{ "best_global_step": null, "best_metric": null, "best_model_checkpoint": null, "epoch": 1.0, "eval_steps": 500, "global_step": 100000, "is_hyper_param_search": false, "is_local_process_zero": true, "is_world_process_zero": true, "log_history": [ { "grad_norm": 1.009881615638733, "learning_rate": 1.8e-07, "loss": 1.5773, "step": 10 }, { "grad_norm": 1.259894847869873, "learning_rate": 3.8e-07, "loss": 1.5971, "step": 20 }, { "grad_norm": 1.0788187980651855, "learning_rate": 5.8e-07, "loss": 1.5405, "step": 30 }, { "grad_norm": 1.416263461112976, "learning_rate": 7.8e-07, "loss": 1.5672, "step": 40 }, { "grad_norm": 1.1153671741485596, "learning_rate": 9.8e-07, "loss": 1.5524, "step": 50 }, { "grad_norm": 1.3952919244766235, "learning_rate": 1.18e-06, "loss": 1.5269, "step": 60 }, { "grad_norm": 1.7639918327331543, "learning_rate": 1.3800000000000001e-06, "loss": 1.4261, "step": 70 }, { "grad_norm": 1.722119927406311, "learning_rate": 1.5800000000000003e-06, "loss": 1.372, "step": 80 }, { "grad_norm": 9.242517471313477, "learning_rate": 1.7800000000000001e-06, "loss": 1.3063, "step": 90 }, { "grad_norm": 1.5149022340774536, "learning_rate": 1.98e-06, "loss": 1.3181, "step": 100 }, { "grad_norm": 1.4771289825439453, "learning_rate": 2.1800000000000003e-06, "loss": 1.2915, "step": 110 }, { "grad_norm": 2.5211093425750732, "learning_rate": 2.38e-06, "loss": 1.2735, "step": 120 }, { "grad_norm": 1.7284237146377563, "learning_rate": 2.5800000000000003e-06, "loss": 1.2586, "step": 130 }, { "grad_norm": 1.9354908466339111, "learning_rate": 2.78e-06, "loss": 1.2273, "step": 140 }, { "grad_norm": 1.5287550687789917, "learning_rate": 2.9800000000000003e-06, "loss": 1.1576, "step": 150 }, { "grad_norm": 2.135711193084717, "learning_rate": 3.1800000000000005e-06, "loss": 1.1841, "step": 160 }, { "grad_norm": 2.1713650226593018, "learning_rate": 3.38e-06, "loss": 1.1266, "step": 170 }, { "grad_norm": 1.606376051902771, "learning_rate": 3.58e-06, "loss": 1.1051, "step": 180 }, { "grad_norm": 1.7015671730041504, "learning_rate": 3.7800000000000002e-06, "loss": 1.0926, "step": 190 }, { "grad_norm": 2.365687370300293, "learning_rate": 3.98e-06, "loss": 1.1253, "step": 200 }, { "grad_norm": 2.0042247772216797, "learning_rate": 4.18e-06, "loss": 1.1214, "step": 210 }, { "grad_norm": 1.8774933815002441, "learning_rate": 4.38e-06, "loss": 1.119, "step": 220 }, { "grad_norm": 1.7164798974990845, "learning_rate": 4.58e-06, "loss": 1.112, "step": 230 }, { "grad_norm": 1.5647478103637695, "learning_rate": 4.780000000000001e-06, "loss": 1.0888, "step": 240 }, { "grad_norm": 1.3002499341964722, "learning_rate": 4.98e-06, "loss": 1.0848, "step": 250 }, { "grad_norm": 2.0346174240112305, "learning_rate": 5.18e-06, "loss": 1.0666, "step": 260 }, { "grad_norm": 1.3889949321746826, "learning_rate": 5.38e-06, "loss": 1.0831, "step": 270 }, { "grad_norm": 1.5031547546386719, "learning_rate": 5.580000000000001e-06, "loss": 1.0704, "step": 280 }, { "grad_norm": 1.537394404411316, "learning_rate": 5.78e-06, "loss": 1.1028, "step": 290 }, { "grad_norm": 1.6205191612243652, "learning_rate": 5.98e-06, "loss": 1.0711, "step": 300 }, { "grad_norm": 1.6051236391067505, "learning_rate": 6.18e-06, "loss": 1.1087, "step": 310 }, { "grad_norm": 1.753230094909668, "learning_rate": 6.38e-06, "loss": 1.0915, "step": 320 }, { "grad_norm": 1.438766360282898, "learning_rate": 6.58e-06, "loss": 1.0734, "step": 330 }, { "grad_norm": 1.848416805267334, "learning_rate": 6.78e-06, "loss": 1.091, "step": 340 }, { "grad_norm": 1.5804449319839478, "learning_rate": 6.98e-06, "loss": 1.0807, "step": 350 }, { "grad_norm": 1.5181145668029785, "learning_rate": 7.180000000000001e-06, "loss": 1.0827, "step": 360 }, { "grad_norm": 1.6110183000564575, "learning_rate": 7.3800000000000005e-06, "loss": 1.0831, "step": 370 }, { "grad_norm": 1.6137192249298096, "learning_rate": 7.580000000000001e-06, "loss": 1.094, "step": 380 }, { "grad_norm": 1.8045647144317627, "learning_rate": 7.78e-06, "loss": 1.0855, "step": 390 }, { "grad_norm": 1.5144387483596802, "learning_rate": 7.98e-06, "loss": 1.0757, "step": 400 }, { "grad_norm": 1.5258187055587769, "learning_rate": 8.18e-06, "loss": 1.088, "step": 410 }, { "grad_norm": 1.4524837732315063, "learning_rate": 8.380000000000001e-06, "loss": 1.0764, "step": 420 }, { "grad_norm": 1.292316198348999, "learning_rate": 8.580000000000001e-06, "loss": 1.0634, "step": 430 }, { "grad_norm": 1.97196626663208, "learning_rate": 8.78e-06, "loss": 1.0602, "step": 440 }, { "grad_norm": 1.2959102392196655, "learning_rate": 8.98e-06, "loss": 1.0242, "step": 450 }, { "grad_norm": 1.8363906145095825, "learning_rate": 9.180000000000002e-06, "loss": 1.019, "step": 460 }, { "grad_norm": 1.7092020511627197, "learning_rate": 9.38e-06, "loss": 1.0271, "step": 470 }, { "grad_norm": 2.938270092010498, "learning_rate": 9.58e-06, "loss": 1.0038, "step": 480 }, { "grad_norm": 1.9421792030334473, "learning_rate": 9.78e-06, "loss": 1.0079, "step": 490 }, { "grad_norm": 2.184401035308838, "learning_rate": 9.980000000000001e-06, "loss": 0.9812, "step": 500 }, { "grad_norm": 2.752075672149658, "learning_rate": 1.018e-05, "loss": 0.9554, "step": 510 }, { "grad_norm": 2.4565112590789795, "learning_rate": 1.038e-05, "loss": 0.9048, "step": 520 }, { "grad_norm": 2.56965708732605, "learning_rate": 1.058e-05, "loss": 0.8969, "step": 530 }, { "grad_norm": 2.3528804779052734, "learning_rate": 1.0780000000000002e-05, "loss": 0.8627, "step": 540 }, { "grad_norm": 2.6987416744232178, "learning_rate": 1.098e-05, "loss": 0.8574, "step": 550 }, { "grad_norm": 2.5286684036254883, "learning_rate": 1.118e-05, "loss": 0.8331, "step": 560 }, { "grad_norm": 2.277613878250122, "learning_rate": 1.1380000000000001e-05, "loss": 0.7978, "step": 570 }, { "grad_norm": 2.1421053409576416, "learning_rate": 1.1580000000000001e-05, "loss": 0.7486, "step": 580 }, { "grad_norm": 2.8423261642456055, "learning_rate": 1.178e-05, "loss": 0.7471, "step": 590 }, { "grad_norm": 2.253740072250366, "learning_rate": 1.198e-05, "loss": 0.7088, "step": 600 }, { "grad_norm": 2.873250961303711, "learning_rate": 1.2180000000000002e-05, "loss": 0.7183, "step": 610 }, { "grad_norm": 2.181898355484009, "learning_rate": 1.238e-05, "loss": 0.6747, "step": 620 }, { "grad_norm": 2.4785451889038086, "learning_rate": 1.258e-05, "loss": 0.6794, "step": 630 }, { "grad_norm": 2.788999080657959, "learning_rate": 1.278e-05, "loss": 0.6097, "step": 640 }, { "grad_norm": 3.2125184535980225, "learning_rate": 1.2980000000000001e-05, "loss": 0.5787, "step": 650 }, { "grad_norm": 3.805216073989868, "learning_rate": 1.3180000000000001e-05, "loss": 0.5837, "step": 660 }, { "grad_norm": 3.5126378536224365, "learning_rate": 1.338e-05, "loss": 0.5223, "step": 670 }, { "grad_norm": 3.1214189529418945, "learning_rate": 1.358e-05, "loss": 0.497, "step": 680 }, { "grad_norm": 3.224828004837036, "learning_rate": 1.3780000000000002e-05, "loss": 0.505, "step": 690 }, { "grad_norm": 3.7324001789093018, "learning_rate": 1.3980000000000002e-05, "loss": 0.4958, "step": 700 }, { "grad_norm": 3.9088408946990967, "learning_rate": 1.4180000000000001e-05, "loss": 0.4206, "step": 710 }, { "grad_norm": 4.0095367431640625, "learning_rate": 1.4380000000000001e-05, "loss": 0.4654, "step": 720 }, { "grad_norm": 3.301685094833374, "learning_rate": 1.4580000000000003e-05, "loss": 0.4064, "step": 730 }, { "grad_norm": 2.4911561012268066, "learning_rate": 1.4779999999999999e-05, "loss": 0.3785, "step": 740 }, { "grad_norm": 2.540182590484619, "learning_rate": 1.4979999999999999e-05, "loss": 0.3951, "step": 750 }, { "grad_norm": 3.2902750968933105, "learning_rate": 1.518e-05, "loss": 0.3701, "step": 760 }, { "grad_norm": 2.894428253173828, "learning_rate": 1.538e-05, "loss": 0.3673, "step": 770 }, { "grad_norm": 2.9845657348632812, "learning_rate": 1.558e-05, "loss": 0.3828, "step": 780 }, { "grad_norm": 2.842771530151367, "learning_rate": 1.578e-05, "loss": 0.3145, "step": 790 }, { "grad_norm": 2.769685745239258, "learning_rate": 1.598e-05, "loss": 0.3006, "step": 800 }, { "grad_norm": 2.841130495071411, "learning_rate": 1.618e-05, "loss": 0.2944, "step": 810 }, { "grad_norm": 2.8243138790130615, "learning_rate": 1.6380000000000002e-05, "loss": 0.2891, "step": 820 }, { "grad_norm": 3.187216281890869, "learning_rate": 1.658e-05, "loss": 0.3182, "step": 830 }, { "grad_norm": 3.423670768737793, "learning_rate": 1.6780000000000002e-05, "loss": 0.3052, "step": 840 }, { "grad_norm": 2.411310911178589, "learning_rate": 1.698e-05, "loss": 0.3019, "step": 850 }, { "grad_norm": 3.228686809539795, "learning_rate": 1.718e-05, "loss": 0.2539, "step": 860 }, { "grad_norm": 2.5975024700164795, "learning_rate": 1.7380000000000003e-05, "loss": 0.2879, "step": 870 }, { "grad_norm": 3.514578104019165, "learning_rate": 1.758e-05, "loss": 0.2661, "step": 880 }, { "grad_norm": 2.811167001724243, "learning_rate": 1.7780000000000003e-05, "loss": 0.2693, "step": 890 }, { "grad_norm": 2.9976494312286377, "learning_rate": 1.798e-05, "loss": 0.2529, "step": 900 }, { "grad_norm": 3.3439524173736572, "learning_rate": 1.818e-05, "loss": 0.232, "step": 910 }, { "grad_norm": 2.453122138977051, "learning_rate": 1.838e-05, "loss": 0.2006, "step": 920 }, { "grad_norm": 3.09694766998291, "learning_rate": 1.858e-05, "loss": 0.2333, "step": 930 }, { "grad_norm": 3.697084903717041, "learning_rate": 1.878e-05, "loss": 0.2397, "step": 940 }, { "grad_norm": 2.729768991470337, "learning_rate": 1.898e-05, "loss": 0.2126, "step": 950 }, { "grad_norm": 2.7003417015075684, "learning_rate": 1.918e-05, "loss": 0.2319, "step": 960 }, { "grad_norm": 3.9751992225646973, "learning_rate": 1.938e-05, "loss": 0.2419, "step": 970 }, { "grad_norm": 3.453742742538452, "learning_rate": 1.9580000000000002e-05, "loss": 0.2165, "step": 980 }, { "grad_norm": 2.857269763946533, "learning_rate": 1.978e-05, "loss": 0.1903, "step": 990 }, { "grad_norm": 3.688100576400757, "learning_rate": 1.9980000000000002e-05, "loss": 0.2034, "step": 1000 }, { "grad_norm": 2.7365951538085938, "learning_rate": 2.0180000000000003e-05, "loss": 0.2224, "step": 1010 }, { "grad_norm": 2.6372408866882324, "learning_rate": 2.038e-05, "loss": 0.2358, "step": 1020 }, { "grad_norm": 3.076603651046753, "learning_rate": 2.0580000000000003e-05, "loss": 0.1845, "step": 1030 }, { "grad_norm": 2.4914908409118652, "learning_rate": 2.078e-05, "loss": 0.221, "step": 1040 }, { "grad_norm": 2.906294822692871, "learning_rate": 2.098e-05, "loss": 0.2093, "step": 1050 }, { "grad_norm": 2.381995677947998, "learning_rate": 2.118e-05, "loss": 0.216, "step": 1060 }, { "grad_norm": 2.1413626670837402, "learning_rate": 2.138e-05, "loss": 0.2157, "step": 1070 }, { "grad_norm": 2.7414536476135254, "learning_rate": 2.158e-05, "loss": 0.2202, "step": 1080 }, { "grad_norm": 1.5640089511871338, "learning_rate": 2.178e-05, "loss": 0.1927, "step": 1090 }, { "grad_norm": 3.453596591949463, "learning_rate": 2.198e-05, "loss": 0.1664, "step": 1100 }, { "grad_norm": 3.390583038330078, "learning_rate": 2.218e-05, "loss": 0.2362, "step": 1110 }, { "grad_norm": 1.8466181755065918, "learning_rate": 2.2380000000000003e-05, "loss": 0.1646, "step": 1120 }, { "grad_norm": 2.411717414855957, "learning_rate": 2.258e-05, "loss": 0.2155, "step": 1130 }, { "grad_norm": 2.780003070831299, "learning_rate": 2.2780000000000002e-05, "loss": 0.25, "step": 1140 }, { "grad_norm": 2.5212554931640625, "learning_rate": 2.298e-05, "loss": 0.2243, "step": 1150 }, { "grad_norm": 1.591482400894165, "learning_rate": 2.318e-05, "loss": 0.1931, "step": 1160 }, { "grad_norm": 1.9623279571533203, "learning_rate": 2.3380000000000003e-05, "loss": 0.1835, "step": 1170 }, { "grad_norm": 2.539191484451294, "learning_rate": 2.358e-05, "loss": 0.1669, "step": 1180 }, { "grad_norm": 2.9760167598724365, "learning_rate": 2.3780000000000003e-05, "loss": 0.1928, "step": 1190 }, { "grad_norm": 4.162106037139893, "learning_rate": 2.398e-05, "loss": 0.2018, "step": 1200 }, { "grad_norm": 4.144816875457764, "learning_rate": 2.418e-05, "loss": 0.2132, "step": 1210 }, { "grad_norm": 2.5186145305633545, "learning_rate": 2.438e-05, "loss": 0.1752, "step": 1220 }, { "grad_norm": 3.1753339767456055, "learning_rate": 2.4580000000000002e-05, "loss": 0.1895, "step": 1230 }, { "grad_norm": 2.759690523147583, "learning_rate": 2.478e-05, "loss": 0.1836, "step": 1240 }, { "grad_norm": 2.7851529121398926, "learning_rate": 2.498e-05, "loss": 0.1881, "step": 1250 }, { "grad_norm": 2.981839418411255, "learning_rate": 2.5180000000000003e-05, "loss": 0.1849, "step": 1260 }, { "grad_norm": 2.6519060134887695, "learning_rate": 2.5380000000000004e-05, "loss": 0.1606, "step": 1270 }, { "grad_norm": 1.4646062850952148, "learning_rate": 2.5580000000000002e-05, "loss": 0.1858, "step": 1280 }, { "grad_norm": 1.5296831130981445, "learning_rate": 2.5779999999999997e-05, "loss": 0.1837, "step": 1290 }, { "grad_norm": 2.0687460899353027, "learning_rate": 2.598e-05, "loss": 0.1668, "step": 1300 }, { "grad_norm": 2.105811834335327, "learning_rate": 2.618e-05, "loss": 0.197, "step": 1310 }, { "grad_norm": 2.480358123779297, "learning_rate": 2.6379999999999998e-05, "loss": 0.1764, "step": 1320 }, { "grad_norm": 1.9675357341766357, "learning_rate": 2.658e-05, "loss": 0.1551, "step": 1330 }, { "grad_norm": 2.034566879272461, "learning_rate": 2.678e-05, "loss": 0.1535, "step": 1340 }, { "grad_norm": 2.4424307346343994, "learning_rate": 2.698e-05, "loss": 0.1914, "step": 1350 }, { "grad_norm": 2.606017827987671, "learning_rate": 2.718e-05, "loss": 0.1744, "step": 1360 }, { "grad_norm": 2.368084192276001, "learning_rate": 2.738e-05, "loss": 0.162, "step": 1370 }, { "grad_norm": 2.159367084503174, "learning_rate": 2.758e-05, "loss": 0.1608, "step": 1380 }, { "grad_norm": 2.5786995887756348, "learning_rate": 2.778e-05, "loss": 0.1724, "step": 1390 }, { "grad_norm": 2.122251033782959, "learning_rate": 2.798e-05, "loss": 0.1578, "step": 1400 }, { "grad_norm": 1.919682502746582, "learning_rate": 2.818e-05, "loss": 0.1751, "step": 1410 }, { "grad_norm": 1.962412714958191, "learning_rate": 2.8380000000000003e-05, "loss": 0.1841, "step": 1420 }, { "grad_norm": 1.664448618888855, "learning_rate": 2.858e-05, "loss": 0.1505, "step": 1430 }, { "grad_norm": 1.7259812355041504, "learning_rate": 2.8780000000000002e-05, "loss": 0.1758, "step": 1440 }, { "grad_norm": 1.817569375038147, "learning_rate": 2.898e-05, "loss": 0.1924, "step": 1450 }, { "grad_norm": 2.0060973167419434, "learning_rate": 2.9180000000000002e-05, "loss": 0.1914, "step": 1460 }, { "grad_norm": 2.2462689876556396, "learning_rate": 2.9380000000000003e-05, "loss": 0.165, "step": 1470 }, { "grad_norm": 1.9764602184295654, "learning_rate": 2.958e-05, "loss": 0.1624, "step": 1480 }, { "grad_norm": 2.0008327960968018, "learning_rate": 2.9780000000000003e-05, "loss": 0.1632, "step": 1490 }, { "grad_norm": 2.020641326904297, "learning_rate": 2.998e-05, "loss": 0.1702, "step": 1500 }, { "grad_norm": 1.612555742263794, "learning_rate": 3.0180000000000002e-05, "loss": 0.1686, "step": 1510 }, { "grad_norm": 2.0490882396698, "learning_rate": 3.0380000000000004e-05, "loss": 0.1774, "step": 1520 }, { "grad_norm": 1.388718843460083, "learning_rate": 3.058e-05, "loss": 0.167, "step": 1530 }, { "grad_norm": 1.8691225051879883, "learning_rate": 3.078e-05, "loss": 0.1579, "step": 1540 }, { "grad_norm": 1.6627726554870605, "learning_rate": 3.0980000000000005e-05, "loss": 0.1395, "step": 1550 }, { "grad_norm": 2.2630114555358887, "learning_rate": 3.118e-05, "loss": 0.183, "step": 1560 }, { "grad_norm": 1.6582595109939575, "learning_rate": 3.138e-05, "loss": 0.145, "step": 1570 }, { "grad_norm": 1.5109256505966187, "learning_rate": 3.1580000000000006e-05, "loss": 0.1472, "step": 1580 }, { "grad_norm": 2.0647196769714355, "learning_rate": 3.1780000000000004e-05, "loss": 0.1798, "step": 1590 }, { "grad_norm": 2.2263152599334717, "learning_rate": 3.198e-05, "loss": 0.1487, "step": 1600 }, { "grad_norm": 1.8444701433181763, "learning_rate": 3.218e-05, "loss": 0.1424, "step": 1610 }, { "grad_norm": 1.9329525232315063, "learning_rate": 3.238e-05, "loss": 0.1637, "step": 1620 }, { "grad_norm": 1.744730830192566, "learning_rate": 3.2579999999999996e-05, "loss": 0.1965, "step": 1630 }, { "grad_norm": 1.6015270948410034, "learning_rate": 3.278e-05, "loss": 0.1602, "step": 1640 }, { "grad_norm": 1.7780007123947144, "learning_rate": 3.298e-05, "loss": 0.1631, "step": 1650 }, { "grad_norm": 1.7094511985778809, "learning_rate": 3.318e-05, "loss": 0.1604, "step": 1660 }, { "grad_norm": 1.6959587335586548, "learning_rate": 3.338e-05, "loss": 0.1279, "step": 1670 }, { "grad_norm": 1.705918550491333, "learning_rate": 3.358e-05, "loss": 0.1427, "step": 1680 }, { "grad_norm": 1.8948055505752563, "learning_rate": 3.378e-05, "loss": 0.1602, "step": 1690 }, { "grad_norm": 1.43036687374115, "learning_rate": 3.398e-05, "loss": 0.1593, "step": 1700 }, { "grad_norm": 2.5530288219451904, "learning_rate": 3.418e-05, "loss": 0.1888, "step": 1710 }, { "grad_norm": 1.2198467254638672, "learning_rate": 3.438e-05, "loss": 0.1725, "step": 1720 }, { "grad_norm": 1.4662491083145142, "learning_rate": 3.4580000000000004e-05, "loss": 0.1554, "step": 1730 }, { "grad_norm": 1.984163761138916, "learning_rate": 3.478e-05, "loss": 0.1395, "step": 1740 }, { "grad_norm": 1.867656946182251, "learning_rate": 3.498e-05, "loss": 0.1597, "step": 1750 }, { "grad_norm": 2.0827674865722656, "learning_rate": 3.518e-05, "loss": 0.1409, "step": 1760 }, { "grad_norm": 1.6779640913009644, "learning_rate": 3.5380000000000003e-05, "loss": 0.1445, "step": 1770 }, { "grad_norm": 1.4428744316101074, "learning_rate": 3.558e-05, "loss": 0.1607, "step": 1780 }, { "grad_norm": 1.933258056640625, "learning_rate": 3.578e-05, "loss": 0.1429, "step": 1790 }, { "grad_norm": 1.3652714490890503, "learning_rate": 3.5980000000000004e-05, "loss": 0.1491, "step": 1800 }, { "grad_norm": 1.49070405960083, "learning_rate": 3.618e-05, "loss": 0.133, "step": 1810 }, { "grad_norm": 2.1000781059265137, "learning_rate": 3.638e-05, "loss": 0.1496, "step": 1820 }, { "grad_norm": 2.2780263423919678, "learning_rate": 3.6580000000000006e-05, "loss": 0.1741, "step": 1830 }, { "grad_norm": 1.5534582138061523, "learning_rate": 3.6780000000000004e-05, "loss": 0.1676, "step": 1840 }, { "grad_norm": 1.5370583534240723, "learning_rate": 3.698e-05, "loss": 0.1818, "step": 1850 }, { "grad_norm": 1.832922101020813, "learning_rate": 3.7180000000000007e-05, "loss": 0.1812, "step": 1860 }, { "grad_norm": 1.0119593143463135, "learning_rate": 3.7380000000000005e-05, "loss": 0.1337, "step": 1870 }, { "grad_norm": 1.3380422592163086, "learning_rate": 3.758e-05, "loss": 0.1282, "step": 1880 }, { "grad_norm": 1.9860190153121948, "learning_rate": 3.778000000000001e-05, "loss": 0.1551, "step": 1890 }, { "grad_norm": 1.2926112413406372, "learning_rate": 3.7980000000000006e-05, "loss": 0.1396, "step": 1900 }, { "grad_norm": 2.0197131633758545, "learning_rate": 3.818e-05, "loss": 0.1659, "step": 1910 }, { "grad_norm": 1.0265758037567139, "learning_rate": 3.838e-05, "loss": 0.1381, "step": 1920 }, { "grad_norm": 1.477844476699829, "learning_rate": 3.858e-05, "loss": 0.1533, "step": 1930 }, { "grad_norm": 2.234030246734619, "learning_rate": 3.878e-05, "loss": 0.1503, "step": 1940 }, { "grad_norm": 1.2991631031036377, "learning_rate": 3.898e-05, "loss": 0.1377, "step": 1950 }, { "grad_norm": 1.6796276569366455, "learning_rate": 3.918e-05, "loss": 0.175, "step": 1960 }, { "grad_norm": 1.4462904930114746, "learning_rate": 3.938e-05, "loss": 0.1376, "step": 1970 }, { "grad_norm": 1.6910706758499146, "learning_rate": 3.958e-05, "loss": 0.1745, "step": 1980 }, { "grad_norm": 2.026552677154541, "learning_rate": 3.978e-05, "loss": 0.1251, "step": 1990 }, { "grad_norm": 1.0838369131088257, "learning_rate": 3.998e-05, "loss": 0.1493, "step": 2000 }, { "grad_norm": 1.8131871223449707, "learning_rate": 4.018e-05, "loss": 0.194, "step": 2010 }, { "grad_norm": 1.7006202936172485, "learning_rate": 4.038e-05, "loss": 0.1374, "step": 2020 }, { "grad_norm": 1.4962269067764282, "learning_rate": 4.058e-05, "loss": 0.1419, "step": 2030 }, { "grad_norm": 1.6430716514587402, "learning_rate": 4.078e-05, "loss": 0.1678, "step": 2040 }, { "grad_norm": 1.5373470783233643, "learning_rate": 4.0980000000000004e-05, "loss": 0.1603, "step": 2050 }, { "grad_norm": 1.5331814289093018, "learning_rate": 4.118e-05, "loss": 0.1413, "step": 2060 }, { "grad_norm": 1.1657742261886597, "learning_rate": 4.138e-05, "loss": 0.1311, "step": 2070 }, { "grad_norm": 1.3034170866012573, "learning_rate": 4.1580000000000005e-05, "loss": 0.1504, "step": 2080 }, { "grad_norm": 1.5379725694656372, "learning_rate": 4.178e-05, "loss": 0.1461, "step": 2090 }, { "grad_norm": 1.4038487672805786, "learning_rate": 4.198e-05, "loss": 0.1476, "step": 2100 }, { "grad_norm": 1.5373846292495728, "learning_rate": 4.2180000000000006e-05, "loss": 0.1675, "step": 2110 }, { "grad_norm": 1.3032809495925903, "learning_rate": 4.2380000000000004e-05, "loss": 0.1415, "step": 2120 }, { "grad_norm": 1.3326354026794434, "learning_rate": 4.258e-05, "loss": 0.1609, "step": 2130 }, { "grad_norm": 1.2883116006851196, "learning_rate": 4.278e-05, "loss": 0.1123, "step": 2140 }, { "grad_norm": 1.3803741931915283, "learning_rate": 4.2980000000000005e-05, "loss": 0.1649, "step": 2150 }, { "grad_norm": 1.193974256515503, "learning_rate": 4.318e-05, "loss": 0.1371, "step": 2160 }, { "grad_norm": 0.9302917122840881, "learning_rate": 4.338e-05, "loss": 0.1532, "step": 2170 }, { "grad_norm": 1.655593991279602, "learning_rate": 4.3580000000000006e-05, "loss": 0.1679, "step": 2180 }, { "grad_norm": 1.3573038578033447, "learning_rate": 4.3780000000000004e-05, "loss": 0.1396, "step": 2190 }, { "grad_norm": 1.728922963142395, "learning_rate": 4.398e-05, "loss": 0.1471, "step": 2200 }, { "grad_norm": 1.6161178350448608, "learning_rate": 4.418000000000001e-05, "loss": 0.1399, "step": 2210 }, { "grad_norm": 1.4912724494934082, "learning_rate": 4.438e-05, "loss": 0.1321, "step": 2220 }, { "grad_norm": 1.2756309509277344, "learning_rate": 4.458e-05, "loss": 0.1455, "step": 2230 }, { "grad_norm": 1.1420340538024902, "learning_rate": 4.478e-05, "loss": 0.1324, "step": 2240 }, { "grad_norm": 1.1801502704620361, "learning_rate": 4.498e-05, "loss": 0.1535, "step": 2250 }, { "grad_norm": 1.5462982654571533, "learning_rate": 4.518e-05, "loss": 0.1259, "step": 2260 }, { "grad_norm": 1.0895618200302124, "learning_rate": 4.538e-05, "loss": 0.1366, "step": 2270 }, { "grad_norm": 1.3234546184539795, "learning_rate": 4.558e-05, "loss": 0.1483, "step": 2280 }, { "grad_norm": 1.4622288942337036, "learning_rate": 4.578e-05, "loss": 0.1883, "step": 2290 }, { "grad_norm": 1.3276995420455933, "learning_rate": 4.5980000000000004e-05, "loss": 0.1279, "step": 2300 }, { "grad_norm": 1.2544394731521606, "learning_rate": 4.618e-05, "loss": 0.1365, "step": 2310 }, { "grad_norm": 1.7331775426864624, "learning_rate": 4.638e-05, "loss": 0.1389, "step": 2320 }, { "grad_norm": 1.1088860034942627, "learning_rate": 4.6580000000000005e-05, "loss": 0.1622, "step": 2330 }, { "grad_norm": 1.0423532724380493, "learning_rate": 4.678e-05, "loss": 0.14, "step": 2340 }, { "grad_norm": 1.0823379755020142, "learning_rate": 4.698e-05, "loss": 0.1409, "step": 2350 }, { "grad_norm": 0.991664707660675, "learning_rate": 4.718e-05, "loss": 0.146, "step": 2360 }, { "grad_norm": 1.1094270944595337, "learning_rate": 4.7380000000000004e-05, "loss": 0.1457, "step": 2370 }, { "grad_norm": 1.2962968349456787, "learning_rate": 4.758e-05, "loss": 0.1612, "step": 2380 }, { "grad_norm": 1.2376526594161987, "learning_rate": 4.778e-05, "loss": 0.1678, "step": 2390 }, { "grad_norm": 1.2085986137390137, "learning_rate": 4.7980000000000005e-05, "loss": 0.142, "step": 2400 }, { "grad_norm": 1.0770319700241089, "learning_rate": 4.818e-05, "loss": 0.1612, "step": 2410 }, { "grad_norm": 1.1725916862487793, "learning_rate": 4.838e-05, "loss": 0.1988, "step": 2420 }, { "grad_norm": 1.4361047744750977, "learning_rate": 4.8580000000000006e-05, "loss": 0.147, "step": 2430 }, { "grad_norm": 1.5015647411346436, "learning_rate": 4.8780000000000004e-05, "loss": 0.1504, "step": 2440 }, { "grad_norm": 1.3237942457199097, "learning_rate": 4.898e-05, "loss": 0.1256, "step": 2450 }, { "grad_norm": 0.949654221534729, "learning_rate": 4.918000000000001e-05, "loss": 0.1438, "step": 2460 }, { "grad_norm": 1.4039239883422852, "learning_rate": 4.9380000000000005e-05, "loss": 0.1217, "step": 2470 }, { "grad_norm": 1.1710916757583618, "learning_rate": 4.958e-05, "loss": 0.1529, "step": 2480 }, { "grad_norm": 1.2593703269958496, "learning_rate": 4.978e-05, "loss": 0.1342, "step": 2490 }, { "grad_norm": 1.2902456521987915, "learning_rate": 4.9980000000000006e-05, "loss": 0.1348, "step": 2500 }, { "grad_norm": 1.0866817235946655, "learning_rate": 5.0180000000000004e-05, "loss": 0.1252, "step": 2510 }, { "grad_norm": 0.9674057960510254, "learning_rate": 5.038e-05, "loss": 0.1374, "step": 2520 }, { "grad_norm": 1.0656646490097046, "learning_rate": 5.058000000000001e-05, "loss": 0.1199, "step": 2530 }, { "grad_norm": 0.9675047993659973, "learning_rate": 5.0780000000000005e-05, "loss": 0.1439, "step": 2540 }, { "grad_norm": 1.2427277565002441, "learning_rate": 5.098e-05, "loss": 0.1526, "step": 2550 }, { "grad_norm": 0.980385959148407, "learning_rate": 5.118000000000001e-05, "loss": 0.1505, "step": 2560 }, { "grad_norm": 0.9844826459884644, "learning_rate": 5.1380000000000006e-05, "loss": 0.1413, "step": 2570 }, { "grad_norm": 1.1286180019378662, "learning_rate": 5.1580000000000004e-05, "loss": 0.1566, "step": 2580 }, { "grad_norm": 0.7535312175750732, "learning_rate": 5.178000000000001e-05, "loss": 0.13, "step": 2590 }, { "grad_norm": 1.5283774137496948, "learning_rate": 5.198000000000001e-05, "loss": 0.1461, "step": 2600 }, { "grad_norm": 1.0752692222595215, "learning_rate": 5.2180000000000005e-05, "loss": 0.1459, "step": 2610 }, { "grad_norm": 1.6843525171279907, "learning_rate": 5.238000000000001e-05, "loss": 0.1257, "step": 2620 }, { "grad_norm": 1.351961374282837, "learning_rate": 5.258000000000001e-05, "loss": 0.138, "step": 2630 }, { "grad_norm": 1.0353034734725952, "learning_rate": 5.2780000000000006e-05, "loss": 0.1408, "step": 2640 }, { "grad_norm": 0.8502292633056641, "learning_rate": 5.2980000000000004e-05, "loss": 0.1389, "step": 2650 }, { "grad_norm": 1.0640288591384888, "learning_rate": 5.318000000000001e-05, "loss": 0.142, "step": 2660 }, { "grad_norm": 1.1499782800674438, "learning_rate": 5.338000000000001e-05, "loss": 0.1472, "step": 2670 }, { "grad_norm": 1.4564071893692017, "learning_rate": 5.3580000000000005e-05, "loss": 0.1645, "step": 2680 }, { "grad_norm": 0.9028648138046265, "learning_rate": 5.378e-05, "loss": 0.1274, "step": 2690 }, { "grad_norm": 1.2858202457427979, "learning_rate": 5.3979999999999995e-05, "loss": 0.1305, "step": 2700 }, { "grad_norm": 1.0816289186477661, "learning_rate": 5.418e-05, "loss": 0.1466, "step": 2710 }, { "grad_norm": 1.243079662322998, "learning_rate": 5.438e-05, "loss": 0.1603, "step": 2720 }, { "grad_norm": 0.9717589020729065, "learning_rate": 5.4579999999999996e-05, "loss": 0.1173, "step": 2730 }, { "grad_norm": 0.9805070757865906, "learning_rate": 5.478e-05, "loss": 0.124, "step": 2740 }, { "grad_norm": 1.1564139127731323, "learning_rate": 5.498e-05, "loss": 0.1513, "step": 2750 }, { "grad_norm": 1.3506481647491455, "learning_rate": 5.518e-05, "loss": 0.1401, "step": 2760 }, { "grad_norm": 1.388710856437683, "learning_rate": 5.538e-05, "loss": 0.1499, "step": 2770 }, { "grad_norm": 1.191598892211914, "learning_rate": 5.558e-05, "loss": 0.1503, "step": 2780 }, { "grad_norm": 0.9047518968582153, "learning_rate": 5.578e-05, "loss": 0.1474, "step": 2790 }, { "grad_norm": 1.3652938604354858, "learning_rate": 5.5979999999999996e-05, "loss": 0.1295, "step": 2800 }, { "grad_norm": 0.9374330043792725, "learning_rate": 5.618e-05, "loss": 0.1637, "step": 2810 }, { "grad_norm": 1.2410961389541626, "learning_rate": 5.638e-05, "loss": 0.1377, "step": 2820 }, { "grad_norm": 0.8729931712150574, "learning_rate": 5.658e-05, "loss": 0.1269, "step": 2830 }, { "grad_norm": 1.277143120765686, "learning_rate": 5.678e-05, "loss": 0.1506, "step": 2840 }, { "grad_norm": 1.0559576749801636, "learning_rate": 5.698e-05, "loss": 0.1494, "step": 2850 }, { "grad_norm": 1.0649340152740479, "learning_rate": 5.718e-05, "loss": 0.1441, "step": 2860 }, { "grad_norm": 0.8106505870819092, "learning_rate": 5.738e-05, "loss": 0.1388, "step": 2870 }, { "grad_norm": 1.1028915643692017, "learning_rate": 5.758e-05, "loss": 0.1267, "step": 2880 }, { "grad_norm": 0.935939371585846, "learning_rate": 5.778e-05, "loss": 0.1352, "step": 2890 }, { "grad_norm": 1.1589796543121338, "learning_rate": 5.7980000000000004e-05, "loss": 0.152, "step": 2900 }, { "grad_norm": 1.321684718132019, "learning_rate": 5.818e-05, "loss": 0.1272, "step": 2910 }, { "grad_norm": 0.8664406538009644, "learning_rate": 5.838e-05, "loss": 0.1429, "step": 2920 }, { "grad_norm": 1.2181752920150757, "learning_rate": 5.858e-05, "loss": 0.114, "step": 2930 }, { "grad_norm": 1.1425237655639648, "learning_rate": 5.878e-05, "loss": 0.1178, "step": 2940 }, { "grad_norm": 1.1624583005905151, "learning_rate": 5.898e-05, "loss": 0.1487, "step": 2950 }, { "grad_norm": 0.8331860303878784, "learning_rate": 5.918e-05, "loss": 0.1521, "step": 2960 }, { "grad_norm": 0.870564341545105, "learning_rate": 5.9380000000000004e-05, "loss": 0.1151, "step": 2970 }, { "grad_norm": 1.0915206670761108, "learning_rate": 5.958e-05, "loss": 0.1437, "step": 2980 }, { "grad_norm": 0.9842650294303894, "learning_rate": 5.978e-05, "loss": 0.1444, "step": 2990 }, { "grad_norm": 1.2216250896453857, "learning_rate": 5.9980000000000005e-05, "loss": 0.141, "step": 3000 }, { "grad_norm": 1.1114015579223633, "learning_rate": 6.018e-05, "loss": 0.1162, "step": 3010 }, { "grad_norm": 0.9805004000663757, "learning_rate": 6.038e-05, "loss": 0.1239, "step": 3020 }, { "grad_norm": 1.9841896295547485, "learning_rate": 6.0580000000000006e-05, "loss": 0.1221, "step": 3030 }, { "grad_norm": 0.7630296349525452, "learning_rate": 6.0780000000000004e-05, "loss": 0.137, "step": 3040 }, { "grad_norm": 0.9451783299446106, "learning_rate": 6.098e-05, "loss": 0.1461, "step": 3050 }, { "grad_norm": 0.884850263595581, "learning_rate": 6.118000000000001e-05, "loss": 0.1474, "step": 3060 }, { "grad_norm": 0.9190703630447388, "learning_rate": 6.138e-05, "loss": 0.1161, "step": 3070 }, { "grad_norm": 1.238025188446045, "learning_rate": 6.158e-05, "loss": 0.1319, "step": 3080 }, { "grad_norm": 1.0712981224060059, "learning_rate": 6.178000000000001e-05, "loss": 0.1427, "step": 3090 }, { "grad_norm": 1.3445574045181274, "learning_rate": 6.198e-05, "loss": 0.1338, "step": 3100 }, { "grad_norm": 0.8485326766967773, "learning_rate": 6.218e-05, "loss": 0.1125, "step": 3110 }, { "grad_norm": 1.0088602304458618, "learning_rate": 6.238000000000001e-05, "loss": 0.1191, "step": 3120 }, { "grad_norm": 1.1196284294128418, "learning_rate": 6.258e-05, "loss": 0.1284, "step": 3130 }, { "grad_norm": 1.3770872354507446, "learning_rate": 6.278e-05, "loss": 0.1348, "step": 3140 }, { "grad_norm": 1.2955726385116577, "learning_rate": 6.298000000000001e-05, "loss": 0.1354, "step": 3150 }, { "grad_norm": 0.8483238816261292, "learning_rate": 6.318e-05, "loss": 0.1362, "step": 3160 }, { "grad_norm": 0.9859054684638977, "learning_rate": 6.338e-05, "loss": 0.1209, "step": 3170 }, { "grad_norm": 0.7468226552009583, "learning_rate": 6.358000000000001e-05, "loss": 0.1308, "step": 3180 }, { "grad_norm": 1.1448923349380493, "learning_rate": 6.378e-05, "loss": 0.1321, "step": 3190 }, { "grad_norm": 0.6875789165496826, "learning_rate": 6.398000000000001e-05, "loss": 0.1229, "step": 3200 }, { "grad_norm": 0.7320196628570557, "learning_rate": 6.418000000000001e-05, "loss": 0.1094, "step": 3210 }, { "grad_norm": 1.0300829410552979, "learning_rate": 6.438e-05, "loss": 0.1477, "step": 3220 }, { "grad_norm": 0.9341477751731873, "learning_rate": 6.458000000000001e-05, "loss": 0.1419, "step": 3230 }, { "grad_norm": 0.7739288806915283, "learning_rate": 6.478000000000001e-05, "loss": 0.1051, "step": 3240 }, { "grad_norm": 0.7093743681907654, "learning_rate": 6.498e-05, "loss": 0.1443, "step": 3250 }, { "grad_norm": 0.8302675485610962, "learning_rate": 6.518000000000001e-05, "loss": 0.1258, "step": 3260 }, { "grad_norm": 0.9137423634529114, "learning_rate": 6.538000000000001e-05, "loss": 0.1149, "step": 3270 }, { "grad_norm": 0.9745979905128479, "learning_rate": 6.558e-05, "loss": 0.1144, "step": 3280 }, { "grad_norm": 0.7991428375244141, "learning_rate": 6.578000000000001e-05, "loss": 0.1287, "step": 3290 }, { "grad_norm": 1.0037037134170532, "learning_rate": 6.598e-05, "loss": 0.1088, "step": 3300 }, { "grad_norm": 0.9385792016983032, "learning_rate": 6.618e-05, "loss": 0.128, "step": 3310 }, { "grad_norm": 0.8189888000488281, "learning_rate": 6.638e-05, "loss": 0.135, "step": 3320 }, { "grad_norm": 0.9176419973373413, "learning_rate": 6.658e-05, "loss": 0.1128, "step": 3330 }, { "grad_norm": 0.8550751805305481, "learning_rate": 6.678e-05, "loss": 0.1094, "step": 3340 }, { "grad_norm": 0.775535523891449, "learning_rate": 6.698e-05, "loss": 0.112, "step": 3350 }, { "grad_norm": 0.8807398080825806, "learning_rate": 6.718e-05, "loss": 0.1161, "step": 3360 }, { "grad_norm": 2.160212278366089, "learning_rate": 6.738e-05, "loss": 0.1249, "step": 3370 }, { "grad_norm": 0.9305012822151184, "learning_rate": 6.758e-05, "loss": 0.1698, "step": 3380 }, { "grad_norm": 1.0523810386657715, "learning_rate": 6.778e-05, "loss": 0.1318, "step": 3390 }, { "grad_norm": 1.3715146780014038, "learning_rate": 6.798e-05, "loss": 0.1347, "step": 3400 }, { "grad_norm": 0.9640893340110779, "learning_rate": 6.818e-05, "loss": 0.1256, "step": 3410 }, { "grad_norm": 1.0416171550750732, "learning_rate": 6.838e-05, "loss": 0.1239, "step": 3420 }, { "grad_norm": 0.8997697830200195, "learning_rate": 6.858e-05, "loss": 0.1147, "step": 3430 }, { "grad_norm": 0.9816014170646667, "learning_rate": 6.878e-05, "loss": 0.1482, "step": 3440 }, { "grad_norm": 0.9251605868339539, "learning_rate": 6.898e-05, "loss": 0.1209, "step": 3450 }, { "grad_norm": 0.6604093313217163, "learning_rate": 6.918e-05, "loss": 0.107, "step": 3460 }, { "grad_norm": 0.8938011527061462, "learning_rate": 6.938e-05, "loss": 0.135, "step": 3470 }, { "grad_norm": 0.7041993737220764, "learning_rate": 6.958e-05, "loss": 0.1475, "step": 3480 }, { "grad_norm": 0.9836302995681763, "learning_rate": 6.978e-05, "loss": 0.1463, "step": 3490 }, { "grad_norm": 0.9868545532226562, "learning_rate": 6.998e-05, "loss": 0.1297, "step": 3500 }, { "grad_norm": 0.6492855548858643, "learning_rate": 7.018e-05, "loss": 0.1164, "step": 3510 }, { "grad_norm": 1.0043702125549316, "learning_rate": 7.038e-05, "loss": 0.1383, "step": 3520 }, { "grad_norm": 1.2943439483642578, "learning_rate": 7.058e-05, "loss": 0.1345, "step": 3530 }, { "grad_norm": 1.049285888671875, "learning_rate": 7.078e-05, "loss": 0.1183, "step": 3540 }, { "grad_norm": 0.8708542585372925, "learning_rate": 7.098e-05, "loss": 0.1367, "step": 3550 }, { "grad_norm": 1.0009750127792358, "learning_rate": 7.118e-05, "loss": 0.1205, "step": 3560 }, { "grad_norm": 0.7614304423332214, "learning_rate": 7.138e-05, "loss": 0.1441, "step": 3570 }, { "grad_norm": 0.7702234983444214, "learning_rate": 7.158e-05, "loss": 0.1413, "step": 3580 }, { "grad_norm": 0.9194754958152771, "learning_rate": 7.178000000000001e-05, "loss": 0.1445, "step": 3590 }, { "grad_norm": 0.9518100619316101, "learning_rate": 7.198e-05, "loss": 0.1205, "step": 3600 }, { "grad_norm": 1.533398985862732, "learning_rate": 7.218e-05, "loss": 0.1282, "step": 3610 }, { "grad_norm": 0.860560953617096, "learning_rate": 7.238000000000001e-05, "loss": 0.1258, "step": 3620 }, { "grad_norm": 0.8664129376411438, "learning_rate": 7.258e-05, "loss": 0.1127, "step": 3630 }, { "grad_norm": 0.9541471004486084, "learning_rate": 7.278e-05, "loss": 0.1224, "step": 3640 }, { "grad_norm": 1.0082863569259644, "learning_rate": 7.298000000000001e-05, "loss": 0.1422, "step": 3650 }, { "grad_norm": 0.7970848679542542, "learning_rate": 7.318e-05, "loss": 0.1235, "step": 3660 }, { "grad_norm": 1.0716497898101807, "learning_rate": 7.338e-05, "loss": 0.1429, "step": 3670 }, { "grad_norm": 0.8613367080688477, "learning_rate": 7.358000000000001e-05, "loss": 0.1343, "step": 3680 }, { "grad_norm": 1.0303003787994385, "learning_rate": 7.378e-05, "loss": 0.1275, "step": 3690 }, { "grad_norm": 0.8043179512023926, "learning_rate": 7.398e-05, "loss": 0.1087, "step": 3700 }, { "grad_norm": 1.0245437622070312, "learning_rate": 7.418000000000001e-05, "loss": 0.1309, "step": 3710 }, { "grad_norm": 0.8720199465751648, "learning_rate": 7.438e-05, "loss": 0.1214, "step": 3720 }, { "grad_norm": 0.8616819381713867, "learning_rate": 7.458000000000001e-05, "loss": 0.1152, "step": 3730 }, { "grad_norm": 0.7980349063873291, "learning_rate": 7.478e-05, "loss": 0.1098, "step": 3740 }, { "grad_norm": 0.9917236566543579, "learning_rate": 7.498e-05, "loss": 0.1392, "step": 3750 }, { "grad_norm": 0.7163695096969604, "learning_rate": 7.518000000000001e-05, "loss": 0.1087, "step": 3760 }, { "grad_norm": 1.0760713815689087, "learning_rate": 7.538e-05, "loss": 0.1295, "step": 3770 }, { "grad_norm": 0.8520299792289734, "learning_rate": 7.558e-05, "loss": 0.1145, "step": 3780 }, { "grad_norm": 0.9375097155570984, "learning_rate": 7.578000000000001e-05, "loss": 0.1202, "step": 3790 }, { "grad_norm": 0.7375433444976807, "learning_rate": 7.598e-05, "loss": 0.1707, "step": 3800 }, { "grad_norm": 1.1758983135223389, "learning_rate": 7.618e-05, "loss": 0.1313, "step": 3810 }, { "grad_norm": 0.8867660164833069, "learning_rate": 7.638000000000001e-05, "loss": 0.1275, "step": 3820 }, { "grad_norm": 1.0412075519561768, "learning_rate": 7.658e-05, "loss": 0.1361, "step": 3830 }, { "grad_norm": 1.4675334692001343, "learning_rate": 7.678000000000001e-05, "loss": 0.117, "step": 3840 }, { "grad_norm": 1.1493767499923706, "learning_rate": 7.698000000000001e-05, "loss": 0.1639, "step": 3850 }, { "grad_norm": 0.9320268034934998, "learning_rate": 7.718e-05, "loss": 0.1144, "step": 3860 }, { "grad_norm": 0.9365837574005127, "learning_rate": 7.738000000000001e-05, "loss": 0.1254, "step": 3870 }, { "grad_norm": 0.7833153605461121, "learning_rate": 7.758000000000001e-05, "loss": 0.1264, "step": 3880 }, { "grad_norm": 0.9971233010292053, "learning_rate": 7.778e-05, "loss": 0.1243, "step": 3890 }, { "grad_norm": 0.8609254360198975, "learning_rate": 7.798000000000001e-05, "loss": 0.1167, "step": 3900 }, { "grad_norm": 1.219224214553833, "learning_rate": 7.818000000000001e-05, "loss": 0.1347, "step": 3910 }, { "grad_norm": 1.176448106765747, "learning_rate": 7.838e-05, "loss": 0.1387, "step": 3920 }, { "grad_norm": 0.6916607022285461, "learning_rate": 7.858000000000001e-05, "loss": 0.1092, "step": 3930 }, { "grad_norm": 1.070788025856018, "learning_rate": 7.878e-05, "loss": 0.142, "step": 3940 }, { "grad_norm": 1.0266536474227905, "learning_rate": 7.897999999999999e-05, "loss": 0.1181, "step": 3950 }, { "grad_norm": 0.8268404006958008, "learning_rate": 7.918e-05, "loss": 0.1433, "step": 3960 }, { "grad_norm": 0.8643967509269714, "learning_rate": 7.938e-05, "loss": 0.1172, "step": 3970 }, { "grad_norm": 0.7308149933815002, "learning_rate": 7.958e-05, "loss": 0.1439, "step": 3980 }, { "grad_norm": 0.9861112236976624, "learning_rate": 7.978e-05, "loss": 0.1391, "step": 3990 }, { "grad_norm": 0.9780738949775696, "learning_rate": 7.998e-05, "loss": 0.1159, "step": 4000 }, { "grad_norm": 0.823293149471283, "learning_rate": 8.018e-05, "loss": 0.1324, "step": 4010 }, { "grad_norm": 0.6461382508277893, "learning_rate": 8.038e-05, "loss": 0.1199, "step": 4020 }, { "grad_norm": 1.1390928030014038, "learning_rate": 8.058e-05, "loss": 0.1454, "step": 4030 }, { "grad_norm": 1.0122706890106201, "learning_rate": 8.078e-05, "loss": 0.108, "step": 4040 }, { "grad_norm": 0.7654885053634644, "learning_rate": 8.098e-05, "loss": 0.1326, "step": 4050 }, { "grad_norm": 0.9022560119628906, "learning_rate": 8.118e-05, "loss": 0.1303, "step": 4060 }, { "grad_norm": 1.101626992225647, "learning_rate": 8.138e-05, "loss": 0.1221, "step": 4070 }, { "grad_norm": 0.8304359912872314, "learning_rate": 8.158e-05, "loss": 0.1274, "step": 4080 }, { "grad_norm": 0.7685043811798096, "learning_rate": 8.178e-05, "loss": 0.1161, "step": 4090 }, { "grad_norm": 0.9391643404960632, "learning_rate": 8.198e-05, "loss": 0.1285, "step": 4100 }, { "grad_norm": 0.739382266998291, "learning_rate": 8.218e-05, "loss": 0.1259, "step": 4110 }, { "grad_norm": 0.7964490652084351, "learning_rate": 8.238000000000001e-05, "loss": 0.1284, "step": 4120 }, { "grad_norm": 0.7545233964920044, "learning_rate": 8.258e-05, "loss": 0.119, "step": 4130 }, { "grad_norm": 0.8217739462852478, "learning_rate": 8.278e-05, "loss": 0.1236, "step": 4140 }, { "grad_norm": 0.7048052549362183, "learning_rate": 8.298000000000001e-05, "loss": 0.1319, "step": 4150 }, { "grad_norm": 0.7765127420425415, "learning_rate": 8.318e-05, "loss": 0.1024, "step": 4160 }, { "grad_norm": 0.7707551121711731, "learning_rate": 8.338e-05, "loss": 0.1094, "step": 4170 }, { "grad_norm": 0.8357105255126953, "learning_rate": 8.358e-05, "loss": 0.1045, "step": 4180 }, { "grad_norm": 1.2052592039108276, "learning_rate": 8.378e-05, "loss": 0.1224, "step": 4190 }, { "grad_norm": 0.7560522556304932, "learning_rate": 8.398e-05, "loss": 0.1255, "step": 4200 }, { "grad_norm": 0.9129976034164429, "learning_rate": 8.418e-05, "loss": 0.1149, "step": 4210 }, { "grad_norm": 0.9543116688728333, "learning_rate": 8.438e-05, "loss": 0.1314, "step": 4220 }, { "grad_norm": 0.8149082660675049, "learning_rate": 8.458e-05, "loss": 0.1397, "step": 4230 }, { "grad_norm": 0.7979441285133362, "learning_rate": 8.478e-05, "loss": 0.1282, "step": 4240 }, { "grad_norm": 1.220233678817749, "learning_rate": 8.498e-05, "loss": 0.1375, "step": 4250 }, { "grad_norm": 0.7171669602394104, "learning_rate": 8.518000000000001e-05, "loss": 0.1339, "step": 4260 }, { "grad_norm": 0.6902512907981873, "learning_rate": 8.538e-05, "loss": 0.133, "step": 4270 }, { "grad_norm": 0.8260260820388794, "learning_rate": 8.558e-05, "loss": 0.1072, "step": 4280 }, { "grad_norm": 0.850919783115387, "learning_rate": 8.578000000000001e-05, "loss": 0.1254, "step": 4290 }, { "grad_norm": 0.629992663860321, "learning_rate": 8.598e-05, "loss": 0.1086, "step": 4300 }, { "grad_norm": 1.0036122798919678, "learning_rate": 8.618e-05, "loss": 0.144, "step": 4310 }, { "grad_norm": 1.217720866203308, "learning_rate": 8.638000000000001e-05, "loss": 0.1329, "step": 4320 }, { "grad_norm": 1.2115391492843628, "learning_rate": 8.658e-05, "loss": 0.1291, "step": 4330 }, { "grad_norm": 1.0515056848526, "learning_rate": 8.678e-05, "loss": 0.1001, "step": 4340 }, { "grad_norm": 1.2596944570541382, "learning_rate": 8.698000000000001e-05, "loss": 0.1193, "step": 4350 }, { "grad_norm": 0.8378982543945312, "learning_rate": 8.718e-05, "loss": 0.1185, "step": 4360 }, { "grad_norm": 0.8526588678359985, "learning_rate": 8.738000000000001e-05, "loss": 0.1504, "step": 4370 }, { "grad_norm": 0.6755622029304504, "learning_rate": 8.758000000000001e-05, "loss": 0.0982, "step": 4380 }, { "grad_norm": 0.8652952313423157, "learning_rate": 8.778e-05, "loss": 0.0998, "step": 4390 }, { "grad_norm": 0.9403833150863647, "learning_rate": 8.798000000000001e-05, "loss": 0.1331, "step": 4400 }, { "grad_norm": 0.5576606392860413, "learning_rate": 8.818000000000001e-05, "loss": 0.1207, "step": 4410 }, { "grad_norm": 0.6226751208305359, "learning_rate": 8.838e-05, "loss": 0.1322, "step": 4420 }, { "grad_norm": 1.0286847352981567, "learning_rate": 8.858000000000001e-05, "loss": 0.1198, "step": 4430 }, { "grad_norm": 0.5896373391151428, "learning_rate": 8.878000000000001e-05, "loss": 0.1181, "step": 4440 }, { "grad_norm": 0.68971848487854, "learning_rate": 8.898e-05, "loss": 0.1457, "step": 4450 }, { "grad_norm": 0.7103783488273621, "learning_rate": 8.918000000000001e-05, "loss": 0.1226, "step": 4460 }, { "grad_norm": 0.6702210307121277, "learning_rate": 8.938e-05, "loss": 0.103, "step": 4470 }, { "grad_norm": 0.6965193748474121, "learning_rate": 8.958e-05, "loss": 0.148, "step": 4480 }, { "grad_norm": 0.7145932912826538, "learning_rate": 8.978000000000001e-05, "loss": 0.1054, "step": 4490 }, { "grad_norm": 0.7475468516349792, "learning_rate": 8.998e-05, "loss": 0.1229, "step": 4500 }, { "grad_norm": 0.7930594682693481, "learning_rate": 9.018000000000001e-05, "loss": 0.147, "step": 4510 }, { "grad_norm": 0.8680574893951416, "learning_rate": 9.038000000000001e-05, "loss": 0.1264, "step": 4520 }, { "grad_norm": 0.7393587231636047, "learning_rate": 9.058e-05, "loss": 0.1017, "step": 4530 }, { "grad_norm": 0.8909984230995178, "learning_rate": 9.078000000000001e-05, "loss": 0.1301, "step": 4540 }, { "grad_norm": 0.7645158171653748, "learning_rate": 9.098000000000001e-05, "loss": 0.1438, "step": 4550 }, { "grad_norm": 0.7856624126434326, "learning_rate": 9.118e-05, "loss": 0.1131, "step": 4560 }, { "grad_norm": 0.8264366984367371, "learning_rate": 9.138e-05, "loss": 0.1431, "step": 4570 }, { "grad_norm": 1.5451103448867798, "learning_rate": 9.158e-05, "loss": 0.1434, "step": 4580 }, { "grad_norm": 1.0954893827438354, "learning_rate": 9.178e-05, "loss": 0.1232, "step": 4590 }, { "grad_norm": 0.7724431753158569, "learning_rate": 9.198e-05, "loss": 0.1263, "step": 4600 }, { "grad_norm": 0.902966320514679, "learning_rate": 9.218e-05, "loss": 0.1422, "step": 4610 }, { "grad_norm": 0.8479042053222656, "learning_rate": 9.238e-05, "loss": 0.1287, "step": 4620 }, { "grad_norm": 1.2860639095306396, "learning_rate": 9.258e-05, "loss": 0.1405, "step": 4630 }, { "grad_norm": 0.9370495676994324, "learning_rate": 9.278e-05, "loss": 0.147, "step": 4640 }, { "grad_norm": 0.8120957612991333, "learning_rate": 9.298e-05, "loss": 0.1205, "step": 4650 }, { "grad_norm": 0.8148022890090942, "learning_rate": 9.318e-05, "loss": 0.1167, "step": 4660 }, { "grad_norm": 0.9422168731689453, "learning_rate": 9.338e-05, "loss": 0.126, "step": 4670 }, { "grad_norm": 0.8738299012184143, "learning_rate": 9.358e-05, "loss": 0.1227, "step": 4680 }, { "grad_norm": 0.9443623423576355, "learning_rate": 9.378e-05, "loss": 0.1511, "step": 4690 }, { "grad_norm": 0.827981173992157, "learning_rate": 9.398e-05, "loss": 0.1217, "step": 4700 }, { "grad_norm": 0.771556556224823, "learning_rate": 9.418e-05, "loss": 0.1208, "step": 4710 }, { "grad_norm": 0.8181405663490295, "learning_rate": 9.438e-05, "loss": 0.1131, "step": 4720 }, { "grad_norm": 0.7427645921707153, "learning_rate": 9.458e-05, "loss": 0.1407, "step": 4730 }, { "grad_norm": 0.8986298441886902, "learning_rate": 9.478e-05, "loss": 0.1454, "step": 4740 }, { "grad_norm": 0.7939658164978027, "learning_rate": 9.498e-05, "loss": 0.1172, "step": 4750 }, { "grad_norm": 0.7588373422622681, "learning_rate": 9.518000000000001e-05, "loss": 0.1294, "step": 4760 }, { "grad_norm": 0.7586398124694824, "learning_rate": 9.538e-05, "loss": 0.1234, "step": 4770 }, { "grad_norm": 0.7255469560623169, "learning_rate": 9.558e-05, "loss": 0.1372, "step": 4780 }, { "grad_norm": 0.8325705528259277, "learning_rate": 9.578000000000001e-05, "loss": 0.1078, "step": 4790 }, { "grad_norm": 0.7984139919281006, "learning_rate": 9.598e-05, "loss": 0.1201, "step": 4800 }, { "grad_norm": 0.8972042202949524, "learning_rate": 9.618e-05, "loss": 0.1123, "step": 4810 }, { "grad_norm": 0.796929657459259, "learning_rate": 9.638000000000001e-05, "loss": 0.1166, "step": 4820 }, { "grad_norm": 0.6709780097007751, "learning_rate": 9.658e-05, "loss": 0.1229, "step": 4830 }, { "grad_norm": 0.835342288017273, "learning_rate": 9.678e-05, "loss": 0.1244, "step": 4840 }, { "grad_norm": 0.925001859664917, "learning_rate": 9.698000000000001e-05, "loss": 0.121, "step": 4850 }, { "grad_norm": 0.8269108533859253, "learning_rate": 9.718e-05, "loss": 0.1033, "step": 4860 }, { "grad_norm": 0.8980498909950256, "learning_rate": 9.738e-05, "loss": 0.1133, "step": 4870 }, { "grad_norm": 0.8397391438484192, "learning_rate": 9.758000000000001e-05, "loss": 0.1291, "step": 4880 }, { "grad_norm": 0.9586042761802673, "learning_rate": 9.778e-05, "loss": 0.1438, "step": 4890 }, { "grad_norm": 1.0461490154266357, "learning_rate": 9.798000000000001e-05, "loss": 0.1275, "step": 4900 }, { "grad_norm": 0.7201167941093445, "learning_rate": 9.818000000000001e-05, "loss": 0.1069, "step": 4910 }, { "grad_norm": 0.9552872776985168, "learning_rate": 9.838e-05, "loss": 0.1175, "step": 4920 }, { "grad_norm": 0.7633842825889587, "learning_rate": 9.858000000000001e-05, "loss": 0.1252, "step": 4930 }, { "grad_norm": 0.6280658841133118, "learning_rate": 9.878e-05, "loss": 0.1194, "step": 4940 }, { "grad_norm": 0.7116129398345947, "learning_rate": 9.898e-05, "loss": 0.1309, "step": 4950 }, { "grad_norm": 0.6978970170021057, "learning_rate": 9.918000000000001e-05, "loss": 0.129, "step": 4960 }, { "grad_norm": 0.7490450143814087, "learning_rate": 9.938e-05, "loss": 0.1403, "step": 4970 }, { "grad_norm": 0.6721053123474121, "learning_rate": 9.958e-05, "loss": 0.1306, "step": 4980 }, { "grad_norm": 0.7695106267929077, "learning_rate": 9.978000000000001e-05, "loss": 0.1158, "step": 4990 }, { "grad_norm": 0.6972319483757019, "learning_rate": 9.998e-05, "loss": 0.1389, "step": 5000 }, { "grad_norm": 0.8363659381866455, "learning_rate": 9.999999778549045e-05, "loss": 0.1342, "step": 5010 }, { "grad_norm": 0.8851209878921509, "learning_rate": 9.999999013039593e-05, "loss": 0.1242, "step": 5020 }, { "grad_norm": 0.7803016901016235, "learning_rate": 9.999997700737766e-05, "loss": 0.135, "step": 5030 }, { "grad_norm": 0.7532033324241638, "learning_rate": 9.999995841643709e-05, "loss": 0.108, "step": 5040 }, { "grad_norm": 0.8297553658485413, "learning_rate": 9.999993435757623e-05, "loss": 0.1329, "step": 5050 }, { "grad_norm": 0.8349153399467468, "learning_rate": 9.999990483079773e-05, "loss": 0.1109, "step": 5060 }, { "grad_norm": 0.7107409238815308, "learning_rate": 9.999986983610481e-05, "loss": 0.1437, "step": 5070 }, { "grad_norm": 0.890874445438385, "learning_rate": 9.99998293735013e-05, "loss": 0.1361, "step": 5080 }, { "grad_norm": 0.8616982698440552, "learning_rate": 9.999978344299161e-05, "loss": 0.1224, "step": 5090 }, { "grad_norm": 1.0104255676269531, "learning_rate": 9.99997320445808e-05, "loss": 0.119, "step": 5100 }, { "grad_norm": 0.8267363905906677, "learning_rate": 9.999967517827444e-05, "loss": 0.1233, "step": 5110 }, { "grad_norm": 0.8050004243850708, "learning_rate": 9.999961284407879e-05, "loss": 0.1358, "step": 5120 }, { "grad_norm": 0.7792283892631531, "learning_rate": 9.999954504200067e-05, "loss": 0.1212, "step": 5130 }, { "grad_norm": 0.7420607805252075, "learning_rate": 9.999947177204744e-05, "loss": 0.1163, "step": 5140 }, { "grad_norm": 0.7124908566474915, "learning_rate": 9.999939303422718e-05, "loss": 0.096, "step": 5150 }, { "grad_norm": 0.6852322220802307, "learning_rate": 9.999930882854847e-05, "loss": 0.1228, "step": 5160 }, { "grad_norm": 0.8853226900100708, "learning_rate": 9.999921915502051e-05, "loss": 0.1247, "step": 5170 }, { "grad_norm": 0.7563617825508118, "learning_rate": 9.99991240136531e-05, "loss": 0.1265, "step": 5180 }, { "grad_norm": 0.7483181953430176, "learning_rate": 9.999902340445668e-05, "loss": 0.1445, "step": 5190 }, { "grad_norm": 0.6392751336097717, "learning_rate": 9.999891732744224e-05, "loss": 0.1148, "step": 5200 }, { "grad_norm": 0.939030110836029, "learning_rate": 9.999880578262135e-05, "loss": 0.1182, "step": 5210 }, { "grad_norm": 1.0182011127471924, "learning_rate": 9.999868877000624e-05, "loss": 0.1143, "step": 5220 }, { "grad_norm": 0.6234009861946106, "learning_rate": 9.99985662896097e-05, "loss": 0.1212, "step": 5230 }, { "grad_norm": 0.6598928570747375, "learning_rate": 9.999843834144513e-05, "loss": 0.1092, "step": 5240 }, { "grad_norm": 1.0272787809371948, "learning_rate": 9.99983049255265e-05, "loss": 0.1357, "step": 5250 }, { "grad_norm": 0.7363941073417664, "learning_rate": 9.999816604186843e-05, "loss": 0.1194, "step": 5260 }, { "grad_norm": 0.6822314262390137, "learning_rate": 9.999802169048609e-05, "loss": 0.1156, "step": 5270 }, { "grad_norm": 0.7433772683143616, "learning_rate": 9.999787187139527e-05, "loss": 0.1218, "step": 5280 }, { "grad_norm": 0.9195886850357056, "learning_rate": 9.999771658461234e-05, "loss": 0.1181, "step": 5290 }, { "grad_norm": 1.0123406648635864, "learning_rate": 9.999755583015431e-05, "loss": 0.1121, "step": 5300 }, { "grad_norm": 1.0442627668380737, "learning_rate": 9.999738960803874e-05, "loss": 0.1246, "step": 5310 }, { "grad_norm": 0.8141539692878723, "learning_rate": 9.99972179182838e-05, "loss": 0.1316, "step": 5320 }, { "grad_norm": 0.9458717107772827, "learning_rate": 9.99970407609083e-05, "loss": 0.1421, "step": 5330 }, { "grad_norm": 0.9512346982955933, "learning_rate": 9.999685813593159e-05, "loss": 0.1247, "step": 5340 }, { "grad_norm": 1.0090619325637817, "learning_rate": 9.999667004337362e-05, "loss": 0.1175, "step": 5350 }, { "grad_norm": 0.8414285778999329, "learning_rate": 9.9996476483255e-05, "loss": 0.1492, "step": 5360 }, { "grad_norm": 0.6911973357200623, "learning_rate": 9.999627745559688e-05, "loss": 0.1046, "step": 5370 }, { "grad_norm": 0.7900015711784363, "learning_rate": 9.999607296042101e-05, "loss": 0.103, "step": 5380 }, { "grad_norm": 0.8508177399635315, "learning_rate": 9.99958629977498e-05, "loss": 0.112, "step": 5390 }, { "grad_norm": 0.7729924321174622, "learning_rate": 9.999564756760615e-05, "loss": 0.1036, "step": 5400 }, { "grad_norm": 0.4753272235393524, "learning_rate": 9.999542667001366e-05, "loss": 0.0964, "step": 5410 }, { "grad_norm": 0.5539396405220032, "learning_rate": 9.999520030499647e-05, "loss": 0.113, "step": 5420 }, { "grad_norm": 1.1467974185943604, "learning_rate": 9.999496847257936e-05, "loss": 0.1036, "step": 5430 }, { "grad_norm": 0.7554905414581299, "learning_rate": 9.999473117278764e-05, "loss": 0.1317, "step": 5440 }, { "grad_norm": 0.6883156299591064, "learning_rate": 9.999448840564731e-05, "loss": 0.1003, "step": 5450 }, { "grad_norm": 0.6659364700317383, "learning_rate": 9.999424017118488e-05, "loss": 0.0981, "step": 5460 }, { "grad_norm": 0.6483915448188782, "learning_rate": 9.999398646942751e-05, "loss": 0.0979, "step": 5470 }, { "grad_norm": 0.6754804849624634, "learning_rate": 9.999372730040296e-05, "loss": 0.0987, "step": 5480 }, { "grad_norm": 0.9053442478179932, "learning_rate": 9.999346266413953e-05, "loss": 0.1179, "step": 5490 }, { "grad_norm": 0.6695641875267029, "learning_rate": 9.99931925606662e-05, "loss": 0.1198, "step": 5500 }, { "grad_norm": 0.8704274296760559, "learning_rate": 9.99929169900125e-05, "loss": 0.1344, "step": 5510 }, { "grad_norm": 0.795042872428894, "learning_rate": 9.999263595220855e-05, "loss": 0.0999, "step": 5520 }, { "grad_norm": 0.8942593336105347, "learning_rate": 9.99923494472851e-05, "loss": 0.116, "step": 5530 }, { "grad_norm": 0.8035354018211365, "learning_rate": 9.999205747527348e-05, "loss": 0.0938, "step": 5540 }, { "grad_norm": 0.8899586796760559, "learning_rate": 9.999176003620561e-05, "loss": 0.1041, "step": 5550 }, { "grad_norm": 1.0396801233291626, "learning_rate": 9.999145713011405e-05, "loss": 0.1264, "step": 5560 }, { "grad_norm": 1.0308122634887695, "learning_rate": 9.999114875703186e-05, "loss": 0.1225, "step": 5570 }, { "grad_norm": 1.0747536420822144, "learning_rate": 9.999083491699281e-05, "loss": 0.1436, "step": 5580 }, { "grad_norm": 0.7227081060409546, "learning_rate": 9.999051561003123e-05, "loss": 0.1295, "step": 5590 }, { "grad_norm": 0.7674605250358582, "learning_rate": 9.999019083618202e-05, "loss": 0.1285, "step": 5600 }, { "grad_norm": 0.702387809753418, "learning_rate": 9.99898605954807e-05, "loss": 0.1205, "step": 5610 }, { "grad_norm": 0.9235619902610779, "learning_rate": 9.998952488796338e-05, "loss": 0.1191, "step": 5620 }, { "grad_norm": 0.9915019273757935, "learning_rate": 9.998918371366676e-05, "loss": 0.1206, "step": 5630 }, { "grad_norm": 0.8473740816116333, "learning_rate": 9.99888370726282e-05, "loss": 0.1202, "step": 5640 }, { "grad_norm": 0.6238826513290405, "learning_rate": 9.998848496488556e-05, "loss": 0.1396, "step": 5650 }, { "grad_norm": 0.8200995326042175, "learning_rate": 9.998812739047736e-05, "loss": 0.116, "step": 5660 }, { "grad_norm": 0.7857980132102966, "learning_rate": 9.99877643494427e-05, "loss": 0.1228, "step": 5670 }, { "grad_norm": 0.7634982466697693, "learning_rate": 9.998739584182128e-05, "loss": 0.1198, "step": 5680 }, { "grad_norm": 0.9117321968078613, "learning_rate": 9.998702186765342e-05, "loss": 0.0989, "step": 5690 }, { "grad_norm": 0.6643564701080322, "learning_rate": 9.998664242698e-05, "loss": 0.109, "step": 5700 }, { "grad_norm": 0.9780733585357666, "learning_rate": 9.998625751984251e-05, "loss": 0.1285, "step": 5710 }, { "grad_norm": 0.6697921752929688, "learning_rate": 9.998586714628307e-05, "loss": 0.1039, "step": 5720 }, { "grad_norm": 0.7188313007354736, "learning_rate": 9.998547130634432e-05, "loss": 0.1213, "step": 5730 }, { "grad_norm": 0.7446991205215454, "learning_rate": 9.99850700000696e-05, "loss": 0.1296, "step": 5740 }, { "grad_norm": 0.922625720500946, "learning_rate": 9.998466322750278e-05, "loss": 0.1173, "step": 5750 }, { "grad_norm": 0.411418080329895, "learning_rate": 9.998425098868834e-05, "loss": 0.1069, "step": 5760 }, { "grad_norm": 0.8589595556259155, "learning_rate": 9.998383328367136e-05, "loss": 0.1277, "step": 5770 }, { "grad_norm": 1.010628342628479, "learning_rate": 9.99834101124975e-05, "loss": 0.132, "step": 5780 }, { "grad_norm": 0.7291847467422485, "learning_rate": 9.998298147521309e-05, "loss": 0.12, "step": 5790 }, { "grad_norm": 0.8249960541725159, "learning_rate": 9.998254737186496e-05, "loss": 0.1321, "step": 5800 }, { "grad_norm": 1.5174444913864136, "learning_rate": 9.99821078025006e-05, "loss": 0.122, "step": 5810 }, { "grad_norm": 0.7573039531707764, "learning_rate": 9.998166276716807e-05, "loss": 0.1231, "step": 5820 }, { "grad_norm": 0.6907941699028015, "learning_rate": 9.998121226591606e-05, "loss": 0.1376, "step": 5830 }, { "grad_norm": 0.8205470442771912, "learning_rate": 9.998075629879382e-05, "loss": 0.122, "step": 5840 }, { "grad_norm": 0.8619305491447449, "learning_rate": 9.99802948658512e-05, "loss": 0.1104, "step": 5850 }, { "grad_norm": 0.5937711596488953, "learning_rate": 9.99798279671387e-05, "loss": 0.1215, "step": 5860 }, { "grad_norm": 0.6072511076927185, "learning_rate": 9.997935560270734e-05, "loss": 0.1012, "step": 5870 }, { "grad_norm": 0.6494302153587341, "learning_rate": 9.997887777260879e-05, "loss": 0.1141, "step": 5880 }, { "grad_norm": 0.5850493907928467, "learning_rate": 9.997839447689532e-05, "loss": 0.1194, "step": 5890 }, { "grad_norm": 1.0442520380020142, "learning_rate": 9.997790571561978e-05, "loss": 0.1369, "step": 5900 }, { "grad_norm": 0.6037907600402832, "learning_rate": 9.99774114888356e-05, "loss": 0.1392, "step": 5910 }, { "grad_norm": 0.796351432800293, "learning_rate": 9.997691179659684e-05, "loss": 0.1266, "step": 5920 }, { "grad_norm": 0.6026914119720459, "learning_rate": 9.997640663895815e-05, "loss": 0.1086, "step": 5930 }, { "grad_norm": 0.7571582198143005, "learning_rate": 9.997589601597477e-05, "loss": 0.1279, "step": 5940 }, { "grad_norm": 0.6309943795204163, "learning_rate": 9.997537992770252e-05, "loss": 0.1017, "step": 5950 }, { "grad_norm": 1.0721940994262695, "learning_rate": 9.997485837419788e-05, "loss": 0.1128, "step": 5960 }, { "grad_norm": 0.8875972032546997, "learning_rate": 9.997433135551786e-05, "loss": 0.1523, "step": 5970 }, { "grad_norm": 0.8785742521286011, "learning_rate": 9.997379887172009e-05, "loss": 0.1133, "step": 5980 }, { "grad_norm": 0.7449256777763367, "learning_rate": 9.997326092286281e-05, "loss": 0.1326, "step": 5990 }, { "grad_norm": 0.7143782377243042, "learning_rate": 9.997271750900486e-05, "loss": 0.1421, "step": 6000 }, { "grad_norm": 0.7353768944740295, "learning_rate": 9.997216863020565e-05, "loss": 0.1279, "step": 6010 }, { "grad_norm": 0.6405795216560364, "learning_rate": 9.99716142865252e-05, "loss": 0.1194, "step": 6020 }, { "grad_norm": 0.8943187594413757, "learning_rate": 9.997105447802415e-05, "loss": 0.12, "step": 6030 }, { "grad_norm": 0.9031051993370056, "learning_rate": 9.997048920476373e-05, "loss": 0.1009, "step": 6040 }, { "grad_norm": 0.7737523317337036, "learning_rate": 9.996991846680572e-05, "loss": 0.1197, "step": 6050 }, { "grad_norm": 0.9304923415184021, "learning_rate": 9.996934226421257e-05, "loss": 0.1112, "step": 6060 }, { "grad_norm": 0.6920255422592163, "learning_rate": 9.996876059704726e-05, "loss": 0.1135, "step": 6070 }, { "grad_norm": 0.6120618581771851, "learning_rate": 9.996817346537343e-05, "loss": 0.1192, "step": 6080 }, { "grad_norm": 0.6717451810836792, "learning_rate": 9.996758086925526e-05, "loss": 0.116, "step": 6090 }, { "grad_norm": 0.6712605357170105, "learning_rate": 9.996698280875759e-05, "loss": 0.1233, "step": 6100 }, { "grad_norm": 0.5125076770782471, "learning_rate": 9.99663792839458e-05, "loss": 0.0979, "step": 6110 }, { "grad_norm": 1.2290927171707153, "learning_rate": 9.99657702948859e-05, "loss": 0.1212, "step": 6120 }, { "grad_norm": 0.5684456825256348, "learning_rate": 9.996515584164448e-05, "loss": 0.0966, "step": 6130 }, { "grad_norm": 0.7110381126403809, "learning_rate": 9.996453592428873e-05, "loss": 0.1155, "step": 6140 }, { "grad_norm": 0.6027089953422546, "learning_rate": 9.996391054288646e-05, "loss": 0.136, "step": 6150 }, { "grad_norm": 0.7586772441864014, "learning_rate": 9.996327969750605e-05, "loss": 0.1277, "step": 6160 }, { "grad_norm": 0.7138927578926086, "learning_rate": 9.996264338821649e-05, "loss": 0.1084, "step": 6170 }, { "grad_norm": 1.0606589317321777, "learning_rate": 9.996200161508735e-05, "loss": 0.0915, "step": 6180 }, { "grad_norm": 0.8069626688957214, "learning_rate": 9.996135437818885e-05, "loss": 0.1193, "step": 6190 }, { "grad_norm": 0.5458461046218872, "learning_rate": 9.996070167759175e-05, "loss": 0.1013, "step": 6200 }, { "grad_norm": 0.7652369141578674, "learning_rate": 9.996004351336743e-05, "loss": 0.1224, "step": 6210 }, { "grad_norm": 0.9108923673629761, "learning_rate": 9.995937988558785e-05, "loss": 0.1102, "step": 6220 }, { "grad_norm": 0.7163881659507751, "learning_rate": 9.995871079432561e-05, "loss": 0.1296, "step": 6230 }, { "grad_norm": 0.7603222131729126, "learning_rate": 9.995803623965389e-05, "loss": 0.1224, "step": 6240 }, { "grad_norm": 0.6208503246307373, "learning_rate": 9.995735622164641e-05, "loss": 0.1158, "step": 6250 }, { "grad_norm": 1.3626817464828491, "learning_rate": 9.995667074037758e-05, "loss": 0.0863, "step": 6260 }, { "grad_norm": 0.9098769426345825, "learning_rate": 9.995597979592232e-05, "loss": 0.1191, "step": 6270 }, { "grad_norm": 0.7792156338691711, "learning_rate": 9.995528338835625e-05, "loss": 0.1047, "step": 6280 }, { "grad_norm": 0.6778807044029236, "learning_rate": 9.995458151775547e-05, "loss": 0.0926, "step": 6290 }, { "grad_norm": 0.4837140440940857, "learning_rate": 9.995387418419677e-05, "loss": 0.105, "step": 6300 }, { "grad_norm": 0.5456592440605164, "learning_rate": 9.99531613877575e-05, "loss": 0.0928, "step": 6310 }, { "grad_norm": 0.6705403923988342, "learning_rate": 9.995244312851559e-05, "loss": 0.1067, "step": 6320 }, { "grad_norm": 1.080539584159851, "learning_rate": 9.995171940654961e-05, "loss": 0.1322, "step": 6330 }, { "grad_norm": 0.7964569330215454, "learning_rate": 9.995099022193871e-05, "loss": 0.1178, "step": 6340 }, { "grad_norm": 0.9378397464752197, "learning_rate": 9.995025557476261e-05, "loss": 0.1002, "step": 6350 }, { "grad_norm": 0.5972092747688293, "learning_rate": 9.994951546510165e-05, "loss": 0.1, "step": 6360 }, { "grad_norm": 0.5181547403335571, "learning_rate": 9.994876989303679e-05, "loss": 0.103, "step": 6370 }, { "grad_norm": 0.6908807754516602, "learning_rate": 9.994801885864955e-05, "loss": 0.1005, "step": 6380 }, { "grad_norm": 1.2713327407836914, "learning_rate": 9.994726236202205e-05, "loss": 0.1105, "step": 6390 }, { "grad_norm": 0.6039512157440186, "learning_rate": 9.994650040323704e-05, "loss": 0.1238, "step": 6400 }, { "grad_norm": 0.5722615122795105, "learning_rate": 9.994573298237784e-05, "loss": 0.0861, "step": 6410 }, { "grad_norm": 0.629410982131958, "learning_rate": 9.994496009952837e-05, "loss": 0.0994, "step": 6420 }, { "grad_norm": 0.8338832259178162, "learning_rate": 9.994418175477316e-05, "loss": 0.1246, "step": 6430 }, { "grad_norm": 0.8082565665245056, "learning_rate": 9.994339794819733e-05, "loss": 0.112, "step": 6440 }, { "grad_norm": 0.6044453978538513, "learning_rate": 9.994260867988658e-05, "loss": 0.1056, "step": 6450 }, { "grad_norm": 0.5027814507484436, "learning_rate": 9.994181394992723e-05, "loss": 0.1055, "step": 6460 }, { "grad_norm": 0.6488783955574036, "learning_rate": 9.994101375840618e-05, "loss": 0.0979, "step": 6470 }, { "grad_norm": 0.7173369526863098, "learning_rate": 9.994020810541098e-05, "loss": 0.0943, "step": 6480 }, { "grad_norm": 0.8103709816932678, "learning_rate": 9.99393969910297e-05, "loss": 0.1283, "step": 6490 }, { "grad_norm": 1.494645595550537, "learning_rate": 9.993858041535104e-05, "loss": 0.1005, "step": 6500 }, { "grad_norm": 0.6495344042778015, "learning_rate": 9.99377583784643e-05, "loss": 0.0982, "step": 6510 }, { "grad_norm": 1.2393442392349243, "learning_rate": 9.993693088045939e-05, "loss": 0.1149, "step": 6520 }, { "grad_norm": 0.7555108070373535, "learning_rate": 9.99360979214268e-05, "loss": 0.097, "step": 6530 }, { "grad_norm": 0.49629494547843933, "learning_rate": 9.99352595014576e-05, "loss": 0.0974, "step": 6540 }, { "grad_norm": 1.1375325918197632, "learning_rate": 9.993441562064354e-05, "loss": 0.1058, "step": 6550 }, { "grad_norm": 0.7465124130249023, "learning_rate": 9.993356627907685e-05, "loss": 0.1007, "step": 6560 }, { "grad_norm": 0.6315445303916931, "learning_rate": 9.99327114768504e-05, "loss": 0.1044, "step": 6570 }, { "grad_norm": 0.6667795777320862, "learning_rate": 9.99318512140577e-05, "loss": 0.086, "step": 6580 }, { "grad_norm": 0.7186805009841919, "learning_rate": 9.993098549079284e-05, "loss": 0.1261, "step": 6590 }, { "grad_norm": 0.6143324971199036, "learning_rate": 9.993011430715047e-05, "loss": 0.1043, "step": 6600 }, { "grad_norm": 0.6278403401374817, "learning_rate": 9.992923766322586e-05, "loss": 0.0991, "step": 6610 }, { "grad_norm": 0.6670648455619812, "learning_rate": 9.99283555591149e-05, "loss": 0.0777, "step": 6620 }, { "grad_norm": 0.53883296251297, "learning_rate": 9.992746799491404e-05, "loss": 0.1045, "step": 6630 }, { "grad_norm": 0.7300519347190857, "learning_rate": 9.992657497072033e-05, "loss": 0.0988, "step": 6640 }, { "grad_norm": 0.6738671064376831, "learning_rate": 9.992567648663147e-05, "loss": 0.1284, "step": 6650 }, { "grad_norm": 0.7095256447792053, "learning_rate": 9.992477254274568e-05, "loss": 0.1049, "step": 6660 }, { "grad_norm": 0.7311720252037048, "learning_rate": 9.992386313916183e-05, "loss": 0.1001, "step": 6670 }, { "grad_norm": 0.6989219784736633, "learning_rate": 9.992294827597934e-05, "loss": 0.1245, "step": 6680 }, { "grad_norm": 0.5730754733085632, "learning_rate": 9.992202795329831e-05, "loss": 0.0969, "step": 6690 }, { "grad_norm": 0.8314914107322693, "learning_rate": 9.992110217121936e-05, "loss": 0.0876, "step": 6700 }, { "grad_norm": 0.7627439498901367, "learning_rate": 9.992017092984372e-05, "loss": 0.1213, "step": 6710 }, { "grad_norm": 0.5281804203987122, "learning_rate": 9.991923422927326e-05, "loss": 0.1155, "step": 6720 }, { "grad_norm": 0.6900732517242432, "learning_rate": 9.991829206961037e-05, "loss": 0.1126, "step": 6730 }, { "grad_norm": 0.609588086605072, "learning_rate": 9.991734445095813e-05, "loss": 0.1342, "step": 6740 }, { "grad_norm": 0.8451758027076721, "learning_rate": 9.991639137342015e-05, "loss": 0.1019, "step": 6750 }, { "grad_norm": 0.7857773303985596, "learning_rate": 9.991543283710064e-05, "loss": 0.0994, "step": 6760 }, { "grad_norm": 0.5673232674598694, "learning_rate": 9.991446884210445e-05, "loss": 0.1082, "step": 6770 }, { "grad_norm": 0.6868008375167847, "learning_rate": 9.9913499388537e-05, "loss": 0.1193, "step": 6780 }, { "grad_norm": 0.45741042494773865, "learning_rate": 9.99125244765043e-05, "loss": 0.0909, "step": 6790 }, { "grad_norm": 0.6515671014785767, "learning_rate": 9.991154410611296e-05, "loss": 0.1221, "step": 6800 }, { "grad_norm": 0.6324724555015564, "learning_rate": 9.99105582774702e-05, "loss": 0.0923, "step": 6810 }, { "grad_norm": 0.9466918706893921, "learning_rate": 9.990956699068384e-05, "loss": 0.123, "step": 6820 }, { "grad_norm": 0.6079122424125671, "learning_rate": 9.990857024586224e-05, "loss": 0.0961, "step": 6830 }, { "grad_norm": 0.7096017003059387, "learning_rate": 9.990756804311446e-05, "loss": 0.1076, "step": 6840 }, { "grad_norm": 0.5982638001441956, "learning_rate": 9.990656038255006e-05, "loss": 0.1153, "step": 6850 }, { "grad_norm": 0.5981314778327942, "learning_rate": 9.990554726427926e-05, "loss": 0.0945, "step": 6860 }, { "grad_norm": 0.6856759190559387, "learning_rate": 9.990452868841284e-05, "loss": 0.1311, "step": 6870 }, { "grad_norm": 0.6912688612937927, "learning_rate": 9.99035046550622e-05, "loss": 0.1182, "step": 6880 }, { "grad_norm": 0.5380834341049194, "learning_rate": 9.99024751643393e-05, "loss": 0.1127, "step": 6890 }, { "grad_norm": 0.7798677682876587, "learning_rate": 9.990144021635677e-05, "loss": 0.1104, "step": 6900 }, { "grad_norm": 0.9761165380477905, "learning_rate": 9.990039981122775e-05, "loss": 0.0922, "step": 6910 }, { "grad_norm": 1.0137526988983154, "learning_rate": 9.989935394906602e-05, "loss": 0.1071, "step": 6920 }, { "grad_norm": 0.6603493690490723, "learning_rate": 9.989830262998598e-05, "loss": 0.1188, "step": 6930 }, { "grad_norm": 0.9013267755508423, "learning_rate": 9.989724585410259e-05, "loss": 0.1026, "step": 6940 }, { "grad_norm": 0.7929587364196777, "learning_rate": 9.989618362153139e-05, "loss": 0.1039, "step": 6950 }, { "grad_norm": 0.6056116223335266, "learning_rate": 9.989511593238859e-05, "loss": 0.1045, "step": 6960 }, { "grad_norm": 0.6414000988006592, "learning_rate": 9.98940427867909e-05, "loss": 0.1185, "step": 6970 }, { "grad_norm": 0.5925828814506531, "learning_rate": 9.989296418485573e-05, "loss": 0.1076, "step": 6980 }, { "grad_norm": 0.437065988779068, "learning_rate": 9.989188012670101e-05, "loss": 0.0874, "step": 6990 }, { "grad_norm": 0.7150040864944458, "learning_rate": 9.989079061244528e-05, "loss": 0.0986, "step": 7000 }, { "grad_norm": 0.44342920184135437, "learning_rate": 9.988969564220769e-05, "loss": 0.1078, "step": 7010 }, { "grad_norm": 0.726161003112793, "learning_rate": 9.988859521610801e-05, "loss": 0.0969, "step": 7020 }, { "grad_norm": 0.7174009680747986, "learning_rate": 9.988748933426656e-05, "loss": 0.1417, "step": 7030 }, { "grad_norm": 0.6055039167404175, "learning_rate": 9.988637799680428e-05, "loss": 0.1039, "step": 7040 }, { "grad_norm": 0.605099081993103, "learning_rate": 9.98852612038427e-05, "loss": 0.0854, "step": 7050 }, { "grad_norm": 0.7901239991188049, "learning_rate": 9.988413895550397e-05, "loss": 0.098, "step": 7060 }, { "grad_norm": 0.7925850749015808, "learning_rate": 9.98830112519108e-05, "loss": 0.1092, "step": 7070 }, { "grad_norm": 0.569531261920929, "learning_rate": 9.98818780931865e-05, "loss": 0.1252, "step": 7080 }, { "grad_norm": 0.7243702411651611, "learning_rate": 9.988073947945502e-05, "loss": 0.0933, "step": 7090 }, { "grad_norm": 0.5920803546905518, "learning_rate": 9.987959541084087e-05, "loss": 0.0892, "step": 7100 }, { "grad_norm": 0.5842676758766174, "learning_rate": 9.987844588746915e-05, "loss": 0.1117, "step": 7110 }, { "grad_norm": 0.6892993450164795, "learning_rate": 9.987729090946558e-05, "loss": 0.0931, "step": 7120 }, { "grad_norm": 0.45074862241744995, "learning_rate": 9.987613047695647e-05, "loss": 0.0754, "step": 7130 }, { "grad_norm": 0.7827214002609253, "learning_rate": 9.987496459006871e-05, "loss": 0.1103, "step": 7140 }, { "grad_norm": 0.6974547505378723, "learning_rate": 9.987379324892982e-05, "loss": 0.1289, "step": 7150 }, { "grad_norm": 0.9706368446350098, "learning_rate": 9.987261645366788e-05, "loss": 0.1172, "step": 7160 }, { "grad_norm": 0.6361007690429688, "learning_rate": 9.987143420441158e-05, "loss": 0.1219, "step": 7170 }, { "grad_norm": 1.004578948020935, "learning_rate": 9.987024650129022e-05, "loss": 0.1131, "step": 7180 }, { "grad_norm": 0.46099942922592163, "learning_rate": 9.986905334443368e-05, "loss": 0.1058, "step": 7190 }, { "grad_norm": 0.709790050983429, "learning_rate": 9.986785473397245e-05, "loss": 0.1027, "step": 7200 }, { "grad_norm": 0.5822070240974426, "learning_rate": 9.98666506700376e-05, "loss": 0.1072, "step": 7210 }, { "grad_norm": 0.6398454308509827, "learning_rate": 9.986544115276081e-05, "loss": 0.1055, "step": 7220 }, { "grad_norm": 0.659543514251709, "learning_rate": 9.986422618227433e-05, "loss": 0.1088, "step": 7230 }, { "grad_norm": 0.8519414663314819, "learning_rate": 9.986300575871106e-05, "loss": 0.1224, "step": 7240 }, { "grad_norm": 0.7153586149215698, "learning_rate": 9.986177988220444e-05, "loss": 0.1116, "step": 7250 }, { "grad_norm": 0.6898037791252136, "learning_rate": 9.986054855288856e-05, "loss": 0.0952, "step": 7260 }, { "grad_norm": 0.6363664865493774, "learning_rate": 9.985931177089802e-05, "loss": 0.104, "step": 7270 }, { "grad_norm": 0.8876185417175293, "learning_rate": 9.985806953636814e-05, "loss": 0.1112, "step": 7280 }, { "grad_norm": 0.6886132955551147, "learning_rate": 9.985682184943471e-05, "loss": 0.1093, "step": 7290 }, { "grad_norm": 0.6098478436470032, "learning_rate": 9.98555687102342e-05, "loss": 0.0965, "step": 7300 }, { "grad_norm": 0.574168860912323, "learning_rate": 9.985431011890367e-05, "loss": 0.1127, "step": 7310 }, { "grad_norm": 0.8741657733917236, "learning_rate": 9.985304607558075e-05, "loss": 0.1091, "step": 7320 }, { "grad_norm": 0.7287672758102417, "learning_rate": 9.985177658040364e-05, "loss": 0.1086, "step": 7330 }, { "grad_norm": 0.795427680015564, "learning_rate": 9.985050163351119e-05, "loss": 0.1126, "step": 7340 }, { "grad_norm": 0.9462711811065674, "learning_rate": 9.984922123504286e-05, "loss": 0.0865, "step": 7350 }, { "grad_norm": 0.9155086874961853, "learning_rate": 9.984793538513862e-05, "loss": 0.1009, "step": 7360 }, { "grad_norm": 0.5647689700126648, "learning_rate": 9.984664408393912e-05, "loss": 0.1007, "step": 7370 }, { "grad_norm": 0.6107422113418579, "learning_rate": 9.984534733158556e-05, "loss": 0.1148, "step": 7380 }, { "grad_norm": 0.840684711933136, "learning_rate": 9.984404512821977e-05, "loss": 0.1112, "step": 7390 }, { "grad_norm": 0.7624855637550354, "learning_rate": 9.984273747398411e-05, "loss": 0.1283, "step": 7400 }, { "grad_norm": 0.7310076951980591, "learning_rate": 9.984142436902165e-05, "loss": 0.1135, "step": 7410 }, { "grad_norm": 0.5680682063102722, "learning_rate": 9.984010581347596e-05, "loss": 0.1151, "step": 7420 }, { "grad_norm": 0.5246697068214417, "learning_rate": 9.983878180749121e-05, "loss": 0.1062, "step": 7430 }, { "grad_norm": 0.6167922616004944, "learning_rate": 9.983745235121222e-05, "loss": 0.0972, "step": 7440 }, { "grad_norm": 0.8895117044448853, "learning_rate": 9.983611744478438e-05, "loss": 0.1037, "step": 7450 }, { "grad_norm": 1.111160159111023, "learning_rate": 9.983477708835365e-05, "loss": 0.0935, "step": 7460 }, { "grad_norm": 0.7489911317825317, "learning_rate": 9.983343128206664e-05, "loss": 0.1269, "step": 7470 }, { "grad_norm": 0.7592513561248779, "learning_rate": 9.983208002607049e-05, "loss": 0.1178, "step": 7480 }, { "grad_norm": 0.7230187058448792, "learning_rate": 9.9830723320513e-05, "loss": 0.1106, "step": 7490 }, { "grad_norm": 1.0397112369537354, "learning_rate": 9.982936116554254e-05, "loss": 0.1035, "step": 7500 }, { "grad_norm": 0.47848039865493774, "learning_rate": 9.982799356130803e-05, "loss": 0.0989, "step": 7510 }, { "grad_norm": 0.5027913451194763, "learning_rate": 9.982662050795908e-05, "loss": 0.1083, "step": 7520 }, { "grad_norm": 0.7136876583099365, "learning_rate": 9.982524200564583e-05, "loss": 0.1129, "step": 7530 }, { "grad_norm": 0.5752906203269958, "learning_rate": 9.982385805451901e-05, "loss": 0.1006, "step": 7540 }, { "grad_norm": 0.6006415486335754, "learning_rate": 9.982246865472998e-05, "loss": 0.081, "step": 7550 }, { "grad_norm": 0.5824446082115173, "learning_rate": 9.982107380643069e-05, "loss": 0.099, "step": 7560 }, { "grad_norm": 0.9410068988800049, "learning_rate": 9.981967350977368e-05, "loss": 0.1224, "step": 7570 }, { "grad_norm": 0.8008447289466858, "learning_rate": 9.981826776491208e-05, "loss": 0.101, "step": 7580 }, { "grad_norm": 0.7193572521209717, "learning_rate": 9.98168565719996e-05, "loss": 0.1252, "step": 7590 }, { "grad_norm": 0.47853004932403564, "learning_rate": 9.98154399311906e-05, "loss": 0.0985, "step": 7600 }, { "grad_norm": 0.6493760347366333, "learning_rate": 9.981401784263997e-05, "loss": 0.1121, "step": 7610 }, { "grad_norm": 0.5800392627716064, "learning_rate": 9.981259030650326e-05, "loss": 0.1056, "step": 7620 }, { "grad_norm": 0.6605609059333801, "learning_rate": 9.981115732293655e-05, "loss": 0.1027, "step": 7630 }, { "grad_norm": 1.1118844747543335, "learning_rate": 9.980971889209659e-05, "loss": 0.1048, "step": 7640 }, { "grad_norm": 0.5222964286804199, "learning_rate": 9.980827501414064e-05, "loss": 0.0968, "step": 7650 }, { "grad_norm": 0.8921046257019043, "learning_rate": 9.980682568922663e-05, "loss": 0.1094, "step": 7660 }, { "grad_norm": 0.8087310194969177, "learning_rate": 9.980537091751304e-05, "loss": 0.1009, "step": 7670 }, { "grad_norm": 0.652947723865509, "learning_rate": 9.980391069915897e-05, "loss": 0.0977, "step": 7680 }, { "grad_norm": 0.6749288439750671, "learning_rate": 9.98024450343241e-05, "loss": 0.1135, "step": 7690 }, { "grad_norm": 0.9120111465454102, "learning_rate": 9.980097392316872e-05, "loss": 0.0963, "step": 7700 }, { "grad_norm": 0.7918119430541992, "learning_rate": 9.97994973658537e-05, "loss": 0.1157, "step": 7710 }, { "grad_norm": 0.5424794554710388, "learning_rate": 9.979801536254054e-05, "loss": 0.1019, "step": 7720 }, { "grad_norm": 0.5972900390625, "learning_rate": 9.979652791339127e-05, "loss": 0.1105, "step": 7730 }, { "grad_norm": 0.5928798317909241, "learning_rate": 9.97950350185686e-05, "loss": 0.0993, "step": 7740 }, { "grad_norm": 0.46752429008483887, "learning_rate": 9.979353667823574e-05, "loss": 0.102, "step": 7750 }, { "grad_norm": 0.5524450540542603, "learning_rate": 9.979203289255658e-05, "loss": 0.1157, "step": 7760 }, { "grad_norm": 0.6718435883522034, "learning_rate": 9.979052366169557e-05, "loss": 0.1006, "step": 7770 }, { "grad_norm": 0.6022260189056396, "learning_rate": 9.978900898581775e-05, "loss": 0.1128, "step": 7780 }, { "grad_norm": 0.7571937441825867, "learning_rate": 9.978748886508875e-05, "loss": 0.0899, "step": 7790 }, { "grad_norm": 0.6159096360206604, "learning_rate": 9.978596329967484e-05, "loss": 0.0952, "step": 7800 }, { "grad_norm": 0.4983973205089569, "learning_rate": 9.978443228974284e-05, "loss": 0.0966, "step": 7810 }, { "grad_norm": 0.6066153645515442, "learning_rate": 9.978289583546015e-05, "loss": 0.0862, "step": 7820 }, { "grad_norm": 1.3125683069229126, "learning_rate": 9.978135393699484e-05, "loss": 0.0938, "step": 7830 }, { "grad_norm": 0.5808900594711304, "learning_rate": 9.977980659451548e-05, "loss": 0.0963, "step": 7840 }, { "grad_norm": 0.6512623429298401, "learning_rate": 9.977825380819135e-05, "loss": 0.0837, "step": 7850 }, { "grad_norm": 0.631701648235321, "learning_rate": 9.97766955781922e-05, "loss": 0.0962, "step": 7860 }, { "grad_norm": 0.6767593622207642, "learning_rate": 9.977513190468848e-05, "loss": 0.0994, "step": 7870 }, { "grad_norm": 0.6230704188346863, "learning_rate": 9.977356278785116e-05, "loss": 0.1039, "step": 7880 }, { "grad_norm": 0.5786467790603638, "learning_rate": 9.977198822785184e-05, "loss": 0.0936, "step": 7890 }, { "grad_norm": 0.6109111905097961, "learning_rate": 9.977040822486273e-05, "loss": 0.1015, "step": 7900 }, { "grad_norm": 0.6067367196083069, "learning_rate": 9.97688227790566e-05, "loss": 0.1064, "step": 7910 }, { "grad_norm": 0.5680882930755615, "learning_rate": 9.976723189060684e-05, "loss": 0.114, "step": 7920 }, { "grad_norm": 0.5146539807319641, "learning_rate": 9.976563555968742e-05, "loss": 0.0817, "step": 7930 }, { "grad_norm": 0.8072332739830017, "learning_rate": 9.976403378647292e-05, "loss": 0.1128, "step": 7940 }, { "grad_norm": 0.7406374216079712, "learning_rate": 9.97624265711385e-05, "loss": 0.1131, "step": 7950 }, { "grad_norm": 0.72047358751297, "learning_rate": 9.976081391385993e-05, "loss": 0.1243, "step": 7960 }, { "grad_norm": 0.7624132633209229, "learning_rate": 9.975919581481356e-05, "loss": 0.106, "step": 7970 }, { "grad_norm": 0.7036441564559937, "learning_rate": 9.975757227417634e-05, "loss": 0.0918, "step": 7980 }, { "grad_norm": 0.8662437200546265, "learning_rate": 9.975594329212586e-05, "loss": 0.1043, "step": 7990 }, { "grad_norm": 0.893536388874054, "learning_rate": 9.97543088688402e-05, "loss": 0.109, "step": 8000 }, { "grad_norm": 1.1607362031936646, "learning_rate": 9.975266900449814e-05, "loss": 0.1021, "step": 8010 }, { "grad_norm": 0.7473620176315308, "learning_rate": 9.975102369927898e-05, "loss": 0.0958, "step": 8020 }, { "grad_norm": 0.5546717643737793, "learning_rate": 9.974937295336269e-05, "loss": 0.1003, "step": 8030 }, { "grad_norm": 0.7718937397003174, "learning_rate": 9.974771676692975e-05, "loss": 0.0953, "step": 8040 }, { "grad_norm": 0.6574953198432922, "learning_rate": 9.974605514016131e-05, "loss": 0.1092, "step": 8050 }, { "grad_norm": 0.9169589877128601, "learning_rate": 9.974438807323907e-05, "loss": 0.089, "step": 8060 }, { "grad_norm": 0.5874967575073242, "learning_rate": 9.974271556634535e-05, "loss": 0.085, "step": 8070 }, { "grad_norm": 0.8421776294708252, "learning_rate": 9.974103761966302e-05, "loss": 0.1137, "step": 8080 }, { "grad_norm": 0.6870169639587402, "learning_rate": 9.973935423337563e-05, "loss": 0.1111, "step": 8090 }, { "grad_norm": 1.0839667320251465, "learning_rate": 9.973766540766722e-05, "loss": 0.1171, "step": 8100 }, { "grad_norm": 0.6874839067459106, "learning_rate": 9.97359711427225e-05, "loss": 0.1087, "step": 8110 }, { "grad_norm": 0.5222720503807068, "learning_rate": 9.973427143872677e-05, "loss": 0.1257, "step": 8120 }, { "grad_norm": 0.9225516319274902, "learning_rate": 9.973256629586589e-05, "loss": 0.1088, "step": 8130 }, { "grad_norm": 0.5717684030532837, "learning_rate": 9.973085571432632e-05, "loss": 0.1129, "step": 8140 }, { "grad_norm": 0.7461307644844055, "learning_rate": 9.972913969429513e-05, "loss": 0.1016, "step": 8150 }, { "grad_norm": 0.5909935832023621, "learning_rate": 9.972741823596e-05, "loss": 0.0984, "step": 8160 }, { "grad_norm": 0.7781122922897339, "learning_rate": 9.972569133950917e-05, "loss": 0.1178, "step": 8170 }, { "grad_norm": 0.7740604281425476, "learning_rate": 9.972395900513151e-05, "loss": 0.0981, "step": 8180 }, { "grad_norm": 0.8099576234817505, "learning_rate": 9.972222123301645e-05, "loss": 0.0983, "step": 8190 }, { "grad_norm": 0.7084090113639832, "learning_rate": 9.972047802335403e-05, "loss": 0.0989, "step": 8200 }, { "grad_norm": 0.5742366313934326, "learning_rate": 9.971872937633488e-05, "loss": 0.0903, "step": 8210 }, { "grad_norm": 0.7892888188362122, "learning_rate": 9.971697529215024e-05, "loss": 0.1083, "step": 8220 }, { "grad_norm": 0.7172331213951111, "learning_rate": 9.971521577099192e-05, "loss": 0.0934, "step": 8230 }, { "grad_norm": 0.7590874433517456, "learning_rate": 9.971345081305236e-05, "loss": 0.1066, "step": 8240 }, { "grad_norm": 0.8088352680206299, "learning_rate": 9.971168041852456e-05, "loss": 0.1032, "step": 8250 }, { "grad_norm": 0.5487443804740906, "learning_rate": 9.970990458760215e-05, "loss": 0.0854, "step": 8260 }, { "grad_norm": 0.5198103785514832, "learning_rate": 9.970812332047929e-05, "loss": 0.0929, "step": 8270 }, { "grad_norm": 0.7410672307014465, "learning_rate": 9.97063366173508e-05, "loss": 0.1052, "step": 8280 }, { "grad_norm": 0.6683383584022522, "learning_rate": 9.970454447841207e-05, "loss": 0.0899, "step": 8290 }, { "grad_norm": 0.48377135396003723, "learning_rate": 9.970274690385909e-05, "loss": 0.0943, "step": 8300 }, { "grad_norm": 0.6085314750671387, "learning_rate": 9.970094389388844e-05, "loss": 0.0803, "step": 8310 }, { "grad_norm": 0.6038764715194702, "learning_rate": 9.969913544869728e-05, "loss": 0.0852, "step": 8320 }, { "grad_norm": 0.8312991857528687, "learning_rate": 9.96973215684834e-05, "loss": 0.0938, "step": 8330 }, { "grad_norm": 0.6899269819259644, "learning_rate": 9.969550225344513e-05, "loss": 0.1146, "step": 8340 }, { "grad_norm": 0.7521376013755798, "learning_rate": 9.969367750378147e-05, "loss": 0.1094, "step": 8350 }, { "grad_norm": 0.8954096436500549, "learning_rate": 9.969184731969194e-05, "loss": 0.0892, "step": 8360 }, { "grad_norm": 0.6137803792953491, "learning_rate": 9.96900117013767e-05, "loss": 0.1055, "step": 8370 }, { "grad_norm": 0.7290326356887817, "learning_rate": 9.96881706490365e-05, "loss": 0.1082, "step": 8380 }, { "grad_norm": 0.6008872389793396, "learning_rate": 9.968632416287265e-05, "loss": 0.0802, "step": 8390 }, { "grad_norm": 0.5940893888473511, "learning_rate": 9.96844722430871e-05, "loss": 0.0846, "step": 8400 }, { "grad_norm": 0.8665472865104675, "learning_rate": 9.968261488988235e-05, "loss": 0.1047, "step": 8410 }, { "grad_norm": 0.6281729936599731, "learning_rate": 9.968075210346155e-05, "loss": 0.094, "step": 8420 }, { "grad_norm": 0.7053804397583008, "learning_rate": 9.967888388402839e-05, "loss": 0.0972, "step": 8430 }, { "grad_norm": 0.6734901666641235, "learning_rate": 9.967701023178717e-05, "loss": 0.0909, "step": 8440 }, { "grad_norm": 0.8738085031509399, "learning_rate": 9.967513114694282e-05, "loss": 0.0965, "step": 8450 }, { "grad_norm": 0.6015689969062805, "learning_rate": 9.967324662970079e-05, "loss": 0.0892, "step": 8460 }, { "grad_norm": 0.6327351927757263, "learning_rate": 9.96713566802672e-05, "loss": 0.0979, "step": 8470 }, { "grad_norm": 0.6199440956115723, "learning_rate": 9.966946129884873e-05, "loss": 0.0961, "step": 8480 }, { "grad_norm": 0.9898268580436707, "learning_rate": 9.966756048565265e-05, "loss": 0.0992, "step": 8490 }, { "grad_norm": 0.7520793676376343, "learning_rate": 9.966565424088681e-05, "loss": 0.1146, "step": 8500 }, { "grad_norm": 0.902530312538147, "learning_rate": 9.96637425647597e-05, "loss": 0.102, "step": 8510 }, { "grad_norm": 0.6565770506858826, "learning_rate": 9.966182545748038e-05, "loss": 0.0966, "step": 8520 }, { "grad_norm": 0.9138883352279663, "learning_rate": 9.96599029192585e-05, "loss": 0.1082, "step": 8530 }, { "grad_norm": 0.6289044618606567, "learning_rate": 9.965797495030428e-05, "loss": 0.1105, "step": 8540 }, { "grad_norm": 0.6484946012496948, "learning_rate": 9.96560415508286e-05, "loss": 0.1059, "step": 8550 }, { "grad_norm": 0.7479314208030701, "learning_rate": 9.965410272104286e-05, "loss": 0.1035, "step": 8560 }, { "grad_norm": 0.72166907787323, "learning_rate": 9.96521584611591e-05, "loss": 0.1316, "step": 8570 }, { "grad_norm": 0.7521912455558777, "learning_rate": 9.965020877138994e-05, "loss": 0.1201, "step": 8580 }, { "grad_norm": 0.557361364364624, "learning_rate": 9.964825365194861e-05, "loss": 0.1132, "step": 8590 }, { "grad_norm": 0.7018749117851257, "learning_rate": 9.96462931030489e-05, "loss": 0.0911, "step": 8600 }, { "grad_norm": 0.6304975748062134, "learning_rate": 9.96443271249052e-05, "loss": 0.0881, "step": 8610 }, { "grad_norm": 0.5149633884429932, "learning_rate": 9.964235571773255e-05, "loss": 0.1071, "step": 8620 }, { "grad_norm": 0.8297896981239319, "learning_rate": 9.96403788817465e-05, "loss": 0.1081, "step": 8630 }, { "grad_norm": 0.615426242351532, "learning_rate": 9.963839661716325e-05, "loss": 0.0981, "step": 8640 }, { "grad_norm": 0.8774893879890442, "learning_rate": 9.963640892419958e-05, "loss": 0.0984, "step": 8650 }, { "grad_norm": 0.8334382772445679, "learning_rate": 9.963441580307286e-05, "loss": 0.0851, "step": 8660 }, { "grad_norm": 0.6264950037002563, "learning_rate": 9.963241725400104e-05, "loss": 0.0978, "step": 8670 }, { "grad_norm": 0.5789666175842285, "learning_rate": 9.963041327720271e-05, "loss": 0.1012, "step": 8680 }, { "grad_norm": 0.6526740789413452, "learning_rate": 9.962840387289697e-05, "loss": 0.1085, "step": 8690 }, { "grad_norm": 0.9680560231208801, "learning_rate": 9.962638904130363e-05, "loss": 0.1066, "step": 8700 }, { "grad_norm": 0.6210917830467224, "learning_rate": 9.962436878264298e-05, "loss": 0.1151, "step": 8710 }, { "grad_norm": 0.6544479131698608, "learning_rate": 9.962234309713598e-05, "loss": 0.112, "step": 8720 }, { "grad_norm": 0.6015482544898987, "learning_rate": 9.962031198500414e-05, "loss": 0.1005, "step": 8730 }, { "grad_norm": 0.5175290107727051, "learning_rate": 9.961827544646958e-05, "loss": 0.0808, "step": 8740 }, { "grad_norm": 0.655426561832428, "learning_rate": 9.961623348175501e-05, "loss": 0.1007, "step": 8750 }, { "grad_norm": 0.5737785696983337, "learning_rate": 9.961418609108377e-05, "loss": 0.0821, "step": 8760 }, { "grad_norm": 0.6903999447822571, "learning_rate": 9.961213327467971e-05, "loss": 0.1065, "step": 8770 }, { "grad_norm": 0.5930440425872803, "learning_rate": 9.961007503276736e-05, "loss": 0.09, "step": 8780 }, { "grad_norm": 0.6147800087928772, "learning_rate": 9.960801136557179e-05, "loss": 0.0748, "step": 8790 }, { "grad_norm": 0.7385327219963074, "learning_rate": 9.960594227331866e-05, "loss": 0.0816, "step": 8800 }, { "grad_norm": 0.4261583983898163, "learning_rate": 9.960386775623429e-05, "loss": 0.0732, "step": 8810 }, { "grad_norm": 0.5831130146980286, "learning_rate": 9.96017878145455e-05, "loss": 0.0889, "step": 8820 }, { "grad_norm": 0.6478028893470764, "learning_rate": 9.959970244847977e-05, "loss": 0.0899, "step": 8830 }, { "grad_norm": 0.6489607691764832, "learning_rate": 9.959761165826518e-05, "loss": 0.0929, "step": 8840 }, { "grad_norm": 0.8064051270484924, "learning_rate": 9.959551544413033e-05, "loss": 0.0947, "step": 8850 }, { "grad_norm": 0.5369306802749634, "learning_rate": 9.959341380630448e-05, "loss": 0.0871, "step": 8860 }, { "grad_norm": 0.691179633140564, "learning_rate": 9.959130674501746e-05, "loss": 0.093, "step": 8870 }, { "grad_norm": 0.7499724626541138, "learning_rate": 9.958919426049968e-05, "loss": 0.1102, "step": 8880 }, { "grad_norm": 0.6880610585212708, "learning_rate": 9.958707635298219e-05, "loss": 0.0875, "step": 8890 }, { "grad_norm": 0.5305365324020386, "learning_rate": 9.958495302269657e-05, "loss": 0.0892, "step": 8900 }, { "grad_norm": 0.8324135541915894, "learning_rate": 9.958282426987503e-05, "loss": 0.0984, "step": 8910 }, { "grad_norm": 0.8099247813224792, "learning_rate": 9.95806900947504e-05, "loss": 0.1048, "step": 8920 }, { "grad_norm": 0.664940357208252, "learning_rate": 9.957855049755604e-05, "loss": 0.0942, "step": 8930 }, { "grad_norm": 0.5960182547569275, "learning_rate": 9.957640547852593e-05, "loss": 0.0955, "step": 8940 }, { "grad_norm": 0.5412540435791016, "learning_rate": 9.957425503789466e-05, "loss": 0.0851, "step": 8950 }, { "grad_norm": 0.7042754292488098, "learning_rate": 9.957209917589738e-05, "loss": 0.0891, "step": 8960 }, { "grad_norm": 0.5389026403427124, "learning_rate": 9.956993789276987e-05, "loss": 0.0946, "step": 8970 }, { "grad_norm": 0.6309081315994263, "learning_rate": 9.956777118874847e-05, "loss": 0.1018, "step": 8980 }, { "grad_norm": 0.8134849667549133, "learning_rate": 9.956559906407016e-05, "loss": 0.1236, "step": 8990 }, { "grad_norm": 0.6768167018890381, "learning_rate": 9.956342151897245e-05, "loss": 0.1002, "step": 9000 }, { "grad_norm": 0.5688517093658447, "learning_rate": 9.956123855369346e-05, "loss": 0.0866, "step": 9010 }, { "grad_norm": 0.5729214549064636, "learning_rate": 9.955905016847196e-05, "loss": 0.0809, "step": 9020 }, { "grad_norm": 0.542509913444519, "learning_rate": 9.955685636354723e-05, "loss": 0.1139, "step": 9030 }, { "grad_norm": 0.5070716738700867, "learning_rate": 9.95546571391592e-05, "loss": 0.1, "step": 9040 }, { "grad_norm": 0.6099133491516113, "learning_rate": 9.955245249554837e-05, "loss": 0.0827, "step": 9050 }, { "grad_norm": 0.45675498247146606, "learning_rate": 9.955024243295582e-05, "loss": 0.1148, "step": 9060 }, { "grad_norm": 0.8353921175003052, "learning_rate": 9.954802695162328e-05, "loss": 0.0863, "step": 9070 }, { "grad_norm": 0.6477458477020264, "learning_rate": 9.954580605179302e-05, "loss": 0.091, "step": 9080 }, { "grad_norm": 0.7598150372505188, "learning_rate": 9.954357973370788e-05, "loss": 0.1025, "step": 9090 }, { "grad_norm": 0.7905017733573914, "learning_rate": 9.954134799761135e-05, "loss": 0.1075, "step": 9100 }, { "grad_norm": 0.6380054354667664, "learning_rate": 9.953911084374748e-05, "loss": 0.1033, "step": 9110 }, { "grad_norm": 0.8428301811218262, "learning_rate": 9.953686827236093e-05, "loss": 0.1207, "step": 9120 }, { "grad_norm": 0.5878851413726807, "learning_rate": 9.953462028369695e-05, "loss": 0.1228, "step": 9130 }, { "grad_norm": 0.6690261960029602, "learning_rate": 9.953236687800136e-05, "loss": 0.1053, "step": 9140 }, { "grad_norm": 0.6564236879348755, "learning_rate": 9.95301080555206e-05, "loss": 0.1116, "step": 9150 }, { "grad_norm": 0.596000075340271, "learning_rate": 9.952784381650171e-05, "loss": 0.0787, "step": 9160 }, { "grad_norm": 0.637140154838562, "learning_rate": 9.952557416119226e-05, "loss": 0.1003, "step": 9170 }, { "grad_norm": 0.7661033272743225, "learning_rate": 9.95232990898405e-05, "loss": 0.0892, "step": 9180 }, { "grad_norm": 0.663573682308197, "learning_rate": 9.95210186026952e-05, "loss": 0.0911, "step": 9190 }, { "grad_norm": 0.5088696479797363, "learning_rate": 9.951873270000576e-05, "loss": 0.0996, "step": 9200 }, { "grad_norm": 0.6549709439277649, "learning_rate": 9.951644138202216e-05, "loss": 0.1129, "step": 9210 }, { "grad_norm": 0.6970807909965515, "learning_rate": 9.951414464899498e-05, "loss": 0.0892, "step": 9220 }, { "grad_norm": 0.6008015871047974, "learning_rate": 9.951184250117538e-05, "loss": 0.0864, "step": 9230 }, { "grad_norm": 1.1846309900283813, "learning_rate": 9.950953493881513e-05, "loss": 0.1059, "step": 9240 }, { "grad_norm": 0.6337588429450989, "learning_rate": 9.950722196216658e-05, "loss": 0.092, "step": 9250 }, { "grad_norm": 0.5814711451530457, "learning_rate": 9.950490357148265e-05, "loss": 0.0814, "step": 9260 }, { "grad_norm": 0.5352410674095154, "learning_rate": 9.950257976701692e-05, "loss": 0.1025, "step": 9270 }, { "grad_norm": 0.7596023678779602, "learning_rate": 9.950025054902348e-05, "loss": 0.1001, "step": 9280 }, { "grad_norm": 0.794547438621521, "learning_rate": 9.949791591775706e-05, "loss": 0.086, "step": 9290 }, { "grad_norm": 0.4618561863899231, "learning_rate": 9.949557587347298e-05, "loss": 0.1012, "step": 9300 }, { "grad_norm": 0.835552990436554, "learning_rate": 9.949323041642713e-05, "loss": 0.0934, "step": 9310 }, { "grad_norm": 0.6086687445640564, "learning_rate": 9.949087954687602e-05, "loss": 0.1137, "step": 9320 }, { "grad_norm": 0.8243016600608826, "learning_rate": 9.948852326507672e-05, "loss": 0.0897, "step": 9330 }, { "grad_norm": 0.7811567783355713, "learning_rate": 9.948616157128694e-05, "loss": 0.0837, "step": 9340 }, { "grad_norm": 0.6112220883369446, "learning_rate": 9.948379446576493e-05, "loss": 0.0906, "step": 9350 }, { "grad_norm": 0.6677262783050537, "learning_rate": 9.948142194876952e-05, "loss": 0.0851, "step": 9360 }, { "grad_norm": 0.4952331781387329, "learning_rate": 9.947904402056024e-05, "loss": 0.0994, "step": 9370 }, { "grad_norm": 0.6933297514915466, "learning_rate": 9.947666068139708e-05, "loss": 0.0974, "step": 9380 }, { "grad_norm": 0.6626335978507996, "learning_rate": 9.947427193154071e-05, "loss": 0.0922, "step": 9390 }, { "grad_norm": 0.5527588725090027, "learning_rate": 9.947187777125233e-05, "loss": 0.1177, "step": 9400 }, { "grad_norm": 0.6219552755355835, "learning_rate": 9.946947820079377e-05, "loss": 0.1012, "step": 9410 }, { "grad_norm": 0.7931899428367615, "learning_rate": 9.946707322042747e-05, "loss": 0.0955, "step": 9420 }, { "grad_norm": 0.9511201977729797, "learning_rate": 9.94646628304164e-05, "loss": 0.0956, "step": 9430 }, { "grad_norm": 0.7198171019554138, "learning_rate": 9.946224703102418e-05, "loss": 0.0884, "step": 9440 }, { "grad_norm": 0.7507266402244568, "learning_rate": 9.945982582251498e-05, "loss": 0.0838, "step": 9450 }, { "grad_norm": 0.6318084597587585, "learning_rate": 9.94573992051536e-05, "loss": 0.0845, "step": 9460 }, { "grad_norm": 0.46475329995155334, "learning_rate": 9.94549671792054e-05, "loss": 0.1024, "step": 9470 }, { "grad_norm": 0.9068580865859985, "learning_rate": 9.945252974493635e-05, "loss": 0.0831, "step": 9480 }, { "grad_norm": 0.4165746569633484, "learning_rate": 9.9450086902613e-05, "loss": 0.0905, "step": 9490 }, { "grad_norm": 0.5813862085342407, "learning_rate": 9.944763865250248e-05, "loss": 0.0871, "step": 9500 }, { "grad_norm": 0.5590550303459167, "learning_rate": 9.944518499487254e-05, "loss": 0.0759, "step": 9510 }, { "grad_norm": 0.7295913100242615, "learning_rate": 9.944272592999151e-05, "loss": 0.1015, "step": 9520 }, { "grad_norm": 0.45298314094543457, "learning_rate": 9.94402614581283e-05, "loss": 0.0887, "step": 9530 }, { "grad_norm": 0.7127010226249695, "learning_rate": 9.943779157955244e-05, "loss": 0.0928, "step": 9540 }, { "grad_norm": 0.5354458093643188, "learning_rate": 9.943531629453403e-05, "loss": 0.0808, "step": 9550 }, { "grad_norm": 0.7757395505905151, "learning_rate": 9.943283560334375e-05, "loss": 0.0935, "step": 9560 }, { "grad_norm": 0.6893711090087891, "learning_rate": 9.943034950625288e-05, "loss": 0.0988, "step": 9570 }, { "grad_norm": 0.49298346042633057, "learning_rate": 9.942785800353332e-05, "loss": 0.0831, "step": 9580 }, { "grad_norm": 0.6142018437385559, "learning_rate": 9.942536109545751e-05, "loss": 0.0834, "step": 9590 }, { "grad_norm": 0.8803045749664307, "learning_rate": 9.942285878229853e-05, "loss": 0.0952, "step": 9600 }, { "grad_norm": 0.8178030252456665, "learning_rate": 9.942035106433001e-05, "loss": 0.1002, "step": 9610 }, { "grad_norm": 1.1434396505355835, "learning_rate": 9.94178379418262e-05, "loss": 0.1238, "step": 9620 }, { "grad_norm": 0.7278323769569397, "learning_rate": 9.941531941506194e-05, "loss": 0.0882, "step": 9630 }, { "grad_norm": 0.772438108921051, "learning_rate": 9.941279548431263e-05, "loss": 0.0876, "step": 9640 }, { "grad_norm": 0.5366286635398865, "learning_rate": 9.941026614985431e-05, "loss": 0.1063, "step": 9650 }, { "grad_norm": 0.4519018530845642, "learning_rate": 9.940773141196357e-05, "loss": 0.0914, "step": 9660 }, { "grad_norm": 0.7294980883598328, "learning_rate": 9.94051912709176e-05, "loss": 0.0934, "step": 9670 }, { "grad_norm": 0.5908989906311035, "learning_rate": 9.940264572699421e-05, "loss": 0.0825, "step": 9680 }, { "grad_norm": 0.6752148270606995, "learning_rate": 9.940009478047174e-05, "loss": 0.0856, "step": 9690 }, { "grad_norm": 0.5647140145301819, "learning_rate": 9.939753843162918e-05, "loss": 0.0817, "step": 9700 }, { "grad_norm": 0.6161364316940308, "learning_rate": 9.939497668074609e-05, "loss": 0.0789, "step": 9710 }, { "grad_norm": 0.7794435620307922, "learning_rate": 9.93924095281026e-05, "loss": 0.0936, "step": 9720 }, { "grad_norm": 0.5598644614219666, "learning_rate": 9.938983697397948e-05, "loss": 0.0922, "step": 9730 }, { "grad_norm": 0.6931707262992859, "learning_rate": 9.938725901865805e-05, "loss": 0.098, "step": 9740 }, { "grad_norm": 0.6073430776596069, "learning_rate": 9.93846756624202e-05, "loss": 0.0724, "step": 9750 }, { "grad_norm": 0.5996362566947937, "learning_rate": 9.938208690554849e-05, "loss": 0.0986, "step": 9760 }, { "grad_norm": 0.571815550327301, "learning_rate": 9.9379492748326e-05, "loss": 0.0925, "step": 9770 }, { "grad_norm": 0.6997599005699158, "learning_rate": 9.937689319103641e-05, "loss": 0.0986, "step": 9780 }, { "grad_norm": 0.6857725977897644, "learning_rate": 9.937428823396404e-05, "loss": 0.0875, "step": 9790 }, { "grad_norm": 0.7181312441825867, "learning_rate": 9.937167787739372e-05, "loss": 0.0726, "step": 9800 }, { "grad_norm": 0.602241039276123, "learning_rate": 9.936906212161095e-05, "loss": 0.0896, "step": 9810 }, { "grad_norm": 0.6406794786453247, "learning_rate": 9.936644096690176e-05, "loss": 0.1004, "step": 9820 }, { "grad_norm": 0.6658282279968262, "learning_rate": 9.936381441355282e-05, "loss": 0.0884, "step": 9830 }, { "grad_norm": 0.7815659046173096, "learning_rate": 9.936118246185136e-05, "loss": 0.1118, "step": 9840 }, { "grad_norm": 1.136521816253662, "learning_rate": 9.935854511208518e-05, "loss": 0.1049, "step": 9850 }, { "grad_norm": 0.6989887356758118, "learning_rate": 9.935590236454272e-05, "loss": 0.0934, "step": 9860 }, { "grad_norm": 0.5394425392150879, "learning_rate": 9.935325421951298e-05, "loss": 0.0701, "step": 9870 }, { "grad_norm": 0.4218885600566864, "learning_rate": 9.935060067728557e-05, "loss": 0.0835, "step": 9880 }, { "grad_norm": 0.6177676916122437, "learning_rate": 9.934794173815067e-05, "loss": 0.0717, "step": 9890 }, { "grad_norm": 0.6528549790382385, "learning_rate": 9.934527740239906e-05, "loss": 0.0902, "step": 9900 }, { "grad_norm": 0.6012160778045654, "learning_rate": 9.934260767032209e-05, "loss": 0.0962, "step": 9910 }, { "grad_norm": 0.6366234421730042, "learning_rate": 9.933993254221172e-05, "loss": 0.0899, "step": 9920 }, { "grad_norm": 1.0889638662338257, "learning_rate": 9.933725201836053e-05, "loss": 0.0851, "step": 9930 }, { "grad_norm": 0.7319002747535706, "learning_rate": 9.933456609906162e-05, "loss": 0.0833, "step": 9940 }, { "grad_norm": 0.5506346821784973, "learning_rate": 9.933187478460875e-05, "loss": 0.0966, "step": 9950 }, { "grad_norm": 0.7259295582771301, "learning_rate": 9.93291780752962e-05, "loss": 0.0944, "step": 9960 }, { "grad_norm": 0.5964837670326233, "learning_rate": 9.932647597141893e-05, "loss": 0.0883, "step": 9970 }, { "grad_norm": 0.6149318218231201, "learning_rate": 9.932376847327239e-05, "loss": 0.0924, "step": 9980 }, { "grad_norm": 0.6615937948226929, "learning_rate": 9.932105558115268e-05, "loss": 0.0936, "step": 9990 }, { "grad_norm": 0.579085648059845, "learning_rate": 9.931833729535651e-05, "loss": 0.0918, "step": 10000 }, { "grad_norm": 0.6954455971717834, "learning_rate": 9.931561361618111e-05, "loss": 0.1016, "step": 10010 }, { "grad_norm": 0.7475769519805908, "learning_rate": 9.931288454392435e-05, "loss": 0.1058, "step": 10020 }, { "grad_norm": 0.7332524657249451, "learning_rate": 9.931015007888467e-05, "loss": 0.0912, "step": 10030 }, { "grad_norm": 0.6252124309539795, "learning_rate": 9.930741022136112e-05, "loss": 0.0939, "step": 10040 }, { "grad_norm": 0.611282229423523, "learning_rate": 9.930466497165333e-05, "loss": 0.0884, "step": 10050 }, { "grad_norm": 0.9351416230201721, "learning_rate": 9.93019143300615e-05, "loss": 0.087, "step": 10060 }, { "grad_norm": 0.5431656241416931, "learning_rate": 9.929915829688644e-05, "loss": 0.0835, "step": 10070 }, { "grad_norm": 0.49141761660575867, "learning_rate": 9.929639687242955e-05, "loss": 0.079, "step": 10080 }, { "grad_norm": 0.6633214354515076, "learning_rate": 9.929363005699281e-05, "loss": 0.091, "step": 10090 }, { "grad_norm": 0.519663393497467, "learning_rate": 9.92908578508788e-05, "loss": 0.0907, "step": 10100 }, { "grad_norm": 0.6544934511184692, "learning_rate": 9.928808025439069e-05, "loss": 0.0968, "step": 10110 }, { "grad_norm": 0.5397289395332336, "learning_rate": 9.928529726783223e-05, "loss": 0.0832, "step": 10120 }, { "grad_norm": 0.8187164664268494, "learning_rate": 9.928250889150774e-05, "loss": 0.0962, "step": 10130 }, { "grad_norm": 0.5687466263771057, "learning_rate": 9.92797151257222e-05, "loss": 0.0918, "step": 10140 }, { "grad_norm": 0.6566798686981201, "learning_rate": 9.927691597078108e-05, "loss": 0.1027, "step": 10150 }, { "grad_norm": 0.7931198477745056, "learning_rate": 9.927411142699053e-05, "loss": 0.0931, "step": 10160 }, { "grad_norm": 0.8178314566612244, "learning_rate": 9.927130149465725e-05, "loss": 0.095, "step": 10170 }, { "grad_norm": 0.6266529560089111, "learning_rate": 9.92684861740885e-05, "loss": 0.1108, "step": 10180 }, { "grad_norm": 0.8314180374145508, "learning_rate": 9.926566546559217e-05, "loss": 0.106, "step": 10190 }, { "grad_norm": 0.5800102949142456, "learning_rate": 9.926283936947673e-05, "loss": 0.0968, "step": 10200 }, { "grad_norm": 1.1444705724716187, "learning_rate": 9.926000788605126e-05, "loss": 0.0956, "step": 10210 }, { "grad_norm": 0.6112441420555115, "learning_rate": 9.92571710156254e-05, "loss": 0.0877, "step": 10220 }, { "grad_norm": 0.5109632015228271, "learning_rate": 9.925432875850936e-05, "loss": 0.0933, "step": 10230 }, { "grad_norm": 0.6982027292251587, "learning_rate": 9.925148111501396e-05, "loss": 0.0931, "step": 10240 }, { "grad_norm": 0.6598369479179382, "learning_rate": 9.924862808545066e-05, "loss": 0.0904, "step": 10250 }, { "grad_norm": 0.6061127185821533, "learning_rate": 9.924576967013141e-05, "loss": 0.0745, "step": 10260 }, { "grad_norm": 1.0508956909179688, "learning_rate": 9.924290586936887e-05, "loss": 0.0809, "step": 10270 }, { "grad_norm": 0.500775933265686, "learning_rate": 9.924003668347614e-05, "loss": 0.0932, "step": 10280 }, { "grad_norm": 0.646136999130249, "learning_rate": 9.923716211276704e-05, "loss": 0.0991, "step": 10290 }, { "grad_norm": 0.8191885948181152, "learning_rate": 9.923428215755594e-05, "loss": 0.0804, "step": 10300 }, { "grad_norm": 0.8680879473686218, "learning_rate": 9.923139681815775e-05, "loss": 0.08, "step": 10310 }, { "grad_norm": 1.0152759552001953, "learning_rate": 9.922850609488801e-05, "loss": 0.0837, "step": 10320 }, { "grad_norm": 1.1181769371032715, "learning_rate": 9.922560998806287e-05, "loss": 0.0909, "step": 10330 }, { "grad_norm": 0.5104462504386902, "learning_rate": 9.922270849799905e-05, "loss": 0.1, "step": 10340 }, { "grad_norm": 0.8135637640953064, "learning_rate": 9.92198016250138e-05, "loss": 0.0929, "step": 10350 }, { "grad_norm": 0.6625185608863831, "learning_rate": 9.921688936942506e-05, "loss": 0.0907, "step": 10360 }, { "grad_norm": 0.5215626358985901, "learning_rate": 9.921397173155129e-05, "loss": 0.0774, "step": 10370 }, { "grad_norm": 1.3672606945037842, "learning_rate": 9.921104871171157e-05, "loss": 0.0806, "step": 10380 }, { "grad_norm": 0.5872976183891296, "learning_rate": 9.920812031022554e-05, "loss": 0.0931, "step": 10390 }, { "grad_norm": 0.8582123517990112, "learning_rate": 9.920518652741348e-05, "loss": 0.0955, "step": 10400 }, { "grad_norm": 0.759380042552948, "learning_rate": 9.920224736359618e-05, "loss": 0.1018, "step": 10410 }, { "grad_norm": 1.0057319402694702, "learning_rate": 9.91993028190951e-05, "loss": 0.1073, "step": 10420 }, { "grad_norm": 0.6194030046463013, "learning_rate": 9.919635289423222e-05, "loss": 0.1071, "step": 10430 }, { "grad_norm": 0.5899195075035095, "learning_rate": 9.919339758933015e-05, "loss": 0.0844, "step": 10440 }, { "grad_norm": 0.4962402880191803, "learning_rate": 9.919043690471209e-05, "loss": 0.0939, "step": 10450 }, { "grad_norm": 0.6787176728248596, "learning_rate": 9.91874708407018e-05, "loss": 0.0865, "step": 10460 }, { "grad_norm": 0.5389747619628906, "learning_rate": 9.918449939762367e-05, "loss": 0.0696, "step": 10470 }, { "grad_norm": 0.4380846917629242, "learning_rate": 9.91815225758026e-05, "loss": 0.0878, "step": 10480 }, { "grad_norm": 0.4553831219673157, "learning_rate": 9.917854037556419e-05, "loss": 0.0767, "step": 10490 }, { "grad_norm": 0.9113444685935974, "learning_rate": 9.917555279723454e-05, "loss": 0.0792, "step": 10500 }, { "grad_norm": 0.6107898354530334, "learning_rate": 9.917255984114036e-05, "loss": 0.105, "step": 10510 }, { "grad_norm": 0.5476572513580322, "learning_rate": 9.916956150760896e-05, "loss": 0.0751, "step": 10520 }, { "grad_norm": 0.512717604637146, "learning_rate": 9.916655779696826e-05, "loss": 0.0938, "step": 10530 }, { "grad_norm": 0.6688931584358215, "learning_rate": 9.916354870954671e-05, "loss": 0.0794, "step": 10540 }, { "grad_norm": 0.6567761898040771, "learning_rate": 9.91605342456734e-05, "loss": 0.1038, "step": 10550 }, { "grad_norm": 0.6602858304977417, "learning_rate": 9.915751440567795e-05, "loss": 0.093, "step": 10560 }, { "grad_norm": 0.6549362540245056, "learning_rate": 9.915448918989066e-05, "loss": 0.0918, "step": 10570 }, { "grad_norm": 0.7620349526405334, "learning_rate": 9.915145859864232e-05, "loss": 0.0935, "step": 10580 }, { "grad_norm": 0.9084587693214417, "learning_rate": 9.914842263226437e-05, "loss": 0.111, "step": 10590 }, { "grad_norm": 0.46949082612991333, "learning_rate": 9.914538129108882e-05, "loss": 0.1053, "step": 10600 }, { "grad_norm": 0.48312342166900635, "learning_rate": 9.914233457544825e-05, "loss": 0.1193, "step": 10610 }, { "grad_norm": 0.6316445469856262, "learning_rate": 9.913928248567586e-05, "loss": 0.1151, "step": 10620 }, { "grad_norm": 0.6741547584533691, "learning_rate": 9.913622502210542e-05, "loss": 0.0747, "step": 10630 }, { "grad_norm": 0.7288582921028137, "learning_rate": 9.913316218507128e-05, "loss": 0.1084, "step": 10640 }, { "grad_norm": 0.5596374273300171, "learning_rate": 9.91300939749084e-05, "loss": 0.0879, "step": 10650 }, { "grad_norm": 0.7031893134117126, "learning_rate": 9.91270203919523e-05, "loss": 0.1154, "step": 10660 }, { "grad_norm": 0.64735347032547, "learning_rate": 9.912394143653912e-05, "loss": 0.1017, "step": 10670 }, { "grad_norm": 0.7628655433654785, "learning_rate": 9.912085710900555e-05, "loss": 0.0896, "step": 10680 }, { "grad_norm": 0.45382553339004517, "learning_rate": 9.911776740968892e-05, "loss": 0.088, "step": 10690 }, { "grad_norm": 0.6853010058403015, "learning_rate": 9.911467233892709e-05, "loss": 0.1111, "step": 10700 }, { "grad_norm": 0.6829540729522705, "learning_rate": 9.911157189705853e-05, "loss": 0.1132, "step": 10710 }, { "grad_norm": 0.5504010319709778, "learning_rate": 9.910846608442229e-05, "loss": 0.0835, "step": 10720 }, { "grad_norm": 0.7723004817962646, "learning_rate": 9.910535490135805e-05, "loss": 0.0802, "step": 10730 }, { "grad_norm": 0.6907552480697632, "learning_rate": 9.910223834820603e-05, "loss": 0.0843, "step": 10740 }, { "grad_norm": 0.655500054359436, "learning_rate": 9.909911642530703e-05, "loss": 0.1052, "step": 10750 }, { "grad_norm": 0.5314609408378601, "learning_rate": 9.909598913300249e-05, "loss": 0.1268, "step": 10760 }, { "grad_norm": 0.7562554478645325, "learning_rate": 9.909285647163438e-05, "loss": 0.1049, "step": 10770 }, { "grad_norm": 0.9115356802940369, "learning_rate": 9.908971844154531e-05, "loss": 0.1136, "step": 10780 }, { "grad_norm": 0.5498230457305908, "learning_rate": 9.908657504307843e-05, "loss": 0.0941, "step": 10790 }, { "grad_norm": 0.6020989418029785, "learning_rate": 9.908342627657751e-05, "loss": 0.1046, "step": 10800 }, { "grad_norm": 0.5496602654457092, "learning_rate": 9.908027214238689e-05, "loss": 0.0813, "step": 10810 }, { "grad_norm": 0.761989951133728, "learning_rate": 9.90771126408515e-05, "loss": 0.0963, "step": 10820 }, { "grad_norm": 0.5325881242752075, "learning_rate": 9.907394777231685e-05, "loss": 0.0903, "step": 10830 }, { "grad_norm": 0.4516233503818512, "learning_rate": 9.907077753712905e-05, "loss": 0.1138, "step": 10840 }, { "grad_norm": 0.6935617923736572, "learning_rate": 9.906760193563482e-05, "loss": 0.1004, "step": 10850 }, { "grad_norm": 0.921842634677887, "learning_rate": 9.906442096818139e-05, "loss": 0.0958, "step": 10860 }, { "grad_norm": 0.7598171234130859, "learning_rate": 9.906123463511665e-05, "loss": 0.1109, "step": 10870 }, { "grad_norm": 0.5349186062812805, "learning_rate": 9.905804293678907e-05, "loss": 0.0873, "step": 10880 }, { "grad_norm": 0.6552234292030334, "learning_rate": 9.905484587354766e-05, "loss": 0.0958, "step": 10890 }, { "grad_norm": 0.5796758532524109, "learning_rate": 9.905164344574205e-05, "loss": 0.0846, "step": 10900 }, { "grad_norm": 0.652521550655365, "learning_rate": 9.904843565372248e-05, "loss": 0.0738, "step": 10910 }, { "grad_norm": 0.4967607259750366, "learning_rate": 9.904522249783972e-05, "loss": 0.098, "step": 10920 }, { "grad_norm": 0.6844785213470459, "learning_rate": 9.904200397844517e-05, "loss": 0.1127, "step": 10930 }, { "grad_norm": 0.6616131663322449, "learning_rate": 9.903878009589078e-05, "loss": 0.0938, "step": 10940 }, { "grad_norm": 0.7113918662071228, "learning_rate": 9.903555085052915e-05, "loss": 0.0832, "step": 10950 }, { "grad_norm": 0.5701022148132324, "learning_rate": 9.903231624271338e-05, "loss": 0.0794, "step": 10960 }, { "grad_norm": 0.5750155448913574, "learning_rate": 9.902907627279724e-05, "loss": 0.0788, "step": 10970 }, { "grad_norm": 0.6445664763450623, "learning_rate": 9.902583094113504e-05, "loss": 0.0821, "step": 10980 }, { "grad_norm": 0.7278185486793518, "learning_rate": 9.902258024808168e-05, "loss": 0.0903, "step": 10990 }, { "grad_norm": 0.6076943278312683, "learning_rate": 9.901932419399264e-05, "loss": 0.0884, "step": 11000 }, { "grad_norm": 0.6604824662208557, "learning_rate": 9.9016062779224e-05, "loss": 0.1006, "step": 11010 }, { "grad_norm": 0.6450875401496887, "learning_rate": 9.901279600413242e-05, "loss": 0.0869, "step": 11020 }, { "grad_norm": 0.9067969918251038, "learning_rate": 9.900952386907518e-05, "loss": 0.0876, "step": 11030 }, { "grad_norm": 0.6899270415306091, "learning_rate": 9.90062463744101e-05, "loss": 0.0824, "step": 11040 }, { "grad_norm": 1.043031096458435, "learning_rate": 9.900296352049558e-05, "loss": 0.0825, "step": 11050 }, { "grad_norm": 0.46811643242836, "learning_rate": 9.899967530769065e-05, "loss": 0.1039, "step": 11060 }, { "grad_norm": 0.8532120585441589, "learning_rate": 9.899638173635489e-05, "loss": 0.0835, "step": 11070 }, { "grad_norm": 0.5176616311073303, "learning_rate": 9.899308280684849e-05, "loss": 0.0851, "step": 11080 }, { "grad_norm": 0.8500572443008423, "learning_rate": 9.898977851953222e-05, "loss": 0.0926, "step": 11090 }, { "grad_norm": 0.5701796412467957, "learning_rate": 9.898646887476741e-05, "loss": 0.0962, "step": 11100 }, { "grad_norm": 0.6873341798782349, "learning_rate": 9.898315387291603e-05, "loss": 0.0959, "step": 11110 }, { "grad_norm": 0.4649241864681244, "learning_rate": 9.89798335143406e-05, "loss": 0.0884, "step": 11120 }, { "grad_norm": 0.6931953430175781, "learning_rate": 9.897650779940419e-05, "loss": 0.0954, "step": 11130 }, { "grad_norm": 0.7933159470558167, "learning_rate": 9.897317672847054e-05, "loss": 0.0821, "step": 11140 }, { "grad_norm": 0.6579645276069641, "learning_rate": 9.89698403019039e-05, "loss": 0.1013, "step": 11150 }, { "grad_norm": 0.7465060353279114, "learning_rate": 9.896649852006917e-05, "loss": 0.0907, "step": 11160 }, { "grad_norm": 0.8238161206245422, "learning_rate": 9.896315138333177e-05, "loss": 0.0851, "step": 11170 }, { "grad_norm": 0.6750006079673767, "learning_rate": 9.895979889205774e-05, "loss": 0.0842, "step": 11180 }, { "grad_norm": 0.6704825758934021, "learning_rate": 9.895644104661372e-05, "loss": 0.0843, "step": 11190 }, { "grad_norm": 0.6361594200134277, "learning_rate": 9.895307784736691e-05, "loss": 0.1026, "step": 11200 }, { "grad_norm": 0.7859727144241333, "learning_rate": 9.894970929468512e-05, "loss": 0.0924, "step": 11210 }, { "grad_norm": 0.8973444700241089, "learning_rate": 9.89463353889367e-05, "loss": 0.0922, "step": 11220 }, { "grad_norm": 0.5688135027885437, "learning_rate": 9.894295613049065e-05, "loss": 0.1136, "step": 11230 }, { "grad_norm": 0.6586962342262268, "learning_rate": 9.893957151971649e-05, "loss": 0.0881, "step": 11240 }, { "grad_norm": 0.7283475995063782, "learning_rate": 9.893618155698436e-05, "loss": 0.0865, "step": 11250 }, { "grad_norm": 0.5772260427474976, "learning_rate": 9.8932786242665e-05, "loss": 0.0924, "step": 11260 }, { "grad_norm": 0.40398797392845154, "learning_rate": 9.89293855771297e-05, "loss": 0.1031, "step": 11270 }, { "grad_norm": 0.8881527185440063, "learning_rate": 9.892597956075036e-05, "loss": 0.0958, "step": 11280 }, { "grad_norm": 0.6212051510810852, "learning_rate": 9.892256819389947e-05, "loss": 0.0956, "step": 11290 }, { "grad_norm": 0.6573490500450134, "learning_rate": 9.891915147695006e-05, "loss": 0.0925, "step": 11300 }, { "grad_norm": 0.6627291440963745, "learning_rate": 9.891572941027577e-05, "loss": 0.1099, "step": 11310 }, { "grad_norm": 0.6178131699562073, "learning_rate": 9.89123019942509e-05, "loss": 0.0997, "step": 11320 }, { "grad_norm": 0.8571581244468689, "learning_rate": 9.89088692292502e-05, "loss": 0.0881, "step": 11330 }, { "grad_norm": 0.8677610754966736, "learning_rate": 9.89054311156491e-05, "loss": 0.082, "step": 11340 }, { "grad_norm": 0.6551342606544495, "learning_rate": 9.890198765382357e-05, "loss": 0.1191, "step": 11350 }, { "grad_norm": 0.729141891002655, "learning_rate": 9.889853884415021e-05, "loss": 0.0882, "step": 11360 }, { "grad_norm": 0.6687241196632385, "learning_rate": 9.889508468700614e-05, "loss": 0.0727, "step": 11370 }, { "grad_norm": 0.46802884340286255, "learning_rate": 9.889162518276915e-05, "loss": 0.0798, "step": 11380 }, { "grad_norm": 0.5036545395851135, "learning_rate": 9.888816033181752e-05, "loss": 0.0842, "step": 11390 }, { "grad_norm": 0.5646862983703613, "learning_rate": 9.888469013453018e-05, "loss": 0.1022, "step": 11400 }, { "grad_norm": 1.0018489360809326, "learning_rate": 9.888121459128663e-05, "loss": 0.0855, "step": 11410 }, { "grad_norm": 0.5454612374305725, "learning_rate": 9.887773370246693e-05, "loss": 0.101, "step": 11420 }, { "grad_norm": 0.7058961987495422, "learning_rate": 9.887424746845177e-05, "loss": 0.0939, "step": 11430 }, { "grad_norm": 0.9560791850090027, "learning_rate": 9.887075588962239e-05, "loss": 0.09, "step": 11440 }, { "grad_norm": 0.5181671380996704, "learning_rate": 9.88672589663606e-05, "loss": 0.0928, "step": 11450 }, { "grad_norm": 0.41570422053337097, "learning_rate": 9.886375669904886e-05, "loss": 0.091, "step": 11460 }, { "grad_norm": 0.6726871132850647, "learning_rate": 9.886024908807014e-05, "loss": 0.0823, "step": 11470 }, { "grad_norm": 0.9765148758888245, "learning_rate": 9.885673613380806e-05, "loss": 0.0916, "step": 11480 }, { "grad_norm": 0.5604873895645142, "learning_rate": 9.885321783664676e-05, "loss": 0.075, "step": 11490 }, { "grad_norm": 0.6934804916381836, "learning_rate": 9.884969419697101e-05, "loss": 0.0804, "step": 11500 }, { "grad_norm": 0.5591869950294495, "learning_rate": 9.884616521516614e-05, "loss": 0.0908, "step": 11510 }, { "grad_norm": 0.5586031675338745, "learning_rate": 9.88426308916181e-05, "loss": 0.0865, "step": 11520 }, { "grad_norm": 0.6591296792030334, "learning_rate": 9.883909122671335e-05, "loss": 0.078, "step": 11530 }, { "grad_norm": 0.7412693500518799, "learning_rate": 9.883554622083904e-05, "loss": 0.0864, "step": 11540 }, { "grad_norm": 0.8586706519126892, "learning_rate": 9.88319958743828e-05, "loss": 0.0972, "step": 11550 }, { "grad_norm": 0.6081297397613525, "learning_rate": 9.882844018773291e-05, "loss": 0.0912, "step": 11560 }, { "grad_norm": 1.035345435142517, "learning_rate": 9.882487916127823e-05, "loss": 0.0953, "step": 11570 }, { "grad_norm": 0.4671396017074585, "learning_rate": 9.882131279540815e-05, "loss": 0.1042, "step": 11580 }, { "grad_norm": 0.7662436366081238, "learning_rate": 9.881774109051271e-05, "loss": 0.0971, "step": 11590 }, { "grad_norm": 0.6263532042503357, "learning_rate": 9.881416404698252e-05, "loss": 0.0999, "step": 11600 }, { "grad_norm": 0.5430188179016113, "learning_rate": 9.881058166520873e-05, "loss": 0.0916, "step": 11610 }, { "grad_norm": 0.6089164018630981, "learning_rate": 9.880699394558311e-05, "loss": 0.1051, "step": 11620 }, { "grad_norm": 0.6123158931732178, "learning_rate": 9.880340088849801e-05, "loss": 0.0989, "step": 11630 }, { "grad_norm": 1.0459327697753906, "learning_rate": 9.879980249434637e-05, "loss": 0.0865, "step": 11640 }, { "grad_norm": 0.5497130155563354, "learning_rate": 9.879619876352168e-05, "loss": 0.0923, "step": 11650 }, { "grad_norm": 0.5731012225151062, "learning_rate": 9.879258969641809e-05, "loss": 0.0878, "step": 11660 }, { "grad_norm": 0.5781258940696716, "learning_rate": 9.878897529343023e-05, "loss": 0.0842, "step": 11670 }, { "grad_norm": 0.6935903429985046, "learning_rate": 9.878535555495338e-05, "loss": 0.0921, "step": 11680 }, { "grad_norm": 0.8796342611312866, "learning_rate": 9.87817304813834e-05, "loss": 0.0996, "step": 11690 }, { "grad_norm": 0.7329862713813782, "learning_rate": 9.877810007311671e-05, "loss": 0.0899, "step": 11700 }, { "grad_norm": 0.6817374229431152, "learning_rate": 9.877446433055035e-05, "loss": 0.1016, "step": 11710 }, { "grad_norm": 0.6884030103683472, "learning_rate": 9.877082325408191e-05, "loss": 0.0947, "step": 11720 }, { "grad_norm": 0.5560138821601868, "learning_rate": 9.876717684410954e-05, "loss": 0.0798, "step": 11730 }, { "grad_norm": 0.6062131524085999, "learning_rate": 9.876352510103204e-05, "loss": 0.082, "step": 11740 }, { "grad_norm": 0.7131356000900269, "learning_rate": 9.875986802524875e-05, "loss": 0.0737, "step": 11750 }, { "grad_norm": 0.5070474743843079, "learning_rate": 9.87562056171596e-05, "loss": 0.0756, "step": 11760 }, { "grad_norm": 0.5189292430877686, "learning_rate": 9.875253787716511e-05, "loss": 0.0972, "step": 11770 }, { "grad_norm": 0.6242278814315796, "learning_rate": 9.874886480566637e-05, "loss": 0.0899, "step": 11780 }, { "grad_norm": 0.5852782130241394, "learning_rate": 9.874518640306507e-05, "loss": 0.0786, "step": 11790 }, { "grad_norm": 0.6524472236633301, "learning_rate": 9.874150266976347e-05, "loss": 0.0989, "step": 11800 }, { "grad_norm": 0.5336146354675293, "learning_rate": 9.873781360616443e-05, "loss": 0.1081, "step": 11810 }, { "grad_norm": 1.1310368776321411, "learning_rate": 9.873411921267137e-05, "loss": 0.0908, "step": 11820 }, { "grad_norm": 0.716774046421051, "learning_rate": 9.873041948968829e-05, "loss": 0.0966, "step": 11830 }, { "grad_norm": 0.5815293788909912, "learning_rate": 9.872671443761981e-05, "loss": 0.085, "step": 11840 }, { "grad_norm": 0.6139611005783081, "learning_rate": 9.872300405687109e-05, "loss": 0.0715, "step": 11850 }, { "grad_norm": 0.6435196995735168, "learning_rate": 9.871928834784792e-05, "loss": 0.0884, "step": 11860 }, { "grad_norm": 0.45932647585868835, "learning_rate": 9.871556731095661e-05, "loss": 0.0953, "step": 11870 }, { "grad_norm": 0.7741960287094116, "learning_rate": 9.871184094660411e-05, "loss": 0.0945, "step": 11880 }, { "grad_norm": 0.5976068377494812, "learning_rate": 9.870810925519791e-05, "loss": 0.0907, "step": 11890 }, { "grad_norm": 0.6548317670822144, "learning_rate": 9.870437223714612e-05, "loss": 0.0854, "step": 11900 }, { "grad_norm": 0.5982529520988464, "learning_rate": 9.87006298928574e-05, "loss": 0.0854, "step": 11910 }, { "grad_norm": 0.4047483503818512, "learning_rate": 9.869688222274103e-05, "loss": 0.0753, "step": 11920 }, { "grad_norm": 0.603080153465271, "learning_rate": 9.869312922720681e-05, "loss": 0.0934, "step": 11930 }, { "grad_norm": 0.5177252888679504, "learning_rate": 9.868937090666521e-05, "loss": 0.0815, "step": 11940 }, { "grad_norm": 0.37943923473358154, "learning_rate": 9.86856072615272e-05, "loss": 0.0733, "step": 11950 }, { "grad_norm": 0.6161454319953918, "learning_rate": 9.868183829220438e-05, "loss": 0.0855, "step": 11960 }, { "grad_norm": 0.7791390419006348, "learning_rate": 9.867806399910893e-05, "loss": 0.0801, "step": 11970 }, { "grad_norm": 0.8104416131973267, "learning_rate": 9.867428438265356e-05, "loss": 0.0857, "step": 11980 }, { "grad_norm": 0.5526025891304016, "learning_rate": 9.867049944325165e-05, "loss": 0.0955, "step": 11990 }, { "grad_norm": 0.5797675251960754, "learning_rate": 9.86667091813171e-05, "loss": 0.102, "step": 12000 }, { "grad_norm": 0.6660271286964417, "learning_rate": 9.866291359726438e-05, "loss": 0.0889, "step": 12010 }, { "grad_norm": 0.6610478162765503, "learning_rate": 9.865911269150861e-05, "loss": 0.093, "step": 12020 }, { "grad_norm": 0.5969988107681274, "learning_rate": 9.865530646446544e-05, "loss": 0.0746, "step": 12030 }, { "grad_norm": 0.6098376512527466, "learning_rate": 9.86514949165511e-05, "loss": 0.079, "step": 12040 }, { "grad_norm": 0.5625382661819458, "learning_rate": 9.864767804818243e-05, "loss": 0.0989, "step": 12050 }, { "grad_norm": 0.6006349325180054, "learning_rate": 9.86438558597768e-05, "loss": 0.0959, "step": 12060 }, { "grad_norm": 0.40218138694763184, "learning_rate": 9.864002835175225e-05, "loss": 0.093, "step": 12070 }, { "grad_norm": 0.6550064086914062, "learning_rate": 9.863619552452734e-05, "loss": 0.0725, "step": 12080 }, { "grad_norm": 0.6242501735687256, "learning_rate": 9.863235737852119e-05, "loss": 0.0595, "step": 12090 }, { "grad_norm": 0.9424892067909241, "learning_rate": 9.862851391415356e-05, "loss": 0.0803, "step": 12100 }, { "grad_norm": 0.5324152112007141, "learning_rate": 9.862466513184477e-05, "loss": 0.087, "step": 12110 }, { "grad_norm": 0.6059129238128662, "learning_rate": 9.86208110320157e-05, "loss": 0.0802, "step": 12120 }, { "grad_norm": 0.9259104132652283, "learning_rate": 9.861695161508784e-05, "loss": 0.0825, "step": 12130 }, { "grad_norm": 0.7278442978858948, "learning_rate": 9.861308688148324e-05, "loss": 0.0826, "step": 12140 }, { "grad_norm": 0.6459093689918518, "learning_rate": 9.860921683162455e-05, "loss": 0.0892, "step": 12150 }, { "grad_norm": 0.7398424744606018, "learning_rate": 9.860534146593499e-05, "loss": 0.1034, "step": 12160 }, { "grad_norm": 0.6791397929191589, "learning_rate": 9.860146078483836e-05, "loss": 0.0907, "step": 12170 }, { "grad_norm": 0.7969271540641785, "learning_rate": 9.859757478875905e-05, "loss": 0.0855, "step": 12180 }, { "grad_norm": 0.6912099719047546, "learning_rate": 9.859368347812204e-05, "loss": 0.0878, "step": 12190 }, { "grad_norm": 0.7312430143356323, "learning_rate": 9.858978685335285e-05, "loss": 0.0859, "step": 12200 }, { "grad_norm": 0.6959239840507507, "learning_rate": 9.858588491487763e-05, "loss": 0.1016, "step": 12210 }, { "grad_norm": 0.49293628334999084, "learning_rate": 9.858197766312308e-05, "loss": 0.0768, "step": 12220 }, { "grad_norm": 0.612618625164032, "learning_rate": 9.857806509851649e-05, "loss": 0.0859, "step": 12230 }, { "grad_norm": 0.489947110414505, "learning_rate": 9.857414722148574e-05, "loss": 0.0881, "step": 12240 }, { "grad_norm": 0.540193498134613, "learning_rate": 9.857022403245928e-05, "loss": 0.095, "step": 12250 }, { "grad_norm": 0.6411373019218445, "learning_rate": 9.856629553186615e-05, "loss": 0.0965, "step": 12260 }, { "grad_norm": 0.5284746885299683, "learning_rate": 9.856236172013595e-05, "loss": 0.0836, "step": 12270 }, { "grad_norm": 0.5400474667549133, "learning_rate": 9.85584225976989e-05, "loss": 0.0846, "step": 12280 }, { "grad_norm": 1.2156387567520142, "learning_rate": 9.855447816498575e-05, "loss": 0.1025, "step": 12290 }, { "grad_norm": 0.7346674799919128, "learning_rate": 9.855052842242787e-05, "loss": 0.0805, "step": 12300 }, { "grad_norm": 0.5978521108627319, "learning_rate": 9.85465733704572e-05, "loss": 0.0879, "step": 12310 }, { "grad_norm": 0.6132931113243103, "learning_rate": 9.854261300950624e-05, "loss": 0.1112, "step": 12320 }, { "grad_norm": 0.5627762079238892, "learning_rate": 9.853864734000813e-05, "loss": 0.0864, "step": 12330 }, { "grad_norm": 0.7303783297538757, "learning_rate": 9.85346763623965e-05, "loss": 0.0994, "step": 12340 }, { "grad_norm": 0.791337788105011, "learning_rate": 9.853070007710564e-05, "loss": 0.083, "step": 12350 }, { "grad_norm": 0.7115092873573303, "learning_rate": 9.85267184845704e-05, "loss": 0.0723, "step": 12360 }, { "grad_norm": 0.6893014311790466, "learning_rate": 9.852273158522616e-05, "loss": 0.094, "step": 12370 }, { "grad_norm": 0.5205362439155579, "learning_rate": 9.851873937950896e-05, "loss": 0.081, "step": 12380 }, { "grad_norm": 0.6678147912025452, "learning_rate": 9.851474186785537e-05, "loss": 0.0869, "step": 12390 }, { "grad_norm": 0.7992581129074097, "learning_rate": 9.851073905070254e-05, "loss": 0.0915, "step": 12400 }, { "grad_norm": 0.8437297344207764, "learning_rate": 9.850673092848824e-05, "loss": 0.1125, "step": 12410 }, { "grad_norm": 0.653643012046814, "learning_rate": 9.850271750165077e-05, "loss": 0.0727, "step": 12420 }, { "grad_norm": 0.7962107062339783, "learning_rate": 9.849869877062902e-05, "loss": 0.0999, "step": 12430 }, { "grad_norm": 0.5988895297050476, "learning_rate": 9.849467473586252e-05, "loss": 0.0892, "step": 12440 }, { "grad_norm": 0.5859716534614563, "learning_rate": 9.849064539779127e-05, "loss": 0.0816, "step": 12450 }, { "grad_norm": 0.5836171507835388, "learning_rate": 9.848661075685594e-05, "loss": 0.0903, "step": 12460 }, { "grad_norm": 1.2143537998199463, "learning_rate": 9.848257081349778e-05, "loss": 0.0971, "step": 12470 }, { "grad_norm": 0.7629567980766296, "learning_rate": 9.847852556815856e-05, "loss": 0.0875, "step": 12480 }, { "grad_norm": 0.6685308218002319, "learning_rate": 9.847447502128067e-05, "loss": 0.1017, "step": 12490 }, { "grad_norm": 0.6864867210388184, "learning_rate": 9.847041917330708e-05, "loss": 0.0851, "step": 12500 }, { "grad_norm": 0.46510186791419983, "learning_rate": 9.846635802468132e-05, "loss": 0.0912, "step": 12510 }, { "grad_norm": 0.6023244857788086, "learning_rate": 9.84622915758475e-05, "loss": 0.0887, "step": 12520 }, { "grad_norm": 0.5804929733276367, "learning_rate": 9.845821982725034e-05, "loss": 0.0803, "step": 12530 }, { "grad_norm": 0.6531669497489929, "learning_rate": 9.845414277933514e-05, "loss": 0.1005, "step": 12540 }, { "grad_norm": 0.5585441589355469, "learning_rate": 9.845006043254771e-05, "loss": 0.093, "step": 12550 }, { "grad_norm": 0.7379462718963623, "learning_rate": 9.844597278733451e-05, "loss": 0.0949, "step": 12560 }, { "grad_norm": 0.604551374912262, "learning_rate": 9.844187984414259e-05, "loss": 0.0844, "step": 12570 }, { "grad_norm": 0.6405412554740906, "learning_rate": 9.84377816034195e-05, "loss": 0.0842, "step": 12580 }, { "grad_norm": 0.7747296690940857, "learning_rate": 9.843367806561345e-05, "loss": 0.0945, "step": 12590 }, { "grad_norm": 0.6387954950332642, "learning_rate": 9.842956923117317e-05, "loss": 0.0775, "step": 12600 }, { "grad_norm": 0.6272313594818115, "learning_rate": 9.842545510054802e-05, "loss": 0.0747, "step": 12610 }, { "grad_norm": 0.6728931069374084, "learning_rate": 9.842133567418792e-05, "loss": 0.0768, "step": 12620 }, { "grad_norm": 0.6115803718566895, "learning_rate": 9.841721095254333e-05, "loss": 0.068, "step": 12630 }, { "grad_norm": 0.60696941614151, "learning_rate": 9.841308093606537e-05, "loss": 0.0806, "step": 12640 }, { "grad_norm": 0.6848500967025757, "learning_rate": 9.840894562520565e-05, "loss": 0.0787, "step": 12650 }, { "grad_norm": 0.765125572681427, "learning_rate": 9.840480502041642e-05, "loss": 0.1007, "step": 12660 }, { "grad_norm": 0.5781741142272949, "learning_rate": 9.840065912215049e-05, "loss": 0.083, "step": 12670 }, { "grad_norm": 0.553991973400116, "learning_rate": 9.839650793086124e-05, "loss": 0.1009, "step": 12680 }, { "grad_norm": 0.4382828176021576, "learning_rate": 9.839235144700265e-05, "loss": 0.0732, "step": 12690 }, { "grad_norm": 0.5213690400123596, "learning_rate": 9.838818967102926e-05, "loss": 0.0799, "step": 12700 }, { "grad_norm": 0.5668008327484131, "learning_rate": 9.83840226033962e-05, "loss": 0.0879, "step": 12710 }, { "grad_norm": 0.5534228682518005, "learning_rate": 9.837985024455918e-05, "loss": 0.0771, "step": 12720 }, { "grad_norm": 0.4276198744773865, "learning_rate": 9.837567259497447e-05, "loss": 0.0738, "step": 12730 }, { "grad_norm": 0.5107090473175049, "learning_rate": 9.837148965509894e-05, "loss": 0.0755, "step": 12740 }, { "grad_norm": 0.5644952058792114, "learning_rate": 9.836730142539001e-05, "loss": 0.0715, "step": 12750 }, { "grad_norm": 0.6281929016113281, "learning_rate": 9.836310790630574e-05, "loss": 0.0992, "step": 12760 }, { "grad_norm": 0.5129491090774536, "learning_rate": 9.83589090983047e-05, "loss": 0.0954, "step": 12770 }, { "grad_norm": 0.637671709060669, "learning_rate": 9.835470500184605e-05, "loss": 0.0823, "step": 12780 }, { "grad_norm": 0.6283777356147766, "learning_rate": 9.835049561738957e-05, "loss": 0.0843, "step": 12790 }, { "grad_norm": 0.6402627825737, "learning_rate": 9.834628094539558e-05, "loss": 0.0813, "step": 12800 }, { "grad_norm": 0.7106264233589172, "learning_rate": 9.834206098632499e-05, "loss": 0.095, "step": 12810 }, { "grad_norm": 0.5908417701721191, "learning_rate": 9.833783574063931e-05, "loss": 0.0962, "step": 12820 }, { "grad_norm": 0.6732028722763062, "learning_rate": 9.833360520880058e-05, "loss": 0.1024, "step": 12830 }, { "grad_norm": 0.5504224896430969, "learning_rate": 9.832936939127144e-05, "loss": 0.0815, "step": 12840 }, { "grad_norm": 0.6568690538406372, "learning_rate": 9.832512828851515e-05, "loss": 0.0979, "step": 12850 }, { "grad_norm": 0.5526666641235352, "learning_rate": 9.832088190099546e-05, "loss": 0.0841, "step": 12860 }, { "grad_norm": 0.6130136847496033, "learning_rate": 9.831663022917679e-05, "loss": 0.0851, "step": 12870 }, { "grad_norm": 0.640504777431488, "learning_rate": 9.831237327352407e-05, "loss": 0.0796, "step": 12880 }, { "grad_norm": 0.612348735332489, "learning_rate": 9.830811103450286e-05, "loss": 0.0806, "step": 12890 }, { "grad_norm": 0.5407823920249939, "learning_rate": 9.830384351257924e-05, "loss": 0.0848, "step": 12900 }, { "grad_norm": 0.793494462966919, "learning_rate": 9.829957070821993e-05, "loss": 0.0865, "step": 12910 }, { "grad_norm": 0.6452816128730774, "learning_rate": 9.829529262189218e-05, "loss": 0.0709, "step": 12920 }, { "grad_norm": 0.5692020654678345, "learning_rate": 9.829100925406385e-05, "loss": 0.082, "step": 12930 }, { "grad_norm": 0.6717889904975891, "learning_rate": 9.828672060520333e-05, "loss": 0.0886, "step": 12940 }, { "grad_norm": 0.6801457405090332, "learning_rate": 9.828242667577966e-05, "loss": 0.0669, "step": 12950 }, { "grad_norm": 0.595852792263031, "learning_rate": 9.82781274662624e-05, "loss": 0.0832, "step": 12960 }, { "grad_norm": 0.624721109867096, "learning_rate": 9.82738229771217e-05, "loss": 0.0736, "step": 12970 }, { "grad_norm": 0.5195843577384949, "learning_rate": 9.826951320882829e-05, "loss": 0.0747, "step": 12980 }, { "grad_norm": 1.0157188177108765, "learning_rate": 9.826519816185351e-05, "loss": 0.08, "step": 12990 }, { "grad_norm": 0.6593079566955566, "learning_rate": 9.826087783666921e-05, "loss": 0.0798, "step": 13000 }, { "grad_norm": 0.6818521618843079, "learning_rate": 9.825655223374787e-05, "loss": 0.097, "step": 13010 }, { "grad_norm": 0.7750303149223328, "learning_rate": 9.825222135356253e-05, "loss": 0.1027, "step": 13020 }, { "grad_norm": 0.5795605182647705, "learning_rate": 9.82478851965868e-05, "loss": 0.1015, "step": 13030 }, { "grad_norm": 0.5676087737083435, "learning_rate": 9.82435437632949e-05, "loss": 0.0753, "step": 13040 }, { "grad_norm": 0.7243655920028687, "learning_rate": 9.823919705416158e-05, "loss": 0.0883, "step": 13050 }, { "grad_norm": 0.8391656279563904, "learning_rate": 9.82348450696622e-05, "loss": 0.0949, "step": 13060 }, { "grad_norm": 0.6524685621261597, "learning_rate": 9.823048781027268e-05, "loss": 0.098, "step": 13070 }, { "grad_norm": 0.5331210494041443, "learning_rate": 9.822612527646953e-05, "loss": 0.0774, "step": 13080 }, { "grad_norm": 0.6151438355445862, "learning_rate": 9.822175746872984e-05, "loss": 0.0935, "step": 13090 }, { "grad_norm": 0.5895125269889832, "learning_rate": 9.821738438753123e-05, "loss": 0.086, "step": 13100 }, { "grad_norm": 0.592450737953186, "learning_rate": 9.821300603335196e-05, "loss": 0.0876, "step": 13110 }, { "grad_norm": 0.6179054379463196, "learning_rate": 9.820862240667085e-05, "loss": 0.0968, "step": 13120 }, { "grad_norm": 0.6349558234214783, "learning_rate": 9.820423350796726e-05, "loss": 0.0807, "step": 13130 }, { "grad_norm": 0.7824263572692871, "learning_rate": 9.819983933772118e-05, "loss": 0.0918, "step": 13140 }, { "grad_norm": 0.7362442016601562, "learning_rate": 9.819543989641314e-05, "loss": 0.085, "step": 13150 }, { "grad_norm": 0.49806728959083557, "learning_rate": 9.819103518452423e-05, "loss": 0.0786, "step": 13160 }, { "grad_norm": 0.7249966859817505, "learning_rate": 9.818662520253618e-05, "loss": 0.0877, "step": 13170 }, { "grad_norm": 0.42193031311035156, "learning_rate": 9.818220995093126e-05, "loss": 0.0763, "step": 13180 }, { "grad_norm": 1.0514620542526245, "learning_rate": 9.817778943019228e-05, "loss": 0.0807, "step": 13190 }, { "grad_norm": 0.7432230114936829, "learning_rate": 9.81733636408027e-05, "loss": 0.0877, "step": 13200 }, { "grad_norm": 0.6951989531517029, "learning_rate": 9.816893258324649e-05, "loss": 0.0716, "step": 13210 }, { "grad_norm": 0.6368913054466248, "learning_rate": 9.816449625800823e-05, "loss": 0.0735, "step": 13220 }, { "grad_norm": 0.5688896775245667, "learning_rate": 9.816005466557308e-05, "loss": 0.0831, "step": 13230 }, { "grad_norm": 0.9792847037315369, "learning_rate": 9.815560780642674e-05, "loss": 0.1082, "step": 13240 }, { "grad_norm": 0.5327269434928894, "learning_rate": 9.815115568105555e-05, "loss": 0.0768, "step": 13250 }, { "grad_norm": 0.6527314186096191, "learning_rate": 9.814669828994638e-05, "loss": 0.091, "step": 13260 }, { "grad_norm": 0.4764961302280426, "learning_rate": 9.814223563358665e-05, "loss": 0.0821, "step": 13270 }, { "grad_norm": 0.6853342056274414, "learning_rate": 9.813776771246443e-05, "loss": 0.0802, "step": 13280 }, { "grad_norm": 0.4622820317745209, "learning_rate": 9.813329452706829e-05, "loss": 0.0796, "step": 13290 }, { "grad_norm": 0.7285516858100891, "learning_rate": 9.812881607788744e-05, "loss": 0.0752, "step": 13300 }, { "grad_norm": 0.40452224016189575, "learning_rate": 9.812433236541163e-05, "loss": 0.0858, "step": 13310 }, { "grad_norm": 0.7357060313224792, "learning_rate": 9.811984339013116e-05, "loss": 0.0954, "step": 13320 }, { "grad_norm": 0.6772493124008179, "learning_rate": 9.811534915253698e-05, "loss": 0.0777, "step": 13330 }, { "grad_norm": 0.9262349605560303, "learning_rate": 9.811084965312056e-05, "loss": 0.1012, "step": 13340 }, { "grad_norm": 0.9320617914199829, "learning_rate": 9.810634489237396e-05, "loss": 0.0838, "step": 13350 }, { "grad_norm": 0.6888753771781921, "learning_rate": 9.81018348707898e-05, "loss": 0.1049, "step": 13360 }, { "grad_norm": 1.1437530517578125, "learning_rate": 9.809731958886131e-05, "loss": 0.0907, "step": 13370 }, { "grad_norm": 0.8316532969474792, "learning_rate": 9.809279904708224e-05, "loss": 0.1076, "step": 13380 }, { "grad_norm": 0.6488223075866699, "learning_rate": 9.808827324594699e-05, "loss": 0.0996, "step": 13390 }, { "grad_norm": 0.4975990951061249, "learning_rate": 9.808374218595046e-05, "loss": 0.1066, "step": 13400 }, { "grad_norm": 0.6951090693473816, "learning_rate": 9.80792058675882e-05, "loss": 0.083, "step": 13410 }, { "grad_norm": 0.6250736117362976, "learning_rate": 9.807466429135627e-05, "loss": 0.088, "step": 13420 }, { "grad_norm": 0.5039610862731934, "learning_rate": 9.807011745775132e-05, "loss": 0.0834, "step": 13430 }, { "grad_norm": 0.5149434804916382, "learning_rate": 9.806556536727061e-05, "loss": 0.0792, "step": 13440 }, { "grad_norm": 0.5813740491867065, "learning_rate": 9.806100802041193e-05, "loss": 0.1, "step": 13450 }, { "grad_norm": 0.45179903507232666, "learning_rate": 9.805644541767368e-05, "loss": 0.09, "step": 13460 }, { "grad_norm": 0.6276927590370178, "learning_rate": 9.805187755955478e-05, "loss": 0.1026, "step": 13470 }, { "grad_norm": 0.5209026336669922, "learning_rate": 9.804730444655483e-05, "loss": 0.0844, "step": 13480 }, { "grad_norm": 0.6999006867408752, "learning_rate": 9.804272607917388e-05, "loss": 0.0983, "step": 13490 }, { "grad_norm": 0.74775230884552, "learning_rate": 9.803814245791265e-05, "loss": 0.0804, "step": 13500 }, { "grad_norm": 0.6172882914543152, "learning_rate": 9.803355358327239e-05, "loss": 0.0972, "step": 13510 }, { "grad_norm": 0.5789906978607178, "learning_rate": 9.802895945575492e-05, "loss": 0.0898, "step": 13520 }, { "grad_norm": 0.688332736492157, "learning_rate": 9.802436007586266e-05, "loss": 0.1084, "step": 13530 }, { "grad_norm": 0.777015745639801, "learning_rate": 9.801975544409858e-05, "loss": 0.0831, "step": 13540 }, { "grad_norm": 0.6939036846160889, "learning_rate": 9.801514556096625e-05, "loss": 0.0921, "step": 13550 }, { "grad_norm": 0.6849474310874939, "learning_rate": 9.801053042696977e-05, "loss": 0.0951, "step": 13560 }, { "grad_norm": 0.508050262928009, "learning_rate": 9.800591004261388e-05, "loss": 0.0786, "step": 13570 }, { "grad_norm": 0.7749068737030029, "learning_rate": 9.800128440840385e-05, "loss": 0.0893, "step": 13580 }, { "grad_norm": 0.8280090093612671, "learning_rate": 9.799665352484552e-05, "loss": 0.1062, "step": 13590 }, { "grad_norm": 0.6985770463943481, "learning_rate": 9.799201739244532e-05, "loss": 0.109, "step": 13600 }, { "grad_norm": 0.6614087224006653, "learning_rate": 9.798737601171025e-05, "loss": 0.0855, "step": 13610 }, { "grad_norm": 0.8155915141105652, "learning_rate": 9.79827293831479e-05, "loss": 0.1036, "step": 13620 }, { "grad_norm": 0.6062188148498535, "learning_rate": 9.797807750726638e-05, "loss": 0.0792, "step": 13630 }, { "grad_norm": 0.5032339096069336, "learning_rate": 9.797342038457446e-05, "loss": 0.0842, "step": 13640 }, { "grad_norm": 0.6383780241012573, "learning_rate": 9.796875801558141e-05, "loss": 0.0957, "step": 13650 }, { "grad_norm": 0.5716334581375122, "learning_rate": 9.79640904007971e-05, "loss": 0.1004, "step": 13660 }, { "grad_norm": 0.6799912452697754, "learning_rate": 9.795941754073199e-05, "loss": 0.0788, "step": 13670 }, { "grad_norm": 0.4188758134841919, "learning_rate": 9.795473943589705e-05, "loss": 0.0727, "step": 13680 }, { "grad_norm": 0.5747365951538086, "learning_rate": 9.795005608680394e-05, "loss": 0.0851, "step": 13690 }, { "grad_norm": 0.8087440729141235, "learning_rate": 9.794536749396477e-05, "loss": 0.1032, "step": 13700 }, { "grad_norm": 0.69086092710495, "learning_rate": 9.79406736578923e-05, "loss": 0.0888, "step": 13710 }, { "grad_norm": 0.6207590103149414, "learning_rate": 9.793597457909984e-05, "loss": 0.1001, "step": 13720 }, { "grad_norm": 0.4396333396434784, "learning_rate": 9.793127025810127e-05, "loss": 0.0814, "step": 13730 }, { "grad_norm": 0.49114954471588135, "learning_rate": 9.792656069541104e-05, "loss": 0.0792, "step": 13740 }, { "grad_norm": 0.6639207005500793, "learning_rate": 9.79218458915442e-05, "loss": 0.0838, "step": 13750 }, { "grad_norm": 0.5593693852424622, "learning_rate": 9.791712584701634e-05, "loss": 0.0946, "step": 13760 }, { "grad_norm": 0.47932639718055725, "learning_rate": 9.791240056234364e-05, "loss": 0.0811, "step": 13770 }, { "grad_norm": 0.5084441900253296, "learning_rate": 9.790767003804283e-05, "loss": 0.085, "step": 13780 }, { "grad_norm": 0.4902383089065552, "learning_rate": 9.790293427463126e-05, "loss": 0.0704, "step": 13790 }, { "grad_norm": 0.6741477847099304, "learning_rate": 9.789819327262684e-05, "loss": 0.0691, "step": 13800 }, { "grad_norm": 0.6043632626533508, "learning_rate": 9.7893447032548e-05, "loss": 0.0883, "step": 13810 }, { "grad_norm": 0.6760991215705872, "learning_rate": 9.78886955549138e-05, "loss": 0.0897, "step": 13820 }, { "grad_norm": 0.5125429630279541, "learning_rate": 9.788393884024387e-05, "loss": 0.0885, "step": 13830 }, { "grad_norm": 0.5873991250991821, "learning_rate": 9.787917688905836e-05, "loss": 0.0946, "step": 13840 }, { "grad_norm": 0.6140190958976746, "learning_rate": 9.787440970187807e-05, "loss": 0.0736, "step": 13850 }, { "grad_norm": 0.740746796131134, "learning_rate": 9.786963727922429e-05, "loss": 0.0884, "step": 13860 }, { "grad_norm": 1.496469259262085, "learning_rate": 9.786485962161897e-05, "loss": 0.0837, "step": 13870 }, { "grad_norm": 0.6970010995864868, "learning_rate": 9.786007672958455e-05, "loss": 0.0755, "step": 13880 }, { "grad_norm": 0.6417742967605591, "learning_rate": 9.78552886036441e-05, "loss": 0.086, "step": 13890 }, { "grad_norm": 0.6106270551681519, "learning_rate": 9.785049524432124e-05, "loss": 0.0708, "step": 13900 }, { "grad_norm": 0.729512631893158, "learning_rate": 9.784569665214016e-05, "loss": 0.0971, "step": 13910 }, { "grad_norm": 0.8528873324394226, "learning_rate": 9.784089282762563e-05, "loss": 0.0878, "step": 13920 }, { "grad_norm": 0.6591704487800598, "learning_rate": 9.7836083771303e-05, "loss": 0.0935, "step": 13930 }, { "grad_norm": 0.7043913006782532, "learning_rate": 9.783126948369817e-05, "loss": 0.0938, "step": 13940 }, { "grad_norm": 0.4686422049999237, "learning_rate": 9.78264499653376e-05, "loss": 0.0772, "step": 13950 }, { "grad_norm": 0.6523698568344116, "learning_rate": 9.782162521674838e-05, "loss": 0.0845, "step": 13960 }, { "grad_norm": 0.6307424306869507, "learning_rate": 9.781679523845812e-05, "loss": 0.1018, "step": 13970 }, { "grad_norm": 0.9975139498710632, "learning_rate": 9.781196003099502e-05, "loss": 0.0898, "step": 13980 }, { "grad_norm": 0.48900189995765686, "learning_rate": 9.780711959488786e-05, "loss": 0.0669, "step": 13990 }, { "grad_norm": 0.46409720182418823, "learning_rate": 9.780227393066599e-05, "loss": 0.0755, "step": 14000 }, { "grad_norm": 0.583935558795929, "learning_rate": 9.77974230388593e-05, "loss": 0.0863, "step": 14010 }, { "grad_norm": 0.5385763645172119, "learning_rate": 9.779256691999829e-05, "loss": 0.0943, "step": 14020 }, { "grad_norm": 0.794115424156189, "learning_rate": 9.778770557461403e-05, "loss": 0.0951, "step": 14030 }, { "grad_norm": 0.5170246362686157, "learning_rate": 9.778283900323812e-05, "loss": 0.0727, "step": 14040 }, { "grad_norm": 0.8844103813171387, "learning_rate": 9.777796720640277e-05, "loss": 0.0928, "step": 14050 }, { "grad_norm": 0.5596650242805481, "learning_rate": 9.777309018464078e-05, "loss": 0.0692, "step": 14060 }, { "grad_norm": 0.5498842000961304, "learning_rate": 9.776820793848547e-05, "loss": 0.0787, "step": 14070 }, { "grad_norm": 0.6152229309082031, "learning_rate": 9.776332046847075e-05, "loss": 0.0726, "step": 14080 }, { "grad_norm": 0.5296191573143005, "learning_rate": 9.775842777513111e-05, "loss": 0.0845, "step": 14090 }, { "grad_norm": 0.47173207998275757, "learning_rate": 9.775352985900163e-05, "loss": 0.0893, "step": 14100 }, { "grad_norm": 0.8818325400352478, "learning_rate": 9.774862672061791e-05, "loss": 0.0856, "step": 14110 }, { "grad_norm": 0.4128448963165283, "learning_rate": 9.774371836051616e-05, "loss": 0.0894, "step": 14120 }, { "grad_norm": 0.585886538028717, "learning_rate": 9.773880477923315e-05, "loss": 0.0819, "step": 14130 }, { "grad_norm": 0.8150635957717896, "learning_rate": 9.773388597730623e-05, "loss": 0.0879, "step": 14140 }, { "grad_norm": 3.813354730606079, "learning_rate": 9.77289619552733e-05, "loss": 0.1093, "step": 14150 }, { "grad_norm": 0.8009932637214661, "learning_rate": 9.772403271367285e-05, "loss": 0.0981, "step": 14160 }, { "grad_norm": 0.5114272832870483, "learning_rate": 9.771909825304396e-05, "loss": 0.0907, "step": 14170 }, { "grad_norm": 0.4745633602142334, "learning_rate": 9.771415857392619e-05, "loss": 0.089, "step": 14180 }, { "grad_norm": 0.7626971006393433, "learning_rate": 9.770921367685978e-05, "loss": 0.1075, "step": 14190 }, { "grad_norm": 0.6451576948165894, "learning_rate": 9.770426356238551e-05, "loss": 0.0885, "step": 14200 }, { "grad_norm": 0.5462473630905151, "learning_rate": 9.769930823104469e-05, "loss": 0.0927, "step": 14210 }, { "grad_norm": 0.5145328640937805, "learning_rate": 9.769434768337926e-05, "loss": 0.0901, "step": 14220 }, { "grad_norm": 0.5081210136413574, "learning_rate": 9.768938191993164e-05, "loss": 0.0808, "step": 14230 }, { "grad_norm": 0.551776111125946, "learning_rate": 9.768441094124494e-05, "loss": 0.0751, "step": 14240 }, { "grad_norm": 0.5495102405548096, "learning_rate": 9.767943474786275e-05, "loss": 0.0891, "step": 14250 }, { "grad_norm": 0.6209490895271301, "learning_rate": 9.767445334032923e-05, "loss": 0.068, "step": 14260 }, { "grad_norm": 0.9738024473190308, "learning_rate": 9.766946671918919e-05, "loss": 0.0974, "step": 14270 }, { "grad_norm": 0.6228636503219604, "learning_rate": 9.766447488498796e-05, "loss": 0.0746, "step": 14280 }, { "grad_norm": 0.5731285810470581, "learning_rate": 9.765947783827139e-05, "loss": 0.0898, "step": 14290 }, { "grad_norm": 0.5721383094787598, "learning_rate": 9.765447557958599e-05, "loss": 0.0753, "step": 14300 }, { "grad_norm": 0.4923485815525055, "learning_rate": 9.764946810947879e-05, "loss": 0.0703, "step": 14310 }, { "grad_norm": 0.4055038392543793, "learning_rate": 9.764445542849738e-05, "loss": 0.0877, "step": 14320 }, { "grad_norm": 0.5246450901031494, "learning_rate": 9.763943753718998e-05, "loss": 0.0768, "step": 14330 }, { "grad_norm": 0.5268251299858093, "learning_rate": 9.76344144361053e-05, "loss": 0.0799, "step": 14340 }, { "grad_norm": 0.7473240494728088, "learning_rate": 9.762938612579269e-05, "loss": 0.0943, "step": 14350 }, { "grad_norm": 1.1141847372055054, "learning_rate": 9.762435260680202e-05, "loss": 0.0884, "step": 14360 }, { "grad_norm": 0.6969948410987854, "learning_rate": 9.761931387968373e-05, "loss": 0.1004, "step": 14370 }, { "grad_norm": 0.4452889859676361, "learning_rate": 9.76142699449889e-05, "loss": 0.0766, "step": 14380 }, { "grad_norm": 0.4913959503173828, "learning_rate": 9.760922080326908e-05, "loss": 0.0847, "step": 14390 }, { "grad_norm": 0.733067512512207, "learning_rate": 9.760416645507644e-05, "loss": 0.0818, "step": 14400 }, { "grad_norm": 0.599712610244751, "learning_rate": 9.759910690096375e-05, "loss": 0.0803, "step": 14410 }, { "grad_norm": 0.6619315147399902, "learning_rate": 9.759404214148429e-05, "loss": 0.0822, "step": 14420 }, { "grad_norm": 0.5321961641311646, "learning_rate": 9.758897217719191e-05, "loss": 0.0844, "step": 14430 }, { "grad_norm": 0.4542900323867798, "learning_rate": 9.758389700864113e-05, "loss": 0.08, "step": 14440 }, { "grad_norm": 0.9488402605056763, "learning_rate": 9.757881663638688e-05, "loss": 0.0923, "step": 14450 }, { "grad_norm": 0.5132675766944885, "learning_rate": 9.757373106098478e-05, "loss": 0.0866, "step": 14460 }, { "grad_norm": 0.5348666906356812, "learning_rate": 9.756864028299097e-05, "loss": 0.0817, "step": 14470 }, { "grad_norm": 0.5375731587409973, "learning_rate": 9.75635443029622e-05, "loss": 0.0717, "step": 14480 }, { "grad_norm": 0.6187760829925537, "learning_rate": 9.755844312145572e-05, "loss": 0.0808, "step": 14490 }, { "grad_norm": 0.6278170347213745, "learning_rate": 9.755333673902941e-05, "loss": 0.1011, "step": 14500 }, { "grad_norm": 0.523739218711853, "learning_rate": 9.75482251562417e-05, "loss": 0.0768, "step": 14510 }, { "grad_norm": 0.6260690093040466, "learning_rate": 9.754310837365155e-05, "loss": 0.0685, "step": 14520 }, { "grad_norm": 0.8518478274345398, "learning_rate": 9.753798639181856e-05, "loss": 0.0982, "step": 14530 }, { "grad_norm": 0.6926760077476501, "learning_rate": 9.753285921130286e-05, "loss": 0.0641, "step": 14540 }, { "grad_norm": 0.6566014289855957, "learning_rate": 9.752772683266512e-05, "loss": 0.0778, "step": 14550 }, { "grad_norm": 0.8064532279968262, "learning_rate": 9.752258925646665e-05, "loss": 0.081, "step": 14560 }, { "grad_norm": 0.8399253487586975, "learning_rate": 9.751744648326926e-05, "loss": 0.0733, "step": 14570 }, { "grad_norm": 1.2437381744384766, "learning_rate": 9.751229851363536e-05, "loss": 0.0869, "step": 14580 }, { "grad_norm": 0.7285233736038208, "learning_rate": 9.750714534812793e-05, "loss": 0.1003, "step": 14590 }, { "grad_norm": 0.9216816425323486, "learning_rate": 9.750198698731053e-05, "loss": 0.0962, "step": 14600 }, { "grad_norm": 0.8692930340766907, "learning_rate": 9.749682343174722e-05, "loss": 0.0866, "step": 14610 }, { "grad_norm": 0.48563826084136963, "learning_rate": 9.749165468200272e-05, "loss": 0.0829, "step": 14620 }, { "grad_norm": 0.6495055556297302, "learning_rate": 9.748648073864229e-05, "loss": 0.0809, "step": 14630 }, { "grad_norm": 0.5537834763526917, "learning_rate": 9.748130160223168e-05, "loss": 0.0734, "step": 14640 }, { "grad_norm": 1.0038461685180664, "learning_rate": 9.747611727333734e-05, "loss": 0.0947, "step": 14650 }, { "grad_norm": 0.7438955903053284, "learning_rate": 9.74709277525262e-05, "loss": 0.0922, "step": 14660 }, { "grad_norm": 0.5169081091880798, "learning_rate": 9.746573304036576e-05, "loss": 0.0901, "step": 14670 }, { "grad_norm": 0.7759031057357788, "learning_rate": 9.746053313742412e-05, "loss": 0.075, "step": 14680 }, { "grad_norm": 0.6889575123786926, "learning_rate": 9.745532804426994e-05, "loss": 0.0912, "step": 14690 }, { "grad_norm": 0.54827481508255, "learning_rate": 9.745011776147242e-05, "loss": 0.0763, "step": 14700 }, { "grad_norm": 0.9758272767066956, "learning_rate": 9.744490228960138e-05, "loss": 0.0933, "step": 14710 }, { "grad_norm": 0.5519786477088928, "learning_rate": 9.743968162922713e-05, "loss": 0.0764, "step": 14720 }, { "grad_norm": 0.5062392950057983, "learning_rate": 9.743445578092064e-05, "loss": 0.097, "step": 14730 }, { "grad_norm": 0.45411643385887146, "learning_rate": 9.742922474525338e-05, "loss": 0.094, "step": 14740 }, { "grad_norm": 0.7346479296684265, "learning_rate": 9.742398852279741e-05, "loss": 0.0724, "step": 14750 }, { "grad_norm": 0.971693217754364, "learning_rate": 9.741874711412535e-05, "loss": 0.0955, "step": 14760 }, { "grad_norm": 0.6468655467033386, "learning_rate": 9.741350051981042e-05, "loss": 0.0748, "step": 14770 }, { "grad_norm": 0.7349188923835754, "learning_rate": 9.740824874042633e-05, "loss": 0.0755, "step": 14780 }, { "grad_norm": 0.7738902568817139, "learning_rate": 9.740299177654746e-05, "loss": 0.0737, "step": 14790 }, { "grad_norm": 0.6949157118797302, "learning_rate": 9.739772962874867e-05, "loss": 0.0857, "step": 14800 }, { "grad_norm": 0.45122072100639343, "learning_rate": 9.739246229760541e-05, "loss": 0.0743, "step": 14810 }, { "grad_norm": 0.6559169292449951, "learning_rate": 9.738718978369376e-05, "loss": 0.087, "step": 14820 }, { "grad_norm": 0.5883896946907043, "learning_rate": 9.738191208759025e-05, "loss": 0.0916, "step": 14830 }, { "grad_norm": 0.7877429723739624, "learning_rate": 9.73766292098721e-05, "loss": 0.092, "step": 14840 }, { "grad_norm": 0.9662805795669556, "learning_rate": 9.737134115111699e-05, "loss": 0.0939, "step": 14850 }, { "grad_norm": 0.5918299555778503, "learning_rate": 9.736604791190323e-05, "loss": 0.0803, "step": 14860 }, { "grad_norm": 0.5544304847717285, "learning_rate": 9.73607494928097e-05, "loss": 0.0662, "step": 14870 }, { "grad_norm": 0.48321211338043213, "learning_rate": 9.735544589441581e-05, "loss": 0.0797, "step": 14880 }, { "grad_norm": 0.7098687291145325, "learning_rate": 9.735013711730154e-05, "loss": 0.0733, "step": 14890 }, { "grad_norm": 0.4790804386138916, "learning_rate": 9.734482316204747e-05, "loss": 0.0718, "step": 14900 }, { "grad_norm": 0.9693362712860107, "learning_rate": 9.733950402923473e-05, "loss": 0.0773, "step": 14910 }, { "grad_norm": 0.503608763217926, "learning_rate": 9.7334179719445e-05, "loss": 0.0749, "step": 14920 }, { "grad_norm": 0.6928145289421082, "learning_rate": 9.732885023326053e-05, "loss": 0.0955, "step": 14930 }, { "grad_norm": 0.5029400587081909, "learning_rate": 9.732351557126418e-05, "loss": 0.1191, "step": 14940 }, { "grad_norm": 0.7508179545402527, "learning_rate": 9.731817573403929e-05, "loss": 0.0932, "step": 14950 }, { "grad_norm": 0.6580794453620911, "learning_rate": 9.731283072216985e-05, "loss": 0.0878, "step": 14960 }, { "grad_norm": 0.5952546000480652, "learning_rate": 9.730748053624039e-05, "loss": 0.0755, "step": 14970 }, { "grad_norm": 0.5625155568122864, "learning_rate": 9.730212517683598e-05, "loss": 0.0943, "step": 14980 }, { "grad_norm": 0.5077638626098633, "learning_rate": 9.729676464454228e-05, "loss": 0.077, "step": 14990 }, { "grad_norm": 0.6750351190567017, "learning_rate": 9.72913989399455e-05, "loss": 0.076, "step": 15000 }, { "grad_norm": 0.4760940968990326, "learning_rate": 9.728602806363242e-05, "loss": 0.0649, "step": 15010 }, { "grad_norm": 0.5592759251594543, "learning_rate": 9.728065201619043e-05, "loss": 0.0833, "step": 15020 }, { "grad_norm": 0.6154934763908386, "learning_rate": 9.727527079820742e-05, "loss": 0.0853, "step": 15030 }, { "grad_norm": 0.5748220086097717, "learning_rate": 9.726988441027186e-05, "loss": 0.0923, "step": 15040 }, { "grad_norm": 0.5063791275024414, "learning_rate": 9.726449285297281e-05, "loss": 0.0935, "step": 15050 }, { "grad_norm": 0.5257989168167114, "learning_rate": 9.72590961268999e-05, "loss": 0.0948, "step": 15060 }, { "grad_norm": 0.4696279764175415, "learning_rate": 9.725369423264328e-05, "loss": 0.0816, "step": 15070 }, { "grad_norm": 0.5846192240715027, "learning_rate": 9.72482871707937e-05, "loss": 0.0747, "step": 15080 }, { "grad_norm": 0.6797550320625305, "learning_rate": 9.724287494194247e-05, "loss": 0.0946, "step": 15090 }, { "grad_norm": 0.43143412470817566, "learning_rate": 9.723745754668147e-05, "loss": 0.0834, "step": 15100 }, { "grad_norm": 0.5215710997581482, "learning_rate": 9.723203498560313e-05, "loss": 0.0975, "step": 15110 }, { "grad_norm": 0.3103964626789093, "learning_rate": 9.722660725930046e-05, "loss": 0.0637, "step": 15120 }, { "grad_norm": 0.5030950903892517, "learning_rate": 9.722117436836702e-05, "loss": 0.079, "step": 15130 }, { "grad_norm": 0.42826181650161743, "learning_rate": 9.721573631339696e-05, "loss": 0.0902, "step": 15140 }, { "grad_norm": 1.0338633060455322, "learning_rate": 9.721029309498494e-05, "loss": 0.1084, "step": 15150 }, { "grad_norm": 0.5289714932441711, "learning_rate": 9.720484471372627e-05, "loss": 0.0893, "step": 15160 }, { "grad_norm": 0.5298082828521729, "learning_rate": 9.719939117021673e-05, "loss": 0.0841, "step": 15170 }, { "grad_norm": 0.7752401232719421, "learning_rate": 9.719393246505275e-05, "loss": 0.0937, "step": 15180 }, { "grad_norm": 0.825082004070282, "learning_rate": 9.718846859883128e-05, "loss": 0.0776, "step": 15190 }, { "grad_norm": 0.5159083604812622, "learning_rate": 9.718299957214982e-05, "loss": 0.0924, "step": 15200 }, { "grad_norm": 0.5599299073219299, "learning_rate": 9.717752538560646e-05, "loss": 0.0869, "step": 15210 }, { "grad_norm": 0.6398415565490723, "learning_rate": 9.717204603979986e-05, "loss": 0.088, "step": 15220 }, { "grad_norm": 0.8119975924491882, "learning_rate": 9.716656153532922e-05, "loss": 0.0778, "step": 15230 }, { "grad_norm": 0.633912205696106, "learning_rate": 9.716107187279434e-05, "loss": 0.0948, "step": 15240 }, { "grad_norm": 0.5900552272796631, "learning_rate": 9.715557705279555e-05, "loss": 0.1029, "step": 15250 }, { "grad_norm": 0.5652599930763245, "learning_rate": 9.715007707593372e-05, "loss": 0.0997, "step": 15260 }, { "grad_norm": 0.6241016387939453, "learning_rate": 9.714457194281036e-05, "loss": 0.0798, "step": 15270 }, { "grad_norm": 0.5459205508232117, "learning_rate": 9.713906165402751e-05, "loss": 0.0831, "step": 15280 }, { "grad_norm": 0.47964638471603394, "learning_rate": 9.713354621018774e-05, "loss": 0.096, "step": 15290 }, { "grad_norm": 0.5124367475509644, "learning_rate": 9.712802561189422e-05, "loss": 0.0671, "step": 15300 }, { "grad_norm": 0.41365036368370056, "learning_rate": 9.712249985975069e-05, "loss": 0.0853, "step": 15310 }, { "grad_norm": 0.4700758159160614, "learning_rate": 9.71169689543614e-05, "loss": 0.0593, "step": 15320 }, { "grad_norm": 0.6113481521606445, "learning_rate": 9.711143289633123e-05, "loss": 0.0832, "step": 15330 }, { "grad_norm": 0.6366254091262817, "learning_rate": 9.710589168626561e-05, "loss": 0.0818, "step": 15340 }, { "grad_norm": 0.5714536905288696, "learning_rate": 9.710034532477048e-05, "loss": 0.0792, "step": 15350 }, { "grad_norm": 0.558681309223175, "learning_rate": 9.709479381245239e-05, "loss": 0.0684, "step": 15360 }, { "grad_norm": 0.5846589803695679, "learning_rate": 9.708923714991847e-05, "loss": 0.0869, "step": 15370 }, { "grad_norm": 0.6501700282096863, "learning_rate": 9.708367533777638e-05, "loss": 0.0794, "step": 15380 }, { "grad_norm": 0.48282602429389954, "learning_rate": 9.707810837663431e-05, "loss": 0.0686, "step": 15390 }, { "grad_norm": 0.6302624940872192, "learning_rate": 9.707253626710113e-05, "loss": 0.0818, "step": 15400 }, { "grad_norm": 0.6587678790092468, "learning_rate": 9.706695900978613e-05, "loss": 0.0702, "step": 15410 }, { "grad_norm": 0.8445385098457336, "learning_rate": 9.706137660529926e-05, "loss": 0.0826, "step": 15420 }, { "grad_norm": 0.6371160745620728, "learning_rate": 9.705578905425101e-05, "loss": 0.0899, "step": 15430 }, { "grad_norm": 0.8680310249328613, "learning_rate": 9.705019635725241e-05, "loss": 0.0827, "step": 15440 }, { "grad_norm": 0.5728322267532349, "learning_rate": 9.704459851491508e-05, "loss": 0.0958, "step": 15450 }, { "grad_norm": 0.7572831511497498, "learning_rate": 9.703899552785118e-05, "loss": 0.0807, "step": 15460 }, { "grad_norm": 0.5274686813354492, "learning_rate": 9.703338739667346e-05, "loss": 0.0682, "step": 15470 }, { "grad_norm": 0.4542377293109894, "learning_rate": 9.70277741219952e-05, "loss": 0.0781, "step": 15480 }, { "grad_norm": 0.6451070308685303, "learning_rate": 9.702215570443027e-05, "loss": 0.0804, "step": 15490 }, { "grad_norm": 0.5530219078063965, "learning_rate": 9.701653214459309e-05, "loss": 0.081, "step": 15500 }, { "grad_norm": 0.5734681487083435, "learning_rate": 9.701090344309865e-05, "loss": 0.0807, "step": 15510 }, { "grad_norm": 0.9585533738136292, "learning_rate": 9.700526960056247e-05, "loss": 0.0743, "step": 15520 }, { "grad_norm": 0.5695812106132507, "learning_rate": 9.699963061760068e-05, "loss": 0.0703, "step": 15530 }, { "grad_norm": 0.44293975830078125, "learning_rate": 9.699398649482997e-05, "loss": 0.0686, "step": 15540 }, { "grad_norm": 0.6702495217323303, "learning_rate": 9.698833723286753e-05, "loss": 0.062, "step": 15550 }, { "grad_norm": 0.49257057905197144, "learning_rate": 9.698268283233118e-05, "loss": 0.077, "step": 15560 }, { "grad_norm": 0.7176018357276917, "learning_rate": 9.697702329383929e-05, "loss": 0.08, "step": 15570 }, { "grad_norm": 0.4802094101905823, "learning_rate": 9.697135861801074e-05, "loss": 0.0872, "step": 15580 }, { "grad_norm": 0.4603065848350525, "learning_rate": 9.696568880546505e-05, "loss": 0.0706, "step": 15590 }, { "grad_norm": 0.7265236377716064, "learning_rate": 9.696001385682223e-05, "loss": 0.0768, "step": 15600 }, { "grad_norm": 0.5483164191246033, "learning_rate": 9.695433377270291e-05, "loss": 0.0754, "step": 15610 }, { "grad_norm": 0.4693377912044525, "learning_rate": 9.694864855372824e-05, "loss": 0.067, "step": 15620 }, { "grad_norm": 0.7349003553390503, "learning_rate": 9.694295820051995e-05, "loss": 0.0808, "step": 15630 }, { "grad_norm": 0.7514433264732361, "learning_rate": 9.693726271370032e-05, "loss": 0.0791, "step": 15640 }, { "grad_norm": 0.6334490180015564, "learning_rate": 9.693156209389221e-05, "loss": 0.0745, "step": 15650 }, { "grad_norm": 0.5800265669822693, "learning_rate": 9.692585634171905e-05, "loss": 0.0784, "step": 15660 }, { "grad_norm": 0.386869877576828, "learning_rate": 9.692014545780476e-05, "loss": 0.075, "step": 15670 }, { "grad_norm": 0.4827163815498352, "learning_rate": 9.691442944277393e-05, "loss": 0.0742, "step": 15680 }, { "grad_norm": 0.47350043058395386, "learning_rate": 9.690870829725162e-05, "loss": 0.0678, "step": 15690 }, { "grad_norm": 0.6200773119926453, "learning_rate": 9.69029820218635e-05, "loss": 0.0826, "step": 15700 }, { "grad_norm": 0.5975066423416138, "learning_rate": 9.689725061723579e-05, "loss": 0.0904, "step": 15710 }, { "grad_norm": 0.5834891200065613, "learning_rate": 9.689151408399527e-05, "loss": 0.0914, "step": 15720 }, { "grad_norm": 0.6101641654968262, "learning_rate": 9.688577242276924e-05, "loss": 0.0973, "step": 15730 }, { "grad_norm": 0.7198936939239502, "learning_rate": 9.688002563418566e-05, "loss": 0.0933, "step": 15740 }, { "grad_norm": 0.6129429340362549, "learning_rate": 9.687427371887293e-05, "loss": 0.0819, "step": 15750 }, { "grad_norm": 0.6923160552978516, "learning_rate": 9.686851667746012e-05, "loss": 0.0838, "step": 15760 }, { "grad_norm": 0.5464374423027039, "learning_rate": 9.686275451057677e-05, "loss": 0.0688, "step": 15770 }, { "grad_norm": 0.726552426815033, "learning_rate": 9.685698721885308e-05, "loss": 0.0791, "step": 15780 }, { "grad_norm": 0.548211932182312, "learning_rate": 9.68512148029197e-05, "loss": 0.0802, "step": 15790 }, { "grad_norm": 0.6118515729904175, "learning_rate": 9.684543726340791e-05, "loss": 0.084, "step": 15800 }, { "grad_norm": 0.891969621181488, "learning_rate": 9.683965460094952e-05, "loss": 0.0739, "step": 15810 }, { "grad_norm": 0.8087490797042847, "learning_rate": 9.683386681617694e-05, "loss": 0.0753, "step": 15820 }, { "grad_norm": 0.6174138188362122, "learning_rate": 9.68280739097231e-05, "loss": 0.0746, "step": 15830 }, { "grad_norm": 0.5855227708816528, "learning_rate": 9.682227588222148e-05, "loss": 0.0674, "step": 15840 }, { "grad_norm": 0.8396157026290894, "learning_rate": 9.681647273430618e-05, "loss": 0.0664, "step": 15850 }, { "grad_norm": 0.8396273255348206, "learning_rate": 9.681066446661182e-05, "loss": 0.0827, "step": 15860 }, { "grad_norm": 0.6536788940429688, "learning_rate": 9.680485107977357e-05, "loss": 0.0837, "step": 15870 }, { "grad_norm": 0.7934635281562805, "learning_rate": 9.679903257442716e-05, "loss": 0.0942, "step": 15880 }, { "grad_norm": 0.7664628624916077, "learning_rate": 9.679320895120891e-05, "loss": 0.0808, "step": 15890 }, { "grad_norm": 0.4039187431335449, "learning_rate": 9.67873802107557e-05, "loss": 0.0805, "step": 15900 }, { "grad_norm": 0.5179488658905029, "learning_rate": 9.67815463537049e-05, "loss": 0.0726, "step": 15910 }, { "grad_norm": 0.601483941078186, "learning_rate": 9.677570738069457e-05, "loss": 0.0724, "step": 15920 }, { "grad_norm": 0.47272244095802307, "learning_rate": 9.676986329236318e-05, "loss": 0.091, "step": 15930 }, { "grad_norm": 0.648162305355072, "learning_rate": 9.676401408934987e-05, "loss": 0.0854, "step": 15940 }, { "grad_norm": 0.6422484517097473, "learning_rate": 9.675815977229428e-05, "loss": 0.0878, "step": 15950 }, { "grad_norm": 0.6860766410827637, "learning_rate": 9.675230034183664e-05, "loss": 0.0712, "step": 15960 }, { "grad_norm": 0.5776480436325073, "learning_rate": 9.674643579861773e-05, "loss": 0.0958, "step": 15970 }, { "grad_norm": 0.5802061557769775, "learning_rate": 9.674056614327886e-05, "loss": 0.0747, "step": 15980 }, { "grad_norm": 0.5554571747779846, "learning_rate": 9.673469137646198e-05, "loss": 0.0856, "step": 15990 }, { "grad_norm": 0.7133740782737732, "learning_rate": 9.67288114988095e-05, "loss": 0.0706, "step": 16000 }, { "grad_norm": 0.6905292272567749, "learning_rate": 9.672292651096447e-05, "loss": 0.0739, "step": 16010 }, { "grad_norm": 0.674646258354187, "learning_rate": 9.671703641357042e-05, "loss": 0.0768, "step": 16020 }, { "grad_norm": 0.5195325016975403, "learning_rate": 9.67111412072715e-05, "loss": 0.0866, "step": 16030 }, { "grad_norm": 0.6415120363235474, "learning_rate": 9.670524089271242e-05, "loss": 0.0811, "step": 16040 }, { "grad_norm": 0.4859398901462555, "learning_rate": 9.669933547053842e-05, "loss": 0.0769, "step": 16050 }, { "grad_norm": 0.4922177493572235, "learning_rate": 9.669342494139531e-05, "loss": 0.084, "step": 16060 }, { "grad_norm": 0.6439933180809021, "learning_rate": 9.668750930592943e-05, "loss": 0.0733, "step": 16070 }, { "grad_norm": 0.5288057923316956, "learning_rate": 9.668158856478775e-05, "loss": 0.0745, "step": 16080 }, { "grad_norm": 0.5831019282341003, "learning_rate": 9.66756627186177e-05, "loss": 0.0817, "step": 16090 }, { "grad_norm": 0.7430217862129211, "learning_rate": 9.666973176806737e-05, "loss": 0.0808, "step": 16100 }, { "grad_norm": 0.5482876300811768, "learning_rate": 9.666379571378534e-05, "loss": 0.089, "step": 16110 }, { "grad_norm": 0.603146493434906, "learning_rate": 9.665785455642076e-05, "loss": 0.0989, "step": 16120 }, { "grad_norm": 0.7487511038780212, "learning_rate": 9.665190829662337e-05, "loss": 0.0945, "step": 16130 }, { "grad_norm": 0.5171171426773071, "learning_rate": 9.664595693504342e-05, "loss": 0.0699, "step": 16140 }, { "grad_norm": 0.5568011403083801, "learning_rate": 9.664000047233175e-05, "loss": 0.1066, "step": 16150 }, { "grad_norm": 0.5126671195030212, "learning_rate": 9.663403890913976e-05, "loss": 0.0785, "step": 16160 }, { "grad_norm": 0.7352534532546997, "learning_rate": 9.662807224611938e-05, "loss": 0.086, "step": 16170 }, { "grad_norm": 1.6497610807418823, "learning_rate": 9.662210048392311e-05, "loss": 0.0839, "step": 16180 }, { "grad_norm": 0.5963031649589539, "learning_rate": 9.661612362320405e-05, "loss": 0.0802, "step": 16190 }, { "grad_norm": 0.7163687944412231, "learning_rate": 9.661014166461579e-05, "loss": 0.0911, "step": 16200 }, { "grad_norm": 0.46795547008514404, "learning_rate": 9.66041546088125e-05, "loss": 0.0824, "step": 16210 }, { "grad_norm": 0.3698340058326721, "learning_rate": 9.659816245644895e-05, "loss": 0.0625, "step": 16220 }, { "grad_norm": 0.6527140736579895, "learning_rate": 9.65921652081804e-05, "loss": 0.0761, "step": 16230 }, { "grad_norm": 0.6725149750709534, "learning_rate": 9.658616286466271e-05, "loss": 0.0789, "step": 16240 }, { "grad_norm": 0.5046895742416382, "learning_rate": 9.65801554265523e-05, "loss": 0.0824, "step": 16250 }, { "grad_norm": 0.932680606842041, "learning_rate": 9.657414289450612e-05, "loss": 0.0752, "step": 16260 }, { "grad_norm": 0.8250831961631775, "learning_rate": 9.656812526918171e-05, "loss": 0.0867, "step": 16270 }, { "grad_norm": 0.6827318072319031, "learning_rate": 9.656210255123712e-05, "loss": 0.0885, "step": 16280 }, { "grad_norm": 0.6254127025604248, "learning_rate": 9.6556074741331e-05, "loss": 0.0823, "step": 16290 }, { "grad_norm": 0.5313261151313782, "learning_rate": 9.655004184012256e-05, "loss": 0.0763, "step": 16300 }, { "grad_norm": 1.0112786293029785, "learning_rate": 9.654400384827152e-05, "loss": 0.0911, "step": 16310 }, { "grad_norm": 0.6630284190177917, "learning_rate": 9.653796076643818e-05, "loss": 0.0696, "step": 16320 }, { "grad_norm": 0.5877612233161926, "learning_rate": 9.653191259528344e-05, "loss": 0.076, "step": 16330 }, { "grad_norm": 0.6738527417182922, "learning_rate": 9.65258593354687e-05, "loss": 0.0822, "step": 16340 }, { "grad_norm": 0.5942484140396118, "learning_rate": 9.651980098765591e-05, "loss": 0.0767, "step": 16350 }, { "grad_norm": 0.5924005508422852, "learning_rate": 9.651373755250765e-05, "loss": 0.0946, "step": 16360 }, { "grad_norm": 0.602979838848114, "learning_rate": 9.650766903068697e-05, "loss": 0.0832, "step": 16370 }, { "grad_norm": 0.7838202714920044, "learning_rate": 9.650159542285753e-05, "loss": 0.0694, "step": 16380 }, { "grad_norm": 0.7981898188591003, "learning_rate": 9.649551672968353e-05, "loss": 0.0974, "step": 16390 }, { "grad_norm": 0.6948804259300232, "learning_rate": 9.648943295182973e-05, "loss": 0.0823, "step": 16400 }, { "grad_norm": 0.5286460518836975, "learning_rate": 9.648334408996144e-05, "loss": 0.078, "step": 16410 }, { "grad_norm": 0.5606731176376343, "learning_rate": 9.647725014474452e-05, "loss": 0.0781, "step": 16420 }, { "grad_norm": 0.3831539452075958, "learning_rate": 9.64711511168454e-05, "loss": 0.0724, "step": 16430 }, { "grad_norm": 0.48631659150123596, "learning_rate": 9.646504700693108e-05, "loss": 0.0782, "step": 16440 }, { "grad_norm": 0.6731683611869812, "learning_rate": 9.645893781566907e-05, "loss": 0.0762, "step": 16450 }, { "grad_norm": 0.7287182807922363, "learning_rate": 9.645282354372744e-05, "loss": 0.0926, "step": 16460 }, { "grad_norm": 1.1036931276321411, "learning_rate": 9.644670419177491e-05, "loss": 0.0825, "step": 16470 }, { "grad_norm": 0.5695368647575378, "learning_rate": 9.644057976048062e-05, "loss": 0.0783, "step": 16480 }, { "grad_norm": 0.809765636920929, "learning_rate": 9.643445025051435e-05, "loss": 0.0889, "step": 16490 }, { "grad_norm": 0.6656861305236816, "learning_rate": 9.642831566254641e-05, "loss": 0.0763, "step": 16500 }, { "grad_norm": 0.7309142351150513, "learning_rate": 9.642217599724769e-05, "loss": 0.0728, "step": 16510 }, { "grad_norm": 0.46131762862205505, "learning_rate": 9.64160312552896e-05, "loss": 0.08, "step": 16520 }, { "grad_norm": 0.45449039340019226, "learning_rate": 9.64098814373441e-05, "loss": 0.0692, "step": 16530 }, { "grad_norm": 0.5640475749969482, "learning_rate": 9.640372654408374e-05, "loss": 0.076, "step": 16540 }, { "grad_norm": 0.5360599756240845, "learning_rate": 9.639756657618162e-05, "loss": 0.0857, "step": 16550 }, { "grad_norm": 0.7156016230583191, "learning_rate": 9.639140153431138e-05, "loss": 0.0793, "step": 16560 }, { "grad_norm": 0.6443936824798584, "learning_rate": 9.638523141914721e-05, "loss": 0.0834, "step": 16570 }, { "grad_norm": 0.7884394526481628, "learning_rate": 9.637905623136388e-05, "loss": 0.0881, "step": 16580 }, { "grad_norm": 0.980488121509552, "learning_rate": 9.637287597163669e-05, "loss": 0.0782, "step": 16590 }, { "grad_norm": 0.8002777099609375, "learning_rate": 9.63666906406415e-05, "loss": 0.0742, "step": 16600 }, { "grad_norm": 0.502001166343689, "learning_rate": 9.636050023905473e-05, "loss": 0.0671, "step": 16610 }, { "grad_norm": 0.658690333366394, "learning_rate": 9.635430476755336e-05, "loss": 0.0675, "step": 16620 }, { "grad_norm": 0.6352140307426453, "learning_rate": 9.63481042268149e-05, "loss": 0.0746, "step": 16630 }, { "grad_norm": 0.6546404957771301, "learning_rate": 9.634189861751745e-05, "loss": 0.0783, "step": 16640 }, { "grad_norm": 0.4329979121685028, "learning_rate": 9.633568794033967e-05, "loss": 0.0827, "step": 16650 }, { "grad_norm": 0.502984344959259, "learning_rate": 9.63294721959607e-05, "loss": 0.0751, "step": 16660 }, { "grad_norm": 0.48392635583877563, "learning_rate": 9.63232513850603e-05, "loss": 0.0928, "step": 16670 }, { "grad_norm": 0.6552484631538391, "learning_rate": 9.631702550831878e-05, "loss": 0.0835, "step": 16680 }, { "grad_norm": 0.5505117177963257, "learning_rate": 9.631079456641698e-05, "loss": 0.0702, "step": 16690 }, { "grad_norm": 0.7678192257881165, "learning_rate": 9.630455856003632e-05, "loss": 0.0708, "step": 16700 }, { "grad_norm": 0.626619279384613, "learning_rate": 9.629831748985876e-05, "loss": 0.0911, "step": 16710 }, { "grad_norm": 0.7499885559082031, "learning_rate": 9.629207135656679e-05, "loss": 0.0922, "step": 16720 }, { "grad_norm": 0.6453942656517029, "learning_rate": 9.628582016084353e-05, "loss": 0.073, "step": 16730 }, { "grad_norm": 0.6900116205215454, "learning_rate": 9.627956390337254e-05, "loss": 0.072, "step": 16740 }, { "grad_norm": 0.5744316577911377, "learning_rate": 9.627330258483802e-05, "loss": 0.0768, "step": 16750 }, { "grad_norm": 0.5451110005378723, "learning_rate": 9.62670362059247e-05, "loss": 0.062, "step": 16760 }, { "grad_norm": 0.48403632640838623, "learning_rate": 9.626076476731786e-05, "loss": 0.0751, "step": 16770 }, { "grad_norm": 0.4461827576160431, "learning_rate": 9.625448826970336e-05, "loss": 0.0878, "step": 16780 }, { "grad_norm": 0.9154906868934631, "learning_rate": 9.624820671376755e-05, "loss": 0.0887, "step": 16790 }, { "grad_norm": 0.4993385076522827, "learning_rate": 9.62419201001974e-05, "loss": 0.0712, "step": 16800 }, { "grad_norm": 0.6026655435562134, "learning_rate": 9.623562842968037e-05, "loss": 0.0731, "step": 16810 }, { "grad_norm": 0.9405809044837952, "learning_rate": 9.622933170290454e-05, "loss": 0.0771, "step": 16820 }, { "grad_norm": 0.6154747009277344, "learning_rate": 9.622302992055849e-05, "loss": 0.0763, "step": 16830 }, { "grad_norm": 0.5055434107780457, "learning_rate": 9.62167230833314e-05, "loss": 0.0634, "step": 16840 }, { "grad_norm": 0.6181569695472717, "learning_rate": 9.621041119191295e-05, "loss": 0.0702, "step": 16850 }, { "grad_norm": 0.6718677878379822, "learning_rate": 9.620409424699342e-05, "loss": 0.0768, "step": 16860 }, { "grad_norm": 0.638919472694397, "learning_rate": 9.619777224926359e-05, "loss": 0.0635, "step": 16870 }, { "grad_norm": 0.5406479835510254, "learning_rate": 9.619144519941485e-05, "loss": 0.0744, "step": 16880 }, { "grad_norm": 0.5319742560386658, "learning_rate": 9.618511309813912e-05, "loss": 0.093, "step": 16890 }, { "grad_norm": 0.816005527973175, "learning_rate": 9.617877594612886e-05, "loss": 0.0779, "step": 16900 }, { "grad_norm": 0.5417620539665222, "learning_rate": 9.617243374407707e-05, "loss": 0.0683, "step": 16910 }, { "grad_norm": 0.6126831769943237, "learning_rate": 9.616608649267736e-05, "loss": 0.0771, "step": 16920 }, { "grad_norm": 0.807256817817688, "learning_rate": 9.615973419262385e-05, "loss": 0.0681, "step": 16930 }, { "grad_norm": 0.730485200881958, "learning_rate": 9.615337684461119e-05, "loss": 0.0659, "step": 16940 }, { "grad_norm": 0.7053149342536926, "learning_rate": 9.614701444933465e-05, "loss": 0.0832, "step": 16950 }, { "grad_norm": 0.6442288756370544, "learning_rate": 9.614064700748997e-05, "loss": 0.0835, "step": 16960 }, { "grad_norm": 0.5569385886192322, "learning_rate": 9.613427451977352e-05, "loss": 0.0657, "step": 16970 }, { "grad_norm": 0.7885040044784546, "learning_rate": 9.612789698688216e-05, "loss": 0.0871, "step": 16980 }, { "grad_norm": 0.7540833353996277, "learning_rate": 9.612151440951334e-05, "loss": 0.0829, "step": 16990 }, { "grad_norm": 0.7247517704963684, "learning_rate": 9.611512678836506e-05, "loss": 0.0821, "step": 17000 }, { "grad_norm": 0.48580068349838257, "learning_rate": 9.610873412413584e-05, "loss": 0.08, "step": 17010 }, { "grad_norm": 2.1985504627227783, "learning_rate": 9.610233641752476e-05, "loss": 0.0913, "step": 17020 }, { "grad_norm": 0.8234118223190308, "learning_rate": 9.609593366923151e-05, "loss": 0.0758, "step": 17030 }, { "grad_norm": 1.047653317451477, "learning_rate": 9.608952587995625e-05, "loss": 0.0779, "step": 17040 }, { "grad_norm": 0.6569225192070007, "learning_rate": 9.608311305039972e-05, "loss": 0.085, "step": 17050 }, { "grad_norm": 0.8283534646034241, "learning_rate": 9.607669518126326e-05, "loss": 0.0978, "step": 17060 }, { "grad_norm": 0.5221315026283264, "learning_rate": 9.607027227324866e-05, "loss": 0.0697, "step": 17070 }, { "grad_norm": 0.6249687671661377, "learning_rate": 9.606384432705837e-05, "loss": 0.0727, "step": 17080 }, { "grad_norm": 0.6744558215141296, "learning_rate": 9.60574113433953e-05, "loss": 0.0878, "step": 17090 }, { "grad_norm": 0.5833021998405457, "learning_rate": 9.6050973322963e-05, "loss": 0.0782, "step": 17100 }, { "grad_norm": 0.5409721732139587, "learning_rate": 9.604453026646547e-05, "loss": 0.0878, "step": 17110 }, { "grad_norm": 0.5599305033683777, "learning_rate": 9.603808217460735e-05, "loss": 0.0736, "step": 17120 }, { "grad_norm": 0.6020180583000183, "learning_rate": 9.603162904809377e-05, "loss": 0.0745, "step": 17130 }, { "grad_norm": 0.4278711676597595, "learning_rate": 9.602517088763045e-05, "loss": 0.064, "step": 17140 }, { "grad_norm": 0.45914992690086365, "learning_rate": 9.601870769392365e-05, "loss": 0.0961, "step": 17150 }, { "grad_norm": 0.5516082048416138, "learning_rate": 9.601223946768017e-05, "loss": 0.0711, "step": 17160 }, { "grad_norm": 0.4507337212562561, "learning_rate": 9.600576620960734e-05, "loss": 0.077, "step": 17170 }, { "grad_norm": 0.7372624278068542, "learning_rate": 9.599928792041308e-05, "loss": 0.0805, "step": 17180 }, { "grad_norm": 0.801733672618866, "learning_rate": 9.599280460080587e-05, "loss": 0.0843, "step": 17190 }, { "grad_norm": 0.5340421199798584, "learning_rate": 9.59863162514947e-05, "loss": 0.085, "step": 17200 }, { "grad_norm": 0.5910202860832214, "learning_rate": 9.597982287318911e-05, "loss": 0.0726, "step": 17210 }, { "grad_norm": 0.47338736057281494, "learning_rate": 9.597332446659923e-05, "loss": 0.0815, "step": 17220 }, { "grad_norm": 0.5426884889602661, "learning_rate": 9.59668210324357e-05, "loss": 0.0721, "step": 17230 }, { "grad_norm": 0.42391496896743774, "learning_rate": 9.596031257140974e-05, "loss": 0.0661, "step": 17240 }, { "grad_norm": 0.5895815491676331, "learning_rate": 9.59537990842331e-05, "loss": 0.0952, "step": 17250 }, { "grad_norm": 0.38414230942726135, "learning_rate": 9.594728057161806e-05, "loss": 0.0665, "step": 17260 }, { "grad_norm": 0.8959840536117554, "learning_rate": 9.594075703427752e-05, "loss": 0.0927, "step": 17270 }, { "grad_norm": 0.677282989025116, "learning_rate": 9.593422847292486e-05, "loss": 0.0747, "step": 17280 }, { "grad_norm": 0.5793978571891785, "learning_rate": 9.592769488827402e-05, "loss": 0.0869, "step": 17290 }, { "grad_norm": 0.5844444036483765, "learning_rate": 9.592115628103952e-05, "loss": 0.0835, "step": 17300 }, { "grad_norm": 0.6186418533325195, "learning_rate": 9.591461265193643e-05, "loss": 0.075, "step": 17310 }, { "grad_norm": 0.8494516611099243, "learning_rate": 9.590806400168032e-05, "loss": 0.0746, "step": 17320 }, { "grad_norm": 0.36167097091674805, "learning_rate": 9.590151033098735e-05, "loss": 0.072, "step": 17330 }, { "grad_norm": 0.5185944437980652, "learning_rate": 9.589495164057423e-05, "loss": 0.0779, "step": 17340 }, { "grad_norm": 0.5346212387084961, "learning_rate": 9.58883879311582e-05, "loss": 0.0769, "step": 17350 }, { "grad_norm": 0.5240914225578308, "learning_rate": 9.588181920345705e-05, "loss": 0.072, "step": 17360 }, { "grad_norm": 0.3964552879333496, "learning_rate": 9.587524545818913e-05, "loss": 0.0731, "step": 17370 }, { "grad_norm": 1.1223474740982056, "learning_rate": 9.586866669607335e-05, "loss": 0.0718, "step": 17380 }, { "grad_norm": 0.9080682992935181, "learning_rate": 9.586208291782915e-05, "loss": 0.0849, "step": 17390 }, { "grad_norm": 0.8501637578010559, "learning_rate": 9.58554941241765e-05, "loss": 0.0951, "step": 17400 }, { "grad_norm": 0.5636930465698242, "learning_rate": 9.584890031583596e-05, "loss": 0.0699, "step": 17410 }, { "grad_norm": 0.6352372765541077, "learning_rate": 9.584230149352861e-05, "loss": 0.077, "step": 17420 }, { "grad_norm": 0.8089883327484131, "learning_rate": 9.58356976579761e-05, "loss": 0.0775, "step": 17430 }, { "grad_norm": 0.46331363916397095, "learning_rate": 9.58290888099006e-05, "loss": 0.0825, "step": 17440 }, { "grad_norm": 0.3960915505886078, "learning_rate": 9.582247495002486e-05, "loss": 0.0795, "step": 17450 }, { "grad_norm": 0.810122549533844, "learning_rate": 9.581585607907214e-05, "loss": 0.076, "step": 17460 }, { "grad_norm": 0.677003026008606, "learning_rate": 9.580923219776628e-05, "loss": 0.0729, "step": 17470 }, { "grad_norm": 1.41922926902771, "learning_rate": 9.580260330683167e-05, "loss": 0.076, "step": 17480 }, { "grad_norm": 0.8478376269340515, "learning_rate": 9.579596940699322e-05, "loss": 0.0844, "step": 17490 }, { "grad_norm": 0.7061793208122253, "learning_rate": 9.578933049897643e-05, "loss": 0.0693, "step": 17500 }, { "grad_norm": 0.6405140161514282, "learning_rate": 9.578268658350728e-05, "loss": 0.0736, "step": 17510 }, { "grad_norm": 0.4855371415615082, "learning_rate": 9.577603766131235e-05, "loss": 0.0775, "step": 17520 }, { "grad_norm": 0.8985607624053955, "learning_rate": 9.576938373311878e-05, "loss": 0.073, "step": 17530 }, { "grad_norm": 0.796748697757721, "learning_rate": 9.576272479965421e-05, "loss": 0.0782, "step": 17540 }, { "grad_norm": 0.5128733515739441, "learning_rate": 9.575606086164687e-05, "loss": 0.0767, "step": 17550 }, { "grad_norm": 0.5207359790802002, "learning_rate": 9.57493919198255e-05, "loss": 0.0898, "step": 17560 }, { "grad_norm": 0.4929557740688324, "learning_rate": 9.57427179749194e-05, "loss": 0.0865, "step": 17570 }, { "grad_norm": 0.5046430230140686, "learning_rate": 9.573603902765846e-05, "loss": 0.0658, "step": 17580 }, { "grad_norm": 0.4742160141468048, "learning_rate": 9.572935507877304e-05, "loss": 0.0678, "step": 17590 }, { "grad_norm": 0.9232434034347534, "learning_rate": 9.57226661289941e-05, "loss": 0.0859, "step": 17600 }, { "grad_norm": 0.7697393298149109, "learning_rate": 9.571597217905315e-05, "loss": 0.088, "step": 17610 }, { "grad_norm": 1.6165716648101807, "learning_rate": 9.57092732296822e-05, "loss": 0.084, "step": 17620 }, { "grad_norm": 0.5315110087394714, "learning_rate": 9.570256928161385e-05, "loss": 0.07, "step": 17630 }, { "grad_norm": 0.5904998183250427, "learning_rate": 9.569586033558126e-05, "loss": 0.0812, "step": 17640 }, { "grad_norm": 0.48092442750930786, "learning_rate": 9.568914639231807e-05, "loss": 0.068, "step": 17650 }, { "grad_norm": 0.8079853653907776, "learning_rate": 9.568242745255852e-05, "loss": 0.077, "step": 17660 }, { "grad_norm": 0.5370391011238098, "learning_rate": 9.567570351703739e-05, "loss": 0.0817, "step": 17670 }, { "grad_norm": 0.7122251391410828, "learning_rate": 9.566897458649001e-05, "loss": 0.0835, "step": 17680 }, { "grad_norm": 0.5795148611068726, "learning_rate": 9.566224066165221e-05, "loss": 0.075, "step": 17690 }, { "grad_norm": 0.6075171232223511, "learning_rate": 9.565550174326043e-05, "loss": 0.082, "step": 17700 }, { "grad_norm": 0.564479649066925, "learning_rate": 9.564875783205162e-05, "loss": 0.0754, "step": 17710 }, { "grad_norm": 0.5292786955833435, "learning_rate": 9.564200892876328e-05, "loss": 0.0819, "step": 17720 }, { "grad_norm": 0.6907364130020142, "learning_rate": 9.563525503413348e-05, "loss": 0.0798, "step": 17730 }, { "grad_norm": 0.5275889039039612, "learning_rate": 9.562849614890079e-05, "loss": 0.0833, "step": 17740 }, { "grad_norm": 0.34658828377723694, "learning_rate": 9.562173227380436e-05, "loss": 0.0854, "step": 17750 }, { "grad_norm": 0.5722072124481201, "learning_rate": 9.561496340958389e-05, "loss": 0.0775, "step": 17760 }, { "grad_norm": 0.6573469042778015, "learning_rate": 9.560818955697959e-05, "loss": 0.081, "step": 17770 }, { "grad_norm": 0.765557587146759, "learning_rate": 9.560141071673228e-05, "loss": 0.0737, "step": 17780 }, { "grad_norm": 1.0484719276428223, "learning_rate": 9.559462688958323e-05, "loss": 0.0511, "step": 17790 }, { "grad_norm": 0.6819642186164856, "learning_rate": 9.558783807627434e-05, "loss": 0.0623, "step": 17800 }, { "grad_norm": 0.7697984576225281, "learning_rate": 9.558104427754801e-05, "loss": 0.0994, "step": 17810 }, { "grad_norm": 0.6646579504013062, "learning_rate": 9.557424549414722e-05, "loss": 0.0945, "step": 17820 }, { "grad_norm": 0.585764467716217, "learning_rate": 9.556744172681546e-05, "loss": 0.0804, "step": 17830 }, { "grad_norm": 0.9744973182678223, "learning_rate": 9.556063297629677e-05, "loss": 0.0823, "step": 17840 }, { "grad_norm": 0.5623778104782104, "learning_rate": 9.555381924333578e-05, "loss": 0.0643, "step": 17850 }, { "grad_norm": 0.5677563548088074, "learning_rate": 9.554700052867758e-05, "loss": 0.0728, "step": 17860 }, { "grad_norm": 0.7848238348960876, "learning_rate": 9.554017683306789e-05, "loss": 0.0761, "step": 17870 }, { "grad_norm": 0.5083346366882324, "learning_rate": 9.553334815725294e-05, "loss": 0.0692, "step": 17880 }, { "grad_norm": 0.5335632562637329, "learning_rate": 9.552651450197949e-05, "loss": 0.0768, "step": 17890 }, { "grad_norm": 0.8969535231590271, "learning_rate": 9.551967586799486e-05, "loss": 0.0717, "step": 17900 }, { "grad_norm": 0.6348199248313904, "learning_rate": 9.551283225604692e-05, "loss": 0.0752, "step": 17910 }, { "grad_norm": 1.0709993839263916, "learning_rate": 9.550598366688406e-05, "loss": 0.0758, "step": 17920 }, { "grad_norm": 0.6327281594276428, "learning_rate": 9.549913010125526e-05, "loss": 0.0781, "step": 17930 }, { "grad_norm": 0.4540155231952667, "learning_rate": 9.549227155990999e-05, "loss": 0.0807, "step": 17940 }, { "grad_norm": 0.6557581424713135, "learning_rate": 9.548540804359828e-05, "loss": 0.0747, "step": 17950 }, { "grad_norm": 0.6202744841575623, "learning_rate": 9.547853955307077e-05, "loss": 0.0835, "step": 17960 }, { "grad_norm": 0.5911651253700256, "learning_rate": 9.547166608907853e-05, "loss": 0.0724, "step": 17970 }, { "grad_norm": 0.5047270655632019, "learning_rate": 9.546478765237326e-05, "loss": 0.083, "step": 17980 }, { "grad_norm": 0.5347998142242432, "learning_rate": 9.545790424370715e-05, "loss": 0.0757, "step": 17990 }, { "grad_norm": 0.4578913152217865, "learning_rate": 9.5451015863833e-05, "loss": 0.0851, "step": 18000 }, { "grad_norm": 0.7932640910148621, "learning_rate": 9.544412251350408e-05, "loss": 0.0813, "step": 18010 }, { "grad_norm": 0.5731373429298401, "learning_rate": 9.543722419347422e-05, "loss": 0.0735, "step": 18020 }, { "grad_norm": 0.5665539503097534, "learning_rate": 9.543032090449788e-05, "loss": 0.0883, "step": 18030 }, { "grad_norm": 0.7877020835876465, "learning_rate": 9.542341264732992e-05, "loss": 0.0778, "step": 18040 }, { "grad_norm": 0.7144089341163635, "learning_rate": 9.541649942272585e-05, "loss": 0.0858, "step": 18050 }, { "grad_norm": 0.5853544473648071, "learning_rate": 9.54095812314417e-05, "loss": 0.0665, "step": 18060 }, { "grad_norm": 0.5525876879692078, "learning_rate": 9.540265807423401e-05, "loss": 0.0735, "step": 18070 }, { "grad_norm": 0.5685207843780518, "learning_rate": 9.53957299518599e-05, "loss": 0.0759, "step": 18080 }, { "grad_norm": 0.44358426332473755, "learning_rate": 9.5388796865077e-05, "loss": 0.0833, "step": 18090 }, { "grad_norm": 0.6221632361412048, "learning_rate": 9.538185881464353e-05, "loss": 0.09, "step": 18100 }, { "grad_norm": 0.5590994954109192, "learning_rate": 9.537491580131821e-05, "loss": 0.0705, "step": 18110 }, { "grad_norm": 0.5300242900848389, "learning_rate": 9.53679678258603e-05, "loss": 0.073, "step": 18120 }, { "grad_norm": 0.836712121963501, "learning_rate": 9.536101488902966e-05, "loss": 0.0719, "step": 18130 }, { "grad_norm": 0.7920121550559998, "learning_rate": 9.535405699158663e-05, "loss": 0.0689, "step": 18140 }, { "grad_norm": 0.6832564473152161, "learning_rate": 9.53470941342921e-05, "loss": 0.0618, "step": 18150 }, { "grad_norm": 0.5353187918663025, "learning_rate": 9.534012631790756e-05, "loss": 0.0668, "step": 18160 }, { "grad_norm": 0.9268457889556885, "learning_rate": 9.533315354319494e-05, "loss": 0.0916, "step": 18170 }, { "grad_norm": 0.605911374092102, "learning_rate": 9.532617581091682e-05, "loss": 0.0806, "step": 18180 }, { "grad_norm": 0.4516974985599518, "learning_rate": 9.531919312183629e-05, "loss": 0.0705, "step": 18190 }, { "grad_norm": 1.1608102321624756, "learning_rate": 9.531220547671688e-05, "loss": 0.0701, "step": 18200 }, { "grad_norm": 0.4893709421157837, "learning_rate": 9.530521287632285e-05, "loss": 0.0768, "step": 18210 }, { "grad_norm": 0.6744621396064758, "learning_rate": 9.529821532141884e-05, "loss": 0.0765, "step": 18220 }, { "grad_norm": 0.4905012547969818, "learning_rate": 9.52912128127701e-05, "loss": 0.0671, "step": 18230 }, { "grad_norm": 0.6084622144699097, "learning_rate": 9.528420535114244e-05, "loss": 0.0705, "step": 18240 }, { "grad_norm": 0.6618199944496155, "learning_rate": 9.527719293730215e-05, "loss": 0.0797, "step": 18250 }, { "grad_norm": 0.4480268061161041, "learning_rate": 9.527017557201611e-05, "loss": 0.081, "step": 18260 }, { "grad_norm": 0.543916642665863, "learning_rate": 9.526315325605176e-05, "loss": 0.086, "step": 18270 }, { "grad_norm": 0.48915427923202515, "learning_rate": 9.525612599017699e-05, "loss": 0.0625, "step": 18280 }, { "grad_norm": 0.5047510266304016, "learning_rate": 9.524909377516033e-05, "loss": 0.0725, "step": 18290 }, { "grad_norm": 0.646248996257782, "learning_rate": 9.524205661177081e-05, "loss": 0.0701, "step": 18300 }, { "grad_norm": 0.4392319321632385, "learning_rate": 9.523501450077801e-05, "loss": 0.0695, "step": 18310 }, { "grad_norm": 0.5185439586639404, "learning_rate": 9.522796744295202e-05, "loss": 0.0819, "step": 18320 }, { "grad_norm": 0.6529541015625, "learning_rate": 9.522091543906352e-05, "loss": 0.0803, "step": 18330 }, { "grad_norm": 0.9537410140037537, "learning_rate": 9.521385848988369e-05, "loss": 0.0813, "step": 18340 }, { "grad_norm": 0.5192062258720398, "learning_rate": 9.520679659618428e-05, "loss": 0.0651, "step": 18350 }, { "grad_norm": 0.5795019865036011, "learning_rate": 9.519972975873754e-05, "loss": 0.0754, "step": 18360 }, { "grad_norm": 0.5191425085067749, "learning_rate": 9.519265797831633e-05, "loss": 0.0665, "step": 18370 }, { "grad_norm": 0.6332674622535706, "learning_rate": 9.518558125569399e-05, "loss": 0.0695, "step": 18380 }, { "grad_norm": 0.5526539087295532, "learning_rate": 9.517849959164442e-05, "loss": 0.0685, "step": 18390 }, { "grad_norm": 0.5254940390586853, "learning_rate": 9.517141298694205e-05, "loss": 0.0805, "step": 18400 }, { "grad_norm": 0.49431440234184265, "learning_rate": 9.516432144236188e-05, "loss": 0.0738, "step": 18410 }, { "grad_norm": 0.5066721439361572, "learning_rate": 9.515722495867941e-05, "loss": 0.0753, "step": 18420 }, { "grad_norm": 0.7685832381248474, "learning_rate": 9.515012353667072e-05, "loss": 0.0746, "step": 18430 }, { "grad_norm": 0.607819676399231, "learning_rate": 9.51430171771124e-05, "loss": 0.0613, "step": 18440 }, { "grad_norm": 0.6671893000602722, "learning_rate": 9.513590588078159e-05, "loss": 0.0753, "step": 18450 }, { "grad_norm": 0.7111877799034119, "learning_rate": 9.512878964845597e-05, "loss": 0.0782, "step": 18460 }, { "grad_norm": 0.5728774666786194, "learning_rate": 9.512166848091377e-05, "loss": 0.0892, "step": 18470 }, { "grad_norm": 0.4926949143409729, "learning_rate": 9.511454237893376e-05, "loss": 0.0799, "step": 18480 }, { "grad_norm": 0.6238998770713806, "learning_rate": 9.51074113432952e-05, "loss": 0.0646, "step": 18490 }, { "grad_norm": 0.5346561670303345, "learning_rate": 9.510027537477797e-05, "loss": 0.0937, "step": 18500 }, { "grad_norm": 0.5518092513084412, "learning_rate": 9.509313447416242e-05, "loss": 0.0728, "step": 18510 }, { "grad_norm": 0.5562000870704651, "learning_rate": 9.508598864222949e-05, "loss": 0.0776, "step": 18520 }, { "grad_norm": 0.760276734828949, "learning_rate": 9.507883787976062e-05, "loss": 0.0708, "step": 18530 }, { "grad_norm": 0.7931352257728577, "learning_rate": 9.507168218753781e-05, "loss": 0.0664, "step": 18540 }, { "grad_norm": 0.6024153828620911, "learning_rate": 9.506452156634362e-05, "loss": 0.0836, "step": 18550 }, { "grad_norm": 0.4829804599285126, "learning_rate": 9.505735601696109e-05, "loss": 0.0853, "step": 18560 }, { "grad_norm": 0.6102399826049805, "learning_rate": 9.505018554017385e-05, "loss": 0.0669, "step": 18570 }, { "grad_norm": 0.8983346819877625, "learning_rate": 9.504301013676604e-05, "loss": 0.0696, "step": 18580 }, { "grad_norm": 0.6152429580688477, "learning_rate": 9.503582980752238e-05, "loss": 0.0807, "step": 18590 }, { "grad_norm": 0.6090443134307861, "learning_rate": 9.502864455322809e-05, "loss": 0.0882, "step": 18600 }, { "grad_norm": 0.6617683172225952, "learning_rate": 9.502145437466891e-05, "loss": 0.0807, "step": 18610 }, { "grad_norm": 0.49380433559417725, "learning_rate": 9.501425927263116e-05, "loss": 0.0811, "step": 18620 }, { "grad_norm": 0.7246612906455994, "learning_rate": 9.500705924790172e-05, "loss": 0.1, "step": 18630 }, { "grad_norm": 0.8021202087402344, "learning_rate": 9.499985430126794e-05, "loss": 0.0773, "step": 18640 }, { "grad_norm": 0.5710193514823914, "learning_rate": 9.499264443351775e-05, "loss": 0.0748, "step": 18650 }, { "grad_norm": 0.7950872182846069, "learning_rate": 9.498542964543961e-05, "loss": 0.0726, "step": 18660 }, { "grad_norm": 0.53908771276474, "learning_rate": 9.497820993782252e-05, "loss": 0.0784, "step": 18670 }, { "grad_norm": 0.4986874759197235, "learning_rate": 9.497098531145601e-05, "loss": 0.0742, "step": 18680 }, { "grad_norm": 0.889297366142273, "learning_rate": 9.496375576713017e-05, "loss": 0.0829, "step": 18690 }, { "grad_norm": 0.5759965777397156, "learning_rate": 9.49565213056356e-05, "loss": 0.0765, "step": 18700 }, { "grad_norm": 0.4839211404323578, "learning_rate": 9.494928192776342e-05, "loss": 0.0788, "step": 18710 }, { "grad_norm": 0.8061370253562927, "learning_rate": 9.494203763430538e-05, "loss": 0.0719, "step": 18720 }, { "grad_norm": 0.43754705786705017, "learning_rate": 9.493478842605366e-05, "loss": 0.0936, "step": 18730 }, { "grad_norm": 0.8749645352363586, "learning_rate": 9.492753430380105e-05, "loss": 0.078, "step": 18740 }, { "grad_norm": 0.6603956818580627, "learning_rate": 9.492027526834083e-05, "loss": 0.0634, "step": 18750 }, { "grad_norm": 0.7339797616004944, "learning_rate": 9.491301132046684e-05, "loss": 0.0779, "step": 18760 }, { "grad_norm": 0.43404799699783325, "learning_rate": 9.490574246097345e-05, "loss": 0.0935, "step": 18770 }, { "grad_norm": 0.6404463648796082, "learning_rate": 9.48984686906556e-05, "loss": 0.0749, "step": 18780 }, { "grad_norm": 0.5626100897789001, "learning_rate": 9.489119001030871e-05, "loss": 0.0639, "step": 18790 }, { "grad_norm": 0.5123383402824402, "learning_rate": 9.488390642072878e-05, "loss": 0.0854, "step": 18800 }, { "grad_norm": 0.7351468801498413, "learning_rate": 9.48766179227123e-05, "loss": 0.0823, "step": 18810 }, { "grad_norm": 0.8334742188453674, "learning_rate": 9.486932451705636e-05, "loss": 0.0931, "step": 18820 }, { "grad_norm": 0.6065106987953186, "learning_rate": 9.486202620455857e-05, "loss": 0.0789, "step": 18830 }, { "grad_norm": 0.5175796151161194, "learning_rate": 9.485472298601704e-05, "loss": 0.0732, "step": 18840 }, { "grad_norm": 1.5982319116592407, "learning_rate": 9.484741486223043e-05, "loss": 0.0869, "step": 18850 }, { "grad_norm": 0.7848303318023682, "learning_rate": 9.484010183399797e-05, "loss": 0.0735, "step": 18860 }, { "grad_norm": 0.4899582862854004, "learning_rate": 9.483278390211938e-05, "loss": 0.0721, "step": 18870 }, { "grad_norm": 0.45719727873802185, "learning_rate": 9.482546106739496e-05, "loss": 0.057, "step": 18880 }, { "grad_norm": 0.629352867603302, "learning_rate": 9.48181333306255e-05, "loss": 0.0676, "step": 18890 }, { "grad_norm": 0.5962671041488647, "learning_rate": 9.481080069261237e-05, "loss": 0.0792, "step": 18900 }, { "grad_norm": 0.7411586046218872, "learning_rate": 9.480346315415745e-05, "loss": 0.0782, "step": 18910 }, { "grad_norm": 2.423921823501587, "learning_rate": 9.479612071606314e-05, "loss": 0.0776, "step": 18920 }, { "grad_norm": 0.40790390968322754, "learning_rate": 9.478877337913244e-05, "loss": 0.0682, "step": 18930 }, { "grad_norm": 0.5378691554069519, "learning_rate": 9.478142114416881e-05, "loss": 0.0625, "step": 18940 }, { "grad_norm": 1.0464645624160767, "learning_rate": 9.47740640119763e-05, "loss": 0.0651, "step": 18950 }, { "grad_norm": 0.6495813131332397, "learning_rate": 9.476670198335947e-05, "loss": 0.0765, "step": 18960 }, { "grad_norm": 0.7725969552993774, "learning_rate": 9.47593350591234e-05, "loss": 0.0767, "step": 18970 }, { "grad_norm": 0.5194699168205261, "learning_rate": 9.475196324007376e-05, "loss": 0.0778, "step": 18980 }, { "grad_norm": 0.5205252170562744, "learning_rate": 9.474458652701669e-05, "loss": 0.0707, "step": 18990 }, { "grad_norm": 0.6316065788269043, "learning_rate": 9.473720492075892e-05, "loss": 0.0773, "step": 19000 }, { "grad_norm": 0.6723541021347046, "learning_rate": 9.472981842210768e-05, "loss": 0.0727, "step": 19010 }, { "grad_norm": 1.2664436101913452, "learning_rate": 9.472242703187074e-05, "loss": 0.0662, "step": 19020 }, { "grad_norm": 0.5708550810813904, "learning_rate": 9.471503075085643e-05, "loss": 0.0649, "step": 19030 }, { "grad_norm": 0.9842525720596313, "learning_rate": 9.470762957987359e-05, "loss": 0.0814, "step": 19040 }, { "grad_norm": 0.5449227094650269, "learning_rate": 9.470022351973158e-05, "loss": 0.0819, "step": 19050 }, { "grad_norm": 0.5743147134780884, "learning_rate": 9.469281257124034e-05, "loss": 0.0745, "step": 19060 }, { "grad_norm": 0.8149358034133911, "learning_rate": 9.46853967352103e-05, "loss": 0.0734, "step": 19070 }, { "grad_norm": 0.6691988706588745, "learning_rate": 9.467797601245246e-05, "loss": 0.0715, "step": 19080 }, { "grad_norm": 1.2314177751541138, "learning_rate": 9.467055040377834e-05, "loss": 0.0782, "step": 19090 }, { "grad_norm": 0.4838065803050995, "learning_rate": 9.466311990999999e-05, "loss": 0.0775, "step": 19100 }, { "grad_norm": 0.749506413936615, "learning_rate": 9.465568453193e-05, "loss": 0.073, "step": 19110 }, { "grad_norm": 0.46312907338142395, "learning_rate": 9.464824427038148e-05, "loss": 0.0652, "step": 19120 }, { "grad_norm": 0.5609905123710632, "learning_rate": 9.46407991261681e-05, "loss": 0.0862, "step": 19130 }, { "grad_norm": 0.5269917845726013, "learning_rate": 9.463334910010404e-05, "loss": 0.0598, "step": 19140 }, { "grad_norm": 0.5351616144180298, "learning_rate": 9.462589419300403e-05, "loss": 0.0555, "step": 19150 }, { "grad_norm": 0.7514714002609253, "learning_rate": 9.461843440568333e-05, "loss": 0.0812, "step": 19160 }, { "grad_norm": 0.5169830918312073, "learning_rate": 9.461096973895773e-05, "loss": 0.0733, "step": 19170 }, { "grad_norm": 0.5969607830047607, "learning_rate": 9.460350019364355e-05, "loss": 0.0773, "step": 19180 }, { "grad_norm": 0.6917380094528198, "learning_rate": 9.459602577055764e-05, "loss": 0.1009, "step": 19190 }, { "grad_norm": 0.4732027053833008, "learning_rate": 9.45885464705174e-05, "loss": 0.0685, "step": 19200 }, { "grad_norm": 0.7900493741035461, "learning_rate": 9.458106229434076e-05, "loss": 0.0916, "step": 19210 }, { "grad_norm": 0.5619314312934875, "learning_rate": 9.457357324284617e-05, "loss": 0.0777, "step": 19220 }, { "grad_norm": 0.6264548897743225, "learning_rate": 9.456607931685262e-05, "loss": 0.0757, "step": 19230 }, { "grad_norm": 0.4697591960430145, "learning_rate": 9.455858051717965e-05, "loss": 0.0723, "step": 19240 }, { "grad_norm": 0.5170140266418457, "learning_rate": 9.45510768446473e-05, "loss": 0.0608, "step": 19250 }, { "grad_norm": 0.4730445146560669, "learning_rate": 9.454356830007618e-05, "loss": 0.0576, "step": 19260 }, { "grad_norm": 2.524696111679077, "learning_rate": 9.45360548842874e-05, "loss": 0.067, "step": 19270 }, { "grad_norm": 0.4618682265281677, "learning_rate": 9.452853659810261e-05, "loss": 0.0645, "step": 19280 }, { "grad_norm": 0.7814725041389465, "learning_rate": 9.452101344234401e-05, "loss": 0.088, "step": 19290 }, { "grad_norm": 0.4669937193393707, "learning_rate": 9.451348541783431e-05, "loss": 0.0727, "step": 19300 }, { "grad_norm": 0.8994338512420654, "learning_rate": 9.450595252539678e-05, "loss": 0.0725, "step": 19310 }, { "grad_norm": 0.7199760675430298, "learning_rate": 9.449841476585518e-05, "loss": 0.0861, "step": 19320 }, { "grad_norm": 0.738205075263977, "learning_rate": 9.449087214003384e-05, "loss": 0.0892, "step": 19330 }, { "grad_norm": 0.6587194800376892, "learning_rate": 9.448332464875765e-05, "loss": 0.0934, "step": 19340 }, { "grad_norm": 0.508217453956604, "learning_rate": 9.447577229285192e-05, "loss": 0.0915, "step": 19350 }, { "grad_norm": 0.5898306369781494, "learning_rate": 9.446821507314261e-05, "loss": 0.0721, "step": 19360 }, { "grad_norm": 0.9479820728302002, "learning_rate": 9.446065299045617e-05, "loss": 0.0804, "step": 19370 }, { "grad_norm": 0.4668600261211395, "learning_rate": 9.445308604561955e-05, "loss": 0.0826, "step": 19380 }, { "grad_norm": 0.7224615812301636, "learning_rate": 9.444551423946028e-05, "loss": 0.0738, "step": 19390 }, { "grad_norm": 0.7754127979278564, "learning_rate": 9.443793757280638e-05, "loss": 0.0864, "step": 19400 }, { "grad_norm": 1.0093727111816406, "learning_rate": 9.443035604648646e-05, "loss": 0.0825, "step": 19410 }, { "grad_norm": 0.7123512625694275, "learning_rate": 9.44227696613296e-05, "loss": 0.0701, "step": 19420 }, { "grad_norm": 0.5615556240081787, "learning_rate": 9.441517841816542e-05, "loss": 0.0964, "step": 19430 }, { "grad_norm": 0.5418074727058411, "learning_rate": 9.440758231782413e-05, "loss": 0.0654, "step": 19440 }, { "grad_norm": 0.4302369952201843, "learning_rate": 9.439998136113639e-05, "loss": 0.0799, "step": 19450 }, { "grad_norm": 0.7084254026412964, "learning_rate": 9.439237554893344e-05, "loss": 0.0768, "step": 19460 }, { "grad_norm": 0.5912802219390869, "learning_rate": 9.438476488204705e-05, "loss": 0.0782, "step": 19470 }, { "grad_norm": 0.6404387950897217, "learning_rate": 9.43771493613095e-05, "loss": 0.0761, "step": 19480 }, { "grad_norm": 0.605553925037384, "learning_rate": 9.436952898755362e-05, "loss": 0.0834, "step": 19490 }, { "grad_norm": 0.6530291438102722, "learning_rate": 9.436190376161276e-05, "loss": 0.0677, "step": 19500 }, { "grad_norm": 0.7162814736366272, "learning_rate": 9.43542736843208e-05, "loss": 0.1025, "step": 19510 }, { "grad_norm": 0.6535533666610718, "learning_rate": 9.434663875651216e-05, "loss": 0.0589, "step": 19520 }, { "grad_norm": 0.6688951849937439, "learning_rate": 9.433899897902177e-05, "loss": 0.0748, "step": 19530 }, { "grad_norm": 0.5291997790336609, "learning_rate": 9.433135435268511e-05, "loss": 0.0552, "step": 19540 }, { "grad_norm": 0.40093380212783813, "learning_rate": 9.432370487833819e-05, "loss": 0.0758, "step": 19550 }, { "grad_norm": 0.4035605788230896, "learning_rate": 9.431605055681756e-05, "loss": 0.0679, "step": 19560 }, { "grad_norm": 0.555280327796936, "learning_rate": 9.430839138896026e-05, "loss": 0.0792, "step": 19570 }, { "grad_norm": 0.6931050419807434, "learning_rate": 9.43007273756039e-05, "loss": 0.0788, "step": 19580 }, { "grad_norm": 0.5204131007194519, "learning_rate": 9.429305851758658e-05, "loss": 0.0792, "step": 19590 }, { "grad_norm": 0.6061282753944397, "learning_rate": 9.428538481574699e-05, "loss": 0.0796, "step": 19600 }, { "grad_norm": 0.8255851864814758, "learning_rate": 9.42777062709243e-05, "loss": 0.0821, "step": 19610 }, { "grad_norm": 0.4616852104663849, "learning_rate": 9.427002288395821e-05, "loss": 0.0718, "step": 19620 }, { "grad_norm": 0.5489121675491333, "learning_rate": 9.426233465568898e-05, "loss": 0.0695, "step": 19630 }, { "grad_norm": 0.45352616906166077, "learning_rate": 9.42546415869574e-05, "loss": 0.0707, "step": 19640 }, { "grad_norm": 0.46539008617401123, "learning_rate": 9.424694367860473e-05, "loss": 0.0561, "step": 19650 }, { "grad_norm": 0.6792424321174622, "learning_rate": 9.423924093147284e-05, "loss": 0.0598, "step": 19660 }, { "grad_norm": 0.552990734577179, "learning_rate": 9.423153334640407e-05, "loss": 0.0739, "step": 19670 }, { "grad_norm": 0.8748123049736023, "learning_rate": 9.42238209242413e-05, "loss": 0.062, "step": 19680 }, { "grad_norm": 1.0373413562774658, "learning_rate": 9.421610366582798e-05, "loss": 0.0877, "step": 19690 }, { "grad_norm": 0.6187617182731628, "learning_rate": 9.420838157200803e-05, "loss": 0.0719, "step": 19700 }, { "grad_norm": 0.6550151705741882, "learning_rate": 9.420065464362594e-05, "loss": 0.088, "step": 19710 }, { "grad_norm": 0.7805719375610352, "learning_rate": 9.419292288152673e-05, "loss": 0.0756, "step": 19720 }, { "grad_norm": 0.5600338578224182, "learning_rate": 9.418518628655588e-05, "loss": 0.0678, "step": 19730 }, { "grad_norm": 0.42017820477485657, "learning_rate": 9.417744485955951e-05, "loss": 0.0692, "step": 19740 }, { "grad_norm": 0.9365131855010986, "learning_rate": 9.41696986013842e-05, "loss": 0.0778, "step": 19750 }, { "grad_norm": 0.5265799760818481, "learning_rate": 9.416194751287705e-05, "loss": 0.0765, "step": 19760 }, { "grad_norm": 0.7565380930900574, "learning_rate": 9.415419159488572e-05, "loss": 0.0606, "step": 19770 }, { "grad_norm": 0.7404882907867432, "learning_rate": 9.414643084825837e-05, "loss": 0.0877, "step": 19780 }, { "grad_norm": 0.8566308617591858, "learning_rate": 9.413866527384372e-05, "loss": 0.0793, "step": 19790 }, { "grad_norm": 0.5854635238647461, "learning_rate": 9.4130894872491e-05, "loss": 0.0822, "step": 19800 }, { "grad_norm": 1.0612430572509766, "learning_rate": 9.412311964504998e-05, "loss": 0.0889, "step": 19810 }, { "grad_norm": 0.5441276431083679, "learning_rate": 9.411533959237091e-05, "loss": 0.087, "step": 19820 }, { "grad_norm": 0.3384670913219452, "learning_rate": 9.410755471530464e-05, "loss": 0.0545, "step": 19830 }, { "grad_norm": 0.7217287421226501, "learning_rate": 9.40997650147025e-05, "loss": 0.0807, "step": 19840 }, { "grad_norm": 0.7106914520263672, "learning_rate": 9.409197049141637e-05, "loss": 0.0899, "step": 19850 }, { "grad_norm": 0.5647422671318054, "learning_rate": 9.408417114629863e-05, "loss": 0.0734, "step": 19860 }, { "grad_norm": 0.472606360912323, "learning_rate": 9.40763669802022e-05, "loss": 0.0653, "step": 19870 }, { "grad_norm": 0.6718759536743164, "learning_rate": 9.406855799398056e-05, "loss": 0.0874, "step": 19880 }, { "grad_norm": 0.7411605715751648, "learning_rate": 9.406074418848767e-05, "loss": 0.0784, "step": 19890 }, { "grad_norm": 0.5499569177627563, "learning_rate": 9.405292556457805e-05, "loss": 0.0846, "step": 19900 }, { "grad_norm": 0.4839414358139038, "learning_rate": 9.404510212310671e-05, "loss": 0.0786, "step": 19910 }, { "grad_norm": 0.36482885479927063, "learning_rate": 9.403727386492924e-05, "loss": 0.0634, "step": 19920 }, { "grad_norm": 0.36770403385162354, "learning_rate": 9.40294407909017e-05, "loss": 0.0656, "step": 19930 }, { "grad_norm": 0.7553935647010803, "learning_rate": 9.40216029018807e-05, "loss": 0.0706, "step": 19940 }, { "grad_norm": 0.6235281825065613, "learning_rate": 9.401376019872338e-05, "loss": 0.0638, "step": 19950 }, { "grad_norm": 0.5905705690383911, "learning_rate": 9.400591268228746e-05, "loss": 0.0675, "step": 19960 }, { "grad_norm": 0.6361359357833862, "learning_rate": 9.399806035343106e-05, "loss": 0.0688, "step": 19970 }, { "grad_norm": 0.4576830267906189, "learning_rate": 9.399020321301294e-05, "loss": 0.0631, "step": 19980 }, { "grad_norm": 0.6986242532730103, "learning_rate": 9.398234126189234e-05, "loss": 0.0834, "step": 19990 }, { "grad_norm": 1.3324246406555176, "learning_rate": 9.397447450092902e-05, "loss": 0.0714, "step": 20000 }, { "grad_norm": 0.38309168815612793, "learning_rate": 9.39666029309833e-05, "loss": 0.0659, "step": 20010 }, { "grad_norm": 0.6019736528396606, "learning_rate": 9.395872655291596e-05, "loss": 0.064, "step": 20020 }, { "grad_norm": 0.585106372833252, "learning_rate": 9.395084536758838e-05, "loss": 0.0787, "step": 20030 }, { "grad_norm": 0.9294666051864624, "learning_rate": 9.394295937586243e-05, "loss": 0.1041, "step": 20040 }, { "grad_norm": 0.4654279053211212, "learning_rate": 9.393506857860052e-05, "loss": 0.0722, "step": 20050 }, { "grad_norm": 0.6946427822113037, "learning_rate": 9.392717297666555e-05, "loss": 0.0586, "step": 20060 }, { "grad_norm": 0.6844024062156677, "learning_rate": 9.391927257092101e-05, "loss": 0.0735, "step": 20070 }, { "grad_norm": 0.5992785096168518, "learning_rate": 9.391136736223085e-05, "loss": 0.0572, "step": 20080 }, { "grad_norm": 0.5386362075805664, "learning_rate": 9.390345735145956e-05, "loss": 0.0772, "step": 20090 }, { "grad_norm": 0.729907751083374, "learning_rate": 9.389554253947219e-05, "loss": 0.0705, "step": 20100 }, { "grad_norm": 0.5718702673912048, "learning_rate": 9.388762292713428e-05, "loss": 0.0753, "step": 20110 }, { "grad_norm": 0.43719014525413513, "learning_rate": 9.38796985153119e-05, "loss": 0.0692, "step": 20120 }, { "grad_norm": 0.5642496347427368, "learning_rate": 9.387176930487169e-05, "loss": 0.0717, "step": 20130 }, { "grad_norm": 1.2700557708740234, "learning_rate": 9.386383529668072e-05, "loss": 0.0721, "step": 20140 }, { "grad_norm": 0.566541314125061, "learning_rate": 9.385589649160669e-05, "loss": 0.0781, "step": 20150 }, { "grad_norm": 0.5467771291732788, "learning_rate": 9.384795289051775e-05, "loss": 0.0788, "step": 20160 }, { "grad_norm": 0.8373836874961853, "learning_rate": 9.384000449428261e-05, "loss": 0.0728, "step": 20170 }, { "grad_norm": 0.8126732707023621, "learning_rate": 9.383205130377048e-05, "loss": 0.0788, "step": 20180 }, { "grad_norm": 0.5008673667907715, "learning_rate": 9.382409331985114e-05, "loss": 0.0658, "step": 20190 }, { "grad_norm": 1.3508187532424927, "learning_rate": 9.381613054339482e-05, "loss": 0.0741, "step": 20200 }, { "grad_norm": 0.5055373907089233, "learning_rate": 9.380816297527235e-05, "loss": 0.0708, "step": 20210 }, { "grad_norm": 0.6131718158721924, "learning_rate": 9.380019061635506e-05, "loss": 0.0761, "step": 20220 }, { "grad_norm": 0.48959192633628845, "learning_rate": 9.379221346751474e-05, "loss": 0.0852, "step": 20230 }, { "grad_norm": 0.6482152938842773, "learning_rate": 9.378423152962382e-05, "loss": 0.0598, "step": 20240 }, { "grad_norm": 0.522408664226532, "learning_rate": 9.377624480355517e-05, "loss": 0.0642, "step": 20250 }, { "grad_norm": 0.44794654846191406, "learning_rate": 9.376825329018219e-05, "loss": 0.0662, "step": 20260 }, { "grad_norm": 0.6629520058631897, "learning_rate": 9.376025699037884e-05, "loss": 0.0723, "step": 20270 }, { "grad_norm": 0.5392124652862549, "learning_rate": 9.37522559050196e-05, "loss": 0.0733, "step": 20280 }, { "grad_norm": 0.49703559279441833, "learning_rate": 9.37442500349794e-05, "loss": 0.0714, "step": 20290 }, { "grad_norm": 0.5470393300056458, "learning_rate": 9.373623938113381e-05, "loss": 0.0696, "step": 20300 }, { "grad_norm": 0.5709894299507141, "learning_rate": 9.372822394435883e-05, "loss": 0.0838, "step": 20310 }, { "grad_norm": 0.46031832695007324, "learning_rate": 9.372020372553102e-05, "loss": 0.0747, "step": 20320 }, { "grad_norm": 0.5219508409500122, "learning_rate": 9.371217872552746e-05, "loss": 0.0708, "step": 20330 }, { "grad_norm": 0.5731662511825562, "learning_rate": 9.370414894522576e-05, "loss": 0.0691, "step": 20340 }, { "grad_norm": 0.5779550671577454, "learning_rate": 9.369611438550406e-05, "loss": 0.076, "step": 20350 }, { "grad_norm": 0.4022309482097626, "learning_rate": 9.368807504724095e-05, "loss": 0.0725, "step": 20360 }, { "grad_norm": 0.7823545932769775, "learning_rate": 9.368003093131565e-05, "loss": 0.0801, "step": 20370 }, { "grad_norm": 0.5589540004730225, "learning_rate": 9.367198203860785e-05, "loss": 0.0804, "step": 20380 }, { "grad_norm": 0.5132343769073486, "learning_rate": 9.366392836999774e-05, "loss": 0.0588, "step": 20390 }, { "grad_norm": 0.7697049379348755, "learning_rate": 9.365586992636607e-05, "loss": 0.0798, "step": 20400 }, { "grad_norm": 0.5177550315856934, "learning_rate": 9.364780670859412e-05, "loss": 0.0681, "step": 20410 }, { "grad_norm": 0.42364028096199036, "learning_rate": 9.363973871756364e-05, "loss": 0.0715, "step": 20420 }, { "grad_norm": 1.3443034887313843, "learning_rate": 9.363166595415696e-05, "loss": 0.0678, "step": 20430 }, { "grad_norm": 0.7353771924972534, "learning_rate": 9.362358841925686e-05, "loss": 0.0778, "step": 20440 }, { "grad_norm": 0.4956705868244171, "learning_rate": 9.361550611374674e-05, "loss": 0.0777, "step": 20450 }, { "grad_norm": 0.46391287446022034, "learning_rate": 9.360741903851043e-05, "loss": 0.1047, "step": 20460 }, { "grad_norm": 0.585292637348175, "learning_rate": 9.359932719443236e-05, "loss": 0.0989, "step": 20470 }, { "grad_norm": 0.35305148363113403, "learning_rate": 9.35912305823974e-05, "loss": 0.0708, "step": 20480 }, { "grad_norm": 0.6817534565925598, "learning_rate": 9.358312920329101e-05, "loss": 0.0692, "step": 20490 }, { "grad_norm": 0.44908276200294495, "learning_rate": 9.357502305799914e-05, "loss": 0.0624, "step": 20500 }, { "grad_norm": 0.5838736295700073, "learning_rate": 9.356691214740824e-05, "loss": 0.0569, "step": 20510 }, { "grad_norm": 0.5714921355247498, "learning_rate": 9.355879647240535e-05, "loss": 0.0633, "step": 20520 }, { "grad_norm": 0.5088648200035095, "learning_rate": 9.355067603387798e-05, "loss": 0.079, "step": 20530 }, { "grad_norm": 0.607487142086029, "learning_rate": 9.354255083271412e-05, "loss": 0.0788, "step": 20540 }, { "grad_norm": 0.5277121067047119, "learning_rate": 9.353442086980239e-05, "loss": 0.0648, "step": 20550 }, { "grad_norm": 0.6068124771118164, "learning_rate": 9.352628614603185e-05, "loss": 0.068, "step": 20560 }, { "grad_norm": 0.6047629117965698, "learning_rate": 9.351814666229209e-05, "loss": 0.072, "step": 20570 }, { "grad_norm": 0.8709841370582581, "learning_rate": 9.351000241947324e-05, "loss": 0.0704, "step": 20580 }, { "grad_norm": 0.5603572726249695, "learning_rate": 9.350185341846594e-05, "loss": 0.0767, "step": 20590 }, { "grad_norm": 0.3825701177120209, "learning_rate": 9.349369966016134e-05, "loss": 0.0722, "step": 20600 }, { "grad_norm": 0.7568562030792236, "learning_rate": 9.348554114545117e-05, "loss": 0.0935, "step": 20610 }, { "grad_norm": 0.7424692511558533, "learning_rate": 9.347737787522758e-05, "loss": 0.0765, "step": 20620 }, { "grad_norm": 0.4880509078502655, "learning_rate": 9.346920985038332e-05, "loss": 0.0845, "step": 20630 }, { "grad_norm": 0.7184958457946777, "learning_rate": 9.346103707181162e-05, "loss": 0.0741, "step": 20640 }, { "grad_norm": 0.46925634145736694, "learning_rate": 9.345285954040626e-05, "loss": 0.0695, "step": 20650 }, { "grad_norm": 0.6275935173034668, "learning_rate": 9.34446772570615e-05, "loss": 0.0682, "step": 20660 }, { "grad_norm": 0.5125834345817566, "learning_rate": 9.343649022267214e-05, "loss": 0.06, "step": 20670 }, { "grad_norm": 0.8632658123970032, "learning_rate": 9.342829843813353e-05, "loss": 0.067, "step": 20680 }, { "grad_norm": 0.5149610638618469, "learning_rate": 9.342010190434149e-05, "loss": 0.0679, "step": 20690 }, { "grad_norm": 0.6126881241798401, "learning_rate": 9.34119006221924e-05, "loss": 0.0762, "step": 20700 }, { "grad_norm": 0.6112223267555237, "learning_rate": 9.340369459258313e-05, "loss": 0.0621, "step": 20710 }, { "grad_norm": 0.6775323748588562, "learning_rate": 9.339548381641106e-05, "loss": 0.0733, "step": 20720 }, { "grad_norm": 0.5266730785369873, "learning_rate": 9.338726829457413e-05, "loss": 0.0826, "step": 20730 }, { "grad_norm": 0.6036523580551147, "learning_rate": 9.337904802797078e-05, "loss": 0.0636, "step": 20740 }, { "grad_norm": 0.6573405861854553, "learning_rate": 9.337082301749993e-05, "loss": 0.0724, "step": 20750 }, { "grad_norm": 0.4133426547050476, "learning_rate": 9.336259326406109e-05, "loss": 0.0688, "step": 20760 }, { "grad_norm": 0.7759695649147034, "learning_rate": 9.335435876855427e-05, "loss": 0.0808, "step": 20770 }, { "grad_norm": 0.585130512714386, "learning_rate": 9.334611953187994e-05, "loss": 0.0731, "step": 20780 }, { "grad_norm": 0.5776684880256653, "learning_rate": 9.333787555493914e-05, "loss": 0.0664, "step": 20790 }, { "grad_norm": 0.6515368223190308, "learning_rate": 9.332962683863345e-05, "loss": 0.0858, "step": 20800 }, { "grad_norm": 0.3605194389820099, "learning_rate": 9.332137338386489e-05, "loss": 0.0654, "step": 20810 }, { "grad_norm": 0.5652406215667725, "learning_rate": 9.33131151915361e-05, "loss": 0.0819, "step": 20820 }, { "grad_norm": 0.5262340307235718, "learning_rate": 9.330485226255012e-05, "loss": 0.0666, "step": 20830 }, { "grad_norm": 0.480140745639801, "learning_rate": 9.329658459781061e-05, "loss": 0.0676, "step": 20840 }, { "grad_norm": 0.31407949328422546, "learning_rate": 9.328831219822172e-05, "loss": 0.0701, "step": 20850 }, { "grad_norm": 0.6733301281929016, "learning_rate": 9.328003506468808e-05, "loss": 0.0845, "step": 20860 }, { "grad_norm": 0.5497675538063049, "learning_rate": 9.327175319811488e-05, "loss": 0.0687, "step": 20870 }, { "grad_norm": 0.5949028730392456, "learning_rate": 9.326346659940781e-05, "loss": 0.0667, "step": 20880 }, { "grad_norm": 0.421586811542511, "learning_rate": 9.325517526947308e-05, "loss": 0.0631, "step": 20890 }, { "grad_norm": 0.532395601272583, "learning_rate": 9.32468792092174e-05, "loss": 0.0691, "step": 20900 }, { "grad_norm": 0.5408346056938171, "learning_rate": 9.323857841954803e-05, "loss": 0.0686, "step": 20910 }, { "grad_norm": 0.6740797758102417, "learning_rate": 9.323027290137276e-05, "loss": 0.0662, "step": 20920 }, { "grad_norm": 0.445829838514328, "learning_rate": 9.322196265559981e-05, "loss": 0.0708, "step": 20930 }, { "grad_norm": 0.8124565482139587, "learning_rate": 9.321364768313803e-05, "loss": 0.072, "step": 20940 }, { "grad_norm": 0.5154606103897095, "learning_rate": 9.32053279848967e-05, "loss": 0.0818, "step": 20950 }, { "grad_norm": 0.4971419870853424, "learning_rate": 9.319700356178567e-05, "loss": 0.0715, "step": 20960 }, { "grad_norm": 0.6455475687980652, "learning_rate": 9.318867441471527e-05, "loss": 0.0649, "step": 20970 }, { "grad_norm": 0.5775684118270874, "learning_rate": 9.318034054459637e-05, "loss": 0.0551, "step": 20980 }, { "grad_norm": 0.49631184339523315, "learning_rate": 9.317200195234034e-05, "loss": 0.0792, "step": 20990 }, { "grad_norm": 0.8313280940055847, "learning_rate": 9.316365863885909e-05, "loss": 0.0865, "step": 21000 }, { "grad_norm": 0.5777172446250916, "learning_rate": 9.315531060506502e-05, "loss": 0.0825, "step": 21010 }, { "grad_norm": 0.5145334601402283, "learning_rate": 9.314695785187108e-05, "loss": 0.0694, "step": 21020 }, { "grad_norm": 0.5080334544181824, "learning_rate": 9.313860038019069e-05, "loss": 0.0669, "step": 21030 }, { "grad_norm": 0.5664353370666504, "learning_rate": 9.313023819093782e-05, "loss": 0.0938, "step": 21040 }, { "grad_norm": 0.5362957715988159, "learning_rate": 9.312187128502695e-05, "loss": 0.0727, "step": 21050 }, { "grad_norm": 0.8703277707099915, "learning_rate": 9.311349966337307e-05, "loss": 0.0772, "step": 21060 }, { "grad_norm": 0.5816182494163513, "learning_rate": 9.310512332689169e-05, "loss": 0.0761, "step": 21070 }, { "grad_norm": 0.6638891696929932, "learning_rate": 9.309674227649883e-05, "loss": 0.0699, "step": 21080 }, { "grad_norm": 0.4957985579967499, "learning_rate": 9.308835651311103e-05, "loss": 0.0698, "step": 21090 }, { "grad_norm": 0.6349040865898132, "learning_rate": 9.307996603764533e-05, "loss": 0.0696, "step": 21100 }, { "grad_norm": 0.6452133059501648, "learning_rate": 9.307157085101932e-05, "loss": 0.0703, "step": 21110 }, { "grad_norm": 0.916531503200531, "learning_rate": 9.306317095415109e-05, "loss": 0.0704, "step": 21120 }, { "grad_norm": 0.5805717706680298, "learning_rate": 9.305476634795922e-05, "loss": 0.0671, "step": 21130 }, { "grad_norm": 0.4949932396411896, "learning_rate": 9.304635703336284e-05, "loss": 0.069, "step": 21140 }, { "grad_norm": 0.5672680139541626, "learning_rate": 9.303794301128157e-05, "loss": 0.0633, "step": 21150 }, { "grad_norm": 0.5047290921211243, "learning_rate": 9.302952428263555e-05, "loss": 0.0723, "step": 21160 }, { "grad_norm": 0.6344391107559204, "learning_rate": 9.302110084834545e-05, "loss": 0.0693, "step": 21170 }, { "grad_norm": 0.4556119740009308, "learning_rate": 9.301267270933245e-05, "loss": 0.0537, "step": 21180 }, { "grad_norm": 0.7282041907310486, "learning_rate": 9.300423986651823e-05, "loss": 0.0649, "step": 21190 }, { "grad_norm": 0.5366691946983337, "learning_rate": 9.299580232082501e-05, "loss": 0.0721, "step": 21200 }, { "grad_norm": 0.5751627087593079, "learning_rate": 9.298736007317547e-05, "loss": 0.0633, "step": 21210 }, { "grad_norm": 0.5105568766593933, "learning_rate": 9.297891312449288e-05, "loss": 0.0742, "step": 21220 }, { "grad_norm": 0.6590508222579956, "learning_rate": 9.297046147570094e-05, "loss": 0.0682, "step": 21230 }, { "grad_norm": 0.6197623014450073, "learning_rate": 9.296200512772396e-05, "loss": 0.0633, "step": 21240 }, { "grad_norm": 0.4932998716831207, "learning_rate": 9.295354408148668e-05, "loss": 0.0678, "step": 21250 }, { "grad_norm": 0.5168628692626953, "learning_rate": 9.294507833791441e-05, "loss": 0.0664, "step": 21260 }, { "grad_norm": 1.082302212715149, "learning_rate": 9.293660789793295e-05, "loss": 0.0827, "step": 21270 }, { "grad_norm": 1.4049524068832397, "learning_rate": 9.292813276246858e-05, "loss": 0.0747, "step": 21280 }, { "grad_norm": 0.5320402979850769, "learning_rate": 9.291965293244816e-05, "loss": 0.0692, "step": 21290 }, { "grad_norm": 0.6486984491348267, "learning_rate": 9.291116840879904e-05, "loss": 0.07, "step": 21300 }, { "grad_norm": 0.5329123735427856, "learning_rate": 9.290267919244904e-05, "loss": 0.0749, "step": 21310 }, { "grad_norm": 0.8801551461219788, "learning_rate": 9.289418528432655e-05, "loss": 0.0664, "step": 21320 }, { "grad_norm": 0.8427965044975281, "learning_rate": 9.288568668536045e-05, "loss": 0.0908, "step": 21330 }, { "grad_norm": 0.5733412504196167, "learning_rate": 9.287718339648013e-05, "loss": 0.0718, "step": 21340 }, { "grad_norm": 0.912251353263855, "learning_rate": 9.28686754186155e-05, "loss": 0.0766, "step": 21350 }, { "grad_norm": 0.5120090842247009, "learning_rate": 9.286016275269698e-05, "loss": 0.09, "step": 21360 }, { "grad_norm": 1.1201881170272827, "learning_rate": 9.285164539965551e-05, "loss": 0.0779, "step": 21370 }, { "grad_norm": 0.45272621512413025, "learning_rate": 9.284312336042251e-05, "loss": 0.0684, "step": 21380 }, { "grad_norm": 0.7768360376358032, "learning_rate": 9.283459663592996e-05, "loss": 0.0723, "step": 21390 }, { "grad_norm": 0.5599491596221924, "learning_rate": 9.282606522711033e-05, "loss": 0.0703, "step": 21400 }, { "grad_norm": 0.5268146991729736, "learning_rate": 9.281752913489657e-05, "loss": 0.081, "step": 21410 }, { "grad_norm": 0.6574068665504456, "learning_rate": 9.280898836022222e-05, "loss": 0.0734, "step": 21420 }, { "grad_norm": 0.502740204334259, "learning_rate": 9.280044290402126e-05, "loss": 0.0685, "step": 21430 }, { "grad_norm": 0.6985101103782654, "learning_rate": 9.279189276722821e-05, "loss": 0.0655, "step": 21440 }, { "grad_norm": 0.8710575699806213, "learning_rate": 9.278333795077812e-05, "loss": 0.0711, "step": 21450 }, { "grad_norm": 0.5984207987785339, "learning_rate": 9.27747784556065e-05, "loss": 0.0589, "step": 21460 }, { "grad_norm": 0.6749893426895142, "learning_rate": 9.276621428264942e-05, "loss": 0.0795, "step": 21470 }, { "grad_norm": 0.49955660104751587, "learning_rate": 9.275764543284345e-05, "loss": 0.0684, "step": 21480 }, { "grad_norm": 0.5313080549240112, "learning_rate": 9.274907190712566e-05, "loss": 0.0635, "step": 21490 }, { "grad_norm": 0.6599202156066895, "learning_rate": 9.274049370643363e-05, "loss": 0.078, "step": 21500 }, { "grad_norm": 0.7030637860298157, "learning_rate": 9.273191083170547e-05, "loss": 0.0797, "step": 21510 }, { "grad_norm": 0.5003170371055603, "learning_rate": 9.27233232838798e-05, "loss": 0.065, "step": 21520 }, { "grad_norm": 0.6184267997741699, "learning_rate": 9.27147310638957e-05, "loss": 0.066, "step": 21530 }, { "grad_norm": 0.3859175741672516, "learning_rate": 9.270613417269286e-05, "loss": 0.0794, "step": 21540 }, { "grad_norm": 0.5306910276412964, "learning_rate": 9.269753261121138e-05, "loss": 0.0721, "step": 21550 }, { "grad_norm": 0.38751575350761414, "learning_rate": 9.268892638039194e-05, "loss": 0.0678, "step": 21560 }, { "grad_norm": 0.5696197152137756, "learning_rate": 9.268031548117569e-05, "loss": 0.0778, "step": 21570 }, { "grad_norm": 0.5089293122291565, "learning_rate": 9.26716999145043e-05, "loss": 0.0623, "step": 21580 }, { "grad_norm": 0.7071417570114136, "learning_rate": 9.266307968131998e-05, "loss": 0.0648, "step": 21590 }, { "grad_norm": 2.082239866256714, "learning_rate": 9.26544547825654e-05, "loss": 0.0727, "step": 21600 }, { "grad_norm": 0.4754975736141205, "learning_rate": 9.264582521918376e-05, "loss": 0.0758, "step": 21610 }, { "grad_norm": 0.6888373494148254, "learning_rate": 9.263719099211881e-05, "loss": 0.0637, "step": 21620 }, { "grad_norm": 0.4887067377567291, "learning_rate": 9.262855210231476e-05, "loss": 0.0804, "step": 21630 }, { "grad_norm": 0.4295872449874878, "learning_rate": 9.261990855071633e-05, "loss": 0.0662, "step": 21640 }, { "grad_norm": 0.6906002759933472, "learning_rate": 9.261126033826878e-05, "loss": 0.0631, "step": 21650 }, { "grad_norm": 0.41437625885009766, "learning_rate": 9.260260746591786e-05, "loss": 0.0629, "step": 21660 }, { "grad_norm": 0.5872179865837097, "learning_rate": 9.259394993460985e-05, "loss": 0.0682, "step": 21670 }, { "grad_norm": 0.5951407551765442, "learning_rate": 9.258528774529151e-05, "loss": 0.0693, "step": 21680 }, { "grad_norm": 0.5376110672950745, "learning_rate": 9.257662089891013e-05, "loss": 0.0679, "step": 21690 }, { "grad_norm": 0.8201741576194763, "learning_rate": 9.25679493964135e-05, "loss": 0.0757, "step": 21700 }, { "grad_norm": 0.40282726287841797, "learning_rate": 9.255927323874994e-05, "loss": 0.0599, "step": 21710 }, { "grad_norm": 0.6260966658592224, "learning_rate": 9.255059242686822e-05, "loss": 0.0855, "step": 21720 }, { "grad_norm": 0.5155765414237976, "learning_rate": 9.254190696171769e-05, "loss": 0.0681, "step": 21730 }, { "grad_norm": 0.6965059638023376, "learning_rate": 9.25332168442482e-05, "loss": 0.1064, "step": 21740 }, { "grad_norm": 0.5741137266159058, "learning_rate": 9.252452207541004e-05, "loss": 0.061, "step": 21750 }, { "grad_norm": 0.5469420552253723, "learning_rate": 9.251582265615409e-05, "loss": 0.0811, "step": 21760 }, { "grad_norm": 0.5300350189208984, "learning_rate": 9.250711858743169e-05, "loss": 0.0828, "step": 21770 }, { "grad_norm": 0.5521745681762695, "learning_rate": 9.24984098701947e-05, "loss": 0.0883, "step": 21780 }, { "grad_norm": 0.717246949672699, "learning_rate": 9.248969650539552e-05, "loss": 0.0755, "step": 21790 }, { "grad_norm": 0.6494486927986145, "learning_rate": 9.2480978493987e-05, "loss": 0.0726, "step": 21800 }, { "grad_norm": 0.6103873252868652, "learning_rate": 9.247225583692256e-05, "loss": 0.0739, "step": 21810 }, { "grad_norm": 0.5856146812438965, "learning_rate": 9.246352853515607e-05, "loss": 0.0729, "step": 21820 }, { "grad_norm": 0.6581018567085266, "learning_rate": 9.245479658964194e-05, "loss": 0.0694, "step": 21830 }, { "grad_norm": 0.792820394039154, "learning_rate": 9.244606000133507e-05, "loss": 0.0749, "step": 21840 }, { "grad_norm": 0.4333091974258423, "learning_rate": 9.24373187711909e-05, "loss": 0.0663, "step": 21850 }, { "grad_norm": 0.7121099233627319, "learning_rate": 9.242857290016537e-05, "loss": 0.0747, "step": 21860 }, { "grad_norm": 0.5402088165283203, "learning_rate": 9.241982238921488e-05, "loss": 0.0814, "step": 21870 }, { "grad_norm": 0.558687686920166, "learning_rate": 9.24110672392964e-05, "loss": 0.0589, "step": 21880 }, { "grad_norm": 0.44465360045433044, "learning_rate": 9.240230745136737e-05, "loss": 0.0709, "step": 21890 }, { "grad_norm": 0.5537315011024475, "learning_rate": 9.239354302638575e-05, "loss": 0.067, "step": 21900 }, { "grad_norm": 0.729340672492981, "learning_rate": 9.238477396531e-05, "loss": 0.075, "step": 21910 }, { "grad_norm": 0.45108792185783386, "learning_rate": 9.23760002690991e-05, "loss": 0.0675, "step": 21920 }, { "grad_norm": 0.4283386468887329, "learning_rate": 9.236722193871252e-05, "loss": 0.0643, "step": 21930 }, { "grad_norm": 0.4822402000427246, "learning_rate": 9.235843897511023e-05, "loss": 0.0702, "step": 21940 }, { "grad_norm": 1.3001490831375122, "learning_rate": 9.234965137925276e-05, "loss": 0.0591, "step": 21950 }, { "grad_norm": 0.5941050052642822, "learning_rate": 9.234085915210108e-05, "loss": 0.0867, "step": 21960 }, { "grad_norm": 0.6044078469276428, "learning_rate": 9.23320622946167e-05, "loss": 0.0774, "step": 21970 }, { "grad_norm": 0.6283558011054993, "learning_rate": 9.232326080776163e-05, "loss": 0.0612, "step": 21980 }, { "grad_norm": 0.3947228193283081, "learning_rate": 9.23144546924984e-05, "loss": 0.0622, "step": 21990 }, { "grad_norm": 0.9034886360168457, "learning_rate": 9.230564394979e-05, "loss": 0.0678, "step": 22000 }, { "grad_norm": 0.7508407235145569, "learning_rate": 9.22968285806e-05, "loss": 0.0564, "step": 22010 }, { "grad_norm": 0.5760527849197388, "learning_rate": 9.228800858589242e-05, "loss": 0.075, "step": 22020 }, { "grad_norm": 0.3585169017314911, "learning_rate": 9.227918396663179e-05, "loss": 0.0713, "step": 22030 }, { "grad_norm": 0.9684298038482666, "learning_rate": 9.227035472378319e-05, "loss": 0.0738, "step": 22040 }, { "grad_norm": 0.5035691857337952, "learning_rate": 9.226152085831213e-05, "loss": 0.0752, "step": 22050 }, { "grad_norm": 0.615294337272644, "learning_rate": 9.22526823711847e-05, "loss": 0.068, "step": 22060 }, { "grad_norm": 0.5224125981330872, "learning_rate": 9.224383926336745e-05, "loss": 0.0629, "step": 22070 }, { "grad_norm": 0.48196837306022644, "learning_rate": 9.223499153582744e-05, "loss": 0.051, "step": 22080 }, { "grad_norm": 0.571281909942627, "learning_rate": 9.222613918953226e-05, "loss": 0.0659, "step": 22090 }, { "grad_norm": 0.5656936764717102, "learning_rate": 9.221728222544999e-05, "loss": 0.0701, "step": 22100 }, { "grad_norm": 0.9329339861869812, "learning_rate": 9.22084206445492e-05, "loss": 0.0814, "step": 22110 }, { "grad_norm": 0.48253196477890015, "learning_rate": 9.2199554447799e-05, "loss": 0.0831, "step": 22120 }, { "grad_norm": 0.6520648002624512, "learning_rate": 9.219068363616897e-05, "loss": 0.0754, "step": 22130 }, { "grad_norm": 0.7930215001106262, "learning_rate": 9.218180821062919e-05, "loss": 0.0744, "step": 22140 }, { "grad_norm": 0.7242578268051147, "learning_rate": 9.21729281721503e-05, "loss": 0.0853, "step": 22150 }, { "grad_norm": 0.7376756072044373, "learning_rate": 9.216404352170339e-05, "loss": 0.0971, "step": 22160 }, { "grad_norm": 0.6041449308395386, "learning_rate": 9.215515426026007e-05, "loss": 0.077, "step": 22170 }, { "grad_norm": 0.47033989429473877, "learning_rate": 9.214626038879246e-05, "loss": 0.0699, "step": 22180 }, { "grad_norm": 0.6678639650344849, "learning_rate": 9.21373619082732e-05, "loss": 0.0665, "step": 22190 }, { "grad_norm": 1.3616173267364502, "learning_rate": 9.212845881967535e-05, "loss": 0.0721, "step": 22200 }, { "grad_norm": 0.7007453441619873, "learning_rate": 9.211955112397262e-05, "loss": 0.0715, "step": 22210 }, { "grad_norm": 0.3933253586292267, "learning_rate": 9.211063882213909e-05, "loss": 0.0726, "step": 22220 }, { "grad_norm": 0.5600783228874207, "learning_rate": 9.210172191514942e-05, "loss": 0.0694, "step": 22230 }, { "grad_norm": 0.5197253227233887, "learning_rate": 9.209280040397874e-05, "loss": 0.0722, "step": 22240 }, { "grad_norm": 0.5320191383361816, "learning_rate": 9.208387428960268e-05, "loss": 0.0535, "step": 22250 }, { "grad_norm": 0.5776503682136536, "learning_rate": 9.20749435729974e-05, "loss": 0.0715, "step": 22260 }, { "grad_norm": 0.6567931771278381, "learning_rate": 9.206600825513957e-05, "loss": 0.0726, "step": 22270 }, { "grad_norm": 0.4489932656288147, "learning_rate": 9.20570683370063e-05, "loss": 0.0793, "step": 22280 }, { "grad_norm": 0.6843447089195251, "learning_rate": 9.204812381957528e-05, "loss": 0.0659, "step": 22290 }, { "grad_norm": 0.615554690361023, "learning_rate": 9.203917470382465e-05, "loss": 0.0629, "step": 22300 }, { "grad_norm": 0.5577470064163208, "learning_rate": 9.203022099073309e-05, "loss": 0.0652, "step": 22310 }, { "grad_norm": 0.7082617878913879, "learning_rate": 9.202126268127976e-05, "loss": 0.0727, "step": 22320 }, { "grad_norm": 0.482645183801651, "learning_rate": 9.20122997764443e-05, "loss": 0.065, "step": 22330 }, { "grad_norm": 1.050502896308899, "learning_rate": 9.200333227720692e-05, "loss": 0.0755, "step": 22340 }, { "grad_norm": 0.840602457523346, "learning_rate": 9.199436018454826e-05, "loss": 0.0785, "step": 22350 }, { "grad_norm": 0.5280869007110596, "learning_rate": 9.198538349944952e-05, "loss": 0.0678, "step": 22360 }, { "grad_norm": 0.6344373822212219, "learning_rate": 9.197640222289234e-05, "loss": 0.0832, "step": 22370 }, { "grad_norm": 0.5259285569190979, "learning_rate": 9.196741635585895e-05, "loss": 0.0808, "step": 22380 }, { "grad_norm": 0.5466430187225342, "learning_rate": 9.195842589933199e-05, "loss": 0.0621, "step": 22390 }, { "grad_norm": 0.5970836281776428, "learning_rate": 9.194943085429466e-05, "loss": 0.081, "step": 22400 }, { "grad_norm": 0.4652928411960602, "learning_rate": 9.194043122173065e-05, "loss": 0.0907, "step": 22410 }, { "grad_norm": 0.5567722320556641, "learning_rate": 9.193142700262413e-05, "loss": 0.0643, "step": 22420 }, { "grad_norm": 0.5099332928657532, "learning_rate": 9.192241819795979e-05, "loss": 0.0582, "step": 22430 }, { "grad_norm": 0.6040326952934265, "learning_rate": 9.191340480872284e-05, "loss": 0.0685, "step": 22440 }, { "grad_norm": 0.5610150098800659, "learning_rate": 9.190438683589895e-05, "loss": 0.0848, "step": 22450 }, { "grad_norm": 0.4758712351322174, "learning_rate": 9.189536428047432e-05, "loss": 0.0698, "step": 22460 }, { "grad_norm": 0.3899894952774048, "learning_rate": 9.188633714343564e-05, "loss": 0.065, "step": 22470 }, { "grad_norm": 0.9366886615753174, "learning_rate": 9.18773054257701e-05, "loss": 0.07, "step": 22480 }, { "grad_norm": 0.6252238750457764, "learning_rate": 9.18682691284654e-05, "loss": 0.0639, "step": 22490 }, { "grad_norm": 0.517499566078186, "learning_rate": 9.185922825250974e-05, "loss": 0.0677, "step": 22500 }, { "grad_norm": 0.6214223504066467, "learning_rate": 9.185018279889181e-05, "loss": 0.052, "step": 22510 }, { "grad_norm": 0.7108891010284424, "learning_rate": 9.184113276860082e-05, "loss": 0.0685, "step": 22520 }, { "grad_norm": 0.6580506563186646, "learning_rate": 9.183207816262645e-05, "loss": 0.0769, "step": 22530 }, { "grad_norm": 0.6170492768287659, "learning_rate": 9.182301898195891e-05, "loss": 0.0731, "step": 22540 }, { "grad_norm": 0.4463775157928467, "learning_rate": 9.181395522758889e-05, "loss": 0.0811, "step": 22550 }, { "grad_norm": 0.6527429223060608, "learning_rate": 9.180488690050759e-05, "loss": 0.0881, "step": 22560 }, { "grad_norm": 0.4577351212501526, "learning_rate": 9.179581400170671e-05, "loss": 0.0671, "step": 22570 }, { "grad_norm": 0.8876394629478455, "learning_rate": 9.178673653217845e-05, "loss": 0.0685, "step": 22580 }, { "grad_norm": 0.5793836116790771, "learning_rate": 9.177765449291551e-05, "loss": 0.0643, "step": 22590 }, { "grad_norm": 1.6239930391311646, "learning_rate": 9.176856788491109e-05, "loss": 0.0859, "step": 22600 }, { "grad_norm": 0.5149936676025391, "learning_rate": 9.175947670915887e-05, "loss": 0.0732, "step": 22610 }, { "grad_norm": 0.9757896661758423, "learning_rate": 9.175038096665309e-05, "loss": 0.0694, "step": 22620 }, { "grad_norm": 0.5944763422012329, "learning_rate": 9.17412806583884e-05, "loss": 0.059, "step": 22630 }, { "grad_norm": 0.6823983788490295, "learning_rate": 9.173217578536002e-05, "loss": 0.0878, "step": 22640 }, { "grad_norm": 0.5810658931732178, "learning_rate": 9.172306634856362e-05, "loss": 0.0727, "step": 22650 }, { "grad_norm": 0.5512166023254395, "learning_rate": 9.171395234899545e-05, "loss": 0.088, "step": 22660 }, { "grad_norm": 0.6423920392990112, "learning_rate": 9.170483378765214e-05, "loss": 0.0801, "step": 22670 }, { "grad_norm": 0.5792227983474731, "learning_rate": 9.169571066553091e-05, "loss": 0.069, "step": 22680 }, { "grad_norm": 0.5673860311508179, "learning_rate": 9.168658298362946e-05, "loss": 0.0796, "step": 22690 }, { "grad_norm": 0.6346651315689087, "learning_rate": 9.167745074294598e-05, "loss": 0.0756, "step": 22700 }, { "grad_norm": 0.5670211911201477, "learning_rate": 9.166831394447913e-05, "loss": 0.0716, "step": 22710 }, { "grad_norm": 0.5113445520401001, "learning_rate": 9.165917258922812e-05, "loss": 0.0579, "step": 22720 }, { "grad_norm": 0.6370770931243896, "learning_rate": 9.165002667819262e-05, "loss": 0.0687, "step": 22730 }, { "grad_norm": 0.5129058361053467, "learning_rate": 9.164087621237282e-05, "loss": 0.0587, "step": 22740 }, { "grad_norm": 0.5612324476242065, "learning_rate": 9.163172119276942e-05, "loss": 0.0798, "step": 22750 }, { "grad_norm": 0.4840638041496277, "learning_rate": 9.162256162038358e-05, "loss": 0.0781, "step": 22760 }, { "grad_norm": 0.6822614669799805, "learning_rate": 9.161339749621698e-05, "loss": 0.071, "step": 22770 }, { "grad_norm": 0.4178991913795471, "learning_rate": 9.160422882127177e-05, "loss": 0.0729, "step": 22780 }, { "grad_norm": 0.5456311106681824, "learning_rate": 9.159505559655069e-05, "loss": 0.0742, "step": 22790 }, { "grad_norm": 0.6804934740066528, "learning_rate": 9.158587782305684e-05, "loss": 0.0694, "step": 22800 }, { "grad_norm": 0.4763110876083374, "learning_rate": 9.157669550179391e-05, "loss": 0.0803, "step": 22810 }, { "grad_norm": 0.36339402198791504, "learning_rate": 9.156750863376609e-05, "loss": 0.0764, "step": 22820 }, { "grad_norm": 0.6534126996994019, "learning_rate": 9.155831721997801e-05, "loss": 0.0802, "step": 22830 }, { "grad_norm": 0.6718156933784485, "learning_rate": 9.154912126143484e-05, "loss": 0.0793, "step": 22840 }, { "grad_norm": 0.635098397731781, "learning_rate": 9.153992075914224e-05, "loss": 0.091, "step": 22850 }, { "grad_norm": 1.3063580989837646, "learning_rate": 9.153071571410635e-05, "loss": 0.0669, "step": 22860 }, { "grad_norm": 0.551820695400238, "learning_rate": 9.152150612733384e-05, "loss": 0.0694, "step": 22870 }, { "grad_norm": 0.5420025587081909, "learning_rate": 9.151229199983184e-05, "loss": 0.0673, "step": 22880 }, { "grad_norm": 0.5504302382469177, "learning_rate": 9.150307333260802e-05, "loss": 0.0723, "step": 22890 }, { "grad_norm": 0.5541456937789917, "learning_rate": 9.149385012667048e-05, "loss": 0.0833, "step": 22900 }, { "grad_norm": 0.5849874019622803, "learning_rate": 9.148462238302788e-05, "loss": 0.086, "step": 22910 }, { "grad_norm": 0.7844039797782898, "learning_rate": 9.147539010268936e-05, "loss": 0.0828, "step": 22920 }, { "grad_norm": 0.8348471522331238, "learning_rate": 9.14661532866645e-05, "loss": 0.0812, "step": 22930 }, { "grad_norm": 0.654809296131134, "learning_rate": 9.145691193596348e-05, "loss": 0.0661, "step": 22940 }, { "grad_norm": 0.5286624431610107, "learning_rate": 9.144766605159691e-05, "loss": 0.0971, "step": 22950 }, { "grad_norm": 0.6891611814498901, "learning_rate": 9.14384156345759e-05, "loss": 0.0758, "step": 22960 }, { "grad_norm": 0.7546586990356445, "learning_rate": 9.142916068591204e-05, "loss": 0.0666, "step": 22970 }, { "grad_norm": 0.38579055666923523, "learning_rate": 9.141990120661746e-05, "loss": 0.0693, "step": 22980 }, { "grad_norm": 0.4272345006465912, "learning_rate": 9.141063719770475e-05, "loss": 0.0738, "step": 22990 }, { "grad_norm": 0.4069660007953644, "learning_rate": 9.140136866018704e-05, "loss": 0.0668, "step": 23000 }, { "grad_norm": 0.4006786644458771, "learning_rate": 9.139209559507788e-05, "loss": 0.0641, "step": 23010 }, { "grad_norm": 0.5314484238624573, "learning_rate": 9.13828180033914e-05, "loss": 0.0622, "step": 23020 }, { "grad_norm": 0.5826640725135803, "learning_rate": 9.137353588614212e-05, "loss": 0.0679, "step": 23030 }, { "grad_norm": 0.4583098888397217, "learning_rate": 9.136424924434519e-05, "loss": 0.0713, "step": 23040 }, { "grad_norm": 0.7545438408851624, "learning_rate": 9.135495807901615e-05, "loss": 0.0743, "step": 23050 }, { "grad_norm": 0.3834001421928406, "learning_rate": 9.134566239117108e-05, "loss": 0.0733, "step": 23060 }, { "grad_norm": 1.0650337934494019, "learning_rate": 9.13363621818265e-05, "loss": 0.0673, "step": 23070 }, { "grad_norm": 0.6203344464302063, "learning_rate": 9.132705745199953e-05, "loss": 0.0779, "step": 23080 }, { "grad_norm": 0.5452418923377991, "learning_rate": 9.131774820270768e-05, "loss": 0.0707, "step": 23090 }, { "grad_norm": 0.7124102711677551, "learning_rate": 9.130843443496901e-05, "loss": 0.0807, "step": 23100 }, { "grad_norm": 1.6085366010665894, "learning_rate": 9.129911614980206e-05, "loss": 0.0756, "step": 23110 }, { "grad_norm": 0.5737872123718262, "learning_rate": 9.128979334822584e-05, "loss": 0.0636, "step": 23120 }, { "grad_norm": 1.0275359153747559, "learning_rate": 9.128046603125992e-05, "loss": 0.0603, "step": 23130 }, { "grad_norm": 0.5517776608467102, "learning_rate": 9.12711341999243e-05, "loss": 0.0676, "step": 23140 }, { "grad_norm": 0.516747236251831, "learning_rate": 9.12617978552395e-05, "loss": 0.0895, "step": 23150 }, { "grad_norm": 0.43453213572502136, "learning_rate": 9.12524569982265e-05, "loss": 0.0613, "step": 23160 }, { "grad_norm": 0.5345939993858337, "learning_rate": 9.124311162990684e-05, "loss": 0.0654, "step": 23170 }, { "grad_norm": 0.4776117503643036, "learning_rate": 9.12337617513025e-05, "loss": 0.0659, "step": 23180 }, { "grad_norm": 0.4859175980091095, "learning_rate": 9.122440736343596e-05, "loss": 0.0573, "step": 23190 }, { "grad_norm": 0.42709818482398987, "learning_rate": 9.12150484673302e-05, "loss": 0.0647, "step": 23200 }, { "grad_norm": 0.43939128518104553, "learning_rate": 9.120568506400873e-05, "loss": 0.0671, "step": 23210 }, { "grad_norm": 0.4509831666946411, "learning_rate": 9.119631715449548e-05, "loss": 0.0711, "step": 23220 }, { "grad_norm": 0.5125412344932556, "learning_rate": 9.118694473981493e-05, "loss": 0.0672, "step": 23230 }, { "grad_norm": 0.5924525856971741, "learning_rate": 9.117756782099203e-05, "loss": 0.0846, "step": 23240 }, { "grad_norm": 0.47413191199302673, "learning_rate": 9.11681863990522e-05, "loss": 0.071, "step": 23250 }, { "grad_norm": 0.963509202003479, "learning_rate": 9.115880047502142e-05, "loss": 0.0699, "step": 23260 }, { "grad_norm": 0.6105724573135376, "learning_rate": 9.114941004992609e-05, "loss": 0.0842, "step": 23270 }, { "grad_norm": 1.0233384370803833, "learning_rate": 9.114001512479317e-05, "loss": 0.0725, "step": 23280 }, { "grad_norm": 0.6097882390022278, "learning_rate": 9.113061570065003e-05, "loss": 0.0867, "step": 23290 }, { "grad_norm": 0.5289575457572937, "learning_rate": 9.112121177852459e-05, "loss": 0.0657, "step": 23300 }, { "grad_norm": 0.5345675945281982, "learning_rate": 9.111180335944527e-05, "loss": 0.0738, "step": 23310 }, { "grad_norm": 0.7447251677513123, "learning_rate": 9.110239044444093e-05, "loss": 0.0696, "step": 23320 }, { "grad_norm": 0.5806578993797302, "learning_rate": 9.109297303454099e-05, "loss": 0.0726, "step": 23330 }, { "grad_norm": 0.5673043131828308, "learning_rate": 9.108355113077526e-05, "loss": 0.0677, "step": 23340 }, { "grad_norm": 0.6962233781814575, "learning_rate": 9.107412473417419e-05, "loss": 0.081, "step": 23350 }, { "grad_norm": 0.574885368347168, "learning_rate": 9.106469384576858e-05, "loss": 0.0742, "step": 23360 }, { "grad_norm": 1.057310938835144, "learning_rate": 9.105525846658978e-05, "loss": 0.0914, "step": 23370 }, { "grad_norm": 0.44544073939323425, "learning_rate": 9.104581859766965e-05, "loss": 0.0851, "step": 23380 }, { "grad_norm": 0.4783451557159424, "learning_rate": 9.10363742400405e-05, "loss": 0.0737, "step": 23390 }, { "grad_norm": 0.488414466381073, "learning_rate": 9.102692539473518e-05, "loss": 0.0763, "step": 23400 }, { "grad_norm": 0.5915192365646362, "learning_rate": 9.101747206278697e-05, "loss": 0.076, "step": 23410 }, { "grad_norm": 0.4540729224681854, "learning_rate": 9.100801424522968e-05, "loss": 0.0657, "step": 23420 }, { "grad_norm": 0.5198405385017395, "learning_rate": 9.099855194309762e-05, "loss": 0.0703, "step": 23430 }, { "grad_norm": 0.5262864828109741, "learning_rate": 9.098908515742554e-05, "loss": 0.0642, "step": 23440 }, { "grad_norm": 0.5190039873123169, "learning_rate": 9.097961388924873e-05, "loss": 0.0689, "step": 23450 }, { "grad_norm": 0.6127952337265015, "learning_rate": 9.097013813960298e-05, "loss": 0.0786, "step": 23460 }, { "grad_norm": 0.8049508333206177, "learning_rate": 9.09606579095245e-05, "loss": 0.0751, "step": 23470 }, { "grad_norm": 0.6065462231636047, "learning_rate": 9.095117320005008e-05, "loss": 0.0682, "step": 23480 }, { "grad_norm": 0.648912250995636, "learning_rate": 9.094168401221691e-05, "loss": 0.0701, "step": 23490 }, { "grad_norm": 0.9909597039222717, "learning_rate": 9.093219034706273e-05, "loss": 0.0886, "step": 23500 }, { "grad_norm": 1.278016448020935, "learning_rate": 9.092269220562577e-05, "loss": 0.084, "step": 23510 }, { "grad_norm": 0.5910999774932861, "learning_rate": 9.09131895889447e-05, "loss": 0.0989, "step": 23520 }, { "grad_norm": 0.4822777807712555, "learning_rate": 9.090368249805873e-05, "loss": 0.0704, "step": 23530 }, { "grad_norm": 0.4572598934173584, "learning_rate": 9.089417093400754e-05, "loss": 0.078, "step": 23540 }, { "grad_norm": 0.43830907344818115, "learning_rate": 9.088465489783131e-05, "loss": 0.0575, "step": 23550 }, { "grad_norm": 0.533790647983551, "learning_rate": 9.087513439057068e-05, "loss": 0.0703, "step": 23560 }, { "grad_norm": 0.5753647685050964, "learning_rate": 9.08656094132668e-05, "loss": 0.0729, "step": 23570 }, { "grad_norm": 0.6675796508789062, "learning_rate": 9.085607996696134e-05, "loss": 0.0701, "step": 23580 }, { "grad_norm": 0.6593713164329529, "learning_rate": 9.084654605269639e-05, "loss": 0.0935, "step": 23590 }, { "grad_norm": 0.5327045321464539, "learning_rate": 9.083700767151457e-05, "loss": 0.0719, "step": 23600 }, { "grad_norm": 0.2997060716152191, "learning_rate": 9.082746482445898e-05, "loss": 0.0622, "step": 23610 }, { "grad_norm": 0.7485368847846985, "learning_rate": 9.081791751257325e-05, "loss": 0.0647, "step": 23620 }, { "grad_norm": 0.5574759244918823, "learning_rate": 9.080836573690142e-05, "loss": 0.0695, "step": 23630 }, { "grad_norm": 0.586402952671051, "learning_rate": 9.079880949848805e-05, "loss": 0.0799, "step": 23640 }, { "grad_norm": 0.5810441970825195, "learning_rate": 9.078924879837822e-05, "loss": 0.0652, "step": 23650 }, { "grad_norm": 0.7350583672523499, "learning_rate": 9.077968363761747e-05, "loss": 0.0784, "step": 23660 }, { "grad_norm": 0.5314158797264099, "learning_rate": 9.077011401725182e-05, "loss": 0.0579, "step": 23670 }, { "grad_norm": 0.5020055770874023, "learning_rate": 9.07605399383278e-05, "loss": 0.0759, "step": 23680 }, { "grad_norm": 0.5487004518508911, "learning_rate": 9.075096140189243e-05, "loss": 0.0757, "step": 23690 }, { "grad_norm": 0.6495032906532288, "learning_rate": 9.074137840899318e-05, "loss": 0.0618, "step": 23700 }, { "grad_norm": 1.1865589618682861, "learning_rate": 9.073179096067804e-05, "loss": 0.0817, "step": 23710 }, { "grad_norm": 0.4871361553668976, "learning_rate": 9.072219905799549e-05, "loss": 0.0645, "step": 23720 }, { "grad_norm": 0.590543270111084, "learning_rate": 9.071260270199447e-05, "loss": 0.0837, "step": 23730 }, { "grad_norm": 0.7273603081703186, "learning_rate": 9.070300189372441e-05, "loss": 0.0739, "step": 23740 }, { "grad_norm": 0.7302895188331604, "learning_rate": 9.069339663423528e-05, "loss": 0.0673, "step": 23750 }, { "grad_norm": 0.628010094165802, "learning_rate": 9.068378692457747e-05, "loss": 0.0769, "step": 23760 }, { "grad_norm": 0.4359516203403473, "learning_rate": 9.067417276580189e-05, "loss": 0.068, "step": 23770 }, { "grad_norm": 0.5062977075576782, "learning_rate": 9.066455415895993e-05, "loss": 0.0584, "step": 23780 }, { "grad_norm": 0.5072382092475891, "learning_rate": 9.065493110510346e-05, "loss": 0.0775, "step": 23790 }, { "grad_norm": 0.6112146377563477, "learning_rate": 9.064530360528484e-05, "loss": 0.0738, "step": 23800 }, { "grad_norm": 0.5129216909408569, "learning_rate": 9.063567166055695e-05, "loss": 0.0562, "step": 23810 }, { "grad_norm": 0.5888888239860535, "learning_rate": 9.062603527197308e-05, "loss": 0.066, "step": 23820 }, { "grad_norm": 0.4272416830062866, "learning_rate": 9.06163944405871e-05, "loss": 0.0662, "step": 23830 }, { "grad_norm": 0.9584103226661682, "learning_rate": 9.060674916745327e-05, "loss": 0.0765, "step": 23840 }, { "grad_norm": 0.43543413281440735, "learning_rate": 9.05970994536264e-05, "loss": 0.063, "step": 23850 }, { "grad_norm": 1.2173842191696167, "learning_rate": 9.05874453001618e-05, "loss": 0.0688, "step": 23860 }, { "grad_norm": 1.0064053535461426, "learning_rate": 9.057778670811517e-05, "loss": 0.0722, "step": 23870 }, { "grad_norm": 0.6481438279151917, "learning_rate": 9.056812367854281e-05, "loss": 0.0671, "step": 23880 }, { "grad_norm": 0.5619135499000549, "learning_rate": 9.055845621250143e-05, "loss": 0.0661, "step": 23890 }, { "grad_norm": 0.46291545033454895, "learning_rate": 9.054878431104825e-05, "loss": 0.07, "step": 23900 }, { "grad_norm": 0.5452505350112915, "learning_rate": 9.0539107975241e-05, "loss": 0.0733, "step": 23910 }, { "grad_norm": 0.680654764175415, "learning_rate": 9.052942720613784e-05, "loss": 0.0676, "step": 23920 }, { "grad_norm": 0.43976157903671265, "learning_rate": 9.051974200479745e-05, "loss": 0.0596, "step": 23930 }, { "grad_norm": 0.5601927638053894, "learning_rate": 9.051005237227901e-05, "loss": 0.0477, "step": 23940 }, { "grad_norm": 0.6342421770095825, "learning_rate": 9.050035830964215e-05, "loss": 0.0559, "step": 23950 }, { "grad_norm": 0.7990930676460266, "learning_rate": 9.049065981794698e-05, "loss": 0.0871, "step": 23960 }, { "grad_norm": 0.6011123657226562, "learning_rate": 9.048095689825414e-05, "loss": 0.0681, "step": 23970 }, { "grad_norm": 0.6046445965766907, "learning_rate": 9.047124955162472e-05, "loss": 0.0624, "step": 23980 }, { "grad_norm": 0.8508060574531555, "learning_rate": 9.046153777912028e-05, "loss": 0.0731, "step": 23990 }, { "grad_norm": 0.8190176486968994, "learning_rate": 9.045182158180292e-05, "loss": 0.0644, "step": 24000 }, { "grad_norm": 0.515546441078186, "learning_rate": 9.044210096073516e-05, "loss": 0.058, "step": 24010 }, { "grad_norm": 0.5194327235221863, "learning_rate": 9.043237591698004e-05, "loss": 0.0652, "step": 24020 }, { "grad_norm": 0.4104349613189697, "learning_rate": 9.04226464516011e-05, "loss": 0.06, "step": 24030 }, { "grad_norm": 0.6642307639122009, "learning_rate": 9.041291256566229e-05, "loss": 0.078, "step": 24040 }, { "grad_norm": 0.5615957975387573, "learning_rate": 9.040317426022814e-05, "loss": 0.0617, "step": 24050 }, { "grad_norm": 0.7897912859916687, "learning_rate": 9.03934315363636e-05, "loss": 0.075, "step": 24060 }, { "grad_norm": 0.5813153386116028, "learning_rate": 9.038368439513409e-05, "loss": 0.0575, "step": 24070 }, { "grad_norm": 0.5457560420036316, "learning_rate": 9.03739328376056e-05, "loss": 0.0684, "step": 24080 }, { "grad_norm": 0.4722915589809418, "learning_rate": 9.036417686484451e-05, "loss": 0.0648, "step": 24090 }, { "grad_norm": 0.4515695869922638, "learning_rate": 9.035441647791773e-05, "loss": 0.058, "step": 24100 }, { "grad_norm": 0.44274789094924927, "learning_rate": 9.034465167789263e-05, "loss": 0.068, "step": 24110 }, { "grad_norm": 0.5339471101760864, "learning_rate": 9.033488246583706e-05, "loss": 0.0648, "step": 24120 }, { "grad_norm": 0.4226186275482178, "learning_rate": 9.032510884281941e-05, "loss": 0.0728, "step": 24130 }, { "grad_norm": 0.3956390917301178, "learning_rate": 9.031533080990848e-05, "loss": 0.0711, "step": 24140 }, { "grad_norm": 0.7249855995178223, "learning_rate": 9.030554836817358e-05, "loss": 0.0612, "step": 24150 }, { "grad_norm": 0.4969235062599182, "learning_rate": 9.029576151868451e-05, "loss": 0.0593, "step": 24160 }, { "grad_norm": 0.5290864706039429, "learning_rate": 9.028597026251155e-05, "loss": 0.0525, "step": 24170 }, { "grad_norm": 0.4007219076156616, "learning_rate": 9.027617460072547e-05, "loss": 0.0755, "step": 24180 }, { "grad_norm": 0.5932977795600891, "learning_rate": 9.026637453439745e-05, "loss": 0.0704, "step": 24190 }, { "grad_norm": 0.5233542323112488, "learning_rate": 9.025657006459927e-05, "loss": 0.0597, "step": 24200 }, { "grad_norm": 0.7257784008979797, "learning_rate": 9.024676119240311e-05, "loss": 0.0652, "step": 24210 }, { "grad_norm": 0.6065930128097534, "learning_rate": 9.023694791888166e-05, "loss": 0.0657, "step": 24220 }, { "grad_norm": 0.43580931425094604, "learning_rate": 9.022713024510808e-05, "loss": 0.0693, "step": 24230 }, { "grad_norm": 0.6735249161720276, "learning_rate": 9.021730817215601e-05, "loss": 0.0613, "step": 24240 }, { "grad_norm": 1.2790626287460327, "learning_rate": 9.02074817010996e-05, "loss": 0.0785, "step": 24250 }, { "grad_norm": 0.5338287949562073, "learning_rate": 9.019765083301342e-05, "loss": 0.0625, "step": 24260 }, { "grad_norm": 0.5074265599250793, "learning_rate": 9.01878155689726e-05, "loss": 0.0678, "step": 24270 }, { "grad_norm": 0.580029308795929, "learning_rate": 9.017797591005268e-05, "loss": 0.0607, "step": 24280 }, { "grad_norm": 0.46633127331733704, "learning_rate": 9.016813185732972e-05, "loss": 0.0806, "step": 24290 }, { "grad_norm": 0.470024973154068, "learning_rate": 9.015828341188027e-05, "loss": 0.079, "step": 24300 }, { "grad_norm": 0.8831984400749207, "learning_rate": 9.01484305747813e-05, "loss": 0.0698, "step": 24310 }, { "grad_norm": 0.5907782912254333, "learning_rate": 9.013857334711033e-05, "loss": 0.0662, "step": 24320 }, { "grad_norm": 1.2957544326782227, "learning_rate": 9.012871172994534e-05, "loss": 0.0652, "step": 24330 }, { "grad_norm": 0.5633066296577454, "learning_rate": 9.011884572436476e-05, "loss": 0.0671, "step": 24340 }, { "grad_norm": 0.8661141395568848, "learning_rate": 9.010897533144754e-05, "loss": 0.0711, "step": 24350 }, { "grad_norm": 1.0541670322418213, "learning_rate": 9.009910055227306e-05, "loss": 0.0644, "step": 24360 }, { "grad_norm": 0.6571032404899597, "learning_rate": 9.008922138792124e-05, "loss": 0.0635, "step": 24370 }, { "grad_norm": 0.685255765914917, "learning_rate": 9.007933783947244e-05, "loss": 0.058, "step": 24380 }, { "grad_norm": 0.7245373129844666, "learning_rate": 9.006944990800752e-05, "loss": 0.0568, "step": 24390 }, { "grad_norm": 0.8792479038238525, "learning_rate": 9.005955759460779e-05, "loss": 0.0739, "step": 24400 }, { "grad_norm": 0.624610185623169, "learning_rate": 9.004966090035508e-05, "loss": 0.0614, "step": 24410 }, { "grad_norm": 0.3586078882217407, "learning_rate": 9.003975982633166e-05, "loss": 0.0596, "step": 24420 }, { "grad_norm": 0.5359655022621155, "learning_rate": 9.00298543736203e-05, "loss": 0.0658, "step": 24430 }, { "grad_norm": 1.1323862075805664, "learning_rate": 9.001994454330427e-05, "loss": 0.059, "step": 24440 }, { "grad_norm": 0.5849613547325134, "learning_rate": 9.001003033646727e-05, "loss": 0.0698, "step": 24450 }, { "grad_norm": 0.5165364146232605, "learning_rate": 9.00001117541935e-05, "loss": 0.0708, "step": 24460 }, { "grad_norm": 0.6299338340759277, "learning_rate": 8.999018879756764e-05, "loss": 0.0704, "step": 24470 }, { "grad_norm": 0.5968087911605835, "learning_rate": 8.998026146767487e-05, "loss": 0.06, "step": 24480 }, { "grad_norm": 0.7722697257995605, "learning_rate": 8.99703297656008e-05, "loss": 0.0678, "step": 24490 }, { "grad_norm": 0.44429200887680054, "learning_rate": 8.996039369243156e-05, "loss": 0.0707, "step": 24500 }, { "grad_norm": 0.49944284558296204, "learning_rate": 8.995045324925378e-05, "loss": 0.0534, "step": 24510 }, { "grad_norm": 0.5118542313575745, "learning_rate": 8.994050843715448e-05, "loss": 0.067, "step": 24520 }, { "grad_norm": 0.6140010356903076, "learning_rate": 8.993055925722121e-05, "loss": 0.0588, "step": 24530 }, { "grad_norm": 0.48402857780456543, "learning_rate": 8.992060571054202e-05, "loss": 0.075, "step": 24540 }, { "grad_norm": 0.49299776554107666, "learning_rate": 8.991064779820542e-05, "loss": 0.0622, "step": 24550 }, { "grad_norm": 0.5084620118141174, "learning_rate": 8.990068552130036e-05, "loss": 0.0652, "step": 24560 }, { "grad_norm": 0.6838477849960327, "learning_rate": 8.989071888091634e-05, "loss": 0.0617, "step": 24570 }, { "grad_norm": 0.47722116112709045, "learning_rate": 8.988074787814329e-05, "loss": 0.0535, "step": 24580 }, { "grad_norm": 1.0827277898788452, "learning_rate": 8.987077251407158e-05, "loss": 0.0669, "step": 24590 }, { "grad_norm": 0.6611756086349487, "learning_rate": 8.986079278979216e-05, "loss": 0.0666, "step": 24600 }, { "grad_norm": 0.6189872026443481, "learning_rate": 8.985080870639635e-05, "loss": 0.0784, "step": 24610 }, { "grad_norm": 0.6275587677955627, "learning_rate": 8.984082026497603e-05, "loss": 0.0589, "step": 24620 }, { "grad_norm": 0.6735484600067139, "learning_rate": 8.98308274666235e-05, "loss": 0.0781, "step": 24630 }, { "grad_norm": 0.5567348599433899, "learning_rate": 8.982083031243155e-05, "loss": 0.0675, "step": 24640 }, { "grad_norm": 0.6562938094139099, "learning_rate": 8.98108288034935e-05, "loss": 0.0616, "step": 24650 }, { "grad_norm": 0.5489308834075928, "learning_rate": 8.980082294090305e-05, "loss": 0.0637, "step": 24660 }, { "grad_norm": 0.6250656247138977, "learning_rate": 8.979081272575443e-05, "loss": 0.0731, "step": 24670 }, { "grad_norm": 0.66073077917099, "learning_rate": 8.978079815914236e-05, "loss": 0.0721, "step": 24680 }, { "grad_norm": 0.5938025712966919, "learning_rate": 8.977077924216202e-05, "loss": 0.0841, "step": 24690 }, { "grad_norm": 1.0298758745193481, "learning_rate": 8.976075597590905e-05, "loss": 0.0778, "step": 24700 }, { "grad_norm": 0.714360237121582, "learning_rate": 8.975072836147958e-05, "loss": 0.0604, "step": 24710 }, { "grad_norm": 1.098227858543396, "learning_rate": 8.974069639997025e-05, "loss": 0.0719, "step": 24720 }, { "grad_norm": 0.5083391666412354, "learning_rate": 8.973066009247808e-05, "loss": 0.073, "step": 24730 }, { "grad_norm": 0.5892356038093567, "learning_rate": 8.972061944010066e-05, "loss": 0.0663, "step": 24740 }, { "grad_norm": 0.5115664601325989, "learning_rate": 8.971057444393603e-05, "loss": 0.0665, "step": 24750 }, { "grad_norm": 0.4252820312976837, "learning_rate": 8.970052510508268e-05, "loss": 0.0624, "step": 24760 }, { "grad_norm": 0.4175870716571808, "learning_rate": 8.969047142463959e-05, "loss": 0.0717, "step": 24770 }, { "grad_norm": 0.5463164448738098, "learning_rate": 8.968041340370621e-05, "loss": 0.065, "step": 24780 }, { "grad_norm": 0.628494918346405, "learning_rate": 8.96703510433825e-05, "loss": 0.0671, "step": 24790 }, { "grad_norm": 1.0203299522399902, "learning_rate": 8.966028434476883e-05, "loss": 0.0625, "step": 24800 }, { "grad_norm": 0.5641735196113586, "learning_rate": 8.96502133089661e-05, "loss": 0.0621, "step": 24810 }, { "grad_norm": 1.0222632884979248, "learning_rate": 8.964013793707564e-05, "loss": 0.0702, "step": 24820 }, { "grad_norm": 0.689217746257782, "learning_rate": 8.963005823019932e-05, "loss": 0.0759, "step": 24830 }, { "grad_norm": 0.5532744526863098, "learning_rate": 8.961997418943939e-05, "loss": 0.0602, "step": 24840 }, { "grad_norm": 0.48511072993278503, "learning_rate": 8.960988581589865e-05, "loss": 0.0741, "step": 24850 }, { "grad_norm": 0.6056707501411438, "learning_rate": 8.959979311068037e-05, "loss": 0.057, "step": 24860 }, { "grad_norm": 0.45925071835517883, "learning_rate": 8.958969607488823e-05, "loss": 0.0774, "step": 24870 }, { "grad_norm": 0.8573199510574341, "learning_rate": 8.957959470962647e-05, "loss": 0.0595, "step": 24880 }, { "grad_norm": 0.48454585671424866, "learning_rate": 8.956948901599971e-05, "loss": 0.0643, "step": 24890 }, { "grad_norm": 0.5047986507415771, "learning_rate": 8.955937899511315e-05, "loss": 0.0664, "step": 24900 }, { "grad_norm": 0.678598940372467, "learning_rate": 8.954926464807238e-05, "loss": 0.0569, "step": 24910 }, { "grad_norm": 0.721947968006134, "learning_rate": 8.953914597598347e-05, "loss": 0.0706, "step": 24920 }, { "grad_norm": 0.6985718607902527, "learning_rate": 8.952902297995303e-05, "loss": 0.0579, "step": 24930 }, { "grad_norm": 0.5653086304664612, "learning_rate": 8.951889566108804e-05, "loss": 0.0738, "step": 24940 }, { "grad_norm": 0.6373744010925293, "learning_rate": 8.950876402049606e-05, "loss": 0.0716, "step": 24950 }, { "grad_norm": 0.9560205340385437, "learning_rate": 8.949862805928504e-05, "loss": 0.0635, "step": 24960 }, { "grad_norm": 0.6365343332290649, "learning_rate": 8.948848777856343e-05, "loss": 0.0672, "step": 24970 }, { "grad_norm": 1.3087055683135986, "learning_rate": 8.947834317944017e-05, "loss": 0.0924, "step": 24980 }, { "grad_norm": 0.7538592219352722, "learning_rate": 8.946819426302466e-05, "loss": 0.0908, "step": 24990 }, { "grad_norm": 0.705133855342865, "learning_rate": 8.945804103042676e-05, "loss": 0.0658, "step": 25000 }, { "grad_norm": 0.9241139888763428, "learning_rate": 8.944788348275681e-05, "loss": 0.0696, "step": 25010 }, { "grad_norm": 0.6247474551200867, "learning_rate": 8.943772162112565e-05, "loss": 0.0742, "step": 25020 }, { "grad_norm": 0.6609522700309753, "learning_rate": 8.942755544664454e-05, "loss": 0.0685, "step": 25030 }, { "grad_norm": 2.229701042175293, "learning_rate": 8.941738496042525e-05, "loss": 0.0752, "step": 25040 }, { "grad_norm": 0.5908825397491455, "learning_rate": 8.940721016357999e-05, "loss": 0.0743, "step": 25050 }, { "grad_norm": 0.5638368725776672, "learning_rate": 8.939703105722148e-05, "loss": 0.0751, "step": 25060 }, { "grad_norm": 0.9063843488693237, "learning_rate": 8.93868476424629e-05, "loss": 0.0662, "step": 25070 }, { "grad_norm": 1.1257550716400146, "learning_rate": 8.937665992041786e-05, "loss": 0.078, "step": 25080 }, { "grad_norm": 1.360400915145874, "learning_rate": 8.93664678922005e-05, "loss": 0.0707, "step": 25090 }, { "grad_norm": 0.40745750069618225, "learning_rate": 8.93562715589254e-05, "loss": 0.0623, "step": 25100 }, { "grad_norm": 0.5349648594856262, "learning_rate": 8.934607092170762e-05, "loss": 0.0764, "step": 25110 }, { "grad_norm": 0.6464384198188782, "learning_rate": 8.933586598166266e-05, "loss": 0.0791, "step": 25120 }, { "grad_norm": 0.5470935106277466, "learning_rate": 8.932565673990655e-05, "loss": 0.0837, "step": 25130 }, { "grad_norm": 0.5800942182540894, "learning_rate": 8.931544319755574e-05, "loss": 0.0624, "step": 25140 }, { "grad_norm": 0.45111504197120667, "learning_rate": 8.930522535572718e-05, "loss": 0.0655, "step": 25150 }, { "grad_norm": 0.3459646999835968, "learning_rate": 8.929500321553826e-05, "loss": 0.0615, "step": 25160 }, { "grad_norm": 0.5289587378501892, "learning_rate": 8.928477677810686e-05, "loss": 0.0745, "step": 25170 }, { "grad_norm": 0.4431512951850891, "learning_rate": 8.927454604455137e-05, "loss": 0.0813, "step": 25180 }, { "grad_norm": 0.6695711612701416, "learning_rate": 8.926431101599053e-05, "loss": 0.0741, "step": 25190 }, { "grad_norm": 0.5910013318061829, "learning_rate": 8.925407169354369e-05, "loss": 0.0614, "step": 25200 }, { "grad_norm": 0.5792476534843445, "learning_rate": 8.92438280783306e-05, "loss": 0.0917, "step": 25210 }, { "grad_norm": 0.7291586399078369, "learning_rate": 8.923358017147146e-05, "loss": 0.0883, "step": 25220 }, { "grad_norm": 0.6093625426292419, "learning_rate": 8.922332797408697e-05, "loss": 0.08, "step": 25230 }, { "grad_norm": 0.46757179498672485, "learning_rate": 8.921307148729831e-05, "loss": 0.0667, "step": 25240 }, { "grad_norm": 0.785561203956604, "learning_rate": 8.920281071222712e-05, "loss": 0.0622, "step": 25250 }, { "grad_norm": 0.5031734108924866, "learning_rate": 8.919254564999548e-05, "loss": 0.0578, "step": 25260 }, { "grad_norm": 0.4092916250228882, "learning_rate": 8.918227630172598e-05, "loss": 0.0603, "step": 25270 }, { "grad_norm": 0.5215057730674744, "learning_rate": 8.917200266854165e-05, "loss": 0.0639, "step": 25280 }, { "grad_norm": 0.6165684461593628, "learning_rate": 8.9161724751566e-05, "loss": 0.0629, "step": 25290 }, { "grad_norm": 0.4417182505130768, "learning_rate": 8.915144255192302e-05, "loss": 0.0802, "step": 25300 }, { "grad_norm": 1.1770856380462646, "learning_rate": 8.914115607073714e-05, "loss": 0.0791, "step": 25310 }, { "grad_norm": 1.0268634557724, "learning_rate": 8.913086530913327e-05, "loss": 0.0663, "step": 25320 }, { "grad_norm": 0.4595615863800049, "learning_rate": 8.912057026823681e-05, "loss": 0.0731, "step": 25330 }, { "grad_norm": 0.5551663041114807, "learning_rate": 8.91102709491736e-05, "loss": 0.0785, "step": 25340 }, { "grad_norm": 0.4692169725894928, "learning_rate": 8.909996735306996e-05, "loss": 0.0724, "step": 25350 }, { "grad_norm": 0.5682493448257446, "learning_rate": 8.908965948105268e-05, "loss": 0.0667, "step": 25360 }, { "grad_norm": 0.6588695645332336, "learning_rate": 8.907934733424901e-05, "loss": 0.0705, "step": 25370 }, { "grad_norm": 0.45810428261756897, "learning_rate": 8.906903091378666e-05, "loss": 0.0689, "step": 25380 }, { "grad_norm": 0.5807210803031921, "learning_rate": 8.905871022079384e-05, "loss": 0.0749, "step": 25390 }, { "grad_norm": 0.46911466121673584, "learning_rate": 8.90483852563992e-05, "loss": 0.0715, "step": 25400 }, { "grad_norm": 0.4980296492576599, "learning_rate": 8.903805602173185e-05, "loss": 0.0665, "step": 25410 }, { "grad_norm": 0.47112640738487244, "learning_rate": 8.902772251792137e-05, "loss": 0.0584, "step": 25420 }, { "grad_norm": 0.6424348950386047, "learning_rate": 8.901738474609786e-05, "loss": 0.0708, "step": 25430 }, { "grad_norm": 0.8724157214164734, "learning_rate": 8.900704270739179e-05, "loss": 0.0783, "step": 25440 }, { "grad_norm": 0.6276055574417114, "learning_rate": 8.89966964029342e-05, "loss": 0.0749, "step": 25450 }, { "grad_norm": 0.5807493329048157, "learning_rate": 8.898634583385652e-05, "loss": 0.0474, "step": 25460 }, { "grad_norm": 0.7132615447044373, "learning_rate": 8.897599100129065e-05, "loss": 0.0797, "step": 25470 }, { "grad_norm": 0.4903959035873413, "learning_rate": 8.896563190636903e-05, "loss": 0.0594, "step": 25480 }, { "grad_norm": 0.4995436668395996, "learning_rate": 8.895526855022448e-05, "loss": 0.0682, "step": 25490 }, { "grad_norm": 0.4741215109825134, "learning_rate": 8.894490093399033e-05, "loss": 0.0733, "step": 25500 }, { "grad_norm": 0.5209914445877075, "learning_rate": 8.893452905880035e-05, "loss": 0.0784, "step": 25510 }, { "grad_norm": 0.5159634351730347, "learning_rate": 8.892415292578883e-05, "loss": 0.078, "step": 25520 }, { "grad_norm": 0.656911313533783, "learning_rate": 8.891377253609046e-05, "loss": 0.0644, "step": 25530 }, { "grad_norm": 0.48464521765708923, "learning_rate": 8.890338789084043e-05, "loss": 0.0673, "step": 25540 }, { "grad_norm": 0.8624019026756287, "learning_rate": 8.88929989911744e-05, "loss": 0.0632, "step": 25550 }, { "grad_norm": 0.6926476359367371, "learning_rate": 8.888260583822847e-05, "loss": 0.0823, "step": 25560 }, { "grad_norm": 0.4822428226470947, "learning_rate": 8.887220843313921e-05, "loss": 0.0699, "step": 25570 }, { "grad_norm": 0.5166053175926208, "learning_rate": 8.88618067770437e-05, "loss": 0.0699, "step": 25580 }, { "grad_norm": 1.2871222496032715, "learning_rate": 8.885140087107942e-05, "loss": 0.0712, "step": 25590 }, { "grad_norm": 0.5187591910362244, "learning_rate": 8.884099071638436e-05, "loss": 0.0687, "step": 25600 }, { "grad_norm": 0.8065744638442993, "learning_rate": 8.883057631409695e-05, "loss": 0.0922, "step": 25610 }, { "grad_norm": 0.569112241268158, "learning_rate": 8.882015766535608e-05, "loss": 0.0676, "step": 25620 }, { "grad_norm": 0.5819387435913086, "learning_rate": 8.880973477130115e-05, "loss": 0.0696, "step": 25630 }, { "grad_norm": 0.5983556509017944, "learning_rate": 8.879930763307197e-05, "loss": 0.0627, "step": 25640 }, { "grad_norm": 0.5606092810630798, "learning_rate": 8.878887625180884e-05, "loss": 0.0704, "step": 25650 }, { "grad_norm": 0.3830035328865051, "learning_rate": 8.877844062865253e-05, "loss": 0.0563, "step": 25660 }, { "grad_norm": 0.5527413487434387, "learning_rate": 8.876800076474424e-05, "loss": 0.0609, "step": 25670 }, { "grad_norm": 0.41931071877479553, "learning_rate": 8.875755666122568e-05, "loss": 0.0563, "step": 25680 }, { "grad_norm": 0.48878544569015503, "learning_rate": 8.8747108319239e-05, "loss": 0.0532, "step": 25690 }, { "grad_norm": 0.47549378871917725, "learning_rate": 8.87366557399268e-05, "loss": 0.0562, "step": 25700 }, { "grad_norm": 0.5544323921203613, "learning_rate": 8.872619892443217e-05, "loss": 0.0644, "step": 25710 }, { "grad_norm": 0.4315697252750397, "learning_rate": 8.871573787389865e-05, "loss": 0.0533, "step": 25720 }, { "grad_norm": 0.6931222081184387, "learning_rate": 8.870527258947024e-05, "loss": 0.071, "step": 25730 }, { "grad_norm": 0.4859151840209961, "learning_rate": 8.869480307229143e-05, "loss": 0.0477, "step": 25740 }, { "grad_norm": 0.530271589756012, "learning_rate": 8.868432932350712e-05, "loss": 0.0614, "step": 25750 }, { "grad_norm": 0.7055520415306091, "learning_rate": 8.867385134426272e-05, "loss": 0.0741, "step": 25760 }, { "grad_norm": 0.5378615856170654, "learning_rate": 8.866336913570407e-05, "loss": 0.0622, "step": 25770 }, { "grad_norm": 0.30700060725212097, "learning_rate": 8.865288269897751e-05, "loss": 0.0554, "step": 25780 }, { "grad_norm": 1.0168143510818481, "learning_rate": 8.864239203522981e-05, "loss": 0.0603, "step": 25790 }, { "grad_norm": 0.4052451550960541, "learning_rate": 8.863189714560822e-05, "loss": 0.0668, "step": 25800 }, { "grad_norm": 0.5619422197341919, "learning_rate": 8.862139803126043e-05, "loss": 0.0581, "step": 25810 }, { "grad_norm": 0.4743097424507141, "learning_rate": 8.861089469333463e-05, "loss": 0.0681, "step": 25820 }, { "grad_norm": 0.3647861182689667, "learning_rate": 8.860038713297944e-05, "loss": 0.049, "step": 25830 }, { "grad_norm": 0.40701401233673096, "learning_rate": 8.858987535134394e-05, "loss": 0.0693, "step": 25840 }, { "grad_norm": 1.2585768699645996, "learning_rate": 8.857935934957769e-05, "loss": 0.0678, "step": 25850 }, { "grad_norm": 0.5286737680435181, "learning_rate": 8.856883912883071e-05, "loss": 0.0654, "step": 25860 }, { "grad_norm": 0.6895859241485596, "learning_rate": 8.855831469025346e-05, "loss": 0.0606, "step": 25870 }, { "grad_norm": 0.6067119836807251, "learning_rate": 8.854778603499689e-05, "loss": 0.0951, "step": 25880 }, { "grad_norm": 0.9637721180915833, "learning_rate": 8.85372531642124e-05, "loss": 0.064, "step": 25890 }, { "grad_norm": 0.5549387335777283, "learning_rate": 8.852671607905185e-05, "loss": 0.0624, "step": 25900 }, { "grad_norm": 0.4439377188682556, "learning_rate": 8.851617478066754e-05, "loss": 0.071, "step": 25910 }, { "grad_norm": 0.5001856088638306, "learning_rate": 8.850562927021227e-05, "loss": 0.0676, "step": 25920 }, { "grad_norm": 0.5642596483230591, "learning_rate": 8.849507954883928e-05, "loss": 0.0644, "step": 25930 }, { "grad_norm": 0.49857595562934875, "learning_rate": 8.848452561770226e-05, "loss": 0.0675, "step": 25940 }, { "grad_norm": 0.7858208417892456, "learning_rate": 8.847396747795538e-05, "loss": 0.0603, "step": 25950 }, { "grad_norm": 0.8479868173599243, "learning_rate": 8.846340513075327e-05, "loss": 0.0665, "step": 25960 }, { "grad_norm": 0.6173948645591736, "learning_rate": 8.845283857725099e-05, "loss": 0.0724, "step": 25970 }, { "grad_norm": 0.5148882269859314, "learning_rate": 8.844226781860409e-05, "loss": 0.0638, "step": 25980 }, { "grad_norm": 0.5908896327018738, "learning_rate": 8.84316928559686e-05, "loss": 0.0802, "step": 25990 }, { "grad_norm": 1.843080997467041, "learning_rate": 8.842111369050094e-05, "loss": 0.0862, "step": 26000 }, { "grad_norm": 0.45894020795822144, "learning_rate": 8.841053032335808e-05, "loss": 0.0709, "step": 26010 }, { "grad_norm": 0.6990311145782471, "learning_rate": 8.839994275569735e-05, "loss": 0.0743, "step": 26020 }, { "grad_norm": 0.691866934299469, "learning_rate": 8.838935098867662e-05, "loss": 0.073, "step": 26030 }, { "grad_norm": 0.5517446398735046, "learning_rate": 8.837875502345418e-05, "loss": 0.0725, "step": 26040 }, { "grad_norm": 0.6865875720977783, "learning_rate": 8.83681548611888e-05, "loss": 0.0857, "step": 26050 }, { "grad_norm": 0.5774704217910767, "learning_rate": 8.835755050303969e-05, "loss": 0.0611, "step": 26060 }, { "grad_norm": 0.6758328676223755, "learning_rate": 8.834694195016653e-05, "loss": 0.0751, "step": 26070 }, { "grad_norm": 0.6685130000114441, "learning_rate": 8.833632920372942e-05, "loss": 0.0998, "step": 26080 }, { "grad_norm": 0.5901865363121033, "learning_rate": 8.832571226488903e-05, "loss": 0.0704, "step": 26090 }, { "grad_norm": 0.6145872473716736, "learning_rate": 8.831509113480634e-05, "loss": 0.0657, "step": 26100 }, { "grad_norm": 0.6183947324752808, "learning_rate": 8.83044658146429e-05, "loss": 0.0843, "step": 26110 }, { "grad_norm": 0.49690186977386475, "learning_rate": 8.829383630556067e-05, "loss": 0.0548, "step": 26120 }, { "grad_norm": 0.5448625683784485, "learning_rate": 8.828320260872207e-05, "loss": 0.0684, "step": 26130 }, { "grad_norm": 0.4374695122241974, "learning_rate": 8.827256472529e-05, "loss": 0.058, "step": 26140 }, { "grad_norm": 0.5069786906242371, "learning_rate": 8.826192265642778e-05, "loss": 0.0626, "step": 26150 }, { "grad_norm": 0.4425819516181946, "learning_rate": 8.825127640329923e-05, "loss": 0.0607, "step": 26160 }, { "grad_norm": 0.4807451665401459, "learning_rate": 8.824062596706861e-05, "loss": 0.0641, "step": 26170 }, { "grad_norm": 0.5372984409332275, "learning_rate": 8.822997134890062e-05, "loss": 0.0719, "step": 26180 }, { "grad_norm": 0.8280799984931946, "learning_rate": 8.821931254996044e-05, "loss": 0.0592, "step": 26190 }, { "grad_norm": 0.49930843710899353, "learning_rate": 8.82086495714137e-05, "loss": 0.0687, "step": 26200 }, { "grad_norm": 0.6351045370101929, "learning_rate": 8.81979824144265e-05, "loss": 0.0801, "step": 26210 }, { "grad_norm": 0.469745934009552, "learning_rate": 8.818731108016536e-05, "loss": 0.0705, "step": 26220 }, { "grad_norm": 0.5310204029083252, "learning_rate": 8.81766355697973e-05, "loss": 0.0759, "step": 26230 }, { "grad_norm": 0.6374649405479431, "learning_rate": 8.816595588448977e-05, "loss": 0.078, "step": 26240 }, { "grad_norm": 0.7955756187438965, "learning_rate": 8.81552720254107e-05, "loss": 0.074, "step": 26250 }, { "grad_norm": 0.659975528717041, "learning_rate": 8.814458399372842e-05, "loss": 0.0835, "step": 26260 }, { "grad_norm": 0.557212769985199, "learning_rate": 8.813389179061181e-05, "loss": 0.0633, "step": 26270 }, { "grad_norm": 0.52077716588974, "learning_rate": 8.812319541723012e-05, "loss": 0.0817, "step": 26280 }, { "grad_norm": 0.49867963790893555, "learning_rate": 8.811249487475309e-05, "loss": 0.0748, "step": 26290 }, { "grad_norm": 0.556988000869751, "learning_rate": 8.810179016435092e-05, "loss": 0.0748, "step": 26300 }, { "grad_norm": 0.6846165657043457, "learning_rate": 8.809108128719428e-05, "loss": 0.0796, "step": 26310 }, { "grad_norm": 0.6450297832489014, "learning_rate": 8.808036824445424e-05, "loss": 0.0701, "step": 26320 }, { "grad_norm": 0.6309899687767029, "learning_rate": 8.806965103730238e-05, "loss": 0.0706, "step": 26330 }, { "grad_norm": 1.2528125047683716, "learning_rate": 8.805892966691074e-05, "loss": 0.0641, "step": 26340 }, { "grad_norm": 0.5926917195320129, "learning_rate": 8.804820413445175e-05, "loss": 0.0626, "step": 26350 }, { "grad_norm": 0.403722882270813, "learning_rate": 8.803747444109837e-05, "loss": 0.0741, "step": 26360 }, { "grad_norm": 0.44319313764572144, "learning_rate": 8.802674058802399e-05, "loss": 0.0495, "step": 26370 }, { "grad_norm": 0.5519322752952576, "learning_rate": 8.801600257640241e-05, "loss": 0.0671, "step": 26380 }, { "grad_norm": 0.5390327572822571, "learning_rate": 8.800526040740795e-05, "loss": 0.0794, "step": 26390 }, { "grad_norm": 0.6136176586151123, "learning_rate": 8.799451408221535e-05, "loss": 0.0657, "step": 26400 }, { "grad_norm": 0.5343246459960938, "learning_rate": 8.798376360199982e-05, "loss": 0.0699, "step": 26410 }, { "grad_norm": 0.5073561072349548, "learning_rate": 8.797300896793701e-05, "loss": 0.0552, "step": 26420 }, { "grad_norm": 0.7040189504623413, "learning_rate": 8.796225018120302e-05, "loss": 0.069, "step": 26430 }, { "grad_norm": 0.5224594473838806, "learning_rate": 8.795148724297444e-05, "loss": 0.0631, "step": 26440 }, { "grad_norm": 0.6658204793930054, "learning_rate": 8.794072015442825e-05, "loss": 0.0678, "step": 26450 }, { "grad_norm": 0.8139907121658325, "learning_rate": 8.792994891674198e-05, "loss": 0.0645, "step": 26460 }, { "grad_norm": 0.963979959487915, "learning_rate": 8.79191735310935e-05, "loss": 0.0708, "step": 26470 }, { "grad_norm": 0.42538246512413025, "learning_rate": 8.790839399866122e-05, "loss": 0.0524, "step": 26480 }, { "grad_norm": 0.720707356929779, "learning_rate": 8.789761032062397e-05, "loss": 0.0584, "step": 26490 }, { "grad_norm": 0.4311786890029907, "learning_rate": 8.788682249816103e-05, "loss": 0.0588, "step": 26500 }, { "grad_norm": 0.5384423732757568, "learning_rate": 8.787603053245215e-05, "loss": 0.0739, "step": 26510 }, { "grad_norm": 0.5991233587265015, "learning_rate": 8.78652344246775e-05, "loss": 0.0615, "step": 26520 }, { "grad_norm": 0.49779045581817627, "learning_rate": 8.785443417601776e-05, "loss": 0.0571, "step": 26530 }, { "grad_norm": 0.5312279462814331, "learning_rate": 8.784362978765401e-05, "loss": 0.0566, "step": 26540 }, { "grad_norm": 0.5135929584503174, "learning_rate": 8.783282126076779e-05, "loss": 0.0617, "step": 26550 }, { "grad_norm": 0.5098806619644165, "learning_rate": 8.782200859654112e-05, "loss": 0.0622, "step": 26560 }, { "grad_norm": 1.3186508417129517, "learning_rate": 8.781119179615646e-05, "loss": 0.0644, "step": 26570 }, { "grad_norm": 0.6504731774330139, "learning_rate": 8.780037086079674e-05, "loss": 0.0645, "step": 26580 }, { "grad_norm": 0.5375447869300842, "learning_rate": 8.778954579164527e-05, "loss": 0.0673, "step": 26590 }, { "grad_norm": 0.6147545576095581, "learning_rate": 8.777871658988588e-05, "loss": 0.0742, "step": 26600 }, { "grad_norm": 0.6510624289512634, "learning_rate": 8.776788325670285e-05, "loss": 0.0589, "step": 26610 }, { "grad_norm": 0.6503321528434753, "learning_rate": 8.775704579328089e-05, "loss": 0.0674, "step": 26620 }, { "grad_norm": 0.5785439610481262, "learning_rate": 8.774620420080517e-05, "loss": 0.0604, "step": 26630 }, { "grad_norm": 0.41220828890800476, "learning_rate": 8.773535848046131e-05, "loss": 0.0494, "step": 26640 }, { "grad_norm": 0.6341250538825989, "learning_rate": 8.772450863343538e-05, "loss": 0.0816, "step": 26650 }, { "grad_norm": 0.7592215538024902, "learning_rate": 8.77136546609139e-05, "loss": 0.0803, "step": 26660 }, { "grad_norm": 1.0503746271133423, "learning_rate": 8.770279656408385e-05, "loss": 0.0613, "step": 26670 }, { "grad_norm": 0.6274832487106323, "learning_rate": 8.769193434413265e-05, "loss": 0.0657, "step": 26680 }, { "grad_norm": 0.5917664170265198, "learning_rate": 8.76810680022482e-05, "loss": 0.066, "step": 26690 }, { "grad_norm": 0.4930553138256073, "learning_rate": 8.767019753961878e-05, "loss": 0.0725, "step": 26700 }, { "grad_norm": 0.6587226986885071, "learning_rate": 8.765932295743321e-05, "loss": 0.0633, "step": 26710 }, { "grad_norm": 0.45868080854415894, "learning_rate": 8.764844425688068e-05, "loss": 0.0713, "step": 26720 }, { "grad_norm": 0.5387156009674072, "learning_rate": 8.763756143915092e-05, "loss": 0.0633, "step": 26730 }, { "grad_norm": 0.8170568346977234, "learning_rate": 8.7626674505434e-05, "loss": 0.0788, "step": 26740 }, { "grad_norm": 0.5131175518035889, "learning_rate": 8.761578345692053e-05, "loss": 0.0817, "step": 26750 }, { "grad_norm": 0.3887965977191925, "learning_rate": 8.760488829480156e-05, "loss": 0.0594, "step": 26760 }, { "grad_norm": 0.5078157782554626, "learning_rate": 8.759398902026854e-05, "loss": 0.0546, "step": 26770 }, { "grad_norm": 0.6185699105262756, "learning_rate": 8.758308563451339e-05, "loss": 0.0801, "step": 26780 }, { "grad_norm": 0.6433870792388916, "learning_rate": 8.75721781387285e-05, "loss": 0.0729, "step": 26790 }, { "grad_norm": 0.535496175289154, "learning_rate": 8.75612665341067e-05, "loss": 0.0614, "step": 26800 }, { "grad_norm": 0.6840435862541199, "learning_rate": 8.755035082184126e-05, "loss": 0.0621, "step": 26810 }, { "grad_norm": 0.4975103437900543, "learning_rate": 8.753943100312592e-05, "loss": 0.0672, "step": 26820 }, { "grad_norm": 0.49721211194992065, "learning_rate": 8.752850707915484e-05, "loss": 0.058, "step": 26830 }, { "grad_norm": 0.8300771713256836, "learning_rate": 8.751757905112264e-05, "loss": 0.0872, "step": 26840 }, { "grad_norm": 0.7638213634490967, "learning_rate": 8.75066469202244e-05, "loss": 0.0758, "step": 26850 }, { "grad_norm": 0.791348934173584, "learning_rate": 8.749571068765567e-05, "loss": 0.0561, "step": 26860 }, { "grad_norm": 0.42699623107910156, "learning_rate": 8.748477035461238e-05, "loss": 0.0657, "step": 26870 }, { "grad_norm": 0.44054898619651794, "learning_rate": 8.747382592229095e-05, "loss": 0.0678, "step": 26880 }, { "grad_norm": 0.7398374080657959, "learning_rate": 8.746287739188828e-05, "loss": 0.0792, "step": 26890 }, { "grad_norm": 0.5561028718948364, "learning_rate": 8.745192476460165e-05, "loss": 0.0582, "step": 26900 }, { "grad_norm": 0.7834759950637817, "learning_rate": 8.744096804162882e-05, "loss": 0.0753, "step": 26910 }, { "grad_norm": 0.8787157535552979, "learning_rate": 8.743000722416804e-05, "loss": 0.0715, "step": 26920 }, { "grad_norm": 0.5001184940338135, "learning_rate": 8.741904231341793e-05, "loss": 0.0577, "step": 26930 }, { "grad_norm": 0.6527646780014038, "learning_rate": 8.740807331057762e-05, "loss": 0.0751, "step": 26940 }, { "grad_norm": 0.636401355266571, "learning_rate": 8.739710021684667e-05, "loss": 0.0644, "step": 26950 }, { "grad_norm": 0.5934635400772095, "learning_rate": 8.738612303342503e-05, "loss": 0.0729, "step": 26960 }, { "grad_norm": 0.483487069606781, "learning_rate": 8.73751417615132e-05, "loss": 0.0692, "step": 26970 }, { "grad_norm": 0.6650215983390808, "learning_rate": 8.736415640231208e-05, "loss": 0.0497, "step": 26980 }, { "grad_norm": 0.46399736404418945, "learning_rate": 8.735316695702297e-05, "loss": 0.0582, "step": 26990 }, { "grad_norm": 0.3743387460708618, "learning_rate": 8.734217342684769e-05, "loss": 0.0541, "step": 27000 }, { "grad_norm": 0.5438827872276306, "learning_rate": 8.733117581298847e-05, "loss": 0.0612, "step": 27010 }, { "grad_norm": 0.6528680324554443, "learning_rate": 8.732017411664796e-05, "loss": 0.0605, "step": 27020 }, { "grad_norm": 0.6631768345832825, "learning_rate": 8.730916833902936e-05, "loss": 0.0571, "step": 27030 }, { "grad_norm": 0.32360848784446716, "learning_rate": 8.729815848133618e-05, "loss": 0.0743, "step": 27040 }, { "grad_norm": 0.47563982009887695, "learning_rate": 8.728714454477247e-05, "loss": 0.051, "step": 27050 }, { "grad_norm": 0.46314114332199097, "learning_rate": 8.727612653054269e-05, "loss": 0.0613, "step": 27060 }, { "grad_norm": 0.4932405650615692, "learning_rate": 8.726510443985176e-05, "loss": 0.0528, "step": 27070 }, { "grad_norm": 0.6349807381629944, "learning_rate": 8.725407827390503e-05, "loss": 0.0613, "step": 27080 }, { "grad_norm": 0.557083249092102, "learning_rate": 8.724304803390833e-05, "loss": 0.0633, "step": 27090 }, { "grad_norm": 0.4101402163505554, "learning_rate": 8.723201372106788e-05, "loss": 0.0549, "step": 27100 }, { "grad_norm": 0.7283057570457458, "learning_rate": 8.722097533659038e-05, "loss": 0.064, "step": 27110 }, { "grad_norm": 0.560431182384491, "learning_rate": 8.720993288168299e-05, "loss": 0.0631, "step": 27120 }, { "grad_norm": 0.6638543605804443, "learning_rate": 8.719888635755327e-05, "loss": 0.0556, "step": 27130 }, { "grad_norm": 0.4218285083770752, "learning_rate": 8.718783576540928e-05, "loss": 0.0637, "step": 27140 }, { "grad_norm": 0.9456207752227783, "learning_rate": 8.717678110645948e-05, "loss": 0.076, "step": 27150 }, { "grad_norm": 0.5810489654541016, "learning_rate": 8.716572238191279e-05, "loss": 0.0617, "step": 27160 }, { "grad_norm": 0.647767961025238, "learning_rate": 8.715465959297857e-05, "loss": 0.0722, "step": 27170 }, { "grad_norm": 0.6643640398979187, "learning_rate": 8.714359274086665e-05, "loss": 0.077, "step": 27180 }, { "grad_norm": 0.4571240544319153, "learning_rate": 8.713252182678726e-05, "loss": 0.0603, "step": 27190 }, { "grad_norm": 0.5629199743270874, "learning_rate": 8.712144685195112e-05, "loss": 0.0871, "step": 27200 }, { "grad_norm": 0.6218168139457703, "learning_rate": 8.711036781756936e-05, "loss": 0.0618, "step": 27210 }, { "grad_norm": 0.7314903140068054, "learning_rate": 8.709928472485357e-05, "loss": 0.0644, "step": 27220 }, { "grad_norm": 0.6166519522666931, "learning_rate": 8.708819757501579e-05, "loss": 0.0536, "step": 27230 }, { "grad_norm": 0.36420536041259766, "learning_rate": 8.707710636926846e-05, "loss": 0.0582, "step": 27240 }, { "grad_norm": 0.5274580121040344, "learning_rate": 8.706601110882455e-05, "loss": 0.0746, "step": 27250 }, { "grad_norm": 0.5190163254737854, "learning_rate": 8.705491179489738e-05, "loss": 0.0587, "step": 27260 }, { "grad_norm": 0.39672017097473145, "learning_rate": 8.704380842870077e-05, "loss": 0.0658, "step": 27270 }, { "grad_norm": 0.4807882606983185, "learning_rate": 8.703270101144895e-05, "loss": 0.0615, "step": 27280 }, { "grad_norm": 0.29785963892936707, "learning_rate": 8.702158954435664e-05, "loss": 0.0671, "step": 27290 }, { "grad_norm": 0.4706408679485321, "learning_rate": 8.701047402863896e-05, "loss": 0.0593, "step": 27300 }, { "grad_norm": 1.0159865617752075, "learning_rate": 8.699935446551148e-05, "loss": 0.0732, "step": 27310 }, { "grad_norm": 0.5672241449356079, "learning_rate": 8.698823085619022e-05, "loss": 0.0613, "step": 27320 }, { "grad_norm": 0.6613131165504456, "learning_rate": 8.697710320189166e-05, "loss": 0.0552, "step": 27330 }, { "grad_norm": 0.8399901986122131, "learning_rate": 8.696597150383268e-05, "loss": 0.0666, "step": 27340 }, { "grad_norm": 0.7942138314247131, "learning_rate": 8.695483576323063e-05, "loss": 0.0694, "step": 27350 }, { "grad_norm": 0.6687741279602051, "learning_rate": 8.69436959813033e-05, "loss": 0.0678, "step": 27360 }, { "grad_norm": 0.6841120719909668, "learning_rate": 8.693255215926892e-05, "loss": 0.0719, "step": 27370 }, { "grad_norm": 0.6272550225257874, "learning_rate": 8.692140429834617e-05, "loss": 0.0843, "step": 27380 }, { "grad_norm": 0.47672733664512634, "learning_rate": 8.691025239975415e-05, "loss": 0.0679, "step": 27390 }, { "grad_norm": 0.657365620136261, "learning_rate": 8.689909646471243e-05, "loss": 0.0731, "step": 27400 }, { "grad_norm": 0.4362635016441345, "learning_rate": 8.688793649444099e-05, "loss": 0.0706, "step": 27410 }, { "grad_norm": 0.4878328740596771, "learning_rate": 8.687677249016029e-05, "loss": 0.0694, "step": 27420 }, { "grad_norm": 0.7190175652503967, "learning_rate": 8.686560445309118e-05, "loss": 0.0724, "step": 27430 }, { "grad_norm": 0.7511278390884399, "learning_rate": 8.685443238445499e-05, "loss": 0.0828, "step": 27440 }, { "grad_norm": 0.48826268315315247, "learning_rate": 8.68432562854735e-05, "loss": 0.0796, "step": 27450 }, { "grad_norm": 0.5193157196044922, "learning_rate": 8.683207615736887e-05, "loss": 0.0636, "step": 27460 }, { "grad_norm": 0.7280276417732239, "learning_rate": 8.682089200136379e-05, "loss": 0.084, "step": 27470 }, { "grad_norm": 0.4639923572540283, "learning_rate": 8.680970381868132e-05, "loss": 0.0723, "step": 27480 }, { "grad_norm": 0.4798910915851593, "learning_rate": 8.679851161054498e-05, "loss": 0.0495, "step": 27490 }, { "grad_norm": 0.4101947546005249, "learning_rate": 8.678731537817873e-05, "loss": 0.0671, "step": 27500 }, { "grad_norm": 0.49330461025238037, "learning_rate": 8.677611512280697e-05, "loss": 0.0685, "step": 27510 }, { "grad_norm": 1.1732847690582275, "learning_rate": 8.676491084565457e-05, "loss": 0.0804, "step": 27520 }, { "grad_norm": 0.4065178632736206, "learning_rate": 8.675370254794678e-05, "loss": 0.079, "step": 27530 }, { "grad_norm": 0.3828677237033844, "learning_rate": 8.674249023090935e-05, "loss": 0.0607, "step": 27540 }, { "grad_norm": 1.0287507772445679, "learning_rate": 8.673127389576843e-05, "loss": 0.0704, "step": 27550 }, { "grad_norm": 0.4345034658908844, "learning_rate": 8.67200535437506e-05, "loss": 0.0567, "step": 27560 }, { "grad_norm": 0.8993280529975891, "learning_rate": 8.670882917608296e-05, "loss": 0.0683, "step": 27570 }, { "grad_norm": 0.43651649355888367, "learning_rate": 8.669760079399292e-05, "loss": 0.0733, "step": 27580 }, { "grad_norm": 0.43477076292037964, "learning_rate": 8.668636839870845e-05, "loss": 0.0606, "step": 27590 }, { "grad_norm": 1.1707079410552979, "learning_rate": 8.667513199145789e-05, "loss": 0.0765, "step": 27600 }, { "grad_norm": 0.5244669318199158, "learning_rate": 8.666389157347002e-05, "loss": 0.075, "step": 27610 }, { "grad_norm": 0.7086758017539978, "learning_rate": 8.66526471459741e-05, "loss": 0.0718, "step": 27620 }, { "grad_norm": 0.45324215292930603, "learning_rate": 8.66413987101998e-05, "loss": 0.0717, "step": 27630 }, { "grad_norm": 0.5440563559532166, "learning_rate": 8.663014626737723e-05, "loss": 0.0587, "step": 27640 }, { "grad_norm": 0.49233096837997437, "learning_rate": 8.661888981873691e-05, "loss": 0.0666, "step": 27650 }, { "grad_norm": 0.5753613114356995, "learning_rate": 8.660762936550988e-05, "loss": 0.0721, "step": 27660 }, { "grad_norm": 0.6820212602615356, "learning_rate": 8.659636490892753e-05, "loss": 0.0734, "step": 27670 }, { "grad_norm": 0.45162901282310486, "learning_rate": 8.658509645022174e-05, "loss": 0.0619, "step": 27680 }, { "grad_norm": 0.5008094310760498, "learning_rate": 8.657382399062481e-05, "loss": 0.0506, "step": 27690 }, { "grad_norm": 1.4025026559829712, "learning_rate": 8.656254753136946e-05, "loss": 0.0698, "step": 27700 }, { "grad_norm": 0.493466317653656, "learning_rate": 8.655126707368891e-05, "loss": 0.051, "step": 27710 }, { "grad_norm": 0.4905868470668793, "learning_rate": 8.653998261881672e-05, "loss": 0.0613, "step": 27720 }, { "grad_norm": 0.4754265248775482, "learning_rate": 8.652869416798699e-05, "loss": 0.0535, "step": 27730 }, { "grad_norm": 0.5427310466766357, "learning_rate": 8.651740172243417e-05, "loss": 0.0646, "step": 27740 }, { "grad_norm": 0.5194096565246582, "learning_rate": 8.65061052833932e-05, "loss": 0.0722, "step": 27750 }, { "grad_norm": 0.5252694487571716, "learning_rate": 8.649480485209945e-05, "loss": 0.074, "step": 27760 }, { "grad_norm": 0.7986653447151184, "learning_rate": 8.64835004297887e-05, "loss": 0.0882, "step": 27770 }, { "grad_norm": 0.6485913395881653, "learning_rate": 8.64721920176972e-05, "loss": 0.0708, "step": 27780 }, { "grad_norm": 0.5362470149993896, "learning_rate": 8.646087961706164e-05, "loss": 0.0712, "step": 27790 }, { "grad_norm": 0.5357070565223694, "learning_rate": 8.644956322911908e-05, "loss": 0.0724, "step": 27800 }, { "grad_norm": 0.4455285370349884, "learning_rate": 8.643824285510709e-05, "loss": 0.0743, "step": 27810 }, { "grad_norm": 0.5362290143966675, "learning_rate": 8.642691849626364e-05, "loss": 0.0525, "step": 27820 }, { "grad_norm": 0.4786578416824341, "learning_rate": 8.641559015382717e-05, "loss": 0.0613, "step": 27830 }, { "grad_norm": 0.7653278112411499, "learning_rate": 8.640425782903649e-05, "loss": 0.0619, "step": 27840 }, { "grad_norm": 0.48523107171058655, "learning_rate": 8.639292152313091e-05, "loss": 0.0617, "step": 27850 }, { "grad_norm": 0.5030706524848938, "learning_rate": 8.638158123735015e-05, "loss": 0.0624, "step": 27860 }, { "grad_norm": 0.40602079033851624, "learning_rate": 8.637023697293436e-05, "loss": 0.0517, "step": 27870 }, { "grad_norm": 0.5453976392745972, "learning_rate": 8.635888873112414e-05, "loss": 0.0764, "step": 27880 }, { "grad_norm": 0.48006710410118103, "learning_rate": 8.634753651316052e-05, "loss": 0.0653, "step": 27890 }, { "grad_norm": 0.4740197956562042, "learning_rate": 8.633618032028496e-05, "loss": 0.0551, "step": 27900 }, { "grad_norm": 0.6705836653709412, "learning_rate": 8.632482015373934e-05, "loss": 0.0642, "step": 27910 }, { "grad_norm": 0.6896576881408691, "learning_rate": 8.6313456014766e-05, "loss": 0.0775, "step": 27920 }, { "grad_norm": 0.7189294099807739, "learning_rate": 8.630208790460771e-05, "loss": 0.0675, "step": 27930 }, { "grad_norm": 1.0125203132629395, "learning_rate": 8.629071582450768e-05, "loss": 0.0748, "step": 27940 }, { "grad_norm": 0.6874539852142334, "learning_rate": 8.62793397757095e-05, "loss": 0.0792, "step": 27950 }, { "grad_norm": 0.7045515775680542, "learning_rate": 8.626795975945729e-05, "loss": 0.0738, "step": 27960 }, { "grad_norm": 0.8654153347015381, "learning_rate": 8.625657577699551e-05, "loss": 0.0483, "step": 27970 }, { "grad_norm": 0.44830113649368286, "learning_rate": 8.624518782956914e-05, "loss": 0.0627, "step": 27980 }, { "grad_norm": 0.586879312992096, "learning_rate": 8.62337959184235e-05, "loss": 0.0663, "step": 27990 }, { "grad_norm": 0.6519105434417725, "learning_rate": 8.622240004480441e-05, "loss": 0.0753, "step": 28000 }, { "grad_norm": 0.6517702341079712, "learning_rate": 8.621100020995814e-05, "loss": 0.0644, "step": 28010 }, { "grad_norm": 0.6806138157844543, "learning_rate": 8.619959641513132e-05, "loss": 0.08, "step": 28020 }, { "grad_norm": 0.6286031007766724, "learning_rate": 8.618818866157105e-05, "loss": 0.0719, "step": 28030 }, { "grad_norm": 0.7754769325256348, "learning_rate": 8.617677695052487e-05, "loss": 0.0684, "step": 28040 }, { "grad_norm": 0.40269559621810913, "learning_rate": 8.616536128324078e-05, "loss": 0.0627, "step": 28050 }, { "grad_norm": 0.5866463780403137, "learning_rate": 8.615394166096712e-05, "loss": 0.0721, "step": 28060 }, { "grad_norm": 0.5746970772743225, "learning_rate": 8.614251808495279e-05, "loss": 0.0593, "step": 28070 }, { "grad_norm": 0.6770668029785156, "learning_rate": 8.6131090556447e-05, "loss": 0.0625, "step": 28080 }, { "grad_norm": 0.6602028608322144, "learning_rate": 8.611965907669947e-05, "loss": 0.0613, "step": 28090 }, { "grad_norm": 0.9233524203300476, "learning_rate": 8.610822364696034e-05, "loss": 0.0708, "step": 28100 }, { "grad_norm": 0.4075932204723358, "learning_rate": 8.609678426848015e-05, "loss": 0.0713, "step": 28110 }, { "grad_norm": 0.6831175088882446, "learning_rate": 8.60853409425099e-05, "loss": 0.0649, "step": 28120 }, { "grad_norm": 0.5197165608406067, "learning_rate": 8.607389367030104e-05, "loss": 0.0667, "step": 28130 }, { "grad_norm": 0.7932625412940979, "learning_rate": 8.606244245310538e-05, "loss": 0.0748, "step": 28140 }, { "grad_norm": 0.7928217649459839, "learning_rate": 8.605098729217525e-05, "loss": 0.0711, "step": 28150 }, { "grad_norm": 0.5535207986831665, "learning_rate": 8.603952818876335e-05, "loss": 0.0741, "step": 28160 }, { "grad_norm": 0.5374224185943604, "learning_rate": 8.602806514412281e-05, "loss": 0.0563, "step": 28170 }, { "grad_norm": 0.5363894701004028, "learning_rate": 8.601659815950726e-05, "loss": 0.0758, "step": 28180 }, { "grad_norm": 0.5528309941291809, "learning_rate": 8.600512723617067e-05, "loss": 0.082, "step": 28190 }, { "grad_norm": 0.5849565863609314, "learning_rate": 8.59936523753675e-05, "loss": 0.0727, "step": 28200 }, { "grad_norm": 0.3982303738594055, "learning_rate": 8.598217357835264e-05, "loss": 0.0665, "step": 28210 }, { "grad_norm": 0.534673810005188, "learning_rate": 8.597069084638135e-05, "loss": 0.0623, "step": 28220 }, { "grad_norm": 0.5331676006317139, "learning_rate": 8.595920418070939e-05, "loss": 0.0592, "step": 28230 }, { "grad_norm": 0.6437534689903259, "learning_rate": 8.594771358259295e-05, "loss": 0.0625, "step": 28240 }, { "grad_norm": 1.0439457893371582, "learning_rate": 8.593621905328858e-05, "loss": 0.076, "step": 28250 }, { "grad_norm": 0.5267831683158875, "learning_rate": 8.592472059405333e-05, "loss": 0.0666, "step": 28260 }, { "grad_norm": 0.5324504375457764, "learning_rate": 8.591321820614464e-05, "loss": 0.0652, "step": 28270 }, { "grad_norm": 0.6260735392570496, "learning_rate": 8.590171189082041e-05, "loss": 0.0774, "step": 28280 }, { "grad_norm": 0.46440839767456055, "learning_rate": 8.589020164933894e-05, "loss": 0.0665, "step": 28290 }, { "grad_norm": 0.48588067293167114, "learning_rate": 8.587868748295898e-05, "loss": 0.0604, "step": 28300 }, { "grad_norm": 0.5412793755531311, "learning_rate": 8.586716939293971e-05, "loss": 0.0571, "step": 28310 }, { "grad_norm": 0.7080075740814209, "learning_rate": 8.58556473805407e-05, "loss": 0.0586, "step": 28320 }, { "grad_norm": 0.3920170068740845, "learning_rate": 8.584412144702202e-05, "loss": 0.0673, "step": 28330 }, { "grad_norm": 0.48528048396110535, "learning_rate": 8.58325915936441e-05, "loss": 0.0723, "step": 28340 }, { "grad_norm": 0.5041676163673401, "learning_rate": 8.582105782166783e-05, "loss": 0.0527, "step": 28350 }, { "grad_norm": 0.33422842621803284, "learning_rate": 8.580952013235455e-05, "loss": 0.0572, "step": 28360 }, { "grad_norm": 1.04315984249115, "learning_rate": 8.579797852696596e-05, "loss": 0.0691, "step": 28370 }, { "grad_norm": 0.452727735042572, "learning_rate": 8.578643300676428e-05, "loss": 0.0634, "step": 28380 }, { "grad_norm": 0.5023826360702515, "learning_rate": 8.577488357301209e-05, "loss": 0.0681, "step": 28390 }, { "grad_norm": 0.5480130910873413, "learning_rate": 8.576333022697242e-05, "loss": 0.0624, "step": 28400 }, { "grad_norm": 0.7231478095054626, "learning_rate": 8.575177296990873e-05, "loss": 0.0675, "step": 28410 }, { "grad_norm": 0.6668936014175415, "learning_rate": 8.574021180308489e-05, "loss": 0.0763, "step": 28420 }, { "grad_norm": 0.7238178849220276, "learning_rate": 8.572864672776523e-05, "loss": 0.0758, "step": 28430 }, { "grad_norm": 0.3645535707473755, "learning_rate": 8.571707774521447e-05, "loss": 0.0572, "step": 28440 }, { "grad_norm": 0.5578731894493103, "learning_rate": 8.57055048566978e-05, "loss": 0.0579, "step": 28450 }, { "grad_norm": 0.9455917477607727, "learning_rate": 8.569392806348078e-05, "loss": 0.0721, "step": 28460 }, { "grad_norm": 0.4756646752357483, "learning_rate": 8.568234736682947e-05, "loss": 0.059, "step": 28470 }, { "grad_norm": 0.5800183415412903, "learning_rate": 8.567076276801029e-05, "loss": 0.092, "step": 28480 }, { "grad_norm": 0.5509396195411682, "learning_rate": 8.565917426829013e-05, "loss": 0.0478, "step": 28490 }, { "grad_norm": 0.8884373903274536, "learning_rate": 8.564758186893628e-05, "loss": 0.0593, "step": 28500 }, { "grad_norm": 0.45690351724624634, "learning_rate": 8.563598557121649e-05, "loss": 0.065, "step": 28510 }, { "grad_norm": 0.5021103620529175, "learning_rate": 8.562438537639888e-05, "loss": 0.0603, "step": 28520 }, { "grad_norm": 0.5744030475616455, "learning_rate": 8.561278128575206e-05, "loss": 0.0691, "step": 28530 }, { "grad_norm": 0.7190417647361755, "learning_rate": 8.5601173300545e-05, "loss": 0.0641, "step": 28540 }, { "grad_norm": 0.9320283532142639, "learning_rate": 8.558956142204717e-05, "loss": 0.079, "step": 28550 }, { "grad_norm": 0.7139034867286682, "learning_rate": 8.55779456515284e-05, "loss": 0.0781, "step": 28560 }, { "grad_norm": 0.46989408135414124, "learning_rate": 8.556632599025898e-05, "loss": 0.0715, "step": 28570 }, { "grad_norm": 0.4319817125797272, "learning_rate": 8.555470243950964e-05, "loss": 0.0567, "step": 28580 }, { "grad_norm": 0.3969988524913788, "learning_rate": 8.554307500055148e-05, "loss": 0.0597, "step": 28590 }, { "grad_norm": 0.6269411444664001, "learning_rate": 8.553144367465609e-05, "loss": 0.0527, "step": 28600 }, { "grad_norm": 0.4808204472064972, "learning_rate": 8.551980846309544e-05, "loss": 0.0682, "step": 28610 }, { "grad_norm": 0.6414273977279663, "learning_rate": 8.550816936714193e-05, "loss": 0.0688, "step": 28620 }, { "grad_norm": 0.5764904022216797, "learning_rate": 8.549652638806841e-05, "loss": 0.0631, "step": 28630 }, { "grad_norm": 0.6862058639526367, "learning_rate": 8.548487952714812e-05, "loss": 0.0832, "step": 28640 }, { "grad_norm": 0.5420055389404297, "learning_rate": 8.547322878565478e-05, "loss": 0.0576, "step": 28650 }, { "grad_norm": 0.5756197571754456, "learning_rate": 8.546157416486245e-05, "loss": 0.0661, "step": 28660 }, { "grad_norm": 0.7312341332435608, "learning_rate": 8.54499156660457e-05, "loss": 0.0637, "step": 28670 }, { "grad_norm": 0.6694369316101074, "learning_rate": 8.543825329047947e-05, "loss": 0.0474, "step": 28680 }, { "grad_norm": 0.6373054385185242, "learning_rate": 8.542658703943913e-05, "loss": 0.0651, "step": 28690 }, { "grad_norm": 0.5663304328918457, "learning_rate": 8.541491691420051e-05, "loss": 0.0706, "step": 28700 }, { "grad_norm": 0.6758229732513428, "learning_rate": 8.54032429160398e-05, "loss": 0.0582, "step": 28710 }, { "grad_norm": 0.7235090136528015, "learning_rate": 8.539156504623369e-05, "loss": 0.0771, "step": 28720 }, { "grad_norm": 0.5180531144142151, "learning_rate": 8.537988330605923e-05, "loss": 0.0845, "step": 28730 }, { "grad_norm": 0.5883924961090088, "learning_rate": 8.536819769679393e-05, "loss": 0.0721, "step": 28740 }, { "grad_norm": 0.4051736295223236, "learning_rate": 8.53565082197157e-05, "loss": 0.0566, "step": 28750 }, { "grad_norm": 0.6515299081802368, "learning_rate": 8.534481487610289e-05, "loss": 0.0796, "step": 28760 }, { "grad_norm": 0.4331109821796417, "learning_rate": 8.533311766723428e-05, "loss": 0.0506, "step": 28770 }, { "grad_norm": 0.4236745834350586, "learning_rate": 8.532141659438901e-05, "loss": 0.0528, "step": 28780 }, { "grad_norm": 1.01369047164917, "learning_rate": 8.530971165884675e-05, "loss": 0.0596, "step": 28790 }, { "grad_norm": 1.7413902282714844, "learning_rate": 8.529800286188752e-05, "loss": 0.0733, "step": 28800 }, { "grad_norm": 0.4449508488178253, "learning_rate": 8.528629020479175e-05, "loss": 0.0559, "step": 28810 }, { "grad_norm": 0.7413952946662903, "learning_rate": 8.527457368884033e-05, "loss": 0.0565, "step": 28820 }, { "grad_norm": 0.5253705978393555, "learning_rate": 8.526285331531458e-05, "loss": 0.0435, "step": 28830 }, { "grad_norm": 0.5866521000862122, "learning_rate": 8.525112908549621e-05, "loss": 0.0542, "step": 28840 }, { "grad_norm": 0.6814342141151428, "learning_rate": 8.523940100066735e-05, "loss": 0.0743, "step": 28850 }, { "grad_norm": 0.3148227334022522, "learning_rate": 8.52276690621106e-05, "loss": 0.0517, "step": 28860 }, { "grad_norm": 0.8775815367698669, "learning_rate": 8.521593327110889e-05, "loss": 0.0661, "step": 28870 }, { "grad_norm": 0.475453644990921, "learning_rate": 8.520419362894569e-05, "loss": 0.0759, "step": 28880 }, { "grad_norm": 0.6136499047279358, "learning_rate": 8.51924501369048e-05, "loss": 0.0562, "step": 28890 }, { "grad_norm": 1.0373705625534058, "learning_rate": 8.518070279627047e-05, "loss": 0.0753, "step": 28900 }, { "grad_norm": 0.7555318474769592, "learning_rate": 8.516895160832737e-05, "loss": 0.0722, "step": 28910 }, { "grad_norm": 0.7090882658958435, "learning_rate": 8.515719657436061e-05, "loss": 0.059, "step": 28920 }, { "grad_norm": 0.6594870090484619, "learning_rate": 8.514543769565568e-05, "loss": 0.0691, "step": 28930 }, { "grad_norm": 0.7793505191802979, "learning_rate": 8.513367497349853e-05, "loss": 0.0782, "step": 28940 }, { "grad_norm": 0.7815800309181213, "learning_rate": 8.51219084091755e-05, "loss": 0.0624, "step": 28950 }, { "grad_norm": 0.5277711153030396, "learning_rate": 8.511013800397338e-05, "loss": 0.0614, "step": 28960 }, { "grad_norm": 0.45200949907302856, "learning_rate": 8.509836375917937e-05, "loss": 0.0714, "step": 28970 }, { "grad_norm": 0.6645366549491882, "learning_rate": 8.508658567608104e-05, "loss": 0.0625, "step": 28980 }, { "grad_norm": 0.4459022283554077, "learning_rate": 8.507480375596647e-05, "loss": 0.0511, "step": 28990 }, { "grad_norm": 0.5081318020820618, "learning_rate": 8.506301800012408e-05, "loss": 0.0536, "step": 29000 }, { "grad_norm": 0.9308363199234009, "learning_rate": 8.505122840984278e-05, "loss": 0.0703, "step": 29010 }, { "grad_norm": 0.45189595222473145, "learning_rate": 8.503943498641182e-05, "loss": 0.0718, "step": 29020 }, { "grad_norm": 0.574515700340271, "learning_rate": 8.502763773112095e-05, "loss": 0.0665, "step": 29030 }, { "grad_norm": 0.633955180644989, "learning_rate": 8.501583664526026e-05, "loss": 0.0598, "step": 29040 }, { "grad_norm": 0.5247575640678406, "learning_rate": 8.500403173012032e-05, "loss": 0.063, "step": 29050 }, { "grad_norm": 0.49658578634262085, "learning_rate": 8.499222298699211e-05, "loss": 0.0683, "step": 29060 }, { "grad_norm": 0.85103839635849, "learning_rate": 8.498041041716701e-05, "loss": 0.0607, "step": 29070 }, { "grad_norm": 0.7494368553161621, "learning_rate": 8.496859402193681e-05, "loss": 0.0636, "step": 29080 }, { "grad_norm": 1.0735480785369873, "learning_rate": 8.495677380259374e-05, "loss": 0.0651, "step": 29090 }, { "grad_norm": 0.4414575695991516, "learning_rate": 8.494494976043045e-05, "loss": 0.0733, "step": 29100 }, { "grad_norm": 0.5937833189964294, "learning_rate": 8.493312189673998e-05, "loss": 0.0628, "step": 29110 }, { "grad_norm": 1.0170694589614868, "learning_rate": 8.492129021281584e-05, "loss": 0.0702, "step": 29120 }, { "grad_norm": 0.5999826788902283, "learning_rate": 8.490945470995188e-05, "loss": 0.0573, "step": 29130 }, { "grad_norm": 0.5015166401863098, "learning_rate": 8.489761538944247e-05, "loss": 0.0618, "step": 29140 }, { "grad_norm": 0.4588626027107239, "learning_rate": 8.48857722525823e-05, "loss": 0.0554, "step": 29150 }, { "grad_norm": 0.3803117871284485, "learning_rate": 8.487392530066652e-05, "loss": 0.0588, "step": 29160 }, { "grad_norm": 0.5183342099189758, "learning_rate": 8.486207453499069e-05, "loss": 0.072, "step": 29170 }, { "grad_norm": 0.5865939259529114, "learning_rate": 8.485021995685082e-05, "loss": 0.0719, "step": 29180 }, { "grad_norm": 0.6077620387077332, "learning_rate": 8.483836156754328e-05, "loss": 0.0562, "step": 29190 }, { "grad_norm": 0.42426058650016785, "learning_rate": 8.482649936836491e-05, "loss": 0.0729, "step": 29200 }, { "grad_norm": 0.5181040167808533, "learning_rate": 8.481463336061293e-05, "loss": 0.0547, "step": 29210 }, { "grad_norm": 0.5022435188293457, "learning_rate": 8.480276354558496e-05, "loss": 0.0733, "step": 29220 }, { "grad_norm": 0.730922520160675, "learning_rate": 8.479088992457913e-05, "loss": 0.0637, "step": 29230 }, { "grad_norm": 0.5434563159942627, "learning_rate": 8.477901249889387e-05, "loss": 0.0672, "step": 29240 }, { "grad_norm": 0.48245877027511597, "learning_rate": 8.47671312698281e-05, "loss": 0.0573, "step": 29250 }, { "grad_norm": 0.809295117855072, "learning_rate": 8.475524623868112e-05, "loss": 0.0616, "step": 29260 }, { "grad_norm": 0.4555227756500244, "learning_rate": 8.474335740675266e-05, "loss": 0.0653, "step": 29270 }, { "grad_norm": 0.569534420967102, "learning_rate": 8.473146477534289e-05, "loss": 0.0659, "step": 29280 }, { "grad_norm": 0.543757975101471, "learning_rate": 8.471956834575232e-05, "loss": 0.0654, "step": 29290 }, { "grad_norm": 0.588533878326416, "learning_rate": 8.470766811928197e-05, "loss": 0.0579, "step": 29300 }, { "grad_norm": 0.464899480342865, "learning_rate": 8.469576409723323e-05, "loss": 0.0531, "step": 29310 }, { "grad_norm": 0.6904110312461853, "learning_rate": 8.468385628090788e-05, "loss": 0.0576, "step": 29320 }, { "grad_norm": 0.46435317397117615, "learning_rate": 8.467194467160815e-05, "loss": 0.0651, "step": 29330 }, { "grad_norm": 0.850078821182251, "learning_rate": 8.466002927063667e-05, "loss": 0.0654, "step": 29340 }, { "grad_norm": 0.9973175525665283, "learning_rate": 8.464811007929651e-05, "loss": 0.0678, "step": 29350 }, { "grad_norm": 0.5075356364250183, "learning_rate": 8.463618709889114e-05, "loss": 0.0702, "step": 29360 }, { "grad_norm": 0.5140700340270996, "learning_rate": 8.462426033072442e-05, "loss": 0.066, "step": 29370 }, { "grad_norm": 0.7430644035339355, "learning_rate": 8.461232977610061e-05, "loss": 0.0636, "step": 29380 }, { "grad_norm": 1.17228364944458, "learning_rate": 8.46003954363245e-05, "loss": 0.0541, "step": 29390 }, { "grad_norm": 0.44810429215431213, "learning_rate": 8.458845731270115e-05, "loss": 0.066, "step": 29400 }, { "grad_norm": 0.494751900434494, "learning_rate": 8.45765154065361e-05, "loss": 0.0586, "step": 29410 }, { "grad_norm": 0.7607948184013367, "learning_rate": 8.456456971913532e-05, "loss": 0.05, "step": 29420 }, { "grad_norm": 0.6225576400756836, "learning_rate": 8.455262025180517e-05, "loss": 0.0578, "step": 29430 }, { "grad_norm": 0.7214946746826172, "learning_rate": 8.454066700585242e-05, "loss": 0.0634, "step": 29440 }, { "grad_norm": 0.4812619090080261, "learning_rate": 8.452870998258423e-05, "loss": 0.0774, "step": 29450 }, { "grad_norm": 0.42940011620521545, "learning_rate": 8.451674918330825e-05, "loss": 0.0599, "step": 29460 }, { "grad_norm": 0.42760905623435974, "learning_rate": 8.450478460933246e-05, "loss": 0.065, "step": 29470 }, { "grad_norm": 0.7255758047103882, "learning_rate": 8.449281626196532e-05, "loss": 0.0781, "step": 29480 }, { "grad_norm": 0.5250855684280396, "learning_rate": 8.448084414251564e-05, "loss": 0.0784, "step": 29490 }, { "grad_norm": 0.589573085308075, "learning_rate": 8.446886825229271e-05, "loss": 0.053, "step": 29500 }, { "grad_norm": 0.552107036113739, "learning_rate": 8.445688859260615e-05, "loss": 0.0624, "step": 29510 }, { "grad_norm": 0.5138055682182312, "learning_rate": 8.444490516476606e-05, "loss": 0.0536, "step": 29520 }, { "grad_norm": 0.6854743361473083, "learning_rate": 8.443291797008293e-05, "loss": 0.0662, "step": 29530 }, { "grad_norm": 0.33573606610298157, "learning_rate": 8.442092700986765e-05, "loss": 0.0688, "step": 29540 }, { "grad_norm": 0.45950937271118164, "learning_rate": 8.440893228543156e-05, "loss": 0.0755, "step": 29550 }, { "grad_norm": 0.9823014736175537, "learning_rate": 8.439693379808638e-05, "loss": 0.0594, "step": 29560 }, { "grad_norm": 0.4708472490310669, "learning_rate": 8.43849315491442e-05, "loss": 0.0648, "step": 29570 }, { "grad_norm": 0.5113913416862488, "learning_rate": 8.437292553991763e-05, "loss": 0.0683, "step": 29580 }, { "grad_norm": 0.9659349322319031, "learning_rate": 8.436091577171959e-05, "loss": 0.0761, "step": 29590 }, { "grad_norm": 0.8044838309288025, "learning_rate": 8.434890224586347e-05, "loss": 0.0649, "step": 29600 }, { "grad_norm": 1.2257927656173706, "learning_rate": 8.433688496366303e-05, "loss": 0.0645, "step": 29610 }, { "grad_norm": 0.4088565409183502, "learning_rate": 8.432486392643248e-05, "loss": 0.0613, "step": 29620 }, { "grad_norm": 0.6549939513206482, "learning_rate": 8.431283913548643e-05, "loss": 0.0644, "step": 29630 }, { "grad_norm": 0.6698103547096252, "learning_rate": 8.430081059213985e-05, "loss": 0.0674, "step": 29640 }, { "grad_norm": 0.6062690019607544, "learning_rate": 8.428877829770823e-05, "loss": 0.0691, "step": 29650 }, { "grad_norm": 1.2442094087600708, "learning_rate": 8.427674225350735e-05, "loss": 0.075, "step": 29660 }, { "grad_norm": 1.2238675355911255, "learning_rate": 8.426470246085347e-05, "loss": 0.0774, "step": 29670 }, { "grad_norm": 0.7180204391479492, "learning_rate": 8.425265892106324e-05, "loss": 0.0732, "step": 29680 }, { "grad_norm": 0.59535151720047, "learning_rate": 8.424061163545374e-05, "loss": 0.0695, "step": 29690 }, { "grad_norm": 0.35037943720817566, "learning_rate": 8.422856060534243e-05, "loss": 0.0668, "step": 29700 }, { "grad_norm": 0.6711537837982178, "learning_rate": 8.421650583204718e-05, "loss": 0.0661, "step": 29710 }, { "grad_norm": 0.3895072937011719, "learning_rate": 8.420444731688633e-05, "loss": 0.0744, "step": 29720 }, { "grad_norm": 0.3854205012321472, "learning_rate": 8.419238506117852e-05, "loss": 0.0861, "step": 29730 }, { "grad_norm": 0.694305419921875, "learning_rate": 8.418031906624289e-05, "loss": 0.0647, "step": 29740 }, { "grad_norm": 0.515608012676239, "learning_rate": 8.416824933339898e-05, "loss": 0.0735, "step": 29750 }, { "grad_norm": 0.5750471353530884, "learning_rate": 8.415617586396667e-05, "loss": 0.0575, "step": 29760 }, { "grad_norm": 1.3703525066375732, "learning_rate": 8.414409865926632e-05, "loss": 0.0706, "step": 29770 }, { "grad_norm": 0.5528045296669006, "learning_rate": 8.413201772061867e-05, "loss": 0.0589, "step": 29780 }, { "grad_norm": 0.3632482886314392, "learning_rate": 8.411993304934488e-05, "loss": 0.0514, "step": 29790 }, { "grad_norm": 0.4223264753818512, "learning_rate": 8.410784464676654e-05, "loss": 0.0596, "step": 29800 }, { "grad_norm": 0.5403691530227661, "learning_rate": 8.409575251420556e-05, "loss": 0.067, "step": 29810 }, { "grad_norm": 0.427034467458725, "learning_rate": 8.408365665298435e-05, "loss": 0.0565, "step": 29820 }, { "grad_norm": 0.6416404843330383, "learning_rate": 8.40715570644257e-05, "loss": 0.0785, "step": 29830 }, { "grad_norm": 0.616169273853302, "learning_rate": 8.40594537498528e-05, "loss": 0.0784, "step": 29840 }, { "grad_norm": 0.455138236284256, "learning_rate": 8.404734671058924e-05, "loss": 0.0702, "step": 29850 }, { "grad_norm": 0.5233464241027832, "learning_rate": 8.403523594795902e-05, "loss": 0.0653, "step": 29860 }, { "grad_norm": 0.6778855323791504, "learning_rate": 8.402312146328659e-05, "loss": 0.0604, "step": 29870 }, { "grad_norm": 0.5262965559959412, "learning_rate": 8.401100325789675e-05, "loss": 0.0644, "step": 29880 }, { "grad_norm": 0.5679575800895691, "learning_rate": 8.399888133311472e-05, "loss": 0.0552, "step": 29890 }, { "grad_norm": 0.4512689411640167, "learning_rate": 8.398675569026613e-05, "loss": 0.0633, "step": 29900 }, { "grad_norm": 0.4572352468967438, "learning_rate": 8.397462633067705e-05, "loss": 0.0598, "step": 29910 }, { "grad_norm": 0.4819413125514984, "learning_rate": 8.396249325567392e-05, "loss": 0.0686, "step": 29920 }, { "grad_norm": 0.6378141045570374, "learning_rate": 8.395035646658357e-05, "loss": 0.0828, "step": 29930 }, { "grad_norm": 0.8358384370803833, "learning_rate": 8.39382159647333e-05, "loss": 0.0764, "step": 29940 }, { "grad_norm": 0.4984806180000305, "learning_rate": 8.392607175145075e-05, "loss": 0.0555, "step": 29950 }, { "grad_norm": 0.6421226263046265, "learning_rate": 8.3913923828064e-05, "loss": 0.0736, "step": 29960 }, { "grad_norm": 0.41200628876686096, "learning_rate": 8.390177219590152e-05, "loss": 0.0529, "step": 29970 }, { "grad_norm": 0.6891482472419739, "learning_rate": 8.388961685629222e-05, "loss": 0.0767, "step": 29980 }, { "grad_norm": 0.8273330330848694, "learning_rate": 8.387745781056536e-05, "loss": 0.06, "step": 29990 }, { "grad_norm": 0.41806796193122864, "learning_rate": 8.386529506005065e-05, "loss": 0.0481, "step": 30000 }, { "grad_norm": 0.4418148398399353, "learning_rate": 8.38531286060782e-05, "loss": 0.058, "step": 30010 }, { "grad_norm": 0.5759233236312866, "learning_rate": 8.384095844997849e-05, "loss": 0.0485, "step": 30020 }, { "grad_norm": 0.5887139439582825, "learning_rate": 8.382878459308245e-05, "loss": 0.0527, "step": 30030 }, { "grad_norm": 0.6080780029296875, "learning_rate": 8.381660703672138e-05, "loss": 0.0564, "step": 30040 }, { "grad_norm": 0.5515627264976501, "learning_rate": 8.380442578222702e-05, "loss": 0.0642, "step": 30050 }, { "grad_norm": 0.5107333660125732, "learning_rate": 8.379224083093146e-05, "loss": 0.0577, "step": 30060 }, { "grad_norm": 0.42001092433929443, "learning_rate": 8.378005218416727e-05, "loss": 0.0612, "step": 30070 }, { "grad_norm": 0.5178819298744202, "learning_rate": 8.376785984326735e-05, "loss": 0.0621, "step": 30080 }, { "grad_norm": 0.4915545582771301, "learning_rate": 8.375566380956506e-05, "loss": 0.0568, "step": 30090 }, { "grad_norm": 0.8408312797546387, "learning_rate": 8.374346408439411e-05, "loss": 0.0737, "step": 30100 }, { "grad_norm": 0.5237038731575012, "learning_rate": 8.373126066908868e-05, "loss": 0.0618, "step": 30110 }, { "grad_norm": 0.5651694536209106, "learning_rate": 8.371905356498326e-05, "loss": 0.0594, "step": 30120 }, { "grad_norm": 0.3758513331413269, "learning_rate": 8.370684277341288e-05, "loss": 0.051, "step": 30130 }, { "grad_norm": 0.5201574563980103, "learning_rate": 8.369462829571282e-05, "loss": 0.0568, "step": 30140 }, { "grad_norm": 0.7787262797355652, "learning_rate": 8.36824101332189e-05, "loss": 0.0503, "step": 30150 }, { "grad_norm": 0.35685351490974426, "learning_rate": 8.367018828726721e-05, "loss": 0.0601, "step": 30160 }, { "grad_norm": 0.5992498993873596, "learning_rate": 8.365796275919438e-05, "loss": 0.0731, "step": 30170 }, { "grad_norm": 1.1499664783477783, "learning_rate": 8.364573355033734e-05, "loss": 0.0496, "step": 30180 }, { "grad_norm": 0.4419664442539215, "learning_rate": 8.363350066203346e-05, "loss": 0.0483, "step": 30190 }, { "grad_norm": 0.829486608505249, "learning_rate": 8.362126409562053e-05, "loss": 0.0651, "step": 30200 }, { "grad_norm": 0.4842569828033447, "learning_rate": 8.360902385243667e-05, "loss": 0.0507, "step": 30210 }, { "grad_norm": 0.41224411129951477, "learning_rate": 8.359677993382052e-05, "loss": 0.0467, "step": 30220 }, { "grad_norm": 0.41672107577323914, "learning_rate": 8.358453234111103e-05, "loss": 0.0578, "step": 30230 }, { "grad_norm": 0.5629432201385498, "learning_rate": 8.357228107564756e-05, "loss": 0.0601, "step": 30240 }, { "grad_norm": 0.5333870649337769, "learning_rate": 8.356002613876993e-05, "loss": 0.057, "step": 30250 }, { "grad_norm": 0.3391995131969452, "learning_rate": 8.35477675318183e-05, "loss": 0.0475, "step": 30260 }, { "grad_norm": 0.7257750630378723, "learning_rate": 8.353550525613323e-05, "loss": 0.0639, "step": 30270 }, { "grad_norm": 0.46474045515060425, "learning_rate": 8.352323931305572e-05, "loss": 0.0642, "step": 30280 }, { "grad_norm": 0.5488659143447876, "learning_rate": 8.351096970392717e-05, "loss": 0.0604, "step": 30290 }, { "grad_norm": 0.5263434648513794, "learning_rate": 8.349869643008937e-05, "loss": 0.065, "step": 30300 }, { "grad_norm": 0.8157851696014404, "learning_rate": 8.348641949288449e-05, "loss": 0.0721, "step": 30310 }, { "grad_norm": 0.4401721656322479, "learning_rate": 8.34741388936551e-05, "loss": 0.0661, "step": 30320 }, { "grad_norm": 0.6069239974021912, "learning_rate": 8.346185463374423e-05, "loss": 0.0774, "step": 30330 }, { "grad_norm": 0.34848275780677795, "learning_rate": 8.344956671449524e-05, "loss": 0.0645, "step": 30340 }, { "grad_norm": 0.5752788782119751, "learning_rate": 8.343727513725192e-05, "loss": 0.0623, "step": 30350 }, { "grad_norm": 0.5380358099937439, "learning_rate": 8.342497990335847e-05, "loss": 0.0593, "step": 30360 }, { "grad_norm": 1.4469166994094849, "learning_rate": 8.341268101415946e-05, "loss": 0.0686, "step": 30370 }, { "grad_norm": 0.895359456539154, "learning_rate": 8.34003784709999e-05, "loss": 0.0554, "step": 30380 }, { "grad_norm": 0.3926224410533905, "learning_rate": 8.338807227522516e-05, "loss": 0.0648, "step": 30390 }, { "grad_norm": 0.4790317118167877, "learning_rate": 8.337576242818103e-05, "loss": 0.0587, "step": 30400 }, { "grad_norm": 0.43643510341644287, "learning_rate": 8.336344893121372e-05, "loss": 0.0591, "step": 30410 }, { "grad_norm": 0.38932907581329346, "learning_rate": 8.335113178566977e-05, "loss": 0.054, "step": 30420 }, { "grad_norm": 0.5880573987960815, "learning_rate": 8.33388109928962e-05, "loss": 0.0714, "step": 30430 }, { "grad_norm": 0.4062415361404419, "learning_rate": 8.33264865542404e-05, "loss": 0.0637, "step": 30440 }, { "grad_norm": 0.2878071367740631, "learning_rate": 8.331415847105013e-05, "loss": 0.0583, "step": 30450 }, { "grad_norm": 0.5273362398147583, "learning_rate": 8.330182674467357e-05, "loss": 0.0518, "step": 30460 }, { "grad_norm": 0.71452796459198, "learning_rate": 8.32894913764593e-05, "loss": 0.0653, "step": 30470 }, { "grad_norm": 0.9336433410644531, "learning_rate": 8.327715236775633e-05, "loss": 0.0624, "step": 30480 }, { "grad_norm": 0.5457044839859009, "learning_rate": 8.326480971991398e-05, "loss": 0.0545, "step": 30490 }, { "grad_norm": 0.5606716871261597, "learning_rate": 8.325246343428206e-05, "loss": 0.0663, "step": 30500 }, { "grad_norm": 0.4425448179244995, "learning_rate": 8.324011351221072e-05, "loss": 0.0608, "step": 30510 }, { "grad_norm": 0.5132929086685181, "learning_rate": 8.322775995505057e-05, "loss": 0.0642, "step": 30520 }, { "grad_norm": 0.4810095727443695, "learning_rate": 8.321540276415254e-05, "loss": 0.0613, "step": 30530 }, { "grad_norm": 0.4656406044960022, "learning_rate": 8.320304194086798e-05, "loss": 0.0589, "step": 30540 }, { "grad_norm": 0.5993728637695312, "learning_rate": 8.31906774865487e-05, "loss": 0.0517, "step": 30550 }, { "grad_norm": 0.5574751496315002, "learning_rate": 8.317830940254682e-05, "loss": 0.0641, "step": 30560 }, { "grad_norm": 0.5150285959243774, "learning_rate": 8.316593769021491e-05, "loss": 0.0688, "step": 30570 }, { "grad_norm": 0.5526305437088013, "learning_rate": 8.315356235090592e-05, "loss": 0.0617, "step": 30580 }, { "grad_norm": 0.9663777947425842, "learning_rate": 8.314118338597319e-05, "loss": 0.064, "step": 30590 }, { "grad_norm": 1.5009067058563232, "learning_rate": 8.312880079677048e-05, "loss": 0.0627, "step": 30600 }, { "grad_norm": 0.5993451476097107, "learning_rate": 8.311641458465191e-05, "loss": 0.0754, "step": 30610 }, { "grad_norm": 0.7434232234954834, "learning_rate": 8.310402475097205e-05, "loss": 0.0665, "step": 30620 }, { "grad_norm": 1.05253005027771, "learning_rate": 8.309163129708581e-05, "loss": 0.0674, "step": 30630 }, { "grad_norm": 1.554256558418274, "learning_rate": 8.307923422434852e-05, "loss": 0.0631, "step": 30640 }, { "grad_norm": 0.9912412166595459, "learning_rate": 8.30668335341159e-05, "loss": 0.0749, "step": 30650 }, { "grad_norm": 0.5401330590248108, "learning_rate": 8.305442922774408e-05, "loss": 0.069, "step": 30660 }, { "grad_norm": 0.5003570318222046, "learning_rate": 8.304202130658959e-05, "loss": 0.0562, "step": 30670 }, { "grad_norm": 0.44652217626571655, "learning_rate": 8.302960977200931e-05, "loss": 0.0495, "step": 30680 }, { "grad_norm": 0.9144473075866699, "learning_rate": 8.301719462536058e-05, "loss": 0.0601, "step": 30690 }, { "grad_norm": 0.6388041377067566, "learning_rate": 8.300477586800108e-05, "loss": 0.0691, "step": 30700 }, { "grad_norm": 0.3733232617378235, "learning_rate": 8.299235350128892e-05, "loss": 0.0608, "step": 30710 }, { "grad_norm": 0.5023977160453796, "learning_rate": 8.297992752658258e-05, "loss": 0.0528, "step": 30720 }, { "grad_norm": 0.61318039894104, "learning_rate": 8.296749794524094e-05, "loss": 0.0625, "step": 30730 }, { "grad_norm": 0.5483152270317078, "learning_rate": 8.295506475862332e-05, "loss": 0.0467, "step": 30740 }, { "grad_norm": 0.5686722993850708, "learning_rate": 8.294262796808933e-05, "loss": 0.0509, "step": 30750 }, { "grad_norm": 0.5043702125549316, "learning_rate": 8.293018757499909e-05, "loss": 0.0644, "step": 30760 }, { "grad_norm": 0.6699415445327759, "learning_rate": 8.291774358071305e-05, "loss": 0.0589, "step": 30770 }, { "grad_norm": 0.6416455507278442, "learning_rate": 8.290529598659205e-05, "loss": 0.0721, "step": 30780 }, { "grad_norm": 0.949528694152832, "learning_rate": 8.289284479399738e-05, "loss": 0.0599, "step": 30790 }, { "grad_norm": 0.4191134572029114, "learning_rate": 8.288039000429064e-05, "loss": 0.0641, "step": 30800 }, { "grad_norm": 1.96027672290802, "learning_rate": 8.286793161883388e-05, "loss": 0.0677, "step": 30810 }, { "grad_norm": 0.846208930015564, "learning_rate": 8.285546963898954e-05, "loss": 0.0635, "step": 30820 }, { "grad_norm": 0.5608425140380859, "learning_rate": 8.284300406612044e-05, "loss": 0.0581, "step": 30830 }, { "grad_norm": 0.6369857788085938, "learning_rate": 8.283053490158978e-05, "loss": 0.0521, "step": 30840 }, { "grad_norm": 0.45156049728393555, "learning_rate": 8.28180621467612e-05, "loss": 0.0567, "step": 30850 }, { "grad_norm": 0.4820041060447693, "learning_rate": 8.280558580299867e-05, "loss": 0.0611, "step": 30860 }, { "grad_norm": 0.6054560542106628, "learning_rate": 8.279310587166661e-05, "loss": 0.0562, "step": 30870 }, { "grad_norm": 0.40808776021003723, "learning_rate": 8.278062235412978e-05, "loss": 0.0533, "step": 30880 }, { "grad_norm": 0.5562430024147034, "learning_rate": 8.276813525175339e-05, "loss": 0.0653, "step": 30890 }, { "grad_norm": 0.5394701957702637, "learning_rate": 8.2755644565903e-05, "loss": 0.0646, "step": 30900 }, { "grad_norm": 0.5351611971855164, "learning_rate": 8.274315029794454e-05, "loss": 0.052, "step": 30910 }, { "grad_norm": 0.8609890341758728, "learning_rate": 8.273065244924443e-05, "loss": 0.0633, "step": 30920 }, { "grad_norm": 0.7012892365455627, "learning_rate": 8.271815102116936e-05, "loss": 0.0587, "step": 30930 }, { "grad_norm": 0.3751330077648163, "learning_rate": 8.270564601508648e-05, "loss": 0.0565, "step": 30940 }, { "grad_norm": 0.5645127296447754, "learning_rate": 8.269313743236333e-05, "loss": 0.0516, "step": 30950 }, { "grad_norm": 0.29254239797592163, "learning_rate": 8.268062527436783e-05, "loss": 0.0482, "step": 30960 }, { "grad_norm": 0.4530227780342102, "learning_rate": 8.266810954246828e-05, "loss": 0.0578, "step": 30970 }, { "grad_norm": 0.4914531707763672, "learning_rate": 8.265559023803338e-05, "loss": 0.0425, "step": 30980 }, { "grad_norm": 0.4409205913543701, "learning_rate": 8.264306736243223e-05, "loss": 0.0557, "step": 30990 }, { "grad_norm": 1.1543058156967163, "learning_rate": 8.263054091703432e-05, "loss": 0.0722, "step": 31000 }, { "grad_norm": 0.6030990481376648, "learning_rate": 8.26180109032095e-05, "loss": 0.0552, "step": 31010 }, { "grad_norm": 0.7374757528305054, "learning_rate": 8.260547732232807e-05, "loss": 0.0563, "step": 31020 }, { "grad_norm": 0.5537574887275696, "learning_rate": 8.259294017576063e-05, "loss": 0.0561, "step": 31030 }, { "grad_norm": 0.4862825274467468, "learning_rate": 8.258039946487828e-05, "loss": 0.053, "step": 31040 }, { "grad_norm": 0.6401965618133545, "learning_rate": 8.256785519105241e-05, "loss": 0.0572, "step": 31050 }, { "grad_norm": 0.6942881941795349, "learning_rate": 8.255530735565488e-05, "loss": 0.0643, "step": 31060 }, { "grad_norm": 0.7059262990951538, "learning_rate": 8.254275596005787e-05, "loss": 0.0617, "step": 31070 }, { "grad_norm": 0.7259795665740967, "learning_rate": 8.2530201005634e-05, "loss": 0.0722, "step": 31080 }, { "grad_norm": 0.39795562624931335, "learning_rate": 8.251764249375626e-05, "loss": 0.0638, "step": 31090 }, { "grad_norm": 0.5820974111557007, "learning_rate": 8.250508042579803e-05, "loss": 0.0585, "step": 31100 }, { "grad_norm": 1.3624937534332275, "learning_rate": 8.249251480313307e-05, "loss": 0.0626, "step": 31110 }, { "grad_norm": 0.8644423484802246, "learning_rate": 8.247994562713555e-05, "loss": 0.0685, "step": 31120 }, { "grad_norm": 0.9594868421554565, "learning_rate": 8.246737289917998e-05, "loss": 0.0761, "step": 31130 }, { "grad_norm": 0.4765201210975647, "learning_rate": 8.245479662064135e-05, "loss": 0.0623, "step": 31140 }, { "grad_norm": 1.02260160446167, "learning_rate": 8.244221679289496e-05, "loss": 0.0528, "step": 31150 }, { "grad_norm": 0.8103088140487671, "learning_rate": 8.242963341731652e-05, "loss": 0.0533, "step": 31160 }, { "grad_norm": 0.5529189705848694, "learning_rate": 8.24170464952821e-05, "loss": 0.0624, "step": 31170 }, { "grad_norm": 0.6057983636856079, "learning_rate": 8.240445602816824e-05, "loss": 0.0556, "step": 31180 }, { "grad_norm": 0.7227347493171692, "learning_rate": 8.239186201735179e-05, "loss": 0.0677, "step": 31190 }, { "grad_norm": 0.38194864988327026, "learning_rate": 8.237926446420998e-05, "loss": 0.0572, "step": 31200 }, { "grad_norm": 0.6776657104492188, "learning_rate": 8.236666337012052e-05, "loss": 0.064, "step": 31210 }, { "grad_norm": 0.5091224908828735, "learning_rate": 8.23540587364614e-05, "loss": 0.0636, "step": 31220 }, { "grad_norm": 0.3513563275337219, "learning_rate": 8.234145056461107e-05, "loss": 0.0567, "step": 31230 }, { "grad_norm": 0.35923081636428833, "learning_rate": 8.232883885594832e-05, "loss": 0.0653, "step": 31240 }, { "grad_norm": 0.557509183883667, "learning_rate": 8.231622361185236e-05, "loss": 0.0556, "step": 31250 }, { "grad_norm": 0.7025154829025269, "learning_rate": 8.230360483370278e-05, "loss": 0.0849, "step": 31260 }, { "grad_norm": 0.7090117931365967, "learning_rate": 8.229098252287953e-05, "loss": 0.0636, "step": 31270 }, { "grad_norm": 0.6077250838279724, "learning_rate": 8.2278356680763e-05, "loss": 0.0663, "step": 31280 }, { "grad_norm": 0.8602210879325867, "learning_rate": 8.22657273087339e-05, "loss": 0.067, "step": 31290 }, { "grad_norm": 0.6391111016273499, "learning_rate": 8.225309440817336e-05, "loss": 0.0716, "step": 31300 }, { "grad_norm": 0.5045356154441833, "learning_rate": 8.224045798046293e-05, "loss": 0.0848, "step": 31310 }, { "grad_norm": 0.44719213247299194, "learning_rate": 8.22278180269845e-05, "loss": 0.069, "step": 31320 }, { "grad_norm": 0.4780941307544708, "learning_rate": 8.221517454912031e-05, "loss": 0.056, "step": 31330 }, { "grad_norm": 0.4820147752761841, "learning_rate": 8.220252754825308e-05, "loss": 0.0443, "step": 31340 }, { "grad_norm": 0.6303141713142395, "learning_rate": 8.218987702576586e-05, "loss": 0.0533, "step": 31350 }, { "grad_norm": 0.48218873143196106, "learning_rate": 8.217722298304207e-05, "loss": 0.0563, "step": 31360 }, { "grad_norm": 0.37946617603302, "learning_rate": 8.216456542146557e-05, "loss": 0.0587, "step": 31370 }, { "grad_norm": 0.43997934460639954, "learning_rate": 8.215190434242055e-05, "loss": 0.047, "step": 31380 }, { "grad_norm": 0.6932762265205383, "learning_rate": 8.213923974729161e-05, "loss": 0.0616, "step": 31390 }, { "grad_norm": 0.6034875512123108, "learning_rate": 8.212657163746373e-05, "loss": 0.0592, "step": 31400 }, { "grad_norm": 0.4288690984249115, "learning_rate": 8.211390001432227e-05, "loss": 0.0529, "step": 31410 }, { "grad_norm": 0.4363257586956024, "learning_rate": 8.210122487925297e-05, "loss": 0.0611, "step": 31420 }, { "grad_norm": 0.629804790019989, "learning_rate": 8.208854623364202e-05, "loss": 0.054, "step": 31430 }, { "grad_norm": 0.3011800944805145, "learning_rate": 8.207586407887589e-05, "loss": 0.0534, "step": 31440 }, { "grad_norm": 0.5163931250572205, "learning_rate": 8.206317841634148e-05, "loss": 0.0574, "step": 31450 }, { "grad_norm": 0.4426804184913635, "learning_rate": 8.205048924742609e-05, "loss": 0.0592, "step": 31460 }, { "grad_norm": 0.8727700114250183, "learning_rate": 8.203779657351738e-05, "loss": 0.0612, "step": 31470 }, { "grad_norm": 0.3968464732170105, "learning_rate": 8.20251003960034e-05, "loss": 0.0624, "step": 31480 }, { "grad_norm": 0.4782809615135193, "learning_rate": 8.201240071627258e-05, "loss": 0.0612, "step": 31490 }, { "grad_norm": 0.7803506255149841, "learning_rate": 8.199969753571377e-05, "loss": 0.0674, "step": 31500 }, { "grad_norm": 1.0067123174667358, "learning_rate": 8.198699085571615e-05, "loss": 0.0596, "step": 31510 }, { "grad_norm": 0.7031679749488831, "learning_rate": 8.197428067766928e-05, "loss": 0.0629, "step": 31520 }, { "grad_norm": 0.8801400065422058, "learning_rate": 8.196156700296316e-05, "loss": 0.0885, "step": 31530 }, { "grad_norm": 0.5300276875495911, "learning_rate": 8.194884983298814e-05, "loss": 0.0513, "step": 31540 }, { "grad_norm": 0.4622477889060974, "learning_rate": 8.193612916913491e-05, "loss": 0.0519, "step": 31550 }, { "grad_norm": 0.5722819566726685, "learning_rate": 8.192340501279463e-05, "loss": 0.0557, "step": 31560 }, { "grad_norm": 0.42342159152030945, "learning_rate": 8.191067736535876e-05, "loss": 0.0609, "step": 31570 }, { "grad_norm": 0.5127863883972168, "learning_rate": 8.18979462282192e-05, "loss": 0.0647, "step": 31580 }, { "grad_norm": 0.5620582699775696, "learning_rate": 8.188521160276819e-05, "loss": 0.0629, "step": 31590 }, { "grad_norm": 0.4599693715572357, "learning_rate": 8.187247349039837e-05, "loss": 0.0632, "step": 31600 }, { "grad_norm": 0.5614412426948547, "learning_rate": 8.185973189250278e-05, "loss": 0.069, "step": 31610 }, { "grad_norm": 1.1717214584350586, "learning_rate": 8.184698681047482e-05, "loss": 0.0748, "step": 31620 }, { "grad_norm": 0.4474904537200928, "learning_rate": 8.183423824570827e-05, "loss": 0.0641, "step": 31630 }, { "grad_norm": 0.5029829740524292, "learning_rate": 8.182148619959725e-05, "loss": 0.0612, "step": 31640 }, { "grad_norm": 0.5019099116325378, "learning_rate": 8.180873067353636e-05, "loss": 0.0612, "step": 31650 }, { "grad_norm": 0.5972543954849243, "learning_rate": 8.179597166892052e-05, "loss": 0.0549, "step": 31660 }, { "grad_norm": 0.6281343102455139, "learning_rate": 8.178320918714501e-05, "loss": 0.0577, "step": 31670 }, { "grad_norm": 1.3616031408309937, "learning_rate": 8.177044322960554e-05, "loss": 0.0635, "step": 31680 }, { "grad_norm": 0.5163766741752625, "learning_rate": 8.175767379769816e-05, "loss": 0.0638, "step": 31690 }, { "grad_norm": 0.5396339893341064, "learning_rate": 8.174490089281932e-05, "loss": 0.058, "step": 31700 }, { "grad_norm": 0.5029247999191284, "learning_rate": 8.173212451636584e-05, "loss": 0.0794, "step": 31710 }, { "grad_norm": 0.7359030842781067, "learning_rate": 8.171934466973493e-05, "loss": 0.0733, "step": 31720 }, { "grad_norm": 0.3116709589958191, "learning_rate": 8.170656135432418e-05, "loss": 0.0562, "step": 31730 }, { "grad_norm": 0.4809320271015167, "learning_rate": 8.169377457153155e-05, "loss": 0.0616, "step": 31740 }, { "grad_norm": 0.49817201495170593, "learning_rate": 8.168098432275539e-05, "loss": 0.0647, "step": 31750 }, { "grad_norm": 0.802686870098114, "learning_rate": 8.166819060939442e-05, "loss": 0.0662, "step": 31760 }, { "grad_norm": 0.4966445565223694, "learning_rate": 8.165539343284772e-05, "loss": 0.066, "step": 31770 }, { "grad_norm": 0.8241668939590454, "learning_rate": 8.16425927945148e-05, "loss": 0.0745, "step": 31780 }, { "grad_norm": 0.5937339663505554, "learning_rate": 8.162978869579551e-05, "loss": 0.0606, "step": 31790 }, { "grad_norm": 0.4813052713871002, "learning_rate": 8.161698113809007e-05, "loss": 0.0506, "step": 31800 }, { "grad_norm": 0.5352398753166199, "learning_rate": 8.160417012279911e-05, "loss": 0.0774, "step": 31810 }, { "grad_norm": 0.36431393027305603, "learning_rate": 8.159135565132363e-05, "loss": 0.0649, "step": 31820 }, { "grad_norm": 1.3180958032608032, "learning_rate": 8.157853772506498e-05, "loss": 0.0767, "step": 31830 }, { "grad_norm": 0.5452778339385986, "learning_rate": 8.156571634542494e-05, "loss": 0.0675, "step": 31840 }, { "grad_norm": 0.7317880392074585, "learning_rate": 8.15528915138056e-05, "loss": 0.0581, "step": 31850 }, { "grad_norm": 0.3637947142124176, "learning_rate": 8.154006323160949e-05, "loss": 0.0519, "step": 31860 }, { "grad_norm": 0.5453394055366516, "learning_rate": 8.152723150023949e-05, "loss": 0.0679, "step": 31870 }, { "grad_norm": 0.5266157984733582, "learning_rate": 8.151439632109886e-05, "loss": 0.07, "step": 31880 }, { "grad_norm": 0.5318577885627747, "learning_rate": 8.150155769559122e-05, "loss": 0.0693, "step": 31890 }, { "grad_norm": 0.5791153907775879, "learning_rate": 8.148871562512058e-05, "loss": 0.0606, "step": 31900 }, { "grad_norm": 0.5915215611457825, "learning_rate": 8.147587011109136e-05, "loss": 0.0763, "step": 31910 }, { "grad_norm": 0.6306318640708923, "learning_rate": 8.14630211549083e-05, "loss": 0.0754, "step": 31920 }, { "grad_norm": 0.734525203704834, "learning_rate": 8.145016875797655e-05, "loss": 0.0513, "step": 31930 }, { "grad_norm": 0.4040755331516266, "learning_rate": 8.143731292170164e-05, "loss": 0.0632, "step": 31940 }, { "grad_norm": 1.188353180885315, "learning_rate": 8.142445364748944e-05, "loss": 0.0701, "step": 31950 }, { "grad_norm": 0.6023337244987488, "learning_rate": 8.141159093674624e-05, "loss": 0.0718, "step": 31960 }, { "grad_norm": 0.5888872146606445, "learning_rate": 8.139872479087869e-05, "loss": 0.056, "step": 31970 }, { "grad_norm": 0.481998473405838, "learning_rate": 8.13858552112938e-05, "loss": 0.0486, "step": 31980 }, { "grad_norm": 0.4752010107040405, "learning_rate": 8.137298219939895e-05, "loss": 0.0678, "step": 31990 }, { "grad_norm": 0.5924907326698303, "learning_rate": 8.136010575660196e-05, "loss": 0.0495, "step": 32000 }, { "grad_norm": 0.5815079808235168, "learning_rate": 8.134722588431095e-05, "loss": 0.0569, "step": 32010 }, { "grad_norm": 0.4628806710243225, "learning_rate": 8.133434258393444e-05, "loss": 0.0713, "step": 32020 }, { "grad_norm": 0.73814457654953, "learning_rate": 8.132145585688134e-05, "loss": 0.0638, "step": 32030 }, { "grad_norm": 0.5606095790863037, "learning_rate": 8.130856570456093e-05, "loss": 0.0693, "step": 32040 }, { "grad_norm": 0.40908029675483704, "learning_rate": 8.129567212838283e-05, "loss": 0.0641, "step": 32050 }, { "grad_norm": 0.5050182342529297, "learning_rate": 8.128277512975708e-05, "loss": 0.0687, "step": 32060 }, { "grad_norm": 0.5989733338356018, "learning_rate": 8.126987471009408e-05, "loss": 0.0643, "step": 32070 }, { "grad_norm": 0.3691343665122986, "learning_rate": 8.125697087080459e-05, "loss": 0.0546, "step": 32080 }, { "grad_norm": 0.8693310022354126, "learning_rate": 8.124406361329976e-05, "loss": 0.0655, "step": 32090 }, { "grad_norm": 0.5273008942604065, "learning_rate": 8.123115293899111e-05, "loss": 0.0563, "step": 32100 }, { "grad_norm": 0.44166964292526245, "learning_rate": 8.121823884929055e-05, "loss": 0.0533, "step": 32110 }, { "grad_norm": 0.8655091524124146, "learning_rate": 8.12053213456103e-05, "loss": 0.0509, "step": 32120 }, { "grad_norm": 0.3668372333049774, "learning_rate": 8.119240042936303e-05, "loss": 0.044, "step": 32130 }, { "grad_norm": 0.6155951619148254, "learning_rate": 8.117947610196175e-05, "loss": 0.0535, "step": 32140 }, { "grad_norm": 1.0675684213638306, "learning_rate": 8.116654836481982e-05, "loss": 0.0552, "step": 32150 }, { "grad_norm": 0.5141671299934387, "learning_rate": 8.115361721935106e-05, "loss": 0.0657, "step": 32160 }, { "grad_norm": 0.6711562871932983, "learning_rate": 8.114068266696953e-05, "loss": 0.0636, "step": 32170 }, { "grad_norm": 0.3591022491455078, "learning_rate": 8.112774470908978e-05, "loss": 0.0449, "step": 32180 }, { "grad_norm": 0.548295259475708, "learning_rate": 8.111480334712665e-05, "loss": 0.0464, "step": 32190 }, { "grad_norm": 0.40029850602149963, "learning_rate": 8.110185858249542e-05, "loss": 0.0555, "step": 32200 }, { "grad_norm": 0.6490981578826904, "learning_rate": 8.108891041661168e-05, "loss": 0.0638, "step": 32210 }, { "grad_norm": 0.5296074748039246, "learning_rate": 8.107595885089146e-05, "loss": 0.063, "step": 32220 }, { "grad_norm": 0.37173137068748474, "learning_rate": 8.106300388675108e-05, "loss": 0.0598, "step": 32230 }, { "grad_norm": 0.5059314966201782, "learning_rate": 8.105004552560731e-05, "loss": 0.0612, "step": 32240 }, { "grad_norm": 0.5468874573707581, "learning_rate": 8.103708376887724e-05, "loss": 0.0568, "step": 32250 }, { "grad_norm": 0.46221235394477844, "learning_rate": 8.102411861797836e-05, "loss": 0.0499, "step": 32260 }, { "grad_norm": 0.40487515926361084, "learning_rate": 8.101115007432851e-05, "loss": 0.0471, "step": 32270 }, { "grad_norm": 0.4649102985858917, "learning_rate": 8.09981781393459e-05, "loss": 0.0565, "step": 32280 }, { "grad_norm": 0.5750274658203125, "learning_rate": 8.098520281444915e-05, "loss": 0.0525, "step": 32290 }, { "grad_norm": 0.6186044812202454, "learning_rate": 8.09722241010572e-05, "loss": 0.046, "step": 32300 }, { "grad_norm": 0.5155160427093506, "learning_rate": 8.095924200058939e-05, "loss": 0.0555, "step": 32310 }, { "grad_norm": 0.4017265737056732, "learning_rate": 8.094625651446541e-05, "loss": 0.0456, "step": 32320 }, { "grad_norm": 0.3035799562931061, "learning_rate": 8.093326764410536e-05, "loss": 0.0585, "step": 32330 }, { "grad_norm": 0.5298442840576172, "learning_rate": 8.092027539092966e-05, "loss": 0.0616, "step": 32340 }, { "grad_norm": 0.592048704624176, "learning_rate": 8.090727975635913e-05, "loss": 0.0538, "step": 32350 }, { "grad_norm": 0.6615852117538452, "learning_rate": 8.089428074181497e-05, "loss": 0.0596, "step": 32360 }, { "grad_norm": 0.41837677359580994, "learning_rate": 8.088127834871871e-05, "loss": 0.0473, "step": 32370 }, { "grad_norm": 0.6399111151695251, "learning_rate": 8.086827257849226e-05, "loss": 0.0549, "step": 32380 }, { "grad_norm": 0.4828062057495117, "learning_rate": 8.085526343255795e-05, "loss": 0.0472, "step": 32390 }, { "grad_norm": 1.435814619064331, "learning_rate": 8.084225091233842e-05, "loss": 0.0649, "step": 32400 }, { "grad_norm": 0.9133020639419556, "learning_rate": 8.082923501925668e-05, "loss": 0.0464, "step": 32410 }, { "grad_norm": 0.7999880909919739, "learning_rate": 8.081621575473617e-05, "loss": 0.069, "step": 32420 }, { "grad_norm": 0.439078152179718, "learning_rate": 8.080319312020064e-05, "loss": 0.0646, "step": 32430 }, { "grad_norm": 0.5602756142616272, "learning_rate": 8.079016711707421e-05, "loss": 0.061, "step": 32440 }, { "grad_norm": 0.5063182711601257, "learning_rate": 8.077713774678139e-05, "loss": 0.0597, "step": 32450 }, { "grad_norm": 0.6806217432022095, "learning_rate": 8.076410501074707e-05, "loss": 0.0597, "step": 32460 }, { "grad_norm": 0.6050848364830017, "learning_rate": 8.075106891039647e-05, "loss": 0.0709, "step": 32470 }, { "grad_norm": 0.48047691583633423, "learning_rate": 8.073802944715523e-05, "loss": 0.0624, "step": 32480 }, { "grad_norm": 0.570381760597229, "learning_rate": 8.072498662244929e-05, "loss": 0.0571, "step": 32490 }, { "grad_norm": 0.42404988408088684, "learning_rate": 8.0711940437705e-05, "loss": 0.0563, "step": 32500 }, { "grad_norm": 0.5422581434249878, "learning_rate": 8.06988908943491e-05, "loss": 0.0518, "step": 32510 }, { "grad_norm": 0.7010546922683716, "learning_rate": 8.068583799380863e-05, "loss": 0.0467, "step": 32520 }, { "grad_norm": 0.6589838862419128, "learning_rate": 8.067278173751104e-05, "loss": 0.0554, "step": 32530 }, { "grad_norm": 0.5232466459274292, "learning_rate": 8.065972212688417e-05, "loss": 0.0564, "step": 32540 }, { "grad_norm": 0.5980812311172485, "learning_rate": 8.064665916335618e-05, "loss": 0.0677, "step": 32550 }, { "grad_norm": 0.3247774541378021, "learning_rate": 8.063359284835564e-05, "loss": 0.067, "step": 32560 }, { "grad_norm": 0.4787070155143738, "learning_rate": 8.062052318331142e-05, "loss": 0.0743, "step": 32570 }, { "grad_norm": 0.5737035274505615, "learning_rate": 8.060745016965283e-05, "loss": 0.0652, "step": 32580 }, { "grad_norm": 0.4289534389972687, "learning_rate": 8.059437380880952e-05, "loss": 0.0516, "step": 32590 }, { "grad_norm": 0.35883596539497375, "learning_rate": 8.058129410221146e-05, "loss": 0.0532, "step": 32600 }, { "grad_norm": 0.44093385338783264, "learning_rate": 8.056821105128908e-05, "loss": 0.0606, "step": 32610 }, { "grad_norm": 0.56224524974823, "learning_rate": 8.05551246574731e-05, "loss": 0.0528, "step": 32620 }, { "grad_norm": 0.6649956703186035, "learning_rate": 8.05420349221946e-05, "loss": 0.0562, "step": 32630 }, { "grad_norm": 0.37617817521095276, "learning_rate": 8.05289418468851e-05, "loss": 0.04, "step": 32640 }, { "grad_norm": 0.45690789818763733, "learning_rate": 8.051584543297642e-05, "loss": 0.0603, "step": 32650 }, { "grad_norm": 0.4130665361881256, "learning_rate": 8.050274568190074e-05, "loss": 0.0589, "step": 32660 }, { "grad_norm": 0.652391254901886, "learning_rate": 8.048964259509067e-05, "loss": 0.0566, "step": 32670 }, { "grad_norm": 0.581281840801239, "learning_rate": 8.047653617397914e-05, "loss": 0.0546, "step": 32680 }, { "grad_norm": 0.5378548502922058, "learning_rate": 8.046342641999941e-05, "loss": 0.0476, "step": 32690 }, { "grad_norm": 0.4302235245704651, "learning_rate": 8.045031333458517e-05, "loss": 0.0559, "step": 32700 }, { "grad_norm": 0.6036633849143982, "learning_rate": 8.043719691917047e-05, "loss": 0.0549, "step": 32710 }, { "grad_norm": 0.5198573470115662, "learning_rate": 8.042407717518966e-05, "loss": 0.0549, "step": 32720 }, { "grad_norm": 0.900503396987915, "learning_rate": 8.041095410407751e-05, "loss": 0.0545, "step": 32730 }, { "grad_norm": 0.41239362955093384, "learning_rate": 8.039782770726913e-05, "loss": 0.0502, "step": 32740 }, { "grad_norm": 0.48063841462135315, "learning_rate": 8.038469798620004e-05, "loss": 0.0593, "step": 32750 }, { "grad_norm": 0.41475072503089905, "learning_rate": 8.037156494230604e-05, "loss": 0.0621, "step": 32760 }, { "grad_norm": 0.45973411202430725, "learning_rate": 8.035842857702338e-05, "loss": 0.046, "step": 32770 }, { "grad_norm": 0.5618531703948975, "learning_rate": 8.03452888917886e-05, "loss": 0.0598, "step": 32780 }, { "grad_norm": 0.46561765670776367, "learning_rate": 8.033214588803866e-05, "loss": 0.0552, "step": 32790 }, { "grad_norm": 0.5256884098052979, "learning_rate": 8.031899956721083e-05, "loss": 0.0509, "step": 32800 }, { "grad_norm": 0.5032817721366882, "learning_rate": 8.030584993074282e-05, "loss": 0.0577, "step": 32810 }, { "grad_norm": 0.41420450806617737, "learning_rate": 8.02926969800726e-05, "loss": 0.0519, "step": 32820 }, { "grad_norm": 0.6233036518096924, "learning_rate": 8.027954071663859e-05, "loss": 0.0635, "step": 32830 }, { "grad_norm": 0.7902959585189819, "learning_rate": 8.026638114187954e-05, "loss": 0.0603, "step": 32840 }, { "grad_norm": 0.6621007919311523, "learning_rate": 8.025321825723456e-05, "loss": 0.0625, "step": 32850 }, { "grad_norm": 0.6839601397514343, "learning_rate": 8.02400520641431e-05, "loss": 0.0735, "step": 32860 }, { "grad_norm": 0.3949538469314575, "learning_rate": 8.022688256404501e-05, "loss": 0.0561, "step": 32870 }, { "grad_norm": 0.6714335083961487, "learning_rate": 8.02137097583805e-05, "loss": 0.0596, "step": 32880 }, { "grad_norm": 0.5535971522331238, "learning_rate": 8.02005336485901e-05, "loss": 0.0647, "step": 32890 }, { "grad_norm": 0.5630702972412109, "learning_rate": 8.018735423611476e-05, "loss": 0.0662, "step": 32900 }, { "grad_norm": 0.5642964243888855, "learning_rate": 8.017417152239574e-05, "loss": 0.0557, "step": 32910 }, { "grad_norm": 0.6484436392784119, "learning_rate": 8.01609855088747e-05, "loss": 0.066, "step": 32920 }, { "grad_norm": 0.5748017430305481, "learning_rate": 8.014779619699362e-05, "loss": 0.0538, "step": 32930 }, { "grad_norm": 0.3928404748439789, "learning_rate": 8.013460358819489e-05, "loss": 0.0542, "step": 32940 }, { "grad_norm": 0.6996949315071106, "learning_rate": 8.01214076839212e-05, "loss": 0.0615, "step": 32950 }, { "grad_norm": 0.5693107843399048, "learning_rate": 8.010820848561565e-05, "loss": 0.0534, "step": 32960 }, { "grad_norm": 0.5507225394248962, "learning_rate": 8.009500599472171e-05, "loss": 0.0493, "step": 32970 }, { "grad_norm": 0.9171202182769775, "learning_rate": 8.008180021268314e-05, "loss": 0.0613, "step": 32980 }, { "grad_norm": 0.5149894952774048, "learning_rate": 8.006859114094414e-05, "loss": 0.0567, "step": 32990 }, { "grad_norm": 0.514141857624054, "learning_rate": 8.005537878094921e-05, "loss": 0.0645, "step": 33000 }, { "grad_norm": 0.4283270537853241, "learning_rate": 8.004216313414323e-05, "loss": 0.0623, "step": 33010 }, { "grad_norm": 0.39192816615104675, "learning_rate": 8.002894420197149e-05, "loss": 0.0527, "step": 33020 }, { "grad_norm": 0.4826791286468506, "learning_rate": 8.001572198587954e-05, "loss": 0.0521, "step": 33030 }, { "grad_norm": 2.0709481239318848, "learning_rate": 8.000249648731338e-05, "loss": 0.0632, "step": 33040 }, { "grad_norm": 0.5439820885658264, "learning_rate": 7.998926770771928e-05, "loss": 0.0534, "step": 33050 }, { "grad_norm": 0.5970486402511597, "learning_rate": 7.997603564854397e-05, "loss": 0.0622, "step": 33060 }, { "grad_norm": 0.43660399317741394, "learning_rate": 7.996280031123448e-05, "loss": 0.0507, "step": 33070 }, { "grad_norm": 0.7319742441177368, "learning_rate": 7.994956169723818e-05, "loss": 0.0708, "step": 33080 }, { "grad_norm": 0.4790610671043396, "learning_rate": 7.993631980800285e-05, "loss": 0.0498, "step": 33090 }, { "grad_norm": 0.5353416204452515, "learning_rate": 7.992307464497659e-05, "loss": 0.0703, "step": 33100 }, { "grad_norm": 0.38838034868240356, "learning_rate": 7.990982620960787e-05, "loss": 0.0507, "step": 33110 }, { "grad_norm": 0.7627564072608948, "learning_rate": 7.989657450334554e-05, "loss": 0.0633, "step": 33120 }, { "grad_norm": 0.5267699956893921, "learning_rate": 7.988331952763877e-05, "loss": 0.0585, "step": 33130 }, { "grad_norm": 0.4953901171684265, "learning_rate": 7.987006128393709e-05, "loss": 0.0556, "step": 33140 }, { "grad_norm": 1.0917409658432007, "learning_rate": 7.985679977369043e-05, "loss": 0.049, "step": 33150 }, { "grad_norm": 0.5971428155899048, "learning_rate": 7.984353499834902e-05, "loss": 0.0519, "step": 33160 }, { "grad_norm": 0.5668661594390869, "learning_rate": 7.983026695936351e-05, "loss": 0.0604, "step": 33170 }, { "grad_norm": 0.5030097961425781, "learning_rate": 7.981699565818486e-05, "loss": 0.058, "step": 33180 }, { "grad_norm": 0.4551691710948944, "learning_rate": 7.980372109626437e-05, "loss": 0.0535, "step": 33190 }, { "grad_norm": 0.3932743966579437, "learning_rate": 7.979044327505375e-05, "loss": 0.0468, "step": 33200 }, { "grad_norm": 0.5332281589508057, "learning_rate": 7.977716219600506e-05, "loss": 0.0696, "step": 33210 }, { "grad_norm": 0.5805884599685669, "learning_rate": 7.97638778605707e-05, "loss": 0.0569, "step": 33220 }, { "grad_norm": 0.6536905765533447, "learning_rate": 7.975059027020338e-05, "loss": 0.0691, "step": 33230 }, { "grad_norm": 0.5756696462631226, "learning_rate": 7.973729942635623e-05, "loss": 0.058, "step": 33240 }, { "grad_norm": 0.6205094456672668, "learning_rate": 7.972400533048273e-05, "loss": 0.0654, "step": 33250 }, { "grad_norm": 0.42691919207572937, "learning_rate": 7.97107079840367e-05, "loss": 0.0642, "step": 33260 }, { "grad_norm": 0.6371417045593262, "learning_rate": 7.969740738847231e-05, "loss": 0.0534, "step": 33270 }, { "grad_norm": 0.45104125142097473, "learning_rate": 7.968410354524411e-05, "loss": 0.0526, "step": 33280 }, { "grad_norm": 0.37069788575172424, "learning_rate": 7.967079645580697e-05, "loss": 0.0478, "step": 33290 }, { "grad_norm": 0.4740864932537079, "learning_rate": 7.965748612161612e-05, "loss": 0.0502, "step": 33300 }, { "grad_norm": 0.6291195154190063, "learning_rate": 7.96441725441272e-05, "loss": 0.0651, "step": 33310 }, { "grad_norm": 1.3458194732666016, "learning_rate": 7.963085572479614e-05, "loss": 0.0608, "step": 33320 }, { "grad_norm": 0.44770586490631104, "learning_rate": 7.961753566507924e-05, "loss": 0.0534, "step": 33330 }, { "grad_norm": 0.5048211812973022, "learning_rate": 7.960421236643316e-05, "loss": 0.0632, "step": 33340 }, { "grad_norm": 0.4411725401878357, "learning_rate": 7.959088583031496e-05, "loss": 0.0545, "step": 33350 }, { "grad_norm": 0.5205578804016113, "learning_rate": 7.957755605818194e-05, "loss": 0.0563, "step": 33360 }, { "grad_norm": 0.3639248311519623, "learning_rate": 7.956422305149185e-05, "loss": 0.061, "step": 33370 }, { "grad_norm": 0.7015015482902527, "learning_rate": 7.95508868117028e-05, "loss": 0.0736, "step": 33380 }, { "grad_norm": 0.6064584851264954, "learning_rate": 7.953754734027318e-05, "loss": 0.0665, "step": 33390 }, { "grad_norm": 0.35887423157691956, "learning_rate": 7.952420463866182e-05, "loss": 0.0606, "step": 33400 }, { "grad_norm": 0.7727649807929993, "learning_rate": 7.951085870832782e-05, "loss": 0.0735, "step": 33410 }, { "grad_norm": 0.48945456743240356, "learning_rate": 7.949750955073067e-05, "loss": 0.0596, "step": 33420 }, { "grad_norm": 0.7497824430465698, "learning_rate": 7.948415716733022e-05, "loss": 0.0685, "step": 33430 }, { "grad_norm": 0.5924727320671082, "learning_rate": 7.947080155958669e-05, "loss": 0.0618, "step": 33440 }, { "grad_norm": 0.6121377348899841, "learning_rate": 7.94574427289606e-05, "loss": 0.0626, "step": 33450 }, { "grad_norm": 0.5538105964660645, "learning_rate": 7.944408067691284e-05, "loss": 0.0606, "step": 33460 }, { "grad_norm": 0.6274104118347168, "learning_rate": 7.943071540490473e-05, "loss": 0.055, "step": 33470 }, { "grad_norm": 0.587581217288971, "learning_rate": 7.94173469143978e-05, "loss": 0.0603, "step": 33480 }, { "grad_norm": 0.2996588945388794, "learning_rate": 7.940397520685406e-05, "loss": 0.0475, "step": 33490 }, { "grad_norm": 0.6255294680595398, "learning_rate": 7.939060028373577e-05, "loss": 0.0644, "step": 33500 }, { "grad_norm": 0.6436875462532043, "learning_rate": 7.937722214650565e-05, "loss": 0.0754, "step": 33510 }, { "grad_norm": 0.42025405168533325, "learning_rate": 7.936384079662666e-05, "loss": 0.061, "step": 33520 }, { "grad_norm": 0.6317248940467834, "learning_rate": 7.93504562355622e-05, "loss": 0.0595, "step": 33530 }, { "grad_norm": 0.44892218708992004, "learning_rate": 7.933706846477599e-05, "loss": 0.0531, "step": 33540 }, { "grad_norm": 0.5336245894432068, "learning_rate": 7.932367748573206e-05, "loss": 0.0598, "step": 33550 }, { "grad_norm": 0.5240049958229065, "learning_rate": 7.931028329989485e-05, "loss": 0.0514, "step": 33560 }, { "grad_norm": 0.6433811783790588, "learning_rate": 7.929688590872913e-05, "loss": 0.0807, "step": 33570 }, { "grad_norm": 0.5252797603607178, "learning_rate": 7.928348531370003e-05, "loss": 0.0547, "step": 33580 }, { "grad_norm": 0.6084861159324646, "learning_rate": 7.927008151627297e-05, "loss": 0.0606, "step": 33590 }, { "grad_norm": 0.8127127289772034, "learning_rate": 7.925667451791383e-05, "loss": 0.0607, "step": 33600 }, { "grad_norm": 0.7507365345954895, "learning_rate": 7.924326432008874e-05, "loss": 0.0628, "step": 33610 }, { "grad_norm": 0.46076837182044983, "learning_rate": 7.922985092426422e-05, "loss": 0.0614, "step": 33620 }, { "grad_norm": 0.7596030235290527, "learning_rate": 7.921643433190717e-05, "loss": 0.0773, "step": 33630 }, { "grad_norm": 0.9823193550109863, "learning_rate": 7.920301454448478e-05, "loss": 0.0697, "step": 33640 }, { "grad_norm": 0.4081360697746277, "learning_rate": 7.918959156346461e-05, "loss": 0.0543, "step": 33650 }, { "grad_norm": 0.6357094645500183, "learning_rate": 7.91761653903146e-05, "loss": 0.0512, "step": 33660 }, { "grad_norm": 0.6267735362052917, "learning_rate": 7.916273602650302e-05, "loss": 0.0635, "step": 33670 }, { "grad_norm": 0.6251793503761292, "learning_rate": 7.914930347349847e-05, "loss": 0.0633, "step": 33680 }, { "grad_norm": 0.664313793182373, "learning_rate": 7.913586773276992e-05, "loss": 0.0611, "step": 33690 }, { "grad_norm": 0.44110116362571716, "learning_rate": 7.912242880578667e-05, "loss": 0.0558, "step": 33700 }, { "grad_norm": 0.5825774073600769, "learning_rate": 7.910898669401839e-05, "loss": 0.0654, "step": 33710 }, { "grad_norm": 0.37359848618507385, "learning_rate": 7.909554139893511e-05, "loss": 0.0564, "step": 33720 }, { "grad_norm": 0.3518601357936859, "learning_rate": 7.908209292200715e-05, "loss": 0.0495, "step": 33730 }, { "grad_norm": 0.6769595742225647, "learning_rate": 7.906864126470523e-05, "loss": 0.0515, "step": 33740 }, { "grad_norm": 0.4446370005607605, "learning_rate": 7.905518642850041e-05, "loss": 0.0655, "step": 33750 }, { "grad_norm": 0.40196511149406433, "learning_rate": 7.904172841486409e-05, "loss": 0.067, "step": 33760 }, { "grad_norm": 0.813896656036377, "learning_rate": 7.902826722526801e-05, "loss": 0.0541, "step": 33770 }, { "grad_norm": 0.4181349575519562, "learning_rate": 7.901480286118427e-05, "loss": 0.0593, "step": 33780 }, { "grad_norm": 0.5122050642967224, "learning_rate": 7.900133532408531e-05, "loss": 0.0581, "step": 33790 }, { "grad_norm": 0.48879754543304443, "learning_rate": 7.898786461544395e-05, "loss": 0.0588, "step": 33800 }, { "grad_norm": 0.7890384197235107, "learning_rate": 7.897439073673325e-05, "loss": 0.0712, "step": 33810 }, { "grad_norm": 0.4728259742259979, "learning_rate": 7.896091368942677e-05, "loss": 0.068, "step": 33820 }, { "grad_norm": 0.7508488297462463, "learning_rate": 7.894743347499832e-05, "loss": 0.0541, "step": 33830 }, { "grad_norm": 0.4992750585079193, "learning_rate": 7.893395009492203e-05, "loss": 0.0682, "step": 33840 }, { "grad_norm": 0.47680598497390747, "learning_rate": 7.892046355067248e-05, "loss": 0.0628, "step": 33850 }, { "grad_norm": 0.586635172367096, "learning_rate": 7.890697384372451e-05, "loss": 0.0546, "step": 33860 }, { "grad_norm": 0.421977162361145, "learning_rate": 7.889348097555336e-05, "loss": 0.0523, "step": 33870 }, { "grad_norm": 0.48112475872039795, "learning_rate": 7.887998494763455e-05, "loss": 0.0543, "step": 33880 }, { "grad_norm": 0.36588120460510254, "learning_rate": 7.886648576144404e-05, "loss": 0.0628, "step": 33890 }, { "grad_norm": 0.5044573545455933, "learning_rate": 7.885298341845802e-05, "loss": 0.0529, "step": 33900 }, { "grad_norm": 0.5233827829360962, "learning_rate": 7.883947792015311e-05, "loss": 0.0593, "step": 33910 }, { "grad_norm": 0.6368498206138611, "learning_rate": 7.882596926800628e-05, "loss": 0.0808, "step": 33920 }, { "grad_norm": 0.6441311836242676, "learning_rate": 7.881245746349477e-05, "loss": 0.0648, "step": 33930 }, { "grad_norm": 0.7456949353218079, "learning_rate": 7.879894250809623e-05, "loss": 0.0549, "step": 33940 }, { "grad_norm": 0.4931800663471222, "learning_rate": 7.878542440328865e-05, "loss": 0.0706, "step": 33950 }, { "grad_norm": 0.47626611590385437, "learning_rate": 7.877190315055031e-05, "loss": 0.0482, "step": 33960 }, { "grad_norm": 0.5727335214614868, "learning_rate": 7.875837875135991e-05, "loss": 0.0557, "step": 33970 }, { "grad_norm": 0.5066584944725037, "learning_rate": 7.874485120719646e-05, "loss": 0.0677, "step": 33980 }, { "grad_norm": 0.4199373424053192, "learning_rate": 7.873132051953928e-05, "loss": 0.0643, "step": 33990 }, { "grad_norm": 0.9853399395942688, "learning_rate": 7.87177866898681e-05, "loss": 0.0615, "step": 34000 }, { "grad_norm": 0.4305740296840668, "learning_rate": 7.870424971966294e-05, "loss": 0.0555, "step": 34010 }, { "grad_norm": 0.3736969828605652, "learning_rate": 7.869070961040419e-05, "loss": 0.0476, "step": 34020 }, { "grad_norm": 0.5843911170959473, "learning_rate": 7.867716636357257e-05, "loss": 0.066, "step": 34030 }, { "grad_norm": 0.40793848037719727, "learning_rate": 7.866361998064915e-05, "loss": 0.0617, "step": 34040 }, { "grad_norm": 0.5342947840690613, "learning_rate": 7.865007046311534e-05, "loss": 0.0575, "step": 34050 }, { "grad_norm": 0.5667257905006409, "learning_rate": 7.86365178124529e-05, "loss": 0.0608, "step": 34060 }, { "grad_norm": 0.7109208703041077, "learning_rate": 7.862296203014394e-05, "loss": 0.0606, "step": 34070 }, { "grad_norm": 0.6003761887550354, "learning_rate": 7.860940311767088e-05, "loss": 0.0686, "step": 34080 }, { "grad_norm": 0.5703616738319397, "learning_rate": 7.85958410765165e-05, "loss": 0.0551, "step": 34090 }, { "grad_norm": 0.7996758222579956, "learning_rate": 7.858227590816394e-05, "loss": 0.0722, "step": 34100 }, { "grad_norm": 0.6561159491539001, "learning_rate": 7.856870761409664e-05, "loss": 0.0602, "step": 34110 }, { "grad_norm": 0.6985490918159485, "learning_rate": 7.855513619579846e-05, "loss": 0.0633, "step": 34120 }, { "grad_norm": 0.8791680335998535, "learning_rate": 7.85415616547535e-05, "loss": 0.0634, "step": 34130 }, { "grad_norm": 0.4123602509498596, "learning_rate": 7.852798399244627e-05, "loss": 0.0471, "step": 34140 }, { "grad_norm": 0.5135371088981628, "learning_rate": 7.851440321036161e-05, "loss": 0.0559, "step": 34150 }, { "grad_norm": 0.40155208110809326, "learning_rate": 7.850081930998468e-05, "loss": 0.0508, "step": 34160 }, { "grad_norm": 1.0135328769683838, "learning_rate": 7.8487232292801e-05, "loss": 0.0645, "step": 34170 }, { "grad_norm": 0.6539138555526733, "learning_rate": 7.847364216029642e-05, "loss": 0.0639, "step": 34180 }, { "grad_norm": 0.5220224857330322, "learning_rate": 7.846004891395716e-05, "loss": 0.0511, "step": 34190 }, { "grad_norm": 0.5895302295684814, "learning_rate": 7.844645255526972e-05, "loss": 0.0631, "step": 34200 }, { "grad_norm": 0.6312253475189209, "learning_rate": 7.843285308572101e-05, "loss": 0.057, "step": 34210 }, { "grad_norm": 0.3988078832626343, "learning_rate": 7.841925050679823e-05, "loss": 0.0428, "step": 34220 }, { "grad_norm": 0.8711577653884888, "learning_rate": 7.840564481998895e-05, "loss": 0.0639, "step": 34230 }, { "grad_norm": 0.7096341848373413, "learning_rate": 7.839203602678105e-05, "loss": 0.0704, "step": 34240 }, { "grad_norm": 0.45675233006477356, "learning_rate": 7.837842412866279e-05, "loss": 0.0708, "step": 34250 }, { "grad_norm": 0.4053291976451874, "learning_rate": 7.836480912712272e-05, "loss": 0.0637, "step": 34260 }, { "grad_norm": 0.6225526928901672, "learning_rate": 7.835119102364976e-05, "loss": 0.0551, "step": 34270 }, { "grad_norm": 0.5010768175125122, "learning_rate": 7.83375698197332e-05, "loss": 0.052, "step": 34280 }, { "grad_norm": 0.46718108654022217, "learning_rate": 7.83239455168626e-05, "loss": 0.057, "step": 34290 }, { "grad_norm": 0.5842958092689514, "learning_rate": 7.83103181165279e-05, "loss": 0.0565, "step": 34300 }, { "grad_norm": 0.5786224007606506, "learning_rate": 7.829668762021937e-05, "loss": 0.0704, "step": 34310 }, { "grad_norm": 0.8434241414070129, "learning_rate": 7.828305402942764e-05, "loss": 0.0568, "step": 34320 }, { "grad_norm": 0.8124279379844666, "learning_rate": 7.826941734564363e-05, "loss": 0.0555, "step": 34330 }, { "grad_norm": 0.415902704000473, "learning_rate": 7.825577757035865e-05, "loss": 0.0626, "step": 34340 }, { "grad_norm": 0.8394750952720642, "learning_rate": 7.824213470506431e-05, "loss": 0.0632, "step": 34350 }, { "grad_norm": 0.786805272102356, "learning_rate": 7.822848875125257e-05, "loss": 0.0573, "step": 34360 }, { "grad_norm": 0.3970300853252411, "learning_rate": 7.821483971041576e-05, "loss": 0.064, "step": 34370 }, { "grad_norm": 0.35905876755714417, "learning_rate": 7.820118758404649e-05, "loss": 0.0483, "step": 34380 }, { "grad_norm": 0.3717955946922302, "learning_rate": 7.818753237363776e-05, "loss": 0.0595, "step": 34390 }, { "grad_norm": 0.4562132656574249, "learning_rate": 7.817387408068286e-05, "loss": 0.0587, "step": 34400 }, { "grad_norm": 0.37019047141075134, "learning_rate": 7.816021270667544e-05, "loss": 0.0533, "step": 34410 }, { "grad_norm": 0.44066375494003296, "learning_rate": 7.81465482531095e-05, "loss": 0.0514, "step": 34420 }, { "grad_norm": 0.5696035623550415, "learning_rate": 7.813288072147938e-05, "loss": 0.0602, "step": 34430 }, { "grad_norm": 0.543287992477417, "learning_rate": 7.811921011327972e-05, "loss": 0.0616, "step": 34440 }, { "grad_norm": 0.3093728721141815, "learning_rate": 7.810553643000549e-05, "loss": 0.0554, "step": 34450 }, { "grad_norm": 0.6607248783111572, "learning_rate": 7.809185967315206e-05, "loss": 0.0589, "step": 34460 }, { "grad_norm": 0.5126156210899353, "learning_rate": 7.80781798442151e-05, "loss": 0.0632, "step": 34470 }, { "grad_norm": 0.7118066549301147, "learning_rate": 7.806449694469061e-05, "loss": 0.046, "step": 34480 }, { "grad_norm": 0.4350542724132538, "learning_rate": 7.805081097607492e-05, "loss": 0.0613, "step": 34490 }, { "grad_norm": 0.6186133027076721, "learning_rate": 7.803712193986474e-05, "loss": 0.0503, "step": 34500 }, { "grad_norm": 0.5505813360214233, "learning_rate": 7.802342983755702e-05, "loss": 0.0484, "step": 34510 }, { "grad_norm": 0.5985414981842041, "learning_rate": 7.800973467064918e-05, "loss": 0.0584, "step": 34520 }, { "grad_norm": 0.8905320763587952, "learning_rate": 7.799603644063884e-05, "loss": 0.0512, "step": 34530 }, { "grad_norm": 0.4029393792152405, "learning_rate": 7.798233514902405e-05, "loss": 0.0452, "step": 34540 }, { "grad_norm": 0.6236472725868225, "learning_rate": 7.796863079730318e-05, "loss": 0.068, "step": 34550 }, { "grad_norm": 0.5442001223564148, "learning_rate": 7.795492338697488e-05, "loss": 0.0584, "step": 34560 }, { "grad_norm": 0.5098721981048584, "learning_rate": 7.794121291953819e-05, "loss": 0.0679, "step": 34570 }, { "grad_norm": 0.8425030708312988, "learning_rate": 7.792749939649246e-05, "loss": 0.073, "step": 34580 }, { "grad_norm": 0.3993195593357086, "learning_rate": 7.79137828193374e-05, "loss": 0.0461, "step": 34590 }, { "grad_norm": 0.738804042339325, "learning_rate": 7.790006318957301e-05, "loss": 0.0672, "step": 34600 }, { "grad_norm": 0.45986270904541016, "learning_rate": 7.788634050869965e-05, "loss": 0.0509, "step": 34610 }, { "grad_norm": 0.6076272130012512, "learning_rate": 7.787261477821803e-05, "loss": 0.0607, "step": 34620 }, { "grad_norm": 0.6121752262115479, "learning_rate": 7.785888599962916e-05, "loss": 0.0486, "step": 34630 }, { "grad_norm": 0.5858614444732666, "learning_rate": 7.784515417443439e-05, "loss": 0.0567, "step": 34640 }, { "grad_norm": 0.5946411490440369, "learning_rate": 7.783141930413545e-05, "loss": 0.0558, "step": 34650 }, { "grad_norm": 0.5879522562026978, "learning_rate": 7.78176813902343e-05, "loss": 0.0604, "step": 34660 }, { "grad_norm": 0.4505295157432556, "learning_rate": 7.780394043423336e-05, "loss": 0.0478, "step": 34670 }, { "grad_norm": 0.5057253241539001, "learning_rate": 7.77901964376353e-05, "loss": 0.0665, "step": 34680 }, { "grad_norm": 0.5296807885169983, "learning_rate": 7.777644940194316e-05, "loss": 0.0557, "step": 34690 }, { "grad_norm": 0.5672580003738403, "learning_rate": 7.776269932866023e-05, "loss": 0.0618, "step": 34700 }, { "grad_norm": 0.43210548162460327, "learning_rate": 7.774894621929026e-05, "loss": 0.0574, "step": 34710 }, { "grad_norm": 0.6829585433006287, "learning_rate": 7.773519007533725e-05, "loss": 0.0695, "step": 34720 }, { "grad_norm": 0.37069302797317505, "learning_rate": 7.772143089830556e-05, "loss": 0.0499, "step": 34730 }, { "grad_norm": 0.41730213165283203, "learning_rate": 7.770766868969985e-05, "loss": 0.0508, "step": 34740 }, { "grad_norm": 0.40005528926849365, "learning_rate": 7.769390345102518e-05, "loss": 0.063, "step": 34750 }, { "grad_norm": 0.699494481086731, "learning_rate": 7.768013518378683e-05, "loss": 0.0688, "step": 34760 }, { "grad_norm": 0.45584481954574585, "learning_rate": 7.766636388949053e-05, "loss": 0.0417, "step": 34770 }, { "grad_norm": 0.48846060037612915, "learning_rate": 7.765258956964229e-05, "loss": 0.0741, "step": 34780 }, { "grad_norm": 0.4316442310810089, "learning_rate": 7.76388122257484e-05, "loss": 0.0596, "step": 34790 }, { "grad_norm": 0.401927649974823, "learning_rate": 7.762503185931558e-05, "loss": 0.0498, "step": 34800 }, { "grad_norm": 0.5692430138587952, "learning_rate": 7.76112484718508e-05, "loss": 0.0577, "step": 34810 }, { "grad_norm": 0.470138818025589, "learning_rate": 7.75974620648614e-05, "loss": 0.064, "step": 34820 }, { "grad_norm": 0.543760359287262, "learning_rate": 7.758367263985503e-05, "loss": 0.0505, "step": 34830 }, { "grad_norm": 0.40854600071907043, "learning_rate": 7.75698801983397e-05, "loss": 0.0476, "step": 34840 }, { "grad_norm": 0.3385355472564697, "learning_rate": 7.755608474182372e-05, "loss": 0.0613, "step": 34850 }, { "grad_norm": 0.6422481536865234, "learning_rate": 7.754228627181574e-05, "loss": 0.0647, "step": 34860 }, { "grad_norm": 0.42651328444480896, "learning_rate": 7.752848478982476e-05, "loss": 0.0534, "step": 34870 }, { "grad_norm": 0.4511811435222626, "learning_rate": 7.751468029736006e-05, "loss": 0.0536, "step": 34880 }, { "grad_norm": 0.4828243553638458, "learning_rate": 7.750087279593129e-05, "loss": 0.0628, "step": 34890 }, { "grad_norm": 0.448903888463974, "learning_rate": 7.748706228704843e-05, "loss": 0.0442, "step": 34900 }, { "grad_norm": 0.43390461802482605, "learning_rate": 7.747324877222176e-05, "loss": 0.0644, "step": 34910 }, { "grad_norm": 0.5969648361206055, "learning_rate": 7.745943225296188e-05, "loss": 0.0651, "step": 34920 }, { "grad_norm": 0.5390621423721313, "learning_rate": 7.744561273077981e-05, "loss": 0.0436, "step": 34930 }, { "grad_norm": 0.3036506175994873, "learning_rate": 7.743179020718678e-05, "loss": 0.0472, "step": 34940 }, { "grad_norm": 0.5111516714096069, "learning_rate": 7.741796468369443e-05, "loss": 0.0544, "step": 34950 }, { "grad_norm": 0.6178280711174011, "learning_rate": 7.740413616181466e-05, "loss": 0.0658, "step": 34960 }, { "grad_norm": 0.41311702132225037, "learning_rate": 7.739030464305978e-05, "loss": 0.0624, "step": 34970 }, { "grad_norm": 1.0132739543914795, "learning_rate": 7.737647012894235e-05, "loss": 0.072, "step": 34980 }, { "grad_norm": 0.6039823293685913, "learning_rate": 7.736263262097532e-05, "loss": 0.0598, "step": 34990 }, { "grad_norm": 0.4101405441761017, "learning_rate": 7.734879212067192e-05, "loss": 0.0613, "step": 35000 }, { "grad_norm": 0.5907109975814819, "learning_rate": 7.733494862954573e-05, "loss": 0.0547, "step": 35010 }, { "grad_norm": 0.45483556389808655, "learning_rate": 7.732110214911066e-05, "loss": 0.0461, "step": 35020 }, { "grad_norm": 0.7746742963790894, "learning_rate": 7.730725268088094e-05, "loss": 0.0764, "step": 35030 }, { "grad_norm": 0.4958050549030304, "learning_rate": 7.729340022637112e-05, "loss": 0.0599, "step": 35040 }, { "grad_norm": 0.5083584189414978, "learning_rate": 7.727954478709607e-05, "loss": 0.0594, "step": 35050 }, { "grad_norm": 0.6621259450912476, "learning_rate": 7.726568636457103e-05, "loss": 0.0644, "step": 35060 }, { "grad_norm": 0.6125110387802124, "learning_rate": 7.725182496031153e-05, "loss": 0.0615, "step": 35070 }, { "grad_norm": 0.5490039587020874, "learning_rate": 7.723796057583341e-05, "loss": 0.0685, "step": 35080 }, { "grad_norm": 0.6518492102622986, "learning_rate": 7.722409321265287e-05, "loss": 0.0704, "step": 35090 }, { "grad_norm": 0.467899352312088, "learning_rate": 7.721022287228645e-05, "loss": 0.0695, "step": 35100 }, { "grad_norm": 0.6701252460479736, "learning_rate": 7.719634955625094e-05, "loss": 0.048, "step": 35110 }, { "grad_norm": 0.9008123278617859, "learning_rate": 7.718247326606355e-05, "loss": 0.0526, "step": 35120 }, { "grad_norm": 0.5403781533241272, "learning_rate": 7.716859400324175e-05, "loss": 0.0577, "step": 35130 }, { "grad_norm": 0.7043761610984802, "learning_rate": 7.715471176930335e-05, "loss": 0.0792, "step": 35140 }, { "grad_norm": 0.4140250086784363, "learning_rate": 7.714082656576651e-05, "loss": 0.0659, "step": 35150 }, { "grad_norm": 0.5477021336555481, "learning_rate": 7.712693839414968e-05, "loss": 0.0559, "step": 35160 }, { "grad_norm": 0.32867512106895447, "learning_rate": 7.711304725597164e-05, "loss": 0.0433, "step": 35170 }, { "grad_norm": 0.46555477380752563, "learning_rate": 7.709915315275151e-05, "loss": 0.0712, "step": 35180 }, { "grad_norm": 0.46281081438064575, "learning_rate": 7.708525608600876e-05, "loss": 0.0537, "step": 35190 }, { "grad_norm": 0.455773264169693, "learning_rate": 7.707135605726311e-05, "loss": 0.0618, "step": 35200 }, { "grad_norm": 0.5285277366638184, "learning_rate": 7.705745306803466e-05, "loss": 0.0838, "step": 35210 }, { "grad_norm": 0.4708910882472992, "learning_rate": 7.704354711984383e-05, "loss": 0.0576, "step": 35220 }, { "grad_norm": 0.4342029392719269, "learning_rate": 7.702963821421133e-05, "loss": 0.0554, "step": 35230 }, { "grad_norm": 0.5504549741744995, "learning_rate": 7.701572635265826e-05, "loss": 0.0566, "step": 35240 }, { "grad_norm": 0.5363555550575256, "learning_rate": 7.700181153670596e-05, "loss": 0.0434, "step": 35250 }, { "grad_norm": 0.6302659511566162, "learning_rate": 7.698789376787616e-05, "loss": 0.0604, "step": 35260 }, { "grad_norm": 0.5731320381164551, "learning_rate": 7.697397304769085e-05, "loss": 0.0546, "step": 35270 }, { "grad_norm": 0.5831170082092285, "learning_rate": 7.696004937767241e-05, "loss": 0.0609, "step": 35280 }, { "grad_norm": 0.3839678466320038, "learning_rate": 7.694612275934352e-05, "loss": 0.0589, "step": 35290 }, { "grad_norm": 0.4432443380355835, "learning_rate": 7.693219319422714e-05, "loss": 0.0557, "step": 35300 }, { "grad_norm": 0.36945730447769165, "learning_rate": 7.69182606838466e-05, "loss": 0.0532, "step": 35310 }, { "grad_norm": 0.5058515667915344, "learning_rate": 7.690432522972558e-05, "loss": 0.0473, "step": 35320 }, { "grad_norm": 0.3705805540084839, "learning_rate": 7.689038683338796e-05, "loss": 0.0506, "step": 35330 }, { "grad_norm": 0.45046675205230713, "learning_rate": 7.687644549635808e-05, "loss": 0.0545, "step": 35340 }, { "grad_norm": 2.6937623023986816, "learning_rate": 7.686250122016053e-05, "loss": 0.0474, "step": 35350 }, { "grad_norm": 0.4986196756362915, "learning_rate": 7.684855400632022e-05, "loss": 0.0615, "step": 35360 }, { "grad_norm": 0.8020352125167847, "learning_rate": 7.683460385636243e-05, "loss": 0.044, "step": 35370 }, { "grad_norm": 0.45995691418647766, "learning_rate": 7.68206507718127e-05, "loss": 0.0592, "step": 35380 }, { "grad_norm": 0.5603251457214355, "learning_rate": 7.680669475419693e-05, "loss": 0.0676, "step": 35390 }, { "grad_norm": 0.6251128911972046, "learning_rate": 7.679273580504132e-05, "loss": 0.0442, "step": 35400 }, { "grad_norm": 0.6459055542945862, "learning_rate": 7.67787739258724e-05, "loss": 0.0636, "step": 35410 }, { "grad_norm": 0.7758822441101074, "learning_rate": 7.676480911821705e-05, "loss": 0.0646, "step": 35420 }, { "grad_norm": 0.5322344899177551, "learning_rate": 7.675084138360238e-05, "loss": 0.054, "step": 35430 }, { "grad_norm": 0.7338405847549438, "learning_rate": 7.673687072355592e-05, "loss": 0.0607, "step": 35440 }, { "grad_norm": 0.566721498966217, "learning_rate": 7.67228971396055e-05, "loss": 0.0505, "step": 35450 }, { "grad_norm": 0.6068481802940369, "learning_rate": 7.670892063327922e-05, "loss": 0.0587, "step": 35460 }, { "grad_norm": 0.7651936411857605, "learning_rate": 7.669494120610552e-05, "loss": 0.0509, "step": 35470 }, { "grad_norm": 0.3994849920272827, "learning_rate": 7.66809588596132e-05, "loss": 0.0432, "step": 35480 }, { "grad_norm": 0.3691488206386566, "learning_rate": 7.666697359533132e-05, "loss": 0.0614, "step": 35490 }, { "grad_norm": 0.8888694047927856, "learning_rate": 7.665298541478932e-05, "loss": 0.0649, "step": 35500 }, { "grad_norm": 0.370389461517334, "learning_rate": 7.66389943195169e-05, "loss": 0.0554, "step": 35510 }, { "grad_norm": 0.7792260646820068, "learning_rate": 7.66250003110441e-05, "loss": 0.0647, "step": 35520 }, { "grad_norm": 0.7031961679458618, "learning_rate": 7.66110033909013e-05, "loss": 0.0578, "step": 35530 }, { "grad_norm": 0.6399801969528198, "learning_rate": 7.659700356061918e-05, "loss": 0.0581, "step": 35540 }, { "grad_norm": 0.4989542067050934, "learning_rate": 7.658300082172875e-05, "loss": 0.0487, "step": 35550 }, { "grad_norm": 0.46626630425453186, "learning_rate": 7.65689951757613e-05, "loss": 0.0528, "step": 35560 }, { "grad_norm": 0.5413011908531189, "learning_rate": 7.655498662424849e-05, "loss": 0.0567, "step": 35570 }, { "grad_norm": 0.42916157841682434, "learning_rate": 7.654097516872227e-05, "loss": 0.0484, "step": 35580 }, { "grad_norm": 0.6198318600654602, "learning_rate": 7.65269608107149e-05, "loss": 0.0593, "step": 35590 }, { "grad_norm": 0.5975087285041809, "learning_rate": 7.6512943551759e-05, "loss": 0.0709, "step": 35600 }, { "grad_norm": 1.2434344291687012, "learning_rate": 7.649892339338743e-05, "loss": 0.0541, "step": 35610 }, { "grad_norm": 0.49795565009117126, "learning_rate": 7.648490033713344e-05, "loss": 0.0429, "step": 35620 }, { "grad_norm": 0.8866866827011108, "learning_rate": 7.647087438453058e-05, "loss": 0.0581, "step": 35630 }, { "grad_norm": 0.4903174638748169, "learning_rate": 7.64568455371127e-05, "loss": 0.0589, "step": 35640 }, { "grad_norm": 0.5415489077568054, "learning_rate": 7.644281379641396e-05, "loss": 0.0509, "step": 35650 }, { "grad_norm": 0.5119814872741699, "learning_rate": 7.642877916396887e-05, "loss": 0.0594, "step": 35660 }, { "grad_norm": 0.49675431847572327, "learning_rate": 7.641474164131221e-05, "loss": 0.0559, "step": 35670 }, { "grad_norm": 0.4919893443584442, "learning_rate": 7.640070122997913e-05, "loss": 0.0518, "step": 35680 }, { "grad_norm": 0.8496274948120117, "learning_rate": 7.638665793150505e-05, "loss": 0.0483, "step": 35690 }, { "grad_norm": 1.2773687839508057, "learning_rate": 7.637261174742574e-05, "loss": 0.0723, "step": 35700 }, { "grad_norm": 0.5436763763427734, "learning_rate": 7.635856267927725e-05, "loss": 0.0464, "step": 35710 }, { "grad_norm": 0.536876380443573, "learning_rate": 7.634451072859597e-05, "loss": 0.0502, "step": 35720 }, { "grad_norm": 0.6109923124313354, "learning_rate": 7.633045589691863e-05, "loss": 0.0435, "step": 35730 }, { "grad_norm": 0.4461679756641388, "learning_rate": 7.63163981857822e-05, "loss": 0.0495, "step": 35740 }, { "grad_norm": 0.3361296057701111, "learning_rate": 7.630233759672403e-05, "loss": 0.0456, "step": 35750 }, { "grad_norm": 0.5239328145980835, "learning_rate": 7.628827413128177e-05, "loss": 0.0645, "step": 35760 }, { "grad_norm": 0.4790849983692169, "learning_rate": 7.627420779099338e-05, "loss": 0.0581, "step": 35770 }, { "grad_norm": 1.0660384893417358, "learning_rate": 7.626013857739711e-05, "loss": 0.0573, "step": 35780 }, { "grad_norm": 0.6130247712135315, "learning_rate": 7.624606649203158e-05, "loss": 0.0534, "step": 35790 }, { "grad_norm": 0.38959866762161255, "learning_rate": 7.623199153643569e-05, "loss": 0.0449, "step": 35800 }, { "grad_norm": 0.5372145771980286, "learning_rate": 7.621791371214863e-05, "loss": 0.0551, "step": 35810 }, { "grad_norm": 0.37590694427490234, "learning_rate": 7.620383302070995e-05, "loss": 0.048, "step": 35820 }, { "grad_norm": 0.47428011894226074, "learning_rate": 7.61897494636595e-05, "loss": 0.0441, "step": 35830 }, { "grad_norm": 0.8022124171257019, "learning_rate": 7.617566304253739e-05, "loss": 0.0548, "step": 35840 }, { "grad_norm": 0.5296559929847717, "learning_rate": 7.616157375888416e-05, "loss": 0.0492, "step": 35850 }, { "grad_norm": 0.4261783957481384, "learning_rate": 7.614748161424053e-05, "loss": 0.0443, "step": 35860 }, { "grad_norm": 0.5528519749641418, "learning_rate": 7.613338661014763e-05, "loss": 0.048, "step": 35870 }, { "grad_norm": 0.735512375831604, "learning_rate": 7.611928874814686e-05, "loss": 0.056, "step": 35880 }, { "grad_norm": 0.49674269556999207, "learning_rate": 7.610518802977996e-05, "loss": 0.0455, "step": 35890 }, { "grad_norm": 0.5099405646324158, "learning_rate": 7.609108445658893e-05, "loss": 0.0461, "step": 35900 }, { "grad_norm": 0.4949479103088379, "learning_rate": 7.607697803011612e-05, "loss": 0.057, "step": 35910 }, { "grad_norm": 0.6316192150115967, "learning_rate": 7.606286875190421e-05, "loss": 0.0589, "step": 35920 }, { "grad_norm": 0.4867188334465027, "learning_rate": 7.604875662349617e-05, "loss": 0.0428, "step": 35930 }, { "grad_norm": 0.44584494829177856, "learning_rate": 7.603464164643525e-05, "loss": 0.0589, "step": 35940 }, { "grad_norm": 0.717316210269928, "learning_rate": 7.602052382226507e-05, "loss": 0.0608, "step": 35950 }, { "grad_norm": 0.7266101837158203, "learning_rate": 7.600640315252954e-05, "loss": 0.0562, "step": 35960 }, { "grad_norm": 0.47883808612823486, "learning_rate": 7.599227963877284e-05, "loss": 0.0646, "step": 35970 }, { "grad_norm": 0.661903977394104, "learning_rate": 7.597815328253954e-05, "loss": 0.078, "step": 35980 }, { "grad_norm": 0.5521270036697388, "learning_rate": 7.596402408537443e-05, "loss": 0.0563, "step": 35990 }, { "grad_norm": 0.5462560057640076, "learning_rate": 7.59498920488227e-05, "loss": 0.0598, "step": 36000 }, { "grad_norm": 0.9379068613052368, "learning_rate": 7.593575717442979e-05, "loss": 0.0578, "step": 36010 }, { "grad_norm": 0.5129972696304321, "learning_rate": 7.592161946374147e-05, "loss": 0.0623, "step": 36020 }, { "grad_norm": 0.31057772040367126, "learning_rate": 7.590747891830381e-05, "loss": 0.0479, "step": 36030 }, { "grad_norm": 0.7496331930160522, "learning_rate": 7.58933355396632e-05, "loss": 0.0518, "step": 36040 }, { "grad_norm": 0.5976462364196777, "learning_rate": 7.587918932936636e-05, "loss": 0.053, "step": 36050 }, { "grad_norm": 0.40590640902519226, "learning_rate": 7.586504028896028e-05, "loss": 0.0585, "step": 36060 }, { "grad_norm": 0.5547118186950684, "learning_rate": 7.585088841999228e-05, "loss": 0.0469, "step": 36070 }, { "grad_norm": 0.42669543623924255, "learning_rate": 7.583673372400999e-05, "loss": 0.0528, "step": 36080 }, { "grad_norm": 0.7351099848747253, "learning_rate": 7.582257620256134e-05, "loss": 0.0502, "step": 36090 }, { "grad_norm": 0.5372759699821472, "learning_rate": 7.580841585719458e-05, "loss": 0.0565, "step": 36100 }, { "grad_norm": 0.774197518825531, "learning_rate": 7.579425268945825e-05, "loss": 0.0661, "step": 36110 }, { "grad_norm": 0.4456097185611725, "learning_rate": 7.578008670090127e-05, "loss": 0.0485, "step": 36120 }, { "grad_norm": 0.39518654346466064, "learning_rate": 7.576591789307272e-05, "loss": 0.0483, "step": 36130 }, { "grad_norm": 0.49748694896698, "learning_rate": 7.575174626752216e-05, "loss": 0.0385, "step": 36140 }, { "grad_norm": 0.8564813137054443, "learning_rate": 7.573757182579934e-05, "loss": 0.0535, "step": 36150 }, { "grad_norm": 0.5675255060195923, "learning_rate": 7.572339456945435e-05, "loss": 0.0517, "step": 36160 }, { "grad_norm": 0.6661283373832703, "learning_rate": 7.570921450003762e-05, "loss": 0.0629, "step": 36170 }, { "grad_norm": 0.41581985354423523, "learning_rate": 7.569503161909983e-05, "loss": 0.0523, "step": 36180 }, { "grad_norm": 0.6794431209564209, "learning_rate": 7.568084592819202e-05, "loss": 0.0575, "step": 36190 }, { "grad_norm": 0.5497203469276428, "learning_rate": 7.566665742886551e-05, "loss": 0.0506, "step": 36200 }, { "grad_norm": 0.9825537800788879, "learning_rate": 7.565246612267194e-05, "loss": 0.0553, "step": 36210 }, { "grad_norm": 0.7417182326316833, "learning_rate": 7.563827201116325e-05, "loss": 0.0661, "step": 36220 }, { "grad_norm": 0.6531925797462463, "learning_rate": 7.562407509589164e-05, "loss": 0.0585, "step": 36230 }, { "grad_norm": 0.494301974773407, "learning_rate": 7.560987537840973e-05, "loss": 0.0622, "step": 36240 }, { "grad_norm": 0.5665419697761536, "learning_rate": 7.559567286027036e-05, "loss": 0.0479, "step": 36250 }, { "grad_norm": 0.935203492641449, "learning_rate": 7.558146754302668e-05, "loss": 0.0654, "step": 36260 }, { "grad_norm": 0.7438482642173767, "learning_rate": 7.556725942823217e-05, "loss": 0.0495, "step": 36270 }, { "grad_norm": 1.0044370889663696, "learning_rate": 7.555304851744061e-05, "loss": 0.0655, "step": 36280 }, { "grad_norm": 0.6033666133880615, "learning_rate": 7.553883481220608e-05, "loss": 0.0711, "step": 36290 }, { "grad_norm": 0.3934142291545868, "learning_rate": 7.552461831408298e-05, "loss": 0.0444, "step": 36300 }, { "grad_norm": 0.5041322112083435, "learning_rate": 7.551039902462599e-05, "loss": 0.0539, "step": 36310 }, { "grad_norm": 0.34157001972198486, "learning_rate": 7.549617694539011e-05, "loss": 0.0338, "step": 36320 }, { "grad_norm": 0.5552269220352173, "learning_rate": 7.548195207793066e-05, "loss": 0.0522, "step": 36330 }, { "grad_norm": 0.45712757110595703, "learning_rate": 7.546772442380323e-05, "loss": 0.0483, "step": 36340 }, { "grad_norm": 0.3930697739124298, "learning_rate": 7.545349398456376e-05, "loss": 0.0671, "step": 36350 }, { "grad_norm": 0.45418477058410645, "learning_rate": 7.543926076176845e-05, "loss": 0.061, "step": 36360 }, { "grad_norm": 0.3792192339897156, "learning_rate": 7.542502475697385e-05, "loss": 0.0506, "step": 36370 }, { "grad_norm": 0.9415712952613831, "learning_rate": 7.541078597173675e-05, "loss": 0.0619, "step": 36380 }, { "grad_norm": 0.5050264000892639, "learning_rate": 7.539654440761431e-05, "loss": 0.0664, "step": 36390 }, { "grad_norm": 0.5016268491744995, "learning_rate": 7.538230006616395e-05, "loss": 0.0487, "step": 36400 }, { "grad_norm": 0.761355996131897, "learning_rate": 7.536805294894342e-05, "loss": 0.0687, "step": 36410 }, { "grad_norm": 0.6255768537521362, "learning_rate": 7.535380305751076e-05, "loss": 0.0533, "step": 36420 }, { "grad_norm": 0.48565274477005005, "learning_rate": 7.533955039342431e-05, "loss": 0.0619, "step": 36430 }, { "grad_norm": 0.7919872403144836, "learning_rate": 7.532529495824274e-05, "loss": 0.0606, "step": 36440 }, { "grad_norm": 0.5473651885986328, "learning_rate": 7.531103675352497e-05, "loss": 0.0519, "step": 36450 }, { "grad_norm": 0.7224080562591553, "learning_rate": 7.529677578083029e-05, "loss": 0.0585, "step": 36460 }, { "grad_norm": 0.7680691480636597, "learning_rate": 7.528251204171823e-05, "loss": 0.0584, "step": 36470 }, { "grad_norm": 0.5202397704124451, "learning_rate": 7.526824553774867e-05, "loss": 0.058, "step": 36480 }, { "grad_norm": 0.7257019281387329, "learning_rate": 7.525397627048178e-05, "loss": 0.0567, "step": 36490 }, { "grad_norm": 0.8211466073989868, "learning_rate": 7.523970424147802e-05, "loss": 0.0509, "step": 36500 }, { "grad_norm": 0.6562297344207764, "learning_rate": 7.522542945229813e-05, "loss": 0.0457, "step": 36510 }, { "grad_norm": 1.155990719795227, "learning_rate": 7.521115190450321e-05, "loss": 0.0559, "step": 36520 }, { "grad_norm": 0.5971108078956604, "learning_rate": 7.519687159965462e-05, "loss": 0.0512, "step": 36530 }, { "grad_norm": 0.6099902391433716, "learning_rate": 7.518258853931403e-05, "loss": 0.0543, "step": 36540 }, { "grad_norm": 0.40527471899986267, "learning_rate": 7.516830272504342e-05, "loss": 0.0447, "step": 36550 }, { "grad_norm": 0.562796950340271, "learning_rate": 7.515401415840509e-05, "loss": 0.07, "step": 36560 }, { "grad_norm": 0.6786022186279297, "learning_rate": 7.513972284096155e-05, "loss": 0.0495, "step": 36570 }, { "grad_norm": 0.3993529975414276, "learning_rate": 7.512542877427574e-05, "loss": 0.0544, "step": 36580 }, { "grad_norm": 0.4420737326145172, "learning_rate": 7.51111319599108e-05, "loss": 0.0531, "step": 36590 }, { "grad_norm": 0.5276815295219421, "learning_rate": 7.509683239943024e-05, "loss": 0.0602, "step": 36600 }, { "grad_norm": 0.511654794216156, "learning_rate": 7.50825300943978e-05, "loss": 0.0674, "step": 36610 }, { "grad_norm": 0.5927472710609436, "learning_rate": 7.50682250463776e-05, "loss": 0.0564, "step": 36620 }, { "grad_norm": 0.4597487449645996, "learning_rate": 7.5053917256934e-05, "loss": 0.0478, "step": 36630 }, { "grad_norm": 0.6259806156158447, "learning_rate": 7.503960672763166e-05, "loss": 0.0601, "step": 36640 }, { "grad_norm": 0.4202451705932617, "learning_rate": 7.502529346003559e-05, "loss": 0.059, "step": 36650 }, { "grad_norm": 0.5208409428596497, "learning_rate": 7.501097745571106e-05, "loss": 0.0587, "step": 36660 }, { "grad_norm": 0.39982566237449646, "learning_rate": 7.499665871622361e-05, "loss": 0.0704, "step": 36670 }, { "grad_norm": 0.6179673671722412, "learning_rate": 7.498233724313917e-05, "loss": 0.0718, "step": 36680 }, { "grad_norm": 0.4552319049835205, "learning_rate": 7.496801303802387e-05, "loss": 0.0693, "step": 36690 }, { "grad_norm": 0.5022062659263611, "learning_rate": 7.495368610244423e-05, "loss": 0.0512, "step": 36700 }, { "grad_norm": 0.7962685823440552, "learning_rate": 7.493935643796697e-05, "loss": 0.0681, "step": 36710 }, { "grad_norm": 0.5108117461204529, "learning_rate": 7.49250240461592e-05, "loss": 0.0562, "step": 36720 }, { "grad_norm": 0.45669791102409363, "learning_rate": 7.491068892858826e-05, "loss": 0.0654, "step": 36730 }, { "grad_norm": 0.4664015471935272, "learning_rate": 7.489635108682184e-05, "loss": 0.0563, "step": 36740 }, { "grad_norm": 0.565549373626709, "learning_rate": 7.48820105224279e-05, "loss": 0.0497, "step": 36750 }, { "grad_norm": 0.8146399855613708, "learning_rate": 7.48676672369747e-05, "loss": 0.0649, "step": 36760 }, { "grad_norm": 0.6357666850090027, "learning_rate": 7.485332123203079e-05, "loss": 0.0658, "step": 36770 }, { "grad_norm": 0.4702226221561432, "learning_rate": 7.483897250916504e-05, "loss": 0.063, "step": 36780 }, { "grad_norm": 1.4708833694458008, "learning_rate": 7.48246210699466e-05, "loss": 0.0601, "step": 36790 }, { "grad_norm": 0.7063269019126892, "learning_rate": 7.481026691594492e-05, "loss": 0.0523, "step": 36800 }, { "grad_norm": 0.342427134513855, "learning_rate": 7.479591004872975e-05, "loss": 0.0457, "step": 36810 }, { "grad_norm": 0.3479376435279846, "learning_rate": 7.478155046987114e-05, "loss": 0.0567, "step": 36820 }, { "grad_norm": 0.5700193643569946, "learning_rate": 7.476718818093944e-05, "loss": 0.0647, "step": 36830 }, { "grad_norm": 0.5744131207466125, "learning_rate": 7.475282318350527e-05, "loss": 0.0544, "step": 36840 }, { "grad_norm": 0.4460200071334839, "learning_rate": 7.473845547913957e-05, "loss": 0.0455, "step": 36850 }, { "grad_norm": 0.6247230172157288, "learning_rate": 7.472408506941357e-05, "loss": 0.0757, "step": 36860 }, { "grad_norm": 0.3910330832004547, "learning_rate": 7.470971195589881e-05, "loss": 0.0532, "step": 36870 }, { "grad_norm": 0.9853628873825073, "learning_rate": 7.46953361401671e-05, "loss": 0.0729, "step": 36880 }, { "grad_norm": 0.8115820288658142, "learning_rate": 7.468095762379055e-05, "loss": 0.0477, "step": 36890 }, { "grad_norm": 0.4576278030872345, "learning_rate": 7.466657640834158e-05, "loss": 0.0629, "step": 36900 }, { "grad_norm": 0.7147277593612671, "learning_rate": 7.46521924953929e-05, "loss": 0.0558, "step": 36910 }, { "grad_norm": 0.46918249130249023, "learning_rate": 7.463780588651752e-05, "loss": 0.0536, "step": 36920 }, { "grad_norm": 0.6222795844078064, "learning_rate": 7.462341658328873e-05, "loss": 0.059, "step": 36930 }, { "grad_norm": 0.4359229505062103, "learning_rate": 7.460902458728012e-05, "loss": 0.053, "step": 36940 }, { "grad_norm": 0.5814335346221924, "learning_rate": 7.459462990006558e-05, "loss": 0.0499, "step": 36950 }, { "grad_norm": 0.5629039406776428, "learning_rate": 7.458023252321929e-05, "loss": 0.0483, "step": 36960 }, { "grad_norm": 0.767063319683075, "learning_rate": 7.456583245831573e-05, "loss": 0.0457, "step": 36970 }, { "grad_norm": 0.568842887878418, "learning_rate": 7.455142970692968e-05, "loss": 0.0728, "step": 36980 }, { "grad_norm": 0.5678025484085083, "learning_rate": 7.453702427063615e-05, "loss": 0.0475, "step": 36990 }, { "grad_norm": 0.7072295546531677, "learning_rate": 7.452261615101057e-05, "loss": 0.0498, "step": 37000 }, { "grad_norm": 1.0280007123947144, "learning_rate": 7.450820534962856e-05, "loss": 0.0719, "step": 37010 }, { "grad_norm": 0.45358335971832275, "learning_rate": 7.449379186806604e-05, "loss": 0.0513, "step": 37020 }, { "grad_norm": 0.8264983296394348, "learning_rate": 7.44793757078993e-05, "loss": 0.0626, "step": 37030 }, { "grad_norm": 0.9390788078308105, "learning_rate": 7.44649568707048e-05, "loss": 0.0526, "step": 37040 }, { "grad_norm": 0.6138522624969482, "learning_rate": 7.445053535805942e-05, "loss": 0.0619, "step": 37050 }, { "grad_norm": 1.185752034187317, "learning_rate": 7.443611117154024e-05, "loss": 0.0523, "step": 37060 }, { "grad_norm": 1.954039216041565, "learning_rate": 7.442168431272469e-05, "loss": 0.055, "step": 37070 }, { "grad_norm": 0.5201994776725769, "learning_rate": 7.440725478319044e-05, "loss": 0.0606, "step": 37080 }, { "grad_norm": 0.5548028945922852, "learning_rate": 7.439282258451553e-05, "loss": 0.0611, "step": 37090 }, { "grad_norm": 0.5342440605163574, "learning_rate": 7.43783877182782e-05, "loss": 0.0614, "step": 37100 }, { "grad_norm": 0.6147887706756592, "learning_rate": 7.436395018605707e-05, "loss": 0.0476, "step": 37110 }, { "grad_norm": 0.49798789620399475, "learning_rate": 7.434950998943094e-05, "loss": 0.049, "step": 37120 }, { "grad_norm": 0.5172694325447083, "learning_rate": 7.433506712997902e-05, "loss": 0.0571, "step": 37130 }, { "grad_norm": 0.4553876221179962, "learning_rate": 7.432062160928077e-05, "loss": 0.0542, "step": 37140 }, { "grad_norm": 0.3487605154514313, "learning_rate": 7.430617342891588e-05, "loss": 0.052, "step": 37150 }, { "grad_norm": 0.34945547580718994, "learning_rate": 7.429172259046442e-05, "loss": 0.0515, "step": 37160 }, { "grad_norm": 0.6215502619743347, "learning_rate": 7.42772690955067e-05, "loss": 0.0522, "step": 37170 }, { "grad_norm": 0.49039706587791443, "learning_rate": 7.426281294562332e-05, "loss": 0.0698, "step": 37180 }, { "grad_norm": 0.4491634666919708, "learning_rate": 7.424835414239519e-05, "loss": 0.0602, "step": 37190 }, { "grad_norm": 0.37230223417282104, "learning_rate": 7.423389268740352e-05, "loss": 0.0517, "step": 37200 }, { "grad_norm": 0.7218837738037109, "learning_rate": 7.42194285822298e-05, "loss": 0.0469, "step": 37210 }, { "grad_norm": 0.44182536005973816, "learning_rate": 7.420496182845576e-05, "loss": 0.0589, "step": 37220 }, { "grad_norm": 0.464758962392807, "learning_rate": 7.419049242766348e-05, "loss": 0.0661, "step": 37230 }, { "grad_norm": 0.7804550528526306, "learning_rate": 7.417602038143533e-05, "loss": 0.0661, "step": 37240 }, { "grad_norm": 1.0003718137741089, "learning_rate": 7.416154569135393e-05, "loss": 0.0708, "step": 37250 }, { "grad_norm": 0.6051241159439087, "learning_rate": 7.414706835900222e-05, "loss": 0.0506, "step": 37260 }, { "grad_norm": 0.7959907054901123, "learning_rate": 7.413258838596345e-05, "loss": 0.0597, "step": 37270 }, { "grad_norm": 0.45591744780540466, "learning_rate": 7.411810577382106e-05, "loss": 0.0679, "step": 37280 }, { "grad_norm": 0.7258987426757812, "learning_rate": 7.410362052415889e-05, "loss": 0.0643, "step": 37290 }, { "grad_norm": 0.6863278150558472, "learning_rate": 7.408913263856102e-05, "loss": 0.0633, "step": 37300 }, { "grad_norm": 0.9103356599807739, "learning_rate": 7.407464211861183e-05, "loss": 0.0686, "step": 37310 }, { "grad_norm": 0.529535710811615, "learning_rate": 7.406014896589597e-05, "loss": 0.0573, "step": 37320 }, { "grad_norm": 0.5165889263153076, "learning_rate": 7.404565318199838e-05, "loss": 0.0522, "step": 37330 }, { "grad_norm": 0.46052029728889465, "learning_rate": 7.403115476850432e-05, "loss": 0.0701, "step": 37340 }, { "grad_norm": 0.5093674659729004, "learning_rate": 7.401665372699932e-05, "loss": 0.0605, "step": 37350 }, { "grad_norm": 0.5227267146110535, "learning_rate": 7.400215005906916e-05, "loss": 0.0652, "step": 37360 }, { "grad_norm": 0.7489084601402283, "learning_rate": 7.398764376629996e-05, "loss": 0.0608, "step": 37370 }, { "grad_norm": 0.43915122747421265, "learning_rate": 7.39731348502781e-05, "loss": 0.072, "step": 37380 }, { "grad_norm": 0.5653843879699707, "learning_rate": 7.395862331259028e-05, "loss": 0.0551, "step": 37390 }, { "grad_norm": 0.772832453250885, "learning_rate": 7.394410915482341e-05, "loss": 0.0606, "step": 37400 }, { "grad_norm": 0.4364195466041565, "learning_rate": 7.392959237856476e-05, "loss": 0.0551, "step": 37410 }, { "grad_norm": 0.7278015613555908, "learning_rate": 7.391507298540187e-05, "loss": 0.0532, "step": 37420 }, { "grad_norm": 0.5867460370063782, "learning_rate": 7.390055097692257e-05, "loss": 0.0481, "step": 37430 }, { "grad_norm": 0.5720579028129578, "learning_rate": 7.388602635471494e-05, "loss": 0.0607, "step": 37440 }, { "grad_norm": 0.6740710735321045, "learning_rate": 7.387149912036738e-05, "loss": 0.0642, "step": 37450 }, { "grad_norm": 0.6516276001930237, "learning_rate": 7.385696927546854e-05, "loss": 0.054, "step": 37460 }, { "grad_norm": 0.5495119690895081, "learning_rate": 7.384243682160743e-05, "loss": 0.0804, "step": 37470 }, { "grad_norm": 0.6068182587623596, "learning_rate": 7.382790176037329e-05, "loss": 0.0595, "step": 37480 }, { "grad_norm": 0.5634384751319885, "learning_rate": 7.381336409335562e-05, "loss": 0.0627, "step": 37490 }, { "grad_norm": 0.858460009098053, "learning_rate": 7.379882382214426e-05, "loss": 0.0602, "step": 37500 }, { "grad_norm": 0.35465750098228455, "learning_rate": 7.378428094832931e-05, "loss": 0.056, "step": 37510 }, { "grad_norm": 0.6430433988571167, "learning_rate": 7.376973547350114e-05, "loss": 0.0592, "step": 37520 }, { "grad_norm": 0.47950759530067444, "learning_rate": 7.375518739925046e-05, "loss": 0.056, "step": 37530 }, { "grad_norm": 0.5634523630142212, "learning_rate": 7.374063672716818e-05, "loss": 0.0468, "step": 37540 }, { "grad_norm": 0.7708344459533691, "learning_rate": 7.372608345884558e-05, "loss": 0.0625, "step": 37550 }, { "grad_norm": 0.8091192841529846, "learning_rate": 7.371152759587415e-05, "loss": 0.063, "step": 37560 }, { "grad_norm": 0.5651447176933289, "learning_rate": 7.369696913984572e-05, "loss": 0.0439, "step": 37570 }, { "grad_norm": 0.41879764199256897, "learning_rate": 7.368240809235237e-05, "loss": 0.059, "step": 37580 }, { "grad_norm": 0.40865448117256165, "learning_rate": 7.366784445498649e-05, "loss": 0.0605, "step": 37590 }, { "grad_norm": 0.3640391528606415, "learning_rate": 7.36532782293407e-05, "loss": 0.0504, "step": 37600 }, { "grad_norm": 0.5021989345550537, "learning_rate": 7.363870941700797e-05, "loss": 0.0568, "step": 37610 }, { "grad_norm": 0.4989710748195648, "learning_rate": 7.362413801958152e-05, "loss": 0.0491, "step": 37620 }, { "grad_norm": 0.43416815996170044, "learning_rate": 7.360956403865486e-05, "loss": 0.0557, "step": 37630 }, { "grad_norm": 0.48118263483047485, "learning_rate": 7.359498747582177e-05, "loss": 0.0565, "step": 37640 }, { "grad_norm": 0.5649029016494751, "learning_rate": 7.358040833267634e-05, "loss": 0.0524, "step": 37650 }, { "grad_norm": 0.503991961479187, "learning_rate": 7.356582661081288e-05, "loss": 0.0572, "step": 37660 }, { "grad_norm": 0.5738114714622498, "learning_rate": 7.355124231182607e-05, "loss": 0.0399, "step": 37670 }, { "grad_norm": 0.562954306602478, "learning_rate": 7.353665543731083e-05, "loss": 0.0489, "step": 37680 }, { "grad_norm": 0.4303019642829895, "learning_rate": 7.352206598886229e-05, "loss": 0.049, "step": 37690 }, { "grad_norm": 0.3927629590034485, "learning_rate": 7.350747396807601e-05, "loss": 0.0516, "step": 37700 }, { "grad_norm": 0.557200014591217, "learning_rate": 7.349287937654772e-05, "loss": 0.0574, "step": 37710 }, { "grad_norm": 0.38092389702796936, "learning_rate": 7.347828221587345e-05, "loss": 0.0592, "step": 37720 }, { "grad_norm": 0.4953969717025757, "learning_rate": 7.346368248764955e-05, "loss": 0.05, "step": 37730 }, { "grad_norm": 0.4324978291988373, "learning_rate": 7.34490801934726e-05, "loss": 0.0553, "step": 37740 }, { "grad_norm": 0.5097509026527405, "learning_rate": 7.343447533493947e-05, "loss": 0.0672, "step": 37750 }, { "grad_norm": 0.36957335472106934, "learning_rate": 7.341986791364739e-05, "loss": 0.0717, "step": 37760 }, { "grad_norm": 0.42693278193473816, "learning_rate": 7.340525793119374e-05, "loss": 0.048, "step": 37770 }, { "grad_norm": 0.5178359746932983, "learning_rate": 7.33906453891763e-05, "loss": 0.055, "step": 37780 }, { "grad_norm": 0.8029922246932983, "learning_rate": 7.337603028919303e-05, "loss": 0.0553, "step": 37790 }, { "grad_norm": 0.5406728982925415, "learning_rate": 7.336141263284225e-05, "loss": 0.0608, "step": 37800 }, { "grad_norm": 1.3339546918869019, "learning_rate": 7.334679242172249e-05, "loss": 0.062, "step": 37810 }, { "grad_norm": 0.48740988969802856, "learning_rate": 7.33321696574326e-05, "loss": 0.0525, "step": 37820 }, { "grad_norm": 0.4188334047794342, "learning_rate": 7.331754434157174e-05, "loss": 0.0498, "step": 37830 }, { "grad_norm": 0.5219718813896179, "learning_rate": 7.330291647573928e-05, "loss": 0.0595, "step": 37840 }, { "grad_norm": 0.5766226649284363, "learning_rate": 7.32882860615349e-05, "loss": 0.0429, "step": 37850 }, { "grad_norm": 0.45186951756477356, "learning_rate": 7.327365310055858e-05, "loss": 0.0463, "step": 37860 }, { "grad_norm": 0.4514310359954834, "learning_rate": 7.325901759441055e-05, "loss": 0.0548, "step": 37870 }, { "grad_norm": 0.5470743179321289, "learning_rate": 7.324437954469132e-05, "loss": 0.0448, "step": 37880 }, { "grad_norm": 0.8433236479759216, "learning_rate": 7.32297389530017e-05, "loss": 0.0503, "step": 37890 }, { "grad_norm": 0.3768702745437622, "learning_rate": 7.321509582094274e-05, "loss": 0.0513, "step": 37900 }, { "grad_norm": 0.41546452045440674, "learning_rate": 7.320045015011582e-05, "loss": 0.0609, "step": 37910 }, { "grad_norm": 0.3336855173110962, "learning_rate": 7.318580194212253e-05, "loss": 0.0506, "step": 37920 }, { "grad_norm": 0.4021228849887848, "learning_rate": 7.317115119856483e-05, "loss": 0.0553, "step": 37930 }, { "grad_norm": 1.2821086645126343, "learning_rate": 7.315649792104485e-05, "loss": 0.0526, "step": 37940 }, { "grad_norm": 0.607326865196228, "learning_rate": 7.314184211116507e-05, "loss": 0.0619, "step": 37950 }, { "grad_norm": 0.7267943024635315, "learning_rate": 7.312718377052823e-05, "loss": 0.0602, "step": 37960 }, { "grad_norm": 0.8149744272232056, "learning_rate": 7.311252290073737e-05, "loss": 0.0557, "step": 37970 }, { "grad_norm": 0.4637778699398041, "learning_rate": 7.309785950339572e-05, "loss": 0.059, "step": 37980 }, { "grad_norm": 0.6047093868255615, "learning_rate": 7.308319358010692e-05, "loss": 0.0483, "step": 37990 }, { "grad_norm": 0.46598756313323975, "learning_rate": 7.306852513247474e-05, "loss": 0.0791, "step": 38000 }, { "grad_norm": 0.7212250232696533, "learning_rate": 7.305385416210336e-05, "loss": 0.0691, "step": 38010 }, { "grad_norm": 0.47623443603515625, "learning_rate": 7.303918067059711e-05, "loss": 0.0388, "step": 38020 }, { "grad_norm": 0.3453197777271271, "learning_rate": 7.302450465956075e-05, "loss": 0.0505, "step": 38030 }, { "grad_norm": 0.44859111309051514, "learning_rate": 7.300982613059914e-05, "loss": 0.0475, "step": 38040 }, { "grad_norm": 0.9661349058151245, "learning_rate": 7.299514508531757e-05, "loss": 0.0643, "step": 38050 }, { "grad_norm": 0.479918509721756, "learning_rate": 7.298046152532148e-05, "loss": 0.0541, "step": 38060 }, { "grad_norm": 0.5733094215393066, "learning_rate": 7.296577545221668e-05, "loss": 0.0576, "step": 38070 }, { "grad_norm": 0.5274555683135986, "learning_rate": 7.29510868676092e-05, "loss": 0.0584, "step": 38080 }, { "grad_norm": 0.526019275188446, "learning_rate": 7.293639577310538e-05, "loss": 0.0536, "step": 38090 }, { "grad_norm": 0.5410826206207275, "learning_rate": 7.292170217031179e-05, "loss": 0.0624, "step": 38100 }, { "grad_norm": 0.6016628742218018, "learning_rate": 7.290700606083532e-05, "loss": 0.0612, "step": 38110 }, { "grad_norm": 0.5204123854637146, "learning_rate": 7.28923074462831e-05, "loss": 0.0552, "step": 38120 }, { "grad_norm": 0.6410219073295593, "learning_rate": 7.287760632826255e-05, "loss": 0.0504, "step": 38130 }, { "grad_norm": 0.5290796160697937, "learning_rate": 7.286290270838139e-05, "loss": 0.0597, "step": 38140 }, { "grad_norm": 0.8492798805236816, "learning_rate": 7.284819658824756e-05, "loss": 0.0487, "step": 38150 }, { "grad_norm": 0.5215325355529785, "learning_rate": 7.283348796946929e-05, "loss": 0.057, "step": 38160 }, { "grad_norm": 0.7134202718734741, "learning_rate": 7.28187768536551e-05, "loss": 0.0538, "step": 38170 }, { "grad_norm": 0.6950613856315613, "learning_rate": 7.280406324241379e-05, "loss": 0.0615, "step": 38180 }, { "grad_norm": 0.45260027050971985, "learning_rate": 7.278934713735441e-05, "loss": 0.0586, "step": 38190 }, { "grad_norm": 0.7437261939048767, "learning_rate": 7.277462854008629e-05, "loss": 0.0548, "step": 38200 }, { "grad_norm": 0.7191012501716614, "learning_rate": 7.275990745221904e-05, "loss": 0.0508, "step": 38210 }, { "grad_norm": 0.41587328910827637, "learning_rate": 7.274518387536252e-05, "loss": 0.0507, "step": 38220 }, { "grad_norm": 0.6006278395652771, "learning_rate": 7.273045781112689e-05, "loss": 0.0615, "step": 38230 }, { "grad_norm": 0.35142430663108826, "learning_rate": 7.271572926112257e-05, "loss": 0.063, "step": 38240 }, { "grad_norm": 0.5381499528884888, "learning_rate": 7.270099822696024e-05, "loss": 0.0531, "step": 38250 }, { "grad_norm": 0.31894487142562866, "learning_rate": 7.26862647102509e-05, "loss": 0.0507, "step": 38260 }, { "grad_norm": 0.34442195296287537, "learning_rate": 7.267152871260573e-05, "loss": 0.0493, "step": 38270 }, { "grad_norm": 0.5774183869361877, "learning_rate": 7.265679023563629e-05, "loss": 0.0688, "step": 38280 }, { "grad_norm": 0.9082382917404175, "learning_rate": 7.264204928095433e-05, "loss": 0.0661, "step": 38290 }, { "grad_norm": 0.5625177621841431, "learning_rate": 7.262730585017188e-05, "loss": 0.0537, "step": 38300 }, { "grad_norm": 0.3463577628135681, "learning_rate": 7.261255994490131e-05, "loss": 0.059, "step": 38310 }, { "grad_norm": 0.5317549705505371, "learning_rate": 7.259781156675517e-05, "loss": 0.0595, "step": 38320 }, { "grad_norm": 0.8695464134216309, "learning_rate": 7.258306071734633e-05, "loss": 0.0521, "step": 38330 }, { "grad_norm": 0.5572136044502258, "learning_rate": 7.256830739828792e-05, "loss": 0.06, "step": 38340 }, { "grad_norm": 0.4805430769920349, "learning_rate": 7.255355161119336e-05, "loss": 0.0463, "step": 38350 }, { "grad_norm": 1.613109827041626, "learning_rate": 7.253879335767628e-05, "loss": 0.0586, "step": 38360 }, { "grad_norm": 0.7204080820083618, "learning_rate": 7.252403263935067e-05, "loss": 0.0727, "step": 38370 }, { "grad_norm": 0.49600985646247864, "learning_rate": 7.25092694578307e-05, "loss": 0.068, "step": 38380 }, { "grad_norm": 0.6922250986099243, "learning_rate": 7.249450381473087e-05, "loss": 0.0667, "step": 38390 }, { "grad_norm": 0.7595468759536743, "learning_rate": 7.247973571166593e-05, "loss": 0.0642, "step": 38400 }, { "grad_norm": 0.546093761920929, "learning_rate": 7.246496515025089e-05, "loss": 0.0542, "step": 38410 }, { "grad_norm": 0.6137871146202087, "learning_rate": 7.245019213210101e-05, "loss": 0.0516, "step": 38420 }, { "grad_norm": 0.38362836837768555, "learning_rate": 7.24354166588319e-05, "loss": 0.0567, "step": 38430 }, { "grad_norm": 0.550947368144989, "learning_rate": 7.242063873205935e-05, "loss": 0.0672, "step": 38440 }, { "grad_norm": 0.590234637260437, "learning_rate": 7.240585835339946e-05, "loss": 0.0566, "step": 38450 }, { "grad_norm": 0.427433580160141, "learning_rate": 7.239107552446857e-05, "loss": 0.0565, "step": 38460 }, { "grad_norm": 0.5470910668373108, "learning_rate": 7.237629024688334e-05, "loss": 0.0567, "step": 38470 }, { "grad_norm": 0.53558349609375, "learning_rate": 7.236150252226064e-05, "loss": 0.0559, "step": 38480 }, { "grad_norm": 0.44715049862861633, "learning_rate": 7.234671235221765e-05, "loss": 0.0528, "step": 38490 }, { "grad_norm": 0.5247056484222412, "learning_rate": 7.233191973837179e-05, "loss": 0.0624, "step": 38500 }, { "grad_norm": 0.6313949823379517, "learning_rate": 7.231712468234078e-05, "loss": 0.0546, "step": 38510 }, { "grad_norm": 0.6350990533828735, "learning_rate": 7.230232718574254e-05, "loss": 0.0631, "step": 38520 }, { "grad_norm": 0.28416818380355835, "learning_rate": 7.228752725019535e-05, "loss": 0.0592, "step": 38530 }, { "grad_norm": 0.39253705739974976, "learning_rate": 7.227272487731769e-05, "loss": 0.0515, "step": 38540 }, { "grad_norm": 0.49312835931777954, "learning_rate": 7.225792006872831e-05, "loss": 0.05, "step": 38550 }, { "grad_norm": 0.7514711022377014, "learning_rate": 7.224311282604628e-05, "loss": 0.0533, "step": 38560 }, { "grad_norm": 0.5473617315292358, "learning_rate": 7.222830315089085e-05, "loss": 0.0548, "step": 38570 }, { "grad_norm": 0.6133010387420654, "learning_rate": 7.22134910448816e-05, "loss": 0.0562, "step": 38580 }, { "grad_norm": 0.5486267805099487, "learning_rate": 7.219867650963839e-05, "loss": 0.0538, "step": 38590 }, { "grad_norm": 0.5758900046348572, "learning_rate": 7.21838595467813e-05, "loss": 0.0437, "step": 38600 }, { "grad_norm": 0.39117494225502014, "learning_rate": 7.216904015793067e-05, "loss": 0.0527, "step": 38610 }, { "grad_norm": 0.44008153676986694, "learning_rate": 7.215421834470713e-05, "loss": 0.0533, "step": 38620 }, { "grad_norm": 0.6339516043663025, "learning_rate": 7.213939410873158e-05, "loss": 0.0554, "step": 38630 }, { "grad_norm": 1.417202115058899, "learning_rate": 7.212456745162519e-05, "loss": 0.0453, "step": 38640 }, { "grad_norm": 1.5872858762741089, "learning_rate": 7.210973837500937e-05, "loss": 0.0627, "step": 38650 }, { "grad_norm": 0.7541807889938354, "learning_rate": 7.209490688050578e-05, "loss": 0.0515, "step": 38660 }, { "grad_norm": 0.5062749981880188, "learning_rate": 7.208007296973641e-05, "loss": 0.0533, "step": 38670 }, { "grad_norm": 0.5808671712875366, "learning_rate": 7.206523664432345e-05, "loss": 0.0723, "step": 38680 }, { "grad_norm": 0.954418957233429, "learning_rate": 7.205039790588939e-05, "loss": 0.0428, "step": 38690 }, { "grad_norm": 0.456589013338089, "learning_rate": 7.203555675605697e-05, "loss": 0.0549, "step": 38700 }, { "grad_norm": 0.5099046230316162, "learning_rate": 7.202071319644917e-05, "loss": 0.067, "step": 38710 }, { "grad_norm": 0.5002079606056213, "learning_rate": 7.200586722868932e-05, "loss": 0.051, "step": 38720 }, { "grad_norm": 0.40319111943244934, "learning_rate": 7.199101885440088e-05, "loss": 0.0575, "step": 38730 }, { "grad_norm": 0.4934159219264984, "learning_rate": 7.197616807520771e-05, "loss": 0.0519, "step": 38740 }, { "grad_norm": 0.5028888583183289, "learning_rate": 7.196131489273381e-05, "loss": 0.0465, "step": 38750 }, { "grad_norm": 0.43286797404289246, "learning_rate": 7.194645930860355e-05, "loss": 0.0439, "step": 38760 }, { "grad_norm": 1.1088519096374512, "learning_rate": 7.19316013244415e-05, "loss": 0.0537, "step": 38770 }, { "grad_norm": 0.5154114365577698, "learning_rate": 7.191674094187248e-05, "loss": 0.0632, "step": 38780 }, { "grad_norm": 0.5930035710334778, "learning_rate": 7.190187816252165e-05, "loss": 0.0454, "step": 38790 }, { "grad_norm": 0.8992857336997986, "learning_rate": 7.188701298801435e-05, "loss": 0.068, "step": 38800 }, { "grad_norm": 0.4788806736469269, "learning_rate": 7.18721454199762e-05, "loss": 0.0613, "step": 38810 }, { "grad_norm": 1.1606401205062866, "learning_rate": 7.185727546003312e-05, "loss": 0.0595, "step": 38820 }, { "grad_norm": 0.41003844141960144, "learning_rate": 7.184240310981126e-05, "loss": 0.0459, "step": 38830 }, { "grad_norm": 0.44349396228790283, "learning_rate": 7.182752837093704e-05, "loss": 0.0589, "step": 38840 }, { "grad_norm": 0.49730244278907776, "learning_rate": 7.181265124503711e-05, "loss": 0.0686, "step": 38850 }, { "grad_norm": 0.4548114538192749, "learning_rate": 7.179777173373847e-05, "loss": 0.049, "step": 38860 }, { "grad_norm": 0.531856894493103, "learning_rate": 7.178288983866826e-05, "loss": 0.0473, "step": 38870 }, { "grad_norm": 0.5917401909828186, "learning_rate": 7.176800556145397e-05, "loss": 0.0634, "step": 38880 }, { "grad_norm": 0.47700071334838867, "learning_rate": 7.175311890372334e-05, "loss": 0.052, "step": 38890 }, { "grad_norm": 0.7006648778915405, "learning_rate": 7.17382298671043e-05, "loss": 0.0583, "step": 38900 }, { "grad_norm": 0.5496140718460083, "learning_rate": 7.172333845322515e-05, "loss": 0.0502, "step": 38910 }, { "grad_norm": 0.5101029872894287, "learning_rate": 7.170844466371436e-05, "loss": 0.0573, "step": 38920 }, { "grad_norm": 0.5099411606788635, "learning_rate": 7.16935485002007e-05, "loss": 0.0526, "step": 38930 }, { "grad_norm": 0.6005198359489441, "learning_rate": 7.167864996431319e-05, "loss": 0.0699, "step": 38940 }, { "grad_norm": 0.44614407420158386, "learning_rate": 7.166374905768111e-05, "loss": 0.0593, "step": 38950 }, { "grad_norm": 0.5378091335296631, "learning_rate": 7.164884578193401e-05, "loss": 0.0445, "step": 38960 }, { "grad_norm": 0.5099169015884399, "learning_rate": 7.163394013870168e-05, "loss": 0.0616, "step": 38970 }, { "grad_norm": 0.7556114196777344, "learning_rate": 7.161903212961419e-05, "loss": 0.0596, "step": 38980 }, { "grad_norm": 0.4938572347164154, "learning_rate": 7.160412175630183e-05, "loss": 0.055, "step": 38990 }, { "grad_norm": 0.3098125159740448, "learning_rate": 7.158920902039521e-05, "loss": 0.0499, "step": 39000 }, { "grad_norm": 0.5823092460632324, "learning_rate": 7.157429392352514e-05, "loss": 0.0473, "step": 39010 }, { "grad_norm": 0.5446298718452454, "learning_rate": 7.155937646732275e-05, "loss": 0.0522, "step": 39020 }, { "grad_norm": 0.5315865874290466, "learning_rate": 7.154445665341933e-05, "loss": 0.0512, "step": 39030 }, { "grad_norm": 0.6779699921607971, "learning_rate": 7.152953448344654e-05, "loss": 0.0473, "step": 39040 }, { "grad_norm": 1.126837134361267, "learning_rate": 7.151460995903624e-05, "loss": 0.0486, "step": 39050 }, { "grad_norm": 0.649665892124176, "learning_rate": 7.149968308182052e-05, "loss": 0.0513, "step": 39060 }, { "grad_norm": 0.5428369641304016, "learning_rate": 7.14847538534318e-05, "loss": 0.0586, "step": 39070 }, { "grad_norm": 0.49140068888664246, "learning_rate": 7.14698222755027e-05, "loss": 0.0527, "step": 39080 }, { "grad_norm": 0.5338363647460938, "learning_rate": 7.14548883496661e-05, "loss": 0.0474, "step": 39090 }, { "grad_norm": 0.8312867283821106, "learning_rate": 7.143995207755517e-05, "loss": 0.0591, "step": 39100 }, { "grad_norm": 0.6190202236175537, "learning_rate": 7.142501346080333e-05, "loss": 0.0465, "step": 39110 }, { "grad_norm": 0.6757678389549255, "learning_rate": 7.141007250104421e-05, "loss": 0.0518, "step": 39120 }, { "grad_norm": 2.8074657917022705, "learning_rate": 7.139512919991176e-05, "loss": 0.0595, "step": 39130 }, { "grad_norm": 0.9162814021110535, "learning_rate": 7.138018355904014e-05, "loss": 0.0586, "step": 39140 }, { "grad_norm": 0.7475929260253906, "learning_rate": 7.13652355800638e-05, "loss": 0.0755, "step": 39150 }, { "grad_norm": 0.5667994618415833, "learning_rate": 7.135028526461743e-05, "loss": 0.0569, "step": 39160 }, { "grad_norm": 0.71302729845047, "learning_rate": 7.133533261433594e-05, "loss": 0.0556, "step": 39170 }, { "grad_norm": 0.9944559931755066, "learning_rate": 7.132037763085457e-05, "loss": 0.0563, "step": 39180 }, { "grad_norm": 0.4589068293571472, "learning_rate": 7.130542031580875e-05, "loss": 0.0611, "step": 39190 }, { "grad_norm": 0.4288880228996277, "learning_rate": 7.12904606708342e-05, "loss": 0.0563, "step": 39200 }, { "grad_norm": 0.4369710683822632, "learning_rate": 7.127549869756687e-05, "loss": 0.0662, "step": 39210 }, { "grad_norm": 0.5116899013519287, "learning_rate": 7.126053439764299e-05, "loss": 0.0534, "step": 39220 }, { "grad_norm": 0.45683929324150085, "learning_rate": 7.124556777269904e-05, "loss": 0.0624, "step": 39230 }, { "grad_norm": 0.7295219898223877, "learning_rate": 7.123059882437174e-05, "loss": 0.0712, "step": 39240 }, { "grad_norm": 0.8503839373588562, "learning_rate": 7.121562755429807e-05, "loss": 0.0546, "step": 39250 }, { "grad_norm": 0.44744306802749634, "learning_rate": 7.120065396411528e-05, "loss": 0.052, "step": 39260 }, { "grad_norm": 0.46703776717185974, "learning_rate": 7.118567805546084e-05, "loss": 0.0487, "step": 39270 }, { "grad_norm": 0.668727695941925, "learning_rate": 7.117069982997248e-05, "loss": 0.0545, "step": 39280 }, { "grad_norm": 0.4757644534111023, "learning_rate": 7.115571928928825e-05, "loss": 0.0745, "step": 39290 }, { "grad_norm": 0.6288663744926453, "learning_rate": 7.114073643504635e-05, "loss": 0.0617, "step": 39300 }, { "grad_norm": 0.870993971824646, "learning_rate": 7.11257512688853e-05, "loss": 0.058, "step": 39310 }, { "grad_norm": 0.41519248485565186, "learning_rate": 7.111076379244384e-05, "loss": 0.0447, "step": 39320 }, { "grad_norm": 0.8705571889877319, "learning_rate": 7.109577400736101e-05, "loss": 0.0619, "step": 39330 }, { "grad_norm": 0.38426586985588074, "learning_rate": 7.108078191527605e-05, "loss": 0.0599, "step": 39340 }, { "grad_norm": 0.5610134601593018, "learning_rate": 7.106578751782847e-05, "loss": 0.0643, "step": 39350 }, { "grad_norm": 0.31522703170776367, "learning_rate": 7.105079081665803e-05, "loss": 0.0496, "step": 39360 }, { "grad_norm": 0.4357529282569885, "learning_rate": 7.103579181340476e-05, "loss": 0.0667, "step": 39370 }, { "grad_norm": 0.7764650583267212, "learning_rate": 7.102079050970893e-05, "loss": 0.05, "step": 39380 }, { "grad_norm": 0.47559407353401184, "learning_rate": 7.100578690721105e-05, "loss": 0.0573, "step": 39390 }, { "grad_norm": 0.5322181582450867, "learning_rate": 7.09907810075519e-05, "loss": 0.0588, "step": 39400 }, { "grad_norm": 0.43542417883872986, "learning_rate": 7.097577281237249e-05, "loss": 0.0603, "step": 39410 }, { "grad_norm": 1.1573375463485718, "learning_rate": 7.09607623233141e-05, "loss": 0.0407, "step": 39420 }, { "grad_norm": 0.5463812351226807, "learning_rate": 7.094574954201827e-05, "loss": 0.0557, "step": 39430 }, { "grad_norm": 0.5460500121116638, "learning_rate": 7.093073447012675e-05, "loss": 0.0639, "step": 39440 }, { "grad_norm": 0.7212914824485779, "learning_rate": 7.09157171092816e-05, "loss": 0.0529, "step": 39450 }, { "grad_norm": 0.39841559529304504, "learning_rate": 7.090069746112504e-05, "loss": 0.0635, "step": 39460 }, { "grad_norm": 0.32514575123786926, "learning_rate": 7.088567552729965e-05, "loss": 0.0506, "step": 39470 }, { "grad_norm": 0.4717951714992523, "learning_rate": 7.087065130944818e-05, "loss": 0.0626, "step": 39480 }, { "grad_norm": 0.630269467830658, "learning_rate": 7.085562480921366e-05, "loss": 0.0664, "step": 39490 }, { "grad_norm": 0.4951692819595337, "learning_rate": 7.084059602823937e-05, "loss": 0.0553, "step": 39500 }, { "grad_norm": 0.6482915282249451, "learning_rate": 7.082556496816882e-05, "loss": 0.0504, "step": 39510 }, { "grad_norm": 0.45025569200515747, "learning_rate": 7.081053163064582e-05, "loss": 0.0627, "step": 39520 }, { "grad_norm": 0.7045347690582275, "learning_rate": 7.079549601731434e-05, "loss": 0.0551, "step": 39530 }, { "grad_norm": 0.6165891885757446, "learning_rate": 7.07804581298187e-05, "loss": 0.0503, "step": 39540 }, { "grad_norm": 0.5664123892784119, "learning_rate": 7.07654179698034e-05, "loss": 0.0607, "step": 39550 }, { "grad_norm": 0.6168689131736755, "learning_rate": 7.075037553891321e-05, "loss": 0.066, "step": 39560 }, { "grad_norm": 0.49818918108940125, "learning_rate": 7.073533083879315e-05, "loss": 0.0516, "step": 39570 }, { "grad_norm": 0.5184632539749146, "learning_rate": 7.072028387108849e-05, "loss": 0.0515, "step": 39580 }, { "grad_norm": 0.4231504201889038, "learning_rate": 7.070523463744472e-05, "loss": 0.0673, "step": 39590 }, { "grad_norm": 0.31352272629737854, "learning_rate": 7.069018313950763e-05, "loss": 0.0543, "step": 39600 }, { "grad_norm": 0.4382774829864502, "learning_rate": 7.067512937892322e-05, "loss": 0.0524, "step": 39610 }, { "grad_norm": 0.5439382791519165, "learning_rate": 7.066007335733775e-05, "loss": 0.0435, "step": 39620 }, { "grad_norm": 0.9433297514915466, "learning_rate": 7.064501507639772e-05, "loss": 0.0555, "step": 39630 }, { "grad_norm": 0.652681827545166, "learning_rate": 7.062995453774987e-05, "loss": 0.07, "step": 39640 }, { "grad_norm": 0.5188747048377991, "learning_rate": 7.061489174304121e-05, "loss": 0.0701, "step": 39650 }, { "grad_norm": 0.3361961245536804, "learning_rate": 7.0599826693919e-05, "loss": 0.0577, "step": 39660 }, { "grad_norm": 0.6167629361152649, "learning_rate": 7.05847593920307e-05, "loss": 0.051, "step": 39670 }, { "grad_norm": 0.6221704483032227, "learning_rate": 7.056968983902406e-05, "loss": 0.0518, "step": 39680 }, { "grad_norm": 0.6016175150871277, "learning_rate": 7.055461803654706e-05, "loss": 0.0508, "step": 39690 }, { "grad_norm": 0.4641375243663788, "learning_rate": 7.053954398624794e-05, "loss": 0.0426, "step": 39700 }, { "grad_norm": 0.3537746071815491, "learning_rate": 7.052446768977518e-05, "loss": 0.0488, "step": 39710 }, { "grad_norm": 0.44746753573417664, "learning_rate": 7.050938914877748e-05, "loss": 0.0472, "step": 39720 }, { "grad_norm": 0.6432476043701172, "learning_rate": 7.049430836490382e-05, "loss": 0.069, "step": 39730 }, { "grad_norm": 0.4703616499900818, "learning_rate": 7.047922533980342e-05, "loss": 0.0522, "step": 39740 }, { "grad_norm": 0.6747845411300659, "learning_rate": 7.046414007512571e-05, "loss": 0.05, "step": 39750 }, { "grad_norm": 0.6870049238204956, "learning_rate": 7.044905257252042e-05, "loss": 0.0465, "step": 39760 }, { "grad_norm": 0.7098677158355713, "learning_rate": 7.04339628336375e-05, "loss": 0.0653, "step": 39770 }, { "grad_norm": 0.47169533371925354, "learning_rate": 7.041887086012711e-05, "loss": 0.0477, "step": 39780 }, { "grad_norm": 0.6767202615737915, "learning_rate": 7.040377665363969e-05, "loss": 0.053, "step": 39790 }, { "grad_norm": 0.49323219060897827, "learning_rate": 7.038868021582594e-05, "loss": 0.0611, "step": 39800 }, { "grad_norm": 0.4824155271053314, "learning_rate": 7.037358154833679e-05, "loss": 0.0525, "step": 39810 }, { "grad_norm": 0.28571343421936035, "learning_rate": 7.035848065282339e-05, "loss": 0.0566, "step": 39820 }, { "grad_norm": 0.42833706736564636, "learning_rate": 7.034337753093714e-05, "loss": 0.0504, "step": 39830 }, { "grad_norm": 0.5226733684539795, "learning_rate": 7.032827218432972e-05, "loss": 0.0551, "step": 39840 }, { "grad_norm": 0.45019474625587463, "learning_rate": 7.031316461465302e-05, "loss": 0.0605, "step": 39850 }, { "grad_norm": 0.5668237209320068, "learning_rate": 7.029805482355915e-05, "loss": 0.0543, "step": 39860 }, { "grad_norm": 0.4707053601741791, "learning_rate": 7.028294281270055e-05, "loss": 0.0435, "step": 39870 }, { "grad_norm": 0.41525229811668396, "learning_rate": 7.026782858372982e-05, "loss": 0.0495, "step": 39880 }, { "grad_norm": 0.5322439670562744, "learning_rate": 7.025271213829982e-05, "loss": 0.0501, "step": 39890 }, { "grad_norm": 0.40736502408981323, "learning_rate": 7.023759347806366e-05, "loss": 0.0454, "step": 39900 }, { "grad_norm": 0.5662975907325745, "learning_rate": 7.022247260467472e-05, "loss": 0.0565, "step": 39910 }, { "grad_norm": 0.4999246299266815, "learning_rate": 7.020734951978658e-05, "loss": 0.051, "step": 39920 }, { "grad_norm": 0.5619690418243408, "learning_rate": 7.019222422505307e-05, "loss": 0.0501, "step": 39930 }, { "grad_norm": 0.42206451296806335, "learning_rate": 7.017709672212827e-05, "loss": 0.0674, "step": 39940 }, { "grad_norm": 0.5413761734962463, "learning_rate": 7.016196701266652e-05, "loss": 0.0567, "step": 39950 }, { "grad_norm": 0.6183328032493591, "learning_rate": 7.014683509832238e-05, "loss": 0.0502, "step": 39960 }, { "grad_norm": 0.5939422845840454, "learning_rate": 7.013170098075063e-05, "loss": 0.0651, "step": 39970 }, { "grad_norm": 0.39942634105682373, "learning_rate": 7.011656466160632e-05, "loss": 0.0577, "step": 39980 }, { "grad_norm": 0.4845411479473114, "learning_rate": 7.010142614254475e-05, "loss": 0.0483, "step": 39990 }, { "grad_norm": 0.40536943078041077, "learning_rate": 7.008628542522147e-05, "loss": 0.06, "step": 40000 }, { "grad_norm": 0.5127610564231873, "learning_rate": 7.007114251129218e-05, "loss": 0.0501, "step": 40010 }, { "grad_norm": 0.8791177272796631, "learning_rate": 7.005599740241293e-05, "loss": 0.0422, "step": 40020 }, { "grad_norm": 0.4774782061576843, "learning_rate": 7.004085010023996e-05, "loss": 0.0583, "step": 40030 }, { "grad_norm": 0.6426105499267578, "learning_rate": 7.002570060642976e-05, "loss": 0.0449, "step": 40040 }, { "grad_norm": 0.6477228403091431, "learning_rate": 7.001054892263903e-05, "loss": 0.0434, "step": 40050 }, { "grad_norm": 0.5850817561149597, "learning_rate": 6.999539505052477e-05, "loss": 0.0602, "step": 40060 }, { "grad_norm": 0.5230039358139038, "learning_rate": 6.998023899174418e-05, "loss": 0.0461, "step": 40070 }, { "grad_norm": 0.472710520029068, "learning_rate": 6.996508074795467e-05, "loss": 0.0492, "step": 40080 }, { "grad_norm": 0.5297344923019409, "learning_rate": 6.994992032081396e-05, "loss": 0.0604, "step": 40090 }, { "grad_norm": 0.5002930164337158, "learning_rate": 6.993475771197995e-05, "loss": 0.0554, "step": 40100 }, { "grad_norm": 0.6173412203788757, "learning_rate": 6.991959292311082e-05, "loss": 0.0522, "step": 40110 }, { "grad_norm": 0.5094457864761353, "learning_rate": 6.990442595586495e-05, "loss": 0.0555, "step": 40120 }, { "grad_norm": 0.8052211403846741, "learning_rate": 6.988925681190098e-05, "loss": 0.0595, "step": 40130 }, { "grad_norm": 0.6948457360267639, "learning_rate": 6.987408549287778e-05, "loss": 0.0653, "step": 40140 }, { "grad_norm": 0.755089282989502, "learning_rate": 6.985891200045449e-05, "loss": 0.0684, "step": 40150 }, { "grad_norm": 0.53725266456604, "learning_rate": 6.984373633629045e-05, "loss": 0.0719, "step": 40160 }, { "grad_norm": 1.2571425437927246, "learning_rate": 6.98285585020452e-05, "loss": 0.0556, "step": 40170 }, { "grad_norm": 0.3352161645889282, "learning_rate": 6.981337849937864e-05, "loss": 0.0527, "step": 40180 }, { "grad_norm": 0.7479858994483948, "learning_rate": 6.979819632995078e-05, "loss": 0.0565, "step": 40190 }, { "grad_norm": 0.3836667537689209, "learning_rate": 6.978301199542193e-05, "loss": 0.0398, "step": 40200 }, { "grad_norm": 0.5473276972770691, "learning_rate": 6.976782549745263e-05, "loss": 0.0455, "step": 40210 }, { "grad_norm": 0.46570107340812683, "learning_rate": 6.975263683770367e-05, "loss": 0.056, "step": 40220 }, { "grad_norm": 0.40403103828430176, "learning_rate": 6.973744601783604e-05, "loss": 0.0614, "step": 40230 }, { "grad_norm": 0.5217891335487366, "learning_rate": 6.972225303951097e-05, "loss": 0.0514, "step": 40240 }, { "grad_norm": 0.3605613708496094, "learning_rate": 6.970705790438998e-05, "loss": 0.0457, "step": 40250 }, { "grad_norm": 0.5996297001838684, "learning_rate": 6.969186061413477e-05, "loss": 0.0522, "step": 40260 }, { "grad_norm": 0.5869225263595581, "learning_rate": 6.967666117040727e-05, "loss": 0.0683, "step": 40270 }, { "grad_norm": 0.4564352035522461, "learning_rate": 6.966145957486972e-05, "loss": 0.051, "step": 40280 }, { "grad_norm": 0.45479515194892883, "learning_rate": 6.964625582918449e-05, "loss": 0.0552, "step": 40290 }, { "grad_norm": 0.3458164930343628, "learning_rate": 6.963104993501425e-05, "loss": 0.0514, "step": 40300 }, { "grad_norm": 0.4540257453918457, "learning_rate": 6.961584189402192e-05, "loss": 0.0403, "step": 40310 }, { "grad_norm": 0.5648849606513977, "learning_rate": 6.96006317078706e-05, "loss": 0.0387, "step": 40320 }, { "grad_norm": 1.0471673011779785, "learning_rate": 6.958541937822367e-05, "loss": 0.0505, "step": 40330 }, { "grad_norm": 0.40781551599502563, "learning_rate": 6.957020490674473e-05, "loss": 0.0547, "step": 40340 }, { "grad_norm": 0.4373512268066406, "learning_rate": 6.95549882950976e-05, "loss": 0.05, "step": 40350 }, { "grad_norm": 0.5782313346862793, "learning_rate": 6.953976954494635e-05, "loss": 0.0385, "step": 40360 }, { "grad_norm": 0.4457709789276123, "learning_rate": 6.952454865795528e-05, "loss": 0.0487, "step": 40370 }, { "grad_norm": 0.6463590264320374, "learning_rate": 6.950932563578892e-05, "loss": 0.0534, "step": 40380 }, { "grad_norm": 0.4717680811882019, "learning_rate": 6.949410048011206e-05, "loss": 0.0503, "step": 40390 }, { "grad_norm": 0.6808131337165833, "learning_rate": 6.947887319258966e-05, "loss": 0.0614, "step": 40400 }, { "grad_norm": 0.4497283101081848, "learning_rate": 6.946364377488696e-05, "loss": 0.0489, "step": 40410 }, { "grad_norm": 0.5084958672523499, "learning_rate": 6.944841222866947e-05, "loss": 0.0628, "step": 40420 }, { "grad_norm": 0.4173315167427063, "learning_rate": 6.943317855560284e-05, "loss": 0.0603, "step": 40430 }, { "grad_norm": 0.8093646764755249, "learning_rate": 6.941794275735302e-05, "loss": 0.0534, "step": 40440 }, { "grad_norm": 0.4293493926525116, "learning_rate": 6.94027048355862e-05, "loss": 0.0453, "step": 40450 }, { "grad_norm": 0.6931309103965759, "learning_rate": 6.938746479196871e-05, "loss": 0.0547, "step": 40460 }, { "grad_norm": 0.5853438973426819, "learning_rate": 6.937222262816724e-05, "loss": 0.0475, "step": 40470 }, { "grad_norm": 0.300969660282135, "learning_rate": 6.935697834584865e-05, "loss": 0.0455, "step": 40480 }, { "grad_norm": 0.627765953540802, "learning_rate": 6.934173194667998e-05, "loss": 0.052, "step": 40490 }, { "grad_norm": 0.7212657928466797, "learning_rate": 6.93264834323286e-05, "loss": 0.0579, "step": 40500 }, { "grad_norm": 0.7226983308792114, "learning_rate": 6.931123280446206e-05, "loss": 0.0567, "step": 40510 }, { "grad_norm": 0.5422793030738831, "learning_rate": 6.929598006474811e-05, "loss": 0.055, "step": 40520 }, { "grad_norm": 0.32501664757728577, "learning_rate": 6.92807252148548e-05, "loss": 0.0577, "step": 40530 }, { "grad_norm": 0.49461546540260315, "learning_rate": 6.926546825645037e-05, "loss": 0.0423, "step": 40540 }, { "grad_norm": 0.304010808467865, "learning_rate": 6.92502091912033e-05, "loss": 0.0486, "step": 40550 }, { "grad_norm": 0.4416493773460388, "learning_rate": 6.923494802078229e-05, "loss": 0.0568, "step": 40560 }, { "grad_norm": 0.7356517910957336, "learning_rate": 6.921968474685628e-05, "loss": 0.0549, "step": 40570 }, { "grad_norm": 0.7177410125732422, "learning_rate": 6.920441937109445e-05, "loss": 0.0642, "step": 40580 }, { "grad_norm": 0.4110654592514038, "learning_rate": 6.91891518951662e-05, "loss": 0.0456, "step": 40590 }, { "grad_norm": 1.8127546310424805, "learning_rate": 6.917388232074114e-05, "loss": 0.0523, "step": 40600 }, { "grad_norm": 0.5129222869873047, "learning_rate": 6.915861064948914e-05, "loss": 0.0614, "step": 40610 }, { "grad_norm": 0.436022013425827, "learning_rate": 6.914333688308025e-05, "loss": 0.0446, "step": 40620 }, { "grad_norm": 0.45353710651397705, "learning_rate": 6.912806102318487e-05, "loss": 0.0473, "step": 40630 }, { "grad_norm": 0.3489135503768921, "learning_rate": 6.911278307147347e-05, "loss": 0.0493, "step": 40640 }, { "grad_norm": 0.4249739944934845, "learning_rate": 6.909750302961684e-05, "loss": 0.0515, "step": 40650 }, { "grad_norm": 1.0907821655273438, "learning_rate": 6.9082220899286e-05, "loss": 0.0579, "step": 40660 }, { "grad_norm": 0.8862417340278625, "learning_rate": 6.906693668215219e-05, "loss": 0.0699, "step": 40670 }, { "grad_norm": 0.5064948797225952, "learning_rate": 6.905165037988683e-05, "loss": 0.0476, "step": 40680 }, { "grad_norm": 0.5856199264526367, "learning_rate": 6.903636199416164e-05, "loss": 0.0544, "step": 40690 }, { "grad_norm": 0.44162383675575256, "learning_rate": 6.902107152664851e-05, "loss": 0.0517, "step": 40700 }, { "grad_norm": 0.6820310354232788, "learning_rate": 6.90057789790196e-05, "loss": 0.0517, "step": 40710 }, { "grad_norm": 0.5247237086296082, "learning_rate": 6.899048435294728e-05, "loss": 0.0624, "step": 40720 }, { "grad_norm": 0.5051069259643555, "learning_rate": 6.897518765010415e-05, "loss": 0.0557, "step": 40730 }, { "grad_norm": 0.3378465473651886, "learning_rate": 6.895988887216302e-05, "loss": 0.0566, "step": 40740 }, { "grad_norm": 0.5290533304214478, "learning_rate": 6.894458802079694e-05, "loss": 0.056, "step": 40750 }, { "grad_norm": 0.674639105796814, "learning_rate": 6.892928509767922e-05, "loss": 0.0473, "step": 40760 }, { "grad_norm": 0.5621660351753235, "learning_rate": 6.891398010448333e-05, "loss": 0.0532, "step": 40770 }, { "grad_norm": 0.7998679280281067, "learning_rate": 6.8898673042883e-05, "loss": 0.0693, "step": 40780 }, { "grad_norm": 0.44926998019218445, "learning_rate": 6.888336391455222e-05, "loss": 0.0519, "step": 40790 }, { "grad_norm": 0.41099610924720764, "learning_rate": 6.886805272116513e-05, "loss": 0.0512, "step": 40800 }, { "grad_norm": 0.5351479053497314, "learning_rate": 6.885273946439617e-05, "loss": 0.0661, "step": 40810 }, { "grad_norm": 0.7082638144493103, "learning_rate": 6.883742414591998e-05, "loss": 0.0607, "step": 40820 }, { "grad_norm": 0.6872256994247437, "learning_rate": 6.88221067674114e-05, "loss": 0.0526, "step": 40830 }, { "grad_norm": 0.7341225147247314, "learning_rate": 6.88067873305455e-05, "loss": 0.053, "step": 40840 }, { "grad_norm": 0.6806560754776001, "learning_rate": 6.879146583699765e-05, "loss": 0.0516, "step": 40850 }, { "grad_norm": 0.43318653106689453, "learning_rate": 6.877614228844334e-05, "loss": 0.0537, "step": 40860 }, { "grad_norm": 0.47184041142463684, "learning_rate": 6.876081668655832e-05, "loss": 0.0518, "step": 40870 }, { "grad_norm": 0.49564629793167114, "learning_rate": 6.874548903301861e-05, "loss": 0.0625, "step": 40880 }, { "grad_norm": 0.5719206929206848, "learning_rate": 6.873015932950039e-05, "loss": 0.0438, "step": 40890 }, { "grad_norm": 0.5010266304016113, "learning_rate": 6.871482757768012e-05, "loss": 0.058, "step": 40900 }, { "grad_norm": 0.46964284777641296, "learning_rate": 6.869949377923442e-05, "loss": 0.059, "step": 40910 }, { "grad_norm": 0.6359380483627319, "learning_rate": 6.868415793584022e-05, "loss": 0.0581, "step": 40920 }, { "grad_norm": 0.40092194080352783, "learning_rate": 6.86688200491746e-05, "loss": 0.0417, "step": 40930 }, { "grad_norm": 0.33573487401008606, "learning_rate": 6.865348012091488e-05, "loss": 0.0468, "step": 40940 }, { "grad_norm": 0.43758654594421387, "learning_rate": 6.86381381527386e-05, "loss": 0.0456, "step": 40950 }, { "grad_norm": 0.6764233708381653, "learning_rate": 6.862279414632359e-05, "loss": 0.0502, "step": 40960 }, { "grad_norm": 0.4612078070640564, "learning_rate": 6.860744810334778e-05, "loss": 0.0447, "step": 40970 }, { "grad_norm": 0.33713921904563904, "learning_rate": 6.859210002548943e-05, "loss": 0.0526, "step": 40980 }, { "grad_norm": 0.4005267918109894, "learning_rate": 6.8576749914427e-05, "loss": 0.0462, "step": 40990 }, { "grad_norm": 0.6456696391105652, "learning_rate": 6.85613977718391e-05, "loss": 0.0491, "step": 41000 }, { "grad_norm": 0.4725490212440491, "learning_rate": 6.854604359940464e-05, "loss": 0.0447, "step": 41010 }, { "grad_norm": 0.5809876918792725, "learning_rate": 6.853068739880276e-05, "loss": 0.0513, "step": 41020 }, { "grad_norm": 0.618849515914917, "learning_rate": 6.851532917171276e-05, "loss": 0.0521, "step": 41030 }, { "grad_norm": 0.5495668649673462, "learning_rate": 6.849996891981418e-05, "loss": 0.0666, "step": 41040 }, { "grad_norm": 0.6903206706047058, "learning_rate": 6.848460664478684e-05, "loss": 0.0559, "step": 41050 }, { "grad_norm": 0.5353668332099915, "learning_rate": 6.846924234831069e-05, "loss": 0.0478, "step": 41060 }, { "grad_norm": 1.0667818784713745, "learning_rate": 6.845387603206597e-05, "loss": 0.061, "step": 41070 }, { "grad_norm": 0.32912227511405945, "learning_rate": 6.843850769773312e-05, "loss": 0.0453, "step": 41080 }, { "grad_norm": 0.502159833908081, "learning_rate": 6.842313734699278e-05, "loss": 0.0601, "step": 41090 }, { "grad_norm": 0.6689754724502563, "learning_rate": 6.840776498152584e-05, "loss": 0.0522, "step": 41100 }, { "grad_norm": 0.6522342562675476, "learning_rate": 6.83923906030134e-05, "loss": 0.0442, "step": 41110 }, { "grad_norm": 0.7912876009941101, "learning_rate": 6.837701421313677e-05, "loss": 0.0723, "step": 41120 }, { "grad_norm": 0.6495282053947449, "learning_rate": 6.83616358135775e-05, "loss": 0.0519, "step": 41130 }, { "grad_norm": 0.7674503922462463, "learning_rate": 6.834625540601734e-05, "loss": 0.069, "step": 41140 }, { "grad_norm": 0.6036444306373596, "learning_rate": 6.833087299213829e-05, "loss": 0.0418, "step": 41150 }, { "grad_norm": 0.8456528186798096, "learning_rate": 6.83154885736225e-05, "loss": 0.0626, "step": 41160 }, { "grad_norm": 0.5229440927505493, "learning_rate": 6.830010215215244e-05, "loss": 0.054, "step": 41170 }, { "grad_norm": 0.6699286103248596, "learning_rate": 6.828471372941072e-05, "loss": 0.0604, "step": 41180 }, { "grad_norm": 0.4920129179954529, "learning_rate": 6.826932330708018e-05, "loss": 0.0431, "step": 41190 }, { "grad_norm": 0.5365797877311707, "learning_rate": 6.825393088684393e-05, "loss": 0.0467, "step": 41200 }, { "grad_norm": 0.5286691188812256, "learning_rate": 6.823853647038524e-05, "loss": 0.0459, "step": 41210 }, { "grad_norm": 0.33112791180610657, "learning_rate": 6.822314005938762e-05, "loss": 0.0436, "step": 41220 }, { "grad_norm": 0.7491744756698608, "learning_rate": 6.820774165553479e-05, "loss": 0.0584, "step": 41230 }, { "grad_norm": 0.7064136862754822, "learning_rate": 6.819234126051072e-05, "loss": 0.0455, "step": 41240 }, { "grad_norm": 0.5271412134170532, "learning_rate": 6.817693887599956e-05, "loss": 0.0593, "step": 41250 }, { "grad_norm": 0.4080163836479187, "learning_rate": 6.816153450368568e-05, "loss": 0.0583, "step": 41260 }, { "grad_norm": 0.5119894742965698, "learning_rate": 6.814612814525372e-05, "loss": 0.0594, "step": 41270 }, { "grad_norm": 0.6308553814888, "learning_rate": 6.813071980238846e-05, "loss": 0.0521, "step": 41280 }, { "grad_norm": 0.5188915729522705, "learning_rate": 6.811530947677492e-05, "loss": 0.0508, "step": 41290 }, { "grad_norm": 0.45713165402412415, "learning_rate": 6.809989717009839e-05, "loss": 0.0513, "step": 41300 }, { "grad_norm": 0.5223844051361084, "learning_rate": 6.808448288404431e-05, "loss": 0.0642, "step": 41310 }, { "grad_norm": 0.4775847792625427, "learning_rate": 6.806906662029838e-05, "loss": 0.0548, "step": 41320 }, { "grad_norm": 0.6814786791801453, "learning_rate": 6.805364838054647e-05, "loss": 0.0504, "step": 41330 }, { "grad_norm": 0.4264069199562073, "learning_rate": 6.803822816647474e-05, "loss": 0.0432, "step": 41340 }, { "grad_norm": 0.7065721154212952, "learning_rate": 6.802280597976949e-05, "loss": 0.0576, "step": 41350 }, { "grad_norm": 0.512365996837616, "learning_rate": 6.800738182211727e-05, "loss": 0.0481, "step": 41360 }, { "grad_norm": 0.6277287602424622, "learning_rate": 6.799195569520485e-05, "loss": 0.0535, "step": 41370 }, { "grad_norm": 0.47648879885673523, "learning_rate": 6.79765276007192e-05, "loss": 0.0558, "step": 41380 }, { "grad_norm": 0.5140426158905029, "learning_rate": 6.796109754034753e-05, "loss": 0.0476, "step": 41390 }, { "grad_norm": 0.408782958984375, "learning_rate": 6.794566551577724e-05, "loss": 0.0474, "step": 41400 }, { "grad_norm": 1.3153183460235596, "learning_rate": 6.793023152869594e-05, "loss": 0.051, "step": 41410 }, { "grad_norm": 0.5620173215866089, "learning_rate": 6.791479558079147e-05, "loss": 0.049, "step": 41420 }, { "grad_norm": 0.4096064865589142, "learning_rate": 6.789935767375191e-05, "loss": 0.0526, "step": 41430 }, { "grad_norm": 0.3580051362514496, "learning_rate": 6.78839178092655e-05, "loss": 0.0351, "step": 41440 }, { "grad_norm": 0.42130470275878906, "learning_rate": 6.786847598902072e-05, "loss": 0.0553, "step": 41450 }, { "grad_norm": 0.662505567073822, "learning_rate": 6.785303221470627e-05, "loss": 0.0573, "step": 41460 }, { "grad_norm": 0.43330860137939453, "learning_rate": 6.783758648801108e-05, "loss": 0.0441, "step": 41470 }, { "grad_norm": 0.7150783538818359, "learning_rate": 6.782213881062422e-05, "loss": 0.0541, "step": 41480 }, { "grad_norm": 0.6234450340270996, "learning_rate": 6.780668918423508e-05, "loss": 0.0502, "step": 41490 }, { "grad_norm": 0.7269213199615479, "learning_rate": 6.779123761053317e-05, "loss": 0.0498, "step": 41500 }, { "grad_norm": 0.5469293594360352, "learning_rate": 6.777578409120826e-05, "loss": 0.0631, "step": 41510 }, { "grad_norm": 0.5711106061935425, "learning_rate": 6.776032862795033e-05, "loss": 0.0606, "step": 41520 }, { "grad_norm": 0.5853442549705505, "learning_rate": 6.774487122244956e-05, "loss": 0.0517, "step": 41530 }, { "grad_norm": 0.5714892148971558, "learning_rate": 6.772941187639637e-05, "loss": 0.0408, "step": 41540 }, { "grad_norm": 0.3810364603996277, "learning_rate": 6.771395059148134e-05, "loss": 0.0562, "step": 41550 }, { "grad_norm": 0.5057798624038696, "learning_rate": 6.76984873693953e-05, "loss": 0.05, "step": 41560 }, { "grad_norm": 0.5899947881698608, "learning_rate": 6.768302221182931e-05, "loss": 0.0611, "step": 41570 }, { "grad_norm": 0.6307139992713928, "learning_rate": 6.766755512047457e-05, "loss": 0.0633, "step": 41580 }, { "grad_norm": 0.6322001814842224, "learning_rate": 6.765208609702259e-05, "loss": 0.0586, "step": 41590 }, { "grad_norm": 0.49982714653015137, "learning_rate": 6.763661514316499e-05, "loss": 0.0459, "step": 41600 }, { "grad_norm": 0.3081877529621124, "learning_rate": 6.76211422605937e-05, "loss": 0.0528, "step": 41610 }, { "grad_norm": 0.4689837694168091, "learning_rate": 6.760566745100076e-05, "loss": 0.0514, "step": 41620 }, { "grad_norm": 0.6938157081604004, "learning_rate": 6.75901907160785e-05, "loss": 0.0525, "step": 41630 }, { "grad_norm": 0.2670300304889679, "learning_rate": 6.757471205751943e-05, "loss": 0.0424, "step": 41640 }, { "grad_norm": 0.5710458755493164, "learning_rate": 6.75592314770163e-05, "loss": 0.0565, "step": 41650 }, { "grad_norm": 0.718573808670044, "learning_rate": 6.7543748976262e-05, "loss": 0.0582, "step": 41660 }, { "grad_norm": 1.0017651319503784, "learning_rate": 6.752826455694968e-05, "loss": 0.0522, "step": 41670 }, { "grad_norm": 0.5096123814582825, "learning_rate": 6.751277822077271e-05, "loss": 0.05, "step": 41680 }, { "grad_norm": 0.285455584526062, "learning_rate": 6.749728996942465e-05, "loss": 0.0437, "step": 41690 }, { "grad_norm": 0.3360357880592346, "learning_rate": 6.748179980459924e-05, "loss": 0.0466, "step": 41700 }, { "grad_norm": 0.324491947889328, "learning_rate": 6.746630772799052e-05, "loss": 0.0413, "step": 41710 }, { "grad_norm": 0.44017401337623596, "learning_rate": 6.745081374129262e-05, "loss": 0.0479, "step": 41720 }, { "grad_norm": 0.6362271904945374, "learning_rate": 6.743531784619998e-05, "loss": 0.0583, "step": 41730 }, { "grad_norm": 0.6229111552238464, "learning_rate": 6.741982004440719e-05, "loss": 0.0544, "step": 41740 }, { "grad_norm": 0.5944031476974487, "learning_rate": 6.740432033760907e-05, "loss": 0.0566, "step": 41750 }, { "grad_norm": 0.7041306495666504, "learning_rate": 6.738881872750066e-05, "loss": 0.0482, "step": 41760 }, { "grad_norm": 0.4453220069408417, "learning_rate": 6.737331521577715e-05, "loss": 0.0578, "step": 41770 }, { "grad_norm": 0.575386643409729, "learning_rate": 6.735780980413403e-05, "loss": 0.0519, "step": 41780 }, { "grad_norm": 1.0070931911468506, "learning_rate": 6.734230249426692e-05, "loss": 0.0534, "step": 41790 }, { "grad_norm": 0.37170732021331787, "learning_rate": 6.732679328787168e-05, "loss": 0.0559, "step": 41800 }, { "grad_norm": 0.37991416454315186, "learning_rate": 6.731128218664438e-05, "loss": 0.0418, "step": 41810 }, { "grad_norm": 0.7595049142837524, "learning_rate": 6.729576919228129e-05, "loss": 0.039, "step": 41820 }, { "grad_norm": 0.4539426863193512, "learning_rate": 6.728025430647888e-05, "loss": 0.0427, "step": 41830 }, { "grad_norm": 0.5898580551147461, "learning_rate": 6.726473753093383e-05, "loss": 0.0517, "step": 41840 }, { "grad_norm": 0.4458160102367401, "learning_rate": 6.724921886734305e-05, "loss": 0.0457, "step": 41850 }, { "grad_norm": 0.6250565648078918, "learning_rate": 6.723369831740363e-05, "loss": 0.0608, "step": 41860 }, { "grad_norm": 0.31170907616615295, "learning_rate": 6.721817588281288e-05, "loss": 0.0579, "step": 41870 }, { "grad_norm": 0.628436267375946, "learning_rate": 6.720265156526828e-05, "loss": 0.0513, "step": 41880 }, { "grad_norm": 0.4277060031890869, "learning_rate": 6.718712536646758e-05, "loss": 0.0477, "step": 41890 }, { "grad_norm": 0.7350761890411377, "learning_rate": 6.71715972881087e-05, "loss": 0.0513, "step": 41900 }, { "grad_norm": 0.4274984300136566, "learning_rate": 6.715606733188974e-05, "loss": 0.0483, "step": 41910 }, { "grad_norm": 0.3925864100456238, "learning_rate": 6.714053549950909e-05, "loss": 0.0424, "step": 41920 }, { "grad_norm": 0.5419528484344482, "learning_rate": 6.71250017926652e-05, "loss": 0.0545, "step": 41930 }, { "grad_norm": 0.5034242272377014, "learning_rate": 6.71094662130569e-05, "loss": 0.0485, "step": 41940 }, { "grad_norm": 0.7609957456588745, "learning_rate": 6.709392876238307e-05, "loss": 0.0552, "step": 41950 }, { "grad_norm": 0.7530171871185303, "learning_rate": 6.707838944234289e-05, "loss": 0.0588, "step": 41960 }, { "grad_norm": 0.46725016832351685, "learning_rate": 6.706284825463574e-05, "loss": 0.0482, "step": 41970 }, { "grad_norm": 0.5606685280799866, "learning_rate": 6.704730520096114e-05, "loss": 0.0529, "step": 41980 }, { "grad_norm": 0.7039974927902222, "learning_rate": 6.703176028301888e-05, "loss": 0.0521, "step": 41990 }, { "grad_norm": 0.42099952697753906, "learning_rate": 6.701621350250892e-05, "loss": 0.05, "step": 42000 }, { "grad_norm": 0.7346627712249756, "learning_rate": 6.700066486113144e-05, "loss": 0.0535, "step": 42010 }, { "grad_norm": 1.2269922494888306, "learning_rate": 6.69851143605868e-05, "loss": 0.0599, "step": 42020 }, { "grad_norm": 0.5822352170944214, "learning_rate": 6.696956200257561e-05, "loss": 0.0463, "step": 42030 }, { "grad_norm": 0.3257177472114563, "learning_rate": 6.695400778879863e-05, "loss": 0.0547, "step": 42040 }, { "grad_norm": 0.6067512035369873, "learning_rate": 6.693845172095683e-05, "loss": 0.0687, "step": 42050 }, { "grad_norm": 0.4299551248550415, "learning_rate": 6.692289380075142e-05, "loss": 0.0461, "step": 42060 }, { "grad_norm": 0.47895821928977966, "learning_rate": 6.690733402988379e-05, "loss": 0.0475, "step": 42070 }, { "grad_norm": 0.4675147831439972, "learning_rate": 6.689177241005553e-05, "loss": 0.0514, "step": 42080 }, { "grad_norm": 0.5066433548927307, "learning_rate": 6.687620894296844e-05, "loss": 0.0509, "step": 42090 }, { "grad_norm": 0.5148975253105164, "learning_rate": 6.686064363032451e-05, "loss": 0.0558, "step": 42100 }, { "grad_norm": 0.9087363481521606, "learning_rate": 6.684507647382596e-05, "loss": 0.0495, "step": 42110 }, { "grad_norm": 0.6736709475517273, "learning_rate": 6.682950747517513e-05, "loss": 0.0542, "step": 42120 }, { "grad_norm": 0.5555005073547363, "learning_rate": 6.68139366360747e-05, "loss": 0.0527, "step": 42130 }, { "grad_norm": 0.41629791259765625, "learning_rate": 6.679836395822744e-05, "loss": 0.0487, "step": 42140 }, { "grad_norm": 0.5898157954216003, "learning_rate": 6.678278944333633e-05, "loss": 0.0428, "step": 42150 }, { "grad_norm": 1.3810147047042847, "learning_rate": 6.676721309310462e-05, "loss": 0.0463, "step": 42160 }, { "grad_norm": 0.5701326727867126, "learning_rate": 6.67516349092357e-05, "loss": 0.0547, "step": 42170 }, { "grad_norm": 0.4387252628803253, "learning_rate": 6.673605489343315e-05, "loss": 0.0401, "step": 42180 }, { "grad_norm": 0.377468079328537, "learning_rate": 6.672047304740081e-05, "loss": 0.0485, "step": 42190 }, { "grad_norm": 0.7386459708213806, "learning_rate": 6.670488937284268e-05, "loss": 0.0465, "step": 42200 }, { "grad_norm": 0.5496640801429749, "learning_rate": 6.668930387146294e-05, "loss": 0.0465, "step": 42210 }, { "grad_norm": 0.968454897403717, "learning_rate": 6.667371654496605e-05, "loss": 0.0468, "step": 42220 }, { "grad_norm": 0.5750849843025208, "learning_rate": 6.665812739505659e-05, "loss": 0.0512, "step": 42230 }, { "grad_norm": 0.5786619782447815, "learning_rate": 6.664253642343935e-05, "loss": 0.0543, "step": 42240 }, { "grad_norm": 0.699519693851471, "learning_rate": 6.662694363181935e-05, "loss": 0.0495, "step": 42250 }, { "grad_norm": 0.3683057427406311, "learning_rate": 6.66113490219018e-05, "loss": 0.0668, "step": 42260 }, { "grad_norm": 0.468872994184494, "learning_rate": 6.659575259539207e-05, "loss": 0.0521, "step": 42270 }, { "grad_norm": 0.5354913473129272, "learning_rate": 6.65801543539958e-05, "loss": 0.0552, "step": 42280 }, { "grad_norm": 0.5916379690170288, "learning_rate": 6.65645542994188e-05, "loss": 0.0484, "step": 42290 }, { "grad_norm": 0.45630329847335815, "learning_rate": 6.654895243336702e-05, "loss": 0.0455, "step": 42300 }, { "grad_norm": 0.5594344735145569, "learning_rate": 6.65333487575467e-05, "loss": 0.0371, "step": 42310 }, { "grad_norm": 0.7005147337913513, "learning_rate": 6.651774327366422e-05, "loss": 0.0494, "step": 42320 }, { "grad_norm": 0.5604430437088013, "learning_rate": 6.650213598342613e-05, "loss": 0.0501, "step": 42330 }, { "grad_norm": 0.3085300922393799, "learning_rate": 6.648652688853931e-05, "loss": 0.0558, "step": 42340 }, { "grad_norm": 0.5448929667472839, "learning_rate": 6.647091599071066e-05, "loss": 0.0532, "step": 42350 }, { "grad_norm": 0.4294745624065399, "learning_rate": 6.645530329164743e-05, "loss": 0.0429, "step": 42360 }, { "grad_norm": 0.4375154674053192, "learning_rate": 6.643968879305694e-05, "loss": 0.0491, "step": 42370 }, { "grad_norm": 0.43474212288856506, "learning_rate": 6.642407249664683e-05, "loss": 0.0469, "step": 42380 }, { "grad_norm": 0.5069403648376465, "learning_rate": 6.640845440412483e-05, "loss": 0.0649, "step": 42390 }, { "grad_norm": 0.5864128470420837, "learning_rate": 6.639283451719893e-05, "loss": 0.0471, "step": 42400 }, { "grad_norm": 0.5028694868087769, "learning_rate": 6.637721283757729e-05, "loss": 0.0361, "step": 42410 }, { "grad_norm": 0.4001007080078125, "learning_rate": 6.636158936696828e-05, "loss": 0.0478, "step": 42420 }, { "grad_norm": 1.0417555570602417, "learning_rate": 6.634596410708047e-05, "loss": 0.0615, "step": 42430 }, { "grad_norm": 0.5204204320907593, "learning_rate": 6.633033705962257e-05, "loss": 0.0496, "step": 42440 }, { "grad_norm": 0.5774695873260498, "learning_rate": 6.631470822630359e-05, "loss": 0.0579, "step": 42450 }, { "grad_norm": 1.1113884449005127, "learning_rate": 6.629907760883263e-05, "loss": 0.0528, "step": 42460 }, { "grad_norm": 0.4481617212295532, "learning_rate": 6.628344520891905e-05, "loss": 0.0444, "step": 42470 }, { "grad_norm": 0.37108275294303894, "learning_rate": 6.62678110282724e-05, "loss": 0.0401, "step": 42480 }, { "grad_norm": 0.6051341891288757, "learning_rate": 6.625217506860237e-05, "loss": 0.058, "step": 42490 }, { "grad_norm": 0.3895435035228729, "learning_rate": 6.623653733161892e-05, "loss": 0.0621, "step": 42500 }, { "grad_norm": 0.4779379665851593, "learning_rate": 6.622089781903217e-05, "loss": 0.0532, "step": 42510 }, { "grad_norm": 0.5421006083488464, "learning_rate": 6.620525653255241e-05, "loss": 0.049, "step": 42520 }, { "grad_norm": 0.48496708273887634, "learning_rate": 6.618961347389015e-05, "loss": 0.0489, "step": 42530 }, { "grad_norm": 0.5768619775772095, "learning_rate": 6.617396864475613e-05, "loss": 0.0548, "step": 42540 }, { "grad_norm": 0.5196482539176941, "learning_rate": 6.61583220468612e-05, "loss": 0.0457, "step": 42550 }, { "grad_norm": 0.501318097114563, "learning_rate": 6.614267368191645e-05, "loss": 0.0486, "step": 42560 }, { "grad_norm": 0.4739047586917877, "learning_rate": 6.61270235516332e-05, "loss": 0.0566, "step": 42570 }, { "grad_norm": 0.5166022181510925, "learning_rate": 6.611137165772287e-05, "loss": 0.0384, "step": 42580 }, { "grad_norm": 0.37898319959640503, "learning_rate": 6.60957180018972e-05, "loss": 0.05, "step": 42590 }, { "grad_norm": 0.46922194957733154, "learning_rate": 6.608006258586797e-05, "loss": 0.0428, "step": 42600 }, { "grad_norm": 0.46261832118034363, "learning_rate": 6.606440541134731e-05, "loss": 0.0542, "step": 42610 }, { "grad_norm": 0.7014391422271729, "learning_rate": 6.60487464800474e-05, "loss": 0.0621, "step": 42620 }, { "grad_norm": 0.6835283041000366, "learning_rate": 6.603308579368071e-05, "loss": 0.0552, "step": 42630 }, { "grad_norm": 0.628324031829834, "learning_rate": 6.601742335395987e-05, "loss": 0.0549, "step": 42640 }, { "grad_norm": 0.5211963653564453, "learning_rate": 6.600175916259769e-05, "loss": 0.056, "step": 42650 }, { "grad_norm": 0.552193284034729, "learning_rate": 6.598609322130718e-05, "loss": 0.0432, "step": 42660 }, { "grad_norm": 0.5840925574302673, "learning_rate": 6.597042553180155e-05, "loss": 0.042, "step": 42670 }, { "grad_norm": 0.4066926836967468, "learning_rate": 6.59547560957942e-05, "loss": 0.0452, "step": 42680 }, { "grad_norm": 0.4488692879676819, "learning_rate": 6.59390849149987e-05, "loss": 0.0586, "step": 42690 }, { "grad_norm": 0.8345471024513245, "learning_rate": 6.592341199112886e-05, "loss": 0.0438, "step": 42700 }, { "grad_norm": 0.4672270119190216, "learning_rate": 6.590773732589861e-05, "loss": 0.0528, "step": 42710 }, { "grad_norm": 0.4717266857624054, "learning_rate": 6.589206092102213e-05, "loss": 0.0619, "step": 42720 }, { "grad_norm": 0.4705473482608795, "learning_rate": 6.587638277821375e-05, "loss": 0.0432, "step": 42730 }, { "grad_norm": 0.3750990331172943, "learning_rate": 6.586070289918803e-05, "loss": 0.0513, "step": 42740 }, { "grad_norm": 0.5390259623527527, "learning_rate": 6.584502128565968e-05, "loss": 0.0542, "step": 42750 }, { "grad_norm": 0.37747687101364136, "learning_rate": 6.582933793934363e-05, "loss": 0.0475, "step": 42760 }, { "grad_norm": 0.38890281319618225, "learning_rate": 6.581365286195498e-05, "loss": 0.0491, "step": 42770 }, { "grad_norm": 0.373991996049881, "learning_rate": 6.579796605520903e-05, "loss": 0.0495, "step": 42780 }, { "grad_norm": 0.7258316278457642, "learning_rate": 6.578227752082127e-05, "loss": 0.0588, "step": 42790 }, { "grad_norm": 0.47517696022987366, "learning_rate": 6.576658726050735e-05, "loss": 0.0424, "step": 42800 }, { "grad_norm": 0.5364022254943848, "learning_rate": 6.575089527598316e-05, "loss": 0.0427, "step": 42810 }, { "grad_norm": 0.3515520691871643, "learning_rate": 6.573520156896476e-05, "loss": 0.0579, "step": 42820 }, { "grad_norm": 0.47619563341140747, "learning_rate": 6.571950614116835e-05, "loss": 0.0498, "step": 42830 }, { "grad_norm": 0.7493143081665039, "learning_rate": 6.570380899431039e-05, "loss": 0.0707, "step": 42840 }, { "grad_norm": 0.6639305949211121, "learning_rate": 6.568811013010749e-05, "loss": 0.0517, "step": 42850 }, { "grad_norm": 0.456342488527298, "learning_rate": 6.567240955027645e-05, "loss": 0.0545, "step": 42860 }, { "grad_norm": 0.6778859496116638, "learning_rate": 6.565670725653427e-05, "loss": 0.0364, "step": 42870 }, { "grad_norm": 0.36943310499191284, "learning_rate": 6.564100325059812e-05, "loss": 0.0419, "step": 42880 }, { "grad_norm": 0.4637584388256073, "learning_rate": 6.562529753418538e-05, "loss": 0.0455, "step": 42890 }, { "grad_norm": 0.4078487157821655, "learning_rate": 6.56095901090136e-05, "loss": 0.045, "step": 42900 }, { "grad_norm": 0.8642622828483582, "learning_rate": 6.55938809768005e-05, "loss": 0.0438, "step": 42910 }, { "grad_norm": 0.9430252313613892, "learning_rate": 6.557817013926402e-05, "loss": 0.0557, "step": 42920 }, { "grad_norm": 0.4801388084888458, "learning_rate": 6.556245759812227e-05, "loss": 0.0548, "step": 42930 }, { "grad_norm": 0.5629636645317078, "learning_rate": 6.554674335509357e-05, "loss": 0.0517, "step": 42940 }, { "grad_norm": 0.4408738613128662, "learning_rate": 6.553102741189638e-05, "loss": 0.0418, "step": 42950 }, { "grad_norm": 0.35554900765419006, "learning_rate": 6.55153097702494e-05, "loss": 0.0527, "step": 42960 }, { "grad_norm": 0.4924887418746948, "learning_rate": 6.549959043187146e-05, "loss": 0.0475, "step": 42970 }, { "grad_norm": 0.7093451023101807, "learning_rate": 6.54838693984816e-05, "loss": 0.062, "step": 42980 }, { "grad_norm": 0.9844540953636169, "learning_rate": 6.546814667179909e-05, "loss": 0.0536, "step": 42990 }, { "grad_norm": 0.44746407866477966, "learning_rate": 6.545242225354328e-05, "loss": 0.0524, "step": 43000 }, { "grad_norm": 0.5048463344573975, "learning_rate": 6.543669614543383e-05, "loss": 0.0479, "step": 43010 }, { "grad_norm": 0.6613062024116516, "learning_rate": 6.542096834919049e-05, "loss": 0.053, "step": 43020 }, { "grad_norm": 0.6799914836883545, "learning_rate": 6.540523886653324e-05, "loss": 0.0516, "step": 43030 }, { "grad_norm": 0.4071614444255829, "learning_rate": 6.538950769918222e-05, "loss": 0.0554, "step": 43040 }, { "grad_norm": 0.48715388774871826, "learning_rate": 6.537377484885779e-05, "loss": 0.044, "step": 43050 }, { "grad_norm": 0.5338031649589539, "learning_rate": 6.535804031728045e-05, "loss": 0.0471, "step": 43060 }, { "grad_norm": 0.4605352580547333, "learning_rate": 6.534230410617092e-05, "loss": 0.0498, "step": 43070 }, { "grad_norm": 0.8949229121208191, "learning_rate": 6.532656621725007e-05, "loss": 0.0512, "step": 43080 }, { "grad_norm": 0.5130307674407959, "learning_rate": 6.531082665223898e-05, "loss": 0.0602, "step": 43090 }, { "grad_norm": 0.46990343928337097, "learning_rate": 6.529508541285889e-05, "loss": 0.0506, "step": 43100 }, { "grad_norm": 0.7646616697311401, "learning_rate": 6.527934250083128e-05, "loss": 0.0593, "step": 43110 }, { "grad_norm": 0.7950694561004639, "learning_rate": 6.526359791787775e-05, "loss": 0.0566, "step": 43120 }, { "grad_norm": 0.8841094970703125, "learning_rate": 6.524785166572009e-05, "loss": 0.0472, "step": 43130 }, { "grad_norm": 0.5785642266273499, "learning_rate": 6.523210374608029e-05, "loss": 0.0563, "step": 43140 }, { "grad_norm": 0.36692529916763306, "learning_rate": 6.521635416068054e-05, "loss": 0.0474, "step": 43150 }, { "grad_norm": 0.4487832486629486, "learning_rate": 6.520060291124317e-05, "loss": 0.0686, "step": 43160 }, { "grad_norm": 0.31952545046806335, "learning_rate": 6.518484999949072e-05, "loss": 0.0468, "step": 43170 }, { "grad_norm": 0.4639260768890381, "learning_rate": 6.516909542714591e-05, "loss": 0.0513, "step": 43180 }, { "grad_norm": 0.5997328162193298, "learning_rate": 6.515333919593161e-05, "loss": 0.0425, "step": 43190 }, { "grad_norm": 0.5552041530609131, "learning_rate": 6.513758130757094e-05, "loss": 0.0449, "step": 43200 }, { "grad_norm": 0.7446303963661194, "learning_rate": 6.512182176378713e-05, "loss": 0.051, "step": 43210 }, { "grad_norm": 0.5814357995986938, "learning_rate": 6.510606056630362e-05, "loss": 0.0458, "step": 43220 }, { "grad_norm": 0.42522504925727844, "learning_rate": 6.509029771684405e-05, "loss": 0.0466, "step": 43230 }, { "grad_norm": 0.5634058117866516, "learning_rate": 6.50745332171322e-05, "loss": 0.0502, "step": 43240 }, { "grad_norm": 0.43632176518440247, "learning_rate": 6.505876706889207e-05, "loss": 0.0603, "step": 43250 }, { "grad_norm": 0.4966789186000824, "learning_rate": 6.504299927384781e-05, "loss": 0.0462, "step": 43260 }, { "grad_norm": 0.6193351149559021, "learning_rate": 6.502722983372378e-05, "loss": 0.044, "step": 43270 }, { "grad_norm": 0.6525517702102661, "learning_rate": 6.501145875024446e-05, "loss": 0.0564, "step": 43280 }, { "grad_norm": 0.5693353414535522, "learning_rate": 6.499568602513459e-05, "loss": 0.0541, "step": 43290 }, { "grad_norm": 0.4196535050868988, "learning_rate": 6.497991166011903e-05, "loss": 0.0477, "step": 43300 }, { "grad_norm": 0.48655465245246887, "learning_rate": 6.49641356569229e-05, "loss": 0.0449, "step": 43310 }, { "grad_norm": 0.26737329363822937, "learning_rate": 6.494835801727135e-05, "loss": 0.0445, "step": 43320 }, { "grad_norm": 0.508700966835022, "learning_rate": 6.493257874288986e-05, "loss": 0.0567, "step": 43330 }, { "grad_norm": 0.33160847425460815, "learning_rate": 6.491679783550399e-05, "loss": 0.0501, "step": 43340 }, { "grad_norm": 0.4103374481201172, "learning_rate": 6.490101529683957e-05, "loss": 0.0456, "step": 43350 }, { "grad_norm": 0.3176514208316803, "learning_rate": 6.488523112862248e-05, "loss": 0.0494, "step": 43360 }, { "grad_norm": 0.3305668830871582, "learning_rate": 6.486944533257891e-05, "loss": 0.0473, "step": 43370 }, { "grad_norm": 0.5772157311439514, "learning_rate": 6.485365791043517e-05, "loss": 0.0564, "step": 43380 }, { "grad_norm": 0.4352125823497772, "learning_rate": 6.483786886391772e-05, "loss": 0.0509, "step": 43390 }, { "grad_norm": 0.3274185061454773, "learning_rate": 6.482207819475323e-05, "loss": 0.0419, "step": 43400 }, { "grad_norm": 0.46036824584007263, "learning_rate": 6.480628590466857e-05, "loss": 0.0432, "step": 43410 }, { "grad_norm": 0.5360816121101379, "learning_rate": 6.479049199539074e-05, "loss": 0.0452, "step": 43420 }, { "grad_norm": 0.5385856628417969, "learning_rate": 6.477469646864693e-05, "loss": 0.0551, "step": 43430 }, { "grad_norm": 0.81321781873703, "learning_rate": 6.475889932616454e-05, "loss": 0.0578, "step": 43440 }, { "grad_norm": 0.5335119962692261, "learning_rate": 6.474310056967111e-05, "loss": 0.0543, "step": 43450 }, { "grad_norm": 0.6082566976547241, "learning_rate": 6.472730020089437e-05, "loss": 0.0637, "step": 43460 }, { "grad_norm": 0.3934541940689087, "learning_rate": 6.471149822156222e-05, "loss": 0.0533, "step": 43470 }, { "grad_norm": 0.588323175907135, "learning_rate": 6.469569463340275e-05, "loss": 0.0472, "step": 43480 }, { "grad_norm": 0.41259896755218506, "learning_rate": 6.467988943814418e-05, "loss": 0.0466, "step": 43490 }, { "grad_norm": 0.45284023880958557, "learning_rate": 6.4664082637515e-05, "loss": 0.0407, "step": 43500 }, { "grad_norm": 0.5493537783622742, "learning_rate": 6.464827423324377e-05, "loss": 0.0475, "step": 43510 }, { "grad_norm": 0.4704609513282776, "learning_rate": 6.463246422705931e-05, "loss": 0.0446, "step": 43520 }, { "grad_norm": 0.5331157445907593, "learning_rate": 6.461665262069055e-05, "loss": 0.0556, "step": 43530 }, { "grad_norm": 0.5149337649345398, "learning_rate": 6.460083941586665e-05, "loss": 0.0481, "step": 43540 }, { "grad_norm": 0.521420419216156, "learning_rate": 6.45850246143169e-05, "loss": 0.0498, "step": 43550 }, { "grad_norm": 0.5796550512313843, "learning_rate": 6.456920821777077e-05, "loss": 0.0538, "step": 43560 }, { "grad_norm": 0.4699636399745941, "learning_rate": 6.455339022795796e-05, "loss": 0.0621, "step": 43570 }, { "grad_norm": 0.38048404455184937, "learning_rate": 6.453757064660825e-05, "loss": 0.0455, "step": 43580 }, { "grad_norm": 0.4069731831550598, "learning_rate": 6.452174947545169e-05, "loss": 0.0402, "step": 43590 }, { "grad_norm": 0.42199912667274475, "learning_rate": 6.450592671621842e-05, "loss": 0.0429, "step": 43600 }, { "grad_norm": 0.6135024428367615, "learning_rate": 6.449010237063882e-05, "loss": 0.0504, "step": 43610 }, { "grad_norm": 0.4992164671421051, "learning_rate": 6.447427644044343e-05, "loss": 0.0539, "step": 43620 }, { "grad_norm": 0.3692421317100525, "learning_rate": 6.44584489273629e-05, "loss": 0.047, "step": 43630 }, { "grad_norm": 0.6016609072685242, "learning_rate": 6.444261983312817e-05, "loss": 0.0497, "step": 43640 }, { "grad_norm": 0.38327354192733765, "learning_rate": 6.442678915947023e-05, "loss": 0.0501, "step": 43650 }, { "grad_norm": 0.6448785066604614, "learning_rate": 6.441095690812032e-05, "loss": 0.0496, "step": 43660 }, { "grad_norm": 0.5326748490333557, "learning_rate": 6.439512308080982e-05, "loss": 0.0471, "step": 43670 }, { "grad_norm": 0.3902905285358429, "learning_rate": 6.43792876792703e-05, "loss": 0.051, "step": 43680 }, { "grad_norm": 0.5442156195640564, "learning_rate": 6.43634507052335e-05, "loss": 0.0517, "step": 43690 }, { "grad_norm": 0.5632613301277161, "learning_rate": 6.434761216043135e-05, "loss": 0.0501, "step": 43700 }, { "grad_norm": 0.4871600866317749, "learning_rate": 6.433177204659589e-05, "loss": 0.049, "step": 43710 }, { "grad_norm": 0.43331220746040344, "learning_rate": 6.431593036545938e-05, "loss": 0.0479, "step": 43720 }, { "grad_norm": 0.6746350526809692, "learning_rate": 6.430008711875426e-05, "loss": 0.0384, "step": 43730 }, { "grad_norm": 0.7918664813041687, "learning_rate": 6.42842423082131e-05, "loss": 0.051, "step": 43740 }, { "grad_norm": 0.6329689025878906, "learning_rate": 6.426839593556868e-05, "loss": 0.0422, "step": 43750 }, { "grad_norm": 0.35188138484954834, "learning_rate": 6.425254800255392e-05, "loss": 0.0464, "step": 43760 }, { "grad_norm": 0.5532768964767456, "learning_rate": 6.423669851090196e-05, "loss": 0.0484, "step": 43770 }, { "grad_norm": 0.5753403306007385, "learning_rate": 6.422084746234604e-05, "loss": 0.0404, "step": 43780 }, { "grad_norm": 0.5598087906837463, "learning_rate": 6.420499485861965e-05, "loss": 0.0338, "step": 43790 }, { "grad_norm": 0.3420179486274719, "learning_rate": 6.418914070145636e-05, "loss": 0.0567, "step": 43800 }, { "grad_norm": 0.5023845434188843, "learning_rate": 6.417328499258996e-05, "loss": 0.0398, "step": 43810 }, { "grad_norm": 0.583392322063446, "learning_rate": 6.415742773375446e-05, "loss": 0.0444, "step": 43820 }, { "grad_norm": 0.31748586893081665, "learning_rate": 6.414156892668393e-05, "loss": 0.05, "step": 43830 }, { "grad_norm": 0.7392582297325134, "learning_rate": 6.412570857311267e-05, "loss": 0.0523, "step": 43840 }, { "grad_norm": 0.435412734746933, "learning_rate": 6.410984667477518e-05, "loss": 0.0515, "step": 43850 }, { "grad_norm": 0.5562236905097961, "learning_rate": 6.409398323340607e-05, "loss": 0.0438, "step": 43860 }, { "grad_norm": 0.6359451413154602, "learning_rate": 6.407811825074012e-05, "loss": 0.0534, "step": 43870 }, { "grad_norm": 0.49582698941230774, "learning_rate": 6.406225172851234e-05, "loss": 0.0456, "step": 43880 }, { "grad_norm": 0.5021716356277466, "learning_rate": 6.404638366845786e-05, "loss": 0.0522, "step": 43890 }, { "grad_norm": 0.5586097836494446, "learning_rate": 6.403051407231196e-05, "loss": 0.0572, "step": 43900 }, { "grad_norm": 0.5955579280853271, "learning_rate": 6.401464294181016e-05, "loss": 0.047, "step": 43910 }, { "grad_norm": 0.5501436591148376, "learning_rate": 6.399877027868808e-05, "loss": 0.0477, "step": 43920 }, { "grad_norm": 0.5752478837966919, "learning_rate": 6.39828960846815e-05, "loss": 0.0516, "step": 43930 }, { "grad_norm": 0.3692147433757782, "learning_rate": 6.396702036152644e-05, "loss": 0.0393, "step": 43940 }, { "grad_norm": 0.3278125822544098, "learning_rate": 6.395114311095905e-05, "loss": 0.061, "step": 43950 }, { "grad_norm": 0.4271550476551056, "learning_rate": 6.393526433471559e-05, "loss": 0.0483, "step": 43960 }, { "grad_norm": 0.3924233615398407, "learning_rate": 6.391938403453262e-05, "loss": 0.0645, "step": 43970 }, { "grad_norm": 0.5874158143997192, "learning_rate": 6.390350221214671e-05, "loss": 0.055, "step": 43980 }, { "grad_norm": 0.4950332045555115, "learning_rate": 6.38876188692947e-05, "loss": 0.0588, "step": 43990 }, { "grad_norm": 1.68781578540802, "learning_rate": 6.387173400771358e-05, "loss": 0.0578, "step": 44000 }, { "grad_norm": 0.3891233503818512, "learning_rate": 6.385584762914047e-05, "loss": 0.0499, "step": 44010 }, { "grad_norm": 0.35734522342681885, "learning_rate": 6.38399597353127e-05, "loss": 0.0507, "step": 44020 }, { "grad_norm": 0.532283365726471, "learning_rate": 6.382407032796775e-05, "loss": 0.0453, "step": 44030 }, { "grad_norm": 0.36085906624794006, "learning_rate": 6.380817940884325e-05, "loss": 0.042, "step": 44040 }, { "grad_norm": 1.4545197486877441, "learning_rate": 6.379228697967702e-05, "loss": 0.048, "step": 44050 }, { "grad_norm": 0.366818904876709, "learning_rate": 6.3776393042207e-05, "loss": 0.0496, "step": 44060 }, { "grad_norm": 0.5772387981414795, "learning_rate": 6.376049759817137e-05, "loss": 0.0507, "step": 44070 }, { "grad_norm": 0.5512857437133789, "learning_rate": 6.37446006493084e-05, "loss": 0.0654, "step": 44080 }, { "grad_norm": 0.3916266858577728, "learning_rate": 6.372870219735656e-05, "loss": 0.0574, "step": 44090 }, { "grad_norm": 0.686968982219696, "learning_rate": 6.371280224405452e-05, "loss": 0.0547, "step": 44100 }, { "grad_norm": 0.5101865530014038, "learning_rate": 6.369690079114101e-05, "loss": 0.0458, "step": 44110 }, { "grad_norm": 0.551486074924469, "learning_rate": 6.368099784035504e-05, "loss": 0.0541, "step": 44120 }, { "grad_norm": 0.5036001801490784, "learning_rate": 6.366509339343572e-05, "loss": 0.0437, "step": 44130 }, { "grad_norm": 0.6075497269630432, "learning_rate": 6.364918745212232e-05, "loss": 0.0427, "step": 44140 }, { "grad_norm": 0.5412997603416443, "learning_rate": 6.363328001815431e-05, "loss": 0.0504, "step": 44150 }, { "grad_norm": 0.7693723440170288, "learning_rate": 6.361737109327128e-05, "loss": 0.0495, "step": 44160 }, { "grad_norm": 0.5429217219352722, "learning_rate": 6.360146067921301e-05, "loss": 0.0554, "step": 44170 }, { "grad_norm": 0.5169814825057983, "learning_rate": 6.35855487777195e-05, "loss": 0.0556, "step": 44180 }, { "grad_norm": 0.3965255320072174, "learning_rate": 6.356963539053076e-05, "loss": 0.0503, "step": 44190 }, { "grad_norm": 0.2983075976371765, "learning_rate": 6.355372051938711e-05, "loss": 0.0398, "step": 44200 }, { "grad_norm": 0.39574557542800903, "learning_rate": 6.353780416602894e-05, "loss": 0.0465, "step": 44210 }, { "grad_norm": 0.5442743897438049, "learning_rate": 6.352188633219689e-05, "loss": 0.0497, "step": 44220 }, { "grad_norm": 0.4582701027393341, "learning_rate": 6.350596701963166e-05, "loss": 0.038, "step": 44230 }, { "grad_norm": 0.5581745505332947, "learning_rate": 6.349004623007419e-05, "loss": 0.0434, "step": 44240 }, { "grad_norm": 0.4201377332210541, "learning_rate": 6.347412396526555e-05, "loss": 0.0432, "step": 44250 }, { "grad_norm": 0.7650616765022278, "learning_rate": 6.345820022694696e-05, "loss": 0.0626, "step": 44260 }, { "grad_norm": 1.0742862224578857, "learning_rate": 6.344227501685984e-05, "loss": 0.0596, "step": 44270 }, { "grad_norm": 0.5107795000076294, "learning_rate": 6.342634833674572e-05, "loss": 0.0501, "step": 44280 }, { "grad_norm": 0.7894645929336548, "learning_rate": 6.341042018834635e-05, "loss": 0.0636, "step": 44290 }, { "grad_norm": 0.5150954127311707, "learning_rate": 6.339449057340359e-05, "loss": 0.0531, "step": 44300 }, { "grad_norm": 0.3771496117115021, "learning_rate": 6.337855949365945e-05, "loss": 0.0511, "step": 44310 }, { "grad_norm": 0.9497362375259399, "learning_rate": 6.336262695085619e-05, "loss": 0.0538, "step": 44320 }, { "grad_norm": 0.39291051030158997, "learning_rate": 6.334669294673612e-05, "loss": 0.0494, "step": 44330 }, { "grad_norm": 0.6627101898193359, "learning_rate": 6.333075748304179e-05, "loss": 0.0542, "step": 44340 }, { "grad_norm": 0.6454407572746277, "learning_rate": 6.331482056151585e-05, "loss": 0.0541, "step": 44350 }, { "grad_norm": 0.6474065780639648, "learning_rate": 6.329888218390117e-05, "loss": 0.0548, "step": 44360 }, { "grad_norm": 0.5689963698387146, "learning_rate": 6.328294235194072e-05, "loss": 0.0453, "step": 44370 }, { "grad_norm": 0.48535245656967163, "learning_rate": 6.326700106737765e-05, "loss": 0.0504, "step": 44380 }, { "grad_norm": 0.38965103030204773, "learning_rate": 6.325105833195531e-05, "loss": 0.0464, "step": 44390 }, { "grad_norm": 0.792096734046936, "learning_rate": 6.323511414741715e-05, "loss": 0.0452, "step": 44400 }, { "grad_norm": 0.556726336479187, "learning_rate": 6.321916851550678e-05, "loss": 0.0606, "step": 44410 }, { "grad_norm": 0.9178048372268677, "learning_rate": 6.320322143796806e-05, "loss": 0.0621, "step": 44420 }, { "grad_norm": 0.8463513851165771, "learning_rate": 6.318727291654488e-05, "loss": 0.0459, "step": 44430 }, { "grad_norm": 0.5130673050880432, "learning_rate": 6.317132295298134e-05, "loss": 0.0551, "step": 44440 }, { "grad_norm": 0.48910728096961975, "learning_rate": 6.315537154902173e-05, "loss": 0.0467, "step": 44450 }, { "grad_norm": 0.6066758632659912, "learning_rate": 6.313941870641048e-05, "loss": 0.049, "step": 44460 }, { "grad_norm": 0.346860408782959, "learning_rate": 6.312346442689214e-05, "loss": 0.049, "step": 44470 }, { "grad_norm": 1.664687156677246, "learning_rate": 6.310750871221147e-05, "loss": 0.043, "step": 44480 }, { "grad_norm": 0.3919779062271118, "learning_rate": 6.309155156411335e-05, "loss": 0.0448, "step": 44490 }, { "grad_norm": 0.47342103719711304, "learning_rate": 6.307559298434284e-05, "loss": 0.0436, "step": 44500 }, { "grad_norm": 0.398485004901886, "learning_rate": 6.305963297464512e-05, "loss": 0.0511, "step": 44510 }, { "grad_norm": 0.7865027189254761, "learning_rate": 6.304367153676561e-05, "loss": 0.0543, "step": 44520 }, { "grad_norm": 0.6480133533477783, "learning_rate": 6.302770867244978e-05, "loss": 0.0499, "step": 44530 }, { "grad_norm": 0.4357796013355255, "learning_rate": 6.301174438344328e-05, "loss": 0.0403, "step": 44540 }, { "grad_norm": 0.5281748175621033, "learning_rate": 6.299577867149202e-05, "loss": 0.0544, "step": 44550 }, { "grad_norm": 0.4174513518810272, "learning_rate": 6.297981153834192e-05, "loss": 0.0462, "step": 44560 }, { "grad_norm": 0.43972641229629517, "learning_rate": 6.296384298573916e-05, "loss": 0.0554, "step": 44570 }, { "grad_norm": 0.4255842864513397, "learning_rate": 6.294787301543001e-05, "loss": 0.0711, "step": 44580 }, { "grad_norm": 0.5746822953224182, "learning_rate": 6.293190162916095e-05, "loss": 0.0464, "step": 44590 }, { "grad_norm": 0.5740575790405273, "learning_rate": 6.291592882867855e-05, "loss": 0.0424, "step": 44600 }, { "grad_norm": 0.43472620844841003, "learning_rate": 6.28999546157296e-05, "loss": 0.039, "step": 44610 }, { "grad_norm": 0.4020138084888458, "learning_rate": 6.288397899206102e-05, "loss": 0.0431, "step": 44620 }, { "grad_norm": 0.7131125330924988, "learning_rate": 6.286800195941984e-05, "loss": 0.0554, "step": 44630 }, { "grad_norm": 0.42084622383117676, "learning_rate": 6.285202351955334e-05, "loss": 0.0432, "step": 44640 }, { "grad_norm": 0.5851677656173706, "learning_rate": 6.283604367420887e-05, "loss": 0.0563, "step": 44650 }, { "grad_norm": 0.4258369505405426, "learning_rate": 6.282006242513394e-05, "loss": 0.0552, "step": 44660 }, { "grad_norm": 0.5796555876731873, "learning_rate": 6.280407977407628e-05, "loss": 0.0547, "step": 44670 }, { "grad_norm": 0.5114135146141052, "learning_rate": 6.27880957227837e-05, "loss": 0.0464, "step": 44680 }, { "grad_norm": 0.5492531061172485, "learning_rate": 6.27721102730042e-05, "loss": 0.0508, "step": 44690 }, { "grad_norm": 0.689143717288971, "learning_rate": 6.275612342648592e-05, "loss": 0.053, "step": 44700 }, { "grad_norm": 0.5347728729248047, "learning_rate": 6.274013518497716e-05, "loss": 0.0511, "step": 44710 }, { "grad_norm": 0.4949626326560974, "learning_rate": 6.272414555022636e-05, "loss": 0.0413, "step": 44720 }, { "grad_norm": 0.6814123392105103, "learning_rate": 6.270815452398215e-05, "loss": 0.0445, "step": 44730 }, { "grad_norm": 0.8532877564430237, "learning_rate": 6.269216210799326e-05, "loss": 0.0443, "step": 44740 }, { "grad_norm": 0.5576585531234741, "learning_rate": 6.26761683040086e-05, "loss": 0.0513, "step": 44750 }, { "grad_norm": 0.36477604508399963, "learning_rate": 6.266017311377723e-05, "loss": 0.0468, "step": 44760 }, { "grad_norm": 0.3414878249168396, "learning_rate": 6.264417653904839e-05, "loss": 0.0564, "step": 44770 }, { "grad_norm": 0.5818560719490051, "learning_rate": 6.262817858157139e-05, "loss": 0.059, "step": 44780 }, { "grad_norm": 0.5940812230110168, "learning_rate": 6.261217924309576e-05, "loss": 0.0714, "step": 44790 }, { "grad_norm": 0.4189929962158203, "learning_rate": 6.259617852537118e-05, "loss": 0.0515, "step": 44800 }, { "grad_norm": 0.5944344997406006, "learning_rate": 6.258017643014747e-05, "loss": 0.0502, "step": 44810 }, { "grad_norm": 0.6301035284996033, "learning_rate": 6.256417295917456e-05, "loss": 0.0553, "step": 44820 }, { "grad_norm": 0.4342222809791565, "learning_rate": 6.254816811420258e-05, "loss": 0.0594, "step": 44830 }, { "grad_norm": 0.7005230188369751, "learning_rate": 6.253216189698183e-05, "loss": 0.0474, "step": 44840 }, { "grad_norm": 0.24435517191886902, "learning_rate": 6.251615430926267e-05, "loss": 0.039, "step": 44850 }, { "grad_norm": 0.4068748652935028, "learning_rate": 6.25001453527957e-05, "loss": 0.0417, "step": 44860 }, { "grad_norm": 0.4160313904285431, "learning_rate": 6.248413502933164e-05, "loss": 0.0528, "step": 44870 }, { "grad_norm": 0.4673183262348175, "learning_rate": 6.246812334062133e-05, "loss": 0.046, "step": 44880 }, { "grad_norm": 0.7335435152053833, "learning_rate": 6.245211028841579e-05, "loss": 0.0557, "step": 44890 }, { "grad_norm": 0.3904763460159302, "learning_rate": 6.24360958744662e-05, "loss": 0.0468, "step": 44900 }, { "grad_norm": 0.5813251733779907, "learning_rate": 6.242008010052387e-05, "loss": 0.0561, "step": 44910 }, { "grad_norm": 0.5252216458320618, "learning_rate": 6.240406296834024e-05, "loss": 0.0508, "step": 44920 }, { "grad_norm": 0.36944207549095154, "learning_rate": 6.238804447966694e-05, "loss": 0.0451, "step": 44930 }, { "grad_norm": 0.36473768949508667, "learning_rate": 6.237202463625573e-05, "loss": 0.0392, "step": 44940 }, { "grad_norm": 0.35276854038238525, "learning_rate": 6.235600343985848e-05, "loss": 0.0446, "step": 44950 }, { "grad_norm": 0.6196567416191101, "learning_rate": 6.233998089222729e-05, "loss": 0.0433, "step": 44960 }, { "grad_norm": 0.39528512954711914, "learning_rate": 6.232395699511433e-05, "loss": 0.0535, "step": 44970 }, { "grad_norm": 0.6031641364097595, "learning_rate": 6.230793175027199e-05, "loss": 0.0531, "step": 44980 }, { "grad_norm": 0.6508851647377014, "learning_rate": 6.22919051594527e-05, "loss": 0.0461, "step": 44990 }, { "grad_norm": 1.262271523475647, "learning_rate": 6.227587722440917e-05, "loss": 0.0558, "step": 45000 }, { "grad_norm": 0.42018410563468933, "learning_rate": 6.225984794689414e-05, "loss": 0.0307, "step": 45010 }, { "grad_norm": 0.9161219596862793, "learning_rate": 6.22438173286606e-05, "loss": 0.0644, "step": 45020 }, { "grad_norm": 0.5527089238166809, "learning_rate": 6.222778537146159e-05, "loss": 0.0453, "step": 45030 }, { "grad_norm": 0.4869990646839142, "learning_rate": 6.221175207705037e-05, "loss": 0.0477, "step": 45040 }, { "grad_norm": 0.7696365714073181, "learning_rate": 6.219571744718027e-05, "loss": 0.0488, "step": 45050 }, { "grad_norm": 0.3902416527271271, "learning_rate": 6.217968148360487e-05, "loss": 0.0436, "step": 45060 }, { "grad_norm": 0.6559630036354065, "learning_rate": 6.216364418807782e-05, "loss": 0.051, "step": 45070 }, { "grad_norm": 0.5586362481117249, "learning_rate": 6.214760556235292e-05, "loss": 0.0406, "step": 45080 }, { "grad_norm": 0.5108618140220642, "learning_rate": 6.213156560818414e-05, "loss": 0.0402, "step": 45090 }, { "grad_norm": 0.5822756886482239, "learning_rate": 6.211552432732559e-05, "loss": 0.0534, "step": 45100 }, { "grad_norm": 0.5466113686561584, "learning_rate": 6.20994817215315e-05, "loss": 0.053, "step": 45110 }, { "grad_norm": 0.5108882784843445, "learning_rate": 6.208343779255628e-05, "loss": 0.0526, "step": 45120 }, { "grad_norm": 0.6619426608085632, "learning_rate": 6.206739254215449e-05, "loss": 0.0549, "step": 45130 }, { "grad_norm": 0.7086274027824402, "learning_rate": 6.205134597208077e-05, "loss": 0.0631, "step": 45140 }, { "grad_norm": 0.4382135570049286, "learning_rate": 6.203529808409e-05, "loss": 0.0484, "step": 45150 }, { "grad_norm": 0.45332974195480347, "learning_rate": 6.201924887993712e-05, "loss": 0.0537, "step": 45160 }, { "grad_norm": 0.6090180277824402, "learning_rate": 6.200319836137725e-05, "loss": 0.0521, "step": 45170 }, { "grad_norm": 0.4363826811313629, "learning_rate": 6.198714653016565e-05, "loss": 0.0451, "step": 45180 }, { "grad_norm": 0.44236963987350464, "learning_rate": 6.197109338805774e-05, "loss": 0.0497, "step": 45190 }, { "grad_norm": 0.5662668943405151, "learning_rate": 6.195503893680903e-05, "loss": 0.0491, "step": 45200 }, { "grad_norm": 0.4218870997428894, "learning_rate": 6.193898317817524e-05, "loss": 0.0473, "step": 45210 }, { "grad_norm": 1.397348165512085, "learning_rate": 6.192292611391221e-05, "loss": 0.0485, "step": 45220 }, { "grad_norm": 0.3450496196746826, "learning_rate": 6.190686774577591e-05, "loss": 0.0461, "step": 45230 }, { "grad_norm": 0.48045846819877625, "learning_rate": 6.189080807552245e-05, "loss": 0.0563, "step": 45240 }, { "grad_norm": 0.7606565952301025, "learning_rate": 6.187474710490809e-05, "loss": 0.0531, "step": 45250 }, { "grad_norm": 0.43658995628356934, "learning_rate": 6.185868483568926e-05, "loss": 0.0481, "step": 45260 }, { "grad_norm": 0.5300836563110352, "learning_rate": 6.184262126962245e-05, "loss": 0.056, "step": 45270 }, { "grad_norm": 0.5196239948272705, "learning_rate": 6.182655640846442e-05, "loss": 0.0333, "step": 45280 }, { "grad_norm": 0.4556358754634857, "learning_rate": 6.181049025397196e-05, "loss": 0.0479, "step": 45290 }, { "grad_norm": 0.4207109212875366, "learning_rate": 6.179442280790202e-05, "loss": 0.0537, "step": 45300 }, { "grad_norm": 0.47373202443122864, "learning_rate": 6.177835407201174e-05, "loss": 0.0532, "step": 45310 }, { "grad_norm": 0.5810307264328003, "learning_rate": 6.176228404805839e-05, "loss": 0.0493, "step": 45320 }, { "grad_norm": 0.6391007900238037, "learning_rate": 6.174621273779932e-05, "loss": 0.0495, "step": 45330 }, { "grad_norm": 0.4942755401134491, "learning_rate": 6.17301401429921e-05, "loss": 0.0431, "step": 45340 }, { "grad_norm": 0.34983718395233154, "learning_rate": 6.171406626539436e-05, "loss": 0.0534, "step": 45350 }, { "grad_norm": 0.7228038907051086, "learning_rate": 6.169799110676398e-05, "loss": 0.0635, "step": 45360 }, { "grad_norm": 0.45146462321281433, "learning_rate": 6.168191466885885e-05, "loss": 0.0448, "step": 45370 }, { "grad_norm": 0.7430742979049683, "learning_rate": 6.166583695343711e-05, "loss": 0.0525, "step": 45380 }, { "grad_norm": 0.35402339696884155, "learning_rate": 6.164975796225698e-05, "loss": 0.0566, "step": 45390 }, { "grad_norm": 0.39977163076400757, "learning_rate": 6.163367769707683e-05, "loss": 0.0398, "step": 45400 }, { "grad_norm": 0.7149611711502075, "learning_rate": 6.161759615965519e-05, "loss": 0.05, "step": 45410 }, { "grad_norm": 0.41038671135902405, "learning_rate": 6.16015133517507e-05, "loss": 0.0498, "step": 45420 }, { "grad_norm": 0.45095178484916687, "learning_rate": 6.158542927512214e-05, "loss": 0.0397, "step": 45430 }, { "grad_norm": 0.5042413473129272, "learning_rate": 6.156934393152846e-05, "loss": 0.0455, "step": 45440 }, { "grad_norm": 0.4843246638774872, "learning_rate": 6.15532573227287e-05, "loss": 0.0616, "step": 45450 }, { "grad_norm": 0.49700555205345154, "learning_rate": 6.153716945048212e-05, "loss": 0.0608, "step": 45460 }, { "grad_norm": 0.5536076426506042, "learning_rate": 6.152108031654802e-05, "loss": 0.0556, "step": 45470 }, { "grad_norm": 0.37857794761657715, "learning_rate": 6.15049899226859e-05, "loss": 0.0455, "step": 45480 }, { "grad_norm": 0.5192883014678955, "learning_rate": 6.148889827065537e-05, "loss": 0.048, "step": 45490 }, { "grad_norm": 0.42082253098487854, "learning_rate": 6.147280536221622e-05, "loss": 0.0511, "step": 45500 }, { "grad_norm": 0.8697906732559204, "learning_rate": 6.145671119912832e-05, "loss": 0.0472, "step": 45510 }, { "grad_norm": 0.5180875062942505, "learning_rate": 6.144061578315169e-05, "loss": 0.0489, "step": 45520 }, { "grad_norm": 0.40306395292282104, "learning_rate": 6.142451911604654e-05, "loss": 0.0509, "step": 45530 }, { "grad_norm": 0.5495306253433228, "learning_rate": 6.140842119957315e-05, "loss": 0.0512, "step": 45540 }, { "grad_norm": 0.6461553573608398, "learning_rate": 6.139232203549197e-05, "loss": 0.0472, "step": 45550 }, { "grad_norm": 0.5842080116271973, "learning_rate": 6.137622162556357e-05, "loss": 0.0554, "step": 45560 }, { "grad_norm": 0.3361794054508209, "learning_rate": 6.13601199715487e-05, "loss": 0.0384, "step": 45570 }, { "grad_norm": 0.4494541585445404, "learning_rate": 6.134401707520816e-05, "loss": 0.0503, "step": 45580 }, { "grad_norm": 0.45919185876846313, "learning_rate": 6.132791293830298e-05, "loss": 0.0405, "step": 45590 }, { "grad_norm": 0.3879046142101288, "learning_rate": 6.131180756259428e-05, "loss": 0.0477, "step": 45600 }, { "grad_norm": 0.366319864988327, "learning_rate": 6.129570094984331e-05, "loss": 0.0394, "step": 45610 }, { "grad_norm": 0.7418840527534485, "learning_rate": 6.127959310181145e-05, "loss": 0.0573, "step": 45620 }, { "grad_norm": 0.679800271987915, "learning_rate": 6.126348402026026e-05, "loss": 0.0511, "step": 45630 }, { "grad_norm": 0.571960985660553, "learning_rate": 6.12473737069514e-05, "loss": 0.041, "step": 45640 }, { "grad_norm": 0.5810232758522034, "learning_rate": 6.123126216364665e-05, "loss": 0.0498, "step": 45650 }, { "grad_norm": 0.5704169273376465, "learning_rate": 6.121514939210797e-05, "loss": 0.0343, "step": 45660 }, { "grad_norm": 0.5166617631912231, "learning_rate": 6.119903539409741e-05, "loss": 0.056, "step": 45670 }, { "grad_norm": 0.5457667112350464, "learning_rate": 6.118292017137716e-05, "loss": 0.0453, "step": 45680 }, { "grad_norm": 0.4903540313243866, "learning_rate": 6.116680372570959e-05, "loss": 0.0515, "step": 45690 }, { "grad_norm": 0.5223370790481567, "learning_rate": 6.115068605885713e-05, "loss": 0.0344, "step": 45700 }, { "grad_norm": 0.5448052287101746, "learning_rate": 6.113456717258243e-05, "loss": 0.0561, "step": 45710 }, { "grad_norm": 0.34912610054016113, "learning_rate": 6.11184470686482e-05, "loss": 0.0468, "step": 45720 }, { "grad_norm": 0.4979635775089264, "learning_rate": 6.110232574881731e-05, "loss": 0.044, "step": 45730 }, { "grad_norm": 1.7630329132080078, "learning_rate": 6.108620321485277e-05, "loss": 0.058, "step": 45740 }, { "grad_norm": 0.45556551218032837, "learning_rate": 6.107007946851773e-05, "loss": 0.0554, "step": 45750 }, { "grad_norm": 0.5107593536376953, "learning_rate": 6.105395451157542e-05, "loss": 0.0511, "step": 45760 }, { "grad_norm": 0.3323166072368622, "learning_rate": 6.103782834578928e-05, "loss": 0.0415, "step": 45770 }, { "grad_norm": 0.5581294894218445, "learning_rate": 6.102170097292281e-05, "loss": 0.0474, "step": 45780 }, { "grad_norm": 0.7383692264556885, "learning_rate": 6.1005572394739716e-05, "loss": 0.057, "step": 45790 }, { "grad_norm": 1.2137483358383179, "learning_rate": 6.0989442613003765e-05, "loss": 0.0432, "step": 45800 }, { "grad_norm": 0.47936785221099854, "learning_rate": 6.097331162947888e-05, "loss": 0.0451, "step": 45810 }, { "grad_norm": 1.9233535528182983, "learning_rate": 6.095717944592914e-05, "loss": 0.0525, "step": 45820 }, { "grad_norm": 0.7048292756080627, "learning_rate": 6.094104606411873e-05, "loss": 0.0556, "step": 45830 }, { "grad_norm": 1.5497735738754272, "learning_rate": 6.0924911485811966e-05, "loss": 0.0556, "step": 45840 }, { "grad_norm": 0.7185392379760742, "learning_rate": 6.090877571277331e-05, "loss": 0.0571, "step": 45850 }, { "grad_norm": 0.4267336130142212, "learning_rate": 6.089263874676736e-05, "loss": 0.0447, "step": 45860 }, { "grad_norm": 0.9695307612419128, "learning_rate": 6.0876500589558796e-05, "loss": 0.0502, "step": 45870 }, { "grad_norm": 1.1903709173202515, "learning_rate": 6.086036124291248e-05, "loss": 0.0512, "step": 45880 }, { "grad_norm": 0.4922686815261841, "learning_rate": 6.084422070859339e-05, "loss": 0.0457, "step": 45890 }, { "grad_norm": 0.8755475282669067, "learning_rate": 6.082807898836663e-05, "loss": 0.047, "step": 45900 }, { "grad_norm": 0.5448103547096252, "learning_rate": 6.081193608399742e-05, "loss": 0.0469, "step": 45910 }, { "grad_norm": 0.45668840408325195, "learning_rate": 6.0795791997251164e-05, "loss": 0.0416, "step": 45920 }, { "grad_norm": 0.6181842684745789, "learning_rate": 6.0779646729893294e-05, "loss": 0.0479, "step": 45930 }, { "grad_norm": 0.41467922925949097, "learning_rate": 6.0763500283689476e-05, "loss": 0.0361, "step": 45940 }, { "grad_norm": 0.5361596345901489, "learning_rate": 6.0747352660405455e-05, "loss": 0.0551, "step": 45950 }, { "grad_norm": 0.43851983547210693, "learning_rate": 6.073120386180709e-05, "loss": 0.0715, "step": 45960 }, { "grad_norm": 0.6201990842819214, "learning_rate": 6.0715053889660425e-05, "loss": 0.0478, "step": 45970 }, { "grad_norm": 0.5055099725723267, "learning_rate": 6.069890274573157e-05, "loss": 0.0396, "step": 45980 }, { "grad_norm": 0.7041274905204773, "learning_rate": 6.068275043178679e-05, "loss": 0.0597, "step": 45990 }, { "grad_norm": 0.7908263206481934, "learning_rate": 6.066659694959248e-05, "loss": 0.0458, "step": 46000 }, { "grad_norm": 0.30086496472358704, "learning_rate": 6.0650442300915176e-05, "loss": 0.0445, "step": 46010 }, { "grad_norm": 0.4873450994491577, "learning_rate": 6.063428648752152e-05, "loss": 0.0591, "step": 46020 }, { "grad_norm": 0.5329539775848389, "learning_rate": 6.0618129511178266e-05, "loss": 0.0428, "step": 46030 }, { "grad_norm": 0.3305453658103943, "learning_rate": 6.060197137365234e-05, "loss": 0.0599, "step": 46040 }, { "grad_norm": 0.5333024263381958, "learning_rate": 6.058581207671077e-05, "loss": 0.0421, "step": 46050 }, { "grad_norm": 0.47789183259010315, "learning_rate": 6.056965162212072e-05, "loss": 0.0566, "step": 46060 }, { "grad_norm": 0.28449293971061707, "learning_rate": 6.055349001164943e-05, "loss": 0.047, "step": 46070 }, { "grad_norm": 0.3572053611278534, "learning_rate": 6.0537327247064347e-05, "loss": 0.0519, "step": 46080 }, { "grad_norm": 0.6154623031616211, "learning_rate": 6.0521163330133014e-05, "loss": 0.0545, "step": 46090 }, { "grad_norm": 0.3574887216091156, "learning_rate": 6.050499826262306e-05, "loss": 0.0453, "step": 46100 }, { "grad_norm": 0.6096853613853455, "learning_rate": 6.0488832046302294e-05, "loss": 0.0593, "step": 46110 }, { "grad_norm": 0.40814346075057983, "learning_rate": 6.0472664682938626e-05, "loss": 0.0482, "step": 46120 }, { "grad_norm": 0.5553181171417236, "learning_rate": 6.045649617430009e-05, "loss": 0.0474, "step": 46130 }, { "grad_norm": 0.5721511840820312, "learning_rate": 6.0440326522154866e-05, "loss": 0.0454, "step": 46140 }, { "grad_norm": 0.4666137397289276, "learning_rate": 6.0424155728271224e-05, "loss": 0.0515, "step": 46150 }, { "grad_norm": 0.33752161264419556, "learning_rate": 6.040798379441758e-05, "loss": 0.0478, "step": 46160 }, { "grad_norm": 0.5505781769752502, "learning_rate": 6.0391810722362485e-05, "loss": 0.0536, "step": 46170 }, { "grad_norm": 0.5403416156768799, "learning_rate": 6.037563651387458e-05, "loss": 0.0439, "step": 46180 }, { "grad_norm": 0.5072623491287231, "learning_rate": 6.0359461170722666e-05, "loss": 0.0418, "step": 46190 }, { "grad_norm": 0.5499431490898132, "learning_rate": 6.034328469467566e-05, "loss": 0.0418, "step": 46200 }, { "grad_norm": 0.43821313977241516, "learning_rate": 6.0327107087502596e-05, "loss": 0.046, "step": 46210 }, { "grad_norm": 0.5611942410469055, "learning_rate": 6.031092835097262e-05, "loss": 0.0493, "step": 46220 }, { "grad_norm": 0.3925071060657501, "learning_rate": 6.0294748486855024e-05, "loss": 0.045, "step": 46230 }, { "grad_norm": 0.44835159182548523, "learning_rate": 6.0278567496919216e-05, "loss": 0.0554, "step": 46240 }, { "grad_norm": 0.5643123984336853, "learning_rate": 6.026238538293472e-05, "loss": 0.0479, "step": 46250 }, { "grad_norm": 0.4051038324832916, "learning_rate": 6.024620214667118e-05, "loss": 0.0456, "step": 46260 }, { "grad_norm": 0.4713001847267151, "learning_rate": 6.0230017789898384e-05, "loss": 0.046, "step": 46270 }, { "grad_norm": 0.48029065132141113, "learning_rate": 6.021383231438622e-05, "loss": 0.0426, "step": 46280 }, { "grad_norm": 0.4186421036720276, "learning_rate": 6.0197645721904704e-05, "loss": 0.0435, "step": 46290 }, { "grad_norm": 0.6472364664077759, "learning_rate": 6.0181458014224e-05, "loss": 0.0442, "step": 46300 }, { "grad_norm": 0.8824816942214966, "learning_rate": 6.016526919311435e-05, "loss": 0.0498, "step": 46310 }, { "grad_norm": 0.3395428955554962, "learning_rate": 6.014907926034613e-05, "loss": 0.0399, "step": 46320 }, { "grad_norm": 0.4221200942993164, "learning_rate": 6.0132888217689875e-05, "loss": 0.0443, "step": 46330 }, { "grad_norm": 0.524925947189331, "learning_rate": 6.01166960669162e-05, "loss": 0.0524, "step": 46340 }, { "grad_norm": 0.5107522010803223, "learning_rate": 6.0100502809795845e-05, "loss": 0.0456, "step": 46350 }, { "grad_norm": 0.3032577931880951, "learning_rate": 6.00843084480997e-05, "loss": 0.0487, "step": 46360 }, { "grad_norm": 0.35107240080833435, "learning_rate": 6.0068112983598736e-05, "loss": 0.0397, "step": 46370 }, { "grad_norm": 0.384326696395874, "learning_rate": 6.005191641806407e-05, "loss": 0.0384, "step": 46380 }, { "grad_norm": 0.502927839756012, "learning_rate": 6.003571875326694e-05, "loss": 0.045, "step": 46390 }, { "grad_norm": 0.6093994379043579, "learning_rate": 6.00195199909787e-05, "loss": 0.0476, "step": 46400 }, { "grad_norm": 0.579138457775116, "learning_rate": 6.000332013297082e-05, "loss": 0.0408, "step": 46410 }, { "grad_norm": 0.32543399930000305, "learning_rate": 5.998711918101487e-05, "loss": 0.0437, "step": 46420 }, { "grad_norm": 0.47535014152526855, "learning_rate": 5.997091713688261e-05, "loss": 0.0527, "step": 46430 }, { "grad_norm": 0.258209764957428, "learning_rate": 5.9954714002345836e-05, "loss": 0.0437, "step": 46440 }, { "grad_norm": 0.4952743947505951, "learning_rate": 5.993850977917649e-05, "loss": 0.0484, "step": 46450 }, { "grad_norm": 0.9399233460426331, "learning_rate": 5.992230446914667e-05, "loss": 0.0442, "step": 46460 }, { "grad_norm": 0.42096248269081116, "learning_rate": 5.990609807402855e-05, "loss": 0.0499, "step": 46470 }, { "grad_norm": 0.6244490146636963, "learning_rate": 5.988989059559443e-05, "loss": 0.0509, "step": 46480 }, { "grad_norm": 0.5840440392494202, "learning_rate": 5.987368203561675e-05, "loss": 0.0456, "step": 46490 }, { "grad_norm": 0.5225292444229126, "learning_rate": 5.9857472395868055e-05, "loss": 0.0549, "step": 46500 }, { "grad_norm": 0.494352251291275, "learning_rate": 5.9841261678120983e-05, "loss": 0.0489, "step": 46510 }, { "grad_norm": 0.49342891573905945, "learning_rate": 5.982504988414834e-05, "loss": 0.0374, "step": 46520 }, { "grad_norm": 0.5096253752708435, "learning_rate": 5.9808837015723015e-05, "loss": 0.0585, "step": 46530 }, { "grad_norm": 0.3339466154575348, "learning_rate": 5.9792623074618016e-05, "loss": 0.0449, "step": 46540 }, { "grad_norm": 0.4255692660808563, "learning_rate": 5.977640806260648e-05, "loss": 0.0492, "step": 46550 }, { "grad_norm": 0.5244811177253723, "learning_rate": 5.976019198146166e-05, "loss": 0.0457, "step": 46560 }, { "grad_norm": 0.671513020992279, "learning_rate": 5.974397483295692e-05, "loss": 0.0431, "step": 46570 }, { "grad_norm": 0.377056747674942, "learning_rate": 5.972775661886575e-05, "loss": 0.0428, "step": 46580 }, { "grad_norm": 0.4511156678199768, "learning_rate": 5.971153734096173e-05, "loss": 0.0463, "step": 46590 }, { "grad_norm": 0.5364383459091187, "learning_rate": 5.969531700101859e-05, "loss": 0.0386, "step": 46600 }, { "grad_norm": 0.39502808451652527, "learning_rate": 5.9679095600810155e-05, "loss": 0.0502, "step": 46610 }, { "grad_norm": 0.44379347562789917, "learning_rate": 5.9662873142110384e-05, "loss": 0.0401, "step": 46620 }, { "grad_norm": 0.4728422164916992, "learning_rate": 5.964664962669333e-05, "loss": 0.0412, "step": 46630 }, { "grad_norm": 0.560775637626648, "learning_rate": 5.9630425056333186e-05, "loss": 0.0482, "step": 46640 }, { "grad_norm": 0.29725876450538635, "learning_rate": 5.961419943280422e-05, "loss": 0.0542, "step": 46650 }, { "grad_norm": 0.40984755754470825, "learning_rate": 5.959797275788087e-05, "loss": 0.0387, "step": 46660 }, { "grad_norm": 0.6409299373626709, "learning_rate": 5.958174503333765e-05, "loss": 0.0529, "step": 46670 }, { "grad_norm": 0.44429007172584534, "learning_rate": 5.9565516260949195e-05, "loss": 0.052, "step": 46680 }, { "grad_norm": 0.6717647910118103, "learning_rate": 5.9549286442490273e-05, "loss": 0.0463, "step": 46690 }, { "grad_norm": 0.6028324961662292, "learning_rate": 5.953305557973572e-05, "loss": 0.0456, "step": 46700 }, { "grad_norm": 0.6602747440338135, "learning_rate": 5.9516823674460564e-05, "loss": 0.0441, "step": 46710 }, { "grad_norm": 0.8935216665267944, "learning_rate": 5.9500590728439875e-05, "loss": 0.0497, "step": 46720 }, { "grad_norm": 0.42763203382492065, "learning_rate": 5.948435674344886e-05, "loss": 0.0514, "step": 46730 }, { "grad_norm": 0.755182147026062, "learning_rate": 5.946812172126285e-05, "loss": 0.0468, "step": 46740 }, { "grad_norm": 0.43455755710601807, "learning_rate": 5.945188566365729e-05, "loss": 0.0407, "step": 46750 }, { "grad_norm": 0.40818291902542114, "learning_rate": 5.943564857240773e-05, "loss": 0.0519, "step": 46760 }, { "grad_norm": 0.7722665667533875, "learning_rate": 5.941941044928981e-05, "loss": 0.0511, "step": 46770 }, { "grad_norm": 0.5082001686096191, "learning_rate": 5.940317129607935e-05, "loss": 0.0423, "step": 46780 }, { "grad_norm": 0.4468514919281006, "learning_rate": 5.9386931114552204e-05, "loss": 0.0455, "step": 46790 }, { "grad_norm": 0.4206448495388031, "learning_rate": 5.937068990648438e-05, "loss": 0.0465, "step": 46800 }, { "grad_norm": 0.5626499652862549, "learning_rate": 5.935444767365199e-05, "loss": 0.07, "step": 46810 }, { "grad_norm": 0.5008993148803711, "learning_rate": 5.933820441783129e-05, "loss": 0.0662, "step": 46820 }, { "grad_norm": 0.4645707309246063, "learning_rate": 5.932196014079857e-05, "loss": 0.0497, "step": 46830 }, { "grad_norm": 0.4198070466518402, "learning_rate": 5.930571484433032e-05, "loss": 0.0372, "step": 46840 }, { "grad_norm": 0.46471500396728516, "learning_rate": 5.928946853020309e-05, "loss": 0.0451, "step": 46850 }, { "grad_norm": 0.44830864667892456, "learning_rate": 5.927322120019352e-05, "loss": 0.0451, "step": 46860 }, { "grad_norm": 0.6557577252388, "learning_rate": 5.925697285607844e-05, "loss": 0.062, "step": 46870 }, { "grad_norm": 0.8192325234413147, "learning_rate": 5.924072349963472e-05, "loss": 0.0541, "step": 46880 }, { "grad_norm": 0.4288097023963928, "learning_rate": 5.922447313263938e-05, "loss": 0.0447, "step": 46890 }, { "grad_norm": 0.5624406933784485, "learning_rate": 5.920822175686952e-05, "loss": 0.0573, "step": 46900 }, { "grad_norm": 0.54914790391922, "learning_rate": 5.919196937410239e-05, "loss": 0.0596, "step": 46910 }, { "grad_norm": 0.4312898814678192, "learning_rate": 5.917571598611529e-05, "loss": 0.046, "step": 46920 }, { "grad_norm": 0.5917773842811584, "learning_rate": 5.9159461594685686e-05, "loss": 0.0501, "step": 46930 }, { "grad_norm": 0.45644834637641907, "learning_rate": 5.914320620159114e-05, "loss": 0.0479, "step": 46940 }, { "grad_norm": 0.5292416214942932, "learning_rate": 5.912694980860931e-05, "loss": 0.0525, "step": 46950 }, { "grad_norm": 0.4138244390487671, "learning_rate": 5.9110692417517964e-05, "loss": 0.0509, "step": 46960 }, { "grad_norm": 0.5646077990531921, "learning_rate": 5.9094434030095e-05, "loss": 0.0626, "step": 46970 }, { "grad_norm": 0.9526076912879944, "learning_rate": 5.907817464811841e-05, "loss": 0.0446, "step": 46980 }, { "grad_norm": 0.6687705516815186, "learning_rate": 5.906191427336627e-05, "loss": 0.0482, "step": 46990 }, { "grad_norm": 0.7239371538162231, "learning_rate": 5.9045652907616835e-05, "loss": 0.0518, "step": 47000 }, { "grad_norm": 0.5786877870559692, "learning_rate": 5.902939055264838e-05, "loss": 0.0496, "step": 47010 }, { "grad_norm": 0.50784832239151, "learning_rate": 5.901312721023935e-05, "loss": 0.0553, "step": 47020 }, { "grad_norm": 0.3890427350997925, "learning_rate": 5.8996862882168294e-05, "loss": 0.0414, "step": 47030 }, { "grad_norm": 0.424697607755661, "learning_rate": 5.8980597570213826e-05, "loss": 0.0625, "step": 47040 }, { "grad_norm": 0.8507677912712097, "learning_rate": 5.896433127615471e-05, "loss": 0.0544, "step": 47050 }, { "grad_norm": 0.3948633372783661, "learning_rate": 5.894806400176981e-05, "loss": 0.0378, "step": 47060 }, { "grad_norm": 0.6216955780982971, "learning_rate": 5.893179574883808e-05, "loss": 0.0427, "step": 47070 }, { "grad_norm": 0.441211462020874, "learning_rate": 5.8915526519138585e-05, "loss": 0.0508, "step": 47080 }, { "grad_norm": 0.6427065134048462, "learning_rate": 5.889925631445053e-05, "loss": 0.0551, "step": 47090 }, { "grad_norm": 0.41307124495506287, "learning_rate": 5.888298513655318e-05, "loss": 0.0367, "step": 47100 }, { "grad_norm": 0.4232921004295349, "learning_rate": 5.886671298722592e-05, "loss": 0.0378, "step": 47110 }, { "grad_norm": 0.8639710545539856, "learning_rate": 5.885043986824828e-05, "loss": 0.0431, "step": 47120 }, { "grad_norm": 0.9492754936218262, "learning_rate": 5.883416578139982e-05, "loss": 0.0514, "step": 47130 }, { "grad_norm": 0.34471410512924194, "learning_rate": 5.881789072846029e-05, "loss": 0.0489, "step": 47140 }, { "grad_norm": 0.509741485118866, "learning_rate": 5.8801614711209474e-05, "loss": 0.0402, "step": 47150 }, { "grad_norm": 0.4289140999317169, "learning_rate": 5.878533773142732e-05, "loss": 0.0427, "step": 47160 }, { "grad_norm": 0.3724318742752075, "learning_rate": 5.8769059790893834e-05, "loss": 0.0481, "step": 47170 }, { "grad_norm": 0.655093252658844, "learning_rate": 5.8752780891389145e-05, "loss": 0.0576, "step": 47180 }, { "grad_norm": 0.7893022894859314, "learning_rate": 5.873650103469351e-05, "loss": 0.0448, "step": 47190 }, { "grad_norm": 0.28385064005851746, "learning_rate": 5.872022022258725e-05, "loss": 0.0368, "step": 47200 }, { "grad_norm": 0.38303041458129883, "learning_rate": 5.8703938456850814e-05, "loss": 0.0415, "step": 47210 }, { "grad_norm": 0.4888629615306854, "learning_rate": 5.868765573926477e-05, "loss": 0.0517, "step": 47220 }, { "grad_norm": 0.4432007074356079, "learning_rate": 5.867137207160974e-05, "loss": 0.0617, "step": 47230 }, { "grad_norm": 0.5911728739738464, "learning_rate": 5.86550874556665e-05, "loss": 0.0503, "step": 47240 }, { "grad_norm": 0.302446573972702, "learning_rate": 5.863880189321592e-05, "loss": 0.0454, "step": 47250 }, { "grad_norm": 0.5246943831443787, "learning_rate": 5.862251538603894e-05, "loss": 0.0375, "step": 47260 }, { "grad_norm": 0.3302302658557892, "learning_rate": 5.8606227935916656e-05, "loss": 0.0381, "step": 47270 }, { "grad_norm": 0.3557806611061096, "learning_rate": 5.858993954463021e-05, "loss": 0.043, "step": 47280 }, { "grad_norm": 0.4260231852531433, "learning_rate": 5.8573650213960896e-05, "loss": 0.05, "step": 47290 }, { "grad_norm": 0.3810003995895386, "learning_rate": 5.8557359945690084e-05, "loss": 0.0481, "step": 47300 }, { "grad_norm": 0.5606212019920349, "learning_rate": 5.8541068741599246e-05, "loss": 0.0525, "step": 47310 }, { "grad_norm": 0.5222567915916443, "learning_rate": 5.8524776603469985e-05, "loss": 0.0373, "step": 47320 }, { "grad_norm": 0.4378889799118042, "learning_rate": 5.850848353308397e-05, "loss": 0.0419, "step": 47330 }, { "grad_norm": 0.9388659000396729, "learning_rate": 5.849218953222297e-05, "loss": 0.0564, "step": 47340 }, { "grad_norm": 0.4637584090232849, "learning_rate": 5.847589460266891e-05, "loss": 0.0438, "step": 47350 }, { "grad_norm": 0.46307045221328735, "learning_rate": 5.845959874620375e-05, "loss": 0.047, "step": 47360 }, { "grad_norm": 0.30840811133384705, "learning_rate": 5.84433019646096e-05, "loss": 0.0525, "step": 47370 }, { "grad_norm": 0.7115793228149414, "learning_rate": 5.842700425966863e-05, "loss": 0.0493, "step": 47380 }, { "grad_norm": 0.35361167788505554, "learning_rate": 5.841070563316315e-05, "loss": 0.0452, "step": 47390 }, { "grad_norm": 0.8702054619789124, "learning_rate": 5.839440608687554e-05, "loss": 0.0491, "step": 47400 }, { "grad_norm": 0.7237820029258728, "learning_rate": 5.837810562258831e-05, "loss": 0.0407, "step": 47410 }, { "grad_norm": 0.5652170777320862, "learning_rate": 5.836180424208405e-05, "loss": 0.0684, "step": 47420 }, { "grad_norm": 0.46458348631858826, "learning_rate": 5.834550194714543e-05, "loss": 0.0425, "step": 47430 }, { "grad_norm": 0.5219388008117676, "learning_rate": 5.832919873955528e-05, "loss": 0.0617, "step": 47440 }, { "grad_norm": 0.4961548149585724, "learning_rate": 5.831289462109648e-05, "loss": 0.0534, "step": 47450 }, { "grad_norm": 0.23711659014225006, "learning_rate": 5.829658959355202e-05, "loss": 0.0506, "step": 47460 }, { "grad_norm": 0.3883431851863861, "learning_rate": 5.8280283658704995e-05, "loss": 0.056, "step": 47470 }, { "grad_norm": 0.5456571578979492, "learning_rate": 5.826397681833859e-05, "loss": 0.0501, "step": 47480 }, { "grad_norm": 0.4856874346733093, "learning_rate": 5.824766907423612e-05, "loss": 0.0444, "step": 47490 }, { "grad_norm": 0.4377448260784149, "learning_rate": 5.8231360428180945e-05, "loss": 0.0466, "step": 47500 }, { "grad_norm": 0.4662648141384125, "learning_rate": 5.821505088195658e-05, "loss": 0.0465, "step": 47510 }, { "grad_norm": 0.47360897064208984, "learning_rate": 5.819874043734661e-05, "loss": 0.054, "step": 47520 }, { "grad_norm": 0.5842567086219788, "learning_rate": 5.8182429096134695e-05, "loss": 0.0427, "step": 47530 }, { "grad_norm": 0.7282580137252808, "learning_rate": 5.816611686010465e-05, "loss": 0.0454, "step": 47540 }, { "grad_norm": 0.3430377244949341, "learning_rate": 5.814980373104033e-05, "loss": 0.0525, "step": 47550 }, { "grad_norm": 0.5528621673583984, "learning_rate": 5.813348971072572e-05, "loss": 0.0597, "step": 47560 }, { "grad_norm": 0.3505125939846039, "learning_rate": 5.811717480094492e-05, "loss": 0.043, "step": 47570 }, { "grad_norm": 0.4914354681968689, "learning_rate": 5.810085900348209e-05, "loss": 0.0458, "step": 47580 }, { "grad_norm": 1.0669066905975342, "learning_rate": 5.8084542320121483e-05, "loss": 0.0614, "step": 47590 }, { "grad_norm": 0.6197124123573303, "learning_rate": 5.8068224752647497e-05, "loss": 0.0657, "step": 47600 }, { "grad_norm": 0.7864397764205933, "learning_rate": 5.805190630284457e-05, "loss": 0.0353, "step": 47610 }, { "grad_norm": 0.5665559768676758, "learning_rate": 5.803558697249729e-05, "loss": 0.0426, "step": 47620 }, { "grad_norm": 0.600027859210968, "learning_rate": 5.8019266763390276e-05, "loss": 0.0592, "step": 47630 }, { "grad_norm": 0.24249176681041718, "learning_rate": 5.800294567730833e-05, "loss": 0.0425, "step": 47640 }, { "grad_norm": 0.5054630041122437, "learning_rate": 5.798662371603628e-05, "loss": 0.047, "step": 47650 }, { "grad_norm": 0.5404293537139893, "learning_rate": 5.797030088135904e-05, "loss": 0.0477, "step": 47660 }, { "grad_norm": 0.5143184065818787, "learning_rate": 5.7953977175061705e-05, "loss": 0.0378, "step": 47670 }, { "grad_norm": 0.5174359679222107, "learning_rate": 5.793765259892938e-05, "loss": 0.0561, "step": 47680 }, { "grad_norm": 0.31154167652130127, "learning_rate": 5.792132715474729e-05, "loss": 0.0405, "step": 47690 }, { "grad_norm": 0.5017186999320984, "learning_rate": 5.790500084430078e-05, "loss": 0.0389, "step": 47700 }, { "grad_norm": 0.3654184341430664, "learning_rate": 5.7888673669375274e-05, "loss": 0.0427, "step": 47710 }, { "grad_norm": 0.40238508582115173, "learning_rate": 5.787234563175625e-05, "loss": 0.0463, "step": 47720 }, { "grad_norm": 0.6637774109840393, "learning_rate": 5.7856016733229355e-05, "loss": 0.0513, "step": 47730 }, { "grad_norm": 0.46288561820983887, "learning_rate": 5.7839686975580297e-05, "loss": 0.0574, "step": 47740 }, { "grad_norm": 0.42581772804260254, "learning_rate": 5.782335636059484e-05, "loss": 0.0446, "step": 47750 }, { "grad_norm": 0.4592531621456146, "learning_rate": 5.780702489005889e-05, "loss": 0.0439, "step": 47760 }, { "grad_norm": 0.3747580349445343, "learning_rate": 5.779069256575845e-05, "loss": 0.0514, "step": 47770 }, { "grad_norm": 0.517117440700531, "learning_rate": 5.7774359389479574e-05, "loss": 0.0405, "step": 47780 }, { "grad_norm": 0.73057621717453, "learning_rate": 5.775802536300845e-05, "loss": 0.0579, "step": 47790 }, { "grad_norm": 0.764045000076294, "learning_rate": 5.774169048813134e-05, "loss": 0.0468, "step": 47800 }, { "grad_norm": 0.5964531898498535, "learning_rate": 5.77253547666346e-05, "loss": 0.0468, "step": 47810 }, { "grad_norm": 0.5383287072181702, "learning_rate": 5.770901820030465e-05, "loss": 0.0538, "step": 47820 }, { "grad_norm": 0.44657739996910095, "learning_rate": 5.769268079092809e-05, "loss": 0.047, "step": 47830 }, { "grad_norm": 0.4921571910381317, "learning_rate": 5.767634254029151e-05, "loss": 0.0345, "step": 47840 }, { "grad_norm": 0.6258965730667114, "learning_rate": 5.7660003450181655e-05, "loss": 0.0637, "step": 47850 }, { "grad_norm": 0.5643960237503052, "learning_rate": 5.764366352238534e-05, "loss": 0.0444, "step": 47860 }, { "grad_norm": 0.5256535410881042, "learning_rate": 5.7627322758689474e-05, "loss": 0.0392, "step": 47870 }, { "grad_norm": 0.4635198414325714, "learning_rate": 5.761098116088105e-05, "loss": 0.0507, "step": 47880 }, { "grad_norm": 0.5803474187850952, "learning_rate": 5.759463873074717e-05, "loss": 0.043, "step": 47890 }, { "grad_norm": 0.4864705801010132, "learning_rate": 5.757829547007504e-05, "loss": 0.0453, "step": 47900 }, { "grad_norm": 0.5954249501228333, "learning_rate": 5.756195138065189e-05, "loss": 0.0451, "step": 47910 }, { "grad_norm": 0.4360894560813904, "learning_rate": 5.754560646426511e-05, "loss": 0.0488, "step": 47920 }, { "grad_norm": 0.363862544298172, "learning_rate": 5.752926072270216e-05, "loss": 0.0395, "step": 47930 }, { "grad_norm": 0.5800789594650269, "learning_rate": 5.7512914157750563e-05, "loss": 0.0406, "step": 47940 }, { "grad_norm": 0.32241129875183105, "learning_rate": 5.749656677119798e-05, "loss": 0.0364, "step": 47950 }, { "grad_norm": 0.6808487772941589, "learning_rate": 5.7480218564832125e-05, "loss": 0.0566, "step": 47960 }, { "grad_norm": 0.4595644772052765, "learning_rate": 5.746386954044082e-05, "loss": 0.0392, "step": 47970 }, { "grad_norm": 0.8041648864746094, "learning_rate": 5.744751969981195e-05, "loss": 0.0432, "step": 47980 }, { "grad_norm": 0.3605642020702362, "learning_rate": 5.7431169044733526e-05, "loss": 0.0464, "step": 47990 }, { "grad_norm": 0.6664701700210571, "learning_rate": 5.741481757699364e-05, "loss": 0.0409, "step": 48000 }, { "grad_norm": 0.5570117831230164, "learning_rate": 5.7398465298380434e-05, "loss": 0.0476, "step": 48010 }, { "grad_norm": 0.3739665150642395, "learning_rate": 5.7382112210682193e-05, "loss": 0.0602, "step": 48020 }, { "grad_norm": 0.570816695690155, "learning_rate": 5.7365758315687266e-05, "loss": 0.0468, "step": 48030 }, { "grad_norm": 0.5395732522010803, "learning_rate": 5.734940361518407e-05, "loss": 0.054, "step": 48040 }, { "grad_norm": 0.5807451009750366, "learning_rate": 5.733304811096116e-05, "loss": 0.0463, "step": 48050 }, { "grad_norm": 0.3455738127231598, "learning_rate": 5.731669180480713e-05, "loss": 0.0508, "step": 48060 }, { "grad_norm": 0.523878812789917, "learning_rate": 5.730033469851067e-05, "loss": 0.057, "step": 48070 }, { "grad_norm": 0.7002742886543274, "learning_rate": 5.7283976793860607e-05, "loss": 0.0499, "step": 48080 }, { "grad_norm": 0.24846495687961578, "learning_rate": 5.726761809264577e-05, "loss": 0.0403, "step": 48090 }, { "grad_norm": 0.9537909030914307, "learning_rate": 5.7251258596655155e-05, "loss": 0.0438, "step": 48100 }, { "grad_norm": 1.112400770187378, "learning_rate": 5.72348983076778e-05, "loss": 0.0491, "step": 48110 }, { "grad_norm": 0.439163476228714, "learning_rate": 5.7218537227502854e-05, "loss": 0.0457, "step": 48120 }, { "grad_norm": 0.3025103211402893, "learning_rate": 5.720217535791951e-05, "loss": 0.0372, "step": 48130 }, { "grad_norm": 0.5995495319366455, "learning_rate": 5.718581270071711e-05, "loss": 0.0542, "step": 48140 }, { "grad_norm": 0.49608540534973145, "learning_rate": 5.716944925768505e-05, "loss": 0.0404, "step": 48150 }, { "grad_norm": 0.3580847978591919, "learning_rate": 5.7153085030612786e-05, "loss": 0.0435, "step": 48160 }, { "grad_norm": 0.36488404870033264, "learning_rate": 5.713672002128989e-05, "loss": 0.0615, "step": 48170 }, { "grad_norm": 0.528386652469635, "learning_rate": 5.712035423150602e-05, "loss": 0.052, "step": 48180 }, { "grad_norm": 0.4923468232154846, "learning_rate": 5.710398766305094e-05, "loss": 0.0461, "step": 48190 }, { "grad_norm": 0.45698586106300354, "learning_rate": 5.708762031771442e-05, "loss": 0.0426, "step": 48200 }, { "grad_norm": 0.5967830419540405, "learning_rate": 5.707125219728642e-05, "loss": 0.0384, "step": 48210 }, { "grad_norm": 0.4185928702354431, "learning_rate": 5.7054883303556905e-05, "loss": 0.045, "step": 48220 }, { "grad_norm": 0.5349538326263428, "learning_rate": 5.703851363831595e-05, "loss": 0.0571, "step": 48230 }, { "grad_norm": 0.5726323127746582, "learning_rate": 5.702214320335374e-05, "loss": 0.0613, "step": 48240 }, { "grad_norm": 0.5749265551567078, "learning_rate": 5.7005772000460514e-05, "loss": 0.0532, "step": 48250 }, { "grad_norm": 0.4233984351158142, "learning_rate": 5.698940003142656e-05, "loss": 0.0495, "step": 48260 }, { "grad_norm": 0.4362339973449707, "learning_rate": 5.697302729804236e-05, "loss": 0.0588, "step": 48270 }, { "grad_norm": 0.8618937730789185, "learning_rate": 5.695665380209837e-05, "loss": 0.0503, "step": 48280 }, { "grad_norm": 0.50276118516922, "learning_rate": 5.6940279545385165e-05, "loss": 0.0389, "step": 48290 }, { "grad_norm": 0.6867548227310181, "learning_rate": 5.692390452969344e-05, "loss": 0.0499, "step": 48300 }, { "grad_norm": 0.9002609252929688, "learning_rate": 5.690752875681392e-05, "loss": 0.0404, "step": 48310 }, { "grad_norm": 0.3449605405330658, "learning_rate": 5.6891152228537435e-05, "loss": 0.0465, "step": 48320 }, { "grad_norm": 0.6800221800804138, "learning_rate": 5.687477494665492e-05, "loss": 0.0409, "step": 48330 }, { "grad_norm": 0.4240587055683136, "learning_rate": 5.685839691295733e-05, "loss": 0.0433, "step": 48340 }, { "grad_norm": 0.4741365313529968, "learning_rate": 5.6842018129235786e-05, "loss": 0.0434, "step": 48350 }, { "grad_norm": 0.35313668847084045, "learning_rate": 5.6825638597281404e-05, "loss": 0.038, "step": 48360 }, { "grad_norm": 0.43719029426574707, "learning_rate": 5.680925831888546e-05, "loss": 0.0595, "step": 48370 }, { "grad_norm": 0.4722763001918793, "learning_rate": 5.6792877295839274e-05, "loss": 0.0479, "step": 48380 }, { "grad_norm": 0.7491981387138367, "learning_rate": 5.6776495529934224e-05, "loss": 0.0456, "step": 48390 }, { "grad_norm": 0.5750075578689575, "learning_rate": 5.6760113022961824e-05, "loss": 0.0453, "step": 48400 }, { "grad_norm": 0.5994870066642761, "learning_rate": 5.6743729776713617e-05, "loss": 0.0526, "step": 48410 }, { "grad_norm": 0.7095215320587158, "learning_rate": 5.672734579298126e-05, "loss": 0.0454, "step": 48420 }, { "grad_norm": 0.5030589699745178, "learning_rate": 5.671096107355649e-05, "loss": 0.0474, "step": 48430 }, { "grad_norm": 0.46993061900138855, "learning_rate": 5.66945756202311e-05, "loss": 0.0412, "step": 48440 }, { "grad_norm": 0.4890495240688324, "learning_rate": 5.667818943479699e-05, "loss": 0.0548, "step": 48450 }, { "grad_norm": 0.5522494912147522, "learning_rate": 5.666180251904612e-05, "loss": 0.0499, "step": 48460 }, { "grad_norm": 0.4559914171695709, "learning_rate": 5.6645414874770555e-05, "loss": 0.043, "step": 48470 }, { "grad_norm": 0.5505487322807312, "learning_rate": 5.66290265037624e-05, "loss": 0.0536, "step": 48480 }, { "grad_norm": 0.6636212468147278, "learning_rate": 5.661263740781386e-05, "loss": 0.038, "step": 48490 }, { "grad_norm": 0.6439868807792664, "learning_rate": 5.6596247588717254e-05, "loss": 0.0436, "step": 48500 }, { "grad_norm": 0.4668419361114502, "learning_rate": 5.6579857048264926e-05, "loss": 0.0418, "step": 48510 }, { "grad_norm": 0.6322882175445557, "learning_rate": 5.6563465788249314e-05, "loss": 0.0398, "step": 48520 }, { "grad_norm": 0.7024990320205688, "learning_rate": 5.6547073810462956e-05, "loss": 0.0441, "step": 48530 }, { "grad_norm": 0.4188385009765625, "learning_rate": 5.653068111669846e-05, "loss": 0.0451, "step": 48540 }, { "grad_norm": 0.37919142842292786, "learning_rate": 5.651428770874848e-05, "loss": 0.0493, "step": 48550 }, { "grad_norm": 0.3499290645122528, "learning_rate": 5.64978935884058e-05, "loss": 0.0412, "step": 48560 }, { "grad_norm": 0.5075868964195251, "learning_rate": 5.6481498757463244e-05, "loss": 0.0439, "step": 48570 }, { "grad_norm": 0.4626539945602417, "learning_rate": 5.646510321771373e-05, "loss": 0.0471, "step": 48580 }, { "grad_norm": 0.546334981918335, "learning_rate": 5.644870697095024e-05, "loss": 0.045, "step": 48590 }, { "grad_norm": 0.48649245500564575, "learning_rate": 5.643231001896586e-05, "loss": 0.0404, "step": 48600 }, { "grad_norm": 0.41718271374702454, "learning_rate": 5.6415912363553726e-05, "loss": 0.0674, "step": 48610 }, { "grad_norm": 0.523724377155304, "learning_rate": 5.639951400650706e-05, "loss": 0.0566, "step": 48620 }, { "grad_norm": 0.5668164491653442, "learning_rate": 5.6383114949619165e-05, "loss": 0.0399, "step": 48630 }, { "grad_norm": 0.4112843871116638, "learning_rate": 5.636671519468342e-05, "loss": 0.0498, "step": 48640 }, { "grad_norm": 0.35615694522857666, "learning_rate": 5.635031474349327e-05, "loss": 0.0429, "step": 48650 }, { "grad_norm": 0.4928550124168396, "learning_rate": 5.6333913597842246e-05, "loss": 0.0419, "step": 48660 }, { "grad_norm": 0.5827411413192749, "learning_rate": 5.6317511759523955e-05, "loss": 0.0491, "step": 48670 }, { "grad_norm": 0.4435907006263733, "learning_rate": 5.630110923033207e-05, "loss": 0.0447, "step": 48680 }, { "grad_norm": 0.6156228184700012, "learning_rate": 5.628470601206036e-05, "loss": 0.0484, "step": 48690 }, { "grad_norm": 0.7270771861076355, "learning_rate": 5.626830210650266e-05, "loss": 0.0489, "step": 48700 }, { "grad_norm": 0.4192240834236145, "learning_rate": 5.625189751545285e-05, "loss": 0.0526, "step": 48710 }, { "grad_norm": 0.31247034668922424, "learning_rate": 5.6235492240704936e-05, "loss": 0.0347, "step": 48720 }, { "grad_norm": 0.3966791331768036, "learning_rate": 5.621908628405296e-05, "loss": 0.0467, "step": 48730 }, { "grad_norm": 0.9312430024147034, "learning_rate": 5.620267964729106e-05, "loss": 0.0517, "step": 48740 }, { "grad_norm": 0.35633137822151184, "learning_rate": 5.618627233221344e-05, "loss": 0.0382, "step": 48750 }, { "grad_norm": 0.8448773622512817, "learning_rate": 5.616986434061438e-05, "loss": 0.0515, "step": 48760 }, { "grad_norm": 0.896153450012207, "learning_rate": 5.615345567428822e-05, "loss": 0.0469, "step": 48770 }, { "grad_norm": 0.36587169766426086, "learning_rate": 5.613704633502941e-05, "loss": 0.0353, "step": 48780 }, { "grad_norm": 0.45331689715385437, "learning_rate": 5.612063632463245e-05, "loss": 0.0474, "step": 48790 }, { "grad_norm": 0.40347498655319214, "learning_rate": 5.610422564489188e-05, "loss": 0.0414, "step": 48800 }, { "grad_norm": 0.4503191113471985, "learning_rate": 5.608781429760239e-05, "loss": 0.0424, "step": 48810 }, { "grad_norm": 0.4327108860015869, "learning_rate": 5.607140228455866e-05, "loss": 0.0368, "step": 48820 }, { "grad_norm": 0.43982645869255066, "learning_rate": 5.605498960755553e-05, "loss": 0.0555, "step": 48830 }, { "grad_norm": 0.5485641360282898, "learning_rate": 5.603857626838782e-05, "loss": 0.0436, "step": 48840 }, { "grad_norm": 0.5223122239112854, "learning_rate": 5.60221622688505e-05, "loss": 0.0438, "step": 48850 }, { "grad_norm": 0.4917372167110443, "learning_rate": 5.6005747610738565e-05, "loss": 0.0393, "step": 48860 }, { "grad_norm": 0.4295201897621155, "learning_rate": 5.59893322958471e-05, "loss": 0.041, "step": 48870 }, { "grad_norm": 0.6647982597351074, "learning_rate": 5.5972916325971256e-05, "loss": 0.0529, "step": 48880 }, { "grad_norm": 0.39476221799850464, "learning_rate": 5.595649970290628e-05, "loss": 0.0441, "step": 48890 }, { "grad_norm": 0.4075770080089569, "learning_rate": 5.5940082428447426e-05, "loss": 0.0451, "step": 48900 }, { "grad_norm": 0.6660475730895996, "learning_rate": 5.592366450439012e-05, "loss": 0.0514, "step": 48910 }, { "grad_norm": 0.7163017392158508, "learning_rate": 5.590724593252975e-05, "loss": 0.0505, "step": 48920 }, { "grad_norm": 0.38194483518600464, "learning_rate": 5.589082671466184e-05, "loss": 0.0394, "step": 48930 }, { "grad_norm": 0.4891251027584076, "learning_rate": 5.587440685258199e-05, "loss": 0.0524, "step": 48940 }, { "grad_norm": 0.6328452825546265, "learning_rate": 5.585798634808583e-05, "loss": 0.0448, "step": 48950 }, { "grad_norm": 0.5307134389877319, "learning_rate": 5.584156520296909e-05, "loss": 0.0491, "step": 48960 }, { "grad_norm": 0.594730794429779, "learning_rate": 5.582514341902757e-05, "loss": 0.0495, "step": 48970 }, { "grad_norm": 0.41898563504219055, "learning_rate": 5.580872099805713e-05, "loss": 0.0415, "step": 48980 }, { "grad_norm": 0.29090842604637146, "learning_rate": 5.5792297941853674e-05, "loss": 0.0463, "step": 48990 }, { "grad_norm": 0.41651439666748047, "learning_rate": 5.5775874252213247e-05, "loss": 0.0458, "step": 49000 }, { "grad_norm": 0.46881306171417236, "learning_rate": 5.575944993093189e-05, "loss": 0.0465, "step": 49010 }, { "grad_norm": 0.5780851244926453, "learning_rate": 5.574302497980574e-05, "loss": 0.0418, "step": 49020 }, { "grad_norm": 0.687965989112854, "learning_rate": 5.5726599400631005e-05, "loss": 0.043, "step": 49030 }, { "grad_norm": 0.5149145722389221, "learning_rate": 5.571017319520399e-05, "loss": 0.0483, "step": 49040 }, { "grad_norm": 0.4045940637588501, "learning_rate": 5.5693746365320984e-05, "loss": 0.0406, "step": 49050 }, { "grad_norm": 0.6104233264923096, "learning_rate": 5.567731891277848e-05, "loss": 0.0522, "step": 49060 }, { "grad_norm": 0.4031237065792084, "learning_rate": 5.566089083937288e-05, "loss": 0.048, "step": 49070 }, { "grad_norm": 0.33325645327568054, "learning_rate": 5.564446214690079e-05, "loss": 0.0489, "step": 49080 }, { "grad_norm": 0.5615314245223999, "learning_rate": 5.56280328371588e-05, "loss": 0.0597, "step": 49090 }, { "grad_norm": 0.47109800577163696, "learning_rate": 5.5611602911943596e-05, "loss": 0.0416, "step": 49100 }, { "grad_norm": 0.8440603613853455, "learning_rate": 5.559517237305194e-05, "loss": 0.0561, "step": 49110 }, { "grad_norm": 0.6042574644088745, "learning_rate": 5.557874122228064e-05, "loss": 0.0508, "step": 49120 }, { "grad_norm": 0.5168505311012268, "learning_rate": 5.55623094614266e-05, "loss": 0.047, "step": 49130 }, { "grad_norm": 0.666347861289978, "learning_rate": 5.5545877092286743e-05, "loss": 0.0523, "step": 49140 }, { "grad_norm": 0.521165132522583, "learning_rate": 5.552944411665811e-05, "loss": 0.0406, "step": 49150 }, { "grad_norm": 0.37899067997932434, "learning_rate": 5.5513010536337795e-05, "loss": 0.052, "step": 49160 }, { "grad_norm": 0.5444997549057007, "learning_rate": 5.5496576353122934e-05, "loss": 0.045, "step": 49170 }, { "grad_norm": 0.5248764753341675, "learning_rate": 5.548014156881074e-05, "loss": 0.0546, "step": 49180 }, { "grad_norm": 0.5206273198127747, "learning_rate": 5.5463706185198494e-05, "loss": 0.0519, "step": 49190 }, { "grad_norm": 0.5005355477333069, "learning_rate": 5.544727020408358e-05, "loss": 0.0487, "step": 49200 }, { "grad_norm": 0.3401598334312439, "learning_rate": 5.5430833627263366e-05, "loss": 0.0372, "step": 49210 }, { "grad_norm": 0.34791481494903564, "learning_rate": 5.5414396456535354e-05, "loss": 0.0417, "step": 49220 }, { "grad_norm": 0.4515964686870575, "learning_rate": 5.539795869369709e-05, "loss": 0.0347, "step": 49230 }, { "grad_norm": 0.6437493562698364, "learning_rate": 5.538152034054618e-05, "loss": 0.0398, "step": 49240 }, { "grad_norm": 0.4971570670604706, "learning_rate": 5.536508139888028e-05, "loss": 0.0462, "step": 49250 }, { "grad_norm": 0.4755154252052307, "learning_rate": 5.534864187049716e-05, "loss": 0.0462, "step": 49260 }, { "grad_norm": 0.4311789870262146, "learning_rate": 5.5332201757194615e-05, "loss": 0.0405, "step": 49270 }, { "grad_norm": 0.3331347107887268, "learning_rate": 5.531576106077049e-05, "loss": 0.0386, "step": 49280 }, { "grad_norm": 0.41375237703323364, "learning_rate": 5.529931978302272e-05, "loss": 0.0374, "step": 49290 }, { "grad_norm": 0.4239288866519928, "learning_rate": 5.528287792574932e-05, "loss": 0.0381, "step": 49300 }, { "grad_norm": 0.44130274653434753, "learning_rate": 5.5266435490748324e-05, "loss": 0.0415, "step": 49310 }, { "grad_norm": 0.3669307827949524, "learning_rate": 5.524999247981787e-05, "loss": 0.0425, "step": 49320 }, { "grad_norm": 0.7541332840919495, "learning_rate": 5.523354889475613e-05, "loss": 0.0464, "step": 49330 }, { "grad_norm": 0.4063931703567505, "learning_rate": 5.521710473736134e-05, "loss": 0.0523, "step": 49340 }, { "grad_norm": 0.4255822002887726, "learning_rate": 5.5200660009431826e-05, "loss": 0.0417, "step": 49350 }, { "grad_norm": 0.435565322637558, "learning_rate": 5.518421471276596e-05, "loss": 0.0402, "step": 49360 }, { "grad_norm": 0.4254908561706543, "learning_rate": 5.5167768849162174e-05, "loss": 0.0357, "step": 49370 }, { "grad_norm": 0.5250228643417358, "learning_rate": 5.515132242041893e-05, "loss": 0.0436, "step": 49380 }, { "grad_norm": 0.40949803590774536, "learning_rate": 5.513487542833483e-05, "loss": 0.0427, "step": 49390 }, { "grad_norm": 0.48342201113700867, "learning_rate": 5.5118427874708466e-05, "loss": 0.0497, "step": 49400 }, { "grad_norm": 0.3954506516456604, "learning_rate": 5.510197976133853e-05, "loss": 0.0355, "step": 49410 }, { "grad_norm": 0.6176831722259521, "learning_rate": 5.508553109002376e-05, "loss": 0.0429, "step": 49420 }, { "grad_norm": 0.5740959644317627, "learning_rate": 5.5069081862562957e-05, "loss": 0.0432, "step": 49430 }, { "grad_norm": 0.406322717666626, "learning_rate": 5.5052632080754965e-05, "loss": 0.0436, "step": 49440 }, { "grad_norm": 2.1662466526031494, "learning_rate": 5.503618174639874e-05, "loss": 0.0506, "step": 49450 }, { "grad_norm": 0.6207571625709534, "learning_rate": 5.501973086129325e-05, "loss": 0.0403, "step": 49460 }, { "grad_norm": 0.7368949055671692, "learning_rate": 5.500327942723753e-05, "loss": 0.0396, "step": 49470 }, { "grad_norm": 0.4114035665988922, "learning_rate": 5.498682744603071e-05, "loss": 0.047, "step": 49480 }, { "grad_norm": 0.5214865207672119, "learning_rate": 5.497037491947194e-05, "loss": 0.0434, "step": 49490 }, { "grad_norm": 0.5148059725761414, "learning_rate": 5.4953921849360424e-05, "loss": 0.0512, "step": 49500 }, { "grad_norm": 0.38185498118400574, "learning_rate": 5.493746823749547e-05, "loss": 0.0469, "step": 49510 }, { "grad_norm": 0.5702016949653625, "learning_rate": 5.49210140856764e-05, "loss": 0.0476, "step": 49520 }, { "grad_norm": 0.3779994547367096, "learning_rate": 5.4904559395702647e-05, "loss": 0.0451, "step": 49530 }, { "grad_norm": 0.37519919872283936, "learning_rate": 5.488810416937364e-05, "loss": 0.0472, "step": 49540 }, { "grad_norm": 0.38249123096466064, "learning_rate": 5.4871648408488926e-05, "loss": 0.0469, "step": 49550 }, { "grad_norm": 0.7285451889038086, "learning_rate": 5.485519211484807e-05, "loss": 0.045, "step": 49560 }, { "grad_norm": 0.572478175163269, "learning_rate": 5.4838735290250676e-05, "loss": 0.0531, "step": 49570 }, { "grad_norm": 0.42123743891716003, "learning_rate": 5.482227793649649e-05, "loss": 0.0535, "step": 49580 }, { "grad_norm": 0.4864569306373596, "learning_rate": 5.480582005538524e-05, "loss": 0.0476, "step": 49590 }, { "grad_norm": 0.42679059505462646, "learning_rate": 5.478936164871671e-05, "loss": 0.0502, "step": 49600 }, { "grad_norm": 0.40770721435546875, "learning_rate": 5.477290271829081e-05, "loss": 0.0475, "step": 49610 }, { "grad_norm": 0.3828468918800354, "learning_rate": 5.475644326590744e-05, "loss": 0.0402, "step": 49620 }, { "grad_norm": 0.4929128885269165, "learning_rate": 5.473998329336658e-05, "loss": 0.0454, "step": 49630 }, { "grad_norm": 0.36983367800712585, "learning_rate": 5.4723522802468286e-05, "loss": 0.0408, "step": 49640 }, { "grad_norm": 0.964473307132721, "learning_rate": 5.470706179501264e-05, "loss": 0.0445, "step": 49650 }, { "grad_norm": 0.431944876909256, "learning_rate": 5.4690600272799776e-05, "loss": 0.052, "step": 49660 }, { "grad_norm": 0.8032799959182739, "learning_rate": 5.467413823762993e-05, "loss": 0.0575, "step": 49670 }, { "grad_norm": 0.40332889556884766, "learning_rate": 5.465767569130335e-05, "loss": 0.0461, "step": 49680 }, { "grad_norm": 0.8357374668121338, "learning_rate": 5.464121263562036e-05, "loss": 0.0468, "step": 49690 }, { "grad_norm": 0.3625034987926483, "learning_rate": 5.4624749072381343e-05, "loss": 0.0445, "step": 49700 }, { "grad_norm": 0.2932693362236023, "learning_rate": 5.460828500338672e-05, "loss": 0.0434, "step": 49710 }, { "grad_norm": 0.376954585313797, "learning_rate": 5.459182043043698e-05, "loss": 0.0451, "step": 49720 }, { "grad_norm": 0.3498789668083191, "learning_rate": 5.457535535533265e-05, "loss": 0.0442, "step": 49730 }, { "grad_norm": 0.5448550581932068, "learning_rate": 5.4558889779874334e-05, "loss": 0.0532, "step": 49740 }, { "grad_norm": 0.46261098980903625, "learning_rate": 5.454242370586269e-05, "loss": 0.0515, "step": 49750 }, { "grad_norm": 0.460105836391449, "learning_rate": 5.452595713509843e-05, "loss": 0.0368, "step": 49760 }, { "grad_norm": 0.46560919284820557, "learning_rate": 5.4509490069382274e-05, "loss": 0.0459, "step": 49770 }, { "grad_norm": 0.35117435455322266, "learning_rate": 5.449302251051509e-05, "loss": 0.0389, "step": 49780 }, { "grad_norm": 0.28698304295539856, "learning_rate": 5.44765544602977e-05, "loss": 0.0595, "step": 49790 }, { "grad_norm": 0.3111191391944885, "learning_rate": 5.446008592053107e-05, "loss": 0.0478, "step": 49800 }, { "grad_norm": 0.32714706659317017, "learning_rate": 5.4443616893016135e-05, "loss": 0.0376, "step": 49810 }, { "grad_norm": 0.50120609998703, "learning_rate": 5.4427147379553934e-05, "loss": 0.0504, "step": 49820 }, { "grad_norm": 0.382021427154541, "learning_rate": 5.441067738194556e-05, "loss": 0.0353, "step": 49830 }, { "grad_norm": 0.3894084692001343, "learning_rate": 5.439420690199214e-05, "loss": 0.0514, "step": 49840 }, { "grad_norm": 0.39243289828300476, "learning_rate": 5.437773594149484e-05, "loss": 0.0395, "step": 49850 }, { "grad_norm": 0.3734551668167114, "learning_rate": 5.436126450225495e-05, "loss": 0.0462, "step": 49860 }, { "grad_norm": 0.4019026458263397, "learning_rate": 5.434479258607371e-05, "loss": 0.0365, "step": 49870 }, { "grad_norm": 0.33877798914909363, "learning_rate": 5.43283201947525e-05, "loss": 0.0506, "step": 49880 }, { "grad_norm": 0.5172692537307739, "learning_rate": 5.431184733009268e-05, "loss": 0.0512, "step": 49890 }, { "grad_norm": 0.37825754284858704, "learning_rate": 5.4295373993895736e-05, "loss": 0.0484, "step": 49900 }, { "grad_norm": 0.3837696313858032, "learning_rate": 5.4278900187963157e-05, "loss": 0.0482, "step": 49910 }, { "grad_norm": 0.4429924190044403, "learning_rate": 5.426242591409646e-05, "loss": 0.0411, "step": 49920 }, { "grad_norm": 0.363731324672699, "learning_rate": 5.42459511740973e-05, "loss": 0.0497, "step": 49930 }, { "grad_norm": 0.5332508087158203, "learning_rate": 5.422947596976729e-05, "loss": 0.043, "step": 49940 }, { "grad_norm": 0.6481026411056519, "learning_rate": 5.4213000302908134e-05, "loss": 0.0445, "step": 49950 }, { "grad_norm": 0.32227376103401184, "learning_rate": 5.419652417532162e-05, "loss": 0.0436, "step": 49960 }, { "grad_norm": 0.7067036032676697, "learning_rate": 5.4180047588809534e-05, "loss": 0.0385, "step": 49970 }, { "grad_norm": 0.7831299304962158, "learning_rate": 5.4163570545173704e-05, "loss": 0.0453, "step": 49980 }, { "grad_norm": 0.4286963641643524, "learning_rate": 5.414709304621608e-05, "loss": 0.0425, "step": 49990 }, { "grad_norm": 0.4635007381439209, "learning_rate": 5.413061509373858e-05, "loss": 0.0461, "step": 50000 }, { "grad_norm": 0.38776934146881104, "learning_rate": 5.411413668954324e-05, "loss": 0.0431, "step": 50010 }, { "grad_norm": 0.3618409335613251, "learning_rate": 5.4097657835432083e-05, "loss": 0.0461, "step": 50020 }, { "grad_norm": 0.4115748405456543, "learning_rate": 5.408117853320723e-05, "loss": 0.0484, "step": 50030 }, { "grad_norm": 0.544844925403595, "learning_rate": 5.406469878467084e-05, "loss": 0.0346, "step": 50040 }, { "grad_norm": 1.104065179824829, "learning_rate": 5.404821859162509e-05, "loss": 0.0481, "step": 50050 }, { "grad_norm": 0.5366735458374023, "learning_rate": 5.403173795587225e-05, "loss": 0.0502, "step": 50060 }, { "grad_norm": 0.5117091536521912, "learning_rate": 5.4015256879214606e-05, "loss": 0.0498, "step": 50070 }, { "grad_norm": 0.38047969341278076, "learning_rate": 5.39987753634545e-05, "loss": 0.0414, "step": 50080 }, { "grad_norm": 0.29848712682724, "learning_rate": 5.3982293410394335e-05, "loss": 0.0571, "step": 50090 }, { "grad_norm": 0.37201976776123047, "learning_rate": 5.396581102183655e-05, "loss": 0.0419, "step": 50100 }, { "grad_norm": 0.5401666164398193, "learning_rate": 5.3949328199583634e-05, "loss": 0.0435, "step": 50110 }, { "grad_norm": 0.4988485872745514, "learning_rate": 5.393284494543813e-05, "loss": 0.051, "step": 50120 }, { "grad_norm": 0.3826623260974884, "learning_rate": 5.391636126120262e-05, "loss": 0.0409, "step": 50130 }, { "grad_norm": 0.5157546401023865, "learning_rate": 5.389987714867971e-05, "loss": 0.042, "step": 50140 }, { "grad_norm": 0.4322167634963989, "learning_rate": 5.3883392609672114e-05, "loss": 0.0448, "step": 50150 }, { "grad_norm": 0.4119834899902344, "learning_rate": 5.3866907645982545e-05, "loss": 0.0442, "step": 50160 }, { "grad_norm": 0.761766254901886, "learning_rate": 5.385042225941375e-05, "loss": 0.0487, "step": 50170 }, { "grad_norm": 0.3805169463157654, "learning_rate": 5.3833936451768585e-05, "loss": 0.0486, "step": 50180 }, { "grad_norm": 0.4330335557460785, "learning_rate": 5.3817450224849884e-05, "loss": 0.0493, "step": 50190 }, { "grad_norm": 0.5811826586723328, "learning_rate": 5.380096358046056e-05, "loss": 0.0576, "step": 50200 }, { "grad_norm": 0.40192341804504395, "learning_rate": 5.378447652040359e-05, "loss": 0.0441, "step": 50210 }, { "grad_norm": 0.3618273138999939, "learning_rate": 5.376798904648195e-05, "loss": 0.0337, "step": 50220 }, { "grad_norm": 0.44884994626045227, "learning_rate": 5.375150116049869e-05, "loss": 0.0356, "step": 50230 }, { "grad_norm": 0.5404634475708008, "learning_rate": 5.373501286425691e-05, "loss": 0.0352, "step": 50240 }, { "grad_norm": 0.43440327048301697, "learning_rate": 5.3718524159559726e-05, "loss": 0.0367, "step": 50250 }, { "grad_norm": 0.45911934971809387, "learning_rate": 5.370203504821034e-05, "loss": 0.047, "step": 50260 }, { "grad_norm": 0.24376064538955688, "learning_rate": 5.368554553201196e-05, "loss": 0.0483, "step": 50270 }, { "grad_norm": 0.5267446637153625, "learning_rate": 5.3669055612767874e-05, "loss": 0.0637, "step": 50280 }, { "grad_norm": 0.5271745920181274, "learning_rate": 5.3652565292281376e-05, "loss": 0.044, "step": 50290 }, { "grad_norm": 0.6104482412338257, "learning_rate": 5.3636074572355824e-05, "loss": 0.0457, "step": 50300 }, { "grad_norm": 0.4477216601371765, "learning_rate": 5.361958345479463e-05, "loss": 0.046, "step": 50310 }, { "grad_norm": 0.3411906361579895, "learning_rate": 5.360309194140123e-05, "loss": 0.0394, "step": 50320 }, { "grad_norm": 0.6810415983200073, "learning_rate": 5.3586600033979106e-05, "loss": 0.0405, "step": 50330 }, { "grad_norm": 0.3943089246749878, "learning_rate": 5.357010773433181e-05, "loss": 0.0424, "step": 50340 }, { "grad_norm": 0.477078914642334, "learning_rate": 5.3553615044262895e-05, "loss": 0.0329, "step": 50350 }, { "grad_norm": 0.3839012682437897, "learning_rate": 5.3537121965575964e-05, "loss": 0.0428, "step": 50360 }, { "grad_norm": 0.5484325289726257, "learning_rate": 5.352062850007471e-05, "loss": 0.0587, "step": 50370 }, { "grad_norm": 0.22804486751556396, "learning_rate": 5.350413464956282e-05, "loss": 0.0407, "step": 50380 }, { "grad_norm": 0.6990364193916321, "learning_rate": 5.348764041584403e-05, "loss": 0.0479, "step": 50390 }, { "grad_norm": 0.5389370322227478, "learning_rate": 5.347114580072212e-05, "loss": 0.0455, "step": 50400 }, { "grad_norm": 2.817976236343384, "learning_rate": 5.345465080600093e-05, "loss": 0.0467, "step": 50410 }, { "grad_norm": 0.403915673494339, "learning_rate": 5.3438155433484326e-05, "loss": 0.045, "step": 50420 }, { "grad_norm": 0.7021260261535645, "learning_rate": 5.3421659684976197e-05, "loss": 0.0467, "step": 50430 }, { "grad_norm": 0.8190227150917053, "learning_rate": 5.340516356228052e-05, "loss": 0.0477, "step": 50440 }, { "grad_norm": 0.3883157968521118, "learning_rate": 5.338866706720128e-05, "loss": 0.0481, "step": 50450 }, { "grad_norm": 0.5660489797592163, "learning_rate": 5.337217020154249e-05, "loss": 0.0461, "step": 50460 }, { "grad_norm": 0.5278846025466919, "learning_rate": 5.335567296710825e-05, "loss": 0.0449, "step": 50470 }, { "grad_norm": 0.4902447760105133, "learning_rate": 5.333917536570265e-05, "loss": 0.0452, "step": 50480 }, { "grad_norm": 0.6076785922050476, "learning_rate": 5.332267739912986e-05, "loss": 0.0463, "step": 50490 }, { "grad_norm": 0.4000087380409241, "learning_rate": 5.330617906919405e-05, "loss": 0.0485, "step": 50500 }, { "grad_norm": 0.4447571635246277, "learning_rate": 5.328968037769949e-05, "loss": 0.0431, "step": 50510 }, { "grad_norm": 0.4053336977958679, "learning_rate": 5.327318132645042e-05, "loss": 0.0459, "step": 50520 }, { "grad_norm": 0.46819284558296204, "learning_rate": 5.325668191725116e-05, "loss": 0.0357, "step": 50530 }, { "grad_norm": 0.45024538040161133, "learning_rate": 5.324018215190606e-05, "loss": 0.0474, "step": 50540 }, { "grad_norm": 0.5789436101913452, "learning_rate": 5.3223682032219515e-05, "loss": 0.0434, "step": 50550 }, { "grad_norm": 0.2992861866950989, "learning_rate": 5.320718155999595e-05, "loss": 0.0479, "step": 50560 }, { "grad_norm": 0.37842506170272827, "learning_rate": 5.3190680737039835e-05, "loss": 0.0561, "step": 50570 }, { "grad_norm": 0.35805565118789673, "learning_rate": 5.317417956515567e-05, "loss": 0.0368, "step": 50580 }, { "grad_norm": 0.3899225890636444, "learning_rate": 5.315767804614798e-05, "loss": 0.0399, "step": 50590 }, { "grad_norm": 0.5111238360404968, "learning_rate": 5.3141176181821395e-05, "loss": 0.0423, "step": 50600 }, { "grad_norm": 0.3537796437740326, "learning_rate": 5.31246739739805e-05, "loss": 0.0391, "step": 50610 }, { "grad_norm": 0.46785968542099, "learning_rate": 5.310817142442994e-05, "loss": 0.0566, "step": 50620 }, { "grad_norm": 0.6767316460609436, "learning_rate": 5.309166853497445e-05, "loss": 0.0469, "step": 50630 }, { "grad_norm": 0.46818748116493225, "learning_rate": 5.307516530741873e-05, "loss": 0.051, "step": 50640 }, { "grad_norm": 0.6144831776618958, "learning_rate": 5.305866174356754e-05, "loss": 0.0497, "step": 50650 }, { "grad_norm": 0.47151631116867065, "learning_rate": 5.304215784522571e-05, "loss": 0.0388, "step": 50660 }, { "grad_norm": 0.5887369513511658, "learning_rate": 5.302565361419808e-05, "loss": 0.0589, "step": 50670 }, { "grad_norm": 0.38371455669403076, "learning_rate": 5.3009149052289507e-05, "loss": 0.0423, "step": 50680 }, { "grad_norm": 0.4414808750152588, "learning_rate": 5.299264416130493e-05, "loss": 0.0441, "step": 50690 }, { "grad_norm": 0.5897269248962402, "learning_rate": 5.297613894304928e-05, "loss": 0.0408, "step": 50700 }, { "grad_norm": 0.49959930777549744, "learning_rate": 5.2959633399327534e-05, "loss": 0.0386, "step": 50710 }, { "grad_norm": 0.43061724305152893, "learning_rate": 5.294312753194476e-05, "loss": 0.0349, "step": 50720 }, { "grad_norm": 0.5477217435836792, "learning_rate": 5.292662134270596e-05, "loss": 0.0434, "step": 50730 }, { "grad_norm": 0.3992651104927063, "learning_rate": 5.291011483341626e-05, "loss": 0.0446, "step": 50740 }, { "grad_norm": 1.0045311450958252, "learning_rate": 5.2893608005880767e-05, "loss": 0.0592, "step": 50750 }, { "grad_norm": 0.36467039585113525, "learning_rate": 5.287710086190467e-05, "loss": 0.0441, "step": 50760 }, { "grad_norm": 0.35232648253440857, "learning_rate": 5.286059340329314e-05, "loss": 0.043, "step": 50770 }, { "grad_norm": 0.8951385021209717, "learning_rate": 5.284408563185141e-05, "loss": 0.0605, "step": 50780 }, { "grad_norm": 0.41552993655204773, "learning_rate": 5.282757754938476e-05, "loss": 0.0476, "step": 50790 }, { "grad_norm": 0.5699262022972107, "learning_rate": 5.281106915769849e-05, "loss": 0.0535, "step": 50800 }, { "grad_norm": 0.7962037324905396, "learning_rate": 5.2794560458597897e-05, "loss": 0.0504, "step": 50810 }, { "grad_norm": 0.4130844473838806, "learning_rate": 5.27780514538884e-05, "loss": 0.0374, "step": 50820 }, { "grad_norm": 0.535386323928833, "learning_rate": 5.2761542145375365e-05, "loss": 0.0541, "step": 50830 }, { "grad_norm": 0.7011036276817322, "learning_rate": 5.274503253486421e-05, "loss": 0.0387, "step": 50840 }, { "grad_norm": 0.6482967138290405, "learning_rate": 5.272852262416046e-05, "loss": 0.0492, "step": 50850 }, { "grad_norm": 0.45669522881507874, "learning_rate": 5.2712012415069555e-05, "loss": 0.0477, "step": 50860 }, { "grad_norm": 0.5947309732437134, "learning_rate": 5.269550190939705e-05, "loss": 0.0455, "step": 50870 }, { "grad_norm": 0.4956717789173126, "learning_rate": 5.267899110894852e-05, "loss": 0.0391, "step": 50880 }, { "grad_norm": 0.3379408121109009, "learning_rate": 5.266248001552955e-05, "loss": 0.0458, "step": 50890 }, { "grad_norm": 0.34320294857025146, "learning_rate": 5.264596863094575e-05, "loss": 0.0502, "step": 50900 }, { "grad_norm": 0.7620213031768799, "learning_rate": 5.262945695700282e-05, "loss": 0.0441, "step": 50910 }, { "grad_norm": 0.9475850462913513, "learning_rate": 5.261294499550643e-05, "loss": 0.0452, "step": 50920 }, { "grad_norm": 0.42556220293045044, "learning_rate": 5.25964327482623e-05, "loss": 0.0394, "step": 50930 }, { "grad_norm": 0.5567404627799988, "learning_rate": 5.257992021707617e-05, "loss": 0.0439, "step": 50940 }, { "grad_norm": 1.5892176628112793, "learning_rate": 5.256340740375387e-05, "loss": 0.0491, "step": 50950 }, { "grad_norm": 0.40696609020233154, "learning_rate": 5.254689431010117e-05, "loss": 0.0445, "step": 50960 }, { "grad_norm": 0.40118342638015747, "learning_rate": 5.253038093792395e-05, "loss": 0.0363, "step": 50970 }, { "grad_norm": 0.37561386823654175, "learning_rate": 5.251386728902806e-05, "loss": 0.0423, "step": 50980 }, { "grad_norm": 0.4031572639942169, "learning_rate": 5.2497353365219446e-05, "loss": 0.0484, "step": 50990 }, { "grad_norm": 0.3544655442237854, "learning_rate": 5.2480839168304e-05, "loss": 0.0362, "step": 51000 }, { "grad_norm": 1.2084527015686035, "learning_rate": 5.246432470008773e-05, "loss": 0.0387, "step": 51010 }, { "grad_norm": 0.6859352588653564, "learning_rate": 5.2447809962376616e-05, "loss": 0.0513, "step": 51020 }, { "grad_norm": 0.6318339109420776, "learning_rate": 5.243129495697666e-05, "loss": 0.0347, "step": 51030 }, { "grad_norm": 0.8795046210289001, "learning_rate": 5.241477968569397e-05, "loss": 0.0572, "step": 51040 }, { "grad_norm": 0.5514506101608276, "learning_rate": 5.23982641503346e-05, "loss": 0.0469, "step": 51050 }, { "grad_norm": 0.6153069138526917, "learning_rate": 5.238174835270465e-05, "loss": 0.0438, "step": 51060 }, { "grad_norm": 0.379794180393219, "learning_rate": 5.23652322946103e-05, "loss": 0.0347, "step": 51070 }, { "grad_norm": 0.5213199257850647, "learning_rate": 5.234871597785771e-05, "loss": 0.0378, "step": 51080 }, { "grad_norm": 0.7433441281318665, "learning_rate": 5.233219940425306e-05, "loss": 0.0526, "step": 51090 }, { "grad_norm": 1.268591046333313, "learning_rate": 5.2315682575602576e-05, "loss": 0.0739, "step": 51100 }, { "grad_norm": 0.6513797044754028, "learning_rate": 5.229916549371254e-05, "loss": 0.0402, "step": 51110 }, { "grad_norm": 0.38369020819664, "learning_rate": 5.228264816038924e-05, "loss": 0.0436, "step": 51120 }, { "grad_norm": 0.5503397583961487, "learning_rate": 5.2266130577438935e-05, "loss": 0.0374, "step": 51130 }, { "grad_norm": 0.38049888610839844, "learning_rate": 5.224961274666802e-05, "loss": 0.0387, "step": 51140 }, { "grad_norm": 0.6831022500991821, "learning_rate": 5.223309466988283e-05, "loss": 0.0464, "step": 51150 }, { "grad_norm": 0.46948859095573425, "learning_rate": 5.221657634888976e-05, "loss": 0.0586, "step": 51160 }, { "grad_norm": 0.5525541305541992, "learning_rate": 5.2200057785495235e-05, "loss": 0.059, "step": 51170 }, { "grad_norm": 0.944000780582428, "learning_rate": 5.2183538981505706e-05, "loss": 0.0481, "step": 51180 }, { "grad_norm": 0.34698089957237244, "learning_rate": 5.216701993872762e-05, "loss": 0.0424, "step": 51190 }, { "grad_norm": 0.5133081674575806, "learning_rate": 5.215050065896752e-05, "loss": 0.0423, "step": 51200 }, { "grad_norm": 0.41031399369239807, "learning_rate": 5.2133981144031875e-05, "loss": 0.0593, "step": 51210 }, { "grad_norm": 0.9861788749694824, "learning_rate": 5.211746139572727e-05, "loss": 0.0507, "step": 51220 }, { "grad_norm": 0.35426339507102966, "learning_rate": 5.210094141586026e-05, "loss": 0.0398, "step": 51230 }, { "grad_norm": 0.8320302367210388, "learning_rate": 5.2084421206237476e-05, "loss": 0.0577, "step": 51240 }, { "grad_norm": 0.38457268476486206, "learning_rate": 5.20679007686655e-05, "loss": 0.051, "step": 51250 }, { "grad_norm": 0.47946789860725403, "learning_rate": 5.205138010495102e-05, "loss": 0.0494, "step": 51260 }, { "grad_norm": 0.43398234248161316, "learning_rate": 5.20348592169007e-05, "loss": 0.045, "step": 51270 }, { "grad_norm": 0.8362836241722107, "learning_rate": 5.201833810632123e-05, "loss": 0.0408, "step": 51280 }, { "grad_norm": 0.48460787534713745, "learning_rate": 5.200181677501933e-05, "loss": 0.0502, "step": 51290 }, { "grad_norm": 0.3171692490577698, "learning_rate": 5.1985295224801777e-05, "loss": 0.0502, "step": 51300 }, { "grad_norm": 0.46644115447998047, "learning_rate": 5.196877345747531e-05, "loss": 0.0528, "step": 51310 }, { "grad_norm": 0.6887357831001282, "learning_rate": 5.195225147484674e-05, "loss": 0.0477, "step": 51320 }, { "grad_norm": 0.7561548948287964, "learning_rate": 5.193572927872288e-05, "loss": 0.0473, "step": 51330 }, { "grad_norm": 0.5401852130889893, "learning_rate": 5.191920687091059e-05, "loss": 0.0434, "step": 51340 }, { "grad_norm": 0.38162535429000854, "learning_rate": 5.1902684253216695e-05, "loss": 0.0434, "step": 51350 }, { "grad_norm": 0.30576616525650024, "learning_rate": 5.1886161427448134e-05, "loss": 0.0442, "step": 51360 }, { "grad_norm": 0.32764142751693726, "learning_rate": 5.186963839541179e-05, "loss": 0.0377, "step": 51370 }, { "grad_norm": 0.5377243161201477, "learning_rate": 5.185311515891459e-05, "loss": 0.034, "step": 51380 }, { "grad_norm": 0.7958558201789856, "learning_rate": 5.1836591719763516e-05, "loss": 0.039, "step": 51390 }, { "grad_norm": 0.44714534282684326, "learning_rate": 5.182006807976554e-05, "loss": 0.0477, "step": 51400 }, { "grad_norm": 0.2971920967102051, "learning_rate": 5.180354424072763e-05, "loss": 0.0365, "step": 51410 }, { "grad_norm": 0.5932517051696777, "learning_rate": 5.178702020445686e-05, "loss": 0.0456, "step": 51420 }, { "grad_norm": 0.3979111611843109, "learning_rate": 5.177049597276024e-05, "loss": 0.0486, "step": 51430 }, { "grad_norm": 0.4933662712574005, "learning_rate": 5.175397154744482e-05, "loss": 0.0461, "step": 51440 }, { "grad_norm": 0.7824925184249878, "learning_rate": 5.1737446930317736e-05, "loss": 0.0384, "step": 51450 }, { "grad_norm": 0.6436455845832825, "learning_rate": 5.172092212318604e-05, "loss": 0.0514, "step": 51460 }, { "grad_norm": 0.6985926628112793, "learning_rate": 5.170439712785692e-05, "loss": 0.0489, "step": 51470 }, { "grad_norm": 0.6058222055435181, "learning_rate": 5.1687871946137467e-05, "loss": 0.0421, "step": 51480 }, { "grad_norm": 0.34733089804649353, "learning_rate": 5.16713465798349e-05, "loss": 0.0385, "step": 51490 }, { "grad_norm": 0.4902783930301666, "learning_rate": 5.165482103075637e-05, "loss": 0.0569, "step": 51500 }, { "grad_norm": 0.31673309206962585, "learning_rate": 5.163829530070909e-05, "loss": 0.0389, "step": 51510 }, { "grad_norm": 0.5053362846374512, "learning_rate": 5.162176939150032e-05, "loss": 0.0429, "step": 51520 }, { "grad_norm": 1.4258555173873901, "learning_rate": 5.160524330493728e-05, "loss": 0.0431, "step": 51530 }, { "grad_norm": 0.3994423747062683, "learning_rate": 5.158871704282724e-05, "loss": 0.0492, "step": 51540 }, { "grad_norm": 0.6999521851539612, "learning_rate": 5.15721906069775e-05, "loss": 0.0533, "step": 51550 }, { "grad_norm": 0.39191001653671265, "learning_rate": 5.155566399919536e-05, "loss": 0.0403, "step": 51560 }, { "grad_norm": 0.5360322594642639, "learning_rate": 5.153913722128813e-05, "loss": 0.0424, "step": 51570 }, { "grad_norm": 0.6556647419929504, "learning_rate": 5.1522610275063176e-05, "loss": 0.0446, "step": 51580 }, { "grad_norm": 0.3640734851360321, "learning_rate": 5.150608316232786e-05, "loss": 0.0381, "step": 51590 }, { "grad_norm": 0.6192983984947205, "learning_rate": 5.148955588488955e-05, "loss": 0.0349, "step": 51600 }, { "grad_norm": 0.31768375635147095, "learning_rate": 5.147302844455564e-05, "loss": 0.0395, "step": 51610 }, { "grad_norm": 0.40398865938186646, "learning_rate": 5.145650084313356e-05, "loss": 0.0439, "step": 51620 }, { "grad_norm": 0.46158039569854736, "learning_rate": 5.1439973082430734e-05, "loss": 0.0437, "step": 51630 }, { "grad_norm": 0.30717456340789795, "learning_rate": 5.142344516425461e-05, "loss": 0.0376, "step": 51640 }, { "grad_norm": 0.3127287030220032, "learning_rate": 5.1406917090412673e-05, "loss": 0.0541, "step": 51650 }, { "grad_norm": 0.43956461548805237, "learning_rate": 5.1390388862712404e-05, "loss": 0.0375, "step": 51660 }, { "grad_norm": 0.582261323928833, "learning_rate": 5.137386048296129e-05, "loss": 0.0471, "step": 51670 }, { "grad_norm": 0.4253338575363159, "learning_rate": 5.135733195296686e-05, "loss": 0.0517, "step": 51680 }, { "grad_norm": 0.3006550967693329, "learning_rate": 5.134080327453664e-05, "loss": 0.0434, "step": 51690 }, { "grad_norm": 0.3203488886356354, "learning_rate": 5.1324274449478204e-05, "loss": 0.0456, "step": 51700 }, { "grad_norm": 0.4371453523635864, "learning_rate": 5.130774547959909e-05, "loss": 0.0523, "step": 51710 }, { "grad_norm": 0.42448654770851135, "learning_rate": 5.129121636670692e-05, "loss": 0.0468, "step": 51720 }, { "grad_norm": 0.23206253349781036, "learning_rate": 5.1274687112609246e-05, "loss": 0.0264, "step": 51730 }, { "grad_norm": 0.38212108612060547, "learning_rate": 5.125815771911372e-05, "loss": 0.042, "step": 51740 }, { "grad_norm": 0.5946088433265686, "learning_rate": 5.1241628188027956e-05, "loss": 0.0358, "step": 51750 }, { "grad_norm": 0.5192737579345703, "learning_rate": 5.1225098521159585e-05, "loss": 0.0389, "step": 51760 }, { "grad_norm": 0.43993517756462097, "learning_rate": 5.12085687203163e-05, "loss": 0.0359, "step": 51770 }, { "grad_norm": 0.7048720121383667, "learning_rate": 5.119203878730575e-05, "loss": 0.0535, "step": 51780 }, { "grad_norm": 0.37417346239089966, "learning_rate": 5.117550872393564e-05, "loss": 0.0484, "step": 51790 }, { "grad_norm": 0.6237667202949524, "learning_rate": 5.1158978532013645e-05, "loss": 0.0516, "step": 51800 }, { "grad_norm": 0.4855859875679016, "learning_rate": 5.114244821334751e-05, "loss": 0.0511, "step": 51810 }, { "grad_norm": 0.455463707447052, "learning_rate": 5.112591776974497e-05, "loss": 0.0392, "step": 51820 }, { "grad_norm": 0.43562981486320496, "learning_rate": 5.110938720301374e-05, "loss": 0.0397, "step": 51830 }, { "grad_norm": 0.2376970797777176, "learning_rate": 5.1092856514961595e-05, "loss": 0.0475, "step": 51840 }, { "grad_norm": 0.4890397787094116, "learning_rate": 5.107632570739631e-05, "loss": 0.0425, "step": 51850 }, { "grad_norm": 0.34243685007095337, "learning_rate": 5.1059794782125645e-05, "loss": 0.054, "step": 51860 }, { "grad_norm": 0.6622103452682495, "learning_rate": 5.104326374095744e-05, "loss": 0.0544, "step": 51870 }, { "grad_norm": 0.33528003096580505, "learning_rate": 5.102673258569947e-05, "loss": 0.0445, "step": 51880 }, { "grad_norm": 0.7202751040458679, "learning_rate": 5.101020131815956e-05, "loss": 0.0482, "step": 51890 }, { "grad_norm": 0.640747606754303, "learning_rate": 5.099366994014556e-05, "loss": 0.0408, "step": 51900 }, { "grad_norm": 0.5559027194976807, "learning_rate": 5.097713845346532e-05, "loss": 0.0464, "step": 51910 }, { "grad_norm": 0.6673837900161743, "learning_rate": 5.096060685992668e-05, "loss": 0.0355, "step": 51920 }, { "grad_norm": 0.6827932596206665, "learning_rate": 5.0944075161337514e-05, "loss": 0.0581, "step": 51930 }, { "grad_norm": 1.8648685216903687, "learning_rate": 5.09275433595057e-05, "loss": 0.0451, "step": 51940 }, { "grad_norm": 0.6336968541145325, "learning_rate": 5.0911011456239157e-05, "loss": 0.0435, "step": 51950 }, { "grad_norm": 0.6298949122428894, "learning_rate": 5.089447945334574e-05, "loss": 0.0437, "step": 51960 }, { "grad_norm": 0.4096868336200714, "learning_rate": 5.087794735263343e-05, "loss": 0.0367, "step": 51970 }, { "grad_norm": 0.3604183495044708, "learning_rate": 5.086141515591011e-05, "loss": 0.0412, "step": 51980 }, { "grad_norm": 0.4709017872810364, "learning_rate": 5.084488286498371e-05, "loss": 0.0378, "step": 51990 }, { "grad_norm": 0.45936480164527893, "learning_rate": 5.08283504816622e-05, "loss": 0.06, "step": 52000 }, { "grad_norm": 0.3890804052352905, "learning_rate": 5.081181800775353e-05, "loss": 0.0477, "step": 52010 }, { "grad_norm": 0.4225451648235321, "learning_rate": 5.079528544506565e-05, "loss": 0.0501, "step": 52020 }, { "grad_norm": 0.3914782702922821, "learning_rate": 5.077875279540657e-05, "loss": 0.0384, "step": 52030 }, { "grad_norm": 0.26935774087905884, "learning_rate": 5.076222006058424e-05, "loss": 0.0487, "step": 52040 }, { "grad_norm": 0.36714690923690796, "learning_rate": 5.074568724240667e-05, "loss": 0.0326, "step": 52050 }, { "grad_norm": 0.32355594635009766, "learning_rate": 5.072915434268187e-05, "loss": 0.0403, "step": 52060 }, { "grad_norm": 0.286207914352417, "learning_rate": 5.071262136321785e-05, "loss": 0.0432, "step": 52070 }, { "grad_norm": 0.8129168152809143, "learning_rate": 5.069608830582262e-05, "loss": 0.0458, "step": 52080 }, { "grad_norm": 0.2114054560661316, "learning_rate": 5.067955517230423e-05, "loss": 0.0456, "step": 52090 }, { "grad_norm": 0.36830934882164, "learning_rate": 5.06630219644707e-05, "loss": 0.0489, "step": 52100 }, { "grad_norm": 0.47867804765701294, "learning_rate": 5.064648868413008e-05, "loss": 0.0328, "step": 52110 }, { "grad_norm": 0.47120991349220276, "learning_rate": 5.062995533309044e-05, "loss": 0.0395, "step": 52120 }, { "grad_norm": 0.39382204413414, "learning_rate": 5.061342191315983e-05, "loss": 0.0544, "step": 52130 }, { "grad_norm": 0.5664030909538269, "learning_rate": 5.0596888426146325e-05, "loss": 0.0468, "step": 52140 }, { "grad_norm": 0.4499123990535736, "learning_rate": 5.058035487385798e-05, "loss": 0.0417, "step": 52150 }, { "grad_norm": 0.7189743518829346, "learning_rate": 5.0563821258102915e-05, "loss": 0.0496, "step": 52160 }, { "grad_norm": 0.6393262147903442, "learning_rate": 5.05472875806892e-05, "loss": 0.0506, "step": 52170 }, { "grad_norm": 0.4794674813747406, "learning_rate": 5.053075384342494e-05, "loss": 0.05, "step": 52180 }, { "grad_norm": 0.37221741676330566, "learning_rate": 5.0514220048118234e-05, "loss": 0.047, "step": 52190 }, { "grad_norm": 0.4283928871154785, "learning_rate": 5.049768619657721e-05, "loss": 0.0519, "step": 52200 }, { "grad_norm": 0.3408101499080658, "learning_rate": 5.048115229060996e-05, "loss": 0.0424, "step": 52210 }, { "grad_norm": 0.8666333556175232, "learning_rate": 5.046461833202462e-05, "loss": 0.0408, "step": 52220 }, { "grad_norm": 1.3481078147888184, "learning_rate": 5.044808432262932e-05, "loss": 0.0506, "step": 52230 }, { "grad_norm": 0.37874165177345276, "learning_rate": 5.043155026423219e-05, "loss": 0.0419, "step": 52240 }, { "grad_norm": 0.38142555952072144, "learning_rate": 5.041501615864137e-05, "loss": 0.0381, "step": 52250 }, { "grad_norm": 0.37232670187950134, "learning_rate": 5.039848200766502e-05, "loss": 0.0355, "step": 52260 }, { "grad_norm": 0.5390467643737793, "learning_rate": 5.038194781311125e-05, "loss": 0.0459, "step": 52270 }, { "grad_norm": 0.33522024750709534, "learning_rate": 5.0365413576788264e-05, "loss": 0.0325, "step": 52280 }, { "grad_norm": 0.5365256667137146, "learning_rate": 5.0348879300504194e-05, "loss": 0.0377, "step": 52290 }, { "grad_norm": 0.37365221977233887, "learning_rate": 5.033234498606721e-05, "loss": 0.0373, "step": 52300 }, { "grad_norm": 0.6024498343467712, "learning_rate": 5.0315810635285476e-05, "loss": 0.0534, "step": 52310 }, { "grad_norm": 0.5122926235198975, "learning_rate": 5.0299276249967164e-05, "loss": 0.044, "step": 52320 }, { "grad_norm": 0.4786442518234253, "learning_rate": 5.0282741831920454e-05, "loss": 0.0516, "step": 52330 }, { "grad_norm": 0.49717357754707336, "learning_rate": 5.026620738295351e-05, "loss": 0.0397, "step": 52340 }, { "grad_norm": 0.43697261810302734, "learning_rate": 5.024967290487455e-05, "loss": 0.036, "step": 52350 }, { "grad_norm": 0.6103284955024719, "learning_rate": 5.023313839949172e-05, "loss": 0.0445, "step": 52360 }, { "grad_norm": 0.3836541473865509, "learning_rate": 5.0216603868613224e-05, "loss": 0.0404, "step": 52370 }, { "grad_norm": 0.4724642336368561, "learning_rate": 5.020006931404726e-05, "loss": 0.0521, "step": 52380 }, { "grad_norm": 0.4742184579372406, "learning_rate": 5.0183534737602014e-05, "loss": 0.0313, "step": 52390 }, { "grad_norm": 0.38664710521698, "learning_rate": 5.0167000141085686e-05, "loss": 0.0381, "step": 52400 }, { "grad_norm": 0.5368133187294006, "learning_rate": 5.015046552630647e-05, "loss": 0.0397, "step": 52410 }, { "grad_norm": 0.5256296396255493, "learning_rate": 5.013393089507258e-05, "loss": 0.0385, "step": 52420 }, { "grad_norm": 0.6117122173309326, "learning_rate": 5.011739624919219e-05, "loss": 0.0376, "step": 52430 }, { "grad_norm": 0.5104063153266907, "learning_rate": 5.0100861590473545e-05, "loss": 0.0448, "step": 52440 }, { "grad_norm": 0.38622409105300903, "learning_rate": 5.008432692072482e-05, "loss": 0.0356, "step": 52450 }, { "grad_norm": 0.6879815459251404, "learning_rate": 5.006779224175423e-05, "loss": 0.0629, "step": 52460 }, { "grad_norm": 0.5003141760826111, "learning_rate": 5.005125755536998e-05, "loss": 0.0505, "step": 52470 }, { "grad_norm": 0.9840788841247559, "learning_rate": 5.003472286338029e-05, "loss": 0.0527, "step": 52480 }, { "grad_norm": 0.3497917950153351, "learning_rate": 5.001818816759335e-05, "loss": 0.0667, "step": 52490 }, { "grad_norm": 0.9743133187294006, "learning_rate": 5.000165346981738e-05, "loss": 0.0441, "step": 52500 }, { "grad_norm": 0.5497400760650635, "learning_rate": 4.9985118771860595e-05, "loss": 0.0439, "step": 52510 }, { "grad_norm": 0.6020117998123169, "learning_rate": 4.996858407553119e-05, "loss": 0.0473, "step": 52520 }, { "grad_norm": 0.32824477553367615, "learning_rate": 4.99520493826374e-05, "loss": 0.0435, "step": 52530 }, { "grad_norm": 0.4496435523033142, "learning_rate": 4.9935514694987384e-05, "loss": 0.0524, "step": 52540 }, { "grad_norm": 0.5757879614830017, "learning_rate": 4.991898001438941e-05, "loss": 0.0494, "step": 52550 }, { "grad_norm": 0.42742398381233215, "learning_rate": 4.990244534265166e-05, "loss": 0.0413, "step": 52560 }, { "grad_norm": 0.5374178290367126, "learning_rate": 4.988591068158234e-05, "loss": 0.0545, "step": 52570 }, { "grad_norm": 0.5148611664772034, "learning_rate": 4.986937603298965e-05, "loss": 0.0491, "step": 52580 }, { "grad_norm": 0.5288200974464417, "learning_rate": 4.985284139868181e-05, "loss": 0.0478, "step": 52590 }, { "grad_norm": 0.4053238332271576, "learning_rate": 4.9836306780467e-05, "loss": 0.0423, "step": 52600 }, { "grad_norm": 0.5310356020927429, "learning_rate": 4.981977218015343e-05, "loss": 0.0447, "step": 52610 }, { "grad_norm": 0.6497471332550049, "learning_rate": 4.980323759954932e-05, "loss": 0.044, "step": 52620 }, { "grad_norm": 0.5653111934661865, "learning_rate": 4.978670304046284e-05, "loss": 0.0394, "step": 52630 }, { "grad_norm": 0.5311776995658875, "learning_rate": 4.9770168504702195e-05, "loss": 0.0429, "step": 52640 }, { "grad_norm": 0.5631177425384521, "learning_rate": 4.975363399407559e-05, "loss": 0.0506, "step": 52650 }, { "grad_norm": 0.7440244555473328, "learning_rate": 4.9737099510391194e-05, "loss": 0.0492, "step": 52660 }, { "grad_norm": 0.3922041952610016, "learning_rate": 4.972056505545719e-05, "loss": 0.0474, "step": 52670 }, { "grad_norm": 0.5054014325141907, "learning_rate": 4.9704030631081775e-05, "loss": 0.0398, "step": 52680 }, { "grad_norm": 0.516917884349823, "learning_rate": 4.968749623907313e-05, "loss": 0.0354, "step": 52690 }, { "grad_norm": 0.357194185256958, "learning_rate": 4.967096188123943e-05, "loss": 0.0473, "step": 52700 }, { "grad_norm": 0.3881068825721741, "learning_rate": 4.965442755938884e-05, "loss": 0.039, "step": 52710 }, { "grad_norm": 0.5389822721481323, "learning_rate": 4.963789327532954e-05, "loss": 0.0486, "step": 52720 }, { "grad_norm": 0.4507213234901428, "learning_rate": 4.962135903086968e-05, "loss": 0.0562, "step": 52730 }, { "grad_norm": 0.40465763211250305, "learning_rate": 4.960482482781741e-05, "loss": 0.0386, "step": 52740 }, { "grad_norm": 0.6488927602767944, "learning_rate": 4.958829066798092e-05, "loss": 0.0447, "step": 52750 }, { "grad_norm": 0.4462316632270813, "learning_rate": 4.957175655316835e-05, "loss": 0.0413, "step": 52760 }, { "grad_norm": 0.3651719093322754, "learning_rate": 4.9555222485187834e-05, "loss": 0.045, "step": 52770 }, { "grad_norm": 0.32725635170936584, "learning_rate": 4.953868846584752e-05, "loss": 0.0373, "step": 52780 }, { "grad_norm": 0.4279814064502716, "learning_rate": 4.9522154496955544e-05, "loss": 0.044, "step": 52790 }, { "grad_norm": 0.32946857810020447, "learning_rate": 4.950562058032001e-05, "loss": 0.0511, "step": 52800 }, { "grad_norm": 0.4815588593482971, "learning_rate": 4.948908671774909e-05, "loss": 0.0468, "step": 52810 }, { "grad_norm": 0.3294120132923126, "learning_rate": 4.947255291105087e-05, "loss": 0.0454, "step": 52820 }, { "grad_norm": 0.5265173316001892, "learning_rate": 4.945601916203348e-05, "loss": 0.0469, "step": 52830 }, { "grad_norm": 0.3939288854598999, "learning_rate": 4.9439485472505016e-05, "loss": 0.0424, "step": 52840 }, { "grad_norm": 0.44797128438949585, "learning_rate": 4.942295184427356e-05, "loss": 0.0446, "step": 52850 }, { "grad_norm": 0.5279996991157532, "learning_rate": 4.9406418279147234e-05, "loss": 0.0479, "step": 52860 }, { "grad_norm": 0.5297189354896545, "learning_rate": 4.9389884778934116e-05, "loss": 0.0473, "step": 52870 }, { "grad_norm": 0.514297604560852, "learning_rate": 4.937335134544228e-05, "loss": 0.041, "step": 52880 }, { "grad_norm": 0.2859746515750885, "learning_rate": 4.935681798047979e-05, "loss": 0.0631, "step": 52890 }, { "grad_norm": 0.5118865370750427, "learning_rate": 4.934028468585473e-05, "loss": 0.0421, "step": 52900 }, { "grad_norm": 0.5103829503059387, "learning_rate": 4.932375146337514e-05, "loss": 0.0466, "step": 52910 }, { "grad_norm": 0.46904274821281433, "learning_rate": 4.930721831484908e-05, "loss": 0.0394, "step": 52920 }, { "grad_norm": 0.8143083453178406, "learning_rate": 4.929068524208456e-05, "loss": 0.0519, "step": 52930 }, { "grad_norm": 0.5593643188476562, "learning_rate": 4.927415224688965e-05, "loss": 0.054, "step": 52940 }, { "grad_norm": 0.3258592188358307, "learning_rate": 4.9257619331072375e-05, "loss": 0.0468, "step": 52950 }, { "grad_norm": 1.6785908937454224, "learning_rate": 4.924108649644072e-05, "loss": 0.055, "step": 52960 }, { "grad_norm": 0.37963882088661194, "learning_rate": 4.922455374480271e-05, "loss": 0.0545, "step": 52970 }, { "grad_norm": 0.5160059928894043, "learning_rate": 4.920802107796634e-05, "loss": 0.0438, "step": 52980 }, { "grad_norm": 0.43497854471206665, "learning_rate": 4.9191488497739564e-05, "loss": 0.049, "step": 52990 }, { "grad_norm": 0.42930835485458374, "learning_rate": 4.917495600593042e-05, "loss": 0.0486, "step": 53000 }, { "grad_norm": 0.40361422300338745, "learning_rate": 4.915842360434684e-05, "loss": 0.0411, "step": 53010 }, { "grad_norm": 0.9991806745529175, "learning_rate": 4.914189129479679e-05, "loss": 0.0593, "step": 53020 }, { "grad_norm": 0.34371721744537354, "learning_rate": 4.912535907908822e-05, "loss": 0.0452, "step": 53030 }, { "grad_norm": 1.2354227304458618, "learning_rate": 4.910882695902907e-05, "loss": 0.0479, "step": 53040 }, { "grad_norm": 0.6453925371170044, "learning_rate": 4.9092294936427234e-05, "loss": 0.0394, "step": 53050 }, { "grad_norm": 0.5020402073860168, "learning_rate": 4.9075763013090685e-05, "loss": 0.041, "step": 53060 }, { "grad_norm": 0.7330597043037415, "learning_rate": 4.9059231190827305e-05, "loss": 0.0385, "step": 53070 }, { "grad_norm": 0.3952038586139679, "learning_rate": 4.904269947144499e-05, "loss": 0.0509, "step": 53080 }, { "grad_norm": 0.5695346593856812, "learning_rate": 4.902616785675161e-05, "loss": 0.048, "step": 53090 }, { "grad_norm": 0.413882851600647, "learning_rate": 4.900963634855505e-05, "loss": 0.0462, "step": 53100 }, { "grad_norm": 0.5210175514221191, "learning_rate": 4.8993104948663173e-05, "loss": 0.032, "step": 53110 }, { "grad_norm": 0.3749910891056061, "learning_rate": 4.8976573658883823e-05, "loss": 0.0415, "step": 53120 }, { "grad_norm": 0.7921019792556763, "learning_rate": 4.896004248102484e-05, "loss": 0.0442, "step": 53130 }, { "grad_norm": 0.461300253868103, "learning_rate": 4.8943511416894045e-05, "loss": 0.047, "step": 53140 }, { "grad_norm": 0.48834264278411865, "learning_rate": 4.892698046829926e-05, "loss": 0.0333, "step": 53150 }, { "grad_norm": 0.7231852412223816, "learning_rate": 4.8910449637048275e-05, "loss": 0.0537, "step": 53160 }, { "grad_norm": 0.5689529180526733, "learning_rate": 4.8893918924948884e-05, "loss": 0.0441, "step": 53170 }, { "grad_norm": 0.3945341110229492, "learning_rate": 4.8877388333808835e-05, "loss": 0.0472, "step": 53180 }, { "grad_norm": 0.35593414306640625, "learning_rate": 4.8860857865435925e-05, "loss": 0.0423, "step": 53190 }, { "grad_norm": 0.4402616620063782, "learning_rate": 4.884432752163789e-05, "loss": 0.0419, "step": 53200 }, { "grad_norm": 0.3235759735107422, "learning_rate": 4.8827797304222466e-05, "loss": 0.0418, "step": 53210 }, { "grad_norm": 0.34097975492477417, "learning_rate": 4.8811267214997366e-05, "loss": 0.04, "step": 53220 }, { "grad_norm": 0.5500618815422058, "learning_rate": 4.879473725577029e-05, "loss": 0.0511, "step": 53230 }, { "grad_norm": 0.5040116906166077, "learning_rate": 4.877820742834891e-05, "loss": 0.042, "step": 53240 }, { "grad_norm": 0.5188037157058716, "learning_rate": 4.8761677734540956e-05, "loss": 0.0487, "step": 53250 }, { "grad_norm": 0.7784205675125122, "learning_rate": 4.874514817615407e-05, "loss": 0.044, "step": 53260 }, { "grad_norm": 0.39629265666007996, "learning_rate": 4.872861875499588e-05, "loss": 0.0527, "step": 53270 }, { "grad_norm": 0.40944230556488037, "learning_rate": 4.871208947287404e-05, "loss": 0.0423, "step": 53280 }, { "grad_norm": 0.5670465230941772, "learning_rate": 4.869556033159615e-05, "loss": 0.0504, "step": 53290 }, { "grad_norm": 0.31652021408081055, "learning_rate": 4.8679031332969835e-05, "loss": 0.0367, "step": 53300 }, { "grad_norm": 0.2848188281059265, "learning_rate": 4.8662502478802635e-05, "loss": 0.0633, "step": 53310 }, { "grad_norm": 0.528400182723999, "learning_rate": 4.8645973770902176e-05, "loss": 0.0429, "step": 53320 }, { "grad_norm": 0.40403419733047485, "learning_rate": 4.862944521107599e-05, "loss": 0.0308, "step": 53330 }, { "grad_norm": 0.7029519081115723, "learning_rate": 4.861291680113162e-05, "loss": 0.0485, "step": 53340 }, { "grad_norm": 0.42524605989456177, "learning_rate": 4.8596388542876563e-05, "loss": 0.0414, "step": 53350 }, { "grad_norm": 0.7630360722541809, "learning_rate": 4.857986043811836e-05, "loss": 0.045, "step": 53360 }, { "grad_norm": 0.48212605714797974, "learning_rate": 4.856333248866448e-05, "loss": 0.0412, "step": 53370 }, { "grad_norm": 0.5842990875244141, "learning_rate": 4.85468046963224e-05, "loss": 0.04, "step": 53380 }, { "grad_norm": 0.43242186307907104, "learning_rate": 4.853027706289957e-05, "loss": 0.0347, "step": 53390 }, { "grad_norm": 0.5283592939376831, "learning_rate": 4.851374959020344e-05, "loss": 0.0399, "step": 53400 }, { "grad_norm": 0.3230290412902832, "learning_rate": 4.849722228004141e-05, "loss": 0.0395, "step": 53410 }, { "grad_norm": 0.8752551078796387, "learning_rate": 4.848069513422091e-05, "loss": 0.0469, "step": 53420 }, { "grad_norm": 0.32637593150138855, "learning_rate": 4.8464168154549294e-05, "loss": 0.0385, "step": 53430 }, { "grad_norm": 0.42155277729034424, "learning_rate": 4.844764134283392e-05, "loss": 0.045, "step": 53440 }, { "grad_norm": 0.5188475847244263, "learning_rate": 4.843111470088218e-05, "loss": 0.0514, "step": 53450 }, { "grad_norm": 0.5604643225669861, "learning_rate": 4.841458823050137e-05, "loss": 0.0413, "step": 53460 }, { "grad_norm": 0.5063894987106323, "learning_rate": 4.8398061933498816e-05, "loss": 0.0365, "step": 53470 }, { "grad_norm": 0.9842407703399658, "learning_rate": 4.83815358116818e-05, "loss": 0.0418, "step": 53480 }, { "grad_norm": 0.43407008051872253, "learning_rate": 4.8365009866857584e-05, "loss": 0.0389, "step": 53490 }, { "grad_norm": 0.33778226375579834, "learning_rate": 4.834848410083342e-05, "loss": 0.0396, "step": 53500 }, { "grad_norm": 0.5917669534683228, "learning_rate": 4.833195851541657e-05, "loss": 0.0503, "step": 53510 }, { "grad_norm": 0.3969860076904297, "learning_rate": 4.831543311241423e-05, "loss": 0.045, "step": 53520 }, { "grad_norm": 0.8186385035514832, "learning_rate": 4.829890789363359e-05, "loss": 0.0591, "step": 53530 }, { "grad_norm": 0.43067818880081177, "learning_rate": 4.828238286088182e-05, "loss": 0.0465, "step": 53540 }, { "grad_norm": 0.3225332796573639, "learning_rate": 4.826585801596609e-05, "loss": 0.0431, "step": 53550 }, { "grad_norm": 0.40759989619255066, "learning_rate": 4.8249333360693486e-05, "loss": 0.0378, "step": 53560 }, { "grad_norm": 0.36254772543907166, "learning_rate": 4.8232808896871185e-05, "loss": 0.0368, "step": 53570 }, { "grad_norm": 0.2811983525753021, "learning_rate": 4.821628462630624e-05, "loss": 0.0405, "step": 53580 }, { "grad_norm": 0.5369959473609924, "learning_rate": 4.8199760550805735e-05, "loss": 0.038, "step": 53590 }, { "grad_norm": 0.5776994824409485, "learning_rate": 4.818323667217669e-05, "loss": 0.0442, "step": 53600 }, { "grad_norm": 0.465190589427948, "learning_rate": 4.8166712992226165e-05, "loss": 0.0454, "step": 53610 }, { "grad_norm": 0.41321828961372375, "learning_rate": 4.8150189512761154e-05, "loss": 0.0495, "step": 53620 }, { "grad_norm": 0.4220404624938965, "learning_rate": 4.813366623558862e-05, "loss": 0.0385, "step": 53630 }, { "grad_norm": 0.5198636651039124, "learning_rate": 4.811714316251554e-05, "loss": 0.0393, "step": 53640 }, { "grad_norm": 0.6735748052597046, "learning_rate": 4.810062029534886e-05, "loss": 0.0384, "step": 53650 }, { "grad_norm": 0.38976579904556274, "learning_rate": 4.8084097635895505e-05, "loss": 0.0442, "step": 53660 }, { "grad_norm": 0.588792085647583, "learning_rate": 4.8067575185962335e-05, "loss": 0.0442, "step": 53670 }, { "grad_norm": 0.49065420031547546, "learning_rate": 4.805105294735625e-05, "loss": 0.0416, "step": 53680 }, { "grad_norm": 0.5800109505653381, "learning_rate": 4.803453092188406e-05, "loss": 0.0415, "step": 53690 }, { "grad_norm": 0.6448007225990295, "learning_rate": 4.801800911135263e-05, "loss": 0.0408, "step": 53700 }, { "grad_norm": 0.6658352017402649, "learning_rate": 4.8001487517568744e-05, "loss": 0.0513, "step": 53710 }, { "grad_norm": 0.6586609482765198, "learning_rate": 4.798496614233918e-05, "loss": 0.0512, "step": 53720 }, { "grad_norm": 0.4502851068973541, "learning_rate": 4.796844498747069e-05, "loss": 0.0558, "step": 53730 }, { "grad_norm": 0.5512513518333435, "learning_rate": 4.795192405477e-05, "loss": 0.0378, "step": 53740 }, { "grad_norm": 0.7356105446815491, "learning_rate": 4.79354033460438e-05, "loss": 0.0536, "step": 53750 }, { "grad_norm": 0.4750708043575287, "learning_rate": 4.79188828630988e-05, "loss": 0.0503, "step": 53760 }, { "grad_norm": 0.8011246919631958, "learning_rate": 4.790236260774165e-05, "loss": 0.0413, "step": 53770 }, { "grad_norm": 0.4047788679599762, "learning_rate": 4.788584258177896e-05, "loss": 0.0455, "step": 53780 }, { "grad_norm": 0.405392050743103, "learning_rate": 4.7869322787017356e-05, "loss": 0.0391, "step": 53790 }, { "grad_norm": 0.4647906422615051, "learning_rate": 4.78528032252634e-05, "loss": 0.0462, "step": 53800 }, { "grad_norm": 0.38199418783187866, "learning_rate": 4.783628389832366e-05, "loss": 0.0468, "step": 53810 }, { "grad_norm": 4.907114505767822, "learning_rate": 4.7819764808004644e-05, "loss": 0.0511, "step": 53820 }, { "grad_norm": 0.43308401107788086, "learning_rate": 4.780324595611289e-05, "loss": 0.0453, "step": 53830 }, { "grad_norm": 0.5045636892318726, "learning_rate": 4.7786727344454854e-05, "loss": 0.0545, "step": 53840 }, { "grad_norm": 0.5261188745498657, "learning_rate": 4.777020897483697e-05, "loss": 0.0475, "step": 53850 }, { "grad_norm": 0.39582541584968567, "learning_rate": 4.775369084906569e-05, "loss": 0.0397, "step": 53860 }, { "grad_norm": 0.44823724031448364, "learning_rate": 4.773717296894742e-05, "loss": 0.0548, "step": 53870 }, { "grad_norm": 0.4993719756603241, "learning_rate": 4.772065533628848e-05, "loss": 0.0481, "step": 53880 }, { "grad_norm": 0.39619532227516174, "learning_rate": 4.7704137952895265e-05, "loss": 0.0428, "step": 53890 }, { "grad_norm": 0.4034239649772644, "learning_rate": 4.768762082057407e-05, "loss": 0.0323, "step": 53900 }, { "grad_norm": 0.41682615876197815, "learning_rate": 4.7671103941131186e-05, "loss": 0.046, "step": 53910 }, { "grad_norm": 0.4386787712574005, "learning_rate": 4.765458731637287e-05, "loss": 0.0412, "step": 53920 }, { "grad_norm": 0.32747143507003784, "learning_rate": 4.763807094810537e-05, "loss": 0.0482, "step": 53930 }, { "grad_norm": 0.4839111566543579, "learning_rate": 4.762155483813487e-05, "loss": 0.0554, "step": 53940 }, { "grad_norm": 0.5438752770423889, "learning_rate": 4.760503898826754e-05, "loss": 0.0537, "step": 53950 }, { "grad_norm": 0.5125327706336975, "learning_rate": 4.758852340030957e-05, "loss": 0.0548, "step": 53960 }, { "grad_norm": 0.504775881767273, "learning_rate": 4.757200807606705e-05, "loss": 0.0466, "step": 53970 }, { "grad_norm": 0.9567219018936157, "learning_rate": 4.755549301734607e-05, "loss": 0.0483, "step": 53980 }, { "grad_norm": 0.4457468092441559, "learning_rate": 4.753897822595271e-05, "loss": 0.0616, "step": 53990 }, { "grad_norm": 0.6694880127906799, "learning_rate": 4.7522463703692994e-05, "loss": 0.0402, "step": 54000 }, { "grad_norm": 0.3074791133403778, "learning_rate": 4.75059494523729e-05, "loss": 0.0379, "step": 54010 }, { "grad_norm": 0.4016331732273102, "learning_rate": 4.748943547379843e-05, "loss": 0.0418, "step": 54020 }, { "grad_norm": 0.5601034760475159, "learning_rate": 4.7472921769775535e-05, "loss": 0.0516, "step": 54030 }, { "grad_norm": 0.4871809482574463, "learning_rate": 4.745640834211011e-05, "loss": 0.0557, "step": 54040 }, { "grad_norm": 0.45278477668762207, "learning_rate": 4.7439895192608054e-05, "loss": 0.0433, "step": 54050 }, { "grad_norm": 0.572461724281311, "learning_rate": 4.7423382323075186e-05, "loss": 0.0428, "step": 54060 }, { "grad_norm": 0.4968744218349457, "learning_rate": 4.740686973531737e-05, "loss": 0.0388, "step": 54070 }, { "grad_norm": 0.6295723915100098, "learning_rate": 4.7390357431140375e-05, "loss": 0.0457, "step": 54080 }, { "grad_norm": 0.5254248976707458, "learning_rate": 4.737384541234998e-05, "loss": 0.0511, "step": 54090 }, { "grad_norm": 0.40975403785705566, "learning_rate": 4.735733368075189e-05, "loss": 0.0325, "step": 54100 }, { "grad_norm": 0.4164230525493622, "learning_rate": 4.734082223815182e-05, "loss": 0.0447, "step": 54110 }, { "grad_norm": 0.4609706699848175, "learning_rate": 4.7324311086355436e-05, "loss": 0.0413, "step": 54120 }, { "grad_norm": 0.4752761125564575, "learning_rate": 4.730780022716836e-05, "loss": 0.0442, "step": 54130 }, { "grad_norm": 0.42476382851600647, "learning_rate": 4.729128966239619e-05, "loss": 0.0379, "step": 54140 }, { "grad_norm": 0.5876315236091614, "learning_rate": 4.7274779393844536e-05, "loss": 0.0357, "step": 54150 }, { "grad_norm": 0.7664980888366699, "learning_rate": 4.725826942331891e-05, "loss": 0.058, "step": 54160 }, { "grad_norm": 0.6737721562385559, "learning_rate": 4.724175975262481e-05, "loss": 0.049, "step": 54170 }, { "grad_norm": 0.5371242165565491, "learning_rate": 4.722525038356772e-05, "loss": 0.0402, "step": 54180 }, { "grad_norm": 0.48399046063423157, "learning_rate": 4.7208741317953075e-05, "loss": 0.0429, "step": 54190 }, { "grad_norm": 0.3837973475456238, "learning_rate": 4.7192232557586264e-05, "loss": 0.0471, "step": 54200 }, { "grad_norm": 0.5747647881507874, "learning_rate": 4.717572410427269e-05, "loss": 0.0458, "step": 54210 }, { "grad_norm": 0.6780873537063599, "learning_rate": 4.71592159598177e-05, "loss": 0.0476, "step": 54220 }, { "grad_norm": 0.49050265550613403, "learning_rate": 4.714270812602657e-05, "loss": 0.039, "step": 54230 }, { "grad_norm": 0.3141813576221466, "learning_rate": 4.712620060470458e-05, "loss": 0.053, "step": 54240 }, { "grad_norm": 1.334918737411499, "learning_rate": 4.7109693397656986e-05, "loss": 0.0508, "step": 54250 }, { "grad_norm": 0.9238873720169067, "learning_rate": 4.709318650668894e-05, "loss": 0.0401, "step": 54260 }, { "grad_norm": 0.5974313616752625, "learning_rate": 4.7076679933605676e-05, "loss": 0.0456, "step": 54270 }, { "grad_norm": 0.4242924451828003, "learning_rate": 4.70601736802123e-05, "loss": 0.037, "step": 54280 }, { "grad_norm": 0.6616755723953247, "learning_rate": 4.704366774831391e-05, "loss": 0.059, "step": 54290 }, { "grad_norm": 0.4341789782047272, "learning_rate": 4.702716213971557e-05, "loss": 0.0404, "step": 54300 }, { "grad_norm": 0.5603156089782715, "learning_rate": 4.70106568562223e-05, "loss": 0.047, "step": 54310 }, { "grad_norm": 0.5484780669212341, "learning_rate": 4.6994151899639105e-05, "loss": 0.0404, "step": 54320 }, { "grad_norm": 0.4252859354019165, "learning_rate": 4.697764727177093e-05, "loss": 0.0622, "step": 54330 }, { "grad_norm": 0.46238505840301514, "learning_rate": 4.696114297442272e-05, "loss": 0.0418, "step": 54340 }, { "grad_norm": 0.47670233249664307, "learning_rate": 4.694463900939933e-05, "loss": 0.0396, "step": 54350 }, { "grad_norm": 0.3968823552131653, "learning_rate": 4.692813537850564e-05, "loss": 0.0382, "step": 54360 }, { "grad_norm": 0.6413598656654358, "learning_rate": 4.6911632083546445e-05, "loss": 0.0335, "step": 54370 }, { "grad_norm": 0.34730759263038635, "learning_rate": 4.6895129126326525e-05, "loss": 0.0436, "step": 54380 }, { "grad_norm": 0.44867485761642456, "learning_rate": 4.687862650865059e-05, "loss": 0.0384, "step": 54390 }, { "grad_norm": 0.3932102620601654, "learning_rate": 4.686212423232339e-05, "loss": 0.035, "step": 54400 }, { "grad_norm": 0.3640306890010834, "learning_rate": 4.6845622299149565e-05, "loss": 0.0492, "step": 54410 }, { "grad_norm": 0.3774716258049011, "learning_rate": 4.682912071093374e-05, "loss": 0.0359, "step": 54420 }, { "grad_norm": 0.9852705001831055, "learning_rate": 4.681261946948052e-05, "loss": 0.0529, "step": 54430 }, { "grad_norm": 1.242893099784851, "learning_rate": 4.679611857659443e-05, "loss": 0.044, "step": 54440 }, { "grad_norm": 0.24852856993675232, "learning_rate": 4.6779618034079975e-05, "loss": 0.0406, "step": 54450 }, { "grad_norm": 0.35020074248313904, "learning_rate": 4.6763117843741675e-05, "loss": 0.0542, "step": 54460 }, { "grad_norm": 0.6434475183486938, "learning_rate": 4.674661800738394e-05, "loss": 0.0425, "step": 54470 }, { "grad_norm": 0.4214041829109192, "learning_rate": 4.673011852681117e-05, "loss": 0.0359, "step": 54480 }, { "grad_norm": 0.3288770318031311, "learning_rate": 4.671361940382772e-05, "loss": 0.049, "step": 54490 }, { "grad_norm": 0.49775826930999756, "learning_rate": 4.669712064023791e-05, "loss": 0.0452, "step": 54500 }, { "grad_norm": 0.47859472036361694, "learning_rate": 4.668062223784602e-05, "loss": 0.0388, "step": 54510 }, { "grad_norm": 0.4839022755622864, "learning_rate": 4.6664124198456274e-05, "loss": 0.0448, "step": 54520 }, { "grad_norm": 0.44872981309890747, "learning_rate": 4.6647626523872905e-05, "loss": 0.0482, "step": 54530 }, { "grad_norm": 0.3971443176269531, "learning_rate": 4.6631129215900066e-05, "loss": 0.0422, "step": 54540 }, { "grad_norm": 0.5654045939445496, "learning_rate": 4.661463227634187e-05, "loss": 0.0658, "step": 54550 }, { "grad_norm": 0.5244407057762146, "learning_rate": 4.659813570700238e-05, "loss": 0.036, "step": 54560 }, { "grad_norm": 0.5543638467788696, "learning_rate": 4.658163950968568e-05, "loss": 0.0374, "step": 54570 }, { "grad_norm": 0.3868657946586609, "learning_rate": 4.6565143686195726e-05, "loss": 0.0438, "step": 54580 }, { "grad_norm": 0.8485687375068665, "learning_rate": 4.654864823833649e-05, "loss": 0.0372, "step": 54590 }, { "grad_norm": 0.507976770401001, "learning_rate": 4.65321531679119e-05, "loss": 0.0387, "step": 54600 }, { "grad_norm": 0.4335758090019226, "learning_rate": 4.6515658476725834e-05, "loss": 0.0472, "step": 54610 }, { "grad_norm": 0.5552670359611511, "learning_rate": 4.649916416658212e-05, "loss": 0.0467, "step": 54620 }, { "grad_norm": 0.6399109959602356, "learning_rate": 4.648267023928455e-05, "loss": 0.0446, "step": 54630 }, { "grad_norm": 0.6111172437667847, "learning_rate": 4.646617669663687e-05, "loss": 0.0385, "step": 54640 }, { "grad_norm": 0.560375452041626, "learning_rate": 4.644968354044277e-05, "loss": 0.0379, "step": 54650 }, { "grad_norm": 0.7619701027870178, "learning_rate": 4.643319077250596e-05, "loss": 0.0476, "step": 54660 }, { "grad_norm": 0.6348981261253357, "learning_rate": 4.6416698394630055e-05, "loss": 0.0321, "step": 54670 }, { "grad_norm": 0.6253263354301453, "learning_rate": 4.640020640861862e-05, "loss": 0.0371, "step": 54680 }, { "grad_norm": 0.3687412440776825, "learning_rate": 4.6383714816275195e-05, "loss": 0.0409, "step": 54690 }, { "grad_norm": 0.3191627860069275, "learning_rate": 4.636722361940329e-05, "loss": 0.035, "step": 54700 }, { "grad_norm": 0.4561704993247986, "learning_rate": 4.635073281980632e-05, "loss": 0.0484, "step": 54710 }, { "grad_norm": 0.33659619092941284, "learning_rate": 4.633424241928775e-05, "loss": 0.0513, "step": 54720 }, { "grad_norm": 0.4235410988330841, "learning_rate": 4.6317752419650914e-05, "loss": 0.0381, "step": 54730 }, { "grad_norm": 0.6600834727287292, "learning_rate": 4.630126282269914e-05, "loss": 0.0518, "step": 54740 }, { "grad_norm": 0.47733035683631897, "learning_rate": 4.628477363023569e-05, "loss": 0.0392, "step": 54750 }, { "grad_norm": 0.49999383091926575, "learning_rate": 4.626828484406383e-05, "loss": 0.0437, "step": 54760 }, { "grad_norm": 0.553095817565918, "learning_rate": 4.625179646598669e-05, "loss": 0.0492, "step": 54770 }, { "grad_norm": 0.3634408414363861, "learning_rate": 4.6235308497807487e-05, "loss": 0.0422, "step": 54780 }, { "grad_norm": 0.5113325119018555, "learning_rate": 4.6218820941329276e-05, "loss": 0.0502, "step": 54790 }, { "grad_norm": 0.31346365809440613, "learning_rate": 4.620233379835513e-05, "loss": 0.0348, "step": 54800 }, { "grad_norm": 0.6167066097259521, "learning_rate": 4.618584707068804e-05, "loss": 0.0407, "step": 54810 }, { "grad_norm": 0.45277702808380127, "learning_rate": 4.616936076013099e-05, "loss": 0.0457, "step": 54820 }, { "grad_norm": 0.6618108153343201, "learning_rate": 4.615287486848688e-05, "loss": 0.0428, "step": 54830 }, { "grad_norm": 0.5397660136222839, "learning_rate": 4.6136389397558584e-05, "loss": 0.0485, "step": 54840 }, { "grad_norm": 0.568410336971283, "learning_rate": 4.6119904349148935e-05, "loss": 0.0338, "step": 54850 }, { "grad_norm": 0.5626286864280701, "learning_rate": 4.6103419725060726e-05, "loss": 0.0348, "step": 54860 }, { "grad_norm": 0.3888207674026489, "learning_rate": 4.608693552709667e-05, "loss": 0.0281, "step": 54870 }, { "grad_norm": 0.5209181904792786, "learning_rate": 4.607045175705947e-05, "loss": 0.0444, "step": 54880 }, { "grad_norm": 0.44693949818611145, "learning_rate": 4.605396841675176e-05, "loss": 0.042, "step": 54890 }, { "grad_norm": 0.4352359175682068, "learning_rate": 4.6037485507976094e-05, "loss": 0.0447, "step": 54900 }, { "grad_norm": 0.27770838141441345, "learning_rate": 4.602100303253509e-05, "loss": 0.0401, "step": 54910 }, { "grad_norm": 0.41123437881469727, "learning_rate": 4.600452099223122e-05, "loss": 0.0542, "step": 54920 }, { "grad_norm": 0.3547321557998657, "learning_rate": 4.598803938886691e-05, "loss": 0.037, "step": 54930 }, { "grad_norm": 0.39593201875686646, "learning_rate": 4.597155822424459e-05, "loss": 0.0358, "step": 54940 }, { "grad_norm": 0.5753024220466614, "learning_rate": 4.59550775001666e-05, "loss": 0.0497, "step": 54950 }, { "grad_norm": 0.5323091745376587, "learning_rate": 4.5938597218435234e-05, "loss": 0.0445, "step": 54960 }, { "grad_norm": 0.5933874249458313, "learning_rate": 4.592211738085279e-05, "loss": 0.0614, "step": 54970 }, { "grad_norm": 0.4961380660533905, "learning_rate": 4.590563798922146e-05, "loss": 0.0318, "step": 54980 }, { "grad_norm": 0.39696574211120605, "learning_rate": 4.5889159045343404e-05, "loss": 0.0557, "step": 54990 }, { "grad_norm": 0.45139697194099426, "learning_rate": 4.5872680551020734e-05, "loss": 0.0436, "step": 55000 }, { "grad_norm": 0.46819162368774414, "learning_rate": 4.585620250805552e-05, "loss": 0.0323, "step": 55010 }, { "grad_norm": 0.5404031872749329, "learning_rate": 4.583972491824975e-05, "loss": 0.0378, "step": 55020 }, { "grad_norm": 0.4170624911785126, "learning_rate": 4.582324778340541e-05, "loss": 0.046, "step": 55030 }, { "grad_norm": 0.5470190644264221, "learning_rate": 4.5806771105324424e-05, "loss": 0.0474, "step": 55040 }, { "grad_norm": 0.3870680034160614, "learning_rate": 4.579029488580865e-05, "loss": 0.0476, "step": 55050 }, { "grad_norm": 0.34262919425964355, "learning_rate": 4.5773819126659876e-05, "loss": 0.0289, "step": 55060 }, { "grad_norm": 0.5137686133384705, "learning_rate": 4.575734382967991e-05, "loss": 0.042, "step": 55070 }, { "grad_norm": 0.40711119771003723, "learning_rate": 4.574086899667044e-05, "loss": 0.0413, "step": 55080 }, { "grad_norm": 0.5604112148284912, "learning_rate": 4.5724394629433106e-05, "loss": 0.0458, "step": 55090 }, { "grad_norm": 0.30467402935028076, "learning_rate": 4.570792072976958e-05, "loss": 0.0419, "step": 55100 }, { "grad_norm": 0.4078710973262787, "learning_rate": 4.569144729948138e-05, "loss": 0.0439, "step": 55110 }, { "grad_norm": 0.49530208110809326, "learning_rate": 4.5674974340370033e-05, "loss": 0.032, "step": 55120 }, { "grad_norm": 0.3733024001121521, "learning_rate": 4.565850185423699e-05, "loss": 0.042, "step": 55130 }, { "grad_norm": 0.4569512605667114, "learning_rate": 4.564202984288365e-05, "loss": 0.0422, "step": 55140 }, { "grad_norm": 0.6488254070281982, "learning_rate": 4.5625558308111356e-05, "loss": 0.0421, "step": 55150 }, { "grad_norm": 0.37405192852020264, "learning_rate": 4.560908725172145e-05, "loss": 0.0449, "step": 55160 }, { "grad_norm": 0.4380039870738983, "learning_rate": 4.559261667551516e-05, "loss": 0.0438, "step": 55170 }, { "grad_norm": 0.3735252916812897, "learning_rate": 4.5576146581293685e-05, "loss": 0.0509, "step": 55180 }, { "grad_norm": 0.8963037729263306, "learning_rate": 4.5559676970858164e-05, "loss": 0.0453, "step": 55190 }, { "grad_norm": 1.4094126224517822, "learning_rate": 4.5543207846009705e-05, "loss": 0.0462, "step": 55200 }, { "grad_norm": 0.43723493814468384, "learning_rate": 4.552673920854933e-05, "loss": 0.0529, "step": 55210 }, { "grad_norm": 0.9149580001831055, "learning_rate": 4.5510271060278006e-05, "loss": 0.0438, "step": 55220 }, { "grad_norm": 0.45959389209747314, "learning_rate": 4.5493803402996704e-05, "loss": 0.0482, "step": 55230 }, { "grad_norm": 0.2642734944820404, "learning_rate": 4.5477336238506295e-05, "loss": 0.0425, "step": 55240 }, { "grad_norm": 0.4270033538341522, "learning_rate": 4.546086956860759e-05, "loss": 0.0432, "step": 55250 }, { "grad_norm": 0.5189295411109924, "learning_rate": 4.5444403395101374e-05, "loss": 0.0362, "step": 55260 }, { "grad_norm": 0.6192681789398193, "learning_rate": 4.542793771978834e-05, "loss": 0.0385, "step": 55270 }, { "grad_norm": 0.8190445899963379, "learning_rate": 4.541147254446916e-05, "loss": 0.0453, "step": 55280 }, { "grad_norm": 0.40494289994239807, "learning_rate": 4.539500787094445e-05, "loss": 0.0451, "step": 55290 }, { "grad_norm": 0.4530128240585327, "learning_rate": 4.537854370101475e-05, "loss": 0.0461, "step": 55300 }, { "grad_norm": 0.3188979923725128, "learning_rate": 4.5362080036480566e-05, "loss": 0.0394, "step": 55310 }, { "grad_norm": 0.4295729994773865, "learning_rate": 4.534561687914235e-05, "loss": 0.037, "step": 55320 }, { "grad_norm": 0.40445175766944885, "learning_rate": 4.5329154230800466e-05, "loss": 0.0436, "step": 55330 }, { "grad_norm": 0.2063402533531189, "learning_rate": 4.531269209325525e-05, "loss": 0.0409, "step": 55340 }, { "grad_norm": 0.972963809967041, "learning_rate": 4.529623046830696e-05, "loss": 0.0327, "step": 55350 }, { "grad_norm": 0.6511141657829285, "learning_rate": 4.5279769357755866e-05, "loss": 0.0499, "step": 55360 }, { "grad_norm": 0.6759310364723206, "learning_rate": 4.5263308763402084e-05, "loss": 0.0355, "step": 55370 }, { "grad_norm": 0.2587810754776001, "learning_rate": 4.524684868704574e-05, "loss": 0.0448, "step": 55380 }, { "grad_norm": 0.5038281679153442, "learning_rate": 4.523038913048688e-05, "loss": 0.0408, "step": 55390 }, { "grad_norm": 0.4788075089454651, "learning_rate": 4.521393009552549e-05, "loss": 0.0462, "step": 55400 }, { "grad_norm": 0.35246598720550537, "learning_rate": 4.5197471583961484e-05, "loss": 0.0465, "step": 55410 }, { "grad_norm": 0.28364822268486023, "learning_rate": 4.51810135975948e-05, "loss": 0.0434, "step": 55420 }, { "grad_norm": 0.5158937573432922, "learning_rate": 4.516455613822521e-05, "loss": 0.0405, "step": 55430 }, { "grad_norm": 0.4013519585132599, "learning_rate": 4.514809920765249e-05, "loss": 0.0458, "step": 55440 }, { "grad_norm": 0.4814970791339874, "learning_rate": 4.5131642807676356e-05, "loss": 0.0389, "step": 55450 }, { "grad_norm": 2.0625009536743164, "learning_rate": 4.511518694009644e-05, "loss": 0.049, "step": 55460 }, { "grad_norm": 0.35493069887161255, "learning_rate": 4.509873160671231e-05, "loss": 0.0459, "step": 55470 }, { "grad_norm": 0.49248892068862915, "learning_rate": 4.508227680932356e-05, "loss": 0.0387, "step": 55480 }, { "grad_norm": 0.9648799300193787, "learning_rate": 4.5065822549729605e-05, "loss": 0.0394, "step": 55490 }, { "grad_norm": 0.41701045632362366, "learning_rate": 4.504936882972989e-05, "loss": 0.0378, "step": 55500 }, { "grad_norm": 0.453467458486557, "learning_rate": 4.503291565112374e-05, "loss": 0.0305, "step": 55510 }, { "grad_norm": 0.5703264474868774, "learning_rate": 4.501646301571048e-05, "loss": 0.0479, "step": 55520 }, { "grad_norm": 0.5661087036132812, "learning_rate": 4.500001092528933e-05, "loss": 0.0467, "step": 55530 }, { "grad_norm": 0.3429136872291565, "learning_rate": 4.4983559381659455e-05, "loss": 0.0373, "step": 55540 }, { "grad_norm": 0.4185387194156647, "learning_rate": 4.496710838661999e-05, "loss": 0.0407, "step": 55550 }, { "grad_norm": 0.35068845748901367, "learning_rate": 4.495065794196999e-05, "loss": 0.0444, "step": 55560 }, { "grad_norm": 0.5015273094177246, "learning_rate": 4.4934208049508455e-05, "loss": 0.0348, "step": 55570 }, { "grad_norm": 0.46328628063201904, "learning_rate": 4.4917758711034303e-05, "loss": 0.035, "step": 55580 }, { "grad_norm": 0.46694454550743103, "learning_rate": 4.490130992834642e-05, "loss": 0.0382, "step": 55590 }, { "grad_norm": 0.748101532459259, "learning_rate": 4.488486170324359e-05, "loss": 0.0654, "step": 55600 }, { "grad_norm": 0.5473750233650208, "learning_rate": 4.486841403752463e-05, "loss": 0.0357, "step": 55610 }, { "grad_norm": 0.36161530017852783, "learning_rate": 4.4851966932988174e-05, "loss": 0.0308, "step": 55620 }, { "grad_norm": 0.5023218393325806, "learning_rate": 4.4835520391432894e-05, "loss": 0.0382, "step": 55630 }, { "grad_norm": 0.3794914782047272, "learning_rate": 4.481907441465733e-05, "loss": 0.0475, "step": 55640 }, { "grad_norm": 0.44564884901046753, "learning_rate": 4.480262900446e-05, "loss": 0.0417, "step": 55650 }, { "grad_norm": 0.2743391692638397, "learning_rate": 4.478618416263933e-05, "loss": 0.0455, "step": 55660 }, { "grad_norm": 0.4102920591831207, "learning_rate": 4.476973989099373e-05, "loss": 0.0392, "step": 55670 }, { "grad_norm": 0.614769697189331, "learning_rate": 4.475329619132153e-05, "loss": 0.0529, "step": 55680 }, { "grad_norm": 0.4480802118778229, "learning_rate": 4.4736853065420956e-05, "loss": 0.0582, "step": 55690 }, { "grad_norm": 0.5067984461784363, "learning_rate": 4.472041051509023e-05, "loss": 0.0406, "step": 55700 }, { "grad_norm": 0.3685310482978821, "learning_rate": 4.470396854212746e-05, "loss": 0.0424, "step": 55710 }, { "grad_norm": 0.7008262276649475, "learning_rate": 4.468752714833072e-05, "loss": 0.0449, "step": 55720 }, { "grad_norm": 0.30157241225242615, "learning_rate": 4.467108633549803e-05, "loss": 0.0387, "step": 55730 }, { "grad_norm": 0.43719038367271423, "learning_rate": 4.4654646105427335e-05, "loss": 0.06, "step": 55740 }, { "grad_norm": 0.4729308784008026, "learning_rate": 4.463820645991651e-05, "loss": 0.038, "step": 55750 }, { "grad_norm": 0.3757542669773102, "learning_rate": 4.462176740076334e-05, "loss": 0.0344, "step": 55760 }, { "grad_norm": 0.3749951422214508, "learning_rate": 4.4605328929765624e-05, "loss": 0.054, "step": 55770 }, { "grad_norm": 0.3911379277706146, "learning_rate": 4.458889104872102e-05, "loss": 0.0377, "step": 55780 }, { "grad_norm": 0.6302512884140015, "learning_rate": 4.457245375942715e-05, "loss": 0.0453, "step": 55790 }, { "grad_norm": 0.3918750286102295, "learning_rate": 4.455601706368157e-05, "loss": 0.0458, "step": 55800 }, { "grad_norm": 0.44637635350227356, "learning_rate": 4.453958096328179e-05, "loss": 0.0375, "step": 55810 }, { "grad_norm": 0.6242620944976807, "learning_rate": 4.452314546002523e-05, "loss": 0.0453, "step": 55820 }, { "grad_norm": 0.326248437166214, "learning_rate": 4.4506710555709246e-05, "loss": 0.05, "step": 55830 }, { "grad_norm": 0.438048779964447, "learning_rate": 4.449027625213114e-05, "loss": 0.0332, "step": 55840 }, { "grad_norm": 0.34625160694122314, "learning_rate": 4.4473842551088135e-05, "loss": 0.04, "step": 55850 }, { "grad_norm": 1.482204794883728, "learning_rate": 4.445740945437738e-05, "loss": 0.0492, "step": 55860 }, { "grad_norm": 0.32633715867996216, "learning_rate": 4.444097696379602e-05, "loss": 0.0402, "step": 55870 }, { "grad_norm": 0.34383341670036316, "learning_rate": 4.442454508114106e-05, "loss": 0.0411, "step": 55880 }, { "grad_norm": 0.48322024941444397, "learning_rate": 4.4408113808209465e-05, "loss": 0.0475, "step": 55890 }, { "grad_norm": 0.4611573815345764, "learning_rate": 4.439168314679813e-05, "loss": 0.0433, "step": 55900 }, { "grad_norm": 1.9004815816879272, "learning_rate": 4.43752530987039e-05, "loss": 0.0437, "step": 55910 }, { "grad_norm": 0.5973731875419617, "learning_rate": 4.4358823665723515e-05, "loss": 0.0362, "step": 55920 }, { "grad_norm": 0.48252391815185547, "learning_rate": 4.434239484965371e-05, "loss": 0.0516, "step": 55930 }, { "grad_norm": 0.41965922713279724, "learning_rate": 4.4325966652291103e-05, "loss": 0.0456, "step": 55940 }, { "grad_norm": 0.47774145007133484, "learning_rate": 4.430953907543225e-05, "loss": 0.0359, "step": 55950 }, { "grad_norm": 0.47459980845451355, "learning_rate": 4.4293112120873645e-05, "loss": 0.0325, "step": 55960 }, { "grad_norm": 0.5020702481269836, "learning_rate": 4.427668579041172e-05, "loss": 0.033, "step": 55970 }, { "grad_norm": 0.4660432040691376, "learning_rate": 4.4260260085842816e-05, "loss": 0.043, "step": 55980 }, { "grad_norm": 0.5019123554229736, "learning_rate": 4.4243835008963256e-05, "loss": 0.0383, "step": 55990 }, { "grad_norm": 0.35232532024383545, "learning_rate": 4.422741056156925e-05, "loss": 0.0417, "step": 56000 }, { "grad_norm": 0.3798207640647888, "learning_rate": 4.421098674545693e-05, "loss": 0.0309, "step": 56010 }, { "grad_norm": 2.469130039215088, "learning_rate": 4.419456356242241e-05, "loss": 0.0434, "step": 56020 }, { "grad_norm": 0.4217148721218109, "learning_rate": 4.4178141014261685e-05, "loss": 0.0392, "step": 56030 }, { "grad_norm": 0.5993359684944153, "learning_rate": 4.416171910277071e-05, "loss": 0.041, "step": 56040 }, { "grad_norm": 0.5266096591949463, "learning_rate": 4.414529782974533e-05, "loss": 0.0388, "step": 56050 }, { "grad_norm": 0.2620704472064972, "learning_rate": 4.41288771969814e-05, "loss": 0.0384, "step": 56060 }, { "grad_norm": 0.49464571475982666, "learning_rate": 4.411245720627464e-05, "loss": 0.0413, "step": 56070 }, { "grad_norm": 0.6779748201370239, "learning_rate": 4.40960378594207e-05, "loss": 0.0414, "step": 56080 }, { "grad_norm": 0.6818798780441284, "learning_rate": 4.407961915821519e-05, "loss": 0.0489, "step": 56090 }, { "grad_norm": 0.49665752053260803, "learning_rate": 4.406320110445362e-05, "loss": 0.039, "step": 56100 }, { "grad_norm": 0.506746232509613, "learning_rate": 4.404678369993144e-05, "loss": 0.04, "step": 56110 }, { "grad_norm": 0.4243427515029907, "learning_rate": 4.403036694644406e-05, "loss": 0.0477, "step": 56120 }, { "grad_norm": 0.37792855501174927, "learning_rate": 4.4013950845786764e-05, "loss": 0.0403, "step": 56130 }, { "grad_norm": 0.4744201600551605, "learning_rate": 4.399753539975482e-05, "loss": 0.0523, "step": 56140 }, { "grad_norm": 0.4102094769477844, "learning_rate": 4.398112061014337e-05, "loss": 0.0411, "step": 56150 }, { "grad_norm": 0.2815297842025757, "learning_rate": 4.396470647874753e-05, "loss": 0.039, "step": 56160 }, { "grad_norm": 0.7076255083084106, "learning_rate": 4.394829300736229e-05, "loss": 0.034, "step": 56170 }, { "grad_norm": 0.45135217905044556, "learning_rate": 4.393188019778265e-05, "loss": 0.0441, "step": 56180 }, { "grad_norm": 0.2596147358417511, "learning_rate": 4.391546805180347e-05, "loss": 0.0435, "step": 56190 }, { "grad_norm": 0.8836647272109985, "learning_rate": 4.389905657121955e-05, "loss": 0.0394, "step": 56200 }, { "grad_norm": 0.6050487756729126, "learning_rate": 4.3882645757825635e-05, "loss": 0.039, "step": 56210 }, { "grad_norm": 0.3231319189071655, "learning_rate": 4.386623561341637e-05, "loss": 0.0372, "step": 56220 }, { "grad_norm": 0.8433903455734253, "learning_rate": 4.384982613978637e-05, "loss": 0.0396, "step": 56230 }, { "grad_norm": 0.4580671489238739, "learning_rate": 4.383341733873012e-05, "loss": 0.0372, "step": 56240 }, { "grad_norm": 0.47208186984062195, "learning_rate": 4.381700921204209e-05, "loss": 0.0385, "step": 56250 }, { "grad_norm": 0.7441823482513428, "learning_rate": 4.380060176151663e-05, "loss": 0.0348, "step": 56260 }, { "grad_norm": 0.3472980260848999, "learning_rate": 4.378419498894805e-05, "loss": 0.051, "step": 56270 }, { "grad_norm": 0.43054211139678955, "learning_rate": 4.376778889613056e-05, "loss": 0.0404, "step": 56280 }, { "grad_norm": 0.41301000118255615, "learning_rate": 4.375138348485831e-05, "loss": 0.0493, "step": 56290 }, { "grad_norm": 0.38998064398765564, "learning_rate": 4.3734978756925346e-05, "loss": 0.0377, "step": 56300 }, { "grad_norm": 0.4793663024902344, "learning_rate": 4.3718574714125706e-05, "loss": 0.0506, "step": 56310 }, { "grad_norm": 0.6934526562690735, "learning_rate": 4.370217135825329e-05, "loss": 0.0434, "step": 56320 }, { "grad_norm": 0.46316221356391907, "learning_rate": 4.368576869110194e-05, "loss": 0.0488, "step": 56330 }, { "grad_norm": 0.48680830001831055, "learning_rate": 4.366936671446544e-05, "loss": 0.0485, "step": 56340 }, { "grad_norm": 0.4280892312526703, "learning_rate": 4.3652965430137474e-05, "loss": 0.0417, "step": 56350 }, { "grad_norm": 0.34738534688949585, "learning_rate": 4.3636564839911646e-05, "loss": 0.0338, "step": 56360 }, { "grad_norm": 0.35767310857772827, "learning_rate": 4.3620164945581545e-05, "loss": 0.0385, "step": 56370 }, { "grad_norm": 0.5000163912773132, "learning_rate": 4.360376574894061e-05, "loss": 0.0397, "step": 56380 }, { "grad_norm": 0.2600632309913635, "learning_rate": 4.358736725178224e-05, "loss": 0.0396, "step": 56390 }, { "grad_norm": 0.5486404895782471, "learning_rate": 4.357096945589974e-05, "loss": 0.0448, "step": 56400 }, { "grad_norm": 0.3625187873840332, "learning_rate": 4.355457236308636e-05, "loss": 0.0327, "step": 56410 }, { "grad_norm": 0.4629276692867279, "learning_rate": 4.353817597513526e-05, "loss": 0.0376, "step": 56420 }, { "grad_norm": 0.6900123953819275, "learning_rate": 4.352178029383948e-05, "loss": 0.0348, "step": 56430 }, { "grad_norm": 0.44768163561820984, "learning_rate": 4.3505385320992105e-05, "loss": 0.0468, "step": 56440 }, { "grad_norm": 0.39896637201309204, "learning_rate": 4.348899105838602e-05, "loss": 0.0344, "step": 56450 }, { "grad_norm": 0.4429287314414978, "learning_rate": 4.3472597507814087e-05, "loss": 0.0392, "step": 56460 }, { "grad_norm": 0.6058270931243896, "learning_rate": 4.3456204671069066e-05, "loss": 0.0429, "step": 56470 }, { "grad_norm": 0.45724615454673767, "learning_rate": 4.343981254994367e-05, "loss": 0.0342, "step": 56480 }, { "grad_norm": 0.4109070599079132, "learning_rate": 4.34234211462305e-05, "loss": 0.04, "step": 56490 }, { "grad_norm": 0.46131807565689087, "learning_rate": 4.340703046172213e-05, "loss": 0.0399, "step": 56500 }, { "grad_norm": 0.23823614418506622, "learning_rate": 4.339064049821097e-05, "loss": 0.0521, "step": 56510 }, { "grad_norm": 0.3632013201713562, "learning_rate": 4.3374251257489446e-05, "loss": 0.0463, "step": 56520 }, { "grad_norm": 0.9875940084457397, "learning_rate": 4.335786274134984e-05, "loss": 0.0489, "step": 56530 }, { "grad_norm": 0.5262277126312256, "learning_rate": 4.334147495158438e-05, "loss": 0.0402, "step": 56540 }, { "grad_norm": 0.4633327126502991, "learning_rate": 4.33250878899852e-05, "loss": 0.0466, "step": 56550 }, { "grad_norm": 0.3772134482860565, "learning_rate": 4.330870155834437e-05, "loss": 0.04, "step": 56560 }, { "grad_norm": 0.5508706569671631, "learning_rate": 4.3292315958453884e-05, "loss": 0.0441, "step": 56570 }, { "grad_norm": 0.47372719645500183, "learning_rate": 4.3275931092105646e-05, "loss": 0.0373, "step": 56580 }, { "grad_norm": 0.3277741074562073, "learning_rate": 4.325954696109147e-05, "loss": 0.0456, "step": 56590 }, { "grad_norm": 0.5044453740119934, "learning_rate": 4.32431635672031e-05, "loss": 0.0515, "step": 56600 }, { "grad_norm": 0.4908428490161896, "learning_rate": 4.3226780912232214e-05, "loss": 0.0435, "step": 56610 }, { "grad_norm": 0.2914363443851471, "learning_rate": 4.321039899797036e-05, "loss": 0.0441, "step": 56620 }, { "grad_norm": 0.3303290009498596, "learning_rate": 4.319401782620908e-05, "loss": 0.0396, "step": 56630 }, { "grad_norm": 0.5442796349525452, "learning_rate": 4.317763739873978e-05, "loss": 0.0377, "step": 56640 }, { "grad_norm": 0.45932427048683167, "learning_rate": 4.31612577173538e-05, "loss": 0.0395, "step": 56650 }, { "grad_norm": 0.3889574110507965, "learning_rate": 4.314487878384239e-05, "loss": 0.0455, "step": 56660 }, { "grad_norm": 0.5608391761779785, "learning_rate": 4.3128500599996736e-05, "loss": 0.0468, "step": 56670 }, { "grad_norm": 1.5547653436660767, "learning_rate": 4.31121231676079e-05, "loss": 0.0408, "step": 56680 }, { "grad_norm": 0.5323914885520935, "learning_rate": 4.309574648846694e-05, "loss": 0.0482, "step": 56690 }, { "grad_norm": 0.9033905267715454, "learning_rate": 4.3079370564364755e-05, "loss": 0.0558, "step": 56700 }, { "grad_norm": 0.31537967920303345, "learning_rate": 4.306299539709221e-05, "loss": 0.0429, "step": 56710 }, { "grad_norm": 0.32613110542297363, "learning_rate": 4.3046620988440046e-05, "loss": 0.0435, "step": 56720 }, { "grad_norm": 0.41228199005126953, "learning_rate": 4.303024734019897e-05, "loss": 0.0457, "step": 56730 }, { "grad_norm": 0.41606149077415466, "learning_rate": 4.3013874454159565e-05, "loss": 0.0437, "step": 56740 }, { "grad_norm": 0.7527840733528137, "learning_rate": 4.299750233211233e-05, "loss": 0.0551, "step": 56750 }, { "grad_norm": 0.5307308435440063, "learning_rate": 4.2981130975847715e-05, "loss": 0.0367, "step": 56760 }, { "grad_norm": 0.4433451294898987, "learning_rate": 4.2964760387156076e-05, "loss": 0.042, "step": 56770 }, { "grad_norm": 0.8744753003120422, "learning_rate": 4.294839056782767e-05, "loss": 0.0416, "step": 56780 }, { "grad_norm": 1.0330437421798706, "learning_rate": 4.293202151965266e-05, "loss": 0.0635, "step": 56790 }, { "grad_norm": 0.3337947726249695, "learning_rate": 4.291565324442115e-05, "loss": 0.0373, "step": 56800 }, { "grad_norm": 0.39111655950546265, "learning_rate": 4.2899285743923135e-05, "loss": 0.0412, "step": 56810 }, { "grad_norm": 0.2866142690181732, "learning_rate": 4.288291901994857e-05, "loss": 0.0373, "step": 56820 }, { "grad_norm": 0.3644070029258728, "learning_rate": 4.286655307428729e-05, "loss": 0.0396, "step": 56830 }, { "grad_norm": 0.49139493703842163, "learning_rate": 4.285018790872903e-05, "loss": 0.0479, "step": 56840 }, { "grad_norm": 0.5280561447143555, "learning_rate": 4.283382352506347e-05, "loss": 0.0563, "step": 56850 }, { "grad_norm": 0.300312340259552, "learning_rate": 4.2817459925080205e-05, "loss": 0.0574, "step": 56860 }, { "grad_norm": 0.49687913060188293, "learning_rate": 4.28010971105687e-05, "loss": 0.0485, "step": 56870 }, { "grad_norm": 0.9180229902267456, "learning_rate": 4.278473508331841e-05, "loss": 0.0457, "step": 56880 }, { "grad_norm": 0.5614893436431885, "learning_rate": 4.276837384511864e-05, "loss": 0.0483, "step": 56890 }, { "grad_norm": 0.3060745596885681, "learning_rate": 4.275201339775864e-05, "loss": 0.0392, "step": 56900 }, { "grad_norm": 0.46133357286453247, "learning_rate": 4.273565374302756e-05, "loss": 0.0428, "step": 56910 }, { "grad_norm": 0.5042287707328796, "learning_rate": 4.2719294882714454e-05, "loss": 0.0372, "step": 56920 }, { "grad_norm": 0.34391623735427856, "learning_rate": 4.270293681860831e-05, "loss": 0.0367, "step": 56930 }, { "grad_norm": 0.49822133779525757, "learning_rate": 4.268657955249802e-05, "loss": 0.0373, "step": 56940 }, { "grad_norm": 0.5781489014625549, "learning_rate": 4.267022308617241e-05, "loss": 0.0368, "step": 56950 }, { "grad_norm": 0.6885257959365845, "learning_rate": 4.265386742142018e-05, "loss": 0.0462, "step": 56960 }, { "grad_norm": 0.5362452864646912, "learning_rate": 4.263751256002996e-05, "loss": 0.0387, "step": 56970 }, { "grad_norm": 0.5408272743225098, "learning_rate": 4.26211585037903e-05, "loss": 0.0475, "step": 56980 }, { "grad_norm": 0.4518430829048157, "learning_rate": 4.2604805254489656e-05, "loss": 0.0433, "step": 56990 }, { "grad_norm": 0.43886223435401917, "learning_rate": 4.258845281391638e-05, "loss": 0.034, "step": 57000 }, { "grad_norm": 0.352512925863266, "learning_rate": 4.257210118385876e-05, "loss": 0.0435, "step": 57010 }, { "grad_norm": 0.6007649898529053, "learning_rate": 4.2555750366104994e-05, "loss": 0.048, "step": 57020 }, { "grad_norm": 0.4124850034713745, "learning_rate": 4.2539400362443184e-05, "loss": 0.0354, "step": 57030 }, { "grad_norm": 0.5317369699478149, "learning_rate": 4.252305117466133e-05, "loss": 0.0528, "step": 57040 }, { "grad_norm": 0.37457743287086487, "learning_rate": 4.2506702804547356e-05, "loss": 0.0421, "step": 57050 }, { "grad_norm": 1.2171770334243774, "learning_rate": 4.249035525388907e-05, "loss": 0.044, "step": 57060 }, { "grad_norm": 0.560752809047699, "learning_rate": 4.247400852447427e-05, "loss": 0.0431, "step": 57070 }, { "grad_norm": 0.39033153653144836, "learning_rate": 4.245766261809059e-05, "loss": 0.0434, "step": 57080 }, { "grad_norm": 0.377248615026474, "learning_rate": 4.244131753652558e-05, "loss": 0.0383, "step": 57090 }, { "grad_norm": 0.356408953666687, "learning_rate": 4.242497328156672e-05, "loss": 0.0381, "step": 57100 }, { "grad_norm": 0.5944132208824158, "learning_rate": 4.240862985500139e-05, "loss": 0.0469, "step": 57110 }, { "grad_norm": 0.4016793370246887, "learning_rate": 4.2392287258616894e-05, "loss": 0.0362, "step": 57120 }, { "grad_norm": 0.23683220148086548, "learning_rate": 4.2375945494200396e-05, "loss": 0.0322, "step": 57130 }, { "grad_norm": 0.4169709086418152, "learning_rate": 4.2359604563539054e-05, "loss": 0.0458, "step": 57140 }, { "grad_norm": 0.30366191267967224, "learning_rate": 4.2343264468419885e-05, "loss": 0.0399, "step": 57150 }, { "grad_norm": 0.5004364252090454, "learning_rate": 4.232692521062979e-05, "loss": 0.0286, "step": 57160 }, { "grad_norm": 0.47013723850250244, "learning_rate": 4.231058679195562e-05, "loss": 0.0378, "step": 57170 }, { "grad_norm": 0.46790042519569397, "learning_rate": 4.229424921418411e-05, "loss": 0.0398, "step": 57180 }, { "grad_norm": 0.5053375363349915, "learning_rate": 4.2277912479101916e-05, "loss": 0.0423, "step": 57190 }, { "grad_norm": 0.6884238719940186, "learning_rate": 4.226157658849562e-05, "loss": 0.0436, "step": 57200 }, { "grad_norm": 0.29672858119010925, "learning_rate": 4.2245241544151664e-05, "loss": 0.0366, "step": 57210 }, { "grad_norm": 0.6992620825767517, "learning_rate": 4.2228907347856425e-05, "loss": 0.0547, "step": 57220 }, { "grad_norm": 0.6402353644371033, "learning_rate": 4.2212574001396214e-05, "loss": 0.0458, "step": 57230 }, { "grad_norm": 0.8188613653182983, "learning_rate": 4.21962415065572e-05, "loss": 0.04, "step": 57240 }, { "grad_norm": 0.47027748823165894, "learning_rate": 4.217990986512548e-05, "loss": 0.0408, "step": 57250 }, { "grad_norm": 0.4398171603679657, "learning_rate": 4.216357907888704e-05, "loss": 0.041, "step": 57260 }, { "grad_norm": 0.6584342122077942, "learning_rate": 4.2147249149627824e-05, "loss": 0.0411, "step": 57270 }, { "grad_norm": 0.39244794845581055, "learning_rate": 4.213092007913364e-05, "loss": 0.0384, "step": 57280 }, { "grad_norm": 0.23213312029838562, "learning_rate": 4.2114591869190205e-05, "loss": 0.0447, "step": 57290 }, { "grad_norm": 0.3005918264389038, "learning_rate": 4.209826452158315e-05, "loss": 0.0475, "step": 57300 }, { "grad_norm": 0.5835704803466797, "learning_rate": 4.2081938038098e-05, "loss": 0.0443, "step": 57310 }, { "grad_norm": 0.5545641779899597, "learning_rate": 4.2065612420520185e-05, "loss": 0.0358, "step": 57320 }, { "grad_norm": 0.4264589548110962, "learning_rate": 4.204928767063509e-05, "loss": 0.0443, "step": 57330 }, { "grad_norm": 0.3641602694988251, "learning_rate": 4.2032963790227944e-05, "loss": 0.0396, "step": 57340 }, { "grad_norm": 0.29407235980033875, "learning_rate": 4.2016640781083887e-05, "loss": 0.0391, "step": 57350 }, { "grad_norm": 0.38733720779418945, "learning_rate": 4.2000318644988004e-05, "loss": 0.0594, "step": 57360 }, { "grad_norm": 0.7390644550323486, "learning_rate": 4.198399738372524e-05, "loss": 0.0371, "step": 57370 }, { "grad_norm": 0.2869701087474823, "learning_rate": 4.196767699908045e-05, "loss": 0.036, "step": 57380 }, { "grad_norm": 0.5401166081428528, "learning_rate": 4.195135749283845e-05, "loss": 0.0409, "step": 57390 }, { "grad_norm": 0.3470911383628845, "learning_rate": 4.1935038866783894e-05, "loss": 0.0366, "step": 57400 }, { "grad_norm": 0.25715264678001404, "learning_rate": 4.1918721122701355e-05, "loss": 0.035, "step": 57410 }, { "grad_norm": 0.6103876233100891, "learning_rate": 4.190240426237533e-05, "loss": 0.0537, "step": 57420 }, { "grad_norm": 0.3036309778690338, "learning_rate": 4.188608828759019e-05, "loss": 0.043, "step": 57430 }, { "grad_norm": 0.582626223564148, "learning_rate": 4.1869773200130246e-05, "loss": 0.0698, "step": 57440 }, { "grad_norm": 0.6509023904800415, "learning_rate": 4.185345900177967e-05, "loss": 0.051, "step": 57450 }, { "grad_norm": 0.30081769824028015, "learning_rate": 4.183714569432258e-05, "loss": 0.0407, "step": 57460 }, { "grad_norm": 0.6162611842155457, "learning_rate": 4.182083327954295e-05, "loss": 0.042, "step": 57470 }, { "grad_norm": 0.332327276468277, "learning_rate": 4.1804521759224714e-05, "loss": 0.0379, "step": 57480 }, { "grad_norm": 0.316574364900589, "learning_rate": 4.178821113515165e-05, "loss": 0.0414, "step": 57490 }, { "grad_norm": 0.412325918674469, "learning_rate": 4.1771901409107476e-05, "loss": 0.0459, "step": 57500 }, { "grad_norm": 0.632709264755249, "learning_rate": 4.175559258287577e-05, "loss": 0.0384, "step": 57510 }, { "grad_norm": 0.6713753342628479, "learning_rate": 4.173928465824008e-05, "loss": 0.0375, "step": 57520 }, { "grad_norm": 0.612429141998291, "learning_rate": 4.17229776369838e-05, "loss": 0.0417, "step": 57530 }, { "grad_norm": 0.6066552996635437, "learning_rate": 4.170667152089025e-05, "loss": 0.0502, "step": 57540 }, { "grad_norm": 0.538979172706604, "learning_rate": 4.1690366311742626e-05, "loss": 0.0389, "step": 57550 }, { "grad_norm": 0.5711348652839661, "learning_rate": 4.1674062011324055e-05, "loss": 0.0394, "step": 57560 }, { "grad_norm": 0.29344889521598816, "learning_rate": 4.165775862141752e-05, "loss": 0.0315, "step": 57570 }, { "grad_norm": 0.5538257956504822, "learning_rate": 4.164145614380598e-05, "loss": 0.0361, "step": 57580 }, { "grad_norm": 0.31320780515670776, "learning_rate": 4.162515458027223e-05, "loss": 0.0329, "step": 57590 }, { "grad_norm": 0.5804953575134277, "learning_rate": 4.160885393259899e-05, "loss": 0.0442, "step": 57600 }, { "grad_norm": 0.2977253496646881, "learning_rate": 4.159255420256886e-05, "loss": 0.0436, "step": 57610 }, { "grad_norm": 0.4685066044330597, "learning_rate": 4.157625539196436e-05, "loss": 0.0353, "step": 57620 }, { "grad_norm": 0.3608221411705017, "learning_rate": 4.155995750256791e-05, "loss": 0.0471, "step": 57630 }, { "grad_norm": 0.3586113452911377, "learning_rate": 4.15436605361618e-05, "loss": 0.0458, "step": 57640 }, { "grad_norm": 0.4317810535430908, "learning_rate": 4.152736449452827e-05, "loss": 0.0525, "step": 57650 }, { "grad_norm": 0.6780093312263489, "learning_rate": 4.151106937944943e-05, "loss": 0.0413, "step": 57660 }, { "grad_norm": 0.6045551896095276, "learning_rate": 4.149477519270727e-05, "loss": 0.0382, "step": 57670 }, { "grad_norm": 0.7968471646308899, "learning_rate": 4.147848193608371e-05, "loss": 0.0534, "step": 57680 }, { "grad_norm": 0.4560297429561615, "learning_rate": 4.146218961136056e-05, "loss": 0.038, "step": 57690 }, { "grad_norm": 0.32403022050857544, "learning_rate": 4.14458982203195e-05, "loss": 0.0524, "step": 57700 }, { "grad_norm": 0.34827977418899536, "learning_rate": 4.1429607764742165e-05, "loss": 0.0475, "step": 57710 }, { "grad_norm": 0.36794576048851013, "learning_rate": 4.141331824641003e-05, "loss": 0.0468, "step": 57720 }, { "grad_norm": 0.5607069134712219, "learning_rate": 4.139702966710452e-05, "loss": 0.0499, "step": 57730 }, { "grad_norm": 0.387060284614563, "learning_rate": 4.13807420286069e-05, "loss": 0.0449, "step": 57740 }, { "grad_norm": 0.4007674753665924, "learning_rate": 4.136445533269837e-05, "loss": 0.0471, "step": 57750 }, { "grad_norm": 0.3118516504764557, "learning_rate": 4.134816958116003e-05, "loss": 0.0303, "step": 57760 }, { "grad_norm": 0.5922039747238159, "learning_rate": 4.133188477577283e-05, "loss": 0.0431, "step": 57770 }, { "grad_norm": 0.42906129360198975, "learning_rate": 4.131560091831771e-05, "loss": 0.0357, "step": 57780 }, { "grad_norm": 0.6153445839881897, "learning_rate": 4.1299318010575405e-05, "loss": 0.0341, "step": 57790 }, { "grad_norm": 0.5163223147392273, "learning_rate": 4.128303605432661e-05, "loss": 0.0649, "step": 57800 }, { "grad_norm": 0.497296541929245, "learning_rate": 4.126675505135188e-05, "loss": 0.0412, "step": 57810 }, { "grad_norm": 0.44441017508506775, "learning_rate": 4.125047500343169e-05, "loss": 0.0394, "step": 57820 }, { "grad_norm": 0.5131333470344543, "learning_rate": 4.1234195912346376e-05, "loss": 0.0415, "step": 57830 }, { "grad_norm": 0.587757408618927, "learning_rate": 4.1217917779876235e-05, "loss": 0.0428, "step": 57840 }, { "grad_norm": 0.3598572015762329, "learning_rate": 4.120164060780141e-05, "loss": 0.05, "step": 57850 }, { "grad_norm": 0.367369145154953, "learning_rate": 4.1185364397901935e-05, "loss": 0.0404, "step": 57860 }, { "grad_norm": 0.3281102180480957, "learning_rate": 4.116908915195776e-05, "loss": 0.035, "step": 57870 }, { "grad_norm": 0.4205820560455322, "learning_rate": 4.115281487174872e-05, "loss": 0.036, "step": 57880 }, { "grad_norm": 1.3363350629806519, "learning_rate": 4.113654155905452e-05, "loss": 0.0409, "step": 57890 }, { "grad_norm": 0.46778061985969543, "learning_rate": 4.112026921565483e-05, "loss": 0.0368, "step": 57900 }, { "grad_norm": 0.5722357034683228, "learning_rate": 4.1103997843329147e-05, "loss": 0.0378, "step": 57910 }, { "grad_norm": 0.40327557921409607, "learning_rate": 4.108772744385689e-05, "loss": 0.0552, "step": 57920 }, { "grad_norm": 0.463484525680542, "learning_rate": 4.1071458019017346e-05, "loss": 0.0376, "step": 57930 }, { "grad_norm": 0.4083722233772278, "learning_rate": 4.1055189570589734e-05, "loss": 0.0462, "step": 57940 }, { "grad_norm": 0.4597293734550476, "learning_rate": 4.103892210035315e-05, "loss": 0.0311, "step": 57950 }, { "grad_norm": 0.9864412546157837, "learning_rate": 4.102265561008656e-05, "loss": 0.0466, "step": 57960 }, { "grad_norm": 0.4216652810573578, "learning_rate": 4.100639010156885e-05, "loss": 0.036, "step": 57970 }, { "grad_norm": 0.5149017572402954, "learning_rate": 4.099012557657882e-05, "loss": 0.0481, "step": 57980 }, { "grad_norm": 0.4870668351650238, "learning_rate": 4.09738620368951e-05, "loss": 0.0418, "step": 57990 }, { "grad_norm": 0.22310535609722137, "learning_rate": 4.095759948429626e-05, "loss": 0.0349, "step": 58000 }, { "grad_norm": 0.5311328768730164, "learning_rate": 4.094133792056075e-05, "loss": 0.0477, "step": 58010 }, { "grad_norm": 0.4241655766963959, "learning_rate": 4.092507734746688e-05, "loss": 0.0462, "step": 58020 }, { "grad_norm": 0.7022945284843445, "learning_rate": 4.090881776679293e-05, "loss": 0.0385, "step": 58030 }, { "grad_norm": 0.4395521283149719, "learning_rate": 4.0892559180316994e-05, "loss": 0.0395, "step": 58040 }, { "grad_norm": 0.5195900201797485, "learning_rate": 4.08763015898171e-05, "loss": 0.0396, "step": 58050 }, { "grad_norm": 0.48832446336746216, "learning_rate": 4.0860044997071145e-05, "loss": 0.0436, "step": 58060 }, { "grad_norm": 0.607245147228241, "learning_rate": 4.084378940385693e-05, "loss": 0.0461, "step": 58070 }, { "grad_norm": 0.4756297171115875, "learning_rate": 4.082753481195211e-05, "loss": 0.0382, "step": 58080 }, { "grad_norm": 0.3795802891254425, "learning_rate": 4.0811281223134324e-05, "loss": 0.0393, "step": 58090 }, { "grad_norm": 0.29816514253616333, "learning_rate": 4.0795028639181e-05, "loss": 0.0473, "step": 58100 }, { "grad_norm": 0.6829971075057983, "learning_rate": 4.077877706186951e-05, "loss": 0.0457, "step": 58110 }, { "grad_norm": 0.47655820846557617, "learning_rate": 4.076252649297709e-05, "loss": 0.0408, "step": 58120 }, { "grad_norm": 0.45476898550987244, "learning_rate": 4.07462769342809e-05, "loss": 0.0387, "step": 58130 }, { "grad_norm": 0.5086646676063538, "learning_rate": 4.073002838755793e-05, "loss": 0.0486, "step": 58140 }, { "grad_norm": 0.6004117727279663, "learning_rate": 4.0713780854585115e-05, "loss": 0.0457, "step": 58150 }, { "grad_norm": 0.425080806016922, "learning_rate": 4.069753433713927e-05, "loss": 0.0495, "step": 58160 }, { "grad_norm": 0.48823219537734985, "learning_rate": 4.068128883699709e-05, "loss": 0.0337, "step": 58170 }, { "grad_norm": 0.5868196487426758, "learning_rate": 4.0665044355935155e-05, "loss": 0.0369, "step": 58180 }, { "grad_norm": 0.3283546566963196, "learning_rate": 4.064880089572993e-05, "loss": 0.0389, "step": 58190 }, { "grad_norm": 0.5137825012207031, "learning_rate": 4.063255845815779e-05, "loss": 0.0414, "step": 58200 }, { "grad_norm": 0.3071942925453186, "learning_rate": 4.0616317044994945e-05, "loss": 0.0454, "step": 58210 }, { "grad_norm": 0.3045174181461334, "learning_rate": 4.0600076658017585e-05, "loss": 0.0331, "step": 58220 }, { "grad_norm": 0.4495464563369751, "learning_rate": 4.0583837299001706e-05, "loss": 0.0389, "step": 58230 }, { "grad_norm": 0.32534390687942505, "learning_rate": 4.056759896972322e-05, "loss": 0.0363, "step": 58240 }, { "grad_norm": 0.407268762588501, "learning_rate": 4.055136167195793e-05, "loss": 0.0415, "step": 58250 }, { "grad_norm": 0.3605334758758545, "learning_rate": 4.053512540748152e-05, "loss": 0.0402, "step": 58260 }, { "grad_norm": 0.9305031299591064, "learning_rate": 4.051889017806955e-05, "loss": 0.0488, "step": 58270 }, { "grad_norm": 0.7285686135292053, "learning_rate": 4.0502655985497516e-05, "loss": 0.0396, "step": 58280 }, { "grad_norm": 0.5373605489730835, "learning_rate": 4.048642283154074e-05, "loss": 0.0487, "step": 58290 }, { "grad_norm": 0.2941879630088806, "learning_rate": 4.0470190717974464e-05, "loss": 0.0435, "step": 58300 }, { "grad_norm": 0.3529393970966339, "learning_rate": 4.045395964657381e-05, "loss": 0.043, "step": 58310 }, { "grad_norm": 0.6196691393852234, "learning_rate": 4.043772961911376e-05, "loss": 0.0511, "step": 58320 }, { "grad_norm": 0.48576653003692627, "learning_rate": 4.042150063736924e-05, "loss": 0.0426, "step": 58330 }, { "grad_norm": 0.48329219222068787, "learning_rate": 4.0405272703114976e-05, "loss": 0.039, "step": 58340 }, { "grad_norm": 0.5925858020782471, "learning_rate": 4.038904581812569e-05, "loss": 0.0431, "step": 58350 }, { "grad_norm": 0.3159629702568054, "learning_rate": 4.03728199841759e-05, "loss": 0.0383, "step": 58360 }, { "grad_norm": 0.4282558262348175, "learning_rate": 4.0356595203040044e-05, "loss": 0.0444, "step": 58370 }, { "grad_norm": 0.29753628373146057, "learning_rate": 4.0340371476492426e-05, "loss": 0.0457, "step": 58380 }, { "grad_norm": 0.4431384801864624, "learning_rate": 4.032414880630727e-05, "loss": 0.0376, "step": 58390 }, { "grad_norm": 0.5725922584533691, "learning_rate": 4.0307927194258646e-05, "loss": 0.034, "step": 58400 }, { "grad_norm": 0.46632400155067444, "learning_rate": 4.029170664212054e-05, "loss": 0.0397, "step": 58410 }, { "grad_norm": 0.37928590178489685, "learning_rate": 4.0275487151666794e-05, "loss": 0.0491, "step": 58420 }, { "grad_norm": 0.45572710037231445, "learning_rate": 4.025926872467116e-05, "loss": 0.0446, "step": 58430 }, { "grad_norm": 0.5081309080123901, "learning_rate": 4.024305136290725e-05, "loss": 0.0489, "step": 58440 }, { "grad_norm": 0.3260208070278168, "learning_rate": 4.022683506814858e-05, "loss": 0.0461, "step": 58450 }, { "grad_norm": 0.5412026643753052, "learning_rate": 4.021061984216853e-05, "loss": 0.0442, "step": 58460 }, { "grad_norm": 0.45452582836151123, "learning_rate": 4.019440568674035e-05, "loss": 0.0393, "step": 58470 }, { "grad_norm": 0.7145681381225586, "learning_rate": 4.017819260363724e-05, "loss": 0.0387, "step": 58480 }, { "grad_norm": 0.5178419351577759, "learning_rate": 4.016198059463223e-05, "loss": 0.0567, "step": 58490 }, { "grad_norm": 0.2146233469247818, "learning_rate": 4.0145769661498233e-05, "loss": 0.0375, "step": 58500 }, { "grad_norm": 0.5409200191497803, "learning_rate": 4.012955980600804e-05, "loss": 0.0399, "step": 58510 }, { "grad_norm": 0.7299923300743103, "learning_rate": 4.0113351029934355e-05, "loss": 0.0443, "step": 58520 }, { "grad_norm": 0.3268757164478302, "learning_rate": 4.0097143335049714e-05, "loss": 0.0397, "step": 58530 }, { "grad_norm": 0.6480310559272766, "learning_rate": 4.0080936723126606e-05, "loss": 0.0329, "step": 58540 }, { "grad_norm": 0.3314215838909149, "learning_rate": 4.006473119593735e-05, "loss": 0.0391, "step": 58550 }, { "grad_norm": 0.5264586806297302, "learning_rate": 4.004852675525415e-05, "loss": 0.0428, "step": 58560 }, { "grad_norm": 0.4121890664100647, "learning_rate": 4.00323234028491e-05, "loss": 0.038, "step": 58570 }, { "grad_norm": 0.4035692811012268, "learning_rate": 4.0016121140494186e-05, "loss": 0.0402, "step": 58580 }, { "grad_norm": 0.34951168298721313, "learning_rate": 3.999991996996122e-05, "loss": 0.0397, "step": 58590 }, { "grad_norm": 0.37812215089797974, "learning_rate": 3.9983719893022e-05, "loss": 0.0371, "step": 58600 }, { "grad_norm": 0.5301382541656494, "learning_rate": 3.996752091144812e-05, "loss": 0.0434, "step": 58610 }, { "grad_norm": 0.478074848651886, "learning_rate": 3.9951323027011055e-05, "loss": 0.041, "step": 58620 }, { "grad_norm": 0.4515411853790283, "learning_rate": 3.993512624148219e-05, "loss": 0.039, "step": 58630 }, { "grad_norm": 0.43539246916770935, "learning_rate": 3.99189305566328e-05, "loss": 0.0393, "step": 58640 }, { "grad_norm": 0.2967379093170166, "learning_rate": 3.990273597423401e-05, "loss": 0.0479, "step": 58650 }, { "grad_norm": 0.30877116322517395, "learning_rate": 3.988654249605681e-05, "loss": 0.0413, "step": 58660 }, { "grad_norm": 0.5475128293037415, "learning_rate": 3.987035012387213e-05, "loss": 0.0402, "step": 58670 }, { "grad_norm": 0.42128491401672363, "learning_rate": 3.9854158859450735e-05, "loss": 0.0405, "step": 58680 }, { "grad_norm": 0.368977427482605, "learning_rate": 3.983796870456328e-05, "loss": 0.0434, "step": 58690 }, { "grad_norm": 0.23965053260326385, "learning_rate": 3.982177966098028e-05, "loss": 0.0323, "step": 58700 }, { "grad_norm": 0.4496389627456665, "learning_rate": 3.980559173047216e-05, "loss": 0.0545, "step": 58710 }, { "grad_norm": 0.3778967261314392, "learning_rate": 3.9789404914809176e-05, "loss": 0.0458, "step": 58720 }, { "grad_norm": 0.6024903655052185, "learning_rate": 3.9773219215761544e-05, "loss": 0.0422, "step": 58730 }, { "grad_norm": 0.3349936902523041, "learning_rate": 3.975703463509928e-05, "loss": 0.031, "step": 58740 }, { "grad_norm": 0.8399305939674377, "learning_rate": 3.974085117459231e-05, "loss": 0.0371, "step": 58750 }, { "grad_norm": 0.48100605607032776, "learning_rate": 3.972466883601043e-05, "loss": 0.0457, "step": 58760 }, { "grad_norm": 0.33742958307266235, "learning_rate": 3.9708487621123316e-05, "loss": 0.0283, "step": 58770 }, { "grad_norm": 0.4231133460998535, "learning_rate": 3.969230753170051e-05, "loss": 0.0462, "step": 58780 }, { "grad_norm": 1.458587408065796, "learning_rate": 3.967612856951146e-05, "loss": 0.0424, "step": 58790 }, { "grad_norm": 0.49606814980506897, "learning_rate": 3.9659950736325476e-05, "loss": 0.0365, "step": 58800 }, { "grad_norm": 1.2344549894332886, "learning_rate": 3.964377403391174e-05, "loss": 0.0728, "step": 58810 }, { "grad_norm": 0.6861191987991333, "learning_rate": 3.9627598464039303e-05, "loss": 0.0613, "step": 58820 }, { "grad_norm": 0.256349116563797, "learning_rate": 3.9611424028477096e-05, "loss": 0.0326, "step": 58830 }, { "grad_norm": 0.6513186097145081, "learning_rate": 3.959525072899394e-05, "loss": 0.0442, "step": 58840 }, { "grad_norm": 0.4059513509273529, "learning_rate": 3.9579078567358506e-05, "loss": 0.0375, "step": 58850 }, { "grad_norm": 0.6345998644828796, "learning_rate": 3.956290754533939e-05, "loss": 0.0451, "step": 58860 }, { "grad_norm": 1.0530680418014526, "learning_rate": 3.954673766470501e-05, "loss": 0.0459, "step": 58870 }, { "grad_norm": 0.31800439953804016, "learning_rate": 3.9530568927223674e-05, "loss": 0.0403, "step": 58880 }, { "grad_norm": 0.34576737880706787, "learning_rate": 3.95144013346636e-05, "loss": 0.0503, "step": 58890 }, { "grad_norm": 0.5849835276603699, "learning_rate": 3.949823488879282e-05, "loss": 0.0308, "step": 58900 }, { "grad_norm": 0.42438361048698425, "learning_rate": 3.9482069591379276e-05, "loss": 0.0467, "step": 58910 }, { "grad_norm": 0.39609625935554504, "learning_rate": 3.9465905444190784e-05, "loss": 0.0341, "step": 58920 }, { "grad_norm": 0.2682783305644989, "learning_rate": 3.944974244899505e-05, "loss": 0.0452, "step": 58930 }, { "grad_norm": 0.49721014499664307, "learning_rate": 3.9433580607559616e-05, "loss": 0.0407, "step": 58940 }, { "grad_norm": 0.3208382725715637, "learning_rate": 3.941741992165193e-05, "loss": 0.0373, "step": 58950 }, { "grad_norm": 0.45743197202682495, "learning_rate": 3.940126039303928e-05, "loss": 0.0416, "step": 58960 }, { "grad_norm": 0.576265811920166, "learning_rate": 3.938510202348886e-05, "loss": 0.0448, "step": 58970 }, { "grad_norm": 0.36163079738616943, "learning_rate": 3.93689448147677e-05, "loss": 0.0379, "step": 58980 }, { "grad_norm": 0.3573300540447235, "learning_rate": 3.935278876864278e-05, "loss": 0.0376, "step": 58990 }, { "grad_norm": 0.34776777029037476, "learning_rate": 3.933663388688087e-05, "loss": 0.0431, "step": 59000 }, { "grad_norm": 0.5069819092750549, "learning_rate": 3.9320480171248645e-05, "loss": 0.0369, "step": 59010 }, { "grad_norm": 0.5851929187774658, "learning_rate": 3.930432762351265e-05, "loss": 0.0432, "step": 59020 }, { "grad_norm": 0.4361882507801056, "learning_rate": 3.928817624543931e-05, "loss": 0.0436, "step": 59030 }, { "grad_norm": 0.3378859758377075, "learning_rate": 3.9272026038794885e-05, "loss": 0.0346, "step": 59040 }, { "grad_norm": 0.4722306430339813, "learning_rate": 3.9255877005345586e-05, "loss": 0.0397, "step": 59050 }, { "grad_norm": 0.5782789587974548, "learning_rate": 3.923972914685743e-05, "loss": 0.0404, "step": 59060 }, { "grad_norm": 0.3903593122959137, "learning_rate": 3.922358246509631e-05, "loss": 0.0363, "step": 59070 }, { "grad_norm": 0.626870334148407, "learning_rate": 3.920743696182801e-05, "loss": 0.044, "step": 59080 }, { "grad_norm": 0.6088133454322815, "learning_rate": 3.9191292638818164e-05, "loss": 0.0407, "step": 59090 }, { "grad_norm": 0.5769241452217102, "learning_rate": 3.9175149497832316e-05, "loss": 0.0357, "step": 59100 }, { "grad_norm": 0.4466109275817871, "learning_rate": 3.915900754063584e-05, "loss": 0.0327, "step": 59110 }, { "grad_norm": 0.38832888007164, "learning_rate": 3.9142866768994014e-05, "loss": 0.0368, "step": 59120 }, { "grad_norm": 0.3949759006500244, "learning_rate": 3.912672718467193e-05, "loss": 0.0481, "step": 59130 }, { "grad_norm": 0.5562381744384766, "learning_rate": 3.911058878943463e-05, "loss": 0.0398, "step": 59140 }, { "grad_norm": 0.3235515058040619, "learning_rate": 3.9094451585046974e-05, "loss": 0.0325, "step": 59150 }, { "grad_norm": 0.34038248658180237, "learning_rate": 3.90783155732737e-05, "loss": 0.0379, "step": 59160 }, { "grad_norm": 0.4449388384819031, "learning_rate": 3.9062180755879373e-05, "loss": 0.0361, "step": 59170 }, { "grad_norm": 0.7081820368766785, "learning_rate": 3.904604713462855e-05, "loss": 0.047, "step": 59180 }, { "grad_norm": 0.3176577389240265, "learning_rate": 3.902991471128554e-05, "loss": 0.031, "step": 59190 }, { "grad_norm": 0.21339353919029236, "learning_rate": 3.901378348761456e-05, "loss": 0.0376, "step": 59200 }, { "grad_norm": 0.3568498492240906, "learning_rate": 3.899765346537971e-05, "loss": 0.0307, "step": 59210 }, { "grad_norm": 0.31691595911979675, "learning_rate": 3.898152464634493e-05, "loss": 0.0295, "step": 59220 }, { "grad_norm": 0.35965538024902344, "learning_rate": 3.896539703227402e-05, "loss": 0.0362, "step": 59230 }, { "grad_norm": 0.5955542922019958, "learning_rate": 3.894927062493073e-05, "loss": 0.0403, "step": 59240 }, { "grad_norm": 0.5178760886192322, "learning_rate": 3.893314542607859e-05, "loss": 0.0461, "step": 59250 }, { "grad_norm": 0.39135104417800903, "learning_rate": 3.891702143748103e-05, "loss": 0.0366, "step": 59260 }, { "grad_norm": 0.5194774270057678, "learning_rate": 3.890089866090133e-05, "loss": 0.0396, "step": 59270 }, { "grad_norm": 0.4101332724094391, "learning_rate": 3.888477709810267e-05, "loss": 0.0345, "step": 59280 }, { "grad_norm": 0.4431515634059906, "learning_rate": 3.886865675084806e-05, "loss": 0.0382, "step": 59290 }, { "grad_norm": 0.46652770042419434, "learning_rate": 3.885253762090043e-05, "loss": 0.0416, "step": 59300 }, { "grad_norm": 0.5224973559379578, "learning_rate": 3.8836419710022534e-05, "loss": 0.056, "step": 59310 }, { "grad_norm": 0.4874669313430786, "learning_rate": 3.882030301997698e-05, "loss": 0.0415, "step": 59320 }, { "grad_norm": 0.5274742841720581, "learning_rate": 3.880418755252629e-05, "loss": 0.0363, "step": 59330 }, { "grad_norm": 0.7610743045806885, "learning_rate": 3.87880733094328e-05, "loss": 0.0478, "step": 59340 }, { "grad_norm": 0.7349393963813782, "learning_rate": 3.877196029245877e-05, "loss": 0.0325, "step": 59350 }, { "grad_norm": 0.41319739818573, "learning_rate": 3.875584850336627e-05, "loss": 0.0546, "step": 59360 }, { "grad_norm": 0.38360685110092163, "learning_rate": 3.873973794391728e-05, "loss": 0.0354, "step": 59370 }, { "grad_norm": 0.28082865476608276, "learning_rate": 3.872362861587361e-05, "loss": 0.029, "step": 59380 }, { "grad_norm": 0.4199638068675995, "learning_rate": 3.870752052099698e-05, "loss": 0.0413, "step": 59390 }, { "grad_norm": 0.28088971972465515, "learning_rate": 3.8691413661048916e-05, "loss": 0.046, "step": 59400 }, { "grad_norm": 1.6822929382324219, "learning_rate": 3.867530803779085e-05, "loss": 0.0431, "step": 59410 }, { "grad_norm": 0.46530139446258545, "learning_rate": 3.8659203652984054e-05, "loss": 0.0374, "step": 59420 }, { "grad_norm": 0.3888809382915497, "learning_rate": 3.864310050838972e-05, "loss": 0.0414, "step": 59430 }, { "grad_norm": 0.26552045345306396, "learning_rate": 3.8626998605768835e-05, "loss": 0.0403, "step": 59440 }, { "grad_norm": 0.4721772074699402, "learning_rate": 3.8610897946882294e-05, "loss": 0.0446, "step": 59450 }, { "grad_norm": 0.6237738132476807, "learning_rate": 3.859479853349082e-05, "loss": 0.0449, "step": 59460 }, { "grad_norm": 1.1822772026062012, "learning_rate": 3.857870036735503e-05, "loss": 0.0435, "step": 59470 }, { "grad_norm": 0.5174084901809692, "learning_rate": 3.856260345023538e-05, "loss": 0.048, "step": 59480 }, { "grad_norm": 0.2932215929031372, "learning_rate": 3.854650778389224e-05, "loss": 0.0356, "step": 59490 }, { "grad_norm": 0.5666347146034241, "learning_rate": 3.8530413370085786e-05, "loss": 0.0319, "step": 59500 }, { "grad_norm": 0.35444846749305725, "learning_rate": 3.851432021057608e-05, "loss": 0.0345, "step": 59510 }, { "grad_norm": 0.45538702607154846, "learning_rate": 3.849822830712306e-05, "loss": 0.0401, "step": 59520 }, { "grad_norm": 0.4693867266178131, "learning_rate": 3.848213766148649e-05, "loss": 0.04, "step": 59530 }, { "grad_norm": 0.4577007293701172, "learning_rate": 3.846604827542602e-05, "loss": 0.0337, "step": 59540 }, { "grad_norm": 0.5080556869506836, "learning_rate": 3.844996015070117e-05, "loss": 0.0416, "step": 59550 }, { "grad_norm": 0.3549930453300476, "learning_rate": 3.843387328907132e-05, "loss": 0.0364, "step": 59560 }, { "grad_norm": 0.4181787073612213, "learning_rate": 3.8417787692295696e-05, "loss": 0.0455, "step": 59570 }, { "grad_norm": 0.39931225776672363, "learning_rate": 3.84017033621334e-05, "loss": 0.0358, "step": 59580 }, { "grad_norm": 0.24037601053714752, "learning_rate": 3.838562030034338e-05, "loss": 0.0327, "step": 59590 }, { "grad_norm": 0.41707056760787964, "learning_rate": 3.836953850868447e-05, "loss": 0.0355, "step": 59600 }, { "grad_norm": 0.3782923221588135, "learning_rate": 3.835345798891533e-05, "loss": 0.0326, "step": 59610 }, { "grad_norm": 0.36718952655792236, "learning_rate": 3.833737874279452e-05, "loss": 0.0402, "step": 59620 }, { "grad_norm": 0.4256573021411896, "learning_rate": 3.8321300772080433e-05, "loss": 0.0245, "step": 59630 }, { "grad_norm": 0.8767553567886353, "learning_rate": 3.8305224078531345e-05, "loss": 0.0416, "step": 59640 }, { "grad_norm": 0.48548269271850586, "learning_rate": 3.828914866390537e-05, "loss": 0.0475, "step": 59650 }, { "grad_norm": 0.37689968943595886, "learning_rate": 3.827307452996048e-05, "loss": 0.0398, "step": 59660 }, { "grad_norm": 0.47216200828552246, "learning_rate": 3.825700167845454e-05, "loss": 0.0485, "step": 59670 }, { "grad_norm": 0.5472368597984314, "learning_rate": 3.8240930111145214e-05, "loss": 0.0505, "step": 59680 }, { "grad_norm": 0.44049668312072754, "learning_rate": 3.8224859829790116e-05, "loss": 0.0356, "step": 59690 }, { "grad_norm": 1.4613242149353027, "learning_rate": 3.820879083614664e-05, "loss": 0.0501, "step": 59700 }, { "grad_norm": 0.5342957377433777, "learning_rate": 3.819272313197208e-05, "loss": 0.0428, "step": 59710 }, { "grad_norm": 1.9031908512115479, "learning_rate": 3.817665671902355e-05, "loss": 0.0402, "step": 59720 }, { "grad_norm": 0.4784861207008362, "learning_rate": 3.816059159905808e-05, "loss": 0.0436, "step": 59730 }, { "grad_norm": 0.5659122467041016, "learning_rate": 3.8144527773832476e-05, "loss": 0.0414, "step": 59740 }, { "grad_norm": 0.4481152892112732, "learning_rate": 3.812846524510352e-05, "loss": 0.0301, "step": 59750 }, { "grad_norm": 0.46399909257888794, "learning_rate": 3.811240401462775e-05, "loss": 0.0455, "step": 59760 }, { "grad_norm": 0.49090033769607544, "learning_rate": 3.809634408416162e-05, "loss": 0.0388, "step": 59770 }, { "grad_norm": 0.37893885374069214, "learning_rate": 3.8080285455461394e-05, "loss": 0.0385, "step": 59780 }, { "grad_norm": 0.3648793399333954, "learning_rate": 3.806422813028323e-05, "loss": 0.0335, "step": 59790 }, { "grad_norm": 0.46742144227027893, "learning_rate": 3.8048172110383114e-05, "loss": 0.0344, "step": 59800 }, { "grad_norm": 0.2740267515182495, "learning_rate": 3.8032117397516944e-05, "loss": 0.0376, "step": 59810 }, { "grad_norm": 0.4570203125476837, "learning_rate": 3.801606399344042e-05, "loss": 0.0426, "step": 59820 }, { "grad_norm": 0.3363385498523712, "learning_rate": 3.800001189990913e-05, "loss": 0.04, "step": 59830 }, { "grad_norm": 0.4488678574562073, "learning_rate": 3.798396111867847e-05, "loss": 0.0403, "step": 59840 }, { "grad_norm": 0.6947852373123169, "learning_rate": 3.796791165150378e-05, "loss": 0.0393, "step": 59850 }, { "grad_norm": 0.6288233995437622, "learning_rate": 3.795186350014018e-05, "loss": 0.0395, "step": 59860 }, { "grad_norm": 0.32478898763656616, "learning_rate": 3.793581666634266e-05, "loss": 0.0353, "step": 59870 }, { "grad_norm": 0.7482414245605469, "learning_rate": 3.791977115186609e-05, "loss": 0.0431, "step": 59880 }, { "grad_norm": 0.44588014483451843, "learning_rate": 3.79037269584652e-05, "loss": 0.0323, "step": 59890 }, { "grad_norm": 0.49905624985694885, "learning_rate": 3.7887684087894545e-05, "loss": 0.0408, "step": 59900 }, { "grad_norm": 0.39088961482048035, "learning_rate": 3.787164254190854e-05, "loss": 0.0385, "step": 59910 }, { "grad_norm": 0.3509613275527954, "learning_rate": 3.7855602322261473e-05, "loss": 0.0371, "step": 59920 }, { "grad_norm": 0.32688990235328674, "learning_rate": 3.783956343070746e-05, "loss": 0.0494, "step": 59930 }, { "grad_norm": 0.9409580826759338, "learning_rate": 3.7823525869000526e-05, "loss": 0.0417, "step": 59940 }, { "grad_norm": 0.32428857684135437, "learning_rate": 3.78074896388945e-05, "loss": 0.0398, "step": 59950 }, { "grad_norm": 0.5260035991668701, "learning_rate": 3.7791454742143075e-05, "loss": 0.047, "step": 59960 }, { "grad_norm": 0.5362885594367981, "learning_rate": 3.7775421180499795e-05, "loss": 0.0401, "step": 59970 }, { "grad_norm": 0.26692116260528564, "learning_rate": 3.7759388955718086e-05, "loss": 0.0394, "step": 59980 }, { "grad_norm": 0.296377569437027, "learning_rate": 3.774335806955117e-05, "loss": 0.0384, "step": 59990 }, { "grad_norm": 0.39217349886894226, "learning_rate": 3.7727328523752215e-05, "loss": 0.0386, "step": 60000 }, { "grad_norm": 0.46248701214790344, "learning_rate": 3.771130032007415e-05, "loss": 0.0391, "step": 60010 }, { "grad_norm": 0.4522985517978668, "learning_rate": 3.769527346026982e-05, "loss": 0.0287, "step": 60020 }, { "grad_norm": 0.3896081745624542, "learning_rate": 3.767924794609188e-05, "loss": 0.0357, "step": 60030 }, { "grad_norm": 0.5004574656486511, "learning_rate": 3.7663223779292854e-05, "loss": 0.041, "step": 60040 }, { "grad_norm": 0.3287942111492157, "learning_rate": 3.764720096162512e-05, "loss": 0.0415, "step": 60050 }, { "grad_norm": 0.5397960543632507, "learning_rate": 3.763117949484092e-05, "loss": 0.038, "step": 60060 }, { "grad_norm": 0.7098709940910339, "learning_rate": 3.7615159380692344e-05, "loss": 0.0434, "step": 60070 }, { "grad_norm": 0.5144539475440979, "learning_rate": 3.759914062093132e-05, "loss": 0.0288, "step": 60080 }, { "grad_norm": 0.49948301911354065, "learning_rate": 3.7583123217309615e-05, "loss": 0.0367, "step": 60090 }, { "grad_norm": 0.34713292121887207, "learning_rate": 3.7567107171578904e-05, "loss": 0.0497, "step": 60100 }, { "grad_norm": 0.6989310383796692, "learning_rate": 3.755109248549066e-05, "loss": 0.0423, "step": 60110 }, { "grad_norm": 0.49116402864456177, "learning_rate": 3.75350791607962e-05, "loss": 0.0339, "step": 60120 }, { "grad_norm": 0.7639616131782532, "learning_rate": 3.751906719924676e-05, "loss": 0.0429, "step": 60130 }, { "grad_norm": 0.4130742847919464, "learning_rate": 3.750305660259337e-05, "loss": 0.0352, "step": 60140 }, { "grad_norm": 0.45101261138916016, "learning_rate": 3.748704737258693e-05, "loss": 0.0381, "step": 60150 }, { "grad_norm": 0.43494337797164917, "learning_rate": 3.747103951097816e-05, "loss": 0.0532, "step": 60160 }, { "grad_norm": 0.4735088348388672, "learning_rate": 3.745503301951768e-05, "loss": 0.0448, "step": 60170 }, { "grad_norm": 0.46550455689430237, "learning_rate": 3.743902789995592e-05, "loss": 0.0368, "step": 60180 }, { "grad_norm": 0.31907200813293457, "learning_rate": 3.7423024154043183e-05, "loss": 0.0309, "step": 60190 }, { "grad_norm": 0.5946813821792603, "learning_rate": 3.740702178352963e-05, "loss": 0.0332, "step": 60200 }, { "grad_norm": 0.7955459952354431, "learning_rate": 3.739102079016523e-05, "loss": 0.0439, "step": 60210 }, { "grad_norm": 0.5735348463058472, "learning_rate": 3.7375021175699846e-05, "loss": 0.0311, "step": 60220 }, { "grad_norm": 0.40983545780181885, "learning_rate": 3.735902294188317e-05, "loss": 0.0391, "step": 60230 }, { "grad_norm": 0.5142786502838135, "learning_rate": 3.734302609046473e-05, "loss": 0.0389, "step": 60240 }, { "grad_norm": 0.5192404389381409, "learning_rate": 3.732703062319391e-05, "loss": 0.044, "step": 60250 }, { "grad_norm": 2.4838454723358154, "learning_rate": 3.731103654181999e-05, "loss": 0.0448, "step": 60260 }, { "grad_norm": 0.6073894500732422, "learning_rate": 3.729504384809203e-05, "loss": 0.0377, "step": 60270 }, { "grad_norm": 0.40086594223976135, "learning_rate": 3.727905254375898e-05, "loss": 0.0337, "step": 60280 }, { "grad_norm": 0.3937610685825348, "learning_rate": 3.7263062630569614e-05, "loss": 0.0337, "step": 60290 }, { "grad_norm": 0.8403246998786926, "learning_rate": 3.724707411027256e-05, "loss": 0.0399, "step": 60300 }, { "grad_norm": 0.4202292859554291, "learning_rate": 3.723108698461631e-05, "loss": 0.0335, "step": 60310 }, { "grad_norm": 0.46840962767601013, "learning_rate": 3.72151012553492e-05, "loss": 0.0399, "step": 60320 }, { "grad_norm": 0.3916466236114502, "learning_rate": 3.719911692421939e-05, "loss": 0.0284, "step": 60330 }, { "grad_norm": 0.7560684680938721, "learning_rate": 3.71831339929749e-05, "loss": 0.0394, "step": 60340 }, { "grad_norm": 0.30702054500579834, "learning_rate": 3.716715246336362e-05, "loss": 0.0332, "step": 60350 }, { "grad_norm": 0.33705899119377136, "learning_rate": 3.715117233713324e-05, "loss": 0.03, "step": 60360 }, { "grad_norm": 0.32440197467803955, "learning_rate": 3.713519361603135e-05, "loss": 0.0379, "step": 60370 }, { "grad_norm": 0.3843730092048645, "learning_rate": 3.711921630180532e-05, "loss": 0.0455, "step": 60380 }, { "grad_norm": 0.3443371653556824, "learning_rate": 3.710324039620245e-05, "loss": 0.0349, "step": 60390 }, { "grad_norm": 0.43545714020729065, "learning_rate": 3.708726590096982e-05, "loss": 0.0395, "step": 60400 }, { "grad_norm": 0.3865772783756256, "learning_rate": 3.707129281785437e-05, "loss": 0.0363, "step": 60410 }, { "grad_norm": 0.617761492729187, "learning_rate": 3.705532114860291e-05, "loss": 0.0419, "step": 60420 }, { "grad_norm": 0.5018903017044067, "learning_rate": 3.703935089496207e-05, "loss": 0.0552, "step": 60430 }, { "grad_norm": 0.3379257619380951, "learning_rate": 3.702338205867829e-05, "loss": 0.0339, "step": 60440 }, { "grad_norm": 0.45155709981918335, "learning_rate": 3.7007414641497976e-05, "loss": 0.0492, "step": 60450 }, { "grad_norm": 0.5308862328529358, "learning_rate": 3.699144864516726e-05, "loss": 0.0335, "step": 60460 }, { "grad_norm": 0.5120449066162109, "learning_rate": 3.6975484071432153e-05, "loss": 0.0388, "step": 60470 }, { "grad_norm": 0.3989478051662445, "learning_rate": 3.695952092203854e-05, "loss": 0.0459, "step": 60480 }, { "grad_norm": 0.43292033672332764, "learning_rate": 3.69435591987321e-05, "loss": 0.045, "step": 60490 }, { "grad_norm": 0.5874362587928772, "learning_rate": 3.6927598903258374e-05, "loss": 0.0387, "step": 60500 }, { "grad_norm": 1.2374444007873535, "learning_rate": 3.691164003736281e-05, "loss": 0.0293, "step": 60510 }, { "grad_norm": 0.34710973501205444, "learning_rate": 3.689568260279061e-05, "loss": 0.0377, "step": 60520 }, { "grad_norm": 0.32473206520080566, "learning_rate": 3.687972660128686e-05, "loss": 0.0337, "step": 60530 }, { "grad_norm": 0.5765471458435059, "learning_rate": 3.686377203459648e-05, "loss": 0.0437, "step": 60540 }, { "grad_norm": 0.3368348479270935, "learning_rate": 3.6847818904464226e-05, "loss": 0.0383, "step": 60550 }, { "grad_norm": 0.45880112051963806, "learning_rate": 3.683186721263474e-05, "loss": 0.0344, "step": 60560 }, { "grad_norm": 0.5227305889129639, "learning_rate": 3.6815916960852435e-05, "loss": 0.0467, "step": 60570 }, { "grad_norm": 0.4367727041244507, "learning_rate": 3.679996815086165e-05, "loss": 0.0373, "step": 60580 }, { "grad_norm": 0.4015030264854431, "learning_rate": 3.678402078440649e-05, "loss": 0.0404, "step": 60590 }, { "grad_norm": 0.4050463140010834, "learning_rate": 3.676807486323096e-05, "loss": 0.0442, "step": 60600 }, { "grad_norm": 0.6855015158653259, "learning_rate": 3.6752130389078864e-05, "loss": 0.0381, "step": 60610 }, { "grad_norm": 0.53281170129776, "learning_rate": 3.673618736369388e-05, "loss": 0.03, "step": 60620 }, { "grad_norm": 0.41921746730804443, "learning_rate": 3.6720245788819474e-05, "loss": 0.0366, "step": 60630 }, { "grad_norm": 0.5236595273017883, "learning_rate": 3.6704305666199044e-05, "loss": 0.0463, "step": 60640 }, { "grad_norm": 0.484528124332428, "learning_rate": 3.668836699757576e-05, "loss": 0.0368, "step": 60650 }, { "grad_norm": 0.7029252648353577, "learning_rate": 3.667242978469265e-05, "loss": 0.0411, "step": 60660 }, { "grad_norm": 0.5895231366157532, "learning_rate": 3.6656494029292575e-05, "loss": 0.0465, "step": 60670 }, { "grad_norm": 0.5177581310272217, "learning_rate": 3.664055973311825e-05, "loss": 0.0388, "step": 60680 }, { "grad_norm": 0.5445086359977722, "learning_rate": 3.662462689791221e-05, "loss": 0.0311, "step": 60690 }, { "grad_norm": 0.32896608114242554, "learning_rate": 3.660869552541689e-05, "loss": 0.0318, "step": 60700 }, { "grad_norm": 0.5936279892921448, "learning_rate": 3.659276561737448e-05, "loss": 0.0383, "step": 60710 }, { "grad_norm": 0.430612176656723, "learning_rate": 3.657683717552708e-05, "loss": 0.0385, "step": 60720 }, { "grad_norm": 0.6473100781440735, "learning_rate": 3.6560910201616574e-05, "loss": 0.0291, "step": 60730 }, { "grad_norm": 0.25990554690361023, "learning_rate": 3.654498469738472e-05, "loss": 0.0404, "step": 60740 }, { "grad_norm": 0.5583845973014832, "learning_rate": 3.6529060664573126e-05, "loss": 0.0473, "step": 60750 }, { "grad_norm": 0.38767117261886597, "learning_rate": 3.6513138104923176e-05, "loss": 0.0408, "step": 60760 }, { "grad_norm": 0.45713865756988525, "learning_rate": 3.6497217020176176e-05, "loss": 0.0382, "step": 60770 }, { "grad_norm": 0.5163827538490295, "learning_rate": 3.648129741207323e-05, "loss": 0.0422, "step": 60780 }, { "grad_norm": 0.38627171516418457, "learning_rate": 3.646537928235527e-05, "loss": 0.0302, "step": 60790 }, { "grad_norm": 0.3609920144081116, "learning_rate": 3.6449462632763075e-05, "loss": 0.0384, "step": 60800 }, { "grad_norm": 0.4277161955833435, "learning_rate": 3.643354746503729e-05, "loss": 0.0324, "step": 60810 }, { "grad_norm": 0.6786975264549255, "learning_rate": 3.6417633780918345e-05, "loss": 0.0295, "step": 60820 }, { "grad_norm": 0.4414404332637787, "learning_rate": 3.640172158214655e-05, "loss": 0.0435, "step": 60830 }, { "grad_norm": 0.9213718771934509, "learning_rate": 3.6385810870462035e-05, "loss": 0.0318, "step": 60840 }, { "grad_norm": 0.7078632116317749, "learning_rate": 3.636990164760479e-05, "loss": 0.0341, "step": 60850 }, { "grad_norm": 0.5395662784576416, "learning_rate": 3.6353993915314596e-05, "loss": 0.0383, "step": 60860 }, { "grad_norm": 0.4795633852481842, "learning_rate": 3.6338087675331124e-05, "loss": 0.0312, "step": 60870 }, { "grad_norm": 0.408577561378479, "learning_rate": 3.632218292939383e-05, "loss": 0.0309, "step": 60880 }, { "grad_norm": 0.47211796045303345, "learning_rate": 3.6306279679242033e-05, "loss": 0.032, "step": 60890 }, { "grad_norm": 0.4592103660106659, "learning_rate": 3.629037792661492e-05, "loss": 0.0464, "step": 60900 }, { "grad_norm": 0.39272770285606384, "learning_rate": 3.627447767325147e-05, "loss": 0.04, "step": 60910 }, { "grad_norm": 0.5458256602287292, "learning_rate": 3.62585789208905e-05, "loss": 0.0358, "step": 60920 }, { "grad_norm": 0.4200141429901123, "learning_rate": 3.624268167127068e-05, "loss": 0.0588, "step": 60930 }, { "grad_norm": 0.40616974234580994, "learning_rate": 3.622678592613051e-05, "loss": 0.0378, "step": 60940 }, { "grad_norm": 0.8316538333892822, "learning_rate": 3.6210891687208304e-05, "loss": 0.0419, "step": 60950 }, { "grad_norm": 0.8622133731842041, "learning_rate": 3.619499895624228e-05, "loss": 0.0356, "step": 60960 }, { "grad_norm": 0.45934221148490906, "learning_rate": 3.61791077349704e-05, "loss": 0.041, "step": 60970 }, { "grad_norm": 0.4504179358482361, "learning_rate": 3.616321802513053e-05, "loss": 0.0374, "step": 60980 }, { "grad_norm": 0.6833789348602295, "learning_rate": 3.6147329828460325e-05, "loss": 0.0488, "step": 60990 }, { "grad_norm": 0.5699121356010437, "learning_rate": 3.6131443146697306e-05, "loss": 0.0384, "step": 61000 }, { "grad_norm": 0.40706923604011536, "learning_rate": 3.611555798157879e-05, "loss": 0.036, "step": 61010 }, { "grad_norm": 0.5342714190483093, "learning_rate": 3.6099674334842e-05, "loss": 0.0403, "step": 61020 }, { "grad_norm": 0.3690778911113739, "learning_rate": 3.6083792208223934e-05, "loss": 0.0467, "step": 61030 }, { "grad_norm": 0.37663549184799194, "learning_rate": 3.606791160346142e-05, "loss": 0.0431, "step": 61040 }, { "grad_norm": 0.28315967321395874, "learning_rate": 3.6052032522291134e-05, "loss": 0.0392, "step": 61050 }, { "grad_norm": 0.7150601148605347, "learning_rate": 3.603615496644962e-05, "loss": 0.0523, "step": 61060 }, { "grad_norm": 0.6600306034088135, "learning_rate": 3.60202789376732e-05, "loss": 0.0411, "step": 61070 }, { "grad_norm": 0.39958229660987854, "learning_rate": 3.6004404437698036e-05, "loss": 0.0409, "step": 61080 }, { "grad_norm": 0.7368254065513611, "learning_rate": 3.5988531468260184e-05, "loss": 0.0385, "step": 61090 }, { "grad_norm": 0.46725204586982727, "learning_rate": 3.597266003109546e-05, "loss": 0.0349, "step": 61100 }, { "grad_norm": 0.4175495207309723, "learning_rate": 3.5956790127939557e-05, "loss": 0.0386, "step": 61110 }, { "grad_norm": 0.38730883598327637, "learning_rate": 3.594092176052796e-05, "loss": 0.0375, "step": 61120 }, { "grad_norm": 0.34233975410461426, "learning_rate": 3.592505493059603e-05, "loss": 0.0409, "step": 61130 }, { "grad_norm": 0.5083874464035034, "learning_rate": 3.5909189639878905e-05, "loss": 0.0427, "step": 61140 }, { "grad_norm": 0.467553973197937, "learning_rate": 3.5893325890111626e-05, "loss": 0.0346, "step": 61150 }, { "grad_norm": 0.5405405759811401, "learning_rate": 3.5877463683029034e-05, "loss": 0.0503, "step": 61160 }, { "grad_norm": 0.25066402554512024, "learning_rate": 3.586160302036578e-05, "loss": 0.0393, "step": 61170 }, { "grad_norm": 0.30877524614334106, "learning_rate": 3.584574390385636e-05, "loss": 0.0336, "step": 61180 }, { "grad_norm": 0.3244086503982544, "learning_rate": 3.582988633523511e-05, "loss": 0.0449, "step": 61190 }, { "grad_norm": 0.3246506154537201, "learning_rate": 3.581403031623616e-05, "loss": 0.0402, "step": 61200 }, { "grad_norm": 0.5227992534637451, "learning_rate": 3.579817584859354e-05, "loss": 0.0398, "step": 61210 }, { "grad_norm": 0.4987413287162781, "learning_rate": 3.578232293404107e-05, "loss": 0.0369, "step": 61220 }, { "grad_norm": 0.5655258297920227, "learning_rate": 3.5766471574312385e-05, "loss": 0.0341, "step": 61230 }, { "grad_norm": 0.40145161747932434, "learning_rate": 3.5750621771140964e-05, "loss": 0.042, "step": 61240 }, { "grad_norm": 0.4642578661441803, "learning_rate": 3.5734773526260115e-05, "loss": 0.0335, "step": 61250 }, { "grad_norm": 0.49575942754745483, "learning_rate": 3.5718926841402996e-05, "loss": 0.0375, "step": 61260 }, { "grad_norm": 0.3673560321331024, "learning_rate": 3.5703081718302555e-05, "loss": 0.0434, "step": 61270 }, { "grad_norm": 0.46533262729644775, "learning_rate": 3.568723815869162e-05, "loss": 0.0385, "step": 61280 }, { "grad_norm": 0.28938981890678406, "learning_rate": 3.5671396164302777e-05, "loss": 0.0567, "step": 61290 }, { "grad_norm": 0.9681481122970581, "learning_rate": 3.565555573686853e-05, "loss": 0.0339, "step": 61300 }, { "grad_norm": 0.48570743203163147, "learning_rate": 3.563971687812113e-05, "loss": 0.0331, "step": 61310 }, { "grad_norm": 0.3268490135669708, "learning_rate": 3.5623879589792706e-05, "loss": 0.0361, "step": 61320 }, { "grad_norm": 0.23099444806575775, "learning_rate": 3.5608043873615185e-05, "loss": 0.0431, "step": 61330 }, { "grad_norm": 0.35344550013542175, "learning_rate": 3.559220973132035e-05, "loss": 0.033, "step": 61340 }, { "grad_norm": 0.410835862159729, "learning_rate": 3.557637716463981e-05, "loss": 0.0368, "step": 61350 }, { "grad_norm": 1.0181304216384888, "learning_rate": 3.556054617530497e-05, "loss": 0.0439, "step": 61360 }, { "grad_norm": 0.335877925157547, "learning_rate": 3.5544716765047106e-05, "loss": 0.0371, "step": 61370 }, { "grad_norm": 0.36569756269454956, "learning_rate": 3.5528888935597274e-05, "loss": 0.0371, "step": 61380 }, { "grad_norm": 0.40705597400665283, "learning_rate": 3.551306268868636e-05, "loss": 0.0352, "step": 61390 }, { "grad_norm": 0.31069639325141907, "learning_rate": 3.5497238026045166e-05, "loss": 0.0358, "step": 61400 }, { "grad_norm": 0.36518174409866333, "learning_rate": 3.548141494940421e-05, "loss": 0.0412, "step": 61410 }, { "grad_norm": 3.2618680000305176, "learning_rate": 3.546559346049388e-05, "loss": 0.0399, "step": 61420 }, { "grad_norm": 0.4189063310623169, "learning_rate": 3.544977356104441e-05, "loss": 0.0344, "step": 61430 }, { "grad_norm": 0.37442174553871155, "learning_rate": 3.543395525278582e-05, "loss": 0.0409, "step": 61440 }, { "grad_norm": 0.4506463408470154, "learning_rate": 3.5418138537447985e-05, "loss": 0.0409, "step": 61450 }, { "grad_norm": 0.2184102088212967, "learning_rate": 3.5402323416760574e-05, "loss": 0.0318, "step": 61460 }, { "grad_norm": 0.2470797449350357, "learning_rate": 3.538650989245316e-05, "loss": 0.0319, "step": 61470 }, { "grad_norm": 0.607216477394104, "learning_rate": 3.537069796625504e-05, "loss": 0.0321, "step": 61480 }, { "grad_norm": 0.5196051597595215, "learning_rate": 3.53548876398954e-05, "loss": 0.0408, "step": 61490 }, { "grad_norm": 0.5418910980224609, "learning_rate": 3.5339078915103214e-05, "loss": 0.0386, "step": 61500 }, { "grad_norm": 0.39755553007125854, "learning_rate": 3.532327179360733e-05, "loss": 0.0448, "step": 61510 }, { "grad_norm": 0.6406497359275818, "learning_rate": 3.530746627713636e-05, "loss": 0.0346, "step": 61520 }, { "grad_norm": 0.5046061873435974, "learning_rate": 3.5291662367418805e-05, "loss": 0.0372, "step": 61530 }, { "grad_norm": 0.47053632140159607, "learning_rate": 3.527586006618293e-05, "loss": 0.0554, "step": 61540 }, { "grad_norm": 0.9085835814476013, "learning_rate": 3.526005937515687e-05, "loss": 0.0508, "step": 61550 }, { "grad_norm": 0.4609076678752899, "learning_rate": 3.524426029606856e-05, "loss": 0.0526, "step": 61560 }, { "grad_norm": 0.35106414556503296, "learning_rate": 3.5228462830645746e-05, "loss": 0.0405, "step": 61570 }, { "grad_norm": 0.5268316268920898, "learning_rate": 3.5212666980616036e-05, "loss": 0.038, "step": 61580 }, { "grad_norm": 0.4600226879119873, "learning_rate": 3.519687274770681e-05, "loss": 0.0457, "step": 61590 }, { "grad_norm": 0.3424615263938904, "learning_rate": 3.518108013364535e-05, "loss": 0.0509, "step": 61600 }, { "grad_norm": 0.46881571412086487, "learning_rate": 3.516528914015868e-05, "loss": 0.0442, "step": 61610 }, { "grad_norm": 0.39271265268325806, "learning_rate": 3.5149499768973694e-05, "loss": 0.0533, "step": 61620 }, { "grad_norm": 0.26391083002090454, "learning_rate": 3.513371202181708e-05, "loss": 0.0383, "step": 61630 }, { "grad_norm": 0.6175339818000793, "learning_rate": 3.511792590041537e-05, "loss": 0.0423, "step": 61640 }, { "grad_norm": 0.6363688707351685, "learning_rate": 3.510214140649488e-05, "loss": 0.0451, "step": 61650 }, { "grad_norm": 0.3261979818344116, "learning_rate": 3.508635854178184e-05, "loss": 0.0446, "step": 61660 }, { "grad_norm": 0.9334842562675476, "learning_rate": 3.50705773080022e-05, "loss": 0.043, "step": 61670 }, { "grad_norm": 0.5367076396942139, "learning_rate": 3.505479770688178e-05, "loss": 0.0442, "step": 61680 }, { "grad_norm": 0.3925241529941559, "learning_rate": 3.503901974014622e-05, "loss": 0.0313, "step": 61690 }, { "grad_norm": 0.39198920130729675, "learning_rate": 3.502324340952095e-05, "loss": 0.0411, "step": 61700 }, { "grad_norm": 0.4813362956047058, "learning_rate": 3.500746871673126e-05, "loss": 0.0511, "step": 61710 }, { "grad_norm": 0.35553979873657227, "learning_rate": 3.499169566350227e-05, "loss": 0.0489, "step": 61720 }, { "grad_norm": 0.40196123719215393, "learning_rate": 3.4975924251558876e-05, "loss": 0.0347, "step": 61730 }, { "grad_norm": 0.2862522602081299, "learning_rate": 3.496015448262582e-05, "loss": 0.0367, "step": 61740 }, { "grad_norm": 0.5397033095359802, "learning_rate": 3.494438635842764e-05, "loss": 0.0391, "step": 61750 }, { "grad_norm": 0.2991580665111542, "learning_rate": 3.492861988068874e-05, "loss": 0.0382, "step": 61760 }, { "grad_norm": 0.30670166015625, "learning_rate": 3.491285505113332e-05, "loss": 0.0501, "step": 61770 }, { "grad_norm": 0.4393632113933563, "learning_rate": 3.4897091871485376e-05, "loss": 0.0513, "step": 61780 }, { "grad_norm": 0.4177812337875366, "learning_rate": 3.488133034346875e-05, "loss": 0.0379, "step": 61790 }, { "grad_norm": 0.768813967704773, "learning_rate": 3.486557046880713e-05, "loss": 0.0331, "step": 61800 }, { "grad_norm": 0.4585607051849365, "learning_rate": 3.484981224922395e-05, "loss": 0.0445, "step": 61810 }, { "grad_norm": 0.40999624133110046, "learning_rate": 3.4834055686442544e-05, "loss": 0.0264, "step": 61820 }, { "grad_norm": 0.6685409545898438, "learning_rate": 3.4818300782186e-05, "loss": 0.0397, "step": 61830 }, { "grad_norm": 0.4588403105735779, "learning_rate": 3.4802547538177236e-05, "loss": 0.0372, "step": 61840 }, { "grad_norm": 0.6748811602592468, "learning_rate": 3.478679595613905e-05, "loss": 0.0344, "step": 61850 }, { "grad_norm": 0.2977169156074524, "learning_rate": 3.4771046037793975e-05, "loss": 0.0336, "step": 61860 }, { "grad_norm": 0.32393679022789, "learning_rate": 3.4755297784864416e-05, "loss": 0.0327, "step": 61870 }, { "grad_norm": 0.5145464539527893, "learning_rate": 3.473955119907257e-05, "loss": 0.0362, "step": 61880 }, { "grad_norm": 0.48446887731552124, "learning_rate": 3.472380628214046e-05, "loss": 0.0385, "step": 61890 }, { "grad_norm": 0.37621554732322693, "learning_rate": 3.4708063035789906e-05, "loss": 0.0388, "step": 61900 }, { "grad_norm": 0.4094819128513336, "learning_rate": 3.4692321461742614e-05, "loss": 0.0357, "step": 61910 }, { "grad_norm": 0.4741421341896057, "learning_rate": 3.4676581561720023e-05, "loss": 0.044, "step": 61920 }, { "grad_norm": 0.6424590349197388, "learning_rate": 3.466084333744344e-05, "loss": 0.036, "step": 61930 }, { "grad_norm": 0.4107155501842499, "learning_rate": 3.4645106790633964e-05, "loss": 0.0448, "step": 61940 }, { "grad_norm": 0.7295455932617188, "learning_rate": 3.4629371923012526e-05, "loss": 0.0509, "step": 61950 }, { "grad_norm": 0.4528985023498535, "learning_rate": 3.4613638736299856e-05, "loss": 0.0412, "step": 61960 }, { "grad_norm": 0.26384711265563965, "learning_rate": 3.459790723221652e-05, "loss": 0.0246, "step": 61970 }, { "grad_norm": 0.7308763861656189, "learning_rate": 3.45821774124829e-05, "loss": 0.0447, "step": 61980 }, { "grad_norm": 0.4315704107284546, "learning_rate": 3.456644927881918e-05, "loss": 0.0387, "step": 61990 }, { "grad_norm": 0.37440717220306396, "learning_rate": 3.4550722832945354e-05, "loss": 0.0394, "step": 62000 }, { "grad_norm": 0.4434318542480469, "learning_rate": 3.453499807658126e-05, "loss": 0.0391, "step": 62010 }, { "grad_norm": 0.26172566413879395, "learning_rate": 3.451927501144653e-05, "loss": 0.0305, "step": 62020 }, { "grad_norm": 0.4874175190925598, "learning_rate": 3.450355363926059e-05, "loss": 0.0512, "step": 62030 }, { "grad_norm": 0.5329251289367676, "learning_rate": 3.448783396174273e-05, "loss": 0.0428, "step": 62040 }, { "grad_norm": 0.563797116279602, "learning_rate": 3.4472115980612034e-05, "loss": 0.0385, "step": 62050 }, { "grad_norm": 0.6415877938270569, "learning_rate": 3.4456399697587396e-05, "loss": 0.0413, "step": 62060 }, { "grad_norm": 0.5981557965278625, "learning_rate": 3.4440685114387506e-05, "loss": 0.0493, "step": 62070 }, { "grad_norm": 0.6411703824996948, "learning_rate": 3.442497223273091e-05, "loss": 0.0366, "step": 62080 }, { "grad_norm": 0.5957815647125244, "learning_rate": 3.4409261054335926e-05, "loss": 0.0475, "step": 62090 }, { "grad_norm": 0.5703765749931335, "learning_rate": 3.439355158092069e-05, "loss": 0.0495, "step": 62100 }, { "grad_norm": 0.4368351995944977, "learning_rate": 3.437784381420322e-05, "loss": 0.0448, "step": 62110 }, { "grad_norm": 0.2921172082424164, "learning_rate": 3.436213775590126e-05, "loss": 0.0325, "step": 62120 }, { "grad_norm": 0.40504759550094604, "learning_rate": 3.4346433407732405e-05, "loss": 0.0367, "step": 62130 }, { "grad_norm": 0.29614925384521484, "learning_rate": 3.433073077141405e-05, "loss": 0.024, "step": 62140 }, { "grad_norm": 0.7429966330528259, "learning_rate": 3.431502984866343e-05, "loss": 0.0424, "step": 62150 }, { "grad_norm": 0.43940263986587524, "learning_rate": 3.429933064119753e-05, "loss": 0.0573, "step": 62160 }, { "grad_norm": 0.8091412782669067, "learning_rate": 3.4283633150733255e-05, "loss": 0.0401, "step": 62170 }, { "grad_norm": 0.4715498685836792, "learning_rate": 3.4267937378987225e-05, "loss": 0.0446, "step": 62180 }, { "grad_norm": 0.4886559545993805, "learning_rate": 3.425224332767592e-05, "loss": 0.0389, "step": 62190 }, { "grad_norm": 0.6076890230178833, "learning_rate": 3.4236550998515595e-05, "loss": 0.0488, "step": 62200 }, { "grad_norm": 0.506870687007904, "learning_rate": 3.422086039322235e-05, "loss": 0.0366, "step": 62210 }, { "grad_norm": 0.4557082951068878, "learning_rate": 3.420517151351209e-05, "loss": 0.0414, "step": 62220 }, { "grad_norm": 0.35633400082588196, "learning_rate": 3.418948436110053e-05, "loss": 0.0391, "step": 62230 }, { "grad_norm": 0.3459213376045227, "learning_rate": 3.417379893770319e-05, "loss": 0.0384, "step": 62240 }, { "grad_norm": 0.5162327289581299, "learning_rate": 3.41581152450354e-05, "loss": 0.0326, "step": 62250 }, { "grad_norm": 0.2920830249786377, "learning_rate": 3.414243328481231e-05, "loss": 0.0434, "step": 62260 }, { "grad_norm": 0.529227077960968, "learning_rate": 3.412675305874888e-05, "loss": 0.0388, "step": 62270 }, { "grad_norm": 0.5568601489067078, "learning_rate": 3.4111074568559876e-05, "loss": 0.0348, "step": 62280 }, { "grad_norm": 0.6151740550994873, "learning_rate": 3.4095397815959846e-05, "loss": 0.0468, "step": 62290 }, { "grad_norm": 0.6533836722373962, "learning_rate": 3.407972280266322e-05, "loss": 0.0376, "step": 62300 }, { "grad_norm": 0.31509196758270264, "learning_rate": 3.4064049530384166e-05, "loss": 0.0381, "step": 62310 }, { "grad_norm": 0.6287221908569336, "learning_rate": 3.40483780008367e-05, "loss": 0.0391, "step": 62320 }, { "grad_norm": 0.21540473401546478, "learning_rate": 3.4032708215734635e-05, "loss": 0.0393, "step": 62330 }, { "grad_norm": 0.5714072585105896, "learning_rate": 3.4017040176791584e-05, "loss": 0.0375, "step": 62340 }, { "grad_norm": 0.5091317296028137, "learning_rate": 3.4001373885720976e-05, "loss": 0.0485, "step": 62350 }, { "grad_norm": 0.3485267460346222, "learning_rate": 3.398570934423608e-05, "loss": 0.038, "step": 62360 }, { "grad_norm": 0.4275312125682831, "learning_rate": 3.397004655404994e-05, "loss": 0.051, "step": 62370 }, { "grad_norm": 0.41337302327156067, "learning_rate": 3.395438551687541e-05, "loss": 0.0422, "step": 62380 }, { "grad_norm": 0.5021601319313049, "learning_rate": 3.393872623442515e-05, "loss": 0.0373, "step": 62390 }, { "grad_norm": 0.39490458369255066, "learning_rate": 3.392306870841164e-05, "loss": 0.0415, "step": 62400 }, { "grad_norm": 0.425979346036911, "learning_rate": 3.390741294054715e-05, "loss": 0.0354, "step": 62410 }, { "grad_norm": 0.764802098274231, "learning_rate": 3.38917589325438e-05, "loss": 0.0364, "step": 62420 }, { "grad_norm": 0.6708556413650513, "learning_rate": 3.387610668611349e-05, "loss": 0.048, "step": 62430 }, { "grad_norm": 0.7505524754524231, "learning_rate": 3.38604562029679e-05, "loss": 0.0472, "step": 62440 }, { "grad_norm": 0.30079472064971924, "learning_rate": 3.384480748481855e-05, "loss": 0.0408, "step": 62450 }, { "grad_norm": 0.3797996938228607, "learning_rate": 3.382916053337676e-05, "loss": 0.0386, "step": 62460 }, { "grad_norm": 0.3889804184436798, "learning_rate": 3.381351535035366e-05, "loss": 0.0495, "step": 62470 }, { "grad_norm": 0.5080941319465637, "learning_rate": 3.379787193746019e-05, "loss": 0.0444, "step": 62480 }, { "grad_norm": 0.35451024770736694, "learning_rate": 3.378223029640709e-05, "loss": 0.0371, "step": 62490 }, { "grad_norm": 0.41092824935913086, "learning_rate": 3.376659042890489e-05, "loss": 0.0438, "step": 62500 }, { "grad_norm": 0.3795204758644104, "learning_rate": 3.375095233666397e-05, "loss": 0.0422, "step": 62510 }, { "grad_norm": 0.38566312193870544, "learning_rate": 3.373531602139447e-05, "loss": 0.0366, "step": 62520 }, { "grad_norm": 0.47220227122306824, "learning_rate": 3.371968148480635e-05, "loss": 0.0416, "step": 62530 }, { "grad_norm": 0.7831631898880005, "learning_rate": 3.370404872860937e-05, "loss": 0.038, "step": 62540 }, { "grad_norm": 0.3224536180496216, "learning_rate": 3.3688417754513145e-05, "loss": 0.0374, "step": 62550 }, { "grad_norm": 0.3504081070423126, "learning_rate": 3.367278856422703e-05, "loss": 0.0364, "step": 62560 }, { "grad_norm": 0.9408565759658813, "learning_rate": 3.365716115946021e-05, "loss": 0.0518, "step": 62570 }, { "grad_norm": 0.5925040245056152, "learning_rate": 3.3641535541921666e-05, "loss": 0.0564, "step": 62580 }, { "grad_norm": 2.9175403118133545, "learning_rate": 3.36259117133202e-05, "loss": 0.0416, "step": 62590 }, { "grad_norm": 0.4812473654747009, "learning_rate": 3.36102896753644e-05, "loss": 0.0309, "step": 62600 }, { "grad_norm": 0.45804280042648315, "learning_rate": 3.35946694297627e-05, "loss": 0.0318, "step": 62610 }, { "grad_norm": 0.3222407400608063, "learning_rate": 3.357905097822328e-05, "loss": 0.0443, "step": 62620 }, { "grad_norm": 0.8594435453414917, "learning_rate": 3.356343432245416e-05, "loss": 0.0405, "step": 62630 }, { "grad_norm": 0.7475888729095459, "learning_rate": 3.354781946416315e-05, "loss": 0.0569, "step": 62640 }, { "grad_norm": 0.433752179145813, "learning_rate": 3.3532206405057864e-05, "loss": 0.0465, "step": 62650 }, { "grad_norm": 0.7322909235954285, "learning_rate": 3.351659514684573e-05, "loss": 0.0401, "step": 62660 }, { "grad_norm": 0.4891132116317749, "learning_rate": 3.350098569123395e-05, "loss": 0.0344, "step": 62670 }, { "grad_norm": 0.41276824474334717, "learning_rate": 3.348537803992958e-05, "loss": 0.0358, "step": 62680 }, { "grad_norm": 0.40956762433052063, "learning_rate": 3.346977219463945e-05, "loss": 0.0422, "step": 62690 }, { "grad_norm": 0.7112048268318176, "learning_rate": 3.345416815707017e-05, "loss": 0.0341, "step": 62700 }, { "grad_norm": 0.589191734790802, "learning_rate": 3.3438565928928185e-05, "loss": 0.0429, "step": 62710 }, { "grad_norm": 0.6365994811058044, "learning_rate": 3.3422965511919736e-05, "loss": 0.0372, "step": 62720 }, { "grad_norm": 0.38722914457321167, "learning_rate": 3.3407366907750846e-05, "loss": 0.0366, "step": 62730 }, { "grad_norm": 0.6451515555381775, "learning_rate": 3.339177011812737e-05, "loss": 0.0512, "step": 62740 }, { "grad_norm": 0.3884196877479553, "learning_rate": 3.3376175144754935e-05, "loss": 0.0333, "step": 62750 }, { "grad_norm": 0.49822646379470825, "learning_rate": 3.3360581989339004e-05, "loss": 0.0404, "step": 62760 }, { "grad_norm": 0.23401165008544922, "learning_rate": 3.3344990653584804e-05, "loss": 0.0369, "step": 62770 }, { "grad_norm": 0.42279767990112305, "learning_rate": 3.332940113919739e-05, "loss": 0.0451, "step": 62780 }, { "grad_norm": 0.41276320815086365, "learning_rate": 3.331381344788161e-05, "loss": 0.03, "step": 62790 }, { "grad_norm": 0.46774011850357056, "learning_rate": 3.3298227581342076e-05, "loss": 0.0391, "step": 62800 }, { "grad_norm": 0.6577802300453186, "learning_rate": 3.328264354128329e-05, "loss": 0.0484, "step": 62810 }, { "grad_norm": 0.43668243288993835, "learning_rate": 3.3267061329409465e-05, "loss": 0.045, "step": 62820 }, { "grad_norm": 0.4154280722141266, "learning_rate": 3.325148094742466e-05, "loss": 0.0376, "step": 62830 }, { "grad_norm": 0.678442656993866, "learning_rate": 3.323590239703271e-05, "loss": 0.0343, "step": 62840 }, { "grad_norm": 0.4297681450843811, "learning_rate": 3.322032567993728e-05, "loss": 0.0339, "step": 62850 }, { "grad_norm": 0.5361529588699341, "learning_rate": 3.3204750797841774e-05, "loss": 0.0382, "step": 62860 }, { "grad_norm": 0.34000471234321594, "learning_rate": 3.31891777524495e-05, "loss": 0.0388, "step": 62870 }, { "grad_norm": 0.4500409662723541, "learning_rate": 3.317360654546347e-05, "loss": 0.0364, "step": 62880 }, { "grad_norm": 0.35125812888145447, "learning_rate": 3.3158037178586534e-05, "loss": 0.0463, "step": 62890 }, { "grad_norm": 0.40760400891304016, "learning_rate": 3.3142469653521316e-05, "loss": 0.0342, "step": 62900 }, { "grad_norm": 0.9802320003509521, "learning_rate": 3.3126903971970284e-05, "loss": 0.0498, "step": 62910 }, { "grad_norm": 0.7891654372215271, "learning_rate": 3.311134013563564e-05, "loss": 0.0445, "step": 62920 }, { "grad_norm": 0.7441649436950684, "learning_rate": 3.309577814621947e-05, "loss": 0.0343, "step": 62930 }, { "grad_norm": 0.44725093245506287, "learning_rate": 3.308021800542359e-05, "loss": 0.035, "step": 62940 }, { "grad_norm": 0.4130576550960541, "learning_rate": 3.306465971494963e-05, "loss": 0.0399, "step": 62950 }, { "grad_norm": 0.731831967830658, "learning_rate": 3.304910327649901e-05, "loss": 0.0425, "step": 62960 }, { "grad_norm": 0.22340360283851624, "learning_rate": 3.303354869177297e-05, "loss": 0.0349, "step": 62970 }, { "grad_norm": 0.4229244291782379, "learning_rate": 3.301799596247255e-05, "loss": 0.0335, "step": 62980 }, { "grad_norm": 0.31090882420539856, "learning_rate": 3.300244509029854e-05, "loss": 0.0315, "step": 62990 }, { "grad_norm": 0.41307732462882996, "learning_rate": 3.2986896076951585e-05, "loss": 0.0346, "step": 63000 }, { "grad_norm": 0.410024493932724, "learning_rate": 3.297134892413211e-05, "loss": 0.0435, "step": 63010 }, { "grad_norm": 0.4950999915599823, "learning_rate": 3.295580363354031e-05, "loss": 0.044, "step": 63020 }, { "grad_norm": 0.5329328179359436, "learning_rate": 3.2940260206876194e-05, "loss": 0.0338, "step": 63030 }, { "grad_norm": 0.559548020362854, "learning_rate": 3.292471864583958e-05, "loss": 0.0416, "step": 63040 }, { "grad_norm": 0.29461243748664856, "learning_rate": 3.290917895213004e-05, "loss": 0.0371, "step": 63050 }, { "grad_norm": 0.4879845380783081, "learning_rate": 3.289364112744701e-05, "loss": 0.0422, "step": 63060 }, { "grad_norm": 0.28966715931892395, "learning_rate": 3.287810517348967e-05, "loss": 0.0405, "step": 63070 }, { "grad_norm": 0.4402751922607422, "learning_rate": 3.2862571091957014e-05, "loss": 0.0402, "step": 63080 }, { "grad_norm": 0.5071405172348022, "learning_rate": 3.284703888454781e-05, "loss": 0.0334, "step": 63090 }, { "grad_norm": 0.3948226869106293, "learning_rate": 3.2831508552960646e-05, "loss": 0.0282, "step": 63100 }, { "grad_norm": 0.472928524017334, "learning_rate": 3.281598009889387e-05, "loss": 0.0416, "step": 63110 }, { "grad_norm": 0.8377196192741394, "learning_rate": 3.28004535240457e-05, "loss": 0.0506, "step": 63120 }, { "grad_norm": 0.3055596351623535, "learning_rate": 3.278492883011407e-05, "loss": 0.0454, "step": 63130 }, { "grad_norm": 0.31316593289375305, "learning_rate": 3.276940601879675e-05, "loss": 0.0461, "step": 63140 }, { "grad_norm": 0.672028660774231, "learning_rate": 3.275388509179127e-05, "loss": 0.0429, "step": 63150 }, { "grad_norm": 0.514619767665863, "learning_rate": 3.2738366050794985e-05, "loss": 0.0513, "step": 63160 }, { "grad_norm": 0.3954029083251953, "learning_rate": 3.2722848897505034e-05, "loss": 0.0352, "step": 63170 }, { "grad_norm": 0.47300437092781067, "learning_rate": 3.2707333633618356e-05, "loss": 0.0346, "step": 63180 }, { "grad_norm": 0.5155367255210876, "learning_rate": 3.269182026083167e-05, "loss": 0.0358, "step": 63190 }, { "grad_norm": 0.4121062159538269, "learning_rate": 3.2676308780841516e-05, "loss": 0.0411, "step": 63200 }, { "grad_norm": 0.35218504071235657, "learning_rate": 3.266079919534416e-05, "loss": 0.0386, "step": 63210 }, { "grad_norm": 0.34834355115890503, "learning_rate": 3.264529150603575e-05, "loss": 0.0362, "step": 63220 }, { "grad_norm": 0.44151821732521057, "learning_rate": 3.262978571461217e-05, "loss": 0.0396, "step": 63230 }, { "grad_norm": 0.46659132838249207, "learning_rate": 3.26142818227691e-05, "loss": 0.0499, "step": 63240 }, { "grad_norm": 0.593883752822876, "learning_rate": 3.259877983220203e-05, "loss": 0.0442, "step": 63250 }, { "grad_norm": 0.48710229992866516, "learning_rate": 3.2583279744606246e-05, "loss": 0.0409, "step": 63260 }, { "grad_norm": 0.4890286922454834, "learning_rate": 3.25677815616768e-05, "loss": 0.0253, "step": 63270 }, { "grad_norm": 1.2218492031097412, "learning_rate": 3.255228528510856e-05, "loss": 0.0494, "step": 63280 }, { "grad_norm": 0.40817707777023315, "learning_rate": 3.253679091659617e-05, "loss": 0.0325, "step": 63290 }, { "grad_norm": 0.6426534056663513, "learning_rate": 3.252129845783404e-05, "loss": 0.0402, "step": 63300 }, { "grad_norm": 0.29746946692466736, "learning_rate": 3.2505807910516456e-05, "loss": 0.0354, "step": 63310 }, { "grad_norm": 0.6492722034454346, "learning_rate": 3.249031927633742e-05, "loss": 0.0461, "step": 63320 }, { "grad_norm": 0.3228980004787445, "learning_rate": 3.2474832556990737e-05, "loss": 0.031, "step": 63330 }, { "grad_norm": 0.41123631596565247, "learning_rate": 3.245934775417002e-05, "loss": 0.0348, "step": 63340 }, { "grad_norm": 0.711793839931488, "learning_rate": 3.244386486956866e-05, "loss": 0.039, "step": 63350 }, { "grad_norm": 0.5608478784561157, "learning_rate": 3.2428383904879846e-05, "loss": 0.0442, "step": 63360 }, { "grad_norm": 1.2509697675704956, "learning_rate": 3.241290486179652e-05, "loss": 0.053, "step": 63370 }, { "grad_norm": 0.37206193804740906, "learning_rate": 3.239742774201151e-05, "loss": 0.0362, "step": 63380 }, { "grad_norm": 0.37298324704170227, "learning_rate": 3.238195254721733e-05, "loss": 0.0337, "step": 63390 }, { "grad_norm": 0.4371233880519867, "learning_rate": 3.236647927910633e-05, "loss": 0.0439, "step": 63400 }, { "grad_norm": 0.22479422390460968, "learning_rate": 3.2351007939370646e-05, "loss": 0.0321, "step": 63410 }, { "grad_norm": 0.6019447445869446, "learning_rate": 3.2335538529702184e-05, "loss": 0.0385, "step": 63420 }, { "grad_norm": 0.4993065893650055, "learning_rate": 3.232007105179268e-05, "loss": 0.0387, "step": 63430 }, { "grad_norm": 0.5541757345199585, "learning_rate": 3.230460550733363e-05, "loss": 0.0292, "step": 63440 }, { "grad_norm": 0.26866623759269714, "learning_rate": 3.2289141898016315e-05, "loss": 0.0337, "step": 63450 }, { "grad_norm": 0.4763132929801941, "learning_rate": 3.227368022553181e-05, "loss": 0.0379, "step": 63460 }, { "grad_norm": 0.49859368801116943, "learning_rate": 3.225822049157099e-05, "loss": 0.0333, "step": 63470 }, { "grad_norm": 0.5426397919654846, "learning_rate": 3.2242762697824494e-05, "loss": 0.0315, "step": 63480 }, { "grad_norm": 0.9404671788215637, "learning_rate": 3.222730684598279e-05, "loss": 0.0435, "step": 63490 }, { "grad_norm": 0.3907446265220642, "learning_rate": 3.221185293773605e-05, "loss": 0.0302, "step": 63500 }, { "grad_norm": 0.3359763026237488, "learning_rate": 3.2196400974774354e-05, "loss": 0.0293, "step": 63510 }, { "grad_norm": 0.3893328905105591, "learning_rate": 3.2180950958787485e-05, "loss": 0.0382, "step": 63520 }, { "grad_norm": 0.2434712052345276, "learning_rate": 3.2165502891465024e-05, "loss": 0.0256, "step": 63530 }, { "grad_norm": 0.6345919370651245, "learning_rate": 3.215005677449636e-05, "loss": 0.0387, "step": 63540 }, { "grad_norm": 0.6572302579879761, "learning_rate": 3.2134612609570644e-05, "loss": 0.0304, "step": 63550 }, { "grad_norm": 0.3383174538612366, "learning_rate": 3.211917039837682e-05, "loss": 0.0359, "step": 63560 }, { "grad_norm": 0.41471758484840393, "learning_rate": 3.210373014260365e-05, "loss": 0.051, "step": 63570 }, { "grad_norm": 0.4012701213359833, "learning_rate": 3.208829184393965e-05, "loss": 0.0354, "step": 63580 }, { "grad_norm": 0.587740421295166, "learning_rate": 3.207285550407312e-05, "loss": 0.0477, "step": 63590 }, { "grad_norm": 0.3317582607269287, "learning_rate": 3.205742112469217e-05, "loss": 0.0359, "step": 63600 }, { "grad_norm": 0.443952739238739, "learning_rate": 3.204198870748465e-05, "loss": 0.0424, "step": 63610 }, { "grad_norm": 0.5573933720588684, "learning_rate": 3.202655825413824e-05, "loss": 0.0366, "step": 63620 }, { "grad_norm": 0.31460294127464294, "learning_rate": 3.2011129766340416e-05, "loss": 0.0534, "step": 63630 }, { "grad_norm": 0.37722402811050415, "learning_rate": 3.1995703245778384e-05, "loss": 0.0301, "step": 63640 }, { "grad_norm": 0.29581013321876526, "learning_rate": 3.1980278694139186e-05, "loss": 0.0317, "step": 63650 }, { "grad_norm": 0.484496533870697, "learning_rate": 3.196485611310962e-05, "loss": 0.0337, "step": 63660 }, { "grad_norm": 0.5662159323692322, "learning_rate": 3.1949435504376257e-05, "loss": 0.0435, "step": 63670 }, { "grad_norm": 0.8926900029182434, "learning_rate": 3.1934016869625505e-05, "loss": 0.0323, "step": 63680 }, { "grad_norm": 0.566296398639679, "learning_rate": 3.191860021054349e-05, "loss": 0.0519, "step": 63690 }, { "grad_norm": 0.5139876008033752, "learning_rate": 3.190318552881618e-05, "loss": 0.0466, "step": 63700 }, { "grad_norm": 0.8960133194923401, "learning_rate": 3.1887772826129285e-05, "loss": 0.0368, "step": 63710 }, { "grad_norm": 0.30879536271095276, "learning_rate": 3.1872362104168335e-05, "loss": 0.0413, "step": 63720 }, { "grad_norm": 0.416795939207077, "learning_rate": 3.185695336461861e-05, "loss": 0.0398, "step": 63730 }, { "grad_norm": 0.5052143335342407, "learning_rate": 3.184154660916519e-05, "loss": 0.0304, "step": 63740 }, { "grad_norm": 0.4227924346923828, "learning_rate": 3.1826141839492906e-05, "loss": 0.0345, "step": 63750 }, { "grad_norm": 0.574547529220581, "learning_rate": 3.181073905728644e-05, "loss": 0.0331, "step": 63760 }, { "grad_norm": 0.4718492329120636, "learning_rate": 3.1795338264230214e-05, "loss": 0.026, "step": 63770 }, { "grad_norm": 0.505924642086029, "learning_rate": 3.1779939462008426e-05, "loss": 0.0433, "step": 63780 }, { "grad_norm": 1.3265491724014282, "learning_rate": 3.1764542652305066e-05, "loss": 0.0342, "step": 63790 }, { "grad_norm": 0.4125988781452179, "learning_rate": 3.17491478368039e-05, "loss": 0.0435, "step": 63800 }, { "grad_norm": 0.7193500995635986, "learning_rate": 3.173375501718847e-05, "loss": 0.0372, "step": 63810 }, { "grad_norm": 0.33514463901519775, "learning_rate": 3.171836419514215e-05, "loss": 0.0305, "step": 63820 }, { "grad_norm": 0.588046133518219, "learning_rate": 3.170297537234803e-05, "loss": 0.0526, "step": 63830 }, { "grad_norm": 0.41963574290275574, "learning_rate": 3.168758855048901e-05, "loss": 0.0288, "step": 63840 }, { "grad_norm": 0.4794602692127228, "learning_rate": 3.1672203731247785e-05, "loss": 0.036, "step": 63850 }, { "grad_norm": 0.4103451371192932, "learning_rate": 3.16568209163068e-05, "loss": 0.0442, "step": 63860 }, { "grad_norm": 0.38943585753440857, "learning_rate": 3.1641440107348295e-05, "loss": 0.0374, "step": 63870 }, { "grad_norm": 0.45701852440834045, "learning_rate": 3.162606130605429e-05, "loss": 0.0344, "step": 63880 }, { "grad_norm": 0.36999669671058655, "learning_rate": 3.16106845141066e-05, "loss": 0.0341, "step": 63890 }, { "grad_norm": 0.4388168156147003, "learning_rate": 3.159530973318682e-05, "loss": 0.0376, "step": 63900 }, { "grad_norm": 0.42208409309387207, "learning_rate": 3.1579936964976286e-05, "loss": 0.0369, "step": 63910 }, { "grad_norm": 0.34481140971183777, "learning_rate": 3.156456621115615e-05, "loss": 0.0467, "step": 63920 }, { "grad_norm": 0.4860634207725525, "learning_rate": 3.154919747340734e-05, "loss": 0.0324, "step": 63930 }, { "grad_norm": 0.5979064702987671, "learning_rate": 3.153383075341054e-05, "loss": 0.0382, "step": 63940 }, { "grad_norm": 0.397009938955307, "learning_rate": 3.1518466052846264e-05, "loss": 0.0316, "step": 63950 }, { "grad_norm": 0.5267108082771301, "learning_rate": 3.150310337339474e-05, "loss": 0.037, "step": 63960 }, { "grad_norm": 0.7942047715187073, "learning_rate": 3.148774271673603e-05, "loss": 0.0333, "step": 63970 }, { "grad_norm": 0.4206879436969757, "learning_rate": 3.147238408454995e-05, "loss": 0.0469, "step": 63980 }, { "grad_norm": 0.3001401722431183, "learning_rate": 3.1457027478516086e-05, "loss": 0.0372, "step": 63990 }, { "grad_norm": 0.45193570852279663, "learning_rate": 3.144167290031382e-05, "loss": 0.0368, "step": 64000 }, { "grad_norm": 0.7590293288230896, "learning_rate": 3.142632035162227e-05, "loss": 0.0441, "step": 64010 }, { "grad_norm": 0.40013450384140015, "learning_rate": 3.141096983412044e-05, "loss": 0.0281, "step": 64020 }, { "grad_norm": 0.3877811133861542, "learning_rate": 3.139562134948699e-05, "loss": 0.0358, "step": 64030 }, { "grad_norm": 1.0957529544830322, "learning_rate": 3.1380274899400424e-05, "loss": 0.0378, "step": 64040 }, { "grad_norm": 0.36947423219680786, "learning_rate": 3.1364930485539e-05, "loss": 0.0338, "step": 64050 }, { "grad_norm": 0.3998362421989441, "learning_rate": 3.134958810958076e-05, "loss": 0.0318, "step": 64060 }, { "grad_norm": 0.5675708055496216, "learning_rate": 3.13342477732035e-05, "loss": 0.0399, "step": 64070 }, { "grad_norm": 0.3716046214103699, "learning_rate": 3.131890947808487e-05, "loss": 0.0299, "step": 64080 }, { "grad_norm": 0.4169718623161316, "learning_rate": 3.13035732259022e-05, "loss": 0.0439, "step": 64090 }, { "grad_norm": 0.387136310338974, "learning_rate": 3.1288239018332656e-05, "loss": 0.0316, "step": 64100 }, { "grad_norm": 0.6462476849555969, "learning_rate": 3.1272906857053164e-05, "loss": 0.0339, "step": 64110 }, { "grad_norm": 0.6353253722190857, "learning_rate": 3.125757674374042e-05, "loss": 0.0361, "step": 64120 }, { "grad_norm": 0.34124091267585754, "learning_rate": 3.124224868007088e-05, "loss": 0.0406, "step": 64130 }, { "grad_norm": 0.39005500078201294, "learning_rate": 3.1226922667720854e-05, "loss": 0.0341, "step": 64140 }, { "grad_norm": 0.6306522488594055, "learning_rate": 3.121159870836633e-05, "loss": 0.0432, "step": 64150 }, { "grad_norm": 0.4558095932006836, "learning_rate": 3.119627680368312e-05, "loss": 0.0369, "step": 64160 }, { "grad_norm": 0.2958529591560364, "learning_rate": 3.1180956955346816e-05, "loss": 0.0372, "step": 64170 }, { "grad_norm": 0.3860623240470886, "learning_rate": 3.116563916503276e-05, "loss": 0.0379, "step": 64180 }, { "grad_norm": 0.3519398868083954, "learning_rate": 3.1150323434416095e-05, "loss": 0.0397, "step": 64190 }, { "grad_norm": 0.336589515209198, "learning_rate": 3.1135009765171696e-05, "loss": 0.0351, "step": 64200 }, { "grad_norm": 0.3808240294456482, "learning_rate": 3.111969815897429e-05, "loss": 0.041, "step": 64210 }, { "grad_norm": 0.34181085228919983, "learning_rate": 3.1104388617498295e-05, "loss": 0.0356, "step": 64220 }, { "grad_norm": 0.34777727723121643, "learning_rate": 3.1089081142417966e-05, "loss": 0.0412, "step": 64230 }, { "grad_norm": 0.33937564492225647, "learning_rate": 3.1073775735407285e-05, "loss": 0.0352, "step": 64240 }, { "grad_norm": 0.6391646862030029, "learning_rate": 3.1058472398140035e-05, "loss": 0.0427, "step": 64250 }, { "grad_norm": 0.436799556016922, "learning_rate": 3.1043171132289736e-05, "loss": 0.0363, "step": 64260 }, { "grad_norm": 0.3176327645778656, "learning_rate": 3.1027871939529774e-05, "loss": 0.0271, "step": 64270 }, { "grad_norm": 0.5754393935203552, "learning_rate": 3.1012574821533214e-05, "loss": 0.0439, "step": 64280 }, { "grad_norm": 0.8458645939826965, "learning_rate": 3.099727977997291e-05, "loss": 0.036, "step": 64290 }, { "grad_norm": 0.1917036473751068, "learning_rate": 3.098198681652154e-05, "loss": 0.0508, "step": 64300 }, { "grad_norm": 0.40528666973114014, "learning_rate": 3.0966695932851475e-05, "loss": 0.0285, "step": 64310 }, { "grad_norm": 0.612489640712738, "learning_rate": 3.0951407130634926e-05, "loss": 0.0376, "step": 64320 }, { "grad_norm": 0.3472277820110321, "learning_rate": 3.0936120411543864e-05, "loss": 0.0338, "step": 64330 }, { "grad_norm": 0.5681430101394653, "learning_rate": 3.0920835777250015e-05, "loss": 0.0404, "step": 64340 }, { "grad_norm": 0.5753549933433533, "learning_rate": 3.090555322942488e-05, "loss": 0.0453, "step": 64350 }, { "grad_norm": 0.47514957189559937, "learning_rate": 3.089027276973973e-05, "loss": 0.0356, "step": 64360 }, { "grad_norm": 0.3392306864261627, "learning_rate": 3.0874994399865615e-05, "loss": 0.0316, "step": 64370 }, { "grad_norm": 0.4385487735271454, "learning_rate": 3.085971812147337e-05, "loss": 0.0329, "step": 64380 }, { "grad_norm": 0.39808568358421326, "learning_rate": 3.0844443936233557e-05, "loss": 0.0376, "step": 64390 }, { "grad_norm": 0.4341810941696167, "learning_rate": 3.082917184581657e-05, "loss": 0.0408, "step": 64400 }, { "grad_norm": 0.38220706582069397, "learning_rate": 3.0813901851892513e-05, "loss": 0.0364, "step": 64410 }, { "grad_norm": 1.2237507104873657, "learning_rate": 3.0798633956131316e-05, "loss": 0.0446, "step": 64420 }, { "grad_norm": 0.1823601871728897, "learning_rate": 3.078336816020264e-05, "loss": 0.03, "step": 64430 }, { "grad_norm": 0.5923796892166138, "learning_rate": 3.0768104465775935e-05, "loss": 0.0411, "step": 64440 }, { "grad_norm": 0.7111577987670898, "learning_rate": 3.0752842874520396e-05, "loss": 0.036, "step": 64450 }, { "grad_norm": 0.5008810758590698, "learning_rate": 3.073758338810503e-05, "loss": 0.0351, "step": 64460 }, { "grad_norm": 0.8138347864151001, "learning_rate": 3.072232600819859e-05, "loss": 0.0393, "step": 64470 }, { "grad_norm": 0.4374401867389679, "learning_rate": 3.070707073646959e-05, "loss": 0.0408, "step": 64480 }, { "grad_norm": 0.3702978789806366, "learning_rate": 3.069181757458633e-05, "loss": 0.0406, "step": 64490 }, { "grad_norm": 0.34884563088417053, "learning_rate": 3.067656652421687e-05, "loss": 0.0327, "step": 64500 }, { "grad_norm": 0.29727858304977417, "learning_rate": 3.0661317587029025e-05, "loss": 0.0396, "step": 64510 }, { "grad_norm": 0.6915820837020874, "learning_rate": 3.064607076469043e-05, "loss": 0.0301, "step": 64520 }, { "grad_norm": 0.40122538805007935, "learning_rate": 3.063082605886845e-05, "loss": 0.0405, "step": 64530 }, { "grad_norm": 0.41640836000442505, "learning_rate": 3.061558347123019e-05, "loss": 0.0367, "step": 64540 }, { "grad_norm": 0.3131866157054901, "learning_rate": 3.060034300344259e-05, "loss": 0.0458, "step": 64550 }, { "grad_norm": 0.3857223093509674, "learning_rate": 3.058510465717231e-05, "loss": 0.0433, "step": 64560 }, { "grad_norm": 0.3262036144733429, "learning_rate": 3.0569868434085804e-05, "loss": 0.0275, "step": 64570 }, { "grad_norm": 0.9376828074455261, "learning_rate": 3.055463433584924e-05, "loss": 0.0275, "step": 64580 }, { "grad_norm": 0.5588920712471008, "learning_rate": 3.0539402364128656e-05, "loss": 0.0309, "step": 64590 }, { "grad_norm": 0.7093711495399475, "learning_rate": 3.052417252058977e-05, "loss": 0.0421, "step": 64600 }, { "grad_norm": 0.4794381856918335, "learning_rate": 3.0508944806898092e-05, "loss": 0.0439, "step": 64610 }, { "grad_norm": 0.3694056272506714, "learning_rate": 3.04937192247189e-05, "loss": 0.0446, "step": 64620 }, { "grad_norm": 1.2853796482086182, "learning_rate": 3.047849577571725e-05, "loss": 0.0453, "step": 64630 }, { "grad_norm": 0.42195582389831543, "learning_rate": 3.046327446155795e-05, "loss": 0.0334, "step": 64640 }, { "grad_norm": 0.3556697964668274, "learning_rate": 3.0448055283905587e-05, "loss": 0.0298, "step": 64650 }, { "grad_norm": 0.3610936403274536, "learning_rate": 3.0432838244424494e-05, "loss": 0.0311, "step": 64660 }, { "grad_norm": 0.39042115211486816, "learning_rate": 3.04176233447788e-05, "loss": 0.0243, "step": 64670 }, { "grad_norm": 0.2797546982765198, "learning_rate": 3.040241058663238e-05, "loss": 0.0299, "step": 64680 }, { "grad_norm": 0.37077009677886963, "learning_rate": 3.038719997164887e-05, "loss": 0.045, "step": 64690 }, { "grad_norm": 0.3036039471626282, "learning_rate": 3.0371991501491683e-05, "loss": 0.0389, "step": 64700 }, { "grad_norm": 0.34732216596603394, "learning_rate": 3.0356785177823975e-05, "loss": 0.0332, "step": 64710 }, { "grad_norm": 0.3013753592967987, "learning_rate": 3.0341581002308726e-05, "loss": 0.0308, "step": 64720 }, { "grad_norm": 0.4917478859424591, "learning_rate": 3.0326378976608627e-05, "loss": 0.0336, "step": 64730 }, { "grad_norm": 0.2812492847442627, "learning_rate": 3.0311179102386146e-05, "loss": 0.0463, "step": 64740 }, { "grad_norm": 0.6077269315719604, "learning_rate": 3.0295981381303513e-05, "loss": 0.0397, "step": 64750 }, { "grad_norm": 0.4704364538192749, "learning_rate": 3.0280785815022727e-05, "loss": 0.0333, "step": 64760 }, { "grad_norm": 0.5063214302062988, "learning_rate": 3.0265592405205544e-05, "loss": 0.0367, "step": 64770 }, { "grad_norm": 0.565886914730072, "learning_rate": 3.0250401153513517e-05, "loss": 0.0335, "step": 64780 }, { "grad_norm": 0.8796241283416748, "learning_rate": 3.0235212061607936e-05, "loss": 0.0502, "step": 64790 }, { "grad_norm": 0.263078510761261, "learning_rate": 3.0220025131149842e-05, "loss": 0.0327, "step": 64800 }, { "grad_norm": 0.320352166891098, "learning_rate": 3.020484036380007e-05, "loss": 0.0302, "step": 64810 }, { "grad_norm": 0.39173611998558044, "learning_rate": 3.0189657761219188e-05, "loss": 0.0473, "step": 64820 }, { "grad_norm": 0.3110910654067993, "learning_rate": 3.017447732506753e-05, "loss": 0.0411, "step": 64830 }, { "grad_norm": 0.3416835069656372, "learning_rate": 3.015929905700524e-05, "loss": 0.0296, "step": 64840 }, { "grad_norm": 0.33232152462005615, "learning_rate": 3.014412295869218e-05, "loss": 0.0345, "step": 64850 }, { "grad_norm": 0.5721332430839539, "learning_rate": 3.0128949031787977e-05, "loss": 0.0413, "step": 64860 }, { "grad_norm": 0.29635581374168396, "learning_rate": 3.0113777277952022e-05, "loss": 0.0434, "step": 64870 }, { "grad_norm": 0.5646743774414062, "learning_rate": 3.0098607698843496e-05, "loss": 0.0346, "step": 64880 }, { "grad_norm": 0.7078467011451721, "learning_rate": 3.0083440296121312e-05, "loss": 0.0364, "step": 64890 }, { "grad_norm": 0.40477728843688965, "learning_rate": 3.0068275071444142e-05, "loss": 0.0322, "step": 64900 }, { "grad_norm": 0.4463489353656769, "learning_rate": 3.0053112026470442e-05, "loss": 0.0376, "step": 64910 }, { "grad_norm": 0.41845598816871643, "learning_rate": 3.0037951162858434e-05, "loss": 0.0333, "step": 64920 }, { "grad_norm": 0.355415940284729, "learning_rate": 3.002279248226607e-05, "loss": 0.0336, "step": 64930 }, { "grad_norm": 0.37071385979652405, "learning_rate": 3.000763598635109e-05, "loss": 0.0363, "step": 64940 }, { "grad_norm": 0.6606887578964233, "learning_rate": 2.999248167677098e-05, "loss": 0.0486, "step": 64950 }, { "grad_norm": 0.43322089314460754, "learning_rate": 2.9977329555182976e-05, "loss": 0.0445, "step": 64960 }, { "grad_norm": 0.48035213351249695, "learning_rate": 2.9962179623244124e-05, "loss": 0.0395, "step": 64970 }, { "grad_norm": 0.5284363031387329, "learning_rate": 2.9947031882611175e-05, "loss": 0.0335, "step": 64980 }, { "grad_norm": 0.4835333824157715, "learning_rate": 2.9931886334940685e-05, "loss": 0.0354, "step": 64990 }, { "grad_norm": 0.30802690982818604, "learning_rate": 2.9916742981888923e-05, "loss": 0.0355, "step": 65000 }, { "grad_norm": 0.46421581506729126, "learning_rate": 2.990160182511196e-05, "loss": 0.0445, "step": 65010 }, { "grad_norm": 0.9807371497154236, "learning_rate": 2.9886462866265574e-05, "loss": 0.0452, "step": 65020 }, { "grad_norm": 0.3466736078262329, "learning_rate": 2.9871326107005392e-05, "loss": 0.0283, "step": 65030 }, { "grad_norm": 0.4910122752189636, "learning_rate": 2.9856191548986735e-05, "loss": 0.0494, "step": 65040 }, { "grad_norm": 0.29562824964523315, "learning_rate": 2.9841059193864672e-05, "loss": 0.0389, "step": 65050 }, { "grad_norm": 0.3177710771560669, "learning_rate": 2.982592904329407e-05, "loss": 0.0302, "step": 65060 }, { "grad_norm": 0.20345519483089447, "learning_rate": 2.981080109892954e-05, "loss": 0.0262, "step": 65070 }, { "grad_norm": 0.6145067811012268, "learning_rate": 2.9795675362425434e-05, "loss": 0.0368, "step": 65080 }, { "grad_norm": 0.34526684880256653, "learning_rate": 2.97805518354359e-05, "loss": 0.0318, "step": 65090 }, { "grad_norm": 0.5459616780281067, "learning_rate": 2.9765430519614825e-05, "loss": 0.0307, "step": 65100 }, { "grad_norm": 0.5999563932418823, "learning_rate": 2.975031141661584e-05, "loss": 0.0329, "step": 65110 }, { "grad_norm": 0.3684505224227905, "learning_rate": 2.9735194528092347e-05, "loss": 0.0342, "step": 65120 }, { "grad_norm": 0.389225035905838, "learning_rate": 2.9720079855697525e-05, "loss": 0.0529, "step": 65130 }, { "grad_norm": 1.2983516454696655, "learning_rate": 2.970496740108427e-05, "loss": 0.0507, "step": 65140 }, { "grad_norm": 0.4636008143424988, "learning_rate": 2.968985716590527e-05, "loss": 0.0336, "step": 65150 }, { "grad_norm": 0.3235859274864197, "learning_rate": 2.9674749151812953e-05, "loss": 0.0359, "step": 65160 }, { "grad_norm": 0.9600730538368225, "learning_rate": 2.965964336045951e-05, "loss": 0.0335, "step": 65170 }, { "grad_norm": 0.46008267998695374, "learning_rate": 2.9644539793496894e-05, "loss": 0.0317, "step": 65180 }, { "grad_norm": 0.5797571539878845, "learning_rate": 2.9629438452576807e-05, "loss": 0.0343, "step": 65190 }, { "grad_norm": 0.46600374579429626, "learning_rate": 2.9614339339350698e-05, "loss": 0.0312, "step": 65200 }, { "grad_norm": 0.767505407333374, "learning_rate": 2.959924245546979e-05, "loss": 0.0329, "step": 65210 }, { "grad_norm": 0.43160656094551086, "learning_rate": 2.9584147802585038e-05, "loss": 0.0492, "step": 65220 }, { "grad_norm": 0.392126202583313, "learning_rate": 2.956905538234721e-05, "loss": 0.0374, "step": 65230 }, { "grad_norm": 0.5248947143554688, "learning_rate": 2.9553965196406764e-05, "loss": 0.0382, "step": 65240 }, { "grad_norm": 0.3826681077480316, "learning_rate": 2.9538877246413943e-05, "loss": 0.0287, "step": 65250 }, { "grad_norm": 0.4254073202610016, "learning_rate": 2.952379153401874e-05, "loss": 0.0367, "step": 65260 }, { "grad_norm": 0.36722224950790405, "learning_rate": 2.9508708060870914e-05, "loss": 0.0337, "step": 65270 }, { "grad_norm": 0.3507959842681885, "learning_rate": 2.9493626828619948e-05, "loss": 0.0327, "step": 65280 }, { "grad_norm": 0.7528233528137207, "learning_rate": 2.9478547838915137e-05, "loss": 0.0319, "step": 65290 }, { "grad_norm": 0.6267837285995483, "learning_rate": 2.946347109340548e-05, "loss": 0.0351, "step": 65300 }, { "grad_norm": 0.4903465807437897, "learning_rate": 2.944839659373976e-05, "loss": 0.0348, "step": 65310 }, { "grad_norm": 0.5041782855987549, "learning_rate": 2.943332434156647e-05, "loss": 0.0361, "step": 65320 }, { "grad_norm": 0.4563165009021759, "learning_rate": 2.9418254338533912e-05, "loss": 0.0413, "step": 65330 }, { "grad_norm": 0.8155187964439392, "learning_rate": 2.940318658629011e-05, "loss": 0.0385, "step": 65340 }, { "grad_norm": 0.4537910223007202, "learning_rate": 2.938812108648287e-05, "loss": 0.0387, "step": 65350 }, { "grad_norm": 0.4042745530605316, "learning_rate": 2.937305784075972e-05, "loss": 0.0307, "step": 65360 }, { "grad_norm": 1.1121853590011597, "learning_rate": 2.9357996850767937e-05, "loss": 0.0315, "step": 65370 }, { "grad_norm": 0.49915045499801636, "learning_rate": 2.9342938118154595e-05, "loss": 0.0317, "step": 65380 }, { "grad_norm": 0.6436570882797241, "learning_rate": 2.9327881644566485e-05, "loss": 0.0468, "step": 65390 }, { "grad_norm": 0.3899194300174713, "learning_rate": 2.9312827431650157e-05, "loss": 0.0354, "step": 65400 }, { "grad_norm": 0.2012491226196289, "learning_rate": 2.92977754810519e-05, "loss": 0.052, "step": 65410 }, { "grad_norm": 0.46721842885017395, "learning_rate": 2.928272579441781e-05, "loss": 0.0394, "step": 65420 }, { "grad_norm": 0.40529000759124756, "learning_rate": 2.9267678373393676e-05, "loss": 0.0472, "step": 65430 }, { "grad_norm": 0.3421226143836975, "learning_rate": 2.925263321962507e-05, "loss": 0.0398, "step": 65440 }, { "grad_norm": 0.4314251244068146, "learning_rate": 2.9237590334757292e-05, "loss": 0.033, "step": 65450 }, { "grad_norm": 0.4515446424484253, "learning_rate": 2.922254972043543e-05, "loss": 0.0385, "step": 65460 }, { "grad_norm": 0.7039005756378174, "learning_rate": 2.9207511378304265e-05, "loss": 0.0364, "step": 65470 }, { "grad_norm": 0.8294596672058105, "learning_rate": 2.9192475310008428e-05, "loss": 0.0457, "step": 65480 }, { "grad_norm": 0.44456467032432556, "learning_rate": 2.917744151719219e-05, "loss": 0.0304, "step": 65490 }, { "grad_norm": 0.47872164845466614, "learning_rate": 2.916241000149964e-05, "loss": 0.0361, "step": 65500 }, { "grad_norm": 0.35438913106918335, "learning_rate": 2.9147380764574622e-05, "loss": 0.0279, "step": 65510 }, { "grad_norm": 0.4814554452896118, "learning_rate": 2.9132353808060674e-05, "loss": 0.0478, "step": 65520 }, { "grad_norm": 0.40950220823287964, "learning_rate": 2.9117329133601134e-05, "loss": 0.0347, "step": 65530 }, { "grad_norm": 0.3313286006450653, "learning_rate": 2.9102306742839093e-05, "loss": 0.0396, "step": 65540 }, { "grad_norm": 0.3339052200317383, "learning_rate": 2.9087286637417375e-05, "loss": 0.0342, "step": 65550 }, { "grad_norm": 0.4836842119693756, "learning_rate": 2.9072268818978537e-05, "loss": 0.0455, "step": 65560 }, { "grad_norm": 0.556890070438385, "learning_rate": 2.905725328916493e-05, "loss": 0.0384, "step": 65570 }, { "grad_norm": 0.620067298412323, "learning_rate": 2.9042240049618608e-05, "loss": 0.0375, "step": 65580 }, { "grad_norm": 0.5531938076019287, "learning_rate": 2.9027229101981413e-05, "loss": 0.0364, "step": 65590 }, { "grad_norm": 0.31997162103652954, "learning_rate": 2.901222044789488e-05, "loss": 0.0407, "step": 65600 }, { "grad_norm": 0.3420173227787018, "learning_rate": 2.89972140890004e-05, "loss": 0.0314, "step": 65610 }, { "grad_norm": 0.5415268540382385, "learning_rate": 2.8982210026938988e-05, "loss": 0.0391, "step": 65620 }, { "grad_norm": 0.5123199820518494, "learning_rate": 2.896720826335151e-05, "loss": 0.0539, "step": 65630 }, { "grad_norm": 0.9438050389289856, "learning_rate": 2.8952208799878493e-05, "loss": 0.0437, "step": 65640 }, { "grad_norm": 0.42458364367485046, "learning_rate": 2.89372116381603e-05, "loss": 0.0355, "step": 65650 }, { "grad_norm": 0.513439953327179, "learning_rate": 2.8922216779836932e-05, "loss": 0.0421, "step": 65660 }, { "grad_norm": 0.5004306435585022, "learning_rate": 2.8907224226548285e-05, "loss": 0.0308, "step": 65670 }, { "grad_norm": 0.5291873812675476, "learning_rate": 2.8892233979933868e-05, "loss": 0.0353, "step": 65680 }, { "grad_norm": 0.745408296585083, "learning_rate": 2.8877246041633026e-05, "loss": 0.0397, "step": 65690 }, { "grad_norm": 0.6606958508491516, "learning_rate": 2.886226041328478e-05, "loss": 0.0375, "step": 65700 }, { "grad_norm": 0.40757983922958374, "learning_rate": 2.8847277096527957e-05, "loss": 0.0313, "step": 65710 }, { "grad_norm": 0.48730212450027466, "learning_rate": 2.8832296093001105e-05, "loss": 0.0285, "step": 65720 }, { "grad_norm": 0.2730180323123932, "learning_rate": 2.8817317404342547e-05, "loss": 0.0315, "step": 65730 }, { "grad_norm": 0.48534536361694336, "learning_rate": 2.880234103219029e-05, "loss": 0.0427, "step": 65740 }, { "grad_norm": 0.8564878106117249, "learning_rate": 2.878736697818215e-05, "loss": 0.0322, "step": 65750 }, { "grad_norm": 0.5523233413696289, "learning_rate": 2.877239524395568e-05, "loss": 0.0364, "step": 65760 }, { "grad_norm": 0.2777637541294098, "learning_rate": 2.8757425831148134e-05, "loss": 0.0378, "step": 65770 }, { "grad_norm": 0.39207401871681213, "learning_rate": 2.874245874139658e-05, "loss": 0.0446, "step": 65780 }, { "grad_norm": 0.37939682602882385, "learning_rate": 2.8727493976337737e-05, "loss": 0.0261, "step": 65790 }, { "grad_norm": 0.46753421425819397, "learning_rate": 2.8712531537608218e-05, "loss": 0.0293, "step": 65800 }, { "grad_norm": 0.43204423785209656, "learning_rate": 2.8697571426844217e-05, "loss": 0.0439, "step": 65810 }, { "grad_norm": 0.5158511996269226, "learning_rate": 2.86826136456818e-05, "loss": 0.0352, "step": 65820 }, { "grad_norm": 0.3814935088157654, "learning_rate": 2.8667658195756685e-05, "loss": 0.0409, "step": 65830 }, { "grad_norm": 0.5590732097625732, "learning_rate": 2.8652705078704412e-05, "loss": 0.047, "step": 65840 }, { "grad_norm": 0.5182101726531982, "learning_rate": 2.863775429616018e-05, "loss": 0.0491, "step": 65850 }, { "grad_norm": 0.3869653046131134, "learning_rate": 2.862280584975906e-05, "loss": 0.0328, "step": 65860 }, { "grad_norm": 0.3362019658088684, "learning_rate": 2.8607859741135735e-05, "loss": 0.0373, "step": 65870 }, { "grad_norm": 0.28827428817749023, "learning_rate": 2.8592915971924723e-05, "loss": 0.046, "step": 65880 }, { "grad_norm": 0.440719872713089, "learning_rate": 2.8577974543760212e-05, "loss": 0.0402, "step": 65890 }, { "grad_norm": 0.4221699833869934, "learning_rate": 2.8563035458276222e-05, "loss": 0.0347, "step": 65900 }, { "grad_norm": 0.2516860365867615, "learning_rate": 2.8548098717106424e-05, "loss": 0.0378, "step": 65910 }, { "grad_norm": 0.3926306962966919, "learning_rate": 2.85331643218843e-05, "loss": 0.0367, "step": 65920 }, { "grad_norm": 0.49555957317352295, "learning_rate": 2.8518232274243046e-05, "loss": 0.0428, "step": 65930 }, { "grad_norm": 0.9546762704849243, "learning_rate": 2.8503302575815638e-05, "loss": 0.0533, "step": 65940 }, { "grad_norm": 0.2717278003692627, "learning_rate": 2.8488375228234716e-05, "loss": 0.039, "step": 65950 }, { "grad_norm": 0.5408127307891846, "learning_rate": 2.8473450233132738e-05, "loss": 0.037, "step": 65960 }, { "grad_norm": 0.40675389766693115, "learning_rate": 2.8458527592141898e-05, "loss": 0.0336, "step": 65970 }, { "grad_norm": 0.27473345398902893, "learning_rate": 2.8443607306894073e-05, "loss": 0.0401, "step": 65980 }, { "grad_norm": 0.28360024094581604, "learning_rate": 2.8428689379020946e-05, "loss": 0.0398, "step": 65990 }, { "grad_norm": 0.3685778081417084, "learning_rate": 2.841377381015391e-05, "loss": 0.0422, "step": 66000 }, { "grad_norm": 0.3389087915420532, "learning_rate": 2.8398860601924144e-05, "loss": 0.0323, "step": 66010 }, { "grad_norm": 0.4310758411884308, "learning_rate": 2.838394975596248e-05, "loss": 0.0302, "step": 66020 }, { "grad_norm": 0.4610023498535156, "learning_rate": 2.83690412738996e-05, "loss": 0.0449, "step": 66030 }, { "grad_norm": 0.28204345703125, "learning_rate": 2.835413515736579e-05, "loss": 0.0411, "step": 66040 }, { "grad_norm": 0.4317043125629425, "learning_rate": 2.8339231407991267e-05, "loss": 0.0366, "step": 66050 }, { "grad_norm": 0.6277415156364441, "learning_rate": 2.832433002740581e-05, "loss": 0.0329, "step": 66060 }, { "grad_norm": 0.47248584032058716, "learning_rate": 2.8309431017239047e-05, "loss": 0.0306, "step": 66070 }, { "grad_norm": 0.32660773396492004, "learning_rate": 2.829453437912028e-05, "loss": 0.0348, "step": 66080 }, { "grad_norm": 0.3767576217651367, "learning_rate": 2.8279640114678617e-05, "loss": 0.0283, "step": 66090 }, { "grad_norm": 0.5863749384880066, "learning_rate": 2.8264748225542836e-05, "loss": 0.0443, "step": 66100 }, { "grad_norm": 0.2862873375415802, "learning_rate": 2.8249858713341516e-05, "loss": 0.0332, "step": 66110 }, { "grad_norm": 0.5873562097549438, "learning_rate": 2.823497157970294e-05, "loss": 0.0401, "step": 66120 }, { "grad_norm": 0.3245088458061218, "learning_rate": 2.8220086826255164e-05, "loss": 0.0279, "step": 66130 }, { "grad_norm": 0.5545581579208374, "learning_rate": 2.8205204454625934e-05, "loss": 0.0411, "step": 66140 }, { "grad_norm": 0.4236770272254944, "learning_rate": 2.8190324466442798e-05, "loss": 0.0327, "step": 66150 }, { "grad_norm": 0.8384367823600769, "learning_rate": 2.8175446863332965e-05, "loss": 0.0337, "step": 66160 }, { "grad_norm": 0.4002092182636261, "learning_rate": 2.816057164692345e-05, "loss": 0.0408, "step": 66170 }, { "grad_norm": 0.38675838708877563, "learning_rate": 2.8145698818840992e-05, "loss": 0.0455, "step": 66180 }, { "grad_norm": 0.6154663562774658, "learning_rate": 2.8130828380712065e-05, "loss": 0.0344, "step": 66190 }, { "grad_norm": 0.5918500423431396, "learning_rate": 2.811596033416285e-05, "loss": 0.0366, "step": 66200 }, { "grad_norm": 0.5183285474777222, "learning_rate": 2.8101094680819307e-05, "loss": 0.0366, "step": 66210 }, { "grad_norm": 0.9462196230888367, "learning_rate": 2.808623142230715e-05, "loss": 0.036, "step": 66220 }, { "grad_norm": 0.39583393931388855, "learning_rate": 2.8071370560251757e-05, "loss": 0.03, "step": 66230 }, { "grad_norm": 0.579078197479248, "learning_rate": 2.805651209627831e-05, "loss": 0.0402, "step": 66240 }, { "grad_norm": 0.7793547511100769, "learning_rate": 2.804165603201171e-05, "loss": 0.0378, "step": 66250 }, { "grad_norm": 0.4501941502094269, "learning_rate": 2.802680236907661e-05, "loss": 0.0313, "step": 66260 }, { "grad_norm": 0.44451361894607544, "learning_rate": 2.8011951109097344e-05, "loss": 0.0319, "step": 66270 }, { "grad_norm": 2.029386281967163, "learning_rate": 2.7997102253698072e-05, "loss": 0.0376, "step": 66280 }, { "grad_norm": 0.3421858251094818, "learning_rate": 2.7982255804502588e-05, "loss": 0.0342, "step": 66290 }, { "grad_norm": 0.33308887481689453, "learning_rate": 2.7967411763134516e-05, "loss": 0.0401, "step": 66300 }, { "grad_norm": 0.4418693482875824, "learning_rate": 2.7952570131217166e-05, "loss": 0.0304, "step": 66310 }, { "grad_norm": 0.41588741540908813, "learning_rate": 2.7937730910373618e-05, "loss": 0.0374, "step": 66320 }, { "grad_norm": 0.2556426525115967, "learning_rate": 2.7922894102226628e-05, "loss": 0.0297, "step": 66330 }, { "grad_norm": 0.3673595190048218, "learning_rate": 2.790805970839876e-05, "loss": 0.033, "step": 66340 }, { "grad_norm": 0.3478824198246002, "learning_rate": 2.789322773051225e-05, "loss": 0.0262, "step": 66350 }, { "grad_norm": 0.44691163301467896, "learning_rate": 2.7878398170189124e-05, "loss": 0.0331, "step": 66360 }, { "grad_norm": 0.45304420590400696, "learning_rate": 2.7863571029051112e-05, "loss": 0.0377, "step": 66370 }, { "grad_norm": 0.28725430369377136, "learning_rate": 2.784874630871971e-05, "loss": 0.0393, "step": 66380 }, { "grad_norm": 0.6998292803764343, "learning_rate": 2.7833924010816086e-05, "loss": 0.0324, "step": 66390 }, { "grad_norm": 0.2731587886810303, "learning_rate": 2.7819104136961226e-05, "loss": 0.0391, "step": 66400 }, { "grad_norm": 0.1566474586725235, "learning_rate": 2.780428668877577e-05, "loss": 0.04, "step": 66410 }, { "grad_norm": 0.51982182264328, "learning_rate": 2.7789471667880152e-05, "loss": 0.0387, "step": 66420 }, { "grad_norm": 0.38417962193489075, "learning_rate": 2.7774659075894523e-05, "loss": 0.0395, "step": 66430 }, { "grad_norm": 0.4742718040943146, "learning_rate": 2.7759848914438778e-05, "loss": 0.0475, "step": 66440 }, { "grad_norm": 0.42926886677742004, "learning_rate": 2.7745041185132496e-05, "loss": 0.0397, "step": 66450 }, { "grad_norm": 0.4651087820529938, "learning_rate": 2.773023588959506e-05, "loss": 0.0448, "step": 66460 }, { "grad_norm": 0.37559306621551514, "learning_rate": 2.7715433029445555e-05, "loss": 0.0402, "step": 66470 }, { "grad_norm": 0.35455650091171265, "learning_rate": 2.7700632606302766e-05, "loss": 0.0333, "step": 66480 }, { "grad_norm": 0.49852126836776733, "learning_rate": 2.768583462178528e-05, "loss": 0.0369, "step": 66490 }, { "grad_norm": 0.5982773900032043, "learning_rate": 2.767103907751136e-05, "loss": 0.0332, "step": 66500 }, { "grad_norm": 0.760187029838562, "learning_rate": 2.7656245975099065e-05, "loss": 0.0347, "step": 66510 }, { "grad_norm": 0.3819600045681, "learning_rate": 2.7641455316166093e-05, "loss": 0.0372, "step": 66520 }, { "grad_norm": 0.34834903478622437, "learning_rate": 2.7626667102329973e-05, "loss": 0.0376, "step": 66530 }, { "grad_norm": 0.3210391104221344, "learning_rate": 2.761188133520788e-05, "loss": 0.0307, "step": 66540 }, { "grad_norm": 0.9482568502426147, "learning_rate": 2.7597098016416784e-05, "loss": 0.0381, "step": 66550 }, { "grad_norm": 0.3789004385471344, "learning_rate": 2.7582317147573368e-05, "loss": 0.0313, "step": 66560 }, { "grad_norm": 0.708365797996521, "learning_rate": 2.7567538730294064e-05, "loss": 0.0398, "step": 66570 }, { "grad_norm": 0.58854740858078, "learning_rate": 2.7552762766194973e-05, "loss": 0.0419, "step": 66580 }, { "grad_norm": 0.4328307807445526, "learning_rate": 2.753798925689202e-05, "loss": 0.0402, "step": 66590 }, { "grad_norm": 0.47437265515327454, "learning_rate": 2.7523218204000774e-05, "loss": 0.0335, "step": 66600 }, { "grad_norm": 0.3905391991138458, "learning_rate": 2.7508449609136612e-05, "loss": 0.03, "step": 66610 }, { "grad_norm": 0.8783509135246277, "learning_rate": 2.749368347391454e-05, "loss": 0.0424, "step": 66620 }, { "grad_norm": 0.8037173748016357, "learning_rate": 2.7478919799949443e-05, "loss": 0.052, "step": 66630 }, { "grad_norm": 0.5077887773513794, "learning_rate": 2.7464158588855803e-05, "loss": 0.0365, "step": 66640 }, { "grad_norm": 0.5593652129173279, "learning_rate": 2.7449399842247918e-05, "loss": 0.0317, "step": 66650 }, { "grad_norm": 0.3310602307319641, "learning_rate": 2.7434643561739736e-05, "loss": 0.0347, "step": 66660 }, { "grad_norm": 0.5159420967102051, "learning_rate": 2.7419889748945016e-05, "loss": 0.0316, "step": 66670 }, { "grad_norm": 0.35997316241264343, "learning_rate": 2.7405138405477203e-05, "loss": 0.0403, "step": 66680 }, { "grad_norm": 0.36620187759399414, "learning_rate": 2.73903895329495e-05, "loss": 0.0406, "step": 66690 }, { "grad_norm": 0.5103565454483032, "learning_rate": 2.737564313297478e-05, "loss": 0.0416, "step": 66700 }, { "grad_norm": 0.45747649669647217, "learning_rate": 2.736089920716572e-05, "loss": 0.0352, "step": 66710 }, { "grad_norm": 0.4757232367992401, "learning_rate": 2.7346157757134706e-05, "loss": 0.0361, "step": 66720 }, { "grad_norm": 0.45375320315361023, "learning_rate": 2.733141878449379e-05, "loss": 0.0301, "step": 66730 }, { "grad_norm": 0.32375508546829224, "learning_rate": 2.7316682290854844e-05, "loss": 0.041, "step": 66740 }, { "grad_norm": 0.34431353211402893, "learning_rate": 2.7301948277829414e-05, "loss": 0.0391, "step": 66750 }, { "grad_norm": 0.4584016501903534, "learning_rate": 2.7287216747028814e-05, "loss": 0.0385, "step": 66760 }, { "grad_norm": 0.5582137703895569, "learning_rate": 2.7272487700064025e-05, "loss": 0.0362, "step": 66770 }, { "grad_norm": 1.129568338394165, "learning_rate": 2.725776113854583e-05, "loss": 0.0453, "step": 66780 }, { "grad_norm": 0.5998116731643677, "learning_rate": 2.7243037064084664e-05, "loss": 0.0341, "step": 66790 }, { "grad_norm": 0.4053252339363098, "learning_rate": 2.7228315478290766e-05, "loss": 0.0424, "step": 66800 }, { "grad_norm": 0.3431435227394104, "learning_rate": 2.721359638277402e-05, "loss": 0.0351, "step": 66810 }, { "grad_norm": 0.37718847393989563, "learning_rate": 2.7198879779144148e-05, "loss": 0.0272, "step": 66820 }, { "grad_norm": 0.33559921383857727, "learning_rate": 2.7184165669010485e-05, "loss": 0.0349, "step": 66830 }, { "grad_norm": 0.4126876890659332, "learning_rate": 2.716945405398218e-05, "loss": 0.0389, "step": 66840 }, { "grad_norm": 0.4845593273639679, "learning_rate": 2.715474493566803e-05, "loss": 0.0356, "step": 66850 }, { "grad_norm": 0.48123419284820557, "learning_rate": 2.7140038315676652e-05, "loss": 0.0277, "step": 66860 }, { "grad_norm": 0.276673823595047, "learning_rate": 2.7125334195616275e-05, "loss": 0.0621, "step": 66870 }, { "grad_norm": 0.37286069989204407, "learning_rate": 2.7110632577094997e-05, "loss": 0.0392, "step": 66880 }, { "grad_norm": 0.35993924736976624, "learning_rate": 2.709593346172051e-05, "loss": 0.0423, "step": 66890 }, { "grad_norm": 0.4058111011981964, "learning_rate": 2.708123685110032e-05, "loss": 0.0342, "step": 66900 }, { "grad_norm": 0.31732919812202454, "learning_rate": 2.7066542746841588e-05, "loss": 0.021, "step": 66910 }, { "grad_norm": 1.337466835975647, "learning_rate": 2.7051851150551267e-05, "loss": 0.043, "step": 66920 }, { "grad_norm": 0.37407422065734863, "learning_rate": 2.7037162063835996e-05, "loss": 0.032, "step": 66930 }, { "grad_norm": 0.6736093759536743, "learning_rate": 2.702247548830218e-05, "loss": 0.04, "step": 66940 }, { "grad_norm": 0.6007930040359497, "learning_rate": 2.7007791425555883e-05, "loss": 0.0291, "step": 66950 }, { "grad_norm": 0.25127917528152466, "learning_rate": 2.6993109877202945e-05, "loss": 0.034, "step": 66960 }, { "grad_norm": 0.6849625110626221, "learning_rate": 2.6978430844848935e-05, "loss": 0.0367, "step": 66970 }, { "grad_norm": 0.431149423122406, "learning_rate": 2.69637543300991e-05, "loss": 0.0329, "step": 66980 }, { "grad_norm": 0.34208986163139343, "learning_rate": 2.694908033455848e-05, "loss": 0.0334, "step": 66990 }, { "grad_norm": 0.42593246698379517, "learning_rate": 2.6934408859831733e-05, "loss": 0.0402, "step": 67000 }, { "grad_norm": 0.47903046011924744, "learning_rate": 2.69197399075234e-05, "loss": 0.0409, "step": 67010 }, { "grad_norm": 0.455973744392395, "learning_rate": 2.6905073479237584e-05, "loss": 0.0352, "step": 67020 }, { "grad_norm": 1.0511151552200317, "learning_rate": 2.6890409576578235e-05, "loss": 0.0433, "step": 67030 }, { "grad_norm": 0.5707403421401978, "learning_rate": 2.687574820114892e-05, "loss": 0.0345, "step": 67040 }, { "grad_norm": 0.6949731707572937, "learning_rate": 2.686108935455305e-05, "loss": 0.0391, "step": 67050 }, { "grad_norm": 0.4849386513233185, "learning_rate": 2.6846433038393603e-05, "loss": 0.0379, "step": 67060 }, { "grad_norm": 0.44073575735092163, "learning_rate": 2.6831779254273482e-05, "loss": 0.0333, "step": 67070 }, { "grad_norm": 1.5468335151672363, "learning_rate": 2.6817128003795122e-05, "loss": 0.0325, "step": 67080 }, { "grad_norm": 0.4830724895000458, "learning_rate": 2.680247928856081e-05, "loss": 0.0352, "step": 67090 }, { "grad_norm": 0.46390581130981445, "learning_rate": 2.6787833110172466e-05, "loss": 0.0419, "step": 67100 }, { "grad_norm": 0.26752549409866333, "learning_rate": 2.6773189470231807e-05, "loss": 0.0347, "step": 67110 }, { "grad_norm": 0.41809090971946716, "learning_rate": 2.675854837034022e-05, "loss": 0.0354, "step": 67120 }, { "grad_norm": 0.36867907643318176, "learning_rate": 2.6743909812098833e-05, "loss": 0.0335, "step": 67130 }, { "grad_norm": 0.32116758823394775, "learning_rate": 2.67292737971085e-05, "loss": 0.0374, "step": 67140 }, { "grad_norm": 0.4792124629020691, "learning_rate": 2.671464032696982e-05, "loss": 0.0368, "step": 67150 }, { "grad_norm": 0.8749402761459351, "learning_rate": 2.6700009403283055e-05, "loss": 0.0427, "step": 67160 }, { "grad_norm": 0.4368797838687897, "learning_rate": 2.6685381027648214e-05, "loss": 0.0396, "step": 67170 }, { "grad_norm": 0.46280232071876526, "learning_rate": 2.6670755201665086e-05, "loss": 0.0472, "step": 67180 }, { "grad_norm": 0.44184598326683044, "learning_rate": 2.6656131926933066e-05, "loss": 0.041, "step": 67190 }, { "grad_norm": 0.68365478515625, "learning_rate": 2.664151120505136e-05, "loss": 0.0329, "step": 67200 }, { "grad_norm": 0.3082832396030426, "learning_rate": 2.6626893037618873e-05, "loss": 0.0383, "step": 67210 }, { "grad_norm": 0.5392576456069946, "learning_rate": 2.6612277426234243e-05, "loss": 0.0512, "step": 67220 }, { "grad_norm": 0.6426820755004883, "learning_rate": 2.659766437249577e-05, "loss": 0.053, "step": 67230 }, { "grad_norm": 0.3739866316318512, "learning_rate": 2.6583053878001557e-05, "loss": 0.0335, "step": 67240 }, { "grad_norm": 0.5315961837768555, "learning_rate": 2.6568445944349323e-05, "loss": 0.0404, "step": 67250 }, { "grad_norm": 0.4417579472064972, "learning_rate": 2.6553840573136657e-05, "loss": 0.0354, "step": 67260 }, { "grad_norm": 0.4738309979438782, "learning_rate": 2.6539237765960706e-05, "loss": 0.0395, "step": 67270 }, { "grad_norm": 0.9260633587837219, "learning_rate": 2.6524637524418465e-05, "loss": 0.0398, "step": 67280 }, { "grad_norm": 0.2911798357963562, "learning_rate": 2.651003985010655e-05, "loss": 0.0298, "step": 67290 }, { "grad_norm": 0.45543813705444336, "learning_rate": 2.649544474462138e-05, "loss": 0.042, "step": 67300 }, { "grad_norm": 0.31195151805877686, "learning_rate": 2.6480852209559015e-05, "loss": 0.0391, "step": 67310 }, { "grad_norm": 0.4098207950592041, "learning_rate": 2.6466262246515296e-05, "loss": 0.0287, "step": 67320 }, { "grad_norm": 0.4976691007614136, "learning_rate": 2.6451674857085746e-05, "loss": 0.0394, "step": 67330 }, { "grad_norm": 0.5472710728645325, "learning_rate": 2.6437090042865655e-05, "loss": 0.0446, "step": 67340 }, { "grad_norm": 0.2786963880062103, "learning_rate": 2.6422507805449947e-05, "loss": 0.0333, "step": 67350 }, { "grad_norm": 0.319307416677475, "learning_rate": 2.6407928146433357e-05, "loss": 0.0344, "step": 67360 }, { "grad_norm": 1.0296173095703125, "learning_rate": 2.639335106741026e-05, "loss": 0.0347, "step": 67370 }, { "grad_norm": 0.33380427956581116, "learning_rate": 2.6378776569974796e-05, "loss": 0.0441, "step": 67380 }, { "grad_norm": 2.269620418548584, "learning_rate": 2.6364204655720814e-05, "loss": 0.0347, "step": 67390 }, { "grad_norm": 0.4412243664264679, "learning_rate": 2.6349635326241894e-05, "loss": 0.0338, "step": 67400 }, { "grad_norm": 0.420764684677124, "learning_rate": 2.633506858313127e-05, "loss": 0.0412, "step": 67410 }, { "grad_norm": 0.3153972625732422, "learning_rate": 2.6320504427981975e-05, "loss": 0.0361, "step": 67420 }, { "grad_norm": 0.34216997027397156, "learning_rate": 2.6305942862386734e-05, "loss": 0.0309, "step": 67430 }, { "grad_norm": 0.47034528851509094, "learning_rate": 2.629138388793793e-05, "loss": 0.0325, "step": 67440 }, { "grad_norm": 0.3618665933609009, "learning_rate": 2.6276827506227743e-05, "loss": 0.0432, "step": 67450 }, { "grad_norm": 0.34748050570487976, "learning_rate": 2.6262273718848037e-05, "loss": 0.0313, "step": 67460 }, { "grad_norm": 0.4074849486351013, "learning_rate": 2.6247722527390407e-05, "loss": 0.0318, "step": 67470 }, { "grad_norm": 0.38971367478370667, "learning_rate": 2.6233173933446104e-05, "loss": 0.0374, "step": 67480 }, { "grad_norm": 0.8137485980987549, "learning_rate": 2.6218627938606188e-05, "loss": 0.0367, "step": 67490 }, { "grad_norm": 0.28931769728660583, "learning_rate": 2.620408454446135e-05, "loss": 0.0395, "step": 67500 }, { "grad_norm": 0.248046413064003, "learning_rate": 2.6189543752602046e-05, "loss": 0.0365, "step": 67510 }, { "grad_norm": 1.1518263816833496, "learning_rate": 2.6175005564618444e-05, "loss": 0.0357, "step": 67520 }, { "grad_norm": 0.34731560945510864, "learning_rate": 2.6160469982100428e-05, "loss": 0.0361, "step": 67530 }, { "grad_norm": 0.5532610416412354, "learning_rate": 2.6145937006637555e-05, "loss": 0.0555, "step": 67540 }, { "grad_norm": 0.3438444137573242, "learning_rate": 2.6131406639819166e-05, "loss": 0.0386, "step": 67550 }, { "grad_norm": 0.27120277285575867, "learning_rate": 2.611687888323424e-05, "loss": 0.0291, "step": 67560 }, { "grad_norm": 0.4641820192337036, "learning_rate": 2.610235373847153e-05, "loss": 0.0401, "step": 67570 }, { "grad_norm": 0.5902366638183594, "learning_rate": 2.6087831207119494e-05, "loss": 0.0454, "step": 67580 }, { "grad_norm": 0.4381672441959381, "learning_rate": 2.60733112907663e-05, "loss": 0.0434, "step": 67590 }, { "grad_norm": 0.6345741152763367, "learning_rate": 2.6058793990999786e-05, "loss": 0.04, "step": 67600 }, { "grad_norm": 0.5595613718032837, "learning_rate": 2.6044279309407593e-05, "loss": 0.0378, "step": 67610 }, { "grad_norm": 0.5207951664924622, "learning_rate": 2.6029767247576976e-05, "loss": 0.042, "step": 67620 }, { "grad_norm": 0.5007409453392029, "learning_rate": 2.601525780709497e-05, "loss": 0.0316, "step": 67630 }, { "grad_norm": 0.6469700932502747, "learning_rate": 2.6000750989548318e-05, "loss": 0.0338, "step": 67640 }, { "grad_norm": 0.4644550383090973, "learning_rate": 2.5986246796523462e-05, "loss": 0.0269, "step": 67650 }, { "grad_norm": 0.3006501793861389, "learning_rate": 2.5971745229606546e-05, "loss": 0.0346, "step": 67660 }, { "grad_norm": 0.5665158033370972, "learning_rate": 2.5957246290383442e-05, "loss": 0.0421, "step": 67670 }, { "grad_norm": 0.5200675129890442, "learning_rate": 2.594274998043975e-05, "loss": 0.0392, "step": 67680 }, { "grad_norm": 0.3064233064651489, "learning_rate": 2.5928256301360733e-05, "loss": 0.0324, "step": 67690 }, { "grad_norm": 1.3621619939804077, "learning_rate": 2.591376525473142e-05, "loss": 0.0286, "step": 67700 }, { "grad_norm": 0.4042647182941437, "learning_rate": 2.5899276842136523e-05, "loss": 0.0385, "step": 67710 }, { "grad_norm": 0.26362186670303345, "learning_rate": 2.5884791065160495e-05, "loss": 0.0478, "step": 67720 }, { "grad_norm": 0.42508625984191895, "learning_rate": 2.5870307925387436e-05, "loss": 0.0281, "step": 67730 }, { "grad_norm": 0.5483607649803162, "learning_rate": 2.585582742440125e-05, "loss": 0.029, "step": 67740 }, { "grad_norm": 0.5001892447471619, "learning_rate": 2.5841349563785455e-05, "loss": 0.0408, "step": 67750 }, { "grad_norm": 0.16979990899562836, "learning_rate": 2.5826874345123355e-05, "loss": 0.029, "step": 67760 }, { "grad_norm": 0.32552093267440796, "learning_rate": 2.581240176999792e-05, "loss": 0.0288, "step": 67770 }, { "grad_norm": 0.7105495929718018, "learning_rate": 2.579793183999189e-05, "loss": 0.0359, "step": 67780 }, { "grad_norm": 0.3813032805919647, "learning_rate": 2.5783464556687627e-05, "loss": 0.0395, "step": 67790 }, { "grad_norm": 0.3904130756855011, "learning_rate": 2.576899992166729e-05, "loss": 0.0326, "step": 67800 }, { "grad_norm": 0.2546674907207489, "learning_rate": 2.5754537936512668e-05, "loss": 0.0278, "step": 67810 }, { "grad_norm": 0.660497784614563, "learning_rate": 2.574007860280535e-05, "loss": 0.0482, "step": 67820 }, { "grad_norm": 0.3541800081729889, "learning_rate": 2.5725621922126518e-05, "loss": 0.0294, "step": 67830 }, { "grad_norm": 0.40454205870628357, "learning_rate": 2.571116789605722e-05, "loss": 0.0357, "step": 67840 }, { "grad_norm": 0.3268132209777832, "learning_rate": 2.5696716526178054e-05, "loss": 0.0399, "step": 67850 }, { "grad_norm": 0.5828756093978882, "learning_rate": 2.568226781406946e-05, "loss": 0.0409, "step": 67860 }, { "grad_norm": 0.466346800327301, "learning_rate": 2.5667821761311474e-05, "loss": 0.0493, "step": 67870 }, { "grad_norm": 0.3710732161998749, "learning_rate": 2.5653378369483917e-05, "loss": 0.029, "step": 67880 }, { "grad_norm": 0.6039983034133911, "learning_rate": 2.5638937640166292e-05, "loss": 0.0316, "step": 67890 }, { "grad_norm": 0.3335379362106323, "learning_rate": 2.5624499574937842e-05, "loss": 0.0276, "step": 67900 }, { "grad_norm": 0.2568524479866028, "learning_rate": 2.5610064175377456e-05, "loss": 0.0379, "step": 67910 }, { "grad_norm": 0.28974026441574097, "learning_rate": 2.5595631443063782e-05, "loss": 0.0347, "step": 67920 }, { "grad_norm": 0.3402034044265747, "learning_rate": 2.5581201379575177e-05, "loss": 0.0302, "step": 67930 }, { "grad_norm": 0.36852264404296875, "learning_rate": 2.5566773986489655e-05, "loss": 0.0364, "step": 67940 }, { "grad_norm": 0.36785751581192017, "learning_rate": 2.5552349265385e-05, "loss": 0.0485, "step": 67950 }, { "grad_norm": 0.5896504521369934, "learning_rate": 2.5537927217838675e-05, "loss": 0.0281, "step": 67960 }, { "grad_norm": 0.600482702255249, "learning_rate": 2.5523507845427864e-05, "loss": 0.0437, "step": 67970 }, { "grad_norm": 0.5893974900245667, "learning_rate": 2.5509091149729413e-05, "loss": 0.039, "step": 67980 }, { "grad_norm": 0.3599942624568939, "learning_rate": 2.5494677132319962e-05, "loss": 0.0359, "step": 67990 }, { "grad_norm": 0.43007439374923706, "learning_rate": 2.548026579477575e-05, "loss": 0.0405, "step": 68000 }, { "grad_norm": 0.442423015832901, "learning_rate": 2.5465857138672834e-05, "loss": 0.0311, "step": 68010 }, { "grad_norm": 0.4039812684059143, "learning_rate": 2.5451451165586848e-05, "loss": 0.043, "step": 68020 }, { "grad_norm": 0.43761706352233887, "learning_rate": 2.543704787709329e-05, "loss": 0.0315, "step": 68030 }, { "grad_norm": 0.5477929711341858, "learning_rate": 2.542264727476723e-05, "loss": 0.0307, "step": 68040 }, { "grad_norm": 0.3777027428150177, "learning_rate": 2.5408249360183535e-05, "loss": 0.0423, "step": 68050 }, { "grad_norm": 0.3577861189842224, "learning_rate": 2.5393854134916694e-05, "loss": 0.032, "step": 68060 }, { "grad_norm": 0.42885130643844604, "learning_rate": 2.5379461600540993e-05, "loss": 0.0417, "step": 68070 }, { "grad_norm": 0.33744490146636963, "learning_rate": 2.5365071758630317e-05, "loss": 0.032, "step": 68080 }, { "grad_norm": 0.3924778997898102, "learning_rate": 2.535068461075839e-05, "loss": 0.0314, "step": 68090 }, { "grad_norm": 0.3669092059135437, "learning_rate": 2.5336300158498516e-05, "loss": 0.0374, "step": 68100 }, { "grad_norm": 0.9974305629730225, "learning_rate": 2.5321918403423793e-05, "loss": 0.039, "step": 68110 }, { "grad_norm": 0.4841557741165161, "learning_rate": 2.5307539347106957e-05, "loss": 0.0326, "step": 68120 }, { "grad_norm": 0.5022715926170349, "learning_rate": 2.529316299112049e-05, "loss": 0.0398, "step": 68130 }, { "grad_norm": 0.4316229224205017, "learning_rate": 2.5278789337036567e-05, "loss": 0.0291, "step": 68140 }, { "grad_norm": 0.5512083768844604, "learning_rate": 2.52644183864271e-05, "loss": 0.035, "step": 68150 }, { "grad_norm": 0.41576600074768066, "learning_rate": 2.525005014086363e-05, "loss": 0.0407, "step": 68160 }, { "grad_norm": 0.5116422176361084, "learning_rate": 2.5235684601917465e-05, "loss": 0.0473, "step": 68170 }, { "grad_norm": 0.40961962938308716, "learning_rate": 2.5221321771159618e-05, "loss": 0.039, "step": 68180 }, { "grad_norm": 0.8359146118164062, "learning_rate": 2.5206961650160743e-05, "loss": 0.0414, "step": 68190 }, { "grad_norm": 0.2858201563358307, "learning_rate": 2.519260424049129e-05, "loss": 0.0375, "step": 68200 }, { "grad_norm": 0.39611467719078064, "learning_rate": 2.5178249543721304e-05, "loss": 0.0421, "step": 68210 }, { "grad_norm": 0.2568466067314148, "learning_rate": 2.5163897561420668e-05, "loss": 0.0347, "step": 68220 }, { "grad_norm": 1.0536963939666748, "learning_rate": 2.514954829515883e-05, "loss": 0.0444, "step": 68230 }, { "grad_norm": 0.49162206053733826, "learning_rate": 2.513520174650505e-05, "loss": 0.0356, "step": 68240 }, { "grad_norm": 0.46980756521224976, "learning_rate": 2.5120857917028207e-05, "loss": 0.0308, "step": 68250 }, { "grad_norm": 0.6256283521652222, "learning_rate": 2.5106516808296953e-05, "loss": 0.0267, "step": 68260 }, { "grad_norm": 0.30235281586647034, "learning_rate": 2.5092178421879552e-05, "loss": 0.0384, "step": 68270 }, { "grad_norm": 0.3883184492588043, "learning_rate": 2.5077842759344116e-05, "loss": 0.0484, "step": 68280 }, { "grad_norm": 0.6564393043518066, "learning_rate": 2.5063509822258304e-05, "loss": 0.0399, "step": 68290 }, { "grad_norm": 0.5479861497879028, "learning_rate": 2.5049179612189576e-05, "loss": 0.0339, "step": 68300 }, { "grad_norm": 0.6879261136054993, "learning_rate": 2.5034852130705043e-05, "loss": 0.041, "step": 68310 }, { "grad_norm": 0.35598546266555786, "learning_rate": 2.5020527379371572e-05, "loss": 0.0409, "step": 68320 }, { "grad_norm": 0.3850798010826111, "learning_rate": 2.500620535975564e-05, "loss": 0.0328, "step": 68330 }, { "grad_norm": 0.44464120268821716, "learning_rate": 2.4991886073423515e-05, "loss": 0.0392, "step": 68340 }, { "grad_norm": 0.2357838749885559, "learning_rate": 2.4977569521941134e-05, "loss": 0.0306, "step": 68350 }, { "grad_norm": 0.35491350293159485, "learning_rate": 2.4963255706874143e-05, "loss": 0.0355, "step": 68360 }, { "grad_norm": 0.4995427131652832, "learning_rate": 2.4948944629787853e-05, "loss": 0.0364, "step": 68370 }, { "grad_norm": 0.5569459795951843, "learning_rate": 2.493463629224731e-05, "loss": 0.0393, "step": 68380 }, { "grad_norm": 0.3072482943534851, "learning_rate": 2.492033069581728e-05, "loss": 0.0374, "step": 68390 }, { "grad_norm": 0.40167516469955444, "learning_rate": 2.4906027842062137e-05, "loss": 0.0324, "step": 68400 }, { "grad_norm": 0.46770644187927246, "learning_rate": 2.48917277325461e-05, "loss": 0.0375, "step": 68410 }, { "grad_norm": 0.3656076192855835, "learning_rate": 2.487743036883295e-05, "loss": 0.0289, "step": 68420 }, { "grad_norm": 0.4589535892009735, "learning_rate": 2.486313575248626e-05, "loss": 0.029, "step": 68430 }, { "grad_norm": 0.4546028673648834, "learning_rate": 2.4848843885069235e-05, "loss": 0.0348, "step": 68440 }, { "grad_norm": 0.44085031747817993, "learning_rate": 2.483455476814484e-05, "loss": 0.0347, "step": 68450 }, { "grad_norm": 0.3836946487426758, "learning_rate": 2.4820268403275672e-05, "loss": 0.0326, "step": 68460 }, { "grad_norm": 0.542392909526825, "learning_rate": 2.480598479202413e-05, "loss": 0.0398, "step": 68470 }, { "grad_norm": 0.7268736958503723, "learning_rate": 2.4791703935952193e-05, "loss": 0.044, "step": 68480 }, { "grad_norm": 0.3780544102191925, "learning_rate": 2.477742583662163e-05, "loss": 0.0393, "step": 68490 }, { "grad_norm": 0.43273261189460754, "learning_rate": 2.476315049559384e-05, "loss": 0.0311, "step": 68500 }, { "grad_norm": 1.1475857496261597, "learning_rate": 2.4748877914429987e-05, "loss": 0.0501, "step": 68510 }, { "grad_norm": 0.27301114797592163, "learning_rate": 2.4734608094690857e-05, "loss": 0.0362, "step": 68520 }, { "grad_norm": 0.24693553149700165, "learning_rate": 2.4720341037937006e-05, "loss": 0.0465, "step": 68530 }, { "grad_norm": 0.3630271852016449, "learning_rate": 2.4706076745728645e-05, "loss": 0.0417, "step": 68540 }, { "grad_norm": 0.33529072999954224, "learning_rate": 2.469181521962573e-05, "loss": 0.0372, "step": 68550 }, { "grad_norm": 0.5712027549743652, "learning_rate": 2.467755646118783e-05, "loss": 0.0268, "step": 68560 }, { "grad_norm": 0.5722665786743164, "learning_rate": 2.46633004719743e-05, "loss": 0.0368, "step": 68570 }, { "grad_norm": 0.5816320180892944, "learning_rate": 2.4649047253544123e-05, "loss": 0.0413, "step": 68580 }, { "grad_norm": 0.21993625164031982, "learning_rate": 2.4634796807456016e-05, "loss": 0.0302, "step": 68590 }, { "grad_norm": 0.31185200810432434, "learning_rate": 2.46205491352684e-05, "loss": 0.0397, "step": 68600 }, { "grad_norm": 0.3684571087360382, "learning_rate": 2.4606304238539375e-05, "loss": 0.03, "step": 68610 }, { "grad_norm": 1.4642913341522217, "learning_rate": 2.4592062118826753e-05, "loss": 0.0373, "step": 68620 }, { "grad_norm": 0.4552222490310669, "learning_rate": 2.4577822777688004e-05, "loss": 0.0248, "step": 68630 }, { "grad_norm": 0.4952127933502197, "learning_rate": 2.4563586216680347e-05, "loss": 0.0362, "step": 68640 }, { "grad_norm": 0.35136309266090393, "learning_rate": 2.4549352437360622e-05, "loss": 0.0262, "step": 68650 }, { "grad_norm": 0.48872077465057373, "learning_rate": 2.4535121441285493e-05, "loss": 0.0406, "step": 68660 }, { "grad_norm": 0.7881930470466614, "learning_rate": 2.4520893230011172e-05, "loss": 0.0356, "step": 68670 }, { "grad_norm": 0.31330394744873047, "learning_rate": 2.450666780509368e-05, "loss": 0.0298, "step": 68680 }, { "grad_norm": 0.5243250727653503, "learning_rate": 2.4492445168088656e-05, "loss": 0.0352, "step": 68690 }, { "grad_norm": 0.40285927057266235, "learning_rate": 2.447822532055149e-05, "loss": 0.0448, "step": 68700 }, { "grad_norm": 0.7229301333427429, "learning_rate": 2.4464008264037212e-05, "loss": 0.0398, "step": 68710 }, { "grad_norm": 0.6767676472663879, "learning_rate": 2.4449794000100605e-05, "loss": 0.0467, "step": 68720 }, { "grad_norm": 0.3089878559112549, "learning_rate": 2.4435582530296108e-05, "loss": 0.0342, "step": 68730 }, { "grad_norm": 0.8003421425819397, "learning_rate": 2.4421373856177887e-05, "loss": 0.0332, "step": 68740 }, { "grad_norm": 0.4180993437767029, "learning_rate": 2.440716797929975e-05, "loss": 0.0287, "step": 68750 }, { "grad_norm": 0.676776647567749, "learning_rate": 2.439296490121526e-05, "loss": 0.0277, "step": 68760 }, { "grad_norm": 0.30323097109794617, "learning_rate": 2.4378764623477608e-05, "loss": 0.0318, "step": 68770 }, { "grad_norm": 0.3390173017978668, "learning_rate": 2.436456714763974e-05, "loss": 0.0299, "step": 68780 }, { "grad_norm": 0.3830207586288452, "learning_rate": 2.4350372475254262e-05, "loss": 0.0402, "step": 68790 }, { "grad_norm": 0.602725088596344, "learning_rate": 2.433618060787351e-05, "loss": 0.0362, "step": 68800 }, { "grad_norm": 0.3183201551437378, "learning_rate": 2.432199154704944e-05, "loss": 0.0386, "step": 68810 }, { "grad_norm": 0.47235530614852905, "learning_rate": 2.430780529433377e-05, "loss": 0.0382, "step": 68820 }, { "grad_norm": 0.3845216929912567, "learning_rate": 2.42936218512779e-05, "loss": 0.0355, "step": 68830 }, { "grad_norm": 0.33627766370773315, "learning_rate": 2.427944121943288e-05, "loss": 0.0299, "step": 68840 }, { "grad_norm": 0.4454176127910614, "learning_rate": 2.4265263400349497e-05, "loss": 0.0398, "step": 68850 }, { "grad_norm": 0.7337167859077454, "learning_rate": 2.4251088395578213e-05, "loss": 0.0354, "step": 68860 }, { "grad_norm": 0.4524880647659302, "learning_rate": 2.4236916206669214e-05, "loss": 0.0435, "step": 68870 }, { "grad_norm": 0.6819097399711609, "learning_rate": 2.4222746835172304e-05, "loss": 0.0457, "step": 68880 }, { "grad_norm": 0.5165082812309265, "learning_rate": 2.4208580282637065e-05, "loss": 0.0405, "step": 68890 }, { "grad_norm": 0.480369508266449, "learning_rate": 2.4194416550612687e-05, "loss": 0.032, "step": 68900 }, { "grad_norm": 0.2750324010848999, "learning_rate": 2.418025564064812e-05, "loss": 0.0305, "step": 68910 }, { "grad_norm": 1.3144885301589966, "learning_rate": 2.4166097554291978e-05, "loss": 0.0273, "step": 68920 }, { "grad_norm": 0.2378184050321579, "learning_rate": 2.4151942293092583e-05, "loss": 0.028, "step": 68930 }, { "grad_norm": 0.17574115097522736, "learning_rate": 2.41377898585979e-05, "loss": 0.0269, "step": 68940 }, { "grad_norm": 0.47002723813056946, "learning_rate": 2.4123640252355652e-05, "loss": 0.0363, "step": 68950 }, { "grad_norm": 0.3631378412246704, "learning_rate": 2.410949347591318e-05, "loss": 0.04, "step": 68960 }, { "grad_norm": 0.534595251083374, "learning_rate": 2.4095349530817578e-05, "loss": 0.0533, "step": 68970 }, { "grad_norm": 0.41081878542900085, "learning_rate": 2.4081208418615603e-05, "loss": 0.0369, "step": 68980 }, { "grad_norm": 0.6819378733634949, "learning_rate": 2.4067070140853725e-05, "loss": 0.0455, "step": 68990 }, { "grad_norm": 0.2254696786403656, "learning_rate": 2.405293469907805e-05, "loss": 0.0393, "step": 69000 }, { "grad_norm": 0.521300196647644, "learning_rate": 2.4038802094834446e-05, "loss": 0.0334, "step": 69010 }, { "grad_norm": 0.36331692337989807, "learning_rate": 2.4024672329668395e-05, "loss": 0.0371, "step": 69020 }, { "grad_norm": 0.40673431754112244, "learning_rate": 2.4010545405125125e-05, "loss": 0.0288, "step": 69030 }, { "grad_norm": 0.576367199420929, "learning_rate": 2.3996421322749528e-05, "loss": 0.0347, "step": 69040 }, { "grad_norm": 0.40141335129737854, "learning_rate": 2.3982300084086224e-05, "loss": 0.0337, "step": 69050 }, { "grad_norm": 0.8876836895942688, "learning_rate": 2.396818169067944e-05, "loss": 0.0335, "step": 69060 }, { "grad_norm": 0.6324887871742249, "learning_rate": 2.3954066144073178e-05, "loss": 0.0389, "step": 69070 }, { "grad_norm": 0.5331494808197021, "learning_rate": 2.39399534458111e-05, "loss": 0.0416, "step": 69080 }, { "grad_norm": 0.28181880712509155, "learning_rate": 2.392584359743651e-05, "loss": 0.0428, "step": 69090 }, { "grad_norm": 0.3712071180343628, "learning_rate": 2.3911736600492463e-05, "loss": 0.0253, "step": 69100 }, { "grad_norm": 0.5241031646728516, "learning_rate": 2.3897632456521672e-05, "loss": 0.0339, "step": 69110 }, { "grad_norm": 0.36345189809799194, "learning_rate": 2.388353116706658e-05, "loss": 0.0448, "step": 69120 }, { "grad_norm": 0.5817629098892212, "learning_rate": 2.3869432733669232e-05, "loss": 0.0361, "step": 69130 }, { "grad_norm": 0.4128390848636627, "learning_rate": 2.385533715787145e-05, "loss": 0.0434, "step": 69140 }, { "grad_norm": 0.339239239692688, "learning_rate": 2.3841244441214673e-05, "loss": 0.0323, "step": 69150 }, { "grad_norm": 0.45460349321365356, "learning_rate": 2.3827154585240073e-05, "loss": 0.0335, "step": 69160 }, { "grad_norm": 0.428120881319046, "learning_rate": 2.3813067591488498e-05, "loss": 0.047, "step": 69170 }, { "grad_norm": 0.5933500528335571, "learning_rate": 2.3798983461500494e-05, "loss": 0.0443, "step": 69180 }, { "grad_norm": 0.8242058157920837, "learning_rate": 2.3784902196816256e-05, "loss": 0.0441, "step": 69190 }, { "grad_norm": 0.39921805262565613, "learning_rate": 2.3770823798975716e-05, "loss": 0.0347, "step": 69200 }, { "grad_norm": 0.44562503695487976, "learning_rate": 2.3756748269518424e-05, "loss": 0.0399, "step": 69210 }, { "grad_norm": 0.39141905307769775, "learning_rate": 2.374267560998371e-05, "loss": 0.0405, "step": 69220 }, { "grad_norm": 0.46820777654647827, "learning_rate": 2.3728605821910477e-05, "loss": 0.037, "step": 69230 }, { "grad_norm": 0.507562518119812, "learning_rate": 2.3714538906837453e-05, "loss": 0.0408, "step": 69240 }, { "grad_norm": 0.4076148271560669, "learning_rate": 2.3700474866302913e-05, "loss": 0.0341, "step": 69250 }, { "grad_norm": 0.358707070350647, "learning_rate": 2.3686413701844917e-05, "loss": 0.0305, "step": 69260 }, { "grad_norm": 0.38871148228645325, "learning_rate": 2.367235541500114e-05, "loss": 0.0292, "step": 69270 }, { "grad_norm": 0.22639387845993042, "learning_rate": 2.3658300007308993e-05, "loss": 0.031, "step": 69280 }, { "grad_norm": 0.44307681918144226, "learning_rate": 2.364424748030555e-05, "loss": 0.0267, "step": 69290 }, { "grad_norm": 0.4209083020687103, "learning_rate": 2.3630197835527595e-05, "loss": 0.0396, "step": 69300 }, { "grad_norm": 0.42981991171836853, "learning_rate": 2.3616151074511546e-05, "loss": 0.0385, "step": 69310 }, { "grad_norm": 0.3668144643306732, "learning_rate": 2.3602107198793545e-05, "loss": 0.0378, "step": 69320 }, { "grad_norm": 0.4417128562927246, "learning_rate": 2.3588066209909427e-05, "loss": 0.034, "step": 69330 }, { "grad_norm": 0.5294080376625061, "learning_rate": 2.3574028109394662e-05, "loss": 0.0368, "step": 69340 }, { "grad_norm": 0.4237736463546753, "learning_rate": 2.3559992898784454e-05, "loss": 0.0362, "step": 69350 }, { "grad_norm": 0.3936921954154968, "learning_rate": 2.354596057961367e-05, "loss": 0.0296, "step": 69360 }, { "grad_norm": 0.3398755192756653, "learning_rate": 2.353193115341688e-05, "loss": 0.0369, "step": 69370 }, { "grad_norm": 0.297795832157135, "learning_rate": 2.3517904621728294e-05, "loss": 0.0276, "step": 69380 }, { "grad_norm": 0.32367071509361267, "learning_rate": 2.3503880986081855e-05, "loss": 0.0341, "step": 69390 }, { "grad_norm": 0.5368133187294006, "learning_rate": 2.3489860248011136e-05, "loss": 0.0324, "step": 69400 }, { "grad_norm": 0.3192571997642517, "learning_rate": 2.3475842409049464e-05, "loss": 0.0369, "step": 69410 }, { "grad_norm": 0.5134454965591431, "learning_rate": 2.346182747072975e-05, "loss": 0.043, "step": 69420 }, { "grad_norm": 0.3724326193332672, "learning_rate": 2.3447815434584718e-05, "loss": 0.0288, "step": 69430 }, { "grad_norm": 0.6550209522247314, "learning_rate": 2.3433806302146656e-05, "loss": 0.0297, "step": 69440 }, { "grad_norm": 0.4963982105255127, "learning_rate": 2.3419800074947612e-05, "loss": 0.0374, "step": 69450 }, { "grad_norm": 0.5095170140266418, "learning_rate": 2.3405796754519248e-05, "loss": 0.0387, "step": 69460 }, { "grad_norm": 0.6811015009880066, "learning_rate": 2.3391796342392992e-05, "loss": 0.0421, "step": 69470 }, { "grad_norm": 0.49907997250556946, "learning_rate": 2.337779884009984e-05, "loss": 0.0385, "step": 69480 }, { "grad_norm": 0.2846969664096832, "learning_rate": 2.3363804249170624e-05, "loss": 0.0375, "step": 69490 }, { "grad_norm": 0.43692970275878906, "learning_rate": 2.3349812571135703e-05, "loss": 0.0337, "step": 69500 }, { "grad_norm": 0.3240213692188263, "learning_rate": 2.333582380752523e-05, "loss": 0.0273, "step": 69510 }, { "grad_norm": 0.3862476944923401, "learning_rate": 2.3321837959868963e-05, "loss": 0.0296, "step": 69520 }, { "grad_norm": 0.5559130907058716, "learning_rate": 2.3307855029696383e-05, "loss": 0.0303, "step": 69530 }, { "grad_norm": 0.5049617886543274, "learning_rate": 2.3293875018536648e-05, "loss": 0.0362, "step": 69540 }, { "grad_norm": 0.3090835511684418, "learning_rate": 2.327989792791861e-05, "loss": 0.0293, "step": 69550 }, { "grad_norm": 0.6896260976791382, "learning_rate": 2.326592375937074e-05, "loss": 0.034, "step": 69560 }, { "grad_norm": 0.42168739438056946, "learning_rate": 2.3251952514421248e-05, "loss": 0.0268, "step": 69570 }, { "grad_norm": 0.5092635750770569, "learning_rate": 2.323798419459804e-05, "loss": 0.035, "step": 69580 }, { "grad_norm": 0.4110814929008484, "learning_rate": 2.3224018801428625e-05, "loss": 0.0301, "step": 69590 }, { "grad_norm": 0.4089163839817047, "learning_rate": 2.3210056336440273e-05, "loss": 0.0265, "step": 69600 }, { "grad_norm": 0.3835153579711914, "learning_rate": 2.3196096801159844e-05, "loss": 0.0321, "step": 69610 }, { "grad_norm": 0.4215521812438965, "learning_rate": 2.3182140197114012e-05, "loss": 0.0347, "step": 69620 }, { "grad_norm": 0.5906819701194763, "learning_rate": 2.3168186525828985e-05, "loss": 0.0435, "step": 69630 }, { "grad_norm": 0.4083251655101776, "learning_rate": 2.3154235788830757e-05, "loss": 0.0322, "step": 69640 }, { "grad_norm": 0.44758784770965576, "learning_rate": 2.3140287987644927e-05, "loss": 0.0395, "step": 69650 }, { "grad_norm": 0.26670515537261963, "learning_rate": 2.3126343123796833e-05, "loss": 0.0302, "step": 69660 }, { "grad_norm": 0.45088711380958557, "learning_rate": 2.3112401198811418e-05, "loss": 0.0323, "step": 69670 }, { "grad_norm": 0.4977339208126068, "learning_rate": 2.3098462214213418e-05, "loss": 0.0389, "step": 69680 }, { "grad_norm": 0.5359702110290527, "learning_rate": 2.308452617152712e-05, "loss": 0.0347, "step": 69690 }, { "grad_norm": 0.6573030948638916, "learning_rate": 2.3070593072276603e-05, "loss": 0.0383, "step": 69700 }, { "grad_norm": 0.37533941864967346, "learning_rate": 2.3056662917985518e-05, "loss": 0.0297, "step": 69710 }, { "grad_norm": 0.2792300879955292, "learning_rate": 2.3042735710177283e-05, "loss": 0.0425, "step": 69720 }, { "grad_norm": 0.3062328100204468, "learning_rate": 2.302881145037492e-05, "loss": 0.0276, "step": 69730 }, { "grad_norm": 0.25411686301231384, "learning_rate": 2.3014890140101192e-05, "loss": 0.0312, "step": 69740 }, { "grad_norm": 0.37684541940689087, "learning_rate": 2.30009717808785e-05, "loss": 0.0283, "step": 69750 }, { "grad_norm": 0.27796655893325806, "learning_rate": 2.2987056374228967e-05, "loss": 0.0367, "step": 69760 }, { "grad_norm": 0.3677716851234436, "learning_rate": 2.297314392167431e-05, "loss": 0.0376, "step": 69770 }, { "grad_norm": 0.32509714365005493, "learning_rate": 2.295923442473601e-05, "loss": 0.0365, "step": 69780 }, { "grad_norm": 0.38721901178359985, "learning_rate": 2.294532788493519e-05, "loss": 0.0277, "step": 69790 }, { "grad_norm": 0.5196292400360107, "learning_rate": 2.293142430379262e-05, "loss": 0.0321, "step": 69800 }, { "grad_norm": 1.4917242527008057, "learning_rate": 2.291752368282879e-05, "loss": 0.0391, "step": 69810 }, { "grad_norm": 0.34731942415237427, "learning_rate": 2.290362602356385e-05, "loss": 0.0325, "step": 69820 }, { "grad_norm": 0.7228786945343018, "learning_rate": 2.288973132751765e-05, "loss": 0.0433, "step": 69830 }, { "grad_norm": 0.4684847295284271, "learning_rate": 2.287583959620965e-05, "loss": 0.0427, "step": 69840 }, { "grad_norm": 0.3125835061073303, "learning_rate": 2.2861950831159074e-05, "loss": 0.0311, "step": 69850 }, { "grad_norm": 0.3544423580169678, "learning_rate": 2.284806503388472e-05, "loss": 0.0301, "step": 69860 }, { "grad_norm": 0.47320619225502014, "learning_rate": 2.283418220590518e-05, "loss": 0.032, "step": 69870 }, { "grad_norm": 0.43213018774986267, "learning_rate": 2.282030234873862e-05, "loss": 0.0414, "step": 69880 }, { "grad_norm": 0.4548986256122589, "learning_rate": 2.2806425463902948e-05, "loss": 0.023, "step": 69890 }, { "grad_norm": 0.4024406969547272, "learning_rate": 2.279255155291568e-05, "loss": 0.0376, "step": 69900 }, { "grad_norm": 0.2619764804840088, "learning_rate": 2.277868061729409e-05, "loss": 0.027, "step": 69910 }, { "grad_norm": 0.5399813652038574, "learning_rate": 2.276481265855504e-05, "loss": 0.0252, "step": 69920 }, { "grad_norm": 0.3381878435611725, "learning_rate": 2.2750947678215128e-05, "loss": 0.0302, "step": 69930 }, { "grad_norm": 0.6016213893890381, "learning_rate": 2.2737085677790616e-05, "loss": 0.0323, "step": 69940 }, { "grad_norm": 0.4594634175300598, "learning_rate": 2.2723226658797443e-05, "loss": 0.0332, "step": 69950 }, { "grad_norm": 0.38397741317749023, "learning_rate": 2.270937062275117e-05, "loss": 0.0282, "step": 69960 }, { "grad_norm": 0.23841045796871185, "learning_rate": 2.2695517571167113e-05, "loss": 0.0264, "step": 69970 }, { "grad_norm": 0.40969520807266235, "learning_rate": 2.268166750556019e-05, "loss": 0.037, "step": 69980 }, { "grad_norm": 0.26844874024391174, "learning_rate": 2.266782042744503e-05, "loss": 0.026, "step": 69990 }, { "grad_norm": 0.4330791234970093, "learning_rate": 2.2653976338335935e-05, "loss": 0.037, "step": 70000 }, { "grad_norm": 0.5439633727073669, "learning_rate": 2.2640135239746895e-05, "loss": 0.0397, "step": 70010 }, { "grad_norm": 0.6945688724517822, "learning_rate": 2.2626297133191515e-05, "loss": 0.0299, "step": 70020 }, { "grad_norm": 0.5424614548683167, "learning_rate": 2.261246202018312e-05, "loss": 0.0358, "step": 70030 }, { "grad_norm": 0.7063108086585999, "learning_rate": 2.2598629902234718e-05, "loss": 0.0287, "step": 70040 }, { "grad_norm": 0.31798872351646423, "learning_rate": 2.258480078085894e-05, "loss": 0.0354, "step": 70050 }, { "grad_norm": 0.7970062494277954, "learning_rate": 2.2570974657568126e-05, "loss": 0.0362, "step": 70060 }, { "grad_norm": 0.2846689224243164, "learning_rate": 2.2557151533874286e-05, "loss": 0.0338, "step": 70070 }, { "grad_norm": 0.3586626350879669, "learning_rate": 2.2543331411289116e-05, "loss": 0.0301, "step": 70080 }, { "grad_norm": 0.47687047719955444, "learning_rate": 2.252951429132392e-05, "loss": 0.0454, "step": 70090 }, { "grad_norm": 0.4414891302585602, "learning_rate": 2.2515700175489762e-05, "loss": 0.0405, "step": 70100 }, { "grad_norm": 0.511324405670166, "learning_rate": 2.2501889065297288e-05, "loss": 0.0319, "step": 70110 }, { "grad_norm": 0.33498427271842957, "learning_rate": 2.2488080962256887e-05, "loss": 0.0353, "step": 70120 }, { "grad_norm": 0.43388548493385315, "learning_rate": 2.2474275867878585e-05, "loss": 0.0464, "step": 70130 }, { "grad_norm": 0.4698392152786255, "learning_rate": 2.246047378367211e-05, "loss": 0.0494, "step": 70140 }, { "grad_norm": 0.29654553532600403, "learning_rate": 2.2446674711146798e-05, "loss": 0.032, "step": 70150 }, { "grad_norm": 0.43905577063560486, "learning_rate": 2.2432878651811734e-05, "loss": 0.0335, "step": 70160 }, { "grad_norm": 0.3439559042453766, "learning_rate": 2.2419085607175594e-05, "loss": 0.0386, "step": 70170 }, { "grad_norm": 0.35044562816619873, "learning_rate": 2.240529557874678e-05, "loss": 0.0331, "step": 70180 }, { "grad_norm": 0.5832408666610718, "learning_rate": 2.2391508568033358e-05, "loss": 0.0332, "step": 70190 }, { "grad_norm": 0.3710576593875885, "learning_rate": 2.237772457654307e-05, "loss": 0.0398, "step": 70200 }, { "grad_norm": 0.3082684874534607, "learning_rate": 2.236394360578327e-05, "loss": 0.0376, "step": 70210 }, { "grad_norm": 0.4437499940395355, "learning_rate": 2.2350165657261073e-05, "loss": 0.0353, "step": 70220 }, { "grad_norm": 0.5534912943840027, "learning_rate": 2.2336390732483164e-05, "loss": 0.0377, "step": 70230 }, { "grad_norm": 0.6631354689598083, "learning_rate": 2.2322618832955976e-05, "loss": 0.0472, "step": 70240 }, { "grad_norm": 0.61593097448349, "learning_rate": 2.2308849960185578e-05, "loss": 0.0365, "step": 70250 }, { "grad_norm": 0.3797697424888611, "learning_rate": 2.229508411567774e-05, "loss": 0.0335, "step": 70260 }, { "grad_norm": 0.4723663628101349, "learning_rate": 2.2281321300937823e-05, "loss": 0.044, "step": 70270 }, { "grad_norm": 0.3102642297744751, "learning_rate": 2.2267561517470937e-05, "loss": 0.031, "step": 70280 }, { "grad_norm": 0.6614684462547302, "learning_rate": 2.2253804766781845e-05, "loss": 0.0336, "step": 70290 }, { "grad_norm": 0.4631696343421936, "learning_rate": 2.2240051050374933e-05, "loss": 0.0371, "step": 70300 }, { "grad_norm": 0.5222879648208618, "learning_rate": 2.22263003697543e-05, "loss": 0.035, "step": 70310 }, { "grad_norm": 0.3281940221786499, "learning_rate": 2.2212552726423702e-05, "loss": 0.035, "step": 70320 }, { "grad_norm": 0.31549957394599915, "learning_rate": 2.2198808121886578e-05, "loss": 0.0369, "step": 70330 }, { "grad_norm": 0.3932192921638489, "learning_rate": 2.218506655764598e-05, "loss": 0.0327, "step": 70340 }, { "grad_norm": 1.7357263565063477, "learning_rate": 2.2171328035204704e-05, "loss": 0.0307, "step": 70350 }, { "grad_norm": 0.3798592686653137, "learning_rate": 2.2157592556065133e-05, "loss": 0.0284, "step": 70360 }, { "grad_norm": 0.40497657656669617, "learning_rate": 2.2143860121729382e-05, "loss": 0.029, "step": 70370 }, { "grad_norm": 0.29712745547294617, "learning_rate": 2.2130130733699206e-05, "loss": 0.0344, "step": 70380 }, { "grad_norm": 0.37818044424057007, "learning_rate": 2.2116404393476052e-05, "loss": 0.0364, "step": 70390 }, { "grad_norm": 2.3358891010284424, "learning_rate": 2.2102681102560967e-05, "loss": 0.0351, "step": 70400 }, { "grad_norm": 0.6965969800949097, "learning_rate": 2.208896086245476e-05, "loss": 0.0303, "step": 70410 }, { "grad_norm": 0.3260401487350464, "learning_rate": 2.2075243674657808e-05, "loss": 0.0352, "step": 70420 }, { "grad_norm": 0.36788785457611084, "learning_rate": 2.206152954067024e-05, "loss": 0.027, "step": 70430 }, { "grad_norm": 0.3169759213924408, "learning_rate": 2.2047818461991758e-05, "loss": 0.0357, "step": 70440 }, { "grad_norm": 0.3675026595592499, "learning_rate": 2.2034110440121864e-05, "loss": 0.0339, "step": 70450 }, { "grad_norm": 0.36884286999702454, "learning_rate": 2.2020405476559585e-05, "loss": 0.0304, "step": 70460 }, { "grad_norm": 0.4554636776447296, "learning_rate": 2.2006703572803722e-05, "loss": 0.0365, "step": 70470 }, { "grad_norm": 0.2271479368209839, "learning_rate": 2.1993004730352646e-05, "loss": 0.0377, "step": 70480 }, { "grad_norm": 0.29923444986343384, "learning_rate": 2.1979308950704464e-05, "loss": 0.0255, "step": 70490 }, { "grad_norm": 0.304526686668396, "learning_rate": 2.1965616235356923e-05, "loss": 0.0427, "step": 70500 }, { "grad_norm": 0.35200008749961853, "learning_rate": 2.1951926585807465e-05, "loss": 0.029, "step": 70510 }, { "grad_norm": 0.2696358859539032, "learning_rate": 2.193824000355313e-05, "loss": 0.0353, "step": 70520 }, { "grad_norm": 0.5037028193473816, "learning_rate": 2.192455649009067e-05, "loss": 0.0423, "step": 70530 }, { "grad_norm": 0.5601984262466431, "learning_rate": 2.1910876046916523e-05, "loss": 0.0347, "step": 70540 }, { "grad_norm": 0.18047696352005005, "learning_rate": 2.1897198675526725e-05, "loss": 0.0288, "step": 70550 }, { "grad_norm": 0.41046324372291565, "learning_rate": 2.1883524377417024e-05, "loss": 0.0303, "step": 70560 }, { "grad_norm": 0.37494033575057983, "learning_rate": 2.1869853154082825e-05, "loss": 0.0324, "step": 70570 }, { "grad_norm": 0.3692529797554016, "learning_rate": 2.1856185007019215e-05, "loss": 0.0374, "step": 70580 }, { "grad_norm": 0.35389721393585205, "learning_rate": 2.1842519937720874e-05, "loss": 0.0404, "step": 70590 }, { "grad_norm": 0.47214075922966003, "learning_rate": 2.182885794768224e-05, "loss": 0.0484, "step": 70600 }, { "grad_norm": 0.39974498748779297, "learning_rate": 2.1815199038397326e-05, "loss": 0.0367, "step": 70610 }, { "grad_norm": 0.9678022861480713, "learning_rate": 2.1801543211359893e-05, "loss": 0.0335, "step": 70620 }, { "grad_norm": 0.40549930930137634, "learning_rate": 2.178789046806326e-05, "loss": 0.0373, "step": 70630 }, { "grad_norm": 1.0643454790115356, "learning_rate": 2.1774240810000547e-05, "loss": 0.0335, "step": 70640 }, { "grad_norm": 0.5629715919494629, "learning_rate": 2.17605942386644e-05, "loss": 0.0473, "step": 70650 }, { "grad_norm": 0.4299582242965698, "learning_rate": 2.174695075554723e-05, "loss": 0.0488, "step": 70660 }, { "grad_norm": 0.44156867265701294, "learning_rate": 2.173331036214103e-05, "loss": 0.0431, "step": 70670 }, { "grad_norm": 0.5417277812957764, "learning_rate": 2.171967305993753e-05, "loss": 0.0328, "step": 70680 }, { "grad_norm": 0.6890495419502258, "learning_rate": 2.1706038850428024e-05, "loss": 0.0361, "step": 70690 }, { "grad_norm": 0.5435839295387268, "learning_rate": 2.169240773510361e-05, "loss": 0.0278, "step": 70700 }, { "grad_norm": 0.48682862520217896, "learning_rate": 2.1678779715454906e-05, "loss": 0.0391, "step": 70710 }, { "grad_norm": 0.3444499969482422, "learning_rate": 2.166515479297229e-05, "loss": 0.032, "step": 70720 }, { "grad_norm": 0.41234439611434937, "learning_rate": 2.1651532969145732e-05, "loss": 0.0291, "step": 70730 }, { "grad_norm": 0.2729909121990204, "learning_rate": 2.1637914245464903e-05, "loss": 0.0367, "step": 70740 }, { "grad_norm": 0.3815028965473175, "learning_rate": 2.1624298623419133e-05, "loss": 0.0285, "step": 70750 }, { "grad_norm": 0.45609748363494873, "learning_rate": 2.1610686104497413e-05, "loss": 0.0368, "step": 70760 }, { "grad_norm": 0.5534377098083496, "learning_rate": 2.1597076690188366e-05, "loss": 0.0403, "step": 70770 }, { "grad_norm": 0.3991175591945648, "learning_rate": 2.1583470381980315e-05, "loss": 0.0386, "step": 70780 }, { "grad_norm": 0.47713756561279297, "learning_rate": 2.156986718136123e-05, "loss": 0.0384, "step": 70790 }, { "grad_norm": 0.33436349034309387, "learning_rate": 2.155626708981871e-05, "loss": 0.0284, "step": 70800 }, { "grad_norm": 0.5648617148399353, "learning_rate": 2.1542670108840075e-05, "loss": 0.0352, "step": 70810 }, { "grad_norm": 0.571382999420166, "learning_rate": 2.152907623991221e-05, "loss": 0.0388, "step": 70820 }, { "grad_norm": 0.29920104146003723, "learning_rate": 2.1515485484521802e-05, "loss": 0.033, "step": 70830 }, { "grad_norm": 0.229373961687088, "learning_rate": 2.150189784415506e-05, "loss": 0.0289, "step": 70840 }, { "grad_norm": 0.25205543637275696, "learning_rate": 2.148831332029794e-05, "loss": 0.0274, "step": 70850 }, { "grad_norm": 0.38372236490249634, "learning_rate": 2.147473191443599e-05, "loss": 0.0285, "step": 70860 }, { "grad_norm": 0.23575808107852936, "learning_rate": 2.1461153628054492e-05, "loss": 0.0258, "step": 70870 }, { "grad_norm": 0.2639261782169342, "learning_rate": 2.1447578462638284e-05, "loss": 0.0395, "step": 70880 }, { "grad_norm": 0.6811462640762329, "learning_rate": 2.143400641967201e-05, "loss": 0.0261, "step": 70890 }, { "grad_norm": 0.2796837091445923, "learning_rate": 2.1420437500639823e-05, "loss": 0.0252, "step": 70900 }, { "grad_norm": 0.9662849307060242, "learning_rate": 2.1406871707025646e-05, "loss": 0.0368, "step": 70910 }, { "grad_norm": 0.5007302761077881, "learning_rate": 2.1393309040312963e-05, "loss": 0.0449, "step": 70920 }, { "grad_norm": 0.29466110467910767, "learning_rate": 2.137974950198502e-05, "loss": 0.033, "step": 70930 }, { "grad_norm": 0.40898239612579346, "learning_rate": 2.1366193093524617e-05, "loss": 0.026, "step": 70940 }, { "grad_norm": 0.4488505721092224, "learning_rate": 2.1352639816414288e-05, "loss": 0.0315, "step": 70950 }, { "grad_norm": 0.2642219066619873, "learning_rate": 2.1339089672136202e-05, "loss": 0.0322, "step": 70960 }, { "grad_norm": 0.23511724174022675, "learning_rate": 2.1325542662172193e-05, "loss": 0.0323, "step": 70970 }, { "grad_norm": 0.403952419757843, "learning_rate": 2.1311998788003712e-05, "loss": 0.027, "step": 70980 }, { "grad_norm": 0.43745940923690796, "learning_rate": 2.1298458051111907e-05, "loss": 0.0385, "step": 70990 }, { "grad_norm": 0.7181980609893799, "learning_rate": 2.1284920452977602e-05, "loss": 0.0284, "step": 71000 }, { "grad_norm": 0.46554145216941833, "learning_rate": 2.12713859950812e-05, "loss": 0.0339, "step": 71010 }, { "grad_norm": 0.3328849673271179, "learning_rate": 2.1257854678902832e-05, "loss": 0.0295, "step": 71020 }, { "grad_norm": 0.4982967674732208, "learning_rate": 2.124432650592227e-05, "loss": 0.0407, "step": 71030 }, { "grad_norm": 0.5547215938568115, "learning_rate": 2.1230801477618937e-05, "loss": 0.0406, "step": 71040 }, { "grad_norm": 0.41384148597717285, "learning_rate": 2.121727959547189e-05, "loss": 0.0325, "step": 71050 }, { "grad_norm": 0.2604469954967499, "learning_rate": 2.1203760860959886e-05, "loss": 0.0293, "step": 71060 }, { "grad_norm": 0.6442835927009583, "learning_rate": 2.119024527556127e-05, "loss": 0.0397, "step": 71070 }, { "grad_norm": 0.49408307671546936, "learning_rate": 2.117673284075415e-05, "loss": 0.0262, "step": 71080 }, { "grad_norm": 0.3571600615978241, "learning_rate": 2.1163223558016165e-05, "loss": 0.0286, "step": 71090 }, { "grad_norm": 0.3733293414115906, "learning_rate": 2.1149717428824722e-05, "loss": 0.0341, "step": 71100 }, { "grad_norm": 0.552108108997345, "learning_rate": 2.1136214454656784e-05, "loss": 0.0396, "step": 71110 }, { "grad_norm": 1.1567044258117676, "learning_rate": 2.1122714636989056e-05, "loss": 0.0351, "step": 71120 }, { "grad_norm": 0.8707807660102844, "learning_rate": 2.1109217977297823e-05, "loss": 0.0394, "step": 71130 }, { "grad_norm": 0.3637327253818512, "learning_rate": 2.1095724477059077e-05, "loss": 0.025, "step": 71140 }, { "grad_norm": 0.37388598918914795, "learning_rate": 2.1082234137748434e-05, "loss": 0.022, "step": 71150 }, { "grad_norm": 0.37528079748153687, "learning_rate": 2.106874696084122e-05, "loss": 0.0382, "step": 71160 }, { "grad_norm": 0.42413756251335144, "learning_rate": 2.1055262947812315e-05, "loss": 0.0378, "step": 71170 }, { "grad_norm": 0.33006927371025085, "learning_rate": 2.104178210013636e-05, "loss": 0.0493, "step": 71180 }, { "grad_norm": 0.5768751502037048, "learning_rate": 2.1028304419287554e-05, "loss": 0.0385, "step": 71190 }, { "grad_norm": 0.4751824140548706, "learning_rate": 2.101482990673982e-05, "loss": 0.0338, "step": 71200 }, { "grad_norm": 0.3860228359699249, "learning_rate": 2.1001358563966707e-05, "loss": 0.0388, "step": 71210 }, { "grad_norm": 0.6503908038139343, "learning_rate": 2.098789039244144e-05, "loss": 0.0301, "step": 71220 }, { "grad_norm": 0.3316677510738373, "learning_rate": 2.097442539363685e-05, "loss": 0.0291, "step": 71230 }, { "grad_norm": 0.3501395285129547, "learning_rate": 2.0960963569025448e-05, "loss": 0.0285, "step": 71240 }, { "grad_norm": 0.24410304427146912, "learning_rate": 2.0947504920079435e-05, "loss": 0.0254, "step": 71250 }, { "grad_norm": 0.3094714879989624, "learning_rate": 2.093404944827059e-05, "loss": 0.029, "step": 71260 }, { "grad_norm": 0.39068177342414856, "learning_rate": 2.0920597155070394e-05, "loss": 0.037, "step": 71270 }, { "grad_norm": 0.7310323119163513, "learning_rate": 2.090714804194997e-05, "loss": 0.0419, "step": 71280 }, { "grad_norm": 0.21089914441108704, "learning_rate": 2.0893702110380114e-05, "loss": 0.029, "step": 71290 }, { "grad_norm": 0.4810129702091217, "learning_rate": 2.0880259361831216e-05, "loss": 0.0341, "step": 71300 }, { "grad_norm": 0.4165888726711273, "learning_rate": 2.0866819797773397e-05, "loss": 0.0315, "step": 71310 }, { "grad_norm": 0.30267030000686646, "learning_rate": 2.085338341967634e-05, "loss": 0.0322, "step": 71320 }, { "grad_norm": 0.4452109932899475, "learning_rate": 2.083995022900946e-05, "loss": 0.0313, "step": 71330 }, { "grad_norm": 0.5178880095481873, "learning_rate": 2.0826520227241768e-05, "loss": 0.0291, "step": 71340 }, { "grad_norm": 0.6268301606178284, "learning_rate": 2.081309341584199e-05, "loss": 0.0367, "step": 71350 }, { "grad_norm": 0.2543618679046631, "learning_rate": 2.0799669796278413e-05, "loss": 0.0295, "step": 71360 }, { "grad_norm": 0.5253681540489197, "learning_rate": 2.078624937001906e-05, "loss": 0.0429, "step": 71370 }, { "grad_norm": 0.6364444494247437, "learning_rate": 2.0772832138531544e-05, "loss": 0.0615, "step": 71380 }, { "grad_norm": 0.37905213236808777, "learning_rate": 2.0759418103283158e-05, "loss": 0.0319, "step": 71390 }, { "grad_norm": 0.7938554286956787, "learning_rate": 2.0746007265740846e-05, "loss": 0.0324, "step": 71400 }, { "grad_norm": 0.6452445387840271, "learning_rate": 2.0732599627371215e-05, "loss": 0.0399, "step": 71410 }, { "grad_norm": 0.3462200462818146, "learning_rate": 2.0719195189640468e-05, "loss": 0.0274, "step": 71420 }, { "grad_norm": 0.7186207175254822, "learning_rate": 2.0705793954014528e-05, "loss": 0.0368, "step": 71430 }, { "grad_norm": 0.31263816356658936, "learning_rate": 2.06923959219589e-05, "loss": 0.0269, "step": 71440 }, { "grad_norm": 0.8129107356071472, "learning_rate": 2.0679001094938793e-05, "loss": 0.0288, "step": 71450 }, { "grad_norm": 0.35424086451530457, "learning_rate": 2.066560947441904e-05, "loss": 0.0331, "step": 71460 }, { "grad_norm": 0.4433184266090393, "learning_rate": 2.0652221061864152e-05, "loss": 0.0257, "step": 71470 }, { "grad_norm": 0.42782774567604065, "learning_rate": 2.0638835858738226e-05, "loss": 0.0372, "step": 71480 }, { "grad_norm": 0.324945867061615, "learning_rate": 2.0625453866505068e-05, "loss": 0.0281, "step": 71490 }, { "grad_norm": 0.34904399514198303, "learning_rate": 2.0612075086628123e-05, "loss": 0.0316, "step": 71500 }, { "grad_norm": 0.44919297099113464, "learning_rate": 2.059869952057045e-05, "loss": 0.0319, "step": 71510 }, { "grad_norm": 0.5237553715705872, "learning_rate": 2.0585327169794793e-05, "loss": 0.033, "step": 71520 }, { "grad_norm": 0.38648736476898193, "learning_rate": 2.0571958035763534e-05, "loss": 0.0288, "step": 71530 }, { "grad_norm": 0.5121671557426453, "learning_rate": 2.0558592119938715e-05, "loss": 0.0354, "step": 71540 }, { "grad_norm": 0.22615888714790344, "learning_rate": 2.054522942378198e-05, "loss": 0.0231, "step": 71550 }, { "grad_norm": 0.8045787811279297, "learning_rate": 2.053186994875469e-05, "loss": 0.0399, "step": 71560 }, { "grad_norm": 0.4593663215637207, "learning_rate": 2.0518513696317775e-05, "loss": 0.0408, "step": 71570 }, { "grad_norm": 0.3589583933353424, "learning_rate": 2.050516066793188e-05, "loss": 0.0313, "step": 71580 }, { "grad_norm": 0.28906261920928955, "learning_rate": 2.0491810865057276e-05, "loss": 0.0304, "step": 71590 }, { "grad_norm": 0.5990573167800903, "learning_rate": 2.0478464289153882e-05, "loss": 0.0325, "step": 71600 }, { "grad_norm": 0.6023444533348083, "learning_rate": 2.0465120941681232e-05, "loss": 0.032, "step": 71610 }, { "grad_norm": 0.392555832862854, "learning_rate": 2.0451780824098566e-05, "loss": 0.0329, "step": 71620 }, { "grad_norm": 0.2515050768852234, "learning_rate": 2.043844393786471e-05, "loss": 0.032, "step": 71630 }, { "grad_norm": 1.2806040048599243, "learning_rate": 2.0425110284438193e-05, "loss": 0.034, "step": 71640 }, { "grad_norm": 0.29124927520751953, "learning_rate": 2.041177986527712e-05, "loss": 0.0313, "step": 71650 }, { "grad_norm": 0.6285495758056641, "learning_rate": 2.0398452681839347e-05, "loss": 0.0297, "step": 71660 }, { "grad_norm": 0.3864491283893585, "learning_rate": 2.0385128735582272e-05, "loss": 0.0441, "step": 71670 }, { "grad_norm": 0.34091058373451233, "learning_rate": 2.0371808027963003e-05, "loss": 0.0303, "step": 71680 }, { "grad_norm": 0.5879014134407043, "learning_rate": 2.035849056043825e-05, "loss": 0.0368, "step": 71690 }, { "grad_norm": 0.6453394889831543, "learning_rate": 2.03451763344644e-05, "loss": 0.0377, "step": 71700 }, { "grad_norm": 0.5645972490310669, "learning_rate": 2.0331865351497482e-05, "loss": 0.0408, "step": 71710 }, { "grad_norm": 0.5218676924705505, "learning_rate": 2.0318557612993185e-05, "loss": 0.0335, "step": 71720 }, { "grad_norm": 0.37916749715805054, "learning_rate": 2.0305253120406782e-05, "loss": 0.0308, "step": 71730 }, { "grad_norm": 0.34139716625213623, "learning_rate": 2.029195187519325e-05, "loss": 0.0329, "step": 71740 }, { "grad_norm": 0.3714253902435303, "learning_rate": 2.0278653878807217e-05, "loss": 0.0304, "step": 71750 }, { "grad_norm": 0.45108717679977417, "learning_rate": 2.0265359132702892e-05, "loss": 0.0399, "step": 71760 }, { "grad_norm": 0.36148619651794434, "learning_rate": 2.0252067638334182e-05, "loss": 0.0286, "step": 71770 }, { "grad_norm": 0.5303459167480469, "learning_rate": 2.023877939715464e-05, "loss": 0.0456, "step": 71780 }, { "grad_norm": 0.5446840524673462, "learning_rate": 2.0225494410617448e-05, "loss": 0.0399, "step": 71790 }, { "grad_norm": 0.27581584453582764, "learning_rate": 2.02122126801754e-05, "loss": 0.0367, "step": 71800 }, { "grad_norm": 0.34562355279922485, "learning_rate": 2.0198934207281018e-05, "loss": 0.0387, "step": 71810 }, { "grad_norm": 0.47262707352638245, "learning_rate": 2.0185658993386357e-05, "loss": 0.0358, "step": 71820 }, { "grad_norm": 0.5069800019264221, "learning_rate": 2.0172387039943226e-05, "loss": 0.0324, "step": 71830 }, { "grad_norm": 0.19941329956054688, "learning_rate": 2.0159118348402968e-05, "loss": 0.0391, "step": 71840 }, { "grad_norm": 2.042872905731201, "learning_rate": 2.0145852920216697e-05, "loss": 0.0302, "step": 71850 }, { "grad_norm": 0.530582845211029, "learning_rate": 2.0132590756835045e-05, "loss": 0.0337, "step": 71860 }, { "grad_norm": 0.30329468846321106, "learning_rate": 2.0119331859708386e-05, "loss": 0.0472, "step": 71870 }, { "grad_norm": 0.8272104263305664, "learning_rate": 2.0106076230286654e-05, "loss": 0.0337, "step": 71880 }, { "grad_norm": 0.4467805027961731, "learning_rate": 2.0092823870019488e-05, "loss": 0.0404, "step": 71890 }, { "grad_norm": 0.2924744486808777, "learning_rate": 2.0079574780356115e-05, "loss": 0.0295, "step": 71900 }, { "grad_norm": 0.4517669081687927, "learning_rate": 2.0066328962745486e-05, "loss": 0.0366, "step": 71910 }, { "grad_norm": 0.288533091545105, "learning_rate": 2.00530864186361e-05, "loss": 0.0261, "step": 71920 }, { "grad_norm": 0.5328848361968994, "learning_rate": 2.003984714947618e-05, "loss": 0.0435, "step": 71930 }, { "grad_norm": 0.3140541613101959, "learning_rate": 2.0026611156713515e-05, "loss": 0.0378, "step": 71940 }, { "grad_norm": 0.41959795355796814, "learning_rate": 2.0013378441795584e-05, "loss": 0.0337, "step": 71950 }, { "grad_norm": 0.4090059995651245, "learning_rate": 2.0000149006169506e-05, "loss": 0.0294, "step": 71960 }, { "grad_norm": 0.43775108456611633, "learning_rate": 1.9986922851282048e-05, "loss": 0.0339, "step": 71970 }, { "grad_norm": 0.4241075813770294, "learning_rate": 1.997369997857956e-05, "loss": 0.0325, "step": 71980 }, { "grad_norm": 0.4502977728843689, "learning_rate": 1.9960480389508103e-05, "loss": 0.0321, "step": 71990 }, { "grad_norm": 0.40848052501678467, "learning_rate": 1.9947264085513363e-05, "loss": 0.0463, "step": 72000 }, { "grad_norm": 0.4194602966308594, "learning_rate": 1.993405106804062e-05, "loss": 0.0354, "step": 72010 }, { "grad_norm": 0.6761909127235413, "learning_rate": 1.9920841338534867e-05, "loss": 0.0415, "step": 72020 }, { "grad_norm": 0.41013002395629883, "learning_rate": 1.9907634898440646e-05, "loss": 0.0266, "step": 72030 }, { "grad_norm": 0.2874845564365387, "learning_rate": 1.989443174920227e-05, "loss": 0.0292, "step": 72040 }, { "grad_norm": 0.5560673475265503, "learning_rate": 1.988123189226355e-05, "loss": 0.0434, "step": 72050 }, { "grad_norm": 0.9677086472511292, "learning_rate": 1.986803532906805e-05, "loss": 0.0347, "step": 72060 }, { "grad_norm": 0.3647485375404358, "learning_rate": 1.985484206105889e-05, "loss": 0.0345, "step": 72070 }, { "grad_norm": 0.3057914674282074, "learning_rate": 1.9841652089678887e-05, "loss": 0.0303, "step": 72080 }, { "grad_norm": 0.2744540274143219, "learning_rate": 1.9828465416370435e-05, "loss": 0.0443, "step": 72090 }, { "grad_norm": 0.37029364705085754, "learning_rate": 1.9815282042575685e-05, "loss": 0.0341, "step": 72100 }, { "grad_norm": 0.2712135314941406, "learning_rate": 1.9802101969736285e-05, "loss": 0.0427, "step": 72110 }, { "grad_norm": 0.3348310589790344, "learning_rate": 1.9788925199293633e-05, "loss": 0.0347, "step": 72120 }, { "grad_norm": 0.2817079722881317, "learning_rate": 1.977575173268868e-05, "loss": 0.0311, "step": 72130 }, { "grad_norm": 0.5645169615745544, "learning_rate": 1.9762581571362094e-05, "loss": 0.0366, "step": 72140 }, { "grad_norm": 0.8839913010597229, "learning_rate": 1.9749414716754112e-05, "loss": 0.0385, "step": 72150 }, { "grad_norm": 0.2879522442817688, "learning_rate": 1.973625117030465e-05, "loss": 0.0313, "step": 72160 }, { "grad_norm": 0.27774107456207275, "learning_rate": 1.972309093345326e-05, "loss": 0.0284, "step": 72170 }, { "grad_norm": 0.4423077404499054, "learning_rate": 1.970993400763915e-05, "loss": 0.0362, "step": 72180 }, { "grad_norm": 0.664219319820404, "learning_rate": 1.969678039430109e-05, "loss": 0.0426, "step": 72190 }, { "grad_norm": 0.5924000144004822, "learning_rate": 1.9683630094877564e-05, "loss": 0.0298, "step": 72200 }, { "grad_norm": 0.3289203643798828, "learning_rate": 1.9670483110806696e-05, "loss": 0.0321, "step": 72210 }, { "grad_norm": 0.6575362086296082, "learning_rate": 1.9657339443526172e-05, "loss": 0.0374, "step": 72220 }, { "grad_norm": 0.22045868635177612, "learning_rate": 1.964419909447339e-05, "loss": 0.0381, "step": 72230 }, { "grad_norm": 0.5436654686927795, "learning_rate": 1.963106206508536e-05, "loss": 0.0359, "step": 72240 }, { "grad_norm": 0.35343241691589355, "learning_rate": 1.9617928356798738e-05, "loss": 0.0295, "step": 72250 }, { "grad_norm": 0.3376626670360565, "learning_rate": 1.960479797104977e-05, "loss": 0.0317, "step": 72260 }, { "grad_norm": 1.1395035982131958, "learning_rate": 1.9591670909274418e-05, "loss": 0.0365, "step": 72270 }, { "grad_norm": 0.3213244378566742, "learning_rate": 1.9578547172908184e-05, "loss": 0.034, "step": 72280 }, { "grad_norm": 0.8315117359161377, "learning_rate": 1.9565426763386325e-05, "loss": 0.036, "step": 72290 }, { "grad_norm": 0.28545647859573364, "learning_rate": 1.9552309682143622e-05, "loss": 0.0364, "step": 72300 }, { "grad_norm": 0.21788755059242249, "learning_rate": 1.9539195930614574e-05, "loss": 0.0314, "step": 72310 }, { "grad_norm": 0.38413071632385254, "learning_rate": 1.952608551023324e-05, "loss": 0.0366, "step": 72320 }, { "grad_norm": 0.7097280025482178, "learning_rate": 1.9512978422433397e-05, "loss": 0.0341, "step": 72330 }, { "grad_norm": 0.35120251774787903, "learning_rate": 1.9499874668648376e-05, "loss": 0.0246, "step": 72340 }, { "grad_norm": 0.34170955419540405, "learning_rate": 1.9486774250311213e-05, "loss": 0.0275, "step": 72350 }, { "grad_norm": 0.5729190111160278, "learning_rate": 1.9473677168854537e-05, "loss": 0.0448, "step": 72360 }, { "grad_norm": 0.3096023499965668, "learning_rate": 1.946058342571065e-05, "loss": 0.0413, "step": 72370 }, { "grad_norm": 0.34678158164024353, "learning_rate": 1.9447493022311424e-05, "loss": 0.0341, "step": 72380 }, { "grad_norm": 0.788720428943634, "learning_rate": 1.9434405960088442e-05, "loss": 0.0319, "step": 72390 }, { "grad_norm": 0.39787232875823975, "learning_rate": 1.9421322240472857e-05, "loss": 0.0261, "step": 72400 }, { "grad_norm": 0.4386383891105652, "learning_rate": 1.9408241864895494e-05, "loss": 0.0314, "step": 72410 }, { "grad_norm": 0.6387067437171936, "learning_rate": 1.939516483478681e-05, "loss": 0.0307, "step": 72420 }, { "grad_norm": 1.075711965560913, "learning_rate": 1.9382091151576902e-05, "loss": 0.0343, "step": 72430 }, { "grad_norm": 0.3649296760559082, "learning_rate": 1.9369020816695454e-05, "loss": 0.026, "step": 72440 }, { "grad_norm": 0.2797543704509735, "learning_rate": 1.935595383157184e-05, "loss": 0.0338, "step": 72450 }, { "grad_norm": 0.4575192630290985, "learning_rate": 1.9342890197635057e-05, "loss": 0.0284, "step": 72460 }, { "grad_norm": 0.2724844217300415, "learning_rate": 1.9329829916313685e-05, "loss": 0.0318, "step": 72470 }, { "grad_norm": 0.32852113246917725, "learning_rate": 1.9316772989036013e-05, "loss": 0.036, "step": 72480 }, { "grad_norm": 0.40467700362205505, "learning_rate": 1.930371941722991e-05, "loss": 0.0353, "step": 72490 }, { "grad_norm": 0.32802847027778625, "learning_rate": 1.929066920232292e-05, "loss": 0.0284, "step": 72500 }, { "grad_norm": 0.5321845412254333, "learning_rate": 1.9277622345742152e-05, "loss": 0.0368, "step": 72510 }, { "grad_norm": 0.37328609824180603, "learning_rate": 1.926457884891444e-05, "loss": 0.0291, "step": 72520 }, { "grad_norm": 0.4890820384025574, "learning_rate": 1.9251538713266147e-05, "loss": 0.0367, "step": 72530 }, { "grad_norm": 0.3991664946079254, "learning_rate": 1.9238501940223358e-05, "loss": 0.0287, "step": 72540 }, { "grad_norm": 0.7352210283279419, "learning_rate": 1.922546853121175e-05, "loss": 0.0433, "step": 72550 }, { "grad_norm": 0.3815334439277649, "learning_rate": 1.9212438487656647e-05, "loss": 0.0344, "step": 72560 }, { "grad_norm": 0.27147752046585083, "learning_rate": 1.9199411810982965e-05, "loss": 0.0303, "step": 72570 }, { "grad_norm": 0.24406304955482483, "learning_rate": 1.918638850261532e-05, "loss": 0.0299, "step": 72580 }, { "grad_norm": 0.4496554434299469, "learning_rate": 1.9173368563977885e-05, "loss": 0.0342, "step": 72590 }, { "grad_norm": 0.4702962338924408, "learning_rate": 1.9160351996494513e-05, "loss": 0.0343, "step": 72600 }, { "grad_norm": 0.36122098565101624, "learning_rate": 1.9147338801588677e-05, "loss": 0.0327, "step": 72610 }, { "grad_norm": 0.4010940194129944, "learning_rate": 1.913432898068351e-05, "loss": 0.0372, "step": 72620 }, { "grad_norm": 0.31508851051330566, "learning_rate": 1.912132253520169e-05, "loss": 0.04, "step": 72630 }, { "grad_norm": 0.9349544644355774, "learning_rate": 1.910831946656564e-05, "loss": 0.0399, "step": 72640 }, { "grad_norm": 0.3381021320819855, "learning_rate": 1.909531977619731e-05, "loss": 0.0468, "step": 72650 }, { "grad_norm": 0.4163961112499237, "learning_rate": 1.908232346551834e-05, "loss": 0.0288, "step": 72660 }, { "grad_norm": 0.5716952681541443, "learning_rate": 1.9069330535949993e-05, "loss": 0.0397, "step": 72670 }, { "grad_norm": 0.4430185854434967, "learning_rate": 1.9056340988913167e-05, "loss": 0.0235, "step": 72680 }, { "grad_norm": 0.43627703189849854, "learning_rate": 1.9043354825828357e-05, "loss": 0.0333, "step": 72690 }, { "grad_norm": 0.18934379518032074, "learning_rate": 1.9030372048115724e-05, "loss": 0.0308, "step": 72700 }, { "grad_norm": 0.4120364785194397, "learning_rate": 1.9017392657195056e-05, "loss": 0.0349, "step": 72710 }, { "grad_norm": 0.6166672706604004, "learning_rate": 1.9004416654485724e-05, "loss": 0.0374, "step": 72720 }, { "grad_norm": 0.3639483153820038, "learning_rate": 1.8991444041406785e-05, "loss": 0.0365, "step": 72730 }, { "grad_norm": 0.37311404943466187, "learning_rate": 1.8978474819376912e-05, "loss": 0.0385, "step": 72740 }, { "grad_norm": 0.63038569688797, "learning_rate": 1.8965508989814406e-05, "loss": 0.0361, "step": 72750 }, { "grad_norm": 0.39700615406036377, "learning_rate": 1.8952546554137158e-05, "loss": 0.0315, "step": 72760 }, { "grad_norm": 0.35664400458335876, "learning_rate": 1.8939587513762757e-05, "loss": 0.0263, "step": 72770 }, { "grad_norm": 0.4180748760700226, "learning_rate": 1.892663187010835e-05, "loss": 0.0249, "step": 72780 }, { "grad_norm": 0.23812688887119293, "learning_rate": 1.891367962459077e-05, "loss": 0.0326, "step": 72790 }, { "grad_norm": 0.5862299203872681, "learning_rate": 1.890073077862644e-05, "loss": 0.0428, "step": 72800 }, { "grad_norm": 0.6739661693572998, "learning_rate": 1.888778533363146e-05, "loss": 0.0283, "step": 72810 }, { "grad_norm": 0.478264719247818, "learning_rate": 1.8874843291021478e-05, "loss": 0.0337, "step": 72820 }, { "grad_norm": 0.3850334584712982, "learning_rate": 1.8861904652211853e-05, "loss": 0.0372, "step": 72830 }, { "grad_norm": 0.35791292786598206, "learning_rate": 1.8848969418617507e-05, "loss": 0.033, "step": 72840 }, { "grad_norm": 0.42078450322151184, "learning_rate": 1.8836037591653044e-05, "loss": 0.0353, "step": 72850 }, { "grad_norm": 0.49908173084259033, "learning_rate": 1.882310917273262e-05, "loss": 0.037, "step": 72860 }, { "grad_norm": 0.4931023418903351, "learning_rate": 1.881018416327014e-05, "loss": 0.033, "step": 72870 }, { "grad_norm": 0.4077194631099701, "learning_rate": 1.8797262564679014e-05, "loss": 0.0353, "step": 72880 }, { "grad_norm": 0.409874826669693, "learning_rate": 1.8784344378372355e-05, "loss": 0.0284, "step": 72890 }, { "grad_norm": 0.3622044026851654, "learning_rate": 1.8771429605762843e-05, "loss": 0.0297, "step": 72900 }, { "grad_norm": 0.3425561487674713, "learning_rate": 1.8758518248262835e-05, "loss": 0.0333, "step": 72910 }, { "grad_norm": 0.3321978747844696, "learning_rate": 1.8745610307284305e-05, "loss": 0.0437, "step": 72920 }, { "grad_norm": 0.3884405195713043, "learning_rate": 1.8732705784238862e-05, "loss": 0.0335, "step": 72930 }, { "grad_norm": 0.5146559476852417, "learning_rate": 1.8719804680537682e-05, "loss": 0.0376, "step": 72940 }, { "grad_norm": 0.6180335879325867, "learning_rate": 1.870690699759164e-05, "loss": 0.0312, "step": 72950 }, { "grad_norm": 0.41872408986091614, "learning_rate": 1.8694012736811213e-05, "loss": 0.0359, "step": 72960 }, { "grad_norm": 0.43824705481529236, "learning_rate": 1.8681121899606463e-05, "loss": 0.0251, "step": 72970 }, { "grad_norm": 0.3108324110507965, "learning_rate": 1.8668234487387144e-05, "loss": 0.0364, "step": 72980 }, { "grad_norm": 0.3256624639034271, "learning_rate": 1.8655350501562592e-05, "loss": 0.0329, "step": 72990 }, { "grad_norm": 0.5515984296798706, "learning_rate": 1.8642469943541796e-05, "loss": 0.0443, "step": 73000 }, { "grad_norm": 0.28988945484161377, "learning_rate": 1.8629592814733328e-05, "loss": 0.0319, "step": 73010 }, { "grad_norm": 0.877364993095398, "learning_rate": 1.8616719116545438e-05, "loss": 0.0429, "step": 73020 }, { "grad_norm": 0.2987446188926697, "learning_rate": 1.8603848850385952e-05, "loss": 0.0281, "step": 73030 }, { "grad_norm": 0.43209412693977356, "learning_rate": 1.859098201766236e-05, "loss": 0.0296, "step": 73040 }, { "grad_norm": 0.31971636414527893, "learning_rate": 1.8578118619781726e-05, "loss": 0.0297, "step": 73050 }, { "grad_norm": 0.17929033935070038, "learning_rate": 1.856525865815083e-05, "loss": 0.0237, "step": 73060 }, { "grad_norm": 0.29010382294654846, "learning_rate": 1.8552402134175966e-05, "loss": 0.0273, "step": 73070 }, { "grad_norm": 0.31562432646751404, "learning_rate": 1.8539549049263143e-05, "loss": 0.0317, "step": 73080 }, { "grad_norm": 0.267502099275589, "learning_rate": 1.8526699404817917e-05, "loss": 0.0324, "step": 73090 }, { "grad_norm": 0.3771274983882904, "learning_rate": 1.8513853202245546e-05, "loss": 0.0278, "step": 73100 }, { "grad_norm": 0.42854756116867065, "learning_rate": 1.850101044295081e-05, "loss": 0.0359, "step": 73110 }, { "grad_norm": 0.43807321786880493, "learning_rate": 1.848817112833825e-05, "loss": 0.0212, "step": 73120 }, { "grad_norm": 0.556083619594574, "learning_rate": 1.8475335259811905e-05, "loss": 0.0271, "step": 73130 }, { "grad_norm": 0.724449872970581, "learning_rate": 1.8462502838775508e-05, "loss": 0.0382, "step": 73140 }, { "grad_norm": 0.29180341958999634, "learning_rate": 1.8449673866632378e-05, "loss": 0.0405, "step": 73150 }, { "grad_norm": 0.30049005150794983, "learning_rate": 1.8436848344785473e-05, "loss": 0.0396, "step": 73160 }, { "grad_norm": 0.34475433826446533, "learning_rate": 1.8424026274637378e-05, "loss": 0.0288, "step": 73170 }, { "grad_norm": 0.6591408848762512, "learning_rate": 1.8411207657590312e-05, "loss": 0.0375, "step": 73180 }, { "grad_norm": 0.28376320004463196, "learning_rate": 1.839839249504607e-05, "loss": 0.0344, "step": 73190 }, { "grad_norm": 0.3681468367576599, "learning_rate": 1.8385580788406108e-05, "loss": 0.0364, "step": 73200 }, { "grad_norm": 0.4613167345523834, "learning_rate": 1.837277253907152e-05, "loss": 0.0329, "step": 73210 }, { "grad_norm": 0.36252647638320923, "learning_rate": 1.8359967748442957e-05, "loss": 0.0326, "step": 73220 }, { "grad_norm": 0.4117736220359802, "learning_rate": 1.8347166417920774e-05, "loss": 0.0324, "step": 73230 }, { "grad_norm": 0.7419973611831665, "learning_rate": 1.833436854890484e-05, "loss": 0.035, "step": 73240 }, { "grad_norm": 0.4860438108444214, "learning_rate": 1.8321574142794796e-05, "loss": 0.0286, "step": 73250 }, { "grad_norm": 0.4976462423801422, "learning_rate": 1.8308783200989748e-05, "loss": 0.0347, "step": 73260 }, { "grad_norm": 0.8548027276992798, "learning_rate": 1.829599572488855e-05, "loss": 0.0339, "step": 73270 }, { "grad_norm": 0.19203664362430573, "learning_rate": 1.828321171588957e-05, "loss": 0.0332, "step": 73280 }, { "grad_norm": 0.4713972508907318, "learning_rate": 1.827043117539089e-05, "loss": 0.0386, "step": 73290 }, { "grad_norm": 0.1939171701669693, "learning_rate": 1.8257654104790116e-05, "loss": 0.0239, "step": 73300 }, { "grad_norm": 0.4858643710613251, "learning_rate": 1.82448805054846e-05, "loss": 0.0321, "step": 73310 }, { "grad_norm": 0.3896019458770752, "learning_rate": 1.823211037887119e-05, "loss": 0.0406, "step": 73320 }, { "grad_norm": 0.965914785861969, "learning_rate": 1.821934372634645e-05, "loss": 0.0385, "step": 73330 }, { "grad_norm": 0.29861006140708923, "learning_rate": 1.8206580549306472e-05, "loss": 0.0268, "step": 73340 }, { "grad_norm": 0.27376389503479004, "learning_rate": 1.819382084914707e-05, "loss": 0.0324, "step": 73350 }, { "grad_norm": 0.3611204922199249, "learning_rate": 1.8181064627263577e-05, "loss": 0.0238, "step": 73360 }, { "grad_norm": 0.39982977509498596, "learning_rate": 1.8168311885051022e-05, "loss": 0.0345, "step": 73370 }, { "grad_norm": 0.5052004456520081, "learning_rate": 1.815556262390402e-05, "loss": 0.0309, "step": 73380 }, { "grad_norm": 0.4702332615852356, "learning_rate": 1.8142816845216826e-05, "loss": 0.0325, "step": 73390 }, { "grad_norm": 0.28838661313056946, "learning_rate": 1.8130074550383276e-05, "loss": 0.0294, "step": 73400 }, { "grad_norm": 0.46683305501937866, "learning_rate": 1.8117335740796848e-05, "loss": 0.0254, "step": 73410 }, { "grad_norm": 0.5256631970405579, "learning_rate": 1.810460041785067e-05, "loss": 0.0342, "step": 73420 }, { "grad_norm": 1.5710721015930176, "learning_rate": 1.809186858293742e-05, "loss": 0.0456, "step": 73430 }, { "grad_norm": 0.3398524820804596, "learning_rate": 1.8079140237449448e-05, "loss": 0.0252, "step": 73440 }, { "grad_norm": 0.6356343626976013, "learning_rate": 1.8066415382778707e-05, "loss": 0.0405, "step": 73450 }, { "grad_norm": 0.46074867248535156, "learning_rate": 1.8053694020316786e-05, "loss": 0.031, "step": 73460 }, { "grad_norm": 0.31176313757896423, "learning_rate": 1.8040976151454837e-05, "loss": 0.0347, "step": 73470 }, { "grad_norm": 0.5008653402328491, "learning_rate": 1.8028261777583704e-05, "loss": 0.0416, "step": 73480 }, { "grad_norm": 0.3991178572177887, "learning_rate": 1.8015550900093758e-05, "loss": 0.0393, "step": 73490 }, { "grad_norm": 0.6176121234893799, "learning_rate": 1.800284352037512e-05, "loss": 0.0265, "step": 73500 }, { "grad_norm": 0.25410616397857666, "learning_rate": 1.799013963981738e-05, "loss": 0.0369, "step": 73510 }, { "grad_norm": 0.4152541756629944, "learning_rate": 1.797743925980987e-05, "loss": 0.0301, "step": 73520 }, { "grad_norm": 0.5434384346008301, "learning_rate": 1.7964742381741435e-05, "loss": 0.0331, "step": 73530 }, { "grad_norm": 1.5801024436950684, "learning_rate": 1.7952049007000627e-05, "loss": 0.0457, "step": 73540 }, { "grad_norm": 0.5176923871040344, "learning_rate": 1.7939359136975535e-05, "loss": 0.0271, "step": 73550 }, { "grad_norm": 0.4875311255455017, "learning_rate": 1.7926672773053932e-05, "loss": 0.0289, "step": 73560 }, { "grad_norm": 0.39456430077552795, "learning_rate": 1.791398991662317e-05, "loss": 0.0336, "step": 73570 }, { "grad_norm": 0.2564780116081238, "learning_rate": 1.7901310569070244e-05, "loss": 0.032, "step": 73580 }, { "grad_norm": 0.5003393292427063, "learning_rate": 1.7888634731781724e-05, "loss": 0.0288, "step": 73590 }, { "grad_norm": 0.9736695885658264, "learning_rate": 1.7875962406143844e-05, "loss": 0.0264, "step": 73600 }, { "grad_norm": 0.33674564957618713, "learning_rate": 1.7863293593542403e-05, "loss": 0.0399, "step": 73610 }, { "grad_norm": 0.32363659143447876, "learning_rate": 1.7850628295362853e-05, "loss": 0.0263, "step": 73620 }, { "grad_norm": 0.4067623019218445, "learning_rate": 1.7837966512990263e-05, "loss": 0.0371, "step": 73630 }, { "grad_norm": 0.3512195646762848, "learning_rate": 1.782530824780931e-05, "loss": 0.0299, "step": 73640 }, { "grad_norm": 0.4570187032222748, "learning_rate": 1.7812653501204263e-05, "loss": 0.0316, "step": 73650 }, { "grad_norm": 0.372125506401062, "learning_rate": 1.780000227455903e-05, "loss": 0.0325, "step": 73660 }, { "grad_norm": 0.500018835067749, "learning_rate": 1.778735456925716e-05, "loss": 0.0353, "step": 73670 }, { "grad_norm": 0.26078560948371887, "learning_rate": 1.7774710386681737e-05, "loss": 0.0259, "step": 73680 }, { "grad_norm": 0.30851566791534424, "learning_rate": 1.7762069728215545e-05, "loss": 0.0352, "step": 73690 }, { "grad_norm": 0.374988853931427, "learning_rate": 1.7749432595240932e-05, "loss": 0.0315, "step": 73700 }, { "grad_norm": 0.382744699716568, "learning_rate": 1.773679898913991e-05, "loss": 0.0358, "step": 73710 }, { "grad_norm": 0.5372465252876282, "learning_rate": 1.7724168911294015e-05, "loss": 0.0381, "step": 73720 }, { "grad_norm": 0.41267797350883484, "learning_rate": 1.7711542363084504e-05, "loss": 0.0281, "step": 73730 }, { "grad_norm": 0.39085260033607483, "learning_rate": 1.7698919345892158e-05, "loss": 0.0278, "step": 73740 }, { "grad_norm": 0.3250258266925812, "learning_rate": 1.768629986109743e-05, "loss": 0.0366, "step": 73750 }, { "grad_norm": 0.47185230255126953, "learning_rate": 1.7673683910080357e-05, "loss": 0.0286, "step": 73760 }, { "grad_norm": 0.2379751354455948, "learning_rate": 1.7661071494220638e-05, "loss": 0.0325, "step": 73770 }, { "grad_norm": 0.23418913781642914, "learning_rate": 1.7648462614897498e-05, "loss": 0.0271, "step": 73780 }, { "grad_norm": 1.376186490058899, "learning_rate": 1.7635857273489863e-05, "loss": 0.0296, "step": 73790 }, { "grad_norm": 0.511520504951477, "learning_rate": 1.7623255471376198e-05, "loss": 0.0328, "step": 73800 }, { "grad_norm": 0.5374942421913147, "learning_rate": 1.7610657209934646e-05, "loss": 0.0466, "step": 73810 }, { "grad_norm": 0.268912672996521, "learning_rate": 1.7598062490542915e-05, "loss": 0.035, "step": 73820 }, { "grad_norm": 0.3653312921524048, "learning_rate": 1.7585471314578377e-05, "loss": 0.0338, "step": 73830 }, { "grad_norm": 0.40068319439888, "learning_rate": 1.7572883683417945e-05, "loss": 0.0422, "step": 73840 }, { "grad_norm": 0.48054933547973633, "learning_rate": 1.7560299598438212e-05, "loss": 0.0431, "step": 73850 }, { "grad_norm": 0.23445534706115723, "learning_rate": 1.7547719061015334e-05, "loss": 0.0303, "step": 73860 }, { "grad_norm": 0.3337237536907196, "learning_rate": 1.75351420725251e-05, "loss": 0.0269, "step": 73870 }, { "grad_norm": 0.27766016125679016, "learning_rate": 1.7522568634342923e-05, "loss": 0.0306, "step": 73880 }, { "grad_norm": 0.30078327655792236, "learning_rate": 1.7509998747843825e-05, "loss": 0.0295, "step": 73890 }, { "grad_norm": 0.33642125129699707, "learning_rate": 1.7497432414402403e-05, "loss": 0.0299, "step": 73900 }, { "grad_norm": 1.0099012851715088, "learning_rate": 1.7484869635392903e-05, "loss": 0.0277, "step": 73910 }, { "grad_norm": 0.4401356279850006, "learning_rate": 1.747231041218919e-05, "loss": 0.0286, "step": 73920 }, { "grad_norm": 0.39749494194984436, "learning_rate": 1.7459754746164692e-05, "loss": 0.0322, "step": 73930 }, { "grad_norm": 0.952297568321228, "learning_rate": 1.744720263869249e-05, "loss": 0.038, "step": 73940 }, { "grad_norm": 1.0279324054718018, "learning_rate": 1.7434654091145263e-05, "loss": 0.0305, "step": 73950 }, { "grad_norm": 0.4103679060935974, "learning_rate": 1.742210910489532e-05, "loss": 0.0285, "step": 73960 }, { "grad_norm": 0.17879480123519897, "learning_rate": 1.7409567681314532e-05, "loss": 0.0297, "step": 73970 }, { "grad_norm": 0.23230788111686707, "learning_rate": 1.7397029821774434e-05, "loss": 0.0331, "step": 73980 }, { "grad_norm": 0.3928362727165222, "learning_rate": 1.7384495527646126e-05, "loss": 0.0328, "step": 73990 }, { "grad_norm": 0.3298207223415375, "learning_rate": 1.737196480030035e-05, "loss": 0.0309, "step": 74000 }, { "grad_norm": 0.3703271746635437, "learning_rate": 1.7359437641107447e-05, "loss": 0.0345, "step": 74010 }, { "grad_norm": 0.3902604579925537, "learning_rate": 1.7346914051437387e-05, "loss": 0.0277, "step": 74020 }, { "grad_norm": 0.42021843791007996, "learning_rate": 1.733439403265969e-05, "loss": 0.0287, "step": 74030 }, { "grad_norm": 0.3216956555843353, "learning_rate": 1.732187758614357e-05, "loss": 0.0394, "step": 74040 }, { "grad_norm": 0.3852483928203583, "learning_rate": 1.730936471325777e-05, "loss": 0.033, "step": 74050 }, { "grad_norm": 0.3010738492012024, "learning_rate": 1.7296855415370718e-05, "loss": 0.0266, "step": 74060 }, { "grad_norm": 0.7257303595542908, "learning_rate": 1.728434969385035e-05, "loss": 0.0279, "step": 74070 }, { "grad_norm": 0.848342776298523, "learning_rate": 1.7271847550064346e-05, "loss": 0.0343, "step": 74080 }, { "grad_norm": 0.6787903904914856, "learning_rate": 1.725934898537987e-05, "loss": 0.0349, "step": 74090 }, { "grad_norm": 0.555660605430603, "learning_rate": 1.7246854001163783e-05, "loss": 0.024, "step": 74100 }, { "grad_norm": 0.4990488886833191, "learning_rate": 1.7234362598782476e-05, "loss": 0.0311, "step": 74110 }, { "grad_norm": 0.8306172490119934, "learning_rate": 1.722187477960202e-05, "loss": 0.035, "step": 74120 }, { "grad_norm": 0.7823688983917236, "learning_rate": 1.7209390544988057e-05, "loss": 0.0324, "step": 74130 }, { "grad_norm": 0.3648071587085724, "learning_rate": 1.7196909896305856e-05, "loss": 0.0426, "step": 74140 }, { "grad_norm": 0.3286972939968109, "learning_rate": 1.718443283492026e-05, "loss": 0.0345, "step": 74150 }, { "grad_norm": 0.5184528827667236, "learning_rate": 1.7171959362195746e-05, "loss": 0.0348, "step": 74160 }, { "grad_norm": 0.24113303422927856, "learning_rate": 1.715948947949642e-05, "loss": 0.0288, "step": 74170 }, { "grad_norm": 0.3568591773509979, "learning_rate": 1.714702318818593e-05, "loss": 0.0399, "step": 74180 }, { "grad_norm": 0.3292202055454254, "learning_rate": 1.7134560489627592e-05, "loss": 0.0282, "step": 74190 }, { "grad_norm": 0.31820157170295715, "learning_rate": 1.712210138518431e-05, "loss": 0.0271, "step": 74200 }, { "grad_norm": 0.3967777490615845, "learning_rate": 1.71096458762186e-05, "loss": 0.0421, "step": 74210 }, { "grad_norm": 0.3012077510356903, "learning_rate": 1.7097193964092552e-05, "loss": 0.0477, "step": 74220 }, { "grad_norm": 0.3949719965457916, "learning_rate": 1.708474565016792e-05, "loss": 0.0336, "step": 74230 }, { "grad_norm": 0.36261147260665894, "learning_rate": 1.7072300935806e-05, "loss": 0.0342, "step": 74240 }, { "grad_norm": 0.4565669894218445, "learning_rate": 1.705985982236776e-05, "loss": 0.0311, "step": 74250 }, { "grad_norm": 0.37671881914138794, "learning_rate": 1.7047422311213697e-05, "loss": 0.0274, "step": 74260 }, { "grad_norm": 0.5104070901870728, "learning_rate": 1.703498840370401e-05, "loss": 0.0387, "step": 74270 }, { "grad_norm": 0.4233083426952362, "learning_rate": 1.702255810119842e-05, "loss": 0.027, "step": 74280 }, { "grad_norm": 0.4068509638309479, "learning_rate": 1.7010131405056305e-05, "loss": 0.0257, "step": 74290 }, { "grad_norm": 0.335518479347229, "learning_rate": 1.6997708316636613e-05, "loss": 0.0297, "step": 74300 }, { "grad_norm": 0.36270248889923096, "learning_rate": 1.6985288837297932e-05, "loss": 0.0325, "step": 74310 }, { "grad_norm": 0.663993775844574, "learning_rate": 1.6972872968398394e-05, "loss": 0.0407, "step": 74320 }, { "grad_norm": 0.3865037262439728, "learning_rate": 1.6960460711295845e-05, "loss": 0.0338, "step": 74330 }, { "grad_norm": 0.2948168218135834, "learning_rate": 1.6948052067347623e-05, "loss": 0.0249, "step": 74340 }, { "grad_norm": 0.6504384279251099, "learning_rate": 1.693564703791075e-05, "loss": 0.031, "step": 74350 }, { "grad_norm": 0.23362158238887787, "learning_rate": 1.6923245624341793e-05, "loss": 0.0383, "step": 74360 }, { "grad_norm": 0.32325661182403564, "learning_rate": 1.6910847827996962e-05, "loss": 0.0257, "step": 74370 }, { "grad_norm": 0.27377036213874817, "learning_rate": 1.689845365023206e-05, "loss": 0.0277, "step": 74380 }, { "grad_norm": 0.20667055249214172, "learning_rate": 1.6886063092402505e-05, "loss": 0.0307, "step": 74390 }, { "grad_norm": 0.5437473654747009, "learning_rate": 1.6873676155863317e-05, "loss": 0.0428, "step": 74400 }, { "grad_norm": 0.24394117295742035, "learning_rate": 1.6861292841969085e-05, "loss": 0.0269, "step": 74410 }, { "grad_norm": 0.6344557404518127, "learning_rate": 1.684891315207406e-05, "loss": 0.0315, "step": 74420 }, { "grad_norm": 0.4625888764858246, "learning_rate": 1.6836537087532034e-05, "loss": 0.036, "step": 74430 }, { "grad_norm": 0.2946072220802307, "learning_rate": 1.6824164649696465e-05, "loss": 0.033, "step": 74440 }, { "grad_norm": 0.8075947761535645, "learning_rate": 1.681179583992034e-05, "loss": 0.029, "step": 74450 }, { "grad_norm": 0.35896992683410645, "learning_rate": 1.679943065955636e-05, "loss": 0.0286, "step": 74460 }, { "grad_norm": 0.7769936919212341, "learning_rate": 1.678706910995671e-05, "loss": 0.0274, "step": 74470 }, { "grad_norm": 0.380501925945282, "learning_rate": 1.6774711192473262e-05, "loss": 0.0303, "step": 74480 }, { "grad_norm": 0.37038400769233704, "learning_rate": 1.676235690845743e-05, "loss": 0.0256, "step": 74490 }, { "grad_norm": 0.5873676538467407, "learning_rate": 1.6750006259260285e-05, "loss": 0.0369, "step": 74500 }, { "grad_norm": 0.41580018401145935, "learning_rate": 1.673765924623244e-05, "loss": 0.0365, "step": 74510 }, { "grad_norm": 0.35550951957702637, "learning_rate": 1.67253158707242e-05, "loss": 0.025, "step": 74520 }, { "grad_norm": 0.5958740711212158, "learning_rate": 1.671297613408537e-05, "loss": 0.0311, "step": 74530 }, { "grad_norm": 0.5527830719947815, "learning_rate": 1.6700640037665444e-05, "loss": 0.037, "step": 74540 }, { "grad_norm": 0.41018590331077576, "learning_rate": 1.668830758281344e-05, "loss": 0.0355, "step": 74550 }, { "grad_norm": 0.33601856231689453, "learning_rate": 1.6675978770878052e-05, "loss": 0.0294, "step": 74560 }, { "grad_norm": 0.2756241261959076, "learning_rate": 1.6663653603207506e-05, "loss": 0.0282, "step": 74570 }, { "grad_norm": 0.4540868103504181, "learning_rate": 1.6651332081149685e-05, "loss": 0.0428, "step": 74580 }, { "grad_norm": 0.5401943325996399, "learning_rate": 1.6639014206052046e-05, "loss": 0.0297, "step": 74590 }, { "grad_norm": 0.27722620964050293, "learning_rate": 1.6626699979261657e-05, "loss": 0.0252, "step": 74600 }, { "grad_norm": 0.3868255913257599, "learning_rate": 1.6614389402125197e-05, "loss": 0.0396, "step": 74610 }, { "grad_norm": 0.33794814348220825, "learning_rate": 1.66020824759889e-05, "loss": 0.0276, "step": 74620 }, { "grad_norm": 0.446992427110672, "learning_rate": 1.6589779202198675e-05, "loss": 0.0278, "step": 74630 }, { "grad_norm": 0.502346932888031, "learning_rate": 1.6577479582099924e-05, "loss": 0.0363, "step": 74640 }, { "grad_norm": 0.5981279611587524, "learning_rate": 1.6565183617037788e-05, "loss": 0.0377, "step": 74650 }, { "grad_norm": 0.47179368138313293, "learning_rate": 1.6552891308356894e-05, "loss": 0.0323, "step": 74660 }, { "grad_norm": 0.43669450283050537, "learning_rate": 1.6540602657401526e-05, "loss": 0.0384, "step": 74670 }, { "grad_norm": 0.3496713936328888, "learning_rate": 1.6528317665515537e-05, "loss": 0.0452, "step": 74680 }, { "grad_norm": 0.9951441884040833, "learning_rate": 1.6516036334042416e-05, "loss": 0.0496, "step": 74690 }, { "grad_norm": 0.3113156855106354, "learning_rate": 1.650375866432519e-05, "loss": 0.0474, "step": 74700 }, { "grad_norm": 0.4001193642616272, "learning_rate": 1.649148465770659e-05, "loss": 0.036, "step": 74710 }, { "grad_norm": 0.8356979489326477, "learning_rate": 1.647921431552883e-05, "loss": 0.0423, "step": 74720 }, { "grad_norm": 1.3818656206130981, "learning_rate": 1.646694763913381e-05, "loss": 0.0431, "step": 74730 }, { "grad_norm": 0.5428155064582825, "learning_rate": 1.6454684629862966e-05, "loss": 0.0342, "step": 74740 }, { "grad_norm": 0.44383561611175537, "learning_rate": 1.6442425289057388e-05, "loss": 0.0289, "step": 74750 }, { "grad_norm": 0.35553234815597534, "learning_rate": 1.6430169618057718e-05, "loss": 0.0316, "step": 74760 }, { "grad_norm": 0.20946384966373444, "learning_rate": 1.641791761820422e-05, "loss": 0.0266, "step": 74770 }, { "grad_norm": 0.4067873954772949, "learning_rate": 1.640566929083676e-05, "loss": 0.0304, "step": 74780 }, { "grad_norm": 0.6155193448066711, "learning_rate": 1.6393424637294823e-05, "loss": 0.0379, "step": 74790 }, { "grad_norm": 0.2962836027145386, "learning_rate": 1.638118365891742e-05, "loss": 0.0264, "step": 74800 }, { "grad_norm": 0.45464593172073364, "learning_rate": 1.636894635704323e-05, "loss": 0.0362, "step": 74810 }, { "grad_norm": 0.2997189164161682, "learning_rate": 1.6356712733010517e-05, "loss": 0.0351, "step": 74820 }, { "grad_norm": 0.9992758631706238, "learning_rate": 1.6344482788157096e-05, "loss": 0.0297, "step": 74830 }, { "grad_norm": 0.3666987121105194, "learning_rate": 1.6332256523820444e-05, "loss": 0.0285, "step": 74840 }, { "grad_norm": 0.7832924127578735, "learning_rate": 1.6320033941337597e-05, "loss": 0.0396, "step": 74850 }, { "grad_norm": 0.39922258257865906, "learning_rate": 1.6307815042045216e-05, "loss": 0.0299, "step": 74860 }, { "grad_norm": 0.45969340205192566, "learning_rate": 1.6295599827279513e-05, "loss": 0.0295, "step": 74870 }, { "grad_norm": 0.48936229944229126, "learning_rate": 1.6283388298376356e-05, "loss": 0.0343, "step": 74880 }, { "grad_norm": 0.4310975670814514, "learning_rate": 1.6271180456671125e-05, "loss": 0.0293, "step": 74890 }, { "grad_norm": 0.3625926673412323, "learning_rate": 1.625897630349893e-05, "loss": 0.0258, "step": 74900 }, { "grad_norm": 0.4372216463088989, "learning_rate": 1.6246775840194335e-05, "loss": 0.0313, "step": 74910 }, { "grad_norm": 0.27009886503219604, "learning_rate": 1.6234579068091604e-05, "loss": 0.0241, "step": 74920 }, { "grad_norm": 0.41027310490608215, "learning_rate": 1.6222385988524524e-05, "loss": 0.033, "step": 74930 }, { "grad_norm": 0.2961721420288086, "learning_rate": 1.6210196602826544e-05, "loss": 0.0342, "step": 74940 }, { "grad_norm": 0.20988836884498596, "learning_rate": 1.6198010912330646e-05, "loss": 0.0289, "step": 74950 }, { "grad_norm": 0.4420551061630249, "learning_rate": 1.6185828918369456e-05, "loss": 0.0359, "step": 74960 }, { "grad_norm": 0.49048346281051636, "learning_rate": 1.617365062227517e-05, "loss": 0.0346, "step": 74970 }, { "grad_norm": 0.22720497846603394, "learning_rate": 1.616147602537962e-05, "loss": 0.0415, "step": 74980 }, { "grad_norm": 0.3283463716506958, "learning_rate": 1.614930512901415e-05, "loss": 0.0301, "step": 74990 }, { "grad_norm": 0.4282599985599518, "learning_rate": 1.6137137934509804e-05, "loss": 0.0226, "step": 75000 }, { "grad_norm": 0.3377009928226471, "learning_rate": 1.6124974443197123e-05, "loss": 0.0309, "step": 75010 }, { "grad_norm": 0.3677823841571808, "learning_rate": 1.6112814656406306e-05, "loss": 0.0344, "step": 75020 }, { "grad_norm": 0.4239729642868042, "learning_rate": 1.6100658575467125e-05, "loss": 0.0349, "step": 75030 }, { "grad_norm": 0.4188910722732544, "learning_rate": 1.6088506201708974e-05, "loss": 0.0267, "step": 75040 }, { "grad_norm": 0.5312337875366211, "learning_rate": 1.6076357536460785e-05, "loss": 0.0383, "step": 75050 }, { "grad_norm": 0.38167935609817505, "learning_rate": 1.606421258105113e-05, "loss": 0.0283, "step": 75060 }, { "grad_norm": 0.45178958773612976, "learning_rate": 1.6052071336808183e-05, "loss": 0.0334, "step": 75070 }, { "grad_norm": 0.27836501598358154, "learning_rate": 1.6039933805059658e-05, "loss": 0.0353, "step": 75080 }, { "grad_norm": 0.3519189655780792, "learning_rate": 1.6027799987132915e-05, "loss": 0.0318, "step": 75090 }, { "grad_norm": 0.3790695071220398, "learning_rate": 1.6015669884354888e-05, "loss": 0.0453, "step": 75100 }, { "grad_norm": 0.46472784876823425, "learning_rate": 1.6003543498052125e-05, "loss": 0.0372, "step": 75110 }, { "grad_norm": 0.4130105674266815, "learning_rate": 1.5991420829550715e-05, "loss": 0.0303, "step": 75120 }, { "grad_norm": 0.3006988763809204, "learning_rate": 1.5979301880176405e-05, "loss": 0.0387, "step": 75130 }, { "grad_norm": 0.27481281757354736, "learning_rate": 1.5967186651254485e-05, "loss": 0.0327, "step": 75140 }, { "grad_norm": 0.5046192407608032, "learning_rate": 1.5955075144109853e-05, "loss": 0.0281, "step": 75150 }, { "grad_norm": 0.5007975101470947, "learning_rate": 1.594296736006703e-05, "loss": 0.0309, "step": 75160 }, { "grad_norm": 0.24052993953227997, "learning_rate": 1.59308633004501e-05, "loss": 0.0388, "step": 75170 }, { "grad_norm": 0.40764835476875305, "learning_rate": 1.5918762966582724e-05, "loss": 0.0407, "step": 75180 }, { "grad_norm": 0.40363436937332153, "learning_rate": 1.5906666359788203e-05, "loss": 0.0311, "step": 75190 }, { "grad_norm": 0.28849244117736816, "learning_rate": 1.589457348138938e-05, "loss": 0.0322, "step": 75200 }, { "grad_norm": 0.4686276316642761, "learning_rate": 1.588248433270872e-05, "loss": 0.0319, "step": 75210 }, { "grad_norm": 0.4031059443950653, "learning_rate": 1.587039891506828e-05, "loss": 0.032, "step": 75220 }, { "grad_norm": 0.2708645462989807, "learning_rate": 1.5858317229789715e-05, "loss": 0.0322, "step": 75230 }, { "grad_norm": 0.521250307559967, "learning_rate": 1.584623927819423e-05, "loss": 0.0283, "step": 75240 }, { "grad_norm": 0.7426726818084717, "learning_rate": 1.583416506160269e-05, "loss": 0.0429, "step": 75250 }, { "grad_norm": 0.4154389500617981, "learning_rate": 1.5822094581335472e-05, "loss": 0.0409, "step": 75260 }, { "grad_norm": 0.3274163007736206, "learning_rate": 1.5810027838712604e-05, "loss": 0.0359, "step": 75270 }, { "grad_norm": 0.6095073223114014, "learning_rate": 1.5797964835053687e-05, "loss": 0.0294, "step": 75280 }, { "grad_norm": 0.2604439854621887, "learning_rate": 1.5785905571677933e-05, "loss": 0.0276, "step": 75290 }, { "grad_norm": 0.4575482904911041, "learning_rate": 1.5773850049904086e-05, "loss": 0.0332, "step": 75300 }, { "grad_norm": 0.4995575249195099, "learning_rate": 1.5761798271050543e-05, "loss": 0.0337, "step": 75310 }, { "grad_norm": 0.4572439193725586, "learning_rate": 1.5749750236435277e-05, "loss": 0.0315, "step": 75320 }, { "grad_norm": 0.6608771681785583, "learning_rate": 1.5737705947375815e-05, "loss": 0.0337, "step": 75330 }, { "grad_norm": 0.39753609895706177, "learning_rate": 1.5725665405189322e-05, "loss": 0.0359, "step": 75340 }, { "grad_norm": 0.4844231903553009, "learning_rate": 1.571362861119253e-05, "loss": 0.0326, "step": 75350 }, { "grad_norm": 0.29755109548568726, "learning_rate": 1.570159556670177e-05, "loss": 0.0314, "step": 75360 }, { "grad_norm": 0.19432367384433746, "learning_rate": 1.5689566273032946e-05, "loss": 0.0355, "step": 75370 }, { "grad_norm": 0.3394545614719391, "learning_rate": 1.5677540731501583e-05, "loss": 0.0362, "step": 75380 }, { "grad_norm": 0.46512219309806824, "learning_rate": 1.5665518943422746e-05, "loss": 0.0328, "step": 75390 }, { "grad_norm": 0.36333513259887695, "learning_rate": 1.5653500910111134e-05, "loss": 0.0374, "step": 75400 }, { "grad_norm": 0.4788365662097931, "learning_rate": 1.5641486632881023e-05, "loss": 0.036, "step": 75410 }, { "grad_norm": 0.4063061773777008, "learning_rate": 1.56294761130463e-05, "loss": 0.0303, "step": 75420 }, { "grad_norm": 0.4072888195514679, "learning_rate": 1.5617469351920373e-05, "loss": 0.034, "step": 75430 }, { "grad_norm": 0.3015904724597931, "learning_rate": 1.5605466350816315e-05, "loss": 0.0321, "step": 75440 }, { "grad_norm": 0.34885162115097046, "learning_rate": 1.559346711104674e-05, "loss": 0.0319, "step": 75450 }, { "grad_norm": 0.42260777950286865, "learning_rate": 1.5581471633923883e-05, "loss": 0.0362, "step": 75460 }, { "grad_norm": 0.4487461447715759, "learning_rate": 1.5569479920759505e-05, "loss": 0.0359, "step": 75470 }, { "grad_norm": 0.37231841683387756, "learning_rate": 1.5557491972865072e-05, "loss": 0.0318, "step": 75480 }, { "grad_norm": 0.25787779688835144, "learning_rate": 1.554550779155152e-05, "loss": 0.0228, "step": 75490 }, { "grad_norm": 0.33720728754997253, "learning_rate": 1.5533527378129453e-05, "loss": 0.0237, "step": 75500 }, { "grad_norm": 0.43027663230895996, "learning_rate": 1.552155073390899e-05, "loss": 0.0363, "step": 75510 }, { "grad_norm": 0.3887423574924469, "learning_rate": 1.5509577860199903e-05, "loss": 0.0447, "step": 75520 }, { "grad_norm": 0.31853243708610535, "learning_rate": 1.5497608758311525e-05, "loss": 0.0334, "step": 75530 }, { "grad_norm": 0.3690298795700073, "learning_rate": 1.5485643429552797e-05, "loss": 0.0325, "step": 75540 }, { "grad_norm": 0.42743927240371704, "learning_rate": 1.5473681875232197e-05, "loss": 0.0345, "step": 75550 }, { "grad_norm": 0.6080257892608643, "learning_rate": 1.5461724096657837e-05, "loss": 0.0287, "step": 75560 }, { "grad_norm": 0.3735201060771942, "learning_rate": 1.5449770095137416e-05, "loss": 0.0305, "step": 75570 }, { "grad_norm": 0.41039568185806274, "learning_rate": 1.5437819871978178e-05, "loss": 0.0302, "step": 75580 }, { "grad_norm": 0.66998291015625, "learning_rate": 1.5425873428486992e-05, "loss": 0.0346, "step": 75590 }, { "grad_norm": 0.2129909098148346, "learning_rate": 1.5413930765970312e-05, "loss": 0.0308, "step": 75600 }, { "grad_norm": 0.4082423448562622, "learning_rate": 1.5401991885734175e-05, "loss": 0.0345, "step": 75610 }, { "grad_norm": 0.43198102712631226, "learning_rate": 1.5390056789084174e-05, "loss": 0.03, "step": 75620 }, { "grad_norm": 0.5367146134376526, "learning_rate": 1.5378125477325545e-05, "loss": 0.0281, "step": 75630 }, { "grad_norm": 0.2881925106048584, "learning_rate": 1.5366197951763046e-05, "loss": 0.0246, "step": 75640 }, { "grad_norm": 0.6098467111587524, "learning_rate": 1.5354274213701075e-05, "loss": 0.0348, "step": 75650 }, { "grad_norm": 0.3431016206741333, "learning_rate": 1.5342354264443566e-05, "loss": 0.0332, "step": 75660 }, { "grad_norm": 0.5450731515884399, "learning_rate": 1.533043810529411e-05, "loss": 0.034, "step": 75670 }, { "grad_norm": 0.385525643825531, "learning_rate": 1.5318525737555807e-05, "loss": 0.0419, "step": 75680 }, { "grad_norm": 0.4338020384311676, "learning_rate": 1.53066171625314e-05, "loss": 0.0371, "step": 75690 }, { "grad_norm": 0.3214133083820343, "learning_rate": 1.529471238152317e-05, "loss": 0.0299, "step": 75700 }, { "grad_norm": 0.5294890403747559, "learning_rate": 1.5282811395833023e-05, "loss": 0.0305, "step": 75710 }, { "grad_norm": 0.5052487254142761, "learning_rate": 1.5270914206762397e-05, "loss": 0.0338, "step": 75720 }, { "grad_norm": 0.3825357258319855, "learning_rate": 1.525902081561241e-05, "loss": 0.0303, "step": 75730 }, { "grad_norm": 0.3247159421443939, "learning_rate": 1.5247131223683664e-05, "loss": 0.0391, "step": 75740 }, { "grad_norm": 0.5562264919281006, "learning_rate": 1.5235245432276408e-05, "loss": 0.0329, "step": 75750 }, { "grad_norm": 0.48397696018218994, "learning_rate": 1.5223363442690424e-05, "loss": 0.0332, "step": 75760 }, { "grad_norm": 0.38202983140945435, "learning_rate": 1.5211485256225133e-05, "loss": 0.0291, "step": 75770 }, { "grad_norm": 0.6967182755470276, "learning_rate": 1.519961087417951e-05, "loss": 0.0253, "step": 75780 }, { "grad_norm": 0.24710020422935486, "learning_rate": 1.5187740297852137e-05, "loss": 0.0319, "step": 75790 }, { "grad_norm": 0.2502228617668152, "learning_rate": 1.517587352854113e-05, "loss": 0.0227, "step": 75800 }, { "grad_norm": 0.9152805209159851, "learning_rate": 1.516401056754423e-05, "loss": 0.0377, "step": 75810 }, { "grad_norm": 0.5368936061859131, "learning_rate": 1.5152151416158783e-05, "loss": 0.0357, "step": 75820 }, { "grad_norm": 0.33135831356048584, "learning_rate": 1.5140296075681642e-05, "loss": 0.0365, "step": 75830 }, { "grad_norm": 0.37217462062835693, "learning_rate": 1.5128444547409331e-05, "loss": 0.0293, "step": 75840 }, { "grad_norm": 0.4807779788970947, "learning_rate": 1.5116596832637864e-05, "loss": 0.0281, "step": 75850 }, { "grad_norm": 0.29983022809028625, "learning_rate": 1.5104752932662952e-05, "loss": 0.037, "step": 75860 }, { "grad_norm": 0.4750562012195587, "learning_rate": 1.5092912848779778e-05, "loss": 0.0365, "step": 75870 }, { "grad_norm": 0.4245259165763855, "learning_rate": 1.5081076582283194e-05, "loss": 0.0283, "step": 75880 }, { "grad_norm": 0.2609073519706726, "learning_rate": 1.5069244134467553e-05, "loss": 0.0267, "step": 75890 }, { "grad_norm": 0.27531898021698, "learning_rate": 1.5057415506626876e-05, "loss": 0.034, "step": 75900 }, { "grad_norm": 0.515151858329773, "learning_rate": 1.5045590700054668e-05, "loss": 0.0369, "step": 75910 }, { "grad_norm": 0.6398363709449768, "learning_rate": 1.5033769716044144e-05, "loss": 0.0276, "step": 75920 }, { "grad_norm": 0.5256572365760803, "learning_rate": 1.5021952555887975e-05, "loss": 0.0269, "step": 75930 }, { "grad_norm": 0.28607094287872314, "learning_rate": 1.50101392208785e-05, "loss": 0.0309, "step": 75940 }, { "grad_norm": 0.41823452711105347, "learning_rate": 1.4998329712307579e-05, "loss": 0.0247, "step": 75950 }, { "grad_norm": 0.4742841422557831, "learning_rate": 1.4986524031466708e-05, "loss": 0.0355, "step": 75960 }, { "grad_norm": 0.4780590832233429, "learning_rate": 1.4974722179646916e-05, "loss": 0.0222, "step": 75970 }, { "grad_norm": 0.5141295790672302, "learning_rate": 1.4962924158138847e-05, "loss": 0.0393, "step": 75980 }, { "grad_norm": 0.4309326410293579, "learning_rate": 1.4951129968232714e-05, "loss": 0.0311, "step": 75990 }, { "grad_norm": 0.6440114974975586, "learning_rate": 1.4939339611218329e-05, "loss": 0.0326, "step": 76000 }, { "grad_norm": 0.43044742941856384, "learning_rate": 1.4927553088385032e-05, "loss": 0.0351, "step": 76010 }, { "grad_norm": 0.48067614436149597, "learning_rate": 1.4915770401021801e-05, "loss": 0.0286, "step": 76020 }, { "grad_norm": 0.36127588152885437, "learning_rate": 1.4903991550417185e-05, "loss": 0.0434, "step": 76030 }, { "grad_norm": 0.37745401263237, "learning_rate": 1.4892216537859277e-05, "loss": 0.038, "step": 76040 }, { "grad_norm": 0.6840916872024536, "learning_rate": 1.4880445364635775e-05, "loss": 0.0313, "step": 76050 }, { "grad_norm": 1.015904188156128, "learning_rate": 1.4868678032033973e-05, "loss": 0.0368, "step": 76060 }, { "grad_norm": 0.36782538890838623, "learning_rate": 1.4856914541340739e-05, "loss": 0.0306, "step": 76070 }, { "grad_norm": 0.44072961807250977, "learning_rate": 1.4845154893842472e-05, "loss": 0.0301, "step": 76080 }, { "grad_norm": 0.6424534916877747, "learning_rate": 1.4833399090825234e-05, "loss": 0.0416, "step": 76090 }, { "grad_norm": 0.518117368221283, "learning_rate": 1.4821647133574562e-05, "loss": 0.0403, "step": 76100 }, { "grad_norm": 0.49168896675109863, "learning_rate": 1.4809899023375706e-05, "loss": 0.0296, "step": 76110 }, { "grad_norm": 0.46380332112312317, "learning_rate": 1.4798154761513366e-05, "loss": 0.0374, "step": 76120 }, { "grad_norm": 0.39901188015937805, "learning_rate": 1.4786414349271916e-05, "loss": 0.0335, "step": 76130 }, { "grad_norm": 0.6874129772186279, "learning_rate": 1.4774677787935226e-05, "loss": 0.0327, "step": 76140 }, { "grad_norm": 0.7198959589004517, "learning_rate": 1.4762945078786838e-05, "loss": 0.0324, "step": 76150 }, { "grad_norm": 0.4483817517757416, "learning_rate": 1.4751216223109777e-05, "loss": 0.0258, "step": 76160 }, { "grad_norm": 0.45157697796821594, "learning_rate": 1.4739491222186714e-05, "loss": 0.0352, "step": 76170 }, { "grad_norm": 0.5423084497451782, "learning_rate": 1.4727770077299879e-05, "loss": 0.0336, "step": 76180 }, { "grad_norm": 0.25968679785728455, "learning_rate": 1.4716052789731094e-05, "loss": 0.0357, "step": 76190 }, { "grad_norm": 0.6235288977622986, "learning_rate": 1.470433936076171e-05, "loss": 0.0308, "step": 76200 }, { "grad_norm": 0.4956217110157013, "learning_rate": 1.4692629791672719e-05, "loss": 0.0288, "step": 76210 }, { "grad_norm": 2.1745898723602295, "learning_rate": 1.468092408374464e-05, "loss": 0.0334, "step": 76220 }, { "grad_norm": 0.5589667558670044, "learning_rate": 1.4669222238257602e-05, "loss": 0.0434, "step": 76230 }, { "grad_norm": 0.3043499290943146, "learning_rate": 1.4657524256491301e-05, "loss": 0.0249, "step": 76240 }, { "grad_norm": 0.41780006885528564, "learning_rate": 1.464583013972502e-05, "loss": 0.0275, "step": 76250 }, { "grad_norm": 0.3188695013523102, "learning_rate": 1.4634139889237591e-05, "loss": 0.0281, "step": 76260 }, { "grad_norm": 0.35185369849205017, "learning_rate": 1.4622453506307448e-05, "loss": 0.0344, "step": 76270 }, { "grad_norm": 0.4298737049102783, "learning_rate": 1.461077099221262e-05, "loss": 0.0279, "step": 76280 }, { "grad_norm": 0.4517979621887207, "learning_rate": 1.4599092348230652e-05, "loss": 0.0321, "step": 76290 }, { "grad_norm": 0.5997466444969177, "learning_rate": 1.4587417575638724e-05, "loss": 0.0265, "step": 76300 }, { "grad_norm": 0.9529848694801331, "learning_rate": 1.4575746675713563e-05, "loss": 0.0313, "step": 76310 }, { "grad_norm": 0.49278056621551514, "learning_rate": 1.4564079649731504e-05, "loss": 0.0319, "step": 76320 }, { "grad_norm": 0.4449906647205353, "learning_rate": 1.4552416498968402e-05, "loss": 0.0281, "step": 76330 }, { "grad_norm": 0.5201361775398254, "learning_rate": 1.4540757224699747e-05, "loss": 0.0249, "step": 76340 }, { "grad_norm": 0.2863228917121887, "learning_rate": 1.452910182820056e-05, "loss": 0.0325, "step": 76350 }, { "grad_norm": 0.5809187293052673, "learning_rate": 1.4517450310745468e-05, "loss": 0.0306, "step": 76360 }, { "grad_norm": 0.3779900372028351, "learning_rate": 1.450580267360866e-05, "loss": 0.0281, "step": 76370 }, { "grad_norm": 0.1798432618379593, "learning_rate": 1.4494158918063922e-05, "loss": 0.0288, "step": 76380 }, { "grad_norm": 0.4514516592025757, "learning_rate": 1.4482519045384568e-05, "loss": 0.0275, "step": 76390 }, { "grad_norm": 0.48297977447509766, "learning_rate": 1.4470883056843548e-05, "loss": 0.0308, "step": 76400 }, { "grad_norm": 0.45935484766960144, "learning_rate": 1.4459250953713322e-05, "loss": 0.0364, "step": 76410 }, { "grad_norm": 0.3970798850059509, "learning_rate": 1.444762273726598e-05, "loss": 0.0308, "step": 76420 }, { "grad_norm": 0.3392438590526581, "learning_rate": 1.443599840877316e-05, "loss": 0.033, "step": 76430 }, { "grad_norm": 0.35142412781715393, "learning_rate": 1.4424377969506103e-05, "loss": 0.0309, "step": 76440 }, { "grad_norm": 0.6625918745994568, "learning_rate": 1.4412761420735565e-05, "loss": 0.0343, "step": 76450 }, { "grad_norm": 0.598503589630127, "learning_rate": 1.4401148763731953e-05, "loss": 0.0348, "step": 76460 }, { "grad_norm": 0.5774337649345398, "learning_rate": 1.4389539999765173e-05, "loss": 0.0318, "step": 76470 }, { "grad_norm": 0.4905831217765808, "learning_rate": 1.4377935130104758e-05, "loss": 0.0299, "step": 76480 }, { "grad_norm": 0.3179452121257782, "learning_rate": 1.4366334156019801e-05, "loss": 0.0331, "step": 76490 }, { "grad_norm": 0.42747223377227783, "learning_rate": 1.435473707877898e-05, "loss": 0.0336, "step": 76500 }, { "grad_norm": 0.3952957093715668, "learning_rate": 1.4343143899650508e-05, "loss": 0.0351, "step": 76510 }, { "grad_norm": 0.30435711145401, "learning_rate": 1.4331554619902204e-05, "loss": 0.0319, "step": 76520 }, { "grad_norm": 0.4542679190635681, "learning_rate": 1.4319969240801484e-05, "loss": 0.0263, "step": 76530 }, { "grad_norm": 0.4170825481414795, "learning_rate": 1.430838776361526e-05, "loss": 0.0357, "step": 76540 }, { "grad_norm": 0.3379678726196289, "learning_rate": 1.4296810189610093e-05, "loss": 0.043, "step": 76550 }, { "grad_norm": 0.35396769642829895, "learning_rate": 1.4285236520052087e-05, "loss": 0.0286, "step": 76560 }, { "grad_norm": 0.44837161898612976, "learning_rate": 1.4273666756206933e-05, "loss": 0.0288, "step": 76570 }, { "grad_norm": 0.49173492193222046, "learning_rate": 1.4262100899339848e-05, "loss": 0.0324, "step": 76580 }, { "grad_norm": 0.25905144214630127, "learning_rate": 1.4250538950715696e-05, "loss": 0.0277, "step": 76590 }, { "grad_norm": 0.4547574520111084, "learning_rate": 1.4238980911598842e-05, "loss": 0.0405, "step": 76600 }, { "grad_norm": 0.30852046608924866, "learning_rate": 1.4227426783253267e-05, "loss": 0.0333, "step": 76610 }, { "grad_norm": 0.4172898530960083, "learning_rate": 1.4215876566942521e-05, "loss": 0.0318, "step": 76620 }, { "grad_norm": 0.6478328108787537, "learning_rate": 1.4204330263929721e-05, "loss": 0.0262, "step": 76630 }, { "grad_norm": 0.4266428053379059, "learning_rate": 1.4192787875477537e-05, "loss": 0.0379, "step": 76640 }, { "grad_norm": 0.24768896400928497, "learning_rate": 1.4181249402848246e-05, "loss": 0.0289, "step": 76650 }, { "grad_norm": 0.357065349817276, "learning_rate": 1.4169714847303656e-05, "loss": 0.0309, "step": 76660 }, { "grad_norm": 0.6750878691673279, "learning_rate": 1.4158184210105197e-05, "loss": 0.0338, "step": 76670 }, { "grad_norm": 0.32825347781181335, "learning_rate": 1.4146657492513787e-05, "loss": 0.0303, "step": 76680 }, { "grad_norm": 0.26692116260528564, "learning_rate": 1.4135134695790047e-05, "loss": 0.0256, "step": 76690 }, { "grad_norm": 0.45018741488456726, "learning_rate": 1.4123615821194037e-05, "loss": 0.0364, "step": 76700 }, { "grad_norm": 0.37746870517730713, "learning_rate": 1.4112100869985473e-05, "loss": 0.0302, "step": 76710 }, { "grad_norm": 0.26450684666633606, "learning_rate": 1.4100589843423589e-05, "loss": 0.0309, "step": 76720 }, { "grad_norm": 0.3306693732738495, "learning_rate": 1.408908274276723e-05, "loss": 0.0367, "step": 76730 }, { "grad_norm": 0.261565238237381, "learning_rate": 1.4077579569274786e-05, "loss": 0.0307, "step": 76740 }, { "grad_norm": 0.40696778893470764, "learning_rate": 1.4066080324204245e-05, "loss": 0.0277, "step": 76750 }, { "grad_norm": 0.5079786777496338, "learning_rate": 1.405458500881312e-05, "loss": 0.036, "step": 76760 }, { "grad_norm": 0.44482940435409546, "learning_rate": 1.4043093624358538e-05, "loss": 0.0294, "step": 76770 }, { "grad_norm": 0.3290223479270935, "learning_rate": 1.4031606172097189e-05, "loss": 0.0452, "step": 76780 }, { "grad_norm": 0.4398788809776306, "learning_rate": 1.40201226532853e-05, "loss": 0.0375, "step": 76790 }, { "grad_norm": 0.2537786662578583, "learning_rate": 1.4008643069178701e-05, "loss": 0.0275, "step": 76800 }, { "grad_norm": 0.3629096746444702, "learning_rate": 1.399716742103278e-05, "loss": 0.0301, "step": 76810 }, { "grad_norm": 0.4052208364009857, "learning_rate": 1.398569571010252e-05, "loss": 0.0294, "step": 76820 }, { "grad_norm": 0.4535866677761078, "learning_rate": 1.3974227937642414e-05, "loss": 0.0386, "step": 76830 }, { "grad_norm": 0.360180139541626, "learning_rate": 1.3962764104906595e-05, "loss": 0.0242, "step": 76840 }, { "grad_norm": 0.30666810274124146, "learning_rate": 1.3951304213148696e-05, "loss": 0.0345, "step": 76850 }, { "grad_norm": 0.2949858605861664, "learning_rate": 1.3939848263621979e-05, "loss": 0.0262, "step": 76860 }, { "grad_norm": 0.8946612477302551, "learning_rate": 1.3928396257579219e-05, "loss": 0.0411, "step": 76870 }, { "grad_norm": 0.45161065459251404, "learning_rate": 1.3916948196272833e-05, "loss": 0.0367, "step": 76880 }, { "grad_norm": 0.47680675983428955, "learning_rate": 1.3905504080954734e-05, "loss": 0.0402, "step": 76890 }, { "grad_norm": 0.3428748846054077, "learning_rate": 1.3894063912876454e-05, "loss": 0.0339, "step": 76900 }, { "grad_norm": 0.4421502649784088, "learning_rate": 1.3882627693289046e-05, "loss": 0.0274, "step": 76910 }, { "grad_norm": 0.4230584502220154, "learning_rate": 1.3871195423443183e-05, "loss": 0.0318, "step": 76920 }, { "grad_norm": 0.38484832644462585, "learning_rate": 1.3859767104589044e-05, "loss": 0.031, "step": 76930 }, { "grad_norm": 0.6003599166870117, "learning_rate": 1.3848342737976467e-05, "loss": 0.0317, "step": 76940 }, { "grad_norm": 0.46838825941085815, "learning_rate": 1.3836922324854757e-05, "loss": 0.0324, "step": 76950 }, { "grad_norm": 0.6967800855636597, "learning_rate": 1.382550586647287e-05, "loss": 0.038, "step": 76960 }, { "grad_norm": 0.2181871086359024, "learning_rate": 1.3814093364079257e-05, "loss": 0.0304, "step": 76970 }, { "grad_norm": 0.5590296983718872, "learning_rate": 1.3802684818921991e-05, "loss": 0.0355, "step": 76980 }, { "grad_norm": 0.21708063781261444, "learning_rate": 1.3791280232248687e-05, "loss": 0.0337, "step": 76990 }, { "grad_norm": 0.29942968487739563, "learning_rate": 1.3779879605306555e-05, "loss": 0.0324, "step": 77000 }, { "grad_norm": 0.4301784038543701, "learning_rate": 1.376848293934232e-05, "loss": 0.0338, "step": 77010 }, { "grad_norm": 0.257730633020401, "learning_rate": 1.3757090235602316e-05, "loss": 0.0309, "step": 77020 }, { "grad_norm": 0.47157272696495056, "learning_rate": 1.3745701495332447e-05, "loss": 0.0346, "step": 77030 }, { "grad_norm": 0.45682892203330994, "learning_rate": 1.3734316719778145e-05, "loss": 0.0256, "step": 77040 }, { "grad_norm": 0.26910269260406494, "learning_rate": 1.3722935910184454e-05, "loss": 0.0269, "step": 77050 }, { "grad_norm": 0.38370606303215027, "learning_rate": 1.3711559067795921e-05, "loss": 0.0288, "step": 77060 }, { "grad_norm": 0.35077163577079773, "learning_rate": 1.3700186193856756e-05, "loss": 0.0366, "step": 77070 }, { "grad_norm": 0.41735315322875977, "learning_rate": 1.368881728961064e-05, "loss": 0.0272, "step": 77080 }, { "grad_norm": 0.37546172738075256, "learning_rate": 1.3677452356300885e-05, "loss": 0.0303, "step": 77090 }, { "grad_norm": 0.38252341747283936, "learning_rate": 1.3666091395170321e-05, "loss": 0.0358, "step": 77100 }, { "grad_norm": 0.46775007247924805, "learning_rate": 1.3654734407461384e-05, "loss": 0.033, "step": 77110 }, { "grad_norm": 0.590140163898468, "learning_rate": 1.3643381394416022e-05, "loss": 0.0277, "step": 77120 }, { "grad_norm": 0.4683465361595154, "learning_rate": 1.3632032357275838e-05, "loss": 0.034, "step": 77130 }, { "grad_norm": 0.31103000044822693, "learning_rate": 1.3620687297281898e-05, "loss": 0.0305, "step": 77140 }, { "grad_norm": 0.9186984896659851, "learning_rate": 1.3609346215674917e-05, "loss": 0.024, "step": 77150 }, { "grad_norm": 0.5705386996269226, "learning_rate": 1.3598009113695109e-05, "loss": 0.036, "step": 77160 }, { "grad_norm": 0.229298397898674, "learning_rate": 1.358667599258231e-05, "loss": 0.0282, "step": 77170 }, { "grad_norm": 0.3443889617919922, "learning_rate": 1.357534685357586e-05, "loss": 0.0354, "step": 77180 }, { "grad_norm": 0.4964650869369507, "learning_rate": 1.3564021697914715e-05, "loss": 0.0313, "step": 77190 }, { "grad_norm": 0.30828535556793213, "learning_rate": 1.3552700526837381e-05, "loss": 0.0318, "step": 77200 }, { "grad_norm": 0.27935415506362915, "learning_rate": 1.3541383341581925e-05, "loss": 0.026, "step": 77210 }, { "grad_norm": 0.553726077079773, "learning_rate": 1.3530070143385965e-05, "loss": 0.0278, "step": 77220 }, { "grad_norm": 0.3656569719314575, "learning_rate": 1.3518760933486702e-05, "loss": 0.0233, "step": 77230 }, { "grad_norm": 0.37652459740638733, "learning_rate": 1.3507455713120908e-05, "loss": 0.0296, "step": 77240 }, { "grad_norm": 0.28979969024658203, "learning_rate": 1.3496154483524876e-05, "loss": 0.0273, "step": 77250 }, { "grad_norm": 0.6698070168495178, "learning_rate": 1.3484857245934506e-05, "loss": 0.0362, "step": 77260 }, { "grad_norm": 0.27715548872947693, "learning_rate": 1.3473564001585248e-05, "loss": 0.0321, "step": 77270 }, { "grad_norm": 0.402254194021225, "learning_rate": 1.3462274751712129e-05, "loss": 0.037, "step": 77280 }, { "grad_norm": 0.3800174593925476, "learning_rate": 1.3450989497549698e-05, "loss": 0.0286, "step": 77290 }, { "grad_norm": 0.5826817154884338, "learning_rate": 1.3439708240332115e-05, "loss": 0.0373, "step": 77300 }, { "grad_norm": 0.4283222258090973, "learning_rate": 1.3428430981293043e-05, "loss": 0.0332, "step": 77310 }, { "grad_norm": 0.2879314720630646, "learning_rate": 1.3417157721665808e-05, "loss": 0.0347, "step": 77320 }, { "grad_norm": 0.3230898082256317, "learning_rate": 1.3405888462683186e-05, "loss": 0.0261, "step": 77330 }, { "grad_norm": 0.4223904609680176, "learning_rate": 1.3394623205577595e-05, "loss": 0.0325, "step": 77340 }, { "grad_norm": 0.6463868021965027, "learning_rate": 1.3383361951580963e-05, "loss": 0.0293, "step": 77350 }, { "grad_norm": 0.5646897554397583, "learning_rate": 1.337210470192483e-05, "loss": 0.0322, "step": 77360 }, { "grad_norm": 0.4109353721141815, "learning_rate": 1.3360851457840245e-05, "loss": 0.0319, "step": 77370 }, { "grad_norm": 0.28079041838645935, "learning_rate": 1.3349602220557856e-05, "loss": 0.0261, "step": 77380 }, { "grad_norm": 0.6010035276412964, "learning_rate": 1.3338356991307865e-05, "loss": 0.0391, "step": 77390 }, { "grad_norm": 0.5853558778762817, "learning_rate": 1.3327115771320048e-05, "loss": 0.0327, "step": 77400 }, { "grad_norm": 0.3343031704425812, "learning_rate": 1.3315878561823697e-05, "loss": 0.0244, "step": 77410 }, { "grad_norm": 0.6583582758903503, "learning_rate": 1.330464536404773e-05, "loss": 0.0295, "step": 77420 }, { "grad_norm": 0.3622463345527649, "learning_rate": 1.3293416179220563e-05, "loss": 0.035, "step": 77430 }, { "grad_norm": 0.5953729152679443, "learning_rate": 1.328219100857021e-05, "loss": 0.0249, "step": 77440 }, { "grad_norm": 0.35873153805732727, "learning_rate": 1.327096985332425e-05, "loss": 0.03, "step": 77450 }, { "grad_norm": 0.5142326354980469, "learning_rate": 1.3259752714709816e-05, "loss": 0.0299, "step": 77460 }, { "grad_norm": 0.7788637280464172, "learning_rate": 1.3248539593953569e-05, "loss": 0.0298, "step": 77470 }, { "grad_norm": 0.5935180187225342, "learning_rate": 1.323733049228178e-05, "loss": 0.0403, "step": 77480 }, { "grad_norm": 0.473471075296402, "learning_rate": 1.3226125410920276e-05, "loss": 0.0358, "step": 77490 }, { "grad_norm": 0.4474855363368988, "learning_rate": 1.3214924351094388e-05, "loss": 0.0303, "step": 77500 }, { "grad_norm": 0.3009568750858307, "learning_rate": 1.3203727314029069e-05, "loss": 0.0268, "step": 77510 }, { "grad_norm": 0.5112428069114685, "learning_rate": 1.3192534300948811e-05, "loss": 0.0363, "step": 77520 }, { "grad_norm": 0.43220531940460205, "learning_rate": 1.3181345313077675e-05, "loss": 0.0373, "step": 77530 }, { "grad_norm": 0.5204796195030212, "learning_rate": 1.3170160351639243e-05, "loss": 0.0339, "step": 77540 }, { "grad_norm": 0.4896087646484375, "learning_rate": 1.3158979417856721e-05, "loss": 0.0381, "step": 77550 }, { "grad_norm": 0.3987995982170105, "learning_rate": 1.31478025129528e-05, "loss": 0.0313, "step": 77560 }, { "grad_norm": 0.322073757648468, "learning_rate": 1.3136629638149794e-05, "loss": 0.029, "step": 77570 }, { "grad_norm": 0.43571212887763977, "learning_rate": 1.3125460794669547e-05, "loss": 0.0309, "step": 77580 }, { "grad_norm": 0.4069550037384033, "learning_rate": 1.3114295983733483e-05, "loss": 0.0288, "step": 77590 }, { "grad_norm": 0.3569151759147644, "learning_rate": 1.3103135206562534e-05, "loss": 0.0359, "step": 77600 }, { "grad_norm": 0.49020916223526, "learning_rate": 1.3091978464377264e-05, "loss": 0.0362, "step": 77610 }, { "grad_norm": 0.33369889855384827, "learning_rate": 1.3080825758397724e-05, "loss": 0.0312, "step": 77620 }, { "grad_norm": 0.6735764741897583, "learning_rate": 1.3069677089843568e-05, "loss": 0.0325, "step": 77630 }, { "grad_norm": 0.48267853260040283, "learning_rate": 1.305853245993401e-05, "loss": 0.035, "step": 77640 }, { "grad_norm": 0.6731436848640442, "learning_rate": 1.3047391869887809e-05, "loss": 0.0297, "step": 77650 }, { "grad_norm": 0.49862930178642273, "learning_rate": 1.3036255320923263e-05, "loss": 0.0281, "step": 77660 }, { "grad_norm": 0.3623940944671631, "learning_rate": 1.3025122814258273e-05, "loss": 0.0293, "step": 77670 }, { "grad_norm": 0.6067607998847961, "learning_rate": 1.3013994351110249e-05, "loss": 0.0338, "step": 77680 }, { "grad_norm": 0.36518198251724243, "learning_rate": 1.3002869932696194e-05, "loss": 0.0337, "step": 77690 }, { "grad_norm": 0.47057148814201355, "learning_rate": 1.2991749560232664e-05, "loss": 0.0316, "step": 77700 }, { "grad_norm": 0.24412265419960022, "learning_rate": 1.2980633234935768e-05, "loss": 0.0205, "step": 77710 }, { "grad_norm": 0.34879347681999207, "learning_rate": 1.2969520958021153e-05, "loss": 0.0328, "step": 77720 }, { "grad_norm": 0.3477097153663635, "learning_rate": 1.295841273070405e-05, "loss": 0.0286, "step": 77730 }, { "grad_norm": 0.4932372570037842, "learning_rate": 1.2947308554199255e-05, "loss": 0.0329, "step": 77740 }, { "grad_norm": 0.49863871932029724, "learning_rate": 1.2936208429721075e-05, "loss": 0.0398, "step": 77750 }, { "grad_norm": 0.34028440713882446, "learning_rate": 1.2925112358483416e-05, "loss": 0.0249, "step": 77760 }, { "grad_norm": 0.6062234044075012, "learning_rate": 1.2914020341699728e-05, "loss": 0.0339, "step": 77770 }, { "grad_norm": 0.3960340917110443, "learning_rate": 1.2902932380583032e-05, "loss": 0.0278, "step": 77780 }, { "grad_norm": 0.32077643275260925, "learning_rate": 1.2891848476345863e-05, "loss": 0.0339, "step": 77790 }, { "grad_norm": 0.36531364917755127, "learning_rate": 1.2880768630200368e-05, "loss": 0.0322, "step": 77800 }, { "grad_norm": 0.30429691076278687, "learning_rate": 1.286969284335819e-05, "loss": 0.0263, "step": 77810 }, { "grad_norm": 0.4626238942146301, "learning_rate": 1.2858621117030578e-05, "loss": 0.0248, "step": 77820 }, { "grad_norm": 0.42844921350479126, "learning_rate": 1.2847553452428324e-05, "loss": 0.0225, "step": 77830 }, { "grad_norm": 0.29855722188949585, "learning_rate": 1.283648985076178e-05, "loss": 0.0236, "step": 77840 }, { "grad_norm": 0.5350841283798218, "learning_rate": 1.2825430313240817e-05, "loss": 0.0366, "step": 77850 }, { "grad_norm": 0.5903857350349426, "learning_rate": 1.281437484107491e-05, "loss": 0.0307, "step": 77860 }, { "grad_norm": 0.42377251386642456, "learning_rate": 1.2803323435473058e-05, "loss": 0.0303, "step": 77870 }, { "grad_norm": 0.40059056878089905, "learning_rate": 1.2792276097643841e-05, "loss": 0.0295, "step": 77880 }, { "grad_norm": 0.5644316077232361, "learning_rate": 1.2781232828795342e-05, "loss": 0.0313, "step": 77890 }, { "grad_norm": 0.3211537301540375, "learning_rate": 1.2770193630135292e-05, "loss": 0.0363, "step": 77900 }, { "grad_norm": 0.6060666441917419, "learning_rate": 1.2759158502870877e-05, "loss": 0.0362, "step": 77910 }, { "grad_norm": 0.330778568983078, "learning_rate": 1.2748127448208913e-05, "loss": 0.0265, "step": 77920 }, { "grad_norm": 0.2594374716281891, "learning_rate": 1.2737100467355706e-05, "loss": 0.036, "step": 77930 }, { "grad_norm": 0.42926421761512756, "learning_rate": 1.2726077561517169e-05, "loss": 0.0384, "step": 77940 }, { "grad_norm": 0.28081634640693665, "learning_rate": 1.2715058731898755e-05, "loss": 0.0301, "step": 77950 }, { "grad_norm": 0.5421275496482849, "learning_rate": 1.2704043979705471e-05, "loss": 0.0365, "step": 77960 }, { "grad_norm": 0.50267493724823, "learning_rate": 1.269303330614185e-05, "loss": 0.0319, "step": 77970 }, { "grad_norm": 0.575829267501831, "learning_rate": 1.2682026712412015e-05, "loss": 0.0335, "step": 77980 }, { "grad_norm": 0.3387613594532013, "learning_rate": 1.2671024199719644e-05, "loss": 0.0267, "step": 77990 }, { "grad_norm": 0.5323688387870789, "learning_rate": 1.2660025769267936e-05, "loss": 0.0319, "step": 78000 }, { "grad_norm": 0.70014888048172, "learning_rate": 1.2649031422259667e-05, "loss": 0.0321, "step": 78010 }, { "grad_norm": 0.43772366642951965, "learning_rate": 1.2638041159897163e-05, "loss": 0.0287, "step": 78020 }, { "grad_norm": 0.20890918374061584, "learning_rate": 1.262705498338232e-05, "loss": 0.0346, "step": 78030 }, { "grad_norm": 0.33345112204551697, "learning_rate": 1.2616072893916542e-05, "loss": 0.0376, "step": 78040 }, { "grad_norm": 0.3291636109352112, "learning_rate": 1.2605094892700836e-05, "loss": 0.0338, "step": 78050 }, { "grad_norm": 0.3488782048225403, "learning_rate": 1.259412098093572e-05, "loss": 0.0312, "step": 78060 }, { "grad_norm": 0.22351887822151184, "learning_rate": 1.258315115982131e-05, "loss": 0.0306, "step": 78070 }, { "grad_norm": 0.49476873874664307, "learning_rate": 1.2572185430557203e-05, "loss": 0.0261, "step": 78080 }, { "grad_norm": 0.7208927273750305, "learning_rate": 1.2561223794342659e-05, "loss": 0.028, "step": 78090 }, { "grad_norm": 0.39858561754226685, "learning_rate": 1.255026625237637e-05, "loss": 0.037, "step": 78100 }, { "grad_norm": 0.43002212047576904, "learning_rate": 1.2539312805856678e-05, "loss": 0.0387, "step": 78110 }, { "grad_norm": 0.47872692346572876, "learning_rate": 1.25283634559814e-05, "loss": 0.0302, "step": 78120 }, { "grad_norm": 0.22423765063285828, "learning_rate": 1.2517418203947972e-05, "loss": 0.032, "step": 78130 }, { "grad_norm": 0.332162469625473, "learning_rate": 1.2506477050953308e-05, "loss": 0.0279, "step": 78140 }, { "grad_norm": 0.35009831190109253, "learning_rate": 1.2495539998193972e-05, "loss": 0.0328, "step": 78150 }, { "grad_norm": 0.4040219485759735, "learning_rate": 1.2484607046865975e-05, "loss": 0.0277, "step": 78160 }, { "grad_norm": 0.3713156282901764, "learning_rate": 1.2473678198164968e-05, "loss": 0.0268, "step": 78170 }, { "grad_norm": 0.6118651032447815, "learning_rate": 1.2462753453286075e-05, "loss": 0.0316, "step": 78180 }, { "grad_norm": 0.2986626923084259, "learning_rate": 1.2451832813424031e-05, "loss": 0.0539, "step": 78190 }, { "grad_norm": 0.5008013844490051, "learning_rate": 1.2440916279773095e-05, "loss": 0.0316, "step": 78200 }, { "grad_norm": 0.2868485450744629, "learning_rate": 1.2430003853527101e-05, "loss": 0.0396, "step": 78210 }, { "grad_norm": 0.7168808579444885, "learning_rate": 1.2419095535879383e-05, "loss": 0.0478, "step": 78220 }, { "grad_norm": 0.25537681579589844, "learning_rate": 1.2408191328022878e-05, "loss": 0.0346, "step": 78230 }, { "grad_norm": 0.508975088596344, "learning_rate": 1.2397291231150066e-05, "loss": 0.0311, "step": 78240 }, { "grad_norm": 0.868192195892334, "learning_rate": 1.2386395246452937e-05, "loss": 0.0337, "step": 78250 }, { "grad_norm": 0.28632181882858276, "learning_rate": 1.2375503375123083e-05, "loss": 0.0311, "step": 78260 }, { "grad_norm": 0.20851798355579376, "learning_rate": 1.2364615618351583e-05, "loss": 0.0392, "step": 78270 }, { "grad_norm": 0.4523394703865051, "learning_rate": 1.2353731977329169e-05, "loss": 0.0301, "step": 78280 }, { "grad_norm": 0.5083028078079224, "learning_rate": 1.2342852453246007e-05, "loss": 0.022, "step": 78290 }, { "grad_norm": 0.47218918800354004, "learning_rate": 1.2331977047291898e-05, "loss": 0.0333, "step": 78300 }, { "grad_norm": 0.6678805351257324, "learning_rate": 1.2321105760656137e-05, "loss": 0.0322, "step": 78310 }, { "grad_norm": 0.2232256531715393, "learning_rate": 1.2310238594527613e-05, "loss": 0.0366, "step": 78320 }, { "grad_norm": 0.7765194177627563, "learning_rate": 1.2299375550094693e-05, "loss": 0.0297, "step": 78330 }, { "grad_norm": 0.8437613844871521, "learning_rate": 1.2288516628545421e-05, "loss": 0.0319, "step": 78340 }, { "grad_norm": 0.3270348906517029, "learning_rate": 1.2277661831067255e-05, "loss": 0.0324, "step": 78350 }, { "grad_norm": 0.2726561427116394, "learning_rate": 1.2266811158847286e-05, "loss": 0.034, "step": 78360 }, { "grad_norm": 0.25197893381118774, "learning_rate": 1.2255964613072107e-05, "loss": 0.0237, "step": 78370 }, { "grad_norm": 0.4044759273529053, "learning_rate": 1.2245122194927905e-05, "loss": 0.0229, "step": 78380 }, { "grad_norm": 0.498393177986145, "learning_rate": 1.2234283905600364e-05, "loss": 0.0335, "step": 78390 }, { "grad_norm": 0.732778012752533, "learning_rate": 1.222344974627475e-05, "loss": 0.0391, "step": 78400 }, { "grad_norm": 0.5081837177276611, "learning_rate": 1.2212619718135875e-05, "loss": 0.0282, "step": 78410 }, { "grad_norm": 0.2839878797531128, "learning_rate": 1.2201793822368118e-05, "loss": 0.0363, "step": 78420 }, { "grad_norm": 0.5177971720695496, "learning_rate": 1.2190972060155337e-05, "loss": 0.0271, "step": 78430 }, { "grad_norm": 0.27854660153388977, "learning_rate": 1.2180154432681013e-05, "loss": 0.0309, "step": 78440 }, { "grad_norm": 0.32162219285964966, "learning_rate": 1.2169340941128148e-05, "loss": 0.0344, "step": 78450 }, { "grad_norm": 0.46838030219078064, "learning_rate": 1.2158531586679267e-05, "loss": 0.0255, "step": 78460 }, { "grad_norm": 0.45511874556541443, "learning_rate": 1.2147726370516477e-05, "loss": 0.0294, "step": 78470 }, { "grad_norm": 0.26179438829421997, "learning_rate": 1.2136925293821422e-05, "loss": 0.0357, "step": 78480 }, { "grad_norm": 0.6479784250259399, "learning_rate": 1.2126128357775306e-05, "loss": 0.0348, "step": 78490 }, { "grad_norm": 0.6835219264030457, "learning_rate": 1.2115335563558838e-05, "loss": 0.0346, "step": 78500 }, { "grad_norm": 0.27392256259918213, "learning_rate": 1.2104546912352327e-05, "loss": 0.0338, "step": 78510 }, { "grad_norm": 0.46804335713386536, "learning_rate": 1.2093762405335557e-05, "loss": 0.0247, "step": 78520 }, { "grad_norm": 0.40518444776535034, "learning_rate": 1.2082982043687974e-05, "loss": 0.0286, "step": 78530 }, { "grad_norm": 0.37950944900512695, "learning_rate": 1.207220582858845e-05, "loss": 0.0315, "step": 78540 }, { "grad_norm": 0.3525064289569855, "learning_rate": 1.206143376121549e-05, "loss": 0.0265, "step": 78550 }, { "grad_norm": 0.8842167854309082, "learning_rate": 1.2050665842747078e-05, "loss": 0.0284, "step": 78560 }, { "grad_norm": 0.5341871976852417, "learning_rate": 1.2039902074360804e-05, "loss": 0.0245, "step": 78570 }, { "grad_norm": 0.512498676776886, "learning_rate": 1.2029142457233755e-05, "loss": 0.0311, "step": 78580 }, { "grad_norm": 0.7272303104400635, "learning_rate": 1.2018386992542601e-05, "loss": 0.033, "step": 78590 }, { "grad_norm": 0.2963065505027771, "learning_rate": 1.2007635681463542e-05, "loss": 0.0377, "step": 78600 }, { "grad_norm": 0.18027977645397186, "learning_rate": 1.1996888525172334e-05, "loss": 0.0312, "step": 78610 }, { "grad_norm": 0.3850761353969574, "learning_rate": 1.198614552484425e-05, "loss": 0.0246, "step": 78620 }, { "grad_norm": 0.46249210834503174, "learning_rate": 1.197540668165415e-05, "loss": 0.0325, "step": 78630 }, { "grad_norm": 0.6515191793441772, "learning_rate": 1.1964671996776395e-05, "loss": 0.0389, "step": 78640 }, { "grad_norm": 0.4843474328517914, "learning_rate": 1.1953941471384922e-05, "loss": 0.0433, "step": 78650 }, { "grad_norm": 0.4895349442958832, "learning_rate": 1.1943215106653205e-05, "loss": 0.0368, "step": 78660 }, { "grad_norm": 0.4285663664340973, "learning_rate": 1.193249290375429e-05, "loss": 0.0294, "step": 78670 }, { "grad_norm": 0.39419907331466675, "learning_rate": 1.1921774863860696e-05, "loss": 0.0334, "step": 78680 }, { "grad_norm": 0.46529698371887207, "learning_rate": 1.1911060988144556e-05, "loss": 0.0374, "step": 78690 }, { "grad_norm": 0.35146623849868774, "learning_rate": 1.1900351277777533e-05, "loss": 0.0295, "step": 78700 }, { "grad_norm": 0.49736905097961426, "learning_rate": 1.1889645733930798e-05, "loss": 0.031, "step": 78710 }, { "grad_norm": 0.7303369045257568, "learning_rate": 1.1878944357775107e-05, "loss": 0.031, "step": 78720 }, { "grad_norm": 0.6416711807250977, "learning_rate": 1.1868247150480749e-05, "loss": 0.0238, "step": 78730 }, { "grad_norm": 0.6523575186729431, "learning_rate": 1.1857554113217567e-05, "loss": 0.0289, "step": 78740 }, { "grad_norm": 0.32737016677856445, "learning_rate": 1.1846865247154903e-05, "loss": 0.0336, "step": 78750 }, { "grad_norm": 0.277383029460907, "learning_rate": 1.1836180553461711e-05, "loss": 0.0304, "step": 78760 }, { "grad_norm": 0.39875736832618713, "learning_rate": 1.1825500033306419e-05, "loss": 0.0432, "step": 78770 }, { "grad_norm": 0.4688059389591217, "learning_rate": 1.1814823687857052e-05, "loss": 0.0376, "step": 78780 }, { "grad_norm": 0.4329470992088318, "learning_rate": 1.1804151518281153e-05, "loss": 0.0338, "step": 78790 }, { "grad_norm": 0.5570707321166992, "learning_rate": 1.179348352574583e-05, "loss": 0.0415, "step": 78800 }, { "grad_norm": 0.5387415289878845, "learning_rate": 1.1782819711417697e-05, "loss": 0.0337, "step": 78810 }, { "grad_norm": 0.26063936948776245, "learning_rate": 1.1772160076462951e-05, "loss": 0.0229, "step": 78820 }, { "grad_norm": 0.3870513141155243, "learning_rate": 1.1761504622047293e-05, "loss": 0.0361, "step": 78830 }, { "grad_norm": 0.4974122941493988, "learning_rate": 1.1750853349335999e-05, "loss": 0.0368, "step": 78840 }, { "grad_norm": 0.4955226182937622, "learning_rate": 1.1740206259493874e-05, "loss": 0.0331, "step": 78850 }, { "grad_norm": 0.31296661496162415, "learning_rate": 1.1729563353685286e-05, "loss": 0.0256, "step": 78860 }, { "grad_norm": 0.4902748465538025, "learning_rate": 1.1718924633074091e-05, "loss": 0.0309, "step": 78870 }, { "grad_norm": 0.4018852114677429, "learning_rate": 1.1708290098823765e-05, "loss": 0.0326, "step": 78880 }, { "grad_norm": 0.5998216271400452, "learning_rate": 1.1697659752097245e-05, "loss": 0.0313, "step": 78890 }, { "grad_norm": 0.4222221076488495, "learning_rate": 1.1687033594057068e-05, "loss": 0.043, "step": 78900 }, { "grad_norm": 0.38232555985450745, "learning_rate": 1.1676411625865297e-05, "loss": 0.0381, "step": 78910 }, { "grad_norm": 0.389360636472702, "learning_rate": 1.1665793848683542e-05, "loss": 0.0368, "step": 78920 }, { "grad_norm": 0.5637901425361633, "learning_rate": 1.1655180263672927e-05, "loss": 0.0454, "step": 78930 }, { "grad_norm": 0.4163879156112671, "learning_rate": 1.1644570871994142e-05, "loss": 0.034, "step": 78940 }, { "grad_norm": 0.5572828054428101, "learning_rate": 1.163396567480744e-05, "loss": 0.0284, "step": 78950 }, { "grad_norm": 0.49044373631477356, "learning_rate": 1.1623364673272552e-05, "loss": 0.0332, "step": 78960 }, { "grad_norm": 0.3133862316608429, "learning_rate": 1.1612767868548807e-05, "loss": 0.0292, "step": 78970 }, { "grad_norm": 0.43070098757743835, "learning_rate": 1.1602175261795057e-05, "loss": 0.031, "step": 78980 }, { "grad_norm": 0.39520400762557983, "learning_rate": 1.1591586854169696e-05, "loss": 0.0364, "step": 78990 }, { "grad_norm": 0.28559035062789917, "learning_rate": 1.158100264683064e-05, "loss": 0.0275, "step": 79000 }, { "grad_norm": 0.3737851083278656, "learning_rate": 1.1570422640935386e-05, "loss": 0.0353, "step": 79010 }, { "grad_norm": 0.2712760865688324, "learning_rate": 1.155984683764092e-05, "loss": 0.036, "step": 79020 }, { "grad_norm": 0.6925023794174194, "learning_rate": 1.1549275238103812e-05, "loss": 0.0319, "step": 79030 }, { "grad_norm": 0.28103911876678467, "learning_rate": 1.1538707843480152e-05, "loss": 0.0333, "step": 79040 }, { "grad_norm": 0.362923264503479, "learning_rate": 1.1528144654925593e-05, "loss": 0.0356, "step": 79050 }, { "grad_norm": 0.34411346912384033, "learning_rate": 1.1517585673595277e-05, "loss": 0.0289, "step": 79060 }, { "grad_norm": 0.3419989347457886, "learning_rate": 1.150703090064395e-05, "loss": 0.0355, "step": 79070 }, { "grad_norm": 0.32314568758010864, "learning_rate": 1.1496480337225835e-05, "loss": 0.0282, "step": 79080 }, { "grad_norm": 0.2893626093864441, "learning_rate": 1.1485933984494751e-05, "loss": 0.0246, "step": 79090 }, { "grad_norm": 0.5919041633605957, "learning_rate": 1.1475391843604e-05, "loss": 0.028, "step": 79100 }, { "grad_norm": 0.45102784037590027, "learning_rate": 1.1464853915706503e-05, "loss": 0.0389, "step": 79110 }, { "grad_norm": 0.5014251470565796, "learning_rate": 1.1454320201954626e-05, "loss": 0.0314, "step": 79120 }, { "grad_norm": 0.3258114159107208, "learning_rate": 1.1443790703500356e-05, "loss": 0.0332, "step": 79130 }, { "grad_norm": 0.38103416562080383, "learning_rate": 1.1433265421495154e-05, "loss": 0.036, "step": 79140 }, { "grad_norm": 0.42047855257987976, "learning_rate": 1.1422744357090065e-05, "loss": 0.0379, "step": 79150 }, { "grad_norm": 0.5679035782814026, "learning_rate": 1.1412227511435647e-05, "loss": 0.0343, "step": 79160 }, { "grad_norm": 0.47324231266975403, "learning_rate": 1.1401714885682025e-05, "loss": 0.0282, "step": 79170 }, { "grad_norm": 0.3258911073207855, "learning_rate": 1.1391206480978823e-05, "loss": 0.0251, "step": 79180 }, { "grad_norm": 0.33251476287841797, "learning_rate": 1.1380702298475227e-05, "loss": 0.035, "step": 79190 }, { "grad_norm": 0.3481500744819641, "learning_rate": 1.1370202339319985e-05, "loss": 0.0299, "step": 79200 }, { "grad_norm": 0.3017886281013489, "learning_rate": 1.1359706604661319e-05, "loss": 0.0404, "step": 79210 }, { "grad_norm": 0.5491799116134644, "learning_rate": 1.1349215095647042e-05, "loss": 0.0424, "step": 79220 }, { "grad_norm": 0.3424995541572571, "learning_rate": 1.1338727813424493e-05, "loss": 0.0255, "step": 79230 }, { "grad_norm": 0.30564913153648376, "learning_rate": 1.132824475914056e-05, "loss": 0.0269, "step": 79240 }, { "grad_norm": 0.7402077913284302, "learning_rate": 1.131776593394162e-05, "loss": 0.0279, "step": 79250 }, { "grad_norm": 0.4254230856895447, "learning_rate": 1.1307291338973657e-05, "loss": 0.0369, "step": 79260 }, { "grad_norm": 0.17996133863925934, "learning_rate": 1.1296820975382121e-05, "loss": 0.0302, "step": 79270 }, { "grad_norm": 0.32810547947883606, "learning_rate": 1.1286354844312074e-05, "loss": 0.0385, "step": 79280 }, { "grad_norm": 0.34153905510902405, "learning_rate": 1.1275892946908023e-05, "loss": 0.0355, "step": 79290 }, { "grad_norm": 0.39010173082351685, "learning_rate": 1.126543528431413e-05, "loss": 0.0457, "step": 79300 }, { "grad_norm": 0.4306032359600067, "learning_rate": 1.125498185767398e-05, "loss": 0.0294, "step": 79310 }, { "grad_norm": 0.31790241599082947, "learning_rate": 1.1244532668130781e-05, "loss": 0.0407, "step": 79320 }, { "grad_norm": 0.27800241112709045, "learning_rate": 1.1234087716827203e-05, "loss": 0.0325, "step": 79330 }, { "grad_norm": 0.5605385899543762, "learning_rate": 1.1223647004905524e-05, "loss": 0.0426, "step": 79340 }, { "grad_norm": 0.6443110108375549, "learning_rate": 1.1213210533507485e-05, "loss": 0.0391, "step": 79350 }, { "grad_norm": 0.3335920572280884, "learning_rate": 1.1202778303774447e-05, "loss": 0.0244, "step": 79360 }, { "grad_norm": 0.7271055579185486, "learning_rate": 1.1192350316847234e-05, "loss": 0.034, "step": 79370 }, { "grad_norm": 0.4773853123188019, "learning_rate": 1.1181926573866258e-05, "loss": 0.0347, "step": 79380 }, { "grad_norm": 0.38100939989089966, "learning_rate": 1.1171507075971416e-05, "loss": 0.0354, "step": 79390 }, { "grad_norm": 0.2804346978664398, "learning_rate": 1.1161091824302177e-05, "loss": 0.0273, "step": 79400 }, { "grad_norm": 0.4055361747741699, "learning_rate": 1.115068081999755e-05, "loss": 0.0369, "step": 79410 }, { "grad_norm": 0.6217944622039795, "learning_rate": 1.1140274064196071e-05, "loss": 0.0391, "step": 79420 }, { "grad_norm": 0.345188707113266, "learning_rate": 1.112987155803578e-05, "loss": 0.0305, "step": 79430 }, { "grad_norm": 0.8720867037773132, "learning_rate": 1.1119473302654298e-05, "loss": 0.0346, "step": 79440 }, { "grad_norm": 1.4961286783218384, "learning_rate": 1.1109079299188769e-05, "loss": 0.0372, "step": 79450 }, { "grad_norm": 0.36215940117836, "learning_rate": 1.1098689548775847e-05, "loss": 0.0343, "step": 79460 }, { "grad_norm": 0.5495557188987732, "learning_rate": 1.1088304052551762e-05, "loss": 0.0306, "step": 79470 }, { "grad_norm": 0.22979673743247986, "learning_rate": 1.1077922811652215e-05, "loss": 0.0261, "step": 79480 }, { "grad_norm": 0.46097585558891296, "learning_rate": 1.1067545827212534e-05, "loss": 0.0285, "step": 79490 }, { "grad_norm": 0.9432956576347351, "learning_rate": 1.1057173100367496e-05, "loss": 0.0265, "step": 79500 }, { "grad_norm": 0.39080679416656494, "learning_rate": 1.1046804632251474e-05, "loss": 0.0315, "step": 79510 }, { "grad_norm": 0.5352514982223511, "learning_rate": 1.1036440423998318e-05, "loss": 0.0264, "step": 79520 }, { "grad_norm": 0.5615611672401428, "learning_rate": 1.1026080476741469e-05, "loss": 0.0345, "step": 79530 }, { "grad_norm": 0.38689565658569336, "learning_rate": 1.1015724791613836e-05, "loss": 0.0267, "step": 79540 }, { "grad_norm": 0.5434698462486267, "learning_rate": 1.1005373369747951e-05, "loss": 0.0399, "step": 79550 }, { "grad_norm": 0.24356205761432648, "learning_rate": 1.09950262122758e-05, "loss": 0.0258, "step": 79560 }, { "grad_norm": 0.4151577055454254, "learning_rate": 1.0984683320328948e-05, "loss": 0.0335, "step": 79570 }, { "grad_norm": 0.23431003093719482, "learning_rate": 1.0974344695038458e-05, "loss": 0.0299, "step": 79580 }, { "grad_norm": 0.2853161692619324, "learning_rate": 1.0964010337534968e-05, "loss": 0.0356, "step": 79590 }, { "grad_norm": 0.49007338285446167, "learning_rate": 1.0953680248948611e-05, "loss": 0.0337, "step": 79600 }, { "grad_norm": 0.1489156186580658, "learning_rate": 1.0943354430409069e-05, "loss": 0.0256, "step": 79610 }, { "grad_norm": 0.4806749224662781, "learning_rate": 1.0933032883045574e-05, "loss": 0.0258, "step": 79620 }, { "grad_norm": 0.6833391189575195, "learning_rate": 1.0922715607986872e-05, "loss": 0.0422, "step": 79630 }, { "grad_norm": 0.29212355613708496, "learning_rate": 1.0912402606361234e-05, "loss": 0.0315, "step": 79640 }, { "grad_norm": 0.35349607467651367, "learning_rate": 1.090209387929647e-05, "loss": 0.0235, "step": 79650 }, { "grad_norm": 0.4567812383174896, "learning_rate": 1.0891789427919957e-05, "loss": 0.0264, "step": 79660 }, { "grad_norm": 0.2973170876502991, "learning_rate": 1.088148925335854e-05, "loss": 0.0231, "step": 79670 }, { "grad_norm": 0.4895362854003906, "learning_rate": 1.087119335673864e-05, "loss": 0.0387, "step": 79680 }, { "grad_norm": 0.4278155267238617, "learning_rate": 1.0860901739186208e-05, "loss": 0.0265, "step": 79690 }, { "grad_norm": 0.3186306953430176, "learning_rate": 1.085061440182673e-05, "loss": 0.0278, "step": 79700 }, { "grad_norm": 0.33250710368156433, "learning_rate": 1.0840331345785193e-05, "loss": 0.0294, "step": 79710 }, { "grad_norm": 0.2830451428890228, "learning_rate": 1.083005257218615e-05, "loss": 0.0254, "step": 79720 }, { "grad_norm": 0.3866153657436371, "learning_rate": 1.0819778082153642e-05, "loss": 0.0269, "step": 79730 }, { "grad_norm": 0.3340511918067932, "learning_rate": 1.0809507876811325e-05, "loss": 0.0351, "step": 79740 }, { "grad_norm": 0.2856810390949249, "learning_rate": 1.0799241957282292e-05, "loss": 0.0306, "step": 79750 }, { "grad_norm": 0.3961736261844635, "learning_rate": 1.078898032468923e-05, "loss": 0.0293, "step": 79760 }, { "grad_norm": 0.48773786425590515, "learning_rate": 1.0778722980154315e-05, "loss": 0.0337, "step": 79770 }, { "grad_norm": 0.3913592994213104, "learning_rate": 1.0768469924799302e-05, "loss": 0.0269, "step": 79780 }, { "grad_norm": 0.5714824795722961, "learning_rate": 1.0758221159745419e-05, "loss": 0.0402, "step": 79790 }, { "grad_norm": 0.45353826880455017, "learning_rate": 1.074797668611347e-05, "loss": 0.0283, "step": 79800 }, { "grad_norm": 0.24862554669380188, "learning_rate": 1.0737736505023777e-05, "loss": 0.0305, "step": 79810 }, { "grad_norm": 1.573638677597046, "learning_rate": 1.0727500617596203e-05, "loss": 0.0319, "step": 79820 }, { "grad_norm": 0.7453432083129883, "learning_rate": 1.0717269024950104e-05, "loss": 0.029, "step": 79830 }, { "grad_norm": 0.2472814917564392, "learning_rate": 1.070704172820442e-05, "loss": 0.0329, "step": 79840 }, { "grad_norm": 0.26759737730026245, "learning_rate": 1.0696818728477559e-05, "loss": 0.0285, "step": 79850 }, { "grad_norm": 0.22692617774009705, "learning_rate": 1.0686600026887511e-05, "loss": 0.0289, "step": 79860 }, { "grad_norm": 0.4018995463848114, "learning_rate": 1.0676385624551777e-05, "loss": 0.0363, "step": 79870 }, { "grad_norm": 1.7821606397628784, "learning_rate": 1.0666175522587402e-05, "loss": 0.0377, "step": 79880 }, { "grad_norm": 0.25485631823539734, "learning_rate": 1.0655969722110925e-05, "loss": 0.0259, "step": 79890 }, { "grad_norm": 1.0366158485412598, "learning_rate": 1.0645768224238439e-05, "loss": 0.0302, "step": 79900 }, { "grad_norm": 0.4960014224052429, "learning_rate": 1.063557103008559e-05, "loss": 0.0302, "step": 79910 }, { "grad_norm": 0.23966404795646667, "learning_rate": 1.0625378140767494e-05, "loss": 0.0252, "step": 79920 }, { "grad_norm": 0.38823601603507996, "learning_rate": 1.0615189557398846e-05, "loss": 0.0232, "step": 79930 }, { "grad_norm": 0.5271137952804565, "learning_rate": 1.060500528109386e-05, "loss": 0.0328, "step": 79940 }, { "grad_norm": 0.28249895572662354, "learning_rate": 1.0594825312966272e-05, "loss": 0.0282, "step": 79950 }, { "grad_norm": 0.38177067041397095, "learning_rate": 1.0584649654129336e-05, "loss": 0.0353, "step": 79960 }, { "grad_norm": 0.28517305850982666, "learning_rate": 1.0574478305695863e-05, "loss": 0.0188, "step": 79970 }, { "grad_norm": 0.3117709755897522, "learning_rate": 1.0564311268778154e-05, "loss": 0.0301, "step": 79980 }, { "grad_norm": 0.3877246677875519, "learning_rate": 1.0554148544488073e-05, "loss": 0.0322, "step": 79990 }, { "grad_norm": 0.5487785935401917, "learning_rate": 1.0543990133937004e-05, "loss": 0.0341, "step": 80000 }, { "grad_norm": 0.5173984169960022, "learning_rate": 1.0533836038235868e-05, "loss": 0.0307, "step": 80010 }, { "grad_norm": 0.5568438172340393, "learning_rate": 1.052368625849507e-05, "loss": 0.0257, "step": 80020 }, { "grad_norm": 0.4879275858402252, "learning_rate": 1.05135407958246e-05, "loss": 0.0259, "step": 80030 }, { "grad_norm": 0.40542683005332947, "learning_rate": 1.0503399651333934e-05, "loss": 0.0324, "step": 80040 }, { "grad_norm": 0.6521413326263428, "learning_rate": 1.04932628261321e-05, "loss": 0.0302, "step": 80050 }, { "grad_norm": 0.4285833537578583, "learning_rate": 1.0483130321327644e-05, "loss": 0.0294, "step": 80060 }, { "grad_norm": 0.20721328258514404, "learning_rate": 1.0473002138028654e-05, "loss": 0.0255, "step": 80070 }, { "grad_norm": 0.5643351674079895, "learning_rate": 1.0462878277342713e-05, "loss": 0.0279, "step": 80080 }, { "grad_norm": 0.39431053400039673, "learning_rate": 1.0452758740376972e-05, "loss": 0.0278, "step": 80090 }, { "grad_norm": 0.3293011486530304, "learning_rate": 1.0442643528238067e-05, "loss": 0.0265, "step": 80100 }, { "grad_norm": 0.3692523241043091, "learning_rate": 1.0432532642032188e-05, "loss": 0.0293, "step": 80110 }, { "grad_norm": 0.23743319511413574, "learning_rate": 1.042242608286505e-05, "loss": 0.0374, "step": 80120 }, { "grad_norm": 1.1262599229812622, "learning_rate": 1.041232385184191e-05, "loss": 0.031, "step": 80130 }, { "grad_norm": 0.4690139591693878, "learning_rate": 1.0402225950067495e-05, "loss": 0.0297, "step": 80140 }, { "grad_norm": 0.2790491282939911, "learning_rate": 1.0392132378646119e-05, "loss": 0.03, "step": 80150 }, { "grad_norm": 0.7185840606689453, "learning_rate": 1.0382043138681607e-05, "loss": 0.0381, "step": 80160 }, { "grad_norm": 0.35823509097099304, "learning_rate": 1.0371958231277284e-05, "loss": 0.0193, "step": 80170 }, { "grad_norm": 0.25553539395332336, "learning_rate": 1.0361877657536024e-05, "loss": 0.0296, "step": 80180 }, { "grad_norm": 0.43468233942985535, "learning_rate": 1.035180141856023e-05, "loss": 0.0381, "step": 80190 }, { "grad_norm": 0.4012661278247833, "learning_rate": 1.0341729515451836e-05, "loss": 0.0292, "step": 80200 }, { "grad_norm": 0.3554151952266693, "learning_rate": 1.0331661949312265e-05, "loss": 0.0312, "step": 80210 }, { "grad_norm": 0.45327189564704895, "learning_rate": 1.0321598721242514e-05, "loss": 0.0348, "step": 80220 }, { "grad_norm": 0.25908875465393066, "learning_rate": 1.0311539832343064e-05, "loss": 0.0335, "step": 80230 }, { "grad_norm": 0.32916200160980225, "learning_rate": 1.0301485283713942e-05, "loss": 0.0315, "step": 80240 }, { "grad_norm": 0.46503081917762756, "learning_rate": 1.0291435076454703e-05, "loss": 0.036, "step": 80250 }, { "grad_norm": 0.6081011891365051, "learning_rate": 1.028138921166444e-05, "loss": 0.0343, "step": 80260 }, { "grad_norm": 0.33145585656166077, "learning_rate": 1.0271347690441719e-05, "loss": 0.0254, "step": 80270 }, { "grad_norm": 0.3021058142185211, "learning_rate": 1.0261310513884698e-05, "loss": 0.0356, "step": 80280 }, { "grad_norm": 0.3769415020942688, "learning_rate": 1.0251277683090998e-05, "loss": 0.0312, "step": 80290 }, { "grad_norm": 0.39706090092658997, "learning_rate": 1.0241249199157826e-05, "loss": 0.0383, "step": 80300 }, { "grad_norm": 0.23200544714927673, "learning_rate": 1.0231225063181837e-05, "loss": 0.0244, "step": 80310 }, { "grad_norm": 0.6041087508201599, "learning_rate": 1.0221205276259316e-05, "loss": 0.0317, "step": 80320 }, { "grad_norm": 0.43763694167137146, "learning_rate": 1.0211189839485958e-05, "loss": 0.0339, "step": 80330 }, { "grad_norm": 0.8028150200843811, "learning_rate": 1.0201178753957075e-05, "loss": 0.0336, "step": 80340 }, { "grad_norm": 0.39948970079421997, "learning_rate": 1.0191172020767432e-05, "loss": 0.0308, "step": 80350 }, { "grad_norm": 0.28366637229919434, "learning_rate": 1.0181169641011362e-05, "loss": 0.0257, "step": 80360 }, { "grad_norm": 0.4046427309513092, "learning_rate": 1.0171171615782721e-05, "loss": 0.0352, "step": 80370 }, { "grad_norm": 0.3505054712295532, "learning_rate": 1.0161177946174866e-05, "loss": 0.0364, "step": 80380 }, { "grad_norm": 0.27755653858184814, "learning_rate": 1.0151188633280706e-05, "loss": 0.026, "step": 80390 }, { "grad_norm": 0.3529011607170105, "learning_rate": 1.0141203678192635e-05, "loss": 0.0366, "step": 80400 }, { "grad_norm": 0.49263378977775574, "learning_rate": 1.0131223082002612e-05, "loss": 0.0372, "step": 80410 }, { "grad_norm": 0.46115773916244507, "learning_rate": 1.012124684580208e-05, "loss": 0.0258, "step": 80420 }, { "grad_norm": 0.4325755834579468, "learning_rate": 1.0111274970682039e-05, "loss": 0.0311, "step": 80430 }, { "grad_norm": 0.39826998114585876, "learning_rate": 1.010130745773299e-05, "loss": 0.029, "step": 80440 }, { "grad_norm": 0.4604857563972473, "learning_rate": 1.0091344308044986e-05, "loss": 0.0334, "step": 80450 }, { "grad_norm": 0.3761232793331146, "learning_rate": 1.0081385522707554e-05, "loss": 0.0238, "step": 80460 }, { "grad_norm": 0.5407357811927795, "learning_rate": 1.0071431102809803e-05, "loss": 0.0323, "step": 80470 }, { "grad_norm": 0.2874642610549927, "learning_rate": 1.0061481049440297e-05, "loss": 0.0337, "step": 80480 }, { "grad_norm": 0.3692770302295685, "learning_rate": 1.0051535363687187e-05, "loss": 0.0225, "step": 80490 }, { "grad_norm": 0.411050021648407, "learning_rate": 1.0041594046638087e-05, "loss": 0.0268, "step": 80500 }, { "grad_norm": 0.3512764573097229, "learning_rate": 1.0031657099380204e-05, "loss": 0.0262, "step": 80510 }, { "grad_norm": 0.36743104457855225, "learning_rate": 1.0021724523000204e-05, "loss": 0.0334, "step": 80520 }, { "grad_norm": 0.43136394023895264, "learning_rate": 1.001179631858431e-05, "loss": 0.0325, "step": 80530 }, { "grad_norm": 0.41413265466690063, "learning_rate": 1.0001872487218245e-05, "loss": 0.0286, "step": 80540 }, { "grad_norm": 0.19149009883403778, "learning_rate": 9.991953029987272e-06, "loss": 0.0223, "step": 80550 }, { "grad_norm": 0.39625856280326843, "learning_rate": 9.982037947976147e-06, "loss": 0.0371, "step": 80560 }, { "grad_norm": 0.5232436060905457, "learning_rate": 9.972127242269208e-06, "loss": 0.0358, "step": 80570 }, { "grad_norm": 0.30488836765289307, "learning_rate": 9.962220913950238e-06, "loss": 0.0428, "step": 80580 }, { "grad_norm": 0.4424018859863281, "learning_rate": 9.952318964102591e-06, "loss": 0.0469, "step": 80590 }, { "grad_norm": 0.5359187126159668, "learning_rate": 9.942421393809148e-06, "loss": 0.0375, "step": 80600 }, { "grad_norm": 0.3146466910839081, "learning_rate": 9.932528204152264e-06, "loss": 0.0283, "step": 80610 }, { "grad_norm": 0.5098817348480225, "learning_rate": 9.922639396213856e-06, "loss": 0.0529, "step": 80620 }, { "grad_norm": 0.6026419401168823, "learning_rate": 9.912754971075344e-06, "loss": 0.0385, "step": 80630 }, { "grad_norm": 0.3945809006690979, "learning_rate": 9.902874929817696e-06, "loss": 0.0309, "step": 80640 }, { "grad_norm": 0.30868667364120483, "learning_rate": 9.892999273521342e-06, "loss": 0.0285, "step": 80650 }, { "grad_norm": 0.3839153051376343, "learning_rate": 9.883128003266302e-06, "loss": 0.0321, "step": 80660 }, { "grad_norm": 0.3765662610530853, "learning_rate": 9.87326112013206e-06, "loss": 0.0248, "step": 80670 }, { "grad_norm": 0.4475438594818115, "learning_rate": 9.863398625197662e-06, "loss": 0.0308, "step": 80680 }, { "grad_norm": 0.5669218301773071, "learning_rate": 9.853540519541615e-06, "loss": 0.0257, "step": 80690 }, { "grad_norm": 0.7931339740753174, "learning_rate": 9.843686804242053e-06, "loss": 0.0391, "step": 80700 }, { "grad_norm": 0.43118324875831604, "learning_rate": 9.833837480376506e-06, "loss": 0.0398, "step": 80710 }, { "grad_norm": 0.48360541462898254, "learning_rate": 9.82399254902212e-06, "loss": 0.0407, "step": 80720 }, { "grad_norm": 0.4722409248352051, "learning_rate": 9.814152011255494e-06, "loss": 0.0308, "step": 80730 }, { "grad_norm": 0.3648313581943512, "learning_rate": 9.804315868152797e-06, "loss": 0.0288, "step": 80740 }, { "grad_norm": 0.49412891268730164, "learning_rate": 9.79448412078966e-06, "loss": 0.0308, "step": 80750 }, { "grad_norm": 0.5667486786842346, "learning_rate": 9.784656770241312e-06, "loss": 0.0301, "step": 80760 }, { "grad_norm": 0.2981190085411072, "learning_rate": 9.77483381758243e-06, "loss": 0.0229, "step": 80770 }, { "grad_norm": 0.4103641211986542, "learning_rate": 9.76501526388725e-06, "loss": 0.0327, "step": 80780 }, { "grad_norm": 0.26754283905029297, "learning_rate": 9.7552011102295e-06, "loss": 0.0457, "step": 80790 }, { "grad_norm": 0.32675617933273315, "learning_rate": 9.745391357682449e-06, "loss": 0.0307, "step": 80800 }, { "grad_norm": 0.14526911079883575, "learning_rate": 9.735586007318875e-06, "loss": 0.0411, "step": 80810 }, { "grad_norm": 0.3038763701915741, "learning_rate": 9.725785060211096e-06, "loss": 0.0285, "step": 80820 }, { "grad_norm": 0.66888028383255, "learning_rate": 9.715988517430896e-06, "loss": 0.0357, "step": 80830 }, { "grad_norm": 0.5392675399780273, "learning_rate": 9.706196380049625e-06, "loss": 0.0242, "step": 80840 }, { "grad_norm": 0.3037625253200531, "learning_rate": 9.69640864913815e-06, "loss": 0.0295, "step": 80850 }, { "grad_norm": 0.3134514391422272, "learning_rate": 9.686625325766818e-06, "loss": 0.027, "step": 80860 }, { "grad_norm": 0.20749443769454956, "learning_rate": 9.67684641100554e-06, "loss": 0.0293, "step": 80870 }, { "grad_norm": 0.36477044224739075, "learning_rate": 9.667071905923692e-06, "loss": 0.0386, "step": 80880 }, { "grad_norm": 0.5679478049278259, "learning_rate": 9.657301811590242e-06, "loss": 0.028, "step": 80890 }, { "grad_norm": 0.3976731598377228, "learning_rate": 9.647536129073603e-06, "loss": 0.0291, "step": 80900 }, { "grad_norm": 0.3178142309188843, "learning_rate": 9.637774859441751e-06, "loss": 0.024, "step": 80910 }, { "grad_norm": 0.5887531042098999, "learning_rate": 9.628018003762152e-06, "loss": 0.0256, "step": 80920 }, { "grad_norm": 0.5892062187194824, "learning_rate": 9.618265563101813e-06, "loss": 0.0318, "step": 80930 }, { "grad_norm": 0.3944726586341858, "learning_rate": 9.608517538527211e-06, "loss": 0.0297, "step": 80940 }, { "grad_norm": 0.33325135707855225, "learning_rate": 9.598773931104431e-06, "loss": 0.0357, "step": 80950 }, { "grad_norm": 0.7904537320137024, "learning_rate": 9.589034741898983e-06, "loss": 0.0331, "step": 80960 }, { "grad_norm": 0.43218183517456055, "learning_rate": 9.579299971975947e-06, "loss": 0.0278, "step": 80970 }, { "grad_norm": 0.28571370244026184, "learning_rate": 9.569569622399882e-06, "loss": 0.0261, "step": 80980 }, { "grad_norm": 0.30541178584098816, "learning_rate": 9.559843694234916e-06, "loss": 0.0216, "step": 80990 }, { "grad_norm": 0.44647839665412903, "learning_rate": 9.550122188544625e-06, "loss": 0.0301, "step": 81000 }, { "grad_norm": 0.38281816244125366, "learning_rate": 9.540405106392152e-06, "loss": 0.0276, "step": 81010 }, { "grad_norm": 0.3651387393474579, "learning_rate": 9.53069244884015e-06, "loss": 0.0313, "step": 81020 }, { "grad_norm": 0.34953683614730835, "learning_rate": 9.52098421695079e-06, "loss": 0.0266, "step": 81030 }, { "grad_norm": 0.7259478569030762, "learning_rate": 9.511280411785728e-06, "loss": 0.0279, "step": 81040 }, { "grad_norm": 0.34760889410972595, "learning_rate": 9.50158103440616e-06, "loss": 0.0343, "step": 81050 }, { "grad_norm": 0.32060229778289795, "learning_rate": 9.491886085872815e-06, "loss": 0.0269, "step": 81060 }, { "grad_norm": 0.43093156814575195, "learning_rate": 9.482195567245895e-06, "loss": 0.0295, "step": 81070 }, { "grad_norm": 0.43632057309150696, "learning_rate": 9.472509479585151e-06, "loss": 0.0374, "step": 81080 }, { "grad_norm": 0.3977482318878174, "learning_rate": 9.462827823949833e-06, "loss": 0.0322, "step": 81090 }, { "grad_norm": 0.13897274434566498, "learning_rate": 9.453150601398735e-06, "loss": 0.0284, "step": 81100 }, { "grad_norm": 0.2761220932006836, "learning_rate": 9.44347781299011e-06, "loss": 0.0264, "step": 81110 }, { "grad_norm": 0.3233075737953186, "learning_rate": 9.43380945978179e-06, "loss": 0.0308, "step": 81120 }, { "grad_norm": 0.4242502450942993, "learning_rate": 9.424145542831047e-06, "loss": 0.0234, "step": 81130 }, { "grad_norm": 0.3247656226158142, "learning_rate": 9.414486063194778e-06, "loss": 0.0297, "step": 81140 }, { "grad_norm": 0.5973108410835266, "learning_rate": 9.404831021929273e-06, "loss": 0.0314, "step": 81150 }, { "grad_norm": 0.426374226808548, "learning_rate": 9.395180420090427e-06, "loss": 0.0287, "step": 81160 }, { "grad_norm": 0.2917173504829407, "learning_rate": 9.385534258733591e-06, "loss": 0.0221, "step": 81170 }, { "grad_norm": 0.4479687809944153, "learning_rate": 9.375892538913677e-06, "loss": 0.0265, "step": 81180 }, { "grad_norm": 0.18459387123584747, "learning_rate": 9.36625526168507e-06, "loss": 0.0302, "step": 81190 }, { "grad_norm": 0.3367881774902344, "learning_rate": 9.356622428101686e-06, "loss": 0.0315, "step": 81200 }, { "grad_norm": 0.39399856328964233, "learning_rate": 9.346994039216972e-06, "loss": 0.0266, "step": 81210 }, { "grad_norm": 0.2396334558725357, "learning_rate": 9.337370096083886e-06, "loss": 0.0216, "step": 81220 }, { "grad_norm": 0.28525570034980774, "learning_rate": 9.327750599754854e-06, "loss": 0.0299, "step": 81230 }, { "grad_norm": 0.26815739274024963, "learning_rate": 9.318135551281875e-06, "loss": 0.032, "step": 81240 }, { "grad_norm": 0.32001370191574097, "learning_rate": 9.308524951716424e-06, "loss": 0.0329, "step": 81250 }, { "grad_norm": 0.4900056719779968, "learning_rate": 9.298918802109497e-06, "loss": 0.03, "step": 81260 }, { "grad_norm": 0.7920191884040833, "learning_rate": 9.289317103511625e-06, "loss": 0.0313, "step": 81270 }, { "grad_norm": 0.5075933337211609, "learning_rate": 9.279719856972835e-06, "loss": 0.0278, "step": 81280 }, { "grad_norm": 0.3329657018184662, "learning_rate": 9.270127063542645e-06, "loss": 0.0349, "step": 81290 }, { "grad_norm": 0.34033241868019104, "learning_rate": 9.260538724270124e-06, "loss": 0.0282, "step": 81300 }, { "grad_norm": 0.9556959867477417, "learning_rate": 9.250954840203852e-06, "loss": 0.0291, "step": 81310 }, { "grad_norm": 0.43816646933555603, "learning_rate": 9.241375412391878e-06, "loss": 0.0295, "step": 81320 }, { "grad_norm": 0.2877780497074127, "learning_rate": 9.23180044188181e-06, "loss": 0.0272, "step": 81330 }, { "grad_norm": 0.3749512732028961, "learning_rate": 9.222229929720755e-06, "loss": 0.0268, "step": 81340 }, { "grad_norm": 0.44056427478790283, "learning_rate": 9.212663876955336e-06, "loss": 0.0432, "step": 81350 }, { "grad_norm": 0.6672977209091187, "learning_rate": 9.20310228463166e-06, "loss": 0.0415, "step": 81360 }, { "grad_norm": 0.3663373589515686, "learning_rate": 9.19354515379539e-06, "loss": 0.0282, "step": 81370 }, { "grad_norm": 0.4092133045196533, "learning_rate": 9.183992485491666e-06, "loss": 0.0277, "step": 81380 }, { "grad_norm": 0.7877443432807922, "learning_rate": 9.174444280765154e-06, "loss": 0.0416, "step": 81390 }, { "grad_norm": 0.5103201270103455, "learning_rate": 9.164900540660032e-06, "loss": 0.0283, "step": 81400 }, { "grad_norm": 0.754573404788971, "learning_rate": 9.155361266220008e-06, "loss": 0.0345, "step": 81410 }, { "grad_norm": 0.3973429799079895, "learning_rate": 9.145826458488254e-06, "loss": 0.03, "step": 81420 }, { "grad_norm": 0.3118398189544678, "learning_rate": 9.136296118507509e-06, "loss": 0.0305, "step": 81430 }, { "grad_norm": 0.33026692271232605, "learning_rate": 9.126770247319966e-06, "loss": 0.0298, "step": 81440 }, { "grad_norm": 0.3533974885940552, "learning_rate": 9.11724884596738e-06, "loss": 0.0181, "step": 81450 }, { "grad_norm": 0.513103723526001, "learning_rate": 9.107731915490996e-06, "loss": 0.0311, "step": 81460 }, { "grad_norm": 0.3857732117176056, "learning_rate": 9.098219456931579e-06, "loss": 0.0235, "step": 81470 }, { "grad_norm": 0.5338658094406128, "learning_rate": 9.088711471329376e-06, "loss": 0.035, "step": 81480 }, { "grad_norm": 0.900566816329956, "learning_rate": 9.079207959724184e-06, "loss": 0.0373, "step": 81490 }, { "grad_norm": 0.32396477460861206, "learning_rate": 9.069708923155279e-06, "loss": 0.0335, "step": 81500 }, { "grad_norm": 0.6154868006706238, "learning_rate": 9.060214362661468e-06, "loss": 0.032, "step": 81510 }, { "grad_norm": 0.28560149669647217, "learning_rate": 9.050724279281064e-06, "loss": 0.0298, "step": 81520 }, { "grad_norm": 0.4288840889930725, "learning_rate": 9.041238674051894e-06, "loss": 0.0329, "step": 81530 }, { "grad_norm": 0.5824549198150635, "learning_rate": 9.031757548011277e-06, "loss": 0.0397, "step": 81540 }, { "grad_norm": 1.1498088836669922, "learning_rate": 9.022280902196051e-06, "loss": 0.0278, "step": 81550 }, { "grad_norm": 0.24793240427970886, "learning_rate": 9.012808737642592e-06, "loss": 0.0282, "step": 81560 }, { "grad_norm": 0.3180687427520752, "learning_rate": 9.003341055386739e-06, "loss": 0.0365, "step": 81570 }, { "grad_norm": 0.47846174240112305, "learning_rate": 8.993877856463862e-06, "loss": 0.0271, "step": 81580 }, { "grad_norm": 0.17551074922084808, "learning_rate": 8.984419141908856e-06, "loss": 0.0236, "step": 81590 }, { "grad_norm": 0.2825135588645935, "learning_rate": 8.974964912756117e-06, "loss": 0.0284, "step": 81600 }, { "grad_norm": 0.6791524887084961, "learning_rate": 8.965515170039523e-06, "loss": 0.0353, "step": 81610 }, { "grad_norm": 0.23630815744400024, "learning_rate": 8.956069914792514e-06, "loss": 0.0237, "step": 81620 }, { "grad_norm": 0.31102970242500305, "learning_rate": 8.946629148047969e-06, "loss": 0.0297, "step": 81630 }, { "grad_norm": 0.4533514380455017, "learning_rate": 8.93719287083834e-06, "loss": 0.0368, "step": 81640 }, { "grad_norm": 0.3656459152698517, "learning_rate": 8.927761084195569e-06, "loss": 0.0309, "step": 81650 }, { "grad_norm": 0.34867343306541443, "learning_rate": 8.918333789151096e-06, "loss": 0.0291, "step": 81660 }, { "grad_norm": 0.40242865681648254, "learning_rate": 8.90891098673587e-06, "loss": 0.0279, "step": 81670 }, { "grad_norm": 0.2052965760231018, "learning_rate": 8.89949267798037e-06, "loss": 0.0241, "step": 81680 }, { "grad_norm": 0.24520714581012726, "learning_rate": 8.890078863914542e-06, "loss": 0.0311, "step": 81690 }, { "grad_norm": 0.32365578413009644, "learning_rate": 8.880669545567894e-06, "loss": 0.0258, "step": 81700 }, { "grad_norm": 0.25229907035827637, "learning_rate": 8.871264723969381e-06, "loss": 0.0312, "step": 81710 }, { "grad_norm": 0.7192583084106445, "learning_rate": 8.86186440014754e-06, "loss": 0.0414, "step": 81720 }, { "grad_norm": 1.8721672296524048, "learning_rate": 8.852468575130351e-06, "loss": 0.0341, "step": 81730 }, { "grad_norm": 0.23455457389354706, "learning_rate": 8.843077249945347e-06, "loss": 0.0271, "step": 81740 }, { "grad_norm": 0.506359875202179, "learning_rate": 8.833690425619523e-06, "loss": 0.0263, "step": 81750 }, { "grad_norm": 0.3600434362888336, "learning_rate": 8.82430810317942e-06, "loss": 0.0349, "step": 81760 }, { "grad_norm": 0.518632709980011, "learning_rate": 8.814930283651074e-06, "loss": 0.0271, "step": 81770 }, { "grad_norm": 0.21645256876945496, "learning_rate": 8.805556968060048e-06, "loss": 0.0305, "step": 81780 }, { "grad_norm": 0.45832058787345886, "learning_rate": 8.796188157431362e-06, "loss": 0.034, "step": 81790 }, { "grad_norm": 0.3977256417274475, "learning_rate": 8.786823852789595e-06, "loss": 0.0238, "step": 81800 }, { "grad_norm": 0.5387120842933655, "learning_rate": 8.777464055158813e-06, "loss": 0.0343, "step": 81810 }, { "grad_norm": 0.18682895600795746, "learning_rate": 8.768108765562578e-06, "loss": 0.0294, "step": 81820 }, { "grad_norm": 0.3427610695362091, "learning_rate": 8.758757985023975e-06, "loss": 0.0292, "step": 81830 }, { "grad_norm": 0.19304369390010834, "learning_rate": 8.749411714565598e-06, "loss": 0.0214, "step": 81840 }, { "grad_norm": 0.44681787490844727, "learning_rate": 8.74006995520954e-06, "loss": 0.0361, "step": 81850 }, { "grad_norm": 0.7993739247322083, "learning_rate": 8.730732707977397e-06, "loss": 0.0339, "step": 81860 }, { "grad_norm": 0.5893653035163879, "learning_rate": 8.721399973890281e-06, "loss": 0.0289, "step": 81870 }, { "grad_norm": 0.2908964455127716, "learning_rate": 8.712071753968792e-06, "loss": 0.0289, "step": 81880 }, { "grad_norm": 0.5534257888793945, "learning_rate": 8.702748049233072e-06, "loss": 0.0327, "step": 81890 }, { "grad_norm": 0.3270368278026581, "learning_rate": 8.693428860702718e-06, "loss": 0.0243, "step": 81900 }, { "grad_norm": 0.47653621435165405, "learning_rate": 8.684114189396898e-06, "loss": 0.0288, "step": 81910 }, { "grad_norm": 0.19805096089839935, "learning_rate": 8.674804036334222e-06, "loss": 0.026, "step": 81920 }, { "grad_norm": 0.522064208984375, "learning_rate": 8.665498402532857e-06, "loss": 0.0253, "step": 81930 }, { "grad_norm": 0.4551360607147217, "learning_rate": 8.656197289010432e-06, "loss": 0.0318, "step": 81940 }, { "grad_norm": 0.4162828326225281, "learning_rate": 8.646900696784122e-06, "loss": 0.0222, "step": 81950 }, { "grad_norm": 0.6562318801879883, "learning_rate": 8.637608626870553e-06, "loss": 0.0309, "step": 81960 }, { "grad_norm": 0.42027854919433594, "learning_rate": 8.628321080285944e-06, "loss": 0.0302, "step": 81970 }, { "grad_norm": 0.5380266308784485, "learning_rate": 8.619038058045925e-06, "loss": 0.0459, "step": 81980 }, { "grad_norm": 0.3723871409893036, "learning_rate": 8.609759561165704e-06, "loss": 0.0257, "step": 81990 }, { "grad_norm": 0.5168229937553406, "learning_rate": 8.600485590659934e-06, "loss": 0.0321, "step": 82000 }, { "grad_norm": 0.913411021232605, "learning_rate": 8.591216147542813e-06, "loss": 0.029, "step": 82010 }, { "grad_norm": 0.35801824927330017, "learning_rate": 8.581951232828045e-06, "loss": 0.0445, "step": 82020 }, { "grad_norm": 0.45217376947402954, "learning_rate": 8.572690847528825e-06, "loss": 0.0286, "step": 82030 }, { "grad_norm": 0.6624726057052612, "learning_rate": 8.563434992657838e-06, "loss": 0.0284, "step": 82040 }, { "grad_norm": 0.4847862720489502, "learning_rate": 8.554183669227307e-06, "loss": 0.0267, "step": 82050 }, { "grad_norm": 0.8092252612113953, "learning_rate": 8.544936878248937e-06, "loss": 0.0324, "step": 82060 }, { "grad_norm": 0.6224797964096069, "learning_rate": 8.535694620733942e-06, "loss": 0.0318, "step": 82070 }, { "grad_norm": 0.39319536089897156, "learning_rate": 8.526456897693053e-06, "loss": 0.0309, "step": 82080 }, { "grad_norm": 0.42592519521713257, "learning_rate": 8.517223710136456e-06, "loss": 0.033, "step": 82090 }, { "grad_norm": 0.9940540194511414, "learning_rate": 8.507995059073931e-06, "loss": 0.0324, "step": 82100 }, { "grad_norm": 0.3690299987792969, "learning_rate": 8.498770945514673e-06, "loss": 0.0374, "step": 82110 }, { "grad_norm": 0.286346435546875, "learning_rate": 8.489551370467436e-06, "loss": 0.0254, "step": 82120 }, { "grad_norm": 0.307566374540329, "learning_rate": 8.480336334940442e-06, "loss": 0.0346, "step": 82130 }, { "grad_norm": 0.27422183752059937, "learning_rate": 8.471125839941458e-06, "loss": 0.026, "step": 82140 }, { "grad_norm": 0.5197246074676514, "learning_rate": 8.461919886477682e-06, "loss": 0.0309, "step": 82150 }, { "grad_norm": 0.7240541577339172, "learning_rate": 8.452718475555927e-06, "loss": 0.0322, "step": 82160 }, { "grad_norm": 0.26441875100135803, "learning_rate": 8.443521608182398e-06, "loss": 0.0252, "step": 82170 }, { "grad_norm": 0.4132573902606964, "learning_rate": 8.43432928536288e-06, "loss": 0.0325, "step": 82180 }, { "grad_norm": 0.32976287603378296, "learning_rate": 8.425141508102607e-06, "loss": 0.0335, "step": 82190 }, { "grad_norm": 0.25544673204421997, "learning_rate": 8.415958277406365e-06, "loss": 0.0252, "step": 82200 }, { "grad_norm": 0.242432102560997, "learning_rate": 8.406779594278392e-06, "loss": 0.0318, "step": 82210 }, { "grad_norm": 0.4827437996864319, "learning_rate": 8.397605459722468e-06, "loss": 0.0287, "step": 82220 }, { "grad_norm": 0.42155054211616516, "learning_rate": 8.388435874741863e-06, "loss": 0.0365, "step": 82230 }, { "grad_norm": 0.45374441146850586, "learning_rate": 8.37927084033936e-06, "loss": 0.0359, "step": 82240 }, { "grad_norm": 0.5214115381240845, "learning_rate": 8.370110357517208e-06, "loss": 0.0323, "step": 82250 }, { "grad_norm": 0.37065133452415466, "learning_rate": 8.360954427277201e-06, "loss": 0.0348, "step": 82260 }, { "grad_norm": 0.18059559166431427, "learning_rate": 8.351803050620627e-06, "loss": 0.0289, "step": 82270 }, { "grad_norm": 0.4385329484939575, "learning_rate": 8.34265622854824e-06, "loss": 0.0298, "step": 82280 }, { "grad_norm": 0.283586710691452, "learning_rate": 8.333513962060335e-06, "loss": 0.0429, "step": 82290 }, { "grad_norm": 0.31816017627716064, "learning_rate": 8.324376252156701e-06, "loss": 0.0182, "step": 82300 }, { "grad_norm": 0.40071967244148254, "learning_rate": 8.315243099836633e-06, "loss": 0.0291, "step": 82310 }, { "grad_norm": 0.41584983468055725, "learning_rate": 8.3061145060989e-06, "loss": 0.038, "step": 82320 }, { "grad_norm": 0.35032108426094055, "learning_rate": 8.296990471941806e-06, "loss": 0.0387, "step": 82330 }, { "grad_norm": 0.4184361696243286, "learning_rate": 8.287870998363112e-06, "loss": 0.0347, "step": 82340 }, { "grad_norm": 0.26256197690963745, "learning_rate": 8.278756086360156e-06, "loss": 0.0297, "step": 82350 }, { "grad_norm": 0.4632318615913391, "learning_rate": 8.269645736929693e-06, "loss": 0.0298, "step": 82360 }, { "grad_norm": 0.4512469172477722, "learning_rate": 8.260539951068048e-06, "loss": 0.0302, "step": 82370 }, { "grad_norm": 0.4150925874710083, "learning_rate": 8.251438729770983e-06, "loss": 0.0398, "step": 82380 }, { "grad_norm": 0.36406272649765015, "learning_rate": 8.242342074033826e-06, "loss": 0.0289, "step": 82390 }, { "grad_norm": 0.46225813031196594, "learning_rate": 8.233249984851344e-06, "loss": 0.0323, "step": 82400 }, { "grad_norm": 0.6833818554878235, "learning_rate": 8.224162463217849e-06, "loss": 0.0404, "step": 82410 }, { "grad_norm": 0.35334286093711853, "learning_rate": 8.215079510127139e-06, "loss": 0.0273, "step": 82420 }, { "grad_norm": 0.5790374279022217, "learning_rate": 8.206001126572521e-06, "loss": 0.038, "step": 82430 }, { "grad_norm": 0.5246878266334534, "learning_rate": 8.196927313546776e-06, "loss": 0.036, "step": 82440 }, { "grad_norm": 0.2220926433801651, "learning_rate": 8.18785807204222e-06, "loss": 0.0341, "step": 82450 }, { "grad_norm": 0.43998652696609497, "learning_rate": 8.178793403050627e-06, "loss": 0.0288, "step": 82460 }, { "grad_norm": 0.3324064314365387, "learning_rate": 8.16973330756331e-06, "loss": 0.0304, "step": 82470 }, { "grad_norm": 0.4823618233203888, "learning_rate": 8.16067778657107e-06, "loss": 0.0353, "step": 82480 }, { "grad_norm": 0.805895209312439, "learning_rate": 8.151626841064214e-06, "loss": 0.0264, "step": 82490 }, { "grad_norm": 0.770523726940155, "learning_rate": 8.142580472032518e-06, "loss": 0.0287, "step": 82500 }, { "grad_norm": 0.524257242679596, "learning_rate": 8.133538680465285e-06, "loss": 0.0253, "step": 82510 }, { "grad_norm": 0.467680424451828, "learning_rate": 8.124501467351336e-06, "loss": 0.0339, "step": 82520 }, { "grad_norm": 0.8448049426078796, "learning_rate": 8.11546883367893e-06, "loss": 0.0258, "step": 82530 }, { "grad_norm": 0.25253841280937195, "learning_rate": 8.106440780435882e-06, "loss": 0.0289, "step": 82540 }, { "grad_norm": 0.22377566993236542, "learning_rate": 8.097417308609485e-06, "loss": 0.0279, "step": 82550 }, { "grad_norm": 0.26996180415153503, "learning_rate": 8.088398419186544e-06, "loss": 0.0282, "step": 82560 }, { "grad_norm": 0.6883817911148071, "learning_rate": 8.079384113153321e-06, "loss": 0.0329, "step": 82570 }, { "grad_norm": 0.4558555483818054, "learning_rate": 8.070374391495644e-06, "loss": 0.0277, "step": 82580 }, { "grad_norm": 0.3536602854728699, "learning_rate": 8.061369255198764e-06, "loss": 0.0235, "step": 82590 }, { "grad_norm": 0.712976336479187, "learning_rate": 8.052368705247493e-06, "loss": 0.0379, "step": 82600 }, { "grad_norm": 0.7617502808570862, "learning_rate": 8.043372742626114e-06, "loss": 0.0451, "step": 82610 }, { "grad_norm": 1.4748769998550415, "learning_rate": 8.034381368318422e-06, "loss": 0.0331, "step": 82620 }, { "grad_norm": 0.5052531957626343, "learning_rate": 8.025394583307677e-06, "loss": 0.0298, "step": 82630 }, { "grad_norm": 1.1013561487197876, "learning_rate": 8.016412388576683e-06, "loss": 0.0356, "step": 82640 }, { "grad_norm": 0.6882063746452332, "learning_rate": 8.007434785107699e-06, "loss": 0.0262, "step": 82650 }, { "grad_norm": 0.5162066221237183, "learning_rate": 7.998461773882509e-06, "loss": 0.0328, "step": 82660 }, { "grad_norm": 0.43166443705558777, "learning_rate": 7.989493355882393e-06, "loss": 0.0302, "step": 82670 }, { "grad_norm": 0.4297500252723694, "learning_rate": 7.980529532088137e-06, "loss": 0.0326, "step": 82680 }, { "grad_norm": 0.5792945623397827, "learning_rate": 7.97157030347998e-06, "loss": 0.0305, "step": 82690 }, { "grad_norm": 0.368621289730072, "learning_rate": 7.96261567103772e-06, "loss": 0.0225, "step": 82700 }, { "grad_norm": 0.5615940690040588, "learning_rate": 7.953665635740593e-06, "loss": 0.0247, "step": 82710 }, { "grad_norm": 0.2934889495372772, "learning_rate": 7.94472019856738e-06, "loss": 0.0239, "step": 82720 }, { "grad_norm": 0.4376870393753052, "learning_rate": 7.935779360496337e-06, "loss": 0.0286, "step": 82730 }, { "grad_norm": 0.5094372630119324, "learning_rate": 7.92684312250523e-06, "loss": 0.0336, "step": 82740 }, { "grad_norm": 0.3233363926410675, "learning_rate": 7.917911485571289e-06, "loss": 0.0279, "step": 82750 }, { "grad_norm": 0.7422325611114502, "learning_rate": 7.90898445067128e-06, "loss": 0.0294, "step": 82760 }, { "grad_norm": 0.5290586948394775, "learning_rate": 7.900062018781462e-06, "loss": 0.0363, "step": 82770 }, { "grad_norm": 0.32430464029312134, "learning_rate": 7.891144190877553e-06, "loss": 0.0262, "step": 82780 }, { "grad_norm": 0.18135586380958557, "learning_rate": 7.882230967934806e-06, "loss": 0.0262, "step": 82790 }, { "grad_norm": 0.373781681060791, "learning_rate": 7.873322350927959e-06, "loss": 0.0317, "step": 82800 }, { "grad_norm": 0.7666438221931458, "learning_rate": 7.864418340831253e-06, "loss": 0.0287, "step": 82810 }, { "grad_norm": 0.48850592970848083, "learning_rate": 7.855518938618395e-06, "loss": 0.0373, "step": 82820 }, { "grad_norm": 0.512430727481842, "learning_rate": 7.846624145262633e-06, "loss": 0.0293, "step": 82830 }, { "grad_norm": 0.23771867156028748, "learning_rate": 7.837733961736672e-06, "loss": 0.0313, "step": 82840 }, { "grad_norm": 0.41541486978530884, "learning_rate": 7.828848389012727e-06, "loss": 0.0257, "step": 82850 }, { "grad_norm": 0.3881905674934387, "learning_rate": 7.819967428062524e-06, "loss": 0.0294, "step": 82860 }, { "grad_norm": 0.4285934269428253, "learning_rate": 7.811091079857274e-06, "loss": 0.0351, "step": 82870 }, { "grad_norm": 0.4385014474391937, "learning_rate": 7.802219345367667e-06, "loss": 0.0279, "step": 82880 }, { "grad_norm": 0.7068995237350464, "learning_rate": 7.793352225563916e-06, "loss": 0.0258, "step": 82890 }, { "grad_norm": 0.4847243130207062, "learning_rate": 7.7844897214157e-06, "loss": 0.0308, "step": 82900 }, { "grad_norm": 0.3009692430496216, "learning_rate": 7.775631833892228e-06, "loss": 0.0383, "step": 82910 }, { "grad_norm": 0.27133625745773315, "learning_rate": 7.76677856396215e-06, "loss": 0.0218, "step": 82920 }, { "grad_norm": 0.740820586681366, "learning_rate": 7.757929912593697e-06, "loss": 0.0273, "step": 82930 }, { "grad_norm": 0.6259267330169678, "learning_rate": 7.749085880754509e-06, "loss": 0.0333, "step": 82940 }, { "grad_norm": 0.28570234775543213, "learning_rate": 7.740246469411777e-06, "loss": 0.0332, "step": 82950 }, { "grad_norm": 0.3466055393218994, "learning_rate": 7.731411679532142e-06, "loss": 0.0288, "step": 82960 }, { "grad_norm": 0.3298392593860626, "learning_rate": 7.722581512081779e-06, "loss": 0.0376, "step": 82970 }, { "grad_norm": 0.5511367917060852, "learning_rate": 7.713755968026331e-06, "loss": 0.0267, "step": 82980 }, { "grad_norm": 0.7245579957962036, "learning_rate": 7.704935048330969e-06, "loss": 0.0308, "step": 82990 }, { "grad_norm": 0.4990362524986267, "learning_rate": 7.696118753960313e-06, "loss": 0.0287, "step": 83000 }, { "grad_norm": 0.45019999146461487, "learning_rate": 7.687307085878503e-06, "loss": 0.0259, "step": 83010 }, { "grad_norm": 0.27085795998573303, "learning_rate": 7.678500045049186e-06, "loss": 0.0283, "step": 83020 }, { "grad_norm": 1.8979554176330566, "learning_rate": 7.669697632435463e-06, "loss": 0.0311, "step": 83030 }, { "grad_norm": 0.42993226647377014, "learning_rate": 7.660899848999963e-06, "loss": 0.0244, "step": 83040 }, { "grad_norm": 0.31424036622047424, "learning_rate": 7.652106695704803e-06, "loss": 0.0305, "step": 83050 }, { "grad_norm": 0.5412424206733704, "learning_rate": 7.643318173511598e-06, "loss": 0.0242, "step": 83060 }, { "grad_norm": 0.49123337864875793, "learning_rate": 7.634534283381423e-06, "loss": 0.0325, "step": 83070 }, { "grad_norm": 0.6807538866996765, "learning_rate": 7.625755026274889e-06, "loss": 0.0293, "step": 83080 }, { "grad_norm": 0.4848451614379883, "learning_rate": 7.616980403152069e-06, "loss": 0.0341, "step": 83090 }, { "grad_norm": 0.5288952589035034, "learning_rate": 7.608210414972561e-06, "loss": 0.032, "step": 83100 }, { "grad_norm": 0.34463685750961304, "learning_rate": 7.599445062695404e-06, "loss": 0.0341, "step": 83110 }, { "grad_norm": 0.471279114484787, "learning_rate": 7.590684347279214e-06, "loss": 0.031, "step": 83120 }, { "grad_norm": 0.355556845664978, "learning_rate": 7.581928269682004e-06, "loss": 0.026, "step": 83130 }, { "grad_norm": 0.23887398838996887, "learning_rate": 7.573176830861361e-06, "loss": 0.0255, "step": 83140 }, { "grad_norm": 0.31187430024147034, "learning_rate": 7.5644300317743e-06, "loss": 0.0321, "step": 83150 }, { "grad_norm": 0.38140881061553955, "learning_rate": 7.555687873377376e-06, "loss": 0.0352, "step": 83160 }, { "grad_norm": 0.31337279081344604, "learning_rate": 7.546950356626592e-06, "loss": 0.0364, "step": 83170 }, { "grad_norm": 0.4002666175365448, "learning_rate": 7.538217482477516e-06, "loss": 0.0339, "step": 83180 }, { "grad_norm": 0.4263489842414856, "learning_rate": 7.529489251885119e-06, "loss": 0.0318, "step": 83190 }, { "grad_norm": 3.406857967376709, "learning_rate": 7.520765665803941e-06, "loss": 0.0317, "step": 83200 }, { "grad_norm": 0.23480777442455292, "learning_rate": 7.512046725187949e-06, "loss": 0.0256, "step": 83210 }, { "grad_norm": 0.692685067653656, "learning_rate": 7.503332430990645e-06, "loss": 0.0272, "step": 83220 }, { "grad_norm": 0.3842483460903168, "learning_rate": 7.494622784165017e-06, "loss": 0.0332, "step": 83230 }, { "grad_norm": 0.2809140682220459, "learning_rate": 7.485917785663543e-06, "loss": 0.0353, "step": 83240 }, { "grad_norm": 0.7925679087638855, "learning_rate": 7.477217436438172e-06, "loss": 0.0268, "step": 83250 }, { "grad_norm": 0.43037694692611694, "learning_rate": 7.46852173744037e-06, "loss": 0.0209, "step": 83260 }, { "grad_norm": 1.6426563262939453, "learning_rate": 7.4598306896211004e-06, "loss": 0.0403, "step": 83270 }, { "grad_norm": 0.33383458852767944, "learning_rate": 7.451144293930773e-06, "loss": 0.0297, "step": 83280 }, { "grad_norm": 0.19376865029335022, "learning_rate": 7.4424625513193455e-06, "loss": 0.033, "step": 83290 }, { "grad_norm": 0.46176114678382874, "learning_rate": 7.433785462736209e-06, "loss": 0.0314, "step": 83300 }, { "grad_norm": 0.2810991108417511, "learning_rate": 7.425113029130315e-06, "loss": 0.0304, "step": 83310 }, { "grad_norm": 0.5213607549667358, "learning_rate": 7.416445251450044e-06, "loss": 0.0296, "step": 83320 }, { "grad_norm": 0.29412075877189636, "learning_rate": 7.4077821306433035e-06, "loss": 0.0362, "step": 83330 }, { "grad_norm": 0.22371631860733032, "learning_rate": 7.399123667657465e-06, "loss": 0.0231, "step": 83340 }, { "grad_norm": 0.5116925835609436, "learning_rate": 7.390469863439426e-06, "loss": 0.0228, "step": 83350 }, { "grad_norm": 0.9331400394439697, "learning_rate": 7.381820718935512e-06, "loss": 0.0299, "step": 83360 }, { "grad_norm": 0.3467881977558136, "learning_rate": 7.373176235091633e-06, "loss": 0.0355, "step": 83370 }, { "grad_norm": 0.3159867823123932, "learning_rate": 7.364536412853101e-06, "loss": 0.0263, "step": 83380 }, { "grad_norm": 1.372725009918213, "learning_rate": 7.355901253164776e-06, "loss": 0.0307, "step": 83390 }, { "grad_norm": 0.3242562413215637, "learning_rate": 7.34727075697097e-06, "loss": 0.0229, "step": 83400 }, { "grad_norm": 0.3289305567741394, "learning_rate": 7.3386449252155095e-06, "loss": 0.029, "step": 83410 }, { "grad_norm": 0.9977010488510132, "learning_rate": 7.330023758841692e-06, "loss": 0.034, "step": 83420 }, { "grad_norm": 0.6487022638320923, "learning_rate": 7.321407258792323e-06, "loss": 0.0264, "step": 83430 }, { "grad_norm": 0.453213632106781, "learning_rate": 7.312795426009694e-06, "loss": 0.0277, "step": 83440 }, { "grad_norm": 0.5081258416175842, "learning_rate": 7.304188261435585e-06, "loss": 0.0304, "step": 83450 }, { "grad_norm": 0.8226735591888428, "learning_rate": 7.295585766011242e-06, "loss": 0.0287, "step": 83460 }, { "grad_norm": 0.6503532528877258, "learning_rate": 7.286987940677436e-06, "loss": 0.0346, "step": 83470 }, { "grad_norm": 0.3128100335597992, "learning_rate": 7.278394786374426e-06, "loss": 0.0227, "step": 83480 }, { "grad_norm": 0.23435047268867493, "learning_rate": 7.269806304041915e-06, "loss": 0.0274, "step": 83490 }, { "grad_norm": 0.7456159591674805, "learning_rate": 7.261222494619147e-06, "loss": 0.0274, "step": 83500 }, { "grad_norm": 0.5207486748695374, "learning_rate": 7.252643359044825e-06, "loss": 0.0349, "step": 83510 }, { "grad_norm": 0.331582248210907, "learning_rate": 7.244068898257172e-06, "loss": 0.0238, "step": 83520 }, { "grad_norm": 0.21583719551563263, "learning_rate": 7.2354991131938495e-06, "loss": 0.0254, "step": 83530 }, { "grad_norm": 0.3777824938297272, "learning_rate": 7.2269340047920574e-06, "loss": 0.0194, "step": 83540 }, { "grad_norm": 0.5923568606376648, "learning_rate": 7.218373573988435e-06, "loss": 0.0333, "step": 83550 }, { "grad_norm": 0.7570068836212158, "learning_rate": 7.209817821719178e-06, "loss": 0.0393, "step": 83560 }, { "grad_norm": 0.2547169327735901, "learning_rate": 7.201266748919894e-06, "loss": 0.0366, "step": 83570 }, { "grad_norm": 0.3047088384628296, "learning_rate": 7.19272035652575e-06, "loss": 0.0281, "step": 83580 }, { "grad_norm": 0.3538009822368622, "learning_rate": 7.184178645471334e-06, "loss": 0.0286, "step": 83590 }, { "grad_norm": 0.5314264297485352, "learning_rate": 7.175641616690775e-06, "loss": 0.0361, "step": 83600 }, { "grad_norm": 0.5069778561592102, "learning_rate": 7.167109271117661e-06, "loss": 0.0352, "step": 83610 }, { "grad_norm": 0.9831804037094116, "learning_rate": 7.158581609685072e-06, "loss": 0.0389, "step": 83620 }, { "grad_norm": 0.23791739344596863, "learning_rate": 7.150058633325591e-06, "loss": 0.0267, "step": 83630 }, { "grad_norm": 0.6082636117935181, "learning_rate": 7.141540342971287e-06, "loss": 0.0333, "step": 83640 }, { "grad_norm": 0.38257667422294617, "learning_rate": 7.133026739553677e-06, "loss": 0.0246, "step": 83650 }, { "grad_norm": 0.2844611406326294, "learning_rate": 7.124517824003829e-06, "loss": 0.0416, "step": 83660 }, { "grad_norm": 0.5765508413314819, "learning_rate": 7.116013597252236e-06, "loss": 0.0265, "step": 83670 }, { "grad_norm": 0.2673681974411011, "learning_rate": 7.107514060228921e-06, "loss": 0.0289, "step": 83680 }, { "grad_norm": 0.3390315771102905, "learning_rate": 7.099019213863378e-06, "loss": 0.0318, "step": 83690 }, { "grad_norm": 0.7796434164047241, "learning_rate": 7.090529059084611e-06, "loss": 0.0211, "step": 83700 }, { "grad_norm": 0.24710945785045624, "learning_rate": 7.082043596821058e-06, "loss": 0.0232, "step": 83710 }, { "grad_norm": 0.870329737663269, "learning_rate": 7.07356282800069e-06, "loss": 0.0325, "step": 83720 }, { "grad_norm": 0.4709210693836212, "learning_rate": 7.0650867535509645e-06, "loss": 0.0256, "step": 83730 }, { "grad_norm": 0.44161397218704224, "learning_rate": 7.0566153743987895e-06, "loss": 0.0242, "step": 83740 }, { "grad_norm": 0.25152891874313354, "learning_rate": 7.04814869147059e-06, "loss": 0.0334, "step": 83750 }, { "grad_norm": 0.44268810749053955, "learning_rate": 7.0396867056922725e-06, "loss": 0.0344, "step": 83760 }, { "grad_norm": 0.35779377818107605, "learning_rate": 7.031229417989244e-06, "loss": 0.0317, "step": 83770 }, { "grad_norm": 0.5575042366981506, "learning_rate": 7.022776829286348e-06, "loss": 0.0389, "step": 83780 }, { "grad_norm": 0.731086254119873, "learning_rate": 7.014328940507975e-06, "loss": 0.0307, "step": 83790 }, { "grad_norm": 0.4804249703884125, "learning_rate": 7.005885752577945e-06, "loss": 0.0325, "step": 83800 }, { "grad_norm": 0.44538047909736633, "learning_rate": 6.997447266419615e-06, "loss": 0.0359, "step": 83810 }, { "grad_norm": 0.30347931385040283, "learning_rate": 6.989013482955797e-06, "loss": 0.0359, "step": 83820 }, { "grad_norm": 1.2005033493041992, "learning_rate": 6.980584403108808e-06, "loss": 0.0307, "step": 83830 }, { "grad_norm": 0.508815586566925, "learning_rate": 6.972160027800417e-06, "loss": 0.029, "step": 83840 }, { "grad_norm": 0.44251906871795654, "learning_rate": 6.963740357951931e-06, "loss": 0.0358, "step": 83850 }, { "grad_norm": 0.3367398977279663, "learning_rate": 6.95532539448408e-06, "loss": 0.032, "step": 83860 }, { "grad_norm": 0.45096302032470703, "learning_rate": 6.946915138317128e-06, "loss": 0.0301, "step": 83870 }, { "grad_norm": 0.30600616335868835, "learning_rate": 6.9385095903708065e-06, "loss": 0.0358, "step": 83880 }, { "grad_norm": 0.5201416611671448, "learning_rate": 6.930108751564346e-06, "loss": 0.039, "step": 83890 }, { "grad_norm": 0.36280080676078796, "learning_rate": 6.921712622816428e-06, "loss": 0.0305, "step": 83900 }, { "grad_norm": 0.3878750205039978, "learning_rate": 6.913321205045259e-06, "loss": 0.0329, "step": 83910 }, { "grad_norm": 0.5672051906585693, "learning_rate": 6.904934499168491e-06, "loss": 0.024, "step": 83920 }, { "grad_norm": 0.24400906264781952, "learning_rate": 6.896552506103299e-06, "loss": 0.0325, "step": 83930 }, { "grad_norm": 0.5637053847312927, "learning_rate": 6.888175226766313e-06, "loss": 0.0389, "step": 83940 }, { "grad_norm": 0.4748346209526062, "learning_rate": 6.879802662073686e-06, "loss": 0.032, "step": 83950 }, { "grad_norm": 0.3590448498725891, "learning_rate": 6.871434812940991e-06, "loss": 0.0367, "step": 83960 }, { "grad_norm": 0.3237929940223694, "learning_rate": 6.863071680283345e-06, "loss": 0.0225, "step": 83970 }, { "grad_norm": 0.39935606718063354, "learning_rate": 6.854713265015339e-06, "loss": 0.0427, "step": 83980 }, { "grad_norm": 0.9550517201423645, "learning_rate": 6.8463595680510025e-06, "loss": 0.037, "step": 83990 }, { "grad_norm": 0.545988917350769, "learning_rate": 6.838010590303911e-06, "loss": 0.0333, "step": 84000 }, { "grad_norm": 0.9107654094696045, "learning_rate": 6.829666332687085e-06, "loss": 0.0245, "step": 84010 }, { "grad_norm": 0.519900918006897, "learning_rate": 6.821326796113059e-06, "loss": 0.0252, "step": 84020 }, { "grad_norm": 0.3309771716594696, "learning_rate": 6.812991981493799e-06, "loss": 0.0275, "step": 84030 }, { "grad_norm": 0.5139305591583252, "learning_rate": 6.804661889740821e-06, "loss": 0.0346, "step": 84040 }, { "grad_norm": 0.8303739428520203, "learning_rate": 6.7963365217650625e-06, "loss": 0.0361, "step": 84050 }, { "grad_norm": 0.31790193915367126, "learning_rate": 6.788015878476983e-06, "loss": 0.0324, "step": 84060 }, { "grad_norm": 0.4316120445728302, "learning_rate": 6.779699960786523e-06, "loss": 0.0358, "step": 84070 }, { "grad_norm": 1.119322657585144, "learning_rate": 6.771388769603104e-06, "loss": 0.0265, "step": 84080 }, { "grad_norm": 0.5442484617233276, "learning_rate": 6.763082305835605e-06, "loss": 0.0449, "step": 84090 }, { "grad_norm": 0.3579372763633728, "learning_rate": 6.754780570392433e-06, "loss": 0.0272, "step": 84100 }, { "grad_norm": 0.5466105937957764, "learning_rate": 6.7464835641814215e-06, "loss": 0.0351, "step": 84110 }, { "grad_norm": 0.1514660269021988, "learning_rate": 6.7381912881099515e-06, "loss": 0.0252, "step": 84120 }, { "grad_norm": 0.7223902344703674, "learning_rate": 6.72990374308482e-06, "loss": 0.0361, "step": 84130 }, { "grad_norm": 0.4198493957519531, "learning_rate": 6.721620930012379e-06, "loss": 0.0323, "step": 84140 }, { "grad_norm": 0.4672493636608124, "learning_rate": 6.713342849798393e-06, "loss": 0.0252, "step": 84150 }, { "grad_norm": 0.3639281988143921, "learning_rate": 6.705069503348166e-06, "loss": 0.0274, "step": 84160 }, { "grad_norm": 0.6384933590888977, "learning_rate": 6.696800891566435e-06, "loss": 0.0354, "step": 84170 }, { "grad_norm": 0.2482864111661911, "learning_rate": 6.688537015357449e-06, "loss": 0.0265, "step": 84180 }, { "grad_norm": 0.4064949154853821, "learning_rate": 6.680277875624941e-06, "loss": 0.0266, "step": 84190 }, { "grad_norm": 0.8193874955177307, "learning_rate": 6.672023473272121e-06, "loss": 0.0346, "step": 84200 }, { "grad_norm": 0.416337251663208, "learning_rate": 6.663773809201667e-06, "loss": 0.0385, "step": 84210 }, { "grad_norm": 0.37534818053245544, "learning_rate": 6.655528884315748e-06, "loss": 0.0296, "step": 84220 }, { "grad_norm": 0.47828319668769836, "learning_rate": 6.647288699516036e-06, "loss": 0.0381, "step": 84230 }, { "grad_norm": 0.6314563155174255, "learning_rate": 6.639053255703642e-06, "loss": 0.0349, "step": 84240 }, { "grad_norm": 0.49758124351501465, "learning_rate": 6.6308225537791925e-06, "loss": 0.0367, "step": 84250 }, { "grad_norm": 0.40582558512687683, "learning_rate": 6.622596594642777e-06, "loss": 0.0271, "step": 84260 }, { "grad_norm": 1.1016508340835571, "learning_rate": 6.614375379193999e-06, "loss": 0.0252, "step": 84270 }, { "grad_norm": 1.0577551126480103, "learning_rate": 6.606158908331889e-06, "loss": 0.0269, "step": 84280 }, { "grad_norm": 0.3482016921043396, "learning_rate": 6.597947182955005e-06, "loss": 0.0354, "step": 84290 }, { "grad_norm": 0.6038007736206055, "learning_rate": 6.589740203961359e-06, "loss": 0.041, "step": 84300 }, { "grad_norm": 1.0876413583755493, "learning_rate": 6.581537972248464e-06, "loss": 0.0371, "step": 84310 }, { "grad_norm": 0.7622758150100708, "learning_rate": 6.573340488713275e-06, "loss": 0.0313, "step": 84320 }, { "grad_norm": 0.7229092717170715, "learning_rate": 6.565147754252304e-06, "loss": 0.036, "step": 84330 }, { "grad_norm": 0.3313325047492981, "learning_rate": 6.556959769761462e-06, "loss": 0.037, "step": 84340 }, { "grad_norm": 1.5095402002334595, "learning_rate": 6.548776536136192e-06, "loss": 0.0265, "step": 84350 }, { "grad_norm": 0.5983250737190247, "learning_rate": 6.54059805427138e-06, "loss": 0.0293, "step": 84360 }, { "grad_norm": 0.3651491403579712, "learning_rate": 6.532424325061437e-06, "loss": 0.0454, "step": 84370 }, { "grad_norm": 0.4955118000507355, "learning_rate": 6.5242553494001935e-06, "loss": 0.0267, "step": 84380 }, { "grad_norm": 0.4644823372364044, "learning_rate": 6.516091128181046e-06, "loss": 0.0254, "step": 84390 }, { "grad_norm": 0.4195861518383026, "learning_rate": 6.507931662296785e-06, "loss": 0.0307, "step": 84400 }, { "grad_norm": 1.2233242988586426, "learning_rate": 6.4997769526397354e-06, "loss": 0.0292, "step": 84410 }, { "grad_norm": 0.3315410614013672, "learning_rate": 6.491627000101663e-06, "loss": 0.03, "step": 84420 }, { "grad_norm": 0.5106527209281921, "learning_rate": 6.4834818055738545e-06, "loss": 0.0298, "step": 84430 }, { "grad_norm": 0.35175609588623047, "learning_rate": 6.4753413699470465e-06, "loss": 0.0226, "step": 84440 }, { "grad_norm": 0.5706942677497864, "learning_rate": 6.467205694111478e-06, "loss": 0.0278, "step": 84450 }, { "grad_norm": 0.26608335971832275, "learning_rate": 6.459074778956836e-06, "loss": 0.0319, "step": 84460 }, { "grad_norm": 0.4021580219268799, "learning_rate": 6.450948625372311e-06, "loss": 0.0263, "step": 84470 }, { "grad_norm": 0.30553096532821655, "learning_rate": 6.442827234246585e-06, "loss": 0.0208, "step": 84480 }, { "grad_norm": 0.3695995509624481, "learning_rate": 6.43471060646777e-06, "loss": 0.0323, "step": 84490 }, { "grad_norm": 0.22397828102111816, "learning_rate": 6.426598742923512e-06, "loss": 0.0273, "step": 84500 }, { "grad_norm": 0.7144094705581665, "learning_rate": 6.418491644500885e-06, "loss": 0.0342, "step": 84510 }, { "grad_norm": 0.3424965739250183, "learning_rate": 6.410389312086512e-06, "loss": 0.0322, "step": 84520 }, { "grad_norm": 0.43071249127388, "learning_rate": 6.4022917465664135e-06, "loss": 0.0284, "step": 84530 }, { "grad_norm": 0.5788045525550842, "learning_rate": 6.3941989488261575e-06, "loss": 0.0332, "step": 84540 }, { "grad_norm": 0.2604577839374542, "learning_rate": 6.386110919750726e-06, "loss": 0.0299, "step": 84550 }, { "grad_norm": 0.2917165458202362, "learning_rate": 6.378027660224651e-06, "loss": 0.026, "step": 84560 }, { "grad_norm": 0.29377201199531555, "learning_rate": 6.369949171131856e-06, "loss": 0.0234, "step": 84570 }, { "grad_norm": 0.3668208718299866, "learning_rate": 6.361875453355848e-06, "loss": 0.0238, "step": 84580 }, { "grad_norm": 0.779508113861084, "learning_rate": 6.353806507779525e-06, "loss": 0.0288, "step": 84590 }, { "grad_norm": 0.31513288617134094, "learning_rate": 6.3457423352853044e-06, "loss": 0.0316, "step": 84600 }, { "grad_norm": 0.6845664978027344, "learning_rate": 6.337682936755062e-06, "loss": 0.0381, "step": 84610 }, { "grad_norm": 0.302367627620697, "learning_rate": 6.329628313070179e-06, "loss": 0.0313, "step": 84620 }, { "grad_norm": 0.2814056873321533, "learning_rate": 6.321578465111477e-06, "loss": 0.0287, "step": 84630 }, { "grad_norm": 0.30864065885543823, "learning_rate": 6.31353339375928e-06, "loss": 0.0296, "step": 84640 }, { "grad_norm": 0.36573368310928345, "learning_rate": 6.305493099893395e-06, "loss": 0.0197, "step": 84650 }, { "grad_norm": 0.4522995948791504, "learning_rate": 6.297457584393096e-06, "loss": 0.0291, "step": 84660 }, { "grad_norm": 0.5151695609092712, "learning_rate": 6.289426848137126e-06, "loss": 0.0299, "step": 84670 }, { "grad_norm": 0.4358444809913635, "learning_rate": 6.281400892003713e-06, "loss": 0.0238, "step": 84680 }, { "grad_norm": 0.3314712345600128, "learning_rate": 6.273379716870581e-06, "loss": 0.028, "step": 84690 }, { "grad_norm": 0.5556342601776123, "learning_rate": 6.265363323614887e-06, "loss": 0.0298, "step": 84700 }, { "grad_norm": 0.7441123127937317, "learning_rate": 6.257351713113308e-06, "loss": 0.0294, "step": 84710 }, { "grad_norm": 0.6004700064659119, "learning_rate": 6.249344886241981e-06, "loss": 0.0363, "step": 84720 }, { "grad_norm": 0.3875994384288788, "learning_rate": 6.2413428438765295e-06, "loss": 0.0223, "step": 84730 }, { "grad_norm": 0.2879999876022339, "learning_rate": 6.233345586892026e-06, "loss": 0.0269, "step": 84740 }, { "grad_norm": 0.25205928087234497, "learning_rate": 6.225353116163052e-06, "loss": 0.0219, "step": 84750 }, { "grad_norm": 0.3716414272785187, "learning_rate": 6.217365432563627e-06, "loss": 0.0267, "step": 84760 }, { "grad_norm": 0.3011830151081085, "learning_rate": 6.209382536967318e-06, "loss": 0.027, "step": 84770 }, { "grad_norm": 0.6334471106529236, "learning_rate": 6.201404430247082e-06, "loss": 0.0337, "step": 84780 }, { "grad_norm": 0.32573166489601135, "learning_rate": 6.1934311132754165e-06, "loss": 0.0215, "step": 84790 }, { "grad_norm": 1.105362892150879, "learning_rate": 6.185462586924257e-06, "loss": 0.0359, "step": 84800 }, { "grad_norm": 0.6136109232902527, "learning_rate": 6.177498852065039e-06, "loss": 0.0321, "step": 84810 }, { "grad_norm": 0.3363182842731476, "learning_rate": 6.169539909568655e-06, "loss": 0.0291, "step": 84820 }, { "grad_norm": 0.40183141827583313, "learning_rate": 6.161585760305488e-06, "loss": 0.0378, "step": 84830 }, { "grad_norm": 0.7323612570762634, "learning_rate": 6.153636405145391e-06, "loss": 0.0338, "step": 84840 }, { "grad_norm": 1.0606738328933716, "learning_rate": 6.145691844957702e-06, "loss": 0.0293, "step": 84850 }, { "grad_norm": 0.26894131302833557, "learning_rate": 6.1377520806112145e-06, "loss": 0.027, "step": 84860 }, { "grad_norm": 0.3299722969532013, "learning_rate": 6.129817112974218e-06, "loss": 0.0252, "step": 84870 }, { "grad_norm": 0.48441487550735474, "learning_rate": 6.121886942914457e-06, "loss": 0.0242, "step": 84880 }, { "grad_norm": 0.565912127494812, "learning_rate": 6.11396157129917e-06, "loss": 0.0308, "step": 84890 }, { "grad_norm": 0.46100887656211853, "learning_rate": 6.106040998995066e-06, "loss": 0.0313, "step": 84900 }, { "grad_norm": 0.2377411425113678, "learning_rate": 6.098125226868329e-06, "loss": 0.0298, "step": 84910 }, { "grad_norm": 0.32708021998405457, "learning_rate": 6.090214255784604e-06, "loss": 0.031, "step": 84920 }, { "grad_norm": 0.3498588502407074, "learning_rate": 6.0823080866090346e-06, "loss": 0.0324, "step": 84930 }, { "grad_norm": 0.4296483099460602, "learning_rate": 6.074406720206233e-06, "loss": 0.0257, "step": 84940 }, { "grad_norm": 0.4166399836540222, "learning_rate": 6.066510157440259e-06, "loss": 0.0257, "step": 84950 }, { "grad_norm": 0.25065916776657104, "learning_rate": 6.058618399174676e-06, "loss": 0.0271, "step": 84960 }, { "grad_norm": 0.39001020789146423, "learning_rate": 6.050731446272528e-06, "loss": 0.0346, "step": 84970 }, { "grad_norm": 1.2099350690841675, "learning_rate": 6.042849299596321e-06, "loss": 0.0356, "step": 84980 }, { "grad_norm": 0.3686612546443939, "learning_rate": 6.034971960008013e-06, "loss": 0.0299, "step": 84990 }, { "grad_norm": 0.4067167341709137, "learning_rate": 6.027099428369082e-06, "loss": 0.0287, "step": 85000 }, { "grad_norm": 0.8295887112617493, "learning_rate": 6.019231705540435e-06, "loss": 0.0357, "step": 85010 }, { "grad_norm": 0.4290846884250641, "learning_rate": 6.011368792382488e-06, "loss": 0.0268, "step": 85020 }, { "grad_norm": 0.28398096561431885, "learning_rate": 6.00351068975511e-06, "loss": 0.0275, "step": 85030 }, { "grad_norm": 0.37634557485580444, "learning_rate": 5.9956573985176674e-06, "loss": 0.0316, "step": 85040 }, { "grad_norm": 0.355917751789093, "learning_rate": 5.987808919528959e-06, "loss": 0.0305, "step": 85050 }, { "grad_norm": 0.4833033084869385, "learning_rate": 5.979965253647307e-06, "loss": 0.0424, "step": 85060 }, { "grad_norm": 0.5162011981010437, "learning_rate": 5.9721264017304655e-06, "loss": 0.0314, "step": 85070 }, { "grad_norm": 0.3254619240760803, "learning_rate": 5.964292364635682e-06, "loss": 0.0341, "step": 85080 }, { "grad_norm": 0.3520101308822632, "learning_rate": 5.956463143219682e-06, "loss": 0.0327, "step": 85090 }, { "grad_norm": 0.3184966444969177, "learning_rate": 5.948638738338663e-06, "loss": 0.0301, "step": 85100 }, { "grad_norm": 0.5135281085968018, "learning_rate": 5.940819150848271e-06, "loss": 0.0333, "step": 85110 }, { "grad_norm": 0.21790871024131775, "learning_rate": 5.933004381603663e-06, "loss": 0.0282, "step": 85120 }, { "grad_norm": 0.5035083293914795, "learning_rate": 5.92519443145943e-06, "loss": 0.029, "step": 85130 }, { "grad_norm": 0.21323569118976593, "learning_rate": 5.917389301269671e-06, "loss": 0.0249, "step": 85140 }, { "grad_norm": 0.38938337564468384, "learning_rate": 5.9095889918879435e-06, "loss": 0.0236, "step": 85150 }, { "grad_norm": 0.3460606038570404, "learning_rate": 5.901793504167286e-06, "loss": 0.0221, "step": 85160 }, { "grad_norm": 0.5148134231567383, "learning_rate": 5.894002838960172e-06, "loss": 0.0295, "step": 85170 }, { "grad_norm": 0.39497247338294983, "learning_rate": 5.886216997118604e-06, "loss": 0.0241, "step": 85180 }, { "grad_norm": 0.4272262454032898, "learning_rate": 5.878435979494023e-06, "loss": 0.031, "step": 85190 }, { "grad_norm": 0.4770191013813019, "learning_rate": 5.870659786937344e-06, "loss": 0.0323, "step": 85200 }, { "grad_norm": 0.501624345779419, "learning_rate": 5.862888420298962e-06, "loss": 0.0315, "step": 85210 }, { "grad_norm": 0.8664355874061584, "learning_rate": 5.855121880428743e-06, "loss": 0.0277, "step": 85220 }, { "grad_norm": 0.5868890285491943, "learning_rate": 5.84736016817603e-06, "loss": 0.0281, "step": 85230 }, { "grad_norm": 0.3081628978252411, "learning_rate": 5.839603284389622e-06, "loss": 0.0212, "step": 85240 }, { "grad_norm": 0.3403080999851227, "learning_rate": 5.831851229917812e-06, "loss": 0.0246, "step": 85250 }, { "grad_norm": 0.28415971994400024, "learning_rate": 5.824104005608338e-06, "loss": 0.0293, "step": 85260 }, { "grad_norm": 0.25964289903640747, "learning_rate": 5.81636161230843e-06, "loss": 0.0337, "step": 85270 }, { "grad_norm": 0.43607717752456665, "learning_rate": 5.808624050864791e-06, "loss": 0.0321, "step": 85280 }, { "grad_norm": 0.3056131899356842, "learning_rate": 5.800891322123592e-06, "loss": 0.0297, "step": 85290 }, { "grad_norm": 0.9931634664535522, "learning_rate": 5.793163426930454e-06, "loss": 0.0297, "step": 85300 }, { "grad_norm": 0.3031926155090332, "learning_rate": 5.785440366130512e-06, "loss": 0.0241, "step": 85310 }, { "grad_norm": 0.23661020398139954, "learning_rate": 5.777722140568326e-06, "loss": 0.0304, "step": 85320 }, { "grad_norm": 0.416465699672699, "learning_rate": 5.7700087510879685e-06, "loss": 0.0291, "step": 85330 }, { "grad_norm": 0.5955694913864136, "learning_rate": 5.762300198532933e-06, "loss": 0.027, "step": 85340 }, { "grad_norm": 0.4193984568119049, "learning_rate": 5.7545964837462606e-06, "loss": 0.0275, "step": 85350 }, { "grad_norm": 0.43477579951286316, "learning_rate": 5.746897607570384e-06, "loss": 0.0279, "step": 85360 }, { "grad_norm": 0.3254978358745575, "learning_rate": 5.7392035708472626e-06, "loss": 0.023, "step": 85370 }, { "grad_norm": 0.7127143740653992, "learning_rate": 5.731514374418278e-06, "loss": 0.0285, "step": 85380 }, { "grad_norm": 0.34084194898605347, "learning_rate": 5.7238300191243325e-06, "loss": 0.0316, "step": 85390 }, { "grad_norm": 1.0260313749313354, "learning_rate": 5.716150505805762e-06, "loss": 0.0297, "step": 85400 }, { "grad_norm": 0.8641685843467712, "learning_rate": 5.708475835302407e-06, "loss": 0.0244, "step": 85410 }, { "grad_norm": 0.3063722252845764, "learning_rate": 5.70080600845353e-06, "loss": 0.0394, "step": 85420 }, { "grad_norm": 0.7325237989425659, "learning_rate": 5.693141026097909e-06, "loss": 0.0254, "step": 85430 }, { "grad_norm": 0.7780413627624512, "learning_rate": 5.6854808890737745e-06, "loss": 0.0316, "step": 85440 }, { "grad_norm": 0.3421507477760315, "learning_rate": 5.6778255982188225e-06, "loss": 0.0325, "step": 85450 }, { "grad_norm": 0.9551219940185547, "learning_rate": 5.670175154370217e-06, "loss": 0.0321, "step": 85460 }, { "grad_norm": 0.4775649905204773, "learning_rate": 5.662529558364616e-06, "loss": 0.0328, "step": 85470 }, { "grad_norm": 0.8179839253425598, "learning_rate": 5.654888811038128e-06, "loss": 0.0275, "step": 85480 }, { "grad_norm": 0.5902443528175354, "learning_rate": 5.647252913226314e-06, "loss": 0.0243, "step": 85490 }, { "grad_norm": 0.5260780453681946, "learning_rate": 5.639621865764255e-06, "loss": 0.0361, "step": 85500 }, { "grad_norm": 0.4869716167449951, "learning_rate": 5.63199566948644e-06, "loss": 0.0279, "step": 85510 }, { "grad_norm": 0.4057789146900177, "learning_rate": 5.6243743252268835e-06, "loss": 0.0315, "step": 85520 }, { "grad_norm": 0.4611128270626068, "learning_rate": 5.616757833819009e-06, "loss": 0.0342, "step": 85530 }, { "grad_norm": 0.3822740316390991, "learning_rate": 5.609146196095794e-06, "loss": 0.025, "step": 85540 }, { "grad_norm": 0.3823794424533844, "learning_rate": 5.6015394128896005e-06, "loss": 0.0251, "step": 85550 }, { "grad_norm": 0.3379172682762146, "learning_rate": 5.593937485032308e-06, "loss": 0.0219, "step": 85560 }, { "grad_norm": 0.33209189772605896, "learning_rate": 5.586340413355246e-06, "loss": 0.0304, "step": 85570 }, { "grad_norm": 0.2768736481666565, "learning_rate": 5.578748198689227e-06, "loss": 0.029, "step": 85580 }, { "grad_norm": 0.42783159017562866, "learning_rate": 5.571160841864504e-06, "loss": 0.0268, "step": 85590 }, { "grad_norm": 0.4238987863063812, "learning_rate": 5.563578343710846e-06, "loss": 0.0248, "step": 85600 }, { "grad_norm": 0.4035227596759796, "learning_rate": 5.556000705057446e-06, "loss": 0.0253, "step": 85610 }, { "grad_norm": 0.24545039236545563, "learning_rate": 5.548427926732996e-06, "loss": 0.0259, "step": 85620 }, { "grad_norm": 0.502693772315979, "learning_rate": 5.540860009565629e-06, "loss": 0.0186, "step": 85630 }, { "grad_norm": 0.3599339425563812, "learning_rate": 5.533296954382966e-06, "loss": 0.0317, "step": 85640 }, { "grad_norm": 0.2627791464328766, "learning_rate": 5.525738762012095e-06, "loss": 0.0275, "step": 85650 }, { "grad_norm": 0.9769049286842346, "learning_rate": 5.518185433279577e-06, "loss": 0.0342, "step": 85660 }, { "grad_norm": 0.33588626980781555, "learning_rate": 5.510636969011407e-06, "loss": 0.0366, "step": 85670 }, { "grad_norm": 0.29601067304611206, "learning_rate": 5.503093370033096e-06, "loss": 0.0266, "step": 85680 }, { "grad_norm": 1.1459623575210571, "learning_rate": 5.495554637169603e-06, "loss": 0.0276, "step": 85690 }, { "grad_norm": 0.8995686769485474, "learning_rate": 5.488020771245328e-06, "loss": 0.0347, "step": 85700 }, { "grad_norm": 0.4017139673233032, "learning_rate": 5.480491773084195e-06, "loss": 0.0317, "step": 85710 }, { "grad_norm": 0.2713452875614166, "learning_rate": 5.472967643509525e-06, "loss": 0.0304, "step": 85720 }, { "grad_norm": 0.2500673830509186, "learning_rate": 5.46544838334419e-06, "loss": 0.0256, "step": 85730 }, { "grad_norm": 0.2682695686817169, "learning_rate": 5.4579339934104525e-06, "loss": 0.0263, "step": 85740 }, { "grad_norm": 0.4250544309616089, "learning_rate": 5.450424474530091e-06, "loss": 0.0472, "step": 85750 }, { "grad_norm": 0.4236310124397278, "learning_rate": 5.442919827524323e-06, "loss": 0.0272, "step": 85760 }, { "grad_norm": 0.31301531195640564, "learning_rate": 5.435420053213863e-06, "loss": 0.027, "step": 85770 }, { "grad_norm": 0.39696457982063293, "learning_rate": 5.427925152418845e-06, "loss": 0.0421, "step": 85780 }, { "grad_norm": 0.4999596178531647, "learning_rate": 5.420435125958934e-06, "loss": 0.0317, "step": 85790 }, { "grad_norm": 0.3229896128177643, "learning_rate": 5.412949974653209e-06, "loss": 0.0254, "step": 85800 }, { "grad_norm": 0.38278523087501526, "learning_rate": 5.405469699320248e-06, "loss": 0.024, "step": 85810 }, { "grad_norm": 0.5216889381408691, "learning_rate": 5.397994300778064e-06, "loss": 0.026, "step": 85820 }, { "grad_norm": 0.45352211594581604, "learning_rate": 5.39052377984417e-06, "loss": 0.0322, "step": 85830 }, { "grad_norm": 0.3119920790195465, "learning_rate": 5.383058137335523e-06, "loss": 0.0293, "step": 85840 }, { "grad_norm": 0.44056132435798645, "learning_rate": 5.375597374068553e-06, "loss": 0.0355, "step": 85850 }, { "grad_norm": 0.5120478868484497, "learning_rate": 5.368141490859163e-06, "loss": 0.0246, "step": 85860 }, { "grad_norm": 0.45017558336257935, "learning_rate": 5.360690488522729e-06, "loss": 0.034, "step": 85870 }, { "grad_norm": 0.2603963613510132, "learning_rate": 5.353244367874055e-06, "loss": 0.031, "step": 85880 }, { "grad_norm": 0.38236501812934875, "learning_rate": 5.345803129727456e-06, "loss": 0.0315, "step": 85890 }, { "grad_norm": 0.29904553294181824, "learning_rate": 5.338366774896697e-06, "loss": 0.0174, "step": 85900 }, { "grad_norm": 0.48907271027565, "learning_rate": 5.330935304194995e-06, "loss": 0.0304, "step": 85910 }, { "grad_norm": 0.5067557692527771, "learning_rate": 5.323508718435044e-06, "loss": 0.0304, "step": 85920 }, { "grad_norm": 0.2842908203601837, "learning_rate": 5.316087018429011e-06, "loss": 0.0228, "step": 85930 }, { "grad_norm": 0.23123662173748016, "learning_rate": 5.30867020498853e-06, "loss": 0.0215, "step": 85940 }, { "grad_norm": 0.4830101728439331, "learning_rate": 5.30125827892467e-06, "loss": 0.0287, "step": 85950 }, { "grad_norm": 0.6042002439498901, "learning_rate": 5.293851241048015e-06, "loss": 0.0412, "step": 85960 }, { "grad_norm": 0.3502817153930664, "learning_rate": 5.286449092168555e-06, "loss": 0.0334, "step": 85970 }, { "grad_norm": 0.6502758264541626, "learning_rate": 5.2790518330958105e-06, "loss": 0.0328, "step": 85980 }, { "grad_norm": 0.3904159665107727, "learning_rate": 5.271659464638717e-06, "loss": 0.0257, "step": 85990 }, { "grad_norm": 0.43013590574264526, "learning_rate": 5.264271987605701e-06, "loss": 0.0297, "step": 86000 }, { "grad_norm": 0.24681438505649567, "learning_rate": 5.256889402804632e-06, "loss": 0.0278, "step": 86010 }, { "grad_norm": 0.37273290753364563, "learning_rate": 5.249511711042881e-06, "loss": 0.0277, "step": 86020 }, { "grad_norm": 0.3541703224182129, "learning_rate": 5.242138913127237e-06, "loss": 0.0229, "step": 86030 }, { "grad_norm": 0.46922361850738525, "learning_rate": 5.234771009863987e-06, "loss": 0.0273, "step": 86040 }, { "grad_norm": 0.5026370286941528, "learning_rate": 5.227408002058881e-06, "loss": 0.0367, "step": 86050 }, { "grad_norm": 0.3974703550338745, "learning_rate": 5.220049890517126e-06, "loss": 0.035, "step": 86060 }, { "grad_norm": 0.42278754711151123, "learning_rate": 5.212696676043377e-06, "loss": 0.0254, "step": 86070 }, { "grad_norm": 0.3999587595462799, "learning_rate": 5.205348359441798e-06, "loss": 0.0307, "step": 86080 }, { "grad_norm": 0.4163150191307068, "learning_rate": 5.198004941515966e-06, "loss": 0.0333, "step": 86090 }, { "grad_norm": 0.6097733378410339, "learning_rate": 5.190666423068952e-06, "loss": 0.0274, "step": 86100 }, { "grad_norm": 0.46530023217201233, "learning_rate": 5.183332804903284e-06, "loss": 0.0216, "step": 86110 }, { "grad_norm": 0.34642255306243896, "learning_rate": 5.176004087820979e-06, "loss": 0.0272, "step": 86120 }, { "grad_norm": 0.6235291361808777, "learning_rate": 5.168680272623455e-06, "loss": 0.0302, "step": 86130 }, { "grad_norm": 0.30060747265815735, "learning_rate": 5.161361360111661e-06, "loss": 0.0275, "step": 86140 }, { "grad_norm": 0.23869292438030243, "learning_rate": 5.154047351085984e-06, "loss": 0.0211, "step": 86150 }, { "grad_norm": 0.3963160514831543, "learning_rate": 5.1467382463462385e-06, "loss": 0.0309, "step": 86160 }, { "grad_norm": 1.1118801832199097, "learning_rate": 5.139434046691788e-06, "loss": 0.0318, "step": 86170 }, { "grad_norm": 0.5147987008094788, "learning_rate": 5.13213475292137e-06, "loss": 0.0295, "step": 86180 }, { "grad_norm": 0.4197653830051422, "learning_rate": 5.12484036583325e-06, "loss": 0.022, "step": 86190 }, { "grad_norm": 0.2529968321323395, "learning_rate": 5.117550886225109e-06, "loss": 0.0229, "step": 86200 }, { "grad_norm": 0.4844144284725189, "learning_rate": 5.11026631489413e-06, "loss": 0.0266, "step": 86210 }, { "grad_norm": 0.40640774369239807, "learning_rate": 5.102986652636926e-06, "loss": 0.0324, "step": 86220 }, { "grad_norm": 0.3477059304714203, "learning_rate": 5.095711900249606e-06, "loss": 0.0234, "step": 86230 }, { "grad_norm": 0.6655848026275635, "learning_rate": 5.088442058527715e-06, "loss": 0.0332, "step": 86240 }, { "grad_norm": 0.304798424243927, "learning_rate": 5.08117712826629e-06, "loss": 0.0285, "step": 86250 }, { "grad_norm": 0.42791762948036194, "learning_rate": 5.073917110259785e-06, "loss": 0.0329, "step": 86260 }, { "grad_norm": 0.3483229875564575, "learning_rate": 5.066662005302175e-06, "loss": 0.0245, "step": 86270 }, { "grad_norm": 0.6382343173027039, "learning_rate": 5.059411814186838e-06, "loss": 0.0282, "step": 86280 }, { "grad_norm": 0.4002556800842285, "learning_rate": 5.05216653770666e-06, "loss": 0.0353, "step": 86290 }, { "grad_norm": 0.42120081186294556, "learning_rate": 5.044926176653969e-06, "loss": 0.0279, "step": 86300 }, { "grad_norm": 0.6399075984954834, "learning_rate": 5.037690731820571e-06, "loss": 0.0274, "step": 86310 }, { "grad_norm": 0.2982192039489746, "learning_rate": 5.030460203997711e-06, "loss": 0.027, "step": 86320 }, { "grad_norm": 0.3615283966064453, "learning_rate": 5.0232345939761174e-06, "loss": 0.0257, "step": 86330 }, { "grad_norm": 0.4430600106716156, "learning_rate": 5.016013902545957e-06, "loss": 0.0312, "step": 86340 }, { "grad_norm": 1.1604095697402954, "learning_rate": 5.008798130496889e-06, "loss": 0.0397, "step": 86350 }, { "grad_norm": 0.646428108215332, "learning_rate": 5.001587278618008e-06, "loss": 0.0339, "step": 86360 }, { "grad_norm": 0.40552109479904175, "learning_rate": 4.994381347697891e-06, "loss": 0.0275, "step": 86370 }, { "grad_norm": 0.3542039394378662, "learning_rate": 4.987180338524572e-06, "loss": 0.0263, "step": 86380 }, { "grad_norm": 0.4730452299118042, "learning_rate": 4.9799842518855275e-06, "loss": 0.0292, "step": 86390 }, { "grad_norm": 0.5668905973434448, "learning_rate": 4.9727930885677285e-06, "loss": 0.0382, "step": 86400 }, { "grad_norm": 0.24656058847904205, "learning_rate": 4.965606849357573e-06, "loss": 0.0312, "step": 86410 }, { "grad_norm": 0.2814961373806, "learning_rate": 4.9584255350409445e-06, "loss": 0.0253, "step": 86420 }, { "grad_norm": 0.46461158990859985, "learning_rate": 4.951249146403175e-06, "loss": 0.0302, "step": 86430 }, { "grad_norm": 0.36861035227775574, "learning_rate": 4.9440776842290815e-06, "loss": 0.0306, "step": 86440 }, { "grad_norm": 0.38617220520973206, "learning_rate": 4.936911149302903e-06, "loss": 0.0339, "step": 86450 }, { "grad_norm": 0.45155179500579834, "learning_rate": 4.929749542408379e-06, "loss": 0.0288, "step": 86460 }, { "grad_norm": 0.2883332669734955, "learning_rate": 4.922592864328673e-06, "loss": 0.0254, "step": 86470 }, { "grad_norm": 1.5732938051223755, "learning_rate": 4.915441115846436e-06, "loss": 0.0444, "step": 86480 }, { "grad_norm": 0.3458216190338135, "learning_rate": 4.908294297743782e-06, "loss": 0.0337, "step": 86490 }, { "grad_norm": 0.4014202654361725, "learning_rate": 4.901152410802273e-06, "loss": 0.0276, "step": 86500 }, { "grad_norm": 0.8540729880332947, "learning_rate": 4.894015455802919e-06, "loss": 0.0355, "step": 86510 }, { "grad_norm": 0.552038311958313, "learning_rate": 4.8868834335262346e-06, "loss": 0.0368, "step": 86520 }, { "grad_norm": 0.27179425954818726, "learning_rate": 4.87975634475214e-06, "loss": 0.0268, "step": 86530 }, { "grad_norm": 0.39836031198501587, "learning_rate": 4.8726341902600614e-06, "loss": 0.0307, "step": 86540 }, { "grad_norm": 0.3593890070915222, "learning_rate": 4.865516970828843e-06, "loss": 0.0289, "step": 86550 }, { "grad_norm": 0.6308100819587708, "learning_rate": 4.85840468723685e-06, "loss": 0.032, "step": 86560 }, { "grad_norm": 0.44149139523506165, "learning_rate": 4.85129734026184e-06, "loss": 0.0294, "step": 86570 }, { "grad_norm": 0.4435039460659027, "learning_rate": 4.844194930681073e-06, "loss": 0.0194, "step": 86580 }, { "grad_norm": 0.26087671518325806, "learning_rate": 4.837097459271267e-06, "loss": 0.0394, "step": 86590 }, { "grad_norm": 0.3997843861579895, "learning_rate": 4.830004926808573e-06, "loss": 0.0259, "step": 86600 }, { "grad_norm": 0.6063441038131714, "learning_rate": 4.822917334068627e-06, "loss": 0.0259, "step": 86610 }, { "grad_norm": 2.0228092670440674, "learning_rate": 4.81583468182652e-06, "loss": 0.0242, "step": 86620 }, { "grad_norm": 0.22323288023471832, "learning_rate": 4.808756970856809e-06, "loss": 0.0339, "step": 86630 }, { "grad_norm": 0.3909667134284973, "learning_rate": 4.801684201933476e-06, "loss": 0.0245, "step": 86640 }, { "grad_norm": 0.30702632665634155, "learning_rate": 4.794616375830019e-06, "loss": 0.0251, "step": 86650 }, { "grad_norm": 0.5978437066078186, "learning_rate": 4.787553493319341e-06, "loss": 0.0269, "step": 86660 }, { "grad_norm": 0.19164316356182098, "learning_rate": 4.780495555173836e-06, "loss": 0.0237, "step": 86670 }, { "grad_norm": 0.2785320580005646, "learning_rate": 4.7734425621653475e-06, "loss": 0.0293, "step": 86680 }, { "grad_norm": 0.5633301138877869, "learning_rate": 4.766394515065198e-06, "loss": 0.0228, "step": 86690 }, { "grad_norm": 0.2789432108402252, "learning_rate": 4.75935141464412e-06, "loss": 0.0303, "step": 86700 }, { "grad_norm": 1.3491077423095703, "learning_rate": 4.75231326167237e-06, "loss": 0.0272, "step": 86710 }, { "grad_norm": 0.32674872875213623, "learning_rate": 4.745280056919599e-06, "loss": 0.0249, "step": 86720 }, { "grad_norm": 0.3210064172744751, "learning_rate": 4.738251801154975e-06, "loss": 0.0246, "step": 86730 }, { "grad_norm": 0.2095457762479782, "learning_rate": 4.731228495147061e-06, "loss": 0.0288, "step": 86740 }, { "grad_norm": 0.5426259636878967, "learning_rate": 4.7242101396639595e-06, "loss": 0.0306, "step": 86750 }, { "grad_norm": 0.5163301229476929, "learning_rate": 4.717196735473151e-06, "loss": 0.0268, "step": 86760 }, { "grad_norm": 0.34102821350097656, "learning_rate": 4.710188283341643e-06, "loss": 0.0305, "step": 86770 }, { "grad_norm": 0.43077901005744934, "learning_rate": 4.703184784035835e-06, "loss": 0.0363, "step": 86780 }, { "grad_norm": 0.3681461811065674, "learning_rate": 4.696186238321642e-06, "loss": 0.0256, "step": 86790 }, { "grad_norm": 0.4636329710483551, "learning_rate": 4.689192646964408e-06, "loss": 0.0327, "step": 86800 }, { "grad_norm": 0.68118816614151, "learning_rate": 4.682204010728952e-06, "loss": 0.0326, "step": 86810 }, { "grad_norm": 0.392211377620697, "learning_rate": 4.67522033037952e-06, "loss": 0.0185, "step": 86820 }, { "grad_norm": 0.32667243480682373, "learning_rate": 4.668241606679846e-06, "loss": 0.0341, "step": 86830 }, { "grad_norm": 0.30809321999549866, "learning_rate": 4.66126784039313e-06, "loss": 0.0267, "step": 86840 }, { "grad_norm": 0.3973626494407654, "learning_rate": 4.6542990322819836e-06, "loss": 0.0346, "step": 86850 }, { "grad_norm": 0.2128410041332245, "learning_rate": 4.64733518310852e-06, "loss": 0.0224, "step": 86860 }, { "grad_norm": 0.5127125382423401, "learning_rate": 4.640376293634291e-06, "loss": 0.0264, "step": 86870 }, { "grad_norm": 0.5385788083076477, "learning_rate": 4.633422364620327e-06, "loss": 0.0347, "step": 86880 }, { "grad_norm": 0.5046197175979614, "learning_rate": 4.6264733968270745e-06, "loss": 0.036, "step": 86890 }, { "grad_norm": 0.4687856435775757, "learning_rate": 4.619529391014482e-06, "loss": 0.0293, "step": 86900 }, { "grad_norm": 0.3322535753250122, "learning_rate": 4.61259034794192e-06, "loss": 0.0266, "step": 86910 }, { "grad_norm": 0.7081454396247864, "learning_rate": 4.605656268368247e-06, "loss": 0.0347, "step": 86920 }, { "grad_norm": 0.28216153383255005, "learning_rate": 4.598727153051735e-06, "loss": 0.0292, "step": 86930 }, { "grad_norm": 0.28024524450302124, "learning_rate": 4.591803002750178e-06, "loss": 0.0327, "step": 86940 }, { "grad_norm": 0.3101784288883209, "learning_rate": 4.58488381822077e-06, "loss": 0.026, "step": 86950 }, { "grad_norm": 1.0704995393753052, "learning_rate": 4.5779696002201946e-06, "loss": 0.0288, "step": 86960 }, { "grad_norm": 0.38423633575439453, "learning_rate": 4.571060349504563e-06, "loss": 0.03, "step": 86970 }, { "grad_norm": 0.519639253616333, "learning_rate": 4.564156066829478e-06, "loss": 0.0267, "step": 86980 }, { "grad_norm": 0.3690911531448364, "learning_rate": 4.557256752949957e-06, "loss": 0.0346, "step": 86990 }, { "grad_norm": 0.6871488094329834, "learning_rate": 4.550362408620534e-06, "loss": 0.0263, "step": 87000 }, { "grad_norm": 0.22262631356716156, "learning_rate": 4.543473034595136e-06, "loss": 0.0274, "step": 87010 }, { "grad_norm": 0.5424175262451172, "learning_rate": 4.536588631627192e-06, "loss": 0.034, "step": 87020 }, { "grad_norm": 0.4053415358066559, "learning_rate": 4.5297092004695565e-06, "loss": 0.0303, "step": 87030 }, { "grad_norm": 0.27019572257995605, "learning_rate": 4.522834741874555e-06, "loss": 0.0299, "step": 87040 }, { "grad_norm": 0.37672165036201477, "learning_rate": 4.515965256593974e-06, "loss": 0.0275, "step": 87050 }, { "grad_norm": 0.35376086831092834, "learning_rate": 4.509100745379058e-06, "loss": 0.0298, "step": 87060 }, { "grad_norm": 0.25610437989234924, "learning_rate": 4.502241208980479e-06, "loss": 0.0296, "step": 87070 }, { "grad_norm": 0.7082431316375732, "learning_rate": 4.495386648148398e-06, "loss": 0.0252, "step": 87080 }, { "grad_norm": 0.31983134150505066, "learning_rate": 4.488537063632431e-06, "loss": 0.0301, "step": 87090 }, { "grad_norm": 0.5561509728431702, "learning_rate": 4.481692456181608e-06, "loss": 0.0338, "step": 87100 }, { "grad_norm": 0.5083333253860474, "learning_rate": 4.474852826544479e-06, "loss": 0.028, "step": 87110 }, { "grad_norm": 0.5208888649940491, "learning_rate": 4.468018175468974e-06, "loss": 0.0245, "step": 87120 }, { "grad_norm": 0.3684374988079071, "learning_rate": 4.461188503702568e-06, "loss": 0.0371, "step": 87130 }, { "grad_norm": 0.9621937274932861, "learning_rate": 4.454363811992107e-06, "loss": 0.0373, "step": 87140 }, { "grad_norm": 0.5995425581932068, "learning_rate": 4.447544101083956e-06, "loss": 0.0275, "step": 87150 }, { "grad_norm": 1.546789526939392, "learning_rate": 4.440729371723884e-06, "loss": 0.0343, "step": 87160 }, { "grad_norm": 0.3109491169452667, "learning_rate": 4.433919624657162e-06, "loss": 0.0232, "step": 87170 }, { "grad_norm": 0.776978611946106, "learning_rate": 4.427114860628462e-06, "loss": 0.032, "step": 87180 }, { "grad_norm": 0.38608279824256897, "learning_rate": 4.420315080381982e-06, "loss": 0.0273, "step": 87190 }, { "grad_norm": 0.2833280861377716, "learning_rate": 4.4135202846613045e-06, "loss": 0.0297, "step": 87200 }, { "grad_norm": 0.49752625823020935, "learning_rate": 4.406730474209525e-06, "loss": 0.0287, "step": 87210 }, { "grad_norm": 0.3732800781726837, "learning_rate": 4.399945649769138e-06, "loss": 0.0378, "step": 87220 }, { "grad_norm": 0.4993225932121277, "learning_rate": 4.393165812082151e-06, "loss": 0.0244, "step": 87230 }, { "grad_norm": 0.34085536003112793, "learning_rate": 4.386390961889969e-06, "loss": 0.036, "step": 87240 }, { "grad_norm": 0.567152202129364, "learning_rate": 4.3796210999334895e-06, "loss": 0.0214, "step": 87250 }, { "grad_norm": 0.5499762892723083, "learning_rate": 4.372856226953065e-06, "loss": 0.0281, "step": 87260 }, { "grad_norm": 1.8843556642532349, "learning_rate": 4.366096343688486e-06, "loss": 0.0265, "step": 87270 }, { "grad_norm": 0.39179471135139465, "learning_rate": 4.359341450878995e-06, "loss": 0.0213, "step": 87280 }, { "grad_norm": 0.4637346863746643, "learning_rate": 4.352591549263302e-06, "loss": 0.032, "step": 87290 }, { "grad_norm": 0.3889254629611969, "learning_rate": 4.345846639579582e-06, "loss": 0.0278, "step": 87300 }, { "grad_norm": 0.6992796659469604, "learning_rate": 4.339106722565417e-06, "loss": 0.0294, "step": 87310 }, { "grad_norm": 0.4118526875972748, "learning_rate": 4.332371798957896e-06, "loss": 0.0371, "step": 87320 }, { "grad_norm": 0.3717088997364044, "learning_rate": 4.3256418694935295e-06, "loss": 0.0361, "step": 87330 }, { "grad_norm": 0.369171679019928, "learning_rate": 4.3189169349083125e-06, "loss": 0.0242, "step": 87340 }, { "grad_norm": 0.37537792325019836, "learning_rate": 4.312196995937645e-06, "loss": 0.0282, "step": 87350 }, { "grad_norm": 0.34645718336105347, "learning_rate": 4.305482053316434e-06, "loss": 0.0295, "step": 87360 }, { "grad_norm": 0.4262024760246277, "learning_rate": 4.2987721077789925e-06, "loss": 0.0294, "step": 87370 }, { "grad_norm": 0.45588600635528564, "learning_rate": 4.2920671600591325e-06, "loss": 0.0271, "step": 87380 }, { "grad_norm": 0.3455909788608551, "learning_rate": 4.285367210890084e-06, "loss": 0.0264, "step": 87390 }, { "grad_norm": 0.21383140981197357, "learning_rate": 4.278672261004551e-06, "loss": 0.0299, "step": 87400 }, { "grad_norm": 0.3675417900085449, "learning_rate": 4.2719823111346704e-06, "loss": 0.0318, "step": 87410 }, { "grad_norm": 0.30192553997039795, "learning_rate": 4.265297362012061e-06, "loss": 0.0201, "step": 87420 }, { "grad_norm": 0.30682578682899475, "learning_rate": 4.258617414367766e-06, "loss": 0.027, "step": 87430 }, { "grad_norm": 0.2640503942966461, "learning_rate": 4.251942468932296e-06, "loss": 0.025, "step": 87440 }, { "grad_norm": 0.6380057334899902, "learning_rate": 4.2452725264356175e-06, "loss": 0.0239, "step": 87450 }, { "grad_norm": 0.756993293762207, "learning_rate": 4.2386075876071575e-06, "loss": 0.0258, "step": 87460 }, { "grad_norm": 0.30476853251457214, "learning_rate": 4.231947653175755e-06, "loss": 0.0321, "step": 87470 }, { "grad_norm": 0.6330248117446899, "learning_rate": 4.225292723869761e-06, "loss": 0.0358, "step": 87480 }, { "grad_norm": 0.2713213860988617, "learning_rate": 4.218642800416916e-06, "loss": 0.0305, "step": 87490 }, { "grad_norm": 0.6381987929344177, "learning_rate": 4.211997883544472e-06, "loss": 0.0221, "step": 87500 }, { "grad_norm": 0.3640121817588806, "learning_rate": 4.205357973979096e-06, "loss": 0.0298, "step": 87510 }, { "grad_norm": 0.2758731544017792, "learning_rate": 4.198723072446925e-06, "loss": 0.0242, "step": 87520 }, { "grad_norm": 0.4272240400314331, "learning_rate": 4.192093179673534e-06, "loss": 0.0266, "step": 87530 }, { "grad_norm": 0.31426072120666504, "learning_rate": 4.185468296383959e-06, "loss": 0.0312, "step": 87540 }, { "grad_norm": 0.7490639686584473, "learning_rate": 4.1788484233027005e-06, "loss": 0.0397, "step": 87550 }, { "grad_norm": 0.5795135498046875, "learning_rate": 4.172233561153677e-06, "loss": 0.0225, "step": 87560 }, { "grad_norm": 0.2630649209022522, "learning_rate": 4.165623710660288e-06, "loss": 0.0281, "step": 87570 }, { "grad_norm": 0.3308180868625641, "learning_rate": 4.1590188725453835e-06, "loss": 0.0276, "step": 87580 }, { "grad_norm": 1.1118558645248413, "learning_rate": 4.152419047531259e-06, "loss": 0.0317, "step": 87590 }, { "grad_norm": 0.2715725302696228, "learning_rate": 4.145824236339651e-06, "loss": 0.0193, "step": 87600 }, { "grad_norm": 0.31870904564857483, "learning_rate": 4.1392344396917695e-06, "loss": 0.0199, "step": 87610 }, { "grad_norm": 0.40937042236328125, "learning_rate": 4.132649658308252e-06, "loss": 0.0267, "step": 87620 }, { "grad_norm": 0.4527651071548462, "learning_rate": 4.126069892909213e-06, "loss": 0.0211, "step": 87630 }, { "grad_norm": 0.2974901795387268, "learning_rate": 4.119495144214191e-06, "loss": 0.0298, "step": 87640 }, { "grad_norm": 0.3441283404827118, "learning_rate": 4.112925412942214e-06, "loss": 0.0269, "step": 87650 }, { "grad_norm": 0.4076322019100189, "learning_rate": 4.106360699811712e-06, "loss": 0.0273, "step": 87660 }, { "grad_norm": 0.3566869795322418, "learning_rate": 4.099801005540616e-06, "loss": 0.0254, "step": 87670 }, { "grad_norm": 0.3622196316719055, "learning_rate": 4.093246330846256e-06, "loss": 0.0234, "step": 87680 }, { "grad_norm": 0.35677361488342285, "learning_rate": 4.086696676445467e-06, "loss": 0.0261, "step": 87690 }, { "grad_norm": 0.36698487401008606, "learning_rate": 4.0801520430544914e-06, "loss": 0.0196, "step": 87700 }, { "grad_norm": 0.38632333278656006, "learning_rate": 4.073612431389062e-06, "loss": 0.0242, "step": 87710 }, { "grad_norm": 0.4535605013370514, "learning_rate": 4.067077842164319e-06, "loss": 0.0225, "step": 87720 }, { "grad_norm": 0.4970906376838684, "learning_rate": 4.0605482760948945e-06, "loss": 0.0352, "step": 87730 }, { "grad_norm": 0.36993667483329773, "learning_rate": 4.054023733894829e-06, "loss": 0.0301, "step": 87740 }, { "grad_norm": 0.7102916836738586, "learning_rate": 4.047504216277653e-06, "loss": 0.0305, "step": 87750 }, { "grad_norm": 0.34810227155685425, "learning_rate": 4.040989723956323e-06, "loss": 0.0293, "step": 87760 }, { "grad_norm": 0.44347283244132996, "learning_rate": 4.03448025764327e-06, "loss": 0.0399, "step": 87770 }, { "grad_norm": 0.33761027455329895, "learning_rate": 4.027975818050339e-06, "loss": 0.0288, "step": 87780 }, { "grad_norm": 0.4814301133155823, "learning_rate": 4.021476405888858e-06, "loss": 0.0297, "step": 87790 }, { "grad_norm": 0.511761486530304, "learning_rate": 4.01498202186959e-06, "loss": 0.0255, "step": 87800 }, { "grad_norm": 0.7459480166435242, "learning_rate": 4.008492666702751e-06, "loss": 0.0316, "step": 87810 }, { "grad_norm": 0.3247740566730499, "learning_rate": 4.002008341098001e-06, "loss": 0.0322, "step": 87820 }, { "grad_norm": 0.3527098298072815, "learning_rate": 3.995529045764462e-06, "loss": 0.0239, "step": 87830 }, { "grad_norm": 0.2241102159023285, "learning_rate": 3.989054781410712e-06, "loss": 0.0234, "step": 87840 }, { "grad_norm": 1.5127015113830566, "learning_rate": 3.982585548744744e-06, "loss": 0.03, "step": 87850 }, { "grad_norm": 0.3391381502151489, "learning_rate": 3.9761213484740435e-06, "loss": 0.0269, "step": 87860 }, { "grad_norm": 0.4083825945854187, "learning_rate": 3.9696621813055055e-06, "loss": 0.0257, "step": 87870 }, { "grad_norm": 0.19733640551567078, "learning_rate": 3.9632080479455044e-06, "loss": 0.0293, "step": 87880 }, { "grad_norm": 0.5668600797653198, "learning_rate": 3.956758949099859e-06, "loss": 0.0318, "step": 87890 }, { "grad_norm": 0.3103303611278534, "learning_rate": 3.9503148854738385e-06, "loss": 0.0303, "step": 87900 }, { "grad_norm": 0.6722822785377502, "learning_rate": 3.943875857772139e-06, "loss": 0.0373, "step": 87910 }, { "grad_norm": 0.5553415417671204, "learning_rate": 3.9374418666989375e-06, "loss": 0.0318, "step": 87920 }, { "grad_norm": 0.38061773777008057, "learning_rate": 3.931012912957838e-06, "loss": 0.0266, "step": 87930 }, { "grad_norm": 0.2493537813425064, "learning_rate": 3.9245889972519045e-06, "loss": 0.0246, "step": 87940 }, { "grad_norm": 0.334529846906662, "learning_rate": 3.9181701202836265e-06, "loss": 0.0333, "step": 87950 }, { "grad_norm": 0.40403851866722107, "learning_rate": 3.911756282755003e-06, "loss": 0.0296, "step": 87960 }, { "grad_norm": 0.689141035079956, "learning_rate": 3.905347485367411e-06, "loss": 0.0293, "step": 87970 }, { "grad_norm": 0.2585696578025818, "learning_rate": 3.898943728821725e-06, "loss": 0.0214, "step": 87980 }, { "grad_norm": 0.3530465066432953, "learning_rate": 3.892545013818227e-06, "loss": 0.0265, "step": 87990 }, { "grad_norm": 1.067083477973938, "learning_rate": 3.886151341056687e-06, "loss": 0.027, "step": 88000 }, { "grad_norm": 0.41470423340797424, "learning_rate": 3.8797627112363066e-06, "loss": 0.0326, "step": 88010 }, { "grad_norm": 0.5250041484832764, "learning_rate": 3.873379125055743e-06, "loss": 0.0246, "step": 88020 }, { "grad_norm": 0.5838326811790466, "learning_rate": 3.867000583213082e-06, "loss": 0.0252, "step": 88030 }, { "grad_norm": 0.3729144036769867, "learning_rate": 3.860627086405871e-06, "loss": 0.0349, "step": 88040 }, { "grad_norm": 0.42788830399513245, "learning_rate": 3.854258635331126e-06, "loss": 0.0307, "step": 88050 }, { "grad_norm": 0.2899990975856781, "learning_rate": 3.8478952306852665e-06, "loss": 0.0275, "step": 88060 }, { "grad_norm": 0.35289204120635986, "learning_rate": 3.841536873164198e-06, "loss": 0.0278, "step": 88070 }, { "grad_norm": 0.455063134431839, "learning_rate": 3.8351835634632626e-06, "loss": 0.0297, "step": 88080 }, { "grad_norm": 0.8337031006813049, "learning_rate": 3.828835302277245e-06, "loss": 0.0398, "step": 88090 }, { "grad_norm": 1.5120437145233154, "learning_rate": 3.822492090300378e-06, "loss": 0.0349, "step": 88100 }, { "grad_norm": 0.333878755569458, "learning_rate": 3.8161539282263566e-06, "loss": 0.0241, "step": 88110 }, { "grad_norm": 0.5100154280662537, "learning_rate": 3.8098208167482995e-06, "loss": 0.0256, "step": 88120 }, { "grad_norm": 0.6996377110481262, "learning_rate": 3.803492756558796e-06, "loss": 0.0296, "step": 88130 }, { "grad_norm": 0.33308374881744385, "learning_rate": 3.797169748349855e-06, "loss": 0.0371, "step": 88140 }, { "grad_norm": 0.4497467279434204, "learning_rate": 3.790851792812977e-06, "loss": 0.0232, "step": 88150 }, { "grad_norm": 0.14701996743679047, "learning_rate": 3.784538890639072e-06, "loss": 0.0235, "step": 88160 }, { "grad_norm": 0.3215484619140625, "learning_rate": 3.7782310425185153e-06, "loss": 0.0297, "step": 88170 }, { "grad_norm": 0.7538844347000122, "learning_rate": 3.77192824914111e-06, "loss": 0.0178, "step": 88180 }, { "grad_norm": 0.37187567353248596, "learning_rate": 3.7656305111961333e-06, "loss": 0.0291, "step": 88190 }, { "grad_norm": 0.3897874355316162, "learning_rate": 3.7593378293722782e-06, "loss": 0.033, "step": 88200 }, { "grad_norm": 0.44965529441833496, "learning_rate": 3.7530502043577333e-06, "loss": 0.0278, "step": 88210 }, { "grad_norm": 0.5956541299819946, "learning_rate": 3.746767636840076e-06, "loss": 0.0264, "step": 88220 }, { "grad_norm": 0.2289106249809265, "learning_rate": 3.74049012750638e-06, "loss": 0.028, "step": 88230 }, { "grad_norm": 1.267748475074768, "learning_rate": 3.734217677043128e-06, "loss": 0.0217, "step": 88240 }, { "grad_norm": 0.4380769431591034, "learning_rate": 3.727950286136267e-06, "loss": 0.027, "step": 88250 }, { "grad_norm": 0.2248060703277588, "learning_rate": 3.7216879554711925e-06, "loss": 0.0279, "step": 88260 }, { "grad_norm": 0.46912288665771484, "learning_rate": 3.715430685732757e-06, "loss": 0.0338, "step": 88270 }, { "grad_norm": 0.5837416052818298, "learning_rate": 3.7091784776052243e-06, "loss": 0.0237, "step": 88280 }, { "grad_norm": 0.4136675000190735, "learning_rate": 3.7029313317723356e-06, "loss": 0.0264, "step": 88290 }, { "grad_norm": 0.24386215209960938, "learning_rate": 3.6966892489172787e-06, "loss": 0.0315, "step": 88300 }, { "grad_norm": 0.8904357552528381, "learning_rate": 3.6904522297226575e-06, "loss": 0.0337, "step": 88310 }, { "grad_norm": 0.2800257205963135, "learning_rate": 3.6842202748705647e-06, "loss": 0.0267, "step": 88320 }, { "grad_norm": 0.385834664106369, "learning_rate": 3.677993385042494e-06, "loss": 0.0316, "step": 88330 }, { "grad_norm": 0.5158926844596863, "learning_rate": 3.67177156091944e-06, "loss": 0.0272, "step": 88340 }, { "grad_norm": 0.5846030116081238, "learning_rate": 3.6655548031817866e-06, "loss": 0.0221, "step": 88350 }, { "grad_norm": 0.5871592164039612, "learning_rate": 3.6593431125094e-06, "loss": 0.0309, "step": 88360 }, { "grad_norm": 0.3669874966144562, "learning_rate": 3.6531364895815766e-06, "loss": 0.0235, "step": 88370 }, { "grad_norm": 0.6693108677864075, "learning_rate": 3.6469349350770677e-06, "loss": 0.0255, "step": 88380 }, { "grad_norm": 0.4156823456287384, "learning_rate": 3.6407384496740525e-06, "loss": 0.0318, "step": 88390 }, { "grad_norm": 0.7497588396072388, "learning_rate": 3.6345470340501954e-06, "loss": 0.0306, "step": 88400 }, { "grad_norm": 0.31973952054977417, "learning_rate": 3.6283606888825537e-06, "loss": 0.0255, "step": 88410 }, { "grad_norm": 0.4795289635658264, "learning_rate": 3.622179414847676e-06, "loss": 0.029, "step": 88420 }, { "grad_norm": 0.7867243885993958, "learning_rate": 3.616003212621527e-06, "loss": 0.0329, "step": 88430 }, { "grad_norm": 0.34609344601631165, "learning_rate": 3.609832082879533e-06, "loss": 0.0246, "step": 88440 }, { "grad_norm": 0.4148257076740265, "learning_rate": 3.6036660262965485e-06, "loss": 0.0306, "step": 88450 }, { "grad_norm": 0.2948029935359955, "learning_rate": 3.597505043546895e-06, "loss": 0.0259, "step": 88460 }, { "grad_norm": 0.5583876371383667, "learning_rate": 3.5913491353043227e-06, "loss": 0.0202, "step": 88470 }, { "grad_norm": 0.3064345121383667, "learning_rate": 3.5851983022420375e-06, "loss": 0.0307, "step": 88480 }, { "grad_norm": 0.5108581185340881, "learning_rate": 3.5790525450326797e-06, "loss": 0.023, "step": 88490 }, { "grad_norm": 0.3065962791442871, "learning_rate": 3.572911864348344e-06, "loss": 0.0233, "step": 88500 }, { "grad_norm": 0.4139716625213623, "learning_rate": 3.5667762608605717e-06, "loss": 0.0305, "step": 88510 }, { "grad_norm": 0.39892756938934326, "learning_rate": 3.5606457352403255e-06, "loss": 0.0256, "step": 88520 }, { "grad_norm": 0.8163975477218628, "learning_rate": 3.554520288158042e-06, "loss": 0.0199, "step": 88530 }, { "grad_norm": 0.2406110167503357, "learning_rate": 3.5483999202835903e-06, "loss": 0.0246, "step": 88540 }, { "grad_norm": 0.38900327682495117, "learning_rate": 3.5422846322862913e-06, "loss": 0.0277, "step": 88550 }, { "grad_norm": 0.20497609674930573, "learning_rate": 3.5361744248348925e-06, "loss": 0.0254, "step": 88560 }, { "grad_norm": 0.285065621137619, "learning_rate": 3.5300692985976047e-06, "loss": 0.0237, "step": 88570 }, { "grad_norm": 0.5046427249908447, "learning_rate": 3.52396925424206e-06, "loss": 0.0224, "step": 88580 }, { "grad_norm": 0.4562508165836334, "learning_rate": 3.5178742924353747e-06, "loss": 0.0288, "step": 88590 }, { "grad_norm": 0.29934456944465637, "learning_rate": 3.511784413844066e-06, "loss": 0.0294, "step": 88600 }, { "grad_norm": 0.6528132557868958, "learning_rate": 3.5056996191341287e-06, "loss": 0.0421, "step": 88610 }, { "grad_norm": 0.5281955599784851, "learning_rate": 3.4996199089709692e-06, "loss": 0.0282, "step": 88620 }, { "grad_norm": 0.17416517436504364, "learning_rate": 3.4935452840194725e-06, "loss": 0.0281, "step": 88630 }, { "grad_norm": 0.4846268892288208, "learning_rate": 3.48747574494393e-06, "loss": 0.0283, "step": 88640 }, { "grad_norm": 0.48847880959510803, "learning_rate": 3.4814112924081154e-06, "loss": 0.0299, "step": 88650 }, { "grad_norm": 0.7369963526725769, "learning_rate": 3.475351927075221e-06, "loss": 0.0289, "step": 88660 }, { "grad_norm": 0.467933714389801, "learning_rate": 3.469297649607894e-06, "loss": 0.0212, "step": 88670 }, { "grad_norm": 0.6555014252662659, "learning_rate": 3.463248460668217e-06, "loss": 0.0281, "step": 88680 }, { "grad_norm": 0.5954755544662476, "learning_rate": 3.457204360917726e-06, "loss": 0.032, "step": 88690 }, { "grad_norm": 0.5634835958480835, "learning_rate": 3.4511653510173827e-06, "loss": 0.0264, "step": 88700 }, { "grad_norm": 0.2809293866157532, "learning_rate": 3.4451314316276074e-06, "loss": 0.0264, "step": 88710 }, { "grad_norm": 0.5360924005508423, "learning_rate": 3.4391026034082674e-06, "loss": 0.0345, "step": 88720 }, { "grad_norm": 0.5123598575592041, "learning_rate": 3.433078867018674e-06, "loss": 0.0274, "step": 88730 }, { "grad_norm": 0.767970860004425, "learning_rate": 3.427060223117551e-06, "loss": 0.0264, "step": 88740 }, { "grad_norm": 0.4740496575832367, "learning_rate": 3.4210466723631042e-06, "loss": 0.0279, "step": 88750 }, { "grad_norm": 0.29921644926071167, "learning_rate": 3.415038215412969e-06, "loss": 0.0204, "step": 88760 }, { "grad_norm": 0.5666257739067078, "learning_rate": 3.4090348529242088e-06, "loss": 0.039, "step": 88770 }, { "grad_norm": 0.7909278273582458, "learning_rate": 3.4030365855533487e-06, "loss": 0.0352, "step": 88780 }, { "grad_norm": 0.5411272048950195, "learning_rate": 3.397043413956347e-06, "loss": 0.0397, "step": 88790 }, { "grad_norm": 0.24106058478355408, "learning_rate": 3.3910553387886237e-06, "loss": 0.0236, "step": 88800 }, { "grad_norm": 0.35171255469322205, "learning_rate": 3.3850723607049996e-06, "loss": 0.0317, "step": 88810 }, { "grad_norm": 0.4438855051994324, "learning_rate": 3.3790944803597902e-06, "loss": 0.0255, "step": 88820 }, { "grad_norm": 0.3013531565666199, "learning_rate": 3.373121698406706e-06, "loss": 0.0331, "step": 88830 }, { "grad_norm": 0.5378166437149048, "learning_rate": 3.3671540154989302e-06, "loss": 0.0233, "step": 88840 }, { "grad_norm": 0.7330234050750732, "learning_rate": 3.361191432289079e-06, "loss": 0.0337, "step": 88850 }, { "grad_norm": 0.6007440686225891, "learning_rate": 3.3552339494292207e-06, "loss": 0.0361, "step": 88860 }, { "grad_norm": 0.5130218863487244, "learning_rate": 3.349281567570839e-06, "loss": 0.0324, "step": 88870 }, { "grad_norm": 0.43605148792266846, "learning_rate": 3.3433342873648965e-06, "loss": 0.0289, "step": 88880 }, { "grad_norm": 0.39564642310142517, "learning_rate": 3.337392109461762e-06, "loss": 0.0275, "step": 88890 }, { "grad_norm": 0.3092896640300751, "learning_rate": 3.331455034511266e-06, "loss": 0.027, "step": 88900 }, { "grad_norm": 0.27425462007522583, "learning_rate": 3.325523063162689e-06, "loss": 0.0319, "step": 88910 }, { "grad_norm": 0.4051603376865387, "learning_rate": 3.3195961960647393e-06, "loss": 0.0227, "step": 88920 }, { "grad_norm": 0.30915871262550354, "learning_rate": 3.31367443386556e-06, "loss": 0.0176, "step": 88930 }, { "grad_norm": 0.24988794326782227, "learning_rate": 3.3077577772127556e-06, "loss": 0.0283, "step": 88940 }, { "grad_norm": 0.3990849256515503, "learning_rate": 3.301846226753358e-06, "loss": 0.0248, "step": 88950 }, { "grad_norm": 0.4904240369796753, "learning_rate": 3.2959397831338444e-06, "loss": 0.0326, "step": 88960 }, { "grad_norm": 0.46958988904953003, "learning_rate": 3.2900384470001432e-06, "loss": 0.0271, "step": 88970 }, { "grad_norm": 0.3829829692840576, "learning_rate": 3.28414221899761e-06, "loss": 0.0274, "step": 88980 }, { "grad_norm": 0.6395407915115356, "learning_rate": 3.278251099771046e-06, "loss": 0.0327, "step": 88990 }, { "grad_norm": 0.6767784357070923, "learning_rate": 3.2723650899646906e-06, "loss": 0.0344, "step": 89000 }, { "grad_norm": 0.4351170063018799, "learning_rate": 3.2664841902222466e-06, "loss": 0.0267, "step": 89010 }, { "grad_norm": 0.2876165211200714, "learning_rate": 3.2606084011868208e-06, "loss": 0.0315, "step": 89020 }, { "grad_norm": 0.291553258895874, "learning_rate": 3.254737723500989e-06, "loss": 0.0247, "step": 89030 }, { "grad_norm": 0.29540959000587463, "learning_rate": 3.2488721578067595e-06, "loss": 0.0251, "step": 89040 }, { "grad_norm": 0.2880606949329376, "learning_rate": 3.243011704745591e-06, "loss": 0.0235, "step": 89050 }, { "grad_norm": 0.6413300633430481, "learning_rate": 3.237156364958355e-06, "loss": 0.0261, "step": 89060 }, { "grad_norm": 0.4432308077812195, "learning_rate": 3.231306139085394e-06, "loss": 0.0273, "step": 89070 }, { "grad_norm": 1.3860514163970947, "learning_rate": 3.2254610277664797e-06, "loss": 0.0441, "step": 89080 }, { "grad_norm": 0.39817994832992554, "learning_rate": 3.2196210316408183e-06, "loss": 0.0274, "step": 89090 }, { "grad_norm": 0.2844561040401459, "learning_rate": 3.2137861513470644e-06, "loss": 0.0255, "step": 89100 }, { "grad_norm": 0.31813034415245056, "learning_rate": 3.2079563875233252e-06, "loss": 0.0299, "step": 89110 }, { "grad_norm": 0.43635237216949463, "learning_rate": 3.2021317408071184e-06, "loss": 0.0213, "step": 89120 }, { "grad_norm": 0.6041378974914551, "learning_rate": 3.196312211835434e-06, "loss": 0.0296, "step": 89130 }, { "grad_norm": 0.6133792400360107, "learning_rate": 3.1904978012446686e-06, "loss": 0.0195, "step": 89140 }, { "grad_norm": 0.5493641495704651, "learning_rate": 3.184688509670691e-06, "loss": 0.0216, "step": 89150 }, { "grad_norm": 0.2265411913394928, "learning_rate": 3.178884337748783e-06, "loss": 0.0255, "step": 89160 }, { "grad_norm": 0.44558313488960266, "learning_rate": 3.173085286113697e-06, "loss": 0.0326, "step": 89170 }, { "grad_norm": 0.3752911388874054, "learning_rate": 3.1672913553995986e-06, "loss": 0.038, "step": 89180 }, { "grad_norm": 0.16333293914794922, "learning_rate": 3.1615025462401138e-06, "loss": 0.0177, "step": 89190 }, { "grad_norm": 0.7166647911071777, "learning_rate": 3.1557188592682808e-06, "loss": 0.0318, "step": 89200 }, { "grad_norm": 0.32363057136535645, "learning_rate": 3.1499402951166045e-06, "loss": 0.0232, "step": 89210 }, { "grad_norm": 0.33097371459007263, "learning_rate": 3.144166854417019e-06, "loss": 0.0222, "step": 89220 }, { "grad_norm": 0.5123064517974854, "learning_rate": 3.1383985378009073e-06, "loss": 0.0269, "step": 89230 }, { "grad_norm": 0.613421618938446, "learning_rate": 3.1326353458990654e-06, "loss": 0.0288, "step": 89240 }, { "grad_norm": 0.5088548064231873, "learning_rate": 3.126877279341761e-06, "loss": 0.0278, "step": 89250 }, { "grad_norm": 0.3284057378768921, "learning_rate": 3.1211243387586964e-06, "loss": 0.027, "step": 89260 }, { "grad_norm": 0.39479202032089233, "learning_rate": 3.115376524778979e-06, "loss": 0.0329, "step": 89270 }, { "grad_norm": 0.1550133377313614, "learning_rate": 3.109633838031195e-06, "loss": 0.0294, "step": 89280 }, { "grad_norm": 0.2191474884748459, "learning_rate": 3.103896279143359e-06, "loss": 0.0277, "step": 89290 }, { "grad_norm": 0.7918566465377808, "learning_rate": 3.098163848742924e-06, "loss": 0.0366, "step": 89300 }, { "grad_norm": 0.4570348262786865, "learning_rate": 3.0924365474567675e-06, "loss": 0.0302, "step": 89310 }, { "grad_norm": 0.4077915847301483, "learning_rate": 3.086714375911237e-06, "loss": 0.0302, "step": 89320 }, { "grad_norm": 0.3936634361743927, "learning_rate": 3.0809973347320775e-06, "loss": 0.0328, "step": 89330 }, { "grad_norm": 0.6023260951042175, "learning_rate": 3.075285424544516e-06, "loss": 0.0341, "step": 89340 }, { "grad_norm": 0.9573518633842468, "learning_rate": 3.069578645973176e-06, "loss": 0.0284, "step": 89350 }, { "grad_norm": 0.5207419991493225, "learning_rate": 3.0638769996421735e-06, "loss": 0.0306, "step": 89360 }, { "grad_norm": 0.9150513410568237, "learning_rate": 3.058180486175005e-06, "loss": 0.033, "step": 89370 }, { "grad_norm": 0.2356780618429184, "learning_rate": 3.052489106194645e-06, "loss": 0.0231, "step": 89380 }, { "grad_norm": 0.3632730543613434, "learning_rate": 3.0468028603234946e-06, "loss": 0.0303, "step": 89390 }, { "grad_norm": 0.49353092908859253, "learning_rate": 3.04112174918339e-06, "loss": 0.022, "step": 89400 }, { "grad_norm": 0.542984664440155, "learning_rate": 3.0354457733955954e-06, "loss": 0.0314, "step": 89410 }, { "grad_norm": 0.3180455267429352, "learning_rate": 3.029774933580859e-06, "loss": 0.022, "step": 89420 }, { "grad_norm": 0.4995102286338806, "learning_rate": 3.0241092303593065e-06, "loss": 0.0248, "step": 89430 }, { "grad_norm": 0.3875858187675476, "learning_rate": 3.018448664350554e-06, "loss": 0.0286, "step": 89440 }, { "grad_norm": 0.2620018720626831, "learning_rate": 3.0127932361736055e-06, "loss": 0.0317, "step": 89450 }, { "grad_norm": 0.812731146812439, "learning_rate": 3.00714294644695e-06, "loss": 0.026, "step": 89460 }, { "grad_norm": 0.6240590810775757, "learning_rate": 3.0014977957884828e-06, "loss": 0.0316, "step": 89470 }, { "grad_norm": 0.39167991280555725, "learning_rate": 2.99585778481557e-06, "loss": 0.0229, "step": 89480 }, { "grad_norm": 0.30972784757614136, "learning_rate": 2.990222914144969e-06, "loss": 0.0321, "step": 89490 }, { "grad_norm": 0.3410223424434662, "learning_rate": 2.984593184392914e-06, "loss": 0.0308, "step": 89500 }, { "grad_norm": 0.2298588901758194, "learning_rate": 2.978968596175069e-06, "loss": 0.0279, "step": 89510 }, { "grad_norm": 0.4436182379722595, "learning_rate": 2.9733491501065137e-06, "loss": 0.0263, "step": 89520 }, { "grad_norm": 0.2270323485136032, "learning_rate": 2.9677348468018005e-06, "loss": 0.0259, "step": 89530 }, { "grad_norm": 0.21003109216690063, "learning_rate": 2.9621256868748726e-06, "loss": 0.0228, "step": 89540 }, { "grad_norm": 0.16650670766830444, "learning_rate": 2.956521670939183e-06, "loss": 0.0288, "step": 89550 }, { "grad_norm": 0.24903912842273712, "learning_rate": 2.950922799607542e-06, "loss": 0.0285, "step": 89560 }, { "grad_norm": 0.6844896078109741, "learning_rate": 2.9453290734922535e-06, "loss": 0.0245, "step": 89570 }, { "grad_norm": 0.548392653465271, "learning_rate": 2.939740493205029e-06, "loss": 0.0393, "step": 89580 }, { "grad_norm": 0.18685585260391235, "learning_rate": 2.9341570593570345e-06, "loss": 0.0251, "step": 89590 }, { "grad_norm": 0.4010055363178253, "learning_rate": 2.928578772558843e-06, "loss": 0.0228, "step": 89600 }, { "grad_norm": 0.4770984351634979, "learning_rate": 2.923005633420528e-06, "loss": 0.0286, "step": 89610 }, { "grad_norm": 0.32501232624053955, "learning_rate": 2.9174376425515237e-06, "loss": 0.0307, "step": 89620 }, { "grad_norm": 1.2159479856491089, "learning_rate": 2.91187480056076e-06, "loss": 0.0319, "step": 89630 }, { "grad_norm": 0.4936597943305969, "learning_rate": 2.9063171080565664e-06, "loss": 0.0261, "step": 89640 }, { "grad_norm": 0.8104962706565857, "learning_rate": 2.900764565646741e-06, "loss": 0.0259, "step": 89650 }, { "grad_norm": 0.35684266686439514, "learning_rate": 2.895217173938475e-06, "loss": 0.0308, "step": 89660 }, { "grad_norm": 0.5094836950302124, "learning_rate": 2.8896749335384445e-06, "loss": 0.0248, "step": 89670 }, { "grad_norm": 0.30957624316215515, "learning_rate": 2.884137845052731e-06, "loss": 0.0371, "step": 89680 }, { "grad_norm": 0.3955601453781128, "learning_rate": 2.8786059090868734e-06, "loss": 0.0259, "step": 89690 }, { "grad_norm": 0.625275731086731, "learning_rate": 2.87307912624582e-06, "loss": 0.0431, "step": 89700 }, { "grad_norm": 0.2631215751171112, "learning_rate": 2.867557497133977e-06, "loss": 0.0307, "step": 89710 }, { "grad_norm": 0.9086897373199463, "learning_rate": 2.862041022355194e-06, "loss": 0.0266, "step": 89720 }, { "grad_norm": 0.6559167504310608, "learning_rate": 2.8565297025127226e-06, "loss": 0.0358, "step": 89730 }, { "grad_norm": 0.37656331062316895, "learning_rate": 2.8510235382092853e-06, "loss": 0.0255, "step": 89740 }, { "grad_norm": 0.4512268900871277, "learning_rate": 2.845522530047029e-06, "loss": 0.0256, "step": 89750 }, { "grad_norm": 0.6205833554267883, "learning_rate": 2.8400266786275387e-06, "loss": 0.0285, "step": 89760 }, { "grad_norm": 1.118811011314392, "learning_rate": 2.834535984551817e-06, "loss": 0.0298, "step": 89770 }, { "grad_norm": 1.0590072870254517, "learning_rate": 2.8290504484203397e-06, "loss": 0.0387, "step": 89780 }, { "grad_norm": 0.4019656479358673, "learning_rate": 2.8235700708329705e-06, "loss": 0.0289, "step": 89790 }, { "grad_norm": 0.8700124621391296, "learning_rate": 2.818094852389058e-06, "loss": 0.0222, "step": 89800 }, { "grad_norm": 0.4068082571029663, "learning_rate": 2.8126247936873516e-06, "loss": 0.0249, "step": 89810 }, { "grad_norm": 0.4271751940250397, "learning_rate": 2.8071598953260614e-06, "loss": 0.0268, "step": 89820 }, { "grad_norm": 0.5802510380744934, "learning_rate": 2.8017001579028035e-06, "loss": 0.0271, "step": 89830 }, { "grad_norm": 0.3963097929954529, "learning_rate": 2.7962455820146617e-06, "loss": 0.0356, "step": 89840 }, { "grad_norm": 0.4392485022544861, "learning_rate": 2.7907961682581253e-06, "loss": 0.0275, "step": 89850 }, { "grad_norm": 0.3041103482246399, "learning_rate": 2.7853519172291453e-06, "loss": 0.0376, "step": 89860 }, { "grad_norm": 0.4208146035671234, "learning_rate": 2.7799128295230904e-06, "loss": 0.029, "step": 89870 }, { "grad_norm": 0.23131319880485535, "learning_rate": 2.774478905734784e-06, "loss": 0.0294, "step": 89880 }, { "grad_norm": 0.5442308783531189, "learning_rate": 2.7690501464584516e-06, "loss": 0.029, "step": 89890 }, { "grad_norm": 1.1061015129089355, "learning_rate": 2.763626552287796e-06, "loss": 0.0289, "step": 89900 }, { "grad_norm": 0.4098883271217346, "learning_rate": 2.7582081238159083e-06, "loss": 0.0284, "step": 89910 }, { "grad_norm": 0.2709091603755951, "learning_rate": 2.7527948616353548e-06, "loss": 0.0341, "step": 89920 }, { "grad_norm": 0.2824145555496216, "learning_rate": 2.7473867663381226e-06, "loss": 0.0238, "step": 89930 }, { "grad_norm": 0.20942559838294983, "learning_rate": 2.7419838385156383e-06, "loss": 0.0252, "step": 89940 }, { "grad_norm": 0.5793076753616333, "learning_rate": 2.7365860787587407e-06, "loss": 0.031, "step": 89950 }, { "grad_norm": 0.8164101243019104, "learning_rate": 2.7311934876577295e-06, "loss": 0.0354, "step": 89960 }, { "grad_norm": 0.3488631248474121, "learning_rate": 2.725806065802339e-06, "loss": 0.0281, "step": 89970 }, { "grad_norm": 0.30861523747444153, "learning_rate": 2.720423813781714e-06, "loss": 0.0335, "step": 89980 }, { "grad_norm": 0.4190726578235626, "learning_rate": 2.715046732184462e-06, "loss": 0.0231, "step": 89990 }, { "grad_norm": 0.5404759645462036, "learning_rate": 2.7096748215986013e-06, "loss": 0.0246, "step": 90000 }, { "grad_norm": 0.7476239204406738, "learning_rate": 2.704308082611612e-06, "loss": 0.0357, "step": 90010 }, { "grad_norm": 0.45957693457603455, "learning_rate": 2.698946515810374e-06, "loss": 0.0259, "step": 90020 }, { "grad_norm": 1.0517072677612305, "learning_rate": 2.6935901217812367e-06, "loss": 0.031, "step": 90030 }, { "grad_norm": 0.48491522669792175, "learning_rate": 2.688238901109952e-06, "loss": 0.0283, "step": 90040 }, { "grad_norm": 0.5589406490325928, "learning_rate": 2.68289285438173e-06, "loss": 0.0335, "step": 90050 }, { "grad_norm": 0.39929112792015076, "learning_rate": 2.6775519821812033e-06, "loss": 0.029, "step": 90060 }, { "grad_norm": 0.3479135036468506, "learning_rate": 2.672216285092449e-06, "loss": 0.0341, "step": 90070 }, { "grad_norm": 0.353275328874588, "learning_rate": 2.666885763698962e-06, "loss": 0.023, "step": 90080 }, { "grad_norm": 0.5068755149841309, "learning_rate": 2.661560418583692e-06, "loss": 0.0368, "step": 90090 }, { "grad_norm": 0.315758615732193, "learning_rate": 2.656240250328995e-06, "loss": 0.0276, "step": 90100 }, { "grad_norm": 0.49633121490478516, "learning_rate": 2.650925259516679e-06, "loss": 0.0381, "step": 90110 }, { "grad_norm": 0.286169171333313, "learning_rate": 2.645615446727995e-06, "loss": 0.0218, "step": 90120 }, { "grad_norm": 1.0538504123687744, "learning_rate": 2.6403108125436115e-06, "loss": 0.0218, "step": 90130 }, { "grad_norm": 0.3450103998184204, "learning_rate": 2.6350113575436264e-06, "loss": 0.0261, "step": 90140 }, { "grad_norm": 0.36009520292282104, "learning_rate": 2.629717082307598e-06, "loss": 0.0243, "step": 90150 }, { "grad_norm": 0.42349982261657715, "learning_rate": 2.6244279874144793e-06, "loss": 0.0261, "step": 90160 }, { "grad_norm": 0.1629074066877365, "learning_rate": 2.6191440734426862e-06, "loss": 0.0223, "step": 90170 }, { "grad_norm": 0.845169186592102, "learning_rate": 2.613865340970062e-06, "loss": 0.0292, "step": 90180 }, { "grad_norm": 0.44315093755722046, "learning_rate": 2.608591790573889e-06, "loss": 0.0309, "step": 90190 }, { "grad_norm": 0.33406299352645874, "learning_rate": 2.603323422830861e-06, "loss": 0.0313, "step": 90200 }, { "grad_norm": 0.36995530128479004, "learning_rate": 2.5980602383171183e-06, "loss": 0.0342, "step": 90210 }, { "grad_norm": 0.46383941173553467, "learning_rate": 2.5928022376082496e-06, "loss": 0.0337, "step": 90220 }, { "grad_norm": 0.4562436044216156, "learning_rate": 2.5875494212792505e-06, "loss": 0.0272, "step": 90230 }, { "grad_norm": 0.3752162754535675, "learning_rate": 2.5823017899045565e-06, "loss": 0.0296, "step": 90240 }, { "grad_norm": 0.4037987291812897, "learning_rate": 2.577059344058047e-06, "loss": 0.0352, "step": 90250 }, { "grad_norm": 0.28286364674568176, "learning_rate": 2.5718220843130415e-06, "loss": 0.025, "step": 90260 }, { "grad_norm": 0.5651338696479797, "learning_rate": 2.5665900112422535e-06, "loss": 0.0299, "step": 90270 }, { "grad_norm": 0.49561047554016113, "learning_rate": 2.561363125417876e-06, "loss": 0.0415, "step": 90280 }, { "grad_norm": 0.2643241286277771, "learning_rate": 2.556141427411496e-06, "loss": 0.019, "step": 90290 }, { "grad_norm": 0.3741667866706848, "learning_rate": 2.550924917794156e-06, "loss": 0.0288, "step": 90300 }, { "grad_norm": 0.3895649015903473, "learning_rate": 2.545713597136329e-06, "loss": 0.0241, "step": 90310 }, { "grad_norm": 0.15475641191005707, "learning_rate": 2.5405074660079298e-06, "loss": 0.0228, "step": 90320 }, { "grad_norm": 0.34706056118011475, "learning_rate": 2.5353065249782647e-06, "loss": 0.028, "step": 90330 }, { "grad_norm": 0.3428986370563507, "learning_rate": 2.530110774616129e-06, "loss": 0.027, "step": 90340 }, { "grad_norm": 0.34201809763908386, "learning_rate": 2.524920215489701e-06, "loss": 0.03, "step": 90350 }, { "grad_norm": 0.23565122485160828, "learning_rate": 2.519734848166627e-06, "loss": 0.0276, "step": 90360 }, { "grad_norm": 0.18468700349330902, "learning_rate": 2.514554673213948e-06, "loss": 0.0282, "step": 90370 }, { "grad_norm": 0.3479185402393341, "learning_rate": 2.5093796911981936e-06, "loss": 0.0221, "step": 90380 }, { "grad_norm": 0.46145936846733093, "learning_rate": 2.5042099026852672e-06, "loss": 0.0364, "step": 90390 }, { "grad_norm": 0.3608398139476776, "learning_rate": 2.499045308240544e-06, "loss": 0.0244, "step": 90400 }, { "grad_norm": 0.20545469224452972, "learning_rate": 2.4938859084288003e-06, "loss": 0.0211, "step": 90410 }, { "grad_norm": 0.287483274936676, "learning_rate": 2.4887317038142734e-06, "loss": 0.0282, "step": 90420 }, { "grad_norm": 0.6439750790596008, "learning_rate": 2.4835826949606124e-06, "loss": 0.025, "step": 90430 }, { "grad_norm": 0.6716420650482178, "learning_rate": 2.4784388824309167e-06, "loss": 0.0307, "step": 90440 }, { "grad_norm": 0.3534201383590698, "learning_rate": 2.473300266787687e-06, "loss": 0.0265, "step": 90450 }, { "grad_norm": 1.6888056993484497, "learning_rate": 2.46816684859289e-06, "loss": 0.0328, "step": 90460 }, { "grad_norm": 0.3768811523914337, "learning_rate": 2.4630386284079043e-06, "loss": 0.0242, "step": 90470 }, { "grad_norm": 0.250296950340271, "learning_rate": 2.4579156067935427e-06, "loss": 0.0261, "step": 90480 }, { "grad_norm": 0.29399916529655457, "learning_rate": 2.452797784310046e-06, "loss": 0.0317, "step": 90490 }, { "grad_norm": 0.6428265571594238, "learning_rate": 2.447685161517105e-06, "loss": 0.0283, "step": 90500 }, { "grad_norm": 0.4666912853717804, "learning_rate": 2.4425777389738224e-06, "loss": 0.0225, "step": 90510 }, { "grad_norm": 0.5143164396286011, "learning_rate": 2.43747551723873e-06, "loss": 0.022, "step": 90520 }, { "grad_norm": 0.2066619098186493, "learning_rate": 2.4323784968698136e-06, "loss": 0.0419, "step": 90530 }, { "grad_norm": 0.6297153830528259, "learning_rate": 2.4272866784244565e-06, "loss": 0.0317, "step": 90540 }, { "grad_norm": 0.38834190368652344, "learning_rate": 2.422200062459518e-06, "loss": 0.0259, "step": 90550 }, { "grad_norm": 1.423296570777893, "learning_rate": 2.41711864953123e-06, "loss": 0.0332, "step": 90560 }, { "grad_norm": 1.0634769201278687, "learning_rate": 2.412042440195322e-06, "loss": 0.0365, "step": 90570 }, { "grad_norm": 0.6793293952941895, "learning_rate": 2.4069714350068993e-06, "loss": 0.0307, "step": 90580 }, { "grad_norm": 0.4260857105255127, "learning_rate": 2.4019056345205294e-06, "loss": 0.0226, "step": 90590 }, { "grad_norm": 0.4837662875652313, "learning_rate": 2.396845039290191e-06, "loss": 0.027, "step": 90600 }, { "grad_norm": 0.3740062713623047, "learning_rate": 2.3917896498693194e-06, "loss": 0.0348, "step": 90610 }, { "grad_norm": 0.504353940486908, "learning_rate": 2.3867394668107335e-06, "loss": 0.0321, "step": 90620 }, { "grad_norm": 0.18786239624023438, "learning_rate": 2.3816944906667527e-06, "loss": 0.0274, "step": 90630 }, { "grad_norm": 0.5666664242744446, "learning_rate": 2.3766547219890634e-06, "loss": 0.026, "step": 90640 }, { "grad_norm": 0.29737991094589233, "learning_rate": 2.371620161328819e-06, "loss": 0.0238, "step": 90650 }, { "grad_norm": 0.5451220870018005, "learning_rate": 2.366590809236574e-06, "loss": 0.03, "step": 90660 }, { "grad_norm": 0.47412624955177307, "learning_rate": 2.3615666662623494e-06, "loss": 0.027, "step": 90670 }, { "grad_norm": 0.5439901947975159, "learning_rate": 2.3565477329555664e-06, "loss": 0.0276, "step": 90680 }, { "grad_norm": 0.8674603700637817, "learning_rate": 2.3515340098650975e-06, "loss": 0.0306, "step": 90690 }, { "grad_norm": 0.8124867677688599, "learning_rate": 2.3465254975392258e-06, "loss": 0.0398, "step": 90700 }, { "grad_norm": 0.28203120827674866, "learning_rate": 2.3415221965256805e-06, "loss": 0.0401, "step": 90710 }, { "grad_norm": 0.4630458652973175, "learning_rate": 2.3365241073716237e-06, "loss": 0.0288, "step": 90720 }, { "grad_norm": 0.315020889043808, "learning_rate": 2.3315312306236183e-06, "loss": 0.0287, "step": 90730 }, { "grad_norm": 0.43792617321014404, "learning_rate": 2.326543566827699e-06, "loss": 0.0338, "step": 90740 }, { "grad_norm": 0.26301541924476624, "learning_rate": 2.3215611165292862e-06, "loss": 0.03, "step": 90750 }, { "grad_norm": 0.3824368715286255, "learning_rate": 2.3165838802732765e-06, "loss": 0.0297, "step": 90760 }, { "grad_norm": 0.24125589430332184, "learning_rate": 2.311611858603957e-06, "loss": 0.0252, "step": 90770 }, { "grad_norm": 0.7730367183685303, "learning_rate": 2.3066450520650763e-06, "loss": 0.0342, "step": 90780 }, { "grad_norm": 0.19337034225463867, "learning_rate": 2.3016834611997773e-06, "loss": 0.0328, "step": 90790 }, { "grad_norm": 0.37922996282577515, "learning_rate": 2.296727086550665e-06, "loss": 0.0321, "step": 90800 }, { "grad_norm": 0.3401414155960083, "learning_rate": 2.2917759286597507e-06, "loss": 0.0357, "step": 90810 }, { "grad_norm": 0.32368043065071106, "learning_rate": 2.2868299880685006e-06, "loss": 0.0292, "step": 90820 }, { "grad_norm": 0.4330683648586273, "learning_rate": 2.2818892653177827e-06, "loss": 0.0255, "step": 90830 }, { "grad_norm": 1.041884422302246, "learning_rate": 2.2769537609479196e-06, "loss": 0.0263, "step": 90840 }, { "grad_norm": 0.5190716981887817, "learning_rate": 2.2720234754986357e-06, "loss": 0.0349, "step": 90850 }, { "grad_norm": 0.17824363708496094, "learning_rate": 2.2670984095091108e-06, "loss": 0.0273, "step": 90860 }, { "grad_norm": 0.28770843148231506, "learning_rate": 2.262178563517936e-06, "loss": 0.0331, "step": 90870 }, { "grad_norm": 0.25709009170532227, "learning_rate": 2.257263938063137e-06, "loss": 0.0325, "step": 90880 }, { "grad_norm": 1.1126788854599, "learning_rate": 2.2523545336821715e-06, "loss": 0.0432, "step": 90890 }, { "grad_norm": 0.7585437893867493, "learning_rate": 2.247450350911939e-06, "loss": 0.0391, "step": 90900 }, { "grad_norm": 0.19463878870010376, "learning_rate": 2.2425513902887328e-06, "loss": 0.0243, "step": 90910 }, { "grad_norm": 0.565909206867218, "learning_rate": 2.2376576523483015e-06, "loss": 0.0233, "step": 90920 }, { "grad_norm": 0.5665329694747925, "learning_rate": 2.232769137625829e-06, "loss": 0.0283, "step": 90930 }, { "grad_norm": 0.8017600774765015, "learning_rate": 2.227885846655903e-06, "loss": 0.0375, "step": 90940 }, { "grad_norm": 0.42231452465057373, "learning_rate": 2.2230077799725523e-06, "loss": 0.0281, "step": 90950 }, { "grad_norm": 0.2758890986442566, "learning_rate": 2.2181349381092396e-06, "loss": 0.0297, "step": 90960 }, { "grad_norm": 0.3200659155845642, "learning_rate": 2.2132673215988552e-06, "loss": 0.0247, "step": 90970 }, { "grad_norm": 0.35291430354118347, "learning_rate": 2.2084049309737065e-06, "loss": 0.0264, "step": 90980 }, { "grad_norm": 1.029447317123413, "learning_rate": 2.2035477667655513e-06, "loss": 0.027, "step": 90990 }, { "grad_norm": 0.5285378694534302, "learning_rate": 2.1986958295055315e-06, "loss": 0.0298, "step": 91000 }, { "grad_norm": 0.7822498679161072, "learning_rate": 2.193849119724284e-06, "loss": 0.0302, "step": 91010 }, { "grad_norm": 0.35940268635749817, "learning_rate": 2.1890076379518175e-06, "loss": 0.0237, "step": 91020 }, { "grad_norm": 0.6202257871627808, "learning_rate": 2.1841713847175925e-06, "loss": 0.0287, "step": 91030 }, { "grad_norm": 0.3893384635448456, "learning_rate": 2.1793403605504968e-06, "loss": 0.0364, "step": 91040 }, { "grad_norm": 0.43411943316459656, "learning_rate": 2.1745145659788413e-06, "loss": 0.0308, "step": 91050 }, { "grad_norm": 0.7956129908561707, "learning_rate": 2.1696940015303645e-06, "loss": 0.0243, "step": 91060 }, { "grad_norm": 0.5591883659362793, "learning_rate": 2.1648786677322397e-06, "loss": 0.037, "step": 91070 }, { "grad_norm": 0.31403982639312744, "learning_rate": 2.1600685651110673e-06, "loss": 0.0321, "step": 91080 }, { "grad_norm": 0.3805180788040161, "learning_rate": 2.1552636941928717e-06, "loss": 0.0265, "step": 91090 }, { "grad_norm": 0.5032737851142883, "learning_rate": 2.1504640555031043e-06, "loss": 0.0217, "step": 91100 }, { "grad_norm": 0.40936973690986633, "learning_rate": 2.1456696495666506e-06, "loss": 0.0234, "step": 91110 }, { "grad_norm": 0.3355347216129303, "learning_rate": 2.1408804769078084e-06, "loss": 0.0355, "step": 91120 }, { "grad_norm": 0.8411651849746704, "learning_rate": 2.136096538050325e-06, "loss": 0.0273, "step": 91130 }, { "grad_norm": 0.604232132434845, "learning_rate": 2.13131783351736e-06, "loss": 0.0353, "step": 91140 }, { "grad_norm": 0.2896086275577545, "learning_rate": 2.126544363831512e-06, "loss": 0.0318, "step": 91150 }, { "grad_norm": 0.2233293503522873, "learning_rate": 2.121776129514791e-06, "loss": 0.0254, "step": 91160 }, { "grad_norm": 0.24136972427368164, "learning_rate": 2.1170131310886523e-06, "loss": 0.0237, "step": 91170 }, { "grad_norm": 0.8608152866363525, "learning_rate": 2.112255369073968e-06, "loss": 0.0342, "step": 91180 }, { "grad_norm": 0.3761659562587738, "learning_rate": 2.1075028439910328e-06, "loss": 0.0246, "step": 91190 }, { "grad_norm": 0.5912306308746338, "learning_rate": 2.1027555563595813e-06, "loss": 0.0305, "step": 91200 }, { "grad_norm": 0.33506399393081665, "learning_rate": 2.0980135066987706e-06, "loss": 0.0243, "step": 91210 }, { "grad_norm": 0.30551719665527344, "learning_rate": 2.093276695527191e-06, "loss": 0.0414, "step": 91220 }, { "grad_norm": 0.34229883551597595, "learning_rate": 2.0885451233628396e-06, "loss": 0.0317, "step": 91230 }, { "grad_norm": 0.6966550350189209, "learning_rate": 2.0838187907231633e-06, "loss": 0.0368, "step": 91240 }, { "grad_norm": 0.28484243154525757, "learning_rate": 2.0790976981250153e-06, "loss": 0.028, "step": 91250 }, { "grad_norm": 0.8030365705490112, "learning_rate": 2.0743818460847e-06, "loss": 0.0294, "step": 91260 }, { "grad_norm": 0.5775840282440186, "learning_rate": 2.069671235117926e-06, "loss": 0.0345, "step": 91270 }, { "grad_norm": 0.4500201642513275, "learning_rate": 2.0649658657398484e-06, "loss": 0.027, "step": 91280 }, { "grad_norm": 0.7068238258361816, "learning_rate": 2.060265738465034e-06, "loss": 0.0261, "step": 91290 }, { "grad_norm": 0.45753395557403564, "learning_rate": 2.055570853807487e-06, "loss": 0.02, "step": 91300 }, { "grad_norm": 0.46120432019233704, "learning_rate": 2.0508812122806196e-06, "loss": 0.0264, "step": 91310 }, { "grad_norm": 0.3954220712184906, "learning_rate": 2.046196814397294e-06, "loss": 0.0257, "step": 91320 }, { "grad_norm": 0.5147157311439514, "learning_rate": 2.0415176606697896e-06, "loss": 0.0293, "step": 91330 }, { "grad_norm": 0.40573394298553467, "learning_rate": 2.0368437516098126e-06, "loss": 0.0282, "step": 91340 }, { "grad_norm": 0.573101282119751, "learning_rate": 2.032175087728483e-06, "loss": 0.0294, "step": 91350 }, { "grad_norm": 0.44766372442245483, "learning_rate": 2.027511669536375e-06, "loss": 0.0319, "step": 91360 }, { "grad_norm": 0.3043466806411743, "learning_rate": 2.0228534975434643e-06, "loss": 0.0239, "step": 91370 }, { "grad_norm": 0.32467424869537354, "learning_rate": 2.0182005722591548e-06, "loss": 0.0297, "step": 91380 }, { "grad_norm": 0.43098995089530945, "learning_rate": 2.0135528941922944e-06, "loss": 0.0275, "step": 91390 }, { "grad_norm": 0.31146925687789917, "learning_rate": 2.008910463851149e-06, "loss": 0.0239, "step": 91400 }, { "grad_norm": 0.3849792182445526, "learning_rate": 2.0042732817433963e-06, "loss": 0.0274, "step": 91410 }, { "grad_norm": 0.4740139842033386, "learning_rate": 1.999641348376158e-06, "loss": 0.022, "step": 91420 }, { "grad_norm": 0.6784924864768982, "learning_rate": 1.9950146642559787e-06, "loss": 0.0341, "step": 91430 }, { "grad_norm": 1.0982956886291504, "learning_rate": 1.990393229888815e-06, "loss": 0.0302, "step": 91440 }, { "grad_norm": 0.3396476209163666, "learning_rate": 1.9857770457800684e-06, "loss": 0.0273, "step": 91450 }, { "grad_norm": 0.298735648393631, "learning_rate": 1.9811661124345517e-06, "loss": 0.0258, "step": 91460 }, { "grad_norm": 0.3925785422325134, "learning_rate": 1.976560430356522e-06, "loss": 0.0238, "step": 91470 }, { "grad_norm": 0.6178544759750366, "learning_rate": 1.9719600000496383e-06, "loss": 0.034, "step": 91480 }, { "grad_norm": 0.4405500292778015, "learning_rate": 1.967364822017004e-06, "loss": 0.0325, "step": 91490 }, { "grad_norm": 0.23370353877544403, "learning_rate": 1.962774896761127e-06, "loss": 0.025, "step": 91500 }, { "grad_norm": 0.2894018888473511, "learning_rate": 1.958190224783968e-06, "loss": 0.0212, "step": 91510 }, { "grad_norm": 0.3173631727695465, "learning_rate": 1.9536108065868975e-06, "loss": 0.0294, "step": 91520 }, { "grad_norm": 0.43732118606567383, "learning_rate": 1.949036642670715e-06, "loss": 0.0339, "step": 91530 }, { "grad_norm": 0.32105910778045654, "learning_rate": 1.9444677335356377e-06, "loss": 0.0307, "step": 91540 }, { "grad_norm": 0.7443557381629944, "learning_rate": 1.939904079681326e-06, "loss": 0.0389, "step": 91550 }, { "grad_norm": 0.22411690652370453, "learning_rate": 1.935345681606837e-06, "loss": 0.0247, "step": 91560 }, { "grad_norm": 0.4910981357097626, "learning_rate": 1.930792539810694e-06, "loss": 0.0246, "step": 91570 }, { "grad_norm": 0.32747331261634827, "learning_rate": 1.9262446547907877e-06, "loss": 0.028, "step": 91580 }, { "grad_norm": 0.3646402955055237, "learning_rate": 1.921702027044503e-06, "loss": 0.0338, "step": 91590 }, { "grad_norm": 0.47275665402412415, "learning_rate": 1.917164657068593e-06, "loss": 0.033, "step": 91600 }, { "grad_norm": 0.8356645703315735, "learning_rate": 1.912632545359272e-06, "loss": 0.0303, "step": 91610 }, { "grad_norm": 0.17036959528923035, "learning_rate": 1.9081056924121488e-06, "loss": 0.025, "step": 91620 }, { "grad_norm": 0.2293647825717926, "learning_rate": 1.9035840987222887e-06, "loss": 0.0279, "step": 91630 }, { "grad_norm": 0.18466271460056305, "learning_rate": 1.8990677647841515e-06, "loss": 0.0296, "step": 91640 }, { "grad_norm": 0.41787317395210266, "learning_rate": 1.8945566910916534e-06, "loss": 0.0329, "step": 91650 }, { "grad_norm": 0.5569785237312317, "learning_rate": 1.8900508781381054e-06, "loss": 0.0259, "step": 91660 }, { "grad_norm": 0.22953397035598755, "learning_rate": 1.8855503264162633e-06, "loss": 0.0192, "step": 91670 }, { "grad_norm": 0.439572274684906, "learning_rate": 1.8810550364182998e-06, "loss": 0.0282, "step": 91680 }, { "grad_norm": 0.6858137845993042, "learning_rate": 1.8765650086358112e-06, "loss": 0.0279, "step": 91690 }, { "grad_norm": 0.740278959274292, "learning_rate": 1.8720802435598151e-06, "loss": 0.0334, "step": 91700 }, { "grad_norm": 0.32341623306274414, "learning_rate": 1.8676007416807694e-06, "loss": 0.0311, "step": 91710 }, { "grad_norm": 0.691636860370636, "learning_rate": 1.8631265034885437e-06, "loss": 0.0359, "step": 91720 }, { "grad_norm": 0.2830236554145813, "learning_rate": 1.858657529472424e-06, "loss": 0.0232, "step": 91730 }, { "grad_norm": 0.34733057022094727, "learning_rate": 1.8541938201211473e-06, "loss": 0.0247, "step": 91740 }, { "grad_norm": 0.6053295135498047, "learning_rate": 1.8497353759228398e-06, "loss": 0.023, "step": 91750 }, { "grad_norm": 0.39465823769569397, "learning_rate": 1.8452821973650835e-06, "loss": 0.0234, "step": 91760 }, { "grad_norm": 0.19266153872013092, "learning_rate": 1.840834284934856e-06, "loss": 0.0265, "step": 91770 }, { "grad_norm": 0.2926989793777466, "learning_rate": 1.8363916391185954e-06, "loss": 0.0288, "step": 91780 }, { "grad_norm": 0.2859906256198883, "learning_rate": 1.8319542604021245e-06, "loss": 0.0356, "step": 91790 }, { "grad_norm": 0.41159921884536743, "learning_rate": 1.8275221492707273e-06, "loss": 0.0244, "step": 91800 }, { "grad_norm": 0.32515066862106323, "learning_rate": 1.8230953062090717e-06, "loss": 0.0378, "step": 91810 }, { "grad_norm": 0.3391071856021881, "learning_rate": 1.8186737317012926e-06, "loss": 0.021, "step": 91820 }, { "grad_norm": 0.3610917925834656, "learning_rate": 1.8142574262308977e-06, "loss": 0.0308, "step": 91830 }, { "grad_norm": 0.3603079617023468, "learning_rate": 1.8098463902808727e-06, "loss": 0.0312, "step": 91840 }, { "grad_norm": 0.6812006235122681, "learning_rate": 1.805440624333593e-06, "loss": 0.0212, "step": 91850 }, { "grad_norm": 1.3715842962265015, "learning_rate": 1.8010401288708734e-06, "loss": 0.0331, "step": 91860 }, { "grad_norm": 0.3603564500808716, "learning_rate": 1.7966449043739343e-06, "loss": 0.0322, "step": 91870 }, { "grad_norm": 0.4961816668510437, "learning_rate": 1.79225495132343e-06, "loss": 0.0249, "step": 91880 }, { "grad_norm": 0.4471239149570465, "learning_rate": 1.7878702701994487e-06, "loss": 0.0334, "step": 91890 }, { "grad_norm": 1.5895884037017822, "learning_rate": 1.783490861481496e-06, "loss": 0.0217, "step": 91900 }, { "grad_norm": 0.2499620020389557, "learning_rate": 1.7791167256484832e-06, "loss": 0.0244, "step": 91910 }, { "grad_norm": 0.6584236025810242, "learning_rate": 1.774747863178766e-06, "loss": 0.039, "step": 91920 }, { "grad_norm": 0.30762842297554016, "learning_rate": 1.7703842745501186e-06, "loss": 0.0261, "step": 91930 }, { "grad_norm": 0.4486810266971588, "learning_rate": 1.7660259602397366e-06, "loss": 0.0387, "step": 91940 }, { "grad_norm": 0.3562350273132324, "learning_rate": 1.7616729207242388e-06, "loss": 0.0284, "step": 91950 }, { "grad_norm": 0.3159610629081726, "learning_rate": 1.7573251564796556e-06, "loss": 0.037, "step": 91960 }, { "grad_norm": 0.5201898813247681, "learning_rate": 1.7529826679814676e-06, "loss": 0.0333, "step": 91970 }, { "grad_norm": 0.4526881277561188, "learning_rate": 1.7486454557045562e-06, "loss": 0.0343, "step": 91980 }, { "grad_norm": 0.2926460802555084, "learning_rate": 1.7443135201232364e-06, "loss": 0.024, "step": 91990 }, { "grad_norm": 0.506901741027832, "learning_rate": 1.7399868617112347e-06, "loss": 0.0348, "step": 92000 }, { "grad_norm": 0.7957027554512024, "learning_rate": 1.7356654809417171e-06, "loss": 0.026, "step": 92010 }, { "grad_norm": 0.31628820300102234, "learning_rate": 1.7313493782872448e-06, "loss": 0.0275, "step": 92020 }, { "grad_norm": 0.6340765953063965, "learning_rate": 1.7270385542198453e-06, "loss": 0.0258, "step": 92030 }, { "grad_norm": 0.40485143661499023, "learning_rate": 1.7227330092109307e-06, "loss": 0.0263, "step": 92040 }, { "grad_norm": 0.20543403923511505, "learning_rate": 1.718432743731352e-06, "loss": 0.0242, "step": 92050 }, { "grad_norm": 0.3814133107662201, "learning_rate": 1.7141377582513718e-06, "loss": 0.0273, "step": 92060 }, { "grad_norm": 0.3857421576976776, "learning_rate": 1.709848053240698e-06, "loss": 0.0276, "step": 92070 }, { "grad_norm": 0.41529685258865356, "learning_rate": 1.7055636291684328e-06, "loss": 0.0183, "step": 92080 }, { "grad_norm": 0.31461358070373535, "learning_rate": 1.7012844865031185e-06, "loss": 0.0274, "step": 92090 }, { "grad_norm": 0.3117507994174957, "learning_rate": 1.6970106257127138e-06, "loss": 0.0344, "step": 92100 }, { "grad_norm": 0.3929849863052368, "learning_rate": 1.6927420472646117e-06, "loss": 0.0217, "step": 92110 }, { "grad_norm": 0.8645080924034119, "learning_rate": 1.6884787516256106e-06, "loss": 0.0366, "step": 92120 }, { "grad_norm": 0.4500882923603058, "learning_rate": 1.6842207392619325e-06, "loss": 0.0329, "step": 92130 }, { "grad_norm": 0.35495051741600037, "learning_rate": 1.679968010639238e-06, "loss": 0.0315, "step": 92140 }, { "grad_norm": 0.4263829290866852, "learning_rate": 1.6757205662225829e-06, "loss": 0.0365, "step": 92150 }, { "grad_norm": 0.5002217888832092, "learning_rate": 1.6714784064764844e-06, "loss": 0.0296, "step": 92160 }, { "grad_norm": 0.29544833302497864, "learning_rate": 1.6672415318648383e-06, "loss": 0.0282, "step": 92170 }, { "grad_norm": 0.5767799615859985, "learning_rate": 1.6630099428510016e-06, "loss": 0.0369, "step": 92180 }, { "grad_norm": 0.31314289569854736, "learning_rate": 1.6587836398977153e-06, "loss": 0.0358, "step": 92190 }, { "grad_norm": 0.23992934823036194, "learning_rate": 1.6545626234671762e-06, "loss": 0.0281, "step": 92200 }, { "grad_norm": 0.4125428795814514, "learning_rate": 1.650346894020971e-06, "loss": 0.0285, "step": 92210 }, { "grad_norm": 0.2101561725139618, "learning_rate": 1.6461364520201472e-06, "loss": 0.029, "step": 92220 }, { "grad_norm": 0.5855650305747986, "learning_rate": 1.6419312979251366e-06, "loss": 0.0392, "step": 92230 }, { "grad_norm": 0.2855778932571411, "learning_rate": 1.6377314321958214e-06, "loss": 0.0328, "step": 92240 }, { "grad_norm": 0.4829443395137787, "learning_rate": 1.6335368552914787e-06, "loss": 0.0312, "step": 92250 }, { "grad_norm": 0.41533127427101135, "learning_rate": 1.6293475676708358e-06, "loss": 0.0275, "step": 92260 }, { "grad_norm": 1.0182534456253052, "learning_rate": 1.6251635697920097e-06, "loss": 0.0333, "step": 92270 }, { "grad_norm": 0.5627737641334534, "learning_rate": 1.6209848621125624e-06, "loss": 0.0352, "step": 92280 }, { "grad_norm": 0.2878866493701935, "learning_rate": 1.6168114450894778e-06, "loss": 0.0317, "step": 92290 }, { "grad_norm": 0.5017517805099487, "learning_rate": 1.612643319179158e-06, "loss": 0.0235, "step": 92300 }, { "grad_norm": 0.6091774702072144, "learning_rate": 1.6084804848374046e-06, "loss": 0.0302, "step": 92310 }, { "grad_norm": 0.3644339144229889, "learning_rate": 1.6043229425194817e-06, "loss": 0.0262, "step": 92320 }, { "grad_norm": 0.31879961490631104, "learning_rate": 1.6001706926800308e-06, "loss": 0.0232, "step": 92330 }, { "grad_norm": 0.4888949990272522, "learning_rate": 1.5960237357731444e-06, "loss": 0.0272, "step": 92340 }, { "grad_norm": 0.5336184501647949, "learning_rate": 1.5918820722523264e-06, "loss": 0.0335, "step": 92350 }, { "grad_norm": 0.28653275966644287, "learning_rate": 1.587745702570509e-06, "loss": 0.0356, "step": 92360 }, { "grad_norm": 0.3871764838695526, "learning_rate": 1.5836146271800356e-06, "loss": 0.0268, "step": 92370 }, { "grad_norm": 0.3296046257019043, "learning_rate": 1.5794888465326728e-06, "loss": 0.0266, "step": 92380 }, { "grad_norm": 0.2060074806213379, "learning_rate": 1.5753683610796155e-06, "loss": 0.0319, "step": 92390 }, { "grad_norm": 0.6102744340896606, "learning_rate": 1.5712531712714528e-06, "loss": 0.0334, "step": 92400 }, { "grad_norm": 0.2689288258552551, "learning_rate": 1.5671432775582474e-06, "loss": 0.0347, "step": 92410 }, { "grad_norm": 0.5291520357131958, "learning_rate": 1.563038680389428e-06, "loss": 0.0301, "step": 92420 }, { "grad_norm": 0.4498847424983978, "learning_rate": 1.5589393802138808e-06, "loss": 0.0268, "step": 92430 }, { "grad_norm": 0.28804507851600647, "learning_rate": 1.554845377479891e-06, "loss": 0.0305, "step": 92440 }, { "grad_norm": 0.5044127106666565, "learning_rate": 1.550756672635173e-06, "loss": 0.035, "step": 92450 }, { "grad_norm": 0.5457752346992493, "learning_rate": 1.5466732661268635e-06, "loss": 0.0264, "step": 92460 }, { "grad_norm": 0.5715937614440918, "learning_rate": 1.5425951584015163e-06, "loss": 0.0238, "step": 92470 }, { "grad_norm": 0.2246595025062561, "learning_rate": 1.5385223499051026e-06, "loss": 0.0266, "step": 92480 }, { "grad_norm": 0.5598774552345276, "learning_rate": 1.5344548410830328e-06, "loss": 0.0242, "step": 92490 }, { "grad_norm": 0.5497739315032959, "learning_rate": 1.5303926323801121e-06, "loss": 0.0267, "step": 92500 }, { "grad_norm": 0.30684223771095276, "learning_rate": 1.5263357242405795e-06, "loss": 0.0208, "step": 92510 }, { "grad_norm": 0.37794700264930725, "learning_rate": 1.5222841171080915e-06, "loss": 0.031, "step": 92520 }, { "grad_norm": 0.4550018310546875, "learning_rate": 1.5182378114257268e-06, "loss": 0.0293, "step": 92530 }, { "grad_norm": 0.2915138900279999, "learning_rate": 1.5141968076359869e-06, "loss": 0.0272, "step": 92540 }, { "grad_norm": 0.7324023246765137, "learning_rate": 1.5101611061807908e-06, "loss": 0.0286, "step": 92550 }, { "grad_norm": 0.6079736351966858, "learning_rate": 1.506130707501463e-06, "loss": 0.0233, "step": 92560 }, { "grad_norm": 0.46170786023139954, "learning_rate": 1.5021056120387734e-06, "loss": 0.036, "step": 92570 }, { "grad_norm": 0.4494604766368866, "learning_rate": 1.4980858202329084e-06, "loss": 0.0264, "step": 92580 }, { "grad_norm": 0.26517313718795776, "learning_rate": 1.4940713325234445e-06, "loss": 0.0264, "step": 92590 }, { "grad_norm": 0.22131773829460144, "learning_rate": 1.4900621493494137e-06, "loss": 0.0215, "step": 92600 }, { "grad_norm": 0.8040284514427185, "learning_rate": 1.4860582711492544e-06, "loss": 0.0388, "step": 92610 }, { "grad_norm": 0.482711523771286, "learning_rate": 1.4820596983608325e-06, "loss": 0.0294, "step": 92620 }, { "grad_norm": 0.2834545075893402, "learning_rate": 1.4780664314214044e-06, "loss": 0.0234, "step": 92630 }, { "grad_norm": 0.2851465940475464, "learning_rate": 1.474078470767687e-06, "loss": 0.0354, "step": 92640 }, { "grad_norm": 0.3060874342918396, "learning_rate": 1.470095816835787e-06, "loss": 0.0246, "step": 92650 }, { "grad_norm": 0.5030096173286438, "learning_rate": 1.4661184700612395e-06, "loss": 0.0321, "step": 92660 }, { "grad_norm": 0.3001086711883545, "learning_rate": 1.4621464308790133e-06, "loss": 0.027, "step": 92670 }, { "grad_norm": 0.5067752003669739, "learning_rate": 1.4581796997234775e-06, "loss": 0.0337, "step": 92680 }, { "grad_norm": 0.19609957933425903, "learning_rate": 1.4542182770284297e-06, "loss": 0.025, "step": 92690 }, { "grad_norm": 0.294752836227417, "learning_rate": 1.450262163227084e-06, "loss": 0.0243, "step": 92700 }, { "grad_norm": 0.5848268270492554, "learning_rate": 1.4463113587520671e-06, "loss": 0.0393, "step": 92710 }, { "grad_norm": 0.28056827187538147, "learning_rate": 1.4423658640354443e-06, "loss": 0.0273, "step": 92720 }, { "grad_norm": 0.4797437787055969, "learning_rate": 1.438425679508687e-06, "loss": 0.0268, "step": 92730 }, { "grad_norm": 0.4978063404560089, "learning_rate": 1.4344908056026895e-06, "loss": 0.0268, "step": 92740 }, { "grad_norm": 0.23428644239902496, "learning_rate": 1.4305612427477521e-06, "loss": 0.0214, "step": 92750 }, { "grad_norm": 0.5417212843894958, "learning_rate": 1.42663699137362e-06, "loss": 0.0215, "step": 92760 }, { "grad_norm": 0.6427485942840576, "learning_rate": 1.4227180519094386e-06, "loss": 0.0394, "step": 92770 }, { "grad_norm": 0.4435737729072571, "learning_rate": 1.4188044247837706e-06, "loss": 0.0239, "step": 92780 }, { "grad_norm": 0.32844147086143494, "learning_rate": 1.4148961104246128e-06, "loss": 0.0244, "step": 92790 }, { "grad_norm": 0.9084274172782898, "learning_rate": 1.4109931092593731e-06, "loss": 0.024, "step": 92800 }, { "grad_norm": 0.8571322560310364, "learning_rate": 1.4070954217148658e-06, "loss": 0.0319, "step": 92810 }, { "grad_norm": 0.633050262928009, "learning_rate": 1.4032030482173497e-06, "loss": 0.0327, "step": 92820 }, { "grad_norm": 0.6627122759819031, "learning_rate": 1.3993159891924901e-06, "loss": 0.0316, "step": 92830 }, { "grad_norm": 0.7267526984214783, "learning_rate": 1.3954342450653524e-06, "loss": 0.0355, "step": 92840 }, { "grad_norm": 0.45999523997306824, "learning_rate": 1.3915578162604582e-06, "loss": 0.0264, "step": 92850 }, { "grad_norm": 0.3602335751056671, "learning_rate": 1.3876867032017126e-06, "loss": 0.0333, "step": 92860 }, { "grad_norm": 1.0352635383605957, "learning_rate": 1.383820906312472e-06, "loss": 0.0297, "step": 92870 }, { "grad_norm": 0.405151903629303, "learning_rate": 1.3799604260154752e-06, "loss": 0.0275, "step": 92880 }, { "grad_norm": 0.46475252509117126, "learning_rate": 1.376105262732913e-06, "loss": 0.0378, "step": 92890 }, { "grad_norm": 0.48817941546440125, "learning_rate": 1.37225541688637e-06, "loss": 0.0354, "step": 92900 }, { "grad_norm": 0.28638118505477905, "learning_rate": 1.3684108888968594e-06, "loss": 0.0338, "step": 92910 }, { "grad_norm": 0.5059488415718079, "learning_rate": 1.364571679184823e-06, "loss": 0.0295, "step": 92920 }, { "grad_norm": 0.3901987671852112, "learning_rate": 1.3607377881701077e-06, "loss": 0.0402, "step": 92930 }, { "grad_norm": 0.375582218170166, "learning_rate": 1.3569092162719787e-06, "loss": 0.0333, "step": 92940 }, { "grad_norm": 0.37950220704078674, "learning_rate": 1.353085963909123e-06, "loss": 0.0261, "step": 92950 }, { "grad_norm": 0.31468266248703003, "learning_rate": 1.3492680314996454e-06, "loss": 0.0294, "step": 92960 }, { "grad_norm": 1.15804922580719, "learning_rate": 1.345455419461078e-06, "loss": 0.0359, "step": 92970 }, { "grad_norm": 0.1862914115190506, "learning_rate": 1.3416481282103431e-06, "loss": 0.0276, "step": 92980 }, { "grad_norm": 0.3518690764904022, "learning_rate": 1.337846158163819e-06, "loss": 0.0362, "step": 92990 }, { "grad_norm": 0.27150285243988037, "learning_rate": 1.334049509737273e-06, "loss": 0.0324, "step": 93000 }, { "grad_norm": 0.3347766697406769, "learning_rate": 1.3302581833459116e-06, "loss": 0.0317, "step": 93010 }, { "grad_norm": 0.4918016791343689, "learning_rate": 1.3264721794043366e-06, "loss": 0.0305, "step": 93020 }, { "grad_norm": 0.5907204747200012, "learning_rate": 1.3226914983265836e-06, "loss": 0.0438, "step": 93030 }, { "grad_norm": 0.5272192358970642, "learning_rate": 1.3189161405260997e-06, "loss": 0.0337, "step": 93040 }, { "grad_norm": 0.4121353328227997, "learning_rate": 1.3151461064157656e-06, "loss": 0.024, "step": 93050 }, { "grad_norm": 0.7526387572288513, "learning_rate": 1.3113813964078459e-06, "loss": 0.0336, "step": 93060 }, { "grad_norm": 0.26572540402412415, "learning_rate": 1.3076220109140614e-06, "loss": 0.0296, "step": 93070 }, { "grad_norm": 0.3910471498966217, "learning_rate": 1.3038679503455275e-06, "loss": 0.0325, "step": 93080 }, { "grad_norm": 0.3172161281108856, "learning_rate": 1.3001192151127773e-06, "loss": 0.0252, "step": 93090 }, { "grad_norm": 0.43785107135772705, "learning_rate": 1.2963758056257657e-06, "loss": 0.0273, "step": 93100 }, { "grad_norm": 0.24762609601020813, "learning_rate": 1.2926377222938768e-06, "loss": 0.0198, "step": 93110 }, { "grad_norm": 0.674121081829071, "learning_rate": 1.2889049655259e-06, "loss": 0.0348, "step": 93120 }, { "grad_norm": 0.39448514580726624, "learning_rate": 1.2851775357300312e-06, "loss": 0.039, "step": 93130 }, { "grad_norm": 0.42274460196495056, "learning_rate": 1.281455433313916e-06, "loss": 0.0295, "step": 93140 }, { "grad_norm": 0.5264642238616943, "learning_rate": 1.2777386586845796e-06, "loss": 0.0291, "step": 93150 }, { "grad_norm": 0.6372518539428711, "learning_rate": 1.2740272122484964e-06, "loss": 0.0353, "step": 93160 }, { "grad_norm": 0.3895500600337982, "learning_rate": 1.2703210944115308e-06, "loss": 0.0254, "step": 93170 }, { "grad_norm": 0.464165061712265, "learning_rate": 1.2666203055789916e-06, "loss": 0.0226, "step": 93180 }, { "grad_norm": 0.3730927109718323, "learning_rate": 1.2629248461555832e-06, "loss": 0.0272, "step": 93190 }, { "grad_norm": 0.48682162165641785, "learning_rate": 1.2592347165454431e-06, "loss": 0.0258, "step": 93200 }, { "grad_norm": 0.3198140561580658, "learning_rate": 1.2555499171521102e-06, "loss": 0.0407, "step": 93210 }, { "grad_norm": 0.24105021357536316, "learning_rate": 1.251870448378556e-06, "loss": 0.0241, "step": 93220 }, { "grad_norm": 0.660546600818634, "learning_rate": 1.2481963106271478e-06, "loss": 0.0326, "step": 93230 }, { "grad_norm": 0.5960095524787903, "learning_rate": 1.2445275042997084e-06, "loss": 0.0373, "step": 93240 }, { "grad_norm": 0.38205161690711975, "learning_rate": 1.2408640297974283e-06, "loss": 0.0328, "step": 93250 }, { "grad_norm": 0.3864727318286896, "learning_rate": 1.2372058875209535e-06, "loss": 0.0257, "step": 93260 }, { "grad_norm": 0.502136766910553, "learning_rate": 1.2335530778703252e-06, "loss": 0.0253, "step": 93270 }, { "grad_norm": 0.6118265986442566, "learning_rate": 1.2299056012450182e-06, "loss": 0.0303, "step": 93280 }, { "grad_norm": 0.6947848796844482, "learning_rate": 1.2262634580439026e-06, "loss": 0.0335, "step": 93290 }, { "grad_norm": 0.513526976108551, "learning_rate": 1.2226266486652926e-06, "loss": 0.0384, "step": 93300 }, { "grad_norm": 0.44589436054229736, "learning_rate": 1.2189951735068927e-06, "loss": 0.0316, "step": 93310 }, { "grad_norm": 0.8373662233352661, "learning_rate": 1.2153690329658407e-06, "loss": 0.0265, "step": 93320 }, { "grad_norm": 0.4179782271385193, "learning_rate": 1.2117482274386859e-06, "loss": 0.036, "step": 93330 }, { "grad_norm": 0.3697015345096588, "learning_rate": 1.2081327573213953e-06, "loss": 0.0257, "step": 93340 }, { "grad_norm": 0.6615483164787292, "learning_rate": 1.2045226230093465e-06, "loss": 0.0319, "step": 93350 }, { "grad_norm": 0.3971458375453949, "learning_rate": 1.2009178248973296e-06, "loss": 0.0323, "step": 93360 }, { "grad_norm": 0.45302900671958923, "learning_rate": 1.1973183633795849e-06, "loss": 0.0256, "step": 93370 }, { "grad_norm": 0.13886874914169312, "learning_rate": 1.193724238849725e-06, "loss": 0.0322, "step": 93380 }, { "grad_norm": 0.38081616163253784, "learning_rate": 1.190135451700808e-06, "loss": 0.0244, "step": 93390 }, { "grad_norm": 0.7397712469100952, "learning_rate": 1.186552002325292e-06, "loss": 0.0336, "step": 93400 }, { "grad_norm": 0.44669684767723083, "learning_rate": 1.1829738911150578e-06, "loss": 0.0231, "step": 93410 }, { "grad_norm": 0.40750938653945923, "learning_rate": 1.1794011184613985e-06, "loss": 0.0267, "step": 93420 }, { "grad_norm": 0.730907678604126, "learning_rate": 1.17583368475504e-06, "loss": 0.0387, "step": 93430 }, { "grad_norm": 1.1528419256210327, "learning_rate": 1.1722715903860982e-06, "loss": 0.0262, "step": 93440 }, { "grad_norm": 0.41020745038986206, "learning_rate": 1.168714835744128e-06, "loss": 0.0318, "step": 93450 }, { "grad_norm": 0.22791722416877747, "learning_rate": 1.165163421218085e-06, "loss": 0.0274, "step": 93460 }, { "grad_norm": 0.5432770848274231, "learning_rate": 1.1616173471963477e-06, "loss": 0.0398, "step": 93470 }, { "grad_norm": 0.39132583141326904, "learning_rate": 1.1580766140667109e-06, "loss": 0.0219, "step": 93480 }, { "grad_norm": 0.288349986076355, "learning_rate": 1.1545412222163764e-06, "loss": 0.0297, "step": 93490 }, { "grad_norm": 0.31680309772491455, "learning_rate": 1.1510111720319795e-06, "loss": 0.0308, "step": 93500 }, { "grad_norm": 0.28162848949432373, "learning_rate": 1.1474864638995663e-06, "loss": 0.0281, "step": 93510 }, { "grad_norm": 0.3927813470363617, "learning_rate": 1.1439670982045735e-06, "loss": 0.0342, "step": 93520 }, { "grad_norm": 0.2775668501853943, "learning_rate": 1.1404530753318877e-06, "loss": 0.0338, "step": 93530 }, { "grad_norm": 0.287570983171463, "learning_rate": 1.136944395665801e-06, "loss": 0.03, "step": 93540 }, { "grad_norm": 0.5418385863304138, "learning_rate": 1.1334410595900014e-06, "loss": 0.0316, "step": 93550 }, { "grad_norm": 0.3483165502548218, "learning_rate": 1.129943067487621e-06, "loss": 0.0344, "step": 93560 }, { "grad_norm": 0.46049103140830994, "learning_rate": 1.126450419741193e-06, "loss": 0.022, "step": 93570 }, { "grad_norm": 0.3971078097820282, "learning_rate": 1.1229631167326671e-06, "loss": 0.0244, "step": 93580 }, { "grad_norm": 0.517203688621521, "learning_rate": 1.1194811588434052e-06, "loss": 0.0293, "step": 93590 }, { "grad_norm": 0.5080922842025757, "learning_rate": 1.1160045464542023e-06, "loss": 0.028, "step": 93600 }, { "grad_norm": 0.3798581659793854, "learning_rate": 1.1125332799452325e-06, "loss": 0.0284, "step": 93610 }, { "grad_norm": 0.5881783962249756, "learning_rate": 1.109067359696131e-06, "loss": 0.0272, "step": 93620 }, { "grad_norm": 0.2708470821380615, "learning_rate": 1.1056067860859166e-06, "loss": 0.0237, "step": 93630 }, { "grad_norm": 0.5993431806564331, "learning_rate": 1.102151559493031e-06, "loss": 0.0267, "step": 93640 }, { "grad_norm": 0.4537312388420105, "learning_rate": 1.0987016802953332e-06, "loss": 0.03, "step": 93650 }, { "grad_norm": 0.7405815720558167, "learning_rate": 1.0952571488701048e-06, "loss": 0.026, "step": 93660 }, { "grad_norm": 0.5076814293861389, "learning_rate": 1.0918179655940163e-06, "loss": 0.0315, "step": 93670 }, { "grad_norm": 0.7511070370674133, "learning_rate": 1.088384130843184e-06, "loss": 0.029, "step": 93680 }, { "grad_norm": 0.6040933728218079, "learning_rate": 1.0849556449931296e-06, "loss": 0.035, "step": 93690 }, { "grad_norm": 0.3008043169975281, "learning_rate": 1.0815325084187866e-06, "loss": 0.0252, "step": 93700 }, { "grad_norm": 0.42894065380096436, "learning_rate": 1.0781147214944943e-06, "loss": 0.0235, "step": 93710 }, { "grad_norm": 0.3500257134437561, "learning_rate": 1.0747022845940313e-06, "loss": 0.0244, "step": 93720 }, { "grad_norm": 0.2940085828304291, "learning_rate": 1.0712951980905605e-06, "loss": 0.024, "step": 93730 }, { "grad_norm": 0.39297163486480713, "learning_rate": 1.0678934623566839e-06, "loss": 0.0294, "step": 93740 }, { "grad_norm": 0.5221898555755615, "learning_rate": 1.0644970777644091e-06, "loss": 0.0271, "step": 93750 }, { "grad_norm": 0.7073183655738831, "learning_rate": 1.0611060446851673e-06, "loss": 0.0286, "step": 93760 }, { "grad_norm": 0.4048362970352173, "learning_rate": 1.0577203634897836e-06, "loss": 0.0273, "step": 93770 }, { "grad_norm": 0.48556777834892273, "learning_rate": 1.0543400345485178e-06, "loss": 0.0268, "step": 93780 }, { "grad_norm": 1.1351481676101685, "learning_rate": 1.0509650582310404e-06, "loss": 0.0312, "step": 93790 }, { "grad_norm": 0.4782143235206604, "learning_rate": 1.0475954349064287e-06, "loss": 0.0336, "step": 93800 }, { "grad_norm": 0.161302387714386, "learning_rate": 1.0442311649431825e-06, "loss": 0.0257, "step": 93810 }, { "grad_norm": 0.32896149158477783, "learning_rate": 1.0408722487092071e-06, "loss": 0.0391, "step": 93820 }, { "grad_norm": 0.41547343134880066, "learning_rate": 1.0375186865718424e-06, "loss": 0.0202, "step": 93830 }, { "grad_norm": 0.3842616677284241, "learning_rate": 1.0341704788978113e-06, "loss": 0.0289, "step": 93840 }, { "grad_norm": 0.2583194375038147, "learning_rate": 1.0308276260532879e-06, "loss": 0.0306, "step": 93850 }, { "grad_norm": 0.6263231039047241, "learning_rate": 1.0274901284038241e-06, "loss": 0.0293, "step": 93860 }, { "grad_norm": 0.32054632902145386, "learning_rate": 1.0241579863144114e-06, "loss": 0.0256, "step": 93870 }, { "grad_norm": 0.9101496934890747, "learning_rate": 1.0208312001494413e-06, "loss": 0.0314, "step": 93880 }, { "grad_norm": 0.500681459903717, "learning_rate": 1.0175097702727453e-06, "loss": 0.0339, "step": 93890 }, { "grad_norm": 0.2651211619377136, "learning_rate": 1.0141936970475275e-06, "loss": 0.0304, "step": 93900 }, { "grad_norm": 0.2945508360862732, "learning_rate": 1.0108829808364417e-06, "loss": 0.0309, "step": 93910 }, { "grad_norm": 0.313614159822464, "learning_rate": 1.0075776220015376e-06, "loss": 0.0319, "step": 93920 }, { "grad_norm": 0.5496768355369568, "learning_rate": 1.0042776209042816e-06, "loss": 0.0259, "step": 93930 }, { "grad_norm": 0.3289669454097748, "learning_rate": 1.000982977905568e-06, "loss": 0.0292, "step": 93940 }, { "grad_norm": 0.5571934580802917, "learning_rate": 9.976936933656867e-07, "loss": 0.0286, "step": 93950 }, { "grad_norm": 0.4598204791545868, "learning_rate": 9.944097676443442e-07, "loss": 0.0286, "step": 93960 }, { "grad_norm": 0.22114932537078857, "learning_rate": 9.911312011006757e-07, "loss": 0.0308, "step": 93970 }, { "grad_norm": 0.621036171913147, "learning_rate": 9.878579940932109e-07, "loss": 0.0286, "step": 93980 }, { "grad_norm": 0.3127249479293823, "learning_rate": 9.84590146979908e-07, "loss": 0.0345, "step": 93990 }, { "grad_norm": 0.26117178797721863, "learning_rate": 9.813276601181365e-07, "loss": 0.0263, "step": 94000 }, { "grad_norm": 1.0896856784820557, "learning_rate": 9.780705338646724e-07, "loss": 0.024, "step": 94010 }, { "grad_norm": 0.3357087969779968, "learning_rate": 9.748187685757083e-07, "loss": 0.029, "step": 94020 }, { "grad_norm": 0.7165668606758118, "learning_rate": 9.715723646068543e-07, "loss": 0.029, "step": 94030 }, { "grad_norm": 0.4823177754878998, "learning_rate": 9.68331322313143e-07, "loss": 0.0256, "step": 94040 }, { "grad_norm": 0.5558852553367615, "learning_rate": 9.650956420489909e-07, "loss": 0.0312, "step": 94050 }, { "grad_norm": 0.29609692096710205, "learning_rate": 9.618653241682595e-07, "loss": 0.0293, "step": 94060 }, { "grad_norm": 1.228115200996399, "learning_rate": 9.586403690242053e-07, "loss": 0.0316, "step": 94070 }, { "grad_norm": 0.755013108253479, "learning_rate": 9.554207769695122e-07, "loss": 0.0266, "step": 94080 }, { "grad_norm": 0.33705106377601624, "learning_rate": 9.522065483562659e-07, "loss": 0.0313, "step": 94090 }, { "grad_norm": 0.35151901841163635, "learning_rate": 9.489976835359681e-07, "loss": 0.0221, "step": 94100 }, { "grad_norm": 1.0321992635726929, "learning_rate": 9.457941828595329e-07, "loss": 0.0318, "step": 94110 }, { "grad_norm": 0.5489938855171204, "learning_rate": 9.425960466772965e-07, "loss": 0.0274, "step": 94120 }, { "grad_norm": 0.4542785584926605, "learning_rate": 9.394032753390014e-07, "loss": 0.0307, "step": 94130 }, { "grad_norm": 0.2708258628845215, "learning_rate": 9.362158691938017e-07, "loss": 0.0215, "step": 94140 }, { "grad_norm": 0.7606090307235718, "learning_rate": 9.330338285902685e-07, "loss": 0.027, "step": 94150 }, { "grad_norm": 0.30155491828918457, "learning_rate": 9.298571538763845e-07, "loss": 0.0264, "step": 94160 }, { "grad_norm": 0.4289332926273346, "learning_rate": 9.266858453995441e-07, "loss": 0.0259, "step": 94170 }, { "grad_norm": 0.7278295755386353, "learning_rate": 9.235199035065645e-07, "loss": 0.0312, "step": 94180 }, { "grad_norm": 0.3238494396209717, "learning_rate": 9.203593285436574e-07, "loss": 0.0245, "step": 94190 }, { "grad_norm": 0.36248910427093506, "learning_rate": 9.172041208564686e-07, "loss": 0.029, "step": 94200 }, { "grad_norm": 0.3180142939090729, "learning_rate": 9.140542807900332e-07, "loss": 0.0234, "step": 94210 }, { "grad_norm": 0.32264241576194763, "learning_rate": 9.109098086888313e-07, "loss": 0.0285, "step": 94220 }, { "grad_norm": 0.2609904706478119, "learning_rate": 9.077707048967265e-07, "loss": 0.0373, "step": 94230 }, { "grad_norm": 0.3740611970424652, "learning_rate": 9.046369697570057e-07, "loss": 0.0253, "step": 94240 }, { "grad_norm": 0.42483749985694885, "learning_rate": 9.015086036123721e-07, "loss": 0.0223, "step": 94250 }, { "grad_norm": 0.42600470781326294, "learning_rate": 8.983856068049413e-07, "loss": 0.0272, "step": 94260 }, { "grad_norm": 0.46443769335746765, "learning_rate": 8.952679796762398e-07, "loss": 0.0354, "step": 94270 }, { "grad_norm": 0.28300490975379944, "learning_rate": 8.921557225671951e-07, "loss": 0.0252, "step": 94280 }, { "grad_norm": 0.553408682346344, "learning_rate": 8.890488358181792e-07, "loss": 0.0303, "step": 94290 }, { "grad_norm": 0.45092537999153137, "learning_rate": 8.859473197689372e-07, "loss": 0.0309, "step": 94300 }, { "grad_norm": 0.48959484696388245, "learning_rate": 8.828511747586588e-07, "loss": 0.0255, "step": 94310 }, { "grad_norm": 0.3060360848903656, "learning_rate": 8.797604011259287e-07, "loss": 0.0324, "step": 94320 }, { "grad_norm": 0.37122493982315063, "learning_rate": 8.766749992087487e-07, "loss": 0.0294, "step": 94330 }, { "grad_norm": 0.5673037171363831, "learning_rate": 8.735949693445378e-07, "loss": 0.0245, "step": 94340 }, { "grad_norm": 0.46549150347709656, "learning_rate": 8.705203118701211e-07, "loss": 0.0224, "step": 94350 }, { "grad_norm": 1.0594555139541626, "learning_rate": 8.674510271217295e-07, "loss": 0.038, "step": 94360 }, { "grad_norm": 0.5635350942611694, "learning_rate": 8.643871154350336e-07, "loss": 0.0306, "step": 94370 }, { "grad_norm": 0.3428395688533783, "learning_rate": 8.613285771450818e-07, "loss": 0.0335, "step": 94380 }, { "grad_norm": 0.28480494022369385, "learning_rate": 8.582754125863623e-07, "loss": 0.0204, "step": 94390 }, { "grad_norm": 0.36379581689834595, "learning_rate": 8.55227622092758e-07, "loss": 0.0248, "step": 94400 }, { "grad_norm": 0.5937604904174805, "learning_rate": 8.521852059975743e-07, "loss": 0.0282, "step": 94410 }, { "grad_norm": 0.3094923198223114, "learning_rate": 8.491481646335231e-07, "loss": 0.028, "step": 94420 }, { "grad_norm": 0.7043126821517944, "learning_rate": 8.46116498332733e-07, "loss": 0.0249, "step": 94430 }, { "grad_norm": 0.467394083738327, "learning_rate": 8.430902074267333e-07, "loss": 0.0321, "step": 94440 }, { "grad_norm": 0.8564512729644775, "learning_rate": 8.400692922464925e-07, "loss": 0.0322, "step": 94450 }, { "grad_norm": 0.41511937975883484, "learning_rate": 8.370537531223577e-07, "loss": 0.0286, "step": 94460 }, { "grad_norm": 0.30390724539756775, "learning_rate": 8.340435903841092e-07, "loss": 0.0213, "step": 94470 }, { "grad_norm": 0.27915313839912415, "learning_rate": 8.310388043609341e-07, "loss": 0.0294, "step": 94480 }, { "grad_norm": 0.31220319867134094, "learning_rate": 8.280393953814303e-07, "loss": 0.0334, "step": 94490 }, { "grad_norm": 0.6182557344436646, "learning_rate": 8.250453637736077e-07, "loss": 0.03, "step": 94500 }, { "grad_norm": 0.4516739249229431, "learning_rate": 8.22056709864899e-07, "loss": 0.0269, "step": 94510 }, { "grad_norm": 0.2298170030117035, "learning_rate": 8.190734339821204e-07, "loss": 0.0235, "step": 94520 }, { "grad_norm": 0.6978858113288879, "learning_rate": 8.160955364515333e-07, "loss": 0.0287, "step": 94530 }, { "grad_norm": 0.6560424566268921, "learning_rate": 8.131230175987936e-07, "loss": 0.023, "step": 94540 }, { "grad_norm": 0.48711925745010376, "learning_rate": 8.101558777489693e-07, "loss": 0.0397, "step": 94550 }, { "grad_norm": 0.5823308825492859, "learning_rate": 8.071941172265396e-07, "loss": 0.0377, "step": 94560 }, { "grad_norm": 0.2974233031272888, "learning_rate": 8.042377363553955e-07, "loss": 0.0286, "step": 94570 }, { "grad_norm": 0.3066963255405426, "learning_rate": 8.01286735458856e-07, "loss": 0.0318, "step": 94580 }, { "grad_norm": 0.721886932849884, "learning_rate": 7.983411148596298e-07, "loss": 0.0339, "step": 94590 }, { "grad_norm": 0.2818848788738251, "learning_rate": 7.954008748798425e-07, "loss": 0.0265, "step": 94600 }, { "grad_norm": 0.3178733289241791, "learning_rate": 7.924660158410369e-07, "loss": 0.0267, "step": 94610 }, { "grad_norm": 0.5483134388923645, "learning_rate": 7.895365380641728e-07, "loss": 0.0326, "step": 94620 }, { "grad_norm": 0.5033985376358032, "learning_rate": 7.866124418695941e-07, "loss": 0.032, "step": 94630 }, { "grad_norm": 0.6223183870315552, "learning_rate": 7.83693727577095e-07, "loss": 0.0317, "step": 94640 }, { "grad_norm": 0.34112533926963806, "learning_rate": 7.807803955058535e-07, "loss": 0.0277, "step": 94650 }, { "grad_norm": 0.29344531893730164, "learning_rate": 7.778724459744702e-07, "loss": 0.0232, "step": 94660 }, { "grad_norm": 0.39482414722442627, "learning_rate": 7.749698793009464e-07, "loss": 0.032, "step": 94670 }, { "grad_norm": 0.39974889159202576, "learning_rate": 7.720726958027169e-07, "loss": 0.0302, "step": 94680 }, { "grad_norm": 0.6263110637664795, "learning_rate": 7.691808957965951e-07, "loss": 0.0307, "step": 94690 }, { "grad_norm": 0.24497003853321075, "learning_rate": 7.662944795988336e-07, "loss": 0.0199, "step": 94700 }, { "grad_norm": 0.2914082109928131, "learning_rate": 7.634134475250854e-07, "loss": 0.0279, "step": 94710 }, { "grad_norm": 0.28904154896736145, "learning_rate": 7.605377998904262e-07, "loss": 0.0288, "step": 94720 }, { "grad_norm": 0.7317439317703247, "learning_rate": 7.5766753700931e-07, "loss": 0.0264, "step": 94730 }, { "grad_norm": 0.3644276559352875, "learning_rate": 7.54802659195647e-07, "loss": 0.0295, "step": 94740 }, { "grad_norm": 0.19376567006111145, "learning_rate": 7.519431667627253e-07, "loss": 0.0214, "step": 94750 }, { "grad_norm": 0.5287134647369385, "learning_rate": 7.490890600232503e-07, "loss": 0.0394, "step": 94760 }, { "grad_norm": 0.4261608421802521, "learning_rate": 7.462403392893502e-07, "loss": 0.0296, "step": 94770 }, { "grad_norm": 0.49394264817237854, "learning_rate": 7.433970048725536e-07, "loss": 0.0296, "step": 94780 }, { "grad_norm": 0.20789915323257446, "learning_rate": 7.405590570838062e-07, "loss": 0.0244, "step": 94790 }, { "grad_norm": 0.14072486758232117, "learning_rate": 7.377264962334596e-07, "loss": 0.0217, "step": 94800 }, { "grad_norm": 0.28765565156936646, "learning_rate": 7.348993226312773e-07, "loss": 0.0227, "step": 94810 }, { "grad_norm": 0.4734954535961151, "learning_rate": 7.32077536586434e-07, "loss": 0.0282, "step": 94820 }, { "grad_norm": 0.5424951910972595, "learning_rate": 7.292611384075221e-07, "loss": 0.0255, "step": 94830 }, { "grad_norm": 0.32403984665870667, "learning_rate": 7.264501284025338e-07, "loss": 0.031, "step": 94840 }, { "grad_norm": 0.8550960421562195, "learning_rate": 7.236445068788789e-07, "loss": 0.0279, "step": 94850 }, { "grad_norm": 0.2794398367404938, "learning_rate": 7.20844274143373e-07, "loss": 0.029, "step": 94860 }, { "grad_norm": 0.3262815773487091, "learning_rate": 7.180494305022489e-07, "loss": 0.0287, "step": 94870 }, { "grad_norm": 0.3565445840358734, "learning_rate": 7.152599762611456e-07, "loss": 0.0287, "step": 94880 }, { "grad_norm": 0.3406681716442108, "learning_rate": 7.124759117251079e-07, "loss": 0.0458, "step": 94890 }, { "grad_norm": 0.5569135546684265, "learning_rate": 7.096972371986033e-07, "loss": 0.0331, "step": 94900 }, { "grad_norm": 0.8638321757316589, "learning_rate": 7.069239529855054e-07, "loss": 0.0357, "step": 94910 }, { "grad_norm": 0.5349089503288269, "learning_rate": 7.041560593890939e-07, "loss": 0.0253, "step": 94920 }, { "grad_norm": 0.23673954606056213, "learning_rate": 7.013935567120599e-07, "loss": 0.0302, "step": 94930 }, { "grad_norm": 0.3871620297431946, "learning_rate": 6.986364452565009e-07, "loss": 0.0201, "step": 94940 }, { "grad_norm": 1.0631047487258911, "learning_rate": 6.95884725323942e-07, "loss": 0.0411, "step": 94950 }, { "grad_norm": 0.18613356351852417, "learning_rate": 6.931383972152982e-07, "loss": 0.0276, "step": 94960 }, { "grad_norm": 0.2924112379550934, "learning_rate": 6.903974612309127e-07, "loss": 0.0235, "step": 94970 }, { "grad_norm": 0.2895548343658447, "learning_rate": 6.876619176705179e-07, "loss": 0.0268, "step": 94980 }, { "grad_norm": 0.2765880227088928, "learning_rate": 6.8493176683328e-07, "loss": 0.0289, "step": 94990 }, { "grad_norm": 0.8958009481430054, "learning_rate": 6.822070090177657e-07, "loss": 0.0314, "step": 95000 }, { "grad_norm": 0.3685862421989441, "learning_rate": 6.794876445219367e-07, "loss": 0.032, "step": 95010 }, { "grad_norm": 0.41256046295166016, "learning_rate": 6.767736736431885e-07, "loss": 0.0258, "step": 95020 }, { "grad_norm": 0.8058810830116272, "learning_rate": 6.740650966783113e-07, "loss": 0.0319, "step": 95030 }, { "grad_norm": 0.2884167432785034, "learning_rate": 6.713619139235239e-07, "loss": 0.0229, "step": 95040 }, { "grad_norm": 0.4392375349998474, "learning_rate": 6.686641256744286e-07, "loss": 0.0253, "step": 95050 }, { "grad_norm": 0.21699193120002747, "learning_rate": 6.659717322260562e-07, "loss": 0.0292, "step": 95060 }, { "grad_norm": 0.24056941270828247, "learning_rate": 6.632847338728377e-07, "loss": 0.0248, "step": 95070 }, { "grad_norm": 0.405390202999115, "learning_rate": 6.60603130908627e-07, "loss": 0.0282, "step": 95080 }, { "grad_norm": 0.39423730969429016, "learning_rate": 6.579269236266783e-07, "loss": 0.0295, "step": 95090 }, { "grad_norm": 0.3840769827365875, "learning_rate": 6.552561123196632e-07, "loss": 0.0215, "step": 95100 }, { "grad_norm": 0.4979211091995239, "learning_rate": 6.525906972796425e-07, "loss": 0.0315, "step": 95110 }, { "grad_norm": 0.4589809477329254, "learning_rate": 6.499306787981108e-07, "loss": 0.0399, "step": 95120 }, { "grad_norm": 0.6230538487434387, "learning_rate": 6.472760571659631e-07, "loss": 0.0277, "step": 95130 }, { "grad_norm": 0.4362897276878357, "learning_rate": 6.446268326735061e-07, "loss": 0.034, "step": 95140 }, { "grad_norm": 0.26359128952026367, "learning_rate": 6.41983005610458e-07, "loss": 0.0232, "step": 95150 }, { "grad_norm": 0.6029804944992065, "learning_rate": 6.393445762659378e-07, "loss": 0.0327, "step": 95160 }, { "grad_norm": 0.3624967336654663, "learning_rate": 6.36711544928481e-07, "loss": 0.0208, "step": 95170 }, { "grad_norm": 0.6769362688064575, "learning_rate": 6.340839118860298e-07, "loss": 0.0397, "step": 95180 }, { "grad_norm": 0.5083233714103699, "learning_rate": 6.314616774259485e-07, "loss": 0.0298, "step": 95190 }, { "grad_norm": 0.44430747628211975, "learning_rate": 6.288448418349858e-07, "loss": 0.0253, "step": 95200 }, { "grad_norm": 0.3180598020553589, "learning_rate": 6.262334053993235e-07, "loss": 0.0361, "step": 95210 }, { "grad_norm": 0.30039143562316895, "learning_rate": 6.236273684045501e-07, "loss": 0.0208, "step": 95220 }, { "grad_norm": 0.39494192600250244, "learning_rate": 6.210267311356488e-07, "loss": 0.0295, "step": 95230 }, { "grad_norm": 0.42069682478904724, "learning_rate": 6.184314938770197e-07, "loss": 0.0338, "step": 95240 }, { "grad_norm": 0.234044149518013, "learning_rate": 6.158416569124858e-07, "loss": 0.0244, "step": 95250 }, { "grad_norm": 0.5699834227561951, "learning_rate": 6.132572205252541e-07, "loss": 0.0177, "step": 95260 }, { "grad_norm": 0.5376375913619995, "learning_rate": 6.106781849979648e-07, "loss": 0.0297, "step": 95270 }, { "grad_norm": 0.45018237829208374, "learning_rate": 6.081045506126482e-07, "loss": 0.0222, "step": 95280 }, { "grad_norm": 0.416282057762146, "learning_rate": 6.055363176507678e-07, "loss": 0.0407, "step": 95290 }, { "grad_norm": 0.7443749308586121, "learning_rate": 6.029734863931602e-07, "loss": 0.0304, "step": 95300 }, { "grad_norm": 0.2829856872558594, "learning_rate": 6.004160571201179e-07, "loss": 0.0306, "step": 95310 }, { "grad_norm": 0.5858530402183533, "learning_rate": 5.978640301112947e-07, "loss": 0.0301, "step": 95320 }, { "grad_norm": 0.5831321477890015, "learning_rate": 5.9531740564579e-07, "loss": 0.0305, "step": 95330 }, { "grad_norm": 0.8366267681121826, "learning_rate": 5.927761840020918e-07, "loss": 0.0259, "step": 95340 }, { "grad_norm": 0.2681007385253906, "learning_rate": 5.902403654581113e-07, "loss": 0.0243, "step": 95350 }, { "grad_norm": 0.48508530855178833, "learning_rate": 5.877099502911542e-07, "loss": 0.0328, "step": 95360 }, { "grad_norm": 0.5925968885421753, "learning_rate": 5.851849387779496e-07, "loss": 0.0381, "step": 95370 }, { "grad_norm": 0.33011287450790405, "learning_rate": 5.826653311946262e-07, "loss": 0.0281, "step": 95380 }, { "grad_norm": 0.5076659917831421, "learning_rate": 5.801511278167249e-07, "loss": 0.0255, "step": 95390 }, { "grad_norm": 0.3105544447898865, "learning_rate": 5.776423289191868e-07, "loss": 0.0323, "step": 95400 }, { "grad_norm": 0.29429227113723755, "learning_rate": 5.751389347763869e-07, "loss": 0.031, "step": 95410 }, { "grad_norm": 0.4798443913459778, "learning_rate": 5.726409456620841e-07, "loss": 0.0254, "step": 95420 }, { "grad_norm": 0.3479304611682892, "learning_rate": 5.701483618494541e-07, "loss": 0.024, "step": 95430 }, { "grad_norm": 0.34760284423828125, "learning_rate": 5.67661183611079e-07, "loss": 0.0288, "step": 95440 }, { "grad_norm": 0.319103479385376, "learning_rate": 5.651794112189579e-07, "loss": 0.0216, "step": 95450 }, { "grad_norm": 0.5184982419013977, "learning_rate": 5.627030449444903e-07, "loss": 0.0332, "step": 95460 }, { "grad_norm": 0.35830655694007874, "learning_rate": 5.60232085058493e-07, "loss": 0.0337, "step": 95470 }, { "grad_norm": 0.6370763778686523, "learning_rate": 5.577665318311776e-07, "loss": 0.0388, "step": 95480 }, { "grad_norm": 0.38028639554977417, "learning_rate": 5.553063855321838e-07, "loss": 0.029, "step": 95490 }, { "grad_norm": 0.37815216183662415, "learning_rate": 5.528516464305467e-07, "loss": 0.0359, "step": 95500 }, { "grad_norm": 0.27978846430778503, "learning_rate": 5.504023147947068e-07, "loss": 0.025, "step": 95510 }, { "grad_norm": 0.5074881315231323, "learning_rate": 5.479583908925279e-07, "loss": 0.0393, "step": 95520 }, { "grad_norm": 0.7881489992141724, "learning_rate": 5.455198749912626e-07, "loss": 0.0253, "step": 95530 }, { "grad_norm": 0.4632844030857086, "learning_rate": 5.430867673575979e-07, "loss": 0.0371, "step": 95540 }, { "grad_norm": 0.336268812417984, "learning_rate": 5.406590682575985e-07, "loss": 0.0234, "step": 95550 }, { "grad_norm": 0.2634357810020447, "learning_rate": 5.382367779567687e-07, "loss": 0.0266, "step": 95560 }, { "grad_norm": 0.5196207165718079, "learning_rate": 5.358198967200023e-07, "loss": 0.0236, "step": 95570 }, { "grad_norm": 0.480832040309906, "learning_rate": 5.334084248115989e-07, "loss": 0.0355, "step": 95580 }, { "grad_norm": 0.3162485361099243, "learning_rate": 5.310023624952754e-07, "loss": 0.0216, "step": 95590 }, { "grad_norm": 0.3969908356666565, "learning_rate": 5.286017100341656e-07, "loss": 0.0266, "step": 95600 }, { "grad_norm": 0.39855843782424927, "learning_rate": 5.262064676907874e-07, "loss": 0.0359, "step": 95610 }, { "grad_norm": 0.2593149244785309, "learning_rate": 5.238166357270924e-07, "loss": 0.0269, "step": 95620 }, { "grad_norm": 0.6706838011741638, "learning_rate": 5.214322144044215e-07, "loss": 0.0194, "step": 95630 }, { "grad_norm": 0.36266234517097473, "learning_rate": 5.190532039835327e-07, "loss": 0.0353, "step": 95640 }, { "grad_norm": 0.4382358491420746, "learning_rate": 5.166796047245903e-07, "loss": 0.0325, "step": 95650 }, { "grad_norm": 0.6447535157203674, "learning_rate": 5.143114168871699e-07, "loss": 0.0307, "step": 95660 }, { "grad_norm": 0.5678912997245789, "learning_rate": 5.119486407302532e-07, "loss": 0.0305, "step": 95670 }, { "grad_norm": 0.4753772020339966, "learning_rate": 5.095912765122335e-07, "loss": 0.034, "step": 95680 }, { "grad_norm": 0.9036566019058228, "learning_rate": 5.07239324490899e-07, "loss": 0.0369, "step": 95690 }, { "grad_norm": 0.3295515775680542, "learning_rate": 5.048927849234608e-07, "loss": 0.0372, "step": 95700 }, { "grad_norm": 0.9332335591316223, "learning_rate": 5.025516580665358e-07, "loss": 0.0349, "step": 95710 }, { "grad_norm": 0.47228795289993286, "learning_rate": 5.002159441761412e-07, "loss": 0.0229, "step": 95720 }, { "grad_norm": 0.2542532682418823, "learning_rate": 4.978856435077062e-07, "loss": 0.0276, "step": 95730 }, { "grad_norm": 0.7534341216087341, "learning_rate": 4.955607563160769e-07, "loss": 0.0317, "step": 95740 }, { "grad_norm": 0.6102871298789978, "learning_rate": 4.932412828554944e-07, "loss": 0.0285, "step": 95750 }, { "grad_norm": 0.30267032980918884, "learning_rate": 4.909272233796114e-07, "loss": 0.025, "step": 95760 }, { "grad_norm": 0.31262749433517456, "learning_rate": 4.88618578141492e-07, "loss": 0.0343, "step": 95770 }, { "grad_norm": 0.35120657086372375, "learning_rate": 4.863153473936011e-07, "loss": 0.0313, "step": 95780 }, { "grad_norm": 0.2936708629131317, "learning_rate": 4.840175313878259e-07, "loss": 0.023, "step": 95790 }, { "grad_norm": 0.2530580461025238, "learning_rate": 4.817251303754433e-07, "loss": 0.0245, "step": 95800 }, { "grad_norm": 0.3795787990093231, "learning_rate": 4.794381446071583e-07, "loss": 0.0285, "step": 95810 }, { "grad_norm": 0.5820637345314026, "learning_rate": 4.771565743330597e-07, "loss": 0.032, "step": 95820 }, { "grad_norm": 0.33511990308761597, "learning_rate": 4.7488041980266463e-07, "loss": 0.0313, "step": 95830 }, { "grad_norm": 0.35061877965927124, "learning_rate": 4.726096812648795e-07, "loss": 0.0321, "step": 95840 }, { "grad_norm": 0.20380747318267822, "learning_rate": 4.703443589680445e-07, "loss": 0.0282, "step": 95850 }, { "grad_norm": 0.8284111022949219, "learning_rate": 4.6808445315987826e-07, "loss": 0.033, "step": 95860 }, { "grad_norm": 0.3540158271789551, "learning_rate": 4.658299640875274e-07, "loss": 0.0247, "step": 95870 }, { "grad_norm": 0.27642562985420227, "learning_rate": 4.635808919975393e-07, "loss": 0.0211, "step": 95880 }, { "grad_norm": 0.4845517873764038, "learning_rate": 4.6133723713586706e-07, "loss": 0.0277, "step": 95890 }, { "grad_norm": 1.5361965894699097, "learning_rate": 4.590989997478756e-07, "loss": 0.0275, "step": 95900 }, { "grad_norm": 0.26896974444389343, "learning_rate": 4.5686618007833027e-07, "loss": 0.0299, "step": 95910 }, { "grad_norm": 0.6275424957275391, "learning_rate": 4.54638778371419e-07, "loss": 0.0284, "step": 95920 }, { "grad_norm": 0.9484860301017761, "learning_rate": 4.524167948707192e-07, "loss": 0.0366, "step": 95930 }, { "grad_norm": 0.48327305912971497, "learning_rate": 4.5020022981921986e-07, "loss": 0.0283, "step": 95940 }, { "grad_norm": 0.3349365293979645, "learning_rate": 4.47989083459327e-07, "loss": 0.0395, "step": 95950 }, { "grad_norm": 0.5116857290267944, "learning_rate": 4.457833560328528e-07, "loss": 0.026, "step": 95960 }, { "grad_norm": 0.3129952847957611, "learning_rate": 4.4358304778100434e-07, "loss": 0.0309, "step": 95970 }, { "grad_norm": 0.2609630823135376, "learning_rate": 4.413881589444058e-07, "loss": 0.03, "step": 95980 }, { "grad_norm": 0.3309369385242462, "learning_rate": 4.391986897630873e-07, "loss": 0.0265, "step": 95990 }, { "grad_norm": 0.4102660119533539, "learning_rate": 4.370146404764908e-07, "loss": 0.0476, "step": 96000 }, { "grad_norm": 0.2485511600971222, "learning_rate": 4.3483601132345285e-07, "loss": 0.0326, "step": 96010 }, { "grad_norm": 0.98744136095047, "learning_rate": 4.32662802542233e-07, "loss": 0.0329, "step": 96020 }, { "grad_norm": 0.3329700529575348, "learning_rate": 4.3049501437047444e-07, "loss": 0.0277, "step": 96030 }, { "grad_norm": 0.5790112614631653, "learning_rate": 4.283326470452653e-07, "loss": 0.0292, "step": 96040 }, { "grad_norm": 0.2043563276529312, "learning_rate": 4.2617570080306647e-07, "loss": 0.022, "step": 96050 }, { "grad_norm": 0.7578836679458618, "learning_rate": 4.2402417587975585e-07, "loss": 0.0346, "step": 96060 }, { "grad_norm": 0.5134776830673218, "learning_rate": 4.218780725106286e-07, "loss": 0.0267, "step": 96070 }, { "grad_norm": 0.4291696548461914, "learning_rate": 4.197373909303748e-07, "loss": 0.0306, "step": 96080 }, { "grad_norm": 0.35493534803390503, "learning_rate": 4.17602131373096e-07, "loss": 0.0223, "step": 96090 }, { "grad_norm": 0.6549144983291626, "learning_rate": 4.154722940723055e-07, "loss": 0.0366, "step": 96100 }, { "grad_norm": 0.45007675886154175, "learning_rate": 4.1334787926091136e-07, "loss": 0.0263, "step": 96110 }, { "grad_norm": 1.0761739015579224, "learning_rate": 4.1122888717124444e-07, "loss": 0.0275, "step": 96120 }, { "grad_norm": 0.23106883466243744, "learning_rate": 4.091153180350249e-07, "loss": 0.0329, "step": 96130 }, { "grad_norm": 0.38778793811798096, "learning_rate": 4.0700717208340125e-07, "loss": 0.0291, "step": 96140 }, { "grad_norm": 0.45098692178726196, "learning_rate": 4.049044495469112e-07, "loss": 0.028, "step": 96150 }, { "grad_norm": 0.5360594987869263, "learning_rate": 4.0280715065549866e-07, "loss": 0.029, "step": 96160 }, { "grad_norm": 0.3734237849712372, "learning_rate": 4.0071527563853575e-07, "loss": 0.0275, "step": 96170 }, { "grad_norm": 0.5700699687004089, "learning_rate": 3.9862882472477823e-07, "loss": 0.0209, "step": 96180 }, { "grad_norm": 0.7116773128509521, "learning_rate": 3.9654779814239375e-07, "loss": 0.0287, "step": 96190 }, { "grad_norm": 0.2293711006641388, "learning_rate": 3.944721961189668e-07, "loss": 0.0282, "step": 96200 }, { "grad_norm": 0.4038659632205963, "learning_rate": 3.9240201888147697e-07, "loss": 0.0303, "step": 96210 }, { "grad_norm": 0.46940621733665466, "learning_rate": 3.9033726665632096e-07, "loss": 0.0352, "step": 96220 }, { "grad_norm": 0.42898595333099365, "learning_rate": 3.882779396692959e-07, "loss": 0.0243, "step": 96230 }, { "grad_norm": 0.37876349687576294, "learning_rate": 3.8622403814559947e-07, "loss": 0.0337, "step": 96240 }, { "grad_norm": 0.32417958974838257, "learning_rate": 3.841755623098575e-07, "loss": 0.0212, "step": 96250 }, { "grad_norm": 0.31032848358154297, "learning_rate": 3.821325123860742e-07, "loss": 0.0272, "step": 96260 }, { "grad_norm": 0.651177167892456, "learning_rate": 3.800948885976818e-07, "loss": 0.027, "step": 96270 }, { "grad_norm": 0.3124264180660248, "learning_rate": 3.780626911675078e-07, "loss": 0.0368, "step": 96280 }, { "grad_norm": 0.2670263648033142, "learning_rate": 3.76035920317791e-07, "loss": 0.0368, "step": 96290 }, { "grad_norm": 0.36238303780555725, "learning_rate": 3.740145762701819e-07, "loss": 0.0258, "step": 96300 }, { "grad_norm": 0.45873889327049255, "learning_rate": 3.7199865924572585e-07, "loss": 0.0252, "step": 96310 }, { "grad_norm": 0.26796954870224, "learning_rate": 3.699881694648799e-07, "loss": 0.0214, "step": 96320 }, { "grad_norm": 0.4460088014602661, "learning_rate": 3.679831071475126e-07, "loss": 0.0309, "step": 96330 }, { "grad_norm": 0.34949710965156555, "learning_rate": 3.65983472512893e-07, "loss": 0.0268, "step": 96340 }, { "grad_norm": 0.3800469636917114, "learning_rate": 3.6398926577969616e-07, "loss": 0.0236, "step": 96350 }, { "grad_norm": 0.5734344124794006, "learning_rate": 3.620004871660032e-07, "loss": 0.0219, "step": 96360 }, { "grad_norm": 0.30759432911872864, "learning_rate": 3.600171368893124e-07, "loss": 0.0201, "step": 96370 }, { "grad_norm": 0.5542364716529846, "learning_rate": 3.580392151665113e-07, "loss": 0.0311, "step": 96380 }, { "grad_norm": 0.2569882869720459, "learning_rate": 3.5606672221391024e-07, "loss": 0.0183, "step": 96390 }, { "grad_norm": 0.4872235655784607, "learning_rate": 3.540996582472089e-07, "loss": 0.0259, "step": 96400 }, { "grad_norm": 0.3159734308719635, "learning_rate": 3.521380234815297e-07, "loss": 0.0282, "step": 96410 }, { "grad_norm": 0.3338097929954529, "learning_rate": 3.5018181813138985e-07, "loss": 0.0289, "step": 96420 }, { "grad_norm": 0.44021114706993103, "learning_rate": 3.482310424107238e-07, "loss": 0.0298, "step": 96430 }, { "grad_norm": 0.7020577192306519, "learning_rate": 3.4628569653285537e-07, "loss": 0.0246, "step": 96440 }, { "grad_norm": 0.24294699728488922, "learning_rate": 3.44345780710531e-07, "loss": 0.0302, "step": 96450 }, { "grad_norm": 0.5821937918663025, "learning_rate": 3.424112951558978e-07, "loss": 0.0427, "step": 96460 }, { "grad_norm": 0.4777607023715973, "learning_rate": 3.40482240080503e-07, "loss": 0.0247, "step": 96470 }, { "grad_norm": 0.38478991389274597, "learning_rate": 3.3855861569531133e-07, "loss": 0.0242, "step": 96480 }, { "grad_norm": 0.4386526644229889, "learning_rate": 3.3664042221068226e-07, "loss": 0.0272, "step": 96490 }, { "grad_norm": 0.37768593430519104, "learning_rate": 3.347276598363924e-07, "loss": 0.0269, "step": 96500 }, { "grad_norm": 0.5164248943328857, "learning_rate": 3.328203287816134e-07, "loss": 0.0293, "step": 96510 }, { "grad_norm": 0.45897161960601807, "learning_rate": 3.3091842925493386e-07, "loss": 0.0354, "step": 96520 }, { "grad_norm": 0.5602344870567322, "learning_rate": 3.2902196146433196e-07, "loss": 0.032, "step": 96530 }, { "grad_norm": 0.32588568329811096, "learning_rate": 3.2713092561720836e-07, "loss": 0.0268, "step": 96540 }, { "grad_norm": 0.4162142276763916, "learning_rate": 3.2524532192036996e-07, "loss": 0.0302, "step": 96550 }, { "grad_norm": 0.137081116437912, "learning_rate": 3.2336515058001836e-07, "loss": 0.0242, "step": 96560 }, { "grad_norm": 0.47975972294807434, "learning_rate": 3.2149041180176143e-07, "loss": 0.0228, "step": 96570 }, { "grad_norm": 0.4593862295150757, "learning_rate": 3.1962110579062954e-07, "loss": 0.0178, "step": 96580 }, { "grad_norm": 0.38966435194015503, "learning_rate": 3.177572327510314e-07, "loss": 0.0228, "step": 96590 }, { "grad_norm": 0.40339478850364685, "learning_rate": 3.158987928868151e-07, "loss": 0.031, "step": 96600 }, { "grad_norm": 0.4251595735549927, "learning_rate": 3.1404578640120144e-07, "loss": 0.0424, "step": 96610 }, { "grad_norm": 0.28113818168640137, "learning_rate": 3.1219821349683933e-07, "loss": 0.0202, "step": 96620 }, { "grad_norm": 0.44166871905326843, "learning_rate": 3.103560743757783e-07, "loss": 0.0272, "step": 96630 }, { "grad_norm": 0.33540937304496765, "learning_rate": 3.0851936923946834e-07, "loss": 0.0242, "step": 96640 }, { "grad_norm": 0.2997804880142212, "learning_rate": 3.06688098288771e-07, "loss": 0.0202, "step": 96650 }, { "grad_norm": 0.3976697325706482, "learning_rate": 3.0486226172394825e-07, "loss": 0.0226, "step": 96660 }, { "grad_norm": 0.1956901252269745, "learning_rate": 3.030418597446738e-07, "loss": 0.0209, "step": 96670 }, { "grad_norm": 0.28203701972961426, "learning_rate": 3.012268925500272e-07, "loss": 0.0275, "step": 96680 }, { "grad_norm": 0.36821258068084717, "learning_rate": 2.994173603384831e-07, "loss": 0.021, "step": 96690 }, { "grad_norm": 0.4351835548877716, "learning_rate": 2.9761326330793317e-07, "loss": 0.0348, "step": 96700 }, { "grad_norm": 0.5286737680435181, "learning_rate": 2.958146016556751e-07, "loss": 0.0271, "step": 96710 }, { "grad_norm": 0.19759291410446167, "learning_rate": 2.94021375578396e-07, "loss": 0.0362, "step": 96720 }, { "grad_norm": 0.3459673225879669, "learning_rate": 2.9223358527221135e-07, "loss": 0.0306, "step": 96730 }, { "grad_norm": 0.3667552173137665, "learning_rate": 2.904512309326257e-07, "loss": 0.0257, "step": 96740 }, { "grad_norm": 0.5340272188186646, "learning_rate": 2.8867431275455545e-07, "loss": 0.033, "step": 96750 }, { "grad_norm": 0.4080890119075775, "learning_rate": 2.86902830932323e-07, "loss": 0.0421, "step": 96760 }, { "grad_norm": 0.3625898063182831, "learning_rate": 2.851367856596565e-07, "loss": 0.0358, "step": 96770 }, { "grad_norm": 0.6748320460319519, "learning_rate": 2.83376177129685e-07, "loss": 0.0288, "step": 96780 }, { "grad_norm": 0.3633280396461487, "learning_rate": 2.8162100553494884e-07, "loss": 0.0341, "step": 96790 }, { "grad_norm": 0.40036648511886597, "learning_rate": 2.798712710673834e-07, "loss": 0.0258, "step": 96800 }, { "grad_norm": 0.3173726797103882, "learning_rate": 2.7812697391834676e-07, "loss": 0.0286, "step": 96810 }, { "grad_norm": 0.9882375597953796, "learning_rate": 2.7638811427858627e-07, "loss": 0.0279, "step": 96820 }, { "grad_norm": 0.5560776591300964, "learning_rate": 2.746546923382665e-07, "loss": 0.0272, "step": 96830 }, { "grad_norm": 0.39023342728614807, "learning_rate": 2.729267082869469e-07, "loss": 0.0301, "step": 96840 }, { "grad_norm": 0.4095713794231415, "learning_rate": 2.712041623135986e-07, "loss": 0.0287, "step": 96850 }, { "grad_norm": 0.3712014853954315, "learning_rate": 2.694870546065931e-07, "loss": 0.0362, "step": 96860 }, { "grad_norm": 0.5586471557617188, "learning_rate": 2.677753853537246e-07, "loss": 0.0338, "step": 96870 }, { "grad_norm": 0.49503475427627563, "learning_rate": 2.660691547421601e-07, "loss": 0.0317, "step": 96880 }, { "grad_norm": 0.853046715259552, "learning_rate": 2.6436836295850033e-07, "loss": 0.0246, "step": 96890 }, { "grad_norm": 0.7879041433334351, "learning_rate": 2.6267301018874093e-07, "loss": 0.028, "step": 96900 }, { "grad_norm": 0.20116208493709564, "learning_rate": 2.6098309661828355e-07, "loss": 0.0219, "step": 96910 }, { "grad_norm": 0.40541401505470276, "learning_rate": 2.592986224319305e-07, "loss": 0.0244, "step": 96920 }, { "grad_norm": 0.4834361970424652, "learning_rate": 2.576195878139009e-07, "loss": 0.0287, "step": 96930 }, { "grad_norm": 0.3823040723800659, "learning_rate": 2.5594599294780364e-07, "loss": 0.0267, "step": 96940 }, { "grad_norm": 0.2316703200340271, "learning_rate": 2.542778380166644e-07, "loss": 0.0358, "step": 96950 }, { "grad_norm": 0.2281533181667328, "learning_rate": 2.526151232029095e-07, "loss": 0.0315, "step": 96960 }, { "grad_norm": 0.3053134083747864, "learning_rate": 2.5095784868836567e-07, "loss": 0.0225, "step": 96970 }, { "grad_norm": 0.4139156937599182, "learning_rate": 2.493060146542825e-07, "loss": 0.0299, "step": 96980 }, { "grad_norm": 0.26271283626556396, "learning_rate": 2.476596212812876e-07, "loss": 0.0212, "step": 96990 }, { "grad_norm": 0.35253602266311646, "learning_rate": 2.4601866874943703e-07, "loss": 0.0304, "step": 97000 }, { "grad_norm": 0.4297088086605072, "learning_rate": 2.4438315723818163e-07, "loss": 0.0271, "step": 97010 }, { "grad_norm": 0.22416000068187714, "learning_rate": 2.4275308692637833e-07, "loss": 0.0235, "step": 97020 }, { "grad_norm": 0.3313128352165222, "learning_rate": 2.411284579922901e-07, "loss": 0.0243, "step": 97030 }, { "grad_norm": 0.41653090715408325, "learning_rate": 2.395092706135804e-07, "loss": 0.0262, "step": 97040 }, { "grad_norm": 0.2394402176141739, "learning_rate": 2.3789552496731872e-07, "loss": 0.0319, "step": 97050 }, { "grad_norm": 0.5070376992225647, "learning_rate": 2.3628722122999158e-07, "loss": 0.0315, "step": 97060 }, { "grad_norm": 0.524830162525177, "learning_rate": 2.3468435957747503e-07, "loss": 0.0273, "step": 97070 }, { "grad_norm": 0.13284848630428314, "learning_rate": 2.330869401850566e-07, "loss": 0.0256, "step": 97080 }, { "grad_norm": 0.4427497684955597, "learning_rate": 2.3149496322742437e-07, "loss": 0.0198, "step": 97090 }, { "grad_norm": 0.6000155210494995, "learning_rate": 2.2990842887868346e-07, "loss": 0.0256, "step": 97100 }, { "grad_norm": 0.2083643674850464, "learning_rate": 2.2832733731232292e-07, "loss": 0.0293, "step": 97110 }, { "grad_norm": 0.42228707671165466, "learning_rate": 2.2675168870124886e-07, "loss": 0.0267, "step": 97120 }, { "grad_norm": 0.35813483595848083, "learning_rate": 2.2518148321778455e-07, "loss": 0.0304, "step": 97130 }, { "grad_norm": 0.4775719940662384, "learning_rate": 2.2361672103363706e-07, "loss": 0.0295, "step": 97140 }, { "grad_norm": 0.8142023086547852, "learning_rate": 2.220574023199251e-07, "loss": 0.024, "step": 97150 }, { "grad_norm": 0.8231428265571594, "learning_rate": 2.2050352724717894e-07, "loss": 0.0303, "step": 97160 }, { "grad_norm": 0.4111548662185669, "learning_rate": 2.1895509598532372e-07, "loss": 0.0233, "step": 97170 }, { "grad_norm": 0.38627898693084717, "learning_rate": 2.1741210870369066e-07, "loss": 0.0324, "step": 97180 }, { "grad_norm": 1.1563855409622192, "learning_rate": 2.1587456557102814e-07, "loss": 0.0307, "step": 97190 }, { "grad_norm": 0.3370778560638428, "learning_rate": 2.143424667554683e-07, "loss": 0.0407, "step": 97200 }, { "grad_norm": 0.24571804702281952, "learning_rate": 2.128158124245716e-07, "loss": 0.0316, "step": 97210 }, { "grad_norm": 0.2422311156988144, "learning_rate": 2.1129460274527668e-07, "loss": 0.0253, "step": 97220 }, { "grad_norm": 0.25570082664489746, "learning_rate": 2.0977883788395046e-07, "loss": 0.0308, "step": 97230 }, { "grad_norm": 0.3646000027656555, "learning_rate": 2.0826851800634928e-07, "loss": 0.024, "step": 97240 }, { "grad_norm": 0.5156513452529907, "learning_rate": 2.0676364327764653e-07, "loss": 0.042, "step": 97250 }, { "grad_norm": 0.5758365392684937, "learning_rate": 2.0526421386240502e-07, "loss": 0.0317, "step": 97260 }, { "grad_norm": 0.3781643211841583, "learning_rate": 2.0377022992461025e-07, "loss": 0.0333, "step": 97270 }, { "grad_norm": 0.4431934952735901, "learning_rate": 2.0228169162763156e-07, "loss": 0.0334, "step": 97280 }, { "grad_norm": 0.25177764892578125, "learning_rate": 2.0079859913425536e-07, "loss": 0.0296, "step": 97290 }, { "grad_norm": 0.4499782621860504, "learning_rate": 1.9932095260667417e-07, "loss": 0.0407, "step": 97300 }, { "grad_norm": 0.47130945324897766, "learning_rate": 1.9784875220648092e-07, "loss": 0.0269, "step": 97310 }, { "grad_norm": 0.255332887172699, "learning_rate": 1.9638199809466907e-07, "loss": 0.0169, "step": 97320 }, { "grad_norm": 0.4990460276603699, "learning_rate": 1.9492069043164364e-07, "loss": 0.0333, "step": 97330 }, { "grad_norm": 0.5191272497177124, "learning_rate": 1.9346482937721567e-07, "loss": 0.0259, "step": 97340 }, { "grad_norm": 0.41290372610092163, "learning_rate": 1.920144150905856e-07, "loss": 0.0301, "step": 97350 }, { "grad_norm": 0.6543121933937073, "learning_rate": 1.9056944773037655e-07, "loss": 0.0366, "step": 97360 }, { "grad_norm": 0.4566629230976105, "learning_rate": 1.8912992745460102e-07, "loss": 0.0277, "step": 97370 }, { "grad_norm": 0.7001051902770996, "learning_rate": 1.876958544206886e-07, "loss": 0.0237, "step": 97380 }, { "grad_norm": 0.48073989152908325, "learning_rate": 1.8626722878546942e-07, "loss": 0.0355, "step": 97390 }, { "grad_norm": 0.597562849521637, "learning_rate": 1.8484405070516851e-07, "loss": 0.031, "step": 97400 }, { "grad_norm": 0.27303969860076904, "learning_rate": 1.8342632033542805e-07, "loss": 0.0287, "step": 97410 }, { "grad_norm": 0.3160664737224579, "learning_rate": 1.8201403783129066e-07, "loss": 0.0223, "step": 97420 }, { "grad_norm": 0.6077839136123657, "learning_rate": 1.806072033471884e-07, "loss": 0.0273, "step": 97430 }, { "grad_norm": 0.3287796676158905, "learning_rate": 1.79205817036987e-07, "loss": 0.0277, "step": 97440 }, { "grad_norm": 0.474536657333374, "learning_rate": 1.7780987905393065e-07, "loss": 0.0226, "step": 97450 }, { "grad_norm": 0.31639912724494934, "learning_rate": 1.764193895506805e-07, "loss": 0.0269, "step": 97460 }, { "grad_norm": 0.3464231491088867, "learning_rate": 1.7503434867929824e-07, "loss": 0.0287, "step": 97470 }, { "grad_norm": 0.49773067235946655, "learning_rate": 1.7365475659125165e-07, "loss": 0.0225, "step": 97480 }, { "grad_norm": 0.40786445140838623, "learning_rate": 1.7228061343740332e-07, "loss": 0.0247, "step": 97490 }, { "grad_norm": 0.2743377983570099, "learning_rate": 1.7091191936803308e-07, "loss": 0.0271, "step": 97500 }, { "grad_norm": 0.23257669806480408, "learning_rate": 1.6954867453281563e-07, "loss": 0.0297, "step": 97510 }, { "grad_norm": 0.27728065848350525, "learning_rate": 1.6819087908084285e-07, "loss": 0.0287, "step": 97520 }, { "grad_norm": 0.5498661994934082, "learning_rate": 1.6683853316058485e-07, "loss": 0.0271, "step": 97530 }, { "grad_norm": 0.5398039221763611, "learning_rate": 1.6549163691995106e-07, "loss": 0.0294, "step": 97540 }, { "grad_norm": 0.17278917133808136, "learning_rate": 1.641501905062237e-07, "loss": 0.0367, "step": 97550 }, { "grad_norm": 0.34935262799263, "learning_rate": 1.628141940661021e-07, "loss": 0.028, "step": 97560 }, { "grad_norm": 0.29489004611968994, "learning_rate": 1.6148364774569158e-07, "loss": 0.0283, "step": 97570 }, { "grad_norm": 0.3647957146167755, "learning_rate": 1.6015855169050355e-07, "loss": 0.0444, "step": 97580 }, { "grad_norm": 0.5405967235565186, "learning_rate": 1.5883890604543873e-07, "loss": 0.0286, "step": 97590 }, { "grad_norm": 0.3624627888202667, "learning_rate": 1.5752471095482057e-07, "loss": 0.0242, "step": 97600 }, { "grad_norm": 0.3942628800868988, "learning_rate": 1.5621596656235638e-07, "loss": 0.0247, "step": 97610 }, { "grad_norm": 0.4164932370185852, "learning_rate": 1.5491267301117608e-07, "loss": 0.027, "step": 97620 }, { "grad_norm": 0.3361794650554657, "learning_rate": 1.536148304438101e-07, "loss": 0.0291, "step": 97630 }, { "grad_norm": 0.39175617694854736, "learning_rate": 1.5232243900217823e-07, "loss": 0.0255, "step": 97640 }, { "grad_norm": 0.27643445134162903, "learning_rate": 1.51035498827623e-07, "loss": 0.0276, "step": 97650 }, { "grad_norm": 0.8918566703796387, "learning_rate": 1.497540100608763e-07, "loss": 0.0418, "step": 97660 }, { "grad_norm": 0.35247930884361267, "learning_rate": 1.4847797284208153e-07, "loss": 0.0237, "step": 97670 }, { "grad_norm": 0.4075736105442047, "learning_rate": 1.472073873107882e-07, "loss": 0.0387, "step": 97680 }, { "grad_norm": 0.22366231679916382, "learning_rate": 1.459422536059407e-07, "loss": 0.0348, "step": 97690 }, { "grad_norm": 0.9965150356292725, "learning_rate": 1.4468257186589508e-07, "loss": 0.0263, "step": 97700 }, { "grad_norm": 0.3797411024570465, "learning_rate": 1.4342834222840217e-07, "loss": 0.0308, "step": 97710 }, { "grad_norm": 0.27040261030197144, "learning_rate": 1.4217956483063566e-07, "loss": 0.0213, "step": 97720 }, { "grad_norm": 0.4586521089076996, "learning_rate": 1.409362398091474e-07, "loss": 0.0283, "step": 97730 }, { "grad_norm": 0.34435510635375977, "learning_rate": 1.3969836729990638e-07, "loss": 0.0296, "step": 97740 }, { "grad_norm": 0.48594439029693604, "learning_rate": 1.384659474382932e-07, "loss": 0.0306, "step": 97750 }, { "grad_norm": 0.7477761507034302, "learning_rate": 1.372389803590779e-07, "loss": 0.0328, "step": 97760 }, { "grad_norm": 0.33355799317359924, "learning_rate": 1.360174661964364e-07, "loss": 0.0352, "step": 97770 }, { "grad_norm": 0.9151043891906738, "learning_rate": 1.3480140508396188e-07, "loss": 0.0256, "step": 97780 }, { "grad_norm": 0.9115147590637207, "learning_rate": 1.335907971546313e-07, "loss": 0.0318, "step": 97790 }, { "grad_norm": 0.5558196902275085, "learning_rate": 1.3238564254083875e-07, "loss": 0.0337, "step": 97800 }, { "grad_norm": 0.3340330123901367, "learning_rate": 1.3118594137437323e-07, "loss": 0.0206, "step": 97810 }, { "grad_norm": 0.37344980239868164, "learning_rate": 1.2999169378644094e-07, "loss": 0.0332, "step": 97820 }, { "grad_norm": 0.543254017829895, "learning_rate": 1.2880289990763738e-07, "loss": 0.033, "step": 97830 }, { "grad_norm": 0.18973678350448608, "learning_rate": 1.2761955986796968e-07, "loss": 0.0304, "step": 97840 }, { "grad_norm": 0.14830709993839264, "learning_rate": 1.264416737968399e-07, "loss": 0.0305, "step": 97850 }, { "grad_norm": 0.502285361289978, "learning_rate": 1.2526924182307275e-07, "loss": 0.0239, "step": 97860 }, { "grad_norm": 0.38178327679634094, "learning_rate": 1.2410226407487124e-07, "loss": 0.035, "step": 97870 }, { "grad_norm": 0.307911217212677, "learning_rate": 1.229407406798555e-07, "loss": 0.0296, "step": 97880 }, { "grad_norm": 0.27722951769828796, "learning_rate": 1.2178467176505726e-07, "loss": 0.0356, "step": 97890 }, { "grad_norm": 0.33285191655158997, "learning_rate": 1.2063405745689205e-07, "loss": 0.0235, "step": 97900 }, { "grad_norm": 0.3298037052154541, "learning_rate": 1.1948889788119812e-07, "loss": 0.024, "step": 97910 }, { "grad_norm": 0.6375606656074524, "learning_rate": 1.1834919316320303e-07, "loss": 0.0328, "step": 97920 }, { "grad_norm": 0.7297864556312561, "learning_rate": 1.1721494342754048e-07, "loss": 0.0368, "step": 97930 }, { "grad_norm": 0.4731164276599884, "learning_rate": 1.1608614879825563e-07, "loss": 0.0234, "step": 97940 }, { "grad_norm": 0.3995837867259979, "learning_rate": 1.1496280939879422e-07, "loss": 0.0271, "step": 97950 }, { "grad_norm": 0.3848179876804352, "learning_rate": 1.1384492535199687e-07, "loss": 0.0269, "step": 97960 }, { "grad_norm": 0.2819427251815796, "learning_rate": 1.1273249678011578e-07, "loss": 0.0238, "step": 97970 }, { "grad_norm": 0.48594218492507935, "learning_rate": 1.1162552380480362e-07, "loss": 0.0279, "step": 97980 }, { "grad_norm": 0.3763049244880676, "learning_rate": 1.1052400654711359e-07, "loss": 0.0247, "step": 97990 }, { "grad_norm": 0.41082486510276794, "learning_rate": 1.0942794512751597e-07, "loss": 0.033, "step": 98000 }, { "grad_norm": 0.33044251799583435, "learning_rate": 1.0833733966587045e-07, "loss": 0.0248, "step": 98010 }, { "grad_norm": 0.31802788376808167, "learning_rate": 1.0725219028143718e-07, "loss": 0.0294, "step": 98020 }, { "grad_norm": 0.5257931351661682, "learning_rate": 1.0617249709289345e-07, "loss": 0.0228, "step": 98030 }, { "grad_norm": 0.4754531681537628, "learning_rate": 1.0509826021831703e-07, "loss": 0.03, "step": 98040 }, { "grad_norm": 0.36077648401260376, "learning_rate": 1.0402947977517508e-07, "loss": 0.0346, "step": 98050 }, { "grad_norm": 0.28735876083374023, "learning_rate": 1.0296615588035185e-07, "loss": 0.026, "step": 98060 }, { "grad_norm": 0.4507361948490143, "learning_rate": 1.0190828865012659e-07, "loss": 0.0256, "step": 98070 }, { "grad_norm": 0.24853165447711945, "learning_rate": 1.008558782001956e-07, "loss": 0.0314, "step": 98080 }, { "grad_norm": 0.7367147207260132, "learning_rate": 9.980892464563906e-08, "loss": 0.0319, "step": 98090 }, { "grad_norm": 0.266605943441391, "learning_rate": 9.876742810095985e-08, "loss": 0.0215, "step": 98100 }, { "grad_norm": 0.27578094601631165, "learning_rate": 9.77313886800446e-08, "loss": 0.0228, "step": 98110 }, { "grad_norm": 0.6190391778945923, "learning_rate": 9.670080649619717e-08, "loss": 0.0383, "step": 98120 }, { "grad_norm": 0.31774279475212097, "learning_rate": 9.567568166212737e-08, "loss": 0.0334, "step": 98130 }, { "grad_norm": 0.4255790710449219, "learning_rate": 9.465601428992888e-08, "loss": 0.0291, "step": 98140 }, { "grad_norm": 0.3388785123825073, "learning_rate": 9.364180449111803e-08, "loss": 0.03, "step": 98150 }, { "grad_norm": 0.4032879173755646, "learning_rate": 9.263305237661169e-08, "loss": 0.0331, "step": 98160 }, { "grad_norm": 0.47010260820388794, "learning_rate": 9.162975805671603e-08, "loss": 0.0294, "step": 98170 }, { "grad_norm": 0.3928118050098419, "learning_rate": 9.063192164115442e-08, "loss": 0.025, "step": 98180 }, { "grad_norm": 0.5288949608802795, "learning_rate": 8.963954323904511e-08, "loss": 0.0239, "step": 98190 }, { "grad_norm": 0.35018089413642883, "learning_rate": 8.865262295891796e-08, "loss": 0.0274, "step": 98200 }, { "grad_norm": 0.7280361652374268, "learning_rate": 8.767116090870331e-08, "loss": 0.0262, "step": 98210 }, { "grad_norm": 0.6735731363296509, "learning_rate": 8.669515719572086e-08, "loss": 0.0236, "step": 98220 }, { "grad_norm": 0.44975772500038147, "learning_rate": 8.572461192671855e-08, "loss": 0.0315, "step": 98230 }, { "grad_norm": 0.6061518788337708, "learning_rate": 8.475952520782815e-08, "loss": 0.0269, "step": 98240 }, { "grad_norm": 0.2039724588394165, "learning_rate": 8.379989714458747e-08, "loss": 0.0199, "step": 98250 }, { "grad_norm": 0.4767468571662903, "learning_rate": 8.284572784194034e-08, "loss": 0.0273, "step": 98260 }, { "grad_norm": 0.36622384190559387, "learning_rate": 8.189701740424216e-08, "loss": 0.0348, "step": 98270 }, { "grad_norm": 0.6076016426086426, "learning_rate": 8.095376593522663e-08, "loss": 0.0278, "step": 98280 }, { "grad_norm": 0.5144572854042053, "learning_rate": 8.001597353806123e-08, "loss": 0.0328, "step": 98290 }, { "grad_norm": 0.3858102858066559, "learning_rate": 7.90836403152917e-08, "loss": 0.0261, "step": 98300 }, { "grad_norm": 0.31122148036956787, "learning_rate": 7.815676636888092e-08, "loss": 0.034, "step": 98310 }, { "grad_norm": 1.0829730033874512, "learning_rate": 7.723535180019226e-08, "loss": 0.0279, "step": 98320 }, { "grad_norm": 0.393899142742157, "learning_rate": 7.6319396709984e-08, "loss": 0.0274, "step": 98330 }, { "grad_norm": 0.5035431385040283, "learning_rate": 7.540890119843158e-08, "loss": 0.0317, "step": 98340 }, { "grad_norm": 0.20033147931098938, "learning_rate": 7.450386536509979e-08, "loss": 0.0222, "step": 98350 }, { "grad_norm": 0.4426089823246002, "learning_rate": 7.360428930895947e-08, "loss": 0.0337, "step": 98360 }, { "grad_norm": 0.41429510712623596, "learning_rate": 7.271017312839302e-08, "loss": 0.0324, "step": 98370 }, { "grad_norm": 0.31097134947776794, "learning_rate": 7.182151692118333e-08, "loss": 0.0343, "step": 98380 }, { "grad_norm": 0.43364834785461426, "learning_rate": 7.093832078449713e-08, "loss": 0.0327, "step": 98390 }, { "grad_norm": 0.3671814203262329, "learning_rate": 7.006058481493493e-08, "loss": 0.0288, "step": 98400 }, { "grad_norm": 0.7937098145484924, "learning_rate": 6.918830910847552e-08, "loss": 0.0288, "step": 98410 }, { "grad_norm": 0.3243396282196045, "learning_rate": 6.832149376052033e-08, "loss": 0.0237, "step": 98420 }, { "grad_norm": 0.24079382419586182, "learning_rate": 6.746013886584912e-08, "loss": 0.0282, "step": 98430 }, { "grad_norm": 0.32471105456352234, "learning_rate": 6.660424451866431e-08, "loss": 0.03, "step": 98440 }, { "grad_norm": 0.2960105240345001, "learning_rate": 6.575381081256327e-08, "loss": 0.0337, "step": 98450 }, { "grad_norm": 0.7780781388282776, "learning_rate": 6.49088378405549e-08, "loss": 0.0298, "step": 98460 }, { "grad_norm": 0.26415640115737915, "learning_rate": 6.406932569503755e-08, "loss": 0.0233, "step": 98470 }, { "grad_norm": 0.5005241632461548, "learning_rate": 6.323527446782662e-08, "loss": 0.0282, "step": 98480 }, { "grad_norm": 0.6963798999786377, "learning_rate": 6.240668425012142e-08, "loss": 0.0278, "step": 98490 }, { "grad_norm": 0.3033718466758728, "learning_rate": 6.158355513254388e-08, "loss": 0.0199, "step": 98500 }, { "grad_norm": 0.5663948059082031, "learning_rate": 6.076588720510534e-08, "loss": 0.0326, "step": 98510 }, { "grad_norm": 0.6511107087135315, "learning_rate": 5.995368055722316e-08, "loss": 0.0305, "step": 98520 }, { "grad_norm": 0.5063977837562561, "learning_rate": 5.914693527773185e-08, "loss": 0.0377, "step": 98530 }, { "grad_norm": 0.8422616124153137, "learning_rate": 5.834565145483861e-08, "loss": 0.0283, "step": 98540 }, { "grad_norm": 0.31772372126579285, "learning_rate": 5.754982917618446e-08, "loss": 0.0237, "step": 98550 }, { "grad_norm": 0.5136826634407043, "learning_rate": 5.675946852878866e-08, "loss": 0.0347, "step": 98560 }, { "grad_norm": 0.4745316207408905, "learning_rate": 5.59745695990932e-08, "loss": 0.0267, "step": 98570 }, { "grad_norm": 0.4138401746749878, "learning_rate": 5.51951324729294e-08, "loss": 0.029, "step": 98580 }, { "grad_norm": 0.6418965458869934, "learning_rate": 5.4421157235529093e-08, "loss": 0.0244, "step": 98590 }, { "grad_norm": 0.35318291187286377, "learning_rate": 5.365264397154679e-08, "loss": 0.0279, "step": 98600 }, { "grad_norm": 0.4040830731391907, "learning_rate": 5.288959276501526e-08, "loss": 0.0326, "step": 98610 }, { "grad_norm": 0.6795858144760132, "learning_rate": 5.2132003699378876e-08, "loss": 0.0413, "step": 98620 }, { "grad_norm": 0.2054973542690277, "learning_rate": 5.137987685749912e-08, "loss": 0.0243, "step": 98630 }, { "grad_norm": 0.46567192673683167, "learning_rate": 5.0633212321610227e-08, "loss": 0.0265, "step": 98640 }, { "grad_norm": 0.47659239172935486, "learning_rate": 4.989201017338019e-08, "loss": 0.0313, "step": 98650 }, { "grad_norm": 0.3063170611858368, "learning_rate": 4.9156270493860846e-08, "loss": 0.0287, "step": 98660 }, { "grad_norm": 0.5879001617431641, "learning_rate": 4.842599336351561e-08, "loss": 0.0218, "step": 98670 }, { "grad_norm": 0.28466781973838806, "learning_rate": 4.7701178862197274e-08, "loss": 0.0257, "step": 98680 }, { "grad_norm": 0.3711429238319397, "learning_rate": 4.698182706918131e-08, "loss": 0.0215, "step": 98690 }, { "grad_norm": 0.42123639583587646, "learning_rate": 4.6267938063121466e-08, "loss": 0.0285, "step": 98700 }, { "grad_norm": 0.4340958893299103, "learning_rate": 4.555951192209973e-08, "loss": 0.0322, "step": 98710 }, { "grad_norm": 0.2944087088108063, "learning_rate": 4.485654872358747e-08, "loss": 0.0259, "step": 98720 }, { "grad_norm": 1.97234046459198, "learning_rate": 4.415904854445097e-08, "loss": 0.034, "step": 98730 }, { "grad_norm": 0.380139023065567, "learning_rate": 4.34670114609792e-08, "loss": 0.0304, "step": 98740 }, { "grad_norm": 0.39194172620773315, "learning_rate": 4.278043754884498e-08, "loss": 0.0224, "step": 98750 }, { "grad_norm": 0.4492689073085785, "learning_rate": 4.209932688312712e-08, "loss": 0.0289, "step": 98760 }, { "grad_norm": 0.49138379096984863, "learning_rate": 4.14236795383216e-08, "loss": 0.0313, "step": 98770 }, { "grad_norm": 0.38401177525520325, "learning_rate": 4.075349558830821e-08, "loss": 0.0322, "step": 98780 }, { "grad_norm": 0.39303797483444214, "learning_rate": 4.008877510638387e-08, "loss": 0.0162, "step": 98790 }, { "grad_norm": 0.29694071412086487, "learning_rate": 3.942951816523488e-08, "loss": 0.0304, "step": 98800 }, { "grad_norm": 0.565390408039093, "learning_rate": 3.8775724836959125e-08, "loss": 0.043, "step": 98810 }, { "grad_norm": 0.5324273109436035, "learning_rate": 3.812739519305497e-08, "loss": 0.03, "step": 98820 }, { "grad_norm": 0.3775072395801544, "learning_rate": 3.748452930442126e-08, "loss": 0.0287, "step": 98830 }, { "grad_norm": 0.6795530319213867, "learning_rate": 3.6847127241362855e-08, "loss": 0.028, "step": 98840 }, { "grad_norm": 0.6239791512489319, "learning_rate": 3.621518907358512e-08, "loss": 0.0379, "step": 98850 }, { "grad_norm": 0.45359164476394653, "learning_rate": 3.558871487019388e-08, "loss": 0.0285, "step": 98860 }, { "grad_norm": 0.4736345112323761, "learning_rate": 3.4967704699701007e-08, "loss": 0.0269, "step": 98870 }, { "grad_norm": 0.2826395332813263, "learning_rate": 3.435215863001884e-08, "loss": 0.0243, "step": 98880 }, { "grad_norm": 0.31503862142562866, "learning_rate": 3.374207672846019e-08, "loss": 0.035, "step": 98890 }, { "grad_norm": 0.42159777879714966, "learning_rate": 3.313745906174392e-08, "loss": 0.0228, "step": 98900 }, { "grad_norm": 0.5417509078979492, "learning_rate": 3.253830569599492e-08, "loss": 0.0315, "step": 98910 }, { "grad_norm": 0.34296733140945435, "learning_rate": 3.1944616696727436e-08, "loss": 0.0237, "step": 98920 }, { "grad_norm": 0.3981340825557709, "learning_rate": 3.135639212887287e-08, "loss": 0.0257, "step": 98930 }, { "grad_norm": 0.47360873222351074, "learning_rate": 3.077363205675754e-08, "loss": 0.0336, "step": 98940 }, { "grad_norm": 0.7217515707015991, "learning_rate": 3.0196336544113804e-08, "loss": 0.0247, "step": 98950 }, { "grad_norm": 1.005605697631836, "learning_rate": 2.9624505654063383e-08, "loss": 0.0317, "step": 98960 }, { "grad_norm": 0.27505043148994446, "learning_rate": 2.90581394491507e-08, "loss": 0.0308, "step": 98970 }, { "grad_norm": 0.5438758134841919, "learning_rate": 2.8497237991309545e-08, "loss": 0.0327, "step": 98980 }, { "grad_norm": 0.5470739006996155, "learning_rate": 2.7941801341879735e-08, "loss": 0.0313, "step": 98990 }, { "grad_norm": 0.5114954710006714, "learning_rate": 2.7391829561601578e-08, "loss": 0.035, "step": 99000 }, { "grad_norm": 0.8214856386184692, "learning_rate": 2.6847322710621403e-08, "loss": 0.0254, "step": 99010 }, { "grad_norm": 0.6739429831504822, "learning_rate": 2.6308280848486022e-08, "loss": 0.0214, "step": 99020 }, { "grad_norm": 0.2856065630912781, "learning_rate": 2.5774704034137177e-08, "loss": 0.0275, "step": 99030 }, { "grad_norm": 0.4420067071914673, "learning_rate": 2.524659232593929e-08, "loss": 0.0256, "step": 99040 }, { "grad_norm": 0.2635957896709442, "learning_rate": 2.472394578163506e-08, "loss": 0.0242, "step": 99050 }, { "grad_norm": 0.4603162109851837, "learning_rate": 2.4206764458378772e-08, "loss": 0.0223, "step": 99060 }, { "grad_norm": 0.5563380122184753, "learning_rate": 2.3695048412736285e-08, "loss": 0.0263, "step": 99070 }, { "grad_norm": 0.476759672164917, "learning_rate": 2.3188797700668395e-08, "loss": 0.0275, "step": 99080 }, { "grad_norm": 0.6176877021789551, "learning_rate": 2.268801237753082e-08, "loss": 0.0232, "step": 99090 }, { "grad_norm": 0.14505454897880554, "learning_rate": 2.2192692498090862e-08, "loss": 0.0226, "step": 99100 }, { "grad_norm": 0.6049925684928894, "learning_rate": 2.170283811652185e-08, "loss": 0.0333, "step": 99110 }, { "grad_norm": 0.6272227764129639, "learning_rate": 2.12184492863865e-08, "loss": 0.0276, "step": 99120 }, { "grad_norm": 0.3717878460884094, "learning_rate": 2.073952606066465e-08, "loss": 0.0293, "step": 99130 }, { "grad_norm": 0.2699768543243408, "learning_rate": 2.026606849171997e-08, "loss": 0.0237, "step": 99140 }, { "grad_norm": 0.4904513359069824, "learning_rate": 1.9798076631333262e-08, "loss": 0.0294, "step": 99150 }, { "grad_norm": 0.44286566972732544, "learning_rate": 1.933555053069136e-08, "loss": 0.0298, "step": 99160 }, { "grad_norm": 0.3200644552707672, "learning_rate": 1.887849024036492e-08, "loss": 0.023, "step": 99170 }, { "grad_norm": 0.7623475193977356, "learning_rate": 1.8426895810341736e-08, "loss": 0.0318, "step": 99180 }, { "grad_norm": 0.37277981638908386, "learning_rate": 1.798076729000453e-08, "loss": 0.0228, "step": 99190 }, { "grad_norm": 0.3715451657772064, "learning_rate": 1.75401047281476e-08, "loss": 0.0233, "step": 99200 }, { "grad_norm": 0.2875046133995056, "learning_rate": 1.7104908172954626e-08, "loss": 0.0278, "step": 99210 }, { "grad_norm": 0.23954805731773376, "learning_rate": 1.6675177672020868e-08, "loss": 0.0213, "step": 99220 }, { "grad_norm": 0.9695614576339722, "learning_rate": 1.6250913272342073e-08, "loss": 0.0357, "step": 99230 }, { "grad_norm": 0.4274553656578064, "learning_rate": 1.5832115020314453e-08, "loss": 0.0387, "step": 99240 }, { "grad_norm": 0.25109317898750305, "learning_rate": 1.5418782961734712e-08, "loss": 0.0275, "step": 99250 }, { "grad_norm": 0.8978539705276489, "learning_rate": 1.501091714181113e-08, "loss": 0.0296, "step": 99260 }, { "grad_norm": 0.34707149863243103, "learning_rate": 1.460851760513582e-08, "loss": 0.0248, "step": 99270 }, { "grad_norm": 0.42155712842941284, "learning_rate": 1.4211584395723566e-08, "loss": 0.0332, "step": 99280 }, { "grad_norm": 0.9438552260398865, "learning_rate": 1.382011755697854e-08, "loss": 0.0225, "step": 99290 }, { "grad_norm": 0.31456273794174194, "learning_rate": 1.3434117131716494e-08, "loss": 0.0192, "step": 99300 }, { "grad_norm": 0.6219499707221985, "learning_rate": 1.3053583162142557e-08, "loss": 0.0246, "step": 99310 }, { "grad_norm": 0.33631211519241333, "learning_rate": 1.2678515689873438e-08, "loss": 0.0304, "step": 99320 }, { "grad_norm": 0.3695807158946991, "learning_rate": 1.2308914755931877e-08, "loss": 0.0285, "step": 99330 }, { "grad_norm": 0.37459954619407654, "learning_rate": 1.1944780400729993e-08, "loss": 0.0353, "step": 99340 }, { "grad_norm": 0.583222508430481, "learning_rate": 1.1586112664085935e-08, "loss": 0.0303, "step": 99350 }, { "grad_norm": 0.31228235363960266, "learning_rate": 1.1232911585234984e-08, "loss": 0.0278, "step": 99360 }, { "grad_norm": 0.5015744566917419, "learning_rate": 1.0885177202790697e-08, "loss": 0.0307, "step": 99370 }, { "grad_norm": 0.31506508588790894, "learning_rate": 1.0542909554789315e-08, "loss": 0.03, "step": 99380 }, { "grad_norm": 0.31778228282928467, "learning_rate": 1.0206108678656456e-08, "loss": 0.0238, "step": 99390 }, { "grad_norm": 0.5841041207313538, "learning_rate": 9.874774611223769e-09, "loss": 0.0368, "step": 99400 }, { "grad_norm": 0.539911687374115, "learning_rate": 9.548907388728933e-09, "loss": 0.0259, "step": 99410 }, { "grad_norm": 1.4311718940734863, "learning_rate": 9.228507046804557e-09, "loss": 0.027, "step": 99420 }, { "grad_norm": 0.20169079303741455, "learning_rate": 8.91357362048928e-09, "loss": 0.0371, "step": 99430 }, { "grad_norm": 0.63807213306427, "learning_rate": 8.604107144227769e-09, "loss": 0.025, "step": 99440 }, { "grad_norm": 0.24412758648395538, "learning_rate": 8.300107651859623e-09, "loss": 0.0253, "step": 99450 }, { "grad_norm": 0.5500485301017761, "learning_rate": 8.001575176630472e-09, "loss": 0.03, "step": 99460 }, { "grad_norm": 0.6431572437286377, "learning_rate": 7.708509751186422e-09, "loss": 0.0202, "step": 99470 }, { "grad_norm": 0.3076460063457489, "learning_rate": 7.420911407579611e-09, "loss": 0.027, "step": 99480 }, { "grad_norm": 0.4865913987159729, "learning_rate": 7.13878017725711e-09, "loss": 0.0169, "step": 99490 }, { "grad_norm": 0.31541121006011963, "learning_rate": 6.8621160910720125e-09, "loss": 0.0352, "step": 99500 }, { "grad_norm": 0.38732433319091797, "learning_rate": 6.59091917928345e-09, "loss": 0.0282, "step": 99510 }, { "grad_norm": 0.42827460169792175, "learning_rate": 6.32518947155103e-09, "loss": 0.0298, "step": 99520 }, { "grad_norm": 0.3599807918071747, "learning_rate": 6.064926996929288e-09, "loss": 0.026, "step": 99530 }, { "grad_norm": 0.1732877641916275, "learning_rate": 5.810131783884343e-09, "loss": 0.0207, "step": 99540 }, { "grad_norm": 0.36290279030799866, "learning_rate": 5.5608038602772415e-09, "loss": 0.0326, "step": 99550 }, { "grad_norm": 0.5217518210411072, "learning_rate": 5.316943253375062e-09, "loss": 0.0219, "step": 99560 }, { "grad_norm": 0.32802852988243103, "learning_rate": 5.07854998984536e-09, "loss": 0.0259, "step": 99570 }, { "grad_norm": 0.3053826689720154, "learning_rate": 4.845624095756174e-09, "loss": 0.0238, "step": 99580 }, { "grad_norm": 0.3675936460494995, "learning_rate": 4.618165596587121e-09, "loss": 0.0307, "step": 99590 }, { "grad_norm": 0.3129911720752716, "learning_rate": 4.396174517207197e-09, "loss": 0.0202, "step": 99600 }, { "grad_norm": 0.2646632790565491, "learning_rate": 4.179650881896979e-09, "loss": 0.0278, "step": 99610 }, { "grad_norm": 0.24620531499385834, "learning_rate": 3.968594714331975e-09, "loss": 0.0229, "step": 99620 }, { "grad_norm": 0.35193851590156555, "learning_rate": 3.763006037593719e-09, "loss": 0.0274, "step": 99630 }, { "grad_norm": 0.3782629370689392, "learning_rate": 3.5628848741586786e-09, "loss": 0.0243, "step": 99640 }, { "grad_norm": 0.5502843856811523, "learning_rate": 3.368231245926001e-09, "loss": 0.0276, "step": 99650 }, { "grad_norm": 0.2842017710208893, "learning_rate": 3.179045174167561e-09, "loss": 0.0252, "step": 99660 }, { "grad_norm": 0.3260875642299652, "learning_rate": 2.9953266795834657e-09, "loss": 0.0305, "step": 99670 }, { "grad_norm": 0.43785151839256287, "learning_rate": 2.817075782263201e-09, "loss": 0.0287, "step": 99680 }, { "grad_norm": 0.3178096413612366, "learning_rate": 2.6442925016967325e-09, "loss": 0.0223, "step": 99690 }, { "grad_norm": 0.4096972346305847, "learning_rate": 2.4769768567800555e-09, "loss": 0.032, "step": 99700 }, { "grad_norm": 0.8040130734443665, "learning_rate": 2.315128865809646e-09, "loss": 0.0262, "step": 99710 }, { "grad_norm": 0.34379857778549194, "learning_rate": 2.1587485464880097e-09, "loss": 0.0244, "step": 99720 }, { "grad_norm": 0.34024351835250854, "learning_rate": 2.0078359159125813e-09, "loss": 0.0322, "step": 99730 }, { "grad_norm": 0.3157147169113159, "learning_rate": 1.8623909905923776e-09, "loss": 0.0307, "step": 99740 }, { "grad_norm": 0.30667880177497864, "learning_rate": 1.7224137864257916e-09, "loss": 0.0291, "step": 99750 }, { "grad_norm": 0.32006892561912537, "learning_rate": 1.5879043187283505e-09, "loss": 0.0312, "step": 99760 }, { "grad_norm": 0.3516169786453247, "learning_rate": 1.458862602204958e-09, "loss": 0.0234, "step": 99770 }, { "grad_norm": 0.36107853055000305, "learning_rate": 1.3352886509720996e-09, "loss": 0.0257, "step": 99780 }, { "grad_norm": 0.3441742956638336, "learning_rate": 1.2171824785356389e-09, "loss": 0.0362, "step": 99790 }, { "grad_norm": 0.4515939950942993, "learning_rate": 1.1045440978185718e-09, "loss": 0.0285, "step": 99800 }, { "grad_norm": 0.2719559073448181, "learning_rate": 9.973735211332714e-10, "loss": 0.0295, "step": 99810 }, { "grad_norm": 0.44136402010917664, "learning_rate": 8.956707602036929e-10, "loss": 0.0272, "step": 99820 }, { "grad_norm": 0.6481277942657471, "learning_rate": 7.994358261542712e-10, "loss": 0.0235, "step": 99830 }, { "grad_norm": 0.2719009816646576, "learning_rate": 7.086687295043692e-10, "loss": 0.019, "step": 99840 }, { "grad_norm": 0.6461162567138672, "learning_rate": 6.23369480179381e-10, "loss": 0.0236, "step": 99850 }, { "grad_norm": 0.40340709686279297, "learning_rate": 5.43538087510731e-10, "loss": 0.0336, "step": 99860 }, { "grad_norm": 1.3987404108047485, "learning_rate": 4.691745602303233e-10, "loss": 0.0286, "step": 99870 }, { "grad_norm": 0.8323236703872681, "learning_rate": 4.0027890646499034e-10, "loss": 0.0482, "step": 99880 }, { "grad_norm": 0.21508808434009552, "learning_rate": 3.368511337531466e-10, "loss": 0.0199, "step": 99890 }, { "grad_norm": 0.6386926174163818, "learning_rate": 2.788912490281348e-10, "loss": 0.0326, "step": 99900 }, { "grad_norm": 0.17308871448040009, "learning_rate": 2.2639925863487954e-10, "loss": 0.0258, "step": 99910 }, { "grad_norm": 0.3367500901222229, "learning_rate": 1.7937516830213163e-10, "loss": 0.0244, "step": 99920 }, { "grad_norm": 0.8386011123657227, "learning_rate": 1.3781898318687702e-10, "loss": 0.0299, "step": 99930 }, { "grad_norm": 0.5884502530097961, "learning_rate": 1.0173070781882566e-10, "loss": 0.0314, "step": 99940 }, { "grad_norm": 0.30457091331481934, "learning_rate": 7.111034615592261e-11, "loss": 0.0321, "step": 99950 }, { "grad_norm": 0.25780248641967773, "learning_rate": 4.595790153993918e-11, "loss": 0.0219, "step": 99960 }, { "grad_norm": 0.7307007908821106, "learning_rate": 2.6273376729779585e-11, "loss": 0.0283, "step": 99970 }, { "grad_norm": 2.2071590423583984, "learning_rate": 1.2056773868174276e-11, "loss": 0.025, "step": 99980 }, { "grad_norm": 0.24369598925113678, "learning_rate": 3.3080945149865926e-12, "loss": 0.0221, "step": 99990 }, { "grad_norm": 0.3432290256023407, "learning_rate": 2.7339625008337977e-14, "loss": 0.0251, "step": 100000 } ], "logging_steps": 10, "max_steps": 100000, "num_input_tokens_seen": 0, "num_train_epochs": 9223372036854775807, "save_steps": 20000, "stateful_callbacks": { "TrainerControl": { "args": { "should_epoch_stop": false, "should_evaluate": false, "should_log": false, "should_save": true, "should_training_stop": true }, "attributes": {} } }, "total_flos": 0.0, "train_batch_size": 8, "trial_name": null, "trial_params": null }