "step": 13900 }, { "grad_norm": 0.395079106092453, "learning_rate": 9.586389395685033e-05, "loss": 0.0444, "step": 13910 }, { "grad_norm": 0.5113591551780701, "learning_rate": 9.585565890513733e-05, "loss": 0.0586, "step": 13920 }, { "grad_norm": 0.41459548473358154, "learning_rate": 9.584741601794636e-05, "loss": 0.045, "step": 13930 }, { "grad_norm": 0.5632772445678711, "learning_rate": 9.58391652966859e-05, "loss": 0.0554, "step": 13940 }, { "grad_norm": 0.541700541973114, "learning_rate": 9.583090674276583e-05, "loss": 0.0483, "step": 13950 }, { "grad_norm": 0.3829869329929352, "learning_rate": 9.582264035759726e-05, "loss": 0.0444, "step": 13960 }, { "grad_norm": 0.6300448775291443, "learning_rate": 9.58143661425927e-05, "loss": 0.0779, "step": 13970 }, { "grad_norm": 0.3800579905509949, "learning_rate": 9.580608409916601e-05, "loss": 0.0578, "step": 13980 }, { "grad_norm": 0.3323085606098175, "learning_rate": 9.579779422873233e-05, "loss": 0.0563, "step": 13990 }, { "grad_norm": 0.45604199171066284, "learning_rate": 9.578949653270819e-05, "loss": 0.0494, "step": 14000 }, { "grad_norm": 0.5251496434211731, "learning_rate": 9.578119101251144e-05, "loss": 0.0491, "step": 14010 }, { "grad_norm": 0.4047832190990448, "learning_rate": 9.577287766956127e-05, "loss": 0.0589, "step": 14020 }, { "grad_norm": 0.42268213629722595, "learning_rate": 9.57645565052782e-05, "loss": 0.0537, "step": 14030 }, { "grad_norm": 0.5514361262321472, "learning_rate": 9.575622752108407e-05, "loss": 0.0467, "step": 14040 }, { "grad_norm": 0.4795500934123993, "learning_rate": 9.57478907184021e-05, "loss": 0.0459, "step": 14050 }, { "grad_norm": 0.4880014955997467, "learning_rate": 9.573954609865681e-05, "loss": 0.0583, "step": 14060 }, { "grad_norm": 0.4189659357070923, "learning_rate": 9.573119366327408e-05, "loss": 0.053, "step": 14070 }, { "grad_norm": 0.42991599440574646, "learning_rate": 9.57228334136811e-05, "loss": 0.0472, "step": 14080 }, { "grad_norm": 0.38633736968040466, "learning_rate": 9.571446535130641e-05, "loss": 0.0538, "step": 14090 }, { "grad_norm": 0.49039220809936523, "learning_rate": 9.570608947757988e-05, "loss": 0.0552, "step": 14100 }, { "grad_norm": 0.43747207522392273, "learning_rate": 9.569770579393274e-05, "loss": 0.0529, "step": 14110 }, { "grad_norm": 0.4230055809020996, "learning_rate": 9.56893143017975e-05, "loss": 0.0474, "step": 14120 }, { "grad_norm": 0.39722612500190735, "learning_rate": 9.568091500260806e-05, "loss": 0.0649, "step": 14130 }, { "grad_norm": 0.5962480306625366, "learning_rate": 9.567250789779961e-05, "loss": 0.0606, "step": 14140 }, { "grad_norm": 0.4724593758583069, "learning_rate": 9.566409298880872e-05, "loss": 0.0443, "step": 14150 }, { "grad_norm": 0.5891101956367493, "learning_rate": 9.565567027707326e-05, "loss": 0.0687, "step": 14160 }, { "grad_norm": 0.5638573169708252, "learning_rate": 9.56472397640324e-05, "loss": 0.055, "step": 14170 }, { "grad_norm": 0.3664485812187195, "learning_rate": 9.563880145112675e-05, "loss": 0.05, "step": 14180 }, { "grad_norm": 0.6362250447273254, "learning_rate": 9.563035533979814e-05, "loss": 0.0555, "step": 14190 }, { "grad_norm": 0.4274928867816925, "learning_rate": 9.562190143148981e-05, "loss": 0.049, "step": 14200 }, { "grad_norm": 0.3567175567150116, "learning_rate": 9.561343972764627e-05, "loss": 0.0496, "step": 14210 }, { "grad_norm": 0.3737016022205353, "learning_rate": 9.560497022971343e-05, "loss": 0.0564, "step": 14220 }, { "grad_norm": 0.5143758654594421, "learning_rate": 9.559649293913847e-05, "loss": 0.0428, "step": 14230 }, { "grad_norm": 0.5343770384788513, "learning_rate": 9.558800785736993e-05, "loss": 0.0485, "step": 14240 }, { "grad_norm": 0.32351234555244446, "learning_rate": 9.557951498585767e-05, "loss": 0.052, "step": 14250 }, { "grad_norm": 0.3517768085002899, "learning_rate": 9.557101432605293e-05, "loss": 0.0429, "step": 14260 }, { "grad_norm": 0.44137367606163025, "learning_rate": 9.556250587940818e-05, "loss": 0.0479, "step": 14270 }, { "grad_norm": 0.5605353713035583, "learning_rate": 9.555398964737734e-05, "loss": 0.0434, "step": 14280 }, { "grad_norm": 0.668316662311554, "learning_rate": 9.554546563141555e-05, "loss": 0.0447, "step": 14290 }, { "grad_norm": 0.48652908205986023, "learning_rate": 9.553693383297937e-05, "loss": 0.062, "step": 14300 }, { "grad_norm": 0.4810357689857483, "learning_rate": 9.552839425352663e-05, "loss": 0.0481, "step": 14310 }, { "grad_norm": 0.38934430480003357, "learning_rate": 9.551984689451652e-05, "loss": 0.0546, "step": 14320 }, { "grad_norm": 0.401573121547699, "learning_rate": 9.551129175740953e-05, "loss": 0.0475, "step": 14330 }, { "grad_norm": 0.37007057666778564, "learning_rate": 9.550272884366754e-05, "loss": 0.0489, "step": 14340 }, { "grad_norm": 0.5134921669960022, "learning_rate": 9.549415815475369e-05, "loss": 0.0529, "step": 14350 }, { "grad_norm": 0.3460010588169098, "learning_rate": 9.548557969213247e-05, "loss": 0.0438, "step": 14360 }, { "grad_norm": 0.46647587418556213, "learning_rate": 9.547699345726972e-05, "loss": 0.0538, "step": 14370 }, { "grad_norm": 0.3549369275569916, "learning_rate": 9.546839945163257e-05, "loss": 0.0458, "step": 14380 }, { "grad_norm": 0.335822731256485, "learning_rate": 9.545979767668953e-05, "loss": 0.0554, "step": 14390 }, { "grad_norm": 0.4232298731803894, "learning_rate": 9.54511881339104e-05, "loss": 0.0449, "step": 14400 }, { "grad_norm": 0.4029057025909424, "learning_rate": 9.54425708247663e-05, "loss": 0.0372, "step": 14410 }, { "grad_norm": 0.44473743438720703, "learning_rate": 9.543394575072972e-05, "loss": 0.0425, "step": 14420 }, { "grad_norm": 0.3823705315589905, "learning_rate": 9.542531291327441e-05, "loss": 0.0448, "step": 14430 }, { "grad_norm": 0.41251423954963684, "learning_rate": 9.541667231387552e-05, "loss": 0.0402, "step": 14440 }, { "grad_norm": 0.3931434452533722, "learning_rate": 9.540802395400949e-05, "loss": 0.0446, "step": 14450 }, { "grad_norm": 0.5140925645828247, "learning_rate": 9.539936783515406e-05, "loss": 0.0514, "step": 14460 }, { "grad_norm": 0.46386322379112244, "learning_rate": 9.539070395878835e-05, "loss": 0.0589, "step": 14470 }, { "grad_norm": 0.3898293375968933, "learning_rate": 9.538203232639277e-05, "loss": 0.0453, "step": 14480 }, { "grad_norm": 0.4270358383655548, "learning_rate": 9.537335293944907e-05, "loss": 0.0545, "step": 14490 }, { "grad_norm": 0.3958742916584015, "learning_rate": 9.536466579944032e-05, "loss": 0.0511, "step": 14500 }, { "grad_norm": 0.3573915958404541, "learning_rate": 9.535597090785091e-05, "loss": 0.0487, "step": 14510 }, { "grad_norm": 0.37973061203956604, "learning_rate": 9.534726826616656e-05, "loss": 0.0439, "step": 14520 }, { "grad_norm": 0.3908436894416809, "learning_rate": 9.53385578758743e-05, "loss": 0.0652, "step": 14530 }, { "grad_norm": 0.3763621151447296, "learning_rate": 9.532983973846252e-05, "loss": 0.0467, "step": 14540 }, { "grad_norm": 0.3563619554042816, "learning_rate": 9.53211138554209e-05, "loss": 0.0492, "step": 14550 }, { "grad_norm": 0.5410149097442627, "learning_rate": 9.531238022824047e-05, "loss": 0.0482, "step": 14560 }, { "grad_norm": 0.371535062789917, "learning_rate": 9.530363885841355e-05, "loss": 0.0433, "step": 14570 }, { "grad_norm": 0.6386649012565613, "learning_rate": 9.52948897474338e-05, "loss": 0.0579, "step": 14580 }, { "grad_norm": 0.3637639284133911, "learning_rate": 9.528613289679622e-05, "loss": 0.0578, "step": 14590 }, { "grad_norm": 0.27464842796325684, "learning_rate": 9.52773683079971e-05, "loss": 0.046, "step": 14600 }, { "grad_norm": 0.6259921789169312, "learning_rate": 9.526859598253407e-05, "loss": 0.0497, "step": 14610 }, { "grad_norm": 0.3236149251461029, "learning_rate": 9.525981592190609e-05, "loss": 0.0447, "step": 14620 }, { "grad_norm": 0.5421450734138489, "learning_rate": 9.525102812761342e-05, "loss": 0.046, "step": 14630 }, { "grad_norm": 0.5637420415878296, "learning_rate": 9.524223260115768e-05, "loss": 0.0451, "step": 14640 }, { "grad_norm": 0.43396374583244324, "learning_rate": 9.523342934404175e-05, "loss": 0.0473, "step": 14650 }, { "grad_norm": 0.40132713317871094, "learning_rate": 9.522461835776989e-05, "loss": 0.0413, "step": 14660 }, { "grad_norm": 0.5744123458862305, "learning_rate": 9.521579964384764e-05, "loss": 0.0589, "step": 14670 }, { "grad_norm": 0.36545059084892273, "learning_rate": 9.52069732037819e-05, "loss": 0.0434, "step": 14680 }, { "grad_norm": 0.4444975256919861, "learning_rate": 9.519813903908083e-05, "loss": 0.0479, "step": 14690 }, { "grad_norm": 0.4383026361465454, "learning_rate": 9.5189297151254e-05, "loss": 0.0592, "step": 14700 }, { "grad_norm": 0.5224928855895996, "learning_rate": 9.518044754181218e-05, "loss": 0.0608, "step": 14710 }, { "grad_norm": 0.48081764578819275, "learning_rate": 9.51715902122676e-05, "loss": 0.0565, "step": 14720 }, { "grad_norm": 0.34417471289634705, "learning_rate": 9.516272516413368e-05, "loss": 0.0502, "step": 14730 }, { "grad_norm": 0.3987349569797516, "learning_rate": 9.515385239892525e-05, "loss": 0.0565, "step": 14740 }, { "grad_norm": 0.5265222191810608, "learning_rate": 9.514497191815839e-05, "loss": 0.058, "step": 14750 }, { "grad_norm": 0.48988908529281616, "learning_rate": 9.513608372335055e-05, "loss": 0.0567, "step": 14760 }, { "grad_norm": 0.5276113748550415, "learning_rate": 9.512718781602045e-05, "loss": 0.0524, "step": 14770 }, { "grad_norm": 0.4319136142730713, "learning_rate": 9.511828419768823e-05, "loss": 0.0506, "step": 14780 }, { "grad_norm": 0.354397714138031, "learning_rate": 9.510937286987521e-05, "loss": 0.0356, "step": 14790 }, { "grad_norm": 0.5350556969642639, "learning_rate": 9.510045383410408e-05, "loss": 0.0484, "step": 14800 }, { "grad_norm": 0.355334609746933, "learning_rate": 9.509152709189892e-05, "loss": 0.0435, "step": 14810 }, { "grad_norm": 0.31612756848335266, "learning_rate": 9.508259264478504e-05, "loss": 0.0483, "step": 14820 }, { "grad_norm": 0.39792823791503906, "learning_rate": 9.507365049428909e-05, "loss": 0.0564, "step": 14830 }, { "grad_norm": 0.4788868725299835, "learning_rate": 9.506470064193902e-05, "loss": 0.0604, "step": 14840 }, { "grad_norm": 0.48969921469688416, "learning_rate": 9.505574308926414e-05, "loss": 0.0585, "step": 14850 }, { "grad_norm": 0.42324474453926086, "learning_rate": 9.504677783779505e-05, "loss": 0.0423, "step": 14860 }, { "grad_norm": 0.5292576551437378, "learning_rate": 9.503780488906365e-05, "loss": 0.0437, "step": 14870 }, { "grad_norm": 0.4574626684188843, "learning_rate": 9.502882424460319e-05, "loss": 0.0478, "step": 14880 }, { "grad_norm": 0.3597344160079956, "learning_rate": 9.501983590594821e-05, "loss": 0.0396, "step": 14890 }, { "grad_norm": 0.42552924156188965, "learning_rate": 9.501083987463455e-05, "loss": 0.0399, "step": 14900 }, { "grad_norm": 0.4066599905490875, "learning_rate": 9.500183615219942e-05, "loss": 0.0435, "step": 14910 }, { "grad_norm": 0.42511945962905884, "learning_rate": 9.49928247401813e-05, "loss": 0.041, "step": 14920 }, { "grad_norm": 0.4580169916152954, "learning_rate": 9.498380564011997e-05, "loss": 0.041, "step": 14930 }, { "grad_norm": 0.311347097158432, "learning_rate": 9.497477885355656e-05, "loss": 0.0456, "step": 14940 }, { "grad_norm": 0.46323731541633606, "learning_rate": 9.496574438203353e-05, "loss": 0.0608, "step": 14950 }, { "grad_norm": 0.38777294754981995, "learning_rate": 9.495670222709459e-05, "loss": 0.0514, "step": 14960 }, { "grad_norm": 0.6246946454048157, "learning_rate": 9.494765239028483e-05, "loss": 0.0515, "step": 14970 }, { "grad_norm": 0.3539639711380005, "learning_rate": 9.493859487315057e-05, "loss": 0.048, "step": 14980 }, { "grad_norm": 0.45446455478668213, "learning_rate": 9.492952967723953e-05, "loss": 0.0516, "step": 14990 }, { "grad_norm": 0.5493901968002319, "learning_rate": 9.492045680410068e-05, "loss": 0.0543, "step": 15000 }, { "grad_norm": 0.3574017584323883, "learning_rate": 9.491137625528436e-05, "loss": 0.0461, "step": 15010 }, { "grad_norm": 0.5267545580863953, "learning_rate": 9.490228803234215e-05, "loss": 0.0439, "step": 15020 }, { "grad_norm": 0.4680495858192444, "learning_rate": 9.489319213682701e-05, "loss": 0.0439, "step": 15030 }, { "grad_norm": 0.3226523995399475, "learning_rate": 9.488408857029316e-05, "loss": 0.0595, "step": 15040 }, { "grad_norm": 0.3285203278064728, "learning_rate": 9.487497733429616e-05, "loss": 0.0557, "step": 15050 }, { "grad_norm": 0.4356628358364105, "learning_rate": 9.486585843039286e-05, "loss": 0.0563, "step": 15060 }, { "grad_norm": 0.5293605923652649, "learning_rate": 9.485673186014143e-05, "loss": 0.0503, "step": 15070 }, { "grad_norm": 0.4125559329986572, "learning_rate": 9.484759762510137e-05, "loss": 0.0567, "step": 15080 }, { "grad_norm": 0.38564661145210266, "learning_rate": 9.483845572683346e-05, "loss": 0.0498, "step": 15090 }, { "grad_norm": 0.45755311846733093, "learning_rate": 9.48293061668998e-05, "loss": 0.0459, "step": 15100 }, { "grad_norm": 0.3876831829547882, "learning_rate": 9.48201489468638e-05, "loss": 0.0448, "step": 15110 }, { "grad_norm": 0.33580484986305237, "learning_rate": 9.481098406829016e-05, "loss": 0.044, "step": 15120 }, { "grad_norm": 0.6601547598838806, "learning_rate": 9.480181153274495e-05, "loss": 0.0398, "step": 15130 }, { "grad_norm": 0.5340685248374939, "learning_rate": 9.479263134179548e-05, "loss": 0.0377, "step": 15140 }, { "grad_norm": 0.45181959867477417, "learning_rate": 9.478344349701039e-05, "loss": 0.0523, "step": 15150 }, { "grad_norm": 0.4356025457382202, "learning_rate": 9.477424799995964e-05, "loss": 0.0463, "step": 15160 }, { "grad_norm": 0.42876335978507996, "learning_rate": 9.476504485221448e-05, "loss": 0.0493, "step": 15170 }, { "grad_norm": 0.38738131523132324, "learning_rate": 9.475583405534748e-05, "loss": 0.044, "step": 15180 }, { "grad_norm": 0.38883450627326965, "learning_rate": 9.474661561093251e-05, "loss": 0.0461, "step": 15190 }, { "grad_norm": 0.4665645658969879, "learning_rate": 9.473738952054478e-05, "loss": 0.0531, "step": 15200 }, { "grad_norm": 0.511527955532074, "learning_rate": 9.472815578576073e-05, "loss": 0.0487, "step": 15210 }, { "grad_norm": 0.6149879097938538, "learning_rate": 9.471891440815817e-05, "loss": 0.0518, "step": 15220 }, { "grad_norm": 0.3436959981918335, "learning_rate": 9.470966538931621e-05, "loss": 0.0407, "step": 15230 }, { "grad_norm": 0.2769535481929779, "learning_rate": 9.470040873081525e-05, "loss": 0.0434, "step": 15240 }, { "grad_norm": 0.5073364973068237, "learning_rate": 9.469114443423698e-05, "loss": 0.0585, "step": 15250 }, { "grad_norm": 0.3960091173648834, "learning_rate": 9.468187250116445e-05, "loss": 0.047, "step": 15260 }, { "grad_norm": 0.36339670419692993, "learning_rate": 9.467259293318197e-05, "loss": 0.0435, "step": 15270 }, { "grad_norm": 0.37663698196411133, "learning_rate": 9.466330573187514e-05, "loss": 0.0412, "step": 15280 }, { "grad_norm": 0.3805038630962372, "learning_rate": 9.46540108988309e-05, "loss": 0.0625, "step": 15290 }, { "grad_norm": 0.4792254865169525, "learning_rate": 9.46447084356375e-05, "loss": 0.0576, "step": 15300 }, { "grad_norm": 0.2677626609802246, "learning_rate": 9.463539834388447e-05, "loss": 0.0501, "step": 15310 }, { "grad_norm": 0.4684320390224457, "learning_rate": 9.462608062516263e-05, "loss": 0.035, "step": 15320 }, { "grad_norm": 0.39876022934913635, "learning_rate": 9.461675528106413e-05, "loss": 0.0391, "step": 15330 }, { "grad_norm": 0.3166360557079315, "learning_rate": 9.460742231318244e-05, "loss": 0.0447, "step": 15340 }, { "grad_norm": 0.6665896773338318, "learning_rate": 9.459808172311229e-05, "loss": 0.0655, "step": 15350 }, { "grad_norm": 0.3686244487762451, "learning_rate": 9.458873351244972e-05, "loss": 0.0403, "step": 15360 }, { "grad_norm": 0.3421466052532196, "learning_rate": 9.457937768279211e-05, "loss": 0.0459, "step": 15370 }, { "grad_norm": 0.5345413684844971, "learning_rate": 9.45700142357381e-05, "loss": 0.0426, "step": 15380 }, { "grad_norm": 0.44683557748794556, "learning_rate": 9.456064317288765e-05, "loss": 0.0395, "step": 15390 }, { "grad_norm": 0.4351876676082611, "learning_rate": 9.455126449584201e-05, "loss": 0.0435, "step": 15400 }, { "grad_norm": 0.34717047214508057, "learning_rate": 9.454187820620375e-05, "loss": 0.0327, "step": 15410 }, { "grad_norm": 0.456391304731369, "learning_rate": 9.453248430557673e-05, "loss": 0.0506, "step": 15420 }, { "grad_norm": 0.6570957899093628, "learning_rate": 9.452308279556611e-05, "loss": 0.0646, "step": 15430 }, { "grad_norm": 0.36198702454566956, "learning_rate": 9.451367367777835e-05, "loss": 0.0419, "step": 15440 }, { "grad_norm": 0.36804890632629395, "learning_rate": 9.450425695382122e-05, "loss": 0.0466, "step": 15450 }, { "grad_norm": 0.43664154410362244, "learning_rate": 9.449483262530375e-05, "loss": 0.0461, "step": 15460 }, { "grad_norm": 0.4594056308269501, "learning_rate": 9.448540069383633e-05, "loss": 0.0432, "step": 15470 }, { "grad_norm": 0.4243190884590149, "learning_rate": 9.447596116103061e-05, "loss": 0.0476, "step": 15480 }, { "grad_norm": 0.4383981227874756, "learning_rate": 9.446651402849955e-05, "loss": 0.048, "step": 15490 }, { "grad_norm": 0.3482053279876709, "learning_rate": 9.44570592978574e-05, "loss": 0.0438, "step": 15500 }, { "grad_norm": 0.3677067458629608, "learning_rate": 9.444759697071972e-05, "loss": 0.0448, "step": 15510 }, { "grad_norm": 0.430418461561203, "learning_rate": 9.443812704870336e-05, "loss": 0.0389, "step": 15520 }, { "grad_norm": 0.3952595889568329, "learning_rate": 9.442864953342649e-05, "loss": 0.0408, "step": 15530 }, { "grad_norm": 0.32777300477027893, "learning_rate": 9.441916442650852e-05, "loss": 0.0461, "step": 15540 }, { "grad_norm": 0.27825677394866943, "learning_rate": 9.440967172957023e-05, "loss": 0.0414, "step": 15550 }, { "grad_norm": 0.4883272349834442, "learning_rate": 9.440017144423364e-05, "loss": 0.0481, "step": 15560 }, { "grad_norm": 0.44123831391334534, "learning_rate": 9.439066357212209e-05, "loss": 0.0468, "step": 15570 }, { "grad_norm": 0.40344303846359253, "learning_rate": 9.438114811486022e-05, "loss": 0.0569, "step": 15580 }, { "grad_norm": 0.31499001383781433, "learning_rate": 9.4371625074074e-05, "loss": 0.041, "step": 15590 }, { "grad_norm": 0.375234991312027, "learning_rate": 9.436209445139059e-05, "loss": 0.0427, "step": 15600 }, { "grad_norm": 0.29422852396965027, "learning_rate": 9.435255624843855e-05, "loss": 0.0495, "step": 15610 }, { "grad_norm": 0.32343804836273193, "learning_rate": 9.43430104668477e-05, "loss": 0.04, "step": 15620 }, { "grad_norm": 0.43721866607666016, "learning_rate": 9.433345710824914e-05, "loss": 0.0473, "step": 15630 }, { "grad_norm": 0.4782821834087372, "learning_rate": 9.432389617427529e-05, "loss": 0.0432, "step": 15640 }, { "grad_norm": 0.43257370591163635, "learning_rate": 9.431432766655984e-05, "loss": 0.0395, "step": 15650 }, { "grad_norm": 0.4842003881931305, "learning_rate": 9.430475158673778e-05, "loss": 0.0489, "step": 15660 }, { "grad_norm": 0.48165959119796753, "learning_rate": 9.429516793644542e-05, "loss": 0.0386, "step": 15670 }, { "grad_norm": 0.32613605260849, "learning_rate": 9.428557671732034e-05, "loss": 0.0331, "step": 15680 }, { "grad_norm": 0.2796420753002167, "learning_rate": 9.42759779310014e-05, "loss": 0.0406, "step": 15690 }, { "grad_norm": 0.4236118197441101, "learning_rate": 9.426637157912879e-05, "loss": 0.0431, "step": 15700 }, { "grad_norm": 0.3842046558856964, "learning_rate": 9.425675766334397e-05, "loss": 0.0404, "step": 15710 }, { "grad_norm": 0.39695361256599426, "learning_rate": 9.424713618528968e-05, "loss": 0.0391, "step": 15720 }, { "grad_norm": 0.43220433592796326, "learning_rate": 9.423750714661e-05, "loss": 0.0515, "step": 15730 }, { "grad_norm": 0.4431755244731903, "learning_rate": 9.422787054895022e-05, "loss": 0.0411, "step": 15740 }, { "grad_norm": 0.43833380937576294, "learning_rate": 9.4218226393957e-05, "loss": 0.0576, "step": 15750 }, { "grad_norm": 0.473070353269577, "learning_rate": 9.420857468327828e-05, "loss": 0.0477, "step": 15760 }, { "grad_norm": 0.4044108986854553, "learning_rate": 9.419891541856323e-05, "loss": 0.0442, "step": 15770 }, { "grad_norm": 0.4894278049468994, "learning_rate": 9.41892486014624e-05, "loss": 0.0413, "step": 15780 }, { "grad_norm": 0.4536127746105194, "learning_rate": 9.417957423362756e-05, "loss": 0.0468, "step": 15790 }, { "grad_norm": 0.4407362937927246, "learning_rate": 9.416989231671178e-05, "loss": 0.0485, "step": 15800 }, { "grad_norm": 0.44214892387390137, "learning_rate": 9.416020285236946e-05, "loss": 0.0442, "step": 15810 }, { "grad_norm": 0.34739595651626587, "learning_rate": 9.415050584225626e-05, "loss": 0.0462, "step": 15820 }, { "grad_norm": 0.44970324635505676, "learning_rate": 9.414080128802914e-05, "loss": 0.0449, "step": 15830 }, { "grad_norm": 0.3624149262905121, "learning_rate": 9.413108919134632e-05, "loss": 0.0469, "step": 15840 }, { "grad_norm": 0.5957751274108887, "learning_rate": 9.412136955386734e-05, "loss": 0.0463, "step": 15850 }, { "grad_norm": 0.4215199649333954, "learning_rate": 9.411164237725303e-05, "loss": 0.0421, "step": 15860 }, { "grad_norm": 0.4508938193321228, "learning_rate": 9.41019076631655e-05, "loss": 0.0489, "step": 15870 }, { "grad_norm": 0.34382355213165283, "learning_rate": 9.409216541326815e-05, "loss": 0.0465, "step": 15880 }, { "grad_norm": 0.39293378591537476, "learning_rate": 9.408241562922564e-05, "loss": 0.0488, "step": 15890 }, { "grad_norm": 0.4434188902378082, "learning_rate": 9.407265831270395e-05, "loss": 0.0447, "step": 15900 }, { "grad_norm": 0.5059065222740173, "learning_rate": 9.406289346537035e-05, "loss": 0.0405, "step": 15910 }, { "grad_norm": 0.5186411142349243, "learning_rate": 9.405312108889339e-05, "loss": 0.044, "step": 15920 }, { "grad_norm": 0.5212205648422241, "learning_rate": 9.404334118494288e-05, "loss": 0.0458, "step": 15930 }, { "grad_norm": 0.35700705647468567, "learning_rate": 9.403355375518995e-05, "loss": 0.0392, "step": 15940 }, { "grad_norm": 0.394294410943985, "learning_rate": 9.4023758801307e-05, "loss": 0.0408, "step": 15950 }, { "grad_norm": 0.7917149066925049, "learning_rate": 9.401395632496774e-05, "loss": 0.0384, "step": 15960 }, { "grad_norm": 0.5102378726005554, "learning_rate": 9.400414632784711e-05, "loss": 0.0533, "step": 15970 }, { "grad_norm": 0.33464574813842773, "learning_rate": 9.39943288116214e-05, "loss": 0.0466, "step": 15980 }, { "grad_norm": 0.4184226393699646, "learning_rate": 9.398450377796815e-05, "loss": 0.048, "step": 15990 }, { "grad_norm": 0.36692914366722107, "learning_rate": 9.397467122856616e-05, "loss": 0.0432, "step": 16000 }, { "grad_norm": 0.4035806953907013, "learning_rate": 9.396483116509558e-05, "loss": 0.0465, "step": 16010 }, { "grad_norm": 0.3566799759864807, "learning_rate": 9.39549835892378e-05, "loss": 0.0448, "step": 16020 }, { "grad_norm": 0.3189820945262909, "learning_rate": 9.39451285026755e-05, "loss": 0.0461, "step": 16030 }, { "grad_norm": 0.5422073602676392, "learning_rate": 9.393526590709262e-05, "loss": 0.0442, "step": 16040 }, { "grad_norm": 0.3901299834251404, "learning_rate": 9.392539580417444e-05, "loss": 0.0407, "step": 16050 }, { "grad_norm": 0.3800064027309418, "learning_rate": 9.391551819560747e-05, "loss": 0.0422, "step": 16060 }, { "grad_norm": 0.419847309589386, "learning_rate": 9.390563308307955e-05, "loss": 0.0425, "step": 16070 }, { "grad_norm": 0.47501933574676514, "learning_rate": 9.389574046827974e-05, "loss": 0.0417, "step": 16080 }, { "grad_norm": 0.41062042117118835, "learning_rate": 9.388584035289845e-05, "loss": 0.0505, "step": 16090 }, { "grad_norm": 0.3801901340484619, "learning_rate": 9.387593273862732e-05, "loss": 0.0461, "step": 16100 }, { "grad_norm": 0.33759981393814087, "learning_rate": 9.386601762715929e-05, "loss": 0.0446, "step": 16110 }, { "grad_norm": 0.38685306906700134, "learning_rate": 9.38560950201886e-05, "loss": 0.0536, "step": 16120 }, { "grad_norm": 0.6183090209960938, "learning_rate": 9.384616491941071e-05, "loss": 0.0529, "step": 16130 }, { "grad_norm": 0.4639653265476227, "learning_rate": 9.383622732652245e-05, "loss": 0.0456, "step": 16140 }, { "grad_norm": 0.2995966672897339, "learning_rate": 9.382628224322187e-05, "loss": 0.0497, "step": 16150 }, { "grad_norm": 0.3074832558631897, "learning_rate": 9.381632967120829e-05, "loss": 0.049, "step": 16160 }, { "grad_norm": 0.5083504319190979, "learning_rate": 9.380636961218235e-05, "loss": 0.039, "step": 16170 }, { "grad_norm": 0.3463893532752991, "learning_rate": 9.379640206784597e-05, "loss": 0.037, "step": 16180 }, { "grad_norm": 0.3878813087940216, "learning_rate": 9.378642703990229e-05, "loss": 0.0353, "step": 16190 }, { "grad_norm": 0.4065835475921631, "learning_rate": 9.37764445300558e-05, "loss": 0.0441, "step": 16200 }, { "grad_norm": 0.3030817210674286, "learning_rate": 9.376645454001222e-05, "loss": 0.052, "step": 16210 }, { "grad_norm": 0.3877769112586975, "learning_rate": 9.375645707147858e-05, "loss": 0.0394, "step": 16220 }, { "grad_norm": 0.2964002788066864, "learning_rate": 9.374645212616316e-05, "loss": 0.0504, "step": 16230 }, { "grad_norm": 0.631924033164978, "learning_rate": 9.373643970577555e-05, "loss": 0.0472, "step": 16240 }, { "grad_norm": 0.4764178991317749, "learning_rate": 9.372641981202659e-05, "loss": 0.0413, "step": 16250 }, { "grad_norm": 0.6143308281898499, "learning_rate": 9.37163924466284e-05, "loss": 0.0499, "step": 16260 }, { "grad_norm": 0.4361116588115692, "learning_rate": 9.370635761129438e-05, "loss": 0.0416, "step": 16270 }, { "grad_norm": 0.4838939607143402, "learning_rate": 9.36963153077392e-05, "loss": 0.0446, "step": 16280 }, { "grad_norm": 0.3908381760120392, "learning_rate": 9.368626553767888e-05, "loss": 0.0401, "step": 16290 }, { "grad_norm": 0.3962392807006836, "learning_rate": 9.367620830283057e-05, "loss": 0.0354, "step": 16300 }, { "grad_norm": 0.4220229387283325, "learning_rate": 9.366614360491281e-05, "loss": 0.0427, "step": 16310 }, { "grad_norm": 0.43414464592933655, "learning_rate": 9.365607144564539e-05, "loss": 0.0431, "step": 16320 }, { "grad_norm": 0.4440256357192993, "learning_rate": 9.364599182674934e-05, "loss": 0.0525, "step": 16330 }, { "grad_norm": 0.4151844084262848, "learning_rate": 9.3635904749947e-05, "loss": 0.0434, "step": 16340 }, { "grad_norm": 0.4529193937778473, "learning_rate": 9.362581021696202e-05, "loss": 0.051, "step": 16350 }, { "grad_norm": 0.3377729058265686, "learning_rate": 9.361570822951921e-05, "loss": 0.0458, "step": 16360 }, { "grad_norm": 0.3396523594856262, "learning_rate": 9.360559878934476e-05, "loss": 0.0472, "step": 16370 }, { "grad_norm": 0.3068786859512329, "learning_rate": 9.359548189816611e-05, "loss": 0.038, "step": 16380 }, { "grad_norm": 0.4262719750404358, "learning_rate": 9.358535755771193e-05, "loss": 0.0479, "step": 16390 }, { "grad_norm": 0.521482527256012, "learning_rate": 9.357522576971221e-05, "loss": 0.0467, "step": 16400 }, { "grad_norm": 0.5120548605918884, "learning_rate": 9.356508653589819e-05, "loss": 0.0516, "step": 16410 }, { "grad_norm": 0.4502435624599457, "learning_rate": 9.355493985800237e-05, "loss": 0.0477, "step": 16420 }, { "grad_norm": 0.45179471373558044, "learning_rate": 9.354478573775857e-05, "loss": 0.0416, "step": 16430 }, { "grad_norm": 0.3880103826522827, "learning_rate": 9.353462417690186e-05, "loss": 0.0417, "step": 16440 }, { "grad_norm": 0.450783371925354, "learning_rate": 9.352445517716853e-05, "loss": 0.0448, "step": 16450 }, { "grad_norm": 0.3761644959449768, "learning_rate": 9.351427874029621e-05, "loss": 0.0436, "step": 16460 }, { "grad_norm": 0.47434043884277344, "learning_rate": 9.350409486802379e-05, "loss": 0.0442, "step": 16470 }, { "grad_norm": 0.5922096967697144, "learning_rate": 9.349390356209138e-05, "loss": 0.0516, "step": 16480 }, { "grad_norm": 0.41341906785964966, "learning_rate": 9.348370482424042e-05, "loss": 0.0434, "step": 16490 }, { "grad_norm": 0.501868486404419, "learning_rate": 9.347349865621357e-05, "loss": 0.0414, "step": 16500 }, { "grad_norm": 0.4863445460796356, "learning_rate": 9.346328505975481e-05, "loss": 0.0483, "step": 16510 }, { "grad_norm": 0.5679590702056885, "learning_rate": 9.345306403660936e-05, "loss": 0.0492, "step": 16520 }, { "grad_norm": 0.518574595451355, "learning_rate": 9.344283558852371e-05, "loss": 0.052, "step": 16530 }, { "grad_norm": 0.5879845023155212, "learning_rate": 9.343259971724563e-05, "loss": 0.0488, "step": 16540 }, { "grad_norm": 0.5084226131439209, "learning_rate": 9.342235642452413e-05, "loss": 0.0544, "step": 16550 }, { "grad_norm": 0.5165947675704956, "learning_rate": 9.341210571210954e-05, "loss": 0.0464, "step": 16560 }, { "grad_norm": 0.3577542304992676, "learning_rate": 9.340184758175338e-05, "loss": 0.0446, "step": 16570 }, { "grad_norm": 0.4352652430534363, "learning_rate": 9.339158203520854e-05, "loss": 0.0541, "step": 16580 }, { "grad_norm": 0.5303510427474976, "learning_rate": 9.338130907422908e-05, "loss": 0.0376, "step": 16590 }, { "grad_norm": 0.41136085987091064, "learning_rate": 9.337102870057037e-05, "loss": 0.0364, "step": 16600 }, { "grad_norm": 0.28453487157821655, "learning_rate": 9.336074091598907e-05, "loss": 0.0352, "step": 16610 }, { "grad_norm": 0.2626551687717438, "learning_rate": 9.335044572224306e-05, "loss": 0.0362, "step": 16620 }, { "grad_norm": 0.5358864068984985, "learning_rate": 9.334014312109151e-05, "loss": 0.0387, "step": 16630 }, { "grad_norm": 0.48256969451904297, "learning_rate": 9.332983311429486e-05, "loss": 0.0503, "step": 16640 }, { "grad_norm": 0.3758418560028076, "learning_rate": 9.33195157036148e-05, "loss": 0.0472, "step": 16650 }, { "grad_norm": 0.3942691683769226, "learning_rate": 9.330919089081432e-05, "loss": 0.0436, "step": 16660 }, { "grad_norm": 0.3391691744327545, "learning_rate": 9.32988586776576e-05, "loss": 0.0591, "step": 16670 }, { "grad_norm": 0.5201638340950012, "learning_rate": 9.328851906591016e-05, "loss": 0.0461, "step": 16680 }, { "grad_norm": 0.3976421356201172, "learning_rate": 9.327817205733875e-05, "loss": 0.0467, "step": 16690 }, { "grad_norm": 0.44531333446502686, "learning_rate": 9.326781765371142e-05, "loss": 0.041, "step": 16700 }, { "grad_norm": 0.43326932191848755, "learning_rate": 9.325745585679741e-05, "loss": 0.0422, "step": 16710 }, { "grad_norm": 0.43953657150268555, "learning_rate": 9.32470866683673e-05, "loss": 0.0586, "step": 16720 }, { "grad_norm": 0.5048316717147827, "learning_rate": 9.323671009019288e-05, "loss": 0.0568, "step": 16730 }, { "grad_norm": 0.3681170344352722, "learning_rate": 9.322632612404725e-05, "loss": 0.0426, "step": 16740 }, { "grad_norm": 0.33829477429389954, "learning_rate": 9.321593477170471e-05, "loss": 0.0561, "step": 16750 }, { "grad_norm": 0.3051062226295471, "learning_rate": 9.320553603494088e-05, "loss": 0.0398, "step": 16760 }, { "grad_norm": 0.5027194619178772, "learning_rate": 9.319512991553261e-05, "loss": 0.0415, "step": 16770 }, { "grad_norm": 0.49349668622016907, "learning_rate": 9.318471641525803e-05, "loss": 0.042, "step": 16780 }, { "grad_norm": 0.4551987946033478, "learning_rate": 9.317429553589652e-05, "loss": 0.0365, "step": 16790 }, { "grad_norm": 0.3596382737159729, "learning_rate": 9.316386727922873e-05, "loss": 0.0454, "step": 16800 }, { "grad_norm": 0.4626270830631256, "learning_rate": 9.315343164703656e-05, "loss": 0.0393, "step": 16810 }, { "grad_norm": 0.42464062571525574, "learning_rate": 9.314298864110316e-05, "loss": 0.0352, "step": 16820 }, { "grad_norm": 0.3531777262687683, "learning_rate": 9.313253826321295e-05, "loss": 0.0394, "step": 16830 }, { "grad_norm": 0.364473432302475, "learning_rate": 9.312208051515165e-05, "loss": 0.035, "step": 16840 }, { "grad_norm": 0.43144690990448, "learning_rate": 9.311161539870618e-05, "loss": 0.0364, "step": 16850 }, { "grad_norm": 0.3423345983028412, "learning_rate": 9.310114291566474e-05, "loss": 0.0452, "step": 16860 }, { "grad_norm": 0.4029861390590668, "learning_rate": 9.309066306781679e-05, "loss": 0.0349, "step": 16870 }, { "grad_norm": 0.3593249022960663, "learning_rate": 9.308017585695306e-05, "loss": 0.047, "step": 16880 }, { "grad_norm": 0.36391180753707886, "learning_rate": 9.306968128486552e-05, "loss": 0.0435, "step": 16890 }, { "grad_norm": 0.3166039288043976, "learning_rate": 9.30591793533474e-05, "loss": 0.0332, "step": 16900 }, { "grad_norm": 0.40885618329048157, "learning_rate": 9.304867006419321e-05, "loss": 0.0396, "step": 16910 }, { "grad_norm": 1.0482898950576782, "learning_rate": 9.303815341919868e-05, "loss": 0.0441, "step": 16920 }, { "grad_norm": 0.4519434869289398, "learning_rate": 9.302762942016084e-05, "loss": 0.048, "step": 16930 }, { "grad_norm": 0.3257277011871338, "learning_rate": 9.301709806887792e-05, "loss": 0.0375, "step": 16940 }, { "grad_norm": 0.4436439275741577, "learning_rate": 9.300655936714948e-05, "loss": 0.0611, "step": 16950 }, { "grad_norm": 0.4149447977542877, "learning_rate": 9.299601331677627e-05, "loss": 0.0504, "step": 16960 }, { "grad_norm": 0.42641881108283997, "learning_rate": 9.298545991956033e-05, "loss": 0.0496, "step": 16970 }, { "grad_norm": 0.34324881434440613, "learning_rate": 9.297489917730493e-05, "loss": 0.0464, "step": 16980 }, { "grad_norm": 0.43665653467178345, "learning_rate": 9.296433109181464e-05, "loss": 0.0484, "step": 16990 }, { "grad_norm": 0.583954930305481, "learning_rate": 9.295375566489523e-05, "loss": 0.0423, "step": 17000 }, { "grad_norm": 0.40009480714797974, "learning_rate": 9.294317289835379e-05, "loss": 0.0476, "step": 17010 }, { "grad_norm": 0.42184683680534363, "learning_rate": 9.293258279399859e-05, "loss": 0.0488, "step": 17020 }, { "grad_norm": 0.27476897835731506, "learning_rate": 9.292198535363919e-05, "loss": 0.0356, "step": 17030 }, { "grad_norm": 0.3100835978984833, "learning_rate": 9.291138057908641e-05, "loss": 0.0342, "step": 17040 }, { "grad_norm": 0.41872158646583557, "learning_rate": 9.290076847215234e-05, "loss": 0.0383, "step": 17050 }, { "grad_norm": 0.6465931534767151, "learning_rate": 9.289014903465025e-05, "loss": 0.0557, "step": 17060 }, { "grad_norm": 0.38697266578674316, "learning_rate": 9.287952226839475e-05, "loss": 0.035, "step": 17070 }, { "grad_norm": 0.4759429693222046, "learning_rate": 9.286888817520164e-05, "loss": 0.0385, "step": 17080 }, { "grad_norm": 0.4004027247428894, "learning_rate": 9.285824675688803e-05, "loss": 0.0405, "step": 17090 }, { "grad_norm": 0.39779022336006165, "learning_rate": 9.28475980152722e-05, "loss": 0.0426, "step": 17100 }, { "grad_norm": 0.3896577060222626, "learning_rate": 9.283694195217379e-05, "loss": 0.0503, "step": 17110 }, { "grad_norm": 0.48196175694465637, "learning_rate": 9.282627856941356e-05, "loss": 0.0432, "step": 17120 }, { "grad_norm": 0.38521042466163635, "learning_rate": 9.281560786881363e-05, "loss": 0.0407, "step": 17130 }, { "grad_norm": 0.5807782411575317, "learning_rate": 9.280492985219733e-05, "loss": 0.0344, "step": 17140 }, { "grad_norm": 0.4619711637496948, "learning_rate": 9.279424452138924e-05, "loss": 0.0444, "step": 17150 }, { "grad_norm": 0.33069875836372375, "learning_rate": 9.278355187821517e-05, "loss": 0.0404, "step": 17160 }, { "grad_norm": 0.3418179154396057, "learning_rate": 9.277285192450224e-05, "loss": 0.042, "step": 17170 }, { "grad_norm": 0.46710795164108276, "learning_rate": 9.276214466207875e-05, "loss": 0.0408, "step": 17180 }, { "grad_norm": 0.4858396053314209, "learning_rate": 9.275143009277427e-05, "loss": 0.0422, "step": 17190 }, { "grad_norm": 0.41810932755470276, "learning_rate": 9.274070821841964e-05, "loss": 0.0408, "step": 17200 }, { "grad_norm": 0.3901822865009308, "learning_rate": 9.272997904084696e-05, "loss": 0.0382, "step": 17210 }, { "grad_norm": 0.3573213517665863, "learning_rate": 9.271924256188951e-05, "loss": 0.0468, "step": 17220 }, { "grad_norm": 0.35929742455482483, "learning_rate": 9.270849878338189e-05, "loss": 0.0437, "step": 17230 }, { "grad_norm": 0.5457412004470825, "learning_rate": 9.269774770715991e-05, "loss": 0.0453, "step": 17240 }, { "grad_norm": 0.47812405228614807, "learning_rate": 9.268698933506061e-05, "loss": 0.0583, "step": 17250 }, { "grad_norm": 0.3820277154445648, "learning_rate": 9.267622366892235e-05, "loss": 0.0443, "step": 17260 }, { "grad_norm": 0.2755855619907379, "learning_rate": 9.266545071058465e-05, "loss": 0.0361, "step": 17270 }, { "grad_norm": 0.3304334282875061, "learning_rate": 9.265467046188833e-05, "loss": 0.0341, "step": 17280 }, { "grad_norm": 0.28729045391082764, "learning_rate": 9.264388292467543e-05, "loss": 0.0442, "step": 17290 }, { "grad_norm": 0.39653337001800537, "learning_rate": 9.263308810078926e-05, "loss": 0.0376, "step": 17300 }, { "grad_norm": 0.3473777770996094, "learning_rate": 9.262228599207434e-05, "loss": 0.0367, "step": 17310 }, { "grad_norm": 0.380655437707901, "learning_rate": 9.261147660037647e-05, "loss": 0.0377, "step": 17320 }, { "grad_norm": 0.41621363162994385, "learning_rate": 9.26006599275427e-05, "loss": 0.0403, "step": 17330 }, { "grad_norm": 0.3089645206928253, "learning_rate": 9.258983597542124e-05, "loss": 0.0368, "step": 17340 }, { "grad_norm": 0.4045976400375366, "learning_rate": 9.257900474586167e-05, "loss": 0.0367, "step": 17350 }, { "grad_norm": 0.3870009183883667, "learning_rate": 9.256816624071471e-05, "loss": 0.0488, "step": 17360 }, { "grad_norm": 0.4017009735107422, "learning_rate": 9.25573204618324e-05, "loss": 0.0425, "step": 17370 }, { "grad_norm": 0.4291140139102936, "learning_rate": 9.254646741106796e-05, "loss": 0.0396, "step": 17380 }, { "grad_norm": 0.39610758423805237, "learning_rate": 9.253560709027589e-05, "loss": 0.0371, "step": 17390 }, { "grad_norm": 0.4601064920425415, "learning_rate": 9.252473950131192e-05, "loss": 0.0438, "step": 17400 }, { "grad_norm": 0.45762211084365845, "learning_rate": 9.251386464603302e-05, "loss": 0.0382, "step": 17410 }, { "grad_norm": 0.382019966840744, "learning_rate": 9.250298252629741e-05, "loss": 0.0448, "step": 17420 }, { "grad_norm": 0.43449342250823975, "learning_rate": 9.249209314396454e-05, "loss": 0.0385, "step": 17430 }, { "grad_norm": 0.4427739977836609, "learning_rate": 9.248119650089513e-05, "loss": 0.0449, "step": 17440 }, { "grad_norm": 0.3351946175098419, "learning_rate": 9.247029259895108e-05, "loss": 0.0442, "step": 17450 }, { "grad_norm": 0.3758673667907715, "learning_rate": 9.24593814399956e-05, "loss": 0.0446, "step": 17460 }, { "grad_norm": 0.31631284952163696, "learning_rate": 9.244846302589309e-05, "loss": 0.0362, "step": 17470 }, { "grad_norm": 0.3406713604927063, "learning_rate": 9.243753735850923e-05, "loss": 0.0432, "step": 17480 }, { "grad_norm": 0.36935099959373474, "learning_rate": 9.24266044397109e-05, "loss": 0.0374, "step": 17490 }, { "grad_norm": 0.302031546831131, "learning_rate": 9.241566427136624e-05, "loss": 0.0364, "step": 17500 }, { "grad_norm": 0.4129435420036316, "learning_rate": 9.240471685534463e-05, "loss": 0.0448, "step": 17510 }, { "grad_norm": 0.37162116169929504, "learning_rate": 9.239376219351667e-05, "loss": 0.0468, "step": 17520 }, { "grad_norm": 0.4091261029243469, "learning_rate": 9.238280028775425e-05, "loss": 0.0428, "step": 17530 }, { "grad_norm": 0.34894564747810364, "learning_rate": 9.237183113993041e-05, "loss": 0.0429, "step": 17540 }, { "grad_norm": 0.37347179651260376, "learning_rate": 9.236085475191952e-05, "loss": 0.0377, "step": 17550 }, { "grad_norm": 0.4452196955680847, "learning_rate": 9.234987112559709e-05, "loss": 0.0663, "step": 17560 }, { "grad_norm": 0.5135387182235718, "learning_rate": 9.233888026283999e-05, "loss": 0.0466, "step": 17570 }, { "grad_norm": 0.4426175355911255, "learning_rate": 9.232788216552619e-05, "loss": 0.0487, "step": 17580 }, { "grad_norm": 0.6211991906166077, "learning_rate": 9.231687683553502e-05, "loss": 0.0507, "step": 17590 }, { "grad_norm": 0.4437233507633209, "learning_rate": 9.230586427474698e-05, "loss": 0.0519, "step": 17600 }, { "grad_norm": 0.4256303310394287, "learning_rate": 9.229484448504379e-05, "loss": 0.0508, "step": 17610 }, { "grad_norm": 0.3371683359146118, "learning_rate": 9.228381746830843e-05, "loss": 0.0458, "step": 17620 }, { "grad_norm": 0.3569183647632599, "learning_rate": 9.227278322642514e-05, "loss": 0.0384, "step": 17630 }, { "grad_norm": 0.3691626787185669, "learning_rate": 9.226174176127937e-05, "loss": 0.0394, "step": 17640 }, { "grad_norm": 0.3228568732738495, "learning_rate": 9.22506930747578e-05, "loss": 0.0373, "step": 17650 }, { "grad_norm": 0.5550647377967834, "learning_rate": 9.223963716874831e-05, "loss": 0.0481, "step": 17660 }, { "grad_norm": 0.5510848760604858, "learning_rate": 9.222857404514012e-05, "loss": 0.0468, "step": 17670 }, { "grad_norm": 0.4391331374645233, "learning_rate": 9.221750370582355e-05, "loss": 0.0439, "step": 17680 }, { "grad_norm": 0.36389705538749695, "learning_rate": 9.220642615269028e-05, "loss": 0.0768, "step": 17690 }, { "grad_norm": 0.36464980244636536, "learning_rate": 9.219534138763311e-05, "loss": 0.0618, "step": 17700 }, { "grad_norm": 0.5256492495536804, "learning_rate": 9.218424941254613e-05, "loss": 0.0551, "step": 17710 }, { "grad_norm": 0.3621556758880615, "learning_rate": 9.217315022932468e-05, "loss": 0.0542, "step": 17720 }, { "grad_norm": 0.6147473454475403, "learning_rate": 9.216204383986528e-05, "loss": 0.0486, "step": 17730 }, { "grad_norm": 0.39959216117858887, "learning_rate": 9.215093024606574e-05, "loss": 0.0507, "step": 17740 }, { "grad_norm": 0.4460222125053406, "learning_rate": 9.213980944982506e-05, "loss": 0.0674, "step": 17750 }, { "grad_norm": 0.5578201413154602, "learning_rate": 9.212868145304346e-05, "loss": 0.0487, "step": 17760 }, { "grad_norm": 0.3353289067745209, "learning_rate": 9.211754625762241e-05, "loss": 0.0395, "step": 17770 }, { "grad_norm": 0.38032034039497375, "learning_rate": 9.210640386546463e-05, "loss": 0.0499, "step": 17780 }, { "grad_norm": 0.31305283308029175, "learning_rate": 9.209525427847405e-05, "loss": 0.0323, "step": 17790 }, { "grad_norm": 0.3407708406448364, "learning_rate": 9.208409749855583e-05, "loss": 0.0343, "step": 17800 }, { "grad_norm": 0.4867485463619232, "learning_rate": 9.207293352761633e-05, "loss": 0.0446, "step": 17810 }, { "grad_norm": 0.4683200418949127, "learning_rate": 9.206176236756319e-05, "loss": 0.043, "step": 17820 }, { "grad_norm": 0.37558525800704956, "learning_rate": 9.205058402030525e-05, "loss": 0.0417, "step": 17830 }, { "grad_norm": 0.3811051547527313, "learning_rate": 9.203939848775259e-05, "loss": 0.0374, "step": 17840 }, { "grad_norm": 0.3205406367778778, "learning_rate": 9.202820577181652e-05, "loss": 0.041, "step": 17850 }, { "grad_norm": 0.4971674978733063, "learning_rate": 9.201700587440953e-05, "loss": 0.0417, "step": 17860 }, { "grad_norm": 0.9981386661529541, "learning_rate": 9.200579879744544e-05, "loss": 0.0415, "step": 17870 }, { "grad_norm": 0.3791153132915497, "learning_rate": 9.199458454283918e-05, "loss": 0.041, "step": 17880 }, { "grad_norm": 0.5937963724136353, "learning_rate": 9.198336311250697e-05, "loss": 0.0437, "step": 17890 }, { "grad_norm": 0.4834320545196533, "learning_rate": 9.197213450836626e-05, "loss": 0.0404, "step": 17900 }, { "grad_norm": 0.4908550977706909, "learning_rate": 9.19608987323357e-05, "loss": 0.0523, "step": 17910 }, { "grad_norm": 0.3505353629589081, "learning_rate": 9.194965578633517e-05, "loss": 0.0388, "step": 17920 }, { "grad_norm": 0.8612362742424011, "learning_rate": 9.193840567228582e-05, "loss": 0.04, "step": 17930 }, { "grad_norm": 0.38880887627601624, "learning_rate": 9.192714839210994e-05, "loss": 0.0371, "step": 17940 }, { "grad_norm": 0.44882965087890625, "learning_rate": 9.19158839477311e-05, "loss": 0.0591, "step": 17950 }, { "grad_norm": 0.4485377371311188, "learning_rate": 9.190461234107411e-05, "loss": 0.0472, "step": 17960 }, { "grad_norm": 0.6432567834854126, "learning_rate": 9.189333357406496e-05, "loss": 0.0456, "step": 17970 }, { "grad_norm": 0.36952081322669983, "learning_rate": 9.188204764863089e-05, "loss": 0.0565, "step": 17980 }, { "grad_norm": 0.4345738887786865, "learning_rate": 9.187075456670033e-05, "loss": 0.0449, "step": 17990 }, { "grad_norm": 0.4026283621788025, "learning_rate": 9.1859454330203e-05, "loss": 0.0399, "step": 18000 }, { "grad_norm": 0.3892875611782074, "learning_rate": 9.18481469410698e-05, "loss": 0.0548, "step": 18010 }, { "grad_norm": 0.3452394902706146, "learning_rate": 9.183683240123281e-05, "loss": 0.0463, "step": 18020 }, { "grad_norm": 0.5631292462348938, "learning_rate": 9.182551071262541e-05, "loss": 0.0468, "step": 18030 }, { "grad_norm": 0.4337971806526184, "learning_rate": 9.181418187718218e-05, "loss": 0.0482, "step": 18040 }, { "grad_norm": 0.4812367558479309, "learning_rate": 9.180284589683888e-05, "loss": 0.0472, "step": 18050 }, { "grad_norm": 0.5391245484352112, "learning_rate": 9.17915027735325e-05, "loss": 0.0458, "step": 18060 }, { "grad_norm": 0.40092983841896057, "learning_rate": 9.178015250920133e-05, "loss": 0.0412, "step": 18070 }, { "grad_norm": 0.34256139397621155, "learning_rate": 9.176879510578477e-05, "loss": 0.0428, "step": 18080 }, { "grad_norm": 0.3337135314941406, "learning_rate": 9.17574305652235e-05, "loss": 0.0379, "step": 18090 }, { "grad_norm": 0.4394456744194031, "learning_rate": 9.174605888945942e-05, "loss": 0.0433, "step": 18100 }, { "grad_norm": 0.4418034553527832, "learning_rate": 9.173468008043564e-05, "loss": 0.0404, "step": 18110 }, { "grad_norm": 0.43073606491088867, "learning_rate": 9.172329414009648e-05, "loss": 0.0403, "step": 18120 }, { "grad_norm": 0.4158955514431, "learning_rate": 9.171190107038747e-05, "loss": 0.0421, "step": 18130 }, { "grad_norm": 0.41904377937316895, "learning_rate": 9.170050087325541e-05, "loss": 0.0314, "step": 18140 }, { "grad_norm": 0.5158021450042725, "learning_rate": 9.168909355064824e-05, "loss": 0.0426, "step": 18150 }, { "grad_norm": 0.39893800020217896, "learning_rate": 9.167767910451519e-05, "loss": 0.039, "step": 18160 }, { "grad_norm": 0.36894306540489197, "learning_rate": 9.166625753680669e-05, "loss": 0.0404, "step": 18170 }, { "grad_norm": 0.3936673402786255, "learning_rate": 9.165482884947431e-05, "loss": 0.0327, "step": 18180 }, { "grad_norm": 0.35857903957366943, "learning_rate": 9.164339304447098e-05, "loss": 0.0487, "step": 18190 }, { "grad_norm": 0.39024773240089417, "learning_rate": 9.163195012375072e-05, "loss": 0.0342, "step": 18200 }, { "grad_norm": 0.3914903700351715, "learning_rate": 9.16205000892688e-05, "loss": 0.0502, "step": 18210 }, { "grad_norm": 0.5752549767494202, "learning_rate": 9.160904294298175e-05, "loss": 0.0396, "step": 18220 }, { "grad_norm": 0.34225520491600037, "learning_rate": 9.159757868684727e-05, "loss": 0.0457, "step": 18230 }, { "grad_norm": 0.6955154538154602, "learning_rate": 9.15861073228243e-05, "loss": 0.0424, "step": 18240 }, { "grad_norm": 0.5163877010345459, "learning_rate": 9.157462885287296e-05, "loss": 0.0498, "step": 18250 }, { "grad_norm": 0.3548603653907776, "learning_rate": 9.156314327895461e-05, "loss": 0.04, "step": 18260 }, { "grad_norm": 0.38972151279449463, "learning_rate": 9.155165060303185e-05, "loss": 0.0459, "step": 18270 }, { "grad_norm": 0.41598814725875854, "learning_rate": 9.154015082706841e-05, "loss": 0.037, "step": 18280 }, { "grad_norm": 0.40248802304267883, "learning_rate": 9.152864395302936e-05, "loss": 0.044, "step": 18290 }, { "grad_norm": 0.36982810497283936, "learning_rate": 9.151712998288085e-05, "loss": 0.0349, "step": 18300 }, { "grad_norm": 0.31346267461776733, "learning_rate": 9.150560891859031e-05, "loss": 0.0388, "step": 18310 }, { "grad_norm": 0.4622703194618225, "learning_rate": 9.14940807621264e-05, "loss": 0.0398, "step": 18320 }, { "grad_norm": 0.34020817279815674, "learning_rate": 9.148254551545894e-05, "loss": 0.0412, "step": 18330 }, { "grad_norm": 0.48552337288856506, "learning_rate": 9.147100318055901e-05, "loss": 0.0386, "step": 18340 }, { "grad_norm": 0.5387587547302246, "learning_rate": 9.145945375939888e-05, "loss": 0.0431, "step": 18350 }, { "grad_norm": 0.3615393340587616, "learning_rate": 9.144789725395203e-05, "loss": 0.0441, "step": 18360 }, { "grad_norm": 0.41904065012931824, "learning_rate": 9.14363336661931e-05, "loss": 0.0447, "step": 18370 }, { "grad_norm": 0.567954421043396, "learning_rate": 9.142476299809806e-05, "loss": 0.0471, "step": 18380 }, { "grad_norm": 0.3253026306629181, "learning_rate": 9.1413185251644e-05, "loss": 0.0365, "step": 18390 }, { "grad_norm": 0.41266146302223206, "learning_rate": 9.140160042880923e-05, "loss": 0.0503, "step": 18400 }, { "grad_norm": 0.3201858103275299, "learning_rate": 9.139000853157327e-05, "loss": 0.0444, "step": 18410 }, { "grad_norm": 0.3968993127346039, "learning_rate": 9.137840956191688e-05, "loss": 0.0405, "step": 18420 }, { "grad_norm": 0.2949041426181793, "learning_rate": 9.136680352182199e-05, "loss": 0.0363, "step": 18430 }, { "grad_norm": 0.8908411264419556, "learning_rate": 9.135519041327177e-05, "loss": 0.0407, "step": 18440 }, { "grad_norm": 0.41982361674308777, "learning_rate": 9.134357023825058e-05, "loss": 0.049, "step": 18450 }, { "grad_norm": 0.44696784019470215, "learning_rate": 9.133194299874398e-05, "loss": 0.0399, "step": 18460 }, { "grad_norm": 0.4827864468097687, "learning_rate": 9.132030869673876e-05, "loss": 0.041, "step": 18470 }, { "grad_norm": 0.39074766635894775, "learning_rate": 9.130866733422288e-05, "loss": 0.045, "step": 18480 }, { "grad_norm": 0.6731314063072205, "learning_rate": 9.129701891318556e-05, "loss": 0.0417, "step": 18490 }, { "grad_norm": 0.36466553807258606, "learning_rate": 9.128536343561718e-05, "loss": 0.0492, "step": 18500 }, { "grad_norm": 0.3806072771549225, "learning_rate": 9.127370090350934e-05, "loss": 0.0466, "step": 18510 }, { "grad_norm": 0.458065390586853, "learning_rate": 9.126203131885487e-05, "loss": 0.0471, "step": 18520 }, { "grad_norm": 0.3479658365249634, "learning_rate": 9.125035468364775e-05, "loss": 0.0489, "step": 18530 }, { "grad_norm": 0.4994592070579529, "learning_rate": 9.123867099988322e-05, "loss": 0.0344, "step": 18540 }, { "grad_norm": 0.4191218614578247, "learning_rate": 9.122698026955769e-05, "loss": 0.046, "step": 18550 }, { "grad_norm": 0.22256433963775635, "learning_rate": 9.12152824946688e-05, "loss": 0.061, "step": 18560 }, { "grad_norm": 0.44135382771492004, "learning_rate": 9.120357767721538e-05, "loss": 0.045, "step": 18570 }, { "grad_norm": 0.428611695766449, "learning_rate": 9.119186581919745e-05, "loss": 0.0453, "step": 18580 }, { "grad_norm": 0.33834418654441833, "learning_rate": 9.118014692261624e-05, "loss": 0.0409, "step": 18590 }, { "grad_norm": 0.5126734972000122, "learning_rate": 9.116842098947422e-05, "loss": 0.0615, "step": 18600 }, { "grad_norm": 0.38379281759262085, "learning_rate": 9.115668802177499e-05, "loss": 0.0386, "step": 18610 }, { "grad_norm": 0.24229620397090912, "learning_rate": 9.114494802152342e-05, "loss": 0.045, "step": 18620 }, { "grad_norm": 0.5564348697662354, "learning_rate": 9.113320099072555e-05, "loss": 0.0722, "step": 18630 }, { "grad_norm": 0.35037893056869507, "learning_rate": 9.112144693138864e-05, "loss": 0.0535, "step": 18640 }, { "grad_norm": 0.38711419701576233, "learning_rate": 9.110968584552111e-05, "loss": 0.0459, "step": 18650 }, { "grad_norm": 0.5526251792907715, "learning_rate": 9.109791773513264e-05, "loss": 0.0491, "step": 18660 }, { "grad_norm": 0.44402429461479187, "learning_rate": 9.108614260223403e-05, "loss": 0.0473, "step": 18670 }, { "grad_norm": 0.35868385434150696, "learning_rate": 9.107436044883738e-05, "loss": 0.0493, "step": 18680 }, { "grad_norm": 0.35645008087158203, "learning_rate": 9.10625712769559e-05, "loss": 0.044, "step": 18690 }, { "grad_norm": 0.35701805353164673, "learning_rate": 9.105077508860406e-05, "loss": 0.039, "step": 18700 }, { "grad_norm": 0.2934526801109314, "learning_rate": 9.103897188579751e-05, "loss": 0.0386, "step": 18710 }, { "grad_norm": 0.30219101905822754, "learning_rate": 9.102716167055308e-05, "loss": 0.0328, "step": 18720 }, { "grad_norm": 0.45058363676071167, "learning_rate": 9.10153444448888e-05, "loss": 0.0745, "step": 18730 }, { "grad_norm": 0.434443861246109, "learning_rate": 9.100352021082393e-05, "loss": 0.0365, "step": 18740 }, { "grad_norm": 0.36147794127464294, "learning_rate": 9.099168897037891e-05, "loss": 0.0404, "step": 18750 }, { "grad_norm": 0.4186076819896698, "learning_rate": 9.097985072557538e-05, "loss": 0.0361, "step": 18760 }, { "grad_norm": 0.5373591184616089, "learning_rate": 9.096800547843615e-05, "loss": 0.0632, "step": 18770 }, { "grad_norm": 0.44378066062927246, "learning_rate": 9.095615323098526e-05, "loss": 0.0551, "step": 18780 }, { "grad_norm": 0.49191299080848694, "learning_rate": 9.094429398524795e-05, "loss": 0.039, "step": 18790 }, { "grad_norm": 0.38859590888023376, "learning_rate": 9.093242774325061e-05, "loss": 0.0421, "step": 18800 }, { "grad_norm": 0.40716567635536194, "learning_rate": 9.092055450702088e-05, "loss": 0.0421, "step": 18810 }, { "grad_norm": 0.4091922640800476, "learning_rate": 9.090867427858756e-05, "loss": 0.043, "step": 18820 }, { "grad_norm": 0.32719185948371887, "learning_rate": 9.089678705998066e-05, "loss": 0.0431, "step": 18830 }, { "grad_norm": 0.3887524902820587, "learning_rate": 9.088489285323139e-05, "loss": 0.0446, "step": 18840 }, { "grad_norm": 0.6597418189048767, "learning_rate": 9.087299166037212e-05, "loss": 0.0457, "step": 18850 }, { "grad_norm": 0.3988957405090332, "learning_rate": 9.086108348343647e-05, "loss": 0.0442, "step": 18860 }, { "grad_norm": 0.49797511100769043, "learning_rate": 9.08491683244592e-05, "loss": 0.0478, "step": 18870 }, { "grad_norm": 0.4283333122730255, "learning_rate": 9.08372461854763e-05, "loss": 0.0379, "step": 18880 }, { "grad_norm": 0.4506303668022156, "learning_rate": 9.082531706852492e-05, "loss": 0.0418, "step": 18890 }, { "grad_norm": 0.364088773727417, "learning_rate": 9.081338097564342e-05, "loss": 0.0525, "step": 18900 }, { "grad_norm": 0.26200050115585327, "learning_rate": 9.080143790887137e-05, "loss": 0.0369, "step": 18910 }, { "grad_norm": 0.6547635197639465, "learning_rate": 9.07894878702495e-05, "loss": 0.036, "step": 18920 }, { "grad_norm": 0.8427756428718567, "learning_rate": 9.077753086181974e-05, "loss": 0.05, "step": 18930 }, { "grad_norm": 0.4025103747844696, "learning_rate": 9.076556688562524e-05, "loss": 0.0395, "step": 18940 }, { "grad_norm": 0.3415917456150055, "learning_rate": 9.075359594371029e-05, "loss": 0.0383, "step": 18950 }, { "grad_norm": 0.38415732979774475, "learning_rate": 9.07416180381204e-05, "loss": 0.0392, "step": 18960 }, { "grad_norm": 0.3901430666446686, "learning_rate": 9.072963317090228e-05, "loss": 0.0457, "step": 18970 }, { "grad_norm": 0.3392031490802765, "learning_rate": 9.071764134410382e-05, "loss": 0.0367, "step": 18980 }, { "grad_norm": 0.6006662249565125, "learning_rate": 9.070564255977407e-05, "loss": 0.0357, "step": 18990 }, { "grad_norm": 0.41089072823524475, "learning_rate": 9.06936368199633e-05, "loss": 0.0423, "step": 19000 }, { "grad_norm": 0.5284126996994019, "learning_rate": 9.0681624126723e-05, "loss": 0.0445, "step": 19010 }, { "grad_norm": 0.5051010847091675, "learning_rate": 9.066960448210576e-05, "loss": 0.0462, "step": 19020 }, { "grad_norm": 0.3397650420665741, "learning_rate": 9.065757788816543e-05, "loss": 0.038, "step": 19030 }, { "grad_norm": 0.5800589919090271, "learning_rate": 9.064554434695705e-05, "loss": 0.0497, "step": 19040 }, { "grad_norm": 0.34031111001968384, "learning_rate": 9.063350386053677e-05, "loss": 0.0409, "step": 19050 }, { "grad_norm": 0.3582072854042053, "learning_rate": 9.062145643096202e-05, "loss": 0.0381, "step": 19060 }, { "grad_norm": 0.35749515891075134, "learning_rate": 9.060940206029136e-05, "loss": 0.0459, "step": 19070 }, { "grad_norm": 0.4245346784591675, "learning_rate": 9.059734075058457e-05, "loss": 0.0416, "step": 19080 }, { "grad_norm": 0.37145256996154785, "learning_rate": 9.058527250390257e-05, "loss": 0.0521, "step": 19090 }, { "grad_norm": 0.390747994184494, "learning_rate": 9.057319732230752e-05, "loss": 0.0524, "step": 19100 }, { "grad_norm": 0.34792423248291016, "learning_rate": 9.056111520786273e-05, "loss": 0.0376, "step": 19110 }, { "grad_norm": 0.46056920289993286, "learning_rate": 9.054902616263268e-05, "loss": 0.0409, "step": 19120 }, { "grad_norm": 0.334823876619339, "learning_rate": 9.05369301886831e-05, "loss": 0.0565, "step": 19130 }, { "grad_norm": 0.28689610958099365, "learning_rate": 9.052482728808083e-05, "loss": 0.0334, "step": 19140 }, { "grad_norm": 0.279452919960022, "learning_rate": 9.051271746289391e-05, "loss": 0.0312, "step": 19150 }, { "grad_norm": 0.48333939909935, "learning_rate": 9.050060071519162e-05, "loss": 0.0433, "step": 19160 }, { "grad_norm": 0.3874804377555847, "learning_rate": 9.048847704704437e-05, "loss": 0.0416, "step": 19170 }, { "grad_norm": 0.5726111531257629, "learning_rate": 9.047634646052376e-05, "loss": 0.0463, "step": 19180 }, { "grad_norm": 0.4343814253807068, "learning_rate": 9.046420895770256e-05, "loss": 0.0512, "step": 19190 }, { "grad_norm": 0.4427550435066223, "learning_rate": 9.045206454065473e-05, "loss": 0.0412, "step": 19200 }, { "grad_norm": 0.4175669550895691, "learning_rate": 9.043991321145546e-05, "loss": 0.0394, "step": 19210 }, { "grad_norm": 0.5042750835418701, "learning_rate": 9.042775497218105e-05, "loss": 0.0355, "step": 19220 }, { "grad_norm": 0.40339434146881104, "learning_rate": 9.041558982490901e-05, "loss": 0.0381, "step": 19230 }, { "grad_norm": 0.521183431148529, "learning_rate": 9.040341777171805e-05, "loss": 0.0394, "step": 19240 }, { "grad_norm": 0.35273662209510803, "learning_rate": 9.039123881468802e-05, "loss": 0.0374, "step": 19250 }, { "grad_norm": 0.28878238797187805, "learning_rate": 9.037905295589998e-05, "loss": 0.0345, "step": 19260 }, { "grad_norm": 0.4540098011493683, "learning_rate": 9.036686019743617e-05, "loss": 0.0291, "step": 19270 }, { "grad_norm": 0.3450894057750702, "learning_rate": 9.035466054137997e-05, "loss": 0.0357, "step": 19280 }, { "grad_norm": 0.5308599472045898, "learning_rate": 9.0342453989816e-05, "loss": 0.0531, "step": 19290 }, { "grad_norm": 0.4245809018611908, "learning_rate": 9.033024054483e-05, "loss": 0.0428, "step": 19300 }, { "grad_norm": 0.49515512585639954, "learning_rate": 9.031802020850894e-05, "loss": 0.0397, "step": 19310 }, { "grad_norm": 0.46518391370773315, "learning_rate": 9.030579298294092e-05, "loss": 0.0522, "step": 19320 }, { "grad_norm": 0.36934852600097656, "learning_rate": 9.029355887021524e-05, "loss": 0.0527, "step": 19330 }, { "grad_norm": 0.49372339248657227, "learning_rate": 9.028131787242238e-05, "loss": 0.0618, "step": 19340 }, { "grad_norm": 0.3982819616794586, "learning_rate": 9.026906999165399e-05, "loss": 0.0485, "step": 19350 }, { "grad_norm": 0.3700130581855774, "learning_rate": 9.025681523000291e-05, "loss": 0.0368, "step": 19360 }, { "grad_norm": 0.5362963676452637, "learning_rate": 9.024455358956315e-05, "loss": 0.0455, "step": 19370 }, { "grad_norm": 0.3822580575942993, "learning_rate": 9.023228507242984e-05, "loss": 0.0457, "step": 19380 }, { "grad_norm": 0.39580729603767395, "learning_rate": 9.022000968069937e-05, "loss": 0.0461, "step": 19390 }, { "grad_norm": 0.35900023579597473, "learning_rate": 9.020772741646928e-05, "loss": 0.0411, "step": 19400 }, { "grad_norm": 0.44313424825668335, "learning_rate": 9.019543828183826e-05, "loss": 0.0425, "step": 19410 }, { "grad_norm": 0.4226120114326477, "learning_rate": 9.018314227890616e-05, "loss": 0.0363, "step": 19420 }, { "grad_norm": 0.5131190419197083, "learning_rate": 9.017083940977408e-05, "loss": 0.0606, "step": 19430 }, { "grad_norm": 0.43083491921424866, "learning_rate": 9.015852967654422e-05, "loss": 0.0384, "step": 19440 }, { "grad_norm": 0.3796912133693695, "learning_rate": 9.014621308131996e-05, "loss": 0.0427, "step": 19450 }, { "grad_norm": 0.3892696797847748, "learning_rate": 9.01338896262059e-05, "loss": 0.037, "step": 19460 }, { "grad_norm": 0.2952183485031128, "learning_rate": 9.012155931330777e-05, "loss": 0.0423, "step": 19470 }, { "grad_norm": 0.31607189774513245, "learning_rate": 9.010922214473246e-05, "loss": 0.0351, "step": 19480 }, { "grad_norm": 0.42303600907325745, "learning_rate": 9.009687812258808e-05, "loss": 0.0496, "step": 19490 }, { "grad_norm": 0.6277573704719543, "learning_rate": 9.00845272489839e-05, "loss": 0.0444, "step": 19500 }, { "grad_norm": 0.4310658574104309, "learning_rate": 9.007216952603031e-05, "loss": 0.0521, "step": 19510 }, { "grad_norm": 0.44198742508888245, "learning_rate": 9.005980495583894e-05, "loss": 0.0353, "step": 19520 }, { "grad_norm": 0.40473464131355286, "learning_rate": 9.004743354052252e-05, "loss": 0.0383, "step": 19530 }, { "grad_norm": 0.41374117136001587, "learning_rate": 9.003505528219503e-05, "loss": 0.0363, "step": 19540 }, { "grad_norm": 0.5580984354019165, "learning_rate": 9.002267018297154e-05, "loss": 0.0463, "step": 19550 }, { "grad_norm": 0.5131754875183105, "learning_rate": 9.001027824496834e-05, "loss": 0.0488, "step": 19560 }, { "grad_norm": 0.3219498097896576, "learning_rate": 8.999787947030287e-05, "loss": 0.0392, "step": 19570 }, { "grad_norm": 0.33498093485832214, "learning_rate": 8.998547386109376e-05, "loss": 0.0475, "step": 19580 }, { "grad_norm": 0.4051853120326996, "learning_rate": 8.997306141946073e-05, "loss": 0.0391, "step": 19590 }, { "grad_norm": 0.48608678579330444, "learning_rate": 8.996064214752481e-05, "loss": 0.0459, "step": 19600 }, { "grad_norm": 0.42771753668785095, "learning_rate": 8.994821604740806e-05, "loss": 0.0351, "step": 19610 }, { "grad_norm": 0.37079453468322754, "learning_rate": 8.993578312123377e-05, "loss": 0.0352, "step": 19620 }, { "grad_norm": 0.34381189942359924, "learning_rate": 8.992334337112639e-05, "loss": 0.042, "step": 19630 }, { "grad_norm": 0.3798394203186035, "learning_rate": 8.991089679921154e-05, "loss": 0.0349, "step": 19640 }, { "grad_norm": 0.4353523552417755, "learning_rate": 8.989844340761599e-05, "loss": 0.0517, "step": 19650 }, { "grad_norm": 0.42989933490753174, "learning_rate": 8.988598319846768e-05, "loss": 0.0393, "step": 19660 }, { "grad_norm": 0.36076658964157104, "learning_rate": 8.987351617389574e-05, "loss": 0.0428, "step": 19670 }, { "grad_norm": 0.36550742387771606, "learning_rate": 8.98610423360304e-05, "loss": 0.0371, "step": 19680 }, { "grad_norm": 0.28797048330307007, "learning_rate": 8.984856168700317e-05, "loss": 0.0444, "step": 19690 }, { "grad_norm": 0.3806675374507904, "learning_rate": 8.983607422894658e-05, "loss": 0.0408, "step": 19700 }, { "grad_norm": 0.3094474971294403, "learning_rate": 8.982357996399442e-05, "loss": 0.045, "step": 19710 }, { "grad_norm": 0.4077465832233429, "learning_rate": 8.981107889428164e-05, "loss": 0.0404, "step": 19720 }, { "grad_norm": 0.4002154767513275, "learning_rate": 8.979857102194428e-05, "loss": 0.0337, "step": 19730 }, { "grad_norm": 0.4381018280982971, "learning_rate": 8.978605634911968e-05, "loss": 0.0397, "step": 19740 }, { "grad_norm": 0.5846157073974609, "learning_rate": 8.977353487794616e-05, "loss": 0.0368, "step": 19750 }, { "grad_norm": 0.393171489238739, "learning_rate": 8.976100661056334e-05, "loss": 0.0441, "step": 19760 }, { "grad_norm": 0.45622754096984863, "learning_rate": 8.974847154911197e-05, "loss": 0.0375, "step": 19770 }, { "grad_norm": 0.43446189165115356, "learning_rate": 8.973592969573393e-05, "loss": 0.0442, "step": 19780 }, { "grad_norm": 0.38047993183135986, "learning_rate": 8.972338105257228e-05, "loss": 0.0443, "step": 19790 }, { "grad_norm": 0.2366430163383484, "learning_rate": 8.971082562177125e-05, "loss": 0.0345, "step": 19800 }, { "grad_norm": 0.40020227432250977, "learning_rate": 8.96982634054762e-05, "loss": 0.038, "step": 19810 }, { "grad_norm": 0.32075077295303345, "learning_rate": 8.96856944058337e-05, "loss": 0.0397, "step": 19820 }, { "grad_norm": 0.3629084527492523, "learning_rate": 8.967311862499144e-05, "loss": 0.039, "step": 19830 }, { "grad_norm": 0.419228732585907, "learning_rate": 8.966053606509825e-05, "loss": 0.0349, "step": 19840 }, { "grad_norm": 0.5659062266349792, "learning_rate": 8.964794672830417e-05, "loss": 0.055, "step": 19850 }, { "grad_norm": 0.37536486983299255, "learning_rate": 8.963535061676038e-05, "loss": 0.0496, "step": 19860 }, { "grad_norm": 0.442333847284317, "learning_rate": 8.962274773261918e-05, "loss": 0.0365, "step": 19870 }, { "grad_norm": 0.4628711938858032, "learning_rate": 8.961013807803409e-05, "loss": 0.0475, "step": 19880 }, { "grad_norm": 0.4023626148700714, "learning_rate": 8.959752165515973e-05, "loss": 0.0424, "step": 19890 }, { "grad_norm": 0.5585499405860901, "learning_rate": 8.958489846615193e-05, "loss": 0.0619, "step": 19900 }, { "grad_norm": 0.5344948172569275, "learning_rate": 8.957226851316762e-05, "loss": 0.0467, "step": 19910 }, { "grad_norm": 0.3871253728866577, "learning_rate": 8.955963179836493e-05, "loss": 0.0382, "step": 19920 }, { "grad_norm": 0.4717368185520172, "learning_rate": 8.954698832390312e-05, "loss": 0.0443, "step": 19930 }, { "grad_norm": 0.3721955716609955, "learning_rate": 8.953433809194263e-05, "loss": 0.0417, "step": 19940 }, { "grad_norm": 0.28115203976631165, "learning_rate": 8.9521681104645e-05, "loss": 0.0405, "step": 19950 }, { "grad_norm": 0.46081745624542236, "learning_rate": 8.9509017364173e-05, "loss": 0.044, "step": 19960 }, { "grad_norm": 0.49454739689826965, "learning_rate": 8.949634687269052e-05, "loss": 0.0374, "step": 19970 }, { "grad_norm": 0.4819866716861725, "learning_rate": 8.948366963236259e-05, "loss": 0.0394, "step": 19980 }, { "grad_norm": 0.5404174327850342, "learning_rate": 8.947098564535538e-05, "loss": 0.0434, "step": 19990 }, { "grad_norm": 0.5549030303955078, "learning_rate": 8.945829491383627e-05, "loss": 0.0391, "step": 20000 }, { "grad_norm": 0.33149224519729614, "learning_rate": 8.944559743997374e-05, "loss": 0.0387, "step": 20010 }, { "grad_norm": 0.5526427626609802, "learning_rate": 8.943289322593746e-05, "loss": 0.0391, "step": 20020 }, { "grad_norm": 0.5497727394104004, "learning_rate": 8.942018227389821e-05, "loss": 0.0357, "step": 20030 }, { "grad_norm": 0.8431993126869202, "learning_rate": 8.940746458602795e-05, "loss": 0.0534, "step": 20040 }, { "grad_norm": 0.40910446643829346, "learning_rate": 8.939474016449979e-05, "loss": 0.043, "step": 20050 }, { "grad_norm": 0.31903451681137085, "learning_rate": 8.938200901148799e-05, "loss": 0.0345, "step": 20060 }, { "grad_norm": 0.49801188707351685, "learning_rate": 8.936927112916795e-05, "loss": 0.047, "step": 20070 }, { "grad_norm": 0.36579781770706177, "learning_rate": 8.935652651971622e-05, "loss": 0.04, "step": 20080 }, { "grad_norm": 0.35311415791511536, "learning_rate": 8.934377518531052e-05, "loss": 0.0479, "step": 20090 }, { "grad_norm": 0.4078524708747864, "learning_rate": 8.933101712812967e-05, "loss": 0.0503, "step": 20100 }, { "grad_norm": 0.309855192899704, "learning_rate": 8.931825235035374e-05, "loss": 0.038, "step": 20110 }, { "grad_norm": 0.4736672043800354, "learning_rate": 8.930548085416382e-05, "loss": 0.0392, "step": 20120 }, { "grad_norm": 0.41174647212028503, "learning_rate": 8.92927026417422e-05, "loss": 0.0417, "step": 20130 }, { "grad_norm": 0.4376612603664398, "learning_rate": 8.92799177152724e-05, "loss": 0.0385, "step": 20140 }, { "grad_norm": 0.3302028775215149, "learning_rate": 8.926712607693895e-05, "loss": 0.0384, "step": 20150 }, { "grad_norm": 0.32927483320236206, "learning_rate": 8.925432772892762e-05, "loss": 0.0442, "step": 20160 }, { "grad_norm": 0.4069021940231323, "learning_rate": 8.924152267342529e-05, "loss": 0.0485, "step": 20170 }, { "grad_norm": 0.355151891708374, "learning_rate": 8.922871091261998e-05, "loss": 0.0446, "step": 20180 }, { "grad_norm": 0.35080382227897644, "learning_rate": 8.92158924487009e-05, "loss": 0.0343, "step": 20190 }, { "grad_norm": 0.34027695655822754, "learning_rate": 8.920306728385834e-05, "loss": 0.0338, "step": 20200 }, { "grad_norm": 0.2777833640575409, "learning_rate": 8.919023542028379e-05, "loss": 0.0338, "step": 20210 }, { "grad_norm": 0.3543553352355957, "learning_rate": 8.917739686016988e-05, "loss": 0.0347, "step": 20220 }, { "grad_norm": 0.45109644532203674, "learning_rate": 8.916455160571033e-05, "loss": 0.0471, "step": 20230 }, { "grad_norm": 0.3469614088535309, "learning_rate": 8.915169965910008e-05, "loss": 0.0319, "step": 20240 }, { "grad_norm": 0.3374491333961487, "learning_rate": 8.913884102253514e-05, "loss": 0.0398, "step": 20250 }, { "grad_norm": 0.36243605613708496, "learning_rate": 8.912597569821273e-05, "loss": 0.0296, "step": 20260 }, { "grad_norm": 0.3787655234336853, "learning_rate": 8.911310368833118e-05, "loss": 0.0345, "step": 20270 }, { "grad_norm": 0.33360809087753296, "learning_rate": 8.910022499508994e-05, "loss": 0.0327, "step": 20280 }, { "grad_norm": 0.3269384801387787, "learning_rate": 8.908733962068965e-05, "loss": 0.0432, "step": 20290 }, { "grad_norm": 0.30784961581230164, "learning_rate": 8.907444756733207e-05, "loss": 0.036, "step": 20300 }, { "grad_norm": 0.37230145931243896, "learning_rate": 8.906154883722006e-05, "loss": 0.042, "step": 20310 }, { "grad_norm": 0.3359549641609192, "learning_rate": 8.904864343255773e-05, "loss": 0.0358, "step": 20320 }, { "grad_norm": 0.5069942474365234, "learning_rate": 8.90357313555502e-05, "loss": 0.0382, "step": 20330 }, { "grad_norm": 0.3485570251941681, "learning_rate": 8.90228126084038e-05, "loss": 0.0415, "step": 20340 }, { "grad_norm": 0.3968176543712616, "learning_rate": 8.900988719332601e-05, "loss": 0.0436, "step": 20350 }, { "grad_norm": 0.32882630825042725, "learning_rate": 8.899695511252542e-05, "loss": 0.0377, "step": 20360 }, { "grad_norm": 0.4254451394081116, "learning_rate": 8.898401636821176e-05, "loss": 0.0406, "step": 20370 }, { "grad_norm": 0.4643445312976837, "learning_rate": 8.897107096259593e-05, "loss": 0.0493, "step": 20380 }, { "grad_norm": 0.3911587595939636, "learning_rate": 8.895811889788994e-05, "loss": 0.0389, "step": 20390 }, { "grad_norm": 0.2816261053085327, "learning_rate": 8.894516017630692e-05, "loss": 0.0343, "step": 20400 }, { "grad_norm": 0.37938547134399414, "learning_rate": 8.893219480006118e-05, "loss": 0.0348, "step": 20410 }, { "grad_norm": 0.5443600416183472, "learning_rate": 8.891922277136817e-05, "loss": 0.0421, "step": 20420 }, { "grad_norm": 0.35430875420570374, "learning_rate": 8.890624409244441e-05, "loss": 0.0342, "step": 20430 }, { "grad_norm": 0.37309104204177856, "learning_rate": 8.889325876550763e-05, "loss": 0.0404, "step": 20440 }, { "grad_norm": 0.3720206022262573, "learning_rate": 8.888026679277666e-05, "loss": 0.042, "step": 20450 }, { "grad_norm": 0.24959403276443481, "learning_rate": 8.886726817647147e-05, "loss": 0.0583, "step": 20460 }, { "grad_norm": 0.30155885219573975, "learning_rate": 8.885426291881319e-05, "loss": 0.0493, "step": 20470 }, { "grad_norm": 0.3234788477420807, "learning_rate": 8.884125102202401e-05, "loss": 0.0354, "step": 20480 }, { "grad_norm": 0.38206207752227783, "learning_rate": 8.882823248832736e-05, "loss": 0.0317, "step": 20490 }, { "grad_norm": 0.30662328004837036, "learning_rate": 8.881520731994772e-05, "loss": 0.04, "step": 20500 }, { "grad_norm": 0.3901589810848236, "learning_rate": 8.880217551911077e-05, "loss": 0.0438, "step": 20510 }, { "grad_norm": 0.5472779870033264, "learning_rate": 8.878913708804323e-05, "loss": 0.042, "step": 20520 }, { "grad_norm": 0.3403833508491516, "learning_rate": 8.877609202897308e-05, "loss": 0.0392, "step": 20530 }, { "grad_norm": 0.38392430543899536, "learning_rate": 8.876304034412933e-05, "loss": 0.0399, "step": 20540 }, { "grad_norm": 0.3729916214942932, "learning_rate": 8.874998203574214e-05, "loss": 0.0419, "step": 20550 }, { "grad_norm": 0.45964497327804565, "learning_rate": 8.873691710604284e-05, "loss": 0.0397, "step": 20560 }, { "grad_norm": 0.45445677638053894, "learning_rate": 8.872384555726387e-05, "loss": 0.0351, "step": 20570 }, { "grad_norm": 0.339543879032135, "learning_rate": 8.871076739163878e-05, "loss": 0.0396, "step": 20580 }, { "grad_norm": 0.3592088520526886, "learning_rate": 8.86976826114023e-05, "loss": 0.0458, "step": 20590 }, { "grad_norm": 0.37025341391563416, "learning_rate": 8.868459121879023e-05, "loss": 0.0409, "step": 20600 }, { "grad_norm": 0.4097757935523987, "learning_rate": 8.867149321603956e-05, "loss": 0.0461, "step": 20610 }, { "grad_norm": 0.3962104320526123, "learning_rate": 8.865838860538835e-05, "loss": 0.0348, "step": 20620 }, { "grad_norm": 0.37460875511169434, "learning_rate": 8.864527738907585e-05, "loss": 0.0439, "step": 20630 }, { "grad_norm": 0.4889926016330719, "learning_rate": 8.863215956934239e-05, "loss": 0.0457, "step": 20640 }, { "grad_norm": 0.2942999303340912, "learning_rate": 8.861903514842947e-05, "loss": 0.0304, "step": 20650 }, { "grad_norm": 0.3444899618625641, "learning_rate": 8.860590412857966e-05, "loss": 0.0363, "step": 20660 }, { "grad_norm": 0.2595793306827545, "learning_rate": 8.85927665120367e-05, "loss": 0.0302, "step": 20670 }, { "grad_norm": 0.346626877784729, "learning_rate": 8.857962230104546e-05, "loss": 0.0327, "step": 20680 }, { "grad_norm": 0.36503303050994873, "learning_rate": 8.856647149785193e-05, "loss": 0.0351, "step": 20690 }, { "grad_norm": 0.3677816689014435, "learning_rate": 8.855331410470322e-05, "loss": 0.0372, "step": 20700 }, { "grad_norm": 0.3396622836589813, "learning_rate": 8.854015012384756e-05, "loss": 0.0341, "step": 20710 }, { "grad_norm": 0.3307208716869354, "learning_rate": 8.852697955753433e-05, "loss": 0.0345, "step": 20720 }, { "grad_norm": 0.3122066259384155, "learning_rate": 8.851380240801399e-05, "loss": 0.0433, "step": 20730 }, { "grad_norm": 0.33445286750793457, "learning_rate": 8.850061867753818e-05, "loss": 0.0297, "step": 20740 }, { "grad_norm": 0.48785197734832764, "learning_rate": 8.848742836835963e-05, "loss": 0.0473, "step": 20750 }, { "grad_norm": 0.288435161113739, "learning_rate": 8.847423148273221e-05, "loss": 0.044, "step": 20760 }, { "grad_norm": 0.4179409146308899, "learning_rate": 8.846102802291092e-05, "loss": 0.053, "step": 20770 }, { "grad_norm": 0.35416802763938904, "learning_rate": 8.844781799115183e-05, "loss": 0.0458, "step": 20780 }, { "grad_norm": 0.4387236535549164, "learning_rate": 8.84346013897122e-05, "loss": 0.0411, "step": 20790 }, { "grad_norm": 0.44602108001708984, "learning_rate": 8.842137822085038e-05, "loss": 0.0499, "step": 20800 }, { "grad_norm": 0.32024797797203064, "learning_rate": 8.840814848682585e-05, "loss": 0.0369, "step": 20810 }, { "grad_norm": 0.35211676359176636, "learning_rate": 8.83949121898992e-05, "loss": 0.0525, "step": 20820 }, { "grad_norm": 0.3279659152030945, "learning_rate": 8.838166933233217e-05, "loss": 0.0346, "step": 20830 }, { "grad_norm": 0.31267496943473816, "learning_rate": 8.83684199163876e-05, "loss": 0.0393, "step": 20840 }, { "grad_norm": 0.29671794176101685, "learning_rate": 8.835516394432943e-05, "loss": 0.0386, "step": 20850 }, { "grad_norm": 0.2922188341617584, "learning_rate": 8.834190141842276e-05, "loss": 0.0458, "step": 20860 }, { "grad_norm": 0.4073728919029236, "learning_rate": 8.83286323409338e-05, "loss": 0.0381, "step": 20870 }, { "grad_norm": 0.3905222713947296, "learning_rate": 8.831535671412986e-05, "loss": 0.0376, "step": 20880 }, { "grad_norm": 0.30172985792160034, "learning_rate": 8.830207454027938e-05, "loss": 0.0399, "step": 20890 }, { "grad_norm": 0.45012181997299194, "learning_rate": 8.828878582165192e-05, "loss": 0.0365, "step": 20900 }, { "grad_norm": 0.47062477469444275, "learning_rate": 8.827549056051818e-05, "loss": 0.0368, "step": 20910 }, { "grad_norm": 0.2687174379825592, "learning_rate": 8.826218875914993e-05, "loss": 0.0364, "step": 20920 }, { "grad_norm": 0.28657063841819763, "learning_rate": 8.82488804198201e-05, "loss": 0.0371, "step": 20930 }, { "grad_norm": 0.23140932619571686, "learning_rate": 8.82355655448027e-05, "loss": 0.0364, "step": 20940 }, { "grad_norm": 0.4065771698951721, "learning_rate": 8.822224413637293e-05, "loss": 0.039, "step": 20950 }, { "grad_norm": 0.3337465524673462, "learning_rate": 8.820891619680697e-05, "loss": 0.0368, "step": 20960 }, { "grad_norm": 0.6263610124588013, "learning_rate": 8.819558172838227e-05, "loss": 0.0353, "step": 20970 }, { "grad_norm": 0.43456727266311646, "learning_rate": 8.818224073337731e-05, "loss": 0.0309, "step": 20980 }, { "grad_norm": 0.3388811945915222, "learning_rate": 8.816889321407169e-05, "loss": 0.034, "step": 20990 }, { "grad_norm": 0.2987961769104004, "learning_rate": 8.815553917274615e-05, "loss": 0.0386, "step": 21000 }, { "grad_norm": 0.38154542446136475, "learning_rate": 8.81421786116825e-05, "loss": 0.0456, "step": 21010 }, { "grad_norm": 0.43998146057128906, "learning_rate": 8.812881153316373e-05, "loss": 0.0324, "step": 21020 }, { "grad_norm": 0.41894710063934326, "learning_rate": 8.81154379394739e-05, "loss": 0.0397, "step": 21030 }, { "grad_norm": 0.4314979910850525, "learning_rate": 8.810205783289818e-05, "loss": 0.0347, "step": 21040 }, { "grad_norm": 0.5356560945510864, "learning_rate": 8.808867121572286e-05, "loss": 0.0379, "step": 21050 }, { "grad_norm": 0.37457409501075745, "learning_rate": 8.807527809023537e-05, "loss": 0.042, "step": 21060 }, { "grad_norm": 0.2653178572654724, "learning_rate": 8.80618784587242e-05, "loss": 0.0395, "step": 21070 }, { "grad_norm": 0.4626311957836151, "learning_rate": 8.804847232347902e-05, "loss": 0.0406, "step": 21080 }, { "grad_norm": 0.3006801903247833, "learning_rate": 8.803505968679054e-05, "loss": 0.0433, "step": 21090 }, { "grad_norm": 0.44281521439552307, "learning_rate": 8.802164055095061e-05, "loss": 0.0513, "step": 21100 }, { "grad_norm": 0.3690878748893738, "learning_rate": 8.80082149182522e-05, "loss": 0.0382, "step": 21110 }, { "grad_norm": 0.33685705065727234, "learning_rate": 8.79947827909894e-05, "loss": 0.0332, "step": 21120 }, { "grad_norm": 0.350763738155365, "learning_rate": 8.798134417145738e-05, "loss": 0.0372, "step": 21130 }, { "grad_norm": 0.4123687446117401, "learning_rate": 8.796789906195243e-05, "loss": 0.0421, "step": 21140 }, { "grad_norm": 0.3923521637916565, "learning_rate": 8.795444746477195e-05, "loss": 0.0329, "step": 21150 }, { "grad_norm": 0.41628777980804443, "learning_rate": 8.794098938221446e-05, "loss": 0.0444, "step": 21160 }, { "grad_norm": 0.36355698108673096, "learning_rate": 8.792752481657957e-05, "loss": 0.0443, "step": 21170 }, { "grad_norm": 0.40379947423934937, "learning_rate": 8.791405377016802e-05, "loss": 0.0428, "step": 21180 }, { "grad_norm": 0.36929696798324585, "learning_rate": 8.790057624528163e-05, "loss": 0.0414, "step": 21190 }, { "grad_norm": 0.4474234879016876, "learning_rate": 8.788709224422333e-05, "loss": 0.0423, "step": 21200 }, { "grad_norm": 0.5102596879005432, "learning_rate": 8.787360176929717e-05, "loss": 0.0412, "step": 21210 }, { "grad_norm": 0.2829363942146301, "learning_rate": 8.786010482280834e-05, "loss": 0.0395, "step": 21220 }, { "grad_norm": 0.3719383478164673, "learning_rate": 8.784660140706306e-05, "loss": 0.0411, "step": 21230 }, { "grad_norm": 0.37302908301353455, "learning_rate": 8.783309152436872e-05, "loss": 0.0341, "step": 21240 }, { "grad_norm": 0.3861207664012909, "learning_rate": 8.781957517703375e-05, "loss": 0.0334, "step": 21250 }, { "grad_norm": 0.39261889457702637, "learning_rate": 8.780605236736776e-05, "loss": 0.0415, "step": 21260 }, { "grad_norm": 0.24428574740886688, "learning_rate": 8.779252309768142e-05, "loss": 0.0371, "step": 21270 }, { "grad_norm": 0.3089029788970947, "learning_rate": 8.777898737028652e-05, "loss": 0.0388, "step": 21280 }, { "grad_norm": 0.27195069193840027, "learning_rate": 8.776544518749591e-05, "loss": 0.0341, "step": 21290 }, { "grad_norm": 0.293862909078598, "learning_rate": 8.775189655162364e-05, "loss": 0.0345, "step": 21300 }, { "grad_norm": 0.4079696834087372, "learning_rate": 8.773834146498474e-05, "loss": 0.0351, "step": 21310 }, { "grad_norm": 0.26136258244514465, "learning_rate": 8.772477992989545e-05, "loss": 0.0358, "step": 21320 }, { "grad_norm": 0.4158865213394165, "learning_rate": 8.771121194867304e-05, "loss": 0.0465, "step": 21330 }, { "grad_norm": 0.33220401406288147, "learning_rate": 8.769763752363589e-05, "loss": 0.0344, "step": 21340 }, { "grad_norm": 0.289739727973938, "learning_rate": 8.768405665710352e-05, "loss": 0.0441, "step": 21350 }, { "grad_norm": 0.5160186886787415, "learning_rate": 8.767046935139655e-05, "loss": 0.0401, "step": 21360 }, { "grad_norm": 0.3328648805618286, "learning_rate": 8.765687560883666e-05, "loss": 0.0421, "step": 21370 }, { "grad_norm": 0.32451629638671875, "learning_rate": 8.764327543174664e-05, "loss": 0.0377, "step": 21380 }, { "grad_norm": 0.3272354006767273, "learning_rate": 8.762966882245038e-05, "loss": 0.0375, "step": 21390 }, { "grad_norm": 0.7250377535820007, "learning_rate": 8.761605578327291e-05, "loss": 0.0435, "step": 21400 }, { "grad_norm": 0.277083158493042, "learning_rate": 8.76024363165403e-05, "loss": 0.0429, "step": 21410 }, { "grad_norm": 0.28229814767837524, "learning_rate": 8.758881042457976e-05, "loss": 0.0436, "step": 21420 }, { "grad_norm": 0.4385519027709961, "learning_rate": 8.757517810971957e-05, "loss": 0.037, "step": 21430 }, { "grad_norm": 0.3120078444480896, "learning_rate": 8.756153937428913e-05, "loss": 0.0369, "step": 21440 }, { "grad_norm": 0.4635653793811798, "learning_rate": 8.754789422061889e-05, "loss": 0.046, "step": 21450 }, { "grad_norm": 0.3845532238483429, "learning_rate": 8.753424265104052e-05, "loss": 0.0366, "step": 21460 }, { "grad_norm": 0.482826828956604, "learning_rate": 8.752058466788659e-05, "loss": 0.0487, "step": 21470 }, { "grad_norm": 0.3433237373828888, "learning_rate": 8.750692027349097e-05, "loss": 0.0377, "step": 21480 }, { "grad_norm": 0.3931199312210083, "learning_rate": 8.749324947018847e-05, "loss": 0.0345, "step": 21490 }, { "grad_norm": 0.3656933307647705, "learning_rate": 8.747957226031507e-05, "loss": 0.0386, "step": 21500 }, { "grad_norm": 0.29934340715408325, "learning_rate": 8.746588864620787e-05, "loss": 0.0368, "step": 21510 }, { "grad_norm": 0.599582850933075, "learning_rate": 8.745219863020498e-05, "loss": 0.0335, "step": 21520 }, { "grad_norm": 0.33935433626174927, "learning_rate": 8.743850221464564e-05, "loss": 0.0317, "step": 21530 }, { "grad_norm": 0.3253304064273834, "learning_rate": 8.742479940187026e-05, "loss": 0.0309, "step": 21540 }, { "grad_norm": 0.48181459307670593, "learning_rate": 8.74110901942202e-05, "loss": 0.033, "step": 21550 }, { "grad_norm": 0.2970108687877655, "learning_rate": 8.739737459403803e-05, "loss": 0.0433, "step": 21560 }, { "grad_norm": 0.37270236015319824, "learning_rate": 8.738365260366737e-05, "loss": 0.0445, "step": 21570 }, { "grad_norm": 0.42366525530815125, "learning_rate": 8.736992422545292e-05, "loss": 0.0441, "step": 21580 }, { "grad_norm": 0.48490166664123535, "learning_rate": 8.73561894617405e-05, "loss": 0.0469, "step": 21590 }, { "grad_norm": 0.3056184649467468, "learning_rate": 8.734244831487697e-05, "loss": 0.0324, "step": 21600 }, { "grad_norm": 0.32153600454330444, "learning_rate": 8.732870078721035e-05, "loss": 0.0449, "step": 21610 }, { "grad_norm": 0.2698453962802887, "learning_rate": 8.731494688108972e-05, "loss": 0.0373, "step": 21620 }, { "grad_norm": 0.34820833802223206, "learning_rate": 8.730118659886523e-05, "loss": 0.0416, "step": 21630 }, { "grad_norm": 0.5052912831306458, "learning_rate": 8.728741994288814e-05, "loss": 0.0371, "step": 21640 }, { "grad_norm": 0.5246961712837219, "learning_rate": 8.727364691551079e-05, "loss": 0.041, "step": 21650 }, { "grad_norm": 0.3908124566078186, "learning_rate": 8.725986751908661e-05, "loss": 0.0419, "step": 21660 }, { "grad_norm": 0.4000665545463562, "learning_rate": 8.724608175597016e-05, "loss": 0.0377, "step": 21670 }, { "grad_norm": 0.5714309215545654, "learning_rate": 8.723228962851699e-05, "loss": 0.0369, "step": 21680 }, { "grad_norm": 0.48923230171203613, "learning_rate": 8.721849113908385e-05, "loss": 0.0422, "step": 21690 }, { "grad_norm": 0.45164069533348083, "learning_rate": 8.720468629002848e-05, "loss": 0.0413, "step": 21700 }, { "grad_norm": 0.49512404203414917, "learning_rate": 8.719087508370978e-05, "loss": 0.0376, "step": 21710 }, { "grad_norm": 0.634674608707428, "learning_rate": 8.717705752248772e-05, "loss": 0.053, "step": 21720 }, { "grad_norm": 0.45473358035087585, "learning_rate": 8.71632336087233e-05, "loss": 0.0358, "step": 21730 }, { "grad_norm": 0.4041622579097748, "learning_rate": 8.71494033447787e-05, "loss": 0.0356, "step": 21740 }, { "grad_norm": 0.5245629549026489, "learning_rate": 8.713556673301708e-05, "loss": 0.036, "step": 21750 }, { "grad_norm": 0.44192203879356384, "learning_rate": 8.712172377580278e-05, "loss": 0.0559, "step": 21760 }, { "grad_norm": 0.4277271330356598, "learning_rate": 8.710787447550114e-05, "loss": 0.0418, "step": 21770 }, { "grad_norm": 0.38026708364486694, "learning_rate": 8.70940188344787e-05, "loss": 0.0497, "step": 21780 }, { "grad_norm": 0.48571205139160156, "learning_rate": 8.708015685510293e-05, "loss": 0.0379, "step": 21790 }, { "grad_norm": 0.5698009729385376, "learning_rate": 8.706628853974252e-05, "loss": 0.0426, "step": 21800 }, { "grad_norm": 0.3613743782043457, "learning_rate": 8.705241389076715e-05, "loss": 0.0319, "step": 21810 }, { "grad_norm": 0.3295101523399353, "learning_rate": 8.703853291054764e-05, "loss": 0.0336, "step": 21820 }, { "grad_norm": 0.42778846621513367, "learning_rate": 8.702464560145587e-05, "loss": 0.039, "step": 21830 }, { "grad_norm": 0.40761303901672363, "learning_rate": 8.701075196586476e-05, "loss": 0.0414, "step": 21840 }, { "grad_norm": 0.38706302642822266, "learning_rate": 8.699685200614842e-05, "loss": 0.0375, "step": 21850 }, { "grad_norm": 0.5067746639251709, "learning_rate": 8.698294572468193e-05, "loss": 0.0442, "step": 21860 }, { "grad_norm": 0.32826972007751465, "learning_rate": 8.696903312384148e-05, "loss": 0.0445, "step": 21870 }, { "grad_norm": 0.3807474672794342, "learning_rate": 8.695511420600439e-05, "loss": 0.0353, "step": 21880 }, { "grad_norm": 0.36405205726623535, "learning_rate": 8.694118897354901e-05, "loss": 0.033, "step": 21890 }, { "grad_norm": 0.3376029431819916, "learning_rate": 8.692725742885478e-05, "loss": 0.0295, "step": 21900 }, { "grad_norm": 0.47054237127304077, "learning_rate": 8.691331957430221e-05, "loss": 0.0369, "step": 21910 }, { "grad_norm": 0.4220587909221649, "learning_rate": 8.68993754122729e-05, "loss": 0.0383, "step": 21920 }, { "grad_norm": 0.3092183768749237, "learning_rate": 8.688542494514955e-05, "loss": 0.0316, "step": 21930 }, { "grad_norm": 0.42614564299583435, "learning_rate": 8.68714681753159e-05, "loss": 0.0367, "step": 21940 }, { "grad_norm": 0.3428335189819336, "learning_rate": 8.685750510515676e-05, "loss": 0.0339, "step": 21950 }, { "grad_norm": 0.43880027532577515, "learning_rate": 8.684353573705805e-05, "loss": 0.0351, "step": 21960 }, { "grad_norm": 0.4605080783367157, "learning_rate": 8.682956007340677e-05, "loss": 0.0295, "step": 21970 }, { "grad_norm": 0.299229234457016, "learning_rate": 8.681557811659095e-05, "loss": 0.0356, "step": 21980 }, { "grad_norm": 0.38275766372680664, "learning_rate": 8.680158986899974e-05, "loss": 0.0336, "step": 21990 }, { "grad_norm": 0.3550286293029785, "learning_rate": 8.678759533302335e-05, "loss": 0.035, "step": 22000 }, { "grad_norm": 0.34762322902679443, "learning_rate": 8.677359451105308e-05, "loss": 0.032, "step": 22010 }, { "grad_norm": 0.47152459621429443, "learning_rate": 8.675958740548123e-05, "loss": 0.0409, "step": 22020 }, { "grad_norm": 0.29137736558914185, "learning_rate": 8.674557401870129e-05, "loss": 0.0419, "step": 22030 }, { "grad_norm": 0.3692927062511444, "learning_rate": 8.673155435310775e-05, "loss": 0.0361, "step": 22040 }, { "grad_norm": 0.4797598719596863, "learning_rate": 8.671752841109617e-05, "loss": 0.0359, "step": 22050 }, { "grad_norm": 0.4454215466976166, "learning_rate": 8.670349619506321e-05, "loss": 0.0386, "step": 22060 }, { "grad_norm": 0.382526159286499, "learning_rate": 8.66894577074066e-05, "loss": 0.0356, "step": 22070 }, { "grad_norm": 0.32761356234550476, "learning_rate": 8.667541295052513e-05, "loss": 0.0348, "step": 22080 }, { "grad_norm": 0.4731781482696533, "learning_rate": 8.666136192681865e-05, "loss": 0.0483, "step": 22090 }, { "grad_norm": 0.3482902944087982, "learning_rate": 8.664730463868811e-05, "loss": 0.041, "step": 22100 }, { "grad_norm": 0.4496508836746216, "learning_rate": 8.663324108853552e-05, "loss": 0.0422, "step": 22110 }, { "grad_norm": 0.5681977868080139, "learning_rate": 8.661917127876395e-05, "loss": 0.0521, "step": 22120 }, { "grad_norm": 0.39434975385665894, "learning_rate": 8.660509521177754e-05, "loss": 0.0402, "step": 22130 }, { "grad_norm": 0.27945321798324585, "learning_rate": 8.65910128899815e-05, "loss": 0.0349, "step": 22140 }, { "grad_norm": 0.3960422873497009, "learning_rate": 8.657692431578214e-05, "loss": 0.0422, "step": 22150 }, { "grad_norm": 0.33782070875167847, "learning_rate": 8.656282949158679e-05, "loss": 0.0369, "step": 22160 }, { "grad_norm": 0.3313688337802887, "learning_rate": 8.654872841980388e-05, "loss": 0.0294, "step": 22170 }, { "grad_norm": 0.38991284370422363, "learning_rate": 8.653462110284289e-05, "loss": 0.029, "step": 22180 }, { "grad_norm": 0.27778905630111694, "learning_rate": 8.652050754311437e-05, "loss": 0.0325, "step": 22190 }, { "grad_norm": 0.3619580566883087, "learning_rate": 8.650638774302995e-05, "loss": 0.0333, "step": 22200 }, { "grad_norm": 0.5232003927230835, "learning_rate": 8.649226170500233e-05, "loss": 0.0366, "step": 22210 }, { "grad_norm": 0.390627384185791, "learning_rate": 8.647812943144524e-05, "loss": 0.0319, "step": 22220 }, { "grad_norm": 0.3224712908267975, "learning_rate": 8.646399092477351e-05, "loss": 0.0325, "step": 22230 }, { "grad_norm": 0.5071769952774048, "learning_rate": 8.644984618740301e-05, "loss": 0.0435, "step": 22240 }, { "grad_norm": 0.34626415371894836, "learning_rate": 8.643569522175073e-05, "loss": 0.0318, "step": 22250 }, { "grad_norm": 0.31154200434684753, "learning_rate": 8.642153803023463e-05, "loss": 0.0358, "step": 22260 }, { "grad_norm": 0.4949910044670105, "learning_rate": 8.640737461527383e-05, "loss": 0.0323, "step": 22270 }, { "grad_norm": 0.37991777062416077, "learning_rate": 8.639320497928845e-05, "loss": 0.0348, "step": 22280 }, { "grad_norm": 0.38398733735084534, "learning_rate": 8.637902912469969e-05, "loss": 0.0289, "step": 22290 }, { "grad_norm": 0.44842788577079773, "learning_rate": 8.636484705392982e-05, "loss": 0.0365, "step": 22300 }, { "grad_norm": 0.3044371008872986, "learning_rate": 8.635065876940216e-05, "loss": 0.0358, "step": 22310 }, { "grad_norm": 0.4205302298069, "learning_rate": 8.633646427354112e-05, "loss": 0.0388, "step": 22320 }, { "grad_norm": 0.3108161985874176, "learning_rate": 8.632226356877213e-05, "loss": 0.0351, "step": 22330 }, { "grad_norm": 0.3272413909435272, "learning_rate": 8.630805665752173e-05, "loss": 0.0439, "step": 22340 }, { "grad_norm": 0.5301911234855652, "learning_rate": 8.629384354221748e-05, "loss": 0.0389, "step": 22350 }, { "grad_norm": 0.26830700039863586, "learning_rate": 8.627962422528797e-05, "loss": 0.0356, "step": 22360 }, { "grad_norm": 0.37260422110557556, "learning_rate": 8.626539870916296e-05, "loss": 0.0503, "step": 22370 }, { "grad_norm": 0.3503490388393402, "learning_rate": 8.625116699627317e-05, "loss": 0.0479, "step": 22380 }, { "grad_norm": 0.3178214728832245, "learning_rate": 8.623692908905041e-05, "loss": 0.0331, "step": 22390 }, { "grad_norm": 0.31625741720199585, "learning_rate": 8.622268498992755e-05, "loss": 0.0398, "step": 22400 }, { "grad_norm": 0.29613161087036133, "learning_rate": 8.620843470133851e-05, "loss": 0.048, "step": 22410 }, { "grad_norm": 0.39292919635772705, "learning_rate": 8.619417822571829e-05, "loss": 0.0379, "step": 22420 }, { "grad_norm": 0.36552563309669495, "learning_rate": 8.617991556550292e-05, "loss": 0.0295, "step": 22430 }, { "grad_norm": 0.3721220791339874, "learning_rate": 8.616564672312952e-05, "loss": 0.0408, "step": 22440 }, { "grad_norm": 0.40164613723754883, "learning_rate": 8.61513717010362e-05, "loss": 0.0411, "step": 22450 }, { "grad_norm": 0.38832470774650574, "learning_rate": 8.613709050166221e-05, "loss": 0.0407, "step": 22460 }, { "grad_norm": 0.27288657426834106, "learning_rate": 8.61228031274478e-05, "loss": 0.0449, "step": 22470 }, { "grad_norm": 0.504935085773468, "learning_rate": 8.610850958083431e-05, "loss": 0.0455, "step": 22480 }, { "grad_norm": 0.47945815324783325, "learning_rate": 8.609420986426409e-05, "loss": 0.0408, "step": 22490 }, { "grad_norm": 0.36659181118011475, "learning_rate": 8.60799039801806e-05, "loss": 0.0379, "step": 22500 }, { "grad_norm": 0.2601102590560913, "learning_rate": 8.606559193102828e-05, "loss": 0.0305, "step": 22510 }, { "grad_norm": 0.6070742607116699, "learning_rate": 8.605127371925273e-05, "loss": 0.0465, "step": 22520 }, { "grad_norm": 0.38940373063087463, "learning_rate": 8.603694934730047e-05, "loss": 0.0393, "step": 22530 }, { "grad_norm": 0.368605375289917, "learning_rate": 8.602261881761919e-05, "loss": 0.0269, "step": 22540 }, { "grad_norm": 0.2238048017024994, "learning_rate": 8.600828213265759e-05, "loss": 0.0434, "step": 22550 }, { "grad_norm": 0.45740368962287903, "learning_rate": 8.599393929486539e-05, "loss": 0.0475, "step": 22560 }, { "grad_norm": 0.36005738377571106, "learning_rate": 8.59795903066934e-05, "loss": 0.0349, "step": 22570 }, { "grad_norm": 0.3123902678489685, "learning_rate": 8.596523517059347e-05, "loss": 0.0338, "step": 22580 }, { "grad_norm": 0.30960169434547424, "learning_rate": 8.59508738890185e-05, "loss": 0.0301, "step": 22590 }, { "grad_norm": 0.27032989263534546, "learning_rate": 8.593650646442246e-05, "loss": 0.0315, "step": 22600 }, { "grad_norm": 0.371704638004303, "learning_rate": 8.59221328992603e-05, "loss": 0.0273, "step": 22610 }, { "grad_norm": 0.29466181993484497, "learning_rate": 8.590775319598813e-05, "loss": 0.0308, "step": 22620 }, { "grad_norm": 0.3284270763397217, "learning_rate": 8.589336735706301e-05, "loss": 0.0305, "step": 22630 }, { "grad_norm": 0.33095788955688477, "learning_rate": 8.587897538494307e-05, "loss": 0.0462, "step": 22640 }, { "grad_norm": 0.414468914270401, "learning_rate": 8.586457728208756e-05, "loss": 0.0387, "step": 22650 }, { "grad_norm": 0.5565128922462463, "learning_rate": 8.585017305095667e-05, "loss": 0.0398, "step": 22660 }, { "grad_norm": 0.3435116708278656, "learning_rate": 8.583576269401173e-05, "loss": 0.0381, "step": 22670 }, { "grad_norm": 0.35730457305908203, "learning_rate": 8.582134621371504e-05, "loss": 0.0363, "step": 22680 }, { "grad_norm": 1.0084251165390015, "learning_rate": 8.580692361253e-05, "loss": 0.0372, "step": 22690 }, { "grad_norm": 0.541100263595581, "learning_rate": 8.579249489292104e-05, "loss": 0.0334, "step": 22700 }, { "grad_norm": 0.3483295440673828, "learning_rate": 8.577806005735363e-05, "loss": 0.0422, "step": 22710 }, { "grad_norm": 0.33458319306373596, "learning_rate": 8.576361910829429e-05, "loss": 0.0303, "step": 22720 }, { "grad_norm": 0.6072381734848022, "learning_rate": 8.574917204821057e-05, "loss": 0.0409, "step": 22730 }, { "grad_norm": 0.3648937940597534, "learning_rate": 8.57347188795711e-05, "loss": 0.0365, "step": 22740 }, { "grad_norm": 0.6473912596702576, "learning_rate": 8.572025960484551e-05, "loss": 0.0394, "step": 22750 }, { "grad_norm": 0.4856742024421692, "learning_rate": 8.57057942265045e-05, "loss": 0.0472, "step": 22760 }, { "grad_norm": 0.43286922574043274, "learning_rate": 8.569132274701984e-05, "loss": 0.034, "step": 22770 }, { "grad_norm": 0.3766199052333832, "learning_rate": 8.567684516886427e-05, "loss": 0.0338, "step": 22780 }, { "grad_norm": 0.38779017329216003, "learning_rate": 8.56623614945116e-05, "loss": 0.0315, "step": 22790 }, { "grad_norm": 0.40756332874298096, "learning_rate": 8.564787172643675e-05, "loss": 0.0362, "step": 22800 }, { "grad_norm": 0.4271855354309082, "learning_rate": 8.563337586711559e-05, "loss": 0.0394, "step": 22810 }, { "grad_norm": 0.3853106200695038, "learning_rate": 8.561887391902506e-05, "loss": 0.0425, "step": 22820 }, { "grad_norm": 0.37835660576820374, "learning_rate": 8.560436588464316e-05, "loss": 0.0321, "step": 22830 }, { "grad_norm": 0.3545258045196533, "learning_rate": 8.55898517664489e-05, "loss": 0.044, "step": 22840 }, { "grad_norm": 0.4929758906364441, "learning_rate": 8.557533156692236e-05, "loss": 0.0479, "step": 22850 }, { "grad_norm": 0.37031278014183044, "learning_rate": 8.556080528854467e-05, "loss": 0.0296, "step": 22860 }, { "grad_norm": 0.4169692397117615, "learning_rate": 8.554627293379791e-05, "loss": 0.0323, "step": 22870 }, { "grad_norm": 0.48806944489479065, "learning_rate": 8.553173450516531e-05, "loss": 0.047, "step": 22880 }, { "grad_norm": 0.43234768509864807, "learning_rate": 8.551719000513108e-05, "loss": 0.0326, "step": 22890 }, { "grad_norm": 0.3956391215324402, "learning_rate": 8.550263943618049e-05, "loss": 0.0477, "step": 22900 }, { "grad_norm": 0.4230270981788635, "learning_rate": 8.54880828007998e-05, "loss": 0.0432, "step": 22910 }, { "grad_norm": 0.33883750438690186, "learning_rate": 8.547352010147637e-05, "loss": 0.0386, "step": 22920 }, { "grad_norm": 0.30228501558303833, "learning_rate": 8.545895134069855e-05, "loss": 0.0338, "step": 22930 }, { "grad_norm": 0.3997485935688019, "learning_rate": 8.544437652095576e-05, "loss": 0.0323, "step": 22940 }, { "grad_norm": 0.4203163981437683, "learning_rate": 8.542979564473843e-05, "loss": 0.0449, "step": 22950 }, { "grad_norm": 0.401949018239975, "learning_rate": 8.541520871453802e-05, "loss": 0.034, "step": 22960 }, { "grad_norm": 0.7144026160240173, "learning_rate": 8.540061573284705e-05, "loss": 0.0298, "step": 22970 }, { "grad_norm": 0.3676210343837738, "learning_rate": 8.538601670215906e-05, "loss": 0.0329, "step": 22980 }, { "grad_norm": 0.3287009298801422, "learning_rate": 8.537141162496864e-05, "loss": 0.0363, "step": 22990 }, { "grad_norm": 0.33524852991104126, "learning_rate": 8.535680050377137e-05, "loss": 0.0323, "step": 23000 }, { "grad_norm": 0.2951144874095917, "learning_rate": 8.534218334106391e-05, "loss": 0.0345, "step": 23010 }, { "grad_norm": 0.39526477456092834, "learning_rate": 8.532756013934393e-05, "loss": 0.0402, "step": 23020 }, { "grad_norm": 0.3584884703159332, "learning_rate": 8.531293090111012e-05, "loss": 0.0408, "step": 23030 }, { "grad_norm": 0.36708399653434753, "learning_rate": 8.529829562886225e-05, "loss": 0.035, "step": 23040 }, { "grad_norm": 0.39188262820243835, "learning_rate": 8.528365432510105e-05, "loss": 0.0381, "step": 23050 }, { "grad_norm": 0.31744009256362915, "learning_rate": 8.526900699232833e-05, "loss": 0.0341, "step": 23060 }, { "grad_norm": 0.4604536294937134, "learning_rate": 8.525435363304695e-05, "loss": 0.0332, "step": 23070 }, { "grad_norm": 0.425444632768631, "learning_rate": 8.523969424976072e-05, "loss": 0.0324, "step": 23080 }, { "grad_norm": 0.28866690397262573, "learning_rate": 8.522502884497457e-05, "loss": 0.0304, "step": 23090 }, { "grad_norm": 0.3182816505432129, "learning_rate": 8.521035742119437e-05, "loss": 0.0338, "step": 23100 }, { "grad_norm": 0.272297739982605, "learning_rate": 8.519567998092712e-05, "loss": 0.0469, "step": 23110 }, { "grad_norm": 0.33294275403022766, "learning_rate": 8.518099652668075e-05, "loss": 0.0352, "step": 23120 }, { "grad_norm": 0.24533827602863312, "learning_rate": 8.516630706096429e-05, "loss": 0.0297, "step": 23130 }, { "grad_norm": 0.37927624583244324, "learning_rate": 8.515161158628773e-05, "loss": 0.0276, "step": 23140 }, { "grad_norm": 0.304470419883728, "learning_rate": 8.513691010516216e-05, "loss": 0.0512, "step": 23150 }, { "grad_norm": 0.29319876432418823, "learning_rate": 8.512220262009966e-05, "loss": 0.038, "step": 23160 }, { "grad_norm": 0.5112755298614502, "learning_rate": 8.510748913361332e-05, "loss": 0.0357, "step": 23170 }, { "grad_norm": 0.4239789545536041, "learning_rate": 8.509276964821726e-05, "loss": 0.0297, "step": 23180 }, { "grad_norm": 0.37936931848526, "learning_rate": 8.507804416642669e-05, "loss": 0.0332, "step": 23190 }, { "grad_norm": 0.31101587414741516, "learning_rate": 8.506331269075774e-05, "loss": 0.0318, "step": 23200 }, { "grad_norm": 0.5301318168640137, "learning_rate": 8.504857522372765e-05, "loss": 0.0301, "step": 23210 }, { "grad_norm": 0.4282008111476898, "learning_rate": 8.503383176785461e-05, "loss": 0.0313, "step": 23220 }, { "grad_norm": 0.34159594774246216, "learning_rate": 8.501908232565792e-05, "loss": 0.0323, "step": 23230 }, { "grad_norm": 0.4723248779773712, "learning_rate": 8.50043268996578e-05, "loss": 0.0351, "step": 23240 }, { "grad_norm": 0.3464142084121704, "learning_rate": 8.498956549237562e-05, "loss": 0.0383, "step": 23250 }, { "grad_norm": 0.29743626713752747, "learning_rate": 8.497479810633366e-05, "loss": 0.0321, "step": 23260 }, { "grad_norm": 0.4639873504638672, "learning_rate": 8.496002474405525e-05, "loss": 0.0369, "step": 23270 }, { "grad_norm": 0.4373556673526764, "learning_rate": 8.494524540806478e-05, "loss": 0.0386, "step": 23280 }, { "grad_norm": 0.36152249574661255, "learning_rate": 8.493046010088761e-05, "loss": 0.0511, "step": 23290 }, { "grad_norm": 0.8016798496246338, "learning_rate": 8.491566882505018e-05, "loss": 0.0506, "step": 23300 }, { "grad_norm": 0.5558517575263977, "learning_rate": 8.490087158307988e-05, "loss": 0.0796, "step": 23310 }, { "grad_norm": 0.6523575782775879, "learning_rate": 8.488606837750518e-05, "loss": 0.048, "step": 23320 }, { "grad_norm": 0.3804439604282379, "learning_rate": 8.487125921085552e-05, "loss": 0.0387, "step": 23330 }, { "grad_norm": 0.4459916651248932, "learning_rate": 8.485644408566141e-05, "loss": 0.0405, "step": 23340 }, { "grad_norm": 0.3445441424846649, "learning_rate": 8.484162300445431e-05, "loss": 0.0369, "step": 23350 }, { "grad_norm": 0.40217629075050354, "learning_rate": 8.482679596976676e-05, "loss": 0.0326, "step": 23360 }, { "grad_norm": 0.3349483907222748, "learning_rate": 8.48119629841323e-05, "loss": 0.0404, "step": 23370 }, { "grad_norm": 0.472002238035202, "learning_rate": 8.479712405008547e-05, "loss": 0.0466, "step": 23380 }, { "grad_norm": 0.49185463786125183, "learning_rate": 8.478227917016184e-05, "loss": 0.0498, "step": 23390 }, { "grad_norm": 0.27812016010284424, "learning_rate": 8.476742834689801e-05, "loss": 0.0389, "step": 23400 }, { "grad_norm": 0.3540708124637604, "learning_rate": 8.475257158283157e-05, "loss": 0.045, "step": 23410 }, { "grad_norm": 0.36846959590911865, "learning_rate": 8.473770888050112e-05, "loss": 0.0291, "step": 23420 }, { "grad_norm": 0.6235247850418091, "learning_rate": 8.47228402424463e-05, "loss": 0.038, "step": 23430 }, { "grad_norm": 0.35385024547576904, "learning_rate": 8.470796567120775e-05, "loss": 0.0303, "step": 23440 }, { "grad_norm": 0.40870431065559387, "learning_rate": 8.469308516932714e-05, "loss": 0.0356, "step": 23450 }, { "grad_norm": 0.6022391319274902, "learning_rate": 8.467819873934714e-05, "loss": 0.036, "step": 23460 }, { "grad_norm": 0.37270864844322205, "learning_rate": 8.466330638381143e-05, "loss": 0.0368, "step": 23470 }, { "grad_norm": 0.30746039748191833, "learning_rate": 8.464840810526469e-05, "loss": 0.0324, "step": 23480 }, { "grad_norm": 0.3661373257637024, "learning_rate": 8.463350390625264e-05, "loss": 0.0446, "step": 23490 }, { "grad_norm": 0.3532007038593292, "learning_rate": 8.4618593789322e-05, "loss": 0.066, "step": 23500 }, { "grad_norm": 0.6172322630882263, "learning_rate": 8.46036777570205e-05, "loss": 0.0485, "step": 23510 }, { "grad_norm": 0.4173557162284851, "learning_rate": 8.458875581189688e-05, "loss": 0.059, "step": 23520 }, { "grad_norm": 0.32053738832473755, "learning_rate": 8.457382795650092e-05, "loss": 0.034, "step": 23530 }, { "grad_norm": 0.2597585618495941, "learning_rate": 8.455889419338335e-05, "loss": 0.0363, "step": 23540 }, { "grad_norm": 0.29360559582710266, "learning_rate": 8.454395452509593e-05, "loss": 0.0332, "step": 23550 }, { "grad_norm": 0.339298278093338, "learning_rate": 8.452900895419146e-05, "loss": 0.0336, "step": 23560 }, { "grad_norm": 0.22686687111854553, "learning_rate": 8.451405748322376e-05, "loss": 0.0348, "step": 23570 }, { "grad_norm": 0.33911970257759094, "learning_rate": 8.449910011474759e-05, "loss": 0.0285, "step": 23580 }, { "grad_norm": 0.5031087398529053, "learning_rate": 8.448413685131876e-05, "loss": 0.0458, "step": 23590 }, { "grad_norm": 0.37361767888069153, "learning_rate": 8.446916769549407e-05, "loss": 0.0416, "step": 23600 }, { "grad_norm": 0.4195674955844879, "learning_rate": 8.445419264983136e-05, "loss": 0.0407, "step": 23610 }, { "grad_norm": 0.3262428939342499, "learning_rate": 8.443921171688947e-05, "loss": 0.0365, "step": 23620 }, { "grad_norm": 0.34357061982154846, "learning_rate": 8.442422489922819e-05, "loss": 0.035, "step": 23630 }, { "grad_norm": 0.3937015235424042, "learning_rate": 8.440923219940838e-05, "loss": 0.0445, "step": 23640 }, { "grad_norm": 0.514778196811676, "learning_rate": 8.439423361999189e-05, "loss": 0.039, "step": 23650 }, { "grad_norm": 0.3950197696685791, "learning_rate": 8.437922916354155e-05, "loss": 0.0411, "step": 23660 }, { "grad_norm": 0.24882864952087402, "learning_rate": 8.436421883262123e-05, "loss": 0.0318, "step": 23670 }, { "grad_norm": 0.30667644739151, "learning_rate": 8.434920262979577e-05, "loss": 0.0435, "step": 23680 }, { "grad_norm": 0.3480653166770935, "learning_rate": 8.433418055763104e-05, "loss": 0.0328, "step": 23690 }, { "grad_norm": 0.376293420791626, "learning_rate": 8.431915261869389e-05, "loss": 0.0467, "step": 23700 }, { "grad_norm": 0.2638557553291321, "learning_rate": 8.43041188155522e-05, "loss": 0.034, "step": 23710 }, { "grad_norm": 0.31432637572288513, "learning_rate": 8.428907915077481e-05, "loss": 0.0342, "step": 23720 }, { "grad_norm": 0.39848074316978455, "learning_rate": 8.42740336269316e-05, "loss": 0.0364, "step": 23730 }, { "grad_norm": 0.47237396240234375, "learning_rate": 8.425898224659345e-05, "loss": 0.0352, "step": 23740 }, { "grad_norm": 0.32952624559402466, "learning_rate": 8.42439250123322e-05, "loss": 0.0395, "step": 23750 }, { "grad_norm": 0.4240806996822357, "learning_rate": 8.422886192672076e-05, "loss": 0.0395, "step": 23760 }, { "grad_norm": 0.32569316029548645, "learning_rate": 8.421379299233297e-05, "loss": 0.0353, "step": 23770 }, { "grad_norm": 0.607617974281311, "learning_rate": 8.419871821174371e-05, "loss": 0.0377, "step": 23780 }, { "grad_norm": 0.4549693763256073, "learning_rate": 8.418363758752884e-05, "loss": 0.0372, "step": 23790 }, { "grad_norm": 0.3257063925266266, "learning_rate": 8.416855112226523e-05, "loss": 0.0355, "step": 23800 }, { "grad_norm": 0.34357157349586487, "learning_rate": 8.415345881853075e-05, "loss": 0.0327, "step": 23810 }, { "grad_norm": 0.35048365592956543, "learning_rate": 8.413836067890426e-05, "loss": 0.0306, "step": 23820 }, { "grad_norm": 0.6846439838409424, "learning_rate": 8.41232567059656e-05, "loss": 0.0387, "step": 23830 }, { "grad_norm": 0.39052814245224, "learning_rate": 8.410814690229565e-05, "loss": 0.0408, "step": 23840 }, { "grad_norm": 0.6046937704086304, "learning_rate": 8.409303127047626e-05, "loss": 0.0366, "step": 23850 }, { "grad_norm": 0.43331626057624817, "learning_rate": 8.407790981309028e-05, "loss": 0.037, "step": 23860 }, { "grad_norm": 0.47038042545318604, "learning_rate": 8.406278253272153e-05, "loss": 0.0444, "step": 23870 }, { "grad_norm": 0.5197181105613708, "learning_rate": 8.404764943195487e-05, "loss": 0.0343, "step": 23880 }, { "grad_norm": 0.29145362973213196, "learning_rate": 8.403251051337613e-05, "loss": 0.0431, "step": 23890 }, { "grad_norm": 0.5540750622749329, "learning_rate": 8.401736577957214e-05, "loss": 0.0305, "step": 23900 }, { "grad_norm": 0.3339332044124603, "learning_rate": 8.40022152331307e-05, "loss": 0.0337, "step": 23910 }, { "grad_norm": 0.4900719225406647, "learning_rate": 8.398705887664064e-05, "loss": 0.0401, "step": 23920 }, { "grad_norm": 0.24710170924663544, "learning_rate": 8.397189671269177e-05, "loss": 0.0286, "step": 23930 }, { "grad_norm": 0.375789076089859, "learning_rate": 8.395672874387488e-05, "loss": 0.0315, "step": 23940 }, { "grad_norm": 0.3652206361293793, "learning_rate": 8.394155497278177e-05, "loss": 0.0275, "step": 23950 }, { "grad_norm": 0.6527431011199951, "learning_rate": 8.392637540200523e-05, "loss": 0.0422, "step": 23960 }, { "grad_norm": 0.29486197233200073, "learning_rate": 8.391119003413902e-05, "loss": 0.0342, "step": 23970 }, { "grad_norm": 0.3955511450767517, "learning_rate": 8.38959988717779e-05, "loss": 0.0376, "step": 23980 }, { "grad_norm": 0.3493586778640747, "learning_rate": 8.388080191751764e-05, "loss": 0.0401, "step": 23990 }, { "grad_norm": 0.3850904703140259, "learning_rate": 8.386559917395496e-05, "loss": 0.0328, "step": 24000 }, { "grad_norm": 0.3187893033027649, "learning_rate": 8.385039064368761e-05, "loss": 0.0259, "step": 24010 }, { "grad_norm": 0.34164494276046753, "learning_rate": 8.383517632931431e-05, "loss": 0.0285, "step": 24020 }, { "grad_norm": 0.4657047986984253, "learning_rate": 8.381995623343477e-05, "loss": 0.0313, "step": 24030 }, { "grad_norm": 0.4104115664958954, "learning_rate": 8.380473035864968e-05, "loss": 0.0386, "step": 24040 }, { "grad_norm": 0.38547542691230774, "learning_rate": 8.378949870756076e-05, "loss": 0.0427, "step": 24050 }, { "grad_norm": 0.25340569019317627, "learning_rate": 8.377426128277063e-05, "loss": 0.0383, "step": 24060 }, { "grad_norm": 0.23826098442077637, "learning_rate": 8.375901808688298e-05, "loss": 0.0305, "step": 24070 }, { "grad_norm": 0.45561879873275757, "learning_rate": 8.374376912250246e-05, "loss": 0.0385, "step": 24080 }, { "grad_norm": 0.4182363748550415, "learning_rate": 8.372851439223468e-05, "loss": 0.0343, "step": 24090 }, { "grad_norm": 0.3299487233161926, "learning_rate": 8.371325389868627e-05, "loss": 0.0449, "step": 24100 }, { "grad_norm": 0.44388869404792786, "learning_rate": 8.369798764446482e-05, "loss": 0.0334, "step": 24110 }, { "grad_norm": 0.4987911880016327, "learning_rate": 8.368271563217893e-05, "loss": 0.0337, "step": 24120 }, { "grad_norm": 0.31179502606391907, "learning_rate": 8.366743786443817e-05, "loss": 0.0492, "step": 24130 }, { "grad_norm": 0.40050578117370605, "learning_rate": 8.365215434385309e-05, "loss": 0.0402, "step": 24140 }, { "grad_norm": 0.31826359033584595, "learning_rate": 8.36368650730352e-05, "loss": 0.0316, "step": 24150 }, { "grad_norm": 0.367806613445282, "learning_rate": 8.362157005459705e-05, "loss": 0.0376, "step": 24160 }, { "grad_norm": 0.29467371106147766, "learning_rate": 8.360626929115213e-05, "loss": 0.0316, "step": 24170 }, { "grad_norm": 0.3771342933177948, "learning_rate": 8.359096278531492e-05, "loss": 0.0366, "step": 24180 }, { "grad_norm": 0.3197535574436188, "learning_rate": 8.357565053970088e-05, "loss": 0.0319, "step": 24190 }, { "grad_norm": 0.32972419261932373, "learning_rate": 8.356033255692647e-05, "loss": 0.035, "step": 24200 }, { "grad_norm": 0.46204426884651184, "learning_rate": 8.354500883960911e-05, "loss": 0.031, "step": 24210 }, { "grad_norm": 0.2741570770740509, "learning_rate": 8.352967939036717e-05, "loss": 0.0369, "step": 24220 }, { "grad_norm": 0.4253057539463043, "learning_rate": 8.35143442118201e-05, "loss": 0.0379, "step": 24230 }, { "grad_norm": 0.28053247928619385, "learning_rate": 8.349900330658819e-05, "loss": 0.0331, "step": 24240 }, { "grad_norm": 0.34981000423431396, "learning_rate": 8.348365667729284e-05, "loss": 0.0335, "step": 24250 }, { "grad_norm": 0.35523486137390137, "learning_rate": 8.346830432655633e-05, "loss": 0.0252, "step": 24260 }, { "grad_norm": 0.5161582827568054, "learning_rate": 8.345294625700195e-05, "loss": 0.0367, "step": 24270 }, { "grad_norm": 0.3665285110473633, "learning_rate": 8.343758247125402e-05, "loss": 0.0322, "step": 24280 }, { "grad_norm": 0.5161361694335938, "learning_rate": 8.342221297193776e-05, "loss": 0.0464, "step": 24290 }, { "grad_norm": 0.3646261394023895, "learning_rate": 8.34068377616794e-05, "loss": 0.0416, "step": 24300 }, { "grad_norm": 0.5203297138214111, "learning_rate": 8.339145684310615e-05, "loss": 0.0364, "step": 24310 }, { "grad_norm": 0.5605708956718445, "learning_rate": 8.337607021884618e-05, "loss": 0.0373, "step": 24320 }, { "grad_norm": 0.28379690647125244, "learning_rate": 8.336067789152867e-05, "loss": 0.0281, "step": 24330 }, { "grad_norm": 0.3265000283718109, "learning_rate": 8.334527986378369e-05, "loss": 0.038, "step": 24340 }, { "grad_norm": 0.30758529901504517, "learning_rate": 8.332987613824239e-05, "loss": 0.0425, "step": 24350 }, { "grad_norm": 0.4067215621471405, "learning_rate": 8.331446671753685e-05, "loss": 0.0404, "step": 24360 }, { "grad_norm": 0.33073654770851135, "learning_rate": 8.329905160430007e-05, "loss": 0.0303, "step": 24370 }, { "grad_norm": 0.7984419465065002, "learning_rate": 8.328363080116611e-05, "loss": 0.0341, "step": 24380 }, { "grad_norm": 0.2868640422821045, "learning_rate": 8.326820431076997e-05, "loss": 0.0278, "step": 24390 }, { "grad_norm": 0.399069607257843, "learning_rate": 8.325277213574759e-05, "loss": 0.0394, "step": 24400 }, { "grad_norm": 0.4909047782421112, "learning_rate": 8.32373342787359e-05, "loss": 0.0361, "step": 24410 }, { "grad_norm": 0.5291438102722168, "learning_rate": 8.322189074237285e-05, "loss": 0.0348, "step": 24420 }, { "grad_norm": 0.3189016580581665, "learning_rate": 8.32064415292973e-05, "loss": 0.0331, "step": 24430 }, { "grad_norm": 0.2996911406517029, "learning_rate": 8.319098664214907e-05, "loss": 0.0299, "step": 24440 }, { "grad_norm": 0.26697850227355957, "learning_rate": 8.3175526083569e-05, "loss": 0.0385, "step": 24450 }, { "grad_norm": 0.4153609275817871, "learning_rate": 8.316005985619889e-05, "loss": 0.0363, "step": 24460 }, { "grad_norm": 0.33402568101882935, "learning_rate": 8.314458796268147e-05, "loss": 0.0342, "step": 24470 }, { "grad_norm": 0.3176194131374359, "learning_rate": 8.312911040566047e-05, "loss": 0.0279, "step": 24480 }, { "grad_norm": 0.3693414032459259, "learning_rate": 8.31136271877806e-05, "loss": 0.0329, "step": 24490 }, { "grad_norm": 0.4596292972564697, "learning_rate": 8.309813831168748e-05, "loss": 0.0378, "step": 24500 }, { "grad_norm": 0.3478022515773773, "learning_rate": 8.308264378002777e-05, "loss": 0.0245, "step": 24510 }, { "grad_norm": 0.4631979167461395, "learning_rate": 8.306714359544906e-05, "loss": 0.0425, "step": 24520 }, { "grad_norm": 0.32751035690307617, "learning_rate": 8.30516377605999e-05, "loss": 0.0277, "step": 24530 }, { "grad_norm": 0.2847299575805664, "learning_rate": 8.30361262781298e-05, "loss": 0.0278, "step": 24540 }, { "grad_norm": 0.42948123812675476, "learning_rate": 8.302060915068924e-05, "loss": 0.0386, "step": 24550 }, { "grad_norm": 0.3134571313858032, "learning_rate": 8.300508638092972e-05, "loss": 0.0296, "step": 24560 }, { "grad_norm": 0.4930594563484192, "learning_rate": 8.298955797150361e-05, "loss": 0.0361, "step": 24570 }, { "grad_norm": 0.4244775176048279, "learning_rate": 8.297402392506433e-05, "loss": 0.0337, "step": 24580 }, { "grad_norm": 0.44179633259773254, "learning_rate": 8.295848424426617e-05, "loss": 0.0356, "step": 24590 }, { "grad_norm": 0.3005983233451843, "learning_rate": 8.29429389317645e-05, "loss": 0.0301, "step": 24600 }, { "grad_norm": 0.5451488494873047, "learning_rate": 8.292738799021556e-05, "loss": 0.0339, "step": 24610 }, { "grad_norm": 0.4296862483024597, "learning_rate": 8.291183142227656e-05, "loss": 0.0318, "step": 24620 }, { "grad_norm": 0.3163367509841919, "learning_rate": 8.289626923060572e-05, "loss": 0.0468, "step": 24630 }, { "grad_norm": 0.4533059597015381, "learning_rate": 8.288070141786218e-05, "loss": 0.0376, "step": 24640 }, { "grad_norm": 0.2763329744338989, "learning_rate": 8.286512798670605e-05, "loss": 0.039, "step": 24650 }, { "grad_norm": 0.4279070198535919, "learning_rate": 8.284954893979842e-05, "loss": 0.0386, "step": 24660 }, { "grad_norm": 0.24486766755580902, "learning_rate": 8.283396427980131e-05, "loss": 0.0398, "step": 24670 }, { "grad_norm": 0.3669859766960144, "learning_rate": 8.281837400937771e-05, "loss": 0.0353, "step": 24680 }, { "grad_norm": 0.3991248309612274, "learning_rate": 8.28027781311916e-05, "loss": 0.0308, "step": 24690 }, { "grad_norm": 0.36036235094070435, "learning_rate": 8.278717664790785e-05, "loss": 0.0334, "step": 24700 }, { "grad_norm": 0.40245935320854187, "learning_rate": 8.277156956219234e-05, "loss": 0.0362, "step": 24710 }, { "grad_norm": 0.5230724811553955, "learning_rate": 8.275595687671189e-05, "loss": 0.0293, "step": 24720 }, { "grad_norm": 0.49990639090538025, "learning_rate": 8.27403385941343e-05, "loss": 0.0309, "step": 24730 }, { "grad_norm": 0.40005606412887573, "learning_rate": 8.272471471712828e-05, "loss": 0.0363, "step": 24740 }, { "grad_norm": 0.21881301701068878, "learning_rate": 8.270908524836355e-05, "loss": 0.0339, "step": 24750 }, { "grad_norm": 0.38071495294570923, "learning_rate": 8.269345019051074e-05, "loss": 0.0431, "step": 24760 }, { "grad_norm": 0.3938385844230652, "learning_rate": 8.267780954624147e-05, "loss": 0.0372, "step": 24770 }, { "grad_norm": 0.2998800575733185, "learning_rate": 8.266216331822827e-05, "loss": 0.036, "step": 24780 }, { "grad_norm": 0.4285912811756134, "learning_rate": 8.264651150914469e-05, "loss": 0.0386, "step": 24790 }, { "grad_norm": 0.25407832860946655, "learning_rate": 8.263085412166517e-05, "loss": 0.0298, "step": 24800 }, { "grad_norm": 0.3600650727748871, "learning_rate": 8.261519115846514e-05, "loss": 0.0277, "step": 24810 }, { "grad_norm": 0.2972981631755829, "learning_rate": 8.259952262222096e-05, "loss": 0.0283, "step": 24820 }, { "grad_norm": 0.2813568413257599, "learning_rate": 8.258384851560997e-05, "loss": 0.0316, "step": 24830 }, { "grad_norm": 0.37930113077163696, "learning_rate": 8.256816884131044e-05, "loss": 0.0354, "step": 24840 }, { "grad_norm": 0.3458172380924225, "learning_rate": 8.255248360200159e-05, "loss": 0.0322, "step": 24850 }, { "grad_norm": 0.3401415944099426, "learning_rate": 8.253679280036359e-05, "loss": 0.029, "step": 24860 }, { "grad_norm": 0.6155313849449158, "learning_rate": 8.252109643907762e-05, "loss": 0.0504, "step": 24870 }, { "grad_norm": 0.398642361164093, "learning_rate": 8.250539452082569e-05, "loss": 0.0314, "step": 24880 }, { "grad_norm": 0.3850824236869812, "learning_rate": 8.248968704829087e-05, "loss": 0.0335, "step": 24890 }, { "grad_norm": 0.40874406695365906, "learning_rate": 8.247397402415714e-05, "loss": 0.0368, "step": 24900 }, { "grad_norm": 0.3773896396160126, "learning_rate": 8.24582554511094e-05, "loss": 0.0369, "step": 24910 }, { "grad_norm": 0.3589361310005188, "learning_rate": 8.244253133183355e-05, "loss": 0.0341, "step": 24920 }, { "grad_norm": 0.3754473030567169, "learning_rate": 8.24268016690164e-05, "loss": 0.0368, "step": 24930 }, { "grad_norm": 0.49721091985702515, "learning_rate": 8.241106646534571e-05, "loss": 0.0379, "step": 24940 }, { "grad_norm": 0.3084474503993988, "learning_rate": 8.23953257235102e-05, "loss": 0.0337, "step": 24950 }, { "grad_norm": 0.4026710093021393, "learning_rate": 8.237957944619956e-05, "loss": 0.0338, "step": 24960 }, { "grad_norm": 0.42335274815559387, "learning_rate": 8.236382763610437e-05, "loss": 0.0335, "step": 24970 }, { "grad_norm": 0.43831807374954224, "learning_rate": 8.234807029591619e-05, "loss": 0.0343, "step": 24980 }, { "grad_norm": 0.4106408953666687, "learning_rate": 8.233230742832752e-05, "loss": 0.0479, "step": 24990 }, { "grad_norm": 0.46421632170677185, "learning_rate": 8.231653903603178e-05, "loss": 0.0379, "step": 25000 }, { "grad_norm": 0.2797313332557678, "learning_rate": 8.23007651217234e-05, "loss": 0.034, "step": 25010 }, { "grad_norm": 0.34977954626083374, "learning_rate": 8.228498568809769e-05, "loss": 0.0348, "step": 25020 }, { "grad_norm": 0.4902045428752899, "learning_rate": 8.22692007378509e-05, "loss": 0.0436, "step": 25030 }, { "grad_norm": 0.2723182141780853, "learning_rate": 8.225341027368028e-05, "loss": 0.0313, "step": 25040 }, { "grad_norm": 0.29193195700645447, "learning_rate": 8.223761429828399e-05, "loss": 0.0334, "step": 25050 }, { "grad_norm": 0.34016814827919006, "learning_rate": 8.22218128143611e-05, "loss": 0.0386, "step": 25060 }, { "grad_norm": 0.36899828910827637, "learning_rate": 8.220600582461166e-05, "loss": 0.0302, "step": 25070 }, { "grad_norm": 0.3418249487876892, "learning_rate": 8.219019333173668e-05, "loss": 0.035, "step": 25080 }, { "grad_norm": 0.45469531416893005, "learning_rate": 8.217437533843805e-05, "loss": 0.0418, "step": 25090 }, { "grad_norm": 0.25951772928237915, "learning_rate": 8.215855184741867e-05, "loss": 0.0334, "step": 25100 }, { "grad_norm": 0.2984091341495514, "learning_rate": 8.21427228613823e-05, "loss": 0.0386, "step": 25110 }, { "grad_norm": 0.3656175434589386, "learning_rate": 8.21268883830337e-05, "loss": 0.0393, "step": 25120 }, { "grad_norm": 0.5150753855705261, "learning_rate": 8.211104841507855e-05, "loss": 0.0428, "step": 25130 }, { "grad_norm": 0.31384435296058655, "learning_rate": 8.209520296022346e-05, "loss": 0.0355, "step": 25140 }, { "grad_norm": 0.44201886653900146, "learning_rate": 8.207935202117599e-05, "loss": 0.0314, "step": 25150 }, { "grad_norm": 0.27302488684654236, "learning_rate": 8.206349560064463e-05, "loss": 0.0334, "step": 25160 }, { "grad_norm": 0.3853662312030792, "learning_rate": 8.204763370133881e-05, "loss": 0.0291, "step": 25170 }, { "grad_norm": 0.348827987909317, "learning_rate": 8.203176632596892e-05, "loss": 0.0305, "step": 25180 }, { "grad_norm": 0.6675968170166016, "learning_rate": 8.20158934772462e-05, "loss": 0.0355, "step": 25190 }, { "grad_norm": 0.28294312953948975, "learning_rate": 8.200001515788294e-05, "loss": 0.0312, "step": 25200 }, { "grad_norm": 0.33844637870788574, "learning_rate": 8.198413137059228e-05, "loss": 0.0358, "step": 25210 }, { "grad_norm": 0.2761630117893219, "learning_rate": 8.196824211808835e-05, "loss": 0.0307, "step": 25220 }, { "grad_norm": 0.4019656777381897, "learning_rate": 8.195234740308617e-05, "loss": 0.0428, "step": 25230 }, { "grad_norm": 0.3803117275238037, "learning_rate": 8.193644722830171e-05, "loss": 0.0324, "step": 25240 }, { "grad_norm": 0.41030412912368774, "learning_rate": 8.19205415964519e-05, "loss": 0.0346, "step": 25250 }, { "grad_norm": 0.35395875573158264, "learning_rate": 8.190463051025456e-05, "loss": 0.0391, "step": 25260 }, { "grad_norm": 0.5022005438804626, "learning_rate": 8.188871397242843e-05, "loss": 0.0407, "step": 25270 }, { "grad_norm": 0.29723334312438965, "learning_rate": 8.187279198569326e-05, "loss": 0.0341, "step": 25280 }, { "grad_norm": 0.42000269889831543, "learning_rate": 8.185686455276966e-05, "loss": 0.038, "step": 25290 }, { "grad_norm": 0.28828632831573486, "learning_rate": 8.184093167637921e-05, "loss": 0.0421, "step": 25300 }, { "grad_norm": 0.3486363887786865, "learning_rate": 8.182499335924437e-05, "loss": 0.0375, "step": 25310 }, { "grad_norm": 0.44364115595817566, "learning_rate": 8.18090496040886e-05, "loss": 0.0306, "step": 25320 }, { "grad_norm": 0.3629694879055023, "learning_rate": 8.179310041363621e-05, "loss": 0.0398, "step": 25330 }, { "grad_norm": 0.46763116121292114, "learning_rate": 8.17771457906125e-05, "loss": 0.0288, "step": 25340 }, { "grad_norm": 0.38912367820739746, "learning_rate": 8.176118573774371e-05, "loss": 0.0365, "step": 25350 }, { "grad_norm": 0.4029150605201721, "learning_rate": 8.174522025775692e-05, "loss": 0.0384, "step": 25360 }, { "grad_norm": 0.40827128291130066, "learning_rate": 8.172924935338022e-05, "loss": 0.0345, "step": 25370 }, { "grad_norm": 0.6236591935157776, "learning_rate": 8.171327302734262e-05, "loss": 0.0394, "step": 25380 }, { "grad_norm": 0.3584021031856537, "learning_rate": 8.169729128237401e-05, "loss": 0.0269, "step": 25390 }, { "grad_norm": 0.8878443241119385, "learning_rate": 8.168130412120525e-05, "loss": 0.0304, "step": 25400 }, { "grad_norm": 0.3528975248336792, "learning_rate": 8.16653115465681e-05, "loss": 0.0396, "step": 25410 }, { "grad_norm": 0.4001292884349823, "learning_rate": 8.164931356119526e-05, "loss": 0.0311, "step": 25420 }, { "grad_norm": 0.569641649723053, "learning_rate": 8.163331016782032e-05, "loss": 0.044, "step": 25430 }, { "grad_norm": 0.4238159656524658, "learning_rate": 8.161730136917785e-05, "loss": 0.0261, "step": 25440 }, { "grad_norm": 0.36399713158607483, "learning_rate": 8.160128716800333e-05, "loss": 0.0397, "step": 25450 }, { "grad_norm": 0.4516690969467163, "learning_rate": 8.158526756703313e-05, "loss": 0.0326, "step": 25460 }, { "grad_norm": 0.36399227380752563, "learning_rate": 8.156924256900455e-05, "loss": 0.0406, "step": 25470 }, { "grad_norm": 0.40969809889793396, "learning_rate": 8.155321217665584e-05, "loss": 0.0302, "step": 25480 }, { "grad_norm": 0.42476925253868103, "learning_rate": 8.153717639272614e-05, "loss": 0.0294, "step": 25490 }, { "grad_norm": 0.359747976064682, "learning_rate": 8.152113521995555e-05, "loss": 0.0372, "step": 25500 }, { "grad_norm": 0.3413226902484894, "learning_rate": 8.150508866108505e-05, "loss": 0.0328, "step": 25510 }, { "grad_norm": 0.4285465180873871, "learning_rate": 8.148903671885657e-05, "loss": 0.0285, "step": 25520 }, { "grad_norm": 0.3901117444038391, "learning_rate": 8.147297939601292e-05, "loss": 0.0353, "step": 25530 }, { "grad_norm": 0.41949376463890076, "learning_rate": 8.145691669529792e-05, "loss": 0.0325, "step": 25540 }, { "grad_norm": 0.4031025171279907, "learning_rate": 8.144084861945618e-05, "loss": 0.0363, "step": 25550 }, { "grad_norm": 0.3730195462703705, "learning_rate": 8.142477517123333e-05, "loss": 0.034, "step": 25560 }, { "grad_norm": 0.3537389636039734, "learning_rate": 8.140869635337586e-05, "loss": 0.0359, "step": 25570 }, { "grad_norm": 0.3156534731388092, "learning_rate": 8.139261216863123e-05, "loss": 0.0368, "step": 25580 }, { "grad_norm": 0.30641448497772217, "learning_rate": 8.137652261974776e-05, "loss": 0.0316, "step": 25590 }, { "grad_norm": 0.3756502866744995, "learning_rate": 8.136042770947472e-05, "loss": 0.0321, "step": 25600 }, { "grad_norm": 0.3292789161205292, "learning_rate": 8.134432744056228e-05, "loss": 0.0352, "step": 25610 }, { "grad_norm": 0.3616666793823242, "learning_rate": 8.132822181576158e-05, "loss": 0.0375, "step": 25620 }, { "grad_norm": 0.36896365880966187, "learning_rate": 8.131211083782459e-05, "loss": 0.0376, "step": 25630 }, { "grad_norm": 0.465878427028656, "learning_rate": 8.129599450950424e-05, "loss": 0.0353, "step": 25640 }, { "grad_norm": 0.8351308703422546, "learning_rate": 8.127987283355438e-05, "loss": 0.044, "step": 25650 }, { "grad_norm": 0.2987654507160187, "learning_rate": 8.126374581272976e-05, "loss": 0.0249, "step": 25660 }, { "grad_norm": 0.29127100110054016, "learning_rate": 8.124761344978605e-05, "loss": 0.0483, "step": 25670 }, { "grad_norm": 0.4580705761909485, "learning_rate": 8.12314757474798e-05, "loss": 0.0293, "step": 25680 }, { "grad_norm": 0.2883104383945465, "learning_rate": 8.121533270856856e-05, "loss": 0.0418, "step": 25690 }, { "grad_norm": 0.5834065079689026, "learning_rate": 8.119918433581069e-05, "loss": 0.0315, "step": 25700 }, { "grad_norm": 0.46527886390686035, "learning_rate": 8.118303063196551e-05, "loss": 0.0338, "step": 25710 }, { "grad_norm": 0.4565019905567169, "learning_rate": 8.116687159979326e-05, "loss": 0.0296, "step": 25720 }, { "grad_norm": 0.36540380120277405, "learning_rate": 8.115070724205508e-05, "loss": 0.0384, "step": 25730 }, { "grad_norm": 0.4516771733760834, "learning_rate": 8.113453756151296e-05, "loss": 0.0344, "step": 25740 }, { "grad_norm": 0.7304093241691589, "learning_rate": 8.111836256092995e-05, "loss": 0.0402, "step": 25750 }, { "grad_norm": 0.258853554725647, "learning_rate": 8.110218224306985e-05, "loss": 0.032, "step": 25760 }, { "grad_norm": 0.29626575112342834, "learning_rate": 8.108599661069745e-05, "loss": 0.0332, "step": 25770 }, { "grad_norm": 0.3402685523033142, "learning_rate": 8.106980566657845e-05, "loss": 0.0286, "step": 25780 }, { "grad_norm": 0.2770219147205353, "learning_rate": 8.10536094134794e-05, "loss": 0.0306, "step": 25790 }, { "grad_norm": 0.2791135013103485, "learning_rate": 8.103740785416783e-05, "loss": 0.0304, "step": 25800 }, { "grad_norm": 0.3502403497695923, "learning_rate": 8.102120099141212e-05, "loss": 0.0286, "step": 25810 }, { "grad_norm": 0.3958449363708496, "learning_rate": 8.100498882798163e-05, "loss": 0.0327, "step": 25820 }, { "grad_norm": 0.5185388326644897, "learning_rate": 8.09887713666465e-05, "loss": 0.0337, "step": 25830 }, { "grad_norm": 0.34314703941345215, "learning_rate": 8.09725486101779e-05, "loss": 0.0364, "step": 25840 }, { "grad_norm": 0.34009435772895813, "learning_rate": 8.095632056134784e-05, "loss": 0.0357, "step": 25850 }, { "grad_norm": 0.37538838386535645, "learning_rate": 8.094008722292925e-05, "loss": 0.0324, "step": 25860 }, { "grad_norm": 0.24305756390094757, "learning_rate": 8.092384859769598e-05, "loss": 0.0347, "step": 25870 }, { "grad_norm": 0.42416098713874817, "learning_rate": 8.090760468842275e-05, "loss": 0.0427, "step": 25880 }, { "grad_norm": 0.24549637734889984, "learning_rate": 8.089135549788521e-05, "loss": 0.0339, "step": 25890 }, { "grad_norm": 0.34286242723464966, "learning_rate": 8.087510102885987e-05, "loss": 0.0314, "step": 25900 }, { "grad_norm": 0.39560240507125854, "learning_rate": 8.085884128412422e-05, "loss": 0.0343, "step": 25910 }, { "grad_norm": 0.4840601980686188, "learning_rate": 8.084257626645659e-05, "loss": 0.0298, "step": 25920 }, { "grad_norm": 0.3568911850452423, "learning_rate": 8.08263059786362e-05, "loss": 0.037, "step": 25930 }, { "grad_norm": 0.37369754910469055, "learning_rate": 8.081003042344325e-05, "loss": 0.0328, "step": 25940 }, { "grad_norm": 0.41992464661598206, "learning_rate": 8.079374960365872e-05, "loss": 0.0338, "step": 25950 }, { "grad_norm": 0.2946968972682953, "learning_rate": 8.077746352206463e-05, "loss": 0.0288, "step": 25960 }, { "grad_norm": 0.388785183429718, "learning_rate": 8.076117218144377e-05, "loss": 0.0317, "step": 25970 }, { "grad_norm": 0.45183300971984863, "learning_rate": 8.074487558457991e-05, "loss": 0.0319, "step": 25980 }, { "grad_norm": 0.32832813262939453, "learning_rate": 8.072857373425768e-05, "loss": 0.0344, "step": 25990 }, { "grad_norm": 0.44282400608062744, "learning_rate": 8.071226663326264e-05, "loss": 0.0305, "step": 26000 }, { "grad_norm": 0.399486243724823, "learning_rate": 8.069595428438121e-05, "loss": 0.0388, "step": 26010 }, { "grad_norm": 0.3288556635379791, "learning_rate": 8.067963669040072e-05, "loss": 0.0398, "step": 26020 }, { "grad_norm": 0.3939703106880188, "learning_rate": 8.066331385410942e-05, "loss": 0.036, "step": 26030 }, { "grad_norm": 0.38430991768836975, "learning_rate": 8.064698577829641e-05, "loss": 0.0367, "step": 26040 }, { "grad_norm": 0.3243853449821472, "learning_rate": 8.063065246575175e-05, "loss": 0.0358, "step": 26050 }, { "grad_norm": 0.3628194034099579, "learning_rate": 8.061431391926631e-05, "loss": 0.0308, "step": 26060 }, { "grad_norm": 0.2530753016471863, "learning_rate": 8.059797014163195e-05, "loss": 0.0326, "step": 26070 }, { "grad_norm": 0.3848333954811096, "learning_rate": 8.058162113564133e-05, "loss": 0.0455, "step": 26080 }, { "grad_norm": 0.41601186990737915, "learning_rate": 8.056526690408806e-05, "loss": 0.041, "step": 26090 }, { "grad_norm": 0.3489494025707245, "learning_rate": 8.054890744976666e-05, "loss": 0.0352, "step": 26100 }, { "grad_norm": 0.3112414479255676, "learning_rate": 8.053254277547248e-05, "loss": 0.0309, "step": 26110 }, { "grad_norm": 0.29988494515419006, "learning_rate": 8.051617288400182e-05, "loss": 0.0403, "step": 26120 }, { "grad_norm": 0.2541069984436035, "learning_rate": 8.049979777815182e-05, "loss": 0.0328, "step": 26130 }, { "grad_norm": 0.35696110129356384, "learning_rate": 8.048341746072054e-05, "loss": 0.0344, "step": 26140 }, { "grad_norm": 0.3952264189720154, "learning_rate": 8.046703193450696e-05, "loss": 0.0398, "step": 26150 }, { "grad_norm": 0.4619864821434021, "learning_rate": 8.04506412023109e-05, "loss": 0.0303, "step": 26160 }, { "grad_norm": 0.28446948528289795, "learning_rate": 8.043424526693306e-05, "loss": 0.0332, "step": 26170 }, { "grad_norm": 0.4170897305011749, "learning_rate": 8.04178441311751e-05, "loss": 0.034, "step": 26180 }, { "grad_norm": 0.39486339688301086, "learning_rate": 8.04014377978395e-05, "loss": 0.0268, "step": 26190 }, { "grad_norm": 0.43095746636390686, "learning_rate": 8.038502626972967e-05, "loss": 0.0429, "step": 26200 }, { "grad_norm": 0.44903337955474854, "learning_rate": 8.036860954964989e-05, "loss": 0.0421, "step": 26210 }, { "grad_norm": 0.39752885699272156, "learning_rate": 8.035218764040531e-05, "loss": 0.0357, "step": 26220 }, { "grad_norm": 0.5176788568496704, "learning_rate": 8.033576054480199e-05, "loss": 0.0395, "step": 26230 }, { "grad_norm": 0.3136470913887024, "learning_rate": 8.031932826564688e-05, "loss": 0.0316, "step": 26240 }, { "grad_norm": 0.46588751673698425, "learning_rate": 8.030289080574782e-05, "loss": 0.0353, "step": 26250 }, { "grad_norm": 0.31128790974617004, "learning_rate": 8.028644816791349e-05, "loss": 0.0361, "step": 26260 }, { "grad_norm": 0.4075264632701874, "learning_rate": 8.027000035495351e-05, "loss": 0.0264, "step": 26270 }, { "grad_norm": 0.39514270424842834, "learning_rate": 8.025354736967836e-05, "loss": 0.0436, "step": 26280 }, { "grad_norm": 0.3035340905189514, "learning_rate": 8.023708921489941e-05, "loss": 0.041, "step": 26290 }, { "grad_norm": 0.4405163824558258, "learning_rate": 8.022062589342887e-05, "loss": 0.0404, "step": 26300 }, { "grad_norm": 0.2971169054508209, "learning_rate": 8.020415740807993e-05, "loss": 0.0403, "step": 26310 }, { "grad_norm": 0.32887399196624756, "learning_rate": 8.018768376166656e-05, "loss": 0.032, "step": 26320 }, { "grad_norm": 0.3512718081474304, "learning_rate": 8.017120495700368e-05, "loss": 0.0322, "step": 26330 }, { "grad_norm": 0.5694708824157715, "learning_rate": 8.015472099690704e-05, "loss": 0.0334, "step": 26340 }, { "grad_norm": 0.21190033853054047, "learning_rate": 8.013823188419332e-05, "loss": 0.0299, "step": 26350 }, { "grad_norm": 0.26948827505111694, "learning_rate": 8.012173762168006e-05, "loss": 0.0315, "step": 26360 }, { "grad_norm": 0.2796485722064972, "learning_rate": 8.010523821218567e-05, "loss": 0.0289, "step": 26370 }, { "grad_norm": 0.31473395228385925, "learning_rate": 8.008873365852945e-05, "loss": 0.0312, "step": 26380 }, { "grad_norm": 0.3110569417476654, "learning_rate": 8.007222396353157e-05, "loss": 0.0386, "step": 26390 }, { "grad_norm": 0.42308908700942993, "learning_rate": 8.00557091300131e-05, "loss": 0.0319, "step": 26400 }, { "grad_norm": 0.5836297273635864, "learning_rate": 8.003918916079597e-05, "loss": 0.0445, "step": 26410 }, { "grad_norm": 0.28690654039382935, "learning_rate": 8.002266405870298e-05, "loss": 0.0283, "step": 26420 }, { "grad_norm": 0.25260958075523376, "learning_rate": 8.000613382655782e-05, "loss": 0.0406, "step": 26430 }, { "grad_norm": 0.3587617576122284, "learning_rate": 7.998959846718505e-05, "loss": 0.0384, "step": 26440 }, { "grad_norm": 0.4004899561405182, "learning_rate": 7.997305798341012e-05, "loss": 0.0394, "step": 26450 }, { "grad_norm": 0.386374831199646, "learning_rate": 7.995651237805937e-05, "loss": 0.0361, "step": 26460 }, { "grad_norm": 0.3768899738788605, "learning_rate": 7.993996165395996e-05, "loss": 0.0288, "step": 26470 }, { "grad_norm": 0.3687960207462311, "learning_rate": 7.992340581393996e-05, "loss": 0.0363, "step": 26480 }, { "grad_norm": 0.3438189923763275, "learning_rate": 7.990684486082831e-05, "loss": 0.0325, "step": 26490 }, { "grad_norm": 0.430316686630249, "learning_rate": 7.989027879745482e-05, "loss": 0.037, "step": 26500 }, { "grad_norm": 0.3513772785663605, "learning_rate": 7.98737076266502e-05, "loss": 0.0387, "step": 26510 }, { "grad_norm": 0.2667069137096405, "learning_rate": 7.985713135124598e-05, "loss": 0.0341, "step": 26520 }, { "grad_norm": 0.42496728897094727, "learning_rate": 7.98405499740746e-05, "loss": 0.0337, "step": 26530 }, { "grad_norm": 0.40333664417266846, "learning_rate": 7.98239634979694e-05, "loss": 0.0309, "step": 26540 }, { "grad_norm": 0.4072751998901367, "learning_rate": 7.98073719257645e-05, "loss": 0.0315, "step": 26550 }, { "grad_norm": 0.39914995431900024, "learning_rate": 7.979077526029499e-05, "loss": 0.0335, "step": 26560 }, { "grad_norm": 0.2611021101474762, "learning_rate": 7.977417350439675e-05, "loss": 0.0302, "step": 26570 }, { "grad_norm": 0.41361004114151, "learning_rate": 7.97575666609066e-05, "loss": 0.0314, "step": 26580 }, { "grad_norm": 0.37715017795562744, "learning_rate": 7.974095473266216e-05, "loss": 0.0304, "step": 26590 }, { "grad_norm": 0.3265886604785919, "learning_rate": 7.972433772250198e-05, "loss": 0.0328, "step": 26600 }, { "grad_norm": 0.41020432114601135, "learning_rate": 7.970771563326544e-05, "loss": 0.0339, "step": 26610 }, { "grad_norm": 0.4451770484447479, "learning_rate": 7.96910884677928e-05, "loss": 0.0347, "step": 26620 }, { "grad_norm": 0.37244394421577454, "learning_rate": 7.967445622892523e-05, "loss": 0.0335, "step": 26630 }, { "grad_norm": 0.2400490790605545, "learning_rate": 7.965781891950465e-05, "loss": 0.0306, "step": 26640 }, { "grad_norm": 0.30071011185646057, "learning_rate": 7.964117654237397e-05, "loss": 0.0272, "step": 26650 }, { "grad_norm": 0.44540682435035706, "learning_rate": 7.962452910037692e-05, "loss": 0.0376, "step": 26660 }, { "grad_norm": 0.4930746853351593, "learning_rate": 7.96078765963581e-05, "loss": 0.0297, "step": 26670 }, { "grad_norm": 0.40775591135025024, "learning_rate": 7.95912190331629e-05, "loss": 0.0364, "step": 26680 }, { "grad_norm": 0.3360355496406555, "learning_rate": 7.957455641363772e-05, "loss": 0.0328, "step": 26690 }, { "grad_norm": 0.4365326762199402, "learning_rate": 7.955788874062968e-05, "loss": 0.0414, "step": 26700 }, { "grad_norm": 0.3523740768432617, "learning_rate": 7.95412160169869e-05, "loss": 0.0372, "step": 26710 }, { "grad_norm": 0.29105502367019653, "learning_rate": 7.952453824555824e-05, "loss": 0.0333, "step": 26720 }, { "grad_norm": 0.27355504035949707, "learning_rate": 7.95078554291935e-05, "loss": 0.0236, "step": 26730 }, { "grad_norm": 0.45771732926368713, "learning_rate": 7.94911675707433e-05, "loss": 0.0281, "step": 26740 }, { "grad_norm": 0.3735257685184479, "learning_rate": 7.947447467305915e-05, "loss": 0.0405, "step": 26750 }, { "grad_norm": 0.37776094675064087, "learning_rate": 7.94577767389934e-05, "loss": 0.0328, "step": 26760 }, { "grad_norm": 0.5247722864151001, "learning_rate": 7.944107377139928e-05, "loss": 0.038, "step": 26770 }, { "grad_norm": 0.38210976123809814, "learning_rate": 7.942436577313088e-05, "loss": 0.0407, "step": 26780 }, { "grad_norm": 0.27256426215171814, "learning_rate": 7.940765274704312e-05, "loss": 0.0347, "step": 26790 }, { "grad_norm": 0.3225674033164978, "learning_rate": 7.939093469599181e-05, "loss": 0.031, "step": 26800 }, { "grad_norm": 0.3377862572669983, "learning_rate": 7.93742116228336e-05, "loss": 0.032, "step": 26810 }, { "grad_norm": 0.30720019340515137, "learning_rate": 7.935748353042602e-05, "loss": 0.0345, "step": 26820 }, { "grad_norm": 0.33541303873062134, "learning_rate": 7.934075042162744e-05, "loss": 0.0294, "step": 26830 }, { "grad_norm": 0.4520694613456726, "learning_rate": 7.932401229929705e-05, "loss": 0.0306, "step": 26840 }, { "grad_norm": 0.4168740510940552, "learning_rate": 7.9307269166295e-05, "loss": 0.0317, "step": 26850 }, { "grad_norm": 0.32417020201683044, "learning_rate": 7.92905210254822e-05, "loss": 0.039, "step": 26860 }, { "grad_norm": 0.37325695157051086, "learning_rate": 7.927376787972045e-05, "loss": 0.0325, "step": 26870 }, { "grad_norm": 0.35661864280700684, "learning_rate": 7.92570097318724e-05, "loss": 0.0404, "step": 26880 }, { "grad_norm": 0.254116028547287, "learning_rate": 7.924024658480158e-05, "loss": 0.0291, "step": 26890 }, { "grad_norm": 0.3189743757247925, "learning_rate": 7.922347844137233e-05, "loss": 0.0279, "step": 26900 }, { "grad_norm": 0.23744083940982819, "learning_rate": 7.92067053044499e-05, "loss": 0.0335, "step": 26910 }, { "grad_norm": 0.4312315881252289, "learning_rate": 7.918992717690031e-05, "loss": 0.0355, "step": 26920 }, { "grad_norm": 0.40966668725013733, "learning_rate": 7.917314406159053e-05, "loss": 0.0359, "step": 26930 }, { "grad_norm": 0.5134232640266418, "learning_rate": 7.915635596138832e-05, "loss": 0.0423, "step": 26940 }, { "grad_norm": 0.2654429078102112, "learning_rate": 7.913956287916228e-05, "loss": 0.0322, "step": 26950 }, { "grad_norm": 0.33293595910072327, "learning_rate": 7.912276481778193e-05, "loss": 0.0261, "step": 26960 }, { "grad_norm": 0.2763858437538147, "learning_rate": 7.910596178011759e-05, "loss": 0.0312, "step": 26970 }, { "grad_norm": 0.24590469896793365, "learning_rate": 7.908915376904043e-05, "loss": 0.0257, "step": 26980 }, { "grad_norm": 0.31854915618896484, "learning_rate": 7.907234078742247e-05, "loss": 0.0303, "step": 26990 }, { "grad_norm": 0.28164494037628174, "learning_rate": 7.90555228381366e-05, "loss": 0.0245, "step": 27000 }, { "grad_norm": 0.24045759439468384, "learning_rate": 7.903869992405656e-05, "loss": 0.0246, "step": 27010 }, { "grad_norm": 0.37576961517333984, "learning_rate": 7.902187204805691e-05, "loss": 0.0304, "step": 27020 }, { "grad_norm": 0.3849301040172577, "learning_rate": 7.900503921301308e-05, "loss": 0.0297, "step": 27030 }, { "grad_norm": 0.4472505748271942, "learning_rate": 7.898820142180133e-05, "loss": 0.04, "step": 27040 }, { "grad_norm": 0.3942393660545349, "learning_rate": 7.897135867729879e-05, "loss": 0.0292, "step": 27050 }, { "grad_norm": 0.23710709810256958, "learning_rate": 7.89545109823834e-05, "loss": 0.0232, "step": 27060 }, { "grad_norm": 0.2935239374637604, "learning_rate": 7.8937658339934e-05, "loss": 0.0273, "step": 27070 }, { "grad_norm": 0.30093061923980713, "learning_rate": 7.892080075283026e-05, "loss": 0.0262, "step": 27080 }, { "grad_norm": 0.29139527678489685, "learning_rate": 7.890393822395263e-05, "loss": 0.0272, "step": 27090 }, { "grad_norm": 0.3968313932418823, "learning_rate": 7.88870707561825e-05, "loss": 0.0345, "step": 27100 }, { "grad_norm": 0.493455171585083, "learning_rate": 7.887019835240203e-05, "loss": 0.0329, "step": 27110 }, { "grad_norm": 0.4207088053226471, "learning_rate": 7.885332101549427e-05, "loss": 0.0332, "step": 27120 }, { "grad_norm": 0.31264176964759827, "learning_rate": 7.883643874834308e-05, "loss": 0.0285, "step": 27130 }, { "grad_norm": 0.4122873842716217, "learning_rate": 7.881955155383321e-05, "loss": 0.0385, "step": 27140 }, { "grad_norm": 0.37939727306365967, "learning_rate": 7.880265943485017e-05, "loss": 0.0363, "step": 27150 }, { "grad_norm": 0.24122732877731323, "learning_rate": 7.878576239428038e-05, "loss": 0.0291, "step": 27160 }, { "grad_norm": 0.3437172770500183, "learning_rate": 7.87688604350111e-05, "loss": 0.0441, "step": 27170 }, { "grad_norm": 0.557395875453949, "learning_rate": 7.875195355993042e-05, "loss": 0.0357, "step": 27180 }, { "grad_norm": 0.4575217068195343, "learning_rate": 7.873504177192724e-05, "loss": 0.0368, "step": 27190 }, { "grad_norm": 0.3154529631137848, "learning_rate": 7.87181250738913e-05, "loss": 0.0385, "step": 27200 }, { "grad_norm": 0.33370134234428406, "learning_rate": 7.870120346871324e-05, "loss": 0.0342, "step": 27210 }, { "grad_norm": 0.46763360500335693, "learning_rate": 7.86842769592845e-05, "loss": 0.0305, "step": 27220 }, { "grad_norm": 0.44040432572364807, "learning_rate": 7.866734554849732e-05, "loss": 0.0278, "step": 27230 }, { "grad_norm": 0.29758238792419434, "learning_rate": 7.865040923924486e-05, "loss": 0.0295, "step": 27240 }, { "grad_norm": 0.4056578278541565, "learning_rate": 7.863346803442104e-05, "loss": 0.0359, "step": 27250 }, { "grad_norm": 0.4628210663795471, "learning_rate": 7.861652193692067e-05, "loss": 0.0289, "step": 27260 }, { "grad_norm": 0.32571396231651306, "learning_rate": 7.859957094963937e-05, "loss": 0.0259, "step": 27270 }, { "grad_norm": 0.2728937864303589, "learning_rate": 7.858261507547357e-05, "loss": 0.0278, "step": 27280 }, { "grad_norm": 0.2824426591396332, "learning_rate": 7.856565431732061e-05, "loss": 0.0247, "step": 27290 }, { "grad_norm": 0.35984334349632263, "learning_rate": 7.854868867807859e-05, "loss": 0.0278, "step": 27300 }, { "grad_norm": 0.4314562976360321, "learning_rate": 7.85317181606465e-05, "loss": 0.0361, "step": 27310 }, { "grad_norm": 0.33186137676239014, "learning_rate": 7.85147427679241e-05, "loss": 0.0268, "step": 27320 }, { "grad_norm": 0.29197171330451965, "learning_rate": 7.849776250281205e-05, "loss": 0.0222, "step": 27330 }, { "grad_norm": 0.4246094822883606, "learning_rate": 7.84807773682118e-05, "loss": 0.032, "step": 27340 }, { "grad_norm": 0.2632177174091339, "learning_rate": 7.846378736702565e-05, "loss": 0.0279, "step": 27350 }, { "grad_norm": 0.502140998840332, "learning_rate": 7.844679250215671e-05, "loss": 0.0414, "step": 27360 }, { "grad_norm": 0.35791897773742676, "learning_rate": 7.842979277650898e-05, "loss": 0.0306, "step": 27370 }, { "grad_norm": 0.4212070405483246, "learning_rate": 7.84127881929872e-05, "loss": 0.0303, "step": 27380 }, { "grad_norm": 0.38639405369758606, "learning_rate": 7.839577875449704e-05, "loss": 0.0441, "step": 27390 }, { "grad_norm": 0.29369229078292847, "learning_rate": 7.837876446394489e-05, "loss": 0.0343, "step": 27400 }, { "grad_norm": 0.3806850016117096, "learning_rate": 7.836174532423805e-05, "loss": 0.0432, "step": 27410 }, { "grad_norm": 0.3730678856372833, "learning_rate": 7.834472133828466e-05, "loss": 0.0295, "step": 27420 }, { "grad_norm": 0.47051966190338135, "learning_rate": 7.832769250899359e-05, "loss": 0.0373, "step": 27430 }, { "grad_norm": 0.4183712899684906, "learning_rate": 7.831065883927464e-05, "loss": 0.0375, "step": 27440 }, { "grad_norm": 0.3458259701728821, "learning_rate": 7.829362033203841e-05, "loss": 0.0329, "step": 27450 }, { "grad_norm": 0.3974364995956421, "learning_rate": 7.827657699019628e-05, "loss": 0.0374, "step": 27460 }, { "grad_norm": 0.36642757058143616, "learning_rate": 7.825952881666052e-05, "loss": 0.0466, "step": 27470 }, { "grad_norm": 0.46367689967155457, "learning_rate": 7.824247581434418e-05, "loss": 0.0334, "step": 27480 }, { "grad_norm": 0.5396660566329956, "learning_rate": 7.822541798616116e-05, "loss": 0.0348, "step": 27490 }, { "grad_norm": 0.46120160818099976, "learning_rate": 7.820835533502617e-05, "loss": 0.0393, "step": 27500 }, { "grad_norm": 0.3954753279685974, "learning_rate": 7.819128786385475e-05, "loss": 0.0332, "step": 27510 }, { "grad_norm": 0.33258071541786194, "learning_rate": 7.817421557556329e-05, "loss": 0.0379, "step": 27520 }, { "grad_norm": 0.4761667847633362, "learning_rate": 7.815713847306893e-05, "loss": 0.041, "step": 27530 }, { "grad_norm": 0.34801170229911804, "learning_rate": 7.81400565592897e-05, "loss": 0.0315, "step": 27540 }, { "grad_norm": 0.44125449657440186, "learning_rate": 7.812296983714444e-05, "loss": 0.0372, "step": 27550 }, { "grad_norm": 0.2331209033727646, "learning_rate": 7.810587830955281e-05, "loss": 0.0303, "step": 27560 }, { "grad_norm": 0.3449260890483856, "learning_rate": 7.808878197943528e-05, "loss": 0.0361, "step": 27570 }, { "grad_norm": 0.2536133825778961, "learning_rate": 7.807168084971312e-05, "loss": 0.0344, "step": 27580 }, { "grad_norm": 0.4256661832332611, "learning_rate": 7.805457492330849e-05, "loss": 0.0344, "step": 27590 }, { "grad_norm": 0.4194496273994446, "learning_rate": 7.803746420314428e-05, "loss": 0.035, "step": 27600 }, { "grad_norm": 0.4437173306941986, "learning_rate": 7.802034869214428e-05, "loss": 0.0407, "step": 27610 }, { "grad_norm": 0.4524526298046112, "learning_rate": 7.800322839323303e-05, "loss": 0.03, "step": 27620 }, { "grad_norm": 0.33297401666641235, "learning_rate": 7.798610330933593e-05, "loss": 0.0306, "step": 27630 }, { "grad_norm": 0.23493894934654236, "learning_rate": 7.796897344337922e-05, "loss": 0.0333, "step": 27640 }, { "grad_norm": 0.3742239475250244, "learning_rate": 7.795183879828989e-05, "loss": 0.0352, "step": 27650 }, { "grad_norm": 0.36756420135498047, "learning_rate": 7.793469937699579e-05, "loss": 0.0431, "step": 27660 }, { "grad_norm": 0.3652227818965912, "learning_rate": 7.791755518242558e-05, "loss": 0.0389, "step": 27670 }, { "grad_norm": 0.5275248289108276, "learning_rate": 7.790040621750876e-05, "loss": 0.0363, "step": 27680 }, { "grad_norm": 0.34058400988578796, "learning_rate": 7.788325248517558e-05, "loss": 0.0285, "step": 27690 }, { "grad_norm": 0.3169928193092346, "learning_rate": 7.786609398835715e-05, "loss": 0.0315, "step": 27700 }, { "grad_norm": 0.33843234181404114, "learning_rate": 7.784893072998541e-05, "loss": 0.0307, "step": 27710 }, { "grad_norm": 0.28188031911849976, "learning_rate": 7.783176271299306e-05, "loss": 0.028, "step": 27720 }, { "grad_norm": 0.3141636848449707, "learning_rate": 7.781458994031368e-05, "loss": 0.0254, "step": 27730 }, { "grad_norm": 0.5643221139907837, "learning_rate": 7.779741241488161e-05, "loss": 0.0335, "step": 27740 }, { "grad_norm": 0.4495774805545807, "learning_rate": 7.7780230139632e-05, "loss": 0.034, "step": 27750 }, { "grad_norm": 0.28739607334136963, "learning_rate": 7.776304311750087e-05, "loss": 0.0266, "step": 27760 }, { "grad_norm": 0.5769495964050293, "learning_rate": 7.7745851351425e-05, "loss": 0.0435, "step": 27770 }, { "grad_norm": 0.2807375192642212, "learning_rate": 7.772865484434197e-05, "loss": 0.03, "step": 27780 }, { "grad_norm": 0.4110167920589447, "learning_rate": 7.77114535991902e-05, "loss": 0.0379, "step": 27790 }, { "grad_norm": 0.4598503112792969, "learning_rate": 7.769424761890893e-05, "loss": 0.0318, "step": 27800 }, { "grad_norm": 0.3972029387950897, "learning_rate": 7.767703690643817e-05, "loss": 0.0322, "step": 27810 }, { "grad_norm": 0.2824877202510834, "learning_rate": 7.76598214647188e-05, "loss": 0.0324, "step": 27820 }, { "grad_norm": 0.3539346754550934, "learning_rate": 7.764260129669241e-05, "loss": 0.0333, "step": 27830 }, { "grad_norm": 0.34736350178718567, "learning_rate": 7.76253764053015e-05, "loss": 0.0395, "step": 27840 }, { "grad_norm": 0.34176546335220337, "learning_rate": 7.760814679348932e-05, "loss": 0.0248, "step": 27850 }, { "grad_norm": 0.32658401131629944, "learning_rate": 7.759091246419992e-05, "loss": 0.024, "step": 27860 }, { "grad_norm": 0.5988019108772278, "learning_rate": 7.75736734203782e-05, "loss": 0.0321, "step": 27870 }, { "grad_norm": 0.45710018277168274, "learning_rate": 7.755642966496985e-05, "loss": 0.0392, "step": 27880 }, { "grad_norm": 0.3475942611694336, "learning_rate": 7.753918120092132e-05, "loss": 0.0334, "step": 27890 }, { "grad_norm": 0.27531346678733826, "learning_rate": 7.752192803117993e-05, "loss": 0.0268, "step": 27900 }, { "grad_norm": 0.49730998277664185, "learning_rate": 7.750467015869377e-05, "loss": 0.0351, "step": 27910 }, { "grad_norm": 0.3614085614681244, "learning_rate": 7.748740758641174e-05, "loss": 0.0377, "step": 27920 }, { "grad_norm": 0.31942617893218994, "learning_rate": 7.74701403172835e-05, "loss": 0.036, "step": 27930 }, { "grad_norm": 0.34240198135375977, "learning_rate": 7.745286835425962e-05, "loss": 0.0354, "step": 27940 }, { "grad_norm": 0.4201430678367615, "learning_rate": 7.743559170029138e-05, "loss": 0.0359, "step": 27950 }, { "grad_norm": 0.2764117419719696, "learning_rate": 7.741831035833087e-05, "loss": 0.0414, "step": 27960 }, { "grad_norm": 0.34902870655059814, "learning_rate": 7.740102433133102e-05, "loss": 0.032, "step": 27970 }, { "grad_norm": 0.23201066255569458, "learning_rate": 7.738373362224553e-05, "loss": 0.0244, "step": 27980 }, { "grad_norm": 0.33830806612968445, "learning_rate": 7.73664382340289e-05, "loss": 0.0277, "step": 27990 }, { "grad_norm": 0.38919660449028015, "learning_rate": 7.734913816963647e-05, "loss": 0.0347, "step": 28000 }, { "grad_norm": 0.29927897453308105, "learning_rate": 7.73318334320243e-05, "loss": 0.028, "step": 28010 }, { "grad_norm": 0.3801274299621582, "learning_rate": 7.731452402414934e-05, "loss": 0.0378, "step": 28020 }, { "grad_norm": 0.3915766477584839, "learning_rate": 7.729720994896928e-05, "loss": 0.0408, "step": 28030 }, { "grad_norm": 0.44587817788124084, "learning_rate": 7.727989120944262e-05, "loss": 0.0348, "step": 28040 }, { "grad_norm": 0.40307268500328064, "learning_rate": 7.726256780852865e-05, "loss": 0.0286, "step": 28050 }, { "grad_norm": 0.3298165202140808, "learning_rate": 7.724523974918749e-05, "loss": 0.0298, "step": 28060 }, { "grad_norm": 0.25804030895233154, "learning_rate": 7.722790703438002e-05, "loss": 0.0264, "step": 28070 }, { "grad_norm": 0.19976530969142914, "learning_rate": 7.72105696670679e-05, "loss": 0.0239, "step": 28080 }, { "grad_norm": 0.49688801169395447, "learning_rate": 7.719322765021364e-05, "loss": 0.0324, "step": 28090 }, { "grad_norm": 0.31548425555229187, "learning_rate": 7.717588098678051e-05, "loss": 0.0311, "step": 28100 }, { "grad_norm": 0.34738004207611084, "learning_rate": 7.715852967973258e-05, "loss": 0.0293, "step": 28110 }, { "grad_norm": 0.3209443688392639, "learning_rate": 7.714117373203474e-05, "loss": 0.0324, "step": 28120 }, { "grad_norm": 0.3172980844974518, "learning_rate": 7.712381314665259e-05, "loss": 0.0284, "step": 28130 }, { "grad_norm": 0.40320923924446106, "learning_rate": 7.710644792655261e-05, "loss": 0.0291, "step": 28140 }, { "grad_norm": 0.49548324942588806, "learning_rate": 7.708907807470207e-05, "loss": 0.0343, "step": 28150 }, { "grad_norm": 0.3284149169921875, "learning_rate": 7.707170359406896e-05, "loss": 0.0309, "step": 28160 }, { "grad_norm": 0.3131861686706543, "learning_rate": 7.705432448762213e-05, "loss": 0.0292, "step": 28170 }, { "grad_norm": 0.5814459323883057, "learning_rate": 7.703694075833117e-05, "loss": 0.0395, "step": 28180 }, { "grad_norm": 0.5272874236106873, "learning_rate": 7.70195524091665e-05, "loss": 0.053, "step": 28190 }, { "grad_norm": 0.454086571931839, "learning_rate": 7.70021594430993e-05, "loss": 0.0383, "step": 28200 }, { "grad_norm": 0.3442489802837372, "learning_rate": 7.698476186310157e-05, "loss": 0.0293, "step": 28210 }, { "grad_norm": 0.3990366756916046, "learning_rate": 7.696735967214608e-05, "loss": 0.0287, "step": 28220 }, { "grad_norm": 0.3559223413467407, "learning_rate": 7.694995287320636e-05, "loss": 0.0289, "step": 28230 }, { "grad_norm": 0.27979445457458496, "learning_rate": 7.693254146925679e-05, "loss": 0.0285, "step": 28240 }, { "grad_norm": 0.4381241202354431, "learning_rate": 7.691512546327251e-05, "loss": 0.0375, "step": 28250 }, { "grad_norm": 0.36967575550079346, "learning_rate": 7.689770485822939e-05, "loss": 0.0314, "step": 28260 }, { "grad_norm": 0.33476513624191284, "learning_rate": 7.688027965710416e-05, "loss": 0.0234, "step": 28270 }, { "grad_norm": 0.3999041020870209, "learning_rate": 7.686284986287433e-05, "loss": 0.03, "step": 28280 }, { "grad_norm": 0.37086620926856995, "learning_rate": 7.684541547851817e-05, "loss": 0.0272, "step": 28290 }, { "grad_norm": 0.41841524839401245, "learning_rate": 7.68279765070147e-05, "loss": 0.0315, "step": 28300 }, { "grad_norm": 0.39815518260002136, "learning_rate": 7.68105329513438e-05, "loss": 0.029, "step": 28310 }, { "grad_norm": 0.2606412172317505, "learning_rate": 7.67930848144861e-05, "loss": 0.0286, "step": 28320 }, { "grad_norm": 0.4471575617790222, "learning_rate": 7.6775632099423e-05, "loss": 0.0361, "step": 28330 }, { "grad_norm": 0.27372485399246216, "learning_rate": 7.675817480913667e-05, "loss": 0.0319, "step": 28340 }, { "grad_norm": 0.39468538761138916, "learning_rate": 7.674071294661011e-05, "loss": 0.0355, "step": 28350 }, { "grad_norm": 0.36037442088127136, "learning_rate": 7.672324651482707e-05, "loss": 0.0267, "step": 28360 }, { "grad_norm": 0.5218867659568787, "learning_rate": 7.670577551677209e-05, "loss": 0.0322, "step": 28370 }, { "grad_norm": 0.3781348764896393, "learning_rate": 7.668829995543047e-05, "loss": 0.0333, "step": 28380 }, { "grad_norm": 0.3263038098812103, "learning_rate": 7.667081983378832e-05, "loss": 0.0356, "step": 28390 }, { "grad_norm": 0.5437476634979248, "learning_rate": 7.66533351548325e-05, "loss": 0.0303, "step": 28400 }, { "grad_norm": 0.47407111525535583, "learning_rate": 7.663584592155069e-05, "loss": 0.0325, "step": 28410 }, { "grad_norm": 0.40542352199554443, "learning_rate": 7.661835213693129e-05, "loss": 0.0327, "step": 28420 }, { "grad_norm": 0.4602213501930237, "learning_rate": 7.660085380396353e-05, "loss": 0.0359, "step": 28430 }, { "grad_norm": 0.474869042634964, "learning_rate": 7.658335092563738e-05, "loss": 0.0328, "step": 28440 }, { "grad_norm": 0.35012128949165344, "learning_rate": 7.656584350494362e-05, "loss": 0.0249, "step": 28450 }, { "grad_norm": 0.4013153612613678, "learning_rate": 7.654833154487378e-05, "loss": 0.0317, "step": 28460 }, { "grad_norm": 0.2731168568134308, "learning_rate": 7.653081504842017e-05, "loss": 0.0322, "step": 28470 }, { "grad_norm": 0.40063583850860596, "learning_rate": 7.65132940185759e-05, "loss": 0.0405, "step": 28480 }, { "grad_norm": 0.38130903244018555, "learning_rate": 7.649576845833481e-05, "loss": 0.0289, "step": 28490 }, { "grad_norm": 0.3029012084007263, "learning_rate": 7.647823837069156e-05, "loss": 0.0327, "step": 28500 }, { "grad_norm": 0.29329073429107666, "learning_rate": 7.646070375864156e-05, "loss": 0.0269, "step": 28510 }, { "grad_norm": 0.4617317020893097, "learning_rate": 7.644316462518099e-05, "loss": 0.0335, "step": 28520 }, { "grad_norm": 0.3845706880092621, "learning_rate": 7.642562097330679e-05, "loss": 0.0347, "step": 28530 }, { "grad_norm": 0.2541613280773163, "learning_rate": 7.640807280601671e-05, "loss": 0.03, "step": 28540 }, { "grad_norm": 0.3431868255138397, "learning_rate": 7.639052012630927e-05, "loss": 0.0357, "step": 28550 }, { "grad_norm": 0.3551851212978363, "learning_rate": 7.63729629371837e-05, "loss": 0.0404, "step": 28560 }, { "grad_norm": 0.4791673421859741, "learning_rate": 7.635540124164009e-05, "loss": 0.0413, "step": 28570 }, { "grad_norm": 0.2712848484516144, "learning_rate": 7.633783504267922e-05, "loss": 0.0324, "step": 28580 }, { "grad_norm": 0.2883760333061218, "learning_rate": 7.632026434330269e-05, "loss": 0.0284, "step": 28590 }, { "grad_norm": 0.49154040217399597, "learning_rate": 7.630268914651282e-05, "loss": 0.0287, "step": 28600 }, { "grad_norm": 0.40651991963386536, "learning_rate": 7.628510945531278e-05, "loss": 0.0414, "step": 28610 }, { "grad_norm": 0.2734713554382324, "learning_rate": 7.626752527270641e-05, "loss": 0.0366, "step": 28620 }, { "grad_norm": 0.2692888677120209, "learning_rate": 7.62499366016984e-05, "loss": 0.0362, "step": 28630 }, { "grad_norm": 0.2655770480632782, "learning_rate": 7.623234344529416e-05, "loss": 0.0339, "step": 28640 }, { "grad_norm": 0.39177852869033813, "learning_rate": 7.62147458064999e-05, "loss": 0.0332, "step": 28650 }, { "grad_norm": 0.4585602283477783, "learning_rate": 7.619714368832254e-05, "loss": 0.0371, "step": 28660 }, { "grad_norm": 0.36992311477661133, "learning_rate": 7.61795370937698e-05, "loss": 0.0343, "step": 28670 }, { "grad_norm": 0.39546018838882446, "learning_rate": 7.61619260258502e-05, "loss": 0.035, "step": 28680 }, { "grad_norm": 0.3348652124404907, "learning_rate": 7.614431048757298e-05, "loss": 0.0404, "step": 28690 }, { "grad_norm": 0.3430466651916504, "learning_rate": 7.612669048194814e-05, "loss": 0.0354, "step": 28700 }, { "grad_norm": 0.35329827666282654, "learning_rate": 7.610906601198646e-05, "loss": 0.028, "step": 28710 }, { "grad_norm": 0.6997968554496765, "learning_rate": 7.60914370806995e-05, "loss": 0.0367, "step": 28720 }, { "grad_norm": 0.3304347097873688, "learning_rate": 7.607380369109953e-05, "loss": 0.0503, "step": 28730 }, { "grad_norm": 0.3514820337295532, "learning_rate": 7.605616584619961e-05, "loss": 0.0274, "step": 28740 }, { "grad_norm": 0.33135199546813965, "learning_rate": 7.603852354901362e-05, "loss": 0.0333, "step": 28750 }, { "grad_norm": 0.2984280288219452, "learning_rate": 7.602087680255609e-05, "loss": 0.0462, "step": 28760 }, { "grad_norm": 0.4260050058364868, "learning_rate": 7.600322560984238e-05, "loss": 0.0305, "step": 28770 }, { "grad_norm": 0.3306851387023926, "learning_rate": 7.598556997388863e-05, "loss": 0.03, "step": 28780 }, { "grad_norm": 0.32383838295936584, "learning_rate": 7.596790989771166e-05, "loss": 0.03, "step": 28790 }, { "grad_norm": 0.5516872406005859, "learning_rate": 7.595024538432914e-05, "loss": 0.033, "step": 28800 }, { "grad_norm": 0.45541390776634216, "learning_rate": 7.59325764367594e-05, "loss": 0.0296, "step": 28810 }, { "grad_norm": 0.2557080090045929, "learning_rate": 7.59149030580216e-05, "loss": 0.0308, "step": 28820 }, { "grad_norm": 0.2317110151052475, "learning_rate": 7.589722525113562e-05, "loss": 0.0271, "step": 28830 }, { "grad_norm": 0.40733274817466736, "learning_rate": 7.587954301912216e-05, "loss": 0.0254, "step": 28840 }, { "grad_norm": 0.42028117179870605, "learning_rate": 7.586185636500263e-05, "loss": 0.0317, "step": 28850 }, { "grad_norm": 0.3419438898563385, "learning_rate": 7.584416529179914e-05, "loss": 0.0292, "step": 28860 }, { "grad_norm": 0.3026672601699829, "learning_rate": 7.582646980253465e-05, "loss": 0.0288, "step": 28870 }, { "grad_norm": 0.33506980538368225, "learning_rate": 7.580876990023282e-05, "loss": 0.0412, "step": 28880 }, { "grad_norm": 0.38744840025901794, "learning_rate": 7.579106558791809e-05, "loss": 0.0301, "step": 28890 }, { "grad_norm": 0.45741748809814453, "learning_rate": 7.577335686861565e-05, "loss": 0.0319, "step": 28900 }, { "grad_norm": 0.24321837723255157, "learning_rate": 7.575564374535141e-05, "loss": 0.0353, "step": 28910 }, { "grad_norm": 0.28738072514533997, "learning_rate": 7.573792622115207e-05, "loss": 0.0273, "step": 28920 }, { "grad_norm": 0.2890666425228119, "learning_rate": 7.572020429904507e-05, "loss": 0.0232, "step": 28930 }, { "grad_norm": 0.3621479868888855, "learning_rate": 7.570247798205861e-05, "loss": 0.0336, "step": 28940 }, { "grad_norm": 0.2330654114484787, "learning_rate": 7.568474727322164e-05, "loss": 0.033, "step": 28950 }, { "grad_norm": 0.3534119725227356, "learning_rate": 7.566701217556384e-05, "loss": 0.0262, "step": 28960 }, { "grad_norm": 0.3596115708351135, "learning_rate": 7.564927269211564e-05, "loss": 0.032, "step": 28970 }, { "grad_norm": 0.41135263442993164, "learning_rate": 7.563152882590824e-05, "loss": 0.0264, "step": 28980 }, { "grad_norm": 0.39235082268714905, "learning_rate": 7.56137805799736e-05, "loss": 0.0328, "step": 28990 }, { "grad_norm": 0.33204948902130127, "learning_rate": 7.559602795734439e-05, "loss": 0.0309, "step": 29000 }, { "grad_norm": 0.22869010269641876, "learning_rate": 7.557827096105408e-05, "loss": 0.0279, "step": 29010 }, { "grad_norm": 0.3086181879043579, "learning_rate": 7.55605095941368e-05, "loss": 0.0314, "step": 29020 }, { "grad_norm": 0.3138245940208435, "learning_rate": 7.554274385962752e-05, "loss": 0.034, "step": 29030 }, { "grad_norm": 0.3442644774913788, "learning_rate": 7.552497376056191e-05, "loss": 0.036, "step": 29040 }, { "grad_norm": 0.742292046546936, "learning_rate": 7.550719929997639e-05, "loss": 0.0403, "step": 29050 }, { "grad_norm": 0.4337995946407318, "learning_rate": 7.548942048090813e-05, "loss": 0.0278, "step": 29060 }, { "grad_norm": 0.32509592175483704, "learning_rate": 7.547163730639506e-05, "loss": 0.033, "step": 29070 }, { "grad_norm": 0.37313249707221985, "learning_rate": 7.545384977947583e-05, "loss": 0.0312, "step": 29080 }, { "grad_norm": 0.35089072585105896, "learning_rate": 7.543605790318981e-05, "loss": 0.0355, "step": 29090 }, { "grad_norm": 0.46059712767601013, "learning_rate": 7.54182616805772e-05, "loss": 0.0389, "step": 29100 }, { "grad_norm": 0.3261191248893738, "learning_rate": 7.540046111467885e-05, "loss": 0.0258, "step": 29110 }, { "grad_norm": 0.3798414170742035, "learning_rate": 7.53826562085364e-05, "loss": 0.0322, "step": 29120 }, { "grad_norm": 0.40169841051101685, "learning_rate": 7.536484696519221e-05, "loss": 0.0261, "step": 29130 }, { "grad_norm": 0.38032498955726624, "learning_rate": 7.534703338768942e-05, "loss": 0.0274, "step": 29140 }, { "grad_norm": 0.4439719021320343, "learning_rate": 7.532921547907185e-05, "loss": 0.034, "step": 29150 }, { "grad_norm": 0.5838073492050171, "learning_rate": 7.531139324238412e-05, "loss": 0.0342, "step": 29160 }, { "grad_norm": 0.5843903422355652, "learning_rate": 7.529356668067157e-05, "loss": 0.0415, "step": 29170 }, { "grad_norm": 0.33858147263526917, "learning_rate": 7.527573579698023e-05, "loss": 0.0358, "step": 29180 }, { "grad_norm": 0.5720019340515137, "learning_rate": 7.525790059435693e-05, "loss": 0.0349, "step": 29190 }, { "grad_norm": 0.3321327269077301, "learning_rate": 7.524006107584926e-05, "loss": 0.0306, "step": 29200 }, { "grad_norm": 0.32231318950653076, "learning_rate": 7.522221724450544e-05, "loss": 0.0299, "step": 29210 }, { "grad_norm": 0.2950664460659027, "learning_rate": 7.520436910337451e-05, "loss": 0.0396, "step": 29220 }, { "grad_norm": 0.470050573348999, "learning_rate": 7.518651665550627e-05, "loss": 0.031, "step": 29230 }, { "grad_norm": 0.4737553000450134, "learning_rate": 7.516865990395117e-05, "loss": 0.0399, "step": 29240 }, { "grad_norm": 0.33217233419418335, "learning_rate": 7.515079885176047e-05, "loss": 0.0258, "step": 29250 }, { "grad_norm": 0.29107409715652466, "learning_rate": 7.513293350198612e-05, "loss": 0.0337, "step": 29260 }, { "grad_norm": 0.42872777581214905, "learning_rate": 7.511506385768081e-05, "loss": 0.0327, "step": 29270 }, { "grad_norm": 0.36792489886283875, "learning_rate": 7.509718992189801e-05, "loss": 0.0325, "step": 29280 }, { "grad_norm": 0.39868393540382385, "learning_rate": 7.507931169769182e-05, "loss": 0.0355, "step": 29290 }, { "grad_norm": 0.5356352925300598, "learning_rate": 7.506142918811722e-05, "loss": 0.0338, "step": 29300 }, { "grad_norm": 0.2298928201198578, "learning_rate": 7.504354239622978e-05, "loss": 0.0342, "step": 29310 }, { "grad_norm": 0.2975746691226959, "learning_rate": 7.50256513250859e-05, "loss": 0.0309, "step": 29320 }, { "grad_norm": 0.3826993703842163, "learning_rate": 7.500775597774265e-05, "loss": 0.0313, "step": 29330 }, { "grad_norm": 0.4623074531555176, "learning_rate": 7.498985635725788e-05, "loss": 0.0259, "step": 29340 }, { "grad_norm": 0.4052867591381073, "learning_rate": 7.497195246669012e-05, "loss": 0.0315, "step": 29350 }, { "grad_norm": 0.3169534206390381, "learning_rate": 7.495404430909868e-05, "loss": 0.0289, "step": 29360 }, { "grad_norm": 0.38681745529174805, "learning_rate": 7.493613188754356e-05, "loss": 0.0315, "step": 29370 }, { "grad_norm": 0.30807381868362427, "learning_rate": 7.49182152050855e-05, "loss": 0.0293, "step": 29380 }, { "grad_norm": 0.3115352392196655, "learning_rate": 7.490029426478598e-05, "loss": 0.0287, "step": 29390 }, { "grad_norm": 0.2503395676612854, "learning_rate": 7.488236906970719e-05, "loss": 0.0331, "step": 29400 }, { "grad_norm": 0.3739856779575348, "learning_rate": 7.486443962291207e-05, "loss": 0.0326, "step": 29410 }, { "grad_norm": 0.30239757895469666, "learning_rate": 7.484650592746424e-05, "loss": 0.0289, "step": 29420 }, { "grad_norm": 0.35278886556625366, "learning_rate": 7.482856798642811e-05, "loss": 0.0272, "step": 29430 }, { "grad_norm": 0.3028189241886139, "learning_rate": 7.481062580286878e-05, "loss": 0.0331, "step": 29440 }, { "grad_norm": 0.21555982530117035, "learning_rate": 7.479267937985208e-05, "loss": 0.0262, "step": 29450 }, { "grad_norm": 0.7255986928939819, "learning_rate": 7.477472872044456e-05, "loss": 0.0288, "step": 29460 }, { "grad_norm": 0.3758552670478821, "learning_rate": 7.475677382771347e-05, "loss": 0.0353, "step": 29470 }, { "grad_norm": 0.22737017273902893, "learning_rate": 7.473881470472683e-05, "loss": 0.0296, "step": 29480 }, { "grad_norm": 0.3711288273334503, "learning_rate": 7.472085135455337e-05, "loss": 0.0339, "step": 29490 }, { "grad_norm": 0.32521018385887146, "learning_rate": 7.470288378026256e-05, "loss": 0.0367, "step": 29500 }, { "grad_norm": 0.3546096384525299, "learning_rate": 7.468491198492451e-05, "loss": 0.0266, "step": 29510 }, { "grad_norm": 0.30849194526672363, "learning_rate": 7.466693597161013e-05, "loss": 0.0327, "step": 29520 }, { "grad_norm": 0.2647179961204529, "learning_rate": 7.464895574339104e-05, "loss": 0.0289, "step": 29530 }, { "grad_norm": 0.20104722678661346, "learning_rate": 7.463097130333958e-05, "loss": 0.0369, "step": 29540 }, { "grad_norm": 0.4540316164493561, "learning_rate": 7.461298265452876e-05, "loss": 0.0306, "step": 29550 }, { "grad_norm": 0.2425558865070343, "learning_rate": 7.459498980003239e-05, "loss": 0.0245, "step": 29560 }, { "grad_norm": 0.3304288685321808, "learning_rate": 7.457699274292493e-05, "loss": 0.0351, "step": 29570 }, { "grad_norm": 0.27873608469963074, "learning_rate": 7.455899148628159e-05, "loss": 0.0336, "step": 29580 }, { "grad_norm": 0.38233834505081177, "learning_rate": 7.45409860331783e-05, "loss": 0.0307, "step": 29590 }, { "grad_norm": 0.3272666335105896, "learning_rate": 7.452297638669169e-05, "loss": 0.0282, "step": 29600 }, { "grad_norm": 0.8052488565444946, "learning_rate": 7.450496254989911e-05, "loss": 0.0247, "step": 29610 }, { "grad_norm": 0.24351724982261658, "learning_rate": 7.448694452587866e-05, "loss": 0.0286, "step": 29620 }, { "grad_norm": 0.6625458002090454, "learning_rate": 7.44689223177091e-05, "loss": 0.0304, "step": 29630 }, { "grad_norm": 0.34906578063964844, "learning_rate": 7.445089592846994e-05, "loss": 0.033, "step": 29640 }, { "grad_norm": 0.35847166180610657, "learning_rate": 7.443286536124141e-05, "loss": 0.0407, "step": 29650 }, { "grad_norm": 0.36523208022117615, "learning_rate": 7.441483061910443e-05, "loss": 0.0432, "step": 29660 }, { "grad_norm": 0.44403016567230225, "learning_rate": 7.439679170514064e-05, "loss": 0.0364, "step": 29670 }, { "grad_norm": 0.31179237365722656, "learning_rate": 7.43787486224324e-05, "loss": 0.032, "step": 29680 }, { "grad_norm": 0.4376644790172577, "learning_rate": 7.436070137406276e-05, "loss": 0.029, "step": 29690 }, { "grad_norm": 0.3140794634819031, "learning_rate": 7.434264996311556e-05, "loss": 0.0313, "step": 29700 }, { "grad_norm": 0.7304261922836304, "learning_rate": 7.432459439267525e-05, "loss": 0.0402, "step": 29710 }, { "grad_norm": 0.3030281960964203, "learning_rate": 7.430653466582701e-05, "loss": 0.032, "step": 29720 }, { "grad_norm": 0.3130320906639099, "learning_rate": 7.42884707856568e-05, "loss": 0.0403, "step": 29730 }, { "grad_norm": 0.4099538326263428, "learning_rate": 7.427040275525122e-05, "loss": 0.03, "step": 29740 }, { "grad_norm": 0.34297552704811096, "learning_rate": 7.42523305776976e-05, "loss": 0.0286, "step": 29750 }, { "grad_norm": 0.38515451550483704, "learning_rate": 7.4234254256084e-05, "loss": 0.0247, "step": 29760 }, { "grad_norm": 0.2333897352218628, "learning_rate": 7.421617379349915e-05, "loss": 0.0319, "step": 29770 }, { "grad_norm": 0.2603554129600525, "learning_rate": 7.41980891930325e-05, "loss": 0.0252, "step": 29780 }, { "grad_norm": 0.2175874263048172, "learning_rate": 7.418000045777425e-05, "loss": 0.0258, "step": 29790 }, { "grad_norm": 0.23554112017154694, "learning_rate": 7.416190759081523e-05, "loss": 0.0278, "step": 29800 }, { "grad_norm": 0.2789245843887329, "learning_rate": 7.414381059524704e-05, "loss": 0.0238, "step": 29810 }, { "grad_norm": 0.2650068402290344, "learning_rate": 7.412570947416195e-05, "loss": 0.0229, "step": 29820 }, { "grad_norm": 0.7061565518379211, "learning_rate": 7.410760423065295e-05, "loss": 0.0453, "step": 29830 }, { "grad_norm": 0.45703768730163574, "learning_rate": 7.408949486781372e-05, "loss": 0.038, "step": 29840 }, { "grad_norm": 0.38185742497444153, "learning_rate": 7.407138138873868e-05, "loss": 0.0372, "step": 29850 }, { "grad_norm": 0.40411776304244995, "learning_rate": 7.405326379652292e-05, "loss": 0.0334, "step": 29860 }, { "grad_norm": 0.30229049921035767, "learning_rate": 7.403514209426222e-05, "loss": 0.026, "step": 29870 }, { "grad_norm": 0.39976921677589417, "learning_rate": 7.40170162850531e-05, "loss": 0.0339, "step": 29880 }, { "grad_norm": 0.596678614616394, "learning_rate": 7.399888637199278e-05, "loss": 0.0296, "step": 29890 }, { "grad_norm": 0.402513325214386, "learning_rate": 7.398075235817914e-05, "loss": 0.0421, "step": 29900 }, { "grad_norm": 0.5852786302566528, "learning_rate": 7.39626142467108e-05, "loss": 0.0359, "step": 29910 }, { "grad_norm": 0.29964062571525574, "learning_rate": 7.394447204068706e-05, "loss": 0.0257, "step": 29920 }, { "grad_norm": 0.367186963558197, "learning_rate": 7.392632574320793e-05, "loss": 0.0429, "step": 29930 }, { "grad_norm": 0.2982637286186218, "learning_rate": 7.390817535737411e-05, "loss": 0.0347, "step": 29940 }, { "grad_norm": 0.2760884761810303, "learning_rate": 7.389002088628703e-05, "loss": 0.0286, "step": 29950 }, { "grad_norm": 0.7512465715408325, "learning_rate": 7.387186233304877e-05, "loss": 0.0334, "step": 29960 }, { "grad_norm": 0.4607868194580078, "learning_rate": 7.385369970076212e-05, "loss": 0.0419, "step": 29970 }, { "grad_norm": 0.5264055132865906, "learning_rate": 7.38355329925306e-05, "loss": 0.0335, "step": 29980 }, { "grad_norm": 0.24072737991809845, "learning_rate": 7.381736221145838e-05, "loss": 0.0339, "step": 29990 }, { "grad_norm": 0.29339420795440674, "learning_rate": 7.37991873606504e-05, "loss": 0.0284, "step": 30000 }, { "grad_norm": 0.3374492824077606, "learning_rate": 7.378100844321218e-05, "loss": 0.0283, "step": 30010 }, { "grad_norm": 0.23144303262233734, "learning_rate": 7.376282546225004e-05, "loss": 0.025, "step": 30020 }, { "grad_norm": 0.29011622071266174, "learning_rate": 7.374463842087094e-05, "loss": 0.0226, "step": 30030 }, { "grad_norm": 0.3684373199939728, "learning_rate": 7.372644732218254e-05, "loss": 0.0251, "step": 30040 }, { "grad_norm": 0.49284476041793823, "learning_rate": 7.370825216929322e-05, "loss": 0.0445, "step": 30050 }, { "grad_norm": 0.3388049900531769, "learning_rate": 7.369005296531205e-05, "loss": 0.0318, "step": 30060 }, { "grad_norm": 0.36454638838768005, "learning_rate": 7.367184971334873e-05, "loss": 0.039, "step": 30070 }, { "grad_norm": 0.2545700669288635, "learning_rate": 7.365364241651371e-05, "loss": 0.0276, "step": 30080 }, { "grad_norm": 0.3957785665988922, "learning_rate": 7.363543107791815e-05, "loss": 0.0276, "step": 30090 }, { "grad_norm": 0.30973371863365173, "learning_rate": 7.361721570067384e-05, "loss": 0.0247, "step": 30100 }, { "grad_norm": 0.31925687193870544, "learning_rate": 7.359899628789331e-05, "loss": 0.0283, "step": 30110 }, { "grad_norm": 0.3073224723339081, "learning_rate": 7.358077284268974e-05, "loss": 0.0383, "step": 30120 }, { "grad_norm": 0.439666211605072, "learning_rate": 7.356254536817702e-05, "loss": 0.0261, "step": 30130 }, { "grad_norm": 0.39577192068099976, "learning_rate": 7.354431386746973e-05, "loss": 0.0328, "step": 30140 }, { "grad_norm": 0.474326491355896, "learning_rate": 7.352607834368316e-05, "loss": 0.0307, "step": 30150 }, { "grad_norm": 0.27294009923934937, "learning_rate": 7.350783879993324e-05, "loss": 0.0242, "step": 30160 }, { "grad_norm": 0.37818464636802673, "learning_rate": 7.348959523933658e-05, "loss": 0.0326, "step": 30170 }, { "grad_norm": 0.22386521100997925, "learning_rate": 7.347134766501057e-05, "loss": 0.0219, "step": 30180 }, { "grad_norm": 0.374813437461853, "learning_rate": 7.345309608007315e-05, "loss": 0.0262, "step": 30190 }, { "grad_norm": 0.3678978383541107, "learning_rate": 7.343484048764308e-05, "loss": 0.0325, "step": 30200 }, { "grad_norm": 0.25208404660224915, "learning_rate": 7.341658089083972e-05, "loss": 0.0303, "step": 30210 }, { "grad_norm": 0.30389901995658875, "learning_rate": 7.339831729278313e-05, "loss": 0.0243, "step": 30220 }, { "grad_norm": 0.2880171537399292, "learning_rate": 7.338004969659404e-05, "loss": 0.027, "step": 30230 }, { "grad_norm": 0.327980101108551, "learning_rate": 7.336177810539391e-05, "loss": 0.0347, "step": 30240 }, { "grad_norm": 0.2596932053565979, "learning_rate": 7.334350252230485e-05, "loss": 0.0212, "step": 30250 }, { "grad_norm": 0.28674694895744324, "learning_rate": 7.332522295044965e-05, "loss": 0.0279, "step": 30260 }, { "grad_norm": 0.4987490475177765, "learning_rate": 7.33069393929518e-05, "loss": 0.0327, "step": 30270 }, { "grad_norm": 0.3833676278591156, "learning_rate": 7.328865185293545e-05, "loss": 0.0412, "step": 30280 }, { "grad_norm": 0.3306726813316345, "learning_rate": 7.327036033352546e-05, "loss": 0.0309, "step": 30290 }, { "grad_norm": 0.37674927711486816, "learning_rate": 7.325206483784733e-05, "loss": 0.0285, "step": 30300 }, { "grad_norm": 0.2899705767631531, "learning_rate": 7.323376536902724e-05, "loss": 0.0399, "step": 30310 }, { "grad_norm": 0.35432276129722595, "learning_rate": 7.321546193019213e-05, "loss": 0.031, "step": 30320 }, { "grad_norm": 0.4040840268135071, "learning_rate": 7.31971545244695e-05, "loss": 0.0337, "step": 30330 }, { "grad_norm": 0.28362369537353516, "learning_rate": 7.31788431549876e-05, "loss": 0.0267, "step": 30340 }, { "grad_norm": 0.28843048214912415, "learning_rate": 7.316052782487534e-05, "loss": 0.0242, "step": 30350 }, { "grad_norm": 0.35593104362487793, "learning_rate": 7.314220853726234e-05, "loss": 0.0342, "step": 30360 }, { "grad_norm": 0.3182820975780487, "learning_rate": 7.312388529527884e-05, "loss": 0.0366, "step": 30370 }, { "grad_norm": 0.407372385263443, "learning_rate": 7.310555810205577e-05, "loss": 0.0301, "step": 30380 }, { "grad_norm": 0.3916490375995636, "learning_rate": 7.308722696072476e-05, "loss": 0.0361, "step": 30390 }, { "grad_norm": 0.3372926712036133, "learning_rate": 7.306889187441811e-05, "loss": 0.0338, "step": 30400 }, { "grad_norm": 0.35613560676574707, "learning_rate": 7.305055284626876e-05, "loss": 0.03, "step": 30410 }, { "grad_norm": 0.33911076188087463, "learning_rate": 7.303220987941037e-05, "loss": 0.0304, "step": 30420 }, { "grad_norm": 0.22834797203540802, "learning_rate": 7.301386297697726e-05, "loss": 0.0328, "step": 30430 }, { "grad_norm": 0.36079707741737366, "learning_rate": 7.299551214210438e-05, "loss": 0.0284, "step": 30440 }, { "grad_norm": 0.26379233598709106, "learning_rate": 7.297715737792738e-05, "loss": 0.0299, "step": 30450 }, { "grad_norm": 0.34056609869003296, "learning_rate": 7.295879868758265e-05, "loss": 0.0365, "step": 30460 }, { "grad_norm": 0.27909359335899353, "learning_rate": 7.294043607420713e-05, "loss": 0.0241, "step": 30470 }, { "grad_norm": 0.4404783248901367, "learning_rate": 7.292206954093852e-05, "loss": 0.0326, "step": 30480 }, { "grad_norm": 0.3403978943824768, "learning_rate": 7.290369909091515e-05, "loss": 0.0279, "step": 30490 }, { "grad_norm": 0.27827155590057373, "learning_rate": 7.2885324727276e-05, "loss": 0.0335, "step": 30500 }, { "grad_norm": 0.4784037470817566, "learning_rate": 7.286694645316076e-05, "loss": 0.027, "step": 30510 }, { "grad_norm": 0.3069555163383484, "learning_rate": 7.284856427170982e-05, "loss": 0.0287, "step": 30520 }, { "grad_norm": 0.35724660754203796, "learning_rate": 7.283017818606414e-05, "loss": 0.0335, "step": 30530 }, { "grad_norm": 0.4682333469390869, "learning_rate": 7.28117881993654e-05, "loss": 0.0324, "step": 30540 }, { "grad_norm": 0.28905823826789856, "learning_rate": 7.279339431475598e-05, "loss": 0.0308, "step": 30550 }, { "grad_norm": 0.28026852011680603, "learning_rate": 7.277499653537887e-05, "loss": 0.0252, "step": 30560 }, { "grad_norm": 0.3310869038105011, "learning_rate": 7.275659486437776e-05, "loss": 0.035, "step": 30570 }, { "grad_norm": 0.3753606379032135, "learning_rate": 7.273818930489695e-05, "loss": 0.0378, "step": 30580 }, { "grad_norm": 0.3784131109714508, "learning_rate": 7.271977986008151e-05, "loss": 0.0285, "step": 30590 }, { "grad_norm": 0.3622722029685974, "learning_rate": 7.270136653307705e-05, "loss": 0.0317, "step": 30600 }, { "grad_norm": 0.5232582688331604, "learning_rate": 7.268294932702994e-05, "loss": 0.0305, "step": 30610 }, { "grad_norm": 0.3591683506965637, "learning_rate": 7.266452824508719e-05, "loss": 0.0372, "step": 30620 }, { "grad_norm": 0.3705812096595764, "learning_rate": 7.264610329039643e-05, "loss": 0.027, "step": 30630 }, { "grad_norm": 0.313448429107666, "learning_rate": 7.262767446610599e-05, "loss": 0.0278, "step": 30640 }, { "grad_norm": 0.4535473585128784, "learning_rate": 7.260924177536485e-05, "loss": 0.0347, "step": 30650 }, { "grad_norm": 0.24018573760986328, "learning_rate": 7.259080522132265e-05, "loss": 0.0287, "step": 30660 }, { "grad_norm": 0.26620519161224365, "learning_rate": 7.257236480712972e-05, "loss": 0.027, "step": 30670 }, { "grad_norm": 0.26319101452827454, "learning_rate": 7.255392053593697e-05, "loss": 0.0289, "step": 30680 }, { "grad_norm": 0.3415294885635376, "learning_rate": 7.253547241089607e-05, "loss": 0.0278, "step": 30690 }, { "grad_norm": 0.28170469403266907, "learning_rate": 7.251702043515927e-05, "loss": 0.0328, "step": 30700 }, { "grad_norm": 0.3491610884666443, "learning_rate": 7.249856461187952e-05, "loss": 0.0278, "step": 30710 }, { "grad_norm": 0.3255000710487366, "learning_rate": 7.248010494421042e-05, "loss": 0.0265, "step": 30720 }, { "grad_norm": 0.22239045798778534, "learning_rate": 7.246164143530622e-05, "loss": 0.0369, "step": 30730 }, { "grad_norm": 0.20379918813705444, "learning_rate": 7.244317408832181e-05, "loss": 0.0283, "step": 30740 }, { "grad_norm": 0.29082074761390686, "learning_rate": 7.242470290641279e-05, "loss": 0.0355, "step": 30750 }, { "grad_norm": 0.4032284617424011, "learning_rate": 7.240622789273536e-05, "loss": 0.0295, "step": 30760 }, { "grad_norm": 0.3437013328075409, "learning_rate": 7.238774905044638e-05, "loss": 0.0236, "step": 30770 }, { "grad_norm": 0.369773268699646, "learning_rate": 7.236926638270341e-05, "loss": 0.0279, "step": 30780 }, { "grad_norm": 0.36193254590034485, "learning_rate": 7.23507798926646e-05, "loss": 0.0301, "step": 30790 }, { "grad_norm": 0.334766685962677, "learning_rate": 7.23322895834888e-05, "loss": 0.0323, "step": 30800 }, { "grad_norm": 0.320147842168808, "learning_rate": 7.231379545833552e-05, "loss": 0.029, "step": 30810 }, { "grad_norm": 0.4648274779319763, "learning_rate": 7.229529752036487e-05, "loss": 0.032, "step": 30820 }, { "grad_norm": 0.36368483304977417, "learning_rate": 7.227679577273765e-05, "loss": 0.0299, "step": 30830 }, { "grad_norm": 0.27841702103614807, "learning_rate": 7.225829021861529e-05, "loss": 0.0293, "step": 30840 }, { "grad_norm": 0.24724051356315613, "learning_rate": 7.223978086115992e-05, "loss": 0.0234, "step": 30850 }, { "grad_norm": 0.3604191541671753, "learning_rate": 7.222126770353425e-05, "loss": 0.0367, "step": 30860 }, { "grad_norm": 0.3040475845336914, "learning_rate": 7.22027507489017e-05, "loss": 0.03, "step": 30870 }, { "grad_norm": 0.32406288385391235, "learning_rate": 7.218423000042627e-05, "loss": 0.0266, "step": 30880 }, { "grad_norm": 0.43160802125930786, "learning_rate": 7.216570546127268e-05, "loss": 0.0321, "step": 30890 }, { "grad_norm": 0.33526596426963806, "learning_rate": 7.214717713460626e-05, "loss": 0.0364, "step": 30900 }, { "grad_norm": 0.3162577152252197, "learning_rate": 7.2128645023593e-05, "loss": 0.0337, "step": 30910 }, { "grad_norm": 0.3169631361961365, "learning_rate": 7.211010913139951e-05, "loss": 0.0321, "step": 30920 }, { "grad_norm": 0.38358211517333984, "learning_rate": 7.209156946119308e-05, "loss": 0.0266, "step": 30930 }, { "grad_norm": 0.33625805377960205, "learning_rate": 7.207302601614166e-05, "loss": 0.0267, "step": 30940 }, { "grad_norm": 0.41792821884155273, "learning_rate": 7.205447879941378e-05, "loss": 0.0359, "step": 30950 }, { "grad_norm": 0.30934301018714905, "learning_rate": 7.203592781417866e-05, "loss": 0.0271, "step": 30960 }, { "grad_norm": 0.27510446310043335, "learning_rate": 7.201737306360617e-05, "loss": 0.0271, "step": 30970 }, { "grad_norm": 0.3657975494861603, "learning_rate": 7.19988145508668e-05, "loss": 0.0313, "step": 30980 }, { "grad_norm": 0.3356311023235321, "learning_rate": 7.198025227913168e-05, "loss": 0.0333, "step": 30990 }, { "grad_norm": 0.3656395971775055, "learning_rate": 7.196168625157261e-05, "loss": 0.0323, "step": 31000 }, { "grad_norm": 0.5786651372909546, "learning_rate": 7.194311647136201e-05, "loss": 0.0305, "step": 31010 }, { "grad_norm": 0.42146408557891846, "learning_rate": 7.192454294167297e-05, "loss": 0.025, "step": 31020 }, { "grad_norm": 0.3870835304260254, "learning_rate": 7.190596566567917e-05, "loss": 0.0271, "step": 31030 }, { "grad_norm": 0.3978036344051361, "learning_rate": 7.188738464655496e-05, "loss": 0.0315, "step": 31040 }, { "grad_norm": 0.3208034932613373, "learning_rate": 7.186879988747533e-05, "loss": 0.0262, "step": 31050 }, { "grad_norm": 0.17345814406871796, "learning_rate": 7.185021139161592e-05, "loss": 0.0242, "step": 31060 }, { "grad_norm": 0.2493760883808136, "learning_rate": 7.1831619162153e-05, "loss": 0.025, "step": 31070 }, { "grad_norm": 0.3761810064315796, "learning_rate": 7.181302320226345e-05, "loss": 0.024, "step": 31080 }, { "grad_norm": 0.26081687211990356, "learning_rate": 7.179442351512482e-05, "loss": 0.0339, "step": 31090 }, { "grad_norm": 0.435488760471344, "learning_rate": 7.177582010391528e-05, "loss": 0.0257, "step": 31100 }, { "grad_norm": 0.41787034273147583, "learning_rate": 7.175721297181366e-05, "loss": 0.025, "step": 31110 }, { "grad_norm": 0.49201810359954834, "learning_rate": 7.173860212199942e-05, "loss": 0.0285, "step": 31120 }, { "grad_norm": 0.3094644248485565, "learning_rate": 7.171998755765263e-05, "loss": 0.0283, "step": 31130 }, { "grad_norm": 0.3205461800098419, "learning_rate": 7.170136928195398e-05, "loss": 0.0267, "step": 31140 }, { "grad_norm": 0.29587289690971375, "learning_rate": 7.168274729808489e-05, "loss": 0.0237, "step": 31150 }, { "grad_norm": 0.27786728739738464, "learning_rate": 7.166412160922728e-05, "loss": 0.0308, "step": 31160 }, { "grad_norm": 0.3101310431957245, "learning_rate": 7.164549221856382e-05, "loss": 0.0286, "step": 31170 }, { "grad_norm": 0.3332190215587616, "learning_rate": 7.162685912927775e-05, "loss": 0.0391, "step": 31180 }, { "grad_norm": 0.32962772250175476, "learning_rate": 7.160822234455294e-05, "loss": 0.0308, "step": 31190 }, { "grad_norm": 0.42291343212127686, "learning_rate": 7.158958186757391e-05, "loss": 0.0314, "step": 31200 }, { "grad_norm": 0.37999144196510315, "learning_rate": 7.157093770152582e-05, "loss": 0.0299, "step": 31210 }, { "grad_norm": 0.28903549909591675, "learning_rate": 7.155228984959446e-05, "loss": 0.0284, "step": 31220 }, { "grad_norm": 0.35260772705078125, "learning_rate": 7.153363831496621e-05, "loss": 0.0334, "step": 31230 }, { "grad_norm": 0.3706074059009552, "learning_rate": 7.151498310082811e-05, "loss": 0.0348, "step": 31240 }, { "grad_norm": 0.30731087923049927, "learning_rate": 7.149632421036784e-05, "loss": 0.0242, "step": 31250 }, { "grad_norm": 0.44832825660705566, "learning_rate": 7.147766164677369e-05, "loss": 0.0506, "step": 31260 }, { "grad_norm": 0.25679299235343933, "learning_rate": 7.145899541323459e-05, "loss": 0.0219, "step": 31270 }, { "grad_norm": 0.4315117597579956, "learning_rate": 7.144032551294007e-05, "loss": 0.0376, "step": 31280 }, { "grad_norm": 0.35963770747184753, "learning_rate": 7.14216519490803e-05, "loss": 0.0317, "step": 31290 }, { "grad_norm": 0.5210920572280884, "learning_rate": 7.140297472484609e-05, "loss": 0.0407, "step": 31300 }, { "grad_norm": 0.24775610864162445, "learning_rate": 7.138429384342891e-05, "loss": 0.0466, "step": 31310 }, { "grad_norm": 0.4928927719593048, "learning_rate": 7.136560930802074e-05, "loss": 0.0276, "step": 31320 }, { "grad_norm": 0.32479676604270935, "learning_rate": 7.134692112181431e-05, "loss": 0.0267, "step": 31330 }, { "grad_norm": 0.36805659532546997, "learning_rate": 7.13282292880029e-05, "loss": 0.0253, "step": 31340 }, { "grad_norm": 0.41647225618362427, "learning_rate": 7.130953380978043e-05, "loss": 0.031, "step": 31350 }, { "grad_norm": 0.6539911031723022, "learning_rate": 7.129083469034144e-05, "loss": 0.0376, "step": 31360 }, { "grad_norm": 0.6173170208930969, "learning_rate": 7.127213193288112e-05, "loss": 0.0377, "step": 31370 }, { "grad_norm": 0.4843776226043701, "learning_rate": 7.125342554059522e-05, "loss": 0.0273, "step": 31380 }, { "grad_norm": 0.38432809710502625, "learning_rate": 7.12347155166802e-05, "loss": 0.032, "step": 31390 }, { "grad_norm": 0.34758317470550537, "learning_rate": 7.121600186433306e-05, "loss": 0.0259, "step": 31400 }, { "grad_norm": 0.3599315881729126, "learning_rate": 7.119728458675148e-05, "loss": 0.0264, "step": 31410 }, { "grad_norm": 0.47506701946258545, "learning_rate": 7.117856368713369e-05, "loss": 0.0265, "step": 31420 }, { "grad_norm": 0.43615707755088806, "learning_rate": 7.115983916867861e-05, "loss": 0.0302, "step": 31430 }, { "grad_norm": 0.290658563375473, "learning_rate": 7.114111103458574e-05, "loss": 0.0245, "step": 31440 }, { "grad_norm": 0.24836169183254242, "learning_rate": 7.11223792880552e-05, "loss": 0.0212, "step": 31450 }, { "grad_norm": 0.27781757712364197, "learning_rate": 7.110364393228773e-05, "loss": 0.0298, "step": 31460 }, { "grad_norm": 0.26717308163642883, "learning_rate": 7.108490497048471e-05, "loss": 0.024, "step": 31470 }, { "grad_norm": 0.44767487049102783, "learning_rate": 7.10661624058481e-05, "loss": 0.03, "step": 31480 }, { "grad_norm": 0.25487658381462097, "learning_rate": 7.10474162415805e-05, "loss": 0.031, "step": 31490 }, { "grad_norm": 0.32449889183044434, "learning_rate": 7.102866648088511e-05, "loss": 0.0325, "step": 31500 }, { "grad_norm": 0.3285079002380371, "learning_rate": 7.100991312696576e-05, "loss": 0.0287, "step": 31510 }, { "grad_norm": 0.4173943102359772, "learning_rate": 7.099115618302686e-05, "loss": 0.0266, "step": 31520 }, { "grad_norm": 0.3276234269142151, "learning_rate": 7.097239565227349e-05, "loss": 0.0345, "step": 31530 }, { "grad_norm": 0.696159303188324, "learning_rate": 7.09536315379113e-05, "loss": 0.0273, "step": 31540 }, { "grad_norm": 0.2785853147506714, "learning_rate": 7.093486384314656e-05, "loss": 0.022, "step": 31550 }, { "grad_norm": 0.27889057993888855, "learning_rate": 7.091609257118616e-05, "loss": 0.0247, "step": 31560 }, { "grad_norm": 0.4061129689216614, "learning_rate": 7.08973177252376e-05, "loss": 0.0286, "step": 31570 }, { "grad_norm": 0.410579115152359, "learning_rate": 7.087853930850898e-05, "loss": 0.0262, "step": 31580 }, { "grad_norm": 0.5335333347320557, "learning_rate": 7.085975732420903e-05, "loss": 0.026, "step": 31590 }, { "grad_norm": 0.9001672863960266, "learning_rate": 7.084097177554706e-05, "loss": 0.0211, "step": 31600 }, { "grad_norm": 0.3734598755836487, "learning_rate": 7.082218266573301e-05, "loss": 0.032, "step": 31610 }, { "grad_norm": 0.38588306307792664, "learning_rate": 7.080338999797743e-05, "loss": 0.0291, "step": 31620 }, { "grad_norm": 0.3885919451713562, "learning_rate": 7.07845937754915e-05, "loss": 0.0317, "step": 31630 }, { "grad_norm": 0.23583415150642395, "learning_rate": 7.076579400148693e-05, "loss": 0.0344, "step": 31640 }, { "grad_norm": 0.2596179246902466, "learning_rate": 7.074699067917611e-05, "loss": 0.0279, "step": 31650 }, { "grad_norm": 0.4629731774330139, "learning_rate": 7.072818381177201e-05, "loss": 0.0308, "step": 31660 }, { "grad_norm": 0.46613091230392456, "learning_rate": 7.070937340248823e-05, "loss": 0.0291, "step": 31670 }, { "grad_norm": 0.3054126501083374, "learning_rate": 7.069055945453893e-05, "loss": 0.0297, "step": 31680 }, { "grad_norm": 0.3069074749946594, "learning_rate": 7.067174197113892e-05, "loss": 0.0231, "step": 31690 }, { "grad_norm": 0.30264556407928467, "learning_rate": 7.065292095550355e-05, "loss": 0.0317, "step": 31700 }, { "grad_norm": 0.2484116405248642, "learning_rate": 7.063409641084887e-05, "loss": 0.0289, "step": 31710 }, { "grad_norm": 0.5487305521965027, "learning_rate": 7.061526834039145e-05, "loss": 0.0321, "step": 31720 }, { "grad_norm": 0.2962186336517334, "learning_rate": 7.05964367473485e-05, "loss": 0.0282, "step": 31730 }, { "grad_norm": 0.37934932112693787, "learning_rate": 7.057760163493783e-05, "loss": 0.0336, "step": 31740 }, { "grad_norm": 0.3806195855140686, "learning_rate": 7.055876300637783e-05, "loss": 0.0256, "step": 31750 }, { "grad_norm": 0.29465019702911377, "learning_rate": 7.053992086488753e-05, "loss": 0.0401, "step": 31760 }, { "grad_norm": 0.26929643750190735, "learning_rate": 7.052107521368651e-05, "loss": 0.0284, "step": 31770 }, { "grad_norm": 0.3343212306499481, "learning_rate": 7.0502226055995e-05, "loss": 0.0312, "step": 31780 }, { "grad_norm": 0.3189127445220947, "learning_rate": 7.048337339503379e-05, "loss": 0.0361, "step": 31790 }, { "grad_norm": 0.39348965883255005, "learning_rate": 7.046451723402427e-05, "loss": 0.0296, "step": 31800 }, { "grad_norm": 0.26840609312057495, "learning_rate": 7.044565757618848e-05, "loss": 0.0417, "step": 31810 }, { "grad_norm": 0.3575003445148468, "learning_rate": 7.042679442474899e-05, "loss": 0.0288, "step": 31820 }, { "grad_norm": 0.2285834401845932, "learning_rate": 7.040792778292902e-05, "loss": 0.0342, "step": 31830 }, { "grad_norm": 0.42408716678619385, "learning_rate": 7.038905765395234e-05, "loss": 0.0381, "step": 31840 }, { "grad_norm": 0.34862974286079407, "learning_rate": 7.037018404104334e-05, "loss": 0.0292, "step": 31850 }, { "grad_norm": 0.26601600646972656, "learning_rate": 7.035130694742702e-05, "loss": 0.0289, "step": 31860 }, { "grad_norm": 0.4060831665992737, "learning_rate": 7.033242637632897e-05, "loss": 0.0332, "step": 31870 }, { "grad_norm": 0.4039652943611145, "learning_rate": 7.031354233097534e-05, "loss": 0.0296, "step": 31880 }, { "grad_norm": 0.36963918805122375, "learning_rate": 7.029465481459289e-05, "loss": 0.039, "step": 31890 }, { "grad_norm": 0.29789766669273376, "learning_rate": 7.027576383040898e-05, "loss": 0.0241, "step": 31900 }, { "grad_norm": 0.30010202527046204, "learning_rate": 7.025686938165159e-05, "loss": 0.0296, "step": 31910 }, { "grad_norm": 0.3787395656108856, "learning_rate": 7.023797147154924e-05, "loss": 0.0427, "step": 31920 }, { "grad_norm": 0.3953520953655243, "learning_rate": 7.021907010333111e-05, "loss": 0.0231, "step": 31930 }, { "grad_norm": 0.23642432689666748, "learning_rate": 7.020016528022685e-05, "loss": 0.0252, "step": 31940 }, { "grad_norm": 0.3605647683143616, "learning_rate": 7.018125700546683e-05, "loss": 0.0341, "step": 31950 }, { "grad_norm": 0.42197275161743164, "learning_rate": 7.016234528228196e-05, "loss": 0.0305, "step": 31960 }, { "grad_norm": 0.30926451086997986, "learning_rate": 7.014343011390372e-05, "loss": 0.0337, "step": 31970 }, { "grad_norm": 0.2767103314399719, "learning_rate": 7.01245115035642e-05, "loss": 0.0244, "step": 31980 }, { "grad_norm": 0.384279727935791, "learning_rate": 7.010558945449606e-05, "loss": 0.0388, "step": 31990 }, { "grad_norm": 0.2671428918838501, "learning_rate": 7.008666396993258e-05, "loss": 0.0245, "step": 32000 }, { "grad_norm": 0.4014359712600708, "learning_rate": 7.006773505310759e-05, "loss": 0.0353, "step": 32010 }, { "grad_norm": 0.5049607753753662, "learning_rate": 7.004880270725553e-05, "loss": 0.0285, "step": 32020 }, { "grad_norm": 0.3632066547870636, "learning_rate": 7.002986693561144e-05, "loss": 0.0391, "step": 32030 }, { "grad_norm": 0.3325813114643097, "learning_rate": 7.001092774141089e-05, "loss": 0.0338, "step": 32040 }, { "grad_norm": 0.6639264822006226, "learning_rate": 6.999198512789009e-05, "loss": 0.0341, "step": 32050 }, { "grad_norm": 0.4620513319969177, "learning_rate": 6.997303909828584e-05, "loss": 0.0349, "step": 32060 }, { "grad_norm": 0.30987465381622314, "learning_rate": 6.995408965583544e-05, "loss": 0.0328, "step": 32070 }, { "grad_norm": 0.3543815314769745, "learning_rate": 6.993513680377688e-05, "loss": 0.0206, "step": 32080 }, { "grad_norm": 0.3289962112903595, "learning_rate": 6.991618054534868e-05, "loss": 0.0384, "step": 32090 }, { "grad_norm": 0.37710776925086975, "learning_rate": 6.989722088378991e-05, "loss": 0.0313, "step": 32100 }, { "grad_norm": 0.7929030060768127, "learning_rate": 6.987825782234027e-05, "loss": 0.0337, "step": 32110 }, { "grad_norm": 0.30906808376312256, "learning_rate": 6.985929136424006e-05, "loss": 0.0238, "step": 32120 }, { "grad_norm": 0.5469813942909241, "learning_rate": 6.984032151273012e-05, "loss": 0.0374, "step": 32130 }, { "grad_norm": 0.42479291558265686, "learning_rate": 6.982134827105186e-05, "loss": 0.0361, "step": 32140 }, { "grad_norm": 0.386257529258728, "learning_rate": 6.980237164244729e-05, "loss": 0.0321, "step": 32150 }, { "grad_norm": 0.35838496685028076, "learning_rate": 6.9783391630159e-05, "loss": 0.0251, "step": 32160 }, { "grad_norm": 0.2787940502166748, "learning_rate": 6.976440823743015e-05, "loss": 0.0281, "step": 32170 }, { "grad_norm": 0.299175500869751, "learning_rate": 6.974542146750451e-05, "loss": 0.0224, "step": 32180 }, { "grad_norm": 0.8952967524528503, "learning_rate": 6.972643132362637e-05, "loss": 0.0253, "step": 32190 }, { "grad_norm": 0.28037920594215393, "learning_rate": 6.970743780904064e-05, "loss": 0.0284, "step": 32200 }, { "grad_norm": 0.34492743015289307, "learning_rate": 6.968844092699277e-05, "loss": 0.0281, "step": 32210 }, { "grad_norm": 0.45620232820510864, "learning_rate": 6.966944068072883e-05, "loss": 0.0292, "step": 32220 }, { "grad_norm": 0.34837889671325684, "learning_rate": 6.965043707349545e-05, "loss": 0.0232, "step": 32230 }, { "grad_norm": 0.29331761598587036, "learning_rate": 6.963143010853982e-05, "loss": 0.03, "step": 32240 }, { "grad_norm": 0.39533644914627075, "learning_rate": 6.961241978910971e-05, "loss": 0.0355, "step": 32250 }, { "grad_norm": 0.40821149945259094, "learning_rate": 6.959340611845344e-05, "loss": 0.0291, "step": 32260 }, { "grad_norm": 0.4941357970237732, "learning_rate": 6.957438909981995e-05, "loss": 0.0277, "step": 32270 }, { "grad_norm": 0.32565340399742126, "learning_rate": 6.955536873645872e-05, "loss": 0.0214, "step": 32280 }, { "grad_norm": 0.3618609309196472, "learning_rate": 6.953634503161982e-05, "loss": 0.0261, "step": 32290 }, { "grad_norm": 0.29772618412971497, "learning_rate": 6.951731798855387e-05, "loss": 0.0247, "step": 32300 }, { "grad_norm": 0.29476311802864075, "learning_rate": 6.949828761051208e-05, "loss": 0.0312, "step": 32310 }, { "grad_norm": 0.2729078233242035, "learning_rate": 6.947925390074622e-05, "loss": 0.0229, "step": 32320 }, { "grad_norm": 0.295571506023407, "learning_rate": 6.946021686250863e-05, "loss": 0.0217, "step": 32330 }, { "grad_norm": 0.6180065870285034, "learning_rate": 6.94411764990522e-05, "loss": 0.0283, "step": 32340 }, { "grad_norm": 0.4812890589237213, "learning_rate": 6.942213281363044e-05, "loss": 0.0277, "step": 32350 }, { "grad_norm": 0.5047973990440369, "learning_rate": 6.940308580949737e-05, "loss": 0.033, "step": 32360 }, { "grad_norm": 0.2991088628768921, "learning_rate": 6.93840354899076e-05, "loss": 0.0241, "step": 32370 }, { "grad_norm": 1.5825037956237793, "learning_rate": 6.936498185811633e-05, "loss": 0.0315, "step": 32380 }, { "grad_norm": 0.27907928824424744, "learning_rate": 6.93459249173793e-05, "loss": 0.0229, "step": 32390 }, { "grad_norm": 0.3579094111919403, "learning_rate": 6.932686467095279e-05, "loss": 0.0325, "step": 32400 }, { "grad_norm": 0.4237314462661743, "learning_rate": 6.930780112209373e-05, "loss": 0.0396, "step": 32410 }, { "grad_norm": 0.5430887937545776, "learning_rate": 6.92887342740595e-05, "loss": 0.0294, "step": 32420 }, { "grad_norm": 0.33372262120246887, "learning_rate": 6.926966413010816e-05, "loss": 0.0356, "step": 32430 }, { "grad_norm": 0.6789897084236145, "learning_rate": 6.925059069349824e-05, "loss": 0.0267, "step": 32440 }, { "grad_norm": 0.2832772135734558, "learning_rate": 6.923151396748886e-05, "loss": 0.0294, "step": 32450 }, { "grad_norm": 0.24412646889686584, "learning_rate": 6.921243395533974e-05, "loss": 0.0286, "step": 32460 }, { "grad_norm": 0.30283889174461365, "learning_rate": 6.919335066031109e-05, "loss": 0.0313, "step": 32470 }, { "grad_norm": 0.29201266169548035, "learning_rate": 6.917426408566379e-05, "loss": 0.0305, "step": 32480 }, { "grad_norm": 0.4444900453090668, "learning_rate": 6.915517423465916e-05, "loss": 0.0342, "step": 32490 }, { "grad_norm": 0.30991095304489136, "learning_rate": 6.913608111055914e-05, "loss": 0.0295, "step": 32500 }, { "grad_norm": 0.5343118906021118, "learning_rate": 6.911698471662623e-05, "loss": 0.0239, "step": 32510 }, { "grad_norm": 0.3374629020690918, "learning_rate": 6.90978850561235e-05, "loss": 0.0249, "step": 32520 }, { "grad_norm": 0.32216840982437134, "learning_rate": 6.907878213231454e-05, "loss": 0.0231, "step": 32530 }, { "grad_norm": 0.44704878330230713, "learning_rate": 6.90596759484635e-05, "loss": 0.0283, "step": 32540 }, { "grad_norm": 0.41643422842025757, "learning_rate": 6.904056650783514e-05, "loss": 0.0284, "step": 32550 }, { "grad_norm": 0.3713257610797882, "learning_rate": 6.902145381369471e-05, "loss": 0.031, "step": 32560 }, { "grad_norm": 0.4000242054462433, "learning_rate": 6.900233786930808e-05, "loss": 0.0313, "step": 32570 }, { "grad_norm": 0.3008486330509186, "learning_rate": 6.898321867794161e-05, "loss": 0.03, "step": 32580 }, { "grad_norm": 0.3401726484298706, "learning_rate": 6.896409624286226e-05, "loss": 0.0342, "step": 32590 }, { "grad_norm": 0.37040644884109497, "learning_rate": 6.894497056733754e-05, "loss": 0.0305, "step": 32600 }, { "grad_norm": 0.3483400046825409, "learning_rate": 6.89258416546355e-05, "loss": 0.0324, "step": 32610 }, { "grad_norm": 0.41284653544425964, "learning_rate": 6.890670950802474e-05, "loss": 0.0304, "step": 32620 }, { "grad_norm": 0.43503955006599426, "learning_rate": 6.88875741307744e-05, "loss": 0.0297, "step": 32630 }, { "grad_norm": 0.2432532161474228, "learning_rate": 6.886843552615425e-05, "loss": 0.0266, "step": 32640 }, { "grad_norm": 0.285188764333725, "learning_rate": 6.884929369743451e-05, "loss": 0.0197, "step": 32650 }, { "grad_norm": 0.3169398903846741, "learning_rate": 6.8830148647886e-05, "loss": 0.0266, "step": 32660 }, { "grad_norm": 0.3147471845149994, "learning_rate": 6.88110003807801e-05, "loss": 0.0347, "step": 32670 }, { "grad_norm": 0.32156360149383545, "learning_rate": 6.87918488993887e-05, "loss": 0.0214, "step": 32680 }, { "grad_norm": 0.3922985792160034, "learning_rate": 6.877269420698431e-05, "loss": 0.0306, "step": 32690 }, { "grad_norm": 0.35301971435546875, "learning_rate": 6.875353630683989e-05, "loss": 0.0288, "step": 32700 }, { "grad_norm": 0.4308597147464752, "learning_rate": 6.873437520222905e-05, "loss": 0.0345, "step": 32710 }, { "grad_norm": 0.20913687348365784, "learning_rate": 6.871521089642585e-05, "loss": 0.0264, "step": 32720 }, { "grad_norm": 0.25792986154556274, "learning_rate": 6.869604339270498e-05, "loss": 0.0276, "step": 32730 }, { "grad_norm": 0.4194072186946869, "learning_rate": 6.867687269434164e-05, "loss": 0.0352, "step": 32740 }, { "grad_norm": 0.34618204832077026, "learning_rate": 6.865769880461156e-05, "loss": 0.0269, "step": 32750 }, { "grad_norm": 0.4057060182094574, "learning_rate": 6.863852172679104e-05, "loss": 0.0273, "step": 32760 }, { "grad_norm": 0.30909886956214905, "learning_rate": 6.861934146415693e-05, "loss": 0.0284, "step": 32770 }, { "grad_norm": 0.48226064443588257, "learning_rate": 6.86001580199866e-05, "loss": 0.033, "step": 32780 }, { "grad_norm": 0.30117419362068176, "learning_rate": 6.858097139755798e-05, "loss": 0.0232, "step": 32790 }, { "grad_norm": 0.42170649766921997, "learning_rate": 6.856178160014955e-05, "loss": 0.0246, "step": 32800 }, { "grad_norm": 0.3781229853630066, "learning_rate": 6.85425886310403e-05, "loss": 0.0238, "step": 32810 }, { "grad_norm": 0.3533317446708679, "learning_rate": 6.852339249350979e-05, "loss": 0.0226, "step": 32820 }, { "grad_norm": 0.23201562464237213, "learning_rate": 6.850419319083812e-05, "loss": 0.0328, "step": 32830 }, { "grad_norm": 0.3460427224636078, "learning_rate": 6.848499072630592e-05, "loss": 0.0225, "step": 32840 }, { "grad_norm": 0.46242865920066833, "learning_rate": 6.846578510319439e-05, "loss": 0.0301, "step": 32850 }, { "grad_norm": 0.28087541460990906, "learning_rate": 6.844657632478519e-05, "loss": 0.0294, "step": 32860 }, { "grad_norm": 0.2488866150379181, "learning_rate": 6.842736439436063e-05, "loss": 0.0214, "step": 32870 }, { "grad_norm": 0.2708098888397217, "learning_rate": 6.84081493152035e-05, "loss": 0.024, "step": 32880 }, { "grad_norm": 0.25235018134117126, "learning_rate": 6.83889310905971e-05, "loss": 0.0213, "step": 32890 }, { "grad_norm": 0.35957571864128113, "learning_rate": 6.836970972382533e-05, "loss": 0.0336, "step": 32900 }, { "grad_norm": 0.2316008359193802, "learning_rate": 6.835048521817257e-05, "loss": 0.0285, "step": 32910 }, { "grad_norm": 0.28733155131340027, "learning_rate": 6.833125757692379e-05, "loss": 0.0278, "step": 32920 }, { "grad_norm": 0.3193526864051819, "learning_rate": 6.831202680336441e-05, "loss": 0.0235, "step": 32930 }, { "grad_norm": 0.2695889472961426, "learning_rate": 6.829279290078052e-05, "loss": 0.023, "step": 32940 }, { "grad_norm": 0.30597832798957825, "learning_rate": 6.827355587245863e-05, "loss": 0.0314, "step": 32950 }, { "grad_norm": 0.2284800410270691, "learning_rate": 6.82543157216858e-05, "loss": 0.0242, "step": 32960 }, { "grad_norm": 0.29127681255340576, "learning_rate": 6.823507245174969e-05, "loss": 0.0291, "step": 32970 }, { "grad_norm": 0.4945136606693268, "learning_rate": 6.821582606593841e-05, "loss": 0.0237, "step": 32980 }, { "grad_norm": 0.2909523844718933, "learning_rate": 6.81965765675407e-05, "loss": 0.0196, "step": 32990 }, { "grad_norm": 0.32204654812812805, "learning_rate": 6.81773239598457e-05, "loss": 0.0288, "step": 33000 }, { "grad_norm": 0.2920638918876648, "learning_rate": 6.815806824614319e-05, "loss": 0.0285, "step": 33010 }, { "grad_norm": 0.3411504030227661, "learning_rate": 6.813880942972343e-05, "loss": 0.0247, "step": 33020 }, { "grad_norm": 0.5539326071739197, "learning_rate": 6.811954751387726e-05, "loss": 0.0302, "step": 33030 }, { "grad_norm": 0.3003183901309967, "learning_rate": 6.810028250189598e-05, "loss": 0.0259, "step": 33040 }, { "grad_norm": 0.4573742151260376, "learning_rate": 6.808101439707147e-05, "loss": 0.0349, "step": 33050 }, { "grad_norm": 0.383136123418808, "learning_rate": 6.806174320269609e-05, "loss": 0.0306, "step": 33060 }, { "grad_norm": 0.38805490732192993, "learning_rate": 6.804246892206281e-05, "loss": 0.0323, "step": 33070 }, { "grad_norm": 0.3642010986804962, "learning_rate": 6.802319155846506e-05, "loss": 0.0424, "step": 33080 }, { "grad_norm": 0.45387008786201477, "learning_rate": 6.800391111519679e-05, "loss": 0.0315, "step": 33090 }, { "grad_norm": 0.39184126257896423, "learning_rate": 6.798462759555253e-05, "loss": 0.0346, "step": 33100 }, { "grad_norm": 0.24072131514549255, "learning_rate": 6.79653410028273e-05, "loss": 0.0267, "step": 33110 }, { "grad_norm": 0.222056046128273, "learning_rate": 6.794605134031663e-05, "loss": 0.0275, "step": 33120 }, { "grad_norm": 0.5211431980133057, "learning_rate": 6.792675861131661e-05, "loss": 0.0342, "step": 33130 }, { "grad_norm": 0.34227949380874634, "learning_rate": 6.790746281912386e-05, "loss": 0.0365, "step": 33140 }, { "grad_norm": 0.24937501549720764, "learning_rate": 6.788816396703546e-05, "loss": 0.0257, "step": 33150 }, { "grad_norm": 0.515194296836853, "learning_rate": 6.78688620583491e-05, "loss": 0.0313, "step": 33160 }, { "grad_norm": 0.3960909843444824, "learning_rate": 6.784955709636292e-05, "loss": 0.0364, "step": 33170 }, { "grad_norm": 0.4370267987251282, "learning_rate": 6.783024908437564e-05, "loss": 0.0266, "step": 33180 }, { "grad_norm": 0.5490837693214417, "learning_rate": 6.781093802568641e-05, "loss": 0.0324, "step": 33190 }, { "grad_norm": 0.31698137521743774, "learning_rate": 6.779162392359504e-05, "loss": 0.0265, "step": 33200 }, { "grad_norm": 0.49086886644363403, "learning_rate": 6.777230678140172e-05, "loss": 0.0298, "step": 33210 }, { "grad_norm": 0.34638291597366333, "learning_rate": 6.775298660240726e-05, "loss": 0.033, "step": 33220 }, { "grad_norm": 0.42303094267845154, "learning_rate": 6.773366338991292e-05, "loss": 0.0422, "step": 33230 }, { "grad_norm": 0.28782233595848083, "learning_rate": 6.771433714722052e-05, "loss": 0.0268, "step": 33240 }, { "grad_norm": 0.31980934739112854, "learning_rate": 6.769500787763239e-05, "loss": 0.0349, "step": 33250 }, { "grad_norm": 0.22178365290164948, "learning_rate": 6.76756755844514e-05, "loss": 0.0316, "step": 33260 }, { "grad_norm": 0.41812464594841003, "learning_rate": 6.765634027098087e-05, "loss": 0.0271, "step": 33270 }, { "grad_norm": 0.3282245993614197, "learning_rate": 6.763700194052468e-05, "loss": 0.026, "step": 33280 }, { "grad_norm": 0.26924118399620056, "learning_rate": 6.761766059638723e-05, "loss": 0.0248, "step": 33290 }, { "grad_norm": 0.3778209090232849, "learning_rate": 6.759831624187345e-05, "loss": 0.0278, "step": 33300 }, { "grad_norm": 0.35266637802124023, "learning_rate": 6.757896888028871e-05, "loss": 0.0272, "step": 33310 }, { "grad_norm": 0.2985454201698303, "learning_rate": 6.7559618514939e-05, "loss": 0.0286, "step": 33320 }, { "grad_norm": 0.3229290544986725, "learning_rate": 6.754026514913073e-05, "loss": 0.0272, "step": 33330 }, { "grad_norm": 0.2670467793941498, "learning_rate": 6.752090878617087e-05, "loss": 0.035, "step": 33340 }, { "grad_norm": 0.42218828201293945, "learning_rate": 6.75015494293669e-05, "loss": 0.0307, "step": 33350 }, { "grad_norm": 0.6954575181007385, "learning_rate": 6.74821870820268e-05, "loss": 0.0255, "step": 33360 }, { "grad_norm": 0.2903476655483246, "learning_rate": 6.746282174745907e-05, "loss": 0.0336, "step": 33370 }, { "grad_norm": 0.39897850155830383, "learning_rate": 6.744345342897271e-05, "loss": 0.0356, "step": 33380 }, { "grad_norm": 0.23960712552070618, "learning_rate": 6.742408212987724e-05, "loss": 0.0368, "step": 33390 }, { "grad_norm": 0.29500338435173035, "learning_rate": 6.740470785348269e-05, "loss": 0.0227, "step": 33400 }, { "grad_norm": 0.26380255818367004, "learning_rate": 6.738533060309958e-05, "loss": 0.0301, "step": 33410 }, { "grad_norm": 0.28542566299438477, "learning_rate": 6.736595038203894e-05, "loss": 0.0286, "step": 33420 }, { "grad_norm": 0.28572267293930054, "learning_rate": 6.734656719361236e-05, "loss": 0.0306, "step": 33430 }, { "grad_norm": 0.24923469126224518, "learning_rate": 6.732718104113189e-05, "loss": 0.0296, "step": 33440 }, { "grad_norm": 0.46059420704841614, "learning_rate": 6.730779192791006e-05, "loss": 0.0285, "step": 33450 }, { "grad_norm": 0.28188544511795044, "learning_rate": 6.728839985725997e-05, "loss": 0.0254, "step": 33460 }, { "grad_norm": 0.3178362250328064, "learning_rate": 6.726900483249517e-05, "loss": 0.0237, "step": 33470 }, { "grad_norm": 0.30517706274986267, "learning_rate": 6.724960685692976e-05, "loss": 0.0266, "step": 33480 }, { "grad_norm": 0.24272239208221436, "learning_rate": 6.723020593387833e-05, "loss": 0.0232, "step": 33490 }, { "grad_norm": 0.37299785017967224, "learning_rate": 6.721080206665593e-05, "loss": 0.0263, "step": 33500 }, { "grad_norm": 0.28928688168525696, "learning_rate": 6.719139525857819e-05, "loss": 0.0355, "step": 33510 }, { "grad_norm": 0.2554408609867096, "learning_rate": 6.717198551296117e-05, "loss": 0.0229, "step": 33520 }, { "grad_norm": 0.28265321254730225, "learning_rate": 6.715257283312148e-05, "loss": 0.0299, "step": 33530 }, { "grad_norm": 0.2526961863040924, "learning_rate": 6.713315722237623e-05, "loss": 0.0272, "step": 33540 }, { "grad_norm": 0.287674218416214, "learning_rate": 6.7113738684043e-05, "loss": 0.0313, "step": 33550 }, { "grad_norm": 0.3889859616756439, "learning_rate": 6.709431722143989e-05, "loss": 0.0264, "step": 33560 }, { "grad_norm": 0.375155508518219, "learning_rate": 6.70748928378855e-05, "loss": 0.0344, "step": 33570 }, { "grad_norm": 0.49618831276893616, "learning_rate": 6.705546553669891e-05, "loss": 0.0252, "step": 33580 }, { "grad_norm": 0.2199762761592865, "learning_rate": 6.703603532119974e-05, "loss": 0.0217, "step": 33590 }, { "grad_norm": 0.3614732623100281, "learning_rate": 6.701660219470808e-05, "loss": 0.0301, "step": 33600 }, { "grad_norm": 0.5684870481491089, "learning_rate": 6.69971661605445e-05, "loss": 0.0268, "step": 33610 }, { "grad_norm": 0.3109402060508728, "learning_rate": 6.697772722203008e-05, "loss": 0.0251, "step": 33620 }, { "grad_norm": 0.3909611701965332, "learning_rate": 6.695828538248643e-05, "loss": 0.0281, "step": 33630 }, { "grad_norm": 0.2950866222381592, "learning_rate": 6.693884064523563e-05, "loss": 0.0316, "step": 33640 }, { "grad_norm": 0.24798008799552917, "learning_rate": 6.691939301360023e-05, "loss": 0.0227, "step": 33650 }, { "grad_norm": 0.5389138460159302, "learning_rate": 6.689994249090333e-05, "loss": 0.0286, "step": 33660 }, { "grad_norm": 0.48656195402145386, "learning_rate": 6.688048908046845e-05, "loss": 0.0319, "step": 33670 }, { "grad_norm": 0.24779899418354034, "learning_rate": 6.686103278561969e-05, "loss": 0.0239, "step": 33680 }, { "grad_norm": 0.27403396368026733, "learning_rate": 6.684157360968156e-05, "loss": 0.0331, "step": 33690 }, { "grad_norm": 0.3524640202522278, "learning_rate": 6.682211155597911e-05, "loss": 0.0221, "step": 33700 }, { "grad_norm": 0.27836257219314575, "learning_rate": 6.680264662783789e-05, "loss": 0.0278, "step": 33710 }, { "grad_norm": 0.37532269954681396, "learning_rate": 6.678317882858391e-05, "loss": 0.0302, "step": 33720 }, { "grad_norm": 0.3199727535247803, "learning_rate": 6.67637081615437e-05, "loss": 0.0287, "step": 33730 }, { "grad_norm": 0.3195752799510956, "learning_rate": 6.674423463004427e-05, "loss": 0.0233, "step": 33740 }, { "grad_norm": 0.3479284346103668, "learning_rate": 6.672475823741308e-05, "loss": 0.025, "step": 33750 }, { "grad_norm": 0.4440537393093109, "learning_rate": 6.670527898697811e-05, "loss": 0.0328, "step": 33760 }, { "grad_norm": 0.2990867495536804, "learning_rate": 6.668579688206788e-05, "loss": 0.0327, "step": 33770 }, { "grad_norm": 0.9785739183425903, "learning_rate": 6.666631192601131e-05, "loss": 0.028, "step": 33780 }, { "grad_norm": 0.4335581064224243, "learning_rate": 6.664682412213785e-05, "loss": 0.028, "step": 33790 }, { "grad_norm": 0.3941774368286133, "learning_rate": 6.662733347377745e-05, "loss": 0.0308, "step": 33800 }, { "grad_norm": 0.23155485093593597, "learning_rate": 6.660783998426051e-05, "loss": 0.0316, "step": 33810 }, { "grad_norm": 0.4367557168006897, "learning_rate": 6.658834365691794e-05, "loss": 0.0259, "step": 33820 }, { "grad_norm": 0.2665696442127228, "learning_rate": 6.656884449508115e-05, "loss": 0.0237, "step": 33830 }, { "grad_norm": 0.3488723337650299, "learning_rate": 6.654934250208198e-05, "loss": 0.0214, "step": 33840 }, { "grad_norm": 0.24777977168560028, "learning_rate": 6.65298376812528e-05, "loss": 0.0256, "step": 33850 }, { "grad_norm": 0.24118424952030182, "learning_rate": 6.651033003592646e-05, "loss": 0.0262, "step": 33860 }, { "grad_norm": 0.313561350107193, "learning_rate": 6.649081956943626e-05, "loss": 0.025, "step": 33870 }, { "grad_norm": 0.37861865758895874, "learning_rate": 6.647130628511604e-05, "loss": 0.0269, "step": 33880 }, { "grad_norm": 0.3101986050605774, "learning_rate": 6.645179018630005e-05, "loss": 0.024, "step": 33890 }, { "grad_norm": 0.37314894795417786, "learning_rate": 6.643227127632309e-05, "loss": 0.0275, "step": 33900 }, { "grad_norm": 0.33824458718299866, "learning_rate": 6.641274955852038e-05, "loss": 0.0308, "step": 33910 }, { "grad_norm": 0.3921222686767578, "learning_rate": 6.639322503622768e-05, "loss": 0.0457, "step": 33920 }, { "grad_norm": 0.47757068276405334, "learning_rate": 6.637369771278116e-05, "loss": 0.0301, "step": 33930 }, { "grad_norm": 0.265060693025589, "learning_rate": 6.635416759151751e-05, "loss": 0.0283, "step": 33940 }, { "grad_norm": 0.34517067670822144, "learning_rate": 6.633463467577394e-05, "loss": 0.0415, "step": 33950 }, { "grad_norm": 0.48621219396591187, "learning_rate": 6.631509896888803e-05, "loss": 0.0276, "step": 33960 }, { "grad_norm": 0.24111847579479218, "learning_rate": 6.629556047419794e-05, "loss": 0.0231, "step": 33970 }, { "grad_norm": 0.351963609457016, "learning_rate": 6.627601919504223e-05, "loss": 0.0243, "step": 33980 }, { "grad_norm": 0.3513329327106476, "learning_rate": 6.625647513476001e-05, "loss": 0.0247, "step": 33990 }, { "grad_norm": 0.4553718566894531, "learning_rate": 6.62369282966908e-05, "loss": 0.0401, "step": 34000 }, { "grad_norm": 0.4219479262828827, "learning_rate": 6.621737868417464e-05, "loss": 0.0336, "step": 34010 }, { "grad_norm": 0.27129673957824707, "learning_rate": 6.619782630055198e-05, "loss": 0.031, "step": 34020 }, { "grad_norm": 0.3128332495689392, "learning_rate": 6.617827114916382e-05, "loss": 0.0287, "step": 34030 }, { "grad_norm": 0.2261677086353302, "learning_rate": 6.615871323335161e-05, "loss": 0.0234, "step": 34040 }, { "grad_norm": 0.2978099584579468, "learning_rate": 6.613915255645725e-05, "loss": 0.0255, "step": 34050 }, { "grad_norm": 0.3839651048183441, "learning_rate": 6.611958912182312e-05, "loss": 0.0234, "step": 34060 }, { "grad_norm": 0.3495449125766754, "learning_rate": 6.610002293279207e-05, "loss": 0.0261, "step": 34070 }, { "grad_norm": 0.2760556638240814, "learning_rate": 6.608045399270746e-05, "loss": 0.022, "step": 34080 }, { "grad_norm": 0.2639803886413574, "learning_rate": 6.606088230491304e-05, "loss": 0.0248, "step": 34090 }, { "grad_norm": 0.483664870262146, "learning_rate": 6.604130787275312e-05, "loss": 0.0232, "step": 34100 }, { "grad_norm": 0.2620958089828491, "learning_rate": 6.602173069957242e-05, "loss": 0.0223, "step": 34110 }, { "grad_norm": 0.320480078458786, "learning_rate": 6.600215078871612e-05, "loss": 0.0284, "step": 34120 }, { "grad_norm": 0.31096699833869934, "learning_rate": 6.598256814352992e-05, "loss": 0.0245, "step": 34130 }, { "grad_norm": 0.33425402641296387, "learning_rate": 6.596298276735995e-05, "loss": 0.0249, "step": 34140 }, { "grad_norm": 0.35670384764671326, "learning_rate": 6.594339466355282e-05, "loss": 0.0257, "step": 34150 }, { "grad_norm": 0.3155776262283325, "learning_rate": 6.592380383545558e-05, "loss": 0.0325, "step": 34160 }, { "grad_norm": 0.36477285623550415, "learning_rate": 6.590421028641577e-05, "loss": 0.0301, "step": 34170 }, { "grad_norm": 0.337516188621521, "learning_rate": 6.588461401978143e-05, "loss": 0.0269, "step": 34180 }, { "grad_norm": 0.4753612279891968, "learning_rate": 6.586501503890099e-05, "loss": 0.0299, "step": 34190 }, { "grad_norm": 0.28013792634010315, "learning_rate": 6.584541334712338e-05, "loss": 0.0246, "step": 34200 }, { "grad_norm": 0.321835994720459, "learning_rate": 6.582580894779802e-05, "loss": 0.0255, "step": 34210 }, { "grad_norm": 0.3273105323314667, "learning_rate": 6.580620184427473e-05, "loss": 0.022, "step": 34220 }, { "grad_norm": 0.3347771465778351, "learning_rate": 6.578659203990385e-05, "loss": 0.0253, "step": 34230 }, { "grad_norm": 0.290671706199646, "learning_rate": 6.576697953803615e-05, "loss": 0.0234, "step": 34240 }, { "grad_norm": 0.26692456007003784, "learning_rate": 6.57473643420229e-05, "loss": 0.0287, "step": 34250 }, { "grad_norm": 0.32529115676879883, "learning_rate": 6.572774645521574e-05, "loss": 0.0293, "step": 34260 }, { "grad_norm": 0.47375211119651794, "learning_rate": 6.570812588096688e-05, "loss": 0.0353, "step": 34270 }, { "grad_norm": 0.2692136764526367, "learning_rate": 6.568850262262893e-05, "loss": 0.0233, "step": 34280 }, { "grad_norm": 0.3294224143028259, "learning_rate": 6.566887668355497e-05, "loss": 0.0237, "step": 34290 }, { "grad_norm": 0.3931708037853241, "learning_rate": 6.564924806709851e-05, "loss": 0.0208, "step": 34300 }, { "grad_norm": 0.3003189265727997, "learning_rate": 6.562961677661359e-05, "loss": 0.0314, "step": 34310 }, { "grad_norm": 0.40803998708724976, "learning_rate": 6.56099828154546e-05, "loss": 0.0304, "step": 34320 }, { "grad_norm": 0.9697645306587219, "learning_rate": 6.55903461869765e-05, "loss": 0.0267, "step": 34330 }, { "grad_norm": 0.27885255217552185, "learning_rate": 6.557070689453465e-05, "loss": 0.0232, "step": 34340 }, { "grad_norm": 0.25498098134994507, "learning_rate": 6.555106494148482e-05, "loss": 0.0367, "step": 34350 }, { "grad_norm": 0.3772563934326172, "learning_rate": 6.553142033118333e-05, "loss": 0.0302, "step": 34360 }, { "grad_norm": 0.3802741467952728, "learning_rate": 6.551177306698688e-05, "loss": 0.0405, "step": 34370 }, { "grad_norm": 0.3328087627887726, "learning_rate": 6.549212315225267e-05, "loss": 0.0254, "step": 34380 }, { "grad_norm": 0.24528473615646362, "learning_rate": 6.547247059033833e-05, "loss": 0.0281, "step": 34390 }, { "grad_norm": 0.4122452437877655, "learning_rate": 6.545281538460193e-05, "loss": 0.0306, "step": 34400 }, { "grad_norm": 0.28875622153282166, "learning_rate": 6.543315753840202e-05, "loss": 0.0278, "step": 34410 }, { "grad_norm": 0.2151126265525818, "learning_rate": 6.541349705509758e-05, "loss": 0.0254, "step": 34420 }, { "grad_norm": 0.4577282965183258, "learning_rate": 6.539383393804805e-05, "loss": 0.0262, "step": 34430 }, { "grad_norm": 0.3699389100074768, "learning_rate": 6.537416819061333e-05, "loss": 0.0352, "step": 34440 }, { "grad_norm": 0.2330148071050644, "learning_rate": 6.535449981615375e-05, "loss": 0.0309, "step": 34450 }, { "grad_norm": 0.2671357989311218, "learning_rate": 6.53348288180301e-05, "loss": 0.0265, "step": 34460 }, { "grad_norm": 0.3852793872356415, "learning_rate": 6.531515519960361e-05, "loss": 0.0302, "step": 34470 }, { "grad_norm": 0.3402911126613617, "learning_rate": 6.529547896423597e-05, "loss": 0.0241, "step": 34480 }, { "grad_norm": 0.5075064897537231, "learning_rate": 6.52758001152893e-05, "loss": 0.0314, "step": 34490 }, { "grad_norm": 0.23718905448913574, "learning_rate": 6.525611865612618e-05, "loss": 0.0205, "step": 34500 }, { "grad_norm": 0.3658240735530853, "learning_rate": 6.523643459010966e-05, "loss": 0.0309, "step": 34510 }, { "grad_norm": 0.35743430256843567, "learning_rate": 6.521674792060317e-05, "loss": 0.032, "step": 34520 }, { "grad_norm": 0.35132738947868347, "learning_rate": 6.519705865097063e-05, "loss": 0.0282, "step": 34530 }, { "grad_norm": 0.294929176568985, "learning_rate": 6.517736678457641e-05, "loss": 0.0288, "step": 34540 }, { "grad_norm": 0.3950045704841614, "learning_rate": 6.515767232478534e-05, "loss": 0.0293, "step": 34550 }, { "grad_norm": 0.22962281107902527, "learning_rate": 6.51379752749626e-05, "loss": 0.0209, "step": 34560 }, { "grad_norm": 0.27340298891067505, "learning_rate": 6.511827563847393e-05, "loss": 0.0278, "step": 34570 }, { "grad_norm": 0.3090802729129791, "learning_rate": 6.509857341868542e-05, "loss": 0.0269, "step": 34580 }, { "grad_norm": 0.32507988810539246, "learning_rate": 6.507886861896367e-05, "loss": 0.0283, "step": 34590 }, { "grad_norm": 0.33898618817329407, "learning_rate": 6.505916124267567e-05, "loss": 0.0315, "step": 34600 }, { "grad_norm": 0.3714722692966461, "learning_rate": 6.503945129318891e-05, "loss": 0.0273, "step": 34610 }, { "grad_norm": 0.45056864619255066, "learning_rate": 6.501973877387122e-05, "loss": 0.0223, "step": 34620 }, { "grad_norm": 0.7479884028434753, "learning_rate": 6.500002368809098e-05, "loss": 0.0255, "step": 34630 }, { "grad_norm": 0.424980103969574, "learning_rate": 6.498030603921694e-05, "loss": 0.0286, "step": 34640 }, { "grad_norm": 0.34906208515167236, "learning_rate": 6.496058583061832e-05, "loss": 0.0289, "step": 34650 }, { "grad_norm": 0.3388766646385193, "learning_rate": 6.494086306566475e-05, "loss": 0.0225, "step": 34660 }, { "grad_norm": 0.359037309885025, "learning_rate": 6.492113774772632e-05, "loss": 0.0274, "step": 34670 }, { "grad_norm": 0.4420201778411865, "learning_rate": 6.490140988017354e-05, "loss": 0.0313, "step": 34680 }, { "grad_norm": 0.5823745727539062, "learning_rate": 6.488167946637736e-05, "loss": 0.0303, "step": 34690 }, { "grad_norm": 0.30559104681015015, "learning_rate": 6.486194650970915e-05, "loss": 0.0276, "step": 34700 }, { "grad_norm": 0.4940708875656128, "learning_rate": 6.48422110135408e-05, "loss": 0.0346, "step": 34710 }, { "grad_norm": 0.2529509961605072, "learning_rate": 6.482247298124451e-05, "loss": 0.0329, "step": 34720 }, { "grad_norm": 0.1894005537033081, "learning_rate": 6.480273241619297e-05, "loss": 0.0265, "step": 34730 }, { "grad_norm": 0.3017736077308655, "learning_rate": 6.478298932175933e-05, "loss": 0.0238, "step": 34740 }, { "grad_norm": 0.2855478525161743, "learning_rate": 6.476324370131712e-05, "loss": 0.0302, "step": 34750 }, { "grad_norm": 0.4005982279777527, "learning_rate": 6.474349555824036e-05, "loss": 0.0342, "step": 34760 }, { "grad_norm": 0.4337940216064453, "learning_rate": 6.472374489590342e-05, "loss": 0.0234, "step": 34770 }, { "grad_norm": 0.2606973648071289, "learning_rate": 6.470399171768118e-05, "loss": 0.0422, "step": 34780 }, { "grad_norm": 0.3030053675174713, "learning_rate": 6.468423602694891e-05, "loss": 0.0248, "step": 34790 }, { "grad_norm": 0.233213871717453, "learning_rate": 6.466447782708232e-05, "loss": 0.0246, "step": 34800 }, { "grad_norm": 0.21975630521774292, "learning_rate": 6.464471712145754e-05, "loss": 0.0249, "step": 34810 }, { "grad_norm": 0.2512769401073456, "learning_rate": 6.462495391345114e-05, "loss": 0.0252, "step": 34820 }, { "grad_norm": 0.4317028522491455, "learning_rate": 6.46051882064401e-05, "loss": 0.0237, "step": 34830 }, { "grad_norm": 0.3101250231266022, "learning_rate": 6.458542000380186e-05, "loss": 0.0282, "step": 34840 }, { "grad_norm": 0.3814540207386017, "learning_rate": 6.456564930891424e-05, "loss": 0.0252, "step": 34850 }, { "grad_norm": 0.3218628168106079, "learning_rate": 6.454587612515555e-05, "loss": 0.0345, "step": 34860 }, { "grad_norm": 0.23276862502098083, "learning_rate": 6.452610045590444e-05, "loss": 0.0238, "step": 34870 }, { "grad_norm": 0.3760179877281189, "learning_rate": 6.450632230454005e-05, "loss": 0.0252, "step": 34880 }, { "grad_norm": 0.3162498474121094, "learning_rate": 6.448654167444195e-05, "loss": 0.0303, "step": 34890 }, { "grad_norm": 0.45646288990974426, "learning_rate": 6.446675856899005e-05, "loss": 0.0211, "step": 34900 }, { "grad_norm": 0.1994081735610962, "learning_rate": 6.444697299156481e-05, "loss": 0.023, "step": 34910 }, { "grad_norm": 0.40784353017807007, "learning_rate": 6.442718494554701e-05, "loss": 0.0282, "step": 34920 }, { "grad_norm": 0.7415485978126526, "learning_rate": 6.440739443431787e-05, "loss": 0.019, "step": 34930 }, { "grad_norm": 0.2429780215024948, "learning_rate": 6.438760146125906e-05, "loss": 0.0184, "step": 34940 }, { "grad_norm": 0.31750553846359253, "learning_rate": 6.436780602975267e-05, "loss": 0.0275, "step": 34950 }, { "grad_norm": 0.2937723398208618, "learning_rate": 6.43480081431812e-05, "loss": 0.0251, "step": 34960 }, { "grad_norm": 0.2980746030807495, "learning_rate": 6.432820780492756e-05, "loss": 0.0267, "step": 34970 }, { "grad_norm": 0.39083489775657654, "learning_rate": 6.430840501837506e-05, "loss": 0.0284, "step": 34980 }, { "grad_norm": 0.2382989227771759, "learning_rate": 6.428859978690748e-05, "loss": 0.0281, "step": 34990 }, { "grad_norm": 0.5111172795295715, "learning_rate": 6.426879211390901e-05, "loss": 0.0249, "step": 35000 }, { "grad_norm": 0.41490012407302856, "learning_rate": 6.424898200276422e-05, "loss": 0.0316, "step": 35010 }, { "grad_norm": 0.31684616208076477, "learning_rate": 6.42291694568581e-05, "loss": 0.0245, "step": 35020 }, { "grad_norm": 0.36184361577033997, "learning_rate": 6.42093544795761e-05, "loss": 0.0354, "step": 35030 }, { "grad_norm": 0.46637439727783203, "learning_rate": 6.418953707430403e-05, "loss": 0.0271, "step": 35040 }, { "grad_norm": 0.3968605101108551, "learning_rate": 6.416971724442819e-05, "loss": 0.0221, "step": 35050 }, { "grad_norm": 0.27130433917045593, "learning_rate": 6.414989499333519e-05, "loss": 0.0186, "step": 35060 }, { "grad_norm": 0.37988656759262085, "learning_rate": 6.413007032441214e-05, "loss": 0.0238, "step": 35070 }, { "grad_norm": 0.4838220179080963, "learning_rate": 6.411024324104653e-05, "loss": 0.0295, "step": 35080 }, { "grad_norm": 0.27047133445739746, "learning_rate": 6.409041374662628e-05, "loss": 0.0222, "step": 35090 }, { "grad_norm": 0.31009554862976074, "learning_rate": 6.407058184453967e-05, "loss": 0.0354, "step": 35100 }, { "grad_norm": 0.3904191553592682, "learning_rate": 6.405074753817548e-05, "loss": 0.024, "step": 35110 }, { "grad_norm": 0.33879339694976807, "learning_rate": 6.40309108309228e-05, "loss": 0.03, "step": 35120 }, { "grad_norm": 0.2954208254814148, "learning_rate": 6.401107172617122e-05, "loss": 0.025, "step": 35130 }, { "grad_norm": 0.5227526426315308, "learning_rate": 6.399123022731068e-05, "loss": 0.0432, "step": 35140 }, { "grad_norm": 0.45350247621536255, "learning_rate": 6.397138633773157e-05, "loss": 0.0258, "step": 35150 }, { "grad_norm": 0.37861597537994385, "learning_rate": 6.395154006082463e-05, "loss": 0.0215, "step": 35160 }, { "grad_norm": 0.3270881175994873, "learning_rate": 6.393169139998109e-05, "loss": 0.0253, "step": 35170 }, { "grad_norm": 0.4059363007545471, "learning_rate": 6.39118403585925e-05, "loss": 0.0239, "step": 35180 }, { "grad_norm": 0.2883397936820984, "learning_rate": 6.38919869400509e-05, "loss": 0.0238, "step": 35190 }, { "grad_norm": 0.39695024490356445, "learning_rate": 6.387213114774865e-05, "loss": 0.0258, "step": 35200 }, { "grad_norm": 0.2710381746292114, "learning_rate": 6.385227298507863e-05, "loss": 0.0464, "step": 35210 }, { "grad_norm": 0.3808046281337738, "learning_rate": 6.3832412455434e-05, "loss": 0.0209, "step": 35220 }, { "grad_norm": 0.2332696169614792, "learning_rate": 6.381254956220841e-05, "loss": 0.0237, "step": 35230 }, { "grad_norm": 0.29254150390625, "learning_rate": 6.379268430879586e-05, "loss": 0.0274, "step": 35240 }, { "grad_norm": 0.3450147211551666, "learning_rate": 6.37728166985908e-05, "loss": 0.022, "step": 35250 }, { "grad_norm": 0.49485495686531067, "learning_rate": 6.375294673498804e-05, "loss": 0.021, "step": 35260 }, { "grad_norm": 0.23329930007457733, "learning_rate": 6.373307442138284e-05, "loss": 0.0254, "step": 35270 }, { "grad_norm": 0.38460400700569153, "learning_rate": 6.371319976117081e-05, "loss": 0.0348, "step": 35280 }, { "grad_norm": 0.22300606966018677, "learning_rate": 6.3693322757748e-05, "loss": 0.0292, "step": 35290 }, { "grad_norm": 0.4274935722351074, "learning_rate": 6.367344341451086e-05, "loss": 0.0279, "step": 35300 }, { "grad_norm": 0.41116634011268616, "learning_rate": 6.36535617348562e-05, "loss": 0.0275, "step": 35310 }, { "grad_norm": 0.3221744894981384, "learning_rate": 6.363367772218128e-05, "loss": 0.029, "step": 35320 }, { "grad_norm": 0.4421304762363434, "learning_rate": 6.36137913798837e-05, "loss": 0.0297, "step": 35330 }, { "grad_norm": 0.2670208513736725, "learning_rate": 6.359390271136151e-05, "loss": 0.0216, "step": 35340 }, { "grad_norm": 0.29599428176879883, "learning_rate": 6.357401172001314e-05, "loss": 0.0265, "step": 35350 }, { "grad_norm": 0.28809574246406555, "learning_rate": 6.355411840923742e-05, "loss": 0.0297, "step": 35360 }, { "grad_norm": 0.3492976725101471, "learning_rate": 6.353422278243358e-05, "loss": 0.0253, "step": 35370 }, { "grad_norm": 0.4103521406650543, "learning_rate": 6.351432484300121e-05, "loss": 0.0313, "step": 35380 }, { "grad_norm": 0.38329312205314636, "learning_rate": 6.349442459434036e-05, "loss": 0.0304, "step": 35390 }, { "grad_norm": 0.46062153577804565, "learning_rate": 6.34745220398514e-05, "loss": 0.03, "step": 35400 }, { "grad_norm": 0.3821355700492859, "learning_rate": 6.345461718293518e-05, "loss": 0.0229, "step": 35410 }, { "grad_norm": 0.3424690365791321, "learning_rate": 6.343471002699286e-05, "loss": 0.0321, "step": 35420 }, { "grad_norm": 0.369210422039032, "learning_rate": 6.341480057542602e-05, "loss": 0.0319, "step": 35430 }, { "grad_norm": 0.2858416736125946, "learning_rate": 6.339488883163667e-05, "loss": 0.0234, "step": 35440 }, { "grad_norm": 0.3136653006076813, "learning_rate": 6.337497479902716e-05, "loss": 0.0224, "step": 35450 }, { "grad_norm": 0.29518523812294006, "learning_rate": 6.335505848100027e-05, "loss": 0.0253, "step": 35460 }, { "grad_norm": 0.23268234729766846, "learning_rate": 6.333513988095915e-05, "loss": 0.0219, "step": 35470 }, { "grad_norm": 0.527717649936676, "learning_rate": 6.331521900230735e-05, "loss": 0.0238, "step": 35480 }, { "grad_norm": 0.5188479423522949, "learning_rate": 6.329529584844878e-05, "loss": 0.0264, "step": 35490 }, { "grad_norm": 0.3942407965660095, "learning_rate": 6.327537042278777e-05, "loss": 0.0401, "step": 35500 }, { "grad_norm": 0.28213512897491455, "learning_rate": 6.325544272872905e-05, "loss": 0.0258, "step": 35510 }, { "grad_norm": 1.060743808746338, "learning_rate": 6.323551276967771e-05, "loss": 0.0319, "step": 35520 }, { "grad_norm": 0.3571740388870239, "learning_rate": 6.321558054903922e-05, "loss": 0.0253, "step": 35530 }, { "grad_norm": 0.20174288749694824, "learning_rate": 6.319564607021947e-05, "loss": 0.0214, "step": 35540 }, { "grad_norm": 0.49147361516952515, "learning_rate": 6.31757093366247e-05, "loss": 0.0257, "step": 35550 }, { "grad_norm": 0.34932565689086914, "learning_rate": 6.315577035166154e-05, "loss": 0.0246, "step": 35560 }, { "grad_norm": 0.28358566761016846, "learning_rate": 6.313582911873708e-05, "loss": 0.0229, "step": 35570 }, { "grad_norm": 0.25263452529907227, "learning_rate": 6.311588564125865e-05, "loss": 0.0205, "step": 35580 }, { "grad_norm": 0.9119868278503418, "learning_rate": 6.30959399226341e-05, "loss": 0.0298, "step": 35590 }, { "grad_norm": 0.33413460850715637, "learning_rate": 6.30759919662716e-05, "loss": 0.0348, "step": 35600 }, { "grad_norm": 0.2949215769767761, "learning_rate": 6.30560417755797e-05, "loss": 0.0331, "step": 35610 }, { "grad_norm": 0.33239737153053284, "learning_rate": 6.303608935396735e-05, "loss": 0.0235, "step": 35620 }, { "grad_norm": 0.23737157881259918, "learning_rate": 6.301613470484386e-05, "loss": 0.0209, "step": 35630 }, { "grad_norm": 0.3740137815475464, "learning_rate": 6.299617783161893e-05, "loss": 0.0265, "step": 35640 }, { "grad_norm": 0.22532212734222412, "learning_rate": 6.297621873770266e-05, "loss": 0.0197, "step": 35650 }, { "grad_norm": 0.24886882305145264, "learning_rate": 6.29562574265055e-05, "loss": 0.0251, "step": 35660 }, { "grad_norm": 0.2738933265209198, "learning_rate": 6.293629390143834e-05, "loss": 0.0244, "step": 35670 }, { "grad_norm": 0.23695683479309082, "learning_rate": 6.291632816591232e-05, "loss": 0.0179, "step": 35680 }, { "grad_norm": 0.4335170090198517, "learning_rate": 6.28963602233391e-05, "loss": 0.024, "step": 35690 }, { "grad_norm": 0.3833775520324707, "learning_rate": 6.287639007713062e-05, "loss": 0.0285, "step": 35700 }, { "grad_norm": 0.2156086415052414, "learning_rate": 6.285641773069926e-05, "loss": 0.0244, "step": 35710 }, { "grad_norm": 0.26265791058540344, "learning_rate": 6.283644318745773e-05, "loss": 0.0256, "step": 35720 }, { "grad_norm": 0.28267917037010193, "learning_rate": 6.281646645081912e-05, "loss": 0.0358, "step": 35730 }, { "grad_norm": 0.4095908999443054, "learning_rate": 6.279648752419693e-05, "loss": 0.0235, "step": 35740 }, { "grad_norm": 0.3075137138366699, "learning_rate": 6.2776506411005e-05, "loss": 0.0278, "step": 35750 }, { "grad_norm": 0.3133309781551361, "learning_rate": 6.275652311465758e-05, "loss": 0.0291, "step": 35760 }, { "grad_norm": 0.26292967796325684, "learning_rate": 6.273653763856926e-05, "loss": 0.0248, "step": 35770 }, { "grad_norm": 0.32030558586120605, "learning_rate": 6.271654998615501e-05, "loss": 0.0247, "step": 35780 }, { "grad_norm": 0.3003963530063629, "learning_rate": 6.269656016083013e-05, "loss": 0.0345, "step": 35790 }, { "grad_norm": 0.24948936700820923, "learning_rate": 6.267656816601038e-05, "loss": 0.0311, "step": 35800 }, { "grad_norm": 0.24686381220817566, "learning_rate": 6.265657400511185e-05, "loss": 0.029, "step": 35810 }, { "grad_norm": 0.3513827919960022, "learning_rate": 6.263657768155098e-05, "loss": 0.0282, "step": 35820 }, { "grad_norm": 0.2722513973712921, "learning_rate": 6.261657919874457e-05, "loss": 0.0227, "step": 35830 }, { "grad_norm": 0.33732032775878906, "learning_rate": 6.259657856010986e-05, "loss": 0.0341, "step": 35840 }, { "grad_norm": 0.48739326000213623, "learning_rate": 6.257657576906439e-05, "loss": 0.0289, "step": 35850 }, { "grad_norm": 0.3974800109863281, "learning_rate": 6.255657082902609e-05, "loss": 0.0252, "step": 35860 }, { "grad_norm": 0.26009106636047363, "learning_rate": 6.253656374341325e-05, "loss": 0.0223, "step": 35870 }, { "grad_norm": 0.31855058670043945, "learning_rate": 6.251655451564457e-05, "loss": 0.0248, "step": 35880 }, { "grad_norm": 0.38036292791366577, "learning_rate": 6.249654314913902e-05, "loss": 0.0235, "step": 35890 }, { "grad_norm": 0.25911685824394226, "learning_rate": 6.247652964731604e-05, "loss": 0.0253, "step": 35900 }, { "grad_norm": 0.3221893608570099, "learning_rate": 6.245651401359537e-05, "loss": 0.0385, "step": 35910 }, { "grad_norm": 0.29516732692718506, "learning_rate": 6.243649625139715e-05, "loss": 0.0273, "step": 35920 }, { "grad_norm": 0.42478999495506287, "learning_rate": 6.241647636414185e-05, "loss": 0.0286, "step": 35930 }, { "grad_norm": 0.2981037497520447, "learning_rate": 6.239645435525034e-05, "loss": 0.0344, "step": 35940 }, { "grad_norm": 0.3351275324821472, "learning_rate": 6.237643022814381e-05, "loss": 0.0267, "step": 35950 }, { "grad_norm": 0.2946583032608032, "learning_rate": 6.235640398624386e-05, "loss": 0.0257, "step": 35960 }, { "grad_norm": 0.48868152499198914, "learning_rate": 6.233637563297243e-05, "loss": 0.0318, "step": 35970 }, { "grad_norm": 0.3862161636352539, "learning_rate": 6.23163451717518e-05, "loss": 0.0288, "step": 35980 }, { "grad_norm": 0.41181573271751404, "learning_rate": 6.229631260600463e-05, "loss": 0.0249, "step": 35990 }, { "grad_norm": 0.2318616807460785, "learning_rate": 6.227627793915392e-05, "loss": 0.0351, "step": 36000 }, { "grad_norm": 0.27489134669303894, "learning_rate": 6.225624117462309e-05, "loss": 0.0252, "step": 36010 }, { "grad_norm": 0.31536009907722473, "learning_rate": 6.223620231583586e-05, "loss": 0.0283, "step": 36020 }, { "grad_norm": 0.33439478278160095, "learning_rate": 6.221616136621629e-05, "loss": 0.0213, "step": 36030 }, { "grad_norm": 0.2851017415523529, "learning_rate": 6.219611832918887e-05, "loss": 0.0279, "step": 36040 }, { "grad_norm": 0.38709884881973267, "learning_rate": 6.217607320817838e-05, "loss": 0.0255, "step": 36050 }, { "grad_norm": 0.40997305512428284, "learning_rate": 6.215602600661001e-05, "loss": 0.0245, "step": 36060 }, { "grad_norm": 0.31331878900527954, "learning_rate": 6.213597672790925e-05, "loss": 0.0226, "step": 36070 }, { "grad_norm": 0.28758811950683594, "learning_rate": 6.2115925375502e-05, "loss": 0.0214, "step": 36080 }, { "grad_norm": 0.3812987208366394, "learning_rate": 6.209587195281447e-05, "loss": 0.0214, "step": 36090 }, { "grad_norm": 0.3673131465911865, "learning_rate": 6.207581646327324e-05, "loss": 0.0297, "step": 36100 }, { "grad_norm": 0.38652992248535156, "learning_rate": 6.205575891030526e-05, "loss": 0.0209, "step": 36110 }, { "grad_norm": 0.2696804404258728, "learning_rate": 6.203569929733781e-05, "loss": 0.023, "step": 36120 }, { "grad_norm": 0.3617110550403595, "learning_rate": 6.201563762779852e-05, "loss": 0.0231, "step": 36130 }, { "grad_norm": 0.488235741853714, "learning_rate": 6.199557390511538e-05, "loss": 0.0237, "step": 36140 }, { "grad_norm": 0.33254900574684143, "learning_rate": 6.197550813271675e-05, "loss": 0.0259, "step": 36150 }, { "grad_norm": 0.26749956607818604, "learning_rate": 6.195544031403131e-05, "loss": 0.023, "step": 36160 }, { "grad_norm": 0.4466310143470764, "learning_rate": 6.19353704524881e-05, "loss": 0.0272, "step": 36170 }, { "grad_norm": 0.29893192648887634, "learning_rate": 6.191529855151652e-05, "loss": 0.0217, "step": 36180 }, { "grad_norm": 0.2679327130317688, "learning_rate": 6.189522461454629e-05, "loss": 0.0251, "step": 36190 }, { "grad_norm": 0.2784460484981537, "learning_rate": 6.187514864500752e-05, "loss": 0.0277, "step": 36200 }, { "grad_norm": 0.4042452275753021, "learning_rate": 6.185507064633062e-05, "loss": 0.0249, "step": 36210 }, { "grad_norm": 0.3434291481971741, "learning_rate": 6.18349906219464e-05, "loss": 0.0298, "step": 36220 }, { "grad_norm": 0.3493233621120453, "learning_rate": 6.181490857528596e-05, "loss": 0.0214, "step": 36230 }, { "grad_norm": 0.26932212710380554, "learning_rate": 6.179482450978077e-05, "loss": 0.0174, "step": 36240 }, { "grad_norm": 0.245059996843338, "learning_rate": 6.177473842886269e-05, "loss": 0.0206, "step": 36250 }, { "grad_norm": 0.43860283493995667, "learning_rate": 6.175465033596382e-05, "loss": 0.0292, "step": 36260 }, { "grad_norm": 0.22467133402824402, "learning_rate": 6.173456023451671e-05, "loss": 0.0224, "step": 36270 }, { "grad_norm": 0.29518207907676697, "learning_rate": 6.171446812795422e-05, "loss": 0.0269, "step": 36280 }, { "grad_norm": 0.30844953656196594, "learning_rate": 6.169437401970949e-05, "loss": 0.0313, "step": 36290 }, { "grad_norm": 0.382456511259079, "learning_rate": 6.16742779132161e-05, "loss": 0.0216, "step": 36300 }, { "grad_norm": 0.3403448760509491, "learning_rate": 6.165417981190789e-05, "loss": 0.0307, "step": 36310 }, { "grad_norm": 0.4834063649177551, "learning_rate": 6.16340797192191e-05, "loss": 0.0263, "step": 36320 }, { "grad_norm": 0.33274129033088684, "learning_rate": 6.161397763858427e-05, "loss": 0.0333, "step": 36330 }, { "grad_norm": 0.5493006706237793, "learning_rate": 6.159387357343834e-05, "loss": 0.0222, "step": 36340 }, { "grad_norm": 0.258409708738327, "learning_rate": 6.157376752721648e-05, "loss": 0.03, "step": 36350 }, { "grad_norm": 0.5453574657440186, "learning_rate": 6.155365950335428e-05, "loss": 0.0266, "step": 36360 }, { "grad_norm": 0.6869057416915894, "learning_rate": 6.153354950528768e-05, "loss": 0.0313, "step": 36370 }, { "grad_norm": 0.33031976222991943, "learning_rate": 6.151343753645293e-05, "loss": 0.029, "step": 36380 }, { "grad_norm": 0.3002523183822632, "learning_rate": 6.149332360028657e-05, "loss": 0.0268, "step": 36390 }, { "grad_norm": 0.2706238627433777, "learning_rate": 6.147320770022555e-05, "loss": 0.0271, "step": 36400 }, { "grad_norm": 0.22902265191078186, "learning_rate": 6.145308983970715e-05, "loss": 0.0251, "step": 36410 }, { "grad_norm": 0.3979523181915283, "learning_rate": 6.143297002216892e-05, "loss": 0.0269, "step": 36420 }, { "grad_norm": 0.36858224868774414, "learning_rate": 6.141284825104882e-05, "loss": 0.0266, "step": 36430 }, { "grad_norm": 0.22424042224884033, "learning_rate": 6.13927245297851e-05, "loss": 0.0265, "step": 36440 }, { "grad_norm": 0.24466189742088318, "learning_rate": 6.137259886181633e-05, "loss": 0.0273, "step": 36450 }, { "grad_norm": 0.2902020215988159, "learning_rate": 6.135247125058145e-05, "loss": 0.0265, "step": 36460 }, { "grad_norm": 0.19723664224147797, "learning_rate": 6.133234169951974e-05, "loss": 0.0193, "step": 36470 }, { "grad_norm": 0.3111399710178375, "learning_rate": 6.131221021207078e-05, "loss": 0.0278, "step": 36480 }, { "grad_norm": 0.26143455505371094, "learning_rate": 6.129207679167448e-05, "loss": 0.0215, "step": 36490 }, { "grad_norm": 0.21625471115112305, "learning_rate": 6.127194144177109e-05, "loss": 0.02, "step": 36500 }, { "grad_norm": 0.2936969995498657, "learning_rate": 6.125180416580118e-05, "loss": 0.0214, "step": 36510 }, { "grad_norm": 0.24686160683631897, "learning_rate": 6.123166496720571e-05, "loss": 0.0245, "step": 36520 }, { "grad_norm": 0.4129692614078522, "learning_rate": 6.121152384942588e-05, "loss": 0.0225, "step": 36530 }, { "grad_norm": 0.3406106233596802, "learning_rate": 6.119138081590324e-05, "loss": 0.025, "step": 36540 }, { "grad_norm": 0.2554318904876709, "learning_rate": 6.117123587007971e-05, "loss": 0.0182, "step": 36550 }, { "grad_norm": 0.43489179015159607, "learning_rate": 6.11510890153975e-05, "loss": 0.0268, "step": 36560 }, { "grad_norm": 0.49450400471687317, "learning_rate": 6.113094025529916e-05, "loss": 0.021, "step": 36570 }, { "grad_norm": 0.21294362843036652, "learning_rate": 6.111078959322757e-05, "loss": 0.0347, "step": 36580 }, { "grad_norm": 0.32714036107063293, "learning_rate": 6.109063703262592e-05, "loss": 0.0356, "step": 36590 }, { "grad_norm": 0.3119068741798401, "learning_rate": 6.107048257693772e-05, "loss": 0.0247, "step": 36600 }, { "grad_norm": 0.3293708860874176, "learning_rate": 6.105032622960683e-05, "loss": 0.026, "step": 36610 }, { "grad_norm": 0.28119832277297974, "learning_rate": 6.103016799407743e-05, "loss": 0.0234, "step": 36620 }, { "grad_norm": 0.3450067937374115, "learning_rate": 6.1010007873793984e-05, "loss": 0.0323, "step": 36630 }, { "grad_norm": 0.20771096646785736, "learning_rate": 6.098984587220131e-05, "loss": 0.0233, "step": 36640 }, { "grad_norm": 0.8960378170013428, "learning_rate": 6.096968199274456e-05, "loss": 0.0307, "step": 36650 }, { "grad_norm": 0.32015225291252136, "learning_rate": 6.0949516238869166e-05, "loss": 0.0272, "step": 36660 }, { "grad_norm": 0.3308746814727783, "learning_rate": 6.092934861402092e-05, "loss": 0.035, "step": 36670 }, { "grad_norm": 0.4121895432472229, "learning_rate": 6.0909179121645924e-05, "loss": 0.0363, "step": 36680 }, { "grad_norm": 0.3118487000465393, "learning_rate": 6.0889007765190576e-05, "loss": 0.0343, "step": 36690 }, { "grad_norm": 0.19198444485664368, "learning_rate": 6.086883454810162e-05, "loss": 0.0256, "step": 36700 }, { "grad_norm": 0.2718144357204437, "learning_rate": 6.0848659473826084e-05, "loss": 0.0237, "step": 36710 }, { "grad_norm": 0.47432997822761536, "learning_rate": 6.082848254581138e-05, "loss": 0.0288, "step": 36720 }, { "grad_norm": 0.31313514709472656, "learning_rate": 6.080830376750517e-05, "loss": 0.0235, "step": 36730 }, { "grad_norm": 0.3216199576854706, "learning_rate": 6.0788123142355445e-05, "loss": 0.0222, "step": 36740 }, { "grad_norm": 0.3861330449581146, "learning_rate": 6.076794067381052e-05, "loss": 0.0282, "step": 36750 }, { "grad_norm": 0.38688042759895325, "learning_rate": 6.074775636531905e-05, "loss": 0.0277, "step": 36760 }, { "grad_norm": 0.4261777698993683, "learning_rate": 6.072757022032997e-05, "loss": 0.0305, "step": 36770 }, { "grad_norm": 0.2699369192123413, "learning_rate": 6.070738224229253e-05, "loss": 0.0276, "step": 36780 }, { "grad_norm": 0.34187138080596924, "learning_rate": 6.0687192434656314e-05, "loss": 0.0302, "step": 36790 }, { "grad_norm": 0.24448458850383759, "learning_rate": 6.066700080087121e-05, "loss": 0.0246, "step": 36800 }, { "grad_norm": 0.2641223669052124, "learning_rate": 6.0646807344387424e-05, "loss": 0.0234, "step": 36810 }, { "grad_norm": 0.383884996175766, "learning_rate": 6.062661206865543e-05, "loss": 0.0292, "step": 36820 }, { "grad_norm": 0.3070410490036011, "learning_rate": 6.06064149771261e-05, "loss": 0.0266, "step": 36830 }, { "grad_norm": 0.3197992444038391, "learning_rate": 6.058621607325051e-05, "loss": 0.0241, "step": 36840 }, { "grad_norm": 0.37305107712745667, "learning_rate": 6.056601536048014e-05, "loss": 0.0268, "step": 36850 }, { "grad_norm": 0.35067975521087646, "learning_rate": 6.0545812842266725e-05, "loss": 0.0341, "step": 36860 }, { "grad_norm": 0.2545805275440216, "learning_rate": 6.052560852206232e-05, "loss": 0.0253, "step": 36870 }, { "grad_norm": 0.2752642333507538, "learning_rate": 6.05054024033193e-05, "loss": 0.0386, "step": 36880 }, { "grad_norm": 0.3524945080280304, "learning_rate": 6.048519448949032e-05, "loss": 0.0254, "step": 36890 }, { "grad_norm": 0.22571447491645813, "learning_rate": 6.046498478402839e-05, "loss": 0.0241, "step": 36900 }, { "grad_norm": 0.20550918579101562, "learning_rate": 6.044477329038677e-05, "loss": 0.0224, "step": 36910 }, { "grad_norm": 0.38513845205307007, "learning_rate": 6.042456001201906e-05, "loss": 0.0238, "step": 36920 }, { "grad_norm": 0.5369551777839661, "learning_rate": 6.040434495237917e-05, "loss": 0.0237, "step": 36930 }, { "grad_norm": 0.37327471375465393, "learning_rate": 6.0384128114921256e-05, "loss": 0.0241, "step": 36940 }, { "grad_norm": 0.30632802844047546, "learning_rate": 6.036390950309987e-05, "loss": 0.0236, "step": 36950 }, { "grad_norm": 0.3406730890274048, "learning_rate": 6.0343689120369805e-05, "loss": 0.0299, "step": 36960 }, { "grad_norm": 0.35226544737815857, "learning_rate": 6.032346697018616e-05, "loss": 0.0202, "step": 36970 }, { "grad_norm": 0.3641219437122345, "learning_rate": 6.0303243056004375e-05, "loss": 0.041, "step": 36980 }, { "grad_norm": 0.25829845666885376, "learning_rate": 6.0283017381280136e-05, "loss": 0.0232, "step": 36990 }, { "grad_norm": 0.37243521213531494, "learning_rate": 6.026278994946948e-05, "loss": 0.0217, "step": 37000 }, { "grad_norm": 0.41546356678009033, "learning_rate": 6.02425607640287e-05, "loss": 0.0275, "step": 37010 }, { "grad_norm": 0.34786251187324524, "learning_rate": 6.022232982841441e-05, "loss": 0.0274, "step": 37020 }, { "grad_norm": 0.2765651047229767, "learning_rate": 6.020209714608355e-05, "loss": 0.0253, "step": 37030 }, { "grad_norm": 0.4348048269748688, "learning_rate": 6.018186272049331e-05, "loss": 0.0296, "step": 37040 }, { "grad_norm": 0.24643981456756592, "learning_rate": 6.01616265551012e-05, "loss": 0.0238, "step": 37050 }, { "grad_norm": 0.2317725569009781, "learning_rate": 6.014138865336503e-05, "loss": 0.0228, "step": 37060 }, { "grad_norm": 0.2836092710494995, "learning_rate": 6.0121149018742905e-05, "loss": 0.0258, "step": 37070 }, { "grad_norm": 0.3880479335784912, "learning_rate": 6.010090765469325e-05, "loss": 0.0228, "step": 37080 }, { "grad_norm": 0.2868867814540863, "learning_rate": 6.008066456467473e-05, "loss": 0.0221, "step": 37090 }, { "grad_norm": 0.2532919645309448, "learning_rate": 6.0060419752146335e-05, "loss": 0.0258, "step": 37100 }, { "grad_norm": 0.4596029818058014, "learning_rate": 6.0040173220567353e-05, "loss": 0.0228, "step": 37110 }, { "grad_norm": 0.20067088305950165, "learning_rate": 6.001992497339737e-05, "loss": 0.0185, "step": 37120 }, { "grad_norm": 0.26618316769599915, "learning_rate": 5.999967501409626e-05, "loss": 0.0317, "step": 37130 }, { "grad_norm": 0.20871615409851074, "learning_rate": 5.997942334612418e-05, "loss": 0.0233, "step": 37140 }, { "grad_norm": 0.2562791705131531, "learning_rate": 5.995916997294158e-05, "loss": 0.017, "step": 37150 }, { "grad_norm": 0.2989925742149353, "learning_rate": 5.9938914898009214e-05, "loss": 0.0304, "step": 37160 }, { "grad_norm": 0.255551278591156, "learning_rate": 5.991865812478813e-05, "loss": 0.0198, "step": 37170 }, { "grad_norm": 0.2830502986907959, "learning_rate": 5.989839965673964e-05, "loss": 0.0348, "step": 37180 }, { "grad_norm": 0.27643999457359314, "learning_rate": 5.987813949732539e-05, "loss": 0.0242, "step": 37190 }, { "grad_norm": 0.2469000220298767, "learning_rate": 5.9857877650007255e-05, "loss": 0.024, "step": 37200 }, { "grad_norm": 0.1966453641653061, "learning_rate": 5.983761411824744e-05, "loss": 0.0248, "step": 37210 }, { "grad_norm": 0.23083938658237457, "learning_rate": 5.981734890550844e-05, "loss": 0.0256, "step": 37220 }, { "grad_norm": 0.28512996435165405, "learning_rate": 5.979708201525301e-05, "loss": 0.031, "step": 37230 }, { "grad_norm": 0.34871527552604675, "learning_rate": 5.977681345094422e-05, "loss": 0.0271, "step": 37240 }, { "grad_norm": 0.36937952041625977, "learning_rate": 5.97565432160454e-05, "loss": 0.0254, "step": 37250 }, { "grad_norm": 0.3084045946598053, "learning_rate": 5.9736271314020186e-05, "loss": 0.0285, "step": 37260 }, { "grad_norm": 0.3800307512283325, "learning_rate": 5.971599774833251e-05, "loss": 0.0281, "step": 37270 }, { "grad_norm": 0.37255364656448364, "learning_rate": 5.9695722522446525e-05, "loss": 0.0276, "step": 37280 }, { "grad_norm": 0.3363049626350403, "learning_rate": 5.9675445639826765e-05, "loss": 0.0282, "step": 37290 }, { "grad_norm": 0.2708231508731842, "learning_rate": 5.965516710393796e-05, "loss": 0.0268, "step": 37300 }, { "grad_norm": 0.3107137978076935, "learning_rate": 5.963488691824516e-05, "loss": 0.0295, "step": 37310 }, { "grad_norm": 0.3486868143081665, "learning_rate": 5.96146050862137e-05, "loss": 0.0335, "step": 37320 }, { "grad_norm": 0.31391727924346924, "learning_rate": 5.959432161130919e-05, "loss": 0.0225, "step": 37330 }, { "grad_norm": 0.22477860748767853, "learning_rate": 5.9574036496997545e-05, "loss": 0.0276, "step": 37340 }, { "grad_norm": 0.24778658151626587, "learning_rate": 5.955374974674488e-05, "loss": 0.024, "step": 37350 }, { "grad_norm": 0.20284998416900635, "learning_rate": 5.9533461364017696e-05, "loss": 0.0235, "step": 37360 }, { "grad_norm": 0.23811906576156616, "learning_rate": 5.9513171352282716e-05, "loss": 0.0223, "step": 37370 }, { "grad_norm": 0.27504780888557434, "learning_rate": 5.949287971500692e-05, "loss": 0.025, "step": 37380 }, { "grad_norm": 0.24585570394992828, "learning_rate": 5.947258645565762e-05, "loss": 0.021, "step": 37390 }, { "grad_norm": 0.24856410920619965, "learning_rate": 5.945229157770237e-05, "loss": 0.0209, "step": 37400 }, { "grad_norm": 0.29931744933128357, "learning_rate": 5.9431995084609006e-05, "loss": 0.0254, "step": 37410 }, { "grad_norm": 0.3014700412750244, "learning_rate": 5.941169697984564e-05, "loss": 0.0271, "step": 37420 }, { "grad_norm": 0.37888962030410767, "learning_rate": 5.9391397266880675e-05, "loss": 0.0266, "step": 37430 }, { "grad_norm": 0.3892439603805542, "learning_rate": 5.937109594918279e-05, "loss": 0.0415, "step": 37440 }, { "grad_norm": 0.30304649472236633, "learning_rate": 5.9350793030220884e-05, "loss": 0.0428, "step": 37450 }, { "grad_norm": 0.4278292953968048, "learning_rate": 5.933048851346421e-05, "loss": 0.0247, "step": 37460 }, { "grad_norm": 0.43000173568725586, "learning_rate": 5.931018240238224e-05, "loss": 0.0356, "step": 37470 }, { "grad_norm": 0.2369011640548706, "learning_rate": 5.928987470044471e-05, "loss": 0.0265, "step": 37480 }, { "grad_norm": 0.3696296215057373, "learning_rate": 5.9269565411121695e-05, "loss": 0.0247, "step": 37490 }, { "grad_norm": 0.22555439174175262, "learning_rate": 5.924925453788347e-05, "loss": 0.0304, "step": 37500 }, { "grad_norm": 0.3117510974407196, "learning_rate": 5.92289420842006e-05, "loss": 0.0267, "step": 37510 }, { "grad_norm": 0.3057446777820587, "learning_rate": 5.9208628053543945e-05, "loss": 0.0309, "step": 37520 }, { "grad_norm": 0.3846855163574219, "learning_rate": 5.918831244938462e-05, "loss": 0.0274, "step": 37530 }, { "grad_norm": 0.2985493838787079, "learning_rate": 5.9167995275194e-05, "loss": 0.0246, "step": 37540 }, { "grad_norm": 0.4688616096973419, "learning_rate": 5.914767653444373e-05, "loss": 0.0295, "step": 37550 }, { "grad_norm": 0.4430249333381653, "learning_rate": 5.912735623060572e-05, "loss": 0.0412, "step": 37560 }, { "grad_norm": 0.325592041015625, "learning_rate": 5.910703436715217e-05, "loss": 0.028, "step": 37570 }, { "grad_norm": 0.2771609127521515, "learning_rate": 5.908671094755552e-05, "loss": 0.0236, "step": 37580 }, { "grad_norm": 0.35249263048171997, "learning_rate": 5.906638597528851e-05, "loss": 0.0241, "step": 37590 }, { "grad_norm": 0.2559843361377716, "learning_rate": 5.9046059453824076e-05, "loss": 0.021, "step": 37600 }, { "grad_norm": 0.33167365193367004, "learning_rate": 5.9025731386635505e-05, "loss": 0.0224, "step": 37610 }, { "grad_norm": 0.23998446762561798, "learning_rate": 5.900540177719629e-05, "loss": 0.0214, "step": 37620 }, { "grad_norm": 0.3325767517089844, "learning_rate": 5.898507062898021e-05, "loss": 0.0268, "step": 37630 }, { "grad_norm": 0.36357825994491577, "learning_rate": 5.8964737945461316e-05, "loss": 0.0214, "step": 37640 }, { "grad_norm": 0.2513343393802643, "learning_rate": 5.8944403730113885e-05, "loss": 0.026, "step": 37650 }, { "grad_norm": 0.3329830467700958, "learning_rate": 5.892406798641248e-05, "loss": 0.0236, "step": 37660 }, { "grad_norm": 0.424220472574234, "learning_rate": 5.890373071783193e-05, "loss": 0.0219, "step": 37670 }, { "grad_norm": 0.539763867855072, "learning_rate": 5.888339192784732e-05, "loss": 0.0313, "step": 37680 }, { "grad_norm": 0.31410136818885803, "learning_rate": 5.8863051619934003e-05, "loss": 0.0251, "step": 37690 }, { "grad_norm": 0.3151073157787323, "learning_rate": 5.8842709797567554e-05, "loss": 0.0258, "step": 37700 }, { "grad_norm": 0.40877968072891235, "learning_rate": 5.8822366464223855e-05, "loss": 0.0291, "step": 37710 }, { "grad_norm": 0.22933627665042877, "learning_rate": 5.880202162337901e-05, "loss": 0.0279, "step": 37720 }, { "grad_norm": 0.9355024695396423, "learning_rate": 5.8781675278509405e-05, "loss": 0.0274, "step": 37730 }, { "grad_norm": 0.2245064377784729, "learning_rate": 5.8761327433091696e-05, "loss": 0.0213, "step": 37740 }, { "grad_norm": 0.4517669975757599, "learning_rate": 5.874097809060275e-05, "loss": 0.029, "step": 37750 }, { "grad_norm": 0.26859065890312195, "learning_rate": 5.87206272545197e-05, "loss": 0.0401, "step": 37760 }, { "grad_norm": 0.3963980972766876, "learning_rate": 5.8700274928319955e-05, "loss": 0.0292, "step": 37770 }, { "grad_norm": 0.3378506302833557, "learning_rate": 5.867992111548118e-05, "loss": 0.0263, "step": 37780 }, { "grad_norm": 0.25307363271713257, "learning_rate": 5.865956581948131e-05, "loss": 0.0222, "step": 37790 }, { "grad_norm": 0.23141847550868988, "learning_rate": 5.863920904379845e-05, "loss": 0.0207, "step": 37800 }, { "grad_norm": 0.33719712495803833, "learning_rate": 5.8618850791911064e-05, "loss": 0.0332, "step": 37810 }, { "grad_norm": 0.3028010427951813, "learning_rate": 5.859849106729779e-05, "loss": 0.0264, "step": 37820 }, { "grad_norm": 0.33983930945396423, "learning_rate": 5.857812987343758e-05, "loss": 0.0219, "step": 37830 }, { "grad_norm": 0.5331070423126221, "learning_rate": 5.855776721380957e-05, "loss": 0.0285, "step": 37840 }, { "grad_norm": 0.30878737568855286, "learning_rate": 5.8537403091893217e-05, "loss": 0.0252, "step": 37850 }, { "grad_norm": 0.3059505820274353, "learning_rate": 5.851703751116816e-05, "loss": 0.0261, "step": 37860 }, { "grad_norm": 0.26032790541648865, "learning_rate": 5.8496670475114336e-05, "loss": 0.0235, "step": 37870 }, { "grad_norm": 0.3965316414833069, "learning_rate": 5.84763019872119e-05, "loss": 0.0372, "step": 37880 }, { "grad_norm": 0.34129655361175537, "learning_rate": 5.845593205094131e-05, "loss": 0.0247, "step": 37890 }, { "grad_norm": 0.21657142043113708, "learning_rate": 5.843556066978318e-05, "loss": 0.0206, "step": 37900 }, { "grad_norm": 0.4020937979221344, "learning_rate": 5.8415187847218455e-05, "loss": 0.0271, "step": 37910 }, { "grad_norm": 0.36619552969932556, "learning_rate": 5.839481358672827e-05, "loss": 0.0236, "step": 37920 }, { "grad_norm": 0.29088541865348816, "learning_rate": 5.837443789179407e-05, "loss": 0.0254, "step": 37930 }, { "grad_norm": 0.5619405508041382, "learning_rate": 5.8354060765897445e-05, "loss": 0.0263, "step": 37940 }, { "grad_norm": 0.43682825565338135, "learning_rate": 5.8333682212520334e-05, "loss": 0.0419, "step": 37950 }, { "grad_norm": 0.4383169710636139, "learning_rate": 5.831330223514486e-05, "loss": 0.0249, "step": 37960 }, { "grad_norm": 0.2408452033996582, "learning_rate": 5.8292920837253396e-05, "loss": 0.0226, "step": 37970 }, { "grad_norm": 0.3368094861507416, "learning_rate": 5.827253802232857e-05, "loss": 0.0213, "step": 37980 }, { "grad_norm": 0.46248093247413635, "learning_rate": 5.825215379385327e-05, "loss": 0.0215, "step": 37990 }, { "grad_norm": 0.32863232493400574, "learning_rate": 5.823176815531057e-05, "loss": 0.0348, "step": 38000 }, { "grad_norm": 0.26447927951812744, "learning_rate": 5.8211381110183826e-05, "loss": 0.0351, "step": 38010 }, { "grad_norm": 0.2604312300682068, "learning_rate": 5.8190992661956645e-05, "loss": 0.0221, "step": 38020 }, { "grad_norm": 0.3029189705848694, "learning_rate": 5.817060281411284e-05, "loss": 0.0263, "step": 38030 }, { "grad_norm": 0.29014861583709717, "learning_rate": 5.815021157013647e-05, "loss": 0.0222, "step": 38040 }, { "grad_norm": 0.3257029354572296, "learning_rate": 5.8129818933511856e-05, "loss": 0.0374, "step": 38050 }, { "grad_norm": 0.8912161588668823, "learning_rate": 5.8109424907723544e-05, "loss": 0.0284, "step": 38060 }, { "grad_norm": 0.19652745127677917, "learning_rate": 5.80890294962563e-05, "loss": 0.017, "step": 38070 }, { "grad_norm": 0.3370501399040222, "learning_rate": 5.806863270259515e-05, "loss": 0.0266, "step": 38080 }, { "grad_norm": 0.39116135239601135, "learning_rate": 5.804823453022536e-05, "loss": 0.0285, "step": 38090 }, { "grad_norm": 0.3013947606086731, "learning_rate": 5.80278349826324e-05, "loss": 0.0222, "step": 38100 }, { "grad_norm": 0.26450851559638977, "learning_rate": 5.8007434063302014e-05, "loss": 0.0244, "step": 38110 }, { "grad_norm": 0.30398431420326233, "learning_rate": 5.7987031775720136e-05, "loss": 0.0248, "step": 38120 }, { "grad_norm": 0.2044331133365631, "learning_rate": 5.7966628123372976e-05, "loss": 0.03, "step": 38130 }, { "grad_norm": 0.4451588988304138, "learning_rate": 5.7946223109746956e-05, "loss": 0.0327, "step": 38140 }, { "grad_norm": 0.32344505190849304, "learning_rate": 5.7925816738328754e-05, "loss": 0.0191, "step": 38150 }, { "grad_norm": 0.47824719548225403, "learning_rate": 5.790540901260521e-05, "loss": 0.0241, "step": 38160 }, { "grad_norm": 0.3865256905555725, "learning_rate": 5.788499993606351e-05, "loss": 0.021, "step": 38170 }, { "grad_norm": 0.4304932653903961, "learning_rate": 5.786458951219096e-05, "loss": 0.0284, "step": 38180 }, { "grad_norm": 0.2606970965862274, "learning_rate": 5.784417774447517e-05, "loss": 0.0286, "step": 38190 }, { "grad_norm": 0.6813376545906067, "learning_rate": 5.782376463640393e-05, "loss": 0.0292, "step": 38200 }, { "grad_norm": 0.31323444843292236, "learning_rate": 5.780335019146531e-05, "loss": 0.0213, "step": 38210 }, { "grad_norm": 0.37456104159355164, "learning_rate": 5.778293441314755e-05, "loss": 0.0262, "step": 38220 }, { "grad_norm": 0.33321887254714966, "learning_rate": 5.776251730493917e-05, "loss": 0.0338, "step": 38230 }, { "grad_norm": 0.32036158442497253, "learning_rate": 5.774209887032887e-05, "loss": 0.0241, "step": 38240 }, { "grad_norm": 0.3387269079685211, "learning_rate": 5.772167911280565e-05, "loss": 0.0267, "step": 38250 }, { "grad_norm": 0.43654903769493103, "learning_rate": 5.770125803585864e-05, "loss": 0.0244, "step": 38260 }, { "grad_norm": 0.29501762986183167, "learning_rate": 5.768083564297726e-05, "loss": 0.0313, "step": 38270 }, { "grad_norm": 0.3668825328350067, "learning_rate": 5.766041193765114e-05, "loss": 0.0275, "step": 38280 }, { "grad_norm": 0.2833770215511322, "learning_rate": 5.763998692337015e-05, "loss": 0.0194, "step": 38290 }, { "grad_norm": 0.34805071353912354, "learning_rate": 5.761956060362433e-05, "loss": 0.0279, "step": 38300 }, { "grad_norm": 0.22942762076854706, "learning_rate": 5.7599132981904e-05, "loss": 0.024, "step": 38310 }, { "grad_norm": 0.29079294204711914, "learning_rate": 5.75787040616997e-05, "loss": 0.033, "step": 38320 }, { "grad_norm": 0.3193078935146332, "learning_rate": 5.755827384650212e-05, "loss": 0.0222, "step": 38330 }, { "grad_norm": 0.2674114406108856, "learning_rate": 5.753784233980228e-05, "loss": 0.0205, "step": 38340 }, { "grad_norm": 0.3766392171382904, "learning_rate": 5.751740954509135e-05, "loss": 0.0237, "step": 38350 }, { "grad_norm": 0.29262587428092957, "learning_rate": 5.7496975465860715e-05, "loss": 0.0198, "step": 38360 }, { "grad_norm": 0.1972084492444992, "learning_rate": 5.747654010560202e-05, "loss": 0.0165, "step": 38370 }, { "grad_norm": 0.310259073972702, "learning_rate": 5.7456103467807097e-05, "loss": 0.0269, "step": 38380 }, { "grad_norm": 0.3796439468860626, "learning_rate": 5.7435665555968046e-05, "loss": 0.0214, "step": 38390 }, { "grad_norm": 0.21440505981445312, "learning_rate": 5.74152263735771e-05, "loss": 0.0261, "step": 38400 }, { "grad_norm": 0.37754616141319275, "learning_rate": 5.739478592412677e-05, "loss": 0.027, "step": 38410 }, { "grad_norm": 0.48194700479507446, "learning_rate": 5.7374344211109766e-05, "loss": 0.0282, "step": 38420 }, { "grad_norm": 0.212917298078537, "learning_rate": 5.735390123801904e-05, "loss": 0.0284, "step": 38430 }, { "grad_norm": 0.2919250428676605, "learning_rate": 5.7333457008347704e-05, "loss": 0.0335, "step": 38440 }, { "grad_norm": 0.36379536986351013, "learning_rate": 5.7313011525589156e-05, "loss": 0.0364, "step": 38450 }, { "grad_norm": 0.48653674125671387, "learning_rate": 5.729256479323694e-05, "loss": 0.031, "step": 38460 }, { "grad_norm": 0.4878826141357422, "learning_rate": 5.727211681478485e-05, "loss": 0.0281, "step": 38470 }, { "grad_norm": 0.2776247560977936, "learning_rate": 5.7251667593726886e-05, "loss": 0.0322, "step": 38480 }, { "grad_norm": 0.28243371844291687, "learning_rate": 5.723121713355728e-05, "loss": 0.0373, "step": 38490 }, { "grad_norm": 0.2535753846168518, "learning_rate": 5.721076543777044e-05, "loss": 0.0257, "step": 38500 }, { "grad_norm": 0.24903157353401184, "learning_rate": 5.7190312509860986e-05, "loss": 0.0226, "step": 38510 }, { "grad_norm": 0.9147739410400391, "learning_rate": 5.716985835332379e-05, "loss": 0.0259, "step": 38520 }, { "grad_norm": 0.25976940989494324, "learning_rate": 5.714940297165389e-05, "loss": 0.029, "step": 38530 }, { "grad_norm": 0.35810449719429016, "learning_rate": 5.712894636834656e-05, "loss": 0.0265, "step": 38540 }, { "grad_norm": 0.34683385491371155, "learning_rate": 5.7108488546897276e-05, "loss": 0.0183, "step": 38550 }, { "grad_norm": 0.2758166491985321, "learning_rate": 5.708802951080172e-05, "loss": 0.0185, "step": 38560 }, { "grad_norm": 0.33392828702926636, "learning_rate": 5.706756926355576e-05, "loss": 0.031, "step": 38570 }, { "grad_norm": 0.30399006605148315, "learning_rate": 5.704710780865554e-05, "loss": 0.0224, "step": 38580 }, { "grad_norm": 0.42748594284057617, "learning_rate": 5.7026645149597325e-05, "loss": 0.0259, "step": 38590 }, { "grad_norm": 0.3153951168060303, "learning_rate": 5.700618128987764e-05, "loss": 0.0246, "step": 38600 }, { "grad_norm": 0.2106982320547104, "learning_rate": 5.698571623299317e-05, "loss": 0.0217, "step": 38610 }, { "grad_norm": 0.5373942255973816, "learning_rate": 5.696524998244086e-05, "loss": 0.0208, "step": 38620 }, { "grad_norm": 0.39063069224357605, "learning_rate": 5.6944782541717836e-05, "loss": 0.0305, "step": 38630 }, { "grad_norm": 0.32655155658721924, "learning_rate": 5.69243139143214e-05, "loss": 0.0244, "step": 38640 }, { "grad_norm": 0.3031657934188843, "learning_rate": 5.6903844103749125e-05, "loss": 0.0262, "step": 38650 }, { "grad_norm": 0.27559706568717957, "learning_rate": 5.688337311349869e-05, "loss": 0.0198, "step": 38660 }, { "grad_norm": 0.34953343868255615, "learning_rate": 5.6862900947068074e-05, "loss": 0.0225, "step": 38670 }, { "grad_norm": 0.348797082901001, "learning_rate": 5.6842427607955374e-05, "loss": 0.0285, "step": 38680 }, { "grad_norm": 0.2994300425052643, "learning_rate": 5.682195309965893e-05, "loss": 0.0202, "step": 38690 }, { "grad_norm": 0.3586895167827606, "learning_rate": 5.6801477425677294e-05, "loss": 0.0295, "step": 38700 }, { "grad_norm": 0.40852078795433044, "learning_rate": 5.678100058950917e-05, "loss": 0.0295, "step": 38710 }, { "grad_norm": 0.5597264170646667, "learning_rate": 5.676052259465352e-05, "loss": 0.0444, "step": 38720 }, { "grad_norm": 0.313345730304718, "learning_rate": 5.674004344460945e-05, "loss": 0.0236, "step": 38730 }, { "grad_norm": 0.49011197686195374, "learning_rate": 5.6719563142876295e-05, "loss": 0.0239, "step": 38740 }, { "grad_norm": 0.2954670488834381, "learning_rate": 5.669908169295359e-05, "loss": 0.0206, "step": 38750 }, { "grad_norm": 0.344159871339798, "learning_rate": 5.667859909834105e-05, "loss": 0.0251, "step": 38760 }, { "grad_norm": 0.27945032715797424, "learning_rate": 5.6658115362538565e-05, "loss": 0.0287, "step": 38770 }, { "grad_norm": 0.357340931892395, "learning_rate": 5.663763048904628e-05, "loss": 0.0251, "step": 38780 }, { "grad_norm": 0.38365334272384644, "learning_rate": 5.661714448136447e-05, "loss": 0.0199, "step": 38790 }, { "grad_norm": 0.2540636658668518, "learning_rate": 5.659665734299366e-05, "loss": 0.0178, "step": 38800 }, { "grad_norm": 0.24848538637161255, "learning_rate": 5.6576169077434516e-05, "loss": 0.0278, "step": 38810 }, { "grad_norm": 0.285776823759079, "learning_rate": 5.6555679688187944e-05, "loss": 0.0216, "step": 38820 }, { "grad_norm": 0.41850417852401733, "learning_rate": 5.6535189178755e-05, "loss": 0.0219, "step": 38830 }, { "grad_norm": 0.34492751955986023, "learning_rate": 5.651469755263698e-05, "loss": 0.0244, "step": 38840 }, { "grad_norm": 0.19128458201885223, "learning_rate": 5.6494204813335316e-05, "loss": 0.0327, "step": 38850 }, { "grad_norm": 0.20545251667499542, "learning_rate": 5.647371096435168e-05, "loss": 0.0216, "step": 38860 }, { "grad_norm": 0.2861710488796234, "learning_rate": 5.645321600918788e-05, "loss": 0.0222, "step": 38870 }, { "grad_norm": 0.35910946130752563, "learning_rate": 5.643271995134597e-05, "loss": 0.0249, "step": 38880 }, { "grad_norm": 0.2612021863460541, "learning_rate": 5.641222279432814e-05, "loss": 0.025, "step": 38890 }, { "grad_norm": 0.5212537050247192, "learning_rate": 5.6391724541636834e-05, "loss": 0.0238, "step": 38900 }, { "grad_norm": 0.3921809494495392, "learning_rate": 5.6371225196774605e-05, "loss": 0.0268, "step": 38910 }, { "grad_norm": 0.3001176714897156, "learning_rate": 5.635072476324423e-05, "loss": 0.028, "step": 38920 }, { "grad_norm": 0.40775495767593384, "learning_rate": 5.63302232445487e-05, "loss": 0.0202, "step": 38930 }, { "grad_norm": 0.209340438246727, "learning_rate": 5.6309720644191144e-05, "loss": 0.0298, "step": 38940 }, { "grad_norm": 0.33542197942733765, "learning_rate": 5.628921696567491e-05, "loss": 0.0223, "step": 38950 }, { "grad_norm": 0.37779080867767334, "learning_rate": 5.62687122125035e-05, "loss": 0.0246, "step": 38960 }, { "grad_norm": 0.2635031044483185, "learning_rate": 5.624820638818062e-05, "loss": 0.0208, "step": 38970 }, { "grad_norm": 0.4653298854827881, "learning_rate": 5.6227699496210164e-05, "loss": 0.0357, "step": 38980 }, { "grad_norm": 0.23382075130939484, "learning_rate": 5.6207191540096195e-05, "loss": 0.0259, "step": 38990 }, { "grad_norm": 0.25456106662750244, "learning_rate": 5.618668252334296e-05, "loss": 0.0196, "step": 39000 }, { "grad_norm": 0.23657992482185364, "learning_rate": 5.616617244945488e-05, "loss": 0.0189, "step": 39010 }, { "grad_norm": 0.29159003496170044, "learning_rate": 5.614566132193656e-05, "loss": 0.0206, "step": 39020 }, { "grad_norm": 0.29372572898864746, "learning_rate": 5.612514914429282e-05, "loss": 0.024, "step": 39030 }, { "grad_norm": 0.2582457661628723, "learning_rate": 5.610463592002863e-05, "loss": 0.023, "step": 39040 }, { "grad_norm": 0.32309868931770325, "learning_rate": 5.608412165264909e-05, "loss": 0.0222, "step": 39050 }, { "grad_norm": 0.46047720313072205, "learning_rate": 5.606360634565959e-05, "loss": 0.0205, "step": 39060 }, { "grad_norm": 0.18048331141471863, "learning_rate": 5.604309000256559e-05, "loss": 0.0222, "step": 39070 }, { "grad_norm": 0.32804736495018005, "learning_rate": 5.6022572626872785e-05, "loss": 0.0182, "step": 39080 }, { "grad_norm": 0.3656793534755707, "learning_rate": 5.600205422208704e-05, "loss": 0.0276, "step": 39090 }, { "grad_norm": 0.2275058776140213, "learning_rate": 5.59815347917144e-05, "loss": 0.0253, "step": 39100 }, { "grad_norm": 0.8843688368797302, "learning_rate": 5.596101433926103e-05, "loss": 0.0305, "step": 39110 }, { "grad_norm": 0.30546364188194275, "learning_rate": 5.5940492868233364e-05, "loss": 0.0185, "step": 39120 }, { "grad_norm": 0.4252098500728607, "learning_rate": 5.591997038213793e-05, "loss": 0.0322, "step": 39130 }, { "grad_norm": 0.2824665606021881, "learning_rate": 5.5899446884481475e-05, "loss": 0.0219, "step": 39140 }, { "grad_norm": 0.32146552205085754, "learning_rate": 5.5878922378770906e-05, "loss": 0.0326, "step": 39150 }, { "grad_norm": 0.1993200033903122, "learning_rate": 5.5858396868513285e-05, "loss": 0.0248, "step": 39160 }, { "grad_norm": 0.2935481667518616, "learning_rate": 5.583787035721586e-05, "loss": 0.0319, "step": 39170 }, { "grad_norm": 0.34451648592948914, "learning_rate": 5.581734284838606e-05, "loss": 0.0244, "step": 39180 }, { "grad_norm": 0.31126338243484497, "learning_rate": 5.579681434553147e-05, "loss": 0.0223, "step": 39190 }, { "grad_norm": 0.848181962966919, "learning_rate": 5.5776284852159854e-05, "loss": 0.0246, "step": 39200 }, { "grad_norm": 0.28958526253700256, "learning_rate": 5.575575437177913e-05, "loss": 0.0312, "step": 39210 }, { "grad_norm": 0.4372452199459076, "learning_rate": 5.573522290789742e-05, "loss": 0.0223, "step": 39220 }, { "grad_norm": 0.2258947491645813, "learning_rate": 5.571469046402298e-05, "loss": 0.0242, "step": 39230 }, { "grad_norm": 0.2698385417461395, "learning_rate": 5.5694157043664205e-05, "loss": 0.0265, "step": 39240 }, { "grad_norm": 0.22759343683719635, "learning_rate": 5.567362265032975e-05, "loss": 0.0179, "step": 39250 }, { "grad_norm": 0.30995723605155945, "learning_rate": 5.565308728752836e-05, "loss": 0.0255, "step": 39260 }, { "grad_norm": 0.30256417393684387, "learning_rate": 5.5632550958768956e-05, "loss": 0.0182, "step": 39270 }, { "grad_norm": 0.33726102113723755, "learning_rate": 5.5612013667560636e-05, "loss": 0.0234, "step": 39280 }, { "grad_norm": 0.32635068893432617, "learning_rate": 5.5591475417412676e-05, "loss": 0.0361, "step": 39290 }, { "grad_norm": 0.30643659830093384, "learning_rate": 5.557093621183451e-05, "loss": 0.0275, "step": 39300 }, { "grad_norm": 0.24781009554862976, "learning_rate": 5.55503960543357e-05, "loss": 0.022, "step": 39310 }, { "grad_norm": 0.30895599722862244, "learning_rate": 5.552985494842601e-05, "loss": 0.0228, "step": 39320 }, { "grad_norm": 0.26847943663597107, "learning_rate": 5.550931289761534e-05, "loss": 0.0194, "step": 39330 }, { "grad_norm": 0.41345855593681335, "learning_rate": 5.548876990541378e-05, "loss": 0.023, "step": 39340 }, { "grad_norm": 0.4814675450325012, "learning_rate": 5.5468225975331564e-05, "loss": 0.0335, "step": 39350 }, { "grad_norm": 0.47178754210472107, "learning_rate": 5.544768111087909e-05, "loss": 0.0272, "step": 39360 }, { "grad_norm": 0.23243756592273712, "learning_rate": 5.54271353155669e-05, "loss": 0.027, "step": 39370 }, { "grad_norm": 0.40128016471862793, "learning_rate": 5.5406588592905715e-05, "loss": 0.0247, "step": 39380 }, { "grad_norm": 0.24829889833927155, "learning_rate": 5.5386040946406416e-05, "loss": 0.0276, "step": 39390 }, { "grad_norm": 0.8184577822685242, "learning_rate": 5.536549237958004e-05, "loss": 0.0323, "step": 39400 }, { "grad_norm": 0.33892229199409485, "learning_rate": 5.5344942895937744e-05, "loss": 0.0365, "step": 39410 }, { "grad_norm": 0.2753291726112366, "learning_rate": 5.5324392498990904e-05, "loss": 0.0233, "step": 39420 }, { "grad_norm": 0.3612356185913086, "learning_rate": 5.5303841192251e-05, "loss": 0.0275, "step": 39430 }, { "grad_norm": 0.3282793462276459, "learning_rate": 5.52832889792297e-05, "loss": 0.03, "step": 39440 }, { "grad_norm": 0.3478429615497589, "learning_rate": 5.526273586343881e-05, "loss": 0.0196, "step": 39450 }, { "grad_norm": 0.22424881160259247, "learning_rate": 5.5242181848390306e-05, "loss": 0.0211, "step": 39460 }, { "grad_norm": 0.26992088556289673, "learning_rate": 5.5221626937596285e-05, "loss": 0.0214, "step": 39470 }, { "grad_norm": 0.24062548577785492, "learning_rate": 5.520107113456903e-05, "loss": 0.0335, "step": 39480 }, { "grad_norm": 0.4021455645561218, "learning_rate": 5.5180514442820974e-05, "loss": 0.0317, "step": 39490 }, { "grad_norm": 0.2987028956413269, "learning_rate": 5.515995686586469e-05, "loss": 0.0249, "step": 39500 }, { "grad_norm": 0.3511926233768463, "learning_rate": 5.5139398407212916e-05, "loss": 0.0292, "step": 39510 }, { "grad_norm": 0.34146732091903687, "learning_rate": 5.511883907037849e-05, "loss": 0.0271, "step": 39520 }, { "grad_norm": 0.4504389464855194, "learning_rate": 5.509827885887449e-05, "loss": 0.0243, "step": 39530 }, { "grad_norm": 0.27588143944740295, "learning_rate": 5.507771777621406e-05, "loss": 0.027, "step": 39540 }, { "grad_norm": 0.29239553213119507, "learning_rate": 5.505715582591052e-05, "loss": 0.0386, "step": 39550 }, { "grad_norm": 0.3588120937347412, "learning_rate": 5.50365930114774e-05, "loss": 0.0281, "step": 39560 }, { "grad_norm": 0.26230159401893616, "learning_rate": 5.5016029336428255e-05, "loss": 0.0246, "step": 39570 }, { "grad_norm": 0.47284120321273804, "learning_rate": 5.49954648042769e-05, "loss": 0.0246, "step": 39580 }, { "grad_norm": 0.24884836375713348, "learning_rate": 5.4974899418537226e-05, "loss": 0.0299, "step": 39590 }, { "grad_norm": 0.43212151527404785, "learning_rate": 5.4954333182723316e-05, "loss": 0.0248, "step": 39600 }, { "grad_norm": 0.39322736859321594, "learning_rate": 5.493376610034937e-05, "loss": 0.0221, "step": 39610 }, { "grad_norm": 0.2399945706129074, "learning_rate": 5.4913198174929735e-05, "loss": 0.0249, "step": 39620 }, { "grad_norm": 0.18428660929203033, "learning_rate": 5.489262940997889e-05, "loss": 0.0252, "step": 39630 }, { "grad_norm": 0.23874415457248688, "learning_rate": 5.487205980901151e-05, "loss": 0.0282, "step": 39640 }, { "grad_norm": 0.3456481695175171, "learning_rate": 5.485148937554234e-05, "loss": 0.0421, "step": 39650 }, { "grad_norm": 0.22839957475662231, "learning_rate": 5.483091811308635e-05, "loss": 0.0208, "step": 39660 }, { "grad_norm": 0.30708667635917664, "learning_rate": 5.4810346025158564e-05, "loss": 0.0307, "step": 39670 }, { "grad_norm": 0.27474454045295715, "learning_rate": 5.478977311527421e-05, "loss": 0.0253, "step": 39680 }, { "grad_norm": 0.2153850495815277, "learning_rate": 5.476919938694863e-05, "loss": 0.0231, "step": 39690 }, { "grad_norm": 0.33279603719711304, "learning_rate": 5.474862484369733e-05, "loss": 0.0164, "step": 39700 }, { "grad_norm": 0.2648562490940094, "learning_rate": 5.472804948903589e-05, "loss": 0.0226, "step": 39710 }, { "grad_norm": 0.26686328649520874, "learning_rate": 5.470747332648013e-05, "loss": 0.0183, "step": 39720 }, { "grad_norm": 0.3531084656715393, "learning_rate": 5.468689635954591e-05, "loss": 0.039, "step": 39730 }, { "grad_norm": 0.20066635310649872, "learning_rate": 5.46663185917493e-05, "loss": 0.0223, "step": 39740 }, { "grad_norm": 0.5160815715789795, "learning_rate": 5.464574002660645e-05, "loss": 0.0248, "step": 39750 }, { "grad_norm": 0.29719799757003784, "learning_rate": 5.4625160667633724e-05, "loss": 0.024, "step": 39760 }, { "grad_norm": 0.2231619656085968, "learning_rate": 5.4604580518347505e-05, "loss": 0.0265, "step": 39770 }, { "grad_norm": 0.315513551235199, "learning_rate": 5.458399958226442e-05, "loss": 0.0291, "step": 39780 }, { "grad_norm": 0.32625696063041687, "learning_rate": 5.456341786290119e-05, "loss": 0.0262, "step": 39790 }, { "grad_norm": 0.39794686436653137, "learning_rate": 5.454283536377465e-05, "loss": 0.0266, "step": 39800 }, { "grad_norm": 0.27396515011787415, "learning_rate": 5.452225208840179e-05, "loss": 0.0229, "step": 39810 }, { "grad_norm": 0.35616084933280945, "learning_rate": 5.450166804029975e-05, "loss": 0.0297, "step": 39820 }, { "grad_norm": 0.3360559046268463, "learning_rate": 5.448108322298574e-05, "loss": 0.0193, "step": 39830 }, { "grad_norm": 0.24790287017822266, "learning_rate": 5.446049763997717e-05, "loss": 0.0223, "step": 39840 }, { "grad_norm": 0.2789706587791443, "learning_rate": 5.4439911294791546e-05, "loss": 0.0209, "step": 39850 }, { "grad_norm": 0.2988627552986145, "learning_rate": 5.441932419094652e-05, "loss": 0.0189, "step": 39860 }, { "grad_norm": 0.3326354920864105, "learning_rate": 5.439873633195985e-05, "loss": 0.0235, "step": 39870 }, { "grad_norm": 0.24367550015449524, "learning_rate": 5.437814772134947e-05, "loss": 0.0285, "step": 39880 }, { "grad_norm": 0.26194748282432556, "learning_rate": 5.4357558362633366e-05, "loss": 0.0225, "step": 39890 }, { "grad_norm": 0.1826702356338501, "learning_rate": 5.4336968259329726e-05, "loss": 0.0171, "step": 39900 }, { "grad_norm": 0.2432022988796234, "learning_rate": 5.431637741495681e-05, "loss": 0.0224, "step": 39910 }, { "grad_norm": 0.3174952268600464, "learning_rate": 5.429578583303307e-05, "loss": 0.0267, "step": 39920 }, { "grad_norm": 0.37870749831199646, "learning_rate": 5.427519351707701e-05, "loss": 0.0271, "step": 39930 }, { "grad_norm": 0.27843642234802246, "learning_rate": 5.42546004706073e-05, "loss": 0.023, "step": 39940 }, { "grad_norm": 0.33928757905960083, "learning_rate": 5.4234006697142735e-05, "loss": 0.0198, "step": 39950 }, { "grad_norm": 0.25439712405204773, "learning_rate": 5.421341220020224e-05, "loss": 0.0243, "step": 39960 }, { "grad_norm": 0.27059754729270935, "learning_rate": 5.419281698330482e-05, "loss": 0.0268, "step": 39970 }, { "grad_norm": 0.29456445574760437, "learning_rate": 5.4172221049969665e-05, "loss": 0.0234, "step": 39980 }, { "grad_norm": 0.24195219576358795, "learning_rate": 5.415162440371604e-05, "loss": 0.0199, "step": 39990 }, { "grad_norm": 0.321366548538208, "learning_rate": 5.413102704806334e-05, "loss": 0.0238, "step": 40000 }, { "grad_norm": 0.492747962474823, "learning_rate": 5.41104289865311e-05, "loss": 0.0279, "step": 40010 }, { "grad_norm": 0.2767402231693268, "learning_rate": 5.408983022263898e-05, "loss": 0.0192, "step": 40020 }, { "grad_norm": 0.45034757256507874, "learning_rate": 5.406923075990673e-05, "loss": 0.0269, "step": 40030 }, { "grad_norm": 0.29838448762893677, "learning_rate": 5.404863060185423e-05, "loss": 0.0207, "step": 40040 }, { "grad_norm": 0.3542391359806061, "learning_rate": 5.40280297520015e-05, "loss": 0.0251, "step": 40050 }, { "grad_norm": 0.23746921122074127, "learning_rate": 5.400742821386865e-05, "loss": 0.0214, "step": 40060 }, { "grad_norm": 0.2642849385738373, "learning_rate": 5.398682599097592e-05, "loss": 0.0228, "step": 40070 }, { "grad_norm": 0.25918319821357727, "learning_rate": 5.396622308684367e-05, "loss": 0.0161, "step": 40080 }, { "grad_norm": 0.4407944083213806, "learning_rate": 5.394561950499236e-05, "loss": 0.0258, "step": 40090 }, { "grad_norm": 0.2725197970867157, "learning_rate": 5.39250152489426e-05, "loss": 0.0249, "step": 40100 }, { "grad_norm": 0.3511219918727875, "learning_rate": 5.390441032221507e-05, "loss": 0.0201, "step": 40110 }, { "grad_norm": 0.31623148918151855, "learning_rate": 5.388380472833062e-05, "loss": 0.024, "step": 40120 }, { "grad_norm": 0.2519453167915344, "learning_rate": 5.386319847081016e-05, "loss": 0.0261, "step": 40130 }, { "grad_norm": 0.37614142894744873, "learning_rate": 5.384259155317473e-05, "loss": 0.0256, "step": 40140 }, { "grad_norm": 0.41001251339912415, "learning_rate": 5.38219839789455e-05, "loss": 0.0277, "step": 40150 }, { "grad_norm": 0.23717860877513885, "learning_rate": 5.380137575164376e-05, "loss": 0.033, "step": 40160 }, { "grad_norm": 0.2572578489780426, "learning_rate": 5.378076687479085e-05, "loss": 0.017, "step": 40170 }, { "grad_norm": 0.2152700126171112, "learning_rate": 5.3760157351908305e-05, "loss": 0.0232, "step": 40180 }, { "grad_norm": 0.19991417229175568, "learning_rate": 5.373954718651768e-05, "loss": 0.0195, "step": 40190 }, { "grad_norm": 0.16354693472385406, "learning_rate": 5.371893638214074e-05, "loss": 0.0194, "step": 40200 }, { "grad_norm": 0.2911522090435028, "learning_rate": 5.369832494229927e-05, "loss": 0.02, "step": 40210 }, { "grad_norm": 0.37522512674331665, "learning_rate": 5.367771287051524e-05, "loss": 0.0281, "step": 40220 }, { "grad_norm": 0.26861101388931274, "learning_rate": 5.365710017031065e-05, "loss": 0.0265, "step": 40230 }, { "grad_norm": 0.19462619721889496, "learning_rate": 5.3636486845207654e-05, "loss": 0.0215, "step": 40240 }, { "grad_norm": 0.2610184848308563, "learning_rate": 5.3615872898728526e-05, "loss": 0.0184, "step": 40250 }, { "grad_norm": 0.57681804895401, "learning_rate": 5.3595258334395614e-05, "loss": 0.0284, "step": 40260 }, { "grad_norm": 0.38057488203048706, "learning_rate": 5.3574643155731396e-05, "loss": 0.0289, "step": 40270 }, { "grad_norm": 0.4746171236038208, "learning_rate": 5.3554027366258405e-05, "loss": 0.0268, "step": 40280 }, { "grad_norm": 0.2857808768749237, "learning_rate": 5.3533410969499355e-05, "loss": 0.0254, "step": 40290 }, { "grad_norm": 0.2157643735408783, "learning_rate": 5.3512793968977e-05, "loss": 0.0225, "step": 40300 }, { "grad_norm": 0.3762705624103546, "learning_rate": 5.349217636821423e-05, "loss": 0.0237, "step": 40310 }, { "grad_norm": 0.28476640582084656, "learning_rate": 5.347155817073405e-05, "loss": 0.0239, "step": 40320 }, { "grad_norm": 0.2854296863079071, "learning_rate": 5.34509393800595e-05, "loss": 0.0278, "step": 40330 }, { "grad_norm": 0.3204137980937958, "learning_rate": 5.343031999971381e-05, "loss": 0.0253, "step": 40340 }, { "grad_norm": 0.3690173923969269, "learning_rate": 5.340970003322026e-05, "loss": 0.0242, "step": 40350 }, { "grad_norm": 0.5342605710029602, "learning_rate": 5.33890794841022e-05, "loss": 0.0231, "step": 40360 }, { "grad_norm": 0.2540609836578369, "learning_rate": 5.336845835588318e-05, "loss": 0.0231, "step": 40370 }, { "grad_norm": 0.24927610158920288, "learning_rate": 5.334783665208674e-05, "loss": 0.0223, "step": 40380 }, { "grad_norm": 0.3343549966812134, "learning_rate": 5.332721437623657e-05, "loss": 0.0178, "step": 40390 }, { "grad_norm": 0.6245632171630859, "learning_rate": 5.3306591531856464e-05, "loss": 0.0284, "step": 40400 }, { "grad_norm": 0.28084442019462585, "learning_rate": 5.3285968122470295e-05, "loss": 0.0233, "step": 40410 }, { "grad_norm": 0.2927774488925934, "learning_rate": 5.326534415160207e-05, "loss": 0.0288, "step": 40420 }, { "grad_norm": 0.32009297609329224, "learning_rate": 5.324471962277582e-05, "loss": 0.0348, "step": 40430 }, { "grad_norm": 0.2961690127849579, "learning_rate": 5.3224094539515746e-05, "loss": 0.0214, "step": 40440 }, { "grad_norm": 0.345122754573822, "learning_rate": 5.3203468905346076e-05, "loss": 0.027, "step": 40450 }, { "grad_norm": 0.3129313588142395, "learning_rate": 5.3182842723791195e-05, "loss": 0.0265, "step": 40460 }, { "grad_norm": 0.28083309531211853, "learning_rate": 5.316221599837554e-05, "loss": 0.0227, "step": 40470 }, { "grad_norm": 0.31770527362823486, "learning_rate": 5.314158873262366e-05, "loss": 0.0302, "step": 40480 }, { "grad_norm": 0.27877217531204224, "learning_rate": 5.312096093006018e-05, "loss": 0.0207, "step": 40490 }, { "grad_norm": 0.5013751983642578, "learning_rate": 5.3100332594209846e-05, "loss": 0.0244, "step": 40500 }, { "grad_norm": 0.28331536054611206, "learning_rate": 5.3079703728597454e-05, "loss": 0.028, "step": 40510 }, { "grad_norm": 0.32099786400794983, "learning_rate": 5.305907433674794e-05, "loss": 0.0209, "step": 40520 }, { "grad_norm": 0.2775461971759796, "learning_rate": 5.3038444422186284e-05, "loss": 0.0233, "step": 40530 }, { "grad_norm": 0.28850632905960083, "learning_rate": 5.301781398843757e-05, "loss": 0.0232, "step": 40540 }, { "grad_norm": 0.24066609144210815, "learning_rate": 5.299718303902699e-05, "loss": 0.0223, "step": 40550 }, { "grad_norm": 0.2456623613834381, "learning_rate": 5.297655157747979e-05, "loss": 0.0209, "step": 40560 }, { "grad_norm": 0.41998419165611267, "learning_rate": 5.295591960732136e-05, "loss": 0.0232, "step": 40570 }, { "grad_norm": 0.30856749415397644, "learning_rate": 5.293528713207708e-05, "loss": 0.0271, "step": 40580 }, { "grad_norm": 0.214792862534523, "learning_rate": 5.291465415527253e-05, "loss": 0.024, "step": 40590 }, { "grad_norm": 0.19959810376167297, "learning_rate": 5.2894020680433296e-05, "loss": 0.0207, "step": 40600 }, { "grad_norm": 0.28039970993995667, "learning_rate": 5.287338671108507e-05, "loss": 0.0237, "step": 40610 }, { "grad_norm": 0.6213350892066956, "learning_rate": 5.285275225075367e-05, "loss": 0.025, "step": 40620 }, { "grad_norm": 0.3508724272251129, "learning_rate": 5.283211730296492e-05, "loss": 0.0217, "step": 40630 }, { "grad_norm": 0.2398073524236679, "learning_rate": 5.281148187124477e-05, "loss": 0.0237, "step": 40640 }, { "grad_norm": 0.3062114119529724, "learning_rate": 5.279084595911927e-05, "loss": 0.0256, "step": 40650 }, { "grad_norm": 0.2983595132827759, "learning_rate": 5.277020957011453e-05, "loss": 0.0317, "step": 40660 }, { "grad_norm": 0.5493999123573303, "learning_rate": 5.274957270775673e-05, "loss": 0.0318, "step": 40670 }, { "grad_norm": 0.28218430280685425, "learning_rate": 5.2728935375572164e-05, "loss": 0.0213, "step": 40680 }, { "grad_norm": 0.23611102998256683, "learning_rate": 5.2708297577087176e-05, "loss": 0.0223, "step": 40690 }, { "grad_norm": 0.3889245390892029, "learning_rate": 5.2687659315828196e-05, "loss": 0.021, "step": 40700 }, { "grad_norm": 0.3303447961807251, "learning_rate": 5.266702059532175e-05, "loss": 0.0218, "step": 40710 }, { "grad_norm": 0.3060084879398346, "learning_rate": 5.264638141909444e-05, "loss": 0.0271, "step": 40720 }, { "grad_norm": 0.3705429136753082, "learning_rate": 5.2625741790672925e-05, "loss": 0.0236, "step": 40730 }, { "grad_norm": 0.2539551556110382, "learning_rate": 5.2605101713583935e-05, "loss": 0.0207, "step": 40740 }, { "grad_norm": 0.35827523469924927, "learning_rate": 5.2584461191354315e-05, "loss": 0.0258, "step": 40750 }, { "grad_norm": 0.30779504776000977, "learning_rate": 5.2563820227510964e-05, "loss": 0.0237, "step": 40760 }, { "grad_norm": 0.2087603658437729, "learning_rate": 5.2543178825580864e-05, "loss": 0.0252, "step": 40770 }, { "grad_norm": 0.4115571081638336, "learning_rate": 5.2522536989091054e-05, "loss": 0.0266, "step": 40780 }, { "grad_norm": 0.32170912623405457, "learning_rate": 5.2501894721568655e-05, "loss": 0.0232, "step": 40790 }, { "grad_norm": 0.35833418369293213, "learning_rate": 5.248125202654089e-05, "loss": 0.0223, "step": 40800 }, { "grad_norm": 0.3301600217819214, "learning_rate": 5.246060890753501e-05, "loss": 0.0325, "step": 40810 }, { "grad_norm": 0.1988617479801178, "learning_rate": 5.243996536807837e-05, "loss": 0.0255, "step": 40820 }, { "grad_norm": 0.2587383985519409, "learning_rate": 5.2419321411698384e-05, "loss": 0.0188, "step": 40830 }, { "grad_norm": 0.27128705382347107, "learning_rate": 5.239867704192253e-05, "loss": 0.0251, "step": 40840 }, { "grad_norm": 0.18297342956066132, "learning_rate": 5.237803226227838e-05, "loss": 0.0209, "step": 40850 }, { "grad_norm": 0.23801490664482117, "learning_rate": 5.235738707629354e-05, "loss": 0.0276, "step": 40860 }, { "grad_norm": 0.3548225164413452, "learning_rate": 5.233674148749575e-05, "loss": 0.0291, "step": 40870 }, { "grad_norm": 0.7737565040588379, "learning_rate": 5.231609549941272e-05, "loss": 0.0305, "step": 40880 }, { "grad_norm": 0.45660945773124695, "learning_rate": 5.229544911557233e-05, "loss": 0.0286, "step": 40890 }, { "grad_norm": 0.15097907185554504, "learning_rate": 5.227480233950246e-05, "loss": 0.0267, "step": 40900 }, { "grad_norm": 0.3234502971172333, "learning_rate": 5.22541551747311e-05, "loss": 0.0283, "step": 40910 }, { "grad_norm": 1.7744258642196655, "learning_rate": 5.223350762478626e-05, "loss": 0.0237, "step": 40920 }, { "grad_norm": 0.2694910168647766, "learning_rate": 5.221285969319606e-05, "loss": 0.022, "step": 40930 }, { "grad_norm": 0.2583896219730377, "learning_rate": 5.219221138348865e-05, "loss": 0.0246, "step": 40940 }, { "grad_norm": 0.31811338663101196, "learning_rate": 5.217156269919228e-05, "loss": 0.0246, "step": 40950 }, { "grad_norm": 0.3552875816822052, "learning_rate": 5.215091364383523e-05, "loss": 0.0208, "step": 40960 }, { "grad_norm": 0.4341040253639221, "learning_rate": 5.213026422094588e-05, "loss": 0.0228, "step": 40970 }, { "grad_norm": 0.20928341150283813, "learning_rate": 5.210961443405262e-05, "loss": 0.0199, "step": 40980 }, { "grad_norm": 0.24789288640022278, "learning_rate": 5.208896428668396e-05, "loss": 0.0169, "step": 40990 }, { "grad_norm": 0.29873332381248474, "learning_rate": 5.206831378236845e-05, "loss": 0.0233, "step": 41000 }, { "grad_norm": 0.29982897639274597, "learning_rate": 5.2047662924634666e-05, "loss": 0.0194, "step": 41010 }, { "grad_norm": 0.293069064617157, "learning_rate": 5.20270117170113e-05, "loss": 0.0219, "step": 41020 }, { "grad_norm": 0.43782734870910645, "learning_rate": 5.200636016302707e-05, "loss": 0.0307, "step": 41030 }, { "grad_norm": 0.2931578755378723, "learning_rate": 5.1985708266210754e-05, "loss": 0.0296, "step": 41040 }, { "grad_norm": 0.24031446874141693, "learning_rate": 5.1965056030091206e-05, "loss": 0.0241, "step": 41050 }, { "grad_norm": 0.22649145126342773, "learning_rate": 5.194440345819732e-05, "loss": 0.0217, "step": 41060 }, { "grad_norm": 0.2922728359699249, "learning_rate": 5.1923750554058084e-05, "loss": 0.025, "step": 41070 }, { "grad_norm": 0.24718095362186432, "learning_rate": 5.1903097321202476e-05, "loss": 0.0232, "step": 41080 }, { "grad_norm": 0.3908497095108032, "learning_rate": 5.18824437631596e-05, "loss": 0.0309, "step": 41090 }, { "grad_norm": 0.2835119962692261, "learning_rate": 5.186178988345856e-05, "loss": 0.0239, "step": 41100 }, { "grad_norm": 0.4456649422645569, "learning_rate": 5.184113568562855e-05, "loss": 0.0237, "step": 41110 }, { "grad_norm": 0.26879945397377014, "learning_rate": 5.18204811731988e-05, "loss": 0.0263, "step": 41120 }, { "grad_norm": 0.5811226963996887, "learning_rate": 5.179982634969862e-05, "loss": 0.0215, "step": 41130 }, { "grad_norm": 0.4249708950519562, "learning_rate": 5.177917121865734e-05, "loss": 0.0301, "step": 41140 }, { "grad_norm": 0.49482816457748413, "learning_rate": 5.1758515783604346e-05, "loss": 0.0221, "step": 41150 }, { "grad_norm": 0.4289616644382477, "learning_rate": 5.173786004806911e-05, "loss": 0.0295, "step": 41160 }, { "grad_norm": 0.5313452482223511, "learning_rate": 5.1717204015581135e-05, "loss": 0.0247, "step": 41170 }, { "grad_norm": 0.28403282165527344, "learning_rate": 5.169654768966994e-05, "loss": 0.0241, "step": 41180 }, { "grad_norm": 0.38529106974601746, "learning_rate": 5.1675891073865156e-05, "loss": 0.0257, "step": 41190 }, { "grad_norm": 0.1984928995370865, "learning_rate": 5.1655234171696424e-05, "loss": 0.0174, "step": 41200 }, { "grad_norm": 0.2190743088722229, "learning_rate": 5.163457698669343e-05, "loss": 0.0177, "step": 41210 }, { "grad_norm": 0.2922353446483612, "learning_rate": 5.1613919522385936e-05, "loss": 0.0207, "step": 41220 }, { "grad_norm": 0.4635431170463562, "learning_rate": 5.1593261782303746e-05, "loss": 0.0306, "step": 41230 }, { "grad_norm": 0.36517783999443054, "learning_rate": 5.157260376997669e-05, "loss": 0.0235, "step": 41240 }, { "grad_norm": 0.3827003836631775, "learning_rate": 5.155194548893466e-05, "loss": 0.0298, "step": 41250 }, { "grad_norm": 0.21921339631080627, "learning_rate": 5.153128694270759e-05, "loss": 0.0311, "step": 41260 }, { "grad_norm": 0.29452627897262573, "learning_rate": 5.151062813482548e-05, "loss": 0.0313, "step": 41270 }, { "grad_norm": 0.3104044497013092, "learning_rate": 5.148996906881832e-05, "loss": 0.0177, "step": 41280 }, { "grad_norm": 0.35709014534950256, "learning_rate": 5.1469309748216196e-05, "loss": 0.0205, "step": 41290 }, { "grad_norm": 0.2538996636867523, "learning_rate": 5.144865017654923e-05, "loss": 0.0165, "step": 41300 }, { "grad_norm": 0.27229008078575134, "learning_rate": 5.1427990357347556e-05, "loss": 0.0221, "step": 41310 }, { "grad_norm": 0.1775854527950287, "learning_rate": 5.140733029414139e-05, "loss": 0.0233, "step": 41320 }, { "grad_norm": 0.355319082736969, "learning_rate": 5.138666999046098e-05, "loss": 0.0244, "step": 41330 }, { "grad_norm": 0.4618789553642273, "learning_rate": 5.136600944983658e-05, "loss": 0.022, "step": 41340 }, { "grad_norm": 0.31298837065696716, "learning_rate": 5.134534867579853e-05, "loss": 0.0203, "step": 41350 }, { "grad_norm": 0.38640764355659485, "learning_rate": 5.1324687671877184e-05, "loss": 0.0199, "step": 41360 }, { "grad_norm": 0.28114745020866394, "learning_rate": 5.130402644160296e-05, "loss": 0.028, "step": 41370 }, { "grad_norm": 0.25838759541511536, "learning_rate": 5.128336498850628e-05, "loss": 0.0205, "step": 41380 }, { "grad_norm": 0.6033904552459717, "learning_rate": 5.126270331611761e-05, "loss": 0.0244, "step": 41390 }, { "grad_norm": 0.3822213411331177, "learning_rate": 5.124204142796748e-05, "loss": 0.0213, "step": 41400 }, { "grad_norm": 0.3447285592556, "learning_rate": 5.122137932758644e-05, "loss": 0.0239, "step": 41410 }, { "grad_norm": 0.23774360120296478, "learning_rate": 5.1200717018505074e-05, "loss": 0.0213, "step": 41420 }, { "grad_norm": 0.32598555088043213, "learning_rate": 5.118005450425403e-05, "loss": 0.0216, "step": 41430 }, { "grad_norm": 0.3879969120025635, "learning_rate": 5.115939178836392e-05, "loss": 0.0193, "step": 41440 }, { "grad_norm": 0.3754434585571289, "learning_rate": 5.113872887436547e-05, "loss": 0.022, "step": 41450 }, { "grad_norm": 0.264105886220932, "learning_rate": 5.11180657657894e-05, "loss": 0.0242, "step": 41460 }, { "grad_norm": 0.3917689621448517, "learning_rate": 5.109740246616649e-05, "loss": 0.0211, "step": 41470 }, { "grad_norm": 0.43049657344818115, "learning_rate": 5.10767389790275e-05, "loss": 0.034, "step": 41480 }, { "grad_norm": 0.42228958010673523, "learning_rate": 5.1056075307903265e-05, "loss": 0.0218, "step": 41490 }, { "grad_norm": 0.26305556297302246, "learning_rate": 5.1035411456324634e-05, "loss": 0.023, "step": 41500 }, { "grad_norm": 0.29235365986824036, "learning_rate": 5.1014747427822504e-05, "loss": 0.0257, "step": 41510 }, { "grad_norm": 0.3127252161502838, "learning_rate": 5.0994083225927804e-05, "loss": 0.0281, "step": 41520 }, { "grad_norm": 0.290885329246521, "learning_rate": 5.0973418854171475e-05, "loss": 0.0244, "step": 41530 }, { "grad_norm": 0.3167831599712372, "learning_rate": 5.095275431608447e-05, "loss": 0.0244, "step": 41540 }, { "grad_norm": 0.22539475560188293, "learning_rate": 5.093208961519782e-05, "loss": 0.0201, "step": 41550 }, { "grad_norm": 0.3498346209526062, "learning_rate": 5.091142475504255e-05, "loss": 0.0224, "step": 41560 }, { "grad_norm": 0.3855073153972626, "learning_rate": 5.089075973914971e-05, "loss": 0.0237, "step": 41570 }, { "grad_norm": 0.3734027147293091, "learning_rate": 5.08700945710504e-05, "loss": 0.0286, "step": 41580 }, { "grad_norm": 0.26259130239486694, "learning_rate": 5.0849429254275714e-05, "loss": 0.0243, "step": 41590 }, { "grad_norm": 0.17351631820201874, "learning_rate": 5.0828763792356804e-05, "loss": 0.0184, "step": 41600 }, { "grad_norm": 0.39951834082603455, "learning_rate": 5.080809818882483e-05, "loss": 0.0256, "step": 41610 }, { "grad_norm": 0.2975984811782837, "learning_rate": 5.078743244721097e-05, "loss": 0.0246, "step": 41620 }, { "grad_norm": 0.3600899279117584, "learning_rate": 5.0766766571046455e-05, "loss": 0.0228, "step": 41630 }, { "grad_norm": 0.33330270648002625, "learning_rate": 5.07461005638625e-05, "loss": 0.0232, "step": 41640 }, { "grad_norm": 0.3029423654079437, "learning_rate": 5.072543442919037e-05, "loss": 0.0214, "step": 41650 }, { "grad_norm": 0.31136372685432434, "learning_rate": 5.070476817056132e-05, "loss": 0.0301, "step": 41660 }, { "grad_norm": 0.29870685935020447, "learning_rate": 5.068410179150668e-05, "loss": 0.0163, "step": 41670 }, { "grad_norm": 0.43588942289352417, "learning_rate": 5.066343529555775e-05, "loss": 0.0297, "step": 41680 }, { "grad_norm": 0.4924378991127014, "learning_rate": 5.064276868624587e-05, "loss": 0.0216, "step": 41690 }, { "grad_norm": 0.3931209444999695, "learning_rate": 5.0622101967102396e-05, "loss": 0.0207, "step": 41700 }, { "grad_norm": 0.3047684133052826, "learning_rate": 5.0601435141658714e-05, "loss": 0.0231, "step": 41710 }, { "grad_norm": 0.33736371994018555, "learning_rate": 5.058076821344622e-05, "loss": 0.022, "step": 41720 }, { "grad_norm": 0.38492658734321594, "learning_rate": 5.056010118599632e-05, "loss": 0.0312, "step": 41730 }, { "grad_norm": 0.36750271916389465, "learning_rate": 5.053943406284044e-05, "loss": 0.0228, "step": 41740 }, { "grad_norm": 0.40865644812583923, "learning_rate": 5.051876684751006e-05, "loss": 0.0266, "step": 41750 }, { "grad_norm": 0.33210840821266174, "learning_rate": 5.0498099543536584e-05, "loss": 0.0226, "step": 41760 }, { "grad_norm": 0.31484147906303406, "learning_rate": 5.047743215445152e-05, "loss": 0.0211, "step": 41770 }, { "grad_norm": 0.33598873019218445, "learning_rate": 5.045676468378637e-05, "loss": 0.0256, "step": 41780 }, { "grad_norm": 0.3536732792854309, "learning_rate": 5.0436097135072626e-05, "loss": 0.0281, "step": 41790 }, { "grad_norm": 0.318274587392807, "learning_rate": 5.041542951184181e-05, "loss": 0.0213, "step": 41800 }, { "grad_norm": 0.26343831419944763, "learning_rate": 5.039476181762545e-05, "loss": 0.0209, "step": 41810 }, { "grad_norm": 0.23602595925331116, "learning_rate": 5.037409405595508e-05, "loss": 0.0148, "step": 41820 }, { "grad_norm": 0.2993618845939636, "learning_rate": 5.035342623036229e-05, "loss": 0.0225, "step": 41830 }, { "grad_norm": 0.3106487989425659, "learning_rate": 5.033275834437862e-05, "loss": 0.0243, "step": 41840 }, { "grad_norm": 0.28930383920669556, "learning_rate": 5.031209040153564e-05, "loss": 0.0175, "step": 41850 }, { "grad_norm": 0.5782111883163452, "learning_rate": 5.0291422405364955e-05, "loss": 0.0436, "step": 41860 }, { "grad_norm": 0.325106680393219, "learning_rate": 5.0270754359398133e-05, "loss": 0.0214, "step": 41870 }, { "grad_norm": 0.34948134422302246, "learning_rate": 5.025008626716682e-05, "loss": 0.0213, "step": 41880 }, { "grad_norm": 0.2582939863204956, "learning_rate": 5.0229418132202585e-05, "loss": 0.0197, "step": 41890 }, { "grad_norm": 0.3015553057193756, "learning_rate": 5.020874995803707e-05, "loss": 0.0276, "step": 41900 }, { "grad_norm": 0.29624617099761963, "learning_rate": 5.01880817482019e-05, "loss": 0.0294, "step": 41910 }, { "grad_norm": 0.2581988573074341, "learning_rate": 5.01674135062287e-05, "loss": 0.021, "step": 41920 }, { "grad_norm": 0.3794972002506256, "learning_rate": 5.0146745235649115e-05, "loss": 0.0249, "step": 41930 }, { "grad_norm": 0.34405574202537537, "learning_rate": 5.0126076939994795e-05, "loss": 0.0225, "step": 41940 }, { "grad_norm": 0.17282435297966003, "learning_rate": 5.010540862279736e-05, "loss": 0.0259, "step": 41950 }, { "grad_norm": 0.27578872442245483, "learning_rate": 5.008474028758846e-05, "loss": 0.0259, "step": 41960 }, { "grad_norm": 0.32741299271583557, "learning_rate": 5.0064071937899784e-05, "loss": 0.0264, "step": 41970 }, { "grad_norm": 0.3417796492576599, "learning_rate": 5.004340357726296e-05, "loss": 0.0243, "step": 41980 }, { "grad_norm": 0.5254642963409424, "learning_rate": 5.002273520920965e-05, "loss": 0.021, "step": 41990 }, { "grad_norm": 0.2630573511123657, "learning_rate": 5.000206683727151e-05, "loss": 0.0196, "step": 42000 }, { "grad_norm": 0.23057158291339874, "learning_rate": 4.998139846498021e-05, "loss": 0.0188, "step": 42010 }, { "grad_norm": 0.2988618314266205, "learning_rate": 4.9960730095867405e-05, "loss": 0.0204, "step": 42020 }, { "grad_norm": 0.31732499599456787, "learning_rate": 4.9940061733464755e-05, "loss": 0.0258, "step": 42030 }, { "grad_norm": 0.2710537910461426, "learning_rate": 4.991939338130392e-05, "loss": 0.0295, "step": 42040 }, { "grad_norm": 0.4583108425140381, "learning_rate": 4.989872504291653e-05, "loss": 0.0323, "step": 42050 }, { "grad_norm": 0.364540696144104, "learning_rate": 4.9878056721834273e-05, "loss": 0.0256, "step": 42060 }, { "grad_norm": 0.22178657352924347, "learning_rate": 4.98573884215888e-05, "loss": 0.019, "step": 42070 }, { "grad_norm": 0.4333505630493164, "learning_rate": 4.9836720145711715e-05, "loss": 0.0249, "step": 42080 }, { "grad_norm": 0.30109769105911255, "learning_rate": 4.9816051897734725e-05, "loss": 0.0251, "step": 42090 }, { "grad_norm": 0.3366502523422241, "learning_rate": 4.979538368118941e-05, "loss": 0.0278, "step": 42100 }, { "grad_norm": 0.18605872988700867, "learning_rate": 4.9774715499607446e-05, "loss": 0.0203, "step": 42110 }, { "grad_norm": 0.437063992023468, "learning_rate": 4.9754047356520416e-05, "loss": 0.0307, "step": 42120 }, { "grad_norm": 0.4374529719352722, "learning_rate": 4.973337925545999e-05, "loss": 0.0231, "step": 42130 }, { "grad_norm": 0.26335951685905457, "learning_rate": 4.9712711199957746e-05, "loss": 0.0172, "step": 42140 }, { "grad_norm": 0.3738362491130829, "learning_rate": 4.969204319354531e-05, "loss": 0.0273, "step": 42150 }, { "grad_norm": 0.25990796089172363, "learning_rate": 4.9671375239754267e-05, "loss": 0.0212, "step": 42160 }, { "grad_norm": 0.24040070176124573, "learning_rate": 4.9650707342116195e-05, "loss": 0.0176, "step": 42170 }, { "grad_norm": 0.2382410615682602, "learning_rate": 4.96300395041627e-05, "loss": 0.0197, "step": 42180 }, { "grad_norm": 0.48350751399993896, "learning_rate": 4.960937172942532e-05, "loss": 0.0197, "step": 42190 }, { "grad_norm": 0.18801265954971313, "learning_rate": 4.958870402143565e-05, "loss": 0.0208, "step": 42200 }, { "grad_norm": 0.2726118564605713, "learning_rate": 4.9568036383725186e-05, "loss": 0.0242, "step": 42210 }, { "grad_norm": 0.24752134084701538, "learning_rate": 4.9547368819825516e-05, "loss": 0.0185, "step": 42220 }, { "grad_norm": 0.3592851758003235, "learning_rate": 4.952670133326812e-05, "loss": 0.0227, "step": 42230 }, { "grad_norm": 0.23017825186252594, "learning_rate": 4.950603392758453e-05, "loss": 0.0229, "step": 42240 }, { "grad_norm": 0.2942015826702118, "learning_rate": 4.948536660630621e-05, "loss": 0.0228, "step": 42250 }, { "grad_norm": 0.2810133099555969, "learning_rate": 4.9464699372964685e-05, "loss": 0.0229, "step": 42260 }, { "grad_norm": 0.23967882990837097, "learning_rate": 4.9444032231091395e-05, "loss": 0.0195, "step": 42270 }, { "grad_norm": 0.3920389413833618, "learning_rate": 4.9423365184217776e-05, "loss": 0.0194, "step": 42280 }, { "grad_norm": 0.32940998673439026, "learning_rate": 4.940269823587529e-05, "loss": 0.0227, "step": 42290 }, { "grad_norm": 0.2767801880836487, "learning_rate": 4.938203138959533e-05, "loss": 0.0158, "step": 42300 }, { "grad_norm": 0.44137659668922424, "learning_rate": 4.9361364648909325e-05, "loss": 0.0226, "step": 42310 }, { "grad_norm": 0.20260769128799438, "learning_rate": 4.9340698017348605e-05, "loss": 0.0167, "step": 42320 }, { "grad_norm": 0.30056166648864746, "learning_rate": 4.932003149844458e-05, "loss": 0.0154, "step": 42330 }, { "grad_norm": 0.39665335416793823, "learning_rate": 4.929936509572857e-05, "loss": 0.0311, "step": 42340 }, { "grad_norm": 0.4559462070465088, "learning_rate": 4.927869881273191e-05, "loss": 0.024, "step": 42350 }, { "grad_norm": 0.3444007337093353, "learning_rate": 4.9258032652985894e-05, "loss": 0.0241, "step": 42360 }, { "grad_norm": 0.253884881734848, "learning_rate": 4.9237366620021786e-05, "loss": 0.0269, "step": 42370 }, { "grad_norm": 0.330171674489975, "learning_rate": 4.921670071737089e-05, "loss": 0.02, "step": 42380 }, { "grad_norm": 0.25044289231300354, "learning_rate": 4.91960349485644e-05, "loss": 0.0236, "step": 42390 }, { "grad_norm": 0.30138373374938965, "learning_rate": 4.9175369317133566e-05, "loss": 0.0222, "step": 42400 }, { "grad_norm": 0.21453523635864258, "learning_rate": 4.915470382660953e-05, "loss": 0.0191, "step": 42410 }, { "grad_norm": 0.2124481350183487, "learning_rate": 4.9134038480523524e-05, "loss": 0.0219, "step": 42420 }, { "grad_norm": 0.23069781064987183, "learning_rate": 4.911337328240664e-05, "loss": 0.0217, "step": 42430 }, { "grad_norm": 0.17669424414634705, "learning_rate": 4.909270823579003e-05, "loss": 0.0216, "step": 42440 }, { "grad_norm": 0.30706918239593506, "learning_rate": 4.907204334420476e-05, "loss": 0.0217, "step": 42450 }, { "grad_norm": 0.24244283139705658, "learning_rate": 4.9051378611181884e-05, "loss": 0.0227, "step": 42460 }, { "grad_norm": 0.2396591156721115, "learning_rate": 4.9030714040252486e-05, "loss": 0.0221, "step": 42470 }, { "grad_norm": 0.20379525423049927, "learning_rate": 4.901004963494752e-05, "loss": 0.0239, "step": 42480 }, { "grad_norm": 0.3548405170440674, "learning_rate": 4.898938539879802e-05, "loss": 0.0286, "step": 42490 }, { "grad_norm": 0.35569483041763306, "learning_rate": 4.8968721335334904e-05, "loss": 0.0259, "step": 42500 }, { "grad_norm": 0.35216042399406433, "learning_rate": 4.894805744808912e-05, "loss": 0.0242, "step": 42510 }, { "grad_norm": 0.31081146001815796, "learning_rate": 4.8927393740591524e-05, "loss": 0.0227, "step": 42520 }, { "grad_norm": 0.30189982056617737, "learning_rate": 4.890673021637302e-05, "loss": 0.019, "step": 42530 }, { "grad_norm": 0.5581313967704773, "learning_rate": 4.888606687896442e-05, "loss": 0.0297, "step": 42540 }, { "grad_norm": 0.25180792808532715, "learning_rate": 4.886540373189652e-05, "loss": 0.022, "step": 42550 }, { "grad_norm": 0.3816031217575073, "learning_rate": 4.88447407787001e-05, "loss": 0.0234, "step": 42560 }, { "grad_norm": 0.33550894260406494, "learning_rate": 4.8824078022905864e-05, "loss": 0.0218, "step": 42570 }, { "grad_norm": 0.2743229269981384, "learning_rate": 4.880341546804455e-05, "loss": 0.0175, "step": 42580 }, { "grad_norm": 0.35830798745155334, "learning_rate": 4.878275311764679e-05, "loss": 0.0254, "step": 42590 }, { "grad_norm": 0.37818431854248047, "learning_rate": 4.8762090975243254e-05, "loss": 0.0236, "step": 42600 }, { "grad_norm": 0.23971500992774963, "learning_rate": 4.874142904436448e-05, "loss": 0.0199, "step": 42610 }, { "grad_norm": 0.2851108908653259, "learning_rate": 4.8720767328541094e-05, "loss": 0.0305, "step": 42620 }, { "grad_norm": 0.4214451313018799, "learning_rate": 4.870010583130358e-05, "loss": 0.0207, "step": 42630 }, { "grad_norm": 0.44078677892684937, "learning_rate": 4.867944455618241e-05, "loss": 0.0284, "step": 42640 }, { "grad_norm": 0.30145952105522156, "learning_rate": 4.865878350670807e-05, "loss": 0.0215, "step": 42650 }, { "grad_norm": 0.30782750248908997, "learning_rate": 4.8638122686410914e-05, "loss": 0.0222, "step": 42660 }, { "grad_norm": 0.30469781160354614, "learning_rate": 4.861746209882137e-05, "loss": 0.0162, "step": 42670 }, { "grad_norm": 0.21243174374103546, "learning_rate": 4.859680174746972e-05, "loss": 0.0222, "step": 42680 }, { "grad_norm": 0.3014833331108093, "learning_rate": 4.857614163588629e-05, "loss": 0.0285, "step": 42690 }, { "grad_norm": 0.2944353520870209, "learning_rate": 4.855548176760131e-05, "loss": 0.0214, "step": 42700 }, { "grad_norm": 0.2858824133872986, "learning_rate": 4.8534822146144986e-05, "loss": 0.0233, "step": 42710 }, { "grad_norm": 0.34228140115737915, "learning_rate": 4.851416277504749e-05, "loss": 0.0335, "step": 42720 }, { "grad_norm": 0.25543269515037537, "learning_rate": 4.8493503657838923e-05, "loss": 0.0209, "step": 42730 }, { "grad_norm": 0.2058272659778595, "learning_rate": 4.84728447980494e-05, "loss": 0.02, "step": 42740 }, { "grad_norm": 0.7231220006942749, "learning_rate": 4.8452186199208914e-05, "loss": 0.0243, "step": 42750 }, { "grad_norm": 0.32280629873275757, "learning_rate": 4.843152786484749e-05, "loss": 0.0221, "step": 42760 }, { "grad_norm": 0.2016136795282364, "learning_rate": 4.8410869798495035e-05, "loss": 0.0195, "step": 42770 }, { "grad_norm": 0.26401805877685547, "learning_rate": 4.8390212003681486e-05, "loss": 0.02, "step": 42780 }, { "grad_norm": 0.45901212096214294, "learning_rate": 4.836955448393667e-05, "loss": 0.0305, "step": 42790 }, { "grad_norm": 0.25752347707748413, "learning_rate": 4.8348897242790406e-05, "loss": 0.0193, "step": 42800 }, { "grad_norm": 0.2700982987880707, "learning_rate": 4.832824028377243e-05, "loss": 0.0198, "step": 42810 }, { "grad_norm": 0.5579552054405212, "learning_rate": 4.830758361041249e-05, "loss": 0.0192, "step": 42820 }, { "grad_norm": 0.29521435499191284, "learning_rate": 4.828692722624022e-05, "loss": 0.0236, "step": 42830 }, { "grad_norm": 0.15610279142856598, "learning_rate": 4.826627113478522e-05, "loss": 0.0355, "step": 42840 }, { "grad_norm": 0.20222289860248566, "learning_rate": 4.824561533957708e-05, "loss": 0.0191, "step": 42850 }, { "grad_norm": 0.3037698268890381, "learning_rate": 4.822495984414527e-05, "loss": 0.0216, "step": 42860 }, { "grad_norm": 0.30691540241241455, "learning_rate": 4.8204304652019304e-05, "loss": 0.0194, "step": 42870 }, { "grad_norm": 0.6101873517036438, "learning_rate": 4.8183649766728537e-05, "loss": 0.0216, "step": 42880 }, { "grad_norm": 0.45302489399909973, "learning_rate": 4.816299519180237e-05, "loss": 0.0236, "step": 42890 }, { "grad_norm": 0.1924523413181305, "learning_rate": 4.8142340930770066e-05, "loss": 0.0186, "step": 42900 }, { "grad_norm": 0.2280205488204956, "learning_rate": 4.81216869871609e-05, "loss": 0.0177, "step": 42910 }, { "grad_norm": 0.4106277525424957, "learning_rate": 4.8101033364504064e-05, "loss": 0.0195, "step": 42920 }, { "grad_norm": 0.31206801533699036, "learning_rate": 4.808038006632866e-05, "loss": 0.0289, "step": 42930 }, { "grad_norm": 0.4422774016857147, "learning_rate": 4.805972709616383e-05, "loss": 0.018, "step": 42940 }, { "grad_norm": 0.34962931275367737, "learning_rate": 4.8039074457538556e-05, "loss": 0.0173, "step": 42950 }, { "grad_norm": 0.2387305498123169, "learning_rate": 4.801842215398184e-05, "loss": 0.024, "step": 42960 }, { "grad_norm": 0.3148019313812256, "learning_rate": 4.799777018902256e-05, "loss": 0.0237, "step": 42970 }, { "grad_norm": 0.2685440182685852, "learning_rate": 4.797711856618961e-05, "loss": 0.0168, "step": 42980 }, { "grad_norm": 0.2715560495853424, "learning_rate": 4.7956467289011765e-05, "loss": 0.018, "step": 42990 }, { "grad_norm": 0.36013680696487427, "learning_rate": 4.793581636101778e-05, "loss": 0.0197, "step": 43000 }, { "grad_norm": 0.5862953662872314, "learning_rate": 4.7915165785736326e-05, "loss": 0.0206, "step": 43010 }, { "grad_norm": 0.32472512125968933, "learning_rate": 4.789451556669599e-05, "loss": 0.0185, "step": 43020 }, { "grad_norm": 0.3555588722229004, "learning_rate": 4.7873865707425384e-05, "loss": 0.023, "step": 43030 }, { "grad_norm": 0.25757426023483276, "learning_rate": 4.7853216211452975e-05, "loss": 0.0181, "step": 43040 }, { "grad_norm": 0.3531554639339447, "learning_rate": 4.78325670823072e-05, "loss": 0.0235, "step": 43050 }, { "grad_norm": 0.24965186417102814, "learning_rate": 4.781191832351641e-05, "loss": 0.0228, "step": 43060 }, { "grad_norm": 0.38360291719436646, "learning_rate": 4.7791269938608965e-05, "loss": 0.0205, "step": 43070 }, { "grad_norm": 0.375491201877594, "learning_rate": 4.777062193111305e-05, "loss": 0.021, "step": 43080 }, { "grad_norm": 0.22829224169254303, "learning_rate": 4.77499743045569e-05, "loss": 0.0208, "step": 43090 }, { "grad_norm": 0.2687884271144867, "learning_rate": 4.7729327062468596e-05, "loss": 0.0215, "step": 43100 }, { "grad_norm": 0.39121100306510925, "learning_rate": 4.770868020837619e-05, "loss": 0.0241, "step": 43110 }, { "grad_norm": 0.22696757316589355, "learning_rate": 4.768803374580768e-05, "loss": 0.0166, "step": 43120 }, { "grad_norm": 0.25989973545074463, "learning_rate": 4.766738767829094e-05, "loss": 0.0169, "step": 43130 }, { "grad_norm": 0.26443129777908325, "learning_rate": 4.764674200935388e-05, "loss": 0.0246, "step": 43140 }, { "grad_norm": 0.4056888818740845, "learning_rate": 4.762609674252424e-05, "loss": 0.0198, "step": 43150 }, { "grad_norm": 0.33976688981056213, "learning_rate": 4.760545188132974e-05, "loss": 0.0358, "step": 43160 }, { "grad_norm": 0.2750459909439087, "learning_rate": 4.758480742929801e-05, "loss": 0.0202, "step": 43170 }, { "grad_norm": 0.5517702102661133, "learning_rate": 4.756416338995664e-05, "loss": 0.026, "step": 43180 }, { "grad_norm": 0.36561834812164307, "learning_rate": 4.7543519766833145e-05, "loss": 0.0172, "step": 43190 }, { "grad_norm": 0.19716405868530273, "learning_rate": 4.752287656345492e-05, "loss": 0.0193, "step": 43200 }, { "grad_norm": 0.24023893475532532, "learning_rate": 4.7502233783349355e-05, "loss": 0.026, "step": 43210 }, { "grad_norm": 0.23877669870853424, "learning_rate": 4.7481591430043694e-05, "loss": 0.0151, "step": 43220 }, { "grad_norm": 0.26279139518737793, "learning_rate": 4.746094950706521e-05, "loss": 0.0243, "step": 43230 }, { "grad_norm": 0.2904466390609741, "learning_rate": 4.744030801794099e-05, "loss": 0.0226, "step": 43240 }, { "grad_norm": 0.26670539379119873, "learning_rate": 4.741966696619813e-05, "loss": 0.0253, "step": 43250 }, { "grad_norm": 0.21368533372879028, "learning_rate": 4.739902635536361e-05, "loss": 0.029, "step": 43260 }, { "grad_norm": 0.2674572169780731, "learning_rate": 4.737838618896436e-05, "loss": 0.0214, "step": 43270 }, { "grad_norm": 0.312661737203598, "learning_rate": 4.7357746470527203e-05, "loss": 0.0236, "step": 43280 }, { "grad_norm": 0.3791695535182953, "learning_rate": 4.73371072035789e-05, "loss": 0.0268, "step": 43290 }, { "grad_norm": 0.5440793633460999, "learning_rate": 4.731646839164616e-05, "loss": 0.0245, "step": 43300 }, { "grad_norm": 0.30164623260498047, "learning_rate": 4.729583003825556e-05, "loss": 0.0219, "step": 43310 }, { "grad_norm": 0.29082781076431274, "learning_rate": 4.727519214693367e-05, "loss": 0.0239, "step": 43320 }, { "grad_norm": 0.312008798122406, "learning_rate": 4.725455472120689e-05, "loss": 0.0224, "step": 43330 }, { "grad_norm": 0.5982043147087097, "learning_rate": 4.723391776460164e-05, "loss": 0.0248, "step": 43340 }, { "grad_norm": 0.6400919556617737, "learning_rate": 4.7213281280644186e-05, "loss": 0.0252, "step": 43350 }, { "grad_norm": 0.24386925995349884, "learning_rate": 4.719264527286075e-05, "loss": 0.0262, "step": 43360 }, { "grad_norm": 0.4077545702457428, "learning_rate": 4.717200974477744e-05, "loss": 0.0254, "step": 43370 }, { "grad_norm": 0.2625100910663605, "learning_rate": 4.715137469992034e-05, "loss": 0.0301, "step": 43380 }, { "grad_norm": 0.3908352553844452, "learning_rate": 4.713074014181539e-05, "loss": 0.0232, "step": 43390 }, { "grad_norm": 0.23426665365695953, "learning_rate": 4.711010607398846e-05, "loss": 0.018, "step": 43400 }, { "grad_norm": 0.37307918071746826, "learning_rate": 4.708947249996538e-05, "loss": 0.0195, "step": 43410 }, { "grad_norm": 0.2823632061481476, "learning_rate": 4.706883942327183e-05, "loss": 0.0169, "step": 43420 }, { "grad_norm": 0.3811284005641937, "learning_rate": 4.704820684743347e-05, "loss": 0.0227, "step": 43430 }, { "grad_norm": 0.28727737069129944, "learning_rate": 4.702757477597581e-05, "loss": 0.0209, "step": 43440 }, { "grad_norm": 0.2876661717891693, "learning_rate": 4.700694321242435e-05, "loss": 0.0203, "step": 43450 }, { "grad_norm": 0.3973812162876129, "learning_rate": 4.698631216030441e-05, "loss": 0.0227, "step": 43460 }, { "grad_norm": 0.4867991507053375, "learning_rate": 4.6965681623141314e-05, "loss": 0.0178, "step": 43470 }, { "grad_norm": 0.28719955682754517, "learning_rate": 4.694505160446024e-05, "loss": 0.0155, "step": 43480 }, { "grad_norm": 0.21646307408809662, "learning_rate": 4.692442210778628e-05, "loss": 0.0207, "step": 43490 }, { "grad_norm": 0.367446631193161, "learning_rate": 4.690379313664448e-05, "loss": 0.0213, "step": 43500 }, { "grad_norm": 0.2753453850746155, "learning_rate": 4.688316469455973e-05, "loss": 0.0199, "step": 43510 }, { "grad_norm": 0.24925824999809265, "learning_rate": 4.6862536785056914e-05, "loss": 0.0213, "step": 43520 }, { "grad_norm": 0.4610264003276825, "learning_rate": 4.684190941166073e-05, "loss": 0.0226, "step": 43530 }, { "grad_norm": 0.29073193669319153, "learning_rate": 4.682128257789587e-05, "loss": 0.0189, "step": 43540 }, { "grad_norm": 0.7998402118682861, "learning_rate": 4.680065628728687e-05, "loss": 0.0224, "step": 43550 }, { "grad_norm": 0.27422964572906494, "learning_rate": 4.678003054335822e-05, "loss": 0.0246, "step": 43560 }, { "grad_norm": 0.34418994188308716, "learning_rate": 4.675940534963428e-05, "loss": 0.0271, "step": 43570 }, { "grad_norm": 0.23165322840213776, "learning_rate": 4.673878070963931e-05, "loss": 0.0212, "step": 43580 }, { "grad_norm": 0.26530852913856506, "learning_rate": 4.671815662689756e-05, "loss": 0.0188, "step": 43590 }, { "grad_norm": 0.25116166472435, "learning_rate": 4.669753310493306e-05, "loss": 0.0203, "step": 43600 }, { "grad_norm": 0.20516252517700195, "learning_rate": 4.6676910147269845e-05, "loss": 0.0238, "step": 43610 }, { "grad_norm": 0.38953471183776855, "learning_rate": 4.665628775743177e-05, "loss": 0.0223, "step": 43620 }, { "grad_norm": 0.21342596411705017, "learning_rate": 4.66356659389427e-05, "loss": 0.0225, "step": 43630 }, { "grad_norm": 0.2721222937107086, "learning_rate": 4.661504469532629e-05, "loss": 0.0174, "step": 43640 }, { "grad_norm": 0.298902302980423, "learning_rate": 4.659442403010617e-05, "loss": 0.0184, "step": 43650 }, { "grad_norm": 0.22362959384918213, "learning_rate": 4.6573803946805845e-05, "loss": 0.0232, "step": 43660 }, { "grad_norm": 0.18838399648666382, "learning_rate": 4.6553184448948695e-05, "loss": 0.0244, "step": 43670 }, { "grad_norm": 0.2543455958366394, "learning_rate": 4.653256554005807e-05, "loss": 0.0177, "step": 43680 }, { "grad_norm": 0.2190009504556656, "learning_rate": 4.6511947223657145e-05, "loss": 0.0213, "step": 43690 }, { "grad_norm": 0.4133666753768921, "learning_rate": 4.649132950326906e-05, "loss": 0.028, "step": 43700 }, { "grad_norm": 0.37282595038414, "learning_rate": 4.647071238241679e-05, "loss": 0.0258, "step": 43710 }, { "grad_norm": 0.20719292759895325, "learning_rate": 4.6450095864623264e-05, "loss": 0.0168, "step": 43720 }, { "grad_norm": 0.3122633099555969, "learning_rate": 4.642947995341123e-05, "loss": 0.0148, "step": 43730 }, { "grad_norm": 0.19683079421520233, "learning_rate": 4.640886465230345e-05, "loss": 0.0223, "step": 43740 }, { "grad_norm": 0.22152851521968842, "learning_rate": 4.6388249964822485e-05, "loss": 0.0234, "step": 43750 }, { "grad_norm": 0.285712331533432, "learning_rate": 4.63676358944908e-05, "loss": 0.0178, "step": 43760 }, { "grad_norm": 0.39089199900627136, "learning_rate": 4.634702244483081e-05, "loss": 0.0253, "step": 43770 }, { "grad_norm": 0.20347166061401367, "learning_rate": 4.632640961936475e-05, "loss": 0.0187, "step": 43780 }, { "grad_norm": 0.2644267678260803, "learning_rate": 4.6305797421614835e-05, "loss": 0.0177, "step": 43790 }, { "grad_norm": 0.477522611618042, "learning_rate": 4.6285185855103105e-05, "loss": 0.0249, "step": 43800 }, { "grad_norm": 0.30665352940559387, "learning_rate": 4.626457492335151e-05, "loss": 0.0202, "step": 43810 }, { "grad_norm": 0.26706451177597046, "learning_rate": 4.624396462988188e-05, "loss": 0.0183, "step": 43820 }, { "grad_norm": 0.2746810019016266, "learning_rate": 4.6223354978215985e-05, "loss": 0.0206, "step": 43830 }, { "grad_norm": 0.25998419523239136, "learning_rate": 4.620274597187544e-05, "loss": 0.0237, "step": 43840 }, { "grad_norm": 0.24654433131217957, "learning_rate": 4.6182137614381736e-05, "loss": 0.0194, "step": 43850 }, { "grad_norm": 0.2629770338535309, "learning_rate": 4.616152990925631e-05, "loss": 0.0189, "step": 43860 }, { "grad_norm": 0.27584490180015564, "learning_rate": 4.614092286002041e-05, "loss": 0.0257, "step": 43870 }, { "grad_norm": 0.288711816072464, "learning_rate": 4.612031647019528e-05, "loss": 0.0299, "step": 43880 }, { "grad_norm": 0.5282553434371948, "learning_rate": 4.609971074330193e-05, "loss": 0.0265, "step": 43890 }, { "grad_norm": 0.38298946619033813, "learning_rate": 4.607910568286136e-05, "loss": 0.0219, "step": 43900 }, { "grad_norm": 0.3667490482330322, "learning_rate": 4.605850129239438e-05, "loss": 0.0258, "step": 43910 }, { "grad_norm": 0.27878567576408386, "learning_rate": 4.603789757542174e-05, "loss": 0.0204, "step": 43920 }, { "grad_norm": 0.38847851753234863, "learning_rate": 4.601729453546404e-05, "loss": 0.022, "step": 43930 }, { "grad_norm": 0.3603787422180176, "learning_rate": 4.599669217604177e-05, "loss": 0.0283, "step": 43940 }, { "grad_norm": 0.5205660462379456, "learning_rate": 4.597609050067532e-05, "loss": 0.0286, "step": 43950 }, { "grad_norm": 0.27681487798690796, "learning_rate": 4.595548951288495e-05, "loss": 0.0246, "step": 43960 }, { "grad_norm": 0.2738281786441803, "learning_rate": 4.593488921619081e-05, "loss": 0.0243, "step": 43970 }, { "grad_norm": 0.3558870851993561, "learning_rate": 4.591428961411289e-05, "loss": 0.0309, "step": 43980 }, { "grad_norm": 0.32474374771118164, "learning_rate": 4.589369071017117e-05, "loss": 0.0266, "step": 43990 }, { "grad_norm": 0.1880248337984085, "learning_rate": 4.587309250788538e-05, "loss": 0.0242, "step": 44000 }, { "grad_norm": 0.19201672077178955, "learning_rate": 4.585249501077522e-05, "loss": 0.0192, "step": 44010 }, { "grad_norm": 0.31432613730430603, "learning_rate": 4.583189822236021e-05, "loss": 0.0237, "step": 44020 }, { "grad_norm": 0.3824200928211212, "learning_rate": 4.5811302146159816e-05, "loss": 0.0189, "step": 44030 }, { "grad_norm": 0.24471192061901093, "learning_rate": 4.579070678569332e-05, "loss": 0.0174, "step": 44040 }, { "grad_norm": 0.3519992232322693, "learning_rate": 4.5770112144479904e-05, "loss": 0.0233, "step": 44050 }, { "grad_norm": 0.33455148339271545, "learning_rate": 4.5749518226038645e-05, "loss": 0.0201, "step": 44060 }, { "grad_norm": 0.3199019134044647, "learning_rate": 4.572892503388845e-05, "loss": 0.0196, "step": 44070 }, { "grad_norm": 0.25867530703544617, "learning_rate": 4.570833257154817e-05, "loss": 0.0206, "step": 44080 }, { "grad_norm": 0.2659342586994171, "learning_rate": 4.568774084253646e-05, "loss": 0.0268, "step": 44090 }, { "grad_norm": 0.24806153774261475, "learning_rate": 4.566714985037191e-05, "loss": 0.0278, "step": 44100 }, { "grad_norm": 0.37023699283599854, "learning_rate": 4.564655959857295e-05, "loss": 0.0257, "step": 44110 }, { "grad_norm": 0.2810904383659363, "learning_rate": 4.5625970090657893e-05, "loss": 0.0225, "step": 44120 }, { "grad_norm": 0.2535261809825897, "learning_rate": 4.560538133014491e-05, "loss": 0.0201, "step": 44130 }, { "grad_norm": 0.3695988655090332, "learning_rate": 4.5584793320552055e-05, "loss": 0.02, "step": 44140 }, { "grad_norm": 0.23503093421459198, "learning_rate": 4.556420606539728e-05, "loss": 0.0217, "step": 44150 }, { "grad_norm": 0.20721597969532013, "learning_rate": 4.554361956819836e-05, "loss": 0.0183, "step": 44160 }, { "grad_norm": 0.20004649460315704, "learning_rate": 4.552303383247299e-05, "loss": 0.019, "step": 44170 }, { "grad_norm": 0.20225726068019867, "learning_rate": 4.5502448861738664e-05, "loss": 0.02, "step": 44180 }, { "grad_norm": 0.255569726228714, "learning_rate": 4.548186465951284e-05, "loss": 0.0283, "step": 44190 }, { "grad_norm": 0.42725279927253723, "learning_rate": 4.546128122931277e-05, "loss": 0.0277, "step": 44200 }, { "grad_norm": 0.2750709056854248, "learning_rate": 4.544069857465561e-05, "loss": 0.0238, "step": 44210 }, { "grad_norm": 0.6583999991416931, "learning_rate": 4.542011669905837e-05, "loss": 0.0203, "step": 44220 }, { "grad_norm": 0.3398890495300293, "learning_rate": 4.53995356060379e-05, "loss": 0.0162, "step": 44230 }, { "grad_norm": 0.25720319151878357, "learning_rate": 4.5378955299110994e-05, "loss": 0.0181, "step": 44240 }, { "grad_norm": 0.37559348344802856, "learning_rate": 4.5358375781794233e-05, "loss": 0.0263, "step": 44250 }, { "grad_norm": 0.34240543842315674, "learning_rate": 4.533779705760411e-05, "loss": 0.025, "step": 44260 }, { "grad_norm": 0.38395148515701294, "learning_rate": 4.5317219130056934e-05, "loss": 0.0193, "step": 44270 }, { "grad_norm": 0.27132534980773926, "learning_rate": 4.5296642002668964e-05, "loss": 0.0206, "step": 44280 }, { "grad_norm": 0.4207056164741516, "learning_rate": 4.5276065678956216e-05, "loss": 0.0268, "step": 44290 }, { "grad_norm": 0.26545393466949463, "learning_rate": 4.525549016243466e-05, "loss": 0.0208, "step": 44300 }, { "grad_norm": 0.23701097071170807, "learning_rate": 4.523491545662008e-05, "loss": 0.0208, "step": 44310 }, { "grad_norm": 0.2336420863866806, "learning_rate": 4.5214341565028104e-05, "loss": 0.0274, "step": 44320 }, { "grad_norm": 0.33924469351768494, "learning_rate": 4.519376849117428e-05, "loss": 0.0213, "step": 44330 }, { "grad_norm": 0.21405890583992004, "learning_rate": 4.517319623857395e-05, "loss": 0.026, "step": 44340 }, { "grad_norm": 0.3670746982097626, "learning_rate": 4.5152624810742394e-05, "loss": 0.0224, "step": 44350 }, { "grad_norm": 0.2619069218635559, "learning_rate": 4.5132054211194664e-05, "loss": 0.019, "step": 44360 }, { "grad_norm": 0.40258753299713135, "learning_rate": 4.511148444344574e-05, "loss": 0.0211, "step": 44370 }, { "grad_norm": 0.3214767873287201, "learning_rate": 4.509091551101041e-05, "loss": 0.0258, "step": 44380 }, { "grad_norm": 0.2396133542060852, "learning_rate": 4.507034741740338e-05, "loss": 0.0206, "step": 44390 }, { "grad_norm": 0.4132554233074188, "learning_rate": 4.5049780166139145e-05, "loss": 0.0223, "step": 44400 }, { "grad_norm": 0.34816011786460876, "learning_rate": 4.5029213760732075e-05, "loss": 0.0305, "step": 44410 }, { "grad_norm": 0.27368152141571045, "learning_rate": 4.5008648204696434e-05, "loss": 0.0291, "step": 44420 }, { "grad_norm": 0.42892909049987793, "learning_rate": 4.4988083501546284e-05, "loss": 0.0183, "step": 44430 }, { "grad_norm": 0.26403701305389404, "learning_rate": 4.4967519654795606e-05, "loss": 0.016, "step": 44440 }, { "grad_norm": 0.2594715654850006, "learning_rate": 4.494695666795816e-05, "loss": 0.0168, "step": 44450 }, { "grad_norm": 0.25855186581611633, "learning_rate": 4.4926394544547644e-05, "loss": 0.0189, "step": 44460 }, { "grad_norm": 0.22045302391052246, "learning_rate": 4.490583328807752e-05, "loss": 0.023, "step": 44470 }, { "grad_norm": 0.3085889220237732, "learning_rate": 4.488527290206117e-05, "loss": 0.0231, "step": 44480 }, { "grad_norm": 0.7308365702629089, "learning_rate": 4.48647133900118e-05, "loss": 0.0189, "step": 44490 }, { "grad_norm": 0.23395957052707672, "learning_rate": 4.484415475544244e-05, "loss": 0.0232, "step": 44500 }, { "grad_norm": 0.2364986687898636, "learning_rate": 4.4823597001866035e-05, "loss": 0.0169, "step": 44510 }, { "grad_norm": 0.24600926041603088, "learning_rate": 4.480304013279532e-05, "loss": 0.021, "step": 44520 }, { "grad_norm": 0.47231701016426086, "learning_rate": 4.478248415174292e-05, "loss": 0.0173, "step": 44530 }, { "grad_norm": 0.2782305181026459, "learning_rate": 4.476192906222126e-05, "loss": 0.0263, "step": 44540 }, { "grad_norm": 0.29297685623168945, "learning_rate": 4.474137486774268e-05, "loss": 0.0321, "step": 44550 }, { "grad_norm": 0.34011876583099365, "learning_rate": 4.4720821571819296e-05, "loss": 0.0221, "step": 44560 }, { "grad_norm": 0.2677392065525055, "learning_rate": 4.470026917796314e-05, "loss": 0.0301, "step": 44570 }, { "grad_norm": 0.30210307240486145, "learning_rate": 4.4679717689686005e-05, "loss": 0.0383, "step": 44580 }, { "grad_norm": 0.2662174105644226, "learning_rate": 4.465916711049963e-05, "loss": 0.0187, "step": 44590 }, { "grad_norm": 0.3582603335380554, "learning_rate": 4.4638617443915524e-05, "loss": 0.0236, "step": 44600 }, { "grad_norm": 0.3478166460990906, "learning_rate": 4.4618068693445055e-05, "loss": 0.0245, "step": 44610 }, { "grad_norm": 0.2031162530183792, "learning_rate": 4.459752086259946e-05, "loss": 0.0176, "step": 44620 }, { "grad_norm": 0.411917120218277, "learning_rate": 4.457697395488977e-05, "loss": 0.0328, "step": 44630 }, { "grad_norm": 0.2689400017261505, "learning_rate": 4.455642797382693e-05, "loss": 0.0277, "step": 44640 }, { "grad_norm": 0.2895258069038391, "learning_rate": 4.4535882922921643e-05, "loss": 0.0284, "step": 44650 }, { "grad_norm": 0.3457315266132355, "learning_rate": 4.451533880568455e-05, "loss": 0.0209, "step": 44660 }, { "grad_norm": 0.23348718881607056, "learning_rate": 4.449479562562603e-05, "loss": 0.0255, "step": 44670 }, { "grad_norm": 0.35529279708862305, "learning_rate": 4.447425338625639e-05, "loss": 0.0262, "step": 44680 }, { "grad_norm": 0.26850366592407227, "learning_rate": 4.4453712091085705e-05, "loss": 0.0223, "step": 44690 }, { "grad_norm": 0.3909880220890045, "learning_rate": 4.443317174362392e-05, "loss": 0.024, "step": 44700 }, { "grad_norm": 0.47635796666145325, "learning_rate": 4.441263234738085e-05, "loss": 0.0239, "step": 44710 }, { "grad_norm": 0.25823870301246643, "learning_rate": 4.4392093905866086e-05, "loss": 0.0137, "step": 44720 }, { "grad_norm": 0.35211730003356934, "learning_rate": 4.437155642258911e-05, "loss": 0.0186, "step": 44730 }, { "grad_norm": 0.2173742949962616, "learning_rate": 4.4351019901059177e-05, "loss": 0.0269, "step": 44740 }, { "grad_norm": 0.2833782732486725, "learning_rate": 4.433048434478547e-05, "loss": 0.0331, "step": 44750 }, { "grad_norm": 0.34741345047950745, "learning_rate": 4.4309949757276906e-05, "loss": 0.0247, "step": 44760 }, { "grad_norm": 0.26578158140182495, "learning_rate": 4.428941614204233e-05, "loss": 0.025, "step": 44770 }, { "grad_norm": 0.2797262668609619, "learning_rate": 4.426888350259034e-05, "loss": 0.0225, "step": 44780 }, { "grad_norm": 0.29644981026649475, "learning_rate": 4.424835184242939e-05, "loss": 0.033, "step": 44790 }, { "grad_norm": 0.16671037673950195, "learning_rate": 4.422782116506784e-05, "loss": 0.0182, "step": 44800 }, { "grad_norm": 0.4127904176712036, "learning_rate": 4.4207291474013754e-05, "loss": 0.0241, "step": 44810 }, { "grad_norm": 0.3140081763267517, "learning_rate": 4.418676277277514e-05, "loss": 0.0286, "step": 44820 }, { "grad_norm": 0.22741374373435974, "learning_rate": 4.416623506485976e-05, "loss": 0.0269, "step": 44830 }, { "grad_norm": 0.2620793581008911, "learning_rate": 4.414570835377527e-05, "loss": 0.0246, "step": 44840 }, { "grad_norm": 0.7859954833984375, "learning_rate": 4.412518264302909e-05, "loss": 0.0243, "step": 44850 }, { "grad_norm": 0.2884601354598999, "learning_rate": 4.4104657936128535e-05, "loss": 0.0231, "step": 44860 }, { "grad_norm": 0.3564537465572357, "learning_rate": 4.408413423658071e-05, "loss": 0.0245, "step": 44870 }, { "grad_norm": 0.26121559739112854, "learning_rate": 4.406361154789253e-05, "loss": 0.0233, "step": 44880 }, { "grad_norm": 0.3479999303817749, "learning_rate": 4.4043089873570776e-05, "loss": 0.0227, "step": 44890 }, { "grad_norm": 0.31391268968582153, "learning_rate": 4.4022569217122025e-05, "loss": 0.0178, "step": 44900 }, { "grad_norm": 0.21201249957084656, "learning_rate": 4.400204958205274e-05, "loss": 0.0195, "step": 44910 }, { "grad_norm": 0.349455326795578, "learning_rate": 4.3981530971869125e-05, "loss": 0.0287, "step": 44920 }, { "grad_norm": 0.3077840507030487, "learning_rate": 4.396101339007727e-05, "loss": 0.0168, "step": 44930 }, { "grad_norm": 1.6394500732421875, "learning_rate": 4.394049684018304e-05, "loss": 0.0258, "step": 44940 }, { "grad_norm": 0.4020657241344452, "learning_rate": 4.3919981325692186e-05, "loss": 0.0223, "step": 44950 }, { "grad_norm": 0.17111052572727203, "learning_rate": 4.389946685011024e-05, "loss": 0.0193, "step": 44960 }, { "grad_norm": 0.22200648486614227, "learning_rate": 4.387895341694255e-05, "loss": 0.0254, "step": 44970 }, { "grad_norm": 0.20934638381004333, "learning_rate": 4.3858441029694324e-05, "loss": 0.0271, "step": 44980 }, { "grad_norm": 0.3334435820579529, "learning_rate": 4.3837929691870527e-05, "loss": 0.0198, "step": 44990 }, { "grad_norm": 0.2975621223449707, "learning_rate": 4.381741940697604e-05, "loss": 0.0236, "step": 45000 }, { "grad_norm": 0.2990640103816986, "learning_rate": 4.379691017851547e-05, "loss": 0.0204, "step": 45010 }, { "grad_norm": 0.2992507219314575, "learning_rate": 4.3776402009993304e-05, "loss": 0.0198, "step": 45020 }, { "grad_norm": 0.2290385514497757, "learning_rate": 4.3755894904913794e-05, "loss": 0.0173, "step": 45030 }, { "grad_norm": 0.35541898012161255, "learning_rate": 4.373538886678109e-05, "loss": 0.02, "step": 45040 }, { "grad_norm": 0.2328617423772812, "learning_rate": 4.371488389909909e-05, "loss": 0.0229, "step": 45050 }, { "grad_norm": 0.26240381598472595, "learning_rate": 4.3694380005371515e-05, "loss": 0.0221, "step": 45060 }, { "grad_norm": 0.21723374724388123, "learning_rate": 4.367387718910196e-05, "loss": 0.0188, "step": 45070 }, { "grad_norm": 0.2822559177875519, "learning_rate": 4.3653375453793764e-05, "loss": 0.021, "step": 45080 }, { "grad_norm": 0.2891218066215515, "learning_rate": 4.3632874802950136e-05, "loss": 0.0209, "step": 45090 }, { "grad_norm": 0.4831196069717407, "learning_rate": 4.3612375240074034e-05, "loss": 0.0237, "step": 45100 }, { "grad_norm": 0.20533454418182373, "learning_rate": 4.3591876768668325e-05, "loss": 0.0221, "step": 45110 }, { "grad_norm": 0.22085461020469666, "learning_rate": 4.3571379392235605e-05, "loss": 0.0245, "step": 45120 }, { "grad_norm": 0.3001624345779419, "learning_rate": 4.3550883114278335e-05, "loss": 0.0229, "step": 45130 }, { "grad_norm": 0.2699243724346161, "learning_rate": 4.353038793829876e-05, "loss": 0.0213, "step": 45140 }, { "grad_norm": 0.2760281562805176, "learning_rate": 4.350989386779891e-05, "loss": 0.0168, "step": 45150 }, { "grad_norm": 0.42763128876686096, "learning_rate": 4.3489400906280724e-05, "loss": 0.031, "step": 45160 }, { "grad_norm": 0.2534681558609009, "learning_rate": 4.3468909057245845e-05, "loss": 0.0246, "step": 45170 }, { "grad_norm": 0.21397413313388824, "learning_rate": 4.3448418324195794e-05, "loss": 0.0223, "step": 45180 }, { "grad_norm": 0.5204465389251709, "learning_rate": 4.342792871063184e-05, "loss": 0.0198, "step": 45190 }, { "grad_norm": 0.46713608503341675, "learning_rate": 4.3407440220055145e-05, "loss": 0.0197, "step": 45200 }, { "grad_norm": 0.2365647554397583, "learning_rate": 4.33869528559666e-05, "loss": 0.0171, "step": 45210 }, { "grad_norm": 0.24847851693630219, "learning_rate": 4.336646662186696e-05, "loss": 0.0138, "step": 45220 }, { "grad_norm": 0.2417992502450943, "learning_rate": 4.334598152125672e-05, "loss": 0.0219, "step": 45230 }, { "grad_norm": 0.3157331049442291, "learning_rate": 4.3325497557636276e-05, "loss": 0.0237, "step": 45240 }, { "grad_norm": 0.1986982375383377, "learning_rate": 4.330501473450574e-05, "loss": 0.0248, "step": 45250 }, { "grad_norm": 0.2800470292568207, "learning_rate": 4.328453305536507e-05, "loss": 0.0194, "step": 45260 }, { "grad_norm": 0.4082741141319275, "learning_rate": 4.326405252371404e-05, "loss": 0.0257, "step": 45270 }, { "grad_norm": 0.35334816575050354, "learning_rate": 4.324357314305221e-05, "loss": 0.0218, "step": 45280 }, { "grad_norm": 0.28081685304641724, "learning_rate": 4.3223094916878945e-05, "loss": 0.0228, "step": 45290 }, { "grad_norm": 0.25371021032333374, "learning_rate": 4.320261784869338e-05, "loss": 0.0315, "step": 45300 }, { "grad_norm": 0.16502074897289276, "learning_rate": 4.318214194199455e-05, "loss": 0.0211, "step": 45310 }, { "grad_norm": 0.144168421626091, "learning_rate": 4.316166720028118e-05, "loss": 0.0148, "step": 45320 }, { "grad_norm": 0.31095585227012634, "learning_rate": 4.3141193627051864e-05, "loss": 0.0182, "step": 45330 }, { "grad_norm": 0.3201296627521515, "learning_rate": 4.312072122580496e-05, "loss": 0.0219, "step": 45340 }, { "grad_norm": 0.39565393328666687, "learning_rate": 4.3100250000038646e-05, "loss": 0.0204, "step": 45350 }, { "grad_norm": 0.4167778491973877, "learning_rate": 4.307977995325091e-05, "loss": 0.0358, "step": 45360 }, { "grad_norm": 0.28926321864128113, "learning_rate": 4.30593110889395e-05, "loss": 0.0242, "step": 45370 }, { "grad_norm": 0.40548044443130493, "learning_rate": 4.3038843410602016e-05, "loss": 0.0154, "step": 45380 }, { "grad_norm": 0.16404497623443604, "learning_rate": 4.3018376921735774e-05, "loss": 0.0222, "step": 45390 }, { "grad_norm": 0.15917879343032837, "learning_rate": 4.299791162583799e-05, "loss": 0.0197, "step": 45400 }, { "grad_norm": 0.17568564414978027, "learning_rate": 4.29774475264056e-05, "loss": 0.0252, "step": 45410 }, { "grad_norm": 0.2128058820962906, "learning_rate": 4.2956984626935365e-05, "loss": 0.0187, "step": 45420 }, { "grad_norm": 0.264424592256546, "learning_rate": 4.293652293092383e-05, "loss": 0.0234, "step": 45430 }, { "grad_norm": 0.2555736303329468, "learning_rate": 4.2916062441867324e-05, "loss": 0.0193, "step": 45440 }, { "grad_norm": 0.2963053584098816, "learning_rate": 4.289560316326201e-05, "loss": 0.0282, "step": 45450 }, { "grad_norm": 0.4041806161403656, "learning_rate": 4.28751450986038e-05, "loss": 0.0245, "step": 45460 }, { "grad_norm": 0.4619394838809967, "learning_rate": 4.2854688251388444e-05, "loss": 0.0257, "step": 45470 }, { "grad_norm": 0.21586668491363525, "learning_rate": 4.2834232625111425e-05, "loss": 0.0239, "step": 45480 }, { "grad_norm": 1.1635167598724365, "learning_rate": 4.2813778223268086e-05, "loss": 0.0211, "step": 45490 }, { "grad_norm": 0.932313084602356, "learning_rate": 4.2793325049353477e-05, "loss": 0.018, "step": 45500 }, { "grad_norm": 0.2724444270133972, "learning_rate": 4.2772873106862535e-05, "loss": 0.0212, "step": 45510 }, { "grad_norm": 0.23914365470409393, "learning_rate": 4.275242239928993e-05, "loss": 0.0229, "step": 45520 }, { "grad_norm": 0.35565999150276184, "learning_rate": 4.273197293013009e-05, "loss": 0.0193, "step": 45530 }, { "grad_norm": 0.2237456887960434, "learning_rate": 4.271152470287731e-05, "loss": 0.017, "step": 45540 }, { "grad_norm": 0.19866742193698883, "learning_rate": 4.2691077721025594e-05, "loss": 0.0202, "step": 45550 }, { "grad_norm": 0.22897054255008698, "learning_rate": 4.267063198806883e-05, "loss": 0.0207, "step": 45560 }, { "grad_norm": 0.2538665235042572, "learning_rate": 4.2650187507500574e-05, "loss": 0.0208, "step": 45570 }, { "grad_norm": 0.25133848190307617, "learning_rate": 4.2629744282814275e-05, "loss": 0.0132, "step": 45580 }, { "grad_norm": 0.22265547513961792, "learning_rate": 4.2609302317503074e-05, "loss": 0.0192, "step": 45590 }, { "grad_norm": 0.2354593724012375, "learning_rate": 4.258886161505999e-05, "loss": 0.0181, "step": 45600 }, { "grad_norm": 0.22232040762901306, "learning_rate": 4.2568422178977775e-05, "loss": 0.0191, "step": 45610 }, { "grad_norm": 0.26185041666030884, "learning_rate": 4.254798401274894e-05, "loss": 0.0244, "step": 45620 }, { "grad_norm": 0.32143375277519226, "learning_rate": 4.252754711986583e-05, "loss": 0.0214, "step": 45630 }, { "grad_norm": 0.27977561950683594, "learning_rate": 4.250711150382052e-05, "loss": 0.0174, "step": 45640 }, { "grad_norm": 0.4525716006755829, "learning_rate": 4.248667716810495e-05, "loss": 0.0187, "step": 45650 }, { "grad_norm": 0.8361020684242249, "learning_rate": 4.246624411621074e-05, "loss": 0.0173, "step": 45660 }, { "grad_norm": 0.3757472336292267, "learning_rate": 4.244581235162938e-05, "loss": 0.0237, "step": 45670 }, { "grad_norm": 0.16620223224163055, "learning_rate": 4.2425381877852075e-05, "loss": 0.0236, "step": 45680 }, { "grad_norm": 0.34847429394721985, "learning_rate": 4.2404952698369856e-05, "loss": 0.0192, "step": 45690 }, { "grad_norm": 0.2462521493434906, "learning_rate": 4.238452481667349e-05, "loss": 0.0226, "step": 45700 }, { "grad_norm": 0.3449648320674896, "learning_rate": 4.2364098236253526e-05, "loss": 0.026, "step": 45710 }, { "grad_norm": 0.2639150619506836, "learning_rate": 4.2343672960600356e-05, "loss": 0.0187, "step": 45720 }, { "grad_norm": 0.29871028661727905, "learning_rate": 4.232324899320406e-05, "loss": 0.0166, "step": 45730 }, { "grad_norm": 0.27540165185928345, "learning_rate": 4.230282633755457e-05, "loss": 0.0206, "step": 45740 }, { "grad_norm": 0.21913573145866394, "learning_rate": 4.2282404997141515e-05, "loss": 0.0169, "step": 45750 }, { "grad_norm": 0.25122934579849243, "learning_rate": 4.2261984975454397e-05, "loss": 0.0237, "step": 45760 }, { "grad_norm": 0.27632346749305725, "learning_rate": 4.224156627598239e-05, "loss": 0.0178, "step": 45770 }, { "grad_norm": 0.5166489481925964, "learning_rate": 4.222114890221453e-05, "loss": 0.0247, "step": 45780 }, { "grad_norm": 0.5618883967399597, "learning_rate": 4.2200732857639546e-05, "loss": 0.0251, "step": 45790 }, { "grad_norm": 0.24778640270233154, "learning_rate": 4.2180318145746035e-05, "loss": 0.0217, "step": 45800 }, { "grad_norm": 0.3254264295101166, "learning_rate": 4.215990477002227e-05, "loss": 0.0266, "step": 45810 }, { "grad_norm": 0.2681080996990204, "learning_rate": 4.2139492733956356e-05, "loss": 0.0179, "step": 45820 }, { "grad_norm": 0.3194403648376465, "learning_rate": 4.211908204103615e-05, "loss": 0.031, "step": 45830 }, { "grad_norm": 0.36175185441970825, "learning_rate": 4.2098672694749265e-05, "loss": 0.0346, "step": 45840 }, { "grad_norm": 0.5510208606719971, "learning_rate": 4.2078264698583133e-05, "loss": 0.0226, "step": 45850 }, { "grad_norm": 0.4914219379425049, "learning_rate": 4.205785805602488e-05, "loss": 0.0224, "step": 45860 }, { "grad_norm": 0.32354599237442017, "learning_rate": 4.203745277056149e-05, "loss": 0.0239, "step": 45870 }, { "grad_norm": 0.20150920748710632, "learning_rate": 4.201704884567964e-05, "loss": 0.0266, "step": 45880 }, { "grad_norm": 0.24924716353416443, "learning_rate": 4.1996646284865816e-05, "loss": 0.0317, "step": 45890 }, { "grad_norm": 0.38695213198661804, "learning_rate": 4.197624509160625e-05, "loss": 0.0199, "step": 45900 }, { "grad_norm": 0.19683751463890076, "learning_rate": 4.195584526938692e-05, "loss": 0.0167, "step": 45910 }, { "grad_norm": 0.32722750306129456, "learning_rate": 4.193544682169365e-05, "loss": 0.0214, "step": 45920 }, { "grad_norm": 0.30130305886268616, "learning_rate": 4.1915049752011946e-05, "loss": 0.017, "step": 45930 }, { "grad_norm": 0.22485753893852234, "learning_rate": 4.189465406382712e-05, "loss": 0.0178, "step": 45940 }, { "grad_norm": 0.23946532607078552, "learning_rate": 4.187425976062422e-05, "loss": 0.023, "step": 45950 }, { "grad_norm": 0.4656553566455841, "learning_rate": 4.18538668458881e-05, "loss": 0.023, "step": 45960 }, { "grad_norm": 0.2481619119644165, "learning_rate": 4.183347532310333e-05, "loss": 0.0186, "step": 45970 }, { "grad_norm": 0.20765671133995056, "learning_rate": 4.181308519575429e-05, "loss": 0.0193, "step": 45980 }, { "grad_norm": 0.24700908362865448, "learning_rate": 4.179269646732507e-05, "loss": 0.0303, "step": 45990 }, { "grad_norm": 0.38924962282180786, "learning_rate": 4.177230914129954e-05, "loss": 0.0175, "step": 46000 }, { "grad_norm": 0.2453765571117401, "learning_rate": 4.175192322116136e-05, "loss": 0.0174, "step": 46010 }, { "grad_norm": 0.23809324204921722, "learning_rate": 4.173153871039391e-05, "loss": 0.0158, "step": 46020 }, { "grad_norm": 0.3671773672103882, "learning_rate": 4.171115561248036e-05, "loss": 0.022, "step": 46030 }, { "grad_norm": 0.17082715034484863, "learning_rate": 4.16907739309036e-05, "loss": 0.0154, "step": 46040 }, { "grad_norm": 0.2690776586532593, "learning_rate": 4.167039366914633e-05, "loss": 0.0183, "step": 46050 }, { "grad_norm": 0.21032200753688812, "learning_rate": 4.165001483069096e-05, "loss": 0.0225, "step": 46060 }, { "grad_norm": 0.2759358584880829, "learning_rate": 4.16296374190197e-05, "loss": 0.0262, "step": 46070 }, { "grad_norm": 0.2142600417137146, "learning_rate": 4.1609261437614464e-05, "loss": 0.0195, "step": 46080 }, { "grad_norm": 0.4601636230945587, "learning_rate": 4.158888688995696e-05, "loss": 0.0245, "step": 46090 }, { "grad_norm": 0.45013371109962463, "learning_rate": 4.1568513779528645e-05, "loss": 0.0175, "step": 46100 }, { "grad_norm": 0.25749170780181885, "learning_rate": 4.1548142109810704e-05, "loss": 0.0268, "step": 46110 }, { "grad_norm": 0.15826591849327087, "learning_rate": 4.152777188428414e-05, "loss": 0.0174, "step": 46120 }, { "grad_norm": 0.22374962270259857, "learning_rate": 4.1507403106429646e-05, "loss": 0.0182, "step": 46130 }, { "grad_norm": 0.2514517605304718, "learning_rate": 4.148703577972768e-05, "loss": 0.0176, "step": 46140 }, { "grad_norm": 0.27235713601112366, "learning_rate": 4.146666990765846e-05, "loss": 0.0215, "step": 46150 }, { "grad_norm": 0.20457105338573456, "learning_rate": 4.1446305493701986e-05, "loss": 0.017, "step": 46160 }, { "grad_norm": 0.35331299901008606, "learning_rate": 4.142594254133796e-05, "loss": 0.0205, "step": 46170 }, { "grad_norm": 0.3609808087348938, "learning_rate": 4.1405581054045835e-05, "loss": 0.0211, "step": 46180 }, { "grad_norm": 0.3318350613117218, "learning_rate": 4.1385221035304864e-05, "loss": 0.0193, "step": 46190 }, { "grad_norm": 0.2605534791946411, "learning_rate": 4.136486248859398e-05, "loss": 0.0197, "step": 46200 }, { "grad_norm": 0.42724257707595825, "learning_rate": 4.1344505417391955e-05, "loss": 0.0202, "step": 46210 }, { "grad_norm": 0.2342974841594696, "learning_rate": 4.132414982517721e-05, "loss": 0.0185, "step": 46220 }, { "grad_norm": 0.1976027935743332, "learning_rate": 4.130379571542798e-05, "loss": 0.0197, "step": 46230 }, { "grad_norm": 0.20452074706554413, "learning_rate": 4.12834430916222e-05, "loss": 0.0197, "step": 46240 }, { "grad_norm": 0.3295089602470398, "learning_rate": 4.126309195723763e-05, "loss": 0.0177, "step": 46250 }, { "grad_norm": 0.250516802072525, "learning_rate": 4.124274231575168e-05, "loss": 0.0236, "step": 46260 }, { "grad_norm": 0.21263420581817627, "learning_rate": 4.122239417064154e-05, "loss": 0.015, "step": 46270 }, { "grad_norm": 0.16222919523715973, "learning_rate": 4.1202047525384184e-05, "loss": 0.0184, "step": 46280 }, { "grad_norm": 0.25525474548339844, "learning_rate": 4.118170238345627e-05, "loss": 0.0189, "step": 46290 }, { "grad_norm": 0.3445025682449341, "learning_rate": 4.1161358748334256e-05, "loss": 0.0239, "step": 46300 }, { "grad_norm": 0.2691080570220947, "learning_rate": 4.1141016623494266e-05, "loss": 0.0199, "step": 46310 }, { "grad_norm": 0.3693767189979553, "learning_rate": 4.112067601241227e-05, "loss": 0.018, "step": 46320 }, { "grad_norm": 0.227589413523674, "learning_rate": 4.110033691856387e-05, "loss": 0.0205, "step": 46330 }, { "grad_norm": 0.46016544103622437, "learning_rate": 4.107999934542451e-05, "loss": 0.0317, "step": 46340 }, { "grad_norm": 0.21809078752994537, "learning_rate": 4.105966329646928e-05, "loss": 0.0234, "step": 46350 }, { "grad_norm": 0.21713067591190338, "learning_rate": 4.103932877517308e-05, "loss": 0.0234, "step": 46360 }, { "grad_norm": 0.31849297881126404, "learning_rate": 4.101899578501052e-05, "loss": 0.0207, "step": 46370 }, { "grad_norm": 0.2624698579311371, "learning_rate": 4.099866432945595e-05, "loss": 0.0201, "step": 46380 }, { "grad_norm": 0.3453781008720398, "learning_rate": 4.097833441198346e-05, "loss": 0.0203, "step": 46390 }, { "grad_norm": 0.17390981316566467, "learning_rate": 4.095800603606685e-05, "loss": 0.0219, "step": 46400 }, { "grad_norm": 0.2694758474826813, "learning_rate": 4.093767920517975e-05, "loss": 0.021, "step": 46410 }, { "grad_norm": 0.26021113991737366, "learning_rate": 4.091735392279539e-05, "loss": 0.02, "step": 46420 }, { "grad_norm": 0.1638372838497162, "learning_rate": 4.089703019238685e-05, "loss": 0.0195, "step": 46430 }, { "grad_norm": 0.33449089527130127, "learning_rate": 4.0876708017426866e-05, "loss": 0.0176, "step": 46440 }, { "grad_norm": 0.1917426437139511, "learning_rate": 4.085638740138798e-05, "loss": 0.021, "step": 46450 }, { "grad_norm": 0.27020812034606934, "learning_rate": 4.08360683477424e-05, "loss": 0.0192, "step": 46460 }, { "grad_norm": 0.3380044102668762, "learning_rate": 4.0815750859962085e-05, "loss": 0.022, "step": 46470 }, { "grad_norm": 0.38105660676956177, "learning_rate": 4.079543494151879e-05, "loss": 0.0241, "step": 46480 }, { "grad_norm": 0.3164326846599579, "learning_rate": 4.07751205958839e-05, "loss": 0.0209, "step": 46490 }, { "grad_norm": 0.26675909757614136, "learning_rate": 4.0754807826528615e-05, "loss": 0.0169, "step": 46500 }, { "grad_norm": 0.2742231786251068, "learning_rate": 4.073449663692379e-05, "loss": 0.0233, "step": 46510 }, { "grad_norm": 0.3626425564289093, "learning_rate": 4.07141870305401e-05, "loss": 0.0173, "step": 46520 }, { "grad_norm": 0.35559025406837463, "learning_rate": 4.0693879010847866e-05, "loss": 0.03, "step": 46530 }, { "grad_norm": 0.2688805162906647, "learning_rate": 4.06735725813172e-05, "loss": 0.0199, "step": 46540 }, { "grad_norm": 0.5698767900466919, "learning_rate": 4.0653267745417903e-05, "loss": 0.0162, "step": 46550 }, { "grad_norm": 0.2135317325592041, "learning_rate": 4.063296450661949e-05, "loss": 0.0168, "step": 46560 }, { "grad_norm": 0.1485602706670761, "learning_rate": 4.061266286839128e-05, "loss": 0.0195, "step": 46570 }, { "grad_norm": 0.28357550501823425, "learning_rate": 4.0592362834202225e-05, "loss": 0.0257, "step": 46580 }, { "grad_norm": 0.2332375943660736, "learning_rate": 4.057206440752107e-05, "loss": 0.0208, "step": 46590 }, { "grad_norm": 0.2471853345632553, "learning_rate": 4.0551767591816245e-05, "loss": 0.0186, "step": 46600 }, { "grad_norm": 0.2375062108039856, "learning_rate": 4.0531472390555935e-05, "loss": 0.0245, "step": 46610 }, { "grad_norm": 0.30741819739341736, "learning_rate": 4.051117880720802e-05, "loss": 0.0176, "step": 46620 }, { "grad_norm": 0.3701116442680359, "learning_rate": 4.049088684524015e-05, "loss": 0.0172, "step": 46630 }, { "grad_norm": 0.2317676842212677, "learning_rate": 4.0470596508119636e-05, "loss": 0.0188, "step": 46640 }, { "grad_norm": 0.2280331701040268, "learning_rate": 4.0450307799313524e-05, "loss": 0.0221, "step": 46650 }, { "grad_norm": 0.329159140586853, "learning_rate": 4.0430020722288656e-05, "loss": 0.0231, "step": 46660 }, { "grad_norm": 0.27800601720809937, "learning_rate": 4.040973528051148e-05, "loss": 0.0255, "step": 46670 }, { "grad_norm": 0.20321530103683472, "learning_rate": 4.038945147744827e-05, "loss": 0.0234, "step": 46680 }, { "grad_norm": 0.24064508080482483, "learning_rate": 4.0369169316564945e-05, "loss": 0.0164, "step": 46690 }, { "grad_norm": 0.6345439553260803, "learning_rate": 4.034888880132718e-05, "loss": 0.0244, "step": 46700 }, { "grad_norm": 1.0819437503814697, "learning_rate": 4.032860993520035e-05, "loss": 0.0173, "step": 46710 }, { "grad_norm": 0.24897503852844238, "learning_rate": 4.030833272164959e-05, "loss": 0.0252, "step": 46720 }, { "grad_norm": 0.19294586777687073, "learning_rate": 4.0288057164139705e-05, "loss": 0.0233, "step": 46730 }, { "grad_norm": 0.19559761881828308, "learning_rate": 4.0267783266135205e-05, "loss": 0.0153, "step": 46740 }, { "grad_norm": 0.42388108372688293, "learning_rate": 4.024751103110039e-05, "loss": 0.0206, "step": 46750 }, { "grad_norm": 0.19406460225582123, "learning_rate": 4.0227240462499176e-05, "loss": 0.0215, "step": 46760 }, { "grad_norm": 0.1796404868364334, "learning_rate": 4.020697156379531e-05, "loss": 0.0228, "step": 46770 }, { "grad_norm": 0.2533823847770691, "learning_rate": 4.018670433845215e-05, "loss": 0.0228, "step": 46780 }, { "grad_norm": 0.1859637200832367, "learning_rate": 4.016643878993284e-05, "loss": 0.0213, "step": 46790 }, { "grad_norm": 0.3981045186519623, "learning_rate": 4.014617492170017e-05, "loss": 0.0222, "step": 46800 }, { "grad_norm": 0.25729331374168396, "learning_rate": 4.0125912737216726e-05, "loss": 0.0297, "step": 46810 }, { "grad_norm": 0.3109944760799408, "learning_rate": 4.0105652239944735e-05, "loss": 0.0219, "step": 46820 }, { "grad_norm": 0.21832279860973358, "learning_rate": 4.0085393433346144e-05, "loss": 0.0222, "step": 46830 }, { "grad_norm": 0.22013410925865173, "learning_rate": 4.006513632088268e-05, "loss": 0.0177, "step": 46840 }, { "grad_norm": 0.19798029959201813, "learning_rate": 4.004488090601567e-05, "loss": 0.0148, "step": 46850 }, { "grad_norm": 0.22614480555057526, "learning_rate": 4.002462719220626e-05, "loss": 0.0196, "step": 46860 }, { "grad_norm": 0.19707034528255463, "learning_rate": 4.000437518291522e-05, "loss": 0.0258, "step": 46870 }, { "grad_norm": 0.17292411625385284, "learning_rate": 3.9984124881603094e-05, "loss": 0.0201, "step": 46880 }, { "grad_norm": 0.36853697896003723, "learning_rate": 3.9963876291730086e-05, "loss": 0.0239, "step": 46890 }, { "grad_norm": 0.26433995366096497, "learning_rate": 3.994362941675614e-05, "loss": 0.0192, "step": 46900 }, { "grad_norm": 0.4408012926578522, "learning_rate": 3.992338426014088e-05, "loss": 0.0236, "step": 46910 }, { "grad_norm": 0.43439528346061707, "learning_rate": 3.9903140825343636e-05, "loss": 0.0204, "step": 46920 }, { "grad_norm": 0.2240021675825119, "learning_rate": 3.98828991158235e-05, "loss": 0.0182, "step": 46930 }, { "grad_norm": 0.25556182861328125, "learning_rate": 3.9862659135039185e-05, "loss": 0.022, "step": 46940 }, { "grad_norm": 0.24897252023220062, "learning_rate": 3.984242088644918e-05, "loss": 0.0197, "step": 46950 }, { "grad_norm": 0.27508237957954407, "learning_rate": 3.9822184373511615e-05, "loss": 0.0208, "step": 46960 }, { "grad_norm": 0.2526935935020447, "learning_rate": 3.98019495996844e-05, "loss": 0.0255, "step": 46970 }, { "grad_norm": 0.190490260720253, "learning_rate": 3.978171656842507e-05, "loss": 0.0205, "step": 46980 }, { "grad_norm": 0.27911949157714844, "learning_rate": 3.976148528319091e-05, "loss": 0.018, "step": 46990 }, { "grad_norm": 0.45159071683883667, "learning_rate": 3.974125574743888e-05, "loss": 0.0194, "step": 47000 }, { "grad_norm": 0.3841310441493988, "learning_rate": 3.9721027964625686e-05, "loss": 0.0214, "step": 47010 }, { "grad_norm": 0.22373820841312408, "learning_rate": 3.9700801938207676e-05, "loss": 0.0181, "step": 47020 }, { "grad_norm": 0.327090859413147, "learning_rate": 3.9680577671640916e-05, "loss": 0.0248, "step": 47030 }, { "grad_norm": 0.24234414100646973, "learning_rate": 3.966035516838121e-05, "loss": 0.0288, "step": 47040 }, { "grad_norm": 0.3568413257598877, "learning_rate": 3.9640134431884014e-05, "loss": 0.0206, "step": 47050 }, { "grad_norm": 0.19799058139324188, "learning_rate": 3.961991546560451e-05, "loss": 0.0205, "step": 47060 }, { "grad_norm": 0.2264462560415268, "learning_rate": 3.959969827299753e-05, "loss": 0.0243, "step": 47070 }, { "grad_norm": 0.2846795618534088, "learning_rate": 3.9579482857517684e-05, "loss": 0.0187, "step": 47080 }, { "grad_norm": 0.6459875106811523, "learning_rate": 3.955926922261921e-05, "loss": 0.0191, "step": 47090 }, { "grad_norm": 0.4727562665939331, "learning_rate": 3.9539057371756084e-05, "loss": 0.0234, "step": 47100 }, { "grad_norm": 0.2656565010547638, "learning_rate": 3.951884730838195e-05, "loss": 0.0138, "step": 47110 }, { "grad_norm": 0.2884919345378876, "learning_rate": 3.949863903595012e-05, "loss": 0.017, "step": 47120 }, { "grad_norm": 0.23256677389144897, "learning_rate": 3.947843255791369e-05, "loss": 0.0171, "step": 47130 }, { "grad_norm": 0.23157571256160736, "learning_rate": 3.9458227877725364e-05, "loss": 0.0217, "step": 47140 }, { "grad_norm": 0.29904380440711975, "learning_rate": 3.943802499883758e-05, "loss": 0.0142, "step": 47150 }, { "grad_norm": 0.34742826223373413, "learning_rate": 3.9417823924702437e-05, "loss": 0.0224, "step": 47160 }, { "grad_norm": 0.4074540138244629, "learning_rate": 3.939762465877178e-05, "loss": 0.0225, "step": 47170 }, { "grad_norm": 0.2551708221435547, "learning_rate": 3.937742720449708e-05, "loss": 0.0214, "step": 47180 }, { "grad_norm": 0.27658966183662415, "learning_rate": 3.9357231565329563e-05, "loss": 0.0194, "step": 47190 }, { "grad_norm": 0.2996051013469696, "learning_rate": 3.933703774472008e-05, "loss": 0.0181, "step": 47200 }, { "grad_norm": 0.3060730993747711, "learning_rate": 3.93168457461192e-05, "loss": 0.017, "step": 47210 }, { "grad_norm": 0.4394756257534027, "learning_rate": 3.9296655572977216e-05, "loss": 0.0207, "step": 47220 }, { "grad_norm": 0.2807008922100067, "learning_rate": 3.927646722874404e-05, "loss": 0.0228, "step": 47230 }, { "grad_norm": 0.18558230996131897, "learning_rate": 3.925628071686934e-05, "loss": 0.0174, "step": 47240 }, { "grad_norm": 0.18115422129631042, "learning_rate": 3.9236096040802415e-05, "loss": 0.0211, "step": 47250 }, { "grad_norm": 0.4867985248565674, "learning_rate": 3.9215913203992294e-05, "loss": 0.0276, "step": 47260 }, { "grad_norm": 0.27344807982444763, "learning_rate": 3.9195732209887645e-05, "loss": 0.0197, "step": 47270 }, { "grad_norm": 0.23800237476825714, "learning_rate": 3.9175553061936875e-05, "loss": 0.0179, "step": 47280 }, { "grad_norm": 0.18374599516391754, "learning_rate": 3.9155375763588045e-05, "loss": 0.0237, "step": 47290 }, { "grad_norm": 0.19968891143798828, "learning_rate": 3.913520031828889e-05, "loss": 0.018, "step": 47300 }, { "grad_norm": 0.22373059391975403, "learning_rate": 3.911502672948685e-05, "loss": 0.0186, "step": 47310 }, { "grad_norm": 0.29293307662010193, "learning_rate": 3.9094855000629014e-05, "loss": 0.0216, "step": 47320 }, { "grad_norm": 0.21776923537254333, "learning_rate": 3.907468513516223e-05, "loss": 0.0158, "step": 47330 }, { "grad_norm": 0.272491455078125, "learning_rate": 3.905451713653294e-05, "loss": 0.0145, "step": 47340 }, { "grad_norm": 0.21069182455539703, "learning_rate": 3.903435100818731e-05, "loss": 0.0156, "step": 47350 }, { "grad_norm": 1.0447592735290527, "learning_rate": 3.901418675357117e-05, "loss": 0.0206, "step": 47360 }, { "grad_norm": 0.24975749850273132, "learning_rate": 3.8994024376130075e-05, "loss": 0.0268, "step": 47370 }, { "grad_norm": 0.22858519852161407, "learning_rate": 3.8973863879309194e-05, "loss": 0.021, "step": 47380 }, { "grad_norm": 0.2631314992904663, "learning_rate": 3.8953705266553394e-05, "loss": 0.0211, "step": 47390 }, { "grad_norm": 0.46041247248649597, "learning_rate": 3.893354854130727e-05, "loss": 0.0284, "step": 47400 }, { "grad_norm": 0.36022520065307617, "learning_rate": 3.8913393707015006e-05, "loss": 0.0211, "step": 47410 }, { "grad_norm": 0.30274730920791626, "learning_rate": 3.889324076712056e-05, "loss": 0.0251, "step": 47420 }, { "grad_norm": 0.24039453268051147, "learning_rate": 3.8873089725067476e-05, "loss": 0.0179, "step": 47430 }, { "grad_norm": 0.2932639718055725, "learning_rate": 3.885294058429905e-05, "loss": 0.0262, "step": 47440 }, { "grad_norm": 0.1534269005060196, "learning_rate": 3.8832793348258206e-05, "loss": 0.0182, "step": 47450 }, { "grad_norm": 0.22396732866764069, "learning_rate": 3.881264802038756e-05, "loss": 0.0345, "step": 47460 }, { "grad_norm": 0.2984519302845001, "learning_rate": 3.879250460412939e-05, "loss": 0.0218, "step": 47470 }, { "grad_norm": 0.24547423422336578, "learning_rate": 3.8772363102925644e-05, "loss": 0.0191, "step": 47480 }, { "grad_norm": 0.2669741213321686, "learning_rate": 3.875222352021798e-05, "loss": 0.0167, "step": 47490 }, { "grad_norm": 0.21274054050445557, "learning_rate": 3.8732085859447686e-05, "loss": 0.0202, "step": 47500 }, { "grad_norm": 0.32527685165405273, "learning_rate": 3.871195012405575e-05, "loss": 0.019, "step": 47510 }, { "grad_norm": 0.2794967293739319, "learning_rate": 3.869181631748278e-05, "loss": 0.0214, "step": 47520 }, { "grad_norm": 0.36347147822380066, "learning_rate": 3.867168444316915e-05, "loss": 0.019, "step": 47530 }, { "grad_norm": 0.13106554746627808, "learning_rate": 3.865155450455481e-05, "loss": 0.015, "step": 47540 }, { "grad_norm": 0.3348800837993622, "learning_rate": 3.8631426505079426e-05, "loss": 0.0245, "step": 47550 }, { "grad_norm": 0.401328980922699, "learning_rate": 3.8611300448182304e-05, "loss": 0.0242, "step": 47560 }, { "grad_norm": 0.26813414692878723, "learning_rate": 3.859117633730248e-05, "loss": 0.0172, "step": 47570 }, { "grad_norm": 0.3276962339878082, "learning_rate": 3.857105417587858e-05, "loss": 0.0195, "step": 47580 }, { "grad_norm": 0.2758522629737854, "learning_rate": 3.855093396734894e-05, "loss": 0.0178, "step": 47590 }, { "grad_norm": 0.2358173131942749, "learning_rate": 3.8530815715151545e-05, "loss": 0.0316, "step": 47600 }, { "grad_norm": 0.30951422452926636, "learning_rate": 3.851069942272405e-05, "loss": 0.0147, "step": 47610 }, { "grad_norm": 0.40649059414863586, "learning_rate": 3.849058509350382e-05, "loss": 0.0203, "step": 47620 }, { "grad_norm": 0.21746426820755005, "learning_rate": 3.8470472730927783e-05, "loss": 0.0236, "step": 47630 }, { "grad_norm": 0.48960080742836, "learning_rate": 3.845036233843264e-05, "loss": 0.019, "step": 47640 }, { "grad_norm": 0.20788812637329102, "learning_rate": 3.843025391945469e-05, "loss": 0.0175, "step": 47650 }, { "grad_norm": 0.23399090766906738, "learning_rate": 3.841014747742992e-05, "loss": 0.0165, "step": 47660 }, { "grad_norm": 0.33226531744003296, "learning_rate": 3.839004301579397e-05, "loss": 0.0155, "step": 47670 }, { "grad_norm": 0.2488543540239334, "learning_rate": 3.8369940537982097e-05, "loss": 0.0205, "step": 47680 }, { "grad_norm": 0.2572069764137268, "learning_rate": 3.834984004742933e-05, "loss": 0.0172, "step": 47690 }, { "grad_norm": 0.5573893785476685, "learning_rate": 3.832974154757026e-05, "loss": 0.015, "step": 47700 }, { "grad_norm": 0.26677459478378296, "learning_rate": 3.830964504183919e-05, "loss": 0.0186, "step": 47710 }, { "grad_norm": 0.1541554480791092, "learning_rate": 3.828955053367003e-05, "loss": 0.0181, "step": 47720 }, { "grad_norm": 0.8635732531547546, "learning_rate": 3.826945802649642e-05, "loss": 0.0253, "step": 47730 }, { "grad_norm": 0.22684255242347717, "learning_rate": 3.824936752375159e-05, "loss": 0.0258, "step": 47740 }, { "grad_norm": 0.24142985045909882, "learning_rate": 3.822927902886848e-05, "loss": 0.0175, "step": 47750 }, { "grad_norm": 0.3050825893878937, "learning_rate": 3.8209192545279653e-05, "loss": 0.02, "step": 47760 }, { "grad_norm": 0.1988951861858368, "learning_rate": 3.8189108076417326e-05, "loss": 0.0189, "step": 47770 }, { "grad_norm": 0.24838796257972717, "learning_rate": 3.816902562571342e-05, "loss": 0.0279, "step": 47780 }, { "grad_norm": 0.3037506341934204, "learning_rate": 3.814894519659944e-05, "loss": 0.0189, "step": 47790 }, { "grad_norm": 0.20213106274604797, "learning_rate": 3.812886679250661e-05, "loss": 0.016, "step": 47800 }, { "grad_norm": 0.16691561043262482, "learning_rate": 3.810879041686575e-05, "loss": 0.0171, "step": 47810 }, { "grad_norm": 0.4337902069091797, "learning_rate": 3.808871607310741e-05, "loss": 0.0244, "step": 47820 }, { "grad_norm": 0.2977790832519531, "learning_rate": 3.806864376466169e-05, "loss": 0.0232, "step": 47830 }, { "grad_norm": 0.27103400230407715, "learning_rate": 3.804857349495845e-05, "loss": 0.0143, "step": 47840 }, { "grad_norm": 0.42087429761886597, "learning_rate": 3.802850526742713e-05, "loss": 0.0375, "step": 47850 }, { "grad_norm": 0.16947223246097565, "learning_rate": 3.800843908549683e-05, "loss": 0.0137, "step": 47860 }, { "grad_norm": 0.21285124123096466, "learning_rate": 3.7988374952596325e-05, "loss": 0.0123, "step": 47870 }, { "grad_norm": 0.28090059757232666, "learning_rate": 3.7968312872154e-05, "loss": 0.0211, "step": 47880 }, { "grad_norm": 0.30547797679901123, "learning_rate": 3.7948252847597965e-05, "loss": 0.0201, "step": 47890 }, { "grad_norm": 0.27609384059906006, "learning_rate": 3.7928194882355885e-05, "loss": 0.0201, "step": 47900 }, { "grad_norm": 0.25362274050712585, "learning_rate": 3.790813897985515e-05, "loss": 0.0182, "step": 47910 }, { "grad_norm": 0.32135310769081116, "learning_rate": 3.7888085143522726e-05, "loss": 0.0245, "step": 47920 }, { "grad_norm": 0.24461044371128082, "learning_rate": 3.7868033376785314e-05, "loss": 0.0193, "step": 47930 }, { "grad_norm": 0.47323286533355713, "learning_rate": 3.784798368306919e-05, "loss": 0.0182, "step": 47940 }, { "grad_norm": 0.43137326836586, "learning_rate": 3.782793606580029e-05, "loss": 0.0171, "step": 47950 }, { "grad_norm": 0.29727622866630554, "learning_rate": 3.7807890528404205e-05, "loss": 0.0219, "step": 47960 }, { "grad_norm": 0.3788757920265198, "learning_rate": 3.778784707430616e-05, "loss": 0.0214, "step": 47970 }, { "grad_norm": 0.22284629940986633, "learning_rate": 3.776780570693107e-05, "loss": 0.0162, "step": 47980 }, { "grad_norm": 0.1984013319015503, "learning_rate": 3.774776642970342e-05, "loss": 0.0247, "step": 47990 }, { "grad_norm": 0.19474488496780396, "learning_rate": 3.77277292460474e-05, "loss": 0.0161, "step": 48000 }, { "grad_norm": 0.4918748140335083, "learning_rate": 3.770769415938678e-05, "loss": 0.0186, "step": 48010 }, { "grad_norm": 0.44168633222579956, "learning_rate": 3.768766117314506e-05, "loss": 0.0245, "step": 48020 }, { "grad_norm": 0.7537171840667725, "learning_rate": 3.76676302907453e-05, "loss": 0.0241, "step": 48030 }, { "grad_norm": 1.1090211868286133, "learning_rate": 3.764760151561021e-05, "loss": 0.0163, "step": 48040 }, { "grad_norm": 0.19464543461799622, "learning_rate": 3.76275748511622e-05, "loss": 0.0166, "step": 48050 }, { "grad_norm": 0.40505293011665344, "learning_rate": 3.7607550300823255e-05, "loss": 0.02, "step": 48060 }, { "grad_norm": 0.27833297848701477, "learning_rate": 3.7587527868015044e-05, "loss": 0.0268, "step": 48070 }, { "grad_norm": 0.260621041059494, "learning_rate": 3.756750755615881e-05, "loss": 0.0193, "step": 48080 }, { "grad_norm": 0.17147135734558105, "learning_rate": 3.754748936867553e-05, "loss": 0.0188, "step": 48090 }, { "grad_norm": 0.22133825719356537, "learning_rate": 3.752747330898573e-05, "loss": 0.0181, "step": 48100 }, { "grad_norm": 0.21173758804798126, "learning_rate": 3.750745938050962e-05, "loss": 0.0181, "step": 48110 }, { "grad_norm": 0.22237294912338257, "learning_rate": 3.7487447586667025e-05, "loss": 0.0181, "step": 48120 }, { "grad_norm": 0.20336824655532837, "learning_rate": 3.7467437930877426e-05, "loss": 0.0151, "step": 48130 }, { "grad_norm": 0.23028506338596344, "learning_rate": 3.744743041655992e-05, "loss": 0.0171, "step": 48140 }, { "grad_norm": 0.2695131301879883, "learning_rate": 3.742742504713324e-05, "loss": 0.0216, "step": 48150 }, { "grad_norm": 0.23819883167743683, "learning_rate": 3.740742182601576e-05, "loss": 0.0195, "step": 48160 }, { "grad_norm": 0.19028666615486145, "learning_rate": 3.7387420756625464e-05, "loss": 0.0201, "step": 48170 }, { "grad_norm": 0.3354189395904541, "learning_rate": 3.736742184238002e-05, "loss": 0.0199, "step": 48180 }, { "grad_norm": 0.19862033426761627, "learning_rate": 3.7347425086696684e-05, "loss": 0.0186, "step": 48190 }, { "grad_norm": 0.35877105593681335, "learning_rate": 3.732743049299235e-05, "loss": 0.0192, "step": 48200 }, { "grad_norm": 0.1789042204618454, "learning_rate": 3.730743806468354e-05, "loss": 0.018, "step": 48210 }, { "grad_norm": 0.2962462902069092, "learning_rate": 3.7287447805186436e-05, "loss": 0.0168, "step": 48220 }, { "grad_norm": 0.21655578911304474, "learning_rate": 3.726745971791682e-05, "loss": 0.0213, "step": 48230 }, { "grad_norm": 0.31172630190849304, "learning_rate": 3.724747380629008e-05, "loss": 0.0271, "step": 48240 }, { "grad_norm": 0.18833324313163757, "learning_rate": 3.72274900737213e-05, "loss": 0.0208, "step": 48250 }, { "grad_norm": 0.2118244469165802, "learning_rate": 3.7207508523625123e-05, "loss": 0.0167, "step": 48260 }, { "grad_norm": 0.2109825760126114, "learning_rate": 3.718752915941588e-05, "loss": 0.0178, "step": 48270 }, { "grad_norm": 0.34459540247917175, "learning_rate": 3.7167551984507464e-05, "loss": 0.0184, "step": 48280 }, { "grad_norm": 0.2626105546951294, "learning_rate": 3.714757700231346e-05, "loss": 0.0176, "step": 48290 }, { "grad_norm": 0.20760153234004974, "learning_rate": 3.712760421624703e-05, "loss": 0.0159, "step": 48300 }, { "grad_norm": 0.21568846702575684, "learning_rate": 3.710763362972099e-05, "loss": 0.0166, "step": 48310 }, { "grad_norm": 0.4596996009349823, "learning_rate": 3.708766524614774e-05, "loss": 0.0251, "step": 48320 }, { "grad_norm": 0.2750209867954254, "learning_rate": 3.7067699068939335e-05, "loss": 0.0166, "step": 48330 }, { "grad_norm": 0.18284283578395844, "learning_rate": 3.704773510150748e-05, "loss": 0.0178, "step": 48340 }, { "grad_norm": 0.34099677205085754, "learning_rate": 3.702777334726344e-05, "loss": 0.0208, "step": 48350 }, { "grad_norm": 0.2585214674472809, "learning_rate": 3.7007813809618164e-05, "loss": 0.0134, "step": 48360 }, { "grad_norm": 0.2710273861885071, "learning_rate": 3.6987856491982145e-05, "loss": 0.0204, "step": 48370 }, { "grad_norm": 0.2525096535682678, "learning_rate": 3.696790139776558e-05, "loss": 0.0178, "step": 48380 }, { "grad_norm": 0.22756250202655792, "learning_rate": 3.6947948530378235e-05, "loss": 0.0223, "step": 48390 }, { "grad_norm": 0.15976306796073914, "learning_rate": 3.6927997893229516e-05, "loss": 0.0158, "step": 48400 }, { "grad_norm": 0.2819018065929413, "learning_rate": 3.690804948972845e-05, "loss": 0.0156, "step": 48410 }, { "grad_norm": 0.24133862555027008, "learning_rate": 3.688810332328363e-05, "loss": 0.0173, "step": 48420 }, { "grad_norm": 0.20544984936714172, "learning_rate": 3.686815939730336e-05, "loss": 0.0209, "step": 48430 }, { "grad_norm": 0.2799988389015198, "learning_rate": 3.684821771519548e-05, "loss": 0.0177, "step": 48440 }, { "grad_norm": 0.2689293622970581, "learning_rate": 3.68282782803675e-05, "loss": 0.0202, "step": 48450 }, { "grad_norm": 0.508490264415741, "learning_rate": 3.6808341096226504e-05, "loss": 0.0261, "step": 48460 }, { "grad_norm": 0.30596694350242615, "learning_rate": 3.678840616617924e-05, "loss": 0.0193, "step": 48470 }, { "grad_norm": 0.4094912111759186, "learning_rate": 3.6768473493632e-05, "loss": 0.0207, "step": 48480 }, { "grad_norm": 0.32955482602119446, "learning_rate": 3.6748543081990783e-05, "loss": 0.0212, "step": 48490 }, { "grad_norm": 0.4178602397441864, "learning_rate": 3.672861493466112e-05, "loss": 0.0149, "step": 48500 }, { "grad_norm": 0.1804555207490921, "learning_rate": 3.670868905504818e-05, "loss": 0.0179, "step": 48510 }, { "grad_norm": 0.3691929280757904, "learning_rate": 3.6688765446556784e-05, "loss": 0.02, "step": 48520 }, { "grad_norm": 0.3019929528236389, "learning_rate": 3.6668844112591276e-05, "loss": 0.0188, "step": 48530 }, { "grad_norm": 0.2798726260662079, "learning_rate": 3.664892505655573e-05, "loss": 0.0186, "step": 48540 }, { "grad_norm": 0.3117108643054962, "learning_rate": 3.662900828185373e-05, "loss": 0.0255, "step": 48550 }, { "grad_norm": 0.20113928616046906, "learning_rate": 3.6609093791888516e-05, "loss": 0.0204, "step": 48560 }, { "grad_norm": 0.2970356345176697, "learning_rate": 3.658918159006292e-05, "loss": 0.03, "step": 48570 }, { "grad_norm": 0.3498767018318176, "learning_rate": 3.656927167977942e-05, "loss": 0.0258, "step": 48580 }, { "grad_norm": 0.3818439841270447, "learning_rate": 3.654936406444006e-05, "loss": 0.0193, "step": 48590 }, { "grad_norm": 0.2860354781150818, "learning_rate": 3.65294587474465e-05, "loss": 0.0193, "step": 48600 }, { "grad_norm": 0.21467989683151245, "learning_rate": 3.650955573220002e-05, "loss": 0.0468, "step": 48610 }, { "grad_norm": 0.3122539818286896, "learning_rate": 3.648965502210149e-05, "loss": 0.0172, "step": 48620 }, { "grad_norm": 0.35775673389434814, "learning_rate": 3.646975662055142e-05, "loss": 0.0214, "step": 48630 }, { "grad_norm": 0.2865304946899414, "learning_rate": 3.644986053094987e-05, "loss": 0.0177, "step": 48640 }, { "grad_norm": 0.26484212279319763, "learning_rate": 3.642996675669659e-05, "loss": 0.0202, "step": 48650 }, { "grad_norm": 0.310587614774704, "learning_rate": 3.641007530119083e-05, "loss": 0.015, "step": 48660 }, { "grad_norm": 0.20396552979946136, "learning_rate": 3.639018616783153e-05, "loss": 0.0197, "step": 48670 }, { "grad_norm": 0.2305084466934204, "learning_rate": 3.637029936001719e-05, "loss": 0.0173, "step": 48680 }, { "grad_norm": 0.20320066809654236, "learning_rate": 3.6350414881145886e-05, "loss": 0.0231, "step": 48690 }, { "grad_norm": 0.6212013959884644, "learning_rate": 3.6330532734615386e-05, "loss": 0.022, "step": 48700 }, { "grad_norm": 0.2137977033853531, "learning_rate": 3.6310652923822975e-05, "loss": 0.0205, "step": 48710 }, { "grad_norm": 0.24652042984962463, "learning_rate": 3.629077545216558e-05, "loss": 0.0195, "step": 48720 }, { "grad_norm": 0.3526974320411682, "learning_rate": 3.627090032303969e-05, "loss": 0.02, "step": 48730 }, { "grad_norm": 0.19780530035495758, "learning_rate": 3.625102753984146e-05, "loss": 0.0219, "step": 48740 }, { "grad_norm": 0.3146408498287201, "learning_rate": 3.623115710596659e-05, "loss": 0.0159, "step": 48750 }, { "grad_norm": 0.371413916349411, "learning_rate": 3.6211289024810395e-05, "loss": 0.0227, "step": 48760 }, { "grad_norm": 0.262281596660614, "learning_rate": 3.619142329976777e-05, "loss": 0.017, "step": 48770 }, { "grad_norm": 0.22843582928180695, "learning_rate": 3.6171559934233247e-05, "loss": 0.0178, "step": 48780 }, { "grad_norm": 0.23050859570503235, "learning_rate": 3.615169893160093e-05, "loss": 0.0146, "step": 48790 }, { "grad_norm": 0.32590654492378235, "learning_rate": 3.61318402952645e-05, "loss": 0.0213, "step": 48800 }, { "grad_norm": 0.35277003049850464, "learning_rate": 3.6111984028617285e-05, "loss": 0.0224, "step": 48810 }, { "grad_norm": 0.25854721665382385, "learning_rate": 3.6092130135052134e-05, "loss": 0.0249, "step": 48820 }, { "grad_norm": 0.20605753362178802, "learning_rate": 3.6072278617961584e-05, "loss": 0.0243, "step": 48830 }, { "grad_norm": 0.26999029517173767, "learning_rate": 3.605242948073767e-05, "loss": 0.0212, "step": 48840 }, { "grad_norm": 0.29675382375717163, "learning_rate": 3.603258272677212e-05, "loss": 0.0193, "step": 48850 }, { "grad_norm": 0.2299017459154129, "learning_rate": 3.601273835945616e-05, "loss": 0.0194, "step": 48860 }, { "grad_norm": 0.27948272228240967, "learning_rate": 3.5992896382180664e-05, "loss": 0.0173, "step": 48870 }, { "grad_norm": 0.23396848142147064, "learning_rate": 3.597305679833609e-05, "loss": 0.0211, "step": 48880 }, { "grad_norm": 0.36847490072250366, "learning_rate": 3.595321961131245e-05, "loss": 0.0223, "step": 48890 }, { "grad_norm": 0.5593545436859131, "learning_rate": 3.593338482449942e-05, "loss": 0.0297, "step": 48900 }, { "grad_norm": 0.26623886823654175, "learning_rate": 3.591355244128618e-05, "loss": 0.0281, "step": 48910 }, { "grad_norm": 0.29110774397850037, "learning_rate": 3.589372246506158e-05, "loss": 0.0265, "step": 48920 }, { "grad_norm": 0.3751128911972046, "learning_rate": 3.5873894899213984e-05, "loss": 0.0152, "step": 48930 }, { "grad_norm": 0.3535178005695343, "learning_rate": 3.5854069747131416e-05, "loss": 0.0172, "step": 48940 }, { "grad_norm": 0.37876129150390625, "learning_rate": 3.583424701220143e-05, "loss": 0.0268, "step": 48950 }, { "grad_norm": 0.3147808015346527, "learning_rate": 3.581442669781121e-05, "loss": 0.0237, "step": 48960 }, { "grad_norm": 0.36112940311431885, "learning_rate": 3.579460880734749e-05, "loss": 0.0189, "step": 48970 }, { "grad_norm": 0.2450316697359085, "learning_rate": 3.577479334419657e-05, "loss": 0.0209, "step": 48980 }, { "grad_norm": 0.30655208230018616, "learning_rate": 3.575498031174444e-05, "loss": 0.0171, "step": 48990 }, { "grad_norm": 0.33196499943733215, "learning_rate": 3.573516971337657e-05, "loss": 0.0203, "step": 49000 }, { "grad_norm": 0.36115342378616333, "learning_rate": 3.5715361552478046e-05, "loss": 0.0247, "step": 49010 }, { "grad_norm": 0.23157604038715363, "learning_rate": 3.5695555832433536e-05, "loss": 0.0166, "step": 49020 }, { "grad_norm": 0.37938740849494934, "learning_rate": 3.5675752556627325e-05, "loss": 0.0186, "step": 49030 }, { "grad_norm": 0.2978001534938812, "learning_rate": 3.565595172844322e-05, "loss": 0.0162, "step": 49040 }, { "grad_norm": 0.2707420289516449, "learning_rate": 3.5636153351264666e-05, "loss": 0.0228, "step": 49050 }, { "grad_norm": 0.20246456563472748, "learning_rate": 3.5616357428474655e-05, "loss": 0.0203, "step": 49060 }, { "grad_norm": 0.26243510842323303, "learning_rate": 3.559656396345575e-05, "loss": 0.0224, "step": 49070 }, { "grad_norm": 0.22041262686252594, "learning_rate": 3.5576772959590146e-05, "loss": 0.022, "step": 49080 }, { "grad_norm": 0.17872510850429535, "learning_rate": 3.5556984420259545e-05, "loss": 0.018, "step": 49090 }, { "grad_norm": 0.2582422196865082, "learning_rate": 3.5537198348845305e-05, "loss": 0.0225, "step": 49100 }, { "grad_norm": 0.25173142552375793, "learning_rate": 3.551741474872831e-05, "loss": 0.024, "step": 49110 }, { "grad_norm": 0.5203378200531006, "learning_rate": 3.549763362328903e-05, "loss": 0.0278, "step": 49120 }, { "grad_norm": 0.2875223755836487, "learning_rate": 3.5477854975907515e-05, "loss": 0.0186, "step": 49130 }, { "grad_norm": 0.2274032086133957, "learning_rate": 3.5458078809963416e-05, "loss": 0.0157, "step": 49140 }, { "grad_norm": 0.26642388105392456, "learning_rate": 3.543830512883594e-05, "loss": 0.0163, "step": 49150 }, { "grad_norm": 0.35941624641418457, "learning_rate": 3.5418533935903824e-05, "loss": 0.021, "step": 49160 }, { "grad_norm": 0.411309152841568, "learning_rate": 3.539876523454547e-05, "loss": 0.0202, "step": 49170 }, { "grad_norm": 0.283141553401947, "learning_rate": 3.537899902813878e-05, "loss": 0.0217, "step": 49180 }, { "grad_norm": 0.1884445995092392, "learning_rate": 3.5359235320061293e-05, "loss": 0.0142, "step": 49190 }, { "grad_norm": 0.26312676072120667, "learning_rate": 3.533947411369003e-05, "loss": 0.0168, "step": 49200 }, { "grad_norm": 0.23646152019500732, "learning_rate": 3.53197154124017e-05, "loss": 0.018, "step": 49210 }, { "grad_norm": 0.3696039617061615, "learning_rate": 3.52999592195725e-05, "loss": 0.0191, "step": 49220 }, { "grad_norm": 0.22494259476661682, "learning_rate": 3.5280205538578224e-05, "loss": 0.0166, "step": 49230 }, { "grad_norm": 0.2791041135787964, "learning_rate": 3.5260454372794236e-05, "loss": 0.0159, "step": 49240 }, { "grad_norm": 0.2584582567214966, "learning_rate": 3.524070572559545e-05, "loss": 0.0207, "step": 49250 }, { "grad_norm": 0.26266247034072876, "learning_rate": 3.5220959600356395e-05, "loss": 0.0149, "step": 49260 }, { "grad_norm": 0.1801149994134903, "learning_rate": 3.5201216000451145e-05, "loss": 0.0183, "step": 49270 }, { "grad_norm": 0.36958014965057373, "learning_rate": 3.5181474929253335e-05, "loss": 0.0166, "step": 49280 }, { "grad_norm": 0.23885200917720795, "learning_rate": 3.516173639013615e-05, "loss": 0.0195, "step": 49290 }, { "grad_norm": 0.1493135690689087, "learning_rate": 3.5142000386472406e-05, "loss": 0.0177, "step": 49300 }, { "grad_norm": 0.24874822795391083, "learning_rate": 3.5122266921634427e-05, "loss": 0.0155, "step": 49310 }, { "grad_norm": 0.2667423188686371, "learning_rate": 3.510253599899413e-05, "loss": 0.0173, "step": 49320 }, { "grad_norm": 0.17882139980793, "learning_rate": 3.5082807621922965e-05, "loss": 0.0143, "step": 49330 }, { "grad_norm": 0.33760207891464233, "learning_rate": 3.506308179379201e-05, "loss": 0.0203, "step": 49340 }, { "grad_norm": 0.19100221991539001, "learning_rate": 3.5043358517971844e-05, "loss": 0.0201, "step": 49350 }, { "grad_norm": 0.22942784428596497, "learning_rate": 3.502363779783264e-05, "loss": 0.0228, "step": 49360 }, { "grad_norm": 0.2516992688179016, "learning_rate": 3.500391963674415e-05, "loss": 0.0182, "step": 49370 }, { "grad_norm": 0.25558891892433167, "learning_rate": 3.4984204038075615e-05, "loss": 0.0229, "step": 49380 }, { "grad_norm": 0.24569551646709442, "learning_rate": 3.496449100519595e-05, "loss": 0.022, "step": 49390 }, { "grad_norm": 0.21235285699367523, "learning_rate": 3.494478054147354e-05, "loss": 0.0158, "step": 49400 }, { "grad_norm": 0.19001378118991852, "learning_rate": 3.4925072650276395e-05, "loss": 0.0146, "step": 49410 }, { "grad_norm": 0.38146233558654785, "learning_rate": 3.4905367334972016e-05, "loss": 0.0152, "step": 49420 }, { "grad_norm": 0.3453446328639984, "learning_rate": 3.488566459892752e-05, "loss": 0.0185, "step": 49430 }, { "grad_norm": 0.46782034635543823, "learning_rate": 3.4865964445509585e-05, "loss": 0.0173, "step": 49440 }, { "grad_norm": 0.2390550673007965, "learning_rate": 3.484626687808438e-05, "loss": 0.0147, "step": 49450 }, { "grad_norm": 0.2408626228570938, "learning_rate": 3.4826571900017735e-05, "loss": 0.0131, "step": 49460 }, { "grad_norm": 0.2857845425605774, "learning_rate": 3.480687951467495e-05, "loss": 0.0153, "step": 49470 }, { "grad_norm": 0.27482736110687256, "learning_rate": 3.4787189725420925e-05, "loss": 0.0148, "step": 49480 }, { "grad_norm": 0.21008197963237762, "learning_rate": 3.4767502535620086e-05, "loss": 0.0178, "step": 49490 }, { "grad_norm": 0.27233558893203735, "learning_rate": 3.474781794863648e-05, "loss": 0.0221, "step": 49500 }, { "grad_norm": 0.32539239525794983, "learning_rate": 3.472813596783363e-05, "loss": 0.0236, "step": 49510 }, { "grad_norm": 0.28062528371810913, "learning_rate": 3.470845659657466e-05, "loss": 0.0195, "step": 49520 }, { "grad_norm": 0.14689651131629944, "learning_rate": 3.468877983822223e-05, "loss": 0.0154, "step": 49530 }, { "grad_norm": 0.21382324397563934, "learning_rate": 3.466910569613855e-05, "loss": 0.0163, "step": 49540 }, { "grad_norm": 0.4169554114341736, "learning_rate": 3.464943417368542e-05, "loss": 0.0203, "step": 49550 }, { "grad_norm": 0.3154906928539276, "learning_rate": 3.462976527422415e-05, "loss": 0.0179, "step": 49560 }, { "grad_norm": 0.3274933993816376, "learning_rate": 3.461009900111562e-05, "loss": 0.0229, "step": 49570 }, { "grad_norm": 0.34378233551979065, "learning_rate": 3.459043535772023e-05, "loss": 0.0201, "step": 49580 }, { "grad_norm": 0.3316482901573181, "learning_rate": 3.4570774347398014e-05, "loss": 0.0234, "step": 49590 }, { "grad_norm": 0.41068434715270996, "learning_rate": 3.4551115973508454e-05, "loss": 0.0246, "step": 49600 }, { "grad_norm": 0.3670130968093872, "learning_rate": 3.453146023941066e-05, "loss": 0.0278, "step": 49610 }, { "grad_norm": 0.23522047698497772, "learning_rate": 3.451180714846325e-05, "loss": 0.0199, "step": 49620 }, { "grad_norm": 0.696019172668457, "learning_rate": 3.449215670402438e-05, "loss": 0.0229, "step": 49630 }, { "grad_norm": 0.47368380427360535, "learning_rate": 3.447250890945181e-05, "loss": 0.0207, "step": 49640 }, { "grad_norm": 0.2965738773345947, "learning_rate": 3.4452863768102754e-05, "loss": 0.022, "step": 49650 }, { "grad_norm": 0.24919742345809937, "learning_rate": 3.443322128333409e-05, "loss": 0.0178, "step": 49660 }, { "grad_norm": 0.46854814887046814, "learning_rate": 3.441358145850215e-05, "loss": 0.0214, "step": 49670 }, { "grad_norm": 0.288274347782135, "learning_rate": 3.439394429696286e-05, "loss": 0.0221, "step": 49680 }, { "grad_norm": 0.24875852465629578, "learning_rate": 3.4374309802071644e-05, "loss": 0.0217, "step": 49690 }, { "grad_norm": 0.5342387557029724, "learning_rate": 3.435467797718353e-05, "loss": 0.0156, "step": 49700 }, { "grad_norm": 0.2914935052394867, "learning_rate": 3.433504882565306e-05, "loss": 0.0191, "step": 49710 }, { "grad_norm": 0.27231505513191223, "learning_rate": 3.43154223508343e-05, "loss": 0.0223, "step": 49720 }, { "grad_norm": 0.17878839373588562, "learning_rate": 3.429579855608089e-05, "loss": 0.0161, "step": 49730 }, { "grad_norm": 0.27280837297439575, "learning_rate": 3.427617744474597e-05, "loss": 0.0163, "step": 49740 }, { "grad_norm": 0.2769489288330078, "learning_rate": 3.425655902018231e-05, "loss": 0.0197, "step": 49750 }, { "grad_norm": 0.15121839940547943, "learning_rate": 3.423694328574211e-05, "loss": 0.0182, "step": 49760 }, { "grad_norm": 0.463919997215271, "learning_rate": 3.42173302447772e-05, "loss": 0.0178, "step": 49770 }, { "grad_norm": 0.14793972671031952, "learning_rate": 3.419771990063886e-05, "loss": 0.016, "step": 49780 }, { "grad_norm": 0.7199063897132874, "learning_rate": 3.417811225667803e-05, "loss": 0.0168, "step": 49790 }, { "grad_norm": 0.24636977910995483, "learning_rate": 3.415850731624508e-05, "loss": 0.0193, "step": 49800 }, { "grad_norm": 0.3152816891670227, "learning_rate": 3.4138905082689945e-05, "loss": 0.025, "step": 49810 }, { "grad_norm": 0.2651118040084839, "learning_rate": 3.4119305559362145e-05, "loss": 0.0187, "step": 49820 }, { "grad_norm": 0.21615859866142273, "learning_rate": 3.4099708749610684e-05, "loss": 0.015, "step": 49830 }, { "grad_norm": 0.4402173161506653, "learning_rate": 3.408011465678413e-05, "loss": 0.0245, "step": 49840 }, { "grad_norm": 0.20929652452468872, "learning_rate": 3.406052328423055e-05, "loss": 0.0192, "step": 49850 }, { "grad_norm": 0.1875891238451004, "learning_rate": 3.4040934635297615e-05, "loss": 0.0155, "step": 49860 }, { "grad_norm": 0.3460161089897156, "learning_rate": 3.4021348713332466e-05, "loss": 0.0152, "step": 49870 }, { "grad_norm": 0.32360732555389404, "learning_rate": 3.4001765521681807e-05, "loss": 0.0211, "step": 49880 }, { "grad_norm": 0.23929931223392487, "learning_rate": 3.398218506369188e-05, "loss": 0.0279, "step": 49890 }, { "grad_norm": 0.28809672594070435, "learning_rate": 3.3962607342708404e-05, "loss": 0.0144, "step": 49900 }, { "grad_norm": 0.24890249967575073, "learning_rate": 3.394303236207673e-05, "loss": 0.0137, "step": 49910 }, { "grad_norm": 0.2803521454334259, "learning_rate": 3.392346012514166e-05, "loss": 0.0198, "step": 49920 }, { "grad_norm": 0.2692691683769226, "learning_rate": 3.390389063524757e-05, "loss": 0.0164, "step": 49930 }, { "grad_norm": 0.25871554017066956, "learning_rate": 3.3884323895738324e-05, "loss": 0.0159, "step": 49940 }, { "grad_norm": 0.26687735319137573, "learning_rate": 3.386475990995738e-05, "loss": 0.0212, "step": 49950 }, { "grad_norm": 0.29828065633773804, "learning_rate": 3.384519868124765e-05, "loss": 0.014, "step": 49960 }, { "grad_norm": 0.2084556668996811, "learning_rate": 3.3825640212951645e-05, "loss": 0.0184, "step": 49970 }, { "grad_norm": 0.4832545816898346, "learning_rate": 3.380608450841134e-05, "loss": 0.0181, "step": 49980 }, { "grad_norm": 0.28775259852409363, "learning_rate": 3.3786531570968305e-05, "loss": 0.024, "step": 49990 }, { "grad_norm": 0.2657579183578491, "learning_rate": 3.3766981403963584e-05, "loss": 0.0228, "step": 50000 }, { "grad_norm": 0.19511093199253082, "learning_rate": 3.374743401073775e-05, "loss": 0.0179, "step": 50010 }, { "grad_norm": 0.21400724351406097, "learning_rate": 3.372788939463095e-05, "loss": 0.0151, "step": 50020 }, { "grad_norm": 0.32315823435783386, "learning_rate": 3.370834755898281e-05, "loss": 0.014, "step": 50030 }, { "grad_norm": 0.3116269111633301, "learning_rate": 3.3688808507132493e-05, "loss": 0.0203, "step": 50040 }, { "grad_norm": 0.20183253288269043, "learning_rate": 3.3669272242418685e-05, "loss": 0.0211, "step": 50050 }, { "grad_norm": 0.23991301655769348, "learning_rate": 3.364973876817961e-05, "loss": 0.0171, "step": 50060 }, { "grad_norm": 0.28517740964889526, "learning_rate": 3.363020808775299e-05, "loss": 0.0147, "step": 50070 }, { "grad_norm": 0.194179967045784, "learning_rate": 3.361068020447611e-05, "loss": 0.0157, "step": 50080 }, { "grad_norm": 0.20742778480052948, "learning_rate": 3.3591155121685724e-05, "loss": 0.0128, "step": 50090 }, { "grad_norm": 0.36908912658691406, "learning_rate": 3.357163284271814e-05, "loss": 0.0189, "step": 50100 }, { "grad_norm": 0.30329063534736633, "learning_rate": 3.355211337090919e-05, "loss": 0.0156, "step": 50110 }, { "grad_norm": 0.32336220145225525, "learning_rate": 3.353259670959421e-05, "loss": 0.0244, "step": 50120 }, { "grad_norm": 0.22807660698890686, "learning_rate": 3.351308286210808e-05, "loss": 0.021, "step": 50130 }, { "grad_norm": 0.8907410502433777, "learning_rate": 3.3493571831785156e-05, "loss": 0.0179, "step": 50140 }, { "grad_norm": 0.2767936587333679, "learning_rate": 3.347406362195936e-05, "loss": 0.0143, "step": 50150 }, { "grad_norm": 0.3362113833427429, "learning_rate": 3.345455823596411e-05, "loss": 0.016, "step": 50160 }, { "grad_norm": 0.2721680998802185, "learning_rate": 3.3435055677132346e-05, "loss": 0.0184, "step": 50170 }, { "grad_norm": 0.3074028491973877, "learning_rate": 3.3415555948796505e-05, "loss": 0.0206, "step": 50180 }, { "grad_norm": 0.412997305393219, "learning_rate": 3.3396059054288556e-05, "loss": 0.0184, "step": 50190 }, { "grad_norm": 0.3234202265739441, "learning_rate": 3.3376564996940015e-05, "loss": 0.0291, "step": 50200 }, { "grad_norm": 0.24170970916748047, "learning_rate": 3.3357073780081836e-05, "loss": 0.0161, "step": 50210 }, { "grad_norm": 0.25073322653770447, "learning_rate": 3.333758540704459e-05, "loss": 0.0148, "step": 50220 }, { "grad_norm": 0.2009810209274292, "learning_rate": 3.331809988115827e-05, "loss": 0.0156, "step": 50230 }, { "grad_norm": 0.32593387365341187, "learning_rate": 3.329861720575244e-05, "loss": 0.0202, "step": 50240 }, { "grad_norm": 0.24277375638484955, "learning_rate": 3.3279137384156126e-05, "loss": 0.0181, "step": 50250 }, { "grad_norm": 0.3061363101005554, "learning_rate": 3.3259660419697934e-05, "loss": 0.0279, "step": 50260 }, { "grad_norm": 0.13630411028862, "learning_rate": 3.3240186315705926e-05, "loss": 0.0238, "step": 50270 }, { "grad_norm": 0.32433629035949707, "learning_rate": 3.322071507550769e-05, "loss": 0.0268, "step": 50280 }, { "grad_norm": 0.30195701122283936, "learning_rate": 3.320124670243033e-05, "loss": 0.0168, "step": 50290 }, { "grad_norm": 0.18186667561531067, "learning_rate": 3.318178119980045e-05, "loss": 0.0154, "step": 50300 }, { "grad_norm": 0.25787535309791565, "learning_rate": 3.316231857094421e-05, "loss": 0.0164, "step": 50310 }, { "grad_norm": 0.23179766535758972, "learning_rate": 3.31428588191872e-05, "loss": 0.0205, "step": 50320 }, { "grad_norm": 0.26975899934768677, "learning_rate": 3.312340194785458e-05, "loss": 0.0193, "step": 50330 }, { "grad_norm": 0.248013436794281, "learning_rate": 3.3103947960270975e-05, "loss": 0.014, "step": 50340 }, { "grad_norm": 0.3912910521030426, "learning_rate": 3.308449685976058e-05, "loss": 0.0163, "step": 50350 }, { "grad_norm": 0.22081898152828217, "learning_rate": 3.3065048649647024e-05, "loss": 0.0173, "step": 50360 }, { "grad_norm": 0.29843807220458984, "learning_rate": 3.304560333325348e-05, "loss": 0.0156, "step": 50370 }, { "grad_norm": 0.35470983386039734, "learning_rate": 3.3026160913902624e-05, "loss": 0.0197, "step": 50380 }, { "grad_norm": 0.4187701344490051, "learning_rate": 3.300672139491662e-05, "loss": 0.0242, "step": 50390 }, { "grad_norm": 0.23460225760936737, "learning_rate": 3.298728477961717e-05, "loss": 0.0194, "step": 50400 }, { "grad_norm": 0.20379012823104858, "learning_rate": 3.296785107132544e-05, "loss": 0.0188, "step": 50410 }, { "grad_norm": 0.4328249990940094, "learning_rate": 3.2948420273362156e-05, "loss": 0.0154, "step": 50420 }, { "grad_norm": 0.17255456745624542, "learning_rate": 3.292899238904747e-05, "loss": 0.018, "step": 50430 }, { "grad_norm": 0.355975478887558, "learning_rate": 3.29095674217011e-05, "loss": 0.0208, "step": 50440 }, { "grad_norm": 0.3414525091648102, "learning_rate": 3.289014537464224e-05, "loss": 0.0197, "step": 50450 }, { "grad_norm": 0.21279370784759521, "learning_rate": 3.287072625118955e-05, "loss": 0.0182, "step": 50460 }, { "grad_norm": 0.22940567135810852, "learning_rate": 3.285131005466129e-05, "loss": 0.0189, "step": 50470 }, { "grad_norm": 0.25941863656044006, "learning_rate": 3.2831896788375105e-05, "loss": 0.019, "step": 50480 }, { "grad_norm": 0.3572346866130829, "learning_rate": 3.281248645564822e-05, "loss": 0.0182, "step": 50490 }, { "grad_norm": 0.2022678554058075, "learning_rate": 3.2793079059797306e-05, "loss": 0.025, "step": 50500 }, { "grad_norm": 0.225595623254776, "learning_rate": 3.277367460413859e-05, "loss": 0.0161, "step": 50510 }, { "grad_norm": 0.16327200829982758, "learning_rate": 3.275427309198773e-05, "loss": 0.0202, "step": 50520 }, { "grad_norm": 0.19802120327949524, "learning_rate": 3.273487452665993e-05, "loss": 0.0171, "step": 50530 }, { "grad_norm": 0.26630380749702454, "learning_rate": 3.271547891146986e-05, "loss": 0.0135, "step": 50540 }, { "grad_norm": 0.3052414655685425, "learning_rate": 3.269608624973173e-05, "loss": 0.0154, "step": 50550 }, { "grad_norm": 0.34591034054756165, "learning_rate": 3.26766965447592e-05, "loss": 0.022, "step": 50560 }, { "grad_norm": 0.30096855759620667, "learning_rate": 3.2657309799865424e-05, "loss": 0.0236, "step": 50570 }, { "grad_norm": 0.1590389460325241, "learning_rate": 3.2637926018363084e-05, "loss": 0.0151, "step": 50580 }, { "grad_norm": 0.3312523365020752, "learning_rate": 3.261854520356432e-05, "loss": 0.0201, "step": 50590 }, { "grad_norm": 0.23519274592399597, "learning_rate": 3.2599167358780816e-05, "loss": 0.0227, "step": 50600 }, { "grad_norm": 0.35419368743896484, "learning_rate": 3.2579792487323676e-05, "loss": 0.0157, "step": 50610 }, { "grad_norm": 0.28393977880477905, "learning_rate": 3.256042059250358e-05, "loss": 0.0241, "step": 50620 }, { "grad_norm": 0.5223758816719055, "learning_rate": 3.254105167763062e-05, "loss": 0.0232, "step": 50630 }, { "grad_norm": 0.2292107790708542, "learning_rate": 3.252168574601443e-05, "loss": 0.0173, "step": 50640 }, { "grad_norm": 0.30454516410827637, "learning_rate": 3.2502322800964115e-05, "loss": 0.0202, "step": 50650 }, { "grad_norm": 0.29177844524383545, "learning_rate": 3.248296284578825e-05, "loss": 0.0173, "step": 50660 }, { "grad_norm": 0.7152701020240784, "learning_rate": 3.246360588379497e-05, "loss": 0.025, "step": 50670 }, { "grad_norm": 0.20897792279720306, "learning_rate": 3.24442519182918e-05, "loss": 0.0304, "step": 50680 }, { "grad_norm": 0.18965929746627808, "learning_rate": 3.2424900952585845e-05, "loss": 0.0173, "step": 50690 }, { "grad_norm": 0.28692516684532166, "learning_rate": 3.240555298998362e-05, "loss": 0.0189, "step": 50700 }, { "grad_norm": 0.2968215346336365, "learning_rate": 3.2386208033791194e-05, "loss": 0.0169, "step": 50710 }, { "grad_norm": 0.4141223430633545, "learning_rate": 3.2366866087314064e-05, "loss": 0.0209, "step": 50720 }, { "grad_norm": 0.24534495174884796, "learning_rate": 3.234752715385727e-05, "loss": 0.0214, "step": 50730 }, { "grad_norm": 0.25321170687675476, "learning_rate": 3.232819123672529e-05, "loss": 0.0231, "step": 50740 }, { "grad_norm": 0.4111206829547882, "learning_rate": 3.2308858339222076e-05, "loss": 0.0283, "step": 50750 }, { "grad_norm": 0.33807605504989624, "learning_rate": 3.2289528464651144e-05, "loss": 0.0219, "step": 50760 }, { "grad_norm": 0.41694486141204834, "learning_rate": 3.22702016163154e-05, "loss": 0.0201, "step": 50770 }, { "grad_norm": 0.3689272701740265, "learning_rate": 3.2250877797517295e-05, "loss": 0.0254, "step": 50780 }, { "grad_norm": 1.4602172374725342, "learning_rate": 3.223155701155872e-05, "loss": 0.0184, "step": 50790 }, { "grad_norm": 0.20334307849407196, "learning_rate": 3.2212239261741105e-05, "loss": 0.0215, "step": 50800 }, { "grad_norm": 0.18926562368869781, "learning_rate": 3.219292455136528e-05, "loss": 0.0164, "step": 50810 }, { "grad_norm": 0.19768182933330536, "learning_rate": 3.217361288373165e-05, "loss": 0.0196, "step": 50820 }, { "grad_norm": 0.1675342619419098, "learning_rate": 3.215430426214002e-05, "loss": 0.0223, "step": 50830 }, { "grad_norm": 0.322726309299469, "learning_rate": 3.2134998689889705e-05, "loss": 0.0195, "step": 50840 }, { "grad_norm": 0.5440185070037842, "learning_rate": 3.211569617027952e-05, "loss": 0.0203, "step": 50850 }, { "grad_norm": 0.3819557726383209, "learning_rate": 3.209639670660769e-05, "loss": 0.0206, "step": 50860 }, { "grad_norm": 0.5936469435691833, "learning_rate": 3.207710030217202e-05, "loss": 0.0181, "step": 50870 }, { "grad_norm": 0.5519839525222778, "learning_rate": 3.205780696026972e-05, "loss": 0.0182, "step": 50880 }, { "grad_norm": 0.16833892464637756, "learning_rate": 3.203851668419749e-05, "loss": 0.0158, "step": 50890 }, { "grad_norm": 0.2687104642391205, "learning_rate": 3.201922947725149e-05, "loss": 0.0164, "step": 50900 }, { "grad_norm": 0.2474723607301712, "learning_rate": 3.199994534272742e-05, "loss": 0.019, "step": 50910 }, { "grad_norm": 0.25497350096702576, "learning_rate": 3.1980664283920394e-05, "loss": 0.0219, "step": 50920 }, { "grad_norm": 0.2684515118598938, "learning_rate": 3.196138630412499e-05, "loss": 0.0132, "step": 50930 }, { "grad_norm": 0.3264741003513336, "learning_rate": 3.194211140663532e-05, "loss": 0.0185, "step": 50940 }, { "grad_norm": 0.30043306946754456, "learning_rate": 3.1922839594744914e-05, "loss": 0.023, "step": 50950 }, { "grad_norm": 0.3411480784416199, "learning_rate": 3.190357087174683e-05, "loss": 0.0205, "step": 50960 }, { "grad_norm": 0.1320803463459015, "learning_rate": 3.188430524093353e-05, "loss": 0.0161, "step": 50970 }, { "grad_norm": 0.44077593088150024, "learning_rate": 3.1865042705597014e-05, "loss": 0.0208, "step": 50980 }, { "grad_norm": 0.28604158759117126, "learning_rate": 3.184578326902868e-05, "loss": 0.0166, "step": 50990 }, { "grad_norm": 0.34168004989624023, "learning_rate": 3.182652693451949e-05, "loss": 0.0187, "step": 51000 }, { "grad_norm": 0.35323163866996765, "learning_rate": 3.180727370535981e-05, "loss": 0.0183, "step": 51010 }, { "grad_norm": 0.182695671916008, "learning_rate": 3.178802358483945e-05, "loss": 0.0166, "step": 51020 }, { "grad_norm": 0.19171465933322906, "learning_rate": 3.17687765762478e-05, "loss": 0.0175, "step": 51030 }, { "grad_norm": 0.2419189214706421, "learning_rate": 3.174953268287357e-05, "loss": 0.0204, "step": 51040 }, { "grad_norm": 0.44861090183258057, "learning_rate": 3.173029190800508e-05, "loss": 0.0163, "step": 51050 }, { "grad_norm": 0.36739763617515564, "learning_rate": 3.171105425493e-05, "loss": 0.0189, "step": 51060 }, { "grad_norm": 0.17209312319755554, "learning_rate": 3.169181972693557e-05, "loss": 0.0133, "step": 51070 }, { "grad_norm": 0.3288004696369171, "learning_rate": 3.1672588327308404e-05, "loss": 0.0163, "step": 51080 }, { "grad_norm": 0.3994300365447998, "learning_rate": 3.1653360059334646e-05, "loss": 0.0189, "step": 51090 }, { "grad_norm": 0.42305487394332886, "learning_rate": 3.163413492629985e-05, "loss": 0.0309, "step": 51100 }, { "grad_norm": 0.46580880880355835, "learning_rate": 3.161491293148912e-05, "loss": 0.0182, "step": 51110 }, { "grad_norm": 0.27119049429893494, "learning_rate": 3.1595694078186925e-05, "loss": 0.0205, "step": 51120 }, { "grad_norm": 0.2724308967590332, "learning_rate": 3.157647836967724e-05, "loss": 0.0166, "step": 51130 }, { "grad_norm": 0.2914634644985199, "learning_rate": 3.155726580924354e-05, "loss": 0.0189, "step": 51140 }, { "grad_norm": 0.22420501708984375, "learning_rate": 3.1538056400168676e-05, "loss": 0.025, "step": 51150 }, { "grad_norm": 0.15562035143375397, "learning_rate": 3.151885014573506e-05, "loss": 0.0199, "step": 51160 }, { "grad_norm": 0.3041869103908539, "learning_rate": 3.149964704922448e-05, "loss": 0.0234, "step": 51170 }, { "grad_norm": 0.22673849761486053, "learning_rate": 3.148044711391825e-05, "loss": 0.0171, "step": 51180 }, { "grad_norm": 0.7171097993850708, "learning_rate": 3.146125034309707e-05, "loss": 0.0187, "step": 51190 }, { "grad_norm": 0.1954379677772522, "learning_rate": 3.1442056740041195e-05, "loss": 0.0212, "step": 51200 }, { "grad_norm": 0.18406088650226593, "learning_rate": 3.142286630803026e-05, "loss": 0.0275, "step": 51210 }, { "grad_norm": 0.2669873535633087, "learning_rate": 3.140367905034337e-05, "loss": 0.0164, "step": 51220 }, { "grad_norm": 0.30357739329338074, "learning_rate": 3.138449497025914e-05, "loss": 0.0211, "step": 51230 }, { "grad_norm": 0.4500175714492798, "learning_rate": 3.136531407105557e-05, "loss": 0.0184, "step": 51240 }, { "grad_norm": 0.3040817677974701, "learning_rate": 3.1346136356010184e-05, "loss": 0.022, "step": 51250 }, { "grad_norm": 0.15883632004261017, "learning_rate": 3.132696182839988e-05, "loss": 0.0148, "step": 51260 }, { "grad_norm": 0.2986765205860138, "learning_rate": 3.130779049150111e-05, "loss": 0.0197, "step": 51270 }, { "grad_norm": 0.3972116708755493, "learning_rate": 3.128862234858971e-05, "loss": 0.0173, "step": 51280 }, { "grad_norm": 0.23279662430286407, "learning_rate": 3.1269457402941005e-05, "loss": 0.0135, "step": 51290 }, { "grad_norm": 0.33812570571899414, "learning_rate": 3.125029565782974e-05, "loss": 0.0169, "step": 51300 }, { "grad_norm": 0.4605935513973236, "learning_rate": 3.123113711653013e-05, "loss": 0.023, "step": 51310 }, { "grad_norm": 0.23406796157360077, "learning_rate": 3.121198178231587e-05, "loss": 0.0175, "step": 51320 }, { "grad_norm": 0.19748473167419434, "learning_rate": 3.119282965846006e-05, "loss": 0.0172, "step": 51330 }, { "grad_norm": 0.2624271512031555, "learning_rate": 3.1173680748235303e-05, "loss": 0.0218, "step": 51340 }, { "grad_norm": 0.30861926078796387, "learning_rate": 3.115453505491358e-05, "loss": 0.0236, "step": 51350 }, { "grad_norm": 0.27583765983581543, "learning_rate": 3.1135392581766404e-05, "loss": 0.0173, "step": 51360 }, { "grad_norm": 0.23772995173931122, "learning_rate": 3.111625333206467e-05, "loss": 0.0158, "step": 51370 }, { "grad_norm": 0.2345532327890396, "learning_rate": 3.109711730907881e-05, "loss": 0.0146, "step": 51380 }, { "grad_norm": 0.5813909769058228, "learning_rate": 3.107798451607859e-05, "loss": 0.0206, "step": 51390 }, { "grad_norm": 0.18958914279937744, "learning_rate": 3.105885495633328e-05, "loss": 0.0182, "step": 51400 }, { "grad_norm": 0.11539656668901443, "learning_rate": 3.103972863311163e-05, "loss": 0.0138, "step": 51410 }, { "grad_norm": 0.2751528024673462, "learning_rate": 3.102060554968178e-05, "loss": 0.0199, "step": 51420 }, { "grad_norm": 0.24936720728874207, "learning_rate": 3.100148570931137e-05, "loss": 0.0137, "step": 51430 }, { "grad_norm": 0.31332337856292725, "learning_rate": 3.098236911526744e-05, "loss": 0.0175, "step": 51440 }, { "grad_norm": 0.195281520485878, "learning_rate": 3.0963255770816495e-05, "loss": 0.0127, "step": 51450 }, { "grad_norm": 0.2019820362329483, "learning_rate": 3.094414567922447e-05, "loss": 0.0161, "step": 51460 }, { "grad_norm": 0.20381243526935577, "learning_rate": 3.092503884375678e-05, "loss": 0.0186, "step": 51470 }, { "grad_norm": 0.3511813282966614, "learning_rate": 3.0905935267678255e-05, "loss": 0.018, "step": 51480 }, { "grad_norm": 0.22646473348140717, "learning_rate": 3.0886834954253166e-05, "loss": 0.0154, "step": 51490 }, { "grad_norm": 0.28077301383018494, "learning_rate": 3.086773790674524e-05, "loss": 0.0185, "step": 51500 }, { "grad_norm": 0.4026588797569275, "learning_rate": 3.0848644128417617e-05, "loss": 0.0214, "step": 51510 }, { "grad_norm": 0.2810802161693573, "learning_rate": 3.082955362253294e-05, "loss": 0.0174, "step": 51520 }, { "grad_norm": 0.3036336898803711, "learning_rate": 3.081046639235323e-05, "loss": 0.0182, "step": 51530 }, { "grad_norm": 0.2677766978740692, "learning_rate": 3.0791382441139986e-05, "loss": 0.0128, "step": 51540 }, { "grad_norm": 0.18950976431369781, "learning_rate": 3.077230177215411e-05, "loss": 0.0204, "step": 51550 }, { "grad_norm": 0.1288466900587082, "learning_rate": 3.0753224388656e-05, "loss": 0.015, "step": 51560 }, { "grad_norm": 0.336037814617157, "learning_rate": 3.073415029390544e-05, "loss": 0.0283, "step": 51570 }, { "grad_norm": 0.3819410502910614, "learning_rate": 3.071507949116166e-05, "loss": 0.0178, "step": 51580 }, { "grad_norm": 0.24024488031864166, "learning_rate": 3.069601198368337e-05, "loss": 0.0179, "step": 51590 }, { "grad_norm": 0.1660381406545639, "learning_rate": 3.067694777472864e-05, "loss": 0.0178, "step": 51600 }, { "grad_norm": 0.24119365215301514, "learning_rate": 3.065788686755508e-05, "loss": 0.0187, "step": 51610 }, { "grad_norm": 0.2832754850387573, "learning_rate": 3.063882926541961e-05, "loss": 0.0202, "step": 51620 }, { "grad_norm": 0.218192458152771, "learning_rate": 3.061977497157872e-05, "loss": 0.0172, "step": 51630 }, { "grad_norm": 0.22540971636772156, "learning_rate": 3.0600723989288224e-05, "loss": 0.0175, "step": 51640 }, { "grad_norm": 0.46374747157096863, "learning_rate": 3.058167632180343e-05, "loss": 0.0228, "step": 51650 }, { "grad_norm": 0.511691689491272, "learning_rate": 3.0562631972379074e-05, "loss": 0.0249, "step": 51660 }, { "grad_norm": 0.20378422737121582, "learning_rate": 3.0543590944269276e-05, "loss": 0.0174, "step": 51670 }, { "grad_norm": 0.4259452223777771, "learning_rate": 3.052455324072766e-05, "loss": 0.0203, "step": 51680 }, { "grad_norm": 0.2813781797885895, "learning_rate": 3.0505518865007244e-05, "loss": 0.0235, "step": 51690 }, { "grad_norm": 0.21955130994319916, "learning_rate": 3.048648782036048e-05, "loss": 0.0167, "step": 51700 }, { "grad_norm": 0.5544446110725403, "learning_rate": 3.0467460110039236e-05, "loss": 0.0187, "step": 51710 }, { "grad_norm": 0.276056170463562, "learning_rate": 3.0448435737294866e-05, "loss": 0.0181, "step": 51720 }, { "grad_norm": 0.15402226150035858, "learning_rate": 3.042941470537808e-05, "loss": 0.0155, "step": 51730 }, { "grad_norm": 0.22895413637161255, "learning_rate": 3.0410397017539072e-05, "loss": 0.0151, "step": 51740 }, { "grad_norm": 0.40780675411224365, "learning_rate": 3.039138267702742e-05, "loss": 0.0188, "step": 51750 }, { "grad_norm": 0.5802053213119507, "learning_rate": 3.0372371687092193e-05, "loss": 0.0127, "step": 51760 }, { "grad_norm": 0.28543320298194885, "learning_rate": 3.0353364050981823e-05, "loss": 0.0175, "step": 51770 }, { "grad_norm": 0.5813317894935608, "learning_rate": 3.033435977194418e-05, "loss": 0.0189, "step": 51780 }, { "grad_norm": 0.28758999705314636, "learning_rate": 3.0315358853226618e-05, "loss": 0.0207, "step": 51790 }, { "grad_norm": 0.24718809127807617, "learning_rate": 3.0296361298075844e-05, "loss": 0.017, "step": 51800 }, { "grad_norm": 0.47821974754333496, "learning_rate": 3.027736710973803e-05, "loss": 0.0186, "step": 51810 }, { "grad_norm": 0.19170942902565002, "learning_rate": 3.025837629145875e-05, "loss": 0.0162, "step": 51820 }, { "grad_norm": 0.6956285834312439, "learning_rate": 3.0239388846483048e-05, "loss": 0.0184, "step": 51830 }, { "grad_norm": 0.20324158668518066, "learning_rate": 3.022040477805532e-05, "loss": 0.0216, "step": 51840 }, { "grad_norm": 0.3134949505329132, "learning_rate": 3.020142408941946e-05, "loss": 0.0174, "step": 51850 }, { "grad_norm": 0.19778841733932495, "learning_rate": 3.018244678381873e-05, "loss": 0.0162, "step": 51860 }, { "grad_norm": 0.4499208629131317, "learning_rate": 3.0163472864495812e-05, "loss": 0.0247, "step": 51870 }, { "grad_norm": 0.2686356008052826, "learning_rate": 3.014450233469287e-05, "loss": 0.0213, "step": 51880 }, { "grad_norm": 0.3692992627620697, "learning_rate": 3.0125535197651422e-05, "loss": 0.0241, "step": 51890 }, { "grad_norm": 0.23046821355819702, "learning_rate": 3.0106571456612448e-05, "loss": 0.0218, "step": 51900 }, { "grad_norm": 0.2160881608724594, "learning_rate": 3.0087611114816305e-05, "loss": 0.0212, "step": 51910 }, { "grad_norm": 0.16043373942375183, "learning_rate": 3.006865417550284e-05, "loss": 0.0155, "step": 51920 }, { "grad_norm": 0.2427925318479538, "learning_rate": 3.0049700641911242e-05, "loss": 0.0189, "step": 51930 }, { "grad_norm": 0.2701161205768585, "learning_rate": 3.0030750517280175e-05, "loss": 0.0187, "step": 51940 }, { "grad_norm": 0.2606913149356842, "learning_rate": 3.0011803804847682e-05, "loss": 0.0215, "step": 51950 }, { "grad_norm": 0.6918845772743225, "learning_rate": 2.9992860507851224e-05, "loss": 0.0195, "step": 51960 }, { "grad_norm": 0.23961104452610016, "learning_rate": 2.9973920629527723e-05, "loss": 0.0159, "step": 51970 }, { "grad_norm": 0.11748479306697845, "learning_rate": 2.9954984173113453e-05, "loss": 0.0141, "step": 51980 }, { "grad_norm": 0.3016767203807831, "learning_rate": 2.993605114184418e-05, "loss": 0.0207, "step": 51990 }, { "grad_norm": 0.31431472301483154, "learning_rate": 2.9917121538955005e-05, "loss": 0.0213, "step": 52000 }, { "grad_norm": 0.4687821567058563, "learning_rate": 2.98981953676805e-05, "loss": 0.0231, "step": 52010 }, { "grad_norm": 0.4248002767562866, "learning_rate": 2.9879272631254594e-05, "loss": 0.0191, "step": 52020 }, { "grad_norm": 0.19029879570007324, "learning_rate": 2.986035333291073e-05, "loss": 0.0174, "step": 52030 }, { "grad_norm": 0.21742314100265503, "learning_rate": 2.9841437475881646e-05, "loss": 0.0225, "step": 52040 }, { "grad_norm": 0.298805296421051, "learning_rate": 2.9822525063399555e-05, "loss": 0.0181, "step": 52050 }, { "grad_norm": 0.16404010355472565, "learning_rate": 2.9803616098696087e-05, "loss": 0.0239, "step": 52060 }, { "grad_norm": 0.29812178015708923, "learning_rate": 2.9784710585002228e-05, "loss": 0.0187, "step": 52070 }, { "grad_norm": 0.2943961024284363, "learning_rate": 2.9765808525548467e-05, "loss": 0.0188, "step": 52080 }, { "grad_norm": 0.2527499496936798, "learning_rate": 2.974690992356461e-05, "loss": 0.0176, "step": 52090 }, { "grad_norm": 0.1972600668668747, "learning_rate": 2.9728014782279934e-05, "loss": 0.0199, "step": 52100 }, { "grad_norm": 0.27330154180526733, "learning_rate": 2.970912310492307e-05, "loss": 0.018, "step": 52110 }, { "grad_norm": 0.2752957344055176, "learning_rate": 2.9690234894722134e-05, "loss": 0.0151, "step": 52120 }, { "grad_norm": 0.2272021323442459, "learning_rate": 2.9671350154904577e-05, "loss": 0.0231, "step": 52130 }, { "grad_norm": 0.26751890778541565, "learning_rate": 2.9652468888697282e-05, "loss": 0.0183, "step": 52140 }, { "grad_norm": 1.759502649307251, "learning_rate": 2.9633591099326562e-05, "loss": 0.0196, "step": 52150 }, { "grad_norm": 0.19142477214336395, "learning_rate": 2.9614716790018072e-05, "loss": 0.0169, "step": 52160 }, { "grad_norm": 0.2149658054113388, "learning_rate": 2.959584596399697e-05, "loss": 0.0137, "step": 52170 }, { "grad_norm": 0.30263081192970276, "learning_rate": 2.9576978624487717e-05, "loss": 0.0225, "step": 52180 }, { "grad_norm": 0.2636680006980896, "learning_rate": 2.9558114774714263e-05, "loss": 0.0164, "step": 52190 }, { "grad_norm": 0.3049285411834717, "learning_rate": 2.9539254417899897e-05, "loss": 0.0207, "step": 52200 }, { "grad_norm": 0.24675841629505157, "learning_rate": 2.9520397557267365e-05, "loss": 0.0206, "step": 52210 }, { "grad_norm": 0.28429114818573, "learning_rate": 2.9501544196038765e-05, "loss": 0.0221, "step": 52220 }, { "grad_norm": 0.17302203178405762, "learning_rate": 2.9482694337435614e-05, "loss": 0.0236, "step": 52230 }, { "grad_norm": 0.35576701164245605, "learning_rate": 2.946384798467887e-05, "loss": 0.0152, "step": 52240 }, { "grad_norm": 0.20146571099758148, "learning_rate": 2.9445005140988824e-05, "loss": 0.0149, "step": 52250 }, { "grad_norm": 0.30674290657043457, "learning_rate": 2.942616580958524e-05, "loss": 0.0182, "step": 52260 }, { "grad_norm": 0.48081174492836, "learning_rate": 2.9407329993687193e-05, "loss": 0.0181, "step": 52270 }, { "grad_norm": 0.339038223028183, "learning_rate": 2.938849769651326e-05, "loss": 0.0145, "step": 52280 }, { "grad_norm": 0.24560096859931946, "learning_rate": 2.9369668921281345e-05, "loss": 0.0169, "step": 52290 }, { "grad_norm": 0.19174998998641968, "learning_rate": 2.9350843671208773e-05, "loss": 0.0132, "step": 52300 }, { "grad_norm": 0.2655886709690094, "learning_rate": 2.9332021949512245e-05, "loss": 0.0146, "step": 52310 }, { "grad_norm": 0.39721450209617615, "learning_rate": 2.9313203759407908e-05, "loss": 0.0175, "step": 52320 }, { "grad_norm": 0.2300165444612503, "learning_rate": 2.929438910411127e-05, "loss": 0.0262, "step": 52330 }, { "grad_norm": 0.6562933325767517, "learning_rate": 2.927557798683722e-05, "loss": 0.0165, "step": 52340 }, { "grad_norm": 0.4122924506664276, "learning_rate": 2.9256770410800095e-05, "loss": 0.0178, "step": 52350 }, { "grad_norm": 0.21707770228385925, "learning_rate": 2.9237966379213554e-05, "loss": 0.0196, "step": 52360 }, { "grad_norm": 0.3174677789211273, "learning_rate": 2.9219165895290736e-05, "loss": 0.0194, "step": 52370 }, { "grad_norm": 0.35913681983947754, "learning_rate": 2.9200368962244083e-05, "loss": 0.0197, "step": 52380 }, { "grad_norm": 0.22352401912212372, "learning_rate": 2.9181575583285536e-05, "loss": 0.0145, "step": 52390 }, { "grad_norm": 0.2844807207584381, "learning_rate": 2.916278576162632e-05, "loss": 0.0199, "step": 52400 }, { "grad_norm": 0.2348972111940384, "learning_rate": 2.9143999500477116e-05, "loss": 0.0179, "step": 52410 }, { "grad_norm": 0.26194775104522705, "learning_rate": 2.9125216803048004e-05, "loss": 0.0217, "step": 52420 }, { "grad_norm": 0.20002175867557526, "learning_rate": 2.9106437672548403e-05, "loss": 0.0177, "step": 52430 }, { "grad_norm": 0.49754586815834045, "learning_rate": 2.9087662112187154e-05, "loss": 0.0138, "step": 52440 }, { "grad_norm": 0.2645229697227478, "learning_rate": 2.9068890125172498e-05, "loss": 0.0159, "step": 52450 }, { "grad_norm": 0.4206070005893707, "learning_rate": 2.9050121714712058e-05, "loss": 0.0212, "step": 52460 }, { "grad_norm": 0.37212857604026794, "learning_rate": 2.9031356884012835e-05, "loss": 0.0194, "step": 52470 }, { "grad_norm": 0.16043822467327118, "learning_rate": 2.9012595636281247e-05, "loss": 0.02, "step": 52480 }, { "grad_norm": 0.1814078390598297, "learning_rate": 2.8993837974723016e-05, "loss": 0.0256, "step": 52490 }, { "grad_norm": 0.16242621839046478, "learning_rate": 2.8975083902543394e-05, "loss": 0.0144, "step": 52500 }, { "grad_norm": 0.2279634028673172, "learning_rate": 2.895633342294688e-05, "loss": 0.018, "step": 52510 }, { "grad_norm": 0.30133169889450073, "learning_rate": 2.893758653913744e-05, "loss": 0.0151, "step": 52520 }, { "grad_norm": 0.21141977608203888, "learning_rate": 2.8918843254318423e-05, "loss": 0.0171, "step": 52530 }, { "grad_norm": 0.30085325241088867, "learning_rate": 2.8900103571692483e-05, "loss": 0.0169, "step": 52540 }, { "grad_norm": 0.2383907437324524, "learning_rate": 2.8881367494461796e-05, "loss": 0.0136, "step": 52550 }, { "grad_norm": 0.3107853829860687, "learning_rate": 2.886263502582779e-05, "loss": 0.0224, "step": 52560 }, { "grad_norm": 0.29102835059165955, "learning_rate": 2.8843906168991353e-05, "loss": 0.0189, "step": 52570 }, { "grad_norm": 0.3378238081932068, "learning_rate": 2.8825180927152728e-05, "loss": 0.0199, "step": 52580 }, { "grad_norm": 0.33086958527565, "learning_rate": 2.8806459303511547e-05, "loss": 0.0151, "step": 52590 }, { "grad_norm": 0.40496864914894104, "learning_rate": 2.8787741301266852e-05, "loss": 0.0194, "step": 52600 }, { "grad_norm": 0.2603248059749603, "learning_rate": 2.876902692361697e-05, "loss": 0.0166, "step": 52610 }, { "grad_norm": 0.2850802540779114, "learning_rate": 2.875031617375976e-05, "loss": 0.0177, "step": 52620 }, { "grad_norm": 0.19337080419063568, "learning_rate": 2.8731609054892318e-05, "loss": 0.0163, "step": 52630 }, { "grad_norm": 0.17910587787628174, "learning_rate": 2.871290557021119e-05, "loss": 0.0163, "step": 52640 }, { "grad_norm": 0.2950170040130615, "learning_rate": 2.8694205722912298e-05, "loss": 0.0168, "step": 52650 }, { "grad_norm": 0.4726533591747284, "learning_rate": 2.8675509516190936e-05, "loss": 0.0207, "step": 52660 }, { "grad_norm": 0.29980307817459106, "learning_rate": 2.8656816953241772e-05, "loss": 0.0239, "step": 52670 }, { "grad_norm": 0.2753871977329254, "learning_rate": 2.863812803725887e-05, "loss": 0.0151, "step": 52680 }, { "grad_norm": 0.2467050850391388, "learning_rate": 2.8619442771435623e-05, "loss": 0.0146, "step": 52690 }, { "grad_norm": 0.18143990635871887, "learning_rate": 2.860076115896484e-05, "loss": 0.025, "step": 52700 }, { "grad_norm": 0.3298426866531372, "learning_rate": 2.8582083203038712e-05, "loss": 0.0209, "step": 52710 }, { "grad_norm": 0.23959068953990936, "learning_rate": 2.8563408906848778e-05, "loss": 0.0142, "step": 52720 }, { "grad_norm": 0.2747787535190582, "learning_rate": 2.8544738273585993e-05, "loss": 0.0166, "step": 52730 }, { "grad_norm": 0.24150937795639038, "learning_rate": 2.8526071306440595e-05, "loss": 0.0136, "step": 52740 }, { "grad_norm": 0.2882041931152344, "learning_rate": 2.850740800860233e-05, "loss": 0.0146, "step": 52750 }, { "grad_norm": 0.22863346338272095, "learning_rate": 2.8488748383260178e-05, "loss": 0.0151, "step": 52760 }, { "grad_norm": 0.2943669855594635, "learning_rate": 2.8470092433602635e-05, "loss": 0.0206, "step": 52770 }, { "grad_norm": 0.33757737278938293, "learning_rate": 2.845144016281745e-05, "loss": 0.0172, "step": 52780 }, { "grad_norm": 0.20133332908153534, "learning_rate": 2.8432791574091744e-05, "loss": 0.0231, "step": 52790 }, { "grad_norm": 0.28593137860298157, "learning_rate": 2.8414146670612134e-05, "loss": 0.0173, "step": 52800 }, { "grad_norm": 0.16858837008476257, "learning_rate": 2.8395505455564446e-05, "loss": 0.0231, "step": 52810 }, { "grad_norm": 0.2136404812335968, "learning_rate": 2.837686793213403e-05, "loss": 0.0186, "step": 52820 }, { "grad_norm": 0.2650343179702759, "learning_rate": 2.8358234103505477e-05, "loss": 0.0165, "step": 52830 }, { "grad_norm": 0.33597713708877563, "learning_rate": 2.8339603972862806e-05, "loss": 0.0261, "step": 52840 }, { "grad_norm": 0.3465748131275177, "learning_rate": 2.83209775433894e-05, "loss": 0.0205, "step": 52850 }, { "grad_norm": 0.20344477891921997, "learning_rate": 2.8302354818268013e-05, "loss": 0.0153, "step": 52860 }, { "grad_norm": 0.2600327730178833, "learning_rate": 2.8283735800680754e-05, "loss": 0.023, "step": 52870 }, { "grad_norm": 0.15018242597579956, "learning_rate": 2.826512049380913e-05, "loss": 0.0169, "step": 52880 }, { "grad_norm": 0.24082280695438385, "learning_rate": 2.8246508900833935e-05, "loss": 0.0161, "step": 52890 }, { "grad_norm": 0.3262383043766022, "learning_rate": 2.822790102493541e-05, "loss": 0.0185, "step": 52900 }, { "grad_norm": 0.15278005599975586, "learning_rate": 2.820929686929314e-05, "loss": 0.0194, "step": 52910 }, { "grad_norm": 0.20137055218219757, "learning_rate": 2.819069643708605e-05, "loss": 0.0133, "step": 52920 }, { "grad_norm": 0.2895924150943756, "learning_rate": 2.8172099731492484e-05, "loss": 0.0184, "step": 52930 }, { "grad_norm": 0.23877602815628052, "learning_rate": 2.815350675569003e-05, "loss": 0.0243, "step": 52940 }, { "grad_norm": 0.23970402777194977, "learning_rate": 2.8134917512855807e-05, "loss": 0.0245, "step": 52950 }, { "grad_norm": 0.23495735228061676, "learning_rate": 2.8116332006166134e-05, "loss": 0.0127, "step": 52960 }, { "grad_norm": 0.309909462928772, "learning_rate": 2.809775023879685e-05, "loss": 0.0269, "step": 52970 }, { "grad_norm": 0.299437016248703, "learning_rate": 2.807917221392299e-05, "loss": 0.0212, "step": 52980 }, { "grad_norm": 0.3282430171966553, "learning_rate": 2.8060597934719067e-05, "loss": 0.0157, "step": 52990 }, { "grad_norm": 0.19882044196128845, "learning_rate": 2.8042027404358927e-05, "loss": 0.0177, "step": 53000 }, { "grad_norm": 0.38757044076919556, "learning_rate": 2.802346062601572e-05, "loss": 0.0185, "step": 53010 }, { "grad_norm": 0.4617633819580078, "learning_rate": 2.800489760286207e-05, "loss": 0.03, "step": 53020 }, { "grad_norm": 0.3215133249759674, "learning_rate": 2.7986338338069834e-05, "loss": 0.017, "step": 53030 }, { "grad_norm": 0.3075254261493683, "learning_rate": 2.7967782834810297e-05, "loss": 0.0247, "step": 53040 }, { "grad_norm": 0.26733922958374023, "learning_rate": 2.794923109625409e-05, "loss": 0.0181, "step": 53050 }, { "grad_norm": 0.2057257890701294, "learning_rate": 2.7930683125571205e-05, "loss": 0.0185, "step": 53060 }, { "grad_norm": 0.3706252872943878, "learning_rate": 2.7912138925930997e-05, "loss": 0.025, "step": 53070 }, { "grad_norm": 0.2448081076145172, "learning_rate": 2.7893598500502117e-05, "loss": 0.0281, "step": 53080 }, { "grad_norm": 0.33187949657440186, "learning_rate": 2.7875061852452644e-05, "loss": 0.0168, "step": 53090 }, { "grad_norm": 0.3727461099624634, "learning_rate": 2.7856528984949982e-05, "loss": 0.0223, "step": 53100 }, { "grad_norm": 0.17012961208820343, "learning_rate": 2.7837999901160888e-05, "loss": 0.0153, "step": 53110 }, { "grad_norm": 0.3191143870353699, "learning_rate": 2.7819474604251484e-05, "loss": 0.0191, "step": 53120 }, { "grad_norm": 0.21514630317687988, "learning_rate": 2.780095309738725e-05, "loss": 0.0171, "step": 53130 }, { "grad_norm": 0.28409403562545776, "learning_rate": 2.778243538373294e-05, "loss": 0.02, "step": 53140 }, { "grad_norm": 0.16228055953979492, "learning_rate": 2.7763921466452826e-05, "loss": 0.017, "step": 53150 }, { "grad_norm": 0.3143168091773987, "learning_rate": 2.7745411348710336e-05, "loss": 0.0262, "step": 53160 }, { "grad_norm": 0.33728450536727905, "learning_rate": 2.7726905033668395e-05, "loss": 0.0237, "step": 53170 }, { "grad_norm": 0.2962724566459656, "learning_rate": 2.7708402524489214e-05, "loss": 0.0236, "step": 53180 }, { "grad_norm": 0.2713715732097626, "learning_rate": 2.7689903824334364e-05, "loss": 0.0202, "step": 53190 }, { "grad_norm": 0.18097974359989166, "learning_rate": 2.7671408936364785e-05, "loss": 0.0138, "step": 53200 }, { "grad_norm": 0.24816592037677765, "learning_rate": 2.765291786374069e-05, "loss": 0.0175, "step": 53210 }, { "grad_norm": 0.35499244928359985, "learning_rate": 2.7634430609621775e-05, "loss": 0.0165, "step": 53220 }, { "grad_norm": 0.3874337375164032, "learning_rate": 2.7615947177166956e-05, "loss": 0.0182, "step": 53230 }, { "grad_norm": 0.5010690689086914, "learning_rate": 2.7597467569534553e-05, "loss": 0.0126, "step": 53240 }, { "grad_norm": 0.2293369472026825, "learning_rate": 2.757899178988226e-05, "loss": 0.0162, "step": 53250 }, { "grad_norm": 0.30429649353027344, "learning_rate": 2.7560519841367005e-05, "loss": 0.0183, "step": 53260 }, { "grad_norm": 0.15334029495716095, "learning_rate": 2.7542051727145237e-05, "loss": 0.0178, "step": 53270 }, { "grad_norm": 0.29177480936050415, "learning_rate": 2.7523587450372578e-05, "loss": 0.0214, "step": 53280 }, { "grad_norm": 0.44228118658065796, "learning_rate": 2.750512701420409e-05, "loss": 0.0171, "step": 53290 }, { "grad_norm": 0.4713383913040161, "learning_rate": 2.7486670421794158e-05, "loss": 0.0149, "step": 53300 }, { "grad_norm": 0.23177067935466766, "learning_rate": 2.7468217676296515e-05, "loss": 0.0247, "step": 53310 }, { "grad_norm": 0.3271923065185547, "learning_rate": 2.7449768780864226e-05, "loss": 0.0225, "step": 53320 }, { "grad_norm": 0.32363077998161316, "learning_rate": 2.7431323738649724e-05, "loss": 0.0166, "step": 53330 }, { "grad_norm": 0.3958829343318939, "learning_rate": 2.7412882552804713e-05, "loss": 0.0184, "step": 53340 }, { "grad_norm": 0.21411731839179993, "learning_rate": 2.739444522648032e-05, "loss": 0.0114, "step": 53350 }, { "grad_norm": 0.23596890270709991, "learning_rate": 2.737601176282697e-05, "loss": 0.0158, "step": 53360 }, { "grad_norm": 0.31003227829933167, "learning_rate": 2.735758216499445e-05, "loss": 0.0163, "step": 53370 }, { "grad_norm": 0.2373497039079666, "learning_rate": 2.7339156436131864e-05, "loss": 0.0132, "step": 53380 }, { "grad_norm": 0.32017838954925537, "learning_rate": 2.7320734579387663e-05, "loss": 0.0185, "step": 53390 }, { "grad_norm": 0.4041140675544739, "learning_rate": 2.730231659790966e-05, "loss": 0.0139, "step": 53400 }, { "grad_norm": 0.32393741607666016, "learning_rate": 2.728390249484494e-05, "loss": 0.0228, "step": 53410 }, { "grad_norm": 0.19540417194366455, "learning_rate": 2.7265492273340032e-05, "loss": 0.0144, "step": 53420 }, { "grad_norm": 0.2883269786834717, "learning_rate": 2.7247085936540678e-05, "loss": 0.0144, "step": 53430 }, { "grad_norm": 0.4085145592689514, "learning_rate": 2.722868348759205e-05, "loss": 0.0194, "step": 53440 }, { "grad_norm": 0.19195541739463806, "learning_rate": 2.7210284929638635e-05, "loss": 0.0209, "step": 53450 }, { "grad_norm": 0.20391976833343506, "learning_rate": 2.7191890265824183e-05, "loss": 0.0164, "step": 53460 }, { "grad_norm": 0.2609870135784149, "learning_rate": 2.7173499499291926e-05, "loss": 0.0143, "step": 53470 }, { "grad_norm": 0.38911283016204834, "learning_rate": 2.7155112633184277e-05, "loss": 0.0148, "step": 53480 }, { "grad_norm": 0.28632310032844543, "learning_rate": 2.713672967064307e-05, "loss": 0.0218, "step": 53490 }, { "grad_norm": 0.2036132961511612, "learning_rate": 2.711835061480945e-05, "loss": 0.0116, "step": 53500 }, { "grad_norm": 0.30086544156074524, "learning_rate": 2.7099975468823896e-05, "loss": 0.0203, "step": 53510 }, { "grad_norm": 0.35462066531181335, "learning_rate": 2.708160423582622e-05, "loss": 0.0183, "step": 53520 }, { "grad_norm": 0.5058711171150208, "learning_rate": 2.706323691895557e-05, "loss": 0.0237, "step": 53530 }, { "grad_norm": 0.4786553978919983, "learning_rate": 2.70448735213504e-05, "loss": 0.0151, "step": 53540 }, { "grad_norm": 0.27309954166412354, "learning_rate": 2.702651404614852e-05, "loss": 0.0179, "step": 53550 }, { "grad_norm": 0.27998173236846924, "learning_rate": 2.7008158496487056e-05, "loss": 0.0174, "step": 53560 }, { "grad_norm": 0.2544194161891937, "learning_rate": 2.6989806875502487e-05, "loss": 0.0151, "step": 53570 }, { "grad_norm": 0.3234808146953583, "learning_rate": 2.6971459186330584e-05, "loss": 0.0147, "step": 53580 }, { "grad_norm": 0.3736848533153534, "learning_rate": 2.695311543210648e-05, "loss": 0.0174, "step": 53590 }, { "grad_norm": 0.3043062686920166, "learning_rate": 2.693477561596463e-05, "loss": 0.0139, "step": 53600 }, { "grad_norm": 0.3087843060493469, "learning_rate": 2.6916439741038756e-05, "loss": 0.0162, "step": 53610 }, { "grad_norm": 0.3267471194267273, "learning_rate": 2.689810781046203e-05, "loss": 0.0167, "step": 53620 }, { "grad_norm": 0.1914670765399933, "learning_rate": 2.6879779827366823e-05, "loss": 0.016, "step": 53630 }, { "grad_norm": 0.16493122279644012, "learning_rate": 2.6861455794884904e-05, "loss": 0.0174, "step": 53640 }, { "grad_norm": 0.24862365424633026, "learning_rate": 2.6843135716147373e-05, "loss": 0.0164, "step": 53650 }, { "grad_norm": 0.2274273782968521, "learning_rate": 2.6824819594284556e-05, "loss": 0.0203, "step": 53660 }, { "grad_norm": 0.2649596631526947, "learning_rate": 2.6806507432426275e-05, "loss": 0.018, "step": 53670 }, { "grad_norm": 0.7283565998077393, "learning_rate": 2.6788199233701512e-05, "loss": 0.0173, "step": 53680 }, { "grad_norm": 0.28886809945106506, "learning_rate": 2.6769895001238652e-05, "loss": 0.0212, "step": 53690 }, { "grad_norm": 0.3076779544353485, "learning_rate": 2.67515947381654e-05, "loss": 0.0196, "step": 53700 }, { "grad_norm": 0.16778706014156342, "learning_rate": 2.6733298447608745e-05, "loss": 0.0138, "step": 53710 }, { "grad_norm": 0.40401068329811096, "learning_rate": 2.6715006132695074e-05, "loss": 0.0204, "step": 53720 }, { "grad_norm": 0.30052006244659424, "learning_rate": 2.6696717796549985e-05, "loss": 0.024, "step": 53730 }, { "grad_norm": 0.31065428256988525, "learning_rate": 2.667843344229848e-05, "loss": 0.0152, "step": 53740 }, { "grad_norm": 0.3580126464366913, "learning_rate": 2.6660153073064846e-05, "loss": 0.0202, "step": 53750 }, { "grad_norm": 0.7271293997764587, "learning_rate": 2.664187669197271e-05, "loss": 0.0238, "step": 53760 }, { "grad_norm": 0.3848721385002136, "learning_rate": 2.6623604302145005e-05, "loss": 0.0149, "step": 53770 }, { "grad_norm": 0.31155431270599365, "learning_rate": 2.6605335906703975e-05, "loss": 0.0174, "step": 53780 }, { "grad_norm": 0.28235942125320435, "learning_rate": 2.65870715087712e-05, "loss": 0.019, "step": 53790 }, { "grad_norm": 0.5699800252914429, "learning_rate": 2.6568811111467573e-05, "loss": 0.0186, "step": 53800 }, { "grad_norm": 0.20767544209957123, "learning_rate": 2.6550554717913258e-05, "loss": 0.0154, "step": 53810 }, { "grad_norm": 0.21371662616729736, "learning_rate": 2.6532302331227805e-05, "loss": 0.0167, "step": 53820 }, { "grad_norm": 0.2966039478778839, "learning_rate": 2.651405395453004e-05, "loss": 0.0204, "step": 53830 }, { "grad_norm": 0.22353696823120117, "learning_rate": 2.6495809590938115e-05, "loss": 0.0207, "step": 53840 }, { "grad_norm": 0.15066681802272797, "learning_rate": 2.64775692435695e-05, "loss": 0.0173, "step": 53850 }, { "grad_norm": 0.23965726792812347, "learning_rate": 2.6459332915540928e-05, "loss": 0.0158, "step": 53860 }, { "grad_norm": 0.2776864171028137, "learning_rate": 2.644110060996856e-05, "loss": 0.0218, "step": 53870 }, { "grad_norm": 0.23846447467803955, "learning_rate": 2.642287232996774e-05, "loss": 0.0194, "step": 53880 }, { "grad_norm": 0.28172677755355835, "learning_rate": 2.6404648078653205e-05, "loss": 0.0133, "step": 53890 }, { "grad_norm": 0.41198405623435974, "learning_rate": 2.6386427859139002e-05, "loss": 0.0171, "step": 53900 }, { "grad_norm": 0.30300652980804443, "learning_rate": 2.63682116745384e-05, "loss": 0.0186, "step": 53910 }, { "grad_norm": 0.23370714485645294, "learning_rate": 2.6349999527964138e-05, "loss": 0.0141, "step": 53920 }, { "grad_norm": 0.24148564040660858, "learning_rate": 2.63317914225281e-05, "loss": 0.0198, "step": 53930 }, { "grad_norm": 0.15838034451007843, "learning_rate": 2.6313587361341585e-05, "loss": 0.0182, "step": 53940 }, { "grad_norm": 0.5163163542747498, "learning_rate": 2.6295387347515165e-05, "loss": 0.0221, "step": 53950 }, { "grad_norm": 0.20668445527553558, "learning_rate": 2.6277191384158727e-05, "loss": 0.0237, "step": 53960 }, { "grad_norm": 0.2734909951686859, "learning_rate": 2.6258999474381457e-05, "loss": 0.0183, "step": 53970 }, { "grad_norm": 0.25722289085388184, "learning_rate": 2.624081162129186e-05, "loss": 0.0199, "step": 53980 }, { "grad_norm": 0.19284965097904205, "learning_rate": 2.6222627827997765e-05, "loss": 0.0274, "step": 53990 }, { "grad_norm": 0.23714280128479004, "learning_rate": 2.6204448097606236e-05, "loss": 0.0174, "step": 54000 }, { "grad_norm": 0.21952931582927704, "learning_rate": 2.6186272433223726e-05, "loss": 0.0215, "step": 54010 }, { "grad_norm": 0.1636437177658081, "learning_rate": 2.6168100837955943e-05, "loss": 0.0192, "step": 54020 }, { "grad_norm": 0.19688504934310913, "learning_rate": 2.6149933314907926e-05, "loss": 0.0283, "step": 54030 }, { "grad_norm": 0.3338499367237091, "learning_rate": 2.6131769867184e-05, "loss": 0.0244, "step": 54040 }, { "grad_norm": 0.3374477028846741, "learning_rate": 2.611361049788783e-05, "loss": 0.0193, "step": 54050 }, { "grad_norm": 0.328678160905838, "learning_rate": 2.6095455210122292e-05, "loss": 0.0196, "step": 54060 }, { "grad_norm": 0.13278311491012573, "learning_rate": 2.6077304006989712e-05, "loss": 0.0115, "step": 54070 }, { "grad_norm": 0.462645024061203, "learning_rate": 2.6059156891591562e-05, "loss": 0.0179, "step": 54080 }, { "grad_norm": 0.26169341802597046, "learning_rate": 2.6041013867028718e-05, "loss": 0.0166, "step": 54090 }, { "grad_norm": 0.28015372157096863, "learning_rate": 2.6022874936401347e-05, "loss": 0.013, "step": 54100 }, { "grad_norm": 0.18424324691295624, "learning_rate": 2.6004740102808832e-05, "loss": 0.0194, "step": 54110 }, { "grad_norm": 0.284726083278656, "learning_rate": 2.598660936935e-05, "loss": 0.0115, "step": 54120 }, { "grad_norm": 0.33512917160987854, "learning_rate": 2.5968482739122845e-05, "loss": 0.0171, "step": 54130 }, { "grad_norm": 0.2822161912918091, "learning_rate": 2.595036021522472e-05, "loss": 0.0161, "step": 54140 }, { "grad_norm": 0.411422461271286, "learning_rate": 2.5932241800752278e-05, "loss": 0.02, "step": 54150 }, { "grad_norm": 0.5106940865516663, "learning_rate": 2.5914127498801453e-05, "loss": 0.031, "step": 54160 }, { "grad_norm": 0.20491495728492737, "learning_rate": 2.5896017312467497e-05, "loss": 0.0169, "step": 54170 }, { "grad_norm": 0.31058838963508606, "learning_rate": 2.587791124484493e-05, "loss": 0.0145, "step": 54180 }, { "grad_norm": 0.2410762459039688, "learning_rate": 2.5859809299027615e-05, "loss": 0.0151, "step": 54190 }, { "grad_norm": 0.27911800146102905, "learning_rate": 2.5841711478108632e-05, "loss": 0.0272, "step": 54200 }, { "grad_norm": 0.2942693829536438, "learning_rate": 2.582361778518043e-05, "loss": 0.0164, "step": 54210 }, { "grad_norm": 0.22391429543495178, "learning_rate": 2.580552822333472e-05, "loss": 0.0173, "step": 54220 }, { "grad_norm": 0.4582223892211914, "learning_rate": 2.578744279566252e-05, "loss": 0.0211, "step": 54230 }, { "grad_norm": 0.1571091115474701, "learning_rate": 2.576936150525413e-05, "loss": 0.0168, "step": 54240 }, { "grad_norm": 0.27790290117263794, "learning_rate": 2.5751284355199168e-05, "loss": 0.0146, "step": 54250 }, { "grad_norm": 0.1557525396347046, "learning_rate": 2.573321134858646e-05, "loss": 0.0102, "step": 54260 }, { "grad_norm": 0.16930891573429108, "learning_rate": 2.5715142488504286e-05, "loss": 0.013, "step": 54270 }, { "grad_norm": 0.18745896220207214, "learning_rate": 2.5697077778040042e-05, "loss": 0.0177, "step": 54280 }, { "grad_norm": 0.3104023337364197, "learning_rate": 2.5679017220280522e-05, "loss": 0.0322, "step": 54290 }, { "grad_norm": 0.4150485098361969, "learning_rate": 2.5660960818311796e-05, "loss": 0.0172, "step": 54300 }, { "grad_norm": 0.31452295184135437, "learning_rate": 2.564290857521915e-05, "loss": 0.0161, "step": 54310 }, { "grad_norm": 0.18637312948703766, "learning_rate": 2.5624860494087298e-05, "loss": 0.0176, "step": 54320 }, { "grad_norm": 0.147558331489563, "learning_rate": 2.5606816578000115e-05, "loss": 0.0198, "step": 54330 }, { "grad_norm": 0.2893131971359253, "learning_rate": 2.558877683004082e-05, "loss": 0.0144, "step": 54340 }, { "grad_norm": 0.20164550840854645, "learning_rate": 2.557074125329192e-05, "loss": 0.0151, "step": 54350 }, { "grad_norm": 0.25578686594963074, "learning_rate": 2.5552709850835195e-05, "loss": 0.0151, "step": 54360 }, { "grad_norm": 0.21532978117465973, "learning_rate": 2.5534682625751738e-05, "loss": 0.0104, "step": 54370 }, { "grad_norm": 0.3224022090435028, "learning_rate": 2.551665958112186e-05, "loss": 0.0216, "step": 54380 }, { "grad_norm": 0.18509872257709503, "learning_rate": 2.549864072002527e-05, "loss": 0.0119, "step": 54390 }, { "grad_norm": 0.24535320699214935, "learning_rate": 2.5480626045540858e-05, "loss": 0.0218, "step": 54400 }, { "grad_norm": 0.2320934385061264, "learning_rate": 2.546261556074684e-05, "loss": 0.0162, "step": 54410 }, { "grad_norm": 0.2412865161895752, "learning_rate": 2.5444609268720726e-05, "loss": 0.0155, "step": 54420 }, { "grad_norm": 0.2673164904117584, "learning_rate": 2.5426607172539297e-05, "loss": 0.0154, "step": 54430 }, { "grad_norm": 0.27578821778297424, "learning_rate": 2.5408609275278617e-05, "loss": 0.0229, "step": 54440 }, { "grad_norm": 0.1851266622543335, "learning_rate": 2.5390615580014055e-05, "loss": 0.0158, "step": 54450 }, { "grad_norm": 0.23968756198883057, "learning_rate": 2.5372626089820207e-05, "loss": 0.0211, "step": 54460 }, { "grad_norm": 0.30766773223876953, "learning_rate": 2.5354640807770997e-05, "loss": 0.0159, "step": 54470 }, { "grad_norm": 0.3780606985092163, "learning_rate": 2.5336659736939622e-05, "loss": 0.015, "step": 54480 }, { "grad_norm": 0.20136035978794098, "learning_rate": 2.5318682880398554e-05, "loss": 0.0195, "step": 54490 }, { "grad_norm": 0.21788989007472992, "learning_rate": 2.530071024121956e-05, "loss": 0.0192, "step": 54500 }, { "grad_norm": 0.2292136698961258, "learning_rate": 2.5282741822473627e-05, "loss": 0.014, "step": 54510 }, { "grad_norm": 0.3178017735481262, "learning_rate": 2.526477762723114e-05, "loss": 0.0161, "step": 54520 }, { "grad_norm": 0.22401265799999237, "learning_rate": 2.5246817658561618e-05, "loss": 0.0183, "step": 54530 }, { "grad_norm": 0.15770091116428375, "learning_rate": 2.5228861919533965e-05, "loss": 0.0172, "step": 54540 }, { "grad_norm": 0.20336084067821503, "learning_rate": 2.5210910413216326e-05, "loss": 0.0266, "step": 54550 }, { "grad_norm": 0.3564149737358093, "learning_rate": 2.5192963142676086e-05, "loss": 0.0135, "step": 54560 }, { "grad_norm": 0.2333536148071289, "learning_rate": 2.517502011098001e-05, "loss": 0.0149, "step": 54570 }, { "grad_norm": 0.1865847259759903, "learning_rate": 2.5157081321193987e-05, "loss": 0.0155, "step": 54580 }, { "grad_norm": 0.25180286169052124, "learning_rate": 2.5139146776383356e-05, "loss": 0.0191, "step": 54590 }, { "grad_norm": 0.19040291011333466, "learning_rate": 2.5121216479612575e-05, "loss": 0.0157, "step": 54600 }, { "grad_norm": 0.23314212262630463, "learning_rate": 2.510329043394546e-05, "loss": 0.024, "step": 54610 }, { "grad_norm": 0.3094766139984131, "learning_rate": 2.508536864244508e-05, "loss": 0.023, "step": 54620 }, { "grad_norm": 0.18151073157787323, "learning_rate": 2.5067451108173778e-05, "loss": 0.016, "step": 54630 }, { "grad_norm": 0.2394673228263855, "learning_rate": 2.5049537834193204e-05, "loss": 0.0164, "step": 54640 }, { "grad_norm": 0.8064454793930054, "learning_rate": 2.5031628823564194e-05, "loss": 0.016, "step": 54650 }, { "grad_norm": 0.24409332871437073, "learning_rate": 2.5013724079346933e-05, "loss": 0.0187, "step": 54660 }, { "grad_norm": 0.15804189443588257, "learning_rate": 2.4995823604600854e-05, "loss": 0.0138, "step": 54670 }, { "grad_norm": 0.21270114183425903, "learning_rate": 2.497792740238465e-05, "loss": 0.0152, "step": 54680 }, { "grad_norm": 0.2682551443576813, "learning_rate": 2.49600354757563e-05, "loss": 0.0195, "step": 54690 }, { "grad_norm": 0.27385374903678894, "learning_rate": 2.494214782777306e-05, "loss": 0.0175, "step": 54700 }, { "grad_norm": 0.3367217779159546, "learning_rate": 2.4924264461491386e-05, "loss": 0.0195, "step": 54710 }, { "grad_norm": 0.2937018573284149, "learning_rate": 2.4906385379967133e-05, "loss": 0.0205, "step": 54720 }, { "grad_norm": 0.22061313688755035, "learning_rate": 2.4888510586255285e-05, "loss": 0.012, "step": 54730 }, { "grad_norm": 0.1885804384946823, "learning_rate": 2.487064008341018e-05, "loss": 0.0164, "step": 54740 }, { "grad_norm": 0.46810382604599, "learning_rate": 2.4852773874485407e-05, "loss": 0.0166, "step": 54750 }, { "grad_norm": 0.2591363489627838, "learning_rate": 2.483491196253377e-05, "loss": 0.0164, "step": 54760 }, { "grad_norm": 0.2346961796283722, "learning_rate": 2.4817054350607443e-05, "loss": 0.018, "step": 54770 }, { "grad_norm": 0.4028240144252777, "learning_rate": 2.4799201041757742e-05, "loss": 0.0222, "step": 54780 }, { "grad_norm": 0.3168281614780426, "learning_rate": 2.4781352039035373e-05, "loss": 0.0165, "step": 54790 }, { "grad_norm": 0.19168773293495178, "learning_rate": 2.4763507345490194e-05, "loss": 0.0129, "step": 54800 }, { "grad_norm": 0.19793051481246948, "learning_rate": 2.4745666964171386e-05, "loss": 0.0169, "step": 54810 }, { "grad_norm": 0.6066384315490723, "learning_rate": 2.4727830898127387e-05, "loss": 0.0174, "step": 54820 }, { "grad_norm": 0.20200401544570923, "learning_rate": 2.4709999150405895e-05, "loss": 0.0225, "step": 54830 }, { "grad_norm": 0.3057819902896881, "learning_rate": 2.469217172405388e-05, "loss": 0.0141, "step": 54840 }, { "grad_norm": 0.19915273785591125, "learning_rate": 2.4674348622117527e-05, "loss": 0.0123, "step": 54850 }, { "grad_norm": 1.434127688407898, "learning_rate": 2.465652984764234e-05, "loss": 0.017, "step": 54860 }, { "grad_norm": 0.2194059044122696, "learning_rate": 2.4638715403673056e-05, "loss": 0.0141, "step": 54870 }, { "grad_norm": 0.37039756774902344, "learning_rate": 2.4620905293253676e-05, "loss": 0.0181, "step": 54880 }, { "grad_norm": 0.3219882845878601, "learning_rate": 2.4603099519427463e-05, "loss": 0.0194, "step": 54890 }, { "grad_norm": 0.6222555637359619, "learning_rate": 2.458529808523695e-05, "loss": 0.0178, "step": 54900 }, { "grad_norm": 0.8303351998329163, "learning_rate": 2.456750099372387e-05, "loss": 0.0157, "step": 54910 }, { "grad_norm": 0.2537265717983246, "learning_rate": 2.4549708247929327e-05, "loss": 0.0202, "step": 54920 }, { "grad_norm": 0.2324468046426773, "learning_rate": 2.4531919850893554e-05, "loss": 0.0132, "step": 54930 }, { "grad_norm": 0.15850162506103516, "learning_rate": 2.4514135805656125e-05, "loss": 0.0133, "step": 54940 }, { "grad_norm": 0.28374233841896057, "learning_rate": 2.449635611525587e-05, "loss": 0.0171, "step": 54950 }, { "grad_norm": 0.2549542784690857, "learning_rate": 2.447858078273079e-05, "loss": 0.0185, "step": 54960 }, { "grad_norm": 0.24862094223499298, "learning_rate": 2.4460809811118275e-05, "loss": 0.0235, "step": 54970 }, { "grad_norm": 0.179012268781662, "learning_rate": 2.444304320345483e-05, "loss": 0.0132, "step": 54980 }, { "grad_norm": 0.23324042558670044, "learning_rate": 2.4425280962776354e-05, "loss": 0.018, "step": 54990 }, { "grad_norm": 0.32882338762283325, "learning_rate": 2.4407523092117875e-05, "loss": 0.019, "step": 55000 }, { "grad_norm": 0.45013752579689026, "learning_rate": 2.438976959451374e-05, "loss": 0.0141, "step": 55010 }, { "grad_norm": 0.2856495976448059, "learning_rate": 2.4372020472997565e-05, "loss": 0.014, "step": 55020 }, { "grad_norm": 0.20896901190280914, "learning_rate": 2.4354275730602122e-05, "loss": 0.0142, "step": 55030 }, { "grad_norm": 0.241717129945755, "learning_rate": 2.433653537035958e-05, "loss": 0.0169, "step": 55040 }, { "grad_norm": 0.2591184079647064, "learning_rate": 2.431879939530123e-05, "loss": 0.0186, "step": 55050 }, { "grad_norm": 0.21168501675128937, "learning_rate": 2.4301067808457684e-05, "loss": 0.0148, "step": 55060 }, { "grad_norm": 0.2316589206457138, "learning_rate": 2.428334061285878e-05, "loss": 0.0125, "step": 55070 }, { "grad_norm": 0.34014931321144104, "learning_rate": 2.426561781153361e-05, "loss": 0.0165, "step": 55080 }, { "grad_norm": 0.13373929262161255, "learning_rate": 2.424789940751052e-05, "loss": 0.0166, "step": 55090 }, { "grad_norm": 0.24440495669841766, "learning_rate": 2.423018540381712e-05, "loss": 0.0189, "step": 55100 }, { "grad_norm": 0.13723856210708618, "learning_rate": 2.421247580348021e-05, "loss": 0.0189, "step": 55110 }, { "grad_norm": 0.2819880247116089, "learning_rate": 2.4194770609525897e-05, "loss": 0.0148, "step": 55120 }, { "grad_norm": 0.2245381772518158, "learning_rate": 2.417706982497951e-05, "loss": 0.0156, "step": 55130 }, { "grad_norm": 0.13805197179317474, "learning_rate": 2.415937345286563e-05, "loss": 0.0178, "step": 55140 }, { "grad_norm": 0.27816128730773926, "learning_rate": 2.4141681496208087e-05, "loss": 0.0153, "step": 55150 }, { "grad_norm": 0.26660633087158203, "learning_rate": 2.4123993958029946e-05, "loss": 0.0131, "step": 55160 }, { "grad_norm": 0.41657713055610657, "learning_rate": 2.4106310841353548e-05, "loss": 0.0175, "step": 55170 }, { "grad_norm": 0.16298775374889374, "learning_rate": 2.4088632149200398e-05, "loss": 0.0189, "step": 55180 }, { "grad_norm": 0.3239968717098236, "learning_rate": 2.4070957884591367e-05, "loss": 0.015, "step": 55190 }, { "grad_norm": 0.4747355580329895, "learning_rate": 2.4053288050546464e-05, "loss": 0.0207, "step": 55200 }, { "grad_norm": 0.2009989470243454, "learning_rate": 2.403562265008498e-05, "loss": 0.0204, "step": 55210 }, { "grad_norm": 0.28510621190071106, "learning_rate": 2.4017961686225483e-05, "loss": 0.0152, "step": 55220 }, { "grad_norm": 0.2380455881357193, "learning_rate": 2.400030516198568e-05, "loss": 0.0236, "step": 55230 }, { "grad_norm": 0.25834494829177856, "learning_rate": 2.3982653080382673e-05, "loss": 0.0182, "step": 55240 }, { "grad_norm": 0.44035249948501587, "learning_rate": 2.396500544443266e-05, "loss": 0.016, "step": 55250 }, { "grad_norm": 0.3085809051990509, "learning_rate": 2.3947362257151156e-05, "loss": 0.0141, "step": 55260 }, { "grad_norm": 0.3752869963645935, "learning_rate": 2.39297235215529e-05, "loss": 0.0197, "step": 55270 }, { "grad_norm": 0.26558956503868103, "learning_rate": 2.3912089240651873e-05, "loss": 0.0207, "step": 55280 }, { "grad_norm": 0.14332589507102966, "learning_rate": 2.389445941746129e-05, "loss": 0.0235, "step": 55290 }, { "grad_norm": 0.27923041582107544, "learning_rate": 2.3876834054993625e-05, "loss": 0.0207, "step": 55300 }, { "grad_norm": 0.16350160539150238, "learning_rate": 2.3859213156260522e-05, "loss": 0.015, "step": 55310 }, { "grad_norm": 0.19881956279277802, "learning_rate": 2.3841596724272948e-05, "loss": 0.0159, "step": 55320 }, { "grad_norm": 0.17474037408828735, "learning_rate": 2.382398476204106e-05, "loss": 0.0143, "step": 55330 }, { "grad_norm": 0.23354925215244293, "learning_rate": 2.3806377272574254e-05, "loss": 0.014, "step": 55340 }, { "grad_norm": 0.29332980513572693, "learning_rate": 2.3788774258881174e-05, "loss": 0.0142, "step": 55350 }, { "grad_norm": 0.8086866140365601, "learning_rate": 2.37711757239697e-05, "loss": 0.02, "step": 55360 }, { "grad_norm": 0.29900383949279785, "learning_rate": 2.3753581670846954e-05, "loss": 0.0094, "step": 55370 }, { "grad_norm": 0.17159338295459747, "learning_rate": 2.3735992102519216e-05, "loss": 0.0168, "step": 55380 }, { "grad_norm": 0.24775376915931702, "learning_rate": 2.371840702199215e-05, "loss": 0.0148, "step": 55390 }, { "grad_norm": 0.17046305537223816, "learning_rate": 2.3700826432270494e-05, "loss": 0.0111, "step": 55400 }, { "grad_norm": 0.24852235615253448, "learning_rate": 2.3683250336358326e-05, "loss": 0.0188, "step": 55410 }, { "grad_norm": 0.1768815517425537, "learning_rate": 2.3665678737258923e-05, "loss": 0.023, "step": 55420 }, { "grad_norm": 0.1648702770471573, "learning_rate": 2.3648111637974745e-05, "loss": 0.0139, "step": 55430 }, { "grad_norm": 0.20436711609363556, "learning_rate": 2.36305490415076e-05, "loss": 0.0156, "step": 55440 }, { "grad_norm": 0.3155638873577118, "learning_rate": 2.3612990950858392e-05, "loss": 0.0211, "step": 55450 }, { "grad_norm": 0.14202146232128143, "learning_rate": 2.359543736902735e-05, "loss": 0.0166, "step": 55460 }, { "grad_norm": 0.24171116948127747, "learning_rate": 2.3577888299013896e-05, "loss": 0.0129, "step": 55470 }, { "grad_norm": 0.264051616191864, "learning_rate": 2.3560343743816683e-05, "loss": 0.0144, "step": 55480 }, { "grad_norm": 0.27987974882125854, "learning_rate": 2.354280370643362e-05, "loss": 0.0216, "step": 55490 }, { "grad_norm": 0.45079532265663147, "learning_rate": 2.3525268189861777e-05, "loss": 0.0158, "step": 55500 }, { "grad_norm": 0.3414422273635864, "learning_rate": 2.3507737197097513e-05, "loss": 0.0143, "step": 55510 }, { "grad_norm": 0.2929264307022095, "learning_rate": 2.34902107311364e-05, "loss": 0.0161, "step": 55520 }, { "grad_norm": 0.2069823443889618, "learning_rate": 2.3472688794973225e-05, "loss": 0.0177, "step": 55530 }, { "grad_norm": 0.29864436388015747, "learning_rate": 2.3455171391602016e-05, "loss": 0.02, "step": 55540 }, { "grad_norm": 0.32127314805984497, "learning_rate": 2.343765852401601e-05, "loss": 0.0132, "step": 55550 }, { "grad_norm": 0.20747153460979462, "learning_rate": 2.342015019520768e-05, "loss": 0.0179, "step": 55560 }, { "grad_norm": 0.22506336867809296, "learning_rate": 2.3402646408168742e-05, "loss": 0.0164, "step": 55570 }, { "grad_norm": 0.38793209195137024, "learning_rate": 2.3385147165890074e-05, "loss": 0.0228, "step": 55580 }, { "grad_norm": 0.1972731202840805, "learning_rate": 2.336765247136184e-05, "loss": 0.0144, "step": 55590 }, { "grad_norm": 0.3644729256629944, "learning_rate": 2.33501623275734e-05, "loss": 0.0236, "step": 55600 }, { "grad_norm": 0.34369415044784546, "learning_rate": 2.333267673751334e-05, "loss": 0.0145, "step": 55610 }, { "grad_norm": 0.26777908205986023, "learning_rate": 2.33151957041695e-05, "loss": 0.0116, "step": 55620 }, { "grad_norm": 0.293495237827301, "learning_rate": 2.329771923052884e-05, "loss": 0.0168, "step": 55630 }, { "grad_norm": 0.250511109828949, "learning_rate": 2.3280247319577697e-05, "loss": 0.0191, "step": 55640 }, { "grad_norm": 0.41276800632476807, "learning_rate": 2.3262779974301473e-05, "loss": 0.0242, "step": 55650 }, { "grad_norm": 0.3147904574871063, "learning_rate": 2.3245317197684895e-05, "loss": 0.0153, "step": 55660 }, { "grad_norm": 0.27707940340042114, "learning_rate": 2.322785899271188e-05, "loss": 0.0146, "step": 55670 }, { "grad_norm": 0.24505887925624847, "learning_rate": 2.3210405362365507e-05, "loss": 0.0207, "step": 55680 }, { "grad_norm": 0.2293519228696823, "learning_rate": 2.31929563096282e-05, "loss": 0.0178, "step": 55690 }, { "grad_norm": 0.20190121233463287, "learning_rate": 2.317551183748146e-05, "loss": 0.0145, "step": 55700 }, { "grad_norm": 0.20903350412845612, "learning_rate": 2.3158071948906103e-05, "loss": 0.0164, "step": 55710 }, { "grad_norm": 0.19283179938793182, "learning_rate": 2.314063664688212e-05, "loss": 0.0141, "step": 55720 }, { "grad_norm": 0.5078739523887634, "learning_rate": 2.3123205934388725e-05, "loss": 0.0127, "step": 55730 }, { "grad_norm": 0.1668107509613037, "learning_rate": 2.310577981440436e-05, "loss": 0.02, "step": 55740 }, { "grad_norm": 0.21899859607219696, "learning_rate": 2.3088358289906665e-05, "loss": 0.0138, "step": 55750 }, { "grad_norm": 0.2846710979938507, "learning_rate": 2.307094136387252e-05, "loss": 0.0157, "step": 55760 }, { "grad_norm": 0.30335918068885803, "learning_rate": 2.305352903927796e-05, "loss": 0.0218, "step": 55770 }, { "grad_norm": 0.3344634175300598, "learning_rate": 2.303612131909831e-05, "loss": 0.0147, "step": 55780 }, { "grad_norm": 0.37244120240211487, "learning_rate": 2.3018718206308054e-05, "loss": 0.0215, "step": 55790 }, { "grad_norm": 0.30734699964523315, "learning_rate": 2.3001319703880925e-05, "loss": 0.0188, "step": 55800 }, { "grad_norm": 0.32375678420066833, "learning_rate": 2.2983925814789835e-05, "loss": 0.0148, "step": 55810 }, { "grad_norm": 0.26986026763916016, "learning_rate": 2.2966536542006957e-05, "loss": 0.0189, "step": 55820 }, { "grad_norm": 0.220256507396698, "learning_rate": 2.294915188850358e-05, "loss": 0.0162, "step": 55830 }, { "grad_norm": 0.14003625512123108, "learning_rate": 2.2931771857250333e-05, "loss": 0.0161, "step": 55840 }, { "grad_norm": 0.20718695223331451, "learning_rate": 2.2914396451216946e-05, "loss": 0.0128, "step": 55850 }, { "grad_norm": 0.2798685133457184, "learning_rate": 2.2897025673372412e-05, "loss": 0.0141, "step": 55860 }, { "grad_norm": 0.19084034860134125, "learning_rate": 2.287965952668494e-05, "loss": 0.0138, "step": 55870 }, { "grad_norm": 0.21697074174880981, "learning_rate": 2.2862298014121873e-05, "loss": 0.0158, "step": 55880 }, { "grad_norm": 0.574627697467804, "learning_rate": 2.28449411386499e-05, "loss": 0.015, "step": 55890 }, { "grad_norm": 0.20765267312526703, "learning_rate": 2.282758890323477e-05, "loss": 0.0178, "step": 55900 }, { "grad_norm": 0.1990841180086136, "learning_rate": 2.2810241310841528e-05, "loss": 0.0151, "step": 55910 }, { "grad_norm": 0.27607637643814087, "learning_rate": 2.2792898364434412e-05, "loss": 0.0131, "step": 55920 }, { "grad_norm": 0.37301722168922424, "learning_rate": 2.2775560066976843e-05, "loss": 0.0184, "step": 55930 }, { "grad_norm": 0.2027982622385025, "learning_rate": 2.275822642143147e-05, "loss": 0.0143, "step": 55940 }, { "grad_norm": 0.2662336230278015, "learning_rate": 2.274089743076014e-05, "loss": 0.0139, "step": 55950 }, { "grad_norm": 0.2940053343772888, "learning_rate": 2.2723573097923917e-05, "loss": 0.0162, "step": 55960 }, { "grad_norm": 0.19095371663570404, "learning_rate": 2.2706253425883024e-05, "loss": 0.0129, "step": 55970 }, { "grad_norm": 0.27493682503700256, "learning_rate": 2.2688938417596933e-05, "loss": 0.0173, "step": 55980 }, { "grad_norm": 0.1346290409564972, "learning_rate": 2.2671628076024305e-05, "loss": 0.0128, "step": 55990 }, { "grad_norm": 0.16049450635910034, "learning_rate": 2.2654322404123008e-05, "loss": 0.0122, "step": 56000 }, { "grad_norm": 0.14077810943126678, "learning_rate": 2.2637021404850105e-05, "loss": 0.016, "step": 56010 }, { "grad_norm": 0.334625780582428, "learning_rate": 2.2619725081161885e-05, "loss": 0.0186, "step": 56020 }, { "grad_norm": 0.21721823513507843, "learning_rate": 2.260243343601376e-05, "loss": 0.0177, "step": 56030 }, { "grad_norm": 0.16218705475330353, "learning_rate": 2.2585146472360473e-05, "loss": 0.0204, "step": 56040 }, { "grad_norm": 0.3093504011631012, "learning_rate": 2.2567864193155834e-05, "loss": 0.0191, "step": 56050 }, { "grad_norm": 0.5509445071220398, "learning_rate": 2.255058660135294e-05, "loss": 0.0176, "step": 56060 }, { "grad_norm": 0.40576624870300293, "learning_rate": 2.253331369990407e-05, "loss": 0.0201, "step": 56070 }, { "grad_norm": 0.1344401091337204, "learning_rate": 2.251604549176063e-05, "loss": 0.0197, "step": 56080 }, { "grad_norm": 0.25430700182914734, "learning_rate": 2.249878197987337e-05, "loss": 0.0166, "step": 56090 }, { "grad_norm": 0.4689810872077942, "learning_rate": 2.2481523167192087e-05, "loss": 0.0205, "step": 56100 }, { "grad_norm": 0.1526443511247635, "learning_rate": 2.2464269056665867e-05, "loss": 0.0162, "step": 56110 }, { "grad_norm": 0.22276850044727325, "learning_rate": 2.2447019651242958e-05, "loss": 0.0138, "step": 56120 }, { "grad_norm": 0.33158427476882935, "learning_rate": 2.2429774953870818e-05, "loss": 0.0158, "step": 56130 }, { "grad_norm": 0.3769245147705078, "learning_rate": 2.241253496749611e-05, "loss": 0.0187, "step": 56140 }, { "grad_norm": 0.364692360162735, "learning_rate": 2.2395299695064614e-05, "loss": 0.015, "step": 56150 }, { "grad_norm": 0.6600449681282043, "learning_rate": 2.237806913952145e-05, "loss": 0.0194, "step": 56160 }, { "grad_norm": 0.25388792157173157, "learning_rate": 2.2360843303810798e-05, "loss": 0.0146, "step": 56170 }, { "grad_norm": 0.26557907462120056, "learning_rate": 2.2343622190876084e-05, "loss": 0.0169, "step": 56180 }, { "grad_norm": 0.3734886646270752, "learning_rate": 2.2326405803659935e-05, "loss": 0.0156, "step": 56190 }, { "grad_norm": 0.2690730690956116, "learning_rate": 2.230919414510416e-05, "loss": 0.013, "step": 56200 }, { "grad_norm": 0.3564513623714447, "learning_rate": 2.229198721814976e-05, "loss": 0.0174, "step": 56210 }, { "grad_norm": 0.2494581937789917, "learning_rate": 2.2274785025736948e-05, "loss": 0.0154, "step": 56220 }, { "grad_norm": 0.28570422530174255, "learning_rate": 2.225758757080507e-05, "loss": 0.0176, "step": 56230 }, { "grad_norm": 0.4023815989494324, "learning_rate": 2.2240394856292723e-05, "loss": 0.0158, "step": 56240 }, { "grad_norm": 1.0506821870803833, "learning_rate": 2.2223206885137664e-05, "loss": 0.0154, "step": 56250 }, { "grad_norm": 0.43160101771354675, "learning_rate": 2.2206023660276853e-05, "loss": 0.0145, "step": 56260 }, { "grad_norm": 0.3390093445777893, "learning_rate": 2.218884518464645e-05, "loss": 0.0161, "step": 56270 }, { "grad_norm": 0.3673282563686371, "learning_rate": 2.2171671461181732e-05, "loss": 0.0176, "step": 56280 }, { "grad_norm": 0.35583627223968506, "learning_rate": 2.2154502492817292e-05, "loss": 0.0179, "step": 56290 }, { "grad_norm": 0.1588192582130432, "learning_rate": 2.2137338282486782e-05, "loss": 0.0113, "step": 56300 }, { "grad_norm": 0.10557368397712708, "learning_rate": 2.2120178833123113e-05, "loss": 0.0147, "step": 56310 }, { "grad_norm": 0.4495086967945099, "learning_rate": 2.210302414765838e-05, "loss": 0.0209, "step": 56320 }, { "grad_norm": 0.19018903374671936, "learning_rate": 2.2085874229023808e-05, "loss": 0.0143, "step": 56330 }, { "grad_norm": 0.18090584874153137, "learning_rate": 2.2068729080149907e-05, "loss": 0.015, "step": 56340 }, { "grad_norm": 0.35367584228515625, "learning_rate": 2.205158870396625e-05, "loss": 0.0158, "step": 56350 }, { "grad_norm": 0.20711752772331238, "learning_rate": 2.2034453103401732e-05, "loss": 0.0191, "step": 56360 }, { "grad_norm": 0.10647924989461899, "learning_rate": 2.20173222813843e-05, "loss": 0.0106, "step": 56370 }, { "grad_norm": 0.2187212109565735, "learning_rate": 2.200019624084116e-05, "loss": 0.0175, "step": 56380 }, { "grad_norm": 0.32503220438957214, "learning_rate": 2.1983074984698687e-05, "loss": 0.0153, "step": 56390 }, { "grad_norm": 0.32015401124954224, "learning_rate": 2.1965958515882433e-05, "loss": 0.0163, "step": 56400 }, { "grad_norm": 0.3126705288887024, "learning_rate": 2.1948846837317162e-05, "loss": 0.0153, "step": 56410 }, { "grad_norm": 0.22366128861904144, "learning_rate": 2.1931739951926738e-05, "loss": 0.0114, "step": 56420 }, { "grad_norm": 0.21001270413398743, "learning_rate": 2.191463786263429e-05, "loss": 0.0123, "step": 56430 }, { "grad_norm": 0.31160393357276917, "learning_rate": 2.1897540572362095e-05, "loss": 0.0151, "step": 56440 }, { "grad_norm": 0.2003633975982666, "learning_rate": 2.1880448084031614e-05, "loss": 0.0117, "step": 56450 }, { "grad_norm": 0.24828168749809265, "learning_rate": 2.1863360400563482e-05, "loss": 0.0143, "step": 56460 }, { "grad_norm": 0.39218515157699585, "learning_rate": 2.184627752487754e-05, "loss": 0.0194, "step": 56470 }, { "grad_norm": 0.2708883583545685, "learning_rate": 2.1829199459892725e-05, "loss": 0.0168, "step": 56480 }, { "grad_norm": 0.20349811017513275, "learning_rate": 2.1812126208527282e-05, "loss": 0.0166, "step": 56490 }, { "grad_norm": 0.2221594601869583, "learning_rate": 2.179505777369852e-05, "loss": 0.0201, "step": 56500 }, { "grad_norm": 0.2745051383972168, "learning_rate": 2.1777994158322974e-05, "loss": 0.0155, "step": 56510 }, { "grad_norm": 0.3118213713169098, "learning_rate": 2.176093536531637e-05, "loss": 0.0159, "step": 56520 }, { "grad_norm": 0.19734516739845276, "learning_rate": 2.1743881397593537e-05, "loss": 0.0152, "step": 56530 }, { "grad_norm": 0.4704294800758362, "learning_rate": 2.1726832258068595e-05, "loss": 0.0254, "step": 56540 }, { "grad_norm": 0.3774593770503998, "learning_rate": 2.170978794965472e-05, "loss": 0.0152, "step": 56550 }, { "grad_norm": 0.7972443699836731, "learning_rate": 2.169274847526438e-05, "loss": 0.0161, "step": 56560 }, { "grad_norm": 0.31899386644363403, "learning_rate": 2.1675713837809103e-05, "loss": 0.0144, "step": 56570 }, { "grad_norm": 0.14245973527431488, "learning_rate": 2.1658684040199655e-05, "loss": 0.0114, "step": 56580 }, { "grad_norm": 0.21605683863162994, "learning_rate": 2.1641659085345974e-05, "loss": 0.0134, "step": 56590 }, { "grad_norm": 0.1609055995941162, "learning_rate": 2.1624638976157154e-05, "loss": 0.0183, "step": 56600 }, { "grad_norm": 0.33282285928726196, "learning_rate": 2.1607623715541476e-05, "loss": 0.0238, "step": 56610 }, { "grad_norm": 0.18233785033226013, "learning_rate": 2.159061330640636e-05, "loss": 0.012, "step": 56620 }, { "grad_norm": 0.33479607105255127, "learning_rate": 2.1573607751658425e-05, "loss": 0.0114, "step": 56630 }, { "grad_norm": 0.2147800624370575, "learning_rate": 2.155660705420347e-05, "loss": 0.0182, "step": 56640 }, { "grad_norm": 0.7409994006156921, "learning_rate": 2.153961121694644e-05, "loss": 0.0198, "step": 56650 }, { "grad_norm": 0.2214280217885971, "learning_rate": 2.152262024279145e-05, "loss": 0.0243, "step": 56660 }, { "grad_norm": 0.43802469968795776, "learning_rate": 2.150563413464183e-05, "loss": 0.0165, "step": 56670 }, { "grad_norm": 0.20582641661167145, "learning_rate": 2.1488652895399974e-05, "loss": 0.0144, "step": 56680 }, { "grad_norm": 0.40053844451904297, "learning_rate": 2.1471676527967587e-05, "loss": 0.0178, "step": 56690 }, { "grad_norm": 0.16620488464832306, "learning_rate": 2.1454705035245414e-05, "loss": 0.0211, "step": 56700 }, { "grad_norm": 0.24807998538017273, "learning_rate": 2.143773842013343e-05, "loss": 0.0153, "step": 56710 }, { "grad_norm": 0.24821273982524872, "learning_rate": 2.1420776685530796e-05, "loss": 0.0183, "step": 56720 }, { "grad_norm": 0.30219608545303345, "learning_rate": 2.1403819834335742e-05, "loss": 0.0103, "step": 56730 }, { "grad_norm": 0.19584082067012787, "learning_rate": 2.1386867869445808e-05, "loss": 0.0146, "step": 56740 }, { "grad_norm": 0.16161249577999115, "learning_rate": 2.1369920793757548e-05, "loss": 0.0183, "step": 56750 }, { "grad_norm": 0.3121107220649719, "learning_rate": 2.1352978610166828e-05, "loss": 0.0133, "step": 56760 }, { "grad_norm": 0.2681971788406372, "learning_rate": 2.1336041321568546e-05, "loss": 0.0179, "step": 56770 }, { "grad_norm": 0.20258519053459167, "learning_rate": 2.131910893085684e-05, "loss": 0.0168, "step": 56780 }, { "grad_norm": 0.17336519062519073, "learning_rate": 2.1302181440925012e-05, "loss": 0.0164, "step": 56790 }, { "grad_norm": 0.44679322838783264, "learning_rate": 2.128525885466546e-05, "loss": 0.0159, "step": 56800 }, { "grad_norm": 0.24346193671226501, "learning_rate": 2.1268341174969847e-05, "loss": 0.0177, "step": 56810 }, { "grad_norm": 0.23464494943618774, "learning_rate": 2.1251428404728894e-05, "loss": 0.0158, "step": 56820 }, { "grad_norm": 0.24110294878482819, "learning_rate": 2.123452054683256e-05, "loss": 0.0161, "step": 56830 }, { "grad_norm": 0.30784448981285095, "learning_rate": 2.1217617604169927e-05, "loss": 0.0174, "step": 56840 }, { "grad_norm": 0.2703423798084259, "learning_rate": 2.120071957962924e-05, "loss": 0.0159, "step": 56850 }, { "grad_norm": 0.24046088755130768, "learning_rate": 2.1183826476097917e-05, "loss": 0.0231, "step": 56860 }, { "grad_norm": 0.27601033449172974, "learning_rate": 2.116693829646254e-05, "loss": 0.019, "step": 56870 }, { "grad_norm": 0.3392108976840973, "learning_rate": 2.1150055043608806e-05, "loss": 0.0135, "step": 56880 }, { "grad_norm": 0.3437946140766144, "learning_rate": 2.113317672042162e-05, "loss": 0.0216, "step": 56890 }, { "grad_norm": 0.29197677969932556, "learning_rate": 2.1116303329785025e-05, "loss": 0.0105, "step": 56900 }, { "grad_norm": 0.2239895910024643, "learning_rate": 2.109943487458222e-05, "loss": 0.0162, "step": 56910 }, { "grad_norm": 0.21459880471229553, "learning_rate": 2.1082571357695574e-05, "loss": 0.0143, "step": 56920 }, { "grad_norm": 0.21821065247058868, "learning_rate": 2.1065712782006557e-05, "loss": 0.017, "step": 56930 }, { "grad_norm": 0.11492598056793213, "learning_rate": 2.104885915039591e-05, "loss": 0.0118, "step": 56940 }, { "grad_norm": 0.22118958830833435, "learning_rate": 2.103201046574338e-05, "loss": 0.0153, "step": 56950 }, { "grad_norm": 0.347513347864151, "learning_rate": 2.1015166730928022e-05, "loss": 0.0142, "step": 56960 }, { "grad_norm": 0.283949613571167, "learning_rate": 2.0998327948827912e-05, "loss": 0.0174, "step": 56970 }, { "grad_norm": 0.29264113306999207, "learning_rate": 2.0981494122320363e-05, "loss": 0.0187, "step": 56980 }, { "grad_norm": 0.35072240233421326, "learning_rate": 2.0964665254281822e-05, "loss": 0.0179, "step": 56990 }, { "grad_norm": 0.23612765967845917, "learning_rate": 2.094784134758784e-05, "loss": 0.0145, "step": 57000 }, { "grad_norm": 0.2418658435344696, "learning_rate": 2.0931022405113226e-05, "loss": 0.0262, "step": 57010 }, { "grad_norm": 0.23957952857017517, "learning_rate": 2.091420842973183e-05, "loss": 0.0148, "step": 57020 }, { "grad_norm": 0.1482999622821808, "learning_rate": 2.0897399424316715e-05, "loss": 0.0133, "step": 57030 }, { "grad_norm": 0.26992976665496826, "learning_rate": 2.0880595391740078e-05, "loss": 0.0167, "step": 57040 }, { "grad_norm": 0.30570778250694275, "learning_rate": 2.0863796334873277e-05, "loss": 0.0169, "step": 57050 }, { "grad_norm": 0.24220485985279083, "learning_rate": 2.08470022565868e-05, "loss": 0.0166, "step": 57060 }, { "grad_norm": 0.22706052660942078, "learning_rate": 2.0830213159750317e-05, "loss": 0.0127, "step": 57070 }, { "grad_norm": 0.17058226466178894, "learning_rate": 2.0813429047232596e-05, "loss": 0.0132, "step": 57080 }, { "grad_norm": 0.16778600215911865, "learning_rate": 2.0796649921901594e-05, "loss": 0.0153, "step": 57090 }, { "grad_norm": 0.29765817523002625, "learning_rate": 2.077987578662441e-05, "loss": 0.0146, "step": 57100 }, { "grad_norm": 0.25935325026512146, "learning_rate": 2.0763106644267277e-05, "loss": 0.0166, "step": 57110 }, { "grad_norm": 0.17621953785419464, "learning_rate": 2.0746342497695607e-05, "loss": 0.0076, "step": 57120 }, { "grad_norm": 0.2177039086818695, "learning_rate": 2.0729583349773886e-05, "loss": 0.0115, "step": 57130 }, { "grad_norm": 0.38745054602622986, "learning_rate": 2.0712829203365853e-05, "loss": 0.0134, "step": 57140 }, { "grad_norm": 0.18593165278434753, "learning_rate": 2.0696080061334267e-05, "loss": 0.016, "step": 57150 }, { "grad_norm": 0.17147204279899597, "learning_rate": 2.067933592654117e-05, "loss": 0.0144, "step": 57160 }, { "grad_norm": 0.21396049857139587, "learning_rate": 2.066259680184763e-05, "loss": 0.0166, "step": 57170 }, { "grad_norm": 0.14670799672603607, "learning_rate": 2.0645862690113908e-05, "loss": 0.0113, "step": 57180 }, { "grad_norm": 0.24818144738674164, "learning_rate": 2.0629133594199436e-05, "loss": 0.0217, "step": 57190 }, { "grad_norm": 0.22898411750793457, "learning_rate": 2.0612409516962704e-05, "loss": 0.0166, "step": 57200 }, { "grad_norm": 0.343554824590683, "learning_rate": 2.0595690461261467e-05, "loss": 0.0192, "step": 57210 }, { "grad_norm": 0.3104367256164551, "learning_rate": 2.0578976429952503e-05, "loss": 0.0154, "step": 57220 }, { "grad_norm": 0.3391253650188446, "learning_rate": 2.0562267425891802e-05, "loss": 0.0164, "step": 57230 }, { "grad_norm": 0.2770210802555084, "learning_rate": 2.0545563451934467e-05, "loss": 0.0147, "step": 57240 }, { "grad_norm": 0.20144404470920563, "learning_rate": 2.0528864510934764e-05, "loss": 0.0149, "step": 57250 }, { "grad_norm": 0.24764437973499298, "learning_rate": 2.0512170605746096e-05, "loss": 0.0222, "step": 57260 }, { "grad_norm": 0.3332589566707611, "learning_rate": 2.0495481739220963e-05, "loss": 0.0165, "step": 57270 }, { "grad_norm": 0.22704143822193146, "learning_rate": 2.0478797914211045e-05, "loss": 0.0197, "step": 57280 }, { "grad_norm": 0.22943057119846344, "learning_rate": 2.046211913356716e-05, "loss": 0.0177, "step": 57290 }, { "grad_norm": 0.21810363233089447, "learning_rate": 2.0445445400139247e-05, "loss": 0.0176, "step": 57300 }, { "grad_norm": 0.3325028121471405, "learning_rate": 2.0428776716776405e-05, "loss": 0.0158, "step": 57310 }, { "grad_norm": 0.1824222207069397, "learning_rate": 2.0412113086326856e-05, "loss": 0.0151, "step": 57320 }, { "grad_norm": 0.22271105647087097, "learning_rate": 2.0395454511637918e-05, "loss": 0.0128, "step": 57330 }, { "grad_norm": 0.36595359444618225, "learning_rate": 2.037880099555616e-05, "loss": 0.015, "step": 57340 }, { "grad_norm": 0.3341670036315918, "learning_rate": 2.0362152540927144e-05, "loss": 0.0136, "step": 57350 }, { "grad_norm": 0.221882626414299, "learning_rate": 2.0345509150595666e-05, "loss": 0.0264, "step": 57360 }, { "grad_norm": 0.2331451028585434, "learning_rate": 2.0328870827405617e-05, "loss": 0.0185, "step": 57370 }, { "grad_norm": 0.2132282257080078, "learning_rate": 2.0312237574200043e-05, "loss": 0.0145, "step": 57380 }, { "grad_norm": 0.2733059823513031, "learning_rate": 2.029560939382112e-05, "loss": 0.0133, "step": 57390 }, { "grad_norm": 0.7022546529769897, "learning_rate": 2.0278986289110097e-05, "loss": 0.0141, "step": 57400 }, { "grad_norm": 0.18482455611228943, "learning_rate": 2.0262368262907484e-05, "loss": 0.0149, "step": 57410 }, { "grad_norm": 0.27960607409477234, "learning_rate": 2.024575531805279e-05, "loss": 0.0169, "step": 57420 }, { "grad_norm": 0.881583571434021, "learning_rate": 2.0229147457384735e-05, "loss": 0.016, "step": 57430 }, { "grad_norm": 0.3614596724510193, "learning_rate": 2.0212544683741157e-05, "loss": 0.024, "step": 57440 }, { "grad_norm": 0.24977019429206848, "learning_rate": 2.0195946999958976e-05, "loss": 0.0297, "step": 57450 }, { "grad_norm": 0.21340762078762054, "learning_rate": 2.017935440887434e-05, "loss": 0.0192, "step": 57460 }, { "grad_norm": 0.22013157606124878, "learning_rate": 2.0162766913322423e-05, "loss": 0.0135, "step": 57470 }, { "grad_norm": 0.16332221031188965, "learning_rate": 2.0146184516137588e-05, "loss": 0.0179, "step": 57480 }, { "grad_norm": 0.24926035106182098, "learning_rate": 2.012960722015332e-05, "loss": 0.0163, "step": 57490 }, { "grad_norm": 0.22689609229564667, "learning_rate": 2.0113035028202214e-05, "loss": 0.0188, "step": 57500 }, { "grad_norm": 0.3444695472717285, "learning_rate": 2.009646794311602e-05, "loss": 0.0177, "step": 57510 }, { "grad_norm": 0.1727912724018097, "learning_rate": 2.007990596772559e-05, "loss": 0.0164, "step": 57520 }, { "grad_norm": 0.16642652451992035, "learning_rate": 2.0063349104860923e-05, "loss": 0.0174, "step": 57530 }, { "grad_norm": 0.2942412495613098, "learning_rate": 2.0046797357351116e-05, "loss": 0.0217, "step": 57540 }, { "grad_norm": 0.32274389266967773, "learning_rate": 2.0030250728024412e-05, "loss": 0.0152, "step": 57550 }, { "grad_norm": 0.44710013270378113, "learning_rate": 2.001370921970819e-05, "loss": 0.0139, "step": 57560 }, { "grad_norm": 0.20166954398155212, "learning_rate": 1.9997172835228932e-05, "loss": 0.0144, "step": 57570 }, { "grad_norm": 0.3229691982269287, "learning_rate": 1.9980641577412262e-05, "loss": 0.0163, "step": 57580 }, { "grad_norm": 0.24479204416275024, "learning_rate": 1.9964115449082925e-05, "loss": 0.0147, "step": 57590 }, { "grad_norm": 0.24773342907428741, "learning_rate": 1.9947594453064742e-05, "loss": 0.0167, "step": 57600 }, { "grad_norm": 0.23829354345798492, "learning_rate": 1.9931078592180774e-05, "loss": 0.015, "step": 57610 }, { "grad_norm": 0.30671706795692444, "learning_rate": 1.9914567869253065e-05, "loss": 0.0205, "step": 57620 }, { "grad_norm": 0.16048741340637207, "learning_rate": 1.989806228710287e-05, "loss": 0.0233, "step": 57630 }, { "grad_norm": 0.1975116729736328, "learning_rate": 1.9881561848550555e-05, "loss": 0.0109, "step": 57640 }, { "grad_norm": 0.2681644856929779, "learning_rate": 1.9865066556415544e-05, "loss": 0.0221, "step": 57650 }, { "grad_norm": 0.2803237736225128, "learning_rate": 1.98485764135165e-05, "loss": 0.0137, "step": 57660 }, { "grad_norm": 0.17097948491573334, "learning_rate": 1.983209142267109e-05, "loss": 0.0134, "step": 57670 }, { "grad_norm": 0.22783128917217255, "learning_rate": 1.9815611586696165e-05, "loss": 0.0146, "step": 57680 }, { "grad_norm": 0.27284476161003113, "learning_rate": 1.9799136908407667e-05, "loss": 0.0167, "step": 57690 }, { "grad_norm": 0.26702502369880676, "learning_rate": 1.9782667390620678e-05, "loss": 0.0186, "step": 57700 }, { "grad_norm": 0.21142354607582092, "learning_rate": 1.976620303614939e-05, "loss": 0.014, "step": 57710 }, { "grad_norm": 0.23592081665992737, "learning_rate": 1.9749743847807108e-05, "loss": 0.0154, "step": 57720 }, { "grad_norm": 0.2642042934894562, "learning_rate": 1.9733289828406272e-05, "loss": 0.0125, "step": 57730 }, { "grad_norm": 0.33126693964004517, "learning_rate": 1.9716840980758382e-05, "loss": 0.0154, "step": 57740 }, { "grad_norm": 0.24304334819316864, "learning_rate": 1.9700397307674134e-05, "loss": 0.0169, "step": 57750 }, { "grad_norm": 0.32101407647132874, "learning_rate": 1.968395881196328e-05, "loss": 0.0127, "step": 57760 }, { "grad_norm": 0.3019983768463135, "learning_rate": 1.966752549643473e-05, "loss": 0.0165, "step": 57770 }, { "grad_norm": 0.1745496243238449, "learning_rate": 1.965109736389647e-05, "loss": 0.0159, "step": 57780 }, { "grad_norm": 0.3536350131034851, "learning_rate": 1.9634674417155645e-05, "loss": 0.0169, "step": 57790 }, { "grad_norm": 0.2627715468406677, "learning_rate": 1.9618256659018434e-05, "loss": 0.0325, "step": 57800 }, { "grad_norm": 0.30524325370788574, "learning_rate": 1.9601844092290257e-05, "loss": 0.0126, "step": 57810 }, { "grad_norm": 0.33480337262153625, "learning_rate": 1.9585436719775512e-05, "loss": 0.0151, "step": 57820 }, { "grad_norm": 0.18675029277801514, "learning_rate": 1.9569034544277793e-05, "loss": 0.0147, "step": 57830 }, { "grad_norm": 0.2933728098869324, "learning_rate": 1.9552637568599798e-05, "loss": 0.014, "step": 57840 }, { "grad_norm": 0.41156527400016785, "learning_rate": 1.953624579554327e-05, "loss": 0.0193, "step": 57850 }, { "grad_norm": 0.17694373428821564, "learning_rate": 1.951985922790918e-05, "loss": 0.0196, "step": 57860 }, { "grad_norm": 0.19070236384868622, "learning_rate": 1.9503477868497505e-05, "loss": 0.0097, "step": 57870 }, { "grad_norm": 0.2960088551044464, "learning_rate": 1.9487101720107375e-05, "loss": 0.018, "step": 57880 }, { "grad_norm": 0.2681933343410492, "learning_rate": 1.9470730785537032e-05, "loss": 0.0162, "step": 57890 }, { "grad_norm": 0.15751811861991882, "learning_rate": 1.9454365067583823e-05, "loss": 0.0127, "step": 57900 }, { "grad_norm": 0.1691056489944458, "learning_rate": 1.9438004569044215e-05, "loss": 0.0143, "step": 57910 }, { "grad_norm": 0.3685746490955353, "learning_rate": 1.9421649292713724e-05, "loss": 0.0304, "step": 57920 }, { "grad_norm": 0.21807536482810974, "learning_rate": 1.9405299241387076e-05, "loss": 0.015, "step": 57930 }, { "grad_norm": 0.19937612116336823, "learning_rate": 1.9388954417858007e-05, "loss": 0.011, "step": 57940 }, { "grad_norm": 0.5629754662513733, "learning_rate": 1.9372614824919417e-05, "loss": 0.0157, "step": 57950 }, { "grad_norm": 0.25332146883010864, "learning_rate": 1.9356280465363284e-05, "loss": 0.0182, "step": 57960 }, { "grad_norm": 0.23531514406204224, "learning_rate": 1.9339951341980723e-05, "loss": 0.0143, "step": 57970 }, { "grad_norm": 0.2833098769187927, "learning_rate": 1.9323627457561916e-05, "loss": 0.0152, "step": 57980 }, { "grad_norm": 0.23814500868320465, "learning_rate": 1.9307308814896198e-05, "loss": 0.0122, "step": 57990 }, { "grad_norm": 0.23832926154136658, "learning_rate": 1.9290995416771935e-05, "loss": 0.0134, "step": 58000 }, { "grad_norm": 0.13617154955863953, "learning_rate": 1.9274687265976665e-05, "loss": 0.012, "step": 58010 }, { "grad_norm": 0.2640210688114166, "learning_rate": 1.9258384365297e-05, "loss": 0.0151, "step": 58020 }, { "grad_norm": 0.33328258991241455, "learning_rate": 1.924208671751866e-05, "loss": 0.0132, "step": 58030 }, { "grad_norm": 0.2518268823623657, "learning_rate": 1.9225794325426492e-05, "loss": 0.0121, "step": 58040 }, { "grad_norm": 0.18542753159999847, "learning_rate": 1.920950719180436e-05, "loss": 0.0111, "step": 58050 }, { "grad_norm": 0.27820754051208496, "learning_rate": 1.919322531943536e-05, "loss": 0.0154, "step": 58060 }, { "grad_norm": 0.23347383737564087, "learning_rate": 1.917694871110157e-05, "loss": 0.014, "step": 58070 }, { "grad_norm": 0.3587896525859833, "learning_rate": 1.9160677369584234e-05, "loss": 0.0124, "step": 58080 }, { "grad_norm": 0.2254170924425125, "learning_rate": 1.9144411297663694e-05, "loss": 0.0137, "step": 58090 }, { "grad_norm": 1.240708351135254, "learning_rate": 1.9128150498119328e-05, "loss": 0.02, "step": 58100 }, { "grad_norm": 0.2686663568019867, "learning_rate": 1.9111894973729726e-05, "loss": 0.0126, "step": 58110 }, { "grad_norm": 0.3884676396846771, "learning_rate": 1.9095644727272454e-05, "loss": 0.0176, "step": 58120 }, { "grad_norm": 0.3751671612262726, "learning_rate": 1.907939976152429e-05, "loss": 0.017, "step": 58130 }, { "grad_norm": 0.26715341210365295, "learning_rate": 1.906316007926101e-05, "loss": 0.0204, "step": 58140 }, { "grad_norm": 0.43878644704818726, "learning_rate": 1.904692568325755e-05, "loss": 0.0166, "step": 58150 }, { "grad_norm": 0.3357371985912323, "learning_rate": 1.9030696576287925e-05, "loss": 0.0175, "step": 58160 }, { "grad_norm": 0.3229123055934906, "learning_rate": 1.9014472761125242e-05, "loss": 0.0179, "step": 58170 }, { "grad_norm": 0.17299726605415344, "learning_rate": 1.899825424054172e-05, "loss": 0.0163, "step": 58180 }, { "grad_norm": 0.26925167441368103, "learning_rate": 1.898204101730863e-05, "loss": 0.0162, "step": 58190 }, { "grad_norm": 0.27717745304107666, "learning_rate": 1.8965833094196394e-05, "loss": 0.0259, "step": 58200 }, { "grad_norm": 0.4710676074028015, "learning_rate": 1.8949630473974495e-05, "loss": 0.0136, "step": 58210 }, { "grad_norm": 0.2266102135181427, "learning_rate": 1.8933433159411517e-05, "loss": 0.0148, "step": 58220 }, { "grad_norm": 0.1373896300792694, "learning_rate": 1.891724115327514e-05, "loss": 0.015, "step": 58230 }, { "grad_norm": 0.3421865999698639, "learning_rate": 1.8901054458332152e-05, "loss": 0.02, "step": 58240 }, { "grad_norm": 0.20010650157928467, "learning_rate": 1.8884873077348364e-05, "loss": 0.013, "step": 58250 }, { "grad_norm": 0.19944719970226288, "learning_rate": 1.88686970130888e-05, "loss": 0.017, "step": 58260 }, { "grad_norm": 0.20851917564868927, "learning_rate": 1.8852526268317456e-05, "loss": 0.0153, "step": 58270 }, { "grad_norm": 0.31689196825027466, "learning_rate": 1.883636084579749e-05, "loss": 0.0154, "step": 58280 }, { "grad_norm": 0.278048038482666, "learning_rate": 1.8820200748291145e-05, "loss": 0.0174, "step": 58290 }, { "grad_norm": 0.21339215338230133, "learning_rate": 1.8804045978559686e-05, "loss": 0.016, "step": 58300 }, { "grad_norm": 0.23495812714099884, "learning_rate": 1.8787896539363594e-05, "loss": 0.0161, "step": 58310 }, { "grad_norm": 0.2864631116390228, "learning_rate": 1.8771752433462298e-05, "loss": 0.0134, "step": 58320 }, { "grad_norm": 0.20895670354366302, "learning_rate": 1.8755613663614447e-05, "loss": 0.0179, "step": 58330 }, { "grad_norm": 0.3590003550052643, "learning_rate": 1.873948023257767e-05, "loss": 0.0176, "step": 58340 }, { "grad_norm": 0.21303699910640717, "learning_rate": 1.8723352143108747e-05, "loss": 0.0117, "step": 58350 }, { "grad_norm": 0.3224118649959564, "learning_rate": 1.870722939796352e-05, "loss": 0.0189, "step": 58360 }, { "grad_norm": 0.22830532491207123, "learning_rate": 1.869111199989693e-05, "loss": 0.016, "step": 58370 }, { "grad_norm": 0.17633210122585297, "learning_rate": 1.8674999951663018e-05, "loss": 0.0136, "step": 58380 }, { "grad_norm": 0.38924798369407654, "learning_rate": 1.8658893256014857e-05, "loss": 0.0168, "step": 58390 }, { "grad_norm": 0.19961006939411163, "learning_rate": 1.8642791915704655e-05, "loss": 0.0143, "step": 58400 }, { "grad_norm": 0.3817737400531769, "learning_rate": 1.8626695933483697e-05, "loss": 0.0138, "step": 58410 }, { "grad_norm": 0.2793067395687103, "learning_rate": 1.8610605312102335e-05, "loss": 0.0159, "step": 58420 }, { "grad_norm": 0.3376237750053406, "learning_rate": 1.8594520054310032e-05, "loss": 0.0178, "step": 58430 }, { "grad_norm": 0.17799779772758484, "learning_rate": 1.8578440162855326e-05, "loss": 0.0158, "step": 58440 }, { "grad_norm": 0.21978020668029785, "learning_rate": 1.8562365640485774e-05, "loss": 0.0162, "step": 58450 }, { "grad_norm": 0.14064213633537292, "learning_rate": 1.854629648994815e-05, "loss": 0.0137, "step": 58460 }, { "grad_norm": 0.22880002856254578, "learning_rate": 1.8530232713988183e-05, "loss": 0.0161, "step": 58470 }, { "grad_norm": 0.3761283755302429, "learning_rate": 1.851417431535074e-05, "loss": 0.0186, "step": 58480 }, { "grad_norm": 0.19959181547164917, "learning_rate": 1.8498121296779785e-05, "loss": 0.0217, "step": 58490 }, { "grad_norm": 0.22367630898952484, "learning_rate": 1.848207366101829e-05, "loss": 0.0175, "step": 58500 }, { "grad_norm": 0.15328143537044525, "learning_rate": 1.8466031410808422e-05, "loss": 0.0181, "step": 58510 }, { "grad_norm": 0.26630884408950806, "learning_rate": 1.844999454889129e-05, "loss": 0.0153, "step": 58520 }, { "grad_norm": 0.2666197121143341, "learning_rate": 1.8433963078007237e-05, "loss": 0.0154, "step": 58530 }, { "grad_norm": 0.24252945184707642, "learning_rate": 1.841793700089554e-05, "loss": 0.0166, "step": 58540 }, { "grad_norm": 0.27063292264938354, "learning_rate": 1.8401916320294644e-05, "loss": 0.014, "step": 58550 }, { "grad_norm": 0.2747492790222168, "learning_rate": 1.838590103894205e-05, "loss": 0.0143, "step": 58560 }, { "grad_norm": 0.11082039773464203, "learning_rate": 1.8369891159574292e-05, "loss": 0.014, "step": 58570 }, { "grad_norm": 0.2553757131099701, "learning_rate": 1.835388668492708e-05, "loss": 0.0175, "step": 58580 }, { "grad_norm": 0.2891392707824707, "learning_rate": 1.8337887617735095e-05, "loss": 0.0135, "step": 58590 }, { "grad_norm": 0.1482413411140442, "learning_rate": 1.8321893960732157e-05, "loss": 0.0112, "step": 58600 }, { "grad_norm": 0.22189636528491974, "learning_rate": 1.8305905716651138e-05, "loss": 0.0156, "step": 58610 }, { "grad_norm": 0.21829628944396973, "learning_rate": 1.8289922888224e-05, "loss": 0.0145, "step": 58620 }, { "grad_norm": 0.19217002391815186, "learning_rate": 1.8273945478181765e-05, "loss": 0.0163, "step": 58630 }, { "grad_norm": 0.3222742974758148, "learning_rate": 1.8257973489254558e-05, "loss": 0.0177, "step": 58640 }, { "grad_norm": 0.26339393854141235, "learning_rate": 1.824200692417152e-05, "loss": 0.0124, "step": 58650 }, { "grad_norm": 0.18812361359596252, "learning_rate": 1.8226045785660912e-05, "loss": 0.0124, "step": 58660 }, { "grad_norm": 0.4819086492061615, "learning_rate": 1.821009007645006e-05, "loss": 0.0163, "step": 58670 }, { "grad_norm": 0.2915036976337433, "learning_rate": 1.8194139799265357e-05, "loss": 0.0153, "step": 58680 }, { "grad_norm": 0.2725202739238739, "learning_rate": 1.8178194956832295e-05, "loss": 0.0172, "step": 58690 }, { "grad_norm": 0.16049061715602875, "learning_rate": 1.8162255551875346e-05, "loss": 0.0144, "step": 58700 }, { "grad_norm": 0.23216859996318817, "learning_rate": 1.8146321587118194e-05, "loss": 0.0188, "step": 58710 }, { "grad_norm": 0.3446981608867645, "learning_rate": 1.8130393065283448e-05, "loss": 0.0148, "step": 58720 }, { "grad_norm": 0.12004721164703369, "learning_rate": 1.8114469989092925e-05, "loss": 0.013, "step": 58730 }, { "grad_norm": 0.23932838439941406, "learning_rate": 1.8098552361267397e-05, "loss": 0.0149, "step": 58740 }, { "grad_norm": 0.18309493362903595, "learning_rate": 1.8082640184526763e-05, "loss": 0.0164, "step": 58750 }, { "grad_norm": 0.22189803421497345, "learning_rate": 1.806673346158999e-05, "loss": 0.0176, "step": 58760 }, { "grad_norm": 0.252389132976532, "learning_rate": 1.8050832195175067e-05, "loss": 0.0175, "step": 58770 }, { "grad_norm": 0.3713877499103546, "learning_rate": 1.8034936387999136e-05, "loss": 0.0188, "step": 58780 }, { "grad_norm": 0.1691790670156479, "learning_rate": 1.8019046042778315e-05, "loss": 0.0133, "step": 58790 }, { "grad_norm": 0.16178546845912933, "learning_rate": 1.800316116222785e-05, "loss": 0.0135, "step": 58800 }, { "grad_norm": 0.1786968857049942, "learning_rate": 1.7987281749062018e-05, "loss": 0.0322, "step": 58810 }, { "grad_norm": 0.26829397678375244, "learning_rate": 1.7971407805994195e-05, "loss": 0.0286, "step": 58820 }, { "grad_norm": 0.2265980839729309, "learning_rate": 1.7955539335736787e-05, "loss": 0.0153, "step": 58830 }, { "grad_norm": 0.1414577066898346, "learning_rate": 1.7939676341001304e-05, "loss": 0.0223, "step": 58840 }, { "grad_norm": 0.2485826164484024, "learning_rate": 1.7923818824498275e-05, "loss": 0.0147, "step": 58850 }, { "grad_norm": 0.3304794132709503, "learning_rate": 1.7907966788937315e-05, "loss": 0.0149, "step": 58860 }, { "grad_norm": 0.17839768528938293, "learning_rate": 1.7892120237027116e-05, "loss": 0.0218, "step": 58870 }, { "grad_norm": 0.2145804613828659, "learning_rate": 1.7876279171475413e-05, "loss": 0.0136, "step": 58880 }, { "grad_norm": 0.18441380560398102, "learning_rate": 1.7860443594989028e-05, "loss": 0.017, "step": 58890 }, { "grad_norm": 0.2487783581018448, "learning_rate": 1.784461351027379e-05, "loss": 0.0107, "step": 58900 }, { "grad_norm": 0.5370379090309143, "learning_rate": 1.7828788920034677e-05, "loss": 0.0217, "step": 58910 }, { "grad_norm": 0.25356999039649963, "learning_rate": 1.7812969826975623e-05, "loss": 0.0146, "step": 58920 }, { "grad_norm": 0.17480063438415527, "learning_rate": 1.7797156233799738e-05, "loss": 0.0167, "step": 58930 }, { "grad_norm": 0.17332692444324493, "learning_rate": 1.7781348143209094e-05, "loss": 0.0173, "step": 58940 }, { "grad_norm": 0.2175191044807434, "learning_rate": 1.7765545557904862e-05, "loss": 0.0253, "step": 58950 }, { "grad_norm": 0.23097607493400574, "learning_rate": 1.7749748480587302e-05, "loss": 0.0195, "step": 58960 }, { "grad_norm": 0.19332700967788696, "learning_rate": 1.773395691395564e-05, "loss": 0.0161, "step": 58970 }, { "grad_norm": 0.1360304355621338, "learning_rate": 1.7718170860708305e-05, "loss": 0.0133, "step": 58980 }, { "grad_norm": 0.19706827402114868, "learning_rate": 1.770239032354264e-05, "loss": 0.0123, "step": 58990 }, { "grad_norm": 0.2765327990055084, "learning_rate": 1.7686615305155124e-05, "loss": 0.0252, "step": 59000 }, { "grad_norm": 0.2455165535211563, "learning_rate": 1.767084580824128e-05, "loss": 0.0133, "step": 59010 }, { "grad_norm": 0.48661941289901733, "learning_rate": 1.765508183549569e-05, "loss": 0.0169, "step": 59020 }, { "grad_norm": 0.23866046965122223, "learning_rate": 1.763932338961199e-05, "loss": 0.0172, "step": 59030 }, { "grad_norm": 0.29831838607788086, "learning_rate": 1.762357047328284e-05, "loss": 0.0111, "step": 59040 }, { "grad_norm": 0.2625732123851776, "learning_rate": 1.7607823089199997e-05, "loss": 0.0206, "step": 59050 }, { "grad_norm": 0.14513516426086426, "learning_rate": 1.7592081240054265e-05, "loss": 0.018, "step": 59060 }, { "grad_norm": 0.258242130279541, "learning_rate": 1.7576344928535488e-05, "loss": 0.0157, "step": 59070 }, { "grad_norm": 0.38614580035209656, "learning_rate": 1.7560614157332573e-05, "loss": 0.0194, "step": 59080 }, { "grad_norm": 0.23870672285556793, "learning_rate": 1.7544888929133495e-05, "loss": 0.0205, "step": 59090 }, { "grad_norm": 0.14171887934207916, "learning_rate": 1.752916924662522e-05, "loss": 0.0135, "step": 59100 }, { "grad_norm": 0.15290416777133942, "learning_rate": 1.751345511249387e-05, "loss": 0.0143, "step": 59110 }, { "grad_norm": 0.16355058550834656, "learning_rate": 1.749774652942452e-05, "loss": 0.0154, "step": 59120 }, { "grad_norm": 0.26133808493614197, "learning_rate": 1.748204350010135e-05, "loss": 0.0156, "step": 59130 }, { "grad_norm": 0.17524400353431702, "learning_rate": 1.7466346027207574e-05, "loss": 0.0167, "step": 59140 }, { "grad_norm": 0.16961722075939178, "learning_rate": 1.745065411342547e-05, "loss": 0.0156, "step": 59150 }, { "grad_norm": 0.21771250665187836, "learning_rate": 1.7434967761436366e-05, "loss": 0.0191, "step": 59160 }, { "grad_norm": 0.3058472275733948, "learning_rate": 1.741928697392058e-05, "loss": 0.0151, "step": 59170 }, { "grad_norm": 0.24669483304023743, "learning_rate": 1.7403611753557597e-05, "loss": 0.0129, "step": 59180 }, { "grad_norm": 0.18142199516296387, "learning_rate": 1.738794210302584e-05, "loss": 0.0117, "step": 59190 }, { "grad_norm": 0.12202700227499008, "learning_rate": 1.7372278025002837e-05, "loss": 0.0119, "step": 59200 }, { "grad_norm": 0.250662237405777, "learning_rate": 1.7356619522165164e-05, "loss": 0.0166, "step": 59210 }, { "grad_norm": 0.15536606311798096, "learning_rate": 1.7340966597188378e-05, "loss": 0.0112, "step": 59220 }, { "grad_norm": 0.15689900517463684, "learning_rate": 1.732531925274722e-05, "loss": 0.0135, "step": 59230 }, { "grad_norm": 0.22485975921154022, "learning_rate": 1.7309677491515318e-05, "loss": 0.0149, "step": 59240 }, { "grad_norm": 0.2177211344242096, "learning_rate": 1.7294041316165455e-05, "loss": 0.0117, "step": 59250 }, { "grad_norm": 0.4576673209667206, "learning_rate": 1.727841072936942e-05, "loss": 0.0183, "step": 59260 }, { "grad_norm": 0.3059578835964203, "learning_rate": 1.7262785733798058e-05, "loss": 0.0126, "step": 59270 }, { "grad_norm": 0.2729240953922272, "learning_rate": 1.7247166332121246e-05, "loss": 0.0152, "step": 59280 }, { "grad_norm": 0.23152385652065277, "learning_rate": 1.7231552527007933e-05, "loss": 0.013, "step": 59290 }, { "grad_norm": 0.2598593831062317, "learning_rate": 1.721594432112606e-05, "loss": 0.0131, "step": 59300 }, { "grad_norm": 0.22736448049545288, "learning_rate": 1.7200341717142655e-05, "loss": 0.0188, "step": 59310 }, { "grad_norm": 0.3397277295589447, "learning_rate": 1.718474471772377e-05, "loss": 0.015, "step": 59320 }, { "grad_norm": 0.14528654515743256, "learning_rate": 1.7169153325534528e-05, "loss": 0.0147, "step": 59330 }, { "grad_norm": 0.24736785888671875, "learning_rate": 1.7153567543239045e-05, "loss": 0.0156, "step": 59340 }, { "grad_norm": 0.1976906806230545, "learning_rate": 1.7137987373500525e-05, "loss": 0.0119, "step": 59350 }, { "grad_norm": 0.13975101709365845, "learning_rate": 1.7122412818981198e-05, "loss": 0.0149, "step": 59360 }, { "grad_norm": 0.23596787452697754, "learning_rate": 1.7106843882342283e-05, "loss": 0.0165, "step": 59370 }, { "grad_norm": 0.23493792116641998, "learning_rate": 1.709128056624415e-05, "loss": 0.0157, "step": 59380 }, { "grad_norm": 0.25538939237594604, "learning_rate": 1.70757228733461e-05, "loss": 0.0117, "step": 59390 }, { "grad_norm": 0.1773449182510376, "learning_rate": 1.706017080630653e-05, "loss": 0.0146, "step": 59400 }, { "grad_norm": 0.21150536835193634, "learning_rate": 1.7044624367782873e-05, "loss": 0.0111, "step": 59410 }, { "grad_norm": 0.15663594007492065, "learning_rate": 1.7029083560431553e-05, "loss": 0.025, "step": 59420 }, { "grad_norm": 0.26223477721214294, "learning_rate": 1.701354838690813e-05, "loss": 0.0194, "step": 59430 }, { "grad_norm": 0.4155566990375519, "learning_rate": 1.6998018849867087e-05, "loss": 0.0133, "step": 59440 }, { "grad_norm": 0.1762935221195221, "learning_rate": 1.698249495196202e-05, "loss": 0.0187, "step": 59450 }, { "grad_norm": 0.18507421016693115, "learning_rate": 1.6966976695845528e-05, "loss": 0.0134, "step": 59460 }, { "grad_norm": 0.21654623746871948, "learning_rate": 1.6951464084169268e-05, "loss": 0.0132, "step": 59470 }, { "grad_norm": 0.32931798696517944, "learning_rate": 1.6935957119583916e-05, "loss": 0.0223, "step": 59480 }, { "grad_norm": 0.17786805331707, "learning_rate": 1.6920455804739205e-05, "loss": 0.0137, "step": 59490 }, { "grad_norm": 0.20869623124599457, "learning_rate": 1.6904960142283855e-05, "loss": 0.0155, "step": 59500 }, { "grad_norm": 0.1988275796175003, "learning_rate": 1.6889470134865666e-05, "loss": 0.0118, "step": 59510 }, { "grad_norm": 0.18157021701335907, "learning_rate": 1.687398578513145e-05, "loss": 0.014, "step": 59520 }, { "grad_norm": 0.7104780077934265, "learning_rate": 1.6858507095727066e-05, "loss": 0.0168, "step": 59530 }, { "grad_norm": 0.21644927561283112, "learning_rate": 1.6843034069297403e-05, "loss": 0.012, "step": 59540 }, { "grad_norm": 0.28138867020606995, "learning_rate": 1.682756670848637e-05, "loss": 0.0157, "step": 59550 }, { "grad_norm": 0.24958553910255432, "learning_rate": 1.6812105015936937e-05, "loss": 0.0113, "step": 59560 }, { "grad_norm": 0.4224136471748352, "learning_rate": 1.6796648994291027e-05, "loss": 0.0184, "step": 59570 }, { "grad_norm": 0.2110956311225891, "learning_rate": 1.678119864618973e-05, "loss": 0.0163, "step": 59580 }, { "grad_norm": 0.23510360717773438, "learning_rate": 1.6765753974273023e-05, "loss": 0.0164, "step": 59590 }, { "grad_norm": 0.737060010433197, "learning_rate": 1.675031498118001e-05, "loss": 0.0127, "step": 59600 }, { "grad_norm": 0.376605749130249, "learning_rate": 1.6734881669548804e-05, "loss": 0.0142, "step": 59610 }, { "grad_norm": 0.21397839486598969, "learning_rate": 1.671945404201648e-05, "loss": 0.0177, "step": 59620 }, { "grad_norm": 0.2424723207950592, "learning_rate": 1.670403210121927e-05, "loss": 0.0137, "step": 59630 }, { "grad_norm": 0.26631978154182434, "learning_rate": 1.6688615849792312e-05, "loss": 0.011, "step": 59640 }, { "grad_norm": 0.3421257436275482, "learning_rate": 1.6673205290369832e-05, "loss": 0.0165, "step": 59650 }, { "grad_norm": 0.32054662704467773, "learning_rate": 1.6657800425585086e-05, "loss": 0.0174, "step": 59660 }, { "grad_norm": 0.21225720643997192, "learning_rate": 1.6642401258070327e-05, "loss": 0.0229, "step": 59670 }, { "grad_norm": 0.4286501407623291, "learning_rate": 1.6627007790456884e-05, "loss": 0.0159, "step": 59680 }, { "grad_norm": 0.10885649174451828, "learning_rate": 1.6611620025375037e-05, "loss": 0.0121, "step": 59690 }, { "grad_norm": 0.21797506511211395, "learning_rate": 1.6596237965454154e-05, "loss": 0.016, "step": 59700 }, { "grad_norm": 0.4464711546897888, "learning_rate": 1.6580861613322607e-05, "loss": 0.0162, "step": 59710 }, { "grad_norm": 0.28865545988082886, "learning_rate": 1.6565490971607796e-05, "loss": 0.0126, "step": 59720 }, { "grad_norm": 0.20955024659633636, "learning_rate": 1.655012604293614e-05, "loss": 0.0167, "step": 59730 }, { "grad_norm": 0.30300667881965637, "learning_rate": 1.653476682993309e-05, "loss": 0.0177, "step": 59740 }, { "grad_norm": 0.3280435800552368, "learning_rate": 1.651941333522311e-05, "loss": 0.0147, "step": 59750 }, { "grad_norm": 0.2864086925983429, "learning_rate": 1.6504065561429715e-05, "loss": 0.0144, "step": 59760 }, { "grad_norm": 0.22050970792770386, "learning_rate": 1.6488723511175385e-05, "loss": 0.0212, "step": 59770 }, { "grad_norm": 0.24845148622989655, "learning_rate": 1.6473387187081668e-05, "loss": 0.0157, "step": 59780 }, { "grad_norm": 0.2132890522480011, "learning_rate": 1.6458056591769123e-05, "loss": 0.0159, "step": 59790 }, { "grad_norm": 0.3308625817298889, "learning_rate": 1.6442731727857335e-05, "loss": 0.0173, "step": 59800 }, { "grad_norm": 0.23725761473178864, "learning_rate": 1.642741259796492e-05, "loss": 0.0118, "step": 59810 }, { "grad_norm": 0.24691911041736603, "learning_rate": 1.641209920470944e-05, "loss": 0.0118, "step": 59820 }, { "grad_norm": 0.32113274931907654, "learning_rate": 1.639679155070762e-05, "loss": 0.0127, "step": 59830 }, { "grad_norm": 0.2747572660446167, "learning_rate": 1.6381489638575048e-05, "loss": 0.017, "step": 59840 }, { "grad_norm": 0.13624230027198792, "learning_rate": 1.636619347092643e-05, "loss": 0.0105, "step": 59850 }, { "grad_norm": 0.3013533353805542, "learning_rate": 1.6350903050375476e-05, "loss": 0.0162, "step": 59860 }, { "grad_norm": 0.3053598701953888, "learning_rate": 1.6335618379534856e-05, "loss": 0.0145, "step": 59870 }, { "grad_norm": 0.23723502457141876, "learning_rate": 1.6320339461016364e-05, "loss": 0.0118, "step": 59880 }, { "grad_norm": 0.42629051208496094, "learning_rate": 1.6305066297430687e-05, "loss": 0.0178, "step": 59890 }, { "grad_norm": 0.2575223445892334, "learning_rate": 1.6289798891387654e-05, "loss": 0.0194, "step": 59900 }, { "grad_norm": 0.2681104838848114, "learning_rate": 1.6274537245495995e-05, "loss": 0.0131, "step": 59910 }, { "grad_norm": 0.2380913496017456, "learning_rate": 1.6259281362363527e-05, "loss": 0.0134, "step": 59920 }, { "grad_norm": 0.3453996479511261, "learning_rate": 1.6244031244597068e-05, "loss": 0.0208, "step": 59930 }, { "grad_norm": 0.30963942408561707, "learning_rate": 1.622878689480244e-05, "loss": 0.0128, "step": 59940 }, { "grad_norm": 0.2437107414007187, "learning_rate": 1.6213548315584498e-05, "loss": 0.0199, "step": 59950 }, { "grad_norm": 0.23137931525707245, "learning_rate": 1.6198315509547074e-05, "loss": 0.0154, "step": 59960 }, { "grad_norm": 0.17593862116336823, "learning_rate": 1.6183088479293056e-05, "loss": 0.0134, "step": 59970 }, { "grad_norm": 0.23658615350723267, "learning_rate": 1.6167867227424316e-05, "loss": 0.0139, "step": 59980 }, { "grad_norm": 0.1875712126493454, "learning_rate": 1.615265175654176e-05, "loss": 0.0166, "step": 59990 }, { "grad_norm": 0.31441420316696167, "learning_rate": 1.613744206924529e-05, "loss": 0.017, "step": 60000 } ], "logging_steps": 10, "max_steps": 80000, "num_input_tokens_seen": 0, "num_train_epochs": 9223372036854775807, "save_steps": 60000, "stateful_callbacks": { "TrainerControl": { "args": { "should_epoch_stop": false, "should_evaluate": false, "should_log": false, "should_save": true, "should_training_stop": false }, "attributes": {} } }, "total_flos": 0.0, "train_batch_size": 8, "trial_name": null, "trial_params": null }